From 8bf62ecee58360749c5f0e68bc97d5e02a6816b1 Mon Sep 17 00:00:00 2001 From: Albert Lee Date: Thu, 12 May 2005 15:29:42 -0400 Subject: [libata] C/H/S support, for older devices --- include/linux/ata.h | 14 ++++++++++++++ include/linux/libata.h | 6 ++++++ 2 files changed, 20 insertions(+) (limited to 'include/linux') diff --git a/include/linux/ata.h b/include/linux/ata.h index f178894edd04..d8981402cd5b 100644 --- a/include/linux/ata.h +++ b/include/linux/ata.h @@ -125,6 +125,7 @@ enum { ATA_CMD_PACKET = 0xA0, ATA_CMD_VERIFY = 0x40, ATA_CMD_VERIFY_EXT = 0x42, + ATA_CMD_INIT_DEV_PARAMS = 0x91, /* SETFEATURES stuff */ SETFEATURES_XFER = 0x03, @@ -174,6 +175,7 @@ enum { ATA_TFLAG_ISADDR = (1 << 1), /* enable r/w to nsect/lba regs */ ATA_TFLAG_DEVICE = (1 << 2), /* enable r/w to device reg */ ATA_TFLAG_WRITE = (1 << 3), /* data dir: host->dev==1 (write) */ + ATA_TFLAG_LBA = (1 << 4), /* enable LBA */ }; enum ata_tf_protocols { @@ -242,6 +244,18 @@ struct ata_taskfile { ((u64) (id)[(n) + 1] << 16) | \ ((u64) (id)[(n) + 0]) ) +static inline int ata_id_current_chs_valid(u16 *id) +{ + /* For ATA-1 devices, if the INITIALIZE DEVICE PARAMETERS command + has not been issued to the device then the values of + id[54] to id[56] are vendor specific. */ + return (id[53] & 0x01) && /* Current translation valid */ + id[54] && /* cylinders in current translation */ + id[55] && /* heads in current translation */ + id[55] <= 16 && + id[56]; /* sectors in current translation */ +} + static inline int atapi_cdb_len(u16 *dev_id) { u16 tmp = dev_id[0] & 0x3; diff --git a/include/linux/libata.h b/include/linux/libata.h index 505160ab472b..bd0df84cfd87 100644 --- a/include/linux/libata.h +++ b/include/linux/libata.h @@ -95,6 +95,7 @@ enum { ATA_DFLAG_LBA48 = (1 << 0), /* device supports LBA48 */ ATA_DFLAG_PIO = (1 << 1), /* device currently in PIO mode */ ATA_DFLAG_LOCK_SECTORS = (1 << 2), /* don't adjust max_sectors */ + ATA_DFLAG_LBA = (1 << 3), /* device supports LBA */ ATA_DEV_UNKNOWN = 0, /* unknown device */ ATA_DEV_ATA = 1, /* ATA device */ @@ -278,6 +279,11 @@ struct ata_device { u8 xfer_protocol; /* taskfile xfer protocol */ u8 read_cmd; /* opcode to use on read */ u8 write_cmd; /* opcode to use on write */ + + /* for CHS addressing */ + u16 cylinders; /* Number of cylinders */ + u16 heads; /* Number of heads */ + u16 sectors; /* Number of sectors per track */ }; struct ata_port { -- cgit v1.2.3 From 3173c8907ffb2c64456142da3df2bd0500bd59e0 Mon Sep 17 00:00:00 2001 From: Nishanth Aravamudan Date: Sun, 11 Sep 2005 02:09:55 -0700 Subject: [PATCH] drivers/net: fix-up schedule_timeout() usage Use schedule_timeout_interruptible() instead of set_current_state()/schedule_timeout() to reduce kernel size. Signed-off-by: Nishanth Aravamudan Cc: Jeff Garzik Cc: "David S. Miller" Signed-off-by: Andrew Morton Signed-off-by: Jeff Garzik --- drivers/net/8139cp.c | 3 +- drivers/net/hp100.c | 48 +++++++++++-------------------- drivers/net/irda/stir4200.c | 7 ++--- drivers/net/ixgb/ixgb_ethtool.c | 7 ++--- drivers/net/ns83820.c | 3 +- drivers/net/tokenring/ibmtr.c | 9 +++--- drivers/net/tokenring/olympic.c | 2 +- drivers/net/tokenring/tms380tr.c | 3 +- drivers/net/typhoon.c | 7 ++--- drivers/net/wan/cosa.c | 6 ++-- drivers/net/wan/dscc4.c | 9 ++---- drivers/net/wan/farsync.c | 3 +- drivers/net/wireless/ipw2100.c | 17 ++++------- drivers/net/wireless/prism54/islpci_dev.c | 6 ++-- drivers/net/wireless/prism54/islpci_mgt.c | 5 ++-- include/linux/ibmtr.h | 4 +-- include/linux/netdevice.h | 6 ++-- 17 files changed, 53 insertions(+), 92 deletions(-) (limited to 'include/linux') diff --git a/drivers/net/8139cp.c b/drivers/net/8139cp.c index ebc20d9e7d7b..bd99c268e2da 100644 --- a/drivers/net/8139cp.c +++ b/drivers/net/8139cp.c @@ -1029,8 +1029,7 @@ static void cp_reset_hw (struct cp_private *cp) if (!(cpr8(Cmd) & CmdReset)) return; - set_current_state(TASK_UNINTERRUPTIBLE); - schedule_timeout(10); + schedule_timeout_uninterruptible(10); } printk(KERN_ERR "%s: hardware reset timeout\n", cp->dev->name); diff --git a/drivers/net/hp100.c b/drivers/net/hp100.c index cf0ac6fda1a1..b71fab6e34f4 100644 --- a/drivers/net/hp100.c +++ b/drivers/net/hp100.c @@ -2517,10 +2517,8 @@ static int hp100_down_vg_link(struct net_device *dev) do { if (hp100_inb(VG_LAN_CFG_1) & HP100_LINK_CABLE_ST) break; - if (!in_interrupt()) { - set_current_state(TASK_INTERRUPTIBLE); - schedule_timeout(1); - } + if (!in_interrupt()) + schedule_timeout_interruptible(1); } while (time_after(time, jiffies)); if (time_after_eq(jiffies, time)) /* no signal->no logout */ @@ -2536,10 +2534,8 @@ static int hp100_down_vg_link(struct net_device *dev) do { if (!(hp100_inb(VG_LAN_CFG_1) & HP100_LINK_UP_ST)) break; - if (!in_interrupt()) { - set_current_state(TASK_INTERRUPTIBLE); - schedule_timeout(1); - } + if (!in_interrupt()) + schedule_timeout_interruptible(1); } while (time_after(time, jiffies)); #ifdef HP100_DEBUG @@ -2577,10 +2573,8 @@ static int hp100_down_vg_link(struct net_device *dev) do { if (!(hp100_inb(MAC_CFG_4) & HP100_MAC_SEL_ST)) break; - if (!in_interrupt()) { - set_current_state(TASK_INTERRUPTIBLE); - schedule_timeout(1); - } + if (!in_interrupt()) + schedule_timeout_interruptible(1); } while (time_after(time, jiffies)); hp100_orb(HP100_AUTO_MODE, MAC_CFG_3); /* Autosel back on */ @@ -2591,10 +2585,8 @@ static int hp100_down_vg_link(struct net_device *dev) do { if ((hp100_inb(VG_LAN_CFG_1) & HP100_LINK_CABLE_ST) == 0) break; - if (!in_interrupt()) { - set_current_state(TASK_INTERRUPTIBLE); - schedule_timeout(1); - } + if (!in_interrupt()) + schedule_timeout_interruptible(1); } while (time_after(time, jiffies)); if (time_before_eq(time, jiffies)) { @@ -2606,10 +2598,8 @@ static int hp100_down_vg_link(struct net_device *dev) time = jiffies + (2 * HZ); /* This seems to take a while.... */ do { - if (!in_interrupt()) { - set_current_state(TASK_INTERRUPTIBLE); - schedule_timeout(1); - } + if (!in_interrupt()) + schedule_timeout_interruptible(1); } while (time_after(time, jiffies)); return 0; @@ -2659,10 +2649,8 @@ static int hp100_login_to_vg_hub(struct net_device *dev, u_short force_relogin) do { if (~(hp100_inb(VG_LAN_CFG_1) & HP100_LINK_UP_ST)) break; - if (!in_interrupt()) { - set_current_state(TASK_INTERRUPTIBLE); - schedule_timeout(1); - } + if (!in_interrupt()) + schedule_timeout_interruptible(1); } while (time_after(time, jiffies)); /* Start an addressed training and optionally request promiscuous port */ @@ -2697,10 +2685,8 @@ static int hp100_login_to_vg_hub(struct net_device *dev, u_short force_relogin) do { if (hp100_inb(VG_LAN_CFG_1) & HP100_LINK_CABLE_ST) break; - if (!in_interrupt()) { - set_current_state(TASK_INTERRUPTIBLE); - schedule_timeout(1); - } + if (!in_interrupt()) + schedule_timeout_interruptible(1); } while (time_before(jiffies, time)); if (time_after_eq(jiffies, time)) { @@ -2723,10 +2709,8 @@ static int hp100_login_to_vg_hub(struct net_device *dev, u_short force_relogin) #endif break; } - if (!in_interrupt()) { - set_current_state(TASK_INTERRUPTIBLE); - schedule_timeout(1); - } + if (!in_interrupt()) + schedule_timeout_interruptible(1); } while (time_after(time, jiffies)); } diff --git a/drivers/net/irda/stir4200.c b/drivers/net/irda/stir4200.c index 15f207323d97..3961a754e920 100644 --- a/drivers/net/irda/stir4200.c +++ b/drivers/net/irda/stir4200.c @@ -678,10 +678,9 @@ static void turnaround_delay(const struct stir_cb *stir, long us) return; ticks = us / (1000000 / HZ); - if (ticks > 0) { - current->state = TASK_INTERRUPTIBLE; - schedule_timeout(1 + ticks); - } else + if (ticks > 0) + schedule_timeout_interruptible(1 + ticks); + else udelay(us); } diff --git a/drivers/net/ixgb/ixgb_ethtool.c b/drivers/net/ixgb/ixgb_ethtool.c index 319ee4cd70d8..04e47189d830 100644 --- a/drivers/net/ixgb/ixgb_ethtool.c +++ b/drivers/net/ixgb/ixgb_ethtool.c @@ -645,11 +645,10 @@ ixgb_phys_id(struct net_device *netdev, uint32_t data) mod_timer(&adapter->blink_timer, jiffies); - set_current_state(TASK_INTERRUPTIBLE); - if(data) - schedule_timeout(data * HZ); + if (data) + schedule_timeout_interruptible(data * HZ); else - schedule_timeout(MAX_SCHEDULE_TIMEOUT); + schedule_timeout_interruptible(MAX_SCHEDULE_TIMEOUT); del_timer_sync(&adapter->blink_timer); ixgb_led_off(&adapter->hw); diff --git a/drivers/net/ns83820.c b/drivers/net/ns83820.c index e64df4d0800b..ed72a23c85dd 100644 --- a/drivers/net/ns83820.c +++ b/drivers/net/ns83820.c @@ -1632,8 +1632,7 @@ static void ns83820_run_bist(struct net_device *ndev, const char *name, u32 enab timed_out = 1; break; } - set_current_state(TASK_UNINTERRUPTIBLE); - schedule_timeout(1); + schedule_timeout_uninterruptible(1); } if (status & fail) diff --git a/drivers/net/tokenring/ibmtr.c b/drivers/net/tokenring/ibmtr.c index e7b001017b9a..8154bbbb7792 100644 --- a/drivers/net/tokenring/ibmtr.c +++ b/drivers/net/tokenring/ibmtr.c @@ -318,7 +318,7 @@ static void ibmtr_cleanup_card(struct net_device *dev) if (dev->base_addr) { outb(0,dev->base_addr+ADAPTRESET); - schedule_timeout(TR_RST_TIME); /* wait 50ms */ + schedule_timeout_uninterruptible(TR_RST_TIME); /* wait 50ms */ outb(0,dev->base_addr+ADAPTRESETREL); } @@ -859,8 +859,7 @@ static int tok_init_card(struct net_device *dev) writeb(~INT_ENABLE, ti->mmio + ACA_OFFSET + ACA_RESET + ISRP_EVEN); outb(0, PIOaddr + ADAPTRESET); - current->state=TASK_UNINTERRUPTIBLE; - schedule_timeout(TR_RST_TIME); /* wait 50ms */ + schedule_timeout_uninterruptible(TR_RST_TIME); /* wait 50ms */ outb(0, PIOaddr + ADAPTRESETREL); #ifdef ENABLE_PAGING @@ -908,8 +907,8 @@ static int tok_open(struct net_device *dev) DPRINTK("Adapter is up and running\n"); return 0; } - current->state=TASK_INTERRUPTIBLE; - i=schedule_timeout(TR_RETRY_INTERVAL); /* wait 30 seconds */ + i=schedule_timeout_interruptible(TR_RETRY_INTERVAL); + /* wait 30 seconds */ if(i!=0) break; /*prob. a signal, like the i>24*HZ case above */ } outb(0, dev->base_addr + ADAPTRESET);/* kill pending interrupts*/ diff --git a/drivers/net/tokenring/olympic.c b/drivers/net/tokenring/olympic.c index 9e7923192a49..05477d24fd49 100644 --- a/drivers/net/tokenring/olympic.c +++ b/drivers/net/tokenring/olympic.c @@ -1101,7 +1101,7 @@ static int olympic_close(struct net_device *dev) while(olympic_priv->srb_queued) { - t = schedule_timeout(60*HZ); + t = schedule_timeout_interruptible(60*HZ); if(signal_pending(current)) { printk(KERN_WARNING "%s: SRB timed out.\n",dev->name); diff --git a/drivers/net/tokenring/tms380tr.c b/drivers/net/tokenring/tms380tr.c index 2e39bf1f7462..c1925590a0e1 100644 --- a/drivers/net/tokenring/tms380tr.c +++ b/drivers/net/tokenring/tms380tr.c @@ -1243,8 +1243,7 @@ void tms380tr_wait(unsigned long time) tmp = jiffies + time/(1000000/HZ); do { - current->state = TASK_INTERRUPTIBLE; - tmp = schedule_timeout(tmp); + tmp = schedule_timeout_interruptible(tmp); } while(time_after(tmp, jiffies)); #else udelay(time); diff --git a/drivers/net/typhoon.c b/drivers/net/typhoon.c index ecfa6f8805ce..4c76cb794bfb 100644 --- a/drivers/net/typhoon.c +++ b/drivers/net/typhoon.c @@ -419,10 +419,9 @@ typhoon_reset(void __iomem *ioaddr, int wait_type) TYPHOON_STATUS_WAITING_FOR_HOST) goto out; - if(wait_type == WaitSleep) { - set_current_state(TASK_UNINTERRUPTIBLE); - schedule_timeout(1); - } else + if(wait_type == WaitSleep) + schedule_timeout_uninterruptible(1); + else udelay(TYPHOON_UDELAY); } diff --git a/drivers/net/wan/cosa.c b/drivers/net/wan/cosa.c index 7ff814fd65d0..ae9e897c255e 100644 --- a/drivers/net/wan/cosa.c +++ b/drivers/net/wan/cosa.c @@ -1617,8 +1617,7 @@ static int get_wait_data(struct cosa_data *cosa) return r; } /* sleep if not ready to read */ - set_current_state(TASK_INTERRUPTIBLE); - schedule_timeout(1); + schedule_timeout_interruptible(1); } printk(KERN_INFO "cosa: timeout in get_wait_data (status 0x%x)\n", cosa_getstatus(cosa)); @@ -1644,8 +1643,7 @@ static int put_wait_data(struct cosa_data *cosa, int data) } #if 0 /* sleep if not ready to read */ - current->state = TASK_INTERRUPTIBLE; - schedule_timeout(1); + schedule_timeout_interruptible(1); #endif } printk(KERN_INFO "cosa%d: timeout in put_wait_data (status 0x%x)\n", diff --git a/drivers/net/wan/dscc4.c b/drivers/net/wan/dscc4.c index 520a77a798e2..0c1ab4ac8bdb 100644 --- a/drivers/net/wan/dscc4.c +++ b/drivers/net/wan/dscc4.c @@ -542,8 +542,7 @@ static int dscc4_wait_ack_cec(struct dscc4_dev_priv *dpriv, msg, i); goto done; } - set_current_state(TASK_UNINTERRUPTIBLE); - schedule_timeout(10); + schedule_timeout_uninterruptible(10); rmb(); } while (++i > 0); printk(KERN_ERR "%s: %s timeout\n", dev->name, msg); @@ -588,8 +587,7 @@ static inline int dscc4_xpr_ack(struct dscc4_dev_priv *dpriv) (dpriv->iqtx[cur] & Xpr)) break; smp_rmb(); - set_current_state(TASK_UNINTERRUPTIBLE); - schedule_timeout(10); + schedule_timeout_uninterruptible(10); } while (++i > 0); return (i >= 0 ) ? i : -EAGAIN; @@ -1035,8 +1033,7 @@ static void dscc4_pci_reset(struct pci_dev *pdev, void __iomem *ioaddr) /* Flush posted writes */ readl(ioaddr + GSTAR); - set_current_state(TASK_UNINTERRUPTIBLE); - schedule_timeout(10); + schedule_timeout_uninterruptible(10); for (i = 0; i < 16; i++) pci_write_config_dword(pdev, i << 2, dscc4_pci_config_store[i]); diff --git a/drivers/net/wan/farsync.c b/drivers/net/wan/farsync.c index 2c83cca34b86..10befb02d768 100644 --- a/drivers/net/wan/farsync.c +++ b/drivers/net/wan/farsync.c @@ -980,8 +980,7 @@ fst_issue_cmd(struct fst_port_info *port, unsigned short cmd) /* Wait for any previous command to complete */ while (mbval > NAK) { spin_unlock_irqrestore(&card->card_lock, flags); - set_current_state(TASK_UNINTERRUPTIBLE); - schedule_timeout(1); + schedule_timeout_uninterruptible(1); spin_lock_irqsave(&card->card_lock, flags); if (++safety > 2000) { diff --git a/drivers/net/wireless/ipw2100.c b/drivers/net/wireless/ipw2100.c index 2414e6493aa5..e5cdb5bfabc8 100644 --- a/drivers/net/wireless/ipw2100.c +++ b/drivers/net/wireless/ipw2100.c @@ -800,8 +800,7 @@ static int ipw2100_hw_send_command(struct ipw2100_priv *priv, * doesn't seem to have as many firmware restart cycles... * * As a test, we're sticking in a 1/100s delay here */ - set_current_state(TASK_UNINTERRUPTIBLE); - schedule_timeout(HZ / 100); + schedule_timeout_uninterruptible(msecs_to_jiffies(10)); return 0; @@ -1256,8 +1255,7 @@ static int ipw2100_start_adapter(struct ipw2100_priv *priv) IPW_DEBUG_FW("Waiting for f/w initialization to complete...\n"); i = 5000; do { - set_current_state(TASK_UNINTERRUPTIBLE); - schedule_timeout(40 * HZ / 1000); + schedule_timeout_uninterruptible(msecs_to_jiffies(40)); /* Todo... wait for sync command ... */ read_register(priv->net_dev, IPW_REG_INTA, &inta); @@ -1411,8 +1409,7 @@ static int ipw2100_hw_phy_off(struct ipw2100_priv *priv) (val2 & IPW2100_COMMAND_PHY_OFF)) return 0; - set_current_state(TASK_UNINTERRUPTIBLE); - schedule_timeout(HW_PHY_OFF_LOOP_DELAY); + schedule_timeout_uninterruptible(HW_PHY_OFF_LOOP_DELAY); } return -EIO; @@ -1466,7 +1463,7 @@ fail_up: static int ipw2100_hw_stop_adapter(struct ipw2100_priv *priv) { -#define HW_POWER_DOWN_DELAY (HZ / 10) +#define HW_POWER_DOWN_DELAY (msecs_to_jiffies(100)) struct host_command cmd = { .host_command = HOST_PRE_POWER_DOWN, @@ -1520,10 +1517,8 @@ static int ipw2100_hw_stop_adapter(struct ipw2100_priv *priv) printk(KERN_WARNING DRV_NAME ": " "%s: Power down command failed: Error %d\n", priv->net_dev->name, err); - else { - set_current_state(TASK_UNINTERRUPTIBLE); - schedule_timeout(HW_POWER_DOWN_DELAY); - } + else + schedule_timeout_uninterruptible(HW_POWER_DOWN_DELAY); } priv->status &= ~STATUS_ENABLED; diff --git a/drivers/net/wireless/prism54/islpci_dev.c b/drivers/net/wireless/prism54/islpci_dev.c index 6f13d4a8e2d3..10cce514c15d 100644 --- a/drivers/net/wireless/prism54/islpci_dev.c +++ b/drivers/net/wireless/prism54/islpci_dev.c @@ -439,8 +439,7 @@ prism54_bring_down(islpci_private *priv) wmb(); /* wait a while for the device to reset */ - set_current_state(TASK_UNINTERRUPTIBLE); - schedule_timeout(50*HZ/1000); + schedule_timeout_uninterruptible(msecs_to_jiffies(50)); return 0; } @@ -491,8 +490,7 @@ islpci_reset_if(islpci_private *priv) /* The software reset acknowledge needs about 220 msec here. * Be conservative and wait for up to one second. */ - set_current_state(TASK_UNINTERRUPTIBLE); - remaining = schedule_timeout(HZ); + remaining = schedule_timeout_uninterruptible(HZ); if(remaining > 0) { result = 0; diff --git a/drivers/net/wireless/prism54/islpci_mgt.c b/drivers/net/wireless/prism54/islpci_mgt.c index b6f2e5a223be..4937a5ad4b2c 100644 --- a/drivers/net/wireless/prism54/islpci_mgt.c +++ b/drivers/net/wireless/prism54/islpci_mgt.c @@ -455,7 +455,7 @@ islpci_mgt_transaction(struct net_device *ndev, struct islpci_mgmtframe **recvframe) { islpci_private *priv = netdev_priv(ndev); - const long wait_cycle_jiffies = (ISL38XX_WAIT_CYCLE * 10 * HZ) / 1000; + const long wait_cycle_jiffies = msecs_to_jiffies(ISL38XX_WAIT_CYCLE * 10); long timeout_left = ISL38XX_MAX_WAIT_CYCLES * wait_cycle_jiffies; int err; DEFINE_WAIT(wait); @@ -475,8 +475,7 @@ islpci_mgt_transaction(struct net_device *ndev, int timeleft; struct islpci_mgmtframe *frame; - set_current_state(TASK_UNINTERRUPTIBLE); - timeleft = schedule_timeout(wait_cycle_jiffies); + timeleft = schedule_timeout_uninterruptible(wait_cycle_jiffies); frame = xchg(&priv->mgmt_received, NULL); if (frame) { if (frame->header->oid == oid) { diff --git a/include/linux/ibmtr.h b/include/linux/ibmtr.h index 2ef0b21517fb..1c7a0dd5536a 100644 --- a/include/linux/ibmtr.h +++ b/include/linux/ibmtr.h @@ -7,8 +7,8 @@ /* ported to the Alpha architecture 02/20/96 (just used the HZ macro) */ #define TR_RETRY_INTERVAL (30*HZ) /* 500 on PC = 5 s */ -#define TR_RST_TIME (HZ/20) /* 5 on PC = 50 ms */ -#define TR_BUSY_INTERVAL (HZ/5) /* 5 on PC = 200 ms */ +#define TR_RST_TIME (msecs_to_jiffies(50)) /* 5 on PC = 50 ms */ +#define TR_BUSY_INTERVAL (msecs_to_jiffies(200)) /* 5 on PC = 200 ms */ #define TR_SPIN_INTERVAL (3*HZ) /* 3 seconds before init timeout */ #define TR_ISA 1 diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index 7c717907896d..98c98e6cd4f3 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -852,11 +852,9 @@ static inline void netif_rx_complete(struct net_device *dev) static inline void netif_poll_disable(struct net_device *dev) { - while (test_and_set_bit(__LINK_STATE_RX_SCHED, &dev->state)) { + while (test_and_set_bit(__LINK_STATE_RX_SCHED, &dev->state)) /* No hurry. */ - current->state = TASK_INTERRUPTIBLE; - schedule_timeout(1); - } + schedule_timeout_interruptible(1); } static inline void netif_poll_enable(struct net_device *dev) -- cgit v1.2.3 From 8e18d1f9c9dcbf2de5b79cad771ed639983ab6cd Mon Sep 17 00:00:00 2001 From: Tobias Klauser Date: Sat, 10 Sep 2005 14:45:00 -0700 Subject: [PATCH] Replace drivers/net/wan custom ctype macros with standard ones Replace the custom is_digit()/is_hex_digit() macros with isdigit()/isxdigit() from Additionaly remove unused macro is_alpha() from Signed-off-by: Tobias Klauser Cc: Jeff Garzik Signed-off-by: Andrew Morton Signed-off-by: Jeff Garzik --- drivers/net/wan/cycx_x25.c | 5 +++-- drivers/net/wan/sdla_fr.c | 4 ++-- drivers/net/wan/sdla_x25.c | 8 ++++---- include/linux/cyclomx.h | 2 -- include/linux/wanpipe.h | 9 --------- 5 files changed, 9 insertions(+), 19 deletions(-) (limited to 'include/linux') diff --git a/drivers/net/wan/cycx_x25.c b/drivers/net/wan/cycx_x25.c index 02d57c0b4243..a631d1c2fa14 100644 --- a/drivers/net/wan/cycx_x25.c +++ b/drivers/net/wan/cycx_x25.c @@ -78,6 +78,7 @@ #define CYCLOMX_X25_DEBUG 1 +#include /* isdigit() */ #include /* return codes */ #include /* ARPHRD_HWX25 */ #include /* printk(), and other useful stuff */ @@ -418,7 +419,7 @@ static int cycx_wan_new_if(struct wan_device *wandev, struct net_device *dev, /* Set channel timeouts (default if not specified) */ chan->idle_tmout = conf->idle_timeout ? conf->idle_timeout : 90; - } else if (is_digit(conf->addr[0])) { /* PVC */ + } else if (isdigit(conf->addr[0])) { /* PVC */ s16 lcn = dec_to_uint(conf->addr, 0); if (lcn >= card->u.x.lo_pvc && lcn <= card->u.x.hi_pvc) @@ -1531,7 +1532,7 @@ static unsigned dec_to_uint(u8 *str, int len) if (!len) len = strlen(str); - for (; len && is_digit(*str); ++str, --len) + for (; len && isdigit(*str); ++str, --len) val = (val * 10) + (*str - (unsigned) '0'); return val; diff --git a/drivers/net/wan/sdla_fr.c b/drivers/net/wan/sdla_fr.c index 0497dbdb8631..7f1ce9d4333e 100644 --- a/drivers/net/wan/sdla_fr.c +++ b/drivers/net/wan/sdla_fr.c @@ -822,7 +822,7 @@ static int new_if(struct wan_device* wandev, struct net_device* dev, chan->card = card; /* verify media address */ - if (is_digit(conf->addr[0])) { + if (isdigit(conf->addr[0])) { dlci = dec_to_uint(conf->addr, 0); @@ -3456,7 +3456,7 @@ static unsigned int dec_to_uint (unsigned char* str, int len) if (!len) len = strlen(str); - for (val = 0; len && is_digit(*str); ++str, --len) + for (val = 0; len && isdigit(*str); ++str, --len) val = (val * 10) + (*str - (unsigned)'0'); return val; diff --git a/drivers/net/wan/sdla_x25.c b/drivers/net/wan/sdla_x25.c index 8a95d61a2f8f..63f846d6f3a6 100644 --- a/drivers/net/wan/sdla_x25.c +++ b/drivers/net/wan/sdla_x25.c @@ -957,7 +957,7 @@ static int new_if(struct wan_device* wandev, struct net_device* dev, chan->hold_timeout = (conf->hold_timeout) ? conf->hold_timeout : 10; - }else if (is_digit(conf->addr[0])){ /* PVC */ + }else if (isdigit(conf->addr[0])){ /* PVC */ int lcn = dec_to_uint(conf->addr, 0); if ((lcn >= card->u.x.lo_pvc) && (lcn <= card->u.x.hi_pvc)){ @@ -3875,7 +3875,7 @@ static unsigned int dec_to_uint (unsigned char* str, int len) if (!len) len = strlen(str); - for (val = 0; len && is_digit(*str); ++str, --len) + for (val = 0; len && isdigit(*str); ++str, --len) val = (val * 10) + (*str - (unsigned)'0'); return val; @@ -3896,9 +3896,9 @@ static unsigned int hex_to_uint (unsigned char* str, int len) for (val = 0; len; ++str, --len) { ch = *str; - if (is_digit(ch)) + if (isdigit(ch)) val = (val << 4) + (ch - (unsigned)'0'); - else if (is_hex_digit(ch)) + else if (isxdigit(ch)) val = (val << 4) + ((ch & 0xDF) - (unsigned)'A' + 10); else break; } diff --git a/include/linux/cyclomx.h b/include/linux/cyclomx.h index 04fa7dff079c..300d704bdb9a 100644 --- a/include/linux/cyclomx.h +++ b/include/linux/cyclomx.h @@ -37,8 +37,6 @@ #include #endif -#define is_digit(ch) (((ch)>=(unsigned)'0'&&(ch)<=(unsigned)'9')?1:0) - /* Adapter Data Space. * This structure is needed because we handle multiple cards, otherwise * static data would do it. diff --git a/include/linux/wanpipe.h b/include/linux/wanpipe.h index 167d956c492b..dae9860091dd 100644 --- a/include/linux/wanpipe.h +++ b/include/linux/wanpipe.h @@ -265,15 +265,6 @@ typedef struct { #include #include - -#define is_digit(ch) (((ch)>=(unsigned)'0'&&(ch)<=(unsigned)'9')?1:0) -#define is_alpha(ch) ((((ch)>=(unsigned)'a'&&(ch)<=(unsigned)'z')||\ - ((ch)>=(unsigned)'A'&&(ch)<=(unsigned)'Z'))?1:0) -#define is_hex_digit(ch) ((((ch)>=(unsigned)'0'&&(ch)<=(unsigned)'9')||\ - ((ch)>=(unsigned)'a'&&(ch)<=(unsigned)'f')||\ - ((ch)>=(unsigned)'A'&&(ch)<=(unsigned)'F'))?1:0) - - /****** Data Structures *****************************************************/ /* Adapter Data Space. -- cgit v1.2.3 From 7665a08928f241247afe8c76865cdbe4ef5489bf Mon Sep 17 00:00:00 2001 From: Adrian Bunk Date: Fri, 9 Sep 2005 23:17:28 -0700 Subject: [PATCH] drivers/net/wan/: possible cleanups This patch contains possible cleanups including the following: - make needlessly global code static - #if 0 the following unused global function: - sdladrv.c: sdla_intde - remove the following unused global variable: - lmc_media.c: lmc_t1_cables - remove the following unneeded EXPORT_SYMBOL's: - cycx_drv.c: cycx_inten - sdladrv.c: sdla_inten - sdladrv.c: sdla_intde - sdladrv.c: sdla_intack - sdladrv.c: sdla_intr - syncppp.c: sppp_input - syncppp.c: sppp_change_mtu Signed-off-by: Adrian Bunk Signed-off-by: Andrew Morton Signed-off-by: Jeff Garzik --- drivers/net/wan/cycx_drv.c | 7 ++-- drivers/net/wan/cycx_main.c | 2 +- drivers/net/wan/dscc4.c | 14 ++++--- drivers/net/wan/farsync.c | 24 ++++++------ drivers/net/wan/hdlc_fr.c | 2 +- drivers/net/wan/lmc/lmc_debug.c | 10 ++--- drivers/net/wan/lmc/lmc_media.c | 8 ---- drivers/net/wan/pc300.h | 16 -------- drivers/net/wan/pc300_drv.c | 87 +++++++++++++++++++++-------------------- drivers/net/wan/pc300_tty.c | 18 ++++----- drivers/net/wan/sdla.c | 20 +++++----- drivers/net/wan/sdladrv.c | 16 +++----- drivers/net/wan/syncppp.c | 10 ++--- include/linux/cycx_drv.h | 1 - include/linux/sdladrv.h | 4 -- include/net/syncppp.h | 1 - 16 files changed, 101 insertions(+), 139 deletions(-) (limited to 'include/linux') diff --git a/drivers/net/wan/cycx_drv.c b/drivers/net/wan/cycx_drv.c index 9e56fc346ba4..e6d005726aad 100644 --- a/drivers/net/wan/cycx_drv.c +++ b/drivers/net/wan/cycx_drv.c @@ -109,7 +109,7 @@ static long cycx_2x_irq_options[] = { 7, 3, 5, 9, 10, 11, 12, 15 }; * < 0 error. * Context: process */ -int __init cycx_drv_init(void) +static int __init cycx_drv_init(void) { printk(KERN_INFO "%s v%u.%u %s\n", fullname, MOD_VERSION, MOD_RELEASE, copyright); @@ -119,7 +119,7 @@ int __init cycx_drv_init(void) /* Module 'remove' entry point. * o release all remaining system resources */ -void cycx_drv_cleanup(void) +static void cycx_drv_cleanup(void) { } @@ -184,8 +184,7 @@ int cycx_down(struct cycx_hw *hw) } /* Enable interrupt generation. */ -EXPORT_SYMBOL(cycx_inten); -void cycx_inten(struct cycx_hw *hw) +static void cycx_inten(struct cycx_hw *hw) { writeb(0, hw->dpmbase); } diff --git a/drivers/net/wan/cycx_main.c b/drivers/net/wan/cycx_main.c index 7b48064364dc..430b1f630fb4 100644 --- a/drivers/net/wan/cycx_main.c +++ b/drivers/net/wan/cycx_main.c @@ -103,7 +103,7 @@ static struct cycx_device *cycx_card_array; /* adapter data space */ * < 0 error. * Context: process */ -int __init cycx_init(void) +static int __init cycx_init(void) { int cnt, err = -ENOMEM; diff --git a/drivers/net/wan/dscc4.c b/drivers/net/wan/dscc4.c index 0c1ab4ac8bdb..2f61a47b4716 100644 --- a/drivers/net/wan/dscc4.c +++ b/drivers/net/wan/dscc4.c @@ -446,8 +446,8 @@ static inline unsigned int dscc4_tx_quiescent(struct dscc4_dev_priv *dpriv, return readl(dpriv->base_addr + CH0FTDA + dpriv->dev_id*4) == dpriv->ltda; } -int state_check(u32 state, struct dscc4_dev_priv *dpriv, struct net_device *dev, - const char *msg) +static int state_check(u32 state, struct dscc4_dev_priv *dpriv, + struct net_device *dev, const char *msg) { int ret = 0; @@ -466,8 +466,9 @@ int state_check(u32 state, struct dscc4_dev_priv *dpriv, struct net_device *dev, return ret; } -void dscc4_tx_print(struct net_device *dev, struct dscc4_dev_priv *dpriv, - char *msg) +static void dscc4_tx_print(struct net_device *dev, + struct dscc4_dev_priv *dpriv, + char *msg) { printk(KERN_DEBUG "%s: tx_current=%02d tx_dirty=%02d (%s)\n", dev->name, dpriv->tx_current, dpriv->tx_dirty, msg); @@ -507,7 +508,8 @@ static void dscc4_release_ring(struct dscc4_dev_priv *dpriv) } } -inline int try_get_rx_skb(struct dscc4_dev_priv *dpriv, struct net_device *dev) +static inline int try_get_rx_skb(struct dscc4_dev_priv *dpriv, + struct net_device *dev) { unsigned int dirty = dpriv->rx_dirty%RX_RING_SIZE; struct RxFD *rx_fd = dpriv->rx_fd + dirty; @@ -1891,7 +1893,7 @@ try: * It failed and locked solid. Thus the introduction of a dummy skb. * Problem is acknowledged in errata sheet DS5. Joy :o/ */ -struct sk_buff *dscc4_init_dummy_skb(struct dscc4_dev_priv *dpriv) +static struct sk_buff *dscc4_init_dummy_skb(struct dscc4_dev_priv *dpriv) { struct sk_buff *skb; diff --git a/drivers/net/wan/farsync.c b/drivers/net/wan/farsync.c index 10befb02d768..7981a2c7906e 100644 --- a/drivers/net/wan/farsync.c +++ b/drivers/net/wan/farsync.c @@ -74,11 +74,11 @@ MODULE_LICENSE("GPL"); /* * Modules parameters and associated varaibles */ -int fst_txq_low = FST_LOW_WATER_MARK; -int fst_txq_high = FST_HIGH_WATER_MARK; -int fst_max_reads = 7; -int fst_excluded_cards = 0; -int fst_excluded_list[FST_MAX_CARDS]; +static int fst_txq_low = FST_LOW_WATER_MARK; +static int fst_txq_high = FST_HIGH_WATER_MARK; +static int fst_max_reads = 7; +static int fst_excluded_cards = 0; +static int fst_excluded_list[FST_MAX_CARDS]; module_param(fst_txq_low, int, 0); module_param(fst_txq_high, int, 0); @@ -572,13 +572,13 @@ static void do_bottom_half_rx(struct fst_card_info *card); static void fst_process_tx_work_q(unsigned long work_q); static void fst_process_int_work_q(unsigned long work_q); -DECLARE_TASKLET(fst_tx_task, fst_process_tx_work_q, 0); -DECLARE_TASKLET(fst_int_task, fst_process_int_work_q, 0); +static DECLARE_TASKLET(fst_tx_task, fst_process_tx_work_q, 0); +static DECLARE_TASKLET(fst_int_task, fst_process_int_work_q, 0); -struct fst_card_info *fst_card_array[FST_MAX_CARDS]; -spinlock_t fst_work_q_lock; -u64 fst_work_txq; -u64 fst_work_intq; +static struct fst_card_info *fst_card_array[FST_MAX_CARDS]; +static spinlock_t fst_work_q_lock; +static u64 fst_work_txq; +static u64 fst_work_intq; static void fst_q_work_item(u64 * queue, int card_index) @@ -1497,7 +1497,7 @@ do_bottom_half_rx(struct fst_card_info *card) * The interrupt service routine * Dev_id is our fst_card_info pointer */ -irqreturn_t +static irqreturn_t fst_intr(int irq, void *dev_id, struct pt_regs *regs) { struct fst_card_info *card; diff --git a/drivers/net/wan/hdlc_fr.c b/drivers/net/wan/hdlc_fr.c index a5d6891c9d4c..e1601d35dced 100644 --- a/drivers/net/wan/hdlc_fr.c +++ b/drivers/net/wan/hdlc_fr.c @@ -330,7 +330,7 @@ static int pvc_close(struct net_device *dev) -int pvc_ioctl(struct net_device *dev, struct ifreq *ifr, int cmd) +static int pvc_ioctl(struct net_device *dev, struct ifreq *ifr, int cmd) { pvc_device *pvc = dev_to_pvc(dev); fr_proto_pvc_info info; diff --git a/drivers/net/wan/lmc/lmc_debug.c b/drivers/net/wan/lmc/lmc_debug.c index 9dccd9546a17..3b94352b0d03 100644 --- a/drivers/net/wan/lmc/lmc_debug.c +++ b/drivers/net/wan/lmc/lmc_debug.c @@ -8,10 +8,10 @@ /* * Prints out len, max to 80 octets using printk, 20 per line */ -void lmcConsoleLog(char *type, unsigned char *ucData, int iLen) -{ #ifdef DEBUG #ifdef LMC_PACKET_LOG +void lmcConsoleLog(char *type, unsigned char *ucData, int iLen) +{ int iNewLine = 1; char str[80], *pstr; @@ -43,26 +43,24 @@ void lmcConsoleLog(char *type, unsigned char *ucData, int iLen) } sprintf(pstr, "\n"); printk(str); +} #endif #endif -} #ifdef DEBUG u_int32_t lmcEventLogIndex = 0; u_int32_t lmcEventLogBuf[LMC_EVENTLOGSIZE * LMC_EVENTLOGARGS]; -#endif void lmcEventLog (u_int32_t EventNum, u_int32_t arg2, u_int32_t arg3) { -#ifdef DEBUG lmcEventLogBuf[lmcEventLogIndex++] = EventNum; lmcEventLogBuf[lmcEventLogIndex++] = arg2; lmcEventLogBuf[lmcEventLogIndex++] = arg3; lmcEventLogBuf[lmcEventLogIndex++] = jiffies; lmcEventLogIndex &= (LMC_EVENTLOGSIZE * LMC_EVENTLOGARGS) - 1; -#endif } +#endif /* DEBUG */ void lmc_trace(struct net_device *dev, char *msg){ #ifdef LMC_TRACE diff --git a/drivers/net/wan/lmc/lmc_media.c b/drivers/net/wan/lmc/lmc_media.c index f55ce76b00ed..af8b55fdd9d9 100644 --- a/drivers/net/wan/lmc/lmc_media.c +++ b/drivers/net/wan/lmc/lmc_media.c @@ -47,14 +47,6 @@ * of the GNU General Public License version 2, incorporated herein by reference. */ -/* - * For lack of a better place, put the SSI cable stuff here. - */ -char *lmc_t1_cables[] = { - "V.10/RS423", "EIA530A", "reserved", "X.21", "V.35", - "EIA449/EIA530/V.36", "V.28/EIA232", "none", NULL -}; - /* * protocol independent method. */ diff --git a/drivers/net/wan/pc300.h b/drivers/net/wan/pc300.h index 73401b0f0151..2024b26b99e6 100644 --- a/drivers/net/wan/pc300.h +++ b/drivers/net/wan/pc300.h @@ -472,24 +472,8 @@ enum pc300_loopback_cmds { #ifdef __KERNEL__ /* Function Prototypes */ -int dma_buf_write(pc300_t *, int, ucchar *, int); -int dma_buf_read(pc300_t *, int, struct sk_buff *); void tx_dma_start(pc300_t *, int); -void rx_dma_start(pc300_t *, int); -void tx_dma_stop(pc300_t *, int); -void rx_dma_stop(pc300_t *, int); -int cpc_queue_xmit(struct sk_buff *, struct net_device *); -void cpc_net_rx(struct net_device *); -void cpc_sca_status(pc300_t *, int); -int cpc_change_mtu(struct net_device *, int); -int cpc_ioctl(struct net_device *, struct ifreq *, int); -int ch_config(pc300dev_t *); -int rx_config(pc300dev_t *); -int tx_config(pc300dev_t *); -void cpc_opench(pc300dev_t *); -void cpc_closech(pc300dev_t *); int cpc_open(struct net_device *dev); -int cpc_close(struct net_device *dev); int cpc_set_media(hdlc_device *, int); #endif /* __KERNEL__ */ diff --git a/drivers/net/wan/pc300_drv.c b/drivers/net/wan/pc300_drv.c index 3e7753b10717..a3e65d1bc19b 100644 --- a/drivers/net/wan/pc300_drv.c +++ b/drivers/net/wan/pc300_drv.c @@ -291,6 +291,7 @@ static uclong detect_ram(pc300_t *); static void plx_init(pc300_t *); static void cpc_trace(struct net_device *, struct sk_buff *, char); static int cpc_attach(struct net_device *, unsigned short, unsigned short); +static int cpc_close(struct net_device *dev); #ifdef CONFIG_PC300_MLPPP void cpc_tty_init(pc300dev_t * dev); @@ -437,7 +438,7 @@ static void rx_dma_buf_check(pc300_t * card, int ch) printk("\n"); } -int dma_get_rx_frame_size(pc300_t * card, int ch) +static int dma_get_rx_frame_size(pc300_t * card, int ch) { volatile pcsca_bd_t __iomem *ptdescr; ucshort first_bd = card->chan[ch].rx_first_bd; @@ -462,7 +463,7 @@ int dma_get_rx_frame_size(pc300_t * card, int ch) * dma_buf_write: writes a frame to the Tx DMA buffers * NOTE: this function writes one frame at a time. */ -int dma_buf_write(pc300_t * card, int ch, ucchar * ptdata, int len) +static int dma_buf_write(pc300_t * card, int ch, ucchar * ptdata, int len) { int i, nchar; volatile pcsca_bd_t __iomem *ptdescr; @@ -503,7 +504,7 @@ int dma_buf_write(pc300_t * card, int ch, ucchar * ptdata, int len) * dma_buf_read: reads a frame from the Rx DMA buffers * NOTE: this function reads one frame at a time. */ -int dma_buf_read(pc300_t * card, int ch, struct sk_buff *skb) +static int dma_buf_read(pc300_t * card, int ch, struct sk_buff *skb) { int nchar; pc300ch_t *chan = (pc300ch_t *) & card->chan[ch]; @@ -560,7 +561,7 @@ int dma_buf_read(pc300_t * card, int ch, struct sk_buff *skb) return (rcvd); } -void tx_dma_stop(pc300_t * card, int ch) +static void tx_dma_stop(pc300_t * card, int ch) { void __iomem *scabase = card->hw.scabase; ucchar drr_ena_bit = 1 << (5 + 2 * ch); @@ -571,7 +572,7 @@ void tx_dma_stop(pc300_t * card, int ch) cpc_writeb(scabase + DRR, drr_rst_bit & ~drr_ena_bit); } -void rx_dma_stop(pc300_t * card, int ch) +static void rx_dma_stop(pc300_t * card, int ch) { void __iomem *scabase = card->hw.scabase; ucchar drr_ena_bit = 1 << (4 + 2 * ch); @@ -582,7 +583,7 @@ void rx_dma_stop(pc300_t * card, int ch) cpc_writeb(scabase + DRR, drr_rst_bit & ~drr_ena_bit); } -void rx_dma_start(pc300_t * card, int ch) +static void rx_dma_start(pc300_t * card, int ch) { void __iomem *scabase = card->hw.scabase; pc300ch_t *chan = (pc300ch_t *) & card->chan[ch]; @@ -607,7 +608,7 @@ void rx_dma_start(pc300_t * card, int ch) /*************************/ /*** FALC Routines ***/ /*************************/ -void falc_issue_cmd(pc300_t * card, int ch, ucchar cmd) +static void falc_issue_cmd(pc300_t * card, int ch, ucchar cmd) { void __iomem *falcbase = card->hw.falcbase; unsigned long i = 0; @@ -622,7 +623,7 @@ void falc_issue_cmd(pc300_t * card, int ch, ucchar cmd) cpc_writeb(falcbase + F_REG(CMDR, ch), cmd); } -void falc_intr_enable(pc300_t * card, int ch) +static void falc_intr_enable(pc300_t * card, int ch) { pc300ch_t *chan = (pc300ch_t *) & card->chan[ch]; pc300chconf_t *conf = (pc300chconf_t *) & chan->conf; @@ -672,7 +673,7 @@ void falc_intr_enable(pc300_t * card, int ch) } } -void falc_open_timeslot(pc300_t * card, int ch, int timeslot) +static void falc_open_timeslot(pc300_t * card, int ch, int timeslot) { void __iomem *falcbase = card->hw.falcbase; ucchar tshf = card->chan[ch].falc.offset; @@ -688,7 +689,7 @@ void falc_open_timeslot(pc300_t * card, int ch, int timeslot) (0x80 >> (timeslot & 0x07))); } -void falc_close_timeslot(pc300_t * card, int ch, int timeslot) +static void falc_close_timeslot(pc300_t * card, int ch, int timeslot) { void __iomem *falcbase = card->hw.falcbase; ucchar tshf = card->chan[ch].falc.offset; @@ -704,7 +705,7 @@ void falc_close_timeslot(pc300_t * card, int ch, int timeslot) ~(0x80 >> (timeslot & 0x07))); } -void falc_close_all_timeslots(pc300_t * card, int ch) +static void falc_close_all_timeslots(pc300_t * card, int ch) { pc300ch_t *chan = (pc300ch_t *) & card->chan[ch]; pc300chconf_t *conf = (pc300chconf_t *) & chan->conf; @@ -726,7 +727,7 @@ void falc_close_all_timeslots(pc300_t * card, int ch) } } -void falc_open_all_timeslots(pc300_t * card, int ch) +static void falc_open_all_timeslots(pc300_t * card, int ch) { pc300ch_t *chan = (pc300ch_t *) & card->chan[ch]; pc300chconf_t *conf = (pc300chconf_t *) & chan->conf; @@ -758,7 +759,7 @@ void falc_open_all_timeslots(pc300_t * card, int ch) } } -void falc_init_timeslot(pc300_t * card, int ch) +static void falc_init_timeslot(pc300_t * card, int ch) { pc300ch_t *chan = (pc300ch_t *) & card->chan[ch]; pc300chconf_t *conf = (pc300chconf_t *) & chan->conf; @@ -776,7 +777,7 @@ void falc_init_timeslot(pc300_t * card, int ch) } } -void falc_enable_comm(pc300_t * card, int ch) +static void falc_enable_comm(pc300_t * card, int ch) { pc300ch_t *chan = (pc300ch_t *) & card->chan[ch]; falc_t *pfalc = (falc_t *) & chan->falc; @@ -792,7 +793,7 @@ void falc_enable_comm(pc300_t * card, int ch) ~((CPLD_REG1_FALC_DCD | CPLD_REG1_FALC_CTS) << (2 * ch))); } -void falc_disable_comm(pc300_t * card, int ch) +static void falc_disable_comm(pc300_t * card, int ch) { pc300ch_t *chan = (pc300ch_t *) & card->chan[ch]; falc_t *pfalc = (falc_t *) & chan->falc; @@ -806,7 +807,7 @@ void falc_disable_comm(pc300_t * card, int ch) ((CPLD_REG1_FALC_DCD | CPLD_REG1_FALC_CTS) << (2 * ch))); } -void falc_init_t1(pc300_t * card, int ch) +static void falc_init_t1(pc300_t * card, int ch) { pc300ch_t *chan = (pc300ch_t *) & card->chan[ch]; pc300chconf_t *conf = (pc300chconf_t *) & chan->conf; @@ -975,7 +976,7 @@ void falc_init_t1(pc300_t * card, int ch) falc_close_all_timeslots(card, ch); } -void falc_init_e1(pc300_t * card, int ch) +static void falc_init_e1(pc300_t * card, int ch) { pc300ch_t *chan = (pc300ch_t *) & card->chan[ch]; pc300chconf_t *conf = (pc300chconf_t *) & chan->conf; @@ -1155,7 +1156,7 @@ void falc_init_e1(pc300_t * card, int ch) falc_close_all_timeslots(card, ch); } -void falc_init_hdlc(pc300_t * card, int ch) +static void falc_init_hdlc(pc300_t * card, int ch) { void __iomem *falcbase = card->hw.falcbase; pc300ch_t *chan = (pc300ch_t *) & card->chan[ch]; @@ -1181,7 +1182,7 @@ void falc_init_hdlc(pc300_t * card, int ch) falc_intr_enable(card, ch); } -void te_config(pc300_t * card, int ch) +static void te_config(pc300_t * card, int ch) { pc300ch_t *chan = (pc300ch_t *) & card->chan[ch]; pc300chconf_t *conf = (pc300chconf_t *) & chan->conf; @@ -1241,7 +1242,7 @@ void te_config(pc300_t * card, int ch) CPC_UNLOCK(card, flags); } -void falc_check_status(pc300_t * card, int ch, unsigned char frs0) +static void falc_check_status(pc300_t * card, int ch, unsigned char frs0) { pc300ch_t *chan = (pc300ch_t *) & card->chan[ch]; pc300chconf_t *conf = (pc300chconf_t *) & chan->conf; @@ -1397,7 +1398,7 @@ void falc_check_status(pc300_t * card, int ch, unsigned char frs0) } } -void falc_update_stats(pc300_t * card, int ch) +static void falc_update_stats(pc300_t * card, int ch) { pc300ch_t *chan = (pc300ch_t *) & card->chan[ch]; pc300chconf_t *conf = (pc300chconf_t *) & chan->conf; @@ -1450,7 +1451,7 @@ void falc_update_stats(pc300_t * card, int ch) * the synchronizer and then sent to the system interface. *---------------------------------------------------------------------------- */ -void falc_remote_loop(pc300_t * card, int ch, int loop_on) +static void falc_remote_loop(pc300_t * card, int ch, int loop_on) { pc300ch_t *chan = (pc300ch_t *) & card->chan[ch]; pc300chconf_t *conf = (pc300chconf_t *) & chan->conf; @@ -1495,7 +1496,7 @@ void falc_remote_loop(pc300_t * card, int ch, int loop_on) * coding must be identical. *---------------------------------------------------------------------------- */ -void falc_local_loop(pc300_t * card, int ch, int loop_on) +static void falc_local_loop(pc300_t * card, int ch, int loop_on) { pc300ch_t *chan = (pc300ch_t *) & card->chan[ch]; falc_t *pfalc = (falc_t *) & chan->falc; @@ -1522,7 +1523,7 @@ void falc_local_loop(pc300_t * card, int ch, int loop_on) * looped. They are originated by the FALC-LH transmitter. *---------------------------------------------------------------------------- */ -void falc_payload_loop(pc300_t * card, int ch, int loop_on) +static void falc_payload_loop(pc300_t * card, int ch, int loop_on) { pc300ch_t *chan = (pc300ch_t *) & card->chan[ch]; pc300chconf_t *conf = (pc300chconf_t *) & chan->conf; @@ -1576,7 +1577,7 @@ void falc_payload_loop(pc300_t * card, int ch, int loop_on) * Description: Turns XLU bit off in the proper register *---------------------------------------------------------------------------- */ -void turn_off_xlu(pc300_t * card, int ch) +static void turn_off_xlu(pc300_t * card, int ch) { pc300ch_t *chan = (pc300ch_t *) & card->chan[ch]; pc300chconf_t *conf = (pc300chconf_t *) & chan->conf; @@ -1597,7 +1598,7 @@ void turn_off_xlu(pc300_t * card, int ch) * Description: Turns XLD bit off in the proper register *---------------------------------------------------------------------------- */ -void turn_off_xld(pc300_t * card, int ch) +static void turn_off_xld(pc300_t * card, int ch) { pc300ch_t *chan = (pc300ch_t *) & card->chan[ch]; pc300chconf_t *conf = (pc300chconf_t *) & chan->conf; @@ -1619,7 +1620,7 @@ void turn_off_xld(pc300_t * card, int ch) * to generate a LOOP activation code over a T1/E1 line. *---------------------------------------------------------------------------- */ -void falc_generate_loop_up_code(pc300_t * card, int ch) +static void falc_generate_loop_up_code(pc300_t * card, int ch) { pc300ch_t *chan = (pc300ch_t *) & card->chan[ch]; pc300chconf_t *conf = (pc300chconf_t *) & chan->conf; @@ -1652,7 +1653,7 @@ void falc_generate_loop_up_code(pc300_t * card, int ch) * to generate a LOOP deactivation code over a T1/E1 line. *---------------------------------------------------------------------------- */ -void falc_generate_loop_down_code(pc300_t * card, int ch) +static void falc_generate_loop_down_code(pc300_t * card, int ch) { pc300ch_t *chan = (pc300ch_t *) & card->chan[ch]; pc300chconf_t *conf = (pc300chconf_t *) & chan->conf; @@ -1682,7 +1683,7 @@ void falc_generate_loop_down_code(pc300_t * card, int ch) * it on the reception side. *---------------------------------------------------------------------------- */ -void falc_pattern_test(pc300_t * card, int ch, unsigned int activate) +static void falc_pattern_test(pc300_t * card, int ch, unsigned int activate) { pc300ch_t *chan = (pc300ch_t *) & card->chan[ch]; pc300chconf_t *conf = (pc300chconf_t *) & chan->conf; @@ -1729,7 +1730,7 @@ void falc_pattern_test(pc300_t * card, int ch, unsigned int activate) * Description: This routine returns the bit error counter value *---------------------------------------------------------------------------- */ -ucshort falc_pattern_test_error(pc300_t * card, int ch) +static ucshort falc_pattern_test_error(pc300_t * card, int ch) { pc300ch_t *chan = (pc300ch_t *) & card->chan[ch]; falc_t *pfalc = (falc_t *) & chan->falc; @@ -1769,7 +1770,7 @@ cpc_trace(struct net_device *dev, struct sk_buff *skb_main, char rx_tx) netif_rx(skb); } -void cpc_tx_timeout(struct net_device *dev) +static void cpc_tx_timeout(struct net_device *dev) { pc300dev_t *d = (pc300dev_t *) dev->priv; pc300ch_t *chan = (pc300ch_t *) d->chan; @@ -1797,7 +1798,7 @@ void cpc_tx_timeout(struct net_device *dev) netif_wake_queue(dev); } -int cpc_queue_xmit(struct sk_buff *skb, struct net_device *dev) +static int cpc_queue_xmit(struct sk_buff *skb, struct net_device *dev) { pc300dev_t *d = (pc300dev_t *) dev->priv; pc300ch_t *chan = (pc300ch_t *) d->chan; @@ -1880,7 +1881,7 @@ int cpc_queue_xmit(struct sk_buff *skb, struct net_device *dev) return 0; } -void cpc_net_rx(struct net_device *dev) +static void cpc_net_rx(struct net_device *dev) { pc300dev_t *d = (pc300dev_t *) dev->priv; pc300ch_t *chan = (pc300ch_t *) d->chan; @@ -2403,7 +2404,7 @@ static irqreturn_t cpc_intr(int irq, void *dev_id, struct pt_regs *regs) return IRQ_HANDLED; } -void cpc_sca_status(pc300_t * card, int ch) +static void cpc_sca_status(pc300_t * card, int ch) { ucchar ilar; void __iomem *scabase = card->hw.scabase; @@ -2495,7 +2496,7 @@ void cpc_sca_status(pc300_t * card, int ch) } } -void cpc_falc_status(pc300_t * card, int ch) +static void cpc_falc_status(pc300_t * card, int ch) { pc300ch_t *chan = &card->chan[ch]; falc_t *pfalc = (falc_t *) & chan->falc; @@ -2523,7 +2524,7 @@ void cpc_falc_status(pc300_t * card, int ch) CPC_UNLOCK(card, flags); } -int cpc_change_mtu(struct net_device *dev, int new_mtu) +static int cpc_change_mtu(struct net_device *dev, int new_mtu) { if ((new_mtu < 128) || (new_mtu > PC300_DEF_MTU)) return -EINVAL; @@ -2531,7 +2532,7 @@ int cpc_change_mtu(struct net_device *dev, int new_mtu) return 0; } -int cpc_ioctl(struct net_device *dev, struct ifreq *ifr, int cmd) +static int cpc_ioctl(struct net_device *dev, struct ifreq *ifr, int cmd) { hdlc_device *hdlc = dev_to_hdlc(dev); pc300dev_t *d = (pc300dev_t *) dev->priv; @@ -2856,7 +2857,7 @@ static int clock_rate_calc(uclong rate, uclong clock, int *br_io) } } -int ch_config(pc300dev_t * d) +static int ch_config(pc300dev_t * d) { pc300ch_t *chan = (pc300ch_t *) d->chan; pc300chconf_t *conf = (pc300chconf_t *) & chan->conf; @@ -3004,7 +3005,7 @@ int ch_config(pc300dev_t * d) return 0; } -int rx_config(pc300dev_t * d) +static int rx_config(pc300dev_t * d) { pc300ch_t *chan = (pc300ch_t *) d->chan; pc300_t *card = (pc300_t *) chan->card; @@ -3035,7 +3036,7 @@ int rx_config(pc300dev_t * d) return 0; } -int tx_config(pc300dev_t * d) +static int tx_config(pc300dev_t * d) { pc300ch_t *chan = (pc300ch_t *) d->chan; pc300_t *card = (pc300_t *) chan->card; @@ -3098,7 +3099,7 @@ static int cpc_attach(struct net_device *dev, unsigned short encoding, return 0; } -void cpc_opench(pc300dev_t * d) +static void cpc_opench(pc300dev_t * d) { pc300ch_t *chan = (pc300ch_t *) d->chan; pc300_t *card = (pc300_t *) chan->card; @@ -3116,7 +3117,7 @@ void cpc_opench(pc300dev_t * d) cpc_readb(scabase + M_REG(CTL, ch)) & ~(CTL_RTS | CTL_DTR)); } -void cpc_closech(pc300dev_t * d) +static void cpc_closech(pc300dev_t * d) { pc300ch_t *chan = (pc300ch_t *) d->chan; pc300_t *card = (pc300_t *) chan->card; @@ -3173,7 +3174,7 @@ int cpc_open(struct net_device *dev) return 0; } -int cpc_close(struct net_device *dev) +static int cpc_close(struct net_device *dev) { hdlc_device *hdlc = dev_to_hdlc(dev); pc300dev_t *d = (pc300dev_t *) dev->priv; diff --git a/drivers/net/wan/pc300_tty.c b/drivers/net/wan/pc300_tty.c index 8454bf6caaa7..52f26b9c69d2 100644 --- a/drivers/net/wan/pc300_tty.c +++ b/drivers/net/wan/pc300_tty.c @@ -112,10 +112,10 @@ typedef struct _st_cpc_tty_area { static struct tty_driver serial_drv; /* local variables */ -st_cpc_tty_area cpc_tty_area[CPC_TTY_NPORTS]; +static st_cpc_tty_area cpc_tty_area[CPC_TTY_NPORTS]; -int cpc_tty_cnt=0; /* number of intrfaces configured with MLPPP */ -int cpc_tty_unreg_flag = 0; +static int cpc_tty_cnt = 0; /* number of intrfaces configured with MLPPP */ +static int cpc_tty_unreg_flag = 0; /* TTY functions prototype */ static int cpc_tty_open(struct tty_struct *tty, struct file *flip); @@ -132,9 +132,9 @@ static void cpc_tty_trace(pc300dev_t *dev, char* buf, int len, char rxtx); static void cpc_tty_signal_off(pc300dev_t *pc300dev, unsigned char); static void cpc_tty_signal_on(pc300dev_t *pc300dev, unsigned char); -int pc300_tiocmset(struct tty_struct *, struct file *, - unsigned int, unsigned int); -int pc300_tiocmget(struct tty_struct *, struct file *); +static int pc300_tiocmset(struct tty_struct *, struct file *, + unsigned int, unsigned int); +static int pc300_tiocmget(struct tty_struct *, struct file *); /* functions called by PC300 driver */ void cpc_tty_init(pc300dev_t *dev); @@ -538,8 +538,8 @@ static int cpc_tty_chars_in_buffer(struct tty_struct *tty) return(0); } -int pc300_tiocmset(struct tty_struct *tty, struct file *file, - unsigned int set, unsigned int clear) +static int pc300_tiocmset(struct tty_struct *tty, struct file *file, + unsigned int set, unsigned int clear) { st_cpc_tty_area *cpc_tty; @@ -565,7 +565,7 @@ int pc300_tiocmset(struct tty_struct *tty, struct file *file, return 0; } -int pc300_tiocmget(struct tty_struct *tty, struct file *file) +static int pc300_tiocmget(struct tty_struct *tty, struct file *file) { unsigned int result; unsigned char status; diff --git a/drivers/net/wan/sdla.c b/drivers/net/wan/sdla.c index 3ac9a45b20fa..036adc4f8ba7 100644 --- a/drivers/net/wan/sdla.c +++ b/drivers/net/wan/sdla.c @@ -182,7 +182,7 @@ static char sdla_byte(struct net_device *dev, int addr) return(byte); } -void sdla_stop(struct net_device *dev) +static void sdla_stop(struct net_device *dev) { struct frad_local *flp; @@ -209,7 +209,7 @@ void sdla_stop(struct net_device *dev) } } -void sdla_start(struct net_device *dev) +static void sdla_start(struct net_device *dev) { struct frad_local *flp; @@ -247,7 +247,7 @@ void sdla_start(struct net_device *dev) * ***************************************************/ -int sdla_z80_poll(struct net_device *dev, int z80_addr, int jiffs, char resp1, char resp2) +static int sdla_z80_poll(struct net_device *dev, int z80_addr, int jiffs, char resp1, char resp2) { unsigned long start, done, now; char resp, *temp; @@ -505,7 +505,7 @@ static int sdla_cmd(struct net_device *dev, int cmd, short dlci, short flags, static int sdla_reconfig(struct net_device *dev); -int sdla_activate(struct net_device *slave, struct net_device *master) +static int sdla_activate(struct net_device *slave, struct net_device *master) { struct frad_local *flp; int i; @@ -527,7 +527,7 @@ int sdla_activate(struct net_device *slave, struct net_device *master) return(0); } -int sdla_deactivate(struct net_device *slave, struct net_device *master) +static int sdla_deactivate(struct net_device *slave, struct net_device *master) { struct frad_local *flp; int i; @@ -549,7 +549,7 @@ int sdla_deactivate(struct net_device *slave, struct net_device *master) return(0); } -int sdla_assoc(struct net_device *slave, struct net_device *master) +static int sdla_assoc(struct net_device *slave, struct net_device *master) { struct frad_local *flp; int i; @@ -585,7 +585,7 @@ int sdla_assoc(struct net_device *slave, struct net_device *master) return(0); } -int sdla_deassoc(struct net_device *slave, struct net_device *master) +static int sdla_deassoc(struct net_device *slave, struct net_device *master) { struct frad_local *flp; int i; @@ -613,7 +613,7 @@ int sdla_deassoc(struct net_device *slave, struct net_device *master) return(0); } -int sdla_dlci_conf(struct net_device *slave, struct net_device *master, int get) +static int sdla_dlci_conf(struct net_device *slave, struct net_device *master, int get) { struct frad_local *flp; struct dlci_local *dlp; @@ -1324,7 +1324,7 @@ NOTE: This is rather a useless action right now, as the return(0); } -int sdla_change_mtu(struct net_device *dev, int new_mtu) +static int sdla_change_mtu(struct net_device *dev, int new_mtu) { struct frad_local *flp; @@ -1337,7 +1337,7 @@ int sdla_change_mtu(struct net_device *dev, int new_mtu) return(-EOPNOTSUPP); } -int sdla_set_config(struct net_device *dev, struct ifmap *map) +static int sdla_set_config(struct net_device *dev, struct ifmap *map) { struct frad_local *flp; int i; diff --git a/drivers/net/wan/sdladrv.c b/drivers/net/wan/sdladrv.c index c8bc6da57a41..7c2cf2e76300 100644 --- a/drivers/net/wan/sdladrv.c +++ b/drivers/net/wan/sdladrv.c @@ -642,9 +642,7 @@ int sdla_mapmem (sdlahw_t* hw, unsigned long addr) * Enable interrupt generation. */ -EXPORT_SYMBOL(sdla_inten); - -int sdla_inten (sdlahw_t* hw) +static int sdla_inten (sdlahw_t* hw) { unsigned port = hw->port; int tmp, i; @@ -698,8 +696,7 @@ int sdla_inten (sdlahw_t* hw) * Disable interrupt generation. */ -EXPORT_SYMBOL(sdla_intde); - +#if 0 int sdla_intde (sdlahw_t* hw) { unsigned port = hw->port; @@ -748,14 +745,13 @@ int sdla_intde (sdlahw_t* hw) } return 0; } +#endif /* 0 */ /*============================================================================ * Acknowledge SDLA hardware interrupt. */ -EXPORT_SYMBOL(sdla_intack); - -int sdla_intack (sdlahw_t* hw) +static int sdla_intack (sdlahw_t* hw) { unsigned port = hw->port; int tmp; @@ -827,8 +823,7 @@ void read_S514_int_stat (sdlahw_t* hw, u32* int_status) * Generate an interrupt to adapter's CPU. */ -EXPORT_SYMBOL(sdla_intr); - +#if 0 int sdla_intr (sdlahw_t* hw) { unsigned port = hw->port; @@ -863,6 +858,7 @@ int sdla_intr (sdlahw_t* hw) } return 0; } +#endif /* 0 */ /*============================================================================ * Execute Adapter Command. diff --git a/drivers/net/wan/syncppp.c b/drivers/net/wan/syncppp.c index b56a7b516d24..3731b22f6757 100644 --- a/drivers/net/wan/syncppp.c +++ b/drivers/net/wan/syncppp.c @@ -221,7 +221,7 @@ static void sppp_clear_timeout(struct sppp *p) * here. */ -void sppp_input (struct net_device *dev, struct sk_buff *skb) +static void sppp_input (struct net_device *dev, struct sk_buff *skb) { struct ppp_header *h; struct sppp *sp = (struct sppp *)sppp_of(dev); @@ -355,8 +355,6 @@ done: return; } -EXPORT_SYMBOL(sppp_input); - /* * Handle transmit packets. */ @@ -990,7 +988,7 @@ EXPORT_SYMBOL(sppp_reopen); * the mtu is out of range. */ -int sppp_change_mtu(struct net_device *dev, int new_mtu) +static int sppp_change_mtu(struct net_device *dev, int new_mtu) { if(new_mtu<128||new_mtu>PPP_MTU||(dev->flags&IFF_UP)) return -EINVAL; @@ -998,8 +996,6 @@ int sppp_change_mtu(struct net_device *dev, int new_mtu) return 0; } -EXPORT_SYMBOL(sppp_change_mtu); - /** * sppp_do_ioctl - Ioctl handler for ppp/hdlc * @dev: Device subject to ioctl @@ -1456,7 +1452,7 @@ static int sppp_rcv(struct sk_buff *skb, struct net_device *dev, struct packet_t return 0; } -struct packet_type sppp_packet_type = { +static struct packet_type sppp_packet_type = { .type = __constant_htons(ETH_P_WAN_PPP), .func = sppp_rcv, }; diff --git a/include/linux/cycx_drv.h b/include/linux/cycx_drv.h index 6621df86a748..12fe6b0bfcff 100644 --- a/include/linux/cycx_drv.h +++ b/include/linux/cycx_drv.h @@ -60,6 +60,5 @@ extern int cycx_peek(struct cycx_hw *hw, u32 addr, void *buf, u32 len); extern int cycx_poke(struct cycx_hw *hw, u32 addr, void *buf, u32 len); extern int cycx_exec(void __iomem *addr); -extern void cycx_inten(struct cycx_hw *hw); extern void cycx_intr(struct cycx_hw *hw); #endif /* _CYCX_DRV_H */ diff --git a/include/linux/sdladrv.h b/include/linux/sdladrv.h index 78f634007fc6..c85e103d5e7b 100644 --- a/include/linux/sdladrv.h +++ b/include/linux/sdladrv.h @@ -52,12 +52,8 @@ typedef struct sdlahw extern int sdla_setup (sdlahw_t* hw, void* sfm, unsigned len); extern int sdla_down (sdlahw_t* hw); -extern int sdla_inten (sdlahw_t* hw); -extern int sdla_intde (sdlahw_t* hw); -extern int sdla_intack (sdlahw_t* hw); extern void S514_intack (sdlahw_t* hw, u32 int_status); extern void read_S514_int_stat (sdlahw_t* hw, u32* int_status); -extern int sdla_intr (sdlahw_t* hw); extern int sdla_mapmem (sdlahw_t* hw, unsigned long addr); extern int sdla_peek (sdlahw_t* hw, unsigned long addr, void* buf, unsigned len); diff --git a/include/net/syncppp.h b/include/net/syncppp.h index 614cb6ba564e..877efa434700 100644 --- a/include/net/syncppp.h +++ b/include/net/syncppp.h @@ -86,7 +86,6 @@ static inline struct sppp *sppp_of(struct net_device *dev) void sppp_attach (struct ppp_device *pd); void sppp_detach (struct net_device *dev); -void sppp_input (struct net_device *dev, struct sk_buff *m); int sppp_do_ioctl (struct net_device *dev, struct ifreq *ifr, int cmd); struct sk_buff *sppp_dequeue (struct net_device *dev); int sppp_isempty (struct net_device *dev); -- cgit v1.2.3 From 43ec6e95e4d8a73afc2405a44b955c380aeeb65a Mon Sep 17 00:00:00 2001 From: Dale Farnsworth Date: Tue, 23 Aug 2005 10:30:29 -0700 Subject: [PATCH] mii: Add test for GigE support Signed-off-by: Dale Farnsworth Signed-off-by: Jeff Garzik --- drivers/net/mii.c | 15 +++++++++++++++ include/linux/mii.h | 1 + 2 files changed, 16 insertions(+) (limited to 'include/linux') diff --git a/drivers/net/mii.c b/drivers/net/mii.c index c33cb3dc942b..e42aa797f08b 100644 --- a/drivers/net/mii.c +++ b/drivers/net/mii.c @@ -207,6 +207,20 @@ int mii_ethtool_sset(struct mii_if_info *mii, struct ethtool_cmd *ecmd) return 0; } +int mii_check_gmii_support(struct mii_if_info *mii) +{ + int reg; + + reg = mii->mdio_read(mii->dev, mii->phy_id, MII_BMSR); + if (reg & BMSR_ESTATEN) { + reg = mii->mdio_read(mii->dev, mii->phy_id, MII_ESTATUS); + if (reg & (ESTATUS_1000_TFULL | ESTATUS_1000_THALF)) + return 1; + } + + return 0; +} + int mii_link_ok (struct mii_if_info *mii) { /* first, a dummy read, needed to latch some MII phys */ @@ -394,5 +408,6 @@ EXPORT_SYMBOL(mii_ethtool_gset); EXPORT_SYMBOL(mii_ethtool_sset); EXPORT_SYMBOL(mii_check_link); EXPORT_SYMBOL(mii_check_media); +EXPORT_SYMBOL(mii_check_gmii_support); EXPORT_SYMBOL(generic_mii_ioctl); diff --git a/include/linux/mii.h b/include/linux/mii.h index 9b8d0476988a..68f5a0f392dd 100644 --- a/include/linux/mii.h +++ b/include/linux/mii.h @@ -158,6 +158,7 @@ extern int mii_link_ok (struct mii_if_info *mii); extern int mii_nway_restart (struct mii_if_info *mii); extern int mii_ethtool_gset(struct mii_if_info *mii, struct ethtool_cmd *ecmd); extern int mii_ethtool_sset(struct mii_if_info *mii, struct ethtool_cmd *ecmd); +extern int mii_check_gmii_support(struct mii_if_info *mii); extern void mii_check_link (struct mii_if_info *mii); extern unsigned int mii_check_media (struct mii_if_info *mii, unsigned int ok_to_print, -- cgit v1.2.3 From 3905ec4561b7b049e9c2e27311d072c356dbdee2 Mon Sep 17 00:00:00 2001 From: James Ketrenos Date: Wed, 21 Sep 2005 11:56:42 -0500 Subject: [PATCH] ieee80211: Added ieee80211_radiotap.h tree 383c59b2516a61f2683f02dfebbed0caf6ee5dc3 parent a04948f63fd96c4b875a43f78afad1a0874cc441 author Mike Kershaw 1124447833 -0500 committer James Ketrenos 1127313883 -0500 Added ieee80211_radiotap.h to enhance statistic reporting to user space from wireless drivers. Signed-off-by: Mike Kershaw Signed-off-by: James Ketrenos Signed-off-by: Jeff Garzik --- include/linux/if_arp.h | 1 + include/net/ieee80211_radiotap.h | 231 +++++++++++++++++++++++++++++++++++++++ 2 files changed, 232 insertions(+) create mode 100644 include/net/ieee80211_radiotap.h (limited to 'include/linux') diff --git a/include/linux/if_arp.h b/include/linux/if_arp.h index 0856548a2a08..a8b1a2071838 100644 --- a/include/linux/if_arp.h +++ b/include/linux/if_arp.h @@ -84,6 +84,7 @@ #define ARPHRD_IEEE802_TR 800 /* Magic type ident for TR */ #define ARPHRD_IEEE80211 801 /* IEEE 802.11 */ #define ARPHRD_IEEE80211_PRISM 802 /* IEEE 802.11 + Prism2 header */ +#define ARPHRD_IEEE80211_RADIOTAP 803 /* IEEE 802.11 + radiotap header */ #define ARPHRD_VOID 0xFFFF /* Void type, nothing is known */ #define ARPHRD_NONE 0xFFFE /* zero header length */ diff --git a/include/net/ieee80211_radiotap.h b/include/net/ieee80211_radiotap.h new file mode 100644 index 000000000000..429b73892a5f --- /dev/null +++ b/include/net/ieee80211_radiotap.h @@ -0,0 +1,231 @@ +/* $FreeBSD: src/sys/net80211/ieee80211_radiotap.h,v 1.5 2005/01/22 20:12:05 sam Exp $ */ +/* $NetBSD: ieee80211_radiotap.h,v 1.11 2005/06/22 06:16:02 dyoung Exp $ */ + +/*- + * Copyright (c) 2003, 2004 David Young. All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in the + * documentation and/or other materials provided with the distribution. + * 3. The name of David Young may not be used to endorse or promote + * products derived from this software without specific prior + * written permission. + * + * THIS SOFTWARE IS PROVIDED BY DAVID YOUNG ``AS IS'' AND ANY + * EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, + * THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A + * PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL DAVID + * YOUNG BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, + * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED + * TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND + * ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, + * OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY + * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY + * OF SUCH DAMAGE. + */ + +/* + * Modifications to fit into the linux IEEE 802.11 stack, + * Mike Kershaw (dragorn@kismetwireless.net) + */ + +#ifndef IEEE80211RADIOTAP_H +#define IEEE80211RADIOTAP_H + +#include +#include + +/* Radiotap header version (from official NetBSD feed) */ +#define IEEE80211RADIOTAP_VERSION "1.5" +/* Base version of the radiotap packet header data */ +#define PKTHDR_RADIOTAP_VERSION 0 + +/* A generic radio capture format is desirable. There is one for + * Linux, but it is neither rigidly defined (there were not even + * units given for some fields) nor easily extensible. + * + * I suggest the following extensible radio capture format. It is + * based on a bitmap indicating which fields are present. + * + * I am trying to describe precisely what the application programmer + * should expect in the following, and for that reason I tell the + * units and origin of each measurement (where it applies), or else I + * use sufficiently weaselly language ("is a monotonically nondecreasing + * function of...") that I cannot set false expectations for lawyerly + * readers. + */ + +/* XXX tcpdump/libpcap do not tolerate variable-length headers, + * yet, so we pad every radiotap header to 64 bytes. Ugh. + */ +#define IEEE80211_RADIOTAP_HDRLEN 64 + +/* The radio capture header precedes the 802.11 header. */ +struct ieee80211_radiotap_header { + u8 it_version; /* Version 0. Only increases + * for drastic changes, + * introduction of compatible + * new fields does not count. + */ + u8 it_pad; + u16 it_len; /* length of the whole + * header in bytes, including + * it_version, it_pad, + * it_len, and data fields. + */ + u32 it_present; /* A bitmap telling which + * fields are present. Set bit 31 + * (0x80000000) to extend the + * bitmap by another 32 bits. + * Additional extensions are made + * by setting bit 31. + */ +}; + +/* Name Data type Units + * ---- --------- ----- + * + * IEEE80211_RADIOTAP_TSFT u64 microseconds + * + * Value in microseconds of the MAC's 64-bit 802.11 Time + * Synchronization Function timer when the first bit of the + * MPDU arrived at the MAC. For received frames, only. + * + * IEEE80211_RADIOTAP_CHANNEL 2 x u16 MHz, bitmap + * + * Tx/Rx frequency in MHz, followed by flags (see below). + * + * IEEE80211_RADIOTAP_FHSS u16 see below + * + * For frequency-hopping radios, the hop set (first byte) + * and pattern (second byte). + * + * IEEE80211_RADIOTAP_RATE u8 500kb/s + * + * Tx/Rx data rate + * + * IEEE80211_RADIOTAP_DBM_ANTSIGNAL int8_t decibels from + * one milliwatt (dBm) + * + * RF signal power at the antenna, decibel difference from + * one milliwatt. + * + * IEEE80211_RADIOTAP_DBM_ANTNOISE int8_t decibels from + * one milliwatt (dBm) + * + * RF noise power at the antenna, decibel difference from one + * milliwatt. + * + * IEEE80211_RADIOTAP_DB_ANTSIGNAL u8 decibel (dB) + * + * RF signal power at the antenna, decibel difference from an + * arbitrary, fixed reference. + * + * IEEE80211_RADIOTAP_DB_ANTNOISE u8 decibel (dB) + * + * RF noise power at the antenna, decibel difference from an + * arbitrary, fixed reference point. + * + * IEEE80211_RADIOTAP_LOCK_QUALITY u16 unitless + * + * Quality of Barker code lock. Unitless. Monotonically + * nondecreasing with "better" lock strength. Called "Signal + * Quality" in datasheets. (Is there a standard way to measure + * this?) + * + * IEEE80211_RADIOTAP_TX_ATTENUATION u16 unitless + * + * Transmit power expressed as unitless distance from max + * power set at factory calibration. 0 is max power. + * Monotonically nondecreasing with lower power levels. + * + * IEEE80211_RADIOTAP_DB_TX_ATTENUATION u16 decibels (dB) + * + * Transmit power expressed as decibel distance from max power + * set at factory calibration. 0 is max power. Monotonically + * nondecreasing with lower power levels. + * + * IEEE80211_RADIOTAP_DBM_TX_POWER int8_t decibels from + * one milliwatt (dBm) + * + * Transmit power expressed as dBm (decibels from a 1 milliwatt + * reference). This is the absolute power level measured at + * the antenna port. + * + * IEEE80211_RADIOTAP_FLAGS u8 bitmap + * + * Properties of transmitted and received frames. See flags + * defined below. + * + * IEEE80211_RADIOTAP_ANTENNA u8 antenna index + * + * Unitless indication of the Rx/Tx antenna for this packet. + * The first antenna is antenna 0. + * + * IEEE80211_RADIOTAP_FCS u32 data + * + * FCS from frame in network byte order. + */ +enum ieee80211_radiotap_type { + IEEE80211_RADIOTAP_TSFT = 0, + IEEE80211_RADIOTAP_FLAGS = 1, + IEEE80211_RADIOTAP_RATE = 2, + IEEE80211_RADIOTAP_CHANNEL = 3, + IEEE80211_RADIOTAP_FHSS = 4, + IEEE80211_RADIOTAP_DBM_ANTSIGNAL = 5, + IEEE80211_RADIOTAP_DBM_ANTNOISE = 6, + IEEE80211_RADIOTAP_LOCK_QUALITY = 7, + IEEE80211_RADIOTAP_TX_ATTENUATION = 8, + IEEE80211_RADIOTAP_DB_TX_ATTENUATION = 9, + IEEE80211_RADIOTAP_DBM_TX_POWER = 10, + IEEE80211_RADIOTAP_ANTENNA = 11, + IEEE80211_RADIOTAP_DB_ANTSIGNAL = 12, + IEEE80211_RADIOTAP_DB_ANTNOISE = 13, + IEEE80211_RADIOTAP_EXT = 31, +}; + +/* Channel flags. */ +#define IEEE80211_CHAN_TURBO 0x0010 /* Turbo channel */ +#define IEEE80211_CHAN_CCK 0x0020 /* CCK channel */ +#define IEEE80211_CHAN_OFDM 0x0040 /* OFDM channel */ +#define IEEE80211_CHAN_2GHZ 0x0080 /* 2 GHz spectrum channel. */ +#define IEEE80211_CHAN_5GHZ 0x0100 /* 5 GHz spectrum channel */ +#define IEEE80211_CHAN_PASSIVE 0x0200 /* Only passive scan allowed */ +#define IEEE80211_CHAN_DYN 0x0400 /* Dynamic CCK-OFDM channel */ +#define IEEE80211_CHAN_GFSK 0x0800 /* GFSK channel (FHSS PHY) */ + +/* For IEEE80211_RADIOTAP_FLAGS */ +#define IEEE80211_RADIOTAP_F_CFP 0x01 /* sent/received + * during CFP + */ +#define IEEE80211_RADIOTAP_F_SHORTPRE 0x02 /* sent/received + * with short + * preamble + */ +#define IEEE80211_RADIOTAP_F_WEP 0x04 /* sent/received + * with WEP encryption + */ +#define IEEE80211_RADIOTAP_F_FRAG 0x08 /* sent/received + * with fragmentation + */ +#define IEEE80211_RADIOTAP_F_FCS 0x10 /* frame includes FCS */ +#define IEEE80211_RADIOTAP_F_DATAPAD 0x20 /* frame has padding between + * 802.11 header and payload + * (to 32-bit boundary) + */ + +/* Ugly macro to convert literal channel numbers into their mhz equivalents + * There are certianly some conditions that will break this (like feeding it '30') + * but they shouldn't arise since nothing talks on channel 30. */ +#define ieee80211chan2mhz(x) \ + (((x) <= 14) ? \ + (((x) == 14) ? 2484 : ((x) * 5) + 2407) : \ + ((x) + 1000) * 5) + +#endif /* IEEE80211_RADIOTAP_H */ -- cgit v1.2.3 From 094bb20b9fcab3a1652a77741caba6b78097d622 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 11 Aug 2005 16:25:20 -0400 Subject: [PATCH] RPC: extract socket logic common to both client and server Clean-up: Move some code that is common to both RPC client- and server-side socket transports into its own source file, net/sunrpc/socklib.c. Test-plan: Compile kernel with CONFIG_NFS enabled. Millions of fsx operations over UDP, client and server. Connectathon over UDP. Version: Thu, 11 Aug 2005 16:03:09 -0400 Signed-off-by: Chuck Lever Signed-off-by: Trond Myklebust --- include/linux/sunrpc/xdr.h | 1 + net/sunrpc/Makefile | 2 +- net/sunrpc/socklib.c | 175 +++++++++++++++++++++++++++++++++++++++++++++ net/sunrpc/svcsock.c | 3 - net/sunrpc/xdr.c | 75 ------------------- net/sunrpc/xprt.c | 64 ----------------- 6 files changed, 177 insertions(+), 143 deletions(-) create mode 100644 net/sunrpc/socklib.c (limited to 'include/linux') diff --git a/include/linux/sunrpc/xdr.h b/include/linux/sunrpc/xdr.h index 23448d0fb5bc..d8b7656bca41 100644 --- a/include/linux/sunrpc/xdr.h +++ b/include/linux/sunrpc/xdr.h @@ -161,6 +161,7 @@ typedef struct { typedef size_t (*skb_read_actor_t)(skb_reader_t *desc, void *to, size_t len); +extern int csum_partial_copy_to_xdr(struct xdr_buf *, struct sk_buff *); extern ssize_t xdr_partial_copy_from_skb(struct xdr_buf *, unsigned int, skb_reader_t *, skb_read_actor_t); diff --git a/net/sunrpc/Makefile b/net/sunrpc/Makefile index 46a2ce00a29b..f0a955627177 100644 --- a/net/sunrpc/Makefile +++ b/net/sunrpc/Makefile @@ -6,7 +6,7 @@ obj-$(CONFIG_SUNRPC) += sunrpc.o obj-$(CONFIG_SUNRPC_GSS) += auth_gss/ -sunrpc-y := clnt.o xprt.o sched.o \ +sunrpc-y := clnt.o xprt.o socklib.o sched.o \ auth.o auth_null.o auth_unix.o \ svc.o svcsock.o svcauth.o svcauth_unix.o \ pmap_clnt.o timer.o xdr.o \ diff --git a/net/sunrpc/socklib.c b/net/sunrpc/socklib.c new file mode 100644 index 000000000000..8f97e90f36c8 --- /dev/null +++ b/net/sunrpc/socklib.c @@ -0,0 +1,175 @@ +/* + * linux/net/sunrpc/socklib.c + * + * Common socket helper routines for RPC client and server + * + * Copyright (C) 1995, 1996 Olaf Kirch + */ + +#include +#include +#include +#include + + +/** + * skb_read_bits - copy some data bits from skb to internal buffer + * @desc: sk_buff copy helper + * @to: copy destination + * @len: number of bytes to copy + * + * Possibly called several times to iterate over an sk_buff and copy + * data out of it. + */ +static size_t skb_read_bits(skb_reader_t *desc, void *to, size_t len) +{ + if (len > desc->count) + len = desc->count; + if (skb_copy_bits(desc->skb, desc->offset, to, len)) + return 0; + desc->count -= len; + desc->offset += len; + return len; +} + +/** + * skb_read_and_csum_bits - copy and checksum from skb to buffer + * @desc: sk_buff copy helper + * @to: copy destination + * @len: number of bytes to copy + * + * Same as skb_read_bits, but calculate a checksum at the same time. + */ +static size_t skb_read_and_csum_bits(skb_reader_t *desc, void *to, size_t len) +{ + unsigned int csum2, pos; + + if (len > desc->count) + len = desc->count; + pos = desc->offset; + csum2 = skb_copy_and_csum_bits(desc->skb, pos, to, len, 0); + desc->csum = csum_block_add(desc->csum, csum2, pos); + desc->count -= len; + desc->offset += len; + return len; +} + +/** + * xdr_partial_copy_from_skb - copy data out of an skb + * @xdr: target XDR buffer + * @base: starting offset + * @desc: sk_buff copy helper + * @copy_actor: virtual method for copying data + * + */ +ssize_t xdr_partial_copy_from_skb(struct xdr_buf *xdr, unsigned int base, skb_reader_t *desc, skb_read_actor_t copy_actor) +{ + struct page **ppage = xdr->pages; + unsigned int len, pglen = xdr->page_len; + ssize_t copied = 0; + int ret; + + len = xdr->head[0].iov_len; + if (base < len) { + len -= base; + ret = copy_actor(desc, (char *)xdr->head[0].iov_base + base, len); + copied += ret; + if (ret != len || !desc->count) + goto out; + base = 0; + } else + base -= len; + + if (unlikely(pglen == 0)) + goto copy_tail; + if (unlikely(base >= pglen)) { + base -= pglen; + goto copy_tail; + } + if (base || xdr->page_base) { + pglen -= base; + base += xdr->page_base; + ppage += base >> PAGE_CACHE_SHIFT; + base &= ~PAGE_CACHE_MASK; + } + do { + char *kaddr; + + /* ACL likes to be lazy in allocating pages - ACLs + * are small by default but can get huge. */ + if (unlikely(*ppage == NULL)) { + *ppage = alloc_page(GFP_ATOMIC); + if (unlikely(*ppage == NULL)) { + if (copied == 0) + copied = -ENOMEM; + goto out; + } + } + + len = PAGE_CACHE_SIZE; + kaddr = kmap_atomic(*ppage, KM_SKB_SUNRPC_DATA); + if (base) { + len -= base; + if (pglen < len) + len = pglen; + ret = copy_actor(desc, kaddr + base, len); + base = 0; + } else { + if (pglen < len) + len = pglen; + ret = copy_actor(desc, kaddr, len); + } + flush_dcache_page(*ppage); + kunmap_atomic(kaddr, KM_SKB_SUNRPC_DATA); + copied += ret; + if (ret != len || !desc->count) + goto out; + ppage++; + } while ((pglen -= len) != 0); +copy_tail: + len = xdr->tail[0].iov_len; + if (base < len) + copied += copy_actor(desc, (char *)xdr->tail[0].iov_base + base, len - base); +out: + return copied; +} + +/** + * csum_partial_copy_to_xdr - checksum and copy data + * @xdr: target XDR buffer + * @skb: source skb + * + * We have set things up such that we perform the checksum of the UDP + * packet in parallel with the copies into the RPC client iovec. -DaveM + */ +int csum_partial_copy_to_xdr(struct xdr_buf *xdr, struct sk_buff *skb) +{ + skb_reader_t desc; + + desc.skb = skb; + desc.offset = sizeof(struct udphdr); + desc.count = skb->len - desc.offset; + + if (skb->ip_summed == CHECKSUM_UNNECESSARY) + goto no_checksum; + + desc.csum = csum_partial(skb->data, desc.offset, skb->csum); + if (xdr_partial_copy_from_skb(xdr, 0, &desc, skb_read_and_csum_bits) < 0) + return -1; + if (desc.offset != skb->len) { + unsigned int csum2; + csum2 = skb_checksum(skb, desc.offset, skb->len - desc.offset, 0); + desc.csum = csum_block_add(desc.csum, csum2, desc.offset); + } + if (desc.count) + return -1; + if ((unsigned short)csum_fold(desc.csum)) + return -1; + return 0; +no_checksum: + if (xdr_partial_copy_from_skb(xdr, 0, &desc, skb_read_bits) < 0) + return -1; + if (desc.count) + return -1; + return 0; +} diff --git a/net/sunrpc/svcsock.c b/net/sunrpc/svcsock.c index 30ec3efc48a6..130f2b5d93dd 100644 --- a/net/sunrpc/svcsock.c +++ b/net/sunrpc/svcsock.c @@ -548,9 +548,6 @@ svc_write_space(struct sock *sk) /* * Receive a datagram from a UDP socket. */ -extern int -csum_partial_copy_to_xdr(struct xdr_buf *xdr, struct sk_buff *skb); - static int svc_udp_recvfrom(struct svc_rqst *rqstp) { diff --git a/net/sunrpc/xdr.c b/net/sunrpc/xdr.c index fde16f40a581..9cc12aeed22c 100644 --- a/net/sunrpc/xdr.c +++ b/net/sunrpc/xdr.c @@ -176,81 +176,6 @@ xdr_inline_pages(struct xdr_buf *xdr, unsigned int offset, xdr->buflen += len; } -ssize_t -xdr_partial_copy_from_skb(struct xdr_buf *xdr, unsigned int base, - skb_reader_t *desc, - skb_read_actor_t copy_actor) -{ - struct page **ppage = xdr->pages; - unsigned int len, pglen = xdr->page_len; - ssize_t copied = 0; - int ret; - - len = xdr->head[0].iov_len; - if (base < len) { - len -= base; - ret = copy_actor(desc, (char *)xdr->head[0].iov_base + base, len); - copied += ret; - if (ret != len || !desc->count) - goto out; - base = 0; - } else - base -= len; - - if (pglen == 0) - goto copy_tail; - if (base >= pglen) { - base -= pglen; - goto copy_tail; - } - if (base || xdr->page_base) { - pglen -= base; - base += xdr->page_base; - ppage += base >> PAGE_CACHE_SHIFT; - base &= ~PAGE_CACHE_MASK; - } - do { - char *kaddr; - - /* ACL likes to be lazy in allocating pages - ACLs - * are small by default but can get huge. */ - if (unlikely(*ppage == NULL)) { - *ppage = alloc_page(GFP_ATOMIC); - if (unlikely(*ppage == NULL)) { - if (copied == 0) - copied = -ENOMEM; - goto out; - } - } - - len = PAGE_CACHE_SIZE; - kaddr = kmap_atomic(*ppage, KM_SKB_SUNRPC_DATA); - if (base) { - len -= base; - if (pglen < len) - len = pglen; - ret = copy_actor(desc, kaddr + base, len); - base = 0; - } else { - if (pglen < len) - len = pglen; - ret = copy_actor(desc, kaddr, len); - } - flush_dcache_page(*ppage); - kunmap_atomic(kaddr, KM_SKB_SUNRPC_DATA); - copied += ret; - if (ret != len || !desc->count) - goto out; - ppage++; - } while ((pglen -= len) != 0); -copy_tail: - len = xdr->tail[0].iov_len; - if (base < len) - copied += copy_actor(desc, (char *)xdr->tail[0].iov_base + base, len - base); -out: - return copied; -} - int xdr_sendpages(struct socket *sock, struct sockaddr *addr, int addrlen, diff --git a/net/sunrpc/xprt.c b/net/sunrpc/xprt.c index 0e4ffdaa0129..67444f494fea 100644 --- a/net/sunrpc/xprt.c +++ b/net/sunrpc/xprt.c @@ -691,70 +691,6 @@ xprt_complete_rqst(struct rpc_xprt *xprt, struct rpc_rqst *req, int copied) return; } -static size_t -skb_read_bits(skb_reader_t *desc, void *to, size_t len) -{ - if (len > desc->count) - len = desc->count; - if (skb_copy_bits(desc->skb, desc->offset, to, len)) - return 0; - desc->count -= len; - desc->offset += len; - return len; -} - -static size_t -skb_read_and_csum_bits(skb_reader_t *desc, void *to, size_t len) -{ - unsigned int csum2, pos; - - if (len > desc->count) - len = desc->count; - pos = desc->offset; - csum2 = skb_copy_and_csum_bits(desc->skb, pos, to, len, 0); - desc->csum = csum_block_add(desc->csum, csum2, pos); - desc->count -= len; - desc->offset += len; - return len; -} - -/* - * We have set things up such that we perform the checksum of the UDP - * packet in parallel with the copies into the RPC client iovec. -DaveM - */ -int -csum_partial_copy_to_xdr(struct xdr_buf *xdr, struct sk_buff *skb) -{ - skb_reader_t desc; - - desc.skb = skb; - desc.offset = sizeof(struct udphdr); - desc.count = skb->len - desc.offset; - - if (skb->ip_summed == CHECKSUM_UNNECESSARY) - goto no_checksum; - - desc.csum = csum_partial(skb->data, desc.offset, skb->csum); - if (xdr_partial_copy_from_skb(xdr, 0, &desc, skb_read_and_csum_bits) < 0) - return -1; - if (desc.offset != skb->len) { - unsigned int csum2; - csum2 = skb_checksum(skb, desc.offset, skb->len - desc.offset, 0); - desc.csum = csum_block_add(desc.csum, csum2, desc.offset); - } - if (desc.count) - return -1; - if ((unsigned short)csum_fold(desc.csum)) - return -1; - return 0; -no_checksum: - if (xdr_partial_copy_from_skb(xdr, 0, &desc, skb_read_bits) < 0) - return -1; - if (desc.count) - return -1; - return 0; -} - /* * Input handler for RPC replies. Called from a bottom half and hence * atomic. -- cgit v1.2.3 From a246b0105bbd9a70a698f69baae2042996f2a0e9 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 11 Aug 2005 16:25:23 -0400 Subject: [PATCH] RPC: introduce client-side transport switch Move the bulk of client-side socket-specific code into a separate source file, net/sunrpc/xprtsock.c. Test-plan: Millions of fsx operations. Performance characterization such as "sio" or "iozone". Destructive testing (unplugging the network temporarily, server reboots). Connectathon with v2, v3, and v4. Version: Thu, 11 Aug 2005 16:03:38 -0400 Signed-off-by: Chuck Lever Signed-off-by: Trond Myklebust --- include/linux/sunrpc/xdr.h | 5 - include/linux/sunrpc/xprt.h | 38 +- net/sunrpc/Makefile | 2 +- net/sunrpc/clnt.c | 3 +- net/sunrpc/sysctl.c | 3 + net/sunrpc/xdr.c | 102 +---- net/sunrpc/xprt.c | 916 ++------------------------------------ net/sunrpc/xprtsock.c | 1021 +++++++++++++++++++++++++++++++++++++++++++ 8 files changed, 1101 insertions(+), 989 deletions(-) create mode 100644 net/sunrpc/xprtsock.c (limited to 'include/linux') diff --git a/include/linux/sunrpc/xdr.h b/include/linux/sunrpc/xdr.h index d8b7656bca41..5da968729cf8 100644 --- a/include/linux/sunrpc/xdr.h +++ b/include/linux/sunrpc/xdr.h @@ -165,11 +165,6 @@ extern int csum_partial_copy_to_xdr(struct xdr_buf *, struct sk_buff *); extern ssize_t xdr_partial_copy_from_skb(struct xdr_buf *, unsigned int, skb_reader_t *, skb_read_actor_t); -struct socket; -struct sockaddr; -extern int xdr_sendpages(struct socket *, struct sockaddr *, int, - struct xdr_buf *, unsigned int, int); - extern int xdr_encode_word(struct xdr_buf *, int, u32); extern int xdr_decode_word(struct xdr_buf *, int, u32 *); diff --git a/include/linux/sunrpc/xprt.h b/include/linux/sunrpc/xprt.h index e618c1649814..d82b47ab73cb 100644 --- a/include/linux/sunrpc/xprt.h +++ b/include/linux/sunrpc/xprt.h @@ -59,7 +59,13 @@ extern unsigned int xprt_tcp_slot_table_entries; */ #define RPC_REESTABLISH_TIMEOUT (15*HZ) -/* RPC call and reply header size as number of 32bit words (verifier +/* + * RPC transport idle timeout. + */ +#define RPC_IDLE_DISCONNECT_TIMEOUT (5*60*HZ) + +/* + * RPC call and reply header size as number of 32bit words (verifier * size computed separately) */ #define RPC_CALLHDRSIZE 6 @@ -121,12 +127,19 @@ struct rpc_rqst { #define rq_svec rq_snd_buf.head #define rq_slen rq_snd_buf.len -#define XPRT_LAST_FRAG (1 << 0) -#define XPRT_COPY_RECM (1 << 1) -#define XPRT_COPY_XID (1 << 2) -#define XPRT_COPY_DATA (1 << 3) +struct rpc_task; +struct rpc_xprt; + +struct rpc_xprt_ops { + void (*set_buffer_size)(struct rpc_xprt *xprt); + void (*connect)(struct rpc_task *task); + int (*send_request)(struct rpc_task *task); + void (*close)(struct rpc_xprt *xprt); + void (*destroy)(struct rpc_xprt *xprt); +}; struct rpc_xprt { + struct rpc_xprt_ops * ops; /* transport methods */ struct socket * sock; /* BSD socket layer */ struct sock * inet; /* INET layer */ @@ -199,14 +212,22 @@ struct rpc_xprt { wait_queue_head_t cong_wait; }; +#define XPRT_LAST_FRAG (1 << 0) +#define XPRT_COPY_RECM (1 << 1) +#define XPRT_COPY_XID (1 << 2) +#define XPRT_COPY_DATA (1 << 3) + #ifdef __KERNEL__ struct rpc_xprt * xprt_create_proto(int proto, struct sockaddr_in *addr, struct rpc_timeout *toparms); +void xprt_disconnect(struct rpc_xprt *); int xprt_destroy(struct rpc_xprt *); void xprt_set_timeout(struct rpc_timeout *, unsigned int, unsigned long); - +struct rpc_rqst * xprt_lookup_rqst(struct rpc_xprt *, u32); +void xprt_complete_rqst(struct rpc_xprt *, + struct rpc_rqst *, int); void xprt_reserve(struct rpc_task *); int xprt_prepare_transmit(struct rpc_task *); void xprt_transmit(struct rpc_task *); @@ -214,7 +235,10 @@ void xprt_receive(struct rpc_task *); int xprt_adjust_timeout(struct rpc_rqst *req); void xprt_release(struct rpc_task *); void xprt_connect(struct rpc_task *); -void xprt_sock_setbufsize(struct rpc_xprt *); +int xs_setup_udp(struct rpc_xprt *, + struct rpc_timeout *); +int xs_setup_tcp(struct rpc_xprt *, + struct rpc_timeout *); #define XPRT_LOCKED 0 #define XPRT_CONNECT 1 diff --git a/net/sunrpc/Makefile b/net/sunrpc/Makefile index f0a955627177..cdcab9ca4c60 100644 --- a/net/sunrpc/Makefile +++ b/net/sunrpc/Makefile @@ -6,7 +6,7 @@ obj-$(CONFIG_SUNRPC) += sunrpc.o obj-$(CONFIG_SUNRPC_GSS) += auth_gss/ -sunrpc-y := clnt.o xprt.o socklib.o sched.o \ +sunrpc-y := clnt.o xprt.o socklib.o xprtsock.o sched.o \ auth.o auth_null.o auth_unix.o \ svc.o svcsock.o svcauth.o svcauth_unix.o \ pmap_clnt.o timer.o xdr.o \ diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c index 2d3cf0a52d82..ab50c3c9e6a8 100644 --- a/net/sunrpc/clnt.c +++ b/net/sunrpc/clnt.c @@ -525,8 +525,7 @@ rpc_setbufsize(struct rpc_clnt *clnt, unsigned int sndsize, unsigned int rcvsize xprt->rcvsize = 0; if (rcvsize) xprt->rcvsize = rcvsize + RPC_SLACK_SPACE; - if (xprt_connected(xprt)) - xprt_sock_setbufsize(xprt); + xprt->ops->set_buffer_size(xprt); } /* diff --git a/net/sunrpc/sysctl.c b/net/sunrpc/sysctl.c index 1b9616a12e24..ef483262f17f 100644 --- a/net/sunrpc/sysctl.c +++ b/net/sunrpc/sysctl.c @@ -119,6 +119,9 @@ done: return 0; } +unsigned int xprt_udp_slot_table_entries = RPC_DEF_SLOT_TABLE; +unsigned int xprt_tcp_slot_table_entries = RPC_DEF_SLOT_TABLE; + static unsigned int min_slot_table_size = RPC_MIN_SLOT_TABLE; static unsigned int max_slot_table_size = RPC_MAX_SLOT_TABLE; diff --git a/net/sunrpc/xdr.c b/net/sunrpc/xdr.c index 9cc12aeed22c..32df43372ee9 100644 --- a/net/sunrpc/xdr.c +++ b/net/sunrpc/xdr.c @@ -6,15 +6,12 @@ * Copyright (C) 1995, 1996 Olaf Kirch */ +#include #include -#include #include #include #include #include -#include -#include -#include #include #include @@ -177,103 +174,6 @@ xdr_inline_pages(struct xdr_buf *xdr, unsigned int offset, } -int -xdr_sendpages(struct socket *sock, struct sockaddr *addr, int addrlen, - struct xdr_buf *xdr, unsigned int base, int msgflags) -{ - struct page **ppage = xdr->pages; - unsigned int len, pglen = xdr->page_len; - int err, ret = 0; - ssize_t (*sendpage)(struct socket *, struct page *, int, size_t, int); - - len = xdr->head[0].iov_len; - if (base < len || (addr != NULL && base == 0)) { - struct kvec iov = { - .iov_base = xdr->head[0].iov_base + base, - .iov_len = len - base, - }; - struct msghdr msg = { - .msg_name = addr, - .msg_namelen = addrlen, - .msg_flags = msgflags, - }; - if (xdr->len > len) - msg.msg_flags |= MSG_MORE; - - if (iov.iov_len != 0) - err = kernel_sendmsg(sock, &msg, &iov, 1, iov.iov_len); - else - err = kernel_sendmsg(sock, &msg, NULL, 0, 0); - if (ret == 0) - ret = err; - else if (err > 0) - ret += err; - if (err != iov.iov_len) - goto out; - base = 0; - } else - base -= len; - - if (pglen == 0) - goto copy_tail; - if (base >= pglen) { - base -= pglen; - goto copy_tail; - } - if (base || xdr->page_base) { - pglen -= base; - base += xdr->page_base; - ppage += base >> PAGE_CACHE_SHIFT; - base &= ~PAGE_CACHE_MASK; - } - - sendpage = sock->ops->sendpage ? : sock_no_sendpage; - do { - int flags = msgflags; - - len = PAGE_CACHE_SIZE; - if (base) - len -= base; - if (pglen < len) - len = pglen; - - if (pglen != len || xdr->tail[0].iov_len != 0) - flags |= MSG_MORE; - - /* Hmm... We might be dealing with highmem pages */ - if (PageHighMem(*ppage)) - sendpage = sock_no_sendpage; - err = sendpage(sock, *ppage, base, len, flags); - if (ret == 0) - ret = err; - else if (err > 0) - ret += err; - if (err != len) - goto out; - base = 0; - ppage++; - } while ((pglen -= len) != 0); -copy_tail: - len = xdr->tail[0].iov_len; - if (base < len) { - struct kvec iov = { - .iov_base = xdr->tail[0].iov_base + base, - .iov_len = len - base, - }; - struct msghdr msg = { - .msg_flags = msgflags, - }; - err = kernel_sendmsg(sock, &msg, &iov, 1, iov.iov_len); - if (ret == 0) - ret = err; - else if (err > 0) - ret += err; - } -out: - return ret; -} - - /* * Helper routines for doing 'memmove' like operations on a struct xdr_buf * diff --git a/net/sunrpc/xprt.c b/net/sunrpc/xprt.c index 67444f494fea..4342acf4d1cd 100644 --- a/net/sunrpc/xprt.c +++ b/net/sunrpc/xprt.c @@ -32,37 +32,16 @@ * tasks that rely on callbacks. * * Copyright (C) 1995-1997, Olaf Kirch - * - * TCP callback races fixes (C) 1998 Red Hat Software - * TCP send fixes (C) 1998 Red Hat Software - * TCP NFS related read + write fixes - * (C) 1999 Dave Airlie, University of Limerick, Ireland - * - * Rewrite of larges part of the code in order to stabilize TCP stuff. - * Fix behaviour when socket buffer is full. - * (C) 1999 Trond Myklebust */ +#include + #include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include +#include #include #include -#include -#include -#include -#include +#include /* * Local variables @@ -74,64 +53,17 @@ #endif #define XPRT_MAX_BACKOFF (8) -#define XPRT_IDLE_TIMEOUT (5*60*HZ) -#define XPRT_MAX_RESVPORT (800) /* * Local functions */ static void xprt_request_init(struct rpc_task *, struct rpc_xprt *); static inline void do_xprt_reserve(struct rpc_task *); -static void xprt_disconnect(struct rpc_xprt *); static void xprt_connect_status(struct rpc_task *task); -static struct rpc_xprt * xprt_setup(int proto, struct sockaddr_in *ap, - struct rpc_timeout *to); -static struct socket *xprt_create_socket(struct rpc_xprt *, int, int); -static void xprt_bind_socket(struct rpc_xprt *, struct socket *); static int __xprt_get_cong(struct rpc_xprt *, struct rpc_task *); static int xprt_clear_backlog(struct rpc_xprt *xprt); -#ifdef RPC_DEBUG_DATA -/* - * Print the buffer contents (first 128 bytes only--just enough for - * diropres return). - */ -static void -xprt_pktdump(char *msg, u32 *packet, unsigned int count) -{ - u8 *buf = (u8 *) packet; - int j; - - dprintk("RPC: %s\n", msg); - for (j = 0; j < count && j < 128; j += 4) { - if (!(j & 31)) { - if (j) - dprintk("\n"); - dprintk("0x%04x ", j); - } - dprintk("%02x%02x%02x%02x ", - buf[j], buf[j+1], buf[j+2], buf[j+3]); - } - dprintk("\n"); -} -#else -static inline void -xprt_pktdump(char *msg, u32 *packet, unsigned int count) -{ - /* NOP */ -} -#endif - -/* - * Look up RPC transport given an INET socket - */ -static inline struct rpc_xprt * -xprt_from_sock(struct sock *sk) -{ - return (struct rpc_xprt *) sk->sk_user_data; -} - /* * Serialize write access to sockets, in order to prevent different * requests from interfering with each other. @@ -234,62 +166,6 @@ xprt_release_write(struct rpc_xprt *xprt, struct rpc_task *task) spin_unlock_bh(&xprt->sock_lock); } -/* - * Write data to socket. - */ -static inline int -xprt_sendmsg(struct rpc_xprt *xprt, struct rpc_rqst *req) -{ - struct socket *sock = xprt->sock; - struct xdr_buf *xdr = &req->rq_snd_buf; - struct sockaddr *addr = NULL; - int addrlen = 0; - unsigned int skip; - int result; - - if (!sock) - return -ENOTCONN; - - xprt_pktdump("packet data:", - req->rq_svec->iov_base, - req->rq_svec->iov_len); - - /* For UDP, we need to provide an address */ - if (!xprt->stream) { - addr = (struct sockaddr *) &xprt->addr; - addrlen = sizeof(xprt->addr); - } - /* Dont repeat bytes */ - skip = req->rq_bytes_sent; - - clear_bit(SOCK_ASYNC_NOSPACE, &sock->flags); - result = xdr_sendpages(sock, addr, addrlen, xdr, skip, MSG_DONTWAIT); - - dprintk("RPC: xprt_sendmsg(%d) = %d\n", xdr->len - skip, result); - - if (result >= 0) - return result; - - switch (result) { - case -ECONNREFUSED: - /* When the server has died, an ICMP port unreachable message - * prompts ECONNREFUSED. - */ - case -EAGAIN: - break; - case -ECONNRESET: - case -ENOTCONN: - case -EPIPE: - /* connection broken */ - if (xprt->stream) - result = -ENOTCONN; - break; - default: - printk(KERN_NOTICE "RPC: sendmsg returned error %d\n", -result); - } - return result; -} - /* * Van Jacobson congestion avoidance. Check if the congestion window * overflowed. Put the task to sleep if this is the case. @@ -405,48 +281,20 @@ int xprt_adjust_timeout(struct rpc_rqst *req) return status; } -/* - * Close down a transport socket - */ -static void -xprt_close(struct rpc_xprt *xprt) -{ - struct socket *sock = xprt->sock; - struct sock *sk = xprt->inet; - - if (!sk) - return; - - write_lock_bh(&sk->sk_callback_lock); - xprt->inet = NULL; - xprt->sock = NULL; - - sk->sk_user_data = NULL; - sk->sk_data_ready = xprt->old_data_ready; - sk->sk_state_change = xprt->old_state_change; - sk->sk_write_space = xprt->old_write_space; - write_unlock_bh(&sk->sk_callback_lock); - - sk->sk_no_check = 0; - - sock_release(sock); -} - static void xprt_socket_autoclose(void *args) { struct rpc_xprt *xprt = (struct rpc_xprt *)args; xprt_disconnect(xprt); - xprt_close(xprt); + xprt->ops->close(xprt); xprt_release_write(xprt, NULL); } /* * Mark a transport as disconnected */ -static void -xprt_disconnect(struct rpc_xprt *xprt) +void xprt_disconnect(struct rpc_xprt *xprt) { dprintk("RPC: disconnected transport %p\n", xprt); spin_lock_bh(&xprt->sock_lock); @@ -479,57 +327,6 @@ out_abort: spin_unlock(&xprt->sock_lock); } -static void xprt_socket_connect(void *args) -{ - struct rpc_xprt *xprt = (struct rpc_xprt *)args; - struct socket *sock = xprt->sock; - int status = -EIO; - - if (xprt->shutdown || xprt->addr.sin_port == 0) - goto out; - - /* - * Start by resetting any existing state - */ - xprt_close(xprt); - sock = xprt_create_socket(xprt, xprt->prot, xprt->resvport); - if (sock == NULL) { - /* couldn't create socket or bind to reserved port; - * this is likely a permanent error, so cause an abort */ - goto out; - } - xprt_bind_socket(xprt, sock); - xprt_sock_setbufsize(xprt); - - status = 0; - if (!xprt->stream) - goto out; - - /* - * Tell the socket layer to start connecting... - */ - status = sock->ops->connect(sock, (struct sockaddr *) &xprt->addr, - sizeof(xprt->addr), O_NONBLOCK); - dprintk("RPC: %p connect status %d connected %d sock state %d\n", - xprt, -status, xprt_connected(xprt), sock->sk->sk_state); - if (status < 0) { - switch (status) { - case -EINPROGRESS: - case -EALREADY: - goto out_clear; - } - } -out: - if (status < 0) - rpc_wake_up_status(&xprt->pending, status); - else - rpc_wake_up(&xprt->pending); -out_clear: - smp_mb__before_clear_bit(); - clear_bit(XPRT_CONNECTING, &xprt->sockstate); - smp_mb__after_clear_bit(); -} - /* * Attempt to connect a TCP socket. * @@ -552,30 +349,16 @@ void xprt_connect(struct rpc_task *task) if (!xprt_lock_write(xprt, task)) return; if (xprt_connected(xprt)) - goto out_write; + xprt_release_write(xprt, task); + else { + if (task->tk_rqstp) + task->tk_rqstp->rq_bytes_sent = 0; - if (task->tk_rqstp) - task->tk_rqstp->rq_bytes_sent = 0; - - task->tk_timeout = RPC_CONNECT_TIMEOUT; - rpc_sleep_on(&xprt->pending, task, xprt_connect_status, NULL); - if (!test_and_set_bit(XPRT_CONNECTING, &xprt->sockstate)) { - /* Note: if we are here due to a dropped connection - * we delay reconnecting by RPC_REESTABLISH_TIMEOUT/HZ - * seconds - */ - if (xprt->sock != NULL) - schedule_delayed_work(&xprt->sock_connect, - RPC_REESTABLISH_TIMEOUT); - else { - schedule_work(&xprt->sock_connect); - if (!RPC_IS_ASYNC(task)) - flush_scheduled_work(); - } + task->tk_timeout = RPC_CONNECT_TIMEOUT; + rpc_sleep_on(&xprt->pending, task, xprt_connect_status, NULL); + xprt->ops->connect(task); } return; - out_write: - xprt_release_write(xprt, task); } /* @@ -624,8 +407,7 @@ xprt_connect_status(struct rpc_task *task) /* * Look up the RPC request corresponding to a reply, and then lock it. */ -static inline struct rpc_rqst * -xprt_lookup_rqst(struct rpc_xprt *xprt, u32 xid) +struct rpc_rqst *xprt_lookup_rqst(struct rpc_xprt *xprt, u32 xid) { struct list_head *pos; struct rpc_rqst *req = NULL; @@ -644,8 +426,7 @@ xprt_lookup_rqst(struct rpc_xprt *xprt, u32 xid) * Complete reply received. * The TCP code relies on us to remove the request from xprt->pending. */ -static void -xprt_complete_rqst(struct rpc_xprt *xprt, struct rpc_rqst *req, int copied) +void xprt_complete_rqst(struct rpc_xprt *xprt, struct rpc_rqst *req, int copied) { struct rpc_task *task = req->rq_task; struct rpc_clnt *clnt = task->tk_client; @@ -691,409 +472,6 @@ xprt_complete_rqst(struct rpc_xprt *xprt, struct rpc_rqst *req, int copied) return; } -/* - * Input handler for RPC replies. Called from a bottom half and hence - * atomic. - */ -static void -udp_data_ready(struct sock *sk, int len) -{ - struct rpc_task *task; - struct rpc_xprt *xprt; - struct rpc_rqst *rovr; - struct sk_buff *skb; - int err, repsize, copied; - u32 _xid, *xp; - - read_lock(&sk->sk_callback_lock); - dprintk("RPC: udp_data_ready...\n"); - if (!(xprt = xprt_from_sock(sk))) { - printk("RPC: udp_data_ready request not found!\n"); - goto out; - } - - dprintk("RPC: udp_data_ready client %p\n", xprt); - - if ((skb = skb_recv_datagram(sk, 0, 1, &err)) == NULL) - goto out; - - if (xprt->shutdown) - goto dropit; - - repsize = skb->len - sizeof(struct udphdr); - if (repsize < 4) { - printk("RPC: impossible RPC reply size %d!\n", repsize); - goto dropit; - } - - /* Copy the XID from the skb... */ - xp = skb_header_pointer(skb, sizeof(struct udphdr), - sizeof(_xid), &_xid); - if (xp == NULL) - goto dropit; - - /* Look up and lock the request corresponding to the given XID */ - spin_lock(&xprt->sock_lock); - rovr = xprt_lookup_rqst(xprt, *xp); - if (!rovr) - goto out_unlock; - task = rovr->rq_task; - - dprintk("RPC: %4d received reply\n", task->tk_pid); - - if ((copied = rovr->rq_private_buf.buflen) > repsize) - copied = repsize; - - /* Suck it into the iovec, verify checksum if not done by hw. */ - if (csum_partial_copy_to_xdr(&rovr->rq_private_buf, skb)) - goto out_unlock; - - /* Something worked... */ - dst_confirm(skb->dst); - - xprt_complete_rqst(xprt, rovr, copied); - - out_unlock: - spin_unlock(&xprt->sock_lock); - dropit: - skb_free_datagram(sk, skb); - out: - read_unlock(&sk->sk_callback_lock); -} - -/* - * Copy from an skb into memory and shrink the skb. - */ -static inline size_t -tcp_copy_data(skb_reader_t *desc, void *p, size_t len) -{ - if (len > desc->count) - len = desc->count; - if (skb_copy_bits(desc->skb, desc->offset, p, len)) { - dprintk("RPC: failed to copy %zu bytes from skb. %zu bytes remain\n", - len, desc->count); - return 0; - } - desc->offset += len; - desc->count -= len; - dprintk("RPC: copied %zu bytes from skb. %zu bytes remain\n", - len, desc->count); - return len; -} - -/* - * TCP read fragment marker - */ -static inline void -tcp_read_fraghdr(struct rpc_xprt *xprt, skb_reader_t *desc) -{ - size_t len, used; - char *p; - - p = ((char *) &xprt->tcp_recm) + xprt->tcp_offset; - len = sizeof(xprt->tcp_recm) - xprt->tcp_offset; - used = tcp_copy_data(desc, p, len); - xprt->tcp_offset += used; - if (used != len) - return; - xprt->tcp_reclen = ntohl(xprt->tcp_recm); - if (xprt->tcp_reclen & 0x80000000) - xprt->tcp_flags |= XPRT_LAST_FRAG; - else - xprt->tcp_flags &= ~XPRT_LAST_FRAG; - xprt->tcp_reclen &= 0x7fffffff; - xprt->tcp_flags &= ~XPRT_COPY_RECM; - xprt->tcp_offset = 0; - /* Sanity check of the record length */ - if (xprt->tcp_reclen < 4) { - printk(KERN_ERR "RPC: Invalid TCP record fragment length\n"); - xprt_disconnect(xprt); - } - dprintk("RPC: reading TCP record fragment of length %d\n", - xprt->tcp_reclen); -} - -static void -tcp_check_recm(struct rpc_xprt *xprt) -{ - dprintk("RPC: xprt = %p, tcp_copied = %lu, tcp_offset = %u, tcp_reclen = %u, tcp_flags = %lx\n", - xprt, xprt->tcp_copied, xprt->tcp_offset, xprt->tcp_reclen, xprt->tcp_flags); - if (xprt->tcp_offset == xprt->tcp_reclen) { - xprt->tcp_flags |= XPRT_COPY_RECM; - xprt->tcp_offset = 0; - if (xprt->tcp_flags & XPRT_LAST_FRAG) { - xprt->tcp_flags &= ~XPRT_COPY_DATA; - xprt->tcp_flags |= XPRT_COPY_XID; - xprt->tcp_copied = 0; - } - } -} - -/* - * TCP read xid - */ -static inline void -tcp_read_xid(struct rpc_xprt *xprt, skb_reader_t *desc) -{ - size_t len, used; - char *p; - - len = sizeof(xprt->tcp_xid) - xprt->tcp_offset; - dprintk("RPC: reading XID (%Zu bytes)\n", len); - p = ((char *) &xprt->tcp_xid) + xprt->tcp_offset; - used = tcp_copy_data(desc, p, len); - xprt->tcp_offset += used; - if (used != len) - return; - xprt->tcp_flags &= ~XPRT_COPY_XID; - xprt->tcp_flags |= XPRT_COPY_DATA; - xprt->tcp_copied = 4; - dprintk("RPC: reading reply for XID %08x\n", - ntohl(xprt->tcp_xid)); - tcp_check_recm(xprt); -} - -/* - * TCP read and complete request - */ -static inline void -tcp_read_request(struct rpc_xprt *xprt, skb_reader_t *desc) -{ - struct rpc_rqst *req; - struct xdr_buf *rcvbuf; - size_t len; - ssize_t r; - - /* Find and lock the request corresponding to this xid */ - spin_lock(&xprt->sock_lock); - req = xprt_lookup_rqst(xprt, xprt->tcp_xid); - if (!req) { - xprt->tcp_flags &= ~XPRT_COPY_DATA; - dprintk("RPC: XID %08x request not found!\n", - ntohl(xprt->tcp_xid)); - spin_unlock(&xprt->sock_lock); - return; - } - - rcvbuf = &req->rq_private_buf; - len = desc->count; - if (len > xprt->tcp_reclen - xprt->tcp_offset) { - skb_reader_t my_desc; - - len = xprt->tcp_reclen - xprt->tcp_offset; - memcpy(&my_desc, desc, sizeof(my_desc)); - my_desc.count = len; - r = xdr_partial_copy_from_skb(rcvbuf, xprt->tcp_copied, - &my_desc, tcp_copy_data); - desc->count -= r; - desc->offset += r; - } else - r = xdr_partial_copy_from_skb(rcvbuf, xprt->tcp_copied, - desc, tcp_copy_data); - - if (r > 0) { - xprt->tcp_copied += r; - xprt->tcp_offset += r; - } - if (r != len) { - /* Error when copying to the receive buffer, - * usually because we weren't able to allocate - * additional buffer pages. All we can do now - * is turn off XPRT_COPY_DATA, so the request - * will not receive any additional updates, - * and time out. - * Any remaining data from this record will - * be discarded. - */ - xprt->tcp_flags &= ~XPRT_COPY_DATA; - dprintk("RPC: XID %08x truncated request\n", - ntohl(xprt->tcp_xid)); - dprintk("RPC: xprt = %p, tcp_copied = %lu, tcp_offset = %u, tcp_reclen = %u\n", - xprt, xprt->tcp_copied, xprt->tcp_offset, xprt->tcp_reclen); - goto out; - } - - dprintk("RPC: XID %08x read %Zd bytes\n", - ntohl(xprt->tcp_xid), r); - dprintk("RPC: xprt = %p, tcp_copied = %lu, tcp_offset = %u, tcp_reclen = %u\n", - xprt, xprt->tcp_copied, xprt->tcp_offset, xprt->tcp_reclen); - - if (xprt->tcp_copied == req->rq_private_buf.buflen) - xprt->tcp_flags &= ~XPRT_COPY_DATA; - else if (xprt->tcp_offset == xprt->tcp_reclen) { - if (xprt->tcp_flags & XPRT_LAST_FRAG) - xprt->tcp_flags &= ~XPRT_COPY_DATA; - } - -out: - if (!(xprt->tcp_flags & XPRT_COPY_DATA)) { - dprintk("RPC: %4d received reply complete\n", - req->rq_task->tk_pid); - xprt_complete_rqst(xprt, req, xprt->tcp_copied); - } - spin_unlock(&xprt->sock_lock); - tcp_check_recm(xprt); -} - -/* - * TCP discard extra bytes from a short read - */ -static inline void -tcp_read_discard(struct rpc_xprt *xprt, skb_reader_t *desc) -{ - size_t len; - - len = xprt->tcp_reclen - xprt->tcp_offset; - if (len > desc->count) - len = desc->count; - desc->count -= len; - desc->offset += len; - xprt->tcp_offset += len; - dprintk("RPC: discarded %Zu bytes\n", len); - tcp_check_recm(xprt); -} - -/* - * TCP record receive routine - * We first have to grab the record marker, then the XID, then the data. - */ -static int -tcp_data_recv(read_descriptor_t *rd_desc, struct sk_buff *skb, - unsigned int offset, size_t len) -{ - struct rpc_xprt *xprt = rd_desc->arg.data; - skb_reader_t desc = { - .skb = skb, - .offset = offset, - .count = len, - .csum = 0 - }; - - dprintk("RPC: tcp_data_recv\n"); - do { - /* Read in a new fragment marker if necessary */ - /* Can we ever really expect to get completely empty fragments? */ - if (xprt->tcp_flags & XPRT_COPY_RECM) { - tcp_read_fraghdr(xprt, &desc); - continue; - } - /* Read in the xid if necessary */ - if (xprt->tcp_flags & XPRT_COPY_XID) { - tcp_read_xid(xprt, &desc); - continue; - } - /* Read in the request data */ - if (xprt->tcp_flags & XPRT_COPY_DATA) { - tcp_read_request(xprt, &desc); - continue; - } - /* Skip over any trailing bytes on short reads */ - tcp_read_discard(xprt, &desc); - } while (desc.count); - dprintk("RPC: tcp_data_recv done\n"); - return len - desc.count; -} - -static void tcp_data_ready(struct sock *sk, int bytes) -{ - struct rpc_xprt *xprt; - read_descriptor_t rd_desc; - - read_lock(&sk->sk_callback_lock); - dprintk("RPC: tcp_data_ready...\n"); - if (!(xprt = xprt_from_sock(sk))) { - printk("RPC: tcp_data_ready socket info not found!\n"); - goto out; - } - if (xprt->shutdown) - goto out; - - /* We use rd_desc to pass struct xprt to tcp_data_recv */ - rd_desc.arg.data = xprt; - rd_desc.count = 65536; - tcp_read_sock(sk, &rd_desc, tcp_data_recv); -out: - read_unlock(&sk->sk_callback_lock); -} - -static void -tcp_state_change(struct sock *sk) -{ - struct rpc_xprt *xprt; - - read_lock(&sk->sk_callback_lock); - if (!(xprt = xprt_from_sock(sk))) - goto out; - dprintk("RPC: tcp_state_change client %p...\n", xprt); - dprintk("RPC: state %x conn %d dead %d zapped %d\n", - sk->sk_state, xprt_connected(xprt), - sock_flag(sk, SOCK_DEAD), - sock_flag(sk, SOCK_ZAPPED)); - - switch (sk->sk_state) { - case TCP_ESTABLISHED: - spin_lock_bh(&xprt->sock_lock); - if (!xprt_test_and_set_connected(xprt)) { - /* Reset TCP record info */ - xprt->tcp_offset = 0; - xprt->tcp_reclen = 0; - xprt->tcp_copied = 0; - xprt->tcp_flags = XPRT_COPY_RECM | XPRT_COPY_XID; - rpc_wake_up(&xprt->pending); - } - spin_unlock_bh(&xprt->sock_lock); - break; - case TCP_SYN_SENT: - case TCP_SYN_RECV: - break; - default: - xprt_disconnect(xprt); - break; - } - out: - read_unlock(&sk->sk_callback_lock); -} - -/* - * Called when more output buffer space is available for this socket. - * We try not to wake our writers until they can make "significant" - * progress, otherwise we'll waste resources thrashing sock_sendmsg - * with a bunch of small requests. - */ -static void -xprt_write_space(struct sock *sk) -{ - struct rpc_xprt *xprt; - struct socket *sock; - - read_lock(&sk->sk_callback_lock); - if (!(xprt = xprt_from_sock(sk)) || !(sock = sk->sk_socket)) - goto out; - if (xprt->shutdown) - goto out; - - /* Wait until we have enough socket memory */ - if (xprt->stream) { - /* from net/core/stream.c:sk_stream_write_space */ - if (sk_stream_wspace(sk) < sk_stream_min_wspace(sk)) - goto out; - } else { - /* from net/core/sock.c:sock_def_write_space */ - if (!sock_writeable(sk)) - goto out; - } - - if (!test_and_clear_bit(SOCK_NOSPACE, &sock->flags)) - goto out; - - spin_lock_bh(&xprt->sock_lock); - if (xprt->snd_task) - rpc_wake_up_task(xprt->snd_task); - spin_unlock_bh(&xprt->sock_lock); -out: - read_unlock(&sk->sk_callback_lock); -} - /* * RPC receive timeout handler. */ @@ -1161,19 +539,10 @@ xprt_transmit(struct rpc_task *task) struct rpc_clnt *clnt = task->tk_client; struct rpc_rqst *req = task->tk_rqstp; struct rpc_xprt *xprt = req->rq_xprt; - int status, retry = 0; - + int status; dprintk("RPC: %4d xprt_transmit(%u)\n", task->tk_pid, req->rq_slen); - /* set up everything as needed. */ - /* Write the record marker */ - if (xprt->stream) { - u32 *marker = req->rq_svec[0].iov_base; - - *marker = htonl(0x80000000|(req->rq_slen-sizeof(*marker))); - } - smp_rmb(); if (!req->rq_received) { if (list_empty(&req->rq_list)) { @@ -1191,41 +560,9 @@ xprt_transmit(struct rpc_task *task) } else if (!req->rq_bytes_sent) return; - /* Continue transmitting the packet/record. We must be careful - * to cope with writespace callbacks arriving _after_ we have - * called xprt_sendmsg(). - */ - while (1) { - req->rq_xtime = jiffies; - status = xprt_sendmsg(xprt, req); - - if (status < 0) - break; - - if (xprt->stream) { - req->rq_bytes_sent += status; - - /* If we've sent the entire packet, immediately - * reset the count of bytes sent. */ - if (req->rq_bytes_sent >= req->rq_slen) { - req->rq_bytes_sent = 0; - goto out_receive; - } - } else { - if (status >= req->rq_slen) - goto out_receive; - status = -EAGAIN; - break; - } - - dprintk("RPC: %4d xmit incomplete (%d left of %d)\n", - task->tk_pid, req->rq_slen - req->rq_bytes_sent, - req->rq_slen); - - status = -EAGAIN; - if (retry++ > 50) - break; - } + status = xprt->ops->send_request(task); + if (!status) + goto out_receive; /* Note: at this point, task->tk_sleeping has not yet been set, * hence there is no danger of the waking up task being put on @@ -1234,26 +571,10 @@ xprt_transmit(struct rpc_task *task) task->tk_status = status; switch (status) { - case -EAGAIN: - if (test_bit(SOCK_ASYNC_NOSPACE, &xprt->sock->flags)) { - /* Protect against races with xprt_write_space */ - spin_lock_bh(&xprt->sock_lock); - /* Don't race with disconnect */ - if (!xprt_connected(xprt)) - task->tk_status = -ENOTCONN; - else if (test_bit(SOCK_NOSPACE, &xprt->sock->flags)) { - task->tk_timeout = req->rq_timeout; - rpc_sleep_on(&xprt->pending, task, NULL, NULL); - } - spin_unlock_bh(&xprt->sock_lock); - return; - } - /* Keep holding the socket if it is blocked */ - rpc_delay(task, HZ>>4); - return; case -ECONNREFUSED: task->tk_timeout = RPC_REESTABLISH_TIMEOUT; rpc_sleep_on(&xprt->sending, task, NULL, NULL); + case -EAGAIN: case -ENOTCONN: return; default: @@ -1367,7 +688,8 @@ xprt_release(struct rpc_task *task) list_del(&req->rq_list); xprt->last_used = jiffies; if (list_empty(&xprt->recv) && !xprt->shutdown) - mod_timer(&xprt->timer, xprt->last_used + XPRT_IDLE_TIMEOUT); + mod_timer(&xprt->timer, + xprt->last_used + RPC_IDLE_DISCONNECT_TIMEOUT); spin_unlock_bh(&xprt->sock_lock); task->tk_rqstp = NULL; memset(req, 0, sizeof(*req)); /* mark unused */ @@ -1380,18 +702,6 @@ xprt_release(struct rpc_task *task) spin_unlock(&xprt->xprt_lock); } -/* - * Set default timeout parameters - */ -static void -xprt_default_timeout(struct rpc_timeout *to, int proto) -{ - if (proto == IPPROTO_UDP) - xprt_set_timeout(to, 5, 5 * HZ); - else - xprt_set_timeout(to, 2, 60 * HZ); -} - /* * Set constant timeout */ @@ -1405,68 +715,51 @@ xprt_set_timeout(struct rpc_timeout *to, unsigned int retr, unsigned long incr) to->to_exponential = 0; } -unsigned int xprt_udp_slot_table_entries = RPC_DEF_SLOT_TABLE; -unsigned int xprt_tcp_slot_table_entries = RPC_DEF_SLOT_TABLE; - /* * Initialize an RPC client */ static struct rpc_xprt * xprt_setup(int proto, struct sockaddr_in *ap, struct rpc_timeout *to) { + int result; struct rpc_xprt *xprt; - unsigned int entries; - size_t slot_table_size; struct rpc_rqst *req; - dprintk("RPC: setting up %s transport...\n", - proto == IPPROTO_UDP? "UDP" : "TCP"); - - entries = (proto == IPPROTO_TCP)? - xprt_tcp_slot_table_entries : xprt_udp_slot_table_entries; - if ((xprt = kmalloc(sizeof(struct rpc_xprt), GFP_KERNEL)) == NULL) return ERR_PTR(-ENOMEM); memset(xprt, 0, sizeof(*xprt)); /* Nnnngh! */ - xprt->max_reqs = entries; - slot_table_size = entries * sizeof(xprt->slot[0]); - xprt->slot = kmalloc(slot_table_size, GFP_KERNEL); - if (xprt->slot == NULL) { - kfree(xprt); - return ERR_PTR(-ENOMEM); - } - memset(xprt->slot, 0, slot_table_size); xprt->addr = *ap; - xprt->prot = proto; - xprt->stream = (proto == IPPROTO_TCP)? 1 : 0; - if (xprt->stream) { - xprt->cwnd = RPC_MAXCWND(xprt); - xprt->nocong = 1; - xprt->max_payload = (1U << 31) - 1; - } else { - xprt->cwnd = RPC_INITCWND; - xprt->max_payload = (1U << 16) - (MAX_HEADER << 3); + + switch (proto) { + case IPPROTO_UDP: + result = xs_setup_udp(xprt, to); + break; + case IPPROTO_TCP: + result = xs_setup_tcp(xprt, to); + break; + default: + printk(KERN_ERR "RPC: unrecognized transport protocol: %d\n", + proto); + result = -EIO; + break; + } + if (result) { + kfree(xprt); + return ERR_PTR(result); } + spin_lock_init(&xprt->sock_lock); spin_lock_init(&xprt->xprt_lock); init_waitqueue_head(&xprt->cong_wait); INIT_LIST_HEAD(&xprt->free); INIT_LIST_HEAD(&xprt->recv); - INIT_WORK(&xprt->sock_connect, xprt_socket_connect, xprt); INIT_WORK(&xprt->task_cleanup, xprt_socket_autoclose, xprt); init_timer(&xprt->timer); xprt->timer.function = xprt_init_autodisconnect; xprt->timer.data = (unsigned long) xprt; xprt->last_used = jiffies; - xprt->port = XPRT_MAX_RESVPORT; - - /* Set timeout parameters */ - if (to) { - xprt->timeout = *to; - } else - xprt_default_timeout(&xprt->timeout, xprt->prot); rpc_init_wait_queue(&xprt->pending, "xprt_pending"); rpc_init_wait_queue(&xprt->sending, "xprt_sending"); @@ -1474,134 +767,17 @@ xprt_setup(int proto, struct sockaddr_in *ap, struct rpc_timeout *to) rpc_init_priority_wait_queue(&xprt->backlog, "xprt_backlog"); /* initialize free list */ - for (req = &xprt->slot[entries-1]; req >= &xprt->slot[0]; req--) + for (req = &xprt->slot[xprt->max_reqs-1]; req >= &xprt->slot[0]; req--) list_add(&req->rq_list, &xprt->free); xprt_init_xid(xprt); - /* Check whether we want to use a reserved port */ - xprt->resvport = capable(CAP_NET_BIND_SERVICE) ? 1 : 0; - dprintk("RPC: created transport %p with %u slots\n", xprt, xprt->max_reqs); return xprt; } -/* - * Bind to a reserved port - */ -static inline int xprt_bindresvport(struct rpc_xprt *xprt, struct socket *sock) -{ - struct sockaddr_in myaddr = { - .sin_family = AF_INET, - }; - int err, port; - - /* Were we already bound to a given port? Try to reuse it */ - port = xprt->port; - do { - myaddr.sin_port = htons(port); - err = sock->ops->bind(sock, (struct sockaddr *) &myaddr, - sizeof(myaddr)); - if (err == 0) { - xprt->port = port; - return 0; - } - if (--port == 0) - port = XPRT_MAX_RESVPORT; - } while (err == -EADDRINUSE && port != xprt->port); - - printk("RPC: Can't bind to reserved port (%d).\n", -err); - return err; -} - -static void -xprt_bind_socket(struct rpc_xprt *xprt, struct socket *sock) -{ - struct sock *sk = sock->sk; - - if (xprt->inet) - return; - - write_lock_bh(&sk->sk_callback_lock); - sk->sk_user_data = xprt; - xprt->old_data_ready = sk->sk_data_ready; - xprt->old_state_change = sk->sk_state_change; - xprt->old_write_space = sk->sk_write_space; - if (xprt->prot == IPPROTO_UDP) { - sk->sk_data_ready = udp_data_ready; - sk->sk_no_check = UDP_CSUM_NORCV; - xprt_set_connected(xprt); - } else { - tcp_sk(sk)->nonagle = 1; /* disable Nagle's algorithm */ - sk->sk_data_ready = tcp_data_ready; - sk->sk_state_change = tcp_state_change; - xprt_clear_connected(xprt); - } - sk->sk_write_space = xprt_write_space; - - /* Reset to new socket */ - xprt->sock = sock; - xprt->inet = sk; - write_unlock_bh(&sk->sk_callback_lock); - - return; -} - -/* - * Set socket buffer length - */ -void -xprt_sock_setbufsize(struct rpc_xprt *xprt) -{ - struct sock *sk = xprt->inet; - - if (xprt->stream) - return; - if (xprt->rcvsize) { - sk->sk_userlocks |= SOCK_RCVBUF_LOCK; - sk->sk_rcvbuf = xprt->rcvsize * xprt->max_reqs * 2; - } - if (xprt->sndsize) { - sk->sk_userlocks |= SOCK_SNDBUF_LOCK; - sk->sk_sndbuf = xprt->sndsize * xprt->max_reqs * 2; - sk->sk_write_space(sk); - } -} - -/* - * Datastream sockets are created here, but xprt_connect will create - * and connect stream sockets. - */ -static struct socket * xprt_create_socket(struct rpc_xprt *xprt, int proto, int resvport) -{ - struct socket *sock; - int type, err; - - dprintk("RPC: xprt_create_socket(%s %d)\n", - (proto == IPPROTO_UDP)? "udp" : "tcp", proto); - - type = (proto == IPPROTO_UDP)? SOCK_DGRAM : SOCK_STREAM; - - if ((err = sock_create_kern(PF_INET, type, proto, &sock)) < 0) { - printk("RPC: can't create socket (%d).\n", -err); - return NULL; - } - - /* If the caller has the capability, bind to a reserved port */ - if (resvport && xprt_bindresvport(xprt, sock) < 0) { - printk("RPC: can't bind to reserved port.\n"); - goto failed; - } - - return sock; - -failed: - sock_release(sock); - return NULL; -} - /* * Create an RPC client transport given the protocol and peer address. */ @@ -1631,10 +807,6 @@ xprt_shutdown(struct rpc_xprt *xprt) rpc_wake_up(&xprt->backlog); wake_up(&xprt->cong_wait); del_timer_sync(&xprt->timer); - - /* synchronously wait for connect worker to finish */ - cancel_delayed_work(&xprt->sock_connect); - flush_scheduled_work(); } /* @@ -1655,9 +827,7 @@ xprt_destroy(struct rpc_xprt *xprt) { dprintk("RPC: destroying transport %p\n", xprt); xprt_shutdown(xprt); - xprt_disconnect(xprt); - xprt_close(xprt); - kfree(xprt->slot); + xprt->ops->destroy(xprt); kfree(xprt); return 0; diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c new file mode 100644 index 000000000000..fa1180ac4823 --- /dev/null +++ b/net/sunrpc/xprtsock.c @@ -0,0 +1,1021 @@ +/* + * linux/net/sunrpc/xprtsock.c + * + * Client-side transport implementation for sockets. + * + * TCP callback races fixes (C) 1998 Red Hat Software + * TCP send fixes (C) 1998 Red Hat Software + * TCP NFS related read + write fixes + * (C) 1999 Dave Airlie, University of Limerick, Ireland + * + * Rewrite of larges part of the code in order to stabilize TCP stuff. + * Fix behaviour when socket buffer is full. + * (C) 1999 Trond Myklebust + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include +#include +#include +#include + +#ifdef RPC_DEBUG +# undef RPC_DEBUG_DATA +# define RPCDBG_FACILITY RPCDBG_XPRT +#endif + +#define XPRT_MAX_RESVPORT (800) + +#ifdef RPC_DEBUG_DATA +/* + * Print the buffer contents (first 128 bytes only--just enough for + * diropres return). + */ +static void +xprt_pktdump(char *msg, u32 *packet, unsigned int count) +{ + u8 *buf = (u8 *) packet; + int j; + + dprintk("RPC: %s\n", msg); + for (j = 0; j < count && j < 128; j += 4) { + if (!(j & 31)) { + if (j) + dprintk("\n"); + dprintk("0x%04x ", j); + } + dprintk("%02x%02x%02x%02x ", + buf[j], buf[j+1], buf[j+2], buf[j+3]); + } + dprintk("\n"); +} +#else +static inline void +xprt_pktdump(char *msg, u32 *packet, unsigned int count) +{ + /* NOP */ +} +#endif + +/* + * Look up RPC transport given an INET socket + */ +static inline struct rpc_xprt * +xprt_from_sock(struct sock *sk) +{ + return (struct rpc_xprt *) sk->sk_user_data; +} + +static int +xdr_sendpages(struct socket *sock, struct sockaddr *addr, int addrlen, + struct xdr_buf *xdr, unsigned int base, int msgflags) +{ + struct page **ppage = xdr->pages; + unsigned int len, pglen = xdr->page_len; + int err, ret = 0; + ssize_t (*sendpage)(struct socket *, struct page *, int, size_t, int); + + len = xdr->head[0].iov_len; + if (base < len || (addr != NULL && base == 0)) { + struct kvec iov = { + .iov_base = xdr->head[0].iov_base + base, + .iov_len = len - base, + }; + struct msghdr msg = { + .msg_name = addr, + .msg_namelen = addrlen, + .msg_flags = msgflags, + }; + if (xdr->len > len) + msg.msg_flags |= MSG_MORE; + + if (iov.iov_len != 0) + err = kernel_sendmsg(sock, &msg, &iov, 1, iov.iov_len); + else + err = kernel_sendmsg(sock, &msg, NULL, 0, 0); + if (ret == 0) + ret = err; + else if (err > 0) + ret += err; + if (err != iov.iov_len) + goto out; + base = 0; + } else + base -= len; + + if (pglen == 0) + goto copy_tail; + if (base >= pglen) { + base -= pglen; + goto copy_tail; + } + if (base || xdr->page_base) { + pglen -= base; + base += xdr->page_base; + ppage += base >> PAGE_CACHE_SHIFT; + base &= ~PAGE_CACHE_MASK; + } + + sendpage = sock->ops->sendpage ? : sock_no_sendpage; + do { + int flags = msgflags; + + len = PAGE_CACHE_SIZE; + if (base) + len -= base; + if (pglen < len) + len = pglen; + + if (pglen != len || xdr->tail[0].iov_len != 0) + flags |= MSG_MORE; + + /* Hmm... We might be dealing with highmem pages */ + if (PageHighMem(*ppage)) + sendpage = sock_no_sendpage; + err = sendpage(sock, *ppage, base, len, flags); + if (ret == 0) + ret = err; + else if (err > 0) + ret += err; + if (err != len) + goto out; + base = 0; + ppage++; + } while ((pglen -= len) != 0); +copy_tail: + len = xdr->tail[0].iov_len; + if (base < len) { + struct kvec iov = { + .iov_base = xdr->tail[0].iov_base + base, + .iov_len = len - base, + }; + struct msghdr msg = { + .msg_flags = msgflags, + }; + err = kernel_sendmsg(sock, &msg, &iov, 1, iov.iov_len); + if (ret == 0) + ret = err; + else if (err > 0) + ret += err; + } +out: + return ret; +} + +/* + * Write data to socket. + */ +static inline int +xprt_sendmsg(struct rpc_xprt *xprt, struct rpc_rqst *req) +{ + struct socket *sock = xprt->sock; + struct xdr_buf *xdr = &req->rq_snd_buf; + struct sockaddr *addr = NULL; + int addrlen = 0; + unsigned int skip; + int result; + + if (!sock) + return -ENOTCONN; + + xprt_pktdump("packet data:", + req->rq_svec->iov_base, + req->rq_svec->iov_len); + + /* For UDP, we need to provide an address */ + if (!xprt->stream) { + addr = (struct sockaddr *) &xprt->addr; + addrlen = sizeof(xprt->addr); + } + /* Dont repeat bytes */ + skip = req->rq_bytes_sent; + + clear_bit(SOCK_ASYNC_NOSPACE, &sock->flags); + result = xdr_sendpages(sock, addr, addrlen, xdr, skip, MSG_DONTWAIT); + + dprintk("RPC: xprt_sendmsg(%d) = %d\n", xdr->len - skip, result); + + if (result >= 0) + return result; + + switch (result) { + case -ECONNREFUSED: + /* When the server has died, an ICMP port unreachable message + * prompts ECONNREFUSED. + */ + case -EAGAIN: + break; + case -ECONNRESET: + case -ENOTCONN: + case -EPIPE: + /* connection broken */ + if (xprt->stream) + result = -ENOTCONN; + break; + default: + printk(KERN_NOTICE "RPC: sendmsg returned error %d\n", -result); + } + return result; +} + +static int +xprt_send_request(struct rpc_task *task) +{ + struct rpc_rqst *req = task->tk_rqstp; + struct rpc_xprt *xprt = req->rq_xprt; + int status, retry = 0; + + /* set up everything as needed. */ + /* Write the record marker */ + if (xprt->stream) { + u32 *marker = req->rq_svec[0].iov_base; + + *marker = htonl(0x80000000|(req->rq_slen-sizeof(*marker))); + } + + /* Continue transmitting the packet/record. We must be careful + * to cope with writespace callbacks arriving _after_ we have + * called xprt_sendmsg(). + */ + while (1) { + req->rq_xtime = jiffies; + status = xprt_sendmsg(xprt, req); + + if (status < 0) + break; + + if (xprt->stream) { + req->rq_bytes_sent += status; + + /* If we've sent the entire packet, immediately + * reset the count of bytes sent. */ + if (req->rq_bytes_sent >= req->rq_slen) { + req->rq_bytes_sent = 0; + return 0; + } + } else { + if (status >= req->rq_slen) + return 0; + status = -EAGAIN; + break; + } + + dprintk("RPC: %4d xmit incomplete (%d left of %d)\n", + task->tk_pid, req->rq_slen - req->rq_bytes_sent, + req->rq_slen); + + status = -EAGAIN; + if (retry++ > 50) + break; + } + + if (status == -EAGAIN) { + if (test_bit(SOCK_ASYNC_NOSPACE, &xprt->sock->flags)) { + /* Protect against races with xprt_write_space */ + spin_lock_bh(&xprt->sock_lock); + /* Don't race with disconnect */ + if (!xprt_connected(xprt)) + task->tk_status = -ENOTCONN; + else if (test_bit(SOCK_NOSPACE, &xprt->sock->flags)) { + task->tk_timeout = req->rq_timeout; + rpc_sleep_on(&xprt->pending, task, NULL, NULL); + } + spin_unlock_bh(&xprt->sock_lock); + return status; + } + /* Keep holding the socket if it is blocked */ + rpc_delay(task, HZ>>4); + } + return status; +} + +/* + * Close down a transport socket + */ +static void +xprt_close(struct rpc_xprt *xprt) +{ + struct socket *sock = xprt->sock; + struct sock *sk = xprt->inet; + + if (!sk) + return; + + write_lock_bh(&sk->sk_callback_lock); + xprt->inet = NULL; + xprt->sock = NULL; + + sk->sk_user_data = NULL; + sk->sk_data_ready = xprt->old_data_ready; + sk->sk_state_change = xprt->old_state_change; + sk->sk_write_space = xprt->old_write_space; + write_unlock_bh(&sk->sk_callback_lock); + + sk->sk_no_check = 0; + + sock_release(sock); +} + +static void xprt_socket_destroy(struct rpc_xprt *xprt) +{ + cancel_delayed_work(&xprt->sock_connect); + flush_scheduled_work(); + + xprt_disconnect(xprt); + xprt_close(xprt); + kfree(xprt->slot); +} + +/* + * Input handler for RPC replies. Called from a bottom half and hence + * atomic. + */ +static void +udp_data_ready(struct sock *sk, int len) +{ + struct rpc_task *task; + struct rpc_xprt *xprt; + struct rpc_rqst *rovr; + struct sk_buff *skb; + int err, repsize, copied; + u32 _xid, *xp; + + read_lock(&sk->sk_callback_lock); + dprintk("RPC: udp_data_ready...\n"); + if (!(xprt = xprt_from_sock(sk))) { + printk("RPC: udp_data_ready request not found!\n"); + goto out; + } + + dprintk("RPC: udp_data_ready client %p\n", xprt); + + if ((skb = skb_recv_datagram(sk, 0, 1, &err)) == NULL) + goto out; + + if (xprt->shutdown) + goto dropit; + + repsize = skb->len - sizeof(struct udphdr); + if (repsize < 4) { + printk("RPC: impossible RPC reply size %d!\n", repsize); + goto dropit; + } + + /* Copy the XID from the skb... */ + xp = skb_header_pointer(skb, sizeof(struct udphdr), + sizeof(_xid), &_xid); + if (xp == NULL) + goto dropit; + + /* Look up and lock the request corresponding to the given XID */ + spin_lock(&xprt->sock_lock); + rovr = xprt_lookup_rqst(xprt, *xp); + if (!rovr) + goto out_unlock; + task = rovr->rq_task; + + dprintk("RPC: %4d received reply\n", task->tk_pid); + + if ((copied = rovr->rq_private_buf.buflen) > repsize) + copied = repsize; + + /* Suck it into the iovec, verify checksum if not done by hw. */ + if (csum_partial_copy_to_xdr(&rovr->rq_private_buf, skb)) + goto out_unlock; + + /* Something worked... */ + dst_confirm(skb->dst); + + xprt_complete_rqst(xprt, rovr, copied); + + out_unlock: + spin_unlock(&xprt->sock_lock); + dropit: + skb_free_datagram(sk, skb); + out: + read_unlock(&sk->sk_callback_lock); +} + +/* + * Copy from an skb into memory and shrink the skb. + */ +static inline size_t +tcp_copy_data(skb_reader_t *desc, void *p, size_t len) +{ + if (len > desc->count) + len = desc->count; + if (skb_copy_bits(desc->skb, desc->offset, p, len)) { + dprintk("RPC: failed to copy %zu bytes from skb. %zu bytes remain\n", + len, desc->count); + return 0; + } + desc->offset += len; + desc->count -= len; + dprintk("RPC: copied %zu bytes from skb. %zu bytes remain\n", + len, desc->count); + return len; +} + +/* + * TCP read fragment marker + */ +static inline void +tcp_read_fraghdr(struct rpc_xprt *xprt, skb_reader_t *desc) +{ + size_t len, used; + char *p; + + p = ((char *) &xprt->tcp_recm) + xprt->tcp_offset; + len = sizeof(xprt->tcp_recm) - xprt->tcp_offset; + used = tcp_copy_data(desc, p, len); + xprt->tcp_offset += used; + if (used != len) + return; + xprt->tcp_reclen = ntohl(xprt->tcp_recm); + if (xprt->tcp_reclen & 0x80000000) + xprt->tcp_flags |= XPRT_LAST_FRAG; + else + xprt->tcp_flags &= ~XPRT_LAST_FRAG; + xprt->tcp_reclen &= 0x7fffffff; + xprt->tcp_flags &= ~XPRT_COPY_RECM; + xprt->tcp_offset = 0; + /* Sanity check of the record length */ + if (xprt->tcp_reclen < 4) { + printk(KERN_ERR "RPC: Invalid TCP record fragment length\n"); + xprt_disconnect(xprt); + } + dprintk("RPC: reading TCP record fragment of length %d\n", + xprt->tcp_reclen); +} + +static void +tcp_check_recm(struct rpc_xprt *xprt) +{ + dprintk("RPC: xprt = %p, tcp_copied = %lu, tcp_offset = %u, tcp_reclen = %u, tcp_flags = %lx\n", + xprt, xprt->tcp_copied, xprt->tcp_offset, xprt->tcp_reclen, xprt->tcp_flags); + if (xprt->tcp_offset == xprt->tcp_reclen) { + xprt->tcp_flags |= XPRT_COPY_RECM; + xprt->tcp_offset = 0; + if (xprt->tcp_flags & XPRT_LAST_FRAG) { + xprt->tcp_flags &= ~XPRT_COPY_DATA; + xprt->tcp_flags |= XPRT_COPY_XID; + xprt->tcp_copied = 0; + } + } +} + +/* + * TCP read xid + */ +static inline void +tcp_read_xid(struct rpc_xprt *xprt, skb_reader_t *desc) +{ + size_t len, used; + char *p; + + len = sizeof(xprt->tcp_xid) - xprt->tcp_offset; + dprintk("RPC: reading XID (%Zu bytes)\n", len); + p = ((char *) &xprt->tcp_xid) + xprt->tcp_offset; + used = tcp_copy_data(desc, p, len); + xprt->tcp_offset += used; + if (used != len) + return; + xprt->tcp_flags &= ~XPRT_COPY_XID; + xprt->tcp_flags |= XPRT_COPY_DATA; + xprt->tcp_copied = 4; + dprintk("RPC: reading reply for XID %08x\n", + ntohl(xprt->tcp_xid)); + tcp_check_recm(xprt); +} + +/* + * TCP read and complete request + */ +static inline void +tcp_read_request(struct rpc_xprt *xprt, skb_reader_t *desc) +{ + struct rpc_rqst *req; + struct xdr_buf *rcvbuf; + size_t len; + ssize_t r; + + /* Find and lock the request corresponding to this xid */ + spin_lock(&xprt->sock_lock); + req = xprt_lookup_rqst(xprt, xprt->tcp_xid); + if (!req) { + xprt->tcp_flags &= ~XPRT_COPY_DATA; + dprintk("RPC: XID %08x request not found!\n", + ntohl(xprt->tcp_xid)); + spin_unlock(&xprt->sock_lock); + return; + } + + rcvbuf = &req->rq_private_buf; + len = desc->count; + if (len > xprt->tcp_reclen - xprt->tcp_offset) { + skb_reader_t my_desc; + + len = xprt->tcp_reclen - xprt->tcp_offset; + memcpy(&my_desc, desc, sizeof(my_desc)); + my_desc.count = len; + r = xdr_partial_copy_from_skb(rcvbuf, xprt->tcp_copied, + &my_desc, tcp_copy_data); + desc->count -= r; + desc->offset += r; + } else + r = xdr_partial_copy_from_skb(rcvbuf, xprt->tcp_copied, + desc, tcp_copy_data); + + if (r > 0) { + xprt->tcp_copied += r; + xprt->tcp_offset += r; + } + if (r != len) { + /* Error when copying to the receive buffer, + * usually because we weren't able to allocate + * additional buffer pages. All we can do now + * is turn off XPRT_COPY_DATA, so the request + * will not receive any additional updates, + * and time out. + * Any remaining data from this record will + * be discarded. + */ + xprt->tcp_flags &= ~XPRT_COPY_DATA; + dprintk("RPC: XID %08x truncated request\n", + ntohl(xprt->tcp_xid)); + dprintk("RPC: xprt = %p, tcp_copied = %lu, tcp_offset = %u, tcp_reclen = %u\n", + xprt, xprt->tcp_copied, xprt->tcp_offset, xprt->tcp_reclen); + goto out; + } + + dprintk("RPC: XID %08x read %Zd bytes\n", + ntohl(xprt->tcp_xid), r); + dprintk("RPC: xprt = %p, tcp_copied = %lu, tcp_offset = %u, tcp_reclen = %u\n", + xprt, xprt->tcp_copied, xprt->tcp_offset, xprt->tcp_reclen); + + if (xprt->tcp_copied == req->rq_private_buf.buflen) + xprt->tcp_flags &= ~XPRT_COPY_DATA; + else if (xprt->tcp_offset == xprt->tcp_reclen) { + if (xprt->tcp_flags & XPRT_LAST_FRAG) + xprt->tcp_flags &= ~XPRT_COPY_DATA; + } + +out: + if (!(xprt->tcp_flags & XPRT_COPY_DATA)) { + dprintk("RPC: %4d received reply complete\n", + req->rq_task->tk_pid); + xprt_complete_rqst(xprt, req, xprt->tcp_copied); + } + spin_unlock(&xprt->sock_lock); + tcp_check_recm(xprt); +} + +/* + * TCP discard extra bytes from a short read + */ +static inline void +tcp_read_discard(struct rpc_xprt *xprt, skb_reader_t *desc) +{ + size_t len; + + len = xprt->tcp_reclen - xprt->tcp_offset; + if (len > desc->count) + len = desc->count; + desc->count -= len; + desc->offset += len; + xprt->tcp_offset += len; + dprintk("RPC: discarded %Zu bytes\n", len); + tcp_check_recm(xprt); +} + +/* + * TCP record receive routine + * We first have to grab the record marker, then the XID, then the data. + */ +static int +tcp_data_recv(read_descriptor_t *rd_desc, struct sk_buff *skb, + unsigned int offset, size_t len) +{ + struct rpc_xprt *xprt = rd_desc->arg.data; + skb_reader_t desc = { + .skb = skb, + .offset = offset, + .count = len, + .csum = 0 + }; + + dprintk("RPC: tcp_data_recv\n"); + do { + /* Read in a new fragment marker if necessary */ + /* Can we ever really expect to get completely empty fragments? */ + if (xprt->tcp_flags & XPRT_COPY_RECM) { + tcp_read_fraghdr(xprt, &desc); + continue; + } + /* Read in the xid if necessary */ + if (xprt->tcp_flags & XPRT_COPY_XID) { + tcp_read_xid(xprt, &desc); + continue; + } + /* Read in the request data */ + if (xprt->tcp_flags & XPRT_COPY_DATA) { + tcp_read_request(xprt, &desc); + continue; + } + /* Skip over any trailing bytes on short reads */ + tcp_read_discard(xprt, &desc); + } while (desc.count); + dprintk("RPC: tcp_data_recv done\n"); + return len - desc.count; +} + +static void tcp_data_ready(struct sock *sk, int bytes) +{ + struct rpc_xprt *xprt; + read_descriptor_t rd_desc; + + read_lock(&sk->sk_callback_lock); + dprintk("RPC: tcp_data_ready...\n"); + if (!(xprt = xprt_from_sock(sk))) { + printk("RPC: tcp_data_ready socket info not found!\n"); + goto out; + } + if (xprt->shutdown) + goto out; + + /* We use rd_desc to pass struct xprt to tcp_data_recv */ + rd_desc.arg.data = xprt; + rd_desc.count = 65536; + tcp_read_sock(sk, &rd_desc, tcp_data_recv); +out: + read_unlock(&sk->sk_callback_lock); +} + +static void +tcp_state_change(struct sock *sk) +{ + struct rpc_xprt *xprt; + + read_lock(&sk->sk_callback_lock); + if (!(xprt = xprt_from_sock(sk))) + goto out; + dprintk("RPC: tcp_state_change client %p...\n", xprt); + dprintk("RPC: state %x conn %d dead %d zapped %d\n", + sk->sk_state, xprt_connected(xprt), + sock_flag(sk, SOCK_DEAD), + sock_flag(sk, SOCK_ZAPPED)); + + switch (sk->sk_state) { + case TCP_ESTABLISHED: + spin_lock_bh(&xprt->sock_lock); + if (!xprt_test_and_set_connected(xprt)) { + /* Reset TCP record info */ + xprt->tcp_offset = 0; + xprt->tcp_reclen = 0; + xprt->tcp_copied = 0; + xprt->tcp_flags = XPRT_COPY_RECM | XPRT_COPY_XID; + rpc_wake_up(&xprt->pending); + } + spin_unlock_bh(&xprt->sock_lock); + break; + case TCP_SYN_SENT: + case TCP_SYN_RECV: + break; + default: + xprt_disconnect(xprt); + break; + } + out: + read_unlock(&sk->sk_callback_lock); +} + +/* + * Called when more output buffer space is available for this socket. + * We try not to wake our writers until they can make "significant" + * progress, otherwise we'll waste resources thrashing sock_sendmsg + * with a bunch of small requests. + */ +static void +xprt_write_space(struct sock *sk) +{ + struct rpc_xprt *xprt; + struct socket *sock; + + read_lock(&sk->sk_callback_lock); + if (!(xprt = xprt_from_sock(sk)) || !(sock = sk->sk_socket)) + goto out; + if (xprt->shutdown) + goto out; + + /* Wait until we have enough socket memory */ + if (xprt->stream) { + /* from net/core/stream.c:sk_stream_write_space */ + if (sk_stream_wspace(sk) < sk_stream_min_wspace(sk)) + goto out; + } else { + /* from net/core/sock.c:sock_def_write_space */ + if (!sock_writeable(sk)) + goto out; + } + + if (!test_and_clear_bit(SOCK_NOSPACE, &sock->flags)) + goto out; + + spin_lock_bh(&xprt->sock_lock); + if (xprt->snd_task) + rpc_wake_up_task(xprt->snd_task); + spin_unlock_bh(&xprt->sock_lock); +out: + read_unlock(&sk->sk_callback_lock); +} + +/* + * Set socket buffer length + */ +static void +xprt_sock_setbufsize(struct rpc_xprt *xprt) +{ + struct sock *sk = xprt->inet; + + if (xprt->stream) + return; + if (xprt->rcvsize) { + sk->sk_userlocks |= SOCK_RCVBUF_LOCK; + sk->sk_rcvbuf = xprt->rcvsize * xprt->max_reqs * 2; + } + if (xprt->sndsize) { + sk->sk_userlocks |= SOCK_SNDBUF_LOCK; + sk->sk_sndbuf = xprt->sndsize * xprt->max_reqs * 2; + sk->sk_write_space(sk); + } +} + +/* + * Bind to a reserved port + */ +static inline int xprt_bindresvport(struct rpc_xprt *xprt, struct socket *sock) +{ + struct sockaddr_in myaddr = { + .sin_family = AF_INET, + }; + int err, port; + + /* Were we already bound to a given port? Try to reuse it */ + port = xprt->port; + do { + myaddr.sin_port = htons(port); + err = sock->ops->bind(sock, (struct sockaddr *) &myaddr, + sizeof(myaddr)); + if (err == 0) { + xprt->port = port; + return 0; + } + if (--port == 0) + port = XPRT_MAX_RESVPORT; + } while (err == -EADDRINUSE && port != xprt->port); + + printk("RPC: Can't bind to reserved port (%d).\n", -err); + return err; +} + +static void +xprt_bind_socket(struct rpc_xprt *xprt, struct socket *sock) +{ + struct sock *sk = sock->sk; + + if (xprt->inet) + return; + + write_lock_bh(&sk->sk_callback_lock); + sk->sk_user_data = xprt; + xprt->old_data_ready = sk->sk_data_ready; + xprt->old_state_change = sk->sk_state_change; + xprt->old_write_space = sk->sk_write_space; + if (xprt->prot == IPPROTO_UDP) { + sk->sk_data_ready = udp_data_ready; + sk->sk_no_check = UDP_CSUM_NORCV; + xprt_set_connected(xprt); + } else { + tcp_sk(sk)->nonagle = 1; /* disable Nagle's algorithm */ + sk->sk_data_ready = tcp_data_ready; + sk->sk_state_change = tcp_state_change; + xprt_clear_connected(xprt); + } + sk->sk_write_space = xprt_write_space; + + /* Reset to new socket */ + xprt->sock = sock; + xprt->inet = sk; + write_unlock_bh(&sk->sk_callback_lock); + + return; +} + +/* + * Datastream sockets are created here, but xprt_connect will create + * and connect stream sockets. + */ +static struct socket * xprt_create_socket(struct rpc_xprt *xprt, int proto, int resvport) +{ + struct socket *sock; + int type, err; + + dprintk("RPC: xprt_create_socket(%s %d)\n", + (proto == IPPROTO_UDP)? "udp" : "tcp", proto); + + type = (proto == IPPROTO_UDP)? SOCK_DGRAM : SOCK_STREAM; + + if ((err = sock_create_kern(PF_INET, type, proto, &sock)) < 0) { + printk("RPC: can't create socket (%d).\n", -err); + return NULL; + } + + /* If the caller has the capability, bind to a reserved port */ + if (resvport && xprt_bindresvport(xprt, sock) < 0) { + printk("RPC: can't bind to reserved port.\n"); + goto failed; + } + + return sock; + +failed: + sock_release(sock); + return NULL; +} + +static void xprt_socket_connect(void *args) +{ + struct rpc_xprt *xprt = (struct rpc_xprt *)args; + struct socket *sock = xprt->sock; + int status = -EIO; + + if (xprt->shutdown || xprt->addr.sin_port == 0) + goto out; + + /* + * Start by resetting any existing state + */ + xprt_close(xprt); + sock = xprt_create_socket(xprt, xprt->prot, xprt->resvport); + if (sock == NULL) { + /* couldn't create socket or bind to reserved port; + * this is likely a permanent error, so cause an abort */ + goto out; + } + xprt_bind_socket(xprt, sock); + xprt_sock_setbufsize(xprt); + + status = 0; + if (!xprt->stream) + goto out; + + /* + * Tell the socket layer to start connecting... + */ + status = sock->ops->connect(sock, (struct sockaddr *) &xprt->addr, + sizeof(xprt->addr), O_NONBLOCK); + dprintk("RPC: %p connect status %d connected %d sock state %d\n", + xprt, -status, xprt_connected(xprt), sock->sk->sk_state); + if (status < 0) { + switch (status) { + case -EINPROGRESS: + case -EALREADY: + goto out_clear; + } + } +out: + if (status < 0) + rpc_wake_up_status(&xprt->pending, status); + else + rpc_wake_up(&xprt->pending); +out_clear: + smp_mb__before_clear_bit(); + clear_bit(XPRT_CONNECTING, &xprt->sockstate); + smp_mb__after_clear_bit(); +} + +static void +xprt_connect_sock(struct rpc_task *task) +{ + struct rpc_xprt *xprt = task->tk_xprt; + + if (!test_and_set_bit(XPRT_CONNECTING, &xprt->sockstate)) { + /* Note: if we are here due to a dropped connection + * we delay reconnecting by RPC_REESTABLISH_TIMEOUT/HZ + * seconds + */ + if (xprt->sock != NULL) + schedule_delayed_work(&xprt->sock_connect, + RPC_REESTABLISH_TIMEOUT); + else { + schedule_work(&xprt->sock_connect); + /* flush_scheduled_work can sleep... */ + if (!RPC_IS_ASYNC(task)) + flush_scheduled_work(); + } + } +} + +/* + * Set default timeout parameters + */ +static void +xprt_default_timeout(struct rpc_timeout *to, int proto) +{ + if (proto == IPPROTO_UDP) + xprt_set_timeout(to, 5, 5 * HZ); + else + xprt_set_timeout(to, 2, 60 * HZ); +} + +static struct rpc_xprt_ops xprt_socket_ops = { + .set_buffer_size = xprt_sock_setbufsize, + .connect = xprt_connect_sock, + .send_request = xprt_send_request, + .close = xprt_close, + .destroy = xprt_socket_destroy, +}; + +extern unsigned int xprt_udp_slot_table_entries; +extern unsigned int xprt_tcp_slot_table_entries; + +int xs_setup_udp(struct rpc_xprt *xprt, struct rpc_timeout *to) +{ + size_t slot_table_size; + + dprintk("RPC: setting up udp-ipv4 transport...\n"); + + xprt->max_reqs = xprt_udp_slot_table_entries; + slot_table_size = xprt->max_reqs * sizeof(xprt->slot[0]); + xprt->slot = kmalloc(slot_table_size, GFP_KERNEL); + if (xprt->slot == NULL) + return -ENOMEM; + memset(xprt->slot, 0, slot_table_size); + + xprt->prot = IPPROTO_UDP; + xprt->port = XPRT_MAX_RESVPORT; + xprt->stream = 0; + xprt->nocong = 0; + xprt->cwnd = RPC_INITCWND; + xprt->resvport = capable(CAP_NET_BIND_SERVICE) ? 1 : 0; + /* XXX: header size can vary due to auth type, IPv6, etc. */ + xprt->max_payload = (1U << 16) - (MAX_HEADER << 3); + + INIT_WORK(&xprt->sock_connect, xprt_socket_connect, xprt); + + xprt->ops = &xprt_socket_ops; + + if (to) + xprt->timeout = *to; + else + xprt_default_timeout(to, xprt->prot); + + return 0; +} + +int xs_setup_tcp(struct rpc_xprt *xprt, struct rpc_timeout *to) +{ + size_t slot_table_size; + + dprintk("RPC: setting up tcp-ipv4 transport...\n"); + + xprt->max_reqs = xprt_tcp_slot_table_entries; + slot_table_size = xprt->max_reqs * sizeof(xprt->slot[0]); + xprt->slot = kmalloc(slot_table_size, GFP_KERNEL); + if (xprt->slot == NULL) + return -ENOMEM; + memset(xprt->slot, 0, slot_table_size); + + xprt->prot = IPPROTO_TCP; + xprt->port = XPRT_MAX_RESVPORT; + xprt->stream = 1; + xprt->nocong = 1; + xprt->cwnd = RPC_MAXCWND(xprt); + xprt->resvport = capable(CAP_NET_BIND_SERVICE) ? 1 : 0; + xprt->max_payload = (1U << 31) - 1; + + INIT_WORK(&xprt->sock_connect, xprt_socket_connect, xprt); + + xprt->ops = &xprt_socket_ops; + + if (to) + xprt->timeout = *to; + else + xprt_default_timeout(to, xprt->prot); + + return 0; +} -- cgit v1.2.3 From 9903cd1c27a1f30e8efea75e125be3b2002f7cb9 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 11 Aug 2005 16:25:26 -0400 Subject: [PATCH] RPC: transport switch function naming Introduce block header comments and a function naming convention to the socket transport implementation. Provide a debug setting for transports that is separate from RPCDBG_XPRT. Eliminate xprt_default_timeout(). Provide block comments for exposed interfaces in xprt.c, and eliminate the useless obvious comments. Convert printk's to dprintk's. Test-plan: Compile kernel with CONFIG_NFS enabled. Version: Thu, 11 Aug 2005 16:04:04 -0400 Signed-off-by: Chuck Lever Signed-off-by: Trond Myklebust --- include/linux/sunrpc/debug.h | 1 + net/sunrpc/xprt.c | 147 +++++++------- net/sunrpc/xprtsock.c | 464 ++++++++++++++++++++++--------------------- 3 files changed, 311 insertions(+), 301 deletions(-) (limited to 'include/linux') diff --git a/include/linux/sunrpc/debug.h b/include/linux/sunrpc/debug.h index eadb31e3c198..42d299747956 100644 --- a/include/linux/sunrpc/debug.h +++ b/include/linux/sunrpc/debug.h @@ -32,6 +32,7 @@ #define RPCDBG_AUTH 0x0010 #define RPCDBG_PMAP 0x0020 #define RPCDBG_SCHED 0x0040 +#define RPCDBG_TRANS 0x0080 #define RPCDBG_SVCSOCK 0x0100 #define RPCDBG_SVCDSP 0x0200 #define RPCDBG_MISC 0x0400 diff --git a/net/sunrpc/xprt.c b/net/sunrpc/xprt.c index 4342acf4d1cd..589195e630ef 100644 --- a/net/sunrpc/xprt.c +++ b/net/sunrpc/xprt.c @@ -227,9 +227,6 @@ xprt_adjust_cwnd(struct rpc_xprt *xprt, int result) xprt->cwnd = cwnd; } -/* - * Reset the major timeout value - */ static void xprt_reset_majortimeo(struct rpc_rqst *req) { struct rpc_timeout *to = &req->rq_xprt->timeout; @@ -244,8 +241,10 @@ static void xprt_reset_majortimeo(struct rpc_rqst *req) req->rq_majortimeo += jiffies; } -/* - * Adjust timeout values etc for next retransmit +/** + * xprt_adjust_timeout - adjust timeout values for next retransmit + * @req: RPC request containing parameters to use for the adjustment + * */ int xprt_adjust_timeout(struct rpc_rqst *req) { @@ -291,8 +290,10 @@ xprt_socket_autoclose(void *args) xprt_release_write(xprt, NULL); } -/* - * Mark a transport as disconnected +/** + * xprt_disconnect - mark a transport as disconnected + * @xprt: transport to flag for disconnect + * */ void xprt_disconnect(struct rpc_xprt *xprt) { @@ -303,9 +304,6 @@ void xprt_disconnect(struct rpc_xprt *xprt) spin_unlock_bh(&xprt->sock_lock); } -/* - * Used to allow disconnection when we've been idle - */ static void xprt_init_autodisconnect(unsigned long data) { @@ -327,8 +325,9 @@ out_abort: spin_unlock(&xprt->sock_lock); } -/* - * Attempt to connect a TCP socket. +/** + * xprt_connect - schedule a transport connect operation + * @task: RPC task that is requesting the connect * */ void xprt_connect(struct rpc_task *task) @@ -361,11 +360,7 @@ void xprt_connect(struct rpc_task *task) return; } -/* - * We arrive here when awoken from waiting on connection establishment. - */ -static void -xprt_connect_status(struct rpc_task *task) +static void xprt_connect_status(struct rpc_task *task) { struct rpc_xprt *xprt = task->tk_xprt; @@ -404,8 +399,11 @@ xprt_connect_status(struct rpc_task *task) } } -/* - * Look up the RPC request corresponding to a reply, and then lock it. +/** + * xprt_lookup_rqst - find an RPC request corresponding to an XID + * @xprt: transport on which the original request was transmitted + * @xid: RPC XID of incoming reply + * */ struct rpc_rqst *xprt_lookup_rqst(struct rpc_xprt *xprt, u32 xid) { @@ -422,9 +420,12 @@ struct rpc_rqst *xprt_lookup_rqst(struct rpc_xprt *xprt, u32 xid) return req; } -/* - * Complete reply received. - * The TCP code relies on us to remove the request from xprt->pending. +/** + * xprt_complete_rqst - called when reply processing is complete + * @xprt: controlling transport + * @req: RPC request that just completed + * @copied: actual number of bytes received from the transport + * */ void xprt_complete_rqst(struct rpc_xprt *xprt, struct rpc_rqst *req, int copied) { @@ -498,12 +499,12 @@ out: spin_unlock(&xprt->sock_lock); } -/* - * Place the actual RPC call. - * We have to copy the iovec because sendmsg fiddles with its contents. +/** + * xprt_prepare_transmit - reserve the transport before sending a request + * @task: RPC task about to send a request + * */ -int -xprt_prepare_transmit(struct rpc_task *task) +int xprt_prepare_transmit(struct rpc_task *task) { struct rpc_rqst *req = task->tk_rqstp; struct rpc_xprt *xprt = req->rq_xprt; @@ -533,8 +534,13 @@ out_unlock: return err; } -void -xprt_transmit(struct rpc_task *task) +/** + * xprt_transmit - send an RPC request on a transport + * @task: controlling RPC task + * + * We have to copy the iovec because sendmsg fiddles with its contents. + */ +void xprt_transmit(struct rpc_task *task) { struct rpc_clnt *clnt = task->tk_client; struct rpc_rqst *req = task->tk_rqstp; @@ -604,11 +610,7 @@ xprt_transmit(struct rpc_task *task) spin_unlock_bh(&xprt->sock_lock); } -/* - * Reserve an RPC call slot. - */ -static inline void -do_xprt_reserve(struct rpc_task *task) +static inline void do_xprt_reserve(struct rpc_task *task) { struct rpc_xprt *xprt = task->tk_xprt; @@ -628,8 +630,14 @@ do_xprt_reserve(struct rpc_task *task) rpc_sleep_on(&xprt->backlog, task, NULL, NULL); } -void -xprt_reserve(struct rpc_task *task) +/** + * xprt_reserve - allocate an RPC request slot + * @task: RPC task requesting a slot allocation + * + * If no more slots are available, place the task on the transport's + * backlog queue. + */ +void xprt_reserve(struct rpc_task *task) { struct rpc_xprt *xprt = task->tk_xprt; @@ -641,9 +649,6 @@ xprt_reserve(struct rpc_task *task) } } -/* - * Allocate a 'unique' XID - */ static inline u32 xprt_alloc_xid(struct rpc_xprt *xprt) { return xprt->xid++; @@ -654,11 +659,7 @@ static inline void xprt_init_xid(struct rpc_xprt *xprt) get_random_bytes(&xprt->xid, sizeof(xprt->xid)); } -/* - * Initialize RPC request - */ -static void -xprt_request_init(struct rpc_task *task, struct rpc_xprt *xprt) +static void xprt_request_init(struct rpc_task *task, struct rpc_xprt *xprt) { struct rpc_rqst *req = task->tk_rqstp; @@ -670,11 +671,12 @@ xprt_request_init(struct rpc_task *task, struct rpc_xprt *xprt) req, ntohl(req->rq_xid)); } -/* - * Release an RPC call slot +/** + * xprt_release - release an RPC request slot + * @task: task which is finished with the slot + * */ -void -xprt_release(struct rpc_task *task) +void xprt_release(struct rpc_task *task) { struct rpc_xprt *xprt = task->tk_xprt; struct rpc_rqst *req; @@ -702,11 +704,14 @@ xprt_release(struct rpc_task *task) spin_unlock(&xprt->xprt_lock); } -/* - * Set constant timeout +/** + * xprt_set_timeout - set constant RPC timeout + * @to: RPC timeout parameters to set up + * @retr: number of retries + * @incr: amount of increase after each retry + * */ -void -xprt_set_timeout(struct rpc_timeout *to, unsigned int retr, unsigned long incr) +void xprt_set_timeout(struct rpc_timeout *to, unsigned int retr, unsigned long incr) { to->to_initval = to->to_increment = incr; @@ -715,11 +720,7 @@ xprt_set_timeout(struct rpc_timeout *to, unsigned int retr, unsigned long incr) to->to_exponential = 0; } -/* - * Initialize an RPC client - */ -static struct rpc_xprt * -xprt_setup(int proto, struct sockaddr_in *ap, struct rpc_timeout *to) +static struct rpc_xprt *xprt_setup(int proto, struct sockaddr_in *ap, struct rpc_timeout *to) { int result; struct rpc_xprt *xprt; @@ -778,11 +779,14 @@ xprt_setup(int proto, struct sockaddr_in *ap, struct rpc_timeout *to) return xprt; } -/* - * Create an RPC client transport given the protocol and peer address. +/** + * xprt_create_proto - create an RPC client transport + * @proto: requested transport protocol + * @sap: remote peer's address + * @to: timeout parameters for new transport + * */ -struct rpc_xprt * -xprt_create_proto(int proto, struct sockaddr_in *sap, struct rpc_timeout *to) +struct rpc_xprt *xprt_create_proto(int proto, struct sockaddr_in *sap, struct rpc_timeout *to) { struct rpc_xprt *xprt; @@ -794,11 +798,7 @@ xprt_create_proto(int proto, struct sockaddr_in *sap, struct rpc_timeout *to) return xprt; } -/* - * Prepare for transport shutdown. - */ -static void -xprt_shutdown(struct rpc_xprt *xprt) +static void xprt_shutdown(struct rpc_xprt *xprt) { xprt->shutdown = 1; rpc_wake_up(&xprt->sending); @@ -809,21 +809,18 @@ xprt_shutdown(struct rpc_xprt *xprt) del_timer_sync(&xprt->timer); } -/* - * Clear the xprt backlog queue - */ -static int -xprt_clear_backlog(struct rpc_xprt *xprt) { +static int xprt_clear_backlog(struct rpc_xprt *xprt) { rpc_wake_up_next(&xprt->backlog); wake_up(&xprt->cong_wait); return 1; } -/* - * Destroy an RPC transport, killing off all requests. +/** + * xprt_destroy - destroy an RPC transport, killing off all requests. + * @xprt: transport to destroy + * */ -int -xprt_destroy(struct rpc_xprt *xprt) +int xprt_destroy(struct rpc_xprt *xprt) { dprintk("RPC: destroying transport %p\n", xprt); xprt_shutdown(xprt); diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c index fa1180ac4823..80222de3afa4 100644 --- a/net/sunrpc/xprtsock.c +++ b/net/sunrpc/xprtsock.c @@ -33,23 +33,21 @@ #include #include +/* + * Maximum port number to use when requesting a reserved port. + */ +#define XS_MAX_RESVPORT (800U) + #ifdef RPC_DEBUG # undef RPC_DEBUG_DATA -# define RPCDBG_FACILITY RPCDBG_XPRT +# define RPCDBG_FACILITY RPCDBG_TRANS #endif -#define XPRT_MAX_RESVPORT (800) - #ifdef RPC_DEBUG_DATA -/* - * Print the buffer contents (first 128 bytes only--just enough for - * diropres return). - */ -static void -xprt_pktdump(char *msg, u32 *packet, unsigned int count) +static void xs_pktdump(char *msg, u32 *packet, unsigned int count) { - u8 *buf = (u8 *) packet; - int j; + u8 *buf = (u8 *) packet; + int j; dprintk("RPC: %s\n", msg); for (j = 0; j < count && j < 128; j += 4) { @@ -64,25 +62,22 @@ xprt_pktdump(char *msg, u32 *packet, unsigned int count) dprintk("\n"); } #else -static inline void -xprt_pktdump(char *msg, u32 *packet, unsigned int count) +static inline void xs_pktdump(char *msg, u32 *packet, unsigned int count) { /* NOP */ } #endif -/* - * Look up RPC transport given an INET socket +/** + * xs_sendpages - write pages directly to a socket + * @sock: socket to send on + * @addr: UDP only -- address of destination + * @addrlen: UDP only -- length of destination address + * @xdr: buffer containing this request + * @base: starting position in the buffer + * */ -static inline struct rpc_xprt * -xprt_from_sock(struct sock *sk) -{ - return (struct rpc_xprt *) sk->sk_user_data; -} - -static int -xdr_sendpages(struct socket *sock, struct sockaddr *addr, int addrlen, - struct xdr_buf *xdr, unsigned int base, int msgflags) +static int xs_sendpages(struct socket *sock, struct sockaddr *addr, int addrlen, struct xdr_buf *xdr, unsigned int base, int msgflags) { struct page **ppage = xdr->pages; unsigned int len, pglen = xdr->page_len; @@ -125,7 +120,7 @@ xdr_sendpages(struct socket *sock, struct sockaddr *addr, int addrlen, } if (base || xdr->page_base) { pglen -= base; - base += xdr->page_base; + base += xdr->page_base; ppage += base >> PAGE_CACHE_SHIFT; base &= ~PAGE_CACHE_MASK; } @@ -176,23 +171,25 @@ out: return ret; } -/* - * Write data to socket. +/** + * xs_sendmsg - write an RPC request to a socket + * @xprt: generic transport + * @req: the RPC request to write + * */ -static inline int -xprt_sendmsg(struct rpc_xprt *xprt, struct rpc_rqst *req) +static int xs_sendmsg(struct rpc_xprt *xprt, struct rpc_rqst *req) { - struct socket *sock = xprt->sock; - struct xdr_buf *xdr = &req->rq_snd_buf; + struct socket *sock = xprt->sock; + struct xdr_buf *xdr = &req->rq_snd_buf; struct sockaddr *addr = NULL; int addrlen = 0; - unsigned int skip; - int result; + unsigned int skip; + int result; if (!sock) return -ENOTCONN; - xprt_pktdump("packet data:", + xs_pktdump("packet data:", req->rq_svec->iov_base, req->rq_svec->iov_len); @@ -201,13 +198,13 @@ xprt_sendmsg(struct rpc_xprt *xprt, struct rpc_rqst *req) addr = (struct sockaddr *) &xprt->addr; addrlen = sizeof(xprt->addr); } - /* Dont repeat bytes */ + /* Don't repeat bytes */ skip = req->rq_bytes_sent; clear_bit(SOCK_ASYNC_NOSPACE, &sock->flags); - result = xdr_sendpages(sock, addr, addrlen, xdr, skip, MSG_DONTWAIT); + result = xs_sendpages(sock, addr, addrlen, xdr, skip, MSG_DONTWAIT); - dprintk("RPC: xprt_sendmsg(%d) = %d\n", xdr->len - skip, result); + dprintk("RPC: xs_sendmsg(%d) = %d\n", xdr->len - skip, result); if (result >= 0) return result; @@ -215,8 +212,7 @@ xprt_sendmsg(struct rpc_xprt *xprt, struct rpc_rqst *req) switch (result) { case -ECONNREFUSED: /* When the server has died, an ICMP port unreachable message - * prompts ECONNREFUSED. - */ + * prompts ECONNREFUSED. */ case -EAGAIN: break; case -ECONNRESET: @@ -227,13 +223,25 @@ xprt_sendmsg(struct rpc_xprt *xprt, struct rpc_rqst *req) result = -ENOTCONN; break; default: - printk(KERN_NOTICE "RPC: sendmsg returned error %d\n", -result); + break; } return result; } -static int -xprt_send_request(struct rpc_task *task) +/** + * xs_send_request - write an RPC request to a socket + * @task: address of RPC task that manages the state of an RPC request + * + * Return values: + * 0: The request has been sent + * EAGAIN: The socket was blocked, please call again later to + * complete the request + * other: Some other error occured, the request was not sent + * + * XXX: In the case of soft timeouts, should we eventually give up + * if the socket is not able to make progress? + */ +static int xs_send_request(struct rpc_task *task) { struct rpc_rqst *req = task->tk_rqstp; struct rpc_xprt *xprt = req->rq_xprt; @@ -242,18 +250,18 @@ xprt_send_request(struct rpc_task *task) /* set up everything as needed. */ /* Write the record marker */ if (xprt->stream) { - u32 *marker = req->rq_svec[0].iov_base; + u32 *marker = req->rq_svec[0].iov_base; *marker = htonl(0x80000000|(req->rq_slen-sizeof(*marker))); } /* Continue transmitting the packet/record. We must be careful * to cope with writespace callbacks arriving _after_ we have - * called xprt_sendmsg(). + * called sendmsg(). */ while (1) { req->rq_xtime = jiffies; - status = xprt_sendmsg(xprt, req); + status = xs_sendmsg(xprt, req); if (status < 0) break; @@ -285,7 +293,7 @@ xprt_send_request(struct rpc_task *task) if (status == -EAGAIN) { if (test_bit(SOCK_ASYNC_NOSPACE, &xprt->sock->flags)) { - /* Protect against races with xprt_write_space */ + /* Protect against races with xs_write_space */ spin_lock_bh(&xprt->sock_lock); /* Don't race with disconnect */ if (!xprt_connected(xprt)) @@ -303,65 +311,77 @@ xprt_send_request(struct rpc_task *task) return status; } -/* - * Close down a transport socket +/** + * xs_close - close a socket + * @xprt: transport + * */ -static void -xprt_close(struct rpc_xprt *xprt) +static void xs_close(struct rpc_xprt *xprt) { - struct socket *sock = xprt->sock; - struct sock *sk = xprt->inet; + struct socket *sock = xprt->sock; + struct sock *sk = xprt->inet; if (!sk) return; + dprintk("RPC: xs_close xprt %p\n", xprt); + write_lock_bh(&sk->sk_callback_lock); xprt->inet = NULL; xprt->sock = NULL; - sk->sk_user_data = NULL; - sk->sk_data_ready = xprt->old_data_ready; + sk->sk_user_data = NULL; + sk->sk_data_ready = xprt->old_data_ready; sk->sk_state_change = xprt->old_state_change; - sk->sk_write_space = xprt->old_write_space; + sk->sk_write_space = xprt->old_write_space; write_unlock_bh(&sk->sk_callback_lock); - sk->sk_no_check = 0; + sk->sk_no_check = 0; sock_release(sock); } -static void xprt_socket_destroy(struct rpc_xprt *xprt) +/** + * xs_destroy - prepare to shutdown a transport + * @xprt: doomed transport + * + */ +static void xs_destroy(struct rpc_xprt *xprt) { + dprintk("RPC: xs_destroy xprt %p\n", xprt); + cancel_delayed_work(&xprt->sock_connect); flush_scheduled_work(); xprt_disconnect(xprt); - xprt_close(xprt); + xs_close(xprt); kfree(xprt->slot); } -/* - * Input handler for RPC replies. Called from a bottom half and hence - * atomic. +static inline struct rpc_xprt *xprt_from_sock(struct sock *sk) +{ + return (struct rpc_xprt *) sk->sk_user_data; +} + +/** + * xs_udp_data_ready - "data ready" callback for UDP sockets + * @sk: socket with data to read + * @len: how much data to read + * */ -static void -udp_data_ready(struct sock *sk, int len) +static void xs_udp_data_ready(struct sock *sk, int len) { - struct rpc_task *task; - struct rpc_xprt *xprt; + struct rpc_task *task; + struct rpc_xprt *xprt; struct rpc_rqst *rovr; - struct sk_buff *skb; + struct sk_buff *skb; int err, repsize, copied; u32 _xid, *xp; read_lock(&sk->sk_callback_lock); - dprintk("RPC: udp_data_ready...\n"); - if (!(xprt = xprt_from_sock(sk))) { - printk("RPC: udp_data_ready request not found!\n"); + dprintk("RPC: xs_udp_data_ready...\n"); + if (!(xprt = xprt_from_sock(sk))) goto out; - } - - dprintk("RPC: udp_data_ready client %p\n", xprt); if ((skb = skb_recv_datagram(sk, 0, 1, &err)) == NULL) goto out; @@ -371,7 +391,7 @@ udp_data_ready(struct sock *sk, int len) repsize = skb->len - sizeof(struct udphdr); if (repsize < 4) { - printk("RPC: impossible RPC reply size %d!\n", repsize); + dprintk("RPC: impossible RPC reply size %d!\n", repsize); goto dropit; } @@ -410,11 +430,7 @@ udp_data_ready(struct sock *sk, int len) read_unlock(&sk->sk_callback_lock); } -/* - * Copy from an skb into memory and shrink the skb. - */ -static inline size_t -tcp_copy_data(skb_reader_t *desc, void *p, size_t len) +static inline size_t xs_tcp_copy_data(skb_reader_t *desc, void *p, size_t len) { if (len > desc->count) len = desc->count; @@ -430,18 +446,14 @@ tcp_copy_data(skb_reader_t *desc, void *p, size_t len) return len; } -/* - * TCP read fragment marker - */ -static inline void -tcp_read_fraghdr(struct rpc_xprt *xprt, skb_reader_t *desc) +static inline void xs_tcp_read_fraghdr(struct rpc_xprt *xprt, skb_reader_t *desc) { size_t len, used; char *p; p = ((char *) &xprt->tcp_recm) + xprt->tcp_offset; len = sizeof(xprt->tcp_recm) - xprt->tcp_offset; - used = tcp_copy_data(desc, p, len); + used = xs_tcp_copy_data(desc, p, len); xprt->tcp_offset += used; if (used != len) return; @@ -455,15 +467,15 @@ tcp_read_fraghdr(struct rpc_xprt *xprt, skb_reader_t *desc) xprt->tcp_offset = 0; /* Sanity check of the record length */ if (xprt->tcp_reclen < 4) { - printk(KERN_ERR "RPC: Invalid TCP record fragment length\n"); + dprintk("RPC: invalid TCP record fragment length\n"); xprt_disconnect(xprt); + return; } dprintk("RPC: reading TCP record fragment of length %d\n", xprt->tcp_reclen); } -static void -tcp_check_recm(struct rpc_xprt *xprt) +static void xs_tcp_check_recm(struct rpc_xprt *xprt) { dprintk("RPC: xprt = %p, tcp_copied = %lu, tcp_offset = %u, tcp_reclen = %u, tcp_flags = %lx\n", xprt, xprt->tcp_copied, xprt->tcp_offset, xprt->tcp_reclen, xprt->tcp_flags); @@ -478,11 +490,7 @@ tcp_check_recm(struct rpc_xprt *xprt) } } -/* - * TCP read xid - */ -static inline void -tcp_read_xid(struct rpc_xprt *xprt, skb_reader_t *desc) +static inline void xs_tcp_read_xid(struct rpc_xprt *xprt, skb_reader_t *desc) { size_t len, used; char *p; @@ -490,7 +498,7 @@ tcp_read_xid(struct rpc_xprt *xprt, skb_reader_t *desc) len = sizeof(xprt->tcp_xid) - xprt->tcp_offset; dprintk("RPC: reading XID (%Zu bytes)\n", len); p = ((char *) &xprt->tcp_xid) + xprt->tcp_offset; - used = tcp_copy_data(desc, p, len); + used = xs_tcp_copy_data(desc, p, len); xprt->tcp_offset += used; if (used != len) return; @@ -499,14 +507,10 @@ tcp_read_xid(struct rpc_xprt *xprt, skb_reader_t *desc) xprt->tcp_copied = 4; dprintk("RPC: reading reply for XID %08x\n", ntohl(xprt->tcp_xid)); - tcp_check_recm(xprt); + xs_tcp_check_recm(xprt); } -/* - * TCP read and complete request - */ -static inline void -tcp_read_request(struct rpc_xprt *xprt, skb_reader_t *desc) +static inline void xs_tcp_read_request(struct rpc_xprt *xprt, skb_reader_t *desc) { struct rpc_rqst *req; struct xdr_buf *rcvbuf; @@ -533,12 +537,12 @@ tcp_read_request(struct rpc_xprt *xprt, skb_reader_t *desc) memcpy(&my_desc, desc, sizeof(my_desc)); my_desc.count = len; r = xdr_partial_copy_from_skb(rcvbuf, xprt->tcp_copied, - &my_desc, tcp_copy_data); + &my_desc, xs_tcp_copy_data); desc->count -= r; desc->offset += r; } else r = xdr_partial_copy_from_skb(rcvbuf, xprt->tcp_copied, - desc, tcp_copy_data); + desc, xs_tcp_copy_data); if (r > 0) { xprt->tcp_copied += r; @@ -581,14 +585,10 @@ out: xprt_complete_rqst(xprt, req, xprt->tcp_copied); } spin_unlock(&xprt->sock_lock); - tcp_check_recm(xprt); + xs_tcp_check_recm(xprt); } -/* - * TCP discard extra bytes from a short read - */ -static inline void -tcp_read_discard(struct rpc_xprt *xprt, skb_reader_t *desc) +static inline void xs_tcp_read_discard(struct rpc_xprt *xprt, skb_reader_t *desc) { size_t len; @@ -599,16 +599,10 @@ tcp_read_discard(struct rpc_xprt *xprt, skb_reader_t *desc) desc->offset += len; xprt->tcp_offset += len; dprintk("RPC: discarded %Zu bytes\n", len); - tcp_check_recm(xprt); + xs_tcp_check_recm(xprt); } -/* - * TCP record receive routine - * We first have to grab the record marker, then the XID, then the data. - */ -static int -tcp_data_recv(read_descriptor_t *rd_desc, struct sk_buff *skb, - unsigned int offset, size_t len) +static int xs_tcp_data_recv(read_descriptor_t *rd_desc, struct sk_buff *skb, unsigned int offset, size_t len) { struct rpc_xprt *xprt = rd_desc->arg.data; skb_reader_t desc = { @@ -616,64 +610,72 @@ tcp_data_recv(read_descriptor_t *rd_desc, struct sk_buff *skb, .offset = offset, .count = len, .csum = 0 - }; + }; - dprintk("RPC: tcp_data_recv\n"); + dprintk("RPC: xs_tcp_data_recv started\n"); do { /* Read in a new fragment marker if necessary */ /* Can we ever really expect to get completely empty fragments? */ if (xprt->tcp_flags & XPRT_COPY_RECM) { - tcp_read_fraghdr(xprt, &desc); + xs_tcp_read_fraghdr(xprt, &desc); continue; } /* Read in the xid if necessary */ if (xprt->tcp_flags & XPRT_COPY_XID) { - tcp_read_xid(xprt, &desc); + xs_tcp_read_xid(xprt, &desc); continue; } /* Read in the request data */ if (xprt->tcp_flags & XPRT_COPY_DATA) { - tcp_read_request(xprt, &desc); + xs_tcp_read_request(xprt, &desc); continue; } /* Skip over any trailing bytes on short reads */ - tcp_read_discard(xprt, &desc); + xs_tcp_read_discard(xprt, &desc); } while (desc.count); - dprintk("RPC: tcp_data_recv done\n"); + dprintk("RPC: xs_tcp_data_recv done\n"); return len - desc.count; } -static void tcp_data_ready(struct sock *sk, int bytes) +/** + * xs_tcp_data_ready - "data ready" callback for TCP sockets + * @sk: socket with data to read + * @bytes: how much data to read + * + */ +static void xs_tcp_data_ready(struct sock *sk, int bytes) { struct rpc_xprt *xprt; read_descriptor_t rd_desc; read_lock(&sk->sk_callback_lock); - dprintk("RPC: tcp_data_ready...\n"); - if (!(xprt = xprt_from_sock(sk))) { - printk("RPC: tcp_data_ready socket info not found!\n"); + dprintk("RPC: xs_tcp_data_ready...\n"); + if (!(xprt = xprt_from_sock(sk))) goto out; - } if (xprt->shutdown) goto out; - /* We use rd_desc to pass struct xprt to tcp_data_recv */ + /* We use rd_desc to pass struct xprt to xs_tcp_data_recv */ rd_desc.arg.data = xprt; rd_desc.count = 65536; - tcp_read_sock(sk, &rd_desc, tcp_data_recv); + tcp_read_sock(sk, &rd_desc, xs_tcp_data_recv); out: read_unlock(&sk->sk_callback_lock); } -static void -tcp_state_change(struct sock *sk) +/** + * xs_tcp_state_change - callback to handle TCP socket state changes + * @sk: socket whose state has changed + * + */ +static void xs_tcp_state_change(struct sock *sk) { - struct rpc_xprt *xprt; + struct rpc_xprt *xprt; read_lock(&sk->sk_callback_lock); if (!(xprt = xprt_from_sock(sk))) goto out; - dprintk("RPC: tcp_state_change client %p...\n", xprt); + dprintk("RPC: xs_tcp_state_change client %p...\n", xprt); dprintk("RPC: state %x conn %d dead %d zapped %d\n", sk->sk_state, xprt_connected(xprt), sock_flag(sk, SOCK_DEAD), @@ -703,17 +705,20 @@ tcp_state_change(struct sock *sk) read_unlock(&sk->sk_callback_lock); } -/* +/** + * xs_write_space - callback invoked when socket buffer space becomes + * available + * @sk: socket whose state has changed + * * Called when more output buffer space is available for this socket. * We try not to wake our writers until they can make "significant" * progress, otherwise we'll waste resources thrashing sock_sendmsg * with a bunch of small requests. */ -static void -xprt_write_space(struct sock *sk) +static void xs_write_space(struct sock *sk) { - struct rpc_xprt *xprt; - struct socket *sock; + struct rpc_xprt *xprt; + struct socket *sock; read_lock(&sk->sk_callback_lock); if (!(xprt = xprt_from_sock(sk)) || !(sock = sk->sk_socket)) @@ -743,11 +748,15 @@ out: read_unlock(&sk->sk_callback_lock); } -/* - * Set socket buffer length +/** + * xs_set_buffer_size - set send and receive limits + * @xprt: generic transport + * + * Set socket send and receive limits based on the + * sndsize and rcvsize fields in the generic transport + * structure. This applies only to UDP sockets. */ -static void -xprt_sock_setbufsize(struct rpc_xprt *xprt) +static void xs_set_buffer_size(struct rpc_xprt *xprt) { struct sock *sk = xprt->inet; @@ -764,15 +773,12 @@ xprt_sock_setbufsize(struct rpc_xprt *xprt) } } -/* - * Bind to a reserved port - */ -static inline int xprt_bindresvport(struct rpc_xprt *xprt, struct socket *sock) +static int xs_bindresvport(struct rpc_xprt *xprt, struct socket *sock) { struct sockaddr_in myaddr = { .sin_family = AF_INET, }; - int err, port; + int err, port; /* Were we already bound to a given port? Try to reuse it */ port = xprt->port; @@ -782,20 +788,47 @@ static inline int xprt_bindresvport(struct rpc_xprt *xprt, struct socket *sock) sizeof(myaddr)); if (err == 0) { xprt->port = port; + dprintk("RPC: xs_bindresvport bound to port %u\n", + port); return 0; } if (--port == 0) - port = XPRT_MAX_RESVPORT; + port = XS_MAX_RESVPORT; } while (err == -EADDRINUSE && port != xprt->port); - printk("RPC: Can't bind to reserved port (%d).\n", -err); + dprintk("RPC: can't bind to reserved port (%d).\n", -err); return err; } -static void -xprt_bind_socket(struct rpc_xprt *xprt, struct socket *sock) +static struct socket *xs_create(struct rpc_xprt *xprt, int proto, int resvport) { - struct sock *sk = sock->sk; + struct socket *sock; + int type, err; + + dprintk("RPC: xs_create(%s %d)\n", + (proto == IPPROTO_UDP)? "udp" : "tcp", proto); + + type = (proto == IPPROTO_UDP)? SOCK_DGRAM : SOCK_STREAM; + + if ((err = sock_create_kern(PF_INET, type, proto, &sock)) < 0) { + dprintk("RPC: can't create socket (%d).\n", -err); + return NULL; + } + + /* If the caller has the capability, bind to a reserved port */ + if (resvport && xs_bindresvport(xprt, sock) < 0) + goto failed; + + return sock; + +failed: + sock_release(sock); + return NULL; +} + +static void xs_bind(struct rpc_xprt *xprt, struct socket *sock) +{ + struct sock *sk = sock->sk; if (xprt->inet) return; @@ -806,16 +839,16 @@ xprt_bind_socket(struct rpc_xprt *xprt, struct socket *sock) xprt->old_state_change = sk->sk_state_change; xprt->old_write_space = sk->sk_write_space; if (xprt->prot == IPPROTO_UDP) { - sk->sk_data_ready = udp_data_ready; + sk->sk_data_ready = xs_udp_data_ready; sk->sk_no_check = UDP_CSUM_NORCV; xprt_set_connected(xprt); } else { tcp_sk(sk)->nonagle = 1; /* disable Nagle's algorithm */ - sk->sk_data_ready = tcp_data_ready; - sk->sk_state_change = tcp_state_change; + sk->sk_data_ready = xs_tcp_data_ready; + sk->sk_state_change = xs_tcp_state_change; xprt_clear_connected(xprt); } - sk->sk_write_space = xprt_write_space; + sk->sk_write_space = xs_write_space; /* Reset to new socket */ xprt->sock = sock; @@ -825,39 +858,13 @@ xprt_bind_socket(struct rpc_xprt *xprt, struct socket *sock) return; } -/* - * Datastream sockets are created here, but xprt_connect will create - * and connect stream sockets. +/** + * xs_connect_worker - try to connect a socket to a remote endpoint + * @args: RPC transport to connect + * + * Invoked by a work queue tasklet. */ -static struct socket * xprt_create_socket(struct rpc_xprt *xprt, int proto, int resvport) -{ - struct socket *sock; - int type, err; - - dprintk("RPC: xprt_create_socket(%s %d)\n", - (proto == IPPROTO_UDP)? "udp" : "tcp", proto); - - type = (proto == IPPROTO_UDP)? SOCK_DGRAM : SOCK_STREAM; - - if ((err = sock_create_kern(PF_INET, type, proto, &sock)) < 0) { - printk("RPC: can't create socket (%d).\n", -err); - return NULL; - } - - /* If the caller has the capability, bind to a reserved port */ - if (resvport && xprt_bindresvport(xprt, sock) < 0) { - printk("RPC: can't bind to reserved port.\n"); - goto failed; - } - - return sock; - -failed: - sock_release(sock); - return NULL; -} - -static void xprt_socket_connect(void *args) +static void xs_connect_worker(void *args) { struct rpc_xprt *xprt = (struct rpc_xprt *)args; struct socket *sock = xprt->sock; @@ -866,18 +873,20 @@ static void xprt_socket_connect(void *args) if (xprt->shutdown || xprt->addr.sin_port == 0) goto out; + dprintk("RPC: xs_connect_worker xprt %p\n", xprt); + /* * Start by resetting any existing state */ - xprt_close(xprt); - sock = xprt_create_socket(xprt, xprt->prot, xprt->resvport); + xs_close(xprt); + sock = xs_create(xprt, xprt->prot, xprt->resvport); if (sock == NULL) { /* couldn't create socket or bind to reserved port; * this is likely a permanent error, so cause an abort */ goto out; } - xprt_bind_socket(xprt, sock); - xprt_sock_setbufsize(xprt); + xs_bind(xprt, sock); + xs_set_buffer_size(xprt); status = 0; if (!xprt->stream) @@ -908,20 +917,23 @@ out_clear: smp_mb__after_clear_bit(); } -static void -xprt_connect_sock(struct rpc_task *task) +/** + * xs_connect - connect a socket to a remote endpoint + * @task: address of RPC task that manages state of connect request + * + * TCP: If the remote end dropped the connection, delay reconnecting. + */ +static void xs_connect(struct rpc_task *task) { struct rpc_xprt *xprt = task->tk_xprt; if (!test_and_set_bit(XPRT_CONNECTING, &xprt->sockstate)) { - /* Note: if we are here due to a dropped connection - * we delay reconnecting by RPC_REESTABLISH_TIMEOUT/HZ - * seconds - */ - if (xprt->sock != NULL) + if (xprt->sock != NULL) { + dprintk("RPC: xs_connect delayed xprt %p\n", xprt); schedule_delayed_work(&xprt->sock_connect, RPC_REESTABLISH_TIMEOUT); - else { + } else { + dprintk("RPC: xs_connect scheduled xprt %p\n", xprt); schedule_work(&xprt->sock_connect); /* flush_scheduled_work can sleep... */ if (!RPC_IS_ASYNC(task)) @@ -930,29 +942,23 @@ xprt_connect_sock(struct rpc_task *task) } } -/* - * Set default timeout parameters - */ -static void -xprt_default_timeout(struct rpc_timeout *to, int proto) -{ - if (proto == IPPROTO_UDP) - xprt_set_timeout(to, 5, 5 * HZ); - else - xprt_set_timeout(to, 2, 60 * HZ); -} - -static struct rpc_xprt_ops xprt_socket_ops = { - .set_buffer_size = xprt_sock_setbufsize, - .connect = xprt_connect_sock, - .send_request = xprt_send_request, - .close = xprt_close, - .destroy = xprt_socket_destroy, +static struct rpc_xprt_ops xs_ops = { + .set_buffer_size = xs_set_buffer_size, + .connect = xs_connect, + .send_request = xs_send_request, + .close = xs_close, + .destroy = xs_destroy, }; extern unsigned int xprt_udp_slot_table_entries; extern unsigned int xprt_tcp_slot_table_entries; +/** + * xs_setup_udp - Set up transport to use a UDP socket + * @xprt: transport to set up + * @to: timeout parameters + * + */ int xs_setup_udp(struct rpc_xprt *xprt, struct rpc_timeout *to) { size_t slot_table_size; @@ -967,7 +973,7 @@ int xs_setup_udp(struct rpc_xprt *xprt, struct rpc_timeout *to) memset(xprt->slot, 0, slot_table_size); xprt->prot = IPPROTO_UDP; - xprt->port = XPRT_MAX_RESVPORT; + xprt->port = XS_MAX_RESVPORT; xprt->stream = 0; xprt->nocong = 0; xprt->cwnd = RPC_INITCWND; @@ -975,18 +981,24 @@ int xs_setup_udp(struct rpc_xprt *xprt, struct rpc_timeout *to) /* XXX: header size can vary due to auth type, IPv6, etc. */ xprt->max_payload = (1U << 16) - (MAX_HEADER << 3); - INIT_WORK(&xprt->sock_connect, xprt_socket_connect, xprt); + INIT_WORK(&xprt->sock_connect, xs_connect_worker, xprt); - xprt->ops = &xprt_socket_ops; + xprt->ops = &xs_ops; if (to) xprt->timeout = *to; else - xprt_default_timeout(to, xprt->prot); + xprt_set_timeout(&xprt->timeout, 5, 5 * HZ); return 0; } +/** + * xs_setup_tcp - Set up transport to use a TCP socket + * @xprt: transport to set up + * @to: timeout parameters + * + */ int xs_setup_tcp(struct rpc_xprt *xprt, struct rpc_timeout *to) { size_t slot_table_size; @@ -1001,21 +1013,21 @@ int xs_setup_tcp(struct rpc_xprt *xprt, struct rpc_timeout *to) memset(xprt->slot, 0, slot_table_size); xprt->prot = IPPROTO_TCP; - xprt->port = XPRT_MAX_RESVPORT; + xprt->port = XS_MAX_RESVPORT; xprt->stream = 1; xprt->nocong = 1; xprt->cwnd = RPC_MAXCWND(xprt); xprt->resvport = capable(CAP_NET_BIND_SERVICE) ? 1 : 0; xprt->max_payload = (1U << 31) - 1; - INIT_WORK(&xprt->sock_connect, xprt_socket_connect, xprt); + INIT_WORK(&xprt->sock_connect, xs_connect_worker, xprt); - xprt->ops = &xprt_socket_ops; + xprt->ops = &xs_ops; if (to) xprt->timeout = *to; else - xprt_default_timeout(to, xprt->prot); + xprt_set_timeout(&xprt->timeout, 2, 60 * HZ); return 0; } -- cgit v1.2.3 From 4a0f8c04f2ece949d54a0c4fd7490259cf23a58a Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 11 Aug 2005 16:25:32 -0400 Subject: [PATCH] RPC: Rename sock_lock Clean-up: replace a name reference to sockets in the generic parts of the RPC client by renaming sock_lock in the rpc_xprt structure. Test-plan: Compile kernel with CONFIG_NFS enabled. Version: Thu, 11 Aug 2005 16:05:00 -0400 Signed-off-by: Chuck Lever Signed-off-by: Trond Myklebust --- include/linux/sunrpc/xprt.h | 2 +- net/sunrpc/xprt.c | 44 ++++++++++++++++++++++---------------------- net/sunrpc/xprtsock.c | 22 +++++++++++----------- 3 files changed, 34 insertions(+), 34 deletions(-) (limited to 'include/linux') diff --git a/include/linux/sunrpc/xprt.h b/include/linux/sunrpc/xprt.h index d82b47ab73cb..c4f903f0e17c 100644 --- a/include/linux/sunrpc/xprt.h +++ b/include/linux/sunrpc/xprt.h @@ -198,7 +198,7 @@ struct rpc_xprt { /* * Send stuff */ - spinlock_t sock_lock; /* lock socket info */ + spinlock_t transport_lock; /* lock transport info */ spinlock_t xprt_lock; /* lock xprt info */ struct rpc_task * snd_task; /* Task blocked in send */ diff --git a/net/sunrpc/xprt.c b/net/sunrpc/xprt.c index 589195e630ef..1f0da8c1a3b0 100644 --- a/net/sunrpc/xprt.c +++ b/net/sunrpc/xprt.c @@ -106,9 +106,9 @@ xprt_lock_write(struct rpc_xprt *xprt, struct rpc_task *task) { int retval; - spin_lock_bh(&xprt->sock_lock); + spin_lock_bh(&xprt->transport_lock); retval = __xprt_lock_write(xprt, task); - spin_unlock_bh(&xprt->sock_lock); + spin_unlock_bh(&xprt->transport_lock); return retval; } @@ -161,9 +161,9 @@ __xprt_release_write(struct rpc_xprt *xprt, struct rpc_task *task) static inline void xprt_release_write(struct rpc_xprt *xprt, struct rpc_task *task) { - spin_lock_bh(&xprt->sock_lock); + spin_lock_bh(&xprt->transport_lock); __xprt_release_write(xprt, task); - spin_unlock_bh(&xprt->sock_lock); + spin_unlock_bh(&xprt->transport_lock); } /* @@ -266,9 +266,9 @@ int xprt_adjust_timeout(struct rpc_rqst *req) req->rq_retries = 0; xprt_reset_majortimeo(req); /* Reset the RTT counters == "slow start" */ - spin_lock_bh(&xprt->sock_lock); + spin_lock_bh(&xprt->transport_lock); rpc_init_rtt(req->rq_task->tk_client->cl_rtt, to->to_initval); - spin_unlock_bh(&xprt->sock_lock); + spin_unlock_bh(&xprt->transport_lock); pprintk("RPC: %lu timeout\n", jiffies); status = -ETIMEDOUT; } @@ -298,10 +298,10 @@ xprt_socket_autoclose(void *args) void xprt_disconnect(struct rpc_xprt *xprt) { dprintk("RPC: disconnected transport %p\n", xprt); - spin_lock_bh(&xprt->sock_lock); + spin_lock_bh(&xprt->transport_lock); xprt_clear_connected(xprt); rpc_wake_up_status(&xprt->pending, -ENOTCONN); - spin_unlock_bh(&xprt->sock_lock); + spin_unlock_bh(&xprt->transport_lock); } static void @@ -309,12 +309,12 @@ xprt_init_autodisconnect(unsigned long data) { struct rpc_xprt *xprt = (struct rpc_xprt *)data; - spin_lock(&xprt->sock_lock); + spin_lock(&xprt->transport_lock); if (!list_empty(&xprt->recv) || xprt->shutdown) goto out_abort; if (test_and_set_bit(XPRT_LOCKED, &xprt->sockstate)) goto out_abort; - spin_unlock(&xprt->sock_lock); + spin_unlock(&xprt->transport_lock); /* Let keventd close the socket */ if (test_bit(XPRT_CONNECTING, &xprt->sockstate) != 0) xprt_release_write(xprt, NULL); @@ -322,7 +322,7 @@ xprt_init_autodisconnect(unsigned long data) schedule_work(&xprt->task_cleanup); return; out_abort: - spin_unlock(&xprt->sock_lock); + spin_unlock(&xprt->transport_lock); } /** @@ -482,7 +482,7 @@ xprt_timer(struct rpc_task *task) struct rpc_rqst *req = task->tk_rqstp; struct rpc_xprt *xprt = req->rq_xprt; - spin_lock(&xprt->sock_lock); + spin_lock(&xprt->transport_lock); if (req->rq_received) goto out; @@ -496,7 +496,7 @@ xprt_timer(struct rpc_task *task) out: task->tk_timeout = 0; rpc_wake_up_task(task); - spin_unlock(&xprt->sock_lock); + spin_unlock(&xprt->transport_lock); } /** @@ -515,7 +515,7 @@ int xprt_prepare_transmit(struct rpc_task *task) if (xprt->shutdown) return -EIO; - spin_lock_bh(&xprt->sock_lock); + spin_lock_bh(&xprt->transport_lock); if (req->rq_received && !req->rq_bytes_sent) { err = req->rq_received; goto out_unlock; @@ -530,7 +530,7 @@ int xprt_prepare_transmit(struct rpc_task *task) goto out_unlock; } out_unlock: - spin_unlock_bh(&xprt->sock_lock); + spin_unlock_bh(&xprt->transport_lock); return err; } @@ -552,13 +552,13 @@ void xprt_transmit(struct rpc_task *task) smp_rmb(); if (!req->rq_received) { if (list_empty(&req->rq_list)) { - spin_lock_bh(&xprt->sock_lock); + spin_lock_bh(&xprt->transport_lock); /* Update the softirq receive buffer */ memcpy(&req->rq_private_buf, &req->rq_rcv_buf, sizeof(req->rq_private_buf)); /* Add request to the receive list */ list_add_tail(&req->rq_list, &xprt->recv); - spin_unlock_bh(&xprt->sock_lock); + spin_unlock_bh(&xprt->transport_lock); xprt_reset_majortimeo(req); /* Turn off autodisconnect */ del_singleshot_timer_sync(&xprt->timer); @@ -592,7 +592,7 @@ void xprt_transmit(struct rpc_task *task) out_receive: dprintk("RPC: %4d xmit complete\n", task->tk_pid); /* Set the task's receive timeout value */ - spin_lock_bh(&xprt->sock_lock); + spin_lock_bh(&xprt->transport_lock); if (!xprt->nocong) { int timer = task->tk_msg.rpc_proc->p_timer; task->tk_timeout = rpc_calc_rto(clnt->cl_rtt, timer); @@ -607,7 +607,7 @@ void xprt_transmit(struct rpc_task *task) else if (!req->rq_received) rpc_sleep_on(&xprt->pending, task, NULL, xprt_timer); __xprt_release_write(xprt, task); - spin_unlock_bh(&xprt->sock_lock); + spin_unlock_bh(&xprt->transport_lock); } static inline void do_xprt_reserve(struct rpc_task *task) @@ -683,7 +683,7 @@ void xprt_release(struct rpc_task *task) if (!(req = task->tk_rqstp)) return; - spin_lock_bh(&xprt->sock_lock); + spin_lock_bh(&xprt->transport_lock); __xprt_release_write(xprt, task); __xprt_put_cong(xprt, req); if (!list_empty(&req->rq_list)) @@ -692,7 +692,7 @@ void xprt_release(struct rpc_task *task) if (list_empty(&xprt->recv) && !xprt->shutdown) mod_timer(&xprt->timer, xprt->last_used + RPC_IDLE_DISCONNECT_TIMEOUT); - spin_unlock_bh(&xprt->sock_lock); + spin_unlock_bh(&xprt->transport_lock); task->tk_rqstp = NULL; memset(req, 0, sizeof(*req)); /* mark unused */ @@ -750,7 +750,7 @@ static struct rpc_xprt *xprt_setup(int proto, struct sockaddr_in *ap, struct rpc return ERR_PTR(result); } - spin_lock_init(&xprt->sock_lock); + spin_lock_init(&xprt->transport_lock); spin_lock_init(&xprt->xprt_lock); init_waitqueue_head(&xprt->cong_wait); diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c index a5a04203a6b0..bc90caab6088 100644 --- a/net/sunrpc/xprtsock.c +++ b/net/sunrpc/xprtsock.c @@ -307,7 +307,7 @@ static int xs_send_request(struct rpc_task *task) if (status == -EAGAIN) { if (test_bit(SOCK_ASYNC_NOSPACE, &xprt->sock->flags)) { /* Protect against races with xs_write_space */ - spin_lock_bh(&xprt->sock_lock); + spin_lock_bh(&xprt->transport_lock); /* Don't race with disconnect */ if (!xprt_connected(xprt)) task->tk_status = -ENOTCONN; @@ -315,7 +315,7 @@ static int xs_send_request(struct rpc_task *task) task->tk_timeout = req->rq_timeout; rpc_sleep_on(&xprt->pending, task, NULL, NULL); } - spin_unlock_bh(&xprt->sock_lock); + spin_unlock_bh(&xprt->transport_lock); return status; } /* Keep holding the socket if it is blocked */ @@ -415,7 +415,7 @@ static void xs_udp_data_ready(struct sock *sk, int len) goto dropit; /* Look up and lock the request corresponding to the given XID */ - spin_lock(&xprt->sock_lock); + spin_lock(&xprt->transport_lock); rovr = xprt_lookup_rqst(xprt, *xp); if (!rovr) goto out_unlock; @@ -436,7 +436,7 @@ static void xs_udp_data_ready(struct sock *sk, int len) xprt_complete_rqst(xprt, rovr, copied); out_unlock: - spin_unlock(&xprt->sock_lock); + spin_unlock(&xprt->transport_lock); dropit: skb_free_datagram(sk, skb); out: @@ -531,13 +531,13 @@ static inline void xs_tcp_read_request(struct rpc_xprt *xprt, skb_reader_t *desc ssize_t r; /* Find and lock the request corresponding to this xid */ - spin_lock(&xprt->sock_lock); + spin_lock(&xprt->transport_lock); req = xprt_lookup_rqst(xprt, xprt->tcp_xid); if (!req) { xprt->tcp_flags &= ~XPRT_COPY_DATA; dprintk("RPC: XID %08x request not found!\n", ntohl(xprt->tcp_xid)); - spin_unlock(&xprt->sock_lock); + spin_unlock(&xprt->transport_lock); return; } @@ -597,7 +597,7 @@ out: req->rq_task->tk_pid); xprt_complete_rqst(xprt, req, xprt->tcp_copied); } - spin_unlock(&xprt->sock_lock); + spin_unlock(&xprt->transport_lock); xs_tcp_check_recm(xprt); } @@ -696,7 +696,7 @@ static void xs_tcp_state_change(struct sock *sk) switch (sk->sk_state) { case TCP_ESTABLISHED: - spin_lock_bh(&xprt->sock_lock); + spin_lock_bh(&xprt->transport_lock); if (!xprt_test_and_set_connected(xprt)) { /* Reset TCP record info */ xprt->tcp_offset = 0; @@ -705,7 +705,7 @@ static void xs_tcp_state_change(struct sock *sk) xprt->tcp_flags = XPRT_COPY_RECM | XPRT_COPY_XID; rpc_wake_up(&xprt->pending); } - spin_unlock_bh(&xprt->sock_lock); + spin_unlock_bh(&xprt->transport_lock); break; case TCP_SYN_SENT: case TCP_SYN_RECV: @@ -753,10 +753,10 @@ static void xs_write_space(struct sock *sk) if (!test_and_clear_bit(SOCK_NOSPACE, &sock->flags)) goto out; - spin_lock_bh(&xprt->sock_lock); + spin_lock_bh(&xprt->transport_lock); if (xprt->snd_task) rpc_wake_up_task(xprt->snd_task); - spin_unlock_bh(&xprt->sock_lock); + spin_unlock_bh(&xprt->transport_lock); out: read_unlock(&sk->sk_callback_lock); } -- cgit v1.2.3 From 5dc07727f86b25851e95193a0c484ea21b531c47 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 11 Aug 2005 16:25:35 -0400 Subject: [PATCH] RPC: Rename xprt_lock Clean-up: Replace the xprt_lock with something more aptly named. This lock single-threads the XID and request slot reservation process. Test-plan: Compile kernel with CONFIG_NFS enabled. Version: Thu, 11 Aug 2005 16:05:26 -0400 Signed-off-by: Chuck Lever Signed-off-by: Trond Myklebust --- include/linux/sunrpc/xprt.h | 2 +- net/sunrpc/xprt.c | 10 +++++----- 2 files changed, 6 insertions(+), 6 deletions(-) (limited to 'include/linux') diff --git a/include/linux/sunrpc/xprt.h b/include/linux/sunrpc/xprt.h index c4f903f0e17c..41ce296dded1 100644 --- a/include/linux/sunrpc/xprt.h +++ b/include/linux/sunrpc/xprt.h @@ -199,7 +199,7 @@ struct rpc_xprt { * Send stuff */ spinlock_t transport_lock; /* lock transport info */ - spinlock_t xprt_lock; /* lock xprt info */ + spinlock_t reserve_lock; /* lock slot table */ struct rpc_task * snd_task; /* Task blocked in send */ struct list_head recv; diff --git a/net/sunrpc/xprt.c b/net/sunrpc/xprt.c index 1f0da8c1a3b0..9c45c522e3ef 100644 --- a/net/sunrpc/xprt.c +++ b/net/sunrpc/xprt.c @@ -643,9 +643,9 @@ void xprt_reserve(struct rpc_task *task) task->tk_status = -EIO; if (!xprt->shutdown) { - spin_lock(&xprt->xprt_lock); + spin_lock(&xprt->reserve_lock); do_xprt_reserve(task); - spin_unlock(&xprt->xprt_lock); + spin_unlock(&xprt->reserve_lock); } } @@ -698,10 +698,10 @@ void xprt_release(struct rpc_task *task) dprintk("RPC: %4d release request %p\n", task->tk_pid, req); - spin_lock(&xprt->xprt_lock); + spin_lock(&xprt->reserve_lock); list_add(&req->rq_list, &xprt->free); xprt_clear_backlog(xprt); - spin_unlock(&xprt->xprt_lock); + spin_unlock(&xprt->reserve_lock); } /** @@ -751,7 +751,7 @@ static struct rpc_xprt *xprt_setup(int proto, struct sockaddr_in *ap, struct rpc } spin_lock_init(&xprt->transport_lock); - spin_lock_init(&xprt->xprt_lock); + spin_lock_init(&xprt->reserve_lock); init_waitqueue_head(&xprt->cong_wait); INIT_LIST_HEAD(&xprt->free); -- cgit v1.2.3 From 2226feb6bcd0e5e117a9be3ea3dd3ffc14f3e41e Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 11 Aug 2005 16:25:38 -0400 Subject: [PATCH] RPC: rename the sockstate field Clean-up: get rid of a name reference to sockets in the generic parts of the RPC client by renaming the sockstate field in the rpc_xprt structure. Test-plan: Compile kernel with CONFIG_NFS enabled. Version: Thu, 11 Aug 2005 16:05:53 -0400 Signed-off-by: Chuck Lever Signed-off-by: Trond Myklebust --- include/linux/sunrpc/xprt.h | 60 ++++++++++++++++++++++++++++++++++++--------- net/sunrpc/xprt.c | 14 +++++------ net/sunrpc/xprtsock.c | 6 ++--- 3 files changed, 58 insertions(+), 22 deletions(-) (limited to 'include/linux') diff --git a/include/linux/sunrpc/xprt.h b/include/linux/sunrpc/xprt.h index 41ce296dded1..009a3bb4f997 100644 --- a/include/linux/sunrpc/xprt.h +++ b/include/linux/sunrpc/xprt.h @@ -163,7 +163,7 @@ struct rpc_xprt { struct list_head free; /* free slots */ struct rpc_rqst * slot; /* slot table storage */ unsigned int max_reqs; /* total slots */ - unsigned long sockstate; /* Socket state */ + unsigned long state; /* transport state */ unsigned char shutdown : 1, /* being shut down */ nocong : 1, /* no congestion control */ resvport : 1, /* use a reserved port */ @@ -240,16 +240,54 @@ int xs_setup_udp(struct rpc_xprt *, int xs_setup_tcp(struct rpc_xprt *, struct rpc_timeout *); -#define XPRT_LOCKED 0 -#define XPRT_CONNECT 1 -#define XPRT_CONNECTING 2 - -#define xprt_connected(xp) (test_bit(XPRT_CONNECT, &(xp)->sockstate)) -#define xprt_set_connected(xp) (set_bit(XPRT_CONNECT, &(xp)->sockstate)) -#define xprt_test_and_set_connected(xp) (test_and_set_bit(XPRT_CONNECT, &(xp)->sockstate)) -#define xprt_test_and_clear_connected(xp) \ - (test_and_clear_bit(XPRT_CONNECT, &(xp)->sockstate)) -#define xprt_clear_connected(xp) (clear_bit(XPRT_CONNECT, &(xp)->sockstate)) +/* + * Reserved bit positions in xprt->state + */ +#define XPRT_LOCKED (0) +#define XPRT_CONNECTED (1) +#define XPRT_CONNECTING (2) + +static inline void xprt_set_connected(struct rpc_xprt *xprt) +{ + set_bit(XPRT_CONNECTED, &xprt->state); +} + +static inline void xprt_clear_connected(struct rpc_xprt *xprt) +{ + clear_bit(XPRT_CONNECTED, &xprt->state); +} + +static inline int xprt_connected(struct rpc_xprt *xprt) +{ + return test_bit(XPRT_CONNECTED, &xprt->state); +} + +static inline int xprt_test_and_set_connected(struct rpc_xprt *xprt) +{ + return test_and_set_bit(XPRT_CONNECTED, &xprt->state); +} + +static inline int xprt_test_and_clear_connected(struct rpc_xprt *xprt) +{ + return test_and_clear_bit(XPRT_CONNECTED, &xprt->state); +} + +static inline void xprt_clear_connecting(struct rpc_xprt *xprt) +{ + smp_mb__before_clear_bit(); + clear_bit(XPRT_CONNECTING, &xprt->state); + smp_mb__after_clear_bit(); +} + +static inline int xprt_connecting(struct rpc_xprt *xprt) +{ + return test_bit(XPRT_CONNECTING, &xprt->state); +} + +static inline int xprt_test_and_set_connecting(struct rpc_xprt *xprt) +{ + return test_and_set_bit(XPRT_CONNECTING, &xprt->state); +} #endif /* __KERNEL__*/ diff --git a/net/sunrpc/xprt.c b/net/sunrpc/xprt.c index 9c45c522e3ef..57c5e77b155e 100644 --- a/net/sunrpc/xprt.c +++ b/net/sunrpc/xprt.c @@ -74,7 +74,7 @@ __xprt_lock_write(struct rpc_xprt *xprt, struct rpc_task *task) { struct rpc_rqst *req = task->tk_rqstp; - if (test_and_set_bit(XPRT_LOCKED, &xprt->sockstate)) { + if (test_and_set_bit(XPRT_LOCKED, &xprt->state)) { if (task == xprt->snd_task) return 1; goto out_sleep; @@ -88,7 +88,7 @@ __xprt_lock_write(struct rpc_xprt *xprt, struct rpc_task *task) return 1; } smp_mb__before_clear_bit(); - clear_bit(XPRT_LOCKED, &xprt->sockstate); + clear_bit(XPRT_LOCKED, &xprt->state); smp_mb__after_clear_bit(); out_sleep: dprintk("RPC: %4d failed to lock socket %p\n", task->tk_pid, xprt); @@ -118,7 +118,7 @@ __xprt_lock_write_next(struct rpc_xprt *xprt) { struct rpc_task *task; - if (test_and_set_bit(XPRT_LOCKED, &xprt->sockstate)) + if (test_and_set_bit(XPRT_LOCKED, &xprt->state)) return; if (!xprt->nocong && RPCXPRT_CONGESTED(xprt)) goto out_unlock; @@ -139,7 +139,7 @@ __xprt_lock_write_next(struct rpc_xprt *xprt) } out_unlock: smp_mb__before_clear_bit(); - clear_bit(XPRT_LOCKED, &xprt->sockstate); + clear_bit(XPRT_LOCKED, &xprt->state); smp_mb__after_clear_bit(); } @@ -152,7 +152,7 @@ __xprt_release_write(struct rpc_xprt *xprt, struct rpc_task *task) if (xprt->snd_task == task) { xprt->snd_task = NULL; smp_mb__before_clear_bit(); - clear_bit(XPRT_LOCKED, &xprt->sockstate); + clear_bit(XPRT_LOCKED, &xprt->state); smp_mb__after_clear_bit(); __xprt_lock_write_next(xprt); } @@ -312,11 +312,11 @@ xprt_init_autodisconnect(unsigned long data) spin_lock(&xprt->transport_lock); if (!list_empty(&xprt->recv) || xprt->shutdown) goto out_abort; - if (test_and_set_bit(XPRT_LOCKED, &xprt->sockstate)) + if (test_and_set_bit(XPRT_LOCKED, &xprt->state)) goto out_abort; spin_unlock(&xprt->transport_lock); /* Let keventd close the socket */ - if (test_bit(XPRT_CONNECTING, &xprt->sockstate) != 0) + if (xprt_connecting(xprt)) xprt_release_write(xprt, NULL); else schedule_work(&xprt->task_cleanup); diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c index bc90caab6088..76a33b54f436 100644 --- a/net/sunrpc/xprtsock.c +++ b/net/sunrpc/xprtsock.c @@ -925,9 +925,7 @@ out: else rpc_wake_up(&xprt->pending); out_clear: - smp_mb__before_clear_bit(); - clear_bit(XPRT_CONNECTING, &xprt->sockstate); - smp_mb__after_clear_bit(); + xprt_clear_connecting(xprt); } /** @@ -940,7 +938,7 @@ static void xs_connect(struct rpc_task *task) { struct rpc_xprt *xprt = task->tk_xprt; - if (!test_and_set_bit(XPRT_CONNECTING, &xprt->sockstate)) { + if (!xprt_test_and_set_connecting(xprt)) { if (xprt->sock != NULL) { dprintk("RPC: xs_connect delayed xprt %p\n", xprt); schedule_delayed_work(&xprt->sock_connect, -- cgit v1.2.3 From 44fbac2288dfed6f1963ac00bf922c3bcd779cd1 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 11 Aug 2005 16:25:44 -0400 Subject: [PATCH] RPC: Add helper for waking tasks pending on a transport Clean-up: remove only reference to xprt->pending from the socket transport implementation. This makes a cleaner interface for other transport implementations as well. Test-plan: Compile kernel with CONFIG_NFS enabled. Version: Thu, 11 Aug 2005 16:06:52 -0400 Signed-off-by: Chuck Lever Signed-off-by: Trond Myklebust --- include/linux/sunrpc/xprt.h | 1 + net/sunrpc/xprt.c | 18 ++++++++++++++++-- net/sunrpc/xprtsock.c | 7 ++----- 3 files changed, 19 insertions(+), 7 deletions(-) (limited to 'include/linux') diff --git a/include/linux/sunrpc/xprt.h b/include/linux/sunrpc/xprt.h index 009a3bb4f997..d5223993fca9 100644 --- a/include/linux/sunrpc/xprt.h +++ b/include/linux/sunrpc/xprt.h @@ -232,6 +232,7 @@ void xprt_reserve(struct rpc_task *); int xprt_prepare_transmit(struct rpc_task *); void xprt_transmit(struct rpc_task *); void xprt_receive(struct rpc_task *); +void xprt_wake_pending_tasks(struct rpc_xprt *, int); int xprt_adjust_timeout(struct rpc_rqst *req); void xprt_release(struct rpc_task *); void xprt_connect(struct rpc_task *); diff --git a/net/sunrpc/xprt.c b/net/sunrpc/xprt.c index 57c5e77b155e..2f9cd468b953 100644 --- a/net/sunrpc/xprt.c +++ b/net/sunrpc/xprt.c @@ -227,6 +227,20 @@ xprt_adjust_cwnd(struct rpc_xprt *xprt, int result) xprt->cwnd = cwnd; } +/** + * xprt_wake_pending_tasks - wake all tasks on a transport's pending queue + * @xprt: transport with waiting tasks + * @status: result code to plant in each task before waking it + * + */ +void xprt_wake_pending_tasks(struct rpc_xprt *xprt, int status) +{ + if (status < 0) + rpc_wake_up_status(&xprt->pending, status); + else + rpc_wake_up(&xprt->pending); +} + static void xprt_reset_majortimeo(struct rpc_rqst *req) { struct rpc_timeout *to = &req->rq_xprt->timeout; @@ -300,7 +314,7 @@ void xprt_disconnect(struct rpc_xprt *xprt) dprintk("RPC: disconnected transport %p\n", xprt); spin_lock_bh(&xprt->transport_lock); xprt_clear_connected(xprt); - rpc_wake_up_status(&xprt->pending, -ENOTCONN); + xprt_wake_pending_tasks(xprt, -ENOTCONN); spin_unlock_bh(&xprt->transport_lock); } @@ -803,7 +817,7 @@ static void xprt_shutdown(struct rpc_xprt *xprt) xprt->shutdown = 1; rpc_wake_up(&xprt->sending); rpc_wake_up(&xprt->resend); - rpc_wake_up(&xprt->pending); + xprt_wake_pending_tasks(xprt, -EIO); rpc_wake_up(&xprt->backlog); wake_up(&xprt->cong_wait); del_timer_sync(&xprt->timer); diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c index 76a33b54f436..182da2edf61c 100644 --- a/net/sunrpc/xprtsock.c +++ b/net/sunrpc/xprtsock.c @@ -703,7 +703,7 @@ static void xs_tcp_state_change(struct sock *sk) xprt->tcp_reclen = 0; xprt->tcp_copied = 0; xprt->tcp_flags = XPRT_COPY_RECM | XPRT_COPY_XID; - rpc_wake_up(&xprt->pending); + xprt_wake_pending_tasks(xprt, 0); } spin_unlock_bh(&xprt->transport_lock); break; @@ -920,10 +920,7 @@ static void xs_connect_worker(void *args) } } out: - if (status < 0) - rpc_wake_up_status(&xprt->pending, status); - else - rpc_wake_up(&xprt->pending); + xprt_wake_pending_tasks(xprt, status); out_clear: xprt_clear_connecting(xprt); } -- cgit v1.2.3 From 55aa4f58aa43dc9a51fb80010630d94b96053a2e Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 11 Aug 2005 16:25:47 -0400 Subject: [PATCH] RPC: client-side transport switch cleanup Clean-up: change some comments to reflect the realities of the new RPC transport switch mechanism. Get rid of unused xprt_receive() prototype. Also, organize function prototypes in xprt.h by usage and scope. Test-plan: Compile kernel with CONFIG_NFS enabled. Version: Thu, 11 Aug 2005 16:07:21 -0400 Signed-off-by: Chuck Lever Signed-off-by: Trond Myklebust --- include/linux/sunrpc/xprt.h | 61 ++++++++++++++++++++++++++------------------- net/sunrpc/clnt.c | 2 +- net/sunrpc/xprt.c | 26 +++++++++---------- net/sunrpc/xprtsock.c | 12 +++++---- 4 files changed, 55 insertions(+), 46 deletions(-) (limited to 'include/linux') diff --git a/include/linux/sunrpc/xprt.h b/include/linux/sunrpc/xprt.h index d5223993fca9..bfbc492ae36d 100644 --- a/include/linux/sunrpc/xprt.h +++ b/include/linux/sunrpc/xprt.h @@ -1,5 +1,5 @@ /* - * linux/include/linux/sunrpc/clnt_xprt.h + * linux/include/linux/sunrpc/xprt.h * * Declarations for the RPC transport interface. * @@ -150,8 +150,8 @@ struct rpc_xprt { unsigned long cong; /* current congestion */ unsigned long cwnd; /* congestion window */ - unsigned int rcvsize, /* socket receive buffer size */ - sndsize; /* socket send buffer size */ + unsigned int rcvsize, /* transport rcv buffer size */ + sndsize; /* transport send buffer size */ size_t max_payload; /* largest RPC payload size, in bytes */ @@ -184,12 +184,12 @@ struct rpc_xprt { unsigned long tcp_copied, /* copied to request */ tcp_flags; /* - * Connection of sockets + * Connection of transports */ - struct work_struct sock_connect; + struct work_struct connect_worker; unsigned short port; /* - * Disconnection of idle sockets + * Disconnection of idle transports */ struct work_struct task_cleanup; struct timer_list timer; @@ -219,27 +219,36 @@ struct rpc_xprt { #ifdef __KERNEL__ -struct rpc_xprt * xprt_create_proto(int proto, struct sockaddr_in *addr, - struct rpc_timeout *toparms); -void xprt_disconnect(struct rpc_xprt *); -int xprt_destroy(struct rpc_xprt *); -void xprt_set_timeout(struct rpc_timeout *, unsigned int, - unsigned long); -struct rpc_rqst * xprt_lookup_rqst(struct rpc_xprt *, u32); -void xprt_complete_rqst(struct rpc_xprt *, - struct rpc_rqst *, int); -void xprt_reserve(struct rpc_task *); -int xprt_prepare_transmit(struct rpc_task *); -void xprt_transmit(struct rpc_task *); -void xprt_receive(struct rpc_task *); -void xprt_wake_pending_tasks(struct rpc_xprt *, int); +/* + * Transport operations used by ULPs + */ +struct rpc_xprt * xprt_create_proto(int proto, struct sockaddr_in *addr, struct rpc_timeout *to); +void xprt_set_timeout(struct rpc_timeout *to, unsigned int retr, unsigned long incr); + +/* + * Generic internal transport functions + */ +void xprt_connect(struct rpc_task *task); +void xprt_reserve(struct rpc_task *task); +int xprt_prepare_transmit(struct rpc_task *task); +void xprt_transmit(struct rpc_task *task); int xprt_adjust_timeout(struct rpc_rqst *req); -void xprt_release(struct rpc_task *); -void xprt_connect(struct rpc_task *); -int xs_setup_udp(struct rpc_xprt *, - struct rpc_timeout *); -int xs_setup_tcp(struct rpc_xprt *, - struct rpc_timeout *); +void xprt_release(struct rpc_task *task); +int xprt_destroy(struct rpc_xprt *xprt); + +/* + * Transport switch helper functions + */ +void xprt_wake_pending_tasks(struct rpc_xprt *xprt, int status); +struct rpc_rqst * xprt_lookup_rqst(struct rpc_xprt *xprt, u32 xid); +void xprt_complete_rqst(struct rpc_xprt *xprt, struct rpc_rqst *req, int copied); +void xprt_disconnect(struct rpc_xprt *xprt); + +/* + * Socket transport setup operations + */ +int xs_setup_udp(struct rpc_xprt *xprt, struct rpc_timeout *to); +int xs_setup_tcp(struct rpc_xprt *xprt, struct rpc_timeout *to); /* * Reserved bit positions in xprt->state diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c index 0d1b010a4a01..4677959d2834 100644 --- a/net/sunrpc/clnt.c +++ b/net/sunrpc/clnt.c @@ -1,5 +1,5 @@ /* - * linux/net/sunrpc/rpcclnt.c + * linux/net/sunrpc/clnt.c * * This file contains the high-level RPC interface. * It is modeled as a finite state machine to support both synchronous diff --git a/net/sunrpc/xprt.c b/net/sunrpc/xprt.c index 2f9cd468b953..247fa1ec870c 100644 --- a/net/sunrpc/xprt.c +++ b/net/sunrpc/xprt.c @@ -10,12 +10,12 @@ * one is available. Otherwise, it sleeps on the backlog queue * (xprt_reserve). * - Next, the caller puts together the RPC message, stuffs it into - * the request struct, and calls xprt_call(). - * - xprt_call transmits the message and installs the caller on the - * socket's wait list. At the same time, it installs a timer that + * the request struct, and calls xprt_transmit(). + * - xprt_transmit sends the message and installs the caller on the + * transport's wait list. At the same time, it installs a timer that * is run after the packet's timeout has expired. * - When a packet arrives, the data_ready handler walks the list of - * pending requests for that socket. If a matching XID is found, the + * pending requests for that transport. If a matching XID is found, the * caller is woken up, and the timer removed. * - When no reply arrives within the timeout interval, the timer is * fired by the kernel and runs xprt_timer(). It either adjusts the @@ -32,6 +32,8 @@ * tasks that rely on callbacks. * * Copyright (C) 1995-1997, Olaf Kirch + * + * Transport switch API copyright (C) 2005, Chuck Lever */ #include @@ -52,8 +54,6 @@ # define RPCDBG_FACILITY RPCDBG_XPRT #endif -#define XPRT_MAX_BACKOFF (8) - /* * Local functions */ @@ -65,9 +65,9 @@ static int __xprt_get_cong(struct rpc_xprt *, struct rpc_task *); static int xprt_clear_backlog(struct rpc_xprt *xprt); /* - * Serialize write access to sockets, in order to prevent different + * Serialize write access to transports, in order to prevent different * requests from interfering with each other. - * Also prevents TCP socket connects from colliding with writes. + * Also prevents transport connects from colliding with writes. */ static int __xprt_lock_write(struct rpc_xprt *xprt, struct rpc_task *task) @@ -91,7 +91,7 @@ __xprt_lock_write(struct rpc_xprt *xprt, struct rpc_task *task) clear_bit(XPRT_LOCKED, &xprt->state); smp_mb__after_clear_bit(); out_sleep: - dprintk("RPC: %4d failed to lock socket %p\n", task->tk_pid, xprt); + dprintk("RPC: %4d failed to lock transport %p\n", task->tk_pid, xprt); task->tk_timeout = 0; task->tk_status = -EAGAIN; if (req && req->rq_ntrans) @@ -144,7 +144,7 @@ out_unlock: } /* - * Releases the socket for use by other requests. + * Releases the transport for use by other requests. */ static void __xprt_release_write(struct rpc_xprt *xprt, struct rpc_task *task) @@ -294,8 +294,7 @@ int xprt_adjust_timeout(struct rpc_rqst *req) return status; } -static void -xprt_socket_autoclose(void *args) +static void xprt_autoclose(void *args) { struct rpc_xprt *xprt = (struct rpc_xprt *)args; @@ -329,7 +328,6 @@ xprt_init_autodisconnect(unsigned long data) if (test_and_set_bit(XPRT_LOCKED, &xprt->state)) goto out_abort; spin_unlock(&xprt->transport_lock); - /* Let keventd close the socket */ if (xprt_connecting(xprt)) xprt_release_write(xprt, NULL); else @@ -770,7 +768,7 @@ static struct rpc_xprt *xprt_setup(int proto, struct sockaddr_in *ap, struct rpc INIT_LIST_HEAD(&xprt->free); INIT_LIST_HEAD(&xprt->recv); - INIT_WORK(&xprt->task_cleanup, xprt_socket_autoclose, xprt); + INIT_WORK(&xprt->task_cleanup, xprt_autoclose, xprt); init_timer(&xprt->timer); xprt->timer.function = xprt_init_autodisconnect; xprt->timer.data = (unsigned long) xprt; diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c index 182da2edf61c..7f0b9f7f167b 100644 --- a/net/sunrpc/xprtsock.c +++ b/net/sunrpc/xprtsock.c @@ -11,6 +11,8 @@ * Rewrite of larges part of the code in order to stabilize TCP stuff. * Fix behaviour when socket buffer is full. * (C) 1999 Trond Myklebust + * + * IP socket transport implementation, (C) 2005 Chuck Lever */ #include @@ -363,7 +365,7 @@ static void xs_destroy(struct rpc_xprt *xprt) { dprintk("RPC: xs_destroy xprt %p\n", xprt); - cancel_delayed_work(&xprt->sock_connect); + cancel_delayed_work(&xprt->connect_worker); flush_scheduled_work(); xprt_disconnect(xprt); @@ -938,11 +940,11 @@ static void xs_connect(struct rpc_task *task) if (!xprt_test_and_set_connecting(xprt)) { if (xprt->sock != NULL) { dprintk("RPC: xs_connect delayed xprt %p\n", xprt); - schedule_delayed_work(&xprt->sock_connect, + schedule_delayed_work(&xprt->connect_worker, RPC_REESTABLISH_TIMEOUT); } else { dprintk("RPC: xs_connect scheduled xprt %p\n", xprt); - schedule_work(&xprt->sock_connect); + schedule_work(&xprt->connect_worker); /* flush_scheduled_work can sleep... */ if (!RPC_IS_ASYNC(task)) flush_scheduled_work(); @@ -989,7 +991,7 @@ int xs_setup_udp(struct rpc_xprt *xprt, struct rpc_timeout *to) /* XXX: header size can vary due to auth type, IPv6, etc. */ xprt->max_payload = (1U << 16) - (MAX_HEADER << 3); - INIT_WORK(&xprt->sock_connect, xs_connect_worker, xprt); + INIT_WORK(&xprt->connect_worker, xs_connect_worker, xprt); xprt->ops = &xs_ops; @@ -1028,7 +1030,7 @@ int xs_setup_tcp(struct rpc_xprt *xprt, struct rpc_timeout *to) xprt->resvport = capable(CAP_NET_BIND_SERVICE) ? 1 : 0; xprt->max_payload = (1U << 31) - 1; - INIT_WORK(&xprt->sock_connect, xs_connect_worker, xprt); + INIT_WORK(&xprt->connect_worker, xs_connect_worker, xprt); xprt->ops = &xs_ops; -- cgit v1.2.3 From c7b2cae8a634015b72941ba2fc6c4bc9b8d3a129 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 11 Aug 2005 16:25:50 -0400 Subject: [PATCH] RPC: separate TCP and UDP write space callbacks Split the socket write space callback function into a TCP version and UDP version, eliminating one dependence on the "xprt->stream" variable. Keep the common pieces of this path in xprt.c so other transports can use it too. Test-plan: Write-intensive workload on a single mount point. Version: Thu, 11 Aug 2005 16:07:51 -0400 Signed-off-by: Chuck Lever Signed-off-by: Trond Myklebust --- include/linux/sunrpc/xprt.h | 2 ++ net/sunrpc/xprt.c | 34 ++++++++++++++++++ net/sunrpc/xprtsock.c | 84 ++++++++++++++++++++++++++++----------------- 3 files changed, 89 insertions(+), 31 deletions(-) (limited to 'include/linux') diff --git a/include/linux/sunrpc/xprt.h b/include/linux/sunrpc/xprt.h index bfbc492ae36d..e73174c7e450 100644 --- a/include/linux/sunrpc/xprt.h +++ b/include/linux/sunrpc/xprt.h @@ -240,6 +240,8 @@ int xprt_destroy(struct rpc_xprt *xprt); * Transport switch helper functions */ void xprt_wake_pending_tasks(struct rpc_xprt *xprt, int status); +void xprt_wait_for_buffer_space(struct rpc_task *task); +void xprt_write_space(struct rpc_xprt *xprt); struct rpc_rqst * xprt_lookup_rqst(struct rpc_xprt *xprt, u32 xid); void xprt_complete_rqst(struct rpc_xprt *xprt, struct rpc_rqst *req, int copied); void xprt_disconnect(struct rpc_xprt *xprt); diff --git a/net/sunrpc/xprt.c b/net/sunrpc/xprt.c index 247fa1ec870c..31ef7dc7eed6 100644 --- a/net/sunrpc/xprt.c +++ b/net/sunrpc/xprt.c @@ -241,6 +241,40 @@ void xprt_wake_pending_tasks(struct rpc_xprt *xprt, int status) rpc_wake_up(&xprt->pending); } +/** + * xprt_wait_for_buffer_space - wait for transport output buffer to clear + * @task: task to be put to sleep + * + */ +void xprt_wait_for_buffer_space(struct rpc_task *task) +{ + struct rpc_rqst *req = task->tk_rqstp; + struct rpc_xprt *xprt = req->rq_xprt; + + task->tk_timeout = req->rq_timeout; + rpc_sleep_on(&xprt->pending, task, NULL, NULL); +} + +/** + * xprt_write_space - wake the task waiting for transport output buffer space + * @xprt: transport with waiting tasks + * + * Can be called in a soft IRQ context, so xprt_write_space never sleeps. + */ +void xprt_write_space(struct rpc_xprt *xprt) +{ + if (unlikely(xprt->shutdown)) + return; + + spin_lock_bh(&xprt->transport_lock); + if (xprt->snd_task) { + dprintk("RPC: write space: waking waiting task on xprt %p\n", + xprt); + rpc_wake_up_task(xprt->snd_task); + } + spin_unlock_bh(&xprt->transport_lock); +} + static void xprt_reset_majortimeo(struct rpc_rqst *req) { struct rpc_timeout *to = &req->rq_xprt->timeout; diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c index 7f0b9f7f167b..70a772d7a796 100644 --- a/net/sunrpc/xprtsock.c +++ b/net/sunrpc/xprtsock.c @@ -308,15 +308,13 @@ static int xs_send_request(struct rpc_task *task) if (status == -EAGAIN) { if (test_bit(SOCK_ASYNC_NOSPACE, &xprt->sock->flags)) { - /* Protect against races with xs_write_space */ + /* Protect against races with write_space */ spin_lock_bh(&xprt->transport_lock); /* Don't race with disconnect */ if (!xprt_connected(xprt)) task->tk_status = -ENOTCONN; - else if (test_bit(SOCK_NOSPACE, &xprt->sock->flags)) { - task->tk_timeout = req->rq_timeout; - rpc_sleep_on(&xprt->pending, task, NULL, NULL); - } + else if (test_bit(SOCK_NOSPACE, &xprt->sock->flags)) + xprt_wait_for_buffer_space(task); spin_unlock_bh(&xprt->transport_lock); return status; } @@ -721,45 +719,68 @@ static void xs_tcp_state_change(struct sock *sk) } /** - * xs_write_space - callback invoked when socket buffer space becomes - * available + * xs_udp_write_space - callback invoked when socket buffer space + * becomes available * @sk: socket whose state has changed * * Called when more output buffer space is available for this socket. * We try not to wake our writers until they can make "significant" - * progress, otherwise we'll waste resources thrashing sock_sendmsg + * progress, otherwise we'll waste resources thrashing kernel_sendmsg * with a bunch of small requests. */ -static void xs_write_space(struct sock *sk) +static void xs_udp_write_space(struct sock *sk) { - struct rpc_xprt *xprt; - struct socket *sock; - read_lock(&sk->sk_callback_lock); - if (!(xprt = xprt_from_sock(sk)) || !(sock = sk->sk_socket)) - goto out; - if (xprt->shutdown) - goto out; - /* Wait until we have enough socket memory */ - if (xprt->stream) { - /* from net/core/stream.c:sk_stream_write_space */ - if (sk_stream_wspace(sk) < sk_stream_min_wspace(sk)) + /* from net/core/sock.c:sock_def_write_space */ + if (sock_writeable(sk)) { + struct socket *sock; + struct rpc_xprt *xprt; + + if (unlikely(!(sock = sk->sk_socket))) goto out; - } else { - /* from net/core/sock.c:sock_def_write_space */ - if (!sock_writeable(sk)) + if (unlikely(!(xprt = xprt_from_sock(sk)))) + goto out; + if (unlikely(!test_and_clear_bit(SOCK_NOSPACE, &sock->flags))) goto out; + + xprt_write_space(xprt); } - if (!test_and_clear_bit(SOCK_NOSPACE, &sock->flags)) - goto out; + out: + read_unlock(&sk->sk_callback_lock); +} - spin_lock_bh(&xprt->transport_lock); - if (xprt->snd_task) - rpc_wake_up_task(xprt->snd_task); - spin_unlock_bh(&xprt->transport_lock); -out: +/** + * xs_tcp_write_space - callback invoked when socket buffer space + * becomes available + * @sk: socket whose state has changed + * + * Called when more output buffer space is available for this socket. + * We try not to wake our writers until they can make "significant" + * progress, otherwise we'll waste resources thrashing kernel_sendmsg + * with a bunch of small requests. + */ +static void xs_tcp_write_space(struct sock *sk) +{ + read_lock(&sk->sk_callback_lock); + + /* from net/core/stream.c:sk_stream_write_space */ + if (sk_stream_wspace(sk) >= sk_stream_min_wspace(sk)) { + struct socket *sock; + struct rpc_xprt *xprt; + + if (unlikely(!(sock = sk->sk_socket))) + goto out; + if (unlikely(!(xprt = xprt_from_sock(sk)))) + goto out; + if (unlikely(!test_and_clear_bit(SOCK_NOSPACE, &sock->flags))) + goto out; + + xprt_write_space(xprt); + } + + out: read_unlock(&sk->sk_callback_lock); } @@ -855,15 +876,16 @@ static void xs_bind(struct rpc_xprt *xprt, struct socket *sock) xprt->old_write_space = sk->sk_write_space; if (xprt->prot == IPPROTO_UDP) { sk->sk_data_ready = xs_udp_data_ready; + sk->sk_write_space = xs_udp_write_space; sk->sk_no_check = UDP_CSUM_NORCV; xprt_set_connected(xprt); } else { tcp_sk(sk)->nonagle = 1; /* disable Nagle's algorithm */ sk->sk_data_ready = xs_tcp_data_ready; sk->sk_state_change = xs_tcp_state_change; + sk->sk_write_space = xs_tcp_write_space; xprt_clear_connected(xprt); } - sk->sk_write_space = xs_write_space; /* Reset to new socket */ xprt->sock = sock; -- cgit v1.2.3 From 808012fbb23a52ec59352445d2076d175ad4ab26 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 25 Aug 2005 16:25:49 -0700 Subject: [PATCH] RPC: skip over transport-specific heads automatically Add a generic mechanism for skipping over transport-specific headers when constructing an RPC request. This removes another "xprt->stream" dependency. Test-plan: Write-intensive workload on a single mount point (try both UDP and TCP). Signed-off-by: Chuck Lever Signed-off-by: Trond Myklebust --- include/linux/sunrpc/msg_prot.h | 25 +++++++++++++++++++++++++ include/linux/sunrpc/xprt.h | 7 +++++++ net/sunrpc/auth_gss/auth_gss.c | 6 ++---- net/sunrpc/clnt.c | 5 ++--- net/sunrpc/xprtsock.c | 24 +++++++++++++++++------- 5 files changed, 53 insertions(+), 14 deletions(-) (limited to 'include/linux') diff --git a/include/linux/sunrpc/msg_prot.h b/include/linux/sunrpc/msg_prot.h index 15f115332389..f43f237360ae 100644 --- a/include/linux/sunrpc/msg_prot.h +++ b/include/linux/sunrpc/msg_prot.h @@ -76,5 +76,30 @@ enum rpc_auth_stat { #define RPC_MAXNETNAMELEN 256 +/* + * From RFC 1831: + * + * "A record is composed of one or more record fragments. A record + * fragment is a four-byte header followed by 0 to (2**31) - 1 bytes of + * fragment data. The bytes encode an unsigned binary number; as with + * XDR integers, the byte order is from highest to lowest. The number + * encodes two values -- a boolean which indicates whether the fragment + * is the last fragment of the record (bit value 1 implies the fragment + * is the last fragment) and a 31-bit unsigned binary value which is the + * length in bytes of the fragment's data. The boolean value is the + * highest-order bit of the header; the length is the 31 low-order bits. + * (Note that this record specification is NOT in XDR standard form!)" + * + * The Linux RPC client always sends its requests in a single record + * fragment, limiting the maximum payload size for stream transports to + * 2GB. + */ + +typedef u32 rpc_fraghdr; + +#define RPC_LAST_STREAM_FRAGMENT (1U << 31) +#define RPC_FRAGMENT_SIZE_MASK (~RPC_LAST_STREAM_FRAGMENT) +#define RPC_MAX_FRAGMENT_SIZE ((1U << 31) - 1) + #endif /* __KERNEL__ */ #endif /* _LINUX_SUNRPC_MSGPROT_H_ */ diff --git a/include/linux/sunrpc/xprt.h b/include/linux/sunrpc/xprt.h index e73174c7e450..966c456a0f6d 100644 --- a/include/linux/sunrpc/xprt.h +++ b/include/linux/sunrpc/xprt.h @@ -155,6 +155,8 @@ struct rpc_xprt { size_t max_payload; /* largest RPC payload size, in bytes */ + unsigned int tsh_size; /* size of transport specific + header */ struct rpc_wait_queue sending; /* requests waiting to send */ struct rpc_wait_queue resend; /* requests waiting to resend */ @@ -236,6 +238,11 @@ int xprt_adjust_timeout(struct rpc_rqst *req); void xprt_release(struct rpc_task *task); int xprt_destroy(struct rpc_xprt *xprt); +static inline u32 *xprt_skip_transport_header(struct rpc_xprt *xprt, u32 *p) +{ + return p + xprt->tsh_size; +} + /* * Transport switch helper functions */ diff --git a/net/sunrpc/auth_gss/auth_gss.c b/net/sunrpc/auth_gss/auth_gss.c index 53a030acdf75..d2b08f16c257 100644 --- a/net/sunrpc/auth_gss/auth_gss.c +++ b/net/sunrpc/auth_gss/auth_gss.c @@ -844,10 +844,8 @@ gss_marshal(struct rpc_task *task, u32 *p) /* We compute the checksum for the verifier over the xdr-encoded bytes * starting with the xid and ending at the end of the credential: */ - iov.iov_base = req->rq_snd_buf.head[0].iov_base; - if (task->tk_client->cl_xprt->stream) - /* See clnt.c:call_header() */ - iov.iov_base += 4; + iov.iov_base = xprt_skip_transport_header(task->tk_xprt, + req->rq_snd_buf.head[0].iov_base); iov.iov_len = (u8 *)p - (u8 *)iov.iov_base; xdr_buf_from_iov(&iov, &verf_buf); diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c index 4677959d2834..cc1b773a79d3 100644 --- a/net/sunrpc/clnt.c +++ b/net/sunrpc/clnt.c @@ -1075,13 +1075,12 @@ static u32 * call_header(struct rpc_task *task) { struct rpc_clnt *clnt = task->tk_client; - struct rpc_xprt *xprt = clnt->cl_xprt; struct rpc_rqst *req = task->tk_rqstp; u32 *p = req->rq_svec[0].iov_base; /* FIXME: check buffer size? */ - if (xprt->stream) - *p++ = 0; /* fill in later */ + + p = xprt_skip_transport_header(task->tk_xprt, p); *p++ = req->rq_xid; /* XID */ *p++ = htonl(RPC_CALL); /* CALL */ *p++ = htonl(RPC_VERSION); /* RPC version */ diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c index 57988300640a..aaf053b1a0c4 100644 --- a/net/sunrpc/xprtsock.c +++ b/net/sunrpc/xprtsock.c @@ -282,6 +282,13 @@ static int xs_udp_send_request(struct rpc_task *task) return status; } +static inline void xs_encode_tcp_record_marker(struct xdr_buf *buf) +{ + u32 reclen = buf->len - sizeof(rpc_fraghdr); + rpc_fraghdr *base = buf->head[0].iov_base; + *base = htonl(RPC_LAST_STREAM_FRAGMENT | reclen); +} + /** * xs_tcp_send_request - write an RPC request to a TCP socket * @task: address of RPC task that manages the state of an RPC request @@ -301,11 +308,9 @@ static int xs_tcp_send_request(struct rpc_task *task) struct rpc_rqst *req = task->tk_rqstp; struct rpc_xprt *xprt = req->rq_xprt; struct xdr_buf *xdr = &req->rq_snd_buf; - u32 *marker = req->rq_svec[0].iov_base; int status, retry = 0; - /* Write the record marker */ - *marker = htonl(0x80000000|(req->rq_slen-sizeof(*marker))); + xs_encode_tcp_record_marker(&req->rq_snd_buf); xs_pktdump("packet data:", req->rq_svec->iov_base, @@ -503,16 +508,19 @@ static inline void xs_tcp_read_fraghdr(struct rpc_xprt *xprt, skb_reader_t *desc xprt->tcp_offset += used; if (used != len) return; + xprt->tcp_reclen = ntohl(xprt->tcp_recm); - if (xprt->tcp_reclen & 0x80000000) + if (xprt->tcp_reclen & RPC_LAST_STREAM_FRAGMENT) xprt->tcp_flags |= XPRT_LAST_FRAG; else xprt->tcp_flags &= ~XPRT_LAST_FRAG; - xprt->tcp_reclen &= 0x7fffffff; + xprt->tcp_reclen &= RPC_FRAGMENT_SIZE_MASK; + xprt->tcp_flags &= ~XPRT_COPY_RECM; xprt->tcp_offset = 0; + /* Sanity check of the record length */ - if (xprt->tcp_reclen < 4) { + if (unlikely(xprt->tcp_reclen < 4)) { dprintk("RPC: invalid TCP record fragment length\n"); xprt_disconnect(xprt); return; @@ -1065,6 +1073,7 @@ int xs_setup_udp(struct rpc_xprt *xprt, struct rpc_timeout *to) xprt->prot = IPPROTO_UDP; xprt->port = XS_MAX_RESVPORT; + xprt->tsh_size = 0; xprt->stream = 0; xprt->nocong = 0; xprt->cwnd = RPC_INITCWND; @@ -1105,11 +1114,12 @@ int xs_setup_tcp(struct rpc_xprt *xprt, struct rpc_timeout *to) xprt->prot = IPPROTO_TCP; xprt->port = XS_MAX_RESVPORT; + xprt->tsh_size = sizeof(rpc_fraghdr) / sizeof(u32); xprt->stream = 1; xprt->nocong = 1; xprt->cwnd = RPC_MAXCWND(xprt); xprt->resvport = capable(CAP_NET_BIND_SERVICE) ? 1 : 0; - xprt->max_payload = (1U << 31) - 1; + xprt->max_payload = RPC_MAX_FRAGMENT_SIZE; INIT_WORK(&xprt->connect_worker, xs_tcp_connect_worker, xprt); -- cgit v1.2.3 From 43118c29dea2b23798bd42a147015cceee7fa885 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 25 Aug 2005 16:25:49 -0700 Subject: [PATCH] RPC: get rid of xprt->stream Now we can fix up the last few places that use the "xprt->stream" variable, and get rid of it from the rpc_xprt structure. Test-plan: Destructive testing (unplugging the network temporarily). Connectathon with UDP and TCP. Signed-off-by: Chuck Lever Signed-off-by: Trond Myklebust --- fs/lockd/host.c | 3 +-- include/linux/sunrpc/xprt.h | 3 +-- net/sunrpc/xprt.c | 3 +-- net/sunrpc/xprtsock.c | 28 ++++++++++++++++++---------- 4 files changed, 21 insertions(+), 16 deletions(-) (limited to 'include/linux') diff --git a/fs/lockd/host.c b/fs/lockd/host.c index 82c77df81c5f..7901f5b8092c 100644 --- a/fs/lockd/host.c +++ b/fs/lockd/host.c @@ -173,11 +173,10 @@ nlm_bind_host(struct nlm_host *host) /* If we've already created an RPC client, check whether * RPC rebind is required - * Note: why keep rebinding if we're on a tcp connection? */ if ((clnt = host->h_rpcclnt) != NULL) { xprt = clnt->cl_xprt; - if (!xprt->stream && time_after_eq(jiffies, host->h_nextrebind)) { + if (time_after_eq(jiffies, host->h_nextrebind)) { clnt->cl_port = 0; host->h_nextrebind = jiffies + NLM_HOST_REBIND; dprintk("lockd: next rebind in %ld jiffies\n", diff --git a/include/linux/sunrpc/xprt.h b/include/linux/sunrpc/xprt.h index 966c456a0f6d..c9477f022efb 100644 --- a/include/linux/sunrpc/xprt.h +++ b/include/linux/sunrpc/xprt.h @@ -168,8 +168,7 @@ struct rpc_xprt { unsigned long state; /* transport state */ unsigned char shutdown : 1, /* being shut down */ nocong : 1, /* no congestion control */ - resvport : 1, /* use a reserved port */ - stream : 1; /* TCP */ + resvport : 1; /* use a reserved port */ /* * XID diff --git a/net/sunrpc/xprt.c b/net/sunrpc/xprt.c index 31ef7dc7eed6..43fef7626442 100644 --- a/net/sunrpc/xprt.c +++ b/net/sunrpc/xprt.c @@ -630,8 +630,7 @@ void xprt_transmit(struct rpc_task *task) case -ENOTCONN: return; default: - if (xprt->stream) - xprt_disconnect(xprt); + break; } xprt_release_write(xprt, task); return; diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c index aaf053b1a0c4..5bb6fed3df34 100644 --- a/net/sunrpc/xprtsock.c +++ b/net/sunrpc/xprtsock.c @@ -356,6 +356,7 @@ static int xs_tcp_send_request(struct rpc_task *task) default: dprintk("RPC: sendmsg returned unrecognized error %d\n", -status); + xprt_disconnect(xprt); break; } @@ -826,19 +827,17 @@ static void xs_tcp_write_space(struct sock *sk) } /** - * xs_set_buffer_size - set send and receive limits + * xs_udp_set_buffer_size - set send and receive limits * @xprt: generic transport * * Set socket send and receive limits based on the * sndsize and rcvsize fields in the generic transport - * structure. This applies only to UDP sockets. + * structure. */ -static void xs_set_buffer_size(struct rpc_xprt *xprt) +static void xs_udp_set_buffer_size(struct rpc_xprt *xprt) { struct sock *sk = xprt->inet; - if (xprt->stream) - return; if (xprt->rcvsize) { sk->sk_userlocks |= SOCK_RCVBUF_LOCK; sk->sk_rcvbuf = xprt->rcvsize * xprt->max_reqs * 2; @@ -850,6 +849,17 @@ static void xs_set_buffer_size(struct rpc_xprt *xprt) } } +/** + * xs_tcp_set_buffer_size - set send and receive limits + * @xprt: generic transport + * + * Nothing to do for TCP. + */ +static void xs_tcp_set_buffer_size(struct rpc_xprt *xprt) +{ + return; +} + static int xs_bindresvport(struct rpc_xprt *xprt, struct socket *sock) { struct sockaddr_in myaddr = { @@ -928,7 +938,7 @@ static void xs_udp_connect_worker(void *args) write_unlock_bh(&sk->sk_callback_lock); } - xs_set_buffer_size(xprt); + xs_udp_set_buffer_size(xprt); status = 0; out: xprt_wake_pending_tasks(xprt, status); @@ -1034,7 +1044,7 @@ static void xs_connect(struct rpc_task *task) } static struct rpc_xprt_ops xs_udp_ops = { - .set_buffer_size = xs_set_buffer_size, + .set_buffer_size = xs_udp_set_buffer_size, .connect = xs_connect, .send_request = xs_udp_send_request, .close = xs_close, @@ -1042,7 +1052,7 @@ static struct rpc_xprt_ops xs_udp_ops = { }; static struct rpc_xprt_ops xs_tcp_ops = { - .set_buffer_size = xs_set_buffer_size, + .set_buffer_size = xs_tcp_set_buffer_size, .connect = xs_connect, .send_request = xs_tcp_send_request, .close = xs_close, @@ -1074,7 +1084,6 @@ int xs_setup_udp(struct rpc_xprt *xprt, struct rpc_timeout *to) xprt->prot = IPPROTO_UDP; xprt->port = XS_MAX_RESVPORT; xprt->tsh_size = 0; - xprt->stream = 0; xprt->nocong = 0; xprt->cwnd = RPC_INITCWND; xprt->resvport = capable(CAP_NET_BIND_SERVICE) ? 1 : 0; @@ -1115,7 +1124,6 @@ int xs_setup_tcp(struct rpc_xprt *xprt, struct rpc_timeout *to) xprt->prot = IPPROTO_TCP; xprt->port = XS_MAX_RESVPORT; xprt->tsh_size = sizeof(rpc_fraghdr) / sizeof(u32); - xprt->stream = 1; xprt->nocong = 1; xprt->cwnd = RPC_MAXCWND(xprt); xprt->resvport = capable(CAP_NET_BIND_SERVICE) ? 1 : 0; -- cgit v1.2.3 From fe3aca290f17ae4978bd73d02aa4029f1c9c024c Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 25 Aug 2005 16:25:50 -0700 Subject: [PATCH] RPC: add API to set transport-specific timeouts Prepare the way to remove the "xprt->nocong" variable by adding a callout to the RPC client transport switch API to handle setting RPC retransmit timeouts. Add a pair of generic helper functions that provide the ability to set a simple fixed timeout, or to set a timeout based on the state of a round- trip estimator. Test-plan: Use WAN simulation to cause sporadic bursty packet loss. Look for significant regression in performance or client stability. Signed-off-by: Chuck Lever Signed-off-by: Trond Myklebust --- include/linux/sunrpc/xprt.h | 3 ++ net/sunrpc/xprt.c | 67 ++++++++++++++++++++++++++++++--------------- net/sunrpc/xprtsock.c | 2 ++ 3 files changed, 50 insertions(+), 22 deletions(-) (limited to 'include/linux') diff --git a/include/linux/sunrpc/xprt.h b/include/linux/sunrpc/xprt.h index c9477f022efb..ac08e99a81cb 100644 --- a/include/linux/sunrpc/xprt.h +++ b/include/linux/sunrpc/xprt.h @@ -134,6 +134,7 @@ struct rpc_xprt_ops { void (*set_buffer_size)(struct rpc_xprt *xprt); void (*connect)(struct rpc_task *task); int (*send_request)(struct rpc_task *task); + void (*set_retrans_timeout)(struct rpc_task *task); void (*close)(struct rpc_xprt *xprt); void (*destroy)(struct rpc_xprt *xprt); }; @@ -245,6 +246,8 @@ static inline u32 *xprt_skip_transport_header(struct rpc_xprt *xprt, u32 *p) /* * Transport switch helper functions */ +void xprt_set_retrans_timeout_def(struct rpc_task *task); +void xprt_set_retrans_timeout_rtt(struct rpc_task *task); void xprt_wake_pending_tasks(struct rpc_xprt *xprt, int status); void xprt_wait_for_buffer_space(struct rpc_task *task); void xprt_write_space(struct rpc_xprt *xprt); diff --git a/net/sunrpc/xprt.c b/net/sunrpc/xprt.c index 43fef7626442..1ac2fbe05102 100644 --- a/net/sunrpc/xprt.c +++ b/net/sunrpc/xprt.c @@ -275,6 +275,38 @@ void xprt_write_space(struct rpc_xprt *xprt) spin_unlock_bh(&xprt->transport_lock); } +/** + * xprt_set_retrans_timeout_def - set a request's retransmit timeout + * @task: task whose timeout is to be set + * + * Set a request's retransmit timeout based on the transport's + * default timeout parameters. Used by transports that don't adjust + * the retransmit timeout based on round-trip time estimation. + */ +void xprt_set_retrans_timeout_def(struct rpc_task *task) +{ + task->tk_timeout = task->tk_rqstp->rq_timeout; +} + +/* + * xprt_set_retrans_timeout_rtt - set a request's retransmit timeout + * @task: task whose timeout is to be set + * + * Set a request's retransmit timeout using the RTT estimator. + */ +void xprt_set_retrans_timeout_rtt(struct rpc_task *task) +{ + int timer = task->tk_msg.rpc_proc->p_timer; + struct rpc_rtt *rtt = task->tk_client->cl_rtt; + struct rpc_rqst *req = task->tk_rqstp; + unsigned long max_timeout = req->rq_xprt->timeout.to_maxval; + + task->tk_timeout = rpc_calc_rto(rtt, timer); + task->tk_timeout <<= rpc_ntimeo(rtt, timer) + req->rq_retries; + if (task->tk_timeout > max_timeout || task->tk_timeout == 0) + task->tk_timeout = max_timeout; +} + static void xprt_reset_majortimeo(struct rpc_rqst *req) { struct rpc_timeout *to = &req->rq_xprt->timeout; @@ -588,7 +620,6 @@ out_unlock: */ void xprt_transmit(struct rpc_task *task) { - struct rpc_clnt *clnt = task->tk_client; struct rpc_rqst *req = task->tk_rqstp; struct rpc_xprt *xprt = req->rq_xprt; int status; @@ -613,8 +644,19 @@ void xprt_transmit(struct rpc_task *task) return; status = xprt->ops->send_request(task); - if (!status) - goto out_receive; + if (status == 0) { + dprintk("RPC: %4d xmit complete\n", task->tk_pid); + spin_lock_bh(&xprt->transport_lock); + xprt->ops->set_retrans_timeout(task); + /* Don't race with disconnect */ + if (!xprt_connected(xprt)) + task->tk_status = -ENOTCONN; + else if (!req->rq_received) + rpc_sleep_on(&xprt->pending, task, NULL, xprt_timer); + __xprt_release_write(xprt, task); + spin_unlock_bh(&xprt->transport_lock); + return; + } /* Note: at this point, task->tk_sleeping has not yet been set, * hence there is no danger of the waking up task being put on @@ -634,25 +676,6 @@ void xprt_transmit(struct rpc_task *task) } xprt_release_write(xprt, task); return; - out_receive: - dprintk("RPC: %4d xmit complete\n", task->tk_pid); - /* Set the task's receive timeout value */ - spin_lock_bh(&xprt->transport_lock); - if (!xprt->nocong) { - int timer = task->tk_msg.rpc_proc->p_timer; - task->tk_timeout = rpc_calc_rto(clnt->cl_rtt, timer); - task->tk_timeout <<= rpc_ntimeo(clnt->cl_rtt, timer) + req->rq_retries; - if (task->tk_timeout > xprt->timeout.to_maxval || task->tk_timeout == 0) - task->tk_timeout = xprt->timeout.to_maxval; - } else - task->tk_timeout = req->rq_timeout; - /* Don't race with disconnect */ - if (!xprt_connected(xprt)) - task->tk_status = -ENOTCONN; - else if (!req->rq_received) - rpc_sleep_on(&xprt->pending, task, NULL, xprt_timer); - __xprt_release_write(xprt, task); - spin_unlock_bh(&xprt->transport_lock); } static inline void do_xprt_reserve(struct rpc_task *task) diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c index 5bb6fed3df34..79433ffd1df0 100644 --- a/net/sunrpc/xprtsock.c +++ b/net/sunrpc/xprtsock.c @@ -1047,6 +1047,7 @@ static struct rpc_xprt_ops xs_udp_ops = { .set_buffer_size = xs_udp_set_buffer_size, .connect = xs_connect, .send_request = xs_udp_send_request, + .set_retrans_timeout = xprt_set_retrans_timeout_rtt, .close = xs_close, .destroy = xs_destroy, }; @@ -1055,6 +1056,7 @@ static struct rpc_xprt_ops xs_tcp_ops = { .set_buffer_size = xs_tcp_set_buffer_size, .connect = xs_connect, .send_request = xs_tcp_send_request, + .set_retrans_timeout = xprt_set_retrans_timeout_def, .close = xs_close, .destroy = xs_destroy, }; -- cgit v1.2.3 From 12a804698b29d040b7cdd92e8a44b0e75164dae9 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 25 Aug 2005 16:25:51 -0700 Subject: [PATCH] RPC: expose API for serializing access to RPC transports The next several patches introduce an API that allows transports to choose whether the RPC client provides congestion control or whether the transport itself provides it. The first method we abstract is the one that serializes access to the RPC transport to prevent the bytes from different requests from mingling together. This method provides proper request serialization and the opportunity to prevent new requests from being started because the transport is congested. The normal situation is for the transport to handle congestion control itself. Although NFS over UDP was first, it has been recognized after years of experience that having the transport provide congestion control is much better than doing it in the RPC client. Thus TCP, and probably every future transport implementation, will use the default method, xprt_lock_write, provided in xprt.c, which does not provide any kind of congestion control. UDP can continue using the xprt.c-provided Van Jacobson congestion avoidance implementation. Test-plan: Use WAN simulation to cause sporadic bursty packet loss. Look for significant regression in performance or client stability. Signed-off-by: Chuck Lever Signed-off-by: Trond Myklebust --- include/linux/sunrpc/xprt.h | 3 +++ net/sunrpc/xprt.c | 64 ++++++++++++++++++++++++++++++++++++--------- net/sunrpc/xprtsock.c | 2 ++ 3 files changed, 57 insertions(+), 12 deletions(-) (limited to 'include/linux') diff --git a/include/linux/sunrpc/xprt.h b/include/linux/sunrpc/xprt.h index ac08e99a81cb..eee1c6877851 100644 --- a/include/linux/sunrpc/xprt.h +++ b/include/linux/sunrpc/xprt.h @@ -132,6 +132,7 @@ struct rpc_xprt; struct rpc_xprt_ops { void (*set_buffer_size)(struct rpc_xprt *xprt); + int (*reserve_xprt)(struct rpc_task *task); void (*connect)(struct rpc_task *task); int (*send_request)(struct rpc_task *task); void (*set_retrans_timeout)(struct rpc_task *task); @@ -232,6 +233,8 @@ void xprt_set_timeout(struct rpc_timeout *to, unsigned int retr, unsigned long */ void xprt_connect(struct rpc_task *task); void xprt_reserve(struct rpc_task *task); +int xprt_reserve_xprt(struct rpc_task *task); +int xprt_reserve_xprt_cong(struct rpc_task *task); int xprt_prepare_transmit(struct rpc_task *task); void xprt_transmit(struct rpc_task *task); int xprt_adjust_timeout(struct rpc_rqst *req); diff --git a/net/sunrpc/xprt.c b/net/sunrpc/xprt.c index 1ac2fbe05102..2d1e8b83dd68 100644 --- a/net/sunrpc/xprt.c +++ b/net/sunrpc/xprt.c @@ -64,14 +64,56 @@ static int __xprt_get_cong(struct rpc_xprt *, struct rpc_task *); static int xprt_clear_backlog(struct rpc_xprt *xprt); +/** + * xprt_reserve_xprt - serialize write access to transports + * @task: task that is requesting access to the transport + * + * This prevents mixing the payload of separate requests, and prevents + * transport connects from colliding with writes. No congestion control + * is provided. + */ +int xprt_reserve_xprt(struct rpc_task *task) +{ + struct rpc_xprt *xprt = task->tk_xprt; + struct rpc_rqst *req = task->tk_rqstp; + + if (test_and_set_bit(XPRT_LOCKED, &xprt->state)) { + if (task == xprt->snd_task) + return 1; + if (task == NULL) + return 0; + goto out_sleep; + } + xprt->snd_task = task; + if (req) { + req->rq_bytes_sent = 0; + req->rq_ntrans++; + } + return 1; + +out_sleep: + dprintk("RPC: %4d failed to lock transport %p\n", + task->tk_pid, xprt); + task->tk_timeout = 0; + task->tk_status = -EAGAIN; + if (req && req->rq_ntrans) + rpc_sleep_on(&xprt->resend, task, NULL, NULL); + else + rpc_sleep_on(&xprt->sending, task, NULL, NULL); + return 0; +} + /* - * Serialize write access to transports, in order to prevent different - * requests from interfering with each other. - * Also prevents transport connects from colliding with writes. + * xprt_reserve_xprt_cong - serialize write access to transports + * @task: task that is requesting access to the transport + * + * Same as xprt_reserve_xprt, but Van Jacobson congestion control is + * integrated into the decision of whether a request is allowed to be + * woken up and given access to the transport. */ -static int -__xprt_lock_write(struct rpc_xprt *xprt, struct rpc_task *task) +int xprt_reserve_xprt_cong(struct rpc_task *task) { + struct rpc_xprt *xprt = task->tk_xprt; struct rpc_rqst *req = task->tk_rqstp; if (test_and_set_bit(XPRT_LOCKED, &xprt->state)) { @@ -79,7 +121,7 @@ __xprt_lock_write(struct rpc_xprt *xprt, struct rpc_task *task) return 1; goto out_sleep; } - if (xprt->nocong || __xprt_get_cong(xprt, task)) { + if (__xprt_get_cong(xprt, task)) { xprt->snd_task = task; if (req) { req->rq_bytes_sent = 0; @@ -101,20 +143,18 @@ out_sleep: return 0; } -static inline int -xprt_lock_write(struct rpc_xprt *xprt, struct rpc_task *task) +static inline int xprt_lock_write(struct rpc_xprt *xprt, struct rpc_task *task) { int retval; spin_lock_bh(&xprt->transport_lock); - retval = __xprt_lock_write(xprt, task); + retval = xprt->ops->reserve_xprt(task); spin_unlock_bh(&xprt->transport_lock); return retval; } -static void -__xprt_lock_write_next(struct rpc_xprt *xprt) +static void __xprt_lock_write_next(struct rpc_xprt *xprt) { struct rpc_task *task; @@ -598,7 +638,7 @@ int xprt_prepare_transmit(struct rpc_task *task) err = req->rq_received; goto out_unlock; } - if (!__xprt_lock_write(xprt, task)) { + if (!xprt->ops->reserve_xprt(task)) { err = -EAGAIN; goto out_unlock; } diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c index 79433ffd1df0..fc4fbe8ea346 100644 --- a/net/sunrpc/xprtsock.c +++ b/net/sunrpc/xprtsock.c @@ -1045,6 +1045,7 @@ static void xs_connect(struct rpc_task *task) static struct rpc_xprt_ops xs_udp_ops = { .set_buffer_size = xs_udp_set_buffer_size, + .reserve_xprt = xprt_reserve_xprt_cong, .connect = xs_connect, .send_request = xs_udp_send_request, .set_retrans_timeout = xprt_set_retrans_timeout_rtt, @@ -1054,6 +1055,7 @@ static struct rpc_xprt_ops xs_udp_ops = { static struct rpc_xprt_ops xs_tcp_ops = { .set_buffer_size = xs_tcp_set_buffer_size, + .reserve_xprt = xprt_reserve_xprt, .connect = xs_connect, .send_request = xs_tcp_send_request, .set_retrans_timeout = xprt_set_retrans_timeout_def, -- cgit v1.2.3 From 49e9a89086b3cae784a4868ca852863e4f4ea3fe Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 25 Aug 2005 16:25:51 -0700 Subject: [PATCH] RPC: expose API for serializing access to RPC transports The next method we abstract is the one that releases a transport, allowing another task to have access to the transport. Again, one generic version of this is provided for transports that don't need the RPC client to perform congestion control, and one version is for transports that can use the original Van Jacobson implementation in xprt.c. Test-plan: Use WAN simulation to cause sporadic bursty packet loss. Look for significant regression in performance or client stability. Signed-off-by: Chuck Lever Signed-off-by: Trond Myklebust --- include/linux/sunrpc/xprt.h | 3 ++ net/sunrpc/xprt.c | 77 ++++++++++++++++++++++++++++++++++++--------- net/sunrpc/xprtsock.c | 2 ++ 3 files changed, 68 insertions(+), 14 deletions(-) (limited to 'include/linux') diff --git a/include/linux/sunrpc/xprt.h b/include/linux/sunrpc/xprt.h index eee1c6877851..86833b725bb5 100644 --- a/include/linux/sunrpc/xprt.h +++ b/include/linux/sunrpc/xprt.h @@ -133,6 +133,7 @@ struct rpc_xprt; struct rpc_xprt_ops { void (*set_buffer_size)(struct rpc_xprt *xprt); int (*reserve_xprt)(struct rpc_task *task); + void (*release_xprt)(struct rpc_xprt *xprt, struct rpc_task *task); void (*connect)(struct rpc_task *task); int (*send_request)(struct rpc_task *task); void (*set_retrans_timeout)(struct rpc_task *task); @@ -238,6 +239,8 @@ int xprt_reserve_xprt_cong(struct rpc_task *task); int xprt_prepare_transmit(struct rpc_task *task); void xprt_transmit(struct rpc_task *task); int xprt_adjust_timeout(struct rpc_rqst *req); +void xprt_release_xprt(struct rpc_xprt *xprt, struct rpc_task *task); +void xprt_release_xprt_cong(struct rpc_xprt *xprt, struct rpc_task *task); void xprt_release(struct rpc_task *task); int xprt_destroy(struct rpc_xprt *xprt); diff --git a/net/sunrpc/xprt.c b/net/sunrpc/xprt.c index 2d1e8b83dd68..e92ea99dd318 100644 --- a/net/sunrpc/xprt.c +++ b/net/sunrpc/xprt.c @@ -153,14 +153,42 @@ static inline int xprt_lock_write(struct rpc_xprt *xprt, struct rpc_task *task) return retval; } - static void __xprt_lock_write_next(struct rpc_xprt *xprt) +{ + struct rpc_task *task; + struct rpc_rqst *req; + + if (test_and_set_bit(XPRT_LOCKED, &xprt->state)) + return; + + task = rpc_wake_up_next(&xprt->resend); + if (!task) { + task = rpc_wake_up_next(&xprt->sending); + if (!task) + goto out_unlock; + } + + req = task->tk_rqstp; + xprt->snd_task = task; + if (req) { + req->rq_bytes_sent = 0; + req->rq_ntrans++; + } + return; + +out_unlock: + smp_mb__before_clear_bit(); + clear_bit(XPRT_LOCKED, &xprt->state); + smp_mb__after_clear_bit(); +} + +static void __xprt_lock_write_next_cong(struct rpc_xprt *xprt) { struct rpc_task *task; if (test_and_set_bit(XPRT_LOCKED, &xprt->state)) return; - if (!xprt->nocong && RPCXPRT_CONGESTED(xprt)) + if (RPCXPRT_CONGESTED(xprt)) goto out_unlock; task = rpc_wake_up_next(&xprt->resend); if (!task) { @@ -168,7 +196,7 @@ static void __xprt_lock_write_next(struct rpc_xprt *xprt) if (!task) goto out_unlock; } - if (xprt->nocong || __xprt_get_cong(xprt, task)) { + if (__xprt_get_cong(xprt, task)) { struct rpc_rqst *req = task->tk_rqstp; xprt->snd_task = task; if (req) { @@ -183,11 +211,14 @@ out_unlock: smp_mb__after_clear_bit(); } -/* - * Releases the transport for use by other requests. +/** + * xprt_release_xprt - allow other requests to use a transport + * @xprt: transport with other tasks potentially waiting + * @task: task that is releasing access to the transport + * + * Note that "task" can be NULL. No congestion control is provided. */ -static void -__xprt_release_write(struct rpc_xprt *xprt, struct rpc_task *task) +void xprt_release_xprt(struct rpc_xprt *xprt, struct rpc_task *task) { if (xprt->snd_task == task) { xprt->snd_task = NULL; @@ -198,11 +229,29 @@ __xprt_release_write(struct rpc_xprt *xprt, struct rpc_task *task) } } -static inline void -xprt_release_write(struct rpc_xprt *xprt, struct rpc_task *task) +/** + * xprt_release_xprt_cong - allow other requests to use a transport + * @xprt: transport with other tasks potentially waiting + * @task: task that is releasing access to the transport + * + * Note that "task" can be NULL. Another task is awoken to use the + * transport if the transport's congestion window allows it. + */ +void xprt_release_xprt_cong(struct rpc_xprt *xprt, struct rpc_task *task) +{ + if (xprt->snd_task == task) { + xprt->snd_task = NULL; + smp_mb__before_clear_bit(); + clear_bit(XPRT_LOCKED, &xprt->state); + smp_mb__after_clear_bit(); + __xprt_lock_write_next_cong(xprt); + } +} + +static inline void xprt_release_write(struct rpc_xprt *xprt, struct rpc_task *task) { spin_lock_bh(&xprt->transport_lock); - __xprt_release_write(xprt, task); + xprt->ops->release_xprt(xprt, task); spin_unlock_bh(&xprt->transport_lock); } @@ -237,7 +286,7 @@ __xprt_put_cong(struct rpc_xprt *xprt, struct rpc_rqst *req) return; req->rq_cong = 0; xprt->cong -= RPC_CWNDSCALE; - __xprt_lock_write_next(xprt); + __xprt_lock_write_next_cong(xprt); } /* @@ -256,7 +305,7 @@ xprt_adjust_cwnd(struct rpc_xprt *xprt, int result) cwnd += (RPC_CWNDSCALE * RPC_CWNDSCALE + (cwnd >> 1)) / cwnd; if (cwnd > RPC_MAXCWND(xprt)) cwnd = RPC_MAXCWND(xprt); - __xprt_lock_write_next(xprt); + __xprt_lock_write_next_cong(xprt); } else if (result == -ETIMEDOUT) { cwnd >>= 1; if (cwnd < RPC_CWNDSCALE) @@ -693,7 +742,7 @@ void xprt_transmit(struct rpc_task *task) task->tk_status = -ENOTCONN; else if (!req->rq_received) rpc_sleep_on(&xprt->pending, task, NULL, xprt_timer); - __xprt_release_write(xprt, task); + xprt->ops->release_xprt(xprt, task); spin_unlock_bh(&xprt->transport_lock); return; } @@ -792,7 +841,7 @@ void xprt_release(struct rpc_task *task) if (!(req = task->tk_rqstp)) return; spin_lock_bh(&xprt->transport_lock); - __xprt_release_write(xprt, task); + xprt->ops->release_xprt(xprt, task); __xprt_put_cong(xprt, req); if (!list_empty(&req->rq_list)) list_del(&req->rq_list); diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c index fc4fbe8ea346..8589c1ad55e3 100644 --- a/net/sunrpc/xprtsock.c +++ b/net/sunrpc/xprtsock.c @@ -1046,6 +1046,7 @@ static void xs_connect(struct rpc_task *task) static struct rpc_xprt_ops xs_udp_ops = { .set_buffer_size = xs_udp_set_buffer_size, .reserve_xprt = xprt_reserve_xprt_cong, + .release_xprt = xprt_release_xprt_cong, .connect = xs_connect, .send_request = xs_udp_send_request, .set_retrans_timeout = xprt_set_retrans_timeout_rtt, @@ -1056,6 +1057,7 @@ static struct rpc_xprt_ops xs_udp_ops = { static struct rpc_xprt_ops xs_tcp_ops = { .set_buffer_size = xs_tcp_set_buffer_size, .reserve_xprt = xprt_reserve_xprt, + .release_xprt = xprt_release_xprt, .connect = xs_connect, .send_request = xs_tcp_send_request, .set_retrans_timeout = xprt_set_retrans_timeout_def, -- cgit v1.2.3 From 46c0ee8bc4ad3743de05e8b8b20201df44dcb6d3 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 25 Aug 2005 16:25:52 -0700 Subject: [PATCH] RPC: separate xprt_timer implementations Allow transports to hook the retransmit timer interrupt. Some transports calculate their congestion window here so that a retransmit timeout has immediate effect on the congestion window. Test-plan: Use WAN simulation to cause sporadic bursty packet loss. Look for significant regression in performance or client stability. Signed-off-by: Chuck Lever Signed-off-by: Trond Myklebust --- include/linux/sunrpc/xprt.h | 2 ++ net/sunrpc/xprt.c | 45 ++++++++++++++++++++------------------------- net/sunrpc/xprtsock.c | 12 ++++++++++++ 3 files changed, 34 insertions(+), 25 deletions(-) (limited to 'include/linux') diff --git a/include/linux/sunrpc/xprt.h b/include/linux/sunrpc/xprt.h index 86833b725bb5..443c3f984cf9 100644 --- a/include/linux/sunrpc/xprt.h +++ b/include/linux/sunrpc/xprt.h @@ -137,6 +137,7 @@ struct rpc_xprt_ops { void (*connect)(struct rpc_task *task); int (*send_request)(struct rpc_task *task); void (*set_retrans_timeout)(struct rpc_task *task); + void (*timer)(struct rpc_task *task); void (*close)(struct rpc_xprt *xprt); void (*destroy)(struct rpc_xprt *xprt); }; @@ -257,6 +258,7 @@ void xprt_set_retrans_timeout_rtt(struct rpc_task *task); void xprt_wake_pending_tasks(struct rpc_xprt *xprt, int status); void xprt_wait_for_buffer_space(struct rpc_task *task); void xprt_write_space(struct rpc_xprt *xprt); +void xprt_adjust_cwnd(struct rpc_task *task, int result); struct rpc_rqst * xprt_lookup_rqst(struct rpc_xprt *xprt, u32 xid); void xprt_complete_rqst(struct rpc_xprt *xprt, struct rpc_rqst *req, int copied); void xprt_disconnect(struct rpc_xprt *xprt); diff --git a/net/sunrpc/xprt.c b/net/sunrpc/xprt.c index e92ea99dd318..ffc595592af3 100644 --- a/net/sunrpc/xprt.c +++ b/net/sunrpc/xprt.c @@ -289,16 +289,19 @@ __xprt_put_cong(struct rpc_xprt *xprt, struct rpc_rqst *req) __xprt_lock_write_next_cong(xprt); } -/* - * Adjust RPC congestion window +/** + * xprt_adjust_cwnd - adjust transport congestion window + * @task: recently completed RPC request used to adjust window + * @result: result code of completed RPC request + * * We use a time-smoothed congestion estimator to avoid heavy oscillation. */ -static void -xprt_adjust_cwnd(struct rpc_xprt *xprt, int result) +void xprt_adjust_cwnd(struct rpc_task *task, int result) { - unsigned long cwnd; + struct rpc_rqst *req = task->tk_rqstp; + struct rpc_xprt *xprt = task->tk_xprt; + unsigned long cwnd = xprt->cwnd; - cwnd = xprt->cwnd; if (result >= 0 && cwnd <= xprt->cong) { /* The (cwnd >> 1) term makes sure * the result gets rounded properly. */ @@ -314,6 +317,7 @@ xprt_adjust_cwnd(struct rpc_xprt *xprt, int result) dprintk("RPC: cong %ld, cwnd was %ld, now %ld\n", xprt->cong, xprt->cwnd, cwnd); xprt->cwnd = cwnd; + __xprt_put_cong(xprt, req); } /** @@ -602,8 +606,7 @@ void xprt_complete_rqst(struct rpc_xprt *xprt, struct rpc_rqst *req, int copied) /* Adjust congestion window */ if (!xprt->nocong) { unsigned timer = task->tk_msg.rpc_proc->p_timer; - xprt_adjust_cwnd(xprt, copied); - __xprt_put_cong(xprt, req); + xprt_adjust_cwnd(task, copied); if (timer) { if (req->rq_ntrans == 1) rpc_update_rtt(clnt->cl_rtt, timer, @@ -640,27 +643,19 @@ void xprt_complete_rqst(struct rpc_xprt *xprt, struct rpc_rqst *req, int copied) return; } -/* - * RPC receive timeout handler. - */ -static void -xprt_timer(struct rpc_task *task) +static void xprt_timer(struct rpc_task *task) { - struct rpc_rqst *req = task->tk_rqstp; + struct rpc_rqst *req = task->tk_rqstp; struct rpc_xprt *xprt = req->rq_xprt; - spin_lock(&xprt->transport_lock); - if (req->rq_received) - goto out; - - xprt_adjust_cwnd(req->rq_xprt, -ETIMEDOUT); - __xprt_put_cong(xprt, req); + dprintk("RPC: %4d xprt_timer\n", task->tk_pid); - dprintk("RPC: %4d xprt_timer (%s request)\n", - task->tk_pid, req ? "pending" : "backlogged"); - - task->tk_status = -ETIMEDOUT; -out: + spin_lock(&xprt->transport_lock); + if (!req->rq_received) { + if (xprt->ops->timer) + xprt->ops->timer(task); + task->tk_status = -ETIMEDOUT; + } task->tk_timeout = 0; rpc_wake_up_task(task); spin_unlock(&xprt->transport_lock); diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c index 8589c1ad55e3..c3658ff027a6 100644 --- a/net/sunrpc/xprtsock.c +++ b/net/sunrpc/xprtsock.c @@ -860,6 +860,17 @@ static void xs_tcp_set_buffer_size(struct rpc_xprt *xprt) return; } +/** + * xs_udp_timer - called when a retransmit timeout occurs on a UDP transport + * @task: task that timed out + * + * Adjust the congestion window after a retransmit timeout has occurred. + */ +static void xs_udp_timer(struct rpc_task *task) +{ + xprt_adjust_cwnd(task, -ETIMEDOUT); +} + static int xs_bindresvport(struct rpc_xprt *xprt, struct socket *sock) { struct sockaddr_in myaddr = { @@ -1050,6 +1061,7 @@ static struct rpc_xprt_ops xs_udp_ops = { .connect = xs_connect, .send_request = xs_udp_send_request, .set_retrans_timeout = xprt_set_retrans_timeout_rtt, + .timer = xs_udp_timer, .close = xs_close, .destroy = xs_destroy, }; -- cgit v1.2.3 From 1570c1e41eabf6b7031f3e4322a2cf1cbe319fee Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 25 Aug 2005 16:25:52 -0700 Subject: [PATCH] RPC: add generic interface for adjusting the congestion window A new interface that allows transports to adjust their congestion window using the Van Jacobson implementation in xprt.c is provided. Test-plan: Use WAN simulation to cause sporadic bursty packet loss. Look for significant regression in performance or client stability. Signed-off-by: Chuck Lever Signed-off-by: Trond Myklebust --- include/linux/sunrpc/xprt.h | 3 ++- net/sunrpc/xprt.c | 66 ++++++++++++++++++--------------------------- net/sunrpc/xprtsock.c | 13 ++++----- 3 files changed, 33 insertions(+), 49 deletions(-) (limited to 'include/linux') diff --git a/include/linux/sunrpc/xprt.h b/include/linux/sunrpc/xprt.h index 443c3f984cf9..2e48752d55d9 100644 --- a/include/linux/sunrpc/xprt.h +++ b/include/linux/sunrpc/xprt.h @@ -258,9 +258,10 @@ void xprt_set_retrans_timeout_rtt(struct rpc_task *task); void xprt_wake_pending_tasks(struct rpc_xprt *xprt, int status); void xprt_wait_for_buffer_space(struct rpc_task *task); void xprt_write_space(struct rpc_xprt *xprt); +void xprt_update_rtt(struct rpc_task *task); void xprt_adjust_cwnd(struct rpc_task *task, int result); struct rpc_rqst * xprt_lookup_rqst(struct rpc_xprt *xprt, u32 xid); -void xprt_complete_rqst(struct rpc_xprt *xprt, struct rpc_rqst *req, int copied); +void xprt_complete_rqst(struct rpc_task *task, int copied); void xprt_disconnect(struct rpc_xprt *xprt); /* diff --git a/net/sunrpc/xprt.c b/net/sunrpc/xprt.c index ffc595592af3..707806fe1a23 100644 --- a/net/sunrpc/xprt.c +++ b/net/sunrpc/xprt.c @@ -591,56 +591,42 @@ struct rpc_rqst *xprt_lookup_rqst(struct rpc_xprt *xprt, u32 xid) return req; } +/** + * xprt_update_rtt - update an RPC client's RTT state after receiving a reply + * @task: RPC request that recently completed + * + */ +void xprt_update_rtt(struct rpc_task *task) +{ + struct rpc_rqst *req = task->tk_rqstp; + struct rpc_rtt *rtt = task->tk_client->cl_rtt; + unsigned timer = task->tk_msg.rpc_proc->p_timer; + + if (timer) { + if (req->rq_ntrans == 1) + rpc_update_rtt(rtt, timer, + (long)jiffies - req->rq_xtime); + rpc_set_timeo(rtt, timer, req->rq_ntrans - 1); + } +} + /** * xprt_complete_rqst - called when reply processing is complete - * @xprt: controlling transport - * @req: RPC request that just completed + * @task: RPC request that recently completed * @copied: actual number of bytes received from the transport * + * Caller holds transport lock. */ -void xprt_complete_rqst(struct rpc_xprt *xprt, struct rpc_rqst *req, int copied) -{ - struct rpc_task *task = req->rq_task; - struct rpc_clnt *clnt = task->tk_client; - - /* Adjust congestion window */ - if (!xprt->nocong) { - unsigned timer = task->tk_msg.rpc_proc->p_timer; - xprt_adjust_cwnd(task, copied); - if (timer) { - if (req->rq_ntrans == 1) - rpc_update_rtt(clnt->cl_rtt, timer, - (long)jiffies - req->rq_xtime); - rpc_set_timeo(clnt->cl_rtt, timer, req->rq_ntrans - 1); - } - } +void xprt_complete_rqst(struct rpc_task *task, int copied) +{ + struct rpc_rqst *req = task->tk_rqstp; -#ifdef RPC_PROFILE - /* Profile only reads for now */ - if (copied > 1024) { - static unsigned long nextstat; - static unsigned long pkt_rtt, pkt_len, pkt_cnt; - - pkt_cnt++; - pkt_len += req->rq_slen + copied; - pkt_rtt += jiffies - req->rq_xtime; - if (time_before(nextstat, jiffies)) { - printk("RPC: %lu %ld cwnd\n", jiffies, xprt->cwnd); - printk("RPC: %ld %ld %ld %ld stat\n", - jiffies, pkt_cnt, pkt_len, pkt_rtt); - pkt_rtt = pkt_len = pkt_cnt = 0; - nextstat = jiffies + 5 * HZ; - } - } -#endif + dprintk("RPC: %5u xid %08x complete (%d bytes received)\n", + task->tk_pid, ntohl(req->rq_xid), copied); - dprintk("RPC: %4d has input (%d bytes)\n", task->tk_pid, copied); list_del_init(&req->rq_list); req->rq_received = req->rq_private_buf.len = copied; - - /* ... and wake up the process. */ rpc_wake_up_task(task); - return; } static void xprt_timer(struct rpc_task *task) diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c index c3658ff027a6..980f26504f48 100644 --- a/net/sunrpc/xprtsock.c +++ b/net/sunrpc/xprtsock.c @@ -460,8 +460,6 @@ static void xs_udp_data_ready(struct sock *sk, int len) goto out_unlock; task = rovr->rq_task; - dprintk("RPC: %4d received reply\n", task->tk_pid); - if ((copied = rovr->rq_private_buf.buflen) > repsize) copied = repsize; @@ -472,7 +470,9 @@ static void xs_udp_data_ready(struct sock *sk, int len) /* Something worked... */ dst_confirm(skb->dst); - xprt_complete_rqst(xprt, rovr, copied); + xprt_adjust_cwnd(task, copied); + xprt_update_rtt(task); + xprt_complete_rqst(task, copied); out_unlock: spin_unlock(&xprt->transport_lock); @@ -634,11 +634,8 @@ static inline void xs_tcp_read_request(struct rpc_xprt *xprt, skb_reader_t *desc } out: - if (!(xprt->tcp_flags & XPRT_COPY_DATA)) { - dprintk("RPC: %4d received reply complete\n", - req->rq_task->tk_pid); - xprt_complete_rqst(xprt, req, xprt->tcp_copied); - } + if (!(xprt->tcp_flags & XPRT_COPY_DATA)) + xprt_complete_rqst(req->rq_task, xprt->tcp_copied); spin_unlock(&xprt->transport_lock); xs_tcp_check_recm(xprt); } -- cgit v1.2.3 From a58dd398f5db4f73d5c581069fd70a4304cc4f0a Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 25 Aug 2005 16:25:53 -0700 Subject: [PATCH] RPC: add a release_rqst callout to the RPC transport switch The final place where congestion control state is adjusted is in xprt_release, where each request is finally released. Add a callout there to allow transports to perform additional processing when a request is about to be released. Test-plan: Use WAN simulation to cause sporadic bursty packet loss. Look for significant regression in performance or client stability. Signed-off-by: Chuck Lever Signed-off-by: Trond Myklebust --- include/linux/sunrpc/xprt.h | 2 ++ net/sunrpc/xprt.c | 14 +++++++++++++- net/sunrpc/xprtsock.c | 1 + 3 files changed, 16 insertions(+), 1 deletion(-) (limited to 'include/linux') diff --git a/include/linux/sunrpc/xprt.h b/include/linux/sunrpc/xprt.h index 2e48752d55d9..64e77658fa30 100644 --- a/include/linux/sunrpc/xprt.h +++ b/include/linux/sunrpc/xprt.h @@ -138,6 +138,7 @@ struct rpc_xprt_ops { int (*send_request)(struct rpc_task *task); void (*set_retrans_timeout)(struct rpc_task *task); void (*timer)(struct rpc_task *task); + void (*release_request)(struct rpc_task *task); void (*close)(struct rpc_xprt *xprt); void (*destroy)(struct rpc_xprt *xprt); }; @@ -262,6 +263,7 @@ void xprt_update_rtt(struct rpc_task *task); void xprt_adjust_cwnd(struct rpc_task *task, int result); struct rpc_rqst * xprt_lookup_rqst(struct rpc_xprt *xprt, u32 xid); void xprt_complete_rqst(struct rpc_task *task, int copied); +void xprt_release_rqst_cong(struct rpc_task *task); void xprt_disconnect(struct rpc_xprt *xprt); /* diff --git a/net/sunrpc/xprt.c b/net/sunrpc/xprt.c index 707806fe1a23..e8d11bd6158e 100644 --- a/net/sunrpc/xprt.c +++ b/net/sunrpc/xprt.c @@ -289,6 +289,17 @@ __xprt_put_cong(struct rpc_xprt *xprt, struct rpc_rqst *req) __xprt_lock_write_next_cong(xprt); } +/** + * xprt_release_rqst_cong - housekeeping when request is complete + * @task: RPC request that recently completed + * + * Useful for transports that require congestion control. + */ +void xprt_release_rqst_cong(struct rpc_task *task) +{ + __xprt_put_cong(task->tk_xprt, task->tk_rqstp); +} + /** * xprt_adjust_cwnd - adjust transport congestion window * @task: recently completed RPC request used to adjust window @@ -823,7 +834,8 @@ void xprt_release(struct rpc_task *task) return; spin_lock_bh(&xprt->transport_lock); xprt->ops->release_xprt(xprt, task); - __xprt_put_cong(xprt, req); + if (xprt->ops->release_request) + xprt->ops->release_request(task); if (!list_empty(&req->rq_list)) list_del(&req->rq_list); xprt->last_used = jiffies; diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c index 980f26504f48..6c2f5dcea416 100644 --- a/net/sunrpc/xprtsock.c +++ b/net/sunrpc/xprtsock.c @@ -1059,6 +1059,7 @@ static struct rpc_xprt_ops xs_udp_ops = { .send_request = xs_udp_send_request, .set_retrans_timeout = xprt_set_retrans_timeout_rtt, .timer = xs_udp_timer, + .release_request = xprt_release_rqst_cong, .close = xs_close, .destroy = xs_destroy, }; -- cgit v1.2.3 From ed63c003701a314c4893c11eceb9d68f8f46c662 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 25 Aug 2005 16:25:53 -0700 Subject: [PATCH] RPC: remove xprt->nocong Get rid of the "xprt->nocong" variable. Test-plan: Use WAN simulation to cause sporadic bursty packet loss with UDP mounts. Look for significant regression in performance or client stability. Signed-off-by: Chuck Lever Signed-off-by: Trond Myklebust --- fs/lockd/host.c | 1 - include/linux/sunrpc/xprt.h | 1 - net/sunrpc/xprtsock.c | 2 -- 3 files changed, 4 deletions(-) (limited to 'include/linux') diff --git a/fs/lockd/host.c b/fs/lockd/host.c index 7901f5b8092c..c4c8601096e0 100644 --- a/fs/lockd/host.c +++ b/fs/lockd/host.c @@ -188,7 +188,6 @@ nlm_bind_host(struct nlm_host *host) goto forgetit; xprt_set_timeout(&xprt->timeout, 5, nlmsvc_timeout); - xprt->nocong = 1; /* No congestion control for NLM */ xprt->resvport = 1; /* NLM requires a reserved port */ /* Existing NLM servers accept AUTH_UNIX only */ diff --git a/include/linux/sunrpc/xprt.h b/include/linux/sunrpc/xprt.h index 64e77658fa30..559fb471f6f2 100644 --- a/include/linux/sunrpc/xprt.h +++ b/include/linux/sunrpc/xprt.h @@ -172,7 +172,6 @@ struct rpc_xprt { unsigned int max_reqs; /* total slots */ unsigned long state; /* transport state */ unsigned char shutdown : 1, /* being shut down */ - nocong : 1, /* no congestion control */ resvport : 1; /* use a reserved port */ /* diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c index 6c2f5dcea416..7e5e020fe78d 100644 --- a/net/sunrpc/xprtsock.c +++ b/net/sunrpc/xprtsock.c @@ -1100,7 +1100,6 @@ int xs_setup_udp(struct rpc_xprt *xprt, struct rpc_timeout *to) xprt->prot = IPPROTO_UDP; xprt->port = XS_MAX_RESVPORT; xprt->tsh_size = 0; - xprt->nocong = 0; xprt->cwnd = RPC_INITCWND; xprt->resvport = capable(CAP_NET_BIND_SERVICE) ? 1 : 0; /* XXX: header size can vary due to auth type, IPv6, etc. */ @@ -1140,7 +1139,6 @@ int xs_setup_tcp(struct rpc_xprt *xprt, struct rpc_timeout *to) xprt->prot = IPPROTO_TCP; xprt->port = XS_MAX_RESVPORT; xprt->tsh_size = sizeof(rpc_fraghdr) / sizeof(u32); - xprt->nocong = 1; xprt->cwnd = RPC_MAXCWND(xprt); xprt->resvport = capable(CAP_NET_BIND_SERVICE) ? 1 : 0; xprt->max_payload = RPC_MAX_FRAGMENT_SIZE; -- cgit v1.2.3 From 555ee3af161b037865793bd4bebc06b58daafde6 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 25 Aug 2005 16:25:54 -0700 Subject: [PATCH] RPC: clean up after nocong was removed Clean-up: Move some macros that are specific to the Van Jacobson implementation into xprt.c. Get rid of the cong_wait field in rpc_xprt, which is no longer used. Get rid of xprt_clear_backlog. Test-plan: Compile with CONFIG_NFS enabled. Signed-off-by: Chuck Lever Signed-off-by: Trond Myklebust --- include/linux/sunrpc/xprt.h | 22 ---------------------- net/sunrpc/xprt.c | 29 +++++++++++++++++++---------- net/sunrpc/xprtsock.c | 2 -- 3 files changed, 19 insertions(+), 34 deletions(-) (limited to 'include/linux') diff --git a/include/linux/sunrpc/xprt.h b/include/linux/sunrpc/xprt.h index 559fb471f6f2..dcf0326bda01 100644 --- a/include/linux/sunrpc/xprt.h +++ b/include/linux/sunrpc/xprt.h @@ -15,20 +15,6 @@ #include #include -/* - * The transport code maintains an estimate on the maximum number of out- - * standing RPC requests, using a smoothed version of the congestion - * avoidance implemented in 44BSD. This is basically the Van Jacobson - * congestion algorithm: If a retransmit occurs, the congestion window is - * halved; otherwise, it is incremented by 1/cwnd when - * - * - a reply is received and - * - a full number of requests are outstanding and - * - the congestion window hasn't been updated recently. - * - * Upper procedures may check whether a request would block waiting for - * a free RPC slot by using the RPC_CONGESTED() macro. - */ extern unsigned int xprt_udp_slot_table_entries; extern unsigned int xprt_tcp_slot_table_entries; @@ -36,12 +22,6 @@ extern unsigned int xprt_tcp_slot_table_entries; #define RPC_DEF_SLOT_TABLE (16U) #define RPC_MAX_SLOT_TABLE (128U) -#define RPC_CWNDSHIFT (8U) -#define RPC_CWNDSCALE (1U << RPC_CWNDSHIFT) -#define RPC_INITCWND RPC_CWNDSCALE -#define RPC_MAXCWND(xprt) ((xprt)->max_reqs << RPC_CWNDSHIFT) -#define RPCXPRT_CONGESTED(xprt) ((xprt)->cong >= (xprt)->cwnd) - /* Default timeout values */ #define RPC_MAX_UDP_TIMEOUT (60*HZ) #define RPC_MAX_TCP_TIMEOUT (600*HZ) @@ -213,8 +193,6 @@ struct rpc_xprt { void (*old_data_ready)(struct sock *, int); void (*old_state_change)(struct sock *); void (*old_write_space)(struct sock *); - - wait_queue_head_t cong_wait; }; #define XPRT_LAST_FRAG (1 << 0) diff --git a/net/sunrpc/xprt.c b/net/sunrpc/xprt.c index e8d11bd6158e..0458319a1bdd 100644 --- a/net/sunrpc/xprt.c +++ b/net/sunrpc/xprt.c @@ -62,7 +62,23 @@ static inline void do_xprt_reserve(struct rpc_task *); static void xprt_connect_status(struct rpc_task *task); static int __xprt_get_cong(struct rpc_xprt *, struct rpc_task *); -static int xprt_clear_backlog(struct rpc_xprt *xprt); +/* + * The transport code maintains an estimate on the maximum number of out- + * standing RPC requests, using a smoothed version of the congestion + * avoidance implemented in 44BSD. This is basically the Van Jacobson + * congestion algorithm: If a retransmit occurs, the congestion window is + * halved; otherwise, it is incremented by 1/cwnd when + * + * - a reply is received and + * - a full number of requests are outstanding and + * - the congestion window hasn't been updated recently. + */ +#define RPC_CWNDSHIFT (8U) +#define RPC_CWNDSCALE (1U << RPC_CWNDSHIFT) +#define RPC_INITCWND RPC_CWNDSCALE +#define RPC_MAXCWND(xprt) ((xprt)->max_reqs << RPC_CWNDSHIFT) + +#define RPCXPRT_CONGESTED(xprt) ((xprt)->cong >= (xprt)->cwnd) /** * xprt_reserve_xprt - serialize write access to transports @@ -850,7 +866,7 @@ void xprt_release(struct rpc_task *task) spin_lock(&xprt->reserve_lock); list_add(&req->rq_list, &xprt->free); - xprt_clear_backlog(xprt); + rpc_wake_up_next(&xprt->backlog); spin_unlock(&xprt->reserve_lock); } @@ -902,7 +918,6 @@ static struct rpc_xprt *xprt_setup(int proto, struct sockaddr_in *ap, struct rpc spin_lock_init(&xprt->transport_lock); spin_lock_init(&xprt->reserve_lock); - init_waitqueue_head(&xprt->cong_wait); INIT_LIST_HEAD(&xprt->free); INIT_LIST_HEAD(&xprt->recv); @@ -911,6 +926,7 @@ static struct rpc_xprt *xprt_setup(int proto, struct sockaddr_in *ap, struct rpc xprt->timer.function = xprt_init_autodisconnect; xprt->timer.data = (unsigned long) xprt; xprt->last_used = jiffies; + xprt->cwnd = RPC_INITCWND; rpc_init_wait_queue(&xprt->pending, "xprt_pending"); rpc_init_wait_queue(&xprt->sending, "xprt_sending"); @@ -955,16 +971,9 @@ static void xprt_shutdown(struct rpc_xprt *xprt) rpc_wake_up(&xprt->resend); xprt_wake_pending_tasks(xprt, -EIO); rpc_wake_up(&xprt->backlog); - wake_up(&xprt->cong_wait); del_timer_sync(&xprt->timer); } -static int xprt_clear_backlog(struct rpc_xprt *xprt) { - rpc_wake_up_next(&xprt->backlog); - wake_up(&xprt->cong_wait); - return 1; -} - /** * xprt_destroy - destroy an RPC transport, killing off all requests. * @xprt: transport to destroy diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c index 7e5e020fe78d..26402c063f00 100644 --- a/net/sunrpc/xprtsock.c +++ b/net/sunrpc/xprtsock.c @@ -1100,7 +1100,6 @@ int xs_setup_udp(struct rpc_xprt *xprt, struct rpc_timeout *to) xprt->prot = IPPROTO_UDP; xprt->port = XS_MAX_RESVPORT; xprt->tsh_size = 0; - xprt->cwnd = RPC_INITCWND; xprt->resvport = capable(CAP_NET_BIND_SERVICE) ? 1 : 0; /* XXX: header size can vary due to auth type, IPv6, etc. */ xprt->max_payload = (1U << 16) - (MAX_HEADER << 3); @@ -1139,7 +1138,6 @@ int xs_setup_tcp(struct rpc_xprt *xprt, struct rpc_timeout *to) xprt->prot = IPPROTO_TCP; xprt->port = XS_MAX_RESVPORT; xprt->tsh_size = sizeof(rpc_fraghdr) / sizeof(u32); - xprt->cwnd = RPC_MAXCWND(xprt); xprt->resvport = capable(CAP_NET_BIND_SERVICE) ? 1 : 0; xprt->max_payload = RPC_MAX_FRAGMENT_SIZE; -- cgit v1.2.3 From 529b33c6db0120126b1381faa51406dc463acdc9 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 25 Aug 2005 16:25:54 -0700 Subject: [PATCH] RPC: allow RPC client's port range to be adjustable Select an RPC client source port between 650 and 1023 instead of between 1 and 800. The old range conflicts with a number of network services. Provide sysctls to allow admins to select a different port range. Note that this doesn't affect user-level RPC library behavior, which still uses 1 to 800. Based on a suggestion by Olaf Kirch . Test-plan: Repeated mount and unmount. Destructive testing. Idle timeouts. Signed-off-by: Chuck Lever Signed-off-by: Trond Myklebust --- include/linux/sunrpc/debug.h | 2 ++ include/linux/sunrpc/xprt.h | 17 ++++++++++++++--- net/sunrpc/sysctl.c | 29 +++++++++++++++++++++++++++++ net/sunrpc/xprtsock.c | 23 ++++++++--------------- 4 files changed, 53 insertions(+), 18 deletions(-) (limited to 'include/linux') diff --git a/include/linux/sunrpc/debug.h b/include/linux/sunrpc/debug.h index 42d299747956..1a42d902bc11 100644 --- a/include/linux/sunrpc/debug.h +++ b/include/linux/sunrpc/debug.h @@ -95,6 +95,8 @@ enum { CTL_NLMDEBUG, CTL_SLOTTABLE_UDP, CTL_SLOTTABLE_TCP, + CTL_MIN_RESVPORT, + CTL_MAX_RESVPORT, }; #endif /* _LINUX_SUNRPC_DEBUG_H_ */ diff --git a/include/linux/sunrpc/xprt.h b/include/linux/sunrpc/xprt.h index dcf0326bda01..9d9266cf8a36 100644 --- a/include/linux/sunrpc/xprt.h +++ b/include/linux/sunrpc/xprt.h @@ -51,6 +51,17 @@ extern unsigned int xprt_tcp_slot_table_entries; #define RPC_CALLHDRSIZE 6 #define RPC_REPHDRSIZE 4 +/* + * Parameters for choosing a free port + */ +extern unsigned int xprt_min_resvport; +extern unsigned int xprt_max_resvport; + +#define RPC_MIN_RESVPORT (1U) +#define RPC_MAX_RESVPORT (65535U) +#define RPC_DEF_MIN_RESVPORT (650U) +#define RPC_DEF_MAX_RESVPORT (1023U) + /* * This describes a timeout strategy */ @@ -62,6 +73,9 @@ struct rpc_timeout { unsigned char to_exponential; }; +struct rpc_task; +struct rpc_xprt; + /* * This describes a complete RPC request */ @@ -107,9 +121,6 @@ struct rpc_rqst { #define rq_svec rq_snd_buf.head #define rq_slen rq_snd_buf.len -struct rpc_task; -struct rpc_xprt; - struct rpc_xprt_ops { void (*set_buffer_size)(struct rpc_xprt *xprt); int (*reserve_xprt)(struct rpc_task *task); diff --git a/net/sunrpc/sysctl.c b/net/sunrpc/sysctl.c index ef483262f17f..d0c9f460e411 100644 --- a/net/sunrpc/sysctl.c +++ b/net/sunrpc/sysctl.c @@ -121,9 +121,16 @@ done: unsigned int xprt_udp_slot_table_entries = RPC_DEF_SLOT_TABLE; unsigned int xprt_tcp_slot_table_entries = RPC_DEF_SLOT_TABLE; +unsigned int xprt_min_resvport = RPC_DEF_MIN_RESVPORT; +EXPORT_SYMBOL(xprt_min_resvport); +unsigned int xprt_max_resvport = RPC_DEF_MAX_RESVPORT; +EXPORT_SYMBOL(xprt_max_resvport); + static unsigned int min_slot_table_size = RPC_MIN_SLOT_TABLE; static unsigned int max_slot_table_size = RPC_MAX_SLOT_TABLE; +static unsigned int xprt_min_resvport_limit = RPC_MIN_RESVPORT; +static unsigned int xprt_max_resvport_limit = RPC_MAX_RESVPORT; static ctl_table debug_table[] = { { @@ -180,6 +187,28 @@ static ctl_table debug_table[] = { .extra1 = &min_slot_table_size, .extra2 = &max_slot_table_size }, + { + .ctl_name = CTL_MIN_RESVPORT, + .procname = "min_resvport", + .data = &xprt_min_resvport, + .maxlen = sizeof(unsigned int), + .mode = 0644, + .proc_handler = &proc_dointvec_minmax, + .strategy = &sysctl_intvec, + .extra1 = &xprt_min_resvport_limit, + .extra2 = &xprt_max_resvport_limit + }, + { + .ctl_name = CTL_MAX_RESVPORT, + .procname = "max_resvport", + .data = &xprt_max_resvport, + .maxlen = sizeof(unsigned int), + .mode = 0644, + .proc_handler = &proc_dointvec_minmax, + .strategy = &sysctl_intvec, + .extra1 = &xprt_min_resvport_limit, + .extra2 = &xprt_max_resvport_limit + }, { .ctl_name = 0 } }; diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c index 26402c063f00..62c2e7caa345 100644 --- a/net/sunrpc/xprtsock.c +++ b/net/sunrpc/xprtsock.c @@ -35,11 +35,6 @@ #include #include -/* - * Maximum port number to use when requesting a reserved port. - */ -#define XS_MAX_RESVPORT (800U) - /* * How many times to try sending a request on a socket before waiting * for the socket buffer to clear. @@ -873,10 +868,9 @@ static int xs_bindresvport(struct rpc_xprt *xprt, struct socket *sock) struct sockaddr_in myaddr = { .sin_family = AF_INET, }; - int err, port; + int err; + unsigned short port = xprt->port; - /* Were we already bound to a given port? Try to reuse it */ - port = xprt->port; do { myaddr.sin_port = htons(port); err = sock->ops->bind(sock, (struct sockaddr *) &myaddr, @@ -887,8 +881,10 @@ static int xs_bindresvport(struct rpc_xprt *xprt, struct socket *sock) port); return 0; } - if (--port == 0) - port = XS_MAX_RESVPORT; + if (port <= xprt_min_resvport) + port = xprt_max_resvport; + else + port--; } while (err == -EADDRINUSE && port != xprt->port); dprintk("RPC: can't bind to reserved port (%d).\n", -err); @@ -1075,9 +1071,6 @@ static struct rpc_xprt_ops xs_tcp_ops = { .destroy = xs_destroy, }; -extern unsigned int xprt_udp_slot_table_entries; -extern unsigned int xprt_tcp_slot_table_entries; - /** * xs_setup_udp - Set up transport to use a UDP socket * @xprt: transport to set up @@ -1098,7 +1091,7 @@ int xs_setup_udp(struct rpc_xprt *xprt, struct rpc_timeout *to) memset(xprt->slot, 0, slot_table_size); xprt->prot = IPPROTO_UDP; - xprt->port = XS_MAX_RESVPORT; + xprt->port = xprt_max_resvport; xprt->tsh_size = 0; xprt->resvport = capable(CAP_NET_BIND_SERVICE) ? 1 : 0; /* XXX: header size can vary due to auth type, IPv6, etc. */ @@ -1136,7 +1129,7 @@ int xs_setup_tcp(struct rpc_xprt *xprt, struct rpc_timeout *to) memset(xprt->slot, 0, slot_table_size); xprt->prot = IPPROTO_TCP; - xprt->port = XS_MAX_RESVPORT; + xprt->port = xprt_max_resvport; xprt->tsh_size = sizeof(rpc_fraghdr) / sizeof(u32); xprt->resvport = capable(CAP_NET_BIND_SERVICE) ? 1 : 0; xprt->max_payload = RPC_MAX_FRAGMENT_SIZE; -- cgit v1.2.3 From 03bf4b707eee06706c9db343dd5c905b7ee47ed2 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 25 Aug 2005 16:25:55 -0700 Subject: [PATCH] RPC: parametrize various transport connect timeouts Each transport implementation can now set unique bind, connect, reestablishment, and idle timeout values. These are variables, allowing the values to be modified dynamically. This permits exponential backoff of any of these values, for instance. As an example, we implement exponential backoff for the connection reestablishment timeout. Test-plan: Destructive testing (unplugging the network temporarily). Connectathon with UDP and TCP. Signed-off-by: Chuck Lever Signed-off-by: Trond Myklebust --- fs/nfs/inode.c | 10 +++---- include/linux/nfs_fs.h | 4 +++ include/linux/sunrpc/xprt.h | 29 ++++--------------- net/sunrpc/clnt.c | 2 +- net/sunrpc/xprt.c | 5 ++-- net/sunrpc/xprtsock.c | 68 +++++++++++++++++++++++++++++++++++++++++++-- 6 files changed, 84 insertions(+), 34 deletions(-) (limited to 'include/linux') diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c index b6a1ca508e60..062911e7ceb5 100644 --- a/fs/nfs/inode.c +++ b/fs/nfs/inode.c @@ -369,8 +369,8 @@ static void nfs_init_timeout_values(struct rpc_timeout *to, int proto, unsigned case IPPROTO_TCP: if (!to->to_initval) to->to_initval = 60 * HZ; - if (to->to_initval > RPC_MAX_TCP_TIMEOUT) - to->to_initval = RPC_MAX_TCP_TIMEOUT; + if (to->to_initval > NFS_MAX_TCP_TIMEOUT) + to->to_initval = NFS_MAX_TCP_TIMEOUT; to->to_increment = to->to_initval; to->to_maxval = to->to_initval + (to->to_increment * to->to_retries); to->to_exponential = 0; @@ -379,9 +379,9 @@ static void nfs_init_timeout_values(struct rpc_timeout *to, int proto, unsigned default: if (!to->to_initval) to->to_initval = 11 * HZ / 10; - if (to->to_initval > RPC_MAX_UDP_TIMEOUT) - to->to_initval = RPC_MAX_UDP_TIMEOUT; - to->to_maxval = RPC_MAX_UDP_TIMEOUT; + if (to->to_initval > NFS_MAX_UDP_TIMEOUT) + to->to_initval = NFS_MAX_UDP_TIMEOUT; + to->to_maxval = NFS_MAX_UDP_TIMEOUT; to->to_exponential = 1; break; } diff --git a/include/linux/nfs_fs.h b/include/linux/nfs_fs.h index 9a6047ff1b25..7bac2785c6e4 100644 --- a/include/linux/nfs_fs.h +++ b/include/linux/nfs_fs.h @@ -41,6 +41,10 @@ #define NFS_MAX_FILE_IO_BUFFER_SIZE 32768 #define NFS_DEF_FILE_IO_BUFFER_SIZE 4096 +/* Default timeout values */ +#define NFS_MAX_UDP_TIMEOUT (60*HZ) +#define NFS_MAX_TCP_TIMEOUT (600*HZ) + /* * superblock magic number for NFS */ diff --git a/include/linux/sunrpc/xprt.h b/include/linux/sunrpc/xprt.h index 9d9266cf8a36..2543adf18551 100644 --- a/include/linux/sunrpc/xprt.h +++ b/include/linux/sunrpc/xprt.h @@ -22,28 +22,6 @@ extern unsigned int xprt_tcp_slot_table_entries; #define RPC_DEF_SLOT_TABLE (16U) #define RPC_MAX_SLOT_TABLE (128U) -/* Default timeout values */ -#define RPC_MAX_UDP_TIMEOUT (60*HZ) -#define RPC_MAX_TCP_TIMEOUT (600*HZ) - -/* - * Wait duration for an RPC TCP connection to be established. Solaris - * NFS over TCP uses 60 seconds, for example, which is in line with how - * long a server takes to reboot. - */ -#define RPC_CONNECT_TIMEOUT (60*HZ) - -/* - * Delay an arbitrary number of seconds before attempting to reconnect - * after an error. - */ -#define RPC_REESTABLISH_TIMEOUT (15*HZ) - -/* - * RPC transport idle timeout. - */ -#define RPC_IDLE_DISCONNECT_TIMEOUT (5*60*HZ) - /* * RPC call and reply header size as number of 32bit words (verifier * size computed separately) @@ -182,14 +160,19 @@ struct rpc_xprt { /* * Connection of transports */ + unsigned long connect_timeout, + bind_timeout, + reestablish_timeout; struct work_struct connect_worker; unsigned short port; + /* * Disconnection of idle transports */ struct work_struct task_cleanup; struct timer_list timer; - unsigned long last_used; + unsigned long last_used, + idle_timeout; /* * Send stuff diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c index cc1b773a79d3..24b44e73f391 100644 --- a/net/sunrpc/clnt.c +++ b/net/sunrpc/clnt.c @@ -740,7 +740,7 @@ call_bind(struct rpc_task *task) task->tk_action = call_connect; if (!clnt->cl_port) { task->tk_action = call_bind_status; - task->tk_timeout = RPC_CONNECT_TIMEOUT; + task->tk_timeout = task->tk_xprt->bind_timeout; rpc_getport(task, clnt); } } diff --git a/net/sunrpc/xprt.c b/net/sunrpc/xprt.c index 0458319a1bdd..215be0d0ef6b 100644 --- a/net/sunrpc/xprt.c +++ b/net/sunrpc/xprt.c @@ -551,7 +551,7 @@ void xprt_connect(struct rpc_task *task) if (task->tk_rqstp) task->tk_rqstp->rq_bytes_sent = 0; - task->tk_timeout = RPC_CONNECT_TIMEOUT; + task->tk_timeout = xprt->connect_timeout; rpc_sleep_on(&xprt->pending, task, xprt_connect_status, NULL); xprt->ops->connect(task); } @@ -763,7 +763,6 @@ void xprt_transmit(struct rpc_task *task) switch (status) { case -ECONNREFUSED: - task->tk_timeout = RPC_REESTABLISH_TIMEOUT; rpc_sleep_on(&xprt->sending, task, NULL, NULL); case -EAGAIN: case -ENOTCONN: @@ -857,7 +856,7 @@ void xprt_release(struct rpc_task *task) xprt->last_used = jiffies; if (list_empty(&xprt->recv) && !xprt->shutdown) mod_timer(&xprt->timer, - xprt->last_used + RPC_IDLE_DISCONNECT_TIMEOUT); + xprt->last_used + xprt->idle_timeout); spin_unlock_bh(&xprt->transport_lock); task->tk_rqstp = NULL; memset(req, 0, sizeof(*req)); /* mark unused */ diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c index 88ac71fcd335..06c2d95484e0 100644 --- a/net/sunrpc/xprtsock.c +++ b/net/sunrpc/xprtsock.c @@ -41,6 +41,50 @@ */ #define XS_SENDMSG_RETRY (10U) +/* + * Time out for an RPC UDP socket connect. UDP socket connects are + * synchronous, but we set a timeout anyway in case of resource + * exhaustion on the local host. + */ +#define XS_UDP_CONN_TO (5U * HZ) + +/* + * Wait duration for an RPC TCP connection to be established. Solaris + * NFS over TCP uses 60 seconds, for example, which is in line with how + * long a server takes to reboot. + */ +#define XS_TCP_CONN_TO (60U * HZ) + +/* + * Wait duration for a reply from the RPC portmapper. + */ +#define XS_BIND_TO (60U * HZ) + +/* + * Delay if a UDP socket connect error occurs. This is most likely some + * kind of resource problem on the local host. + */ +#define XS_UDP_REEST_TO (2U * HZ) + +/* + * The reestablish timeout allows clients to delay for a bit before attempting + * to reconnect to a server that just dropped our connection. + * + * We implement an exponential backoff when trying to reestablish a TCP + * transport connection with the server. Some servers like to drop a TCP + * connection when they are overworked, so we start with a short timeout and + * increase over time if the server is down or not responding. + */ +#define XS_TCP_INIT_REEST_TO (3U * HZ) +#define XS_TCP_MAX_REEST_TO (5U * 60 * HZ) + +/* + * TCP idle timeout; client drops the transport socket if it is idle + * for this long. Note that we also timeout UDP sockets to prevent + * holding port numbers when there is no RPC traffic. + */ +#define XS_IDLE_DISC_TO (5U * 60 * HZ) + #ifdef RPC_DEBUG # undef RPC_DEBUG_DATA # define RPCDBG_FACILITY RPCDBG_TRANS @@ -739,6 +783,7 @@ static void xs_tcp_state_change(struct sock *sk) xprt->tcp_reclen = 0; xprt->tcp_copied = 0; xprt->tcp_flags = XPRT_COPY_RECM | XPRT_COPY_XID; + xprt->reestablish_timeout = XS_TCP_INIT_REEST_TO; xprt_wake_pending_tasks(xprt, 0); } spin_unlock_bh(&xprt->transport_lock); @@ -1066,6 +1111,13 @@ out_clear: * @task: address of RPC task that manages state of connect request * * TCP: If the remote end dropped the connection, delay reconnecting. + * + * UDP socket connects are synchronous, but we use a work queue anyway + * to guarantee that even unprivileged user processes can set up a + * socket on a privileged port. + * + * If a UDP socket connect fails, the delay behavior here prevents + * retry floods (hard mounts). */ static void xs_connect(struct rpc_task *task) { @@ -1075,9 +1127,13 @@ static void xs_connect(struct rpc_task *task) return; if (xprt->sock != NULL) { - dprintk("RPC: xs_connect delayed xprt %p\n", xprt); + dprintk("RPC: xs_connect delayed xprt %p for %lu seconds\n", + xprt, xprt->reestablish_timeout / HZ); schedule_delayed_work(&xprt->connect_worker, - RPC_REESTABLISH_TIMEOUT); + xprt->reestablish_timeout); + xprt->reestablish_timeout <<= 1; + if (xprt->reestablish_timeout > XS_TCP_MAX_REEST_TO) + xprt->reestablish_timeout = XS_TCP_MAX_REEST_TO; } else { dprintk("RPC: xs_connect scheduled xprt %p\n", xprt); schedule_work(&xprt->connect_worker); @@ -1139,6 +1195,10 @@ int xs_setup_udp(struct rpc_xprt *xprt, struct rpc_timeout *to) xprt->max_payload = (1U << 16) - (MAX_HEADER << 3); INIT_WORK(&xprt->connect_worker, xs_udp_connect_worker, xprt); + xprt->bind_timeout = XS_BIND_TO; + xprt->connect_timeout = XS_UDP_CONN_TO; + xprt->reestablish_timeout = XS_UDP_REEST_TO; + xprt->idle_timeout = XS_IDLE_DISC_TO; xprt->ops = &xs_udp_ops; @@ -1176,6 +1236,10 @@ int xs_setup_tcp(struct rpc_xprt *xprt, struct rpc_timeout *to) xprt->max_payload = RPC_MAX_FRAGMENT_SIZE; INIT_WORK(&xprt->connect_worker, xs_tcp_connect_worker, xprt); + xprt->bind_timeout = XS_BIND_TO; + xprt->connect_timeout = XS_TCP_CONN_TO; + xprt->reestablish_timeout = XS_TCP_INIT_REEST_TO; + xprt->idle_timeout = XS_IDLE_DISC_TO; xprt->ops = &xs_tcp_ops; -- cgit v1.2.3 From 470056c288334eb0b37be26c9ff8aee37ed1cc7a Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Thu, 25 Aug 2005 16:25:56 -0700 Subject: [PATCH] RPC: rationalize set_buffer_size In fact, ->set_buffer_size should be completely functionless for non-UDP. Test-plan: Check socket buffer size on UDP sockets over time. Signed-off-by: Chuck Lever Signed-off-by: Trond Myklebust --- include/linux/sunrpc/xprt.h | 4 ++-- net/sunrpc/clnt.c | 10 ++-------- net/sunrpc/xprtsock.c | 30 +++++++++++++++--------------- 3 files changed, 19 insertions(+), 25 deletions(-) (limited to 'include/linux') diff --git a/include/linux/sunrpc/xprt.h b/include/linux/sunrpc/xprt.h index 2543adf18551..99cad3ead81d 100644 --- a/include/linux/sunrpc/xprt.h +++ b/include/linux/sunrpc/xprt.h @@ -100,7 +100,7 @@ struct rpc_rqst { #define rq_slen rq_snd_buf.len struct rpc_xprt_ops { - void (*set_buffer_size)(struct rpc_xprt *xprt); + void (*set_buffer_size)(struct rpc_xprt *xprt, size_t sndsize, size_t rcvsize); int (*reserve_xprt)(struct rpc_task *task); void (*release_xprt)(struct rpc_xprt *xprt, struct rpc_task *task); void (*connect)(struct rpc_task *task); @@ -124,7 +124,7 @@ struct rpc_xprt { unsigned long cong; /* current congestion */ unsigned long cwnd; /* congestion window */ - unsigned int rcvsize, /* transport rcv buffer size */ + size_t rcvsize, /* transport rcv buffer size */ sndsize; /* transport send buffer size */ size_t max_payload; /* largest RPC payload size, diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c index 24b44e73f391..5a8f01d726e9 100644 --- a/net/sunrpc/clnt.c +++ b/net/sunrpc/clnt.c @@ -517,14 +517,8 @@ void rpc_setbufsize(struct rpc_clnt *clnt, unsigned int sndsize, unsigned int rcvsize) { struct rpc_xprt *xprt = clnt->cl_xprt; - - xprt->sndsize = 0; - if (sndsize) - xprt->sndsize = sndsize + RPC_SLACK_SPACE; - xprt->rcvsize = 0; - if (rcvsize) - xprt->rcvsize = rcvsize + RPC_SLACK_SPACE; - xprt->ops->set_buffer_size(xprt); + if (xprt->ops->set_buffer_size) + xprt->ops->set_buffer_size(xprt, sndsize, rcvsize); } /* diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c index 06c2d95484e0..2e1529217e65 100644 --- a/net/sunrpc/xprtsock.c +++ b/net/sunrpc/xprtsock.c @@ -865,15 +865,7 @@ static void xs_tcp_write_space(struct sock *sk) read_unlock(&sk->sk_callback_lock); } -/** - * xs_udp_set_buffer_size - set send and receive limits - * @xprt: generic transport - * - * Set socket send and receive limits based on the - * sndsize and rcvsize fields in the generic transport - * structure. - */ -static void xs_udp_set_buffer_size(struct rpc_xprt *xprt) +static void xs_udp_do_set_buffer_size(struct rpc_xprt *xprt) { struct sock *sk = xprt->inet; @@ -889,14 +881,23 @@ static void xs_udp_set_buffer_size(struct rpc_xprt *xprt) } /** - * xs_tcp_set_buffer_size - set send and receive limits + * xs_udp_set_buffer_size - set send and receive limits * @xprt: generic transport + * @sndsize: requested size of send buffer, in bytes + * @rcvsize: requested size of receive buffer, in bytes * - * Nothing to do for TCP. + * Set socket send and receive buffer size limits. */ -static void xs_tcp_set_buffer_size(struct rpc_xprt *xprt) +static void xs_udp_set_buffer_size(struct rpc_xprt *xprt, size_t sndsize, size_t rcvsize) { - return; + xprt->sndsize = 0; + if (sndsize) + xprt->sndsize = sndsize + 1024; + xprt->rcvsize = 0; + if (rcvsize) + xprt->rcvsize = rcvsize + 1024; + + xs_udp_do_set_buffer_size(xprt); } /** @@ -989,7 +990,7 @@ static void xs_udp_connect_worker(void *args) write_unlock_bh(&sk->sk_callback_lock); } - xs_udp_set_buffer_size(xprt); + xs_udp_do_set_buffer_size(xprt); status = 0; out: xprt_wake_pending_tasks(xprt, status); @@ -1158,7 +1159,6 @@ static struct rpc_xprt_ops xs_udp_ops = { }; static struct rpc_xprt_ops xs_tcp_ops = { - .set_buffer_size = xs_tcp_set_buffer_size, .reserve_xprt = xprt_reserve_xprt, .release_xprt = xprt_release_xprt, .connect = xs_connect, -- cgit v1.2.3 From 278c995c8a153bb2a9bc427e931cfb9c8034c9d7 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Sun, 24 Jul 2005 23:53:01 +0100 Subject: [PATCH] RPC,NFS: new rpc_pipefs patch Currently rpc_mkdir/rpc_rmdir and rpc_mkpipe/mk_unlink have an API that's a little unfortunate. They take a path relative to the rpc_pipefs root and thus need to perform a full lookup. If you look at debugfs or usbfs they always store the dentry for directories they created and thus can pass in a dentry + single pathname component pair into their equivalents of the above functions. And in fact rpc_pipefs actually stores a dentry for all but one component so this change not only simplifies the core rpc_pipe code but also the callers. Unfortuntately this code path is only used by the NFS4 idmapper and AUTH_GSSAPI for which I don't have a test enviroment. Could someone give it a spin? It's the last bit needed before we can rework the lookup_hash API Signed-off-by: Christoph Hellwig Signed-off-by: Trond Myklebust --- fs/nfs/idmap.c | 10 +- include/linux/sunrpc/clnt.h | 2 +- include/linux/sunrpc/rpc_pipe_fs.h | 9 +- net/sunrpc/auth_gss/auth_gss.c | 9 +- net/sunrpc/clnt.c | 53 +++++--- net/sunrpc/rpc_pipe.c | 268 +++++++++++++------------------------ 6 files changed, 142 insertions(+), 209 deletions(-) (limited to 'include/linux') diff --git a/fs/nfs/idmap.c b/fs/nfs/idmap.c index ffb8df91dc34..1d0a5bf0d264 100644 --- a/fs/nfs/idmap.c +++ b/fs/nfs/idmap.c @@ -66,7 +66,6 @@ struct idmap_hashtable { }; struct idmap { - char idmap_path[48]; struct dentry *idmap_dentry; wait_queue_head_t idmap_wq; struct idmap_msg idmap_im; @@ -102,11 +101,8 @@ nfs_idmap_new(struct nfs4_client *clp) memset(idmap, 0, sizeof(*idmap)); - snprintf(idmap->idmap_path, sizeof(idmap->idmap_path), - "%s/idmap", clp->cl_rpcclient->cl_pathname); - - idmap->idmap_dentry = rpc_mkpipe(idmap->idmap_path, - idmap, &idmap_upcall_ops, 0); + idmap->idmap_dentry = rpc_mkpipe(clp->cl_rpcclient->cl_dentry, + "idmap", idmap, &idmap_upcall_ops, 0); if (IS_ERR(idmap->idmap_dentry)) { kfree(idmap); return; @@ -128,7 +124,7 @@ nfs_idmap_delete(struct nfs4_client *clp) if (!idmap) return; - rpc_unlink(idmap->idmap_path); + rpc_unlink(idmap->idmap_dentry); clp->cl_idmap = NULL; kfree(idmap); } diff --git a/include/linux/sunrpc/clnt.h b/include/linux/sunrpc/clnt.h index ab151bbb66df..b5b51c196690 100644 --- a/include/linux/sunrpc/clnt.h +++ b/include/linux/sunrpc/clnt.h @@ -59,7 +59,7 @@ struct rpc_clnt { int cl_nodelen; /* nodename length */ char cl_nodename[UNX_MAXNODENAME]; - char cl_pathname[30];/* Path in rpc_pipe_fs */ + struct dentry * __cl_parent_dentry; struct dentry * cl_dentry; /* inode */ struct rpc_clnt * cl_parent; /* Points to parent of clones */ struct rpc_rtt cl_rtt_default; diff --git a/include/linux/sunrpc/rpc_pipe_fs.h b/include/linux/sunrpc/rpc_pipe_fs.h index 63929349571f..63878d05c9a9 100644 --- a/include/linux/sunrpc/rpc_pipe_fs.h +++ b/include/linux/sunrpc/rpc_pipe_fs.h @@ -41,10 +41,11 @@ RPC_I(struct inode *inode) extern int rpc_queue_upcall(struct inode *, struct rpc_pipe_msg *); -extern struct dentry *rpc_mkdir(char *, struct rpc_clnt *); -extern int rpc_rmdir(char *); -extern struct dentry *rpc_mkpipe(char *, void *, struct rpc_pipe_ops *, int flags); -extern int rpc_unlink(char *); +extern struct dentry *rpc_mkdir(struct dentry *, char *, struct rpc_clnt *); +extern void rpc_rmdir(struct dentry *); +extern struct dentry *rpc_mkpipe(struct dentry *, char *, void *, + struct rpc_pipe_ops *, int flags); +extern void rpc_unlink(struct dentry *); #endif #endif diff --git a/net/sunrpc/auth_gss/auth_gss.c b/net/sunrpc/auth_gss/auth_gss.c index d2b08f16c257..bd2555139fa9 100644 --- a/net/sunrpc/auth_gss/auth_gss.c +++ b/net/sunrpc/auth_gss/auth_gss.c @@ -87,7 +87,6 @@ struct gss_auth { struct list_head upcalls; struct rpc_clnt *client; struct dentry *dentry; - char path[48]; spinlock_t lock; }; @@ -690,10 +689,8 @@ gss_create(struct rpc_clnt *clnt, rpc_authflavor_t flavor) if (err) goto err_put_mech; - snprintf(gss_auth->path, sizeof(gss_auth->path), "%s/%s", - clnt->cl_pathname, - gss_auth->mech->gm_name); - gss_auth->dentry = rpc_mkpipe(gss_auth->path, clnt, &gss_upcall_ops, RPC_PIPE_WAIT_FOR_OPEN); + gss_auth->dentry = rpc_mkpipe(clnt->cl_dentry, gss_auth->mech->gm_name, + clnt, &gss_upcall_ops, RPC_PIPE_WAIT_FOR_OPEN); if (IS_ERR(gss_auth->dentry)) { err = PTR_ERR(gss_auth->dentry); goto err_put_mech; @@ -718,7 +715,7 @@ gss_destroy(struct rpc_auth *auth) auth, auth->au_flavor); gss_auth = container_of(auth, struct gss_auth, rpc_auth); - rpc_unlink(gss_auth->path); + rpc_unlink(gss_auth->dentry); gss_mech_put(gss_auth->mech); rpcauth_free_credcache(auth); diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c index 5a8f01d726e9..63bf591310e0 100644 --- a/net/sunrpc/clnt.c +++ b/net/sunrpc/clnt.c @@ -67,26 +67,42 @@ static u32 * call_verify(struct rpc_task *task); static int rpc_setup_pipedir(struct rpc_clnt *clnt, char *dir_name) { - static uint32_t clntid; + static unsigned int clntid; + char name[128]; int error; if (dir_name == NULL) return 0; - for (;;) { - snprintf(clnt->cl_pathname, sizeof(clnt->cl_pathname), - "%s/clnt%x", dir_name, - (unsigned int)clntid++); - clnt->cl_pathname[sizeof(clnt->cl_pathname) - 1] = '\0'; - clnt->cl_dentry = rpc_mkdir(clnt->cl_pathname, clnt); - if (!IS_ERR(clnt->cl_dentry)) - return 0; + + retry_parent: + clnt->__cl_parent_dentry = rpc_mkdir(NULL, dir_name, NULL); + if (IS_ERR(clnt->__cl_parent_dentry)) { + error = PTR_ERR(clnt->__cl_parent_dentry); + if (error == -EEXIST) + goto retry_parent; /* XXX(hch): WTF? */ + + printk(KERN_INFO "RPC: Couldn't create pipefs entry %s, error %d\n", + dir_name, error); + return error; + } + + + retry_child: + snprintf(name, sizeof(name), "clnt%x", clntid++); + name[sizeof(name) - 1] = '\0'; + + clnt->cl_dentry = rpc_mkdir(clnt->__cl_parent_dentry, name, clnt); + if (IS_ERR(clnt->cl_dentry)) { error = PTR_ERR(clnt->cl_dentry); - if (error != -EEXIST) { - printk(KERN_INFO "RPC: Couldn't create pipefs entry %s, error %d\n", - clnt->cl_pathname, error); - return error; - } + if (error == -EEXIST) + goto retry_child; + printk(KERN_INFO "RPC: Couldn't create pipefs entry %s, error %d\n", + name, error); + rpc_rmdir(clnt->__cl_parent_dentry); + return error; } + + return 0; } /* @@ -174,7 +190,8 @@ rpc_new_client(struct rpc_xprt *xprt, char *servname, return clnt; out_no_auth: - rpc_rmdir(clnt->cl_pathname); + rpc_rmdir(clnt->cl_dentry); + rpc_rmdir(clnt->__cl_parent_dentry); out_no_path: if (clnt->cl_server != clnt->cl_inline_name) kfree(clnt->cl_server); @@ -302,8 +319,10 @@ rpc_destroy_client(struct rpc_clnt *clnt) rpc_destroy_client(clnt->cl_parent); goto out_free; } - if (clnt->cl_pathname[0]) - rpc_rmdir(clnt->cl_pathname); + if (clnt->cl_dentry) + rpc_rmdir(clnt->cl_dentry); + if (clnt->__cl_parent_dentry) + rpc_rmdir(clnt->__cl_parent_dentry); if (clnt->cl_xprt) { xprt_destroy(clnt->cl_xprt); clnt->cl_xprt = NULL; diff --git a/net/sunrpc/rpc_pipe.c b/net/sunrpc/rpc_pipe.c index ded6c63f11ec..b382809726d8 100644 --- a/net/sunrpc/rpc_pipe.c +++ b/net/sunrpc/rpc_pipe.c @@ -414,38 +414,6 @@ rpc_put_mount(void) simple_release_fs(&rpc_mount, &rpc_mount_count); } -static int -rpc_lookup_parent(char *path, struct nameidata *nd) -{ - if (path[0] == '\0') - return -ENOENT; - if (rpc_get_mount()) { - printk(KERN_WARNING "%s: %s failed to mount " - "pseudofilesystem \n", __FILE__, __FUNCTION__); - return -ENODEV; - } - nd->mnt = mntget(rpc_mount); - nd->dentry = dget(rpc_mount->mnt_root); - nd->last_type = LAST_ROOT; - nd->flags = LOOKUP_PARENT; - nd->depth = 0; - - if (path_walk(path, nd)) { - printk(KERN_WARNING "%s: %s failed to find path %s\n", - __FILE__, __FUNCTION__, path); - rpc_put_mount(); - return -ENOENT; - } - return 0; -} - -static void -rpc_release_path(struct nameidata *nd) -{ - path_release(nd); - rpc_put_mount(); -} - static struct inode * rpc_get_inode(struct super_block *sb, int mode) { @@ -550,197 +518,149 @@ out_bad: return -ENOMEM; } -static int -__rpc_mkdir(struct inode *dir, struct dentry *dentry) +struct dentry * +rpc_mkdir(struct dentry *parent, char *name, struct rpc_clnt *rpc_client) { + struct inode *dir; + struct dentry *dentry; struct inode *inode; - - inode = rpc_get_inode(dir->i_sb, S_IFDIR | S_IRUSR | S_IXUSR); - if (!inode) - goto out_err; - inode->i_ino = iunique(dir->i_sb, 100); - d_instantiate(dentry, inode); - dir->i_nlink++; - inode_dir_notify(dir, DN_CREATE); - rpc_get_mount(); - return 0; -out_err: - printk(KERN_WARNING "%s: %s failed to allocate inode for dentry %s\n", - __FILE__, __FUNCTION__, dentry->d_name.name); - return -ENOMEM; -} - -static int -__rpc_rmdir(struct inode *dir, struct dentry *dentry) -{ int error; - shrink_dcache_parent(dentry); - if (dentry->d_inode) { - rpc_close_pipes(dentry->d_inode); - rpc_inode_setowner(dentry->d_inode, NULL); - } - if ((error = simple_rmdir(dir, dentry)) != 0) - return error; - if (!error) { - inode_dir_notify(dir, DN_DELETE); - d_drop(dentry); - rpc_put_mount(); - } - return 0; -} + if (!parent) + parent = rpc_mount->mnt_root; -static struct dentry * -rpc_lookup_negative(char *path, struct nameidata *nd) -{ - struct dentry *dentry; - struct inode *dir; - int error; - - if ((error = rpc_lookup_parent(path, nd)) != 0) + dir = parent->d_inode; + + error = rpc_get_mount(); + if (error) return ERR_PTR(error); - dir = nd->dentry->d_inode; + down(&dir->i_sem); - dentry = lookup_hash(&nd->last, nd->dentry); + dentry = lookup_one_len(name, parent, strlen(name)); if (IS_ERR(dentry)) - goto out_err; + goto out_unlock; if (dentry->d_inode) { - dput(dentry); dentry = ERR_PTR(-EEXIST); - goto out_err; + goto out_dput; } - return dentry; -out_err: - up(&dir->i_sem); - rpc_release_path(nd); - return dentry; -} + inode = rpc_get_inode(dir->i_sb, S_IFDIR | S_IRUSR | S_IXUSR); + if (!inode) + goto out_dput; + inode->i_ino = iunique(dir->i_sb, 100); + dir->i_nlink++; + RPC_I(dentry->d_inode)->private = rpc_client; -struct dentry * -rpc_mkdir(char *path, struct rpc_clnt *rpc_client) -{ - struct nameidata nd; - struct dentry *dentry; - struct inode *dir; - int error; + d_instantiate(dentry, inode); + dget(dentry); + up(&dir->i_sem); + + inode_dir_notify(dir, DN_CREATE); - dentry = rpc_lookup_negative(path, &nd); - if (IS_ERR(dentry)) - return dentry; - dir = nd.dentry->d_inode; - if ((error = __rpc_mkdir(dir, dentry)) != 0) - goto err_dput; - RPC_I(dentry->d_inode)->private = rpc_client; error = rpc_populate(dentry, authfiles, RPCAUTH_info, RPCAUTH_EOF); if (error) - goto err_depopulate; -out: - up(&dir->i_sem); - rpc_release_path(&nd); + goto out_depopulate; + return dentry; -err_depopulate: - rpc_depopulate(dentry); - __rpc_rmdir(dir, dentry); -err_dput: + + out_depopulate: + rpc_rmdir(dentry); + out_dput: dput(dentry); - printk(KERN_WARNING "%s: %s() failed to create directory %s (errno = %d)\n", - __FILE__, __FUNCTION__, path, error); - dentry = ERR_PTR(error); - goto out; + out_unlock: + up(&dir->i_sem); + rpc_put_mount(); + return dentry; } -int -rpc_rmdir(char *path) +void +rpc_rmdir(struct dentry *dentry) { - struct nameidata nd; - struct dentry *dentry; - struct inode *dir; - int error; + struct dentry *parent = dentry->d_parent; - if ((error = rpc_lookup_parent(path, &nd)) != 0) - return error; - dir = nd.dentry->d_inode; - down(&dir->i_sem); - dentry = lookup_hash(&nd.last, nd.dentry); - if (IS_ERR(dentry)) { - error = PTR_ERR(dentry); - goto out_release; - } rpc_depopulate(dentry); - error = __rpc_rmdir(dir, dentry); - dput(dentry); -out_release: - up(&dir->i_sem); - rpc_release_path(&nd); - return error; + + down(&parent->d_inode->i_sem); + if (dentry->d_inode) { + rpc_close_pipes(dentry->d_inode); + rpc_inode_setowner(dentry->d_inode, NULL); + simple_rmdir(parent->d_inode, dentry); + } + up(&parent->d_inode->i_sem); + + inode_dir_notify(parent->d_inode, DN_DELETE); + rpc_put_mount(); } struct dentry * -rpc_mkpipe(char *path, void *private, struct rpc_pipe_ops *ops, int flags) +rpc_mkpipe(struct dentry *parent, char *name, void *private, + struct rpc_pipe_ops *ops, int flags) { - struct nameidata nd; + struct inode *dir = parent->d_inode; struct dentry *dentry; - struct inode *dir, *inode; + struct inode *inode; struct rpc_inode *rpci; + int error; + + error = rpc_get_mount(); + if (error) + return ERR_PTR(error); - dentry = rpc_lookup_negative(path, &nd); + down(&parent->d_inode->i_sem); + dentry = lookup_one_len(name, parent, strlen(name)); if (IS_ERR(dentry)) - return dentry; - dir = nd.dentry->d_inode; - inode = rpc_get_inode(dir->i_sb, S_IFSOCK | S_IRUSR | S_IWUSR); - if (!inode) - goto err_dput; + goto out_unlock; + if (dentry->d_inode) { + dentry = ERR_PTR(-EEXIST); + goto out_dput; + } + + inode = rpc_get_inode(parent->d_inode->i_sb, + S_IFSOCK | S_IRUSR | S_IWUSR); + if (!inode) { + dentry = ERR_PTR(-ENOMEM); + goto out_dput; + } + inode->i_ino = iunique(dir->i_sb, 100); inode->i_fop = &rpc_pipe_fops; - d_instantiate(dentry, inode); + rpci = RPC_I(inode); rpci->private = private; rpci->flags = flags; rpci->ops = ops; + + d_instantiate(dentry, inode); + dget(dentry); + up(&parent->d_inode->i_sem); + inode_dir_notify(dir, DN_CREATE); -out: - up(&dir->i_sem); - rpc_release_path(&nd); return dentry; -err_dput: + + out_dput: dput(dentry); - dentry = ERR_PTR(-ENOMEM); - printk(KERN_WARNING "%s: %s() failed to create pipe %s (errno = %d)\n", - __FILE__, __FUNCTION__, path, -ENOMEM); - goto out; + out_unlock: + up(&parent->d_inode->i_sem); + rpc_put_mount(); + return dentry; } -int -rpc_unlink(char *path) +void +rpc_unlink(struct dentry *dentry) { - struct nameidata nd; - struct dentry *dentry; - struct inode *dir; - int error; + struct dentry *parent = dentry->d_parent; - if ((error = rpc_lookup_parent(path, &nd)) != 0) - return error; - dir = nd.dentry->d_inode; - down(&dir->i_sem); - dentry = lookup_hash(&nd.last, nd.dentry); - if (IS_ERR(dentry)) { - error = PTR_ERR(dentry); - goto out_release; - } - d_drop(dentry); + down(&parent->d_inode->i_sem); if (dentry->d_inode) { rpc_close_pipes(dentry->d_inode); rpc_inode_setowner(dentry->d_inode, NULL); - error = simple_unlink(dir, dentry); + simple_unlink(parent->d_inode, dentry); } - dput(dentry); - inode_dir_notify(dir, DN_DELETE); -out_release: - up(&dir->i_sem); - rpc_release_path(&nd); - return error; + up(&parent->d_inode->i_sem); + + inode_dir_notify(parent->d_inode, DN_DELETE); + rpc_put_mount(); } /* -- cgit v1.2.3 From f134585a7343d71f9be7f0cf97e2145f21dd10c6 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Fri, 23 Sep 2005 11:08:25 -0400 Subject: Revert "[PATCH] RPC,NFS: new rpc_pipefs patch" This reverts 17f4e6febca160a9f9dd4bdece9784577a2f4524 commit. --- fs/nfs/idmap.c | 10 +- include/linux/sunrpc/clnt.h | 2 +- include/linux/sunrpc/rpc_pipe_fs.h | 9 +- net/sunrpc/auth_gss/auth_gss.c | 9 +- net/sunrpc/clnt.c | 53 +++----- net/sunrpc/rpc_pipe.c | 268 ++++++++++++++++++++++++------------- 6 files changed, 209 insertions(+), 142 deletions(-) (limited to 'include/linux') diff --git a/fs/nfs/idmap.c b/fs/nfs/idmap.c index 1d0a5bf0d264..ffb8df91dc34 100644 --- a/fs/nfs/idmap.c +++ b/fs/nfs/idmap.c @@ -66,6 +66,7 @@ struct idmap_hashtable { }; struct idmap { + char idmap_path[48]; struct dentry *idmap_dentry; wait_queue_head_t idmap_wq; struct idmap_msg idmap_im; @@ -101,8 +102,11 @@ nfs_idmap_new(struct nfs4_client *clp) memset(idmap, 0, sizeof(*idmap)); - idmap->idmap_dentry = rpc_mkpipe(clp->cl_rpcclient->cl_dentry, - "idmap", idmap, &idmap_upcall_ops, 0); + snprintf(idmap->idmap_path, sizeof(idmap->idmap_path), + "%s/idmap", clp->cl_rpcclient->cl_pathname); + + idmap->idmap_dentry = rpc_mkpipe(idmap->idmap_path, + idmap, &idmap_upcall_ops, 0); if (IS_ERR(idmap->idmap_dentry)) { kfree(idmap); return; @@ -124,7 +128,7 @@ nfs_idmap_delete(struct nfs4_client *clp) if (!idmap) return; - rpc_unlink(idmap->idmap_dentry); + rpc_unlink(idmap->idmap_path); clp->cl_idmap = NULL; kfree(idmap); } diff --git a/include/linux/sunrpc/clnt.h b/include/linux/sunrpc/clnt.h index b5b51c196690..ab151bbb66df 100644 --- a/include/linux/sunrpc/clnt.h +++ b/include/linux/sunrpc/clnt.h @@ -59,7 +59,7 @@ struct rpc_clnt { int cl_nodelen; /* nodename length */ char cl_nodename[UNX_MAXNODENAME]; - struct dentry * __cl_parent_dentry; + char cl_pathname[30];/* Path in rpc_pipe_fs */ struct dentry * cl_dentry; /* inode */ struct rpc_clnt * cl_parent; /* Points to parent of clones */ struct rpc_rtt cl_rtt_default; diff --git a/include/linux/sunrpc/rpc_pipe_fs.h b/include/linux/sunrpc/rpc_pipe_fs.h index 63878d05c9a9..63929349571f 100644 --- a/include/linux/sunrpc/rpc_pipe_fs.h +++ b/include/linux/sunrpc/rpc_pipe_fs.h @@ -41,11 +41,10 @@ RPC_I(struct inode *inode) extern int rpc_queue_upcall(struct inode *, struct rpc_pipe_msg *); -extern struct dentry *rpc_mkdir(struct dentry *, char *, struct rpc_clnt *); -extern void rpc_rmdir(struct dentry *); -extern struct dentry *rpc_mkpipe(struct dentry *, char *, void *, - struct rpc_pipe_ops *, int flags); -extern void rpc_unlink(struct dentry *); +extern struct dentry *rpc_mkdir(char *, struct rpc_clnt *); +extern int rpc_rmdir(char *); +extern struct dentry *rpc_mkpipe(char *, void *, struct rpc_pipe_ops *, int flags); +extern int rpc_unlink(char *); #endif #endif diff --git a/net/sunrpc/auth_gss/auth_gss.c b/net/sunrpc/auth_gss/auth_gss.c index bd2555139fa9..d2b08f16c257 100644 --- a/net/sunrpc/auth_gss/auth_gss.c +++ b/net/sunrpc/auth_gss/auth_gss.c @@ -87,6 +87,7 @@ struct gss_auth { struct list_head upcalls; struct rpc_clnt *client; struct dentry *dentry; + char path[48]; spinlock_t lock; }; @@ -689,8 +690,10 @@ gss_create(struct rpc_clnt *clnt, rpc_authflavor_t flavor) if (err) goto err_put_mech; - gss_auth->dentry = rpc_mkpipe(clnt->cl_dentry, gss_auth->mech->gm_name, - clnt, &gss_upcall_ops, RPC_PIPE_WAIT_FOR_OPEN); + snprintf(gss_auth->path, sizeof(gss_auth->path), "%s/%s", + clnt->cl_pathname, + gss_auth->mech->gm_name); + gss_auth->dentry = rpc_mkpipe(gss_auth->path, clnt, &gss_upcall_ops, RPC_PIPE_WAIT_FOR_OPEN); if (IS_ERR(gss_auth->dentry)) { err = PTR_ERR(gss_auth->dentry); goto err_put_mech; @@ -715,7 +718,7 @@ gss_destroy(struct rpc_auth *auth) auth, auth->au_flavor); gss_auth = container_of(auth, struct gss_auth, rpc_auth); - rpc_unlink(gss_auth->dentry); + rpc_unlink(gss_auth->path); gss_mech_put(gss_auth->mech); rpcauth_free_credcache(auth); diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c index 63bf591310e0..5a8f01d726e9 100644 --- a/net/sunrpc/clnt.c +++ b/net/sunrpc/clnt.c @@ -67,42 +67,26 @@ static u32 * call_verify(struct rpc_task *task); static int rpc_setup_pipedir(struct rpc_clnt *clnt, char *dir_name) { - static unsigned int clntid; - char name[128]; + static uint32_t clntid; int error; if (dir_name == NULL) return 0; - - retry_parent: - clnt->__cl_parent_dentry = rpc_mkdir(NULL, dir_name, NULL); - if (IS_ERR(clnt->__cl_parent_dentry)) { - error = PTR_ERR(clnt->__cl_parent_dentry); - if (error == -EEXIST) - goto retry_parent; /* XXX(hch): WTF? */ - - printk(KERN_INFO "RPC: Couldn't create pipefs entry %s, error %d\n", - dir_name, error); - return error; - } - - - retry_child: - snprintf(name, sizeof(name), "clnt%x", clntid++); - name[sizeof(name) - 1] = '\0'; - - clnt->cl_dentry = rpc_mkdir(clnt->__cl_parent_dentry, name, clnt); - if (IS_ERR(clnt->cl_dentry)) { + for (;;) { + snprintf(clnt->cl_pathname, sizeof(clnt->cl_pathname), + "%s/clnt%x", dir_name, + (unsigned int)clntid++); + clnt->cl_pathname[sizeof(clnt->cl_pathname) - 1] = '\0'; + clnt->cl_dentry = rpc_mkdir(clnt->cl_pathname, clnt); + if (!IS_ERR(clnt->cl_dentry)) + return 0; error = PTR_ERR(clnt->cl_dentry); - if (error == -EEXIST) - goto retry_child; - printk(KERN_INFO "RPC: Couldn't create pipefs entry %s, error %d\n", - name, error); - rpc_rmdir(clnt->__cl_parent_dentry); - return error; + if (error != -EEXIST) { + printk(KERN_INFO "RPC: Couldn't create pipefs entry %s, error %d\n", + clnt->cl_pathname, error); + return error; + } } - - return 0; } /* @@ -190,8 +174,7 @@ rpc_new_client(struct rpc_xprt *xprt, char *servname, return clnt; out_no_auth: - rpc_rmdir(clnt->cl_dentry); - rpc_rmdir(clnt->__cl_parent_dentry); + rpc_rmdir(clnt->cl_pathname); out_no_path: if (clnt->cl_server != clnt->cl_inline_name) kfree(clnt->cl_server); @@ -319,10 +302,8 @@ rpc_destroy_client(struct rpc_clnt *clnt) rpc_destroy_client(clnt->cl_parent); goto out_free; } - if (clnt->cl_dentry) - rpc_rmdir(clnt->cl_dentry); - if (clnt->__cl_parent_dentry) - rpc_rmdir(clnt->__cl_parent_dentry); + if (clnt->cl_pathname[0]) + rpc_rmdir(clnt->cl_pathname); if (clnt->cl_xprt) { xprt_destroy(clnt->cl_xprt); clnt->cl_xprt = NULL; diff --git a/net/sunrpc/rpc_pipe.c b/net/sunrpc/rpc_pipe.c index b382809726d8..ded6c63f11ec 100644 --- a/net/sunrpc/rpc_pipe.c +++ b/net/sunrpc/rpc_pipe.c @@ -414,6 +414,38 @@ rpc_put_mount(void) simple_release_fs(&rpc_mount, &rpc_mount_count); } +static int +rpc_lookup_parent(char *path, struct nameidata *nd) +{ + if (path[0] == '\0') + return -ENOENT; + if (rpc_get_mount()) { + printk(KERN_WARNING "%s: %s failed to mount " + "pseudofilesystem \n", __FILE__, __FUNCTION__); + return -ENODEV; + } + nd->mnt = mntget(rpc_mount); + nd->dentry = dget(rpc_mount->mnt_root); + nd->last_type = LAST_ROOT; + nd->flags = LOOKUP_PARENT; + nd->depth = 0; + + if (path_walk(path, nd)) { + printk(KERN_WARNING "%s: %s failed to find path %s\n", + __FILE__, __FUNCTION__, path); + rpc_put_mount(); + return -ENOENT; + } + return 0; +} + +static void +rpc_release_path(struct nameidata *nd) +{ + path_release(nd); + rpc_put_mount(); +} + static struct inode * rpc_get_inode(struct super_block *sb, int mode) { @@ -518,149 +550,197 @@ out_bad: return -ENOMEM; } -struct dentry * -rpc_mkdir(struct dentry *parent, char *name, struct rpc_clnt *rpc_client) +static int +__rpc_mkdir(struct inode *dir, struct dentry *dentry) { - struct inode *dir; - struct dentry *dentry; struct inode *inode; + + inode = rpc_get_inode(dir->i_sb, S_IFDIR | S_IRUSR | S_IXUSR); + if (!inode) + goto out_err; + inode->i_ino = iunique(dir->i_sb, 100); + d_instantiate(dentry, inode); + dir->i_nlink++; + inode_dir_notify(dir, DN_CREATE); + rpc_get_mount(); + return 0; +out_err: + printk(KERN_WARNING "%s: %s failed to allocate inode for dentry %s\n", + __FILE__, __FUNCTION__, dentry->d_name.name); + return -ENOMEM; +} + +static int +__rpc_rmdir(struct inode *dir, struct dentry *dentry) +{ int error; - if (!parent) - parent = rpc_mount->mnt_root; + shrink_dcache_parent(dentry); + if (dentry->d_inode) { + rpc_close_pipes(dentry->d_inode); + rpc_inode_setowner(dentry->d_inode, NULL); + } + if ((error = simple_rmdir(dir, dentry)) != 0) + return error; + if (!error) { + inode_dir_notify(dir, DN_DELETE); + d_drop(dentry); + rpc_put_mount(); + } + return 0; +} - dir = parent->d_inode; - - error = rpc_get_mount(); - if (error) - return ERR_PTR(error); +static struct dentry * +rpc_lookup_negative(char *path, struct nameidata *nd) +{ + struct dentry *dentry; + struct inode *dir; + int error; + if ((error = rpc_lookup_parent(path, nd)) != 0) + return ERR_PTR(error); + dir = nd->dentry->d_inode; down(&dir->i_sem); - dentry = lookup_one_len(name, parent, strlen(name)); + dentry = lookup_hash(&nd->last, nd->dentry); if (IS_ERR(dentry)) - goto out_unlock; + goto out_err; if (dentry->d_inode) { + dput(dentry); dentry = ERR_PTR(-EEXIST); - goto out_dput; + goto out_err; } - - inode = rpc_get_inode(dir->i_sb, S_IFDIR | S_IRUSR | S_IXUSR); - if (!inode) - goto out_dput; - inode->i_ino = iunique(dir->i_sb, 100); - dir->i_nlink++; - RPC_I(dentry->d_inode)->private = rpc_client; - - d_instantiate(dentry, inode); - dget(dentry); + return dentry; +out_err: up(&dir->i_sem); + rpc_release_path(nd); + return dentry; +} - inode_dir_notify(dir, DN_CREATE); +struct dentry * +rpc_mkdir(char *path, struct rpc_clnt *rpc_client) +{ + struct nameidata nd; + struct dentry *dentry; + struct inode *dir; + int error; + + dentry = rpc_lookup_negative(path, &nd); + if (IS_ERR(dentry)) + return dentry; + dir = nd.dentry->d_inode; + if ((error = __rpc_mkdir(dir, dentry)) != 0) + goto err_dput; + RPC_I(dentry->d_inode)->private = rpc_client; error = rpc_populate(dentry, authfiles, RPCAUTH_info, RPCAUTH_EOF); if (error) - goto out_depopulate; - - return dentry; - - out_depopulate: - rpc_rmdir(dentry); - out_dput: - dput(dentry); - out_unlock: + goto err_depopulate; +out: up(&dir->i_sem); - rpc_put_mount(); + rpc_release_path(&nd); return dentry; +err_depopulate: + rpc_depopulate(dentry); + __rpc_rmdir(dir, dentry); +err_dput: + dput(dentry); + printk(KERN_WARNING "%s: %s() failed to create directory %s (errno = %d)\n", + __FILE__, __FUNCTION__, path, error); + dentry = ERR_PTR(error); + goto out; } -void -rpc_rmdir(struct dentry *dentry) +int +rpc_rmdir(char *path) { - struct dentry *parent = dentry->d_parent; - - rpc_depopulate(dentry); + struct nameidata nd; + struct dentry *dentry; + struct inode *dir; + int error; - down(&parent->d_inode->i_sem); - if (dentry->d_inode) { - rpc_close_pipes(dentry->d_inode); - rpc_inode_setowner(dentry->d_inode, NULL); - simple_rmdir(parent->d_inode, dentry); + if ((error = rpc_lookup_parent(path, &nd)) != 0) + return error; + dir = nd.dentry->d_inode; + down(&dir->i_sem); + dentry = lookup_hash(&nd.last, nd.dentry); + if (IS_ERR(dentry)) { + error = PTR_ERR(dentry); + goto out_release; } - up(&parent->d_inode->i_sem); - - inode_dir_notify(parent->d_inode, DN_DELETE); - rpc_put_mount(); + rpc_depopulate(dentry); + error = __rpc_rmdir(dir, dentry); + dput(dentry); +out_release: + up(&dir->i_sem); + rpc_release_path(&nd); + return error; } struct dentry * -rpc_mkpipe(struct dentry *parent, char *name, void *private, - struct rpc_pipe_ops *ops, int flags) +rpc_mkpipe(char *path, void *private, struct rpc_pipe_ops *ops, int flags) { - struct inode *dir = parent->d_inode; + struct nameidata nd; struct dentry *dentry; - struct inode *inode; + struct inode *dir, *inode; struct rpc_inode *rpci; - int error; - - error = rpc_get_mount(); - if (error) - return ERR_PTR(error); - down(&parent->d_inode->i_sem); - dentry = lookup_one_len(name, parent, strlen(name)); + dentry = rpc_lookup_negative(path, &nd); if (IS_ERR(dentry)) - goto out_unlock; - if (dentry->d_inode) { - dentry = ERR_PTR(-EEXIST); - goto out_dput; - } - - inode = rpc_get_inode(parent->d_inode->i_sb, - S_IFSOCK | S_IRUSR | S_IWUSR); - if (!inode) { - dentry = ERR_PTR(-ENOMEM); - goto out_dput; - } - + return dentry; + dir = nd.dentry->d_inode; + inode = rpc_get_inode(dir->i_sb, S_IFSOCK | S_IRUSR | S_IWUSR); + if (!inode) + goto err_dput; inode->i_ino = iunique(dir->i_sb, 100); inode->i_fop = &rpc_pipe_fops; - + d_instantiate(dentry, inode); rpci = RPC_I(inode); rpci->private = private; rpci->flags = flags; rpci->ops = ops; - - d_instantiate(dentry, inode); - dget(dentry); - up(&parent->d_inode->i_sem); - inode_dir_notify(dir, DN_CREATE); +out: + up(&dir->i_sem); + rpc_release_path(&nd); return dentry; - - out_dput: +err_dput: dput(dentry); - out_unlock: - up(&parent->d_inode->i_sem); - rpc_put_mount(); - return dentry; + dentry = ERR_PTR(-ENOMEM); + printk(KERN_WARNING "%s: %s() failed to create pipe %s (errno = %d)\n", + __FILE__, __FUNCTION__, path, -ENOMEM); + goto out; } -void -rpc_unlink(struct dentry *dentry) +int +rpc_unlink(char *path) { - struct dentry *parent = dentry->d_parent; + struct nameidata nd; + struct dentry *dentry; + struct inode *dir; + int error; - down(&parent->d_inode->i_sem); + if ((error = rpc_lookup_parent(path, &nd)) != 0) + return error; + dir = nd.dentry->d_inode; + down(&dir->i_sem); + dentry = lookup_hash(&nd.last, nd.dentry); + if (IS_ERR(dentry)) { + error = PTR_ERR(dentry); + goto out_release; + } + d_drop(dentry); if (dentry->d_inode) { rpc_close_pipes(dentry->d_inode); rpc_inode_setowner(dentry->d_inode, NULL); - simple_unlink(parent->d_inode, dentry); + error = simple_unlink(dir, dentry); } - up(&parent->d_inode->i_sem); - - inode_dir_notify(parent->d_inode, DN_DELETE); - rpc_put_mount(); + dput(dentry); + inode_dir_notify(dir, DN_DELETE); +out_release: + up(&dir->i_sem); + rpc_release_path(&nd); + return error; } /* -- cgit v1.2.3 From 14be71f4c5c5ad1e222c5202ee6d234e9c8828b7 Mon Sep 17 00:00:00 2001 From: Albert Lee Date: Tue, 27 Sep 2005 17:36:35 +0800 Subject: [PATCH] libata: rename host states Changes: s/PIO_ST_/HSM_ST_/ and s/pio_task_state/hsm_task_state/. Signed-off-by: Albert Lee Signed-off-by: Jeff Garzik --- drivers/scsi/libata-core.c | 78 +++++++++++++++++++++++----------------------- include/linux/libata.h | 20 ++++++------ 2 files changed, 49 insertions(+), 49 deletions(-) (limited to 'include/linux') diff --git a/drivers/scsi/libata-core.c b/drivers/scsi/libata-core.c index cc68f5706acf..c4fcdc30f18c 100644 --- a/drivers/scsi/libata-core.c +++ b/drivers/scsi/libata-core.c @@ -2425,20 +2425,20 @@ void ata_poll_qc_complete(struct ata_queued_cmd *qc, u8 drv_stat) static unsigned long ata_pio_poll(struct ata_port *ap) { u8 status; - unsigned int poll_state = PIO_ST_UNKNOWN; - unsigned int reg_state = PIO_ST_UNKNOWN; - const unsigned int tmout_state = PIO_ST_TMOUT; - - switch (ap->pio_task_state) { - case PIO_ST: - case PIO_ST_POLL: - poll_state = PIO_ST_POLL; - reg_state = PIO_ST; + unsigned int poll_state = HSM_ST_UNKNOWN; + unsigned int reg_state = HSM_ST_UNKNOWN; + const unsigned int tmout_state = HSM_ST_TMOUT; + + switch (ap->hsm_task_state) { + case HSM_ST: + case HSM_ST_POLL: + poll_state = HSM_ST_POLL; + reg_state = HSM_ST; break; - case PIO_ST_LAST: - case PIO_ST_LAST_POLL: - poll_state = PIO_ST_LAST_POLL; - reg_state = PIO_ST_LAST; + case HSM_ST_LAST: + case HSM_ST_LAST_POLL: + poll_state = HSM_ST_LAST_POLL; + reg_state = HSM_ST_LAST; break; default: BUG(); @@ -2448,14 +2448,14 @@ static unsigned long ata_pio_poll(struct ata_port *ap) status = ata_chk_status(ap); if (status & ATA_BUSY) { if (time_after(jiffies, ap->pio_task_timeout)) { - ap->pio_task_state = tmout_state; + ap->hsm_task_state = tmout_state; return 0; } - ap->pio_task_state = poll_state; + ap->hsm_task_state = poll_state; return ATA_SHORT_PAUSE; } - ap->pio_task_state = reg_state; + ap->hsm_task_state = reg_state; return 0; } @@ -2480,14 +2480,14 @@ static int ata_pio_complete (struct ata_port *ap) * we enter, BSY will be cleared in a chk-status or two. If not, * the drive is probably seeking or something. Snooze for a couple * msecs, then chk-status again. If still busy, fall back to - * PIO_ST_POLL state. + * HSM_ST_POLL state. */ drv_stat = ata_busy_wait(ap, ATA_BUSY | ATA_DRQ, 10); if (drv_stat & (ATA_BUSY | ATA_DRQ)) { msleep(2); drv_stat = ata_busy_wait(ap, ATA_BUSY | ATA_DRQ, 10); if (drv_stat & (ATA_BUSY | ATA_DRQ)) { - ap->pio_task_state = PIO_ST_LAST_POLL; + ap->hsm_task_state = HSM_ST_LAST_POLL; ap->pio_task_timeout = jiffies + ATA_TMOUT_PIO; return 0; } @@ -2495,14 +2495,14 @@ static int ata_pio_complete (struct ata_port *ap) drv_stat = ata_wait_idle(ap); if (!ata_ok(drv_stat)) { - ap->pio_task_state = PIO_ST_ERR; + ap->hsm_task_state = HSM_ST_ERR; return 0; } qc = ata_qc_from_tag(ap, ap->active_tag); assert(qc != NULL); - ap->pio_task_state = PIO_ST_IDLE; + ap->hsm_task_state = HSM_ST_IDLE; ata_poll_qc_complete(qc, drv_stat); @@ -2662,7 +2662,7 @@ static void ata_pio_sector(struct ata_queued_cmd *qc) unsigned char *buf; if (qc->cursect == (qc->nsect - 1)) - ap->pio_task_state = PIO_ST_LAST; + ap->hsm_task_state = HSM_ST_LAST; page = sg[qc->cursg].page; offset = sg[qc->cursg].offset + qc->cursg_ofs * ATA_SECT_SIZE; @@ -2712,7 +2712,7 @@ static void __atapi_pio_bytes(struct ata_queued_cmd *qc, unsigned int bytes) unsigned int offset, count; if (qc->curbytes + bytes >= qc->nbytes) - ap->pio_task_state = PIO_ST_LAST; + ap->hsm_task_state = HSM_ST_LAST; next_sg: if (unlikely(qc->cursg >= qc->n_elem)) { @@ -2734,7 +2734,7 @@ next_sg: for (i = 0; i < words; i++) ata_data_xfer(ap, (unsigned char*)pad_buf, 2, do_write); - ap->pio_task_state = PIO_ST_LAST; + ap->hsm_task_state = HSM_ST_LAST; return; } @@ -2815,7 +2815,7 @@ static void atapi_pio_bytes(struct ata_queued_cmd *qc) err_out: printk(KERN_INFO "ata%u: dev %u: ATAPI check failed\n", ap->id, dev->devno); - ap->pio_task_state = PIO_ST_ERR; + ap->hsm_task_state = HSM_ST_ERR; } /** @@ -2837,14 +2837,14 @@ static void ata_pio_block(struct ata_port *ap) * a chk-status or two. If not, the drive is probably seeking * or something. Snooze for a couple msecs, then * chk-status again. If still busy, fall back to - * PIO_ST_POLL state. + * HSM_ST_POLL state. */ status = ata_busy_wait(ap, ATA_BUSY, 5); if (status & ATA_BUSY) { msleep(2); status = ata_busy_wait(ap, ATA_BUSY, 10); if (status & ATA_BUSY) { - ap->pio_task_state = PIO_ST_POLL; + ap->hsm_task_state = HSM_ST_POLL; ap->pio_task_timeout = jiffies + ATA_TMOUT_PIO; return; } @@ -2856,7 +2856,7 @@ static void ata_pio_block(struct ata_port *ap) if (is_atapi_taskfile(&qc->tf)) { /* no more data to transfer or unsupported ATAPI command */ if ((status & ATA_DRQ) == 0) { - ap->pio_task_state = PIO_ST_LAST; + ap->hsm_task_state = HSM_ST_LAST; return; } @@ -2864,7 +2864,7 @@ static void ata_pio_block(struct ata_port *ap) } else { /* handle BSY=0, DRQ=0 as error */ if ((status & ATA_DRQ) == 0) { - ap->pio_task_state = PIO_ST_ERR; + ap->hsm_task_state = HSM_ST_ERR; return; } @@ -2884,7 +2884,7 @@ static void ata_pio_error(struct ata_port *ap) printk(KERN_WARNING "ata%u: PIO error, drv_stat 0x%x\n", ap->id, drv_stat); - ap->pio_task_state = PIO_ST_IDLE; + ap->hsm_task_state = HSM_ST_IDLE; ata_poll_qc_complete(qc, drv_stat | ATA_ERR); } @@ -2899,25 +2899,25 @@ fsm_start: timeout = 0; qc_completed = 0; - switch (ap->pio_task_state) { - case PIO_ST_IDLE: + switch (ap->hsm_task_state) { + case HSM_ST_IDLE: return; - case PIO_ST: + case HSM_ST: ata_pio_block(ap); break; - case PIO_ST_LAST: + case HSM_ST_LAST: qc_completed = ata_pio_complete(ap); break; - case PIO_ST_POLL: - case PIO_ST_LAST_POLL: + case HSM_ST_POLL: + case HSM_ST_LAST_POLL: timeout = ata_pio_poll(ap); break; - case PIO_ST_TMOUT: - case PIO_ST_ERR: + case HSM_ST_TMOUT: + case HSM_ST_ERR: ata_pio_error(ap); return; } @@ -3360,7 +3360,7 @@ int ata_qc_issue_prot(struct ata_queued_cmd *qc) case ATA_PROT_PIO: /* load tf registers, initiate polling pio */ ata_qc_set_polling(qc); ata_tf_to_host_nolock(ap, &qc->tf); - ap->pio_task_state = PIO_ST; + ap->hsm_task_state = HSM_ST; queue_work(ata_wq, &ap->pio_task); break; @@ -3806,7 +3806,7 @@ static void atapi_packet_task(void *_data) ata_data_xfer(ap, qc->cdb, ap->cdb_len, 1); /* PIO commands are handled by polling */ - ap->pio_task_state = PIO_ST; + ap->hsm_task_state = HSM_ST; queue_work(ata_wq, &ap->pio_task); } diff --git a/include/linux/libata.h b/include/linux/libata.h index ceee1fc42c60..bb2d916bce44 100644 --- a/include/linux/libata.h +++ b/include/linux/libata.h @@ -156,15 +156,15 @@ enum { ATA_SHIFT_PIO = 11, }; -enum pio_task_states { - PIO_ST_UNKNOWN, - PIO_ST_IDLE, - PIO_ST_POLL, - PIO_ST_TMOUT, - PIO_ST, - PIO_ST_LAST, - PIO_ST_LAST_POLL, - PIO_ST_ERR, +enum hsm_task_states { + HSM_ST_UNKNOWN, + HSM_ST_IDLE, + HSM_ST_POLL, + HSM_ST_TMOUT, + HSM_ST, + HSM_ST_LAST, + HSM_ST_LAST_POLL, + HSM_ST_ERR, }; /* forward declarations */ @@ -319,7 +319,7 @@ struct ata_port { struct work_struct packet_task; struct work_struct pio_task; - unsigned int pio_task_state; + unsigned int hsm_task_state; unsigned long pio_task_timeout; void *private_data; -- cgit v1.2.3 From b4b52db71529bbe46da914eda772fb574914c94d Mon Sep 17 00:00:00 2001 From: Alan Cox Date: Mon, 26 Sep 2005 12:48:41 +0100 Subject: [PATCH] ata: re-order speeds sensibly. Signed-off-by: Jeff Garzik --- include/linux/ata.h | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'include/linux') diff --git a/include/linux/ata.h b/include/linux/ata.h index 85169ea9eb01..ecb7346d0c16 100644 --- a/include/linux/ata.h +++ b/include/linux/ata.h @@ -147,14 +147,14 @@ enum { XFER_MW_DMA_2 = 0x22, XFER_MW_DMA_1 = 0x21, XFER_MW_DMA_0 = 0x20, + XFER_SW_DMA_2 = 0x12, + XFER_SW_DMA_1 = 0x11, + XFER_SW_DMA_0 = 0x10, XFER_PIO_4 = 0x0C, XFER_PIO_3 = 0x0B, XFER_PIO_2 = 0x0A, XFER_PIO_1 = 0x09, XFER_PIO_0 = 0x08, - XFER_SW_DMA_2 = 0x12, - XFER_SW_DMA_1 = 0x11, - XFER_SW_DMA_0 = 0x10, XFER_PIO_SLOW = 0x00, /* ATAPI stuff */ -- cgit v1.2.3 From 47a8659380d40d5c0786ddb62a89b3f7f1392430 Mon Sep 17 00:00:00 2001 From: Alan Cox Date: Tue, 4 Oct 2005 08:09:19 -0400 Subject: libata: bitmask based pci init functions for one or two ports This redoes the n_ports logic I proposed before as a bitmask. ata_pci_init_native_mode is now used with a mask allowing for mixed mode stuff later on. ata_pci_init_legacy_port is called with port number and does one port now not two. Instead it is called twice by the ata init logic which cleans both of them up. There are stil limits in the original code left over - IRQ/port mapping for legacy mode should be arch specific values - You can have one legacy mode IDE adapter per PCI root bridge on some systems - Doesn't handle mixed mode devices yet (but is now a lot closer to it) --- drivers/scsi/libata-core.c | 133 ++++++++++++++++++++++++++------------------- drivers/scsi/sata_nv.c | 2 +- drivers/scsi/sata_sis.c | 2 +- drivers/scsi/sata_uli.c | 2 +- drivers/scsi/sata_via.c | 2 +- include/linux/libata.h | 6 +- 6 files changed, 85 insertions(+), 62 deletions(-) (limited to 'include/linux') diff --git a/drivers/scsi/libata-core.c b/drivers/scsi/libata-core.c index 03d422e99e58..82ec7f30bf3f 100644 --- a/drivers/scsi/libata-core.c +++ b/drivers/scsi/libata-core.c @@ -4363,85 +4363,87 @@ void ata_pci_host_stop (struct ata_host_set *host_set) * ata_pci_init_native_mode - Initialize native-mode driver * @pdev: pci device to be initialized * @port: array[2] of pointers to port info structures. + * @ports: bitmap of ports present * * Utility function which allocates and initializes an * ata_probe_ent structure for a standard dual-port * PIO-based IDE controller. The returned ata_probe_ent * structure can be passed to ata_device_add(). The returned * ata_probe_ent structure should then be freed with kfree(). + * + * The caller need only pass the address of the primary port, the + * secondary will be deduced automatically. If the device has non + * standard secondary port mappings this function can be called twice, + * once for each interface. */ struct ata_probe_ent * -ata_pci_init_native_mode(struct pci_dev *pdev, struct ata_port_info **port) +ata_pci_init_native_mode(struct pci_dev *pdev, struct ata_port_info **port, int ports) { struct ata_probe_ent *probe_ent = ata_probe_ent_alloc(pci_dev_to_dev(pdev), port[0]); + int p = 0; + if (!probe_ent) return NULL; - probe_ent->n_ports = 2; probe_ent->irq = pdev->irq; probe_ent->irq_flags = SA_SHIRQ; - probe_ent->port[0].cmd_addr = pci_resource_start(pdev, 0); - probe_ent->port[0].altstatus_addr = - probe_ent->port[0].ctl_addr = - pci_resource_start(pdev, 1) | ATA_PCI_CTL_OFS; - probe_ent->port[0].bmdma_addr = pci_resource_start(pdev, 4); - - probe_ent->port[1].cmd_addr = pci_resource_start(pdev, 2); - probe_ent->port[1].altstatus_addr = - probe_ent->port[1].ctl_addr = - pci_resource_start(pdev, 3) | ATA_PCI_CTL_OFS; - probe_ent->port[1].bmdma_addr = pci_resource_start(pdev, 4) + 8; + if (ports & ATA_PORT_PRIMARY) { + probe_ent->port[p].cmd_addr = pci_resource_start(pdev, 0); + probe_ent->port[p].altstatus_addr = + probe_ent->port[p].ctl_addr = + pci_resource_start(pdev, 1) | ATA_PCI_CTL_OFS; + probe_ent->port[p].bmdma_addr = pci_resource_start(pdev, 4); + ata_std_ports(&probe_ent->port[p]); + p++; + } - ata_std_ports(&probe_ent->port[0]); - ata_std_ports(&probe_ent->port[1]); + if (ports & ATA_PORT_SECONDARY) { + probe_ent->port[p].cmd_addr = pci_resource_start(pdev, 2); + probe_ent->port[p].altstatus_addr = + probe_ent->port[p].ctl_addr = + pci_resource_start(pdev, 3) | ATA_PCI_CTL_OFS; + probe_ent->port[p].bmdma_addr = pci_resource_start(pdev, 4) + 8; + ata_std_ports(&probe_ent->port[p]); + p++; + } + probe_ent->n_ports = p; return probe_ent; } -static struct ata_probe_ent * -ata_pci_init_legacy_mode(struct pci_dev *pdev, struct ata_port_info **port, - struct ata_probe_ent **ppe2) +static struct ata_probe_ent *ata_pci_init_legacy_port(struct pci_dev *pdev, struct ata_port_info **port, int port_num) { - struct ata_probe_ent *probe_ent, *probe_ent2; + struct ata_probe_ent *probe_ent; probe_ent = ata_probe_ent_alloc(pci_dev_to_dev(pdev), port[0]); if (!probe_ent) return NULL; - probe_ent2 = ata_probe_ent_alloc(pci_dev_to_dev(pdev), port[1]); - if (!probe_ent2) { - kfree(probe_ent); - return NULL; - } - - probe_ent->n_ports = 1; - probe_ent->irq = 14; - probe_ent->hard_port_no = 0; + probe_ent->legacy_mode = 1; - - probe_ent2->n_ports = 1; - probe_ent2->irq = 15; - - probe_ent2->hard_port_no = 1; - probe_ent2->legacy_mode = 1; - - probe_ent->port[0].cmd_addr = 0x1f0; - probe_ent->port[0].altstatus_addr = - probe_ent->port[0].ctl_addr = 0x3f6; - probe_ent->port[0].bmdma_addr = pci_resource_start(pdev, 4); - - probe_ent2->port[0].cmd_addr = 0x170; - probe_ent2->port[0].altstatus_addr = - probe_ent2->port[0].ctl_addr = 0x376; - probe_ent2->port[0].bmdma_addr = pci_resource_start(pdev, 4)+8; - + probe_ent->n_ports = 1; + probe_ent->hard_port_no = port_num; + + switch(port_num) + { + case 0: + probe_ent->irq = 14; + probe_ent->port[0].cmd_addr = 0x1f0; + probe_ent->port[0].altstatus_addr = + probe_ent->port[0].ctl_addr = 0x3f6; + break; + case 1: + probe_ent->irq = 15; + probe_ent->port[0].cmd_addr = 0x170; + probe_ent->port[0].altstatus_addr = + probe_ent->port[0].ctl_addr = 0x376; + break; + } + probe_ent->port[0].bmdma_addr = pci_resource_start(pdev, 4) + 8 * port_num; ata_std_ports(&probe_ent->port[0]); - ata_std_ports(&probe_ent2->port[0]); - - *ppe2 = probe_ent2; return probe_ent; } @@ -4470,7 +4472,7 @@ ata_pci_init_legacy_mode(struct pci_dev *pdev, struct ata_port_info **port, int ata_pci_init_one (struct pci_dev *pdev, struct ata_port_info **port_info, unsigned int n_ports) { - struct ata_probe_ent *probe_ent, *probe_ent2 = NULL; + struct ata_probe_ent *probe_ent = NULL, *probe_ent2 = NULL; struct ata_port_info *port[2]; u8 tmp8, mask; unsigned int legacy_mode = 0; @@ -4487,7 +4489,7 @@ int ata_pci_init_one (struct pci_dev *pdev, struct ata_port_info **port_info, if ((port[0]->host_flags & ATA_FLAG_NO_LEGACY) == 0 && (pdev->class >> 8) == PCI_CLASS_STORAGE_IDE) { - /* TODO: support transitioning to native mode? */ + /* TODO: What if one channel is in native mode ... */ pci_read_config_byte(pdev, PCI_CLASS_PROG, &tmp8); mask = (1 << 2) | (1 << 0); if ((tmp8 & mask) != mask) @@ -4495,11 +4497,20 @@ int ata_pci_init_one (struct pci_dev *pdev, struct ata_port_info **port_info, } /* FIXME... */ - if ((!legacy_mode) && (n_ports > 1)) { - printk(KERN_ERR "ata: BUG: native mode, n_ports > 1\n"); - return -EINVAL; + if ((!legacy_mode) && (n_ports > 2)) { + printk(KERN_ERR "ata: BUG: native mode, n_ports > 2\n"); + n_ports = 2; + /* For now */ } + /* FIXME: Really for ATA it isn't safe because the device may be + multi-purpose and we want to leave it alone if it was already + enabled. Secondly for shared use as Arjan says we want refcounting + + Checking dev->is_enabled is insufficient as this is not set at + boot for the primary video which is BIOS enabled + */ + rc = pci_enable_device(pdev); if (rc) return rc; @@ -4510,6 +4521,7 @@ int ata_pci_init_one (struct pci_dev *pdev, struct ata_port_info **port_info, goto err_out; } + /* FIXME: Should use platform specific mappers for legacy port ranges */ if (legacy_mode) { if (!request_region(0x1f0, 8, "libata")) { struct resource *conflict, res; @@ -4554,10 +4566,17 @@ int ata_pci_init_one (struct pci_dev *pdev, struct ata_port_info **port_info, goto err_out_regions; if (legacy_mode) { - probe_ent = ata_pci_init_legacy_mode(pdev, port, &probe_ent2); - } else - probe_ent = ata_pci_init_native_mode(pdev, port); - if (!probe_ent) { + if (legacy_mode & (1 << 0)) + probe_ent = ata_pci_init_legacy_port(pdev, port, 0); + if (legacy_mode & (1 << 1)) + probe_ent2 = ata_pci_init_legacy_port(pdev, port, 1); + } else { + if (n_ports == 2) + probe_ent = ata_pci_init_native_mode(pdev, port, ATA_PORT_PRIMARY | ATA_PORT_SECONDARY); + else + probe_ent = ata_pci_init_native_mode(pdev, port, ATA_PORT_PRIMARY); + } + if (!probe_ent && !probe_ent2) { rc = -ENOMEM; goto err_out_regions; } diff --git a/drivers/scsi/sata_nv.c b/drivers/scsi/sata_nv.c index c05653c7779d..749ff92d8c63 100644 --- a/drivers/scsi/sata_nv.c +++ b/drivers/scsi/sata_nv.c @@ -405,7 +405,7 @@ static int nv_init_one (struct pci_dev *pdev, const struct pci_device_id *ent) rc = -ENOMEM; ppi = &nv_port_info; - probe_ent = ata_pci_init_native_mode(pdev, &ppi); + probe_ent = ata_pci_init_native_mode(pdev, &ppi, ATA_PORT_PRIMARY | ATA_PORT_SECONDARY); if (!probe_ent) goto err_out_regions; diff --git a/drivers/scsi/sata_sis.c b/drivers/scsi/sata_sis.c index b227e51d12f4..0761a3234fcf 100644 --- a/drivers/scsi/sata_sis.c +++ b/drivers/scsi/sata_sis.c @@ -263,7 +263,7 @@ static int sis_init_one (struct pci_dev *pdev, const struct pci_device_id *ent) goto err_out_regions; ppi = &sis_port_info; - probe_ent = ata_pci_init_native_mode(pdev, &ppi); + probe_ent = ata_pci_init_native_mode(pdev, &ppi, ATA_PORT_PRIMARY | ATA_PORT_SECONDARY); if (!probe_ent) { rc = -ENOMEM; goto err_out_regions; diff --git a/drivers/scsi/sata_uli.c b/drivers/scsi/sata_uli.c index 4c9fb8b71be1..9c06f2abe7f7 100644 --- a/drivers/scsi/sata_uli.c +++ b/drivers/scsi/sata_uli.c @@ -202,7 +202,7 @@ static int uli_init_one (struct pci_dev *pdev, const struct pci_device_id *ent) goto err_out_regions; ppi = &uli_port_info; - probe_ent = ata_pci_init_native_mode(pdev, &ppi); + probe_ent = ata_pci_init_native_mode(pdev, &ppi, ATA_PORT_PRIMARY | ATA_PORT_SECONDARY); if (!probe_ent) { rc = -ENOMEM; goto err_out_regions; diff --git a/drivers/scsi/sata_via.c b/drivers/scsi/sata_via.c index 128b996b07b7..565872479b9a 100644 --- a/drivers/scsi/sata_via.c +++ b/drivers/scsi/sata_via.c @@ -212,7 +212,7 @@ static struct ata_probe_ent *vt6420_init_probe_ent(struct pci_dev *pdev) struct ata_probe_ent *probe_ent; struct ata_port_info *ppi = &svia_port_info; - probe_ent = ata_pci_init_native_mode(pdev, &ppi); + probe_ent = ata_pci_init_native_mode(pdev, &ppi, ATA_PORT_PRIMARY | ATA_PORT_SECONDARY); if (!probe_ent) return NULL; diff --git a/include/linux/libata.h b/include/linux/libata.h index 4739a75b983d..4d45179872cc 100644 --- a/include/linux/libata.h +++ b/include/linux/libata.h @@ -155,6 +155,10 @@ enum { ATA_SHIFT_UDMA = 0, ATA_SHIFT_MWDMA = 8, ATA_SHIFT_PIO = 11, + + /* Masks for port functions */ + ATA_PORT_PRIMARY = (1 << 0), + ATA_PORT_SECONDARY = (1 << 1), }; enum hsm_task_states { @@ -458,7 +462,7 @@ struct pci_bits { extern void ata_pci_host_stop (struct ata_host_set *host_set); extern struct ata_probe_ent * -ata_pci_init_native_mode(struct pci_dev *pdev, struct ata_port_info **port); +ata_pci_init_native_mode(struct pci_dev *pdev, struct ata_port_info **port, int portmask); extern int pci_test_config_bits(struct pci_dev *pdev, struct pci_bits *bits); #endif /* CONFIG_PCI */ -- cgit v1.2.3 From 67846b30171cc4d706125f630193a76a26bb334a Mon Sep 17 00:00:00 2001 From: Jeff Garzik Date: Wed, 5 Oct 2005 02:58:32 -0400 Subject: libata: add ata_ratelimit(), use it in AHCI driver irq handler --- drivers/scsi/ahci.c | 31 +++++++++++++++++++++++++------ drivers/scsi/libata-core.c | 23 +++++++++++++++++++++++ include/linux/libata.h | 2 ++ 3 files changed, 50 insertions(+), 6 deletions(-) (limited to 'include/linux') diff --git a/drivers/scsi/ahci.c b/drivers/scsi/ahci.c index c2c8fa828e24..5ec866b00479 100644 --- a/drivers/scsi/ahci.c +++ b/drivers/scsi/ahci.c @@ -672,17 +672,36 @@ static irqreturn_t ahci_interrupt (int irq, void *dev_instance, struct pt_regs * for (i = 0; i < host_set->n_ports; i++) { struct ata_port *ap; - u32 tmp; - VPRINTK("port %u\n", i); + if (!(irq_stat & (1 << i))) + continue; + ap = host_set->ports[i]; - tmp = irq_stat & (1 << i); - if (tmp && ap) { + if (ap) { struct ata_queued_cmd *qc; qc = ata_qc_from_tag(ap, ap->active_tag); - if (ahci_host_intr(ap, qc)) - irq_ack |= (1 << i); + if (!ahci_host_intr(ap, qc)) + if (ata_ratelimit()) { + struct pci_dev *pdev = + to_pci_dev(ap->host_set->dev); + printk(KERN_WARNING + "ahci(%s): unhandled interrupt on port %u\n", + pci_name(pdev), i); + } + + VPRINTK("port %u\n", i); + } else { + VPRINTK("port %u (no irq)\n", i); + if (ata_ratelimit()) { + struct pci_dev *pdev = + to_pci_dev(ap->host_set->dev); + printk(KERN_WARNING + "ahci(%s): interrupt on disabled port %u\n", + pci_name(pdev), i); + } } + + irq_ack |= (1 << i); } if (irq_ack) { diff --git a/drivers/scsi/libata-core.c b/drivers/scsi/libata-core.c index f0894bfa908b..ceffaef37d17 100644 --- a/drivers/scsi/libata-core.c +++ b/drivers/scsi/libata-core.c @@ -48,6 +48,7 @@ #include #include #include +#include #include #include "scsi.h" #include "scsi_priv.h" @@ -4688,6 +4689,27 @@ static void __exit ata_exit(void) module_init(ata_init); module_exit(ata_exit); +static unsigned long ratelimit_time; +static spinlock_t ata_ratelimit_lock = SPIN_LOCK_UNLOCKED; + +int ata_ratelimit(void) +{ + int rc; + unsigned long flags; + + spin_lock_irqsave(&ata_ratelimit_lock, flags); + + if (time_after(jiffies, ratelimit_time)) { + rc = 1; + ratelimit_time = jiffies + (HZ/5); + } else + rc = 0; + + spin_unlock_irqrestore(&ata_ratelimit_lock, flags); + + return rc; +} + /* * libata is essentially a library of internal helper functions for * low-level ATA host controller drivers. As such, the API/ABI is @@ -4729,6 +4751,7 @@ EXPORT_SYMBOL_GPL(sata_phy_reset); EXPORT_SYMBOL_GPL(__sata_phy_reset); EXPORT_SYMBOL_GPL(ata_bus_reset); EXPORT_SYMBOL_GPL(ata_port_disable); +EXPORT_SYMBOL_GPL(ata_ratelimit); EXPORT_SYMBOL_GPL(ata_scsi_ioctl); EXPORT_SYMBOL_GPL(ata_scsi_queuecmd); EXPORT_SYMBOL_GPL(ata_scsi_error); diff --git a/include/linux/libata.h b/include/linux/libata.h index 4d45179872cc..7929cfc9318d 100644 --- a/include/linux/libata.h +++ b/include/linux/libata.h @@ -410,6 +410,8 @@ extern int ata_scsi_queuecmd(struct scsi_cmnd *cmd, void (*done)(struct scsi_cmn extern int ata_scsi_error(struct Scsi_Host *host); extern int ata_scsi_release(struct Scsi_Host *host); extern unsigned int ata_host_intr(struct ata_port *ap, struct ata_queued_cmd *qc); +extern int ata_ratelimit(void); + /* * Default driver ops implementations */ -- cgit v1.2.3 From 8cbd6df1f0ce977ab7b61feffa59879bb5e0ed8f Mon Sep 17 00:00:00 2001 From: Albert Lee Date: Wed, 12 Oct 2005 15:06:27 +0800 Subject: [PATCH] libata CHS: calculate read/write commands and protocol on the fly (revise #6) - merge ata_prot_to_cmd() and ata_dev_set_protocol() as ata_rwcmd_protocol() - pave road for read/write multiple support - remove usage of pre-cached command and protocol values and call ata_rwcmd_protocol() instead Signed-off-by: Albert Lee ============== Signed-off-by: Jeff Garzik --- drivers/scsi/libata-core.c | 105 ++++++++++++++++----------------------------- drivers/scsi/libata-scsi.c | 13 +++--- drivers/scsi/libata.h | 1 + include/linux/ata.h | 4 ++ include/linux/libata.h | 6 +-- 5 files changed, 50 insertions(+), 79 deletions(-) (limited to 'include/linux') diff --git a/drivers/scsi/libata-core.c b/drivers/scsi/libata-core.c index 55d4dee133af..19d3d717faf6 100644 --- a/drivers/scsi/libata-core.c +++ b/drivers/scsi/libata-core.c @@ -616,79 +616,53 @@ void ata_tf_from_fis(u8 *fis, struct ata_taskfile *tf) tf->hob_nsect = fis[13]; } -/** - * ata_prot_to_cmd - determine which read/write opcodes to use - * @protocol: ATA_PROT_xxx taskfile protocol - * @lba48: true is lba48 is present - * - * Given necessary input, determine which read/write commands - * to use to transfer data. - * - * LOCKING: - * None. - */ -static int ata_prot_to_cmd(int protocol, int lba48) -{ - int rcmd = 0, wcmd = 0; - - switch (protocol) { - case ATA_PROT_PIO: - if (lba48) { - rcmd = ATA_CMD_PIO_READ_EXT; - wcmd = ATA_CMD_PIO_WRITE_EXT; - } else { - rcmd = ATA_CMD_PIO_READ; - wcmd = ATA_CMD_PIO_WRITE; - } - break; - - case ATA_PROT_DMA: - if (lba48) { - rcmd = ATA_CMD_READ_EXT; - wcmd = ATA_CMD_WRITE_EXT; - } else { - rcmd = ATA_CMD_READ; - wcmd = ATA_CMD_WRITE; - } - break; - - default: - return -1; - } - - return rcmd | (wcmd << 8); -} +static const u8 ata_rw_cmds[] = { + /* pio multi */ + ATA_CMD_READ_MULTI, + ATA_CMD_WRITE_MULTI, + ATA_CMD_READ_MULTI_EXT, + ATA_CMD_WRITE_MULTI_EXT, + /* pio */ + ATA_CMD_PIO_READ, + ATA_CMD_PIO_WRITE, + ATA_CMD_PIO_READ_EXT, + ATA_CMD_PIO_WRITE_EXT, + /* dma */ + ATA_CMD_READ, + ATA_CMD_WRITE, + ATA_CMD_READ_EXT, + ATA_CMD_WRITE_EXT +}; /** - * ata_dev_set_protocol - set taskfile protocol and r/w commands - * @dev: device to examine and configure + * ata_rwcmd_protocol - set taskfile r/w commands and protocol + * @qc: command to examine and configure * - * Examine the device configuration, after we have - * read the identify-device page and configured the - * data transfer mode. Set internal state related to - * the ATA taskfile protocol (pio, pio mult, dma, etc.) - * and calculate the proper read/write commands to use. + * Examine the device configuration and tf->flags to calculate + * the proper read/write commands and protocol to use. * * LOCKING: * caller. */ -static void ata_dev_set_protocol(struct ata_device *dev) +void ata_rwcmd_protocol(struct ata_queued_cmd *qc) { - int pio = (dev->flags & ATA_DFLAG_PIO); - int lba48 = (dev->flags & ATA_DFLAG_LBA48); - int proto, cmd; + struct ata_taskfile *tf = &qc->tf; + struct ata_device *dev = qc->dev; - if (pio) - proto = dev->xfer_protocol = ATA_PROT_PIO; - else - proto = dev->xfer_protocol = ATA_PROT_DMA; + int index, lba48, write; + + lba48 = (tf->flags & ATA_TFLAG_LBA48) ? 2 : 0; + write = (tf->flags & ATA_TFLAG_WRITE) ? 1 : 0; - cmd = ata_prot_to_cmd(proto, lba48); - if (cmd < 0) - BUG(); + if (dev->flags & ATA_DFLAG_PIO) { + tf->protocol = ATA_PROT_PIO; + index = dev->multi_count ? 0 : 4; + } else { + tf->protocol = ATA_PROT_DMA; + index = 8; + } - dev->read_cmd = cmd & 0xff; - dev->write_cmd = (cmd >> 8) & 0xff; + tf->command = ata_rw_cmds[index + lba48 + write]; } static const char * xfer_mode_str[] = { @@ -1641,7 +1615,7 @@ static void ata_host_set_dma(struct ata_port *ap, u8 xfer_mode, */ static void ata_set_mode(struct ata_port *ap) { - unsigned int i, xfer_shift; + unsigned int xfer_shift; u8 xfer_mode; int rc; @@ -1670,11 +1644,6 @@ static void ata_set_mode(struct ata_port *ap) if (ap->ops->post_set_mode) ap->ops->post_set_mode(ap); - for (i = 0; i < 2; i++) { - struct ata_device *dev = &ap->device[i]; - ata_dev_set_protocol(dev); - } - return; err_out: diff --git a/drivers/scsi/libata-scsi.c b/drivers/scsi/libata-scsi.c index ea7a4d8a6fc9..90bf22204668 100644 --- a/drivers/scsi/libata-scsi.c +++ b/drivers/scsi/libata-scsi.c @@ -742,15 +742,10 @@ static unsigned int ata_scsi_rw_xlat(struct ata_queued_cmd *qc, u8 *scsicmd) u32 n_block; tf->flags |= ATA_TFLAG_ISADDR | ATA_TFLAG_DEVICE; - tf->protocol = qc->dev->xfer_protocol; - if (scsicmd[0] == READ_10 || scsicmd[0] == READ_6 || - scsicmd[0] == READ_16) { - tf->command = qc->dev->read_cmd; - } else { - tf->command = qc->dev->write_cmd; + if (scsicmd[0] == WRITE_10 || scsicmd[0] == WRITE_6 || + scsicmd[0] == WRITE_16) tf->flags |= ATA_TFLAG_WRITE; - } /* Calculate the SCSI LBA and transfer length. */ switch (scsicmd[0]) { @@ -812,6 +807,8 @@ static unsigned int ata_scsi_rw_xlat(struct ata_queued_cmd *qc, u8 *scsicmd) tf->device |= (block >> 24) & 0xf; } + ata_rwcmd_protocol(qc); + qc->nsect = n_block; tf->nsect = n_block & 0xff; @@ -828,6 +825,8 @@ static unsigned int ata_scsi_rw_xlat(struct ata_queued_cmd *qc, u8 *scsicmd) if ((block >> 28) || (n_block > 256)) goto out_of_range; + ata_rwcmd_protocol(qc); + /* Convert LBA to CHS */ track = (u32)block / dev->sectors; cyl = track / dev->heads; diff --git a/drivers/scsi/libata.h b/drivers/scsi/libata.h index a18f2ac1d4a1..67d752ca8ae2 100644 --- a/drivers/scsi/libata.h +++ b/drivers/scsi/libata.h @@ -42,6 +42,7 @@ extern int atapi_enabled; extern int ata_qc_complete_noop(struct ata_queued_cmd *qc, u8 drv_stat); extern struct ata_queued_cmd *ata_qc_new_init(struct ata_port *ap, struct ata_device *dev); +extern void ata_rwcmd_protocol(struct ata_queued_cmd *qc); extern void ata_qc_free(struct ata_queued_cmd *qc); extern int ata_qc_issue(struct ata_queued_cmd *qc); extern int ata_check_atapi_dma(struct ata_queued_cmd *qc); diff --git a/include/linux/ata.h b/include/linux/ata.h index ecb7346d0c16..630908c9378b 100644 --- a/include/linux/ata.h +++ b/include/linux/ata.h @@ -128,6 +128,10 @@ enum { ATA_CMD_PIO_READ_EXT = 0x24, ATA_CMD_PIO_WRITE = 0x30, ATA_CMD_PIO_WRITE_EXT = 0x34, + ATA_CMD_READ_MULTI = 0xC4, + ATA_CMD_READ_MULTI_EXT = 0x29, + ATA_CMD_WRITE_MULTI = 0xC5, + ATA_CMD_WRITE_MULTI_EXT = 0x39, ATA_CMD_SET_FEATURES = 0xEF, ATA_CMD_PACKET = 0xA0, ATA_CMD_VERIFY = 0x40, diff --git a/include/linux/libata.h b/include/linux/libata.h index 7929cfc9318d..0261c55f3483 100644 --- a/include/linux/libata.h +++ b/include/linux/libata.h @@ -283,10 +283,8 @@ struct ata_device { u8 xfer_mode; unsigned int xfer_shift; /* ATA_SHIFT_xxx */ - /* cache info about current transfer mode */ - u8 xfer_protocol; /* taskfile xfer protocol */ - u8 read_cmd; /* opcode to use on read */ - u8 write_cmd; /* opcode to use on write */ + unsigned int multi_count; /* sectors count for + READ/WRITE MULTIPLE */ /* for CHS addressing */ u16 cylinders; /* Number of cylinders */ -- cgit v1.2.3 From 5e5ce5be6f0161d2a069a4f8a1154fe639c5c02f Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Tue, 18 Oct 2005 14:20:11 -0700 Subject: RPC: allow call_encode() to delay transmission of an RPC call. Currently, call_encode will cause the entire RPC call to abort if it returns an error. This is unnecessarily rigid, and gets in the way of attempts to allow the NFSv4 layer to order RPC calls that carry sequence ids. Signed-off-by: Trond Myklebust --- include/linux/sunrpc/xprt.h | 1 + net/sunrpc/clnt.c | 23 ++++++++++++----------- net/sunrpc/xprt.c | 8 ++++++++ 3 files changed, 21 insertions(+), 11 deletions(-) (limited to 'include/linux') diff --git a/include/linux/sunrpc/xprt.h b/include/linux/sunrpc/xprt.h index 99cad3ead81d..068e1fb0868b 100644 --- a/include/linux/sunrpc/xprt.h +++ b/include/linux/sunrpc/xprt.h @@ -211,6 +211,7 @@ int xprt_reserve_xprt(struct rpc_task *task); int xprt_reserve_xprt_cong(struct rpc_task *task); int xprt_prepare_transmit(struct rpc_task *task); void xprt_transmit(struct rpc_task *task); +void xprt_abort_transmit(struct rpc_task *task); int xprt_adjust_timeout(struct rpc_rqst *req); void xprt_release_xprt(struct rpc_xprt *xprt, struct rpc_task *task); void xprt_release_xprt_cong(struct rpc_xprt *xprt, struct rpc_task *task); diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c index a5f7029b1daa..534274056329 100644 --- a/net/sunrpc/clnt.c +++ b/net/sunrpc/clnt.c @@ -678,13 +678,11 @@ call_allocate(struct rpc_task *task) static void call_encode(struct rpc_task *task) { - struct rpc_clnt *clnt = task->tk_client; struct rpc_rqst *req = task->tk_rqstp; struct xdr_buf *sndbuf = &req->rq_snd_buf; struct xdr_buf *rcvbuf = &req->rq_rcv_buf; unsigned int bufsiz; kxdrproc_t encode; - int status; u32 *p; dprintk("RPC: %4d call_encode (status %d)\n", @@ -712,12 +710,9 @@ call_encode(struct rpc_task *task) rpc_exit(task, -EIO); return; } - if (encode && (status = rpcauth_wrap_req(task, encode, req, p, - task->tk_msg.rpc_argp)) < 0) { - printk(KERN_WARNING "%s: can't encode arguments: %d\n", - clnt->cl_protname, -status); - rpc_exit(task, status); - } + if (encode != NULL) + task->tk_status = rpcauth_wrap_req(task, encode, req, p, + task->tk_msg.rpc_argp); } /* @@ -865,10 +860,12 @@ call_transmit(struct rpc_task *task) if (task->tk_status != 0) return; /* Encode here so that rpcsec_gss can use correct sequence number. */ - if (!task->tk_rqstp->rq_bytes_sent) + if (task->tk_rqstp->rq_bytes_sent == 0) { call_encode(task); - if (task->tk_status < 0) - return; + /* Did the encode result in an error condition? */ + if (task->tk_status != 0) + goto out_nosend; + } xprt_transmit(task); if (task->tk_status < 0) return; @@ -876,6 +873,10 @@ call_transmit(struct rpc_task *task) task->tk_action = NULL; rpc_wake_up_task(task); } + return; +out_nosend: + /* release socket write lock before attempting to handle error */ + xprt_abort_transmit(task); } /* diff --git a/net/sunrpc/xprt.c b/net/sunrpc/xprt.c index 215be0d0ef6b..1ba55dc38b7a 100644 --- a/net/sunrpc/xprt.c +++ b/net/sunrpc/xprt.c @@ -709,6 +709,14 @@ out_unlock: return err; } +void +xprt_abort_transmit(struct rpc_task *task) +{ + struct rpc_xprt *xprt = task->tk_xprt; + + xprt_release_write(xprt, task); +} + /** * xprt_transmit - send an RPC request on a transport * @task: controlling RPC task -- cgit v1.2.3 From cee54fc944422c44e476736c045a9e8053cb0644 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Tue, 18 Oct 2005 14:20:12 -0700 Subject: NFSv4: Add functions to order RPC calls NFSv4 file state-changing functions such as OPEN, CLOSE, LOCK,... are all labelled with "sequence identifiers" in order to prevent the server from reordering RPC requests, as this could cause its file state to become out of sync with the client. Currently the NFS client code enforces this ordering locally using semaphores to restrict access to structures until the RPC call is done. This, of course, only works with synchronous RPC calls, since the user process must first grab the semaphore. By dropping semaphores, and instead teaching the RPC engine to hold the RPC calls until they are ready to be sent, we can extend this process to work nicely with asynchronous RPC calls too. This patch adds a new list called "rpc_sequence" that defines the order of the RPC calls to be sent. We add one such list for each state_owner. When an RPC call is ready to be sent, it checks if it is top of the rpc_sequence list. If so, it proceeds. If not, it goes back to sleep, and loops until it hits top of the list. Once the RPC call has completed, it can then bump the sequence id counter, and remove itself from the rpc_sequence list, and then wake up the next sleeper. Note that the state_owner sequence ids and lock_owner sequence ids are all indexed to the same rpc_sequence list, so OPEN, LOCK,... requests are all ordered w.r.t. each other. Signed-off-by: Trond Myklebust --- fs/nfs/nfs4_fs.h | 42 ++++++++++++-- fs/nfs/nfs4proc.c | 111 +++++++++++++++++++++++------------- fs/nfs/nfs4state.c | 145 +++++++++++++++++++++++++++++++++++++++--------- fs/nfs/nfs4xdr.c | 43 +++++++++++--- include/linux/nfs_xdr.h | 15 ++--- 5 files changed, 273 insertions(+), 83 deletions(-) (limited to 'include/linux') diff --git a/fs/nfs/nfs4_fs.h b/fs/nfs/nfs4_fs.h index ec1a22d7b876..6ac6708484f5 100644 --- a/fs/nfs/nfs4_fs.h +++ b/fs/nfs/nfs4_fs.h @@ -92,6 +92,35 @@ struct nfs4_client { unsigned char cl_id_uniquifier; }; +/* + * struct rpc_sequence ensures that RPC calls are sent in the exact + * order that they appear on the list. + */ +struct rpc_sequence { + struct rpc_wait_queue wait; /* RPC call delay queue */ + spinlock_t lock; /* Protects the list */ + struct list_head list; /* Defines sequence of RPC calls */ +}; + +#define NFS_SEQID_CONFIRMED 1 +struct nfs_seqid_counter { + struct rpc_sequence *sequence; + int flags; + u32 counter; +}; + +struct nfs_seqid { + struct list_head list; + struct nfs_seqid_counter *sequence; + struct rpc_task *task; +}; + +static inline void nfs_confirm_seqid(struct nfs_seqid_counter *seqid, int status) +{ + if (seqid_mutating_err(-status)) + seqid->flags |= NFS_SEQID_CONFIRMED; +} + /* * NFS4 state_owners and lock_owners are simply labels for ordered * sequences of RPC calls. Their sole purpose is to provide once-only @@ -106,12 +135,13 @@ struct nfs4_state_owner { struct nfs4_client *so_client; u32 so_id; /* 32-bit identifier, unique */ struct semaphore so_sema; - u32 so_seqid; /* protected by so_sema */ atomic_t so_count; struct rpc_cred *so_cred; /* Associated cred */ struct list_head so_states; struct list_head so_delegations; + struct nfs_seqid_counter so_seqid; + struct rpc_sequence so_sequence; }; /* @@ -132,7 +162,7 @@ struct nfs4_lock_state { fl_owner_t ls_owner; /* POSIX lock owner */ #define NFS_LOCK_INITIALIZED 1 int ls_flags; - u32 ls_seqid; + struct nfs_seqid_counter ls_seqid; u32 ls_id; nfs4_stateid ls_stateid; atomic_t ls_count; @@ -224,12 +254,16 @@ extern struct nfs4_state * nfs4_get_open_state(struct inode *, struct nfs4_state extern void nfs4_put_open_state(struct nfs4_state *); extern void nfs4_close_state(struct nfs4_state *, mode_t); extern struct nfs4_state *nfs4_find_state(struct inode *, struct rpc_cred *, mode_t mode); -extern void nfs4_increment_seqid(int status, struct nfs4_state_owner *sp); extern void nfs4_schedule_state_recovery(struct nfs4_client *); extern int nfs4_set_lock_state(struct nfs4_state *state, struct file_lock *fl); -extern void nfs4_increment_lock_seqid(int status, struct nfs4_lock_state *ls); extern void nfs4_copy_stateid(nfs4_stateid *, struct nfs4_state *, fl_owner_t); +extern struct nfs_seqid *nfs_alloc_seqid(struct nfs_seqid_counter *counter); +extern int nfs_wait_on_sequence(struct nfs_seqid *seqid, struct rpc_task *task); +extern void nfs_increment_open_seqid(int status, struct nfs_seqid *seqid); +extern void nfs_increment_lock_seqid(int status, struct nfs_seqid *seqid); +extern void nfs_free_seqid(struct nfs_seqid *seqid); + extern const nfs4_stateid zero_stateid; /* nfs4xdr.c */ diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 9701ca8c9428..9ba89e7cdd28 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -218,7 +218,6 @@ static int _nfs4_open_reclaim(struct nfs4_state_owner *sp, struct nfs4_state *st struct nfs_delegation *delegation = NFS_I(inode)->delegation; struct nfs_openargs o_arg = { .fh = NFS_FH(inode), - .seqid = sp->so_seqid, .id = sp->so_id, .open_flags = state->state, .clientid = server->nfs4_state->cl_clientid, @@ -245,8 +244,13 @@ static int _nfs4_open_reclaim(struct nfs4_state_owner *sp, struct nfs4_state *st } o_arg.u.delegation_type = delegation->type; } + o_arg.seqid = nfs_alloc_seqid(&sp->so_seqid); + if (o_arg.seqid == NULL) + return -ENOMEM; status = rpc_call_sync(server->client, &msg, RPC_TASK_NOINTR); - nfs4_increment_seqid(status, sp); + /* Confirm the sequence as being established */ + nfs_confirm_seqid(&sp->so_seqid, status); + nfs_increment_open_seqid(status, o_arg.seqid); if (status == 0) { memcpy(&state->stateid, &o_res.stateid, sizeof(state->stateid)); if (o_res.delegation_type != 0) { @@ -256,6 +260,7 @@ static int _nfs4_open_reclaim(struct nfs4_state_owner *sp, struct nfs4_state *st nfs_async_inode_return_delegation(inode, &o_res.stateid); } } + nfs_free_seqid(o_arg.seqid); clear_bit(NFS_DELEGATED_STATE, &state->flags); /* Ensure we update the inode attributes */ NFS_CACHEINV(inode); @@ -307,16 +312,20 @@ static int _nfs4_open_delegation_recall(struct dentry *dentry, struct nfs4_state goto out; if (state->state == 0) goto out; - arg.seqid = sp->so_seqid; + arg.seqid = nfs_alloc_seqid(&sp->so_seqid); + status = -ENOMEM; + if (arg.seqid == NULL) + goto out; arg.open_flags = state->state; memcpy(arg.u.delegation.data, state->stateid.data, sizeof(arg.u.delegation.data)); status = rpc_call_sync(server->client, &msg, RPC_TASK_NOINTR); - nfs4_increment_seqid(status, sp); + nfs_increment_open_seqid(status, arg.seqid); if (status >= 0) { memcpy(state->stateid.data, res.stateid.data, sizeof(state->stateid.data)); clear_bit(NFS_DELEGATED_STATE, &state->flags); } + nfs_free_seqid(arg.seqid); out: up(&sp->so_sema); dput(parent); @@ -345,11 +354,11 @@ int nfs4_open_delegation_recall(struct dentry *dentry, struct nfs4_state *state) return err; } -static inline int _nfs4_proc_open_confirm(struct rpc_clnt *clnt, const struct nfs_fh *fh, struct nfs4_state_owner *sp, nfs4_stateid *stateid) +static inline int _nfs4_proc_open_confirm(struct rpc_clnt *clnt, const struct nfs_fh *fh, struct nfs4_state_owner *sp, nfs4_stateid *stateid, struct nfs_seqid *seqid) { struct nfs_open_confirmargs arg = { .fh = fh, - .seqid = sp->so_seqid, + .seqid = seqid, .stateid = *stateid, }; struct nfs_open_confirmres res; @@ -362,7 +371,9 @@ static inline int _nfs4_proc_open_confirm(struct rpc_clnt *clnt, const struct nf int status; status = rpc_call_sync(clnt, &msg, RPC_TASK_NOINTR); - nfs4_increment_seqid(status, sp); + /* Confirm the sequence as being established */ + nfs_confirm_seqid(&sp->so_seqid, status); + nfs_increment_open_seqid(status, seqid); if (status >= 0) memcpy(stateid, &res.stateid, sizeof(*stateid)); return status; @@ -380,21 +391,21 @@ static int _nfs4_proc_open(struct inode *dir, struct nfs4_state_owner *sp, stru int status; /* Update sequence id. The caller must serialize! */ - o_arg->seqid = sp->so_seqid; o_arg->id = sp->so_id; o_arg->clientid = sp->so_client->cl_clientid; status = rpc_call_sync(server->client, &msg, RPC_TASK_NOINTR); - nfs4_increment_seqid(status, sp); + nfs_increment_open_seqid(status, o_arg->seqid); if (status != 0) goto out; update_changeattr(dir, &o_res->cinfo); if(o_res->rflags & NFS4_OPEN_RESULT_CONFIRM) { status = _nfs4_proc_open_confirm(server->client, &o_res->fh, - sp, &o_res->stateid); + sp, &o_res->stateid, o_arg->seqid); if (status != 0) goto out; } + nfs_confirm_seqid(&sp->so_seqid, 0); if (!(o_res->f_attr->valid & NFS_ATTR_FATTR)) status = server->rpc_ops->getattr(server, &o_res->fh, o_res->f_attr); out: @@ -465,6 +476,10 @@ static int _nfs4_open_expired(struct nfs4_state_owner *sp, struct nfs4_state *st set_bit(NFS_DELEGATED_STATE, &state->flags); goto out; } + o_arg.seqid = nfs_alloc_seqid(&sp->so_seqid); + status = -ENOMEM; + if (o_arg.seqid == NULL) + goto out; status = _nfs4_proc_open(dir, sp, &o_arg, &o_res); if (status != 0) goto out_nodeleg; @@ -490,6 +505,7 @@ static int _nfs4_open_expired(struct nfs4_state_owner *sp, struct nfs4_state *st nfs_inode_reclaim_delegation(inode, sp->so_cred, &o_res); } out_nodeleg: + nfs_free_seqid(o_arg.seqid); clear_bit(NFS_DELEGATED_STATE, &state->flags); out: dput(parent); @@ -667,6 +683,9 @@ static int _nfs4_do_open(struct inode *dir, struct dentry *dentry, int flags, st /* Serialization for the sequence id */ down(&sp->so_sema); + o_arg.seqid = nfs_alloc_seqid(&sp->so_seqid); + if (o_arg.seqid == NULL) + return -ENOMEM; status = _nfs4_proc_open(dir, sp, &o_arg, &o_res); if (status != 0) goto out_err; @@ -681,6 +700,7 @@ static int _nfs4_do_open(struct inode *dir, struct dentry *dentry, int flags, st update_open_stateid(state, &o_res.stateid, flags); if (o_res.delegation_type != 0) nfs_inode_set_delegation(inode, cred, &o_res); + nfs_free_seqid(o_arg.seqid); up(&sp->so_sema); nfs4_put_state_owner(sp); up_read(&clp->cl_sem); @@ -690,6 +710,7 @@ out_err: if (sp != NULL) { if (state != NULL) nfs4_put_open_state(state); + nfs_free_seqid(o_arg.seqid); up(&sp->so_sema); nfs4_put_state_owner(sp); } @@ -718,7 +739,7 @@ static struct nfs4_state *nfs4_do_open(struct inode *dir, struct dentry *dentry, * It is actually a sign of a bug on the client or on the server. * * If we receive a BAD_SEQID error in the particular case of - * doing an OPEN, we assume that nfs4_increment_seqid() will + * doing an OPEN, we assume that nfs_increment_open_seqid() will * have unhashed the old state_owner for us, and that we can * therefore safely retry using a new one. We should still warn * the user though... @@ -799,7 +820,7 @@ static void nfs4_close_done(struct rpc_task *task) /* hmm. we are done with the inode, and in the process of freeing * the state_owner. we keep this around to process errors */ - nfs4_increment_seqid(task->tk_status, sp); + nfs_increment_open_seqid(task->tk_status, calldata->arg.seqid); switch (task->tk_status) { case 0: memcpy(&state->stateid, &calldata->res.stateid, @@ -818,6 +839,7 @@ static void nfs4_close_done(struct rpc_task *task) } state->state = calldata->arg.open_flags; nfs4_put_open_state(state); + nfs_free_seqid(calldata->arg.seqid); up(&sp->so_sema); nfs4_put_state_owner(sp); up_read(&server->nfs4_state->cl_sem); @@ -865,7 +887,11 @@ int nfs4_do_close(struct inode *inode, struct nfs4_state *state, mode_t mode) calldata->state = state; calldata->arg.fh = NFS_FH(inode); /* Serialization for the sequence id */ - calldata->arg.seqid = state->owner->so_seqid; + calldata->arg.seqid = nfs_alloc_seqid(&state->owner->so_seqid); + if (calldata->arg.seqid == NULL) { + kfree(calldata); + return -ENOMEM; + } calldata->arg.open_flags = mode; memcpy(&calldata->arg.stateid, &state->stateid, sizeof(calldata->arg.stateid)); @@ -2729,15 +2755,19 @@ static int _nfs4_proc_unlck(struct nfs4_state *state, int cmd, struct file_lock /* We might have lost the locks! */ if ((lsp->ls_flags & NFS_LOCK_INITIALIZED) == 0) goto out; - luargs.seqid = lsp->ls_seqid; - memcpy(&luargs.stateid, &lsp->ls_stateid, sizeof(luargs.stateid)); + luargs.seqid = nfs_alloc_seqid(&lsp->ls_seqid); + status = -ENOMEM; + if (luargs.seqid == NULL) + goto out; + memcpy(luargs.stateid.data, lsp->ls_stateid.data, sizeof(luargs.stateid.data)); arg.u.locku = &luargs; status = rpc_call_sync(server->client, &msg, RPC_TASK_NOINTR); - nfs4_increment_lock_seqid(status, lsp); + nfs_increment_lock_seqid(status, luargs.seqid); if (status == 0) - memcpy(&lsp->ls_stateid, &res.u.stateid, - sizeof(lsp->ls_stateid)); + memcpy(lsp->ls_stateid.data, res.u.stateid.data, + sizeof(lsp->ls_stateid.data)); + nfs_free_seqid(luargs.seqid); out: up(&state->lock_sema); if (status == 0) @@ -2783,9 +2813,13 @@ static int _nfs4_do_setlk(struct nfs4_state *state, int cmd, struct file_lock *r .reclaim = reclaim, .new_lock_owner = 0, }; - int status; + struct nfs_seqid *lock_seqid; + int status = -ENOMEM; - if (!(lsp->ls_flags & NFS_LOCK_INITIALIZED)) { + lock_seqid = nfs_alloc_seqid(&lsp->ls_seqid); + if (lock_seqid == NULL) + return -ENOMEM; + if (!(lsp->ls_seqid.flags & NFS_SEQID_CONFIRMED)) { struct nfs4_state_owner *owner = state->owner; struct nfs_open_to_lock otl = { .lock_owner = { @@ -2793,39 +2827,40 @@ static int _nfs4_do_setlk(struct nfs4_state *state, int cmd, struct file_lock *r }, }; - otl.lock_seqid = lsp->ls_seqid; + otl.lock_seqid = lock_seqid; otl.lock_owner.id = lsp->ls_id; memcpy(&otl.open_stateid, &state->stateid, sizeof(otl.open_stateid)); largs.u.open_lock = &otl; largs.new_lock_owner = 1; arg.u.lock = &largs; down(&owner->so_sema); - otl.open_seqid = owner->so_seqid; - status = rpc_call_sync(server->client, &msg, RPC_TASK_NOINTR); - /* increment open_owner seqid on success, and - * seqid mutating errors */ - nfs4_increment_seqid(status, owner); - up(&owner->so_sema); - if (status == 0) { - lsp->ls_flags |= NFS_LOCK_INITIALIZED; - lsp->ls_seqid++; + otl.open_seqid = nfs_alloc_seqid(&owner->so_seqid); + if (otl.open_seqid != NULL) { + status = rpc_call_sync(server->client, &msg, RPC_TASK_NOINTR); + /* increment seqid on success, and seqid mutating errors */ + nfs_increment_open_seqid(status, otl.open_seqid); + nfs_free_seqid(otl.open_seqid); } + up(&owner->so_sema); + if (status == 0) + nfs_confirm_seqid(&lsp->ls_seqid, 0); } else { - struct nfs_exist_lock el = { - .seqid = lsp->ls_seqid, - }; + struct nfs_exist_lock el; memcpy(&el.stateid, &lsp->ls_stateid, sizeof(el.stateid)); largs.u.exist_lock = ⪙ arg.u.lock = &largs; + el.seqid = lock_seqid; status = rpc_call_sync(server->client, &msg, RPC_TASK_NOINTR); - /* increment seqid on success, and * seqid mutating errors*/ - nfs4_increment_lock_seqid(status, lsp); } + /* increment seqid on success, and seqid mutating errors*/ + nfs_increment_lock_seqid(status, lock_seqid); /* save the returned stateid. */ - if (status == 0) - memcpy(&lsp->ls_stateid, &res.u.stateid, sizeof(nfs4_stateid)); - else if (status == -NFS4ERR_DENIED) + if (status == 0) { + memcpy(lsp->ls_stateid.data, res.u.stateid.data, sizeof(lsp->ls_stateid.data)); + lsp->ls_flags |= NFS_LOCK_INITIALIZED; + } else if (status == -NFS4ERR_DENIED) status = -EAGAIN; + nfs_free_seqid(lock_seqid); return status; } diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c index afe587d82f1e..f535c219cf3a 100644 --- a/fs/nfs/nfs4state.c +++ b/fs/nfs/nfs4state.c @@ -264,13 +264,16 @@ nfs4_alloc_state_owner(void) { struct nfs4_state_owner *sp; - sp = kmalloc(sizeof(*sp),GFP_KERNEL); + sp = kzalloc(sizeof(*sp),GFP_KERNEL); if (!sp) return NULL; init_MUTEX(&sp->so_sema); - sp->so_seqid = 0; /* arbitrary */ INIT_LIST_HEAD(&sp->so_states); INIT_LIST_HEAD(&sp->so_delegations); + rpc_init_wait_queue(&sp->so_sequence.wait, "Seqid_waitqueue"); + sp->so_seqid.sequence = &sp->so_sequence; + spin_lock_init(&sp->so_sequence.lock); + INIT_LIST_HEAD(&sp->so_sequence.list); atomic_set(&sp->so_count, 1); return sp; } @@ -553,12 +556,10 @@ static struct nfs4_lock_state *nfs4_alloc_lock_state(struct nfs4_state *state, f struct nfs4_lock_state *lsp; struct nfs4_client *clp = state->owner->so_client; - lsp = kmalloc(sizeof(*lsp), GFP_KERNEL); + lsp = kzalloc(sizeof(*lsp), GFP_KERNEL); if (lsp == NULL) return NULL; - lsp->ls_flags = 0; - lsp->ls_seqid = 0; /* arbitrary */ - memset(lsp->ls_stateid.data, 0, sizeof(lsp->ls_stateid.data)); + lsp->ls_seqid.sequence = &state->owner->so_sequence; atomic_set(&lsp->ls_count, 1); lsp->ls_owner = fl_owner; spin_lock(&clp->cl_lock); @@ -673,29 +674,102 @@ void nfs4_copy_stateid(nfs4_stateid *dst, struct nfs4_state *state, fl_owner_t f nfs4_put_lock_state(lsp); } -/* -* Called with state->lock_sema and clp->cl_sem held. -*/ -void nfs4_increment_lock_seqid(int status, struct nfs4_lock_state *lsp) +struct nfs_seqid *nfs_alloc_seqid(struct nfs_seqid_counter *counter) +{ + struct rpc_sequence *sequence = counter->sequence; + struct nfs_seqid *new; + + new = kmalloc(sizeof(*new), GFP_KERNEL); + if (new != NULL) { + new->sequence = counter; + new->task = NULL; + spin_lock(&sequence->lock); + list_add_tail(&new->list, &sequence->list); + spin_unlock(&sequence->lock); + } + return new; +} + +void nfs_free_seqid(struct nfs_seqid *seqid) { - if (status == NFS_OK || seqid_mutating_err(-status)) - lsp->ls_seqid++; + struct rpc_sequence *sequence = seqid->sequence->sequence; + struct rpc_task *next = NULL; + + spin_lock(&sequence->lock); + list_del(&seqid->list); + if (!list_empty(&sequence->list)) { + next = list_entry(sequence->list.next, struct nfs_seqid, list)->task; + if (next) + rpc_wake_up_task(next); + } + spin_unlock(&sequence->lock); + kfree(seqid); } /* -* Called with sp->so_sema and clp->cl_sem held. -* -* Increment the seqid if the OPEN/OPEN_DOWNGRADE/CLOSE succeeded, or -* failed with a seqid incrementing error - -* see comments nfs_fs.h:seqid_mutating_error() -*/ -void nfs4_increment_seqid(int status, struct nfs4_state_owner *sp) -{ - if (status == NFS_OK || seqid_mutating_err(-status)) - sp->so_seqid++; - /* If the server returns BAD_SEQID, unhash state_owner here */ - if (status == -NFS4ERR_BAD_SEQID) + * Called with sp->so_sema and clp->cl_sem held. + * + * Increment the seqid if the OPEN/OPEN_DOWNGRADE/CLOSE succeeded, or + * failed with a seqid incrementing error - + * see comments nfs_fs.h:seqid_mutating_error() + */ +static inline void nfs_increment_seqid(int status, struct nfs_seqid *seqid) +{ + switch (status) { + case 0: + break; + case -NFS4ERR_BAD_SEQID: + case -NFS4ERR_STALE_CLIENTID: + case -NFS4ERR_STALE_STATEID: + case -NFS4ERR_BAD_STATEID: + case -NFS4ERR_BADXDR: + case -NFS4ERR_RESOURCE: + case -NFS4ERR_NOFILEHANDLE: + /* Non-seqid mutating errors */ + return; + }; + /* + * Note: no locking needed as we are guaranteed to be first + * on the sequence list + */ + seqid->sequence->counter++; +} + +void nfs_increment_open_seqid(int status, struct nfs_seqid *seqid) +{ + if (status == -NFS4ERR_BAD_SEQID) { + struct nfs4_state_owner *sp = container_of(seqid->sequence, + struct nfs4_state_owner, so_seqid); nfs4_drop_state_owner(sp); + } + return nfs_increment_seqid(status, seqid); +} + +/* + * Called with ls->lock_sema and clp->cl_sem held. + * + * Increment the seqid if the LOCK/LOCKU succeeded, or + * failed with a seqid incrementing error - + * see comments nfs_fs.h:seqid_mutating_error() + */ +void nfs_increment_lock_seqid(int status, struct nfs_seqid *seqid) +{ + return nfs_increment_seqid(status, seqid); +} + +int nfs_wait_on_sequence(struct nfs_seqid *seqid, struct rpc_task *task) +{ + struct rpc_sequence *sequence = seqid->sequence->sequence; + int status = 0; + + spin_lock(&sequence->lock); + if (sequence->list.next != &seqid->list) { + seqid->task = task; + rpc_sleep_on(&sequence->wait, task, NULL, NULL); + status = -EAGAIN; + } + spin_unlock(&sequence->lock); + return status; } static int reclaimer(void *); @@ -791,8 +865,6 @@ static int nfs4_reclaim_open_state(struct nfs4_state_recovery_ops *ops, struct n if (state->state == 0) continue; status = ops->recover_open(sp, state); - list_for_each_entry(lock, &state->lock_states, ls_locks) - lock->ls_flags &= ~NFS_LOCK_INITIALIZED; if (status >= 0) { status = nfs4_reclaim_locks(ops, state); if (status < 0) @@ -831,6 +903,26 @@ out_err: return status; } +static void nfs4_state_mark_reclaim(struct nfs4_client *clp) +{ + struct nfs4_state_owner *sp; + struct nfs4_state *state; + struct nfs4_lock_state *lock; + + /* Reset all sequence ids to zero */ + list_for_each_entry(sp, &clp->cl_state_owners, so_list) { + sp->so_seqid.counter = 0; + sp->so_seqid.flags = 0; + list_for_each_entry(state, &sp->so_states, open_states) { + list_for_each_entry(lock, &state->lock_states, ls_locks) { + lock->ls_seqid.counter = 0; + lock->ls_seqid.flags = 0; + lock->ls_flags &= ~NFS_LOCK_INITIALIZED; + } + } + } +} + static int reclaimer(void *ptr) { struct reclaimer_args *args = (struct reclaimer_args *)ptr; @@ -864,6 +956,7 @@ restart_loop: default: ops = &nfs4_network_partition_recovery_ops; }; + nfs4_state_mark_reclaim(clp); status = __nfs4_init_client(clp); if (status) goto out_error; diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c index 6c564ef9489e..fcd28a29a2f8 100644 --- a/fs/nfs/nfs4xdr.c +++ b/fs/nfs/nfs4xdr.c @@ -604,7 +604,7 @@ static int encode_close(struct xdr_stream *xdr, const struct nfs_closeargs *arg) RESERVE_SPACE(8+sizeof(arg->stateid.data)); WRITE32(OP_CLOSE); - WRITE32(arg->seqid); + WRITE32(arg->seqid->sequence->counter); WRITEMEM(arg->stateid.data, sizeof(arg->stateid.data)); return 0; @@ -732,9 +732,9 @@ static int encode_lock(struct xdr_stream *xdr, const struct nfs_lockargs *arg) struct nfs_open_to_lock *ol = opargs->u.open_lock; RESERVE_SPACE(40); - WRITE32(ol->open_seqid); + WRITE32(ol->open_seqid->sequence->counter); WRITEMEM(&ol->open_stateid, sizeof(ol->open_stateid)); - WRITE32(ol->lock_seqid); + WRITE32(ol->lock_seqid->sequence->counter); WRITE64(ol->lock_owner.clientid); WRITE32(4); WRITE32(ol->lock_owner.id); @@ -744,7 +744,7 @@ static int encode_lock(struct xdr_stream *xdr, const struct nfs_lockargs *arg) RESERVE_SPACE(20); WRITEMEM(&el->stateid, sizeof(el->stateid)); - WRITE32(el->seqid); + WRITE32(el->seqid->sequence->counter); } return 0; @@ -775,7 +775,7 @@ static int encode_locku(struct xdr_stream *xdr, const struct nfs_lockargs *arg) RESERVE_SPACE(44); WRITE32(OP_LOCKU); WRITE32(arg->type); - WRITE32(opargs->seqid); + WRITE32(opargs->seqid->sequence->counter); WRITEMEM(&opargs->stateid, sizeof(opargs->stateid)); WRITE64(arg->offset); WRITE64(arg->length); @@ -826,7 +826,7 @@ static inline void encode_openhdr(struct xdr_stream *xdr, const struct nfs_opena */ RESERVE_SPACE(8); WRITE32(OP_OPEN); - WRITE32(arg->seqid); + WRITE32(arg->seqid->sequence->counter); encode_share_access(xdr, arg->open_flags); RESERVE_SPACE(16); WRITE64(arg->clientid); @@ -941,7 +941,7 @@ static int encode_open_confirm(struct xdr_stream *xdr, const struct nfs_open_con RESERVE_SPACE(8+sizeof(arg->stateid.data)); WRITE32(OP_OPEN_CONFIRM); WRITEMEM(arg->stateid.data, sizeof(arg->stateid.data)); - WRITE32(arg->seqid); + WRITE32(arg->seqid->sequence->counter); return 0; } @@ -953,7 +953,7 @@ static int encode_open_downgrade(struct xdr_stream *xdr, const struct nfs_closea RESERVE_SPACE(8+sizeof(arg->stateid.data)); WRITE32(OP_OPEN_DOWNGRADE); WRITEMEM(arg->stateid.data, sizeof(arg->stateid.data)); - WRITE32(arg->seqid); + WRITE32(arg->seqid->sequence->counter); encode_share_access(xdr, arg->open_flags); return 0; } @@ -1416,6 +1416,9 @@ static int nfs4_xdr_enc_close(struct rpc_rqst *req, uint32_t *p, struct nfs_clos }; int status; + status = nfs_wait_on_sequence(args->seqid, req->rq_task); + if (status != 0) + goto out; xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, &hdr); status = encode_putfh(&xdr, args->fh); @@ -1437,6 +1440,9 @@ static int nfs4_xdr_enc_open(struct rpc_rqst *req, uint32_t *p, struct nfs_opena }; int status; + status = nfs_wait_on_sequence(args->seqid, req->rq_task); + if (status != 0) + goto out; xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, &hdr); status = encode_putfh(&xdr, args->fh); @@ -1464,6 +1470,9 @@ static int nfs4_xdr_enc_open_confirm(struct rpc_rqst *req, uint32_t *p, struct n }; int status; + status = nfs_wait_on_sequence(args->seqid, req->rq_task); + if (status != 0) + goto out; xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, &hdr); status = encode_putfh(&xdr, args->fh); @@ -1485,6 +1494,9 @@ static int nfs4_xdr_enc_open_noattr(struct rpc_rqst *req, uint32_t *p, struct nf }; int status; + status = nfs_wait_on_sequence(args->seqid, req->rq_task); + if (status != 0) + goto out; xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, &hdr); status = encode_putfh(&xdr, args->fh); @@ -1506,6 +1518,9 @@ static int nfs4_xdr_enc_open_downgrade(struct rpc_rqst *req, uint32_t *p, struct }; int status; + status = nfs_wait_on_sequence(args->seqid, req->rq_task); + if (status != 0) + goto out; xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, &hdr); status = encode_putfh(&xdr, args->fh); @@ -1525,8 +1540,17 @@ static int nfs4_xdr_enc_lock(struct rpc_rqst *req, uint32_t *p, struct nfs_locka struct compound_hdr hdr = { .nops = 2, }; + struct nfs_lock_opargs *opargs = args->u.lock; + struct nfs_seqid *seqid; int status; + if (opargs->new_lock_owner) + seqid = opargs->u.open_lock->lock_seqid; + else + seqid = opargs->u.exist_lock->seqid; + status = nfs_wait_on_sequence(seqid, req->rq_task); + if (status != 0) + goto out; xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, &hdr); status = encode_putfh(&xdr, args->fh); @@ -1569,6 +1593,9 @@ static int nfs4_xdr_enc_locku(struct rpc_rqst *req, uint32_t *p, struct nfs_lock }; int status; + status = nfs_wait_on_sequence(args->u.locku->seqid, req->rq_task); + if (status != 0) + goto out; xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, &hdr); status = encode_putfh(&xdr, args->fh); diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h index a2bf6914ff1b..d578912bf9a9 100644 --- a/include/linux/nfs_xdr.h +++ b/include/linux/nfs_xdr.h @@ -96,12 +96,13 @@ struct nfs4_change_info { u64 after; }; +struct nfs_seqid; /* * Arguments to the open call. */ struct nfs_openargs { const struct nfs_fh * fh; - __u32 seqid; + struct nfs_seqid * seqid; int open_flags; __u64 clientid; __u32 id; @@ -136,7 +137,7 @@ struct nfs_openres { struct nfs_open_confirmargs { const struct nfs_fh * fh; nfs4_stateid stateid; - __u32 seqid; + struct nfs_seqid * seqid; }; struct nfs_open_confirmres { @@ -149,7 +150,7 @@ struct nfs_open_confirmres { struct nfs_closeargs { struct nfs_fh * fh; nfs4_stateid stateid; - __u32 seqid; + struct nfs_seqid * seqid; int open_flags; }; @@ -165,15 +166,15 @@ struct nfs_lowner { }; struct nfs_open_to_lock { - __u32 open_seqid; + struct nfs_seqid * open_seqid; nfs4_stateid open_stateid; - __u32 lock_seqid; + struct nfs_seqid * lock_seqid; struct nfs_lowner lock_owner; }; struct nfs_exist_lock { nfs4_stateid stateid; - __u32 seqid; + struct nfs_seqid * seqid; }; struct nfs_lock_opargs { @@ -186,7 +187,7 @@ struct nfs_lock_opargs { }; struct nfs_locku_opargs { - __u32 seqid; + struct nfs_seqid * seqid; nfs4_stateid stateid; }; -- cgit v1.2.3 From 9512135df14f8293b9bc5e8fb22d4279dee5ff66 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Tue, 18 Oct 2005 14:20:12 -0700 Subject: NFSv4: Fix a potential CLOSE race Once the state_owner and lock_owner semaphores get removed, it will be possible for other OPEN requests to reopen the same file if they have lower sequence ids than our CLOSE call. This patch ensures that we recheck the file state once nfs_wait_on_sequence() has completed waiting. Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 110 +++++++++++++++++++++++++++++++++--------------- fs/nfs/nfs4state.c | 6 ++- fs/nfs/nfs4xdr.c | 14 ++---- include/linux/nfs_xdr.h | 2 +- 4 files changed, 87 insertions(+), 45 deletions(-) (limited to 'include/linux') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 9ba89e7cdd28..5154ddf6d9a5 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -189,6 +189,21 @@ static void update_changeattr(struct inode *inode, struct nfs4_change_info *cinf nfsi->change_attr = cinfo->after; } +/* Helper for asynchronous RPC calls */ +static int nfs4_call_async(struct rpc_clnt *clnt, rpc_action tk_begin, + rpc_action tk_exit, void *calldata) +{ + struct rpc_task *task; + + if (!(task = rpc_new_task(clnt, tk_exit, RPC_TASK_ASYNC))) + return -ENOMEM; + + task->tk_calldata = calldata; + task->tk_action = tk_begin; + rpc_execute(task); + return 0; +} + static void update_open_stateid(struct nfs4_state *state, nfs4_stateid *stateid, int open_flags) { struct inode *inode = state->inode; @@ -810,11 +825,24 @@ struct nfs4_closedata { struct nfs_closeres res; }; +static void nfs4_free_closedata(struct nfs4_closedata *calldata) +{ + struct nfs4_state *state = calldata->state; + struct nfs4_state_owner *sp = state->owner; + struct nfs_server *server = NFS_SERVER(calldata->inode); + + nfs4_put_open_state(calldata->state); + nfs_free_seqid(calldata->arg.seqid); + up(&sp->so_sema); + nfs4_put_state_owner(sp); + up_read(&server->nfs4_state->cl_sem); + kfree(calldata); +} + static void nfs4_close_done(struct rpc_task *task) { struct nfs4_closedata *calldata = (struct nfs4_closedata *)task->tk_calldata; struct nfs4_state *state = calldata->state; - struct nfs4_state_owner *sp = state->owner; struct nfs_server *server = NFS_SERVER(calldata->inode); /* hmm. we are done with the inode, and in the process of freeing @@ -838,25 +866,46 @@ static void nfs4_close_done(struct rpc_task *task) } } state->state = calldata->arg.open_flags; - nfs4_put_open_state(state); - nfs_free_seqid(calldata->arg.seqid); - up(&sp->so_sema); - nfs4_put_state_owner(sp); - up_read(&server->nfs4_state->cl_sem); - kfree(calldata); + nfs4_free_closedata(calldata); } -static inline int nfs4_close_call(struct rpc_clnt *clnt, struct nfs4_closedata *calldata) +static void nfs4_close_begin(struct rpc_task *task) { + struct nfs4_closedata *calldata = (struct nfs4_closedata *)task->tk_calldata; + struct nfs4_state *state = calldata->state; struct rpc_message msg = { .rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_CLOSE], .rpc_argp = &calldata->arg, .rpc_resp = &calldata->res, - .rpc_cred = calldata->state->owner->so_cred, + .rpc_cred = state->owner->so_cred, }; - if (calldata->arg.open_flags != 0) + int mode = 0; + int status; + + status = nfs_wait_on_sequence(calldata->arg.seqid, task); + if (status != 0) + return; + /* Don't reorder reads */ + smp_rmb(); + /* Recalculate the new open mode in case someone reopened the file + * while we were waiting in line to be scheduled. + */ + if (state->nreaders != 0) + mode |= FMODE_READ; + if (state->nwriters != 0) + mode |= FMODE_WRITE; + if (test_bit(NFS_DELEGATED_STATE, &state->flags)) + state->state = mode; + if (mode == state->state) { + nfs4_free_closedata(calldata); + task->tk_exit = NULL; + rpc_exit(task, 0); + return; + } + if (mode != 0) msg.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_OPEN_DOWNGRADE]; - return rpc_call_async(clnt, &msg, 0, nfs4_close_done, calldata); + calldata->arg.open_flags = mode; + rpc_call_setup(task, &msg, 0); } /* @@ -873,35 +922,30 @@ static inline int nfs4_close_call(struct rpc_clnt *clnt, struct nfs4_closedata * int nfs4_do_close(struct inode *inode, struct nfs4_state *state, mode_t mode) { struct nfs4_closedata *calldata; - int status; + int status = -ENOMEM; - /* Tell caller we're done */ - if (test_bit(NFS_DELEGATED_STATE, &state->flags)) { - state->state = mode; - return 0; - } - calldata = (struct nfs4_closedata *)kmalloc(sizeof(*calldata), GFP_KERNEL); + calldata = kmalloc(sizeof(*calldata), GFP_KERNEL); if (calldata == NULL) - return -ENOMEM; + goto out; calldata->inode = inode; calldata->state = state; calldata->arg.fh = NFS_FH(inode); + calldata->arg.stateid = &state->stateid; /* Serialization for the sequence id */ calldata->arg.seqid = nfs_alloc_seqid(&state->owner->so_seqid); - if (calldata->arg.seqid == NULL) { - kfree(calldata); - return -ENOMEM; - } - calldata->arg.open_flags = mode; - memcpy(&calldata->arg.stateid, &state->stateid, - sizeof(calldata->arg.stateid)); - status = nfs4_close_call(NFS_SERVER(inode)->client, calldata); - /* - * Return -EINPROGRESS on success in order to indicate to the - * caller that an asynchronous RPC call has been launched, and - * that it will release the semaphores on completion. - */ - return (status == 0) ? -EINPROGRESS : status; + if (calldata->arg.seqid == NULL) + goto out_free_calldata; + + status = nfs4_call_async(NFS_SERVER(inode)->client, nfs4_close_begin, + nfs4_close_done, calldata); + if (status == 0) + goto out; + + nfs_free_seqid(calldata->arg.seqid); +out_free_calldata: + kfree(calldata); +out: + return status; } struct inode * diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c index f535c219cf3a..59c93f37e1b2 100644 --- a/fs/nfs/nfs4state.c +++ b/fs/nfs/nfs4state.c @@ -518,7 +518,11 @@ void nfs4_close_state(struct nfs4_state *state, mode_t mode) newstate |= FMODE_WRITE; if (state->state == newstate) goto out; - if (nfs4_do_close(inode, state, newstate) == -EINPROGRESS) + if (test_bit(NFS_DELEGATED_STATE, &state->flags)) { + state->state = newstate; + goto out; + } + if (nfs4_do_close(inode, state, newstate) == 0) return; } out: diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c index fcd28a29a2f8..934ec50ea6bf 100644 --- a/fs/nfs/nfs4xdr.c +++ b/fs/nfs/nfs4xdr.c @@ -602,10 +602,10 @@ static int encode_close(struct xdr_stream *xdr, const struct nfs_closeargs *arg) { uint32_t *p; - RESERVE_SPACE(8+sizeof(arg->stateid.data)); + RESERVE_SPACE(8+sizeof(arg->stateid->data)); WRITE32(OP_CLOSE); WRITE32(arg->seqid->sequence->counter); - WRITEMEM(arg->stateid.data, sizeof(arg->stateid.data)); + WRITEMEM(arg->stateid->data, sizeof(arg->stateid->data)); return 0; } @@ -950,9 +950,9 @@ static int encode_open_downgrade(struct xdr_stream *xdr, const struct nfs_closea { uint32_t *p; - RESERVE_SPACE(8+sizeof(arg->stateid.data)); + RESERVE_SPACE(8+sizeof(arg->stateid->data)); WRITE32(OP_OPEN_DOWNGRADE); - WRITEMEM(arg->stateid.data, sizeof(arg->stateid.data)); + WRITEMEM(arg->stateid->data, sizeof(arg->stateid->data)); WRITE32(arg->seqid->sequence->counter); encode_share_access(xdr, arg->open_flags); return 0; @@ -1416,9 +1416,6 @@ static int nfs4_xdr_enc_close(struct rpc_rqst *req, uint32_t *p, struct nfs_clos }; int status; - status = nfs_wait_on_sequence(args->seqid, req->rq_task); - if (status != 0) - goto out; xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, &hdr); status = encode_putfh(&xdr, args->fh); @@ -1518,9 +1515,6 @@ static int nfs4_xdr_enc_open_downgrade(struct rpc_rqst *req, uint32_t *p, struct }; int status; - status = nfs_wait_on_sequence(args->seqid, req->rq_task); - if (status != 0) - goto out; xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, &hdr); status = encode_putfh(&xdr, args->fh); diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h index d578912bf9a9..cac0df950c66 100644 --- a/include/linux/nfs_xdr.h +++ b/include/linux/nfs_xdr.h @@ -149,7 +149,7 @@ struct nfs_open_confirmres { */ struct nfs_closeargs { struct nfs_fh * fh; - nfs4_stateid stateid; + nfs4_stateid * stateid; struct nfs_seqid * seqid; int open_flags; }; -- cgit v1.2.3 From faf5f49c2d9c0af2847837c232a432cc146e203b Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Tue, 18 Oct 2005 14:20:15 -0700 Subject: NFSv4: Make NFS clean up byte range locks asynchronously Currently we fail to do so if the process was signalled. Signed-off-by: Trond Myklebust --- fs/nfs/nfs4_fs.h | 1 + fs/nfs/nfs4proc.c | 161 +++++++++++++++++++++++++++++++++--------------- fs/nfs/nfs4state.c | 2 +- fs/nfs/nfs4xdr.c | 9 +-- include/linux/nfs_xdr.h | 2 +- 5 files changed, 116 insertions(+), 59 deletions(-) (limited to 'include/linux') diff --git a/fs/nfs/nfs4_fs.h b/fs/nfs/nfs4_fs.h index d4fcb5d0ce6c..2215cdee43ae 100644 --- a/fs/nfs/nfs4_fs.h +++ b/fs/nfs/nfs4_fs.h @@ -249,6 +249,7 @@ extern void nfs4_put_open_state(struct nfs4_state *); extern void nfs4_close_state(struct nfs4_state *, mode_t); extern struct nfs4_state *nfs4_find_state(struct inode *, struct rpc_cred *, mode_t mode); extern void nfs4_schedule_state_recovery(struct nfs4_client *); +extern void nfs4_put_lock_state(struct nfs4_lock_state *lsp); extern int nfs4_set_lock_state(struct nfs4_state *state, struct file_lock *fl); extern void nfs4_copy_stateid(nfs4_stateid *, struct nfs4_state *, fl_owner_t); diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index f57dba815099..612a9a14aed3 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -58,9 +58,9 @@ static int _nfs4_proc_open_confirm(struct rpc_clnt *clnt, const struct nfs_fh *fh, struct nfs4_state_owner *sp, nfs4_stateid *stateid, struct nfs_seqid *seqid); static int nfs4_do_fsinfo(struct nfs_server *, struct nfs_fh *, struct nfs_fsinfo *); -static int nfs4_async_handle_error(struct rpc_task *, struct nfs_server *); +static int nfs4_async_handle_error(struct rpc_task *, const struct nfs_server *); static int _nfs4_proc_access(struct inode *inode, struct nfs_access_entry *entry); -static int nfs4_handle_exception(struct nfs_server *server, int errorcode, struct nfs4_exception *exception); +static int nfs4_handle_exception(const struct nfs_server *server, int errorcode, struct nfs4_exception *exception); extern u32 *nfs4_decode_dirent(u32 *p, struct nfs_entry *entry, int plus); extern struct rpc_procinfo nfs4_procedures[]; @@ -2422,7 +2422,7 @@ static int nfs4_proc_set_acl(struct inode *inode, const void *buf, size_t buflen } static int -nfs4_async_handle_error(struct rpc_task *task, struct nfs_server *server) +nfs4_async_handle_error(struct rpc_task *task, const struct nfs_server *server) { struct nfs4_client *clp = server->nfs4_state; @@ -2500,7 +2500,7 @@ static int nfs4_delay(struct rpc_clnt *clnt, long *timeout) /* This is the error handling routine for processes that are allowed * to sleep. */ -int nfs4_handle_exception(struct nfs_server *server, int errorcode, struct nfs4_exception *exception) +int nfs4_handle_exception(const struct nfs_server *server, int errorcode, struct nfs4_exception *exception) { struct nfs4_client *clp = server->nfs4_state; int ret = errorcode; @@ -2763,68 +2763,127 @@ static int do_vfs_lock(struct file *file, struct file_lock *fl) return res; } -static int _nfs4_proc_unlck(struct nfs4_state *state, int cmd, struct file_lock *request) +struct nfs4_unlockdata { + struct nfs_lockargs arg; + struct nfs_locku_opargs luargs; + struct nfs_lockres res; + struct nfs4_lock_state *lsp; + struct nfs_open_context *ctx; + atomic_t refcount; + struct completion completion; +}; + +static void nfs4_locku_release_calldata(struct nfs4_unlockdata *calldata) { - struct inode *inode = state->inode; - struct nfs_server *server = NFS_SERVER(inode); - struct nfs_lockargs arg = { - .fh = NFS_FH(inode), - .type = nfs4_lck_type(cmd, request), - .offset = request->fl_start, - .length = nfs4_lck_length(request), - }; - struct nfs_lockres res = { - .server = server, - }; + if (atomic_dec_and_test(&calldata->refcount)) { + nfs_free_seqid(calldata->luargs.seqid); + nfs4_put_lock_state(calldata->lsp); + put_nfs_open_context(calldata->ctx); + kfree(calldata); + } +} + +static void nfs4_locku_complete(struct nfs4_unlockdata *calldata) +{ + complete(&calldata->completion); + nfs4_locku_release_calldata(calldata); +} + +static void nfs4_locku_done(struct rpc_task *task) +{ + struct nfs4_unlockdata *calldata = (struct nfs4_unlockdata *)task->tk_calldata; + + nfs_increment_lock_seqid(task->tk_status, calldata->luargs.seqid); + switch (task->tk_status) { + case 0: + memcpy(calldata->lsp->ls_stateid.data, + calldata->res.u.stateid.data, + sizeof(calldata->lsp->ls_stateid.data)); + break; + case -NFS4ERR_STALE_STATEID: + case -NFS4ERR_EXPIRED: + nfs4_schedule_state_recovery(calldata->res.server->nfs4_state); + break; + default: + if (nfs4_async_handle_error(task, calldata->res.server) == -EAGAIN) { + rpc_restart_call(task); + return; + } + } + nfs4_locku_complete(calldata); +} + +static void nfs4_locku_begin(struct rpc_task *task) +{ + struct nfs4_unlockdata *calldata = (struct nfs4_unlockdata *)task->tk_calldata; struct rpc_message msg = { .rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_LOCKU], - .rpc_argp = &arg, - .rpc_resp = &res, - .rpc_cred = state->owner->so_cred, + .rpc_argp = &calldata->arg, + .rpc_resp = &calldata->res, + .rpc_cred = calldata->lsp->ls_state->owner->so_cred, }; + int status; + + status = nfs_wait_on_sequence(calldata->luargs.seqid, task); + if (status != 0) + return; + if ((calldata->lsp->ls_flags & NFS_LOCK_INITIALIZED) == 0) { + nfs4_locku_complete(calldata); + task->tk_exit = NULL; + rpc_exit(task, 0); + return; + } + rpc_call_setup(task, &msg, 0); +} + +static int nfs4_proc_unlck(struct nfs4_state *state, int cmd, struct file_lock *request) +{ + struct nfs4_unlockdata *calldata; + struct inode *inode = state->inode; + struct nfs_server *server = NFS_SERVER(inode); struct nfs4_lock_state *lsp; - struct nfs_locku_opargs luargs; int status; - + status = nfs4_set_lock_state(state, request); if (status != 0) - goto out; + return status; lsp = request->fl_u.nfs4_fl.owner; /* We might have lost the locks! */ if ((lsp->ls_flags & NFS_LOCK_INITIALIZED) == 0) - goto out; - luargs.seqid = nfs_alloc_seqid(&lsp->ls_seqid); - status = -ENOMEM; - if (luargs.seqid == NULL) - goto out; - memcpy(luargs.stateid.data, lsp->ls_stateid.data, sizeof(luargs.stateid.data)); - arg.u.locku = &luargs; - status = rpc_call_sync(server->client, &msg, RPC_TASK_NOINTR); - nfs_increment_lock_seqid(status, luargs.seqid); - - if (status == 0) - memcpy(lsp->ls_stateid.data, res.u.stateid.data, - sizeof(lsp->ls_stateid.data)); - nfs_free_seqid(luargs.seqid); -out: + return 0; + calldata = kmalloc(sizeof(*calldata), GFP_KERNEL); + if (calldata == NULL) + return -ENOMEM; + calldata->luargs.seqid = nfs_alloc_seqid(&lsp->ls_seqid); + if (calldata->luargs.seqid == NULL) { + kfree(calldata); + return -ENOMEM; + } + calldata->luargs.stateid = &lsp->ls_stateid; + calldata->arg.fh = NFS_FH(inode); + calldata->arg.type = nfs4_lck_type(cmd, request); + calldata->arg.offset = request->fl_start; + calldata->arg.length = nfs4_lck_length(request); + calldata->arg.u.locku = &calldata->luargs; + calldata->res.server = server; + calldata->lsp = lsp; + atomic_inc(&lsp->ls_count); + + /* Ensure we don't close file until we're done freeing locks! */ + calldata->ctx = get_nfs_open_context((struct nfs_open_context*)request->fl_file->private_data); + + atomic_set(&calldata->refcount, 2); + init_completion(&calldata->completion); + + status = nfs4_call_async(NFS_SERVER(inode)->client, nfs4_locku_begin, + nfs4_locku_done, calldata); if (status == 0) - do_vfs_lock(request->fl_file, request); + wait_for_completion_interruptible(&calldata->completion); + do_vfs_lock(request->fl_file, request); + nfs4_locku_release_calldata(calldata); return status; } -static int nfs4_proc_unlck(struct nfs4_state *state, int cmd, struct file_lock *request) -{ - struct nfs4_exception exception = { }; - int err; - - do { - err = nfs4_handle_exception(NFS_SERVER(state->inode), - _nfs4_proc_unlck(state, cmd, request), - &exception); - } while (exception.retry); - return err; -} - static int _nfs4_do_setlk(struct nfs4_state *state, int cmd, struct file_lock *request, int reclaim) { struct inode *inode = state->inode; diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c index bb3574361958..23834c8fb740 100644 --- a/fs/nfs/nfs4state.c +++ b/fs/nfs/nfs4state.c @@ -600,7 +600,7 @@ static struct nfs4_lock_state *nfs4_get_lock_state(struct nfs4_state *state, fl_ * Release reference to lock_state, and free it if we see that * it is no longer in use */ -static void nfs4_put_lock_state(struct nfs4_lock_state *lsp) +void nfs4_put_lock_state(struct nfs4_lock_state *lsp) { struct nfs4_state *state; diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c index 934ec50ea6bf..4706192cfb07 100644 --- a/fs/nfs/nfs4xdr.c +++ b/fs/nfs/nfs4xdr.c @@ -776,7 +776,7 @@ static int encode_locku(struct xdr_stream *xdr, const struct nfs_lockargs *arg) WRITE32(OP_LOCKU); WRITE32(arg->type); WRITE32(opargs->seqid->sequence->counter); - WRITEMEM(&opargs->stateid, sizeof(opargs->stateid)); + WRITEMEM(opargs->stateid->data, sizeof(opargs->stateid->data)); WRITE64(arg->offset); WRITE64(arg->length); @@ -1587,9 +1587,6 @@ static int nfs4_xdr_enc_locku(struct rpc_rqst *req, uint32_t *p, struct nfs_lock }; int status; - status = nfs_wait_on_sequence(args->u.locku->seqid, req->rq_task); - if (status != 0) - goto out; xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, &hdr); status = encode_putfh(&xdr, args->fh); @@ -2934,8 +2931,8 @@ static int decode_locku(struct xdr_stream *xdr, struct nfs_lockres *res) status = decode_op_hdr(xdr, OP_LOCKU); if (status == 0) { - READ_BUF(sizeof(nfs4_stateid)); - COPYMEM(&res->u.stateid, sizeof(res->u.stateid)); + READ_BUF(sizeof(res->u.stateid.data)); + COPYMEM(res->u.stateid.data, sizeof(res->u.stateid.data)); } return status; } diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h index cac0df950c66..849f95c5fae4 100644 --- a/include/linux/nfs_xdr.h +++ b/include/linux/nfs_xdr.h @@ -188,7 +188,7 @@ struct nfs_lock_opargs { struct nfs_locku_opargs { struct nfs_seqid * seqid; - nfs4_stateid stateid; + nfs4_stateid * stateid; }; struct nfs_lockargs { -- cgit v1.2.3 From 06735b3454824bd561decbde46111f144e905923 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Tue, 18 Oct 2005 14:20:15 -0700 Subject: NFSv4: Fix up handling of open_to_lock sequence ids Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 69 +++++++++++++++++++++++-------------------------- fs/nfs/nfs4xdr.c | 32 ++++++++++------------- include/linux/nfs_xdr.h | 19 +++----------- 3 files changed, 49 insertions(+), 71 deletions(-) (limited to 'include/linux') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 612a9a14aed3..35da15342e05 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -2889,11 +2889,23 @@ static int _nfs4_do_setlk(struct nfs4_state *state, int cmd, struct file_lock *r struct inode *inode = state->inode; struct nfs_server *server = NFS_SERVER(inode); struct nfs4_lock_state *lsp = request->fl_u.nfs4_fl.owner; + struct nfs_lock_opargs largs = { + .lock_stateid = &lsp->ls_stateid, + .open_stateid = &state->stateid, + .lock_owner = { + .clientid = server->nfs4_state->cl_clientid, + .id = lsp->ls_id, + }, + .reclaim = reclaim, + }; struct nfs_lockargs arg = { .fh = NFS_FH(inode), .type = nfs4_lck_type(cmd, request), .offset = request->fl_start, .length = nfs4_lck_length(request), + .u = { + .lock = &largs, + }, }; struct nfs_lockres res = { .server = server, @@ -2904,56 +2916,39 @@ static int _nfs4_do_setlk(struct nfs4_state *state, int cmd, struct file_lock *r .rpc_resp = &res, .rpc_cred = state->owner->so_cred, }; - struct nfs_lock_opargs largs = { - .reclaim = reclaim, - .new_lock_owner = 0, - }; - struct nfs_seqid *lock_seqid; int status = -ENOMEM; - lock_seqid = nfs_alloc_seqid(&lsp->ls_seqid); - if (lock_seqid == NULL) + largs.lock_seqid = nfs_alloc_seqid(&lsp->ls_seqid); + if (largs.lock_seqid == NULL) return -ENOMEM; if (!(lsp->ls_seqid.flags & NFS_SEQID_CONFIRMED)) { struct nfs4_state_owner *owner = state->owner; - struct nfs_open_to_lock otl = { - .lock_owner = { - .clientid = server->nfs4_state->cl_clientid, - }, - }; - - otl.lock_seqid = lock_seqid; - otl.lock_owner.id = lsp->ls_id; - memcpy(&otl.open_stateid, &state->stateid, sizeof(otl.open_stateid)); - largs.u.open_lock = &otl; + + largs.open_seqid = nfs_alloc_seqid(&owner->so_seqid); + if (largs.open_seqid == NULL) + goto out; largs.new_lock_owner = 1; - arg.u.lock = &largs; - otl.open_seqid = nfs_alloc_seqid(&owner->so_seqid); - if (otl.open_seqid != NULL) { - status = rpc_call_sync(server->client, &msg, RPC_TASK_NOINTR); - /* increment seqid on success, and seqid mutating errors */ - nfs_increment_open_seqid(status, otl.open_seqid); - nfs_free_seqid(otl.open_seqid); + status = rpc_call_sync(server->client, &msg, RPC_TASK_NOINTR); + /* increment open seqid on success, and seqid mutating errors */ + if (largs.new_lock_owner != 0) { + nfs_increment_open_seqid(status, largs.open_seqid); + if (status == 0) + nfs_confirm_seqid(&lsp->ls_seqid, 0); } - if (status == 0) - nfs_confirm_seqid(&lsp->ls_seqid, 0); - } else { - struct nfs_exist_lock el; - memcpy(&el.stateid, &lsp->ls_stateid, sizeof(el.stateid)); - largs.u.exist_lock = ⪙ - arg.u.lock = &largs; - el.seqid = lock_seqid; + nfs_free_seqid(largs.open_seqid); + } else status = rpc_call_sync(server->client, &msg, RPC_TASK_NOINTR); - } - /* increment seqid on success, and seqid mutating errors*/ - nfs_increment_lock_seqid(status, lock_seqid); + /* increment lock seqid on success, and seqid mutating errors*/ + nfs_increment_lock_seqid(status, largs.lock_seqid); /* save the returned stateid. */ if (status == 0) { - memcpy(lsp->ls_stateid.data, res.u.stateid.data, sizeof(lsp->ls_stateid.data)); + memcpy(lsp->ls_stateid.data, res.u.stateid.data, + sizeof(lsp->ls_stateid.data)); lsp->ls_flags |= NFS_LOCK_INITIALIZED; } else if (status == -NFS4ERR_DENIED) status = -EAGAIN; - nfs_free_seqid(lock_seqid); +out: + nfs_free_seqid(largs.lock_seqid); return status; } diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c index 4706192cfb07..c5c75235c5b8 100644 --- a/fs/nfs/nfs4xdr.c +++ b/fs/nfs/nfs4xdr.c @@ -729,22 +729,18 @@ static int encode_lock(struct xdr_stream *xdr, const struct nfs_lockargs *arg) WRITE64(arg->length); WRITE32(opargs->new_lock_owner); if (opargs->new_lock_owner){ - struct nfs_open_to_lock *ol = opargs->u.open_lock; - RESERVE_SPACE(40); - WRITE32(ol->open_seqid->sequence->counter); - WRITEMEM(&ol->open_stateid, sizeof(ol->open_stateid)); - WRITE32(ol->lock_seqid->sequence->counter); - WRITE64(ol->lock_owner.clientid); + WRITE32(opargs->open_seqid->sequence->counter); + WRITEMEM(opargs->open_stateid->data, sizeof(opargs->open_stateid->data)); + WRITE32(opargs->lock_seqid->sequence->counter); + WRITE64(opargs->lock_owner.clientid); WRITE32(4); - WRITE32(ol->lock_owner.id); + WRITE32(opargs->lock_owner.id); } else { - struct nfs_exist_lock *el = opargs->u.exist_lock; - RESERVE_SPACE(20); - WRITEMEM(&el->stateid, sizeof(el->stateid)); - WRITE32(el->seqid->sequence->counter); + WRITEMEM(opargs->lock_stateid->data, sizeof(opargs->lock_stateid->data)); + WRITE32(opargs->lock_seqid->sequence->counter); } return 0; @@ -1535,16 +1531,14 @@ static int nfs4_xdr_enc_lock(struct rpc_rqst *req, uint32_t *p, struct nfs_locka .nops = 2, }; struct nfs_lock_opargs *opargs = args->u.lock; - struct nfs_seqid *seqid; int status; - if (opargs->new_lock_owner) - seqid = opargs->u.open_lock->lock_seqid; - else - seqid = opargs->u.exist_lock->seqid; - status = nfs_wait_on_sequence(seqid, req->rq_task); + status = nfs_wait_on_sequence(opargs->lock_seqid, req->rq_task); if (status != 0) goto out; + /* Do we need to do an open_to_lock_owner? */ + if (opargs->lock_seqid->sequence->flags & NFS_SEQID_CONFIRMED) + opargs->new_lock_owner = 0; xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, &hdr); status = encode_putfh(&xdr, args->fh); @@ -2908,8 +2902,8 @@ static int decode_lock(struct xdr_stream *xdr, struct nfs_lockres *res) status = decode_op_hdr(xdr, OP_LOCK); if (status == 0) { - READ_BUF(sizeof(nfs4_stateid)); - COPYMEM(&res->u.stateid, sizeof(res->u.stateid)); + READ_BUF(sizeof(res->u.stateid.data)); + COPYMEM(res->u.stateid.data, sizeof(res->u.stateid.data)); } else if (status == -NFS4ERR_DENIED) return decode_lock_denied(xdr, &res->u.denied); return status; diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h index 849f95c5fae4..57efcc27f20b 100644 --- a/include/linux/nfs_xdr.h +++ b/include/linux/nfs_xdr.h @@ -165,25 +165,14 @@ struct nfs_lowner { u32 id; }; -struct nfs_open_to_lock { - struct nfs_seqid * open_seqid; - nfs4_stateid open_stateid; +struct nfs_lock_opargs { struct nfs_seqid * lock_seqid; + nfs4_stateid * lock_stateid; + struct nfs_seqid * open_seqid; + nfs4_stateid * open_stateid; struct nfs_lowner lock_owner; -}; - -struct nfs_exist_lock { - nfs4_stateid stateid; - struct nfs_seqid * seqid; -}; - -struct nfs_lock_opargs { __u32 reclaim; __u32 new_lock_owner; - union { - struct nfs_open_to_lock *open_lock; - struct nfs_exist_lock *exist_lock; - } u; }; struct nfs_locku_opargs { -- cgit v1.2.3 From 834f2a4a1554dc5b2598038b3fe8703defcbe467 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Tue, 18 Oct 2005 14:20:16 -0700 Subject: VFS: Allow the filesystem to return a full file pointer on open intent This is needed by NFSv4 for atomicity reasons: our open command is in fact a lookup+open, so we need to be able to propagate open context information from lookup() into the resulting struct file's private_data field. Signed-off-by: Trond Myklebust --- fs/exec.c | 12 +++---- fs/namei.c | 93 +++++++++++++++++++++++++++++++++++++++++++++++---- fs/open.c | 79 +++++++++++++++++++++++++++++++++++-------- include/linux/namei.h | 8 +++++ 4 files changed, 165 insertions(+), 27 deletions(-) (limited to 'include/linux') diff --git a/fs/exec.c b/fs/exec.c index a04a575ad433..d2208f7c87db 100644 --- a/fs/exec.c +++ b/fs/exec.c @@ -126,8 +126,7 @@ asmlinkage long sys_uselib(const char __user * library) struct nameidata nd; int error; - nd.intent.open.flags = FMODE_READ; - error = __user_walk(library, LOOKUP_FOLLOW|LOOKUP_OPEN, &nd); + error = __user_path_lookup_open(library, LOOKUP_FOLLOW, &nd, FMODE_READ); if (error) goto out; @@ -139,7 +138,7 @@ asmlinkage long sys_uselib(const char __user * library) if (error) goto exit; - file = dentry_open(nd.dentry, nd.mnt, O_RDONLY); + file = nameidata_to_filp(&nd, O_RDONLY); error = PTR_ERR(file); if (IS_ERR(file)) goto out; @@ -167,6 +166,7 @@ asmlinkage long sys_uselib(const char __user * library) out: return error; exit: + release_open_intent(&nd); path_release(&nd); goto out; } @@ -490,8 +490,7 @@ struct file *open_exec(const char *name) int err; struct file *file; - nd.intent.open.flags = FMODE_READ; - err = path_lookup(name, LOOKUP_FOLLOW|LOOKUP_OPEN, &nd); + err = path_lookup_open(name, LOOKUP_FOLLOW, &nd, FMODE_READ); file = ERR_PTR(err); if (!err) { @@ -504,7 +503,7 @@ struct file *open_exec(const char *name) err = -EACCES; file = ERR_PTR(err); if (!err) { - file = dentry_open(nd.dentry, nd.mnt, O_RDONLY); + file = nameidata_to_filp(&nd, O_RDONLY); if (!IS_ERR(file)) { err = deny_write_access(file); if (err) { @@ -516,6 +515,7 @@ out: return file; } } + release_open_intent(&nd); path_release(&nd); } goto out; diff --git a/fs/namei.c b/fs/namei.c index aa62dbda93ac..0d1dff7d3d95 100644 --- a/fs/namei.c +++ b/fs/namei.c @@ -28,6 +28,7 @@ #include #include #include +#include #include #include @@ -317,6 +318,18 @@ void path_release_on_umount(struct nameidata *nd) mntput_no_expire(nd->mnt); } +/** + * release_open_intent - free up open intent resources + * @nd: pointer to nameidata + */ +void release_open_intent(struct nameidata *nd) +{ + if (nd->intent.open.file->f_dentry == NULL) + put_filp(nd->intent.open.file); + else + fput(nd->intent.open.file); +} + /* * Internal lookup() using the new generic dcache. * SMP-safe @@ -1052,6 +1065,70 @@ out: return retval; } +static int __path_lookup_intent_open(const char *name, unsigned int lookup_flags, + struct nameidata *nd, int open_flags, int create_mode) +{ + struct file *filp = get_empty_filp(); + int err; + + if (filp == NULL) + return -ENFILE; + nd->intent.open.file = filp; + nd->intent.open.flags = open_flags; + nd->intent.open.create_mode = create_mode; + err = path_lookup(name, lookup_flags|LOOKUP_OPEN, nd); + if (IS_ERR(nd->intent.open.file)) { + if (err == 0) { + err = PTR_ERR(nd->intent.open.file); + path_release(nd); + } + } else if (err != 0) + release_open_intent(nd); + return err; +} + +/** + * path_lookup_open - lookup a file path with open intent + * @name: pointer to file name + * @lookup_flags: lookup intent flags + * @nd: pointer to nameidata + * @open_flags: open intent flags + */ +int path_lookup_open(const char *name, unsigned int lookup_flags, + struct nameidata *nd, int open_flags) +{ + return __path_lookup_intent_open(name, lookup_flags, nd, + open_flags, 0); +} + +/** + * path_lookup_create - lookup a file path with open + create intent + * @name: pointer to file name + * @lookup_flags: lookup intent flags + * @nd: pointer to nameidata + * @open_flags: open intent flags + * @create_mode: create intent flags + */ +int path_lookup_create(const char *name, unsigned int lookup_flags, + struct nameidata *nd, int open_flags, int create_mode) +{ + return __path_lookup_intent_open(name, lookup_flags|LOOKUP_CREATE, nd, + open_flags, create_mode); +} + +int __user_path_lookup_open(const char __user *name, unsigned int lookup_flags, + struct nameidata *nd, int open_flags) +{ + char *tmp = getname(name); + int err = PTR_ERR(tmp); + + if (!IS_ERR(tmp)) { + err = __path_lookup_intent_open(tmp, lookup_flags, nd, open_flags, 0); + putname(tmp); + } + return err; +} + /* * Restricted form of lookup. Doesn't follow links, single-component only, * needs parent already locked. Doesn't follow mounts. @@ -1416,27 +1493,27 @@ int may_open(struct nameidata *nd, int acc_mode, int flag) */ int open_namei(const char * pathname, int flag, int mode, struct nameidata *nd) { - int acc_mode, error = 0; + int acc_mode, error; struct path path; struct dentry *dir; int count = 0; acc_mode = ACC_MODE(flag); + /* O_TRUNC implies we need access checks for write permissions */ + if (flag & O_TRUNC) + acc_mode |= MAY_WRITE; + /* Allow the LSM permission hook to distinguish append access from general write access. */ if (flag & O_APPEND) acc_mode |= MAY_APPEND; - /* Fill in the open() intent data */ - nd->intent.open.flags = flag; - nd->intent.open.create_mode = mode; - /* * The simplest case - just a plain lookup. */ if (!(flag & O_CREAT)) { - error = path_lookup(pathname, lookup_flags(flag)|LOOKUP_OPEN, nd); + error = path_lookup_open(pathname, lookup_flags(flag), nd, flag); if (error) return error; goto ok; @@ -1445,7 +1522,7 @@ int open_namei(const char * pathname, int flag, int mode, struct nameidata *nd) /* * Create - we need to know the parent. */ - error = path_lookup(pathname, LOOKUP_PARENT|LOOKUP_OPEN|LOOKUP_CREATE, nd); + error = path_lookup_create(pathname, LOOKUP_PARENT, nd, flag, mode); if (error) return error; @@ -1520,6 +1597,8 @@ ok: exit_dput: dput_path(&path, nd); exit: + if (!IS_ERR(nd->intent.open.file)) + release_open_intent(nd); path_release(nd); return error; diff --git a/fs/open.c b/fs/open.c index f0d90cf0495c..8d06ec911fd9 100644 --- a/fs/open.c +++ b/fs/open.c @@ -739,7 +739,8 @@ asmlinkage long sys_fchown(unsigned int fd, uid_t user, gid_t group) } static struct file *__dentry_open(struct dentry *dentry, struct vfsmount *mnt, - int flags, struct file *f) + int flags, struct file *f, + int (*open)(struct inode *, struct file *)) { struct inode *inode; int error; @@ -761,11 +762,14 @@ static struct file *__dentry_open(struct dentry *dentry, struct vfsmount *mnt, f->f_op = fops_get(inode->i_fop); file_move(f, &inode->i_sb->s_files); - if (f->f_op && f->f_op->open) { - error = f->f_op->open(inode,f); + if (!open && f->f_op) + open = f->f_op->open; + if (open) { + error = open(inode, f); if (error) goto cleanup_all; } + f->f_flags &= ~(O_CREAT | O_EXCL | O_NOCTTY | O_TRUNC); file_ra_state_init(&f->f_ra, f->f_mapping->host->i_mapping); @@ -814,28 +818,75 @@ struct file *filp_open(const char * filename, int flags, int mode) { int namei_flags, error; struct nameidata nd; - struct file *f; namei_flags = flags; if ((namei_flags+1) & O_ACCMODE) namei_flags++; - if (namei_flags & O_TRUNC) - namei_flags |= 2; - - error = -ENFILE; - f = get_empty_filp(); - if (f == NULL) - return ERR_PTR(error); error = open_namei(filename, namei_flags, mode, &nd); if (!error) - return __dentry_open(nd.dentry, nd.mnt, flags, f); + return nameidata_to_filp(&nd, flags); - put_filp(f); return ERR_PTR(error); } EXPORT_SYMBOL(filp_open); +/** + * lookup_instantiate_filp - instantiates the open intent filp + * @nd: pointer to nameidata + * @dentry: pointer to dentry + * @open: open callback + * + * Helper for filesystems that want to use lookup open intents and pass back + * a fully instantiated struct file to the caller. + * This function is meant to be called from within a filesystem's + * lookup method. + * Note that in case of error, nd->intent.open.file is destroyed, but the + * path information remains valid. + * If the open callback is set to NULL, then the standard f_op->open() + * filesystem callback is substituted. + */ +struct file *lookup_instantiate_filp(struct nameidata *nd, struct dentry *dentry, + int (*open)(struct inode *, struct file *)) +{ + if (IS_ERR(nd->intent.open.file)) + goto out; + if (IS_ERR(dentry)) + goto out_err; + nd->intent.open.file = __dentry_open(dget(dentry), mntget(nd->mnt), + nd->intent.open.flags - 1, + nd->intent.open.file, + open); +out: + return nd->intent.open.file; +out_err: + release_open_intent(nd); + nd->intent.open.file = (struct file *)dentry; + goto out; +} +EXPORT_SYMBOL_GPL(lookup_instantiate_filp); + +/** + * nameidata_to_filp - convert a nameidata to an open filp. + * @nd: pointer to nameidata + * @flags: open flags + * + * Note that this function destroys the original nameidata + */ +struct file *nameidata_to_filp(struct nameidata *nd, int flags) +{ + struct file *filp; + + /* Pick up the filp from the open intent */ + filp = nd->intent.open.file; + /* Has the filesystem initialised the file for us? */ + if (filp->f_dentry == NULL) + filp = __dentry_open(nd->dentry, nd->mnt, flags, filp, NULL); + else + path_release(nd); + return filp; +} + struct file *dentry_open(struct dentry *dentry, struct vfsmount *mnt, int flags) { int error; @@ -846,7 +897,7 @@ struct file *dentry_open(struct dentry *dentry, struct vfsmount *mnt, int flags) if (f == NULL) return ERR_PTR(error); - return __dentry_open(dentry, mnt, flags, f); + return __dentry_open(dentry, mnt, flags, f, NULL); } EXPORT_SYMBOL(dentry_open); diff --git a/include/linux/namei.h b/include/linux/namei.h index 7db67b008cac..1c975d0d9e94 100644 --- a/include/linux/namei.h +++ b/include/linux/namei.h @@ -8,6 +8,7 @@ struct vfsmount; struct open_intent { int flags; int create_mode; + struct file *file; }; enum { MAX_NESTED_LINKS = 5 }; @@ -65,6 +66,13 @@ extern int FASTCALL(link_path_walk(const char *, struct nameidata *)); extern void path_release(struct nameidata *); extern void path_release_on_umount(struct nameidata *); +extern int __user_path_lookup_open(const char __user *, unsigned lookup_flags, struct nameidata *nd, int open_flags); +extern int path_lookup_open(const char *, unsigned lookup_flags, struct nameidata *, int open_flags); +extern struct file *lookup_instantiate_filp(struct nameidata *nd, struct dentry *dentry, + int (*open)(struct inode *, struct file *)); +extern struct file *nameidata_to_filp(struct nameidata *nd, int flags); +extern void release_open_intent(struct nameidata *); + extern struct dentry * lookup_one_len(const char *, struct dentry *, int); extern struct dentry * lookup_hash(struct qstr *, struct dentry *); -- cgit v1.2.3 From 02a913a73b52071e93f4b76db3e86138d19efffd Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Tue, 18 Oct 2005 14:20:17 -0700 Subject: NFSv4: Eliminate nfsv4 open race... Make NFSv4 return the fully initialized file pointer with the stateid that it created in the lookup w/intent. Signed-off-by: Trond Myklebust --- fs/nfs/dir.c | 29 +++++----- fs/nfs/nfs3proc.c | 2 +- fs/nfs/nfs4_fs.h | 4 +- fs/nfs/nfs4proc.c | 137 ++++++++++++++++++++---------------------------- fs/nfs/proc.c | 2 +- include/linux/nfs_xdr.h | 2 +- 6 files changed, 74 insertions(+), 102 deletions(-) (limited to 'include/linux') diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c index c70eabd6d179..a6e251f21fd8 100644 --- a/fs/nfs/dir.c +++ b/fs/nfs/dir.c @@ -914,7 +914,6 @@ static int is_atomic_open(struct inode *dir, struct nameidata *nd) static struct dentry *nfs_atomic_lookup(struct inode *dir, struct dentry *dentry, struct nameidata *nd) { struct dentry *res = NULL; - struct inode *inode = NULL; int error; /* Check that we are indeed trying to open this file */ @@ -928,8 +927,10 @@ static struct dentry *nfs_atomic_lookup(struct inode *dir, struct dentry *dentry dentry->d_op = NFS_PROTO(dir)->dentry_ops; /* Let vfs_create() deal with O_EXCL */ - if (nd->intent.open.flags & O_EXCL) - goto no_entry; + if (nd->intent.open.flags & O_EXCL) { + d_add(dentry, NULL); + goto out; + } /* Open the file on the server */ lock_kernel(); @@ -943,18 +944,18 @@ static struct dentry *nfs_atomic_lookup(struct inode *dir, struct dentry *dentry if (nd->intent.open.flags & O_CREAT) { nfs_begin_data_update(dir); - inode = nfs4_atomic_open(dir, dentry, nd); + res = nfs4_atomic_open(dir, dentry, nd); nfs_end_data_update(dir); } else - inode = nfs4_atomic_open(dir, dentry, nd); + res = nfs4_atomic_open(dir, dentry, nd); unlock_kernel(); - if (IS_ERR(inode)) { - error = PTR_ERR(inode); + if (IS_ERR(res)) { + error = PTR_ERR(res); switch (error) { /* Make a negative dentry */ case -ENOENT: - inode = NULL; - break; + res = NULL; + goto out; /* This turned out not to be a regular file */ case -ELOOP: if (!(nd->intent.open.flags & O_NOFOLLOW)) @@ -962,13 +963,9 @@ static struct dentry *nfs_atomic_lookup(struct inode *dir, struct dentry *dentry /* case -EISDIR: */ /* case -EINVAL: */ default: - res = ERR_PTR(error); goto out; } - } -no_entry: - res = d_add_unique(dentry, inode); - if (res != NULL) + } else if (res != NULL) dentry = res; nfs_renew_times(dentry); nfs_set_verifier(dentry, nfs_save_change_attribute(dir)); @@ -1012,7 +1009,7 @@ static int nfs_open_revalidate(struct dentry *dentry, struct nameidata *nd) */ lock_kernel(); verifier = nfs_save_change_attribute(dir); - ret = nfs4_open_revalidate(dir, dentry, openflags); + ret = nfs4_open_revalidate(dir, dentry, openflags, nd); if (!ret) nfs_set_verifier(dentry, verifier); unlock_kernel(); @@ -1135,7 +1132,7 @@ static int nfs_create(struct inode *dir, struct dentry *dentry, int mode, lock_kernel(); nfs_begin_data_update(dir); - error = NFS_PROTO(dir)->create(dir, dentry, &attr, open_flags); + error = NFS_PROTO(dir)->create(dir, dentry, &attr, open_flags, nd); nfs_end_data_update(dir); if (error != 0) goto out_err; diff --git a/fs/nfs/nfs3proc.c b/fs/nfs/nfs3proc.c index edc95514046d..df80477c5af4 100644 --- a/fs/nfs/nfs3proc.c +++ b/fs/nfs/nfs3proc.c @@ -299,7 +299,7 @@ static int nfs3_proc_commit(struct nfs_write_data *cdata) */ static int nfs3_proc_create(struct inode *dir, struct dentry *dentry, struct iattr *sattr, - int flags) + int flags, struct nameidata *nd) { struct nfs_fh fhandle; struct nfs_fattr fattr; diff --git a/fs/nfs/nfs4_fs.h b/fs/nfs/nfs4_fs.h index 2215cdee43ae..8a3788199052 100644 --- a/fs/nfs/nfs4_fs.h +++ b/fs/nfs/nfs4_fs.h @@ -215,8 +215,8 @@ extern int nfs4_proc_setclientid_confirm(struct nfs4_client *); extern int nfs4_proc_async_renew(struct nfs4_client *); extern int nfs4_proc_renew(struct nfs4_client *); extern int nfs4_do_close(struct inode *inode, struct nfs4_state *state, mode_t mode); -extern struct inode *nfs4_atomic_open(struct inode *, struct dentry *, struct nameidata *); -extern int nfs4_open_revalidate(struct inode *, struct dentry *, int); +extern struct dentry *nfs4_atomic_open(struct inode *, struct dentry *, struct nameidata *); +extern int nfs4_open_revalidate(struct inode *, struct dentry *, int, struct nameidata *); extern struct nfs4_state_recovery_ops nfs4_reboot_recovery_ops; extern struct nfs4_state_recovery_ops nfs4_network_partition_recovery_ops; diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 35da15342e05..c9ecb8119632 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -47,6 +47,7 @@ #include #include #include +#include #include "nfs4_fs.h" #include "delegation.h" @@ -947,12 +948,26 @@ out: return status; } -struct inode * +static void nfs4_intent_set_file(struct nameidata *nd, struct dentry *dentry, struct nfs4_state *state) +{ + struct file *filp; + + filp = lookup_instantiate_filp(nd, dentry, NULL); + if (!IS_ERR(filp)) { + struct nfs_open_context *ctx; + ctx = (struct nfs_open_context *)filp->private_data; + ctx->state = state; + } else + nfs4_close_state(state, nd->intent.open.flags); +} + +struct dentry * nfs4_atomic_open(struct inode *dir, struct dentry *dentry, struct nameidata *nd) { struct iattr attr; struct rpc_cred *cred; struct nfs4_state *state; + struct dentry *res; if (nd->flags & LOOKUP_CREATE) { attr.ia_mode = nd->intent.open.create_mode; @@ -966,16 +981,23 @@ nfs4_atomic_open(struct inode *dir, struct dentry *dentry, struct nameidata *nd) cred = rpcauth_lookupcred(NFS_SERVER(dir)->client->cl_auth, 0); if (IS_ERR(cred)) - return (struct inode *)cred; + return (struct dentry *)cred; state = nfs4_do_open(dir, dentry, nd->intent.open.flags, &attr, cred); put_rpccred(cred); - if (IS_ERR(state)) - return (struct inode *)state; - return state->inode; + if (IS_ERR(state)) { + if (PTR_ERR(state) == -ENOENT) + d_add(dentry, NULL); + return (struct dentry *)state; + } + res = d_add_unique(dentry, state->inode); + if (res != NULL) + dentry = res; + nfs4_intent_set_file(nd, dentry, state); + return res; } int -nfs4_open_revalidate(struct inode *dir, struct dentry *dentry, int openflags) +nfs4_open_revalidate(struct inode *dir, struct dentry *dentry, int openflags, struct nameidata *nd) { struct rpc_cred *cred; struct nfs4_state *state; @@ -988,18 +1010,30 @@ nfs4_open_revalidate(struct inode *dir, struct dentry *dentry, int openflags) if (IS_ERR(state)) state = nfs4_do_open(dir, dentry, openflags, NULL, cred); put_rpccred(cred); - if (state == ERR_PTR(-ENOENT) && dentry->d_inode == 0) - return 1; - if (IS_ERR(state)) - return 0; + if (IS_ERR(state)) { + switch (PTR_ERR(state)) { + case -EPERM: + case -EACCES: + case -EDQUOT: + case -ENOSPC: + case -EROFS: + lookup_instantiate_filp(nd, (struct dentry *)state, NULL); + return 1; + case -ENOENT: + if (dentry->d_inode == NULL) + return 1; + } + goto out_drop; + } inode = state->inode; + iput(inode); if (inode == dentry->d_inode) { - iput(inode); + nfs4_intent_set_file(nd, dentry, state); return 1; } - d_drop(dentry); nfs4_close_state(state, openflags); - iput(inode); +out_drop: + d_drop(dentry); return 0; } @@ -1500,7 +1534,7 @@ static int nfs4_proc_commit(struct nfs_write_data *cdata) static int nfs4_proc_create(struct inode *dir, struct dentry *dentry, struct iattr *sattr, - int flags) + int flags, struct nameidata *nd) { struct nfs4_state *state; struct rpc_cred *cred; @@ -1522,13 +1556,13 @@ nfs4_proc_create(struct inode *dir, struct dentry *dentry, struct iattr *sattr, struct nfs_fattr fattr; status = nfs4_do_setattr(NFS_SERVER(dir), &fattr, NFS_FH(state->inode), sattr, state); - if (status == 0) { + if (status == 0) nfs_setattr_update_inode(state->inode, sattr); - goto out; - } - } else if (flags != 0) - goto out; - nfs4_close_state(state, flags); + } + if (status == 0 && nd != NULL && (nd->flags & LOOKUP_OPEN)) + nfs4_intent_set_file(nd, dentry, state); + else + nfs4_close_state(state, flags); out: return status; } @@ -2175,65 +2209,6 @@ nfs4_proc_renew(struct nfs4_client *clp) return 0; } -/* - * We will need to arrange for the VFS layer to provide an atomic open. - * Until then, this open method is prone to inefficiency and race conditions - * due to the lookup, potential create, and open VFS calls from sys_open() - * placed on the wire. - */ -static int -nfs4_proc_file_open(struct inode *inode, struct file *filp) -{ - struct dentry *dentry = filp->f_dentry; - struct nfs_open_context *ctx; - struct nfs4_state *state = NULL; - struct rpc_cred *cred; - int status = -ENOMEM; - - dprintk("nfs4_proc_file_open: starting on (%.*s/%.*s)\n", - (int)dentry->d_parent->d_name.len, - dentry->d_parent->d_name.name, - (int)dentry->d_name.len, dentry->d_name.name); - - - /* Find our open stateid */ - cred = rpcauth_lookupcred(NFS_SERVER(inode)->client->cl_auth, 0); - if (IS_ERR(cred)) - return PTR_ERR(cred); - ctx = alloc_nfs_open_context(dentry, cred); - put_rpccred(cred); - if (unlikely(ctx == NULL)) - return -ENOMEM; - status = -EIO; /* ERACE actually */ - state = nfs4_find_state(inode, cred, filp->f_mode); - if (unlikely(state == NULL)) - goto no_state; - ctx->state = state; - nfs4_close_state(state, filp->f_mode); - ctx->mode = filp->f_mode; - nfs_file_set_open_context(filp, ctx); - put_nfs_open_context(ctx); - if (filp->f_mode & FMODE_WRITE) - nfs_begin_data_update(inode); - return 0; -no_state: - printk(KERN_WARNING "NFS: v4 raced in function %s\n", __FUNCTION__); - put_nfs_open_context(ctx); - return status; -} - -/* - * Release our state - */ -static int -nfs4_proc_file_release(struct inode *inode, struct file *filp) -{ - if (filp->f_mode & FMODE_WRITE) - nfs_end_data_update(inode); - nfs_file_clear_open_context(filp); - return 0; -} - static inline int nfs4_server_supports_acls(struct nfs_server *server) { return (server->caps & NFS_CAP_ACLS) @@ -3145,8 +3120,8 @@ struct nfs_rpc_ops nfs_v4_clientops = { .read_setup = nfs4_proc_read_setup, .write_setup = nfs4_proc_write_setup, .commit_setup = nfs4_proc_commit_setup, - .file_open = nfs4_proc_file_open, - .file_release = nfs4_proc_file_release, + .file_open = nfs_open, + .file_release = nfs_release, .lock = nfs4_proc_lock, .clear_acl_cache = nfs4_zap_acl_attr, }; diff --git a/fs/nfs/proc.c b/fs/nfs/proc.c index be23c3fb9260..8fef86523d7f 100644 --- a/fs/nfs/proc.c +++ b/fs/nfs/proc.c @@ -216,7 +216,7 @@ static int nfs_proc_write(struct nfs_write_data *wdata) static int nfs_proc_create(struct inode *dir, struct dentry *dentry, struct iattr *sattr, - int flags) + int flags, struct nameidata *nd) { struct nfs_fh fhandle; struct nfs_fattr fattr; diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h index 57efcc27f20b..60086dac11d5 100644 --- a/include/linux/nfs_xdr.h +++ b/include/linux/nfs_xdr.h @@ -712,7 +712,7 @@ struct nfs_rpc_ops { int (*write) (struct nfs_write_data *); int (*commit) (struct nfs_write_data *); int (*create) (struct inode *, struct dentry *, - struct iattr *, int); + struct iattr *, int, struct nameidata *); int (*remove) (struct inode *, struct qstr *); int (*unlink_setup) (struct rpc_message *, struct dentry *, struct qstr *); -- cgit v1.2.3 From 293f1eb551a77fe5c8956a559a3c0baea95cd9bc Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Thu, 13 Oct 2005 16:54:37 -0400 Subject: SUNRPC: Add support for privacy to generic gss-api code. Add support for privacy to generic gss-api code. This is dead code until we have both a mechanism that supports privacy and code in the client or server that uses it. Signed-off-by: J. Bruce Fields Signed-off-by: Trond Myklebust --- include/linux/sunrpc/gss_api.h | 22 ++++++++++++++++++++++ net/sunrpc/auth_gss/gss_mech_switch.c | 22 ++++++++++++++++++++++ 2 files changed, 44 insertions(+) (limited to 'include/linux') diff --git a/include/linux/sunrpc/gss_api.h b/include/linux/sunrpc/gss_api.h index 689262f63059..e896752ffbf9 100644 --- a/include/linux/sunrpc/gss_api.h +++ b/include/linux/sunrpc/gss_api.h @@ -48,6 +48,17 @@ u32 gss_verify_mic( struct xdr_buf *message, struct xdr_netobj *mic_token, u32 *qstate); +u32 gss_wrap( + struct gss_ctx *ctx_id, + u32 qop, + int offset, + struct xdr_buf *outbuf, + struct page **inpages); +u32 gss_unwrap( + struct gss_ctx *ctx_id, + u32 *qop, + int offset, + struct xdr_buf *inbuf); u32 gss_delete_sec_context( struct gss_ctx **ctx_id); @@ -93,6 +104,17 @@ struct gss_api_ops { struct xdr_buf *message, struct xdr_netobj *mic_token, u32 *qstate); + u32 (*gss_wrap)( + struct gss_ctx *ctx_id, + u32 qop, + int offset, + struct xdr_buf *outbuf, + struct page **inpages); + u32 (*gss_unwrap)( + struct gss_ctx *ctx_id, + u32 *qop, + int offset, + struct xdr_buf *buf); void (*gss_delete_sec_context)( void *internal_ctx_id); }; diff --git a/net/sunrpc/auth_gss/gss_mech_switch.c b/net/sunrpc/auth_gss/gss_mech_switch.c index 58aeaddd8c79..06d97cb3481a 100644 --- a/net/sunrpc/auth_gss/gss_mech_switch.c +++ b/net/sunrpc/auth_gss/gss_mech_switch.c @@ -276,6 +276,28 @@ gss_verify_mic(struct gss_ctx *context_handle, qstate); } +u32 +gss_wrap(struct gss_ctx *ctx_id, + u32 qop, + int offset, + struct xdr_buf *buf, + struct page **inpages) +{ + return ctx_id->mech_type->gm_ops + ->gss_wrap(ctx_id, qop, offset, buf, inpages); +} + +u32 +gss_unwrap(struct gss_ctx *ctx_id, + u32 *qop, + int offset, + struct xdr_buf *buf) +{ + return ctx_id->mech_type->gm_ops + ->gss_unwrap(ctx_id, qop, offset, buf); +} + + /* gss_delete_sec_context: free all resources associated with context_handle. * Note this differs from the RFC 2744-specified prototype in that we don't * bother returning an output token, since it would never be used anyway. */ -- cgit v1.2.3 From ead5e1c26fdcd969cf40c49cb0589d56879d240d Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Thu, 13 Oct 2005 16:54:43 -0400 Subject: SUNRPC: Provide a callback to allow free pages allocated during xdr encoding For privacy, we need to allocate pages to store the encrypted data (passed in pages can't be used without the risk of corrupting data in the page cache). So we need a way to free that memory after the request has been transmitted. Signed-off-by: J. Bruce Fields Signed-off-by: Trond Myklebust --- include/linux/sunrpc/xprt.h | 5 ++++- net/sunrpc/xprt.c | 3 +++ 2 files changed, 7 insertions(+), 1 deletion(-) (limited to 'include/linux') diff --git a/include/linux/sunrpc/xprt.h b/include/linux/sunrpc/xprt.h index 068e1fb0868b..3b8b6e823c70 100644 --- a/include/linux/sunrpc/xprt.h +++ b/include/linux/sunrpc/xprt.h @@ -73,7 +73,10 @@ struct rpc_rqst { int rq_cong; /* has incremented xprt->cong */ int rq_received; /* receive completed */ u32 rq_seqno; /* gss seq no. used on req. */ - + int rq_enc_pages_num; + struct page **rq_enc_pages; /* scratch pages for use by + gss privacy code */ + void (*rq_release_snd_buf)(struct rpc_rqst *); /* release rq_enc_pages */ struct list_head rq_list; struct xdr_buf rq_private_buf; /* The receive buffer diff --git a/net/sunrpc/xprt.c b/net/sunrpc/xprt.c index 1ba55dc38b7a..6dda3860351f 100644 --- a/net/sunrpc/xprt.c +++ b/net/sunrpc/xprt.c @@ -839,6 +839,7 @@ static void xprt_request_init(struct rpc_task *task, struct rpc_xprt *xprt) req->rq_task = task; req->rq_xprt = xprt; req->rq_xid = xprt_alloc_xid(xprt); + req->rq_release_snd_buf = NULL; dprintk("RPC: %4d reserved req %p xid %08x\n", task->tk_pid, req, ntohl(req->rq_xid)); } @@ -867,6 +868,8 @@ void xprt_release(struct rpc_task *task) xprt->last_used + xprt->idle_timeout); spin_unlock_bh(&xprt->transport_lock); task->tk_rqstp = NULL; + if (req->rq_release_snd_buf) + req->rq_release_snd_buf(req); memset(req, 0, sizeof(*req)); /* mark unused */ dprintk("RPC: %4d release request %p\n", task->tk_pid, req); -- cgit v1.2.3 From 24b2605becc10ca63c4c30808fa59a8abbf68727 Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Thu, 13 Oct 2005 16:54:53 -0400 Subject: RPCSEC_GSS: cleanup au_rslack calculation Various xdr encode routines use au_rslack to guess where the reply argument will end up, so we can set up the xdr_buf to recieve data into the right place for zero copy. Currently we calculate the au_rslack estimate when we check the verifier. Normally this only depends on the verifier size. In the integrity case we add a few bytes to allow for a length and sequence number. It's a bit simpler to calculate only the verifier size when we check the verifier, and delay the full calculation till we unwrap. Signed-off-by: J. Bruce Fields Signed-off-by: Trond Myklebust --- include/linux/sunrpc/auth.h | 7 ++++++- net/sunrpc/auth_gss/auth_gss.c | 20 ++++++-------------- 2 files changed, 12 insertions(+), 15 deletions(-) (limited to 'include/linux') diff --git a/include/linux/sunrpc/auth.h b/include/linux/sunrpc/auth.h index 04ebc24db348..b68c11a2d6dd 100644 --- a/include/linux/sunrpc/auth.h +++ b/include/linux/sunrpc/auth.h @@ -66,7 +66,12 @@ struct rpc_cred_cache { struct rpc_auth { unsigned int au_cslack; /* call cred size estimate */ - unsigned int au_rslack; /* reply verf size guess */ + /* guess at number of u32's auth adds before + * reply data; normally the verifier size: */ + unsigned int au_rslack; + /* for gss, used to calculate au_rslack: */ + unsigned int au_verfsize; + unsigned int au_flags; /* various flags */ struct rpc_authops * au_ops; /* operations */ rpc_authflavor_t au_flavor; /* pseudoflavor (note may diff --git a/net/sunrpc/auth_gss/auth_gss.c b/net/sunrpc/auth_gss/auth_gss.c index d2b08f16c257..dc95b797ca65 100644 --- a/net/sunrpc/auth_gss/auth_gss.c +++ b/net/sunrpc/auth_gss/auth_gss.c @@ -886,8 +886,6 @@ static u32 * gss_validate(struct rpc_task *task, u32 *p) { struct rpc_cred *cred = task->tk_msg.rpc_cred; - struct gss_cred *gss_cred = container_of(cred, struct gss_cred, - gc_base); struct gss_cl_ctx *ctx = gss_cred_get_ctx(cred); u32 seq, qop_state; struct kvec iov; @@ -915,18 +913,9 @@ gss_validate(struct rpc_task *task, u32 *p) cred->cr_flags &= ~RPCAUTH_CRED_UPTODATE; if (maj_stat) goto out_bad; - switch (gss_cred->gc_service) { - case RPC_GSS_SVC_NONE: - /* verifier data, flavor, length: */ - task->tk_auth->au_rslack = XDR_QUADLEN(len) + 2; - break; - case RPC_GSS_SVC_INTEGRITY: - /* verifier data, flavor, length, length, sequence number: */ - task->tk_auth->au_rslack = XDR_QUADLEN(len) + 4; - break; - case RPC_GSS_SVC_PRIVACY: - goto out_bad; - } + /* We leave it to unwrap to calculate au_rslack. For now we just + * calculate the length of the verifier: */ + task->tk_auth->au_verfsize = XDR_QUADLEN(len) + 2; gss_put_ctx(ctx); dprintk("RPC: %4u GSS gss_validate: gss_verify_mic succeeded.\n", task->tk_pid); @@ -1067,6 +1056,7 @@ gss_unwrap_resp(struct rpc_task *task, struct gss_cred *gss_cred = container_of(cred, struct gss_cred, gc_base); struct gss_cl_ctx *ctx = gss_cred_get_ctx(cred); + u32 *savedp = p; int status = -EIO; if (ctx->gc_proc != RPC_GSS_PROC_DATA) @@ -1082,6 +1072,8 @@ gss_unwrap_resp(struct rpc_task *task, case RPC_GSS_SVC_PRIVACY: break; } + /* take into account extra slack for integrity and privacy cases: */ + task->tk_auth->au_rslack = task->tk_auth->au_verfsize + (p - savedp); out_decode: status = decode(rqstp, p, obj); out: -- cgit v1.2.3 From bfa91516b57483fc9c81d8d90325fd2c3c16ac48 Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Thu, 13 Oct 2005 16:55:08 -0400 Subject: RPCSEC_GSS: krb5 pre-privacy cleanup The code this was originally derived from processed wrap and mic tokens using the same functions. This required some contortions, and more would be required with the addition of xdr_buf's, so it's better to separate out the two code paths. In preparation for adding privacy support, remove the last vestiges of the old wrap token code. Signed-off-by: J. Bruce Fields Signed-off-by: Trond Myklebust --- include/linux/sunrpc/gss_krb5.h | 5 ++--- net/sunrpc/auth_gss/gss_krb5_mech.c | 5 ++--- net/sunrpc/auth_gss/gss_krb5_seal.c | 38 ++++++----------------------------- net/sunrpc/auth_gss/gss_krb5_unseal.c | 30 ++++++--------------------- 4 files changed, 16 insertions(+), 62 deletions(-) (limited to 'include/linux') diff --git a/include/linux/sunrpc/gss_krb5.h b/include/linux/sunrpc/gss_krb5.h index ffe31d2eb9ec..cb35833e2ae3 100644 --- a/include/linux/sunrpc/gss_krb5.h +++ b/include/linux/sunrpc/gss_krb5.h @@ -121,13 +121,12 @@ make_checksum(s32 cksumtype, char *header, int hdrlen, struct xdr_buf *body, u32 krb5_make_token(struct krb5_ctx *context_handle, int qop_req, struct xdr_buf *input_message_buffer, - struct xdr_netobj *output_message_buffer, int toktype); + struct xdr_netobj *output_message_buffer); u32 krb5_read_token(struct krb5_ctx *context_handle, struct xdr_netobj *input_token_buffer, - struct xdr_buf *message_buffer, - int *qop_state, int toktype); + struct xdr_buf *message_buffer, int *qop_state); u32 krb5_encrypt(struct crypto_tfm * key, diff --git a/net/sunrpc/auth_gss/gss_krb5_mech.c b/net/sunrpc/auth_gss/gss_krb5_mech.c index 462c5b86b073..8b9066fdfda5 100644 --- a/net/sunrpc/auth_gss/gss_krb5_mech.c +++ b/net/sunrpc/auth_gss/gss_krb5_mech.c @@ -199,8 +199,7 @@ gss_verify_mic_kerberos(struct gss_ctx *ctx, int qop_state; struct krb5_ctx *kctx = ctx->internal_ctx_id; - maj_stat = krb5_read_token(kctx, mic_token, message, &qop_state, - KG_TOK_MIC_MSG); + maj_stat = krb5_read_token(kctx, mic_token, message, &qop_state); if (!maj_stat && qop_state) *qstate = qop_state; @@ -216,7 +215,7 @@ gss_get_mic_kerberos(struct gss_ctx *ctx, u32 err = 0; struct krb5_ctx *kctx = ctx->internal_ctx_id; - err = krb5_make_token(kctx, qop, message, mic_token, KG_TOK_MIC_MSG); + err = krb5_make_token(kctx, qop, message, mic_token); dprintk("RPC: gss_get_mic_kerberos returning %d\n",err); diff --git a/net/sunrpc/auth_gss/gss_krb5_seal.c b/net/sunrpc/auth_gss/gss_krb5_seal.c index afeeb8715a77..2511834e6e52 100644 --- a/net/sunrpc/auth_gss/gss_krb5_seal.c +++ b/net/sunrpc/auth_gss/gss_krb5_seal.c @@ -70,22 +70,12 @@ # define RPCDBG_FACILITY RPCDBG_AUTH #endif -static inline int -gss_krb5_padding(int blocksize, int length) { - /* Most of the code is block-size independent but in practice we - * use only 8: */ - BUG_ON(blocksize != 8); - return 8 - (length & 7); -} - u32 krb5_make_token(struct krb5_ctx *ctx, int qop_req, - struct xdr_buf *text, struct xdr_netobj *token, - int toktype) + struct xdr_buf *text, struct xdr_netobj *token) { s32 checksum_type; struct xdr_netobj md5cksum = {.len = 0, .data = NULL}; - int blocksize = 0, tmsglen; unsigned char *ptr, *krb5_hdr, *msg_start; s32 now; @@ -111,21 +101,13 @@ krb5_make_token(struct krb5_ctx *ctx, int qop_req, goto out_err; } - if (toktype == KG_TOK_WRAP_MSG) { - blocksize = crypto_tfm_alg_blocksize(ctx->enc); - tmsglen = blocksize + text->len - + gss_krb5_padding(blocksize, blocksize + text->len); - } else { - tmsglen = 0; - } - - token->len = g_token_size(&ctx->mech_used, 22 + tmsglen); + token->len = g_token_size(&ctx->mech_used, 22); ptr = token->data; - g_make_token_header(&ctx->mech_used, 22 + tmsglen, &ptr); + g_make_token_header(&ctx->mech_used, 22, &ptr); - *ptr++ = (unsigned char) ((toktype>>8)&0xff); - *ptr++ = (unsigned char) (toktype&0xff); + *ptr++ = (unsigned char) ((KG_TOK_MIC_MSG>>8)&0xff); + *ptr++ = (unsigned char) (KG_TOK_MIC_MSG&0xff); /* ptr now at byte 2 of header described in rfc 1964, section 1.2.1: */ krb5_hdr = ptr - 2; @@ -133,17 +115,9 @@ krb5_make_token(struct krb5_ctx *ctx, int qop_req, *(u16 *)(krb5_hdr + 2) = htons(ctx->signalg); memset(krb5_hdr + 4, 0xff, 4); - if (toktype == KG_TOK_WRAP_MSG) - *(u16 *)(krb5_hdr + 4) = htons(ctx->sealalg); - if (toktype == KG_TOK_WRAP_MSG) { - /* XXX removing support for now */ + if (make_checksum(checksum_type, krb5_hdr, 8, text, &md5cksum)) goto out_err; - } else { /* Sign only. */ - if (make_checksum(checksum_type, krb5_hdr, 8, text, - &md5cksum)) - goto out_err; - } switch (ctx->signalg) { case SGN_ALG_DES_MAC_MD5: diff --git a/net/sunrpc/auth_gss/gss_krb5_unseal.c b/net/sunrpc/auth_gss/gss_krb5_unseal.c index 8767fc53183d..19eba3df6607 100644 --- a/net/sunrpc/auth_gss/gss_krb5_unseal.c +++ b/net/sunrpc/auth_gss/gss_krb5_unseal.c @@ -68,20 +68,13 @@ #endif -/* message_buffer is an input if toktype is MIC and an output if it is WRAP: - * If toktype is MIC: read_token is a mic token, and message_buffer is the - * data that the mic was supposedly taken over. - * If toktype is WRAP: read_token is a wrap token, and message_buffer is used - * to return the decrypted data. - */ +/* read_token is a mic token, and message_buffer is the data that the mic was + * supposedly taken over. */ -/* XXX will need to change prototype and/or just split into a separate function - * when we add privacy (because read_token will be in pages too). */ u32 krb5_read_token(struct krb5_ctx *ctx, struct xdr_netobj *read_token, - struct xdr_buf *message_buffer, - int *qop_state, int toktype) + struct xdr_buf *message_buffer, int *qop_state) { int signalg; int sealalg; @@ -100,16 +93,12 @@ krb5_read_token(struct krb5_ctx *ctx, read_token->len)) goto out; - if ((*ptr++ != ((toktype>>8)&0xff)) || (*ptr++ != (toktype&0xff))) + if ((*ptr++ != ((KG_TOK_MIC_MSG>>8)&0xff)) || + (*ptr++ != ( KG_TOK_MIC_MSG &0xff)) ) goto out; /* XXX sanity-check bodysize?? */ - if (toktype == KG_TOK_WRAP_MSG) { - /* XXX gone */ - goto out; - } - /* get the sign and seal algorithms */ signalg = ptr[0] + (ptr[1] << 8); @@ -120,14 +109,7 @@ krb5_read_token(struct krb5_ctx *ctx, if ((ptr[4] != 0xff) || (ptr[5] != 0xff)) goto out; - if (((toktype != KG_TOK_WRAP_MSG) && (sealalg != 0xffff)) || - ((toktype == KG_TOK_WRAP_MSG) && (sealalg == 0xffff))) - goto out; - - /* in the current spec, there is only one valid seal algorithm per - key type, so a simple comparison is ok */ - - if ((toktype == KG_TOK_WRAP_MSG) && !(sealalg == ctx->sealalg)) + if (sealalg != 0xffff) goto out; /* there are several mappings of seal algorithms to sign algorithms, -- cgit v1.2.3 From 14ae162c24d985593d5b19437d7f3d8fd0062b59 Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Thu, 13 Oct 2005 16:55:13 -0400 Subject: RPCSEC_GSS: Add support for privacy to krb5 rpcsec_gss mechanism. Add support for privacy to the krb5 rpcsec_gss mechanism. Signed-off-by: J. Bruce Fields Signed-off-by: Trond Myklebust --- include/linux/sunrpc/gss_krb5.h | 18 +- net/sunrpc/auth_gss/Makefile | 2 +- net/sunrpc/auth_gss/gss_krb5_crypto.c | 156 +++++++++++++- net/sunrpc/auth_gss/gss_krb5_mech.c | 7 + net/sunrpc/auth_gss/gss_krb5_seal.c | 4 +- net/sunrpc/auth_gss/gss_krb5_unseal.c | 2 +- net/sunrpc/auth_gss/gss_krb5_wrap.c | 370 ++++++++++++++++++++++++++++++++++ 7 files changed, 552 insertions(+), 7 deletions(-) create mode 100644 net/sunrpc/auth_gss/gss_krb5_wrap.c (limited to 'include/linux') diff --git a/include/linux/sunrpc/gss_krb5.h b/include/linux/sunrpc/gss_krb5.h index cb35833e2ae3..7f93c2d5ebdb 100644 --- a/include/linux/sunrpc/gss_krb5.h +++ b/include/linux/sunrpc/gss_krb5.h @@ -116,7 +116,7 @@ enum seal_alg { s32 make_checksum(s32 cksumtype, char *header, int hdrlen, struct xdr_buf *body, - struct xdr_netobj *cksum); + int body_offset, struct xdr_netobj *cksum); u32 krb5_make_token(struct krb5_ctx *context_handle, int qop_req, @@ -128,6 +128,15 @@ krb5_read_token(struct krb5_ctx *context_handle, struct xdr_netobj *input_token_buffer, struct xdr_buf *message_buffer, int *qop_state); +u32 +gss_wrap_kerberos(struct gss_ctx *ctx_id, u32 qop, int offset, + struct xdr_buf *outbuf, struct page **pages); + +u32 +gss_unwrap_kerberos(struct gss_ctx *ctx_id, u32 *qop, int offset, + struct xdr_buf *buf); + + u32 krb5_encrypt(struct crypto_tfm * key, void *iv, void *in, void *out, int length); @@ -136,6 +145,13 @@ u32 krb5_decrypt(struct crypto_tfm * key, void *iv, void *in, void *out, int length); +int +gss_encrypt_xdr_buf(struct crypto_tfm *tfm, struct xdr_buf *outbuf, int offset, + struct page **pages); + +int +gss_decrypt_xdr_buf(struct crypto_tfm *tfm, struct xdr_buf *inbuf, int offset); + s32 krb5_make_seq_num(struct crypto_tfm * key, int direction, diff --git a/net/sunrpc/auth_gss/Makefile b/net/sunrpc/auth_gss/Makefile index fe1b874084bc..f3431a7e33da 100644 --- a/net/sunrpc/auth_gss/Makefile +++ b/net/sunrpc/auth_gss/Makefile @@ -10,7 +10,7 @@ auth_rpcgss-objs := auth_gss.o gss_generic_token.o \ obj-$(CONFIG_RPCSEC_GSS_KRB5) += rpcsec_gss_krb5.o rpcsec_gss_krb5-objs := gss_krb5_mech.o gss_krb5_seal.o gss_krb5_unseal.o \ - gss_krb5_seqnum.o + gss_krb5_seqnum.o gss_krb5_wrap.o obj-$(CONFIG_RPCSEC_GSS_SPKM3) += rpcsec_gss_spkm3.o diff --git a/net/sunrpc/auth_gss/gss_krb5_crypto.c b/net/sunrpc/auth_gss/gss_krb5_crypto.c index 2baf93f8b8f5..3f3d5437f02d 100644 --- a/net/sunrpc/auth_gss/gss_krb5_crypto.c +++ b/net/sunrpc/auth_gss/gss_krb5_crypto.c @@ -218,7 +218,7 @@ checksummer(struct scatterlist *sg, void *data) /* checksum the plaintext data and hdrlen bytes of the token header */ s32 make_checksum(s32 cksumtype, char *header, int hdrlen, struct xdr_buf *body, - struct xdr_netobj *cksum) + int body_offset, struct xdr_netobj *cksum) { char *cksumname; struct crypto_tfm *tfm = NULL; /* XXX add to ctx? */ @@ -243,7 +243,8 @@ make_checksum(s32 cksumtype, char *header, int hdrlen, struct xdr_buf *body, crypto_digest_init(tfm); buf_to_sg(sg, header, hdrlen); crypto_digest_update(tfm, sg, 1); - process_xdr_buf(body, 0, body->len, checksummer, tfm); + process_xdr_buf(body, body_offset, body->len - body_offset, + checksummer, tfm); crypto_digest_final(tfm, cksum->data); code = 0; out: @@ -252,3 +253,154 @@ out: } EXPORT_SYMBOL(make_checksum); + +struct encryptor_desc { + u8 iv[8]; /* XXX hard-coded blocksize */ + struct crypto_tfm *tfm; + int pos; + struct xdr_buf *outbuf; + struct page **pages; + struct scatterlist infrags[4]; + struct scatterlist outfrags[4]; + int fragno; + int fraglen; +}; + +static int +encryptor(struct scatterlist *sg, void *data) +{ + struct encryptor_desc *desc = data; + struct xdr_buf *outbuf = desc->outbuf; + struct page *in_page; + int thislen = desc->fraglen + sg->length; + int fraglen, ret; + int page_pos; + + /* Worst case is 4 fragments: head, end of page 1, start + * of page 2, tail. Anything more is a bug. */ + BUG_ON(desc->fragno > 3); + desc->infrags[desc->fragno] = *sg; + desc->outfrags[desc->fragno] = *sg; + + page_pos = desc->pos - outbuf->head[0].iov_len; + if (page_pos >= 0 && page_pos < outbuf->page_len) { + /* pages are not in place: */ + int i = (page_pos + outbuf->page_base) >> PAGE_CACHE_SHIFT; + in_page = desc->pages[i]; + } else { + in_page = sg->page; + } + desc->infrags[desc->fragno].page = in_page; + desc->fragno++; + desc->fraglen += sg->length; + desc->pos += sg->length; + + fraglen = thislen & 7; /* XXX hardcoded blocksize */ + thislen -= fraglen; + + if (thislen == 0) + return 0; + + ret = crypto_cipher_encrypt_iv(desc->tfm, desc->outfrags, desc->infrags, + thislen, desc->iv); + if (ret) + return ret; + if (fraglen) { + desc->outfrags[0].page = sg->page; + desc->outfrags[0].offset = sg->offset + sg->length - fraglen; + desc->outfrags[0].length = fraglen; + desc->infrags[0] = desc->outfrags[0]; + desc->infrags[0].page = in_page; + desc->fragno = 1; + desc->fraglen = fraglen; + } else { + desc->fragno = 0; + desc->fraglen = 0; + } + return 0; +} + +int +gss_encrypt_xdr_buf(struct crypto_tfm *tfm, struct xdr_buf *buf, int offset, + struct page **pages) +{ + int ret; + struct encryptor_desc desc; + + BUG_ON((buf->len - offset) % crypto_tfm_alg_blocksize(tfm) != 0); + + memset(desc.iv, 0, sizeof(desc.iv)); + desc.tfm = tfm; + desc.pos = offset; + desc.outbuf = buf; + desc.pages = pages; + desc.fragno = 0; + desc.fraglen = 0; + + ret = process_xdr_buf(buf, offset, buf->len - offset, encryptor, &desc); + return ret; +} + +EXPORT_SYMBOL(gss_encrypt_xdr_buf); + +struct decryptor_desc { + u8 iv[8]; /* XXX hard-coded blocksize */ + struct crypto_tfm *tfm; + struct scatterlist frags[4]; + int fragno; + int fraglen; +}; + +static int +decryptor(struct scatterlist *sg, void *data) +{ + struct decryptor_desc *desc = data; + int thislen = desc->fraglen + sg->length; + int fraglen, ret; + + /* Worst case is 4 fragments: head, end of page 1, start + * of page 2, tail. Anything more is a bug. */ + BUG_ON(desc->fragno > 3); + desc->frags[desc->fragno] = *sg; + desc->fragno++; + desc->fraglen += sg->length; + + fraglen = thislen & 7; /* XXX hardcoded blocksize */ + thislen -= fraglen; + + if (thislen == 0) + return 0; + + ret = crypto_cipher_decrypt_iv(desc->tfm, desc->frags, desc->frags, + thislen, desc->iv); + if (ret) + return ret; + if (fraglen) { + desc->frags[0].page = sg->page; + desc->frags[0].offset = sg->offset + sg->length - fraglen; + desc->frags[0].length = fraglen; + desc->fragno = 1; + desc->fraglen = fraglen; + } else { + desc->fragno = 0; + desc->fraglen = 0; + } + return 0; +} + +int +gss_decrypt_xdr_buf(struct crypto_tfm *tfm, struct xdr_buf *buf, int offset) +{ + struct decryptor_desc desc; + + /* XXXJBF: */ + BUG_ON((buf->len - offset) % crypto_tfm_alg_blocksize(tfm) != 0); + + memset(desc.iv, 0, sizeof(desc.iv)); + desc.tfm = tfm; + desc.fragno = 0; + desc.fraglen = 0; + return process_xdr_buf(buf, offset, buf->len - offset, decryptor, &desc); +} + +EXPORT_SYMBOL(gss_decrypt_xdr_buf); diff --git a/net/sunrpc/auth_gss/gss_krb5_mech.c b/net/sunrpc/auth_gss/gss_krb5_mech.c index 8b9066fdfda5..37a9ad97ccd4 100644 --- a/net/sunrpc/auth_gss/gss_krb5_mech.c +++ b/net/sunrpc/auth_gss/gss_krb5_mech.c @@ -226,6 +226,8 @@ static struct gss_api_ops gss_kerberos_ops = { .gss_import_sec_context = gss_import_sec_context_kerberos, .gss_get_mic = gss_get_mic_kerberos, .gss_verify_mic = gss_verify_mic_kerberos, + .gss_wrap = gss_wrap_kerberos, + .gss_unwrap = gss_unwrap_kerberos, .gss_delete_sec_context = gss_delete_sec_context_kerberos, }; @@ -240,6 +242,11 @@ static struct pf_desc gss_kerberos_pfs[] = { .service = RPC_GSS_SVC_INTEGRITY, .name = "krb5i", }, + [2] = { + .pseudoflavor = RPC_AUTH_GSS_KRB5P, + .service = RPC_GSS_SVC_PRIVACY, + .name = "krb5p", + }, }; static struct gss_api_mech gss_kerberos_mech = { diff --git a/net/sunrpc/auth_gss/gss_krb5_seal.c b/net/sunrpc/auth_gss/gss_krb5_seal.c index 2511834e6e52..fb852d9ab06f 100644 --- a/net/sunrpc/auth_gss/gss_krb5_seal.c +++ b/net/sunrpc/auth_gss/gss_krb5_seal.c @@ -116,8 +116,8 @@ krb5_make_token(struct krb5_ctx *ctx, int qop_req, *(u16 *)(krb5_hdr + 2) = htons(ctx->signalg); memset(krb5_hdr + 4, 0xff, 4); - if (make_checksum(checksum_type, krb5_hdr, 8, text, &md5cksum)) - goto out_err; + if (make_checksum(checksum_type, krb5_hdr, 8, text, 0, &md5cksum)) + goto out_err; switch (ctx->signalg) { case SGN_ALG_DES_MAC_MD5: diff --git a/net/sunrpc/auth_gss/gss_krb5_unseal.c b/net/sunrpc/auth_gss/gss_krb5_unseal.c index 19eba3df6607..c3d6d1bc100c 100644 --- a/net/sunrpc/auth_gss/gss_krb5_unseal.c +++ b/net/sunrpc/auth_gss/gss_krb5_unseal.c @@ -136,7 +136,7 @@ krb5_read_token(struct krb5_ctx *ctx, switch (signalg) { case SGN_ALG_DES_MAC_MD5: ret = make_checksum(checksum_type, ptr - 2, 8, - message_buffer, &md5cksum); + message_buffer, 0, &md5cksum); if (ret) goto out; diff --git a/net/sunrpc/auth_gss/gss_krb5_wrap.c b/net/sunrpc/auth_gss/gss_krb5_wrap.c new file mode 100644 index 000000000000..ddcde6e42b23 --- /dev/null +++ b/net/sunrpc/auth_gss/gss_krb5_wrap.c @@ -0,0 +1,370 @@ +#include +#include +#include +#include +#include +#include +#include +#include + +#ifdef RPC_DEBUG +# define RPCDBG_FACILITY RPCDBG_AUTH +#endif + +static inline int +gss_krb5_padding(int blocksize, int length) +{ + /* Most of the code is block-size independent but currently we + * use only 8: */ + BUG_ON(blocksize != 8); + return 8 - (length & 7); +} + +static inline void +gss_krb5_add_padding(struct xdr_buf *buf, int offset, int blocksize) +{ + int padding = gss_krb5_padding(blocksize, buf->len - offset); + char *p; + struct kvec *iov; + + if (buf->page_len || buf->tail[0].iov_len) + iov = &buf->tail[0]; + else + iov = &buf->head[0]; + p = iov->iov_base + iov->iov_len; + iov->iov_len += padding; + buf->len += padding; + memset(p, padding, padding); +} + +static inline int +gss_krb5_remove_padding(struct xdr_buf *buf, int blocksize) +{ + u8 *ptr; + u8 pad; + int len = buf->len; + + if (len <= buf->head[0].iov_len) { + pad = *(u8 *)(buf->head[0].iov_base + len - 1); + if (pad > buf->head[0].iov_len) + return -EINVAL; + buf->head[0].iov_len -= pad; + goto out; + } else + len -= buf->head[0].iov_len; + if (len <= buf->page_len) { + int last = (buf->page_base + len - 1) + >>PAGE_CACHE_SHIFT; + int offset = (buf->page_base + len - 1) + & (PAGE_CACHE_SIZE - 1); + ptr = kmap_atomic(buf->pages[last], KM_SKB_SUNRPC_DATA); + pad = *(ptr + offset); + kunmap_atomic(ptr, KM_SKB_SUNRPC_DATA); + goto out; + } else + len -= buf->page_len; + BUG_ON(len > buf->tail[0].iov_len); + pad = *(u8 *)(buf->tail[0].iov_base + len - 1); +out: + /* XXX: NOTE: we do not adjust the page lengths--they represent + * a range of data in the real filesystem page cache, and we need + * to know that range so the xdr code can properly place read data. + * However adjusting the head length, as we do above, is harmless. + * In the case of a request that fits into a single page, the server + * also uses length and head length together to determine the original + * start of the request to copy the request for deferal; so it's + * easier on the server if we adjust head and tail length in tandem. + * It's not really a problem that we don't fool with the page and + * tail lengths, though--at worst badly formed xdr might lead the + * server to attempt to parse the padding. + * XXX: Document all these weird requirements for gss mechanism + * wrap/unwrap functions. */ + if (pad > blocksize) + return -EINVAL; + if (buf->len > pad) + buf->len -= pad; + else + return -EINVAL; + return 0; +} + +static inline void +make_confounder(char *p, int blocksize) +{ + static u64 i = 0; + u64 *q = (u64 *)p; + + /* rfc1964 claims this should be "random". But all that's really + * necessary is that it be unique. And not even that is necessary in + * our case since our "gssapi" implementation exists only to support + * rpcsec_gss, so we know that the only buffers we will ever encrypt + * already begin with a unique sequence number. Just to hedge my bets + * I'll make a half-hearted attempt at something unique, but ensuring + * uniqueness would mean worrying about atomicity and rollover, and I + * don't care enough. */ + + BUG_ON(blocksize != 8); + *q = i++; +} + +/* Assumptions: the head and tail of inbuf are ours to play with. + * The pages, however, may be real pages in the page cache and we replace + * them with scratch pages from **pages before writing to them. */ +/* XXX: obviously the above should be documentation of wrap interface, + * and shouldn't be in this kerberos-specific file. */ + +/* XXX factor out common code with seal/unseal. */ + +u32 +gss_wrap_kerberos(struct gss_ctx *ctx, u32 qop, int offset, + struct xdr_buf *buf, struct page **pages) +{ + struct krb5_ctx *kctx = ctx->internal_ctx_id; + s32 checksum_type; + struct xdr_netobj md5cksum = {.len = 0, .data = NULL}; + int blocksize = 0, plainlen; + unsigned char *ptr, *krb5_hdr, *msg_start; + s32 now; + int headlen; + struct page **tmp_pages; + + dprintk("RPC: gss_wrap_kerberos\n"); + + now = get_seconds(); + + if (qop != 0) + goto out_err; + + switch (kctx->signalg) { + case SGN_ALG_DES_MAC_MD5: + checksum_type = CKSUMTYPE_RSA_MD5; + break; + default: + dprintk("RPC: gss_krb5_seal: kctx->signalg %d not" + " supported\n", kctx->signalg); + goto out_err; + } + if (kctx->sealalg != SEAL_ALG_NONE && kctx->sealalg != SEAL_ALG_DES) { + dprintk("RPC: gss_krb5_seal: kctx->sealalg %d not supported\n", + kctx->sealalg); + goto out_err; + } + + blocksize = crypto_tfm_alg_blocksize(kctx->enc); + gss_krb5_add_padding(buf, offset, blocksize); + BUG_ON((buf->len - offset) % blocksize); + plainlen = blocksize + buf->len - offset; + + headlen = g_token_size(&kctx->mech_used, 22 + plainlen) - + (buf->len - offset); + + ptr = buf->head[0].iov_base + offset; + /* shift data to make room for header. */ + /* XXX Would be cleverer to encrypt while copying. */ + /* XXX bounds checking, slack, etc. */ + memmove(ptr + headlen, ptr, buf->head[0].iov_len - offset); + buf->head[0].iov_len += headlen; + buf->len += headlen; + BUG_ON((buf->len - offset - headlen) % blocksize); + + g_make_token_header(&kctx->mech_used, 22 + plainlen, &ptr); + + + *ptr++ = (unsigned char) ((KG_TOK_WRAP_MSG>>8)&0xff); + *ptr++ = (unsigned char) (KG_TOK_WRAP_MSG&0xff); + + /* ptr now at byte 2 of header described in rfc 1964, section 1.2.1: */ + krb5_hdr = ptr - 2; + msg_start = krb5_hdr + 24; + /* XXXJBF: */ BUG_ON(buf->head[0].iov_base + offset + headlen != msg_start + blocksize); + + *(u16 *)(krb5_hdr + 2) = htons(kctx->signalg); + memset(krb5_hdr + 4, 0xff, 4); + *(u16 *)(krb5_hdr + 4) = htons(kctx->sealalg); + + make_confounder(msg_start, blocksize); + + /* XXXJBF: UGH!: */ + tmp_pages = buf->pages; + buf->pages = pages; + if (make_checksum(checksum_type, krb5_hdr, 8, buf, + offset + headlen - blocksize, &md5cksum)) + goto out_err; + buf->pages = tmp_pages; + + switch (kctx->signalg) { + case SGN_ALG_DES_MAC_MD5: + if (krb5_encrypt(kctx->seq, NULL, md5cksum.data, + md5cksum.data, md5cksum.len)) + goto out_err; + memcpy(krb5_hdr + 16, + md5cksum.data + md5cksum.len - KRB5_CKSUM_LENGTH, + KRB5_CKSUM_LENGTH); + + dprintk("RPC: make_seal_token: cksum data: \n"); + print_hexl((u32 *) (krb5_hdr + 16), KRB5_CKSUM_LENGTH, 0); + break; + default: + BUG(); + } + + kfree(md5cksum.data); + + /* XXX would probably be more efficient to compute checksum + * and encrypt at the same time: */ + if ((krb5_make_seq_num(kctx->seq, kctx->initiate ? 0 : 0xff, + kctx->seq_send, krb5_hdr + 16, krb5_hdr + 8))) + goto out_err; + + if (gss_encrypt_xdr_buf(kctx->enc, buf, offset + headlen - blocksize, + pages)) + goto out_err; + + kctx->seq_send++; + + return ((kctx->endtime < now) ? GSS_S_CONTEXT_EXPIRED : GSS_S_COMPLETE); +out_err: + if (md5cksum.data) kfree(md5cksum.data); + return GSS_S_FAILURE; +} + +u32 +gss_unwrap_kerberos(struct gss_ctx *ctx, u32 *qop, int offset, + struct xdr_buf *buf) +{ + struct krb5_ctx *kctx = ctx->internal_ctx_id; + int signalg; + int sealalg; + s32 checksum_type; + struct xdr_netobj md5cksum = {.len = 0, .data = NULL}; + s32 now; + int direction; + s32 seqnum; + unsigned char *ptr; + int bodysize; + u32 ret = GSS_S_DEFECTIVE_TOKEN; + void *data_start, *orig_start; + int data_len; + int blocksize; + + dprintk("RPC: gss_unwrap_kerberos\n"); + + ptr = (u8 *)buf->head[0].iov_base + offset; + if (g_verify_token_header(&kctx->mech_used, &bodysize, &ptr, + buf->len - offset)) + goto out; + + if ((*ptr++ != ((KG_TOK_WRAP_MSG>>8)&0xff)) || + (*ptr++ != (KG_TOK_WRAP_MSG &0xff)) ) + goto out; + + /* XXX sanity-check bodysize?? */ + + /* get the sign and seal algorithms */ + + signalg = ptr[0] + (ptr[1] << 8); + sealalg = ptr[2] + (ptr[3] << 8); + + /* Sanity checks */ + + if ((ptr[4] != 0xff) || (ptr[5] != 0xff)) + goto out; + + if (sealalg == 0xffff) + goto out; + + /* in the current spec, there is only one valid seal algorithm per + key type, so a simple comparison is ok */ + + if (sealalg != kctx->sealalg) + goto out; + + /* there are several mappings of seal algorithms to sign algorithms, + but few enough that we can try them all. */ + + if ((kctx->sealalg == SEAL_ALG_NONE && signalg > 1) || + (kctx->sealalg == SEAL_ALG_1 && signalg != SGN_ALG_3) || + (kctx->sealalg == SEAL_ALG_DES3KD && + signalg != SGN_ALG_HMAC_SHA1_DES3_KD)) + goto out; + + if (gss_decrypt_xdr_buf(kctx->enc, buf, + ptr + 22 - (unsigned char *)buf->head[0].iov_base)) + goto out; + + /* compute the checksum of the message */ + + /* initialize the the cksum */ + switch (signalg) { + case SGN_ALG_DES_MAC_MD5: + checksum_type = CKSUMTYPE_RSA_MD5; + break; + default: + ret = GSS_S_DEFECTIVE_TOKEN; + goto out; + } + + switch (signalg) { + case SGN_ALG_DES_MAC_MD5: + ret = make_checksum(checksum_type, ptr - 2, 8, buf, + ptr + 22 - (unsigned char *)buf->head[0].iov_base, &md5cksum); + if (ret) + goto out; + + ret = krb5_encrypt(kctx->seq, NULL, md5cksum.data, + md5cksum.data, md5cksum.len); + if (ret) + goto out; + + if (memcmp(md5cksum.data + 8, ptr + 14, 8)) { + ret = GSS_S_BAD_SIG; + goto out; + } + break; + default: + ret = GSS_S_DEFECTIVE_TOKEN; + goto out; + } + + /* it got through unscathed. Make sure the context is unexpired */ + + if (qop) + *qop = GSS_C_QOP_DEFAULT; + + now = get_seconds(); + + ret = GSS_S_CONTEXT_EXPIRED; + if (now > kctx->endtime) + goto out; + + /* do sequencing checks */ + + ret = GSS_S_BAD_SIG; + if ((ret = krb5_get_seq_num(kctx->seq, ptr + 14, ptr + 6, &direction, + &seqnum))) + goto out; + + if ((kctx->initiate && direction != 0xff) || + (!kctx->initiate && direction != 0)) + goto out; + + /* Copy the data back to the right position. XXX: Would probably be + * better to copy and encrypt at the same time. */ + + blocksize = crypto_tfm_alg_blocksize(kctx->enc); + data_start = ptr + 22 + blocksize; + orig_start = buf->head[0].iov_base + offset; + data_len = (buf->head[0].iov_base + buf->head[0].iov_len) - data_start; + memmove(orig_start, data_start, data_len); + buf->head[0].iov_len -= (data_start - orig_start); + buf->len -= (data_start - orig_start); + + ret = GSS_S_DEFECTIVE_TOKEN; + if (gss_krb5_remove_padding(buf, blocksize)) + goto out; + + ret = GSS_S_COMPLETE; +out: + if (md5cksum.data) kfree(md5cksum.data); + return ret; +} -- cgit v1.2.3 From 00fd6e14255fe7a249315746386d640bc4e9e758 Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Thu, 13 Oct 2005 16:55:18 -0400 Subject: RPCSEC_GSS remove all qop parameters Not only are the qop parameters that are passed around throughout the gssapi unused by any currently implemented mechanism, but there appears to be some doubt as to whether they will ever be used. Let's just kill them off for now. Signed-off-by: J. Bruce Fields Signed-off-by: Trond Myklebust --- include/linux/sunrpc/gss_api.h | 13 ++----------- include/linux/sunrpc/gss_err.h | 10 ---------- include/linux/sunrpc/gss_krb5.h | 8 ++++---- include/linux/sunrpc/gss_spkm3.h | 4 ++-- net/sunrpc/auth_gss/auth_gss.c | 20 +++++++------------- net/sunrpc/auth_gss/gss_krb5_mech.c | 12 ++++-------- net/sunrpc/auth_gss/gss_krb5_seal.c | 5 +---- net/sunrpc/auth_gss/gss_krb5_unseal.c | 5 +---- net/sunrpc/auth_gss/gss_krb5_wrap.c | 11 ++--------- net/sunrpc/auth_gss/gss_mech_switch.c | 14 ++++---------- net/sunrpc/auth_gss/gss_spkm3_mech.c | 21 ++++++++------------- net/sunrpc/auth_gss/gss_spkm3_seal.c | 4 +--- net/sunrpc/auth_gss/gss_spkm3_unseal.c | 2 +- net/sunrpc/auth_gss/svcauth_gss.c | 9 ++++----- 14 files changed, 41 insertions(+), 97 deletions(-) (limited to 'include/linux') diff --git a/include/linux/sunrpc/gss_api.h b/include/linux/sunrpc/gss_api.h index e896752ffbf9..9b8bcf125c18 100644 --- a/include/linux/sunrpc/gss_api.h +++ b/include/linux/sunrpc/gss_api.h @@ -40,23 +40,19 @@ int gss_import_sec_context( struct gss_ctx **ctx_id); u32 gss_get_mic( struct gss_ctx *ctx_id, - u32 qop, struct xdr_buf *message, struct xdr_netobj *mic_token); u32 gss_verify_mic( struct gss_ctx *ctx_id, struct xdr_buf *message, - struct xdr_netobj *mic_token, - u32 *qstate); + struct xdr_netobj *mic_token); u32 gss_wrap( struct gss_ctx *ctx_id, - u32 qop, int offset, struct xdr_buf *outbuf, struct page **inpages); u32 gss_unwrap( struct gss_ctx *ctx_id, - u32 *qop, int offset, struct xdr_buf *inbuf); u32 gss_delete_sec_context( @@ -67,7 +63,6 @@ char *gss_service_to_auth_domain_name(struct gss_api_mech *, u32 service); struct pf_desc { u32 pseudoflavor; - u32 qop; u32 service; char *name; char *auth_domain_name; @@ -96,23 +91,19 @@ struct gss_api_ops { struct gss_ctx *ctx_id); u32 (*gss_get_mic)( struct gss_ctx *ctx_id, - u32 qop, struct xdr_buf *message, struct xdr_netobj *mic_token); u32 (*gss_verify_mic)( struct gss_ctx *ctx_id, struct xdr_buf *message, - struct xdr_netobj *mic_token, - u32 *qstate); + struct xdr_netobj *mic_token); u32 (*gss_wrap)( struct gss_ctx *ctx_id, - u32 qop, int offset, struct xdr_buf *outbuf, struct page **inpages); u32 (*gss_unwrap)( struct gss_ctx *ctx_id, - u32 *qop, int offset, struct xdr_buf *buf); void (*gss_delete_sec_context)( diff --git a/include/linux/sunrpc/gss_err.h b/include/linux/sunrpc/gss_err.h index 92608a2e574c..a6807867bd21 100644 --- a/include/linux/sunrpc/gss_err.h +++ b/include/linux/sunrpc/gss_err.h @@ -65,16 +65,6 @@ typedef unsigned int OM_uint32; #define GSS_C_MECH_CODE 2 -/* - * Define the default Quality of Protection for per-message services. Note - * that an implementation that offers multiple levels of QOP may either reserve - * a value (for example zero, as assumed here) to mean "default protection", or - * alternatively may simply equate GSS_C_QOP_DEFAULT to a specific explicit - * QOP value. However a value of 0 should always be interpreted by a GSSAPI - * implementation as a request for the default protection level. - */ -#define GSS_C_QOP_DEFAULT 0 - /* * Expiration time of 2^32-1 seconds means infinite lifetime for a * credential or security context diff --git a/include/linux/sunrpc/gss_krb5.h b/include/linux/sunrpc/gss_krb5.h index 7f93c2d5ebdb..a7bda4edb853 100644 --- a/include/linux/sunrpc/gss_krb5.h +++ b/include/linux/sunrpc/gss_krb5.h @@ -119,21 +119,21 @@ make_checksum(s32 cksumtype, char *header, int hdrlen, struct xdr_buf *body, int body_offset, struct xdr_netobj *cksum); u32 -krb5_make_token(struct krb5_ctx *context_handle, int qop_req, +krb5_make_token(struct krb5_ctx *context_handle, struct xdr_buf *input_message_buffer, struct xdr_netobj *output_message_buffer); u32 krb5_read_token(struct krb5_ctx *context_handle, struct xdr_netobj *input_token_buffer, - struct xdr_buf *message_buffer, int *qop_state); + struct xdr_buf *message_buffer); u32 -gss_wrap_kerberos(struct gss_ctx *ctx_id, u32 qop, int offset, +gss_wrap_kerberos(struct gss_ctx *ctx_id, int offset, struct xdr_buf *outbuf, struct page **pages); u32 -gss_unwrap_kerberos(struct gss_ctx *ctx_id, u32 *qop, int offset, +gss_unwrap_kerberos(struct gss_ctx *ctx_id, int offset, struct xdr_buf *buf); diff --git a/include/linux/sunrpc/gss_spkm3.h b/include/linux/sunrpc/gss_spkm3.h index b5c9968c3c17..0beb2cf00a84 100644 --- a/include/linux/sunrpc/gss_spkm3.h +++ b/include/linux/sunrpc/gss_spkm3.h @@ -41,9 +41,9 @@ struct spkm3_ctx { #define SPKM_WRAP_TOK 5 #define SPKM_DEL_TOK 6 -u32 spkm3_make_token(struct spkm3_ctx *ctx, int qop_req, struct xdr_buf * text, struct xdr_netobj * token, int toktype); +u32 spkm3_make_token(struct spkm3_ctx *ctx, struct xdr_buf * text, struct xdr_netobj * token, int toktype); -u32 spkm3_read_token(struct spkm3_ctx *ctx, struct xdr_netobj *read_token, struct xdr_buf *message_buffer, int *qop_state, int toktype); +u32 spkm3_read_token(struct spkm3_ctx *ctx, struct xdr_netobj *read_token, struct xdr_buf *message_buffer, int toktype); #define CKSUMTYPE_RSA_MD5 0x0007 diff --git a/net/sunrpc/auth_gss/auth_gss.c b/net/sunrpc/auth_gss/auth_gss.c index 5e4872058ec7..f44f46f1d8e0 100644 --- a/net/sunrpc/auth_gss/auth_gss.c +++ b/net/sunrpc/auth_gss/auth_gss.c @@ -854,9 +854,7 @@ gss_marshal(struct rpc_task *task, u32 *p) *p++ = htonl(RPC_AUTH_GSS); mic.data = (u8 *)(p + 1); - maj_stat = gss_get_mic(ctx->gc_gss_ctx, - GSS_C_QOP_DEFAULT, - &verf_buf, &mic); + maj_stat = gss_get_mic(ctx->gc_gss_ctx, &verf_buf, &mic); if (maj_stat == GSS_S_CONTEXT_EXPIRED) { cred->cr_flags &= ~RPCAUTH_CRED_UPTODATE; } else if (maj_stat != 0) { @@ -888,7 +886,7 @@ gss_validate(struct rpc_task *task, u32 *p) { struct rpc_cred *cred = task->tk_msg.rpc_cred; struct gss_cl_ctx *ctx = gss_cred_get_ctx(cred); - u32 seq, qop_state; + u32 seq; struct kvec iov; struct xdr_buf verf_buf; struct xdr_netobj mic; @@ -909,7 +907,7 @@ gss_validate(struct rpc_task *task, u32 *p) mic.data = (u8 *)p; mic.len = len; - maj_stat = gss_verify_mic(ctx->gc_gss_ctx, &verf_buf, &mic, &qop_state); + maj_stat = gss_verify_mic(ctx->gc_gss_ctx, &verf_buf, &mic); if (maj_stat == GSS_S_CONTEXT_EXPIRED) cred->cr_flags &= ~RPCAUTH_CRED_UPTODATE; if (maj_stat) @@ -961,8 +959,7 @@ gss_wrap_req_integ(struct rpc_cred *cred, struct gss_cl_ctx *ctx, p = iov->iov_base + iov->iov_len; mic.data = (u8 *)(p + 1); - maj_stat = gss_get_mic(ctx->gc_gss_ctx, - GSS_C_QOP_DEFAULT, &integ_buf, &mic); + maj_stat = gss_get_mic(ctx->gc_gss_ctx, &integ_buf, &mic); status = -EIO; /* XXX? */ if (maj_stat == GSS_S_CONTEXT_EXPIRED) cred->cr_flags &= ~RPCAUTH_CRED_UPTODATE; @@ -1057,8 +1054,7 @@ gss_wrap_req_priv(struct rpc_cred *cred, struct gss_cl_ctx *ctx, memcpy(tmp, snd_buf->tail[0].iov_base, snd_buf->tail[0].iov_len); snd_buf->tail[0].iov_base = tmp; } - maj_stat = gss_wrap(ctx->gc_gss_ctx, GSS_C_QOP_DEFAULT, offset, - snd_buf, inpages); + maj_stat = gss_wrap(ctx->gc_gss_ctx, offset, snd_buf, inpages); /* RPC_SLACK_SPACE should prevent this ever happening: */ BUG_ON(snd_buf->len > snd_buf->buflen); status = -EIO; @@ -1150,8 +1146,7 @@ gss_unwrap_resp_integ(struct rpc_cred *cred, struct gss_cl_ctx *ctx, if (xdr_buf_read_netobj(rcv_buf, &mic, mic_offset)) return status; - maj_stat = gss_verify_mic(ctx->gc_gss_ctx, &integ_buf, - &mic, NULL); + maj_stat = gss_verify_mic(ctx->gc_gss_ctx, &integ_buf, &mic); if (maj_stat == GSS_S_CONTEXT_EXPIRED) cred->cr_flags &= ~RPCAUTH_CRED_UPTODATE; if (maj_stat != GSS_S_COMPLETE) @@ -1176,8 +1171,7 @@ gss_unwrap_resp_priv(struct rpc_cred *cred, struct gss_cl_ctx *ctx, /* remove padding: */ rcv_buf->len = offset + opaque_len; - maj_stat = gss_unwrap(ctx->gc_gss_ctx, NULL, - offset, rcv_buf); + maj_stat = gss_unwrap(ctx->gc_gss_ctx, offset, rcv_buf); if (maj_stat == GSS_S_CONTEXT_EXPIRED) cred->cr_flags &= ~RPCAUTH_CRED_UPTODATE; if (maj_stat != GSS_S_COMPLETE) diff --git a/net/sunrpc/auth_gss/gss_krb5_mech.c b/net/sunrpc/auth_gss/gss_krb5_mech.c index 37a9ad97ccd4..9ffac2c50b94 100644 --- a/net/sunrpc/auth_gss/gss_krb5_mech.c +++ b/net/sunrpc/auth_gss/gss_krb5_mech.c @@ -193,15 +193,12 @@ gss_delete_sec_context_kerberos(void *internal_ctx) { static u32 gss_verify_mic_kerberos(struct gss_ctx *ctx, struct xdr_buf *message, - struct xdr_netobj *mic_token, - u32 *qstate) { + struct xdr_netobj *mic_token) +{ u32 maj_stat = 0; - int qop_state; struct krb5_ctx *kctx = ctx->internal_ctx_id; - maj_stat = krb5_read_token(kctx, mic_token, message, &qop_state); - if (!maj_stat && qop_state) - *qstate = qop_state; + maj_stat = krb5_read_token(kctx, mic_token, message); dprintk("RPC: gss_verify_mic_kerberos returning %d\n", maj_stat); return maj_stat; @@ -209,13 +206,12 @@ gss_verify_mic_kerberos(struct gss_ctx *ctx, static u32 gss_get_mic_kerberos(struct gss_ctx *ctx, - u32 qop, struct xdr_buf *message, struct xdr_netobj *mic_token) { u32 err = 0; struct krb5_ctx *kctx = ctx->internal_ctx_id; - err = krb5_make_token(kctx, qop, message, mic_token); + err = krb5_make_token(kctx, message, mic_token); dprintk("RPC: gss_get_mic_kerberos returning %d\n",err); diff --git a/net/sunrpc/auth_gss/gss_krb5_seal.c b/net/sunrpc/auth_gss/gss_krb5_seal.c index fb852d9ab06f..15227c727c8b 100644 --- a/net/sunrpc/auth_gss/gss_krb5_seal.c +++ b/net/sunrpc/auth_gss/gss_krb5_seal.c @@ -71,7 +71,7 @@ #endif u32 -krb5_make_token(struct krb5_ctx *ctx, int qop_req, +krb5_make_token(struct krb5_ctx *ctx, struct xdr_buf *text, struct xdr_netobj *token) { s32 checksum_type; @@ -83,9 +83,6 @@ krb5_make_token(struct krb5_ctx *ctx, int qop_req, now = get_seconds(); - if (qop_req != 0) - goto out_err; - switch (ctx->signalg) { case SGN_ALG_DES_MAC_MD5: checksum_type = CKSUMTYPE_RSA_MD5; diff --git a/net/sunrpc/auth_gss/gss_krb5_unseal.c b/net/sunrpc/auth_gss/gss_krb5_unseal.c index c3d6d1bc100c..bcf978627a71 100644 --- a/net/sunrpc/auth_gss/gss_krb5_unseal.c +++ b/net/sunrpc/auth_gss/gss_krb5_unseal.c @@ -74,7 +74,7 @@ u32 krb5_read_token(struct krb5_ctx *ctx, struct xdr_netobj *read_token, - struct xdr_buf *message_buffer, int *qop_state) + struct xdr_buf *message_buffer) { int signalg; int sealalg; @@ -157,9 +157,6 @@ krb5_read_token(struct krb5_ctx *ctx, /* it got through unscathed. Make sure the context is unexpired */ - if (qop_state) - *qop_state = GSS_C_QOP_DEFAULT; - now = get_seconds(); ret = GSS_S_CONTEXT_EXPIRED; diff --git a/net/sunrpc/auth_gss/gss_krb5_wrap.c b/net/sunrpc/auth_gss/gss_krb5_wrap.c index ddcde6e42b23..af777cf9f251 100644 --- a/net/sunrpc/auth_gss/gss_krb5_wrap.c +++ b/net/sunrpc/auth_gss/gss_krb5_wrap.c @@ -116,7 +116,7 @@ make_confounder(char *p, int blocksize) /* XXX factor out common code with seal/unseal. */ u32 -gss_wrap_kerberos(struct gss_ctx *ctx, u32 qop, int offset, +gss_wrap_kerberos(struct gss_ctx *ctx, int offset, struct xdr_buf *buf, struct page **pages) { struct krb5_ctx *kctx = ctx->internal_ctx_id; @@ -132,9 +132,6 @@ gss_wrap_kerberos(struct gss_ctx *ctx, u32 qop, int offset, now = get_seconds(); - if (qop != 0) - goto out_err; - switch (kctx->signalg) { case SGN_ALG_DES_MAC_MD5: checksum_type = CKSUMTYPE_RSA_MD5; @@ -229,8 +226,7 @@ out_err: } u32 -gss_unwrap_kerberos(struct gss_ctx *ctx, u32 *qop, int offset, - struct xdr_buf *buf) +gss_unwrap_kerberos(struct gss_ctx *ctx, int offset, struct xdr_buf *buf) { struct krb5_ctx *kctx = ctx->internal_ctx_id; int signalg; @@ -328,9 +324,6 @@ gss_unwrap_kerberos(struct gss_ctx *ctx, u32 *qop, int offset, /* it got through unscathed. Make sure the context is unexpired */ - if (qop) - *qop = GSS_C_QOP_DEFAULT; - now = get_seconds(); ret = GSS_S_CONTEXT_EXPIRED; diff --git a/net/sunrpc/auth_gss/gss_mech_switch.c b/net/sunrpc/auth_gss/gss_mech_switch.c index 06d97cb3481a..b048bf672da2 100644 --- a/net/sunrpc/auth_gss/gss_mech_switch.c +++ b/net/sunrpc/auth_gss/gss_mech_switch.c @@ -250,13 +250,11 @@ gss_import_sec_context(const void *input_token, size_t bufsize, u32 gss_get_mic(struct gss_ctx *context_handle, - u32 qop, struct xdr_buf *message, struct xdr_netobj *mic_token) { return context_handle->mech_type->gm_ops ->gss_get_mic(context_handle, - qop, message, mic_token); } @@ -266,35 +264,31 @@ gss_get_mic(struct gss_ctx *context_handle, u32 gss_verify_mic(struct gss_ctx *context_handle, struct xdr_buf *message, - struct xdr_netobj *mic_token, - u32 *qstate) + struct xdr_netobj *mic_token) { return context_handle->mech_type->gm_ops ->gss_verify_mic(context_handle, message, - mic_token, - qstate); + mic_token); } u32 gss_wrap(struct gss_ctx *ctx_id, - u32 qop, int offset, struct xdr_buf *buf, struct page **inpages) { return ctx_id->mech_type->gm_ops - ->gss_wrap(ctx_id, qop, offset, buf, inpages); + ->gss_wrap(ctx_id, offset, buf, inpages); } u32 gss_unwrap(struct gss_ctx *ctx_id, - u32 *qop, int offset, struct xdr_buf *buf) { return ctx_id->mech_type->gm_ops - ->gss_unwrap(ctx_id, qop, offset, buf); + ->gss_unwrap(ctx_id, offset, buf); } diff --git a/net/sunrpc/auth_gss/gss_spkm3_mech.c b/net/sunrpc/auth_gss/gss_spkm3_mech.c index 6c97d61baa9b..39b3edc14694 100644 --- a/net/sunrpc/auth_gss/gss_spkm3_mech.c +++ b/net/sunrpc/auth_gss/gss_spkm3_mech.c @@ -224,18 +224,13 @@ gss_delete_sec_context_spkm3(void *internal_ctx) { static u32 gss_verify_mic_spkm3(struct gss_ctx *ctx, struct xdr_buf *signbuf, - struct xdr_netobj *checksum, - u32 *qstate) { + struct xdr_netobj *checksum) +{ u32 maj_stat = 0; - int qop_state = 0; struct spkm3_ctx *sctx = ctx->internal_ctx_id; dprintk("RPC: gss_verify_mic_spkm3 calling spkm3_read_token\n"); - maj_stat = spkm3_read_token(sctx, checksum, signbuf, &qop_state, - SPKM_MIC_TOK); - - if (!maj_stat && qop_state) - *qstate = qop_state; + maj_stat = spkm3_read_token(sctx, checksum, signbuf, SPKM_MIC_TOK); dprintk("RPC: gss_verify_mic_spkm3 returning %d\n", maj_stat); return maj_stat; @@ -243,15 +238,15 @@ gss_verify_mic_spkm3(struct gss_ctx *ctx, static u32 gss_get_mic_spkm3(struct gss_ctx *ctx, - u32 qop, struct xdr_buf *message_buffer, - struct xdr_netobj *message_token) { + struct xdr_netobj *message_token) +{ u32 err = 0; struct spkm3_ctx *sctx = ctx->internal_ctx_id; dprintk("RPC: gss_get_mic_spkm3\n"); - err = spkm3_make_token(sctx, qop, message_buffer, + err = spkm3_make_token(sctx, message_buffer, message_token, SPKM_MIC_TOK); return err; } @@ -264,8 +259,8 @@ static struct gss_api_ops gss_spkm3_ops = { }; static struct pf_desc gss_spkm3_pfs[] = { - {RPC_AUTH_GSS_SPKM, 0, RPC_GSS_SVC_NONE, "spkm3"}, - {RPC_AUTH_GSS_SPKMI, 0, RPC_GSS_SVC_INTEGRITY, "spkm3i"}, + {RPC_AUTH_GSS_SPKM, RPC_GSS_SVC_NONE, "spkm3"}, + {RPC_AUTH_GSS_SPKMI, RPC_GSS_SVC_INTEGRITY, "spkm3i"}, }; static struct gss_api_mech gss_spkm3_mech = { diff --git a/net/sunrpc/auth_gss/gss_spkm3_seal.c b/net/sunrpc/auth_gss/gss_spkm3_seal.c index 25339868d462..148201e929d0 100644 --- a/net/sunrpc/auth_gss/gss_spkm3_seal.c +++ b/net/sunrpc/auth_gss/gss_spkm3_seal.c @@ -51,7 +51,7 @@ */ u32 -spkm3_make_token(struct spkm3_ctx *ctx, int qop_req, +spkm3_make_token(struct spkm3_ctx *ctx, struct xdr_buf * text, struct xdr_netobj * token, int toktype) { @@ -68,8 +68,6 @@ spkm3_make_token(struct spkm3_ctx *ctx, int qop_req, dprintk("RPC: spkm3_make_token\n"); now = jiffies; - if (qop_req != 0) - goto out_err; if (ctx->ctx_id.len != 16) { dprintk("RPC: spkm3_make_token BAD ctx_id.len %d\n", diff --git a/net/sunrpc/auth_gss/gss_spkm3_unseal.c b/net/sunrpc/auth_gss/gss_spkm3_unseal.c index 65ce81bf0bc4..c3c0d9586103 100644 --- a/net/sunrpc/auth_gss/gss_spkm3_unseal.c +++ b/net/sunrpc/auth_gss/gss_spkm3_unseal.c @@ -52,7 +52,7 @@ u32 spkm3_read_token(struct spkm3_ctx *ctx, struct xdr_netobj *read_token, /* checksum */ struct xdr_buf *message_buffer, /* signbuf */ - int *qop_state, int toktype) + int toktype) { s32 code; struct xdr_netobj wire_cksum = {.len =0, .data = NULL}; diff --git a/net/sunrpc/auth_gss/svcauth_gss.c b/net/sunrpc/auth_gss/svcauth_gss.c index e3308195374e..e4ada15ed856 100644 --- a/net/sunrpc/auth_gss/svcauth_gss.c +++ b/net/sunrpc/auth_gss/svcauth_gss.c @@ -566,8 +566,7 @@ gss_verify_header(struct svc_rqst *rqstp, struct rsc *rsci, if (rqstp->rq_deferred) /* skip verification of revisited request */ return SVC_OK; - if (gss_verify_mic(ctx_id, &rpchdr, &checksum, NULL) - != GSS_S_COMPLETE) { + if (gss_verify_mic(ctx_id, &rpchdr, &checksum) != GSS_S_COMPLETE) { *authp = rpcsec_gsserr_credproblem; return SVC_DENIED; } @@ -604,7 +603,7 @@ gss_write_verf(struct svc_rqst *rqstp, struct gss_ctx *ctx_id, u32 seq) xdr_buf_from_iov(&iov, &verf_data); p = rqstp->rq_res.head->iov_base + rqstp->rq_res.head->iov_len; mic.data = (u8 *)(p + 1); - maj_stat = gss_get_mic(ctx_id, 0, &verf_data, &mic); + maj_stat = gss_get_mic(ctx_id, &verf_data, &mic); if (maj_stat != GSS_S_COMPLETE) return -1; *p++ = htonl(mic.len); @@ -710,7 +709,7 @@ unwrap_integ_data(struct xdr_buf *buf, u32 seq, struct gss_ctx *ctx) goto out; if (read_bytes_from_xdr_buf(buf, integ_len + 4, mic.data, mic.len)) goto out; - maj_stat = gss_verify_mic(ctx, &integ_buf, &mic, NULL); + maj_stat = gss_verify_mic(ctx, &integ_buf, &mic); if (maj_stat != GSS_S_COMPLETE) goto out; if (ntohl(svc_getu32(&buf->head[0])) != seq) @@ -1012,7 +1011,7 @@ svcauth_gss_release(struct svc_rqst *rqstp) resv = &resbuf->tail[0]; } mic.data = (u8 *)resv->iov_base + resv->iov_len + 4; - if (gss_get_mic(gsd->rsci->mechctx, 0, &integ_buf, &mic)) + if (gss_get_mic(gsd->rsci->mechctx, &integ_buf, &mic)) goto out_err; svc_putu32(resv, htonl(mic.len)); memset(mic.data + mic.len, 0, -- cgit v1.2.3 From a0857d03b21fa54653c9d2fe7a315381176015b4 Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Thu, 13 Oct 2005 16:55:23 -0400 Subject: RPCSEC_GSS: krb5 cleanup Remove some senseless wrappers. Signed-off-by: J. Bruce Fields Signed-off-by: Trond Myklebust --- include/linux/sunrpc/gss_krb5.h | 12 ++++-------- net/sunrpc/auth_gss/gss_krb5_mech.c | 28 ---------------------------- net/sunrpc/auth_gss/gss_krb5_seal.c | 5 +++-- net/sunrpc/auth_gss/gss_krb5_unseal.c | 6 +++--- 4 files changed, 10 insertions(+), 41 deletions(-) (limited to 'include/linux') diff --git a/include/linux/sunrpc/gss_krb5.h b/include/linux/sunrpc/gss_krb5.h index a7bda4edb853..2c3601d31045 100644 --- a/include/linux/sunrpc/gss_krb5.h +++ b/include/linux/sunrpc/gss_krb5.h @@ -118,15 +118,11 @@ s32 make_checksum(s32 cksumtype, char *header, int hdrlen, struct xdr_buf *body, int body_offset, struct xdr_netobj *cksum); -u32 -krb5_make_token(struct krb5_ctx *context_handle, - struct xdr_buf *input_message_buffer, - struct xdr_netobj *output_message_buffer); +u32 gss_get_mic_kerberos(struct gss_ctx *, struct xdr_buf *, + struct xdr_netobj *); -u32 -krb5_read_token(struct krb5_ctx *context_handle, - struct xdr_netobj *input_token_buffer, - struct xdr_buf *message_buffer); +u32 gss_verify_mic_kerberos(struct gss_ctx *, struct xdr_buf *, + struct xdr_netobj *); u32 gss_wrap_kerberos(struct gss_ctx *ctx_id, int offset, diff --git a/net/sunrpc/auth_gss/gss_krb5_mech.c b/net/sunrpc/auth_gss/gss_krb5_mech.c index 9ffac2c50b94..5f1f806a0b11 100644 --- a/net/sunrpc/auth_gss/gss_krb5_mech.c +++ b/net/sunrpc/auth_gss/gss_krb5_mech.c @@ -190,34 +190,6 @@ gss_delete_sec_context_kerberos(void *internal_ctx) { kfree(kctx); } -static u32 -gss_verify_mic_kerberos(struct gss_ctx *ctx, - struct xdr_buf *message, - struct xdr_netobj *mic_token) -{ - u32 maj_stat = 0; - struct krb5_ctx *kctx = ctx->internal_ctx_id; - - maj_stat = krb5_read_token(kctx, mic_token, message); - - dprintk("RPC: gss_verify_mic_kerberos returning %d\n", maj_stat); - return maj_stat; -} - -static u32 -gss_get_mic_kerberos(struct gss_ctx *ctx, - struct xdr_buf *message, - struct xdr_netobj *mic_token) { - u32 err = 0; - struct krb5_ctx *kctx = ctx->internal_ctx_id; - - err = krb5_make_token(kctx, message, mic_token); - - dprintk("RPC: gss_get_mic_kerberos returning %d\n",err); - - return err; -} - static struct gss_api_ops gss_kerberos_ops = { .gss_import_sec_context = gss_import_sec_context_kerberos, .gss_get_mic = gss_get_mic_kerberos, diff --git a/net/sunrpc/auth_gss/gss_krb5_seal.c b/net/sunrpc/auth_gss/gss_krb5_seal.c index 15227c727c8b..13f8ae979454 100644 --- a/net/sunrpc/auth_gss/gss_krb5_seal.c +++ b/net/sunrpc/auth_gss/gss_krb5_seal.c @@ -71,9 +71,10 @@ #endif u32 -krb5_make_token(struct krb5_ctx *ctx, - struct xdr_buf *text, struct xdr_netobj *token) +gss_get_mic_kerberos(struct gss_ctx *gss_ctx, struct xdr_buf *text, + struct xdr_netobj *token) { + struct krb5_ctx *ctx = gss_ctx->internal_ctx_id; s32 checksum_type; struct xdr_netobj md5cksum = {.len = 0, .data = NULL}; unsigned char *ptr, *krb5_hdr, *msg_start; diff --git a/net/sunrpc/auth_gss/gss_krb5_unseal.c b/net/sunrpc/auth_gss/gss_krb5_unseal.c index bcf978627a71..2030475d98ed 100644 --- a/net/sunrpc/auth_gss/gss_krb5_unseal.c +++ b/net/sunrpc/auth_gss/gss_krb5_unseal.c @@ -72,10 +72,10 @@ * supposedly taken over. */ u32 -krb5_read_token(struct krb5_ctx *ctx, - struct xdr_netobj *read_token, - struct xdr_buf *message_buffer) +gss_verify_mic_kerberos(struct gss_ctx *gss_ctx, + struct xdr_buf *message_buffer, struct xdr_netobj *read_token) { + struct krb5_ctx *ctx = gss_ctx->internal_ctx_id; int signalg; int sealalg; s32 checksum_type; -- cgit v1.2.3 From 11e29e21514517f3022a1f30998ac4c7b1197658 Mon Sep 17 00:00:00 2001 From: Alan Cox Date: Fri, 21 Oct 2005 18:46:32 -0400 Subject: libata: handle early device PIO modes correctly --- drivers/scsi/libata-core.c | 31 +++++++++++++++++++++++++++---- include/linux/ata.h | 13 +++++++++---- 2 files changed, 36 insertions(+), 8 deletions(-) (limited to 'include/linux') diff --git a/drivers/scsi/libata-core.c b/drivers/scsi/libata-core.c index 175d4646333d..09639e7aaa71 100644 --- a/drivers/scsi/libata-core.c +++ b/drivers/scsi/libata-core.c @@ -1082,6 +1082,31 @@ static inline void ata_dump_id(struct ata_device *dev) dev->id[93]); } +/* + * Compute the PIO modes available for this device. This is not as + * trivial as it seems if we must consider early devices correctly. + * + * FIXME: pre IDE drive timing (do we care ?). + */ + +static unsigned int ata_pio_modes(struct ata_device *adev) +{ + u16 modes; + + /* Usual case. Word 53 indicates word 88 is valid */ + if (adev->id[ATA_ID_FIELD_VALID] & (1 << 2)) { + modes = adev->id[ATA_ID_PIO_MODES] & 0x03; + modes <<= 3; + modes |= 0x7; + return modes; + } + + /* If word 88 isn't valid then Word 51 holds the PIO timing number + for the maximum. Turn it into a mask and return it */ + modes = (2 << (adev->id[ATA_ID_OLD_PIO_MODES] & 0xFF)) - 1 ; + return modes; +} + /** * ata_dev_identify - obtain IDENTIFY x DEVICE page * @ap: port on which device we wish to probe resides @@ -1215,10 +1240,8 @@ retry: xfer_modes = dev->id[ATA_ID_UDMA_MODES]; if (!xfer_modes) xfer_modes = (dev->id[ATA_ID_MWDMA_MODES]) << ATA_SHIFT_MWDMA; - if (!xfer_modes) { - xfer_modes = (dev->id[ATA_ID_PIO_MODES]) << (ATA_SHIFT_PIO + 3); - xfer_modes |= (0x7 << ATA_SHIFT_PIO); - } + if (!xfer_modes) + xfer_modes = ata_pio_modes(dev); ata_dump_id(dev); diff --git a/include/linux/ata.h b/include/linux/ata.h index 630908c9378b..33276d1d05d2 100644 --- a/include/linux/ata.h +++ b/include/linux/ata.h @@ -42,13 +42,18 @@ enum { ATA_SECT_SIZE = 512, ATA_ID_WORDS = 256, - ATA_ID_PROD_OFS = 27, - ATA_ID_FW_REV_OFS = 23, ATA_ID_SERNO_OFS = 10, - ATA_ID_MAJOR_VER = 80, - ATA_ID_PIO_MODES = 64, + ATA_ID_FW_REV_OFS = 23, + ATA_ID_PROD_OFS = 27, + ATA_ID_OLD_PIO_MODES = 51, + ATA_ID_FIELD_VALID = 53, ATA_ID_MWDMA_MODES = 63, + ATA_ID_PIO_MODES = 64, + ATA_ID_EIDE_DMA_MIN = 65, + ATA_ID_EIDE_PIO = 67, + ATA_ID_EIDE_PIO_IORDY = 68, ATA_ID_UDMA_MODES = 88, + ATA_ID_MAJOR_VER = 80, ATA_ID_PIO4 = (1 << 1), ATA_PCI_CTL_OFS = 2, -- cgit v1.2.3 From 452503f993feffe96e8cc9fbff4888b96e2c5e40 Mon Sep 17 00:00:00 2001 From: Alan Cox Date: Fri, 21 Oct 2005 19:01:32 -0400 Subject: Add ide-timing functionality to libata. This is needed for full AMD and VIA drivers and possibly more. Functions to turn actual clocking and cycle timings into register values. Also to merge shared timings to compute an optimal timing set. Built from the drivers/ide version by Vojtech Pavlik Signed-off-by: Alan Cox Signed-off-by: Jeff Garzik --- drivers/scsi/libata-core.c | 149 +++++++++++++++++++++++++++++++++++++++++++++ include/linux/libata.h | 39 ++++++++++++ 2 files changed, 188 insertions(+) (limited to 'include/linux') diff --git a/drivers/scsi/libata-core.c b/drivers/scsi/libata-core.c index 09639e7aaa71..9269fd9b814f 100644 --- a/drivers/scsi/libata-core.c +++ b/drivers/scsi/libata-core.c @@ -1538,6 +1538,152 @@ void ata_port_disable(struct ata_port *ap) ap->flags |= ATA_FLAG_PORT_DISABLED; } +/* + * This mode timing computation functionality is ported over from + * drivers/ide/ide-timing.h and was originally written by Vojtech Pavlik + */ +/* + * PIO 0-5, MWDMA 0-2 and UDMA 0-6 timings (in nanoseconds). + * These were taken from ATA/ATAPI-6 standard, rev 0a, except + * for PIO 5, which is a nonstandard extension and UDMA6, which + * is currently supported only by Maxtor drives. + */ + +static const struct ata_timing ata_timing[] = { + + { XFER_UDMA_6, 0, 0, 0, 0, 0, 0, 0, 15 }, + { XFER_UDMA_5, 0, 0, 0, 0, 0, 0, 0, 20 }, + { XFER_UDMA_4, 0, 0, 0, 0, 0, 0, 0, 30 }, + { XFER_UDMA_3, 0, 0, 0, 0, 0, 0, 0, 45 }, + + { XFER_UDMA_2, 0, 0, 0, 0, 0, 0, 0, 60 }, + { XFER_UDMA_1, 0, 0, 0, 0, 0, 0, 0, 80 }, + { XFER_UDMA_0, 0, 0, 0, 0, 0, 0, 0, 120 }, + +/* { XFER_UDMA_SLOW, 0, 0, 0, 0, 0, 0, 0, 150 }, */ + + { XFER_MW_DMA_2, 25, 0, 0, 0, 70, 25, 120, 0 }, + { XFER_MW_DMA_1, 45, 0, 0, 0, 80, 50, 150, 0 }, + { XFER_MW_DMA_0, 60, 0, 0, 0, 215, 215, 480, 0 }, + + { XFER_SW_DMA_2, 60, 0, 0, 0, 120, 120, 240, 0 }, + { XFER_SW_DMA_1, 90, 0, 0, 0, 240, 240, 480, 0 }, + { XFER_SW_DMA_0, 120, 0, 0, 0, 480, 480, 960, 0 }, + +/* { XFER_PIO_5, 20, 50, 30, 100, 50, 30, 100, 0 }, */ + { XFER_PIO_4, 25, 70, 25, 120, 70, 25, 120, 0 }, + { XFER_PIO_3, 30, 80, 70, 180, 80, 70, 180, 0 }, + + { XFER_PIO_2, 30, 290, 40, 330, 100, 90, 240, 0 }, + { XFER_PIO_1, 50, 290, 93, 383, 125, 100, 383, 0 }, + { XFER_PIO_0, 70, 290, 240, 600, 165, 150, 600, 0 }, + +/* { XFER_PIO_SLOW, 120, 290, 240, 960, 290, 240, 960, 0 }, */ + + { 0xFF } +}; + +#define ENOUGH(v,unit) (((v)-1)/(unit)+1) +#define EZ(v,unit) ((v)?ENOUGH(v,unit):0) + +static void ata_timing_quantize(const struct ata_timing *t, struct ata_timing *q, int T, int UT) +{ + q->setup = EZ(t->setup * 1000, T); + q->act8b = EZ(t->act8b * 1000, T); + q->rec8b = EZ(t->rec8b * 1000, T); + q->cyc8b = EZ(t->cyc8b * 1000, T); + q->active = EZ(t->active * 1000, T); + q->recover = EZ(t->recover * 1000, T); + q->cycle = EZ(t->cycle * 1000, T); + q->udma = EZ(t->udma * 1000, UT); +} + +void ata_timing_merge(const struct ata_timing *a, const struct ata_timing *b, + struct ata_timing *m, unsigned int what) +{ + if (what & ATA_TIMING_SETUP ) m->setup = max(a->setup, b->setup); + if (what & ATA_TIMING_ACT8B ) m->act8b = max(a->act8b, b->act8b); + if (what & ATA_TIMING_REC8B ) m->rec8b = max(a->rec8b, b->rec8b); + if (what & ATA_TIMING_CYC8B ) m->cyc8b = max(a->cyc8b, b->cyc8b); + if (what & ATA_TIMING_ACTIVE ) m->active = max(a->active, b->active); + if (what & ATA_TIMING_RECOVER) m->recover = max(a->recover, b->recover); + if (what & ATA_TIMING_CYCLE ) m->cycle = max(a->cycle, b->cycle); + if (what & ATA_TIMING_UDMA ) m->udma = max(a->udma, b->udma); +} + +static const struct ata_timing* ata_timing_find_mode(unsigned short speed) +{ + const struct ata_timing *t; + + for (t = ata_timing; t->mode != speed; t++) + if (t->mode != 0xFF) + return NULL; + return t; +} + +int ata_timing_compute(struct ata_device *adev, unsigned short speed, + struct ata_timing *t, int T, int UT) +{ + const struct ata_timing *s; + struct ata_timing p; + + /* + * Find the mode. + */ + + if (!(s = ata_timing_find_mode(speed))) + return -EINVAL; + + /* + * If the drive is an EIDE drive, it can tell us it needs extended + * PIO/MW_DMA cycle timing. + */ + + if (adev->id[ATA_ID_FIELD_VALID] & 2) { /* EIDE drive */ + memset(&p, 0, sizeof(p)); + if(speed >= XFER_PIO_0 && speed <= XFER_SW_DMA_0) { + if (speed <= XFER_PIO_2) p.cycle = p.cyc8b = adev->id[ATA_ID_EIDE_PIO]; + else p.cycle = p.cyc8b = adev->id[ATA_ID_EIDE_PIO_IORDY]; + } else if(speed >= XFER_MW_DMA_0 && speed <= XFER_MW_DMA_2) { + p.cycle = adev->id[ATA_ID_EIDE_DMA_MIN]; + } + ata_timing_merge(&p, t, t, ATA_TIMING_CYCLE | ATA_TIMING_CYC8B); + } + + /* + * Convert the timing to bus clock counts. + */ + + ata_timing_quantize(s, t, T, UT); + + /* + * Even in DMA/UDMA modes we still use PIO access for IDENTIFY, S.M.A.R.T + * and some other commands. We have to ensure that the DMA cycle timing is + * slower/equal than the fastest PIO timing. + */ + + if (speed > XFER_PIO_4) { + ata_timing_compute(adev, adev->pio_mode, &p, T, UT); + ata_timing_merge(&p, t, t, ATA_TIMING_ALL); + } + + /* + * Lenghten active & recovery time so that cycle time is correct. + */ + + if (t->act8b + t->rec8b < t->cyc8b) { + t->act8b += (t->cyc8b - (t->act8b + t->rec8b)) / 2; + t->rec8b = t->cyc8b - t->act8b; + } + + if (t->active + t->recover < t->cycle) { + t->active += (t->cycle - (t->active + t->recover)) / 2; + t->recover = t->cycle - t->active; + } + + return 0; +} + static struct { unsigned int shift; u8 base; @@ -4764,6 +4910,9 @@ EXPORT_SYMBOL_GPL(ata_dev_id_string); EXPORT_SYMBOL_GPL(ata_dev_config); EXPORT_SYMBOL_GPL(ata_scsi_simulate); +EXPORT_SYMBOL_GPL(ata_timing_compute); +EXPORT_SYMBOL_GPL(ata_timing_merge); + #ifdef CONFIG_PCI EXPORT_SYMBOL_GPL(pci_test_config_bits); EXPORT_SYMBOL_GPL(ata_pci_host_stop); diff --git a/include/linux/libata.h b/include/linux/libata.h index 0261c55f3483..0e214f8c8f9f 100644 --- a/include/linux/libata.h +++ b/include/linux/libata.h @@ -388,6 +388,19 @@ struct ata_port_info { struct ata_port_operations *port_ops; }; +struct ata_timing { + unsigned short mode; /* ATA mode */ + unsigned short setup; /* t1 */ + unsigned short act8b; /* t2 for 8-bit I/O */ + unsigned short rec8b; /* t2i for 8-bit I/O */ + unsigned short cyc8b; /* t0 for 8-bit I/O */ + unsigned short active; /* t2 or tD */ + unsigned short recover; /* t2i or tK */ + unsigned short cycle; /* t0 */ + unsigned short udma; /* t2CYCTYP/2 */ +}; + +#define FIT(v,vmin,vmax) max_t(short,min_t(short,v,vmax),vmin) extern void ata_port_probe(struct ata_port *); extern void __sata_phy_reset(struct ata_port *ap); @@ -451,6 +464,32 @@ extern int ata_std_bios_param(struct scsi_device *sdev, sector_t capacity, int geom[]); extern int ata_scsi_slave_config(struct scsi_device *sdev); +/* + * Timing helpers + */ +extern int ata_timing_compute(struct ata_device *, unsigned short, + struct ata_timing *, int, int); +extern void ata_timing_merge(const struct ata_timing *, + const struct ata_timing *, struct ata_timing *, + unsigned int); + +enum { + ATA_TIMING_SETUP = (1 << 0), + ATA_TIMING_ACT8B = (1 << 1), + ATA_TIMING_REC8B = (1 << 2), + ATA_TIMING_CYC8B = (1 << 3), + ATA_TIMING_8BIT = ATA_TIMING_ACT8B | ATA_TIMING_REC8B | + ATA_TIMING_CYC8B, + ATA_TIMING_ACTIVE = (1 << 4), + ATA_TIMING_RECOVER = (1 << 5), + ATA_TIMING_CYCLE = (1 << 6), + ATA_TIMING_UDMA = (1 << 7), + ATA_TIMING_ALL = ATA_TIMING_SETUP | ATA_TIMING_ACT8B | + ATA_TIMING_REC8B | ATA_TIMING_CYC8B | + ATA_TIMING_ACTIVE | ATA_TIMING_RECOVER | + ATA_TIMING_CYCLE | ATA_TIMING_UDMA, +}; + #ifdef CONFIG_PCI struct pci_bits { -- cgit v1.2.3 From 6ab0f5cd364476fe5cb329fd46ee41bea6d4c69c Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Fri, 21 Oct 2005 22:58:51 -0400 Subject: [PARISC] Update parisc specific input code from parisc tree Update drivers to new input layer changes. Signed-off-by: Helge Deller Signed-off-by: Matthew Wilcox Reorder code in gscps2_interrupt() and only enable ports when opened. This fixes issues with hangs booting an SMP kernel on my C360. Previously serio_interrupt() could be called before the lock in struct serio was initialised. Signed-off-by: Richard Hirst Signed-off-by: Kyle McMartin --- drivers/input/keyboard/hil_kbd.c | 28 ++- drivers/input/keyboard/hilkbd.c | 2 +- drivers/input/mouse/hil_ptr.c | 33 ++- drivers/input/serio/gscps2.c | 13 +- drivers/input/serio/hil_mlc.c | 14 +- include/linux/hil.h | 483 +++++++++++++++++++++++++++++++++++++++ include/linux/hil_mlc.h | 168 ++++++++++++++ include/linux/hp_sdc.h | 300 ++++++++++++++++++++++++ include/linux/input.h | 1 + 9 files changed, 1011 insertions(+), 31 deletions(-) create mode 100644 include/linux/hil.h create mode 100644 include/linux/hil_mlc.h create mode 100644 include/linux/hp_sdc.h (limited to 'include/linux') diff --git a/drivers/input/keyboard/hil_kbd.c b/drivers/input/keyboard/hil_kbd.c index ef78bffed5e7..0a90962c38e7 100644 --- a/drivers/input/keyboard/hil_kbd.c +++ b/drivers/input/keyboard/hil_kbd.c @@ -204,7 +204,7 @@ static irqreturn_t hil_kbd_interrupt(struct serio *serio, hil_packet packet; int idx; - kbd = (struct hil_kbd *)serio->private; + kbd = serio_get_drvdata(serio); if (kbd == NULL) { BUG(); return IRQ_HANDLED; @@ -234,7 +234,7 @@ static void hil_kbd_disconnect(struct serio *serio) { struct hil_kbd *kbd; - kbd = (struct hil_kbd *)serio->private; + kbd = serio_get_drvdata(serio); if (kbd == NULL) { BUG(); return; @@ -245,20 +245,20 @@ static void hil_kbd_disconnect(struct serio *serio) kfree(kbd); } -static void hil_kbd_connect(struct serio *serio, struct serio_driver *drv) +static int hil_kbd_connect(struct serio *serio, struct serio_driver *drv) { struct hil_kbd *kbd; uint8_t did, *idd; int i; - if (serio->type != (SERIO_HIL_MLC | SERIO_HIL)) return; - - if (!(kbd = kmalloc(sizeof(struct hil_kbd), GFP_KERNEL))) return; + kbd = kmalloc(sizeof(*kbd), GFP_KERNEL); + if (!kbd) + return -ENOMEM; memset(kbd, 0, sizeof(struct hil_kbd)); if (serio_open(serio, drv)) goto bail0; - serio->private = kbd; + serio_set_drvdata(serio, kbd); kbd->serio = serio; kbd->dev.private = kbd; @@ -342,19 +342,31 @@ static void hil_kbd_connect(struct serio *serio, struct serio_driver *drv) down(&(kbd->sem)); up(&(kbd->sem)); - return; + return 0; bail1: serio_close(serio); bail0: kfree(kbd); + serio_set_drvdata(serio, NULL); + return -EIO; } +static struct serio_device_id hil_kbd_ids[] = { + { + .type = SERIO_HIL_MLC, + .proto = SERIO_HIL, + .id = SERIO_ANY, + .extra = SERIO_ANY, + }, + { 0 } +}; struct serio_driver hil_kbd_serio_drv = { .driver = { .name = "hil_kbd", }, .description = "HP HIL keyboard driver", + .id_table = hil_kbd_ids, .connect = hil_kbd_connect, .disconnect = hil_kbd_disconnect, .interrupt = hil_kbd_interrupt diff --git a/drivers/input/keyboard/hilkbd.c b/drivers/input/keyboard/hilkbd.c index e7a1e14f8b0c..e95bc052e32a 100644 --- a/drivers/input/keyboard/hilkbd.c +++ b/drivers/input/keyboard/hilkbd.c @@ -22,7 +22,7 @@ #include #include #include -#include +#include #include #include diff --git a/drivers/input/mouse/hil_ptr.c b/drivers/input/mouse/hil_ptr.c index bc22849c6c79..c2bf2ed07dc6 100644 --- a/drivers/input/mouse/hil_ptr.c +++ b/drivers/input/mouse/hil_ptr.c @@ -196,7 +196,7 @@ static irqreturn_t hil_ptr_interrupt(struct serio *serio, hil_packet packet; int idx; - ptr = (struct hil_ptr *)serio->private; + ptr = serio_get_drvdata(serio); if (ptr == NULL) { BUG(); return IRQ_HANDLED; @@ -227,7 +227,7 @@ static void hil_ptr_disconnect(struct serio *serio) { struct hil_ptr *ptr; - ptr = (struct hil_ptr *)serio->private; + ptr = serio_get_drvdata(serio); if (ptr == NULL) { BUG(); return; @@ -238,21 +238,19 @@ static void hil_ptr_disconnect(struct serio *serio) kfree(ptr); } -static void hil_ptr_connect(struct serio *serio, struct serio_driver *driver) +static int hil_ptr_connect(struct serio *serio, struct serio_driver *driver) { struct hil_ptr *ptr; char *txt; unsigned int i, naxsets, btntype; uint8_t did, *idd; - if (serio->type != (SERIO_HIL_MLC | SERIO_HIL)) return; - - if (!(ptr = kmalloc(sizeof(struct hil_ptr), GFP_KERNEL))) return; + if (!(ptr = kmalloc(sizeof(struct hil_ptr), GFP_KERNEL))) return -ENOMEM; memset(ptr, 0, sizeof(struct hil_ptr)); if (serio_open(serio, driver)) goto bail0; - serio->private = ptr; + serio_set_drvdata(serio, ptr); ptr->serio = serio; ptr->dev.private = ptr; @@ -380,23 +378,34 @@ static void hil_ptr_connect(struct serio *serio, struct serio_driver *driver) (btntype == BTN_MOUSE) ? "HIL mouse":"HIL tablet or touchpad", did); - return; + return 0; bail1: serio_close(serio); bail0: kfree(ptr); - return; + serio_set_drvdata(serio, NULL); + return -ENODEV; } +static struct serio_device_id hil_ptr_ids[] = { + { + .type = SERIO_HIL_MLC, + .proto = SERIO_HIL, + .id = SERIO_ANY, + .extra = SERIO_ANY, + }, + { 0 } +}; static struct serio_driver hil_ptr_serio_driver = { .driver = { .name = "hil_ptr", }, .description = "HP HIL mouse/tablet driver", - .connect = hil_ptr_connect, - .disconnect = hil_ptr_disconnect, - .interrupt = hil_ptr_interrupt + .id_table = hil_ptr_ids, + .connect = hil_ptr_connect, + .disconnect = hil_ptr_disconnect, + .interrupt = hil_ptr_interrupt }; static int __init hil_ptr_init(void) diff --git a/drivers/input/serio/gscps2.c b/drivers/input/serio/gscps2.c index 9790d7169a53..a7b0de0f92b2 100644 --- a/drivers/input/serio/gscps2.c +++ b/drivers/input/serio/gscps2.c @@ -211,9 +211,6 @@ static void gscps2_reset(struct gscps2port *ps2port) writeb(0xff, addr+GSC_RESET); gscps2_flush(ps2port); spin_unlock_irqrestore(&ps2port->lock, flags); - - /* enable it */ - gscps2_enable(ps2port, ENABLE); } static LIST_HEAD(ps2port_list); @@ -307,6 +304,9 @@ static int gscps2_open(struct serio *port) gscps2_reset(ps2port); + /* enable it */ + gscps2_enable(ps2port, ENABLE); + gscps2_interrupt(0, NULL, NULL); return 0; @@ -370,8 +370,6 @@ static int __init gscps2_probe(struct parisc_device *dev) serio->port_data = ps2port; serio->dev.parent = &dev->dev; - list_add_tail(&ps2port->node, &ps2port_list); - ret = -EBUSY; if (request_irq(dev->irq, gscps2_interrupt, SA_SHIRQ, ps2port->port->name, ps2port)) goto fail_miserably; @@ -396,15 +394,16 @@ static int __init gscps2_probe(struct parisc_device *dev) serio_register_port(ps2port->port); + list_add_tail(&ps2port->node, &ps2port_list); + return 0; fail: free_irq(dev->irq, ps2port); fail_miserably: - list_del(&ps2port->node); iounmap(ps2port->addr); - release_mem_region(dev->hpa, GSC_STATUS + 4); + release_mem_region(dev->hpa.start, GSC_STATUS + 4); fail_nomem: kfree(ps2port); diff --git a/drivers/input/serio/hil_mlc.c b/drivers/input/serio/hil_mlc.c index c243cb6fdfc4..5704204964a3 100644 --- a/drivers/input/serio/hil_mlc.c +++ b/drivers/input/serio/hil_mlc.c @@ -801,7 +801,8 @@ static int hil_mlc_serio_open(struct serio *serio) { struct hil_mlc_serio_map *map; struct hil_mlc *mlc; - if (serio->private != NULL) return -EBUSY; + if (serio_get_drvdata(serio) != NULL) + return -EBUSY; map = serio->port_data; if (map == NULL) { @@ -832,11 +833,18 @@ static void hil_mlc_serio_close(struct serio *serio) { return; } - serio->private = NULL; + serio_set_drvdata(serio, NULL); serio->drv = NULL; /* TODO wake up interruptable */ } +static struct serio_device_id hil_mlc_serio_id = { + .type = SERIO_HIL_MLC, + .proto = SERIO_HIL, + .extra = SERIO_ANY, + .id = SERIO_ANY, +}; + int hil_mlc_register(hil_mlc *mlc) { int i; unsigned long flags; @@ -867,7 +875,7 @@ int hil_mlc_register(hil_mlc *mlc) { mlc_serio = kmalloc(sizeof(*mlc_serio), GFP_KERNEL); mlc->serio[i] = mlc_serio; memset(mlc_serio, 0, sizeof(*mlc_serio)); - mlc_serio->type = SERIO_HIL | SERIO_HIL_MLC; + mlc_serio->id = hil_mlc_serio_id; mlc_serio->write = hil_mlc_serio_write; mlc_serio->open = hil_mlc_serio_open; mlc_serio->close = hil_mlc_serio_close; diff --git a/include/linux/hil.h b/include/linux/hil.h new file mode 100644 index 000000000000..13352d7d0caf --- /dev/null +++ b/include/linux/hil.h @@ -0,0 +1,483 @@ +#ifndef _HIL_H_ +#define _HIL_H_ + +/* + * Hewlett Packard Human Interface Loop (HP-HIL) Protocol -- header. + * + * Copyright (c) 2001 Brian S. Julin + * All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions, and the following disclaimer, + * without modification. + * 2. The name of the author may not be used to endorse or promote products + * derived from this software without specific prior written permission. + * + * Alternatively, this software may be distributed under the terms of the + * GNU General Public License ("GPL"). + * + * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND + * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE + * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE + * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE FOR + * ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL + * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS + * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) + * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT + * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY + * + * References: + * HP-HIL Technical Reference Manual. Hewlett Packard Product No. 45918A + * + * A note of thanks to HP for providing and shipping reference materials + * free of charge to help in the development of HIL support for Linux. + * + */ + +#include + +/* Physical constants relevant to raw loop/device timing. + */ + +#define HIL_CLOCK 8MHZ +#define HIL_EK1_CLOCK 30HZ +#define HIL_EK2_CLOCK 60HZ + +#define HIL_TIMEOUT_DEV 5 /* ms */ +#define HIL_TIMEOUT_DEVS 10 /* ms */ +#define HIL_TIMEOUT_NORESP 10 /* ms */ +#define HIL_TIMEOUT_DEVS_DATA 16 /* ms */ +#define HIL_TIMEOUT_SELFTEST 200 /* ms */ + + +/* Actual wire line coding. These will only be useful if someone is + * implementing a software MLC to run HIL devices on a non-parisc machine. + */ + +#define HIL_WIRE_PACKET_LEN 15 +enum hil_wire_bitpos { + HIL_WIRE_START = 0, + HIL_WIRE_ADDR2, + HIL_WIRE_ADDR1, + HIL_WIRE_ADDR0, + HIL_WIRE_COMMAND, + HIL_WIRE_DATA7, + HIL_WIRE_DATA6, + HIL_WIRE_DATA5, + HIL_WIRE_DATA4, + HIL_WIRE_DATA3, + HIL_WIRE_DATA2, + HIL_WIRE_DATA1, + HIL_WIRE_DATA0, + HIL_WIRE_PARITY, + HIL_WIRE_STOP +}; + +/* HP documentation uses these bit positions to refer to commands; + * we will call these "packets". + */ +enum hil_pkt_bitpos { + HIL_PKT_CMD = 0x00000800, + HIL_PKT_ADDR2 = 0x00000400, + HIL_PKT_ADDR1 = 0x00000200, + HIL_PKT_ADDR0 = 0x00000100, + HIL_PKT_ADDR_MASK = 0x00000700, + HIL_PKT_ADDR_SHIFT = 8, + HIL_PKT_DATA7 = 0x00000080, + HIL_PKT_DATA6 = 0x00000040, + HIL_PKT_DATA5 = 0x00000020, + HIL_PKT_DATA4 = 0x00000010, + HIL_PKT_DATA3 = 0x00000008, + HIL_PKT_DATA2 = 0x00000004, + HIL_PKT_DATA1 = 0x00000002, + HIL_PKT_DATA0 = 0x00000001, + HIL_PKT_DATA_MASK = 0x000000FF, + HIL_PKT_DATA_SHIFT = 0 +}; + +/* The HIL MLC also has several error/status/control bits. We extend the + * "packet" to include these when direct access to the MLC is available, + * or emulate them in cases where they are not available. + * + * This way the device driver knows that the underlying MLC driver + * has had to deal with loop errors. + */ +enum hil_error_bitpos { + HIL_ERR_OB = 0x00000800, /* MLC is busy sending an auto-poll, + or we have filled up the output + buffer and must wait. */ + HIL_ERR_INT = 0x00010000, /* A normal interrupt has occurred. */ + HIL_ERR_NMI = 0x00020000, /* An NMI has occurred. */ + HIL_ERR_LERR = 0x00040000, /* A poll didn't come back. */ + HIL_ERR_PERR = 0x01000000, /* There was a Parity Error. */ + HIL_ERR_FERR = 0x02000000, /* There was a Framing Error. */ + HIL_ERR_FOF = 0x04000000 /* Input FIFO Overflowed. */ +}; + +enum hil_control_bitpos { + HIL_CTRL_TEST = 0x00010000, + HIL_CTRL_IPF = 0x00040000, + HIL_CTRL_APE = 0x02000000 +}; + +/* Bits 30,31 are unused, we use them to control write behavior. */ +#define HIL_DO_ALTER_CTRL 0x40000000 /* Write MSW of packet to control + before writing LSW to loop */ +#define HIL_CTRL_ONLY 0xc0000000 /* *Only* alter the control registers */ + +/* This gives us a 32-bit "packet" + */ +typedef u32 hil_packet; + + +/* HIL Loop commands + */ +enum hil_command { + HIL_CMD_IFC = 0x00, /* Interface Clear */ + HIL_CMD_EPT = 0x01, /* Enter Pass-Thru Mode */ + HIL_CMD_ELB = 0x02, /* Enter Loop-Back Mode */ + HIL_CMD_IDD = 0x03, /* Identify and Describe */ + HIL_CMD_DSR = 0x04, /* Device Soft Reset */ + HIL_CMD_PST = 0x05, /* Perform Self Test */ + HIL_CMD_RRG = 0x06, /* Read Register */ + HIL_CMD_WRG = 0x07, /* Write Register */ + HIL_CMD_ACF = 0x08, /* Auto Configure */ + HIL_CMDID_ACF = 0x07, /* Auto Configure bits with incremented ID */ + HIL_CMD_POL = 0x10, /* Poll */ + HIL_CMDCT_POL = 0x0f, /* Poll command bits with item count */ + HIL_CMD_RPL = 0x20, /* RePoll */ + HIL_CMDCT_RPL = 0x0f, /* RePoll command bits with item count */ + HIL_CMD_RNM = 0x30, /* Report Name */ + HIL_CMD_RST = 0x31, /* Report Status */ + HIL_CMD_EXD = 0x32, /* Extended Describe */ + HIL_CMD_RSC = 0x33, /* Report Security Code */ + + /* 0x34 to 0x3c reserved for future use */ + + HIL_CMD_DKA = 0x3d, /* Disable Keyswitch Autorepeat */ + HIL_CMD_EK1 = 0x3e, /* Enable Keyswitch Autorepeat 1 */ + HIL_CMD_EK2 = 0x3f, /* Enable Keyswitch Autorepeat 2 */ + HIL_CMD_PR1 = 0x40, /* Prompt1 */ + HIL_CMD_PR2 = 0x41, /* Prompt2 */ + HIL_CMD_PR3 = 0x42, /* Prompt3 */ + HIL_CMD_PR4 = 0x43, /* Prompt4 */ + HIL_CMD_PR5 = 0x44, /* Prompt5 */ + HIL_CMD_PR6 = 0x45, /* Prompt6 */ + HIL_CMD_PR7 = 0x46, /* Prompt7 */ + HIL_CMD_PRM = 0x47, /* Prompt (General Purpose) */ + HIL_CMD_AK1 = 0x48, /* Acknowlege1 */ + HIL_CMD_AK2 = 0x49, /* Acknowlege2 */ + HIL_CMD_AK3 = 0x4a, /* Acknowlege3 */ + HIL_CMD_AK4 = 0x4b, /* Acknowlege4 */ + HIL_CMD_AK5 = 0x4c, /* Acknowlege5 */ + HIL_CMD_AK6 = 0x4d, /* Acknowlege6 */ + HIL_CMD_AK7 = 0x4e, /* Acknowlege7 */ + HIL_CMD_ACK = 0x4f, /* Acknowlege (General Purpose) */ + + /* 0x50 to 0x78 reserved for future use */ + /* 0x80 to 0xEF device-specific commands */ + /* 0xf0 to 0xf9 reserved for future use */ + + HIL_CMD_RIO = 0xfa, /* Register I/O Error */ + HIL_CMD_SHR = 0xfb, /* System Hard Reset */ + HIL_CMD_TER = 0xfc, /* Transmission Error */ + HIL_CMD_CAE = 0xfd, /* Configuration Address Error */ + HIL_CMD_DHR = 0xfe, /* Device Hard Reset */ + + /* 0xff is prohibited from use. */ +}; + + +/* + * Response "records" to HIL commands + */ + +/* Device ID byte + */ +#define HIL_IDD_DID_TYPE_MASK 0xe0 /* Primary type bits */ +#define HIL_IDD_DID_TYPE_KB_INTEGRAL 0xa0 /* Integral keyboard */ +#define HIL_IDD_DID_TYPE_KB_ITF 0xc0 /* ITD keyboard */ +#define HIL_IDD_DID_TYPE_KB_RSVD 0xe0 /* Reserved keyboard type */ +#define HIL_IDD_DID_TYPE_KB_LANG_MASK 0x1f /* Keyboard locale bits */ +#define HIL_IDD_DID_KBLANG_USE_ESD 0x00 /* Use ESD Locale instead */ +#define HIL_IDD_DID_TYPE_ABS 0x80 /* Absolute Positioners */ +#define HIL_IDD_DID_ABS_RSVD1_MASK 0xf8 /* Reserved */ +#define HIL_IDD_DID_ABS_RSVD1 0x98 +#define HIL_IDD_DID_ABS_TABLET_MASK 0xf8 /* Tablets and digitizers */ +#define HIL_IDD_DID_ABS_TABLET 0x90 +#define HIL_IDD_DID_ABS_TSCREEN_MASK 0xfc /* Touch screens */ +#define HIL_IDD_DID_ABS_TSCREEN 0x8c +#define HIL_IDD_DID_ABS_RSVD2_MASK 0xfc /* Reserved */ +#define HIL_IDD_DID_ABS_RSVD2 0x88 +#define HIL_IDD_DID_ABS_RSVD3_MASK 0xfc /* Reserved */ +#define HIL_IDD_DID_ABS_RSVD3 0x80 +#define HIL_IDD_DID_TYPE_REL 0x60 /* Relative Positioners */ +#define HIL_IDD_DID_REL_RSVD1_MASK 0xf0 /* Reserved */ +#define HIL_IDD_DID_REL_RSVD1 0x70 +#define HIL_IDD_DID_REL_RSVD2_MASK 0xfc /* Reserved */ +#define HIL_IDD_DID_REL_RSVD2 0x6c +#define HIL_IDD_DID_REL_MOUSE_MASK 0xfc /* Mouse */ +#define HIL_IDD_DID_REL_MOUSE 0x68 +#define HIL_IDD_DID_REL_QUAD_MASK 0xf8 /* Other Quadrature Devices */ +#define HIL_IDD_DID_REL_QUAD 0x60 +#define HIL_IDD_DID_TYPE_CHAR 0x40 /* Character Entry */ +#define HIL_IDD_DID_CHAR_BARCODE_MASK 0xfc /* Barcode Reader */ +#define HIL_IDD_DID_CHAR_BARCODE 0x5c +#define HIL_IDD_DID_CHAR_RSVD1_MASK 0xfc /* Reserved */ +#define HIL_IDD_DID_CHAR_RSVD1 0x58 +#define HIL_IDD_DID_CHAR_RSVD2_MASK 0xf8 /* Reserved */ +#define HIL_IDD_DID_CHAR_RSVD2 0x50 +#define HIL_IDD_DID_CHAR_RSVD3_MASK 0xf0 /* Reserved */ +#define HIL_IDD_DID_CHAR_RSVD3 0x40 +#define HIL_IDD_DID_TYPE_OTHER 0x20 /* Miscellaneous */ +#define HIL_IDD_DID_OTHER_RSVD1_MASK 0xf0 /* Reserved */ +#define HIL_IDD_DID_OTHER_RSVD1 0x30 +#define HIL_IDD_DID_OTHER_BARCODE_MASK 0xfc /* Tone Generator */ +#define HIL_IDD_DID_OTHER_BARCODE 0x2c +#define HIL_IDD_DID_OTHER_RSVD2_MASK 0xfc /* Reserved */ +#define HIL_IDD_DID_OTHER_RSVD2 0x28 +#define HIL_IDD_DID_OTHER_RSVD3_MASK 0xf8 /* Reserved */ +#define HIL_IDD_DID_OTHER_RSVD3 0x20 +#define HIL_IDD_DID_TYPE_KEYPAD 0x00 /* Vectra Keyboard */ + +/* IDD record header + */ +#define HIL_IDD_HEADER_AXSET_MASK 0x03 /* Number of axis in a set */ +#define HIL_IDD_HEADER_RSC 0x04 /* Supports RSC command */ +#define HIL_IDD_HEADER_EXD 0x08 /* Supports EXD command */ +#define HIL_IDD_HEADER_IOD 0x10 /* IOD byte to follow */ +#define HIL_IDD_HEADER_16BIT 0x20 /* 16 (vs. 8) bit resolution */ +#define HIL_IDD_HEADER_ABS 0x40 /* Reports Absolute Position */ +#define HIL_IDD_HEADER_2X_AXIS 0x80 /* Two sets of 1-3 axis */ + +/* I/O Descriptor + */ +#define HIL_IDD_IOD_NBUTTON_MASK 0x07 /* Number of buttons */ +#define HIL_IDD_IOD_PROXIMITY 0x08 /* Proximity in/out events */ +#define HIL_IDD_IOD_PROMPT_MASK 0x70 /* Number of prompts/acks */ +#define HIL_IDD_IOD_PROMPT_SHIFT 4 +#define HIL_IDD_IOD_PROMPT 0x80 /* Generic prompt/ack */ + +#define HIL_IDD_NUM_AXES_PER_SET(header_packet) \ +((header_packet) & HIL_IDD_HEADER_AXSET_MASK) + +#define HIL_IDD_NUM_AXSETS(header_packet) \ +(2 - !((header_packet) & HIL_IDD_HEADER_2X_AXIS)) + +#define HIL_IDD_LEN(header_packet) \ +((4 - !(header_packet & HIL_IDD_HEADER_IOD) - \ + 2 * !(HIL_IDD_NUM_AXES_PER_SET(header_packet))) + \ + 2 * HIL_IDD_NUM_AXES_PER_SET(header_packet) * \ + !!((header_packet) & HIL_IDD_HEADER_ABS)) + +/* The following HIL_IDD_* macros assume you have an array of + * packets and/or unpacked 8-bit data in the order that they + * were received. + */ + +#define HIL_IDD_AXIS_COUNTS_PER_M(header_ptr) \ +(!(HIL_IDD_NUM_AXSETS(*(header_ptr))) ? -1 : \ +(((*(header_ptr + 1) & HIL_PKT_DATA_MASK) + \ + ((*(header_ptr + 2) & HIL_PKT_DATA_MASK)) << 8) \ +* ((*(header_ptr) & HIL_IDD_HEADER_16BIT) ? 100 : 1))) + +#define HIL_IDD_AXIS_MAX(header_ptr, __axnum) \ +((!(*(header_ptr) & HIL_IDD_HEADER_ABS) || \ + (HIL_IDD_NUM_AXES_PER_SET(*(header_ptr)) <= __axnum)) ? 0 : \ + ((HIL_PKT_DATA_MASK & *((header_ptr) + 3 + 2 * __axnum)) + \ + ((HIL_PKT_DATA_MASK & *((header_ptr) + 4 + 2 * __axnum)) << 8))) + +#define HIL_IDD_IOD(header_ptr) \ +(*(header_ptr + HIL_IDD_LEN((*header_ptr)) - 1)) + +#define HIL_IDD_HAS_GEN_PROMPT(header_ptr) \ +((*header_ptr & HIL_IDD_HEADER_IOD) && \ + (HIL_IDD_IOD(header_ptr) & HIL_IDD_IOD_PROMPT)) + +#define HIL_IDD_HAS_GEN_PROXIMITY(header_ptr) \ +((*header_ptr & HIL_IDD_HEADER_IOD) && \ + (HIL_IDD_IOD(header_ptr) & HIL_IDD_IOD_PROXIMITY)) + +#define HIL_IDD_NUM_BUTTONS(header_ptr) \ +((*header_ptr & HIL_IDD_HEADER_IOD) ? \ + (HIL_IDD_IOD(header_ptr) & HIL_IDD_IOD_NBUTTON_MASK) : 0) + +#define HIL_IDD_NUM_PROMPTS(header_ptr) \ +((*header_ptr & HIL_IDD_HEADER_IOD) ? \ + ((HIL_IDD_IOD(header_ptr) & HIL_IDD_IOD_NPROMPT_MASK) \ + >> HIL_IDD_IOD_PROMPT_SHIFT) : 0) + +/* The response to HIL EXD commands -- the "extended describe record" */ +#define HIL_EXD_HEADER_WRG 0x03 /* Supports type2 WRG */ +#define HIL_EXD_HEADER_WRG_TYPE1 0x01 /* Supports type1 WRG */ +#define HIL_EXD_HEADER_WRG_TYPE2 0x02 /* Supports type2 WRG */ +#define HIL_EXD_HEADER_RRG 0x04 /* Supports RRG command */ +#define HIL_EXD_HEADER_RNM 0x10 /* Supports RNM command */ +#define HIL_EXD_HEADER_RST 0x20 /* Supports RST command */ +#define HIL_EXD_HEADER_LOCALE 0x40 /* Contains locale code */ + +#define HIL_EXD_NUM_RRG(header_ptr) \ +((*header_ptr & HIL_EXD_HEADER_RRG) ? \ + (*(header_ptr + 1) & HIL_PKT_DATA_MASK) : 0) + +#define HIL_EXD_NUM_WWG(header_ptr) \ +((*header_ptr & HIL_EXD_HEADER_WRG) ? \ + (*(header_ptr + 2 - !(*header_ptr & HIL_EXD_HEADER_RRG)) & \ + HIL_PKT_DATA_MASK) : 0) + +#define HIL_EXD_LEN(header_ptr) \ +(!!(*header_ptr & HIL_EXD_HEADER_RRG) + \ + !!(*header_ptr & HIL_EXD_HEADER_WRG) + \ + !!(*header_ptr & HIL_EXD_HEADER_LOCALE) + \ + 2 * !!(*header_ptr & HIL_EXD_HEADER_WRG_TYPE2) + 1) + +#define HIL_EXD_LOCALE(header_ptr) \ +(!(*header_ptr & HIL_EXD_HEADER_LOCALE) ? -1 : \ + (*(header_ptr + HIL_EXD_LEN(header_ptr) - 1) & HIL_PKT_DATA_MASK)) + +#define HIL_EXD_WRG_TYPE2_LEN(header_ptr) \ +(!(*header_ptr & HIL_EXD_HEADER_WRG_TYPE2) ? -1 : \ + (*(header_ptr + HIL_EXD_LEN(header_ptr) - 2 - \ + !!(*header_ptr & HIL_EXD_HEADER_LOCALE)) & HIL_PKT_DATA_MASK) + \ + ((*(header_ptr + HIL_EXD_LEN(header_ptr) - 1 - \ + !!(*header_ptr & HIL_EXD_HEADER_LOCALE)) & HIL_PKT_DATA_MASK) << 8)) + +/* Device locale codes. */ + +/* Last defined locale code. Everything above this is "Reserved", + and note that this same table applies to the Device ID Byte where + keyboards may have a nationality code which is only 5 bits. */ +#define HIL_LOCALE_MAX 0x1f + +/* Map to hopefully useful strings. I was trying to make these look + like locale.aliases strings do; maybe that isn't the right table to + emulate. In either case, I didn't have much to work on. */ +#define HIL_LOCALE_MAP \ +"", /* 0x00 Reserved */ \ +"", /* 0x01 Reserved */ \ +"", /* 0x02 Reserved */ \ +"swiss.french", /* 0x03 Swiss/French */ \ +"portuguese", /* 0x04 Portuguese */ \ +"arabic", /* 0x05 Arabic */ \ +"hebrew", /* 0x06 Hebrew */ \ +"english.canadian", /* 0x07 Canadian English */ \ +"turkish", /* 0x08 Turkish */ \ +"greek", /* 0x09 Greek */ \ +"thai", /* 0x0a Thai (Thailand) */ \ +"italian", /* 0x0b Italian */ \ +"korean", /* 0x0c Hangul (Korea) */ \ +"dutch", /* 0x0d Dutch */ \ +"swedish", /* 0x0e Swedish */ \ +"german", /* 0x0f German */ \ +"chinese", /* 0x10 Chinese-PRC */ \ +"chinese", /* 0x11 Chinese-ROC */ \ +"swiss.french", /* 0x12 Swiss/French II */ \ +"spanish", /* 0x13 Spanish */ \ +"swiss.german", /* 0x14 Swiss/German II */ \ +"flemish", /* 0x15 Belgian (Flemish) */ \ +"finnish", /* 0x16 Finnish */ \ +"english.uk", /* 0x17 United Kingdom */ \ +"french.canadian", /* 0x18 French/Canadian */ \ +"swiss.german", /* 0x19 Swiss/German */ \ +"norwegian", /* 0x1a Norwegian */ \ +"french", /* 0x1b French */ \ +"danish", /* 0x1c Danish */ \ +"japanese", /* 0x1d Katakana */ \ +"spanish", /* 0x1e Latin American/Spanish*/\ +"english.us" /* 0x1f United States */ \ + + +/* HIL keycodes */ +#define HIL_KEYCODES_SET1_TBLSIZE 128 +#define HIL_KEYCODES_SET1 \ + KEY_5, KEY_RESERVED, KEY_RIGHTALT, KEY_LEFTALT, \ + KEY_RIGHTSHIFT, KEY_LEFTSHIFT, KEY_LEFTCTRL, KEY_SYSRQ, \ + KEY_KP4, KEY_KP8, KEY_KP5, KEY_KP9, \ + KEY_KP6, KEY_KP7, KEY_KPCOMMA, KEY_KPENTER, \ + KEY_KP1, KEY_KPSLASH, KEY_KP2, KEY_KPPLUS, \ + KEY_KP3, KEY_KPASTERISK, KEY_KP0, KEY_KPMINUS, \ + KEY_B, KEY_V, KEY_C, KEY_X, \ + KEY_Z, KEY_RESERVED, KEY_RESERVED, KEY_ESC, \ + KEY_6, KEY_F10, KEY_3, KEY_F11, \ + KEY_KPDOT, KEY_F9, KEY_TAB /*KP*/, KEY_F12, \ + KEY_H, KEY_G, KEY_F, KEY_D, \ + KEY_S, KEY_A, KEY_RESERVED, KEY_CAPSLOCK, \ + KEY_U, KEY_Y, KEY_T, KEY_R, \ + KEY_E, KEY_W, KEY_Q, KEY_TAB, \ + KEY_7, KEY_6, KEY_5, KEY_4, \ + KEY_3, KEY_2, KEY_1, KEY_GRAVE, \ + KEY_F13, KEY_F14, KEY_F15, KEY_F16, \ + KEY_F17, KEY_F18, KEY_F19, KEY_F20, \ + KEY_MENU, KEY_F4, KEY_F3, KEY_F2, \ + KEY_F1, KEY_VOLUMEUP, KEY_STOP, KEY_SENDFILE, \ + KEY_SYSRQ, KEY_F5, KEY_F6, KEY_F7, \ + KEY_F8, KEY_VOLUMEDOWN, KEY_DEL_EOL, KEY_DEL_EOS, \ + KEY_8, KEY_9, KEY_0, KEY_MINUS, \ + KEY_EQUAL, KEY_BACKSPACE, KEY_INS_LINE, KEY_DEL_LINE, \ + KEY_I, KEY_O, KEY_P, KEY_LEFTBRACE, \ + KEY_RIGHTBRACE, KEY_BACKSLASH, KEY_INSERT, KEY_DELETE, \ + KEY_J, KEY_K, KEY_L, KEY_SEMICOLON, \ + KEY_APOSTROPHE, KEY_ENTER, KEY_HOME, KEY_PAGEUP, \ + KEY_M, KEY_COMMA, KEY_DOT, KEY_SLASH, \ + KEY_BACKSLASH, KEY_SELECT, KEY_102ND, KEY_PAGEDOWN, \ + KEY_N, KEY_SPACE, KEY_NEXT, KEY_RESERVED, \ + KEY_LEFT, KEY_DOWN, KEY_UP, KEY_RIGHT + + +#define HIL_KEYCODES_SET3_TBLSIZE 128 +#define HIL_KEYCODES_SET3 \ + KEY_RESERVED, KEY_ESC, KEY_1, KEY_2, \ + KEY_3, KEY_4, KEY_5, KEY_6, \ + KEY_7, KEY_8, KEY_9, KEY_0, \ + KEY_MINUS, KEY_EQUAL, KEY_BACKSPACE, KEY_TAB, \ + KEY_Q, KEY_W, KEY_E, KEY_R, \ + KEY_T, KEY_Y, KEY_U, KEY_I, \ + KEY_O, KEY_P, KEY_LEFTBRACE, KEY_RIGHTBRACE, \ + KEY_ENTER, KEY_LEFTCTRL, KEY_A, KEY_S, \ + KEY_D, KEY_F, KEY_G, KEY_H, \ + KEY_J, KEY_K, KEY_L, KEY_SEMICOLON, \ + KEY_APOSTROPHE,KEY_GRAVE, KEY_LEFTSHIFT, KEY_BACKSLASH, \ + KEY_Z, KEY_X, KEY_C, KEY_V, \ + KEY_B, KEY_N, KEY_M, KEY_COMMA, \ + KEY_DOT, KEY_SLASH, KEY_RIGHTSHIFT, KEY_KPASTERISK, \ + KEY_LEFTALT, KEY_SPACE, KEY_CAPSLOCK, KEY_F1, \ + KEY_F2, KEY_F3, KEY_F4, KEY_F5, \ + KEY_F6, KEY_F7, KEY_F8, KEY_F9, \ + KEY_F10, KEY_NUMLOCK, KEY_SCROLLLOCK, KEY_KP7, \ + KEY_KP8, KEY_KP9, KEY_KPMINUS, KEY_KP4, \ + KEY_KP5, KEY_KP6, KEY_KPPLUS, KEY_KP1, \ + KEY_KP2, KEY_KP3, KEY_KP0, KEY_KPDOT, \ + KEY_SYSRQ, KEY_RESERVED, KEY_RESERVED, KEY_RESERVED, \ + KEY_RESERVED, KEY_RESERVED, KEY_RESERVED, KEY_RESERVED, \ + KEY_RESERVED, KEY_RESERVED, KEY_RESERVED, KEY_RESERVED, \ + KEY_UP, KEY_LEFT, KEY_DOWN, KEY_RIGHT, \ + KEY_HOME, KEY_PAGEUP, KEY_END, KEY_PAGEDOWN, \ + KEY_INSERT, KEY_DELETE, KEY_102ND, KEY_RESERVED, \ + KEY_RESERVED, KEY_RESERVED, KEY_RESERVED, KEY_RESERVED, \ + KEY_F1, KEY_F2, KEY_F3, KEY_F4, \ + KEY_F5, KEY_F6, KEY_F7, KEY_F8, \ + KEY_RESERVED, KEY_RESERVED, KEY_RESERVED, KEY_RESERVED, \ + KEY_RESERVED, KEY_RESERVED, KEY_RESERVED, KEY_RESERVED + + +/* Response to POL command, the "poll record header" */ + +#define HIL_POL_NUM_AXES_MASK 0x03 /* Number of axis reported */ +#define HIL_POL_CTS 0x04 /* Device ready to receive data */ +#define HIL_POL_STATUS_PENDING 0x08 /* Device has status to report */ +#define HIL_POL_CHARTYPE_MASK 0x70 /* Type of character data to follow */ +#define HIL_POL_CHARTYPE_NONE 0x00 /* No character data to follow */ +#define HIL_POL_CHARTYPE_RSVD1 0x10 /* Reserved Set 1 */ +#define HIL_POL_CHARTYPE_ASCII 0x20 /* U.S. ASCII */ +#define HIL_POL_CHARTYPE_BINARY 0x30 /* Binary data */ +#define HIL_POL_CHARTYPE_SET1 0x40 /* Keycode Set 1 */ +#define HIL_POL_CHARTYPE_RSVD2 0x50 /* Reserved Set 2 */ +#define HIL_POL_CHARTYPE_SET2 0x60 /* Keycode Set 2 */ +#define HIL_POL_CHARTYPE_SET3 0x70 /* Keycode Set 3 */ +#define HIL_POL_AXIS_ALT 0x80 /* Data is from axis set 2 */ + + +#endif /* _HIL_H_ */ diff --git a/include/linux/hil_mlc.h b/include/linux/hil_mlc.h new file mode 100644 index 000000000000..8df29ca48a13 --- /dev/null +++ b/include/linux/hil_mlc.h @@ -0,0 +1,168 @@ +/* + * HP Human Interface Loop Master Link Controller driver. + * + * Copyright (c) 2001 Brian S. Julin + * All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions, and the following disclaimer, + * without modification. + * 2. The name of the author may not be used to endorse or promote products + * derived from this software without specific prior written permission. + * + * Alternatively, this software may be distributed under the terms of the + * GNU General Public License ("GPL"). + * + * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND + * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE + * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE + * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE FOR + * ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL + * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS + * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) + * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT + * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY + * + * References: + * HP-HIL Technical Reference Manual. Hewlett Packard Product No. 45918A + * + */ + +#include +#include +#include +#include +#include +#include + +typedef struct hil_mlc hil_mlc; + +/* The HIL has a complicated state engine. + * We define the structure of nodes in the state engine here. + */ +enum hilse_act { + /* HILSE_OUT prepares to receive input if the next node + * is an IN or EXPECT, and then sends the given packet. + */ + HILSE_OUT = 0, + + /* HILSE_CTS checks if the loop is busy. */ + HILSE_CTS, + + /* HILSE_OUT_LAST sends the given command packet to + * the last configured/running device on the loop. + */ + HILSE_OUT_LAST, + + /* HILSE_OUT_DISC sends the given command packet to + * the next device past the last configured/running one. + */ + HILSE_OUT_DISC, + + /* HILSE_FUNC runs a callback function with given arguments. + * a positive return value causes the "ugly" branch to be taken. + */ + HILSE_FUNC, + + /* HILSE_IN simply expects any non-errored packet to arrive + * within arg usecs. + */ + HILSE_IN = 0x100, + + /* HILSE_EXPECT expects a particular packet to arrive + * within arg usecs, any other packet is considered an error. + */ + HILSE_EXPECT, + + /* HILSE_EXPECT_LAST as above but dev field should be last + * discovered/operational device. + */ + HILSE_EXPECT_LAST, + + /* HILSE_EXPECT_LAST as above but dev field should be first + * undiscovered/inoperational device. + */ + HILSE_EXPECT_DISC +}; + +typedef int (hilse_func) (hil_mlc *mlc, int arg); +struct hilse_node { + enum hilse_act act; /* How to process this node */ + union { + hilse_func *func; /* Function to call if HILSE_FUNC */ + hil_packet packet; /* Packet to send or to compare */ + } object; + int arg; /* Timeout in usec or parm for func */ + int good; /* Node to jump to on success */ + int bad; /* Node to jump to on error */ + int ugly; /* Node to jump to on timeout */ +}; + +/* Methods for back-end drivers, e.g. hp_sdc_mlc */ +typedef int (hil_mlc_cts) (hil_mlc *mlc); +typedef void (hil_mlc_out) (hil_mlc *mlc); +typedef int (hil_mlc_in) (hil_mlc *mlc, suseconds_t timeout); + +struct hil_mlc_devinfo { + uint8_t idd[16]; /* Device ID Byte and Describe Record */ + uint8_t rsc[16]; /* Security Code Header and Record */ + uint8_t exd[16]; /* Extended Describe Record */ + uint8_t rnm[16]; /* Device name as returned by RNM command */ +}; + +struct hil_mlc_serio_map { + hil_mlc *mlc; + int di_revmap; + int didx; +}; + +/* How many (possibly old/detached) devices the we try to keep track of */ +#define HIL_MLC_DEVMEM 16 + +struct hil_mlc { + struct list_head list; /* hil_mlc is organized as linked list */ + + rwlock_t lock; + + void *priv; /* Data specific to a particular type of MLC */ + + int seidx; /* Current node in state engine */ + int istarted, ostarted; + + hil_mlc_cts *cts; + struct semaphore csem; /* Raised when loop idle */ + + hil_mlc_out *out; + struct semaphore osem; /* Raised when outpacket dispatched */ + hil_packet opacket; + + hil_mlc_in *in; + struct semaphore isem; /* Raised when a packet arrives */ + hil_packet ipacket[16]; + hil_packet imatch; + int icount; + struct timeval instart; + suseconds_t intimeout; + + int ddi; /* Last operational device id */ + int lcv; /* LCV to throttle loops */ + struct timeval lcv_tv; /* Time loop was started */ + + int di_map[7]; /* Maps below items to live devs */ + struct hil_mlc_devinfo di[HIL_MLC_DEVMEM]; + struct serio *serio[HIL_MLC_DEVMEM]; + struct hil_mlc_serio_map serio_map[HIL_MLC_DEVMEM]; + hil_packet serio_opacket[HIL_MLC_DEVMEM]; + int serio_oidx[HIL_MLC_DEVMEM]; + struct hil_mlc_devinfo di_scratch; /* Temporary area */ + + int opercnt; + + struct tasklet_struct *tasklet; +}; + +int hil_mlc_register(hil_mlc *mlc); +int hil_mlc_unregister(hil_mlc *mlc); diff --git a/include/linux/hp_sdc.h b/include/linux/hp_sdc.h new file mode 100644 index 000000000000..debd71515312 --- /dev/null +++ b/include/linux/hp_sdc.h @@ -0,0 +1,300 @@ +/* + * HP i8042 System Device Controller -- header + * + * Copyright (c) 2001 Brian S. Julin + * All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions, and the following disclaimer, + * without modification. + * 2. The name of the author may not be used to endorse or promote products + * derived from this software without specific prior written permission. + * + * Alternatively, this software may be distributed under the terms of the + * GNU General Public License ("GPL"). + * + * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND + * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE + * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE + * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE FOR + * ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL + * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS + * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) + * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT + * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY + * + * References: + * + * HP-HIL Technical Reference Manual. Hewlett Packard Product No. 45918A + * + * System Device Controller Microprocessor Firmware Theory of Operation + * for Part Number 1820-4784 Revision B. Dwg No. A-1820-4784-2 + * + */ + +#ifndef _LINUX_HP_SDC_H +#define _LINUX_HP_SDC_H + +#include +#include +#include +#include +#if defined(__hppa__) +#include +#endif + + +/* No 4X status reads take longer than this (in usec). + */ +#define HP_SDC_MAX_REG_DELAY 20000 + +typedef void (hp_sdc_irqhook) (int irq, void *dev_id, + uint8_t status, uint8_t data); + +int hp_sdc_request_timer_irq(hp_sdc_irqhook *callback); +int hp_sdc_request_hil_irq(hp_sdc_irqhook *callback); +int hp_sdc_request_cooked_irq(hp_sdc_irqhook *callback); +int hp_sdc_release_timer_irq(hp_sdc_irqhook *callback); +int hp_sdc_release_hil_irq(hp_sdc_irqhook *callback); +int hp_sdc_release_cooked_irq(hp_sdc_irqhook *callback); + +typedef struct { + int actidx; /* Start of act. Acts are atomic WRT I/O to SDC */ + int idx; /* Index within the act */ + int endidx; /* transaction is over and done if idx == endidx */ + uint8_t *seq; /* commands/data for the transaction */ + union { + hp_sdc_irqhook *irqhook; /* Callback, isr or tasklet context */ + struct semaphore *semaphore; /* Semaphore to sleep on. */ + } act; +} hp_sdc_transaction; +int hp_sdc_enqueue_transaction(hp_sdc_transaction *this); +int hp_sdc_dequeue_transaction(hp_sdc_transaction *this); + +/* The HP_SDC_ACT* values are peculiar to this driver. + * Nuance: never HP_SDC_ACT_DATAIN | HP_SDC_ACT_DEALLOC, use another + * act to perform the dealloc. + */ +#define HP_SDC_ACT_PRECMD 0x01 /* Send a command first */ +#define HP_SDC_ACT_DATAREG 0x02 /* Set data registers */ +#define HP_SDC_ACT_DATAOUT 0x04 /* Send data bytes */ +#define HP_SDC_ACT_POSTCMD 0x08 /* Send command after */ +#define HP_SDC_ACT_DATAIN 0x10 /* Collect data after */ +#define HP_SDC_ACT_DURING 0x1f +#define HP_SDC_ACT_SEMAPHORE 0x20 /* Raise semaphore after */ +#define HP_SDC_ACT_CALLBACK 0x40 /* Pass data to IRQ handler */ +#define HP_SDC_ACT_DEALLOC 0x80 /* Destroy transaction after */ +#define HP_SDC_ACT_AFTER 0xe0 +#define HP_SDC_ACT_DEAD 0x60 /* Act timed out. */ + +/* Rest of the flags are straightforward representation of the SDC interface */ +#define HP_SDC_STATUS_IBF 0x02 /* Input buffer full */ + +#define HP_SDC_STATUS_IRQMASK 0xf0 /* Bits containing "level 1" irq */ +#define HP_SDC_STATUS_PERIODIC 0x10 /* Periodic 10ms timer */ +#define HP_SDC_STATUS_USERTIMER 0x20 /* "Special purpose" timer */ +#define HP_SDC_STATUS_TIMER 0x30 /* Both PERIODIC and USERTIMER */ +#define HP_SDC_STATUS_REG 0x40 /* Data from an i8042 register */ +#define HP_SDC_STATUS_HILCMD 0x50 /* Command from HIL MLC */ +#define HP_SDC_STATUS_HILDATA 0x60 /* Data from HIL MLC */ +#define HP_SDC_STATUS_PUP 0x70 /* Sucessful power-up self test */ +#define HP_SDC_STATUS_KCOOKED 0x80 /* Key from cooked kbd */ +#define HP_SDC_STATUS_KRPG 0xc0 /* Key from Repeat Gen */ +#define HP_SDC_STATUS_KMOD_SUP 0x10 /* Shift key is up */ +#define HP_SDC_STATUS_KMOD_CUP 0x20 /* Control key is up */ + +#define HP_SDC_NMISTATUS_FHS 0x40 /* NMI is a fast handshake irq */ + +/* Internal i8042 registers (there are more, but they are not too useful). */ + +#define HP_SDC_USE 0x02 /* Resource usage (including OB bit) */ +#define HP_SDC_IM 0x04 /* Interrupt mask */ +#define HP_SDC_CFG 0x11 /* Configuration register */ +#define HP_SDC_KBLANGUAGE 0x12 /* Keyboard language */ + +#define HP_SDC_D0 0x70 /* General purpose data buffer 0 */ +#define HP_SDC_D1 0x71 /* General purpose data buffer 1 */ +#define HP_SDC_D2 0x72 /* General purpose data buffer 2 */ +#define HP_SDC_D3 0x73 /* General purpose data buffer 3 */ +#define HP_SDC_VT1 0x74 /* Timer for voice 1 */ +#define HP_SDC_VT2 0x75 /* Timer for voice 2 */ +#define HP_SDC_VT3 0x76 /* Timer for voice 3 */ +#define HP_SDC_VT4 0x77 /* Timer for voice 4 */ +#define HP_SDC_KBN 0x78 /* Which HIL devs are Nimitz */ +#define HP_SDC_KBC 0x79 /* Which HIL devs are cooked kbds */ +#define HP_SDC_LPS 0x7a /* i8042's view of HIL status */ +#define HP_SDC_LPC 0x7b /* i8042's view of HIL "control" */ +#define HP_SDC_RSV 0x7c /* Reserved "for testing" */ +#define HP_SDC_LPR 0x7d /* i8042 count of HIL reconfigs */ +#define HP_SDC_XTD 0x7e /* "Extended Configuration" register */ +#define HP_SDC_STR 0x7f /* i8042 self-test result */ + +/* Bitfields for above registers */ +#define HP_SDC_USE_LOOP 0x04 /* Command is currently on the loop. */ + +#define HP_SDC_IM_MASK 0x1f /* these bits not part of cmd/status */ +#define HP_SDC_IM_FH 0x10 /* Mask the fast handshake irq */ +#define HP_SDC_IM_PT 0x08 /* Mask the periodic timer irq */ +#define HP_SDC_IM_TIMERS 0x04 /* Mask the MT/DT/CT irq */ +#define HP_SDC_IM_RESET 0x02 /* Mask the reset key irq */ +#define HP_SDC_IM_HIL 0x01 /* Mask the HIL MLC irq */ + +#define HP_SDC_CFG_ROLLOVER 0x08 /* WTF is "N-key rollover"? */ +#define HP_SDC_CFG_KBD 0x10 /* There is a keyboard */ +#define HP_SDC_CFG_NEW 0x20 /* Supports/uses HIL MLC */ +#define HP_SDC_CFG_KBD_OLD 0x03 /* keyboard code for non-HIL */ +#define HP_SDC_CFG_KBD_NEW 0x07 /* keyboard code from HIL autoconfig */ +#define HP_SDC_CFG_REV 0x40 /* Code revision bit */ +#define HP_SDC_CFG_IDPROM 0x80 /* IDPROM present in kbd (not HIL) */ + +#define HP_SDC_LPS_NDEV 0x07 /* # devices autoconfigured on HIL */ +#define HP_SDC_LPS_ACSUCC 0x08 /* loop autoconfigured successfully */ +#define HP_SDC_LPS_ACFAIL 0x80 /* last loop autoconfigure failed */ + +#define HP_SDC_LPC_APE_IPF 0x01 /* HIL MLC APE/IPF (autopoll) set */ +#define HP_SDC_LPC_ARCONERR 0x02 /* i8042 autoreconfigs loop on err */ +#define HP_SDC_LPC_ARCQUIET 0x03 /* i8042 doesn't report autoreconfigs*/ +#define HP_SDC_LPC_COOK 0x10 /* i8042 cooks devices in _KBN */ +#define HP_SDC_LPC_RC 0x80 /* causes autoreconfig */ + +#define HP_SDC_XTD_REV 0x07 /* contains revision code */ +#define HP_SDC_XTD_REV_STRINGS(val, str) \ +switch (val) { \ + case 0x1: str = "1820-3712"; break; \ + case 0x2: str = "1820-4379"; break; \ + case 0x3: str = "1820-4784"; break; \ + default: str = "unknown"; \ +}; +#define HP_SDC_XTD_BEEPER 0x08 /* TI SN76494 beeper available */ +#define HP_SDC_XTD_BBRTC 0x20 /* OKI MSM-58321 BBRTC present */ + +#define HP_SDC_CMD_LOAD_RT 0x31 /* Load real time (from 8042) */ +#define HP_SDC_CMD_LOAD_FHS 0x36 /* Load the fast handshake timer */ +#define HP_SDC_CMD_LOAD_MT 0x38 /* Load the match timer */ +#define HP_SDC_CMD_LOAD_DT 0x3B /* Load the delay timer */ +#define HP_SDC_CMD_LOAD_CT 0x3E /* Load the cycle timer */ + +#define HP_SDC_CMD_SET_IM 0x40 /* 010xxxxx == set irq mask */ + +/* The documents provided do not explicitly state that all registers betweem + * 0x01 and 0x1f inclusive can be read by sending their register index as a + * command, but this is implied and appears to be the case. + */ +#define HP_SDC_CMD_READ_RAM 0x00 /* Load from i8042 RAM (autoinc) */ +#define HP_SDC_CMD_READ_USE 0x02 /* Undocumented! Load from usage reg */ +#define HP_SDC_CMD_READ_IM 0x04 /* Load current interrupt mask */ +#define HP_SDC_CMD_READ_KCC 0x11 /* Load primary kbd config code */ +#define HP_SDC_CMD_READ_KLC 0x12 /* Load primary kbd language code */ +#define HP_SDC_CMD_READ_T1 0x13 /* Load timer output buffer byte 1 */ +#define HP_SDC_CMD_READ_T2 0x14 /* Load timer output buffer byte 1 */ +#define HP_SDC_CMD_READ_T3 0x15 /* Load timer output buffer byte 1 */ +#define HP_SDC_CMD_READ_T4 0x16 /* Load timer output buffer byte 1 */ +#define HP_SDC_CMD_READ_T5 0x17 /* Load timer output buffer byte 1 */ +#define HP_SDC_CMD_READ_D0 0xf0 /* Load from i8042 RAM location 0x70 */ +#define HP_SDC_CMD_READ_D1 0xf1 /* Load from i8042 RAM location 0x71 */ +#define HP_SDC_CMD_READ_D2 0xf2 /* Load from i8042 RAM location 0x72 */ +#define HP_SDC_CMD_READ_D3 0xf3 /* Load from i8042 RAM location 0x73 */ +#define HP_SDC_CMD_READ_VT1 0xf4 /* Load from i8042 RAM location 0x74 */ +#define HP_SDC_CMD_READ_VT2 0xf5 /* Load from i8042 RAM location 0x75 */ +#define HP_SDC_CMD_READ_VT3 0xf6 /* Load from i8042 RAM location 0x76 */ +#define HP_SDC_CMD_READ_VT4 0xf7 /* Load from i8042 RAM location 0x77 */ +#define HP_SDC_CMD_READ_KBN 0xf8 /* Load from i8042 RAM location 0x78 */ +#define HP_SDC_CMD_READ_KBC 0xf9 /* Load from i8042 RAM location 0x79 */ +#define HP_SDC_CMD_READ_LPS 0xfa /* Load from i8042 RAM location 0x7a */ +#define HP_SDC_CMD_READ_LPC 0xfb /* Load from i8042 RAM location 0x7b */ +#define HP_SDC_CMD_READ_RSV 0xfc /* Load from i8042 RAM location 0x7c */ +#define HP_SDC_CMD_READ_LPR 0xfd /* Load from i8042 RAM location 0x7d */ +#define HP_SDC_CMD_READ_XTD 0xfe /* Load from i8042 RAM location 0x7e */ +#define HP_SDC_CMD_READ_STR 0xff /* Load from i8042 RAM location 0x7f */ + +#define HP_SDC_CMD_SET_ARD 0xA0 /* Set emulated autorepeat delay */ +#define HP_SDC_CMD_SET_ARR 0xA2 /* Set emulated autorepeat rate */ +#define HP_SDC_CMD_SET_BELL 0xA3 /* Set voice 3 params for "beep" cmd */ +#define HP_SDC_CMD_SET_RPGR 0xA6 /* Set "RPG" irq rate (doesn't work) */ +#define HP_SDC_CMD_SET_RTMS 0xAD /* Set the RTC time (milliseconds) */ +#define HP_SDC_CMD_SET_RTD 0xAF /* Set the RTC time (days) */ +#define HP_SDC_CMD_SET_FHS 0xB2 /* Set fast handshake timer */ +#define HP_SDC_CMD_SET_MT 0xB4 /* Set match timer */ +#define HP_SDC_CMD_SET_DT 0xB7 /* Set delay timer */ +#define HP_SDC_CMD_SET_CT 0xBA /* Set cycle timer */ +#define HP_SDC_CMD_SET_RAMP 0xC1 /* Reset READ_RAM autoinc counter */ +#define HP_SDC_CMD_SET_D0 0xe0 /* Load to i8042 RAM location 0x70 */ +#define HP_SDC_CMD_SET_D1 0xe1 /* Load to i8042 RAM location 0x71 */ +#define HP_SDC_CMD_SET_D2 0xe2 /* Load to i8042 RAM location 0x72 */ +#define HP_SDC_CMD_SET_D3 0xe3 /* Load to i8042 RAM location 0x73 */ +#define HP_SDC_CMD_SET_VT1 0xe4 /* Load to i8042 RAM location 0x74 */ +#define HP_SDC_CMD_SET_VT2 0xe5 /* Load to i8042 RAM location 0x75 */ +#define HP_SDC_CMD_SET_VT3 0xe6 /* Load to i8042 RAM location 0x76 */ +#define HP_SDC_CMD_SET_VT4 0xe7 /* Load to i8042 RAM location 0x77 */ +#define HP_SDC_CMD_SET_KBN 0xe8 /* Load to i8042 RAM location 0x78 */ +#define HP_SDC_CMD_SET_KBC 0xe9 /* Load to i8042 RAM location 0x79 */ +#define HP_SDC_CMD_SET_LPS 0xea /* Load to i8042 RAM location 0x7a */ +#define HP_SDC_CMD_SET_LPC 0xeb /* Load to i8042 RAM location 0x7b */ +#define HP_SDC_CMD_SET_RSV 0xec /* Load to i8042 RAM location 0x7c */ +#define HP_SDC_CMD_SET_LPR 0xed /* Load to i8042 RAM location 0x7d */ +#define HP_SDC_CMD_SET_XTD 0xee /* Load to i8042 RAM location 0x7e */ +#define HP_SDC_CMD_SET_STR 0xef /* Load to i8042 RAM location 0x7f */ + +#define HP_SDC_CMD_DO_RTCW 0xc2 /* i8042 RAM 0x70 --> RTC */ +#define HP_SDC_CMD_DO_RTCR 0xc3 /* RTC[0x70 0:3] --> irq/status/data */ +#define HP_SDC_CMD_DO_BEEP 0xc4 /* i8042 RAM 0x70-74 --> beeper,VT3 */ +#define HP_SDC_CMD_DO_HIL 0xc5 /* i8042 RAM 0x70-73 --> + HIL MLC R0,R1 i8042 HIL watchdog */ + +/* Values used to (de)mangle input/output to/from the HIL MLC */ +#define HP_SDC_DATA 0x40 /* Data from an 8042 register */ +#define HP_SDC_HIL_CMD 0x50 /* Data from HIL MLC R1/8042 */ +#define HP_SDC_HIL_R1MASK 0x0f /* Contents of HIL MLC R1 0:3 */ +#define HP_SDC_HIL_AUTO 0x10 /* Set if POL results from i8042 */ +#define HP_SDC_HIL_ISERR 0x80 /* Has meaning as in next 4 values */ +#define HP_SDC_HIL_RC_DONE 0x80 /* i8042 auto-configured loop */ +#define HP_SDC_HIL_ERR 0x81 /* HIL MLC R2 had a bit set */ +#define HP_SDC_HIL_TO 0x82 /* i8042 HIL watchdog expired */ +#define HP_SDC_HIL_RC 0x84 /* i8042 is auto-configuring loop */ +#define HP_SDC_HIL_DAT 0x60 /* Data from HIL MLC R0 */ + + +typedef struct { + rwlock_t ibf_lock; + rwlock_t lock; /* user/tasklet lock */ + rwlock_t rtq_lock; /* isr/tasklet lock */ + rwlock_t hook_lock; /* isr/user lock for handler add/del */ + + unsigned int irq, nmi; /* Our IRQ lines */ + unsigned long base_io, status_io, data_io; /* Our IO ports */ + + uint8_t im; /* Interrupt mask */ + int set_im; /* Interrupt mask needs to be set. */ + + int ibf; /* Last known status of IBF flag */ + uint8_t wi; /* current i8042 write index */ + uint8_t r7[4]; /* current i8042[0x70 - 0x74] values */ + uint8_t r11, r7e; /* Values from version/revision regs */ + + hp_sdc_irqhook *timer, *reg, *hil, *pup, *cooked; + +#define HP_SDC_QUEUE_LEN 16 + hp_sdc_transaction *tq[HP_SDC_QUEUE_LEN]; /* All pending read/writes */ + + int rcurr, rqty; /* Current read transact in process */ + struct timeval rtv; /* Time when current read started */ + int wcurr; /* Current write transact in process */ + + int dev_err; /* carries status from registration */ +#if defined(__hppa__) + struct parisc_device *dev; +#elif defined(__mc68000__) + void *dev; +#else +#error No support for device registration on this arch yet. +#endif + + struct timer_list kicker; /* Keeps below task alive */ + struct tasklet_struct task; + +} hp_i8042_sdc; + +#endif /* _LINUX_HP_SDC_H */ diff --git a/include/linux/input.h b/include/linux/input.h index e8c296ff6257..d7836311ada9 100644 --- a/include/linux/input.h +++ b/include/linux/input.h @@ -644,6 +644,7 @@ struct input_absinfo { #define BUS_ADB 0x17 #define BUS_I2C 0x18 #define BUS_HOST 0x19 +#define BUS_GSC 0x1A /* * Values describing the status of an effect -- cgit v1.2.3 From cf482935c6abe5245e481213c6e6df808c976f56 Mon Sep 17 00:00:00 2001 From: Jeff Garzik Date: Sat, 22 Oct 2005 00:19:33 -0400 Subject: libata: turn on block layer clustering --- include/linux/libata.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include/linux') diff --git a/include/linux/libata.h b/include/linux/libata.h index 0e214f8c8f9f..634b5aa0a615 100644 --- a/include/linux/libata.h +++ b/include/linux/libata.h @@ -91,7 +91,7 @@ enum { ATA_SHT_EMULATED = 1, ATA_SHT_CMD_PER_LUN = 1, ATA_SHT_THIS_ID = -1, - ATA_SHT_USE_CLUSTERING = 0, + ATA_SHT_USE_CLUSTERING = 1, /* struct ata_device stuff */ ATA_DFLAG_LBA48 = (1 << 0), /* device supports LBA48 */ -- cgit v1.2.3 From 057ace5e79da9ebf2aa82833cfea825533ac06fb Mon Sep 17 00:00:00 2001 From: Jeff Garzik Date: Sat, 22 Oct 2005 14:27:05 -0400 Subject: libata: const-ification bombing run Enforce access rules where appropriate. If the compiler is smart enough, this may buy us an optimization or two as a side effect. --- drivers/scsi/ahci.c | 2 +- drivers/scsi/ata_piix.c | 4 +-- drivers/scsi/libata-core.c | 68 ++++++++++++++++++++++----------------------- drivers/scsi/libata-scsi.c | 24 ++++++++-------- drivers/scsi/libata.h | 2 +- drivers/scsi/pdc_adma.c | 2 +- drivers/scsi/sata_mv.c | 2 +- drivers/scsi/sata_nv.c | 2 +- drivers/scsi/sata_promise.c | 12 ++++---- drivers/scsi/sata_qstor.c | 2 +- drivers/scsi/sata_sil.c | 2 +- drivers/scsi/sata_sil24.c | 2 +- drivers/scsi/sata_sis.c | 2 +- drivers/scsi/sata_svw.c | 4 +-- drivers/scsi/sata_sx4.c | 10 +++---- drivers/scsi/sata_uli.c | 2 +- drivers/scsi/sata_via.c | 2 +- drivers/scsi/sata_vsc.c | 4 +-- include/linux/ata.h | 6 ++-- include/linux/libata.h | 32 ++++++++++----------- 20 files changed, 93 insertions(+), 93 deletions(-) (limited to 'include/linux') diff --git a/drivers/scsi/ahci.c b/drivers/scsi/ahci.c index cfa22e4ee547..fe8187d6f58b 100644 --- a/drivers/scsi/ahci.c +++ b/drivers/scsi/ahci.c @@ -216,7 +216,7 @@ static Scsi_Host_Template ahci_sht = { .ordered_flush = 1, }; -static struct ata_port_operations ahci_ops = { +static const struct ata_port_operations ahci_ops = { .port_disable = ata_port_disable, .check_status = ahci_check_status, diff --git a/drivers/scsi/ata_piix.c b/drivers/scsi/ata_piix.c index d71cef767cec..be021478f416 100644 --- a/drivers/scsi/ata_piix.c +++ b/drivers/scsi/ata_piix.c @@ -147,7 +147,7 @@ static Scsi_Host_Template piix_sht = { .ordered_flush = 1, }; -static struct ata_port_operations piix_pata_ops = { +static const struct ata_port_operations piix_pata_ops = { .port_disable = ata_port_disable, .set_piomode = piix_set_piomode, .set_dmamode = piix_set_dmamode, @@ -177,7 +177,7 @@ static struct ata_port_operations piix_pata_ops = { .host_stop = ata_host_stop, }; -static struct ata_port_operations piix_sata_ops = { +static const struct ata_port_operations piix_sata_ops = { .port_disable = ata_port_disable, .tf_load = ata_tf_load, diff --git a/drivers/scsi/libata-core.c b/drivers/scsi/libata-core.c index 9269fd9b814f..d55f12dacfcb 100644 --- a/drivers/scsi/libata-core.c +++ b/drivers/scsi/libata-core.c @@ -67,9 +67,9 @@ static void ata_dev_reread_id(struct ata_port *ap, struct ata_device *dev); static void ata_dev_init_params(struct ata_port *ap, struct ata_device *dev); static void ata_set_mode(struct ata_port *ap); static void ata_dev_set_xfermode(struct ata_port *ap, struct ata_device *dev); -static unsigned int ata_get_mode_mask(struct ata_port *ap, int shift); +static unsigned int ata_get_mode_mask(const struct ata_port *ap, int shift); static int fgb(u32 bitmap); -static int ata_choose_xfer_mode(struct ata_port *ap, +static int ata_choose_xfer_mode(const struct ata_port *ap, u8 *xfer_mode_out, unsigned int *xfer_shift_out); static void __ata_qc_complete(struct ata_queued_cmd *qc); @@ -97,7 +97,7 @@ MODULE_VERSION(DRV_VERSION); * Inherited from caller. */ -static void ata_tf_load_pio(struct ata_port *ap, struct ata_taskfile *tf) +static void ata_tf_load_pio(struct ata_port *ap, const struct ata_taskfile *tf) { struct ata_ioports *ioaddr = &ap->ioaddr; unsigned int is_addr = tf->flags & ATA_TFLAG_ISADDR; @@ -155,7 +155,7 @@ static void ata_tf_load_pio(struct ata_port *ap, struct ata_taskfile *tf) * Inherited from caller. */ -static void ata_tf_load_mmio(struct ata_port *ap, struct ata_taskfile *tf) +static void ata_tf_load_mmio(struct ata_port *ap, const struct ata_taskfile *tf) { struct ata_ioports *ioaddr = &ap->ioaddr; unsigned int is_addr = tf->flags & ATA_TFLAG_ISADDR; @@ -224,7 +224,7 @@ static void ata_tf_load_mmio(struct ata_port *ap, struct ata_taskfile *tf) * LOCKING: * Inherited from caller. */ -void ata_tf_load(struct ata_port *ap, struct ata_taskfile *tf) +void ata_tf_load(struct ata_port *ap, const struct ata_taskfile *tf) { if (ap->flags & ATA_FLAG_MMIO) ata_tf_load_mmio(ap, tf); @@ -244,7 +244,7 @@ void ata_tf_load(struct ata_port *ap, struct ata_taskfile *tf) * spin_lock_irqsave(host_set lock) */ -static void ata_exec_command_pio(struct ata_port *ap, struct ata_taskfile *tf) +static void ata_exec_command_pio(struct ata_port *ap, const struct ata_taskfile *tf) { DPRINTK("ata%u: cmd 0x%X\n", ap->id, tf->command); @@ -265,7 +265,7 @@ static void ata_exec_command_pio(struct ata_port *ap, struct ata_taskfile *tf) * spin_lock_irqsave(host_set lock) */ -static void ata_exec_command_mmio(struct ata_port *ap, struct ata_taskfile *tf) +static void ata_exec_command_mmio(struct ata_port *ap, const struct ata_taskfile *tf) { DPRINTK("ata%u: cmd 0x%X\n", ap->id, tf->command); @@ -285,7 +285,7 @@ static void ata_exec_command_mmio(struct ata_port *ap, struct ata_taskfile *tf) * LOCKING: * spin_lock_irqsave(host_set lock) */ -void ata_exec_command(struct ata_port *ap, struct ata_taskfile *tf) +void ata_exec_command(struct ata_port *ap, const struct ata_taskfile *tf) { if (ap->flags & ATA_FLAG_MMIO) ata_exec_command_mmio(ap, tf); @@ -305,7 +305,7 @@ void ata_exec_command(struct ata_port *ap, struct ata_taskfile *tf) * Obtains host_set lock. */ -static inline void ata_exec(struct ata_port *ap, struct ata_taskfile *tf) +static inline void ata_exec(struct ata_port *ap, const struct ata_taskfile *tf) { unsigned long flags; @@ -328,7 +328,7 @@ static inline void ata_exec(struct ata_port *ap, struct ata_taskfile *tf) * Obtains host_set lock. */ -static void ata_tf_to_host(struct ata_port *ap, struct ata_taskfile *tf) +static void ata_tf_to_host(struct ata_port *ap, const struct ata_taskfile *tf) { ap->ops->tf_load(ap, tf); @@ -348,7 +348,7 @@ static void ata_tf_to_host(struct ata_port *ap, struct ata_taskfile *tf) * spin_lock_irqsave(host_set lock) */ -void ata_tf_to_host_nolock(struct ata_port *ap, struct ata_taskfile *tf) +void ata_tf_to_host_nolock(struct ata_port *ap, const struct ata_taskfile *tf) { ap->ops->tf_load(ap, tf); ap->ops->exec_command(ap, tf); @@ -558,7 +558,7 @@ u8 ata_chk_err(struct ata_port *ap) * Inherited from caller. */ -void ata_tf_to_fis(struct ata_taskfile *tf, u8 *fis, u8 pmp) +void ata_tf_to_fis(const struct ata_taskfile *tf, u8 *fis, u8 pmp) { fis[0] = 0x27; /* Register - Host to Device FIS */ fis[1] = (pmp & 0xf) | (1 << 7); /* Port multiplier number, @@ -599,7 +599,7 @@ void ata_tf_to_fis(struct ata_taskfile *tf, u8 *fis, u8 pmp) * Inherited from caller. */ -void ata_tf_from_fis(u8 *fis, struct ata_taskfile *tf) +void ata_tf_from_fis(const u8 *fis, struct ata_taskfile *tf) { tf->command = fis[2]; /* status */ tf->feature = fis[3]; /* error */ @@ -845,7 +845,7 @@ static unsigned int ata_devchk(struct ata_port *ap, * the event of failure. */ -unsigned int ata_dev_classify(struct ata_taskfile *tf) +unsigned int ata_dev_classify(const struct ata_taskfile *tf) { /* Apple's open source Darwin code hints that some devices only * put a proper signature into the LBA mid/high registers, @@ -937,7 +937,7 @@ static u8 ata_dev_try_classify(struct ata_port *ap, unsigned int device) * caller. */ -void ata_dev_id_string(u16 *id, unsigned char *s, +void ata_dev_id_string(const u16 *id, unsigned char *s, unsigned int ofs, unsigned int len) { unsigned int c; @@ -1054,7 +1054,7 @@ void ata_dev_select(struct ata_port *ap, unsigned int device, * caller. */ -static inline void ata_dump_id(struct ata_device *dev) +static inline void ata_dump_id(const struct ata_device *dev) { DPRINTK("49==0x%04x " "53==0x%04x " @@ -1089,7 +1089,7 @@ static inline void ata_dump_id(struct ata_device *dev) * FIXME: pre IDE drive timing (do we care ?). */ -static unsigned int ata_pio_modes(struct ata_device *adev) +static unsigned int ata_pio_modes(const struct ata_device *adev) { u16 modes; @@ -1352,7 +1352,7 @@ err_out: } -static inline u8 ata_dev_knobble(struct ata_port *ap) +static inline u8 ata_dev_knobble(const struct ata_port *ap) { return ((ap->cbl == ATA_CBL_SATA) && (!ata_id_is_sata(ap->device->id))); } @@ -1684,7 +1684,7 @@ int ata_timing_compute(struct ata_device *adev, unsigned short speed, return 0; } -static struct { +static const struct { unsigned int shift; u8 base; } xfer_mode_classes[] = { @@ -2093,7 +2093,8 @@ err_out: DPRINTK("EXIT\n"); } -static void ata_pr_blacklisted(struct ata_port *ap, struct ata_device *dev) +static void ata_pr_blacklisted(const struct ata_port *ap, + const struct ata_device *dev) { printk(KERN_WARNING "ata%u: dev %u is on DMA blacklist, disabling DMA\n", ap->id, dev->devno); @@ -2131,7 +2132,7 @@ static const char * ata_dma_blacklist [] = { "_NEC DV5800A", }; -static int ata_dma_blacklisted(struct ata_port *ap, struct ata_device *dev) +static int ata_dma_blacklisted(const struct ata_device *dev) { unsigned char model_num[40]; char *s; @@ -2156,9 +2157,9 @@ static int ata_dma_blacklisted(struct ata_port *ap, struct ata_device *dev) return 0; } -static unsigned int ata_get_mode_mask(struct ata_port *ap, int shift) +static unsigned int ata_get_mode_mask(const struct ata_port *ap, int shift) { - struct ata_device *master, *slave; + const struct ata_device *master, *slave; unsigned int mask; master = &ap->device[0]; @@ -2170,14 +2171,14 @@ static unsigned int ata_get_mode_mask(struct ata_port *ap, int shift) mask = ap->udma_mask; if (ata_dev_present(master)) { mask &= (master->id[ATA_ID_UDMA_MODES] & 0xff); - if (ata_dma_blacklisted(ap, master)) { + if (ata_dma_blacklisted(master)) { mask = 0; ata_pr_blacklisted(ap, master); } } if (ata_dev_present(slave)) { mask &= (slave->id[ATA_ID_UDMA_MODES] & 0xff); - if (ata_dma_blacklisted(ap, slave)) { + if (ata_dma_blacklisted(slave)) { mask = 0; ata_pr_blacklisted(ap, slave); } @@ -2187,14 +2188,14 @@ static unsigned int ata_get_mode_mask(struct ata_port *ap, int shift) mask = ap->mwdma_mask; if (ata_dev_present(master)) { mask &= (master->id[ATA_ID_MWDMA_MODES] & 0x07); - if (ata_dma_blacklisted(ap, master)) { + if (ata_dma_blacklisted(master)) { mask = 0; ata_pr_blacklisted(ap, master); } } if (ata_dev_present(slave)) { mask &= (slave->id[ATA_ID_MWDMA_MODES] & 0x07); - if (ata_dma_blacklisted(ap, slave)) { + if (ata_dma_blacklisted(slave)) { mask = 0; ata_pr_blacklisted(ap, slave); } @@ -2258,7 +2259,7 @@ static int fgb(u32 bitmap) * Zero on success, negative on error. */ -static int ata_choose_xfer_mode(struct ata_port *ap, +static int ata_choose_xfer_mode(const struct ata_port *ap, u8 *xfer_mode_out, unsigned int *xfer_shift_out) { @@ -4144,7 +4145,7 @@ static void ata_host_remove(struct ata_port *ap, unsigned int do_unregister) static void ata_host_init(struct ata_port *ap, struct Scsi_Host *host, struct ata_host_set *host_set, - struct ata_probe_ent *ent, unsigned int port_no) + const struct ata_probe_ent *ent, unsigned int port_no) { unsigned int i; @@ -4203,7 +4204,7 @@ static void ata_host_init(struct ata_port *ap, struct Scsi_Host *host, * */ -static struct ata_port * ata_host_add(struct ata_probe_ent *ent, +static struct ata_port * ata_host_add(const struct ata_probe_ent *ent, struct ata_host_set *host_set, unsigned int port_no) { @@ -4251,7 +4252,7 @@ err_out: * */ -int ata_device_add(struct ata_probe_ent *ent) +int ata_device_add(const struct ata_probe_ent *ent) { unsigned int count = 0, i; struct device *dev = ent->dev; @@ -4470,7 +4471,7 @@ void ata_std_ports(struct ata_ioports *ioaddr) } static struct ata_probe_ent * -ata_probe_ent_alloc(struct device *dev, struct ata_port_info *port) +ata_probe_ent_alloc(struct device *dev, const struct ata_port_info *port) { struct ata_probe_ent *probe_ent; @@ -4570,7 +4571,6 @@ static struct ata_probe_ent *ata_pci_init_legacy_port(struct pci_dev *pdev, stru if (!probe_ent) return NULL; - probe_ent->legacy_mode = 1; probe_ent->n_ports = 1; probe_ent->hard_port_no = port_num; @@ -4783,7 +4783,7 @@ void ata_pci_remove_one (struct pci_dev *pdev) } /* move to PCI subsystem */ -int pci_test_config_bits(struct pci_dev *pdev, struct pci_bits *bits) +int pci_test_config_bits(struct pci_dev *pdev, const struct pci_bits *bits) { unsigned long tmp = 0; diff --git a/drivers/scsi/libata-scsi.c b/drivers/scsi/libata-scsi.c index 90bf22204668..58858886d751 100644 --- a/drivers/scsi/libata-scsi.c +++ b/drivers/scsi/libata-scsi.c @@ -44,9 +44,9 @@ #include "libata.h" -typedef unsigned int (*ata_xlat_func_t)(struct ata_queued_cmd *qc, u8 *scsicmd); +typedef unsigned int (*ata_xlat_func_t)(struct ata_queued_cmd *qc, const u8 *scsicmd); static struct ata_device * -ata_scsi_find_dev(struct ata_port *ap, struct scsi_device *scsidev); +ata_scsi_find_dev(struct ata_port *ap, const struct scsi_device *scsidev); static void ata_scsi_invalid_field(struct scsi_cmnd *cmd, @@ -418,7 +418,7 @@ int ata_scsi_error(struct Scsi_Host *host) */ static unsigned int ata_scsi_start_stop_xlat(struct ata_queued_cmd *qc, - u8 *scsicmd) + const u8 *scsicmd) { struct ata_taskfile *tf = &qc->tf; @@ -485,7 +485,7 @@ invalid_fld: * Zero on success, non-zero on error. */ -static unsigned int ata_scsi_flush_xlat(struct ata_queued_cmd *qc, u8 *scsicmd) +static unsigned int ata_scsi_flush_xlat(struct ata_queued_cmd *qc, const u8 *scsicmd) { struct ata_taskfile *tf = &qc->tf; @@ -512,7 +512,7 @@ static unsigned int ata_scsi_flush_xlat(struct ata_queued_cmd *qc, u8 *scsicmd) * @plen: the transfer length */ -static void scsi_6_lba_len(u8 *scsicmd, u64 *plba, u32 *plen) +static void scsi_6_lba_len(const u8 *scsicmd, u64 *plba, u32 *plen) { u64 lba = 0; u32 len = 0; @@ -539,7 +539,7 @@ static void scsi_6_lba_len(u8 *scsicmd, u64 *plba, u32 *plen) * @plen: the transfer length */ -static void scsi_10_lba_len(u8 *scsicmd, u64 *plba, u32 *plen) +static void scsi_10_lba_len(const u8 *scsicmd, u64 *plba, u32 *plen) { u64 lba = 0; u32 len = 0; @@ -569,7 +569,7 @@ static void scsi_10_lba_len(u8 *scsicmd, u64 *plba, u32 *plen) * @plen: the transfer length */ -static void scsi_16_lba_len(u8 *scsicmd, u64 *plba, u32 *plen) +static void scsi_16_lba_len(const u8 *scsicmd, u64 *plba, u32 *plen) { u64 lba = 0; u32 len = 0; @@ -608,7 +608,7 @@ static void scsi_16_lba_len(u8 *scsicmd, u64 *plba, u32 *plen) * Zero on success, non-zero on error. */ -static unsigned int ata_scsi_verify_xlat(struct ata_queued_cmd *qc, u8 *scsicmd) +static unsigned int ata_scsi_verify_xlat(struct ata_queued_cmd *qc, const u8 *scsicmd) { struct ata_taskfile *tf = &qc->tf; struct ata_device *dev = qc->dev; @@ -734,7 +734,7 @@ nothing_to_do: * Zero on success, non-zero on error. */ -static unsigned int ata_scsi_rw_xlat(struct ata_queued_cmd *qc, u8 *scsicmd) +static unsigned int ata_scsi_rw_xlat(struct ata_queued_cmd *qc, const u8 *scsicmd) { struct ata_taskfile *tf = &qc->tf; struct ata_device *dev = qc->dev; @@ -1688,7 +1688,7 @@ static int atapi_qc_complete(struct ata_queued_cmd *qc, u8 drv_stat) * Zero on success, non-zero on failure. */ -static unsigned int atapi_xlat(struct ata_queued_cmd *qc, u8 *scsicmd) +static unsigned int atapi_xlat(struct ata_queued_cmd *qc, const u8 *scsicmd) { struct scsi_cmnd *cmd = qc->scsicmd; struct ata_device *dev = qc->dev; @@ -1757,7 +1757,7 @@ static unsigned int atapi_xlat(struct ata_queued_cmd *qc, u8 *scsicmd) */ static struct ata_device * -ata_scsi_find_dev(struct ata_port *ap, struct scsi_device *scsidev) +ata_scsi_find_dev(struct ata_port *ap, const struct scsi_device *scsidev) { struct ata_device *dev; @@ -1914,7 +1914,7 @@ void ata_scsi_simulate(u16 *id, void (*done)(struct scsi_cmnd *)) { struct ata_scsi_args args; - u8 *scsicmd = cmd->cmnd; + const u8 *scsicmd = cmd->cmnd; args.id = id; args.cmd = cmd; diff --git a/drivers/scsi/libata.h b/drivers/scsi/libata.h index 67d752ca8ae2..3d60190584ba 100644 --- a/drivers/scsi/libata.h +++ b/drivers/scsi/libata.h @@ -48,7 +48,7 @@ extern int ata_qc_issue(struct ata_queued_cmd *qc); extern int ata_check_atapi_dma(struct ata_queued_cmd *qc); extern void ata_dev_select(struct ata_port *ap, unsigned int device, unsigned int wait, unsigned int can_sleep); -extern void ata_tf_to_host_nolock(struct ata_port *ap, struct ata_taskfile *tf); +extern void ata_tf_to_host_nolock(struct ata_port *ap, const struct ata_taskfile *tf); extern void swap_buf_le16(u16 *buf, unsigned int buf_words); diff --git a/drivers/scsi/pdc_adma.c b/drivers/scsi/pdc_adma.c index 53b8db4be1a9..9820f272f889 100644 --- a/drivers/scsi/pdc_adma.c +++ b/drivers/scsi/pdc_adma.c @@ -158,7 +158,7 @@ static Scsi_Host_Template adma_ata_sht = { .bios_param = ata_std_bios_param, }; -static struct ata_port_operations adma_ata_ops = { +static const struct ata_port_operations adma_ata_ops = { .port_disable = ata_port_disable, .tf_load = ata_tf_load, .tf_read = ata_tf_read, diff --git a/drivers/scsi/sata_mv.c b/drivers/scsi/sata_mv.c index 9b6213928f7a..422e0b6f603a 100644 --- a/drivers/scsi/sata_mv.c +++ b/drivers/scsi/sata_mv.c @@ -290,7 +290,7 @@ static Scsi_Host_Template mv_sht = { .ordered_flush = 1, }; -static struct ata_port_operations mv_ops = { +static const struct ata_port_operations mv_ops = { .port_disable = ata_port_disable, .tf_load = ata_tf_load, diff --git a/drivers/scsi/sata_nv.c b/drivers/scsi/sata_nv.c index 8866530bc491..1a56d6c79ddd 100644 --- a/drivers/scsi/sata_nv.c +++ b/drivers/scsi/sata_nv.c @@ -238,7 +238,7 @@ static Scsi_Host_Template nv_sht = { .ordered_flush = 1, }; -static struct ata_port_operations nv_ops = { +static const struct ata_port_operations nv_ops = { .port_disable = ata_port_disable, .tf_load = ata_tf_load, .tf_read = ata_tf_read, diff --git a/drivers/scsi/sata_promise.c b/drivers/scsi/sata_promise.c index 9bf8cbd29901..eee93b0016df 100644 --- a/drivers/scsi/sata_promise.c +++ b/drivers/scsi/sata_promise.c @@ -87,8 +87,8 @@ static void pdc_port_stop(struct ata_port *ap); static void pdc_pata_phy_reset(struct ata_port *ap); static void pdc_sata_phy_reset(struct ata_port *ap); static void pdc_qc_prep(struct ata_queued_cmd *qc); -static void pdc_tf_load_mmio(struct ata_port *ap, struct ata_taskfile *tf); -static void pdc_exec_command_mmio(struct ata_port *ap, struct ata_taskfile *tf); +static void pdc_tf_load_mmio(struct ata_port *ap, const struct ata_taskfile *tf); +static void pdc_exec_command_mmio(struct ata_port *ap, const struct ata_taskfile *tf); static void pdc_irq_clear(struct ata_port *ap); static int pdc_qc_issue_prot(struct ata_queued_cmd *qc); @@ -113,7 +113,7 @@ static Scsi_Host_Template pdc_ata_sht = { .ordered_flush = 1, }; -static struct ata_port_operations pdc_sata_ops = { +static const struct ata_port_operations pdc_sata_ops = { .port_disable = ata_port_disable, .tf_load = pdc_tf_load_mmio, .tf_read = ata_tf_read, @@ -136,7 +136,7 @@ static struct ata_port_operations pdc_sata_ops = { .host_stop = ata_pci_host_stop, }; -static struct ata_port_operations pdc_pata_ops = { +static const struct ata_port_operations pdc_pata_ops = { .port_disable = ata_port_disable, .tf_load = pdc_tf_load_mmio, .tf_read = ata_tf_read, @@ -546,7 +546,7 @@ static int pdc_qc_issue_prot(struct ata_queued_cmd *qc) return ata_qc_issue_prot(qc); } -static void pdc_tf_load_mmio(struct ata_port *ap, struct ata_taskfile *tf) +static void pdc_tf_load_mmio(struct ata_port *ap, const struct ata_taskfile *tf) { WARN_ON (tf->protocol == ATA_PROT_DMA || tf->protocol == ATA_PROT_NODATA); @@ -554,7 +554,7 @@ static void pdc_tf_load_mmio(struct ata_port *ap, struct ata_taskfile *tf) } -static void pdc_exec_command_mmio(struct ata_port *ap, struct ata_taskfile *tf) +static void pdc_exec_command_mmio(struct ata_port *ap, const struct ata_taskfile *tf) { WARN_ON (tf->protocol == ATA_PROT_DMA || tf->protocol == ATA_PROT_NODATA); diff --git a/drivers/scsi/sata_qstor.c b/drivers/scsi/sata_qstor.c index e1c1dae27c52..250dafa6bc36 100644 --- a/drivers/scsi/sata_qstor.c +++ b/drivers/scsi/sata_qstor.c @@ -147,7 +147,7 @@ static Scsi_Host_Template qs_ata_sht = { .bios_param = ata_std_bios_param, }; -static struct ata_port_operations qs_ata_ops = { +static const struct ata_port_operations qs_ata_ops = { .port_disable = ata_port_disable, .tf_load = ata_tf_load, .tf_read = ata_tf_read, diff --git a/drivers/scsi/sata_sil.c b/drivers/scsi/sata_sil.c index f6f0184e1ac8..3a056173fb95 100644 --- a/drivers/scsi/sata_sil.c +++ b/drivers/scsi/sata_sil.c @@ -150,7 +150,7 @@ static Scsi_Host_Template sil_sht = { .ordered_flush = 1, }; -static struct ata_port_operations sil_ops = { +static const struct ata_port_operations sil_ops = { .port_disable = ata_port_disable, .dev_config = sil_dev_config, .tf_load = ata_tf_load, diff --git a/drivers/scsi/sata_sil24.c b/drivers/scsi/sata_sil24.c index 19857814d69f..32d730bd5bb6 100644 --- a/drivers/scsi/sata_sil24.c +++ b/drivers/scsi/sata_sil24.c @@ -275,7 +275,7 @@ static Scsi_Host_Template sil24_sht = { .ordered_flush = 1, /* NCQ not supported yet */ }; -static struct ata_port_operations sil24_ops = { +static const struct ata_port_operations sil24_ops = { .port_disable = ata_port_disable, .check_status = sil24_check_status, diff --git a/drivers/scsi/sata_sis.c b/drivers/scsi/sata_sis.c index 0761a3234fcf..057f7b98b6c4 100644 --- a/drivers/scsi/sata_sis.c +++ b/drivers/scsi/sata_sis.c @@ -102,7 +102,7 @@ static Scsi_Host_Template sis_sht = { .ordered_flush = 1, }; -static struct ata_port_operations sis_ops = { +static const struct ata_port_operations sis_ops = { .port_disable = ata_port_disable, .tf_load = ata_tf_load, .tf_read = ata_tf_read, diff --git a/drivers/scsi/sata_svw.c b/drivers/scsi/sata_svw.c index d89d968bedac..e0f9570bc6dd 100644 --- a/drivers/scsi/sata_svw.c +++ b/drivers/scsi/sata_svw.c @@ -102,7 +102,7 @@ static void k2_sata_scr_write (struct ata_port *ap, unsigned int sc_reg, } -static void k2_sata_tf_load(struct ata_port *ap, struct ata_taskfile *tf) +static void k2_sata_tf_load(struct ata_port *ap, const struct ata_taskfile *tf) { struct ata_ioports *ioaddr = &ap->ioaddr; unsigned int is_addr = tf->flags & ATA_TFLAG_ISADDR; @@ -297,7 +297,7 @@ static Scsi_Host_Template k2_sata_sht = { }; -static struct ata_port_operations k2_sata_ops = { +static const struct ata_port_operations k2_sata_ops = { .port_disable = ata_port_disable, .tf_load = k2_sata_tf_load, .tf_read = k2_sata_tf_read, diff --git a/drivers/scsi/sata_sx4.c b/drivers/scsi/sata_sx4.c index d6d350a0b5e9..af08f4f650c1 100644 --- a/drivers/scsi/sata_sx4.c +++ b/drivers/scsi/sata_sx4.c @@ -157,8 +157,8 @@ static void pdc_20621_phy_reset (struct ata_port *ap); static int pdc_port_start(struct ata_port *ap); static void pdc_port_stop(struct ata_port *ap); static void pdc20621_qc_prep(struct ata_queued_cmd *qc); -static void pdc_tf_load_mmio(struct ata_port *ap, struct ata_taskfile *tf); -static void pdc_exec_command_mmio(struct ata_port *ap, struct ata_taskfile *tf); +static void pdc_tf_load_mmio(struct ata_port *ap, const struct ata_taskfile *tf); +static void pdc_exec_command_mmio(struct ata_port *ap, const struct ata_taskfile *tf); static void pdc20621_host_stop(struct ata_host_set *host_set); static unsigned int pdc20621_dimm_init(struct ata_probe_ent *pe); static int pdc20621_detect_dimm(struct ata_probe_ent *pe); @@ -196,7 +196,7 @@ static Scsi_Host_Template pdc_sata_sht = { .ordered_flush = 1, }; -static struct ata_port_operations pdc_20621_ops = { +static const struct ata_port_operations pdc_20621_ops = { .port_disable = ata_port_disable, .tf_load = pdc_tf_load_mmio, .tf_read = ata_tf_read, @@ -899,7 +899,7 @@ out: DPRINTK("EXIT\n"); } -static void pdc_tf_load_mmio(struct ata_port *ap, struct ata_taskfile *tf) +static void pdc_tf_load_mmio(struct ata_port *ap, const struct ata_taskfile *tf) { WARN_ON (tf->protocol == ATA_PROT_DMA || tf->protocol == ATA_PROT_NODATA); @@ -907,7 +907,7 @@ static void pdc_tf_load_mmio(struct ata_port *ap, struct ata_taskfile *tf) } -static void pdc_exec_command_mmio(struct ata_port *ap, struct ata_taskfile *tf) +static void pdc_exec_command_mmio(struct ata_port *ap, const struct ata_taskfile *tf) { WARN_ON (tf->protocol == ATA_PROT_DMA || tf->protocol == ATA_PROT_NODATA); diff --git a/drivers/scsi/sata_uli.c b/drivers/scsi/sata_uli.c index 9c06f2abe7f7..d68dc7d3422c 100644 --- a/drivers/scsi/sata_uli.c +++ b/drivers/scsi/sata_uli.c @@ -90,7 +90,7 @@ static Scsi_Host_Template uli_sht = { .ordered_flush = 1, }; -static struct ata_port_operations uli_ops = { +static const struct ata_port_operations uli_ops = { .port_disable = ata_port_disable, .tf_load = ata_tf_load, diff --git a/drivers/scsi/sata_via.c b/drivers/scsi/sata_via.c index 565872479b9a..80e291a909a9 100644 --- a/drivers/scsi/sata_via.c +++ b/drivers/scsi/sata_via.c @@ -109,7 +109,7 @@ static Scsi_Host_Template svia_sht = { .ordered_flush = 1, }; -static struct ata_port_operations svia_sata_ops = { +static const struct ata_port_operations svia_sata_ops = { .port_disable = ata_port_disable, .tf_load = ata_tf_load, diff --git a/drivers/scsi/sata_vsc.c b/drivers/scsi/sata_vsc.c index 877b9fda3965..5af05fdf8544 100644 --- a/drivers/scsi/sata_vsc.c +++ b/drivers/scsi/sata_vsc.c @@ -115,7 +115,7 @@ static void vsc_intr_mask_update(struct ata_port *ap, u8 ctl) } -static void vsc_sata_tf_load(struct ata_port *ap, struct ata_taskfile *tf) +static void vsc_sata_tf_load(struct ata_port *ap, const struct ata_taskfile *tf) { struct ata_ioports *ioaddr = &ap->ioaddr; unsigned int is_addr = tf->flags & ATA_TFLAG_ISADDR; @@ -231,7 +231,7 @@ static Scsi_Host_Template vsc_sata_sht = { }; -static struct ata_port_operations vsc_sata_ops = { +static const struct ata_port_operations vsc_sata_ops = { .port_disable = ata_port_disable, .tf_load = vsc_sata_tf_load, .tf_read = vsc_sata_tf_read, diff --git a/include/linux/ata.h b/include/linux/ata.h index 33276d1d05d2..d2873b732bb1 100644 --- a/include/linux/ata.h +++ b/include/linux/ata.h @@ -261,7 +261,7 @@ struct ata_taskfile { ((u64) (id)[(n) + 1] << 16) | \ ((u64) (id)[(n) + 0]) ) -static inline int ata_id_current_chs_valid(u16 *id) +static inline int ata_id_current_chs_valid(const u16 *id) { /* For ATA-1 devices, if the INITIALIZE DEVICE PARAMETERS command has not been issued to the device then the values of @@ -273,7 +273,7 @@ static inline int ata_id_current_chs_valid(u16 *id) id[56]; /* sectors in current translation */ } -static inline int atapi_cdb_len(u16 *dev_id) +static inline int atapi_cdb_len(const u16 *dev_id) { u16 tmp = dev_id[0] & 0x3; switch (tmp) { @@ -283,7 +283,7 @@ static inline int atapi_cdb_len(u16 *dev_id) } } -static inline int is_atapi_taskfile(struct ata_taskfile *tf) +static inline int is_atapi_taskfile(const struct ata_taskfile *tf) { return (tf->protocol == ATA_PROT_ATAPI) || (tf->protocol == ATA_PROT_ATAPI_NODATA) || diff --git a/include/linux/libata.h b/include/linux/libata.h index 634b5aa0a615..00a8a5738858 100644 --- a/include/linux/libata.h +++ b/include/linux/libata.h @@ -202,7 +202,7 @@ struct ata_ioports { struct ata_probe_ent { struct list_head node; struct device *dev; - struct ata_port_operations *port_ops; + const struct ata_port_operations *port_ops; Scsi_Host_Template *sht; struct ata_ioports port[ATA_MAX_PORTS]; unsigned int n_ports; @@ -225,7 +225,7 @@ struct ata_host_set { void __iomem *mmio_base; unsigned int n_ports; void *private_data; - struct ata_port_operations *ops; + const struct ata_port_operations *ops; struct ata_port * ports[0]; }; @@ -294,7 +294,7 @@ struct ata_device { struct ata_port { struct Scsi_Host *host; /* our co-allocated scsi host */ - struct ata_port_operations *ops; + const struct ata_port_operations *ops; unsigned long flags; /* ATA_FLAG_xxx */ unsigned int id; /* unique id req'd by scsi midlyr */ unsigned int port_no; /* unique port #; from zero */ @@ -341,10 +341,10 @@ struct ata_port_operations { void (*set_piomode) (struct ata_port *, struct ata_device *); void (*set_dmamode) (struct ata_port *, struct ata_device *); - void (*tf_load) (struct ata_port *ap, struct ata_taskfile *tf); + void (*tf_load) (struct ata_port *ap, const struct ata_taskfile *tf); void (*tf_read) (struct ata_port *ap, struct ata_taskfile *tf); - void (*exec_command)(struct ata_port *ap, struct ata_taskfile *tf); + void (*exec_command)(struct ata_port *ap, const struct ata_taskfile *tf); u8 (*check_status)(struct ata_port *ap); u8 (*check_altstatus)(struct ata_port *ap); u8 (*check_err)(struct ata_port *ap); @@ -385,7 +385,7 @@ struct ata_port_info { unsigned long pio_mask; unsigned long mwdma_mask; unsigned long udma_mask; - struct ata_port_operations *port_ops; + const struct ata_port_operations *port_ops; }; struct ata_timing { @@ -413,7 +413,7 @@ extern int ata_pci_init_one (struct pci_dev *pdev, struct ata_port_info **port_i unsigned int n_ports); extern void ata_pci_remove_one (struct pci_dev *pdev); #endif /* CONFIG_PCI */ -extern int ata_device_add(struct ata_probe_ent *ent); +extern int ata_device_add(const struct ata_probe_ent *ent); extern void ata_host_set_remove(struct ata_host_set *host_set); extern int ata_scsi_detect(Scsi_Host_Template *sht); extern int ata_scsi_ioctl(struct scsi_device *dev, int cmd, void __user *arg); @@ -426,16 +426,16 @@ extern int ata_ratelimit(void); /* * Default driver ops implementations */ -extern void ata_tf_load(struct ata_port *ap, struct ata_taskfile *tf); +extern void ata_tf_load(struct ata_port *ap, const struct ata_taskfile *tf); extern void ata_tf_read(struct ata_port *ap, struct ata_taskfile *tf); -extern void ata_tf_to_fis(struct ata_taskfile *tf, u8 *fis, u8 pmp); -extern void ata_tf_from_fis(u8 *fis, struct ata_taskfile *tf); +extern void ata_tf_to_fis(const struct ata_taskfile *tf, u8 *fis, u8 pmp); +extern void ata_tf_from_fis(const u8 *fis, struct ata_taskfile *tf); extern void ata_noop_dev_select (struct ata_port *ap, unsigned int device); extern void ata_std_dev_select (struct ata_port *ap, unsigned int device); extern u8 ata_check_status(struct ata_port *ap); extern u8 ata_altstatus(struct ata_port *ap); extern u8 ata_chk_err(struct ata_port *ap); -extern void ata_exec_command(struct ata_port *ap, struct ata_taskfile *tf); +extern void ata_exec_command(struct ata_port *ap, const struct ata_taskfile *tf); extern int ata_port_start (struct ata_port *ap); extern void ata_port_stop (struct ata_port *ap); extern void ata_host_stop (struct ata_host_set *host_set); @@ -446,8 +446,8 @@ extern void ata_sg_init_one(struct ata_queued_cmd *qc, void *buf, unsigned int buflen); extern void ata_sg_init(struct ata_queued_cmd *qc, struct scatterlist *sg, unsigned int n_elem); -extern unsigned int ata_dev_classify(struct ata_taskfile *tf); -extern void ata_dev_id_string(u16 *id, unsigned char *s, +extern unsigned int ata_dev_classify(const struct ata_taskfile *tf); +extern void ata_dev_id_string(const u16 *id, unsigned char *s, unsigned int ofs, unsigned int len); extern void ata_dev_config(struct ata_port *ap, unsigned int i); extern void ata_bmdma_setup (struct ata_queued_cmd *qc); @@ -502,7 +502,7 @@ struct pci_bits { extern void ata_pci_host_stop (struct ata_host_set *host_set); extern struct ata_probe_ent * ata_pci_init_native_mode(struct pci_dev *pdev, struct ata_port_info **port, int portmask); -extern int pci_test_config_bits(struct pci_dev *pdev, struct pci_bits *bits); +extern int pci_test_config_bits(struct pci_dev *pdev, const struct pci_bits *bits); #endif /* CONFIG_PCI */ @@ -512,7 +512,7 @@ static inline unsigned int ata_tag_valid(unsigned int tag) return (tag < ATA_MAX_QUEUE) ? 1 : 0; } -static inline unsigned int ata_dev_present(struct ata_device *dev) +static inline unsigned int ata_dev_present(const struct ata_device *dev) { return ((dev->class == ATA_DEV_ATA) || (dev->class == ATA_DEV_ATAPI)); @@ -711,7 +711,7 @@ static inline unsigned int sata_dev_present(struct ata_port *ap) return ((scr_read(ap, SCR_STATUS) & 0xf) == 0x3) ? 1 : 0; } -static inline int ata_try_flush_cache(struct ata_device *dev) +static inline int ata_try_flush_cache(const struct ata_device *dev) { return ata_id_wcache_enabled(dev->id) || ata_id_has_flush(dev->id) || -- cgit v1.2.3 From c83c24861882758b9731e8550225cd1e52a4cd1c Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Tue, 18 Oct 2005 22:07:41 -0700 Subject: [SK_BUFF] kernel-doc: fix skbuff warnings Add kernel-doc to skbuff.h, skbuff.c to eliminate kernel-doc warnings. Signed-off-by: Randy Dunlap Signed-off-by: Arnaldo Carvalho de Melo --- include/linux/skbuff.h | 3 ++- net/core/skbuff.c | 2 ++ 2 files changed, 4 insertions(+), 1 deletion(-) (limited to 'include/linux') diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h index 8f5d9e7f8734..b756935da9c8 100644 --- a/include/linux/skbuff.h +++ b/include/linux/skbuff.h @@ -171,7 +171,6 @@ enum { * struct sk_buff - socket buffer * @next: Next buffer in list * @prev: Previous buffer in list - * @list: List we are on * @sk: Socket we are owned by * @tstamp: Time we arrived * @dev: Device we arrived on/are leaving by @@ -190,6 +189,7 @@ enum { * @cloned: Head may be cloned (check refcnt to be sure) * @nohdr: Payload reference only, must not modify header * @pkt_type: Packet class + * @fclone: skbuff clone status * @ip_summed: Driver fed us an IP checksum * @priority: Packet queueing priority * @users: User count - see {datagram,tcp}.c @@ -202,6 +202,7 @@ enum { * @destructor: Destruct function * @nfmark: Can be used for communication between hooks * @nfct: Associated connection, if any + * @ipvs_property: skbuff is owned by ipvs * @nfctinfo: Relationship of this skb to the connection * @nf_bridge: Saved data about a bridged frame - see br_netfilter.c * @tc_index: Traffic control index diff --git a/net/core/skbuff.c b/net/core/skbuff.c index 02cd4cde2112..ef9d46b91eb9 100644 --- a/net/core/skbuff.c +++ b/net/core/skbuff.c @@ -122,6 +122,8 @@ void skb_under_panic(struct sk_buff *skb, int sz, void *here) * __alloc_skb - allocate a network buffer * @size: size to allocate * @gfp_mask: allocation mask + * @fclone: allocate from fclone cache instead of head cache + * and allocate a cloned (child) skb * * Allocate a new &sk_buff. The returned buffer has no headroom and a * tail room of size bytes. The object has a reference count of one. -- cgit v1.2.3 From 0e574af1be5f569a5d7f2800333b0bfb358a5e34 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Thu, 27 Oct 2005 22:12:38 -0400 Subject: NFS: Cleanup initialisation of struct nfs_fattr Signed-off-by: Trond Myklebust --- fs/nfs/dir.c | 1 + fs/nfs/nfs3proc.c | 58 +++++++++++++++++++++++++------------------------- fs/nfs/nfs4proc.c | 43 +++++++++++++++++-------------------- fs/nfs/proc.c | 26 +++++++++++----------- fs/nfs/read.c | 1 + fs/nfs/write.c | 2 ++ include/linux/nfs_fs.h | 5 +++++ 7 files changed, 71 insertions(+), 65 deletions(-) (limited to 'include/linux') diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c index eb50c19fc253..b8a73045e9a0 100644 --- a/fs/nfs/dir.c +++ b/fs/nfs/dir.c @@ -532,6 +532,7 @@ static int nfs_readdir(struct file *filp, void *dirent, filldir_t filldir) my_entry.eof = 0; my_entry.fh = &fh; my_entry.fattr = &fattr; + nfs_fattr_init(&fattr); desc->entry = &my_entry; while(!desc->entry->eof) { diff --git a/fs/nfs/nfs3proc.c b/fs/nfs/nfs3proc.c index e4a1cd48195e..4b1b48b139f6 100644 --- a/fs/nfs/nfs3proc.c +++ b/fs/nfs/nfs3proc.c @@ -78,7 +78,7 @@ nfs3_proc_get_root(struct nfs_server *server, struct nfs_fh *fhandle, int status; dprintk("%s: call fsinfo\n", __FUNCTION__); - info->fattr->valid = 0; + nfs_fattr_init(info->fattr); status = rpc_call(server->client_sys, NFS3PROC_FSINFO, fhandle, info, 0); dprintk("%s: reply fsinfo: %d\n", __FUNCTION__, status); if (!(info->fattr->valid & NFS_ATTR_FATTR)) { @@ -98,7 +98,7 @@ nfs3_proc_getattr(struct nfs_server *server, struct nfs_fh *fhandle, int status; dprintk("NFS call getattr\n"); - fattr->valid = 0; + nfs_fattr_init(fattr); status = rpc_call(server->client, NFS3PROC_GETATTR, fhandle, fattr, 0); dprintk("NFS reply getattr: %d\n", status); @@ -117,7 +117,7 @@ nfs3_proc_setattr(struct dentry *dentry, struct nfs_fattr *fattr, int status; dprintk("NFS call setattr\n"); - fattr->valid = 0; + nfs_fattr_init(fattr); status = rpc_call(NFS_CLIENT(inode), NFS3PROC_SETATTR, &arg, fattr, 0); if (status == 0) nfs_setattr_update_inode(inode, sattr); @@ -143,8 +143,8 @@ nfs3_proc_lookup(struct inode *dir, struct qstr *name, int status; dprintk("NFS call lookup %s\n", name->name); - dir_attr.valid = 0; - fattr->valid = 0; + nfs_fattr_init(&dir_attr); + nfs_fattr_init(fattr); status = rpc_call(NFS_CLIENT(dir), NFS3PROC_LOOKUP, &arg, &res, 0); if (status >= 0 && !(fattr->valid & NFS_ATTR_FATTR)) status = rpc_call(NFS_CLIENT(dir), NFS3PROC_GETATTR, @@ -174,7 +174,6 @@ static int nfs3_proc_access(struct inode *inode, struct nfs_access_entry *entry) int status; dprintk("NFS call access\n"); - fattr.valid = 0; if (mode & MAY_READ) arg.access |= NFS3_ACCESS_READ; @@ -189,6 +188,7 @@ static int nfs3_proc_access(struct inode *inode, struct nfs_access_entry *entry) if (mode & MAY_EXEC) arg.access |= NFS3_ACCESS_EXECUTE; } + nfs_fattr_init(&fattr); status = rpc_call_sync(NFS_CLIENT(inode), &msg, 0); nfs_refresh_inode(inode, &fattr); if (status == 0) { @@ -217,7 +217,7 @@ static int nfs3_proc_readlink(struct inode *inode, struct page *page, int status; dprintk("NFS call readlink\n"); - fattr.valid = 0; + nfs_fattr_init(&fattr); status = rpc_call(NFS_CLIENT(inode), NFS3PROC_READLINK, &args, &fattr, 0); nfs_refresh_inode(inode, &fattr); @@ -240,7 +240,7 @@ static int nfs3_proc_read(struct nfs_read_data *rdata) dprintk("NFS call read %d @ %Ld\n", rdata->args.count, (long long) rdata->args.offset); - fattr->valid = 0; + nfs_fattr_init(fattr); status = rpc_call_sync(NFS_CLIENT(inode), &msg, flags); if (status >= 0) nfs_refresh_inode(inode, fattr); @@ -263,7 +263,7 @@ static int nfs3_proc_write(struct nfs_write_data *wdata) dprintk("NFS call write %d @ %Ld\n", wdata->args.count, (long long) wdata->args.offset); - fattr->valid = 0; + nfs_fattr_init(fattr); status = rpc_call_sync(NFS_CLIENT(inode), &msg, rpcflags); if (status >= 0) nfs_refresh_inode(inode, fattr); @@ -285,7 +285,7 @@ static int nfs3_proc_commit(struct nfs_write_data *cdata) dprintk("NFS call commit %d @ %Ld\n", cdata->args.count, (long long) cdata->args.offset); - fattr->valid = 0; + nfs_fattr_init(fattr); status = rpc_call_sync(NFS_CLIENT(inode), &msg, 0); if (status >= 0) nfs_refresh_inode(inode, fattr); @@ -329,8 +329,8 @@ nfs3_proc_create(struct inode *dir, struct dentry *dentry, struct iattr *sattr, sattr->ia_mode &= ~current->fs->umask; again: - dir_attr.valid = 0; - fattr.valid = 0; + nfs_fattr_init(&dir_attr); + nfs_fattr_init(&fattr); status = rpc_call(NFS_CLIENT(dir), NFS3PROC_CREATE, &arg, &res, 0); nfs_refresh_inode(dir, &dir_attr); @@ -401,7 +401,7 @@ nfs3_proc_remove(struct inode *dir, struct qstr *name) int status; dprintk("NFS call remove %s\n", name->name); - dir_attr.valid = 0; + nfs_fattr_init(&dir_attr); status = rpc_call_sync(NFS_CLIENT(dir), &msg, 0); nfs_refresh_inode(dir, &dir_attr); dprintk("NFS reply remove: %d\n", status); @@ -422,7 +422,7 @@ nfs3_proc_unlink_setup(struct rpc_message *msg, struct dentry *dir, struct qstr ptr->arg.fh = NFS_FH(dir->d_inode); ptr->arg.name = name->name; ptr->arg.len = name->len; - ptr->res.valid = 0; + nfs_fattr_init(&ptr->res); msg->rpc_proc = &nfs3_procedures[NFS3PROC_REMOVE]; msg->rpc_argp = &ptr->arg; msg->rpc_resp = &ptr->res; @@ -465,8 +465,8 @@ nfs3_proc_rename(struct inode *old_dir, struct qstr *old_name, int status; dprintk("NFS call rename %s -> %s\n", old_name->name, new_name->name); - old_dir_attr.valid = 0; - new_dir_attr.valid = 0; + nfs_fattr_init(&old_dir_attr); + nfs_fattr_init(&new_dir_attr); status = rpc_call(NFS_CLIENT(old_dir), NFS3PROC_RENAME, &arg, &res, 0); nfs_refresh_inode(old_dir, &old_dir_attr); nfs_refresh_inode(new_dir, &new_dir_attr); @@ -491,8 +491,8 @@ nfs3_proc_link(struct inode *inode, struct inode *dir, struct qstr *name) int status; dprintk("NFS call link %s\n", name->name); - dir_attr.valid = 0; - fattr.valid = 0; + nfs_fattr_init(&dir_attr); + nfs_fattr_init(&fattr); status = rpc_call(NFS_CLIENT(inode), NFS3PROC_LINK, &arg, &res, 0); nfs_refresh_inode(dir, &dir_attr); nfs_refresh_inode(inode, &fattr); @@ -524,8 +524,8 @@ nfs3_proc_symlink(struct inode *dir, struct qstr *name, struct qstr *path, if (path->len > NFS3_MAXPATHLEN) return -ENAMETOOLONG; dprintk("NFS call symlink %s -> %s\n", name->name, path->name); - dir_attr.valid = 0; - fattr->valid = 0; + nfs_fattr_init(&dir_attr); + nfs_fattr_init(fattr); status = rpc_call(NFS_CLIENT(dir), NFS3PROC_SYMLINK, &arg, &res, 0); nfs_refresh_inode(dir, &dir_attr); dprintk("NFS reply symlink: %d\n", status); @@ -552,11 +552,11 @@ nfs3_proc_mkdir(struct inode *dir, struct dentry *dentry, struct iattr *sattr) int status; dprintk("NFS call mkdir %s\n", dentry->d_name.name); - dir_attr.valid = 0; - fattr.valid = 0; sattr->ia_mode &= ~current->fs->umask; + nfs_fattr_init(&dir_attr); + nfs_fattr_init(&fattr); status = rpc_call(NFS_CLIENT(dir), NFS3PROC_MKDIR, &arg, &res, 0); nfs_refresh_inode(dir, &dir_attr); if (status != 0) @@ -582,7 +582,7 @@ nfs3_proc_rmdir(struct inode *dir, struct qstr *name) int status; dprintk("NFS call rmdir %s\n", name->name); - dir_attr.valid = 0; + nfs_fattr_init(&dir_attr); status = rpc_call(NFS_CLIENT(dir), NFS3PROC_RMDIR, &arg, &dir_attr, 0); nfs_refresh_inode(dir, &dir_attr); dprintk("NFS reply rmdir: %d\n", status); @@ -634,7 +634,7 @@ nfs3_proc_readdir(struct dentry *dentry, struct rpc_cred *cred, dprintk("NFS call readdir%s %d\n", plus? "plus" : "", (unsigned int) cookie); - dir_attr.valid = 0; + nfs_fattr_init(&dir_attr); status = rpc_call_sync(NFS_CLIENT(dir), &msg, 0); nfs_refresh_inode(dir, &dir_attr); dprintk("NFS reply readdir: %d\n", status); @@ -676,8 +676,8 @@ nfs3_proc_mknod(struct inode *dir, struct dentry *dentry, struct iattr *sattr, sattr->ia_mode &= ~current->fs->umask; - dir_attr.valid = 0; - fattr.valid = 0; + nfs_fattr_init(&dir_attr); + nfs_fattr_init(&fattr); status = rpc_call(NFS_CLIENT(dir), NFS3PROC_MKNOD, &arg, &res, 0); nfs_refresh_inode(dir, &dir_attr); if (status != 0) @@ -698,7 +698,7 @@ nfs3_proc_statfs(struct nfs_server *server, struct nfs_fh *fhandle, int status; dprintk("NFS call fsstat\n"); - stat->fattr->valid = 0; + nfs_fattr_init(stat->fattr); status = rpc_call(server->client, NFS3PROC_FSSTAT, fhandle, stat, 0); dprintk("NFS reply statfs: %d\n", status); return status; @@ -711,7 +711,7 @@ nfs3_proc_fsinfo(struct nfs_server *server, struct nfs_fh *fhandle, int status; dprintk("NFS call fsinfo\n"); - info->fattr->valid = 0; + nfs_fattr_init(info->fattr); status = rpc_call(server->client_sys, NFS3PROC_FSINFO, fhandle, info, 0); dprintk("NFS reply fsinfo: %d\n", status); return status; @@ -724,7 +724,7 @@ nfs3_proc_pathconf(struct nfs_server *server, struct nfs_fh *fhandle, int status; dprintk("NFS call pathconf\n"); - info->fattr->valid = 0; + nfs_fattr_init(info->fattr); status = rpc_call(server->client, NFS3PROC_PATHCONF, fhandle, info, 0); dprintk("NFS reply pathconf: %d\n", status); return status; diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 9c1da34036aa..2a759e8e387c 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -494,9 +494,7 @@ static int _nfs4_open_expired(struct nfs4_state_owner *sp, struct nfs4_state *st struct inode *inode = state->inode; struct nfs_server *server = NFS_SERVER(dir); struct nfs_delegation *delegation = NFS_I(inode)->delegation; - struct nfs_fattr f_attr = { - .valid = 0, - }; + struct nfs_fattr f_attr; struct nfs_openargs o_arg = { .fh = NFS_FH(dir), .open_flags = state->state, @@ -522,6 +520,7 @@ static int _nfs4_open_expired(struct nfs4_state_owner *sp, struct nfs4_state *st status = -ENOMEM; if (o_arg.seqid == NULL) goto out; + nfs_fattr_init(&f_attr); status = _nfs4_proc_open(dir, sp, &o_arg, &o_res); if (status != 0) goto out_nodeleg; @@ -692,9 +691,7 @@ static int _nfs4_do_open(struct inode *dir, struct dentry *dentry, int flags, st struct nfs4_client *clp = server->nfs4_state; struct inode *inode = NULL; int status; - struct nfs_fattr f_attr = { - .valid = 0, - }; + struct nfs_fattr f_attr; struct nfs_openargs o_arg = { .fh = NFS_FH(dir), .open_flags = flags, @@ -726,6 +723,7 @@ static int _nfs4_do_open(struct inode *dir, struct dentry *dentry, int flags, st o_arg.seqid = nfs_alloc_seqid(&sp->so_seqid); if (o_arg.seqid == NULL) return -ENOMEM; + nfs_fattr_init(&f_attr); status = _nfs4_proc_open(dir, sp, &o_arg, &o_res); if (status != 0) goto out_err; @@ -824,7 +822,7 @@ static int _nfs4_do_setattr(struct nfs_server *server, struct nfs_fattr *fattr, }; int status; - fattr->valid = 0; + nfs_fattr_init(fattr); if (state != NULL) { msg.rpc_cred = state->owner->so_cred; @@ -1107,13 +1105,12 @@ static int nfs4_server_capabilities(struct nfs_server *server, struct nfs_fh *fh static int _nfs4_lookup_root(struct nfs_server *server, struct nfs_fh *fhandle, struct nfs_fsinfo *info) { - struct nfs_fattr * fattr = info->fattr; struct nfs4_lookup_root_arg args = { .bitmask = nfs4_fattr_bitmap, }; struct nfs4_lookup_res res = { .server = server, - .fattr = fattr, + .fattr = info->fattr, .fh = fhandle, }; struct rpc_message msg = { @@ -1121,7 +1118,7 @@ static int _nfs4_lookup_root(struct nfs_server *server, struct nfs_fh *fhandle, .rpc_argp = &args, .rpc_resp = &res, }; - fattr->valid = 0; + nfs_fattr_init(info->fattr); return rpc_call_sync(server->client, &msg, 0); } @@ -1184,7 +1181,7 @@ static int nfs4_proc_get_root(struct nfs_server *server, struct nfs_fh *fhandle, q.len = p - q.name; do { - fattr->valid = 0; + nfs_fattr_init(fattr); status = nfs4_handle_exception(server, rpc_call_sync(server->client, &msg, 0), &exception); @@ -1221,7 +1218,7 @@ static int _nfs4_proc_getattr(struct nfs_server *server, struct nfs_fh *fhandle, .rpc_resp = &res, }; - fattr->valid = 0; + nfs_fattr_init(fattr); return rpc_call_sync(server->client, &msg, 0); } @@ -1263,7 +1260,7 @@ nfs4_proc_setattr(struct dentry *dentry, struct nfs_fattr *fattr, struct nfs4_state *state; int status; - fattr->valid = 0; + nfs_fattr_init(fattr); cred = rpcauth_lookupcred(NFS_SERVER(inode)->client->cl_auth, 0); if (IS_ERR(cred)) @@ -1309,7 +1306,7 @@ static int _nfs4_proc_lookup(struct inode *dir, struct qstr *name, .rpc_resp = &res, }; - fattr->valid = 0; + nfs_fattr_init(fattr); dprintk("NFS call lookup %s\n", name->name); status = rpc_call_sync(NFS_CLIENT(dir), &msg, 0); @@ -1458,7 +1455,7 @@ static int _nfs4_proc_read(struct nfs_read_data *rdata) dprintk("NFS call read %d @ %Ld\n", rdata->args.count, (long long) rdata->args.offset); - fattr->valid = 0; + nfs_fattr_init(fattr); status = rpc_call_sync(server->client, &msg, flags); if (!status) renew_lease(server, timestamp); @@ -1495,7 +1492,7 @@ static int _nfs4_proc_write(struct nfs_write_data *wdata) dprintk("NFS call write %d @ %Ld\n", wdata->args.count, (long long) wdata->args.offset); - fattr->valid = 0; + nfs_fattr_init(fattr); status = rpc_call_sync(server->client, &msg, rpcflags); dprintk("NFS reply write: %d\n", status); return status; @@ -1529,7 +1526,7 @@ static int _nfs4_proc_commit(struct nfs_write_data *cdata) dprintk("NFS call commit %d @ %Ld\n", cdata->args.count, (long long) cdata->args.offset); - fattr->valid = 0; + nfs_fattr_init(fattr); status = rpc_call_sync(server->client, &msg, 0); dprintk("NFS reply commit: %d\n", status); return status; @@ -1769,7 +1766,7 @@ static int _nfs4_proc_symlink(struct inode *dir, struct qstr *name, if (path->len > NFS4_MAXPATHLEN) return -ENAMETOOLONG; arg.u.symlink = path; - fattr->valid = 0; + nfs_fattr_init(fattr); status = rpc_call_sync(NFS_CLIENT(dir), &msg, 0); if (!status) @@ -1818,7 +1815,7 @@ static int _nfs4_proc_mkdir(struct inode *dir, struct dentry *dentry, }; int status; - fattr.valid = 0; + nfs_fattr_init(&fattr); status = rpc_call_sync(NFS_CLIENT(dir), &msg, 0); if (!status) { @@ -1916,7 +1913,7 @@ static int _nfs4_proc_mknod(struct inode *dir, struct dentry *dentry, int status; int mode = sattr->ia_mode; - fattr.valid = 0; + nfs_fattr_init(&fattr); BUG_ON(!(sattr->ia_valid & ATTR_MODE)); BUG_ON(!S_ISFIFO(mode) && !S_ISBLK(mode) && !S_ISCHR(mode) && !S_ISSOCK(mode)); @@ -1969,7 +1966,7 @@ static int _nfs4_proc_statfs(struct nfs_server *server, struct nfs_fh *fhandle, .rpc_resp = fsstat, }; - fsstat->fattr->valid = 0; + nfs_fattr_init(fsstat->fattr); return rpc_call_sync(server->client, &msg, 0); } @@ -2016,7 +2013,7 @@ static int nfs4_do_fsinfo(struct nfs_server *server, struct nfs_fh *fhandle, str static int nfs4_proc_fsinfo(struct nfs_server *server, struct nfs_fh *fhandle, struct nfs_fsinfo *fsinfo) { - fsinfo->fattr->valid = 0; + nfs_fattr_init(fsinfo->fattr); return nfs4_do_fsinfo(server, fhandle, fsinfo); } @@ -2039,7 +2036,7 @@ static int _nfs4_proc_pathconf(struct nfs_server *server, struct nfs_fh *fhandle return 0; } - pathconf->fattr->valid = 0; + nfs_fattr_init(pathconf->fattr); return rpc_call_sync(server->client, &msg, 0); } diff --git a/fs/nfs/proc.c b/fs/nfs/proc.c index 8fef86523d7f..5ef28f08f424 100644 --- a/fs/nfs/proc.c +++ b/fs/nfs/proc.c @@ -61,7 +61,7 @@ nfs_proc_get_root(struct nfs_server *server, struct nfs_fh *fhandle, int status; dprintk("%s: call getattr\n", __FUNCTION__); - fattr->valid = 0; + nfs_fattr_init(fattr); status = rpc_call(server->client_sys, NFSPROC_GETATTR, fhandle, fattr, 0); dprintk("%s: reply getattr: %d\n", __FUNCTION__, status); if (status) @@ -93,7 +93,7 @@ nfs_proc_getattr(struct nfs_server *server, struct nfs_fh *fhandle, int status; dprintk("NFS call getattr\n"); - fattr->valid = 0; + nfs_fattr_init(fattr); status = rpc_call(server->client, NFSPROC_GETATTR, fhandle, fattr, 0); dprintk("NFS reply getattr: %d\n", status); @@ -112,7 +112,7 @@ nfs_proc_setattr(struct dentry *dentry, struct nfs_fattr *fattr, int status; dprintk("NFS call setattr\n"); - fattr->valid = 0; + nfs_fattr_init(fattr); status = rpc_call(NFS_CLIENT(inode), NFSPROC_SETATTR, &arg, fattr, 0); if (status == 0) nfs_setattr_update_inode(inode, sattr); @@ -136,7 +136,7 @@ nfs_proc_lookup(struct inode *dir, struct qstr *name, int status; dprintk("NFS call lookup %s\n", name->name); - fattr->valid = 0; + nfs_fattr_init(fattr); status = rpc_call(NFS_CLIENT(dir), NFSPROC_LOOKUP, &arg, &res, 0); dprintk("NFS reply lookup: %d\n", status); return status; @@ -174,7 +174,7 @@ static int nfs_proc_read(struct nfs_read_data *rdata) dprintk("NFS call read %d @ %Ld\n", rdata->args.count, (long long) rdata->args.offset); - fattr->valid = 0; + nfs_fattr_init(fattr); status = rpc_call_sync(NFS_CLIENT(inode), &msg, flags); if (status >= 0) { nfs_refresh_inode(inode, fattr); @@ -203,7 +203,7 @@ static int nfs_proc_write(struct nfs_write_data *wdata) dprintk("NFS call write %d @ %Ld\n", wdata->args.count, (long long) wdata->args.offset); - fattr->valid = 0; + nfs_fattr_init(fattr); status = rpc_call_sync(NFS_CLIENT(inode), &msg, flags); if (status >= 0) { nfs_refresh_inode(inode, fattr); @@ -232,7 +232,7 @@ nfs_proc_create(struct inode *dir, struct dentry *dentry, struct iattr *sattr, }; int status; - fattr.valid = 0; + nfs_fattr_init(&fattr); dprintk("NFS call create %s\n", dentry->d_name.name); status = rpc_call(NFS_CLIENT(dir), NFSPROC_CREATE, &arg, &res, 0); if (status == 0) @@ -273,12 +273,12 @@ nfs_proc_mknod(struct inode *dir, struct dentry *dentry, struct iattr *sattr, sattr->ia_size = new_encode_dev(rdev);/* get out your barf bag */ } - fattr.valid = 0; + nfs_fattr_init(&fattr); status = rpc_call(NFS_CLIENT(dir), NFSPROC_CREATE, &arg, &res, 0); if (status == -EINVAL && S_ISFIFO(mode)) { sattr->ia_mode = mode; - fattr.valid = 0; + nfs_fattr_init(&fattr); status = rpc_call(NFS_CLIENT(dir), NFSPROC_CREATE, &arg, &res, 0); } if (status == 0) @@ -391,7 +391,7 @@ nfs_proc_symlink(struct inode *dir, struct qstr *name, struct qstr *path, if (path->len > NFS2_MAXPATHLEN) return -ENAMETOOLONG; dprintk("NFS call symlink %s -> %s\n", name->name, path->name); - fattr->valid = 0; + nfs_fattr_init(fattr); fhandle->size = 0; status = rpc_call(NFS_CLIENT(dir), NFSPROC_SYMLINK, &arg, NULL, 0); dprintk("NFS reply symlink: %d\n", status); @@ -416,7 +416,7 @@ nfs_proc_mkdir(struct inode *dir, struct dentry *dentry, struct iattr *sattr) int status; dprintk("NFS call mkdir %s\n", dentry->d_name.name); - fattr.valid = 0; + nfs_fattr_init(&fattr); status = rpc_call(NFS_CLIENT(dir), NFSPROC_MKDIR, &arg, &res, 0); if (status == 0) status = nfs_instantiate(dentry, &fhandle, &fattr); @@ -484,7 +484,7 @@ nfs_proc_statfs(struct nfs_server *server, struct nfs_fh *fhandle, int status; dprintk("NFS call statfs\n"); - stat->fattr->valid = 0; + nfs_fattr_init(stat->fattr); status = rpc_call(server->client, NFSPROC_STATFS, fhandle, &fsinfo, 0); dprintk("NFS reply statfs: %d\n", status); if (status) @@ -507,7 +507,7 @@ nfs_proc_fsinfo(struct nfs_server *server, struct nfs_fh *fhandle, int status; dprintk("NFS call fsinfo\n"); - info->fattr->valid = 0; + nfs_fattr_init(info->fattr); status = rpc_call(server->client, NFSPROC_STATFS, fhandle, &fsinfo, 0); dprintk("NFS reply fsinfo: %d\n", status); if (status) diff --git a/fs/nfs/read.c b/fs/nfs/read.c index 9758ebd49905..43b03b19731b 100644 --- a/fs/nfs/read.c +++ b/fs/nfs/read.c @@ -215,6 +215,7 @@ static void nfs_read_rpcsetup(struct nfs_page *req, struct nfs_read_data *data, data->res.fattr = &data->fattr; data->res.count = count; data->res.eof = 0; + nfs_fattr_init(&data->fattr); NFS_PROTO(inode)->read_setup(data); diff --git a/fs/nfs/write.c b/fs/nfs/write.c index 5130eda231d7..819a65f5071f 100644 --- a/fs/nfs/write.c +++ b/fs/nfs/write.c @@ -870,6 +870,7 @@ static void nfs_write_rpcsetup(struct nfs_page *req, data->res.fattr = &data->fattr; data->res.count = count; data->res.verf = &data->verf; + nfs_fattr_init(&data->fattr); NFS_PROTO(inode)->write_setup(data, how); @@ -1237,6 +1238,7 @@ static void nfs_commit_rpcsetup(struct list_head *head, data->res.count = 0; data->res.fattr = &data->fattr; data->res.verf = &data->verf; + nfs_fattr_init(&data->fattr); NFS_PROTO(inode)->commit_setup(data, how); diff --git a/include/linux/nfs_fs.h b/include/linux/nfs_fs.h index 7bac2785c6e4..8120fd68dee5 100644 --- a/include/linux/nfs_fs.h +++ b/include/linux/nfs_fs.h @@ -316,6 +316,11 @@ extern void nfs_file_clear_open_context(struct file *filp); /* linux/net/ipv4/ipconfig.c: trims ip addr off front of name, too. */ extern u32 root_nfs_parse_addr(char *name); /*__init*/ +static inline void nfs_fattr_init(struct nfs_fattr *fattr) +{ + fattr->valid = 0; +} + /* * linux/fs/nfs/file.c */ -- cgit v1.2.3 From 913a70fc170530f7e1ff0693595155457cc6d0ca Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Thu, 27 Oct 2005 22:12:38 -0400 Subject: NFS: Convert cache_change_attribute into a jiffy-based value Signed-off-by: Trond Myklebust --- fs/nfs/inode.c | 8 ++++---- include/linux/nfs_fs.h | 2 +- 2 files changed, 5 insertions(+), 5 deletions(-) (limited to 'include/linux') diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c index 65d5ab45ddc5..449df8c8aa31 100644 --- a/fs/nfs/inode.c +++ b/fs/nfs/inode.c @@ -1135,7 +1135,7 @@ __nfs_revalidate_inode(struct nfs_server *server, struct inode *inode) * We may need to keep the attributes marked as invalid if * we raced with nfs_end_attr_update(). */ - if (verifier == nfsi->cache_change_attribute) + if (time_after_eq(verifier, nfsi->cache_change_attribute)) nfsi->cache_validity &= ~(NFS_INO_INVALID_ATTR|NFS_INO_INVALID_ATIME); spin_unlock(&inode->i_lock); @@ -1202,7 +1202,7 @@ void nfs_revalidate_mapping(struct inode *inode, struct address_space *mapping) if (S_ISDIR(inode->i_mode)) { memset(nfsi->cookieverf, 0, sizeof(nfsi->cookieverf)); /* This ensures we revalidate child dentries */ - nfsi->cache_change_attribute++; + nfsi->cache_change_attribute = jiffies; } spin_unlock(&inode->i_lock); @@ -1242,7 +1242,7 @@ void nfs_end_data_update(struct inode *inode) nfsi->cache_validity |= NFS_INO_INVALID_DATA; spin_unlock(&inode->i_lock); } - nfsi->cache_change_attribute ++; + nfsi->cache_change_attribute = jiffies; atomic_dec(&nfsi->data_updates); } @@ -1391,7 +1391,7 @@ static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr, unsign /* Do we perhaps have any outstanding writes? */ if (nfsi->npages == 0) { /* No, but did we race with nfs_end_data_update()? */ - if (verifier == nfsi->cache_change_attribute) { + if (time_after_eq(verifier, nfsi->cache_change_attribute)) { inode->i_size = new_isize; invalid |= NFS_INO_INVALID_DATA; } diff --git a/include/linux/nfs_fs.h b/include/linux/nfs_fs.h index 8120fd68dee5..abf890f5fbfb 100644 --- a/include/linux/nfs_fs.h +++ b/include/linux/nfs_fs.h @@ -280,7 +280,7 @@ static inline long nfs_save_change_attribute(struct inode *inode) static inline int nfs_verify_change_attribute(struct inode *inode, unsigned long chattr) { return !nfs_caches_unstable(inode) - && chattr == NFS_I(inode)->cache_change_attribute; + && time_after_eq(chattr, NFS_I(inode)->cache_change_attribute); } /* -- cgit v1.2.3 From 33801147a8fda6b04d7e9afe1d42f1c01d3d6837 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Thu, 27 Oct 2005 22:12:39 -0400 Subject: NFS: Optimise inode attribute cache updates Allow nfs_refresh_inode() also to update attributes on the inode if the RPC call was sent after the last call to nfs_update_inode(). Signed-off-by: Trond Myklebust --- fs/nfs/inode.c | 54 ++++++++++++++++++++++++++++++++++++------------- fs/nfs/nfs2xdr.c | 1 - fs/nfs/nfs3xdr.c | 1 - fs/nfs/nfs4xdr.c | 4 +--- include/linux/nfs_fs.h | 2 ++ include/linux/nfs_xdr.h | 2 +- 6 files changed, 44 insertions(+), 20 deletions(-) (limited to 'include/linux') diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c index 449df8c8aa31..b7d4f8f13ac2 100644 --- a/fs/nfs/inode.c +++ b/fs/nfs/inode.c @@ -785,7 +785,8 @@ nfs_fhget(struct super_block *sb, struct nfs_fh *fh, struct nfs_fattr *fattr) else init_special_inode(inode, inode->i_mode, fattr->rdev); - nfsi->read_cache_jiffies = fattr->timestamp; + nfsi->read_cache_jiffies = fattr->time_start; + nfsi->last_updated = jiffies; inode->i_atime = fattr->atime; inode->i_mtime = fattr->mtime; inode->i_ctime = fattr->ctime; @@ -1120,14 +1121,15 @@ __nfs_revalidate_inode(struct nfs_server *server, struct inode *inode) goto out; } + spin_lock(&inode->i_lock); status = nfs_update_inode(inode, &fattr, verifier); if (status) { + spin_unlock(&inode->i_lock); dfprintk(PAGECACHE, "nfs_revalidate_inode: (%s/%Ld) refresh failed, error=%d\n", inode->i_sb->s_id, (long long)NFS_FILEID(inode), status); goto out; } - spin_lock(&inode->i_lock); cache_validity = nfsi->cache_validity; nfsi->cache_validity &= ~NFS_INO_REVAL_PAGECACHE; @@ -1247,7 +1249,7 @@ void nfs_end_data_update(struct inode *inode) } /** - * nfs_refresh_inode - verify consistency of the inode attribute cache + * nfs_check_inode_attributes - verify consistency of the inode attribute cache * @inode - pointer to inode * @fattr - updated attributes * @@ -1255,13 +1257,12 @@ void nfs_end_data_update(struct inode *inode) * so that fattr carries weak cache consistency data, then it may * also update the ctime/mtime/change_attribute. */ -int nfs_refresh_inode(struct inode *inode, struct nfs_fattr *fattr) +static int nfs_check_inode_attributes(struct inode *inode, struct nfs_fattr *fattr) { struct nfs_inode *nfsi = NFS_I(inode); loff_t cur_size, new_isize; int data_unstable; - spin_lock(&inode->i_lock); /* Are we in the process of updating data on the server? */ data_unstable = nfs_caches_unstable(inode); @@ -1325,11 +1326,40 @@ int nfs_refresh_inode(struct inode *inode, struct nfs_fattr *fattr) if (!timespec_equal(&inode->i_atime, &fattr->atime)) nfsi->cache_validity |= NFS_INO_INVALID_ATIME; - nfsi->read_cache_jiffies = fattr->timestamp; - spin_unlock(&inode->i_lock); + nfsi->read_cache_jiffies = fattr->time_start; return 0; } +/** + * nfs_refresh_inode - try to update the inode attribute cache + * @inode - pointer to inode + * @fattr - updated attributes + * + * Check that an RPC call that returned attributes has not overlapped with + * other recent updates of the inode metadata, then decide whether it is + * safe to do a full update of the inode attributes, or whether just to + * call nfs_check_inode_attributes. + */ +int nfs_refresh_inode(struct inode *inode, struct nfs_fattr *fattr) +{ + struct nfs_inode *nfsi = NFS_I(inode); + int status; + + if ((fattr->valid & NFS_ATTR_FATTR) == 0) + return 0; + spin_lock(&inode->i_lock); + nfsi->cache_validity &= ~NFS_INO_REVAL_PAGECACHE; + if (nfs_verify_change_attribute(inode, fattr->time_start)) + nfsi->cache_validity &= ~(NFS_INO_INVALID_ATTR|NFS_INO_INVALID_ATIME); + if (time_after(fattr->time_start, nfsi->last_updated)) + status = nfs_update_inode(inode, fattr, fattr->time_start); + else + status = nfs_check_inode_attributes(inode, fattr); + + spin_unlock(&inode->i_lock); + return status; +} + /* * Many nfs protocol calls return the new file attributes after * an operation. Here we update the inode to reflect the state @@ -1365,20 +1395,17 @@ static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr, unsign goto out_err; } - spin_lock(&inode->i_lock); - /* * Make sure the inode's type hasn't changed. */ - if ((inode->i_mode & S_IFMT) != (fattr->mode & S_IFMT)) { - spin_unlock(&inode->i_lock); + if ((inode->i_mode & S_IFMT) != (fattr->mode & S_IFMT)) goto out_changed; - } /* * Update the read time so we don't revalidate too often. */ - nfsi->read_cache_jiffies = fattr->timestamp; + nfsi->read_cache_jiffies = fattr->time_start; + nfsi->last_updated = jiffies; /* Are we racing with known updates of the metadata on the server? */ data_unstable = ! (nfs_verify_change_attribute(inode, verifier) || @@ -1467,7 +1494,6 @@ static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr, unsign if (!nfs_have_delegation(inode, FMODE_READ)) nfsi->cache_validity |= invalid; - spin_unlock(&inode->i_lock); return 0; out_changed: /* diff --git a/fs/nfs/nfs2xdr.c b/fs/nfs/nfs2xdr.c index d91b69044a4d..59049e864ca7 100644 --- a/fs/nfs/nfs2xdr.c +++ b/fs/nfs/nfs2xdr.c @@ -143,7 +143,6 @@ xdr_decode_fattr(u32 *p, struct nfs_fattr *fattr) fattr->mode = (fattr->mode & ~S_IFMT) | S_IFIFO; fattr->rdev = 0; } - fattr->timestamp = jiffies; return p; } diff --git a/fs/nfs/nfs3xdr.c b/fs/nfs/nfs3xdr.c index db4a904810a4..0498bd36602c 100644 --- a/fs/nfs/nfs3xdr.c +++ b/fs/nfs/nfs3xdr.c @@ -174,7 +174,6 @@ xdr_decode_fattr(u32 *p, struct nfs_fattr *fattr) /* Update the mode bits */ fattr->valid |= (NFS_ATTR_FATTR | NFS_ATTR_FATTR_V3); - fattr->timestamp = jiffies; return p; } diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c index cd762648fa9a..8b21de8a06fa 100644 --- a/fs/nfs/nfs4xdr.c +++ b/fs/nfs/nfs4xdr.c @@ -2799,10 +2799,8 @@ static int decode_getfattr(struct xdr_stream *xdr, struct nfs_fattr *fattr, cons goto xdr_error; if ((status = decode_attr_time_modify(xdr, bitmap, &fattr->mtime)) != 0) goto xdr_error; - if ((status = verify_attr_len(xdr, savep, attrlen)) == 0) { + if ((status = verify_attr_len(xdr, savep, attrlen)) == 0) fattr->valid = NFS_ATTR_FATTR | NFS_ATTR_FATTR_V3 | NFS_ATTR_FATTR_V4; - fattr->timestamp = jiffies; - } xdr_error: if (status != 0) printk(KERN_NOTICE "%s: xdr error %d!\n", __FUNCTION__, -status); diff --git a/include/linux/nfs_fs.h b/include/linux/nfs_fs.h index abf890f5fbfb..faeaad666ca8 100644 --- a/include/linux/nfs_fs.h +++ b/include/linux/nfs_fs.h @@ -141,6 +141,7 @@ struct nfs_inode { unsigned long attrtimeo_timestamp; __u64 change_attr; /* v4 only */ + unsigned long last_updated; /* "Generation counter" for the attribute cache. This is * bumped whenever we update the metadata on the * server. @@ -319,6 +320,7 @@ extern u32 root_nfs_parse_addr(char *name); /*__init*/ static inline void nfs_fattr_init(struct nfs_fattr *fattr) { fattr->valid = 0; + fattr->time_start = jiffies; } /* diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h index 60086dac11d5..aeaee7e7c51d 100644 --- a/include/linux/nfs_xdr.h +++ b/include/linux/nfs_xdr.h @@ -41,7 +41,7 @@ struct nfs_fattr { __u32 bitmap[2]; /* NFSv4 returned attribute bitmap */ __u64 change_attr; /* NFSv4 change attribute */ __u64 pre_change_attr;/* pre-op NFSv4 change attribute */ - unsigned long timestamp; + unsigned long time_start; }; #define NFS_ATTR_WCC 0x0001 /* pre-op WCC data */ -- cgit v1.2.3 From decf491f3076190262d4c649bed877650623903a Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Thu, 27 Oct 2005 22:12:39 -0400 Subject: NFS: Don't let nfs_end_data_update() clobber attribute update information Since we almost always call nfs_end_data_update() after we called nfs_refresh_inode(), we now end up marking the inode metadata as needing revalidation immediately after having updated it. This patch rearranges things so that we mark the inode as needing revalidation _before_ we call nfs_refresh_inode() on those operations that need it. Signed-off-by: Trond Myklebust --- fs/nfs/inode.c | 38 ++++++++++++++++++++++++++++++++------ fs/nfs/nfs3proc.c | 30 +++++++++++++++--------------- fs/nfs/nfs4proc.c | 3 +++ fs/nfs/proc.c | 16 +++++++++++++--- include/linux/nfs_fs.h | 15 ++++++++++----- 5 files changed, 73 insertions(+), 29 deletions(-) (limited to 'include/linux') diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c index b7d4f8f13ac2..6b3156e15350 100644 --- a/fs/nfs/inode.c +++ b/fs/nfs/inode.c @@ -1236,13 +1236,12 @@ void nfs_end_data_update(struct inode *inode) struct nfs_inode *nfsi = NFS_I(inode); if (!nfs_have_delegation(inode, FMODE_READ)) { - /* Mark the attribute cache for revalidation */ - spin_lock(&inode->i_lock); - nfsi->cache_validity |= NFS_INO_INVALID_ATTR; - /* Directories and symlinks: invalidate page cache too */ - if (S_ISDIR(inode->i_mode) || S_ISLNK(inode->i_mode)) + /* Directories and symlinks: invalidate page cache */ + if (S_ISDIR(inode->i_mode) || S_ISLNK(inode->i_mode)) { + spin_lock(&inode->i_lock); nfsi->cache_validity |= NFS_INO_INVALID_DATA; - spin_unlock(&inode->i_lock); + spin_unlock(&inode->i_lock); + } } nfsi->cache_change_attribute = jiffies; atomic_dec(&nfsi->data_updates); @@ -1360,6 +1359,33 @@ int nfs_refresh_inode(struct inode *inode, struct nfs_fattr *fattr) return status; } +/** + * nfs_post_op_update_inode - try to update the inode attribute cache + * @inode - pointer to inode + * @fattr - updated attributes + * + * After an operation that has changed the inode metadata, mark the + * attribute cache as being invalid, then try to update it. + */ +int nfs_post_op_update_inode(struct inode *inode, struct nfs_fattr *fattr) +{ + struct nfs_inode *nfsi = NFS_I(inode); + int status = 0; + + spin_lock(&inode->i_lock); + if (unlikely((fattr->valid & NFS_ATTR_FATTR) == 0)) { + nfsi->cache_validity |= NFS_INO_INVALID_ATTR | NFS_INO_INVALID_ACCESS; + goto out; + } + status = nfs_update_inode(inode, fattr, fattr->time_start); + if (time_after_eq(fattr->time_start, nfsi->cache_change_attribute)) + nfsi->cache_validity &= ~(NFS_INO_INVALID_ATTR|NFS_INO_INVALID_ATIME|NFS_INO_REVAL_PAGECACHE); + nfsi->cache_change_attribute = jiffies; +out: + spin_unlock(&inode->i_lock); + return status; +} + /* * Many nfs protocol calls return the new file attributes after * an operation. Here we update the inode to reflect the state diff --git a/fs/nfs/nfs3proc.c b/fs/nfs/nfs3proc.c index 4b1b48b139f6..92c870d19ccd 100644 --- a/fs/nfs/nfs3proc.c +++ b/fs/nfs/nfs3proc.c @@ -266,7 +266,7 @@ static int nfs3_proc_write(struct nfs_write_data *wdata) nfs_fattr_init(fattr); status = rpc_call_sync(NFS_CLIENT(inode), &msg, rpcflags); if (status >= 0) - nfs_refresh_inode(inode, fattr); + nfs_post_op_update_inode(inode, fattr); dprintk("NFS reply write: %d\n", status); return status < 0? status : wdata->res.count; } @@ -288,7 +288,7 @@ static int nfs3_proc_commit(struct nfs_write_data *cdata) nfs_fattr_init(fattr); status = rpc_call_sync(NFS_CLIENT(inode), &msg, 0); if (status >= 0) - nfs_refresh_inode(inode, fattr); + nfs_post_op_update_inode(inode, fattr); dprintk("NFS reply commit: %d\n", status); return status; } @@ -332,7 +332,7 @@ again: nfs_fattr_init(&dir_attr); nfs_fattr_init(&fattr); status = rpc_call(NFS_CLIENT(dir), NFS3PROC_CREATE, &arg, &res, 0); - nfs_refresh_inode(dir, &dir_attr); + nfs_post_op_update_inode(dir, &dir_attr); /* If the server doesn't support the exclusive creation semantics, * try again with simple 'guarded' mode. */ @@ -403,7 +403,7 @@ nfs3_proc_remove(struct inode *dir, struct qstr *name) dprintk("NFS call remove %s\n", name->name); nfs_fattr_init(&dir_attr); status = rpc_call_sync(NFS_CLIENT(dir), &msg, 0); - nfs_refresh_inode(dir, &dir_attr); + nfs_post_op_update_inode(dir, &dir_attr); dprintk("NFS reply remove: %d\n", status); return status; } @@ -439,7 +439,7 @@ nfs3_proc_unlink_done(struct dentry *dir, struct rpc_task *task) return 1; if (msg->rpc_argp) { dir_attr = (struct nfs_fattr*)msg->rpc_resp; - nfs_refresh_inode(dir->d_inode, dir_attr); + nfs_post_op_update_inode(dir->d_inode, dir_attr); kfree(msg->rpc_argp); } return 0; @@ -468,8 +468,8 @@ nfs3_proc_rename(struct inode *old_dir, struct qstr *old_name, nfs_fattr_init(&old_dir_attr); nfs_fattr_init(&new_dir_attr); status = rpc_call(NFS_CLIENT(old_dir), NFS3PROC_RENAME, &arg, &res, 0); - nfs_refresh_inode(old_dir, &old_dir_attr); - nfs_refresh_inode(new_dir, &new_dir_attr); + nfs_post_op_update_inode(old_dir, &old_dir_attr); + nfs_post_op_update_inode(new_dir, &new_dir_attr); dprintk("NFS reply rename: %d\n", status); return status; } @@ -494,8 +494,8 @@ nfs3_proc_link(struct inode *inode, struct inode *dir, struct qstr *name) nfs_fattr_init(&dir_attr); nfs_fattr_init(&fattr); status = rpc_call(NFS_CLIENT(inode), NFS3PROC_LINK, &arg, &res, 0); - nfs_refresh_inode(dir, &dir_attr); - nfs_refresh_inode(inode, &fattr); + nfs_post_op_update_inode(dir, &dir_attr); + nfs_post_op_update_inode(inode, &fattr); dprintk("NFS reply link: %d\n", status); return status; } @@ -527,7 +527,7 @@ nfs3_proc_symlink(struct inode *dir, struct qstr *name, struct qstr *path, nfs_fattr_init(&dir_attr); nfs_fattr_init(fattr); status = rpc_call(NFS_CLIENT(dir), NFS3PROC_SYMLINK, &arg, &res, 0); - nfs_refresh_inode(dir, &dir_attr); + nfs_post_op_update_inode(dir, &dir_attr); dprintk("NFS reply symlink: %d\n", status); return status; } @@ -558,7 +558,7 @@ nfs3_proc_mkdir(struct inode *dir, struct dentry *dentry, struct iattr *sattr) nfs_fattr_init(&dir_attr); nfs_fattr_init(&fattr); status = rpc_call(NFS_CLIENT(dir), NFS3PROC_MKDIR, &arg, &res, 0); - nfs_refresh_inode(dir, &dir_attr); + nfs_post_op_update_inode(dir, &dir_attr); if (status != 0) goto out; status = nfs_instantiate(dentry, &fhandle, &fattr); @@ -584,7 +584,7 @@ nfs3_proc_rmdir(struct inode *dir, struct qstr *name) dprintk("NFS call rmdir %s\n", name->name); nfs_fattr_init(&dir_attr); status = rpc_call(NFS_CLIENT(dir), NFS3PROC_RMDIR, &arg, &dir_attr, 0); - nfs_refresh_inode(dir, &dir_attr); + nfs_post_op_update_inode(dir, &dir_attr); dprintk("NFS reply rmdir: %d\n", status); return status; } @@ -679,7 +679,7 @@ nfs3_proc_mknod(struct inode *dir, struct dentry *dentry, struct iattr *sattr, nfs_fattr_init(&dir_attr); nfs_fattr_init(&fattr); status = rpc_call(NFS_CLIENT(dir), NFS3PROC_MKNOD, &arg, &res, 0); - nfs_refresh_inode(dir, &dir_attr); + nfs_post_op_update_inode(dir, &dir_attr); if (status != 0) goto out; status = nfs_instantiate(dentry, &fh, &fattr); @@ -775,7 +775,7 @@ nfs3_write_done(struct rpc_task *task) return; data = (struct nfs_write_data *)task->tk_calldata; if (task->tk_status >= 0) - nfs_refresh_inode(data->inode, data->res.fattr); + nfs_post_op_update_inode(data->inode, data->res.fattr); nfs_writeback_done(task); } @@ -819,7 +819,7 @@ nfs3_commit_done(struct rpc_task *task) return; data = (struct nfs_write_data *)task->tk_calldata; if (task->tk_status >= 0) - nfs_refresh_inode(data->inode, data->res.fattr); + nfs_post_op_update_inode(data->inode, data->res.fattr); nfs_commit_done(task); } diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 2a759e8e387c..3274f2d354f3 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -187,8 +187,11 @@ static void update_changeattr(struct inode *inode, struct nfs4_change_info *cinf { struct nfs_inode *nfsi = NFS_I(inode); + spin_lock(&inode->i_lock); + nfsi->cache_validity |= NFS_INO_INVALID_ATTR; if (cinfo->before == nfsi->change_attr && cinfo->atomic) nfsi->change_attr = cinfo->after; + spin_unlock(&inode->i_lock); } /* Helper for asynchronous RPC calls */ diff --git a/fs/nfs/proc.c b/fs/nfs/proc.c index 5ef28f08f424..a48a003242c0 100644 --- a/fs/nfs/proc.c +++ b/fs/nfs/proc.c @@ -206,7 +206,7 @@ static int nfs_proc_write(struct nfs_write_data *wdata) nfs_fattr_init(fattr); status = rpc_call_sync(NFS_CLIENT(inode), &msg, flags); if (status >= 0) { - nfs_refresh_inode(inode, fattr); + nfs_post_op_update_inode(inode, fattr); wdata->res.count = wdata->args.count; wdata->verf.committed = NFS_FILE_SYNC; } @@ -275,6 +275,7 @@ nfs_proc_mknod(struct inode *dir, struct dentry *dentry, struct iattr *sattr, nfs_fattr_init(&fattr); status = rpc_call(NFS_CLIENT(dir), NFSPROC_CREATE, &arg, &res, 0); + nfs_mark_for_revalidate(dir); if (status == -EINVAL && S_ISFIFO(mode)) { sattr->ia_mode = mode; @@ -305,6 +306,7 @@ nfs_proc_remove(struct inode *dir, struct qstr *name) dprintk("NFS call remove %s\n", name->name); status = rpc_call_sync(NFS_CLIENT(dir), &msg, 0); + nfs_mark_for_revalidate(dir); dprintk("NFS reply remove: %d\n", status); return status; @@ -331,8 +333,10 @@ nfs_proc_unlink_done(struct dentry *dir, struct rpc_task *task) { struct rpc_message *msg = &task->tk_msg; - if (msg->rpc_argp) + if (msg->rpc_argp) { + nfs_mark_for_revalidate(dir->d_inode); kfree(msg->rpc_argp); + } return 0; } @@ -352,6 +356,8 @@ nfs_proc_rename(struct inode *old_dir, struct qstr *old_name, dprintk("NFS call rename %s -> %s\n", old_name->name, new_name->name); status = rpc_call(NFS_CLIENT(old_dir), NFSPROC_RENAME, &arg, NULL, 0); + nfs_mark_for_revalidate(old_dir); + nfs_mark_for_revalidate(new_dir); dprintk("NFS reply rename: %d\n", status); return status; } @@ -369,6 +375,7 @@ nfs_proc_link(struct inode *inode, struct inode *dir, struct qstr *name) dprintk("NFS call link %s\n", name->name); status = rpc_call(NFS_CLIENT(inode), NFSPROC_LINK, &arg, NULL, 0); + nfs_mark_for_revalidate(dir); dprintk("NFS reply link: %d\n", status); return status; } @@ -394,6 +401,7 @@ nfs_proc_symlink(struct inode *dir, struct qstr *name, struct qstr *path, nfs_fattr_init(fattr); fhandle->size = 0; status = rpc_call(NFS_CLIENT(dir), NFSPROC_SYMLINK, &arg, NULL, 0); + nfs_mark_for_revalidate(dir); dprintk("NFS reply symlink: %d\n", status); return status; } @@ -418,6 +426,7 @@ nfs_proc_mkdir(struct inode *dir, struct dentry *dentry, struct iattr *sattr) dprintk("NFS call mkdir %s\n", dentry->d_name.name); nfs_fattr_init(&fattr); status = rpc_call(NFS_CLIENT(dir), NFSPROC_MKDIR, &arg, &res, 0); + nfs_mark_for_revalidate(dir); if (status == 0) status = nfs_instantiate(dentry, &fhandle, &fattr); dprintk("NFS reply mkdir: %d\n", status); @@ -436,6 +445,7 @@ nfs_proc_rmdir(struct inode *dir, struct qstr *name) dprintk("NFS call rmdir %s\n", name->name); status = rpc_call(NFS_CLIENT(dir), NFSPROC_RMDIR, &arg, NULL, 0); + nfs_mark_for_revalidate(dir); dprintk("NFS reply rmdir: %d\n", status); return status; } @@ -579,7 +589,7 @@ nfs_write_done(struct rpc_task *task) struct nfs_write_data *data = (struct nfs_write_data *) task->tk_calldata; if (task->tk_status >= 0) - nfs_refresh_inode(data->inode, data->res.fattr); + nfs_post_op_update_inode(data->inode, data->res.fattr); nfs_writeback_done(task); } diff --git a/include/linux/nfs_fs.h b/include/linux/nfs_fs.h index faeaad666ca8..325fe7ae49bb 100644 --- a/include/linux/nfs_fs.h +++ b/include/linux/nfs_fs.h @@ -241,13 +241,17 @@ static inline int nfs_caches_unstable(struct inode *inode) return atomic_read(&NFS_I(inode)->data_updates) != 0; } +static inline void nfs_mark_for_revalidate(struct inode *inode) +{ + spin_lock(&inode->i_lock); + NFS_I(inode)->cache_validity |= NFS_INO_INVALID_ATTR | NFS_INO_INVALID_ACCESS; + spin_unlock(&inode->i_lock); +} + static inline void NFS_CACHEINV(struct inode *inode) { - if (!nfs_caches_unstable(inode)) { - spin_lock(&inode->i_lock); - NFS_I(inode)->cache_validity |= NFS_INO_INVALID_ATTR | NFS_INO_INVALID_ACCESS; - spin_unlock(&inode->i_lock); - } + if (!nfs_caches_unstable(inode)) + nfs_mark_for_revalidate(inode); } static inline int nfs_server_capable(struct inode *inode, int cap) @@ -291,6 +295,7 @@ extern void nfs_zap_caches(struct inode *); extern struct inode *nfs_fhget(struct super_block *, struct nfs_fh *, struct nfs_fattr *); extern int nfs_refresh_inode(struct inode *, struct nfs_fattr *); +extern int nfs_post_op_update_inode(struct inode *inode, struct nfs_fattr *fattr); extern int nfs_getattr(struct vfsmount *, struct dentry *, struct kstat *); extern int nfs_permission(struct inode *, int, struct nameidata *); extern int nfs_access_get_cached(struct inode *, struct rpc_cred *, struct nfs_access_entry *); -- cgit v1.2.3 From 56ae19f38f10aad4f27f7e12138a29b295dff07a Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Thu, 27 Oct 2005 22:12:40 -0400 Subject: NFSv4: Add directory post-op attributes to the CREATE operations. Since the directory attributes change every time we CREATE a file, we might as well pick up the new directory attributes in the same compound. Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 28 ++++++++++++++++---- fs/nfs/nfs4xdr.c | 69 ++++++++++++++++++++++++++++++++++++++++++------- include/linux/nfs_xdr.h | 2 ++ 3 files changed, 84 insertions(+), 15 deletions(-) (limited to 'include/linux') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 3274f2d354f3..f363fd6c7f4d 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -443,7 +443,11 @@ static int _nfs4_proc_open(struct inode *dir, struct nfs4_state_owner *sp, stru nfs_increment_open_seqid(status, o_arg->seqid); if (status != 0) goto out; - update_changeattr(dir, &o_res->cinfo); + if (o_arg->open_flags & O_CREAT) { + update_changeattr(dir, &o_res->cinfo); + nfs_post_op_update_inode(dir, o_res->dir_attr); + } else + nfs_refresh_inode(dir, o_res->dir_attr); if(o_res->rflags & NFS4_OPEN_RESULT_CONFIRM) { status = _nfs4_proc_open_confirm(server->client, &o_res->fh, sp, &o_res->stateid, o_arg->seqid); @@ -497,7 +501,7 @@ static int _nfs4_open_expired(struct nfs4_state_owner *sp, struct nfs4_state *st struct inode *inode = state->inode; struct nfs_server *server = NFS_SERVER(dir); struct nfs_delegation *delegation = NFS_I(inode)->delegation; - struct nfs_fattr f_attr; + struct nfs_fattr f_attr, dir_attr; struct nfs_openargs o_arg = { .fh = NFS_FH(dir), .open_flags = state->state, @@ -507,6 +511,7 @@ static int _nfs4_open_expired(struct nfs4_state_owner *sp, struct nfs4_state *st }; struct nfs_openres o_res = { .f_attr = &f_attr, + .dir_attr = &dir_attr, .server = server, }; int status = 0; @@ -524,6 +529,7 @@ static int _nfs4_open_expired(struct nfs4_state_owner *sp, struct nfs4_state *st if (o_arg.seqid == NULL) goto out; nfs_fattr_init(&f_attr); + nfs_fattr_init(&dir_attr); status = _nfs4_proc_open(dir, sp, &o_arg, &o_res); if (status != 0) goto out_nodeleg; @@ -694,7 +700,7 @@ static int _nfs4_do_open(struct inode *dir, struct dentry *dentry, int flags, st struct nfs4_client *clp = server->nfs4_state; struct inode *inode = NULL; int status; - struct nfs_fattr f_attr; + struct nfs_fattr f_attr, dir_attr; struct nfs_openargs o_arg = { .fh = NFS_FH(dir), .open_flags = flags, @@ -705,6 +711,7 @@ static int _nfs4_do_open(struct inode *dir, struct dentry *dentry, int flags, st }; struct nfs_openres o_res = { .f_attr = &f_attr, + .dir_attr = &dir_attr, .server = server, }; @@ -727,6 +734,7 @@ static int _nfs4_do_open(struct inode *dir, struct dentry *dentry, int flags, st if (o_arg.seqid == NULL) return -ENOMEM; nfs_fattr_init(&f_attr); + nfs_fattr_init(&dir_attr); status = _nfs4_proc_open(dir, sp, &o_arg, &o_res); if (status != 0) goto out_err; @@ -1746,6 +1754,7 @@ static int _nfs4_proc_symlink(struct inode *dir, struct qstr *name, struct nfs_fattr *fattr) { struct nfs_server *server = NFS_SERVER(dir); + struct nfs_fattr dir_fattr; struct nfs4_create_arg arg = { .dir_fh = NFS_FH(dir), .server = server, @@ -1758,6 +1767,7 @@ static int _nfs4_proc_symlink(struct inode *dir, struct qstr *name, .server = server, .fh = fhandle, .fattr = fattr, + .dir_fattr = &dir_fattr, }; struct rpc_message msg = { .rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_SYMLINK], @@ -1770,10 +1780,12 @@ static int _nfs4_proc_symlink(struct inode *dir, struct qstr *name, return -ENAMETOOLONG; arg.u.symlink = path; nfs_fattr_init(fattr); + nfs_fattr_init(&dir_fattr); status = rpc_call_sync(NFS_CLIENT(dir), &msg, 0); if (!status) update_changeattr(dir, &res.dir_cinfo); + nfs_post_op_update_inode(dir, res.dir_fattr); return status; } @@ -1797,7 +1809,7 @@ static int _nfs4_proc_mkdir(struct inode *dir, struct dentry *dentry, { struct nfs_server *server = NFS_SERVER(dir); struct nfs_fh fhandle; - struct nfs_fattr fattr; + struct nfs_fattr fattr, dir_fattr; struct nfs4_create_arg arg = { .dir_fh = NFS_FH(dir), .server = server, @@ -1810,6 +1822,7 @@ static int _nfs4_proc_mkdir(struct inode *dir, struct dentry *dentry, .server = server, .fh = &fhandle, .fattr = &fattr, + .dir_fattr = &dir_fattr, }; struct rpc_message msg = { .rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_CREATE], @@ -1819,10 +1832,12 @@ static int _nfs4_proc_mkdir(struct inode *dir, struct dentry *dentry, int status; nfs_fattr_init(&fattr); + nfs_fattr_init(&dir_fattr); status = rpc_call_sync(NFS_CLIENT(dir), &msg, 0); if (!status) { update_changeattr(dir, &res.dir_cinfo); + nfs_post_op_update_inode(dir, res.dir_fattr); status = nfs_instantiate(dentry, &fhandle, &fattr); } return status; @@ -1895,7 +1910,7 @@ static int _nfs4_proc_mknod(struct inode *dir, struct dentry *dentry, { struct nfs_server *server = NFS_SERVER(dir); struct nfs_fh fh; - struct nfs_fattr fattr; + struct nfs_fattr fattr, dir_fattr; struct nfs4_create_arg arg = { .dir_fh = NFS_FH(dir), .server = server, @@ -1907,6 +1922,7 @@ static int _nfs4_proc_mknod(struct inode *dir, struct dentry *dentry, .server = server, .fh = &fh, .fattr = &fattr, + .dir_fattr = &dir_fattr, }; struct rpc_message msg = { .rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_CREATE], @@ -1917,6 +1933,7 @@ static int _nfs4_proc_mknod(struct inode *dir, struct dentry *dentry, int mode = sattr->ia_mode; nfs_fattr_init(&fattr); + nfs_fattr_init(&dir_fattr); BUG_ON(!(sattr->ia_valid & ATTR_MODE)); BUG_ON(!S_ISFIFO(mode) && !S_ISBLK(mode) && !S_ISCHR(mode) && !S_ISSOCK(mode)); @@ -1938,6 +1955,7 @@ static int _nfs4_proc_mknod(struct inode *dir, struct dentry *dentry, status = rpc_call_sync(NFS_CLIENT(dir), &msg, 0); if (status == 0) { update_changeattr(dir, &res.dir_cinfo); + nfs_post_op_update_inode(dir, res.dir_fattr); status = nfs_instantiate(dentry, &fh, &fattr); } return status; diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c index 8b21de8a06fa..7f91d613d31a 100644 --- a/fs/nfs/nfs4xdr.c +++ b/fs/nfs/nfs4xdr.c @@ -95,6 +95,8 @@ static int nfs_stat_to_errno(int); #define decode_getattr_maxsz (op_decode_hdr_maxsz + nfs4_fattr_maxsz) #define encode_savefh_maxsz (op_encode_hdr_maxsz) #define decode_savefh_maxsz (op_decode_hdr_maxsz) +#define encode_restorefh_maxsz (op_encode_hdr_maxsz) +#define decode_restorefh_maxsz (op_decode_hdr_maxsz) #define encode_fsinfo_maxsz (op_encode_hdr_maxsz + 2) #define decode_fsinfo_maxsz (op_decode_hdr_maxsz + 11) #define encode_renew_maxsz (op_encode_hdr_maxsz + 3) @@ -336,14 +338,20 @@ static int nfs_stat_to_errno(int); decode_getfh_maxsz) #define NFS4_enc_create_sz (compound_encode_hdr_maxsz + \ encode_putfh_maxsz + \ + encode_savefh_maxsz + \ encode_create_maxsz + \ + encode_getfh_maxsz + \ encode_getattr_maxsz + \ - encode_getfh_maxsz) + encode_restorefh_maxsz + \ + encode_getattr_maxsz) #define NFS4_dec_create_sz (compound_decode_hdr_maxsz + \ decode_putfh_maxsz + \ + decode_savefh_maxsz + \ decode_create_maxsz + \ + decode_getfh_maxsz + \ decode_getattr_maxsz + \ - decode_getfh_maxsz) + decode_restorefh_maxsz + \ + decode_getattr_maxsz) #define NFS4_enc_pathconf_sz (compound_encode_hdr_maxsz + \ encode_putfh_maxsz + \ encode_getattr_maxsz) @@ -1112,6 +1120,17 @@ static int encode_renew(struct xdr_stream *xdr, const struct nfs4_client *client return 0; } +static int +encode_restorefh(struct xdr_stream *xdr) +{ + uint32_t *p; + + RESERVE_SPACE(4); + WRITE32(OP_RESTOREFH); + + return 0; +} + static int encode_setacl(struct xdr_stream *xdr, struct nfs_setaclargs *arg) { @@ -1358,7 +1377,7 @@ static int nfs4_xdr_enc_create(struct rpc_rqst *req, uint32_t *p, const struct n { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 4, + .nops = 7, }; int status; @@ -1366,10 +1385,16 @@ static int nfs4_xdr_enc_create(struct rpc_rqst *req, uint32_t *p, const struct n encode_compound_hdr(&xdr, &hdr); if ((status = encode_putfh(&xdr, args->dir_fh)) != 0) goto out; + if ((status = encode_savefh(&xdr)) != 0) + goto out; if ((status = encode_create(&xdr, args)) != 0) goto out; if ((status = encode_getfh(&xdr)) != 0) goto out; + if ((status = encode_getfattr(&xdr, args->bitmask)) != 0) + goto out; + if ((status = encode_restorefh(&xdr)) != 0) + goto out; status = encode_getfattr(&xdr, args->bitmask); out: return status; @@ -1429,7 +1454,7 @@ static int nfs4_xdr_enc_open(struct rpc_rqst *req, uint32_t *p, struct nfs_opena { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 4, + .nops = 7, }; int status; @@ -1439,6 +1464,9 @@ static int nfs4_xdr_enc_open(struct rpc_rqst *req, uint32_t *p, struct nfs_opena xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, &hdr); status = encode_putfh(&xdr, args->fh); + if (status) + goto out; + status = encode_savefh(&xdr); if (status) goto out; status = encode_open(&xdr, args); @@ -1448,6 +1476,12 @@ static int nfs4_xdr_enc_open(struct rpc_rqst *req, uint32_t *p, struct nfs_opena if (status) goto out; status = encode_getfattr(&xdr, args->bitmask); + if (status) + goto out; + status = encode_restorefh(&xdr); + if (status) + goto out; + status = encode_getfattr(&xdr, args->bitmask); out: return status; } @@ -3218,6 +3252,12 @@ static int decode_renew(struct xdr_stream *xdr) return decode_op_hdr(xdr, OP_RENEW); } +static int +decode_restorefh(struct xdr_stream *xdr) +{ + return decode_op_hdr(xdr, OP_RESTOREFH); +} + static int decode_getacl(struct xdr_stream *xdr, struct rpc_rqst *req, size_t *acl_len) { @@ -3510,13 +3550,17 @@ static int nfs4_xdr_dec_create(struct rpc_rqst *rqstp, uint32_t *p, struct nfs4_ goto out; if ((status = decode_putfh(&xdr)) != 0) goto out; + if ((status = decode_savefh(&xdr)) != 0) + goto out; if ((status = decode_create(&xdr,&res->dir_cinfo)) != 0) goto out; if ((status = decode_getfh(&xdr, res->fh)) != 0) goto out; - status = decode_getfattr(&xdr, res->fattr, res->server); - if (status == NFS4ERR_DELAY) - status = 0; + if (decode_getfattr(&xdr, res->fattr, res->server) != 0) + goto out; + if ((status = decode_restorefh(&xdr)) != 0) + goto out; + decode_getfattr(&xdr, res->dir_fattr, res->server); out: return status; } @@ -3654,15 +3698,20 @@ static int nfs4_xdr_dec_open(struct rpc_rqst *rqstp, uint32_t *p, struct nfs_ope status = decode_putfh(&xdr); if (status) goto out; + status = decode_savefh(&xdr); + if (status) + goto out; status = decode_open(&xdr, res); if (status) goto out; status = decode_getfh(&xdr, &res->fh); if (status) goto out; - status = decode_getfattr(&xdr, res->f_attr, res->server); - if (status == NFS4ERR_DELAY) - status = 0; + if (decode_getfattr(&xdr, res->f_attr, res->server) != 0) + goto out; + if ((status = decode_restorefh(&xdr)) != 0) + goto out; + decode_getfattr(&xdr, res->dir_attr, res->server); out: return status; } diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h index aeaee7e7c51d..6485b8b41b83 100644 --- a/include/linux/nfs_xdr.h +++ b/include/linux/nfs_xdr.h @@ -124,6 +124,7 @@ struct nfs_openres { struct nfs4_change_info cinfo; __u32 rflags; struct nfs_fattr * f_attr; + struct nfs_fattr * dir_attr; const struct nfs_server *server; int delegation_type; nfs4_stateid delegation; @@ -540,6 +541,7 @@ struct nfs4_create_res { struct nfs_fh * fh; struct nfs_fattr * fattr; struct nfs4_change_info dir_cinfo; + struct nfs_fattr * dir_fattr; }; struct nfs4_fsinfo_arg { -- cgit v1.2.3 From 516a6af641bb50c608329a5bd751acd0d65cc4ab Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Thu, 27 Oct 2005 22:12:41 -0400 Subject: NFS: Add optional post-op getattr instruction to the NFSv4 file close. "Optional" means that the close call will not fail if the getattr at the end of the compound fails. If it does succeed, try to refresh inode attributes. Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 9 ++++++++- fs/nfs/nfs4xdr.c | 34 ++++++++++++++++++++++++++++------ include/linux/nfs_xdr.h | 3 +++ 3 files changed, 39 insertions(+), 7 deletions(-) (limited to 'include/linux') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index f363fd6c7f4d..7be3d2d15d6f 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -865,6 +865,7 @@ struct nfs4_closedata { struct nfs4_state *state; struct nfs_closeargs arg; struct nfs_closeres res; + struct nfs_fattr fattr; }; static void nfs4_free_closedata(struct nfs4_closedata *calldata) @@ -904,6 +905,7 @@ static void nfs4_close_done(struct rpc_task *task) return; } } + nfs_refresh_inode(calldata->inode, calldata->res.fattr); state->state = calldata->arg.open_flags; nfs4_free_closedata(calldata); } @@ -941,6 +943,7 @@ static void nfs4_close_begin(struct rpc_task *task) rpc_exit(task, 0); return; } + nfs_fattr_init(calldata->res.fattr); if (mode != 0) msg.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_OPEN_DOWNGRADE]; calldata->arg.open_flags = mode; @@ -960,6 +963,7 @@ static void nfs4_close_begin(struct rpc_task *task) */ int nfs4_do_close(struct inode *inode, struct nfs4_state *state, mode_t mode) { + struct nfs_server *server = NFS_SERVER(inode); struct nfs4_closedata *calldata; int status = -ENOMEM; @@ -974,8 +978,11 @@ int nfs4_do_close(struct inode *inode, struct nfs4_state *state, mode_t mode) calldata->arg.seqid = nfs_alloc_seqid(&state->owner->so_seqid); if (calldata->arg.seqid == NULL) goto out_free_calldata; + calldata->arg.bitmask = server->attr_bitmask; + calldata->res.fattr = &calldata->fattr; + calldata->res.server = server; - status = nfs4_call_async(NFS_SERVER(inode)->client, nfs4_close_begin, + status = nfs4_call_async(server->client, nfs4_close_begin, nfs4_close_done, calldata); if (status == 0) goto out; diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c index 7f91d613d31a..cd9e26cfa868 100644 --- a/fs/nfs/nfs4xdr.c +++ b/fs/nfs/nfs4xdr.c @@ -198,17 +198,21 @@ static int nfs_stat_to_errno(int); #define NFS4_enc_open_downgrade_sz \ (compound_encode_hdr_maxsz + \ encode_putfh_maxsz + \ - op_encode_hdr_maxsz + 7) + op_encode_hdr_maxsz + 7 + \ + encode_getattr_maxsz) #define NFS4_dec_open_downgrade_sz \ (compound_decode_hdr_maxsz + \ decode_putfh_maxsz + \ - op_decode_hdr_maxsz + 4) + op_decode_hdr_maxsz + 4 + \ + decode_getattr_maxsz) #define NFS4_enc_close_sz (compound_encode_hdr_maxsz + \ encode_putfh_maxsz + \ - op_encode_hdr_maxsz + 5) + op_encode_hdr_maxsz + 5 + \ + encode_getattr_maxsz) #define NFS4_dec_close_sz (compound_decode_hdr_maxsz + \ decode_putfh_maxsz + \ - op_decode_hdr_maxsz + 4) + op_decode_hdr_maxsz + 4 + \ + decode_getattr_maxsz) #define NFS4_enc_setattr_sz (compound_encode_hdr_maxsz + \ encode_putfh_maxsz + \ op_encode_hdr_maxsz + 4 + \ @@ -1433,7 +1437,7 @@ static int nfs4_xdr_enc_close(struct rpc_rqst *req, uint32_t *p, struct nfs_clos { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 2, + .nops = 3, }; int status; @@ -1443,6 +1447,9 @@ static int nfs4_xdr_enc_close(struct rpc_rqst *req, uint32_t *p, struct nfs_clos if(status) goto out; status = encode_close(&xdr, args); + if (status != 0) + goto out; + status = encode_getfattr(&xdr, args->bitmask); out: return status; } @@ -1541,7 +1548,7 @@ static int nfs4_xdr_enc_open_downgrade(struct rpc_rqst *req, uint32_t *p, struct { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 2, + .nops = 3, }; int status; @@ -1551,6 +1558,9 @@ static int nfs4_xdr_enc_open_downgrade(struct rpc_rqst *req, uint32_t *p, struct if (status) goto out; status = encode_open_downgrade(&xdr, args); + if (status != 0) + goto out; + status = encode_getfattr(&xdr, args->bitmask); out: return status; } @@ -3403,6 +3413,9 @@ static int nfs4_xdr_dec_open_downgrade(struct rpc_rqst *rqstp, uint32_t *p, stru if (status) goto out; status = decode_open_downgrade(&xdr, res); + if (status != 0) + goto out; + decode_getfattr(&xdr, res->fattr, res->server); out: return status; } @@ -3678,6 +3691,15 @@ static int nfs4_xdr_dec_close(struct rpc_rqst *rqstp, uint32_t *p, struct nfs_cl if (status) goto out; status = decode_close(&xdr, res); + if (status != 0) + goto out; + /* + * Note: Server may do delete on close for this file + * in which case the getattr call will fail with + * an ESTALE error. Shouldn't be a problem, + * though, since fattr->valid will remain unset. + */ + decode_getfattr(&xdr, res->fattr, res->server); out: return status; } diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h index 6485b8b41b83..4f03dc21cf4a 100644 --- a/include/linux/nfs_xdr.h +++ b/include/linux/nfs_xdr.h @@ -153,10 +153,13 @@ struct nfs_closeargs { nfs4_stateid * stateid; struct nfs_seqid * seqid; int open_flags; + const u32 * bitmask; }; struct nfs_closeres { nfs4_stateid stateid; + struct nfs_fattr * fattr; + const struct nfs_server *server; }; /* * * Arguments to the lock,lockt, and locku call. -- cgit v1.2.3 From 91ba2eeec5e8e86e054937eb3bf5aec5b22b1830 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Thu, 27 Oct 2005 22:12:42 -0400 Subject: NFSv4: Add post-op attributes to nfs4_proc_link() Optimise attribute revalidation when hardlinking. Add post-op attributes for the directory and the original inode. Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 22 +++++++++++++++++----- fs/nfs/nfs4xdr.c | 34 ++++++++++++++++++++++++++++------ include/linux/nfs_xdr.h | 9 +++++++++ 3 files changed, 54 insertions(+), 11 deletions(-) (limited to 'include/linux') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 7be3d2d15d6f..04995e39e867 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -1724,22 +1724,34 @@ static int nfs4_proc_rename(struct inode *old_dir, struct qstr *old_name, static int _nfs4_proc_link(struct inode *inode, struct inode *dir, struct qstr *name) { + struct nfs_server *server = NFS_SERVER(inode); struct nfs4_link_arg arg = { .fh = NFS_FH(inode), .dir_fh = NFS_FH(dir), .name = name, + .bitmask = server->attr_bitmask, + }; + struct nfs_fattr fattr, dir_attr; + struct nfs4_link_res res = { + .server = server, + .fattr = &fattr, + .dir_attr = &dir_attr, }; - struct nfs4_change_info cinfo = { }; struct rpc_message msg = { .rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_LINK], .rpc_argp = &arg, - .rpc_resp = &cinfo, + .rpc_resp = &res, }; int status; - status = rpc_call_sync(NFS_CLIENT(inode), &msg, 0); - if (!status) - update_changeattr(dir, &cinfo); + nfs_fattr_init(res.fattr); + nfs_fattr_init(res.dir_attr); + status = rpc_call_sync(server->client, &msg, 0); + if (!status) { + update_changeattr(dir, &res.cinfo); + nfs_post_op_update_inode(dir, res.dir_attr); + nfs_refresh_inode(inode, res.fattr); + } return status; } diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c index cd9e26cfa868..f624b693ce21 100644 --- a/fs/nfs/nfs4xdr.c +++ b/fs/nfs/nfs4xdr.c @@ -324,12 +324,18 @@ static int nfs_stat_to_errno(int); encode_putfh_maxsz + \ encode_savefh_maxsz + \ encode_putfh_maxsz + \ - encode_link_maxsz) + encode_link_maxsz + \ + decode_getattr_maxsz + \ + encode_restorefh_maxsz + \ + decode_getattr_maxsz) #define NFS4_dec_link_sz (compound_decode_hdr_maxsz + \ decode_putfh_maxsz + \ decode_savefh_maxsz + \ decode_putfh_maxsz + \ - decode_link_maxsz) + decode_link_maxsz + \ + decode_getattr_maxsz + \ + decode_restorefh_maxsz + \ + decode_getattr_maxsz) #define NFS4_enc_symlink_sz (compound_encode_hdr_maxsz + \ encode_putfh_maxsz + \ encode_symlink_maxsz + \ @@ -1357,7 +1363,7 @@ static int nfs4_xdr_enc_link(struct rpc_rqst *req, uint32_t *p, const struct nfs { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 4, + .nops = 7, }; int status; @@ -1369,7 +1375,13 @@ static int nfs4_xdr_enc_link(struct rpc_rqst *req, uint32_t *p, const struct nfs goto out; if ((status = encode_putfh(&xdr, args->dir_fh)) != 0) goto out; - status = encode_link(&xdr, args->name); + if ((status = encode_link(&xdr, args->name)) != 0) + goto out; + if ((status = encode_getfattr(&xdr, args->bitmask)) != 0) + goto out; + if ((status = encode_restorefh(&xdr)) != 0) + goto out; + status = encode_getfattr(&xdr, args->bitmask); out: return status; } @@ -3529,7 +3541,7 @@ out: /* * Decode LINK response */ -static int nfs4_xdr_dec_link(struct rpc_rqst *rqstp, uint32_t *p, struct nfs4_change_info *cinfo) +static int nfs4_xdr_dec_link(struct rpc_rqst *rqstp, uint32_t *p, struct nfs4_link_res *res) { struct xdr_stream xdr; struct compound_hdr hdr; @@ -3544,7 +3556,17 @@ static int nfs4_xdr_dec_link(struct rpc_rqst *rqstp, uint32_t *p, struct nfs4_ch goto out; if ((status = decode_putfh(&xdr)) != 0) goto out; - status = decode_link(&xdr, cinfo); + if ((status = decode_link(&xdr, &res->cinfo)) != 0) + goto out; + /* + * Note order: OP_LINK leaves the directory as the current + * filehandle. + */ + if (decode_getfattr(&xdr, res->dir_attr, res->server) != 0) + goto out; + if ((status = decode_restorefh(&xdr)) != 0) + goto out; + decode_getfattr(&xdr, res->fattr, res->server); out: return status; } diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h index 4f03dc21cf4a..89238b799cfd 100644 --- a/include/linux/nfs_xdr.h +++ b/include/linux/nfs_xdr.h @@ -566,8 +566,17 @@ struct nfs4_link_arg { const struct nfs_fh * fh; const struct nfs_fh * dir_fh; const struct qstr * name; + const u32 * bitmask; +}; + +struct nfs4_link_res { + const struct nfs_server * server; + struct nfs_fattr * fattr; + struct nfs4_change_info cinfo; + struct nfs_fattr * dir_attr; }; + struct nfs4_lookup_arg { const struct nfs_fh * dir_fh; const struct qstr * name; -- cgit v1.2.3 From 6caf2c8276d371679a798058e8fdf49f5ff831a3 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Thu, 27 Oct 2005 22:12:43 -0400 Subject: NFSv4: Add post-op attributes to nfs4_proc_rename() Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 15 +++++++++++++-- fs/nfs/nfs4xdr.c | 29 ++++++++++++++++++++++++----- include/linux/nfs_xdr.h | 4 ++++ 3 files changed, 41 insertions(+), 7 deletions(-) (limited to 'include/linux') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 04995e39e867..f96bc12c0fa0 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -1685,13 +1685,20 @@ static int nfs4_proc_unlink_done(struct dentry *dir, struct rpc_task *task) static int _nfs4_proc_rename(struct inode *old_dir, struct qstr *old_name, struct inode *new_dir, struct qstr *new_name) { + struct nfs_server *server = NFS_SERVER(old_dir); struct nfs4_rename_arg arg = { .old_dir = NFS_FH(old_dir), .new_dir = NFS_FH(new_dir), .old_name = old_name, .new_name = new_name, + .bitmask = server->attr_bitmask, + }; + struct nfs_fattr old_fattr, new_fattr; + struct nfs4_rename_res res = { + .server = server, + .old_fattr = &old_fattr, + .new_fattr = &new_fattr, }; - struct nfs4_rename_res res = { }; struct rpc_message msg = { .rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_RENAME], .rpc_argp = &arg, @@ -1699,11 +1706,15 @@ static int _nfs4_proc_rename(struct inode *old_dir, struct qstr *old_name, }; int status; - status = rpc_call_sync(NFS_CLIENT(old_dir), &msg, 0); + nfs_fattr_init(res.old_fattr); + nfs_fattr_init(res.new_fattr); + status = rpc_call_sync(server->client, &msg, 0); if (!status) { update_changeattr(old_dir, &res.old_cinfo); + nfs_post_op_update_inode(old_dir, res.old_fattr); update_changeattr(new_dir, &res.new_cinfo); + nfs_post_op_update_inode(new_dir, res.new_fattr); } return status; } diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c index f624b693ce21..2a07755bd347 100644 --- a/fs/nfs/nfs4xdr.c +++ b/fs/nfs/nfs4xdr.c @@ -314,12 +314,18 @@ static int nfs_stat_to_errno(int); encode_putfh_maxsz + \ encode_savefh_maxsz + \ encode_putfh_maxsz + \ - encode_rename_maxsz) + encode_rename_maxsz + \ + encode_getattr_maxsz + \ + encode_restorefh_maxsz + \ + encode_getattr_maxsz) #define NFS4_dec_rename_sz (compound_decode_hdr_maxsz + \ decode_putfh_maxsz + \ decode_savefh_maxsz + \ decode_putfh_maxsz + \ - decode_rename_maxsz) + decode_rename_maxsz + \ + decode_getattr_maxsz + \ + decode_restorefh_maxsz + \ + decode_getattr_maxsz) #define NFS4_enc_link_sz (compound_encode_hdr_maxsz + \ encode_putfh_maxsz + \ encode_savefh_maxsz + \ @@ -1339,7 +1345,7 @@ static int nfs4_xdr_enc_rename(struct rpc_rqst *req, uint32_t *p, const struct n { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 4, + .nops = 7, }; int status; @@ -1351,7 +1357,13 @@ static int nfs4_xdr_enc_rename(struct rpc_rqst *req, uint32_t *p, const struct n goto out; if ((status = encode_putfh(&xdr, args->new_dir)) != 0) goto out; - status = encode_rename(&xdr, args->old_name, args->new_name); + if ((status = encode_rename(&xdr, args->old_name, args->new_name)) != 0) + goto out; + if ((status = encode_getfattr(&xdr, args->bitmask)) != 0) + goto out; + if ((status = encode_restorefh(&xdr)) != 0) + goto out; + status = encode_getfattr(&xdr, args->bitmask); out: return status; } @@ -3533,7 +3545,14 @@ static int nfs4_xdr_dec_rename(struct rpc_rqst *rqstp, uint32_t *p, struct nfs4_ goto out; if ((status = decode_putfh(&xdr)) != 0) goto out; - status = decode_rename(&xdr, &res->old_cinfo, &res->new_cinfo); + if ((status = decode_rename(&xdr, &res->old_cinfo, &res->new_cinfo)) != 0) + goto out; + /* Current FH is target directory */ + if (decode_getfattr(&xdr, res->new_fattr, res->server) != 0) + goto out; + if ((status = decode_restorefh(&xdr)) != 0) + goto out; + decode_getfattr(&xdr, res->old_fattr, res->server); out: return status; } diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h index 89238b799cfd..6f0804280824 100644 --- a/include/linux/nfs_xdr.h +++ b/include/linux/nfs_xdr.h @@ -630,11 +630,15 @@ struct nfs4_rename_arg { const struct nfs_fh * new_dir; const struct qstr * old_name; const struct qstr * new_name; + const u32 * bitmask; }; struct nfs4_rename_res { + const struct nfs_server * server; struct nfs4_change_info old_cinfo; + struct nfs_fattr * old_fattr; struct nfs4_change_info new_cinfo; + struct nfs_fattr * new_fattr; }; struct nfs4_setclientid { -- cgit v1.2.3 From 16e429596dec4d28e16812b3a9be27f18412c567 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Thu, 27 Oct 2005 22:12:44 -0400 Subject: NFSv4: Add post-op attributes to nfs4_proc_remove() Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 27 +++++++++++++++++++++------ fs/nfs/nfs4xdr.c | 25 +++++++++++++++++-------- include/linux/nfs_xdr.h | 7 +++++++ 3 files changed, 45 insertions(+), 14 deletions(-) (limited to 'include/linux') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index f96bc12c0fa0..bab47c4cb41c 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -1614,11 +1614,17 @@ out: static int _nfs4_proc_remove(struct inode *dir, struct qstr *name) { + struct nfs_server *server = NFS_SERVER(dir); struct nfs4_remove_arg args = { .fh = NFS_FH(dir), .name = name, + .bitmask = server->attr_bitmask, + }; + struct nfs_fattr dir_attr; + struct nfs4_remove_res res = { + .server = server, + .dir_attr = &dir_attr, }; - struct nfs4_change_info res; struct rpc_message msg = { .rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_REMOVE], .rpc_argp = &args, @@ -1626,9 +1632,12 @@ static int _nfs4_proc_remove(struct inode *dir, struct qstr *name) }; int status; - status = rpc_call_sync(NFS_CLIENT(dir), &msg, 0); - if (status == 0) - update_changeattr(dir, &res); + nfs_fattr_init(res.dir_attr); + status = rpc_call_sync(server->client, &msg, 0); + if (status == 0) { + update_changeattr(dir, &res.cinfo); + nfs_post_op_update_inode(dir, res.dir_attr); + } return status; } @@ -1646,12 +1655,14 @@ static int nfs4_proc_remove(struct inode *dir, struct qstr *name) struct unlink_desc { struct nfs4_remove_arg args; - struct nfs4_change_info res; + struct nfs4_remove_res res; + struct nfs_fattr dir_attr; }; static int nfs4_proc_unlink_setup(struct rpc_message *msg, struct dentry *dir, struct qstr *name) { + struct nfs_server *server = NFS_SERVER(dir->d_inode); struct unlink_desc *up; up = (struct unlink_desc *) kmalloc(sizeof(*up), GFP_KERNEL); @@ -1660,6 +1671,9 @@ static int nfs4_proc_unlink_setup(struct rpc_message *msg, struct dentry *dir, up->args.fh = NFS_FH(dir->d_inode); up->args.name = name; + up->args.bitmask = server->attr_bitmask; + up->res.server = server; + up->res.dir_attr = &up->dir_attr; msg->rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_REMOVE]; msg->rpc_argp = &up->args; @@ -1674,7 +1688,8 @@ static int nfs4_proc_unlink_done(struct dentry *dir, struct rpc_task *task) if (msg->rpc_resp != NULL) { up = container_of(msg->rpc_resp, struct unlink_desc, res); - update_changeattr(dir->d_inode, &up->res); + update_changeattr(dir->d_inode, &up->res.cinfo); + nfs_post_op_update_inode(dir->d_inode, up->res.dir_attr); kfree(up); msg->rpc_resp = NULL; msg->rpc_argp = NULL; diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c index 2a07755bd347..3ee3a1669d28 100644 --- a/fs/nfs/nfs4xdr.c +++ b/fs/nfs/nfs4xdr.c @@ -306,10 +306,12 @@ static int nfs_stat_to_errno(int); decode_getfh_maxsz) #define NFS4_enc_remove_sz (compound_encode_hdr_maxsz + \ encode_putfh_maxsz + \ - encode_remove_maxsz) + encode_remove_maxsz + \ + encode_getattr_maxsz) #define NFS4_dec_remove_sz (compound_decode_hdr_maxsz + \ decode_putfh_maxsz + \ - op_decode_hdr_maxsz + 5) + op_decode_hdr_maxsz + 5 + \ + decode_getattr_maxsz) #define NFS4_enc_rename_sz (compound_encode_hdr_maxsz + \ encode_putfh_maxsz + \ encode_savefh_maxsz + \ @@ -1327,14 +1329,18 @@ static int nfs4_xdr_enc_remove(struct rpc_rqst *req, uint32_t *p, const struct n { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 2, + .nops = 3, }; int status; xdr_init_encode(&xdr, &req->rq_snd_buf, p); encode_compound_hdr(&xdr, &hdr); - if ((status = encode_putfh(&xdr, args->fh)) == 0) - status = encode_remove(&xdr, args->name); + if ((status = encode_putfh(&xdr, args->fh)) != 0) + goto out; + if ((status = encode_remove(&xdr, args->name)) != 0) + goto out; + status = encode_getfattr(&xdr, args->bitmask); +out: return status; } @@ -3512,7 +3518,7 @@ out: /* * Decode REMOVE response */ -static int nfs4_xdr_dec_remove(struct rpc_rqst *rqstp, uint32_t *p, struct nfs4_change_info *cinfo) +static int nfs4_xdr_dec_remove(struct rpc_rqst *rqstp, uint32_t *p, struct nfs4_remove_res *res) { struct xdr_stream xdr; struct compound_hdr hdr; @@ -3521,8 +3527,11 @@ static int nfs4_xdr_dec_remove(struct rpc_rqst *rqstp, uint32_t *p, struct nfs4_ xdr_init_decode(&xdr, &rqstp->rq_rcv_buf, p); if ((status = decode_compound_hdr(&xdr, &hdr)) != 0) goto out; - if ((status = decode_putfh(&xdr)) == 0) - status = decode_remove(&xdr, cinfo); + if ((status = decode_putfh(&xdr)) != 0) + goto out; + if ((status = decode_remove(&xdr, &res->cinfo)) != 0) + goto out; + decode_getfattr(&xdr, res->dir_attr, res->server); out: return status; } diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h index 6f0804280824..deeba7e2c518 100644 --- a/include/linux/nfs_xdr.h +++ b/include/linux/nfs_xdr.h @@ -623,6 +623,13 @@ struct nfs4_readlink { struct nfs4_remove_arg { const struct nfs_fh * fh; const struct qstr * name; + const u32 * bitmask; +}; + +struct nfs4_remove_res { + const struct nfs_server * server; + struct nfs4_change_info cinfo; + struct nfs_fattr * dir_attr; }; struct nfs4_rename_arg { -- cgit v1.2.3 From 4f9838c7ecd14f31f701f64fa65ded132fc0db8a Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Thu, 27 Oct 2005 22:12:44 -0400 Subject: NFSv4: Add post-op attributes to NFSv4 write and commit callbacks. Signed-off-by: Trond Myklebust --- fs/nfs/nfs4proc.c | 13 ++++++++++++- fs/nfs/nfs4xdr.c | 28 ++++++++++++++++++++++------ include/linux/nfs_xdr.h | 2 ++ 3 files changed, 36 insertions(+), 7 deletions(-) (limited to 'include/linux') diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index bab47c4cb41c..933e13b383f8 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -2169,8 +2169,10 @@ nfs4_write_done(struct rpc_task *task) rpc_restart_call(task); return; } - if (task->tk_status >= 0) + if (task->tk_status >= 0) { renew_lease(NFS_SERVER(inode), data->timestamp); + nfs_post_op_update_inode(inode, data->res.fattr); + } /* Call back common NFS writeback processing */ nfs_writeback_done(task); } @@ -2186,6 +2188,7 @@ nfs4_proc_write_setup(struct nfs_write_data *data, int how) .rpc_cred = data->cred, }; struct inode *inode = data->inode; + struct nfs_server *server = NFS_SERVER(inode); int stable; int flags; @@ -2197,6 +2200,8 @@ nfs4_proc_write_setup(struct nfs_write_data *data, int how) } else stable = NFS_UNSTABLE; data->args.stable = stable; + data->args.bitmask = server->attr_bitmask; + data->res.server = server; data->timestamp = jiffies; @@ -2218,6 +2223,8 @@ nfs4_commit_done(struct rpc_task *task) rpc_restart_call(task); return; } + if (task->tk_status >= 0) + nfs_post_op_update_inode(inode, data->res.fattr); /* Call back common NFS writeback processing */ nfs_commit_done(task); } @@ -2233,8 +2240,12 @@ nfs4_proc_commit_setup(struct nfs_write_data *data, int how) .rpc_cred = data->cred, }; struct inode *inode = data->inode; + struct nfs_server *server = NFS_SERVER(inode); int flags; + data->args.bitmask = server->attr_bitmask; + data->res.server = server; + /* Set the initial flags for the task. */ flags = (how & FLUSH_SYNC) ? 0 : RPC_TASK_ASYNC; diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c index 3ee3a1669d28..6f1bf182e0e0 100644 --- a/fs/nfs/nfs4xdr.c +++ b/fs/nfs/nfs4xdr.c @@ -159,16 +159,20 @@ static int nfs_stat_to_errno(int); op_decode_hdr_maxsz + 2) #define NFS4_enc_write_sz (compound_encode_hdr_maxsz + \ encode_putfh_maxsz + \ - op_encode_hdr_maxsz + 8) + op_encode_hdr_maxsz + 8 + \ + encode_getattr_maxsz) #define NFS4_dec_write_sz (compound_decode_hdr_maxsz + \ decode_putfh_maxsz + \ - op_decode_hdr_maxsz + 4) + op_decode_hdr_maxsz + 4 + \ + decode_getattr_maxsz) #define NFS4_enc_commit_sz (compound_encode_hdr_maxsz + \ encode_putfh_maxsz + \ - op_encode_hdr_maxsz + 3) + op_encode_hdr_maxsz + 3 + \ + encode_getattr_maxsz) #define NFS4_dec_commit_sz (compound_decode_hdr_maxsz + \ decode_putfh_maxsz + \ - op_decode_hdr_maxsz + 2) + op_decode_hdr_maxsz + 2 + \ + decode_getattr_maxsz) #define NFS4_enc_open_sz (compound_encode_hdr_maxsz + \ encode_putfh_maxsz + \ op_encode_hdr_maxsz + \ @@ -1799,7 +1803,7 @@ static int nfs4_xdr_enc_write(struct rpc_rqst *req, uint32_t *p, struct nfs_writ { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 2, + .nops = 3, }; int status; @@ -1809,6 +1813,9 @@ static int nfs4_xdr_enc_write(struct rpc_rqst *req, uint32_t *p, struct nfs_writ if (status) goto out; status = encode_write(&xdr, args); + if (status) + goto out; + status = encode_getfattr(&xdr, args->bitmask); out: return status; } @@ -1820,7 +1827,7 @@ static int nfs4_xdr_enc_commit(struct rpc_rqst *req, uint32_t *p, struct nfs_wri { struct xdr_stream xdr; struct compound_hdr hdr = { - .nops = 2, + .nops = 3, }; int status; @@ -1830,6 +1837,9 @@ static int nfs4_xdr_enc_commit(struct rpc_rqst *req, uint32_t *p, struct nfs_wri if (status) goto out; status = encode_commit(&xdr, args); + if (status) + goto out; + status = encode_getfattr(&xdr, args->bitmask); out: return status; } @@ -4001,6 +4011,9 @@ static int nfs4_xdr_dec_write(struct rpc_rqst *rqstp, uint32_t *p, struct nfs_wr if (status) goto out; status = decode_write(&xdr, res); + if (status) + goto out; + decode_getfattr(&xdr, res->fattr, res->server); if (!status) status = res->count; out: @@ -4024,6 +4037,9 @@ static int nfs4_xdr_dec_commit(struct rpc_rqst *rqstp, uint32_t *p, struct nfs_w if (status) goto out; status = decode_commit(&xdr, res); + if (status) + goto out; + decode_getfattr(&xdr, res->fattr, res->server); out: return status; } diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h index deeba7e2c518..40718669b9c8 100644 --- a/include/linux/nfs_xdr.h +++ b/include/linux/nfs_xdr.h @@ -256,6 +256,7 @@ struct nfs_writeargs { enum nfs3_stable_how stable; unsigned int pgbase; struct page ** pages; + const u32 * bitmask; }; struct nfs_writeverf { @@ -267,6 +268,7 @@ struct nfs_writeres { struct nfs_fattr * fattr; struct nfs_writeverf * verf; __u32 count; + const struct nfs_server *server; }; /* -- cgit v1.2.3 From 20e5c81fcff89535dced2ed71cf24c6c648ff40e Mon Sep 17 00:00:00 2001 From: "Chen, Kenneth W" Date: Thu, 13 Oct 2005 21:48:42 +0200 Subject: [patch] remove gendisk->stamp_idle field struct gendisk has these two fields: stamp, stamp_idle. Update to stamp_idle is always in sync with stamp and they are always the same. Therefore, it does not add any value in having two fields tracking same timestamp. Suggest to remove it. Also, we should only update gendisk stats with non-zero value. Advantage is that we don't have to needlessly calculate memory address, and then add zero to the content. Signed-off-by: Ken Chen Signed-off-by: Jens Axboe --- drivers/block/ll_rw_blk.c | 11 +++++------ fs/partitions/check.c | 2 +- include/linux/genhd.h | 2 +- 3 files changed, 7 insertions(+), 8 deletions(-) (limited to 'include/linux') diff --git a/drivers/block/ll_rw_blk.c b/drivers/block/ll_rw_blk.c index baedac522945..c42071fd2e95 100644 --- a/drivers/block/ll_rw_blk.c +++ b/drivers/block/ll_rw_blk.c @@ -2433,13 +2433,12 @@ void disk_round_stats(struct gendisk *disk) { unsigned long now = jiffies; - __disk_stat_add(disk, time_in_queue, - disk->in_flight * (now - disk->stamp)); + if (disk->in_flight) { + __disk_stat_add(disk, time_in_queue, + disk->in_flight * (now - disk->stamp)); + __disk_stat_add(disk, io_ticks, (now - disk->stamp)); + } disk->stamp = now; - - if (disk->in_flight) - __disk_stat_add(disk, io_ticks, (now - disk->stamp_idle)); - disk->stamp_idle = now; } /* diff --git a/fs/partitions/check.c b/fs/partitions/check.c index 77e178f13162..1e848648a322 100644 --- a/fs/partitions/check.c +++ b/fs/partitions/check.c @@ -430,7 +430,7 @@ void del_gendisk(struct gendisk *disk) disk->flags &= ~GENHD_FL_UP; unlink_gendisk(disk); disk_stat_set_all(disk, 0); - disk->stamp = disk->stamp_idle = 0; + disk->stamp = 0; devfs_remove_disk(disk); diff --git a/include/linux/genhd.h b/include/linux/genhd.h index 01796c41c951..142e1c1e0689 100644 --- a/include/linux/genhd.h +++ b/include/linux/genhd.h @@ -119,7 +119,7 @@ struct gendisk { int policy; atomic_t sync_io; /* RAID */ - unsigned long stamp, stamp_idle; + unsigned long stamp; int in_flight; #ifdef CONFIG_SMP struct disk_stats *dkstats; -- cgit v1.2.3 From 8922e16cf6269e668123acb1ae1fdc62b7a3a4fc Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Thu, 20 Oct 2005 16:23:44 +0200 Subject: [PATCH] 01/05 Implement generic dispatch queue Implements generic dispatch queue which can replace all dispatch queues implemented by each iosched. This reduces code duplication, eases enforcing semantics over dispatch queue, and simplifies specific ioscheds. Signed-off-by: Tejun Heo Signed-off-by: Jens Axboe --- drivers/block/elevator.c | 242 ++++++++++++++++++++++++++++++++-------------- drivers/block/ll_rw_blk.c | 23 +++-- include/linux/blkdev.h | 17 ++-- include/linux/elevator.h | 16 ++- 4 files changed, 201 insertions(+), 97 deletions(-) (limited to 'include/linux') diff --git a/drivers/block/elevator.c b/drivers/block/elevator.c index 4144f30d82a9..a27555908d35 100644 --- a/drivers/block/elevator.c +++ b/drivers/block/elevator.c @@ -40,6 +40,11 @@ static DEFINE_SPINLOCK(elv_list_lock); static LIST_HEAD(elv_list); +static inline sector_t rq_last_sector(struct request *rq) +{ + return rq->sector + rq->nr_sectors; +} + /* * can we safely merge with this request? */ @@ -143,6 +148,9 @@ static int elevator_attach(request_queue_t *q, struct elevator_type *e, INIT_LIST_HEAD(&q->queue_head); q->last_merge = NULL; q->elevator = eq; + q->last_sector = 0; + q->boundary_rq = NULL; + q->max_back_kb = 0; if (eq->ops->elevator_init_fn) ret = eq->ops->elevator_init_fn(q, eq); @@ -225,6 +233,48 @@ void elevator_exit(elevator_t *e) kfree(e); } +/* + * Insert rq into dispatch queue of q. Queue lock must be held on + * entry. If sort != 0, rq is sort-inserted; otherwise, rq will be + * appended to the dispatch queue. To be used by specific elevators. + */ +void elv_dispatch_insert(request_queue_t *q, struct request *rq, int sort) +{ + sector_t boundary; + unsigned max_back; + struct list_head *entry; + + if (!sort) { + /* Specific elevator is performing sort. Step away. */ + q->last_sector = rq_last_sector(rq); + q->boundary_rq = rq; + list_add_tail(&rq->queuelist, &q->queue_head); + return; + } + + boundary = q->last_sector; + max_back = q->max_back_kb * 2; + boundary = boundary > max_back ? boundary - max_back : 0; + + list_for_each_prev(entry, &q->queue_head) { + struct request *pos = list_entry_rq(entry); + + if (pos->flags & (REQ_SOFTBARRIER|REQ_HARDBARRIER|REQ_STARTED)) + break; + if (rq->sector >= boundary) { + if (pos->sector < boundary) + continue; + } else { + if (pos->sector >= boundary) + break; + } + if (rq->sector >= pos->sector) + break; + } + + list_add(&rq->queuelist, entry); +} + int elv_merge(request_queue_t *q, struct request **req, struct bio *bio) { elevator_t *e = q->elevator; @@ -255,13 +305,7 @@ void elv_merge_requests(request_queue_t *q, struct request *rq, e->ops->elevator_merge_req_fn(q, rq, next); } -/* - * For careful internal use by the block layer. Essentially the same as - * a requeue in that it tells the io scheduler that this request is not - * active in the driver or hardware anymore, but we don't want the request - * added back to the scheduler. Function is not exported. - */ -void elv_deactivate_request(request_queue_t *q, struct request *rq) +void elv_requeue_request(request_queue_t *q, struct request *rq) { elevator_t *e = q->elevator; @@ -269,19 +313,14 @@ void elv_deactivate_request(request_queue_t *q, struct request *rq) * it already went through dequeue, we need to decrement the * in_flight count again */ - if (blk_account_rq(rq)) + if (blk_account_rq(rq)) { q->in_flight--; + if (blk_sorted_rq(rq) && e->ops->elevator_deactivate_req_fn) + e->ops->elevator_deactivate_req_fn(q, rq); + } rq->flags &= ~REQ_STARTED; - if (e->ops->elevator_deactivate_req_fn) - e->ops->elevator_deactivate_req_fn(q, rq); -} - -void elv_requeue_request(request_queue_t *q, struct request *rq) -{ - elv_deactivate_request(q, rq); - /* * if this is the flush, requeue the original instead and drop the flush */ @@ -290,55 +329,89 @@ void elv_requeue_request(request_queue_t *q, struct request *rq) rq = rq->end_io_data; } - /* - * the request is prepped and may have some resources allocated. - * allowing unprepped requests to pass this one may cause resource - * deadlock. turn on softbarrier. - */ - rq->flags |= REQ_SOFTBARRIER; - - /* - * if iosched has an explicit requeue hook, then use that. otherwise - * just put the request at the front of the queue - */ - if (q->elevator->ops->elevator_requeue_req_fn) - q->elevator->ops->elevator_requeue_req_fn(q, rq); - else - __elv_add_request(q, rq, ELEVATOR_INSERT_FRONT, 0); + __elv_add_request(q, rq, ELEVATOR_INSERT_FRONT, 0); } void __elv_add_request(request_queue_t *q, struct request *rq, int where, int plug) { - /* - * barriers implicitly indicate back insertion - */ - if (rq->flags & (REQ_SOFTBARRIER | REQ_HARDBARRIER) && - where == ELEVATOR_INSERT_SORT) - where = ELEVATOR_INSERT_BACK; + if (rq->flags & (REQ_SOFTBARRIER | REQ_HARDBARRIER)) { + /* + * barriers implicitly indicate back insertion + */ + if (where == ELEVATOR_INSERT_SORT) + where = ELEVATOR_INSERT_BACK; + + /* + * this request is scheduling boundary, update last_sector + */ + if (blk_fs_request(rq)) { + q->last_sector = rq_last_sector(rq); + q->boundary_rq = rq; + } + } if (plug) blk_plug_device(q); rq->q = q; - if (!test_bit(QUEUE_FLAG_DRAIN, &q->queue_flags)) { - q->elevator->ops->elevator_add_req_fn(q, rq, where); - - if (blk_queue_plugged(q)) { - int nrq = q->rq.count[READ] + q->rq.count[WRITE] - - q->in_flight; - - if (nrq >= q->unplug_thresh) - __generic_unplug_device(q); - } - } else + if (unlikely(test_bit(QUEUE_FLAG_DRAIN, &q->queue_flags))) { /* * if drain is set, store the request "locally". when the drain * is finished, the requests will be handed ordered to the io * scheduler */ list_add_tail(&rq->queuelist, &q->drain_list); + return; + } + + switch (where) { + case ELEVATOR_INSERT_FRONT: + rq->flags |= REQ_SOFTBARRIER; + + list_add(&rq->queuelist, &q->queue_head); + break; + + case ELEVATOR_INSERT_BACK: + rq->flags |= REQ_SOFTBARRIER; + + while (q->elevator->ops->elevator_dispatch_fn(q, 1)) + ; + list_add_tail(&rq->queuelist, &q->queue_head); + /* + * We kick the queue here for the following reasons. + * - The elevator might have returned NULL previously + * to delay requests and returned them now. As the + * queue wasn't empty before this request, ll_rw_blk + * won't run the queue on return, resulting in hang. + * - Usually, back inserted requests won't be merged + * with anything. There's no point in delaying queue + * processing. + */ + blk_remove_plug(q); + q->request_fn(q); + break; + + case ELEVATOR_INSERT_SORT: + BUG_ON(!blk_fs_request(rq)); + rq->flags |= REQ_SORTED; + q->elevator->ops->elevator_add_req_fn(q, rq); + break; + + default: + printk(KERN_ERR "%s: bad insertion point %d\n", + __FUNCTION__, where); + BUG(); + } + + if (blk_queue_plugged(q)) { + int nrq = q->rq.count[READ] + q->rq.count[WRITE] + - q->in_flight; + + if (nrq >= q->unplug_thresh) + __generic_unplug_device(q); + } } void elv_add_request(request_queue_t *q, struct request *rq, int where, @@ -353,13 +426,19 @@ void elv_add_request(request_queue_t *q, struct request *rq, int where, static inline struct request *__elv_next_request(request_queue_t *q) { - struct request *rq = q->elevator->ops->elevator_next_req_fn(q); + struct request *rq; + + if (unlikely(list_empty(&q->queue_head) && + !q->elevator->ops->elevator_dispatch_fn(q, 0))) + return NULL; + + rq = list_entry_rq(q->queue_head.next); /* * if this is a barrier write and the device has to issue a * flush sequence to support it, check how far we are */ - if (rq && blk_fs_request(rq) && blk_barrier_rq(rq)) { + if (blk_fs_request(rq) && blk_barrier_rq(rq)) { BUG_ON(q->ordered == QUEUE_ORDERED_NONE); if (q->ordered == QUEUE_ORDERED_FLUSH && @@ -376,16 +455,34 @@ struct request *elv_next_request(request_queue_t *q) int ret; while ((rq = __elv_next_request(q)) != NULL) { - /* - * just mark as started even if we don't start it, a request - * that has been delayed should not be passed by new incoming - * requests - */ - rq->flags |= REQ_STARTED; + if (!(rq->flags & REQ_STARTED)) { + elevator_t *e = q->elevator; + + /* + * This is the first time the device driver + * sees this request (possibly after + * requeueing). Notify IO scheduler. + */ + if (blk_sorted_rq(rq) && + e->ops->elevator_activate_req_fn) + e->ops->elevator_activate_req_fn(q, rq); + + /* + * just mark as started even if we don't start + * it, a request that has been delayed should + * not be passed by new incoming requests + */ + rq->flags |= REQ_STARTED; + } if (rq == q->last_merge) q->last_merge = NULL; + if (!q->boundary_rq || q->boundary_rq == rq) { + q->last_sector = rq_last_sector(rq); + q->boundary_rq = NULL; + } + if ((rq->flags & REQ_DONTPREP) || !q->prep_rq_fn) break; @@ -396,9 +493,9 @@ struct request *elv_next_request(request_queue_t *q) /* * the request may have been (partially) prepped. * we need to keep this request in the front to - * avoid resource deadlock. turn on softbarrier. + * avoid resource deadlock. REQ_STARTED will + * prevent other fs requests from passing this one. */ - rq->flags |= REQ_SOFTBARRIER; rq = NULL; break; } else if (ret == BLKPREP_KILL) { @@ -421,16 +518,16 @@ struct request *elv_next_request(request_queue_t *q) return rq; } -void elv_remove_request(request_queue_t *q, struct request *rq) +void elv_dequeue_request(request_queue_t *q, struct request *rq) { - elevator_t *e = q->elevator; + BUG_ON(list_empty(&rq->queuelist)); + + list_del_init(&rq->queuelist); /* * the time frame between a request being removed from the lists * and to it is freed is accounted as io that is in progress at - * the driver side. note that we only account requests that the - * driver has seen (REQ_STARTED set), to avoid false accounting - * for request-request merges + * the driver side. */ if (blk_account_rq(rq)) q->in_flight++; @@ -444,19 +541,19 @@ void elv_remove_request(request_queue_t *q, struct request *rq) */ if (rq == q->last_merge) q->last_merge = NULL; - - if (e->ops->elevator_remove_req_fn) - e->ops->elevator_remove_req_fn(q, rq); } int elv_queue_empty(request_queue_t *q) { elevator_t *e = q->elevator; + if (!list_empty(&q->queue_head)) + return 0; + if (e->ops->elevator_queue_empty_fn) return e->ops->elevator_queue_empty_fn(q); - return list_empty(&q->queue_head); + return 1; } struct request *elv_latter_request(request_queue_t *q, struct request *rq) @@ -528,11 +625,11 @@ void elv_completed_request(request_queue_t *q, struct request *rq) /* * request is released from the driver, io must be done */ - if (blk_account_rq(rq)) + if (blk_account_rq(rq)) { q->in_flight--; - - if (e->ops->elevator_completed_req_fn) - e->ops->elevator_completed_req_fn(q, rq); + if (blk_sorted_rq(rq) && e->ops->elevator_completed_req_fn) + e->ops->elevator_completed_req_fn(q, rq); + } } int elv_register_queue(struct request_queue *q) @@ -705,11 +802,12 @@ ssize_t elv_iosched_show(request_queue_t *q, char *name) return len; } +EXPORT_SYMBOL(elv_dispatch_insert); EXPORT_SYMBOL(elv_add_request); EXPORT_SYMBOL(__elv_add_request); EXPORT_SYMBOL(elv_requeue_request); EXPORT_SYMBOL(elv_next_request); -EXPORT_SYMBOL(elv_remove_request); +EXPORT_SYMBOL(elv_dequeue_request); EXPORT_SYMBOL(elv_queue_empty); EXPORT_SYMBOL(elv_completed_request); EXPORT_SYMBOL(elevator_exit); diff --git a/drivers/block/ll_rw_blk.c b/drivers/block/ll_rw_blk.c index 4e2b1b06b411..d2a66fd309c3 100644 --- a/drivers/block/ll_rw_blk.c +++ b/drivers/block/ll_rw_blk.c @@ -353,6 +353,8 @@ static void blk_pre_flush_end_io(struct request *flush_rq) struct request *rq = flush_rq->end_io_data; request_queue_t *q = rq->q; + elv_completed_request(q, flush_rq); + rq->flags |= REQ_BAR_PREFLUSH; if (!flush_rq->errors) @@ -369,6 +371,8 @@ static void blk_post_flush_end_io(struct request *flush_rq) struct request *rq = flush_rq->end_io_data; request_queue_t *q = rq->q; + elv_completed_request(q, flush_rq); + rq->flags |= REQ_BAR_POSTFLUSH; q->end_flush_fn(q, flush_rq); @@ -408,8 +412,6 @@ struct request *blk_start_pre_flush(request_queue_t *q, struct request *rq) if (!list_empty(&rq->queuelist)) blkdev_dequeue_request(rq); - elv_deactivate_request(q, rq); - flush_rq->end_io_data = rq; flush_rq->end_io = blk_pre_flush_end_io; @@ -1040,6 +1042,7 @@ EXPORT_SYMBOL(blk_queue_invalidate_tags); static char *rq_flags[] = { "REQ_RW", "REQ_FAILFAST", + "REQ_SORTED", "REQ_SOFTBARRIER", "REQ_HARDBARRIER", "REQ_CMD", @@ -2456,6 +2459,8 @@ static void __blk_put_request(request_queue_t *q, struct request *req) if (unlikely(--req->ref_count)) return; + elv_completed_request(q, req); + req->rq_status = RQ_INACTIVE; req->rl = NULL; @@ -2466,8 +2471,6 @@ static void __blk_put_request(request_queue_t *q, struct request *req) if (rl) { int rw = rq_data_dir(req); - elv_completed_request(q, req); - BUG_ON(!list_empty(&req->queuelist)); blk_free_request(q, req); @@ -2477,14 +2480,14 @@ static void __blk_put_request(request_queue_t *q, struct request *req) void blk_put_request(struct request *req) { + unsigned long flags; + request_queue_t *q = req->q; + /* - * if req->rl isn't set, this request didnt originate from the - * block layer, so it's safe to just disregard it + * Gee, IDE calls in w/ NULL q. Fix IDE and remove the + * following if (q) test. */ - if (req->rl) { - unsigned long flags; - request_queue_t *q = req->q; - + if (q) { spin_lock_irqsave(q->queue_lock, flags); __blk_put_request(q, req); spin_unlock_irqrestore(q->queue_lock, flags); diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h index efdc9b5bc05c..2c7b9154927a 100644 --- a/include/linux/blkdev.h +++ b/include/linux/blkdev.h @@ -203,6 +203,7 @@ struct request { enum rq_flag_bits { __REQ_RW, /* not set, read. set, write */ __REQ_FAILFAST, /* no low level driver retries */ + __REQ_SORTED, /* elevator knows about this request */ __REQ_SOFTBARRIER, /* may not be passed by ioscheduler */ __REQ_HARDBARRIER, /* may not be passed by drive either */ __REQ_CMD, /* is a regular fs rw request */ @@ -235,6 +236,7 @@ enum rq_flag_bits { #define REQ_RW (1 << __REQ_RW) #define REQ_FAILFAST (1 << __REQ_FAILFAST) +#define REQ_SORTED (1 << __REQ_SORTED) #define REQ_SOFTBARRIER (1 << __REQ_SOFTBARRIER) #define REQ_HARDBARRIER (1 << __REQ_HARDBARRIER) #define REQ_CMD (1 << __REQ_CMD) @@ -332,6 +334,13 @@ struct request_queue prepare_flush_fn *prepare_flush_fn; end_flush_fn *end_flush_fn; + /* + * Dispatch queue sorting + */ + sector_t last_sector; + struct request *boundary_rq; + unsigned int max_back_kb; + /* * Auto-unplugging state */ @@ -454,6 +463,7 @@ enum { #define blk_pm_request(rq) \ ((rq)->flags & (REQ_PM_SUSPEND | REQ_PM_RESUME)) +#define blk_sorted_rq(rq) ((rq)->flags & REQ_SORTED) #define blk_barrier_rq(rq) ((rq)->flags & REQ_HARDBARRIER) #define blk_barrier_preflush(rq) ((rq)->flags & REQ_BAR_PREFLUSH) #define blk_barrier_postflush(rq) ((rq)->flags & REQ_BAR_POSTFLUSH) @@ -611,12 +621,7 @@ extern void end_request(struct request *req, int uptodate); static inline void blkdev_dequeue_request(struct request *req) { - BUG_ON(list_empty(&req->queuelist)); - - list_del_init(&req->queuelist); - - if (req->rl) - elv_remove_request(req->q, req); + elv_dequeue_request(req->q, req); } /* diff --git a/include/linux/elevator.h b/include/linux/elevator.h index ea6bbc2d7407..76f4f6920744 100644 --- a/include/linux/elevator.h +++ b/include/linux/elevator.h @@ -8,18 +8,17 @@ typedef void (elevator_merge_req_fn) (request_queue_t *, struct request *, struc typedef void (elevator_merged_fn) (request_queue_t *, struct request *); -typedef struct request *(elevator_next_req_fn) (request_queue_t *); +typedef int (elevator_dispatch_fn) (request_queue_t *, int); -typedef void (elevator_add_req_fn) (request_queue_t *, struct request *, int); +typedef void (elevator_add_req_fn) (request_queue_t *, struct request *); typedef int (elevator_queue_empty_fn) (request_queue_t *); -typedef void (elevator_remove_req_fn) (request_queue_t *, struct request *); -typedef void (elevator_requeue_req_fn) (request_queue_t *, struct request *); typedef struct request *(elevator_request_list_fn) (request_queue_t *, struct request *); typedef void (elevator_completed_req_fn) (request_queue_t *, struct request *); typedef int (elevator_may_queue_fn) (request_queue_t *, int, struct bio *); typedef int (elevator_set_req_fn) (request_queue_t *, struct request *, struct bio *, int); typedef void (elevator_put_req_fn) (request_queue_t *, struct request *); +typedef void (elevator_activate_req_fn) (request_queue_t *, struct request *); typedef void (elevator_deactivate_req_fn) (request_queue_t *, struct request *); typedef int (elevator_init_fn) (request_queue_t *, elevator_t *); @@ -31,10 +30,9 @@ struct elevator_ops elevator_merged_fn *elevator_merged_fn; elevator_merge_req_fn *elevator_merge_req_fn; - elevator_next_req_fn *elevator_next_req_fn; + elevator_dispatch_fn *elevator_dispatch_fn; elevator_add_req_fn *elevator_add_req_fn; - elevator_remove_req_fn *elevator_remove_req_fn; - elevator_requeue_req_fn *elevator_requeue_req_fn; + elevator_activate_req_fn *elevator_activate_req_fn; elevator_deactivate_req_fn *elevator_deactivate_req_fn; elevator_queue_empty_fn *elevator_queue_empty_fn; @@ -81,15 +79,15 @@ struct elevator_queue /* * block elevator interface */ +extern void elv_dispatch_insert(request_queue_t *, struct request *, int); extern void elv_add_request(request_queue_t *, struct request *, int, int); extern void __elv_add_request(request_queue_t *, struct request *, int, int); extern int elv_merge(request_queue_t *, struct request **, struct bio *); extern void elv_merge_requests(request_queue_t *, struct request *, struct request *); extern void elv_merged_request(request_queue_t *, struct request *); -extern void elv_remove_request(request_queue_t *, struct request *); +extern void elv_dequeue_request(request_queue_t *, struct request *); extern void elv_requeue_request(request_queue_t *, struct request *); -extern void elv_deactivate_request(request_queue_t *, struct request *); extern int elv_queue_empty(request_queue_t *); extern struct request *elv_next_request(struct request_queue *q); extern struct request *elv_former_request(request_queue_t *, struct request *); -- cgit v1.2.3 From 1b47f531e244e339b648dfff071c086f936e49e1 Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Thu, 20 Oct 2005 16:37:00 +0200 Subject: [PATCH] generic dispatch fixes - Split elv_dispatch_insert() into two functions - Rename rq_last_sector() to rq_end_sector() Signed-off-by: Jens Axboe --- drivers/block/elevator.c | 33 ++++++++++++--------------------- include/linux/blkdev.h | 13 ++++++++++++- include/linux/elevator.h | 4 +++- 3 files changed, 27 insertions(+), 23 deletions(-) (limited to 'include/linux') diff --git a/drivers/block/elevator.c b/drivers/block/elevator.c index a27555908d35..237e43d760c8 100644 --- a/drivers/block/elevator.c +++ b/drivers/block/elevator.c @@ -40,11 +40,6 @@ static DEFINE_SPINLOCK(elv_list_lock); static LIST_HEAD(elv_list); -static inline sector_t rq_last_sector(struct request *rq) -{ - return rq->sector + rq->nr_sectors; -} - /* * can we safely merge with this request? */ @@ -148,7 +143,7 @@ static int elevator_attach(request_queue_t *q, struct elevator_type *e, INIT_LIST_HEAD(&q->queue_head); q->last_merge = NULL; q->elevator = eq; - q->last_sector = 0; + q->end_sector = 0; q->boundary_rq = NULL; q->max_back_kb = 0; @@ -233,29 +228,25 @@ void elevator_exit(elevator_t *e) kfree(e); } +void elv_dispatch_insert_tail(request_queue_t *q, struct request *rq) +{ +} + /* * Insert rq into dispatch queue of q. Queue lock must be held on * entry. If sort != 0, rq is sort-inserted; otherwise, rq will be * appended to the dispatch queue. To be used by specific elevators. */ -void elv_dispatch_insert(request_queue_t *q, struct request *rq, int sort) +void elv_dispatch_sort(request_queue_t *q, struct request *rq) { sector_t boundary; unsigned max_back; struct list_head *entry; - if (!sort) { - /* Specific elevator is performing sort. Step away. */ - q->last_sector = rq_last_sector(rq); - q->boundary_rq = rq; - list_add_tail(&rq->queuelist, &q->queue_head); - return; - } - - boundary = q->last_sector; + boundary = q->end_sector; max_back = q->max_back_kb * 2; boundary = boundary > max_back ? boundary - max_back : 0; - + list_for_each_prev(entry, &q->queue_head) { struct request *pos = list_entry_rq(entry); @@ -343,10 +334,10 @@ void __elv_add_request(request_queue_t *q, struct request *rq, int where, where = ELEVATOR_INSERT_BACK; /* - * this request is scheduling boundary, update last_sector + * this request is scheduling boundary, update end_sector */ if (blk_fs_request(rq)) { - q->last_sector = rq_last_sector(rq); + q->end_sector = rq_end_sector(rq); q->boundary_rq = rq; } } @@ -479,7 +470,7 @@ struct request *elv_next_request(request_queue_t *q) q->last_merge = NULL; if (!q->boundary_rq || q->boundary_rq == rq) { - q->last_sector = rq_last_sector(rq); + q->end_sector = rq_end_sector(rq); q->boundary_rq = NULL; } @@ -802,7 +793,7 @@ ssize_t elv_iosched_show(request_queue_t *q, char *name) return len; } -EXPORT_SYMBOL(elv_dispatch_insert); +EXPORT_SYMBOL(elv_dispatch_sort); EXPORT_SYMBOL(elv_add_request); EXPORT_SYMBOL(__elv_add_request); EXPORT_SYMBOL(elv_requeue_request); diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h index 2c7b9154927a..a2e496f285a4 100644 --- a/include/linux/blkdev.h +++ b/include/linux/blkdev.h @@ -337,7 +337,7 @@ struct request_queue /* * Dispatch queue sorting */ - sector_t last_sector; + sector_t end_sector; struct request *boundary_rq; unsigned int max_back_kb; @@ -624,6 +624,17 @@ static inline void blkdev_dequeue_request(struct request *req) elv_dequeue_request(req->q, req); } +/* + * This should be in elevator.h, but that requires pulling in rq and q + */ +static inline void elv_dispatch_add_tail(struct request_queue *q, + struct request *rq) +{ + q->end_sector = rq_end_sector(rq); + q->boundary_rq = rq; + list_add_tail(&rq->queuelist, &q->queue_head); +} + /* * Access functions for manipulating queue properties */ diff --git a/include/linux/elevator.h b/include/linux/elevator.h index 76f4f6920744..f9b5ec38121a 100644 --- a/include/linux/elevator.h +++ b/include/linux/elevator.h @@ -79,7 +79,7 @@ struct elevator_queue /* * block elevator interface */ -extern void elv_dispatch_insert(request_queue_t *, struct request *, int); +extern void elv_dispatch_sort(request_queue_t *, struct request *); extern void elv_add_request(request_queue_t *, struct request *, int, int); extern void __elv_add_request(request_queue_t *, struct request *, int, int); extern int elv_merge(request_queue_t *, struct request **, struct bio *); @@ -140,4 +140,6 @@ enum { ELV_MQUEUE_MUST, }; +#define rq_end_sector(rq) ((rq)->sector + (rq)->nr_sectors) + #endif -- cgit v1.2.3 From 06b86245c052963029bfd9020ca1f08ceb66f85a Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Thu, 20 Oct 2005 16:46:23 +0200 Subject: [PATCH] 03/05 move last_merge handlin into generic elevator code Currently, both generic elevator code and specific ioscheds participate in the management and usage of last_merge. This and the following patches move last_merge handling into generic elevator code. Signed-off-by: Tejun Heo Signed-off-by: Jens Axboe --- drivers/block/elevator.c | 43 ++++++++++++++++++------------------------- include/linux/blkdev.h | 3 +++ 2 files changed, 21 insertions(+), 25 deletions(-) (limited to 'include/linux') diff --git a/drivers/block/elevator.c b/drivers/block/elevator.c index 29d6c8237ab3..415144372c75 100644 --- a/drivers/block/elevator.c +++ b/drivers/block/elevator.c @@ -83,15 +83,6 @@ inline int elv_try_merge(struct request *__rq, struct bio *bio) } EXPORT_SYMBOL(elv_try_merge); -inline int elv_try_last_merge(request_queue_t *q, struct bio *bio) -{ - if (q->last_merge) - return elv_try_merge(q->last_merge, bio); - - return ELEVATOR_NO_MERGE; -} -EXPORT_SYMBOL(elv_try_last_merge); - static struct elevator_type *elevator_find(const char *name) { struct elevator_type *e = NULL; @@ -239,6 +230,9 @@ void elv_dispatch_sort(request_queue_t *q, struct request *rq) unsigned max_back; struct list_head *entry; + if (q->last_merge == rq) + q->last_merge = NULL; + boundary = q->end_sector; max_back = q->max_back_kb * 2; boundary = boundary > max_back ? boundary - max_back : 0; @@ -265,6 +259,15 @@ void elv_dispatch_sort(request_queue_t *q, struct request *rq) int elv_merge(request_queue_t *q, struct request **req, struct bio *bio) { elevator_t *e = q->elevator; + int ret; + + if (q->last_merge) { + ret = elv_try_merge(q->last_merge, bio); + if (ret != ELEVATOR_NO_MERGE) { + *req = q->last_merge; + return ret; + } + } if (e->ops->elevator_merge_fn) return e->ops->elevator_merge_fn(q, req, bio); @@ -278,6 +281,8 @@ void elv_merged_request(request_queue_t *q, struct request *rq) if (e->ops->elevator_merged_fn) e->ops->elevator_merged_fn(q, rq); + + q->last_merge = rq; } void elv_merge_requests(request_queue_t *q, struct request *rq, @@ -285,11 +290,10 @@ void elv_merge_requests(request_queue_t *q, struct request *rq, { elevator_t *e = q->elevator; - if (q->last_merge == next) - q->last_merge = NULL; - if (e->ops->elevator_merge_req_fn) e->ops->elevator_merge_req_fn(q, rq, next); + + q->last_merge = rq; } void elv_requeue_request(request_queue_t *q, struct request *rq) @@ -384,6 +388,8 @@ void __elv_add_request(request_queue_t *q, struct request *rq, int where, BUG_ON(!blk_fs_request(rq)); rq->flags |= REQ_SORTED; q->elevator->ops->elevator_add_req_fn(q, rq); + if (q->last_merge == NULL && rq_mergeable(rq)) + q->last_merge = rq; break; default: @@ -462,9 +468,6 @@ struct request *elv_next_request(request_queue_t *q) rq->flags |= REQ_STARTED; } - if (rq == q->last_merge) - q->last_merge = NULL; - if (!q->boundary_rq || q->boundary_rq == rq) { q->end_sector = rq_end_sector(rq); q->boundary_rq = NULL; @@ -518,16 +521,6 @@ void elv_dequeue_request(request_queue_t *q, struct request *rq) */ if (blk_account_rq(rq)) q->in_flight++; - - /* - * the main clearing point for q->last_merge is on retrieval of - * request by driver (it calls elv_next_request()), but it _can_ - * also happen here if a request is added to the queue but later - * deleted without ever being given to driver (merged with another - * request). - */ - if (rq == q->last_merge) - q->last_merge = NULL; } int elv_queue_empty(request_queue_t *q) diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h index a2e496f285a4..bd55b1df8997 100644 --- a/include/linux/blkdev.h +++ b/include/linux/blkdev.h @@ -630,6 +630,9 @@ static inline void blkdev_dequeue_request(struct request *req) static inline void elv_dispatch_add_tail(struct request_queue *q, struct request *rq) { + if (q->last_merge == rq) + q->last_merge = NULL; + q->end_sector = rq_end_sector(rq); q->boundary_rq = rq; list_add_tail(&rq->queuelist, &q->queue_head); -- cgit v1.2.3 From cb19833dccb32f97cacbfff834b53523915f13f6 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Mon, 24 Oct 2005 08:35:58 +0200 Subject: [BLOCK] kill generic max_back_kb handling This patch kills max_back_kb handling from elv_dispatch_sort() and kills max_back_kb field from struct request_queue. Signed-off-by: Tejun Heo Signed-off-by: Jens Axboe --- drivers/block/elevator.c | 6 +----- include/linux/blkdev.h | 1 - 2 files changed, 1 insertion(+), 6 deletions(-) (limited to 'include/linux') diff --git a/drivers/block/elevator.c b/drivers/block/elevator.c index 415144372c75..af2388e73f61 100644 --- a/drivers/block/elevator.c +++ b/drivers/block/elevator.c @@ -136,7 +136,6 @@ static int elevator_attach(request_queue_t *q, struct elevator_type *e, q->elevator = eq; q->end_sector = 0; q->boundary_rq = NULL; - q->max_back_kb = 0; if (eq->ops->elevator_init_fn) ret = eq->ops->elevator_init_fn(q, eq); @@ -227,16 +226,13 @@ void elevator_exit(elevator_t *e) void elv_dispatch_sort(request_queue_t *q, struct request *rq) { sector_t boundary; - unsigned max_back; struct list_head *entry; if (q->last_merge == rq) q->last_merge = NULL; boundary = q->end_sector; - max_back = q->max_back_kb * 2; - boundary = boundary > max_back ? boundary - max_back : 0; - + list_for_each_prev(entry, &q->queue_head) { struct request *pos = list_entry_rq(entry); diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h index bd55b1df8997..159dbcd2eb59 100644 --- a/include/linux/blkdev.h +++ b/include/linux/blkdev.h @@ -339,7 +339,6 @@ struct request_queue */ sector_t end_sector; struct request *boundary_rq; - unsigned int max_back_kb; /* * Auto-unplugging state -- cgit v1.2.3 From cb98fc8bb9c141009e2bda99c0db39d387e142cf Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Fri, 28 Oct 2005 08:29:39 +0200 Subject: [BLOCK] Reimplement elevator switch This patch reimplements elevator switch. This patch assumes generic dispatch queue patchset is applied. * Each request is tagged with REQ_ELVPRIV flag if it has its elevator private data set. * Requests which doesn't have REQ_ELVPRIV flag set never enter iosched. They are always directly back inserted to dispatch queue. Of course, elevator_put_req_fn is called only for requests which have its REQ_ELVPRIV set. * Request queue maintains the current number of requests which have its elevator data set (elevator_set_req_fn called) in q->rq->elvpriv. * If a request queue has QUEUE_FLAG_BYPASS set, elevator private data is not allocated for new requests. To switch to another iosched, we set QUEUE_FLAG_BYPASS and wait until elvpriv goes to zero; then, we attach the new iosched and clears QUEUE_FLAG_BYPASS. New implementation is much simpler and main code paths are less cluttered, IMHO. Signed-off-by: Tejun Heo Signed-off-by: Jens Axboe --- drivers/block/elevator.c | 78 ++++++++++++------------- drivers/block/ll_rw_blk.c | 142 +++++++++------------------------------------- include/linux/blkdev.h | 10 ++-- 3 files changed, 66 insertions(+), 164 deletions(-) (limited to 'include/linux') diff --git a/drivers/block/elevator.c b/drivers/block/elevator.c index af2388e73f61..272d93946621 100644 --- a/drivers/block/elevator.c +++ b/drivers/block/elevator.c @@ -34,6 +34,7 @@ #include #include #include +#include #include @@ -131,11 +132,7 @@ static int elevator_attach(request_queue_t *q, struct elevator_type *e, eq->ops = &e->ops; eq->elevator_type = e; - INIT_LIST_HEAD(&q->queue_head); - q->last_merge = NULL; q->elevator = eq; - q->end_sector = 0; - q->boundary_rq = NULL; if (eq->ops->elevator_init_fn) ret = eq->ops->elevator_init_fn(q, eq); @@ -184,6 +181,12 @@ int elevator_init(request_queue_t *q, char *name) struct elevator_queue *eq; int ret = 0; + INIT_LIST_HEAD(&q->queue_head); + q->last_merge = NULL; + q->end_sector = 0; + q->boundary_rq = NULL; + q->max_back_kb = 0; + elevator_setup_default(); if (!name) @@ -336,23 +339,14 @@ void __elv_add_request(request_queue_t *q, struct request *rq, int where, q->end_sector = rq_end_sector(rq); q->boundary_rq = rq; } - } + } else if (!(rq->flags & REQ_ELVPRIV) && where == ELEVATOR_INSERT_SORT) + where = ELEVATOR_INSERT_BACK; if (plug) blk_plug_device(q); rq->q = q; - if (unlikely(test_bit(QUEUE_FLAG_DRAIN, &q->queue_flags))) { - /* - * if drain is set, store the request "locally". when the drain - * is finished, the requests will be handed ordered to the io - * scheduler - */ - list_add_tail(&rq->queuelist, &q->drain_list); - return; - } - switch (where) { case ELEVATOR_INSERT_FRONT: rq->flags |= REQ_SOFTBARRIER; @@ -659,25 +653,36 @@ EXPORT_SYMBOL_GPL(elv_unregister); * switch to new_e io scheduler. be careful not to introduce deadlocks - * we don't free the old io scheduler, before we have allocated what we * need for the new one. this way we have a chance of going back to the old - * one, if the new one fails init for some reason. we also do an intermediate - * switch to noop to ensure safety with stack-allocated requests, since they - * don't originate from the block layer allocator. noop is safe here, because - * it never needs to touch the elevator itself for completion events. DRAIN - * flags will make sure we don't touch it for additions either. + * one, if the new one fails init for some reason. */ static void elevator_switch(request_queue_t *q, struct elevator_type *new_e) { - elevator_t *e = kmalloc(sizeof(elevator_t), GFP_KERNEL); - struct elevator_type *noop_elevator = NULL; - elevator_t *old_elevator; + elevator_t *old_elevator, *e; + /* + * Allocate new elevator + */ + e = kmalloc(sizeof(elevator_t), GFP_KERNEL); if (!e) goto error; /* - * first step, drain requests from the block freelist + * Turn on BYPASS and drain all requests w/ elevator private data */ - blk_wait_queue_drained(q, 0); + spin_lock_irq(q->queue_lock); + + set_bit(QUEUE_FLAG_BYPASS, &q->queue_flags); + + while (q->elevator->ops->elevator_dispatch_fn(q, 1)) + ; + + while (q->rq.elvpriv) { + spin_unlock_irq(q->queue_lock); + msleep(100); + spin_lock_irq(q->queue_lock); + } + + spin_unlock_irq(q->queue_lock); /* * unregister old elevator data @@ -685,18 +690,6 @@ static void elevator_switch(request_queue_t *q, struct elevator_type *new_e) elv_unregister_queue(q); old_elevator = q->elevator; - /* - * next step, switch to noop since it uses no private rq structures - * and doesn't allocate any memory for anything. then wait for any - * non-fs requests in-flight - */ - noop_elevator = elevator_get("noop"); - spin_lock_irq(q->queue_lock); - elevator_attach(q, noop_elevator, e); - spin_unlock_irq(q->queue_lock); - - blk_wait_queue_drained(q, 1); - /* * attach and start new elevator */ @@ -707,11 +700,10 @@ static void elevator_switch(request_queue_t *q, struct elevator_type *new_e) goto fail_register; /* - * finally exit old elevator and start queue again + * finally exit old elevator and turn off BYPASS. */ elevator_exit(old_elevator); - blk_finish_queue_drain(q); - elevator_put(noop_elevator); + clear_bit(QUEUE_FLAG_BYPASS, &q->queue_flags); return; fail_register: @@ -720,13 +712,13 @@ fail_register: * one again (along with re-adding the sysfs dir) */ elevator_exit(e); + e = NULL; fail: q->elevator = old_elevator; elv_register_queue(q); - blk_finish_queue_drain(q); + clear_bit(QUEUE_FLAG_BYPASS, &q->queue_flags); + kfree(e); error: - if (noop_elevator) - elevator_put(noop_elevator); elevator_put(new_e); printk(KERN_ERR "elevator: switch to %s failed\n",new_e->elevator_name); } diff --git a/drivers/block/ll_rw_blk.c b/drivers/block/ll_rw_blk.c index d2a66fd309c3..f7c9931cb380 100644 --- a/drivers/block/ll_rw_blk.c +++ b/drivers/block/ll_rw_blk.c @@ -263,8 +263,6 @@ void blk_queue_make_request(request_queue_t * q, make_request_fn * mfn) blk_queue_bounce_limit(q, BLK_BOUNCE_HIGH); blk_queue_activity_fn(q, NULL, NULL); - - INIT_LIST_HEAD(&q->drain_list); } EXPORT_SYMBOL(blk_queue_make_request); @@ -1050,6 +1048,7 @@ static char *rq_flags[] = { "REQ_STARTED", "REQ_DONTPREP", "REQ_QUEUED", + "REQ_ELVPRIV", "REQ_PC", "REQ_BLOCK_PC", "REQ_SENSE", @@ -1640,9 +1639,9 @@ static int blk_init_free_list(request_queue_t *q) rl->count[READ] = rl->count[WRITE] = 0; rl->starved[READ] = rl->starved[WRITE] = 0; + rl->elvpriv = 0; init_waitqueue_head(&rl->wait[READ]); init_waitqueue_head(&rl->wait[WRITE]); - init_waitqueue_head(&rl->drain); rl->rq_pool = mempool_create_node(BLKDEV_MIN_RQ, mempool_alloc_slab, mempool_free_slab, request_cachep, q->node); @@ -1785,12 +1784,14 @@ EXPORT_SYMBOL(blk_get_queue); static inline void blk_free_request(request_queue_t *q, struct request *rq) { - elv_put_request(q, rq); + if (rq->flags & REQ_ELVPRIV) + elv_put_request(q, rq); mempool_free(rq, q->rq.rq_pool); } static inline struct request * -blk_alloc_request(request_queue_t *q, int rw, struct bio *bio, int gfp_mask) +blk_alloc_request(request_queue_t *q, int rw, struct bio *bio, + int priv, int gfp_mask) { struct request *rq = mempool_alloc(q->rq.rq_pool, gfp_mask); @@ -1803,11 +1804,15 @@ blk_alloc_request(request_queue_t *q, int rw, struct bio *bio, int gfp_mask) */ rq->flags = rw; - if (!elv_set_request(q, rq, bio, gfp_mask)) - return rq; + if (priv) { + if (unlikely(elv_set_request(q, rq, bio, gfp_mask))) { + mempool_free(rq, q->rq.rq_pool); + return NULL; + } + rq->flags |= REQ_ELVPRIV; + } - mempool_free(rq, q->rq.rq_pool); - return NULL; + return rq; } /* @@ -1863,22 +1868,18 @@ static void __freed_request(request_queue_t *q, int rw) * A request has just been released. Account for it, update the full and * congestion status, wake up any waiters. Called under q->queue_lock. */ -static void freed_request(request_queue_t *q, int rw) +static void freed_request(request_queue_t *q, int rw, int priv) { struct request_list *rl = &q->rq; rl->count[rw]--; + if (priv) + rl->elvpriv--; __freed_request(q, rw); if (unlikely(rl->starved[rw ^ 1])) __freed_request(q, rw ^ 1); - - if (!rl->count[READ] && !rl->count[WRITE]) { - smp_mb(); - if (unlikely(waitqueue_active(&rl->drain))) - wake_up(&rl->drain); - } } #define blkdev_free_rq(list) list_entry((list)->next, struct request, queuelist) @@ -1893,9 +1894,7 @@ static struct request *get_request(request_queue_t *q, int rw, struct bio *bio, struct request *rq = NULL; struct request_list *rl = &q->rq; struct io_context *ioc = current_io_context(GFP_ATOMIC); - - if (unlikely(test_bit(QUEUE_FLAG_DRAIN, &q->queue_flags))) - goto out; + int priv; if (rl->count[rw]+1 >= q->nr_requests) { /* @@ -1940,9 +1939,14 @@ get_rq: rl->starved[rw] = 0; if (rl->count[rw] >= queue_congestion_on_threshold(q)) set_queue_congested(q, rw); + + priv = !test_bit(QUEUE_FLAG_BYPASS, &q->queue_flags); + if (priv) + rl->elvpriv++; + spin_unlock_irq(q->queue_lock); - rq = blk_alloc_request(q, rw, bio, gfp_mask); + rq = blk_alloc_request(q, rw, bio, priv, gfp_mask); if (!rq) { /* * Allocation failed presumably due to memory. Undo anything @@ -1952,7 +1956,7 @@ get_rq: * wait queue, but this is pretty rare. */ spin_lock_irq(q->queue_lock); - freed_request(q, rw); + freed_request(q, rw, priv); /* * in the very unlikely event that allocation failed and no @@ -2470,11 +2474,12 @@ static void __blk_put_request(request_queue_t *q, struct request *req) */ if (rl) { int rw = rq_data_dir(req); + int priv = req->flags & REQ_ELVPRIV; BUG_ON(!list_empty(&req->queuelist)); blk_free_request(q, req); - freed_request(q, rw); + freed_request(q, rw, priv); } } @@ -2802,97 +2807,6 @@ static inline void blk_partition_remap(struct bio *bio) } } -void blk_finish_queue_drain(request_queue_t *q) -{ - struct request_list *rl = &q->rq; - struct request *rq; - int requeued = 0; - - spin_lock_irq(q->queue_lock); - clear_bit(QUEUE_FLAG_DRAIN, &q->queue_flags); - - while (!list_empty(&q->drain_list)) { - rq = list_entry_rq(q->drain_list.next); - - list_del_init(&rq->queuelist); - elv_requeue_request(q, rq); - requeued++; - } - - if (requeued) - q->request_fn(q); - - spin_unlock_irq(q->queue_lock); - - wake_up(&rl->wait[0]); - wake_up(&rl->wait[1]); - wake_up(&rl->drain); -} - -static int wait_drain(request_queue_t *q, struct request_list *rl, int dispatch) -{ - int wait = rl->count[READ] + rl->count[WRITE]; - - if (dispatch) - wait += !list_empty(&q->queue_head); - - return wait; -} - -/* - * We rely on the fact that only requests allocated through blk_alloc_request() - * have io scheduler private data structures associated with them. Any other - * type of request (allocated on stack or through kmalloc()) should not go - * to the io scheduler core, but be attached to the queue head instead. - */ -void blk_wait_queue_drained(request_queue_t *q, int wait_dispatch) -{ - struct request_list *rl = &q->rq; - DEFINE_WAIT(wait); - - spin_lock_irq(q->queue_lock); - set_bit(QUEUE_FLAG_DRAIN, &q->queue_flags); - - while (wait_drain(q, rl, wait_dispatch)) { - prepare_to_wait(&rl->drain, &wait, TASK_UNINTERRUPTIBLE); - - if (wait_drain(q, rl, wait_dispatch)) { - __generic_unplug_device(q); - spin_unlock_irq(q->queue_lock); - io_schedule(); - spin_lock_irq(q->queue_lock); - } - - finish_wait(&rl->drain, &wait); - } - - spin_unlock_irq(q->queue_lock); -} - -/* - * block waiting for the io scheduler being started again. - */ -static inline void block_wait_queue_running(request_queue_t *q) -{ - DEFINE_WAIT(wait); - - while (unlikely(test_bit(QUEUE_FLAG_DRAIN, &q->queue_flags))) { - struct request_list *rl = &q->rq; - - prepare_to_wait_exclusive(&rl->drain, &wait, - TASK_UNINTERRUPTIBLE); - - /* - * re-check the condition. avoids using prepare_to_wait() - * in the fast path (queue is running) - */ - if (test_bit(QUEUE_FLAG_DRAIN, &q->queue_flags)) - io_schedule(); - - finish_wait(&rl->drain, &wait); - } -} - static void handle_bad_sector(struct bio *bio) { char b[BDEVNAME_SIZE]; @@ -2988,8 +2902,6 @@ end_io: if (unlikely(test_bit(QUEUE_FLAG_DEAD, &q->queue_flags))) goto end_io; - block_wait_queue_running(q); - /* * If this device has partitions, remap block n * of partition p to block n+start(p) of the disk. diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h index 159dbcd2eb59..6186d5e2110f 100644 --- a/include/linux/blkdev.h +++ b/include/linux/blkdev.h @@ -107,9 +107,9 @@ typedef void (rq_end_io_fn)(struct request *); struct request_list { int count[2]; int starved[2]; + int elvpriv; mempool_t *rq_pool; wait_queue_head_t wait[2]; - wait_queue_head_t drain; }; #define BLK_MAX_CDB 16 @@ -211,6 +211,7 @@ enum rq_flag_bits { __REQ_STARTED, /* drive already may have started this one */ __REQ_DONTPREP, /* don't call prep for this one */ __REQ_QUEUED, /* uses queueing */ + __REQ_ELVPRIV, /* elevator private data attached */ /* * for ATA/ATAPI devices */ @@ -244,6 +245,7 @@ enum rq_flag_bits { #define REQ_STARTED (1 << __REQ_STARTED) #define REQ_DONTPREP (1 << __REQ_DONTPREP) #define REQ_QUEUED (1 << __REQ_QUEUED) +#define REQ_ELVPRIV (1 << __REQ_ELVPRIV) #define REQ_PC (1 << __REQ_PC) #define REQ_BLOCK_PC (1 << __REQ_BLOCK_PC) #define REQ_SENSE (1 << __REQ_SENSE) @@ -413,8 +415,6 @@ struct request_queue unsigned int sg_reserved_size; int node; - struct list_head drain_list; - /* * reserved for flush operations */ @@ -442,7 +442,7 @@ enum { #define QUEUE_FLAG_DEAD 5 /* queue being torn down */ #define QUEUE_FLAG_REENTER 6 /* Re-entrancy avoidance */ #define QUEUE_FLAG_PLUGGED 7 /* queue is plugged */ -#define QUEUE_FLAG_DRAIN 8 /* draining queue for sched switch */ +#define QUEUE_FLAG_BYPASS 8 /* don't use elevator, just do FIFO */ #define QUEUE_FLAG_FLUSH 9 /* doing barrier flush sequence */ #define blk_queue_plugged(q) test_bit(QUEUE_FLAG_PLUGGED, &(q)->queue_flags) @@ -668,8 +668,6 @@ extern void blk_dump_rq_flags(struct request *, char *); extern void generic_unplug_device(request_queue_t *); extern void __generic_unplug_device(request_queue_t *); extern long nr_blockdev_pages(void); -extern void blk_wait_queue_drained(request_queue_t *, int); -extern void blk_finish_queue_drain(request_queue_t *); int blk_get_queue(request_queue_t *); request_queue_t *blk_alloc_queue(int gfp_mask); -- cgit v1.2.3 From 64521d1a3ba7cc7fe4ab5640da83d88144f05340 Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Fri, 28 Oct 2005 08:30:39 +0200 Subject: [BLOCK] elevator switch fixes/cleanup - 100msec sleep is a little excessive, lots of requests can complete in that timeframe. Use 10msec instead. - Rename QUEUE_FLAG_BYPASS to QUEUE_FLAG_ELVSWITCH to indicate what is going on. Signed-off-by: Jens Axboe --- drivers/block/elevator.c | 8 ++++---- drivers/block/ll_rw_blk.c | 2 +- include/linux/blkdev.h | 2 +- 3 files changed, 6 insertions(+), 6 deletions(-) (limited to 'include/linux') diff --git a/drivers/block/elevator.c b/drivers/block/elevator.c index 272d93946621..cf9bedbb761b 100644 --- a/drivers/block/elevator.c +++ b/drivers/block/elevator.c @@ -671,14 +671,14 @@ static void elevator_switch(request_queue_t *q, struct elevator_type *new_e) */ spin_lock_irq(q->queue_lock); - set_bit(QUEUE_FLAG_BYPASS, &q->queue_flags); + set_bit(QUEUE_FLAG_ELVSWITCH, &q->queue_flags); while (q->elevator->ops->elevator_dispatch_fn(q, 1)) ; while (q->rq.elvpriv) { spin_unlock_irq(q->queue_lock); - msleep(100); + msleep(10); spin_lock_irq(q->queue_lock); } @@ -703,7 +703,7 @@ static void elevator_switch(request_queue_t *q, struct elevator_type *new_e) * finally exit old elevator and turn off BYPASS. */ elevator_exit(old_elevator); - clear_bit(QUEUE_FLAG_BYPASS, &q->queue_flags); + clear_bit(QUEUE_FLAG_ELVSWITCH, &q->queue_flags); return; fail_register: @@ -716,7 +716,7 @@ fail_register: fail: q->elevator = old_elevator; elv_register_queue(q); - clear_bit(QUEUE_FLAG_BYPASS, &q->queue_flags); + clear_bit(QUEUE_FLAG_ELVSWITCH, &q->queue_flags); kfree(e); error: elevator_put(new_e); diff --git a/drivers/block/ll_rw_blk.c b/drivers/block/ll_rw_blk.c index f7c9931cb380..fc12d019b19d 100644 --- a/drivers/block/ll_rw_blk.c +++ b/drivers/block/ll_rw_blk.c @@ -1940,7 +1940,7 @@ get_rq: if (rl->count[rw] >= queue_congestion_on_threshold(q)) set_queue_congested(q, rw); - priv = !test_bit(QUEUE_FLAG_BYPASS, &q->queue_flags); + priv = !test_bit(QUEUE_FLAG_ELVSWITCH, &q->queue_flags); if (priv) rl->elvpriv++; diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h index 6186d5e2110f..7d9d14d6d8f9 100644 --- a/include/linux/blkdev.h +++ b/include/linux/blkdev.h @@ -442,7 +442,7 @@ enum { #define QUEUE_FLAG_DEAD 5 /* queue being torn down */ #define QUEUE_FLAG_REENTER 6 /* Re-entrancy avoidance */ #define QUEUE_FLAG_PLUGGED 7 /* queue is plugged */ -#define QUEUE_FLAG_BYPASS 8 /* don't use elevator, just do FIFO */ +#define QUEUE_FLAG_ELVSWITCH 8 /* don't use elevator, just do FIFO */ #define QUEUE_FLAG_FLUSH 9 /* doing barrier flush sequence */ #define blk_queue_plugged(q) test_bit(QUEUE_FLAG_PLUGGED, &(q)->queue_flags) -- cgit v1.2.3 From af4ca457eaf2d6682059c18463eb106e2ce58198 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Fri, 21 Oct 2005 02:55:38 -0400 Subject: [PATCH] gfp_t: infrastructure Beginning of gfp_t annotations: - -Wbitwise added to CHECKFLAGS - old __bitwise renamed to __bitwise__ - __bitwise defined to either __bitwise__ or nothing, depending on __CHECK_ENDIAN__ being defined - gfp_t switched from __nocast to __bitwise__ - force cast to gfp_t added to __GFP_... constants - new helper - gfp_zone(); extracts zone bits out of gfp_t value and casts the result to int Signed-off-by: Al Viro Signed-off-by: Linus Torvalds --- Makefile | 2 +- fs/buffer.c | 2 +- include/linux/gfp.h | 39 ++++++++++++++++++++------------------- include/linux/types.h | 9 +++++++-- mm/mempolicy.c | 6 +++--- mm/page_alloc.c | 4 ++-- 6 files changed, 34 insertions(+), 28 deletions(-) (limited to 'include/linux') diff --git a/Makefile b/Makefile index 1fa7e5343464..f1d121f23025 100644 --- a/Makefile +++ b/Makefile @@ -334,7 +334,7 @@ KALLSYMS = scripts/kallsyms PERL = perl CHECK = sparse -CHECKFLAGS := -D__linux__ -Dlinux -D__STDC__ -Dunix -D__unix__ $(CF) +CHECKFLAGS := -D__linux__ -Dlinux -D__STDC__ -Dunix -D__unix__ -Wbitwise $(CF) MODFLAGS = -DMODULE CFLAGS_MODULE = $(MODFLAGS) AFLAGS_MODULE = $(MODFLAGS) diff --git a/fs/buffer.c b/fs/buffer.c index 1216c0d3c8ce..9657696fd6d7 100644 --- a/fs/buffer.c +++ b/fs/buffer.c @@ -502,7 +502,7 @@ static void free_more_memory(void) yield(); for_each_pgdat(pgdat) { - zones = pgdat->node_zonelists[GFP_NOFS&GFP_ZONEMASK].zones; + zones = pgdat->node_zonelists[gfp_zone(GFP_NOFS)].zones; if (*zones) try_to_free_pages(zones, GFP_NOFS); } diff --git a/include/linux/gfp.h b/include/linux/gfp.h index 3010e172394d..c3779432a723 100644 --- a/include/linux/gfp.h +++ b/include/linux/gfp.h @@ -12,8 +12,8 @@ struct vm_area_struct; * GFP bitmasks.. */ /* Zone modifiers in GFP_ZONEMASK (see linux/mmzone.h - low two bits) */ -#define __GFP_DMA 0x01u -#define __GFP_HIGHMEM 0x02u +#define __GFP_DMA ((__force gfp_t)0x01u) +#define __GFP_HIGHMEM ((__force gfp_t)0x02u) /* * Action modifiers - doesn't change the zoning @@ -26,24 +26,24 @@ struct vm_area_struct; * * __GFP_NORETRY: The VM implementation must not retry indefinitely. */ -#define __GFP_WAIT 0x10u /* Can wait and reschedule? */ -#define __GFP_HIGH 0x20u /* Should access emergency pools? */ -#define __GFP_IO 0x40u /* Can start physical IO? */ -#define __GFP_FS 0x80u /* Can call down to low-level FS? */ -#define __GFP_COLD 0x100u /* Cache-cold page required */ -#define __GFP_NOWARN 0x200u /* Suppress page allocation failure warning */ -#define __GFP_REPEAT 0x400u /* Retry the allocation. Might fail */ -#define __GFP_NOFAIL 0x800u /* Retry for ever. Cannot fail */ -#define __GFP_NORETRY 0x1000u /* Do not retry. Might fail */ -#define __GFP_NO_GROW 0x2000u /* Slab internal usage */ -#define __GFP_COMP 0x4000u /* Add compound page metadata */ -#define __GFP_ZERO 0x8000u /* Return zeroed page on success */ -#define __GFP_NOMEMALLOC 0x10000u /* Don't use emergency reserves */ -#define __GFP_NORECLAIM 0x20000u /* No realy zone reclaim during allocation */ -#define __GFP_HARDWALL 0x40000u /* Enforce hardwall cpuset memory allocs */ +#define __GFP_WAIT ((__force gfp_t)0x10u) /* Can wait and reschedule? */ +#define __GFP_HIGH ((__force gfp_t)0x20u) /* Should access emergency pools? */ +#define __GFP_IO ((__force gfp_t)0x40u) /* Can start physical IO? */ +#define __GFP_FS ((__force gfp_t)0x80u) /* Can call down to low-level FS? */ +#define __GFP_COLD ((__force gfp_t)0x100u) /* Cache-cold page required */ +#define __GFP_NOWARN ((__force gfp_t)0x200u) /* Suppress page allocation failure warning */ +#define __GFP_REPEAT ((__force gfp_t)0x400u) /* Retry the allocation. Might fail */ +#define __GFP_NOFAIL ((__force gfp_t)0x800u) /* Retry for ever. Cannot fail */ +#define __GFP_NORETRY ((__force gfp_t)0x1000u)/* Do not retry. Might fail */ +#define __GFP_NO_GROW ((__force gfp_t)0x2000u)/* Slab internal usage */ +#define __GFP_COMP ((__force gfp_t)0x4000u)/* Add compound page metadata */ +#define __GFP_ZERO ((__force gfp_t)0x8000u)/* Return zeroed page on success */ +#define __GFP_NOMEMALLOC ((__force gfp_t)0x10000u) /* Don't use emergency reserves */ +#define __GFP_NORECLAIM ((__force gfp_t)0x20000u) /* No realy zone reclaim during allocation */ +#define __GFP_HARDWALL ((__force gfp_t)0x40000u) /* Enforce hardwall cpuset memory allocs */ #define __GFP_BITS_SHIFT 20 /* Room for 20 __GFP_FOO bits */ -#define __GFP_BITS_MASK ((1 << __GFP_BITS_SHIFT) - 1) +#define __GFP_BITS_MASK ((__force gfp_t)((1 << __GFP_BITS_SHIFT) - 1)) /* if you forget to add the bitmask here kernel will crash, period */ #define GFP_LEVEL_MASK (__GFP_WAIT|__GFP_HIGH|__GFP_IO|__GFP_FS| \ @@ -64,6 +64,7 @@ struct vm_area_struct; #define GFP_DMA __GFP_DMA +#define gfp_zone(mask) ((__force int)((mask) & (__force gfp_t)GFP_ZONEMASK)) /* * There is only one page-allocator function, and two main namespaces to @@ -94,7 +95,7 @@ static inline struct page *alloc_pages_node(int nid, gfp_t gfp_mask, return NULL; return __alloc_pages(gfp_mask, order, - NODE_DATA(nid)->node_zonelists + (gfp_mask & GFP_ZONEMASK)); + NODE_DATA(nid)->node_zonelists + gfp_zone(gfp_mask)); } #ifdef CONFIG_NUMA diff --git a/include/linux/types.h b/include/linux/types.h index 0aee34f9da9f..21b9ce803644 100644 --- a/include/linux/types.h +++ b/include/linux/types.h @@ -151,7 +151,12 @@ typedef unsigned long sector_t; */ #ifdef __CHECKER__ -#define __bitwise __attribute__((bitwise)) +#define __bitwise__ __attribute__((bitwise)) +#else +#define __bitwise__ +#endif +#ifdef __CHECK_ENDIAN__ +#define __bitwise __bitwise__ #else #define __bitwise #endif @@ -166,7 +171,7 @@ typedef __u64 __bitwise __be64; #endif #ifdef __KERNEL__ -typedef unsigned __nocast gfp_t; +typedef unsigned __bitwise__ gfp_t; #endif struct ustat { diff --git a/mm/mempolicy.c b/mm/mempolicy.c index 37af443eb094..1d5c64df1653 100644 --- a/mm/mempolicy.c +++ b/mm/mempolicy.c @@ -700,7 +700,7 @@ static struct zonelist *zonelist_policy(gfp_t gfp, struct mempolicy *policy) case MPOL_BIND: /* Lower zones don't get a policy applied */ /* Careful: current->mems_allowed might have moved */ - if ((gfp & GFP_ZONEMASK) >= policy_zone) + if (gfp_zone(gfp) >= policy_zone) if (cpuset_zonelist_valid_mems_allowed(policy->v.zonelist)) return policy->v.zonelist; /*FALL THROUGH*/ @@ -712,7 +712,7 @@ static struct zonelist *zonelist_policy(gfp_t gfp, struct mempolicy *policy) nd = 0; BUG(); } - return NODE_DATA(nd)->node_zonelists + (gfp & GFP_ZONEMASK); + return NODE_DATA(nd)->node_zonelists + gfp_zone(gfp); } /* Do dynamic interleaving for a process */ @@ -757,7 +757,7 @@ static struct page *alloc_page_interleave(gfp_t gfp, unsigned order, unsigned ni struct page *page; BUG_ON(!node_online(nid)); - zl = NODE_DATA(nid)->node_zonelists + (gfp & GFP_ZONEMASK); + zl = NODE_DATA(nid)->node_zonelists + gfp_zone(gfp); page = __alloc_pages(gfp, order, zl); if (page && page_zone(page) == zl->zones[0]) { zone_pcp(zl->zones[0],get_cpu())->interleave_hit++; diff --git a/mm/page_alloc.c b/mm/page_alloc.c index e1d3d77f4aee..aa43ae3ab8c9 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -1089,7 +1089,7 @@ static unsigned int nr_free_zone_pages(int offset) */ unsigned int nr_free_buffer_pages(void) { - return nr_free_zone_pages(GFP_USER & GFP_ZONEMASK); + return nr_free_zone_pages(gfp_zone(GFP_USER)); } /* @@ -1097,7 +1097,7 @@ unsigned int nr_free_buffer_pages(void) */ unsigned int nr_free_pagecache_pages(void) { - return nr_free_zone_pages(GFP_HIGHUSER & GFP_ZONEMASK); + return nr_free_zone_pages(gfp_zone(GFP_HIGHUSER)); } #ifdef CONFIG_HIGHMEM -- cgit v1.2.3 From 6daa0e28627abf362138244a620a821a9027d816 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Fri, 21 Oct 2005 03:18:50 -0400 Subject: [PATCH] gfp_t: mm/* (easy parts) Signed-off-by: Al Viro Signed-off-by: Linus Torvalds --- include/linux/mm.h | 2 +- include/linux/pagemap.h | 6 +++--- include/linux/slab.h | 2 +- include/linux/swap.h | 4 ++-- mm/filemap.c | 8 ++++---- mm/mempool.c | 2 +- mm/shmem.c | 4 ++-- mm/slab.c | 8 ++++---- mm/vmscan.c | 8 ++++---- 9 files changed, 22 insertions(+), 22 deletions(-) (limited to 'include/linux') diff --git a/include/linux/mm.h b/include/linux/mm.h index 097b3a3c693d..e1649578fb0c 100644 --- a/include/linux/mm.h +++ b/include/linux/mm.h @@ -747,7 +747,7 @@ extern unsigned long do_mremap(unsigned long addr, * The callback will be passed nr_to_scan == 0 when the VM is querying the * cache size, so a fastpath for that case is appropriate. */ -typedef int (*shrinker_t)(int nr_to_scan, unsigned int gfp_mask); +typedef int (*shrinker_t)(int nr_to_scan, gfp_t gfp_mask); /* * Add an aging callback. The int is the number of 'seeks' it takes diff --git a/include/linux/pagemap.h b/include/linux/pagemap.h index acbf31c154f8..efbae53fb078 100644 --- a/include/linux/pagemap.h +++ b/include/linux/pagemap.h @@ -69,7 +69,7 @@ extern struct page * find_lock_page(struct address_space *mapping, extern struct page * find_trylock_page(struct address_space *mapping, unsigned long index); extern struct page * find_or_create_page(struct address_space *mapping, - unsigned long index, unsigned int gfp_mask); + unsigned long index, gfp_t gfp_mask); unsigned find_get_pages(struct address_space *mapping, pgoff_t start, unsigned int nr_pages, struct page **pages); unsigned find_get_pages_tag(struct address_space *mapping, pgoff_t *index, @@ -92,9 +92,9 @@ extern int read_cache_pages(struct address_space *mapping, struct list_head *pages, filler_t *filler, void *data); int add_to_page_cache(struct page *page, struct address_space *mapping, - unsigned long index, int gfp_mask); + unsigned long index, gfp_t gfp_mask); int add_to_page_cache_lru(struct page *page, struct address_space *mapping, - unsigned long index, int gfp_mask); + unsigned long index, gfp_t gfp_mask); extern void remove_from_page_cache(struct page *page); extern void __remove_from_page_cache(struct page *page); diff --git a/include/linux/slab.h b/include/linux/slab.h index 5fc04a16ecb0..09b9aa60063d 100644 --- a/include/linux/slab.h +++ b/include/linux/slab.h @@ -121,7 +121,7 @@ extern unsigned int ksize(const void *); extern void *kmem_cache_alloc_node(kmem_cache_t *, gfp_t flags, int node); extern void *kmalloc_node(size_t size, gfp_t flags, int node); #else -static inline void *kmem_cache_alloc_node(kmem_cache_t *cachep, int flags, int node) +static inline void *kmem_cache_alloc_node(kmem_cache_t *cachep, gfp_t flags, int node) { return kmem_cache_alloc(cachep, flags); } diff --git a/include/linux/swap.h b/include/linux/swap.h index a7bf1a3b1496..20c975642cab 100644 --- a/include/linux/swap.h +++ b/include/linux/swap.h @@ -171,8 +171,8 @@ extern int rotate_reclaimable_page(struct page *page); extern void swap_setup(void); /* linux/mm/vmscan.c */ -extern int try_to_free_pages(struct zone **, unsigned int); -extern int zone_reclaim(struct zone *, unsigned int, unsigned int); +extern int try_to_free_pages(struct zone **, gfp_t); +extern int zone_reclaim(struct zone *, gfp_t, unsigned int); extern int shrink_all_memory(int); extern int vm_swappiness; diff --git a/mm/filemap.c b/mm/filemap.c index b5346576e58d..1c31b2fd2ca5 100644 --- a/mm/filemap.c +++ b/mm/filemap.c @@ -377,7 +377,7 @@ int filemap_write_and_wait_range(struct address_space *mapping, * This function does not add the page to the LRU. The caller must do that. */ int add_to_page_cache(struct page *page, struct address_space *mapping, - pgoff_t offset, int gfp_mask) + pgoff_t offset, gfp_t gfp_mask) { int error = radix_tree_preload(gfp_mask & ~__GFP_HIGHMEM); @@ -401,7 +401,7 @@ int add_to_page_cache(struct page *page, struct address_space *mapping, EXPORT_SYMBOL(add_to_page_cache); int add_to_page_cache_lru(struct page *page, struct address_space *mapping, - pgoff_t offset, int gfp_mask) + pgoff_t offset, gfp_t gfp_mask) { int ret = add_to_page_cache(page, mapping, offset, gfp_mask); if (ret == 0) @@ -591,7 +591,7 @@ EXPORT_SYMBOL(find_lock_page); * memory exhaustion. */ struct page *find_or_create_page(struct address_space *mapping, - unsigned long index, unsigned int gfp_mask) + unsigned long index, gfp_t gfp_mask) { struct page *page, *cached_page = NULL; int err; @@ -683,7 +683,7 @@ struct page * grab_cache_page_nowait(struct address_space *mapping, unsigned long index) { struct page *page = find_get_page(mapping, index); - unsigned int gfp_mask; + gfp_t gfp_mask; if (page) { if (!TestSetPageLocked(page)) diff --git a/mm/mempool.c b/mm/mempool.c index 9e377ea700b2..1a99b80480d3 100644 --- a/mm/mempool.c +++ b/mm/mempool.c @@ -205,7 +205,7 @@ void * mempool_alloc(mempool_t *pool, gfp_t gfp_mask) void *element; unsigned long flags; wait_queue_t wait; - unsigned int gfp_temp; + gfp_t gfp_temp; might_sleep_if(gfp_mask & __GFP_WAIT); diff --git a/mm/shmem.c b/mm/shmem.c index ea064d89cda9..55e04a0734c1 100644 --- a/mm/shmem.c +++ b/mm/shmem.c @@ -85,7 +85,7 @@ enum sgp_type { static int shmem_getpage(struct inode *inode, unsigned long idx, struct page **pagep, enum sgp_type sgp, int *type); -static inline struct page *shmem_dir_alloc(unsigned int gfp_mask) +static inline struct page *shmem_dir_alloc(gfp_t gfp_mask) { /* * The above definition of ENTRIES_PER_PAGE, and the use of @@ -898,7 +898,7 @@ struct page *shmem_swapin(struct shmem_inode_info *info, swp_entry_t entry, } static struct page * -shmem_alloc_page(unsigned long gfp, struct shmem_inode_info *info, +shmem_alloc_page(gfp_t gfp, struct shmem_inode_info *info, unsigned long idx) { struct vm_area_struct pvma; diff --git a/mm/slab.c b/mm/slab.c index d05c678bceb3..d30423f167a2 100644 --- a/mm/slab.c +++ b/mm/slab.c @@ -386,7 +386,7 @@ struct kmem_cache_s { unsigned int gfporder; /* force GFP flags, e.g. GFP_DMA */ - unsigned int gfpflags; + gfp_t gfpflags; size_t colour; /* cache colouring range */ unsigned int colour_off; /* colour offset */ @@ -2117,7 +2117,7 @@ static void cache_init_objs(kmem_cache_t *cachep, slabp->free = 0; } -static void kmem_flagcheck(kmem_cache_t *cachep, unsigned int flags) +static void kmem_flagcheck(kmem_cache_t *cachep, gfp_t flags) { if (flags & SLAB_DMA) { if (!(cachep->gfpflags & GFP_DMA)) @@ -2152,7 +2152,7 @@ static int cache_grow(kmem_cache_t *cachep, gfp_t flags, int nodeid) struct slab *slabp; void *objp; size_t offset; - unsigned int local_flags; + gfp_t local_flags; unsigned long ctor_flags; struct kmem_list3 *l3; @@ -2546,7 +2546,7 @@ static inline void *__cache_alloc(kmem_cache_t *cachep, gfp_t flags) /* * A interface to enable slab creation on nodeid */ -static void *__cache_alloc_node(kmem_cache_t *cachep, int flags, int nodeid) +static void *__cache_alloc_node(kmem_cache_t *cachep, gfp_t flags, int nodeid) { struct list_head *entry; struct slab *slabp; diff --git a/mm/vmscan.c b/mm/vmscan.c index 64f9570cff56..843c87d1e61f 100644 --- a/mm/vmscan.c +++ b/mm/vmscan.c @@ -70,7 +70,7 @@ struct scan_control { unsigned int priority; /* This context's GFP mask */ - unsigned int gfp_mask; + gfp_t gfp_mask; int may_writepage; @@ -186,7 +186,7 @@ EXPORT_SYMBOL(remove_shrinker); * * Returns the number of slab objects which we shrunk. */ -static int shrink_slab(unsigned long scanned, unsigned int gfp_mask, +static int shrink_slab(unsigned long scanned, gfp_t gfp_mask, unsigned long lru_pages) { struct shrinker *shrinker; @@ -926,7 +926,7 @@ shrink_caches(struct zone **zones, struct scan_control *sc) * holds filesystem locks which prevent writeout this might not work, and the * allocation attempt will fail. */ -int try_to_free_pages(struct zone **zones, unsigned int gfp_mask) +int try_to_free_pages(struct zone **zones, gfp_t gfp_mask) { int priority; int ret = 0; @@ -1338,7 +1338,7 @@ module_init(kswapd_init) /* * Try to free up some pages from this zone through reclaim. */ -int zone_reclaim(struct zone *zone, unsigned int gfp_mask, unsigned int order) +int zone_reclaim(struct zone *zone, gfp_t gfp_mask, unsigned int order) { struct scan_control sc; int nr_pages = 1 << order; -- cgit v1.2.3 From fd4f2df24bc23e6b8fc069765b425c7dacf52347 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Fri, 21 Oct 2005 03:18:50 -0400 Subject: [PATCH] gfp_t: lib/* Signed-off-by: Al Viro Signed-off-by: Linus Torvalds --- include/linux/idr.h | 2 +- include/linux/kobject.h | 2 +- include/linux/radix-tree.h | 2 +- include/linux/textsearch.h | 4 ++-- lib/idr.c | 2 +- lib/kobject.c | 2 +- lib/kobject_uevent.c | 4 ++-- lib/textsearch.c | 2 +- 8 files changed, 10 insertions(+), 10 deletions(-) (limited to 'include/linux') diff --git a/include/linux/idr.h b/include/linux/idr.h index 3d5de45f961b..7fb3ff9c7b0e 100644 --- a/include/linux/idr.h +++ b/include/linux/idr.h @@ -71,7 +71,7 @@ struct idr { */ void *idr_find(struct idr *idp, int id); -int idr_pre_get(struct idr *idp, unsigned gfp_mask); +int idr_pre_get(struct idr *idp, gfp_t gfp_mask); int idr_get_new(struct idr *idp, void *ptr, int *id); int idr_get_new_above(struct idr *idp, void *ptr, int starting_id, int *id); void idr_remove(struct idr *idp, int id); diff --git a/include/linux/kobject.h b/include/linux/kobject.h index 3b22304f12fd..7f7403aa4a41 100644 --- a/include/linux/kobject.h +++ b/include/linux/kobject.h @@ -65,7 +65,7 @@ extern void kobject_unregister(struct kobject *); extern struct kobject * kobject_get(struct kobject *); extern void kobject_put(struct kobject *); -extern char * kobject_get_path(struct kobject *, int); +extern char * kobject_get_path(struct kobject *, gfp_t); struct kobj_type { void (*release)(struct kobject *); diff --git a/include/linux/radix-tree.h b/include/linux/radix-tree.h index 045d4761febc..9f0f9281f42a 100644 --- a/include/linux/radix-tree.h +++ b/include/linux/radix-tree.h @@ -24,7 +24,7 @@ struct radix_tree_root { unsigned int height; - unsigned int gfp_mask; + gfp_t gfp_mask; struct radix_tree_node *rnode; }; diff --git a/include/linux/textsearch.h b/include/linux/textsearch.h index 515046d1b2f4..fc5bb4e91a58 100644 --- a/include/linux/textsearch.h +++ b/include/linux/textsearch.h @@ -40,7 +40,7 @@ struct ts_state struct ts_ops { const char *name; - struct ts_config * (*init)(const void *, unsigned int, int); + struct ts_config * (*init)(const void *, unsigned int, gfp_t); unsigned int (*find)(struct ts_config *, struct ts_state *); void (*destroy)(struct ts_config *); @@ -148,7 +148,7 @@ static inline unsigned int textsearch_get_pattern_len(struct ts_config *conf) extern int textsearch_register(struct ts_ops *); extern int textsearch_unregister(struct ts_ops *); extern struct ts_config *textsearch_prepare(const char *, const void *, - unsigned int, int, int); + unsigned int, gfp_t, int); extern void textsearch_destroy(struct ts_config *conf); extern unsigned int textsearch_find_continuous(struct ts_config *, struct ts_state *, diff --git a/lib/idr.c b/lib/idr.c index d4df21debc4d..6414b2fb482d 100644 --- a/lib/idr.c +++ b/lib/idr.c @@ -72,7 +72,7 @@ static void free_layer(struct idr *idp, struct idr_layer *p) * If the system is REALLY out of memory this function returns 0, * otherwise 1. */ -int idr_pre_get(struct idr *idp, unsigned gfp_mask) +int idr_pre_get(struct idr *idp, gfp_t gfp_mask) { while (idp->id_free_cnt < IDR_FREE_MAX) { struct idr_layer *new; diff --git a/lib/kobject.c b/lib/kobject.c index dd0917dd9fa9..253d3004ace9 100644 --- a/lib/kobject.c +++ b/lib/kobject.c @@ -100,7 +100,7 @@ static void fill_kobj_path(struct kobject *kobj, char *path, int length) * @kobj: kobject in question, with which to build the path * @gfp_mask: the allocation type used to allocate the path */ -char *kobject_get_path(struct kobject *kobj, int gfp_mask) +char *kobject_get_path(struct kobject *kobj, gfp_t gfp_mask) { char *path; int len; diff --git a/lib/kobject_uevent.c b/lib/kobject_uevent.c index 04ca4429ddfa..7ef6f6a17aa6 100644 --- a/lib/kobject_uevent.c +++ b/lib/kobject_uevent.c @@ -62,7 +62,7 @@ static struct sock *uevent_sock; * @gfp_mask: */ static int send_uevent(const char *signal, const char *obj, - char **envp, int gfp_mask) + char **envp, gfp_t gfp_mask) { struct sk_buff *skb; char *pos; @@ -98,7 +98,7 @@ static int send_uevent(const char *signal, const char *obj, } static int do_kobject_uevent(struct kobject *kobj, enum kobject_action action, - struct attribute *attr, int gfp_mask) + struct attribute *attr, gfp_t gfp_mask) { char *path; char *attrpath; diff --git a/lib/textsearch.c b/lib/textsearch.c index 1e934c196f0f..6f3093efbd7b 100644 --- a/lib/textsearch.c +++ b/lib/textsearch.c @@ -254,7 +254,7 @@ unsigned int textsearch_find_continuous(struct ts_config *conf, * parameters or a ERR_PTR(). */ struct ts_config *textsearch_prepare(const char *algo, const void *pattern, - unsigned int len, int gfp_mask, int flags) + unsigned int len, gfp_t gfp_mask, int flags) { int err = -ENOENT; struct ts_config *conf; -- cgit v1.2.3 From 7d877f3bda870ab5f001bd92528654471d5966b3 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Fri, 21 Oct 2005 03:20:43 -0400 Subject: [PATCH] gfp_t: net/* Signed-off-by: Al Viro Signed-off-by: Linus Torvalds --- include/linux/security.h | 2 +- include/net/sock.h | 2 +- net/core/sock.c | 2 +- net/dccp/output.c | 2 +- net/netlink/af_netlink.c | 2 +- security/dummy.c | 2 +- security/selinux/hooks.c | 4 ++-- 7 files changed, 8 insertions(+), 8 deletions(-) (limited to 'include/linux') diff --git a/include/linux/security.h b/include/linux/security.h index 627382e74057..dac956ed98f0 100644 --- a/include/linux/security.h +++ b/include/linux/security.h @@ -1210,7 +1210,7 @@ struct security_operations { int (*socket_shutdown) (struct socket * sock, int how); int (*socket_sock_rcv_skb) (struct sock * sk, struct sk_buff * skb); int (*socket_getpeersec) (struct socket *sock, char __user *optval, int __user *optlen, unsigned len); - int (*sk_alloc_security) (struct sock *sk, int family, int priority); + int (*sk_alloc_security) (struct sock *sk, int family, gfp_t priority); void (*sk_free_security) (struct sock *sk); #endif /* CONFIG_SECURITY_NETWORK */ }; diff --git a/include/net/sock.h b/include/net/sock.h index ecb75526cba0..e0498bd36004 100644 --- a/include/net/sock.h +++ b/include/net/sock.h @@ -207,7 +207,7 @@ struct sock { struct sk_buff_head sk_write_queue; int sk_wmem_queued; int sk_forward_alloc; - unsigned int sk_allocation; + gfp_t sk_allocation; int sk_sndbuf; int sk_route_caps; unsigned long sk_flags; diff --git a/net/core/sock.c b/net/core/sock.c index 1c52fe809eda..9602ceb3bac9 100644 --- a/net/core/sock.c +++ b/net/core/sock.c @@ -940,7 +940,7 @@ static struct sk_buff *sock_alloc_send_pskb(struct sock *sk, int noblock, int *errcode) { struct sk_buff *skb; - unsigned int gfp_mask; + gfp_t gfp_mask; long timeo; int err; diff --git a/net/dccp/output.c b/net/dccp/output.c index 29250749f16f..d59f86f7ceab 100644 --- a/net/dccp/output.c +++ b/net/dccp/output.c @@ -495,7 +495,7 @@ void dccp_send_close(struct sock *sk, const int active) { struct dccp_sock *dp = dccp_sk(sk); struct sk_buff *skb; - const unsigned int prio = active ? GFP_KERNEL : GFP_ATOMIC; + const gfp_t prio = active ? GFP_KERNEL : GFP_ATOMIC; skb = alloc_skb(sk->sk_prot->max_header, prio); if (skb == NULL) diff --git a/net/netlink/af_netlink.c b/net/netlink/af_netlink.c index 678c3f2c0d0b..291df2e4c492 100644 --- a/net/netlink/af_netlink.c +++ b/net/netlink/af_netlink.c @@ -827,7 +827,7 @@ struct netlink_broadcast_data { int failure; int congested; int delivered; - unsigned int allocation; + gfp_t allocation; struct sk_buff *skb, *skb2; }; diff --git a/security/dummy.c b/security/dummy.c index 9623a61dfc76..3d34f3de7e82 100644 --- a/security/dummy.c +++ b/security/dummy.c @@ -768,7 +768,7 @@ static int dummy_socket_getpeersec(struct socket *sock, char __user *optval, return -ENOPROTOOPT; } -static inline int dummy_sk_alloc_security (struct sock *sk, int family, int priority) +static inline int dummy_sk_alloc_security (struct sock *sk, int family, gfp_t priority) { return 0; } diff --git a/security/selinux/hooks.c b/security/selinux/hooks.c index b13be15165f5..447a1e0f48cb 100644 --- a/security/selinux/hooks.c +++ b/security/selinux/hooks.c @@ -262,7 +262,7 @@ static void superblock_free_security(struct super_block *sb) } #ifdef CONFIG_SECURITY_NETWORK -static int sk_alloc_security(struct sock *sk, int family, int priority) +static int sk_alloc_security(struct sock *sk, int family, gfp_t priority) { struct sk_security_struct *ssec; @@ -3380,7 +3380,7 @@ out: return err; } -static int selinux_sk_alloc_security(struct sock *sk, int family, int priority) +static int selinux_sk_alloc_security(struct sock *sk, int family, gfp_t priority) { return sk_alloc_security(sk, family, priority); } -- cgit v1.2.3 From 27496a8c67bef4d789d8e3c8317ca35813a507ae Mon Sep 17 00:00:00 2001 From: Al Viro Date: Fri, 21 Oct 2005 03:20:48 -0400 Subject: [PATCH] gfp_t: fs/* - ->releasepage() annotated (s/int/gfp_t), instances updated - missing gfp_t in fs/* added - fixed misannotation from the original sweep caught by bitwise checks: XFS used __nocast both for gfp_t and for flags used by XFS allocator. The latter left with unsigned int __nocast; we might want to add a different type for those but for now let's leave them alone. That, BTW, is a case when __nocast use had been actively confusing - it had been used in the same code for two different and similar types, with no way to catch misuses. Switch of gfp_t to bitwise had caught that immediately... One tricky bit is left alone to be dealt with later - mapping->flags is a mix of gfp_t and error indications. Left alone for now. Signed-off-by: Al Viro Signed-off-by: Linus Torvalds --- fs/afs/file.c | 4 ++-- fs/bio.c | 4 ++-- fs/buffer.c | 2 +- fs/dcache.c | 2 +- fs/dquot.c | 2 +- fs/ext3/inode.c | 2 +- fs/hfs/inode.c | 2 +- fs/hfsplus/inode.c | 2 +- fs/inode.c | 2 +- fs/jbd/journal.c | 2 +- fs/jbd/transaction.c | 2 +- fs/jfs/jfs_metapage.c | 4 ++-- fs/mbcache.c | 6 +++--- fs/reiserfs/fix_node.c | 2 +- fs/reiserfs/inode.c | 2 +- fs/xfs/linux-2.6/kmem.c | 22 +++++++++++----------- fs/xfs/linux-2.6/kmem.h | 18 +++++++++--------- fs/xfs/linux-2.6/xfs_aops.c | 2 +- fs/xfs/linux-2.6/xfs_buf.c | 8 ++++---- include/linux/bio.h | 2 +- include/linux/buffer_head.h | 2 +- include/linux/fs.h | 2 +- include/linux/jbd.h | 4 ++-- include/linux/mbcache.h | 2 +- include/linux/reiserfs_fs.h | 2 +- 25 files changed, 52 insertions(+), 52 deletions(-) (limited to 'include/linux') diff --git a/fs/afs/file.c b/fs/afs/file.c index 23c125128024..0d576987ec67 100644 --- a/fs/afs/file.c +++ b/fs/afs/file.c @@ -29,7 +29,7 @@ static int afs_file_release(struct inode *inode, struct file *file); static int afs_file_readpage(struct file *file, struct page *page); static int afs_file_invalidatepage(struct page *page, unsigned long offset); -static int afs_file_releasepage(struct page *page, int gfp_flags); +static int afs_file_releasepage(struct page *page, gfp_t gfp_flags); static ssize_t afs_file_write(struct file *file, const char __user *buf, size_t size, loff_t *off); @@ -279,7 +279,7 @@ static int afs_file_invalidatepage(struct page *page, unsigned long offset) /* * release a page and cleanup its private data */ -static int afs_file_releasepage(struct page *page, int gfp_flags) +static int afs_file_releasepage(struct page *page, gfp_t gfp_flags) { struct cachefs_page *pageio; diff --git a/fs/bio.c b/fs/bio.c index 7d81a93afd48..460554b07ff9 100644 --- a/fs/bio.c +++ b/fs/bio.c @@ -778,7 +778,7 @@ static int bio_map_kern_endio(struct bio *bio, unsigned int bytes_done, int err) static struct bio *__bio_map_kern(request_queue_t *q, void *data, - unsigned int len, unsigned int gfp_mask) + unsigned int len, gfp_t gfp_mask) { unsigned long kaddr = (unsigned long)data; unsigned long end = (kaddr + len + PAGE_SIZE - 1) >> PAGE_SHIFT; @@ -825,7 +825,7 @@ static struct bio *__bio_map_kern(request_queue_t *q, void *data, * device. Returns an error pointer in case of error. */ struct bio *bio_map_kern(request_queue_t *q, void *data, unsigned int len, - unsigned int gfp_mask) + gfp_t gfp_mask) { struct bio *bio; diff --git a/fs/buffer.c b/fs/buffer.c index 9657696fd6d7..b1667986442f 100644 --- a/fs/buffer.c +++ b/fs/buffer.c @@ -1571,7 +1571,7 @@ static inline void discard_buffer(struct buffer_head * bh) * * NOTE: @gfp_mask may go away, and this function may become non-blocking. */ -int try_to_release_page(struct page *page, int gfp_mask) +int try_to_release_page(struct page *page, gfp_t gfp_mask) { struct address_space * const mapping = page->mapping; diff --git a/fs/dcache.c b/fs/dcache.c index fb10386c59be..e90512ed35a4 100644 --- a/fs/dcache.c +++ b/fs/dcache.c @@ -689,7 +689,7 @@ void shrink_dcache_anon(struct hlist_head *head) * * In this case we return -1 to tell the caller that we baled. */ -static int shrink_dcache_memory(int nr, unsigned int gfp_mask) +static int shrink_dcache_memory(int nr, gfp_t gfp_mask) { if (nr) { if (!(gfp_mask & __GFP_FS)) diff --git a/fs/dquot.c b/fs/dquot.c index b9732335bcdc..05f3327d64a3 100644 --- a/fs/dquot.c +++ b/fs/dquot.c @@ -500,7 +500,7 @@ static void prune_dqcache(int count) * more memory */ -static int shrink_dqcache_memory(int nr, unsigned int gfp_mask) +static int shrink_dqcache_memory(int nr, gfp_t gfp_mask) { if (nr) { spin_lock(&dq_list_lock); diff --git a/fs/ext3/inode.c b/fs/ext3/inode.c index b5177c90d6f1..8b38f2232796 100644 --- a/fs/ext3/inode.c +++ b/fs/ext3/inode.c @@ -1434,7 +1434,7 @@ static int ext3_invalidatepage(struct page *page, unsigned long offset) return journal_invalidatepage(journal, page, offset); } -static int ext3_releasepage(struct page *page, int wait) +static int ext3_releasepage(struct page *page, gfp_t wait) { journal_t *journal = EXT3_JOURNAL(page->mapping->host); diff --git a/fs/hfs/inode.c b/fs/hfs/inode.c index f1570b9f9de3..3f680c5675bf 100644 --- a/fs/hfs/inode.c +++ b/fs/hfs/inode.c @@ -46,7 +46,7 @@ static sector_t hfs_bmap(struct address_space *mapping, sector_t block) return generic_block_bmap(mapping, block, hfs_get_block); } -static int hfs_releasepage(struct page *page, int mask) +static int hfs_releasepage(struct page *page, gfp_t mask) { struct inode *inode = page->mapping->host; struct super_block *sb = inode->i_sb; diff --git a/fs/hfsplus/inode.c b/fs/hfsplus/inode.c index d5642705f633..f205773ddfbe 100644 --- a/fs/hfsplus/inode.c +++ b/fs/hfsplus/inode.c @@ -40,7 +40,7 @@ static sector_t hfsplus_bmap(struct address_space *mapping, sector_t block) return generic_block_bmap(mapping, block, hfsplus_get_block); } -static int hfsplus_releasepage(struct page *page, int mask) +static int hfsplus_releasepage(struct page *page, gfp_t mask) { struct inode *inode = page->mapping->host; struct super_block *sb = inode->i_sb; diff --git a/fs/inode.c b/fs/inode.c index f80a79ff156b..7d3316527767 100644 --- a/fs/inode.c +++ b/fs/inode.c @@ -475,7 +475,7 @@ static void prune_icache(int nr_to_scan) * This function is passed the number of inodes to scan, and it returns the * total number of remaining possibly-reclaimable inodes. */ -static int shrink_icache_memory(int nr, unsigned int gfp_mask) +static int shrink_icache_memory(int nr, gfp_t gfp_mask) { if (nr) { /* diff --git a/fs/jbd/journal.c b/fs/jbd/journal.c index 7ae2c4fe506b..e4b516ac4989 100644 --- a/fs/jbd/journal.c +++ b/fs/jbd/journal.c @@ -1606,7 +1606,7 @@ int journal_blocks_per_page(struct inode *inode) * Simple support for retrying memory allocations. Introduced to help to * debug different VM deadlock avoidance strategies. */ -void * __jbd_kmalloc (const char *where, size_t size, int flags, int retry) +void * __jbd_kmalloc (const char *where, size_t size, gfp_t flags, int retry) { return kmalloc(size, flags | (retry ? __GFP_NOFAIL : 0)); } diff --git a/fs/jbd/transaction.c b/fs/jbd/transaction.c index 49bbc2be3d72..13cb05bf6048 100644 --- a/fs/jbd/transaction.c +++ b/fs/jbd/transaction.c @@ -1621,7 +1621,7 @@ out: * while the data is part of a transaction. Yes? */ int journal_try_to_free_buffers(journal_t *journal, - struct page *page, int unused_gfp_mask) + struct page *page, gfp_t unused_gfp_mask) { struct buffer_head *head; struct buffer_head *bh; diff --git a/fs/jfs/jfs_metapage.c b/fs/jfs/jfs_metapage.c index 13d7e3f1feb4..eeb37d70e650 100644 --- a/fs/jfs/jfs_metapage.c +++ b/fs/jfs/jfs_metapage.c @@ -198,7 +198,7 @@ static void init_once(void *foo, kmem_cache_t *cachep, unsigned long flags) } } -static inline struct metapage *alloc_metapage(unsigned int gfp_mask) +static inline struct metapage *alloc_metapage(gfp_t gfp_mask) { return mempool_alloc(metapage_mempool, gfp_mask); } @@ -534,7 +534,7 @@ add_failed: return -EIO; } -static int metapage_releasepage(struct page *page, int gfp_mask) +static int metapage_releasepage(struct page *page, gfp_t gfp_mask) { struct metapage *mp; int busy = 0; diff --git a/fs/mbcache.c b/fs/mbcache.c index b002a088857d..298997f17475 100644 --- a/fs/mbcache.c +++ b/fs/mbcache.c @@ -116,7 +116,7 @@ mb_cache_indexes(struct mb_cache *cache) * What the mbcache registers as to get shrunk dynamically. */ -static int mb_cache_shrink_fn(int nr_to_scan, unsigned int gfp_mask); +static int mb_cache_shrink_fn(int nr_to_scan, gfp_t gfp_mask); static inline int @@ -140,7 +140,7 @@ __mb_cache_entry_unhash(struct mb_cache_entry *ce) static inline void -__mb_cache_entry_forget(struct mb_cache_entry *ce, int gfp_mask) +__mb_cache_entry_forget(struct mb_cache_entry *ce, gfp_t gfp_mask) { struct mb_cache *cache = ce->e_cache; @@ -193,7 +193,7 @@ forget: * Returns the number of objects which are present in the cache. */ static int -mb_cache_shrink_fn(int nr_to_scan, unsigned int gfp_mask) +mb_cache_shrink_fn(int nr_to_scan, gfp_t gfp_mask) { LIST_HEAD(free_list); struct list_head *l, *ltmp; diff --git a/fs/reiserfs/fix_node.c b/fs/reiserfs/fix_node.c index 2706e2adffab..45829889dcdc 100644 --- a/fs/reiserfs/fix_node.c +++ b/fs/reiserfs/fix_node.c @@ -2022,7 +2022,7 @@ static int get_neighbors(struct tree_balance *p_s_tb, int n_h) } #ifdef CONFIG_REISERFS_CHECK -void *reiserfs_kmalloc(size_t size, int flags, struct super_block *s) +void *reiserfs_kmalloc(size_t size, gfp_t flags, struct super_block *s) { void *vp; static size_t malloced; diff --git a/fs/reiserfs/inode.c b/fs/reiserfs/inode.c index d76ee6c4f9b8..5f82352b97e1 100644 --- a/fs/reiserfs/inode.c +++ b/fs/reiserfs/inode.c @@ -2842,7 +2842,7 @@ static int reiserfs_set_page_dirty(struct page *page) * even in -o notail mode, we can't be sure an old mount without -o notail * didn't create files with tails. */ -static int reiserfs_releasepage(struct page *page, int unused_gfp_flags) +static int reiserfs_releasepage(struct page *page, gfp_t unused_gfp_flags) { struct inode *inode = page->mapping->host; struct reiserfs_journal *j = SB_JOURNAL(inode->i_sb); diff --git a/fs/xfs/linux-2.6/kmem.c b/fs/xfs/linux-2.6/kmem.c index d2653b589b1c..3c92162dc728 100644 --- a/fs/xfs/linux-2.6/kmem.c +++ b/fs/xfs/linux-2.6/kmem.c @@ -45,11 +45,11 @@ void * -kmem_alloc(size_t size, gfp_t flags) +kmem_alloc(size_t size, unsigned int __nocast flags) { - int retries = 0; - unsigned int lflags = kmem_flags_convert(flags); - void *ptr; + int retries = 0; + gfp_t lflags = kmem_flags_convert(flags); + void *ptr; do { if (size < MAX_SLAB_SIZE || retries > MAX_VMALLOCS) @@ -67,7 +67,7 @@ kmem_alloc(size_t size, gfp_t flags) } void * -kmem_zalloc(size_t size, gfp_t flags) +kmem_zalloc(size_t size, unsigned int __nocast flags) { void *ptr; @@ -90,7 +90,7 @@ kmem_free(void *ptr, size_t size) void * kmem_realloc(void *ptr, size_t newsize, size_t oldsize, - gfp_t flags) + unsigned int __nocast flags) { void *new; @@ -105,11 +105,11 @@ kmem_realloc(void *ptr, size_t newsize, size_t oldsize, } void * -kmem_zone_alloc(kmem_zone_t *zone, gfp_t flags) +kmem_zone_alloc(kmem_zone_t *zone, unsigned int __nocast flags) { - int retries = 0; - unsigned int lflags = kmem_flags_convert(flags); - void *ptr; + int retries = 0; + gfp_t lflags = kmem_flags_convert(flags); + void *ptr; do { ptr = kmem_cache_alloc(zone, lflags); @@ -124,7 +124,7 @@ kmem_zone_alloc(kmem_zone_t *zone, gfp_t flags) } void * -kmem_zone_zalloc(kmem_zone_t *zone, gfp_t flags) +kmem_zone_zalloc(kmem_zone_t *zone, unsigned int __nocast flags) { void *ptr; diff --git a/fs/xfs/linux-2.6/kmem.h b/fs/xfs/linux-2.6/kmem.h index ee7010f085bc..f4bb78c268c0 100644 --- a/fs/xfs/linux-2.6/kmem.h +++ b/fs/xfs/linux-2.6/kmem.h @@ -81,9 +81,9 @@ typedef unsigned long xfs_pflags_t; *(NSTATEP) = *(OSTATEP); \ } while (0) -static __inline unsigned int kmem_flags_convert(gfp_t flags) +static __inline gfp_t kmem_flags_convert(unsigned int __nocast flags) { - unsigned int lflags = __GFP_NOWARN; /* we'll report problems, if need be */ + gfp_t lflags = __GFP_NOWARN; /* we'll report problems, if need be */ #ifdef DEBUG if (unlikely(flags & ~(KM_SLEEP|KM_NOSLEEP|KM_NOFS|KM_MAYFAIL))) { @@ -125,16 +125,16 @@ kmem_zone_destroy(kmem_zone_t *zone) BUG(); } -extern void *kmem_zone_zalloc(kmem_zone_t *, gfp_t); -extern void *kmem_zone_alloc(kmem_zone_t *, gfp_t); +extern void *kmem_zone_zalloc(kmem_zone_t *, unsigned int __nocast); +extern void *kmem_zone_alloc(kmem_zone_t *, unsigned int __nocast); -extern void *kmem_alloc(size_t, gfp_t); -extern void *kmem_realloc(void *, size_t, size_t, gfp_t); -extern void *kmem_zalloc(size_t, gfp_t); +extern void *kmem_alloc(size_t, unsigned int __nocast); +extern void *kmem_realloc(void *, size_t, size_t, unsigned int __nocast); +extern void *kmem_zalloc(size_t, unsigned int __nocast); extern void kmem_free(void *, size_t); typedef struct shrinker *kmem_shaker_t; -typedef int (*kmem_shake_func_t)(int, unsigned int); +typedef int (*kmem_shake_func_t)(int, gfp_t); static __inline kmem_shaker_t kmem_shake_register(kmem_shake_func_t sfunc) @@ -149,7 +149,7 @@ kmem_shake_deregister(kmem_shaker_t shrinker) } static __inline int -kmem_shake_allow(unsigned int gfp_mask) +kmem_shake_allow(gfp_t gfp_mask) { return (gfp_mask & __GFP_WAIT); } diff --git a/fs/xfs/linux-2.6/xfs_aops.c b/fs/xfs/linux-2.6/xfs_aops.c index c6c077978fe3..7aa398724706 100644 --- a/fs/xfs/linux-2.6/xfs_aops.c +++ b/fs/xfs/linux-2.6/xfs_aops.c @@ -1296,7 +1296,7 @@ linvfs_invalidate_page( STATIC int linvfs_release_page( struct page *page, - int gfp_mask) + gfp_t gfp_mask) { struct inode *inode = page->mapping->host; int dirty, delalloc, unmapped, unwritten; diff --git a/fs/xfs/linux-2.6/xfs_buf.c b/fs/xfs/linux-2.6/xfs_buf.c index e82cf72ac599..ba4767c04adf 100644 --- a/fs/xfs/linux-2.6/xfs_buf.c +++ b/fs/xfs/linux-2.6/xfs_buf.c @@ -64,7 +64,7 @@ STATIC kmem_cache_t *pagebuf_zone; STATIC kmem_shaker_t pagebuf_shake; -STATIC int xfsbufd_wakeup(int, unsigned int); +STATIC int xfsbufd_wakeup(int, gfp_t); STATIC void pagebuf_delwri_queue(xfs_buf_t *, int); STATIC struct workqueue_struct *xfslogd_workqueue; @@ -383,7 +383,7 @@ _pagebuf_lookup_pages( size_t blocksize = bp->pb_target->pbr_bsize; size_t size = bp->pb_count_desired; size_t nbytes, offset; - int gfp_mask = pb_to_gfp(flags); + gfp_t gfp_mask = pb_to_gfp(flags); unsigned short page_count, i; pgoff_t first; loff_t end; @@ -1749,8 +1749,8 @@ STATIC int xfsbufd_force_sleep; STATIC int xfsbufd_wakeup( - int priority, - unsigned int mask) + int priority, + gfp_t mask) { if (xfsbufd_force_sleep) return 0; diff --git a/include/linux/bio.h b/include/linux/bio.h index 3344b4e8e43a..685fd3720df5 100644 --- a/include/linux/bio.h +++ b/include/linux/bio.h @@ -301,7 +301,7 @@ extern struct bio *bio_map_user_iov(struct request_queue *, struct sg_iovec *, int, int); extern void bio_unmap_user(struct bio *); extern struct bio *bio_map_kern(struct request_queue *, void *, unsigned int, - unsigned int); + gfp_t); extern void bio_set_pages_dirty(struct bio *bio); extern void bio_check_pages_dirty(struct bio *bio); extern struct bio *bio_copy_user(struct request_queue *, unsigned long, unsigned int, int); diff --git a/include/linux/buffer_head.h b/include/linux/buffer_head.h index 6a1d154c0825..88af42f5e04a 100644 --- a/include/linux/buffer_head.h +++ b/include/linux/buffer_head.h @@ -188,7 +188,7 @@ extern int buffer_heads_over_limit; * Generic address_space_operations implementations for buffer_head-backed * address_spaces. */ -int try_to_release_page(struct page * page, int gfp_mask); +int try_to_release_page(struct page * page, gfp_t gfp_mask); int block_invalidatepage(struct page *page, unsigned long offset); int block_write_full_page(struct page *page, get_block_t *get_block, struct writeback_control *wbc); diff --git a/include/linux/fs.h b/include/linux/fs.h index e0b77c5af9a0..f83d997c5582 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -320,7 +320,7 @@ struct address_space_operations { /* Unfortunately this kludge is needed for FIBMAP. Don't use it */ sector_t (*bmap)(struct address_space *, sector_t); int (*invalidatepage) (struct page *, unsigned long); - int (*releasepage) (struct page *, int); + int (*releasepage) (struct page *, gfp_t); ssize_t (*direct_IO)(int, struct kiocb *, const struct iovec *iov, loff_t offset, unsigned long nr_segs); struct page* (*get_xip_page)(struct address_space *, sector_t, diff --git a/include/linux/jbd.h b/include/linux/jbd.h index ff853b3173c6..be197eb90077 100644 --- a/include/linux/jbd.h +++ b/include/linux/jbd.h @@ -69,7 +69,7 @@ extern int journal_enable_debug; #define jbd_debug(f, a...) /**/ #endif -extern void * __jbd_kmalloc (const char *where, size_t size, int flags, int retry); +extern void * __jbd_kmalloc (const char *where, size_t size, gfp_t flags, int retry); #define jbd_kmalloc(size, flags) \ __jbd_kmalloc(__FUNCTION__, (size), (flags), journal_oom_retry) #define jbd_rep_kmalloc(size, flags) \ @@ -890,7 +890,7 @@ extern int journal_forget (handle_t *, struct buffer_head *); extern void journal_sync_buffer (struct buffer_head *); extern int journal_invalidatepage(journal_t *, struct page *, unsigned long); -extern int journal_try_to_free_buffers(journal_t *, struct page *, int); +extern int journal_try_to_free_buffers(journal_t *, struct page *, gfp_t); extern int journal_stop(handle_t *); extern int journal_flush (journal_t *); extern void journal_lock_updates (journal_t *); diff --git a/include/linux/mbcache.h b/include/linux/mbcache.h index 9263d2db2d67..99e044b4efc6 100644 --- a/include/linux/mbcache.h +++ b/include/linux/mbcache.h @@ -22,7 +22,7 @@ struct mb_cache_entry { }; struct mb_cache_op { - int (*free)(struct mb_cache_entry *, int); + int (*free)(struct mb_cache_entry *, gfp_t); }; /* Functions on caches */ diff --git a/include/linux/reiserfs_fs.h b/include/linux/reiserfs_fs.h index af00b10294cd..001ab82df051 100644 --- a/include/linux/reiserfs_fs.h +++ b/include/linux/reiserfs_fs.h @@ -1972,7 +1972,7 @@ extern struct address_space_operations reiserfs_address_space_operations; /* fix_nodes.c */ #ifdef CONFIG_REISERFS_CHECK -void *reiserfs_kmalloc(size_t size, int flags, struct super_block *s); +void *reiserfs_kmalloc(size_t size, gfp_t flags, struct super_block *s); void reiserfs_kfree(const void *vp, size_t size, struct super_block *s); #else static inline void *reiserfs_kmalloc(size_t size, int flags, -- cgit v1.2.3 From 8267e268e0914ac9371d07f711fcf20cc572993c Mon Sep 17 00:00:00 2001 From: Al Viro Date: Fri, 21 Oct 2005 03:20:53 -0400 Subject: [PATCH] gfp_t: block layer core Signed-off-by: Al Viro Signed-off-by: Linus Torvalds --- drivers/block/as-iosched.c | 2 +- drivers/block/cfq-iosched.c | 8 ++++---- drivers/block/deadline-iosched.c | 2 +- drivers/block/elevator.c | 2 +- drivers/block/ll_rw_blk.c | 16 ++++++++-------- include/linux/blkdev.h | 14 +++++++------- include/linux/elevator.h | 4 ++-- 7 files changed, 24 insertions(+), 24 deletions(-) (limited to 'include/linux') diff --git a/drivers/block/as-iosched.c b/drivers/block/as-iosched.c index 95c0a3690b0f..1f08e14697e9 100644 --- a/drivers/block/as-iosched.c +++ b/drivers/block/as-iosched.c @@ -1807,7 +1807,7 @@ static void as_put_request(request_queue_t *q, struct request *rq) } static int as_set_request(request_queue_t *q, struct request *rq, - struct bio *bio, int gfp_mask) + struct bio *bio, gfp_t gfp_mask) { struct as_data *ad = q->elevator->elevator_data; struct as_rq *arq = mempool_alloc(ad->arq_pool, gfp_mask); diff --git a/drivers/block/cfq-iosched.c b/drivers/block/cfq-iosched.c index cd056e7e64ec..d3bfe8cfb039 100644 --- a/drivers/block/cfq-iosched.c +++ b/drivers/block/cfq-iosched.c @@ -1422,7 +1422,7 @@ static void cfq_exit_io_context(struct cfq_io_context *cic) } static struct cfq_io_context * -cfq_alloc_io_context(struct cfq_data *cfqd, int gfp_mask) +cfq_alloc_io_context(struct cfq_data *cfqd, gfp_t gfp_mask) { struct cfq_io_context *cic = kmem_cache_alloc(cfq_ioc_pool, gfp_mask); @@ -1517,7 +1517,7 @@ static int cfq_ioc_set_ioprio(struct io_context *ioc, unsigned int ioprio) static struct cfq_queue * cfq_get_queue(struct cfq_data *cfqd, unsigned int key, unsigned short ioprio, - int gfp_mask) + gfp_t gfp_mask) { const int hashval = hash_long(key, CFQ_QHASH_SHIFT); struct cfq_queue *cfqq, *new_cfqq = NULL; @@ -1578,7 +1578,7 @@ out: * cfqq, so we don't need to worry about it disappearing */ static struct cfq_io_context * -cfq_get_io_context(struct cfq_data *cfqd, pid_t pid, int gfp_mask) +cfq_get_io_context(struct cfq_data *cfqd, pid_t pid, gfp_t gfp_mask) { struct io_context *ioc = NULL; struct cfq_io_context *cic; @@ -2075,7 +2075,7 @@ static void cfq_put_request(request_queue_t *q, struct request *rq) */ static int cfq_set_request(request_queue_t *q, struct request *rq, struct bio *bio, - int gfp_mask) + gfp_t gfp_mask) { struct cfq_data *cfqd = q->elevator->elevator_data; struct task_struct *tsk = current; diff --git a/drivers/block/deadline-iosched.c b/drivers/block/deadline-iosched.c index 52a3ae5289a0..753546ba2262 100644 --- a/drivers/block/deadline-iosched.c +++ b/drivers/block/deadline-iosched.c @@ -756,7 +756,7 @@ static void deadline_put_request(request_queue_t *q, struct request *rq) static int deadline_set_request(request_queue_t *q, struct request *rq, struct bio *bio, - int gfp_mask) + gfp_t gfp_mask) { struct deadline_data *dd = q->elevator->elevator_data; struct deadline_rq *drq; diff --git a/drivers/block/elevator.c b/drivers/block/elevator.c index 98f0126a2deb..c744d2a13062 100644 --- a/drivers/block/elevator.c +++ b/drivers/block/elevator.c @@ -487,7 +487,7 @@ struct request *elv_former_request(request_queue_t *q, struct request *rq) } int elv_set_request(request_queue_t *q, struct request *rq, struct bio *bio, - int gfp_mask) + gfp_t gfp_mask) { elevator_t *e = q->elevator; diff --git a/drivers/block/ll_rw_blk.c b/drivers/block/ll_rw_blk.c index baedac522945..0f64ee7d8d27 100644 --- a/drivers/block/ll_rw_blk.c +++ b/drivers/block/ll_rw_blk.c @@ -1652,13 +1652,13 @@ static int blk_init_free_list(request_queue_t *q) static int __make_request(request_queue_t *, struct bio *); -request_queue_t *blk_alloc_queue(int gfp_mask) +request_queue_t *blk_alloc_queue(gfp_t gfp_mask) { return blk_alloc_queue_node(gfp_mask, -1); } EXPORT_SYMBOL(blk_alloc_queue); -request_queue_t *blk_alloc_queue_node(int gfp_mask, int node_id) +request_queue_t *blk_alloc_queue_node(gfp_t gfp_mask, int node_id) { request_queue_t *q; @@ -1787,7 +1787,7 @@ static inline void blk_free_request(request_queue_t *q, struct request *rq) } static inline struct request * -blk_alloc_request(request_queue_t *q, int rw, struct bio *bio, int gfp_mask) +blk_alloc_request(request_queue_t *q, int rw, struct bio *bio, gfp_t gfp_mask) { struct request *rq = mempool_alloc(q->rq.rq_pool, gfp_mask); @@ -1885,7 +1885,7 @@ static void freed_request(request_queue_t *q, int rw) * Returns !NULL on success, with queue_lock *not held*. */ static struct request *get_request(request_queue_t *q, int rw, struct bio *bio, - int gfp_mask) + gfp_t gfp_mask) { struct request *rq = NULL; struct request_list *rl = &q->rq; @@ -2019,7 +2019,7 @@ static struct request *get_request_wait(request_queue_t *q, int rw, return rq; } -struct request *blk_get_request(request_queue_t *q, int rw, int gfp_mask) +struct request *blk_get_request(request_queue_t *q, int rw, gfp_t gfp_mask) { struct request *rq; @@ -2251,7 +2251,7 @@ EXPORT_SYMBOL(blk_rq_unmap_user); * @gfp_mask: memory allocation flags */ int blk_rq_map_kern(request_queue_t *q, struct request *rq, void *kbuf, - unsigned int len, unsigned int gfp_mask) + unsigned int len, gfp_t gfp_mask) { struct bio *bio; @@ -3393,7 +3393,7 @@ void exit_io_context(void) * but since the current task itself holds a reference, the context can be * used in general code, so long as it stays within `current` context. */ -struct io_context *current_io_context(int gfp_flags) +struct io_context *current_io_context(gfp_t gfp_flags) { struct task_struct *tsk = current; struct io_context *ret; @@ -3424,7 +3424,7 @@ EXPORT_SYMBOL(current_io_context); * * This is always called in the context of the task which submitted the I/O. */ -struct io_context *get_io_context(int gfp_flags) +struct io_context *get_io_context(gfp_t gfp_flags) { struct io_context *ret; ret = current_io_context(gfp_flags); diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h index efdc9b5bc05c..1afbdb2d752c 100644 --- a/include/linux/blkdev.h +++ b/include/linux/blkdev.h @@ -96,8 +96,8 @@ struct io_context { void put_io_context(struct io_context *ioc); void exit_io_context(void); -struct io_context *current_io_context(int gfp_flags); -struct io_context *get_io_context(int gfp_flags); +struct io_context *current_io_context(gfp_t gfp_flags); +struct io_context *get_io_context(gfp_t gfp_flags); void copy_io_context(struct io_context **pdst, struct io_context **psrc); void swap_io_context(struct io_context **ioc1, struct io_context **ioc2); @@ -354,7 +354,7 @@ struct request_queue * queue needs bounce pages for pages above this limit */ unsigned long bounce_pfn; - unsigned int bounce_gfp; + gfp_t bounce_gfp; /* * various queue flags, see QUEUE_* below @@ -550,7 +550,7 @@ extern void generic_make_request(struct bio *bio); extern void blk_put_request(struct request *); extern void blk_end_sync_rq(struct request *rq); extern void blk_attempt_remerge(request_queue_t *, struct request *); -extern struct request *blk_get_request(request_queue_t *, int, int); +extern struct request *blk_get_request(request_queue_t *, int, gfp_t); extern void blk_insert_request(request_queue_t *, struct request *, int, void *); extern void blk_requeue_request(request_queue_t *, struct request *); extern void blk_plug_device(request_queue_t *); @@ -565,7 +565,7 @@ extern void blk_run_queue(request_queue_t *); extern void blk_queue_activity_fn(request_queue_t *, activity_fn *, void *); extern int blk_rq_map_user(request_queue_t *, struct request *, void __user *, unsigned int); extern int blk_rq_unmap_user(struct bio *, unsigned int); -extern int blk_rq_map_kern(request_queue_t *, struct request *, void *, unsigned int, unsigned int); +extern int blk_rq_map_kern(request_queue_t *, struct request *, void *, unsigned int, gfp_t); extern int blk_rq_map_user_iov(request_queue_t *, struct request *, struct sg_iovec *, int); extern int blk_execute_rq(request_queue_t *, struct gendisk *, struct request *, int); @@ -654,8 +654,8 @@ extern void blk_wait_queue_drained(request_queue_t *, int); extern void blk_finish_queue_drain(request_queue_t *); int blk_get_queue(request_queue_t *); -request_queue_t *blk_alloc_queue(int gfp_mask); -request_queue_t *blk_alloc_queue_node(int,int); +request_queue_t *blk_alloc_queue(gfp_t); +request_queue_t *blk_alloc_queue_node(gfp_t, int); #define blk_put_queue(q) blk_cleanup_queue((q)) /* diff --git a/include/linux/elevator.h b/include/linux/elevator.h index ea6bbc2d7407..ed93125c1db5 100644 --- a/include/linux/elevator.h +++ b/include/linux/elevator.h @@ -18,7 +18,7 @@ typedef struct request *(elevator_request_list_fn) (request_queue_t *, struct re typedef void (elevator_completed_req_fn) (request_queue_t *, struct request *); typedef int (elevator_may_queue_fn) (request_queue_t *, int, struct bio *); -typedef int (elevator_set_req_fn) (request_queue_t *, struct request *, struct bio *, int); +typedef int (elevator_set_req_fn) (request_queue_t *, struct request *, struct bio *, gfp_t); typedef void (elevator_put_req_fn) (request_queue_t *, struct request *); typedef void (elevator_deactivate_req_fn) (request_queue_t *, struct request *); @@ -98,7 +98,7 @@ extern int elv_register_queue(request_queue_t *q); extern void elv_unregister_queue(request_queue_t *q); extern int elv_may_queue(request_queue_t *, int, struct bio *); extern void elv_completed_request(request_queue_t *, struct request *); -extern int elv_set_request(request_queue_t *, struct request *, struct bio *, int); +extern int elv_set_request(request_queue_t *, struct request *, struct bio *, gfp_t); extern void elv_put_request(request_queue_t *, struct request *); /* -- cgit v1.2.3 From 55016f10e31bb15b85d8c500f979dfdceb37d548 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Fri, 21 Oct 2005 03:21:58 -0400 Subject: [PATCH] gfp_t: drivers/usb Signed-off-by: Al Viro Signed-off-by: Linus Torvalds --- drivers/usb/core/buffer.c | 2 +- drivers/usb/core/hcd.c | 2 +- drivers/usb/core/hcd.h | 8 ++++---- drivers/usb/core/message.c | 2 +- drivers/usb/core/urb.c | 4 ++-- drivers/usb/core/usb.c | 2 +- drivers/usb/gadget/dummy_hcd.c | 8 ++++---- drivers/usb/gadget/ether.c | 22 +++++++++++----------- drivers/usb/gadget/goku_udc.c | 6 +++--- drivers/usb/gadget/lh7a40x_udc.c | 12 ++++++------ drivers/usb/gadget/net2280.c | 6 +++--- drivers/usb/gadget/omap_udc.c | 6 +++--- drivers/usb/gadget/pxa2xx_udc.c | 6 +++--- drivers/usb/gadget/serial.c | 16 ++++++++-------- drivers/usb/gadget/zero.c | 8 ++++---- drivers/usb/host/ehci-hcd.c | 2 +- drivers/usb/host/ehci-mem.c | 6 +++--- drivers/usb/host/ehci-q.c | 6 +++--- drivers/usb/host/ehci-sched.c | 14 +++++++------- drivers/usb/host/isp116x-hcd.c | 2 +- drivers/usb/host/ohci-hcd.c | 2 +- drivers/usb/host/ohci-mem.c | 4 ++-- drivers/usb/host/sl811-hcd.c | 2 +- drivers/usb/host/uhci-q.c | 2 +- drivers/usb/misc/uss720.c | 6 +++--- drivers/usb/net/asix.c | 2 +- drivers/usb/net/gl620a.c | 2 +- drivers/usb/net/kaweth.c | 6 +++--- drivers/usb/net/net1080.c | 2 +- drivers/usb/net/rndis_host.c | 2 +- drivers/usb/net/usbnet.c | 2 +- drivers/usb/net/usbnet.h | 2 +- drivers/usb/net/zaurus.c | 2 +- drivers/usb/net/zd1201.c | 2 +- include/linux/usb.h | 8 ++++---- include/linux/usb_gadget.h | 12 ++++++------ sound/usb/usbmidi.c | 2 +- 37 files changed, 100 insertions(+), 100 deletions(-) (limited to 'include/linux') diff --git a/drivers/usb/core/buffer.c b/drivers/usb/core/buffer.c index fc15b4acc8af..57e800ac3cee 100644 --- a/drivers/usb/core/buffer.c +++ b/drivers/usb/core/buffer.c @@ -106,7 +106,7 @@ void hcd_buffer_destroy (struct usb_hcd *hcd) void *hcd_buffer_alloc ( struct usb_bus *bus, size_t size, - unsigned mem_flags, + gfp_t mem_flags, dma_addr_t *dma ) { diff --git a/drivers/usb/core/hcd.c b/drivers/usb/core/hcd.c index 1017a97a418b..ff19d64041b5 100644 --- a/drivers/usb/core/hcd.c +++ b/drivers/usb/core/hcd.c @@ -1112,7 +1112,7 @@ static void urb_unlink (struct urb *urb) * expects usb_submit_urb() to have sanity checked and conditioned all * inputs in the urb */ -static int hcd_submit_urb (struct urb *urb, unsigned mem_flags) +static int hcd_submit_urb (struct urb *urb, gfp_t mem_flags) { int status; struct usb_hcd *hcd = urb->dev->bus->hcpriv; diff --git a/drivers/usb/core/hcd.h b/drivers/usb/core/hcd.h index ac451fa7e4d2..1f1ed6211af8 100644 --- a/drivers/usb/core/hcd.h +++ b/drivers/usb/core/hcd.h @@ -142,12 +142,12 @@ struct hcd_timeout { /* timeouts we allocate */ struct usb_operations { int (*get_frame_number) (struct usb_device *usb_dev); - int (*submit_urb) (struct urb *urb, unsigned mem_flags); + int (*submit_urb) (struct urb *urb, gfp_t mem_flags); int (*unlink_urb) (struct urb *urb, int status); /* allocate dma-consistent buffer for URB_DMA_NOMAPPING */ void *(*buffer_alloc)(struct usb_bus *bus, size_t size, - unsigned mem_flags, + gfp_t mem_flags, dma_addr_t *dma); void (*buffer_free)(struct usb_bus *bus, size_t size, void *addr, dma_addr_t dma); @@ -200,7 +200,7 @@ struct hc_driver { int (*urb_enqueue) (struct usb_hcd *hcd, struct usb_host_endpoint *ep, struct urb *urb, - unsigned mem_flags); + gfp_t mem_flags); int (*urb_dequeue) (struct usb_hcd *hcd, struct urb *urb); /* hw synch, freeing endpoint resources that urb_dequeue can't */ @@ -247,7 +247,7 @@ int hcd_buffer_create (struct usb_hcd *hcd); void hcd_buffer_destroy (struct usb_hcd *hcd); void *hcd_buffer_alloc (struct usb_bus *bus, size_t size, - unsigned mem_flags, dma_addr_t *dma); + gfp_t mem_flags, dma_addr_t *dma); void hcd_buffer_free (struct usb_bus *bus, size_t size, void *addr, dma_addr_t dma); diff --git a/drivers/usb/core/message.c b/drivers/usb/core/message.c index f1fb67fe22a8..f9a81e84dbdf 100644 --- a/drivers/usb/core/message.c +++ b/drivers/usb/core/message.c @@ -321,7 +321,7 @@ int usb_sg_init ( struct scatterlist *sg, int nents, size_t length, - unsigned mem_flags + gfp_t mem_flags ) { int i; diff --git a/drivers/usb/core/urb.c b/drivers/usb/core/urb.c index c846fefb7386..b32898e0a27d 100644 --- a/drivers/usb/core/urb.c +++ b/drivers/usb/core/urb.c @@ -60,7 +60,7 @@ void usb_init_urb(struct urb *urb) * * The driver must call usb_free_urb() when it is finished with the urb. */ -struct urb *usb_alloc_urb(int iso_packets, unsigned mem_flags) +struct urb *usb_alloc_urb(int iso_packets, gfp_t mem_flags) { struct urb *urb; @@ -224,7 +224,7 @@ struct urb * usb_get_urb(struct urb *urb) * GFP_NOIO, unless b) or c) apply * */ -int usb_submit_urb(struct urb *urb, unsigned mem_flags) +int usb_submit_urb(struct urb *urb, gfp_t mem_flags) { int pipe, temp, max; struct usb_device *dev; diff --git a/drivers/usb/core/usb.c b/drivers/usb/core/usb.c index 7d131509e419..4c57f3f649ed 100644 --- a/drivers/usb/core/usb.c +++ b/drivers/usb/core/usb.c @@ -1147,7 +1147,7 @@ int __usb_get_extra_descriptor(char *buffer, unsigned size, void *usb_buffer_alloc ( struct usb_device *dev, size_t size, - unsigned mem_flags, + gfp_t mem_flags, dma_addr_t *dma ) { diff --git a/drivers/usb/gadget/dummy_hcd.c b/drivers/usb/gadget/dummy_hcd.c index 583db7c38cf1..8d9d8ee89554 100644 --- a/drivers/usb/gadget/dummy_hcd.c +++ b/drivers/usb/gadget/dummy_hcd.c @@ -470,7 +470,7 @@ static int dummy_disable (struct usb_ep *_ep) } static struct usb_request * -dummy_alloc_request (struct usb_ep *_ep, unsigned mem_flags) +dummy_alloc_request (struct usb_ep *_ep, gfp_t mem_flags) { struct dummy_ep *ep; struct dummy_request *req; @@ -507,7 +507,7 @@ dummy_alloc_buffer ( struct usb_ep *_ep, unsigned bytes, dma_addr_t *dma, - unsigned mem_flags + gfp_t mem_flags ) { char *retval; struct dummy_ep *ep; @@ -541,7 +541,7 @@ fifo_complete (struct usb_ep *ep, struct usb_request *req) static int dummy_queue (struct usb_ep *_ep, struct usb_request *_req, - unsigned mem_flags) + gfp_t mem_flags) { struct dummy_ep *ep; struct dummy_request *req; @@ -999,7 +999,7 @@ static int dummy_urb_enqueue ( struct usb_hcd *hcd, struct usb_host_endpoint *ep, struct urb *urb, - unsigned mem_flags + gfp_t mem_flags ) { struct dummy *dum; struct urbp *urbp; diff --git a/drivers/usb/gadget/ether.c b/drivers/usb/gadget/ether.c index 49459e33e952..f1024e804d5c 100644 --- a/drivers/usb/gadget/ether.c +++ b/drivers/usb/gadget/ether.c @@ -945,11 +945,11 @@ config_buf (enum usb_device_speed speed, /*-------------------------------------------------------------------------*/ -static void eth_start (struct eth_dev *dev, unsigned gfp_flags); -static int alloc_requests (struct eth_dev *dev, unsigned n, unsigned gfp_flags); +static void eth_start (struct eth_dev *dev, gfp_t gfp_flags); +static int alloc_requests (struct eth_dev *dev, unsigned n, gfp_t gfp_flags); static int -set_ether_config (struct eth_dev *dev, unsigned gfp_flags) +set_ether_config (struct eth_dev *dev, gfp_t gfp_flags) { int result = 0; struct usb_gadget *gadget = dev->gadget; @@ -1081,7 +1081,7 @@ static void eth_reset_config (struct eth_dev *dev) * that returns config descriptors, and altsetting code. */ static int -eth_set_config (struct eth_dev *dev, unsigned number, unsigned gfp_flags) +eth_set_config (struct eth_dev *dev, unsigned number, gfp_t gfp_flags) { int result = 0; struct usb_gadget *gadget = dev->gadget; @@ -1598,7 +1598,7 @@ static void defer_kevent (struct eth_dev *dev, int flag) static void rx_complete (struct usb_ep *ep, struct usb_request *req); static int -rx_submit (struct eth_dev *dev, struct usb_request *req, unsigned gfp_flags) +rx_submit (struct eth_dev *dev, struct usb_request *req, gfp_t gfp_flags) { struct sk_buff *skb; int retval = -ENOMEM; @@ -1724,7 +1724,7 @@ clean: } static int prealloc (struct list_head *list, struct usb_ep *ep, - unsigned n, unsigned gfp_flags) + unsigned n, gfp_t gfp_flags) { unsigned i; struct usb_request *req; @@ -1763,7 +1763,7 @@ extra: return 0; } -static int alloc_requests (struct eth_dev *dev, unsigned n, unsigned gfp_flags) +static int alloc_requests (struct eth_dev *dev, unsigned n, gfp_t gfp_flags) { int status; @@ -1779,7 +1779,7 @@ fail: return status; } -static void rx_fill (struct eth_dev *dev, unsigned gfp_flags) +static void rx_fill (struct eth_dev *dev, gfp_t gfp_flags) { struct usb_request *req; unsigned long flags; @@ -1962,7 +1962,7 @@ drop: * normally just one notification will be queued. */ -static struct usb_request *eth_req_alloc (struct usb_ep *, unsigned, unsigned); +static struct usb_request *eth_req_alloc (struct usb_ep *, unsigned, gfp_t); static void eth_req_free (struct usb_ep *ep, struct usb_request *req); static void @@ -2024,7 +2024,7 @@ static int rndis_control_ack (struct net_device *net) #endif /* RNDIS */ -static void eth_start (struct eth_dev *dev, unsigned gfp_flags) +static void eth_start (struct eth_dev *dev, gfp_t gfp_flags) { DEBUG (dev, "%s\n", __FUNCTION__); @@ -2092,7 +2092,7 @@ static int eth_stop (struct net_device *net) /*-------------------------------------------------------------------------*/ static struct usb_request * -eth_req_alloc (struct usb_ep *ep, unsigned size, unsigned gfp_flags) +eth_req_alloc (struct usb_ep *ep, unsigned size, gfp_t gfp_flags) { struct usb_request *req; diff --git a/drivers/usb/gadget/goku_udc.c b/drivers/usb/gadget/goku_udc.c index eaab26f4ed37..b0f3cd63e3b9 100644 --- a/drivers/usb/gadget/goku_udc.c +++ b/drivers/usb/gadget/goku_udc.c @@ -269,7 +269,7 @@ static int goku_ep_disable(struct usb_ep *_ep) /*-------------------------------------------------------------------------*/ static struct usb_request * -goku_alloc_request(struct usb_ep *_ep, unsigned gfp_flags) +goku_alloc_request(struct usb_ep *_ep, gfp_t gfp_flags) { struct goku_request *req; @@ -327,7 +327,7 @@ goku_free_request(struct usb_ep *_ep, struct usb_request *_req) */ static void * goku_alloc_buffer(struct usb_ep *_ep, unsigned bytes, - dma_addr_t *dma, unsigned gfp_flags) + dma_addr_t *dma, gfp_t gfp_flags) { void *retval; struct goku_ep *ep; @@ -789,7 +789,7 @@ finished: /*-------------------------------------------------------------------------*/ static int -goku_queue(struct usb_ep *_ep, struct usb_request *_req, unsigned gfp_flags) +goku_queue(struct usb_ep *_ep, struct usb_request *_req, gfp_t gfp_flags) { struct goku_request *req; struct goku_ep *ep; diff --git a/drivers/usb/gadget/lh7a40x_udc.c b/drivers/usb/gadget/lh7a40x_udc.c index 4842577789c9..012d1e5f1524 100644 --- a/drivers/usb/gadget/lh7a40x_udc.c +++ b/drivers/usb/gadget/lh7a40x_udc.c @@ -71,13 +71,13 @@ static char *state_names[] = { static int lh7a40x_ep_enable(struct usb_ep *ep, const struct usb_endpoint_descriptor *); static int lh7a40x_ep_disable(struct usb_ep *ep); -static struct usb_request *lh7a40x_alloc_request(struct usb_ep *ep, int); +static struct usb_request *lh7a40x_alloc_request(struct usb_ep *ep, gfp_t); static void lh7a40x_free_request(struct usb_ep *ep, struct usb_request *); static void *lh7a40x_alloc_buffer(struct usb_ep *ep, unsigned, dma_addr_t *, - int); + gfp_t); static void lh7a40x_free_buffer(struct usb_ep *ep, void *, dma_addr_t, unsigned); -static int lh7a40x_queue(struct usb_ep *ep, struct usb_request *, int); +static int lh7a40x_queue(struct usb_ep *ep, struct usb_request *, gfp_t); static int lh7a40x_dequeue(struct usb_ep *ep, struct usb_request *); static int lh7a40x_set_halt(struct usb_ep *ep, int); static int lh7a40x_fifo_status(struct usb_ep *ep); @@ -1106,7 +1106,7 @@ static int lh7a40x_ep_disable(struct usb_ep *_ep) } static struct usb_request *lh7a40x_alloc_request(struct usb_ep *ep, - unsigned gfp_flags) + gfp_t gfp_flags) { struct lh7a40x_request *req; @@ -1134,7 +1134,7 @@ static void lh7a40x_free_request(struct usb_ep *ep, struct usb_request *_req) } static void *lh7a40x_alloc_buffer(struct usb_ep *ep, unsigned bytes, - dma_addr_t * dma, unsigned gfp_flags) + dma_addr_t * dma, gfp_t gfp_flags) { char *retval; @@ -1158,7 +1158,7 @@ static void lh7a40x_free_buffer(struct usb_ep *ep, void *buf, dma_addr_t dma, * NOTE: Sets INDEX register */ static int lh7a40x_queue(struct usb_ep *_ep, struct usb_request *_req, - unsigned gfp_flags) + gfp_t gfp_flags) { struct lh7a40x_request *req; struct lh7a40x_ep *ep; diff --git a/drivers/usb/gadget/net2280.c b/drivers/usb/gadget/net2280.c index 477fab2e74d1..c32e1f7476da 100644 --- a/drivers/usb/gadget/net2280.c +++ b/drivers/usb/gadget/net2280.c @@ -376,7 +376,7 @@ static int net2280_disable (struct usb_ep *_ep) /*-------------------------------------------------------------------------*/ static struct usb_request * -net2280_alloc_request (struct usb_ep *_ep, unsigned gfp_flags) +net2280_alloc_request (struct usb_ep *_ep, gfp_t gfp_flags) { struct net2280_ep *ep; struct net2280_request *req; @@ -463,7 +463,7 @@ net2280_alloc_buffer ( struct usb_ep *_ep, unsigned bytes, dma_addr_t *dma, - unsigned gfp_flags + gfp_t gfp_flags ) { void *retval; @@ -897,7 +897,7 @@ done (struct net2280_ep *ep, struct net2280_request *req, int status) /*-------------------------------------------------------------------------*/ static int -net2280_queue (struct usb_ep *_ep, struct usb_request *_req, unsigned gfp_flags) +net2280_queue (struct usb_ep *_ep, struct usb_request *_req, gfp_t gfp_flags) { struct net2280_request *req; struct net2280_ep *ep; diff --git a/drivers/usb/gadget/omap_udc.c b/drivers/usb/gadget/omap_udc.c index ff5533e69560..287c5900fb13 100644 --- a/drivers/usb/gadget/omap_udc.c +++ b/drivers/usb/gadget/omap_udc.c @@ -269,7 +269,7 @@ static int omap_ep_disable(struct usb_ep *_ep) /*-------------------------------------------------------------------------*/ static struct usb_request * -omap_alloc_request(struct usb_ep *ep, unsigned gfp_flags) +omap_alloc_request(struct usb_ep *ep, gfp_t gfp_flags) { struct omap_req *req; @@ -298,7 +298,7 @@ omap_alloc_buffer( struct usb_ep *_ep, unsigned bytes, dma_addr_t *dma, - unsigned gfp_flags + gfp_t gfp_flags ) { void *retval; @@ -937,7 +937,7 @@ static void dma_channel_release(struct omap_ep *ep) /*-------------------------------------------------------------------------*/ static int -omap_ep_queue(struct usb_ep *_ep, struct usb_request *_req, unsigned gfp_flags) +omap_ep_queue(struct usb_ep *_ep, struct usb_request *_req, gfp_t gfp_flags) { struct omap_ep *ep = container_of(_ep, struct omap_ep, ep); struct omap_req *req = container_of(_req, struct omap_req, req); diff --git a/drivers/usb/gadget/pxa2xx_udc.c b/drivers/usb/gadget/pxa2xx_udc.c index 73f8c9404156..6e545393cfff 100644 --- a/drivers/usb/gadget/pxa2xx_udc.c +++ b/drivers/usb/gadget/pxa2xx_udc.c @@ -332,7 +332,7 @@ static int pxa2xx_ep_disable (struct usb_ep *_ep) * pxa2xx_ep_alloc_request - allocate a request data structure */ static struct usb_request * -pxa2xx_ep_alloc_request (struct usb_ep *_ep, unsigned gfp_flags) +pxa2xx_ep_alloc_request (struct usb_ep *_ep, gfp_t gfp_flags) { struct pxa2xx_request *req; @@ -367,7 +367,7 @@ pxa2xx_ep_free_request (struct usb_ep *_ep, struct usb_request *_req) */ static void * pxa2xx_ep_alloc_buffer(struct usb_ep *_ep, unsigned bytes, - dma_addr_t *dma, unsigned gfp_flags) + dma_addr_t *dma, gfp_t gfp_flags) { char *retval; @@ -874,7 +874,7 @@ done: /*-------------------------------------------------------------------------*/ static int -pxa2xx_ep_queue(struct usb_ep *_ep, struct usb_request *_req, unsigned gfp_flags) +pxa2xx_ep_queue(struct usb_ep *_ep, struct usb_request *_req, gfp_t gfp_flags) { struct pxa2xx_request *req; struct pxa2xx_ep *ep; diff --git a/drivers/usb/gadget/serial.c b/drivers/usb/gadget/serial.c index c925d9222f53..b35ac6d334f8 100644 --- a/drivers/usb/gadget/serial.c +++ b/drivers/usb/gadget/serial.c @@ -300,18 +300,18 @@ static int gs_build_config_buf(u8 *buf, enum usb_device_speed speed, u8 type, unsigned int index, int is_otg); static struct usb_request *gs_alloc_req(struct usb_ep *ep, unsigned int len, - unsigned kmalloc_flags); + gfp_t kmalloc_flags); static void gs_free_req(struct usb_ep *ep, struct usb_request *req); static struct gs_req_entry *gs_alloc_req_entry(struct usb_ep *ep, unsigned len, - unsigned kmalloc_flags); + gfp_t kmalloc_flags); static void gs_free_req_entry(struct usb_ep *ep, struct gs_req_entry *req); -static int gs_alloc_ports(struct gs_dev *dev, unsigned kmalloc_flags); +static int gs_alloc_ports(struct gs_dev *dev, gfp_t kmalloc_flags); static void gs_free_ports(struct gs_dev *dev); /* circular buffer */ -static struct gs_buf *gs_buf_alloc(unsigned int size, unsigned kmalloc_flags); +static struct gs_buf *gs_buf_alloc(unsigned int size, gfp_t kmalloc_flags); static void gs_buf_free(struct gs_buf *gb); static void gs_buf_clear(struct gs_buf *gb); static unsigned int gs_buf_data_avail(struct gs_buf *gb); @@ -2091,7 +2091,7 @@ static int gs_build_config_buf(u8 *buf, enum usb_device_speed speed, * usb_request or NULL if there is an error. */ static struct usb_request * -gs_alloc_req(struct usb_ep *ep, unsigned int len, unsigned kmalloc_flags) +gs_alloc_req(struct usb_ep *ep, unsigned int len, gfp_t kmalloc_flags) { struct usb_request *req; @@ -2132,7 +2132,7 @@ static void gs_free_req(struct usb_ep *ep, struct usb_request *req) * endpoint, buffer len, and kmalloc flags. */ static struct gs_req_entry * -gs_alloc_req_entry(struct usb_ep *ep, unsigned len, unsigned kmalloc_flags) +gs_alloc_req_entry(struct usb_ep *ep, unsigned len, gfp_t kmalloc_flags) { struct gs_req_entry *req; @@ -2173,7 +2173,7 @@ static void gs_free_req_entry(struct usb_ep *ep, struct gs_req_entry *req) * * The device lock is normally held when calling this function. */ -static int gs_alloc_ports(struct gs_dev *dev, unsigned kmalloc_flags) +static int gs_alloc_ports(struct gs_dev *dev, gfp_t kmalloc_flags) { int i; struct gs_port *port; @@ -2255,7 +2255,7 @@ static void gs_free_ports(struct gs_dev *dev) * * Allocate a circular buffer and all associated memory. */ -static struct gs_buf *gs_buf_alloc(unsigned int size, unsigned kmalloc_flags) +static struct gs_buf *gs_buf_alloc(unsigned int size, gfp_t kmalloc_flags) { struct gs_buf *gb; diff --git a/drivers/usb/gadget/zero.c b/drivers/usb/gadget/zero.c index 6890e773b2a2..ec9c424f1d97 100644 --- a/drivers/usb/gadget/zero.c +++ b/drivers/usb/gadget/zero.c @@ -612,7 +612,7 @@ static void source_sink_complete (struct usb_ep *ep, struct usb_request *req) } static struct usb_request * -source_sink_start_ep (struct usb_ep *ep, unsigned gfp_flags) +source_sink_start_ep (struct usb_ep *ep, gfp_t gfp_flags) { struct usb_request *req; int status; @@ -640,7 +640,7 @@ source_sink_start_ep (struct usb_ep *ep, unsigned gfp_flags) } static int -set_source_sink_config (struct zero_dev *dev, unsigned gfp_flags) +set_source_sink_config (struct zero_dev *dev, gfp_t gfp_flags) { int result = 0; struct usb_ep *ep; @@ -744,7 +744,7 @@ static void loopback_complete (struct usb_ep *ep, struct usb_request *req) } static int -set_loopback_config (struct zero_dev *dev, unsigned gfp_flags) +set_loopback_config (struct zero_dev *dev, gfp_t gfp_flags) { int result = 0; struct usb_ep *ep; @@ -845,7 +845,7 @@ static void zero_reset_config (struct zero_dev *dev) * by limiting configuration choices (like the pxa2xx). */ static int -zero_set_config (struct zero_dev *dev, unsigned number, unsigned gfp_flags) +zero_set_config (struct zero_dev *dev, unsigned number, gfp_t gfp_flags) { int result = 0; struct usb_gadget *gadget = dev->gadget; diff --git a/drivers/usb/host/ehci-hcd.c b/drivers/usb/host/ehci-hcd.c index b948ffd94f45..f5eb9e7b5b18 100644 --- a/drivers/usb/host/ehci-hcd.c +++ b/drivers/usb/host/ehci-hcd.c @@ -983,7 +983,7 @@ static int ehci_urb_enqueue ( struct usb_hcd *hcd, struct usb_host_endpoint *ep, struct urb *urb, - unsigned mem_flags + gfp_t mem_flags ) { struct ehci_hcd *ehci = hcd_to_ehci (hcd); struct list_head qtd_list; diff --git a/drivers/usb/host/ehci-mem.c b/drivers/usb/host/ehci-mem.c index 5c38ad869485..91c2ab43cbcc 100644 --- a/drivers/usb/host/ehci-mem.c +++ b/drivers/usb/host/ehci-mem.c @@ -45,7 +45,7 @@ static inline void ehci_qtd_init (struct ehci_qtd *qtd, dma_addr_t dma) INIT_LIST_HEAD (&qtd->qtd_list); } -static struct ehci_qtd *ehci_qtd_alloc (struct ehci_hcd *ehci, int flags) +static struct ehci_qtd *ehci_qtd_alloc (struct ehci_hcd *ehci, gfp_t flags) { struct ehci_qtd *qtd; dma_addr_t dma; @@ -79,7 +79,7 @@ static void qh_destroy (struct kref *kref) dma_pool_free (ehci->qh_pool, qh, qh->qh_dma); } -static struct ehci_qh *ehci_qh_alloc (struct ehci_hcd *ehci, int flags) +static struct ehci_qh *ehci_qh_alloc (struct ehci_hcd *ehci, gfp_t flags) { struct ehci_qh *qh; dma_addr_t dma; @@ -161,7 +161,7 @@ static void ehci_mem_cleanup (struct ehci_hcd *ehci) } /* remember to add cleanup code (above) if you add anything here */ -static int ehci_mem_init (struct ehci_hcd *ehci, int flags) +static int ehci_mem_init (struct ehci_hcd *ehci, gfp_t flags) { int i; diff --git a/drivers/usb/host/ehci-q.c b/drivers/usb/host/ehci-q.c index 940d38ca7d91..5bb872c3496d 100644 --- a/drivers/usb/host/ehci-q.c +++ b/drivers/usb/host/ehci-q.c @@ -477,7 +477,7 @@ qh_urb_transaction ( struct ehci_hcd *ehci, struct urb *urb, struct list_head *head, - int flags + gfp_t flags ) { struct ehci_qtd *qtd, *qtd_prev; dma_addr_t buf; @@ -629,7 +629,7 @@ static struct ehci_qh * qh_make ( struct ehci_hcd *ehci, struct urb *urb, - int flags + gfp_t flags ) { struct ehci_qh *qh = ehci_qh_alloc (ehci, flags); u32 info1 = 0, info2 = 0; @@ -906,7 +906,7 @@ submit_async ( struct usb_host_endpoint *ep, struct urb *urb, struct list_head *qtd_list, - unsigned mem_flags + gfp_t mem_flags ) { struct ehci_qtd *qtd; int epnum; diff --git a/drivers/usb/host/ehci-sched.c b/drivers/usb/host/ehci-sched.c index ccc7300baa6d..f0c8aa1ccd5d 100644 --- a/drivers/usb/host/ehci-sched.c +++ b/drivers/usb/host/ehci-sched.c @@ -589,7 +589,7 @@ static int intr_submit ( struct usb_host_endpoint *ep, struct urb *urb, struct list_head *qtd_list, - unsigned mem_flags + gfp_t mem_flags ) { unsigned epnum; unsigned long flags; @@ -634,7 +634,7 @@ done: /* ehci_iso_stream ops work with both ITD and SITD */ static struct ehci_iso_stream * -iso_stream_alloc (unsigned mem_flags) +iso_stream_alloc (gfp_t mem_flags) { struct ehci_iso_stream *stream; @@ -851,7 +851,7 @@ iso_stream_find (struct ehci_hcd *ehci, struct urb *urb) /* ehci_iso_sched ops can be ITD-only or SITD-only */ static struct ehci_iso_sched * -iso_sched_alloc (unsigned packets, unsigned mem_flags) +iso_sched_alloc (unsigned packets, gfp_t mem_flags) { struct ehci_iso_sched *iso_sched; int size = sizeof *iso_sched; @@ -924,7 +924,7 @@ itd_urb_transaction ( struct ehci_iso_stream *stream, struct ehci_hcd *ehci, struct urb *urb, - unsigned mem_flags + gfp_t mem_flags ) { struct ehci_itd *itd; @@ -1418,7 +1418,7 @@ itd_complete ( /*-------------------------------------------------------------------------*/ static int itd_submit (struct ehci_hcd *ehci, struct urb *urb, - unsigned mem_flags) + gfp_t mem_flags) { int status = -EINVAL; unsigned long flags; @@ -1529,7 +1529,7 @@ sitd_urb_transaction ( struct ehci_iso_stream *stream, struct ehci_hcd *ehci, struct urb *urb, - unsigned mem_flags + gfp_t mem_flags ) { struct ehci_sitd *sitd; @@ -1779,7 +1779,7 @@ sitd_complete ( static int sitd_submit (struct ehci_hcd *ehci, struct urb *urb, - unsigned mem_flags) + gfp_t mem_flags) { int status = -EINVAL; unsigned long flags; diff --git a/drivers/usb/host/isp116x-hcd.c b/drivers/usb/host/isp116x-hcd.c index e142056b0d2c..2548d94fcd72 100644 --- a/drivers/usb/host/isp116x-hcd.c +++ b/drivers/usb/host/isp116x-hcd.c @@ -694,7 +694,7 @@ static int balance(struct isp116x *isp116x, u16 period, u16 load) static int isp116x_urb_enqueue(struct usb_hcd *hcd, struct usb_host_endpoint *hep, struct urb *urb, - unsigned mem_flags) + gfp_t mem_flags) { struct isp116x *isp116x = hcd_to_isp116x(hcd); struct usb_device *udev = urb->dev; diff --git a/drivers/usb/host/ohci-hcd.c b/drivers/usb/host/ohci-hcd.c index 67c1aa5eb1c1..f8da8c7af7c6 100644 --- a/drivers/usb/host/ohci-hcd.c +++ b/drivers/usb/host/ohci-hcd.c @@ -180,7 +180,7 @@ static int ohci_urb_enqueue ( struct usb_hcd *hcd, struct usb_host_endpoint *ep, struct urb *urb, - unsigned mem_flags + gfp_t mem_flags ) { struct ohci_hcd *ohci = hcd_to_ohci (hcd); struct ed *ed; diff --git a/drivers/usb/host/ohci-mem.c b/drivers/usb/host/ohci-mem.c index fd3c4d3714bd..9fb83dfb1eb4 100644 --- a/drivers/usb/host/ohci-mem.c +++ b/drivers/usb/host/ohci-mem.c @@ -84,7 +84,7 @@ dma_to_td (struct ohci_hcd *hc, dma_addr_t td_dma) /* TDs ... */ static struct td * -td_alloc (struct ohci_hcd *hc, unsigned mem_flags) +td_alloc (struct ohci_hcd *hc, gfp_t mem_flags) { dma_addr_t dma; struct td *td; @@ -118,7 +118,7 @@ td_free (struct ohci_hcd *hc, struct td *td) /* EDs ... */ static struct ed * -ed_alloc (struct ohci_hcd *hc, unsigned mem_flags) +ed_alloc (struct ohci_hcd *hc, gfp_t mem_flags) { dma_addr_t dma; struct ed *ed; diff --git a/drivers/usb/host/sl811-hcd.c b/drivers/usb/host/sl811-hcd.c index d42a15d10a46..cad858575cea 100644 --- a/drivers/usb/host/sl811-hcd.c +++ b/drivers/usb/host/sl811-hcd.c @@ -818,7 +818,7 @@ static int sl811h_urb_enqueue( struct usb_hcd *hcd, struct usb_host_endpoint *hep, struct urb *urb, - unsigned mem_flags + gfp_t mem_flags ) { struct sl811 *sl811 = hcd_to_sl811(hcd); struct usb_device *udev = urb->dev; diff --git a/drivers/usb/host/uhci-q.c b/drivers/usb/host/uhci-q.c index ea0d168a8c67..4e0fbe2c1a9a 100644 --- a/drivers/usb/host/uhci-q.c +++ b/drivers/usb/host/uhci-q.c @@ -1164,7 +1164,7 @@ static struct urb *uhci_find_urb_ep(struct uhci_hcd *uhci, struct urb *urb) static int uhci_urb_enqueue(struct usb_hcd *hcd, struct usb_host_endpoint *ep, - struct urb *urb, unsigned mem_flags) + struct urb *urb, gfp_t mem_flags) { int ret; struct uhci_hcd *uhci = hcd_to_uhci(hcd); diff --git a/drivers/usb/misc/uss720.c b/drivers/usb/misc/uss720.c index 03fb70ef2eb3..0592cb5e6c4d 100644 --- a/drivers/usb/misc/uss720.c +++ b/drivers/usb/misc/uss720.c @@ -137,7 +137,7 @@ static void async_complete(struct urb *urb, struct pt_regs *ptregs) static struct uss720_async_request *submit_async_request(struct parport_uss720_private *priv, __u8 request, __u8 requesttype, __u16 value, __u16 index, - unsigned int mem_flags) + gfp_t mem_flags) { struct usb_device *usbdev; struct uss720_async_request *rq; @@ -204,7 +204,7 @@ static unsigned int kill_all_async_requests_priv(struct parport_uss720_private * /* --------------------------------------------------------------------- */ -static int get_1284_register(struct parport *pp, unsigned char reg, unsigned char *val, unsigned int mem_flags) +static int get_1284_register(struct parport *pp, unsigned char reg, unsigned char *val, gfp_t mem_flags) { struct parport_uss720_private *priv; struct uss720_async_request *rq; @@ -238,7 +238,7 @@ static int get_1284_register(struct parport *pp, unsigned char reg, unsigned cha return -EIO; } -static int set_1284_register(struct parport *pp, unsigned char reg, unsigned char val, unsigned int mem_flags) +static int set_1284_register(struct parport *pp, unsigned char reg, unsigned char val, gfp_t mem_flags) { struct parport_uss720_private *priv; struct uss720_async_request *rq; diff --git a/drivers/usb/net/asix.c b/drivers/usb/net/asix.c index 861f00a43750..252a34fbb42c 100644 --- a/drivers/usb/net/asix.c +++ b/drivers/usb/net/asix.c @@ -753,7 +753,7 @@ static int ax88772_rx_fixup(struct usbnet *dev, struct sk_buff *skb) } static struct sk_buff *ax88772_tx_fixup(struct usbnet *dev, struct sk_buff *skb, - unsigned flags) + gfp_t flags) { int padlen; int headroom = skb_headroom(skb); diff --git a/drivers/usb/net/gl620a.c b/drivers/usb/net/gl620a.c index c8763ae33c73..c0f263b202a6 100644 --- a/drivers/usb/net/gl620a.c +++ b/drivers/usb/net/gl620a.c @@ -301,7 +301,7 @@ static int genelink_rx_fixup(struct usbnet *dev, struct sk_buff *skb) } static struct sk_buff * -genelink_tx_fixup(struct usbnet *dev, struct sk_buff *skb, unsigned flags) +genelink_tx_fixup(struct usbnet *dev, struct sk_buff *skb, gfp_t flags) { int padlen; int length = skb->len; diff --git a/drivers/usb/net/kaweth.c b/drivers/usb/net/kaweth.c index e04b0ce3611a..c82655d3d448 100644 --- a/drivers/usb/net/kaweth.c +++ b/drivers/usb/net/kaweth.c @@ -477,13 +477,13 @@ static int kaweth_reset(struct kaweth_device *kaweth) } static void kaweth_usb_receive(struct urb *, struct pt_regs *regs); -static int kaweth_resubmit_rx_urb(struct kaweth_device *, unsigned); +static int kaweth_resubmit_rx_urb(struct kaweth_device *, gfp_t); /**************************************************************** int_callback *****************************************************************/ -static void kaweth_resubmit_int_urb(struct kaweth_device *kaweth, int mf) +static void kaweth_resubmit_int_urb(struct kaweth_device *kaweth, gfp_t mf) { int status; @@ -550,7 +550,7 @@ static void kaweth_resubmit_tl(void *d) * kaweth_resubmit_rx_urb ****************************************************************/ static int kaweth_resubmit_rx_urb(struct kaweth_device *kaweth, - unsigned mem_flags) + gfp_t mem_flags) { int result; diff --git a/drivers/usb/net/net1080.c b/drivers/usb/net/net1080.c index a4309c4a491b..cee55f8cf64f 100644 --- a/drivers/usb/net/net1080.c +++ b/drivers/usb/net/net1080.c @@ -500,7 +500,7 @@ static int net1080_rx_fixup(struct usbnet *dev, struct sk_buff *skb) } static struct sk_buff * -net1080_tx_fixup(struct usbnet *dev, struct sk_buff *skb, unsigned flags) +net1080_tx_fixup(struct usbnet *dev, struct sk_buff *skb, gfp_t flags) { int padlen; struct sk_buff *skb2; diff --git a/drivers/usb/net/rndis_host.c b/drivers/usb/net/rndis_host.c index 2ed2e5fb7778..b5a925dc1beb 100644 --- a/drivers/usb/net/rndis_host.c +++ b/drivers/usb/net/rndis_host.c @@ -517,7 +517,7 @@ static int rndis_rx_fixup(struct usbnet *dev, struct sk_buff *skb) } static struct sk_buff * -rndis_tx_fixup(struct usbnet *dev, struct sk_buff *skb, unsigned flags) +rndis_tx_fixup(struct usbnet *dev, struct sk_buff *skb, gfp_t flags) { struct rndis_data_hdr *hdr; struct sk_buff *skb2; diff --git a/drivers/usb/net/usbnet.c b/drivers/usb/net/usbnet.c index 6c460918d54f..fce81d738933 100644 --- a/drivers/usb/net/usbnet.c +++ b/drivers/usb/net/usbnet.c @@ -288,7 +288,7 @@ EXPORT_SYMBOL_GPL(usbnet_defer_kevent); static void rx_complete (struct urb *urb, struct pt_regs *regs); -static void rx_submit (struct usbnet *dev, struct urb *urb, unsigned flags) +static void rx_submit (struct usbnet *dev, struct urb *urb, gfp_t flags) { struct sk_buff *skb; struct skb_data *entry; diff --git a/drivers/usb/net/usbnet.h b/drivers/usb/net/usbnet.h index 7aa0abd1a9bd..89fc4958eecf 100644 --- a/drivers/usb/net/usbnet.h +++ b/drivers/usb/net/usbnet.h @@ -107,7 +107,7 @@ struct driver_info { /* fixup tx packet (add framing) */ struct sk_buff *(*tx_fixup)(struct usbnet *dev, - struct sk_buff *skb, unsigned flags); + struct sk_buff *skb, gfp_t flags); /* for new devices, use the descriptor-reading code instead */ int in; /* rx endpoint */ diff --git a/drivers/usb/net/zaurus.c b/drivers/usb/net/zaurus.c index ee3b892aeabc..5d4b7d55b097 100644 --- a/drivers/usb/net/zaurus.c +++ b/drivers/usb/net/zaurus.c @@ -62,7 +62,7 @@ */ static struct sk_buff * -zaurus_tx_fixup(struct usbnet *dev, struct sk_buff *skb, unsigned flags) +zaurus_tx_fixup(struct usbnet *dev, struct sk_buff *skb, gfp_t flags) { int padlen; struct sk_buff *skb2; diff --git a/drivers/usb/net/zd1201.c b/drivers/usb/net/zd1201.c index c4e479ee926a..2f52261c7cc1 100644 --- a/drivers/usb/net/zd1201.c +++ b/drivers/usb/net/zd1201.c @@ -521,7 +521,7 @@ static int zd1201_setconfig(struct zd1201 *zd, int rid, void *buf, int len, int int reqlen; char seq=0; struct urb *urb; - unsigned int gfp_mask = wait ? GFP_NOIO : GFP_ATOMIC; + gfp_t gfp_mask = wait ? GFP_NOIO : GFP_ATOMIC; len += 4; /* first 4 are for header */ diff --git a/include/linux/usb.h b/include/linux/usb.h index 4dbe580f9335..8f731e8f2821 100644 --- a/include/linux/usb.h +++ b/include/linux/usb.h @@ -933,17 +933,17 @@ static inline void usb_fill_int_urb (struct urb *urb, } extern void usb_init_urb(struct urb *urb); -extern struct urb *usb_alloc_urb(int iso_packets, unsigned mem_flags); +extern struct urb *usb_alloc_urb(int iso_packets, gfp_t mem_flags); extern void usb_free_urb(struct urb *urb); #define usb_put_urb usb_free_urb extern struct urb *usb_get_urb(struct urb *urb); -extern int usb_submit_urb(struct urb *urb, unsigned mem_flags); +extern int usb_submit_urb(struct urb *urb, gfp_t mem_flags); extern int usb_unlink_urb(struct urb *urb); extern void usb_kill_urb(struct urb *urb); #define HAVE_USB_BUFFERS void *usb_buffer_alloc (struct usb_device *dev, size_t size, - unsigned mem_flags, dma_addr_t *dma); + gfp_t mem_flags, dma_addr_t *dma); void usb_buffer_free (struct usb_device *dev, size_t size, void *addr, dma_addr_t dma); @@ -1050,7 +1050,7 @@ int usb_sg_init ( struct scatterlist *sg, int nents, size_t length, - unsigned mem_flags + gfp_t mem_flags ); void usb_sg_cancel (struct usb_sg_request *io); void usb_sg_wait (struct usb_sg_request *io); diff --git a/include/linux/usb_gadget.h b/include/linux/usb_gadget.h index 71e608607324..ff81117eb733 100644 --- a/include/linux/usb_gadget.h +++ b/include/linux/usb_gadget.h @@ -107,18 +107,18 @@ struct usb_ep_ops { int (*disable) (struct usb_ep *ep); struct usb_request *(*alloc_request) (struct usb_ep *ep, - unsigned gfp_flags); + gfp_t gfp_flags); void (*free_request) (struct usb_ep *ep, struct usb_request *req); void *(*alloc_buffer) (struct usb_ep *ep, unsigned bytes, - dma_addr_t *dma, unsigned gfp_flags); + dma_addr_t *dma, gfp_t gfp_flags); void (*free_buffer) (struct usb_ep *ep, void *buf, dma_addr_t dma, unsigned bytes); // NOTE: on 2.6, drivers may also use dma_map() and // dma_sync_single_*() to directly manage dma overhead. int (*queue) (struct usb_ep *ep, struct usb_request *req, - unsigned gfp_flags); + gfp_t gfp_flags); int (*dequeue) (struct usb_ep *ep, struct usb_request *req); int (*set_halt) (struct usb_ep *ep, int value); @@ -214,7 +214,7 @@ usb_ep_disable (struct usb_ep *ep) * Returns the request, or null if one could not be allocated. */ static inline struct usb_request * -usb_ep_alloc_request (struct usb_ep *ep, unsigned gfp_flags) +usb_ep_alloc_request (struct usb_ep *ep, gfp_t gfp_flags) { return ep->ops->alloc_request (ep, gfp_flags); } @@ -254,7 +254,7 @@ usb_ep_free_request (struct usb_ep *ep, struct usb_request *req) */ static inline void * usb_ep_alloc_buffer (struct usb_ep *ep, unsigned len, dma_addr_t *dma, - unsigned gfp_flags) + gfp_t gfp_flags) { return ep->ops->alloc_buffer (ep, len, dma, gfp_flags); } @@ -330,7 +330,7 @@ usb_ep_free_buffer (struct usb_ep *ep, void *buf, dma_addr_t dma, unsigned len) * reported when the usb peripheral is disconnected. */ static inline int -usb_ep_queue (struct usb_ep *ep, struct usb_request *req, unsigned gfp_flags) +usb_ep_queue (struct usb_ep *ep, struct usb_request *req, gfp_t gfp_flags) { return ep->ops->queue (ep, req, gfp_flags); } diff --git a/sound/usb/usbmidi.c b/sound/usb/usbmidi.c index e0d0365453b3..f1a2e2c2e02f 100644 --- a/sound/usb/usbmidi.c +++ b/sound/usb/usbmidi.c @@ -163,7 +163,7 @@ static const uint8_t snd_usbmidi_cin_length[] = { /* * Submits the URB, with error handling. */ -static int snd_usbmidi_submit_urb(struct urb* urb, int flags) +static int snd_usbmidi_submit_urb(struct urb* urb, gfp_t flags) { int err = usb_submit_urb(urb, flags); if (err < 0 && err != -ENODEV) -- cgit v1.2.3 From 9796fdd829da626374458e8706daedcc0e432ddd Mon Sep 17 00:00:00 2001 From: Al Viro Date: Fri, 21 Oct 2005 03:22:03 -0400 Subject: [PATCH] gfp_t: kernel/* Signed-off-by: Al Viro Signed-off-by: Linus Torvalds --- include/linux/audit.h | 4 ++-- include/linux/suspend.h | 2 +- kernel/audit.c | 6 +++--- kernel/auditsc.c | 2 +- kernel/kexec.c | 7 +++---- kernel/power/swsusp.c | 2 +- 6 files changed, 11 insertions(+), 12 deletions(-) (limited to 'include/linux') diff --git a/include/linux/audit.h b/include/linux/audit.h index b2a2509bd7ea..da3c01955f3d 100644 --- a/include/linux/audit.h +++ b/include/linux/audit.h @@ -260,11 +260,11 @@ extern int audit_filter_user(struct netlink_skb_parms *cb, int type); #ifdef CONFIG_AUDIT /* These are defined in audit.c */ /* Public API */ -extern void audit_log(struct audit_context *ctx, int gfp_mask, +extern void audit_log(struct audit_context *ctx, gfp_t gfp_mask, int type, const char *fmt, ...) __attribute__((format(printf,4,5))); -extern struct audit_buffer *audit_log_start(struct audit_context *ctx, int gfp_mask, int type); +extern struct audit_buffer *audit_log_start(struct audit_context *ctx, gfp_t gfp_mask, int type); extern void audit_log_format(struct audit_buffer *ab, const char *fmt, ...) __attribute__((format(printf,2,3))); diff --git a/include/linux/suspend.h b/include/linux/suspend.h index ad15a54806d8..ba448c760168 100644 --- a/include/linux/suspend.h +++ b/include/linux/suspend.h @@ -71,7 +71,7 @@ void restore_processor_state(void); struct saved_context; void __save_processor_state(struct saved_context *ctxt); void __restore_processor_state(struct saved_context *ctxt); -extern unsigned long get_usable_page(unsigned gfp_mask); +extern unsigned long get_usable_page(gfp_t gfp_mask); extern void free_eaten_memory(void); #endif /* _LINUX_SWSUSP_H */ diff --git a/kernel/audit.c b/kernel/audit.c index aefa73a8a586..0c56320d38dc 100644 --- a/kernel/audit.c +++ b/kernel/audit.c @@ -133,7 +133,7 @@ struct audit_buffer { struct list_head list; struct sk_buff *skb; /* formatted skb ready to send */ struct audit_context *ctx; /* NULL or associated context */ - int gfp_mask; + gfp_t gfp_mask; }; static void audit_set_pid(struct audit_buffer *ab, pid_t pid) @@ -647,7 +647,7 @@ static inline void audit_get_stamp(struct audit_context *ctx, * will be written at syscall exit. If there is no associated task, tsk * should be NULL. */ -struct audit_buffer *audit_log_start(struct audit_context *ctx, int gfp_mask, +struct audit_buffer *audit_log_start(struct audit_context *ctx, gfp_t gfp_mask, int type) { struct audit_buffer *ab = NULL; @@ -879,7 +879,7 @@ void audit_log_end(struct audit_buffer *ab) /* Log an audit record. This is a convenience function that calls * audit_log_start, audit_log_vformat, and audit_log_end. It may be * called in any context. */ -void audit_log(struct audit_context *ctx, int gfp_mask, int type, +void audit_log(struct audit_context *ctx, gfp_t gfp_mask, int type, const char *fmt, ...) { struct audit_buffer *ab; diff --git a/kernel/auditsc.c b/kernel/auditsc.c index 88696f639aab..d8a68509e729 100644 --- a/kernel/auditsc.c +++ b/kernel/auditsc.c @@ -803,7 +803,7 @@ static void audit_log_task_info(struct audit_buffer *ab) up_read(&mm->mmap_sem); } -static void audit_log_exit(struct audit_context *context, unsigned int gfp_mask) +static void audit_log_exit(struct audit_context *context, gfp_t gfp_mask) { int i; struct audit_buffer *ab; diff --git a/kernel/kexec.c b/kernel/kexec.c index cdd4dcd8fb63..36c5d9cd4cc1 100644 --- a/kernel/kexec.c +++ b/kernel/kexec.c @@ -90,7 +90,7 @@ int kexec_should_crash(struct task_struct *p) static int kimage_is_destination_range(struct kimage *image, unsigned long start, unsigned long end); static struct page *kimage_alloc_page(struct kimage *image, - unsigned int gfp_mask, + gfp_t gfp_mask, unsigned long dest); static int do_kimage_alloc(struct kimage **rimage, unsigned long entry, @@ -326,8 +326,7 @@ static int kimage_is_destination_range(struct kimage *image, return 0; } -static struct page *kimage_alloc_pages(unsigned int gfp_mask, - unsigned int order) +static struct page *kimage_alloc_pages(gfp_t gfp_mask, unsigned int order) { struct page *pages; @@ -654,7 +653,7 @@ static kimage_entry_t *kimage_dst_used(struct kimage *image, } static struct page *kimage_alloc_page(struct kimage *image, - unsigned int gfp_mask, + gfp_t gfp_mask, unsigned long destination) { /* diff --git a/kernel/power/swsusp.c b/kernel/power/swsusp.c index 2d5c45676442..10bc5ec496d7 100644 --- a/kernel/power/swsusp.c +++ b/kernel/power/swsusp.c @@ -1095,7 +1095,7 @@ static inline void eat_page(void *page) *eaten_memory = c; } -unsigned long get_usable_page(unsigned gfp_mask) +unsigned long get_usable_page(gfp_t gfp_mask) { unsigned long m; -- cgit v1.2.3 From b4e3ca1ab1ae9ae86134126dcdc88da1caaa32ca Mon Sep 17 00:00:00 2001 From: Al Viro Date: Fri, 21 Oct 2005 03:22:34 -0400 Subject: [PATCH] gfp_t: remaining bits of drivers/* Signed-off-by: Al Viro Signed-off-by: Linus Torvalds --- drivers/block/loop.c | 2 +- drivers/block/rd.c | 2 +- drivers/char/n_tty.c | 2 +- drivers/ieee1394/eth1394.c | 2 +- drivers/md/bitmap.c | 2 +- drivers/md/dm-crypt.c | 2 +- drivers/s390/net/fsm.c | 2 +- drivers/s390/net/fsm.h | 2 +- include/linux/i2o.h | 4 ++-- include/linux/loop.h | 2 +- 10 files changed, 11 insertions(+), 11 deletions(-) (limited to 'include/linux') diff --git a/drivers/block/loop.c b/drivers/block/loop.c index b35e08876dd4..96c664af8d06 100644 --- a/drivers/block/loop.c +++ b/drivers/block/loop.c @@ -881,7 +881,7 @@ loop_init_xfer(struct loop_device *lo, struct loop_func_table *xfer, static int loop_clr_fd(struct loop_device *lo, struct block_device *bdev) { struct file *filp = lo->lo_backing_file; - int gfp = lo->old_gfp_mask; + gfp_t gfp = lo->old_gfp_mask; if (lo->lo_state != Lo_bound) return -ENXIO; diff --git a/drivers/block/rd.c b/drivers/block/rd.c index 145c1fbffe01..68c60a5bcdab 100644 --- a/drivers/block/rd.c +++ b/drivers/block/rd.c @@ -348,7 +348,7 @@ static int rd_open(struct inode *inode, struct file *filp) struct block_device *bdev = inode->i_bdev; struct address_space *mapping; unsigned bsize; - int gfp_mask; + gfp_t gfp_mask; inode = igrab(bdev->bd_inode); rd_bdev[unit] = bdev; diff --git a/drivers/char/n_tty.c b/drivers/char/n_tty.c index c9bdf544ed2c..c556f4d3ccd7 100644 --- a/drivers/char/n_tty.c +++ b/drivers/char/n_tty.c @@ -62,7 +62,7 @@ static inline unsigned char *alloc_buf(void) { - unsigned int prio = in_interrupt() ? GFP_ATOMIC : GFP_KERNEL; + gfp_t prio = in_interrupt() ? GFP_ATOMIC : GFP_KERNEL; if (PAGE_SIZE != N_TTY_BUF_SIZE) return kmalloc(N_TTY_BUF_SIZE, prio); diff --git a/drivers/ieee1394/eth1394.c b/drivers/ieee1394/eth1394.c index 4802bbbb6dc9..c9e92d85c893 100644 --- a/drivers/ieee1394/eth1394.c +++ b/drivers/ieee1394/eth1394.c @@ -1630,7 +1630,7 @@ static void ether1394_complete_cb(void *__ptask) /* Transmit a packet (called by kernel) */ static int ether1394_tx (struct sk_buff *skb, struct net_device *dev) { - int kmflags = in_interrupt() ? GFP_ATOMIC : GFP_KERNEL; + gfp_t kmflags = in_interrupt() ? GFP_ATOMIC : GFP_KERNEL; struct eth1394hdr *eth; struct eth1394_priv *priv = netdev_priv(dev); int proto; diff --git a/drivers/md/bitmap.c b/drivers/md/bitmap.c index 2fba2bbe72d8..01654fcabc52 100644 --- a/drivers/md/bitmap.c +++ b/drivers/md/bitmap.c @@ -91,7 +91,7 @@ int bitmap_active(struct bitmap *bitmap) #define WRITE_POOL_SIZE 256 /* mempool for queueing pending writes on the bitmap file */ -static void *write_pool_alloc(unsigned int gfp_flags, void *data) +static void *write_pool_alloc(gfp_t gfp_flags, void *data) { return kmalloc(sizeof(struct page_list), gfp_flags); } diff --git a/drivers/md/dm-crypt.c b/drivers/md/dm-crypt.c index b6148f6f7836..28c1a628621f 100644 --- a/drivers/md/dm-crypt.c +++ b/drivers/md/dm-crypt.c @@ -331,7 +331,7 @@ crypt_alloc_buffer(struct crypt_config *cc, unsigned int size, { struct bio *bio; unsigned int nr_iovecs = (size + PAGE_SIZE - 1) >> PAGE_SHIFT; - int gfp_mask = GFP_NOIO | __GFP_HIGHMEM; + gfp_t gfp_mask = GFP_NOIO | __GFP_HIGHMEM; unsigned int i; /* diff --git a/drivers/s390/net/fsm.c b/drivers/s390/net/fsm.c index fa09440d82e5..38f50b7129a2 100644 --- a/drivers/s390/net/fsm.c +++ b/drivers/s390/net/fsm.c @@ -16,7 +16,7 @@ MODULE_LICENSE("GPL"); fsm_instance * init_fsm(char *name, const char **state_names, const char **event_names, int nr_states, - int nr_events, const fsm_node *tmpl, int tmpl_len, int order) + int nr_events, const fsm_node *tmpl, int tmpl_len, gfp_t order) { int i; fsm_instance *this; diff --git a/drivers/s390/net/fsm.h b/drivers/s390/net/fsm.h index f9a011001eb6..1b8a7e7c34f3 100644 --- a/drivers/s390/net/fsm.h +++ b/drivers/s390/net/fsm.h @@ -110,7 +110,7 @@ extern fsm_instance * init_fsm(char *name, const char **state_names, const char **event_names, int nr_states, int nr_events, const fsm_node *tmpl, - int tmpl_len, int order); + int tmpl_len, gfp_t order); /** * Releases an FSM diff --git a/include/linux/i2o.h b/include/linux/i2o.h index bdc286ec947c..b4af45aad25d 100644 --- a/include/linux/i2o.h +++ b/include/linux/i2o.h @@ -492,7 +492,7 @@ static inline int i2o_dma_map_sg(struct i2o_controller *c, * Returns 0 on success or -ENOMEM on failure. */ static inline int i2o_dma_alloc(struct device *dev, struct i2o_dma *addr, - size_t len, unsigned int gfp_mask) + size_t len, gfp_t gfp_mask) { struct pci_dev *pdev = to_pci_dev(dev); int dma_64 = 0; @@ -551,7 +551,7 @@ static inline void i2o_dma_free(struct device *dev, struct i2o_dma *addr) * Returns the 0 on success or negative error code on failure. */ static inline int i2o_dma_realloc(struct device *dev, struct i2o_dma *addr, - size_t len, unsigned int gfp_mask) + size_t len, gfp_t gfp_mask) { i2o_dma_free(dev, addr); diff --git a/include/linux/loop.h b/include/linux/loop.h index 53fa51595443..40f63c9879d2 100644 --- a/include/linux/loop.h +++ b/include/linux/loop.h @@ -52,7 +52,7 @@ struct loop_device { unsigned lo_blocksize; void *key_data; - int old_gfp_mask; + gfp_t old_gfp_mask; spinlock_t lo_lock; struct bio *lo_bio; -- cgit v1.2.3 From 260b23674fdb570f3235ce55892246bef1c24c2a Mon Sep 17 00:00:00 2001 From: Al Viro Date: Fri, 21 Oct 2005 03:22:44 -0400 Subject: [PATCH] gfp_t: the rest zone handling, mapping->flags handling Signed-off-by: Al Viro Signed-off-by: Linus Torvalds --- include/linux/mmzone.h | 2 +- include/linux/pagemap.h | 7 ++++--- mm/highmem.c | 14 +++++++++----- mm/page_alloc.c | 29 +++++++++++++++-------------- 4 files changed, 29 insertions(+), 23 deletions(-) (limited to 'include/linux') diff --git a/include/linux/mmzone.h b/include/linux/mmzone.h index 5ed471b58f4f..7519eb4191e7 100644 --- a/include/linux/mmzone.h +++ b/include/linux/mmzone.h @@ -302,7 +302,7 @@ void get_zone_counts(unsigned long *active, unsigned long *inactive, void build_all_zonelists(void); void wakeup_kswapd(struct zone *zone, int order); int zone_watermark_ok(struct zone *z, int order, unsigned long mark, - int alloc_type, int can_try_harder, int gfp_high); + int alloc_type, int can_try_harder, gfp_t gfp_high); #ifdef CONFIG_HAVE_MEMORY_PRESENT void memory_present(int nid, unsigned long start, unsigned long end); diff --git a/include/linux/pagemap.h b/include/linux/pagemap.h index efbae53fb078..ba6c310a055f 100644 --- a/include/linux/pagemap.h +++ b/include/linux/pagemap.h @@ -21,16 +21,17 @@ static inline gfp_t mapping_gfp_mask(struct address_space * mapping) { - return mapping->flags & __GFP_BITS_MASK; + return (__force gfp_t)mapping->flags & __GFP_BITS_MASK; } /* * This is non-atomic. Only to be used before the mapping is activated. * Probably needs a barrier... */ -static inline void mapping_set_gfp_mask(struct address_space *m, int mask) +static inline void mapping_set_gfp_mask(struct address_space *m, gfp_t mask) { - m->flags = (m->flags & ~__GFP_BITS_MASK) | mask; + m->flags = (m->flags & ~(__force unsigned long)__GFP_BITS_MASK) | + (__force unsigned long)mask; } /* diff --git a/mm/highmem.c b/mm/highmem.c index 90e1861e2da0..ce2e7e8bbfa7 100644 --- a/mm/highmem.c +++ b/mm/highmem.c @@ -30,11 +30,9 @@ static mempool_t *page_pool, *isa_page_pool; -static void *page_pool_alloc(gfp_t gfp_mask, void *data) +static void *page_pool_alloc_isa(gfp_t gfp_mask, void *data) { - unsigned int gfp = gfp_mask | (unsigned int) (long) data; - - return alloc_page(gfp); + return alloc_page(gfp_mask | GFP_DMA); } static void page_pool_free(void *page, void *data) @@ -51,6 +49,12 @@ static void page_pool_free(void *page, void *data) * n means that there are (n-1) current users of it. */ #ifdef CONFIG_HIGHMEM + +static void *page_pool_alloc(gfp_t gfp_mask, void *data) +{ + return alloc_page(gfp_mask); +} + static int pkmap_count[LAST_PKMAP]; static unsigned int last_pkmap_nr; static __cacheline_aligned_in_smp DEFINE_SPINLOCK(kmap_lock); @@ -267,7 +271,7 @@ int init_emergency_isa_pool(void) if (isa_page_pool) return 0; - isa_page_pool = mempool_create(ISA_POOL_SIZE, page_pool_alloc, page_pool_free, (void *) __GFP_DMA); + isa_page_pool = mempool_create(ISA_POOL_SIZE, page_pool_alloc_isa, page_pool_free, NULL); if (!isa_page_pool) BUG(); diff --git a/mm/page_alloc.c b/mm/page_alloc.c index aa43ae3ab8c9..94c864eac9c4 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -734,7 +734,7 @@ buffered_rmqueue(struct zone *zone, int order, gfp_t gfp_flags) * of the allocation. */ int zone_watermark_ok(struct zone *z, int order, unsigned long mark, - int classzone_idx, int can_try_harder, int gfp_high) + int classzone_idx, int can_try_harder, gfp_t gfp_high) { /* free_pages my go negative - that's OK */ long min = mark, free_pages = z->free_pages - (1 << order) + 1; @@ -777,7 +777,7 @@ struct page * fastcall __alloc_pages(gfp_t gfp_mask, unsigned int order, struct zonelist *zonelist) { - const int wait = gfp_mask & __GFP_WAIT; + const gfp_t wait = gfp_mask & __GFP_WAIT; struct zone **zones, *z; struct page *page; struct reclaim_state reclaim_state; @@ -996,7 +996,7 @@ fastcall unsigned long get_zeroed_page(gfp_t gfp_mask) * get_zeroed_page() returns a 32-bit address, which cannot represent * a highmem page */ - BUG_ON(gfp_mask & __GFP_HIGHMEM); + BUG_ON((gfp_mask & __GFP_HIGHMEM) != 0); page = alloc_pages(gfp_mask | __GFP_ZERO, 0); if (page) @@ -1428,6 +1428,16 @@ static int __init build_zonelists_node(pg_data_t *pgdat, struct zonelist *zoneli return j; } +static inline int highest_zone(int zone_bits) +{ + int res = ZONE_NORMAL; + if (zone_bits & (__force int)__GFP_HIGHMEM) + res = ZONE_HIGHMEM; + if (zone_bits & (__force int)__GFP_DMA) + res = ZONE_DMA; + return res; +} + #ifdef CONFIG_NUMA #define MAX_NODE_LOAD (num_online_nodes()) static int __initdata node_load[MAX_NUMNODES]; @@ -1524,11 +1534,7 @@ static void __init build_zonelists(pg_data_t *pgdat) zonelist = pgdat->node_zonelists + i; for (j = 0; zonelist->zones[j] != NULL; j++); - k = ZONE_NORMAL; - if (i & __GFP_HIGHMEM) - k = ZONE_HIGHMEM; - if (i & __GFP_DMA) - k = ZONE_DMA; + k = highest_zone(i); j = build_zonelists_node(NODE_DATA(node), zonelist, j, k); zonelist->zones[j] = NULL; @@ -1549,12 +1555,7 @@ static void __init build_zonelists(pg_data_t *pgdat) zonelist = pgdat->node_zonelists + i; j = 0; - k = ZONE_NORMAL; - if (i & __GFP_HIGHMEM) - k = ZONE_HIGHMEM; - if (i & __GFP_DMA) - k = ZONE_DMA; - + k = highest_zone(i); j = build_zonelists_node(pgdat, zonelist, j, k); /* * Now we build the zonelist so that it contains the zones -- cgit v1.2.3 From 8573b80f020dce7aefa3237f1e932d562b65323d Mon Sep 17 00:00:00 2001 From: Erik Hovland Date: Fri, 28 Oct 2005 16:28:04 +0100 Subject: [ARM] 3031/1: fix typos in comments of mmc.h Patch from Erik Hovland I noticed that the same typo (i before c in associated) showed up twice in the file kernel/include/linux/mmc/mmc.h. This patch fixes both of the instances I found with this mistake. The typos are in comments and should have no affect on working code. E Signed-off-by: Erik Hovland Signed-off-by: Russell King --- include/linux/mmc/mmc.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'include/linux') diff --git a/include/linux/mmc/mmc.h b/include/linux/mmc/mmc.h index 1ab78e8d6c53..aef6042f8f0b 100644 --- a/include/linux/mmc/mmc.h +++ b/include/linux/mmc/mmc.h @@ -50,7 +50,7 @@ struct mmc_command { #define MMC_ERR_INVALID 5 struct mmc_data *data; /* data segment associated with cmd */ - struct mmc_request *mrq; /* assoicated request */ + struct mmc_request *mrq; /* associated request */ }; struct mmc_data { @@ -68,7 +68,7 @@ struct mmc_data { unsigned int bytes_xfered; struct mmc_command *stop; /* stop command */ - struct mmc_request *mrq; /* assoicated request */ + struct mmc_request *mrq; /* associated request */ unsigned int sg_len; /* size of scatter list */ struct scatterlist *sg; /* I/O scatter list */ -- cgit v1.2.3 From 0ac85241ebc7bf6b86ab498960cc121d53ef69ae Mon Sep 17 00:00:00 2001 From: David Brownell Date: Mon, 12 Sep 2005 19:39:34 -0700 Subject: [PATCH] driver model wakeup flags This is a refresh of an earlier patch to add "wakeup" support to the PM core model. This provides per-device bus-neutral control of the use of wakeup events. * "struct device_pm_info" has two bits that are initialized as part of setting up the enclosing struct device: - "can_wakeup", reflecting hardware capabilities - "may_wakeup", the policy setting (when CONFIG_PM) * There's a writeable sysfs "wakeup" file, with one of two values: - "enabled", when the policy is to allow wakeup - "disabled", when the policy is not to allow it - "" if the device can't currently issue wakeups By default, wakeup is enabled on all devices that support it. If its driver doesn't support it ... treat it as a bug. :) Signed-off-by: David Brownell Signed-off-by: Greg Kroah-Hartman --- drivers/base/core.c | 1 + drivers/base/power/sysfs.c | 73 ++++++++++++++++++++++++++++++++++++++++++++++ include/linux/pm.h | 26 ++++++++++++++++- 3 files changed, 99 insertions(+), 1 deletion(-) (limited to 'include/linux') diff --git a/drivers/base/core.c b/drivers/base/core.c index 6ab73f5c799a..31109193e2c4 100644 --- a/drivers/base/core.c +++ b/drivers/base/core.c @@ -225,6 +225,7 @@ void device_initialize(struct device *dev) klist_children_put); INIT_LIST_HEAD(&dev->dma_pools); init_MUTEX(&dev->sem); + device_init_wakeup(dev, 0); } /** diff --git a/drivers/base/power/sysfs.c b/drivers/base/power/sysfs.c index 8d04fb435c17..89c57875f3e5 100644 --- a/drivers/base/power/sysfs.c +++ b/drivers/base/power/sysfs.c @@ -48,8 +48,81 @@ static ssize_t state_store(struct device * dev, struct device_attribute *attr, c static DEVICE_ATTR(state, 0644, state_show, state_store); +/* + * wakeup - Report/change current wakeup option for device + * + * Some devices support "wakeup" events, which are hardware signals + * used to activate devices from suspended or low power states. Such + * devices have one of three values for the sysfs power/wakeup file: + * + * + "enabled\n" to issue the events; + * + "disabled\n" not to do so; or + * + "\n" for temporary or permanent inability to issue wakeup. + * + * (For example, unconfigured USB devices can't issue wakeups.) + * + * Familiar examples of devices that can issue wakeup events include + * keyboards and mice (both PS2 and USB styles), power buttons, modems, + * "Wake-On-LAN" Ethernet links, GPIO lines, and more. Some events + * will wake the entire system from a suspend state; others may just + * wake up the device (if the system as a whole is already active). + * Some wakeup events use normal IRQ lines; other use special out + * of band signaling. + * + * It is the responsibility of device drivers to enable (or disable) + * wakeup signaling as part of changing device power states, respecting + * the policy choices provided through the driver model. + * + * Devices may not be able to generate wakeup events from all power + * states. Also, the events may be ignored in some configurations; + * for example, they might need help from other devices that aren't + * active, or which may have wakeup disabled. Some drivers rely on + * wakeup events internally (unless they are disabled), keeping + * their hardware in low power modes whenever they're unused. This + * saves runtime power, without requiring system-wide sleep states. + */ + +static const char enabled[] = "enabled"; +static const char disabled[] = "disabled"; + +static ssize_t +wake_show(struct device * dev, struct device_attribute *attr, char * buf) +{ + return sprintf(buf, "%s\n", device_can_wakeup(dev) + ? (device_may_wakeup(dev) ? enabled : disabled) + : ""); +} + +static ssize_t +wake_store(struct device * dev, struct device_attribute *attr, + const char * buf, size_t n) +{ + char *cp; + int len = n; + + if (!device_can_wakeup(dev)) + return -EINVAL; + + cp = memchr(buf, '\n', n); + if (cp) + len = cp - buf; + if (len == sizeof enabled - 1 + && strncmp(buf, enabled, sizeof enabled - 1) == 0) + device_set_wakeup_enable(dev, 1); + else if (len == sizeof disabled - 1 + && strncmp(buf, disabled, sizeof disabled - 1) == 0) + device_set_wakeup_enable(dev, 0); + else + return -EINVAL; + return n; +} + +static DEVICE_ATTR(wakeup, 0644, wake_show, wake_store); + + static struct attribute * power_attrs[] = { &dev_attr_state.attr, + &dev_attr_wakeup.attr, NULL, }; static struct attribute_group pm_attr_group = { diff --git a/include/linux/pm.h b/include/linux/pm.h index 5cfb07648eca..7897cf500c51 100644 --- a/include/linux/pm.h +++ b/include/linux/pm.h @@ -219,7 +219,9 @@ typedef struct pm_message { struct dev_pm_info { pm_message_t power_state; + unsigned can_wakeup:1; #ifdef CONFIG_PM + unsigned should_wakeup:1; pm_message_t prev_state; void * saved_state; atomic_t pm_users; @@ -236,13 +238,35 @@ extern void device_resume(void); #ifdef CONFIG_PM extern int device_suspend(pm_message_t state); -#else + +#define device_set_wakeup_enable(dev,val) \ + ((dev)->power.should_wakeup = !!(val)) +#define device_may_wakeup(dev) \ + (device_can_wakeup(dev) && (dev)->power.should_wakeup) + +#else /* !CONFIG_PM */ + static inline int device_suspend(pm_message_t state) { return 0; } + +#define device_set_wakeup_enable(dev,val) do{}while(0) +#define device_may_wakeup(dev) (0) + #endif +/* changes to device_may_wakeup take effect on the next pm state change. + * by default, devices should wakeup if they can. + */ +#define device_can_wakeup(dev) \ + ((dev)->power.can_wakeup) +#define device_init_wakeup(dev,val) \ + do { \ + device_can_wakeup(dev) = !!(val); \ + device_set_wakeup_enable(dev,val); \ + } while(0) + #endif /* __KERNEL__ */ #endif /* _LINUX_PM_H */ -- cgit v1.2.3 From 607cf4d9aa1d766890f42fc892d39d48cf6d6c16 Mon Sep 17 00:00:00 2001 From: Greg Kroah-Hartman Date: Thu, 27 Oct 2005 22:25:43 -0700 Subject: [PATCH] I2O: Clean up some pretty bad driver model abuses in the i2o code Signed-off-by: Greg Kroah-Hartman --- drivers/message/i2o/iop.c | 22 ++++++++++------------ include/linux/i2o.h | 2 +- 2 files changed, 11 insertions(+), 13 deletions(-) (limited to 'include/linux') diff --git a/drivers/message/i2o/iop.c b/drivers/message/i2o/iop.c index 42f8b810d6e5..15deb45ce995 100644 --- a/drivers/message/i2o/iop.c +++ b/drivers/message/i2o/iop.c @@ -833,6 +833,7 @@ void i2o_iop_remove(struct i2o_controller *c) list_for_each_entry_safe(dev, tmp, &c->devices, list) i2o_device_remove(dev); + class_device_unregister(c->classdev); device_del(&c->device); /* Ask the IOP to switch to RESET state */ @@ -1077,9 +1078,7 @@ static void i2o_iop_release(struct device *dev) }; /* I2O controller class */ -static struct class i2o_controller_class = { - .name = "i2o_controller", -}; +static struct class *i2o_controller_class; /** * i2o_iop_alloc - Allocate and initialize a i2o_controller struct @@ -1110,14 +1109,10 @@ struct i2o_controller *i2o_iop_alloc(void) sprintf(c->name, "iop%d", c->unit); device_initialize(&c->device); - class_device_initialize(&c->classdev); c->device.release = &i2o_iop_release; - c->classdev.class = &i2o_controller_class; - c->classdev.dev = &c->device; snprintf(c->device.bus_id, BUS_ID_SIZE, "iop%d", c->unit); - snprintf(c->classdev.class_id, BUS_ID_SIZE, "iop%d", c->unit); #if BITS_PER_LONG == 64 spin_lock_init(&c->context_list_lock); @@ -1146,7 +1141,9 @@ int i2o_iop_add(struct i2o_controller *c) goto iop_reset; } - if ((rc = class_device_add(&c->classdev))) { + c->classdev = class_device_create(i2o_controller_class, 0, + &c->device, "iop%d", c->unit); + if (IS_ERR(c->classdev)) { osm_err("%s: could not add controller class\n", c->name); goto device_del; } @@ -1184,7 +1181,7 @@ int i2o_iop_add(struct i2o_controller *c) return 0; class_del: - class_device_del(&c->classdev); + class_device_unregister(c->classdev); device_del: device_del(&c->device); @@ -1250,7 +1247,8 @@ static int __init i2o_iop_init(void) if (rc) goto exit; - if ((rc = class_register(&i2o_controller_class))) { + i2o_controller_class = class_create(THIS_MODULE, "i2o_controller"); + if (IS_ERR(i2o_controller_class)) { osm_err("can't register class i2o_controller\n"); goto device_exit; } @@ -1273,7 +1271,7 @@ static int __init i2o_iop_init(void) i2o_driver_exit(); class_exit: - class_unregister(&i2o_controller_class); + class_destroy(i2o_controller_class); device_exit: i2o_device_exit(); @@ -1292,7 +1290,7 @@ static void __exit i2o_iop_exit(void) i2o_pci_exit(); i2o_exec_exit(); i2o_driver_exit(); - class_unregister(&i2o_controller_class); + class_destroy(i2o_controller_class); i2o_device_exit(); }; diff --git a/include/linux/i2o.h b/include/linux/i2o.h index bdc286ec947c..694ea298d4bc 100644 --- a/include/linux/i2o.h +++ b/include/linux/i2o.h @@ -194,7 +194,7 @@ struct i2o_controller { struct resource mem_resource; /* Mem resource allocated to the IOP */ struct device device; - struct class_device classdev; /* I2O controller class */ + struct class_device *classdev; /* I2O controller class device */ struct i2o_device *exec; /* Executive */ #if BITS_PER_LONG == 64 spinlock_t context_list_lock; /* lock for context_list */ -- cgit v1.2.3 From 7bd7b091429705eb281d60c553cc643aada8045a Mon Sep 17 00:00:00 2001 From: Dmitry Torokhov Date: Thu, 29 Sep 2005 00:40:07 -0500 Subject: [PATCH] I2O: remove i2o_device_class I2O: cleanup - remove i2o_device_class I2O devices reside on their own bus so there should be no reason to also have i2c_device class that mirros i2o bus. Signed-off-by: Dmitry Torokhov Signed-off-by: Greg Kroah-Hartman --- drivers/message/i2o/core.h | 3 -- drivers/message/i2o/device.c | 71 ++++++++------------------------------------ drivers/message/i2o/driver.c | 3 ++ drivers/message/i2o/iop.c | 10 +------ include/linux/i2o.h | 2 -- 5 files changed, 17 insertions(+), 72 deletions(-) (limited to 'include/linux') diff --git a/drivers/message/i2o/core.h b/drivers/message/i2o/core.h index c5bcfd70f711..9eefedb16211 100644 --- a/drivers/message/i2o/core.h +++ b/drivers/message/i2o/core.h @@ -36,9 +36,6 @@ extern void __exit i2o_pci_exit(void); extern void i2o_device_remove(struct i2o_device *); extern int i2o_device_parse_lct(struct i2o_controller *); -extern int i2o_device_init(void); -extern void i2o_device_exit(void); - /* IOP */ extern struct i2o_controller *i2o_iop_alloc(void); extern void i2o_iop_free(struct i2o_controller *); diff --git a/drivers/message/i2o/device.c b/drivers/message/i2o/device.c index 551d582e2887..d9879965eb50 100644 --- a/drivers/message/i2o/device.c +++ b/drivers/message/i2o/device.c @@ -138,17 +138,6 @@ static void i2o_device_release(struct device *dev) kfree(i2o_dev); } -/** - * i2o_device_class_release - I2O class device release function - * @cd: I2O class device which is added to the I2O device class - * - * The function is just a stub - memory will be freed when - * associated I2O device is released. - */ -static void i2o_device_class_release(struct class_device *cd) -{ - /* empty */ -} /** * i2o_device_class_show_class_id - Displays class id of I2O device @@ -157,12 +146,13 @@ static void i2o_device_class_release(struct class_device *cd) * * Returns the number of bytes which are printed into the buffer. */ -static ssize_t i2o_device_class_show_class_id(struct class_device *cd, - char *buf) +static ssize_t i2o_device_show_class_id(struct device *dev, + struct device_attribute *attr, + char *buf) { - struct i2o_device *dev = to_i2o_device(cd->dev); + struct i2o_device *i2o_dev = to_i2o_device(dev); - sprintf(buf, "0x%03x\n", dev->lct_data.class_id); + sprintf(buf, "0x%03x\n", i2o_dev->lct_data.class_id); return strlen(buf) + 1; } @@ -173,27 +163,22 @@ static ssize_t i2o_device_class_show_class_id(struct class_device *cd, * * Returns the number of bytes which are printed into the buffer. */ -static ssize_t i2o_device_class_show_tid(struct class_device *cd, char *buf) +static ssize_t i2o_device_show_tid(struct device *dev, + struct device_attribute *attr, + char *buf) { - struct i2o_device *dev = to_i2o_device(cd->dev); + struct i2o_device *i2o_dev = to_i2o_device(dev); - sprintf(buf, "0x%03x\n", dev->lct_data.tid); + sprintf(buf, "0x%03x\n", i2o_dev->lct_data.tid); return strlen(buf) + 1; } -static struct class_device_attribute i2o_device_class_attrs[] = { - __ATTR(class_id, S_IRUGO, i2o_device_class_show_class_id, NULL), - __ATTR(tid, S_IRUGO, i2o_device_class_show_tid, NULL), +struct device_attribute i2o_device_attrs[] = { + __ATTR(class_id, S_IRUGO, i2o_device_show_class_id, NULL), + __ATTR(tid, S_IRUGO, i2o_device_show_tid, NULL), __ATTR_NULL }; -/* I2O device class */ -static struct class i2o_device_class = { - .name = "i2o_device", - .release = i2o_device_class_release, - .class_dev_attrs = i2o_device_class_attrs, -}; - /** * i2o_device_alloc - Allocate a I2O device and initialize it * @@ -217,8 +202,6 @@ static struct i2o_device *i2o_device_alloc(void) dev->device.bus = &i2o_bus_type; dev->device.release = &i2o_device_release; - dev->classdev.class = &i2o_device_class; - dev->classdev.dev = &dev->device; return dev; } @@ -311,17 +294,12 @@ static struct i2o_device *i2o_device_add(struct i2o_controller *c, snprintf(dev->device.bus_id, BUS_ID_SIZE, "%d:%03x", c->unit, dev->lct_data.tid); - snprintf(dev->classdev.class_id, BUS_ID_SIZE, "%d:%03x", c->unit, - dev->lct_data.tid); - dev->device.parent = &c->device; device_register(&dev->device); list_add_tail(&dev->list, &c->devices); - class_device_register(&dev->classdev); - i2o_setup_sysfs_links(dev); i2o_driver_notify_device_add_all(dev); @@ -343,7 +321,6 @@ void i2o_device_remove(struct i2o_device *i2o_dev) { i2o_driver_notify_device_remove_all(i2o_dev); i2o_remove_sysfs_links(i2o_dev); - class_device_unregister(&i2o_dev->classdev); list_del(&i2o_dev->list); device_unregister(&i2o_dev->device); } @@ -598,28 +575,6 @@ int i2o_parm_table_get(struct i2o_device *dev, int oper, int group, return size; } -/** - * i2o_device_init - Initialize I2O devices - * - * Registers the I2O device class. - * - * Returns 0 on success or negative error code on failure. - */ -int i2o_device_init(void) -{ - return class_register(&i2o_device_class); -} - -/** - * i2o_device_exit - I2O devices exit function - * - * Unregisters the I2O device class. - */ -void i2o_device_exit(void) -{ - class_unregister(&i2o_device_class); -} - EXPORT_SYMBOL(i2o_device_claim); EXPORT_SYMBOL(i2o_device_claim_release); EXPORT_SYMBOL(i2o_parm_field_get); diff --git a/drivers/message/i2o/driver.c b/drivers/message/i2o/driver.c index 739bfdef0c6d..0079a4be0af2 100644 --- a/drivers/message/i2o/driver.c +++ b/drivers/message/i2o/driver.c @@ -58,9 +58,12 @@ static int i2o_bus_match(struct device *dev, struct device_driver *drv) }; /* I2O bus type */ +extern struct device_attribute i2o_device_attrs[]; + struct bus_type i2o_bus_type = { .name = "i2o", .match = i2o_bus_match, + .dev_attrs = i2o_device_attrs, }; /** diff --git a/drivers/message/i2o/iop.c b/drivers/message/i2o/iop.c index 15deb45ce995..176fb573ea26 100644 --- a/drivers/message/i2o/iop.c +++ b/drivers/message/i2o/iop.c @@ -1243,14 +1243,10 @@ static int __init i2o_iop_init(void) printk(KERN_INFO OSM_DESCRIPTION " v" OSM_VERSION "\n"); - rc = i2o_device_init(); - if (rc) - goto exit; - i2o_controller_class = class_create(THIS_MODULE, "i2o_controller"); if (IS_ERR(i2o_controller_class)) { osm_err("can't register class i2o_controller\n"); - goto device_exit; + goto exit; } if ((rc = i2o_driver_init())) @@ -1273,9 +1269,6 @@ static int __init i2o_iop_init(void) class_exit: class_destroy(i2o_controller_class); - device_exit: - i2o_device_exit(); - exit: return rc; } @@ -1291,7 +1284,6 @@ static void __exit i2o_iop_exit(void) i2o_exec_exit(); i2o_driver_exit(); class_destroy(i2o_controller_class); - i2o_device_exit(); }; module_init(i2o_iop_init); diff --git a/include/linux/i2o.h b/include/linux/i2o.h index 694ea298d4bc..84db8f6ae7e5 100644 --- a/include/linux/i2o.h +++ b/include/linux/i2o.h @@ -66,8 +66,6 @@ struct i2o_device { struct device device; struct semaphore lock; /* device lock */ - - struct class_device classdev; /* i2o device class */ }; /* -- cgit v1.2.3 From d8539d81aeee4dbdc0624a798321e822fb2df7ae Mon Sep 17 00:00:00 2001 From: Dmitry Torokhov Date: Thu, 15 Sep 2005 02:01:36 -0500 Subject: [PATCH] Driver core: pass interface to class interface methods Driver core: pass interface to class intreface methods Pass interface as argument to add() and remove() class interface methods. This way a subsystem can implement generic add/remove handlers and then call interface-specific ones. Signed-off-by: Dmitry Torokhov Signed-off-by: Greg Kroah-Hartman --- drivers/base/class.c | 8 ++++---- drivers/pcmcia/ds.c | 6 ++++-- drivers/pcmcia/rsrc_nonstatic.c | 6 ++++-- drivers/pcmcia/socket_sysfs.c | 6 ++++-- drivers/scsi/sg.c | 8 ++++---- include/linux/device.h | 4 ++-- 6 files changed, 22 insertions(+), 16 deletions(-) (limited to 'include/linux') diff --git a/drivers/base/class.c b/drivers/base/class.c index 8df58c57dd25..73d44cf53db0 100644 --- a/drivers/base/class.c +++ b/drivers/base/class.c @@ -532,7 +532,7 @@ int class_device_add(struct class_device *class_dev) list_add_tail(&class_dev->node, &parent->children); list_for_each_entry(class_intf, &parent->interfaces, node) if (class_intf->add) - class_intf->add(class_dev); + class_intf->add(class_dev, class_intf); up(&parent->sem); } @@ -612,7 +612,7 @@ void class_device_del(struct class_device *class_dev) list_del_init(&class_dev->node); list_for_each_entry(class_intf, &parent->interfaces, node) if (class_intf->remove) - class_intf->remove(class_dev); + class_intf->remove(class_dev, class_intf); up(&parent->sem); } @@ -729,7 +729,7 @@ int class_interface_register(struct class_interface *class_intf) list_add_tail(&class_intf->node, &parent->interfaces); if (class_intf->add) { list_for_each_entry(class_dev, &parent->children, node) - class_intf->add(class_dev); + class_intf->add(class_dev, class_intf); } up(&parent->sem); @@ -748,7 +748,7 @@ void class_interface_unregister(struct class_interface *class_intf) list_del_init(&class_intf->node); if (class_intf->remove) { list_for_each_entry(class_dev, &parent->children, node) - class_intf->remove(class_dev); + class_intf->remove(class_dev, class_intf); } up(&parent->sem); diff --git a/drivers/pcmcia/ds.c b/drivers/pcmcia/ds.c index 080608c7381a..39d096b52926 100644 --- a/drivers/pcmcia/ds.c +++ b/drivers/pcmcia/ds.c @@ -1157,7 +1157,8 @@ static struct pcmcia_callback pcmcia_bus_callback = { .requery = pcmcia_bus_rescan, }; -static int __devinit pcmcia_bus_add_socket(struct class_device *class_dev) +static int __devinit pcmcia_bus_add_socket(struct class_device *class_dev, + struct class_interface *class_intf) { struct pcmcia_socket *socket = class_get_devdata(class_dev); int ret; @@ -1192,7 +1193,8 @@ static int __devinit pcmcia_bus_add_socket(struct class_device *class_dev) return 0; } -static void pcmcia_bus_remove_socket(struct class_device *class_dev) +static void pcmcia_bus_remove_socket(struct class_device *class_dev, + struct class_interface *class_intf) { struct pcmcia_socket *socket = class_get_devdata(class_dev); diff --git a/drivers/pcmcia/rsrc_nonstatic.c b/drivers/pcmcia/rsrc_nonstatic.c index f9a5c70284b5..fc87e7e2b6b8 100644 --- a/drivers/pcmcia/rsrc_nonstatic.c +++ b/drivers/pcmcia/rsrc_nonstatic.c @@ -994,7 +994,8 @@ static struct class_device_attribute *pccard_rsrc_attributes[] = { NULL, }; -static int __devinit pccard_sysfs_add_rsrc(struct class_device *class_dev) +static int __devinit pccard_sysfs_add_rsrc(struct class_device *class_dev, + struct class_interface *class_intf) { struct pcmcia_socket *s = class_get_devdata(class_dev); struct class_device_attribute **attr; @@ -1011,7 +1012,8 @@ static int __devinit pccard_sysfs_add_rsrc(struct class_device *class_dev) return ret; } -static void __devexit pccard_sysfs_remove_rsrc(struct class_device *class_dev) +static void __devexit pccard_sysfs_remove_rsrc(struct class_device *class_dev, + struct class_interface *class_intf) { struct pcmcia_socket *s = class_get_devdata(class_dev); struct class_device_attribute **attr; diff --git a/drivers/pcmcia/socket_sysfs.c b/drivers/pcmcia/socket_sysfs.c index 1040a6c1a8a4..4a3150a7854c 100644 --- a/drivers/pcmcia/socket_sysfs.c +++ b/drivers/pcmcia/socket_sysfs.c @@ -341,7 +341,8 @@ static struct bin_attribute pccard_cis_attr = { .write = pccard_store_cis, }; -static int __devinit pccard_sysfs_add_socket(struct class_device *class_dev) +static int __devinit pccard_sysfs_add_socket(struct class_device *class_dev, + struct class_interface *class_intf) { struct class_device_attribute **attr; int ret = 0; @@ -357,7 +358,8 @@ static int __devinit pccard_sysfs_add_socket(struct class_device *class_dev) return ret; } -static void __devexit pccard_sysfs_remove_socket(struct class_device *class_dev) +static void __devexit pccard_sysfs_remove_socket(struct class_device *class_dev, + struct class_interface *class_intf) { struct class_device_attribute **attr; diff --git a/drivers/scsi/sg.c b/drivers/scsi/sg.c index ad94367df430..f0d8b4eda5ad 100644 --- a/drivers/scsi/sg.c +++ b/drivers/scsi/sg.c @@ -104,8 +104,8 @@ static int sg_allow_dio = SG_ALLOW_DIO_DEF; #define SG_DEV_ARR_LUMP 32 /* amount to over allocate sg_dev_arr by */ -static int sg_add(struct class_device *); -static void sg_remove(struct class_device *); +static int sg_add(struct class_device *, struct class_interface *); +static void sg_remove(struct class_device *, struct class_interface *); static Scsi_Request *dummy_cmdp; /* only used for sizeof */ @@ -1506,7 +1506,7 @@ static int sg_alloc(struct gendisk *disk, struct scsi_device *scsidp) } static int -sg_add(struct class_device *cl_dev) +sg_add(struct class_device *cl_dev, struct class_interface *cl_intf) { struct scsi_device *scsidp = to_scsi_device(cl_dev->dev); struct gendisk *disk; @@ -1582,7 +1582,7 @@ out: } static void -sg_remove(struct class_device *cl_dev) +sg_remove(struct class_device *cl_dev, struct class_interface *cl_intf) { struct scsi_device *scsidp = to_scsi_device(cl_dev->dev); Sg_device *sdp = NULL; diff --git a/include/linux/device.h b/include/linux/device.h index 95d607a48f06..a53a822c4d16 100644 --- a/include/linux/device.h +++ b/include/linux/device.h @@ -251,8 +251,8 @@ struct class_interface { struct list_head node; struct class *class; - int (*add) (struct class_device *); - void (*remove) (struct class_device *); + int (*add) (struct class_device *, struct class_interface *); + void (*remove) (struct class_device *, struct class_interface *); }; extern int class_interface_register(struct class_interface *); -- cgit v1.2.3 From a7fd67062efc5b0fc9a61368c607fa92d1d57f9e Mon Sep 17 00:00:00 2001 From: Kay Sievers Date: Sat, 1 Oct 2005 14:49:43 +0200 Subject: [PATCH] add sysfs attr to re-emit device hotplug event A "coldplug + udevstart" can be simple like this: for i in /sys/block/*/*/uevent; do echo 1 > $i; done for i in /sys/class/*/*/uevent; do echo 1 > $i; done for i in /sys/bus/*/devices/*/uevent; do echo 1 > $i; done Signed-off-by: Kay Sievers Signed-off-by: Greg Kroah-Hartman --- drivers/base/class.c | 16 ++++++++++++-- drivers/base/core.c | 16 ++++++++++++++ drivers/block/genhd.c | 25 ++++++++++++++++++++++ fs/partitions/check.c | 27 +++++++++++++++++++++++- include/linux/device.h | 57 +++++++++++++++++++++++++------------------------- include/linux/genhd.h | 1 + 6 files changed, 110 insertions(+), 32 deletions(-) (limited to 'include/linux') diff --git a/drivers/base/class.c b/drivers/base/class.c index 73d44cf53db0..3cf6eb36f3d8 100644 --- a/drivers/base/class.c +++ b/drivers/base/class.c @@ -442,6 +442,13 @@ static ssize_t show_dev(struct class_device *class_dev, char *buf) return print_dev_t(buf, class_dev->devt); } +static ssize_t store_uevent(struct class_device *class_dev, + const char *buf, size_t count) +{ + kobject_hotplug(&class_dev->kobj, KOBJ_ADD); + return count; +} + void class_device_initialize(struct class_device *class_dev) { kobj_set_kset_s(class_dev, class_obj_subsys); @@ -497,6 +504,12 @@ int class_device_add(struct class_device *class_dev) goto register_done; /* add the needed attributes to this device */ + class_dev->uevent_attr.attr.name = "uevent"; + class_dev->uevent_attr.attr.mode = S_IWUSR; + class_dev->uevent_attr.attr.owner = parent->owner; + class_dev->uevent_attr.store = store_uevent; + class_device_create_file(class_dev, &class_dev->uevent_attr); + if (MAJOR(class_dev->devt)) { struct class_device_attribute *attr; attr = kzalloc(sizeof(*attr), GFP_KERNEL); @@ -505,12 +518,10 @@ int class_device_add(struct class_device *class_dev) kobject_del(&class_dev->kobj); goto register_done; } - attr->attr.name = "dev"; attr->attr.mode = S_IRUGO; attr->attr.owner = parent->owner; attr->show = show_dev; - attr->store = NULL; class_device_create_file(class_dev, attr); class_dev->devt_attr = attr; } @@ -621,6 +632,7 @@ void class_device_del(struct class_device *class_dev) sysfs_remove_link(&class_dev->kobj, "device"); sysfs_remove_link(&class_dev->dev->kobj, class_name); } + class_device_remove_file(class_dev, &class_dev->uevent_attr); if (class_dev->devt_attr) class_device_remove_file(class_dev, class_dev->devt_attr); class_device_remove_attrs(class_dev); diff --git a/drivers/base/core.c b/drivers/base/core.c index 31109193e2c4..ac4b5fdd95f5 100644 --- a/drivers/base/core.c +++ b/drivers/base/core.c @@ -154,6 +154,13 @@ static struct kset_hotplug_ops device_hotplug_ops = { .hotplug = dev_hotplug, }; +static ssize_t store_uevent(struct device *dev, struct device_attribute *attr, + const char *buf, size_t count) +{ + kobject_hotplug(&dev->kobj, KOBJ_ADD); + return count; +} + /** * device_subsys - structure to be registered with kobject core. */ @@ -259,6 +266,14 @@ int device_add(struct device *dev) if ((error = kobject_add(&dev->kobj))) goto Error; + + dev->uevent_attr.attr.name = "uevent"; + dev->uevent_attr.attr.mode = S_IWUSR; + if (dev->driver) + dev->uevent_attr.attr.owner = dev->driver->owner; + dev->uevent_attr.store = store_uevent; + device_create_file(dev, &dev->uevent_attr); + kobject_hotplug(&dev->kobj, KOBJ_ADD); if ((error = device_pm_add(dev))) goto PMError; @@ -350,6 +365,7 @@ void device_del(struct device * dev) if (parent) klist_del(&dev->knode_parent); + device_remove_file(dev, &dev->uevent_attr); /* Notify the platform of the removal, in case they * need to do anything... diff --git a/drivers/block/genhd.c b/drivers/block/genhd.c index d42840cc0d1d..486ce1fdeb8c 100644 --- a/drivers/block/genhd.c +++ b/drivers/block/genhd.c @@ -337,10 +337,30 @@ static ssize_t disk_attr_show(struct kobject *kobj, struct attribute *attr, return ret; } +static ssize_t disk_attr_store(struct kobject * kobj, struct attribute * attr, + const char *page, size_t count) +{ + struct gendisk *disk = to_disk(kobj); + struct disk_attribute *disk_attr = + container_of(attr,struct disk_attribute,attr); + ssize_t ret = 0; + + if (disk_attr->store) + ret = disk_attr->store(disk, page, count); + return ret; +} + static struct sysfs_ops disk_sysfs_ops = { .show = &disk_attr_show, + .store = &disk_attr_store, }; +static ssize_t disk_uevent_store(struct gendisk * disk, + const char *buf, size_t count) +{ + kobject_hotplug(&disk->kobj, KOBJ_ADD); + return count; +} static ssize_t disk_dev_read(struct gendisk * disk, char *page) { dev_t base = MKDEV(disk->major, disk->first_minor); @@ -382,6 +402,10 @@ static ssize_t disk_stats_read(struct gendisk * disk, char *page) jiffies_to_msecs(disk_stat_read(disk, io_ticks)), jiffies_to_msecs(disk_stat_read(disk, time_in_queue))); } +static struct disk_attribute disk_attr_uevent = { + .attr = {.name = "uevent", .mode = S_IWUSR }, + .store = disk_uevent_store +}; static struct disk_attribute disk_attr_dev = { .attr = {.name = "dev", .mode = S_IRUGO }, .show = disk_dev_read @@ -404,6 +428,7 @@ static struct disk_attribute disk_attr_stat = { }; static struct attribute * default_attrs[] = { + &disk_attr_uevent.attr, &disk_attr_dev.attr, &disk_attr_range.attr, &disk_attr_removable.attr, diff --git a/fs/partitions/check.c b/fs/partitions/check.c index 77e178f13162..d95a110293fa 100644 --- a/fs/partitions/check.c +++ b/fs/partitions/check.c @@ -192,6 +192,7 @@ check_partition(struct gendisk *hd, struct block_device *bdev) struct part_attribute { struct attribute attr; ssize_t (*show)(struct hd_struct *,char *); + ssize_t (*store)(struct hd_struct *,const char *, size_t); }; static ssize_t @@ -201,14 +202,33 @@ part_attr_show(struct kobject * kobj, struct attribute * attr, char * page) struct part_attribute * part_attr = container_of(attr,struct part_attribute,attr); ssize_t ret = 0; if (part_attr->show) - ret = part_attr->show(p,page); + ret = part_attr->show(p, page); + return ret; +} +static ssize_t +part_attr_store(struct kobject * kobj, struct attribute * attr, + const char *page, size_t count) +{ + struct hd_struct * p = container_of(kobj,struct hd_struct,kobj); + struct part_attribute * part_attr = container_of(attr,struct part_attribute,attr); + ssize_t ret = 0; + + if (part_attr->store) + ret = part_attr->store(p, page, count); return ret; } static struct sysfs_ops part_sysfs_ops = { .show = part_attr_show, + .store = part_attr_store, }; +static ssize_t part_uevent_store(struct hd_struct * p, + const char *page, size_t count) +{ + kobject_hotplug(&p->kobj, KOBJ_ADD); + return count; +} static ssize_t part_dev_read(struct hd_struct * p, char *page) { struct gendisk *disk = container_of(p->kobj.parent,struct gendisk,kobj); @@ -229,6 +249,10 @@ static ssize_t part_stat_read(struct hd_struct * p, char *page) p->reads, (unsigned long long)p->read_sectors, p->writes, (unsigned long long)p->write_sectors); } +static struct part_attribute part_attr_uevent = { + .attr = {.name = "uevent", .mode = S_IWUSR }, + .store = part_uevent_store +}; static struct part_attribute part_attr_dev = { .attr = {.name = "dev", .mode = S_IRUGO }, .show = part_dev_read @@ -247,6 +271,7 @@ static struct part_attribute part_attr_stat = { }; static struct attribute * default_attrs[] = { + &part_attr_uevent.attr, &part_attr_dev.attr, &part_attr_start.attr, &part_attr_size.attr, diff --git a/include/linux/device.h b/include/linux/device.h index a53a822c4d16..e86a580b72e1 100644 --- a/include/linux/device.h +++ b/include/linux/device.h @@ -190,6 +190,18 @@ struct class_attribute class_attr_##_name = __ATTR(_name,_mode,_show,_store) extern int class_create_file(struct class *, const struct class_attribute *); extern void class_remove_file(struct class *, const struct class_attribute *); +struct class_device_attribute { + struct attribute attr; + ssize_t (*show)(struct class_device *, char * buf); + ssize_t (*store)(struct class_device *, const char * buf, size_t count); +}; + +#define CLASS_DEVICE_ATTR(_name,_mode,_show,_store) \ +struct class_device_attribute class_device_attr_##_name = \ + __ATTR(_name,_mode,_show,_store) + +extern int class_device_create_file(struct class_device *, + const struct class_device_attribute *); struct class_device { struct list_head node; @@ -198,6 +210,7 @@ struct class_device { struct class * class; /* required */ dev_t devt; /* dev_t, creates the sysfs "dev" */ struct class_device_attribute *devt_attr; + struct class_device_attribute uevent_attr; struct device * dev; /* not necessary, but nice to have */ void * class_data; /* class-specific data */ @@ -228,18 +241,6 @@ extern int class_device_rename(struct class_device *, char *); extern struct class_device * class_device_get(struct class_device *); extern void class_device_put(struct class_device *); -struct class_device_attribute { - struct attribute attr; - ssize_t (*show)(struct class_device *, char * buf); - ssize_t (*store)(struct class_device *, const char * buf, size_t count); -}; - -#define CLASS_DEVICE_ATTR(_name,_mode,_show,_store) \ -struct class_device_attribute class_device_attr_##_name = \ - __ATTR(_name,_mode,_show,_store) - -extern int class_device_create_file(struct class_device *, - const struct class_device_attribute *); extern void class_device_remove_file(struct class_device *, const struct class_device_attribute *); extern int class_device_create_bin_file(struct class_device *, @@ -266,6 +267,20 @@ extern struct class_device *class_device_create(struct class *cls, dev_t devt, extern void class_device_destroy(struct class *cls, dev_t devt); +/* interface for exporting device attributes */ +struct device_attribute { + struct attribute attr; + ssize_t (*show)(struct device *dev, struct device_attribute *attr, + char *buf); + ssize_t (*store)(struct device *dev, struct device_attribute *attr, + const char *buf, size_t count); +}; + +#define DEVICE_ATTR(_name,_mode,_show,_store) \ +struct device_attribute dev_attr_##_name = __ATTR(_name,_mode,_show,_store) + +extern int device_create_file(struct device *device, struct device_attribute * entry); +extern void device_remove_file(struct device * dev, struct device_attribute * attr); struct device { struct klist klist_children; struct klist_node knode_parent; /* node in sibling list */ @@ -275,6 +290,7 @@ struct device { struct kobject kobj; char bus_id[BUS_ID_SIZE]; /* position on parent bus */ + struct device_attribute uevent_attr; struct semaphore sem; /* semaphore to synchronize calls to * its driver. @@ -343,23 +359,6 @@ extern int device_attach(struct device * dev); extern void driver_attach(struct device_driver * drv); -/* driverfs interface for exporting device attributes */ - -struct device_attribute { - struct attribute attr; - ssize_t (*show)(struct device *dev, struct device_attribute *attr, - char *buf); - ssize_t (*store)(struct device *dev, struct device_attribute *attr, - const char *buf, size_t count); -}; - -#define DEVICE_ATTR(_name,_mode,_show,_store) \ -struct device_attribute dev_attr_##_name = __ATTR(_name,_mode,_show,_store) - - -extern int device_create_file(struct device *device, struct device_attribute * entry); -extern void device_remove_file(struct device * dev, struct device_attribute * attr); - /* * Platform "fixup" functions - allow the platform to have their say * about devices and actions that the general device layer doesn't diff --git a/include/linux/genhd.h b/include/linux/genhd.h index 01796c41c951..78af34840c69 100644 --- a/include/linux/genhd.h +++ b/include/linux/genhd.h @@ -132,6 +132,7 @@ struct gendisk { struct disk_attribute { struct attribute attr; ssize_t (*show)(struct gendisk *, char *); + ssize_t (*store)(struct gendisk *, const char *, size_t); }; /* -- cgit v1.2.3 From 51d172d5f3a193e4b8f76179b2e55d7a36b94117 Mon Sep 17 00:00:00 2001 From: Greg Kroah-Hartman Date: Thu, 27 Oct 2005 22:25:43 -0700 Subject: [PATCH] Driver Core: add the ability for class_device structures to be nested This patch allows struct class_device to be nested, so that another struct class_device can be the parent of a new one, instead of only having the struct class be the parent. This will allow us to (hopefully) fix up the input and video class subsystem mess. But please people, don't go crazy and start making huge trees of class devices, you should only need 2 levels deep to get everything to work (remember to use a class_interface to get notification of a new class device being added to the system.) Oh, this also allows us to have the possibility of potentially, someday, moving /sys/block into /sys/class. The main hindrance is that pesky /dev numberspace issue... Signed-off-by: Greg Kroah-Hartman --- drivers/base/class.c | 125 ++++++++++++++++++++++++++++++++----------------- include/linux/device.h | 13 +++-- 2 files changed, 91 insertions(+), 47 deletions(-) (limited to 'include/linux') diff --git a/drivers/base/class.c b/drivers/base/class.c index 3cf6eb36f3d8..c3e569730afe 100644 --- a/drivers/base/class.c +++ b/drivers/base/class.c @@ -99,7 +99,8 @@ struct class * class_get(struct class * cls) void class_put(struct class * cls) { - subsys_put(&cls->subsys); + if (cls) + subsys_put(&cls->subsys); } @@ -165,14 +166,25 @@ void class_unregister(struct class * cls) static void class_create_release(struct class *cls) { + pr_debug("%s called for %s\n", __FUNCTION__, cls->name); kfree(cls); } static void class_device_create_release(struct class_device *class_dev) { + pr_debug("%s called for %s\n", __FUNCTION__, class_dev->class_id); kfree(class_dev); } +/* needed to allow these devices to have parent class devices */ +static int class_device_create_hotplug(struct class_device *class_dev, + char **envp, int num_envp, + char *buffer, int buffer_size) +{ + pr_debug("%s called for %s\n", __FUNCTION__, class_dev->class_id); + return 0; +} + /** * class_create - create a struct class structure * @owner: pointer to the module that is to "own" this struct class @@ -301,10 +313,12 @@ static void class_dev_release(struct kobject * kobj) kfree(cd->devt_attr); cd->devt_attr = NULL; - if (cls->release) + if (cd->release) + cd->release(cd); + else if (cls->release) cls->release(cd); else { - printk(KERN_ERR "Device class '%s' does not have a release() function, " + printk(KERN_ERR "Class Device '%s' does not have a release() function, " "it is broken and must be fixed.\n", cd->class_id); WARN_ON(1); @@ -382,14 +396,18 @@ static int class_hotplug(struct kset *kset, struct kobject *kobj, char **envp, buffer = &buffer[length]; buffer_size -= length; - if (class_dev->class->hotplug) { - /* have the bus specific function add its stuff */ - retval = class_dev->class->hotplug (class_dev, envp, num_envp, - buffer, buffer_size); - if (retval) { - pr_debug ("%s - hotplug() returned %d\n", - __FUNCTION__, retval); - } + if (class_dev->hotplug) { + /* have the class device specific function add its stuff */ + retval = class_dev->hotplug(class_dev, envp, num_envp, + buffer, buffer_size); + if (retval) + pr_debug("class_dev->hotplug() returned %d\n", retval); + } else if (class_dev->class->hotplug) { + /* have the class specific function add its stuff */ + retval = class_dev->class->hotplug(class_dev, envp, num_envp, + buffer, buffer_size); + if (retval) + pr_debug("class->hotplug() returned %d\n", retval); } return retval; @@ -476,37 +494,42 @@ static char *make_class_name(struct class_device *class_dev) int class_device_add(struct class_device *class_dev) { - struct class * parent = NULL; - struct class_interface * class_intf; + struct class *parent_class = NULL; + struct class_device *parent_class_dev = NULL; + struct class_interface *class_intf; char *class_name = NULL; - int error; + int error = -EINVAL; class_dev = class_device_get(class_dev); if (!class_dev) return -EINVAL; - if (!strlen(class_dev->class_id)) { - error = -EINVAL; + if (!strlen(class_dev->class_id)) goto register_done; - } - parent = class_get(class_dev->class); + parent_class = class_get(class_dev->class); + if (!parent_class) + goto register_done; + parent_class_dev = class_device_get(class_dev->parent); pr_debug("CLASS: registering class device: ID = '%s'\n", class_dev->class_id); /* first, register with generic layer. */ kobject_set_name(&class_dev->kobj, "%s", class_dev->class_id); - if (parent) - class_dev->kobj.parent = &parent->subsys.kset.kobj; + if (parent_class_dev) + class_dev->kobj.parent = &parent_class_dev->kobj; + else + class_dev->kobj.parent = &parent_class->subsys.kset.kobj; - if ((error = kobject_add(&class_dev->kobj))) + error = kobject_add(&class_dev->kobj); + if (error) goto register_done; /* add the needed attributes to this device */ class_dev->uevent_attr.attr.name = "uevent"; class_dev->uevent_attr.attr.mode = S_IWUSR; - class_dev->uevent_attr.attr.owner = parent->owner; + class_dev->uevent_attr.attr.owner = parent_class->owner; class_dev->uevent_attr.store = store_uevent; class_device_create_file(class_dev, &class_dev->uevent_attr); @@ -520,7 +543,7 @@ int class_device_add(struct class_device *class_dev) } attr->attr.name = "dev"; attr->attr.mode = S_IRUGO; - attr->attr.owner = parent->owner; + attr->attr.owner = parent_class->owner; attr->show = show_dev; class_device_create_file(class_dev, attr); class_dev->devt_attr = attr; @@ -538,18 +561,20 @@ int class_device_add(struct class_device *class_dev) kobject_hotplug(&class_dev->kobj, KOBJ_ADD); /* notify any interfaces this device is now here */ - if (parent) { - down(&parent->sem); - list_add_tail(&class_dev->node, &parent->children); - list_for_each_entry(class_intf, &parent->interfaces, node) + if (parent_class) { + down(&parent_class->sem); + list_add_tail(&class_dev->node, &parent_class->children); + list_for_each_entry(class_intf, &parent_class->interfaces, node) if (class_intf->add) class_intf->add(class_dev, class_intf); - up(&parent->sem); + up(&parent_class->sem); } register_done: - if (error && parent) - class_put(parent); + if (error) { + class_put(parent_class); + class_device_put(parent_class_dev); + } class_device_put(class_dev); kfree(class_name); return error; @@ -564,21 +589,28 @@ int class_device_register(struct class_device *class_dev) /** * class_device_create - creates a class device and registers it with sysfs * @cs: pointer to the struct class that this device should be registered to. + * @parent: pointer to the parent struct class_device of this new device, if any. * @dev: the dev_t for the char device to be added. * @device: a pointer to a struct device that is assiociated with this class device. * @fmt: string for the class device's name * * This function can be used by char device classes. A struct * class_device will be created in sysfs, registered to the specified - * class. A "dev" file will be created, showing the dev_t for the - * device. The pointer to the struct class_device will be returned from - * the call. Any further sysfs files that might be required can be - * created using this pointer. + * class. + * A "dev" file will be created, showing the dev_t for the device, if + * the dev_t is not 0,0. + * If a pointer to a parent struct class_device is passed in, the newly + * created struct class_device will be a child of that device in sysfs. + * The pointer to the struct class_device will be returned from the + * call. Any further sysfs files that might be required can be created + * using this pointer. * * Note: the struct class passed to this function must have previously * been created with a call to class_create(). */ -struct class_device *class_device_create(struct class *cls, dev_t devt, +struct class_device *class_device_create(struct class *cls, + struct class_device *parent, + dev_t devt, struct device *device, char *fmt, ...) { va_list args; @@ -597,6 +629,9 @@ struct class_device *class_device_create(struct class *cls, dev_t devt, class_dev->devt = devt; class_dev->dev = device; class_dev->class = cls; + class_dev->parent = parent; + class_dev->release = class_device_create_release; + class_dev->hotplug = class_device_create_hotplug; va_start(args, fmt); vsnprintf(class_dev->class_id, BUS_ID_SIZE, fmt, args); @@ -614,17 +649,18 @@ error: void class_device_del(struct class_device *class_dev) { - struct class * parent = class_dev->class; - struct class_interface * class_intf; + struct class *parent_class = class_dev->class; + struct class_device *parent_device = class_dev->parent; + struct class_interface *class_intf; char *class_name = NULL; - if (parent) { - down(&parent->sem); + if (parent_class) { + down(&parent_class->sem); list_del_init(&class_dev->node); - list_for_each_entry(class_intf, &parent->interfaces, node) + list_for_each_entry(class_intf, &parent_class->interfaces, node) if (class_intf->remove) class_intf->remove(class_dev, class_intf); - up(&parent->sem); + up(&parent_class->sem); } if (class_dev->dev) { @@ -640,8 +676,8 @@ void class_device_del(struct class_device *class_dev) kobject_hotplug(&class_dev->kobj, KOBJ_REMOVE); kobject_del(&class_dev->kobj); - if (parent) - class_put(parent); + class_device_put(parent_device); + class_put(parent_class); kfree(class_name); } @@ -721,7 +757,8 @@ struct class_device * class_device_get(struct class_device *class_dev) void class_device_put(struct class_device *class_dev) { - kobject_put(&class_dev->kobj); + if (class_dev) + kobject_put(&class_dev->kobj); } diff --git a/include/linux/device.h b/include/linux/device.h index e86a580b72e1..226e550ae2ea 100644 --- a/include/linux/device.h +++ b/include/linux/device.h @@ -213,7 +213,11 @@ struct class_device { struct class_device_attribute uevent_attr; struct device * dev; /* not necessary, but nice to have */ void * class_data; /* class-specific data */ + struct class_device *parent; /* parent of this child device, if there is one */ + void (*release)(struct class_device *dev); + int (*hotplug)(struct class_device *dev, char **envp, + int num_envp, char *buffer, int buffer_size); char class_id[BUS_ID_SIZE]; /* unique to this class */ }; @@ -261,9 +265,12 @@ extern void class_interface_unregister(struct class_interface *); extern struct class *class_create(struct module *owner, char *name); extern void class_destroy(struct class *cls); -extern struct class_device *class_device_create(struct class *cls, dev_t devt, - struct device *device, char *fmt, ...) - __attribute__((format(printf,4,5))); +extern struct class_device *class_device_create(struct class *cls, + struct class_device *parent, + dev_t devt, + struct device *device, + char *fmt, ...) + __attribute__((format(printf,5,6))); extern void class_device_destroy(struct class *cls, dev_t devt); -- cgit v1.2.3 From 74be227f728ed68bfc270153665b43fc1f0fa845 Mon Sep 17 00:00:00 2001 From: Greg Kroah-Hartman Date: Thu, 27 Oct 2005 22:25:43 -0700 Subject: [PATCH] Driver Core: document struct class_device properly Signed-off-by: Greg Kroah-Hartman --- include/linux/device.h | 24 ++++++++++++++++++++++++ 1 file changed, 24 insertions(+) (limited to 'include/linux') diff --git a/include/linux/device.h b/include/linux/device.h index 226e550ae2ea..10ab7807f8ea 100644 --- a/include/linux/device.h +++ b/include/linux/device.h @@ -203,6 +203,30 @@ struct class_device_attribute class_device_attr_##_name = \ extern int class_device_create_file(struct class_device *, const struct class_device_attribute *); +/** + * struct class_device - class devices + * @class: pointer to the parent class for this class device. This is required. + * @devt: for internal use by the driver core only. + * @node: for internal use by the driver core only. + * @kobj: for internal use by the driver core only. + * @devt_attr: for internal use by the driver core only. + * @dev: if set, a symlink to the struct device is created in the sysfs + * directory for this struct class device. + * @class_data: pointer to whatever you want to store here for this struct + * class_device. Use class_get_devdata() and class_set_devdata() to get and + * set this pointer. + * @parent: pointer to a struct class_device that is the parent of this struct + * class_device. If NULL, this class_device will show up at the root of the + * struct class in sysfs (which is probably what you want to have happen.) + * @release: pointer to a release function for this struct class_device. If + * set, this will be called instead of the class specific release function. + * Only use this if you want to override the default release function, like + * when you are nesting class_device structures. + * @hotplug: pointer to a hotplug function for this struct class_device. If + * set, this will be called instead of the class specific hotplug function. + * Only use this if you want to override the default hotplug function, like + * when you are nesting class_device structures. + */ struct class_device { struct list_head node; -- cgit v1.2.3 From d19fbe8a763634395d4bef40fc88cdb61c4a6274 Mon Sep 17 00:00:00 2001 From: Dmitry Torokhov Date: Thu, 15 Sep 2005 02:01:39 -0500 Subject: [PATCH] Input: prepare to sysfs integration Input: prepare to sysfs integration Add struct class_device to input_dev; add input_allocate_dev() to dynamically allocate input devices; dynamically allocated devices are automatically registered with sysfs. Signed-off-by: Dmitry Torokhov Signed-off-by: Greg Kroah-Hartman --- drivers/input/input.c | 77 +++++++++++++++++++++++++++++++++++++++++++++++---- include/linux/input.h | 24 +++++++++++++++- 2 files changed, 95 insertions(+), 6 deletions(-) (limited to 'include/linux') diff --git a/drivers/input/input.c b/drivers/input/input.c index 072bbf528151..0e2e890c0988 100644 --- a/drivers/input/input.c +++ b/drivers/input/input.c @@ -27,6 +27,7 @@ MODULE_AUTHOR("Vojtech Pavlik "); MODULE_DESCRIPTION("Input core"); MODULE_LICENSE("GPL"); +EXPORT_SYMBOL(input_allocate_device); EXPORT_SYMBOL(input_register_device); EXPORT_SYMBOL(input_unregister_device); EXPORT_SYMBOL(input_register_handler); @@ -605,6 +606,56 @@ static inline int input_proc_init(void) { return 0; } static inline void input_proc_exit(void) { } #endif +static void input_dev_release(struct class_device *class_dev) +{ + struct input_dev *dev = to_input_dev(class_dev); + + kfree(dev); + module_put(THIS_MODULE); +} + +static struct class input_dev_class = { + .name = "input_dev", + .release = input_dev_release, +}; + +struct input_dev *input_allocate_device(void) +{ + struct input_dev *dev; + + dev = kzalloc(sizeof(struct input_dev), GFP_KERNEL); + if (dev) { + dev->dynalloc = 1; + dev->cdev.class = &input_dev_class; + class_device_initialize(&dev->cdev); + INIT_LIST_HEAD(&dev->h_list); + INIT_LIST_HEAD(&dev->node); + } + + return dev; +} + +static void input_register_classdevice(struct input_dev *dev) +{ + static atomic_t input_no = ATOMIC_INIT(0); + const char *path; + + __module_get(THIS_MODULE); + + dev->dev = dev->cdev.dev; + + snprintf(dev->cdev.class_id, sizeof(dev->cdev.class_id), + "input%ld", (unsigned long) atomic_inc_return(&input_no) - 1); + + path = kobject_get_path(&dev->cdev.class->subsys.kset.kobj, GFP_KERNEL); + printk(KERN_INFO "input: %s/%s as %s\n", + dev->name ? dev->name : "Unspecified device", + path ? path : "", dev->cdev.class_id); + kfree(path); + + class_device_add(&dev->cdev); +} + void input_register_device(struct input_dev *dev) { struct input_handle *handle; @@ -637,6 +688,10 @@ void input_register_device(struct input_dev *dev) if ((handle = handler->connect(handler, dev, id))) input_link_handle(handle); + + if (dev->dynalloc) + input_register_classdevice(dev); + #ifdef CONFIG_HOTPLUG input_call_hotplug("add", dev); #endif @@ -665,6 +720,9 @@ void input_unregister_device(struct input_dev *dev) list_del_init(&dev->node); + if (dev->dynalloc) + class_device_unregister(&dev->cdev); + input_wakeup_procfs_readers(); } @@ -753,26 +811,34 @@ static int __init input_init(void) { int err; + err = class_register(&input_dev_class); + if (err) { + printk(KERN_ERR "input: unable to register input_dev class\n"); + return err; + } + input_class = class_create(THIS_MODULE, "input"); if (IS_ERR(input_class)) { printk(KERN_ERR "input: unable to register input class\n"); - return PTR_ERR(input_class); + err = PTR_ERR(input_class); + goto fail1; } err = input_proc_init(); if (err) - goto fail1; + goto fail2; err = register_chrdev(INPUT_MAJOR, "input", &input_fops); if (err) { printk(KERN_ERR "input: unable to register char major %d", INPUT_MAJOR); - goto fail2; + goto fail3; } return 0; - fail2: input_proc_exit(); - fail1: class_destroy(input_class); + fail3: input_proc_exit(); + fail2: class_destroy(input_class); + fail1: class_unregister(&input_dev_class); return err; } @@ -781,6 +847,7 @@ static void __exit input_exit(void) input_proc_exit(); unregister_chrdev(INPUT_MAJOR, "input"); class_destroy(input_class); + class_unregister(&input_dev_class); } subsys_initcall(input_init); diff --git a/include/linux/input.h b/include/linux/input.h index e8c296ff6257..3defa29a17d3 100644 --- a/include/linux/input.h +++ b/include/linux/input.h @@ -12,6 +12,7 @@ #ifdef __KERNEL__ #include #include +#include #else #include #include @@ -889,11 +890,15 @@ struct input_dev { struct semaphore sem; /* serializes open and close operations */ unsigned int users; - struct device *dev; + struct class_device cdev; + struct device *dev; /* will be removed soon */ + + int dynalloc; /* temporarily */ struct list_head h_list; struct list_head node; }; +#define to_input_dev(d) container_of(d, struct input_dev, cdev) /* * Structure for hotplug & device<->driver matching. @@ -984,6 +989,23 @@ static inline void init_input_dev(struct input_dev *dev) INIT_LIST_HEAD(&dev->node); } +struct input_dev *input_allocate_device(void); + +static inline void input_free_device(struct input_dev *dev) +{ + kfree(dev); +} + +static inline struct input_dev *input_get_device(struct input_dev *dev) +{ + return to_input_dev(class_device_get(&dev->cdev)); +} + +static inline void input_put_device(struct input_dev *dev) +{ + class_device_put(&dev->cdev); +} + void input_register_device(struct input_dev *); void input_unregister_device(struct input_dev *); -- cgit v1.2.3 From 23d50901617c2a8bdef509279a42d2e90f523db9 Mon Sep 17 00:00:00 2001 From: Greg Kroah-Hartman Date: Thu, 27 Oct 2005 22:25:43 -0700 Subject: [PATCH] INPUT: export input_dev_class so that input drivers can use it. Signed-off-by: Greg Kroah-Hartman --- drivers/input/input.c | 3 ++- include/linux/input.h | 1 + 2 files changed, 3 insertions(+), 1 deletion(-) (limited to 'include/linux') diff --git a/drivers/input/input.c b/drivers/input/input.c index 03c2ca404f20..b0ede4cc72b7 100644 --- a/drivers/input/input.c +++ b/drivers/input/input.c @@ -40,6 +40,7 @@ EXPORT_SYMBOL(input_accept_process); EXPORT_SYMBOL(input_flush_device); EXPORT_SYMBOL(input_event); EXPORT_SYMBOL(input_class); +EXPORT_SYMBOL_GPL(input_dev_class); #define INPUT_DEVICES 256 @@ -724,7 +725,7 @@ static void input_dev_release(struct class_device *class_dev) module_put(THIS_MODULE); } -static struct class input_dev_class = { +struct class input_dev_class = { .name = "input_dev", .release = input_dev_release, .class_dev_attrs = input_dev_attrs, diff --git a/include/linux/input.h b/include/linux/input.h index 3defa29a17d3..5de844157fa9 100644 --- a/include/linux/input.h +++ b/include/linux/input.h @@ -1075,6 +1075,7 @@ static inline void input_set_abs_params(struct input_dev *dev, int axis, int min } extern struct class *input_class; +extern struct class input_dev_class; #endif #endif -- cgit v1.2.3 From b0fdfebb205fcbf394c3db39679a766b8fc4f07d Mon Sep 17 00:00:00 2001 From: Greg Kroah-Hartman Date: Thu, 27 Oct 2005 22:25:43 -0700 Subject: [PATCH] INPUT: remove the input_class structure, as it is unused. Signed-off-by: Greg Kroah-Hartman --- drivers/input/input.c | 18 +++--------------- include/linux/input.h | 1 - 2 files changed, 3 insertions(+), 16 deletions(-) (limited to 'include/linux') diff --git a/drivers/input/input.c b/drivers/input/input.c index 0d570cf92dc2..5c9044dbf00e 100644 --- a/drivers/input/input.c +++ b/drivers/input/input.c @@ -39,7 +39,6 @@ EXPORT_SYMBOL(input_close_device); EXPORT_SYMBOL(input_accept_process); EXPORT_SYMBOL(input_flush_device); EXPORT_SYMBOL(input_event); -EXPORT_SYMBOL(input_class); EXPORT_SYMBOL_GPL(input_dev_class); #define INPUT_DEVICES 256 @@ -927,8 +926,6 @@ static struct file_operations input_fops = { .open = input_open_file, }; -struct class *input_class; - static int __init input_init(void) { int err; @@ -939,27 +936,19 @@ static int __init input_init(void) return err; } - input_class = class_create(THIS_MODULE, "input"); - if (IS_ERR(input_class)) { - printk(KERN_ERR "input: unable to register input class\n"); - err = PTR_ERR(input_class); - goto fail1; - } - err = input_proc_init(); if (err) - goto fail2; + goto fail1; err = register_chrdev(INPUT_MAJOR, "input", &input_fops); if (err) { printk(KERN_ERR "input: unable to register char major %d", INPUT_MAJOR); - goto fail3; + goto fail2; } return 0; - fail3: input_proc_exit(); - fail2: class_destroy(input_class); + fail2: input_proc_exit(); fail1: class_unregister(&input_dev_class); return err; } @@ -968,7 +957,6 @@ static void __exit input_exit(void) { input_proc_exit(); unregister_chrdev(INPUT_MAJOR, "input"); - class_destroy(input_class); class_unregister(&input_dev_class); } diff --git a/include/linux/input.h b/include/linux/input.h index 5de844157fa9..256e88755f12 100644 --- a/include/linux/input.h +++ b/include/linux/input.h @@ -1074,7 +1074,6 @@ static inline void input_set_abs_params(struct input_dev *dev, int axis, int min dev->absbit[LONG(axis)] |= BIT(axis); } -extern struct class *input_class; extern struct class input_dev_class; #endif -- cgit v1.2.3 From ea9f240bd819f9299703283e5326da606bbb4b05 Mon Sep 17 00:00:00 2001 From: Greg Kroah-Hartman Date: Thu, 27 Oct 2005 22:25:43 -0700 Subject: [PATCH] INPUT: rename input_dev_class to input_class to be correct. Signed-off-by: Greg Kroah-Hartman --- drivers/input/evdev.c | 4 ++-- drivers/input/input.c | 14 +++++++------- drivers/input/joydev.c | 4 ++-- drivers/input/mousedev.c | 8 ++++---- drivers/input/tsdev.c | 4 ++-- include/linux/input.h | 2 +- 6 files changed, 18 insertions(+), 18 deletions(-) (limited to 'include/linux') diff --git a/drivers/input/evdev.c b/drivers/input/evdev.c index 579041ded4be..2a96b260a2a8 100644 --- a/drivers/input/evdev.c +++ b/drivers/input/evdev.c @@ -686,7 +686,7 @@ static struct input_handle *evdev_connect(struct input_handler *handler, struct evdev_table[minor] = evdev; - class_device_create(&input_dev_class, &dev->cdev, + class_device_create(&input_class, &dev->cdev, MKDEV(INPUT_MAJOR, EVDEV_MINOR_BASE + minor), dev->cdev.dev, "event%d", minor); @@ -698,7 +698,7 @@ static void evdev_disconnect(struct input_handle *handle) struct evdev *evdev = handle->private; struct evdev_list *list; - class_device_destroy(&input_dev_class, + class_device_destroy(&input_class, MKDEV(INPUT_MAJOR, EVDEV_MINOR_BASE + evdev->minor)); evdev->exist = 0; diff --git a/drivers/input/input.c b/drivers/input/input.c index 5c9044dbf00e..a8f65fa7e17a 100644 --- a/drivers/input/input.c +++ b/drivers/input/input.c @@ -39,7 +39,7 @@ EXPORT_SYMBOL(input_close_device); EXPORT_SYMBOL(input_accept_process); EXPORT_SYMBOL(input_flush_device); EXPORT_SYMBOL(input_event); -EXPORT_SYMBOL_GPL(input_dev_class); +EXPORT_SYMBOL_GPL(input_class); #define INPUT_DEVICES 256 @@ -729,8 +729,8 @@ static void input_dev_release(struct class_device *class_dev) module_put(THIS_MODULE); } -struct class input_dev_class = { - .name = "input_dev", +struct class input_class = { + .name = "input", .release = input_dev_release, }; @@ -741,7 +741,7 @@ struct input_dev *input_allocate_device(void) dev = kzalloc(sizeof(struct input_dev), GFP_KERNEL); if (dev) { dev->dynalloc = 1; - dev->cdev.class = &input_dev_class; + dev->cdev.class = &input_class; class_device_initialize(&dev->cdev); INIT_LIST_HEAD(&dev->h_list); INIT_LIST_HEAD(&dev->node); @@ -930,7 +930,7 @@ static int __init input_init(void) { int err; - err = class_register(&input_dev_class); + err = class_register(&input_class); if (err) { printk(KERN_ERR "input: unable to register input_dev class\n"); return err; @@ -949,7 +949,7 @@ static int __init input_init(void) return 0; fail2: input_proc_exit(); - fail1: class_unregister(&input_dev_class); + fail1: class_unregister(&input_class); return err; } @@ -957,7 +957,7 @@ static void __exit input_exit(void) { input_proc_exit(); unregister_chrdev(INPUT_MAJOR, "input"); - class_unregister(&input_dev_class); + class_unregister(&input_class); } subsys_initcall(input_init); diff --git a/drivers/input/joydev.c b/drivers/input/joydev.c index 9c17d1add260..25f7eba4761f 100644 --- a/drivers/input/joydev.c +++ b/drivers/input/joydev.c @@ -513,7 +513,7 @@ static struct input_handle *joydev_connect(struct input_handler *handler, struct joydev_table[minor] = joydev; - class_device_create(&input_dev_class, &dev->cdev, + class_device_create(&input_class, &dev->cdev, MKDEV(INPUT_MAJOR, JOYDEV_MINOR_BASE + minor), dev->cdev.dev, "js%d", minor); @@ -525,7 +525,7 @@ static void joydev_disconnect(struct input_handle *handle) struct joydev *joydev = handle->private; struct joydev_list *list; - class_device_destroy(&input_dev_class, MKDEV(INPUT_MAJOR, JOYDEV_MINOR_BASE + joydev->minor)); + class_device_destroy(&input_class, MKDEV(INPUT_MAJOR, JOYDEV_MINOR_BASE + joydev->minor)); joydev->exist = 0; if (joydev->open) { diff --git a/drivers/input/mousedev.c b/drivers/input/mousedev.c index 5ec6291e967f..de2808fc85b3 100644 --- a/drivers/input/mousedev.c +++ b/drivers/input/mousedev.c @@ -648,7 +648,7 @@ static struct input_handle *mousedev_connect(struct input_handler *handler, stru mousedev_table[minor] = mousedev; - class_device_create(&input_dev_class, &dev->cdev, + class_device_create(&input_class, &dev->cdev, MKDEV(INPUT_MAJOR, MOUSEDEV_MINOR_BASE + minor), dev->cdev.dev, "mouse%d", minor); @@ -660,7 +660,7 @@ static void mousedev_disconnect(struct input_handle *handle) struct mousedev *mousedev = handle->private; struct mousedev_list *list; - class_device_destroy(&input_dev_class, + class_device_destroy(&input_class, MKDEV(INPUT_MAJOR, MOUSEDEV_MINOR_BASE + mousedev->minor)); mousedev->exist = 0; @@ -734,7 +734,7 @@ static int __init mousedev_init(void) mousedev_mix.exist = 1; mousedev_mix.minor = MOUSEDEV_MIX; - class_device_create(&input_dev_class, NULL, + class_device_create(&input_class, NULL, MKDEV(INPUT_MAJOR, MOUSEDEV_MINOR_BASE + MOUSEDEV_MIX), NULL, "mice"); #ifdef CONFIG_INPUT_MOUSEDEV_PSAUX @@ -753,7 +753,7 @@ static void __exit mousedev_exit(void) if (psaux_registered) misc_deregister(&psaux_mouse); #endif - class_device_destroy(&input_dev_class, + class_device_destroy(&input_class, MKDEV(INPUT_MAJOR, MOUSEDEV_MINOR_BASE + MOUSEDEV_MIX)); input_unregister_handler(&mousedev_handler); } diff --git a/drivers/input/tsdev.c b/drivers/input/tsdev.c index 0581edb9a560..75e165736655 100644 --- a/drivers/input/tsdev.c +++ b/drivers/input/tsdev.c @@ -409,7 +409,7 @@ static struct input_handle *tsdev_connect(struct input_handler *handler, tsdev_table[minor] = tsdev; - class_device_create(&input_dev_class, &dev->cdev, + class_device_create(&input_class, &dev->cdev, MKDEV(INPUT_MAJOR, TSDEV_MINOR_BASE + minor), dev->cdev.dev, "ts%d", minor); @@ -421,7 +421,7 @@ static void tsdev_disconnect(struct input_handle *handle) struct tsdev *tsdev = handle->private; struct tsdev_list *list; - class_device_destroy(&input_dev_class, + class_device_destroy(&input_class, MKDEV(INPUT_MAJOR, TSDEV_MINOR_BASE + tsdev->minor)); tsdev->exist = 0; diff --git a/include/linux/input.h b/include/linux/input.h index 256e88755f12..e3d9c08b5cbe 100644 --- a/include/linux/input.h +++ b/include/linux/input.h @@ -1074,7 +1074,7 @@ static inline void input_set_abs_params(struct input_dev *dev, int axis, int min dev->absbit[LONG(axis)] |= BIT(axis); } -extern struct class input_dev_class; +extern struct class input_class; #endif #endif -- cgit v1.2.3 From 9480e307cd88ef09ec9294c7d97ebec18e6d2221 Mon Sep 17 00:00:00 2001 From: Russell King Date: Fri, 28 Oct 2005 09:52:56 -0700 Subject: [PATCH] DRIVER MODEL: Get rid of the obsolete tri-level suspend/resume callbacks In PM v1, all devices were called at SUSPEND_DISABLE level. Then all devices were called at SUSPEND_SAVE_STATE level, and finally SUSPEND_POWER_DOWN level. However, with PM v2, to maintain compatibility for platform devices, I arranged for the PM v2 suspend/resume callbacks to call the old PM v1 suspend/resume callbacks three times with each level in order so that existing drivers continued to work. Since this is obsolete infrastructure which is no longer necessary, we can remove it. Here's an (untested) patch to do exactly that. Signed-off-by: Russell King Signed-off-by: Greg Kroah-Hartman --- Documentation/driver-model/driver.txt | 60 ++--------------------------------- arch/arm/common/locomo.c | 10 ++---- arch/arm/common/sa1111.c | 11 ++----- arch/arm/common/scoop.c | 24 +++++++------- arch/arm/mach-pxa/corgi_ssp.c | 24 +++++++------- arch/arm/mach-sa1100/neponset.c | 30 +++++++----------- drivers/base/platform.c | 20 ++++-------- drivers/char/s3c2410-rtc.c | 20 ++++++------ drivers/char/sonypi.c | 14 ++++---- drivers/char/watchdog/s3c2410_wdt.c | 30 ++++++++---------- drivers/hwmon/hdaps.c | 6 ++-- drivers/i2c/busses/i2c-s3c2410.c | 8 ++--- drivers/i2c/i2c-core.c | 4 +-- drivers/ieee1394/nodemgr.c | 4 +-- drivers/input/keyboard/corgikbd.c | 22 ++++++------- drivers/input/keyboard/spitzkbd.c | 40 +++++++++++------------ drivers/input/serio/i8042.c | 13 +++----- drivers/input/touchscreen/corgi_ts.c | 38 +++++++++++----------- drivers/media/video/msp3400.c | 8 ++--- drivers/media/video/tda9887.c | 4 +-- drivers/media/video/tuner-core.c | 4 +-- drivers/mfd/mcp-sa11x0.c | 20 ++++++------ drivers/mmc/pxamci.c | 8 ++--- drivers/mmc/wbsd.c | 4 +-- drivers/mtd/maps/sa1100-flash.c | 8 ++--- drivers/net/dm9000.c | 8 ++--- drivers/net/irda/sa1100_ir.c | 8 ++--- drivers/net/irda/smsc-ircc2.c | 12 +++---- drivers/net/phy/mdio_bus.c | 20 ++++-------- drivers/net/smc91x.c | 8 ++--- drivers/pci/pcie/portdrv_core.c | 4 +-- drivers/pcmcia/au1000_generic.c | 21 ++---------- drivers/pcmcia/hd64465_ss.c | 20 ++---------- drivers/pcmcia/i82365.c | 20 ++---------- drivers/pcmcia/m32r_cfc.c | 21 ++---------- drivers/pcmcia/m32r_pcc.c | 21 ++---------- drivers/pcmcia/omap_cf.c | 18 ++--------- drivers/pcmcia/pxa2xx_base.c | 26 ++++----------- drivers/pcmcia/sa1100_generic.c | 20 ++---------- drivers/pcmcia/tcic.c | 20 ++---------- drivers/pcmcia/vrc4171_card.c | 24 ++------------ drivers/serial/8250.c | 10 ++---- drivers/serial/imx.c | 8 ++--- drivers/serial/mpc52xx_uart.c | 8 ++--- drivers/serial/pxa.c | 8 ++--- drivers/serial/s3c2410.c | 9 +++--- drivers/serial/sa1100.c | 8 ++--- drivers/serial/vr41xx_siu.c | 10 ++---- drivers/usb/gadget/dummy_hcd.c | 22 +++---------- drivers/usb/gadget/omap_udc.c | 9 ++---- drivers/usb/gadget/pxa2xx_udc.c | 17 +++++----- drivers/usb/host/isp116x-hcd.c | 14 +++----- drivers/usb/host/ohci-omap.c | 10 ++---- drivers/usb/host/ohci-pxa27x.c | 4 +-- drivers/usb/host/sl811-hcd.c | 10 ++---- drivers/video/backlight/corgi_bl.c | 10 +++--- drivers/video/imxfb.c | 10 +++--- drivers/video/pxafb.c | 10 +++--- drivers/video/s1d13xxxfb.c | 7 ++-- drivers/video/s3c2410fb.c | 27 +++++++--------- drivers/video/sa1100fb.c | 10 +++--- drivers/video/w100fb.c | 46 +++++++++++++-------------- include/linux/device.h | 17 ++-------- sound/arm/pxa2xx-ac97.c | 8 ++--- sound/core/init.c | 14 +++----- sound/pci/ac97/ac97_bus.c | 6 ++-- 66 files changed, 325 insertions(+), 692 deletions(-) (limited to 'include/linux') diff --git a/Documentation/driver-model/driver.txt b/Documentation/driver-model/driver.txt index fabaca1ab1b0..7c26bfae4ba0 100644 --- a/Documentation/driver-model/driver.txt +++ b/Documentation/driver-model/driver.txt @@ -196,67 +196,11 @@ it into a supported low-power state. int (*suspend) (struct device * dev, pm_message_t state, u32 level); -suspend is called to put the device in a low power state. There are -several stages to successfully suspending a device, which is denoted in -the @level parameter. Breaking the suspend transition into several -stages affords the platform flexibility in performing device power -management based on the requirements of the system and the -user-defined policy. - -SUSPEND_NOTIFY notifies the device that a suspend transition is about -to happen. This happens on system power state transitions to verify -that all devices can successfully suspend. - -A driver may choose to fail on this call, which should cause the -entire suspend transition to fail. A driver should fail only if it -knows that the device will not be able to be resumed properly when the -system wakes up again. It could also fail if it somehow determines it -is in the middle of an operation too important to stop. - -SUSPEND_DISABLE tells the device to stop I/O transactions. When it -stops transactions, or what it should do with unfinished transactions -is a policy of the driver. After this call, the driver should not -accept any other I/O requests. - -SUSPEND_SAVE_STATE tells the device to save the context of the -hardware. This includes any bus-specific hardware state and -device-specific hardware state. A pointer to this saved state can be -stored in the device's saved_state field. - -SUSPEND_POWER_DOWN tells the driver to place the device in the low -power state requested. - -Whether suspend is called with a given level is a policy of the -platform. Some levels may be omitted; drivers must not assume the -reception of any level. However, all levels must be called in the -order above; i.e. notification will always come before disabling; -disabling the device will come before suspending the device. - -All calls are made with interrupts enabled, except for the -SUSPEND_POWER_DOWN level. +suspend is called to put the device in a low power state. int (*resume) (struct device * dev, u32 level); -Resume is used to bring a device back from a low power state. Like the -suspend transition, it happens in several stages. - -RESUME_POWER_ON tells the driver to set the power state to the state -before the suspend call (The device could have already been in a low -power state before the suspend call to put in a lower power state). - -RESUME_RESTORE_STATE tells the driver to restore the state saved by -the SUSPEND_SAVE_STATE suspend call. - -RESUME_ENABLE tells the driver to start accepting I/O transactions -again. Depending on driver policy, the device may already have pending -I/O requests. - -RESUME_POWER_ON is called with interrupts disabled. The other resume -levels are called with interrupts enabled. - -As with the various suspend stages, the driver must not assume that -any other resume calls have been or will be made. Each call should be -self-contained and not dependent on any external state. +Resume is used to bring a device back from a low power state. Attributes diff --git a/arch/arm/common/locomo.c b/arch/arm/common/locomo.c index e8053d16829b..5cdb4122f057 100644 --- a/arch/arm/common/locomo.c +++ b/arch/arm/common/locomo.c @@ -550,15 +550,12 @@ struct locomo_save_data { u16 LCM_SPIMD; }; -static int locomo_suspend(struct device *dev, pm_message_t state, u32 level) +static int locomo_suspend(struct device *dev, pm_message_t state) { struct locomo *lchip = dev_get_drvdata(dev); struct locomo_save_data *save; unsigned long flags; - if (level != SUSPEND_DISABLE) - return 0; - save = kmalloc(sizeof(struct locomo_save_data), GFP_KERNEL); if (!save) return -ENOMEM; @@ -597,16 +594,13 @@ static int locomo_suspend(struct device *dev, pm_message_t state, u32 level) return 0; } -static int locomo_resume(struct device *dev, u32 level) +static int locomo_resume(struct device *dev) { struct locomo *lchip = dev_get_drvdata(dev); struct locomo_save_data *save; unsigned long r; unsigned long flags; - if (level != RESUME_ENABLE) - return 0; - save = (struct locomo_save_data *) dev->power.saved_state; if (!save) return 0; diff --git a/arch/arm/common/sa1111.c b/arch/arm/common/sa1111.c index 1a47fbf9cbbc..21e2a518ad3a 100644 --- a/arch/arm/common/sa1111.c +++ b/arch/arm/common/sa1111.c @@ -801,7 +801,7 @@ struct sa1111_save_data { #ifdef CONFIG_PM -static int sa1111_suspend(struct device *dev, pm_message_t state, u32 level) +static int sa1111_suspend(struct device *dev, pm_message_t state) { struct sa1111 *sachip = dev_get_drvdata(dev); struct sa1111_save_data *save; @@ -809,9 +809,6 @@ static int sa1111_suspend(struct device *dev, pm_message_t state, u32 level) unsigned int val; void __iomem *base; - if (level != SUSPEND_DISABLE) - return 0; - save = kmalloc(sizeof(struct sa1111_save_data), GFP_KERNEL); if (!save) return -ENOMEM; @@ -856,23 +853,19 @@ static int sa1111_suspend(struct device *dev, pm_message_t state, u32 level) /* * sa1111_resume - Restore the SA1111 device state. * @dev: device to restore - * @level: resume level * * Restore the general state of the SA1111; clock control and * interrupt controller. Other parts of the SA1111 must be * restored by their respective drivers, and must be called * via LDM after this function. */ -static int sa1111_resume(struct device *dev, u32 level) +static int sa1111_resume(struct device *dev) { struct sa1111 *sachip = dev_get_drvdata(dev); struct sa1111_save_data *save; unsigned long flags, id; void __iomem *base; - if (level != RESUME_ENABLE) - return 0; - save = (struct sa1111_save_data *)dev->power.saved_state; if (!save) return 0; diff --git a/arch/arm/common/scoop.c b/arch/arm/common/scoop.c index 9e5245c702de..e8356b76d7c6 100644 --- a/arch/arm/common/scoop.c +++ b/arch/arm/common/scoop.c @@ -102,26 +102,24 @@ static void check_scoop_reg(struct scoop_dev *sdev) } #ifdef CONFIG_PM -static int scoop_suspend(struct device *dev, pm_message_t state, uint32_t level) +static int scoop_suspend(struct device *dev, pm_message_t state) { - if (level == SUSPEND_POWER_DOWN) { - struct scoop_dev *sdev = dev_get_drvdata(dev); + struct scoop_dev *sdev = dev_get_drvdata(dev); + + check_scoop_reg(sdev); + sdev->scoop_gpwr = SCOOP_REG(sdev->base, SCOOP_GPWR); + SCOOP_REG(sdev->base, SCOOP_GPWR) = (sdev->scoop_gpwr & ~sdev->suspend_clr) | sdev->suspend_set; - check_scoop_reg(sdev); - sdev->scoop_gpwr = SCOOP_REG(sdev->base, SCOOP_GPWR); - SCOOP_REG(sdev->base, SCOOP_GPWR) = (sdev->scoop_gpwr & ~sdev->suspend_clr) | sdev->suspend_set; - } return 0; } -static int scoop_resume(struct device *dev, uint32_t level) +static int scoop_resume(struct device *dev) { - if (level == RESUME_POWER_ON) { - struct scoop_dev *sdev = dev_get_drvdata(dev); + struct scoop_dev *sdev = dev_get_drvdata(dev); + + check_scoop_reg(sdev); + SCOOP_REG(sdev->base,SCOOP_GPWR) = sdev->scoop_gpwr; - check_scoop_reg(sdev); - SCOOP_REG(sdev->base,SCOOP_GPWR) = sdev->scoop_gpwr; - } return 0; } #else diff --git a/arch/arm/mach-pxa/corgi_ssp.c b/arch/arm/mach-pxa/corgi_ssp.c index 0ef428287055..136c269db0b7 100644 --- a/arch/arm/mach-pxa/corgi_ssp.c +++ b/arch/arm/mach-pxa/corgi_ssp.c @@ -222,24 +222,22 @@ static int corgi_ssp_remove(struct device *dev) return 0; } -static int corgi_ssp_suspend(struct device *dev, pm_message_t state, u32 level) +static int corgi_ssp_suspend(struct device *dev, pm_message_t state) { - if (level == SUSPEND_POWER_DOWN) { - ssp_flush(&corgi_ssp_dev); - ssp_save_state(&corgi_ssp_dev,&corgi_ssp_state); - } + ssp_flush(&corgi_ssp_dev); + ssp_save_state(&corgi_ssp_dev,&corgi_ssp_state); + return 0; } -static int corgi_ssp_resume(struct device *dev, u32 level) +static int corgi_ssp_resume(struct device *dev) { - if (level == RESUME_POWER_ON) { - GPSR(ssp_machinfo->cs_lcdcon) = GPIO_bit(ssp_machinfo->cs_lcdcon); /* High - Disable LCD Control/Timing Gen */ - GPSR(ssp_machinfo->cs_max1111) = GPIO_bit(ssp_machinfo->cs_max1111); /* High - Disable MAX1111*/ - GPSR(ssp_machinfo->cs_ads7846) = GPIO_bit(ssp_machinfo->cs_ads7846); /* High - Disable ADS7846*/ - ssp_restore_state(&corgi_ssp_dev,&corgi_ssp_state); - ssp_enable(&corgi_ssp_dev); - } + GPSR(ssp_machinfo->cs_lcdcon) = GPIO_bit(ssp_machinfo->cs_lcdcon); /* High - Disable LCD Control/Timing Gen */ + GPSR(ssp_machinfo->cs_max1111) = GPIO_bit(ssp_machinfo->cs_max1111); /* High - Disable MAX1111*/ + GPSR(ssp_machinfo->cs_ads7846) = GPIO_bit(ssp_machinfo->cs_ads7846); /* High - Disable ADS7846*/ + ssp_restore_state(&corgi_ssp_dev,&corgi_ssp_state); + ssp_enable(&corgi_ssp_dev); + return 0; } diff --git a/arch/arm/mach-sa1100/neponset.c b/arch/arm/mach-sa1100/neponset.c index fc061641b7be..7609d69cf1cc 100644 --- a/arch/arm/mach-sa1100/neponset.c +++ b/arch/arm/mach-sa1100/neponset.c @@ -178,33 +178,27 @@ static int neponset_probe(struct device *dev) /* * LDM power management. */ -static int neponset_suspend(struct device *dev, pm_message_t state, u32 level) +static int neponset_suspend(struct device *dev, pm_message_t state) { /* * Save state. */ - if (level == SUSPEND_SAVE_STATE || - level == SUSPEND_DISABLE || - level == SUSPEND_POWER_DOWN) { - if (!dev->power.saved_state) - dev->power.saved_state = kmalloc(sizeof(unsigned int), GFP_KERNEL); - if (!dev->power.saved_state) - return -ENOMEM; - - *(unsigned int *)dev->power.saved_state = NCR_0; - } + if (!dev->power.saved_state) + dev->power.saved_state = kmalloc(sizeof(unsigned int), GFP_KERNEL); + if (!dev->power.saved_state) + return -ENOMEM; + + *(unsigned int *)dev->power.saved_state = NCR_0; return 0; } -static int neponset_resume(struct device *dev, u32 level) +static int neponset_resume(struct device *dev) { - if (level == RESUME_RESTORE_STATE || level == RESUME_ENABLE) { - if (dev->power.saved_state) { - NCR_0 = *(unsigned int *)dev->power.saved_state; - kfree(dev->power.saved_state); - dev->power.saved_state = NULL; - } + if (dev->power.saved_state) { + NCR_0 = *(unsigned int *)dev->power.saved_state; + kfree(dev->power.saved_state); + dev->power.saved_state = NULL; } return 0; diff --git a/drivers/base/platform.c b/drivers/base/platform.c index a1a56ff65b76..75ce8711bca5 100644 --- a/drivers/base/platform.c +++ b/drivers/base/platform.c @@ -281,13 +281,9 @@ static int platform_suspend(struct device * dev, pm_message_t state) { int ret = 0; - if (dev->driver && dev->driver->suspend) { - ret = dev->driver->suspend(dev, state, SUSPEND_DISABLE); - if (ret == 0) - ret = dev->driver->suspend(dev, state, SUSPEND_SAVE_STATE); - if (ret == 0) - ret = dev->driver->suspend(dev, state, SUSPEND_POWER_DOWN); - } + if (dev->driver && dev->driver->suspend) + ret = dev->driver->suspend(dev, state); + return ret; } @@ -295,13 +291,9 @@ static int platform_resume(struct device * dev) { int ret = 0; - if (dev->driver && dev->driver->resume) { - ret = dev->driver->resume(dev, RESUME_POWER_ON); - if (ret == 0) - ret = dev->driver->resume(dev, RESUME_RESTORE_STATE); - if (ret == 0) - ret = dev->driver->resume(dev, RESUME_ENABLE); - } + if (dev->driver && dev->driver->resume) + ret = dev->driver->resume(dev); + return ret; } diff --git a/drivers/char/s3c2410-rtc.c b/drivers/char/s3c2410-rtc.c index e1a90d9a8756..887b8b2d7882 100644 --- a/drivers/char/s3c2410-rtc.c +++ b/drivers/char/s3c2410-rtc.c @@ -519,30 +519,28 @@ static struct timespec s3c2410_rtc_delta; static int ticnt_save; -static int s3c2410_rtc_suspend(struct device *dev, pm_message_t state, u32 level) +static int s3c2410_rtc_suspend(struct device *dev, pm_message_t state) { struct rtc_time tm; struct timespec time; time.tv_nsec = 0; - if (level == SUSPEND_POWER_DOWN) { - /* save TICNT for anyone using periodic interrupts */ + /* save TICNT for anyone using periodic interrupts */ - ticnt_save = readb(S3C2410_TICNT); + ticnt_save = readb(S3C2410_TICNT); - /* calculate time delta for suspend */ + /* calculate time delta for suspend */ - s3c2410_rtc_gettime(&tm); - rtc_tm_to_time(&tm, &time.tv_sec); - save_time_delta(&s3c2410_rtc_delta, &time); - s3c2410_rtc_enable(dev, 0); - } + s3c2410_rtc_gettime(&tm); + rtc_tm_to_time(&tm, &time.tv_sec); + save_time_delta(&s3c2410_rtc_delta, &time); + s3c2410_rtc_enable(dev, 0); return 0; } -static int s3c2410_rtc_resume(struct device *dev, u32 level) +static int s3c2410_rtc_resume(struct device *dev) { struct rtc_time tm; struct timespec time; diff --git a/drivers/char/sonypi.c b/drivers/char/sonypi.c index a4873684f22c..f86c15587238 100644 --- a/drivers/char/sonypi.c +++ b/drivers/char/sonypi.c @@ -1167,19 +1167,17 @@ static int sonypi_disable(void) #ifdef CONFIG_PM static int old_camera_power; -static int sonypi_suspend(struct device *dev, pm_message_t state, u32 level) +static int sonypi_suspend(struct device *dev, pm_message_t state) { - if (level == SUSPEND_DISABLE) { - old_camera_power = sonypi_device.camera_power; - sonypi_disable(); - } + old_camera_power = sonypi_device.camera_power; + sonypi_disable(); + return 0; } -static int sonypi_resume(struct device *dev, u32 level) +static int sonypi_resume(struct device *dev) { - if (level == RESUME_ENABLE) - sonypi_enable(old_camera_power); + sonypi_enable(old_camera_power); return 0; } #endif diff --git a/drivers/char/watchdog/s3c2410_wdt.c b/drivers/char/watchdog/s3c2410_wdt.c index 3625b2601b42..b732020acadb 100644 --- a/drivers/char/watchdog/s3c2410_wdt.c +++ b/drivers/char/watchdog/s3c2410_wdt.c @@ -464,32 +464,28 @@ static void s3c2410wdt_shutdown(struct device *dev) static unsigned long wtcon_save; static unsigned long wtdat_save; -static int s3c2410wdt_suspend(struct device *dev, pm_message_t state, u32 level) +static int s3c2410wdt_suspend(struct device *dev, pm_message_t state) { - if (level == SUSPEND_POWER_DOWN) { - /* Save watchdog state, and turn it off. */ - wtcon_save = readl(wdt_base + S3C2410_WTCON); - wtdat_save = readl(wdt_base + S3C2410_WTDAT); + /* Save watchdog state, and turn it off. */ + wtcon_save = readl(wdt_base + S3C2410_WTCON); + wtdat_save = readl(wdt_base + S3C2410_WTDAT); - /* Note that WTCNT doesn't need to be saved. */ - s3c2410wdt_stop(); - } + /* Note that WTCNT doesn't need to be saved. */ + s3c2410wdt_stop(); return 0; } -static int s3c2410wdt_resume(struct device *dev, u32 level) +static int s3c2410wdt_resume(struct device *dev) { - if (level == RESUME_POWER_ON) { - /* Restore watchdog state. */ + /* Restore watchdog state. */ - writel(wtdat_save, wdt_base + S3C2410_WTDAT); - writel(wtdat_save, wdt_base + S3C2410_WTCNT); /* Reset count */ - writel(wtcon_save, wdt_base + S3C2410_WTCON); + writel(wtdat_save, wdt_base + S3C2410_WTDAT); + writel(wtdat_save, wdt_base + S3C2410_WTCNT); /* Reset count */ + writel(wtcon_save, wdt_base + S3C2410_WTCON); - printk(KERN_INFO PFX "watchdog %sabled\n", - (wtcon_save & S3C2410_WTCON_ENABLE) ? "en" : "dis"); - } + printk(KERN_INFO PFX "watchdog %sabled\n", + (wtcon_save & S3C2410_WTCON_ENABLE) ? "en" : "dis"); return 0; } diff --git a/drivers/hwmon/hdaps.c b/drivers/hwmon/hdaps.c index 7f0107613827..0015da5668a1 100644 --- a/drivers/hwmon/hdaps.c +++ b/drivers/hwmon/hdaps.c @@ -296,11 +296,9 @@ static int hdaps_probe(struct device *dev) return 0; } -static int hdaps_resume(struct device *dev, u32 level) +static int hdaps_resume(struct device *dev) { - if (level == RESUME_ENABLE) - return hdaps_device_init(); - return 0; + return hdaps_device_init(); } static struct device_driver hdaps_driver = { diff --git a/drivers/i2c/busses/i2c-s3c2410.c b/drivers/i2c/busses/i2c-s3c2410.c index 73a092fb0e7e..69fa282df2d5 100644 --- a/drivers/i2c/busses/i2c-s3c2410.c +++ b/drivers/i2c/busses/i2c-s3c2410.c @@ -879,14 +879,12 @@ static int s3c24xx_i2c_remove(struct device *dev) } #ifdef CONFIG_PM -static int s3c24xx_i2c_resume(struct device *dev, u32 level) +static int s3c24xx_i2c_resume(struct device *dev) { struct s3c24xx_i2c *i2c = dev_get_drvdata(dev); - - if (i2c != NULL && level == RESUME_ENABLE) { - dev_dbg(dev, "resume: level %d\n", level); + + if (i2c != NULL) s3c24xx_i2c_init(i2c); - } return 0; } diff --git a/drivers/i2c/i2c-core.c b/drivers/i2c/i2c-core.c index dda472e5e8be..45aa0e54e297 100644 --- a/drivers/i2c/i2c-core.c +++ b/drivers/i2c/i2c-core.c @@ -48,7 +48,7 @@ static int i2c_bus_suspend(struct device * dev, pm_message_t state) int rc = 0; if (dev->driver && dev->driver->suspend) - rc = dev->driver->suspend(dev,state,0); + rc = dev->driver->suspend(dev, state); return rc; } @@ -57,7 +57,7 @@ static int i2c_bus_resume(struct device * dev) int rc = 0; if (dev->driver && dev->driver->resume) - rc = dev->driver->resume(dev,0); + rc = dev->driver->resume(dev); return rc; } diff --git a/drivers/ieee1394/nodemgr.c b/drivers/ieee1394/nodemgr.c index 347ece6b583c..7fff5a1d2ea4 100644 --- a/drivers/ieee1394/nodemgr.c +++ b/drivers/ieee1394/nodemgr.c @@ -1292,7 +1292,7 @@ static void nodemgr_suspend_ne(struct node_entry *ne) if (ud->device.driver && (!ud->device.driver->suspend || - ud->device.driver->suspend(&ud->device, PMSG_SUSPEND, 0))) + ud->device.driver->suspend(&ud->device, PMSG_SUSPEND))) device_release_driver(&ud->device); } up_write(&ne->device.bus->subsys.rwsem); @@ -1315,7 +1315,7 @@ static void nodemgr_resume_ne(struct node_entry *ne) continue; if (ud->device.driver && ud->device.driver->resume) - ud->device.driver->resume(&ud->device, 0); + ud->device.driver->resume(&ud->device); } up_read(&ne->device.bus->subsys.rwsem); diff --git a/drivers/input/keyboard/corgikbd.c b/drivers/input/keyboard/corgikbd.c index 564bb365f6fc..3210d298b3bc 100644 --- a/drivers/input/keyboard/corgikbd.c +++ b/drivers/input/keyboard/corgikbd.c @@ -259,24 +259,22 @@ static void corgikbd_hinge_timer(unsigned long data) } #ifdef CONFIG_PM -static int corgikbd_suspend(struct device *dev, pm_message_t state, uint32_t level) +static int corgikbd_suspend(struct device *dev, pm_message_t state) { - if (level == SUSPEND_POWER_DOWN) { - struct corgikbd *corgikbd = dev_get_drvdata(dev); - corgikbd->suspended = 1; - } + struct corgikbd *corgikbd = dev_get_drvdata(dev); + corgikbd->suspended = 1; + return 0; } -static int corgikbd_resume(struct device *dev, uint32_t level) +static int corgikbd_resume(struct device *dev) { - if (level == RESUME_POWER_ON) { - struct corgikbd *corgikbd = dev_get_drvdata(dev); + struct corgikbd *corgikbd = dev_get_drvdata(dev); + + /* Upon resume, ignore the suspend key for a short while */ + corgikbd->suspend_jiffies=jiffies; + corgikbd->suspended = 0; - /* Upon resume, ignore the suspend key for a short while */ - corgikbd->suspend_jiffies=jiffies; - corgikbd->suspended = 0; - } return 0; } #else diff --git a/drivers/input/keyboard/spitzkbd.c b/drivers/input/keyboard/spitzkbd.c index 732fb310e487..cee9c734a048 100644 --- a/drivers/input/keyboard/spitzkbd.c +++ b/drivers/input/keyboard/spitzkbd.c @@ -309,34 +309,32 @@ static void spitzkbd_hinge_timer(unsigned long data) } #ifdef CONFIG_PM -static int spitzkbd_suspend(struct device *dev, pm_message_t state, uint32_t level) +static int spitzkbd_suspend(struct device *dev, pm_message_t state) { - if (level == SUSPEND_POWER_DOWN) { - int i; - struct spitzkbd *spitzkbd = dev_get_drvdata(dev); - spitzkbd->suspended = 1; - - /* Set Strobe lines as inputs - *except* strobe line 0 leave this - enabled so we can detect a power button press for resume */ - for (i = 1; i < SPITZ_KEY_STROBE_NUM; i++) - pxa_gpio_mode(spitz_strobes[i] | GPIO_IN); - } + int i; + struct spitzkbd *spitzkbd = dev_get_drvdata(dev); + spitzkbd->suspended = 1; + + /* Set Strobe lines as inputs - *except* strobe line 0 leave this + enabled so we can detect a power button press for resume */ + for (i = 1; i < SPITZ_KEY_STROBE_NUM; i++) + pxa_gpio_mode(spitz_strobes[i] | GPIO_IN); + return 0; } -static int spitzkbd_resume(struct device *dev, uint32_t level) +static int spitzkbd_resume(struct device *dev) { - if (level == RESUME_POWER_ON) { - int i; - struct spitzkbd *spitzkbd = dev_get_drvdata(dev); + int i; + struct spitzkbd *spitzkbd = dev_get_drvdata(dev); + + for (i = 0; i < SPITZ_KEY_STROBE_NUM; i++) + pxa_gpio_mode(spitz_strobes[i] | GPIO_OUT | GPIO_DFLT_HIGH); - for (i = 0; i < SPITZ_KEY_STROBE_NUM; i++) - pxa_gpio_mode(spitz_strobes[i] | GPIO_OUT | GPIO_DFLT_HIGH); + /* Upon resume, ignore the suspend key for a short while */ + spitzkbd->suspend_jiffies = jiffies; + spitzkbd->suspended = 0; - /* Upon resume, ignore the suspend key for a short while */ - spitzkbd->suspend_jiffies = jiffies; - spitzkbd->suspended = 0; - } return 0; } #else diff --git a/drivers/input/serio/i8042.c b/drivers/input/serio/i8042.c index 40d451ce07ff..4bc40f159996 100644 --- a/drivers/input/serio/i8042.c +++ b/drivers/input/serio/i8042.c @@ -911,12 +911,10 @@ static long i8042_panic_blink(long count) * Here we try to restore the original BIOS settings */ -static int i8042_suspend(struct device *dev, pm_message_t state, u32 level) +static int i8042_suspend(struct device *dev, pm_message_t state) { - if (level == SUSPEND_DISABLE) { - del_timer_sync(&i8042_timer); - i8042_controller_reset(); - } + del_timer_sync(&i8042_timer); + i8042_controller_reset(); return 0; } @@ -926,13 +924,10 @@ static int i8042_suspend(struct device *dev, pm_message_t state, u32 level) * Here we try to reset everything back to a state in which suspended */ -static int i8042_resume(struct device *dev, u32 level) +static int i8042_resume(struct device *dev) { int i; - if (level != RESUME_ENABLE) - return 0; - if (i8042_ctl_test()) return -1; diff --git a/drivers/input/touchscreen/corgi_ts.c b/drivers/input/touchscreen/corgi_ts.c index 40ae183ba1cd..0ba3e6562bff 100644 --- a/drivers/input/touchscreen/corgi_ts.c +++ b/drivers/input/touchscreen/corgi_ts.c @@ -231,34 +231,32 @@ static irqreturn_t ts_interrupt(int irq, void *dev_id, struct pt_regs *regs) } #ifdef CONFIG_PM -static int corgits_suspend(struct device *dev, pm_message_t state, uint32_t level) +static int corgits_suspend(struct device *dev, pm_message_t state) { - if (level == SUSPEND_POWER_DOWN) { - struct corgi_ts *corgi_ts = dev_get_drvdata(dev); - - if (corgi_ts->pendown) { - del_timer_sync(&corgi_ts->timer); - corgi_ts->tc.pressure = 0; - new_data(corgi_ts, NULL); - corgi_ts->pendown = 0; - } - corgi_ts->power_mode = PWR_MODE_SUSPEND; + struct corgi_ts *corgi_ts = dev_get_drvdata(dev); - corgi_ssp_ads7846_putget((1u << ADSCTRL_ADR_SH) | ADSCTRL_STS); + if (corgi_ts->pendown) { + del_timer_sync(&corgi_ts->timer); + corgi_ts->tc.pressure = 0; + new_data(corgi_ts, NULL); + corgi_ts->pendown = 0; } + corgi_ts->power_mode = PWR_MODE_SUSPEND; + + corgi_ssp_ads7846_putget((1u << ADSCTRL_ADR_SH) | ADSCTRL_STS); + return 0; } -static int corgits_resume(struct device *dev, uint32_t level) +static int corgits_resume(struct device *dev) { - if (level == RESUME_POWER_ON) { - struct corgi_ts *corgi_ts = dev_get_drvdata(dev); + struct corgi_ts *corgi_ts = dev_get_drvdata(dev); + + corgi_ssp_ads7846_putget((4u << ADSCTRL_ADR_SH) | ADSCTRL_STS); + /* Enable Falling Edge */ + set_irq_type(corgi_ts->irq_gpio, IRQT_FALLING); + corgi_ts->power_mode = PWR_MODE_ACTIVE; - corgi_ssp_ads7846_putget((4u << ADSCTRL_ADR_SH) | ADSCTRL_STS); - /* Enable Falling Edge */ - set_irq_type(corgi_ts->irq_gpio, IRQT_FALLING); - corgi_ts->power_mode = PWR_MODE_ACTIVE; - } return 0; } #else diff --git a/drivers/media/video/msp3400.c b/drivers/media/video/msp3400.c index f0d43fc2632f..262890cb20a7 100644 --- a/drivers/media/video/msp3400.c +++ b/drivers/media/video/msp3400.c @@ -1420,8 +1420,8 @@ static int msp_detach(struct i2c_client *client); static int msp_probe(struct i2c_adapter *adap); static int msp_command(struct i2c_client *client, unsigned int cmd, void *arg); -static int msp_suspend(struct device * dev, pm_message_t state, u32 level); -static int msp_resume(struct device * dev, u32 level); +static int msp_suspend(struct device * dev, pm_message_t state); +static int msp_resume(struct device * dev); static void msp_wake_thread(struct i2c_client *client); @@ -1821,7 +1821,7 @@ static int msp_command(struct i2c_client *client, unsigned int cmd, void *arg) return 0; } -static int msp_suspend(struct device * dev, pm_message_t state, u32 level) +static int msp_suspend(struct device * dev, pm_message_t state) { struct i2c_client *c = container_of(dev, struct i2c_client, dev); @@ -1830,7 +1830,7 @@ static int msp_suspend(struct device * dev, pm_message_t state, u32 level) return 0; } -static int msp_resume(struct device * dev, u32 level) +static int msp_resume(struct device * dev) { struct i2c_client *c = container_of(dev, struct i2c_client, dev); diff --git a/drivers/media/video/tda9887.c b/drivers/media/video/tda9887.c index 0456dda2624d..94053f149ddf 100644 --- a/drivers/media/video/tda9887.c +++ b/drivers/media/video/tda9887.c @@ -784,13 +784,13 @@ tda9887_command(struct i2c_client *client, unsigned int cmd, void *arg) return 0; } -static int tda9887_suspend(struct device * dev, pm_message_t state, u32 level) +static int tda9887_suspend(struct device * dev, pm_message_t state) { dprintk("tda9887: suspend\n"); return 0; } -static int tda9887_resume(struct device * dev, u32 level) +static int tda9887_resume(struct device * dev) { struct i2c_client *c = container_of(dev, struct i2c_client, dev); struct tda9887 *t = i2c_get_clientdata(c); diff --git a/drivers/media/video/tuner-core.c b/drivers/media/video/tuner-core.c index 05572020af4d..ad85bef1c3d5 100644 --- a/drivers/media/video/tuner-core.c +++ b/drivers/media/video/tuner-core.c @@ -697,7 +697,7 @@ static int tuner_command(struct i2c_client *client, unsigned int cmd, void *arg) return 0; } -static int tuner_suspend(struct device *dev, pm_message_t state, u32 level) +static int tuner_suspend(struct device *dev, pm_message_t state) { struct i2c_client *c = container_of (dev, struct i2c_client, dev); struct tuner *t = i2c_get_clientdata (c); @@ -707,7 +707,7 @@ static int tuner_suspend(struct device *dev, pm_message_t state, u32 level) return 0; } -static int tuner_resume(struct device *dev, u32 level) +static int tuner_resume(struct device *dev) { struct i2c_client *c = container_of (dev, struct i2c_client, dev); struct tuner *t = i2c_get_clientdata (c); diff --git a/drivers/mfd/mcp-sa11x0.c b/drivers/mfd/mcp-sa11x0.c index e9806fbbe696..720e7a326308 100644 --- a/drivers/mfd/mcp-sa11x0.c +++ b/drivers/mfd/mcp-sa11x0.c @@ -219,26 +219,24 @@ static int mcp_sa11x0_remove(struct device *dev) return 0; } -static int mcp_sa11x0_suspend(struct device *dev, pm_message_t state, u32 level) +static int mcp_sa11x0_suspend(struct device *dev, pm_message_t state) { struct mcp *mcp = dev_get_drvdata(dev); - if (level == SUSPEND_DISABLE) { - priv(mcp)->mccr0 = Ser4MCCR0; - priv(mcp)->mccr1 = Ser4MCCR1; - Ser4MCCR0 &= ~MCCR0_MCE; - } + priv(mcp)->mccr0 = Ser4MCCR0; + priv(mcp)->mccr1 = Ser4MCCR1; + Ser4MCCR0 &= ~MCCR0_MCE; + return 0; } -static int mcp_sa11x0_resume(struct device *dev, u32 level) +static int mcp_sa11x0_resume(struct device *dev) { struct mcp *mcp = dev_get_drvdata(dev); - if (level == RESUME_RESTORE_STATE) { - Ser4MCCR1 = priv(mcp)->mccr1; - Ser4MCCR0 = priv(mcp)->mccr0; - } + Ser4MCCR1 = priv(mcp)->mccr1; + Ser4MCCR0 = priv(mcp)->mccr0; + return 0; } diff --git a/drivers/mmc/pxamci.c b/drivers/mmc/pxamci.c index b53af57074e3..8eba373d42d7 100644 --- a/drivers/mmc/pxamci.c +++ b/drivers/mmc/pxamci.c @@ -571,23 +571,23 @@ static int pxamci_remove(struct device *dev) } #ifdef CONFIG_PM -static int pxamci_suspend(struct device *dev, pm_message_t state, u32 level) +static int pxamci_suspend(struct device *dev, pm_message_t state) { struct mmc_host *mmc = dev_get_drvdata(dev); int ret = 0; - if (mmc && level == SUSPEND_DISABLE) + if (mmc) ret = mmc_suspend_host(mmc, state); return ret; } -static int pxamci_resume(struct device *dev, u32 level) +static int pxamci_resume(struct device *dev) { struct mmc_host *mmc = dev_get_drvdata(dev); int ret = 0; - if (mmc && level == RESUME_ENABLE) + if (mmc) ret = mmc_resume_host(mmc); return ret; diff --git a/drivers/mmc/wbsd.c b/drivers/mmc/wbsd.c index 3cbca7cbea80..25f7ce7b3bc0 100644 --- a/drivers/mmc/wbsd.c +++ b/drivers/mmc/wbsd.c @@ -1955,14 +1955,14 @@ static void __devexit wbsd_pnp_remove(struct pnp_dev * dev) */ #ifdef CONFIG_PM -static int wbsd_suspend(struct device *dev, pm_message_t state, u32 level) +static int wbsd_suspend(struct device *dev, pm_message_t state) { DBGF("Not yet supported\n"); return 0; } -static int wbsd_resume(struct device *dev, u32 level) +static int wbsd_resume(struct device *dev) { DBGF("Not yet supported\n"); diff --git a/drivers/mtd/maps/sa1100-flash.c b/drivers/mtd/maps/sa1100-flash.c index 8dcaa357b4bb..55f21ddec3df 100644 --- a/drivers/mtd/maps/sa1100-flash.c +++ b/drivers/mtd/maps/sa1100-flash.c @@ -402,21 +402,21 @@ static int __exit sa1100_mtd_remove(struct device *dev) } #ifdef CONFIG_PM -static int sa1100_mtd_suspend(struct device *dev, pm_message_t state, u32 level) +static int sa1100_mtd_suspend(struct device *dev, pm_message_t state) { struct sa_info *info = dev_get_drvdata(dev); int ret = 0; - if (info && level == SUSPEND_SAVE_STATE) + if (info) ret = info->mtd->suspend(info->mtd); return ret; } -static int sa1100_mtd_resume(struct device *dev, u32 level) +static int sa1100_mtd_resume(struct device *dev) { struct sa_info *info = dev_get_drvdata(dev); - if (info && level == RESUME_RESTORE_STATE) + if (info) info->mtd->resume(info->mtd); return 0; } diff --git a/drivers/net/dm9000.c b/drivers/net/dm9000.c index e54fc10f6846..abce1f730d00 100644 --- a/drivers/net/dm9000.c +++ b/drivers/net/dm9000.c @@ -1140,11 +1140,11 @@ dm9000_phy_write(struct net_device *dev, int phyaddr_unused, int reg, int value) } static int -dm9000_drv_suspend(struct device *dev, pm_message_t state, u32 level) +dm9000_drv_suspend(struct device *dev, pm_message_t state) { struct net_device *ndev = dev_get_drvdata(dev); - if (ndev && level == SUSPEND_DISABLE) { + if (ndev) { if (netif_running(ndev)) { netif_device_detach(ndev); dm9000_shutdown(ndev); @@ -1154,12 +1154,12 @@ dm9000_drv_suspend(struct device *dev, pm_message_t state, u32 level) } static int -dm9000_drv_resume(struct device *dev, u32 level) +dm9000_drv_resume(struct device *dev) { struct net_device *ndev = dev_get_drvdata(dev); board_info_t *db = (board_info_t *) ndev->priv; - if (ndev && level == RESUME_ENABLE) { + if (ndev) { if (netif_running(ndev)) { dm9000_reset(db); diff --git a/drivers/net/irda/sa1100_ir.c b/drivers/net/irda/sa1100_ir.c index 8d34ac60d906..06883309916d 100644 --- a/drivers/net/irda/sa1100_ir.c +++ b/drivers/net/irda/sa1100_ir.c @@ -291,12 +291,12 @@ static void sa1100_irda_shutdown(struct sa1100_irda *si) /* * Suspend the IrDA interface. */ -static int sa1100_irda_suspend(struct device *_dev, pm_message_t state, u32 level) +static int sa1100_irda_suspend(struct device *_dev, pm_message_t state) { struct net_device *dev = dev_get_drvdata(_dev); struct sa1100_irda *si; - if (!dev || level != SUSPEND_DISABLE) + if (!dev) return 0; si = dev->priv; @@ -316,12 +316,12 @@ static int sa1100_irda_suspend(struct device *_dev, pm_message_t state, u32 leve /* * Resume the IrDA interface. */ -static int sa1100_irda_resume(struct device *_dev, u32 level) +static int sa1100_irda_resume(struct device *_dev) { struct net_device *dev = dev_get_drvdata(_dev); struct sa1100_irda *si; - if (!dev || level != RESUME_ENABLE) + if (!dev) return 0; si = dev->priv; diff --git a/drivers/net/irda/smsc-ircc2.c b/drivers/net/irda/smsc-ircc2.c index dd89bda1f131..bbac720cca63 100644 --- a/drivers/net/irda/smsc-ircc2.c +++ b/drivers/net/irda/smsc-ircc2.c @@ -213,8 +213,8 @@ static int smsc_ircc_probe_transceiver_smsc_ircc_atc(int fir_base); /* Power Management */ -static int smsc_ircc_suspend(struct device *dev, pm_message_t state, u32 level); -static int smsc_ircc_resume(struct device *dev, u32 level); +static int smsc_ircc_suspend(struct device *dev, pm_message_t state); +static int smsc_ircc_resume(struct device *dev); static struct device_driver smsc_ircc_driver = { .name = SMSC_IRCC2_DRIVER_NAME, @@ -1646,13 +1646,13 @@ static int smsc_ircc_net_close(struct net_device *dev) return 0; } -static int smsc_ircc_suspend(struct device *dev, pm_message_t state, u32 level) +static int smsc_ircc_suspend(struct device *dev, pm_message_t state) { struct smsc_ircc_cb *self = dev_get_drvdata(dev); IRDA_MESSAGE("%s, Suspending\n", driver_name); - if (level == SUSPEND_DISABLE && !self->io.suspended) { + if (!self->io.suspended) { smsc_ircc_net_close(self->netdev); self->io.suspended = 1; } @@ -1660,11 +1660,11 @@ static int smsc_ircc_suspend(struct device *dev, pm_message_t state, u32 level) return 0; } -static int smsc_ircc_resume(struct device *dev, u32 level) +static int smsc_ircc_resume(struct device *dev) { struct smsc_ircc_cb *self = dev_get_drvdata(dev); - if (level == RESUME_ENABLE && self->io.suspended) { + if (self->io.suspended) { smsc_ircc_net_open(self->netdev); self->io.suspended = 0; diff --git a/drivers/net/phy/mdio_bus.c b/drivers/net/phy/mdio_bus.c index 90630672703d..ad93b0da87f0 100644 --- a/drivers/net/phy/mdio_bus.c +++ b/drivers/net/phy/mdio_bus.c @@ -133,13 +133,9 @@ static int mdio_bus_suspend(struct device * dev, pm_message_t state) int ret = 0; struct device_driver *drv = dev->driver; - if (drv && drv->suspend) { - ret = drv->suspend(dev, state, SUSPEND_DISABLE); - if (ret == 0) - ret = drv->suspend(dev, state, SUSPEND_SAVE_STATE); - if (ret == 0) - ret = drv->suspend(dev, state, SUSPEND_POWER_DOWN); - } + if (drv && drv->suspend) + ret = drv->suspend(dev, state); + return ret; } @@ -148,13 +144,9 @@ static int mdio_bus_resume(struct device * dev) int ret = 0; struct device_driver *drv = dev->driver; - if (drv && drv->resume) { - ret = drv->resume(dev, RESUME_POWER_ON); - if (ret == 0) - ret = drv->resume(dev, RESUME_RESTORE_STATE); - if (ret == 0) - ret = drv->resume(dev, RESUME_ENABLE); - } + if (drv && drv->resume) + ret = drv->resume(dev); + return ret; } diff --git a/drivers/net/smc91x.c b/drivers/net/smc91x.c index 1438fdd20826..0ddaa611cc61 100644 --- a/drivers/net/smc91x.c +++ b/drivers/net/smc91x.c @@ -2291,11 +2291,11 @@ static int smc_drv_remove(struct device *dev) return 0; } -static int smc_drv_suspend(struct device *dev, pm_message_t state, u32 level) +static int smc_drv_suspend(struct device *dev, pm_message_t state) { struct net_device *ndev = dev_get_drvdata(dev); - if (ndev && level == SUSPEND_DISABLE) { + if (ndev) { if (netif_running(ndev)) { netif_device_detach(ndev); smc_shutdown(ndev); @@ -2305,12 +2305,12 @@ static int smc_drv_suspend(struct device *dev, pm_message_t state, u32 level) return 0; } -static int smc_drv_resume(struct device *dev, u32 level) +static int smc_drv_resume(struct device *dev) { struct platform_device *pdev = to_platform_device(dev); struct net_device *ndev = dev_get_drvdata(dev); - if (ndev && level == RESUME_ENABLE) { + if (ndev) { struct smc_local *lp = netdev_priv(ndev); smc_enable_device(pdev); if (netif_running(ndev)) { diff --git a/drivers/pci/pcie/portdrv_core.c b/drivers/pci/pcie/portdrv_core.c index 393e0cee91a9..14f05d22bb70 100644 --- a/drivers/pci/pcie/portdrv_core.c +++ b/drivers/pci/pcie/portdrv_core.c @@ -61,7 +61,7 @@ static int pcie_port_remove_service(struct device *dev) static void pcie_port_shutdown_service(struct device *dev) {} -static int pcie_port_suspend_service(struct device *dev, pm_message_t state, u32 level) +static int pcie_port_suspend_service(struct device *dev, pm_message_t state) { struct pcie_device *pciedev; struct pcie_port_service_driver *driver; @@ -76,7 +76,7 @@ static int pcie_port_suspend_service(struct device *dev, pm_message_t state, u32 return 0; } -static int pcie_port_resume_service(struct device *dev, u32 level) +static int pcie_port_resume_service(struct device *dev) { struct pcie_device *pciedev; struct pcie_port_service_driver *driver; diff --git a/drivers/pcmcia/au1000_generic.c b/drivers/pcmcia/au1000_generic.c index 470ef756252e..d90a634cebf5 100644 --- a/drivers/pcmcia/au1000_generic.c +++ b/drivers/pcmcia/au1000_generic.c @@ -519,30 +519,13 @@ static int au1x00_drv_pcmcia_probe(struct device *dev) } -static int au1x00_drv_pcmcia_suspend(struct device *dev, pm_message_t state, u32 level) -{ - int ret = 0; - if (level == SUSPEND_SAVE_STATE) - ret = pcmcia_socket_dev_suspend(dev, state); - return ret; -} - -static int au1x00_drv_pcmcia_resume(struct device *dev, u32 level) -{ - int ret = 0; - if (level == RESUME_RESTORE_STATE) - ret = pcmcia_socket_dev_resume(dev); - return ret; -} - - static struct device_driver au1x00_pcmcia_driver = { .probe = au1x00_drv_pcmcia_probe, .remove = au1x00_drv_pcmcia_remove, .name = "au1x00-pcmcia", .bus = &platform_bus_type, - .suspend = au1x00_drv_pcmcia_suspend, - .resume = au1x00_drv_pcmcia_resume + .suspend = pcmcia_socket_dev_suspend, + .resume = pcmcia_socket_dev_resume, }; static struct platform_device au1x00_device = { diff --git a/drivers/pcmcia/hd64465_ss.c b/drivers/pcmcia/hd64465_ss.c index 316f8bcc878b..b57a0b98b4d6 100644 --- a/drivers/pcmcia/hd64465_ss.c +++ b/drivers/pcmcia/hd64465_ss.c @@ -844,27 +844,11 @@ static void hs_exit_socket(hs_socket_t *sp) local_irq_restore(flags); } -static int hd64465_suspend(struct device *dev, pm_message_t state, u32 level) -{ - int ret = 0; - if (level == SUSPEND_SAVE_STATE) - ret = pcmcia_socket_dev_suspend(dev, state); - return ret; -} - -static int hd64465_resume(struct device *dev, u32 level) -{ - int ret = 0; - if (level == RESUME_RESTORE_STATE) - ret = pcmcia_socket_dev_resume(dev); - return ret; -} - static struct device_driver hd64465_driver = { .name = "hd64465-pcmcia", .bus = &platform_bus_type, - .suspend = hd64465_suspend, - .resume = hd64465_resume, + .suspend = pcmcia_socket_dev_suspend, + .resume = pcmcia_socket_dev_resume, }; static struct platform_device hd64465_device = { diff --git a/drivers/pcmcia/i82365.c b/drivers/pcmcia/i82365.c index a713015e8228..4a41f67d185d 100644 --- a/drivers/pcmcia/i82365.c +++ b/drivers/pcmcia/i82365.c @@ -1332,27 +1332,11 @@ static struct pccard_operations pcic_operations = { /*====================================================================*/ -static int i82365_suspend(struct device *dev, pm_message_t state, u32 level) -{ - int ret = 0; - if (level == SUSPEND_SAVE_STATE) - ret = pcmcia_socket_dev_suspend(dev, state); - return ret; -} - -static int i82365_resume(struct device *dev, u32 level) -{ - int ret = 0; - if (level == RESUME_RESTORE_STATE) - ret = pcmcia_socket_dev_resume(dev); - return ret; -} - static struct device_driver i82365_driver = { .name = "i82365", .bus = &platform_bus_type, - .suspend = i82365_suspend, - .resume = i82365_resume, + .suspend = pcmcia_socket_dev_suspend, + .resume = pcmcia_socket_dev_resume, }; static struct platform_device i82365_device = { diff --git a/drivers/pcmcia/m32r_cfc.c b/drivers/pcmcia/m32r_cfc.c index 65f3ee3d4d3c..c6ed70ea4812 100644 --- a/drivers/pcmcia/m32r_cfc.c +++ b/drivers/pcmcia/m32r_cfc.c @@ -731,28 +731,11 @@ static struct pccard_operations pcc_operations = { /*====================================================================*/ -static int m32r_pcc_suspend(struct device *dev, pm_message_t state, u32 level) -{ - int ret = 0; - if (level == SUSPEND_SAVE_STATE) - ret = pcmcia_socket_dev_suspend(dev, state); - return ret; -} - -static int m32r_pcc_resume(struct device *dev, u32 level) -{ - int ret = 0; - if (level == RESUME_RESTORE_STATE) - ret = pcmcia_socket_dev_resume(dev); - return ret; -} - - static struct device_driver pcc_driver = { .name = "cfc", .bus = &platform_bus_type, - .suspend = m32r_pcc_suspend, - .resume = m32r_pcc_resume, + .suspend = pcmcia_socket_dev_suspend, + .resume = pcmcia_socket_dev_resume, }; static struct platform_device pcc_device = { diff --git a/drivers/pcmcia/m32r_pcc.c b/drivers/pcmcia/m32r_pcc.c index 7b14d7efd68c..3397ff28de6a 100644 --- a/drivers/pcmcia/m32r_pcc.c +++ b/drivers/pcmcia/m32r_pcc.c @@ -695,28 +695,11 @@ static struct pccard_operations pcc_operations = { /*====================================================================*/ -static int m32r_pcc_suspend(struct device *dev, pm_message_t state, u32 level) -{ - int ret = 0; - if (level == SUSPEND_SAVE_STATE) - ret = pcmcia_socket_dev_suspend(dev, state); - return ret; -} - -static int m32r_pcc_resume(struct device *dev, u32 level) -{ - int ret = 0; - if (level == RESUME_RESTORE_STATE) - ret = pcmcia_socket_dev_resume(dev); - return ret; -} - - static struct device_driver pcc_driver = { .name = "pcc", .bus = &platform_bus_type, - .suspend = m32r_pcc_suspend, - .resume = m32r_pcc_resume, + .suspend = pcmcia_socket_dev_suspend, + .resume = pcmcia_socket_dev_resume, }; static struct platform_device pcc_device = { diff --git a/drivers/pcmcia/omap_cf.c b/drivers/pcmcia/omap_cf.c index 94be9e51654e..2558c3cc91ec 100644 --- a/drivers/pcmcia/omap_cf.c +++ b/drivers/pcmcia/omap_cf.c @@ -329,27 +329,13 @@ static int __devexit omap_cf_remove(struct device *dev) return 0; } -static int omap_cf_suspend(struct device *dev, pm_message_t mesg, u32 level) -{ - if (level != SUSPEND_SAVE_STATE) - return 0; - return pcmcia_socket_dev_suspend(dev, mesg); -} - -static int omap_cf_resume(struct device *dev, u32 level) -{ - if (level != RESUME_RESTORE_STATE) - return 0; - return pcmcia_socket_dev_resume(dev); -} - static struct device_driver omap_cf_driver = { .name = (char *) driver_name, .bus = &platform_bus_type, .probe = omap_cf_probe, .remove = __devexit_p(omap_cf_remove), - .suspend = omap_cf_suspend, - .resume = omap_cf_resume, + .suspend = pcmcia_socket_dev_suspend, + .resume = pcmcia_socket_dev_resume, }; static int __init omap_cf_init(void) diff --git a/drivers/pcmcia/pxa2xx_base.c b/drivers/pcmcia/pxa2xx_base.c index 325c992f7d8f..c2a12d53f6c7 100644 --- a/drivers/pcmcia/pxa2xx_base.c +++ b/drivers/pcmcia/pxa2xx_base.c @@ -205,32 +205,20 @@ int pxa2xx_drv_pcmcia_probe(struct device *dev) } EXPORT_SYMBOL(pxa2xx_drv_pcmcia_probe); -static int pxa2xx_drv_pcmcia_suspend(struct device *dev, pm_message_t state, u32 level) +static int pxa2xx_drv_pcmcia_resume(struct device *dev) { - int ret = 0; - if (level == SUSPEND_SAVE_STATE) - ret = pcmcia_socket_dev_suspend(dev, state); - return ret; -} + struct pcmcia_low_level *ops = dev->platform_data; + int nr = ops ? ops->nr : 0; -static int pxa2xx_drv_pcmcia_resume(struct device *dev, u32 level) -{ - int ret = 0; - if (level == RESUME_RESTORE_STATE) - { - struct pcmcia_low_level *ops = dev->platform_data; - int nr = ops ? ops->nr : 0; - - MECR = nr > 1 ? MECR_CIT | MECR_NOS : (nr > 0 ? MECR_CIT : 0); - ret = pcmcia_socket_dev_resume(dev); - } - return ret; + MECR = nr > 1 ? MECR_CIT | MECR_NOS : (nr > 0 ? MECR_CIT : 0); + + return pcmcia_socket_dev_resume(dev); } static struct device_driver pxa2xx_pcmcia_driver = { .probe = pxa2xx_drv_pcmcia_probe, .remove = soc_common_drv_pcmcia_remove, - .suspend = pxa2xx_drv_pcmcia_suspend, + .suspend = pcmcia_socket_dev_suspend, .resume = pxa2xx_drv_pcmcia_resume, .name = "pxa2xx-pcmcia", .bus = &platform_bus_type, diff --git a/drivers/pcmcia/sa1100_generic.c b/drivers/pcmcia/sa1100_generic.c index d4ed508b38be..b768fa81f043 100644 --- a/drivers/pcmcia/sa1100_generic.c +++ b/drivers/pcmcia/sa1100_generic.c @@ -74,29 +74,13 @@ static int sa11x0_drv_pcmcia_probe(struct device *dev) return ret; } -static int sa11x0_drv_pcmcia_suspend(struct device *dev, pm_message_t state, u32 level) -{ - int ret = 0; - if (level == SUSPEND_SAVE_STATE) - ret = pcmcia_socket_dev_suspend(dev, state); - return ret; -} - -static int sa11x0_drv_pcmcia_resume(struct device *dev, u32 level) -{ - int ret = 0; - if (level == RESUME_RESTORE_STATE) - ret = pcmcia_socket_dev_resume(dev); - return ret; -} - static struct device_driver sa11x0_pcmcia_driver = { .probe = sa11x0_drv_pcmcia_probe, .remove = soc_common_drv_pcmcia_remove, .name = "sa11x0-pcmcia", .bus = &platform_bus_type, - .suspend = sa11x0_drv_pcmcia_suspend, - .resume = sa11x0_drv_pcmcia_resume, + .suspend = pcmcia_socket_dev_suspend, + .resume = pcmcia_socket_dev_resume, }; /* sa11x0_pcmcia_init() diff --git a/drivers/pcmcia/tcic.c b/drivers/pcmcia/tcic.c index d5a61eae6119..f158b67f6610 100644 --- a/drivers/pcmcia/tcic.c +++ b/drivers/pcmcia/tcic.c @@ -372,27 +372,11 @@ static int __init get_tcic_id(void) /*====================================================================*/ -static int tcic_drv_suspend(struct device *dev, pm_message_t state, u32 level) -{ - int ret = 0; - if (level == SUSPEND_SAVE_STATE) - ret = pcmcia_socket_dev_suspend(dev, state); - return ret; -} - -static int tcic_drv_resume(struct device *dev, u32 level) -{ - int ret = 0; - if (level == RESUME_RESTORE_STATE) - ret = pcmcia_socket_dev_resume(dev); - return ret; -} - static struct device_driver tcic_driver = { .name = "tcic-pcmcia", .bus = &platform_bus_type, - .suspend = tcic_drv_suspend, - .resume = tcic_drv_resume, + .suspend = pcmcia_socket_dev_suspend, + .resume = pcmcia_socket_dev_resume, }; static struct platform_device tcic_device = { diff --git a/drivers/pcmcia/vrc4171_card.c b/drivers/pcmcia/vrc4171_card.c index 17bb2da6752b..3d2dca675e02 100644 --- a/drivers/pcmcia/vrc4171_card.c +++ b/drivers/pcmcia/vrc4171_card.c @@ -774,31 +774,11 @@ static int __devinit vrc4171_card_setup(char *options) __setup("vrc4171_card=", vrc4171_card_setup); -static int vrc4171_card_suspend(struct device *dev, pm_message_t state, u32 level) -{ - int retval = 0; - - if (level == SUSPEND_SAVE_STATE) - retval = pcmcia_socket_dev_suspend(dev, state); - - return retval; -} - -static int vrc4171_card_resume(struct device *dev, u32 level) -{ - int retval = 0; - - if (level == RESUME_RESTORE_STATE) - retval = pcmcia_socket_dev_resume(dev); - - return retval; -} - static struct device_driver vrc4171_card_driver = { .name = vrc4171_card_name, .bus = &platform_bus_type, - .suspend = vrc4171_card_suspend, - .resume = vrc4171_card_resume, + .suspend = pcmcia_socket_dev_suspend, + .resume = pcmcia_socket_dev_resume, }; static int __devinit vrc4171_card_init(void) diff --git a/drivers/serial/8250.c b/drivers/serial/8250.c index 4d75cdfa0a0a..afb7ddf200e0 100644 --- a/drivers/serial/8250.c +++ b/drivers/serial/8250.c @@ -2358,13 +2358,10 @@ static int __devexit serial8250_remove(struct device *dev) return 0; } -static int serial8250_suspend(struct device *dev, pm_message_t state, u32 level) +static int serial8250_suspend(struct device *dev, pm_message_t state) { int i; - if (level != SUSPEND_DISABLE) - return 0; - for (i = 0; i < UART_NR; i++) { struct uart_8250_port *up = &serial8250_ports[i]; @@ -2375,13 +2372,10 @@ static int serial8250_suspend(struct device *dev, pm_message_t state, u32 level) return 0; } -static int serial8250_resume(struct device *dev, u32 level) +static int serial8250_resume(struct device *dev) { int i; - if (level != RESUME_ENABLE) - return 0; - for (i = 0; i < UART_NR; i++) { struct uart_8250_port *up = &serial8250_ports[i]; diff --git a/drivers/serial/imx.c b/drivers/serial/imx.c index bdb4e454b8b0..5b3933b0c997 100644 --- a/drivers/serial/imx.c +++ b/drivers/serial/imx.c @@ -921,21 +921,21 @@ static struct uart_driver imx_reg = { .cons = IMX_CONSOLE, }; -static int serial_imx_suspend(struct device *_dev, pm_message_t state, u32 level) +static int serial_imx_suspend(struct device *_dev, pm_message_t state) { struct imx_port *sport = dev_get_drvdata(_dev); - if (sport && level == SUSPEND_DISABLE) + if (sport) uart_suspend_port(&imx_reg, &sport->port); return 0; } -static int serial_imx_resume(struct device *_dev, u32 level) +static int serial_imx_resume(struct device *_dev) { struct imx_port *sport = dev_get_drvdata(_dev); - if (sport && level == RESUME_ENABLE) + if (sport) uart_resume_port(&imx_reg, &sport->port); return 0; diff --git a/drivers/serial/mpc52xx_uart.c b/drivers/serial/mpc52xx_uart.c index 0585ab27ffde..8a79968f8ce1 100644 --- a/drivers/serial/mpc52xx_uart.c +++ b/drivers/serial/mpc52xx_uart.c @@ -781,22 +781,22 @@ mpc52xx_uart_remove(struct device *dev) #ifdef CONFIG_PM static int -mpc52xx_uart_suspend(struct device *dev, pm_message_t state, u32 level) +mpc52xx_uart_suspend(struct device *dev, pm_message_t state) { struct uart_port *port = (struct uart_port *) dev_get_drvdata(dev); - if (sport && level == SUSPEND_DISABLE) + if (sport) uart_suspend_port(&mpc52xx_uart_driver, port); return 0; } static int -mpc52xx_uart_resume(struct device *dev, u32 level) +mpc52xx_uart_resume(struct device *dev) { struct uart_port *port = (struct uart_port *) dev_get_drvdata(dev); - if (port && level == RESUME_ENABLE) + if (port) uart_resume_port(&mpc52xx_uart_driver, port); return 0; diff --git a/drivers/serial/pxa.c b/drivers/serial/pxa.c index 90c2a86c421b..7999686d7b47 100644 --- a/drivers/serial/pxa.c +++ b/drivers/serial/pxa.c @@ -786,21 +786,21 @@ static struct uart_driver serial_pxa_reg = { .cons = PXA_CONSOLE, }; -static int serial_pxa_suspend(struct device *_dev, pm_message_t state, u32 level) +static int serial_pxa_suspend(struct device *_dev, pm_message_t state) { struct uart_pxa_port *sport = dev_get_drvdata(_dev); - if (sport && level == SUSPEND_DISABLE) + if (sport) uart_suspend_port(&serial_pxa_reg, &sport->port); return 0; } -static int serial_pxa_resume(struct device *_dev, u32 level) +static int serial_pxa_resume(struct device *_dev) { struct uart_pxa_port *sport = dev_get_drvdata(_dev); - if (sport && level == RESUME_ENABLE) + if (sport) uart_resume_port(&serial_pxa_reg, &sport->port); return 0; diff --git a/drivers/serial/s3c2410.c b/drivers/serial/s3c2410.c index 52692aa345ec..06a17dff1a73 100644 --- a/drivers/serial/s3c2410.c +++ b/drivers/serial/s3c2410.c @@ -1134,23 +1134,22 @@ static int s3c24xx_serial_remove(struct device *_dev) #ifdef CONFIG_PM -static int s3c24xx_serial_suspend(struct device *dev, pm_message_t state, - u32 level) +static int s3c24xx_serial_suspend(struct device *dev, pm_message_t state) { struct uart_port *port = s3c24xx_dev_to_port(dev); - if (port && level == SUSPEND_DISABLE) + if (port) uart_suspend_port(&s3c24xx_uart_drv, port); return 0; } -static int s3c24xx_serial_resume(struct device *dev, u32 level) +static int s3c24xx_serial_resume(struct device *dev) { struct uart_port *port = s3c24xx_dev_to_port(dev); struct s3c24xx_uart_port *ourport = to_ourport(port); - if (port && level == RESUME_ENABLE) { + if (port) { clk_enable(ourport->clk); s3c24xx_serial_resetport(port, s3c24xx_port_to_cfg(port)); clk_disable(ourport->clk); diff --git a/drivers/serial/sa1100.c b/drivers/serial/sa1100.c index dd8aed242357..c4a789e6af44 100644 --- a/drivers/serial/sa1100.c +++ b/drivers/serial/sa1100.c @@ -834,21 +834,21 @@ static struct uart_driver sa1100_reg = { .cons = SA1100_CONSOLE, }; -static int sa1100_serial_suspend(struct device *_dev, pm_message_t state, u32 level) +static int sa1100_serial_suspend(struct device *_dev, pm_message_t state) { struct sa1100_port *sport = dev_get_drvdata(_dev); - if (sport && level == SUSPEND_DISABLE) + if (sport) uart_suspend_port(&sa1100_reg, &sport->port); return 0; } -static int sa1100_serial_resume(struct device *_dev, u32 level) +static int sa1100_serial_resume(struct device *_dev) { struct sa1100_port *sport = dev_get_drvdata(_dev); - if (sport && level == RESUME_ENABLE) + if (sport) uart_resume_port(&sa1100_reg, &sport->port); return 0; diff --git a/drivers/serial/vr41xx_siu.c b/drivers/serial/vr41xx_siu.c index 0c5d65a08f6e..2b623ab0e36e 100644 --- a/drivers/serial/vr41xx_siu.c +++ b/drivers/serial/vr41xx_siu.c @@ -976,14 +976,11 @@ static int siu_remove(struct device *dev) return 0; } -static int siu_suspend(struct device *dev, pm_message_t state, u32 level) +static int siu_suspend(struct device *dev, pm_message_t state) { struct uart_port *port; int i; - if (level != SUSPEND_DISABLE) - return 0; - for (i = 0; i < siu_uart_driver.nr; i++) { port = &siu_uart_ports[i]; if ((port->type == PORT_VR41XX_SIU || @@ -995,14 +992,11 @@ static int siu_suspend(struct device *dev, pm_message_t state, u32 level) return 0; } -static int siu_resume(struct device *dev, u32 level) +static int siu_resume(struct device *dev) { struct uart_port *port; int i; - if (level != RESUME_ENABLE) - return 0; - for (i = 0; i < siu_uart_driver.nr; i++) { port = &siu_uart_ports[i]; if ((port->type == PORT_VR41XX_SIU || diff --git a/drivers/usb/gadget/dummy_hcd.c b/drivers/usb/gadget/dummy_hcd.c index 583db7c38cf1..f2bdf4e1eb80 100644 --- a/drivers/usb/gadget/dummy_hcd.c +++ b/drivers/usb/gadget/dummy_hcd.c @@ -935,14 +935,10 @@ static int dummy_udc_remove (struct device *dev) return 0; } -static int dummy_udc_suspend (struct device *dev, pm_message_t state, - u32 level) +static int dummy_udc_suspend (struct device *dev, pm_message_t state) { struct dummy *dum = dev_get_drvdata(dev); - if (level != SUSPEND_DISABLE) - return 0; - dev_dbg (dev, "%s\n", __FUNCTION__); spin_lock_irq (&dum->lock); dum->udc_suspended = 1; @@ -954,13 +950,10 @@ static int dummy_udc_suspend (struct device *dev, pm_message_t state, return 0; } -static int dummy_udc_resume (struct device *dev, u32 level) +static int dummy_udc_resume (struct device *dev) { struct dummy *dum = dev_get_drvdata(dev); - if (level != RESUME_ENABLE) - return 0; - dev_dbg (dev, "%s\n", __FUNCTION__); spin_lock_irq (&dum->lock); dum->udc_suspended = 0; @@ -1936,14 +1929,10 @@ static int dummy_hcd_remove (struct device *dev) return 0; } -static int dummy_hcd_suspend (struct device *dev, pm_message_t state, - u32 level) +static int dummy_hcd_suspend (struct device *dev, pm_message_t state) { struct usb_hcd *hcd; - if (level != SUSPEND_DISABLE) - return 0; - dev_dbg (dev, "%s\n", __FUNCTION__); hcd = dev_get_drvdata (dev); @@ -1958,13 +1947,10 @@ static int dummy_hcd_suspend (struct device *dev, pm_message_t state, return 0; } -static int dummy_hcd_resume (struct device *dev, u32 level) +static int dummy_hcd_resume (struct device *dev) { struct usb_hcd *hcd; - if (level != RESUME_ENABLE) - return 0; - dev_dbg (dev, "%s\n", __FUNCTION__); hcd = dev_get_drvdata (dev); hcd->state = HC_STATE_RUNNING; diff --git a/drivers/usb/gadget/omap_udc.c b/drivers/usb/gadget/omap_udc.c index ff5533e69560..58b3ec97fb9a 100644 --- a/drivers/usb/gadget/omap_udc.c +++ b/drivers/usb/gadget/omap_udc.c @@ -2909,12 +2909,10 @@ static int __exit omap_udc_remove(struct device *dev) * may involve talking to an external transceiver (e.g. isp1301). */ -static int omap_udc_suspend(struct device *dev, pm_message_t message, u32 level) +static int omap_udc_suspend(struct device *dev, pm_message_t message) { u32 devstat; - if (level != SUSPEND_POWER_DOWN) - return 0; devstat = UDC_DEVSTAT_REG; /* we're requesting 48 MHz clock if the pullup is enabled @@ -2931,11 +2929,8 @@ static int omap_udc_suspend(struct device *dev, pm_message_t message, u32 level) return 0; } -static int omap_udc_resume(struct device *dev, u32 level) +static int omap_udc_resume(struct device *dev) { - if (level != RESUME_POWER_ON) - return 0; - DBG("resume + wakeup/SRP\n"); omap_pullup(&udc->gadget, 1); diff --git a/drivers/usb/gadget/pxa2xx_udc.c b/drivers/usb/gadget/pxa2xx_udc.c index 73f8c9404156..00dfe42d7a86 100644 --- a/drivers/usb/gadget/pxa2xx_udc.c +++ b/drivers/usb/gadget/pxa2xx_udc.c @@ -2602,24 +2602,23 @@ static int __exit pxa2xx_udc_remove(struct device *_dev) * VBUS IRQs should probably be ignored so that the PXA device just acts * "dead" to USB hosts until system resume. */ -static int pxa2xx_udc_suspend(struct device *dev, pm_message_t state, u32 level) +static int pxa2xx_udc_suspend(struct device *dev, pm_message_t state) { struct pxa2xx_udc *udc = dev_get_drvdata(dev); - if (level == SUSPEND_POWER_DOWN) { - if (!udc->mach->udc_command) - WARN("USB host won't detect disconnect!\n"); - pullup(udc, 0); - } + if (!udc->mach->udc_command) + WARN("USB host won't detect disconnect!\n"); + pullup(udc, 0); + return 0; } -static int pxa2xx_udc_resume(struct device *dev, u32 level) +static int pxa2xx_udc_resume(struct device *dev) { struct pxa2xx_udc *udc = dev_get_drvdata(dev); - if (level == RESUME_POWER_ON) - pullup(udc, 1); + pullup(udc, 1); + return 0; } diff --git a/drivers/usb/host/isp116x-hcd.c b/drivers/usb/host/isp116x-hcd.c index e142056b0d2c..0f6183a829c4 100644 --- a/drivers/usb/host/isp116x-hcd.c +++ b/drivers/usb/host/isp116x-hcd.c @@ -1774,15 +1774,12 @@ static int __init isp116x_probe(struct device *dev) /* Suspend of platform device */ -static int isp116x_suspend(struct device *dev, pm_message_t state, u32 phase) +static int isp116x_suspend(struct device *dev, pm_message_t state) { int ret = 0; struct usb_hcd *hcd = dev_get_drvdata(dev); - VDBG("%s: state %x, phase %x\n", __func__, state, phase); - - if (phase != SUSPEND_DISABLE && phase != SUSPEND_POWER_DOWN) - return 0; + VDBG("%s: state %x\n", __func__, state); ret = usb_suspend_device(hcd->self.root_hub, state); if (!ret) { @@ -1797,15 +1794,12 @@ static int isp116x_suspend(struct device *dev, pm_message_t state, u32 phase) /* Resume platform device */ -static int isp116x_resume(struct device *dev, u32 phase) +static int isp116x_resume(struct device *dev) { int ret = 0; struct usb_hcd *hcd = dev_get_drvdata(dev); - VDBG("%s: state %x, phase %x\n", __func__, dev->power.power_state, - phase); - if (phase != RESUME_POWER_ON) - return 0; + VDBG("%s: state %x\n", __func__, dev->power.power_state); ret = usb_resume_device(hcd->self.root_hub); if (!ret) { diff --git a/drivers/usb/host/ohci-omap.c b/drivers/usb/host/ohci-omap.c index d8f3ba7ad52e..a574216625a0 100644 --- a/drivers/usb/host/ohci-omap.c +++ b/drivers/usb/host/ohci-omap.c @@ -455,14 +455,11 @@ static int ohci_hcd_omap_drv_remove(struct device *dev) #ifdef CONFIG_PM -static int ohci_omap_suspend(struct device *dev, pm_message_t message, u32 level) +static int ohci_omap_suspend(struct device *dev, pm_message_t message) { struct ohci_hcd *ohci = hcd_to_ohci(dev_get_drvdata(dev)); int status = -EINVAL; - if (level != SUSPEND_POWER_DOWN) - return 0; - down(&ohci_to_hcd(ohci)->self.root_hub->serialize); status = ohci_hub_suspend(ohci_to_hcd(ohci)); if (status == 0) { @@ -476,14 +473,11 @@ static int ohci_omap_suspend(struct device *dev, pm_message_t message, u32 level return status; } -static int ohci_omap_resume(struct device *dev, u32 level) +static int ohci_omap_resume(struct device *dev) { struct ohci_hcd *ohci = hcd_to_ohci(dev_get_drvdata(dev)); int status = 0; - if (level != RESUME_POWER_ON) - return 0; - if (time_before(jiffies, ohci->next_statechange)) msleep(5); ohci->next_statechange = jiffies; diff --git a/drivers/usb/host/ohci-pxa27x.c b/drivers/usb/host/ohci-pxa27x.c index 2fdb262d4726..f042261ecb8e 100644 --- a/drivers/usb/host/ohci-pxa27x.c +++ b/drivers/usb/host/ohci-pxa27x.c @@ -309,7 +309,7 @@ static int ohci_hcd_pxa27x_drv_remove(struct device *dev) return 0; } -static int ohci_hcd_pxa27x_drv_suspend(struct device *dev, pm_message_t state, u32 level) +static int ohci_hcd_pxa27x_drv_suspend(struct device *dev, pm_message_t state) { // struct platform_device *pdev = to_platform_device(dev); // struct usb_hcd *hcd = dev_get_drvdata(dev); @@ -318,7 +318,7 @@ static int ohci_hcd_pxa27x_drv_suspend(struct device *dev, pm_message_t state, u return 0; } -static int ohci_hcd_pxa27x_drv_resume(struct device *dev, u32 level) +static int ohci_hcd_pxa27x_drv_resume(struct device *dev) { // struct platform_device *pdev = to_platform_device(dev); // struct usb_hcd *hcd = dev_get_drvdata(dev); diff --git a/drivers/usb/host/sl811-hcd.c b/drivers/usb/host/sl811-hcd.c index d42a15d10a46..03cf6accfe64 100644 --- a/drivers/usb/host/sl811-hcd.c +++ b/drivers/usb/host/sl811-hcd.c @@ -1784,15 +1784,12 @@ sl811h_probe(struct device *dev) */ static int -sl811h_suspend(struct device *dev, pm_message_t state, u32 phase) +sl811h_suspend(struct device *dev, pm_message_t state) { struct usb_hcd *hcd = dev_get_drvdata(dev); struct sl811 *sl811 = hcd_to_sl811(hcd); int retval = 0; - if (phase != SUSPEND_POWER_DOWN) - return retval; - if (state.event == PM_EVENT_FREEZE) retval = sl811h_hub_suspend(hcd); else if (state.event == PM_EVENT_SUSPEND) @@ -1803,14 +1800,11 @@ sl811h_suspend(struct device *dev, pm_message_t state, u32 phase) } static int -sl811h_resume(struct device *dev, u32 phase) +sl811h_resume(struct device *dev) { struct usb_hcd *hcd = dev_get_drvdata(dev); struct sl811 *sl811 = hcd_to_sl811(hcd); - if (phase != RESUME_POWER_ON) - return 0; - /* with no "check to see if VBUS is still powered" board hook, * let's assume it'd only be powered to enable remote wakeup. */ diff --git a/drivers/video/backlight/corgi_bl.c b/drivers/video/backlight/corgi_bl.c index 3c72c627e65e..1991fdb32dfb 100644 --- a/drivers/video/backlight/corgi_bl.c +++ b/drivers/video/backlight/corgi_bl.c @@ -73,17 +73,15 @@ static void corgibl_blank(int blank) } #ifdef CONFIG_PM -static int corgibl_suspend(struct device *dev, pm_message_t state, u32 level) +static int corgibl_suspend(struct device *dev, pm_message_t state) { - if (level == SUSPEND_POWER_DOWN) - corgibl_blank(FB_BLANK_POWERDOWN); + corgibl_blank(FB_BLANK_POWERDOWN); return 0; } -static int corgibl_resume(struct device *dev, u32 level) +static int corgibl_resume(struct device *dev) { - if (level == RESUME_POWER_ON) - corgibl_blank(FB_BLANK_UNBLANK); + corgibl_blank(FB_BLANK_UNBLANK); return 0; } #else diff --git a/drivers/video/imxfb.c b/drivers/video/imxfb.c index 1d54d3d6960b..0b9301facbd3 100644 --- a/drivers/video/imxfb.c +++ b/drivers/video/imxfb.c @@ -424,23 +424,21 @@ static void imxfb_setup_gpio(struct imxfb_info *fbi) * Power management hooks. Note that we won't be called from IRQ context, * unlike the blank functions above, so we may sleep. */ -static int imxfb_suspend(struct device *dev, pm_message_t state, u32 level) +static int imxfb_suspend(struct device *dev, pm_message_t state) { struct imxfb_info *fbi = dev_get_drvdata(dev); pr_debug("%s\n",__FUNCTION__); - if (level == SUSPEND_DISABLE || level == SUSPEND_POWER_DOWN) - imxfb_disable_controller(fbi); + imxfb_disable_controller(fbi); return 0; } -static int imxfb_resume(struct device *dev, u32 level) +static int imxfb_resume(struct device *dev) { struct imxfb_info *fbi = dev_get_drvdata(dev); pr_debug("%s\n",__FUNCTION__); - if (level == RESUME_ENABLE) - imxfb_enable_controller(fbi); + imxfb_enable_controller(fbi); return 0; } #else diff --git a/drivers/video/pxafb.c b/drivers/video/pxafb.c index 194eed0a238c..6206da9dd5da 100644 --- a/drivers/video/pxafb.c +++ b/drivers/video/pxafb.c @@ -981,21 +981,19 @@ pxafb_freq_policy(struct notifier_block *nb, unsigned long val, void *data) * Power management hooks. Note that we won't be called from IRQ context, * unlike the blank functions above, so we may sleep. */ -static int pxafb_suspend(struct device *dev, pm_message_t state, u32 level) +static int pxafb_suspend(struct device *dev, pm_message_t state) { struct pxafb_info *fbi = dev_get_drvdata(dev); - if (level == SUSPEND_DISABLE || level == SUSPEND_POWER_DOWN) - set_ctrlr_state(fbi, C_DISABLE_PM); + set_ctrlr_state(fbi, C_DISABLE_PM); return 0; } -static int pxafb_resume(struct device *dev, u32 level) +static int pxafb_resume(struct device *dev) { struct pxafb_info *fbi = dev_get_drvdata(dev); - if (level == RESUME_ENABLE) - set_ctrlr_state(fbi, C_ENABLE_PM); + set_ctrlr_state(fbi, C_ENABLE_PM); return 0; } #else diff --git a/drivers/video/s1d13xxxfb.c b/drivers/video/s1d13xxxfb.c index fa98d91c42eb..cb2f7a1de947 100644 --- a/drivers/video/s1d13xxxfb.c +++ b/drivers/video/s1d13xxxfb.c @@ -655,7 +655,7 @@ bail: } #ifdef CONFIG_PM -static int s1d13xxxfb_suspend(struct device *dev, pm_message_t state, u32 level) +static int s1d13xxxfb_suspend(struct device *dev, pm_message_t state) { struct fb_info *info = dev_get_drvdata(dev); struct s1d13xxxfb_par *s1dfb = info->par; @@ -702,15 +702,12 @@ static int s1d13xxxfb_suspend(struct device *dev, pm_message_t state, u32 level) return 0; } -static int s1d13xxxfb_resume(struct device *dev, u32 level) +static int s1d13xxxfb_resume(struct device *dev) { struct fb_info *info = dev_get_drvdata(dev); struct s1d13xxxfb_par *s1dfb = info->par; struct s1d13xxxfb_pdata *pdata = NULL; - if (level != RESUME_ENABLE) - return 0; - /* awaken the chip */ s1d13xxxfb_writereg(s1dfb, S1DREG_PS_CNF, 0x10); diff --git a/drivers/video/s3c2410fb.c b/drivers/video/s3c2410fb.c index 5ab79afb53b7..3862d3cb1fb2 100644 --- a/drivers/video/s3c2410fb.c +++ b/drivers/video/s3c2410fb.c @@ -847,37 +847,32 @@ static int s3c2410fb_remove(struct device *dev) /* suspend and resume support for the lcd controller */ -static int s3c2410fb_suspend(struct device *dev, pm_message_t state, u32 level) +static int s3c2410fb_suspend(struct device *dev, pm_message_t state) { struct fb_info *fbinfo = dev_get_drvdata(dev); struct s3c2410fb_info *info = fbinfo->par; - if (level == SUSPEND_DISABLE || level == SUSPEND_POWER_DOWN) { - s3c2410fb_stop_lcd(); + s3c2410fb_stop_lcd(); - /* sleep before disabling the clock, we need to ensure - * the LCD DMA engine is not going to get back on the bus - * before the clock goes off again (bjd) */ + /* sleep before disabling the clock, we need to ensure + * the LCD DMA engine is not going to get back on the bus + * before the clock goes off again (bjd) */ - msleep(1); - clk_disable(info->clk); - } + msleep(1); + clk_disable(info->clk); return 0; } -static int s3c2410fb_resume(struct device *dev, u32 level) +static int s3c2410fb_resume(struct device *dev) { struct fb_info *fbinfo = dev_get_drvdata(dev); struct s3c2410fb_info *info = fbinfo->par; - if (level == RESUME_ENABLE) { - clk_enable(info->clk); - msleep(1); - - s3c2410fb_init_registers(info); + clk_enable(info->clk); + msleep(1); - } + s3c2410fb_init_registers(info); return 0; } diff --git a/drivers/video/sa1100fb.c b/drivers/video/sa1100fb.c index 8000890e4271..78e5f194b0df 100644 --- a/drivers/video/sa1100fb.c +++ b/drivers/video/sa1100fb.c @@ -1309,21 +1309,19 @@ sa1100fb_freq_policy(struct notifier_block *nb, unsigned long val, * Power management hooks. Note that we won't be called from IRQ context, * unlike the blank functions above, so we may sleep. */ -static int sa1100fb_suspend(struct device *dev, pm_message_t state, u32 level) +static int sa1100fb_suspend(struct device *dev, pm_message_t state) { struct sa1100fb_info *fbi = dev_get_drvdata(dev); - if (level == SUSPEND_DISABLE || level == SUSPEND_POWER_DOWN) - set_ctrlr_state(fbi, C_DISABLE_PM); + set_ctrlr_state(fbi, C_DISABLE_PM); return 0; } -static int sa1100fb_resume(struct device *dev, u32 level) +static int sa1100fb_resume(struct device *dev) { struct sa1100fb_info *fbi = dev_get_drvdata(dev); - if (level == RESUME_ENABLE) - set_ctrlr_state(fbi, C_ENABLE_PM); + set_ctrlr_state(fbi, C_ENABLE_PM); return 0; } #else diff --git a/drivers/video/w100fb.c b/drivers/video/w100fb.c index 0030c071da8f..752bf88906a9 100644 --- a/drivers/video/w100fb.c +++ b/drivers/video/w100fb.c @@ -438,36 +438,34 @@ static void w100fb_restore_vidmem(struct w100fb_par *par) } } -static int w100fb_suspend(struct device *dev, pm_message_t state, uint32_t level) +static int w100fb_suspend(struct device *dev, pm_message_t state) { - if (level == SUSPEND_POWER_DOWN) { - struct fb_info *info = dev_get_drvdata(dev); - struct w100fb_par *par=info->par; - struct w100_tg_info *tg = par->mach->tg; - - w100fb_save_vidmem(par); - if(tg && tg->suspend) - tg->suspend(par); - w100_suspend(W100_SUSPEND_ALL); - par->blanked = 1; - } + struct fb_info *info = dev_get_drvdata(dev); + struct w100fb_par *par=info->par; + struct w100_tg_info *tg = par->mach->tg; + + w100fb_save_vidmem(par); + if(tg && tg->suspend) + tg->suspend(par); + w100_suspend(W100_SUSPEND_ALL); + par->blanked = 1; + return 0; } -static int w100fb_resume(struct device *dev, uint32_t level) +static int w100fb_resume(struct device *dev) { - if (level == RESUME_POWER_ON) { - struct fb_info *info = dev_get_drvdata(dev); - struct w100fb_par *par=info->par; - struct w100_tg_info *tg = par->mach->tg; + struct fb_info *info = dev_get_drvdata(dev); + struct w100fb_par *par=info->par; + struct w100_tg_info *tg = par->mach->tg; + + w100_hw_init(par); + w100fb_activate_var(par); + w100fb_restore_vidmem(par); + if(tg && tg->resume) + tg->resume(par); + par->blanked = 0; - w100_hw_init(par); - w100fb_activate_var(par); - w100fb_restore_vidmem(par); - if(tg && tg->resume) - tg->resume(par); - par->blanked = 0; - } return 0; } #else diff --git a/include/linux/device.h b/include/linux/device.h index 10ab7807f8ea..a9e72ac3fb9f 100644 --- a/include/linux/device.h +++ b/include/linux/device.h @@ -28,19 +28,6 @@ #define BUS_ID_SIZE KOBJ_NAME_LEN -enum { - SUSPEND_NOTIFY, - SUSPEND_SAVE_STATE, - SUSPEND_DISABLE, - SUSPEND_POWER_DOWN, -}; - -enum { - RESUME_POWER_ON, - RESUME_RESTORE_STATE, - RESUME_ENABLE, -}; - struct device; struct device_driver; struct class; @@ -115,8 +102,8 @@ struct device_driver { int (*probe) (struct device * dev); int (*remove) (struct device * dev); void (*shutdown) (struct device * dev); - int (*suspend) (struct device * dev, pm_message_t state, u32 level); - int (*resume) (struct device * dev, u32 level); + int (*suspend) (struct device * dev, pm_message_t state); + int (*resume) (struct device * dev); }; diff --git a/sound/arm/pxa2xx-ac97.c b/sound/arm/pxa2xx-ac97.c index 38b20efc9c0b..877bb00d3295 100644 --- a/sound/arm/pxa2xx-ac97.c +++ b/sound/arm/pxa2xx-ac97.c @@ -275,23 +275,23 @@ static int pxa2xx_ac97_do_resume(snd_card_t *card) return 0; } -static int pxa2xx_ac97_suspend(struct device *_dev, pm_message_t state, u32 level) +static int pxa2xx_ac97_suspend(struct device *_dev, pm_message_t state) { snd_card_t *card = dev_get_drvdata(_dev); int ret = 0; - if (card && level == SUSPEND_DISABLE) + if (card) ret = pxa2xx_ac97_do_suspend(card, PMSG_SUSPEND); return ret; } -static int pxa2xx_ac97_resume(struct device *_dev, u32 level) +static int pxa2xx_ac97_resume(struct device *_dev) { snd_card_t *card = dev_get_drvdata(_dev); int ret = 0; - if (card && level == RESUME_ENABLE) + if (card) ret = pxa2xx_ac97_do_resume(card); return ret; diff --git a/sound/core/init.c b/sound/core/init.c index c72a79115cca..59202de1d2ce 100644 --- a/sound/core/init.c +++ b/sound/core/init.c @@ -676,8 +676,8 @@ struct snd_generic_device { #define SND_GENERIC_NAME "snd_generic" #ifdef CONFIG_PM -static int snd_generic_suspend(struct device *dev, pm_message_t state, u32 level); -static int snd_generic_resume(struct device *dev, u32 level); +static int snd_generic_suspend(struct device *dev, pm_message_t state); +static int snd_generic_resume(struct device *dev); #endif /* initialized in sound.c */ @@ -818,13 +818,10 @@ int snd_card_set_pm_callback(snd_card_t *card, #ifdef CONFIG_SND_GENERIC_DRIVER /* suspend/resume callbacks for snd_generic platform device */ -static int snd_generic_suspend(struct device *dev, pm_message_t state, u32 level) +static int snd_generic_suspend(struct device *dev, pm_message_t state) { snd_card_t *card; - if (level != SUSPEND_DISABLE) - return 0; - card = get_snd_generic_card(dev); if (card->power_state == SNDRV_CTL_POWER_D3hot) return 0; @@ -834,13 +831,10 @@ static int snd_generic_suspend(struct device *dev, pm_message_t state, u32 level return 0; } -static int snd_generic_resume(struct device *dev, u32 level) +static int snd_generic_resume(struct device *dev) { snd_card_t *card; - if (level != RESUME_ENABLE) - return 0; - card = get_snd_generic_card(dev); if (card->power_state == SNDRV_CTL_POWER_D0) return 0; diff --git a/sound/pci/ac97/ac97_bus.c b/sound/pci/ac97/ac97_bus.c index becbc420ba41..ec70fadde7d9 100644 --- a/sound/pci/ac97/ac97_bus.c +++ b/sound/pci/ac97/ac97_bus.c @@ -31,7 +31,8 @@ static int ac97_bus_suspend(struct device *dev, pm_message_t state) int ret = 0; if (dev->driver && dev->driver->suspend) - ret = dev->driver->suspend(dev, state, SUSPEND_POWER_DOWN); + ret = dev->driver->suspend(dev, state); + return ret; } @@ -40,7 +41,8 @@ static int ac97_bus_resume(struct device *dev) int ret = 0; if (dev->driver && dev->driver->resume) - ret = dev->driver->resume(dev, RESUME_POWER_ON); + ret = dev->driver->resume(dev); + return ret; } -- cgit v1.2.3 From e89e9cf539a28df7d0eb1d0a545368e9920b34ac Mon Sep 17 00:00:00 2001 From: Ananda Raju Date: Tue, 18 Oct 2005 15:46:41 -0700 Subject: [IPv4/IPv6]: UFO Scatter-gather approach Attached is kernel patch for UDP Fragmentation Offload (UFO) feature. 1. This patch incorporate the review comments by Jeff Garzik. 2. Renamed USO as UFO (UDP Fragmentation Offload) 3. udp sendfile support with UFO This patches uses scatter-gather feature of skb to generate large UDP datagram. Below is a "how-to" on changes required in network device driver to use the UFO interface. UDP Fragmentation Offload (UFO) Interface: ------------------------------------------- UFO is a feature wherein the Linux kernel network stack will offload the IP fragmentation functionality of large UDP datagram to hardware. This will reduce the overhead of stack in fragmenting the large UDP datagram to MTU sized packets 1) Drivers indicate their capability of UFO using dev->features |= NETIF_F_UFO | NETIF_F_HW_CSUM | NETIF_F_SG NETIF_F_HW_CSUM is required for UFO over ipv6. 2) UFO packet will be submitted for transmission using driver xmit routine. UFO packet will have a non-zero value for "skb_shinfo(skb)->ufo_size" skb_shinfo(skb)->ufo_size will indicate the length of data part in each IP fragment going out of the adapter after IP fragmentation by hardware. skb->data will contain MAC/IP/UDP header and skb_shinfo(skb)->frags[] contains the data payload. The skb->ip_summed will be set to CHECKSUM_HW indicating that hardware has to do checksum calculation. Hardware should compute the UDP checksum of complete datagram and also ip header checksum of each fragmented IP packet. For IPV6 the UFO provides the fragment identification-id in skb_shinfo(skb)->ip6_frag_id. The adapter should use this ID for generating IPv6 fragments. Signed-off-by: Ananda Raju Signed-off-by: Rusty Russell (forwarded) Signed-off-by: Arnaldo Carvalho de Melo --- include/linux/ethtool.h | 8 +++++ include/linux/netdevice.h | 1 + include/linux/skbuff.h | 7 ++++ net/core/dev.c | 14 ++++++++ net/core/ethtool.c | 53 ++++++++++++++++++++++++++++++ net/core/skbuff.c | 75 ++++++++++++++++++++++++++++++++++++++++++ net/ipv4/ip_output.c | 83 ++++++++++++++++++++++++++++++++++++++++++++--- net/ipv6/ip6_output.c | 71 +++++++++++++++++++++++++++++++++++++++- 8 files changed, 306 insertions(+), 6 deletions(-) (limited to 'include/linux') diff --git a/include/linux/ethtool.h b/include/linux/ethtool.h index ed1440ea4c91..d2c390eff1b2 100644 --- a/include/linux/ethtool.h +++ b/include/linux/ethtool.h @@ -269,6 +269,8 @@ u32 ethtool_op_get_tso(struct net_device *dev); int ethtool_op_set_tso(struct net_device *dev, u32 data); int ethtool_op_get_perm_addr(struct net_device *dev, struct ethtool_perm_addr *addr, u8 *data); +u32 ethtool_op_get_ufo(struct net_device *dev); +int ethtool_op_set_ufo(struct net_device *dev, u32 data); /** * ðtool_ops - Alter and report network device settings @@ -298,6 +300,8 @@ int ethtool_op_get_perm_addr(struct net_device *dev, * set_sg: Turn scatter-gather on or off * get_tso: Report whether TCP segmentation offload is enabled * set_tso: Turn TCP segmentation offload on or off + * get_ufo: Report whether UDP fragmentation offload is enabled + * set_ufo: Turn UDP fragmentation offload on or off * self_test: Run specified self-tests * get_strings: Return a set of strings that describe the requested objects * phys_id: Identify the device @@ -364,6 +368,8 @@ struct ethtool_ops { int (*get_perm_addr)(struct net_device *, struct ethtool_perm_addr *, u8 *); int (*begin)(struct net_device *); void (*complete)(struct net_device *); + u32 (*get_ufo)(struct net_device *); + int (*set_ufo)(struct net_device *, u32); }; /* CMDs currently supported */ @@ -400,6 +406,8 @@ struct ethtool_ops { #define ETHTOOL_GTSO 0x0000001e /* Get TSO enable (ethtool_value) */ #define ETHTOOL_STSO 0x0000001f /* Set TSO enable (ethtool_value) */ #define ETHTOOL_GPERMADDR 0x00000020 /* Get permanent hardware address */ +#define ETHTOOL_GUFO 0x00000021 /* Get UFO enable (ethtool_value) */ +#define ETHTOOL_SUFO 0x00000022 /* Set UFO enable (ethtool_value) */ /* compatibility with older code */ #define SPARC_ETH_GSET ETHTOOL_GSET diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index a9281b24c40b..c6efce4a04a4 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -308,6 +308,7 @@ struct net_device #define NETIF_F_VLAN_CHALLENGED 1024 /* Device cannot handle VLAN packets */ #define NETIF_F_TSO 2048 /* Can offload TCP/IP segmentation */ #define NETIF_F_LLTX 4096 /* LockLess TX */ +#define NETIF_F_UFO 8192 /* Can offload UDP Large Send*/ struct net_device *next_sched; diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h index b756935da9c8..4286d832166f 100644 --- a/include/linux/skbuff.h +++ b/include/linux/skbuff.h @@ -137,6 +137,8 @@ struct skb_shared_info { unsigned int nr_frags; unsigned short tso_size; unsigned short tso_segs; + unsigned short ufo_size; + unsigned int ip6_frag_id; struct sk_buff *frag_list; skb_frag_t frags[MAX_SKB_FRAGS]; }; @@ -341,6 +343,11 @@ extern void skb_over_panic(struct sk_buff *skb, int len, extern void skb_under_panic(struct sk_buff *skb, int len, void *here); +extern int skb_append_datato_frags(struct sock *sk, struct sk_buff *skb, + int getfrag(void *from, char *to, int offset, + int len,int odd, struct sk_buff *skb), + void *from, int length); + struct skb_seq_state { __u32 lower_offset; diff --git a/net/core/dev.c b/net/core/dev.c index a44eeef24edf..8d1541595277 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -2717,6 +2717,20 @@ int register_netdevice(struct net_device *dev) dev->name); dev->features &= ~NETIF_F_TSO; } + if (dev->features & NETIF_F_UFO) { + if (!(dev->features & NETIF_F_HW_CSUM)) { + printk(KERN_ERR "%s: Dropping NETIF_F_UFO since no " + "NETIF_F_HW_CSUM feature.\n", + dev->name); + dev->features &= ~NETIF_F_UFO; + } + if (!(dev->features & NETIF_F_SG)) { + printk(KERN_ERR "%s: Dropping NETIF_F_UFO since no " + "NETIF_F_SG feature.\n", + dev->name); + dev->features &= ~NETIF_F_UFO; + } + } /* * nil rebuild_header routine, diff --git a/net/core/ethtool.c b/net/core/ethtool.c index 404b761e82ce..0350586e9195 100644 --- a/net/core/ethtool.c +++ b/net/core/ethtool.c @@ -93,6 +93,20 @@ int ethtool_op_get_perm_addr(struct net_device *dev, struct ethtool_perm_addr *a } +u32 ethtool_op_get_ufo(struct net_device *dev) +{ + return (dev->features & NETIF_F_UFO) != 0; +} + +int ethtool_op_set_ufo(struct net_device *dev, u32 data) +{ + if (data) + dev->features |= NETIF_F_UFO; + else + dev->features &= ~NETIF_F_UFO; + return 0; +} + /* Handlers for each ethtool command */ static int ethtool_get_settings(struct net_device *dev, void __user *useraddr) @@ -483,6 +497,11 @@ static int __ethtool_set_sg(struct net_device *dev, u32 data) return err; } + if (!data && dev->ethtool_ops->set_ufo) { + err = dev->ethtool_ops->set_ufo(dev, 0); + if (err) + return err; + } return dev->ethtool_ops->set_sg(dev, data); } @@ -569,6 +588,32 @@ static int ethtool_set_tso(struct net_device *dev, char __user *useraddr) return dev->ethtool_ops->set_tso(dev, edata.data); } +static int ethtool_get_ufo(struct net_device *dev, char __user *useraddr) +{ + struct ethtool_value edata = { ETHTOOL_GTSO }; + + if (!dev->ethtool_ops->get_ufo) + return -EOPNOTSUPP; + edata.data = dev->ethtool_ops->get_ufo(dev); + if (copy_to_user(useraddr, &edata, sizeof(edata))) + return -EFAULT; + return 0; +} +static int ethtool_set_ufo(struct net_device *dev, char __user *useraddr) +{ + struct ethtool_value edata; + + if (!dev->ethtool_ops->set_ufo) + return -EOPNOTSUPP; + if (copy_from_user(&edata, useraddr, sizeof(edata))) + return -EFAULT; + if (edata.data && !(dev->features & NETIF_F_SG)) + return -EINVAL; + if (edata.data && !(dev->features & NETIF_F_HW_CSUM)) + return -EINVAL; + return dev->ethtool_ops->set_ufo(dev, edata.data); +} + static int ethtool_self_test(struct net_device *dev, char __user *useraddr) { struct ethtool_test test; @@ -854,6 +899,12 @@ int dev_ethtool(struct ifreq *ifr) case ETHTOOL_GPERMADDR: rc = ethtool_get_perm_addr(dev, useraddr); break; + case ETHTOOL_GUFO: + rc = ethtool_get_ufo(dev, useraddr); + break; + case ETHTOOL_SUFO: + rc = ethtool_set_ufo(dev, useraddr); + break; default: rc = -EOPNOTSUPP; } @@ -882,3 +933,5 @@ EXPORT_SYMBOL(ethtool_op_set_sg); EXPORT_SYMBOL(ethtool_op_set_tso); EXPORT_SYMBOL(ethtool_op_set_tx_csum); EXPORT_SYMBOL(ethtool_op_set_tx_hw_csum); +EXPORT_SYMBOL(ethtool_op_set_ufo); +EXPORT_SYMBOL(ethtool_op_get_ufo); diff --git a/net/core/skbuff.c b/net/core/skbuff.c index ef9d46b91eb9..95501e40100e 100644 --- a/net/core/skbuff.c +++ b/net/core/skbuff.c @@ -176,6 +176,8 @@ struct sk_buff *__alloc_skb(unsigned int size, gfp_t gfp_mask, skb_shinfo(skb)->tso_size = 0; skb_shinfo(skb)->tso_segs = 0; skb_shinfo(skb)->frag_list = NULL; + skb_shinfo(skb)->ufo_size = 0; + skb_shinfo(skb)->ip6_frag_id = 0; out: return skb; nodata: @@ -1696,6 +1698,78 @@ unsigned int skb_find_text(struct sk_buff *skb, unsigned int from, return textsearch_find(config, state); } +/** + * skb_append_datato_frags: - append the user data to a skb + * @sk: sock structure + * @skb: skb structure to be appened with user data. + * @getfrag: call back function to be used for getting the user data + * @from: pointer to user message iov + * @length: length of the iov message + * + * Description: This procedure append the user data in the fragment part + * of the skb if any page alloc fails user this procedure returns -ENOMEM + */ +int skb_append_datato_frags(struct sock *sk, struct sk_buff *skb, + int getfrag(void *from, char *to, int offset, + int len, int odd, struct sk_buff *skb), + void *from, int length) +{ + int frg_cnt = 0; + skb_frag_t *frag = NULL; + struct page *page = NULL; + int copy, left; + int offset = 0; + int ret; + + do { + /* Return error if we don't have space for new frag */ + frg_cnt = skb_shinfo(skb)->nr_frags; + if (frg_cnt >= MAX_SKB_FRAGS) + return -EFAULT; + + /* allocate a new page for next frag */ + page = alloc_pages(sk->sk_allocation, 0); + + /* If alloc_page fails just return failure and caller will + * free previous allocated pages by doing kfree_skb() + */ + if (page == NULL) + return -ENOMEM; + + /* initialize the next frag */ + sk->sk_sndmsg_page = page; + sk->sk_sndmsg_off = 0; + skb_fill_page_desc(skb, frg_cnt, page, 0, 0); + skb->truesize += PAGE_SIZE; + atomic_add(PAGE_SIZE, &sk->sk_wmem_alloc); + + /* get the new initialized frag */ + frg_cnt = skb_shinfo(skb)->nr_frags; + frag = &skb_shinfo(skb)->frags[frg_cnt - 1]; + + /* copy the user data to page */ + left = PAGE_SIZE - frag->page_offset; + copy = (length > left)? left : length; + + ret = getfrag(from, (page_address(frag->page) + + frag->page_offset + frag->size), + offset, copy, 0, skb); + if (ret < 0) + return -EFAULT; + + /* copy was successful so update the size parameters */ + sk->sk_sndmsg_off += copy; + frag->size += copy; + skb->len += copy; + skb->data_len += copy; + offset += copy; + length -= copy; + + } while (length > 0); + + return 0; +} + void __init skb_init(void) { skbuff_head_cache = kmem_cache_create("skbuff_head_cache", @@ -1747,3 +1821,4 @@ EXPORT_SYMBOL(skb_prepare_seq_read); EXPORT_SYMBOL(skb_seq_read); EXPORT_SYMBOL(skb_abort_seq_read); EXPORT_SYMBOL(skb_find_text); +EXPORT_SYMBOL(skb_append_datato_frags); diff --git a/net/ipv4/ip_output.c b/net/ipv4/ip_output.c index 87e350069abb..17758234a3e3 100644 --- a/net/ipv4/ip_output.c +++ b/net/ipv4/ip_output.c @@ -275,7 +275,8 @@ int ip_output(struct sk_buff *skb) { IP_INC_STATS(IPSTATS_MIB_OUTREQUESTS); - if (skb->len > dst_mtu(skb->dst) && !skb_shinfo(skb)->tso_size) + if (skb->len > dst_mtu(skb->dst) && + !(skb_shinfo(skb)->ufo_size || skb_shinfo(skb)->tso_size)) return ip_fragment(skb, ip_finish_output); else return ip_finish_output(skb); @@ -688,6 +689,60 @@ csum_page(struct page *page, int offset, int copy) return csum; } +inline int ip_ufo_append_data(struct sock *sk, + int getfrag(void *from, char *to, int offset, int len, + int odd, struct sk_buff *skb), + void *from, int length, int hh_len, int fragheaderlen, + int transhdrlen, int mtu,unsigned int flags) +{ + struct sk_buff *skb; + int err; + + /* There is support for UDP fragmentation offload by network + * device, so create one single skb packet containing complete + * udp datagram + */ + if ((skb = skb_peek_tail(&sk->sk_write_queue)) == NULL) { + skb = sock_alloc_send_skb(sk, + hh_len + fragheaderlen + transhdrlen + 20, + (flags & MSG_DONTWAIT), &err); + + if (skb == NULL) + return err; + + /* reserve space for Hardware header */ + skb_reserve(skb, hh_len); + + /* create space for UDP/IP header */ + skb_put(skb,fragheaderlen + transhdrlen); + + /* initialize network header pointer */ + skb->nh.raw = skb->data; + + /* initialize protocol header pointer */ + skb->h.raw = skb->data + fragheaderlen; + + skb->ip_summed = CHECKSUM_HW; + skb->csum = 0; + sk->sk_sndmsg_off = 0; + } + + err = skb_append_datato_frags(sk,skb, getfrag, from, + (length - transhdrlen)); + if (!err) { + /* specify the length of each IP datagram fragment*/ + skb_shinfo(skb)->ufo_size = (mtu - fragheaderlen); + __skb_queue_tail(&sk->sk_write_queue, skb); + + return 0; + } + /* There is not enough support do UFO , + * so follow normal path + */ + kfree_skb(skb); + return err; +} + /* * ip_append_data() and ip_append_page() can make one large IP datagram * from many pieces of data. Each pieces will be holded on the socket @@ -777,6 +832,15 @@ int ip_append_data(struct sock *sk, csummode = CHECKSUM_HW; inet->cork.length += length; + if (((length > mtu) && (sk->sk_protocol == IPPROTO_UDP)) && + (rt->u.dst.dev->features & NETIF_F_UFO)) { + + if(ip_ufo_append_data(sk, getfrag, from, length, hh_len, + fragheaderlen, transhdrlen, mtu, flags)) + goto error; + + return 0; + } /* So, what's going on in the loop below? * @@ -1008,14 +1072,23 @@ ssize_t ip_append_page(struct sock *sk, struct page *page, return -EINVAL; inet->cork.length += size; + if ((sk->sk_protocol == IPPROTO_UDP) && + (rt->u.dst.dev->features & NETIF_F_UFO)) + skb_shinfo(skb)->ufo_size = (mtu - fragheaderlen); + while (size > 0) { int i; - /* Check if the remaining data fits into current packet. */ - len = mtu - skb->len; - if (len < size) - len = maxfraglen - skb->len; + if (skb_shinfo(skb)->ufo_size) + len = size; + else { + + /* Check if the remaining data fits into current packet. */ + len = mtu - skb->len; + if (len < size) + len = maxfraglen - skb->len; + } if (len <= 0) { struct sk_buff *skb_prev; char *data; diff --git a/net/ipv6/ip6_output.c b/net/ipv6/ip6_output.c index 563b442ffab8..614296a920c6 100644 --- a/net/ipv6/ip6_output.c +++ b/net/ipv6/ip6_output.c @@ -147,7 +147,8 @@ static int ip6_output2(struct sk_buff *skb) int ip6_output(struct sk_buff *skb) { - if (skb->len > dst_mtu(skb->dst) || dst_allfrag(skb->dst)) + if ((skb->len > dst_mtu(skb->dst) && !skb_shinfo(skb)->ufo_size) || + dst_allfrag(skb->dst)) return ip6_fragment(skb, ip6_output2); else return ip6_output2(skb); @@ -768,6 +769,65 @@ out_err_release: *dst = NULL; return err; } +inline int ip6_ufo_append_data(struct sock *sk, + int getfrag(void *from, char *to, int offset, int len, + int odd, struct sk_buff *skb), + void *from, int length, int hh_len, int fragheaderlen, + int transhdrlen, int mtu,unsigned int flags) + +{ + struct sk_buff *skb; + int err; + + /* There is support for UDP large send offload by network + * device, so create one single skb packet containing complete + * udp datagram + */ + if ((skb = skb_peek_tail(&sk->sk_write_queue)) == NULL) { + skb = sock_alloc_send_skb(sk, + hh_len + fragheaderlen + transhdrlen + 20, + (flags & MSG_DONTWAIT), &err); + if (skb == NULL) + return -ENOMEM; + + /* reserve space for Hardware header */ + skb_reserve(skb, hh_len); + + /* create space for UDP/IP header */ + skb_put(skb,fragheaderlen + transhdrlen); + + /* initialize network header pointer */ + skb->nh.raw = skb->data; + + /* initialize protocol header pointer */ + skb->h.raw = skb->data + fragheaderlen; + + skb->ip_summed = CHECKSUM_HW; + skb->csum = 0; + sk->sk_sndmsg_off = 0; + } + + err = skb_append_datato_frags(sk,skb, getfrag, from, + (length - transhdrlen)); + if (!err) { + struct frag_hdr fhdr; + + /* specify the length of each IP datagram fragment*/ + skb_shinfo(skb)->ufo_size = (mtu - fragheaderlen) - + sizeof(struct frag_hdr); + ipv6_select_ident(skb, &fhdr); + skb_shinfo(skb)->ip6_frag_id = fhdr.identification; + __skb_queue_tail(&sk->sk_write_queue, skb); + + return 0; + } + /* There is not enough support do UPD LSO, + * so follow normal path + */ + kfree_skb(skb); + + return err; +} int ip6_append_data(struct sock *sk, int getfrag(void *from, char *to, int offset, int len, int odd, struct sk_buff *skb), @@ -860,6 +920,15 @@ int ip6_append_data(struct sock *sk, int getfrag(void *from, char *to, */ inet->cork.length += length; + if (((length > mtu) && (sk->sk_protocol == IPPROTO_UDP)) && + (rt->u.dst.dev->features & NETIF_F_UFO)) { + + if(ip6_ufo_append_data(sk, getfrag, from, length, hh_len, + fragheaderlen, transhdrlen, mtu, flags)) + goto error; + + return 0; + } if ((skb = skb_peek_tail(&sk->sk_write_queue)) == NULL) goto alloc_new_skb; -- cgit v1.2.3 From a4e2b347848bf626b822599329933887dc90e50f Mon Sep 17 00:00:00 2001 From: Michael Chan Date: Wed, 26 Oct 2005 15:46:52 -0700 Subject: [PATCH] tg3: add 5714/5715 support Add complete support for 5714/5715. These chips are very similar to 5780 so the changes are very trivial. A TG3_FLG2_5780_CLASS flag is added to identify these chips. Signed-off-by: Michael Chan Signed-off-by: Jeff Garzik --- drivers/net/tg3.c | 39 +++++++++++++++++++++++++-------------- drivers/net/tg3.h | 11 +++++++++-- include/linux/pci_ids.h | 2 ++ 3 files changed, 36 insertions(+), 16 deletions(-) (limited to 'include/linux') diff --git a/drivers/net/tg3.c b/drivers/net/tg3.c index 1802c3b48799..cf2204f5b7f9 100644 --- a/drivers/net/tg3.c +++ b/drivers/net/tg3.c @@ -219,6 +219,10 @@ static struct pci_device_id tg3_pci_tbl[] = { PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0UL }, { PCI_VENDOR_ID_BROADCOM, PCI_DEVICE_ID_TIGON3_5753F, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0UL }, + { PCI_VENDOR_ID_BROADCOM, PCI_DEVICE_ID_TIGON3_5714, + PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0UL }, + { PCI_VENDOR_ID_BROADCOM, PCI_DEVICE_ID_TIGON3_5715, + PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0UL }, { PCI_VENDOR_ID_BROADCOM, PCI_DEVICE_ID_TIGON3_5780, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0UL }, { PCI_VENDOR_ID_BROADCOM, PCI_DEVICE_ID_TIGON3_5780S, @@ -570,7 +574,7 @@ static void tg3_switch_clocks(struct tg3 *tp) u32 clock_ctrl = tr32(TG3PCI_CLOCK_CTRL); u32 orig_clock_ctrl; - if (GET_ASIC_REV(tp->pci_chip_rev_id) == ASIC_REV_5780) + if (tp->tg3_flags2 & TG3_FLG2_5780_CLASS) return; orig_clock_ctrl = clock_ctrl; @@ -1210,7 +1214,7 @@ static int tg3_set_power_state(struct tg3 *tp, int state) CLOCK_CTRL_ALTCLK | CLOCK_CTRL_PWRDOWN_PLL133); udelay(40); - } else if (GET_ASIC_REV(tp->pci_chip_rev_id) == ASIC_REV_5780) { + } else if (tp->tg3_flags2 & TG3_FLG2_5780_CLASS) { /* do nothing */ } else if (!((tp->tg3_flags2 & TG3_FLG2_5750_PLUS) && (tp->tg3_flags & TG3_FLAG_ENABLE_ASF))) { @@ -3712,14 +3716,14 @@ static inline void tg3_set_mtu(struct net_device *dev, struct tg3 *tp, dev->mtu = new_mtu; if (new_mtu > ETH_DATA_LEN) { - if (GET_ASIC_REV(tp->pci_chip_rev_id) == ASIC_REV_5780) { + if (tp->tg3_flags2 & TG3_FLG2_5780_CLASS) { tp->tg3_flags2 &= ~TG3_FLG2_TSO_CAPABLE; ethtool_op_set_tso(dev, 0); } else tp->tg3_flags |= TG3_FLAG_JUMBO_RING_ENABLE; } else { - if (GET_ASIC_REV(tp->pci_chip_rev_id) == ASIC_REV_5780) + if (tp->tg3_flags2 & TG3_FLG2_5780_CLASS) tp->tg3_flags2 |= TG3_FLG2_TSO_CAPABLE; tp->tg3_flags &= ~TG3_FLAG_JUMBO_RING_ENABLE; } @@ -3850,7 +3854,7 @@ static void tg3_init_rings(struct tg3 *tp) memset(tp->tx_ring, 0, TG3_TX_RING_BYTES); tp->rx_pkt_buf_sz = RX_PKT_BUF_SZ; - if ((GET_ASIC_REV(tp->pci_chip_rev_id) == ASIC_REV_5780) && + if ((tp->tg3_flags2 & TG3_FLG2_5780_CLASS) && (tp->dev->mtu > ETH_DATA_LEN)) tp->rx_pkt_buf_sz = RX_JUMBO_PKT_BUF_SZ; @@ -4347,7 +4351,7 @@ static int tg3_chip_reset(struct tg3 *tp) val &= ~PCIX_CAPS_RELAXED_ORDERING; pci_write_config_dword(tp->pdev, TG3PCI_X_CAPS, val); - if (GET_ASIC_REV(tp->pci_chip_rev_id) == ASIC_REV_5780) { + if (tp->tg3_flags2 & TG3_FLG2_5780_CLASS) { u32 val; /* Chip reset on 5780 will reset MSI enable bit, @@ -6003,7 +6007,7 @@ static int tg3_reset_hw(struct tg3 *tp) tw32(MAC_RCV_VALUE_1, 0xffffffff & RCV_RULE_DISABLE_MASK); if ((tp->tg3_flags2 & TG3_FLG2_5705_PLUS) && - (GET_ASIC_REV(tp->pci_chip_rev_id) != ASIC_REV_5780)) + !(tp->tg3_flags2 & TG3_FLG2_5780_CLASS)) limit = 8; else limit = 16; @@ -7237,7 +7241,7 @@ static int tg3_get_settings(struct net_device *dev, struct ethtool_cmd *cmd) cmd->supported |= (SUPPORTED_1000baseT_Half | SUPPORTED_1000baseT_Full); - if (!(tp->tg3_flags2 & TG3_FLG2_PHY_SERDES)) + if (!(tp->tg3_flags2 & TG3_FLG2_ANY_SERDES)) cmd->supported |= (SUPPORTED_100baseT_Half | SUPPORTED_100baseT_Full | SUPPORTED_10baseT_Half | @@ -8380,7 +8384,7 @@ static void __devinit tg3_get_nvram_info(struct tg3 *tp) } if ((GET_ASIC_REV(tp->pci_chip_rev_id) == ASIC_REV_5750) || - (GET_ASIC_REV(tp->pci_chip_rev_id) == ASIC_REV_5780)) { + (tp->tg3_flags2 & TG3_FLG2_5780_CLASS)) { switch (nvcfg1 & NVRAM_CFG1_VENDOR_MASK) { case FLASH_VENDOR_ATMEL_FLASH_BUFFERED: tp->nvram_jedecnum = JEDEC_ATMEL; @@ -8980,7 +8984,7 @@ static void __devinit tg3_get_eeprom_hw_cfg(struct tg3 *tp) tp->phy_id = eeprom_phy_id; if (eeprom_phy_serdes) { - if (GET_ASIC_REV(tp->pci_chip_rev_id) == ASIC_REV_5780) + if (tp->tg3_flags2 & TG3_FLG2_5780_CLASS) tp->tg3_flags2 |= TG3_FLG2_MII_SERDES; else tp->tg3_flags2 |= TG3_FLG2_PHY_SERDES; @@ -9393,8 +9397,11 @@ static int __devinit tg3_get_invariants(struct tg3 *tp) } /* Find msi capability. */ - if (GET_ASIC_REV(tp->pci_chip_rev_id) == ASIC_REV_5780) + if (GET_ASIC_REV(tp->pci_chip_rev_id) == ASIC_REV_5780 || + GET_ASIC_REV(tp->pci_chip_rev_id) == ASIC_REV_5714) { + tp->tg3_flags2 |= TG3_FLG2_5780_CLASS; tp->msi_cap = pci_find_capability(tp->pdev, PCI_CAP_ID_MSI); + } /* Initialize misc host control in PCI block. */ tp->misc_host_ctrl |= (misc_ctrl_reg & @@ -9412,7 +9419,7 @@ static int __devinit tg3_get_invariants(struct tg3 *tp) if (GET_ASIC_REV(tp->pci_chip_rev_id) == ASIC_REV_5750 || GET_ASIC_REV(tp->pci_chip_rev_id) == ASIC_REV_5752 || - GET_ASIC_REV(tp->pci_chip_rev_id) == ASIC_REV_5780) + (tp->tg3_flags2 & TG3_FLG2_5780_CLASS)) tp->tg3_flags2 |= TG3_FLG2_5750_PLUS; if ((GET_ASIC_REV(tp->pci_chip_rev_id) == ASIC_REV_5705) || @@ -9607,7 +9614,7 @@ static int __devinit tg3_get_invariants(struct tg3 *tp) * ether_setup() via the alloc_etherdev() call */ if (tp->dev->mtu > ETH_DATA_LEN && - GET_ASIC_REV(tp->pci_chip_rev_id) != ASIC_REV_5780) + !(tp->tg3_flags2 & TG3_FLG2_5780_CLASS)) tp->tg3_flags |= TG3_FLAG_JUMBO_RING_ENABLE; /* Determine WakeOnLan speed to use. */ @@ -9830,7 +9837,7 @@ static int __devinit tg3_get_device_address(struct tg3 *tp) mac_offset = 0x7c; if ((GET_ASIC_REV(tp->pci_chip_rev_id) == ASIC_REV_5704 && !(tp->tg3_flags & TG3_FLG2_SUN_570X)) || - GET_ASIC_REV(tp->pci_chip_rev_id) == ASIC_REV_5780) { + (tp->tg3_flags2 & TG3_FLG2_5780_CLASS)) { if (tr32(TG3PCI_DUAL_MAC_CTRL) & DUAL_MAC_CTRL_ID) mac_offset = 0xcc; if (tg3_nvram_lock(tp)) @@ -10148,6 +10155,9 @@ static int __devinit tg3_test_dma(struct tg3 *tp) } else if (GET_ASIC_REV(tp->pci_chip_rev_id) == ASIC_REV_5780) { /* 5780 always in PCIX mode */ tp->dma_rwctrl |= 0x00144000; + } else if (GET_ASIC_REV(tp->pci_chip_rev_id) == ASIC_REV_5714) { + /* 5714 always in PCIX mode */ + tp->dma_rwctrl |= 0x00148000; } else { tp->dma_rwctrl |= 0x001b000f; } @@ -10347,6 +10357,7 @@ static char * __devinit tg3_phy_string(struct tg3 *tp) case PHY_ID_BCM5705: return "5705"; case PHY_ID_BCM5750: return "5750"; case PHY_ID_BCM5752: return "5752"; + case PHY_ID_BCM5714: return "5714"; case PHY_ID_BCM5780: return "5780"; case PHY_ID_BCM8002: return "8002/serdes"; case 0: return "serdes"; diff --git a/drivers/net/tg3.h b/drivers/net/tg3.h index 2e733c60bfa4..456ef2b3d0ef 100644 --- a/drivers/net/tg3.h +++ b/drivers/net/tg3.h @@ -137,6 +137,7 @@ #define ASIC_REV_5750 0x04 #define ASIC_REV_5752 0x06 #define ASIC_REV_5780 0x08 +#define ASIC_REV_5714 0x09 #define GET_CHIP_REV(CHIP_REV_ID) ((CHIP_REV_ID) >> 8) #define CHIPREV_5700_AX 0x70 #define CHIPREV_5700_BX 0x71 @@ -531,6 +532,8 @@ #define MAC_SERDES_CFG_EDGE_SELECT 0x00001000 #define MAC_SERDES_STAT 0x00000594 /* 0x598 --> 0x5b0 unused */ +#define SERDES_RX_CTRL 0x000005b0 /* 5780/5714 only */ +#define SERDES_RX_SIG_DETECT 0x00000400 #define SG_DIG_CTRL 0x000005b0 #define SG_DIG_USING_HW_AUTONEG 0x80000000 #define SG_DIG_SOFT_RESET 0x40000000 @@ -1329,6 +1332,8 @@ #define GRC_LCLCTRL_CLEARINT 0x00000002 #define GRC_LCLCTRL_SETINT 0x00000004 #define GRC_LCLCTRL_INT_ON_ATTN 0x00000008 +#define GRC_LCLCTRL_USE_SIG_DETECT 0x00000010 /* 5714/5780 only */ +#define GRC_LCLCTRL_USE_EXT_SIG_DETECT 0x00000020 /* 5714/5780 only */ #define GRC_LCLCTRL_GPIO_INPUT3 0x00000020 #define GRC_LCLCTRL_GPIO_OE3 0x00000040 #define GRC_LCLCTRL_GPIO_OUTPUT3 0x00000080 @@ -2175,6 +2180,7 @@ struct tg3 { TG3_FLG2_MII_SERDES) #define TG3_FLG2_PARALLEL_DETECT 0x01000000 #define TG3_FLG2_ICH_WORKAROUND 0x02000000 +#define TG3_FLG2_5780_CLASS 0x04000000 u32 split_mode_max_reqs; #define SPLIT_MODE_5704_MAX_REQ 3 @@ -2222,6 +2228,7 @@ struct tg3 { #define PHY_ID_BCM5705 0x600081a0 #define PHY_ID_BCM5750 0x60008180 #define PHY_ID_BCM5752 0x60008100 +#define PHY_ID_BCM5714 0x60008340 #define PHY_ID_BCM5780 0x60008350 #define PHY_ID_BCM8002 0x60010140 #define PHY_ID_INVALID 0xffffffff @@ -2246,8 +2253,8 @@ struct tg3 { (X) == PHY_ID_BCM5411 || (X) == PHY_ID_BCM5701 || \ (X) == PHY_ID_BCM5703 || (X) == PHY_ID_BCM5704 || \ (X) == PHY_ID_BCM5705 || (X) == PHY_ID_BCM5750 || \ - (X) == PHY_ID_BCM5752 || (X) == PHY_ID_BCM5780 || \ - (X) == PHY_ID_BCM8002) + (X) == PHY_ID_BCM5752 || (X) == PHY_ID_BCM5714 || \ + (X) == PHY_ID_BCM5780 || (X) == PHY_ID_BCM8002) struct tg3_hw_stats *hw_stats; dma_addr_t stats_mapping; diff --git a/include/linux/pci_ids.h b/include/linux/pci_ids.h index 71834f05504f..d6865bbd9ea4 100644 --- a/include/linux/pci_ids.h +++ b/include/linux/pci_ids.h @@ -2165,11 +2165,13 @@ #define PCI_DEVICE_ID_TIGON3_5721 0x1659 #define PCI_DEVICE_ID_TIGON3_5705M 0x165d #define PCI_DEVICE_ID_TIGON3_5705M_2 0x165e +#define PCI_DEVICE_ID_TIGON3_5714 0x1668 #define PCI_DEVICE_ID_TIGON3_5780 0x166a #define PCI_DEVICE_ID_TIGON3_5780S 0x166b #define PCI_DEVICE_ID_TIGON3_5705F 0x166e #define PCI_DEVICE_ID_TIGON3_5750 0x1676 #define PCI_DEVICE_ID_TIGON3_5751 0x1677 +#define PCI_DEVICE_ID_TIGON3_5715 0x1678 #define PCI_DEVICE_ID_TIGON3_5750M 0x167c #define PCI_DEVICE_ID_TIGON3_5751M 0x167d #define PCI_DEVICE_ID_TIGON3_5751F 0x167e -- cgit v1.2.3 From 48257c4f168e5d040394aeca4d37b59f68e0d36b Mon Sep 17 00:00:00 2001 From: Pantelis Antoniou Date: Fri, 28 Oct 2005 16:25:58 -0400 Subject: Add fs_enet ethernet network driver, for several embedded platforms. --- drivers/net/Kconfig | 1 + drivers/net/Makefile | 3 + drivers/net/fs_enet/Kconfig | 20 + drivers/net/fs_enet/Makefile | 10 + drivers/net/fs_enet/fs_enet-main.c | 1226 ++++++++++++++++++++++++++++++++++++ drivers/net/fs_enet/fs_enet-mii.c | 507 +++++++++++++++ drivers/net/fs_enet/fs_enet.h | 245 +++++++ drivers/net/fs_enet/mac-fcc.c | 578 +++++++++++++++++ drivers/net/fs_enet/mac-fec.c | 653 +++++++++++++++++++ drivers/net/fs_enet/mac-scc.c | 524 +++++++++++++++ drivers/net/fs_enet/mii-bitbang.c | 405 ++++++++++++ drivers/net/fs_enet/mii-fixed.c | 92 +++ include/linux/fs_enet_pd.h | 136 ++++ 13 files changed, 4400 insertions(+) create mode 100644 drivers/net/fs_enet/Kconfig create mode 100644 drivers/net/fs_enet/Makefile create mode 100644 drivers/net/fs_enet/fs_enet-main.c create mode 100644 drivers/net/fs_enet/fs_enet-mii.c create mode 100644 drivers/net/fs_enet/fs_enet.h create mode 100644 drivers/net/fs_enet/mac-fcc.c create mode 100644 drivers/net/fs_enet/mac-fec.c create mode 100644 drivers/net/fs_enet/mac-scc.c create mode 100644 drivers/net/fs_enet/mii-bitbang.c create mode 100644 drivers/net/fs_enet/mii-fixed.c create mode 100644 include/linux/fs_enet_pd.h (limited to 'include/linux') diff --git a/drivers/net/Kconfig b/drivers/net/Kconfig index 814a93b8c903..27732fdeeea4 100644 --- a/drivers/net/Kconfig +++ b/drivers/net/Kconfig @@ -1775,6 +1775,7 @@ config NE_H8300 controller on the Renesas H8/300 processor. source "drivers/net/fec_8xx/Kconfig" +source "drivers/net/fs_enet/Kconfig" endmenu diff --git a/drivers/net/Makefile b/drivers/net/Makefile index 1a84e0435f64..7c313cb341b8 100644 --- a/drivers/net/Makefile +++ b/drivers/net/Makefile @@ -203,3 +203,6 @@ obj-$(CONFIG_IRDA) += irda/ obj-$(CONFIG_ETRAX_ETHERNET) += cris/ obj-$(CONFIG_NETCONSOLE) += netconsole.o + +obj-$(CONFIG_FS_ENET) += fs_enet/ + diff --git a/drivers/net/fs_enet/Kconfig b/drivers/net/fs_enet/Kconfig new file mode 100644 index 000000000000..6aaee67dd4b7 --- /dev/null +++ b/drivers/net/fs_enet/Kconfig @@ -0,0 +1,20 @@ +config FS_ENET + tristate "Freescale Ethernet Driver" + depends on NET_ETHERNET && (CPM1 || CPM2) + select MII + +config FS_ENET_HAS_SCC + bool "Chip has an SCC usable for ethernet" + depends on FS_ENET && (CPM1 || CPM2) + default y + +config FS_ENET_HAS_FCC + bool "Chip has an FCC usable for ethernet" + depends on FS_ENET && CPM2 + default y + +config FS_ENET_HAS_FEC + bool "Chip has an FEC usable for ethernet" + depends on FS_ENET && CPM1 + default y + diff --git a/drivers/net/fs_enet/Makefile b/drivers/net/fs_enet/Makefile new file mode 100644 index 000000000000..d6dd3f2fb43e --- /dev/null +++ b/drivers/net/fs_enet/Makefile @@ -0,0 +1,10 @@ +# +# Makefile for the Freescale Ethernet controllers +# + +obj-$(CONFIG_FS_ENET) += fs_enet.o + +obj-$(CONFIG_8xx) += mac-fec.o mac-scc.o +obj-$(CONFIG_8260) += mac-fcc.o + +fs_enet-objs := fs_enet-main.o fs_enet-mii.o mii-bitbang.o mii-fixed.o diff --git a/drivers/net/fs_enet/fs_enet-main.c b/drivers/net/fs_enet/fs_enet-main.c new file mode 100644 index 000000000000..44fac7373289 --- /dev/null +++ b/drivers/net/fs_enet/fs_enet-main.c @@ -0,0 +1,1226 @@ +/* + * Combined Ethernet driver for Motorola MPC8xx and MPC82xx. + * + * Copyright (c) 2003 Intracom S.A. + * by Pantelis Antoniou + * + * 2005 (c) MontaVista Software, Inc. + * Vitaly Bordug + * + * Heavily based on original FEC driver by Dan Malek + * and modifications by Joakim Tjernlund + * + * This file is licensed under the terms of the GNU General Public License + * version 2. This program is licensed "as is" without any warranty of any + * kind, whether express or implied. + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include +#include + +#include +#include +#include + +#include "fs_enet.h" + +/*************************************************/ + +static char version[] __devinitdata = + DRV_MODULE_NAME ".c:v" DRV_MODULE_VERSION " (" DRV_MODULE_RELDATE ")" "\n"; + +MODULE_AUTHOR("Pantelis Antoniou "); +MODULE_DESCRIPTION("Freescale Ethernet Driver"); +MODULE_LICENSE("GPL"); +MODULE_VERSION(DRV_MODULE_VERSION); + +MODULE_PARM(fs_enet_debug, "i"); +MODULE_PARM_DESC(fs_enet_debug, + "Freescale bitmapped debugging message enable value"); + +int fs_enet_debug = -1; /* -1 == use FS_ENET_DEF_MSG_ENABLE as value */ + +static void fs_set_multicast_list(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + + (*fep->ops->set_multicast_list)(dev); +} + +/* NAPI receive function */ +static int fs_enet_rx_napi(struct net_device *dev, int *budget) +{ + struct fs_enet_private *fep = netdev_priv(dev); + const struct fs_platform_info *fpi = fep->fpi; + cbd_t *bdp; + struct sk_buff *skb, *skbn, *skbt; + int received = 0; + u16 pkt_len, sc; + int curidx; + int rx_work_limit = 0; /* pacify gcc */ + + rx_work_limit = min(dev->quota, *budget); + + if (!netif_running(dev)) + return 0; + + /* + * First, grab all of the stats for the incoming packet. + * These get messed up if we get called due to a busy condition. + */ + bdp = fep->cur_rx; + + /* clear RX status bits for napi*/ + (*fep->ops->napi_clear_rx_event)(dev); + + while (((sc = CBDR_SC(bdp)) & BD_ENET_RX_EMPTY) == 0) { + + curidx = bdp - fep->rx_bd_base; + + /* + * Since we have allocated space to hold a complete frame, + * the last indicator should be set. + */ + if ((sc & BD_ENET_RX_LAST) == 0) + printk(KERN_WARNING DRV_MODULE_NAME + ": %s rcv is not +last\n", + dev->name); + + /* + * Check for errors. + */ + if (sc & (BD_ENET_RX_LG | BD_ENET_RX_SH | BD_ENET_RX_CL | + BD_ENET_RX_NO | BD_ENET_RX_CR | BD_ENET_RX_OV)) { + fep->stats.rx_errors++; + /* Frame too long or too short. */ + if (sc & (BD_ENET_RX_LG | BD_ENET_RX_SH)) + fep->stats.rx_length_errors++; + /* Frame alignment */ + if (sc & (BD_ENET_RX_NO | BD_ENET_RX_CL)) + fep->stats.rx_frame_errors++; + /* CRC Error */ + if (sc & BD_ENET_RX_CR) + fep->stats.rx_crc_errors++; + /* FIFO overrun */ + if (sc & BD_ENET_RX_OV) + fep->stats.rx_crc_errors++; + + skb = fep->rx_skbuff[curidx]; + + dma_unmap_single(fep->dev, skb->data, + L1_CACHE_ALIGN(PKT_MAXBUF_SIZE), + DMA_FROM_DEVICE); + + skbn = skb; + + } else { + + /* napi, got packet but no quota */ + if (--rx_work_limit < 0) + break; + + skb = fep->rx_skbuff[curidx]; + + dma_unmap_single(fep->dev, skb->data, + L1_CACHE_ALIGN(PKT_MAXBUF_SIZE), + DMA_FROM_DEVICE); + + /* + * Process the incoming frame. + */ + fep->stats.rx_packets++; + pkt_len = CBDR_DATLEN(bdp) - 4; /* remove CRC */ + fep->stats.rx_bytes += pkt_len + 4; + + if (pkt_len <= fpi->rx_copybreak) { + /* +2 to make IP header L1 cache aligned */ + skbn = dev_alloc_skb(pkt_len + 2); + if (skbn != NULL) { + skb_reserve(skbn, 2); /* align IP header */ + memcpy(skbn->data, skb->data, pkt_len); + /* swap */ + skbt = skb; + skb = skbn; + skbn = skbt; + } + } else + skbn = dev_alloc_skb(ENET_RX_FRSIZE); + + if (skbn != NULL) { + skb->dev = dev; + skb_put(skb, pkt_len); /* Make room */ + skb->protocol = eth_type_trans(skb, dev); + received++; + netif_receive_skb(skb); + } else { + printk(KERN_WARNING DRV_MODULE_NAME + ": %s Memory squeeze, dropping packet.\n", + dev->name); + fep->stats.rx_dropped++; + skbn = skb; + } + } + + fep->rx_skbuff[curidx] = skbn; + CBDW_BUFADDR(bdp, dma_map_single(fep->dev, skbn->data, + L1_CACHE_ALIGN(PKT_MAXBUF_SIZE), + DMA_FROM_DEVICE)); + CBDW_DATLEN(bdp, 0); + CBDW_SC(bdp, (sc & ~BD_ENET_RX_STATS) | BD_ENET_RX_EMPTY); + + /* + * Update BD pointer to next entry. + */ + if ((sc & BD_ENET_RX_WRAP) == 0) + bdp++; + else + bdp = fep->rx_bd_base; + + (*fep->ops->rx_bd_done)(dev); + } + + fep->cur_rx = bdp; + + dev->quota -= received; + *budget -= received; + + if (rx_work_limit < 0) + return 1; /* not done */ + + /* done */ + netif_rx_complete(dev); + + (*fep->ops->napi_enable_rx)(dev); + + return 0; +} + +/* non NAPI receive function */ +static int fs_enet_rx_non_napi(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + const struct fs_platform_info *fpi = fep->fpi; + cbd_t *bdp; + struct sk_buff *skb, *skbn, *skbt; + int received = 0; + u16 pkt_len, sc; + int curidx; + /* + * First, grab all of the stats for the incoming packet. + * These get messed up if we get called due to a busy condition. + */ + bdp = fep->cur_rx; + + while (((sc = CBDR_SC(bdp)) & BD_ENET_RX_EMPTY) == 0) { + + curidx = bdp - fep->rx_bd_base; + + /* + * Since we have allocated space to hold a complete frame, + * the last indicator should be set. + */ + if ((sc & BD_ENET_RX_LAST) == 0) + printk(KERN_WARNING DRV_MODULE_NAME + ": %s rcv is not +last\n", + dev->name); + + /* + * Check for errors. + */ + if (sc & (BD_ENET_RX_LG | BD_ENET_RX_SH | BD_ENET_RX_CL | + BD_ENET_RX_NO | BD_ENET_RX_CR | BD_ENET_RX_OV)) { + fep->stats.rx_errors++; + /* Frame too long or too short. */ + if (sc & (BD_ENET_RX_LG | BD_ENET_RX_SH)) + fep->stats.rx_length_errors++; + /* Frame alignment */ + if (sc & (BD_ENET_RX_NO | BD_ENET_RX_CL)) + fep->stats.rx_frame_errors++; + /* CRC Error */ + if (sc & BD_ENET_RX_CR) + fep->stats.rx_crc_errors++; + /* FIFO overrun */ + if (sc & BD_ENET_RX_OV) + fep->stats.rx_crc_errors++; + + skb = fep->rx_skbuff[curidx]; + + dma_unmap_single(fep->dev, skb->data, + L1_CACHE_ALIGN(PKT_MAXBUF_SIZE), + DMA_FROM_DEVICE); + + skbn = skb; + + } else { + + skb = fep->rx_skbuff[curidx]; + + dma_unmap_single(fep->dev, skb->data, + L1_CACHE_ALIGN(PKT_MAXBUF_SIZE), + DMA_FROM_DEVICE); + + /* + * Process the incoming frame. + */ + fep->stats.rx_packets++; + pkt_len = CBDR_DATLEN(bdp) - 4; /* remove CRC */ + fep->stats.rx_bytes += pkt_len + 4; + + if (pkt_len <= fpi->rx_copybreak) { + /* +2 to make IP header L1 cache aligned */ + skbn = dev_alloc_skb(pkt_len + 2); + if (skbn != NULL) { + skb_reserve(skbn, 2); /* align IP header */ + memcpy(skbn->data, skb->data, pkt_len); + /* swap */ + skbt = skb; + skb = skbn; + skbn = skbt; + } + } else + skbn = dev_alloc_skb(ENET_RX_FRSIZE); + + if (skbn != NULL) { + skb->dev = dev; + skb_put(skb, pkt_len); /* Make room */ + skb->protocol = eth_type_trans(skb, dev); + received++; + netif_rx(skb); + } else { + printk(KERN_WARNING DRV_MODULE_NAME + ": %s Memory squeeze, dropping packet.\n", + dev->name); + fep->stats.rx_dropped++; + skbn = skb; + } + } + + fep->rx_skbuff[curidx] = skbn; + CBDW_BUFADDR(bdp, dma_map_single(fep->dev, skbn->data, + L1_CACHE_ALIGN(PKT_MAXBUF_SIZE), + DMA_FROM_DEVICE)); + CBDW_DATLEN(bdp, 0); + CBDW_SC(bdp, (sc & ~BD_ENET_RX_STATS) | BD_ENET_RX_EMPTY); + + /* + * Update BD pointer to next entry. + */ + if ((sc & BD_ENET_RX_WRAP) == 0) + bdp++; + else + bdp = fep->rx_bd_base; + + (*fep->ops->rx_bd_done)(dev); + } + + fep->cur_rx = bdp; + + return 0; +} + +static void fs_enet_tx(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + cbd_t *bdp; + struct sk_buff *skb; + int dirtyidx, do_wake, do_restart; + u16 sc; + + spin_lock(&fep->lock); + bdp = fep->dirty_tx; + + do_wake = do_restart = 0; + while (((sc = CBDR_SC(bdp)) & BD_ENET_TX_READY) == 0) { + + dirtyidx = bdp - fep->tx_bd_base; + + if (fep->tx_free == fep->tx_ring) + break; + + skb = fep->tx_skbuff[dirtyidx]; + + /* + * Check for errors. + */ + if (sc & (BD_ENET_TX_HB | BD_ENET_TX_LC | + BD_ENET_TX_RL | BD_ENET_TX_UN | BD_ENET_TX_CSL)) { + + if (sc & BD_ENET_TX_HB) /* No heartbeat */ + fep->stats.tx_heartbeat_errors++; + if (sc & BD_ENET_TX_LC) /* Late collision */ + fep->stats.tx_window_errors++; + if (sc & BD_ENET_TX_RL) /* Retrans limit */ + fep->stats.tx_aborted_errors++; + if (sc & BD_ENET_TX_UN) /* Underrun */ + fep->stats.tx_fifo_errors++; + if (sc & BD_ENET_TX_CSL) /* Carrier lost */ + fep->stats.tx_carrier_errors++; + + if (sc & (BD_ENET_TX_LC | BD_ENET_TX_RL | BD_ENET_TX_UN)) { + fep->stats.tx_errors++; + do_restart = 1; + } + } else + fep->stats.tx_packets++; + + if (sc & BD_ENET_TX_READY) + printk(KERN_WARNING DRV_MODULE_NAME + ": %s HEY! Enet xmit interrupt and TX_READY.\n", + dev->name); + + /* + * Deferred means some collisions occurred during transmit, + * but we eventually sent the packet OK. + */ + if (sc & BD_ENET_TX_DEF) + fep->stats.collisions++; + + /* unmap */ + dma_unmap_single(fep->dev, skb->data, skb->len, DMA_TO_DEVICE); + + /* + * Free the sk buffer associated with this last transmit. + */ + dev_kfree_skb_irq(skb); + fep->tx_skbuff[dirtyidx] = NULL; + + /* + * Update pointer to next buffer descriptor to be transmitted. + */ + if ((sc & BD_ENET_TX_WRAP) == 0) + bdp++; + else + bdp = fep->tx_bd_base; + + /* + * Since we have freed up a buffer, the ring is no longer + * full. + */ + if (!fep->tx_free++) + do_wake = 1; + } + + fep->dirty_tx = bdp; + + if (do_restart) + (*fep->ops->tx_restart)(dev); + + spin_unlock(&fep->lock); + + if (do_wake) + netif_wake_queue(dev); +} + +/* + * The interrupt handler. + * This is called from the MPC core interrupt. + */ +static irqreturn_t +fs_enet_interrupt(int irq, void *dev_id, struct pt_regs *regs) +{ + struct net_device *dev = dev_id; + struct fs_enet_private *fep; + const struct fs_platform_info *fpi; + u32 int_events; + u32 int_clr_events; + int nr, napi_ok; + int handled; + + fep = netdev_priv(dev); + fpi = fep->fpi; + + nr = 0; + while ((int_events = (*fep->ops->get_int_events)(dev)) != 0) { + + nr++; + + int_clr_events = int_events; + if (fpi->use_napi) + int_clr_events &= ~fep->ev_napi_rx; + + (*fep->ops->clear_int_events)(dev, int_clr_events); + + if (int_events & fep->ev_err) + (*fep->ops->ev_error)(dev, int_events); + + if (int_events & fep->ev_rx) { + if (!fpi->use_napi) + fs_enet_rx_non_napi(dev); + else { + napi_ok = netif_rx_schedule_prep(dev); + + (*fep->ops->napi_disable_rx)(dev); + (*fep->ops->clear_int_events)(dev, fep->ev_napi_rx); + + /* NOTE: it is possible for FCCs in NAPI mode */ + /* to submit a spurious interrupt while in poll */ + if (napi_ok) + __netif_rx_schedule(dev); + } + } + + if (int_events & fep->ev_tx) + fs_enet_tx(dev); + } + + handled = nr > 0; + return IRQ_RETVAL(handled); +} + +void fs_init_bds(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + cbd_t *bdp; + struct sk_buff *skb; + int i; + + fs_cleanup_bds(dev); + + fep->dirty_tx = fep->cur_tx = fep->tx_bd_base; + fep->tx_free = fep->tx_ring; + fep->cur_rx = fep->rx_bd_base; + + /* + * Initialize the receive buffer descriptors. + */ + for (i = 0, bdp = fep->rx_bd_base; i < fep->rx_ring; i++, bdp++) { + skb = dev_alloc_skb(ENET_RX_FRSIZE); + if (skb == NULL) { + printk(KERN_WARNING DRV_MODULE_NAME + ": %s Memory squeeze, unable to allocate skb\n", + dev->name); + break; + } + fep->rx_skbuff[i] = skb; + skb->dev = dev; + CBDW_BUFADDR(bdp, + dma_map_single(fep->dev, skb->data, + L1_CACHE_ALIGN(PKT_MAXBUF_SIZE), + DMA_FROM_DEVICE)); + CBDW_DATLEN(bdp, 0); /* zero */ + CBDW_SC(bdp, BD_ENET_RX_EMPTY | + ((i < fep->rx_ring - 1) ? 0 : BD_SC_WRAP)); + } + /* + * if we failed, fillup remainder + */ + for (; i < fep->rx_ring; i++, bdp++) { + fep->rx_skbuff[i] = NULL; + CBDW_SC(bdp, (i < fep->rx_ring - 1) ? 0 : BD_SC_WRAP); + } + + /* + * ...and the same for transmit. + */ + for (i = 0, bdp = fep->tx_bd_base; i < fep->tx_ring; i++, bdp++) { + fep->tx_skbuff[i] = NULL; + CBDW_BUFADDR(bdp, 0); + CBDW_DATLEN(bdp, 0); + CBDW_SC(bdp, (i < fep->tx_ring - 1) ? 0 : BD_SC_WRAP); + } +} + +void fs_cleanup_bds(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + struct sk_buff *skb; + int i; + + /* + * Reset SKB transmit buffers. + */ + for (i = 0; i < fep->tx_ring; i++) { + if ((skb = fep->tx_skbuff[i]) == NULL) + continue; + + /* unmap */ + dma_unmap_single(fep->dev, skb->data, skb->len, DMA_TO_DEVICE); + + fep->tx_skbuff[i] = NULL; + dev_kfree_skb(skb); + } + + /* + * Reset SKB receive buffers + */ + for (i = 0; i < fep->rx_ring; i++) { + if ((skb = fep->rx_skbuff[i]) == NULL) + continue; + + /* unmap */ + dma_unmap_single(fep->dev, skb->data, + L1_CACHE_ALIGN(PKT_MAXBUF_SIZE), + DMA_FROM_DEVICE); + + fep->rx_skbuff[i] = NULL; + + dev_kfree_skb(skb); + } +} + +/**********************************************************************************/ + +static int fs_enet_start_xmit(struct sk_buff *skb, struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + cbd_t *bdp; + int curidx; + u16 sc; + unsigned long flags; + + spin_lock_irqsave(&fep->tx_lock, flags); + + /* + * Fill in a Tx ring entry + */ + bdp = fep->cur_tx; + + if (!fep->tx_free || (CBDR_SC(bdp) & BD_ENET_TX_READY)) { + netif_stop_queue(dev); + spin_unlock_irqrestore(&fep->tx_lock, flags); + + /* + * Ooops. All transmit buffers are full. Bail out. + * This should not happen, since the tx queue should be stopped. + */ + printk(KERN_WARNING DRV_MODULE_NAME + ": %s tx queue full!.\n", dev->name); + return NETDEV_TX_BUSY; + } + + curidx = bdp - fep->tx_bd_base; + /* + * Clear all of the status flags. + */ + CBDC_SC(bdp, BD_ENET_TX_STATS); + + /* + * Save skb pointer. + */ + fep->tx_skbuff[curidx] = skb; + + fep->stats.tx_bytes += skb->len; + + /* + * Push the data cache so the CPM does not get stale memory data. + */ + CBDW_BUFADDR(bdp, dma_map_single(fep->dev, + skb->data, skb->len, DMA_TO_DEVICE)); + CBDW_DATLEN(bdp, skb->len); + + dev->trans_start = jiffies; + + /* + * If this was the last BD in the ring, start at the beginning again. + */ + if ((CBDR_SC(bdp) & BD_ENET_TX_WRAP) == 0) + fep->cur_tx++; + else + fep->cur_tx = fep->tx_bd_base; + + if (!--fep->tx_free) + netif_stop_queue(dev); + + /* Trigger transmission start */ + sc = BD_ENET_TX_READY | BD_ENET_TX_INTR | + BD_ENET_TX_LAST | BD_ENET_TX_TC; + + /* note that while FEC does not have this bit + * it marks it as available for software use + * yay for hw reuse :) */ + if (skb->len <= 60) + sc |= BD_ENET_TX_PAD; + CBDS_SC(bdp, sc); + + (*fep->ops->tx_kickstart)(dev); + + spin_unlock_irqrestore(&fep->tx_lock, flags); + + return NETDEV_TX_OK; +} + +static int fs_request_irq(struct net_device *dev, int irq, const char *name, + irqreturn_t (*irqf)(int irq, void *dev_id, struct pt_regs *regs)) +{ + struct fs_enet_private *fep = netdev_priv(dev); + + (*fep->ops->pre_request_irq)(dev, irq); + return request_irq(irq, irqf, SA_SHIRQ, name, dev); +} + +static void fs_free_irq(struct net_device *dev, int irq) +{ + struct fs_enet_private *fep = netdev_priv(dev); + + free_irq(irq, dev); + (*fep->ops->post_free_irq)(dev, irq); +} + +/**********************************************************************************/ + +/* This interrupt occurs when the PHY detects a link change. */ +static irqreturn_t +fs_mii_link_interrupt(int irq, void *dev_id, struct pt_regs *regs) +{ + struct net_device *dev = dev_id; + struct fs_enet_private *fep; + const struct fs_platform_info *fpi; + + fep = netdev_priv(dev); + fpi = fep->fpi; + + /* + * Acknowledge the interrupt if possible. If we have not + * found the PHY yet we can't process or acknowledge the + * interrupt now. Instead we ignore this interrupt for now, + * which we can do since it is edge triggered. It will be + * acknowledged later by fs_enet_open(). + */ + if (!fep->phy) + return IRQ_NONE; + + fs_mii_ack_int(dev); + fs_mii_link_status_change_check(dev, 0); + + return IRQ_HANDLED; +} + +static void fs_timeout(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + unsigned long flags; + int wake = 0; + + fep->stats.tx_errors++; + + spin_lock_irqsave(&fep->lock, flags); + + if (dev->flags & IFF_UP) { + (*fep->ops->stop)(dev); + (*fep->ops->restart)(dev); + } + + wake = fep->tx_free && !(CBDR_SC(fep->cur_tx) & BD_ENET_TX_READY); + spin_unlock_irqrestore(&fep->lock, flags); + + if (wake) + netif_wake_queue(dev); +} + +static int fs_enet_open(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + const struct fs_platform_info *fpi = fep->fpi; + int r; + + /* Install our interrupt handler. */ + r = fs_request_irq(dev, fep->interrupt, "fs_enet-mac", fs_enet_interrupt); + if (r != 0) { + printk(KERN_ERR DRV_MODULE_NAME + ": %s Could not allocate FEC IRQ!", dev->name); + return -EINVAL; + } + + /* Install our phy interrupt handler */ + if (fpi->phy_irq != -1) { + + r = fs_request_irq(dev, fpi->phy_irq, "fs_enet-phy", fs_mii_link_interrupt); + if (r != 0) { + printk(KERN_ERR DRV_MODULE_NAME + ": %s Could not allocate PHY IRQ!", dev->name); + fs_free_irq(dev, fep->interrupt); + return -EINVAL; + } + } + + fs_mii_startup(dev); + netif_carrier_off(dev); + fs_mii_link_status_change_check(dev, 1); + + return 0; +} + +static int fs_enet_close(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + const struct fs_platform_info *fpi = fep->fpi; + unsigned long flags; + + netif_stop_queue(dev); + netif_carrier_off(dev); + fs_mii_shutdown(dev); + + spin_lock_irqsave(&fep->lock, flags); + (*fep->ops->stop)(dev); + spin_unlock_irqrestore(&fep->lock, flags); + + /* release any irqs */ + if (fpi->phy_irq != -1) + fs_free_irq(dev, fpi->phy_irq); + fs_free_irq(dev, fep->interrupt); + + return 0; +} + +static struct net_device_stats *fs_enet_get_stats(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + return &fep->stats; +} + +/*************************************************************************/ + +static void fs_get_drvinfo(struct net_device *dev, + struct ethtool_drvinfo *info) +{ + strcpy(info->driver, DRV_MODULE_NAME); + strcpy(info->version, DRV_MODULE_VERSION); +} + +static int fs_get_regs_len(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + + return (*fep->ops->get_regs_len)(dev); +} + +static void fs_get_regs(struct net_device *dev, struct ethtool_regs *regs, + void *p) +{ + struct fs_enet_private *fep = netdev_priv(dev); + unsigned long flags; + int r, len; + + len = regs->len; + + spin_lock_irqsave(&fep->lock, flags); + r = (*fep->ops->get_regs)(dev, p, &len); + spin_unlock_irqrestore(&fep->lock, flags); + + if (r == 0) + regs->version = 0; +} + +static int fs_get_settings(struct net_device *dev, struct ethtool_cmd *cmd) +{ + struct fs_enet_private *fep = netdev_priv(dev); + unsigned long flags; + int rc; + + spin_lock_irqsave(&fep->lock, flags); + rc = mii_ethtool_gset(&fep->mii_if, cmd); + spin_unlock_irqrestore(&fep->lock, flags); + + return rc; +} + +static int fs_set_settings(struct net_device *dev, struct ethtool_cmd *cmd) +{ + struct fs_enet_private *fep = netdev_priv(dev); + unsigned long flags; + int rc; + + spin_lock_irqsave(&fep->lock, flags); + rc = mii_ethtool_sset(&fep->mii_if, cmd); + spin_unlock_irqrestore(&fep->lock, flags); + + return rc; +} + +static int fs_nway_reset(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + return mii_nway_restart(&fep->mii_if); +} + +static u32 fs_get_msglevel(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + return fep->msg_enable; +} + +static void fs_set_msglevel(struct net_device *dev, u32 value) +{ + struct fs_enet_private *fep = netdev_priv(dev); + fep->msg_enable = value; +} + +static struct ethtool_ops fs_ethtool_ops = { + .get_drvinfo = fs_get_drvinfo, + .get_regs_len = fs_get_regs_len, + .get_settings = fs_get_settings, + .set_settings = fs_set_settings, + .nway_reset = fs_nway_reset, + .get_link = ethtool_op_get_link, + .get_msglevel = fs_get_msglevel, + .set_msglevel = fs_set_msglevel, + .get_tx_csum = ethtool_op_get_tx_csum, + .set_tx_csum = ethtool_op_set_tx_csum, /* local! */ + .get_sg = ethtool_op_get_sg, + .set_sg = ethtool_op_set_sg, + .get_regs = fs_get_regs, +}; + +static int fs_ioctl(struct net_device *dev, struct ifreq *rq, int cmd) +{ + struct fs_enet_private *fep = netdev_priv(dev); + struct mii_ioctl_data *mii = (struct mii_ioctl_data *)&rq->ifr_data; + unsigned long flags; + int rc; + + if (!netif_running(dev)) + return -EINVAL; + + spin_lock_irqsave(&fep->lock, flags); + rc = generic_mii_ioctl(&fep->mii_if, mii, cmd, NULL); + spin_unlock_irqrestore(&fep->lock, flags); + return rc; +} + +extern int fs_mii_connect(struct net_device *dev); +extern void fs_mii_disconnect(struct net_device *dev); + +static struct net_device *fs_init_instance(struct device *dev, + const struct fs_platform_info *fpi) +{ + struct net_device *ndev = NULL; + struct fs_enet_private *fep = NULL; + int privsize, i, r, err = 0, registered = 0; + + /* guard */ + if ((unsigned int)fpi->fs_no >= FS_MAX_INDEX) + return ERR_PTR(-EINVAL); + + privsize = sizeof(*fep) + (sizeof(struct sk_buff **) * + (fpi->rx_ring + fpi->tx_ring)); + + ndev = alloc_etherdev(privsize); + if (!ndev) { + err = -ENOMEM; + goto err; + } + SET_MODULE_OWNER(ndev); + + fep = netdev_priv(ndev); + memset(fep, 0, privsize); /* clear everything */ + + fep->dev = dev; + dev_set_drvdata(dev, ndev); + fep->fpi = fpi; + if (fpi->init_ioports) + fpi->init_ioports(); + +#ifdef CONFIG_FS_ENET_HAS_FEC + if (fs_get_fec_index(fpi->fs_no) >= 0) + fep->ops = &fs_fec_ops; +#endif + +#ifdef CONFIG_FS_ENET_HAS_SCC + if (fs_get_scc_index(fpi->fs_no) >=0 ) + fep->ops = &fs_scc_ops; +#endif + +#ifdef CONFIG_FS_ENET_HAS_FCC + if (fs_get_fcc_index(fpi->fs_no) >= 0) + fep->ops = &fs_fcc_ops; +#endif + + if (fep->ops == NULL) { + printk(KERN_ERR DRV_MODULE_NAME + ": %s No matching ops found (%d).\n", + ndev->name, fpi->fs_no); + err = -EINVAL; + goto err; + } + + r = (*fep->ops->setup_data)(ndev); + if (r != 0) { + printk(KERN_ERR DRV_MODULE_NAME + ": %s setup_data failed\n", + ndev->name); + err = r; + goto err; + } + + /* point rx_skbuff, tx_skbuff */ + fep->rx_skbuff = (struct sk_buff **)&fep[1]; + fep->tx_skbuff = fep->rx_skbuff + fpi->rx_ring; + + /* init locks */ + spin_lock_init(&fep->lock); + spin_lock_init(&fep->tx_lock); + + /* + * Set the Ethernet address. + */ + for (i = 0; i < 6; i++) + ndev->dev_addr[i] = fpi->macaddr[i]; + + r = (*fep->ops->allocate_bd)(ndev); + + if (fep->ring_base == NULL) { + printk(KERN_ERR DRV_MODULE_NAME + ": %s buffer descriptor alloc failed (%d).\n", ndev->name, r); + err = r; + goto err; + } + + /* + * Set receive and transmit descriptor base. + */ + fep->rx_bd_base = fep->ring_base; + fep->tx_bd_base = fep->rx_bd_base + fpi->rx_ring; + + /* initialize ring size variables */ + fep->tx_ring = fpi->tx_ring; + fep->rx_ring = fpi->rx_ring; + + /* + * The FEC Ethernet specific entries in the device structure. + */ + ndev->open = fs_enet_open; + ndev->hard_start_xmit = fs_enet_start_xmit; + ndev->tx_timeout = fs_timeout; + ndev->watchdog_timeo = 2 * HZ; + ndev->stop = fs_enet_close; + ndev->get_stats = fs_enet_get_stats; + ndev->set_multicast_list = fs_set_multicast_list; + if (fpi->use_napi) { + ndev->poll = fs_enet_rx_napi; + ndev->weight = fpi->napi_weight; + } + ndev->ethtool_ops = &fs_ethtool_ops; + ndev->do_ioctl = fs_ioctl; + + init_timer(&fep->phy_timer_list); + + netif_carrier_off(ndev); + + err = register_netdev(ndev); + if (err != 0) { + printk(KERN_ERR DRV_MODULE_NAME + ": %s register_netdev failed.\n", ndev->name); + goto err; + } + registered = 1; + + err = fs_mii_connect(ndev); + if (err != 0) { + printk(KERN_ERR DRV_MODULE_NAME + ": %s fs_mii_connect failed.\n", ndev->name); + goto err; + } + + return ndev; + + err: + if (ndev != NULL) { + + if (registered) + unregister_netdev(ndev); + + if (fep != NULL) { + (*fep->ops->free_bd)(ndev); + (*fep->ops->cleanup_data)(ndev); + } + + free_netdev(ndev); + } + + dev_set_drvdata(dev, NULL); + + return ERR_PTR(err); +} + +static int fs_cleanup_instance(struct net_device *ndev) +{ + struct fs_enet_private *fep; + const struct fs_platform_info *fpi; + struct device *dev; + + if (ndev == NULL) + return -EINVAL; + + fep = netdev_priv(ndev); + if (fep == NULL) + return -EINVAL; + + fpi = fep->fpi; + + fs_mii_disconnect(ndev); + + unregister_netdev(ndev); + + dma_free_coherent(fep->dev, (fpi->tx_ring + fpi->rx_ring) * sizeof(cbd_t), + fep->ring_base, fep->ring_mem_addr); + + /* reset it */ + (*fep->ops->cleanup_data)(ndev); + + dev = fep->dev; + if (dev != NULL) { + dev_set_drvdata(dev, NULL); + fep->dev = NULL; + } + + free_netdev(ndev); + + return 0; +} + +/**************************************************************************************/ + +/* handy pointer to the immap */ +void *fs_enet_immap = NULL; + +static int setup_immap(void) +{ + phys_addr_t paddr = 0; + unsigned long size = 0; + +#ifdef CONFIG_CPM1 + paddr = IMAP_ADDR; + size = 0x10000; /* map 64K */ +#endif + +#ifdef CONFIG_CPM2 + paddr = CPM_MAP_ADDR; + size = 0x40000; /* map 256 K */ +#endif + fs_enet_immap = ioremap(paddr, size); + if (fs_enet_immap == NULL) + return -EBADF; /* XXX ahem; maybe just BUG_ON? */ + + return 0; +} + +static void cleanup_immap(void) +{ + if (fs_enet_immap != NULL) { + iounmap(fs_enet_immap); + fs_enet_immap = NULL; + } +} + +/**************************************************************************************/ + +static int __devinit fs_enet_probe(struct device *dev) +{ + struct net_device *ndev; + + /* no fixup - no device */ + if (dev->platform_data == NULL) { + printk(KERN_INFO "fs_enet: " + "probe called with no platform data; " + "remove unused devices\n"); + return -ENODEV; + } + + ndev = fs_init_instance(dev, dev->platform_data); + if (IS_ERR(ndev)) + return PTR_ERR(ndev); + return 0; +} + +static int fs_enet_remove(struct device *dev) +{ + return fs_cleanup_instance(dev_get_drvdata(dev)); +} + +static struct device_driver fs_enet_fec_driver = { + .name = "fsl-cpm-fec", + .bus = &platform_bus_type, + .probe = fs_enet_probe, + .remove = fs_enet_remove, +#ifdef CONFIG_PM +/* .suspend = fs_enet_suspend, TODO */ +/* .resume = fs_enet_resume, TODO */ +#endif +}; + +static struct device_driver fs_enet_scc_driver = { + .name = "fsl-cpm-scc", + .bus = &platform_bus_type, + .probe = fs_enet_probe, + .remove = fs_enet_remove, +#ifdef CONFIG_PM +/* .suspend = fs_enet_suspend, TODO */ +/* .resume = fs_enet_resume, TODO */ +#endif +}; + +static struct device_driver fs_enet_fcc_driver = { + .name = "fsl-cpm-fcc", + .bus = &platform_bus_type, + .probe = fs_enet_probe, + .remove = fs_enet_remove, +#ifdef CONFIG_PM +/* .suspend = fs_enet_suspend, TODO */ +/* .resume = fs_enet_resume, TODO */ +#endif +}; + +static int __init fs_init(void) +{ + int r; + + printk(KERN_INFO + "%s", version); + + r = setup_immap(); + if (r != 0) + return r; + r = driver_register(&fs_enet_fec_driver); + if (r != 0) + goto err; + + r = driver_register(&fs_enet_fcc_driver); + if (r != 0) + goto err; + + r = driver_register(&fs_enet_scc_driver); + if (r != 0) + goto err; + + return 0; +err: + cleanup_immap(); + return r; + +} + +static void __exit fs_cleanup(void) +{ + driver_unregister(&fs_enet_fec_driver); + driver_unregister(&fs_enet_fcc_driver); + driver_unregister(&fs_enet_scc_driver); + cleanup_immap(); +} + +/**************************************************************************************/ + +module_init(fs_init); +module_exit(fs_cleanup); diff --git a/drivers/net/fs_enet/fs_enet-mii.c b/drivers/net/fs_enet/fs_enet-mii.c new file mode 100644 index 000000000000..c6770377ef87 --- /dev/null +++ b/drivers/net/fs_enet/fs_enet-mii.c @@ -0,0 +1,507 @@ +/* + * Combined Ethernet driver for Motorola MPC8xx and MPC82xx. + * + * Copyright (c) 2003 Intracom S.A. + * by Pantelis Antoniou + * + * 2005 (c) MontaVista Software, Inc. + * Vitaly Bordug + * + * Heavily based on original FEC driver by Dan Malek + * and modifications by Joakim Tjernlund + * + * This file is licensed under the terms of the GNU General Public License + * version 2. This program is licensed "as is" without any warranty of any + * kind, whether express or implied. + */ + + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include +#include +#include + +#include "fs_enet.h" + +/*************************************************/ + +/* + * Generic PHY support. + * Should work for all PHYs, but link change is detected by polling + */ + +static void generic_timer_callback(unsigned long data) +{ + struct net_device *dev = (struct net_device *)data; + struct fs_enet_private *fep = netdev_priv(dev); + + fep->phy_timer_list.expires = jiffies + HZ / 2; + + add_timer(&fep->phy_timer_list); + + fs_mii_link_status_change_check(dev, 0); +} + +static void generic_startup(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + + fep->phy_timer_list.expires = jiffies + HZ / 2; /* every 500ms */ + fep->phy_timer_list.data = (unsigned long)dev; + fep->phy_timer_list.function = generic_timer_callback; + add_timer(&fep->phy_timer_list); +} + +static void generic_shutdown(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + + del_timer_sync(&fep->phy_timer_list); +} + +/* ------------------------------------------------------------------------- */ +/* The Davicom DM9161 is used on the NETTA board */ + +/* register definitions */ + +#define MII_DM9161_ANAR 4 /* Aux. Config Register */ +#define MII_DM9161_ACR 16 /* Aux. Config Register */ +#define MII_DM9161_ACSR 17 /* Aux. Config/Status Register */ +#define MII_DM9161_10TCSR 18 /* 10BaseT Config/Status Reg. */ +#define MII_DM9161_INTR 21 /* Interrupt Register */ +#define MII_DM9161_RECR 22 /* Receive Error Counter Reg. */ +#define MII_DM9161_DISCR 23 /* Disconnect Counter Register */ + +static void dm9161_startup(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + + fs_mii_write(dev, fep->mii_if.phy_id, MII_DM9161_INTR, 0x0000); + /* Start autonegotiation */ + fs_mii_write(dev, fep->mii_if.phy_id, MII_BMCR, 0x1200); + + set_current_state(TASK_UNINTERRUPTIBLE); + schedule_timeout(HZ*8); +} + +static void dm9161_ack_int(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + + fs_mii_read(dev, fep->mii_if.phy_id, MII_DM9161_INTR); +} + +static void dm9161_shutdown(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + + fs_mii_write(dev, fep->mii_if.phy_id, MII_DM9161_INTR, 0x0f00); +} + +/**********************************************************************************/ + +static const struct phy_info phy_info[] = { + { + .id = 0x00181b88, + .name = "DM9161", + .startup = dm9161_startup, + .ack_int = dm9161_ack_int, + .shutdown = dm9161_shutdown, + }, { + .id = 0, + .name = "GENERIC", + .startup = generic_startup, + .shutdown = generic_shutdown, + }, +}; + +/**********************************************************************************/ + +static int phy_id_detect(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + const struct fs_platform_info *fpi = fep->fpi; + struct fs_enet_mii_bus *bus = fep->mii_bus; + int i, r, start, end, phytype, physubtype; + const struct phy_info *phy; + int phy_hwid, phy_id; + + phy_hwid = -1; + fep->phy = NULL; + + /* auto-detect? */ + if (fpi->phy_addr == -1) { + start = 1; + end = 32; + } else { /* direct */ + start = fpi->phy_addr; + end = start + 1; + } + + for (phy_id = start; phy_id < end; phy_id++) { + /* skip already used phy addresses on this bus */ + if (bus->usage_map & (1 << phy_id)) + continue; + r = fs_mii_read(dev, phy_id, MII_PHYSID1); + if (r == -1 || (phytype = (r & 0xffff)) == 0xffff) + continue; + r = fs_mii_read(dev, phy_id, MII_PHYSID2); + if (r == -1 || (physubtype = (r & 0xffff)) == 0xffff) + continue; + phy_hwid = (phytype << 16) | physubtype; + if (phy_hwid != -1) + break; + } + + if (phy_hwid == -1) { + printk(KERN_ERR DRV_MODULE_NAME + ": %s No PHY detected! range=0x%02x-0x%02x\n", + dev->name, start, end); + return -1; + } + + for (i = 0, phy = phy_info; i < ARRAY_SIZE(phy_info); i++, phy++) + if (phy->id == (phy_hwid >> 4) || phy->id == 0) + break; + + if (i >= ARRAY_SIZE(phy_info)) { + printk(KERN_ERR DRV_MODULE_NAME + ": %s PHY id 0x%08x is not supported!\n", + dev->name, phy_hwid); + return -1; + } + + fep->phy = phy; + + /* mark this address as used */ + bus->usage_map |= (1 << phy_id); + + printk(KERN_INFO DRV_MODULE_NAME + ": %s Phy @ 0x%x, type %s (0x%08x)%s\n", + dev->name, phy_id, fep->phy->name, phy_hwid, + fpi->phy_addr == -1 ? " (auto-detected)" : ""); + + return phy_id; +} + +void fs_mii_startup(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + + if (fep->phy->startup) + (*fep->phy->startup) (dev); +} + +void fs_mii_shutdown(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + + if (fep->phy->shutdown) + (*fep->phy->shutdown) (dev); +} + +void fs_mii_ack_int(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + + if (fep->phy->ack_int) + (*fep->phy->ack_int) (dev); +} + +#define MII_LINK 0x0001 +#define MII_HALF 0x0002 +#define MII_FULL 0x0004 +#define MII_BASE4 0x0008 +#define MII_10M 0x0010 +#define MII_100M 0x0020 +#define MII_1G 0x0040 +#define MII_10G 0x0080 + +/* return full mii info at one gulp, with a usable form */ +static unsigned int mii_full_status(struct mii_if_info *mii) +{ + unsigned int status; + int bmsr, adv, lpa, neg; + struct fs_enet_private* fep = netdev_priv(mii->dev); + + /* first, a dummy read, needed to latch some MII phys */ + (void)mii->mdio_read(mii->dev, mii->phy_id, MII_BMSR); + bmsr = mii->mdio_read(mii->dev, mii->phy_id, MII_BMSR); + + /* no link */ + if ((bmsr & BMSR_LSTATUS) == 0) + return 0; + + status = MII_LINK; + + /* Lets look what ANEG says if it's supported - otherwize we shall + take the right values from the platform info*/ + if(!mii->force_media) { + /* autoneg not completed; don't bother */ + if ((bmsr & BMSR_ANEGCOMPLETE) == 0) + return 0; + + adv = (*mii->mdio_read)(mii->dev, mii->phy_id, MII_ADVERTISE); + lpa = (*mii->mdio_read)(mii->dev, mii->phy_id, MII_LPA); + + neg = lpa & adv; + } else { + neg = fep->fpi->bus_info->lpa; + } + + if (neg & LPA_100FULL) + status |= MII_FULL | MII_100M; + else if (neg & LPA_100BASE4) + status |= MII_FULL | MII_BASE4 | MII_100M; + else if (neg & LPA_100HALF) + status |= MII_HALF | MII_100M; + else if (neg & LPA_10FULL) + status |= MII_FULL | MII_10M; + else + status |= MII_HALF | MII_10M; + + return status; +} + +void fs_mii_link_status_change_check(struct net_device *dev, int init_media) +{ + struct fs_enet_private *fep = netdev_priv(dev); + struct mii_if_info *mii = &fep->mii_if; + unsigned int mii_status; + int ok_to_print, link, duplex, speed; + unsigned long flags; + + ok_to_print = netif_msg_link(fep); + + mii_status = mii_full_status(mii); + + if (!init_media && mii_status == fep->last_mii_status) + return; + + fep->last_mii_status = mii_status; + + link = !!(mii_status & MII_LINK); + duplex = !!(mii_status & MII_FULL); + speed = (mii_status & MII_100M) ? 100 : 10; + + if (link == 0) { + netif_carrier_off(mii->dev); + netif_stop_queue(dev); + if (!init_media) { + spin_lock_irqsave(&fep->lock, flags); + (*fep->ops->stop)(dev); + spin_unlock_irqrestore(&fep->lock, flags); + } + + if (ok_to_print) + printk(KERN_INFO "%s: link down\n", mii->dev->name); + + } else { + + mii->full_duplex = duplex; + + netif_carrier_on(mii->dev); + + spin_lock_irqsave(&fep->lock, flags); + fep->duplex = duplex; + fep->speed = speed; + (*fep->ops->restart)(dev); + spin_unlock_irqrestore(&fep->lock, flags); + + netif_start_queue(dev); + + if (ok_to_print) + printk(KERN_INFO "%s: link up, %dMbps, %s-duplex\n", + dev->name, speed, duplex ? "full" : "half"); + } +} + +/**********************************************************************************/ + +int fs_mii_read(struct net_device *dev, int phy_id, int location) +{ + struct fs_enet_private *fep = netdev_priv(dev); + struct fs_enet_mii_bus *bus = fep->mii_bus; + + unsigned long flags; + int ret; + + spin_lock_irqsave(&bus->mii_lock, flags); + ret = (*bus->mii_read)(bus, phy_id, location); + spin_unlock_irqrestore(&bus->mii_lock, flags); + + return ret; +} + +void fs_mii_write(struct net_device *dev, int phy_id, int location, int value) +{ + struct fs_enet_private *fep = netdev_priv(dev); + struct fs_enet_mii_bus *bus = fep->mii_bus; + unsigned long flags; + + spin_lock_irqsave(&bus->mii_lock, flags); + (*bus->mii_write)(bus, phy_id, location, value); + spin_unlock_irqrestore(&bus->mii_lock, flags); +} + +/*****************************************************************************/ + +/* list of all registered mii buses */ +static LIST_HEAD(fs_mii_bus_list); + +static struct fs_enet_mii_bus *lookup_bus(int method, int id) +{ + struct list_head *ptr; + struct fs_enet_mii_bus *bus; + + list_for_each(ptr, &fs_mii_bus_list) { + bus = list_entry(ptr, struct fs_enet_mii_bus, list); + if (bus->bus_info->method == method && + bus->bus_info->id == id) + return bus; + } + return NULL; +} + +static struct fs_enet_mii_bus *create_bus(const struct fs_mii_bus_info *bi) +{ + struct fs_enet_mii_bus *bus; + int ret = 0; + + bus = kmalloc(sizeof(*bus), GFP_KERNEL); + if (bus == NULL) { + ret = -ENOMEM; + goto err; + } + memset(bus, 0, sizeof(*bus)); + spin_lock_init(&bus->mii_lock); + bus->bus_info = bi; + bus->refs = 0; + bus->usage_map = 0; + + /* perform initialization */ + switch (bi->method) { + + case fsmii_fixed: + ret = fs_mii_fixed_init(bus); + if (ret != 0) + goto err; + break; + + case fsmii_bitbang: + ret = fs_mii_bitbang_init(bus); + if (ret != 0) + goto err; + break; +#ifdef CONFIG_FS_ENET_HAS_FEC + case fsmii_fec: + ret = fs_mii_fec_init(bus); + if (ret != 0) + goto err; + break; +#endif + default: + ret = -EINVAL; + goto err; + } + + list_add(&bus->list, &fs_mii_bus_list); + + return bus; + +err: + if (bus) + kfree(bus); + return ERR_PTR(ret); +} + +static void destroy_bus(struct fs_enet_mii_bus *bus) +{ + /* remove from bus list */ + list_del(&bus->list); + + /* nothing more needed */ + kfree(bus); +} + +int fs_mii_connect(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + const struct fs_platform_info *fpi = fep->fpi; + struct fs_enet_mii_bus *bus = NULL; + + /* check method validity */ + switch (fpi->bus_info->method) { + case fsmii_fixed: + case fsmii_bitbang: + break; +#ifdef CONFIG_FS_ENET_HAS_FEC + case fsmii_fec: + break; +#endif + default: + printk(KERN_ERR DRV_MODULE_NAME + ": %s Unknown MII bus method (%d)!\n", + dev->name, fpi->bus_info->method); + return -EINVAL; + } + + bus = lookup_bus(fpi->bus_info->method, fpi->bus_info->id); + + /* if not found create new bus */ + if (bus == NULL) { + bus = create_bus(fpi->bus_info); + if (IS_ERR(bus)) { + printk(KERN_ERR DRV_MODULE_NAME + ": %s MII bus creation failure!\n", dev->name); + return PTR_ERR(bus); + } + } + + bus->refs++; + + fep->mii_bus = bus; + + fep->mii_if.dev = dev; + fep->mii_if.phy_id_mask = 0x1f; + fep->mii_if.reg_num_mask = 0x1f; + fep->mii_if.mdio_read = fs_mii_read; + fep->mii_if.mdio_write = fs_mii_write; + fep->mii_if.force_media = fpi->bus_info->disable_aneg; + fep->mii_if.phy_id = phy_id_detect(dev); + + return 0; +} + +void fs_mii_disconnect(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + struct fs_enet_mii_bus *bus = NULL; + + bus = fep->mii_bus; + fep->mii_bus = NULL; + + if (--bus->refs <= 0) + destroy_bus(bus); +} diff --git a/drivers/net/fs_enet/fs_enet.h b/drivers/net/fs_enet/fs_enet.h new file mode 100644 index 000000000000..1105543b9d88 --- /dev/null +++ b/drivers/net/fs_enet/fs_enet.h @@ -0,0 +1,245 @@ +#ifndef FS_ENET_H +#define FS_ENET_H + +#include +#include +#include +#include +#include + +#include + +#include + +#ifdef CONFIG_CPM1 +#include +#endif + +#ifdef CONFIG_CPM2 +#include +#endif + +/* hw driver ops */ +struct fs_ops { + int (*setup_data)(struct net_device *dev); + int (*allocate_bd)(struct net_device *dev); + void (*free_bd)(struct net_device *dev); + void (*cleanup_data)(struct net_device *dev); + void (*set_multicast_list)(struct net_device *dev); + void (*restart)(struct net_device *dev); + void (*stop)(struct net_device *dev); + void (*pre_request_irq)(struct net_device *dev, int irq); + void (*post_free_irq)(struct net_device *dev, int irq); + void (*napi_clear_rx_event)(struct net_device *dev); + void (*napi_enable_rx)(struct net_device *dev); + void (*napi_disable_rx)(struct net_device *dev); + void (*rx_bd_done)(struct net_device *dev); + void (*tx_kickstart)(struct net_device *dev); + u32 (*get_int_events)(struct net_device *dev); + void (*clear_int_events)(struct net_device *dev, u32 int_events); + void (*ev_error)(struct net_device *dev, u32 int_events); + int (*get_regs)(struct net_device *dev, void *p, int *sizep); + int (*get_regs_len)(struct net_device *dev); + void (*tx_restart)(struct net_device *dev); +}; + +struct phy_info { + unsigned int id; + const char *name; + void (*startup) (struct net_device * dev); + void (*shutdown) (struct net_device * dev); + void (*ack_int) (struct net_device * dev); +}; + +/* The FEC stores dest/src/type, data, and checksum for receive packets. + */ +#define MAX_MTU 1508 /* Allow fullsized pppoe packets over VLAN */ +#define MIN_MTU 46 /* this is data size */ +#define CRC_LEN 4 + +#define PKT_MAXBUF_SIZE (MAX_MTU+ETH_HLEN+CRC_LEN) +#define PKT_MINBUF_SIZE (MIN_MTU+ETH_HLEN+CRC_LEN) + +/* Must be a multiple of 32 (to cover both FEC & FCC) */ +#define PKT_MAXBLR_SIZE ((PKT_MAXBUF_SIZE + 31) & ~31) +/* This is needed so that invalidate_xxx wont invalidate too much */ +#define ENET_RX_FRSIZE L1_CACHE_ALIGN(PKT_MAXBUF_SIZE) + +struct fs_enet_mii_bus { + struct list_head list; + spinlock_t mii_lock; + const struct fs_mii_bus_info *bus_info; + int refs; + u32 usage_map; + + int (*mii_read)(struct fs_enet_mii_bus *bus, + int phy_id, int location); + + void (*mii_write)(struct fs_enet_mii_bus *bus, + int phy_id, int location, int value); + + union { + struct { + unsigned int mii_speed; + void *fecp; + } fec; + + struct { + /* note that the actual port size may */ + /* be different; cpm(s) handle it OK */ + u8 mdio_msk; + u8 *mdio_dir; + u8 *mdio_dat; + u8 mdc_msk; + u8 *mdc_dir; + u8 *mdc_dat; + } bitbang; + + struct { + u16 lpa; + } fixed; + }; +}; + +int fs_mii_bitbang_init(struct fs_enet_mii_bus *bus); +int fs_mii_fixed_init(struct fs_enet_mii_bus *bus); +int fs_mii_fec_init(struct fs_enet_mii_bus *bus); + +struct fs_enet_private { + struct device *dev; /* pointer back to the device (must be initialized first) */ + spinlock_t lock; /* during all ops except TX pckt processing */ + spinlock_t tx_lock; /* during fs_start_xmit and fs_tx */ + const struct fs_platform_info *fpi; + const struct fs_ops *ops; + int rx_ring, tx_ring; + dma_addr_t ring_mem_addr; + void *ring_base; + struct sk_buff **rx_skbuff; + struct sk_buff **tx_skbuff; + cbd_t *rx_bd_base; /* Address of Rx and Tx buffers. */ + cbd_t *tx_bd_base; + cbd_t *dirty_tx; /* ring entries to be free()ed. */ + cbd_t *cur_rx; + cbd_t *cur_tx; + int tx_free; + struct net_device_stats stats; + struct timer_list phy_timer_list; + const struct phy_info *phy; + u32 msg_enable; + struct mii_if_info mii_if; + unsigned int last_mii_status; + struct fs_enet_mii_bus *mii_bus; + int interrupt; + + int duplex, speed; /* current settings */ + + /* event masks */ + u32 ev_napi_rx; /* mask of NAPI rx events */ + u32 ev_rx; /* rx event mask */ + u32 ev_tx; /* tx event mask */ + u32 ev_err; /* error event mask */ + + u16 bd_rx_empty; /* mask of BD rx empty */ + u16 bd_rx_err; /* mask of BD rx errors */ + + union { + struct { + int idx; /* FEC1 = 0, FEC2 = 1 */ + void *fecp; /* hw registers */ + u32 hthi, htlo; /* state for multicast */ + } fec; + + struct { + int idx; /* FCC1-3 = 0-2 */ + void *fccp; /* hw registers */ + void *ep; /* parameter ram */ + void *fcccp; /* hw registers cont. */ + void *mem; /* FCC DPRAM */ + u32 gaddrh, gaddrl; /* group address */ + } fcc; + + struct { + int idx; /* FEC1 = 0, FEC2 = 1 */ + void *sccp; /* hw registers */ + void *ep; /* parameter ram */ + u32 hthi, htlo; /* state for multicast */ + } scc; + + }; +}; + +/***************************************************************************/ + +int fs_mii_read(struct net_device *dev, int phy_id, int location); +void fs_mii_write(struct net_device *dev, int phy_id, int location, int value); + +void fs_mii_startup(struct net_device *dev); +void fs_mii_shutdown(struct net_device *dev); +void fs_mii_ack_int(struct net_device *dev); + +void fs_mii_link_status_change_check(struct net_device *dev, int init_media); + +void fs_init_bds(struct net_device *dev); +void fs_cleanup_bds(struct net_device *dev); + +/***************************************************************************/ + +#define DRV_MODULE_NAME "fs_enet" +#define PFX DRV_MODULE_NAME ": " +#define DRV_MODULE_VERSION "1.0" +#define DRV_MODULE_RELDATE "Aug 8, 2005" + +/***************************************************************************/ + +int fs_enet_platform_init(void); +void fs_enet_platform_cleanup(void); + +/***************************************************************************/ + +/* buffer descriptor access macros */ + +/* access macros */ +#if defined(CONFIG_CPM1) +/* for a a CPM1 __raw_xxx's are sufficient */ +#define __cbd_out32(addr, x) __raw_writel(x, addr) +#define __cbd_out16(addr, x) __raw_writew(x, addr) +#define __cbd_in32(addr) __raw_readl(addr) +#define __cbd_in16(addr) __raw_readw(addr) +#else +/* for others play it safe */ +#define __cbd_out32(addr, x) out_be32(addr, x) +#define __cbd_out16(addr, x) out_be16(addr, x) +#define __cbd_in32(addr) in_be32(addr) +#define __cbd_in16(addr) in_be16(addr) +#endif + +/* write */ +#define CBDW_SC(_cbd, _sc) __cbd_out16(&(_cbd)->cbd_sc, (_sc)) +#define CBDW_DATLEN(_cbd, _datlen) __cbd_out16(&(_cbd)->cbd_datlen, (_datlen)) +#define CBDW_BUFADDR(_cbd, _bufaddr) __cbd_out32(&(_cbd)->cbd_bufaddr, (_bufaddr)) + +/* read */ +#define CBDR_SC(_cbd) __cbd_in16(&(_cbd)->cbd_sc) +#define CBDR_DATLEN(_cbd) __cbd_in16(&(_cbd)->cbd_datlen) +#define CBDR_BUFADDR(_cbd) __cbd_in32(&(_cbd)->cbd_bufaddr) + +/* set bits */ +#define CBDS_SC(_cbd, _sc) CBDW_SC(_cbd, CBDR_SC(_cbd) | (_sc)) + +/* clear bits */ +#define CBDC_SC(_cbd, _sc) CBDW_SC(_cbd, CBDR_SC(_cbd) & ~(_sc)) + +/*******************************************************************/ + +extern const struct fs_ops fs_fec_ops; +extern const struct fs_ops fs_fcc_ops; +extern const struct fs_ops fs_scc_ops; + +/*******************************************************************/ + +/* handy pointer to the immap */ +extern void *fs_enet_immap; + +/*******************************************************************/ + +#endif diff --git a/drivers/net/fs_enet/mac-fcc.c b/drivers/net/fs_enet/mac-fcc.c new file mode 100644 index 000000000000..a940b96433c7 --- /dev/null +++ b/drivers/net/fs_enet/mac-fcc.c @@ -0,0 +1,578 @@ +/* + * FCC driver for Motorola MPC82xx (PQ2). + * + * Copyright (c) 2003 Intracom S.A. + * by Pantelis Antoniou + * + * 2005 (c) MontaVista Software, Inc. + * Vitaly Bordug + * + * This file is licensed under the terms of the GNU General Public License + * version 2. This program is licensed "as is" without any warranty of any + * kind, whether express or implied. + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include +#include +#include + +#include +#include +#include + +#include "fs_enet.h" + +/*************************************************/ + +/* FCC access macros */ + +#define __fcc_out32(addr, x) out_be32((unsigned *)addr, x) +#define __fcc_out16(addr, x) out_be16((unsigned short *)addr, x) +#define __fcc_out8(addr, x) out_8((unsigned char *)addr, x) +#define __fcc_in32(addr) in_be32((unsigned *)addr) +#define __fcc_in16(addr) in_be16((unsigned short *)addr) +#define __fcc_in8(addr) in_8((unsigned char *)addr) + +/* parameter space */ + +/* write, read, set bits, clear bits */ +#define W32(_p, _m, _v) __fcc_out32(&(_p)->_m, (_v)) +#define R32(_p, _m) __fcc_in32(&(_p)->_m) +#define S32(_p, _m, _v) W32(_p, _m, R32(_p, _m) | (_v)) +#define C32(_p, _m, _v) W32(_p, _m, R32(_p, _m) & ~(_v)) + +#define W16(_p, _m, _v) __fcc_out16(&(_p)->_m, (_v)) +#define R16(_p, _m) __fcc_in16(&(_p)->_m) +#define S16(_p, _m, _v) W16(_p, _m, R16(_p, _m) | (_v)) +#define C16(_p, _m, _v) W16(_p, _m, R16(_p, _m) & ~(_v)) + +#define W8(_p, _m, _v) __fcc_out8(&(_p)->_m, (_v)) +#define R8(_p, _m) __fcc_in8(&(_p)->_m) +#define S8(_p, _m, _v) W8(_p, _m, R8(_p, _m) | (_v)) +#define C8(_p, _m, _v) W8(_p, _m, R8(_p, _m) & ~(_v)) + +/*************************************************/ + +#define FCC_MAX_MULTICAST_ADDRS 64 + +#define mk_mii_read(REG) (0x60020000 | ((REG & 0x1f) << 18)) +#define mk_mii_write(REG, VAL) (0x50020000 | ((REG & 0x1f) << 18) | (VAL & 0xffff)) +#define mk_mii_end 0 + +#define MAX_CR_CMD_LOOPS 10000 + +static inline int fcc_cr_cmd(struct fs_enet_private *fep, u32 mcn, u32 op) +{ + const struct fs_platform_info *fpi = fep->fpi; + + cpm2_map_t *immap = fs_enet_immap; + cpm_cpm2_t *cpmp = &immap->im_cpm; + u32 v; + int i; + + /* Currently I don't know what feature call will look like. But + I guess there'd be something like do_cpm_cmd() which will require page & sblock */ + v = mk_cr_cmd(fpi->cp_page, fpi->cp_block, mcn, op); + W32(cpmp, cp_cpcr, v | CPM_CR_FLG); + for (i = 0; i < MAX_CR_CMD_LOOPS; i++) + if ((R32(cpmp, cp_cpcr) & CPM_CR_FLG) == 0) + break; + + if (i >= MAX_CR_CMD_LOOPS) { + printk(KERN_ERR "%s(): Not able to issue CPM command\n", + __FUNCTION__); + return 1; + } + + return 0; +} + +static int do_pd_setup(struct fs_enet_private *fep) +{ + struct platform_device *pdev = to_platform_device(fep->dev); + struct resource *r; + + /* Fill out IRQ field */ + fep->interrupt = platform_get_irq(pdev, 0); + + /* Attach the memory for the FCC Parameter RAM */ + r = platform_get_resource_byname(pdev, IORESOURCE_MEM, "fcc_pram"); + fep->fcc.ep = (void *)r->start; + + if (fep->fcc.ep == NULL) + return -EINVAL; + + r = platform_get_resource_byname(pdev, IORESOURCE_MEM, "fcc_regs"); + fep->fcc.fccp = (void *)r->start; + + if (fep->fcc.fccp == NULL) + return -EINVAL; + + fep->fcc.fcccp = (void *)fep->fpi->fcc_regs_c; + + if (fep->fcc.fcccp == NULL) + return -EINVAL; + + return 0; +} + +#define FCC_NAPI_RX_EVENT_MSK (FCC_ENET_RXF | FCC_ENET_RXB) +#define FCC_RX_EVENT (FCC_ENET_RXF) +#define FCC_TX_EVENT (FCC_ENET_TXB) +#define FCC_ERR_EVENT_MSK (FCC_ENET_TXE | FCC_ENET_BSY) + +static int setup_data(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + const struct fs_platform_info *fpi = fep->fpi; + + fep->fcc.idx = fs_get_fcc_index(fpi->fs_no); + if ((unsigned int)fep->fcc.idx >= 3) /* max 3 FCCs */ + return -EINVAL; + + fep->fcc.mem = (void *)fpi->mem_offset; + + if (do_pd_setup(fep) != 0) + return -EINVAL; + + fep->ev_napi_rx = FCC_NAPI_RX_EVENT_MSK; + fep->ev_rx = FCC_RX_EVENT; + fep->ev_tx = FCC_TX_EVENT; + fep->ev_err = FCC_ERR_EVENT_MSK; + + return 0; +} + +static int allocate_bd(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + const struct fs_platform_info *fpi = fep->fpi; + + fep->ring_base = dma_alloc_coherent(fep->dev, + (fpi->tx_ring + fpi->rx_ring) * + sizeof(cbd_t), &fep->ring_mem_addr, + GFP_KERNEL); + if (fep->ring_base == NULL) + return -ENOMEM; + + return 0; +} + +static void free_bd(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + const struct fs_platform_info *fpi = fep->fpi; + + if (fep->ring_base) + dma_free_coherent(fep->dev, + (fpi->tx_ring + fpi->rx_ring) * sizeof(cbd_t), + fep->ring_base, fep->ring_mem_addr); +} + +static void cleanup_data(struct net_device *dev) +{ + /* nothing */ +} + +static void set_promiscuous_mode(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + fcc_t *fccp = fep->fcc.fccp; + + S32(fccp, fcc_fpsmr, FCC_PSMR_PRO); +} + +static void set_multicast_start(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + fcc_enet_t *ep = fep->fcc.ep; + + W32(ep, fen_gaddrh, 0); + W32(ep, fen_gaddrl, 0); +} + +static void set_multicast_one(struct net_device *dev, const u8 *mac) +{ + struct fs_enet_private *fep = netdev_priv(dev); + fcc_enet_t *ep = fep->fcc.ep; + u16 taddrh, taddrm, taddrl; + + taddrh = ((u16)mac[5] << 8) | mac[4]; + taddrm = ((u16)mac[3] << 8) | mac[2]; + taddrl = ((u16)mac[1] << 8) | mac[0]; + + W16(ep, fen_taddrh, taddrh); + W16(ep, fen_taddrm, taddrm); + W16(ep, fen_taddrl, taddrl); + fcc_cr_cmd(fep, 0x0C, CPM_CR_SET_GADDR); +} + +static void set_multicast_finish(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + fcc_t *fccp = fep->fcc.fccp; + fcc_enet_t *ep = fep->fcc.ep; + + /* clear promiscuous always */ + C32(fccp, fcc_fpsmr, FCC_PSMR_PRO); + + /* if all multi or too many multicasts; just enable all */ + if ((dev->flags & IFF_ALLMULTI) != 0 || + dev->mc_count > FCC_MAX_MULTICAST_ADDRS) { + + W32(ep, fen_gaddrh, 0xffffffff); + W32(ep, fen_gaddrl, 0xffffffff); + } + + /* read back */ + fep->fcc.gaddrh = R32(ep, fen_gaddrh); + fep->fcc.gaddrl = R32(ep, fen_gaddrl); +} + +static void set_multicast_list(struct net_device *dev) +{ + struct dev_mc_list *pmc; + + if ((dev->flags & IFF_PROMISC) == 0) { + set_multicast_start(dev); + for (pmc = dev->mc_list; pmc != NULL; pmc = pmc->next) + set_multicast_one(dev, pmc->dmi_addr); + set_multicast_finish(dev); + } else + set_promiscuous_mode(dev); +} + +static void restart(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + const struct fs_platform_info *fpi = fep->fpi; + fcc_t *fccp = fep->fcc.fccp; + fcc_c_t *fcccp = fep->fcc.fcccp; + fcc_enet_t *ep = fep->fcc.ep; + dma_addr_t rx_bd_base_phys, tx_bd_base_phys; + u16 paddrh, paddrm, paddrl; + u16 mem_addr; + const unsigned char *mac; + int i; + + C32(fccp, fcc_gfmr, FCC_GFMR_ENR | FCC_GFMR_ENT); + + /* clear everything (slow & steady does it) */ + for (i = 0; i < sizeof(*ep); i++) + __fcc_out8((char *)ep + i, 0); + + /* get physical address */ + rx_bd_base_phys = fep->ring_mem_addr; + tx_bd_base_phys = rx_bd_base_phys + sizeof(cbd_t) * fpi->rx_ring; + + /* point to bds */ + W32(ep, fen_genfcc.fcc_rbase, rx_bd_base_phys); + W32(ep, fen_genfcc.fcc_tbase, tx_bd_base_phys); + + /* Set maximum bytes per receive buffer. + * It must be a multiple of 32. + */ + W16(ep, fen_genfcc.fcc_mrblr, PKT_MAXBLR_SIZE); + + W32(ep, fen_genfcc.fcc_rstate, (CPMFCR_GBL | CPMFCR_EB) << 24); + W32(ep, fen_genfcc.fcc_tstate, (CPMFCR_GBL | CPMFCR_EB) << 24); + + /* Allocate space in the reserved FCC area of DPRAM for the + * internal buffers. No one uses this space (yet), so we + * can do this. Later, we will add resource management for + * this area. + */ + + mem_addr = (u32) fep->fcc.mem; /* de-fixup dpram offset */ + + W16(ep, fen_genfcc.fcc_riptr, (mem_addr & 0xffff)); + W16(ep, fen_genfcc.fcc_tiptr, ((mem_addr + 32) & 0xffff)); + W16(ep, fen_padptr, mem_addr + 64); + + /* fill with special symbol... */ + memset(fep->fcc.mem + fpi->dpram_offset + 64, 0x88, 32); + + W32(ep, fen_genfcc.fcc_rbptr, 0); + W32(ep, fen_genfcc.fcc_tbptr, 0); + W32(ep, fen_genfcc.fcc_rcrc, 0); + W32(ep, fen_genfcc.fcc_tcrc, 0); + W16(ep, fen_genfcc.fcc_res1, 0); + W32(ep, fen_genfcc.fcc_res2, 0); + + /* no CAM */ + W32(ep, fen_camptr, 0); + + /* Set CRC preset and mask */ + W32(ep, fen_cmask, 0xdebb20e3); + W32(ep, fen_cpres, 0xffffffff); + + W32(ep, fen_crcec, 0); /* CRC Error counter */ + W32(ep, fen_alec, 0); /* alignment error counter */ + W32(ep, fen_disfc, 0); /* discard frame counter */ + W16(ep, fen_retlim, 15); /* Retry limit threshold */ + W16(ep, fen_pper, 0); /* Normal persistence */ + + /* set group address */ + W32(ep, fen_gaddrh, fep->fcc.gaddrh); + W32(ep, fen_gaddrl, fep->fcc.gaddrh); + + /* Clear hash filter tables */ + W32(ep, fen_iaddrh, 0); + W32(ep, fen_iaddrl, 0); + + /* Clear the Out-of-sequence TxBD */ + W16(ep, fen_tfcstat, 0); + W16(ep, fen_tfclen, 0); + W32(ep, fen_tfcptr, 0); + + W16(ep, fen_mflr, PKT_MAXBUF_SIZE); /* maximum frame length register */ + W16(ep, fen_minflr, PKT_MINBUF_SIZE); /* minimum frame length register */ + + /* set address */ + mac = dev->dev_addr; + paddrh = ((u16)mac[5] << 8) | mac[4]; + paddrm = ((u16)mac[3] << 8) | mac[2]; + paddrl = ((u16)mac[1] << 8) | mac[0]; + + W16(ep, fen_paddrh, paddrh); + W16(ep, fen_paddrm, paddrm); + W16(ep, fen_paddrl, paddrl); + + W16(ep, fen_taddrh, 0); + W16(ep, fen_taddrm, 0); + W16(ep, fen_taddrl, 0); + + W16(ep, fen_maxd1, 1520); /* maximum DMA1 length */ + W16(ep, fen_maxd2, 1520); /* maximum DMA2 length */ + + /* Clear stat counters, in case we ever enable RMON */ + W32(ep, fen_octc, 0); + W32(ep, fen_colc, 0); + W32(ep, fen_broc, 0); + W32(ep, fen_mulc, 0); + W32(ep, fen_uspc, 0); + W32(ep, fen_frgc, 0); + W32(ep, fen_ospc, 0); + W32(ep, fen_jbrc, 0); + W32(ep, fen_p64c, 0); + W32(ep, fen_p65c, 0); + W32(ep, fen_p128c, 0); + W32(ep, fen_p256c, 0); + W32(ep, fen_p512c, 0); + W32(ep, fen_p1024c, 0); + + W16(ep, fen_rfthr, 0); /* Suggested by manual */ + W16(ep, fen_rfcnt, 0); + W16(ep, fen_cftype, 0); + + fs_init_bds(dev); + + /* adjust to speed (for RMII mode) */ + if (fpi->use_rmii) { + if (fep->speed == 100) + C8(fcccp, fcc_gfemr, 0x20); + else + S8(fcccp, fcc_gfemr, 0x20); + } + + fcc_cr_cmd(fep, 0x0c, CPM_CR_INIT_TRX); + + /* clear events */ + W16(fccp, fcc_fcce, 0xffff); + + /* Enable interrupts we wish to service */ + W16(fccp, fcc_fccm, FCC_ENET_TXE | FCC_ENET_RXF | FCC_ENET_TXB); + + /* Set GFMR to enable Ethernet operating mode */ + W32(fccp, fcc_gfmr, FCC_GFMR_TCI | FCC_GFMR_MODE_ENET); + + /* set sync/delimiters */ + W16(fccp, fcc_fdsr, 0xd555); + + W32(fccp, fcc_fpsmr, FCC_PSMR_ENCRC); + + if (fpi->use_rmii) + S32(fccp, fcc_fpsmr, FCC_PSMR_RMII); + + /* adjust to duplex mode */ + if (fep->duplex) + S32(fccp, fcc_fpsmr, FCC_PSMR_FDE | FCC_PSMR_LPB); + else + C32(fccp, fcc_fpsmr, FCC_PSMR_FDE | FCC_PSMR_LPB); + + S32(fccp, fcc_gfmr, FCC_GFMR_ENR | FCC_GFMR_ENT); +} + +static void stop(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + fcc_t *fccp = fep->fcc.fccp; + + /* stop ethernet */ + C32(fccp, fcc_gfmr, FCC_GFMR_ENR | FCC_GFMR_ENT); + + /* clear events */ + W16(fccp, fcc_fcce, 0xffff); + + /* clear interrupt mask */ + W16(fccp, fcc_fccm, 0); + + fs_cleanup_bds(dev); +} + +static void pre_request_irq(struct net_device *dev, int irq) +{ + /* nothing */ +} + +static void post_free_irq(struct net_device *dev, int irq) +{ + /* nothing */ +} + +static void napi_clear_rx_event(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + fcc_t *fccp = fep->fcc.fccp; + + W16(fccp, fcc_fcce, FCC_NAPI_RX_EVENT_MSK); +} + +static void napi_enable_rx(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + fcc_t *fccp = fep->fcc.fccp; + + S16(fccp, fcc_fccm, FCC_NAPI_RX_EVENT_MSK); +} + +static void napi_disable_rx(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + fcc_t *fccp = fep->fcc.fccp; + + C16(fccp, fcc_fccm, FCC_NAPI_RX_EVENT_MSK); +} + +static void rx_bd_done(struct net_device *dev) +{ + /* nothing */ +} + +static void tx_kickstart(struct net_device *dev) +{ + /* nothing */ +} + +static u32 get_int_events(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + fcc_t *fccp = fep->fcc.fccp; + + return (u32)R16(fccp, fcc_fcce); +} + +static void clear_int_events(struct net_device *dev, u32 int_events) +{ + struct fs_enet_private *fep = netdev_priv(dev); + fcc_t *fccp = fep->fcc.fccp; + + W16(fccp, fcc_fcce, int_events & 0xffff); +} + +static void ev_error(struct net_device *dev, u32 int_events) +{ + printk(KERN_WARNING DRV_MODULE_NAME + ": %s FS_ENET ERROR(s) 0x%x\n", dev->name, int_events); +} + +int get_regs(struct net_device *dev, void *p, int *sizep) +{ + struct fs_enet_private *fep = netdev_priv(dev); + + if (*sizep < sizeof(fcc_t) + sizeof(fcc_c_t) + sizeof(fcc_enet_t)) + return -EINVAL; + + memcpy_fromio(p, fep->fcc.fccp, sizeof(fcc_t)); + p = (char *)p + sizeof(fcc_t); + + memcpy_fromio(p, fep->fcc.fcccp, sizeof(fcc_c_t)); + p = (char *)p + sizeof(fcc_c_t); + + memcpy_fromio(p, fep->fcc.ep, sizeof(fcc_enet_t)); + + return 0; +} + +int get_regs_len(struct net_device *dev) +{ + return sizeof(fcc_t) + sizeof(fcc_c_t) + sizeof(fcc_enet_t); +} + +/* Some transmit errors cause the transmitter to shut + * down. We now issue a restart transmit. Since the + * errors close the BD and update the pointers, the restart + * _should_ pick up without having to reset any of our + * pointers either. Also, To workaround 8260 device erratum + * CPM37, we must disable and then re-enable the transmitter + * following a Late Collision, Underrun, or Retry Limit error. + */ +void tx_restart(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + fcc_t *fccp = fep->fcc.fccp; + + C32(fccp, fcc_gfmr, FCC_GFMR_ENT); + udelay(10); + S32(fccp, fcc_gfmr, FCC_GFMR_ENT); + + fcc_cr_cmd(fep, 0x0C, CPM_CR_RESTART_TX); +} + +/*************************************************************************/ + +const struct fs_ops fs_fcc_ops = { + .setup_data = setup_data, + .cleanup_data = cleanup_data, + .set_multicast_list = set_multicast_list, + .restart = restart, + .stop = stop, + .pre_request_irq = pre_request_irq, + .post_free_irq = post_free_irq, + .napi_clear_rx_event = napi_clear_rx_event, + .napi_enable_rx = napi_enable_rx, + .napi_disable_rx = napi_disable_rx, + .rx_bd_done = rx_bd_done, + .tx_kickstart = tx_kickstart, + .get_int_events = get_int_events, + .clear_int_events = clear_int_events, + .ev_error = ev_error, + .get_regs = get_regs, + .get_regs_len = get_regs_len, + .tx_restart = tx_restart, + .allocate_bd = allocate_bd, + .free_bd = free_bd, +}; diff --git a/drivers/net/fs_enet/mac-fec.c b/drivers/net/fs_enet/mac-fec.c new file mode 100644 index 000000000000..5ef4e845a387 --- /dev/null +++ b/drivers/net/fs_enet/mac-fec.c @@ -0,0 +1,653 @@ +/* + * Freescale Ethernet controllers + * + * Copyright (c) 2005 Intracom S.A. + * by Pantelis Antoniou + * + * 2005 (c) MontaVista Software, Inc. + * Vitaly Bordug + * + * This file is licensed under the terms of the GNU General Public License + * version 2. This program is licensed "as is" without any warranty of any + * kind, whether express or implied. + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include +#include + +#ifdef CONFIG_8xx +#include +#include +#include +#include +#endif + +#include "fs_enet.h" + +/*************************************************/ + +#if defined(CONFIG_CPM1) +/* for a CPM1 __raw_xxx's are sufficient */ +#define __fs_out32(addr, x) __raw_writel(x, addr) +#define __fs_out16(addr, x) __raw_writew(x, addr) +#define __fs_in32(addr) __raw_readl(addr) +#define __fs_in16(addr) __raw_readw(addr) +#else +/* for others play it safe */ +#define __fs_out32(addr, x) out_be32(addr, x) +#define __fs_out16(addr, x) out_be16(addr, x) +#define __fs_in32(addr) in_be32(addr) +#define __fs_in16(addr) in_be16(addr) +#endif + +/* write */ +#define FW(_fecp, _reg, _v) __fs_out32(&(_fecp)->fec_ ## _reg, (_v)) + +/* read */ +#define FR(_fecp, _reg) __fs_in32(&(_fecp)->fec_ ## _reg) + +/* set bits */ +#define FS(_fecp, _reg, _v) FW(_fecp, _reg, FR(_fecp, _reg) | (_v)) + +/* clear bits */ +#define FC(_fecp, _reg, _v) FW(_fecp, _reg, FR(_fecp, _reg) & ~(_v)) + + +/* CRC polynomium used by the FEC for the multicast group filtering */ +#define FEC_CRC_POLY 0x04C11DB7 + +#define FEC_MAX_MULTICAST_ADDRS 64 + +/* Interrupt events/masks. +*/ +#define FEC_ENET_HBERR 0x80000000U /* Heartbeat error */ +#define FEC_ENET_BABR 0x40000000U /* Babbling receiver */ +#define FEC_ENET_BABT 0x20000000U /* Babbling transmitter */ +#define FEC_ENET_GRA 0x10000000U /* Graceful stop complete */ +#define FEC_ENET_TXF 0x08000000U /* Full frame transmitted */ +#define FEC_ENET_TXB 0x04000000U /* A buffer was transmitted */ +#define FEC_ENET_RXF 0x02000000U /* Full frame received */ +#define FEC_ENET_RXB 0x01000000U /* A buffer was received */ +#define FEC_ENET_MII 0x00800000U /* MII interrupt */ +#define FEC_ENET_EBERR 0x00400000U /* SDMA bus error */ + +#define FEC_ECNTRL_PINMUX 0x00000004 +#define FEC_ECNTRL_ETHER_EN 0x00000002 +#define FEC_ECNTRL_RESET 0x00000001 + +#define FEC_RCNTRL_BC_REJ 0x00000010 +#define FEC_RCNTRL_PROM 0x00000008 +#define FEC_RCNTRL_MII_MODE 0x00000004 +#define FEC_RCNTRL_DRT 0x00000002 +#define FEC_RCNTRL_LOOP 0x00000001 + +#define FEC_TCNTRL_FDEN 0x00000004 +#define FEC_TCNTRL_HBC 0x00000002 +#define FEC_TCNTRL_GTS 0x00000001 + + +/* Make MII read/write commands for the FEC. +*/ +#define mk_mii_read(REG) (0x60020000 | ((REG & 0x1f) << 18)) +#define mk_mii_write(REG, VAL) (0x50020000 | ((REG & 0x1f) << 18) | (VAL & 0xffff)) +#define mk_mii_end 0 + +#define FEC_MII_LOOPS 10000 + +/* + * Delay to wait for FEC reset command to complete (in us) + */ +#define FEC_RESET_DELAY 50 + +static int whack_reset(fec_t * fecp) +{ + int i; + + FW(fecp, ecntrl, FEC_ECNTRL_PINMUX | FEC_ECNTRL_RESET); + for (i = 0; i < FEC_RESET_DELAY; i++) { + if ((FR(fecp, ecntrl) & FEC_ECNTRL_RESET) == 0) + return 0; /* OK */ + udelay(1); + } + + return -1; +} + +static int do_pd_setup(struct fs_enet_private *fep) +{ + struct platform_device *pdev = to_platform_device(fep->dev); + struct resource *r; + + /* Fill out IRQ field */ + fep->interrupt = platform_get_irq_byname(pdev,"interrupt"); + + r = platform_get_resource_byname(pdev, IORESOURCE_MEM, "regs"); + fep->fec.fecp =(void*)r->start; + + if(fep->fec.fecp == NULL) + return -EINVAL; + + return 0; + +} + +#define FEC_NAPI_RX_EVENT_MSK (FEC_ENET_RXF | FEC_ENET_RXB) +#define FEC_RX_EVENT (FEC_ENET_RXF) +#define FEC_TX_EVENT (FEC_ENET_TXF) +#define FEC_ERR_EVENT_MSK (FEC_ENET_HBERR | FEC_ENET_BABR | \ + FEC_ENET_BABT | FEC_ENET_EBERR) + +static int setup_data(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + + if (do_pd_setup(fep) != 0) + return -EINVAL; + + fep->fec.hthi = 0; + fep->fec.htlo = 0; + + fep->ev_napi_rx = FEC_NAPI_RX_EVENT_MSK; + fep->ev_rx = FEC_RX_EVENT; + fep->ev_tx = FEC_TX_EVENT; + fep->ev_err = FEC_ERR_EVENT_MSK; + + return 0; +} + +static int allocate_bd(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + const struct fs_platform_info *fpi = fep->fpi; + + fep->ring_base = dma_alloc_coherent(fep->dev, + (fpi->tx_ring + fpi->rx_ring) * + sizeof(cbd_t), &fep->ring_mem_addr, + GFP_KERNEL); + if (fep->ring_base == NULL) + return -ENOMEM; + + return 0; +} + +static void free_bd(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + const struct fs_platform_info *fpi = fep->fpi; + + if(fep->ring_base) + dma_free_coherent(fep->dev, (fpi->tx_ring + fpi->rx_ring) + * sizeof(cbd_t), + fep->ring_base, + fep->ring_mem_addr); +} + +static void cleanup_data(struct net_device *dev) +{ + /* nothing */ +} + +static void set_promiscuous_mode(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + fec_t *fecp = fep->fec.fecp; + + FS(fecp, r_cntrl, FEC_RCNTRL_PROM); +} + +static void set_multicast_start(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + + fep->fec.hthi = 0; + fep->fec.htlo = 0; +} + +static void set_multicast_one(struct net_device *dev, const u8 *mac) +{ + struct fs_enet_private *fep = netdev_priv(dev); + int temp, hash_index, i, j; + u32 crc, csrVal; + u8 byte, msb; + + crc = 0xffffffff; + for (i = 0; i < 6; i++) { + byte = mac[i]; + for (j = 0; j < 8; j++) { + msb = crc >> 31; + crc <<= 1; + if (msb ^ (byte & 0x1)) + crc ^= FEC_CRC_POLY; + byte >>= 1; + } + } + + temp = (crc & 0x3f) >> 1; + hash_index = ((temp & 0x01) << 4) | + ((temp & 0x02) << 2) | + ((temp & 0x04)) | + ((temp & 0x08) >> 2) | + ((temp & 0x10) >> 4); + csrVal = 1 << hash_index; + if (crc & 1) + fep->fec.hthi |= csrVal; + else + fep->fec.htlo |= csrVal; +} + +static void set_multicast_finish(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + fec_t *fecp = fep->fec.fecp; + + /* if all multi or too many multicasts; just enable all */ + if ((dev->flags & IFF_ALLMULTI) != 0 || + dev->mc_count > FEC_MAX_MULTICAST_ADDRS) { + fep->fec.hthi = 0xffffffffU; + fep->fec.htlo = 0xffffffffU; + } + + FC(fecp, r_cntrl, FEC_RCNTRL_PROM); + FW(fecp, hash_table_high, fep->fec.hthi); + FW(fecp, hash_table_low, fep->fec.htlo); +} + +static void set_multicast_list(struct net_device *dev) +{ + struct dev_mc_list *pmc; + + if ((dev->flags & IFF_PROMISC) == 0) { + set_multicast_start(dev); + for (pmc = dev->mc_list; pmc != NULL; pmc = pmc->next) + set_multicast_one(dev, pmc->dmi_addr); + set_multicast_finish(dev); + } else + set_promiscuous_mode(dev); +} + +static void restart(struct net_device *dev) +{ +#ifdef CONFIG_DUET + immap_t *immap = fs_enet_immap; + u32 cptr; +#endif + struct fs_enet_private *fep = netdev_priv(dev); + fec_t *fecp = fep->fec.fecp; + const struct fs_platform_info *fpi = fep->fpi; + dma_addr_t rx_bd_base_phys, tx_bd_base_phys; + int r; + u32 addrhi, addrlo; + + r = whack_reset(fep->fec.fecp); + if (r != 0) + printk(KERN_ERR DRV_MODULE_NAME + ": %s FEC Reset FAILED!\n", dev->name); + + /* + * Set station address. + */ + addrhi = ((u32) dev->dev_addr[0] << 24) | + ((u32) dev->dev_addr[1] << 16) | + ((u32) dev->dev_addr[2] << 8) | + (u32) dev->dev_addr[3]; + addrlo = ((u32) dev->dev_addr[4] << 24) | + ((u32) dev->dev_addr[5] << 16); + FW(fecp, addr_low, addrhi); + FW(fecp, addr_high, addrlo); + + /* + * Reset all multicast. + */ + FW(fecp, hash_table_high, fep->fec.hthi); + FW(fecp, hash_table_low, fep->fec.htlo); + + /* + * Set maximum receive buffer size. + */ + FW(fecp, r_buff_size, PKT_MAXBLR_SIZE); + FW(fecp, r_hash, PKT_MAXBUF_SIZE); + + /* get physical address */ + rx_bd_base_phys = fep->ring_mem_addr; + tx_bd_base_phys = rx_bd_base_phys + sizeof(cbd_t) * fpi->rx_ring; + + /* + * Set receive and transmit descriptor base. + */ + FW(fecp, r_des_start, rx_bd_base_phys); + FW(fecp, x_des_start, tx_bd_base_phys); + + fs_init_bds(dev); + + /* + * Enable big endian and don't care about SDMA FC. + */ + FW(fecp, fun_code, 0x78000000); + + /* + * Set MII speed. + */ + FW(fecp, mii_speed, fep->mii_bus->fec.mii_speed); + + /* + * Clear any outstanding interrupt. + */ + FW(fecp, ievent, 0xffc0); + FW(fecp, ivec, (fep->interrupt / 2) << 29); + + + /* + * adjust to speed (only for DUET & RMII) + */ +#ifdef CONFIG_DUET + if (fpi->use_rmii) { + cptr = in_be32(&immap->im_cpm.cp_cptr); + switch (fs_get_fec_index(fpi->fs_no)) { + case 0: + cptr |= 0x100; + if (fep->speed == 10) + cptr |= 0x0000010; + else if (fep->speed == 100) + cptr &= ~0x0000010; + break; + case 1: + cptr |= 0x80; + if (fep->speed == 10) + cptr |= 0x0000008; + else if (fep->speed == 100) + cptr &= ~0x0000008; + break; + default: + BUG(); /* should never happen */ + break; + } + out_be32(&immap->im_cpm.cp_cptr, cptr); + } +#endif + + FW(fecp, r_cntrl, FEC_RCNTRL_MII_MODE); /* MII enable */ + /* + * adjust to duplex mode + */ + if (fep->duplex) { + FC(fecp, r_cntrl, FEC_RCNTRL_DRT); + FS(fecp, x_cntrl, FEC_TCNTRL_FDEN); /* FD enable */ + } else { + FS(fecp, r_cntrl, FEC_RCNTRL_DRT); + FC(fecp, x_cntrl, FEC_TCNTRL_FDEN); /* FD disable */ + } + + /* + * Enable interrupts we wish to service. + */ + FW(fecp, imask, FEC_ENET_TXF | FEC_ENET_TXB | + FEC_ENET_RXF | FEC_ENET_RXB); + + /* + * And last, enable the transmit and receive processing. + */ + FW(fecp, ecntrl, FEC_ECNTRL_PINMUX | FEC_ECNTRL_ETHER_EN); + FW(fecp, r_des_active, 0x01000000); +} + +static void stop(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + fec_t *fecp = fep->fec.fecp; + struct fs_enet_mii_bus *bus = fep->mii_bus; + const struct fs_mii_bus_info *bi = bus->bus_info; + int i; + + if ((FR(fecp, ecntrl) & FEC_ECNTRL_ETHER_EN) == 0) + return; /* already down */ + + FW(fecp, x_cntrl, 0x01); /* Graceful transmit stop */ + for (i = 0; ((FR(fecp, ievent) & 0x10000000) == 0) && + i < FEC_RESET_DELAY; i++) + udelay(1); + + if (i == FEC_RESET_DELAY) + printk(KERN_WARNING DRV_MODULE_NAME + ": %s FEC timeout on graceful transmit stop\n", + dev->name); + /* + * Disable FEC. Let only MII interrupts. + */ + FW(fecp, imask, 0); + FC(fecp, ecntrl, FEC_ECNTRL_ETHER_EN); + + fs_cleanup_bds(dev); + + /* shut down FEC1? that's where the mii bus is */ + if (fep->fec.idx == 0 && bus->refs > 1 && bi->method == fsmii_fec) { + FS(fecp, r_cntrl, FEC_RCNTRL_MII_MODE); /* MII enable */ + FS(fecp, ecntrl, FEC_ECNTRL_PINMUX | FEC_ECNTRL_ETHER_EN); + FW(fecp, ievent, FEC_ENET_MII); + FW(fecp, mii_speed, bus->fec.mii_speed); + } +} + +static void pre_request_irq(struct net_device *dev, int irq) +{ + immap_t *immap = fs_enet_immap; + u32 siel; + + /* SIU interrupt */ + if (irq >= SIU_IRQ0 && irq < SIU_LEVEL7) { + + siel = in_be32(&immap->im_siu_conf.sc_siel); + if ((irq & 1) == 0) + siel |= (0x80000000 >> irq); + else + siel &= ~(0x80000000 >> (irq & ~1)); + out_be32(&immap->im_siu_conf.sc_siel, siel); + } +} + +static void post_free_irq(struct net_device *dev, int irq) +{ + /* nothing */ +} + +static void napi_clear_rx_event(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + fec_t *fecp = fep->fec.fecp; + + FW(fecp, ievent, FEC_NAPI_RX_EVENT_MSK); +} + +static void napi_enable_rx(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + fec_t *fecp = fep->fec.fecp; + + FS(fecp, imask, FEC_NAPI_RX_EVENT_MSK); +} + +static void napi_disable_rx(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + fec_t *fecp = fep->fec.fecp; + + FC(fecp, imask, FEC_NAPI_RX_EVENT_MSK); +} + +static void rx_bd_done(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + fec_t *fecp = fep->fec.fecp; + + FW(fecp, r_des_active, 0x01000000); +} + +static void tx_kickstart(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + fec_t *fecp = fep->fec.fecp; + + FW(fecp, x_des_active, 0x01000000); +} + +static u32 get_int_events(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + fec_t *fecp = fep->fec.fecp; + + return FR(fecp, ievent) & FR(fecp, imask); +} + +static void clear_int_events(struct net_device *dev, u32 int_events) +{ + struct fs_enet_private *fep = netdev_priv(dev); + fec_t *fecp = fep->fec.fecp; + + FW(fecp, ievent, int_events); +} + +static void ev_error(struct net_device *dev, u32 int_events) +{ + printk(KERN_WARNING DRV_MODULE_NAME + ": %s FEC ERROR(s) 0x%x\n", dev->name, int_events); +} + +int get_regs(struct net_device *dev, void *p, int *sizep) +{ + struct fs_enet_private *fep = netdev_priv(dev); + + if (*sizep < sizeof(fec_t)) + return -EINVAL; + + memcpy_fromio(p, fep->fec.fecp, sizeof(fec_t)); + + return 0; +} + +int get_regs_len(struct net_device *dev) +{ + return sizeof(fec_t); +} + +void tx_restart(struct net_device *dev) +{ + /* nothing */ +} + +/*************************************************************************/ + +const struct fs_ops fs_fec_ops = { + .setup_data = setup_data, + .cleanup_data = cleanup_data, + .set_multicast_list = set_multicast_list, + .restart = restart, + .stop = stop, + .pre_request_irq = pre_request_irq, + .post_free_irq = post_free_irq, + .napi_clear_rx_event = napi_clear_rx_event, + .napi_enable_rx = napi_enable_rx, + .napi_disable_rx = napi_disable_rx, + .rx_bd_done = rx_bd_done, + .tx_kickstart = tx_kickstart, + .get_int_events = get_int_events, + .clear_int_events = clear_int_events, + .ev_error = ev_error, + .get_regs = get_regs, + .get_regs_len = get_regs_len, + .tx_restart = tx_restart, + .allocate_bd = allocate_bd, + .free_bd = free_bd, +}; + +/***********************************************************************/ + +static int mii_read(struct fs_enet_mii_bus *bus, int phy_id, int location) +{ + fec_t *fecp = bus->fec.fecp; + int i, ret = -1; + + if ((FR(fecp, r_cntrl) & FEC_RCNTRL_MII_MODE) == 0) + BUG(); + + /* Add PHY address to register command. */ + FW(fecp, mii_data, (phy_id << 23) | mk_mii_read(location)); + + for (i = 0; i < FEC_MII_LOOPS; i++) + if ((FR(fecp, ievent) & FEC_ENET_MII) != 0) + break; + + if (i < FEC_MII_LOOPS) { + FW(fecp, ievent, FEC_ENET_MII); + ret = FR(fecp, mii_data) & 0xffff; + } + + return ret; +} + +static void mii_write(struct fs_enet_mii_bus *bus, int phy_id, int location, int value) +{ + fec_t *fecp = bus->fec.fecp; + int i; + + /* this must never happen */ + if ((FR(fecp, r_cntrl) & FEC_RCNTRL_MII_MODE) == 0) + BUG(); + + /* Add PHY address to register command. */ + FW(fecp, mii_data, (phy_id << 23) | mk_mii_write(location, value)); + + for (i = 0; i < FEC_MII_LOOPS; i++) + if ((FR(fecp, ievent) & FEC_ENET_MII) != 0) + break; + + if (i < FEC_MII_LOOPS) + FW(fecp, ievent, FEC_ENET_MII); +} + +int fs_mii_fec_init(struct fs_enet_mii_bus *bus) +{ + bd_t *bd = (bd_t *)__res; + const struct fs_mii_bus_info *bi = bus->bus_info; + fec_t *fecp; + + if (bi->id != 0) + return -1; + + bus->fec.fecp = &((immap_t *)fs_enet_immap)->im_cpm.cp_fec; + bus->fec.mii_speed = ((((bd->bi_intfreq + 4999999) / 2500000) / 2) + & 0x3F) << 1; + + fecp = bus->fec.fecp; + + FS(fecp, r_cntrl, FEC_RCNTRL_MII_MODE); /* MII enable */ + FS(fecp, ecntrl, FEC_ECNTRL_PINMUX | FEC_ECNTRL_ETHER_EN); + FW(fecp, ievent, FEC_ENET_MII); + FW(fecp, mii_speed, bus->fec.mii_speed); + + bus->mii_read = mii_read; + bus->mii_write = mii_write; + + return 0; +} diff --git a/drivers/net/fs_enet/mac-scc.c b/drivers/net/fs_enet/mac-scc.c new file mode 100644 index 000000000000..d8c6e9cadcf5 --- /dev/null +++ b/drivers/net/fs_enet/mac-scc.c @@ -0,0 +1,524 @@ +/* + * Ethernet on Serial Communications Controller (SCC) driver for Motorola MPC8xx and MPC82xx. + * + * Copyright (c) 2003 Intracom S.A. + * by Pantelis Antoniou + * + * 2005 (c) MontaVista Software, Inc. + * Vitaly Bordug + * + * This file is licensed under the terms of the GNU General Public License + * version 2. This program is licensed "as is" without any warranty of any + * kind, whether express or implied. + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include +#include + +#ifdef CONFIG_8xx +#include +#include +#include +#include +#endif + +#include "fs_enet.h" + +/*************************************************/ + +#if defined(CONFIG_CPM1) +/* for a 8xx __raw_xxx's are sufficient */ +#define __fs_out32(addr, x) __raw_writel(x, addr) +#define __fs_out16(addr, x) __raw_writew(x, addr) +#define __fs_out8(addr, x) __raw_writeb(x, addr) +#define __fs_in32(addr) __raw_readl(addr) +#define __fs_in16(addr) __raw_readw(addr) +#define __fs_in8(addr) __raw_readb(addr) +#else +/* for others play it safe */ +#define __fs_out32(addr, x) out_be32(addr, x) +#define __fs_out16(addr, x) out_be16(addr, x) +#define __fs_in32(addr) in_be32(addr) +#define __fs_in16(addr) in_be16(addr) +#endif + +/* write, read, set bits, clear bits */ +#define W32(_p, _m, _v) __fs_out32(&(_p)->_m, (_v)) +#define R32(_p, _m) __fs_in32(&(_p)->_m) +#define S32(_p, _m, _v) W32(_p, _m, R32(_p, _m) | (_v)) +#define C32(_p, _m, _v) W32(_p, _m, R32(_p, _m) & ~(_v)) + +#define W16(_p, _m, _v) __fs_out16(&(_p)->_m, (_v)) +#define R16(_p, _m) __fs_in16(&(_p)->_m) +#define S16(_p, _m, _v) W16(_p, _m, R16(_p, _m) | (_v)) +#define C16(_p, _m, _v) W16(_p, _m, R16(_p, _m) & ~(_v)) + +#define W8(_p, _m, _v) __fs_out8(&(_p)->_m, (_v)) +#define R8(_p, _m) __fs_in8(&(_p)->_m) +#define S8(_p, _m, _v) W8(_p, _m, R8(_p, _m) | (_v)) +#define C8(_p, _m, _v) W8(_p, _m, R8(_p, _m) & ~(_v)) + +#define SCC_MAX_MULTICAST_ADDRS 64 + +/* + * Delay to wait for SCC reset command to complete (in us) + */ +#define SCC_RESET_DELAY 50 +#define MAX_CR_CMD_LOOPS 10000 + +static inline int scc_cr_cmd(struct fs_enet_private *fep, u32 op) +{ + cpm8xx_t *cpmp = &((immap_t *)fs_enet_immap)->im_cpm; + u32 v, ch; + int i = 0; + + ch = fep->scc.idx << 2; + v = mk_cr_cmd(ch, op); + W16(cpmp, cp_cpcr, v | CPM_CR_FLG); + for (i = 0; i < MAX_CR_CMD_LOOPS; i++) + if ((R16(cpmp, cp_cpcr) & CPM_CR_FLG) == 0) + break; + + if (i >= MAX_CR_CMD_LOOPS) { + printk(KERN_ERR "%s(): Not able to issue CPM command\n", + __FUNCTION__); + return 1; + } + return 0; +} + +static int do_pd_setup(struct fs_enet_private *fep) +{ + struct platform_device *pdev = to_platform_device(fep->dev); + struct resource *r; + + /* Fill out IRQ field */ + fep->interrupt = platform_get_irq_byname(pdev, "interrupt"); + + r = platform_get_resource_byname(pdev, IORESOURCE_MEM, "regs"); + fep->scc.sccp = (void *)r->start; + + if (fep->scc.sccp == NULL) + return -EINVAL; + + r = platform_get_resource_byname(pdev, IORESOURCE_MEM, "pram"); + fep->scc.ep = (void *)r->start; + + if (fep->scc.ep == NULL) + return -EINVAL; + + return 0; +} + +#define SCC_NAPI_RX_EVENT_MSK (SCCE_ENET_RXF | SCCE_ENET_RXB) +#define SCC_RX_EVENT (SCCE_ENET_RXF) +#define SCC_TX_EVENT (SCCE_ENET_TXB) +#define SCC_ERR_EVENT_MSK (SCCE_ENET_TXE | SCCE_ENET_BSY) + +static int setup_data(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + const struct fs_platform_info *fpi = fep->fpi; + + fep->scc.idx = fs_get_scc_index(fpi->fs_no); + if ((unsigned int)fep->fcc.idx > 4) /* max 4 SCCs */ + return -EINVAL; + + do_pd_setup(fep); + + fep->scc.hthi = 0; + fep->scc.htlo = 0; + + fep->ev_napi_rx = SCC_NAPI_RX_EVENT_MSK; + fep->ev_rx = SCC_RX_EVENT; + fep->ev_tx = SCC_TX_EVENT; + fep->ev_err = SCC_ERR_EVENT_MSK; + + return 0; +} + +static int allocate_bd(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + const struct fs_platform_info *fpi = fep->fpi; + + fep->ring_mem_addr = cpm_dpalloc((fpi->tx_ring + fpi->rx_ring) * + sizeof(cbd_t), 8); + if (IS_DPERR(fep->ring_mem_addr)) + return -ENOMEM; + + fep->ring_base = cpm_dpram_addr(fep->ring_mem_addr); + + return 0; +} + +static void free_bd(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + + if (fep->ring_base) + cpm_dpfree(fep->ring_mem_addr); +} + +static void cleanup_data(struct net_device *dev) +{ + /* nothing */ +} + +static void set_promiscuous_mode(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + scc_t *sccp = fep->scc.sccp; + + S16(sccp, scc_psmr, SCC_PSMR_PRO); +} + +static void set_multicast_start(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + scc_enet_t *ep = fep->scc.ep; + + W16(ep, sen_gaddr1, 0); + W16(ep, sen_gaddr2, 0); + W16(ep, sen_gaddr3, 0); + W16(ep, sen_gaddr4, 0); +} + +static void set_multicast_one(struct net_device *dev, const u8 * mac) +{ + struct fs_enet_private *fep = netdev_priv(dev); + scc_enet_t *ep = fep->scc.ep; + u16 taddrh, taddrm, taddrl; + + taddrh = ((u16) mac[5] << 8) | mac[4]; + taddrm = ((u16) mac[3] << 8) | mac[2]; + taddrl = ((u16) mac[1] << 8) | mac[0]; + + W16(ep, sen_taddrh, taddrh); + W16(ep, sen_taddrm, taddrm); + W16(ep, sen_taddrl, taddrl); + scc_cr_cmd(fep, CPM_CR_SET_GADDR); +} + +static void set_multicast_finish(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + scc_t *sccp = fep->scc.sccp; + scc_enet_t *ep = fep->scc.ep; + + /* clear promiscuous always */ + C16(sccp, scc_psmr, SCC_PSMR_PRO); + + /* if all multi or too many multicasts; just enable all */ + if ((dev->flags & IFF_ALLMULTI) != 0 || + dev->mc_count > SCC_MAX_MULTICAST_ADDRS) { + + W16(ep, sen_gaddr1, 0xffff); + W16(ep, sen_gaddr2, 0xffff); + W16(ep, sen_gaddr3, 0xffff); + W16(ep, sen_gaddr4, 0xffff); + } +} + +static void set_multicast_list(struct net_device *dev) +{ + struct dev_mc_list *pmc; + + if ((dev->flags & IFF_PROMISC) == 0) { + set_multicast_start(dev); + for (pmc = dev->mc_list; pmc != NULL; pmc = pmc->next) + set_multicast_one(dev, pmc->dmi_addr); + set_multicast_finish(dev); + } else + set_promiscuous_mode(dev); +} + +/* + * This function is called to start or restart the FEC during a link + * change. This only happens when switching between half and full + * duplex. + */ +static void restart(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + scc_t *sccp = fep->scc.sccp; + scc_enet_t *ep = fep->scc.ep; + const struct fs_platform_info *fpi = fep->fpi; + u16 paddrh, paddrm, paddrl; + const unsigned char *mac; + int i; + + C32(sccp, scc_gsmrl, SCC_GSMRL_ENR | SCC_GSMRL_ENT); + + /* clear everything (slow & steady does it) */ + for (i = 0; i < sizeof(*ep); i++) + __fs_out8((char *)ep + i, 0); + + /* point to bds */ + W16(ep, sen_genscc.scc_rbase, fep->ring_mem_addr); + W16(ep, sen_genscc.scc_tbase, + fep->ring_mem_addr + sizeof(cbd_t) * fpi->rx_ring); + + /* Initialize function code registers for big-endian. + */ + W8(ep, sen_genscc.scc_rfcr, SCC_EB); + W8(ep, sen_genscc.scc_tfcr, SCC_EB); + + /* Set maximum bytes per receive buffer. + * This appears to be an Ethernet frame size, not the buffer + * fragment size. It must be a multiple of four. + */ + W16(ep, sen_genscc.scc_mrblr, 0x5f0); + + /* Set CRC preset and mask. + */ + W32(ep, sen_cpres, 0xffffffff); + W32(ep, sen_cmask, 0xdebb20e3); + + W32(ep, sen_crcec, 0); /* CRC Error counter */ + W32(ep, sen_alec, 0); /* alignment error counter */ + W32(ep, sen_disfc, 0); /* discard frame counter */ + + W16(ep, sen_pads, 0x8888); /* Tx short frame pad character */ + W16(ep, sen_retlim, 15); /* Retry limit threshold */ + + W16(ep, sen_maxflr, 0x5ee); /* maximum frame length register */ + + W16(ep, sen_minflr, PKT_MINBUF_SIZE); /* minimum frame length register */ + + W16(ep, sen_maxd1, 0x000005f0); /* maximum DMA1 length */ + W16(ep, sen_maxd2, 0x000005f0); /* maximum DMA2 length */ + + /* Clear hash tables. + */ + W16(ep, sen_gaddr1, 0); + W16(ep, sen_gaddr2, 0); + W16(ep, sen_gaddr3, 0); + W16(ep, sen_gaddr4, 0); + W16(ep, sen_iaddr1, 0); + W16(ep, sen_iaddr2, 0); + W16(ep, sen_iaddr3, 0); + W16(ep, sen_iaddr4, 0); + + /* set address + */ + mac = dev->dev_addr; + paddrh = ((u16) mac[5] << 8) | mac[4]; + paddrm = ((u16) mac[3] << 8) | mac[2]; + paddrl = ((u16) mac[1] << 8) | mac[0]; + + W16(ep, sen_paddrh, paddrh); + W16(ep, sen_paddrm, paddrm); + W16(ep, sen_paddrl, paddrl); + + W16(ep, sen_pper, 0); + W16(ep, sen_taddrl, 0); + W16(ep, sen_taddrm, 0); + W16(ep, sen_taddrh, 0); + + fs_init_bds(dev); + + scc_cr_cmd(fep, CPM_CR_INIT_TRX); + + W16(sccp, scc_scce, 0xffff); + + /* Enable interrupts we wish to service. + */ + W16(sccp, scc_sccm, SCCE_ENET_TXE | SCCE_ENET_RXF | SCCE_ENET_TXB); + + /* Set GSMR_H to enable all normal operating modes. + * Set GSMR_L to enable Ethernet to MC68160. + */ + W32(sccp, scc_gsmrh, 0); + W32(sccp, scc_gsmrl, + SCC_GSMRL_TCI | SCC_GSMRL_TPL_48 | SCC_GSMRL_TPP_10 | + SCC_GSMRL_MODE_ENET); + + /* Set sync/delimiters. + */ + W16(sccp, scc_dsr, 0xd555); + + /* Set processing mode. Use Ethernet CRC, catch broadcast, and + * start frame search 22 bit times after RENA. + */ + W16(sccp, scc_psmr, SCC_PSMR_ENCRC | SCC_PSMR_NIB22); + + /* Set full duplex mode if needed */ + if (fep->duplex) + S16(sccp, scc_psmr, SCC_PSMR_LPB | SCC_PSMR_FDE); + + S32(sccp, scc_gsmrl, SCC_GSMRL_ENR | SCC_GSMRL_ENT); +} + +static void stop(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + scc_t *sccp = fep->scc.sccp; + int i; + + for (i = 0; (R16(sccp, scc_sccm) == 0) && i < SCC_RESET_DELAY; i++) + udelay(1); + + if (i == SCC_RESET_DELAY) + printk(KERN_WARNING DRV_MODULE_NAME + ": %s SCC timeout on graceful transmit stop\n", + dev->name); + + W16(sccp, scc_sccm, 0); + C32(sccp, scc_gsmrl, SCC_GSMRL_ENR | SCC_GSMRL_ENT); + + fs_cleanup_bds(dev); +} + +static void pre_request_irq(struct net_device *dev, int irq) +{ + immap_t *immap = fs_enet_immap; + u32 siel; + + /* SIU interrupt */ + if (irq >= SIU_IRQ0 && irq < SIU_LEVEL7) { + + siel = in_be32(&immap->im_siu_conf.sc_siel); + if ((irq & 1) == 0) + siel |= (0x80000000 >> irq); + else + siel &= ~(0x80000000 >> (irq & ~1)); + out_be32(&immap->im_siu_conf.sc_siel, siel); + } +} + +static void post_free_irq(struct net_device *dev, int irq) +{ + /* nothing */ +} + +static void napi_clear_rx_event(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + scc_t *sccp = fep->scc.sccp; + + W16(sccp, scc_scce, SCC_NAPI_RX_EVENT_MSK); +} + +static void napi_enable_rx(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + scc_t *sccp = fep->scc.sccp; + + S16(sccp, scc_sccm, SCC_NAPI_RX_EVENT_MSK); +} + +static void napi_disable_rx(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + scc_t *sccp = fep->scc.sccp; + + C16(sccp, scc_sccm, SCC_NAPI_RX_EVENT_MSK); +} + +static void rx_bd_done(struct net_device *dev) +{ + /* nothing */ +} + +static void tx_kickstart(struct net_device *dev) +{ + /* nothing */ +} + +static u32 get_int_events(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + scc_t *sccp = fep->scc.sccp; + + return (u32) R16(sccp, scc_scce); +} + +static void clear_int_events(struct net_device *dev, u32 int_events) +{ + struct fs_enet_private *fep = netdev_priv(dev); + scc_t *sccp = fep->scc.sccp; + + W16(sccp, scc_scce, int_events & 0xffff); +} + +static void ev_error(struct net_device *dev, u32 int_events) +{ + printk(KERN_WARNING DRV_MODULE_NAME + ": %s SCC ERROR(s) 0x%x\n", dev->name, int_events); +} + +static int get_regs(struct net_device *dev, void *p, int *sizep) +{ + struct fs_enet_private *fep = netdev_priv(dev); + + if (*sizep < sizeof(scc_t) + sizeof(scc_enet_t)) + return -EINVAL; + + memcpy_fromio(p, fep->scc.sccp, sizeof(scc_t)); + p = (char *)p + sizeof(scc_t); + + memcpy_fromio(p, fep->scc.ep, sizeof(scc_enet_t)); + + return 0; +} + +static int get_regs_len(struct net_device *dev) +{ + return sizeof(scc_t) + sizeof(scc_enet_t); +} + +static void tx_restart(struct net_device *dev) +{ + struct fs_enet_private *fep = netdev_priv(dev); + + scc_cr_cmd(fep, CPM_CR_RESTART_TX); +} + +/*************************************************************************/ + +const struct fs_ops fs_scc_ops = { + .setup_data = setup_data, + .cleanup_data = cleanup_data, + .set_multicast_list = set_multicast_list, + .restart = restart, + .stop = stop, + .pre_request_irq = pre_request_irq, + .post_free_irq = post_free_irq, + .napi_clear_rx_event = napi_clear_rx_event, + .napi_enable_rx = napi_enable_rx, + .napi_disable_rx = napi_disable_rx, + .rx_bd_done = rx_bd_done, + .tx_kickstart = tx_kickstart, + .get_int_events = get_int_events, + .clear_int_events = clear_int_events, + .ev_error = ev_error, + .get_regs = get_regs, + .get_regs_len = get_regs_len, + .tx_restart = tx_restart, + .allocate_bd = allocate_bd, + .free_bd = free_bd, +}; diff --git a/drivers/net/fs_enet/mii-bitbang.c b/drivers/net/fs_enet/mii-bitbang.c new file mode 100644 index 000000000000..24a5e2e23d18 --- /dev/null +++ b/drivers/net/fs_enet/mii-bitbang.c @@ -0,0 +1,405 @@ +/* + * Combined Ethernet driver for Motorola MPC8xx and MPC82xx. + * + * Copyright (c) 2003 Intracom S.A. + * by Pantelis Antoniou + * + * 2005 (c) MontaVista Software, Inc. + * Vitaly Bordug + * + * This file is licensed under the terms of the GNU General Public License + * version 2. This program is licensed "as is" without any warranty of any + * kind, whether express or implied. + */ + + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include +#include +#include + +#include "fs_enet.h" + +#ifdef CONFIG_8xx +static int bitbang_prep_bit(u8 **dirp, u8 **datp, u8 *mskp, int port, int bit) +{ + immap_t *im = (immap_t *)fs_enet_immap; + void *dir, *dat, *ppar; + int adv; + u8 msk; + + switch (port) { + case fsiop_porta: + dir = &im->im_ioport.iop_padir; + dat = &im->im_ioport.iop_padat; + ppar = &im->im_ioport.iop_papar; + break; + + case fsiop_portb: + dir = &im->im_cpm.cp_pbdir; + dat = &im->im_cpm.cp_pbdat; + ppar = &im->im_cpm.cp_pbpar; + break; + + case fsiop_portc: + dir = &im->im_ioport.iop_pcdir; + dat = &im->im_ioport.iop_pcdat; + ppar = &im->im_ioport.iop_pcpar; + break; + + case fsiop_portd: + dir = &im->im_ioport.iop_pddir; + dat = &im->im_ioport.iop_pddat; + ppar = &im->im_ioport.iop_pdpar; + break; + + case fsiop_porte: + dir = &im->im_cpm.cp_pedir; + dat = &im->im_cpm.cp_pedat; + ppar = &im->im_cpm.cp_pepar; + break; + + default: + printk(KERN_ERR DRV_MODULE_NAME + "Illegal port value %d!\n", port); + return -EINVAL; + } + + adv = bit >> 3; + dir = (char *)dir + adv; + dat = (char *)dat + adv; + ppar = (char *)ppar + adv; + + msk = 1 << (7 - (bit & 7)); + if ((in_8(ppar) & msk) != 0) { + printk(KERN_ERR DRV_MODULE_NAME + "pin %d on port %d is not general purpose!\n", bit, port); + return -EINVAL; + } + + *dirp = dir; + *datp = dat; + *mskp = msk; + + return 0; +} +#endif + +#ifdef CONFIG_8260 +static int bitbang_prep_bit(u8 **dirp, u8 **datp, u8 *mskp, int port, int bit) +{ + iop_cpm2_t *io = &((cpm2_map_t *)fs_enet_immap)->im_ioport; + void *dir, *dat, *ppar; + int adv; + u8 msk; + + switch (port) { + case fsiop_porta: + dir = &io->iop_pdira; + dat = &io->iop_pdata; + ppar = &io->iop_ppara; + break; + + case fsiop_portb: + dir = &io->iop_pdirb; + dat = &io->iop_pdatb; + ppar = &io->iop_pparb; + break; + + case fsiop_portc: + dir = &io->iop_pdirc; + dat = &io->iop_pdatc; + ppar = &io->iop_pparc; + break; + + case fsiop_portd: + dir = &io->iop_pdird; + dat = &io->iop_pdatd; + ppar = &io->iop_ppard; + break; + + default: + printk(KERN_ERR DRV_MODULE_NAME + "Illegal port value %d!\n", port); + return -EINVAL; + } + + adv = bit >> 3; + dir = (char *)dir + adv; + dat = (char *)dat + adv; + ppar = (char *)ppar + adv; + + msk = 1 << (7 - (bit & 7)); + if ((in_8(ppar) & msk) != 0) { + printk(KERN_ERR DRV_MODULE_NAME + "pin %d on port %d is not general purpose!\n", bit, port); + return -EINVAL; + } + + *dirp = dir; + *datp = dat; + *mskp = msk; + + return 0; +} +#endif + +static inline void bb_set(u8 *p, u8 m) +{ + out_8(p, in_8(p) | m); +} + +static inline void bb_clr(u8 *p, u8 m) +{ + out_8(p, in_8(p) & ~m); +} + +static inline int bb_read(u8 *p, u8 m) +{ + return (in_8(p) & m) != 0; +} + +static inline void mdio_active(struct fs_enet_mii_bus *bus) +{ + bb_set(bus->bitbang.mdio_dir, bus->bitbang.mdio_msk); +} + +static inline void mdio_tristate(struct fs_enet_mii_bus *bus) +{ + bb_clr(bus->bitbang.mdio_dir, bus->bitbang.mdio_msk); +} + +static inline int mdio_read(struct fs_enet_mii_bus *bus) +{ + return bb_read(bus->bitbang.mdio_dat, bus->bitbang.mdio_msk); +} + +static inline void mdio(struct fs_enet_mii_bus *bus, int what) +{ + if (what) + bb_set(bus->bitbang.mdio_dat, bus->bitbang.mdio_msk); + else + bb_clr(bus->bitbang.mdio_dat, bus->bitbang.mdio_msk); +} + +static inline void mdc(struct fs_enet_mii_bus *bus, int what) +{ + if (what) + bb_set(bus->bitbang.mdc_dat, bus->bitbang.mdc_msk); + else + bb_clr(bus->bitbang.mdc_dat, bus->bitbang.mdc_msk); +} + +static inline void mii_delay(struct fs_enet_mii_bus *bus) +{ + udelay(bus->bus_info->i.bitbang.delay); +} + +/* Utility to send the preamble, address, and register (common to read and write). */ +static void bitbang_pre(struct fs_enet_mii_bus *bus, int read, u8 addr, u8 reg) +{ + int j; + + /* + * Send a 32 bit preamble ('1's) with an extra '1' bit for good measure. + * The IEEE spec says this is a PHY optional requirement. The AMD + * 79C874 requires one after power up and one after a MII communications + * error. This means that we are doing more preambles than we need, + * but it is safer and will be much more robust. + */ + + mdio_active(bus); + mdio(bus, 1); + for (j = 0; j < 32; j++) { + mdc(bus, 0); + mii_delay(bus); + mdc(bus, 1); + mii_delay(bus); + } + + /* send the start bit (01) and the read opcode (10) or write (10) */ + mdc(bus, 0); + mdio(bus, 0); + mii_delay(bus); + mdc(bus, 1); + mii_delay(bus); + mdc(bus, 0); + mdio(bus, 1); + mii_delay(bus); + mdc(bus, 1); + mii_delay(bus); + mdc(bus, 0); + mdio(bus, read); + mii_delay(bus); + mdc(bus, 1); + mii_delay(bus); + mdc(bus, 0); + mdio(bus, !read); + mii_delay(bus); + mdc(bus, 1); + mii_delay(bus); + + /* send the PHY address */ + for (j = 0; j < 5; j++) { + mdc(bus, 0); + mdio(bus, (addr & 0x10) != 0); + mii_delay(bus); + mdc(bus, 1); + mii_delay(bus); + addr <<= 1; + } + + /* send the register address */ + for (j = 0; j < 5; j++) { + mdc(bus, 0); + mdio(bus, (reg & 0x10) != 0); + mii_delay(bus); + mdc(bus, 1); + mii_delay(bus); + reg <<= 1; + } +} + +static int mii_read(struct fs_enet_mii_bus *bus, int phy_id, int location) +{ + u16 rdreg; + int ret, j; + u8 addr = phy_id & 0xff; + u8 reg = location & 0xff; + + bitbang_pre(bus, 1, addr, reg); + + /* tri-state our MDIO I/O pin so we can read */ + mdc(bus, 0); + mdio_tristate(bus); + mii_delay(bus); + mdc(bus, 1); + mii_delay(bus); + + /* check the turnaround bit: the PHY should be driving it to zero */ + if (mdio_read(bus) != 0) { + /* PHY didn't drive TA low */ + for (j = 0; j < 32; j++) { + mdc(bus, 0); + mii_delay(bus); + mdc(bus, 1); + mii_delay(bus); + } + ret = -1; + goto out; + } + + mdc(bus, 0); + mii_delay(bus); + + /* read 16 bits of register data, MSB first */ + rdreg = 0; + for (j = 0; j < 16; j++) { + mdc(bus, 1); + mii_delay(bus); + rdreg <<= 1; + rdreg |= mdio_read(bus); + mdc(bus, 0); + mii_delay(bus); + } + + mdc(bus, 1); + mii_delay(bus); + mdc(bus, 0); + mii_delay(bus); + mdc(bus, 1); + mii_delay(bus); + + ret = rdreg; +out: + return ret; +} + +static void mii_write(struct fs_enet_mii_bus *bus, int phy_id, int location, int val) +{ + int j; + u8 addr = phy_id & 0xff; + u8 reg = location & 0xff; + u16 value = val & 0xffff; + + bitbang_pre(bus, 0, addr, reg); + + /* send the turnaround (10) */ + mdc(bus, 0); + mdio(bus, 1); + mii_delay(bus); + mdc(bus, 1); + mii_delay(bus); + mdc(bus, 0); + mdio(bus, 0); + mii_delay(bus); + mdc(bus, 1); + mii_delay(bus); + + /* write 16 bits of register data, MSB first */ + for (j = 0; j < 16; j++) { + mdc(bus, 0); + mdio(bus, (value & 0x8000) != 0); + mii_delay(bus); + mdc(bus, 1); + mii_delay(bus); + value <<= 1; + } + + /* + * Tri-state the MDIO line. + */ + mdio_tristate(bus); + mdc(bus, 0); + mii_delay(bus); + mdc(bus, 1); + mii_delay(bus); +} + +int fs_mii_bitbang_init(struct fs_enet_mii_bus *bus) +{ + const struct fs_mii_bus_info *bi = bus->bus_info; + int r; + + r = bitbang_prep_bit(&bus->bitbang.mdio_dir, + &bus->bitbang.mdio_dat, + &bus->bitbang.mdio_msk, + bi->i.bitbang.mdio_port, + bi->i.bitbang.mdio_bit); + if (r != 0) + return r; + + r = bitbang_prep_bit(&bus->bitbang.mdc_dir, + &bus->bitbang.mdc_dat, + &bus->bitbang.mdc_msk, + bi->i.bitbang.mdc_port, + bi->i.bitbang.mdc_bit); + if (r != 0) + return r; + + bus->mii_read = mii_read; + bus->mii_write = mii_write; + + return 0; +} diff --git a/drivers/net/fs_enet/mii-fixed.c b/drivers/net/fs_enet/mii-fixed.c new file mode 100644 index 000000000000..b3e192d612e5 --- /dev/null +++ b/drivers/net/fs_enet/mii-fixed.c @@ -0,0 +1,92 @@ +/* + * Combined Ethernet driver for Motorola MPC8xx and MPC82xx. + * + * Copyright (c) 2003 Intracom S.A. + * by Pantelis Antoniou + * + * 2005 (c) MontaVista Software, Inc. + * Vitaly Bordug + * + * This file is licensed under the terms of the GNU General Public License + * version 2. This program is licensed "as is" without any warranty of any + * kind, whether express or implied. + */ + + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include +#include +#include + +#include "fs_enet.h" + +static const u16 mii_regs[7] = { + 0x3100, + 0x786d, + 0x0fff, + 0x0fff, + 0x01e1, + 0x45e1, + 0x0003, +}; + +static int mii_read(struct fs_enet_mii_bus *bus, int phy_id, int location) +{ + int ret = 0; + + if ((unsigned int)location >= ARRAY_SIZE(mii_regs)) + return -1; + + if (location != 5) + ret = mii_regs[location]; + else + ret = bus->fixed.lpa; + + return ret; +} + +static void mii_write(struct fs_enet_mii_bus *bus, int phy_id, int location, int val) +{ + /* do nothing */ +} + +int fs_mii_fixed_init(struct fs_enet_mii_bus *bus) +{ + const struct fs_mii_bus_info *bi = bus->bus_info; + + bus->fixed.lpa = 0x45e1; /* default 100Mb, full duplex */ + + /* if speed is fixed at 10Mb, remove 100Mb modes */ + if (bi->i.fixed.speed == 10) + bus->fixed.lpa &= ~LPA_100; + + /* if duplex is half, remove full duplex modes */ + if (bi->i.fixed.duplex == 0) + bus->fixed.lpa &= ~LPA_DUPLEX; + + bus->mii_read = mii_read; + bus->mii_write = mii_write; + + return 0; +} diff --git a/include/linux/fs_enet_pd.h b/include/linux/fs_enet_pd.h new file mode 100644 index 000000000000..bef23bbf8690 --- /dev/null +++ b/include/linux/fs_enet_pd.h @@ -0,0 +1,136 @@ +/* + * Platform information definitions for the + * universal Freescale Ethernet driver. + * + * Copyright (c) 2003 Intracom S.A. + * by Pantelis Antoniou + * + * 2005 (c) MontaVista Software, Inc. + * Vitaly Bordug + * + * This file is licensed under the terms of the GNU General Public License + * version 2. This program is licensed "as is" without any warranty of any + * kind, whether express or implied. + */ + +#ifndef FS_ENET_PD_H +#define FS_ENET_PD_H + +#include +#include + +#define FS_ENET_NAME "fs_enet" + +enum fs_id { + fsid_fec1, + fsid_fec2, + fsid_fcc1, + fsid_fcc2, + fsid_fcc3, + fsid_scc1, + fsid_scc2, + fsid_scc3, + fsid_scc4, +}; + +#define FS_MAX_INDEX 9 + +static inline int fs_get_fec_index(enum fs_id id) +{ + if (id >= fsid_fec1 && id <= fsid_fec2) + return id - fsid_fec1; + return -1; +} + +static inline int fs_get_fcc_index(enum fs_id id) +{ + if (id >= fsid_fcc1 && id <= fsid_fcc3) + return id - fsid_fcc1; + return -1; +} + +static inline int fs_get_scc_index(enum fs_id id) +{ + if (id >= fsid_scc1 && id <= fsid_scc4) + return id - fsid_scc1; + return -1; +} + +enum fs_mii_method { + fsmii_fixed, + fsmii_fec, + fsmii_bitbang, +}; + +enum fs_ioport { + fsiop_porta, + fsiop_portb, + fsiop_portc, + fsiop_portd, + fsiop_porte, +}; + +struct fs_mii_bus_info { + int method; /* mii method */ + int id; /* the id of the mii_bus */ + int disable_aneg; /* if the controller needs to negothiate speed & duplex */ + int lpa; /* the default board-specific vallues will be applied otherwise */ + + union { + struct { + int duplex; + int speed; + } fixed; + + struct { + /* nothing */ + } fec; + + struct { + /* nothing */ + } scc; + + struct { + int mdio_port; /* port & bit for MDIO */ + int mdio_bit; + int mdc_port; /* port & bit for MDC */ + int mdc_bit; + int delay; /* delay in us */ + } bitbang; + } i; +}; + +struct fs_platform_info { + + void(*init_ioports)(void); + /* device specific information */ + int fs_no; /* controller index */ + + u32 cp_page; /* CPM page */ + u32 cp_block; /* CPM sblock */ + + u32 clk_trx; /* some stuff for pins & mux configuration*/ + u32 clk_route; + u32 clk_mask; + + u32 mem_offset; + u32 dpram_offset; + u32 fcc_regs_c; + + u32 device_flags; + + int phy_addr; /* the phy address (-1 no phy) */ + int phy_irq; /* the phy irq (if it exists) */ + + const struct fs_mii_bus_info *bus_info; + + int rx_ring, tx_ring; /* number of buffers on rx */ + __u8 macaddr[6]; /* mac address */ + int rx_copybreak; /* limit we copy small frames */ + int use_napi; /* use NAPI */ + int napi_weight; /* NAPI weight */ + + int use_rmii; /* use RMII mode */ +}; + +#endif -- cgit v1.2.3 From 332bf92b3338e140cbcfc25f69911e8ca59788c7 Mon Sep 17 00:00:00 2001 From: Hideki Iwamoto Date: Sun, 25 Sep 2005 16:56:43 +0200 Subject: [PATCH] i2c: Fix union i2c_smbus_data definition The i2c_smbus_data union block member has a comment stating that an extra byte is required for SMBus Block Process Call transactions. This has been true for three weeks around June 2002, but no more since, so it is about time that we drop this comment and fix the definition. From: Hideki Iwamoto Signed-off-by: Jean Delvare Signed-off-by: Greg Kroah-Hartman include/linux/i2c.h | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) --- include/linux/i2c.h | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) (limited to 'include/linux') diff --git a/include/linux/i2c.h b/include/linux/i2c.h index 3d49a305bf88..17d63c1efdce 100644 --- a/include/linux/i2c.h +++ b/include/linux/i2c.h @@ -454,8 +454,7 @@ struct i2c_msg { union i2c_smbus_data { __u8 byte; __u16 word; - __u8 block[I2C_SMBUS_BLOCK_MAX + 3]; /* block[0] is used for length */ - /* one more for read length in block process call */ + __u8 block[I2C_SMBUS_BLOCK_MAX + 2]; /* block[0] is used for length */ /* and one more for PEC */ }; -- cgit v1.2.3 From bf813b314a2271c3f3903eb3279ebf5e09b3d27a Mon Sep 17 00:00:00 2001 From: Jean Delvare Date: Fri, 7 Oct 2005 23:09:04 +0200 Subject: [PATCH] i2c: Drop useless CVS revision IDs CVS revision IDs are totally useless and irrelevant by now. Signed-off-by: Jean Delvare Signed-off-by: Greg Kroah-Hartman --- include/linux/i2c-algo-bit.h | 2 -- include/linux/i2c-algo-pcf.h | 2 -- include/linux/i2c-dev.h | 2 -- include/linux/i2c.h | 2 -- 4 files changed, 8 deletions(-) (limited to 'include/linux') diff --git a/include/linux/i2c-algo-bit.h b/include/linux/i2c-algo-bit.h index 110904481238..fb592147caec 100644 --- a/include/linux/i2c-algo-bit.h +++ b/include/linux/i2c-algo-bit.h @@ -21,8 +21,6 @@ /* With some changes from Kyösti Mälkki and even Frodo Looijaard */ -/* $Id: i2c-algo-bit.h,v 1.10 2003/01/21 08:08:16 kmalkki Exp $ */ - #ifndef _LINUX_I2C_ALGO_BIT_H #define _LINUX_I2C_ALGO_BIT_H diff --git a/include/linux/i2c-algo-pcf.h b/include/linux/i2c-algo-pcf.h index 2a508562255f..a0e534b334a9 100644 --- a/include/linux/i2c-algo-pcf.h +++ b/include/linux/i2c-algo-pcf.h @@ -22,8 +22,6 @@ /* With some changes from Kyösti Mälkki and even Frodo Looijaard */ -/* $Id: i2c-algo-pcf.h,v 1.8 2003/01/21 08:08:16 kmalkki Exp $ */ - #ifndef _LINUX_I2C_ALGO_PCF_H #define _LINUX_I2C_ALGO_PCF_H diff --git a/include/linux/i2c-dev.h b/include/linux/i2c-dev.h index 541695679762..81c229a0fbca 100644 --- a/include/linux/i2c-dev.h +++ b/include/linux/i2c-dev.h @@ -19,8 +19,6 @@ Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. */ -/* $Id: i2c-dev.h,v 1.13 2003/01/21 08:08:16 kmalkki Exp $ */ - #ifndef _LINUX_I2C_DEV_H #define _LINUX_I2C_DEV_H diff --git a/include/linux/i2c.h b/include/linux/i2c.h index 17d63c1efdce..ec8188414506 100644 --- a/include/linux/i2c.h +++ b/include/linux/i2c.h @@ -23,8 +23,6 @@ /* With some changes from Kyösti Mälkki and Frodo Looijaard */ -/* $Id: i2c.h,v 1.68 2003/01/21 08:08:16 kmalkki Exp $ */ - #ifndef _LINUX_I2C_H #define _LINUX_I2C_H -- cgit v1.2.3 From 31ec5bc57146a479fac6f6878146059180413e43 Mon Sep 17 00:00:00 2001 From: Jean Delvare Date: Sat, 8 Oct 2005 00:04:13 +0200 Subject: [PATCH] i2c: Fix misplaced i2c.h comment Fix a misplaced comment in i2c.h. Spotted by Hideki Iwamoto. Signed-off-by: Jean Delvare Signed-off-by: Greg Kroah-Hartman --- include/linux/i2c.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include/linux') diff --git a/include/linux/i2c.h b/include/linux/i2c.h index ec8188414506..6a1f3424f10f 100644 --- a/include/linux/i2c.h +++ b/include/linux/i2c.h @@ -92,10 +92,10 @@ extern s32 i2c_smbus_write_byte_data(struct i2c_client * client, extern s32 i2c_smbus_read_word_data(struct i2c_client * client, u8 command); extern s32 i2c_smbus_write_word_data(struct i2c_client * client, u8 command, u16 value); -/* Returns the number of bytes transferred */ extern s32 i2c_smbus_write_block_data(struct i2c_client * client, u8 command, u8 length, u8 *values); +/* Returns the number of read bytes */ extern s32 i2c_smbus_read_i2c_block_data(struct i2c_client * client, u8 command, u8 *values); -- cgit v1.2.3 From 80ce3b7d0f52877b80cddc3ace8b332d888f0131 Mon Sep 17 00:00:00 2001 From: Jean Delvare Date: Sat, 8 Oct 2005 00:06:09 +0200 Subject: [PATCH] i2c: Drop out-of-date, colliding ioctl definitions Delete 2 out-of-date, colliding ioctl defines. I2C_UDELAY and I2C_MDELAY are supposed to be used by i2c-algo-bit, but actually aren't (and I suspect never were). Moreover, their values are the same as I2C_FUNCS and I2C_SLAVE_FORCE, respectively, which *are* widely used. Signed-off-by: Jean Delvare Signed-off-by: Greg Kroah-Hartman --- include/linux/i2c.h | 5 ----- 1 file changed, 5 deletions(-) (limited to 'include/linux') diff --git a/include/linux/i2c.h b/include/linux/i2c.h index 6a1f3424f10f..74103d0a8cd2 100644 --- a/include/linux/i2c.h +++ b/include/linux/i2c.h @@ -503,11 +503,6 @@ union i2c_smbus_data { #define I2C_SMBUS 0x0720 /* SMBus-level access */ -/* ... algo-bit.c recognizes */ -#define I2C_UDELAY 0x0705 /* set delay in microsecs between each */ - /* written byte (except address) */ -#define I2C_MDELAY 0x0706 /* millisec delay between written bytes */ - /* ----- I2C-DEV: char device interface stuff ------------------------- */ #define I2C_MAJOR 89 /* Device major number */ -- cgit v1.2.3 From d3554b4a2fb0e2229eb0d3fa9ece5b2f0b906d3e Mon Sep 17 00:00:00 2001 From: Jean Delvare Date: Sat, 8 Oct 2005 00:14:17 +0200 Subject: [PATCH] i2c: Drop unused per-i2c-algorithm adapter max There are no more per-i2c-algorithm adapter max. Last time there were was in July 1999. Signed-off-by: Jean Delvare Signed-off-by: Greg Kroah-Hartman --- include/linux/i2c-algo-bit.h | 2 -- include/linux/i2c-algo-pca.h | 2 -- include/linux/i2c-algo-pcf.h | 2 -- 3 files changed, 6 deletions(-) (limited to 'include/linux') diff --git a/include/linux/i2c-algo-bit.h b/include/linux/i2c-algo-bit.h index fb592147caec..c0e7fab28ce3 100644 --- a/include/linux/i2c-algo-bit.h +++ b/include/linux/i2c-algo-bit.h @@ -44,8 +44,6 @@ struct i2c_algo_bit_data { int timeout; /* in jiffies */ }; -#define I2C_BIT_ADAP_MAX 16 - int i2c_bit_add_bus(struct i2c_adapter *); int i2c_bit_del_bus(struct i2c_adapter *); diff --git a/include/linux/i2c-algo-pca.h b/include/linux/i2c-algo-pca.h index 941b786c5732..226693e0d88b 100644 --- a/include/linux/i2c-algo-pca.h +++ b/include/linux/i2c-algo-pca.h @@ -9,8 +9,6 @@ struct i2c_algo_pca_data { int (*wait_for_interrupt) (struct i2c_algo_pca_data *adap); }; -#define I2C_PCA_ADAP_MAX 16 - int i2c_pca_add_bus(struct i2c_adapter *); int i2c_pca_del_bus(struct i2c_adapter *); diff --git a/include/linux/i2c-algo-pcf.h b/include/linux/i2c-algo-pcf.h index a0e534b334a9..18b0adf57a3d 100644 --- a/include/linux/i2c-algo-pcf.h +++ b/include/linux/i2c-algo-pcf.h @@ -39,8 +39,6 @@ struct i2c_algo_pcf_data { int timeout; }; -#define I2C_PCF_ADAP_MAX 16 - int i2c_pcf_add_bus(struct i2c_adapter *); int i2c_pcf_del_bus(struct i2c_adapter *); -- cgit v1.2.3 From 30dac7469741906436b50f9413dccd446366d371 Mon Sep 17 00:00:00 2001 From: Jean Delvare Date: Sat, 8 Oct 2005 00:15:59 +0200 Subject: [PATCH] i2c: Drop I2C_SMBUS_I2C_BLOCK_MAX Drop I2C_SMBUS_I2C_BLOCK_MAX, use I2C_SMBUS_BLOCK_MAX instead. I2C_SMBUS_I2C_BLOCK_MAX has always been defined to the same value as I2C_SMBUS_BLOCK_MAX, and this will never change: setting it to a lower value would make no sense, setting it to a higher value would break i2c_smbus_data compatibility. There is no point in changing i2c_smbus_data to support larger block transactions in SMBus mode, as no SMBus hardware supports more than 32 byte blocks. Thus, for larger transactions, direct I2C transfers are the way to go. Signed-off-by: Jean Delvare Signed-off-by: Greg Kroah-Hartman --- drivers/i2c/chips/eeprom.c | 4 ++-- drivers/i2c/i2c-core.c | 8 ++++---- include/linux/i2c.h | 1 - 3 files changed, 6 insertions(+), 7 deletions(-) (limited to 'include/linux') diff --git a/drivers/i2c/chips/eeprom.c b/drivers/i2c/chips/eeprom.c index d58403a47908..7fb739c43935 100644 --- a/drivers/i2c/chips/eeprom.c +++ b/drivers/i2c/chips/eeprom.c @@ -88,8 +88,8 @@ static void eeprom_update_client(struct i2c_client *client, u8 slice) dev_dbg(&client->dev, "Starting eeprom update, slice %u\n", slice); if (i2c_check_functionality(client->adapter, I2C_FUNC_SMBUS_READ_I2C_BLOCK)) { - for (i = slice << 5; i < (slice + 1) << 5; i += I2C_SMBUS_I2C_BLOCK_MAX) - if (i2c_smbus_read_i2c_block_data(client, i, data->data + i) != I2C_SMBUS_I2C_BLOCK_MAX) + for (i = slice << 5; i < (slice + 1) << 5; i += I2C_SMBUS_BLOCK_MAX) + if (i2c_smbus_read_i2c_block_data(client, i, data->data + i) != I2C_SMBUS_BLOCK_MAX) goto exit; } else { if (i2c_smbus_write_byte(client, slice << 5)) { diff --git a/drivers/i2c/i2c-core.c b/drivers/i2c/i2c-core.c index 9583a54ce16b..15354176ff63 100644 --- a/drivers/i2c/i2c-core.c +++ b/drivers/i2c/i2c-core.c @@ -1108,10 +1108,10 @@ static s32 i2c_smbus_xfer_emulated(struct i2c_adapter * adapter, u16 addr, return -1; case I2C_SMBUS_I2C_BLOCK_DATA: if (read_write == I2C_SMBUS_READ) { - msg[1].len = I2C_SMBUS_I2C_BLOCK_MAX; + msg[1].len = I2C_SMBUS_BLOCK_MAX; } else { msg[0].len = data->block[0] + 1; - if (msg[0].len > I2C_SMBUS_I2C_BLOCK_MAX + 1) { + if (msg[0].len > I2C_SMBUS_BLOCK_MAX + 1) { dev_err(&adapter->dev, "i2c_smbus_xfer_emulated called with " "invalid block write size (%d)\n", data->block[0]); @@ -1144,8 +1144,8 @@ static s32 i2c_smbus_xfer_emulated(struct i2c_adapter * adapter, u16 addr, break; case I2C_SMBUS_I2C_BLOCK_DATA: /* fixed at 32 for now */ - data->block[0] = I2C_SMBUS_I2C_BLOCK_MAX; - for (i = 0; i < I2C_SMBUS_I2C_BLOCK_MAX; i++) + data->block[0] = I2C_SMBUS_BLOCK_MAX; + for (i = 0; i < I2C_SMBUS_BLOCK_MAX; i++) data->block[i+1] = msgbuf1[i]; break; } diff --git a/include/linux/i2c.h b/include/linux/i2c.h index 74103d0a8cd2..64c13c042f9d 100644 --- a/include/linux/i2c.h +++ b/include/linux/i2c.h @@ -448,7 +448,6 @@ struct i2c_msg { * Data for SMBus Messages */ #define I2C_SMBUS_BLOCK_MAX 32 /* As specified in SMBus standard */ -#define I2C_SMBUS_I2C_BLOCK_MAX 32 /* Not specified but we use same structure */ union i2c_smbus_data { __u8 byte; __u16 word; -- cgit v1.2.3 From 4d4e5ce8648561b964699afb2df5e7268a84599b Mon Sep 17 00:00:00 2001 From: Alessandro Zummo Date: Mon, 17 Oct 2005 23:04:42 +0200 Subject: [PATCH] i2c: New Xicor X1205 RTC driver New driver for the Xicor X1205 RTC chip. Signed-off-by: Alessandro Zummo Signed-off-by: Jean Delvare Signed-off-by: Greg Kroah-Hartman --- Documentation/i2c/chips/x1205 | 38 +++ drivers/i2c/chips/Kconfig | 9 + drivers/i2c/chips/Makefile | 1 + drivers/i2c/chips/x1205.c | 698 ++++++++++++++++++++++++++++++++++++++++++ include/linux/x1205.h | 31 ++ 5 files changed, 777 insertions(+) create mode 100644 Documentation/i2c/chips/x1205 create mode 100644 drivers/i2c/chips/x1205.c create mode 100644 include/linux/x1205.h (limited to 'include/linux') diff --git a/Documentation/i2c/chips/x1205 b/Documentation/i2c/chips/x1205 new file mode 100644 index 000000000000..09407c991fe5 --- /dev/null +++ b/Documentation/i2c/chips/x1205 @@ -0,0 +1,38 @@ +Kernel driver x1205 +=================== + +Supported chips: + * Xicor X1205 RTC + Prefix: 'x1205' + Addresses scanned: none + Datasheet: http://www.intersil.com/cda/deviceinfo/0,1477,X1205,00.html + +Authors: + Karen Spearel , + Alessandro Zummo + +Description +----------- + +This module aims to provide complete access to the Xicor X1205 RTC. +Recently Xicor has merged with Intersil, but the chip is +still sold under the Xicor brand. + +This chip is located at address 0x6f and uses a 2-byte register addressing. +Two bytes need to be written to read a single register, while most +other chips just require one and take the second one as the data +to be written. To prevent corrupting unknown chips, the user must +explicitely set the probe parameter. + +example: + +modprobe x1205 probe=0,0x6f + +The module supports one more option, hctosys, which is used to set the +software clock from the x1205. On systems where the x1205 is the +only hardware rtc, this parameter could be used to achieve a correct +date/time earlier in the system boot sequence. + +example: + +modprobe x1205 probe=0,0x6f hctosys=1 diff --git a/drivers/i2c/chips/Kconfig b/drivers/i2c/chips/Kconfig index 6bd44a44cd28..f9fae28f5612 100644 --- a/drivers/i2c/chips/Kconfig +++ b/drivers/i2c/chips/Kconfig @@ -126,4 +126,13 @@ config SENSORS_MAX6875 This driver can also be built as a module. If so, the module will be called max6875. +config RTC_X1205_I2C + tristate "Xicor X1205 RTC chip" + depends on I2C && EXPERIMENTAL + help + If you say yes here you get support for the Xicor X1205 RTC chip. + + This driver can also be built as a module. If so, the module + will be called x1205. + endmenu diff --git a/drivers/i2c/chips/Makefile b/drivers/i2c/chips/Makefile index a876dd42b860..46178b57b1f1 100644 --- a/drivers/i2c/chips/Makefile +++ b/drivers/i2c/chips/Makefile @@ -13,6 +13,7 @@ obj-$(CONFIG_SENSORS_PCF8591) += pcf8591.o obj-$(CONFIG_SENSORS_RTC8564) += rtc8564.o obj-$(CONFIG_ISP1301_OMAP) += isp1301_omap.o obj-$(CONFIG_TPS65010) += tps65010.o +obj-$(CONFIG_RTC_X1205_I2C) += x1205.o ifeq ($(CONFIG_I2C_DEBUG_CHIP),y) EXTRA_CFLAGS += -DDEBUG diff --git a/drivers/i2c/chips/x1205.c b/drivers/i2c/chips/x1205.c new file mode 100644 index 000000000000..7da366cdc18c --- /dev/null +++ b/drivers/i2c/chips/x1205.c @@ -0,0 +1,698 @@ +/* + * x1205.c - An i2c driver for the Xicor X1205 RTC + * Copyright 2004 Karen Spearel + * Copyright 2005 Alessandro Zummo + * + * please send all reports to: + * kas11 at tampabay dot rr dot com + * a dot zummo at towertech dot it + * + * based on the other drivers in this same directory. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + */ + +#include +#include +#include +#include +#include +#include +#include +#include + +#include + +#define DRV_VERSION "0.9.9" + +/* Addresses to scan: none. This chip is located at + * 0x6f and uses a two bytes register addressing. + * Two bytes need to be written to read a single register, + * while most other chips just require one and take the second + * one as the data to be written. To prevent corrupting + * unknown chips, the user must explicitely set the probe parameter. + */ + +static unsigned short normal_i2c[] = { I2C_CLIENT_END }; + +/* Insmod parameters */ +I2C_CLIENT_INSMOD; +I2C_CLIENT_MODULE_PARM(hctosys, + "Set the system time from the hardware clock upon initialization"); + +/* offsets into CCR area */ + +#define CCR_SEC 0 +#define CCR_MIN 1 +#define CCR_HOUR 2 +#define CCR_MDAY 3 +#define CCR_MONTH 4 +#define CCR_YEAR 5 +#define CCR_WDAY 6 +#define CCR_Y2K 7 + +#define X1205_REG_SR 0x3F /* status register */ +#define X1205_REG_Y2K 0x37 +#define X1205_REG_DW 0x36 +#define X1205_REG_YR 0x35 +#define X1205_REG_MO 0x34 +#define X1205_REG_DT 0x33 +#define X1205_REG_HR 0x32 +#define X1205_REG_MN 0x31 +#define X1205_REG_SC 0x30 +#define X1205_REG_DTR 0x13 +#define X1205_REG_ATR 0x12 +#define X1205_REG_INT 0x11 +#define X1205_REG_0 0x10 +#define X1205_REG_Y2K1 0x0F +#define X1205_REG_DWA1 0x0E +#define X1205_REG_YRA1 0x0D +#define X1205_REG_MOA1 0x0C +#define X1205_REG_DTA1 0x0B +#define X1205_REG_HRA1 0x0A +#define X1205_REG_MNA1 0x09 +#define X1205_REG_SCA1 0x08 +#define X1205_REG_Y2K0 0x07 +#define X1205_REG_DWA0 0x06 +#define X1205_REG_YRA0 0x05 +#define X1205_REG_MOA0 0x04 +#define X1205_REG_DTA0 0x03 +#define X1205_REG_HRA0 0x02 +#define X1205_REG_MNA0 0x01 +#define X1205_REG_SCA0 0x00 + +#define X1205_CCR_BASE 0x30 /* Base address of CCR */ +#define X1205_ALM0_BASE 0x00 /* Base address of ALARM0 */ + +#define X1205_SR_RTCF 0x01 /* Clock failure */ +#define X1205_SR_WEL 0x02 /* Write Enable Latch */ +#define X1205_SR_RWEL 0x04 /* Register Write Enable */ + +#define X1205_DTR_DTR0 0x01 +#define X1205_DTR_DTR1 0x02 +#define X1205_DTR_DTR2 0x04 + +#define X1205_HR_MIL 0x80 /* Set in ccr.hour for 24 hr mode */ + +/* Prototypes */ +static int x1205_attach(struct i2c_adapter *adapter); +static int x1205_detach(struct i2c_client *client); +static int x1205_probe(struct i2c_adapter *adapter, int address, int kind); +static int x1205_command(struct i2c_client *client, unsigned int cmd, + void *arg); + +static struct i2c_driver x1205_driver = { + .owner = THIS_MODULE, + .name = "x1205", + .flags = I2C_DF_NOTIFY, + .attach_adapter = &x1205_attach, + .detach_client = &x1205_detach, +}; + +struct x1205_data { + struct i2c_client client; + struct list_head list; + unsigned int epoch; +}; + +static const unsigned char days_in_mo[] = + { 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 }; + +static LIST_HEAD(x1205_clients); + +/* Workaround until the I2C subsytem will allow to send + * commands to a specific client. This function will send the command + * to the first client. + */ +int x1205_do_command(unsigned int cmd, void *arg) +{ + struct list_head *walk; + struct list_head *tmp; + struct x1205_data *data; + + list_for_each_safe(walk, tmp, &x1205_clients) { + data = list_entry(walk, struct x1205_data, list); + return x1205_command(&data->client, cmd, arg); + } + + return -ENODEV; +} + +#define is_leap(year) \ + ((year) % 4 == 0 && ((year) % 100 != 0 || (year) % 400 == 0)) + +/* make sure the rtc_time values are in bounds */ +static int x1205_validate_tm(struct rtc_time *tm) +{ + int year = tm->tm_year + 1900; + + if ((tm->tm_year < 70) || (tm->tm_year > 255)) + return -EINVAL; + + if ((tm->tm_mon > 11) || (tm->tm_mday == 0)) + return -EINVAL; + + if (tm->tm_mday > days_in_mo[tm->tm_mon] + + ((tm->tm_mon == 1) && is_leap(year))) + return -EINVAL; + + if ((tm->tm_hour >= 24) || (tm->tm_min >= 60) || (tm->tm_sec >= 60)) + return -EINVAL; + + return 0; +} + +/* + * In the routines that deal directly with the x1205 hardware, we use + * rtc_time -- month 0-11, hour 0-23, yr = calendar year-epoch + * Epoch is initialized as 2000. Time is set to UTC. + */ +static int x1205_get_datetime(struct i2c_client *client, struct rtc_time *tm, + u8 reg_base) +{ + unsigned char dt_addr[2] = { 0, reg_base }; + static unsigned char sr_addr[2] = { 0, X1205_REG_SR }; + + unsigned char buf[8], sr; + + struct i2c_msg msgs[] = { + { client->addr, 0, 2, sr_addr }, /* setup read ptr */ + { client->addr, I2C_M_RD, 1, &sr }, /* read status */ + { client->addr, 0, 2, dt_addr }, /* setup read ptr */ + { client->addr, I2C_M_RD, 8, buf }, /* read date */ + }; + + struct x1205_data *data = i2c_get_clientdata(client); + + /* read status register */ + if ((i2c_transfer(client->adapter, &msgs[0], 2)) != 2) { + dev_err(&client->dev, "%s: read error\n", __FUNCTION__); + return -EIO; + } + + /* check for battery failure */ + if (sr & X1205_SR_RTCF) { + dev_warn(&client->dev, + "Clock had a power failure, you must set the date.\n"); + return -EINVAL; + } + + /* read date registers */ + if ((i2c_transfer(client->adapter, &msgs[2], 2)) != 2) { + dev_err(&client->dev, "%s: read error\n", __FUNCTION__); + return -EIO; + } + + dev_dbg(&client->dev, + "%s: raw read data - sec=%02x, min=%02x, hr=%02x, " + "mday=%02x, mon=%02x, year=%02x, wday=%02x, y2k=%02x\n", + __FUNCTION__, + buf[0], buf[1], buf[2], buf[3], + buf[4], buf[5], buf[6], buf[7]); + + tm->tm_sec = BCD2BIN(buf[CCR_SEC]); + tm->tm_min = BCD2BIN(buf[CCR_MIN]); + tm->tm_hour = BCD2BIN(buf[CCR_HOUR] & 0x3F); /* hr is 0-23 */ + tm->tm_mday = BCD2BIN(buf[CCR_MDAY]); + tm->tm_mon = BCD2BIN(buf[CCR_MONTH]); + data->epoch = BCD2BIN(buf[CCR_Y2K]) * 100; + tm->tm_year = BCD2BIN(buf[CCR_YEAR]) + data->epoch - 1900; + tm->tm_wday = buf[CCR_WDAY]; + + dev_dbg(&client->dev, "%s: tm is secs=%d, mins=%d, hours=%d, " + "mday=%d, mon=%d, year=%d, wday=%d\n", + __FUNCTION__, + tm->tm_sec, tm->tm_min, tm->tm_hour, + tm->tm_mday, tm->tm_mon, tm->tm_year, tm->tm_wday); + + return 0; +} + +static int x1205_set_datetime(struct i2c_client *client, struct rtc_time *tm, + int datetoo, u8 reg_base) +{ + int i, err, xfer; + + unsigned char buf[8]; + + static const unsigned char wel[3] = { 0, X1205_REG_SR, + X1205_SR_WEL }; + + static const unsigned char rwel[3] = { 0, X1205_REG_SR, + X1205_SR_WEL | X1205_SR_RWEL }; + + static const unsigned char diswe[3] = { 0, X1205_REG_SR, 0 }; + + struct x1205_data *data = i2c_get_clientdata(client); + + /* check if all values in the tm struct are correct */ + if ((err = x1205_validate_tm(tm)) < 0) + return err; + + dev_dbg(&client->dev, "%s: secs=%d, mins=%d, hours=%d, " + "mday=%d, mon=%d, year=%d, wday=%d\n", + __FUNCTION__, + tm->tm_sec, tm->tm_min, tm->tm_hour, + tm->tm_mday, tm->tm_mon, tm->tm_year, tm->tm_wday); + + buf[CCR_SEC] = BIN2BCD(tm->tm_sec); + buf[CCR_MIN] = BIN2BCD(tm->tm_min); + + /* set hour and 24hr bit */ + buf[CCR_HOUR] = BIN2BCD(tm->tm_hour) | X1205_HR_MIL; + + /* should we also set the date? */ + if (datetoo) { + buf[CCR_MDAY] = BIN2BCD(tm->tm_mday); + + /* month, 0 - 11 */ + buf[CCR_MONTH] = BIN2BCD(tm->tm_mon); + + /* year, since 1900 */ + buf[CCR_YEAR] = BIN2BCD(tm->tm_year + 1900 - data->epoch); + buf[CCR_WDAY] = tm->tm_wday & 0x07; + buf[CCR_Y2K] = BIN2BCD(data->epoch / 100); + } + + /* this sequence is required to unlock the chip */ + xfer = i2c_master_send(client, wel, 3); + if (xfer != 3) { + dev_err(&client->dev, "%s: wel - %d\n", __FUNCTION__, xfer); + return -EIO; + } + + xfer = i2c_master_send(client, rwel, 3); + if (xfer != 3) { + dev_err(&client->dev, "%s: rwel - %d\n", __FUNCTION__, xfer); + return -EIO; + } + + /* write register's data */ + for (i = 0; i < (datetoo ? 8 : 3); i++) { + unsigned char rdata[3] = { 0, reg_base + i, buf[i] }; + + xfer = i2c_master_send(client, rdata, 3); + if (xfer != 3) { + dev_err(&client->dev, + "%s: xfer=%d addr=%02x, data=%02x\n", + __FUNCTION__, + xfer, rdata[1], rdata[2]); + return -EIO; + } + }; + + /* disable further writes */ + xfer = i2c_master_send(client, diswe, 3); + if (xfer != 3) { + dev_err(&client->dev, "%s: diswe - %d\n", __FUNCTION__, xfer); + return -EIO; + } + + return 0; +} + +static int x1205_get_dtrim(struct i2c_client *client, int *trim) +{ + unsigned char dtr; + static unsigned char dtr_addr[2] = { 0, X1205_REG_DTR }; + + struct i2c_msg msgs[] = { + { client->addr, 0, 2, dtr_addr }, /* setup read ptr */ + { client->addr, I2C_M_RD, 1, &dtr }, /* read dtr */ + }; + + /* read dtr register */ + if ((i2c_transfer(client->adapter, &msgs[0], 2)) != 2) { + dev_err(&client->dev, "%s: read error\n", __FUNCTION__); + return -EIO; + } + + dev_dbg(&client->dev, "%s: raw dtr=%x\n", __FUNCTION__, dtr); + + *trim = 0; + + if (dtr & X1205_DTR_DTR0) + *trim += 20; + + if (dtr & X1205_DTR_DTR1) + *trim += 10; + + if (dtr & X1205_DTR_DTR2) + *trim = -*trim; + + return 0; +} + +static int x1205_get_atrim(struct i2c_client *client, int *trim) +{ + s8 atr; + static unsigned char atr_addr[2] = { 0, X1205_REG_ATR }; + + struct i2c_msg msgs[] = { + { client->addr, 0, 2, atr_addr }, /* setup read ptr */ + { client->addr, I2C_M_RD, 1, &atr }, /* read atr */ + }; + + /* read atr register */ + if ((i2c_transfer(client->adapter, &msgs[0], 2)) != 2) { + dev_err(&client->dev, "%s: read error\n", __FUNCTION__); + return -EIO; + } + + dev_dbg(&client->dev, "%s: raw atr=%x\n", __FUNCTION__, atr); + + /* atr is a two's complement value on 6 bits, + * perform sign extension. The formula is + * Catr = (atr * 0.25pF) + 11.00pF. + */ + if (atr & 0x20) + atr |= 0xC0; + + dev_dbg(&client->dev, "%s: raw atr=%x (%d)\n", __FUNCTION__, atr, atr); + + *trim = (atr * 250) + 11000; + + dev_dbg(&client->dev, "%s: real=%d\n", __FUNCTION__, *trim); + + return 0; +} + +static int x1205_hctosys(struct i2c_client *client) +{ + int err; + + struct rtc_time tm; + struct timespec tv; + + err = x1205_command(client, X1205_CMD_GETDATETIME, &tm); + + if (err) { + dev_err(&client->dev, + "Unable to set the system clock\n"); + return err; + } + + /* IMPORTANT: the RTC only stores whole seconds. It is arbitrary + * whether it stores the most close value or the value with partial + * seconds truncated. However, it is important that we use it to store + * the truncated value. This is because otherwise it is necessary, + * in an rtc sync function, to read both xtime.tv_sec and + * xtime.tv_nsec. On some processors (i.e. ARM), an atomic read + * of >32bits is not possible. So storing the most close value would + * slow down the sync API. So here we have the truncated value and + * the best guess is to add 0.5s. + */ + + tv.tv_nsec = NSEC_PER_SEC >> 1; + + /* WARNING: this is not the C library 'mktime' call, it is a built in + * inline function from include/linux/time.h. It expects (requires) + * the month to be in the range 1-12 + */ + + tv.tv_sec = mktime(tm.tm_year + 1900, tm.tm_mon + 1, + tm.tm_mday, tm.tm_hour, + tm.tm_min, tm.tm_sec); + + do_settimeofday(&tv); + + dev_info(&client->dev, + "setting the system clock to %d-%d-%d %d:%d:%d\n", + tm.tm_year + 1900, tm.tm_mon + 1, + tm.tm_mday, tm.tm_hour, tm.tm_min, + tm.tm_sec); + + return 0; +} + +struct x1205_limit +{ + unsigned char reg; + unsigned char mask; + unsigned char min; + unsigned char max; +}; + +static int x1205_validate_client(struct i2c_client *client) +{ + int i, xfer; + + /* Probe array. We will read the register at the specified + * address and check if the given bits are zero. + */ + static const unsigned char probe_zero_pattern[] = { + /* register, mask */ + X1205_REG_SR, 0x18, + X1205_REG_DTR, 0xF8, + X1205_REG_ATR, 0xC0, + X1205_REG_INT, 0x18, + X1205_REG_0, 0xFF, + }; + + static const struct x1205_limit probe_limits_pattern[] = { + /* register, mask, min, max */ + { X1205_REG_Y2K, 0xFF, 19, 20 }, + { X1205_REG_DW, 0xFF, 0, 6 }, + { X1205_REG_YR, 0xFF, 0, 99 }, + { X1205_REG_MO, 0xFF, 0, 12 }, + { X1205_REG_DT, 0xFF, 0, 31 }, + { X1205_REG_HR, 0x7F, 0, 23 }, + { X1205_REG_MN, 0xFF, 0, 59 }, + { X1205_REG_SC, 0xFF, 0, 59 }, + { X1205_REG_Y2K1, 0xFF, 19, 20 }, + { X1205_REG_Y2K0, 0xFF, 19, 20 }, + }; + + /* check that registers have bits a 0 where expected */ + for (i = 0; i < ARRAY_SIZE(probe_zero_pattern); i += 2) { + unsigned char buf; + + unsigned char addr[2] = { 0, probe_zero_pattern[i] }; + + struct i2c_msg msgs[2] = { + { client->addr, 0, 2, addr }, + { client->addr, I2C_M_RD, 1, &buf }, + }; + + xfer = i2c_transfer(client->adapter, msgs, 2); + if (xfer != 2) { + dev_err(&client->adapter->dev, + "%s: could not read register %x\n", + __FUNCTION__, addr[1]); + + return -EIO; + } + + if ((buf & probe_zero_pattern[i+1]) != 0) { + dev_err(&client->adapter->dev, + "%s: register=%02x, zero pattern=%d, value=%x\n", + __FUNCTION__, addr[1], i, buf); + + return -ENODEV; + } + } + + /* check limits (only registers with bcd values) */ + for (i = 0; i < ARRAY_SIZE(probe_limits_pattern); i++) { + unsigned char reg, value; + + unsigned char addr[2] = { 0, probe_limits_pattern[i].reg }; + + struct i2c_msg msgs[2] = { + { client->addr, 0, 2, addr }, + { client->addr, I2C_M_RD, 1, ® }, + }; + + xfer = i2c_transfer(client->adapter, msgs, 2); + + if (xfer != 2) { + dev_err(&client->adapter->dev, + "%s: could not read register %x\n", + __FUNCTION__, addr[1]); + + return -EIO; + } + + value = BCD2BIN(reg & probe_limits_pattern[i].mask); + + if (value > probe_limits_pattern[i].max || + value < probe_limits_pattern[i].min) { + dev_dbg(&client->adapter->dev, + "%s: register=%x, lim pattern=%d, value=%d\n", + __FUNCTION__, addr[1], i, value); + + return -ENODEV; + } + } + + return 0; +} + +static int x1205_attach(struct i2c_adapter *adapter) +{ + dev_dbg(&adapter->dev, "%s\n", __FUNCTION__); + + return i2c_probe(adapter, &addr_data, x1205_probe); +} + +int x1205_direct_attach(int adapter_id, + struct i2c_client_address_data *address_data) +{ + int err; + struct i2c_adapter *adapter = i2c_get_adapter(adapter_id); + + if (adapter) { + err = i2c_probe(adapter, + address_data, x1205_probe); + + i2c_put_adapter(adapter); + + return err; + } + + return -ENODEV; +} + +static int x1205_probe(struct i2c_adapter *adapter, int address, int kind) +{ + struct i2c_client *client; + struct x1205_data *data; + + int err = 0; + + dev_dbg(&adapter->dev, "%s\n", __FUNCTION__); + + if (!i2c_check_functionality(adapter, I2C_FUNC_I2C)) { + err = -ENODEV; + goto exit; + } + + if (!(data = kzalloc(sizeof(struct x1205_data), GFP_KERNEL))) { + err = -ENOMEM; + goto exit; + } + + /* Initialize our structures */ + data->epoch = 2000; + + client = &data->client; + client->addr = address; + client->driver = &x1205_driver; + client->adapter = adapter; + + strlcpy(client->name, "x1205", I2C_NAME_SIZE); + + i2c_set_clientdata(client, data); + + /* Verify the chip is really an X1205 */ + if (kind < 0) { + if (x1205_validate_client(client) < 0) { + err = -ENODEV; + goto exit_kfree; + } + } + + /* Inform the i2c layer */ + if ((err = i2c_attach_client(client))) + goto exit_kfree; + + list_add(&data->list, &x1205_clients); + + dev_info(&client->dev, "chip found, driver version " DRV_VERSION "\n"); + + /* If requested, set the system time */ + if (hctosys) + x1205_hctosys(client); + + return 0; + +exit_kfree: + kfree(data); + +exit: + return err; +} + +static int x1205_detach(struct i2c_client *client) +{ + int err; + struct x1205_data *data = i2c_get_clientdata(client); + + dev_dbg(&client->dev, "%s\n", __FUNCTION__); + + if ((err = i2c_detach_client(client))) + return err; + + list_del(&data->list); + + kfree(data); + + return 0; +} + +static int x1205_command(struct i2c_client *client, unsigned int cmd, + void *param) +{ + if (param == NULL) + return -EINVAL; + + if (!capable(CAP_SYS_TIME)) + return -EACCES; + + dev_dbg(&client->dev, "%s: cmd=%d\n", __FUNCTION__, cmd); + + switch (cmd) { + case X1205_CMD_GETDATETIME: + return x1205_get_datetime(client, param, X1205_CCR_BASE); + + case X1205_CMD_SETTIME: + return x1205_set_datetime(client, param, 0, + X1205_CCR_BASE); + + case X1205_CMD_SETDATETIME: + return x1205_set_datetime(client, param, 1, + X1205_CCR_BASE); + + case X1205_CMD_GETALARM: + return x1205_get_datetime(client, param, X1205_ALM0_BASE); + + case X1205_CMD_SETALARM: + return x1205_set_datetime(client, param, 1, + X1205_ALM0_BASE); + + case X1205_CMD_GETDTRIM: + return x1205_get_dtrim(client, param); + + case X1205_CMD_GETATRIM: + return x1205_get_atrim(client, param); + + default: + return -EINVAL; + } +} + +static int __init x1205_init(void) +{ + return i2c_add_driver(&x1205_driver); +} + +static void __exit x1205_exit(void) +{ + i2c_del_driver(&x1205_driver); +} + +MODULE_AUTHOR( + "Karen Spearel , " + "Alessandro Zummo "); +MODULE_DESCRIPTION("Xicor X1205 RTC driver"); +MODULE_LICENSE("GPL"); +MODULE_VERSION(DRV_VERSION); + +EXPORT_SYMBOL_GPL(x1205_do_command); +EXPORT_SYMBOL_GPL(x1205_direct_attach); + +module_init(x1205_init); +module_exit(x1205_exit); diff --git a/include/linux/x1205.h b/include/linux/x1205.h new file mode 100644 index 000000000000..64fd3af894a5 --- /dev/null +++ b/include/linux/x1205.h @@ -0,0 +1,31 @@ +/* + * x1205.h - defines for drivers/i2c/chips/x1205.c + * Copyright 2004 Karen Spearel + * Copyright 2005 Alessandro Zummo + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + */ + +#ifndef __LINUX_X1205_H__ +#define __LINUX_X1205_H__ + +/* commands */ + +#define X1205_CMD_GETDATETIME 0 +#define X1205_CMD_SETTIME 1 +#define X1205_CMD_SETDATETIME 2 +#define X1205_CMD_GETALARM 3 +#define X1205_CMD_SETALARM 4 +#define X1205_CMD_GETDTRIM 5 +#define X1205_CMD_SETDTRIM 6 +#define X1205_CMD_GETATRIM 7 +#define X1205_CMD_SETATRIM 8 + +extern int x1205_do_command(unsigned int cmd, void *arg); +extern int x1205_direct_attach(int adapter_id, + struct i2c_client_address_data *address_data); + +#endif /* __LINUX_X1205_H__ */ -- cgit v1.2.3 From eb00a28ae1a8fc4b3914f94ab1944396b8dda9fc Mon Sep 17 00:00:00 2001 From: Jean Delvare Date: Wed, 26 Oct 2005 21:20:17 +0200 Subject: [PATCH] i2c: Drop unused parport i2c IDs Drop unused i2c-over-parallel-port i2c IDs: * I2C_HW_B_LPC was never actually used as far as I could search. * I2C_HW_B_ELV and I2C_HW_B_VELLE are no more used since the introduction of the unified i2c-parport driver in Linux 2.6.2. Signed-off-by: Jean Delvare Signed-off-by: Greg Kroah-Hartman --- include/linux/i2c-id.h | 3 --- 1 file changed, 3 deletions(-) (limited to 'include/linux') diff --git a/include/linux/i2c-id.h b/include/linux/i2c-id.h index 44f30876a1c9..1ce4b54caa21 100644 --- a/include/linux/i2c-id.h +++ b/include/linux/i2c-id.h @@ -164,10 +164,7 @@ /* --- Bit algorithm adapters */ #define I2C_HW_B_LP 0x010000 /* Parallel port Philips style */ -#define I2C_HW_B_LPC 0x010001 /* Parallel port control reg. */ #define I2C_HW_B_SER 0x010002 /* Serial line interface */ -#define I2C_HW_B_ELV 0x010003 /* ELV Card */ -#define I2C_HW_B_VELLE 0x010004 /* Vellemann K8000 */ #define I2C_HW_B_BT848 0x010005 /* BT848 video boards */ #define I2C_HW_B_WNV 0x010006 /* Winnov Videums */ #define I2C_HW_B_VIA 0x010007 /* Via vt82c586b */ -- cgit v1.2.3 From b8095544bc27044a7aa79e1e073b781a249910c3 Mon Sep 17 00:00:00 2001 From: Jean Delvare Date: Wed, 26 Oct 2005 21:25:04 +0200 Subject: [PATCH] i2c: SMBus PEC support rewrite, 1 of 3 Discard I2C_FUNC_SMBUS_*_PEC defines. i2c clients are not supposed to check for PEC support of i2c bus drivers on individual SMBus transactions, and i2c bus drivers are not supposed to advertise them. Signed-off-by: Jean Delvare Signed-off-by: Greg Kroah-Hartman --- drivers/i2c/busses/i2c-i801.c | 3 +-- include/linux/i2c.h | 18 ------------------ 2 files changed, 1 insertion(+), 20 deletions(-) (limited to 'include/linux') diff --git a/drivers/i2c/busses/i2c-i801.c b/drivers/i2c/busses/i2c-i801.c index e90b6c4e7da3..b05e045f770c 100644 --- a/drivers/i2c/busses/i2c-i801.c +++ b/drivers/i2c/busses/i2c-i801.c @@ -513,8 +513,7 @@ static u32 i801_func(struct i2c_adapter *adapter) return I2C_FUNC_SMBUS_QUICK | I2C_FUNC_SMBUS_BYTE | I2C_FUNC_SMBUS_BYTE_DATA | I2C_FUNC_SMBUS_WORD_DATA | I2C_FUNC_SMBUS_BLOCK_DATA | I2C_FUNC_SMBUS_WRITE_I2C_BLOCK - | (isich4 ? I2C_FUNC_SMBUS_BLOCK_DATA_PEC | - I2C_FUNC_SMBUS_HWPEC_CALC : 0); + | (isich4 ? I2C_FUNC_SMBUS_HWPEC_CALC : 0); } static struct i2c_algorithm smbus_algorithm = { diff --git a/include/linux/i2c.h b/include/linux/i2c.h index 64c13c042f9d..bcd4bb1b450c 100644 --- a/include/linux/i2c.h +++ b/include/linux/i2c.h @@ -389,10 +389,6 @@ struct i2c_msg { #define I2C_FUNC_10BIT_ADDR 0x00000002 #define I2C_FUNC_PROTOCOL_MANGLING 0x00000004 /* I2C_M_{REV_DIR_ADDR,NOSTART,..} */ #define I2C_FUNC_SMBUS_HWPEC_CALC 0x00000008 /* SMBus 2.0 */ -#define I2C_FUNC_SMBUS_READ_WORD_DATA_PEC 0x00000800 /* SMBus 2.0 */ -#define I2C_FUNC_SMBUS_WRITE_WORD_DATA_PEC 0x00001000 /* SMBus 2.0 */ -#define I2C_FUNC_SMBUS_PROC_CALL_PEC 0x00002000 /* SMBus 2.0 */ -#define I2C_FUNC_SMBUS_BLOCK_PROC_CALL_PEC 0x00004000 /* SMBus 2.0 */ #define I2C_FUNC_SMBUS_BLOCK_PROC_CALL 0x00008000 /* SMBus 2.0 */ #define I2C_FUNC_SMBUS_QUICK 0x00010000 #define I2C_FUNC_SMBUS_READ_BYTE 0x00020000 @@ -408,8 +404,6 @@ struct i2c_msg { #define I2C_FUNC_SMBUS_WRITE_I2C_BLOCK 0x08000000 /* w/ 1-byte reg. addr. */ #define I2C_FUNC_SMBUS_READ_I2C_BLOCK_2 0x10000000 /* I2C-like block xfer */ #define I2C_FUNC_SMBUS_WRITE_I2C_BLOCK_2 0x20000000 /* w/ 2-byte reg. addr. */ -#define I2C_FUNC_SMBUS_READ_BLOCK_DATA_PEC 0x40000000 /* SMBus 2.0 */ -#define I2C_FUNC_SMBUS_WRITE_BLOCK_DATA_PEC 0x80000000 /* SMBus 2.0 */ #define I2C_FUNC_SMBUS_BYTE (I2C_FUNC_SMBUS_READ_BYTE | \ I2C_FUNC_SMBUS_WRITE_BYTE) @@ -423,17 +417,6 @@ struct i2c_msg { I2C_FUNC_SMBUS_WRITE_I2C_BLOCK) #define I2C_FUNC_SMBUS_I2C_BLOCK_2 (I2C_FUNC_SMBUS_READ_I2C_BLOCK_2 | \ I2C_FUNC_SMBUS_WRITE_I2C_BLOCK_2) -#define I2C_FUNC_SMBUS_BLOCK_DATA_PEC (I2C_FUNC_SMBUS_READ_BLOCK_DATA_PEC | \ - I2C_FUNC_SMBUS_WRITE_BLOCK_DATA_PEC) -#define I2C_FUNC_SMBUS_WORD_DATA_PEC (I2C_FUNC_SMBUS_READ_WORD_DATA_PEC | \ - I2C_FUNC_SMBUS_WRITE_WORD_DATA_PEC) - -#define I2C_FUNC_SMBUS_READ_BYTE_PEC I2C_FUNC_SMBUS_READ_BYTE_DATA -#define I2C_FUNC_SMBUS_WRITE_BYTE_PEC I2C_FUNC_SMBUS_WRITE_BYTE_DATA -#define I2C_FUNC_SMBUS_READ_BYTE_DATA_PEC I2C_FUNC_SMBUS_READ_WORD_DATA -#define I2C_FUNC_SMBUS_WRITE_BYTE_DATA_PEC I2C_FUNC_SMBUS_WRITE_WORD_DATA -#define I2C_FUNC_SMBUS_BYTE_PEC I2C_FUNC_SMBUS_BYTE_DATA -#define I2C_FUNC_SMBUS_BYTE_DATA_PEC I2C_FUNC_SMBUS_WORD_DATA #define I2C_FUNC_SMBUS_EMUL (I2C_FUNC_SMBUS_QUICK | \ I2C_FUNC_SMBUS_BYTE | \ @@ -441,7 +424,6 @@ struct i2c_msg { I2C_FUNC_SMBUS_WORD_DATA | \ I2C_FUNC_SMBUS_PROC_CALL | \ I2C_FUNC_SMBUS_WRITE_BLOCK_DATA | \ - I2C_FUNC_SMBUS_WRITE_BLOCK_DATA_PEC | \ I2C_FUNC_SMBUS_I2C_BLOCK) /* -- cgit v1.2.3 From 421ef47be20c5454b12ae0ec918d5073a9d2b938 Mon Sep 17 00:00:00 2001 From: Jean Delvare Date: Wed, 26 Oct 2005 21:28:55 +0200 Subject: [PATCH] i2c: SMBus PEC support rewrite, 2 of 3 This is my rewrite of the SMBus PEC support. The original implementation was known to have bugs (credits go to Hideki Iwamoto for reporting many of them recently), and was incomplete due to a conceptual limitation. The rewrite affects only software PEC. Hardware PEC needs very little code and is mostly untouched. Technically, both implementations differ in that the original one was emulating PEC in software by modifying the contents of an i2c_smbus_data union (changing the transaction to a different type), while the new one works one level lower, on i2c_msg structures (working on message contents). Due to the definition of the i2c_smbus_data union, not all SMBus transactions could be handled (at least not without changing the definition of this union, which would break user-space compatibility), and those which could had to be implemented individually. At the opposite, adding PEC to an i2c_msg structure can be done on any SMBus transaction with common code. Advantages of the new implementation: * It's about twice as small (from ~136 lines before to ~70 now, only counting i2c-core, including blank and comment lines). The memory used by i2c-core is down by ~640 bytes (~3.5%). * Easier to validate, less tricky code. The code being common to all transactions by design, the risk that a bug can stay uncovered is lower. * All SMBus transactions have PEC support in I2C emulation mode (providing the non-PEC transaction is also implemented). Transactions which have no emulation code right now will get PEC support for free when they finally get implemented. * Allows for code simplifications in header files and bus drivers (patch follows). Drawbacks (I guess there had to be at least one): * PEC emulation for non-PEC capable non-I2C SMBus masters was dropped. It was based on SMBus tricks and doesn't quite fit in the new design. I don't think it's really a problem, as the benefit was certainly not worth the additional complexity, but it's only fair that I at least mention it. Lastly, let's note that the new implementation does slightly affect compatibility (both in kernel and user-space), but doesn't actually break it. Some defines will be dropped, but the code can always be changed in a way that will work with both the old and the new implementations. It shouldn't be a problem as there doesn't seem to be many users of SMBus PEC to date anyway. Signed-off-by: Jean Delvare Signed-off-by: Greg Kroah-Hartman --- drivers/i2c/i2c-core.c | 162 +++++++++++++++---------------------------------- include/linux/i2c.h | 2 +- 2 files changed, 49 insertions(+), 115 deletions(-) (limited to 'include/linux') diff --git a/drivers/i2c/i2c-core.c b/drivers/i2c/i2c-core.c index 0040981b6698..02e335a04f09 100644 --- a/drivers/i2c/i2c-core.c +++ b/drivers/i2c/i2c-core.c @@ -19,7 +19,8 @@ /* With some changes from Kyösti Mälkki . All SMBus-related things are written by Frodo Looijaard - SMBus 2.0 support by Mark Studebaker */ + SMBus 2.0 support by Mark Studebaker and + Jean Delvare */ #include #include @@ -830,101 +831,44 @@ crc8(u16 data) return (u8)(data >> 8); } -/* CRC over count bytes in the first array plus the bytes in the rest - array if it is non-null. rest[0] is the (length of rest) - 1 - and is included. */ -static u8 i2c_smbus_partial_pec(u8 crc, int count, u8 *first, u8 *rest) +/* Incremental CRC8 over count bytes in the array pointed to by p */ +static u8 i2c_smbus_pec(u8 crc, u8 *p, size_t count) { int i; for(i = 0; i < count; i++) - crc = crc8((crc ^ first[i]) << 8); - if(rest != NULL) - for(i = 0; i <= rest[0]; i++) - crc = crc8((crc ^ rest[i]) << 8); + crc = crc8((crc ^ p[i]) << 8); return crc; } -static u8 i2c_smbus_pec(int count, u8 *first, u8 *rest) +/* Assume a 7-bit address, which is reasonable for SMBus */ +static u8 i2c_smbus_msg_pec(u8 pec, struct i2c_msg *msg) { - return i2c_smbus_partial_pec(0, count, first, rest); + /* The address will be sent first */ + u8 addr = (msg->addr << 1) | !!(msg->flags & I2C_M_RD); + pec = i2c_smbus_pec(pec, &addr, 1); + + /* The data buffer follows */ + return i2c_smbus_pec(pec, msg->buf, msg->len); } -/* Returns new "size" (transaction type) - Note that we convert byte to byte_data and byte_data to word_data - rather than invent new xxx_PEC transactions. */ -static int i2c_smbus_add_pec(u16 addr, u8 command, int size, - union i2c_smbus_data *data) +/* Used for write only transactions */ +static inline void i2c_smbus_add_pec(struct i2c_msg *msg) { - u8 buf[3]; - - buf[0] = addr << 1; - buf[1] = command; - switch(size) { - case I2C_SMBUS_BYTE: - data->byte = i2c_smbus_pec(2, buf, NULL); - size = I2C_SMBUS_BYTE_DATA; - break; - case I2C_SMBUS_BYTE_DATA: - buf[2] = data->byte; - data->word = buf[2] | - (i2c_smbus_pec(3, buf, NULL) << 8); - size = I2C_SMBUS_WORD_DATA; - break; - case I2C_SMBUS_WORD_DATA: - /* unsupported */ - break; - case I2C_SMBUS_BLOCK_DATA: - data->block[data->block[0] + 1] = - i2c_smbus_pec(2, buf, data->block); - size = I2C_SMBUS_BLOCK_DATA_PEC; - break; - } - return size; + msg->buf[msg->len] = i2c_smbus_msg_pec(0, msg); + msg->len++; } -static int i2c_smbus_check_pec(u16 addr, u8 command, int size, u8 partial, - union i2c_smbus_data *data) +/* Return <0 on CRC error + If there was a write before this read (most cases) we need to take the + partial CRC from the write part into account. + Note that this function does modify the message (we need to decrease the + message length to hide the CRC byte from the caller). */ +static int i2c_smbus_check_pec(u8 cpec, struct i2c_msg *msg) { - u8 buf[3], rpec, cpec; + u8 rpec = msg->buf[--msg->len]; + cpec = i2c_smbus_msg_pec(cpec, msg); - buf[1] = command; - switch(size) { - case I2C_SMBUS_BYTE_DATA: - buf[0] = (addr << 1) | 1; - cpec = i2c_smbus_pec(2, buf, NULL); - rpec = data->byte; - break; - case I2C_SMBUS_WORD_DATA: - buf[0] = (addr << 1) | 1; - buf[2] = data->word & 0xff; - cpec = i2c_smbus_pec(3, buf, NULL); - rpec = data->word >> 8; - break; - case I2C_SMBUS_WORD_DATA_PEC: - /* unsupported */ - cpec = rpec = 0; - break; - case I2C_SMBUS_PROC_CALL_PEC: - /* unsupported */ - cpec = rpec = 0; - break; - case I2C_SMBUS_BLOCK_DATA_PEC: - buf[0] = (addr << 1); - buf[2] = (addr << 1) | 1; - cpec = i2c_smbus_pec(3, buf, data->block); - rpec = data->block[data->block[0] + 1]; - break; - case I2C_SMBUS_BLOCK_PROC_CALL_PEC: - buf[0] = (addr << 1) | 1; - rpec = i2c_smbus_partial_pec(partial, 1, - buf, data->block); - cpec = data->block[data->block[0] + 1]; - break; - default: - cpec = rpec = 0; - break; - } if (rpec != cpec) { pr_debug("i2c-core: Bad PEC 0x%02x vs. 0x%02x\n", rpec, cpec); @@ -951,9 +895,8 @@ s32 i2c_smbus_read_byte(struct i2c_client *client) s32 i2c_smbus_write_byte(struct i2c_client *client, u8 value) { - union i2c_smbus_data data; /* only for PEC */ return i2c_smbus_xfer(client->adapter,client->addr,client->flags, - I2C_SMBUS_WRITE,value, I2C_SMBUS_BYTE,&data); + I2C_SMBUS_WRITE, value, I2C_SMBUS_BYTE, NULL); } s32 i2c_smbus_read_byte_data(struct i2c_client *client, u8 command) @@ -1043,6 +986,7 @@ static s32 i2c_smbus_xfer_emulated(struct i2c_adapter * adapter, u16 addr, { addr, flags | I2C_M_RD, 0, msgbuf1 } }; int i; + u8 partial_pec = 0; msgbuf0[0] = command; switch(size) { @@ -1085,7 +1029,6 @@ static s32 i2c_smbus_xfer_emulated(struct i2c_adapter * adapter, u16 addr, msgbuf0[2] = (data->word >> 8) & 0xff; break; case I2C_SMBUS_BLOCK_DATA: - case I2C_SMBUS_BLOCK_DATA_PEC: if (read_write == I2C_SMBUS_READ) { dev_err(&adapter->dev, "Block read not supported " "under I2C emulation!\n"); @@ -1098,14 +1041,11 @@ static s32 i2c_smbus_xfer_emulated(struct i2c_adapter * adapter, u16 addr, data->block[0]); return -1; } - if(size == I2C_SMBUS_BLOCK_DATA_PEC) - (msg[0].len)++; for (i = 1; i < msg[0].len; i++) msgbuf0[i] = data->block[i-1]; } break; case I2C_SMBUS_BLOCK_PROC_CALL: - case I2C_SMBUS_BLOCK_PROC_CALL_PEC: dev_dbg(&adapter->dev, "Block process call not supported " "under I2C emulation!\n"); return -1; @@ -1130,9 +1070,30 @@ static s32 i2c_smbus_xfer_emulated(struct i2c_adapter * adapter, u16 addr, return -1; } + i = ((flags & I2C_CLIENT_PEC) && size != I2C_SMBUS_QUICK + && size != I2C_SMBUS_I2C_BLOCK_DATA); + if (i) { + /* Compute PEC if first message is a write */ + if (!(msg[0].flags & I2C_M_RD)) { + if (num == 1) /* Write only */ + i2c_smbus_add_pec(&msg[0]); + else /* Write followed by read */ + partial_pec = i2c_smbus_msg_pec(0, &msg[0]); + } + /* Ask for PEC if last message is a read */ + if (msg[num-1].flags & I2C_M_RD) + msg[num-1].len++; + } + if (i2c_transfer(adapter, msg, num) < 0) return -1; + /* Check PEC if last message is a read */ + if (i && (msg[num-1].flags & I2C_M_RD)) { + if (i2c_smbus_check_pec(partial_pec, &msg[num-1]) < 0) + return -1; + } + if (read_write == I2C_SMBUS_READ) switch(size) { case I2C_SMBUS_BYTE: @@ -1161,28 +1122,8 @@ s32 i2c_smbus_xfer(struct i2c_adapter * adapter, u16 addr, unsigned short flags, union i2c_smbus_data * data) { s32 res; - int swpec = 0; - u8 partial = 0; flags &= I2C_M_TEN | I2C_CLIENT_PEC; - if((flags & I2C_CLIENT_PEC) && - !(i2c_check_functionality(adapter, I2C_FUNC_SMBUS_HWPEC_CALC))) { - swpec = 1; - if(read_write == I2C_SMBUS_READ && - size == I2C_SMBUS_BLOCK_DATA) - size = I2C_SMBUS_BLOCK_DATA_PEC; - else if(size == I2C_SMBUS_PROC_CALL) - size = I2C_SMBUS_PROC_CALL_PEC; - else if(size == I2C_SMBUS_BLOCK_PROC_CALL) { - i2c_smbus_add_pec(addr, command, - I2C_SMBUS_BLOCK_DATA, data); - partial = data->block[data->block[0] + 1]; - size = I2C_SMBUS_BLOCK_PROC_CALL_PEC; - } else if(read_write == I2C_SMBUS_WRITE && - size != I2C_SMBUS_QUICK && - size != I2C_SMBUS_I2C_BLOCK_DATA) - size = i2c_smbus_add_pec(addr, command, size, data); - } if (adapter->algo->smbus_xfer) { down(&adapter->bus_lock); @@ -1193,13 +1134,6 @@ s32 i2c_smbus_xfer(struct i2c_adapter * adapter, u16 addr, unsigned short flags, res = i2c_smbus_xfer_emulated(adapter,addr,flags,read_write, command,size,data); - if(res >= 0 && swpec && - size != I2C_SMBUS_QUICK && size != I2C_SMBUS_I2C_BLOCK_DATA && - (read_write == I2C_SMBUS_READ || size == I2C_SMBUS_PROC_CALL_PEC || - size == I2C_SMBUS_BLOCK_PROC_CALL_PEC)) { - if(i2c_smbus_check_pec(addr, command, size, partial, data)) - return -1; - } return res; } diff --git a/include/linux/i2c.h b/include/linux/i2c.h index bcd4bb1b450c..78c64f7b3105 100644 --- a/include/linux/i2c.h +++ b/include/linux/i2c.h @@ -434,7 +434,7 @@ union i2c_smbus_data { __u8 byte; __u16 word; __u8 block[I2C_SMBUS_BLOCK_MAX + 2]; /* block[0] is used for length */ - /* and one more for PEC */ + /* and one more for user-space compatibility */ }; /* smbus_access read or write markers */ -- cgit v1.2.3 From 585b3160f8212e58325bc1c0292c2ec01ac5db84 Mon Sep 17 00:00:00 2001 From: Jean Delvare Date: Wed, 26 Oct 2005 21:31:15 +0200 Subject: [PATCH] i2c: SMBus PEC support rewrite, 3 of 3 The new SMBus PEC implementation doesn't support PEC emulation on non-PEC non-I2C SMBus masters, so we can drop all related code. Signed-off-by: Jean Delvare Signed-off-by: Greg Kroah-Hartman --- drivers/i2c/busses/i2c-amd8111.c | 7 ------- drivers/i2c/busses/i2c-i801.c | 13 +++++-------- drivers/i2c/busses/i2c-nforce2.c | 7 ------- include/linux/i2c.h | 4 ---- 4 files changed, 5 insertions(+), 26 deletions(-) (limited to 'include/linux') diff --git a/drivers/i2c/busses/i2c-amd8111.c b/drivers/i2c/busses/i2c-amd8111.c index aface7938a5b..f3b79a68dbec 100644 --- a/drivers/i2c/busses/i2c-amd8111.c +++ b/drivers/i2c/busses/i2c-amd8111.c @@ -253,13 +253,6 @@ static s32 amd8111_access(struct i2c_adapter * adap, u16 addr, unsigned short fl read_write = I2C_SMBUS_READ; break; - case I2C_SMBUS_WORD_DATA_PEC: - case I2C_SMBUS_BLOCK_DATA_PEC: - case I2C_SMBUS_PROC_CALL_PEC: - case I2C_SMBUS_BLOCK_PROC_CALL_PEC: - dev_warn(&adap->dev, "Unexpected software PEC transaction %d\n.", size); - return -1; - default: dev_warn(&adap->dev, "Unsupported transaction %d\n", size); return -1; diff --git a/drivers/i2c/busses/i2c-i801.c b/drivers/i2c/busses/i2c-i801.c index b05e045f770c..27e7894a9d1c 100644 --- a/drivers/i2c/busses/i2c-i801.c +++ b/drivers/i2c/busses/i2c-i801.c @@ -102,8 +102,8 @@ MODULE_PARM_DESC(force_addr, "EXTREMELY DANGEROUS!"); static int i801_transaction(void); -static int i801_block_transaction(union i2c_smbus_data *data, - char read_write, int command); +static int i801_block_transaction(union i2c_smbus_data *data, char read_write, + int command, int hwpec); static unsigned short i801_smba; static struct pci_driver i801_driver; @@ -249,7 +249,7 @@ static int i801_transaction(void) /* All-inclusive block transaction function */ static int i801_block_transaction(union i2c_smbus_data *data, char read_write, - int command) + int command, int hwpec) { int i, len; int smbcmd; @@ -388,7 +388,7 @@ static int i801_block_transaction(union i2c_smbus_data *data, char read_write, goto END; } - if(isich4 && command == I2C_SMBUS_BLOCK_DATA_PEC) { + if (hwpec && command == I2C_SMBUS_BLOCK_DATA) { /* wait for INTR bit as advised by Intel */ timeout = 0; do { @@ -456,9 +456,6 @@ static s32 i801_access(struct i2c_adapter * adap, u16 addr, break; case I2C_SMBUS_BLOCK_DATA: case I2C_SMBUS_I2C_BLOCK_DATA: - case I2C_SMBUS_BLOCK_DATA_PEC: - if(hwpec && size == I2C_SMBUS_BLOCK_DATA) - size = I2C_SMBUS_BLOCK_DATA_PEC; outb_p(((addr & 0x7f) << 1) | (read_write & 0x01), SMBHSTADD); outb_p(command, SMBHSTCMD); @@ -476,7 +473,7 @@ static s32 i801_access(struct i2c_adapter * adap, u16 addr, outb_p(1, SMBAUXCTL); /* enable HW PEC */ } if(block) - ret = i801_block_transaction(data, read_write, size); + ret = i801_block_transaction(data, read_write, size, hwpec); else { outb_p(xact | ENABLE_INT9, SMBHSTCNT); ret = i801_transaction(); diff --git a/drivers/i2c/busses/i2c-nforce2.c b/drivers/i2c/busses/i2c-nforce2.c index 9b4247610815..fd26036e68a3 100644 --- a/drivers/i2c/busses/i2c-nforce2.c +++ b/drivers/i2c/busses/i2c-nforce2.c @@ -188,13 +188,6 @@ static s32 nforce2_access(struct i2c_adapter * adap, u16 addr, dev_err(&adap->dev, "I2C_SMBUS_BLOCK_PROC_CALL not supported!\n"); return -1; - case I2C_SMBUS_WORD_DATA_PEC: - case I2C_SMBUS_BLOCK_DATA_PEC: - case I2C_SMBUS_PROC_CALL_PEC: - case I2C_SMBUS_BLOCK_PROC_CALL_PEC: - dev_err(&adap->dev, "Unexpected software PEC transaction %d\n.", size); - return -1; - default: dev_err(&adap->dev, "Unsupported transaction %d\n", size); return -1; diff --git a/include/linux/i2c.h b/include/linux/i2c.h index 78c64f7b3105..32977fb1d99a 100644 --- a/include/linux/i2c.h +++ b/include/linux/i2c.h @@ -451,10 +451,6 @@ union i2c_smbus_data { #define I2C_SMBUS_BLOCK_DATA 5 #define I2C_SMBUS_I2C_BLOCK_DATA 6 #define I2C_SMBUS_BLOCK_PROC_CALL 7 /* SMBus 2.0 */ -#define I2C_SMBUS_BLOCK_DATA_PEC 8 /* SMBus 2.0 */ -#define I2C_SMBUS_PROC_CALL_PEC 9 /* SMBus 2.0 */ -#define I2C_SMBUS_BLOCK_PROC_CALL_PEC 10 /* SMBus 2.0 */ -#define I2C_SMBUS_WORD_DATA_PEC 11 /* SMBus 2.0 */ /* ----- commands for the ioctl like i2c_command call: -- cgit v1.2.3 From a9d1b24d91f91b77db3da8aeacb414764f789b9c Mon Sep 17 00:00:00 2001 From: Greg Kroah-Hartman Date: Sat, 22 Oct 2005 00:23:27 +0200 Subject: [PATCH] I2C: add i2c module alias for i2c drivers to use This is the start of adding hotplug-like support for i2c devices. Signed-off-by: Greg Kroah-Hartman --- include/linux/i2c.h | 1 + include/linux/mod_devicetable.h | 5 +++++ scripts/mod/file2alias.c | 10 ++++++++++ 3 files changed, 16 insertions(+) (limited to 'include/linux') diff --git a/include/linux/i2c.h b/include/linux/i2c.h index 32977fb1d99a..f88577ca3b3a 100644 --- a/include/linux/i2c.h +++ b/include/linux/i2c.h @@ -29,6 +29,7 @@ #include #include #include +#include #include /* for struct device */ #include diff --git a/include/linux/mod_devicetable.h b/include/linux/mod_devicetable.h index 2f0299a448f6..7b08c11ec4cc 100644 --- a/include/linux/mod_devicetable.h +++ b/include/linux/mod_devicetable.h @@ -244,4 +244,9 @@ struct pcmcia_device_id { #define PCMCIA_DEV_ID_MATCH_FAKE_CIS 0x0200 #define PCMCIA_DEV_ID_MATCH_ANONYMOUS 0x0400 +/* I2C */ +struct i2c_device_id { + __u16 id; +}; + #endif /* LINUX_MOD_DEVICETABLE_H */ diff --git a/scripts/mod/file2alias.c b/scripts/mod/file2alias.c index f2ee673329a7..e3d144a3f10b 100644 --- a/scripts/mod/file2alias.c +++ b/scripts/mod/file2alias.c @@ -359,6 +359,13 @@ static int do_vio_entry(const char *filename, struct vio_device_id *vio, return 1; } +static int do_i2c_entry(const char *filename, struct i2c_device_id *i2c, char *alias) +{ + strcpy(alias, "i2c:"); + ADD(alias, "id", 1, i2c->id); + return 1; +} + /* Ignore any prefix, eg. v850 prepends _ */ static inline int sym_is(const char *symbol, const char *name) { @@ -443,6 +450,9 @@ void handle_moddevtable(struct module *mod, struct elf_info *info, else if (sym_is(symname, "__mod_vio_device_table")) do_table(symval, sym->st_size, sizeof(struct vio_device_id), do_vio_entry, mod); + else if (sym_is(symname, "__mod_i2c_device_table")) + do_table(symval, sym->st_size, sizeof(struct i2c_device_id), + do_i2c_entry, mod); } -- cgit v1.2.3 From e04b0ea2e0f9c1bb0d874db4493fc7f7a623116b Mon Sep 17 00:00:00 2001 From: Brian King Date: Tue, 27 Sep 2005 01:21:55 -0700 Subject: [PATCH] PCI: Block config access during BIST Some PCI adapters (eg. ipr scsi adapters) have an exposure today in that they issue BIST to the adapter to reset the card. If, during the time it takes to complete BIST, userspace attempts to access PCI config space, the host bus bridge will master abort the access since the ipr adapter does not respond on the PCI bus for a brief period of time when running BIST. On PPC64 hardware, this master abort results in the host PCI bridge isolating that PCI device from the rest of the system, making the device unusable until Linux is rebooted. This patch is an attempt to close that exposure by introducing some blocking code in the PCI code. When blocked, writes will be humored and reads will return the cached value. Ben Herrenschmidt has also mentioned that he plans to use this in PPC power management. Signed-off-by: Brian King Cc: Benjamin Herrenschmidt Signed-off-by: Andrew Morton Signed-off-by: Greg Kroah-Hartman drivers/pci/access.c | 89 ++++++++++++++++++++++++++++++++++++++++++++++++ drivers/pci/pci-sysfs.c | 20 +++++----- drivers/pci/pci.h | 7 +++ drivers/pci/proc.c | 28 +++++++-------- drivers/pci/syscall.c | 14 +++---- include/linux/pci.h | 7 +++ 6 files changed, 134 insertions(+), 31 deletions(-) --- drivers/pci/access.c | 89 +++++++++++++++++++++++++++++++++++++++++++++++++ drivers/pci/pci-sysfs.c | 20 +++++------ drivers/pci/pci.h | 7 ++++ drivers/pci/proc.c | 28 ++++++++-------- drivers/pci/syscall.c | 14 ++++---- include/linux/pci.h | 7 ++++ 6 files changed, 134 insertions(+), 31 deletions(-) (limited to 'include/linux') diff --git a/drivers/pci/access.c b/drivers/pci/access.c index 24a76de49f41..2a42add7f563 100644 --- a/drivers/pci/access.c +++ b/drivers/pci/access.c @@ -60,3 +60,92 @@ EXPORT_SYMBOL(pci_bus_read_config_dword); EXPORT_SYMBOL(pci_bus_write_config_byte); EXPORT_SYMBOL(pci_bus_write_config_word); EXPORT_SYMBOL(pci_bus_write_config_dword); + +static u32 pci_user_cached_config(struct pci_dev *dev, int pos) +{ + u32 data; + + data = dev->saved_config_space[pos/sizeof(dev->saved_config_space[0])]; + data >>= (pos % sizeof(dev->saved_config_space[0])) * 8; + return data; +} + +#define PCI_USER_READ_CONFIG(size,type) \ +int pci_user_read_config_##size \ + (struct pci_dev *dev, int pos, type *val) \ +{ \ + unsigned long flags; \ + int ret = 0; \ + u32 data = -1; \ + if (PCI_##size##_BAD) return PCIBIOS_BAD_REGISTER_NUMBER; \ + spin_lock_irqsave(&pci_lock, flags); \ + if (likely(!dev->block_ucfg_access)) \ + ret = dev->bus->ops->read(dev->bus, dev->devfn, \ + pos, sizeof(type), &data); \ + else if (pos < sizeof(dev->saved_config_space)) \ + data = pci_user_cached_config(dev, pos); \ + spin_unlock_irqrestore(&pci_lock, flags); \ + *val = (type)data; \ + return ret; \ +} + +#define PCI_USER_WRITE_CONFIG(size,type) \ +int pci_user_write_config_##size \ + (struct pci_dev *dev, int pos, type val) \ +{ \ + unsigned long flags; \ + int ret = -EIO; \ + if (PCI_##size##_BAD) return PCIBIOS_BAD_REGISTER_NUMBER; \ + spin_lock_irqsave(&pci_lock, flags); \ + if (likely(!dev->block_ucfg_access)) \ + ret = dev->bus->ops->write(dev->bus, dev->devfn, \ + pos, sizeof(type), val); \ + spin_unlock_irqrestore(&pci_lock, flags); \ + return ret; \ +} + +PCI_USER_READ_CONFIG(byte, u8) +PCI_USER_READ_CONFIG(word, u16) +PCI_USER_READ_CONFIG(dword, u32) +PCI_USER_WRITE_CONFIG(byte, u8) +PCI_USER_WRITE_CONFIG(word, u16) +PCI_USER_WRITE_CONFIG(dword, u32) + +/** + * pci_block_user_cfg_access - Block userspace PCI config reads/writes + * @dev: pci device struct + * + * This function blocks any userspace PCI config accesses from occurring. + * When blocked, any writes will be bit bucketed and reads will return the + * data saved using pci_save_state for the first 64 bytes of config + * space and return 0xff for all other config reads. + **/ +void pci_block_user_cfg_access(struct pci_dev *dev) +{ + unsigned long flags; + + pci_save_state(dev); + + /* spinlock to synchronize with anyone reading config space now */ + spin_lock_irqsave(&pci_lock, flags); + dev->block_ucfg_access = 1; + spin_unlock_irqrestore(&pci_lock, flags); +} +EXPORT_SYMBOL_GPL(pci_block_user_cfg_access); + +/** + * pci_unblock_user_cfg_access - Unblock userspace PCI config reads/writes + * @dev: pci device struct + * + * This function allows userspace PCI config accesses to resume. + **/ +void pci_unblock_user_cfg_access(struct pci_dev *dev) +{ + unsigned long flags; + + /* spinlock to synchronize with anyone reading saved config space */ + spin_lock_irqsave(&pci_lock, flags); + dev->block_ucfg_access = 0; + spin_unlock_irqrestore(&pci_lock, flags); +} +EXPORT_SYMBOL_GPL(pci_unblock_user_cfg_access); diff --git a/drivers/pci/pci-sysfs.c b/drivers/pci/pci-sysfs.c index 2898830c496f..965a5934623a 100644 --- a/drivers/pci/pci-sysfs.c +++ b/drivers/pci/pci-sysfs.c @@ -130,7 +130,7 @@ pci_read_config(struct kobject *kobj, char *buf, loff_t off, size_t count) if ((off & 1) && size) { u8 val; - pci_read_config_byte(dev, off, &val); + pci_user_read_config_byte(dev, off, &val); data[off - init_off] = val; off++; size--; @@ -138,7 +138,7 @@ pci_read_config(struct kobject *kobj, char *buf, loff_t off, size_t count) if ((off & 3) && size > 2) { u16 val; - pci_read_config_word(dev, off, &val); + pci_user_read_config_word(dev, off, &val); data[off - init_off] = val & 0xff; data[off - init_off + 1] = (val >> 8) & 0xff; off += 2; @@ -147,7 +147,7 @@ pci_read_config(struct kobject *kobj, char *buf, loff_t off, size_t count) while (size > 3) { u32 val; - pci_read_config_dword(dev, off, &val); + pci_user_read_config_dword(dev, off, &val); data[off - init_off] = val & 0xff; data[off - init_off + 1] = (val >> 8) & 0xff; data[off - init_off + 2] = (val >> 16) & 0xff; @@ -158,7 +158,7 @@ pci_read_config(struct kobject *kobj, char *buf, loff_t off, size_t count) if (size >= 2) { u16 val; - pci_read_config_word(dev, off, &val); + pci_user_read_config_word(dev, off, &val); data[off - init_off] = val & 0xff; data[off - init_off + 1] = (val >> 8) & 0xff; off += 2; @@ -167,7 +167,7 @@ pci_read_config(struct kobject *kobj, char *buf, loff_t off, size_t count) if (size > 0) { u8 val; - pci_read_config_byte(dev, off, &val); + pci_user_read_config_byte(dev, off, &val); data[off - init_off] = val; off++; --size; @@ -192,7 +192,7 @@ pci_write_config(struct kobject *kobj, char *buf, loff_t off, size_t count) } if ((off & 1) && size) { - pci_write_config_byte(dev, off, data[off - init_off]); + pci_user_write_config_byte(dev, off, data[off - init_off]); off++; size--; } @@ -200,7 +200,7 @@ pci_write_config(struct kobject *kobj, char *buf, loff_t off, size_t count) if ((off & 3) && size > 2) { u16 val = data[off - init_off]; val |= (u16) data[off - init_off + 1] << 8; - pci_write_config_word(dev, off, val); + pci_user_write_config_word(dev, off, val); off += 2; size -= 2; } @@ -210,7 +210,7 @@ pci_write_config(struct kobject *kobj, char *buf, loff_t off, size_t count) val |= (u32) data[off - init_off + 1] << 8; val |= (u32) data[off - init_off + 2] << 16; val |= (u32) data[off - init_off + 3] << 24; - pci_write_config_dword(dev, off, val); + pci_user_write_config_dword(dev, off, val); off += 4; size -= 4; } @@ -218,13 +218,13 @@ pci_write_config(struct kobject *kobj, char *buf, loff_t off, size_t count) if (size >= 2) { u16 val = data[off - init_off]; val |= (u16) data[off - init_off + 1] << 8; - pci_write_config_word(dev, off, val); + pci_user_write_config_word(dev, off, val); off += 2; size -= 2; } if (size) { - pci_write_config_byte(dev, off, data[off - init_off]); + pci_user_write_config_byte(dev, off, data[off - init_off]); off++; --size; } diff --git a/drivers/pci/pci.h b/drivers/pci/pci.h index d3f3dd42240d..6527b36c9a61 100644 --- a/drivers/pci/pci.h +++ b/drivers/pci/pci.h @@ -15,6 +15,13 @@ extern int pci_bus_alloc_resource(struct pci_bus *bus, struct resource *res, extern int (*platform_pci_choose_state)(struct pci_dev *dev, pm_message_t state); extern int (*platform_pci_set_power_state)(struct pci_dev *dev, pci_power_t state); +extern int pci_user_read_config_byte(struct pci_dev *dev, int where, u8 *val); +extern int pci_user_read_config_word(struct pci_dev *dev, int where, u16 *val); +extern int pci_user_read_config_dword(struct pci_dev *dev, int where, u32 *val); +extern int pci_user_write_config_byte(struct pci_dev *dev, int where, u8 val); +extern int pci_user_write_config_word(struct pci_dev *dev, int where, u16 val); +extern int pci_user_write_config_dword(struct pci_dev *dev, int where, u32 val); + /* PCI /proc functions */ #ifdef CONFIG_PROC_FS extern int pci_proc_attach_device(struct pci_dev *dev); diff --git a/drivers/pci/proc.c b/drivers/pci/proc.c index 9613f666c110..9eb465727fce 100644 --- a/drivers/pci/proc.c +++ b/drivers/pci/proc.c @@ -80,7 +80,7 @@ proc_bus_pci_read(struct file *file, char __user *buf, size_t nbytes, loff_t *pp if ((pos & 1) && cnt) { unsigned char val; - pci_read_config_byte(dev, pos, &val); + pci_user_read_config_byte(dev, pos, &val); __put_user(val, buf); buf++; pos++; @@ -89,7 +89,7 @@ proc_bus_pci_read(struct file *file, char __user *buf, size_t nbytes, loff_t *pp if ((pos & 3) && cnt > 2) { unsigned short val; - pci_read_config_word(dev, pos, &val); + pci_user_read_config_word(dev, pos, &val); __put_user(cpu_to_le16(val), (unsigned short __user *) buf); buf += 2; pos += 2; @@ -98,7 +98,7 @@ proc_bus_pci_read(struct file *file, char __user *buf, size_t nbytes, loff_t *pp while (cnt >= 4) { unsigned int val; - pci_read_config_dword(dev, pos, &val); + pci_user_read_config_dword(dev, pos, &val); __put_user(cpu_to_le32(val), (unsigned int __user *) buf); buf += 4; pos += 4; @@ -107,7 +107,7 @@ proc_bus_pci_read(struct file *file, char __user *buf, size_t nbytes, loff_t *pp if (cnt >= 2) { unsigned short val; - pci_read_config_word(dev, pos, &val); + pci_user_read_config_word(dev, pos, &val); __put_user(cpu_to_le16(val), (unsigned short __user *) buf); buf += 2; pos += 2; @@ -116,7 +116,7 @@ proc_bus_pci_read(struct file *file, char __user *buf, size_t nbytes, loff_t *pp if (cnt) { unsigned char val; - pci_read_config_byte(dev, pos, &val); + pci_user_read_config_byte(dev, pos, &val); __put_user(val, buf); buf++; pos++; @@ -151,7 +151,7 @@ proc_bus_pci_write(struct file *file, const char __user *buf, size_t nbytes, lof if ((pos & 1) && cnt) { unsigned char val; __get_user(val, buf); - pci_write_config_byte(dev, pos, val); + pci_user_write_config_byte(dev, pos, val); buf++; pos++; cnt--; @@ -160,7 +160,7 @@ proc_bus_pci_write(struct file *file, const char __user *buf, size_t nbytes, lof if ((pos & 3) && cnt > 2) { unsigned short val; __get_user(val, (unsigned short __user *) buf); - pci_write_config_word(dev, pos, le16_to_cpu(val)); + pci_user_write_config_word(dev, pos, le16_to_cpu(val)); buf += 2; pos += 2; cnt -= 2; @@ -169,7 +169,7 @@ proc_bus_pci_write(struct file *file, const char __user *buf, size_t nbytes, lof while (cnt >= 4) { unsigned int val; __get_user(val, (unsigned int __user *) buf); - pci_write_config_dword(dev, pos, le32_to_cpu(val)); + pci_user_write_config_dword(dev, pos, le32_to_cpu(val)); buf += 4; pos += 4; cnt -= 4; @@ -178,7 +178,7 @@ proc_bus_pci_write(struct file *file, const char __user *buf, size_t nbytes, lof if (cnt >= 2) { unsigned short val; __get_user(val, (unsigned short __user *) buf); - pci_write_config_word(dev, pos, le16_to_cpu(val)); + pci_user_write_config_word(dev, pos, le16_to_cpu(val)); buf += 2; pos += 2; cnt -= 2; @@ -187,7 +187,7 @@ proc_bus_pci_write(struct file *file, const char __user *buf, size_t nbytes, lof if (cnt) { unsigned char val; __get_user(val, buf); - pci_write_config_byte(dev, pos, val); + pci_user_write_config_byte(dev, pos, val); buf++; pos++; cnt--; @@ -484,10 +484,10 @@ static int show_dev_config(struct seq_file *m, void *v) drv = pci_dev_driver(dev); - pci_read_config_dword(dev, PCI_CLASS_REVISION, &class_rev); - pci_read_config_byte (dev, PCI_LATENCY_TIMER, &latency); - pci_read_config_byte (dev, PCI_MIN_GNT, &min_gnt); - pci_read_config_byte (dev, PCI_MAX_LAT, &max_lat); + pci_user_read_config_dword(dev, PCI_CLASS_REVISION, &class_rev); + pci_user_read_config_byte (dev, PCI_LATENCY_TIMER, &latency); + pci_user_read_config_byte (dev, PCI_MIN_GNT, &min_gnt); + pci_user_read_config_byte (dev, PCI_MAX_LAT, &max_lat); seq_printf(m, " Bus %2d, device %3d, function %2d:\n", dev->bus->number, PCI_SLOT(dev->devfn), PCI_FUNC(dev->devfn)); seq_printf(m, " Class %04x", class_rev >> 16); diff --git a/drivers/pci/syscall.c b/drivers/pci/syscall.c index c071790cc983..87fafc08cb9d 100644 --- a/drivers/pci/syscall.c +++ b/drivers/pci/syscall.c @@ -13,7 +13,7 @@ #include #include #include - +#include "pci.h" asmlinkage long sys_pciconfig_read(unsigned long bus, unsigned long dfn, @@ -38,13 +38,13 @@ sys_pciconfig_read(unsigned long bus, unsigned long dfn, lock_kernel(); switch (len) { case 1: - cfg_ret = pci_read_config_byte(dev, off, &byte); + cfg_ret = pci_user_read_config_byte(dev, off, &byte); break; case 2: - cfg_ret = pci_read_config_word(dev, off, &word); + cfg_ret = pci_user_read_config_word(dev, off, &word); break; case 4: - cfg_ret = pci_read_config_dword(dev, off, &dword); + cfg_ret = pci_user_read_config_dword(dev, off, &dword); break; default: err = -EINVAL; @@ -112,7 +112,7 @@ sys_pciconfig_write(unsigned long bus, unsigned long dfn, err = get_user(byte, (u8 __user *)buf); if (err) break; - err = pci_write_config_byte(dev, off, byte); + err = pci_user_write_config_byte(dev, off, byte); if (err != PCIBIOS_SUCCESSFUL) err = -EIO; break; @@ -121,7 +121,7 @@ sys_pciconfig_write(unsigned long bus, unsigned long dfn, err = get_user(word, (u16 __user *)buf); if (err) break; - err = pci_write_config_word(dev, off, word); + err = pci_user_write_config_word(dev, off, word); if (err != PCIBIOS_SUCCESSFUL) err = -EIO; break; @@ -130,7 +130,7 @@ sys_pciconfig_write(unsigned long bus, unsigned long dfn, err = get_user(dword, (u32 __user *)buf); if (err) break; - err = pci_write_config_dword(dev, off, dword); + err = pci_user_write_config_dword(dev, off, dword); if (err != PCIBIOS_SUCCESSFUL) err = -EIO; break; diff --git a/include/linux/pci.h b/include/linux/pci.h index 7349058ed778..3596ac94ecff 100644 --- a/include/linux/pci.h +++ b/include/linux/pci.h @@ -132,6 +132,7 @@ struct pci_dev { unsigned int is_enabled:1; /* pci_enable_device has been called */ unsigned int is_busmaster:1; /* device is busmaster */ unsigned int no_msi:1; /* device may not use msi */ + unsigned int block_ucfg_access:1; /* userspace config space access is blocked */ u32 saved_config_space[16]; /* config space saved at suspend time */ struct bin_attribute *rom_attr; /* attribute descriptor for sysfs ROM entry */ @@ -490,6 +491,9 @@ extern void pci_disable_msix(struct pci_dev *dev); extern void msi_remove_pci_irq_vectors(struct pci_dev *dev); #endif +extern void pci_block_user_cfg_access(struct pci_dev *dev); +extern void pci_unblock_user_cfg_access(struct pci_dev *dev); + /* * PCI domain support. Sometimes called PCI segment (eg by ACPI), * a PCI domain is defined to be a set of PCI busses which share @@ -560,6 +564,9 @@ static inline int pci_enable_wake(struct pci_dev *dev, pci_power_t state, int en #define pci_dma_burst_advice(pdev, strat, strategy_parameter) do { } while (0) +static inline void pci_block_user_cfg_access(struct pci_dev *dev) { } +static inline void pci_unblock_user_cfg_access(struct pci_dev *dev) { } + #endif /* CONFIG_PCI */ /* Include architecture-dependent settings and functions */ -- cgit v1.2.3 From b135c4815051bad6b2472e4ad0152f205918d2c5 Mon Sep 17 00:00:00 2001 From: Grant Coady Date: Thu, 29 Sep 2005 10:39:46 +1000 Subject: [PATCH] pci_ids: remove duplicates from pci_ids.h pci_ids.h cleanup: remove duplicated entries and change some defines to explicit value rather than in terms of another constant, preparation for removing unused symbols Signed-off-by: Grant Coady Signed-off-by: Greg Kroah-Hartman include/linux/pci_ids.h | 28 +++++++++------------------- 1 file changed, 9 insertions(+), 19 deletions(-) --- include/linux/pci_ids.h | 28 +++++++++------------------- 1 file changed, 9 insertions(+), 19 deletions(-) (limited to 'include/linux') diff --git a/include/linux/pci_ids.h b/include/linux/pci_ids.h index 71834f05504f..0631074954f7 100644 --- a/include/linux/pci_ids.h +++ b/include/linux/pci_ids.h @@ -274,7 +274,6 @@ #define PCI_DEVICE_ID_ATI_RAGE128_PP 0x5050 #define PCI_DEVICE_ID_ATI_RAGE128_PQ 0x5051 #define PCI_DEVICE_ID_ATI_RAGE128_PR 0x5052 -#define PCI_DEVICE_ID_ATI_RAGE128_TR 0x5452 #define PCI_DEVICE_ID_ATI_RAGE128_PS 0x5053 #define PCI_DEVICE_ID_ATI_RAGE128_PT 0x5054 #define PCI_DEVICE_ID_ATI_RAGE128_PU 0x5055 @@ -517,16 +516,16 @@ #define PCI_DEVICE_ID_AMD_VIPER_7413 0x7413 #define PCI_DEVICE_ID_AMD_VIPER_7414 0x7414 #define PCI_DEVICE_ID_AMD_OPUS_7440 0x7440 -# define PCI_DEVICE_ID_AMD_VIPER_7440 PCI_DEVICE_ID_AMD_OPUS_7440 +#define PCI_DEVICE_ID_AMD_VIPER_7440 0x7440 #define PCI_DEVICE_ID_AMD_OPUS_7441 0x7441 -# define PCI_DEVICE_ID_AMD_VIPER_7441 PCI_DEVICE_ID_AMD_OPUS_7441 +#define PCI_DEVICE_ID_AMD_VIPER_7441 0x7441 #define PCI_DEVICE_ID_AMD_OPUS_7443 0x7443 -# define PCI_DEVICE_ID_AMD_VIPER_7443 PCI_DEVICE_ID_AMD_OPUS_7443 +#define PCI_DEVICE_ID_AMD_VIPER_7443 0x7443 #define PCI_DEVICE_ID_AMD_OPUS_7445 0x7445 #define PCI_DEVICE_ID_AMD_OPUS_7448 0x7448 -# define PCI_DEVICE_ID_AMD_VIPER_7448 PCI_DEVICE_ID_AMD_OPUS_7448 +#define PCI_DEVICE_ID_AMD_VIPER_7448 0x7448 #define PCI_DEVICE_ID_AMD_OPUS_7449 0x7449 -# define PCI_DEVICE_ID_AMD_VIPER_7449 PCI_DEVICE_ID_AMD_OPUS_7449 +#define PCI_DEVICE_ID_AMD_VIPER_7449 0x7449 #define PCI_DEVICE_ID_AMD_8111_LAN 0x7462 #define PCI_DEVICE_ID_AMD_8111_LPC 0x7468 #define PCI_DEVICE_ID_AMD_8111_IDE 0x7469 @@ -585,7 +584,6 @@ #define PCI_DEVICE_ID_CT_65550 0x00e0 #define PCI_DEVICE_ID_CT_65554 0x00e4 #define PCI_DEVICE_ID_CT_65555 0x00e5 -#define PCI_DEVICE_ID_CT_69000 0x00c0 #define PCI_VENDOR_ID_MIRO 0x1031 #define PCI_DEVICE_ID_MIRO_36050 0x5601 @@ -1197,7 +1195,6 @@ #define PCI_DEVICE_ID_QUADRO_FX_GO1400 0x00cc #define PCI_DEVICE_ID_QUADRO_FX_1400 0x00ce #define PCI_DEVICE_ID_NVIDIA_NFORCE3 0x00d1 -#define PCI_DEVICE_ID_NVIDIA_MCP3_AUDIO 0x00da #define PCI_DEVICE_ID_NVIDIA_NFORCE3_SMBUS 0x00d4 #define PCI_DEVICE_ID_NVIDIA_NFORCE3_IDE 0x00d5 #define PCI_DEVICE_ID_NVIDIA_NVENET_3 0x00d6 @@ -1623,7 +1620,7 @@ #define PCI_DEVICE_ID_SERVERWORKS_HT1000IDE 0x0214 #define PCI_DEVICE_ID_SERVERWORKS_CSB6IDE2 0x0217 #define PCI_DEVICE_ID_SERVERWORKS_OSB4USB 0x0220 -#define PCI_DEVICE_ID_SERVERWORKS_CSB5USB PCI_DEVICE_ID_SERVERWORKS_OSB4USB +#define PCI_DEVICE_ID_SERVERWORKS_CSB5USB 0x0220 #define PCI_DEVICE_ID_SERVERWORKS_CSB6USB 0x0221 #define PCI_DEVICE_ID_SERVERWORKS_CSB6LPC 0x0227 #define PCI_DEVICE_ID_SERVERWORKS_GCLE 0x0225 @@ -1792,10 +1789,6 @@ #define PCI_DEVICE_ID_PC300_TE_M_2 0x0320 #define PCI_DEVICE_ID_PC300_TE_M_1 0x0321 -/* Allied Telesyn */ -#define PCI_VENDOR_ID_AT 0x1259 -#define PCI_SUBDEVICE_ID_AT_2701FX 0x2703 - #define PCI_VENDOR_ID_ESSENTIAL 0x120f #define PCI_DEVICE_ID_ESSENTIAL_ROADRUNNER 0x0001 @@ -2340,16 +2333,13 @@ #define PCI_VENDOR_ID_DUNORD 0x5544 #define PCI_DEVICE_ID_DUNORD_I3000 0x0001 +#define PCI_VENDOR_ID_GENROCO 0x5555 +#define PCI_DEVICE_ID_GENROCO_HFP832 0x0003 + #define PCI_VENDOR_ID_DCI 0x6666 #define PCI_DEVICE_ID_DCI_PCCOM4 0x0001 #define PCI_DEVICE_ID_DCI_PCCOM8 0x0002 -#define PCI_VENDOR_ID_DUNORD 0x5544 -#define PCI_DEVICE_ID_DUNORD_I3000 0x0001 - -#define PCI_VENDOR_ID_GENROCO 0x5555 -#define PCI_DEVICE_ID_GENROCO_HFP832 0x0003 - #define PCI_VENDOR_ID_INTEL 0x8086 #define PCI_DEVICE_ID_INTEL_EESSC 0x0008 #define PCI_DEVICE_ID_INTEL_21145 0x0039 -- cgit v1.2.3 From b7924c38c9a9c6151213bc0d91776eace614e761 Mon Sep 17 00:00:00 2001 From: Grant Coady Date: Thu, 29 Sep 2005 11:06:40 +1000 Subject: [PATCH] pci_ids: remove non-referenced symbols from pci_ids.h pci_ids.h cleanup: removed non-referenced symbols, compile tested with 'make allmodconfig' Signed-off-by: Grant Coady Signed-off-by: Greg Kroah-Hartman include/linux/pci_ids.h | 540 ------------------------------------------------ 1 file changed, 540 deletions(-) --- include/linux/pci_ids.h | 540 ------------------------------------------------ 1 file changed, 540 deletions(-) (limited to 'include/linux') diff --git a/include/linux/pci_ids.h b/include/linux/pci_ids.h index 0631074954f7..11fed60a928d 100644 --- a/include/linux/pci_ids.h +++ b/include/linux/pci_ids.h @@ -132,9 +132,6 @@ #define PCI_VENDOR_ID_COMPAQ 0x0e11 #define PCI_DEVICE_ID_COMPAQ_TOKENRING 0x0508 -#define PCI_DEVICE_ID_COMPAQ_1280 0x3033 -#define PCI_DEVICE_ID_COMPAQ_TRIFLEX 0x4000 -#define PCI_DEVICE_ID_COMPAQ_6010 0x6010 #define PCI_DEVICE_ID_COMPAQ_TACHYON 0xa0fc #define PCI_DEVICE_ID_COMPAQ_SMART2P 0xae10 #define PCI_DEVICE_ID_COMPAQ_NETEL100 0xae32 @@ -281,8 +278,6 @@ #define PCI_DEVICE_ID_ATI_RAGE128_PW 0x5057 #define PCI_DEVICE_ID_ATI_RAGE128_PX 0x5058 /* Rage128 M4 */ -#define PCI_DEVICE_ID_ATI_RADEON_LE 0x4d45 -#define PCI_DEVICE_ID_ATI_RADEON_LF 0x4d46 /* Radeon R100 */ #define PCI_DEVICE_ID_ATI_RADEON_QD 0x5144 #define PCI_DEVICE_ID_ATI_RADEON_QE 0x5145 @@ -303,32 +298,22 @@ #define PCI_DEVICE_ID_ATI_RADEON_QW 0x5157 #define PCI_DEVICE_ID_ATI_RADEON_QX 0x5158 /* Radeon NV-100 */ -#define PCI_DEVICE_ID_ATI_RADEON_N1 0x5159 -#define PCI_DEVICE_ID_ATI_RADEON_N2 0x515a /* Radeon RV250 (9000) */ #define PCI_DEVICE_ID_ATI_RADEON_Id 0x4964 #define PCI_DEVICE_ID_ATI_RADEON_Ie 0x4965 #define PCI_DEVICE_ID_ATI_RADEON_If 0x4966 #define PCI_DEVICE_ID_ATI_RADEON_Ig 0x4967 /* Radeon RV280 (9200) */ -#define PCI_DEVICE_ID_ATI_RADEON_Y_ 0x5960 #define PCI_DEVICE_ID_ATI_RADEON_Ya 0x5961 #define PCI_DEVICE_ID_ATI_RADEON_Yd 0x5964 /* Radeon R300 (9500) */ -#define PCI_DEVICE_ID_ATI_RADEON_AD 0x4144 /* Radeon R300 (9700) */ #define PCI_DEVICE_ID_ATI_RADEON_ND 0x4e44 #define PCI_DEVICE_ID_ATI_RADEON_NE 0x4e45 #define PCI_DEVICE_ID_ATI_RADEON_NF 0x4e46 #define PCI_DEVICE_ID_ATI_RADEON_NG 0x4e47 -#define PCI_DEVICE_ID_ATI_RADEON_AE 0x4145 -#define PCI_DEVICE_ID_ATI_RADEON_AF 0x4146 /* Radeon R350 (9800) */ -#define PCI_DEVICE_ID_ATI_RADEON_NH 0x4e48 -#define PCI_DEVICE_ID_ATI_RADEON_NI 0x4e49 /* Radeon RV350 (9600) */ -#define PCI_DEVICE_ID_ATI_RADEON_AP 0x4150 -#define PCI_DEVICE_ID_ATI_RADEON_AR 0x4152 /* Radeon M6 */ #define PCI_DEVICE_ID_ATI_RADEON_LY 0x4c59 #define PCI_DEVICE_ID_ATI_RADEON_LZ 0x4c5a @@ -341,10 +326,6 @@ #define PCI_DEVICE_ID_ATI_RADEON_Lf 0x4c66 #define PCI_DEVICE_ID_ATI_RADEON_Lg 0x4c67 /* Radeon */ -#define PCI_DEVICE_ID_ATI_RADEON_RA 0x5144 -#define PCI_DEVICE_ID_ATI_RADEON_RB 0x5145 -#define PCI_DEVICE_ID_ATI_RADEON_RC 0x5146 -#define PCI_DEVICE_ID_ATI_RADEON_RD 0x5147 /* RadeonIGP */ #define PCI_DEVICE_ID_ATI_RS100 0xcab0 #define PCI_DEVICE_ID_ATI_RS200 0xcab2 @@ -445,34 +426,18 @@ #define PCI_DEVICE_ID_CIRRUS_5465 0x00d6 #define PCI_DEVICE_ID_CIRRUS_6729 0x1100 #define PCI_DEVICE_ID_CIRRUS_6832 0x1110 -#define PCI_DEVICE_ID_CIRRUS_7542 0x1200 #define PCI_DEVICE_ID_CIRRUS_7543 0x1202 -#define PCI_DEVICE_ID_CIRRUS_7541 0x1204 #define PCI_DEVICE_ID_CIRRUS_4610 0x6001 #define PCI_DEVICE_ID_CIRRUS_4612 0x6003 #define PCI_DEVICE_ID_CIRRUS_4615 0x6004 -#define PCI_DEVICE_ID_CIRRUS_4281 0x6005 #define PCI_VENDOR_ID_IBM 0x1014 -#define PCI_DEVICE_ID_IBM_FIRE_CORAL 0x000a #define PCI_DEVICE_ID_IBM_TR 0x0018 -#define PCI_DEVICE_ID_IBM_82G2675 0x001d -#define PCI_DEVICE_ID_IBM_MCA 0x0020 -#define PCI_DEVICE_ID_IBM_82351 0x0022 -#define PCI_DEVICE_ID_IBM_PYTHON 0x002d -#define PCI_DEVICE_ID_IBM_SERVERAID 0x002e #define PCI_DEVICE_ID_IBM_TR_WAKE 0x003e -#define PCI_DEVICE_ID_IBM_MPIC 0x0046 -#define PCI_DEVICE_ID_IBM_3780IDSP 0x007d -#define PCI_DEVICE_ID_IBM_CHUKAR 0x0096 #define PCI_DEVICE_ID_IBM_CPC710_PCI64 0x00fc -#define PCI_DEVICE_ID_IBM_CPC710_PCI32 0x0105 -#define PCI_DEVICE_ID_IBM_405GP 0x0156 #define PCI_DEVICE_ID_IBM_SNIPE 0x0180 -#define PCI_DEVICE_ID_IBM_SERVERAIDI960 0x01bd #define PCI_DEVICE_ID_IBM_CITRINE 0x028C #define PCI_DEVICE_ID_IBM_GEMSTONE 0xB166 -#define PCI_DEVICE_ID_IBM_MPIC_2 0xffff #define PCI_DEVICE_ID_IBM_ICOM_DEV_ID_1 0x0031 #define PCI_DEVICE_ID_IBM_ICOM_DEV_ID_2 0x0219 #define PCI_DEVICE_ID_IBM_ICOM_V2_TWO_PORTS_RVX 0x021A @@ -483,7 +448,6 @@ #define PCI_DEVICE_ID_COMPEX2_100VG 0x0005 #define PCI_VENDOR_ID_WD 0x101c -#define PCI_DEVICE_ID_WD_7197 0x3296 #define PCI_DEVICE_ID_WD_90C 0xc24a #define PCI_VENDOR_ID_AMI 0x101e @@ -500,33 +464,18 @@ #define PCI_DEVICE_ID_AMD_FE_GATE_7006 0x7006 #define PCI_DEVICE_ID_AMD_FE_GATE_7007 0x7007 #define PCI_DEVICE_ID_AMD_FE_GATE_700C 0x700C -#define PCI_DEVICE_ID_AMD_FE_GATE_700D 0x700D #define PCI_DEVICE_ID_AMD_FE_GATE_700E 0x700E -#define PCI_DEVICE_ID_AMD_FE_GATE_700F 0x700F -#define PCI_DEVICE_ID_AMD_COBRA_7400 0x7400 #define PCI_DEVICE_ID_AMD_COBRA_7401 0x7401 -#define PCI_DEVICE_ID_AMD_COBRA_7403 0x7403 -#define PCI_DEVICE_ID_AMD_COBRA_7404 0x7404 -#define PCI_DEVICE_ID_AMD_VIPER_7408 0x7408 #define PCI_DEVICE_ID_AMD_VIPER_7409 0x7409 #define PCI_DEVICE_ID_AMD_VIPER_740B 0x740B -#define PCI_DEVICE_ID_AMD_VIPER_740C 0x740C #define PCI_DEVICE_ID_AMD_VIPER_7410 0x7410 #define PCI_DEVICE_ID_AMD_VIPER_7411 0x7411 #define PCI_DEVICE_ID_AMD_VIPER_7413 0x7413 -#define PCI_DEVICE_ID_AMD_VIPER_7414 0x7414 -#define PCI_DEVICE_ID_AMD_OPUS_7440 0x7440 #define PCI_DEVICE_ID_AMD_VIPER_7440 0x7440 #define PCI_DEVICE_ID_AMD_OPUS_7441 0x7441 -#define PCI_DEVICE_ID_AMD_VIPER_7441 0x7441 #define PCI_DEVICE_ID_AMD_OPUS_7443 0x7443 #define PCI_DEVICE_ID_AMD_VIPER_7443 0x7443 #define PCI_DEVICE_ID_AMD_OPUS_7445 0x7445 -#define PCI_DEVICE_ID_AMD_OPUS_7448 0x7448 -#define PCI_DEVICE_ID_AMD_VIPER_7448 0x7448 -#define PCI_DEVICE_ID_AMD_OPUS_7449 0x7449 -#define PCI_DEVICE_ID_AMD_VIPER_7449 0x7449 -#define PCI_DEVICE_ID_AMD_8111_LAN 0x7462 #define PCI_DEVICE_ID_AMD_8111_LPC 0x7468 #define PCI_DEVICE_ID_AMD_8111_IDE 0x7469 #define PCI_DEVICE_ID_AMD_8111_SMBUS2 0x746a @@ -637,7 +586,6 @@ #define PCI_DEVICE_ID_SI_550 0x0550 #define PCI_DEVICE_ID_SI_540_VGA 0x5300 #define PCI_DEVICE_ID_SI_550_VGA 0x5315 -#define PCI_DEVICE_ID_SI_601 0x0601 #define PCI_DEVICE_ID_SI_620 0x0620 #define PCI_DEVICE_ID_SI_630 0x0630 #define PCI_DEVICE_ID_SI_633 0x0633 @@ -648,30 +596,22 @@ #define PCI_DEVICE_ID_SI_648 0x0648 #define PCI_DEVICE_ID_SI_650 0x0650 #define PCI_DEVICE_ID_SI_651 0x0651 -#define PCI_DEVICE_ID_SI_652 0x0652 #define PCI_DEVICE_ID_SI_655 0x0655 #define PCI_DEVICE_ID_SI_661 0x0661 #define PCI_DEVICE_ID_SI_730 0x0730 #define PCI_DEVICE_ID_SI_733 0x0733 #define PCI_DEVICE_ID_SI_630_VGA 0x6300 -#define PCI_DEVICE_ID_SI_730_VGA 0x7300 #define PCI_DEVICE_ID_SI_735 0x0735 #define PCI_DEVICE_ID_SI_740 0x0740 #define PCI_DEVICE_ID_SI_741 0x0741 #define PCI_DEVICE_ID_SI_745 0x0745 #define PCI_DEVICE_ID_SI_746 0x0746 -#define PCI_DEVICE_ID_SI_748 0x0748 -#define PCI_DEVICE_ID_SI_750 0x0750 -#define PCI_DEVICE_ID_SI_751 0x0751 -#define PCI_DEVICE_ID_SI_752 0x0752 #define PCI_DEVICE_ID_SI_755 0x0755 #define PCI_DEVICE_ID_SI_760 0x0760 #define PCI_DEVICE_ID_SI_900 0x0900 #define PCI_DEVICE_ID_SI_961 0x0961 #define PCI_DEVICE_ID_SI_962 0x0962 #define PCI_DEVICE_ID_SI_963 0x0963 -#define PCI_DEVICE_ID_SI_5107 0x5107 -#define PCI_DEVICE_ID_SI_5300 0x5300 #define PCI_DEVICE_ID_SI_5511 0x5511 #define PCI_DEVICE_ID_SI_5513 0x5513 #define PCI_DEVICE_ID_SI_5518 0x5518 @@ -683,10 +623,6 @@ #define PCI_DEVICE_ID_SI_5597 0x5597 #define PCI_DEVICE_ID_SI_5598 0x5598 #define PCI_DEVICE_ID_SI_5600 0x5600 -#define PCI_DEVICE_ID_SI_6300 0x6300 -#define PCI_DEVICE_ID_SI_6306 0x6306 -#define PCI_DEVICE_ID_SI_6326 0x6326 -#define PCI_DEVICE_ID_SI_7001 0x7001 #define PCI_DEVICE_ID_SI_7012 0x7012 #define PCI_DEVICE_ID_SI_7013 0x7013 #define PCI_DEVICE_ID_SI_7016 0x7016 @@ -707,14 +643,11 @@ #define PCI_DEVICE_ID_HP_DIVA_TOSCA1 0x1049 #define PCI_DEVICE_ID_HP_DIVA_TOSCA2 0x104A #define PCI_DEVICE_ID_HP_DIVA_MAESTRO 0x104B -#define PCI_DEVICE_ID_HP_PCI_LBA 0x1054 -#define PCI_DEVICE_ID_HP_REO_SBA 0x10f0 #define PCI_DEVICE_ID_HP_REO_IOC 0x10f1 #define PCI_DEVICE_ID_HP_VISUALIZE_FXE 0x108b #define PCI_DEVICE_ID_HP_DIVA_HALFDOME 0x1223 #define PCI_DEVICE_ID_HP_DIVA_KEYSTONE 0x1226 #define PCI_DEVICE_ID_HP_DIVA_POWERBAR 0x1227 -#define PCI_DEVICE_ID_HP_ZX1_SBA 0x1229 #define PCI_DEVICE_ID_HP_ZX1_IOC 0x122a #define PCI_DEVICE_ID_HP_PCIX_LBA 0x122e #define PCI_DEVICE_ID_HP_SX1000_IOC 0x127c @@ -722,9 +655,7 @@ #define PCI_DEVICE_ID_HP_DIVA_AUX 0x1290 #define PCI_DEVICE_ID_HP_DIVA_RMP3 0x1301 #define PCI_DEVICE_ID_HP_DIVA_HURRICANE 0x132a -#define PCI_DEVICE_ID_HP_CISS 0x3210 #define PCI_DEVICE_ID_HP_CISSA 0x3220 -#define PCI_DEVICE_ID_HP_CISSB 0x3222 #define PCI_DEVICE_ID_HP_CISSC 0x3230 #define PCI_DEVICE_ID_HP_CISSD 0x3238 #define PCI_DEVICE_ID_HP_ZX2_IOC 0x4031 @@ -732,8 +663,6 @@ #define PCI_VENDOR_ID_PCTECH 0x1042 #define PCI_DEVICE_ID_PCTECH_RZ1000 0x1000 #define PCI_DEVICE_ID_PCTECH_RZ1001 0x1001 -#define PCI_DEVICE_ID_PCTECH_SAMURAI_0 0x3000 -#define PCI_DEVICE_ID_PCTECH_SAMURAI_1 0x3010 #define PCI_DEVICE_ID_PCTECH_SAMURAI_IDE 0x3020 #define PCI_VENDOR_ID_ASUSTEK 0x1043 @@ -743,24 +672,15 @@ #define PCI_DEVICE_ID_DPT 0xa400 #define PCI_VENDOR_ID_OPTI 0x1045 -#define PCI_DEVICE_ID_OPTI_92C178 0xc178 -#define PCI_DEVICE_ID_OPTI_82C557 0xc557 #define PCI_DEVICE_ID_OPTI_82C558 0xc558 #define PCI_DEVICE_ID_OPTI_82C621 0xc621 #define PCI_DEVICE_ID_OPTI_82C700 0xc700 -#define PCI_DEVICE_ID_OPTI_82C701 0xc701 -#define PCI_DEVICE_ID_OPTI_82C814 0xc814 -#define PCI_DEVICE_ID_OPTI_82C822 0xc822 -#define PCI_DEVICE_ID_OPTI_82C861 0xc861 #define PCI_DEVICE_ID_OPTI_82C825 0xd568 #define PCI_VENDOR_ID_ELSA 0x1048 #define PCI_DEVICE_ID_ELSA_MICROLINK 0x1000 #define PCI_DEVICE_ID_ELSA_QS3000 0x3000 -#define PCI_VENDOR_ID_SGS 0x104a -#define PCI_DEVICE_ID_SGS_2000 0x0008 -#define PCI_DEVICE_ID_SGS_1764 0x0009 #define PCI_VENDOR_ID_BUSLOGIC 0x104B #define PCI_DEVICE_ID_BUSLOGIC_MULTIMASTER_NC 0x0140 @@ -768,7 +688,6 @@ #define PCI_DEVICE_ID_BUSLOGIC_FLASHPOINT 0x8130 #define PCI_VENDOR_ID_TI 0x104c -#define PCI_DEVICE_ID_TI_TVP4010 0x3d04 #define PCI_DEVICE_ID_TI_TVP4020 0x3d07 #define PCI_DEVICE_ID_TI_4450 0x8011 #define PCI_DEVICE_ID_TI_XX21_XX11 0x8031 @@ -802,14 +721,10 @@ #define PCI_DEVICE_ID_TI_X420 0xac8e #define PCI_VENDOR_ID_SONY 0x104d -#define PCI_DEVICE_ID_SONY_CXD3222 0x8039 -#define PCI_VENDOR_ID_OAK 0x104e -#define PCI_DEVICE_ID_OAK_OTI107 0x0107 /* Winbond have two vendor IDs! See 0x10ad as well */ #define PCI_VENDOR_ID_WINBOND2 0x1050 -#define PCI_DEVICE_ID_WINBOND2_89C940 0x0940 #define PCI_DEVICE_ID_WINBOND2_89C940F 0x5a5a #define PCI_DEVICE_ID_WINBOND2_6692 0x6692 @@ -818,19 +733,15 @@ #define PCI_VENDOR_ID_EFAR 0x1055 #define PCI_DEVICE_ID_EFAR_SLC90E66_1 0x9130 -#define PCI_DEVICE_ID_EFAR_SLC90E66_0 0x9460 -#define PCI_DEVICE_ID_EFAR_SLC90E66_2 0x9462 #define PCI_DEVICE_ID_EFAR_SLC90E66_3 0x9463 #define PCI_VENDOR_ID_MOTOROLA 0x1057 -#define PCI_VENDOR_ID_MOTOROLA_OOPS 0x1507 #define PCI_DEVICE_ID_MOTOROLA_MPC105 0x0001 #define PCI_DEVICE_ID_MOTOROLA_MPC106 0x0002 #define PCI_DEVICE_ID_MOTOROLA_MPC107 0x0004 #define PCI_DEVICE_ID_MOTOROLA_RAVEN 0x4801 #define PCI_DEVICE_ID_MOTOROLA_FALCON 0x4802 #define PCI_DEVICE_ID_MOTOROLA_HAWK 0x4803 -#define PCI_DEVICE_ID_MOTOROLA_CPX8216 0x4806 #define PCI_DEVICE_ID_MOTOROLA_HARRIER 0x480b #define PCI_DEVICE_ID_MOTOROLA_MPC5200 0x5803 @@ -841,33 +752,19 @@ #define PCI_DEVICE_ID_PROMISE_20262 0x4d38 #define PCI_DEVICE_ID_PROMISE_20263 0x0D38 #define PCI_DEVICE_ID_PROMISE_20268 0x4d68 -#define PCI_DEVICE_ID_PROMISE_20268R 0x6268 #define PCI_DEVICE_ID_PROMISE_20269 0x4d69 #define PCI_DEVICE_ID_PROMISE_20270 0x6268 #define PCI_DEVICE_ID_PROMISE_20271 0x6269 #define PCI_DEVICE_ID_PROMISE_20275 0x1275 #define PCI_DEVICE_ID_PROMISE_20276 0x5275 #define PCI_DEVICE_ID_PROMISE_20277 0x7275 -#define PCI_DEVICE_ID_PROMISE_5300 0x5300 -#define PCI_VENDOR_ID_N9 0x105d -#define PCI_DEVICE_ID_N9_I128 0x2309 -#define PCI_DEVICE_ID_N9_I128_2 0x2339 -#define PCI_DEVICE_ID_N9_I128_T2R 0x493d #define PCI_VENDOR_ID_UMC 0x1060 #define PCI_DEVICE_ID_UMC_UM8673F 0x0101 -#define PCI_DEVICE_ID_UMC_UM8891A 0x0891 #define PCI_DEVICE_ID_UMC_UM8886BF 0x673a #define PCI_DEVICE_ID_UMC_UM8886A 0x886a -#define PCI_DEVICE_ID_UMC_UM8881F 0x8881 -#define PCI_DEVICE_ID_UMC_UM8886F 0x8886 -#define PCI_DEVICE_ID_UMC_UM9017F 0x9017 -#define PCI_DEVICE_ID_UMC_UM8886N 0xe886 -#define PCI_DEVICE_ID_UMC_UM8891N 0xe891 -#define PCI_VENDOR_ID_X 0x1061 -#define PCI_DEVICE_ID_X_AGX016 0x0001 #define PCI_VENDOR_ID_MYLEX 0x1069 #define PCI_DEVICE_ID_MYLEX_DAC960_P 0x0001 @@ -878,37 +775,26 @@ #define PCI_DEVICE_ID_MYLEX_DAC960_BA 0xBA56 #define PCI_DEVICE_ID_MYLEX_DAC960_GEM 0xB166 -#define PCI_VENDOR_ID_PICOP 0x1066 -#define PCI_DEVICE_ID_PICOP_PT86C52X 0x0001 -#define PCI_DEVICE_ID_PICOP_PT80C524 0x8002 #define PCI_VENDOR_ID_APPLE 0x106b #define PCI_DEVICE_ID_APPLE_BANDIT 0x0001 -#define PCI_DEVICE_ID_APPLE_GC 0x0002 #define PCI_DEVICE_ID_APPLE_HYDRA 0x000e #define PCI_DEVICE_ID_APPLE_UNI_N_FW 0x0018 -#define PCI_DEVICE_ID_APPLE_KL_USB 0x0019 #define PCI_DEVICE_ID_APPLE_UNI_N_AGP 0x0020 #define PCI_DEVICE_ID_APPLE_UNI_N_GMAC 0x0021 -#define PCI_DEVICE_ID_APPLE_KEYLARGO 0x0022 #define PCI_DEVICE_ID_APPLE_UNI_N_GMACP 0x0024 -#define PCI_DEVICE_ID_APPLE_KEYLARGO_P 0x0025 -#define PCI_DEVICE_ID_APPLE_KL_USB_P 0x0026 #define PCI_DEVICE_ID_APPLE_UNI_N_AGP_P 0x0027 #define PCI_DEVICE_ID_APPLE_UNI_N_AGP15 0x002d #define PCI_DEVICE_ID_APPLE_UNI_N_PCI15 0x002e -#define PCI_DEVICE_ID_APPLE_UNI_N_FW2 0x0030 #define PCI_DEVICE_ID_APPLE_UNI_N_GMAC2 0x0032 #define PCI_DEVICE_ID_APPLE_UNI_N_ATA 0x0033 #define PCI_DEVICE_ID_APPLE_UNI_N_AGP2 0x0034 #define PCI_DEVICE_ID_APPLE_IPID_ATA100 0x003b -#define PCI_DEVICE_ID_APPLE_KEYLARGO_I 0x003e #define PCI_DEVICE_ID_APPLE_K2_ATA100 0x0043 #define PCI_DEVICE_ID_APPLE_U3_AGP 0x004b #define PCI_DEVICE_ID_APPLE_K2_GMAC 0x004c #define PCI_DEVICE_ID_APPLE_SH_ATA 0x0050 #define PCI_DEVICE_ID_APPLE_SH_SUNGEM 0x0051 -#define PCI_DEVICE_ID_APPLE_SH_FW 0x0052 #define PCI_DEVICE_ID_APPLE_U3L_AGP 0x0058 #define PCI_DEVICE_ID_APPLE_U3H_AGP 0x0059 #define PCI_DEVICE_ID_APPLE_TIGON3 0x1645 @@ -921,12 +807,9 @@ #define PCI_DEVICE_ID_YAMAHA_744 0x0010 #define PCI_DEVICE_ID_YAMAHA_754 0x0012 -#define PCI_VENDOR_ID_NEXGEN 0x1074 -#define PCI_DEVICE_ID_NEXGEN_82C501 0x4e78 #define PCI_VENDOR_ID_QLOGIC 0x1077 #define PCI_DEVICE_ID_QLOGIC_ISP1020 0x1020 -#define PCI_DEVICE_ID_QLOGIC_ISP1022 0x1022 #define PCI_DEVICE_ID_QLOGIC_ISP2100 0x2100 #define PCI_DEVICE_ID_QLOGIC_ISP2200 0x2200 #define PCI_DEVICE_ID_QLOGIC_ISP2300 0x2300 @@ -944,32 +827,20 @@ #define PCI_DEVICE_ID_CYRIX_PCI_MASTER 0x0001 #define PCI_DEVICE_ID_CYRIX_5520 0x0002 #define PCI_DEVICE_ID_CYRIX_5530_LEGACY 0x0100 -#define PCI_DEVICE_ID_CYRIX_5530_SMI 0x0101 #define PCI_DEVICE_ID_CYRIX_5530_IDE 0x0102 #define PCI_DEVICE_ID_CYRIX_5530_AUDIO 0x0103 #define PCI_DEVICE_ID_CYRIX_5530_VIDEO 0x0104 -#define PCI_VENDOR_ID_LEADTEK 0x107d -#define PCI_DEVICE_ID_LEADTEK_805 0x0000 -#define PCI_VENDOR_ID_INTERPHASE 0x107e -#define PCI_DEVICE_ID_INTERPHASE_5526 0x0004 -#define PCI_DEVICE_ID_INTERPHASE_55x6 0x0005 -#define PCI_DEVICE_ID_INTERPHASE_5575 0x0008 #define PCI_VENDOR_ID_CONTAQ 0x1080 -#define PCI_DEVICE_ID_CONTAQ_82C599 0x0600 #define PCI_DEVICE_ID_CONTAQ_82C693 0xc693 -#define PCI_VENDOR_ID_FOREX 0x1083 #define PCI_VENDOR_ID_OLICOM 0x108d -#define PCI_DEVICE_ID_OLICOM_OC3136 0x0001 -#define PCI_DEVICE_ID_OLICOM_OC2315 0x0011 #define PCI_DEVICE_ID_OLICOM_OC2325 0x0012 #define PCI_DEVICE_ID_OLICOM_OC2183 0x0013 #define PCI_DEVICE_ID_OLICOM_OC2326 0x0014 -#define PCI_DEVICE_ID_OLICOM_OC6151 0x0021 #define PCI_VENDOR_ID_SUN 0x108e #define PCI_DEVICE_ID_SUN_EBUS 0x1000 @@ -988,49 +859,31 @@ #define PCI_DEVICE_ID_SUN_CASSINI 0xabba #define PCI_VENDOR_ID_CMD 0x1095 -#define PCI_DEVICE_ID_CMD_640 0x0640 #define PCI_DEVICE_ID_CMD_643 0x0643 #define PCI_DEVICE_ID_CMD_646 0x0646 -#define PCI_DEVICE_ID_CMD_647 0x0647 #define PCI_DEVICE_ID_CMD_648 0x0648 #define PCI_DEVICE_ID_CMD_649 0x0649 -#define PCI_DEVICE_ID_CMD_670 0x0670 -#define PCI_DEVICE_ID_CMD_680 0x0680 #define PCI_DEVICE_ID_SII_680 0x0680 #define PCI_DEVICE_ID_SII_3112 0x3112 #define PCI_DEVICE_ID_SII_1210SA 0x0240 -#define PCI_VENDOR_ID_VISION 0x1098 -#define PCI_DEVICE_ID_VISION_QD8500 0x0001 -#define PCI_DEVICE_ID_VISION_QD8580 0x0002 #define PCI_VENDOR_ID_BROOKTREE 0x109e -#define PCI_DEVICE_ID_BROOKTREE_848 0x0350 -#define PCI_DEVICE_ID_BROOKTREE_849A 0x0351 -#define PCI_DEVICE_ID_BROOKTREE_878_1 0x036e #define PCI_DEVICE_ID_BROOKTREE_878 0x0878 #define PCI_DEVICE_ID_BROOKTREE_879 0x0879 -#define PCI_DEVICE_ID_BROOKTREE_8474 0x8474 -#define PCI_VENDOR_ID_SIERRA 0x10a8 -#define PCI_DEVICE_ID_SIERRA_STB 0x0000 #define PCI_VENDOR_ID_SGI 0x10a9 #define PCI_DEVICE_ID_SGI_IOC3 0x0003 #define PCI_DEVICE_ID_SGI_IOC4 0x100a #define PCI_VENDOR_ID_SGI_LITHIUM 0x1002 -#define PCI_VENDOR_ID_ACC 0x10aa -#define PCI_DEVICE_ID_ACC_2056 0x0000 #define PCI_VENDOR_ID_WINBOND 0x10ad -#define PCI_DEVICE_ID_WINBOND_83769 0x0001 #define PCI_DEVICE_ID_WINBOND_82C105 0x0105 #define PCI_DEVICE_ID_WINBOND_83C553 0x0565 -#define PCI_VENDOR_ID_DATABOOK 0x10b3 -#define PCI_DEVICE_ID_DATABOOK_87144 0xb106 #define PCI_VENDOR_ID_PLX 0x10b5 #define PCI_DEVICE_ID_PLX_R685 0x1030 @@ -1041,33 +894,19 @@ #define PCI_DEVICE_ID_PLX_DJINN_ITOO 0x1151 #define PCI_DEVICE_ID_PLX_R753 0x1152 #define PCI_DEVICE_ID_PLX_OLITEC 0x1187 -#define PCI_DEVICE_ID_PLX_9030 0x9030 #define PCI_DEVICE_ID_PLX_9050 0x9050 -#define PCI_DEVICE_ID_PLX_9060 0x9060 -#define PCI_DEVICE_ID_PLX_9060ES 0x906E -#define PCI_DEVICE_ID_PLX_9060SD 0x906D #define PCI_DEVICE_ID_PLX_9080 0x9080 #define PCI_DEVICE_ID_PLX_GTEK_SERIAL2 0xa001 #define PCI_VENDOR_ID_MADGE 0x10b6 #define PCI_DEVICE_ID_MADGE_MK2 0x0002 -#define PCI_DEVICE_ID_MADGE_C155S 0x1001 #define PCI_VENDOR_ID_3COM 0x10b7 #define PCI_DEVICE_ID_3COM_3C985 0x0001 #define PCI_DEVICE_ID_3COM_3C940 0x1700 #define PCI_DEVICE_ID_3COM_3C339 0x3390 #define PCI_DEVICE_ID_3COM_3C359 0x3590 -#define PCI_DEVICE_ID_3COM_3C590 0x5900 -#define PCI_DEVICE_ID_3COM_3C595TX 0x5950 -#define PCI_DEVICE_ID_3COM_3C595T4 0x5951 -#define PCI_DEVICE_ID_3COM_3C595MII 0x5952 #define PCI_DEVICE_ID_3COM_3C940B 0x80eb -#define PCI_DEVICE_ID_3COM_3C900TPO 0x9000 -#define PCI_DEVICE_ID_3COM_3C900COMBO 0x9001 -#define PCI_DEVICE_ID_3COM_3C905TX 0x9050 -#define PCI_DEVICE_ID_3COM_3C905T4 0x9051 -#define PCI_DEVICE_ID_3COM_3C905B_TX 0x9055 #define PCI_DEVICE_ID_3COM_3CR990 0x9900 #define PCI_DEVICE_ID_3COM_3CR990_TX_95 0x9902 #define PCI_DEVICE_ID_3COM_3CR990_TX_97 0x9903 @@ -1077,24 +916,11 @@ #define PCI_DEVICE_ID_3COM_3CR990SVR97 0x9909 #define PCI_DEVICE_ID_3COM_3CR990SVR 0x990a -#define PCI_VENDOR_ID_SMC 0x10b8 -#define PCI_DEVICE_ID_SMC_EPIC100 0x0005 #define PCI_VENDOR_ID_AL 0x10b9 -#define PCI_DEVICE_ID_AL_M1445 0x1445 -#define PCI_DEVICE_ID_AL_M1449 0x1449 -#define PCI_DEVICE_ID_AL_M1451 0x1451 -#define PCI_DEVICE_ID_AL_M1461 0x1461 -#define PCI_DEVICE_ID_AL_M1489 0x1489 -#define PCI_DEVICE_ID_AL_M1511 0x1511 -#define PCI_DEVICE_ID_AL_M1513 0x1513 -#define PCI_DEVICE_ID_AL_M1521 0x1521 -#define PCI_DEVICE_ID_AL_M1523 0x1523 -#define PCI_DEVICE_ID_AL_M1531 0x1531 #define PCI_DEVICE_ID_AL_M1533 0x1533 #define PCI_DEVICE_ID_AL_M1535 0x1535 #define PCI_DEVICE_ID_AL_M1541 0x1541 -#define PCI_DEVICE_ID_AL_M1543 0x1543 #define PCI_DEVICE_ID_AL_M1563 0x1563 #define PCI_DEVICE_ID_AL_M1621 0x1621 #define PCI_DEVICE_ID_AL_M1631 0x1631 @@ -1107,49 +933,23 @@ #define PCI_DEVICE_ID_AL_M1681 0x1681 #define PCI_DEVICE_ID_AL_M1683 0x1683 #define PCI_DEVICE_ID_AL_M1689 0x1689 -#define PCI_DEVICE_ID_AL_M3307 0x3307 -#define PCI_DEVICE_ID_AL_M4803 0x5215 #define PCI_DEVICE_ID_AL_M5219 0x5219 #define PCI_DEVICE_ID_AL_M5228 0x5228 #define PCI_DEVICE_ID_AL_M5229 0x5229 -#define PCI_DEVICE_ID_AL_M5237 0x5237 -#define PCI_DEVICE_ID_AL_M5243 0x5243 #define PCI_DEVICE_ID_AL_M5451 0x5451 #define PCI_DEVICE_ID_AL_M7101 0x7101 -#define PCI_VENDOR_ID_MITSUBISHI 0x10ba -#define PCI_VENDOR_ID_SURECOM 0x10bd -#define PCI_DEVICE_ID_SURECOM_NE34 0x0e34 #define PCI_VENDOR_ID_NEOMAGIC 0x10c8 -#define PCI_DEVICE_ID_NEOMAGIC_MAGICGRAPH_NM2070 0x0001 -#define PCI_DEVICE_ID_NEOMAGIC_MAGICGRAPH_128V 0x0002 -#define PCI_DEVICE_ID_NEOMAGIC_MAGICGRAPH_128ZV 0x0003 -#define PCI_DEVICE_ID_NEOMAGIC_MAGICGRAPH_NM2160 0x0004 -#define PCI_DEVICE_ID_NEOMAGIC_MAGICMEDIA_256AV 0x0005 -#define PCI_DEVICE_ID_NEOMAGIC_MAGICGRAPH_128ZVPLUS 0x0083 #define PCI_DEVICE_ID_NEOMAGIC_NM256AV_AUDIO 0x8005 #define PCI_DEVICE_ID_NEOMAGIC_NM256ZX_AUDIO 0x8006 #define PCI_DEVICE_ID_NEOMAGIC_NM256XL_PLUS_AUDIO 0x8016 -#define PCI_VENDOR_ID_ASP 0x10cd -#define PCI_DEVICE_ID_ASP_ABP940 0x1200 -#define PCI_DEVICE_ID_ASP_ABP940U 0x1300 -#define PCI_DEVICE_ID_ASP_ABP940UW 0x2300 - -#define PCI_VENDOR_ID_MACRONIX 0x10d9 -#define PCI_DEVICE_ID_MACRONIX_MX98713 0x0512 -#define PCI_DEVICE_ID_MACRONIX_MX987x5 0x0531 #define PCI_VENDOR_ID_TCONRAD 0x10da #define PCI_DEVICE_ID_TCONRAD_TOKENRING 0x0508 -#define PCI_VENDOR_ID_CERN 0x10dc -#define PCI_DEVICE_ID_CERN_SPSB_PMC 0x0001 -#define PCI_DEVICE_ID_CERN_SPSB_PCI 0x0002 -#define PCI_DEVICE_ID_CERN_HIPPI_DST 0x0021 -#define PCI_DEVICE_ID_CERN_HIPPI_SRC 0x0022 #define PCI_VENDOR_ID_NVIDIA 0x10de #define PCI_DEVICE_ID_NVIDIA_TNT 0x0020 @@ -1281,7 +1081,6 @@ #define PCI_DEVICE_ID_NVIDIA_NFORCE_MCP55_SATA2 0x037F #define PCI_DEVICE_ID_NVIDIA_NVENET_12 0x0268 #define PCI_DEVICE_ID_NVIDIA_NVENET_13 0x0269 -#define PCI_DEVICE_ID_NVIDIA_MCP51_AUDIO 0x026B #define PCI_DEVICE_ID_NVIDIA_GEFORCE4_TI_4800 0x0280 #define PCI_DEVICE_ID_NVIDIA_GEFORCE4_TI_4800_8X 0x0281 #define PCI_DEVICE_ID_NVIDIA_GEFORCE4_TI_4800SE 0x0282 @@ -1332,24 +1131,13 @@ #define PCI_DEVICE_ID_NVIDIA_NVENET_15 0x0373 #define PCI_VENDOR_ID_IMS 0x10e0 -#define PCI_DEVICE_ID_IMS_8849 0x8849 #define PCI_DEVICE_ID_IMS_TT128 0x9128 #define PCI_DEVICE_ID_IMS_TT3D 0x9135 -#define PCI_VENDOR_ID_TEKRAM2 0x10e1 -#define PCI_DEVICE_ID_TEKRAM2_690c 0x690c -#define PCI_VENDOR_ID_TUNDRA 0x10e3 -#define PCI_DEVICE_ID_TUNDRA_CA91C042 0x0000 -#define PCI_VENDOR_ID_AMCC 0x10e8 -#define PCI_DEVICE_ID_AMCC_MYRINET 0x8043 -#define PCI_DEVICE_ID_AMCC_PARASTATION 0x8062 -#define PCI_DEVICE_ID_AMCC_S5933 0x807d -#define PCI_DEVICE_ID_AMCC_S5933_HEPC3 0x809c #define PCI_VENDOR_ID_INTERG 0x10ea -#define PCI_DEVICE_ID_INTERG_1680 0x1680 #define PCI_DEVICE_ID_INTERG_1682 0x1682 #define PCI_DEVICE_ID_INTERG_2000 0x2000 #define PCI_DEVICE_ID_INTERG_2010 0x2010 @@ -1357,27 +1145,18 @@ #define PCI_DEVICE_ID_INTERG_5050 0x5050 #define PCI_VENDOR_ID_REALTEK 0x10ec -#define PCI_DEVICE_ID_REALTEK_8029 0x8029 -#define PCI_DEVICE_ID_REALTEK_8129 0x8129 #define PCI_DEVICE_ID_REALTEK_8139 0x8139 -#define PCI_DEVICE_ID_REALTEK_8169 0x8169 #define PCI_VENDOR_ID_XILINX 0x10ee #define PCI_DEVICE_ID_RME_DIGI96 0x3fc0 #define PCI_DEVICE_ID_RME_DIGI96_8 0x3fc1 #define PCI_DEVICE_ID_RME_DIGI96_8_PRO 0x3fc2 #define PCI_DEVICE_ID_RME_DIGI96_8_PAD_OR_PST 0x3fc3 -#define PCI_DEVICE_ID_XILINX_HAMMERFALL 0x3fc4 #define PCI_DEVICE_ID_XILINX_HAMMERFALL_DSP 0x3fc5 #define PCI_DEVICE_ID_XILINX_HAMMERFALL_DSP_MADI 0x3fc6 -#define PCI_DEVICE_ID_TURBOPAM 0x4020 -#define PCI_VENDOR_ID_TRUEVISION 0x10fa -#define PCI_DEVICE_ID_TRUEVISION_T1000 0x000c #define PCI_VENDOR_ID_INIT 0x1101 -#define PCI_DEVICE_ID_INIT_320P 0x9100 -#define PCI_DEVICE_ID_INIT_360P 0x9500 #define PCI_VENDOR_ID_CREATIVE 0x1102 // duplicate: ECTIVA #define PCI_DEVICE_ID_CREATIVE_EMU10K1 0x0002 @@ -1405,36 +1184,25 @@ #define PCI_DEVICE_ID_VIA_8363_0 0x0305 #define PCI_DEVICE_ID_VIA_8371_0 0x0391 #define PCI_DEVICE_ID_VIA_8501_0 0x0501 -#define PCI_DEVICE_ID_VIA_82C505 0x0505 #define PCI_DEVICE_ID_VIA_82C561 0x0561 #define PCI_DEVICE_ID_VIA_82C586_1 0x0571 #define PCI_DEVICE_ID_VIA_82C576 0x0576 -#define PCI_DEVICE_ID_VIA_82C585 0x0585 #define PCI_DEVICE_ID_VIA_82C586_0 0x0586 -#define PCI_DEVICE_ID_VIA_82C595 0x0595 #define PCI_DEVICE_ID_VIA_82C596 0x0596 #define PCI_DEVICE_ID_VIA_82C597_0 0x0597 #define PCI_DEVICE_ID_VIA_82C598_0 0x0598 #define PCI_DEVICE_ID_VIA_8601_0 0x0601 #define PCI_DEVICE_ID_VIA_8605_0 0x0605 -#define PCI_DEVICE_ID_VIA_82C680 0x0680 #define PCI_DEVICE_ID_VIA_82C686 0x0686 #define PCI_DEVICE_ID_VIA_82C691_0 0x0691 -#define PCI_DEVICE_ID_VIA_82C693 0x0693 -#define PCI_DEVICE_ID_VIA_82C693_1 0x0698 -#define PCI_DEVICE_ID_VIA_82C926 0x0926 #define PCI_DEVICE_ID_VIA_82C576_1 0x1571 -#define PCI_DEVICE_ID_VIA_82C595_97 0x1595 #define PCI_DEVICE_ID_VIA_82C586_2 0x3038 #define PCI_DEVICE_ID_VIA_82C586_3 0x3040 -#define PCI_DEVICE_ID_VIA_6305 0x3044 #define PCI_DEVICE_ID_VIA_82C596_3 0x3050 #define PCI_DEVICE_ID_VIA_82C596B_3 0x3051 #define PCI_DEVICE_ID_VIA_82C686_4 0x3057 #define PCI_DEVICE_ID_VIA_82C686_5 0x3058 #define PCI_DEVICE_ID_VIA_8233_5 0x3059 -#define PCI_DEVICE_ID_VIA_8233_7 0x3065 -#define PCI_DEVICE_ID_VIA_82C686_6 0x3068 #define PCI_DEVICE_ID_VIA_8233_0 0x3074 #define PCI_DEVICE_ID_VIA_8633_0 0x3091 #define PCI_DEVICE_ID_VIA_8367_0 0x3099 @@ -1452,38 +1220,23 @@ #define PCI_DEVICE_ID_VIA_XN266 0x3156 #define PCI_DEVICE_ID_VIA_8754C_0 0x3168 #define PCI_DEVICE_ID_VIA_8235 0x3177 -#define PCI_DEVICE_ID_VIA_P4N333 0x3178 #define PCI_DEVICE_ID_VIA_8385_0 0x3188 #define PCI_DEVICE_ID_VIA_8377_0 0x3189 #define PCI_DEVICE_ID_VIA_8378_0 0x3205 #define PCI_DEVICE_ID_VIA_8783_0 0x3208 -#define PCI_DEVICE_ID_VIA_P4M400 0x3209 #define PCI_DEVICE_ID_VIA_8237 0x3227 #define PCI_DEVICE_ID_VIA_3296_0 0x0296 -#define PCI_DEVICE_ID_VIA_86C100A 0x6100 #define PCI_DEVICE_ID_VIA_8231 0x8231 #define PCI_DEVICE_ID_VIA_8231_4 0x8235 #define PCI_DEVICE_ID_VIA_8365_1 0x8305 #define PCI_DEVICE_ID_VIA_8371_1 0x8391 -#define PCI_DEVICE_ID_VIA_8501_1 0x8501 -#define PCI_DEVICE_ID_VIA_82C597_1 0x8597 #define PCI_DEVICE_ID_VIA_82C598_1 0x8598 -#define PCI_DEVICE_ID_VIA_8601_1 0x8601 -#define PCI_DEVICE_ID_VIA_8505_1 0x8605 -#define PCI_DEVICE_ID_VIA_8633_1 0xB091 -#define PCI_DEVICE_ID_VIA_8367_1 0xB099 -#define PCI_DEVICE_ID_VIA_P4X266_1 0xB101 -#define PCI_DEVICE_ID_VIA_8615_1 0xB103 -#define PCI_DEVICE_ID_VIA_8361_1 0xB112 -#define PCI_DEVICE_ID_VIA_8235_1 0xB168 #define PCI_DEVICE_ID_VIA_838X_1 0xB188 #define PCI_DEVICE_ID_VIA_83_87XX_1 0xB198 #define PCI_VENDOR_ID_SIEMENS 0x110A #define PCI_DEVICE_ID_SIEMENS_DSCC4 0x2102 -#define PCI_VENDOR_ID_SMC2 0x1113 -#define PCI_DEVICE_ID_SMC2_1211TX 0x1211 #define PCI_VENDOR_ID_VORTEX 0x1119 #define PCI_DEVICE_ID_VORTEX_GDT60x0 0x0000 @@ -1506,18 +1259,6 @@ #define PCI_DEVICE_ID_VORTEX_GDT6557RP 0x0103 #define PCI_DEVICE_ID_VORTEX_GDT6x11RP 0x0104 #define PCI_DEVICE_ID_VORTEX_GDT6x21RP 0x0105 -#define PCI_DEVICE_ID_VORTEX_GDT6x17RP1 0x0110 -#define PCI_DEVICE_ID_VORTEX_GDT6x27RP1 0x0111 -#define PCI_DEVICE_ID_VORTEX_GDT6537RP1 0x0112 -#define PCI_DEVICE_ID_VORTEX_GDT6557RP1 0x0113 -#define PCI_DEVICE_ID_VORTEX_GDT6x11RP1 0x0114 -#define PCI_DEVICE_ID_VORTEX_GDT6x21RP1 0x0115 -#define PCI_DEVICE_ID_VORTEX_GDT6x17RP2 0x0120 -#define PCI_DEVICE_ID_VORTEX_GDT6x27RP2 0x0121 -#define PCI_DEVICE_ID_VORTEX_GDT6537RP2 0x0122 -#define PCI_DEVICE_ID_VORTEX_GDT6557RP2 0x0123 -#define PCI_DEVICE_ID_VORTEX_GDT6x11RP2 0x0124 -#define PCI_DEVICE_ID_VORTEX_GDT6x21RP2 0x0125 #define PCI_VENDOR_ID_EF 0x111a #define PCI_DEVICE_ID_EF_ATM_FPGA 0x0000 @@ -1529,21 +1270,15 @@ #define PCI_DEVICE_ID_IDT_IDT77201 0x0001 #define PCI_VENDOR_ID_FORE 0x1127 -#define PCI_DEVICE_ID_FORE_PCA200PC 0x0210 #define PCI_DEVICE_ID_FORE_PCA200E 0x0300 -#define PCI_VENDOR_ID_IMAGINGTECH 0x112f -#define PCI_DEVICE_ID_IMAGINGTECH_ICPCI 0x0000 #define PCI_VENDOR_ID_PHILIPS 0x1131 -#define PCI_DEVICE_ID_PHILIPS_SAA7145 0x7145 #define PCI_DEVICE_ID_PHILIPS_SAA7146 0x7146 #define PCI_DEVICE_ID_PHILIPS_SAA9730 0x9730 #define PCI_VENDOR_ID_EICON 0x1133 -#define PCI_DEVICE_ID_EICON_DIVA20PRO 0xe001 #define PCI_DEVICE_ID_EICON_DIVA20 0xe002 -#define PCI_DEVICE_ID_EICON_DIVA20PRO_U 0xe003 #define PCI_DEVICE_ID_EICON_DIVA20_U 0xe004 #define PCI_DEVICE_ID_EICON_DIVA201 0xe005 #define PCI_DEVICE_ID_EICON_DIVA202 0xe00b @@ -1555,35 +1290,17 @@ #define PCI_VENDOR_ID_ZIATECH 0x1138 #define PCI_DEVICE_ID_ZIATECH_5550_HC 0x5550 -#define PCI_VENDOR_ID_CYCLONE 0x113c -#define PCI_DEVICE_ID_CYCLONE_SDK 0x0001 -#define PCI_VENDOR_ID_ALLIANCE 0x1142 -#define PCI_DEVICE_ID_ALLIANCE_PROMOTIO 0x3210 -#define PCI_DEVICE_ID_ALLIANCE_PROVIDEO 0x6422 -#define PCI_DEVICE_ID_ALLIANCE_AT24 0x6424 -#define PCI_DEVICE_ID_ALLIANCE_AT3D 0x643d #define PCI_VENDOR_ID_SYSKONNECT 0x1148 -#define PCI_DEVICE_ID_SYSKONNECT_FP 0x4000 #define PCI_DEVICE_ID_SYSKONNECT_TR 0x4200 #define PCI_DEVICE_ID_SYSKONNECT_GE 0x4300 #define PCI_DEVICE_ID_SYSKONNECT_YU 0x4320 #define PCI_DEVICE_ID_SYSKONNECT_9DXX 0x4400 #define PCI_DEVICE_ID_SYSKONNECT_9MXX 0x4500 -#define PCI_VENDOR_ID_VMIC 0x114a -#define PCI_DEVICE_ID_VMIC_VME 0x7587 #define PCI_VENDOR_ID_DIGI 0x114f -#define PCI_DEVICE_ID_DIGI_EPC 0x0002 -#define PCI_DEVICE_ID_DIGI_RIGHTSWITCH 0x0003 -#define PCI_DEVICE_ID_DIGI_XEM 0x0004 -#define PCI_DEVICE_ID_DIGI_XR 0x0005 -#define PCI_DEVICE_ID_DIGI_CX 0x0006 -#define PCI_DEVICE_ID_DIGI_XRJ 0x0009 -#define PCI_DEVICE_ID_DIGI_EPCJ 0x000a -#define PCI_DEVICE_ID_DIGI_XR_920 0x0027 #define PCI_DEVICE_ID_DIGI_DF_M_IOM2_E 0x0070 #define PCI_DEVICE_ID_DIGI_DF_M_E 0x0071 #define PCI_DEVICE_ID_DIGI_DF_M_IOM2_A 0x0072 @@ -1593,23 +1310,15 @@ #define PCI_DEVICE_ID_NEO_2RJ45 0x00CA #define PCI_DEVICE_ID_NEO_2RJ45PRI 0x00CB -#define PCI_VENDOR_ID_MUTECH 0x1159 -#define PCI_DEVICE_ID_MUTECH_MV1000 0x0001 #define PCI_VENDOR_ID_XIRCOM 0x115d -#define PCI_DEVICE_ID_XIRCOM_X3201_ETH 0x0003 #define PCI_DEVICE_ID_XIRCOM_RBM56G 0x0101 #define PCI_DEVICE_ID_XIRCOM_X3201_MDM 0x0103 -#define PCI_VENDOR_ID_RENDITION 0x1163 -#define PCI_DEVICE_ID_RENDITION_VERITE 0x0001 -#define PCI_DEVICE_ID_RENDITION_VERITE2100 0x2000 #define PCI_VENDOR_ID_SERVERWORKS 0x1166 #define PCI_DEVICE_ID_SERVERWORKS_HE 0x0008 #define PCI_DEVICE_ID_SERVERWORKS_LE 0x0009 -#define PCI_DEVICE_ID_SERVERWORKS_CIOB30 0x0010 -#define PCI_DEVICE_ID_SERVERWORKS_CMIC_HE 0x0011 #define PCI_DEVICE_ID_SERVERWORKS_GCNB_LE 0x0017 #define PCI_DEVICE_ID_SERVERWORKS_OSB4 0x0200 #define PCI_DEVICE_ID_SERVERWORKS_CSB5 0x0201 @@ -1619,13 +1328,7 @@ #define PCI_DEVICE_ID_SERVERWORKS_CSB6IDE 0x0213 #define PCI_DEVICE_ID_SERVERWORKS_HT1000IDE 0x0214 #define PCI_DEVICE_ID_SERVERWORKS_CSB6IDE2 0x0217 -#define PCI_DEVICE_ID_SERVERWORKS_OSB4USB 0x0220 -#define PCI_DEVICE_ID_SERVERWORKS_CSB5USB 0x0220 -#define PCI_DEVICE_ID_SERVERWORKS_CSB6USB 0x0221 #define PCI_DEVICE_ID_SERVERWORKS_CSB6LPC 0x0227 -#define PCI_DEVICE_ID_SERVERWORKS_GCLE 0x0225 -#define PCI_DEVICE_ID_SERVERWORKS_GCLE2 0x0227 -#define PCI_DEVICE_ID_SERVERWORKS_CSB5ISA 0x0230 #define PCI_VENDOR_ID_SBE 0x1176 #define PCI_DEVICE_ID_SBE_WANXL100 0x0301 @@ -1636,17 +1339,12 @@ #define PCI_DEVICE_ID_TOSHIBA_PICCOLO 0x0102 #define PCI_DEVICE_ID_TOSHIBA_PICCOLO_1 0x0103 #define PCI_DEVICE_ID_TOSHIBA_PICCOLO_2 0x0105 -#define PCI_DEVICE_ID_TOSHIBA_601 0x0601 #define PCI_DEVICE_ID_TOSHIBA_TOPIC95 0x060a -#define PCI_DEVICE_ID_TOSHIBA_TOPIC95_A 0x0603 -#define PCI_DEVICE_ID_TOSHIBA_TOPIC95_B 0x060a #define PCI_DEVICE_ID_TOSHIBA_TOPIC97 0x060f #define PCI_DEVICE_ID_TOSHIBA_TOPIC100 0x0617 #define PCI_VENDOR_ID_TOSHIBA_2 0x102f -#define PCI_DEVICE_ID_TOSHIBA_TX3927 0x000a #define PCI_DEVICE_ID_TOSHIBA_TC35815CF 0x0030 -#define PCI_DEVICE_ID_TOSHIBA_TX4927 0x0180 #define PCI_DEVICE_ID_TOSHIBA_TC86C001_MISC 0x0108 #define PCI_DEVICE_ID_TOSHIBA_SPIDER_NET 0x01b3 @@ -1661,7 +1359,6 @@ #define PCI_DEVICE_ID_DLINK_DGE510T 0x4c00 #define PCI_VENDOR_ID_ARTOP 0x1191 -#define PCI_DEVICE_ID_ARTOP_ATP8400 0x0004 #define PCI_DEVICE_ID_ARTOP_ATP850UF 0x0005 #define PCI_DEVICE_ID_ARTOP_ATP860 0x0006 #define PCI_DEVICE_ID_ARTOP_ATP860R 0x0007 @@ -1674,16 +1371,11 @@ #define PCI_DEVICE_ID_ARTOP_AEC7612D 0x8040 #define PCI_DEVICE_ID_ARTOP_AEC7612SUW 0x8050 #define PCI_DEVICE_ID_ARTOP_8060 0x8060 -#define PCI_DEVICE_ID_ARTOP_AEC67160 0x8080 -#define PCI_DEVICE_ID_ARTOP_AEC67160_2 0x8081 -#define PCI_DEVICE_ID_ARTOP_AEC67162 0x808a #define PCI_VENDOR_ID_ZEITNET 0x1193 #define PCI_DEVICE_ID_ZEITNET_1221 0x0001 #define PCI_DEVICE_ID_ZEITNET_1225 0x0002 -#define PCI_VENDOR_ID_OMEGA 0x119b -#define PCI_DEVICE_ID_OMEGA_82C092G 0x1221 #define PCI_VENDOR_ID_FUJITSU_ME 0x119e #define PCI_DEVICE_ID_FUJITSU_FS155 0x0001 @@ -1693,61 +1385,41 @@ #define PCI_SUBDEVICE_ID_KEYSPAN_SX2 0x5334 #define PCI_VENDOR_ID_MARVELL 0x11ab -#define PCI_DEVICE_ID_MARVELL_GT64011 0x4146 -#define PCI_DEVICE_ID_MARVELL_GT64111 0x4146 #define PCI_DEVICE_ID_MARVELL_GT64260 0x6430 #define PCI_DEVICE_ID_MARVELL_MV64360 0x6460 #define PCI_DEVICE_ID_MARVELL_MV64460 0x6480 #define PCI_DEVICE_ID_MARVELL_GT96100 0x9652 #define PCI_DEVICE_ID_MARVELL_GT96100A 0x9653 -#define PCI_VENDOR_ID_LITEON 0x11ad -#define PCI_DEVICE_ID_LITEON_LNE100TX 0x0002 #define PCI_VENDOR_ID_V3 0x11b0 #define PCI_DEVICE_ID_V3_V960 0x0001 -#define PCI_DEVICE_ID_V3_V350 0x0001 -#define PCI_DEVICE_ID_V3_V961 0x0002 #define PCI_DEVICE_ID_V3_V351 0x0002 -#define PCI_VENDOR_ID_NP 0x11bc -#define PCI_DEVICE_ID_NP_PCI_FDDI 0x0001 #define PCI_VENDOR_ID_ATT 0x11c1 -#define PCI_DEVICE_ID_ATT_L56XMF 0x0440 #define PCI_DEVICE_ID_ATT_VENUS_MODEM 0x480 -#define PCI_VENDOR_ID_NEC2 0x11c3 /* NEC (2nd) */ #define PCI_VENDOR_ID_SPECIALIX 0x11cb #define PCI_DEVICE_ID_SPECIALIX_IO8 0x2000 -#define PCI_DEVICE_ID_SPECIALIX_XIO 0x4000 #define PCI_DEVICE_ID_SPECIALIX_RIO 0x8000 #define PCI_SUBDEVICE_ID_SPECIALIX_SPEED4 0xa004 -#define PCI_VENDOR_ID_AURAVISION 0x11d1 -#define PCI_DEVICE_ID_AURAVISION_VXP524 0x01f7 #define PCI_VENDOR_ID_ANALOG_DEVICES 0x11d4 #define PCI_DEVICE_ID_AD1889JS 0x1889 -#define PCI_VENDOR_ID_IKON 0x11d5 -#define PCI_DEVICE_ID_IKON_10115 0x0115 -#define PCI_DEVICE_ID_IKON_10117 0x0117 -#define PCI_VENDOR_ID_SEGA 0x11db #define PCI_DEVICE_ID_SEGA_BBA 0x1234 #define PCI_VENDOR_ID_ZORAN 0x11de #define PCI_DEVICE_ID_ZORAN_36057 0x6057 #define PCI_DEVICE_ID_ZORAN_36120 0x6120 -#define PCI_VENDOR_ID_KINETIC 0x11f4 -#define PCI_DEVICE_ID_KINETIC_2915 0x2915 #define PCI_VENDOR_ID_COMPEX 0x11f6 #define PCI_DEVICE_ID_COMPEX_ENET100VG4 0x0112 -#define PCI_DEVICE_ID_COMPEX_RL2000 0x1401 #define PCI_VENDOR_ID_RP 0x11fe #define PCI_DEVICE_ID_RP32INTF 0x0001 @@ -1761,7 +1433,6 @@ #define PCI_DEVICE_ID_RP16SNI 0x0009 #define PCI_DEVICE_ID_RPP4 0x000A #define PCI_DEVICE_ID_RPP8 0x000B -#define PCI_DEVICE_ID_RP8M 0x000C #define PCI_DEVICE_ID_RP4M 0x000D #define PCI_DEVICE_ID_RP2_232 0x000E #define PCI_DEVICE_ID_RP2_422 0x000F @@ -1805,10 +1476,7 @@ #define PCI_DEVICE_ID_3DFX_VOODOO3 0x0005 #define PCI_DEVICE_ID_3DFX_VOODOO5 0x0009 -#define PCI_VENDOR_ID_SIGMADES 0x1236 -#define PCI_DEVICE_ID_SIGMADES_6425 0x6401 -#define PCI_VENDOR_ID_CCUBE 0x123f #define PCI_VENDOR_ID_AVM 0x1244 #define PCI_DEVICE_ID_AVM_B1 0x0700 @@ -1818,19 +1486,8 @@ #define PCI_DEVICE_ID_AVM_C2 0x1100 #define PCI_DEVICE_ID_AVM_T1 0x1200 -#define PCI_VENDOR_ID_DIPIX 0x1246 #define PCI_VENDOR_ID_STALLION 0x124d -#define PCI_DEVICE_ID_STALLION_ECHPCI832 0x0000 -#define PCI_DEVICE_ID_STALLION_ECHPCI864 0x0002 -#define PCI_DEVICE_ID_STALLION_EIOPCI 0x0003 - -#define PCI_VENDOR_ID_OPTIBASE 0x1255 -#define PCI_DEVICE_ID_OPTIBASE_FORGE 0x1110 -#define PCI_DEVICE_ID_OPTIBASE_FUSION 0x1210 -#define PCI_DEVICE_ID_OPTIBASE_VPLEX 0x2110 -#define PCI_DEVICE_ID_OPTIBASE_VPLEXCC 0x2120 -#define PCI_DEVICE_ID_OPTIBASE_VQUEST 0x2130 /* Allied Telesyn */ #define PCI_VENDOR_ID_AT 0x1259 @@ -1839,7 +1496,6 @@ #define PCI_VENDOR_ID_ESS 0x125d #define PCI_DEVICE_ID_ESS_ESS1968 0x1968 -#define PCI_DEVICE_ID_ESS_AUDIOPCI 0x1969 #define PCI_DEVICE_ID_ESS_ESS1978 0x1978 #define PCI_DEVICE_ID_ESS_ALLEGRO_1 0x1988 #define PCI_DEVICE_ID_ESS_ALLEGRO 0x1989 @@ -1852,11 +1508,7 @@ #define PCI_VENDOR_ID_SATSAGEM 0x1267 #define PCI_DEVICE_ID_SATSAGEM_NICCY 0x1016 -#define PCI_DEVICE_ID_SATSAGEM_PCR2101 0x5352 -#define PCI_DEVICE_ID_SATSAGEM_TELSATTURBO 0x5a4b -#define PCI_VENDOR_ID_HUGHES 0x1273 -#define PCI_DEVICE_ID_HUGHES_DIRECPC 0x0002 #define PCI_VENDOR_ID_ENSONIQ 0x1274 #define PCI_DEVICE_ID_ENSONIQ_CT5880 0x5880 @@ -1877,13 +1529,10 @@ #define PCI_DEVICE_ID_ITE_IT8330G_0 0xe886 /* formerly Platform Tech */ -#define PCI_VENDOR_ID_ESS_OLD 0x1285 #define PCI_DEVICE_ID_ESS_ESS0100 0x0100 #define PCI_VENDOR_ID_ALTEON 0x12ae -#define PCI_DEVICE_ID_ALTEON_ACENIC 0x0001 -#define PCI_VENDOR_ID_USR 0x12B9 #define PCI_SUBVENDOR_ID_CONNECT_TECH 0x12c4 #define PCI_SUBDEVICE_ID_CONNECT_TECH_BH8_232 0x0001 @@ -1898,8 +1547,6 @@ #define PCI_SUBDEVICE_ID_CONNECT_TECH_BH081101V1 0x000A #define PCI_SUBDEVICE_ID_CONNECT_TECH_BH041101V1 0x000B -#define PCI_VENDOR_ID_PICTUREL 0x12c5 -#define PCI_DEVICE_ID_PICTUREL_PCIVST 0x0081 #define PCI_VENDOR_ID_NVIDIA_SGS 0x12d2 #define PCI_DEVICE_ID_NVIDIA_SGS_RIVA128 0x0018 @@ -1921,8 +1568,6 @@ #define PCI_VENDOR_ID_ELECTRONICDESIGNGMBH 0x12f8 #define PCI_DEVICE_ID_LML_33R10 0x8a02 -#define PCI_VENDOR_ID_CBOARDS 0x1307 -#define PCI_DEVICE_ID_CBOARDS_DAS1602_16 0x0001 #define PCI_VENDOR_ID_SIIG 0x131f #define PCI_SUBVENDOR_ID_SIIG 0x131f @@ -1965,8 +1610,6 @@ #define PCI_DEVICE_ID_SIIG_2S1P_20x_850 0x2062 #define PCI_SUBDEVICE_ID_SIIG_QUARTET_SERIAL 0x2050 -#define PCI_VENDOR_ID_RADISYS 0x1331 -#define PCI_DEVICE_ID_RADISYS_ENP2611 0x0030 #define PCI_VENDOR_ID_DOMEX 0x134a #define PCI_DEVICE_ID_DOMEX_DMX3191D 0x0001 @@ -1974,8 +1617,6 @@ #define PCI_VENDOR_ID_QUATECH 0x135C #define PCI_DEVICE_ID_QUATECH_QSC100 0x0010 #define PCI_DEVICE_ID_QUATECH_DSC100 0x0020 -#define PCI_DEVICE_ID_QUATECH_DSC200 0x0030 -#define PCI_DEVICE_ID_QUATECH_QSC200 0x0040 #define PCI_DEVICE_ID_QUATECH_ESC100D 0x0050 #define PCI_DEVICE_ID_QUATECH_ESC100M 0x0060 @@ -1994,7 +1635,6 @@ #define PCI_SUBDEVICE_ID_HYPERCOPE_ERGO 0x0106 #define PCI_SUBDEVICE_ID_HYPERCOPE_METRO 0x0107 #define PCI_SUBDEVICE_ID_HYPERCOPE_CHAMP2 0x0108 -#define PCI_SUBDEVICE_ID_HYPERCOPE_PLEXUS 0x0109 #define PCI_VENDOR_ID_KAWASAKI 0x136b #define PCI_DEVICE_ID_MCHIP_KL5A72002 0xff01 @@ -2008,12 +1648,9 @@ #define PCI_DEVICE_ID_LMC_SSI 0x0005 #define PCI_DEVICE_ID_LMC_T1 0x0006 -#define PCI_VENDOR_ID_MARIAN 0x1382 -#define PCI_DEVICE_ID_MARIAN_PRODIF_PLUS 0x2048 #define PCI_VENDOR_ID_NETGEAR 0x1385 #define PCI_DEVICE_ID_NETGEAR_GA620 0x620a -#define PCI_DEVICE_ID_NETGEAR_GA622 0x622a #define PCI_VENDOR_ID_APPLICOM 0x1389 #define PCI_DEVICE_ID_APPLICOM_PCIGENERIC 0x0001 @@ -2036,9 +1673,6 @@ #define PCI_DEVICE_ID_MOXA_CP134U 0x1340 #define PCI_DEVICE_ID_MOXA_C168 0x1680 #define PCI_DEVICE_ID_MOXA_CP168U 0x1681 -#define PCI_DEVICE_ID_MOXA_CP204J 0x2040 -#define PCI_DEVICE_ID_MOXA_C218 0x2180 -#define PCI_DEVICE_ID_MOXA_C320 0x3200 #define PCI_VENDOR_ID_CCD 0x1397 #define PCI_DEVICE_ID_CCD_2BD0 0x2bd0 @@ -2059,9 +1693,7 @@ #define PCI_VENDOR_ID_MICROGATE 0x13c0 #define PCI_DEVICE_ID_MICROGATE_USC 0x0010 -#define PCI_DEVICE_ID_MICROGATE_SCC 0x0020 #define PCI_DEVICE_ID_MICROGATE_SCA 0x0030 -#define PCI_DEVICE_ID_MICROGATE_USC2 0x0210 #define PCI_VENDOR_ID_3WARE 0x13C1 #define PCI_DEVICE_ID_3WARE_1000 0x1000 @@ -2112,10 +1744,6 @@ #define PCI_VENDOR_ID_SAMSUNG 0x144d -#define PCI_VENDOR_ID_AIRONET 0x14b9 -#define PCI_DEVICE_ID_AIRONET_4800_1 0x0001 -#define PCI_DEVICE_ID_AIRONET_4800 0x4500 // values switched? see -#define PCI_DEVICE_ID_AIRONET_4500 0x4800 // drivers/net/aironet4500_card.c #define PCI_VENDOR_ID_TITAN 0x14D2 #define PCI_DEVICE_ID_TITAN_010L 0x8001 @@ -2134,8 +1762,6 @@ #define PCI_DEVICE_ID_PANACOM_QUADMODEM 0x0400 #define PCI_DEVICE_ID_PANACOM_DUALMODEM 0x0402 -#define PCI_VENDOR_ID_SIPACKETS 0x14d9 -#define PCI_DEVICE_ID_SP_HT 0x0010 #define PCI_VENDOR_ID_AFAVLAB 0x14db #define PCI_DEVICE_ID_AFAVLAB_P028 0x2180 @@ -2200,8 +1826,6 @@ #define PCI_VENDOR_ID_CHELSIO 0x1425 -#define PCI_VENDOR_ID_MIPS 0x153f -#define PCI_DEVICE_ID_SOC_IT 0x0001 #define PCI_VENDOR_ID_SYBA 0x1592 #define PCI_DEVICE_ID_SYBA_2P_EPP 0x0782 @@ -2221,15 +1845,7 @@ #define PCI_DEVICE_ID_MELLANOX_SINAI 0x6274 #define PCI_VENDOR_ID_PDC 0x15e9 -#define PCI_DEVICE_ID_PDC_1841 0x1841 -#define PCI_VENDOR_ID_MACROLINK 0x15ed -#define PCI_DEVICE_ID_MACROLINK_MCCS8 0x1000 -#define PCI_DEVICE_ID_MACROLINK_MCCS 0x1001 -#define PCI_DEVICE_ID_MACROLINK_MCCS8H 0x1002 -#define PCI_DEVICE_ID_MACROLINK_MCCSH 0x1003 -#define PCI_DEVICE_ID_MACROLINK_MCCR8 0x2000 -#define PCI_DEVICE_ID_MACROLINK_MCCR 0x2001 #define PCI_VENDOR_ID_FARSITE 0x1619 #define PCI_DEVICE_ID_FARSITE_T2P 0x0400 @@ -2247,7 +1863,6 @@ #define PCI_DEVICE_ID_REVOLUTION 0x0044 #define PCI_VENDOR_ID_LINKSYS 0x1737 -#define PCI_DEVICE_ID_LINKSYS_EG1032 0x1032 #define PCI_DEVICE_ID_LINKSYS_EG1064 0x1064 #define PCI_VENDOR_ID_ALTIMA 0x173b @@ -2262,7 +1877,6 @@ #define PCI_DEVICE_ID_HERC_WIN 0x5732 #define PCI_DEVICE_ID_HERC_UNI 0x5832 -#define PCI_VENDOR_ID_INFINICON 0x1820 #define PCI_VENDOR_ID_SITECOM 0x182d #define PCI_DEVICE_ID_SITECOM_DC105V2 0x3069 @@ -2272,8 +1886,6 @@ #define PCI_VENDOR_ID_TDI 0x192E #define PCI_DEVICE_ID_TDI_EHCI 0x0101 -#define PCI_VENDOR_ID_SYMPHONY 0x1c1c -#define PCI_DEVICE_ID_SYMPHONY_101 0x0001 #define PCI_VENDOR_ID_TEKRAM 0x1de1 #define PCI_DEVICE_ID_TEKRAM_DC290 0xdc29 @@ -2282,59 +1894,26 @@ #define PCI_DEVICE_ID_HINT_VXPROII_IDE 0x8013 #define PCI_VENDOR_ID_3DLABS 0x3d3d -#define PCI_DEVICE_ID_3DLABS_300SX 0x0001 -#define PCI_DEVICE_ID_3DLABS_500TX 0x0002 -#define PCI_DEVICE_ID_3DLABS_DELTA 0x0003 -#define PCI_DEVICE_ID_3DLABS_PERMEDIA 0x0004 -#define PCI_DEVICE_ID_3DLABS_MX 0x0006 #define PCI_DEVICE_ID_3DLABS_PERMEDIA2 0x0007 -#define PCI_DEVICE_ID_3DLABS_GAMMA 0x0008 #define PCI_DEVICE_ID_3DLABS_PERMEDIA2V 0x0009 -#define PCI_VENDOR_ID_AVANCE 0x4005 -#define PCI_DEVICE_ID_AVANCE_ALG2064 0x2064 -#define PCI_DEVICE_ID_AVANCE_2302 0x2302 #define PCI_VENDOR_ID_AKS 0x416c #define PCI_DEVICE_ID_AKS_ALADDINCARD 0x0100 -#define PCI_DEVICE_ID_AKS_CPC 0x0200 -#define PCI_VENDOR_ID_REDCREEK 0x4916 -#define PCI_DEVICE_ID_RC45 0x1960 -#define PCI_VENDOR_ID_NETVIN 0x4a14 -#define PCI_DEVICE_ID_NETVIN_NV5000SC 0x5000 #define PCI_VENDOR_ID_S3 0x5333 -#define PCI_DEVICE_ID_S3_PLATO_PXS 0x0551 -#define PCI_DEVICE_ID_S3_ViRGE 0x5631 #define PCI_DEVICE_ID_S3_TRIO 0x8811 -#define PCI_DEVICE_ID_S3_AURORA64VP 0x8812 -#define PCI_DEVICE_ID_S3_TRIO64UVP 0x8814 -#define PCI_DEVICE_ID_S3_ViRGE_VX 0x883d #define PCI_DEVICE_ID_S3_868 0x8880 -#define PCI_DEVICE_ID_S3_928 0x88b0 -#define PCI_DEVICE_ID_S3_864_1 0x88c0 -#define PCI_DEVICE_ID_S3_864_2 0x88c1 -#define PCI_DEVICE_ID_S3_964_1 0x88d0 -#define PCI_DEVICE_ID_S3_964_2 0x88d1 #define PCI_DEVICE_ID_S3_968 0x88f0 -#define PCI_DEVICE_ID_S3_TRIO64V2 0x8901 -#define PCI_DEVICE_ID_S3_PLATO_PXG 0x8902 -#define PCI_DEVICE_ID_S3_ViRGE_DXGX 0x8a01 -#define PCI_DEVICE_ID_S3_ViRGE_GX2 0x8a10 #define PCI_DEVICE_ID_S3_SAVAGE4 0x8a25 -#define PCI_DEVICE_ID_S3_ViRGE_MX 0x8c01 -#define PCI_DEVICE_ID_S3_ViRGE_MXP 0x8c02 -#define PCI_DEVICE_ID_S3_ViRGE_MXPMV 0x8c03 #define PCI_DEVICE_ID_S3_PROSAVAGE8 0x8d04 #define PCI_DEVICE_ID_S3_SONICVIBES 0xca00 #define PCI_VENDOR_ID_DUNORD 0x5544 #define PCI_DEVICE_ID_DUNORD_I3000 0x0001 -#define PCI_VENDOR_ID_GENROCO 0x5555 -#define PCI_DEVICE_ID_GENROCO_HFP832 0x0003 #define PCI_VENDOR_ID_DCI 0x6666 #define PCI_DEVICE_ID_DCI_PCCOM4 0x0001 @@ -2342,7 +1921,6 @@ #define PCI_VENDOR_ID_INTEL 0x8086 #define PCI_DEVICE_ID_INTEL_EESSC 0x0008 -#define PCI_DEVICE_ID_INTEL_21145 0x0039 #define PCI_DEVICE_ID_INTEL_PXHD_0 0x0320 #define PCI_DEVICE_ID_INTEL_PXHD_1 0x0321 #define PCI_DEVICE_ID_INTEL_PXH_0 0x0329 @@ -2351,30 +1929,17 @@ #define PCI_DEVICE_ID_INTEL_82375 0x0482 #define PCI_DEVICE_ID_INTEL_82424 0x0483 #define PCI_DEVICE_ID_INTEL_82378 0x0484 -#define PCI_DEVICE_ID_INTEL_82430 0x0486 -#define PCI_DEVICE_ID_INTEL_82434 0x04a3 #define PCI_DEVICE_ID_INTEL_I960 0x0960 #define PCI_DEVICE_ID_INTEL_I960RM 0x0962 -#define PCI_DEVICE_ID_INTEL_82562ET 0x1031 -#define PCI_DEVICE_ID_INTEL_82801CAM 0x1038 #define PCI_DEVICE_ID_INTEL_82815_MC 0x1130 -#define PCI_DEVICE_ID_INTEL_82815_AB 0x1131 #define PCI_DEVICE_ID_INTEL_82815_CGC 0x1132 -#define PCI_DEVICE_ID_INTEL_82559ER 0x1209 #define PCI_DEVICE_ID_INTEL_82092AA_0 0x1221 -#define PCI_DEVICE_ID_INTEL_82092AA_1 0x1222 -#define PCI_DEVICE_ID_INTEL_7116 0x1223 #define PCI_DEVICE_ID_INTEL_7505_0 0x2550 -#define PCI_DEVICE_ID_INTEL_7505_1 0x2552 #define PCI_DEVICE_ID_INTEL_7205_0 0x255d -#define PCI_DEVICE_ID_INTEL_82596 0x1226 -#define PCI_DEVICE_ID_INTEL_82865 0x1227 -#define PCI_DEVICE_ID_INTEL_82557 0x1229 #define PCI_DEVICE_ID_INTEL_82437 0x122d #define PCI_DEVICE_ID_INTEL_82371FB_0 0x122e #define PCI_DEVICE_ID_INTEL_82371FB_1 0x1230 #define PCI_DEVICE_ID_INTEL_82371MX 0x1234 -#define PCI_DEVICE_ID_INTEL_82437MX 0x1235 #define PCI_DEVICE_ID_INTEL_82441 0x1237 #define PCI_DEVICE_ID_INTEL_82380FB 0x124b #define PCI_DEVICE_ID_INTEL_82439 0x1250 @@ -2383,83 +1948,53 @@ #define PCI_DEVICE_ID_INTEL_82845_HB 0x1a30 #define PCI_DEVICE_ID_INTEL_82801AA_0 0x2410 #define PCI_DEVICE_ID_INTEL_82801AA_1 0x2411 -#define PCI_DEVICE_ID_INTEL_82801AA_2 0x2412 #define PCI_DEVICE_ID_INTEL_82801AA_3 0x2413 #define PCI_DEVICE_ID_INTEL_82801AA_5 0x2415 #define PCI_DEVICE_ID_INTEL_82801AA_6 0x2416 #define PCI_DEVICE_ID_INTEL_82801AA_8 0x2418 #define PCI_DEVICE_ID_INTEL_82801AB_0 0x2420 #define PCI_DEVICE_ID_INTEL_82801AB_1 0x2421 -#define PCI_DEVICE_ID_INTEL_82801AB_2 0x2422 #define PCI_DEVICE_ID_INTEL_82801AB_3 0x2423 #define PCI_DEVICE_ID_INTEL_82801AB_5 0x2425 #define PCI_DEVICE_ID_INTEL_82801AB_6 0x2426 #define PCI_DEVICE_ID_INTEL_82801AB_8 0x2428 #define PCI_DEVICE_ID_INTEL_82801BA_0 0x2440 -#define PCI_DEVICE_ID_INTEL_82801BA_1 0x2442 #define PCI_DEVICE_ID_INTEL_82801BA_2 0x2443 -#define PCI_DEVICE_ID_INTEL_82801BA_3 0x2444 #define PCI_DEVICE_ID_INTEL_82801BA_4 0x2445 -#define PCI_DEVICE_ID_INTEL_82801BA_5 0x2446 #define PCI_DEVICE_ID_INTEL_82801BA_6 0x2448 -#define PCI_DEVICE_ID_INTEL_82801BA_7 0x2449 #define PCI_DEVICE_ID_INTEL_82801BA_8 0x244a #define PCI_DEVICE_ID_INTEL_82801BA_9 0x244b #define PCI_DEVICE_ID_INTEL_82801BA_10 0x244c #define PCI_DEVICE_ID_INTEL_82801BA_11 0x244e #define PCI_DEVICE_ID_INTEL_82801E_0 0x2450 -#define PCI_DEVICE_ID_INTEL_82801E_2 0x2452 -#define PCI_DEVICE_ID_INTEL_82801E_3 0x2453 -#define PCI_DEVICE_ID_INTEL_82801E_9 0x2459 #define PCI_DEVICE_ID_INTEL_82801E_11 0x245b -#define PCI_DEVICE_ID_INTEL_82801E_13 0x245d -#define PCI_DEVICE_ID_INTEL_82801E_14 0x245e #define PCI_DEVICE_ID_INTEL_82801CA_0 0x2480 -#define PCI_DEVICE_ID_INTEL_82801CA_2 0x2482 #define PCI_DEVICE_ID_INTEL_82801CA_3 0x2483 -#define PCI_DEVICE_ID_INTEL_82801CA_4 0x2484 #define PCI_DEVICE_ID_INTEL_82801CA_5 0x2485 #define PCI_DEVICE_ID_INTEL_82801CA_6 0x2486 -#define PCI_DEVICE_ID_INTEL_82801CA_7 0x2487 #define PCI_DEVICE_ID_INTEL_82801CA_10 0x248a #define PCI_DEVICE_ID_INTEL_82801CA_11 0x248b #define PCI_DEVICE_ID_INTEL_82801CA_12 0x248c #define PCI_DEVICE_ID_INTEL_82801DB_0 0x24c0 #define PCI_DEVICE_ID_INTEL_82801DB_1 0x24c1 -#define PCI_DEVICE_ID_INTEL_82801DB_2 0x24c2 #define PCI_DEVICE_ID_INTEL_82801DB_3 0x24c3 -#define PCI_DEVICE_ID_INTEL_82801DB_4 0x24c4 #define PCI_DEVICE_ID_INTEL_82801DB_5 0x24c5 #define PCI_DEVICE_ID_INTEL_82801DB_6 0x24c6 -#define PCI_DEVICE_ID_INTEL_82801DB_7 0x24c7 #define PCI_DEVICE_ID_INTEL_82801DB_9 0x24c9 #define PCI_DEVICE_ID_INTEL_82801DB_10 0x24ca #define PCI_DEVICE_ID_INTEL_82801DB_11 0x24cb #define PCI_DEVICE_ID_INTEL_82801DB_12 0x24cc -#define PCI_DEVICE_ID_INTEL_82801DB_13 0x24cd #define PCI_DEVICE_ID_INTEL_82801EB_0 0x24d0 #define PCI_DEVICE_ID_INTEL_82801EB_1 0x24d1 -#define PCI_DEVICE_ID_INTEL_82801EB_2 0x24d2 #define PCI_DEVICE_ID_INTEL_82801EB_3 0x24d3 -#define PCI_DEVICE_ID_INTEL_82801EB_4 0x24d4 #define PCI_DEVICE_ID_INTEL_82801EB_5 0x24d5 #define PCI_DEVICE_ID_INTEL_82801EB_6 0x24d6 -#define PCI_DEVICE_ID_INTEL_82801EB_7 0x24d7 #define PCI_DEVICE_ID_INTEL_82801EB_11 0x24db -#define PCI_DEVICE_ID_INTEL_82801EB_13 0x24dd #define PCI_DEVICE_ID_INTEL_ESB_1 0x25a1 #define PCI_DEVICE_ID_INTEL_ESB_2 0x25a2 -#define PCI_DEVICE_ID_INTEL_ESB_3 0x25a3 -#define PCI_DEVICE_ID_INTEL_ESB_31 0x25b0 #define PCI_DEVICE_ID_INTEL_ESB_4 0x25a4 #define PCI_DEVICE_ID_INTEL_ESB_5 0x25a6 -#define PCI_DEVICE_ID_INTEL_ESB_6 0x25a7 -#define PCI_DEVICE_ID_INTEL_ESB_7 0x25a9 -#define PCI_DEVICE_ID_INTEL_ESB_8 0x25aa #define PCI_DEVICE_ID_INTEL_ESB_9 0x25ab -#define PCI_DEVICE_ID_INTEL_ESB_11 0x25ac -#define PCI_DEVICE_ID_INTEL_ESB_12 0x25ad -#define PCI_DEVICE_ID_INTEL_ESB_13 0x25ae #define PCI_DEVICE_ID_INTEL_82820_HB 0x2500 #define PCI_DEVICE_ID_INTEL_82820_UP_HB 0x2501 #define PCI_DEVICE_ID_INTEL_82850_HB 0x2530 @@ -2469,7 +2004,6 @@ #define PCI_DEVICE_ID_INTEL_82865_HB 0x2570 #define PCI_DEVICE_ID_INTEL_82865_IG 0x2572 #define PCI_DEVICE_ID_INTEL_82875_HB 0x2578 -#define PCI_DEVICE_ID_INTEL_82875_IG 0x257b #define PCI_DEVICE_ID_INTEL_82915G_HB 0x2580 #define PCI_DEVICE_ID_INTEL_82915G_IG 0x2582 #define PCI_DEVICE_ID_INTEL_82915GM_HB 0x2590 @@ -2479,80 +2013,23 @@ #define PCI_DEVICE_ID_INTEL_ICH6_0 0x2640 #define PCI_DEVICE_ID_INTEL_ICH6_1 0x2641 #define PCI_DEVICE_ID_INTEL_ICH6_2 0x2642 -#define PCI_DEVICE_ID_INTEL_ICH6_3 0x2651 -#define PCI_DEVICE_ID_INTEL_ICH6_4 0x2652 -#define PCI_DEVICE_ID_INTEL_ICH6_5 0x2653 -#define PCI_DEVICE_ID_INTEL_ICH6_6 0x2658 -#define PCI_DEVICE_ID_INTEL_ICH6_7 0x2659 -#define PCI_DEVICE_ID_INTEL_ICH6_8 0x265a -#define PCI_DEVICE_ID_INTEL_ICH6_9 0x265b -#define PCI_DEVICE_ID_INTEL_ICH6_10 0x265c -#define PCI_DEVICE_ID_INTEL_ICH6_11 0x2660 -#define PCI_DEVICE_ID_INTEL_ICH6_12 0x2662 -#define PCI_DEVICE_ID_INTEL_ICH6_13 0x2664 -#define PCI_DEVICE_ID_INTEL_ICH6_14 0x2666 -#define PCI_DEVICE_ID_INTEL_ICH6_15 0x2668 #define PCI_DEVICE_ID_INTEL_ICH6_16 0x266a #define PCI_DEVICE_ID_INTEL_ICH6_17 0x266d #define PCI_DEVICE_ID_INTEL_ICH6_18 0x266e #define PCI_DEVICE_ID_INTEL_ICH6_19 0x266f #define PCI_DEVICE_ID_INTEL_ESB2_0 0x2670 -#define PCI_DEVICE_ID_INTEL_ESB2_1 0x2680 -#define PCI_DEVICE_ID_INTEL_ESB2_2 0x2681 -#define PCI_DEVICE_ID_INTEL_ESB2_3 0x2682 -#define PCI_DEVICE_ID_INTEL_ESB2_4 0x2683 -#define PCI_DEVICE_ID_INTEL_ESB2_5 0x2688 -#define PCI_DEVICE_ID_INTEL_ESB2_6 0x2689 -#define PCI_DEVICE_ID_INTEL_ESB2_7 0x268a -#define PCI_DEVICE_ID_INTEL_ESB2_8 0x268b -#define PCI_DEVICE_ID_INTEL_ESB2_9 0x268c -#define PCI_DEVICE_ID_INTEL_ESB2_10 0x2690 -#define PCI_DEVICE_ID_INTEL_ESB2_11 0x2692 -#define PCI_DEVICE_ID_INTEL_ESB2_12 0x2694 -#define PCI_DEVICE_ID_INTEL_ESB2_13 0x2696 #define PCI_DEVICE_ID_INTEL_ESB2_14 0x2698 -#define PCI_DEVICE_ID_INTEL_ESB2_15 0x2699 -#define PCI_DEVICE_ID_INTEL_ESB2_16 0x269a #define PCI_DEVICE_ID_INTEL_ESB2_17 0x269b #define PCI_DEVICE_ID_INTEL_ESB2_18 0x269e #define PCI_DEVICE_ID_INTEL_ICH7_0 0x27b8 #define PCI_DEVICE_ID_INTEL_ICH7_1 0x27b9 -#define PCI_DEVICE_ID_INTEL_ICH7_2 0x27c0 -#define PCI_DEVICE_ID_INTEL_ICH7_3 0x27c1 #define PCI_DEVICE_ID_INTEL_ICH7_30 0x27b0 #define PCI_DEVICE_ID_INTEL_ICH7_31 0x27bd -#define PCI_DEVICE_ID_INTEL_ICH7_5 0x27c4 -#define PCI_DEVICE_ID_INTEL_ICH7_6 0x27c5 -#define PCI_DEVICE_ID_INTEL_ICH7_7 0x27c8 -#define PCI_DEVICE_ID_INTEL_ICH7_8 0x27c9 -#define PCI_DEVICE_ID_INTEL_ICH7_9 0x27ca -#define PCI_DEVICE_ID_INTEL_ICH7_10 0x27cb -#define PCI_DEVICE_ID_INTEL_ICH7_11 0x27cc -#define PCI_DEVICE_ID_INTEL_ICH7_12 0x27d0 -#define PCI_DEVICE_ID_INTEL_ICH7_13 0x27d2 -#define PCI_DEVICE_ID_INTEL_ICH7_14 0x27d4 -#define PCI_DEVICE_ID_INTEL_ICH7_15 0x27d6 -#define PCI_DEVICE_ID_INTEL_ICH7_16 0x27d8 #define PCI_DEVICE_ID_INTEL_ICH7_17 0x27da -#define PCI_DEVICE_ID_INTEL_ICH7_18 0x27dc #define PCI_DEVICE_ID_INTEL_ICH7_19 0x27dd #define PCI_DEVICE_ID_INTEL_ICH7_20 0x27de #define PCI_DEVICE_ID_INTEL_ICH7_21 0x27df -#define PCI_DEVICE_ID_INTEL_ICH7_22 0x27e0 -#define PCI_DEVICE_ID_INTEL_ICH7_23 0x27e2 #define PCI_DEVICE_ID_INTEL_82855PM_HB 0x3340 -#define PCI_DEVICE_ID_INTEL_ESB2_19 0x3500 -#define PCI_DEVICE_ID_INTEL_ESB2_20 0x3501 -#define PCI_DEVICE_ID_INTEL_ESB2_21 0x3504 -#define PCI_DEVICE_ID_INTEL_ESB2_22 0x3505 -#define PCI_DEVICE_ID_INTEL_ESB2_23 0x350c -#define PCI_DEVICE_ID_INTEL_ESB2_24 0x350d -#define PCI_DEVICE_ID_INTEL_ESB2_25 0x3510 -#define PCI_DEVICE_ID_INTEL_ESB2_26 0x3511 -#define PCI_DEVICE_ID_INTEL_ESB2_27 0x3514 -#define PCI_DEVICE_ID_INTEL_ESB2_28 0x3515 -#define PCI_DEVICE_ID_INTEL_ESB2_29 0x3518 -#define PCI_DEVICE_ID_INTEL_ESB2_30 0x3519 #define PCI_DEVICE_ID_INTEL_82830_HB 0x3575 #define PCI_DEVICE_ID_INTEL_82830_CGC 0x3577 #define PCI_DEVICE_ID_INTEL_82855GM_HB 0x3580 @@ -2566,7 +2043,6 @@ #define PCI_DEVICE_ID_INTEL_MCH_PC 0x3599 #define PCI_DEVICE_ID_INTEL_MCH_PC1 0x359a #define PCI_DEVICE_ID_INTEL_E7525_MCH 0x359e -#define PCI_DEVICE_ID_INTEL_80310 0x530d #define PCI_DEVICE_ID_INTEL_82371SB_0 0x7000 #define PCI_DEVICE_ID_INTEL_82371SB_1 0x7010 #define PCI_DEVICE_ID_INTEL_82371SB_2 0x7020 @@ -2591,22 +2067,15 @@ #define PCI_DEVICE_ID_INTEL_440MX_6 0x7196 #define PCI_DEVICE_ID_INTEL_82443MX_0 0x7198 #define PCI_DEVICE_ID_INTEL_82443MX_1 0x7199 -#define PCI_DEVICE_ID_INTEL_82443MX_2 0x719a #define PCI_DEVICE_ID_INTEL_82443MX_3 0x719b #define PCI_DEVICE_ID_INTEL_82443GX_0 0x71a0 -#define PCI_DEVICE_ID_INTEL_82443GX_1 0x71a1 #define PCI_DEVICE_ID_INTEL_82443GX_2 0x71a2 -#define PCI_DEVICE_ID_INTEL_82372FB_0 0x7600 #define PCI_DEVICE_ID_INTEL_82372FB_1 0x7601 -#define PCI_DEVICE_ID_INTEL_82372FB_2 0x7602 -#define PCI_DEVICE_ID_INTEL_82372FB_3 0x7603 #define PCI_DEVICE_ID_INTEL_82454GX 0x84c4 -#define PCI_DEVICE_ID_INTEL_82450GX 0x84c5 #define PCI_DEVICE_ID_INTEL_82451NX 0x84ca #define PCI_DEVICE_ID_INTEL_82454NX 0x84cb #define PCI_DEVICE_ID_INTEL_84460GX 0x84ea #define PCI_DEVICE_ID_INTEL_IXP4XX 0x8500 -#define PCI_DEVICE_ID_INTEL_IXP2400 0x9001 #define PCI_DEVICE_ID_INTEL_IXP2800 0x9004 #define PCI_DEVICE_ID_INTEL_S21152BB 0xb152 @@ -2619,7 +2088,6 @@ #define PCI_SUBDEVICE_ID_COMPUTONE_PG6 0x0003 #define PCI_VENDOR_ID_KTI 0x8e2e -#define PCI_DEVICE_ID_KTI_ET32P2 0x3000 #define PCI_VENDOR_ID_ADAPTEC 0x9004 #define PCI_DEVICE_ID_ADAPTEC_7810 0x1078 @@ -2627,7 +2095,6 @@ #define PCI_DEVICE_ID_ADAPTEC_38602 0x3860 #define PCI_DEVICE_ID_ADAPTEC_7850 0x5078 #define PCI_DEVICE_ID_ADAPTEC_7855 0x5578 -#define PCI_DEVICE_ID_ADAPTEC_5800 0x5800 #define PCI_DEVICE_ID_ADAPTEC_3860 0x6038 #define PCI_DEVICE_ID_ADAPTEC_1480A 0x6075 #define PCI_DEVICE_ID_ADAPTEC_7860 0x6078 @@ -2647,7 +2114,6 @@ #define PCI_DEVICE_ID_ADAPTEC_7886 0x8678 #define PCI_DEVICE_ID_ADAPTEC_7887 0x8778 #define PCI_DEVICE_ID_ADAPTEC_7888 0x8878 -#define PCI_DEVICE_ID_ADAPTEC_1030 0x8b78 #define PCI_VENDOR_ID_ADAPTEC2 0x9005 #define PCI_DEVICE_ID_ADAPTEC2_2940U2 0x0010 @@ -2667,8 +2133,6 @@ #define PCI_DEVICE_ID_ADAPTEC2_7899P 0x00cf #define PCI_DEVICE_ID_ADAPTEC2_SCAMP 0x0503 -#define PCI_VENDOR_ID_ATRONICS 0x907f -#define PCI_DEVICE_ID_ATRONICS_2015 0x2015 #define PCI_VENDOR_ID_HOLTEK 0x9412 #define PCI_DEVICE_ID_HOLTEK_6565 0x6565 @@ -2701,7 +2165,3 @@ #define PCI_DEVICE_ID_RME_DIGI32_PRO 0x9897 #define PCI_DEVICE_ID_RME_DIGI32_8 0x9898 -#define PCI_VENDOR_ID_ARK 0xedd8 -#define PCI_DEVICE_ID_ARK_STING 0xa091 -#define PCI_DEVICE_ID_ARK_STINGARK 0xa099 -#define PCI_DEVICE_ID_ARK_2000MT 0xa0a1 -- cgit v1.2.3 From d10211b278c70fbebb83fb9e4ee717ca13f69b2d Mon Sep 17 00:00:00 2001 From: Andrew Morton Date: Sat, 15 Oct 2005 22:02:19 -0700 Subject: [PATCH] PCI: fix edac drivers for radisys 82600 borkage I told you that the pci_ids.h cleanup was a bad idea ;) Cc: Alan Cox Signed-off-by: Andrew Morton Signed-off-by: Greg Kroah-Hartman --- include/linux/pci_ids.h | 1 + 1 file changed, 1 insertion(+) (limited to 'include/linux') diff --git a/include/linux/pci_ids.h b/include/linux/pci_ids.h index 11fed60a928d..611e48a4e679 100644 --- a/include/linux/pci_ids.h +++ b/include/linux/pci_ids.h @@ -1610,6 +1610,7 @@ #define PCI_DEVICE_ID_SIIG_2S1P_20x_850 0x2062 #define PCI_SUBDEVICE_ID_SIIG_QUARTET_SERIAL 0x2050 +#define PCI_VENDOR_ID_RADISYS 0x1331 #define PCI_VENDOR_ID_DOMEX 0x134a #define PCI_DEVICE_ID_DOMEX_DMX3191D 0x0001 -- cgit v1.2.3 From 2f028234f2c7f31dc0ff0784e20f14be11f7035c Mon Sep 17 00:00:00 2001 From: Grant Coady Date: Thu, 20 Oct 2005 09:55:32 +1000 Subject: [PATCH] pci_ids: cleanup comments pci_ids.h cleanup: convert // comment to /* comment */ Signed-off-by: Grant Coady Signed-off-by: Greg Kroah-Hartman --- include/linux/pci_ids.h | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) (limited to 'include/linux') diff --git a/include/linux/pci_ids.h b/include/linux/pci_ids.h index 611e48a4e679..7d300f7469e3 100644 --- a/include/linux/pci_ids.h +++ b/include/linux/pci_ids.h @@ -444,7 +444,7 @@ #define PCI_DEVICE_ID_IBM_ICOM_V2_ONE_PORT_RVX_ONE_PORT_MDM 0x0251 #define PCI_DEVICE_ID_IBM_ICOM_FOUR_PORT_MODEL 0x252 -#define PCI_VENDOR_ID_COMPEX2 0x101a // pci.ids says "AT&T GIS (NCR)" +#define PCI_VENDOR_ID_COMPEX2 0x101a /* pci.ids says "AT&T GIS (NCR)" */ #define PCI_DEVICE_ID_COMPEX2_100VG 0x0005 #define PCI_VENDOR_ID_WD 0x101c @@ -1158,10 +1158,10 @@ #define PCI_VENDOR_ID_INIT 0x1101 -#define PCI_VENDOR_ID_CREATIVE 0x1102 // duplicate: ECTIVA +#define PCI_VENDOR_ID_CREATIVE 0x1102 /* duplicate: ECTIVA */ #define PCI_DEVICE_ID_CREATIVE_EMU10K1 0x0002 -#define PCI_VENDOR_ID_ECTIVA 0x1102 // duplicate: CREATIVE +#define PCI_VENDOR_ID_ECTIVA 0x1102 /* duplicate: CREATIVE */ #define PCI_DEVICE_ID_ECTIVA_EV1938 0x8938 #define PCI_VENDOR_ID_TTI 0x1103 @@ -1171,7 +1171,7 @@ #define PCI_DEVICE_ID_TTI_HPT302 0x0006 #define PCI_DEVICE_ID_TTI_HPT371 0x0007 #define PCI_DEVICE_ID_TTI_HPT374 0x0008 -#define PCI_DEVICE_ID_TTI_HPT372N 0x0009 // apparently a 372N variant? +#define PCI_DEVICE_ID_TTI_HPT372N 0x0009 /* apparently a 372N variant? */ #define PCI_VENDOR_ID_VIA 0x1106 #define PCI_DEVICE_ID_VIA_8763_0 0x0198 -- cgit v1.2.3 From 094f1649577dfc7f2c7407a8380e05a506b31f7f Mon Sep 17 00:00:00 2001 From: Greg Kroah-Hartman Date: Mon, 20 Jun 2005 21:15:16 -0700 Subject: [PATCH] USB: add endpoint information to sysfs This patch adds endpoint information for both devices and interfaces to sysfs. Previously it was only possible to get the endpoint information from usbfs, and never possible to get any information on endpoint 0. Signed-off-by: Greg Kroah-Hartman drivers/usb/core/sysfs.c | 195 ++++++++++++++++++++++++++++++++++++++++++++++- include/linux/usb.h | 4 2 files changed, 197 insertions(+), 2 deletions(-) --- drivers/usb/core/sysfs.c | 195 ++++++++++++++++++++++++++++++++++++++++++++++- include/linux/usb.h | 4 + 2 files changed, 197 insertions(+), 2 deletions(-) (limited to 'include/linux') diff --git a/drivers/usb/core/sysfs.c b/drivers/usb/core/sysfs.c index 00297f113849..eae413bf8c2a 100644 --- a/drivers/usb/core/sysfs.c +++ b/drivers/usb/core/sysfs.c @@ -22,6 +22,174 @@ #include "usb.h" +/* endpoint stuff */ +struct endpoint_attribute { + struct device_attribute dev_attr; + struct usb_endpoint_descriptor *endpoint; + struct usb_device *udev; +}; +#define to_endpoint_attr(_dev_attr) \ + container_of(_dev_attr, struct endpoint_attribute, dev_attr) + +#define usb_ep_attr(field, format_string) \ +static ssize_t show_ep_##field(struct device *dev, struct device_attribute *attr, \ + char *buf) \ +{ \ + struct endpoint_attribute *endpoint_attr = to_endpoint_attr(attr); \ + \ + return sprintf(buf, format_string, endpoint_attr->endpoint->field); \ +} +usb_ep_attr(bLength, "%02x\n") +usb_ep_attr(bDescriptorType, "%02x\n") +usb_ep_attr(bEndpointAddress, "%02x\n") +usb_ep_attr(bmAttributes, "%02x\n") +usb_ep_attr(bInterval, "%02x\n") + +static ssize_t show_ep_wMaxPacketSize(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct endpoint_attribute *endpoint_attr = to_endpoint_attr(attr); + + return sprintf(buf, "%04x\n", + le16_to_cpu(endpoint_attr->endpoint->wMaxPacketSize) & 0x07ff); +} + +static ssize_t show_ep_type(struct device *dev, struct device_attribute *attr, char *buf) +{ + struct endpoint_attribute *endpoint_attr = to_endpoint_attr(attr); + char *type = "unknown"; + + switch (endpoint_attr->endpoint->bmAttributes & USB_ENDPOINT_XFERTYPE_MASK) { + case USB_ENDPOINT_XFER_CONTROL: + type = "Control"; + break; + case USB_ENDPOINT_XFER_ISOC: + type = "Isoc"; + break; + case USB_ENDPOINT_XFER_BULK: + type = "Bulk"; + break; + case USB_ENDPOINT_XFER_INT: + type = "Interrupt"; + break; + } + return sprintf(buf, "%s\n", type); +} + +static ssize_t show_ep_interval(struct device *dev, struct device_attribute *attr, char *buf) +{ + struct endpoint_attribute *endpoint_attr = to_endpoint_attr(attr); + struct usb_device *udev = endpoint_attr->udev; + struct usb_endpoint_descriptor *endpoint = endpoint_attr->endpoint; + char unit; + unsigned interval = 0; + unsigned in; + + in = (endpoint->bEndpointAddress & USB_DIR_IN); + + switch (endpoint->bmAttributes & USB_ENDPOINT_XFERTYPE_MASK) { + case USB_ENDPOINT_XFER_CONTROL: + if (udev->speed == USB_SPEED_HIGH) /* uframes per NAK */ + interval = endpoint->bInterval; + break; + case USB_ENDPOINT_XFER_ISOC: + interval = 1 << (endpoint->bInterval - 1); + break; + case USB_ENDPOINT_XFER_BULK: + if (udev->speed == USB_SPEED_HIGH && !in) /* uframes per NAK */ + interval = endpoint->bInterval; + break; + case USB_ENDPOINT_XFER_INT: + if (udev->speed == USB_SPEED_HIGH) { + interval = 1 << (endpoint->bInterval - 1); + } else + interval = endpoint->bInterval; + break; + } + interval *= (udev->speed == USB_SPEED_HIGH) ? 125 : 1000; + if (interval % 1000) + unit = 'u'; + else { + unit = 'm'; + interval /= 1000; + } + + return sprintf(buf, "%d%cs\n", interval, unit); +} + +static ssize_t show_ep_direction(struct device *dev, struct device_attribute *attr, char *buf) +{ + struct endpoint_attribute *endpoint_attr = to_endpoint_attr(attr); + char *direction; + + if ((endpoint_attr->endpoint->bmAttributes & USB_ENDPOINT_XFERTYPE_MASK) == + USB_ENDPOINT_XFER_CONTROL) + direction = "both"; + else if (endpoint_attr->endpoint->bEndpointAddress & USB_DIR_IN) + direction = "in"; + else + direction = "out"; + return sprintf(buf, "%s\n", direction); +} + +static struct endpoint_attribute *create_ep_attr(struct usb_endpoint_descriptor *endpoint, + struct usb_device *udev, char *name, + ssize_t (*show)(struct device *dev, struct device_attribute *attr, char *buf)) +{ + struct endpoint_attribute *ep_attr; + + ep_attr = kzalloc(sizeof(*ep_attr), GFP_KERNEL); + if (ep_attr) { + ep_attr->endpoint = endpoint; + ep_attr->udev = udev; + ep_attr->dev_attr.attr.name = name; + ep_attr->dev_attr.attr.mode = 0444; + ep_attr->dev_attr.attr.owner = THIS_MODULE; + ep_attr->dev_attr.show = show; + } + return ep_attr; +} + +static void usb_create_ep_files(struct kobject *kobj, struct usb_host_endpoint *endpoint, struct usb_device *udev) +{ + struct usb_endpoint_descriptor *ep; + + ep = &endpoint->desc; + + endpoint->attrs = kzalloc(sizeof(struct attribute *) * 10, GFP_KERNEL); + endpoint->attrs[0] = &(create_ep_attr(ep, udev, "direction", show_ep_direction)->dev_attr.attr); + endpoint->attrs[1] = &(create_ep_attr(ep, udev, "type", show_ep_type)->dev_attr.attr); + endpoint->attrs[2] = &(create_ep_attr(ep, udev, "bLength", show_ep_bLength)->dev_attr.attr); + endpoint->attrs[3] = &(create_ep_attr(ep, udev, "bDescriptorType", show_ep_bDescriptorType)->dev_attr.attr); + endpoint->attrs[4] = &(create_ep_attr(ep, udev, "bEndpointAddress", show_ep_bEndpointAddress)->dev_attr.attr); + endpoint->attrs[5] = &(create_ep_attr(ep, udev, "bmAttributes", show_ep_bmAttributes)->dev_attr.attr); + endpoint->attrs[6] = &(create_ep_attr(ep, udev, "wMaxPacketSize", show_ep_wMaxPacketSize)->dev_attr.attr); + endpoint->attrs[7] = &(create_ep_attr(ep, udev, "bInterval", show_ep_bInterval)->dev_attr.attr); + endpoint->attrs[8] = &(create_ep_attr(ep, udev, "interval", show_ep_interval)->dev_attr.attr); + endpoint->attrs[9] = NULL; + endpoint->num_attrs = 9; + + endpoint->attr_group = kzalloc(sizeof(*endpoint->attr_group), GFP_KERNEL); + endpoint->attr_name = kzalloc(10, GFP_KERNEL); + sprintf(endpoint->attr_name, "ep_%02x", endpoint->desc.bEndpointAddress); + + endpoint->attr_group->attrs = endpoint->attrs; + endpoint->attr_group->name = endpoint->attr_name; + sysfs_create_group(kobj, endpoint->attr_group); +} + +static void usb_remove_ep_files(struct kobject *kobj, struct usb_host_endpoint *endpoint) +{ + int i; + + sysfs_remove_group(kobj, endpoint->attr_group); + kfree(endpoint->attr_group); + kfree(endpoint->attr_name); + for (i = 0; i < endpoint->num_attrs; ++i) + kfree(endpoint->attrs[i]); + kfree(endpoint->attrs); +} + /* Active configuration fields */ #define usb_actconfig_show(field, multiplier, format_string) \ static ssize_t show_##field (struct device *dev, struct device_attribute *attr, char *buf) \ @@ -236,12 +404,14 @@ void usb_create_sysfs_dev_files (struct usb_device *udev) if (udev->serial) device_create_file (dev, &dev_attr_serial); device_create_file (dev, &dev_attr_configuration); + usb_create_ep_files(&dev->kobj, &udev->ep0, udev); } void usb_remove_sysfs_dev_files (struct usb_device *udev) { struct device *dev = &udev->dev; + usb_remove_ep_files(&dev->kobj, &udev->ep0); sysfs_remove_group(&dev->kobj, &dev_attr_grp); if (udev->descriptor.iManufacturer) @@ -333,20 +503,41 @@ static struct attribute_group intf_attr_grp = { .attrs = intf_attrs, }; +static void usb_create_intf_ep_files(struct usb_interface *intf) +{ + struct usb_host_interface *iface_desc; + int i; + + iface_desc = intf->cur_altsetting; + for (i = 0; i < iface_desc->desc.bNumEndpoints; ++i) + usb_create_ep_files(&intf->dev.kobj, &iface_desc->endpoint[i], + interface_to_usbdev(intf)); +} + +static void usb_remove_intf_ep_files(struct usb_interface *intf) +{ + struct usb_host_interface *iface_desc; + int i; + + iface_desc = intf->cur_altsetting; + for (i = 0; i < iface_desc->desc.bNumEndpoints; ++i) + usb_remove_ep_files(&intf->dev.kobj, &iface_desc->endpoint[i]); +} + void usb_create_sysfs_intf_files (struct usb_interface *intf) { sysfs_create_group(&intf->dev.kobj, &intf_attr_grp); if (intf->cur_altsetting->string) device_create_file(&intf->dev, &dev_attr_interface); - + usb_create_intf_ep_files(intf); } void usb_remove_sysfs_intf_files (struct usb_interface *intf) { + usb_remove_intf_ep_files(intf); sysfs_remove_group(&intf->dev.kobj, &intf_attr_grp); if (intf->cur_altsetting->string) device_remove_file(&intf->dev, &dev_attr_interface); - } diff --git a/include/linux/usb.h b/include/linux/usb.h index 8f731e8f2821..4512210e97e7 100644 --- a/include/linux/usb.h +++ b/include/linux/usb.h @@ -57,6 +57,10 @@ struct usb_host_endpoint { struct usb_endpoint_descriptor desc; struct list_head urb_list; void *hcpriv; + char *attr_name; + struct attribute_group *attr_group; + struct attribute **attrs; + int num_attrs; unsigned char *extra; /* Extra descriptors */ int extralen; -- cgit v1.2.3 From d6e5bcf4a7ebbe258124a931f1449338340a99b5 Mon Sep 17 00:00:00 2001 From: Greg Kroah-Hartman Date: Mon, 20 Jun 2005 21:15:16 -0700 Subject: [PATCH] devfs: Remove the mode field from usb_class_driver as it's no longer needed Also fixes all drivers that set this field, and removes some other devfs specfic USB logic. Signed-off-by: Greg Kroah-Hartman drivers/usb/class/usblp.c | 3 +-- drivers/usb/core/file.c | 19 ++++--------------- drivers/usb/image/mdc800.c | 3 +-- drivers/usb/input/aiptek.c | 2 +- drivers/usb/input/hiddev.c | 3 +-- drivers/usb/media/dabusb.c | 3 +-- drivers/usb/misc/auerswald.c | 3 +-- drivers/usb/misc/idmouse.c | 5 ++--- drivers/usb/misc/legousbtower.c | 5 ++--- drivers/usb/misc/rio500.c | 3 +-- drivers/usb/misc/sisusbvga/sisusb.c | 5 ----- drivers/usb/misc/usblcd.c | 9 ++++----- drivers/usb/usb-skeleton.c | 3 +-- include/linux/usb.h | 7 ++----- 14 files changed, 22 insertions(+), 51 deletions(-) --- drivers/usb/class/usblp.c | 3 +-- drivers/usb/core/file.c | 19 ++++--------------- drivers/usb/image/mdc800.c | 3 +-- drivers/usb/input/aiptek.c | 2 +- drivers/usb/input/hiddev.c | 3 +-- drivers/usb/media/dabusb.c | 3 +-- drivers/usb/misc/auerswald.c | 3 +-- drivers/usb/misc/idmouse.c | 5 ++--- drivers/usb/misc/legousbtower.c | 5 ++--- drivers/usb/misc/rio500.c | 3 +-- drivers/usb/misc/sisusbvga/sisusb.c | 5 ----- drivers/usb/misc/usblcd.c | 9 ++++----- drivers/usb/usb-skeleton.c | 3 +-- include/linux/usb.h | 7 ++----- 14 files changed, 22 insertions(+), 51 deletions(-) (limited to 'include/linux') diff --git a/drivers/usb/class/usblp.c b/drivers/usb/class/usblp.c index e195709c9c7f..357e75335f17 100644 --- a/drivers/usb/class/usblp.c +++ b/drivers/usb/class/usblp.c @@ -844,9 +844,8 @@ static struct file_operations usblp_fops = { }; static struct usb_class_driver usblp_class = { - .name = "usb/lp%d", + .name = "lp%d", .fops = &usblp_fops, - .mode = S_IFCHR | S_IRUSR | S_IWUSR | S_IRGRP | S_IWGRP, .minor_base = USBLP_MINOR_BASE, }; diff --git a/drivers/usb/core/file.c b/drivers/usb/core/file.c index 78cb4be9529f..e695308095ae 100644 --- a/drivers/usb/core/file.c +++ b/drivers/usb/core/file.c @@ -17,7 +17,6 @@ #include #include -#include #include #include @@ -88,8 +87,6 @@ int usb_major_init(void) goto out; } - devfs_mk_dir("usb"); - out: return error; } @@ -97,7 +94,6 @@ out: void usb_major_cleanup(void) { class_destroy(usb_class); - devfs_remove("usb"); unregister_chrdev(USB_MAJOR, "usb"); } @@ -112,8 +108,7 @@ void usb_major_cleanup(void) * enabled, the minor number will be based on the next available free minor, * starting at the class_driver->minor_base. * - * This function also creates the devfs file for the usb device, if devfs - * is enabled, and creates a usb class device in the sysfs tree. + * This function also creates a usb class device in the sysfs tree. * * usb_deregister_dev() must be called when the driver is done with * the minor numbers given out by this function. @@ -162,11 +157,8 @@ int usb_register_dev(struct usb_interface *intf, intf->minor = minor; - /* handle the devfs registration */ - snprintf(name, BUS_ID_SIZE, class_driver->name, minor - minor_base); - devfs_mk_cdev(MKDEV(USB_MAJOR, minor), class_driver->mode, name); - /* create a usb class device for this usb interface */ + snprintf(name, BUS_ID_SIZE, class_driver->name, minor - minor_base); temp = strrchr(name, '/'); if (temp && (temp[1] != 0x00)) ++temp; @@ -179,7 +171,6 @@ int usb_register_dev(struct usb_interface *intf, spin_lock (&minor_lock); usb_minors[intf->minor] = NULL; spin_unlock (&minor_lock); - devfs_remove (name); retval = PTR_ERR(intf->class_dev); } exit: @@ -197,9 +188,8 @@ EXPORT_SYMBOL(usb_register_dev); * call to usb_register_dev() (usually when the device is disconnected * from the system.) * - * This function also cleans up the devfs file for the usb device, if devfs - * is enabled, and removes the usb class device from the sysfs tree. - * + * This function also removes the usb class device from the sysfs tree. + * * This should be called by all drivers that use the USB major number. */ void usb_deregister_dev(struct usb_interface *intf, @@ -222,7 +212,6 @@ void usb_deregister_dev(struct usb_interface *intf, spin_unlock (&minor_lock); snprintf(name, BUS_ID_SIZE, class_driver->name, intf->minor - minor_base); - devfs_remove (name); class_device_destroy(usb_class, MKDEV(USB_MAJOR, intf->minor)); intf->class_dev = NULL; intf->minor = -1; diff --git a/drivers/usb/image/mdc800.c b/drivers/usb/image/mdc800.c index a330a4b50e16..e4ec133ee74a 100644 --- a/drivers/usb/image/mdc800.c +++ b/drivers/usb/image/mdc800.c @@ -425,9 +425,8 @@ static void mdc800_usb_download_notify (struct urb *urb, struct pt_regs *res) static struct usb_driver mdc800_usb_driver; static struct file_operations mdc800_device_ops; static struct usb_class_driver mdc800_class = { - .name = "usb/mdc800%d", + .name = "mdc800%d", .fops = &mdc800_device_ops, - .mode = S_IFCHR | S_IRUSR | S_IWUSR | S_IRGRP | S_IWGRP, .minor_base = MDC800_DEVICE_MINOR_BASE, }; diff --git a/drivers/usb/input/aiptek.c b/drivers/usb/input/aiptek.c index 1c5205321d83..1c3b472a3bca 100644 --- a/drivers/usb/input/aiptek.c +++ b/drivers/usb/input/aiptek.c @@ -2154,7 +2154,7 @@ aiptek_probe(struct usb_interface *intf, const struct usb_device_id *id) * input_handles associated with this input device. * What identifies an evdev input_handler is that it begins * with 'event', continues with a digit, and that in turn - * is mapped to /{devfs}/input/eventN. + * is mapped to input/eventN. */ list_for_each_safe(node, next, &inputdev->h_list) { inputhandle = to_handle(node); diff --git a/drivers/usb/input/hiddev.c b/drivers/usb/input/hiddev.c index d32427818af7..440377c7a0da 100644 --- a/drivers/usb/input/hiddev.c +++ b/drivers/usb/input/hiddev.c @@ -732,9 +732,8 @@ static struct file_operations hiddev_fops = { }; static struct usb_class_driver hiddev_class = { - .name = "usb/hid/hiddev%d", + .name = "hiddev%d", .fops = &hiddev_fops, - .mode = S_IFCHR | S_IRUGO | S_IWUSR, .minor_base = HIDDEV_MINOR_BASE, }; diff --git a/drivers/usb/media/dabusb.c b/drivers/usb/media/dabusb.c index 6ca2fae99d2d..27b23c55bbc7 100644 --- a/drivers/usb/media/dabusb.c +++ b/drivers/usb/media/dabusb.c @@ -707,9 +707,8 @@ static struct file_operations dabusb_fops = }; static struct usb_class_driver dabusb_class = { - .name = "usb/dabusb%d", + .name = "dabusb%d", .fops = &dabusb_fops, - .mode = S_IFCHR | S_IRUSR | S_IWUSR | S_IRGRP | S_IWGRP, .minor_base = DABUSB_MINOR, }; diff --git a/drivers/usb/misc/auerswald.c b/drivers/usb/misc/auerswald.c index ae4681f9f0ea..5f33f7c64885 100644 --- a/drivers/usb/misc/auerswald.c +++ b/drivers/usb/misc/auerswald.c @@ -1873,9 +1873,8 @@ static struct file_operations auerswald_fops = }; static struct usb_class_driver auerswald_class = { - .name = "usb/auer%d", + .name = "auer%d", .fops = &auerswald_fops, - .mode = S_IFCHR | S_IRUGO | S_IWUGO, .minor_base = AUER_MINOR_BASE, }; diff --git a/drivers/usb/misc/idmouse.c b/drivers/usb/misc/idmouse.c index 733acc213726..3944a55ed74c 100644 --- a/drivers/usb/misc/idmouse.c +++ b/drivers/usb/misc/idmouse.c @@ -105,11 +105,10 @@ static struct file_operations idmouse_fops = { .release = idmouse_release, }; -/* class driver information for devfs */ +/* class driver information */ static struct usb_class_driver idmouse_class = { - .name = "usb/idmouse%d", + .name = "idmouse%d", .fops = &idmouse_fops, - .mode = S_IFCHR | S_IRUSR | S_IRGRP | S_IROTH, /* filemode (char, 444) */ .minor_base = USB_IDMOUSE_MINOR_BASE, }; diff --git a/drivers/usb/misc/legousbtower.c b/drivers/usb/misc/legousbtower.c index 7d06105763d4..2703e205bc8f 100644 --- a/drivers/usb/misc/legousbtower.c +++ b/drivers/usb/misc/legousbtower.c @@ -271,12 +271,11 @@ static struct file_operations tower_fops = { /* * usb class driver info in order to get a minor number from the usb core, - * and to have the device registered with devfs and the driver core + * and to have the device registered with the driver core */ static struct usb_class_driver tower_class = { - .name = "usb/legousbtower%d", + .name = "legousbtower%d", .fops = &tower_fops, - .mode = S_IFCHR | S_IRUSR | S_IWUSR | S_IRGRP | S_IWGRP | S_IROTH, .minor_base = LEGO_USB_TOWER_MINOR_BASE, }; diff --git a/drivers/usb/misc/rio500.c b/drivers/usb/misc/rio500.c index 26f77e29c7a6..7d02d8ec6b1a 100644 --- a/drivers/usb/misc/rio500.c +++ b/drivers/usb/misc/rio500.c @@ -443,9 +443,8 @@ file_operations usb_rio_fops = { }; static struct usb_class_driver usb_rio_class = { - .name = "usb/rio500%d", + .name = "rio500%d", .fops = &usb_rio_fops, - .mode = S_IFCHR | S_IRUSR | S_IWUSR | S_IRGRP | S_IWGRP, .minor_base = RIO_MINOR, }; diff --git a/drivers/usb/misc/sisusbvga/sisusb.c b/drivers/usb/misc/sisusbvga/sisusb.c index 39db3155723a..222ae9cd4c7c 100644 --- a/drivers/usb/misc/sisusbvga/sisusb.c +++ b/drivers/usb/misc/sisusbvga/sisusb.c @@ -3239,12 +3239,7 @@ static struct file_operations usb_sisusb_fops = { }; static struct usb_class_driver usb_sisusb_class = { -#if LINUX_VERSION_CODE <= KERNEL_VERSION(2,6,13) - .name = "usb/sisusbvga%d", - .mode = S_IFCHR | S_IRUSR | S_IWUSR | S_IRGRP | S_IWGRP, -#else .name = "sisusbvga%d", -#endif .fops = &usb_sisusb_fops, .minor_base = SISUSB_MINOR }; diff --git a/drivers/usb/misc/usblcd.c b/drivers/usb/misc/usblcd.c index 096ab3029676..85f3725334b0 100644 --- a/drivers/usb/misc/usblcd.c +++ b/drivers/usb/misc/usblcd.c @@ -251,13 +251,12 @@ static struct file_operations lcd_fops = { }; /* - * * usb class driver info in order to get a minor number from the usb core, - * * and to have the device registered with devfs and the driver core - * */ + * usb class driver info in order to get a minor number from the usb core, + * and to have the device registered with the driver core + */ static struct usb_class_driver lcd_class = { - .name = "usb/lcd%d", + .name = "lcd%d", .fops = &lcd_fops, - .mode = S_IFCHR | S_IRUSR | S_IWUSR | S_IRGRP | S_IWGRP | S_IROTH, .minor_base = USBLCD_MINOR, }; diff --git a/drivers/usb/usb-skeleton.c b/drivers/usb/usb-skeleton.c index 353f24d45bc1..6c3a53f8f26c 100644 --- a/drivers/usb/usb-skeleton.c +++ b/drivers/usb/usb-skeleton.c @@ -223,9 +223,8 @@ static struct file_operations skel_fops = { * and to have the device registered with devfs and the driver core */ static struct usb_class_driver skel_class = { - .name = "usb/skel%d", + .name = "skel%d", .fops = &skel_fops, - .mode = S_IFCHR | S_IRUSR | S_IWUSR | S_IRGRP | S_IWGRP | S_IROTH, .minor_base = USB_SKEL_MINOR_BASE, }; diff --git a/include/linux/usb.h b/include/linux/usb.h index 4512210e97e7..04502e183dd1 100644 --- a/include/linux/usb.h +++ b/include/linux/usb.h @@ -576,10 +576,8 @@ extern struct bus_type usb_bus_type; /** * struct usb_class_driver - identifies a USB driver that wants to use the USB major number - * @name: devfs name for this driver. Will also be used by the driver - * class code to create a usb class device. + * @name: the usb class device name for this driver. Will show up in sysfs. * @fops: pointer to the struct file_operations of this driver. - * @mode: the mode for the devfs file to be created for this driver. * @minor_base: the start of the minor range for this driver. * * This structure is used for the usb_register_dev() and @@ -589,8 +587,7 @@ extern struct bus_type usb_bus_type; struct usb_class_driver { char *name; struct file_operations *fops; - mode_t mode; - int minor_base; + int minor_base; }; /* -- cgit v1.2.3 From 390a8c345e6415cbf811232feedac70b56c9fc8d Mon Sep 17 00:00:00 2001 From: David Brownell Date: Tue, 13 Sep 2005 19:57:27 -0700 Subject: [PATCH] remove usb_suspend_device() parameter This patch removes the extra usb_suspend_device() parameter. The original reason to pass that parameter was so that this routine could suspend any active children. A previous patch removed that functionality ... leaving no reason to pass the parameter. A close analogy is pci_set_power_state, which doesn't need a pm_message_t either. On the internal code path that comes through the driver model, the parameter is now used to distinguish cases where USB devices need to "freeze" but not suspend. It also checks for an error case that's accessible through sysfs: attempting to suspend a device before its interfaces (or for hubs, ports). Signed-off-by: David Brownell Signed-off-by: Greg Kroah-Hartman drivers/usb/core/hub.c | 34 +++++++++++++++++++++------------- drivers/usb/core/usb.c | 23 +++++++++++++++++++++-- drivers/usb/host/ehci-hcd.c | 2 +- drivers/usb/host/isp116x-hcd.c | 2 +- drivers/usb/host/ohci-pci.c | 2 +- include/linux/usb.h | 2 +- 6 files changed, 46 insertions(+), 19 deletions(-) --- drivers/usb/core/hub.c | 34 +++++++++++++++++++++------------- drivers/usb/core/usb.c | 23 +++++++++++++++++++++-- drivers/usb/host/ehci-hcd.c | 2 +- drivers/usb/host/isp116x-hcd.c | 2 +- drivers/usb/host/ohci-pci.c | 2 +- include/linux/usb.h | 2 +- 6 files changed, 46 insertions(+), 19 deletions(-) (limited to 'include/linux') diff --git a/drivers/usb/core/hub.c b/drivers/usb/core/hub.c index d3337d9c31dc..33127b828d60 100644 --- a/drivers/usb/core/hub.c +++ b/drivers/usb/core/hub.c @@ -1323,11 +1323,9 @@ int usb_new_device(struct usb_device *udev) * (Includes HNP test device.) */ if (udev->bus->b_hnp_enable || udev->bus->is_b_host) { - static int __usb_suspend_device (struct usb_device *, - int port1, pm_message_t state); - err = __usb_suspend_device(udev, - udev->bus->otg_port, - PMSG_SUSPEND); + static int __usb_suspend_device(struct usb_device *, + int port1); + err = __usb_suspend_device(udev, udev->bus->otg_port); if (err < 0) dev_dbg(&udev->dev, "HNP fail, %d\n", err); } @@ -1517,7 +1515,7 @@ static void hub_port_logical_disconnect(struct usb_hub *hub, int port1) /* FIXME let caller ask to power down the port: * - some devices won't enumerate without a VBUS power cycle * - SRP saves power that way - * - usb_suspend_device(dev, PMSG_SUSPEND) + * - ... new call, TBD ... * That's easy if this hub can switch power per-port, and * khubd reactivates the port later (timer, SRP, etc). * Powerdown must be optional, because of reset/DFU. @@ -1599,9 +1597,12 @@ static int hub_port_suspend(struct usb_hub *hub, int port1, * Other than re-initializing the hub (plug/unplug, except for root hubs), * Linux (2.6) currently has NO mechanisms to initiate that: no khubd * timer, no SRP, no requests through sysfs. + * + * If CONFIG_USB_SUSPEND isn't enabled, devices only really suspend when + * the root hub for their bus goes into global suspend ... so we don't + * (falsely) update the device power state to say it suspended. */ -static int __usb_suspend_device (struct usb_device *udev, int port1, - pm_message_t state) +static int __usb_suspend_device (struct usb_device *udev, int port1) { int status; @@ -1648,14 +1649,13 @@ static int __usb_suspend_device (struct usb_device *udev, int port1, udev); if (status == 0) - udev->dev.power.power_state = state; + udev->dev.power.power_state = PMSG_SUSPEND; return status; } /** * usb_suspend_device - suspend a usb device * @udev: device that's no longer in active use - * @state: PMSG_SUSPEND to suspend * Context: must be able to sleep; device not locked * * Suspends a USB device that isn't in active use, conserving power. @@ -1664,13 +1664,16 @@ static int __usb_suspend_device (struct usb_device *udev, int port1, * suspend by the host, using usb_resume_device(). It's also routine * to disconnect devices while they are suspended. * + * This only affects the USB hardware for a device; its interfaces + * (and, for hubs, child devices) must already have been suspended. + * * Suspending OTG devices may trigger HNP, if that's been enabled * between a pair of dual-role devices. That will change roles, such * as from A-Host to A-Peripheral or from B-Host back to B-Peripheral. * * Returns 0 on success, else negative errno. */ -int usb_suspend_device(struct usb_device *udev, pm_message_t state) +int usb_suspend_device(struct usb_device *udev) { int port1, status; @@ -1678,12 +1681,15 @@ int usb_suspend_device(struct usb_device *udev, pm_message_t state) if (port1 < 0) return port1; - status = __usb_suspend_device(udev, port1, state); + status = __usb_suspend_device(udev, port1); usb_unlock_device(udev); return status; } /* + * If the USB "suspend" state is in use (rather than "global suspend"), + * many devices will be individually taken out of suspend state using + * special" resume" signaling. These routines kick in shortly after * hardware resume signaling is finished, either because of selective * resume (by host) or remote wakeup (by device) ... now see what changed * in the tree that's rooted at this device. @@ -1986,13 +1992,15 @@ void usb_resume_root_hub(struct usb_device *hdev) #else /* !CONFIG_USB_SUSPEND */ -int usb_suspend_device(struct usb_device *udev, pm_message_t state) +int usb_suspend_device(struct usb_device *udev) { + /* state does NOT lie by saying it's USB_STATE_SUSPENDED! */ return 0; } int usb_resume_device(struct usb_device *udev) { + udev->dev.power_state.event = PM_EVENT_ON; return 0; } diff --git a/drivers/usb/core/usb.c b/drivers/usb/core/usb.c index 6ecfdce4f848..e89dbd43e952 100644 --- a/drivers/usb/core/usb.c +++ b/drivers/usb/core/usb.c @@ -1414,14 +1414,33 @@ void usb_buffer_unmap_sg (struct usb_device *dev, unsigned pipe, usb_pipein (pipe) ? DMA_FROM_DEVICE : DMA_TO_DEVICE); } +static int verify_suspended(struct device *dev, void *unused) +{ + return (dev->power.power_state.event == PM_EVENT_ON) ? -EBUSY : 0; +} + static int usb_generic_suspend(struct device *dev, pm_message_t message) { struct usb_interface *intf; struct usb_driver *driver; int status; - if (dev->driver == &usb_generic_driver) - return usb_suspend_device (to_usb_device(dev), message); + /* USB devices enter SUSPEND state through their hubs, but can be + * marked for FREEZE as soon as their children are already idled. + */ + if (dev->driver == &usb_generic_driver) { + if (dev->power.power_state.event == message.event) + return 0; + /* we need to rule out bogus requests through sysfs */ + status = device_for_each_child(dev, NULL, verify_suspended); + if (status) + return status; + if (message.event == PM_EVENT_FREEZE) { + dev->power.power_state = message; + return 0; + } + return usb_suspend_device (to_usb_device(dev)); + } if ((dev->driver == NULL) || (dev->driver_data == &usb_generic_driver_data)) diff --git a/drivers/usb/host/ehci-hcd.c b/drivers/usb/host/ehci-hcd.c index b3eb02613bff..513fccbb8e43 100644 --- a/drivers/usb/host/ehci-hcd.c +++ b/drivers/usb/host/ehci-hcd.c @@ -759,7 +759,7 @@ static int ehci_suspend (struct usb_hcd *hcd, pm_message_t message) msleep (100); #ifdef CONFIG_USB_SUSPEND - (void) usb_suspend_device (hcd->self.root_hub, message); + (void) usb_suspend_device (hcd->self.root_hub); #else usb_lock_device (hcd->self.root_hub); (void) ehci_hub_suspend (hcd); diff --git a/drivers/usb/host/isp116x-hcd.c b/drivers/usb/host/isp116x-hcd.c index 642f35068ce2..554d60282a9d 100644 --- a/drivers/usb/host/isp116x-hcd.c +++ b/drivers/usb/host/isp116x-hcd.c @@ -1781,7 +1781,7 @@ static int isp116x_suspend(struct device *dev, pm_message_t state) VDBG("%s: state %x\n", __func__, state); - ret = usb_suspend_device(hcd->self.root_hub, state); + ret = usb_suspend_device(hcd->self.root_hub); if (!ret) { dev->power.power_state = state; INFO("%s suspended\n", hcd_name); diff --git a/drivers/usb/host/ohci-pci.c b/drivers/usb/host/ohci-pci.c index eede6be098d2..41e85980fa7a 100644 --- a/drivers/usb/host/ohci-pci.c +++ b/drivers/usb/host/ohci-pci.c @@ -119,7 +119,7 @@ static int ohci_pci_suspend (struct usb_hcd *hcd, pm_message_t message) msleep (100); #ifdef CONFIG_USB_SUSPEND - (void) usb_suspend_device (hcd->self.root_hub, message); + (void) usb_suspend_device (hcd->self.root_hub); #else usb_lock_device (hcd->self.root_hub); (void) ohci_hub_suspend (hcd); diff --git a/include/linux/usb.h b/include/linux/usb.h index 04502e183dd1..25ec91ddcd04 100644 --- a/include/linux/usb.h +++ b/include/linux/usb.h @@ -976,7 +976,7 @@ extern int usb_bulk_msg(struct usb_device *usb_dev, unsigned int pipe, int timeout); /* selective suspend/resume */ -extern int usb_suspend_device(struct usb_device *dev, pm_message_t message); +extern int usb_suspend_device(struct usb_device *dev); extern int usb_resume_device(struct usb_device *dev); -- cgit v1.2.3 From e9b7bd4ee7f6e3ee002dc72c5211cd97c7186d00 Mon Sep 17 00:00:00 2001 From: David Brownell Date: Thu, 22 Sep 2005 22:30:48 -0700 Subject: [PATCH] one less word in struct device This saves a word from "struct device" ... there's a refcounting mechanism stub that's rather ineffective (the values are never even tested!), which can safely be deleted. With this patch it uses normal device refcounting, so any potential users of the pm_parent mechanism will be more correct. (That mechanism is actually unusable for now though; it does nothing.) Signed-off-by: David Brownell Signed-off-by: Greg Kroah-Hartman drivers/base/power/main.c | 26 +++----------------------- include/linux/pm.h | 1 - 2 files changed, 3 insertions(+), 24 deletions(-) --- drivers/base/power/main.c | 26 +++----------------------- include/linux/pm.h | 1 - 2 files changed, 3 insertions(+), 24 deletions(-) (limited to 'include/linux') diff --git a/drivers/base/power/main.c b/drivers/base/power/main.c index 15e6a8f951f1..0d2e101e4f15 100644 --- a/drivers/base/power/main.c +++ b/drivers/base/power/main.c @@ -30,23 +30,6 @@ LIST_HEAD(dpm_off_irq); DECLARE_MUTEX(dpm_sem); DECLARE_MUTEX(dpm_list_sem); -/* - * PM Reference Counting. - */ - -static inline void device_pm_hold(struct device * dev) -{ - if (dev) - atomic_inc(&dev->power.pm_users); -} - -static inline void device_pm_release(struct device * dev) -{ - if (dev) - atomic_dec(&dev->power.pm_users); -} - - /** * device_pm_set_parent - Specify power dependency. * @dev: Device who needs power. @@ -62,10 +45,8 @@ static inline void device_pm_release(struct device * dev) void device_pm_set_parent(struct device * dev, struct device * parent) { - struct device * old_parent = dev->power.pm_parent; - device_pm_release(old_parent); - dev->power.pm_parent = parent; - device_pm_hold(parent); + put_device(dev->power.pm_parent); + dev->power.pm_parent = get_device(parent); } EXPORT_SYMBOL_GPL(device_pm_set_parent); @@ -75,7 +56,6 @@ int device_pm_add(struct device * dev) pr_debug("PM: Adding info for %s:%s\n", dev->bus ? dev->bus->name : "No Bus", dev->kobj.name); - atomic_set(&dev->power.pm_users, 0); down(&dpm_list_sem); list_add_tail(&dev->power.entry, &dpm_active); device_pm_set_parent(dev, dev->parent); @@ -91,7 +71,7 @@ void device_pm_remove(struct device * dev) dev->bus ? dev->bus->name : "No Bus", dev->kobj.name); down(&dpm_list_sem); dpm_sysfs_remove(dev); - device_pm_release(dev->power.pm_parent); + put_device(dev->power.pm_parent); list_del_init(&dev->power.entry); up(&dpm_list_sem); } diff --git a/include/linux/pm.h b/include/linux/pm.h index 7897cf500c51..83bae1cbe55b 100644 --- a/include/linux/pm.h +++ b/include/linux/pm.h @@ -224,7 +224,6 @@ struct dev_pm_info { unsigned should_wakeup:1; pm_message_t prev_state; void * saved_state; - atomic_t pm_users; struct device * pm_parent; struct list_head entry; #endif -- cgit v1.2.3 From 5edbfb7c8af0577097dae87cdd4dfdba82bb9579 Mon Sep 17 00:00:00 2001 From: David Brownell Date: Thu, 22 Sep 2005 22:45:26 -0700 Subject: [PATCH] stop exporting two functions The way we're looking at USB suspend lately doesn't expect drivers to call usb_suspend_device() or usb_resume_device() directly; that'll be implicit when no interfaces are in use. This patch removes those APIs from visibility outside usbcore. Signed-off-by: David Brownell drivers/usb/core/hub.c | 12 ++++-------- drivers/usb/core/usb.h | 4 ++++ include/linux/usb.h | 5 ----- 3 files changed, 8 insertions(+), 13 deletions(-) --- drivers/usb/core/hub.c | 12 ++++-------- drivers/usb/core/usb.h | 4 ++++ include/linux/usb.h | 5 ----- 3 files changed, 8 insertions(+), 13 deletions(-) (limited to 'include/linux') diff --git a/drivers/usb/core/hub.c b/drivers/usb/core/hub.c index 6a601a4547e7..15db5e490aa2 100644 --- a/drivers/usb/core/hub.c +++ b/drivers/usb/core/hub.c @@ -1652,10 +1652,10 @@ static int __usb_suspend_device (struct usb_device *udev, int port1) #endif -/** +/* * usb_suspend_device - suspend a usb device * @udev: device that's no longer in active use - * Context: must be able to sleep; device not locked + * Context: must be able to sleep; device not locked; pm locks held * * Suspends a USB device that isn't in active use, conserving power. * Devices may wake out of a suspend, if anything important happens, @@ -1820,10 +1820,10 @@ hub_port_resume(struct usb_hub *hub, int port1, struct usb_device *udev) #endif -/** +/* * usb_resume_device - re-activate a suspended usb device * @udev: device to re-activate - * Context: must be able to sleep; device not locked + * Context: must be able to sleep; device not locked; pm locks held * * This will re-activate the suspended device, increasing power usage * while letting drivers communicate again with its endpoints. @@ -2024,10 +2024,6 @@ void usb_resume_root_hub(struct usb_device *hdev) kick_khubd(hub); } -EXPORT_SYMBOL(usb_suspend_device); -EXPORT_SYMBOL(usb_resume_device); - - /* USB 2.0 spec, 7.1.7.3 / fig 7-29: * diff --git a/drivers/usb/core/usb.h b/drivers/usb/core/usb.h index 7add46ecc6a2..4d59f6e07999 100644 --- a/drivers/usb/core/usb.h +++ b/drivers/usb/core/usb.h @@ -29,6 +29,10 @@ extern void usb_major_cleanup(void); extern int usb_host_init(void); extern void usb_host_cleanup(void); +extern int usb_suspend_device(struct usb_device *dev); +extern int usb_resume_device(struct usb_device *dev); + + /* Interfaces and their "power state" are owned by usbcore */ static inline void mark_active(struct usb_interface *f) diff --git a/include/linux/usb.h b/include/linux/usb.h index 25ec91ddcd04..207b1ad9d990 100644 --- a/include/linux/usb.h +++ b/include/linux/usb.h @@ -975,11 +975,6 @@ extern int usb_bulk_msg(struct usb_device *usb_dev, unsigned int pipe, void *data, int len, int *actual_length, int timeout); -/* selective suspend/resume */ -extern int usb_suspend_device(struct usb_device *dev); -extern int usb_resume_device(struct usb_device *dev); - - /* wrappers around usb_control_msg() for the most common standard requests */ extern int usb_get_descriptor(struct usb_device *dev, unsigned char desctype, unsigned char descindex, void *buf, int size); -- cgit v1.2.3 From 4e67185a7ac3ecb1710e636e2f7e318e99298c7a Mon Sep 17 00:00:00 2001 From: Juha Yrj?l? Date: Sun, 16 Oct 2005 15:47:04 -0700 Subject: [PATCH] add usb transceiver set_suspend() method When a USB device is put into suspend mode, the current drawn from VBUS has to be less than 500 uA. Some transceivers need to be put into a special power-saving mode to accomplish this, and won't have a separate OTG driver handling that. This adds a suspend method to the "otg_transceiver" struct -- misnamed, it's not only for OTG -- and calls it from the OMAP UDC driver. Signed-off-by: Juha Yrj?l? Signed-off-by: David Brownell Signed-off-by: Greg Kroah-Hartman --- drivers/usb/gadget/omap_udc.c | 4 ++++ include/linux/usb_otg.h | 13 +++++++++++++ 2 files changed, 17 insertions(+) (limited to 'include/linux') diff --git a/drivers/usb/gadget/omap_udc.c b/drivers/usb/gadget/omap_udc.c index de8a89ae4bc3..b42799bc747a 100644 --- a/drivers/usb/gadget/omap_udc.c +++ b/drivers/usb/gadget/omap_udc.c @@ -1788,8 +1788,12 @@ static void devstate_irq(struct omap_udc *udc, u16 irq_src) udc->driver->suspend(&udc->gadget); spin_lock(&udc->lock); } + if (udc->transceiver) + otg_set_suspend(udc->transceiver, 1); } else { VDBG("resume\n"); + if (udc->transceiver) + otg_set_suspend(udc->transceiver, 0); if (udc->gadget.speed == USB_SPEED_FULL && udc->driver->resume) { spin_unlock(&udc->lock); diff --git a/include/linux/usb_otg.h b/include/linux/usb_otg.h index c6683146e9b0..f827f6e203c2 100644 --- a/include/linux/usb_otg.h +++ b/include/linux/usb_otg.h @@ -63,6 +63,10 @@ struct otg_transceiver { int (*set_power)(struct otg_transceiver *otg, unsigned mA); + /* for non-OTG B devices: set transceiver into suspend mode */ + int (*set_suspend)(struct otg_transceiver *otg, + int suspend); + /* for B devices only: start session with A-Host */ int (*start_srp)(struct otg_transceiver *otg); @@ -107,6 +111,15 @@ otg_set_power(struct otg_transceiver *otg, unsigned mA) return otg->set_power(otg, mA); } +static inline int +otg_set_suspend(struct otg_transceiver *otg, int suspend) +{ + if (otg->set_suspend != NULL) + return otg->set_suspend(otg, suspend); + else + return 0; +} + static inline int otg_start_srp(struct otg_transceiver *otg) { -- cgit v1.2.3 From 3099e75a7ccc3c5b0a4cf988a76d9c4a7fa5e91a Mon Sep 17 00:00:00 2001 From: Greg Kroah-Hartman Date: Mon, 20 Jun 2005 21:15:16 -0700 Subject: [PATCH] USB: add notifier functions to the USB core for devices and busses This should let us get rid of all of the different hooks in the USB core for when something has changed. Also, some other parts of the kernel have wanted to know this kind of information at times. Signed-off-by: Greg Kroah-Hartman --- drivers/usb/core/Makefile | 2 +- drivers/usb/core/hcd.c | 2 + drivers/usb/core/hub.c | 3 ++ drivers/usb/core/notify.c | 120 ++++++++++++++++++++++++++++++++++++++++++++++ drivers/usb/core/usb.h | 6 +++ include/linux/usb.h | 8 ++++ 6 files changed, 140 insertions(+), 1 deletion(-) create mode 100644 drivers/usb/core/notify.c (limited to 'include/linux') diff --git a/drivers/usb/core/Makefile b/drivers/usb/core/Makefile index d5503cf0bf74..dd1c4d2a0c31 100644 --- a/drivers/usb/core/Makefile +++ b/drivers/usb/core/Makefile @@ -3,7 +3,7 @@ # usbcore-objs := usb.o hub.o hcd.o urb.o message.o \ - config.o file.o buffer.o sysfs.o devio.o + config.o file.o buffer.o sysfs.o devio.o notify.o ifeq ($(CONFIG_PCI),y) usbcore-objs += hcd-pci.o diff --git a/drivers/usb/core/hcd.c b/drivers/usb/core/hcd.c index 9ad3912a5ed7..b700b6cdb683 100644 --- a/drivers/usb/core/hcd.c +++ b/drivers/usb/core/hcd.c @@ -792,6 +792,7 @@ static int usb_register_bus(struct usb_bus *bus) list_add (&bus->bus_list, &usb_bus_list); up (&usb_bus_list_lock); + usb_notify_add_bus(bus); usbfs_add_bus (bus); usbmon_notify_bus_add (bus); @@ -820,6 +821,7 @@ static void usb_deregister_bus (struct usb_bus *bus) list_del (&bus->bus_list); up (&usb_bus_list_lock); + usb_notify_remove_bus(bus); usbmon_notify_bus_remove (bus); usbfs_remove_bus (bus); diff --git a/drivers/usb/core/hub.c b/drivers/usb/core/hub.c index 273e6ccca213..4f1a8c8cf92b 100644 --- a/drivers/usb/core/hub.c +++ b/drivers/usb/core/hub.c @@ -1128,6 +1128,8 @@ void usb_disconnect(struct usb_device **pdev) */ usb_disable_device(udev, 0); + usb_notify_remove_device(udev); + /* Free the device number, remove the /proc/bus/usb entry and * the sysfs attributes, and delete the parent's children[] * (or root_hub) pointer. @@ -1371,6 +1373,7 @@ int usb_new_device(struct usb_device *udev) } /* USB device state == configured ... usable */ + usb_notify_add_device(udev); /* add a /proc/bus/usb entry */ usbdev_add(udev); diff --git a/drivers/usb/core/notify.c b/drivers/usb/core/notify.c new file mode 100644 index 000000000000..37da059eced7 --- /dev/null +++ b/drivers/usb/core/notify.c @@ -0,0 +1,120 @@ +/* + * All the USB notify logic + * + * (C) Copyright 2005 Greg Kroah-Hartman + * + * notifier functions originally based on those in kernel/sys.c + * but fixed up to not be so broken. + * + */ + + +#include +#include +#include +#ifdef CONFIG_USB_DEBUG + #define DEBUG +#else + #undef DEBUG +#endif +#include + +#include "usb.h" + + +static struct notifier_block *usb_notifier_list; +static DECLARE_MUTEX(usb_notifier_lock); + +static void usb_notifier_chain_register(struct notifier_block **list, + struct notifier_block *n) +{ + down(&usb_notifier_lock); + while (*list) { + if (n->priority > (*list)->priority) + break; + list = &((*list)->next); + } + n->next = *list; + *list = n; + up(&usb_notifier_lock); +} + +static void usb_notifier_chain_unregister(struct notifier_block **nl, + struct notifier_block *n) +{ + down(&usb_notifier_lock); + while ((*nl)!=NULL) { + if ((*nl)==n) { + *nl = n->next; + goto exit; + } + nl=&((*nl)->next); + } +exit: + up(&usb_notifier_lock); +} + +static int usb_notifier_call_chain(struct notifier_block **n, + unsigned long val, void *v) +{ + int ret=NOTIFY_DONE; + struct notifier_block *nb = *n; + + down(&usb_notifier_lock); + while (nb) { + ret = nb->notifier_call(nb,val,v); + if (ret&NOTIFY_STOP_MASK) { + goto exit; + } + nb = nb->next; + } +exit: + up(&usb_notifier_lock); + return ret; +} + +/** + * usb_register_notify - register a notifier callback whenever a usb change happens + * @nb: pointer to the notifier block for the callback events. + * + * These changes are either USB devices or busses being added or removed. + */ +void usb_register_notify(struct notifier_block *nb) +{ + usb_notifier_chain_register(&usb_notifier_list, nb); +} +EXPORT_SYMBOL_GPL(usb_register_notify); + +/** + * usb_unregister_notify - unregister a notifier callback + * @nb: pointer to the notifier block for the callback events. + * + * usb_register_notifier() must have been previously called for this function + * to work properly. + */ +void usb_unregister_notify(struct notifier_block *nb) +{ + usb_notifier_chain_unregister(&usb_notifier_list, nb); +} +EXPORT_SYMBOL_GPL(usb_unregister_notify); + + +void usb_notify_add_device(struct usb_device *udev) +{ + usb_notifier_call_chain(&usb_notifier_list, USB_DEVICE_ADD, udev); +} + +void usb_notify_remove_device(struct usb_device *udev) +{ + usb_notifier_call_chain(&usb_notifier_list, USB_DEVICE_REMOVE, udev); +} + +void usb_notify_add_bus(struct usb_bus *ubus) +{ + usb_notifier_call_chain(&usb_notifier_list, USB_BUS_ADD, ubus); +} + +void usb_notify_remove_bus(struct usb_bus *ubus) +{ + usb_notifier_call_chain(&usb_notifier_list, USB_BUS_REMOVE, ubus); +} diff --git a/drivers/usb/core/usb.h b/drivers/usb/core/usb.h index 3dc8096c0f92..811cf4482e0d 100644 --- a/drivers/usb/core/usb.h +++ b/drivers/usb/core/usb.h @@ -80,3 +80,9 @@ struct dev_state { unsigned long ifclaimed; }; +/* internal notify stuff */ +extern void usb_notify_add_device(struct usb_device *udev); +extern void usb_notify_remove_device(struct usb_device *udev); +extern void usb_notify_add_bus(struct usb_bus *ubus); +extern void usb_notify_remove_bus(struct usb_bus *ubus); + diff --git a/include/linux/usb.h b/include/linux/usb.h index 207b1ad9d990..a2d923fd54f9 100644 --- a/include/linux/usb.h +++ b/include/linux/usb.h @@ -1135,6 +1135,14 @@ usb_maxpacket(struct usb_device *udev, int pipe, int is_out) /* -------------------------------------------------------------------------- */ +/* Events from the usb core */ +#define USB_DEVICE_ADD 0x0001 +#define USB_DEVICE_REMOVE 0x0002 +#define USB_BUS_ADD 0x0003 +#define USB_BUS_REMOVE 0x0004 +extern void usb_register_notify(struct notifier_block *nb); +extern void usb_unregister_notify(struct notifier_block *nb); + #ifdef DEBUG #define dbg(format, arg...) printk(KERN_DEBUG "%s: " format "\n" , __FILE__ , ## arg) #else -- cgit v1.2.3 From c36fc889b5a4fd66cfd9ba80d9e038745d349567 Mon Sep 17 00:00:00 2001 From: Pete Zaitcev Date: Mon, 17 Oct 2005 18:15:54 -0700 Subject: [PATCH] usb: Patch for USBDEVFS_IOCTL from 32-bit programs Dell supplied me with the following test: #include #include #include #include #include main(int argc,char*argv[]) { struct usbdevfs_hub_portinfo hubPortInfo = {0}; struct usbdevfs_ioctl command = {0}; command.ifno = 0; command.ioctl_code = USBDEVFS_HUB_PORTINFO; command.data = (void*)&hubPortInfo; int fd, ret; if(argc != 2) { fprintf(stderr,"Usage: %s /proc/bus/usb//\n",argv[0]); fprintf(stderr,"Example: %s /proc/bus/usb/001/001\n",argv[0]); exit(1); } errno = 0; fd = open(argv[1],O_RDWR); if(fd < 0) { perror("open failed:"); exit(errno); } errno = 0; ret = ioctl(fd,USBDEVFS_IOCTL,&command); printf("IOCTL return status:%d\n",ret); if(ret<0) { perror("IOCTL failed:"); close(fd); exit(3); } else { printf("IOCTL passed:Num of ports %d\n",hubPortInfo.nports); close(fd); exit(0); } return 0; } I have verified that it breaks if built in 32 bit mode on x86_64 and that the patch below fixes it. Signed-off-by: Pete Zaitcev Signed-off-by: Greg Kroah-Hartman --- drivers/usb/core/devio.c | 56 +++++++++++++++++++++++++++++++++----------- fs/compat_ioctl.c | 1 + include/linux/usbdevice_fs.h | 7 ++++++ 3 files changed, 50 insertions(+), 14 deletions(-) (limited to 'include/linux') diff --git a/drivers/usb/core/devio.c b/drivers/usb/core/devio.c index 2bd742ba812d..ffb2e242b100 100644 --- a/drivers/usb/core/devio.c +++ b/drivers/usb/core/devio.c @@ -1301,23 +1301,20 @@ static int proc_releaseinterface(struct dev_state *ps, void __user *arg) return 0; } -static int proc_ioctl (struct dev_state *ps, void __user *arg) +static int proc_ioctl(struct dev_state *ps, struct usbdevfs_ioctl *ctl) { - struct usbdevfs_ioctl ctrl; int size; void *buf = NULL; int retval = 0; struct usb_interface *intf = NULL; struct usb_driver *driver = NULL; - /* get input parameters and alloc buffer */ - if (copy_from_user(&ctrl, arg, sizeof (ctrl))) - return -EFAULT; - if ((size = _IOC_SIZE (ctrl.ioctl_code)) > 0) { + /* alloc buffer */ + if ((size = _IOC_SIZE (ctl->ioctl_code)) > 0) { if ((buf = kmalloc (size, GFP_KERNEL)) == NULL) return -ENOMEM; - if ((_IOC_DIR(ctrl.ioctl_code) & _IOC_WRITE)) { - if (copy_from_user (buf, ctrl.data, size)) { + if ((_IOC_DIR(ctl->ioctl_code) & _IOC_WRITE)) { + if (copy_from_user (buf, ctl->data, size)) { kfree(buf); return -EFAULT; } @@ -1333,9 +1330,9 @@ static int proc_ioctl (struct dev_state *ps, void __user *arg) if (ps->dev->state != USB_STATE_CONFIGURED) retval = -EHOSTUNREACH; - else if (!(intf = usb_ifnum_to_if (ps->dev, ctrl.ifno))) + else if (!(intf = usb_ifnum_to_if (ps->dev, ctl->ifno))) retval = -EINVAL; - else switch (ctrl.ioctl_code) { + else switch (ctl->ioctl_code) { /* disconnect kernel driver from interface */ case USBDEVFS_DISCONNECT: @@ -1367,7 +1364,7 @@ static int proc_ioctl (struct dev_state *ps, void __user *arg) if (driver == NULL || driver->ioctl == NULL) { retval = -ENOTTY; } else { - retval = driver->ioctl (intf, ctrl.ioctl_code, buf); + retval = driver->ioctl (intf, ctl->ioctl_code, buf); if (retval == -ENOIOCTLCMD) retval = -ENOTTY; } @@ -1376,15 +1373,42 @@ static int proc_ioctl (struct dev_state *ps, void __user *arg) /* cleanup and return */ if (retval >= 0 - && (_IOC_DIR (ctrl.ioctl_code) & _IOC_READ) != 0 + && (_IOC_DIR (ctl->ioctl_code) & _IOC_READ) != 0 && size > 0 - && copy_to_user (ctrl.data, buf, size) != 0) + && copy_to_user (ctl->data, buf, size) != 0) retval = -EFAULT; kfree(buf); return retval; } +static int proc_ioctl_default(struct dev_state *ps, void __user *arg) +{ + struct usbdevfs_ioctl ctrl; + + if (copy_from_user(&ctrl, arg, sizeof (ctrl))) + return -EFAULT; + return proc_ioctl(ps, &ctrl); +} + +#ifdef CONFIG_COMPAT +static int proc_ioctl_compat(struct dev_state *ps, void __user *arg) +{ + struct usbdevfs_ioctl32 __user *uioc; + struct usbdevfs_ioctl ctrl; + u32 udata; + + uioc = compat_ptr(arg); + if (get_user(ctrl.ifno, &uioc->ifno) || + get_user(ctrl.ioctl_code, &uioc->ioctl_code) || + __get_user(udata, &uioc->data)) + return -EFAULT; + ctrl.data = compat_ptr(udata); + + return proc_ioctl(ps, &ctrl); +} +#endif + /* * NOTE: All requests here that have interface numbers as parameters * are assuming that somehow the configuration has been prevented from @@ -1485,6 +1509,10 @@ static int usbdev_ioctl(struct inode *inode, struct file *file, unsigned int cmd ret = proc_reapurbnonblock_compat(ps, p); break; + case USBDEVFS_IOCTL32: + snoop(&dev->dev, "%s: IOCTL\n", __FUNCTION__); + ret = proc_ioctl_compat(ps, p); + break; #endif case USBDEVFS_DISCARDURB: @@ -1519,7 +1547,7 @@ static int usbdev_ioctl(struct inode *inode, struct file *file, unsigned int cmd case USBDEVFS_IOCTL: snoop(&dev->dev, "%s: IOCTL\n", __FUNCTION__); - ret = proc_ioctl(ps, p); + ret = proc_ioctl_default(ps, p); break; } usb_unlock_device(dev); diff --git a/fs/compat_ioctl.c b/fs/compat_ioctl.c index e28a74203f3b..a327e03753ac 100644 --- a/fs/compat_ioctl.c +++ b/fs/compat_ioctl.c @@ -3050,6 +3050,7 @@ HANDLE_IOCTL(TIOCSSERIAL, serial_struct_ioctl) HANDLE_IOCTL(USBDEVFS_CONTROL32, do_usbdevfs_control) HANDLE_IOCTL(USBDEVFS_BULK32, do_usbdevfs_bulk) HANDLE_IOCTL(USBDEVFS_DISCSIGNAL32, do_usbdevfs_discsignal) +COMPATIBLE_IOCTL(USBDEVFS_IOCTL32) /* i2c */ HANDLE_IOCTL(I2C_FUNCS, w_long) HANDLE_IOCTL(I2C_RDWR, do_i2c_rdwr_ioctl) diff --git a/include/linux/usbdevice_fs.h b/include/linux/usbdevice_fs.h index 9facf733800c..8859f0b41543 100644 --- a/include/linux/usbdevice_fs.h +++ b/include/linux/usbdevice_fs.h @@ -140,6 +140,12 @@ struct usbdevfs_urb32 { compat_caddr_t usercontext; /* unused */ struct usbdevfs_iso_packet_desc iso_frame_desc[0]; }; + +struct usbdevfs_ioctl32 { + s32 ifno; + s32 ioctl_code; + compat_caddr_t data; +}; #endif #define USBDEVFS_CONTROL _IOWR('U', 0, struct usbdevfs_ctrltransfer) @@ -160,6 +166,7 @@ struct usbdevfs_urb32 { #define USBDEVFS_RELEASEINTERFACE _IOR('U', 16, unsigned int) #define USBDEVFS_CONNECTINFO _IOW('U', 17, struct usbdevfs_connectinfo) #define USBDEVFS_IOCTL _IOWR('U', 18, struct usbdevfs_ioctl) +#define USBDEVFS_IOCTL32 _IOWR('U', 18, struct usbdevfs_ioctl32) #define USBDEVFS_HUB_PORTINFO _IOR('U', 19, struct usbdevfs_hub_portinfo) #define USBDEVFS_RESET _IO('U', 20) #define USBDEVFS_CLEAR_HALT _IOR('U', 21, unsigned int) -- cgit v1.2.3 From 478a3bab8c87a9ba4a4ba338314e32bb0c378e62 Mon Sep 17 00:00:00 2001 From: Alan Stern Date: Wed, 19 Oct 2005 12:52:02 -0400 Subject: [PATCH] USB: Always do usb-handoff This revised patch (as586b) makes usb-handoff permanently true and no longer a kernel boot parameter. It also removes the piix3_usb quirk code; that was nothing more than an early version of the USB handoff code (written at a time when Intel's PIIX3 was about the only motherboard with USB support). And it adds identifiers for the three PCI USB controller classes to pci_ids.h. Signed-off-by: Alan Stern Signed-off-by: Greg Kroah-Hartman --- Documentation/kernel-parameters.txt | 2 -- drivers/usb/host/pci-quirks.c | 46 ++--------------------------- include/asm-i386/mach-summit/mach_mpparse.h | 3 -- include/linux/pci_ids.h | 3 ++ 4 files changed, 6 insertions(+), 48 deletions(-) (limited to 'include/linux') diff --git a/Documentation/kernel-parameters.txt b/Documentation/kernel-parameters.txt index 971589a9752d..90766b75d1b7 100644 --- a/Documentation/kernel-parameters.txt +++ b/Documentation/kernel-parameters.txt @@ -1517,8 +1517,6 @@ running once the system is up. uart6850= [HW,OSS] Format: , - usb-handoff [HW] Enable early USB BIOS -> OS handoff - usbhid.mousepoll= [USBHID] The interval which mice are to be polled at. diff --git a/drivers/usb/host/pci-quirks.c b/drivers/usb/host/pci-quirks.c index f7411ca48835..b7fd3f644e1e 100644 --- a/drivers/usb/host/pci-quirks.c +++ b/drivers/usb/host/pci-quirks.c @@ -23,33 +23,6 @@ #include -/* - * PIIX3 USB: We have to disable USB interrupts that are - * hardwired to PIRQD# and may be shared with an - * external device. - * - * Legacy Support Register (LEGSUP): - * bit13: USB PIRQ Enable (USBPIRQDEN), - * bit4: Trap/SMI On IRQ Enable (USBSMIEN). - * - * We mask out all r/wc bits, too. - */ -static void __devinit quirk_piix3_usb(struct pci_dev *dev) -{ - u16 legsup; - - pci_read_config_word(dev, 0xc0, &legsup); - legsup &= 0x50ef; - pci_write_config_word(dev, 0xc0, legsup); -} -DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_82371SB_2, quirk_piix3_usb ); -DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_82371AB_2, quirk_piix3_usb ); - - -/* FIXME these should be the guts of hcd->reset() methods; resolve all - * the differences between this version and the HCD's version. - */ - #define UHCI_USBLEGSUP 0xc0 /* legacy support */ #define UHCI_USBCMD 0 /* command register */ #define UHCI_USBINTR 4 /* interrupt register */ @@ -83,13 +56,6 @@ DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_82371AB_2, qui #define EHCI_USBLEGCTLSTS 4 /* legacy control/status */ #define EHCI_USBLEGCTLSTS_SOOE (1 << 13) /* SMI on ownership change */ -int usb_early_handoff __devinitdata = 0; -static int __init usb_handoff_early(char *str) -{ - usb_early_handoff = 1; - return 0; -} -__setup("usb-handoff", usb_handoff_early); /* * Make sure the controller is completely inactive, unable to @@ -320,17 +286,11 @@ static void __devinit quirk_usb_disable_ehci(struct pci_dev *pdev) static void __devinit quirk_usb_early_handoff(struct pci_dev *pdev) { - if (!usb_early_handoff) - return; - - if (pdev->class == ((PCI_CLASS_SERIAL_USB << 8) | 0x00)) { /* UHCI */ + if (pdev->class == PCI_CLASS_SERIAL_USB_UHCI) quirk_usb_handoff_uhci(pdev); - } else if (pdev->class == ((PCI_CLASS_SERIAL_USB << 8) | 0x10)) { /* OHCI */ + else if (pdev->class == PCI_CLASS_SERIAL_USB_OHCI) quirk_usb_handoff_ohci(pdev); - } else if (pdev->class == ((PCI_CLASS_SERIAL_USB << 8) | 0x20)) { /* EHCI */ + else if (pdev->class == PCI_CLASS_SERIAL_USB_EHCI) quirk_usb_disable_ehci(pdev); - } - - return; } DECLARE_PCI_FIXUP_HEADER(PCI_ANY_ID, PCI_ANY_ID, quirk_usb_early_handoff); diff --git a/include/asm-i386/mach-summit/mach_mpparse.h b/include/asm-i386/mach-summit/mach_mpparse.h index 2b9e6d55bef1..1cce2b924a80 100644 --- a/include/asm-i386/mach-summit/mach_mpparse.h +++ b/include/asm-i386/mach-summit/mach_mpparse.h @@ -22,7 +22,6 @@ static inline void mpc_oem_pci_bus(struct mpc_config_bus *m, { } -extern int usb_early_handoff; static inline int mps_oem_check(struct mp_config_table *mpc, char *oem, char *productid) { @@ -32,7 +31,6 @@ static inline int mps_oem_check(struct mp_config_table *mpc, char *oem, || !strncmp(productid, "RUTHLESS SMP", 12))){ use_cyclone = 1; /*enable cyclone-timer*/ setup_summit(); - usb_early_handoff = 1; return 1; } return 0; @@ -46,7 +44,6 @@ static inline int acpi_madt_oem_check(char *oem_id, char *oem_table_id) || !strncmp(oem_table_id, "EXA", 3))){ use_cyclone = 1; /*enable cyclone-timer*/ setup_summit(); - usb_early_handoff = 1; return 1; } return 0; diff --git a/include/linux/pci_ids.h b/include/linux/pci_ids.h index 7d300f7469e3..467a096c3b81 100644 --- a/include/linux/pci_ids.h +++ b/include/linux/pci_ids.h @@ -96,6 +96,9 @@ #define PCI_CLASS_SERIAL_ACCESS 0x0c01 #define PCI_CLASS_SERIAL_SSA 0x0c02 #define PCI_CLASS_SERIAL_USB 0x0c03 +#define PCI_CLASS_SERIAL_USB_UHCI 0x0c0300 +#define PCI_CLASS_SERIAL_USB_OHCI 0x0c0310 +#define PCI_CLASS_SERIAL_USB_EHCI 0x0c0320 #define PCI_CLASS_SERIAL_FIBER 0x0c04 #define PCI_CLASS_SERIAL_SMBUS 0x0c05 -- cgit v1.2.3 From be69e5b1900a19a545becda822b18d6f09168ba5 Mon Sep 17 00:00:00 2001 From: Alan Stern Date: Tue, 25 Oct 2005 15:56:06 -0400 Subject: [PATCH] usbcore: Improve endpoint sysfs file handling This revised patch (as587b) improves the implementation of USB endpoint sysfs files. Instead of storing a whole bunch of attributes for every single endpoint, each endpoint now gets its own kobject and they can share a static list of attributes. The number of extra fields added to struct usb_host_endpoint has been reduced from 4 to 1. The bEndpointAddress field is retained even though it is redundant (it repeats the same information as the attributes' directory name). The code avoids calling kobject_register, to prevent generating unwanted hotplug events. Signed-off-by: Alan Stern Signed-off-by: Greg Kroah-Hartman --- drivers/usb/core/sysfs.c | 219 +++++++++++++++++++++++++++-------------------- include/linux/usb.h | 5 +- 2 files changed, 125 insertions(+), 99 deletions(-) (limited to 'include/linux') diff --git a/drivers/usb/core/sysfs.c b/drivers/usb/core/sysfs.c index 4bdbc9df6e03..f18317fb49ee 100644 --- a/drivers/usb/core/sysfs.c +++ b/drivers/usb/core/sysfs.c @@ -23,43 +23,56 @@ #include "usb.h" /* endpoint stuff */ -struct endpoint_attribute { - struct device_attribute dev_attr; - struct usb_endpoint_descriptor *endpoint; +struct ep_object { + struct usb_endpoint_descriptor *desc; struct usb_device *udev; + struct kobject kobj; }; -#define to_endpoint_attr(_dev_attr) \ - container_of(_dev_attr, struct endpoint_attribute, dev_attr) - -#define usb_ep_attr(field, format_string) \ -static ssize_t show_ep_##field(struct device *dev, struct device_attribute *attr, \ - char *buf) \ -{ \ - struct endpoint_attribute *endpoint_attr = to_endpoint_attr(attr); \ - \ - return sprintf(buf, format_string, endpoint_attr->endpoint->field); \ -} +#define to_ep_object(_kobj) \ + container_of(_kobj, struct ep_object, kobj) + +struct ep_attribute { + struct attribute attr; + ssize_t (*show)(struct usb_device *, + struct usb_endpoint_descriptor *, char *); +}; +#define to_ep_attribute(_attr) \ + container_of(_attr, struct ep_attribute, attr) + +#define EP_ATTR(_name) \ +struct ep_attribute ep_##_name = { \ + .attr = {.name = #_name, .owner = THIS_MODULE, \ + .mode = S_IRUGO}, \ + .show = show_ep_##_name} + +#define usb_ep_attr(field, format_string) \ +static ssize_t show_ep_##field(struct usb_device *udev, \ + struct usb_endpoint_descriptor *desc, \ + char *buf) \ +{ \ + return sprintf(buf, format_string, desc->field); \ +} \ +static EP_ATTR(field); + usb_ep_attr(bLength, "%02x\n") -usb_ep_attr(bDescriptorType, "%02x\n") usb_ep_attr(bEndpointAddress, "%02x\n") usb_ep_attr(bmAttributes, "%02x\n") usb_ep_attr(bInterval, "%02x\n") -static ssize_t show_ep_wMaxPacketSize(struct device *dev, - struct device_attribute *attr, char *buf) +static ssize_t show_ep_wMaxPacketSize(struct usb_device *udev, + struct usb_endpoint_descriptor *desc, char *buf) { - struct endpoint_attribute *endpoint_attr = to_endpoint_attr(attr); - return sprintf(buf, "%04x\n", - le16_to_cpu(endpoint_attr->endpoint->wMaxPacketSize) & 0x07ff); + le16_to_cpu(desc->wMaxPacketSize) & 0x07ff); } +static EP_ATTR(wMaxPacketSize); -static ssize_t show_ep_type(struct device *dev, struct device_attribute *attr, char *buf) +static ssize_t show_ep_type(struct usb_device *udev, + struct usb_endpoint_descriptor *desc, char *buf) { - struct endpoint_attribute *endpoint_attr = to_endpoint_attr(attr); char *type = "unknown"; - switch (endpoint_attr->endpoint->bmAttributes & USB_ENDPOINT_XFERTYPE_MASK) { + switch (desc->bmAttributes & USB_ENDPOINT_XFERTYPE_MASK) { case USB_ENDPOINT_XFER_CONTROL: type = "Control"; break; @@ -75,35 +88,34 @@ static ssize_t show_ep_type(struct device *dev, struct device_attribute *attr, c } return sprintf(buf, "%s\n", type); } +static EP_ATTR(type); -static ssize_t show_ep_interval(struct device *dev, struct device_attribute *attr, char *buf) +static ssize_t show_ep_interval(struct usb_device *udev, + struct usb_endpoint_descriptor *desc, char *buf) { - struct endpoint_attribute *endpoint_attr = to_endpoint_attr(attr); - struct usb_device *udev = endpoint_attr->udev; - struct usb_endpoint_descriptor *endpoint = endpoint_attr->endpoint; char unit; unsigned interval = 0; unsigned in; - in = (endpoint->bEndpointAddress & USB_DIR_IN); + in = (desc->bEndpointAddress & USB_DIR_IN); - switch (endpoint->bmAttributes & USB_ENDPOINT_XFERTYPE_MASK) { + switch (desc->bmAttributes & USB_ENDPOINT_XFERTYPE_MASK) { case USB_ENDPOINT_XFER_CONTROL: if (udev->speed == USB_SPEED_HIGH) /* uframes per NAK */ - interval = endpoint->bInterval; + interval = desc->bInterval; break; case USB_ENDPOINT_XFER_ISOC: - interval = 1 << (endpoint->bInterval - 1); + interval = 1 << (desc->bInterval - 1); break; case USB_ENDPOINT_XFER_BULK: - if (udev->speed == USB_SPEED_HIGH && !in) /* uframes per NAK */ - interval = endpoint->bInterval; + if (udev->speed == USB_SPEED_HIGH && !in) /* uframes per NAK */ + interval = desc->bInterval; break; case USB_ENDPOINT_XFER_INT: - if (udev->speed == USB_SPEED_HIGH) { - interval = 1 << (endpoint->bInterval - 1); - } else - interval = endpoint->bInterval; + if (udev->speed == USB_SPEED_HIGH) + interval = 1 << (desc->bInterval - 1); + else + interval = desc->bInterval; break; } interval *= (udev->speed == USB_SPEED_HIGH) ? 125 : 1000; @@ -116,78 +128,95 @@ static ssize_t show_ep_interval(struct device *dev, struct device_attribute *att return sprintf(buf, "%d%cs\n", interval, unit); } +static EP_ATTR(interval); -static ssize_t show_ep_direction(struct device *dev, struct device_attribute *attr, char *buf) +static ssize_t show_ep_direction(struct usb_device *udev, + struct usb_endpoint_descriptor *desc, char *buf) { - struct endpoint_attribute *endpoint_attr = to_endpoint_attr(attr); char *direction; - if ((endpoint_attr->endpoint->bmAttributes & USB_ENDPOINT_XFERTYPE_MASK) == - USB_ENDPOINT_XFER_CONTROL) + if ((desc->bmAttributes & USB_ENDPOINT_XFERTYPE_MASK) == + USB_ENDPOINT_XFER_CONTROL) direction = "both"; - else if (endpoint_attr->endpoint->bEndpointAddress & USB_DIR_IN) + else if (desc->bEndpointAddress & USB_DIR_IN) direction = "in"; else direction = "out"; return sprintf(buf, "%s\n", direction); } +static EP_ATTR(direction); + +static struct attribute *ep_attrs[] = { + &ep_bLength.attr, + &ep_bEndpointAddress.attr, + &ep_bmAttributes.attr, + &ep_bInterval.attr, + &ep_wMaxPacketSize.attr, + &ep_type.attr, + &ep_interval.attr, + &ep_direction.attr, + NULL, +}; -static struct endpoint_attribute *create_ep_attr(struct usb_endpoint_descriptor *endpoint, - struct usb_device *udev, char *name, - ssize_t (*show)(struct device *dev, struct device_attribute *attr, char *buf)) +static void ep_object_release(struct kobject *kobj) { - struct endpoint_attribute *ep_attr; - - ep_attr = kzalloc(sizeof(*ep_attr), GFP_KERNEL); - if (ep_attr) { - ep_attr->endpoint = endpoint; - ep_attr->udev = udev; - ep_attr->dev_attr.attr.name = name; - ep_attr->dev_attr.attr.mode = 0444; - ep_attr->dev_attr.attr.owner = THIS_MODULE; - ep_attr->dev_attr.show = show; - } - return ep_attr; + kfree(to_ep_object(kobj)); } -static void usb_create_ep_files(struct kobject *kobj, struct usb_host_endpoint *endpoint, struct usb_device *udev) +static ssize_t ep_object_show(struct kobject *kobj, struct attribute *attr, + char *buf) { - struct usb_endpoint_descriptor *ep; - - ep = &endpoint->desc; - - endpoint->attrs = kzalloc(sizeof(struct attribute *) * 10, GFP_KERNEL); - endpoint->attrs[0] = &(create_ep_attr(ep, udev, "direction", show_ep_direction)->dev_attr.attr); - endpoint->attrs[1] = &(create_ep_attr(ep, udev, "type", show_ep_type)->dev_attr.attr); - endpoint->attrs[2] = &(create_ep_attr(ep, udev, "bLength", show_ep_bLength)->dev_attr.attr); - endpoint->attrs[3] = &(create_ep_attr(ep, udev, "bDescriptorType", show_ep_bDescriptorType)->dev_attr.attr); - endpoint->attrs[4] = &(create_ep_attr(ep, udev, "bEndpointAddress", show_ep_bEndpointAddress)->dev_attr.attr); - endpoint->attrs[5] = &(create_ep_attr(ep, udev, "bmAttributes", show_ep_bmAttributes)->dev_attr.attr); - endpoint->attrs[6] = &(create_ep_attr(ep, udev, "wMaxPacketSize", show_ep_wMaxPacketSize)->dev_attr.attr); - endpoint->attrs[7] = &(create_ep_attr(ep, udev, "bInterval", show_ep_bInterval)->dev_attr.attr); - endpoint->attrs[8] = &(create_ep_attr(ep, udev, "interval", show_ep_interval)->dev_attr.attr); - endpoint->attrs[9] = NULL; - endpoint->num_attrs = 9; - - endpoint->attr_group = kzalloc(sizeof(*endpoint->attr_group), GFP_KERNEL); - endpoint->attr_name = kzalloc(10, GFP_KERNEL); - sprintf(endpoint->attr_name, "ep_%02x", endpoint->desc.bEndpointAddress); - - endpoint->attr_group->attrs = endpoint->attrs; - endpoint->attr_group->name = endpoint->attr_name; - sysfs_create_group(kobj, endpoint->attr_group); + struct ep_object *ep_obj = to_ep_object(kobj); + struct ep_attribute *ep_attr = to_ep_attribute(attr); + + return (ep_attr->show)(ep_obj->udev, ep_obj->desc, buf); } -static void usb_remove_ep_files(struct kobject *kobj, struct usb_host_endpoint *endpoint) +static struct sysfs_ops ep_object_sysfs_ops = { + .show = ep_object_show, +}; + +static struct kobj_type ep_object_ktype = { + .release = ep_object_release, + .sysfs_ops = &ep_object_sysfs_ops, + .default_attrs = ep_attrs, +}; + +static void usb_create_ep_files(struct kobject *parent, + struct usb_host_endpoint *endpoint, + struct usb_device *udev) { - int i; + struct ep_object *ep_obj; + struct kobject *kobj; + + ep_obj = kzalloc(sizeof(struct ep_object), GFP_KERNEL); + if (!ep_obj) + return; - sysfs_remove_group(kobj, endpoint->attr_group); - kfree(endpoint->attr_group); - kfree(endpoint->attr_name); - for (i = 0; i < endpoint->num_attrs; ++i) - kfree(endpoint->attrs[i]); - kfree(endpoint->attrs); + ep_obj->desc = &endpoint->desc; + ep_obj->udev = udev; + + kobj = &ep_obj->kobj; + kobject_set_name(kobj, "ep_%02x", endpoint->desc.bEndpointAddress); + kobj->parent = parent; + kobj->ktype = &ep_object_ktype; + + /* Don't use kobject_register, because it generates a hotplug event */ + kobject_init(kobj); + if (kobject_add(kobj) == 0) + endpoint->kobj = kobj; + else + kobject_put(kobj); +} + +static void usb_remove_ep_files(struct usb_host_endpoint *endpoint) +{ + + if (endpoint->kobj) { + kobject_del(endpoint->kobj); + kobject_put(endpoint->kobj); + endpoint->kobj = NULL; + } } /* Active configuration fields */ @@ -411,7 +440,7 @@ void usb_remove_sysfs_dev_files (struct usb_device *udev) { struct device *dev = &udev->dev; - usb_remove_ep_files(&dev->kobj, &udev->ep0); + usb_remove_ep_files(&udev->ep0); sysfs_remove_group(&dev->kobj, &dev_attr_grp); if (udev->descriptor.iManufacturer) @@ -496,7 +525,7 @@ static struct attribute_group intf_attr_grp = { .attrs = intf_attrs, }; -static void usb_create_intf_ep_files(struct usb_interface *intf) +static inline void usb_create_intf_ep_files(struct usb_interface *intf) { struct usb_host_interface *iface_desc; int i; @@ -504,17 +533,17 @@ static void usb_create_intf_ep_files(struct usb_interface *intf) iface_desc = intf->cur_altsetting; for (i = 0; i < iface_desc->desc.bNumEndpoints; ++i) usb_create_ep_files(&intf->dev.kobj, &iface_desc->endpoint[i], - interface_to_usbdev(intf)); + interface_to_usbdev(intf)); } -static void usb_remove_intf_ep_files(struct usb_interface *intf) +static inline void usb_remove_intf_ep_files(struct usb_interface *intf) { struct usb_host_interface *iface_desc; int i; iface_desc = intf->cur_altsetting; for (i = 0; i < iface_desc->desc.bNumEndpoints; ++i) - usb_remove_ep_files(&intf->dev.kobj, &iface_desc->endpoint[i]); + usb_remove_ep_files(&iface_desc->endpoint[i]); } void usb_create_sysfs_intf_files (struct usb_interface *intf) diff --git a/include/linux/usb.h b/include/linux/usb.h index a2d923fd54f9..465ff4585ca5 100644 --- a/include/linux/usb.h +++ b/include/linux/usb.h @@ -57,10 +57,7 @@ struct usb_host_endpoint { struct usb_endpoint_descriptor desc; struct list_head urb_list; void *hcpriv; - char *attr_name; - struct attribute_group *attr_group; - struct attribute **attrs; - int num_attrs; + struct kobject *kobj; /* For sysfs info */ unsigned char *extra; /* Extra descriptors */ int extralen; -- cgit v1.2.3 From b724ae77969fd832be71419dca74bece9af287ff Mon Sep 17 00:00:00 2001 From: Alan Stern Date: Mon, 24 Oct 2005 15:36:00 -0400 Subject: [PATCH] usbcore: Wrap lines before column 80 I can't stand text lines that wrap-around in my 80-column windows. This patch (as589) makes cosmetic changes to a couple of source files. Signed-off-by: Alan Stern Signed-off-by: Greg Kroah-Hartman --- drivers/usb/core/sysfs.c | 30 +++++++---- include/linux/usb.h | 132 ++++++++++++++++++++++++++++++----------------- 2 files changed, 106 insertions(+), 56 deletions(-) (limited to 'include/linux') diff --git a/drivers/usb/core/sysfs.c b/drivers/usb/core/sysfs.c index f18317fb49ee..4cca77cf0c48 100644 --- a/drivers/usb/core/sysfs.c +++ b/drivers/usb/core/sysfs.c @@ -221,7 +221,8 @@ static void usb_remove_ep_files(struct usb_host_endpoint *endpoint) /* Active configuration fields */ #define usb_actconfig_show(field, multiplier, format_string) \ -static ssize_t show_##field (struct device *dev, struct device_attribute *attr, char *buf) \ +static ssize_t show_##field (struct device *dev, \ + struct device_attribute *attr, char *buf) \ { \ struct usb_device *udev; \ struct usb_host_config *actconfig; \ @@ -243,7 +244,8 @@ usb_actconfig_attr (bNumInterfaces, 1, "%2d\n") usb_actconfig_attr (bmAttributes, 1, "%2x\n") usb_actconfig_attr (bMaxPower, 2, "%3dmA\n") -static ssize_t show_configuration_string(struct device *dev, struct device_attribute *attr, char *buf) +static ssize_t show_configuration_string(struct device *dev, + struct device_attribute *attr, char *buf) { struct usb_device *udev; struct usb_host_config *actconfig; @@ -266,7 +268,8 @@ static DEVICE_ATTR(configuration, S_IRUGO, show_configuration_string, NULL); usb_actconfig_show(bConfigurationValue, 1, "%u\n"); static ssize_t -set_bConfigurationValue (struct device *dev, struct device_attribute *attr, const char *buf, size_t count) +set_bConfigurationValue (struct device *dev, struct device_attribute *attr, + const char *buf, size_t count) { struct usb_device *udev = udev = to_usb_device (dev); int config, value; @@ -284,7 +287,8 @@ static DEVICE_ATTR(bConfigurationValue, S_IRUGO | S_IWUSR, /* String fields */ #define usb_string_attr(name) \ -static ssize_t show_##name(struct device *dev, struct device_attribute *attr, char *buf) \ +static ssize_t show_##name(struct device *dev, \ + struct device_attribute *attr, char *buf) \ { \ struct usb_device *udev; \ int len; \ @@ -364,7 +368,8 @@ static DEVICE_ATTR(maxchild, S_IRUGO, show_maxchild, NULL); /* Descriptor fields */ #define usb_descriptor_attr_le16(field, format_string) \ static ssize_t \ -show_##field (struct device *dev, struct device_attribute *attr, char *buf) \ +show_##field (struct device *dev, struct device_attribute *attr, \ + char *buf) \ { \ struct usb_device *udev; \ \ @@ -380,7 +385,8 @@ usb_descriptor_attr_le16(bcdDevice, "%04x\n") #define usb_descriptor_attr(field, format_string) \ static ssize_t \ -show_##field (struct device *dev, struct device_attribute *attr, char *buf) \ +show_##field (struct device *dev, struct device_attribute *attr, \ + char *buf) \ { \ struct usb_device *udev; \ \ @@ -455,11 +461,13 @@ void usb_remove_sysfs_dev_files (struct usb_device *udev) /* Interface fields */ #define usb_intf_attr(field, format_string) \ static ssize_t \ -show_##field (struct device *dev, struct device_attribute *attr, char *buf) \ +show_##field (struct device *dev, struct device_attribute *attr, \ + char *buf) \ { \ struct usb_interface *intf = to_usb_interface (dev); \ \ - return sprintf (buf, format_string, intf->cur_altsetting->desc.field); \ + return sprintf (buf, format_string, \ + intf->cur_altsetting->desc.field); \ } \ static DEVICE_ATTR(field, S_IRUGO, show_##field, NULL); @@ -470,7 +478,8 @@ usb_intf_attr (bInterfaceClass, "%02x\n") usb_intf_attr (bInterfaceSubClass, "%02x\n") usb_intf_attr (bInterfaceProtocol, "%02x\n") -static ssize_t show_interface_string(struct device *dev, struct device_attribute *attr, char *buf) +static ssize_t show_interface_string(struct device *dev, + struct device_attribute *attr, char *buf) { struct usb_interface *intf; struct usb_device *udev; @@ -487,7 +496,8 @@ static ssize_t show_interface_string(struct device *dev, struct device_attribute } static DEVICE_ATTR(interface, S_IRUGO, show_interface_string, NULL); -static ssize_t show_modalias(struct device *dev, struct device_attribute *attr, char *buf) +static ssize_t show_modalias(struct device *dev, + struct device_attribute *attr, char *buf) { struct usb_interface *intf; struct usb_device *udev; diff --git a/include/linux/usb.h b/include/linux/usb.h index 465ff4585ca5..c500d6b5a16d 100644 --- a/include/linux/usb.h +++ b/include/linux/usb.h @@ -137,7 +137,8 @@ struct usb_interface { * active alternate setting */ unsigned num_altsetting; /* number of alternate settings */ - int minor; /* minor number this interface is bound to */ + int minor; /* minor number this interface is + * bound to */ enum usb_interface_condition condition; /* state of binding */ struct device dev; /* interface specific device info */ struct class_device *class_dev; @@ -249,7 +250,7 @@ int __usb_get_extra_descriptor(char *buffer, unsigned size, __usb_get_extra_descriptor((ifpoint)->extra,(ifpoint)->extralen,\ type,(void**)ptr) -/* -------------------------------------------------------------------------- */ +/* ----------------------------------------------------------------------- */ struct usb_operations; @@ -269,7 +270,8 @@ struct usb_bus { unsigned is_b_host:1; /* true during some HNP roleswitches */ unsigned b_hnp_enable:1; /* OTG: did A-Host enable HNP? */ - int devnum_next; /* Next open device number in round-robin allocation */ + int devnum_next; /* Next open device number in + * round-robin allocation */ struct usb_devmap devmap; /* device address allocation map */ struct usb_operations *op; /* Operations (specific to the HC) */ @@ -290,15 +292,16 @@ struct usb_bus { struct dentry *usbfs_dentry; /* usbfs dentry entry for the bus */ struct class_device *class_dev; /* class device for this bus */ - struct kref kref; /* handles reference counting this bus */ - void (*release)(struct usb_bus *bus); /* function to destroy this bus's memory */ + struct kref kref; /* reference counting for this bus */ + void (*release)(struct usb_bus *bus); + #if defined(CONFIG_USB_MON) struct mon_bus *mon_bus; /* non-null when associated */ int monitored; /* non-zero when monitored */ #endif }; -/* -------------------------------------------------------------------------- */ +/* ----------------------------------------------------------------------- */ /* This is arbitrary. * From USB 2.0 spec Table 11-13, offset 7, a hub can @@ -327,7 +330,8 @@ struct usb_device { struct semaphore serialize; - unsigned int toggle[2]; /* one bit for each endpoint ([0] = IN, [1] = OUT) */ + unsigned int toggle[2]; /* one bit for each endpoint + * ([0] = IN, [1] = OUT) */ struct usb_device *parent; /* our hub, unless we're the root */ struct usb_bus *bus; /* Bus we're part of */ @@ -344,7 +348,7 @@ struct usb_device { char **rawdescriptors; /* Raw descriptors for each config */ - int have_langid; /* whether string_langid is valid yet */ + int have_langid; /* whether string_langid is valid */ int string_langid; /* language ID for strings */ char *product; @@ -441,22 +445,31 @@ extern struct usb_host_interface *usb_altnum_to_altsetting( * USB 2.0 root hubs (EHCI host controllers) will get one path ID if they are * high speed, and a different one if they are full or low speed. */ -static inline int usb_make_path (struct usb_device *dev, char *buf, size_t size) +static inline int usb_make_path (struct usb_device *dev, char *buf, + size_t size) { int actual; - actual = snprintf (buf, size, "usb-%s-%s", dev->bus->bus_name, dev->devpath); + actual = snprintf (buf, size, "usb-%s-%s", dev->bus->bus_name, + dev->devpath); return (actual >= (int)size) ? -1 : actual; } /*-------------------------------------------------------------------------*/ -#define USB_DEVICE_ID_MATCH_DEVICE (USB_DEVICE_ID_MATCH_VENDOR | USB_DEVICE_ID_MATCH_PRODUCT) -#define USB_DEVICE_ID_MATCH_DEV_RANGE (USB_DEVICE_ID_MATCH_DEV_LO | USB_DEVICE_ID_MATCH_DEV_HI) -#define USB_DEVICE_ID_MATCH_DEVICE_AND_VERSION (USB_DEVICE_ID_MATCH_DEVICE | USB_DEVICE_ID_MATCH_DEV_RANGE) +#define USB_DEVICE_ID_MATCH_DEVICE \ + (USB_DEVICE_ID_MATCH_VENDOR | USB_DEVICE_ID_MATCH_PRODUCT) +#define USB_DEVICE_ID_MATCH_DEV_RANGE \ + (USB_DEVICE_ID_MATCH_DEV_LO | USB_DEVICE_ID_MATCH_DEV_HI) +#define USB_DEVICE_ID_MATCH_DEVICE_AND_VERSION \ + (USB_DEVICE_ID_MATCH_DEVICE | USB_DEVICE_ID_MATCH_DEV_RANGE) #define USB_DEVICE_ID_MATCH_DEV_INFO \ - (USB_DEVICE_ID_MATCH_DEV_CLASS | USB_DEVICE_ID_MATCH_DEV_SUBCLASS | USB_DEVICE_ID_MATCH_DEV_PROTOCOL) + (USB_DEVICE_ID_MATCH_DEV_CLASS | \ + USB_DEVICE_ID_MATCH_DEV_SUBCLASS | \ + USB_DEVICE_ID_MATCH_DEV_PROTOCOL) #define USB_DEVICE_ID_MATCH_INT_INFO \ - (USB_DEVICE_ID_MATCH_INT_CLASS | USB_DEVICE_ID_MATCH_INT_SUBCLASS | USB_DEVICE_ID_MATCH_INT_PROTOCOL) + (USB_DEVICE_ID_MATCH_INT_CLASS | \ + USB_DEVICE_ID_MATCH_INT_SUBCLASS | \ + USB_DEVICE_ID_MATCH_INT_PROTOCOL) /** * USB_DEVICE - macro used to describe a specific usb device @@ -467,9 +480,11 @@ static inline int usb_make_path (struct usb_device *dev, char *buf, size_t size) * specific device. */ #define USB_DEVICE(vend,prod) \ - .match_flags = USB_DEVICE_ID_MATCH_DEVICE, .idVendor = (vend), .idProduct = (prod) + .match_flags = USB_DEVICE_ID_MATCH_DEVICE, .idVendor = (vend), \ + .idProduct = (prod) /** - * USB_DEVICE_VER - macro used to describe a specific usb device with a version range + * USB_DEVICE_VER - macro used to describe a specific usb device with a + * version range * @vend: the 16 bit USB Vendor ID * @prod: the 16 bit USB Product ID * @lo: the bcdDevice_lo value @@ -479,7 +494,9 @@ static inline int usb_make_path (struct usb_device *dev, char *buf, size_t size) * specific device, with a version range. */ #define USB_DEVICE_VER(vend,prod,lo,hi) \ - .match_flags = USB_DEVICE_ID_MATCH_DEVICE_AND_VERSION, .idVendor = (vend), .idProduct = (prod), .bcdDevice_lo = (lo), .bcdDevice_hi = (hi) + .match_flags = USB_DEVICE_ID_MATCH_DEVICE_AND_VERSION, \ + .idVendor = (vend), .idProduct = (prod), \ + .bcdDevice_lo = (lo), .bcdDevice_hi = (hi) /** * USB_DEVICE_INFO - macro used to describe a class of usb devices @@ -491,7 +508,8 @@ static inline int usb_make_path (struct usb_device *dev, char *buf, size_t size) * specific class of devices. */ #define USB_DEVICE_INFO(cl,sc,pr) \ - .match_flags = USB_DEVICE_ID_MATCH_DEV_INFO, .bDeviceClass = (cl), .bDeviceSubClass = (sc), .bDeviceProtocol = (pr) + .match_flags = USB_DEVICE_ID_MATCH_DEV_INFO, .bDeviceClass = (cl), \ + .bDeviceSubClass = (sc), .bDeviceProtocol = (pr) /** * USB_INTERFACE_INFO - macro used to describe a class of usb interfaces @@ -503,9 +521,10 @@ static inline int usb_make_path (struct usb_device *dev, char *buf, size_t size) * specific class of interfaces. */ #define USB_INTERFACE_INFO(cl,sc,pr) \ - .match_flags = USB_DEVICE_ID_MATCH_INT_INFO, .bInterfaceClass = (cl), .bInterfaceSubClass = (sc), .bInterfaceProtocol = (pr) + .match_flags = USB_DEVICE_ID_MATCH_INT_INFO, .bInterfaceClass = (cl), \ + .bInterfaceSubClass = (sc), .bInterfaceProtocol = (pr) -/* -------------------------------------------------------------------------- */ +/* ----------------------------------------------------------------------- */ /** * struct usb_driver - identifies USB driver to usbcore @@ -558,7 +577,8 @@ struct usb_driver { void (*disconnect) (struct usb_interface *intf); - int (*ioctl) (struct usb_interface *intf, unsigned int code, void *buf); + int (*ioctl) (struct usb_interface *intf, unsigned int code, + void *buf); int (*suspend) (struct usb_interface *intf, pm_message_t message); int (*resume) (struct usb_interface *intf); @@ -601,7 +621,7 @@ extern void usb_deregister_dev(struct usb_interface *intf, extern int usb_disabled(void); -/* -------------------------------------------------------------------------- */ +/* ----------------------------------------------------------------------- */ /* * URB support, for asynchronous request completions @@ -611,12 +631,14 @@ extern int usb_disabled(void); * urb->transfer_flags: */ #define URB_SHORT_NOT_OK 0x0001 /* report short reads as errors */ -#define URB_ISO_ASAP 0x0002 /* iso-only, urb->start_frame ignored */ +#define URB_ISO_ASAP 0x0002 /* iso-only, urb->start_frame + * ignored */ #define URB_NO_TRANSFER_DMA_MAP 0x0004 /* urb->transfer_dma valid on submit */ #define URB_NO_SETUP_DMA_MAP 0x0008 /* urb->setup_dma valid on submit */ #define URB_NO_FSBR 0x0020 /* UHCI-specific */ -#define URB_ZERO_PACKET 0x0040 /* Finish bulk OUTs with short packet */ -#define URB_NO_INTERRUPT 0x0080 /* HINT: no non-error interrupt needed */ +#define URB_ZERO_PACKET 0x0040 /* Finish bulk OUT with short packet */ +#define URB_NO_INTERRUPT 0x0080 /* HINT: no non-error interrupt + * needed */ struct usb_iso_packet_descriptor { unsigned int offset; @@ -804,7 +826,8 @@ struct urb u8 reject; /* submissions will fail */ /* public, documented fields in the urb that can be used by drivers */ - struct list_head urb_list; /* list head for use by the urb owner */ + struct list_head urb_list; /* list head for use by the urb's + * current owner */ struct usb_device *dev; /* (in) pointer to associated device */ unsigned int pipe; /* (in) pipe information */ int status; /* (return) non-ISO status */ @@ -817,14 +840,16 @@ struct urb dma_addr_t setup_dma; /* (in) dma addr for setup_packet */ int start_frame; /* (modify) start frame (ISO) */ int number_of_packets; /* (in) number of ISO packets */ - int interval; /* (modify) transfer interval (INT/ISO) */ + int interval; /* (modify) transfer interval + * (INT/ISO) */ int error_count; /* (return) number of ISO errors */ void *context; /* (in) context for completion */ usb_complete_t complete; /* (in) completion routine */ - struct usb_iso_packet_descriptor iso_frame_desc[0]; /* (in) ISO ONLY */ + struct usb_iso_packet_descriptor iso_frame_desc[0]; + /* (in) ISO ONLY */ }; -/* -------------------------------------------------------------------------- */ +/* ----------------------------------------------------------------------- */ /** * usb_fill_control_urb - initializes a control urb @@ -1049,7 +1074,7 @@ void usb_sg_cancel (struct usb_sg_request *io); void usb_sg_wait (struct usb_sg_request *io); -/* -------------------------------------------------------------------------- */ +/* ----------------------------------------------------------------------- */ /* * For various legacy reasons, Linux has a small cookie that's paired with @@ -1090,23 +1115,34 @@ void usb_sg_wait (struct usb_sg_request *io); /* The D0/D1 toggle bits ... USE WITH CAUTION (they're almost hcd-internal) */ #define usb_gettoggle(dev, ep, out) (((dev)->toggle[out] >> (ep)) & 1) #define usb_dotoggle(dev, ep, out) ((dev)->toggle[out] ^= (1 << (ep))) -#define usb_settoggle(dev, ep, out, bit) ((dev)->toggle[out] = ((dev)->toggle[out] & ~(1 << (ep))) | ((bit) << (ep))) +#define usb_settoggle(dev, ep, out, bit) \ + ((dev)->toggle[out] = ((dev)->toggle[out] & ~(1 << (ep))) | \ + ((bit) << (ep))) -static inline unsigned int __create_pipe(struct usb_device *dev, unsigned int endpoint) +static inline unsigned int __create_pipe(struct usb_device *dev, + unsigned int endpoint) { return (dev->devnum << 8) | (endpoint << 15); } /* Create various pipes... */ -#define usb_sndctrlpipe(dev,endpoint) ((PIPE_CONTROL << 30) | __create_pipe(dev,endpoint)) -#define usb_rcvctrlpipe(dev,endpoint) ((PIPE_CONTROL << 30) | __create_pipe(dev,endpoint) | USB_DIR_IN) -#define usb_sndisocpipe(dev,endpoint) ((PIPE_ISOCHRONOUS << 30) | __create_pipe(dev,endpoint)) -#define usb_rcvisocpipe(dev,endpoint) ((PIPE_ISOCHRONOUS << 30) | __create_pipe(dev,endpoint) | USB_DIR_IN) -#define usb_sndbulkpipe(dev,endpoint) ((PIPE_BULK << 30) | __create_pipe(dev,endpoint)) -#define usb_rcvbulkpipe(dev,endpoint) ((PIPE_BULK << 30) | __create_pipe(dev,endpoint) | USB_DIR_IN) -#define usb_sndintpipe(dev,endpoint) ((PIPE_INTERRUPT << 30) | __create_pipe(dev,endpoint)) -#define usb_rcvintpipe(dev,endpoint) ((PIPE_INTERRUPT << 30) | __create_pipe(dev,endpoint) | USB_DIR_IN) +#define usb_sndctrlpipe(dev,endpoint) \ + ((PIPE_CONTROL << 30) | __create_pipe(dev,endpoint)) +#define usb_rcvctrlpipe(dev,endpoint) \ + ((PIPE_CONTROL << 30) | __create_pipe(dev,endpoint) | USB_DIR_IN) +#define usb_sndisocpipe(dev,endpoint) \ + ((PIPE_ISOCHRONOUS << 30) | __create_pipe(dev,endpoint)) +#define usb_rcvisocpipe(dev,endpoint) \ + ((PIPE_ISOCHRONOUS << 30) | __create_pipe(dev,endpoint) | USB_DIR_IN) +#define usb_sndbulkpipe(dev,endpoint) \ + ((PIPE_BULK << 30) | __create_pipe(dev,endpoint)) +#define usb_rcvbulkpipe(dev,endpoint) \ + ((PIPE_BULK << 30) | __create_pipe(dev,endpoint) | USB_DIR_IN) +#define usb_sndintpipe(dev,endpoint) \ + ((PIPE_INTERRUPT << 30) | __create_pipe(dev,endpoint)) +#define usb_rcvintpipe(dev,endpoint) \ + ((PIPE_INTERRUPT << 30) | __create_pipe(dev,endpoint) | USB_DIR_IN) /*-------------------------------------------------------------------------*/ @@ -1130,7 +1166,7 @@ usb_maxpacket(struct usb_device *udev, int pipe, int is_out) return le16_to_cpu(ep->desc.wMaxPacketSize); } -/* -------------------------------------------------------------------------- */ +/* ----------------------------------------------------------------------- */ /* Events from the usb core */ #define USB_DEVICE_ADD 0x0001 @@ -1141,14 +1177,18 @@ extern void usb_register_notify(struct notifier_block *nb); extern void usb_unregister_notify(struct notifier_block *nb); #ifdef DEBUG -#define dbg(format, arg...) printk(KERN_DEBUG "%s: " format "\n" , __FILE__ , ## arg) +#define dbg(format, arg...) printk(KERN_DEBUG "%s: " format "\n" , \ + __FILE__ , ## arg) #else #define dbg(format, arg...) do {} while (0) #endif -#define err(format, arg...) printk(KERN_ERR "%s: " format "\n" , __FILE__ , ## arg) -#define info(format, arg...) printk(KERN_INFO "%s: " format "\n" , __FILE__ , ## arg) -#define warn(format, arg...) printk(KERN_WARNING "%s: " format "\n" , __FILE__ , ## arg) +#define err(format, arg...) printk(KERN_ERR "%s: " format "\n" , \ + __FILE__ , ## arg) +#define info(format, arg...) printk(KERN_INFO "%s: " format "\n" , \ + __FILE__ , ## arg) +#define warn(format, arg...) printk(KERN_WARNING "%s: " format "\n" , \ + __FILE__ , ## arg) #endif /* __KERNEL__ */ -- cgit v1.2.3 From 4f62efe67f077db17dad03a1d4c9665000a3eb45 Mon Sep 17 00:00:00 2001 From: Alan Stern Date: Mon, 24 Oct 2005 16:24:14 -0400 Subject: [PATCH] usbcore: Fix handling of sysfs strings and other attributes This patch (as592) makes a few small improvements to the way device strings are handled, and it fixes some bugs in a couple of other sysfs attribute routines. (Look at show_configuration_string() to see what I mean.) Signed-off-by: Alan Stern Signed-off-by: Greg Kroah-Hartman --- drivers/usb/core/config.c | 10 ++++++---- drivers/usb/core/hub.c | 22 ++++----------------- drivers/usb/core/message.c | 48 +++++++++++++++++++++++++++++----------------- drivers/usb/core/sysfs.c | 36 +++++++++++++++------------------- drivers/usb/core/usb.h | 1 + include/linux/usb.h | 10 ++++++---- 6 files changed, 62 insertions(+), 65 deletions(-) (limited to 'include/linux') diff --git a/drivers/usb/core/config.c b/drivers/usb/core/config.c index 63f374e62db2..993019500cc3 100644 --- a/drivers/usb/core/config.c +++ b/drivers/usb/core/config.c @@ -112,8 +112,12 @@ void usb_release_interface_cache(struct kref *ref) struct usb_interface_cache *intfc = ref_to_usb_interface_cache(ref); int j; - for (j = 0; j < intfc->num_altsetting; j++) - kfree(intfc->altsetting[j].endpoint); + for (j = 0; j < intfc->num_altsetting; j++) { + struct usb_host_interface *alt = &intfc->altsetting[j]; + + kfree(alt->endpoint); + kfree(alt->string); + } kfree(intfc); } @@ -420,8 +424,6 @@ void usb_destroy_configuration(struct usb_device *dev) struct usb_host_config *cf = &dev->config[c]; kfree(cf->string); - cf->string = NULL; - for (i = 0; i < cf->desc.bNumInterfaces; i++) { if (cf->intf_cache[i]) kref_put(&cf->intf_cache[i]->ref, diff --git a/drivers/usb/core/hub.c b/drivers/usb/core/hub.c index 8ba5854e5387..1bacb374b007 100644 --- a/drivers/usb/core/hub.c +++ b/drivers/usb/core/hub.c @@ -1204,21 +1204,6 @@ static inline void show_string(struct usb_device *udev, char *id, char *string) {} #endif -static void get_string(struct usb_device *udev, char **string, int index) -{ - char *buf; - - if (!index) - return; - buf = kmalloc(256, GFP_KERNEL); - if (!buf) - return; - if (usb_string(udev, index, buf, 256) > 0) - *string = buf; - else - kfree(buf); -} - #ifdef CONFIG_USB_OTG #include "otg_whitelist.h" @@ -1257,9 +1242,10 @@ int usb_new_device(struct usb_device *udev) } /* read the standard strings and cache them if present */ - get_string(udev, &udev->product, udev->descriptor.iProduct); - get_string(udev, &udev->manufacturer, udev->descriptor.iManufacturer); - get_string(udev, &udev->serial, udev->descriptor.iSerialNumber); + udev->product = usb_cache_string(udev, udev->descriptor.iProduct); + udev->manufacturer = usb_cache_string(udev, + udev->descriptor.iManufacturer); + udev->serial = usb_cache_string(udev, udev->descriptor.iSerialNumber); /* Tell the world! */ dev_dbg(&udev->dev, "new device strings: Mfr=%d, Product=%d, " diff --git a/drivers/usb/core/message.c b/drivers/usb/core/message.c index 3519f317898e..644a3d4f12aa 100644 --- a/drivers/usb/core/message.c +++ b/drivers/usb/core/message.c @@ -787,6 +787,31 @@ int usb_string(struct usb_device *dev, int index, char *buf, size_t size) return err; } +/** + * usb_cache_string - read a string descriptor and cache it for later use + * @udev: the device whose string descriptor is being read + * @index: the descriptor index + * + * Returns a pointer to a kmalloc'ed buffer containing the descriptor string, + * or NULL if the index is 0 or the string could not be read. + */ +char *usb_cache_string(struct usb_device *udev, int index) +{ + char *buf; + char *smallbuf = NULL; + int len; + + if (index > 0 && (buf = kmalloc(256, GFP_KERNEL)) != NULL) { + if ((len = usb_string(udev, index, buf, 256)) > 0) { + if ((smallbuf = kmalloc(++len, GFP_KERNEL)) == NULL) + return buf; + memcpy(smallbuf, buf, len); + } + kfree(buf); + } + return smallbuf; +} + /* * usb_get_device_descriptor - (re)reads the device descriptor (usbcore) * @dev: the device whose device descriptor is being updated @@ -1008,8 +1033,6 @@ void usb_disable_device(struct usb_device *dev, int skip_ep0) dev_dbg (&dev->dev, "unregistering interface %s\n", interface->dev.bus_id); usb_remove_sysfs_intf_files(interface); - kfree(interface->cur_altsetting->string); - interface->cur_altsetting->string = NULL; device_del (&interface->dev); } @@ -1422,12 +1445,9 @@ free_interfaces: } kfree(new_interfaces); - if ((cp->desc.iConfiguration) && - (cp->string == NULL)) { - cp->string = kmalloc(256, GFP_KERNEL); - if (cp->string) - usb_string(dev, cp->desc.iConfiguration, cp->string, 256); - } + if (cp->string == NULL) + cp->string = usb_cache_string(dev, + cp->desc.iConfiguration); /* Now that all the interfaces are set up, register them * to trigger binding of drivers to interfaces. probe() @@ -1437,13 +1457,12 @@ free_interfaces: */ for (i = 0; i < nintf; ++i) { struct usb_interface *intf = cp->interface[i]; - struct usb_interface_descriptor *desc; + struct usb_host_interface *alt = intf->cur_altsetting; - desc = &intf->altsetting [0].desc; dev_dbg (&dev->dev, "adding %s (config #%d, interface %d)\n", intf->dev.bus_id, configuration, - desc->bInterfaceNumber); + alt->desc.bInterfaceNumber); ret = device_add (&intf->dev); if (ret != 0) { dev_err(&dev->dev, @@ -1452,13 +1471,6 @@ free_interfaces: ret); continue; } - if ((intf->cur_altsetting->desc.iInterface) && - (intf->cur_altsetting->string == NULL)) { - intf->cur_altsetting->string = kmalloc(256, GFP_KERNEL); - if (intf->cur_altsetting->string) - usb_string(dev, intf->cur_altsetting->desc.iInterface, - intf->cur_altsetting->string, 256); - } usb_create_sysfs_intf_files (intf); } } diff --git a/drivers/usb/core/sysfs.c b/drivers/usb/core/sysfs.c index 4cca77cf0c48..edd83e014452 100644 --- a/drivers/usb/core/sysfs.c +++ b/drivers/usb/core/sysfs.c @@ -249,18 +249,12 @@ static ssize_t show_configuration_string(struct device *dev, { struct usb_device *udev; struct usb_host_config *actconfig; - int len; udev = to_usb_device (dev); actconfig = udev->actconfig; if ((!actconfig) || (!actconfig->string)) return 0; - len = sprintf(buf, actconfig->string, PAGE_SIZE); - if (len < 0) - return 0; - buf[len] = '\n'; - buf[len+1] = 0; - return len+1; + return sprintf(buf, "%s\n", actconfig->string); } static DEVICE_ATTR(configuration, S_IRUGO, show_configuration_string, NULL); @@ -291,15 +285,9 @@ static ssize_t show_##name(struct device *dev, \ struct device_attribute *attr, char *buf) \ { \ struct usb_device *udev; \ - int len; \ \ udev = to_usb_device (dev); \ - len = snprintf(buf, 256, "%s", udev->name); \ - if (len < 0) \ - return 0; \ - buf[len] = '\n'; \ - buf[len+1] = 0; \ - return len+1; \ + return sprintf(buf, "%s\n", udev->name); \ } \ static DEVICE_ATTR(name, S_IRUGO, show_##name, NULL); @@ -449,11 +437,11 @@ void usb_remove_sysfs_dev_files (struct usb_device *udev) usb_remove_ep_files(&udev->ep0); sysfs_remove_group(&dev->kobj, &dev_attr_grp); - if (udev->descriptor.iManufacturer) + if (udev->manufacturer) device_remove_file(dev, &dev_attr_manufacturer); - if (udev->descriptor.iProduct) + if (udev->product) device_remove_file(dev, &dev_attr_product); - if (udev->descriptor.iSerialNumber) + if (udev->serial) device_remove_file(dev, &dev_attr_serial); device_remove_file (dev, &dev_attr_configuration); } @@ -535,7 +523,8 @@ static struct attribute_group intf_attr_grp = { .attrs = intf_attrs, }; -static inline void usb_create_intf_ep_files(struct usb_interface *intf) +static inline void usb_create_intf_ep_files(struct usb_interface *intf, + struct usb_device *udev) { struct usb_host_interface *iface_desc; int i; @@ -543,7 +532,7 @@ static inline void usb_create_intf_ep_files(struct usb_interface *intf) iface_desc = intf->cur_altsetting; for (i = 0; i < iface_desc->desc.bNumEndpoints; ++i) usb_create_ep_files(&intf->dev.kobj, &iface_desc->endpoint[i], - interface_to_usbdev(intf)); + udev); } static inline void usb_remove_intf_ep_files(struct usb_interface *intf) @@ -558,11 +547,16 @@ static inline void usb_remove_intf_ep_files(struct usb_interface *intf) void usb_create_sysfs_intf_files (struct usb_interface *intf) { + struct usb_device *udev = interface_to_usbdev(intf); + struct usb_host_interface *alt = intf->cur_altsetting; + sysfs_create_group(&intf->dev.kobj, &intf_attr_grp); - if (intf->cur_altsetting->string) + if (alt->string == NULL) + alt->string = usb_cache_string(udev, alt->desc.iInterface); + if (alt->string) device_create_file(&intf->dev, &dev_attr_interface); - usb_create_intf_ep_files(intf); + usb_create_intf_ep_files(intf, udev); } void usb_remove_sysfs_intf_files (struct usb_interface *intf) diff --git a/drivers/usb/core/usb.h b/drivers/usb/core/usb.h index 888dbe443695..1c4a68499dce 100644 --- a/drivers/usb/core/usb.h +++ b/drivers/usb/core/usb.h @@ -13,6 +13,7 @@ extern void usb_disable_device (struct usb_device *dev, int skip_ep0); extern int usb_get_device_descriptor(struct usb_device *dev, unsigned int size); +extern char *usb_cache_string(struct usb_device *udev, int index); extern int usb_set_configuration(struct usb_device *dev, int configuration); extern void usb_lock_all_devices(void); diff --git a/include/linux/usb.h b/include/linux/usb.h index c500d6b5a16d..748d04385256 100644 --- a/include/linux/usb.h +++ b/include/linux/usb.h @@ -231,7 +231,7 @@ struct usb_interface_cache { struct usb_host_config { struct usb_config_descriptor desc; - char *string; + char *string; /* iConfiguration string, if present */ /* the interfaces associated with this configuration, * stored in no particular order */ struct usb_interface *interface[USB_MAXINTERFACES]; @@ -351,9 +351,11 @@ struct usb_device { int have_langid; /* whether string_langid is valid */ int string_langid; /* language ID for strings */ - char *product; - char *manufacturer; - char *serial; /* static strings from the device */ + /* static strings from the device */ + char *product; /* iProduct string, if present */ + char *manufacturer; /* iManufacturer string, if present */ + char *serial; /* iSerialNumber string, if present */ + struct list_head filelist; struct class_device *class_dev; struct dentry *usbfs_dentry; /* usbfs dentry entry for the device */ -- cgit v1.2.3 From 9a7834d06d553d02cc6e659e94772f69a8b5367f Mon Sep 17 00:00:00 2001 From: Andrew Morton Date: Sun, 23 Oct 2005 23:02:20 -0700 Subject: [PATCH] USB: fix pm patches with CONFIG_PM off part 2 With CONFIG_PM=n: drivers/built-in.o(.text+0x1098c): In function `hub_thread': drivers/usb/core/hub.c:2673: undefined reference to `.dpm_runtime_resume' drivers/built-in.o(.text+0x10998):drivers/usb/core/hub.c:2674: undefined reference to `.dpm_runtime_resume' Please, never ever ever put extern decls into .c files. Use the darn header files :( Cc: David Brownell Signed-off-by: Andrew Morton Signed-off-by: Greg Kroah-Hartman --- drivers/base/power/power.h | 13 ------------- drivers/usb/core/hub.c | 2 -- include/linux/pm.h | 13 +++++++++++++ 3 files changed, 13 insertions(+), 15 deletions(-) (limited to 'include/linux') diff --git a/drivers/base/power/power.h b/drivers/base/power/power.h index 2e700d795cf1..fb3d35a9e101 100644 --- a/drivers/base/power/power.h +++ b/drivers/base/power/power.h @@ -67,9 +67,6 @@ extern int suspend_device(struct device *, pm_message_t); * runtime.c */ -extern int dpm_runtime_suspend(struct device *, pm_message_t); -extern void dpm_runtime_resume(struct device *); - #else /* CONFIG_PM */ @@ -82,14 +79,4 @@ static inline void device_pm_remove(struct device * dev) } -static inline int dpm_runtime_suspend(struct device * dev, pm_message_t state) -{ - return 0; -} - -static inline void dpm_runtime_resume(struct device * dev) -{ - -} - #endif diff --git a/drivers/usb/core/hub.c b/drivers/usb/core/hub.c index 1bacb374b007..9660a8909b7c 100644 --- a/drivers/usb/core/hub.c +++ b/drivers/usb/core/hub.c @@ -2652,8 +2652,6 @@ static void hub_events(void) * stub "device" node was never suspended. */ if (i) { - extern void dpm_runtime_resume(struct device *); - dpm_runtime_resume(&hdev->dev); dpm_runtime_resume(&intf->dev); } diff --git a/include/linux/pm.h b/include/linux/pm.h index 83bae1cbe55b..c61d5de837ef 100644 --- a/include/linux/pm.h +++ b/include/linux/pm.h @@ -243,6 +243,9 @@ extern int device_suspend(pm_message_t state); #define device_may_wakeup(dev) \ (device_can_wakeup(dev) && (dev)->power.should_wakeup) +extern int dpm_runtime_suspend(struct device *, pm_message_t); +extern void dpm_runtime_resume(struct device *); + #else /* !CONFIG_PM */ static inline int device_suspend(pm_message_t state) @@ -253,6 +256,16 @@ static inline int device_suspend(pm_message_t state) #define device_set_wakeup_enable(dev,val) do{}while(0) #define device_may_wakeup(dev) (0) +static inline int dpm_runtime_suspend(struct device * dev, pm_message_t state) +{ + return 0; +} + +static inline void dpm_runtime_resume(struct device * dev) +{ + +} + #endif /* changes to device_may_wakeup take effect on the next pm state change. -- cgit v1.2.3 From 360ac8e2f1a38c3497739636c3b702352d1ad0ae Mon Sep 17 00:00:00 2001 From: Stephen Hemminger Date: Tue, 25 Oct 2005 15:03:41 -0700 Subject: [ETH]: ether address compare Expose faster ether compare for use by protocols and other driver. And change name to be more consistent with other ether address manipulation routines in same file Signed-off-by: Stephen Hemminger Signed-off-by: Arnaldo Carvalho de Melo --- include/linux/etherdevice.h | 16 ++++++++++++++++ net/ethernet/eth.c | 17 ++--------------- 2 files changed, 18 insertions(+), 15 deletions(-) (limited to 'include/linux') diff --git a/include/linux/etherdevice.h b/include/linux/etherdevice.h index 4522c7186bf3..cc84934f9059 100644 --- a/include/linux/etherdevice.h +++ b/include/linux/etherdevice.h @@ -104,6 +104,22 @@ static inline void random_ether_addr(u8 *addr) addr [0] &= 0xfe; /* clear multicast bit */ addr [0] |= 0x02; /* set local assignment bit (IEEE802) */ } + +/** + * compare_ether_addr - Compare two Ethernet addresses + * @addr1: Pointer to a six-byte array containing the Ethernet address + * @addr2 Pointer other six-byte array containing the Ethernet address + * + * Compare two ethernet addresses, returns 0 if equal + */ +static inline unsigned compare_ether_addr(const u8 *_a, const u8 *_b) +{ + const u16 *a = (const u16 *) _a; + const u16 *b = (const u16 *) _b; + + BUILD_BUG_ON(ETH_ALEN != 6); + return ((a[0] ^ b[0]) | (a[1] ^ b[1]) | (a[2] ^ b[2])) != 0; +} #endif /* __KERNEL__ */ #endif /* _LINUX_ETHERDEVICE_H */ diff --git a/net/ethernet/eth.c b/net/ethernet/eth.c index 68a5ca866442..e24577367274 100644 --- a/net/ethernet/eth.c +++ b/net/ethernet/eth.c @@ -146,19 +146,6 @@ int eth_rebuild_header(struct sk_buff *skb) return 0; } -static inline unsigned int compare_eth_addr(const unsigned char *__a, const unsigned char *__b) -{ - const unsigned short *dest = (unsigned short *) __a; - const unsigned short *devaddr = (unsigned short *) __b; - unsigned int res; - - BUILD_BUG_ON(ETH_ALEN != 6); - res = ((dest[0] ^ devaddr[0]) | - (dest[1] ^ devaddr[1]) | - (dest[2] ^ devaddr[2])) != 0; - - return res; -} /* * Determine the packet's protocol ID. The rule here is that we @@ -176,7 +163,7 @@ __be16 eth_type_trans(struct sk_buff *skb, struct net_device *dev) eth = eth_hdr(skb); if (*eth->h_dest&1) { - if (!compare_eth_addr(eth->h_dest, dev->broadcast)) + if (!compare_ether_addr(eth->h_dest, dev->broadcast)) skb->pkt_type = PACKET_BROADCAST; else skb->pkt_type = PACKET_MULTICAST; @@ -191,7 +178,7 @@ __be16 eth_type_trans(struct sk_buff *skb, struct net_device *dev) */ else if(1 /*dev->flags&IFF_PROMISC*/) { - if (unlikely(compare_eth_addr(eth->h_dest, dev->dev_addr))) + if (unlikely(compare_ether_addr(eth->h_dest, dev->dev_addr))) skb->pkt_type = PACKET_OTHERHOST; } -- cgit v1.2.3 From bdf21b18b4abf983db38f04ef7fec88f47389867 Mon Sep 17 00:00:00 2001 From: Pete Popov Date: Thu, 14 Jul 2005 17:47:57 +0000 Subject: Philips PNX8550 support: MIPS32-like core with 2 Trimedias on it. Signed-off-by: Ralf Baechle --- arch/mips/Kconfig | 21 ++ arch/mips/Makefile | 14 ++ arch/mips/kernel/cpu-probe.c | 19 ++ arch/mips/kernel/proc.c | 3 +- arch/mips/kernel/time.c | 3 + arch/mips/mm/tlbex.c | 1 + arch/mips/pci/Makefile | 1 + arch/mips/pci/fixup-pnx8550.c | 57 +++++ arch/mips/pci/ops-pnx8550.c | 284 +++++++++++++++++++++ arch/mips/philips/pnx8550/common/Kconfig | 1 + arch/mips/philips/pnx8550/common/Makefile | 27 ++ arch/mips/philips/pnx8550/common/gdb_hook.c | 109 ++++++++ arch/mips/philips/pnx8550/common/int.c | 293 ++++++++++++++++++++++ arch/mips/philips/pnx8550/common/mipsIRQ.S | 76 ++++++ arch/mips/philips/pnx8550/common/pci.c | 133 ++++++++++ arch/mips/philips/pnx8550/common/platform.c | 135 ++++++++++ arch/mips/philips/pnx8550/common/proc.c | 113 +++++++++ arch/mips/philips/pnx8550/common/prom.c | 138 ++++++++++ arch/mips/philips/pnx8550/common/reset.c | 49 ++++ arch/mips/philips/pnx8550/common/setup.c | 149 +++++++++++ arch/mips/philips/pnx8550/common/time.c | 105 ++++++++ arch/mips/philips/pnx8550/jbs/Makefile | 4 + arch/mips/philips/pnx8550/jbs/board_setup.c | 65 +++++ arch/mips/philips/pnx8550/jbs/init.c | 57 +++++ arch/mips/philips/pnx8550/jbs/irqmap.c | 36 +++ include/asm-mips/bootinfo.h | 1 + include/asm-mips/cpu.h | 4 +- include/asm-mips/mach-pnx8550/cm.h | 43 ++++ include/asm-mips/mach-pnx8550/glb.h | 86 +++++++ include/asm-mips/mach-pnx8550/int.h | 140 +++++++++++ include/asm-mips/mach-pnx8550/kernel-entry-init.h | 262 +++++++++++++++++++ include/asm-mips/mach-pnx8550/nand.h | 121 +++++++++ include/asm-mips/mach-pnx8550/pci.h | 185 ++++++++++++++ include/asm-mips/mach-pnx8550/uart.h | 16 ++ include/asm-mips/mach-pnx8550/usb.h | 32 +++ include/asm-mips/mipsregs.h | 12 + include/linux/serial_core.h | 3 +- include/linux/serial_ip3106.h | 81 ++++++ 38 files changed, 2876 insertions(+), 3 deletions(-) create mode 100644 arch/mips/pci/fixup-pnx8550.c create mode 100644 arch/mips/pci/ops-pnx8550.c create mode 100644 arch/mips/philips/pnx8550/common/Kconfig create mode 100644 arch/mips/philips/pnx8550/common/Makefile create mode 100644 arch/mips/philips/pnx8550/common/gdb_hook.c create mode 100644 arch/mips/philips/pnx8550/common/int.c create mode 100644 arch/mips/philips/pnx8550/common/mipsIRQ.S create mode 100644 arch/mips/philips/pnx8550/common/pci.c create mode 100644 arch/mips/philips/pnx8550/common/platform.c create mode 100644 arch/mips/philips/pnx8550/common/proc.c create mode 100644 arch/mips/philips/pnx8550/common/prom.c create mode 100644 arch/mips/philips/pnx8550/common/reset.c create mode 100644 arch/mips/philips/pnx8550/common/setup.c create mode 100644 arch/mips/philips/pnx8550/common/time.c create mode 100644 arch/mips/philips/pnx8550/jbs/Makefile create mode 100644 arch/mips/philips/pnx8550/jbs/board_setup.c create mode 100644 arch/mips/philips/pnx8550/jbs/init.c create mode 100644 arch/mips/philips/pnx8550/jbs/irqmap.c create mode 100644 include/asm-mips/mach-pnx8550/cm.h create mode 100644 include/asm-mips/mach-pnx8550/glb.h create mode 100644 include/asm-mips/mach-pnx8550/int.h create mode 100644 include/asm-mips/mach-pnx8550/kernel-entry-init.h create mode 100644 include/asm-mips/mach-pnx8550/nand.h create mode 100644 include/asm-mips/mach-pnx8550/pci.h create mode 100644 include/asm-mips/mach-pnx8550/uart.h create mode 100644 include/asm-mips/mach-pnx8550/usb.h create mode 100644 include/linux/serial_ip3106.h (limited to 'include/linux') diff --git a/arch/mips/Kconfig b/arch/mips/Kconfig index 4ef015f580f9..f6b25ae1861b 100644 --- a/arch/mips/Kconfig +++ b/arch/mips/Kconfig @@ -489,6 +489,16 @@ config HYPERTRANSPORT bool "Hypertransport Support for PMC-Sierra Yosemite" depends on PMC_YOSEMITE +config PNX8550_V2PCI + bool "Support for Philips PNX8550 based Viper2-PCI board" + select PNX8550 + select SYS_SUPPORTS_LITTLE_ENDIAN + +config PNX8550_JBS + bool "Support for Philips PNX8550 based JBS board" + select PNX8550 + select SYS_SUPPORTS_LITTLE_ENDIAN + config DDB5074 bool "Support for NEC DDB Vrc-5074 (EXPERIMENTAL)" depends on EXPERIMENTAL @@ -827,6 +837,7 @@ config TOSHIBA_FPCIB0 source "arch/mips/sgi-ip27/Kconfig" source "arch/mips/sibyte/Kconfig" +source "arch/mips/philips/pnx8550/common/Kconfig" config RWSEM_GENERIC_SPINLOCK bool @@ -954,6 +965,16 @@ config ITE_BOARD_GEN depends on MIPS_IVR || MIPS_ITE8172 default y +config PNX8550 + bool + select SOC_PNX8550 + +config SOC_PNX8550 + bool + select SYS_SUPPORTS_32BIT_KERNEL + select DMA_NONCOHERENT + select HW_HAS_PCI + config SWAP_IO_SPACE bool diff --git a/arch/mips/Makefile b/arch/mips/Makefile index 6780d115a7dc..d62787ab9fff 100644 --- a/arch/mips/Makefile +++ b/arch/mips/Makefile @@ -228,6 +228,7 @@ cflags-$(CONFIG_CPU_RM9000) += \ $(call set_gccflags,rm9000,mips4,r5000,mips4,mips2) \ -Wa,--trap + cflags-$(CONFIG_CPU_SB1) += \ $(call set_gccflags,sb1,mips64,r5000,mips4,mips2) \ -Wa,--trap @@ -560,6 +561,19 @@ load-$(CONFIG_CASIO_E55) += 0xffffffff80004000 # load-$(CONFIG_TANBAC_TB022X) += 0xffffffff80000000 +# +# Common Philips PNX8550 +# +core-$(CONFIG_SOC_PNX8550) += arch/mips/philips/pnx8550/common/ +cflags-$(CONFIG_SOC_PNX8550) += -Iinclude/asm-mips/mach-pnx8550 + +# +# Philips PNX8550 JBS board +# +libs-$(CONFIG_PNX8550_JBS) += arch/mips/philips/pnx8550/jbs/ +#cflags-$(CONFIG_PNX8550_JBS) += -Iinclude/asm-mips/mach-pnx8550 +load-$(CONFIG_PNX8550_JBS) += 0xffffffff80060000 + # # SGI IP22 (Indy/Indigo2) # diff --git a/arch/mips/kernel/cpu-probe.c b/arch/mips/kernel/cpu-probe.c index 844126b39ed3..70c8ad9bc8fc 100644 --- a/arch/mips/kernel/cpu-probe.c +++ b/arch/mips/kernel/cpu-probe.c @@ -121,6 +121,7 @@ static inline void check_wait(void) case CPU_24K: case CPU_25KF: case CPU_34K: + case CPU_PR4450: cpu_wait = r4k_wait; printk(" available.\n"); break; @@ -624,6 +625,21 @@ static inline void cpu_probe_sandcraft(struct cpuinfo_mips *c) } } +static inline void cpu_probe_philips(struct cpuinfo_mips *c) +{ + decode_configs(c); + switch (c->processor_id & 0xff00) { + case PRID_IMP_PR4450: + c->cputype = CPU_PR4450; + c->isa_level = MIPS_CPU_ISA_M32; + break; + default: + panic("Unknown Philips Core!"); /* REVISIT: die? */ + break; + } +} + + __init void cpu_probe(void) { struct cpuinfo_mips *c = ¤t_cpu_data; @@ -649,6 +665,9 @@ __init void cpu_probe(void) case PRID_COMP_SANDCRAFT: cpu_probe_sandcraft(c); break; + case PRID_COMP_PHILIPS: + cpu_probe_philips(c); + break; default: c->cputype = CPU_UNKNOWN; } diff --git a/arch/mips/kernel/proc.c b/arch/mips/kernel/proc.c index 1bd40af508ed..e46a92d01d51 100644 --- a/arch/mips/kernel/proc.c +++ b/arch/mips/kernel/proc.c @@ -80,7 +80,8 @@ static const char *cpu_name[] = { [CPU_VR4133] = "NEC VR4133", [CPU_VR4181] = "NEC VR4181", [CPU_VR4181A] = "NEC VR4181A", - [CPU_SR71000] = "Sandcraft SR71000" + [CPU_SR71000] = "Sandcraft SR71000", + [CPU_PR4450] = "Philips PR4450", }; diff --git a/arch/mips/kernel/time.c b/arch/mips/kernel/time.c index fbc153c8f833..a24651dfaaba 100644 --- a/arch/mips/kernel/time.c +++ b/arch/mips/kernel/time.c @@ -11,6 +11,7 @@ * Free Software Foundation; either version 2 of the License, or (at your * option) any later version. */ +#include #include #include #include @@ -112,8 +113,10 @@ static void c0_timer_ack(void) { unsigned int count; +#ifndef CONFIG_SOC_PNX8550 /* pnx8550 resets to zero */ /* Ack this timer interrupt and set the next one. */ expirelo += cycles_per_jiffy; +#endif write_c0_compare(expirelo); /* Check to see if we have missed any timer interrupts. */ diff --git a/arch/mips/mm/tlbex.c b/arch/mips/mm/tlbex.c index c1d394d36f6c..a876ed6cde24 100644 --- a/arch/mips/mm/tlbex.c +++ b/arch/mips/mm/tlbex.c @@ -844,6 +844,7 @@ static __init void build_tlb_write_entry(u32 **p, struct label **l, case CPU_AU1500: case CPU_AU1550: case CPU_AU1200: + case CPU_PR4450: i_nop(p); tlbw(p); break; diff --git a/arch/mips/pci/Makefile b/arch/mips/pci/Makefile index 83d81c9cdc2b..ea8438b81fed 100644 --- a/arch/mips/pci/Makefile +++ b/arch/mips/pci/Makefile @@ -34,6 +34,7 @@ obj-$(CONFIG_MIPS_ITE8172) += fixup-ite8172g.o obj-$(CONFIG_MIPS_IVR) += fixup-ivr.o obj-$(CONFIG_SOC_AU1500) += fixup-au1000.o ops-au1000.o obj-$(CONFIG_SOC_AU1550) += fixup-au1000.o ops-au1000.o +obj-$(CONFIG_SOC_PNX8550) += fixup-pnx8550.o ops-pnx8550.o obj-$(CONFIG_MIPS_MALTA) += fixup-malta.o obj-$(CONFIG_MOMENCO_JAGUAR_ATX)+= fixup-jaguar.o obj-$(CONFIG_MOMENCO_OCELOT) += fixup-ocelot.o pci-ocelot.o diff --git a/arch/mips/pci/fixup-pnx8550.c b/arch/mips/pci/fixup-pnx8550.c new file mode 100644 index 000000000000..4256b3b30b77 --- /dev/null +++ b/arch/mips/pci/fixup-pnx8550.c @@ -0,0 +1,57 @@ +/* + * Philips PNX8550 pci fixups. + * + * Copyright 2005 Embedded Alley Solutions, Inc + * source@embeddealley.com + * + * This program is free software; you can distribute it and/or modify it + * under the terms of the GNU General Public License (Version 2) as + * published by the Free Software Foundation. + * + * This program is distributed in the hope it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License + * for more details. + * + * You should have received a copy of the GNU General Public License along + * with this program; if not, write to the Free Software Foundation, Inc., + * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. + */ +#include +#include +#include +#include + +#include +#include + + +#undef DEBUG +#ifdef DEBUG +#define DBG(x...) printk(x) +#else +#define DBG(x...) +#endif + +extern char irq_tab_jbs[][5]; + +void __init pcibios_fixup_resources(struct pci_dev *dev) +{ + /* no need to fixup IO resources */ +} + +void __init pcibios_fixup(void) +{ + /* nothing to do here */ +} + +int __init pcibios_map_irq(struct pci_dev *dev, u8 slot, u8 pin) +{ + return irq_tab_jbs[slot][pin]; +} + +/* Do platform specific device initialization at pci_enable_device() time */ +int pcibios_plat_dev_init(struct pci_dev *dev) +{ + return 0; +} diff --git a/arch/mips/pci/ops-pnx8550.c b/arch/mips/pci/ops-pnx8550.c new file mode 100644 index 000000000000..454b65cc3354 --- /dev/null +++ b/arch/mips/pci/ops-pnx8550.c @@ -0,0 +1,284 @@ +/* + * + * BRIEF MODULE DESCRIPTION + * + * 2.6 port, Embedded Alley Solutions, Inc + * + * Based on: + * Author: source@mvista.com + * + * This program is free software; you can distribute it and/or modify it + * under the terms of the GNU General Public License (Version 2) as + * published by the Free Software Foundation. + * + * This program is distributed in the hope it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License + * for more details. + * + * You should have received a copy of the GNU General Public License along + * with this program; if not, write to the Free Software Foundation, Inc., + * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. + */ +#include +#include +#include +#include +#include +#include + +#include +#include +#include + + +static inline void clear_status(void) +{ + unsigned long pci_stat; + + pci_stat = inl(PCI_BASE | PCI_GPPM_STATUS); + outl(pci_stat, PCI_BASE | PCI_GPPM_ICLR); +} + +static inline unsigned int +calc_cfg_addr(struct pci_bus *bus, unsigned int devfn, int where) +{ + unsigned int addr; + + addr = ((bus->number > 0) ? (((bus->number & 0xff) << PCI_CFG_BUS_SHIFT) | 1) : 0); + addr |= ((devfn & 0xff) << PCI_CFG_FUNC_SHIFT) | (where & 0xfc); + + return addr; +} + +static int +config_access(unsigned int pci_cmd, struct pci_bus *bus, unsigned int devfn, int where, unsigned int pci_mode, unsigned int *val) +{ + unsigned int flags; + unsigned long loops = 0; + unsigned long ioaddr = calc_cfg_addr(bus, devfn, where); + + local_irq_save(flags); + /*Clear pending interrupt status */ + if (inl(PCI_BASE | PCI_GPPM_STATUS)) { + clear_status(); + while (!(inl(PCI_BASE | PCI_GPPM_STATUS) == 0)) ; + } + + outl(ioaddr, PCI_BASE | PCI_GPPM_ADDR); + + if ((pci_cmd == PCI_CMD_IOW) || (pci_cmd == PCI_CMD_CONFIG_WRITE)) + outl(*val, PCI_BASE | PCI_GPPM_WDAT); + + outl(INIT_PCI_CYCLE | pci_cmd | (pci_mode & PCI_BYTE_ENABLE_MASK), + PCI_BASE | PCI_GPPM_CTRL); + + loops = + ((loops_per_jiffy * + PCI_IO_JIFFIES_TIMEOUT) >> (PCI_IO_JIFFIES_SHIFT)); + while (1) { + if (inl(PCI_BASE | PCI_GPPM_STATUS) & GPPM_DONE) { + if ((pci_cmd == PCI_CMD_IOR) || + (pci_cmd == PCI_CMD_CONFIG_READ)) + *val = inl(PCI_BASE | PCI_GPPM_RDAT); + clear_status(); + local_irq_restore(flags); + return PCIBIOS_SUCCESSFUL; + } else if (inl(PCI_BASE | PCI_GPPM_STATUS) & GPPM_R_MABORT) { + break; + } + + loops--; + if (loops == 0) { + printk("%s : Arbiter Locked.\n", __FUNCTION__); + } + } + + clear_status(); + if ((pci_cmd == PCI_CMD_IOR) || (pci_cmd == PCI_CMD_IOW)) { + printk("%s timeout (GPPM_CTRL=%X) ioaddr %lX pci_cmd %X\n", + __FUNCTION__, inl(PCI_BASE | PCI_GPPM_CTRL), ioaddr, + pci_cmd); + } + + if ((pci_cmd == PCI_CMD_IOR) || (pci_cmd == PCI_CMD_CONFIG_READ)) + *val = 0xffffffff; + local_irq_restore(flags); + return PCIBIOS_DEVICE_NOT_FOUND; +} + +/* + * We can't address 8 and 16 bit words directly. Instead we have to + * read/write a 32bit word and mask/modify the data we actually want. + */ +static int +read_config_byte(struct pci_bus *bus, unsigned int devfn, int where, u8 * val) +{ + unsigned int data = 0; + int err; + + if (bus == 0) + return -1; + + err = config_access(PCI_CMD_CONFIG_READ, bus, devfn, where, ~(1 << (where & 3)), &data); + switch (where & 0x03) { + case 0: + *val = (unsigned char)(data & 0x000000ff); + break; + case 1: + *val = (unsigned char)((data & 0x0000ff00) >> 8); + break; + case 2: + *val = (unsigned char)((data & 0x00ff0000) >> 16); + break; + case 3: + *val = (unsigned char)((data & 0xff000000) >> 24); + break; + } + + return err; +} + +static int +read_config_word(struct pci_bus *bus, unsigned int devfn, int where, u16 * val) +{ + unsigned int data = 0; + int err; + + if (bus == 0) + return -1; + + if (where & 0x01) + return PCIBIOS_BAD_REGISTER_NUMBER; + + err = config_access(PCI_CMD_CONFIG_READ, bus, devfn, where, ~(3 << (where & 3)), &data); + switch (where & 0x02) { + case 0: + *val = (unsigned short)(data & 0x0000ffff); + break; + case 2: + *val = (unsigned short)((data & 0xffff0000) >> 16); + break; + } + + return err; +} + +static int +read_config_dword(struct pci_bus *bus, unsigned int devfn, int where, u32 * val) +{ + int err; + if (bus == 0) + return -1; + + if (where & 0x03) + return PCIBIOS_BAD_REGISTER_NUMBER; + + err = config_access(PCI_CMD_CONFIG_READ, bus, devfn, where, 0, val); + + return err; +} + +static int +write_config_byte(struct pci_bus *bus, unsigned int devfn, int where, u8 val) +{ + unsigned int data = (unsigned int)val; + int err; + + if (bus == 0) + return -1; + + switch (where & 0x03) { + case 1: + data = (data << 8); + break; + case 2: + data = (data << 16); + break; + case 3: + data = (data << 24); + break; + default: + break; + } + + err = config_access(PCI_CMD_CONFIG_READ, bus, devfn, where, ~(1 << (where & 3)), &data); + + return err; +} + +static int +write_config_word(struct pci_bus *bus, unsigned int devfn, int where, u16 val) +{ + unsigned int data = (unsigned int)val; + int err; + + if (bus == 0) + return -1; + + if (where & 0x01) + return PCIBIOS_BAD_REGISTER_NUMBER; + + switch (where & 0x02) { + case 2: + data = (data << 16); + break; + default: + break; + } + err = config_access(PCI_CMD_CONFIG_WRITE, bus, devfn, where, ~(3 << (where & 3)), &data); + + return err; +} + +static int +write_config_dword(struct pci_bus *bus, unsigned int devfn, int where, u32 val) +{ + int err; + if (bus == 0) + return -1; + + if (where & 0x03) + return PCIBIOS_BAD_REGISTER_NUMBER; + + err = config_access(PCI_CMD_CONFIG_WRITE, bus, devfn, where, 0, &val); + + return err; +} + +static int config_read(struct pci_bus *bus, unsigned int devfn, int where, int size, u32 * val) +{ + switch (size) { + case 1: { + u8 _val; + int rc = read_config_byte(bus, devfn, where, &_val); + *val = _val; + return rc; + } + case 2: { + u16 _val; + int rc = read_config_word(bus, devfn, where, &_val); + *val = _val; + return rc; + } + default: + return read_config_dword(bus, devfn, where, val); + } +} + +static int config_write(struct pci_bus *bus, unsigned int devfn, int where, int size, u32 val) +{ + switch (size) { + case 1: + return write_config_byte(bus, devfn, where, (u8) val); + case 2: + return write_config_word(bus, devfn, where, (u16) val); + default: + return write_config_dword(bus, devfn, where, val); + } +} + +struct pci_ops pnx8550_pci_ops = { + config_read, + config_write +}; diff --git a/arch/mips/philips/pnx8550/common/Kconfig b/arch/mips/philips/pnx8550/common/Kconfig new file mode 100644 index 000000000000..072572d173cc --- /dev/null +++ b/arch/mips/philips/pnx8550/common/Kconfig @@ -0,0 +1 @@ +# Place holder diff --git a/arch/mips/philips/pnx8550/common/Makefile b/arch/mips/philips/pnx8550/common/Makefile new file mode 100644 index 000000000000..6e38f3bc443c --- /dev/null +++ b/arch/mips/philips/pnx8550/common/Makefile @@ -0,0 +1,27 @@ +# +# Per Hallsmark, per.hallsmark@mvista.com +# +# ######################################################################## +# +# This program is free software; you can distribute it and/or modify it +# under the terms of the GNU General Public License (Version 2) as +# published by the Free Software Foundation. +# +# This program is distributed in the hope it will be useful, but WITHOUT +# ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or +# FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License +# for more details. +# +# You should have received a copy of the GNU General Public License along +# with this program; if not, write to the Free Software Foundation, Inc., +# 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. +# +# ####################################################################### +# +# Makefile for the PNX8550 specific kernel interface routines +# under Linux. +# + +obj-y := setup.o prom.o mipsIRQ.o int.o reset.o time.o proc.o platform.o +obj-$(CONFIG_PCI) += pci.o +obj-$(CONFIG_KGDB) += gdb_hook.o diff --git a/arch/mips/philips/pnx8550/common/gdb_hook.c b/arch/mips/philips/pnx8550/common/gdb_hook.c new file mode 100644 index 000000000000..ad4624f6d9bc --- /dev/null +++ b/arch/mips/philips/pnx8550/common/gdb_hook.c @@ -0,0 +1,109 @@ +/* + * Carsten Langgaard, carstenl@mips.com + * Copyright (C) 2000 MIPS Technologies, Inc. All rights reserved. + * + * ######################################################################## + * + * This program is free software; you can distribute it and/or modify it + * under the terms of the GNU General Public License (Version 2) as + * published by the Free Software Foundation. + * + * This program is distributed in the hope it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License + * for more details. + * + * You should have received a copy of the GNU General Public License along + * with this program; if not, write to the Free Software Foundation, Inc., + * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. + * + * ######################################################################## + * + * This is the interface to the remote debugger stub. + * + */ +#include +#include +#include +#include +#include + +#include +#include + +#include + +static struct serial_state rs_table[IP3106_NR_PORTS] = { +}; +static struct async_struct kdb_port_info = {0}; + +void rs_kgdb_hook(int tty_no) +{ + struct serial_state *ser = &rs_table[tty_no]; + + kdb_port_info.state = ser; + kdb_port_info.magic = SERIAL_MAGIC; + kdb_port_info.port = tty_no; + kdb_port_info.flags = ser->flags; + + /* + * Clear all interrupts + */ + /* Clear all the transmitter FIFO counters (pointer and status) */ + ip3106_lcr(UART_BASE, tty_no) |= IP3106_UART_LCR_TX_RST; + /* Clear all the receiver FIFO counters (pointer and status) */ + ip3106_lcr(UART_BASE, tty_no) |= IP3106_UART_LCR_RX_RST; + /* Clear all interrupts */ + ip3106_iclr(UART_BASE, tty_no) = IP3106_UART_INT_ALLRX | + IP3106_UART_INT_ALLTX; + + /* + * Now, initialize the UART + */ + ip3106_lcr(UART_BASE, tty_no) = IP3106_UART_LCR_8BIT; + ip3106_baud(UART_BASE, tty_no) = 5; // 38400 Baud +} + +int putDebugChar(char c) +{ + /* Wait until FIFO not full */ + while (((ip3106_fifo(UART_BASE, kdb_port_info.port) & IP3106_UART_FIFO_TXFIFO) >> 16) >= 16) + ; + /* Send one char */ + ip3106_fifo(UART_BASE, kdb_port_info.port) = c; + + return 1; +} + +char getDebugChar(void) +{ + char ch; + + /* Wait until there is a char in the FIFO */ + while (!((ip3106_fifo(UART_BASE, kdb_port_info.port) & + IP3106_UART_FIFO_RXFIFO) >> 8)) + ; + /* Read one char */ + ch = ip3106_fifo(UART_BASE, kdb_port_info.port) & + IP3106_UART_FIFO_RBRTHR; + /* Advance the RX FIFO read pointer */ + ip3106_lcr(UART_BASE, kdb_port_info.port) |= IP3106_UART_LCR_RX_NEXT; + return (ch); +} + +void rs_disable_debug_interrupts(void) +{ + ip3106_ien(UART_BASE, kdb_port_info.port) = 0; /* Disable all interrupts */ +} + +void rs_enable_debug_interrupts(void) +{ + /* Clear all the transmitter FIFO counters (pointer and status) */ + ip3106_lcr(UART_BASE, kdb_port_info.port) |= IP3106_UART_LCR_TX_RST; + /* Clear all the receiver FIFO counters (pointer and status) */ + ip3106_lcr(UART_BASE, kdb_port_info.port) |= IP3106_UART_LCR_RX_RST; + /* Clear all interrupts */ + ip3106_iclr(UART_BASE, kdb_port_info.port) = IP3106_UART_INT_ALLRX | + IP3106_UART_INT_ALLTX; + ip3106_ien(UART_BASE, kdb_port_info.port) = IP3106_UART_INT_ALLRX; /* Enable RX interrupts */ +} diff --git a/arch/mips/philips/pnx8550/common/int.c b/arch/mips/philips/pnx8550/common/int.c new file mode 100644 index 000000000000..546144988bf5 --- /dev/null +++ b/arch/mips/philips/pnx8550/common/int.c @@ -0,0 +1,293 @@ +/* + * + * Copyright (C) 2005 Embedded Alley Solutions, Inc + * Ported to 2.6. + * + * Per Hallsmark, per.hallsmark@mvista.com + * Copyright (C) 2000, 2001 MIPS Technologies, Inc. + * Copyright (C) 2001 Ralf Baechle + * + * Cleaned up and bug fixing: Pete Popov, ppopov@embeddedalley.com + * + * This program is free software; you can distribute it and/or modify it + * under the terms of the GNU General Public License (Version 2) as + * published by the Free Software Foundation. + * + * This program is distributed in the hope it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License + * for more details. + * + * You should have received a copy of the GNU General Public License along + * with this program; if not, write to the Free Software Foundation, Inc., + * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. + * + */ +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include +#include +#include +#include + +extern asmlinkage void cp0_irqdispatch(void); + +static DEFINE_SPINLOCK(irq_lock); + +/* default prio for interrupts */ +/* first one is a no-no so therefore always prio 0 (disabled) */ +static char gic_prio[PNX8550_INT_GIC_TOTINT] = { + 0, 1, 1, 1, 1, 15, 1, 1, 1, 1, // 0 - 9 + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, // 10 - 19 + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, // 20 - 29 + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, // 30 - 39 + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, // 40 - 49 + 1, 1, 1, 1, 1, 1, 1, 1, 2, 1, // 50 - 59 + 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, // 60 - 69 + 1 // 70 +}; + +void hw0_irqdispatch(int irq, struct pt_regs *regs) +{ + /* find out which interrupt */ + irq = PNX8550_GIC_VECTOR_0 >> 3; + + if (irq == 0) { + printk("hw0_irqdispatch: irq 0, spurious interrupt?\n"); + return; + } + do_IRQ(PNX8550_INT_GIC_MIN + irq, regs); +} + + +void timer_irqdispatch(int irq, struct pt_regs *regs) +{ + irq = (0x01c0 & read_c0_config7()) >> 6; + + if (irq == 0) { + printk("timer_irqdispatch: irq 0, spurious interrupt?\n"); + return; + } + + if (irq & 0x1) { + do_IRQ(PNX8550_INT_TIMER1, regs); + } + if (irq & 0x2) { + do_IRQ(PNX8550_INT_TIMER2, regs); + } + if (irq & 0x4) { + do_IRQ(PNX8550_INT_TIMER3, regs); + } +} + +static inline void modify_cp0_intmask(unsigned clr_mask, unsigned set_mask) +{ + unsigned long status = read_c0_status(); + + status &= ~((clr_mask & 0xFF) << 8); + status |= (set_mask & 0xFF) << 8; + + write_c0_status(status); +} + +static inline void mask_gic_int(unsigned int irq_nr) +{ + /* interrupt disabled, bit 26(WE_ENABLE)=1 and bit 16(enable)=0 */ + PNX8550_GIC_REQ(irq_nr) = 1<<28; /* set priority to 0 */ +} + +static inline void unmask_gic_int(unsigned int irq_nr) +{ + /* set prio mask to lower four bits and enable interrupt */ + PNX8550_GIC_REQ(irq_nr) = (1<<26 | 1<<16) | (1<<28) | gic_prio[irq_nr]; +} + +static inline void mask_irq(unsigned int irq_nr) +{ + if ((PNX8550_INT_CP0_MIN <= irq_nr) && (irq_nr <= PNX8550_INT_CP0_MAX)) { + modify_cp0_intmask(1 << irq_nr, 0); + } else if ((PNX8550_INT_GIC_MIN <= irq_nr) && + (irq_nr <= PNX8550_INT_GIC_MAX)) { + mask_gic_int(irq_nr - PNX8550_INT_GIC_MIN); + } else if ((PNX8550_INT_TIMER_MIN <= irq_nr) && + (irq_nr <= PNX8550_INT_TIMER_MAX)) { + modify_cp0_intmask(1 << 7, 0); + } else { + printk("mask_irq: irq %d doesn't exist!\n", irq_nr); + } +} + +static inline void unmask_irq(unsigned int irq_nr) +{ + if ((PNX8550_INT_CP0_MIN <= irq_nr) && (irq_nr <= PNX8550_INT_CP0_MAX)) { + modify_cp0_intmask(0, 1 << irq_nr); + } else if ((PNX8550_INT_GIC_MIN <= irq_nr) && + (irq_nr <= PNX8550_INT_GIC_MAX)) { + unmask_gic_int(irq_nr - PNX8550_INT_GIC_MIN); + } else if ((PNX8550_INT_TIMER_MIN <= irq_nr) && + (irq_nr <= PNX8550_INT_TIMER_MAX)) { + modify_cp0_intmask(0, 1 << 7); + } else { + printk("mask_irq: irq %d doesn't exist!\n", irq_nr); + } +} + +#define pnx8550_disable pnx8550_ack +static void pnx8550_ack(unsigned int irq) +{ + unsigned long flags; + + spin_lock_irqsave(&irq_lock, flags); + mask_irq(irq); + spin_unlock_irqrestore(&irq_lock, flags); +} + +#define pnx8550_enable pnx8550_unmask +static void pnx8550_unmask(unsigned int irq) +{ + unsigned long flags; + + spin_lock_irqsave(&irq_lock, flags); + unmask_irq(irq); + spin_unlock_irqrestore(&irq_lock, flags); +} + +static unsigned int startup_irq(unsigned int irq_nr) +{ + pnx8550_unmask(irq_nr); + return 0; +} + +static void shutdown_irq(unsigned int irq_nr) +{ + pnx8550_ack(irq_nr); + return; +} + +int pnx8550_set_gic_priority(int irq, int priority) +{ + int gic_irq = irq-PNX8550_INT_GIC_MIN; + int prev_priority = PNX8550_GIC_REQ(gic_irq) & 0xf; + + gic_prio[gic_irq] = priority; + PNX8550_GIC_REQ(gic_irq) |= (0x10000000 | gic_prio[gic_irq]); + + return prev_priority; +} + +static inline void mask_and_ack_level_irq(unsigned int irq) +{ + pnx8550_disable(irq); + return; +} + +static void end_irq(unsigned int irq) +{ + if (!(irq_desc[irq].status & (IRQ_DISABLED|IRQ_INPROGRESS))) { + pnx8550_enable(irq); + } +} + +static struct hw_interrupt_type level_irq_type = { + .typename = "PNX Level IRQ", + .startup = startup_irq, + .shutdown = shutdown_irq, + .enable = pnx8550_enable, + .disable = pnx8550_disable, + .ack = mask_and_ack_level_irq, + .end = end_irq, +}; + +static struct irqaction gic_action = { + .handler = no_action, + .flags = SA_INTERRUPT, + .name = "GIC", +}; + +static struct irqaction timer_action = { + .handler = no_action, + .flags = SA_INTERRUPT, + .name = "Timer", +}; + +void __init arch_init_irq(void) +{ + int i; + int configPR; + + /* init of cp0 interrupts */ + set_except_vector(0, cp0_irqdispatch); + + for (i = 0; i < PNX8550_INT_CP0_TOTINT; i++) { + irq_desc[i].handler = &level_irq_type; + pnx8550_ack(i); /* mask the irq just in case */ + } + + /* init of GIC/IPC interrupts */ + /* should be done before cp0 since cp0 init enables the GIC int */ + for (i = PNX8550_INT_GIC_MIN; i <= PNX8550_INT_GIC_MAX; i++) { + int gic_int_line = i - PNX8550_INT_GIC_MIN; + if (gic_int_line == 0 ) + continue; // don't fiddle with int 0 + /* + * enable change of TARGET, ENABLE and ACTIVE_LOW bits + * set TARGET 0 to route through hw0 interrupt + * set ACTIVE_LOW 0 active high (correct?) + * + * We really should setup an interrupt description table + * to do this nicely. + * Note, PCI INTA is active low on the bus, but inverted + * in the GIC, so to us it's active high. + */ +#ifdef CONFIG_PNX8550_V2PCI + if (gic_int_line == (PNX8550_INT_GPIO0 - PNX8550_INT_GIC_MIN)) { + /* PCI INT through gpio 8, which is setup in + * pnx8550_setup.c and routed to GPIO + * Interrupt Level 0 (GPIO Connection 58). + * Set it active low. */ + + PNX8550_GIC_REQ(gic_int_line) = 0x1E020000; + } else +#endif + { + PNX8550_GIC_REQ(i - PNX8550_INT_GIC_MIN) = 0x1E000000; + } + + /* mask/priority is still 0 so we will not get any + * interrupts until it is unmasked */ + + irq_desc[i].handler = &level_irq_type; + } + + /* Priority level 0 */ + PNX8550_GIC_PRIMASK_0 = PNX8550_GIC_PRIMASK_1 = 0; + + /* Set int vector table address */ + PNX8550_GIC_VECTOR_0 = PNX8550_GIC_VECTOR_1 = 0; + + irq_desc[MIPS_CPU_GIC_IRQ].handler = &level_irq_type; + setup_irq(MIPS_CPU_GIC_IRQ, &gic_action); + + /* init of Timer interrupts */ + for (i = PNX8550_INT_TIMER_MIN; i <= PNX8550_INT_TIMER_MAX; i++) { + irq_desc[i].handler = &level_irq_type; + } + + /* Stop Timer 1-3 */ + configPR = read_c0_config7(); + configPR |= 0x00000038; + write_c0_config7(configPR); + + irq_desc[MIPS_CPU_TIMER_IRQ].handler = &level_irq_type; + setup_irq(MIPS_CPU_TIMER_IRQ, &timer_action); +} + +EXPORT_SYMBOL(pnx8550_set_gic_priority); diff --git a/arch/mips/philips/pnx8550/common/mipsIRQ.S b/arch/mips/philips/pnx8550/common/mipsIRQ.S new file mode 100644 index 000000000000..338bffda3fab --- /dev/null +++ b/arch/mips/philips/pnx8550/common/mipsIRQ.S @@ -0,0 +1,76 @@ +/* + * Copyright (c) 2002 Philips, Inc. All rights. + * Copyright (c) 2002 Red Hat, Inc. All rights. + * + * This software may be freely redistributed under the terms of the + * GNU General Public License. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. + * + * Based upon arch/mips/galileo-boards/ev64240/int-handler.S + * + */ +#include +#include +#include +#include +#include + +/* + * cp0_irqdispatch + * + * Code to handle in-core interrupt exception. + */ + + .align 5 + .set reorder + .set noat + NESTED(cp0_irqdispatch, PT_SIZE, sp) + SAVE_ALL + CLI + .set at + mfc0 t0,CP0_CAUSE + mfc0 t2,CP0_STATUS + + and t0,t2 + + andi t1,t0,STATUSF_IP2 /* int0 hardware line */ + bnez t1,ll_hw0_irq + nop + + andi t1,t0,STATUSF_IP7 /* int5 hardware line */ + bnez t1,ll_timer_irq + nop + + /* wrong alarm or masked ... */ + + j spurious_interrupt + nop + END(cp0_irqdispatch) + + .align 5 + .set reorder +ll_hw0_irq: + li a0,2 + move a1,sp + jal hw0_irqdispatch + nop + j ret_from_irq + nop + + .align 5 + .set reorder +ll_timer_irq: + mfc0 t3,CP0_CONFIG,7 + andi t4,t3,0x01c0 + beqz t4,ll_timer_out + nop + li a0,7 + move a1,sp + jal timer_irqdispatch + nop + +ll_timer_out: j ret_from_irq + nop diff --git a/arch/mips/philips/pnx8550/common/pci.c b/arch/mips/philips/pnx8550/common/pci.c new file mode 100644 index 000000000000..baa6905f649f --- /dev/null +++ b/arch/mips/philips/pnx8550/common/pci.c @@ -0,0 +1,133 @@ +/* + * + * BRIEF MODULE DESCRIPTION + * + * Author: source@mvista.com + * + * This program is free software; you can distribute it and/or modify it + * under the terms of the GNU General Public License (Version 2) as + * published by the Free Software Foundation. + * + * This program is distributed in the hope it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License + * for more details. + * + * You should have received a copy of the GNU General Public License along + * with this program; if not, write to the Free Software Foundation, Inc., + * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. + */ +#include +#include +#include +#include + +#include +#include +#include + +static struct resource pci_io_resource = { + "pci IO space", + (u32)(PNX8550_PCIIO + 0x1000), /* reserve regacy I/O space */ + (u32)(PNX8550_PCIIO + PNX8550_PCIIO_SIZE), + IORESOURCE_IO +}; + +static struct resource pci_mem_resource = { + "pci memory space", + (u32)(PNX8550_PCIMEM), + (u32)(PNX8550_PCIMEM + PNX8550_PCIMEM_SIZE - 1), + IORESOURCE_MEM +}; + +extern struct pci_ops pnx8550_pci_ops; + +static struct pci_controller pnx8550_controller = { + .pci_ops = &pnx8550_pci_ops, + .io_resource = &pci_io_resource, + .mem_resource = &pci_mem_resource, +}; + +/* Return the total size of DRAM-memory, (RANK0 + RANK1) */ +static inline unsigned long get_system_mem_size(void) +{ + /* Read IP2031_RANK0_ADDR_LO */ + unsigned long dram_r0_lo = inl(PCI_BASE | 0x65010); + /* Read IP2031_RANK1_ADDR_HI */ + unsigned long dram_r1_hi = inl(PCI_BASE | 0x65018); + + return dram_r1_hi - dram_r0_lo + 1; +} + +static int __init pnx8550_pci_setup(void) +{ + int pci_mem_code; + int mem_size = get_system_mem_size() >> 20; + + /* Clear the Global 2 Register, PCI Inta Output Enable Registers + Bit 1:Enable DAC Powerdown + -> 0:DACs are enabled and are working normally + 1:DACs are powerdown + Bit 0:Enable of PCI inta output + -> 0 = Disable PCI inta output + 1 = Enable PCI inta output + */ + PNX8550_GLB2_ENAB_INTA_O = 0; + + /* Calc the PCI mem size code */ + if (mem_size >= 128) + pci_mem_code = SIZE_128M; + else if (mem_size >= 64) + pci_mem_code = SIZE_64M; + else if (mem_size >= 32) + pci_mem_code = SIZE_32M; + else + pci_mem_code = SIZE_16M; + + /* Set PCI_XIO registers */ + outl(pci_mem_resource.start, PCI_BASE | PCI_BASE1_LO); + outl(pci_mem_resource.end + 1, PCI_BASE | PCI_BASE1_HI); + outl(pci_io_resource.start, PCI_BASE | PCI_BASE2_LO); + outl(pci_io_resource.end, PCI_BASE | PCI_BASE2_HI); + + /* Send memory transaction via PCI_BASE2 */ + outl(0x00000001, PCI_BASE | PCI_IO); + + /* Unlock the setup register */ + outl(0xca, PCI_BASE | PCI_UNLOCKREG); + + /* + * BAR0 of PNX8550 (pci base 10) must be zero in order for ide + * to work, and in order for bus_to_baddr to work without any + * hacks. + */ + outl(0x00000000, PCI_BASE | PCI_BASE10); + + /* + *These two bars are set by default or the boot code. + * However, it's safer to set them here so we're not boot + * code dependent. + */ + outl(0x1be00000, PCI_BASE | PCI_BASE14); /* PNX MMIO */ + outl(PNX8550_NAND_BASE_ADDR, PCI_BASE | PCI_BASE18); /* XIO */ + + outl(PCI_EN_TA | + PCI_EN_PCI2MMI | + PCI_EN_XIO | + PCI_SETUP_BASE18_SIZE(SIZE_32M) | + PCI_SETUP_BASE18_EN | + PCI_SETUP_BASE14_EN | + PCI_SETUP_BASE10_PREF | + PCI_SETUP_BASE10_SIZE(pci_mem_code) | + PCI_SETUP_CFGMANAGE_EN | + PCI_SETUP_PCIARB_EN, + PCI_BASE | + PCI_SETUP); /* PCI_SETUP */ + outl(0x00000000, PCI_BASE | PCI_CTRL); /* PCI_CONTROL */ + + register_pci_controller(&pnx8550_controller); + + return 0; +} + +arch_initcall(pnx8550_pci_setup); diff --git a/arch/mips/philips/pnx8550/common/platform.c b/arch/mips/philips/pnx8550/common/platform.c new file mode 100644 index 000000000000..8aa9bd65b45e --- /dev/null +++ b/arch/mips/philips/pnx8550/common/platform.c @@ -0,0 +1,135 @@ +/* + * Platform device support for Philips PNX8550 SoCs + * + * Copyright 2005, Embedded Alley Solutions, Inc + * + * Based on arch/mips/au1000/common/platform.c + * Platform device support for Au1x00 SoCs. + * + * Copyright 2004, Matt Porter + * + * This file is licensed under the terms of the GNU General Public + * License version 2. This program is licensed "as is" without any + * warranty of any kind, whether express or implied. + */ +#include +#include +#include +#include +#include +#include + +#include +#include +#include + +extern struct uart_ops ip3106_pops; + +static struct resource pnx8550_usb_ohci_resources[] = { + [0] = { + .start = PNX8550_USB_OHCI_OP_BASE, + .end = PNX8550_USB_OHCI_OP_BASE + + PNX8550_USB_OHCI_OP_LEN, + .flags = IORESOURCE_MEM, + }, + [1] = { + .start = PNX8550_INT_USB, + .end = PNX8550_INT_USB, + .flags = IORESOURCE_IRQ, + }, +}; + +static struct resource pnx8550_uart_resources[] = { + [0] = { + .start = PNX8550_UART_PORT0, + .end = PNX8550_UART_PORT0 + 0xfff, + .flags = IORESOURCE_MEM, + }, + [1] = { + .start = PNX8550_UART_INT(0), + .end = PNX8550_UART_INT(0), + .flags = IORESOURCE_IRQ, + }, + [2] = { + .start = PNX8550_UART_PORT1, + .end = PNX8550_UART_PORT1 + 0xfff, + .flags = IORESOURCE_MEM, + }, + [3] = { + .start = PNX8550_UART_INT(1), + .end = PNX8550_UART_INT(1), + .flags = IORESOURCE_IRQ, + }, +}; + +struct ip3106_port ip3106_ports[] = { + [0] = { + .port = { + .type = PORT_IP3106, + .iotype = SERIAL_IO_MEM, + .membase = (void __iomem *)PNX8550_UART_PORT0, + .mapbase = PNX8550_UART_PORT0, + .irq = PNX8550_UART_INT(0), + .uartclk = 3692300, + .fifosize = 16, + .ops = &ip3106_pops, + .flags = ASYNC_BOOT_AUTOCONF, + .line = 0, + }, + }, + [1] = { + .port = { + .type = PORT_IP3106, + .iotype = SERIAL_IO_MEM, + .membase = (void __iomem *)PNX8550_UART_PORT1, + .mapbase = PNX8550_UART_PORT1, + .irq = PNX8550_UART_INT(1), + .uartclk = 3692300, + .fifosize = 16, + .ops = &ip3106_pops, + .flags = ASYNC_BOOT_AUTOCONF, + .line = 1, + }, + }, +}; + +/* The dmamask must be set for OHCI to work */ +static u64 ohci_dmamask = ~(u32)0; + +static u64 uart_dmamask = ~(u32)0; + +static struct platform_device pnx8550_usb_ohci_device = { + .name = "pnx8550-ohci", + .id = -1, + .dev = { + .dma_mask = &ohci_dmamask, + .coherent_dma_mask = 0xffffffff, + }, + .num_resources = ARRAY_SIZE(pnx8550_usb_ohci_resources), + .resource = pnx8550_usb_ohci_resources, +}; + +static struct platform_device pnx8550_uart_device = { + .name = "ip3106-uart", + .id = -1, + .dev = { + .dma_mask = &uart_dmamask, + .coherent_dma_mask = 0xffffffff, + .platform_data = ip3106_ports, + }, + .num_resources = ARRAY_SIZE(pnx8550_uart_resources), + .resource = pnx8550_uart_resources, +}; + +static struct platform_device *pnx8550_platform_devices[] __initdata = { + &pnx8550_usb_ohci_device, + &pnx8550_uart_device, +}; + +int pnx8550_platform_init(void) +{ + return platform_add_devices(pnx8550_platform_devices, + ARRAY_SIZE(pnx8550_platform_devices)); +} + +arch_initcall(pnx8550_platform_init); diff --git a/arch/mips/philips/pnx8550/common/proc.c b/arch/mips/philips/pnx8550/common/proc.c new file mode 100644 index 000000000000..72a016767e09 --- /dev/null +++ b/arch/mips/philips/pnx8550/common/proc.c @@ -0,0 +1,113 @@ +/* + * This program is free software; you can distribute it and/or modify it + * under the terms of the GNU General Public License (Version 2) as + * published by the Free Software Foundation. + * + * This program is distributed in the hope it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License + * for more details. + * + * You should have received a copy of the GNU General Public License along + * with this program; if not, write to the Free Software Foundation, Inc., + * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. + */ +#include +#include +#include +#include +#include +#include +#include +#include + +#include +#include +#include +#include + + +static int pnx8550_timers_read (char* page, char** start, off_t offset, int count, int* eof, void* data) +{ + int len = 0; + int configPR = read_c0_config7(); + + if (offset==0) { + len += sprintf(&page[len],"Timer: count, compare, tc, status\n"); + len += sprintf(&page[len]," 1: %11i, %8i, %1i, %s\n", + read_c0_count(), read_c0_compare(), + (configPR>>6)&0x1, ((configPR>>3)&0x1)? "off":"on"); + len += sprintf(&page[len]," 2: %11i, %8i, %1i, %s\n", + read_c0_count2(), read_c0_compare2(), + (configPR>>7)&0x1, ((configPR>>4)&0x1)? "off":"on"); + len += sprintf(&page[len]," 3: %11i, %8i, %1i, %s\n", + read_c0_count3(), read_c0_compare3(), + (configPR>>8)&0x1, ((configPR>>5)&0x1)? "off":"on"); + } + + return len; +} + +static int pnx8550_registers_read (char* page, char** start, off_t offset, int count, int* eof, void* data) +{ + int len = 0; + + if (offset==0) { + len += sprintf(&page[len],"config1: %#10.8x\n",read_c0_config1()); + len += sprintf(&page[len],"config2: %#10.8x\n",read_c0_config2()); + len += sprintf(&page[len],"config3: %#10.8x\n",read_c0_config3()); + len += sprintf(&page[len],"configPR: %#10.8x\n",read_c0_config7()); + len += sprintf(&page[len],"status: %#10.8x\n",read_c0_status()); + len += sprintf(&page[len],"cause: %#10.8x\n",read_c0_cause()); + len += sprintf(&page[len],"count: %#10.8x\n",read_c0_count()); + len += sprintf(&page[len],"count_2: %#10.8x\n",read_c0_count2()); + len += sprintf(&page[len],"count_3: %#10.8x\n",read_c0_count3()); + len += sprintf(&page[len],"compare: %#10.8x\n",read_c0_compare()); + len += sprintf(&page[len],"compare_2: %#10.8x\n",read_c0_compare2()); + len += sprintf(&page[len],"compare_3: %#10.8x\n",read_c0_compare3()); + } + + return len; +} + +static struct proc_dir_entry* pnx8550_dir = NULL; +static struct proc_dir_entry* pnx8550_timers = NULL; +static struct proc_dir_entry* pnx8550_registers = NULL; + +static int pnx8550_proc_init( void ) +{ + + // Create /proc/pnx8550 + pnx8550_dir = create_proc_entry("pnx8550", S_IFDIR|S_IRUGO, NULL); + if (pnx8550_dir){ + pnx8550_dir->nlink = 1; + } + else { + printk(KERN_ERR "Can't create pnx8550 proc dir\n"); + return -1; + } + + // Create /proc/pnx8550/timers + pnx8550_timers = create_proc_entry("timers", S_IFREG|S_IRUGO, pnx8550_dir ); + if (pnx8550_timers){ + pnx8550_timers->nlink = 1; + pnx8550_timers->read_proc = pnx8550_timers_read; + } + else { + printk(KERN_ERR "Can't create pnx8550 timers proc file\n"); + } + + // Create /proc/pnx8550/registers + pnx8550_registers = create_proc_entry("registers", S_IFREG|S_IRUGO, pnx8550_dir ); + if (pnx8550_registers){ + pnx8550_registers->nlink = 1; + pnx8550_registers->read_proc = pnx8550_registers_read; + } + else { + printk(KERN_ERR "Can't create pnx8550 registers proc file\n"); + } + + return 0; +} + +__initcall(pnx8550_proc_init); diff --git a/arch/mips/philips/pnx8550/common/prom.c b/arch/mips/philips/pnx8550/common/prom.c new file mode 100644 index 000000000000..70aac9759412 --- /dev/null +++ b/arch/mips/philips/pnx8550/common/prom.c @@ -0,0 +1,138 @@ +/* + * + * Per Hallsmark, per.hallsmark@mvista.com + * + * Based on jmr3927/common/prom.c + * + * 2004 (c) MontaVista Software, Inc. This file is licensed under the + * terms of the GNU General Public License version 2. This program is + * licensed "as is" without any warranty of any kind, whether express + * or implied. + */ +#include +#include +#include +#include +#include + +#include +#include + +/* #define DEBUG_CMDLINE */ + +extern int prom_argc; +extern char **prom_argv, **prom_envp; + +typedef struct +{ + char *name; +/* char *val; */ +}t_env_var; + + +char * prom_getcmdline(void) +{ + return &(arcs_cmdline[0]); +} + +void prom_init_cmdline(void) +{ + char *cp; + int actr; + + actr = 1; /* Always ignore argv[0] */ + + cp = &(arcs_cmdline[0]); + while(actr < prom_argc) { + strcpy(cp, prom_argv[actr]); + cp += strlen(prom_argv[actr]); + *cp++ = ' '; + actr++; + } + if (cp != &(arcs_cmdline[0])) /* get rid of trailing space */ + --cp; + *cp = '\0'; +} + +char *prom_getenv(char *envname) +{ + /* + * Return a pointer to the given environment variable. + * Environment variables are stored in the form of "memsize=64". + */ + + t_env_var *env = (t_env_var *)prom_envp; + int i; + + i = strlen(envname); + + while(env->name) { + if(strncmp(envname, env->name, i) == 0) { + return(env->name + strlen(envname) + 1); + } + env++; + } + return(NULL); +} + +inline unsigned char str2hexnum(unsigned char c) +{ + if(c >= '0' && c <= '9') + return c - '0'; + if(c >= 'a' && c <= 'f') + return c - 'a' + 10; + if(c >= 'A' && c <= 'F') + return c - 'A' + 10; + return 0; /* foo */ +} + +inline void str2eaddr(unsigned char *ea, unsigned char *str) +{ + int i; + + for(i = 0; i < 6; i++) { + unsigned char num; + + if((*str == '.') || (*str == ':')) + str++; + num = str2hexnum(*str++) << 4; + num |= (str2hexnum(*str++)); + ea[i] = num; + } +} + +int get_ethernet_addr(char *ethernet_addr) +{ + char *ethaddr_str; + + ethaddr_str = prom_getenv("ethaddr"); + if (!ethaddr_str) { + printk("ethaddr not set in boot prom\n"); + return -1; + } + str2eaddr(ethernet_addr, ethaddr_str); + return 0; +} + +unsigned long __init prom_free_prom_memory(void) +{ + return 0; +} + +extern int pnx8550_console_port; + +/* used by prom_printf */ +void prom_putchar(char c) +{ + if (pnx8550_console_port != -1) { + /* Wait until FIFO not full */ + while( ((ip3106_fifo(UART_BASE, pnx8550_console_port) & IP3106_UART_FIFO_TXFIFO) >> 16) >= 16) + ; + /* Send one char */ + ip3106_fifo(UART_BASE, pnx8550_console_port) = c; + } +} + +EXPORT_SYMBOL(prom_getcmdline); +EXPORT_SYMBOL(get_ethernet_addr); +EXPORT_SYMBOL(str2eaddr); diff --git a/arch/mips/philips/pnx8550/common/reset.c b/arch/mips/philips/pnx8550/common/reset.c new file mode 100644 index 000000000000..7b2cbc5b2c7c --- /dev/null +++ b/arch/mips/philips/pnx8550/common/reset.c @@ -0,0 +1,49 @@ +/*. + * + * ######################################################################## + * + * This program is free software; you can distribute it and/or modify it + * under the terms of the GNU General Public License (Version 2) as + * published by the Free Software Foundation. + * + * This program is distributed in the hope it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License + * for more details. + * + * You should have received a copy of the GNU General Public License along + * with this program; if not, write to the Free Software Foundation, Inc., + * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. + * + * ######################################################################## + * + * Reset the PNX8550 board. + * + */ +#include +#include +#include + +void pnx8550_machine_restart(char *command) +{ + char head[] = "************* Machine restart *************"; + char foot[] = "*******************************************"; + + printk("\n\n"); + printk("%s\n", head); + if (command != NULL) + printk("* %s\n", command); + printk("%s\n", foot); + + PNX8550_RST_CTL = PNX8550_RST_DO_SW_RST; +} + +void pnx8550_machine_halt(void) +{ + printk("*** Machine halt. (Not implemented) ***\n"); +} + +void pnx8550_machine_power_off(void) +{ + printk("*** Machine power off. (Not implemented) ***\n"); +} diff --git a/arch/mips/philips/pnx8550/common/setup.c b/arch/mips/philips/pnx8550/common/setup.c new file mode 100644 index 000000000000..ee6bf72094f6 --- /dev/null +++ b/arch/mips/philips/pnx8550/common/setup.c @@ -0,0 +1,149 @@ +/* + * + * 2.6 port, Embedded Alley Solutions, Inc + * + * Based on Per Hallsmark, per.hallsmark@mvista.com + * + * This program is free software; you can distribute it and/or modify it + * under the terms of the GNU General Public License (Version 2) as + * published by the Free Software Foundation. + * + * This program is distributed in the hope it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License + * for more details. + * + * You should have received a copy of the GNU General Public License along + * with this program; if not, write to the Free Software Foundation, Inc., + * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. + */ +#include +#include +#include +#include +#include +#include +#include +#include + +#include +#include +#include +#include +#include +#include +#include + +#include +#include +#include +#include +#include + +extern void prom_printf(char *fmt, ...); + +extern void __init board_setup(void); +extern void pnx8550_machine_restart(char *); +extern void pnx8550_machine_halt(void); +extern void pnx8550_machine_power_off(void); +extern struct resource ioport_resource; +extern struct resource iomem_resource; +extern void (*board_time_init)(void); +extern void pnx8550_time_init(void); +extern void (*board_timer_setup)(struct irqaction *irq); +extern void pnx8550_timer_setup(struct irqaction *irq); +extern void rs_kgdb_hook(int tty_no); +extern void prom_printf(char *fmt, ...); +extern char *prom_getcmdline(void); + +struct resource standard_io_resources[] = { + {"dma1", 0x00, 0x1f, IORESOURCE_BUSY}, + {"timer", 0x40, 0x5f, IORESOURCE_BUSY}, + {"dma page reg", 0x80, 0x8f, IORESOURCE_BUSY}, + {"dma2", 0xc0, 0xdf, IORESOURCE_BUSY}, +}; + +#define STANDARD_IO_RESOURCES (sizeof(standard_io_resources)/sizeof(struct resource)) + +extern struct resource pci_io_resource; +extern struct resource pci_mem_resource; + +/* Return the total size of DRAM-memory, (RANK0 + RANK1) */ +unsigned long get_system_mem_size(void) +{ + /* Read IP2031_RANK0_ADDR_LO */ + unsigned long dram_r0_lo = inl(PCI_BASE | 0x65010); + /* Read IP2031_RANK1_ADDR_HI */ + unsigned long dram_r1_hi = inl(PCI_BASE | 0x65018); + + return dram_r1_hi - dram_r0_lo + 1; +} + +int pnx8550_console_port = -1; + +void __init plat_setup(void) +{ + int i; + char* argptr; + + board_setup(); /* board specific setup */ + + _machine_restart = pnx8550_machine_restart; + _machine_halt = pnx8550_machine_halt; + _machine_power_off = pnx8550_machine_power_off; + + board_time_init = pnx8550_time_init; + board_timer_setup = pnx8550_timer_setup; + + /* Clear the Global 2 Register, PCI Inta Output Enable Registers + Bit 1:Enable DAC Powerdown + -> 0:DACs are enabled and are working normally + 1:DACs are powerdown + Bit 0:Enable of PCI inta output + -> 0 = Disable PCI inta output + 1 = Enable PCI inta output + */ + PNX8550_GLB2_ENAB_INTA_O = 0; + + /* IO/MEM resources. */ + set_io_port_base(KSEG1); + ioport_resource.start = 0; + ioport_resource.end = ~0; + iomem_resource.start = 0; + iomem_resource.end = ~0; + + /* Request I/O space for devices on this board */ + for (i = 0; i < STANDARD_IO_RESOURCES; i++) + request_resource(&ioport_resource, standard_io_resources + i); + + /* Place the Mode Control bit for GPIO pin 16 in primary function */ + /* Pin 16 is used by UART1, UA1_TX */ + outl((PNX8550_GPIO_MODE_PRIMOP << PNX8550_GPIO_MC_16_BIT) | + (PNX8550_GPIO_MODE_PRIMOP << PNX8550_GPIO_MC_17_BIT), + PNX8550_GPIO_MC1); + + argptr = prom_getcmdline(); + if ((argptr = strstr(argptr, "console=ttyS")) != NULL) { + argptr += strlen("console=ttyS"); + pnx8550_console_port = *argptr == '0' ? 0 : 1; + + /* We must initialize the UART (console) before prom_printf */ + /* Set LCR to 8-bit and BAUD to 38400 (no 5) */ + ip3106_lcr(UART_BASE, pnx8550_console_port) = + IP3106_UART_LCR_8BIT; + ip3106_baud(UART_BASE, pnx8550_console_port) = 5; + } + +#ifdef CONFIG_KGDB + argptr = prom_getcmdline(); + if ((argptr = strstr(argptr, "kgdb=ttyS")) != NULL) { + int line; + argptr += strlen("kgdb=ttyS"); + line = *argptr == '0' ? 0 : 1; + rs_kgdb_hook(line); + prom_printf("KGDB: Using ttyS%i for session, " + "please connect your debugger\n", line ? 1 : 0); + } +#endif + return; +} diff --git a/arch/mips/philips/pnx8550/common/time.c b/arch/mips/philips/pnx8550/common/time.c new file mode 100644 index 000000000000..70664ea96b92 --- /dev/null +++ b/arch/mips/philips/pnx8550/common/time.c @@ -0,0 +1,105 @@ +/* + * Copyright 2001, 2002, 2003 MontaVista Software Inc. + * Author: Jun Sun, jsun@mvista.com or jsun@junsun.net + * + * Common time service routines for MIPS machines. See + * Documents/MIPS/README.txt. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License as published by the + * Free Software Foundation; either version 2 of the License, or (at your + * option) any later version. + */ +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include +#include +#include +#include +#include +#include + +#include +#include + +extern unsigned int mips_hpt_frequency; + +/* + * pnx8550_time_init() - it does the following things: + * + * 1) board_time_init() - + * a) (optional) set up RTC routines, + * b) (optional) calibrate and set the mips_hpt_frequency + * (only needed if you intended to use fixed_rate_gettimeoffset + * or use cpu counter as timer interrupt source) + */ + +void pnx8550_time_init(void) +{ + unsigned int n; + unsigned int m; + unsigned int p; + unsigned int pow2p; + + /* PLL0 sets MIPS clock (PLL1 <=> TM1, PLL6 <=> TM2, PLL5 <=> mem) */ + /* (but only if CLK_MIPS_CTL select value [bits 3:1] is 1: FIXME) */ + + n = (PNX8550_CM_PLL0_CTL & PNX8550_CM_PLL_N_MASK) >> 16; + m = (PNX8550_CM_PLL0_CTL & PNX8550_CM_PLL_M_MASK) >> 8; + p = (PNX8550_CM_PLL0_CTL & PNX8550_CM_PLL_P_MASK) >> 2; + pow2p = (1 << p); + + db_assert(m != 0 && pow2p != 0); + + /* + * Compute the frequency as in the PNX8550 User Manual 1.0, p.186 + * (a.k.a. 8-10). Divide by HZ for a timer offset that results in + * HZ timer interrupts per second. + */ + mips_hpt_frequency = 27UL * ((1000000UL * n)/(m * pow2p)); +} + +/* + * pnx8550_timer_setup() - it does the following things: + * + * 5) board_timer_setup() - + * a) (optional) over-write any choices made above by time_init(). + * b) machine specific code should setup the timer irqaction. + * c) enable the timer interrupt + */ + +void __init pnx8550_timer_setup(struct irqaction *irq) +{ + int configPR; + + setup_irq(PNX8550_INT_TIMER1, irq); + + /* Start timer1 */ + configPR = read_c0_config7(); + configPR &= ~0x00000008; + write_c0_config7(configPR); + + /* Timer 2 stop */ + configPR = read_c0_config7(); + configPR |= 0x00000010; + write_c0_config7(configPR); + + write_c0_count2(0); + write_c0_compare2(0xffffffff); + + /* Timer 3 stop */ + configPR = read_c0_config7(); + configPR |= 0x00000020; + write_c0_config7(configPR); +} diff --git a/arch/mips/philips/pnx8550/jbs/Makefile b/arch/mips/philips/pnx8550/jbs/Makefile new file mode 100644 index 000000000000..e8228dbca8f6 --- /dev/null +++ b/arch/mips/philips/pnx8550/jbs/Makefile @@ -0,0 +1,4 @@ + +# Makefile for the Philips JBS Board. + +lib-y := init.o board_setup.o irqmap.o diff --git a/arch/mips/philips/pnx8550/jbs/board_setup.c b/arch/mips/philips/pnx8550/jbs/board_setup.c new file mode 100644 index 000000000000..f92826e0096d --- /dev/null +++ b/arch/mips/philips/pnx8550/jbs/board_setup.c @@ -0,0 +1,65 @@ +/* + * JBS Specific board startup routines. + * + * Copyright 2005, Embedded Alley Solutions, Inc + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License as published by the + * Free Software Foundation; either version 2 of the License, or (at your + * option) any later version. + * + * THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESS OR IMPLIED + * WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF + * MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN + * NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT, + * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT + * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF + * USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON + * ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF + * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + * You should have received a copy of the GNU General Public License along + * with this program; if not, write to the Free Software Foundation, Inc., + * 675 Mass Ave, Cambridge, MA 02139, USA. + */ +#include +#include +#include +#include +#include +#include +#include + +#include +#include +#include +#include +#include +#include + +#include + +/* CP0 hazard avoidance. */ +#define BARRIER __asm__ __volatile__(".set noreorder\n\t" \ + "nop; nop; nop; nop; nop; nop;\n\t" \ + ".set reorder\n\t") + +void __init board_setup(void) +{ + unsigned long config0, configpr; + + config0 = read_c0_config(); + + /* clear all three cache coherency fields */ + config0 &= ~(0x7 | (7<<25) | (7<<28)); + config0 |= (CONF_CM_DEFAULT | (CONF_CM_DEFAULT<<25) | + (CONF_CM_DEFAULT<<28)); + write_c0_config(config0); + BARRIER; + + configpr = read_c0_config7(); + configpr |= (1<<19); /* enable tlb */ + write_c0_config7(configpr); + BARRIER; +} diff --git a/arch/mips/philips/pnx8550/jbs/init.c b/arch/mips/philips/pnx8550/jbs/init.c new file mode 100644 index 000000000000..85f449174bc3 --- /dev/null +++ b/arch/mips/philips/pnx8550/jbs/init.c @@ -0,0 +1,57 @@ +/* + * + * Copyright 2005 Embedded Alley Solutions, Inc + * source@embeddedalley.com + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License as published by the + * Free Software Foundation; either version 2 of the License, or (at your + * option) any later version. + * + * THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESS OR IMPLIED + * WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF + * MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN + * NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT, + * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT + * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF + * USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON + * ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF + * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + * You should have received a copy of the GNU General Public License along + * with this program; if not, write to the Free Software Foundation, Inc., + * 675 Mass Ave, Cambridge, MA 02139, USA. + */ + +#include +#include +#include +#include +#include +#include +#include +#include + +int prom_argc; +char **prom_argv, **prom_envp; +extern void __init prom_init_cmdline(void); +extern char *prom_getenv(char *envname); + +const char *get_system_type(void) +{ + return "Philips PNX8550/JBS"; +} + +void __init prom_init(void) +{ + + unsigned long memsize; + + mips_machgroup = MACH_GROUP_PHILIPS; + mips_machtype = MACH_PHILIPS_JBS; + + //memsize = 0x02800000; /* Trimedia uses memory above */ + memsize = 0x08000000; /* Trimedia uses memory above */ + add_memory_region(0, memsize, BOOT_MEM_RAM); +} diff --git a/arch/mips/philips/pnx8550/jbs/irqmap.c b/arch/mips/philips/pnx8550/jbs/irqmap.c new file mode 100644 index 000000000000..f78e0423dc98 --- /dev/null +++ b/arch/mips/philips/pnx8550/jbs/irqmap.c @@ -0,0 +1,36 @@ +/* + * Philips JBS board irqmap. + * + * Copyright 2005 Embedded Alley Solutions, Inc + * source@embeddealley.com + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License as published by the + * Free Software Foundation; either version 2 of the License, or (at your + * option) any later version. + * + * THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESS OR IMPLIED + * WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF + * MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN + * NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT, + * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT + * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF + * USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON + * ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF + * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + * You should have received a copy of the GNU General Public License along + * with this program; if not, write to the Free Software Foundation, Inc., + * 675 Mass Ave, Cambridge, MA 02139, USA. + */ + +#include +#include + +char irq_tab_jbs[][5] __initdata = { + [8] = { -1, PNX8550_INT_PCI_INTA, 0xff, 0xff, 0xff}, + [9] = { -1, PNX8550_INT_PCI_INTA, 0xff, 0xff, 0xff}, + [17] = { -1, PNX8550_INT_PCI_INTA, 0xff, 0xff, 0xff}, +}; + diff --git a/include/asm-mips/bootinfo.h b/include/asm-mips/bootinfo.h index e5c03c525bda..3870a76efc3b 100644 --- a/include/asm-mips/bootinfo.h +++ b/include/asm-mips/bootinfo.h @@ -137,6 +137,7 @@ #define MACH_GROUP_PHILIPS 14 #define MACH_PHILIPS_NINO 0 /* Nino */ #define MACH_PHILIPS_VELO 1 /* Velo */ +#define MACH_PHILIPS_JBS 2 /* JBS */ /* * Valid machtype for group Globespan diff --git a/include/asm-mips/cpu.h b/include/asm-mips/cpu.h index 3bbb6431d218..ed92c0b14410 100644 --- a/include/asm-mips/cpu.h +++ b/include/asm-mips/cpu.h @@ -52,6 +52,7 @@ #define PRID_IMP_VR41XX 0x0c00 #define PRID_IMP_R12000 0x0e00 #define PRID_IMP_R8000 0x1000 +#define PRID_IMP_PR4450 0x1200 #define PRID_IMP_R4600 0x2000 #define PRID_IMP_R4700 0x2100 #define PRID_IMP_TX39 0x2200 @@ -187,7 +188,8 @@ #define CPU_24K 58 #define CPU_AU1200 59 #define CPU_34K 60 -#define CPU_LAST 60 +#define CPU_PR4450 61 +#define CPU_LAST 61 /* * ISA Level encodings diff --git a/include/asm-mips/mach-pnx8550/cm.h b/include/asm-mips/mach-pnx8550/cm.h new file mode 100644 index 000000000000..bb0a56c7d011 --- /dev/null +++ b/include/asm-mips/mach-pnx8550/cm.h @@ -0,0 +1,43 @@ +/* + * + * BRIEF MODULE DESCRIPTION + * Clock module specific definitions + * + * Author: source@mvista.com + * + * This program is free software; you can distribute it and/or modify it + * under the terms of the GNU General Public License (Version 2) as + * published by the Free Software Foundation. + * + * This program is distributed in the hope it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License + * for more details. + * + * You should have received a copy of the GNU General Public License along + * with this program; if not, write to the Free Software Foundation, Inc., + * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. + */ + +#ifndef __PNX8550_CM_H +#define __PNX8550_CM_H + +#define PNX8550_CM_BASE 0xBBE47000 + +#define PNX8550_CM_PLL0_CTL *(volatile unsigned long *)(PNX8550_CM_BASE + 0x000) +#define PNX8550_CM_PLL1_CTL *(volatile unsigned long *)(PNX8550_CM_BASE + 0x004) +#define PNX8550_CM_PLL2_CTL *(volatile unsigned long *)(PNX8550_CM_BASE + 0x008) +#define PNX8550_CM_PLL3_CTL *(volatile unsigned long *)(PNX8550_CM_BASE + 0x00C) + +// Table not complete..... + +#define PNX8550_CM_PLL_BLOCKED_MASK 0x80000000 +#define PNX8550_CM_PLL_LOCK_MASK 0x40000000 +#define PNX8550_CM_PLL_CURRENT_ADJ_MASK 0x3c000000 +#define PNX8550_CM_PLL_N_MASK 0x01ff0000 +#define PNX8550_CM_PLL_M_MASK 0x00003f00 +#define PNX8550_CM_PLL_P_MASK 0x0000000c +#define PNX8550_CM_PLL_PD_MASK 0x00000002 + + +#endif diff --git a/include/asm-mips/mach-pnx8550/glb.h b/include/asm-mips/mach-pnx8550/glb.h new file mode 100644 index 000000000000..07aa85e609bc --- /dev/null +++ b/include/asm-mips/mach-pnx8550/glb.h @@ -0,0 +1,86 @@ +/* + * + * BRIEF MODULE DESCRIPTION + * PNX8550 global definitions + * + * Author: source@mvista.com + * + * This program is free software; you can distribute it and/or modify it + * under the terms of the GNU General Public License (Version 2) as + * published by the Free Software Foundation. + * + * This program is distributed in the hope it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License + * for more details. + * + * You should have received a copy of the GNU General Public License along + * with this program; if not, write to the Free Software Foundation, Inc., + * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. + */ + +#ifndef __PNX8550_GLB_H +#define __PNX8550_GLB_H + +#define PNX8550_GLB1_BASE 0xBBE63000 +#define PNX8550_GLB2_BASE 0xBBE4d000 +#define PNX8550_RESET_BASE 0xBBE60000 + +/* PCI Inta Output Enable Registers */ +#define PNX8550_GLB2_ENAB_INTA_O *(volatile unsigned long *)(PNX8550_GLB2_BASE + 0x050) + +/* Bit 1:Enable DAC Powerdown + 0:DACs are enabled and are working normally + 1:DACs are powerdown +*/ +#define PNX8550_GLB_DAC_PD 0x2 +/* Bit 0:Enable of PCI inta output + 0 = Disable PCI inta output + 1 = Enable PCI inta output +*/ +#define PNX8550_GLB_ENABLE_INTA_O 0x1 + +/* PCI Direct Mappings */ +#define PNX8550_PCIMEM 0x12000000 +#define PNX8550_PCIMEM_SIZE 0x08000000 +#define PNX8550_PCIIO 0x1c000000 +#define PNX8550_PCIIO_SIZE 0x02000000 /* 32M */ + +#define PNX8550_PORT_BASE KSEG1 + +// GPIO def +#define PNX8550_GPIO_BASE 0x1Be00000 + +#define PNX8550_GPIO_DIRQ0 (PNX8550_GPIO_BASE + 0x104500) +#define PNX8550_GPIO_MC1 (PNX8550_GPIO_BASE + 0x104004) +#define PNX8550_GPIO_MC_31_BIT 30 +#define PNX8550_GPIO_MC_30_BIT 28 +#define PNX8550_GPIO_MC_29_BIT 26 +#define PNX8550_GPIO_MC_28_BIT 24 +#define PNX8550_GPIO_MC_27_BIT 22 +#define PNX8550_GPIO_MC_26_BIT 20 +#define PNX8550_GPIO_MC_25_BIT 18 +#define PNX8550_GPIO_MC_24_BIT 16 +#define PNX8550_GPIO_MC_23_BIT 14 +#define PNX8550_GPIO_MC_22_BIT 12 +#define PNX8550_GPIO_MC_21_BIT 10 +#define PNX8550_GPIO_MC_20_BIT 8 +#define PNX8550_GPIO_MC_19_BIT 6 +#define PNX8550_GPIO_MC_18_BIT 4 +#define PNX8550_GPIO_MC_17_BIT 2 +#define PNX8550_GPIO_MC_16_BIT 0 + +#define PNX8550_GPIO_MODE_PRIMOP 0x1 +#define PNX8550_GPIO_MODE_NO_OPENDR 0x2 +#define PNX8550_GPIO_MODE_OPENDR 0x3 + +// RESET module +#define PNX8550_RST_CTL *(volatile unsigned long *)(PNX8550_RESET_BASE + 0x0) +#define PNX8550_RST_CAUSE *(volatile unsigned long *)(PNX8550_RESET_BASE + 0x4) +#define PNX8550_RST_EN_WATCHDOG *(volatile unsigned long *)(PNX8550_RESET_BASE + 0x8) + +#define PNX8550_RST_REL_MIPS_RST_N 0x8 +#define PNX8550_RST_DO_SW_RST 0x4 +#define PNX8550_RST_REL_SYS_RST_OUT 0x2 +#define PNX8550_RST_ASSERT_SYS_RST_OUT 0x1 +#endif diff --git a/include/asm-mips/mach-pnx8550/int.h b/include/asm-mips/mach-pnx8550/int.h new file mode 100644 index 000000000000..0e0668b524f4 --- /dev/null +++ b/include/asm-mips/mach-pnx8550/int.h @@ -0,0 +1,140 @@ +/* + * + * BRIEF MODULE DESCRIPTION + * Interrupt specific definitions + * + * Author: source@mvista.com + * + * This program is free software; you can distribute it and/or modify it + * under the terms of the GNU General Public License (Version 2) as + * published by the Free Software Foundation. + * + * This program is distributed in the hope it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License + * for more details. + * + * You should have received a copy of the GNU General Public License along + * with this program; if not, write to the Free Software Foundation, Inc., + * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. + */ + +#ifndef __PNX8550_INT_H +#define __PNX8550_INT_H + +#define PNX8550_GIC_BASE 0xBBE3E000 + +#define PNX8550_GIC_PRIMASK_0 *(volatile unsigned long *)(PNX8550_GIC_BASE + 0x000) +#define PNX8550_GIC_PRIMASK_1 *(volatile unsigned long *)(PNX8550_GIC_BASE + 0x004) +#define PNX8550_GIC_VECTOR_0 *(volatile unsigned long *)(PNX8550_GIC_BASE + 0x100) +#define PNX8550_GIC_VECTOR_1 *(volatile unsigned long *)(PNX8550_GIC_BASE + 0x104) +#define PNX8550_GIC_PEND_1_31 *(volatile unsigned long *)(PNX8550_GIC_BASE + 0x200) +#define PNX8550_GIC_PEND_32_63 *(volatile unsigned long *)(PNX8550_GIC_BASE + 0x204) +#define PNX8550_GIC_PEND_64_70 *(volatile unsigned long *)(PNX8550_GIC_BASE + 0x208) +#define PNX8550_GIC_FEATURES *(volatile unsigned long *)(PNX8550_GIC_BASE + 0x300) +#define PNX8550_GIC_REQ(x) *(volatile unsigned long *)(PNX8550_GIC_BASE + 0x400 + (x)*4) +#define PNX8550_GIC_MOD_ID *(volatile unsigned long *)(PNX8550_GIC_BASE + 0xFFC) + +// cp0 is two software + six hw exceptions +#define PNX8550_INT_CP0_TOTINT 8 +#define PNX8550_INT_CP0_MIN 0 +#define PNX8550_INT_CP0_MAX (PNX8550_INT_CP0_MIN + PNX8550_INT_CP0_TOTINT - 1) + +#define MIPS_CPU_GIC_IRQ 2 +#define MIPS_CPU_TIMER_IRQ 7 + +// GIC are 71 exceptions connected to cp0's first hardware exception +#define PNX8550_INT_GIC_TOTINT 71 +#define PNX8550_INT_GIC_MIN (PNX8550_INT_CP0_MAX+1) +#define PNX8550_INT_GIC_MAX (PNX8550_INT_GIC_MIN + PNX8550_INT_GIC_TOTINT - 1) + +#define PNX8550_INT_UNDEF (PNX8550_INT_GIC_MIN+0) +#define PNX8550_INT_IPC_TARGET0_MIPS (PNX8550_INT_GIC_MIN+1) +#define PNX8550_INT_IPC_TARGET1_TM32_1 (PNX8550_INT_GIC_MIN+2) +#define PNX8550_INT_IPC_TARGET1_TM32_2 (PNX8550_INT_GIC_MIN+3) +#define PNX8550_INT_RESERVED_4 (PNX8550_INT_GIC_MIN+4) +#define PNX8550_INT_USB (PNX8550_INT_GIC_MIN+5) +#define PNX8550_INT_GPIO_EQ1 (PNX8550_INT_GIC_MIN+6) +#define PNX8550_INT_GPIO_EQ2 (PNX8550_INT_GIC_MIN+7) +#define PNX8550_INT_GPIO_EQ3 (PNX8550_INT_GIC_MIN+8) +#define PNX8550_INT_GPIO_EQ4 (PNX8550_INT_GIC_MIN+9) + +#define PNX8550_INT_GPIO_EQ5 (PNX8550_INT_GIC_MIN+10) +#define PNX8550_INT_GPIO_EQ6 (PNX8550_INT_GIC_MIN+11) +#define PNX8550_INT_RESERVED_12 (PNX8550_INT_GIC_MIN+12) +#define PNX8550_INT_QVCP1 (PNX8550_INT_GIC_MIN+13) +#define PNX8550_INT_QVCP2 (PNX8550_INT_GIC_MIN+14) +#define PNX8550_INT_I2C1 (PNX8550_INT_GIC_MIN+15) +#define PNX8550_INT_I2C2 (PNX8550_INT_GIC_MIN+16) +#define PNX8550_INT_ISO_UART1 (PNX8550_INT_GIC_MIN+17) +#define PNX8550_INT_ISO_UART2 (PNX8550_INT_GIC_MIN+18) +#define PNX8550_INT_UART1 (PNX8550_INT_GIC_MIN+19) + +#define PNX8550_INT_UART2 (PNX8550_INT_GIC_MIN+20) +#define PNX8550_INT_QNTR (PNX8550_INT_GIC_MIN+21) +#define PNX8550_INT_RESERVED22 (PNX8550_INT_GIC_MIN+22) +#define PNX8550_INT_T_DSC (PNX8550_INT_GIC_MIN+23) +#define PNX8550_INT_M_DSC (PNX8550_INT_GIC_MIN+24) +#define PNX8550_INT_RESERVED25 (PNX8550_INT_GIC_MIN+25) +#define PNX8550_INT_2D_DRAW_ENG (PNX8550_INT_GIC_MIN+26) +#define PNX8550_INT_MEM_BASED_SCALAR1 (PNX8550_INT_GIC_MIN+27) +#define PNX8550_INT_VIDEO_MPEG (PNX8550_INT_GIC_MIN+28) +#define PNX8550_INT_VIDEO_INPUT_P1 (PNX8550_INT_GIC_MIN+29) + +#define PNX8550_INT_VIDEO_INPUT_P2 (PNX8550_INT_GIC_MIN+30) +#define PNX8550_INT_SPDI1 (PNX8550_INT_GIC_MIN+31) +#define PNX8550_INT_SPDO (PNX8550_INT_GIC_MIN+32) +#define PNX8550_INT_AUDIO_INPUT1 (PNX8550_INT_GIC_MIN+33) +#define PNX8550_INT_AUDIO_OUTPUT1 (PNX8550_INT_GIC_MIN+34) +#define PNX8550_INT_AUDIO_INPUT2 (PNX8550_INT_GIC_MIN+35) +#define PNX8550_INT_AUDIO_OUTPUT2 (PNX8550_INT_GIC_MIN+36) +#define PNX8550_INT_MEMBASED_SCALAR2 (PNX8550_INT_GIC_MIN+37) +#define PNX8550_INT_VPK (PNX8550_INT_GIC_MIN+38) +#define PNX8550_INT_MPEG1_MIPS (PNX8550_INT_GIC_MIN+39) + +#define PNX8550_INT_MPEG1_TM (PNX8550_INT_GIC_MIN+40) +#define PNX8550_INT_MPEG2_MIPS (PNX8550_INT_GIC_MIN+41) +#define PNX8550_INT_MPEG2_TM (PNX8550_INT_GIC_MIN+42) +#define PNX8550_INT_TS_DMA (PNX8550_INT_GIC_MIN+43) +#define PNX8550_INT_EDMA (PNX8550_INT_GIC_MIN+44) +#define PNX8550_INT_TM_DEBUG1 (PNX8550_INT_GIC_MIN+45) +#define PNX8550_INT_TM_DEBUG2 (PNX8550_INT_GIC_MIN+46) +#define PNX8550_INT_PCI_INTA (PNX8550_INT_GIC_MIN+47) +#define PNX8550_INT_CLOCK_MODULE (PNX8550_INT_GIC_MIN+48) +#define PNX8550_INT_PCI_XIO_INTA_PCI (PNX8550_INT_GIC_MIN+49) + +#define PNX8550_INT_PCI_XIO_INTB_DMA (PNX8550_INT_GIC_MIN+50) +#define PNX8550_INT_PCI_XIO_INTC_GPPM (PNX8550_INT_GIC_MIN+51) +#define PNX8550_INT_PCI_XIO_INTD_GPXIO (PNX8550_INT_GIC_MIN+52) +#define PNX8550_INT_DVD_CSS (PNX8550_INT_GIC_MIN+53) +#define PNX8550_INT_VLD (PNX8550_INT_GIC_MIN+54) +#define PNX8550_INT_GPIO_TSU_7_0 (PNX8550_INT_GIC_MIN+55) +#define PNX8550_INT_GPIO_TSU_15_8 (PNX8550_INT_GIC_MIN+56) +#define PNX8550_INT_GPIO_CTU_IR (PNX8550_INT_GIC_MIN+57) +#define PNX8550_INT_GPIO0 (PNX8550_INT_GIC_MIN+58) +#define PNX8550_INT_GPIO1 (PNX8550_INT_GIC_MIN+59) + +#define PNX8550_INT_GPIO2 (PNX8550_INT_GIC_MIN+60) +#define PNX8550_INT_GPIO3 (PNX8550_INT_GIC_MIN+61) +#define PNX8550_INT_GPIO4 (PNX8550_INT_GIC_MIN+62) +#define PNX8550_INT_GPIO5 (PNX8550_INT_GIC_MIN+63) +#define PNX8550_INT_GPIO6 (PNX8550_INT_GIC_MIN+64) +#define PNX8550_INT_GPIO7 (PNX8550_INT_GIC_MIN+65) +#define PNX8550_INT_PMAN_SECURITY (PNX8550_INT_GIC_MIN+66) +#define PNX8550_INT_I2C3 (PNX8550_INT_GIC_MIN+67) +#define PNX8550_INT_RESERVED_68 (PNX8550_INT_GIC_MIN+68) +#define PNX8550_INT_SPDI2 (PNX8550_INT_GIC_MIN+69) + +#define PNX8550_INT_I2C4 (PNX8550_INT_GIC_MIN+70) + +// Timer are 3 exceptions connected to cp0's 7th hardware exception +#define PNX8550_INT_TIMER_TOTINT 3 +#define PNX8550_INT_TIMER_MIN (PNX8550_INT_GIC_MAX+1) +#define PNX8550_INT_TIMER_MAX (PNX8550_INT_TIMER_MIN + PNX8550_INT_TIMER_TOTINT - 1) + +#define PNX8550_INT_TIMER1 (PNX8550_INT_TIMER_MIN+0) +#define PNX8550_INT_TIMER2 (PNX8550_INT_TIMER_MIN+1) +#define PNX8550_INT_TIMER3 (PNX8550_INT_TIMER_MIN+2) +#define PNX8550_INT_WATCHDOG PNX8550_INT_TIMER3 + +#endif diff --git a/include/asm-mips/mach-pnx8550/kernel-entry-init.h b/include/asm-mips/mach-pnx8550/kernel-entry-init.h new file mode 100644 index 000000000000..57102fa9da51 --- /dev/null +++ b/include/asm-mips/mach-pnx8550/kernel-entry-init.h @@ -0,0 +1,262 @@ +/* + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + * + * Copyright (C) 2005 Embedded Alley Solutions, Inc + */ +#ifndef __ASM_MACH_KERNEL_ENTRY_INIT_H +#define __ASM_MACH_KERNEL_ENTRY_INIT_H + +#include +#include + +#define CO_CONFIGPR_VALID 0x3F1F41FF /* valid bits to write to ConfigPR */ +#define HAZARD_CP0 nop; nop; nop; nop; nop; nop; nop; nop; nop; nop; nop; nop; +#define CACHE_OPC 0xBC000000 /* MIPS cache instruction opcode */ +#define ICACHE_LINE_SIZE 32 /* Instruction cache line size bytes */ +#define DCACHE_LINE_SIZE 32 /* Data cache line size in bytes */ + +#define ICACHE_SET_COUNT 256 /* Instruction cache set count */ +#define DCACHE_SET_COUNT 128 /* Data cache set count */ + +#define ICACHE_SET_SIZE (ICACHE_SET_COUNT * ICACHE_LINE_SIZE) +#define DCACHE_SET_SIZE (DCACHE_SET_COUNT * DCACHE_LINE_SIZE) + + .macro kernel_entry_setup + .set push + .set noreorder + /* + * PNX8550 entry point, when running a non compressed + * kernel. When loading a zImage, the head.S code in + * arch/mips/zboot/pnx8550 will init the caches and, + * decompress the kernel, and branch to kernel_entry. + */ +cache_begin: li t0, (1<<28) + mtc0 t0, CP0_STATUS /* cp0 usable */ + HAZARD_CP0 + + mtc0 zero, CP0_CAUSE + HAZARD_CP0 + + + /* Set static virtual to phys address translation and TLB disabled */ + mfc0 t0, CP0_CONFIG, 7 + HAZARD_CP0 + + and t0,~((1<<19) | (1<<20)) /* TLB/MAP cleared */ + mtc0 t0, CP0_CONFIG, 7 + HAZARD_CP0 + + /* CPU boots with kseg0 cache algo set to 0x2 -- uncached */ + + init_icache + nop + init_dcache + nop + + cachePr4450ICReset + nop + + cachePr4450DCReset + nop + + /* read ConfigPR into t0 */ + mfc0 t0, CP0_CONFIG, 7 + HAZARD_CP0 + + /* enable the TLB */ + or t0, (1<<19) + + /* disable the ICACHE: at least 10x slower */ + /* or t0, (1<<26) */ + + /* disable the DCACHE; CONFIG_CPU_HAS_LLSC should not be set */ + /* or t0, (1<<27) */ + + and t0, CO_CONFIGPR_VALID + + /* enable TLB. */ + mtc0 t0, CP0_CONFIG, 7 + HAZARD_CP0 +cache_end: + /* Setup CMEM_0 to MMIO address space, 2MB */ + lui t0, 0x1BE0 + addi t0, t0, 0x3 + mtc0 $8, $22, 4 + nop + + /* Setup CMEM_1, 128MB */ + lui t0, 0x1000 + addi t0, t0, 0xf + mtc0 $8, $22, 5 + nop + + + /* Setup CMEM_2, 32MB */ + lui t0, 0x1C00 + addi t0, t0, 0xb + mtc0 $8, $22, 6 + nop + + /* Setup CMEM_3, 0MB */ + lui t0, 0x0 + addi t0, t0, 0x0 + mtc0 $8, $22, 7 + nop + + /* Enable cache */ + mfc0 t0, CP0_CONFIG + HAZARD_CP0 + and t0, t0, 0xFFFFFFF8 + or t0, t0, 3 + mtc0 t0, CP0_CONFIG + HAZARD_CP0 + .set pop + .endm + + .macro init_icache + .set push + .set noreorder + + /* Get Cache Configuration */ + mfc0 t3, CP0_CONFIG, 1 + HAZARD_CP0 + + /* get cache Line size */ + + srl t1, t3, 19 /* C0_CONFIGPR_IL_SHIFT */ + andi t1, t1, 0x7 /* C0_CONFIGPR_IL_MASK */ + beq t1, zero, pr4450_instr_cache_invalidated /* if zero instruction cache is absent */ + nop + addiu t0, t1, 1 + ori t1, zero, 1 + sllv t1, t1, t0 + + /* get max cache Index */ + srl t2, t3, 22 /* C0_CONFIGPR_IS_SHIFT */ + andi t2, t2, 0x7 /* C0_CONFIGPR_IS_MASK */ + addiu t0, t2, 6 + ori t2, zero, 1 + sllv t2, t2, t0 + + /* get max cache way */ + srl t3, t3, 16 /* C0_CONFIGPR_IA_SHIFT */ + andi t3, t3, 0x7 /* C0_CONFIGPR_IA_MASK */ + addiu t3, t3, 1 + + /* total no of cache lines */ + multu t2, t3 /* max index * max way */ + mflo t2 + addiu t2, t2, -1 + + move t0, zero +pr4450_next_instruction_cache_set: + cache Index_Invalidate_I, 0(t0) + addu t0, t0, t1 /* add bytes in a line */ + bne t2, zero, pr4450_next_instruction_cache_set + addiu t2, t2, -1 /* reduce no of lines to invalidate by one */ +pr4450_instr_cache_invalidated: + .set pop + .endm + + .macro init_dcache + .set push + .set noreorder + move t1, zero + + /* Store Tag Information */ + mtc0 zero, CP0_TAGLO, 0 + HAZARD_CP0 + + mtc0 zero, CP0_TAGHI, 0 + HAZARD_CP0 + + /* Cache size is 16384 = 512 lines x 32 bytes per line */ + or t2, zero, (128*4)-1 /* 512 lines */ + /* Invalidate all lines */ +2: + cache Index_Store_Tag_D, 0(t1) + addiu t2, t2, -1 + bne t2, zero, 2b + addiu t1, t1, 32 /* 32 bytes in a line */ + .set pop + .endm + + .macro cachePr4450ICReset + .set push + .set noreorder + + /* Save CP0 status reg on entry; */ + /* disable interrupts during cache reset */ + mfc0 t0, CP0_STATUS /* T0 = interrupt status on entry */ + HAZARD_CP0 + + mtc0 zero, CP0_STATUS /* disable CPU interrupts */ + HAZARD_CP0 + + or t1, zero, zero /* T1 = starting cache index (0) */ + ori t2, zero, (256 - 1) /* T2 = inst cache set cnt - 1 */ + + icache_invd_loop: + /* 9 == register t1 */ + .word (CACHE_OPC | (9 << 21) | (Index_Invalidate_I << 16) | \ + (0 * ICACHE_SET_SIZE)) /* invalidate inst cache WAY0 */ + .word (CACHE_OPC | (9 << 21) | (Index_Invalidate_I << 16) | \ + (1 * ICACHE_SET_SIZE)) /* invalidate inst cache WAY1 */ + + addiu t1, t1, ICACHE_LINE_SIZE /* T1 = next cache line index */ + bne t2, zero, icache_invd_loop /* T2 = 0 if all sets invalidated */ + addiu t2, t2, -1 /* decrement T2 set cnt (delay slot) */ + + /* Initialize the latches in the instruction cache tag */ + /* that drive the way selection tri-state bus drivers, by doing a */ + /* dummy load while the instruction cache is still disabled. */ + /* TODO: Is this needed ? */ + la t1, KSEG0 /* T1 = cached memory base address */ + lw zero, 0x0000(t1) /* (dummy read of first memory word) */ + + mtc0 t0, CP0_STATUS /* restore interrupt status on entry */ + HAZARD_CP0 + .set pop + .endm + + .macro cachePr4450DCReset + .set push + .set noreorder + mfc0 t0, CP0_STATUS /* T0 = interrupt status on entry */ + HAZARD_CP0 + mtc0 zero, CP0_STATUS /* disable CPU interrupts */ + HAZARD_CP0 + + /* Writeback/invalidate entire data cache sets/ways/lines */ + or t1, zero, zero /* T1 = starting cache index (0) */ + ori t2, zero, (DCACHE_SET_COUNT - 1) /* T2 = data cache set cnt - 1 */ + + dcache_wbinvd_loop: + /* 9 == register t1 */ + .word (CACHE_OPC | (9 << 21) | (Index_Writeback_Inv_D << 16) | \ + (0 * DCACHE_SET_SIZE)) /* writeback/invalidate WAY0 */ + .word (CACHE_OPC | (9 << 21) | (Index_Writeback_Inv_D << 16) | \ + (1 * DCACHE_SET_SIZE)) /* writeback/invalidate WAY1 */ + .word (CACHE_OPC | (9 << 21) | (Index_Writeback_Inv_D << 16) | \ + (2 * DCACHE_SET_SIZE)) /* writeback/invalidate WAY2 */ + .word (CACHE_OPC | (9 << 21) | (Index_Writeback_Inv_D << 16) | \ + (3 * DCACHE_SET_SIZE)) /* writeback/invalidate WAY3 */ + + addiu t1, t1, DCACHE_LINE_SIZE /* T1 = next data cache line index */ + bne t2, zero, dcache_wbinvd_loop /* T2 = 0 when wbinvd entire cache */ + addiu t2, t2, -1 /* decrement T2 set cnt (delay slot) */ + + /* Initialize the latches in the data cache tag that drive the way + selection tri-state bus drivers, by doing a dummy load while the + data cache is still in the disabled mode. TODO: Is this needed ? */ + la t1, KSEG0 /* T1 = cached memory base address */ + lw zero, 0x0000(t1) /* (dummy read of first memory word) */ + + mtc0 t0, CP0_STATUS /* restore interrupt status on entry */ + HAZARD_CP0 + .set pop + .endm + +#endif /* __ASM_MACH_KERNEL_ENTRY_INIT_H */ diff --git a/include/asm-mips/mach-pnx8550/nand.h b/include/asm-mips/mach-pnx8550/nand.h new file mode 100644 index 000000000000..aefbc514ab09 --- /dev/null +++ b/include/asm-mips/mach-pnx8550/nand.h @@ -0,0 +1,121 @@ +#ifndef __PNX8550_NAND_H +#define __PNX8550_NAND_H + +#define PNX8550_NAND_BASE_ADDR 0x10000000 +#define PNX8550_PCIXIO_BASE 0xBBE40000 + +#define PNX8550_DMA_EXT_ADDR *(volatile unsigned long *)(PNX8550_PCIXIO_BASE + 0x800) +#define PNX8550_DMA_INT_ADDR *(volatile unsigned long *)(PNX8550_PCIXIO_BASE + 0x804) +#define PNX8550_DMA_TRANS_SIZE *(volatile unsigned long *)(PNX8550_PCIXIO_BASE + 0x808) +#define PNX8550_DMA_CTRL *(volatile unsigned long *)(PNX8550_PCIXIO_BASE + 0x80c) +#define PNX8550_XIO_SEL0 *(volatile unsigned long *)(PNX8550_PCIXIO_BASE + 0x814) +#define PNX8550_GPXIO_ADDR *(volatile unsigned long *)(PNX8550_PCIXIO_BASE + 0x820) +#define PNX8550_GPXIO_WR *(volatile unsigned long *)(PNX8550_PCIXIO_BASE + 0x824) +#define PNX8550_GPXIO_RD *(volatile unsigned long *)(PNX8550_PCIXIO_BASE + 0x828) +#define PNX8550_GPXIO_CTRL *(volatile unsigned long *)(PNX8550_PCIXIO_BASE + 0x82C) +#define PNX8550_XIO_FLASH_CTRL *(volatile unsigned long *)(PNX8550_PCIXIO_BASE + 0x830) +#define PNX8550_GPXIO_INT_STATUS *(volatile unsigned long *)(PNX8550_PCIXIO_BASE + 0xfb0) +#define PNX8550_GPXIO_INT_ENABLE *(volatile unsigned long *)(PNX8550_PCIXIO_BASE + 0xfb4) +#define PNX8550_GPXIO_INT_CLEAR *(volatile unsigned long *)(PNX8550_PCIXIO_BASE + 0xfb8) +#define PNX8550_DMA_INT_STATUS *(volatile unsigned long *)(PNX8550_PCIXIO_BASE + 0xfd0) +#define PNX8550_DMA_INT_ENABLE *(volatile unsigned long *)(PNX8550_PCIXIO_BASE + 0xfd4) +#define PNX8550_DMA_INT_CLEAR *(volatile unsigned long *)(PNX8550_PCIXIO_BASE + 0xfd8) + +#define PNX8550_XIO_SEL0_EN_16BIT 0x00800000 +#define PNX8550_XIO_SEL0_USE_ACK 0x00400000 +#define PNX8550_XIO_SEL0_REN_HIGH 0x00100000 +#define PNX8550_XIO_SEL0_REN_LOW 0x00040000 +#define PNX8550_XIO_SEL0_WEN_HIGH 0x00010000 +#define PNX8550_XIO_SEL0_WEN_LOW 0x00004000 +#define PNX8550_XIO_SEL0_WAIT 0x00000200 +#define PNX8550_XIO_SEL0_OFFSET 0x00000020 +#define PNX8550_XIO_SEL0_TYPE_68360 0x00000000 +#define PNX8550_XIO_SEL0_TYPE_NOR 0x00000008 +#define PNX8550_XIO_SEL0_TYPE_NAND 0x00000010 +#define PNX8550_XIO_SEL0_TYPE_IDE 0x00000018 +#define PNX8550_XIO_SEL0_SIZE_8MB 0x00000000 +#define PNX8550_XIO_SEL0_SIZE_16MB 0x00000002 +#define PNX8550_XIO_SEL0_SIZE_32MB 0x00000004 +#define PNX8550_XIO_SEL0_SIZE_64MB 0x00000006 +#define PNX8550_XIO_SEL0_ENAB 0x00000001 + +#define PNX8550_SEL0_DEFAULT ((PNX8550_XIO_SEL0_EN_16BIT) | \ + (PNX8550_XIO_SEL0_REN_HIGH*0)| \ + (PNX8550_XIO_SEL0_REN_LOW*2) | \ + (PNX8550_XIO_SEL0_WEN_HIGH*0)| \ + (PNX8550_XIO_SEL0_WEN_LOW*2) | \ + (PNX8550_XIO_SEL0_WAIT*4) | \ + (PNX8550_XIO_SEL0_OFFSET*0) | \ + (PNX8550_XIO_SEL0_TYPE_NAND) | \ + (PNX8550_XIO_SEL0_SIZE_32MB) | \ + (PNX8550_XIO_SEL0_ENAB)) + +#define PNX8550_GPXIO_PENDING 0x00000200 +#define PNX8550_GPXIO_DONE 0x00000100 +#define PNX8550_GPXIO_CLR_DONE 0x00000080 +#define PNX8550_GPXIO_INIT 0x00000040 +#define PNX8550_GPXIO_READ_CMD 0x00000010 +#define PNX8550_GPXIO_BEN 0x0000000F + +#define PNX8550_XIO_FLASH_64MB 0x00200000 +#define PNX8550_XIO_FLASH_INC_DATA 0x00100000 +#define PNX8550_XIO_FLASH_CMD_PH 0x000C0000 +#define PNX8550_XIO_FLASH_CMD_PH2 0x00080000 +#define PNX8550_XIO_FLASH_CMD_PH1 0x00040000 +#define PNX8550_XIO_FLASH_CMD_PH0 0x00000000 +#define PNX8550_XIO_FLASH_ADR_PH 0x00030000 +#define PNX8550_XIO_FLASH_ADR_PH3 0x00030000 +#define PNX8550_XIO_FLASH_ADR_PH2 0x00020000 +#define PNX8550_XIO_FLASH_ADR_PH1 0x00010000 +#define PNX8550_XIO_FLASH_ADR_PH0 0x00000000 +#define PNX8550_XIO_FLASH_CMD_B(x) ((x<<8) & 0x0000FF00) +#define PNX8550_XIO_FLASH_CMD_A(x) (x & 0x000000FF) + +#define PNX8550_XIO_INT_ACK 0x00004000 +#define PNX8550_XIO_INT_COMPL 0x00002000 +#define PNX8550_XIO_INT_NONSUP 0x00000200 +#define PNX8550_XIO_INT_ABORT 0x00000004 + +#define PNX8550_DMA_CTRL_SINGLE_DATA 0x00000400 +#define PNX8550_DMA_CTRL_SND2XIO 0x00000200 +#define PNX8550_DMA_CTRL_FIX_ADDR 0x00000100 +#define PNX8550_DMA_CTRL_BURST_8 0x00000000 +#define PNX8550_DMA_CTRL_BURST_16 0x00000020 +#define PNX8550_DMA_CTRL_BURST_32 0x00000040 +#define PNX8550_DMA_CTRL_BURST_64 0x00000060 +#define PNX8550_DMA_CTRL_BURST_128 0x00000080 +#define PNX8550_DMA_CTRL_BURST_256 0x000000A0 +#define PNX8550_DMA_CTRL_BURST_512 0x000000C0 +#define PNX8550_DMA_CTRL_BURST_NORES 0x000000E0 +#define PNX8550_DMA_CTRL_INIT_DMA 0x00000010 +#define PNX8550_DMA_CTRL_CMD_TYPE 0x0000000F + +/* see PCI system arch, page 100 for the full list: */ +#define PNX8550_DMA_CTRL_PCI_CMD_READ 0x00000006 +#define PNX8550_DMA_CTRL_PCI_CMD_WRITE 0x00000007 + +#define PNX8550_DMA_INT_STAT_ACK_DONE (1<<14) +#define PNX8550_DMA_INT_STAT_DMA_DONE (1<<12) +#define PNX8550_DMA_INT_STAT_DMA_ERR (1<<9) +#define PNX8550_DMA_INT_STAT_PERR5 (1<<5) +#define PNX8550_DMA_INT_STAT_PERR4 (1<<4) +#define PNX8550_DMA_INT_STAT_M_ABORT (1<<2) +#define PNX8550_DMA_INT_STAT_T_ABORT (1<<1) + +#define PNX8550_DMA_INT_EN_ACK_DONE (1<<14) +#define PNX8550_DMA_INT_EN_DMA_DONE (1<<12) +#define PNX8550_DMA_INT_EN_DMA_ERR (1<<9) +#define PNX8550_DMA_INT_EN_PERR5 (1<<5) +#define PNX8550_DMA_INT_EN_PERR4 (1<<4) +#define PNX8550_DMA_INT_EN_M_ABORT (1<<2) +#define PNX8550_DMA_INT_EN_T_ABORT (1<<1) + +#define PNX8550_DMA_INT_CLR_ACK_DONE (1<<14) +#define PNX8550_DMA_INT_CLR_DMA_DONE (1<<12) +#define PNX8550_DMA_INT_CLR_DMA_ERR (1<<9) +#define PNX8550_DMA_INT_CLR_PERR5 (1<<5) +#define PNX8550_DMA_INT_CLR_PERR4 (1<<4) +#define PNX8550_DMA_INT_CLR_M_ABORT (1<<2) +#define PNX8550_DMA_INT_CLR_T_ABORT (1<<1) + +#endif diff --git a/include/asm-mips/mach-pnx8550/pci.h b/include/asm-mips/mach-pnx8550/pci.h new file mode 100644 index 000000000000..b921508d701b --- /dev/null +++ b/include/asm-mips/mach-pnx8550/pci.h @@ -0,0 +1,185 @@ +/* + * + * BRIEF MODULE DESCRIPTION + * PCI specific definitions + * + * Author: source@mvista.com + * + * This program is free software; you can distribute it and/or modify it + * under the terms of the GNU General Public License (Version 2) as + * published by the Free Software Foundation. + * + * This program is distributed in the hope it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License + * for more details. + * + * You should have received a copy of the GNU General Public License along + * with this program; if not, write to the Free Software Foundation, Inc., + * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. + */ + +#ifndef __PNX8550_PCI_H +#define __PNX8550_PCI_H + +#include +#include +#include +#include + +#define PCI_ACCESS_READ 0 +#define PCI_ACCESS_WRITE 1 + +#define PCI_CMD_IOR 0x20 +#define PCI_CMD_IOW 0x30 +#define PCI_CMD_CONFIG_READ 0xa0 +#define PCI_CMD_CONFIG_WRITE 0xb0 + +#define PCI_IO_TIMEOUT 1000 +#define PCI_IO_RETRY 5 +/* Timeout for IO and CFG accesses. + This is in 1/1024 th of a jiffie(=10ms) + i.e. approx 10us */ +#define PCI_IO_JIFFIES_TIMEOUT 40 +#define PCI_IO_JIFFIES_SHIFT 10 + +#define PCI_BYTE_ENABLE_MASK 0x0000000f +#define PCI_CFG_BUS_SHIFT 16 +#define PCI_CFG_FUNC_SHIFT 8 +#define PCI_CFG_REG_SHIFT 2 + +#define PCI_BASE 0x1be00000 +#define PCI_SETUP 0x00040010 +#define PCI_DIS_REQGNT (1<<30) +#define PCI_DIS_REQGNTA (1<<29) +#define PCI_DIS_REQGNTB (1<<28) +#define PCI_D2_SUPPORT (1<<27) +#define PCI_D1_SUPPORT (1<<26) +#define PCI_EN_TA (1<<24) +#define PCI_EN_PCI2MMI (1<<23) +#define PCI_EN_XIO (1<<22) +#define PCI_BASE18_PREF (1<<21) +#define SIZE_16M 0x3 +#define SIZE_32M 0x4 +#define SIZE_64M 0x5 +#define SIZE_128M 0x6 +#define PCI_SETUP_BASE18_SIZE(X) (X<<18) +#define PCI_SETUP_BASE18_EN (1<<17) +#define PCI_SETUP_BASE14_PREF (1<<16) +#define PCI_SETUP_BASE14_SIZE(X) (X<<12) +#define PCI_SETUP_BASE14_EN (1<<11) +#define PCI_SETUP_BASE10_PREF (1<<10) +#define PCI_SETUP_BASE10_SIZE(X) (X<<7) +#define PCI_SETUP_CFGMANAGE_EN (1<<1) +#define PCI_SETUP_PCIARB_EN (1<<0) + +#define PCI_CTRL 0x040014 +#define PCI_SWPB_DCS_PCI (1<<16) +#define PCI_SWPB_PCI_PCI (1<<15) +#define PCI_SWPB_PCI_DCS (1<<14) +#define PCI_REG_WR_POST (1<<13) +#define PCI_XIO_WR_POST (1<<12) +#define PCI_PCI2_WR_POST (1<<13) +#define PCI_PCI1_WR_POST (1<<12) +#define PCI_SERR_SEEN (1<<11) +#define PCI_B10_SPEC_RD (1<<6) +#define PCI_B14_SPEC_RD (1<<5) +#define PCI_B18_SPEC_RD (1<<4) +#define PCI_B10_NOSUBWORD (1<<3) +#define PCI_B14_NOSUBWORD (1<<2) +#define PCI_B18_NOSUBWORD (1<<1) +#define PCI_RETRY_TMREN (1<<0) + +#define PCI_BASE1_LO 0x040018 +#define PCI_BASE1_HI 0x04001C +#define PCI_BASE2_LO 0x040020 +#define PCI_BASE2_HI 0x040024 +#define PCI_RDLIFETIM 0x040028 +#define PCI_GPPM_ADDR 0x04002C +#define PCI_GPPM_WDAT 0x040030 +#define PCI_GPPM_RDAT 0x040034 +#define PCI_GPPM_CTRL 0x040038 +#define GPPM_DONE (1<<10) +#define INIT_PCI_CYCLE (1<<9) +#define GPPM_CMD(X) (((X)&0xf)<<4) +#define GPPM_BYTEEN(X) ((X)&0xf) +#define PCI_UNLOCKREG 0x04003C +#define UNLOCK_SSID(X) (((X)&0xff)<<8) +#define UNLOCK_SETUP(X) (((X)&0xff)<<0) +#define UNLOCK_MAGIC 0xCA +#define PCI_DEV_VEND_ID 0x040040 +#define DEVICE_ID(X) (((X)>>16)&0xffff) +#define VENDOR_ID(X) (((X)&0xffff)) +#define PCI_CFG_CMDSTAT 0x040044 +#define PCI_CFG_STATUS(X) (((X)>>16)&0xffff) +#define PCI_CFG_COMMAND(X) ((X)&0xffff) +#define PCI_CLASS_REV 0x040048 +#define PCI_CLASSCODE(X) (((X)>>8)&0xffffff) +#define PCI_REVID(X) ((X)&0xff) +#define PCI_LAT_TMR 0x04004c +#define PCI_BASE10 0x040050 +#define PCI_BASE14 0x040054 +#define PCI_BASE18 0x040058 +#define PCI_SUBSYS_ID 0x04006c +#define PCI_CAP_PTR 0x040074 +#define PCI_CFG_MISC 0x04007c +#define PCI_PMC 0x040080 +#define PCI_PWR_STATE 0x040084 +#define PCI_IO 0x040088 +#define PCI_SLVTUNING 0x04008C +#define PCI_DMATUNING 0x040090 +#define PCI_DMAEADDR 0x040800 +#define PCI_DMAIADDR 0x040804 +#define PCI_DMALEN 0x040808 +#define PCI_DMACTRL 0x04080C +#define PCI_XIOCTRL 0x040810 +#define PCI_SEL0PROF 0x040814 +#define PCI_SEL1PROF 0x040818 +#define PCI_SEL2PROF 0x04081C +#define PCI_GPXIOADDR 0x040820 +#define PCI_NANDCTRLS 0x400830 +#define PCI_SEL3PROF 0x040834 +#define PCI_SEL4PROF 0x040838 +#define PCI_GPXIO_STAT 0x040FB0 +#define PCI_GPXIO_IMASK 0x040FB4 +#define PCI_GPXIO_ICLR 0x040FB8 +#define PCI_GPXIO_ISET 0x040FBC +#define PCI_GPPM_STATUS 0x040FC0 +#define GPPM_DONE (1<<10) +#define GPPM_ERR (1<<9) +#define GPPM_MPAR_ERR (1<<8) +#define GPPM_PAR_ERR (1<<7) +#define GPPM_R_MABORT (1<<2) +#define GPPM_R_TABORT (1<<1) +#define PCI_GPPM_IMASK 0x040FC4 +#define PCI_GPPM_ICLR 0x040FC8 +#define PCI_GPPM_ISET 0x040FCC +#define PCI_DMA_STATUS 0x040FD0 +#define PCI_DMA_IMASK 0x040FD4 +#define PCI_DMA_ICLR 0x040FD8 +#define PCI_DMA_ISET 0x040FDC +#define PCI_ISTATUS 0x040FE0 +#define PCI_IMASK 0x040FE4 +#define PCI_ICLR 0x040FE8 +#define PCI_ISET 0x040FEC +#define PCI_MOD_ID 0x040FFC + +/* + * PCI configuration cycle AD bus definition + */ +/* Type 0 */ +#define PCI_CFG_TYPE0_REG_SHF 0 +#define PCI_CFG_TYPE0_FUNC_SHF 8 + +/* Type 1 */ +#define PCI_CFG_TYPE1_REG_SHF 0 +#define PCI_CFG_TYPE1_FUNC_SHF 8 +#define PCI_CFG_TYPE1_DEV_SHF 11 +#define PCI_CFG_TYPE1_BUS_SHF 16 + +/* + * Ethernet device DP83816 definition + */ +#define DP83816_IRQ_ETHER 66 + +#endif diff --git a/include/asm-mips/mach-pnx8550/uart.h b/include/asm-mips/mach-pnx8550/uart.h new file mode 100644 index 000000000000..e32b9a23d70e --- /dev/null +++ b/include/asm-mips/mach-pnx8550/uart.h @@ -0,0 +1,16 @@ +#ifndef __IP3106_UART_H +#define __IP3106_UART_H + +#include + +/* early macros for kgdb use. fixme: clean this up */ + +#define UART_BASE 0xbbe4a000 /* PNX8550 */ + +#define PNX8550_UART_PORT0 (UART_BASE) +#define PNX8550_UART_PORT1 (UART_BASE + 0x1000) + +#define PNX8550_UART_INT(x) (PNX8550_INT_GIC_MIN+19+x) +#define IRQ_TO_UART(x) (x-PNX8550_INT_GIC_MIN-19) + +#endif diff --git a/include/asm-mips/mach-pnx8550/usb.h b/include/asm-mips/mach-pnx8550/usb.h new file mode 100644 index 000000000000..483b7fc65d41 --- /dev/null +++ b/include/asm-mips/mach-pnx8550/usb.h @@ -0,0 +1,32 @@ +/* + * + * BRIEF MODULE DESCRIPTION + * USB specific definitions + * + * Author: source@mvista.com + * + * This program is free software; you can distribute it and/or modify it + * under the terms of the GNU General Public License (Version 2) as + * published by the Free Software Foundation. + * + * This program is distributed in the hope it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License + * for more details. + * + * You should have received a copy of the GNU General Public License along + * with this program; if not, write to the Free Software Foundation, Inc., + * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. + */ + +#ifndef __PNX8550_USB_H +#define __PNX8550_USB_H + +/* + * USB Host controller + */ + +#define PNX8550_USB_OHCI_OP_BASE 0x1be48000 +#define PNX8550_USB_OHCI_OP_LEN 0x1000 + +#endif diff --git a/include/asm-mips/mipsregs.h b/include/asm-mips/mipsregs.h index a7a43ff8c5cf..08aa231b4dea 100644 --- a/include/asm-mips/mipsregs.h +++ b/include/asm-mips/mipsregs.h @@ -841,12 +841,24 @@ do { \ #define read_c0_count() __read_32bit_c0_register($9, 0) #define write_c0_count(val) __write_32bit_c0_register($9, 0, val) +#define read_c0_count2() __read_32bit_c0_register($9, 6) /* pnx8550 */ +#define write_c0_count2(val) __write_32bit_c0_register($9, 6, val) + +#define read_c0_count3() __read_32bit_c0_register($9, 7) /* pnx8550 */ +#define write_c0_count3(val) __write_32bit_c0_register($9, 7, val) + #define read_c0_entryhi() __read_ulong_c0_register($10, 0) #define write_c0_entryhi(val) __write_ulong_c0_register($10, 0, val) #define read_c0_compare() __read_32bit_c0_register($11, 0) #define write_c0_compare(val) __write_32bit_c0_register($11, 0, val) +#define read_c0_compare2() __read_32bit_c0_register($11, 6) /* pnx8550 */ +#define write_c0_compare2(val) __write_32bit_c0_register($11, 6, val) + +#define read_c0_compare3() __read_32bit_c0_register($11, 7) /* pnx8550 */ +#define write_c0_compare3(val) __write_32bit_c0_register($11, 7, val) + #define read_c0_status() __read_32bit_c0_register($12, 0) #define write_c0_status(val) __write_32bit_c0_register($12, 0, val) diff --git a/include/linux/serial_core.h b/include/linux/serial_core.h index 27db8da43aa4..2b0401b93f2b 100644 --- a/include/linux/serial_core.h +++ b/include/linux/serial_core.h @@ -39,7 +39,8 @@ #define PORT_RSA 13 #define PORT_NS16550A 14 #define PORT_XSCALE 15 -#define PORT_MAX_8250 15 /* max port ID */ +#define PORT_IP3106 16 +#define PORT_MAX_8250 16 /* max port ID */ /* * ARM specific type numbers. These are not currently guaranteed diff --git a/include/linux/serial_ip3106.h b/include/linux/serial_ip3106.h new file mode 100644 index 000000000000..f500ac602c5c --- /dev/null +++ b/include/linux/serial_ip3106.h @@ -0,0 +1,81 @@ +/* + * Embedded Alley Solutions, source@embeddedalley.com. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + +#ifndef _LINUX_SERIAL_IP3106_H +#define _LINUX_SERIAL_IP3106_H + +#include +#include + +#define IP3106_NR_PORTS 2 + +struct ip3106_port { + struct uart_port port; + struct timer_list timer; + unsigned int old_status; +}; + +/* register offsets */ +#define IP3106_LCR 0 +#define IP3106_MCR 0x004 +#define IP3106_BAUD 0x008 +#define IP3106_CFG 0x00c +#define IP3106_FIFO 0x028 +#define IP3106_ISTAT 0xfe0 +#define IP3106_IEN 0xfe4 +#define IP3106_ICLR 0xfe8 +#define IP3106_ISET 0xfec +#define IP3106_PD 0xff4 +#define IP3106_MID 0xffc + +#define IP3106_UART_LCR_TXBREAK (1<<30) +#define IP3106_UART_LCR_PAREVN 0x10000000 +#define IP3106_UART_LCR_PAREN 0x08000000 +#define IP3106_UART_LCR_2STOPB 0x04000000 +#define IP3106_UART_LCR_8BIT 0x01000000 +#define IP3106_UART_LCR_TX_RST 0x00040000 +#define IP3106_UART_LCR_RX_RST 0x00020000 +#define IP3106_UART_LCR_RX_NEXT 0x00010000 + +#define IP3106_UART_MCR_SCR 0xFF000000 +#define IP3106_UART_MCR_DCD 0x00800000 +#define IP3106_UART_MCR_CTS 0x00100000 +#define IP3106_UART_MCR_LOOP 0x00000010 +#define IP3106_UART_MCR_RTS 0x00000002 +#define IP3106_UART_MCR_DTR 0x00000001 + +#define IP3106_UART_INT_TX 0x00000080 +#define IP3106_UART_INT_EMPTY 0x00000040 +#define IP3106_UART_INT_RCVTO 0x00000020 +#define IP3106_UART_INT_RX 0x00000010 +#define IP3106_UART_INT_RXOVRN 0x00000008 +#define IP3106_UART_INT_FRERR 0x00000004 +#define IP3106_UART_INT_BREAK 0x00000002 +#define IP3106_UART_INT_PARITY 0x00000001 +#define IP3106_UART_INT_ALLRX 0x0000003F +#define IP3106_UART_INT_ALLTX 0x000000C0 + +#define IP3106_UART_FIFO_TXFIFO 0x001F0000 +#define IP3106_UART_FIFO_TXFIFO_STA (0x1f<<16) +#define IP3106_UART_FIFO_RXBRK 0x00008000 +#define IP3106_UART_FIFO_RXFE 0x00004000 +#define IP3106_UART_FIFO_RXPAR 0x00002000 +#define IP3106_UART_FIFO_RXFIFO 0x00001F00 +#define IP3106_UART_FIFO_RBRTHR 0x000000FF + +#endif -- cgit v1.2.3 From 26a940e21752e0de8f068f77dad606a7d1986937 Mon Sep 17 00:00:00 2001 From: Pete Popov Date: Thu, 15 Sep 2005 08:03:12 +0000 Subject: Cleaned up AMD Au1200 IDE driver: - converted to platform bus - removed pci dependencies - removed virt_to_phys/phys_to_virt calls System now can root off of a disk. Signed-off-by: Ralf Baechle diff --git a/Documentation/mips/AU1xxx_IDE.README b/Documentation/mips/AU1xxx_IDE.README new file mode 100644 --- Documentation/mips/AU1xxx_IDE.README | 168 ++++ arch/mips/Kconfig | 2 +- arch/mips/au1000/common/dbdma.c | 6 + arch/mips/au1000/common/platform.c | 32 +- arch/mips/au1000/pb1200/irqmap.c | 6 +- drivers/ide/Kconfig | 31 +- drivers/ide/ide-proc.c | 1 + drivers/ide/mips/au1xxx-ide.c | 1250 +++++++++++++++++++++++++++ include/asm-mips/mach-au1x00/au1xxx.h | 44 + include/asm-mips/mach-au1x00/au1xxx_dbdma.h | 4 + include/asm-mips/mach-au1x00/au1xxx_ide.h | 301 +++++++ include/linux/ide.h | 2 +- 12 files changed, 1840 insertions(+), 7 deletions(-) create mode 100644 Documentation/mips/AU1xxx_IDE.README create mode 100644 drivers/ide/mips/au1xxx-ide.c create mode 100644 include/asm-mips/mach-au1x00/au1xxx.h create mode 100644 include/asm-mips/mach-au1x00/au1xxx_ide.h (limited to 'include/linux') diff --git a/Documentation/mips/AU1xxx_IDE.README b/Documentation/mips/AU1xxx_IDE.README new file mode 100644 index 000000000000..a7e4c4ea3560 --- /dev/null +++ b/Documentation/mips/AU1xxx_IDE.README @@ -0,0 +1,168 @@ +README for MIPS AU1XXX IDE driver - Released 2005-07-15 + +ABOUT +----- +This file describes the 'drivers/ide/mips/au1xxx-ide.c', related files and the +services they provide. + +If you are short in patience and just want to know how to add your hard disc to +the white or black list, go to the 'ADD NEW HARD DISC TO WHITE OR BLACK LIST' +section. + + +LICENSE +------- + +Copyright (c) 2003-2005 AMD, Personal Connectivity Solutions + +This program is free software; you can redistribute it and/or modify it under +the terms of the GNU General Public License as published by the Free Software +Foundation; either version 2 of the License, or (at your option) any later +version. + +THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, +INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND +FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR +BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR +CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF +SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS +INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN +CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) +ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE +POSSIBILITY OF SUCH DAMAGE. + +You should have received a copy of the GNU General Public License along with +this program; if not, write to the Free Software Foundation, Inc., +675 Mass Ave, Cambridge, MA 02139, USA. + +Note: for more information, please refer "AMD Alchemy Au1200/Au1550 IDE + Interface and Linux Device Driver" Application Note. + + +FILES, CONFIGS AND COMPATABILITY +-------------------------------- + +Two files are introduced: + + a) 'include/asm-mips/mach-au1x00/au1xxx_ide.h' + containes : struct _auide_hwif + struct drive_list_entry dma_white_list + struct drive_list_entry dma_black_list + timing parameters for PIO mode 0/1/2/3/4 + timing parameters for MWDMA 0/1/2 + + b) 'drivers/ide/mips/au1xxx-ide.c' + contains the functionality of the AU1XXX IDE driver + +Four configs variables are introduced: + + CONFIG_BLK_DEV_IDE_AU1XXX_PIO_DBDMA - enable the PIO+DBDMA mode + CONFIG_BLK_DEV_IDE_AU1XXX_MDMA2_DBDMA - enable the MWDMA mode + CONFIG_BLK_DEV_IDE_AU1XXX_BURSTABLE_ON - set Burstable FIFO in DBDMA + controler + CONFIG_BLK_DEV_IDE_AU1XXX_SEQTS_PER_RQ - maximum transfer size + per descriptor + +If MWDMA is enabled and the connected hard disc is not on the white list, the +kernel switches to a "safe mwdma mode" at boot time. In this mode the IDE +performance is substantial slower then in full speed mwdma. In this case +please add your hard disc to the white list (follow instruction from 'ADD NEW +HARD DISC TO WHITE OR BLACK LIST' section). + + +SUPPORTED IDE MODES +------------------- + +The AU1XXX IDE driver supported all PIO modes - PIO mode 0/1/2/3/4 - and all +MWDMA modes - MWDMA 0/1/2 -. There is no support for SWDMA and UDMA mode. + +To change the PIO mode use the program hdparm with option -p, e.g. +'hdparm -p0 [device]' for PIO mode 0. To enable the MWDMA mode use the option +-X, e.g. 'hdparm -X32 [device]' for MWDMA mode 0. + + +PERFORMANCE CONFIGURATIONS +-------------------------- + +If the used system doesn't need USB support enable the following kernel configs: + +CONFIG_IDE=y +CONFIG_BLK_DEV_IDE=y +CONFIG_IDE_GENERIC=y +CONFIG_BLK_DEV_IDEPCI=y +CONFIG_BLK_DEV_GENERIC=y +CONFIG_BLK_DEV_IDEDMA_PCI=y +CONFIG_IDEDMA_PCI_AUTO=y +CONFIG_BLK_DEV_IDE_AU1XXX=y +CONFIG_BLK_DEV_IDE_AU1XXX_MDMA2_DBDMA=y +CONFIG_BLK_DEV_IDE_AU1XXX_BURSTABLE_ON=y +CONFIG_BLK_DEV_IDE_AU1XXX_SEQTS_PER_RQ=128 +CONFIG_BLK_DEV_IDEDMA=y +CONFIG_IDEDMA_AUTO=y + +If the used system need the USB support enable the following kernel configs for +high IDE to USB throughput. + +CONFIG_BLK_DEV_IDEDISK=y +CONFIG_IDE_GENERIC=y +CONFIG_BLK_DEV_IDEPCI=y +CONFIG_BLK_DEV_GENERIC=y +CONFIG_BLK_DEV_IDEDMA_PCI=y +CONFIG_IDEDMA_PCI_AUTO=y +CONFIG_BLK_DEV_IDE_AU1XXX=y +CONFIG_BLK_DEV_IDE_AU1XXX_MDMA2_DBDMA=y +CONFIG_BLK_DEV_IDE_AU1XXX_SEQTS_PER_RQ=128 +CONFIG_BLK_DEV_IDEDMA=y +CONFIG_IDEDMA_AUTO=y + + +ADD NEW HARD DISC TO WHITE OR BLACK LIST +---------------------------------------- + +Step 1 : detect the model name of your hard disc + + a) connect your hard disc to the AU1XXX + + b) boot your kernel and get the hard disc model. + + Example boot log: + + --snipped-- + Uniform Multi-Platform E-IDE driver Revision: 7.00alpha2 + ide: Assuming 50MHz system bus speed for PIO modes; override with idebus=xx + Au1xxx IDE(builtin) configured for MWDMA2 + Probing IDE interface ide0... + hda: Maxtor 6E040L0, ATA DISK drive + ide0 at 0xac800000-0xac800007,0xac8001c0 on irq 64 + hda: max request size: 64KiB + hda: 80293248 sectors (41110 MB) w/2048KiB Cache, CHS=65535/16/63, (U)DMA + --snipped-- + + In this example 'Maxtor 6E040L0'. + +Step 2 : edit 'include/asm-mips/mach-au1x00/au1xxx_ide.h' + + Add your hard disc to the dma_white_list or dma_black_list structur. + +Step 3 : Recompile the kernel + + Enable MWDMA support in the kernel configuration. Recompile the kernel and + reboot. + +Step 4 : Tests + + If you have add a hard disc to the white list, please run some stress tests + for verification. + + +ACKNOWLEDGMENTS +--------------- + +These drivers wouldn't have been done without the base of kernel 2.4.x AU1XXX +IDE driver from AMD. + +Additional input also from: +Matthias Lenk + +Happy hacking! +Enrico Walther diff --git a/arch/mips/Kconfig b/arch/mips/Kconfig index 8746e1b1621b..dc5c629af106 100644 --- a/arch/mips/Kconfig +++ b/arch/mips/Kconfig @@ -799,7 +799,7 @@ config MIPS_BOSPORUS config MIPS_DB1200 bool "AMD Alchemy DB1200 board" select SOC_AU1200 - select DMA_NONCOHERENT + select DMA_COHERENT select MIPS_DISABLE_OBSOLETE_IDE select SYS_SUPPORTS_LITTLE_ENDIAN diff --git a/arch/mips/au1000/common/dbdma.c b/arch/mips/au1000/common/dbdma.c index 8f78c2fe7cf5..eb5803aed91b 100644 --- a/arch/mips/au1000/common/dbdma.c +++ b/arch/mips/au1000/common/dbdma.c @@ -197,6 +197,12 @@ find_dbdev_id (u32 id) return NULL; } +void * au1xxx_ddma_get_nextptr_virt(au1x_ddma_desc_t *dp) +{ + return phys_to_virt(DSCR_GET_NXTPTR(dp->dscr_nxtptr)); +} +EXPORT_SYMBOL(au1xxx_ddma_get_nextptr_virt); + u32 au1xxx_ddma_add_device(dbdev_tab_t *dev) { diff --git a/arch/mips/au1000/common/platform.c b/arch/mips/au1000/common/platform.c index 0c3fd726c3d6..4aca18f0e290 100644 --- a/arch/mips/au1000/common/platform.c +++ b/arch/mips/au1000/common/platform.c @@ -12,7 +12,7 @@ #include #include -#include +#include /* OHCI (USB full speed host controller) */ static struct resource au1xxx_usb_ohci_resources[] = { @@ -154,7 +154,6 @@ static struct platform_device au1xxx_usb_otg_device = { .resource = au1xxx_usb_otg_resources, }; -/*** AU1200 LCD controller ***/ static struct resource au1200_lcd_resources[] = { [0] = { .start = LCD_PHYS_ADDR, @@ -168,6 +167,19 @@ static struct resource au1200_lcd_resources[] = { } }; +static struct resource au1200_ide0_resources[] = { + [0] = { + .start = AU1XXX_ATA_PHYS_ADDR, + .end = AU1XXX_ATA_PHYS_ADDR + AU1XXX_ATA_PHYS_LEN, + .flags = IORESOURCE_MEM, + }, + [1] = { + .start = AU1XXX_ATA_INT, + .end = AU1XXX_ATA_INT, + .flags = IORESOURCE_IRQ, + } +}; + static u64 au1200_lcd_dmamask = ~(u32)0; static struct platform_device au1200_lcd_device = { @@ -180,6 +192,21 @@ static struct platform_device au1200_lcd_device = { .num_resources = ARRAY_SIZE(au1200_lcd_resources), .resource = au1200_lcd_resources, }; + + +static u64 ide0_dmamask = ~(u32)0; + +static struct platform_device au1200_ide0_device = { + .name = "au1200-ide", + .id = 0, + .dev = { + .dma_mask = &ide0_dmamask, + .coherent_dma_mask = 0xffffffff, + }, + .num_resources = ARRAY_SIZE(au1200_ide0_resources), + .resource = au1200_ide0_resources, +}; + #endif static struct platform_device *au1xxx_platform_devices[] __initdata = { @@ -194,6 +221,7 @@ static struct platform_device *au1xxx_platform_devices[] __initdata = { &au1xxx_usb_gdt_device, &au1xxx_usb_otg_device, &au1200_lcd_device, + &au1200_ide0_device, #endif }; diff --git a/arch/mips/au1000/pb1200/irqmap.c b/arch/mips/au1000/pb1200/irqmap.c index 2ec64e78aa01..59e70e5cf325 100644 --- a/arch/mips/au1000/pb1200/irqmap.c +++ b/arch/mips/au1000/pb1200/irqmap.c @@ -22,6 +22,7 @@ * with this program; if not, write to the Free Software Foundation, Inc., * 675 Mass Ave, Cambridge, MA 02139, USA. */ +#include #include #include #include @@ -65,7 +66,7 @@ int au1xxx_nr_irqs = sizeof(au1xxx_irq_map)/sizeof(au1xxx_irq_map_t); */ static volatile int pb1200_cascade_en=0; -void pb1200_cascade_handler( int irq, void *dev_id, struct pt_regs *regs) +irqreturn_t pb1200_cascade_handler( int irq, void *dev_id, struct pt_regs *regs) { unsigned short bisr = bcsr->int_status; int extirq_nr = 0; @@ -78,6 +79,7 @@ void pb1200_cascade_handler( int irq, void *dev_id, struct pt_regs *regs) /* Ack and dispatch IRQ */ do_IRQ(extirq_nr,regs); } + return IRQ_RETVAL(1); } inline void pb1200_enable_irq(unsigned int irq_nr) @@ -97,7 +99,7 @@ static unsigned int pb1200_startup_irq( unsigned int irq_nr ) if (++pb1200_cascade_en == 1) { request_irq(AU1000_GPIO_7, &pb1200_cascade_handler, - 0, "Pb1200 Cascade", &pb1200_cascade_handler ); + 0, "Pb1200 Cascade", (void *)&pb1200_cascade_handler ); #ifdef CONFIG_MIPS_PB1200 /* We have a problem with CPLD rev3. Enable a workaround */ if( ((bcsr->whoami & BCSR_WHOAMI_CPLD)>>4) <= 3) diff --git a/drivers/ide/Kconfig b/drivers/ide/Kconfig index 1cadd2c3cadd..a737886e39d1 100644 --- a/drivers/ide/Kconfig +++ b/drivers/ide/Kconfig @@ -778,6 +778,35 @@ config BLK_DEV_IDE_PMAC_BLINK This option enables the use of the sleep LED as a hard drive activity LED. +config BLK_DEV_IDE_AU1XXX + bool "IDE for AMD Alchemy Au1200" + depends on SOC_AU1200 +choice + prompt "IDE Mode for AMD Alchemy Au1200" + default CONFIG_BLK_DEV_IDE_AU1XXX_PIO_DBDMA + depends on SOC_AU1200 && BLK_DEV_IDE_AU1XXX + +config BLK_DEV_IDE_AU1XXX_PIO_DBDMA + bool "PIO+DbDMA IDE for AMD Alchemy Au1200" + +config BLK_DEV_IDE_AU1XXX_MDMA2_DBDMA + bool "MDMA2+DbDMA IDE for AMD Alchemy Au1200" + depends on SOC_AU1200 && BLK_DEV_IDE_AU1XXX +endchoice + +config BLK_DEV_IDE_AU1XXX_BURSTABLE_ON + bool "Enable burstable Mode on DbDMA" + default false + depends BLK_DEV_IDE_AU1XXX + help + This option enable the burstable Flag on DbDMA controller + (cf. "AMD Alchemy 'Au1200' Processor Data Book - PRELIMINARY"). + +config BLK_DEV_IDE_AU1XXX_SEQTS_PER_RQ + int "Maximum transfer size (KB) per request (up to 128)" + default "128" + depends BLK_DEV_IDE_AU1XXX + config IDE_ARM def_bool ARM && (ARCH_A5K || ARCH_CLPS7500 || ARCH_RPC || ARCH_SHARK) @@ -1013,7 +1042,7 @@ config BLK_DEV_UMC8672 endif config BLK_DEV_IDEDMA - def_bool BLK_DEV_IDEDMA_PCI || BLK_DEV_IDEDMA_PMAC || BLK_DEV_IDEDMA_ICS + def_bool BLK_DEV_IDEDMA_PCI || BLK_DEV_IDEDMA_PMAC || BLK_DEV_IDEDMA_ICS || BLK_DEV_IDE_AU1XXX_MDMA2_DBDMA config IDEDMA_IVB bool "IGNORE word93 Validation BITS" diff --git a/drivers/ide/ide-proc.c b/drivers/ide/ide-proc.c index 4063d2c34e3d..84665e2ba3c8 100644 --- a/drivers/ide/ide-proc.c +++ b/drivers/ide/ide-proc.c @@ -64,6 +64,7 @@ static int proc_ide_read_imodel case ide_cy82c693: name = "cy82c693"; break; case ide_4drives: name = "4drives"; break; case ide_pmac: name = "mac-io"; break; + case ide_au1xxx: name = "au1xxx"; break; default: name = "(unknown)"; break; } len = sprintf(page, "%s\n", name); diff --git a/drivers/ide/mips/au1xxx-ide.c b/drivers/ide/mips/au1xxx-ide.c new file mode 100644 index 000000000000..2b6327c576b9 --- /dev/null +++ b/drivers/ide/mips/au1xxx-ide.c @@ -0,0 +1,1250 @@ +/* + * linux/drivers/ide/mips/au1xxx-ide.c version 01.30.00 Aug. 02 2005 + * + * BRIEF MODULE DESCRIPTION + * AMD Alchemy Au1xxx IDE interface routines over the Static Bus + * + * Copyright (c) 2003-2005 AMD, Personal Connectivity Solutions + * + * This program is free software; you can redistribute it and/or modify it under + * the terms of the GNU General Public License as published by the Free Software + * Foundation; either version 2 of the License, or (at your option) any later + * version. + * + * THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, + * INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND + * FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR + * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR + * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF + * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS + * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN + * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) + * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE + * POSSIBILITY OF SUCH DAMAGE. + * + * You should have received a copy of the GNU General Public License along with + * this program; if not, write to the Free Software Foundation, Inc., + * 675 Mass Ave, Cambridge, MA 02139, USA. + * + * Note: for more information, please refer "AMD Alchemy Au1200/Au1550 IDE + * Interface and Linux Device Driver" Application Note. + */ +#undef REALLY_SLOW_IO /* most systems can safely undef this */ + +#include /* for CONFIG_BLK_DEV_IDEPCI */ +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include + +#include +#include +#include + +#if CONFIG_PM +#include +#endif + +#include + +#define DRV_NAME "au1200-ide" +#define DRV_VERSION "1.0" +#define DRV_AUTHOR "AMD PCS / Pete Popov " +#define DRV_DESC "Au1200 IDE" + +static _auide_hwif auide_hwif; +static spinlock_t ide_tune_drive_spin_lock = SPIN_LOCK_UNLOCKED; +static spinlock_t ide_tune_chipset_spin_lock = SPIN_LOCK_UNLOCKED; +static int dbdma_init_done = 0; + +/* + * local I/O functions + */ +u8 auide_inb(unsigned long port) +{ + return (au_readb(port)); +} + +u16 auide_inw(unsigned long port) +{ + return (au_readw(port)); +} + +u32 auide_inl(unsigned long port) +{ + return (au_readl(port)); +} + +void auide_insw(unsigned long port, void *addr, u32 count) +{ +#if defined(CONFIG_BLK_DEV_IDE_AU1XXX_PIO_DBDMA) + + _auide_hwif *ahwif = &auide_hwif; + chan_tab_t *ctp; + au1x_ddma_desc_t *dp; + + if(!put_dest_flags(ahwif->rx_chan, (void*)addr, count << 1, + DDMA_FLAGS_NOIE)) { + printk(KERN_ERR "%s failed %d\n", __FUNCTION__, __LINE__); + return; + } + ctp = *((chan_tab_t **)ahwif->rx_chan); + dp = ctp->cur_ptr; + while (dp->dscr_cmd0 & DSCR_CMD0_V) + ; + ctp->cur_ptr = au1xxx_ddma_get_nextptr_virt(dp); +#else + while (count--) + { + *(u16 *)addr = au_readw(port); + addr +=2 ; + } +#endif +} + +void auide_insl(unsigned long port, void *addr, u32 count) +{ + while (count--) + { + *(u32 *)addr = au_readl(port); + /* NOTE: For IDE interfaces over PCMCIA, + * 32-bit access does not work + */ + addr += 4; + } +} + +void auide_outb(u8 addr, unsigned long port) +{ + return (au_writeb(addr, port)); +} + +void auide_outbsync(ide_drive_t *drive, u8 addr, unsigned long port) +{ + return (au_writeb(addr, port)); +} + +void auide_outw(u16 addr, unsigned long port) +{ + return (au_writew(addr, port)); +} + +void auide_outl(u32 addr, unsigned long port) +{ + return (au_writel(addr, port)); +} + +void auide_outsw(unsigned long port, void *addr, u32 count) +{ +#if defined(CONFIG_BLK_DEV_IDE_AU1XXX_PIO_DBDMA) + _auide_hwif *ahwif = &auide_hwif; + chan_tab_t *ctp; + au1x_ddma_desc_t *dp; + + if(!put_source_flags(ahwif->tx_chan, (void*)addr, + count << 1, DDMA_FLAGS_NOIE)) { + printk(KERN_ERR "%s failed %d\n", __FUNCTION__, __LINE__); + return; + } + ctp = *((chan_tab_t **)ahwif->tx_chan); + dp = ctp->cur_ptr; + while (dp->dscr_cmd0 & DSCR_CMD0_V) + ; + ctp->cur_ptr = au1xxx_ddma_get_nextptr_virt(dp); +#else + while (count--) + { + au_writew(*(u16 *)addr, port); + addr += 2; + } +#endif +} + +void auide_outsl(unsigned long port, void *addr, u32 count) +{ + while (count--) + { + au_writel(*(u32 *)addr, port); + /* NOTE: For IDE interfaces over PCMCIA, + * 32-bit access does not work + */ + addr += 4; + } +} + +static void auide_tune_drive(ide_drive_t *drive, byte pio) +{ + int mem_sttime; + int mem_stcfg; + unsigned long flags; + u8 speed; + + /* get the best pio mode for the drive */ + pio = ide_get_best_pio_mode(drive, pio, 4, NULL); + + printk("%s: setting Au1XXX IDE to PIO mode%d\n", + drive->name, pio); + + spin_lock_irqsave(&ide_tune_drive_spin_lock, flags); + + mem_sttime = 0; + mem_stcfg = au_readl(MEM_STCFG2); + + /* set pio mode! */ + switch(pio) { + case 0: + /* set timing parameters for RCS2# */ + mem_sttime = SBC_IDE_PIO0_TWCS + | SBC_IDE_PIO0_TCSH + | SBC_IDE_PIO0_TCSOFF + | SBC_IDE_PIO0_TWP + | SBC_IDE_PIO0_TCSW + | SBC_IDE_PIO0_TPM + | SBC_IDE_PIO0_TA; + /* set configuration for RCS2# */ + mem_stcfg |= TS_MASK; + mem_stcfg &= ~TCSOE_MASK; + mem_stcfg &= ~TOECS_MASK; + mem_stcfg |= SBC_IDE_PIO0_TCSOE | SBC_IDE_PIO0_TOECS; + + au_writel(mem_sttime,MEM_STTIME2); + au_writel(mem_stcfg,MEM_STCFG2); + break; + + case 1: + /* set timing parameters for RCS2# */ + mem_sttime = SBC_IDE_PIO1_TWCS + | SBC_IDE_PIO1_TCSH + | SBC_IDE_PIO1_TCSOFF + | SBC_IDE_PIO1_TWP + | SBC_IDE_PIO1_TCSW + | SBC_IDE_PIO1_TPM + | SBC_IDE_PIO1_TA; + /* set configuration for RCS2# */ + mem_stcfg |= TS_MASK; + mem_stcfg &= ~TCSOE_MASK; + mem_stcfg &= ~TOECS_MASK; + mem_stcfg |= SBC_IDE_PIO1_TCSOE | SBC_IDE_PIO1_TOECS; + break; + + case 2: + /* set timing parameters for RCS2# */ + mem_sttime = SBC_IDE_PIO2_TWCS + | SBC_IDE_PIO2_TCSH + | SBC_IDE_PIO2_TCSOFF + | SBC_IDE_PIO2_TWP + | SBC_IDE_PIO2_TCSW + | SBC_IDE_PIO2_TPM + | SBC_IDE_PIO2_TA; + /* set configuration for RCS2# */ + mem_stcfg &= ~TS_MASK; + mem_stcfg &= ~TCSOE_MASK; + mem_stcfg &= ~TOECS_MASK; + mem_stcfg |= SBC_IDE_PIO2_TCSOE | SBC_IDE_PIO2_TOECS; + break; + + case 3: + /* set timing parameters for RCS2# */ + mem_sttime = SBC_IDE_PIO3_TWCS + | SBC_IDE_PIO3_TCSH + | SBC_IDE_PIO3_TCSOFF + | SBC_IDE_PIO3_TWP + | SBC_IDE_PIO3_TCSW + | SBC_IDE_PIO3_TPM + | SBC_IDE_PIO3_TA; + /* set configuration for RCS2# */ + mem_stcfg |= TS_MASK; + mem_stcfg &= ~TS_MASK; + mem_stcfg &= ~TCSOE_MASK; + mem_stcfg &= ~TOECS_MASK; + mem_stcfg |= SBC_IDE_PIO3_TCSOE | SBC_IDE_PIO3_TOECS; + + break; + + case 4: + /* set timing parameters for RCS2# */ + mem_sttime = SBC_IDE_PIO4_TWCS + | SBC_IDE_PIO4_TCSH + | SBC_IDE_PIO4_TCSOFF + | SBC_IDE_PIO4_TWP + | SBC_IDE_PIO4_TCSW + | SBC_IDE_PIO4_TPM + | SBC_IDE_PIO4_TA; + /* set configuration for RCS2# */ + mem_stcfg &= ~TS_MASK; + mem_stcfg &= ~TCSOE_MASK; + mem_stcfg &= ~TOECS_MASK; + mem_stcfg |= SBC_IDE_PIO4_TCSOE | SBC_IDE_PIO4_TOECS; + break; + } + + au_writel(mem_sttime,MEM_STTIME2); + au_writel(mem_stcfg,MEM_STCFG2); + + spin_unlock_irqrestore(&ide_tune_drive_spin_lock, flags); + + speed = pio + XFER_PIO_0; + ide_config_drive_speed(drive, speed); +} + +static int auide_tune_chipset (ide_drive_t *drive, u8 speed) +{ + u8 mode = 0; + int mem_sttime; + int mem_stcfg; + unsigned long flags; +#ifdef CONFIG_BLK_DEV_IDE_AU1XXX_MDMA2_DBDMA + struct hd_driveid *id = drive->id; + + /* + * Now see what the current drive is capable of, + * selecting UDMA only if the mate said it was ok. + */ + if (id && (id->capability & 1) && drive->autodma && + !__ide_dma_bad_drive(drive)) { + if (!mode && (id->field_valid & 2) && (id->dma_mword & 7)) { + if (id->dma_mword & 4) + mode = XFER_MW_DMA_2; + else if (id->dma_mword & 2) + mode = XFER_MW_DMA_1; + else if (id->dma_mword & 1) + mode = XFER_MW_DMA_0; + } + } +#endif + + spin_lock_irqsave(&ide_tune_chipset_spin_lock, flags); + + mem_sttime = 0; + mem_stcfg = au_readl(MEM_STCFG2); + + switch(speed) { + case XFER_PIO_4: + case XFER_PIO_3: + case XFER_PIO_2: + case XFER_PIO_1: + case XFER_PIO_0: + auide_tune_drive(drive, (speed - XFER_PIO_0)); + break; +#ifdef CONFIG_BLK_DEV_IDE_AU1XXX_MDMA2_DBDMA + case XFER_MW_DMA_2: + /* set timing parameters for RCS2# */ + mem_sttime = SBC_IDE_MDMA2_TWCS + | SBC_IDE_MDMA2_TCSH + | SBC_IDE_MDMA2_TCSOFF + | SBC_IDE_MDMA2_TWP + | SBC_IDE_MDMA2_TCSW + | SBC_IDE_MDMA2_TPM + | SBC_IDE_MDMA2_TA; + /* set configuration for RCS2# */ + mem_stcfg &= ~TS_MASK; + mem_stcfg &= ~TCSOE_MASK; + mem_stcfg &= ~TOECS_MASK; + mem_stcfg |= SBC_IDE_MDMA2_TCSOE | SBC_IDE_MDMA2_TOECS; + + mode = XFER_MW_DMA_2; + break; + case XFER_MW_DMA_1: + /* set timing parameters for RCS2# */ + mem_sttime = SBC_IDE_MDMA1_TWCS + | SBC_IDE_MDMA1_TCSH + | SBC_IDE_MDMA1_TCSOFF + | SBC_IDE_MDMA1_TWP + | SBC_IDE_MDMA1_TCSW + | SBC_IDE_MDMA1_TPM + | SBC_IDE_MDMA1_TA; + /* set configuration for RCS2# */ + mem_stcfg &= ~TS_MASK; + mem_stcfg &= ~TCSOE_MASK; + mem_stcfg &= ~TOECS_MASK; + mem_stcfg |= SBC_IDE_MDMA1_TCSOE | SBC_IDE_MDMA1_TOECS; + + mode = XFER_MW_DMA_1; + break; + case XFER_MW_DMA_0: + /* set timing parameters for RCS2# */ + mem_sttime = SBC_IDE_MDMA0_TWCS + | SBC_IDE_MDMA0_TCSH + | SBC_IDE_MDMA0_TCSOFF + | SBC_IDE_MDMA0_TWP + | SBC_IDE_MDMA0_TCSW + | SBC_IDE_MDMA0_TPM + | SBC_IDE_MDMA0_TA; + /* set configuration for RCS2# */ + mem_stcfg |= TS_MASK; + mem_stcfg &= ~TCSOE_MASK; + mem_stcfg &= ~TOECS_MASK; + mem_stcfg |= SBC_IDE_MDMA0_TCSOE | SBC_IDE_MDMA0_TOECS; + + mode = XFER_MW_DMA_0; + break; +#endif + default: + return 1; + } + + /* + * Tell the drive to switch to the new mode; abort on failure. + */ + if (!mode || ide_config_drive_speed(drive, mode)) + { + return 1; /* failure */ + } + + + au_writel(mem_sttime,MEM_STTIME2); + au_writel(mem_stcfg,MEM_STCFG2); + + spin_unlock_irqrestore(&ide_tune_chipset_spin_lock, flags); + + return 0; +} + +/* + * Multi-Word DMA + DbDMA functions + */ +#ifdef CONFIG_BLK_DEV_IDE_AU1XXX_MDMA2_DBDMA + +static int in_drive_list(struct hd_driveid *id, + const struct drive_list_entry *drive_table) +{ + for ( ; drive_table->id_model ; drive_table++){ + if ((!strcmp(drive_table->id_model, id->model)) && + ((strstr(drive_table->id_firmware, id->fw_rev)) || + (!strcmp(drive_table->id_firmware, "ALL"))) + ) + return 1; + } + return 0; +} + +static int auide_build_sglist(ide_drive_t *drive, struct request *rq) +{ + ide_hwif_t *hwif = drive->hwif; + _auide_hwif *ahwif = (_auide_hwif*)hwif->hwif_data; + struct scatterlist *sg = hwif->sg_table; + + ide_map_sg(drive, rq); + + if (rq_data_dir(rq) == READ) + hwif->sg_dma_direction = DMA_FROM_DEVICE; + else + hwif->sg_dma_direction = DMA_TO_DEVICE; + + return dma_map_sg(ahwif->dev, sg, hwif->sg_nents, + hwif->sg_dma_direction); +} + +static int auide_build_dmatable(ide_drive_t *drive) +{ + int i, iswrite, count = 0; + ide_hwif_t *hwif = HWIF(drive); + + struct request *rq = HWGROUP(drive)->rq; + + _auide_hwif *ahwif = (_auide_hwif*)hwif->hwif_data; + struct scatterlist *sg; + + iswrite = (rq_data_dir(rq) == WRITE); + /* Save for interrupt context */ + ahwif->drive = drive; + + /* Build sglist */ + hwif->sg_nents = i = auide_build_sglist(drive, rq); + + if (!i) + return 0; + + /* fill the descriptors */ + sg = hwif->sg_table; + while (i && sg_dma_len(sg)) { + u32 cur_addr; + u32 cur_len; + + cur_addr = sg_dma_address(sg); + cur_len = sg_dma_len(sg); + + while (cur_len) { + u32 flags = DDMA_FLAGS_NOIE; + unsigned int tc = (cur_len < 0xfe00)? cur_len: 0xfe00; + + if (++count >= PRD_ENTRIES) { + printk(KERN_WARNING "%s: DMA table too small\n", + drive->name); + goto use_pio_instead; + } + + /* Lets enable intr for the last descriptor only */ + if (1==i) + flags = DDMA_FLAGS_IE; + else + flags = DDMA_FLAGS_NOIE; + + if (iswrite) { + if(!put_source_flags(ahwif->tx_chan, + (void*)(page_address(sg->page) + + sg->offset), + tc, flags)) { + printk(KERN_ERR "%s failed %d\n", + __FUNCTION__, __LINE__); + } + } else + { + if(!put_dest_flags(ahwif->rx_chan, + (void*)(page_address(sg->page) + + sg->offset), + tc, flags)) { + printk(KERN_ERR "%s failed %d\n", + __FUNCTION__, __LINE__); + } + } + + cur_addr += tc; + cur_len -= tc; + } + sg++; + i--; + } + + if (count) + return 1; + +use_pio_instead: + dma_unmap_sg(ahwif->dev, + hwif->sg_table, + hwif->sg_nents, + hwif->sg_dma_direction); + + return 0; /* revert to PIO for this request */ +} + +static int auide_dma_end(ide_drive_t *drive) +{ + ide_hwif_t *hwif = HWIF(drive); + _auide_hwif *ahwif = (_auide_hwif*)hwif->hwif_data; + + if (hwif->sg_nents) { + dma_unmap_sg(ahwif->dev, hwif->sg_table, hwif->sg_nents, + hwif->sg_dma_direction); + hwif->sg_nents = 0; + } + + return 0; +} + +static void auide_dma_start(ide_drive_t *drive ) +{ +// printk("%s\n", __FUNCTION__); +} + +ide_startstop_t auide_dma_intr(ide_drive_t *drive) +{ + //printk("%s\n", __FUNCTION__); + + u8 stat = 0, dma_stat = 0; + + dma_stat = HWIF(drive)->ide_dma_end(drive); + stat = HWIF(drive)->INB(IDE_STATUS_REG); /* get drive status */ + if (OK_STAT(stat,DRIVE_READY,drive->bad_wstat|DRQ_STAT)) { + if (!dma_stat) { + struct request *rq = HWGROUP(drive)->rq; + + ide_end_request(drive, 1, rq->nr_sectors); + return ide_stopped; + } + printk(KERN_ERR "%s: dma_intr: bad DMA status (dma_stat=%x)\n", + drive->name, dma_stat); + } + return ide_error(drive, "dma_intr", stat); +} + +static void auide_dma_exec_cmd(ide_drive_t *drive, u8 command) +{ + //printk("%s\n", __FUNCTION__); + + /* issue cmd to drive */ + ide_execute_command(drive, command, &auide_dma_intr, + (2*WAIT_CMD), NULL); +} + +static int auide_dma_setup(ide_drive_t *drive) +{ +// printk("%s\n", __FUNCTION__); + + if (drive->media != ide_disk) + return 1; + + if (!auide_build_dmatable(drive)) + /* try PIO instead of DMA */ + return 1; + + drive->waiting_for_dma = 1; + + return 0; +} + +static int auide_dma_check(ide_drive_t *drive) +{ +// printk("%s\n", __FUNCTION__); + +#ifdef CONFIG_BLK_DEV_IDE_AU1XXX_MDMA2_DBDMA + if( !dbdma_init_done ){ + auide_hwif.white_list = in_drive_list(drive->id, + dma_white_list); + auide_hwif.black_list = in_drive_list(drive->id, + dma_black_list); + auide_hwif.drive = drive; + auide_ddma_init(&auide_hwif); + dbdma_init_done = 1; + } +#endif + + /* Is the drive in our DMA black list? */ + if ( auide_hwif.black_list ) { + drive->using_dma = 0; + printk("%s found in dma_blacklist[]! Disabling DMA.\n", + drive->id->model); + } + else + drive->using_dma = 1; + + return HWIF(drive)->ide_dma_host_on(drive); +} + +static int auide_dma_test_irq(ide_drive_t *drive) +{ +// printk("%s\n", __FUNCTION__); + + if (!drive->waiting_for_dma) + printk(KERN_WARNING "%s: ide_dma_test_irq \ + called while not waiting\n", drive->name); + + /* If dbdma didn't execute the STOP command yet, the + * active bit is still set + */ + drive->waiting_for_dma++; + if (drive->waiting_for_dma >= DMA_WAIT_TIMEOUT) { + printk(KERN_WARNING "%s: timeout waiting for ddma to \ + complete\n", drive->name); + return 1; + } + udelay(10); + return 0; +} + +static int auide_dma_host_on(ide_drive_t *drive) +{ +// printk("%s\n", __FUNCTION__); + return 0; +} + +static int auide_dma_on(ide_drive_t *drive) +{ +// printk("%s\n", __FUNCTION__); + drive->using_dma = 1; + return auide_dma_host_on(drive); +} + + +static int auide_dma_host_off(ide_drive_t *drive) +{ +// printk("%s\n", __FUNCTION__); + return 0; +} + +static int auide_dma_off_quietly(ide_drive_t *drive) +{ +// printk("%s\n", __FUNCTION__); + drive->using_dma = 0; + return auide_dma_host_off(drive); +} + +static int auide_dma_lostirq(ide_drive_t *drive) +{ +// printk("%s\n", __FUNCTION__); + + printk(KERN_ERR "%s: IRQ lost\n", drive->name); + return 0; +} + +static void auide_ddma_tx_callback(int irq, void *param, struct pt_regs *regs) +{ +// printk("%s\n", __FUNCTION__); + + _auide_hwif *ahwif = (_auide_hwif*)param; + ahwif->drive->waiting_for_dma = 0; + return; +} + +static void auide_ddma_rx_callback(int irq, void *param, struct pt_regs *regs) +{ +// printk("%s\n", __FUNCTION__); + + _auide_hwif *ahwif = (_auide_hwif*)param; + ahwif->drive->waiting_for_dma = 0; + return; +} + +static int auide_dma_timeout(ide_drive_t *drive) +{ +// printk("%s\n", __FUNCTION__); + + printk(KERN_ERR "%s: DMA timeout occurred: ", drive->name); + + if (HWIF(drive)->ide_dma_test_irq(drive)) + return 0; + + return HWIF(drive)->ide_dma_end(drive); +} +#endif /* end CONFIG_BLK_DEV_IDE_AU1XXX_MDMA2_DBDMA */ + + +static int auide_ddma_init( _auide_hwif *auide ) +{ +// printk("%s\n", __FUNCTION__); + + dbdev_tab_t source_dev_tab; +#if defined(CONFIG_BLK_DEV_IDE_AU1XXX_MDMA2_DBDMA) + dbdev_tab_t target_dev_tab; + ide_hwif_t *hwif = auide->hwif; + char warning_output [2][80]; + int i; +#endif + + /* Add our custom device to DDMA device table */ + /* Create our new device entries in the table */ +#if defined(CONFIG_BLK_DEV_IDE_AU1XXX_MDMA2_DBDMA) + source_dev_tab.dev_id = AU1XXX_ATA_DDMA_REQ; + + if( auide->white_list || auide->black_list ){ + source_dev_tab.dev_tsize = 8; + source_dev_tab.dev_devwidth = 32; + source_dev_tab.dev_physaddr = (u32)AU1XXX_ATA_PHYS_ADDR; + source_dev_tab.dev_intlevel = 0; + source_dev_tab.dev_intpolarity = 0; + + /* init device table for target - static bus controller - */ + target_dev_tab.dev_id = DSCR_CMD0_ALWAYS; + target_dev_tab.dev_tsize = 8; + target_dev_tab.dev_devwidth = 32; + target_dev_tab.dev_physaddr = (u32)AU1XXX_ATA_PHYS_ADDR; + target_dev_tab.dev_intlevel = 0; + target_dev_tab.dev_intpolarity = 0; + target_dev_tab.dev_flags = DEV_FLAGS_ANYUSE; + } + else{ + source_dev_tab.dev_tsize = 1; + source_dev_tab.dev_devwidth = 16; + source_dev_tab.dev_physaddr = (u32)AU1XXX_ATA_PHYS_ADDR; + source_dev_tab.dev_intlevel = 0; + source_dev_tab.dev_intpolarity = 0; + + /* init device table for target - static bus controller - */ + target_dev_tab.dev_id = DSCR_CMD0_ALWAYS; + target_dev_tab.dev_tsize = 1; + target_dev_tab.dev_devwidth = 16; + target_dev_tab.dev_physaddr = (u32)AU1XXX_ATA_PHYS_ADDR; + target_dev_tab.dev_intlevel = 0; + target_dev_tab.dev_intpolarity = 0; + target_dev_tab.dev_flags = DEV_FLAGS_ANYUSE; + + sprintf(&warning_output[0][0], + "%s is not on ide driver white list.", + auide_hwif.drive->id->model); + for ( i=strlen(&warning_output[0][0]) ; i<76; i++ ){ + sprintf(&warning_output[0][i]," "); + } + + sprintf(&warning_output[1][0], + "To add %s please read 'Documentation/mips/AU1xxx_IDE.README'.", + auide_hwif.drive->id->model); + for ( i=strlen(&warning_output[1][0]) ; i<76; i++ ){ + sprintf(&warning_output[1][i]," "); + } + + printk("\n****************************************"); + printk("****************************************\n"); + printk("* %s *\n",&warning_output[0][0]); + printk("* Switch to safe MWDMA Mode! "); + printk(" *\n"); + printk("* %s *\n",&warning_output[1][0]); + printk("****************************************"); + printk("****************************************\n\n"); + } +#else + source_dev_tab.dev_id = DSCR_CMD0_ALWAYS; + source_dev_tab.dev_tsize = 8; + source_dev_tab.dev_devwidth = 32; + source_dev_tab.dev_physaddr = (u32)AU1XXX_ATA_PHYS_ADDR; + source_dev_tab.dev_intlevel = 0; + source_dev_tab.dev_intpolarity = 0; +#endif + +#if CONFIG_BLK_DEV_IDE_AU1XXX_BURSTABLE_ON + /* set flags for tx channel */ + source_dev_tab.dev_flags = DEV_FLAGS_OUT + | DEV_FLAGS_SYNC + | DEV_FLAGS_BURSTABLE; + auide->tx_dev_id = au1xxx_ddma_add_device( &source_dev_tab ); + /* set flags for rx channel */ + source_dev_tab.dev_flags = DEV_FLAGS_IN + | DEV_FLAGS_SYNC + | DEV_FLAGS_BURSTABLE; + auide->rx_dev_id = au1xxx_ddma_add_device( &source_dev_tab ); +#else + /* set flags for tx channel */ + source_dev_tab.dev_flags = DEV_FLAGS_OUT | DEV_FLAGS_SYNC; + auide->tx_dev_id = au1xxx_ddma_add_device( &source_dev_tab ); + /* set flags for rx channel */ + source_dev_tab.dev_flags = DEV_FLAGS_IN | DEV_FLAGS_SYNC; + auide->rx_dev_id = au1xxx_ddma_add_device( &source_dev_tab ); +#endif + +#if defined(CONFIG_BLK_DEV_IDE_AU1XXX_MDMA2_DBDMA) + + auide->target_dev_id = au1xxx_ddma_add_device(&target_dev_tab); + + /* Get a channel for TX */ + auide->tx_chan = au1xxx_dbdma_chan_alloc(auide->target_dev_id, + auide->tx_dev_id, + auide_ddma_tx_callback, + (void*)auide); + /* Get a channel for RX */ + auide->rx_chan = au1xxx_dbdma_chan_alloc(auide->rx_dev_id, + auide->target_dev_id, + auide_ddma_rx_callback, + (void*)auide); +#else /* CONFIG_BLK_DEV_IDE_AU1XXX_PIO_DBDMA */ + /* + * Note: if call back is not enabled, update ctp->cur_ptr manually + */ + auide->tx_chan = au1xxx_dbdma_chan_alloc(DSCR_CMD0_ALWAYS, + auide->tx_dev_id, + NULL, + (void*)auide); + auide->rx_chan = au1xxx_dbdma_chan_alloc(auide->rx_dev_id, + DSCR_CMD0_ALWAYS, + NULL, + (void*)auide); +#endif + auide->tx_desc_head = (void*)au1xxx_dbdma_ring_alloc(auide->tx_chan, + NUM_DESCRIPTORS); + auide->rx_desc_head = (void*)au1xxx_dbdma_ring_alloc(auide->rx_chan, + NUM_DESCRIPTORS); + +#if defined(CONFIG_BLK_DEV_IDE_AU1XXX_MDMA2_DBDMA) + hwif->dmatable_cpu = dma_alloc_coherent(auide->dev, + PRD_ENTRIES * PRD_BYTES, /* 1 Page */ + &hwif->dmatable_dma, GFP_KERNEL); + + auide->sg_table = kmalloc(sizeof(struct scatterlist) * PRD_ENTRIES, + GFP_KERNEL|GFP_DMA); + if (auide->sg_table == NULL) { + return -ENOMEM; + } +#endif + au1xxx_dbdma_start( auide->tx_chan ); + au1xxx_dbdma_start( auide->rx_chan ); + return 0; +} + +static void auide_setup_ports(hw_regs_t *hw, _auide_hwif *ahwif) +{ + int i; +#define ide_ioreg_t unsigned long + ide_ioreg_t *ata_regs = hw->io_ports; + + /* fixme */ + for (i = 0; i < IDE_CONTROL_OFFSET; i++) { + *ata_regs++ = (ide_ioreg_t) ahwif->regbase + + (ide_ioreg_t)(i << AU1XXX_ATA_REG_OFFSET); + } + + /* set the Alternative Status register */ + *ata_regs = (ide_ioreg_t) ahwif->regbase + + (ide_ioreg_t)(14 << AU1XXX_ATA_REG_OFFSET); +} + +static int au_ide_probe(struct device *dev) +{ + struct platform_device *pdev = to_platform_device(dev); + _auide_hwif *ahwif = &auide_hwif; + ide_hwif_t *hwif; + struct resource *res; + int ret = 0; + +#if defined(CONFIG_BLK_DEV_IDE_AU1XXX_MDMA2_DBDMA) + char *mode = "MWDMA2"; +#elif defined(CONFIG_BLK_DEV_IDE_AU1XXX_PIO_DBDMA) + char *mode = "PIO+DDMA(offload)"; +#endif + + memset(&auide_hwif, 0, sizeof(_auide_hwif)); + auide_hwif.dev = 0; + + ahwif->dev = dev; + ahwif->irq = platform_get_irq(pdev, 0); + + res = platform_get_resource(pdev, IORESOURCE_MEM, 0); + + if (res == NULL) { + pr_debug("%s %d: no base address\n", DRV_NAME, pdev->id); + ret = -ENODEV; + goto out; + } + + if (!request_mem_region (res->start, res->end-res->start, pdev->name)) { + pr_debug("%s: request_mem_region failed\n", DRV_NAME); + ret = -EBUSY; + goto out; + } + + ahwif->regbase = (u32)ioremap(res->start, res->end-res->start); + if (ahwif->regbase == 0) { + ret = -ENOMEM; + goto out; + } + + hwif = &ide_hwifs[pdev->id]; + hw_regs_t *hw = &hwif->hw; + hwif->irq = hw->irq = ahwif->irq; + hwif->chipset = ide_au1xxx; + + auide_setup_ports(hw, ahwif); + memcpy(hwif->io_ports, hw->io_ports, sizeof(hwif->io_ports)); + +#ifdef CONFIG_BLK_DEV_IDE_AU1XXX_SEQTS_PER_RQ + hwif->rqsize = CONFIG_BLK_DEV_IDE_AU1XXX_SEQTS_PER_RQ; + hwif->rqsize = ((hwif->rqsize > AU1XXX_ATA_RQSIZE) + || (hwif->rqsize < 32)) ? AU1XXX_ATA_RQSIZE : hwif->rqsize; +#else /* if kernel config is not set */ + hwif->rqsize = AU1XXX_ATA_RQSIZE; +#endif + + hwif->ultra_mask = 0x0; /* Disable Ultra DMA */ +#ifdef CONFIG_BLK_DEV_IDE_AU1XXX_MDMA2_DBDMA + hwif->mwdma_mask = 0x07; /* Multimode-2 DMA */ + hwif->swdma_mask = 0x07; +#else + hwif->mwdma_mask = 0x0; + hwif->swdma_mask = 0x0; +#endif + //hwif->noprobe = !hwif->io_ports[IDE_DATA_OFFSET]; + hwif->noprobe = 0; + hwif->drives[0].unmask = 1; + hwif->drives[1].unmask = 1; + + /* hold should be on in all cases */ + hwif->hold = 1; + hwif->mmio = 2; + + /* set up local I/O function entry points */ + hwif->INB = auide_inb; + hwif->INW = auide_inw; + hwif->INL = auide_inl; + hwif->INSW = auide_insw; + hwif->INSL = auide_insl; + hwif->OUTB = auide_outb; + hwif->OUTBSYNC = auide_outbsync; + hwif->OUTW = auide_outw; + hwif->OUTL = auide_outl; + hwif->OUTSW = auide_outsw; + hwif->OUTSL = auide_outsl; + + hwif->tuneproc = &auide_tune_drive; + hwif->speedproc = &auide_tune_chipset; + +#ifdef CONFIG_BLK_DEV_IDE_AU1XXX_MDMA2_DBDMA + hwif->ide_dma_off_quietly = &auide_dma_off_quietly; + hwif->ide_dma_timeout = &auide_dma_timeout; + + hwif->ide_dma_check = &auide_dma_check; + hwif->dma_exec_cmd = &auide_dma_exec_cmd; + hwif->dma_start = &auide_dma_start; + hwif->ide_dma_end = &auide_dma_end; + hwif->dma_setup = &auide_dma_setup; + hwif->ide_dma_test_irq = &auide_dma_test_irq; + hwif->ide_dma_host_off = &auide_dma_host_off; + hwif->ide_dma_host_on = &auide_dma_host_on; + hwif->ide_dma_lostirq = &auide_dma_lostirq; + hwif->ide_dma_on = &auide_dma_on; + + hwif->autodma = 1; + hwif->drives[0].autodma = hwif->autodma; + hwif->drives[1].autodma = hwif->autodma; + hwif->atapi_dma = 1; + hwif->drives[0].using_dma = 1; + hwif->drives[1].using_dma = 1; +#else /* !CONFIG_BLK_DEV_IDE_AU1XXX_MDMA2_DBDMA */ + hwif->autodma = 0; + hwif->channel = 0; + hwif->hold = 1; + hwif->select_data = 0; /* no chipset-specific code */ + hwif->config_data = 0; /* no chipset-specific code */ + + hwif->drives[0].autodma = 0; + hwif->drives[0].drive_data = 0; /* no drive data */ + hwif->drives[0].using_dma = 0; + hwif->drives[0].waiting_for_dma = 0; + hwif->drives[0].autotune = 1; /* 1=autotune, 2=noautotune, 0=default */ + /* secondary hdd not supported */ + hwif->drives[1].autodma = 0; + + hwif->drives[1].drive_data = 0; + hwif->drives[1].using_dma = 0; + hwif->drives[1].waiting_for_dma = 0; + hwif->drives[1].autotune = 2; /* 1=autotune, 2=noautotune, 0=default */ +#endif + hwif->drives[0].io_32bit = 0; /* 0=16-bit, 1=32-bit, 2/3=32bit+sync */ + hwif->drives[1].io_32bit = 0; /* 0=16-bit, 1=32-bit, 2/3=32bit+sync */ + + /*Register Driver with PM Framework*/ +#ifdef CONFIG_PM + auide_hwif.pm.lock = SPIN_LOCK_UNLOCKED; + auide_hwif.pm.stopped = 0; + + auide_hwif.pm.dev = new_au1xxx_power_device( "ide", + &au1200ide_pm_callback, + NULL); + if ( auide_hwif.pm.dev == NULL ) + printk(KERN_INFO "Unable to create a power management \ + device entry for the au1200-IDE.\n"); + else + printk(KERN_INFO "Power management device entry for the \ + au1200-IDE loaded.\n"); +#endif + + auide_hwif.hwif = hwif; + hwif->hwif_data = &auide_hwif; + +#ifdef CONFIG_BLK_DEV_IDE_AU1XXX_PIO_DBDMA + auide_ddma_init(&auide_hwif); + dbdma_init_done = 1; +#endif + + probe_hwif_init(hwif); + dev_set_drvdata(dev, hwif); + + printk(KERN_INFO "Au1xxx IDE(builtin) configured for %s\n", mode ); + +out: + return ret; +} + +static int au_ide_remove(struct device *dev) +{ + struct platform_device *pdev = to_platform_device(dev); + struct resource *res; + ide_hwif_t *hwif = dev_get_drvdata(dev); + _auide_hwif *ahwif = &auide_hwif; + + ide_unregister(hwif - ide_hwifs); + + iounmap((void *)ahwif->regbase); + + res = platform_get_resource(pdev, IORESOURCE_MEM, 0); + release_mem_region(res->start, res->end - res->start); + + return 0; +} + +static struct device_driver au1200_ide_driver = { + .name = "au1200-ide", + .bus = &platform_bus_type, + .probe = au_ide_probe, + .remove = au_ide_remove, +}; + +static int __init au_ide_init(void) +{ + return driver_register(&au1200_ide_driver); +} + +static void __init au_ide_exit(void) +{ + driver_unregister(&au1200_ide_driver); +} + +#ifdef CONFIG_PM +int au1200ide_pm_callback( au1xxx_power_dev_t *dev,\ + au1xxx_request_t request, void *data) { + + unsigned int d, err = 0; + unsigned long flags; + + spin_lock_irqsave(auide_hwif.pm.lock, flags); + + switch (request){ + case AU1XXX_PM_SLEEP: + err = au1xxxide_pm_sleep(dev); + break; + case AU1XXX_PM_WAKEUP: + d = *((unsigned int*)data); + if ( d > 0 && d <= 99) { + err = au1xxxide_pm_standby(dev); + } + else { + err = au1xxxide_pm_resume(dev); + } + break; + case AU1XXX_PM_GETSTATUS: + err = au1xxxide_pm_getstatus(dev); + break; + case AU1XXX_PM_ACCESS: + err = au1xxxide_pm_access(dev); + break; + case AU1XXX_PM_IDLE: + err = au1xxxide_pm_idle(dev); + break; + case AU1XXX_PM_CLEANUP: + err = au1xxxide_pm_cleanup(dev); + break; + default: + err = -1; + break; + } + + spin_unlock_irqrestore(auide_hwif.pm.lock, flags); + + return err; +} + +static int au1xxxide_pm_standby( au1xxx_power_dev_t *dev ) { + return 0; +} + +static int au1xxxide_pm_sleep( au1xxx_power_dev_t *dev ) { + + int retval; + ide_hwif_t *hwif = auide_hwif.hwif; + struct request rq; + struct request_pm_state rqpm; + ide_task_t args; + + if(auide_hwif.pm.stopped) + return -1; + + /* + * wait until hard disc is ready + */ + if ( wait_for_ready(&hwif->drives[0], 35000) ) { + printk("Wait for drive sleep timeout!\n"); + retval = -1; + } + + /* + * sequenz to tell the high level ide driver that pm is resuming + */ + memset(&rq, 0, sizeof(rq)); + memset(&rqpm, 0, sizeof(rqpm)); + memset(&args, 0, sizeof(args)); + rq.flags = REQ_PM_SUSPEND; + rq.special = &args; + rq.pm = &rqpm; + rqpm.pm_step = ide_pm_state_start_suspend; + rqpm.pm_state = PMSG_SUSPEND; + + retval = ide_do_drive_cmd(&hwif->drives[0], &rq, ide_wait); + + if (wait_for_ready (&hwif->drives[0], 35000)) { + printk("Wait for drive sleep timeout!\n"); + retval = -1; + } + + /* + * stop dbdma channels + */ + au1xxx_dbdma_reset(auide_hwif.tx_chan); + au1xxx_dbdma_reset(auide_hwif.rx_chan); + + auide_hwif.pm.stopped = 1; + + return retval; +} + +static int au1xxxide_pm_resume( au1xxx_power_dev_t *dev ) { + + int retval; + ide_hwif_t *hwif = auide_hwif.hwif; + struct request rq; + struct request_pm_state rqpm; + ide_task_t args; + + if(!auide_hwif.pm.stopped) + return -1; + + /* + * start dbdma channels + */ + au1xxx_dbdma_start(auide_hwif.tx_chan); + au1xxx_dbdma_start(auide_hwif.rx_chan); + + /* + * wait until hard disc is ready + */ + if (wait_for_ready ( &hwif->drives[0], 35000)) { + printk("Wait for drive wake up timeout!\n"); + retval = -1; + } + + /* + * sequenz to tell the high level ide driver that pm is resuming + */ + memset(&rq, 0, sizeof(rq)); + memset(&rqpm, 0, sizeof(rqpm)); + memset(&args, 0, sizeof(args)); + rq.flags = REQ_PM_RESUME; + rq.special = &args; + rq.pm = &rqpm; + rqpm.pm_step = ide_pm_state_start_resume; + rqpm.pm_state = PMSG_ON; + + retval = ide_do_drive_cmd(&hwif->drives[0], &rq, ide_head_wait); + + /* + * wait for hard disc + */ + if ( wait_for_ready(&hwif->drives[0], 35000) ) { + printk("Wait for drive wake up timeout!\n"); + retval = -1; + } + + auide_hwif.pm.stopped = 0; + + return retval; +} + +static int au1xxxide_pm_getstatus( au1xxx_power_dev_t *dev ) { + return dev->cur_state; +} + +static int au1xxxide_pm_access( au1xxx_power_dev_t *dev ) { + if (dev->cur_state != AWAKE_STATE) + return 0; + else + return -1; +} + +static int au1xxxide_pm_idle( au1xxx_power_dev_t *dev ) { + return 0; +} + +static int au1xxxide_pm_cleanup( au1xxx_power_dev_t *dev ) { + return 0; +} +#endif /* CONFIG_PM */ + +MODULE_LICENSE("GPL"); +MODULE_DESCRIPTION("AU1200 IDE driver"); + +module_init(au_ide_init); +module_exit(au_ide_exit); diff --git a/include/asm-mips/mach-au1x00/au1xxx.h b/include/asm-mips/mach-au1x00/au1xxx.h new file mode 100644 index 000000000000..b7b46dd9b929 --- /dev/null +++ b/include/asm-mips/mach-au1x00/au1xxx.h @@ -0,0 +1,44 @@ +/* + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License as published by the + * Free Software Foundation; either version 2 of the License, or (at your + * option) any later version. + * + * THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESS OR IMPLIED + * WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF + * MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN + * NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT, + * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT + * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF + * USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON + * ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF + * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + * You should have received a copy of the GNU General Public License along + * with this program; if not, write to the Free Software Foundation, Inc., + * 675 Mass Ave, Cambridge, MA 02139, USA. + */ + +#ifndef _AU1XXX_H_ +#define _AU1XXX_H_ + +#include + +#include + +#if defined(CONFIG_MIPS_DB1000) || defined(CONFIG_MIPS_DB1100) || defined(CONFIG_MIPS_DB1500) || defined(CONFIG_MIPS_DB1550) +#include + +#elif defined(CONFIG_MIPS_PB1550) +#include + +#elif defined(CONFIG_MIPS_PB1200) +#include + +#elif defined(CONFIG_MIPS_DB1200) +#include + +#endif + +#endif /* _AU1XXX_H_ */ diff --git a/include/asm-mips/mach-au1x00/au1xxx_dbdma.h b/include/asm-mips/mach-au1x00/au1xxx_dbdma.h index ddbd9f5a2489..b327bcd3fee1 100644 --- a/include/asm-mips/mach-au1x00/au1xxx_dbdma.h +++ b/include/asm-mips/mach-au1x00/au1xxx_dbdma.h @@ -368,6 +368,7 @@ void au1xxx_dbdma_dump(u32 chanid); u32 au1xxx_dbdma_put_dscr(u32 chanid, au1x_ddma_desc_t *dscr ); u32 au1xxx_ddma_add_device( dbdev_tab_t *dev ); +void * au1xxx_ddma_get_nextptr_virt(au1x_ddma_desc_t *dp); /* Some compatibilty macros -- @@ -375,9 +376,12 @@ u32 au1xxx_ddma_add_device( dbdev_tab_t *dev ); */ #define au1xxx_dbdma_put_source(chanid,buf,nbytes)_au1xxx_dbdma_put_source(chanid, buf, nbytes, DDMA_FLAGS_IE) #define au1xxx_dbdma_put_source_flags(chanid,buf,nbytes,flags) _au1xxx_dbdma_put_source(chanid, buf, nbytes, flags) +#define put_source_flags(chanid,buf,nbytes,flags) au1xxx_dbdma_put_source_flags(chanid,buf,nbytes,flags) + #define au1xxx_dbdma_put_dest(chanid,buf,nbytes) _au1xxx_dbdma_put_dest(chanid, buf, nbytes, DDMA_FLAGS_IE) #define au1xxx_dbdma_put_dest_flags(chanid,buf,nbytes,flags) _au1xxx_dbdma_put_dest(chanid, buf, nbytes, flags) +#define put_dest_flags(chanid,buf,nbytes,flags) au1xxx_dbdma_put_dest_flags(chanid,buf,nbytes,flags) /* * Flags for the put_source/put_dest functions. diff --git a/include/asm-mips/mach-au1x00/au1xxx_ide.h b/include/asm-mips/mach-au1x00/au1xxx_ide.h new file mode 100644 index 000000000000..33d275c3b84c --- /dev/null +++ b/include/asm-mips/mach-au1x00/au1xxx_ide.h @@ -0,0 +1,301 @@ +/* + * include/asm-mips/mach-au1x00/au1xxx_ide.h version 01.30.00 Aug. 02 2005 + * + * BRIEF MODULE DESCRIPTION + * AMD Alchemy Au1xxx IDE interface routines over the Static Bus + * + * Copyright (c) 2003-2005 AMD, Personal Connectivity Solutions + * + * This program is free software; you can redistribute it and/or modify it under + * the terms of the GNU General Public License as published by the Free Software + * Foundation; either version 2 of the License, or (at your option) any later + * version. + * + * THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, + * INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND + * FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR + * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR + * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF + * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS + * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN + * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) + * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE + * POSSIBILITY OF SUCH DAMAGE. + * + * You should have received a copy of the GNU General Public License along with + * this program; if not, write to the Free Software Foundation, Inc., + * 675 Mass Ave, Cambridge, MA 02139, USA. + * + * Note: for more information, please refer "AMD Alchemy Au1200/Au1550 IDE + * Interface and Linux Device Driver" Application Note. + */ +#include + +#ifdef CONFIG_BLK_DEV_IDE_AU1XXX_MDMA2_DBDMA + #define DMA_WAIT_TIMEOUT 100 + #define NUM_DESCRIPTORS PRD_ENTRIES +#else /* CONFIG_BLK_DEV_IDE_AU1XXX_PIO_DBDMA */ + #define NUM_DESCRIPTORS 2 +#endif + +#ifndef AU1XXX_ATA_RQSIZE + #define AU1XXX_ATA_RQSIZE 128 +#endif + +/* Disable Burstable-Support for DBDMA */ +#ifndef CONFIG_BLK_DEV_IDE_AU1XXX_BURSTABLE_ON + #define CONFIG_BLK_DEV_IDE_AU1XXX_BURSTABLE_ON 0 +#endif + +#ifdef CONFIG_PM +/* +* This will enable the device to be powered up when write() or read() +* is called. If this is not defined, the driver will return -EBUSY. +*/ +#define WAKE_ON_ACCESS 1 + +typedef struct +{ + spinlock_t lock; /* Used to block on state transitions */ + au1xxx_power_dev_t *dev; /* Power Managers device structure */ + unsigned stopped; /* USed to signaling device is stopped */ +} pm_state; +#endif + + +typedef struct +{ + u32 tx_dev_id, rx_dev_id, target_dev_id; + u32 tx_chan, rx_chan; + void *tx_desc_head, *rx_desc_head; + ide_hwif_t *hwif; +#ifdef CONFIG_BLK_DEV_IDE_AU1XXX_MDMA2_DBDMA + ide_drive_t *drive; + u8 white_list, black_list; + struct dbdma_cmd *dma_table_cpu; + dma_addr_t dma_table_dma; + struct scatterlist *sg_table; + int sg_nents; + int sg_dma_direction; +#endif + struct device *dev; + int irq; + u32 regbase; +#ifdef CONFIG_PM + pm_state pm; +#endif +} _auide_hwif; + +#ifdef CONFIG_BLK_DEV_IDE_AU1XXX_MDMA2_DBDMA +struct drive_list_entry { + const char * id_model; + const char * id_firmware; +}; + +/* HD white list */ +static const struct drive_list_entry dma_white_list [] = { +/* + * Hitachi + */ + { "HITACHI_DK14FA-20" , "ALL" }, + { "HTS726060M9AT00" , "ALL" }, +/* + * Maxtor + */ + { "Maxtor 6E040L0" , "ALL" }, + { "Maxtor 6Y080P0" , "ALL" }, + { "Maxtor 6Y160P0" , "ALL" }, +/* + * Seagate + */ + { "ST3120026A" , "ALL" }, + { "ST320014A" , "ALL" }, + { "ST94011A" , "ALL" }, + { "ST340016A" , "ALL" }, +/* + * Western Digital + */ + { "WDC WD400UE-00HCT0" , "ALL" }, + { "WDC WD400JB-00JJC0" , "ALL" }, + { NULL , NULL } +}; + +/* HD black list */ +static const struct drive_list_entry dma_black_list [] = { +/* + * Western Digital + */ + { "WDC WD100EB-00CGH0" , "ALL" }, + { "WDC WD200BB-00AUA1" , "ALL" }, + { "WDC AC24300L" , "ALL" }, + { NULL , NULL } +}; +#endif + +/* function prototyping */ +u8 auide_inb(unsigned long port); +u16 auide_inw(unsigned long port); +u32 auide_inl(unsigned long port); +void auide_insw(unsigned long port, void *addr, u32 count); +void auide_insl(unsigned long port, void *addr, u32 count); +void auide_outb(u8 addr, unsigned long port); +void auide_outbsync(ide_drive_t *drive, u8 addr, unsigned long port); +void auide_outw(u16 addr, unsigned long port); +void auide_outl(u32 addr, unsigned long port); +void auide_outsw(unsigned long port, void *addr, u32 count); +void auide_outsl(unsigned long port, void *addr, u32 count); +static void auide_tune_drive(ide_drive_t *drive, byte pio); +static int auide_tune_chipset (ide_drive_t *drive, u8 speed); +static int auide_ddma_init( _auide_hwif *auide ); +static void auide_setup_ports(hw_regs_t *hw, _auide_hwif *ahwif); +int __init auide_probe(void); + +#ifdef CONFIG_PM + int au1200ide_pm_callback( au1xxx_power_dev_t *dev, + au1xxx_request_t request, void *data); + static int au1xxxide_pm_standby( au1xxx_power_dev_t *dev ); + static int au1xxxide_pm_sleep( au1xxx_power_dev_t *dev ); + static int au1xxxide_pm_resume( au1xxx_power_dev_t *dev ); + static int au1xxxide_pm_getstatus( au1xxx_power_dev_t *dev ); + static int au1xxxide_pm_access( au1xxx_power_dev_t *dev ); + static int au1xxxide_pm_idle( au1xxx_power_dev_t *dev ); + static int au1xxxide_pm_cleanup( au1xxx_power_dev_t *dev ); +#endif + + +/* + * Multi-Word DMA + DbDMA functions + */ +#ifdef CONFIG_BLK_DEV_IDE_AU1XXX_MDMA2_DBDMA + + static int in_drive_list(struct hd_driveid *id, + const struct drive_list_entry *drive_table); + static int auide_build_sglist(ide_drive_t *drive, struct request *rq); + static int auide_build_dmatable(ide_drive_t *drive); + static int auide_dma_end(ide_drive_t *drive); + static void auide_dma_start(ide_drive_t *drive ); + ide_startstop_t auide_dma_intr (ide_drive_t *drive); + static void auide_dma_exec_cmd(ide_drive_t *drive, u8 command); + static int auide_dma_setup(ide_drive_t *drive); + static int auide_dma_check(ide_drive_t *drive); + static int auide_dma_test_irq(ide_drive_t *drive); + static int auide_dma_host_off(ide_drive_t *drive); + static int auide_dma_host_on(ide_drive_t *drive); + static int auide_dma_lostirq(ide_drive_t *drive); + static int auide_dma_on(ide_drive_t *drive); + static void auide_ddma_tx_callback(int irq, void *param, + struct pt_regs *regs); + static void auide_ddma_rx_callback(int irq, void *param, + struct pt_regs *regs); + static int auide_dma_off_quietly(ide_drive_t *drive); + static int auide_dma_timeout(ide_drive_t *drive); + +#endif /* end CONFIG_BLK_DEV_IDE_AU1XXX_MDMA2_DBDMA */ + +/******************************************************************************* +* PIO Mode timing calculation : * +* * +* Static Bus Spec ATA Spec * +* Tcsoe = t1 * +* Toecs = t9 * +* Twcs = t9 * +* Tcsh = t2i | t2 * +* Tcsoff = t2i | t2 * +* Twp = t2 * +* Tcsw = t1 * +* Tpm = 0 * +* Ta = t1+t2 * +*******************************************************************************/ + +#define TCSOE_MASK (0x07<<29) +#define TOECS_MASK (0x07<<26) +#define TWCS_MASK (0x07<<28) +#define TCSH_MASK (0x0F<<24) +#define TCSOFF_MASK (0x07<<20) +#define TWP_MASK (0x3F<<14) +#define TCSW_MASK (0x0F<<10) +#define TPM_MASK (0x0F<<6) +#define TA_MASK (0x3F<<0) +#define TS_MASK (1<<8) + +/* Timing parameters PIO mode 0 */ +#define SBC_IDE_PIO0_TCSOE (0x04<<29) +#define SBC_IDE_PIO0_TOECS (0x01<<26) +#define SBC_IDE_PIO0_TWCS (0x02<<28) +#define SBC_IDE_PIO0_TCSH (0x08<<24) +#define SBC_IDE_PIO0_TCSOFF (0x07<<20) +#define SBC_IDE_PIO0_TWP (0x10<<14) +#define SBC_IDE_PIO0_TCSW (0x04<<10) +#define SBC_IDE_PIO0_TPM (0x0<<6) +#define SBC_IDE_PIO0_TA (0x15<<0) +/* Timing parameters PIO mode 1 */ +#define SBC_IDE_PIO1_TCSOE (0x03<<29) +#define SBC_IDE_PIO1_TOECS (0x01<<26) +#define SBC_IDE_PIO1_TWCS (0x01<<28) +#define SBC_IDE_PIO1_TCSH (0x06<<24) +#define SBC_IDE_PIO1_TCSOFF (0x06<<20) +#define SBC_IDE_PIO1_TWP (0x08<<14) +#define SBC_IDE_PIO1_TCSW (0x03<<10) +#define SBC_IDE_PIO1_TPM (0x00<<6) +#define SBC_IDE_PIO1_TA (0x0B<<0) +/* Timing parameters PIO mode 2 */ +#define SBC_IDE_PIO2_TCSOE (0x05<<29) +#define SBC_IDE_PIO2_TOECS (0x01<<26) +#define SBC_IDE_PIO2_TWCS (0x01<<28) +#define SBC_IDE_PIO2_TCSH (0x07<<24) +#define SBC_IDE_PIO2_TCSOFF (0x07<<20) +#define SBC_IDE_PIO2_TWP (0x1F<<14) +#define SBC_IDE_PIO2_TCSW (0x05<<10) +#define SBC_IDE_PIO2_TPM (0x00<<6) +#define SBC_IDE_PIO2_TA (0x22<<0) +/* Timing parameters PIO mode 3 */ +#define SBC_IDE_PIO3_TCSOE (0x05<<29) +#define SBC_IDE_PIO3_TOECS (0x01<<26) +#define SBC_IDE_PIO3_TWCS (0x01<<28) +#define SBC_IDE_PIO3_TCSH (0x0D<<24) +#define SBC_IDE_PIO3_TCSOFF (0x0D<<20) +#define SBC_IDE_PIO3_TWP (0x15<<14) +#define SBC_IDE_PIO3_TCSW (0x05<<10) +#define SBC_IDE_PIO3_TPM (0x00<<6) +#define SBC_IDE_PIO3_TA (0x1A<<0) +/* Timing parameters PIO mode 4 */ +#define SBC_IDE_PIO4_TCSOE (0x04<<29) +#define SBC_IDE_PIO4_TOECS (0x01<<26) +#define SBC_IDE_PIO4_TWCS (0x01<<28) +#define SBC_IDE_PIO4_TCSH (0x04<<24) +#define SBC_IDE_PIO4_TCSOFF (0x04<<20) +#define SBC_IDE_PIO4_TWP (0x0D<<14) +#define SBC_IDE_PIO4_TCSW (0x03<<10) +#define SBC_IDE_PIO4_TPM (0x00<<6) +#define SBC_IDE_PIO4_TA (0x12<<0) +/* Timing parameters MDMA mode 0 */ +#define SBC_IDE_MDMA0_TCSOE (0x03<<29) +#define SBC_IDE_MDMA0_TOECS (0x01<<26) +#define SBC_IDE_MDMA0_TWCS (0x01<<28) +#define SBC_IDE_MDMA0_TCSH (0x07<<24) +#define SBC_IDE_MDMA0_TCSOFF (0x07<<20) +#define SBC_IDE_MDMA0_TWP (0x0C<<14) +#define SBC_IDE_MDMA0_TCSW (0x03<<10) +#define SBC_IDE_MDMA0_TPM (0x00<<6) +#define SBC_IDE_MDMA0_TA (0x0F<<0) +/* Timing parameters MDMA mode 1 */ +#define SBC_IDE_MDMA1_TCSOE (0x05<<29) +#define SBC_IDE_MDMA1_TOECS (0x01<<26) +#define SBC_IDE_MDMA1_TWCS (0x01<<28) +#define SBC_IDE_MDMA1_TCSH (0x05<<24) +#define SBC_IDE_MDMA1_TCSOFF (0x05<<20) +#define SBC_IDE_MDMA1_TWP (0x0F<<14) +#define SBC_IDE_MDMA1_TCSW (0x05<<10) +#define SBC_IDE_MDMA1_TPM (0x00<<6) +#define SBC_IDE_MDMA1_TA (0x15<<0) +/* Timing parameters MDMA mode 2 */ +#define SBC_IDE_MDMA2_TCSOE (0x04<<29) +#define SBC_IDE_MDMA2_TOECS (0x01<<26) +#define SBC_IDE_MDMA2_TWCS (0x01<<28) +#define SBC_IDE_MDMA2_TCSH (0x04<<24) +#define SBC_IDE_MDMA2_TCSOFF (0x04<<20) +#define SBC_IDE_MDMA2_TWP (0x0D<<14) +#define SBC_IDE_MDMA2_TCSW (0x04<<10) +#define SBC_IDE_MDMA2_TPM (0x00<<6) +#define SBC_IDE_MDMA2_TA (0x12<<0) + diff --git a/include/linux/ide.h b/include/linux/ide.h index a6dbb51ecd7b..3461abc1e854 100644 --- a/include/linux/ide.h +++ b/include/linux/ide.h @@ -218,7 +218,7 @@ typedef enum { ide_unknown, ide_generic, ide_pci, ide_rz1000, ide_trm290, ide_cmd646, ide_cy82c693, ide_4drives, ide_pmac, ide_etrax100, ide_acorn, - ide_forced + ide_au1xxx, ide_forced } hwif_chipset_t; /* -- cgit v1.2.3 From d32311fed70d12f14e585feb4653571b1e2b0e6d Mon Sep 17 00:00:00 2001 From: Herbert Xu Date: Sat, 17 Sep 2005 14:41:40 +1000 Subject: [PATCH] Introduce sg_set_buf sg_init_one is a nice tool for the block layer. However, users of struct scatterlist in other subsystems don't usually need the DMA attributes. For them it's a waste of time and space to initialise the whole struct scatterlist structure. Therefore this patch adds a new function sg_set_buf to initialise a scatterlist without zeroing the DMA attributes. Signed-off-by: Herbert Xu --- include/linux/scatterlist.h | 17 +++++++++++++---- 1 file changed, 13 insertions(+), 4 deletions(-) (limited to 'include/linux') diff --git a/include/linux/scatterlist.h b/include/linux/scatterlist.h index 7f717e95ae37..66ff545552f7 100644 --- a/include/linux/scatterlist.h +++ b/include/linux/scatterlist.h @@ -1,14 +1,23 @@ #ifndef _LINUX_SCATTERLIST_H #define _LINUX_SCATTERLIST_H -static inline void sg_init_one(struct scatterlist *sg, - u8 *buf, unsigned int buflen) -{ - memset(sg, 0, sizeof(*sg)); +#include +#include +#include +static inline void sg_set_buf(struct scatterlist *sg, void *buf, + unsigned int buflen) +{ sg->page = virt_to_page(buf); sg->offset = offset_in_page(buf); sg->length = buflen; } +static inline void sg_init_one(struct scatterlist *sg, void *buf, + unsigned int buflen) +{ + memset(sg, 0, sizeof(*sg)); + sg_set_buf(sg, buf, buflen); +} + #endif /* _LINUX_SCATTERLIST_H */ -- cgit v1.2.3 From 0169e284f6b6b263cc7c2ed25986b96cd6fda610 Mon Sep 17 00:00:00 2001 From: Jeff Garzik Date: Sat, 29 Oct 2005 21:25:10 -0400 Subject: [libata] remove ata_chk_err(), ->check_err() hook. We now depend on ->tf_read() to provide us with the contents of the Error shadow register. --- drivers/scsi/ahci.c | 9 --------- drivers/scsi/libata-core.c | 41 +++++++++-------------------------------- drivers/scsi/sata_mv.c | 18 ------------------ drivers/scsi/sata_sil24.c | 8 -------- include/linux/libata.h | 2 -- 5 files changed, 9 insertions(+), 69 deletions(-) (limited to 'include/linux') diff --git a/drivers/scsi/ahci.c b/drivers/scsi/ahci.c index fe8187d6f58b..03829aedfd39 100644 --- a/drivers/scsi/ahci.c +++ b/drivers/scsi/ahci.c @@ -192,7 +192,6 @@ static void ahci_port_stop(struct ata_port *ap); static void ahci_tf_read(struct ata_port *ap, struct ata_taskfile *tf); static void ahci_qc_prep(struct ata_queued_cmd *qc); static u8 ahci_check_status(struct ata_port *ap); -static u8 ahci_check_err(struct ata_port *ap); static inline int ahci_host_intr(struct ata_port *ap, struct ata_queued_cmd *qc); static void ahci_remove_one (struct pci_dev *pdev); @@ -221,7 +220,6 @@ static const struct ata_port_operations ahci_ops = { .check_status = ahci_check_status, .check_altstatus = ahci_check_status, - .check_err = ahci_check_err, .dev_select = ata_noop_dev_select, .tf_read = ahci_tf_read, @@ -458,13 +456,6 @@ static u8 ahci_check_status(struct ata_port *ap) return readl(mmio + PORT_TFDATA) & 0xFF; } -static u8 ahci_check_err(struct ata_port *ap) -{ - void __iomem *mmio = (void __iomem *) ap->ioaddr.cmd_addr; - - return (readl(mmio + PORT_TFDATA) >> 8) & 0xFF; -} - static void ahci_tf_read(struct ata_port *ap, struct ata_taskfile *tf) { struct ahci_port_priv *pp = ap->private_data; diff --git a/drivers/scsi/libata-core.c b/drivers/scsi/libata-core.c index b1b1c6f01419..d2f71a2331bb 100644 --- a/drivers/scsi/libata-core.c +++ b/drivers/scsi/libata-core.c @@ -371,7 +371,7 @@ static void ata_tf_read_pio(struct ata_port *ap, struct ata_taskfile *tf) struct ata_ioports *ioaddr = &ap->ioaddr; tf->command = ata_check_status(ap); - tf->feature = ata_chk_err(ap); + tf->feature = inb(ioaddr->error_addr); tf->nsect = inb(ioaddr->nsect_addr); tf->lbal = inb(ioaddr->lbal_addr); tf->lbam = inb(ioaddr->lbam_addr); @@ -405,7 +405,7 @@ static void ata_tf_read_mmio(struct ata_port *ap, struct ata_taskfile *tf) struct ata_ioports *ioaddr = &ap->ioaddr; tf->command = ata_check_status(ap); - tf->feature = ata_chk_err(ap); + tf->feature = readb((void __iomem *)ioaddr->error_addr); tf->nsect = readb((void __iomem *)ioaddr->nsect_addr); tf->lbal = readb((void __iomem *)ioaddr->lbal_addr); tf->lbam = readb((void __iomem *)ioaddr->lbam_addr); @@ -525,30 +525,6 @@ u8 ata_altstatus(struct ata_port *ap) } -/** - * ata_chk_err - Read device error reg - * @ap: port where the device is - * - * Reads ATA taskfile error register for - * currently-selected device and return its value. - * - * Note: may NOT be used as the check_err() entry in - * ata_port_operations. - * - * LOCKING: - * Inherited from caller. - */ -u8 ata_chk_err(struct ata_port *ap) -{ - if (ap->ops->check_err) - return ap->ops->check_err(ap); - - if (ap->flags & ATA_FLAG_MMIO) { - return readb((void __iomem *) ap->ioaddr.error_addr); - } - return inb(ap->ioaddr.error_addr); -} - /** * ata_tf_to_fis - Convert ATA taskfile to SATA FIS structure * @tf: Taskfile to convert @@ -901,8 +877,8 @@ static u8 ata_dev_try_classify(struct ata_port *ap, unsigned int device) memset(&tf, 0, sizeof(tf)); - err = ata_chk_err(ap); ap->ops->tf_read(ap, &tf); + err = tf.feature; dev->class = ATA_DEV_NONE; @@ -1139,7 +1115,6 @@ static void ata_dev_identify(struct ata_port *ap, unsigned int device) unsigned int major_version; u16 tmp; unsigned long xfer_modes; - u8 status; unsigned int using_edd; DECLARE_COMPLETION(wait); struct ata_queued_cmd *qc; @@ -1193,8 +1168,11 @@ retry: else wait_for_completion(&wait); - status = ata_chk_status(ap); - if (status & ATA_ERR) { + spin_lock_irqsave(&ap->host_set->lock, flags); + ap->ops->tf_read(ap, &qc->tf); + spin_unlock_irqrestore(&ap->host_set->lock, flags); + + if (qc->tf.command & ATA_ERR) { /* * arg! EDD works for all test cases, but seems to return * the ATA signature for some ATAPI devices. Until the @@ -1207,7 +1185,7 @@ retry: * to have this problem. */ if ((using_edd) && (qc->tf.command == ATA_CMD_ID_ATA)) { - u8 err = ata_chk_err(ap); + u8 err = qc->tf.feature; if (err & ATA_ABORTED) { dev->class = ATA_DEV_ATAPI; qc->cursg = 0; @@ -4873,7 +4851,6 @@ EXPORT_SYMBOL_GPL(ata_tf_to_fis); EXPORT_SYMBOL_GPL(ata_tf_from_fis); EXPORT_SYMBOL_GPL(ata_check_status); EXPORT_SYMBOL_GPL(ata_altstatus); -EXPORT_SYMBOL_GPL(ata_chk_err); EXPORT_SYMBOL_GPL(ata_exec_command); EXPORT_SYMBOL_GPL(ata_port_start); EXPORT_SYMBOL_GPL(ata_port_stop); diff --git a/drivers/scsi/sata_mv.c b/drivers/scsi/sata_mv.c index 422e0b6f603a..dcef5fe8600b 100644 --- a/drivers/scsi/sata_mv.c +++ b/drivers/scsi/sata_mv.c @@ -258,7 +258,6 @@ struct mv_host_priv { static void mv_irq_clear(struct ata_port *ap); static u32 mv_scr_read(struct ata_port *ap, unsigned int sc_reg_in); static void mv_scr_write(struct ata_port *ap, unsigned int sc_reg_in, u32 val); -static u8 mv_check_err(struct ata_port *ap); static void mv_phy_reset(struct ata_port *ap); static void mv_host_stop(struct ata_host_set *host_set); static int mv_port_start(struct ata_port *ap); @@ -296,7 +295,6 @@ static const struct ata_port_operations mv_ops = { .tf_load = ata_tf_load, .tf_read = ata_tf_read, .check_status = ata_check_status, - .check_err = mv_check_err, .exec_command = ata_exec_command, .dev_select = ata_std_dev_select, @@ -1184,22 +1182,6 @@ static irqreturn_t mv_interrupt(int irq, void *dev_instance, return IRQ_RETVAL(handled); } -/** - * mv_check_err - Return the error shadow register to caller. - * @ap: ATA channel to manipulate - * - * Marvell requires DMA to be stopped before accessing shadow - * registers. So we do that, then return the needed register. - * - * LOCKING: - * Inherited from caller. FIXME: protect mv_stop_dma with lock? - */ -static u8 mv_check_err(struct ata_port *ap) -{ - mv_stop_dma(ap); /* can't read shadow regs if DMA on */ - return readb((void __iomem *) ap->ioaddr.error_addr); -} - /** * mv_phy_reset - Perform eDMA reset followed by COMRESET * @ap: ATA channel to manipulate diff --git a/drivers/scsi/sata_sil24.c b/drivers/scsi/sata_sil24.c index 51855d3bac64..e18a1e2bb65e 100644 --- a/drivers/scsi/sata_sil24.c +++ b/drivers/scsi/sata_sil24.c @@ -225,7 +225,6 @@ struct sil24_host_priv { }; static u8 sil24_check_status(struct ata_port *ap); -static u8 sil24_check_err(struct ata_port *ap); static u32 sil24_scr_read(struct ata_port *ap, unsigned sc_reg); static void sil24_scr_write(struct ata_port *ap, unsigned sc_reg, u32 val); static void sil24_tf_read(struct ata_port *ap, struct ata_taskfile *tf); @@ -280,7 +279,6 @@ static const struct ata_port_operations sil24_ops = { .check_status = sil24_check_status, .check_altstatus = sil24_check_status, - .check_err = sil24_check_err, .dev_select = ata_noop_dev_select, .tf_read = sil24_tf_read, @@ -363,12 +361,6 @@ static u8 sil24_check_status(struct ata_port *ap) return pp->tf.command; } -static u8 sil24_check_err(struct ata_port *ap) -{ - struct sil24_port_priv *pp = ap->private_data; - return pp->tf.feature; -} - static int sil24_scr_map[] = { [SCR_CONTROL] = 0, [SCR_STATUS] = 1, diff --git a/include/linux/libata.h b/include/linux/libata.h index 00a8a5738858..a4cce9936a80 100644 --- a/include/linux/libata.h +++ b/include/linux/libata.h @@ -347,7 +347,6 @@ struct ata_port_operations { void (*exec_command)(struct ata_port *ap, const struct ata_taskfile *tf); u8 (*check_status)(struct ata_port *ap); u8 (*check_altstatus)(struct ata_port *ap); - u8 (*check_err)(struct ata_port *ap); void (*dev_select)(struct ata_port *ap, unsigned int device); void (*phy_reset) (struct ata_port *ap); @@ -434,7 +433,6 @@ extern void ata_noop_dev_select (struct ata_port *ap, unsigned int device); extern void ata_std_dev_select (struct ata_port *ap, unsigned int device); extern u8 ata_check_status(struct ata_port *ap); extern u8 ata_altstatus(struct ata_port *ap); -extern u8 ata_chk_err(struct ata_port *ap); extern void ata_exec_command(struct ata_port *ap, const struct ata_taskfile *tf); extern int ata_port_start (struct ata_port *ap); extern void ata_port_stop (struct ata_port *ap); -- cgit v1.2.3 From 930fc45a49ddebe7555cc5c837d82b9c27e65ff4 Mon Sep 17 00:00:00 2001 From: Christoph Lameter Date: Sat, 29 Oct 2005 18:15:41 -0700 Subject: [PATCH] vmalloc_node This patch adds vmalloc_node(size, node) -> Allocate necessary memory on the specified node and get_vm_area_node(size, flags, node) and the other functions that it depends on. Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/vmalloc.h | 8 +++++- mm/vmalloc.c | 73 ++++++++++++++++++++++++++++++++++++++----------- 2 files changed, 64 insertions(+), 17 deletions(-) (limited to 'include/linux') diff --git a/include/linux/vmalloc.h b/include/linux/vmalloc.h index 3701a0673d2c..1d5577b2b752 100644 --- a/include/linux/vmalloc.h +++ b/include/linux/vmalloc.h @@ -32,10 +32,14 @@ struct vm_struct { * Highlevel APIs for driver use */ extern void *vmalloc(unsigned long size); +extern void *vmalloc_node(unsigned long size, int node); extern void *vmalloc_exec(unsigned long size); extern void *vmalloc_32(unsigned long size); extern void *__vmalloc(unsigned long size, gfp_t gfp_mask, pgprot_t prot); -extern void *__vmalloc_area(struct vm_struct *area, gfp_t gfp_mask, pgprot_t prot); +extern void *__vmalloc_area(struct vm_struct *area, gfp_t gfp_mask, + pgprot_t prot); +extern void *__vmalloc_node(unsigned long size, gfp_t gfp_mask, + pgprot_t prot, int node); extern void vfree(void *addr); extern void *vmap(struct page **pages, unsigned int count, @@ -48,6 +52,8 @@ extern void vunmap(void *addr); extern struct vm_struct *get_vm_area(unsigned long size, unsigned long flags); extern struct vm_struct *__get_vm_area(unsigned long size, unsigned long flags, unsigned long start, unsigned long end); +extern struct vm_struct *get_vm_area_node(unsigned long size, + unsigned long flags, int node); extern struct vm_struct *remove_vm_area(void *addr); extern struct vm_struct *__remove_vm_area(void *addr); extern int map_vm_area(struct vm_struct *area, pgprot_t prot, diff --git a/mm/vmalloc.c b/mm/vmalloc.c index 1150229b6366..5e9120598799 100644 --- a/mm/vmalloc.c +++ b/mm/vmalloc.c @@ -5,6 +5,7 @@ * Support of BIGMEM added by Gerhard Wichert, Siemens AG, July 1999 * SMP-safe vmalloc/vfree/ioremap, Tigran Aivazian , May 2000 * Major rework to support vmap/vunmap, Christoph Hellwig, SGI, August 2002 + * Numa awareness, Christoph Lameter, SGI, June 2005 */ #include @@ -158,8 +159,8 @@ int map_vm_area(struct vm_struct *area, pgprot_t prot, struct page ***pages) return err; } -struct vm_struct *__get_vm_area(unsigned long size, unsigned long flags, - unsigned long start, unsigned long end) +struct vm_struct *__get_vm_area_node(unsigned long size, unsigned long flags, + unsigned long start, unsigned long end, int node) { struct vm_struct **p, *tmp, *area; unsigned long align = 1; @@ -178,7 +179,7 @@ struct vm_struct *__get_vm_area(unsigned long size, unsigned long flags, addr = ALIGN(start, align); size = PAGE_ALIGN(size); - area = kmalloc(sizeof(*area), GFP_KERNEL); + area = kmalloc_node(sizeof(*area), GFP_KERNEL, node); if (unlikely(!area)) return NULL; @@ -231,6 +232,12 @@ out: return NULL; } +struct vm_struct *__get_vm_area(unsigned long size, unsigned long flags, + unsigned long start, unsigned long end) +{ + return __get_vm_area_node(size, flags, start, end, -1); +} + /** * get_vm_area - reserve a contingous kernel virtual area * @@ -246,6 +253,11 @@ struct vm_struct *get_vm_area(unsigned long size, unsigned long flags) return __get_vm_area(size, flags, VMALLOC_START, VMALLOC_END); } +struct vm_struct *get_vm_area_node(unsigned long size, unsigned long flags, int node) +{ + return __get_vm_area_node(size, flags, VMALLOC_START, VMALLOC_END, node); +} + /* Caller must hold vmlist_lock */ struct vm_struct *__remove_vm_area(void *addr) { @@ -342,7 +354,6 @@ void vfree(void *addr) BUG_ON(in_interrupt()); __vunmap(addr, 1); } - EXPORT_SYMBOL(vfree); /** @@ -360,7 +371,6 @@ void vunmap(void *addr) BUG_ON(in_interrupt()); __vunmap(addr, 0); } - EXPORT_SYMBOL(vunmap); /** @@ -392,10 +402,10 @@ void *vmap(struct page **pages, unsigned int count, return area->addr; } - EXPORT_SYMBOL(vmap); -void *__vmalloc_area(struct vm_struct *area, gfp_t gfp_mask, pgprot_t prot) +void *__vmalloc_area_node(struct vm_struct *area, gfp_t gfp_mask, + pgprot_t prot, int node) { struct page **pages; unsigned int nr_pages, array_size, i; @@ -406,9 +416,9 @@ void *__vmalloc_area(struct vm_struct *area, gfp_t gfp_mask, pgprot_t prot) area->nr_pages = nr_pages; /* Please note that the recursion is strictly bounded. */ if (array_size > PAGE_SIZE) - pages = __vmalloc(array_size, gfp_mask, PAGE_KERNEL); + pages = __vmalloc_node(array_size, gfp_mask, PAGE_KERNEL, node); else - pages = kmalloc(array_size, (gfp_mask & ~__GFP_HIGHMEM)); + pages = kmalloc_node(array_size, (gfp_mask & ~__GFP_HIGHMEM), node); area->pages = pages; if (!area->pages) { remove_vm_area(area->addr); @@ -418,7 +428,10 @@ void *__vmalloc_area(struct vm_struct *area, gfp_t gfp_mask, pgprot_t prot) memset(area->pages, 0, array_size); for (i = 0; i < area->nr_pages; i++) { - area->pages[i] = alloc_page(gfp_mask); + if (node < 0) + area->pages[i] = alloc_page(gfp_mask); + else + area->pages[i] = alloc_pages_node(node, gfp_mask, 0); if (unlikely(!area->pages[i])) { /* Successfully allocated i pages, free them in __vunmap() */ area->nr_pages = i; @@ -435,18 +448,25 @@ fail: return NULL; } +void *__vmalloc_area(struct vm_struct *area, gfp_t gfp_mask, pgprot_t prot) +{ + return __vmalloc_area_node(area, gfp_mask, prot, -1); +} + /** - * __vmalloc - allocate virtually contiguous memory + * __vmalloc_node - allocate virtually contiguous memory * * @size: allocation size * @gfp_mask: flags for the page level allocator * @prot: protection mask for the allocated pages + * @node node to use for allocation or -1 * * Allocate enough pages to cover @size from the page level * allocator with @gfp_mask flags. Map them into contiguous * kernel virtual space, using a pagetable protection of @prot. */ -void *__vmalloc(unsigned long size, gfp_t gfp_mask, pgprot_t prot) +void *__vmalloc_node(unsigned long size, gfp_t gfp_mask, pgprot_t prot, + int node) { struct vm_struct *area; @@ -454,13 +474,18 @@ void *__vmalloc(unsigned long size, gfp_t gfp_mask, pgprot_t prot) if (!size || (size >> PAGE_SHIFT) > num_physpages) return NULL; - area = get_vm_area(size, VM_ALLOC); + area = get_vm_area_node(size, VM_ALLOC, node); if (!area) return NULL; - return __vmalloc_area(area, gfp_mask, prot); + return __vmalloc_area_node(area, gfp_mask, prot, node); } +EXPORT_SYMBOL(__vmalloc_node); +void *__vmalloc(unsigned long size, gfp_t gfp_mask, pgprot_t prot) +{ + return __vmalloc_node(size, gfp_mask, prot, -1); +} EXPORT_SYMBOL(__vmalloc); /** @@ -478,9 +503,26 @@ void *vmalloc(unsigned long size) { return __vmalloc(size, GFP_KERNEL | __GFP_HIGHMEM, PAGE_KERNEL); } - EXPORT_SYMBOL(vmalloc); +/** + * vmalloc_node - allocate memory on a specific node + * + * @size: allocation size + * @node; numa node + * + * Allocate enough pages to cover @size from the page level + * allocator and map them into contiguous kernel virtual space. + * + * For tight cotrol over page level allocator and protection flags + * use __vmalloc() instead. + */ +void *vmalloc_node(unsigned long size, int node) +{ + return __vmalloc_node(size, GFP_KERNEL | __GFP_HIGHMEM, PAGE_KERNEL, node); +} +EXPORT_SYMBOL(vmalloc_node); + #ifndef PAGE_KERNEL_EXEC # define PAGE_KERNEL_EXEC PAGE_KERNEL #endif @@ -515,7 +557,6 @@ void *vmalloc_32(unsigned long size) { return __vmalloc(size, GFP_KERNEL, PAGE_KERNEL); } - EXPORT_SYMBOL(vmalloc_32); long vread(char *buf, char *addr, unsigned long count) -- cgit v1.2.3 From eb92f4ef320b738e41ad43476a5d05c8a20d5cc7 Mon Sep 17 00:00:00 2001 From: Rik Van Riel Date: Sat, 29 Oct 2005 18:15:44 -0700 Subject: [PATCH] add sem_is_read/write_locked() Add sem_is_read/write_locked functions to the read/write semaphores, along the same lines of the *_is_locked spinlock functions. The swap token tuning patch uses sem_is_read_locked; sem_is_write_locked is added for completeness. Signed-off-by: Rik van Riel Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/asm-alpha/rwsem.h | 5 +++++ include/asm-i386/rwsem.h | 5 +++++ include/asm-ia64/rwsem.h | 5 +++++ include/asm-ppc/rwsem.h | 5 +++++ include/asm-ppc64/rwsem.h | 5 +++++ include/asm-s390/rwsem.h | 5 +++++ include/asm-sh/rwsem.h | 5 +++++ include/asm-sparc64/rwsem.h | 5 +++++ include/asm-x86_64/rwsem.h | 5 +++++ include/linux/rwsem-spinlock.h | 5 +++++ 10 files changed, 50 insertions(+) (limited to 'include/linux') diff --git a/include/asm-alpha/rwsem.h b/include/asm-alpha/rwsem.h index 8e058a67c9a4..fafdd4f7010a 100644 --- a/include/asm-alpha/rwsem.h +++ b/include/asm-alpha/rwsem.h @@ -262,5 +262,10 @@ static inline long rwsem_atomic_update(long val, struct rw_semaphore *sem) #endif } +static inline int rwsem_is_locked(struct rw_semaphore *sem) +{ + return (sem->count != 0); +} + #endif /* __KERNEL__ */ #endif /* _ALPHA_RWSEM_H */ diff --git a/include/asm-i386/rwsem.h b/include/asm-i386/rwsem.h index 7625a675852f..be4ab859238e 100644 --- a/include/asm-i386/rwsem.h +++ b/include/asm-i386/rwsem.h @@ -284,5 +284,10 @@ LOCK_PREFIX "xadd %0,(%2)" return tmp+delta; } +static inline int rwsem_is_locked(struct rw_semaphore *sem) +{ + return (sem->count != 0); +} + #endif /* __KERNEL__ */ #endif /* _I386_RWSEM_H */ diff --git a/include/asm-ia64/rwsem.h b/include/asm-ia64/rwsem.h index e18b5ab0cb75..1327c91ea39c 100644 --- a/include/asm-ia64/rwsem.h +++ b/include/asm-ia64/rwsem.h @@ -186,4 +186,9 @@ __downgrade_write (struct rw_semaphore *sem) #define rwsem_atomic_add(delta, sem) atomic64_add(delta, (atomic64_t *)(&(sem)->count)) #define rwsem_atomic_update(delta, sem) atomic64_add_return(delta, (atomic64_t *)(&(sem)->count)) +static inline int rwsem_is_locked(struct rw_semaphore *sem) +{ + return (sem->count != 0); +} + #endif /* _ASM_IA64_RWSEM_H */ diff --git a/include/asm-ppc/rwsem.h b/include/asm-ppc/rwsem.h index 3e738f483c11..3501ea72f88c 100644 --- a/include/asm-ppc/rwsem.h +++ b/include/asm-ppc/rwsem.h @@ -168,5 +168,10 @@ static inline int rwsem_atomic_update(int delta, struct rw_semaphore *sem) return atomic_add_return(delta, (atomic_t *)(&sem->count)); } +static inline int rwsem_is_locked(struct rw_semaphore *sem) +{ + return (sem->count != 0); +} + #endif /* __KERNEL__ */ #endif /* _PPC_RWSEM_XADD_H */ diff --git a/include/asm-ppc64/rwsem.h b/include/asm-ppc64/rwsem.h index bd5c2f093575..7a647fae3765 100644 --- a/include/asm-ppc64/rwsem.h +++ b/include/asm-ppc64/rwsem.h @@ -163,5 +163,10 @@ static inline int rwsem_atomic_update(int delta, struct rw_semaphore *sem) return atomic_add_return(delta, (atomic_t *)(&sem->count)); } +static inline int rwsem_is_locked(struct rw_semaphore *sem) +{ + return (sem->count != 0); +} + #endif /* __KERNEL__ */ #endif /* _PPC_RWSEM_XADD_H */ diff --git a/include/asm-s390/rwsem.h b/include/asm-s390/rwsem.h index 8c0cebbfc034..0422a085dd56 100644 --- a/include/asm-s390/rwsem.h +++ b/include/asm-s390/rwsem.h @@ -351,5 +351,10 @@ static inline long rwsem_atomic_update(long delta, struct rw_semaphore *sem) return new; } +static inline int rwsem_is_locked(struct rw_semaphore *sem) +{ + return (sem->count != 0); +} + #endif /* __KERNEL__ */ #endif /* _S390_RWSEM_H */ diff --git a/include/asm-sh/rwsem.h b/include/asm-sh/rwsem.h index 1be4337f5259..0262d3d1e5e0 100644 --- a/include/asm-sh/rwsem.h +++ b/include/asm-sh/rwsem.h @@ -166,5 +166,10 @@ static inline int rwsem_atomic_update(int delta, struct rw_semaphore *sem) return atomic_add_return(delta, (atomic_t *)(&sem->count)); } +static inline int rwsem_is_locked(struct rw_semaphore *sem) +{ + return (sem->count != 0); +} + #endif /* __KERNEL__ */ #endif /* _ASM_SH_RWSEM_H */ diff --git a/include/asm-sparc64/rwsem.h b/include/asm-sparc64/rwsem.h index 4568ee4022df..cef5e8270421 100644 --- a/include/asm-sparc64/rwsem.h +++ b/include/asm-sparc64/rwsem.h @@ -56,6 +56,11 @@ static inline void rwsem_atomic_add(int delta, struct rw_semaphore *sem) atomic_add(delta, (atomic_t *)(&sem->count)); } +static inline int rwsem_is_locked(struct rw_semaphore *sem) +{ + return (sem->count != 0); +} + #endif /* __KERNEL__ */ #endif /* _SPARC64_RWSEM_H */ diff --git a/include/asm-x86_64/rwsem.h b/include/asm-x86_64/rwsem.h index c002175b6e82..46077e9c1910 100644 --- a/include/asm-x86_64/rwsem.h +++ b/include/asm-x86_64/rwsem.h @@ -274,5 +274,10 @@ LOCK_PREFIX "xaddl %0,(%2)" return tmp+delta; } +static inline int rwsem_is_locked(struct rw_semaphore *sem) +{ + return (sem->count != 0); +} + #endif /* __KERNEL__ */ #endif /* _X8664_RWSEM_H */ diff --git a/include/linux/rwsem-spinlock.h b/include/linux/rwsem-spinlock.h index b52a2af25f1f..f30f805080ae 100644 --- a/include/linux/rwsem-spinlock.h +++ b/include/linux/rwsem-spinlock.h @@ -61,5 +61,10 @@ extern void FASTCALL(__up_read(struct rw_semaphore *sem)); extern void FASTCALL(__up_write(struct rw_semaphore *sem)); extern void FASTCALL(__downgrade_write(struct rw_semaphore *sem)); +static inline int rwsem_is_locked(struct rw_semaphore *sem) +{ + return (sem->activity != 0); +} + #endif /* __KERNEL__ */ #endif /* _LINUX_RWSEM_SPINLOCK_H */ -- cgit v1.2.3 From dfcd3c0dc426bb75770c34b40e14f2da8845ea62 Mon Sep 17 00:00:00 2001 From: Andi Kleen Date: Sat, 29 Oct 2005 18:15:48 -0700 Subject: [PATCH] Convert mempolicies to nodemask_t The NUMA policy code predated nodemask_t so it used open coded bitmaps. Convert everything to nodemask_t. Big patch, but shouldn't have any actual behaviour changes (except I removed one unnecessary check against node_online_map and one unnecessary BUG_ON) Signed-off-by: "Andi Kleen" Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/proc/task_mmu.c | 2 +- include/linux/mempolicy.h | 4 +- mm/mempolicy.c | 120 ++++++++++++++++++++-------------------------- 3 files changed, 56 insertions(+), 70 deletions(-) (limited to 'include/linux') diff --git a/fs/proc/task_mmu.c b/fs/proc/task_mmu.c index c7ef3e48e35b..994612bc72d0 100644 --- a/fs/proc/task_mmu.c +++ b/fs/proc/task_mmu.c @@ -469,7 +469,7 @@ static int show_numa_map(struct seq_file *m, void *v) seq_printf(m, " interleave={"); first = 1; for_each_node(n) { - if (test_bit(n, pol->v.nodes)) { + if (node_isset(n, pol->v.nodes)) { if (!first) seq_putc(m,','); else diff --git a/include/linux/mempolicy.h b/include/linux/mempolicy.h index 58385ee1c0ac..38e60a099399 100644 --- a/include/linux/mempolicy.h +++ b/include/linux/mempolicy.h @@ -27,10 +27,10 @@ #include #include -#include #include #include #include +#include struct vm_area_struct; @@ -63,7 +63,7 @@ struct mempolicy { union { struct zonelist *zonelist; /* bind */ short preferred_node; /* preferred */ - DECLARE_BITMAP(nodes, MAX_NUMNODES); /* interleave */ + nodemask_t nodes; /* interleave */ /* undefined for default */ } v; }; diff --git a/mm/mempolicy.c b/mm/mempolicy.c index 1d5c64df1653..8bc0be1c9efd 100644 --- a/mm/mempolicy.c +++ b/mm/mempolicy.c @@ -93,23 +93,10 @@ struct mempolicy default_policy = { .policy = MPOL_DEFAULT, }; -/* Check if all specified nodes are online */ -static int nodes_online(unsigned long *nodes) -{ - DECLARE_BITMAP(online2, MAX_NUMNODES); - - bitmap_copy(online2, nodes_addr(node_online_map), MAX_NUMNODES); - if (bitmap_empty(online2, MAX_NUMNODES)) - set_bit(0, online2); - if (!bitmap_subset(nodes, online2, MAX_NUMNODES)) - return -EINVAL; - return 0; -} - /* Do sanity checking on a policy */ -static int mpol_check_policy(int mode, unsigned long *nodes) +static int mpol_check_policy(int mode, nodemask_t *nodes) { - int empty = bitmap_empty(nodes, MAX_NUMNODES); + int empty = nodes_empty(*nodes); switch (mode) { case MPOL_DEFAULT: @@ -124,11 +111,11 @@ static int mpol_check_policy(int mode, unsigned long *nodes) return -EINVAL; break; } - return nodes_online(nodes); + return nodes_subset(*nodes, node_online_map) ? 0 : -EINVAL; } /* Copy a node mask from user space. */ -static int get_nodes(unsigned long *nodes, unsigned long __user *nmask, +static int get_nodes(nodemask_t *nodes, unsigned long __user *nmask, unsigned long maxnode, int mode) { unsigned long k; @@ -136,7 +123,7 @@ static int get_nodes(unsigned long *nodes, unsigned long __user *nmask, unsigned long endmask; --maxnode; - bitmap_zero(nodes, MAX_NUMNODES); + nodes_clear(*nodes); if (maxnode == 0 || !nmask) return 0; @@ -153,7 +140,7 @@ static int get_nodes(unsigned long *nodes, unsigned long __user *nmask, return -EINVAL; for (k = BITS_TO_LONGS(MAX_NUMNODES); k < nlongs; k++) { unsigned long t; - if (get_user(t, nmask + k)) + if (get_user(t, nmask + k)) return -EFAULT; if (k == nlongs - 1) { if (t & endmask) @@ -165,30 +152,29 @@ static int get_nodes(unsigned long *nodes, unsigned long __user *nmask, endmask = ~0UL; } - if (copy_from_user(nodes, nmask, nlongs*sizeof(unsigned long))) + if (copy_from_user(nodes_addr(*nodes), nmask, nlongs*sizeof(unsigned long))) return -EFAULT; - nodes[nlongs-1] &= endmask; + nodes_addr(*nodes)[nlongs-1] &= endmask; /* Update current mems_allowed */ cpuset_update_current_mems_allowed(); /* Ignore nodes not set in current->mems_allowed */ - cpuset_restrict_to_mems_allowed(nodes); + /* AK: shouldn't this error out instead? */ + cpuset_restrict_to_mems_allowed(nodes_addr(*nodes)); return mpol_check_policy(mode, nodes); } /* Generate a custom zonelist for the BIND policy. */ -static struct zonelist *bind_zonelist(unsigned long *nodes) +static struct zonelist *bind_zonelist(nodemask_t *nodes) { struct zonelist *zl; int num, max, nd; - max = 1 + MAX_NR_ZONES * bitmap_weight(nodes, MAX_NUMNODES); + max = 1 + MAX_NR_ZONES * nodes_weight(*nodes); zl = kmalloc(sizeof(void *) * max, GFP_KERNEL); if (!zl) return NULL; num = 0; - for (nd = find_first_bit(nodes, MAX_NUMNODES); - nd < MAX_NUMNODES; - nd = find_next_bit(nodes, MAX_NUMNODES, 1+nd)) { + for_each_node_mask(nd, *nodes) { int k; for (k = MAX_NR_ZONES-1; k >= 0; k--) { struct zone *z = &NODE_DATA(nd)->node_zones[k]; @@ -205,11 +191,11 @@ static struct zonelist *bind_zonelist(unsigned long *nodes) } /* Create a new policy */ -static struct mempolicy *mpol_new(int mode, unsigned long *nodes) +static struct mempolicy *mpol_new(int mode, nodemask_t *nodes) { struct mempolicy *policy; - PDprintk("setting mode %d nodes[0] %lx\n", mode, nodes[0]); + PDprintk("setting mode %d nodes[0] %lx\n", mode, nodes_addr(*nodes)[0]); if (mode == MPOL_DEFAULT) return NULL; policy = kmem_cache_alloc(policy_cache, GFP_KERNEL); @@ -218,10 +204,10 @@ static struct mempolicy *mpol_new(int mode, unsigned long *nodes) atomic_set(&policy->refcnt, 1); switch (mode) { case MPOL_INTERLEAVE: - bitmap_copy(policy->v.nodes, nodes, MAX_NUMNODES); + policy->v.nodes = *nodes; break; case MPOL_PREFERRED: - policy->v.preferred_node = find_first_bit(nodes, MAX_NUMNODES); + policy->v.preferred_node = first_node(*nodes); if (policy->v.preferred_node >= MAX_NUMNODES) policy->v.preferred_node = -1; break; @@ -239,7 +225,7 @@ static struct mempolicy *mpol_new(int mode, unsigned long *nodes) /* Ensure all existing pages follow the policy. */ static int check_pte_range(struct mm_struct *mm, pmd_t *pmd, - unsigned long addr, unsigned long end, unsigned long *nodes) + unsigned long addr, unsigned long end, nodemask_t *nodes) { pte_t *orig_pte; pte_t *pte; @@ -256,7 +242,7 @@ static int check_pte_range(struct mm_struct *mm, pmd_t *pmd, if (!pfn_valid(pfn)) continue; nid = pfn_to_nid(pfn); - if (!test_bit(nid, nodes)) + if (!node_isset(nid, *nodes)) break; } while (pte++, addr += PAGE_SIZE, addr != end); pte_unmap(orig_pte); @@ -265,7 +251,7 @@ static int check_pte_range(struct mm_struct *mm, pmd_t *pmd, } static inline int check_pmd_range(struct mm_struct *mm, pud_t *pud, - unsigned long addr, unsigned long end, unsigned long *nodes) + unsigned long addr, unsigned long end, nodemask_t *nodes) { pmd_t *pmd; unsigned long next; @@ -282,7 +268,7 @@ static inline int check_pmd_range(struct mm_struct *mm, pud_t *pud, } static inline int check_pud_range(struct mm_struct *mm, pgd_t *pgd, - unsigned long addr, unsigned long end, unsigned long *nodes) + unsigned long addr, unsigned long end, nodemask_t *nodes) { pud_t *pud; unsigned long next; @@ -299,7 +285,7 @@ static inline int check_pud_range(struct mm_struct *mm, pgd_t *pgd, } static inline int check_pgd_range(struct mm_struct *mm, - unsigned long addr, unsigned long end, unsigned long *nodes) + unsigned long addr, unsigned long end, nodemask_t *nodes) { pgd_t *pgd; unsigned long next; @@ -318,7 +304,7 @@ static inline int check_pgd_range(struct mm_struct *mm, /* Step 1: check the range */ static struct vm_area_struct * check_range(struct mm_struct *mm, unsigned long start, unsigned long end, - unsigned long *nodes, unsigned long flags) + nodemask_t *nodes, unsigned long flags) { int err; struct vm_area_struct *first, *vma, *prev; @@ -403,7 +389,7 @@ asmlinkage long sys_mbind(unsigned long start, unsigned long len, struct mm_struct *mm = current->mm; struct mempolicy *new; unsigned long end; - DECLARE_BITMAP(nodes, MAX_NUMNODES); + nodemask_t nodes; int err; if ((flags & ~(unsigned long)(MPOL_MF_STRICT)) || mode > MPOL_MAX) @@ -419,19 +405,19 @@ asmlinkage long sys_mbind(unsigned long start, unsigned long len, if (end == start) return 0; - err = get_nodes(nodes, nmask, maxnode, mode); + err = get_nodes(&nodes, nmask, maxnode, mode); if (err) return err; - new = mpol_new(mode, nodes); + new = mpol_new(mode, &nodes); if (IS_ERR(new)) return PTR_ERR(new); PDprintk("mbind %lx-%lx mode:%ld nodes:%lx\n",start,start+len, - mode,nodes[0]); + mode,nodes_addr(nodes)[0]); down_write(&mm->mmap_sem); - vma = check_range(mm, start, end, nodes, flags); + vma = check_range(mm, start, end, &nodes, flags); err = PTR_ERR(vma); if (!IS_ERR(vma)) err = mbind_range(vma, start, end, new); @@ -446,45 +432,45 @@ asmlinkage long sys_set_mempolicy(int mode, unsigned long __user *nmask, { int err; struct mempolicy *new; - DECLARE_BITMAP(nodes, MAX_NUMNODES); + nodemask_t nodes; if (mode < 0 || mode > MPOL_MAX) return -EINVAL; - err = get_nodes(nodes, nmask, maxnode, mode); + err = get_nodes(&nodes, nmask, maxnode, mode); if (err) return err; - new = mpol_new(mode, nodes); + new = mpol_new(mode, &nodes); if (IS_ERR(new)) return PTR_ERR(new); mpol_free(current->mempolicy); current->mempolicy = new; if (new && new->policy == MPOL_INTERLEAVE) - current->il_next = find_first_bit(new->v.nodes, MAX_NUMNODES); + current->il_next = first_node(new->v.nodes); return 0; } /* Fill a zone bitmap for a policy */ -static void get_zonemask(struct mempolicy *p, unsigned long *nodes) +static void get_zonemask(struct mempolicy *p, nodemask_t *nodes) { int i; - bitmap_zero(nodes, MAX_NUMNODES); + nodes_clear(*nodes); switch (p->policy) { case MPOL_BIND: for (i = 0; p->v.zonelist->zones[i]; i++) - __set_bit(p->v.zonelist->zones[i]->zone_pgdat->node_id, nodes); + node_set(p->v.zonelist->zones[i]->zone_pgdat->node_id, *nodes); break; case MPOL_DEFAULT: break; case MPOL_INTERLEAVE: - bitmap_copy(nodes, p->v.nodes, MAX_NUMNODES); + *nodes = p->v.nodes; break; case MPOL_PREFERRED: /* or use current node instead of online map? */ if (p->v.preferred_node < 0) - bitmap_copy(nodes, nodes_addr(node_online_map), MAX_NUMNODES); + *nodes = node_online_map; else - __set_bit(p->v.preferred_node, nodes); + node_set(p->v.preferred_node, *nodes); break; default: BUG(); @@ -506,9 +492,10 @@ static int lookup_node(struct mm_struct *mm, unsigned long addr) /* Copy a kernel node mask to user space */ static int copy_nodes_to_user(unsigned long __user *mask, unsigned long maxnode, - void *nodes, unsigned nbytes) + nodemask_t *nodes) { unsigned long copy = ALIGN(maxnode-1, 64) / 8; + const int nbytes = BITS_TO_LONGS(MAX_NUMNODES) * sizeof(long); if (copy > nbytes) { if (copy > PAGE_SIZE) @@ -517,7 +504,7 @@ static int copy_nodes_to_user(unsigned long __user *mask, unsigned long maxnode, return -EFAULT; copy = nbytes; } - return copy_to_user(mask, nodes, copy) ? -EFAULT : 0; + return copy_to_user(mask, nodes_addr(*nodes), copy) ? -EFAULT : 0; } /* Retrieve NUMA policy */ @@ -578,9 +565,9 @@ asmlinkage long sys_get_mempolicy(int __user *policy, err = 0; if (nmask) { - DECLARE_BITMAP(nodes, MAX_NUMNODES); - get_zonemask(pol, nodes); - err = copy_nodes_to_user(nmask, maxnode, nodes, sizeof(nodes)); + nodemask_t nodes; + get_zonemask(pol, &nodes); + err = copy_nodes_to_user(nmask, maxnode, &nodes); } out: @@ -649,15 +636,15 @@ asmlinkage long compat_sys_mbind(compat_ulong_t start, compat_ulong_t len, long err = 0; unsigned long __user *nm = NULL; unsigned long nr_bits, alloc_size; - DECLARE_BITMAP(bm, MAX_NUMNODES); + nodemask_t bm; nr_bits = min_t(unsigned long, maxnode-1, MAX_NUMNODES); alloc_size = ALIGN(nr_bits, BITS_PER_LONG) / 8; if (nmask) { - err = compat_get_bitmap(bm, nmask, nr_bits); + err = compat_get_bitmap(nodes_addr(bm), nmask, nr_bits); nm = compat_alloc_user_space(alloc_size); - err |= copy_to_user(nm, bm, alloc_size); + err |= copy_to_user(nm, nodes_addr(bm), alloc_size); } if (err) @@ -723,9 +710,9 @@ static unsigned interleave_nodes(struct mempolicy *policy) nid = me->il_next; BUG_ON(nid >= MAX_NUMNODES); - next = find_next_bit(policy->v.nodes, MAX_NUMNODES, 1+nid); + next = next_node(nid, policy->v.nodes); if (next >= MAX_NUMNODES) - next = find_first_bit(policy->v.nodes, MAX_NUMNODES); + next = first_node(policy->v.nodes); me->il_next = next; return nid; } @@ -734,18 +721,17 @@ static unsigned interleave_nodes(struct mempolicy *policy) static unsigned offset_il_node(struct mempolicy *pol, struct vm_area_struct *vma, unsigned long off) { - unsigned nnodes = bitmap_weight(pol->v.nodes, MAX_NUMNODES); + unsigned nnodes = nodes_weight(pol->v.nodes); unsigned target = (unsigned)off % nnodes; int c; int nid = -1; c = 0; do { - nid = find_next_bit(pol->v.nodes, MAX_NUMNODES, nid+1); + nid = next_node(nid, pol->v.nodes); c++; } while (c <= target); BUG_ON(nid >= MAX_NUMNODES); - BUG_ON(!test_bit(nid, pol->v.nodes)); return nid; } @@ -878,7 +864,7 @@ int __mpol_equal(struct mempolicy *a, struct mempolicy *b) case MPOL_DEFAULT: return 1; case MPOL_INTERLEAVE: - return bitmap_equal(a->v.nodes, b->v.nodes, MAX_NUMNODES); + return nodes_equal(a->v.nodes, b->v.nodes); case MPOL_PREFERRED: return a->v.preferred_node == b->v.preferred_node; case MPOL_BIND: { @@ -1117,7 +1103,7 @@ int mpol_set_shared_policy(struct shared_policy *info, PDprintk("set_shared_policy %lx sz %lu %d %lx\n", vma->vm_pgoff, sz, npol? npol->policy : -1, - npol ? npol->v.nodes[0] : -1); + npol ? nodes_addr(npol->v.nodes)[0] : -1); if (npol) { new = sp_alloc(vma->vm_pgoff, vma->vm_pgoff + sz, npol); -- cgit v1.2.3 From ab50b8ed818016cfecd747d6d4bb9139986bc029 Mon Sep 17 00:00:00 2001 From: Hugh Dickins Date: Sat, 29 Oct 2005 18:15:56 -0700 Subject: [PATCH] mm: vm_stat_account unshackled The original vm_stat_account has fallen into disuse, with only one user, and only one user of vm_stat_unaccount. It's easier to keep track if we convert them all to __vm_stat_account, then free it from its __shackles. Signed-off-by: Hugh Dickins Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/ia64/kernel/perfmon.c | 3 ++- arch/ia64/mm/fault.c | 2 +- include/linux/mm.h | 16 ++-------------- kernel/fork.c | 2 +- mm/mmap.c | 20 ++++++++++---------- mm/mprotect.c | 4 ++-- mm/mremap.c | 4 ++-- 7 files changed, 20 insertions(+), 31 deletions(-) (limited to 'include/linux') diff --git a/arch/ia64/kernel/perfmon.c b/arch/ia64/kernel/perfmon.c index d71731ee5b61..f7dfc107cb7b 100644 --- a/arch/ia64/kernel/perfmon.c +++ b/arch/ia64/kernel/perfmon.c @@ -2352,7 +2352,8 @@ pfm_smpl_buffer_alloc(struct task_struct *task, pfm_context_t *ctx, unsigned lon insert_vm_struct(mm, vma); mm->total_vm += size >> PAGE_SHIFT; - vm_stat_account(vma); + vm_stat_account(vma->vm_mm, vma->vm_flags, vma->vm_file, + vma_pages(vma)); up_write(&task->mm->mmap_sem); /* diff --git a/arch/ia64/mm/fault.c b/arch/ia64/mm/fault.c index 3c32af910d60..f21b55549787 100644 --- a/arch/ia64/mm/fault.c +++ b/arch/ia64/mm/fault.c @@ -41,7 +41,7 @@ expand_backing_store (struct vm_area_struct *vma, unsigned long address) vma->vm_mm->total_vm += grow; if (vma->vm_flags & VM_LOCKED) vma->vm_mm->locked_vm += grow; - __vm_stat_account(vma->vm_mm, vma->vm_flags, vma->vm_file, grow); + vm_stat_account(vma->vm_mm, vma->vm_flags, vma->vm_file, grow); return 0; } diff --git a/include/linux/mm.h b/include/linux/mm.h index e1649578fb0c..376a466743bc 100644 --- a/include/linux/mm.h +++ b/include/linux/mm.h @@ -928,26 +928,14 @@ int remap_pfn_range(struct vm_area_struct *, unsigned long, unsigned long, unsigned long, pgprot_t); #ifdef CONFIG_PROC_FS -void __vm_stat_account(struct mm_struct *, unsigned long, struct file *, long); +void vm_stat_account(struct mm_struct *, unsigned long, struct file *, long); #else -static inline void __vm_stat_account(struct mm_struct *mm, +static inline void vm_stat_account(struct mm_struct *mm, unsigned long flags, struct file *file, long pages) { } #endif /* CONFIG_PROC_FS */ -static inline void vm_stat_account(struct vm_area_struct *vma) -{ - __vm_stat_account(vma->vm_mm, vma->vm_flags, vma->vm_file, - vma_pages(vma)); -} - -static inline void vm_stat_unaccount(struct vm_area_struct *vma) -{ - __vm_stat_account(vma->vm_mm, vma->vm_flags, vma->vm_file, - -vma_pages(vma)); -} - /* update per process rss and vm hiwater data */ extern void update_mem_hiwater(struct task_struct *tsk); diff --git a/kernel/fork.c b/kernel/fork.c index 280bd44ac441..e2ff11f8c1b0 100644 --- a/kernel/fork.c +++ b/kernel/fork.c @@ -212,7 +212,7 @@ static inline int dup_mmap(struct mm_struct * mm, struct mm_struct * oldmm) if (mpnt->vm_flags & VM_DONTCOPY) { long pages = vma_pages(mpnt); mm->total_vm -= pages; - __vm_stat_account(mm, mpnt->vm_flags, mpnt->vm_file, + vm_stat_account(mm, mpnt->vm_flags, mpnt->vm_file, -pages); continue; } diff --git a/mm/mmap.c b/mm/mmap.c index fa11d91242e8..e1780266ac7d 100644 --- a/mm/mmap.c +++ b/mm/mmap.c @@ -832,7 +832,7 @@ none: } #ifdef CONFIG_PROC_FS -void __vm_stat_account(struct mm_struct *mm, unsigned long flags, +void vm_stat_account(struct mm_struct *mm, unsigned long flags, struct file *file, long pages) { const unsigned long stack_flags @@ -1110,7 +1110,7 @@ munmap_back: } out: mm->total_vm += len >> PAGE_SHIFT; - __vm_stat_account(mm, vm_flags, file, len >> PAGE_SHIFT); + vm_stat_account(mm, vm_flags, file, len >> PAGE_SHIFT); if (vm_flags & VM_LOCKED) { mm->locked_vm += len >> PAGE_SHIFT; make_pages_present(addr, addr + len); @@ -1475,7 +1475,7 @@ static int acct_stack_growth(struct vm_area_struct * vma, unsigned long size, un mm->total_vm += grow; if (vma->vm_flags & VM_LOCKED) mm->locked_vm += grow; - __vm_stat_account(mm, vma->vm_flags, vma->vm_file, grow); + vm_stat_account(mm, vma->vm_flags, vma->vm_file, grow); return 0; } @@ -1610,15 +1610,15 @@ find_extend_vma(struct mm_struct * mm, unsigned long addr) * By the time this function is called, the area struct has been * removed from the process mapping list. */ -static void unmap_vma(struct mm_struct *mm, struct vm_area_struct *area) +static void unmap_vma(struct mm_struct *mm, struct vm_area_struct *vma) { - size_t len = area->vm_end - area->vm_start; + long nrpages = vma_pages(vma); - area->vm_mm->total_vm -= len >> PAGE_SHIFT; - if (area->vm_flags & VM_LOCKED) - area->vm_mm->locked_vm -= len >> PAGE_SHIFT; - vm_stat_unaccount(area); - remove_vm_struct(area); + mm->total_vm -= nrpages; + if (vma->vm_flags & VM_LOCKED) + mm->locked_vm -= nrpages; + vm_stat_account(mm, vma->vm_flags, vma->vm_file, -nrpages); + remove_vm_struct(vma); } /* diff --git a/mm/mprotect.c b/mm/mprotect.c index 57577f63b305..b426f01c5e9c 100644 --- a/mm/mprotect.c +++ b/mm/mprotect.c @@ -168,8 +168,8 @@ success: vma->vm_flags = newflags; vma->vm_page_prot = newprot; change_protection(vma, start, end, newprot); - __vm_stat_account(mm, oldflags, vma->vm_file, -nrpages); - __vm_stat_account(mm, newflags, vma->vm_file, nrpages); + vm_stat_account(mm, oldflags, vma->vm_file, -nrpages); + vm_stat_account(mm, newflags, vma->vm_file, nrpages); return 0; fail: diff --git a/mm/mremap.c b/mm/mremap.c index f343fc73a8bd..55df8f53e84d 100644 --- a/mm/mremap.c +++ b/mm/mremap.c @@ -233,7 +233,7 @@ static unsigned long move_vma(struct vm_area_struct *vma, * since do_munmap() will decrement it by old_len == new_len */ mm->total_vm += new_len >> PAGE_SHIFT; - __vm_stat_account(mm, vma->vm_flags, vma->vm_file, new_len>>PAGE_SHIFT); + vm_stat_account(mm, vma->vm_flags, vma->vm_file, new_len>>PAGE_SHIFT); if (do_munmap(mm, old_addr, old_len) < 0) { /* OOM: unable to split vma, just get accounts right */ @@ -384,7 +384,7 @@ unsigned long do_mremap(unsigned long addr, addr + new_len, vma->vm_pgoff, NULL); current->mm->total_vm += pages; - __vm_stat_account(vma->vm_mm, vma->vm_flags, + vm_stat_account(vma->vm_mm, vma->vm_flags, vma->vm_file, pages); if (vma->vm_flags & VM_LOCKED) { current->mm->locked_vm += pages; -- cgit v1.2.3 From a8fb5618dab7e45c8990f3155628d772a9ed45f9 Mon Sep 17 00:00:00 2001 From: Hugh Dickins Date: Sat, 29 Oct 2005 18:15:57 -0700 Subject: [PATCH] mm: unlink_file_vma, remove_vma Divide remove_vm_struct into two parts: first anon_vma_unlink plus unlink_file_vma, to unlink the vma from the list and tree by which rmap or vmtruncate might find it; then remove_vma to close, fput and free. The intention here is to do the anon_vma_unlink and unlink_file_vma earlier, in free_pgtables before freeing any page tables: so we can be sure that any page tables traversed by rmap and vmtruncate are stable (and other, ordinary cases are stabilized by holding mmap_sem). This will be crucial to traversing pgd,pud,pmd without page_table_lock. But testing the split-out patch showed that lifting the page_table_lock is symbiotically necessary to make this change - the lock ordering is wrong to move those unlinks into free_pgtables while it's under ptlock. Signed-off-by: Hugh Dickins Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/mm.h | 1 + mm/mmap.c | 41 +++++++++++++++++++++++++++-------------- 2 files changed, 28 insertions(+), 14 deletions(-) (limited to 'include/linux') diff --git a/include/linux/mm.h b/include/linux/mm.h index 376a466743bc..0c64484d8ae0 100644 --- a/include/linux/mm.h +++ b/include/linux/mm.h @@ -834,6 +834,7 @@ extern int split_vma(struct mm_struct *, extern int insert_vm_struct(struct mm_struct *, struct vm_area_struct *); extern void __vma_link_rb(struct mm_struct *, struct vm_area_struct *, struct rb_node **, struct rb_node *); +extern void unlink_file_vma(struct vm_area_struct *); extern struct vm_area_struct *copy_vma(struct vm_area_struct **, unsigned long addr, unsigned long len, pgoff_t pgoff); extern void exit_mmap(struct mm_struct *); diff --git a/mm/mmap.c b/mm/mmap.c index eeefe19a0fac..a3984fad3fc2 100644 --- a/mm/mmap.c +++ b/mm/mmap.c @@ -181,26 +181,44 @@ static void __remove_shared_vm_struct(struct vm_area_struct *vma, } /* - * Remove one vm structure and free it. + * Unlink a file-based vm structure from its prio_tree, to hide + * vma from rmap and vmtruncate before freeing its page tables. */ -static void remove_vm_struct(struct vm_area_struct *vma) +void unlink_file_vma(struct vm_area_struct *vma) { struct file *file = vma->vm_file; - might_sleep(); if (file) { struct address_space *mapping = file->f_mapping; spin_lock(&mapping->i_mmap_lock); __remove_shared_vm_struct(vma, file, mapping); spin_unlock(&mapping->i_mmap_lock); } +} + +/* + * Close a vm structure and free it, returning the next. + */ +static struct vm_area_struct *remove_vma(struct vm_area_struct *vma) +{ + struct vm_area_struct *next = vma->vm_next; + + /* + * Hide vma from rmap and vmtruncate before freeing page tables: + * to be moved into free_pgtables once page_table_lock is lifted + * from it, but until then lock ordering forbids that move. + */ + anon_vma_unlink(vma); + unlink_file_vma(vma); + + might_sleep(); if (vma->vm_ops && vma->vm_ops->close) vma->vm_ops->close(vma); - if (file) - fput(file); - anon_vma_unlink(vma); + if (vma->vm_file) + fput(vma->vm_file); mpol_free(vma_policy(vma)); kmem_cache_free(vm_area_cachep, vma); + return next; } asmlinkage unsigned long sys_brk(unsigned long brk) @@ -1612,15 +1630,13 @@ find_extend_vma(struct mm_struct * mm, unsigned long addr) static void remove_vma_list(struct mm_struct *mm, struct vm_area_struct *vma) { do { - struct vm_area_struct *next = vma->vm_next; long nrpages = vma_pages(vma); mm->total_vm -= nrpages; if (vma->vm_flags & VM_LOCKED) mm->locked_vm -= nrpages; vm_stat_account(mm, vma->vm_flags, vma->vm_file, -nrpages); - remove_vm_struct(vma); - vma = next; + vma = remove_vma(vma); } while (vma); validate_mm(mm); } @@ -1944,11 +1960,8 @@ void exit_mmap(struct mm_struct *mm) * Walk the list again, actually closing and freeing it * without holding any MM locks. */ - while (vma) { - struct vm_area_struct *next = vma->vm_next; - remove_vm_struct(vma); - vma = next; - } + while (vma) + vma = remove_vma(vma); BUG_ON(mm->nr_ptes > (FIRST_USER_ADDRESS+PMD_SIZE-1)>>PMD_SHIFT); } -- cgit v1.2.3 From 4294621f41a85497019fae64341aa5351a1921b7 Mon Sep 17 00:00:00 2001 From: Hugh Dickins Date: Sat, 29 Oct 2005 18:16:05 -0700 Subject: [PATCH] mm: rss = file_rss + anon_rss I was lazy when we added anon_rss, and chose to change as few places as possible. So currently each anonymous page has to be counted twice, in rss and in anon_rss. Which won't be so good if those are atomic counts in some configurations. Change that around: keep file_rss and anon_rss separately, and add them together (with get_mm_rss macro) when the total is needed - reading two atomics is much cheaper than updating two atomics. And update anon_rss upfront, typically in memory.c, not tucked away in page_add_anon_rmap. Signed-off-by: Hugh Dickins Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/exec.c | 2 +- fs/proc/array.c | 2 +- fs/proc/task_mmu.c | 8 +++----- include/linux/sched.h | 4 +++- kernel/acct.c | 2 +- kernel/fork.c | 4 ++-- mm/fremap.c | 4 ++-- mm/hugetlb.c | 6 +++--- mm/memory.c | 31 +++++++++++++++++-------------- mm/nommu.c | 2 +- mm/rmap.c | 8 +++----- mm/swapfile.c | 2 +- 12 files changed, 38 insertions(+), 37 deletions(-) (limited to 'include/linux') diff --git a/fs/exec.c b/fs/exec.c index d2208f7c87db..cefadf5ab83b 100644 --- a/fs/exec.c +++ b/fs/exec.c @@ -330,7 +330,7 @@ void install_arg_page(struct vm_area_struct *vma, pte_unmap(pte); goto out; } - inc_mm_counter(mm, rss); + inc_mm_counter(mm, anon_rss); lru_cache_add_active(page); set_pte_at(mm, address, pte, pte_mkdirty(pte_mkwrite(mk_pte( page, vma->vm_page_prot)))); diff --git a/fs/proc/array.c b/fs/proc/array.c index d84eecacbeaf..3e1239e4b303 100644 --- a/fs/proc/array.c +++ b/fs/proc/array.c @@ -438,7 +438,7 @@ static int do_task_stat(struct task_struct *task, char * buffer, int whole) jiffies_to_clock_t(it_real_value), start_time, vsize, - mm ? get_mm_counter(mm, rss) : 0, /* you might want to shift this left 3 */ + mm ? get_mm_rss(mm) : 0, rsslim, mm ? mm->start_code : 0, mm ? mm->end_code : 0, diff --git a/fs/proc/task_mmu.c b/fs/proc/task_mmu.c index 994612bc72d0..bccee7cf9ccd 100644 --- a/fs/proc/task_mmu.c +++ b/fs/proc/task_mmu.c @@ -29,7 +29,7 @@ char *task_mem(struct mm_struct *mm, char *buffer) "VmPTE:\t%8lu kB\n", (mm->total_vm - mm->reserved_vm) << (PAGE_SHIFT-10), mm->locked_vm << (PAGE_SHIFT-10), - get_mm_counter(mm, rss) << (PAGE_SHIFT-10), + get_mm_rss(mm) << (PAGE_SHIFT-10), data << (PAGE_SHIFT-10), mm->stack_vm << (PAGE_SHIFT-10), text, lib, (PTRS_PER_PTE*sizeof(pte_t)*mm->nr_ptes) >> 10); @@ -44,13 +44,11 @@ unsigned long task_vsize(struct mm_struct *mm) int task_statm(struct mm_struct *mm, int *shared, int *text, int *data, int *resident) { - int rss = get_mm_counter(mm, rss); - - *shared = rss - get_mm_counter(mm, anon_rss); + *shared = get_mm_counter(mm, file_rss); *text = (PAGE_ALIGN(mm->end_code) - (mm->start_code & PAGE_MASK)) >> PAGE_SHIFT; *data = mm->total_vm - mm->shared_vm; - *resident = rss; + *resident = *shared + get_mm_counter(mm, anon_rss); return mm->total_vm; } diff --git a/include/linux/sched.h b/include/linux/sched.h index 27519df0f987..afcaac66cbd5 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -254,6 +254,8 @@ extern void arch_unmap_area_topdown(struct mm_struct *, unsigned long); #define add_mm_counter(mm, member, value) (mm)->_##member += (value) #define inc_mm_counter(mm, member) (mm)->_##member++ #define dec_mm_counter(mm, member) (mm)->_##member-- +#define get_mm_rss(mm) ((mm)->_file_rss + (mm)->_anon_rss) + typedef unsigned long mm_counter_t; struct mm_struct { @@ -286,7 +288,7 @@ struct mm_struct { unsigned long exec_vm, stack_vm, reserved_vm, def_flags, nr_ptes; /* Special counters protected by the page_table_lock */ - mm_counter_t _rss; + mm_counter_t _file_rss; mm_counter_t _anon_rss; unsigned long saved_auxv[AT_VECTOR_SIZE]; /* for /proc/PID/auxv */ diff --git a/kernel/acct.c b/kernel/acct.c index b756f527497e..2e3f4a47e7d0 100644 --- a/kernel/acct.c +++ b/kernel/acct.c @@ -553,7 +553,7 @@ void acct_update_integrals(struct task_struct *tsk) if (delta == 0) return; tsk->acct_stimexpd = tsk->stime; - tsk->acct_rss_mem1 += delta * get_mm_counter(tsk->mm, rss); + tsk->acct_rss_mem1 += delta * get_mm_rss(tsk->mm); tsk->acct_vm_mem1 += delta * tsk->mm->total_vm; } } diff --git a/kernel/fork.c b/kernel/fork.c index 25caa02e2eac..2048ed7b5872 100644 --- a/kernel/fork.c +++ b/kernel/fork.c @@ -321,7 +321,7 @@ static struct mm_struct * mm_init(struct mm_struct * mm) INIT_LIST_HEAD(&mm->mmlist); mm->core_waiters = 0; mm->nr_ptes = 0; - set_mm_counter(mm, rss, 0); + set_mm_counter(mm, file_rss, 0); set_mm_counter(mm, anon_rss, 0); spin_lock_init(&mm->page_table_lock); rwlock_init(&mm->ioctx_list_lock); @@ -499,7 +499,7 @@ static int copy_mm(unsigned long clone_flags, struct task_struct * tsk) if (retval) goto free_pt; - mm->hiwater_rss = get_mm_counter(mm,rss); + mm->hiwater_rss = get_mm_rss(mm); mm->hiwater_vm = mm->total_vm; good_mm: diff --git a/mm/fremap.c b/mm/fremap.c index ab23a0673c35..fd7f2a17ff3e 100644 --- a/mm/fremap.c +++ b/mm/fremap.c @@ -39,7 +39,7 @@ static inline void zap_pte(struct mm_struct *mm, struct vm_area_struct *vma, set_page_dirty(page); page_remove_rmap(page); page_cache_release(page); - dec_mm_counter(mm, rss); + dec_mm_counter(mm, file_rss); } } } else { @@ -95,7 +95,7 @@ int install_page(struct mm_struct *mm, struct vm_area_struct *vma, zap_pte(mm, vma, addr, pte); - inc_mm_counter(mm,rss); + inc_mm_counter(mm, file_rss); flush_icache_page(vma, page); set_pte_at(mm, addr, pte, mk_pte(page, prot)); page_add_file_rmap(page); diff --git a/mm/hugetlb.c b/mm/hugetlb.c index 61d380678030..094455bcbbf7 100644 --- a/mm/hugetlb.c +++ b/mm/hugetlb.c @@ -286,7 +286,7 @@ int copy_hugetlb_page_range(struct mm_struct *dst, struct mm_struct *src, entry = *src_pte; ptepage = pte_page(entry); get_page(ptepage); - add_mm_counter(dst, rss, HPAGE_SIZE / PAGE_SIZE); + add_mm_counter(dst, file_rss, HPAGE_SIZE / PAGE_SIZE); set_huge_pte_at(dst, addr, dst_pte, entry); } spin_unlock(&src->page_table_lock); @@ -324,7 +324,7 @@ void unmap_hugepage_range(struct vm_area_struct *vma, unsigned long start, page = pte_page(pte); put_page(page); - add_mm_counter(mm, rss, - (HPAGE_SIZE / PAGE_SIZE)); + add_mm_counter(mm, file_rss, (int) -(HPAGE_SIZE / PAGE_SIZE)); } flush_tlb_range(vma, start, end); } @@ -386,7 +386,7 @@ int hugetlb_prefault(struct address_space *mapping, struct vm_area_struct *vma) goto out; } } - add_mm_counter(mm, rss, HPAGE_SIZE / PAGE_SIZE); + add_mm_counter(mm, file_rss, HPAGE_SIZE / PAGE_SIZE); set_huge_pte_at(mm, addr, pte, make_huge_pte(vma, page)); } out: diff --git a/mm/memory.c b/mm/memory.c index 51eb38574830..59d42e50fa53 100644 --- a/mm/memory.c +++ b/mm/memory.c @@ -397,9 +397,10 @@ copy_one_pte(struct mm_struct *dst_mm, struct mm_struct *src_mm, pte = pte_mkclean(pte); pte = pte_mkold(pte); get_page(page); - inc_mm_counter(dst_mm, rss); if (PageAnon(page)) inc_mm_counter(dst_mm, anon_rss); + else + inc_mm_counter(dst_mm, file_rss); set_pte_at(dst_mm, addr, dst_pte, pte); page_dup_rmap(page); } @@ -581,8 +582,8 @@ static void zap_pte_range(struct mmu_gather *tlb, pmd_t *pmd, set_page_dirty(page); if (pte_young(ptent)) mark_page_accessed(page); + dec_mm_counter(tlb->mm, file_rss); } - dec_mm_counter(tlb->mm, rss); page_remove_rmap(page); tlb_remove_page(tlb, page); continue; @@ -1290,13 +1291,15 @@ static int do_wp_page(struct mm_struct *mm, struct vm_area_struct *vma, spin_lock(&mm->page_table_lock); page_table = pte_offset_map(pmd, address); if (likely(pte_same(*page_table, orig_pte))) { - if (PageAnon(old_page)) - dec_mm_counter(mm, anon_rss); if (PageReserved(old_page)) - inc_mm_counter(mm, rss); - else + inc_mm_counter(mm, anon_rss); + else { page_remove_rmap(old_page); - + if (!PageAnon(old_page)) { + inc_mm_counter(mm, anon_rss); + dec_mm_counter(mm, file_rss); + } + } flush_cache_page(vma, address, pfn); entry = mk_pte(new_page, vma->vm_page_prot); entry = maybe_mkwrite(pte_mkdirty(entry), vma); @@ -1701,7 +1704,7 @@ static int do_swap_page(struct mm_struct *mm, struct vm_area_struct *vma, /* The page isn't present yet, go ahead with the fault. */ - inc_mm_counter(mm, rss); + inc_mm_counter(mm, anon_rss); pte = mk_pte(page, vma->vm_page_prot); if (write_access && can_share_swap_page(page)) { pte = maybe_mkwrite(pte_mkdirty(pte), vma); @@ -1774,7 +1777,7 @@ static int do_anonymous_page(struct mm_struct *mm, struct vm_area_struct *vma, page_cache_release(page); goto unlock; } - inc_mm_counter(mm, rss); + inc_mm_counter(mm, anon_rss); entry = mk_pte(page, vma->vm_page_prot); entry = maybe_mkwrite(pte_mkdirty(entry), vma); lru_cache_add_active(page); @@ -1887,19 +1890,19 @@ retry: */ /* Only go through if we didn't race with anybody else... */ if (pte_none(*page_table)) { - if (!PageReserved(new_page)) - inc_mm_counter(mm, rss); - flush_icache_page(vma, new_page); entry = mk_pte(new_page, vma->vm_page_prot); if (write_access) entry = maybe_mkwrite(pte_mkdirty(entry), vma); set_pte_at(mm, address, page_table, entry); if (anon) { + inc_mm_counter(mm, anon_rss); lru_cache_add_active(new_page); page_add_anon_rmap(new_page, vma, address); - } else + } else if (!PageReserved(new_page)) { + inc_mm_counter(mm, file_rss); page_add_file_rmap(new_page); + } } else { /* One of our sibling threads was faster, back out. */ page_cache_release(new_page); @@ -2192,7 +2195,7 @@ EXPORT_SYMBOL(vmalloc_to_pfn); void update_mem_hiwater(struct task_struct *tsk) { if (tsk->mm) { - unsigned long rss = get_mm_counter(tsk->mm, rss); + unsigned long rss = get_mm_rss(tsk->mm); if (tsk->mm->hiwater_rss < rss) tsk->mm->hiwater_rss = rss; diff --git a/mm/nommu.c b/mm/nommu.c index 0ef241ae3763..599924886eb5 100644 --- a/mm/nommu.c +++ b/mm/nommu.c @@ -1083,7 +1083,7 @@ void update_mem_hiwater(struct task_struct *tsk) unsigned long rss; if (likely(tsk->mm)) { - rss = get_mm_counter(tsk->mm, rss); + rss = get_mm_rss(tsk->mm); if (tsk->mm->hiwater_rss < rss) tsk->mm->hiwater_rss = rss; if (tsk->mm->hiwater_vm < tsk->mm->total_vm) diff --git a/mm/rmap.c b/mm/rmap.c index 1fc559e09ca8..504757624cce 100644 --- a/mm/rmap.c +++ b/mm/rmap.c @@ -445,8 +445,6 @@ void page_add_anon_rmap(struct page *page, { BUG_ON(PageReserved(page)); - inc_mm_counter(vma->vm_mm, anon_rss); - if (atomic_inc_and_test(&page->_mapcount)) { struct anon_vma *anon_vma = vma->anon_vma; @@ -561,9 +559,9 @@ static int try_to_unmap_one(struct page *page, struct vm_area_struct *vma) set_pte_at(mm, address, pte, swp_entry_to_pte(entry)); BUG_ON(pte_file(*pte)); dec_mm_counter(mm, anon_rss); - } + } else + dec_mm_counter(mm, file_rss); - dec_mm_counter(mm, rss); page_remove_rmap(page); page_cache_release(page); @@ -667,7 +665,7 @@ static void try_to_unmap_cluster(unsigned long cursor, page_remove_rmap(page); page_cache_release(page); - dec_mm_counter(mm, rss); + dec_mm_counter(mm, file_rss); (*mapcount)--; } diff --git a/mm/swapfile.c b/mm/swapfile.c index 05c851291241..296e0bbf7836 100644 --- a/mm/swapfile.c +++ b/mm/swapfile.c @@ -407,7 +407,7 @@ void free_swap_and_cache(swp_entry_t entry) static void unuse_pte(struct vm_area_struct *vma, pte_t *pte, unsigned long addr, swp_entry_t entry, struct page *page) { - inc_mm_counter(vma->vm_mm, rss); + inc_mm_counter(vma->vm_mm, anon_rss); get_page(page); set_pte_at(vma->vm_mm, addr, pte, pte_mkold(mk_pte(page, vma->vm_page_prot))); -- cgit v1.2.3 From b5810039a54e5babf428e9a1e89fc1940fabff11 Mon Sep 17 00:00:00 2001 From: Nick Piggin Date: Sat, 29 Oct 2005 18:16:12 -0700 Subject: [PATCH] core remove PageReserved Remove PageReserved() calls from core code by tightening VM_RESERVED handling in mm/ to cover PageReserved functionality. PageReserved special casing is removed from get_page and put_page. All setting and clearing of PageReserved is retained, and it is now flagged in the page_alloc checks to help ensure we don't introduce any refcount based freeing of Reserved pages. MAP_PRIVATE, PROT_WRITE of VM_RESERVED regions is tentatively being deprecated. We never completely handled it correctly anyway, and is be reintroduced in future if required (Hugh has a proof of concept). Once PageReserved() calls are removed from kernel/power/swsusp.c, and all arch/ and driver code, the Set and Clear calls, and the PG_reserved bit can be trivially removed. Last real user of PageReserved is swsusp, which uses PageReserved to determine whether a struct page points to valid memory or not. This still needs to be addressed (a generic page_is_ram() should work). A last caveat: the ZERO_PAGE is now refcounted and managed with rmap (and thus mapcounted and count towards shared rss). These writes to the struct page could cause excessive cacheline bouncing on big systems. There are a number of ways this could be addressed if it is an issue. Signed-off-by: Nick Piggin Refcount bug fix for filemap_xip.c Signed-off-by: Carsten Otte Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/ppc64/kernel/vdso.c | 12 +++-- arch/sparc/mm/generic.c | 3 ++ arch/sparc64/mm/generic.c | 3 ++ drivers/scsi/sg.c | 12 +++-- drivers/scsi/st.c | 10 ++-- fs/direct-io.c | 4 +- include/linux/mm.h | 5 +- kernel/power/swsusp.c | 25 +++++---- mm/bootmem.c | 1 + mm/filemap_xip.c | 11 ++-- mm/fremap.c | 23 ++++---- mm/madvise.c | 2 +- mm/memory.c | 131 ++++++++++++++++++++++++++++------------------ mm/mempolicy.c | 29 +++++----- mm/mmap.c | 11 ++++ mm/mprotect.c | 8 +++ mm/msync.c | 17 +++--- mm/page_alloc.c | 14 ++--- mm/rmap.c | 14 ++--- mm/shmem.c | 4 +- mm/swap.c | 4 +- sound/core/pcm_native.c | 9 ++-- 22 files changed, 218 insertions(+), 134 deletions(-) (limited to 'include/linux') diff --git a/arch/ppc64/kernel/vdso.c b/arch/ppc64/kernel/vdso.c index efa985f05aca..4aacf521e3e4 100644 --- a/arch/ppc64/kernel/vdso.c +++ b/arch/ppc64/kernel/vdso.c @@ -176,13 +176,13 @@ static struct page * vdso_vma_nopage(struct vm_area_struct * vma, return NOPAGE_SIGBUS; /* - * Last page is systemcfg, special handling here, no get_page() a - * this is a reserved page + * Last page is systemcfg. */ if ((vma->vm_end - address) <= PAGE_SIZE) - return virt_to_page(systemcfg); + pg = virt_to_page(systemcfg); + else + pg = virt_to_page(vbase + offset); - pg = virt_to_page(vbase + offset); get_page(pg); DBG(" ->page count: %d\n", page_count(pg)); @@ -259,7 +259,7 @@ int arch_setup_additional_pages(struct linux_binprm *bprm, int executable_stack) * gettimeofday will be totally dead. It's fine to use that for setting * breakpoints in the vDSO code pages though */ - vma->vm_flags = VM_READ | VM_EXEC | VM_MAYREAD | VM_MAYWRITE | VM_MAYEXEC; + vma->vm_flags = VM_READ | VM_EXEC | VM_MAYREAD | VM_MAYWRITE | VM_MAYEXEC | VM_RESERVED; vma->vm_flags |= mm->def_flags; vma->vm_page_prot = protection_map[vma->vm_flags & 0x7]; vma->vm_ops = &vdso_vmops; @@ -603,6 +603,8 @@ void __init vdso_init(void) ClearPageReserved(pg); get_page(pg); } + + get_page(virt_to_page(systemcfg)); } int in_gate_area_no_task(unsigned long addr) diff --git a/arch/sparc/mm/generic.c b/arch/sparc/mm/generic.c index 20ccb957fb77..659c9a71f867 100644 --- a/arch/sparc/mm/generic.c +++ b/arch/sparc/mm/generic.c @@ -73,6 +73,9 @@ int io_remap_pfn_range(struct vm_area_struct *vma, unsigned long from, int space = GET_IOSPACE(pfn); unsigned long offset = GET_PFN(pfn) << PAGE_SHIFT; + /* See comment in mm/memory.c remap_pfn_range */ + vma->vm_flags |= VM_IO | VM_RESERVED; + prot = __pgprot(pg_iobits); offset -= from; dir = pgd_offset(mm, from); diff --git a/arch/sparc64/mm/generic.c b/arch/sparc64/mm/generic.c index c954d91f01d0..afc01cec701f 100644 --- a/arch/sparc64/mm/generic.c +++ b/arch/sparc64/mm/generic.c @@ -127,6 +127,9 @@ int io_remap_pfn_range(struct vm_area_struct *vma, unsigned long from, int space = GET_IOSPACE(pfn); unsigned long offset = GET_PFN(pfn) << PAGE_SHIFT; + /* See comment in mm/memory.c remap_pfn_range */ + vma->vm_flags |= VM_IO | VM_RESERVED; + prot = __pgprot(pg_iobits); offset -= from; dir = pgd_offset(mm, from); diff --git a/drivers/scsi/sg.c b/drivers/scsi/sg.c index 861e51375d70..2d30b46806bf 100644 --- a/drivers/scsi/sg.c +++ b/drivers/scsi/sg.c @@ -1886,13 +1886,17 @@ st_unmap_user_pages(struct scatterlist *sgl, const unsigned int nr_pages, int i; for (i=0; i < nr_pages; i++) { - if (dirtied && !PageReserved(sgl[i].page)) - SetPageDirty(sgl[i].page); - /* unlock_page(sgl[i].page); */ + struct page *page = sgl[i].page; + + /* XXX: just for debug. Remove when PageReserved is removed */ + BUG_ON(PageReserved(page)); + if (dirtied) + SetPageDirty(page); + /* unlock_page(page); */ /* FIXME: cache flush missing for rw==READ * FIXME: call the correct reference counting function */ - page_cache_release(sgl[i].page); + page_cache_release(page); } return 0; diff --git a/drivers/scsi/st.c b/drivers/scsi/st.c index 5eb54d8019b4..da9766283bd7 100644 --- a/drivers/scsi/st.c +++ b/drivers/scsi/st.c @@ -4526,12 +4526,16 @@ static int sgl_unmap_user_pages(struct scatterlist *sgl, const unsigned int nr_p int i; for (i=0; i < nr_pages; i++) { - if (dirtied && !PageReserved(sgl[i].page)) - SetPageDirty(sgl[i].page); + struct page *page = sgl[i].page; + + /* XXX: just for debug. Remove when PageReserved is removed */ + BUG_ON(PageReserved(page)); + if (dirtied) + SetPageDirty(page); /* FIXME: cache flush missing for rw==READ * FIXME: call the correct reference counting function */ - page_cache_release(sgl[i].page); + page_cache_release(page); } return 0; diff --git a/fs/direct-io.c b/fs/direct-io.c index 0d06097bc995..3931e7f1e6bf 100644 --- a/fs/direct-io.c +++ b/fs/direct-io.c @@ -162,6 +162,7 @@ static int dio_refill_pages(struct dio *dio) up_read(¤t->mm->mmap_sem); if (ret < 0 && dio->blocks_available && (dio->rw == WRITE)) { + struct page *page = ZERO_PAGE(dio->curr_user_address); /* * A memory fault, but the filesystem has some outstanding * mapped blocks. We need to use those blocks up to avoid @@ -169,7 +170,8 @@ static int dio_refill_pages(struct dio *dio) */ if (dio->page_errors == 0) dio->page_errors = ret; - dio->pages[0] = ZERO_PAGE(dio->curr_user_address); + page_cache_get(page); + dio->pages[0] = page; dio->head = 0; dio->tail = 1; ret = 0; diff --git a/include/linux/mm.h b/include/linux/mm.h index 0c64484d8ae0..da42093250c3 100644 --- a/include/linux/mm.h +++ b/include/linux/mm.h @@ -157,7 +157,7 @@ extern unsigned int kobjsize(const void *objp); #define VM_DONTCOPY 0x00020000 /* Do not copy this vma on fork */ #define VM_DONTEXPAND 0x00040000 /* Cannot expand with mremap() */ -#define VM_RESERVED 0x00080000 /* Don't unmap it from swap_out */ +#define VM_RESERVED 0x00080000 /* Pages managed in a special way */ #define VM_ACCOUNT 0x00100000 /* Is a VM accounted object */ #define VM_HUGETLB 0x00400000 /* Huge TLB Page VM */ #define VM_NONLINEAR 0x00800000 /* Is non-linear (remap_file_pages) */ @@ -338,7 +338,7 @@ static inline void get_page(struct page *page) static inline void put_page(struct page *page) { - if (!PageReserved(page) && put_page_testzero(page)) + if (put_page_testzero(page)) __page_cache_release(page); } @@ -723,6 +723,7 @@ void install_arg_page(struct vm_area_struct *, struct page *, unsigned long); int get_user_pages(struct task_struct *tsk, struct mm_struct *mm, unsigned long start, int len, int write, int force, struct page **pages, struct vm_area_struct **vmas); +void print_bad_pte(struct vm_area_struct *, pte_t, unsigned long); int __set_page_dirty_buffers(struct page *page); int __set_page_dirty_nobuffers(struct page *page); diff --git a/kernel/power/swsusp.c b/kernel/power/swsusp.c index 10bc5ec496d7..016504ccfccf 100644 --- a/kernel/power/swsusp.c +++ b/kernel/power/swsusp.c @@ -578,15 +578,23 @@ static int save_highmem_zone(struct zone *zone) continue; page = pfn_to_page(pfn); /* - * This condition results from rvmalloc() sans vmalloc_32() - * and architectural memory reservations. This should be - * corrected eventually when the cases giving rise to this - * are better understood. + * PageReserved results from rvmalloc() sans vmalloc_32() + * and architectural memory reservations. + * + * rvmalloc should not cause this, because all implementations + * appear to always be using vmalloc_32 on architectures with + * highmem. This is a good thing, because we would like to save + * rvmalloc pages. + * + * It appears to be triggered by pages which do not point to + * valid memory (see arch/i386/mm/init.c:one_highpage_init(), + * which sets PageReserved if the page does not point to valid + * RAM. + * + * XXX: must remove usage of PageReserved! */ - if (PageReserved(page)) { - printk("highmem reserved page?!\n"); + if (PageReserved(page)) continue; - } BUG_ON(PageNosave(page)); if (PageNosaveFree(page)) continue; @@ -672,10 +680,9 @@ static int saveable(struct zone * zone, unsigned long * zone_pfn) return 0; page = pfn_to_page(pfn); - BUG_ON(PageReserved(page) && PageNosave(page)); if (PageNosave(page)) return 0; - if (PageReserved(page) && pfn_is_nosave(pfn)) { + if (pfn_is_nosave(pfn)) { pr_debug("[nosave pfn 0x%lx]", pfn); return 0; } diff --git a/mm/bootmem.c b/mm/bootmem.c index a58699b6579e..e8c567177dcf 100644 --- a/mm/bootmem.c +++ b/mm/bootmem.c @@ -305,6 +305,7 @@ static unsigned long __init free_all_bootmem_core(pg_data_t *pgdat) if (j + 16 < BITS_PER_LONG) prefetchw(page + j + 16); __ClearPageReserved(page + j); + set_page_count(page + j, 0); } __free_pages(page, order); i += BITS_PER_LONG; diff --git a/mm/filemap_xip.c b/mm/filemap_xip.c index 8c199f537732..9354ee279b13 100644 --- a/mm/filemap_xip.c +++ b/mm/filemap_xip.c @@ -174,6 +174,7 @@ __xip_unmap (struct address_space * mapping, unsigned long address; pte_t *pte; pte_t pteval; + struct page *page = ZERO_PAGE(address); spin_lock(&mapping->i_mmap_lock); vma_prio_tree_foreach(vma, &iter, &mapping->i_mmap, pgoff, pgoff) { @@ -185,15 +186,17 @@ __xip_unmap (struct address_space * mapping, * We need the page_table_lock to protect us from page faults, * munmap, fork, etc... */ - pte = page_check_address(ZERO_PAGE(address), mm, - address); + pte = page_check_address(page, mm, address); if (!IS_ERR(pte)) { /* Nuke the page table entry. */ flush_cache_page(vma, address, pte_pfn(*pte)); pteval = ptep_clear_flush(vma, address, pte); + page_remove_rmap(page); + dec_mm_counter(mm, file_rss); BUG_ON(pte_dirty(pteval)); pte_unmap(pte); spin_unlock(&mm->page_table_lock); + page_cache_release(page); } } spin_unlock(&mapping->i_mmap_lock); @@ -228,7 +231,7 @@ xip_file_nopage(struct vm_area_struct * area, page = mapping->a_ops->get_xip_page(mapping, pgoff*(PAGE_SIZE/512), 0); if (!IS_ERR(page)) { - return page; + goto out; } if (PTR_ERR(page) != -ENODATA) return NULL; @@ -249,6 +252,8 @@ xip_file_nopage(struct vm_area_struct * area, page = ZERO_PAGE(address); } +out: + page_cache_get(page); return page; } diff --git a/mm/fremap.c b/mm/fremap.c index fd7f2a17ff3e..224cc1598b35 100644 --- a/mm/fremap.c +++ b/mm/fremap.c @@ -29,19 +29,20 @@ static inline void zap_pte(struct mm_struct *mm, struct vm_area_struct *vma, return; if (pte_present(pte)) { unsigned long pfn = pte_pfn(pte); + struct page *page; flush_cache_page(vma, addr, pfn); pte = ptep_clear_flush(vma, addr, ptep); - if (pfn_valid(pfn)) { - struct page *page = pfn_to_page(pfn); - if (!PageReserved(page)) { - if (pte_dirty(pte)) - set_page_dirty(page); - page_remove_rmap(page); - page_cache_release(page); - dec_mm_counter(mm, file_rss); - } + if (unlikely(!pfn_valid(pfn))) { + print_bad_pte(vma, pte, addr); + return; } + page = pfn_to_page(pfn); + if (pte_dirty(pte)) + set_page_dirty(page); + page_remove_rmap(page); + page_cache_release(page); + dec_mm_counter(mm, file_rss); } else { if (!pte_file(pte)) free_swap_and_cache(pte_to_swp_entry(pte)); @@ -65,6 +66,8 @@ int install_page(struct mm_struct *mm, struct vm_area_struct *vma, pgd_t *pgd; pte_t pte_val; + BUG_ON(vma->vm_flags & VM_RESERVED); + pgd = pgd_offset(mm, addr); spin_lock(&mm->page_table_lock); @@ -125,6 +128,8 @@ int install_file_pte(struct mm_struct *mm, struct vm_area_struct *vma, pgd_t *pgd; pte_t pte_val; + BUG_ON(vma->vm_flags & VM_RESERVED); + pgd = pgd_offset(mm, addr); spin_lock(&mm->page_table_lock); diff --git a/mm/madvise.c b/mm/madvise.c index 20e075d1c64c..17aaf3e16449 100644 --- a/mm/madvise.c +++ b/mm/madvise.c @@ -126,7 +126,7 @@ static long madvise_dontneed(struct vm_area_struct * vma, unsigned long start, unsigned long end) { *prev = vma; - if ((vma->vm_flags & VM_LOCKED) || is_vm_hugetlb_page(vma)) + if (vma->vm_flags & (VM_LOCKED|VM_HUGETLB|VM_RESERVED)) return -EINVAL; if (unlikely(vma->vm_flags & VM_NONLINEAR)) { diff --git a/mm/memory.c b/mm/memory.c index da642b5528fa..e83f9440bb66 100644 --- a/mm/memory.c +++ b/mm/memory.c @@ -342,6 +342,23 @@ static inline void add_mm_rss(struct mm_struct *mm, int file_rss, int anon_rss) #define NO_RSS 2 /* Increment neither file_rss nor anon_rss */ +/* + * This function is called to print an error when a pte in a + * !VM_RESERVED region is found pointing to an invalid pfn (which + * is an error. + * + * The calling function must still handle the error. + */ +void print_bad_pte(struct vm_area_struct *vma, pte_t pte, unsigned long vaddr) +{ + printk(KERN_ERR "Bad pte = %08llx, process = %s, " + "vm_flags = %lx, vaddr = %lx\n", + (long long)pte_val(pte), + (vma->vm_mm == current->mm ? current->comm : "???"), + vma->vm_flags, vaddr); + dump_stack(); +} + /* * copy one vm_area from one task to the other. Assumes the page tables * already present in the new task to be cleared in the whole range @@ -353,9 +370,10 @@ static inline void add_mm_rss(struct mm_struct *mm, int file_rss, int anon_rss) static inline int copy_one_pte(struct mm_struct *dst_mm, struct mm_struct *src_mm, - pte_t *dst_pte, pte_t *src_pte, unsigned long vm_flags, + pte_t *dst_pte, pte_t *src_pte, struct vm_area_struct *vma, unsigned long addr) { + unsigned long vm_flags = vma->vm_flags; pte_t pte = *src_pte; struct page *page; unsigned long pfn; @@ -375,18 +393,22 @@ copy_one_pte(struct mm_struct *dst_mm, struct mm_struct *src_mm, goto out_set_pte; } + /* If the region is VM_RESERVED, the mapping is not + * mapped via rmap - duplicate the pte as is. + */ + if (vm_flags & VM_RESERVED) + goto out_set_pte; + pfn = pte_pfn(pte); - /* the pte points outside of valid memory, the - * mapping is assumed to be good, meaningful - * and not mapped via rmap - duplicate the - * mapping as is. + /* If the pte points outside of valid memory but + * the region is not VM_RESERVED, we have a problem. */ - page = NULL; - if (pfn_valid(pfn)) - page = pfn_to_page(pfn); + if (unlikely(!pfn_valid(pfn))) { + print_bad_pte(vma, pte, addr); + goto out_set_pte; /* try to do something sane */ + } - if (!page || PageReserved(page)) - goto out_set_pte; + page = pfn_to_page(pfn); /* * If it's a COW mapping, write protect it both @@ -418,7 +440,6 @@ static int copy_pte_range(struct mm_struct *dst_mm, struct mm_struct *src_mm, unsigned long addr, unsigned long end) { pte_t *src_pte, *dst_pte; - unsigned long vm_flags = vma->vm_flags; int progress = 0; int rss[NO_RSS+1], anon; @@ -446,8 +467,7 @@ again: progress++; continue; } - anon = copy_one_pte(dst_mm, src_mm, dst_pte, src_pte, - vm_flags, addr); + anon = copy_one_pte(dst_mm, src_mm, dst_pte, src_pte, vma,addr); rss[anon]++; progress += 8; } while (dst_pte++, src_pte++, addr += PAGE_SIZE, addr != end); @@ -541,10 +561,12 @@ int copy_page_range(struct mm_struct *dst_mm, struct mm_struct *src_mm, return 0; } -static void zap_pte_range(struct mmu_gather *tlb, pmd_t *pmd, +static void zap_pte_range(struct mmu_gather *tlb, + struct vm_area_struct *vma, pmd_t *pmd, unsigned long addr, unsigned long end, struct zap_details *details) { + struct mm_struct *mm = tlb->mm; pte_t *pte; int file_rss = 0; int anon_rss = 0; @@ -556,11 +578,12 @@ static void zap_pte_range(struct mmu_gather *tlb, pmd_t *pmd, continue; if (pte_present(ptent)) { struct page *page = NULL; - unsigned long pfn = pte_pfn(ptent); - if (pfn_valid(pfn)) { - page = pfn_to_page(pfn); - if (PageReserved(page)) - page = NULL; + if (!(vma->vm_flags & VM_RESERVED)) { + unsigned long pfn = pte_pfn(ptent); + if (unlikely(!pfn_valid(pfn))) + print_bad_pte(vma, ptent, addr); + else + page = pfn_to_page(pfn); } if (unlikely(details) && page) { /* @@ -580,7 +603,7 @@ static void zap_pte_range(struct mmu_gather *tlb, pmd_t *pmd, page->index > details->last_index)) continue; } - ptent = ptep_get_and_clear_full(tlb->mm, addr, pte, + ptent = ptep_get_and_clear_full(mm, addr, pte, tlb->fullmm); tlb_remove_tlb_entry(tlb, pte, addr); if (unlikely(!page)) @@ -588,7 +611,7 @@ static void zap_pte_range(struct mmu_gather *tlb, pmd_t *pmd, if (unlikely(details) && details->nonlinear_vma && linear_page_index(details->nonlinear_vma, addr) != page->index) - set_pte_at(tlb->mm, addr, pte, + set_pte_at(mm, addr, pte, pgoff_to_pte(page->index)); if (PageAnon(page)) anon_rss++; @@ -611,14 +634,15 @@ static void zap_pte_range(struct mmu_gather *tlb, pmd_t *pmd, continue; if (!pte_file(ptent)) free_swap_and_cache(pte_to_swp_entry(ptent)); - pte_clear_full(tlb->mm, addr, pte, tlb->fullmm); + pte_clear_full(mm, addr, pte, tlb->fullmm); } while (pte++, addr += PAGE_SIZE, addr != end); - add_mm_rss(tlb->mm, -file_rss, -anon_rss); + add_mm_rss(mm, -file_rss, -anon_rss); pte_unmap(pte - 1); } -static inline void zap_pmd_range(struct mmu_gather *tlb, pud_t *pud, +static inline void zap_pmd_range(struct mmu_gather *tlb, + struct vm_area_struct *vma, pud_t *pud, unsigned long addr, unsigned long end, struct zap_details *details) { @@ -630,11 +654,12 @@ static inline void zap_pmd_range(struct mmu_gather *tlb, pud_t *pud, next = pmd_addr_end(addr, end); if (pmd_none_or_clear_bad(pmd)) continue; - zap_pte_range(tlb, pmd, addr, next, details); + zap_pte_range(tlb, vma, pmd, addr, next, details); } while (pmd++, addr = next, addr != end); } -static inline void zap_pud_range(struct mmu_gather *tlb, pgd_t *pgd, +static inline void zap_pud_range(struct mmu_gather *tlb, + struct vm_area_struct *vma, pgd_t *pgd, unsigned long addr, unsigned long end, struct zap_details *details) { @@ -646,7 +671,7 @@ static inline void zap_pud_range(struct mmu_gather *tlb, pgd_t *pgd, next = pud_addr_end(addr, end); if (pud_none_or_clear_bad(pud)) continue; - zap_pmd_range(tlb, pud, addr, next, details); + zap_pmd_range(tlb, vma, pud, addr, next, details); } while (pud++, addr = next, addr != end); } @@ -667,7 +692,7 @@ static void unmap_page_range(struct mmu_gather *tlb, struct vm_area_struct *vma, next = pgd_addr_end(addr, end); if (pgd_none_or_clear_bad(pgd)) continue; - zap_pud_range(tlb, pgd, addr, next, details); + zap_pud_range(tlb, vma, pgd, addr, next, details); } while (pgd++, addr = next, addr != end); tlb_end_vma(tlb, vma); } @@ -967,7 +992,7 @@ int get_user_pages(struct task_struct *tsk, struct mm_struct *mm, continue; } - if (!vma || (vma->vm_flags & VM_IO) + if (!vma || (vma->vm_flags & (VM_IO | VM_RESERVED)) || !(flags & vma->vm_flags)) return i ? : -EFAULT; @@ -1027,8 +1052,7 @@ int get_user_pages(struct task_struct *tsk, struct mm_struct *mm, if (pages) { pages[i] = page; flush_dcache_page(page); - if (!PageReserved(page)) - page_cache_get(page); + page_cache_get(page); } if (vmas) vmas[i] = vma; @@ -1051,7 +1075,11 @@ static int zeromap_pte_range(struct mm_struct *mm, pmd_t *pmd, if (!pte) return -ENOMEM; do { - pte_t zero_pte = pte_wrprotect(mk_pte(ZERO_PAGE(addr), prot)); + struct page *page = ZERO_PAGE(addr); + pte_t zero_pte = pte_wrprotect(mk_pte(page, prot)); + page_cache_get(page); + page_add_file_rmap(page); + inc_mm_counter(mm, file_rss); BUG_ON(!pte_none(*pte)); set_pte_at(mm, addr, pte, zero_pte); } while (pte++, addr += PAGE_SIZE, addr != end); @@ -1132,8 +1160,7 @@ static int remap_pte_range(struct mm_struct *mm, pmd_t *pmd, return -ENOMEM; do { BUG_ON(!pte_none(*pte)); - if (!pfn_valid(pfn) || PageReserved(pfn_to_page(pfn))) - set_pte_at(mm, addr, pte, pfn_pte(pfn, prot)); + set_pte_at(mm, addr, pte, pfn_pte(pfn, prot)); pfn++; } while (pte++, addr += PAGE_SIZE, addr != end); pte_unmap(pte - 1); @@ -1195,8 +1222,8 @@ int remap_pfn_range(struct vm_area_struct *vma, unsigned long addr, * rest of the world about it: * VM_IO tells people not to look at these pages * (accesses can have side effects). - * VM_RESERVED tells swapout not to try to touch - * this region. + * VM_RESERVED tells the core MM not to "manage" these pages + * (e.g. refcount, mapcount, try to swap them out). */ vma->vm_flags |= VM_IO | VM_RESERVED; @@ -1256,11 +1283,13 @@ static int do_wp_page(struct mm_struct *mm, struct vm_area_struct *vma, pte_t entry; int ret = VM_FAULT_MINOR; + BUG_ON(vma->vm_flags & VM_RESERVED); + if (unlikely(!pfn_valid(pfn))) { /* * Page table corrupted: show pte and kill process. */ - pte_ERROR(orig_pte); + print_bad_pte(vma, orig_pte, address); ret = VM_FAULT_OOM; goto unlock; } @@ -1284,8 +1313,7 @@ static int do_wp_page(struct mm_struct *mm, struct vm_area_struct *vma, /* * Ok, we need to copy. Oh, well.. */ - if (!PageReserved(old_page)) - page_cache_get(old_page); + page_cache_get(old_page); pte_unmap(page_table); spin_unlock(&mm->page_table_lock); @@ -1308,14 +1336,10 @@ static int do_wp_page(struct mm_struct *mm, struct vm_area_struct *vma, spin_lock(&mm->page_table_lock); page_table = pte_offset_map(pmd, address); if (likely(pte_same(*page_table, orig_pte))) { - if (PageReserved(old_page)) + page_remove_rmap(old_page); + if (!PageAnon(old_page)) { inc_mm_counter(mm, anon_rss); - else { - page_remove_rmap(old_page); - if (!PageAnon(old_page)) { - inc_mm_counter(mm, anon_rss); - dec_mm_counter(mm, file_rss); - } + dec_mm_counter(mm, file_rss); } flush_cache_page(vma, address, pfn); entry = mk_pte(new_page, vma->vm_page_prot); @@ -1769,14 +1793,13 @@ static int do_anonymous_page(struct mm_struct *mm, struct vm_area_struct *vma, unsigned long address, pte_t *page_table, pmd_t *pmd, int write_access) { + struct page *page = ZERO_PAGE(addr); pte_t entry; /* Mapping of ZERO_PAGE - vm_page_prot is readonly */ - entry = mk_pte(ZERO_PAGE(addr), vma->vm_page_prot); + entry = mk_pte(page, vma->vm_page_prot); if (write_access) { - struct page *page; - /* Allocate our own private page. */ pte_unmap(page_table); spin_unlock(&mm->page_table_lock); @@ -1800,6 +1823,10 @@ static int do_anonymous_page(struct mm_struct *mm, struct vm_area_struct *vma, lru_cache_add_active(page); SetPageReferenced(page); page_add_anon_rmap(page, vma, address); + } else { + inc_mm_counter(mm, file_rss); + page_add_file_rmap(page); + page_cache_get(page); } set_pte_at(mm, address, page_table, entry); @@ -1916,7 +1943,7 @@ retry: inc_mm_counter(mm, anon_rss); lru_cache_add_active(new_page); page_add_anon_rmap(new_page, vma, address); - } else if (!PageReserved(new_page)) { + } else if (!(vma->vm_flags & VM_RESERVED)) { inc_mm_counter(mm, file_rss); page_add_file_rmap(new_page); } @@ -1957,7 +1984,7 @@ static int do_file_page(struct mm_struct *mm, struct vm_area_struct *vma, /* * Page table corrupted: show pte and kill process. */ - pte_ERROR(orig_pte); + print_bad_pte(vma, orig_pte, address); return VM_FAULT_OOM; } /* We can then assume vm->vm_ops && vma->vm_ops->populate */ @@ -2232,7 +2259,7 @@ static int __init gate_vma_init(void) gate_vma.vm_start = FIXADDR_USER_START; gate_vma.vm_end = FIXADDR_USER_END; gate_vma.vm_page_prot = PAGE_READONLY; - gate_vma.vm_flags = 0; + gate_vma.vm_flags = VM_RESERVED; return 0; } __initcall(gate_vma_init); diff --git a/mm/mempolicy.c b/mm/mempolicy.c index 43b1199af591..11d824f282f1 100644 --- a/mm/mempolicy.c +++ b/mm/mempolicy.c @@ -223,13 +223,13 @@ static struct mempolicy *mpol_new(int mode, nodemask_t *nodes) } /* Ensure all existing pages follow the policy. */ -static int check_pte_range(struct mm_struct *mm, pmd_t *pmd, +static int check_pte_range(struct vm_area_struct *vma, pmd_t *pmd, unsigned long addr, unsigned long end, nodemask_t *nodes) { pte_t *orig_pte; pte_t *pte; - spin_lock(&mm->page_table_lock); + spin_lock(&vma->vm_mm->page_table_lock); orig_pte = pte = pte_offset_map(pmd, addr); do { unsigned long pfn; @@ -238,18 +238,20 @@ static int check_pte_range(struct mm_struct *mm, pmd_t *pmd, if (!pte_present(*pte)) continue; pfn = pte_pfn(*pte); - if (!pfn_valid(pfn)) + if (!pfn_valid(pfn)) { + print_bad_pte(vma, *pte, addr); continue; + } nid = pfn_to_nid(pfn); if (!node_isset(nid, *nodes)) break; } while (pte++, addr += PAGE_SIZE, addr != end); pte_unmap(orig_pte); - spin_unlock(&mm->page_table_lock); + spin_unlock(&vma->vm_mm->page_table_lock); return addr != end; } -static inline int check_pmd_range(struct mm_struct *mm, pud_t *pud, +static inline int check_pmd_range(struct vm_area_struct *vma, pud_t *pud, unsigned long addr, unsigned long end, nodemask_t *nodes) { pmd_t *pmd; @@ -260,13 +262,13 @@ static inline int check_pmd_range(struct mm_struct *mm, pud_t *pud, next = pmd_addr_end(addr, end); if (pmd_none_or_clear_bad(pmd)) continue; - if (check_pte_range(mm, pmd, addr, next, nodes)) + if (check_pte_range(vma, pmd, addr, next, nodes)) return -EIO; } while (pmd++, addr = next, addr != end); return 0; } -static inline int check_pud_range(struct mm_struct *mm, pgd_t *pgd, +static inline int check_pud_range(struct vm_area_struct *vma, pgd_t *pgd, unsigned long addr, unsigned long end, nodemask_t *nodes) { pud_t *pud; @@ -277,24 +279,24 @@ static inline int check_pud_range(struct mm_struct *mm, pgd_t *pgd, next = pud_addr_end(addr, end); if (pud_none_or_clear_bad(pud)) continue; - if (check_pmd_range(mm, pud, addr, next, nodes)) + if (check_pmd_range(vma, pud, addr, next, nodes)) return -EIO; } while (pud++, addr = next, addr != end); return 0; } -static inline int check_pgd_range(struct mm_struct *mm, +static inline int check_pgd_range(struct vm_area_struct *vma, unsigned long addr, unsigned long end, nodemask_t *nodes) { pgd_t *pgd; unsigned long next; - pgd = pgd_offset(mm, addr); + pgd = pgd_offset(vma->vm_mm, addr); do { next = pgd_addr_end(addr, end); if (pgd_none_or_clear_bad(pgd)) continue; - if (check_pud_range(mm, pgd, addr, next, nodes)) + if (check_pud_range(vma, pgd, addr, next, nodes)) return -EIO; } while (pgd++, addr = next, addr != end); return 0; @@ -311,6 +313,8 @@ check_range(struct mm_struct *mm, unsigned long start, unsigned long end, first = find_vma(mm, start); if (!first) return ERR_PTR(-EFAULT); + if (first->vm_flags & VM_RESERVED) + return ERR_PTR(-EACCES); prev = NULL; for (vma = first; vma && vma->vm_start < end; vma = vma->vm_next) { if (!vma->vm_next && vma->vm_end < end) @@ -323,8 +327,7 @@ check_range(struct mm_struct *mm, unsigned long start, unsigned long end, endvma = end; if (vma->vm_start > start) start = vma->vm_start; - err = check_pgd_range(vma->vm_mm, - start, endvma, nodes); + err = check_pgd_range(vma, start, endvma, nodes); if (err) { first = ERR_PTR(err); break; diff --git a/mm/mmap.c b/mm/mmap.c index 459b9f068ad7..8a111792b8db 100644 --- a/mm/mmap.c +++ b/mm/mmap.c @@ -1088,6 +1088,17 @@ munmap_back: error = file->f_op->mmap(file, vma); if (error) goto unmap_and_free_vma; + if ((vma->vm_flags & (VM_SHARED | VM_WRITE | VM_RESERVED)) + == (VM_WRITE | VM_RESERVED)) { + printk(KERN_WARNING "program %s is using MAP_PRIVATE, " + "PROT_WRITE mmap of VM_RESERVED memory, which " + "is deprecated. Please report this to " + "linux-kernel@vger.kernel.org\n",current->comm); + if (vma->vm_ops && vma->vm_ops->close) + vma->vm_ops->close(vma); + error = -EACCES; + goto unmap_and_free_vma; + } } else if (vm_flags & VM_SHARED) { error = shmem_zero_setup(vma); if (error) diff --git a/mm/mprotect.c b/mm/mprotect.c index b426f01c5e9c..672a76fddd5e 100644 --- a/mm/mprotect.c +++ b/mm/mprotect.c @@ -125,6 +125,14 @@ mprotect_fixup(struct vm_area_struct *vma, struct vm_area_struct **pprev, * a MAP_NORESERVE private mapping to writable will now reserve. */ if (newflags & VM_WRITE) { + if (oldflags & VM_RESERVED) { + BUG_ON(oldflags & VM_WRITE); + printk(KERN_WARNING "program %s is using MAP_PRIVATE, " + "PROT_WRITE mprotect of VM_RESERVED memory, " + "which is deprecated. Please report this to " + "linux-kernel@vger.kernel.org\n",current->comm); + return -EACCES; + } if (!(oldflags & (VM_ACCOUNT|VM_WRITE|VM_SHARED|VM_HUGETLB))) { charged = nrpages; if (security_vm_enough_memory(charged)) diff --git a/mm/msync.c b/mm/msync.c index 3b5f1c521d4b..860395486060 100644 --- a/mm/msync.c +++ b/mm/msync.c @@ -25,6 +25,7 @@ static void msync_pte_range(struct vm_area_struct *vma, pmd_t *pmd, unsigned long addr, unsigned long end) { + struct mm_struct *mm = vma->vm_mm; pte_t *pte; int progress = 0; @@ -37,7 +38,7 @@ again: if (progress >= 64) { progress = 0; if (need_resched() || - need_lockbreak(&vma->vm_mm->page_table_lock)) + need_lockbreak(&mm->page_table_lock)) break; } progress++; @@ -46,11 +47,11 @@ again: if (!pte_maybe_dirty(*pte)) continue; pfn = pte_pfn(*pte); - if (!pfn_valid(pfn)) + if (unlikely(!pfn_valid(pfn))) { + print_bad_pte(vma, *pte, addr); continue; + } page = pfn_to_page(pfn); - if (PageReserved(page)) - continue; if (ptep_clear_flush_dirty(vma, addr, pte) || page_test_and_clear_dirty(page)) @@ -58,7 +59,7 @@ again: progress += 3; } while (pte++, addr += PAGE_SIZE, addr != end); pte_unmap(pte - 1); - cond_resched_lock(&vma->vm_mm->page_table_lock); + cond_resched_lock(&mm->page_table_lock); if (addr != end) goto again; } @@ -102,8 +103,10 @@ static void msync_page_range(struct vm_area_struct *vma, /* For hugepages we can't go walking the page table normally, * but that's ok, hugetlbfs is memory based, so we don't need - * to do anything more on an msync() */ - if (is_vm_hugetlb_page(vma)) + * to do anything more on an msync(). + * Can't do anything with VM_RESERVED regions either. + */ + if (vma->vm_flags & (VM_HUGETLB|VM_RESERVED)) return; BUG_ON(addr >= end); diff --git a/mm/page_alloc.c b/mm/page_alloc.c index 60663232fbb2..0541288ebf4b 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -114,7 +114,8 @@ static void bad_page(const char *function, struct page *page) 1 << PG_reclaim | 1 << PG_slab | 1 << PG_swapcache | - 1 << PG_writeback); + 1 << PG_writeback | + 1 << PG_reserved ); set_page_count(page, 0); reset_page_mapcount(page); page->mapping = NULL; @@ -244,7 +245,6 @@ static inline int page_is_buddy(struct page *page, int order) { if (PagePrivate(page) && (page_order(page) == order) && - !PageReserved(page) && page_count(page) == 0) return 1; return 0; @@ -327,7 +327,8 @@ static inline void free_pages_check(const char *function, struct page *page) 1 << PG_reclaim | 1 << PG_slab | 1 << PG_swapcache | - 1 << PG_writeback ))) + 1 << PG_writeback | + 1 << PG_reserved ))) bad_page(function, page); if (PageDirty(page)) __ClearPageDirty(page); @@ -455,7 +456,8 @@ static void prep_new_page(struct page *page, int order) 1 << PG_reclaim | 1 << PG_slab | 1 << PG_swapcache | - 1 << PG_writeback ))) + 1 << PG_writeback | + 1 << PG_reserved ))) bad_page(__FUNCTION__, page); page->flags &= ~(1 << PG_uptodate | 1 << PG_error | @@ -1016,7 +1018,7 @@ void __pagevec_free(struct pagevec *pvec) fastcall void __free_pages(struct page *page, unsigned int order) { - if (!PageReserved(page) && put_page_testzero(page)) { + if (put_page_testzero(page)) { if (order == 0) free_hot_page(page); else @@ -1674,7 +1676,7 @@ void __init memmap_init_zone(unsigned long size, int nid, unsigned long zone, continue; page = pfn_to_page(pfn); set_page_links(page, zone, nid, pfn); - set_page_count(page, 0); + set_page_count(page, 1); reset_page_mapcount(page); SetPageReserved(page); INIT_LIST_HEAD(&page->lru); diff --git a/mm/rmap.c b/mm/rmap.c index 504757624cce..f69d5342ce7f 100644 --- a/mm/rmap.c +++ b/mm/rmap.c @@ -443,8 +443,6 @@ int page_referenced(struct page *page, int is_locked, int ignore_token) void page_add_anon_rmap(struct page *page, struct vm_area_struct *vma, unsigned long address) { - BUG_ON(PageReserved(page)); - if (atomic_inc_and_test(&page->_mapcount)) { struct anon_vma *anon_vma = vma->anon_vma; @@ -468,8 +466,7 @@ void page_add_anon_rmap(struct page *page, void page_add_file_rmap(struct page *page) { BUG_ON(PageAnon(page)); - if (!pfn_valid(page_to_pfn(page)) || PageReserved(page)) - return; + BUG_ON(!pfn_valid(page_to_pfn(page))); if (atomic_inc_and_test(&page->_mapcount)) inc_page_state(nr_mapped); @@ -483,8 +480,6 @@ void page_add_file_rmap(struct page *page) */ void page_remove_rmap(struct page *page) { - BUG_ON(PageReserved(page)); - if (atomic_add_negative(-1, &page->_mapcount)) { BUG_ON(page_mapcount(page) < 0); /* @@ -640,13 +635,13 @@ static void try_to_unmap_cluster(unsigned long cursor, continue; pfn = pte_pfn(*pte); - if (!pfn_valid(pfn)) + if (unlikely(!pfn_valid(pfn))) { + print_bad_pte(vma, *pte, address); continue; + } page = pfn_to_page(pfn); BUG_ON(PageAnon(page)); - if (PageReserved(page)) - continue; if (ptep_clear_flush_young(vma, address, pte)) continue; @@ -808,7 +803,6 @@ int try_to_unmap(struct page *page) { int ret; - BUG_ON(PageReserved(page)); BUG_ON(!PageLocked(page)); if (PageAnon(page)) diff --git a/mm/shmem.c b/mm/shmem.c index 6796311a23ef..37777f4c11f8 100644 --- a/mm/shmem.c +++ b/mm/shmem.c @@ -1506,8 +1506,10 @@ static void do_shmem_file_read(struct file *filp, loff_t *ppos, read_descriptor_ */ if (!offset) mark_page_accessed(page); - } else + } else { page = ZERO_PAGE(0); + page_cache_get(page); + } /* * Ok, we have the page, and it's up-to-date, so diff --git a/mm/swap.c b/mm/swap.c index 7771d2803f62..21d15f99805c 100644 --- a/mm/swap.c +++ b/mm/swap.c @@ -48,7 +48,7 @@ void put_page(struct page *page) } return; } - if (!PageReserved(page) && put_page_testzero(page)) + if (put_page_testzero(page)) __page_cache_release(page); } EXPORT_SYMBOL(put_page); @@ -215,7 +215,7 @@ void release_pages(struct page **pages, int nr, int cold) struct page *page = pages[i]; struct zone *pagezone; - if (PageReserved(page) || !put_page_testzero(page)) + if (!put_page_testzero(page)) continue; pagezone = page_zone(page); diff --git a/sound/core/pcm_native.c b/sound/core/pcm_native.c index 67abebabf83e..e97b2d162cc7 100644 --- a/sound/core/pcm_native.c +++ b/sound/core/pcm_native.c @@ -2949,8 +2949,7 @@ static struct page * snd_pcm_mmap_status_nopage(struct vm_area_struct *area, uns return NOPAGE_OOM; runtime = substream->runtime; page = virt_to_page(runtime->status); - if (!PageReserved(page)) - get_page(page); + get_page(page); if (type) *type = VM_FAULT_MINOR; return page; @@ -2992,8 +2991,7 @@ static struct page * snd_pcm_mmap_control_nopage(struct vm_area_struct *area, un return NOPAGE_OOM; runtime = substream->runtime; page = virt_to_page(runtime->control); - if (!PageReserved(page)) - get_page(page); + get_page(page); if (type) *type = VM_FAULT_MINOR; return page; @@ -3066,8 +3064,7 @@ static struct page *snd_pcm_mmap_data_nopage(struct vm_area_struct *area, unsign vaddr = runtime->dma_area + offset; page = virt_to_page(vaddr); } - if (!PageReserved(page)) - get_page(page); + get_page(page); if (type) *type = VM_FAULT_MINOR; return page; -- cgit v1.2.3 From 365e9c87a982c03d0af3886e29d877f581b59611 Mon Sep 17 00:00:00 2001 From: Hugh Dickins Date: Sat, 29 Oct 2005 18:16:18 -0700 Subject: [PATCH] mm: update_hiwaters just in time update_mem_hiwater has attracted various criticisms, in particular from those concerned with mm scalability. Originally it was called whenever rss or total_vm got raised. Then many of those callsites were replaced by a timer tick call from account_system_time. Now Frank van Maarseveen reports that to be found inadequate. How about this? Works for Frank. Replace update_mem_hiwater, a poor combination of two unrelated ops, by macros update_hiwater_rss and update_hiwater_vm. Don't attempt to keep mm->hiwater_rss up to date at timer tick, nor every time we raise rss (usually by 1): those are hot paths. Do the opposite, update only when about to lower rss (usually by many), or just before final accounting in do_exit. Handle mm->hiwater_vm in the same way, though it's much less of an issue. Demand that whoever collects these hiwater statistics do the work of taking the maximum with rss or total_vm. And there has been no collector of these hiwater statistics in the tree. The new convention needs an example, so match Frank's usage by adding a VmPeak line above VmSize to /proc//status, and also a VmHWM line above VmRSS (High-Water-Mark or High-Water-Memory). There was a particular anomaly during mremap move, that hiwater_vm might be captured too high. A fleeting such anomaly remains, but it's quickly corrected now, whereas before it would stick. What locking? None: if the app is racy then these statistics will be racy, it's not worth any overhead to make them exact. But whenever it suits, hiwater_vm is updated under exclusive mmap_sem, and hiwater_rss under page_table_lock (for now) or with preemption disabled (later on): without going to any trouble, minimize the time between reading current values and updating, to minimize those occasions when a racing thread bumps a count up and back down in between. Signed-off-by: Hugh Dickins Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/compat.c | 1 - fs/exec.c | 1 - fs/proc/task_mmu.c | 23 +++++++++++++++++++++-- include/linux/mm.h | 3 --- include/linux/sched.h | 10 ++++++++++ kernel/exit.c | 5 ++++- kernel/sched.c | 2 -- mm/fremap.c | 4 +++- mm/hugetlb.c | 3 +++ mm/memory.c | 17 +---------------- mm/mmap.c | 4 ++++ mm/mremap.c | 12 ++++++++++-- mm/nommu.c | 15 ++------------- mm/rmap.c | 6 ++++++ 14 files changed, 64 insertions(+), 42 deletions(-) (limited to 'include/linux') diff --git a/fs/compat.c b/fs/compat.c index a719e158e002..8e71cdbecc7c 100644 --- a/fs/compat.c +++ b/fs/compat.c @@ -1490,7 +1490,6 @@ int compat_do_execve(char * filename, /* execve success */ security_bprm_free(bprm); acct_update_integrals(current); - update_mem_hiwater(current); kfree(bprm); return retval; } diff --git a/fs/exec.c b/fs/exec.c index cefadf5ab83b..9bb55c8cf224 100644 --- a/fs/exec.c +++ b/fs/exec.c @@ -1207,7 +1207,6 @@ int do_execve(char * filename, /* execve success */ security_bprm_free(bprm); acct_update_integrals(current); - update_mem_hiwater(current); kfree(bprm); return retval; } diff --git a/fs/proc/task_mmu.c b/fs/proc/task_mmu.c index bccee7cf9ccd..7c89b4549049 100644 --- a/fs/proc/task_mmu.c +++ b/fs/proc/task_mmu.c @@ -14,22 +14,41 @@ char *task_mem(struct mm_struct *mm, char *buffer) { unsigned long data, text, lib; + unsigned long hiwater_vm, total_vm, hiwater_rss, total_rss; + + /* + * Note: to minimize their overhead, mm maintains hiwater_vm and + * hiwater_rss only when about to *lower* total_vm or rss. Any + * collector of these hiwater stats must therefore get total_vm + * and rss too, which will usually be the higher. Barriers? not + * worth the effort, such snapshots can always be inconsistent. + */ + hiwater_vm = total_vm = mm->total_vm; + if (hiwater_vm < mm->hiwater_vm) + hiwater_vm = mm->hiwater_vm; + hiwater_rss = total_rss = get_mm_rss(mm); + if (hiwater_rss < mm->hiwater_rss) + hiwater_rss = mm->hiwater_rss; data = mm->total_vm - mm->shared_vm - mm->stack_vm; text = (PAGE_ALIGN(mm->end_code) - (mm->start_code & PAGE_MASK)) >> 10; lib = (mm->exec_vm << (PAGE_SHIFT-10)) - text; buffer += sprintf(buffer, + "VmPeak:\t%8lu kB\n" "VmSize:\t%8lu kB\n" "VmLck:\t%8lu kB\n" + "VmHWM:\t%8lu kB\n" "VmRSS:\t%8lu kB\n" "VmData:\t%8lu kB\n" "VmStk:\t%8lu kB\n" "VmExe:\t%8lu kB\n" "VmLib:\t%8lu kB\n" "VmPTE:\t%8lu kB\n", - (mm->total_vm - mm->reserved_vm) << (PAGE_SHIFT-10), + hiwater_vm << (PAGE_SHIFT-10), + (total_vm - mm->reserved_vm) << (PAGE_SHIFT-10), mm->locked_vm << (PAGE_SHIFT-10), - get_mm_rss(mm) << (PAGE_SHIFT-10), + hiwater_rss << (PAGE_SHIFT-10), + total_rss << (PAGE_SHIFT-10), data << (PAGE_SHIFT-10), mm->stack_vm << (PAGE_SHIFT-10), text, lib, (PTRS_PER_PTE*sizeof(pte_t)*mm->nr_ptes) >> 10); diff --git a/include/linux/mm.h b/include/linux/mm.h index da42093250c3..7d4552fe0864 100644 --- a/include/linux/mm.h +++ b/include/linux/mm.h @@ -938,9 +938,6 @@ static inline void vm_stat_account(struct mm_struct *mm, } #endif /* CONFIG_PROC_FS */ -/* update per process rss and vm hiwater data */ -extern void update_mem_hiwater(struct task_struct *tsk); - #ifndef CONFIG_DEBUG_PAGEALLOC static inline void kernel_map_pages(struct page *page, int numpages, int enable) diff --git a/include/linux/sched.h b/include/linux/sched.h index afcaac66cbd5..a9c0b7d26303 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -256,6 +256,16 @@ extern void arch_unmap_area_topdown(struct mm_struct *, unsigned long); #define dec_mm_counter(mm, member) (mm)->_##member-- #define get_mm_rss(mm) ((mm)->_file_rss + (mm)->_anon_rss) +#define update_hiwater_rss(mm) do { \ + unsigned long _rss = get_mm_rss(mm); \ + if ((mm)->hiwater_rss < _rss) \ + (mm)->hiwater_rss = _rss; \ +} while (0) +#define update_hiwater_vm(mm) do { \ + if ((mm)->hiwater_vm < (mm)->total_vm) \ + (mm)->hiwater_vm = (mm)->total_vm; \ +} while (0) + typedef unsigned long mm_counter_t; struct mm_struct { diff --git a/kernel/exit.c b/kernel/exit.c index 3b25b182d2be..79f52b85d6ed 100644 --- a/kernel/exit.c +++ b/kernel/exit.c @@ -839,7 +839,10 @@ fastcall NORET_TYPE void do_exit(long code) preempt_count()); acct_update_integrals(tsk); - update_mem_hiwater(tsk); + if (tsk->mm) { + update_hiwater_rss(tsk->mm); + update_hiwater_vm(tsk->mm); + } group_dead = atomic_dec_and_test(&tsk->signal->live); if (group_dead) { del_timer_sync(&tsk->signal->real_timer); diff --git a/kernel/sched.c b/kernel/sched.c index 1e5cafdf4e27..4f26c544d02c 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -2511,8 +2511,6 @@ void account_system_time(struct task_struct *p, int hardirq_offset, cpustat->idle = cputime64_add(cpustat->idle, tmp); /* Account for system time used */ acct_update_integrals(p); - /* Update rss highwater mark */ - update_mem_hiwater(p); } /* diff --git a/mm/fremap.c b/mm/fremap.c index 7f08d10ceaff..49719a35769a 100644 --- a/mm/fremap.c +++ b/mm/fremap.c @@ -143,8 +143,10 @@ int install_file_pte(struct mm_struct *mm, struct vm_area_struct *vma, if (!pte) goto err_unlock; - if (!pte_none(*pte) && zap_pte(mm, vma, addr, pte)) + if (!pte_none(*pte) && zap_pte(mm, vma, addr, pte)) { + update_hiwater_rss(mm); dec_mm_counter(mm, file_rss); + } set_pte_at(mm, addr, pte, pgoff_to_pte(pgoff)); pte_val = *pte; diff --git a/mm/hugetlb.c b/mm/hugetlb.c index 094455bcbbf7..ac5f044bf514 100644 --- a/mm/hugetlb.c +++ b/mm/hugetlb.c @@ -310,6 +310,9 @@ void unmap_hugepage_range(struct vm_area_struct *vma, unsigned long start, BUG_ON(start & ~HPAGE_MASK); BUG_ON(end & ~HPAGE_MASK); + /* Update high watermark before we lower rss */ + update_hiwater_rss(mm); + for (address = start; address < end; address += HPAGE_SIZE) { ptep = huge_pte_offset(mm, address); if (! ptep) diff --git a/mm/memory.c b/mm/memory.c index a25ee1d3e20a..692ad810263d 100644 --- a/mm/memory.c +++ b/mm/memory.c @@ -820,6 +820,7 @@ unsigned long zap_page_range(struct vm_area_struct *vma, unsigned long address, lru_add_drain(); spin_lock(&mm->page_table_lock); tlb = tlb_gather_mmu(mm, 0); + update_hiwater_rss(mm); end = unmap_vmas(&tlb, mm, vma, address, end, &nr_accounted, details); tlb_finish_mmu(tlb, address, end); spin_unlock(&mm->page_table_lock); @@ -2225,22 +2226,6 @@ unsigned long vmalloc_to_pfn(void * vmalloc_addr) EXPORT_SYMBOL(vmalloc_to_pfn); -/* - * update_mem_hiwater - * - update per process rss and vm high water data - */ -void update_mem_hiwater(struct task_struct *tsk) -{ - if (tsk->mm) { - unsigned long rss = get_mm_rss(tsk->mm); - - if (tsk->mm->hiwater_rss < rss) - tsk->mm->hiwater_rss = rss; - if (tsk->mm->hiwater_vm < tsk->mm->total_vm) - tsk->mm->hiwater_vm = tsk->mm->total_vm; - } -} - #if !defined(__HAVE_ARCH_GATE_AREA) #if defined(AT_SYSINFO_EHDR) diff --git a/mm/mmap.c b/mm/mmap.c index 8a111792b8db..c43b28457007 100644 --- a/mm/mmap.c +++ b/mm/mmap.c @@ -1640,6 +1640,8 @@ find_extend_vma(struct mm_struct * mm, unsigned long addr) */ static void remove_vma_list(struct mm_struct *mm, struct vm_area_struct *vma) { + /* Update high watermark before we lower total_vm */ + update_hiwater_vm(mm); do { long nrpages = vma_pages(vma); @@ -1668,6 +1670,7 @@ static void unmap_region(struct mm_struct *mm, lru_add_drain(); spin_lock(&mm->page_table_lock); tlb = tlb_gather_mmu(mm, 0); + update_hiwater_rss(mm); unmap_vmas(&tlb, mm, vma, start, end, &nr_accounted, NULL); vm_unacct_memory(nr_accounted); free_pgtables(&tlb, vma, prev? prev->vm_end: FIRST_USER_ADDRESS, @@ -1953,6 +1956,7 @@ void exit_mmap(struct mm_struct *mm) flush_cache_mm(mm); tlb = tlb_gather_mmu(mm, 1); + /* Don't update_hiwater_rss(mm) here, do_exit already did */ /* Use -1 here to ensure all VMAs in the mm are unmapped */ end = unmap_vmas(&tlb, mm, vma, 0, -1, &nr_accounted, NULL); vm_unacct_memory(nr_accounted); diff --git a/mm/mremap.c b/mm/mremap.c index 318eea5467a0..ccf456477020 100644 --- a/mm/mremap.c +++ b/mm/mremap.c @@ -167,6 +167,7 @@ static unsigned long move_vma(struct vm_area_struct *vma, unsigned long new_pgoff; unsigned long moved_len; unsigned long excess = 0; + unsigned long hiwater_vm; int split = 0; /* @@ -205,9 +206,15 @@ static unsigned long move_vma(struct vm_area_struct *vma, } /* - * if we failed to move page tables we still do total_vm increment - * since do_munmap() will decrement it by old_len == new_len + * If we failed to move page tables we still do total_vm increment + * since do_munmap() will decrement it by old_len == new_len. + * + * Since total_vm is about to be raised artificially high for a + * moment, we need to restore high watermark afterwards: if stats + * are taken meanwhile, total_vm and hiwater_vm appear too high. + * If this were a serious issue, we'd add a flag to do_munmap(). */ + hiwater_vm = mm->hiwater_vm; mm->total_vm += new_len >> PAGE_SHIFT; vm_stat_account(mm, vma->vm_flags, vma->vm_file, new_len>>PAGE_SHIFT); @@ -216,6 +223,7 @@ static unsigned long move_vma(struct vm_area_struct *vma, vm_unacct_memory(excess >> PAGE_SHIFT); excess = 0; } + mm->hiwater_vm = hiwater_vm; /* Restore VM_ACCOUNT if one or two pieces of vma left */ if (excess) { diff --git a/mm/nommu.c b/mm/nommu.c index 599924886eb5..dfb124ffb9be 100644 --- a/mm/nommu.c +++ b/mm/nommu.c @@ -931,6 +931,8 @@ int do_munmap(struct mm_struct *mm, unsigned long addr, size_t len) realalloc -= kobjsize(vml); askedalloc -= sizeof(*vml); kfree(vml); + + update_hiwater_vm(mm); mm->total_vm -= len >> PAGE_SHIFT; #ifdef DEBUG @@ -1078,19 +1080,6 @@ void arch_unmap_area(struct mm_struct *mm, unsigned long addr) { } -void update_mem_hiwater(struct task_struct *tsk) -{ - unsigned long rss; - - if (likely(tsk->mm)) { - rss = get_mm_rss(tsk->mm); - if (tsk->mm->hiwater_rss < rss) - tsk->mm->hiwater_rss = rss; - if (tsk->mm->hiwater_vm < tsk->mm->total_vm) - tsk->mm->hiwater_vm = tsk->mm->total_vm; - } -} - void unmap_mapping_range(struct address_space *mapping, loff_t const holebegin, loff_t const holelen, int even_cows) diff --git a/mm/rmap.c b/mm/rmap.c index f69d5342ce7f..4c52c56c9905 100644 --- a/mm/rmap.c +++ b/mm/rmap.c @@ -538,6 +538,9 @@ static int try_to_unmap_one(struct page *page, struct vm_area_struct *vma) if (pte_dirty(pteval)) set_page_dirty(page); + /* Update high watermark before we lower rss */ + update_hiwater_rss(mm); + if (PageAnon(page)) { swp_entry_t entry = { .val = page->private }; /* @@ -628,6 +631,9 @@ static void try_to_unmap_cluster(unsigned long cursor, if (!pmd_present(*pmd)) goto out_unlock; + /* Update high watermark before we lower rss */ + update_hiwater_rss(mm); + for (original_pte = pte = pte_offset_map(pmd, address); address < end; pte++, address += PAGE_SIZE) { -- cgit v1.2.3 From f449952bc8bde7fbc73c6d20dff92b627a21f8b9 Mon Sep 17 00:00:00 2001 From: Hugh Dickins Date: Sat, 29 Oct 2005 18:16:19 -0700 Subject: [PATCH] mm: mm_struct hiwaters moved Slight and timid rearrangement of mm_struct: hiwater_rss and hiwater_vm were tacked on the end, but it seems better to keep them near _file_rss, _anon_rss and total_vm, in the same cacheline on those arches verified. There are likely to be more profitable rearrangements, but less obvious (is it good or bad that saved_auxv[AT_VECTOR_SIZE] isolates cpu_vm_mask and context from many others?), needing serious instrumentation. Signed-off-by: Hugh Dickins Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/sched.h | 19 +++++++++---------- 1 file changed, 9 insertions(+), 10 deletions(-) (limited to 'include/linux') diff --git a/include/linux/sched.h b/include/linux/sched.h index a9c0b7d26303..292cb57ce38f 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -291,16 +291,19 @@ struct mm_struct { * by mmlist_lock */ - unsigned long start_code, end_code, start_data, end_data; - unsigned long start_brk, brk, start_stack; - unsigned long arg_start, arg_end, env_start, env_end; - unsigned long total_vm, locked_vm, shared_vm; - unsigned long exec_vm, stack_vm, reserved_vm, def_flags, nr_ptes; - /* Special counters protected by the page_table_lock */ mm_counter_t _file_rss; mm_counter_t _anon_rss; + unsigned long hiwater_rss; /* High-watermark of RSS usage */ + unsigned long hiwater_vm; /* High-water virtual memory usage */ + + unsigned long total_vm, locked_vm, shared_vm, exec_vm; + unsigned long stack_vm, reserved_vm, def_flags, nr_ptes; + unsigned long start_code, end_code, start_data, end_data; + unsigned long start_brk, brk, start_stack; + unsigned long arg_start, arg_end, env_start, env_end; + unsigned long saved_auxv[AT_VECTOR_SIZE]; /* for /proc/PID/auxv */ unsigned dumpable:2; @@ -320,11 +323,7 @@ struct mm_struct { /* aio bits */ rwlock_t ioctx_list_lock; struct kioctx *ioctx_list; - struct kioctx default_kioctx; - - unsigned long hiwater_rss; /* High-water RSS usage */ - unsigned long hiwater_vm; /* High-water virtual memory usage */ }; struct sighand_struct { -- cgit v1.2.3 From 46dea3d092d23a58b42499cc8a21de0fad079f4a Mon Sep 17 00:00:00 2001 From: Hugh Dickins Date: Sat, 29 Oct 2005 18:16:20 -0700 Subject: [PATCH] mm: ia64 use expand_upwards ia64 has expand_backing_store function for growing its Register Backing Store vma upwards. But more complete code for this purpose is found in the CONFIG_STACK_GROWSUP part of mm/mmap.c. Uglify its #ifdefs further to provide expand_upwards for ia64 as well as expand_stack for parisc. The Register Backing Store vma should be marked VM_ACCOUNT. Implement the intention of growing it only a page at a time, instead of passing an address outside of the vma to handle_mm_fault, with unknown consequences. Signed-off-by: Hugh Dickins Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/ia64/mm/fault.c | 34 +++++++--------------------------- arch/ia64/mm/init.c | 2 +- include/linux/mm.h | 3 ++- mm/mmap.c | 17 ++++++++++++++--- 4 files changed, 24 insertions(+), 32 deletions(-) (limited to 'include/linux') diff --git a/arch/ia64/mm/fault.c b/arch/ia64/mm/fault.c index f21b55549787..af7eb087dca7 100644 --- a/arch/ia64/mm/fault.c +++ b/arch/ia64/mm/fault.c @@ -19,32 +19,6 @@ extern void die (char *, struct pt_regs *, long); -/* - * This routine is analogous to expand_stack() but instead grows the - * register backing store (which grows towards higher addresses). - * Since the register backing store is access sequentially, we - * disallow growing the RBS by more than a page at a time. Note that - * the VM_GROWSUP flag can be set on any VM area but that's fine - * because the total process size is still limited by RLIMIT_STACK and - * RLIMIT_AS. - */ -static inline long -expand_backing_store (struct vm_area_struct *vma, unsigned long address) -{ - unsigned long grow; - - grow = PAGE_SIZE >> PAGE_SHIFT; - if (address - vma->vm_start > current->signal->rlim[RLIMIT_STACK].rlim_cur - || (((vma->vm_mm->total_vm + grow) << PAGE_SHIFT) > current->signal->rlim[RLIMIT_AS].rlim_cur)) - return -ENOMEM; - vma->vm_end += PAGE_SIZE; - vma->vm_mm->total_vm += grow; - if (vma->vm_flags & VM_LOCKED) - vma->vm_mm->locked_vm += grow; - vm_stat_account(vma->vm_mm, vma->vm_flags, vma->vm_file, grow); - return 0; -} - /* * Return TRUE if ADDRESS points at a page in the kernel's mapped segment * (inside region 5, on ia64) and that page is present. @@ -185,7 +159,13 @@ ia64_do_page_fault (unsigned long address, unsigned long isr, struct pt_regs *re if (REGION_NUMBER(address) != REGION_NUMBER(vma->vm_start) || REGION_OFFSET(address) >= RGN_MAP_LIMIT) goto bad_area; - if (expand_backing_store(vma, address)) + /* + * Since the register backing store is accessed sequentially, + * we disallow growing it by more than a page at a time. + */ + if (address > vma->vm_end + PAGE_SIZE - sizeof(long)) + goto bad_area; + if (expand_upwards(vma, address)) goto bad_area; } goto good_area; diff --git a/arch/ia64/mm/init.c b/arch/ia64/mm/init.c index 98246acd4991..0063b2c50908 100644 --- a/arch/ia64/mm/init.c +++ b/arch/ia64/mm/init.c @@ -158,7 +158,7 @@ ia64_init_addr_space (void) vma->vm_start = current->thread.rbs_bot & PAGE_MASK; vma->vm_end = vma->vm_start + PAGE_SIZE; vma->vm_page_prot = protection_map[VM_DATA_DEFAULT_FLAGS & 0x7]; - vma->vm_flags = VM_DATA_DEFAULT_FLAGS | VM_GROWSUP; + vma->vm_flags = VM_DATA_DEFAULT_FLAGS|VM_GROWSUP|VM_ACCOUNT; down_write(¤t->mm->mmap_sem); if (insert_vm_struct(current->mm, vma)) { up_write(¤t->mm->mmap_sem); diff --git a/include/linux/mm.h b/include/linux/mm.h index 7d4552fe0864..89398032bc4b 100644 --- a/include/linux/mm.h +++ b/include/linux/mm.h @@ -896,7 +896,8 @@ void handle_ra_miss(struct address_space *mapping, unsigned long max_sane_readahead(unsigned long nr); /* Do stack extension */ -extern int expand_stack(struct vm_area_struct * vma, unsigned long address); +extern int expand_stack(struct vm_area_struct *vma, unsigned long address); +extern int expand_upwards(struct vm_area_struct *vma, unsigned long address); /* Look up the first VMA which satisfies addr < vm_end, NULL if none. */ extern struct vm_area_struct * find_vma(struct mm_struct * mm, unsigned long addr); diff --git a/mm/mmap.c b/mm/mmap.c index c43b28457007..d931d7e49ac9 100644 --- a/mm/mmap.c +++ b/mm/mmap.c @@ -1508,11 +1508,15 @@ static int acct_stack_growth(struct vm_area_struct * vma, unsigned long size, un return 0; } -#ifdef CONFIG_STACK_GROWSUP +#if defined(CONFIG_STACK_GROWSUP) || defined(CONFIG_IA64) /* - * vma is the first one with address > vma->vm_end. Have to extend vma. + * PA-RISC uses this for its stack; IA64 for its Register Backing Store. + * vma is the last one with address > vma->vm_end. Have to extend vma. */ -int expand_stack(struct vm_area_struct * vma, unsigned long address) +#ifdef CONFIG_STACK_GROWSUP +static inline +#endif +int expand_upwards(struct vm_area_struct *vma, unsigned long address) { int error; @@ -1550,6 +1554,13 @@ int expand_stack(struct vm_area_struct * vma, unsigned long address) anon_vma_unlock(vma); return error; } +#endif /* CONFIG_STACK_GROWSUP || CONFIG_IA64 */ + +#ifdef CONFIG_STACK_GROWSUP +int expand_stack(struct vm_area_struct *vma, unsigned long address) +{ + return expand_upwards(vma, address); +} struct vm_area_struct * find_extend_vma(struct mm_struct *mm, unsigned long addr) -- cgit v1.2.3 From 872fec16d9a0ed3b75b8893aa217e49cca575ee5 Mon Sep 17 00:00:00 2001 From: Hugh Dickins Date: Sat, 29 Oct 2005 18:16:21 -0700 Subject: [PATCH] mm: init_mm without ptlock First step in pushing down the page_table_lock. init_mm.page_table_lock has been used throughout the architectures (usually for ioremap): not to serialize kernel address space allocation (that's usually vmlist_lock), but because pud_alloc,pmd_alloc,pte_alloc_kernel expect caller holds it. Reverse that: don't lock or unlock init_mm.page_table_lock in any of the architectures; instead rely on pud_alloc,pmd_alloc,pte_alloc_kernel to take and drop it when allocating a new one, to check lest a racing task already did. Similarly no page_table_lock in vmalloc's map_vm_area. Some temporary ugliness in __pud_alloc and __pmd_alloc: since they also handle user mms, which are converted only by a later patch, for now they have to lock differently according to whether or not it's init_mm. If sources get muddled, there's a danger that an arch source taking init_mm.page_table_lock will be mixed with common source also taking it (or neither take it). So break the rules and make another change, which should break the build for such a mismatch: remove the redundant mm arg from pte_alloc_kernel (ppc64 scrapped its distinct ioremap_mm in 2.6.13). Exceptions: arm26 used pte_alloc_kernel on user mm, now pte_alloc_map; ia64 used pte_alloc_map on init_mm, now pte_alloc_kernel; parisc had bad args to pmd_alloc and pte_alloc_kernel in unused USE_HPPA_IOREMAP code; ppc64 map_io_page forgot to unlock on failure; ppc mmu_mapin_ram and ppc64 im_free took page_table_lock for no good reason. Signed-off-by: Hugh Dickins Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/alpha/mm/remap.c | 6 +---- arch/arm/mm/consistent.c | 6 +---- arch/arm/mm/ioremap.c | 4 +-- arch/arm26/mm/memc.c | 3 ++- arch/cris/mm/ioremap.c | 4 +-- arch/frv/mm/dma-alloc.c | 5 +--- arch/i386/mm/ioremap.c | 4 +-- arch/ia64/mm/init.c | 11 +++----- arch/m32r/mm/ioremap.c | 4 +-- arch/m68k/mm/kmap.c | 2 +- arch/m68k/sun3x/dvma.c | 2 +- arch/mips/mm/ioremap.c | 4 +-- arch/parisc/kernel/pci-dma.c | 2 +- arch/parisc/mm/ioremap.c | 6 ++--- arch/ppc/kernel/dma-mapping.c | 6 +---- arch/ppc/mm/4xx_mmu.c | 4 --- arch/ppc/mm/pgtable.c | 4 +-- arch/ppc64/mm/imalloc.c | 5 ---- arch/ppc64/mm/init.c | 4 +-- arch/s390/mm/ioremap.c | 4 +-- arch/sh/mm/ioremap.c | 4 +-- arch/sh64/mm/ioremap.c | 4 +-- arch/x86_64/mm/ioremap.c | 4 +-- include/linux/mm.h | 2 +- mm/memory.c | 60 +++++++++++++++++++------------------------ mm/vmalloc.c | 4 +-- 26 files changed, 54 insertions(+), 114 deletions(-) (limited to 'include/linux') diff --git a/arch/alpha/mm/remap.c b/arch/alpha/mm/remap.c index 19817ad3d89b..a78356c3ead5 100644 --- a/arch/alpha/mm/remap.c +++ b/arch/alpha/mm/remap.c @@ -2,7 +2,6 @@ #include #include -/* called with the page_table_lock held */ static inline void remap_area_pte(pte_t * pte, unsigned long address, unsigned long size, unsigned long phys_addr, unsigned long flags) @@ -31,7 +30,6 @@ remap_area_pte(pte_t * pte, unsigned long address, unsigned long size, } while (address && (address < end)); } -/* called with the page_table_lock held */ static inline int remap_area_pmd(pmd_t * pmd, unsigned long address, unsigned long size, unsigned long phys_addr, unsigned long flags) @@ -46,7 +44,7 @@ remap_area_pmd(pmd_t * pmd, unsigned long address, unsigned long size, if (address >= end) BUG(); do { - pte_t * pte = pte_alloc_kernel(&init_mm, pmd, address); + pte_t * pte = pte_alloc_kernel(pmd, address); if (!pte) return -ENOMEM; remap_area_pte(pte, address, end - address, @@ -70,7 +68,6 @@ __alpha_remap_area_pages(unsigned long address, unsigned long phys_addr, flush_cache_all(); if (address >= end) BUG(); - spin_lock(&init_mm.page_table_lock); do { pmd_t *pmd; pmd = pmd_alloc(&init_mm, dir, address); @@ -84,7 +81,6 @@ __alpha_remap_area_pages(unsigned long address, unsigned long phys_addr, address = (address + PGDIR_SIZE) & PGDIR_MASK; dir++; } while (address && (address < end)); - spin_unlock(&init_mm.page_table_lock); return error; } diff --git a/arch/arm/mm/consistent.c b/arch/arm/mm/consistent.c index 82f4d5e27c54..47b0b767f080 100644 --- a/arch/arm/mm/consistent.c +++ b/arch/arm/mm/consistent.c @@ -397,8 +397,6 @@ static int __init consistent_init(void) pte_t *pte; int ret = 0; - spin_lock(&init_mm.page_table_lock); - do { pgd = pgd_offset(&init_mm, CONSISTENT_BASE); pmd = pmd_alloc(&init_mm, pgd, CONSISTENT_BASE); @@ -409,7 +407,7 @@ static int __init consistent_init(void) } WARN_ON(!pmd_none(*pmd)); - pte = pte_alloc_kernel(&init_mm, pmd, CONSISTENT_BASE); + pte = pte_alloc_kernel(pmd, CONSISTENT_BASE); if (!pte) { printk(KERN_ERR "%s: no pte tables\n", __func__); ret = -ENOMEM; @@ -419,8 +417,6 @@ static int __init consistent_init(void) consistent_pte = pte; } while (0); - spin_unlock(&init_mm.page_table_lock); - return ret; } diff --git a/arch/arm/mm/ioremap.c b/arch/arm/mm/ioremap.c index 6fb1258df1b5..0f128c28fee4 100644 --- a/arch/arm/mm/ioremap.c +++ b/arch/arm/mm/ioremap.c @@ -75,7 +75,7 @@ remap_area_pmd(pmd_t * pmd, unsigned long address, unsigned long size, pgprot = __pgprot(L_PTE_PRESENT | L_PTE_YOUNG | L_PTE_DIRTY | L_PTE_WRITE | flags); do { - pte_t * pte = pte_alloc_kernel(&init_mm, pmd, address); + pte_t * pte = pte_alloc_kernel(pmd, address); if (!pte) return -ENOMEM; remap_area_pte(pte, address, end - address, address + phys_addr, pgprot); @@ -97,7 +97,6 @@ remap_area_pages(unsigned long start, unsigned long phys_addr, phys_addr -= address; dir = pgd_offset(&init_mm, address); BUG_ON(address >= end); - spin_lock(&init_mm.page_table_lock); do { pmd_t *pmd = pmd_alloc(&init_mm, dir, address); if (!pmd) { @@ -114,7 +113,6 @@ remap_area_pages(unsigned long start, unsigned long phys_addr, dir++; } while (address && (address < end)); - spin_unlock(&init_mm.page_table_lock); flush_cache_vmap(start, end); return err; } diff --git a/arch/arm26/mm/memc.c b/arch/arm26/mm/memc.c index 8e8a2bb2487d..d6b008b8db76 100644 --- a/arch/arm26/mm/memc.c +++ b/arch/arm26/mm/memc.c @@ -92,7 +92,7 @@ pgd_t *get_pgd_slow(struct mm_struct *mm) if (!new_pmd) goto no_pmd; - new_pte = pte_alloc_kernel(mm, new_pmd, 0); + new_pte = pte_alloc_map(mm, new_pmd, 0); if (!new_pte) goto no_pte; @@ -101,6 +101,7 @@ pgd_t *get_pgd_slow(struct mm_struct *mm) init_pte = pte_offset(init_pmd, 0); set_pte(new_pte, *init_pte); + pte_unmap(new_pte); /* * the page table entries are zeroed diff --git a/arch/cris/mm/ioremap.c b/arch/cris/mm/ioremap.c index ebba11e270fa..a92ac9877582 100644 --- a/arch/cris/mm/ioremap.c +++ b/arch/cris/mm/ioremap.c @@ -52,7 +52,7 @@ static inline int remap_area_pmd(pmd_t * pmd, unsigned long address, unsigned lo if (address >= end) BUG(); do { - pte_t * pte = pte_alloc_kernel(&init_mm, pmd, address); + pte_t * pte = pte_alloc_kernel(pmd, address); if (!pte) return -ENOMEM; remap_area_pte(pte, address, end - address, address + phys_addr, prot); @@ -74,7 +74,6 @@ static int remap_area_pages(unsigned long address, unsigned long phys_addr, flush_cache_all(); if (address >= end) BUG(); - spin_lock(&init_mm.page_table_lock); do { pud_t *pud; pmd_t *pmd; @@ -94,7 +93,6 @@ static int remap_area_pages(unsigned long address, unsigned long phys_addr, address = (address + PGDIR_SIZE) & PGDIR_MASK; dir++; } while (address && (address < end)); - spin_unlock(&init_mm.page_table_lock); flush_tlb_all(); return error; } diff --git a/arch/frv/mm/dma-alloc.c b/arch/frv/mm/dma-alloc.c index cfc4f97490c6..342823aad758 100644 --- a/arch/frv/mm/dma-alloc.c +++ b/arch/frv/mm/dma-alloc.c @@ -55,21 +55,18 @@ static int map_page(unsigned long va, unsigned long pa, pgprot_t prot) pte_t *pte; int err = -ENOMEM; - spin_lock(&init_mm.page_table_lock); - /* Use upper 10 bits of VA to index the first level map */ pge = pgd_offset_k(va); pue = pud_offset(pge, va); pme = pmd_offset(pue, va); /* Use middle 10 bits of VA to index the second-level map */ - pte = pte_alloc_kernel(&init_mm, pme, va); + pte = pte_alloc_kernel(pme, va); if (pte != 0) { err = 0; set_pte(pte, mk_pte_phys(pa & PAGE_MASK, prot)); } - spin_unlock(&init_mm.page_table_lock); return err; } diff --git a/arch/i386/mm/ioremap.c b/arch/i386/mm/ioremap.c index f379b8d67558..5d09de8d1c6b 100644 --- a/arch/i386/mm/ioremap.c +++ b/arch/i386/mm/ioremap.c @@ -28,7 +28,7 @@ static int ioremap_pte_range(pmd_t *pmd, unsigned long addr, unsigned long pfn; pfn = phys_addr >> PAGE_SHIFT; - pte = pte_alloc_kernel(&init_mm, pmd, addr); + pte = pte_alloc_kernel(pmd, addr); if (!pte) return -ENOMEM; do { @@ -87,14 +87,12 @@ static int ioremap_page_range(unsigned long addr, flush_cache_all(); phys_addr -= addr; pgd = pgd_offset_k(addr); - spin_lock(&init_mm.page_table_lock); do { next = pgd_addr_end(addr, end); err = ioremap_pud_range(pgd, addr, next, phys_addr+addr, flags); if (err) break; } while (pgd++, addr = next, addr != end); - spin_unlock(&init_mm.page_table_lock); flush_tlb_all(); return err; } diff --git a/arch/ia64/mm/init.c b/arch/ia64/mm/init.c index 0063b2c50908..e3215ba64ffd 100644 --- a/arch/ia64/mm/init.c +++ b/arch/ia64/mm/init.c @@ -275,26 +275,21 @@ put_kernel_page (struct page *page, unsigned long address, pgprot_t pgprot) pgd = pgd_offset_k(address); /* note: this is NOT pgd_offset()! */ - spin_lock(&init_mm.page_table_lock); { pud = pud_alloc(&init_mm, pgd, address); if (!pud) goto out; - pmd = pmd_alloc(&init_mm, pud, address); if (!pmd) goto out; - pte = pte_alloc_map(&init_mm, pmd, address); + pte = pte_alloc_kernel(pmd, address); if (!pte) goto out; - if (!pte_none(*pte)) { - pte_unmap(pte); + if (!pte_none(*pte)) goto out; - } set_pte(pte, mk_pte(page, pgprot)); - pte_unmap(pte); } - out: spin_unlock(&init_mm.page_table_lock); + out: /* no need for flush_tlb */ return page; } diff --git a/arch/m32r/mm/ioremap.c b/arch/m32r/mm/ioremap.c index 70c59055c19c..a151849a605e 100644 --- a/arch/m32r/mm/ioremap.c +++ b/arch/m32r/mm/ioremap.c @@ -67,7 +67,7 @@ remap_area_pmd(pmd_t * pmd, unsigned long address, unsigned long size, if (address >= end) BUG(); do { - pte_t * pte = pte_alloc_kernel(&init_mm, pmd, address); + pte_t * pte = pte_alloc_kernel(pmd, address); if (!pte) return -ENOMEM; remap_area_pte(pte, address, end - address, address + phys_addr, flags); @@ -90,7 +90,6 @@ remap_area_pages(unsigned long address, unsigned long phys_addr, flush_cache_all(); if (address >= end) BUG(); - spin_lock(&init_mm.page_table_lock); do { pmd_t *pmd; pmd = pmd_alloc(&init_mm, dir, address); @@ -104,7 +103,6 @@ remap_area_pages(unsigned long address, unsigned long phys_addr, address = (address + PGDIR_SIZE) & PGDIR_MASK; dir++; } while (address && (address < end)); - spin_unlock(&init_mm.page_table_lock); flush_tlb_all(); return error; } diff --git a/arch/m68k/mm/kmap.c b/arch/m68k/mm/kmap.c index 5dcb3fa35ea9..fe2383e36b06 100644 --- a/arch/m68k/mm/kmap.c +++ b/arch/m68k/mm/kmap.c @@ -201,7 +201,7 @@ void *__ioremap(unsigned long physaddr, unsigned long size, int cacheflag) virtaddr += PTRTREESIZE; size -= PTRTREESIZE; } else { - pte_dir = pte_alloc_kernel(&init_mm, pmd_dir, virtaddr); + pte_dir = pte_alloc_kernel(pmd_dir, virtaddr); if (!pte_dir) { printk("ioremap: no mem for pte_dir\n"); return NULL; diff --git a/arch/m68k/sun3x/dvma.c b/arch/m68k/sun3x/dvma.c index 32e55adfeb8e..117481e86305 100644 --- a/arch/m68k/sun3x/dvma.c +++ b/arch/m68k/sun3x/dvma.c @@ -116,7 +116,7 @@ inline int dvma_map_cpu(unsigned long kaddr, pte_t *pte; unsigned long end3; - if((pte = pte_alloc_kernel(&init_mm, pmd, vaddr)) == NULL) { + if((pte = pte_alloc_kernel(pmd, vaddr)) == NULL) { ret = -ENOMEM; goto out; } diff --git a/arch/mips/mm/ioremap.c b/arch/mips/mm/ioremap.c index 9c44ca70befa..3101d1db5592 100644 --- a/arch/mips/mm/ioremap.c +++ b/arch/mips/mm/ioremap.c @@ -55,7 +55,7 @@ static inline int remap_area_pmd(pmd_t * pmd, unsigned long address, if (address >= end) BUG(); do { - pte_t * pte = pte_alloc_kernel(&init_mm, pmd, address); + pte_t * pte = pte_alloc_kernel(pmd, address); if (!pte) return -ENOMEM; remap_area_pte(pte, address, end - address, address + phys_addr, flags); @@ -77,7 +77,6 @@ static int remap_area_pages(unsigned long address, phys_t phys_addr, flush_cache_all(); if (address >= end) BUG(); - spin_lock(&init_mm.page_table_lock); do { pud_t *pud; pmd_t *pmd; @@ -96,7 +95,6 @@ static int remap_area_pages(unsigned long address, phys_t phys_addr, address = (address + PGDIR_SIZE) & PGDIR_MASK; dir++; } while (address && (address < end)); - spin_unlock(&init_mm.page_table_lock); flush_tlb_all(); return error; } diff --git a/arch/parisc/kernel/pci-dma.c b/arch/parisc/kernel/pci-dma.c index ae6213d71670..f94a02ef3d95 100644 --- a/arch/parisc/kernel/pci-dma.c +++ b/arch/parisc/kernel/pci-dma.c @@ -114,7 +114,7 @@ static inline int map_pmd_uncached(pmd_t * pmd, unsigned long vaddr, if (end > PGDIR_SIZE) end = PGDIR_SIZE; do { - pte_t * pte = pte_alloc_kernel(&init_mm, pmd, vaddr); + pte_t * pte = pte_alloc_kernel(pmd, vaddr); if (!pte) return -ENOMEM; if (map_pte_uncached(pte, orig_vaddr, end - vaddr, paddr_ptr)) diff --git a/arch/parisc/mm/ioremap.c b/arch/parisc/mm/ioremap.c index f2df502cdae3..5c7a1b3b9326 100644 --- a/arch/parisc/mm/ioremap.c +++ b/arch/parisc/mm/ioremap.c @@ -52,7 +52,7 @@ static inline int remap_area_pmd(pmd_t * pmd, unsigned long address, unsigned lo if (address >= end) BUG(); do { - pte_t * pte = pte_alloc_kernel(NULL, pmd, address); + pte_t * pte = pte_alloc_kernel(pmd, address); if (!pte) return -ENOMEM; remap_area_pte(pte, address, end - address, address + phys_addr, flags); @@ -75,10 +75,9 @@ static int remap_area_pages(unsigned long address, unsigned long phys_addr, flush_cache_all(); if (address >= end) BUG(); - spin_lock(&init_mm.page_table_lock); do { pmd_t *pmd; - pmd = pmd_alloc(dir, address); + pmd = pmd_alloc(&init_mm, dir, address); error = -ENOMEM; if (!pmd) break; @@ -89,7 +88,6 @@ static int remap_area_pages(unsigned long address, unsigned long phys_addr, address = (address + PGDIR_SIZE) & PGDIR_MASK; dir++; } while (address && (address < end)); - spin_unlock(&init_mm.page_table_lock); flush_tlb_all(); return error; } diff --git a/arch/ppc/kernel/dma-mapping.c b/arch/ppc/kernel/dma-mapping.c index 0f710d2baec6..685fd0defe23 100644 --- a/arch/ppc/kernel/dma-mapping.c +++ b/arch/ppc/kernel/dma-mapping.c @@ -335,8 +335,6 @@ static int __init dma_alloc_init(void) pte_t *pte; int ret = 0; - spin_lock(&init_mm.page_table_lock); - do { pgd = pgd_offset(&init_mm, CONSISTENT_BASE); pmd = pmd_alloc(&init_mm, pgd, CONSISTENT_BASE); @@ -347,7 +345,7 @@ static int __init dma_alloc_init(void) } WARN_ON(!pmd_none(*pmd)); - pte = pte_alloc_kernel(&init_mm, pmd, CONSISTENT_BASE); + pte = pte_alloc_kernel(pmd, CONSISTENT_BASE); if (!pte) { printk(KERN_ERR "%s: no pte tables\n", __func__); ret = -ENOMEM; @@ -357,8 +355,6 @@ static int __init dma_alloc_init(void) consistent_pte = pte; } while (0); - spin_unlock(&init_mm.page_table_lock); - return ret; } diff --git a/arch/ppc/mm/4xx_mmu.c b/arch/ppc/mm/4xx_mmu.c index b7bcbc232f39..4d006aa1a0d1 100644 --- a/arch/ppc/mm/4xx_mmu.c +++ b/arch/ppc/mm/4xx_mmu.c @@ -110,13 +110,11 @@ unsigned long __init mmu_mapin_ram(void) pmd_t *pmdp; unsigned long val = p | _PMD_SIZE_16M | _PAGE_HWEXEC | _PAGE_HWWRITE; - spin_lock(&init_mm.page_table_lock); pmdp = pmd_offset(pgd_offset_k(v), v); pmd_val(*pmdp++) = val; pmd_val(*pmdp++) = val; pmd_val(*pmdp++) = val; pmd_val(*pmdp++) = val; - spin_unlock(&init_mm.page_table_lock); v += LARGE_PAGE_SIZE_16M; p += LARGE_PAGE_SIZE_16M; @@ -127,10 +125,8 @@ unsigned long __init mmu_mapin_ram(void) pmd_t *pmdp; unsigned long val = p | _PMD_SIZE_4M | _PAGE_HWEXEC | _PAGE_HWWRITE; - spin_lock(&init_mm.page_table_lock); pmdp = pmd_offset(pgd_offset_k(v), v); pmd_val(*pmdp) = val; - spin_unlock(&init_mm.page_table_lock); v += LARGE_PAGE_SIZE_4M; p += LARGE_PAGE_SIZE_4M; diff --git a/arch/ppc/mm/pgtable.c b/arch/ppc/mm/pgtable.c index 43505b1fc5d8..6ea9185fd120 100644 --- a/arch/ppc/mm/pgtable.c +++ b/arch/ppc/mm/pgtable.c @@ -280,18 +280,16 @@ map_page(unsigned long va, phys_addr_t pa, int flags) pte_t *pg; int err = -ENOMEM; - spin_lock(&init_mm.page_table_lock); /* Use upper 10 bits of VA to index the first level map */ pd = pmd_offset(pgd_offset_k(va), va); /* Use middle 10 bits of VA to index the second-level map */ - pg = pte_alloc_kernel(&init_mm, pd, va); + pg = pte_alloc_kernel(pd, va); if (pg != 0) { err = 0; set_pte_at(&init_mm, va, pg, pfn_pte(pa >> PAGE_SHIFT, __pgprot(flags))); if (mem_init_done) flush_HPTE(0, va, pmd_val(*pd)); } - spin_unlock(&init_mm.page_table_lock); return err; } diff --git a/arch/ppc64/mm/imalloc.c b/arch/ppc64/mm/imalloc.c index c65b87b92756..f4ca29cf5364 100644 --- a/arch/ppc64/mm/imalloc.c +++ b/arch/ppc64/mm/imalloc.c @@ -300,12 +300,7 @@ void im_free(void * addr) for (p = &imlist ; (tmp = *p) ; p = &tmp->next) { if (tmp->addr == addr) { *p = tmp->next; - - /* XXX: do we need the lock? */ - spin_lock(&init_mm.page_table_lock); unmap_vm_area(tmp); - spin_unlock(&init_mm.page_table_lock); - kfree(tmp); up(&imlist_sem); return; diff --git a/arch/ppc64/mm/init.c b/arch/ppc64/mm/init.c index be64b157afce..a45584b3440c 100644 --- a/arch/ppc64/mm/init.c +++ b/arch/ppc64/mm/init.c @@ -155,7 +155,6 @@ static int map_io_page(unsigned long ea, unsigned long pa, int flags) unsigned long vsid; if (mem_init_done) { - spin_lock(&init_mm.page_table_lock); pgdp = pgd_offset_k(ea); pudp = pud_alloc(&init_mm, pgdp, ea); if (!pudp) @@ -163,12 +162,11 @@ static int map_io_page(unsigned long ea, unsigned long pa, int flags) pmdp = pmd_alloc(&init_mm, pudp, ea); if (!pmdp) return -ENOMEM; - ptep = pte_alloc_kernel(&init_mm, pmdp, ea); + ptep = pte_alloc_kernel(pmdp, ea); if (!ptep) return -ENOMEM; set_pte_at(&init_mm, ea, ptep, pfn_pte(pa >> PAGE_SHIFT, __pgprot(flags))); - spin_unlock(&init_mm.page_table_lock); } else { unsigned long va, vpn, hash, hpteg; diff --git a/arch/s390/mm/ioremap.c b/arch/s390/mm/ioremap.c index c6c39d868bc8..0f6e9ecbefe2 100644 --- a/arch/s390/mm/ioremap.c +++ b/arch/s390/mm/ioremap.c @@ -58,7 +58,7 @@ static inline int remap_area_pmd(pmd_t * pmd, unsigned long address, unsigned lo if (address >= end) BUG(); do { - pte_t * pte = pte_alloc_kernel(&init_mm, pmd, address); + pte_t * pte = pte_alloc_kernel(pmd, address); if (!pte) return -ENOMEM; remap_area_pte(pte, address, end - address, address + phys_addr, flags); @@ -80,7 +80,6 @@ static int remap_area_pages(unsigned long address, unsigned long phys_addr, flush_cache_all(); if (address >= end) BUG(); - spin_lock(&init_mm.page_table_lock); do { pmd_t *pmd; pmd = pmd_alloc(&init_mm, dir, address); @@ -94,7 +93,6 @@ static int remap_area_pages(unsigned long address, unsigned long phys_addr, address = (address + PGDIR_SIZE) & PGDIR_MASK; dir++; } while (address && (address < end)); - spin_unlock(&init_mm.page_table_lock); flush_tlb_all(); return 0; } diff --git a/arch/sh/mm/ioremap.c b/arch/sh/mm/ioremap.c index 9f490c2742f0..e794e27a72f1 100644 --- a/arch/sh/mm/ioremap.c +++ b/arch/sh/mm/ioremap.c @@ -57,7 +57,7 @@ static inline int remap_area_pmd(pmd_t * pmd, unsigned long address, if (address >= end) BUG(); do { - pte_t * pte = pte_alloc_kernel(&init_mm, pmd, address); + pte_t * pte = pte_alloc_kernel(pmd, address); if (!pte) return -ENOMEM; remap_area_pte(pte, address, end - address, address + phys_addr, flags); @@ -79,7 +79,6 @@ int remap_area_pages(unsigned long address, unsigned long phys_addr, flush_cache_all(); if (address >= end) BUG(); - spin_lock(&init_mm.page_table_lock); do { pmd_t *pmd; pmd = pmd_alloc(&init_mm, dir, address); @@ -93,7 +92,6 @@ int remap_area_pages(unsigned long address, unsigned long phys_addr, address = (address + PGDIR_SIZE) & PGDIR_MASK; dir++; } while (address && (address < end)); - spin_unlock(&init_mm.page_table_lock); flush_tlb_all(); return error; } diff --git a/arch/sh64/mm/ioremap.c b/arch/sh64/mm/ioremap.c index f4003da556bc..fb1866fa2c9d 100644 --- a/arch/sh64/mm/ioremap.c +++ b/arch/sh64/mm/ioremap.c @@ -79,7 +79,7 @@ static inline int remap_area_pmd(pmd_t * pmd, unsigned long address, unsigned lo BUG(); do { - pte_t * pte = pte_alloc_kernel(&init_mm, pmd, address); + pte_t * pte = pte_alloc_kernel(pmd, address); if (!pte) return -ENOMEM; remap_area_pte(pte, address, end - address, address + phys_addr, flags); @@ -101,7 +101,6 @@ static int remap_area_pages(unsigned long address, unsigned long phys_addr, flush_cache_all(); if (address >= end) BUG(); - spin_lock(&init_mm.page_table_lock); do { pmd_t *pmd = pmd_alloc(&init_mm, dir, address); error = -ENOMEM; @@ -115,7 +114,6 @@ static int remap_area_pages(unsigned long address, unsigned long phys_addr, address = (address + PGDIR_SIZE) & PGDIR_MASK; dir++; } while (address && (address < end)); - spin_unlock(&init_mm.page_table_lock); flush_tlb_all(); return 0; } diff --git a/arch/x86_64/mm/ioremap.c b/arch/x86_64/mm/ioremap.c index 6972df480d2b..ecf7acb5db9b 100644 --- a/arch/x86_64/mm/ioremap.c +++ b/arch/x86_64/mm/ioremap.c @@ -60,7 +60,7 @@ static inline int remap_area_pmd(pmd_t * pmd, unsigned long address, unsigned lo if (address >= end) BUG(); do { - pte_t * pte = pte_alloc_kernel(&init_mm, pmd, address); + pte_t * pte = pte_alloc_kernel(pmd, address); if (!pte) return -ENOMEM; remap_area_pte(pte, address, end - address, address + phys_addr, flags); @@ -105,7 +105,6 @@ static int remap_area_pages(unsigned long address, unsigned long phys_addr, flush_cache_all(); if (address >= end) BUG(); - spin_lock(&init_mm.page_table_lock); do { pud_t *pud; pud = pud_alloc(&init_mm, pgd, address); @@ -119,7 +118,6 @@ static int remap_area_pages(unsigned long address, unsigned long phys_addr, address = (address + PGDIR_SIZE) & PGDIR_MASK; pgd++; } while (address && (address < end)); - spin_unlock(&init_mm.page_table_lock); flush_tlb_all(); return error; } diff --git a/include/linux/mm.h b/include/linux/mm.h index 89398032bc4b..b9fa82b96d9e 100644 --- a/include/linux/mm.h +++ b/include/linux/mm.h @@ -706,7 +706,7 @@ static inline void unmap_shared_mapping_range(struct address_space *mapping, extern int vmtruncate(struct inode * inode, loff_t offset); extern pud_t *FASTCALL(__pud_alloc(struct mm_struct *mm, pgd_t *pgd, unsigned long address)); extern pmd_t *FASTCALL(__pmd_alloc(struct mm_struct *mm, pud_t *pud, unsigned long address)); -extern pte_t *FASTCALL(pte_alloc_kernel(struct mm_struct *mm, pmd_t *pmd, unsigned long address)); +extern pte_t *FASTCALL(pte_alloc_kernel(pmd_t *pmd, unsigned long address)); extern pte_t *FASTCALL(pte_alloc_map(struct mm_struct *mm, pmd_t *pmd, unsigned long address)); extern int install_page(struct mm_struct *mm, struct vm_area_struct *vma, unsigned long addr, struct page *page, pgprot_t prot); extern int install_file_pte(struct mm_struct *mm, struct vm_area_struct *vma, unsigned long addr, unsigned long pgoff, pgprot_t prot); diff --git a/mm/memory.c b/mm/memory.c index 692ad810263d..95a4553c75f7 100644 --- a/mm/memory.c +++ b/mm/memory.c @@ -307,28 +307,22 @@ out: return pte_offset_map(pmd, address); } -pte_t fastcall * pte_alloc_kernel(struct mm_struct *mm, pmd_t *pmd, unsigned long address) +pte_t fastcall * pte_alloc_kernel(pmd_t *pmd, unsigned long address) { if (!pmd_present(*pmd)) { pte_t *new; - spin_unlock(&mm->page_table_lock); - new = pte_alloc_one_kernel(mm, address); - spin_lock(&mm->page_table_lock); + new = pte_alloc_one_kernel(&init_mm, address); if (!new) return NULL; - /* - * Because we dropped the lock, we should re-check the - * entry, as somebody else could have populated it.. - */ - if (pmd_present(*pmd)) { + spin_lock(&init_mm.page_table_lock); + if (pmd_present(*pmd)) pte_free_kernel(new); - goto out; - } - pmd_populate_kernel(mm, pmd, new); + else + pmd_populate_kernel(&init_mm, pmd, new); + spin_unlock(&init_mm.page_table_lock); } -out: return pte_offset_kernel(pmd, address); } @@ -2097,30 +2091,30 @@ int __handle_mm_fault(struct mm_struct *mm, struct vm_area_struct *vma, #ifndef __PAGETABLE_PUD_FOLDED /* * Allocate page upper directory. - * - * We've already handled the fast-path in-line, and we own the - * page table lock. + * We've already handled the fast-path in-line. */ pud_t fastcall *__pud_alloc(struct mm_struct *mm, pgd_t *pgd, unsigned long address) { pud_t *new; - spin_unlock(&mm->page_table_lock); + if (mm != &init_mm) /* Temporary bridging hack */ + spin_unlock(&mm->page_table_lock); new = pud_alloc_one(mm, address); - spin_lock(&mm->page_table_lock); - if (!new) + if (!new) { + if (mm != &init_mm) /* Temporary bridging hack */ + spin_lock(&mm->page_table_lock); return NULL; + } - /* - * Because we dropped the lock, we should re-check the - * entry, as somebody else could have populated it.. - */ + spin_lock(&mm->page_table_lock); if (pgd_present(*pgd)) { pud_free(new); goto out; } pgd_populate(mm, pgd, new); out: + if (mm == &init_mm) /* Temporary bridging hack */ + spin_unlock(&mm->page_table_lock); return pud_offset(pgd, address); } #endif /* __PAGETABLE_PUD_FOLDED */ @@ -2128,24 +2122,22 @@ pud_t fastcall *__pud_alloc(struct mm_struct *mm, pgd_t *pgd, unsigned long addr #ifndef __PAGETABLE_PMD_FOLDED /* * Allocate page middle directory. - * - * We've already handled the fast-path in-line, and we own the - * page table lock. + * We've already handled the fast-path in-line. */ pmd_t fastcall *__pmd_alloc(struct mm_struct *mm, pud_t *pud, unsigned long address) { pmd_t *new; - spin_unlock(&mm->page_table_lock); + if (mm != &init_mm) /* Temporary bridging hack */ + spin_unlock(&mm->page_table_lock); new = pmd_alloc_one(mm, address); - spin_lock(&mm->page_table_lock); - if (!new) + if (!new) { + if (mm != &init_mm) /* Temporary bridging hack */ + spin_lock(&mm->page_table_lock); return NULL; + } - /* - * Because we dropped the lock, we should re-check the - * entry, as somebody else could have populated it.. - */ + spin_lock(&mm->page_table_lock); #ifndef __ARCH_HAS_4LEVEL_HACK if (pud_present(*pud)) { pmd_free(new); @@ -2161,6 +2153,8 @@ pmd_t fastcall *__pmd_alloc(struct mm_struct *mm, pud_t *pud, unsigned long addr #endif /* __ARCH_HAS_4LEVEL_HACK */ out: + if (mm == &init_mm) /* Temporary bridging hack */ + spin_unlock(&mm->page_table_lock); return pmd_offset(pud, address); } #endif /* __PAGETABLE_PMD_FOLDED */ diff --git a/mm/vmalloc.c b/mm/vmalloc.c index 5e9120598799..54a90e83cb31 100644 --- a/mm/vmalloc.c +++ b/mm/vmalloc.c @@ -89,7 +89,7 @@ static int vmap_pte_range(pmd_t *pmd, unsigned long addr, { pte_t *pte; - pte = pte_alloc_kernel(&init_mm, pmd, addr); + pte = pte_alloc_kernel(pmd, addr); if (!pte) return -ENOMEM; do { @@ -147,14 +147,12 @@ int map_vm_area(struct vm_struct *area, pgprot_t prot, struct page ***pages) BUG_ON(addr >= end); pgd = pgd_offset_k(addr); - spin_lock(&init_mm.page_table_lock); do { next = pgd_addr_end(addr, end); err = vmap_pud_range(pgd, addr, next, prot, pages); if (err) break; } while (pgd++, addr = next, addr != end); - spin_unlock(&init_mm.page_table_lock); flush_cache_vmap((unsigned long) area->addr, end); return err; } -- cgit v1.2.3 From 1bb3630e89cb8a7b3d3807629c20c5bad88290ff Mon Sep 17 00:00:00 2001 From: Hugh Dickins Date: Sat, 29 Oct 2005 18:16:22 -0700 Subject: [PATCH] mm: ptd_alloc inline and out It seems odd to me that, whereas pud_alloc and pmd_alloc test inline, only calling out-of-line __pud_alloc __pmd_alloc if allocation needed, pte_alloc_map and pte_alloc_kernel are entirely out-of-line. Though it does add a little to kernel size, change them to macros testing inline, calling __pte_alloc or __pte_alloc_kernel to allocate out-of-line. Mark none of them as fastcalls, leave that to CONFIG_REGPARM or not. It also seems more natural for the out-of-line functions to leave the offset calculation and map to the inline, which has to do it anyway for the common case. At least mremap move wants __pte_alloc without _map. Macros rather than inline functions, certainly to avoid the header file issues which arise from CONFIG_HIGHPTE needing kmap_types.h, but also in case any architectures I haven't built would have other such problems. Signed-off-by: Hugh Dickins Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/asm-generic/4level-fixup.h | 11 ++--- include/linux/mm.h | 38 +++++++-------- mm/memory.c | 95 ++++++++++++++++---------------------- mm/mremap.c | 7 +-- 4 files changed, 62 insertions(+), 89 deletions(-) (limited to 'include/linux') diff --git a/include/asm-generic/4level-fixup.h b/include/asm-generic/4level-fixup.h index c20ec257ecc0..68c6fea994d9 100644 --- a/include/asm-generic/4level-fixup.h +++ b/include/asm-generic/4level-fixup.h @@ -10,14 +10,9 @@ #define pud_t pgd_t -#define pmd_alloc(mm, pud, address) \ -({ pmd_t *ret; \ - if (pgd_none(*pud)) \ - ret = __pmd_alloc(mm, pud, address); \ - else \ - ret = pmd_offset(pud, address); \ - ret; \ -}) +#define pmd_alloc(mm, pud, address) \ + ((unlikely(pgd_none(*(pud))) && __pmd_alloc(mm, pud, address))? \ + NULL: pmd_offset(pud, address)) #define pud_alloc(mm, pgd, address) (pgd) #define pud_offset(pgd, start) (pgd) diff --git a/include/linux/mm.h b/include/linux/mm.h index b9fa82b96d9e..22c2d6922c0e 100644 --- a/include/linux/mm.h +++ b/include/linux/mm.h @@ -704,10 +704,6 @@ static inline void unmap_shared_mapping_range(struct address_space *mapping, } extern int vmtruncate(struct inode * inode, loff_t offset); -extern pud_t *FASTCALL(__pud_alloc(struct mm_struct *mm, pgd_t *pgd, unsigned long address)); -extern pmd_t *FASTCALL(__pmd_alloc(struct mm_struct *mm, pud_t *pud, unsigned long address)); -extern pte_t *FASTCALL(pte_alloc_kernel(pmd_t *pmd, unsigned long address)); -extern pte_t *FASTCALL(pte_alloc_map(struct mm_struct *mm, pmd_t *pmd, unsigned long address)); extern int install_page(struct mm_struct *mm, struct vm_area_struct *vma, unsigned long addr, struct page *page, pgprot_t prot); extern int install_file_pte(struct mm_struct *mm, struct vm_area_struct *vma, unsigned long addr, unsigned long pgoff, pgprot_t prot); extern int __handle_mm_fault(struct mm_struct *mm,struct vm_area_struct *vma, unsigned long address, int write_access); @@ -760,32 +756,36 @@ struct shrinker; extern struct shrinker *set_shrinker(int, shrinker_t); extern void remove_shrinker(struct shrinker *shrinker); -/* - * On a two-level or three-level page table, this ends up being trivial. Thus - * the inlining and the symmetry break with pte_alloc_map() that does all - * of this out-of-line. - */ +int __pud_alloc(struct mm_struct *mm, pgd_t *pgd, unsigned long address); +int __pmd_alloc(struct mm_struct *mm, pud_t *pud, unsigned long address); +int __pte_alloc(struct mm_struct *mm, pmd_t *pmd, unsigned long address); +int __pte_alloc_kernel(pmd_t *pmd, unsigned long address); + /* * The following ifdef needed to get the 4level-fixup.h header to work. * Remove it when 4level-fixup.h has been removed. */ -#ifdef CONFIG_MMU -#ifndef __ARCH_HAS_4LEVEL_HACK +#if defined(CONFIG_MMU) && !defined(__ARCH_HAS_4LEVEL_HACK) static inline pud_t *pud_alloc(struct mm_struct *mm, pgd_t *pgd, unsigned long address) { - if (pgd_none(*pgd)) - return __pud_alloc(mm, pgd, address); - return pud_offset(pgd, address); + return (unlikely(pgd_none(*pgd)) && __pud_alloc(mm, pgd, address))? + NULL: pud_offset(pgd, address); } static inline pmd_t *pmd_alloc(struct mm_struct *mm, pud_t *pud, unsigned long address) { - if (pud_none(*pud)) - return __pmd_alloc(mm, pud, address); - return pmd_offset(pud, address); + return (unlikely(pud_none(*pud)) && __pmd_alloc(mm, pud, address))? + NULL: pmd_offset(pud, address); } -#endif -#endif /* CONFIG_MMU */ +#endif /* CONFIG_MMU && !__ARCH_HAS_4LEVEL_HACK */ + +#define pte_alloc_map(mm, pmd, address) \ + ((unlikely(!pmd_present(*(pmd))) && __pte_alloc(mm, pmd, address))? \ + NULL: pte_offset_map(pmd, address)) + +#define pte_alloc_kernel(pmd, address) \ + ((unlikely(!pmd_present(*(pmd))) && __pte_alloc_kernel(pmd, address))? \ + NULL: pte_offset_kernel(pmd, address)) extern void free_area_init(unsigned long * zones_size); extern void free_area_init_node(int nid, pg_data_t *pgdat, diff --git a/mm/memory.c b/mm/memory.c index 95a4553c75f7..4bdd1186b43b 100644 --- a/mm/memory.c +++ b/mm/memory.c @@ -280,50 +280,39 @@ void free_pgtables(struct mmu_gather **tlb, struct vm_area_struct *vma, } } -pte_t fastcall *pte_alloc_map(struct mm_struct *mm, pmd_t *pmd, - unsigned long address) +int __pte_alloc(struct mm_struct *mm, pmd_t *pmd, unsigned long address) { - if (!pmd_present(*pmd)) { - struct page *new; + struct page *new; - spin_unlock(&mm->page_table_lock); - new = pte_alloc_one(mm, address); - spin_lock(&mm->page_table_lock); - if (!new) - return NULL; - /* - * Because we dropped the lock, we should re-check the - * entry, as somebody else could have populated it.. - */ - if (pmd_present(*pmd)) { - pte_free(new); - goto out; - } + spin_unlock(&mm->page_table_lock); + new = pte_alloc_one(mm, address); + spin_lock(&mm->page_table_lock); + if (!new) + return -ENOMEM; + + if (pmd_present(*pmd)) /* Another has populated it */ + pte_free(new); + else { mm->nr_ptes++; inc_page_state(nr_page_table_pages); pmd_populate(mm, pmd, new); } -out: - return pte_offset_map(pmd, address); + return 0; } -pte_t fastcall * pte_alloc_kernel(pmd_t *pmd, unsigned long address) +int __pte_alloc_kernel(pmd_t *pmd, unsigned long address) { - if (!pmd_present(*pmd)) { - pte_t *new; - - new = pte_alloc_one_kernel(&init_mm, address); - if (!new) - return NULL; - - spin_lock(&init_mm.page_table_lock); - if (pmd_present(*pmd)) - pte_free_kernel(new); - else - pmd_populate_kernel(&init_mm, pmd, new); - spin_unlock(&init_mm.page_table_lock); - } - return pte_offset_kernel(pmd, address); + pte_t *new = pte_alloc_one_kernel(&init_mm, address); + if (!new) + return -ENOMEM; + + spin_lock(&init_mm.page_table_lock); + if (pmd_present(*pmd)) /* Another has populated it */ + pte_free_kernel(new); + else + pmd_populate_kernel(&init_mm, pmd, new); + spin_unlock(&init_mm.page_table_lock); + return 0; } static inline void add_mm_rss(struct mm_struct *mm, int file_rss, int anon_rss) @@ -2093,7 +2082,7 @@ int __handle_mm_fault(struct mm_struct *mm, struct vm_area_struct *vma, * Allocate page upper directory. * We've already handled the fast-path in-line. */ -pud_t fastcall *__pud_alloc(struct mm_struct *mm, pgd_t *pgd, unsigned long address) +int __pud_alloc(struct mm_struct *mm, pgd_t *pgd, unsigned long address) { pud_t *new; @@ -2103,19 +2092,17 @@ pud_t fastcall *__pud_alloc(struct mm_struct *mm, pgd_t *pgd, unsigned long addr if (!new) { if (mm != &init_mm) /* Temporary bridging hack */ spin_lock(&mm->page_table_lock); - return NULL; + return -ENOMEM; } spin_lock(&mm->page_table_lock); - if (pgd_present(*pgd)) { + if (pgd_present(*pgd)) /* Another has populated it */ pud_free(new); - goto out; - } - pgd_populate(mm, pgd, new); - out: + else + pgd_populate(mm, pgd, new); if (mm == &init_mm) /* Temporary bridging hack */ spin_unlock(&mm->page_table_lock); - return pud_offset(pgd, address); + return 0; } #endif /* __PAGETABLE_PUD_FOLDED */ @@ -2124,7 +2111,7 @@ pud_t fastcall *__pud_alloc(struct mm_struct *mm, pgd_t *pgd, unsigned long addr * Allocate page middle directory. * We've already handled the fast-path in-line. */ -pmd_t fastcall *__pmd_alloc(struct mm_struct *mm, pud_t *pud, unsigned long address) +int __pmd_alloc(struct mm_struct *mm, pud_t *pud, unsigned long address) { pmd_t *new; @@ -2134,28 +2121,24 @@ pmd_t fastcall *__pmd_alloc(struct mm_struct *mm, pud_t *pud, unsigned long addr if (!new) { if (mm != &init_mm) /* Temporary bridging hack */ spin_lock(&mm->page_table_lock); - return NULL; + return -ENOMEM; } spin_lock(&mm->page_table_lock); #ifndef __ARCH_HAS_4LEVEL_HACK - if (pud_present(*pud)) { + if (pud_present(*pud)) /* Another has populated it */ pmd_free(new); - goto out; - } - pud_populate(mm, pud, new); + else + pud_populate(mm, pud, new); #else - if (pgd_present(*pud)) { + if (pgd_present(*pud)) /* Another has populated it */ pmd_free(new); - goto out; - } - pgd_populate(mm, pud, new); + else + pgd_populate(mm, pud, new); #endif /* __ARCH_HAS_4LEVEL_HACK */ - - out: if (mm == &init_mm) /* Temporary bridging hack */ spin_unlock(&mm->page_table_lock); - return pmd_offset(pud, address); + return 0; } #endif /* __PAGETABLE_PMD_FOLDED */ diff --git a/mm/mremap.c b/mm/mremap.c index ccf456477020..616facc3d28a 100644 --- a/mm/mremap.c +++ b/mm/mremap.c @@ -51,7 +51,6 @@ static pmd_t *alloc_new_pmd(struct mm_struct *mm, unsigned long addr) pgd_t *pgd; pud_t *pud; pmd_t *pmd = NULL; - pte_t *pte; /* * We do need page_table_lock: because allocators expect that. @@ -66,12 +65,8 @@ static pmd_t *alloc_new_pmd(struct mm_struct *mm, unsigned long addr) if (!pmd) goto out; - pte = pte_alloc_map(mm, pmd, addr); - if (!pte) { + if (!pmd_present(*pmd) && __pte_alloc(mm, pmd, addr)) pmd = NULL; - goto out; - } - pte_unmap(pte); out: spin_unlock(&mm->page_table_lock); return pmd; -- cgit v1.2.3 From c74df32c724a1652ad8399b4891bb02c9d43743a Mon Sep 17 00:00:00 2001 From: Hugh Dickins Date: Sat, 29 Oct 2005 18:16:23 -0700 Subject: [PATCH] mm: ptd_alloc take ptlock Second step in pushing down the page_table_lock. Remove the temporary bridging hack from __pud_alloc, __pmd_alloc, __pte_alloc: expect callers not to hold page_table_lock, whether it's on init_mm or a user mm; take page_table_lock internally to check if a racing task already allocated. Convert their callers from common code. But avoid coming back to change them again later: instead of moving the spin_lock(&mm->page_table_lock) down, switch over to new macros pte_alloc_map_lock and pte_unmap_unlock, which encapsulate the mapping+locking and unlocking+unmapping together, and in the end may use alternatives to the mm page_table_lock itself. These callers all hold mmap_sem (some exclusively, some not), so at no level can a page table be whipped away from beneath them; and pte_alloc uses the "atomic" pmd_present to test whether it needs to allocate. It appears that on all arches we can safely descend without page_table_lock. Signed-off-by: Hugh Dickins Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/exec.c | 14 +++----- include/linux/mm.h | 18 ++++++++++ kernel/fork.c | 2 -- mm/fremap.c | 48 ++++++++++--------------- mm/hugetlb.c | 12 ++++--- mm/memory.c | 104 +++++++++++++++++------------------------------------ mm/mremap.c | 27 +++++--------- 7 files changed, 90 insertions(+), 135 deletions(-) (limited to 'include/linux') diff --git a/fs/exec.c b/fs/exec.c index 9bb55c8cf224..ba73797eb4cb 100644 --- a/fs/exec.c +++ b/fs/exec.c @@ -309,25 +309,24 @@ void install_arg_page(struct vm_area_struct *vma, pud_t * pud; pmd_t * pmd; pte_t * pte; + spinlock_t *ptl; if (unlikely(anon_vma_prepare(vma))) - goto out_sig; + goto out; flush_dcache_page(page); pgd = pgd_offset(mm, address); - - spin_lock(&mm->page_table_lock); pud = pud_alloc(mm, pgd, address); if (!pud) goto out; pmd = pmd_alloc(mm, pud, address); if (!pmd) goto out; - pte = pte_alloc_map(mm, pmd, address); + pte = pte_alloc_map_lock(mm, pmd, address, &ptl); if (!pte) goto out; if (!pte_none(*pte)) { - pte_unmap(pte); + pte_unmap_unlock(pte, ptl); goto out; } inc_mm_counter(mm, anon_rss); @@ -335,14 +334,11 @@ void install_arg_page(struct vm_area_struct *vma, set_pte_at(mm, address, pte, pte_mkdirty(pte_mkwrite(mk_pte( page, vma->vm_page_prot)))); page_add_anon_rmap(page, vma, address); - pte_unmap(pte); - spin_unlock(&mm->page_table_lock); + pte_unmap_unlock(pte, ptl); /* no need for flush_tlb */ return; out: - spin_unlock(&mm->page_table_lock); -out_sig: __free_page(page); force_sig(SIGKILL, current); } diff --git a/include/linux/mm.h b/include/linux/mm.h index 22c2d6922c0e..d4c3512e7db4 100644 --- a/include/linux/mm.h +++ b/include/linux/mm.h @@ -779,10 +779,28 @@ static inline pmd_t *pmd_alloc(struct mm_struct *mm, pud_t *pud, unsigned long a } #endif /* CONFIG_MMU && !__ARCH_HAS_4LEVEL_HACK */ +#define pte_offset_map_lock(mm, pmd, address, ptlp) \ +({ \ + spinlock_t *__ptl = &(mm)->page_table_lock; \ + pte_t *__pte = pte_offset_map(pmd, address); \ + *(ptlp) = __ptl; \ + spin_lock(__ptl); \ + __pte; \ +}) + +#define pte_unmap_unlock(pte, ptl) do { \ + spin_unlock(ptl); \ + pte_unmap(pte); \ +} while (0) + #define pte_alloc_map(mm, pmd, address) \ ((unlikely(!pmd_present(*(pmd))) && __pte_alloc(mm, pmd, address))? \ NULL: pte_offset_map(pmd, address)) +#define pte_alloc_map_lock(mm, pmd, address, ptlp) \ + ((unlikely(!pmd_present(*(pmd))) && __pte_alloc(mm, pmd, address))? \ + NULL: pte_offset_map_lock(mm, pmd, address, ptlp)) + #define pte_alloc_kernel(pmd, address) \ ((unlikely(!pmd_present(*(pmd))) && __pte_alloc_kernel(pmd, address))? \ NULL: pte_offset_kernel(pmd, address)) diff --git a/kernel/fork.c b/kernel/fork.c index 2a587b3224e3..8a069612eac3 100644 --- a/kernel/fork.c +++ b/kernel/fork.c @@ -255,7 +255,6 @@ static inline int dup_mmap(struct mm_struct *mm, struct mm_struct *oldmm) /* * Link in the new vma and copy the page table entries. */ - spin_lock(&mm->page_table_lock); *pprev = tmp; pprev = &tmp->vm_next; @@ -265,7 +264,6 @@ static inline int dup_mmap(struct mm_struct *mm, struct mm_struct *oldmm) mm->map_count++; retval = copy_page_range(mm, oldmm, tmp); - spin_unlock(&mm->page_table_lock); if (tmp->vm_ops && tmp->vm_ops->open) tmp->vm_ops->open(tmp); diff --git a/mm/fremap.c b/mm/fremap.c index 49719a35769a..d862be3bc3e3 100644 --- a/mm/fremap.c +++ b/mm/fremap.c @@ -63,23 +63,20 @@ int install_page(struct mm_struct *mm, struct vm_area_struct *vma, pud_t *pud; pgd_t *pgd; pte_t pte_val; + spinlock_t *ptl; BUG_ON(vma->vm_flags & VM_RESERVED); pgd = pgd_offset(mm, addr); - spin_lock(&mm->page_table_lock); - pud = pud_alloc(mm, pgd, addr); if (!pud) - goto err_unlock; - + goto out; pmd = pmd_alloc(mm, pud, addr); if (!pmd) - goto err_unlock; - - pte = pte_alloc_map(mm, pmd, addr); + goto out; + pte = pte_alloc_map_lock(mm, pmd, addr, &ptl); if (!pte) - goto err_unlock; + goto out; /* * This page may have been truncated. Tell the @@ -89,10 +86,10 @@ int install_page(struct mm_struct *mm, struct vm_area_struct *vma, inode = vma->vm_file->f_mapping->host; size = (i_size_read(inode) + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT; if (!page->mapping || page->index >= size) - goto err_unlock; + goto unlock; err = -ENOMEM; if (page_mapcount(page) > INT_MAX/2) - goto err_unlock; + goto unlock; if (pte_none(*pte) || !zap_pte(mm, vma, addr, pte)) inc_mm_counter(mm, file_rss); @@ -101,17 +98,15 @@ int install_page(struct mm_struct *mm, struct vm_area_struct *vma, set_pte_at(mm, addr, pte, mk_pte(page, prot)); page_add_file_rmap(page); pte_val = *pte; - pte_unmap(pte); update_mmu_cache(vma, addr, pte_val); - err = 0; -err_unlock: - spin_unlock(&mm->page_table_lock); +unlock: + pte_unmap_unlock(pte, ptl); +out: return err; } EXPORT_SYMBOL(install_page); - /* * Install a file pte to a given virtual memory address, release any * previously existing mapping. @@ -125,23 +120,20 @@ int install_file_pte(struct mm_struct *mm, struct vm_area_struct *vma, pud_t *pud; pgd_t *pgd; pte_t pte_val; + spinlock_t *ptl; BUG_ON(vma->vm_flags & VM_RESERVED); pgd = pgd_offset(mm, addr); - spin_lock(&mm->page_table_lock); - pud = pud_alloc(mm, pgd, addr); if (!pud) - goto err_unlock; - + goto out; pmd = pmd_alloc(mm, pud, addr); if (!pmd) - goto err_unlock; - - pte = pte_alloc_map(mm, pmd, addr); + goto out; + pte = pte_alloc_map_lock(mm, pmd, addr, &ptl); if (!pte) - goto err_unlock; + goto out; if (!pte_none(*pte) && zap_pte(mm, vma, addr, pte)) { update_hiwater_rss(mm); @@ -150,17 +142,13 @@ int install_file_pte(struct mm_struct *mm, struct vm_area_struct *vma, set_pte_at(mm, addr, pte, pgoff_to_pte(pgoff)); pte_val = *pte; - pte_unmap(pte); update_mmu_cache(vma, addr, pte_val); - spin_unlock(&mm->page_table_lock); - return 0; - -err_unlock: - spin_unlock(&mm->page_table_lock); + pte_unmap_unlock(pte, ptl); + err = 0; +out: return err; } - /*** * sys_remap_file_pages - remap arbitrary pages of a shared backing store * file within an existing vma. diff --git a/mm/hugetlb.c b/mm/hugetlb.c index ac5f044bf514..ea0826ff2663 100644 --- a/mm/hugetlb.c +++ b/mm/hugetlb.c @@ -277,12 +277,15 @@ int copy_hugetlb_page_range(struct mm_struct *dst, struct mm_struct *src, unsigned long addr; for (addr = vma->vm_start; addr < vma->vm_end; addr += HPAGE_SIZE) { + src_pte = huge_pte_offset(src, addr); + if (!src_pte) + continue; dst_pte = huge_pte_alloc(dst, addr); if (!dst_pte) goto nomem; + spin_lock(&dst->page_table_lock); spin_lock(&src->page_table_lock); - src_pte = huge_pte_offset(src, addr); - if (src_pte && !pte_none(*src_pte)) { + if (!pte_none(*src_pte)) { entry = *src_pte; ptepage = pte_page(entry); get_page(ptepage); @@ -290,6 +293,7 @@ int copy_hugetlb_page_range(struct mm_struct *dst, struct mm_struct *src, set_huge_pte_at(dst, addr, dst_pte, entry); } spin_unlock(&src->page_table_lock); + spin_unlock(&dst->page_table_lock); } return 0; @@ -354,7 +358,6 @@ int hugetlb_prefault(struct address_space *mapping, struct vm_area_struct *vma) hugetlb_prefault_arch_hook(mm); - spin_lock(&mm->page_table_lock); for (addr = vma->vm_start; addr < vma->vm_end; addr += HPAGE_SIZE) { unsigned long idx; pte_t *pte = huge_pte_alloc(mm, addr); @@ -389,11 +392,12 @@ int hugetlb_prefault(struct address_space *mapping, struct vm_area_struct *vma) goto out; } } + spin_lock(&mm->page_table_lock); add_mm_counter(mm, file_rss, HPAGE_SIZE / PAGE_SIZE); set_huge_pte_at(mm, addr, pte, make_huge_pte(vma, page)); + spin_unlock(&mm->page_table_lock); } out: - spin_unlock(&mm->page_table_lock); return ret; } diff --git a/mm/memory.c b/mm/memory.c index 4bdd1186b43b..a40e4b1cee4f 100644 --- a/mm/memory.c +++ b/mm/memory.c @@ -282,14 +282,11 @@ void free_pgtables(struct mmu_gather **tlb, struct vm_area_struct *vma, int __pte_alloc(struct mm_struct *mm, pmd_t *pmd, unsigned long address) { - struct page *new; - - spin_unlock(&mm->page_table_lock); - new = pte_alloc_one(mm, address); - spin_lock(&mm->page_table_lock); + struct page *new = pte_alloc_one(mm, address); if (!new) return -ENOMEM; + spin_lock(&mm->page_table_lock); if (pmd_present(*pmd)) /* Another has populated it */ pte_free(new); else { @@ -297,6 +294,7 @@ int __pte_alloc(struct mm_struct *mm, pmd_t *pmd, unsigned long address) inc_page_state(nr_page_table_pages); pmd_populate(mm, pmd, new); } + spin_unlock(&mm->page_table_lock); return 0; } @@ -344,9 +342,6 @@ void print_bad_pte(struct vm_area_struct *vma, pte_t pte, unsigned long vaddr) * copy one vm_area from one task to the other. Assumes the page tables * already present in the new task to be cleared in the whole range * covered by this vma. - * - * dst->page_table_lock is held on entry and exit, - * but may be dropped within p[mg]d_alloc() and pte_alloc_map(). */ static inline void @@ -419,17 +414,19 @@ static int copy_pte_range(struct mm_struct *dst_mm, struct mm_struct *src_mm, unsigned long addr, unsigned long end) { pte_t *src_pte, *dst_pte; + spinlock_t *src_ptl, *dst_ptl; int progress = 0; int rss[2]; again: rss[1] = rss[0] = 0; - dst_pte = pte_alloc_map(dst_mm, dst_pmd, addr); + dst_pte = pte_alloc_map_lock(dst_mm, dst_pmd, addr, &dst_ptl); if (!dst_pte) return -ENOMEM; src_pte = pte_offset_map_nested(src_pmd, addr); + src_ptl = &src_mm->page_table_lock; + spin_lock(src_ptl); - spin_lock(&src_mm->page_table_lock); do { /* * We are holding two locks at this point - either of them @@ -438,8 +435,8 @@ again: if (progress >= 32) { progress = 0; if (need_resched() || - need_lockbreak(&src_mm->page_table_lock) || - need_lockbreak(&dst_mm->page_table_lock)) + need_lockbreak(src_ptl) || + need_lockbreak(dst_ptl)) break; } if (pte_none(*src_pte)) { @@ -449,12 +446,12 @@ again: copy_one_pte(dst_mm, src_mm, dst_pte, src_pte, vma, addr, rss); progress += 8; } while (dst_pte++, src_pte++, addr += PAGE_SIZE, addr != end); - spin_unlock(&src_mm->page_table_lock); + spin_unlock(src_ptl); pte_unmap_nested(src_pte - 1); - pte_unmap(dst_pte - 1); add_mm_rss(dst_mm, rss[0], rss[1]); - cond_resched_lock(&dst_mm->page_table_lock); + pte_unmap_unlock(dst_pte - 1, dst_ptl); + cond_resched(); if (addr != end) goto again; return 0; @@ -1049,8 +1046,9 @@ static int zeromap_pte_range(struct mm_struct *mm, pmd_t *pmd, unsigned long addr, unsigned long end, pgprot_t prot) { pte_t *pte; + spinlock_t *ptl; - pte = pte_alloc_map(mm, pmd, addr); + pte = pte_alloc_map_lock(mm, pmd, addr, &ptl); if (!pte) return -ENOMEM; do { @@ -1062,7 +1060,7 @@ static int zeromap_pte_range(struct mm_struct *mm, pmd_t *pmd, BUG_ON(!pte_none(*pte)); set_pte_at(mm, addr, pte, zero_pte); } while (pte++, addr += PAGE_SIZE, addr != end); - pte_unmap(pte - 1); + pte_unmap_unlock(pte - 1, ptl); return 0; } @@ -1112,14 +1110,12 @@ int zeromap_page_range(struct vm_area_struct *vma, BUG_ON(addr >= end); pgd = pgd_offset(mm, addr); flush_cache_range(vma, addr, end); - spin_lock(&mm->page_table_lock); do { next = pgd_addr_end(addr, end); err = zeromap_pud_range(mm, pgd, addr, next, prot); if (err) break; } while (pgd++, addr = next, addr != end); - spin_unlock(&mm->page_table_lock); return err; } @@ -1133,8 +1129,9 @@ static int remap_pte_range(struct mm_struct *mm, pmd_t *pmd, unsigned long pfn, pgprot_t prot) { pte_t *pte; + spinlock_t *ptl; - pte = pte_alloc_map(mm, pmd, addr); + pte = pte_alloc_map_lock(mm, pmd, addr, &ptl); if (!pte) return -ENOMEM; do { @@ -1142,7 +1139,7 @@ static int remap_pte_range(struct mm_struct *mm, pmd_t *pmd, set_pte_at(mm, addr, pte, pfn_pte(pfn, prot)); pfn++; } while (pte++, addr += PAGE_SIZE, addr != end); - pte_unmap(pte - 1); + pte_unmap_unlock(pte - 1, ptl); return 0; } @@ -1210,7 +1207,6 @@ int remap_pfn_range(struct vm_area_struct *vma, unsigned long addr, pfn -= addr >> PAGE_SHIFT; pgd = pgd_offset(mm, addr); flush_cache_range(vma, addr, end); - spin_lock(&mm->page_table_lock); do { next = pgd_addr_end(addr, end); err = remap_pud_range(mm, pgd, addr, next, @@ -1218,7 +1214,6 @@ int remap_pfn_range(struct vm_area_struct *vma, unsigned long addr, if (err) break; } while (pgd++, addr = next, addr != end); - spin_unlock(&mm->page_table_lock); return err; } EXPORT_SYMBOL(remap_pfn_range); @@ -1985,17 +1980,9 @@ static int do_file_page(struct mm_struct *mm, struct vm_area_struct *vma, * with external mmu caches can use to update those (ie the Sparc or * PowerPC hashed page tables that act as extended TLBs). * - * Note the "page_table_lock". It is to protect against kswapd removing - * pages from under us. Note that kswapd only ever _removes_ pages, never - * adds them. As such, once we have noticed that the page is not present, - * we can drop the lock early. - * - * The adding of pages is protected by the MM semaphore (which we hold), - * so we don't need to worry about a page being suddenly been added into - * our VM. - * - * We enter with the pagetable spinlock held, we are supposed to - * release it when done. + * We enter with non-exclusive mmap_sem (to exclude vma changes, + * but allow concurrent faults), and pte mapped but not yet locked. + * We return with mmap_sem still held, but pte unmapped and unlocked. */ static inline int handle_pte_fault(struct mm_struct *mm, struct vm_area_struct *vma, unsigned long address, @@ -2003,6 +1990,7 @@ static inline int handle_pte_fault(struct mm_struct *mm, { pte_t entry; + spin_lock(&mm->page_table_lock); entry = *pte; if (!pte_present(entry)) { if (pte_none(entry)) { @@ -2051,30 +2039,18 @@ int __handle_mm_fault(struct mm_struct *mm, struct vm_area_struct *vma, if (unlikely(is_vm_hugetlb_page(vma))) return hugetlb_fault(mm, vma, address, write_access); - /* - * We need the page table lock to synchronize with kswapd - * and the SMP-safe atomic PTE updates. - */ pgd = pgd_offset(mm, address); - spin_lock(&mm->page_table_lock); - pud = pud_alloc(mm, pgd, address); if (!pud) - goto oom; - + return VM_FAULT_OOM; pmd = pmd_alloc(mm, pud, address); if (!pmd) - goto oom; - + return VM_FAULT_OOM; pte = pte_alloc_map(mm, pmd, address); if (!pte) - goto oom; - - return handle_pte_fault(mm, vma, address, pte, pmd, write_access); + return VM_FAULT_OOM; - oom: - spin_unlock(&mm->page_table_lock); - return VM_FAULT_OOM; + return handle_pte_fault(mm, vma, address, pte, pmd, write_access); } #ifndef __PAGETABLE_PUD_FOLDED @@ -2084,24 +2060,16 @@ int __handle_mm_fault(struct mm_struct *mm, struct vm_area_struct *vma, */ int __pud_alloc(struct mm_struct *mm, pgd_t *pgd, unsigned long address) { - pud_t *new; - - if (mm != &init_mm) /* Temporary bridging hack */ - spin_unlock(&mm->page_table_lock); - new = pud_alloc_one(mm, address); - if (!new) { - if (mm != &init_mm) /* Temporary bridging hack */ - spin_lock(&mm->page_table_lock); + pud_t *new = pud_alloc_one(mm, address); + if (!new) return -ENOMEM; - } spin_lock(&mm->page_table_lock); if (pgd_present(*pgd)) /* Another has populated it */ pud_free(new); else pgd_populate(mm, pgd, new); - if (mm == &init_mm) /* Temporary bridging hack */ - spin_unlock(&mm->page_table_lock); + spin_unlock(&mm->page_table_lock); return 0; } #endif /* __PAGETABLE_PUD_FOLDED */ @@ -2113,16 +2081,9 @@ int __pud_alloc(struct mm_struct *mm, pgd_t *pgd, unsigned long address) */ int __pmd_alloc(struct mm_struct *mm, pud_t *pud, unsigned long address) { - pmd_t *new; - - if (mm != &init_mm) /* Temporary bridging hack */ - spin_unlock(&mm->page_table_lock); - new = pmd_alloc_one(mm, address); - if (!new) { - if (mm != &init_mm) /* Temporary bridging hack */ - spin_lock(&mm->page_table_lock); + pmd_t *new = pmd_alloc_one(mm, address); + if (!new) return -ENOMEM; - } spin_lock(&mm->page_table_lock); #ifndef __ARCH_HAS_4LEVEL_HACK @@ -2136,8 +2097,7 @@ int __pmd_alloc(struct mm_struct *mm, pud_t *pud, unsigned long address) else pgd_populate(mm, pud, new); #endif /* __ARCH_HAS_4LEVEL_HACK */ - if (mm == &init_mm) /* Temporary bridging hack */ - spin_unlock(&mm->page_table_lock); + spin_unlock(&mm->page_table_lock); return 0; } #endif /* __PAGETABLE_PMD_FOLDED */ diff --git a/mm/mremap.c b/mm/mremap.c index 616facc3d28a..8de77b632a20 100644 --- a/mm/mremap.c +++ b/mm/mremap.c @@ -28,9 +28,6 @@ static pmd_t *get_old_pmd(struct mm_struct *mm, unsigned long addr) pud_t *pud; pmd_t *pmd; - /* - * We don't need page_table_lock: we have mmap_sem exclusively. - */ pgd = pgd_offset(mm, addr); if (pgd_none_or_clear_bad(pgd)) return NULL; @@ -50,25 +47,20 @@ static pmd_t *alloc_new_pmd(struct mm_struct *mm, unsigned long addr) { pgd_t *pgd; pud_t *pud; - pmd_t *pmd = NULL; + pmd_t *pmd; - /* - * We do need page_table_lock: because allocators expect that. - */ - spin_lock(&mm->page_table_lock); pgd = pgd_offset(mm, addr); pud = pud_alloc(mm, pgd, addr); if (!pud) - goto out; + return NULL; pmd = pmd_alloc(mm, pud, addr); if (!pmd) - goto out; + return NULL; if (!pmd_present(*pmd) && __pte_alloc(mm, pmd, addr)) - pmd = NULL; -out: - spin_unlock(&mm->page_table_lock); + return NULL; + return pmd; } @@ -80,6 +72,7 @@ static void move_ptes(struct vm_area_struct *vma, pmd_t *old_pmd, struct address_space *mapping = NULL; struct mm_struct *mm = vma->vm_mm; pte_t *old_pte, *new_pte, pte; + spinlock_t *old_ptl; if (vma->vm_file) { /* @@ -95,9 +88,8 @@ static void move_ptes(struct vm_area_struct *vma, pmd_t *old_pmd, new_vma->vm_truncate_count = 0; } - spin_lock(&mm->page_table_lock); - old_pte = pte_offset_map(old_pmd, old_addr); - new_pte = pte_offset_map_nested(new_pmd, new_addr); + old_pte = pte_offset_map_lock(mm, old_pmd, old_addr, &old_ptl); + new_pte = pte_offset_map_nested(new_pmd, new_addr); for (; old_addr < old_end; old_pte++, old_addr += PAGE_SIZE, new_pte++, new_addr += PAGE_SIZE) { @@ -110,8 +102,7 @@ static void move_ptes(struct vm_area_struct *vma, pmd_t *old_pmd, } pte_unmap_nested(new_pte - 1); - pte_unmap(old_pte - 1); - spin_unlock(&mm->page_table_lock); + pte_unmap_unlock(old_pte - 1, old_ptl); if (mapping) spin_unlock(&mapping->i_mmap_lock); } -- cgit v1.2.3 From 508034a32b819a2d40aa7ac0dbc8cd2e044c2de6 Mon Sep 17 00:00:00 2001 From: Hugh Dickins Date: Sat, 29 Oct 2005 18:16:30 -0700 Subject: [PATCH] mm: unmap_vmas with inner ptlock Remove the page_table_lock from around the calls to unmap_vmas, and replace the pte_offset_map in zap_pte_range by pte_offset_map_lock: all callers are now safe to descend without page_table_lock. Don't attempt fancy locking for hugepages, just take page_table_lock in unmap_hugepage_range. Which makes zap_hugepage_range, and the hugetlb test in zap_page_range, redundant: unmap_vmas calls unmap_hugepage_range anyway. Nor does unmap_vmas have much use for its mm arg now. The tlb_start_vma and tlb_end_vma in unmap_page_range are now called without page_table_lock: if they're implemented at all, they typically come down to flush_cache_range (usually done outside page_table_lock) and flush_tlb_range (which we already audited for the mprotect case). Signed-off-by: Hugh Dickins Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/hugetlbfs/inode.c | 10 +++------- include/linux/hugetlb.h | 2 -- include/linux/mm.h | 2 +- mm/hugetlb.c | 12 +++--------- mm/memory.c | 41 ++++++++++++----------------------------- mm/mmap.c | 8 ++------ 6 files changed, 21 insertions(+), 54 deletions(-) (limited to 'include/linux') diff --git a/fs/hugetlbfs/inode.c b/fs/hugetlbfs/inode.c index 3a9b6d179cbd..a826a8add5e3 100644 --- a/fs/hugetlbfs/inode.c +++ b/fs/hugetlbfs/inode.c @@ -92,7 +92,7 @@ out: } /* - * Called under down_write(mmap_sem), page_table_lock is not held + * Called under down_write(mmap_sem). */ #ifdef HAVE_ARCH_HUGETLB_UNMAPPED_AREA @@ -308,7 +308,6 @@ hugetlb_vmtruncate_list(struct prio_tree_root *root, unsigned long h_pgoff) vma_prio_tree_foreach(vma, &iter, root, h_pgoff, ULONG_MAX) { unsigned long h_vm_pgoff; - unsigned long v_length; unsigned long v_offset; h_vm_pgoff = vma->vm_pgoff >> (HPAGE_SHIFT - PAGE_SHIFT); @@ -319,11 +318,8 @@ hugetlb_vmtruncate_list(struct prio_tree_root *root, unsigned long h_pgoff) if (h_vm_pgoff >= h_pgoff) v_offset = 0; - v_length = vma->vm_end - vma->vm_start; - - zap_hugepage_range(vma, - vma->vm_start + v_offset, - v_length - v_offset); + unmap_hugepage_range(vma, + vma->vm_start + v_offset, vma->vm_end); } } diff --git a/include/linux/hugetlb.h b/include/linux/hugetlb.h index d664330d900e..0cea162b08c0 100644 --- a/include/linux/hugetlb.h +++ b/include/linux/hugetlb.h @@ -16,7 +16,6 @@ static inline int is_vm_hugetlb_page(struct vm_area_struct *vma) int hugetlb_sysctl_handler(struct ctl_table *, int, struct file *, void __user *, size_t *, loff_t *); int copy_hugetlb_page_range(struct mm_struct *, struct mm_struct *, struct vm_area_struct *); int follow_hugetlb_page(struct mm_struct *, struct vm_area_struct *, struct page **, struct vm_area_struct **, unsigned long *, int *, int); -void zap_hugepage_range(struct vm_area_struct *, unsigned long, unsigned long); void unmap_hugepage_range(struct vm_area_struct *, unsigned long, unsigned long); int hugetlb_prefault(struct address_space *, struct vm_area_struct *); int hugetlb_report_meminfo(char *); @@ -87,7 +86,6 @@ static inline unsigned long hugetlb_total_pages(void) #define follow_huge_addr(mm, addr, write) ERR_PTR(-EINVAL) #define copy_hugetlb_page_range(src, dst, vma) ({ BUG(); 0; }) #define hugetlb_prefault(mapping, vma) ({ BUG(); 0; }) -#define zap_hugepage_range(vma, start, len) BUG() #define unmap_hugepage_range(vma, start, end) BUG() #define is_hugepage_mem_enough(size) 0 #define hugetlb_report_meminfo(buf) 0 diff --git a/include/linux/mm.h b/include/linux/mm.h index d4c3512e7db4..972e2ce8e07c 100644 --- a/include/linux/mm.h +++ b/include/linux/mm.h @@ -682,7 +682,7 @@ struct zap_details { unsigned long zap_page_range(struct vm_area_struct *vma, unsigned long address, unsigned long size, struct zap_details *); -unsigned long unmap_vmas(struct mmu_gather **tlb, struct mm_struct *mm, +unsigned long unmap_vmas(struct mmu_gather **tlb, struct vm_area_struct *start_vma, unsigned long start_addr, unsigned long end_addr, unsigned long *nr_accounted, struct zap_details *); diff --git a/mm/hugetlb.c b/mm/hugetlb.c index ea0826ff2663..f29b7dc02c39 100644 --- a/mm/hugetlb.c +++ b/mm/hugetlb.c @@ -314,6 +314,8 @@ void unmap_hugepage_range(struct vm_area_struct *vma, unsigned long start, BUG_ON(start & ~HPAGE_MASK); BUG_ON(end & ~HPAGE_MASK); + spin_lock(&mm->page_table_lock); + /* Update high watermark before we lower rss */ update_hiwater_rss(mm); @@ -333,17 +335,9 @@ void unmap_hugepage_range(struct vm_area_struct *vma, unsigned long start, put_page(page); add_mm_counter(mm, file_rss, (int) -(HPAGE_SIZE / PAGE_SIZE)); } - flush_tlb_range(vma, start, end); -} -void zap_hugepage_range(struct vm_area_struct *vma, - unsigned long start, unsigned long length) -{ - struct mm_struct *mm = vma->vm_mm; - - spin_lock(&mm->page_table_lock); - unmap_hugepage_range(vma, start, start + length); spin_unlock(&mm->page_table_lock); + flush_tlb_range(vma, start, end); } int hugetlb_prefault(struct address_space *mapping, struct vm_area_struct *vma) diff --git a/mm/memory.c b/mm/memory.c index 4ea89a2e3a83..622a4ef5409f 100644 --- a/mm/memory.c +++ b/mm/memory.c @@ -551,10 +551,11 @@ static void zap_pte_range(struct mmu_gather *tlb, { struct mm_struct *mm = tlb->mm; pte_t *pte; + spinlock_t *ptl; int file_rss = 0; int anon_rss = 0; - pte = pte_offset_map(pmd, addr); + pte = pte_offset_map_lock(mm, pmd, addr, &ptl); do { pte_t ptent = *pte; if (pte_none(ptent)) @@ -621,7 +622,7 @@ static void zap_pte_range(struct mmu_gather *tlb, } while (pte++, addr += PAGE_SIZE, addr != end); add_mm_rss(mm, file_rss, anon_rss); - pte_unmap(pte - 1); + pte_unmap_unlock(pte - 1, ptl); } static inline void zap_pmd_range(struct mmu_gather *tlb, @@ -690,7 +691,6 @@ static void unmap_page_range(struct mmu_gather *tlb, struct vm_area_struct *vma, /** * unmap_vmas - unmap a range of memory covered by a list of vma's * @tlbp: address of the caller's struct mmu_gather - * @mm: the controlling mm_struct * @vma: the starting vma * @start_addr: virtual address at which to start unmapping * @end_addr: virtual address at which to end unmapping @@ -699,10 +699,10 @@ static void unmap_page_range(struct mmu_gather *tlb, struct vm_area_struct *vma, * * Returns the end address of the unmapping (restart addr if interrupted). * - * Unmap all pages in the vma list. Called under page_table_lock. + * Unmap all pages in the vma list. * - * We aim to not hold page_table_lock for too long (for scheduling latency - * reasons). So zap pages in ZAP_BLOCK_SIZE bytecounts. This means we need to + * We aim to not hold locks for too long (for scheduling latency reasons). + * So zap pages in ZAP_BLOCK_SIZE bytecounts. This means we need to * return the ending mmu_gather to the caller. * * Only addresses between `start' and `end' will be unmapped. @@ -714,7 +714,7 @@ static void unmap_page_range(struct mmu_gather *tlb, struct vm_area_struct *vma, * ensure that any thus-far unmapped pages are flushed before unmap_vmas() * drops the lock and schedules. */ -unsigned long unmap_vmas(struct mmu_gather **tlbp, struct mm_struct *mm, +unsigned long unmap_vmas(struct mmu_gather **tlbp, struct vm_area_struct *vma, unsigned long start_addr, unsigned long end_addr, unsigned long *nr_accounted, struct zap_details *details) @@ -764,19 +764,15 @@ unsigned long unmap_vmas(struct mmu_gather **tlbp, struct mm_struct *mm, tlb_finish_mmu(*tlbp, tlb_start, start); if (need_resched() || - need_lockbreak(&mm->page_table_lock) || (i_mmap_lock && need_lockbreak(i_mmap_lock))) { if (i_mmap_lock) { - /* must reset count of rss freed */ - *tlbp = tlb_gather_mmu(mm, fullmm); + *tlbp = NULL; goto out; } - spin_unlock(&mm->page_table_lock); cond_resched(); - spin_lock(&mm->page_table_lock); } - *tlbp = tlb_gather_mmu(mm, fullmm); + *tlbp = tlb_gather_mmu(vma->vm_mm, fullmm); tlb_start_valid = 0; zap_bytes = ZAP_BLOCK_SIZE; } @@ -800,18 +796,12 @@ unsigned long zap_page_range(struct vm_area_struct *vma, unsigned long address, unsigned long end = address + size; unsigned long nr_accounted = 0; - if (is_vm_hugetlb_page(vma)) { - zap_hugepage_range(vma, address, size); - return end; - } - lru_add_drain(); tlb = tlb_gather_mmu(mm, 0); update_hiwater_rss(mm); - spin_lock(&mm->page_table_lock); - end = unmap_vmas(&tlb, mm, vma, address, end, &nr_accounted, details); - spin_unlock(&mm->page_table_lock); - tlb_finish_mmu(tlb, address, end); + end = unmap_vmas(&tlb, vma, address, end, &nr_accounted, details); + if (tlb) + tlb_finish_mmu(tlb, address, end); return end; } @@ -1434,13 +1424,6 @@ again: restart_addr = zap_page_range(vma, start_addr, end_addr - start_addr, details); - - /* - * We cannot rely on the break test in unmap_vmas: - * on the one hand, we don't want to restart our loop - * just because that broke out for the page_table_lock; - * on the other hand, it does no test when vma is small. - */ need_break = need_resched() || need_lockbreak(details->i_mmap_lock); diff --git a/mm/mmap.c b/mm/mmap.c index fa35323a3c5b..5ecc2cf3e1d7 100644 --- a/mm/mmap.c +++ b/mm/mmap.c @@ -1673,9 +1673,7 @@ static void unmap_region(struct mm_struct *mm, lru_add_drain(); tlb = tlb_gather_mmu(mm, 0); update_hiwater_rss(mm); - spin_lock(&mm->page_table_lock); - unmap_vmas(&tlb, mm, vma, start, end, &nr_accounted, NULL); - spin_unlock(&mm->page_table_lock); + unmap_vmas(&tlb, vma, start, end, &nr_accounted, NULL); vm_unacct_memory(nr_accounted); free_pgtables(&tlb, vma, prev? prev->vm_end: FIRST_USER_ADDRESS, next? next->vm_start: 0); @@ -1958,9 +1956,7 @@ void exit_mmap(struct mm_struct *mm) tlb = tlb_gather_mmu(mm, 1); /* Don't update_hiwater_rss(mm) here, do_exit already did */ /* Use -1 here to ensure all VMAs in the mm are unmapped */ - spin_lock(&mm->page_table_lock); - end = unmap_vmas(&tlb, mm, vma, 0, -1, &nr_accounted, NULL); - spin_unlock(&mm->page_table_lock); + end = unmap_vmas(&tlb, vma, 0, -1, &nr_accounted, NULL); vm_unacct_memory(nr_accounted); free_pgtables(&tlb, vma, FIRST_USER_ADDRESS, 0); tlb_finish_mmu(tlb, 0, end); -- cgit v1.2.3 From c0718806cf955d5eb51ea77bffb5b21d9bba4972 Mon Sep 17 00:00:00 2001 From: Hugh Dickins Date: Sat, 29 Oct 2005 18:16:31 -0700 Subject: [PATCH] mm: rmap with inner ptlock rmap's page_check_address descend without page_table_lock. First just pte_offset_map in case there's no pte present worth locking for, then take page_table_lock for the full check, and pass ptl back to caller in the same style as pte_offset_map_lock. __xip_unmap, page_referenced_one and try_to_unmap_one use pte_unmap_unlock. try_to_unmap_cluster also. page_check_address reformatted to avoid progressive indentation. No use is made of its one error code, return NULL when it fails. Signed-off-by: Hugh Dickins Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/rmap.h | 4 +- mm/filemap_xip.c | 12 ++---- mm/rmap.c | 109 +++++++++++++++++++++++++-------------------------- 3 files changed, 60 insertions(+), 65 deletions(-) (limited to 'include/linux') diff --git a/include/linux/rmap.h b/include/linux/rmap.h index e80fb7ee6efd..35b30e6c8cf8 100644 --- a/include/linux/rmap.h +++ b/include/linux/rmap.h @@ -95,8 +95,8 @@ int try_to_unmap(struct page *); /* * Called from mm/filemap_xip.c to unmap empty zero page */ -pte_t *page_check_address(struct page *, struct mm_struct *, unsigned long); - +pte_t *page_check_address(struct page *, struct mm_struct *, + unsigned long, spinlock_t **); /* * Used by swapoff to help locate where page is expected in vma. diff --git a/mm/filemap_xip.c b/mm/filemap_xip.c index 4e74ad60339a..9cf687e4a29a 100644 --- a/mm/filemap_xip.c +++ b/mm/filemap_xip.c @@ -174,6 +174,7 @@ __xip_unmap (struct address_space * mapping, unsigned long address; pte_t *pte; pte_t pteval; + spinlock_t *ptl; struct page *page; spin_lock(&mapping->i_mmap_lock); @@ -183,20 +184,15 @@ __xip_unmap (struct address_space * mapping, ((pgoff - vma->vm_pgoff) << PAGE_SHIFT); BUG_ON(address < vma->vm_start || address >= vma->vm_end); page = ZERO_PAGE(address); - /* - * We need the page_table_lock to protect us from page faults, - * munmap, fork, etc... - */ - pte = page_check_address(page, mm, address); - if (!IS_ERR(pte)) { + pte = page_check_address(page, mm, address, &ptl); + if (pte) { /* Nuke the page table entry. */ flush_cache_page(vma, address, pte_pfn(*pte)); pteval = ptep_clear_flush(vma, address, pte); page_remove_rmap(page); dec_mm_counter(mm, file_rss); BUG_ON(pte_dirty(pteval)); - pte_unmap(pte); - spin_unlock(&mm->page_table_lock); + pte_unmap_unlock(pte, ptl); page_cache_release(page); } } diff --git a/mm/rmap.c b/mm/rmap.c index 4c52c56c9905..a84bdfe582c0 100644 --- a/mm/rmap.c +++ b/mm/rmap.c @@ -247,34 +247,41 @@ unsigned long page_address_in_vma(struct page *page, struct vm_area_struct *vma) * On success returns with mapped pte and locked mm->page_table_lock. */ pte_t *page_check_address(struct page *page, struct mm_struct *mm, - unsigned long address) + unsigned long address, spinlock_t **ptlp) { pgd_t *pgd; pud_t *pud; pmd_t *pmd; pte_t *pte; + spinlock_t *ptl; - /* - * We need the page_table_lock to protect us from page faults, - * munmap, fork, etc... - */ - spin_lock(&mm->page_table_lock); pgd = pgd_offset(mm, address); - if (likely(pgd_present(*pgd))) { - pud = pud_offset(pgd, address); - if (likely(pud_present(*pud))) { - pmd = pmd_offset(pud, address); - if (likely(pmd_present(*pmd))) { - pte = pte_offset_map(pmd, address); - if (likely(pte_present(*pte) && - page_to_pfn(page) == pte_pfn(*pte))) - return pte; - pte_unmap(pte); - } - } + if (!pgd_present(*pgd)) + return NULL; + + pud = pud_offset(pgd, address); + if (!pud_present(*pud)) + return NULL; + + pmd = pmd_offset(pud, address); + if (!pmd_present(*pmd)) + return NULL; + + pte = pte_offset_map(pmd, address); + /* Make a quick check before getting the lock */ + if (!pte_present(*pte)) { + pte_unmap(pte); + return NULL; } - spin_unlock(&mm->page_table_lock); - return ERR_PTR(-ENOENT); + + ptl = &mm->page_table_lock; + spin_lock(ptl); + if (pte_present(*pte) && page_to_pfn(page) == pte_pfn(*pte)) { + *ptlp = ptl; + return pte; + } + pte_unmap_unlock(pte, ptl); + return NULL; } /* @@ -287,28 +294,28 @@ static int page_referenced_one(struct page *page, struct mm_struct *mm = vma->vm_mm; unsigned long address; pte_t *pte; + spinlock_t *ptl; int referenced = 0; address = vma_address(page, vma); if (address == -EFAULT) goto out; - pte = page_check_address(page, mm, address); - if (!IS_ERR(pte)) { - if (ptep_clear_flush_young(vma, address, pte)) - referenced++; + pte = page_check_address(page, mm, address, &ptl); + if (!pte) + goto out; - /* Pretend the page is referenced if the task has the - swap token and is in the middle of a page fault. */ - if (mm != current->mm && !ignore_token && - has_swap_token(mm) && - rwsem_is_locked(&mm->mmap_sem)) - referenced++; + if (ptep_clear_flush_young(vma, address, pte)) + referenced++; - (*mapcount)--; - pte_unmap(pte); - spin_unlock(&mm->page_table_lock); - } + /* Pretend the page is referenced if the task has the + swap token and is in the middle of a page fault. */ + if (mm != current->mm && !ignore_token && has_swap_token(mm) && + rwsem_is_locked(&mm->mmap_sem)) + referenced++; + + (*mapcount)--; + pte_unmap_unlock(pte, ptl); out: return referenced; } @@ -507,14 +514,15 @@ static int try_to_unmap_one(struct page *page, struct vm_area_struct *vma) unsigned long address; pte_t *pte; pte_t pteval; + spinlock_t *ptl; int ret = SWAP_AGAIN; address = vma_address(page, vma); if (address == -EFAULT) goto out; - pte = page_check_address(page, mm, address); - if (IS_ERR(pte)) + pte = page_check_address(page, mm, address, &ptl); + if (!pte) goto out; /* @@ -564,8 +572,7 @@ static int try_to_unmap_one(struct page *page, struct vm_area_struct *vma) page_cache_release(page); out_unmap: - pte_unmap(pte); - spin_unlock(&mm->page_table_lock); + pte_unmap_unlock(pte, ptl); out: return ret; } @@ -599,19 +606,14 @@ static void try_to_unmap_cluster(unsigned long cursor, pgd_t *pgd; pud_t *pud; pmd_t *pmd; - pte_t *pte, *original_pte; + pte_t *pte; pte_t pteval; + spinlock_t *ptl; struct page *page; unsigned long address; unsigned long end; unsigned long pfn; - /* - * We need the page_table_lock to protect us from page faults, - * munmap, fork, etc... - */ - spin_lock(&mm->page_table_lock); - address = (vma->vm_start + cursor) & CLUSTER_MASK; end = address + CLUSTER_SIZE; if (address < vma->vm_start) @@ -621,22 +623,22 @@ static void try_to_unmap_cluster(unsigned long cursor, pgd = pgd_offset(mm, address); if (!pgd_present(*pgd)) - goto out_unlock; + return; pud = pud_offset(pgd, address); if (!pud_present(*pud)) - goto out_unlock; + return; pmd = pmd_offset(pud, address); if (!pmd_present(*pmd)) - goto out_unlock; + return; + + pte = pte_offset_map_lock(mm, pmd, address, &ptl); /* Update high watermark before we lower rss */ update_hiwater_rss(mm); - for (original_pte = pte = pte_offset_map(pmd, address); - address < end; pte++, address += PAGE_SIZE) { - + for (; address < end; pte++, address += PAGE_SIZE) { if (!pte_present(*pte)) continue; @@ -669,10 +671,7 @@ static void try_to_unmap_cluster(unsigned long cursor, dec_mm_counter(mm, file_rss); (*mapcount)--; } - - pte_unmap(original_pte); -out_unlock: - spin_unlock(&mm->page_table_lock); + pte_unmap_unlock(pte - 1, ptl); } static int try_to_unmap_anon(struct page *page) -- cgit v1.2.3 From c34d1b4d165c67b966bca4aba026443d7ff161eb Mon Sep 17 00:00:00 2001 From: Hugh Dickins Date: Sat, 29 Oct 2005 18:16:32 -0700 Subject: [PATCH] mm: kill check_user_page_readable check_user_page_readable is a problematic variant of follow_page. It's used only by oprofile's i386 and arm backtrace code, at interrupt time, to establish whether a userspace stackframe is currently readable. This is problematic, because we want to push the page_table_lock down inside follow_page, and later split it; whereas oprofile is doing a spin_trylock on it (in the i386 case, forgotten in the arm case), and needs that to pin perhaps two pages spanned by the stackframe (which might be covered by different locks when we split). I think oprofile is going about this in the wrong way: it doesn't need to know the area is readable (neither i386 nor arm uses read protection of user pages), it doesn't need to pin the memory, it should simply __copy_from_user_inatomic, and see if that succeeds or not. Sorry, but I've not got around to devising the sparse __user annotations for this. Then we can eliminate check_user_page_readable, and return to a single follow_page without the __follow_page variants. Signed-off-by: Hugh Dickins Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/arm/oprofile/backtrace.c | 46 +++++++++--------------------------------- arch/i386/oprofile/backtrace.c | 38 ++++++++++++---------------------- include/linux/mm.h | 1 - mm/memory.c | 29 ++++---------------------- 4 files changed, 26 insertions(+), 88 deletions(-) (limited to 'include/linux') diff --git a/arch/arm/oprofile/backtrace.c b/arch/arm/oprofile/backtrace.c index df35c452a8bf..7c22c12618cc 100644 --- a/arch/arm/oprofile/backtrace.c +++ b/arch/arm/oprofile/backtrace.c @@ -49,42 +49,22 @@ static struct frame_tail* kernel_backtrace(struct frame_tail *tail) static struct frame_tail* user_backtrace(struct frame_tail *tail) { - struct frame_tail buftail; + struct frame_tail buftail[2]; - /* hardware pte might not be valid due to dirty/accessed bit emulation - * so we use copy_from_user and benefit from exception fixups */ - if (copy_from_user(&buftail, tail, sizeof(struct frame_tail))) + /* Also check accessibility of one struct frame_tail beyond */ + if (!access_ok(VERIFY_READ, tail, sizeof(buftail))) + return NULL; + if (__copy_from_user_inatomic(buftail, tail, sizeof(buftail))) return NULL; - oprofile_add_trace(buftail.lr); + oprofile_add_trace(buftail[0].lr); /* frame pointers should strictly progress back up the stack * (towards higher addresses) */ - if (tail >= buftail.fp) + if (tail >= buftail[0].fp) return NULL; - return buftail.fp-1; -} - -/* Compare two addresses and see if they're on the same page */ -#define CMP_ADDR_EQUAL(x,y,offset) ((((unsigned long) x) >> PAGE_SHIFT) \ - == ((((unsigned long) y) + offset) >> PAGE_SHIFT)) - -/* check that the page(s) containing the frame tail are present */ -static int pages_present(struct frame_tail *tail) -{ - struct mm_struct * mm = current->mm; - - if (!check_user_page_readable(mm, (unsigned long)tail)) - return 0; - - if (CMP_ADDR_EQUAL(tail, tail, 8)) - return 1; - - if (!check_user_page_readable(mm, ((unsigned long)tail) + 8)) - return 0; - - return 1; + return buftail[0].fp-1; } /* @@ -118,7 +98,6 @@ static int valid_kernel_stack(struct frame_tail *tail, struct pt_regs *regs) void arm_backtrace(struct pt_regs * const regs, unsigned int depth) { struct frame_tail *tail; - unsigned long last_address = 0; tail = ((struct frame_tail *) regs->ARM_fp) - 1; @@ -132,13 +111,6 @@ void arm_backtrace(struct pt_regs * const regs, unsigned int depth) return; } - while (depth-- && tail && !((unsigned long) tail & 3)) { - if ((!CMP_ADDR_EQUAL(last_address, tail, 0) - || !CMP_ADDR_EQUAL(last_address, tail, 8)) - && !pages_present(tail)) - return; - last_address = (unsigned long) tail; + while (depth-- && tail && !((unsigned long) tail & 3)) tail = user_backtrace(tail); - } } - diff --git a/arch/i386/oprofile/backtrace.c b/arch/i386/oprofile/backtrace.c index 65dfd2edb671..21654be3f73f 100644 --- a/arch/i386/oprofile/backtrace.c +++ b/arch/i386/oprofile/backtrace.c @@ -12,6 +12,7 @@ #include #include #include +#include struct frame_head { struct frame_head * ebp; @@ -21,26 +22,22 @@ struct frame_head { static struct frame_head * dump_backtrace(struct frame_head * head) { - oprofile_add_trace(head->ret); + struct frame_head bufhead[2]; - /* frame pointers should strictly progress back up the stack - * (towards higher addresses) */ - if (head >= head->ebp) + /* Also check accessibility of one struct frame_head beyond */ + if (!access_ok(VERIFY_READ, head, sizeof(bufhead))) + return NULL; + if (__copy_from_user_inatomic(bufhead, head, sizeof(bufhead))) return NULL; - return head->ebp; -} - -/* check that the page(s) containing the frame head are present */ -static int pages_present(struct frame_head * head) -{ - struct mm_struct * mm = current->mm; + oprofile_add_trace(bufhead[0].ret); - /* FIXME: only necessary once per page */ - if (!check_user_page_readable(mm, (unsigned long)head)) - return 0; + /* frame pointers should strictly progress back up the stack + * (towards higher addresses) */ + if (head >= bufhead[0].ebp) + return NULL; - return check_user_page_readable(mm, (unsigned long)(head + 1)); + return bufhead[0].ebp; } /* @@ -97,15 +94,6 @@ x86_backtrace(struct pt_regs * const regs, unsigned int depth) return; } -#ifdef CONFIG_SMP - if (!spin_trylock(¤t->mm->page_table_lock)) - return; -#endif - - while (depth-- && head && pages_present(head)) + while (depth-- && head) head = dump_backtrace(head); - -#ifdef CONFIG_SMP - spin_unlock(¤t->mm->page_table_lock); -#endif } diff --git a/include/linux/mm.h b/include/linux/mm.h index 972e2ce8e07c..aa8de20e2e80 100644 --- a/include/linux/mm.h +++ b/include/linux/mm.h @@ -944,7 +944,6 @@ extern struct page * vmalloc_to_page(void *addr); extern unsigned long vmalloc_to_pfn(void *addr); extern struct page * follow_page(struct mm_struct *mm, unsigned long address, int write); -extern int check_user_page_readable(struct mm_struct *mm, unsigned long address); int remap_pfn_range(struct vm_area_struct *, unsigned long, unsigned long, unsigned long, pgprot_t); diff --git a/mm/memory.c b/mm/memory.c index 622a4ef5409f..51f7c0a220d4 100644 --- a/mm/memory.c +++ b/mm/memory.c @@ -809,8 +809,7 @@ unsigned long zap_page_range(struct vm_area_struct *vma, unsigned long address, * Do a quick page-table lookup for a single page. * mm->page_table_lock must be held. */ -static struct page *__follow_page(struct mm_struct *mm, unsigned long address, - int read, int write, int accessed) +struct page *follow_page(struct mm_struct *mm, unsigned long address, int write) { pgd_t *pgd; pud_t *pud; @@ -846,16 +845,12 @@ static struct page *__follow_page(struct mm_struct *mm, unsigned long address, if (pte_present(pte)) { if (write && !pte_write(pte)) goto out; - if (read && !pte_read(pte)) - goto out; pfn = pte_pfn(pte); if (pfn_valid(pfn)) { page = pfn_to_page(pfn); - if (accessed) { - if (write && !pte_dirty(pte) &&!PageDirty(page)) - set_page_dirty(page); - mark_page_accessed(page); - } + if (write && !pte_dirty(pte) &&!PageDirty(page)) + set_page_dirty(page); + mark_page_accessed(page); return page; } } @@ -864,22 +859,6 @@ out: return NULL; } -inline struct page * -follow_page(struct mm_struct *mm, unsigned long address, int write) -{ - return __follow_page(mm, address, 0, write, 1); -} - -/* - * check_user_page_readable() can be called frm niterrupt context by oprofile, - * so we need to avoid taking any non-irq-safe locks - */ -int check_user_page_readable(struct mm_struct *mm, unsigned long address) -{ - return __follow_page(mm, address, 1, 0, 0) != NULL; -} -EXPORT_SYMBOL(check_user_page_readable); - static inline int untouched_anonymous_page(struct mm_struct* mm, struct vm_area_struct *vma, unsigned long address) -- cgit v1.2.3 From deceb6cd17e6dfafe4c4f81b1b4153bc41b2cb70 Mon Sep 17 00:00:00 2001 From: Hugh Dickins Date: Sat, 29 Oct 2005 18:16:33 -0700 Subject: [PATCH] mm: follow_page with inner ptlock Final step in pushing down common core's page_table_lock. follow_page no longer wants caller to hold page_table_lock, uses pte_offset_map_lock itself; and so no page_table_lock is taken in get_user_pages itself. But get_user_pages (and get_futex_key) do then need follow_page to pin the page for them: take Daniel's suggestion of bitflags to follow_page. Need one for WRITE, another for TOUCH (it was the accessed flag before: vanished along with check_user_page_readable, but surely get_numa_maps is wrong to mark every page it finds as accessed), another for GET. And another, ANON to dispose of untouched_anonymous_page: it seems silly for that to descend a second time, let follow_page observe if there was no page table and return ZERO_PAGE if so. Fix minor bug in that: check VM_LOCKED - make_pages_present ought to make readonly anonymous present. Give get_numa_maps a cond_resched while we're there. Signed-off-by: Hugh Dickins Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/proc/task_mmu.c | 3 +- include/linux/mm.h | 20 ++++--- kernel/futex.c | 6 +-- mm/memory.c | 152 +++++++++++++++++++++++++---------------------------- mm/nommu.c | 3 +- 5 files changed, 88 insertions(+), 96 deletions(-) (limited to 'include/linux') diff --git a/fs/proc/task_mmu.c b/fs/proc/task_mmu.c index 7e5e7ec2e36d..d2fa42006d8f 100644 --- a/fs/proc/task_mmu.c +++ b/fs/proc/task_mmu.c @@ -419,7 +419,6 @@ static struct numa_maps *get_numa_maps(const struct vm_area_struct *vma) for_each_node(i) md->node[i] =0; - spin_lock(&mm->page_table_lock); for (vaddr = vma->vm_start; vaddr < vma->vm_end; vaddr += PAGE_SIZE) { page = follow_page(mm, vaddr, 0); if (page) { @@ -434,8 +433,8 @@ static struct numa_maps *get_numa_maps(const struct vm_area_struct *vma) md->anon++; md->node[page_to_nid(page)]++; } + cond_resched(); } - spin_unlock(&mm->page_table_lock); return md; } diff --git a/include/linux/mm.h b/include/linux/mm.h index aa8de20e2e80..e8d1424153bb 100644 --- a/include/linux/mm.h +++ b/include/linux/mm.h @@ -938,14 +938,18 @@ static inline unsigned long vma_pages(struct vm_area_struct *vma) return (vma->vm_end - vma->vm_start) >> PAGE_SHIFT; } -extern struct vm_area_struct *find_extend_vma(struct mm_struct *mm, unsigned long addr); - -extern struct page * vmalloc_to_page(void *addr); -extern unsigned long vmalloc_to_pfn(void *addr); -extern struct page * follow_page(struct mm_struct *mm, unsigned long address, - int write); -int remap_pfn_range(struct vm_area_struct *, unsigned long, - unsigned long, unsigned long, pgprot_t); +struct vm_area_struct *find_extend_vma(struct mm_struct *, unsigned long addr); +struct page *vmalloc_to_page(void *addr); +unsigned long vmalloc_to_pfn(void *addr); +int remap_pfn_range(struct vm_area_struct *, unsigned long addr, + unsigned long pfn, unsigned long size, pgprot_t); + +struct page *follow_page(struct mm_struct *, unsigned long address, + unsigned int foll_flags); +#define FOLL_WRITE 0x01 /* check pte is writable */ +#define FOLL_TOUCH 0x02 /* mark page accessed */ +#define FOLL_GET 0x04 /* do get_page on page */ +#define FOLL_ANON 0x08 /* give ZERO_PAGE if no pgtable */ #ifdef CONFIG_PROC_FS void vm_stat_account(struct mm_struct *, unsigned long, struct file *, long); diff --git a/kernel/futex.c b/kernel/futex.c index ca05fe6a70b2..3b4d5ad44cc6 100644 --- a/kernel/futex.c +++ b/kernel/futex.c @@ -205,15 +205,13 @@ static int get_futex_key(unsigned long uaddr, union futex_key *key) /* * Do a quick atomic lookup first - this is the fastpath. */ - spin_lock(¤t->mm->page_table_lock); - page = follow_page(mm, uaddr, 0); + page = follow_page(mm, uaddr, FOLL_TOUCH|FOLL_GET); if (likely(page != NULL)) { key->shared.pgoff = page->index << (PAGE_CACHE_SHIFT - PAGE_SHIFT); - spin_unlock(¤t->mm->page_table_lock); + put_page(page); return 0; } - spin_unlock(¤t->mm->page_table_lock); /* * Do it the general way. diff --git a/mm/memory.c b/mm/memory.c index 51f7c0a220d4..8461e2dd91d7 100644 --- a/mm/memory.c +++ b/mm/memory.c @@ -807,86 +807,82 @@ unsigned long zap_page_range(struct vm_area_struct *vma, unsigned long address, /* * Do a quick page-table lookup for a single page. - * mm->page_table_lock must be held. */ -struct page *follow_page(struct mm_struct *mm, unsigned long address, int write) +struct page *follow_page(struct mm_struct *mm, unsigned long address, + unsigned int flags) { pgd_t *pgd; pud_t *pud; pmd_t *pmd; pte_t *ptep, pte; + spinlock_t *ptl; unsigned long pfn; struct page *page; - page = follow_huge_addr(mm, address, write); - if (! IS_ERR(page)) - return page; + page = follow_huge_addr(mm, address, flags & FOLL_WRITE); + if (!IS_ERR(page)) { + BUG_ON(flags & FOLL_GET); + goto out; + } + page = NULL; pgd = pgd_offset(mm, address); if (pgd_none(*pgd) || unlikely(pgd_bad(*pgd))) - goto out; + goto no_page_table; pud = pud_offset(pgd, address); if (pud_none(*pud) || unlikely(pud_bad(*pud))) - goto out; + goto no_page_table; pmd = pmd_offset(pud, address); if (pmd_none(*pmd) || unlikely(pmd_bad(*pmd))) + goto no_page_table; + + if (pmd_huge(*pmd)) { + BUG_ON(flags & FOLL_GET); + page = follow_huge_pmd(mm, address, pmd, flags & FOLL_WRITE); goto out; - if (pmd_huge(*pmd)) - return follow_huge_pmd(mm, address, pmd, write); + } - ptep = pte_offset_map(pmd, address); + ptep = pte_offset_map_lock(mm, pmd, address, &ptl); if (!ptep) goto out; pte = *ptep; - pte_unmap(ptep); - if (pte_present(pte)) { - if (write && !pte_write(pte)) - goto out; - pfn = pte_pfn(pte); - if (pfn_valid(pfn)) { - page = pfn_to_page(pfn); - if (write && !pte_dirty(pte) &&!PageDirty(page)) - set_page_dirty(page); - mark_page_accessed(page); - return page; - } - } + if (!pte_present(pte)) + goto unlock; + if ((flags & FOLL_WRITE) && !pte_write(pte)) + goto unlock; + pfn = pte_pfn(pte); + if (!pfn_valid(pfn)) + goto unlock; + page = pfn_to_page(pfn); + if (flags & FOLL_GET) + get_page(page); + if (flags & FOLL_TOUCH) { + if ((flags & FOLL_WRITE) && + !pte_dirty(pte) && !PageDirty(page)) + set_page_dirty(page); + mark_page_accessed(page); + } +unlock: + pte_unmap_unlock(ptep, ptl); out: - return NULL; -} - -static inline int -untouched_anonymous_page(struct mm_struct* mm, struct vm_area_struct *vma, - unsigned long address) -{ - pgd_t *pgd; - pud_t *pud; - pmd_t *pmd; - - /* Check if the vma is for an anonymous mapping. */ - if (vma->vm_ops && vma->vm_ops->nopage) - return 0; - - /* Check if page directory entry exists. */ - pgd = pgd_offset(mm, address); - if (pgd_none(*pgd) || unlikely(pgd_bad(*pgd))) - return 1; - - pud = pud_offset(pgd, address); - if (pud_none(*pud) || unlikely(pud_bad(*pud))) - return 1; - - /* Check if page middle directory entry exists. */ - pmd = pmd_offset(pud, address); - if (pmd_none(*pmd) || unlikely(pmd_bad(*pmd))) - return 1; + return page; - /* There is a pte slot for 'address' in 'mm'. */ - return 0; +no_page_table: + /* + * When core dumping an enormous anonymous area that nobody + * has touched so far, we don't want to allocate page tables. + */ + if (flags & FOLL_ANON) { + page = ZERO_PAGE(address); + if (flags & FOLL_GET) + get_page(page); + BUG_ON(flags & FOLL_WRITE); + } + return page; } int get_user_pages(struct task_struct *tsk, struct mm_struct *mm, @@ -894,18 +890,19 @@ int get_user_pages(struct task_struct *tsk, struct mm_struct *mm, struct page **pages, struct vm_area_struct **vmas) { int i; - unsigned int flags; + unsigned int vm_flags; /* * Require read or write permissions. * If 'force' is set, we only require the "MAY" flags. */ - flags = write ? (VM_WRITE | VM_MAYWRITE) : (VM_READ | VM_MAYREAD); - flags &= force ? (VM_MAYREAD | VM_MAYWRITE) : (VM_READ | VM_WRITE); + vm_flags = write ? (VM_WRITE | VM_MAYWRITE) : (VM_READ | VM_MAYREAD); + vm_flags &= force ? (VM_MAYREAD | VM_MAYWRITE) : (VM_READ | VM_WRITE); i = 0; do { - struct vm_area_struct * vma; + struct vm_area_struct *vma; + unsigned int foll_flags; vma = find_extend_vma(mm, start); if (!vma && in_gate_area(tsk, start)) { @@ -946,7 +943,7 @@ int get_user_pages(struct task_struct *tsk, struct mm_struct *mm, } if (!vma || (vma->vm_flags & (VM_IO | VM_RESERVED)) - || !(flags & vma->vm_flags)) + || !(vm_flags & vma->vm_flags)) return i ? : -EFAULT; if (is_vm_hugetlb_page(vma)) { @@ -954,29 +951,25 @@ int get_user_pages(struct task_struct *tsk, struct mm_struct *mm, &start, &len, i); continue; } - spin_lock(&mm->page_table_lock); + + foll_flags = FOLL_TOUCH; + if (pages) + foll_flags |= FOLL_GET; + if (!write && !(vma->vm_flags & VM_LOCKED) && + (!vma->vm_ops || !vma->vm_ops->nopage)) + foll_flags |= FOLL_ANON; + do { - int write_access = write; struct page *page; - cond_resched_lock(&mm->page_table_lock); - while (!(page = follow_page(mm, start, write_access))) { - int ret; - - /* - * Shortcut for anonymous pages. We don't want - * to force the creation of pages tables for - * insanely big anonymously mapped areas that - * nobody touched so far. This is important - * for doing a core dump for these mappings. - */ - if (!write && untouched_anonymous_page(mm,vma,start)) { - page = ZERO_PAGE(start); - break; - } - spin_unlock(&mm->page_table_lock); - ret = __handle_mm_fault(mm, vma, start, write_access); + if (write) + foll_flags |= FOLL_WRITE; + cond_resched(); + while (!(page = follow_page(mm, start, foll_flags))) { + int ret; + ret = __handle_mm_fault(mm, vma, start, + foll_flags & FOLL_WRITE); /* * The VM_FAULT_WRITE bit tells us that do_wp_page has * broken COW when necessary, even if maybe_mkwrite @@ -984,7 +977,7 @@ int get_user_pages(struct task_struct *tsk, struct mm_struct *mm, * subsequent page lookups as if they were reads. */ if (ret & VM_FAULT_WRITE) - write_access = 0; + foll_flags &= ~FOLL_WRITE; switch (ret & ~VM_FAULT_WRITE) { case VM_FAULT_MINOR: @@ -1000,12 +993,10 @@ int get_user_pages(struct task_struct *tsk, struct mm_struct *mm, default: BUG(); } - spin_lock(&mm->page_table_lock); } if (pages) { pages[i] = page; flush_dcache_page(page); - page_cache_get(page); } if (vmas) vmas[i] = vma; @@ -1013,7 +1004,6 @@ int get_user_pages(struct task_struct *tsk, struct mm_struct *mm, start += PAGE_SIZE; len--; } while (len && start < vma->vm_end); - spin_unlock(&mm->page_table_lock); } while (len); return i; } diff --git a/mm/nommu.c b/mm/nommu.c index dfb124ffb9be..d1e076a487cb 100644 --- a/mm/nommu.c +++ b/mm/nommu.c @@ -1049,7 +1049,8 @@ struct vm_area_struct *find_vma(struct mm_struct *mm, unsigned long addr) EXPORT_SYMBOL(find_vma); -struct page * follow_page(struct mm_struct *mm, unsigned long addr, int write) +struct page *follow_page(struct mm_struct *mm, unsigned long address, + unsigned int foll_flags) { return NULL; } -- cgit v1.2.3 From 4c21e2f2441dc5fbb957b030333f5a3f2d02dea7 Mon Sep 17 00:00:00 2001 From: Hugh Dickins Date: Sat, 29 Oct 2005 18:16:40 -0700 Subject: [PATCH] mm: split page table lock Christoph Lameter demonstrated very poor scalability on the SGI 512-way, with a many-threaded application which concurrently initializes different parts of a large anonymous area. This patch corrects that, by using a separate spinlock per page table page, to guard the page table entries in that page, instead of using the mm's single page_table_lock. (But even then, page_table_lock is still used to guard page table allocation, and anon_vma allocation.) In this implementation, the spinlock is tucked inside the struct page of the page table page: with a BUILD_BUG_ON in case it overflows - which it would in the case of 32-bit PA-RISC with spinlock debugging enabled. Splitting the lock is not quite for free: another cacheline access. Ideally, I suppose we would use split ptlock only for multi-threaded processes on multi-cpu machines; but deciding that dynamically would have its own costs. So for now enable it by config, at some number of cpus - since the Kconfig language doesn't support inequalities, let preprocessor compare that with NR_CPUS. But I don't think it's worth being user-configurable: for good testing of both split and unsplit configs, split now at 4 cpus, and perhaps change that to 8 later. There is a benefit even for singly threaded processes: kswapd can be attacking one part of the mm while another part is busy faulting. Signed-off-by: Hugh Dickins Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/arm/mm/mm-armv.c | 1 + arch/frv/mm/pgalloc.c | 4 ++-- arch/i386/mm/pgtable.c | 8 ++++---- arch/um/kernel/skas/mmu.c | 1 + fs/afs/file.c | 4 ++-- fs/buffer.c | 2 +- fs/jfs/jfs_metapage.c | 12 ++++++------ fs/xfs/linux-2.6/xfs_buf.c | 7 ++++--- include/linux/buffer_head.h | 6 +++--- include/linux/mm.h | 46 +++++++++++++++++++++++++++++++++++++-------- kernel/kexec.c | 4 ++-- mm/Kconfig | 13 +++++++++++++ mm/filemap.c | 2 +- mm/memory.c | 24 +++++++++++++---------- mm/mremap.c | 11 ++++++++++- mm/page_alloc.c | 16 ++++++++-------- mm/page_io.c | 6 ++++-- mm/rmap.c | 4 ++-- mm/shmem.c | 22 ++++++++++------------ mm/swap.c | 2 +- mm/swap_state.c | 8 ++++---- mm/swapfile.c | 12 ++++++------ mm/vmscan.c | 2 +- 23 files changed, 138 insertions(+), 79 deletions(-) (limited to 'include/linux') diff --git a/arch/arm/mm/mm-armv.c b/arch/arm/mm/mm-armv.c index 60f3e039bac2..1221fdde1769 100644 --- a/arch/arm/mm/mm-armv.c +++ b/arch/arm/mm/mm-armv.c @@ -229,6 +229,7 @@ void free_pgd_slow(pgd_t *pgd) pte = pmd_page(*pmd); pmd_clear(pmd); dec_page_state(nr_page_table_pages); + pte_lock_deinit(pte); pte_free(pte); pmd_free(pmd); free: diff --git a/arch/frv/mm/pgalloc.c b/arch/frv/mm/pgalloc.c index 4eaec0f3525b..2c67dfe5a6b3 100644 --- a/arch/frv/mm/pgalloc.c +++ b/arch/frv/mm/pgalloc.c @@ -87,14 +87,14 @@ static inline void pgd_list_add(pgd_t *pgd) if (pgd_list) pgd_list->private = (unsigned long) &page->index; pgd_list = page; - page->private = (unsigned long) &pgd_list; + set_page_private(page, (unsigned long)&pgd_list); } static inline void pgd_list_del(pgd_t *pgd) { struct page *next, **pprev, *page = virt_to_page(pgd); next = (struct page *) page->index; - pprev = (struct page **) page->private; + pprev = (struct page **)page_private(page); *pprev = next; if (next) next->private = (unsigned long) pprev; diff --git a/arch/i386/mm/pgtable.c b/arch/i386/mm/pgtable.c index dcdce2c6c532..39c099f15b5f 100644 --- a/arch/i386/mm/pgtable.c +++ b/arch/i386/mm/pgtable.c @@ -188,19 +188,19 @@ static inline void pgd_list_add(pgd_t *pgd) struct page *page = virt_to_page(pgd); page->index = (unsigned long)pgd_list; if (pgd_list) - pgd_list->private = (unsigned long)&page->index; + set_page_private(pgd_list, (unsigned long)&page->index); pgd_list = page; - page->private = (unsigned long)&pgd_list; + set_page_private(page, (unsigned long)&pgd_list); } static inline void pgd_list_del(pgd_t *pgd) { struct page *next, **pprev, *page = virt_to_page(pgd); next = (struct page *)page->index; - pprev = (struct page **)page->private; + pprev = (struct page **)page_private(page); *pprev = next; if (next) - next->private = (unsigned long)pprev; + set_page_private(next, (unsigned long)pprev); } void pgd_ctor(void *pgd, kmem_cache_t *cache, unsigned long unused) diff --git a/arch/um/kernel/skas/mmu.c b/arch/um/kernel/skas/mmu.c index 02cf36e0331a..9e5e39cea821 100644 --- a/arch/um/kernel/skas/mmu.c +++ b/arch/um/kernel/skas/mmu.c @@ -144,6 +144,7 @@ void destroy_context_skas(struct mm_struct *mm) if(!proc_mm || !ptrace_faultinfo){ free_page(mmu->id.stack); + pte_lock_deinit(virt_to_page(mmu->last_page_table)); pte_free_kernel((pte_t *) mmu->last_page_table); dec_page_state(nr_page_table_pages); #ifdef CONFIG_3_LEVEL_PGTABLES diff --git a/fs/afs/file.c b/fs/afs/file.c index 0d576987ec67..4975c9c193dd 100644 --- a/fs/afs/file.c +++ b/fs/afs/file.c @@ -291,8 +291,8 @@ static int afs_file_releasepage(struct page *page, gfp_t gfp_flags) cachefs_uncache_page(vnode->cache, page); #endif - pageio = (struct cachefs_page *) page->private; - page->private = 0; + pageio = (struct cachefs_page *) page_private(page); + set_page_private(page, 0); ClearPagePrivate(page); if (pageio) diff --git a/fs/buffer.c b/fs/buffer.c index b1667986442f..2066e4cb700c 100644 --- a/fs/buffer.c +++ b/fs/buffer.c @@ -96,7 +96,7 @@ static void __clear_page_buffers(struct page *page) { ClearPagePrivate(page); - page->private = 0; + set_page_private(page, 0); page_cache_release(page); } diff --git a/fs/jfs/jfs_metapage.c b/fs/jfs/jfs_metapage.c index 26091a5f88d4..8a53981f9f27 100644 --- a/fs/jfs/jfs_metapage.c +++ b/fs/jfs/jfs_metapage.c @@ -86,7 +86,7 @@ struct meta_anchor { atomic_t io_count; struct metapage *mp[MPS_PER_PAGE]; }; -#define mp_anchor(page) ((struct meta_anchor *)page->private) +#define mp_anchor(page) ((struct meta_anchor *)page_private(page)) static inline struct metapage *page_to_mp(struct page *page, uint offset) { @@ -108,7 +108,7 @@ static inline int insert_metapage(struct page *page, struct metapage *mp) if (!a) return -ENOMEM; memset(a, 0, sizeof(struct meta_anchor)); - page->private = (unsigned long)a; + set_page_private(page, (unsigned long)a); SetPagePrivate(page); kmap(page); } @@ -136,7 +136,7 @@ static inline void remove_metapage(struct page *page, struct metapage *mp) a->mp[index] = NULL; if (--a->mp_count == 0) { kfree(a); - page->private = 0; + set_page_private(page, 0); ClearPagePrivate(page); kunmap(page); } @@ -156,13 +156,13 @@ static inline void dec_io(struct page *page, void (*handler) (struct page *)) #else static inline struct metapage *page_to_mp(struct page *page, uint offset) { - return PagePrivate(page) ? (struct metapage *)page->private : NULL; + return PagePrivate(page) ? (struct metapage *)page_private(page) : NULL; } static inline int insert_metapage(struct page *page, struct metapage *mp) { if (mp) { - page->private = (unsigned long)mp; + set_page_private(page, (unsigned long)mp); SetPagePrivate(page); kmap(page); } @@ -171,7 +171,7 @@ static inline int insert_metapage(struct page *page, struct metapage *mp) static inline void remove_metapage(struct page *page, struct metapage *mp) { - page->private = 0; + set_page_private(page, 0); ClearPagePrivate(page); kunmap(page); } diff --git a/fs/xfs/linux-2.6/xfs_buf.c b/fs/xfs/linux-2.6/xfs_buf.c index ba4767c04adf..4cd46abe8434 100644 --- a/fs/xfs/linux-2.6/xfs_buf.c +++ b/fs/xfs/linux-2.6/xfs_buf.c @@ -181,8 +181,9 @@ set_page_region( size_t offset, size_t length) { - page->private |= page_region_mask(offset, length); - if (page->private == ~0UL) + set_page_private(page, + page_private(page) | page_region_mask(offset, length)); + if (page_private(page) == ~0UL) SetPageUptodate(page); } @@ -194,7 +195,7 @@ test_page_region( { unsigned long mask = page_region_mask(offset, length); - return (mask && (page->private & mask) == mask); + return (mask && (page_private(page) & mask) == mask); } /* diff --git a/include/linux/buffer_head.h b/include/linux/buffer_head.h index 88af42f5e04a..c937d6e65502 100644 --- a/include/linux/buffer_head.h +++ b/include/linux/buffer_head.h @@ -126,8 +126,8 @@ BUFFER_FNS(Eopnotsupp, eopnotsupp) /* If we *know* page->private refers to buffer_heads */ #define page_buffers(page) \ ({ \ - BUG_ON(!PagePrivate(page)); \ - ((struct buffer_head *)(page)->private); \ + BUG_ON(!PagePrivate(page)); \ + ((struct buffer_head *)page_private(page)); \ }) #define page_has_buffers(page) PagePrivate(page) @@ -219,7 +219,7 @@ static inline void attach_page_buffers(struct page *page, { page_cache_get(page); SetPagePrivate(page); - page->private = (unsigned long)head; + set_page_private(page, (unsigned long)head); } static inline void get_bh(struct buffer_head *bh) diff --git a/include/linux/mm.h b/include/linux/mm.h index e8d1424153bb..8a514eca40d5 100644 --- a/include/linux/mm.h +++ b/include/linux/mm.h @@ -226,13 +226,18 @@ struct page { * to show when page is mapped * & limit reverse map searches. */ - unsigned long private; /* Mapping-private opaque data: + union { + unsigned long private; /* Mapping-private opaque data: * usually used for buffer_heads * if PagePrivate set; used for * swp_entry_t if PageSwapCache * When page is free, this indicates * order in the buddy system. */ +#if NR_CPUS >= CONFIG_SPLIT_PTLOCK_CPUS + spinlock_t ptl; +#endif + } u; struct address_space *mapping; /* If low bit clear, points to * inode address_space, or NULL. * If page mapped as anonymous @@ -260,6 +265,9 @@ struct page { #endif /* WANT_PAGE_VIRTUAL */ }; +#define page_private(page) ((page)->u.private) +#define set_page_private(page, v) ((page)->u.private = (v)) + /* * FIXME: take this include out, include page-flags.h in * files which need it (119 of them) @@ -311,17 +319,17 @@ extern void FASTCALL(__page_cache_release(struct page *)); #ifdef CONFIG_HUGETLB_PAGE -static inline int page_count(struct page *p) +static inline int page_count(struct page *page) { - if (PageCompound(p)) - p = (struct page *)p->private; - return atomic_read(&(p)->_count) + 1; + if (PageCompound(page)) + page = (struct page *)page_private(page); + return atomic_read(&page->_count) + 1; } static inline void get_page(struct page *page) { if (unlikely(PageCompound(page))) - page = (struct page *)page->private; + page = (struct page *)page_private(page); atomic_inc(&page->_count); } @@ -587,7 +595,7 @@ static inline int PageAnon(struct page *page) static inline pgoff_t page_index(struct page *page) { if (unlikely(PageSwapCache(page))) - return page->private; + return page_private(page); return page->index; } @@ -779,9 +787,31 @@ static inline pmd_t *pmd_alloc(struct mm_struct *mm, pud_t *pud, unsigned long a } #endif /* CONFIG_MMU && !__ARCH_HAS_4LEVEL_HACK */ +#if NR_CPUS >= CONFIG_SPLIT_PTLOCK_CPUS +/* + * We tuck a spinlock to guard each pagetable page into its struct page, + * at page->private, with BUILD_BUG_ON to make sure that this will not + * overflow into the next struct page (as it might with DEBUG_SPINLOCK). + * When freeing, reset page->mapping so free_pages_check won't complain. + */ +#define __pte_lockptr(page) &((page)->u.ptl) +#define pte_lock_init(_page) do { \ + spin_lock_init(__pte_lockptr(_page)); \ +} while (0) +#define pte_lock_deinit(page) ((page)->mapping = NULL) +#define pte_lockptr(mm, pmd) ({(void)(mm); __pte_lockptr(pmd_page(*(pmd)));}) +#else +/* + * We use mm->page_table_lock to guard all pagetable pages of the mm. + */ +#define pte_lock_init(page) do {} while (0) +#define pte_lock_deinit(page) do {} while (0) +#define pte_lockptr(mm, pmd) ({(void)(pmd); &(mm)->page_table_lock;}) +#endif /* NR_CPUS < CONFIG_SPLIT_PTLOCK_CPUS */ + #define pte_offset_map_lock(mm, pmd, address, ptlp) \ ({ \ - spinlock_t *__ptl = &(mm)->page_table_lock; \ + spinlock_t *__ptl = pte_lockptr(mm, pmd); \ pte_t *__pte = pte_offset_map(pmd, address); \ *(ptlp) = __ptl; \ spin_lock(__ptl); \ diff --git a/kernel/kexec.c b/kernel/kexec.c index 36c5d9cd4cc1..2c95848fbce8 100644 --- a/kernel/kexec.c +++ b/kernel/kexec.c @@ -334,7 +334,7 @@ static struct page *kimage_alloc_pages(gfp_t gfp_mask, unsigned int order) if (pages) { unsigned int count, i; pages->mapping = NULL; - pages->private = order; + set_page_private(pages, order); count = 1 << order; for (i = 0; i < count; i++) SetPageReserved(pages + i); @@ -347,7 +347,7 @@ static void kimage_free_pages(struct page *page) { unsigned int order, count, i; - order = page->private; + order = page_private(page); count = 1 << order; for (i = 0; i < count; i++) ClearPageReserved(page + i); diff --git a/mm/Kconfig b/mm/Kconfig index 391ffc54d136..f35a550ba4b9 100644 --- a/mm/Kconfig +++ b/mm/Kconfig @@ -111,3 +111,16 @@ config SPARSEMEM_STATIC config SPARSEMEM_EXTREME def_bool y depends on SPARSEMEM && !SPARSEMEM_STATIC + +# Heavily threaded applications may benefit from splitting the mm-wide +# page_table_lock, so that faults on different parts of the user address +# space can be handled with less contention: split it at this NR_CPUS. +# Default to 4 for wider testing, though 8 might be more appropriate. +# ARM's adjust_pte (unused if VIPT) depends on mm-wide page_table_lock. +# PA-RISC's debug spinlock_t is too large for the 32-bit struct page. +# +config SPLIT_PTLOCK_CPUS + int + default "4096" if ARM && !CPU_CACHE_VIPT + default "4096" if PARISC && DEBUG_SPINLOCK && !64BIT + default "4" diff --git a/mm/filemap.c b/mm/filemap.c index 8aa344e88489..f560b41c8f61 100644 --- a/mm/filemap.c +++ b/mm/filemap.c @@ -152,7 +152,7 @@ static int sync_page(void *word) * in the ->sync_page() methods make essential use of the * page_mapping(), merely passing the page down to the backing * device's unplug functions when it's non-NULL, which in turn - * ignore it for all cases but swap, where only page->private is + * ignore it for all cases but swap, where only page_private(page) is * of interest. When page_mapping() does go NULL, the entire * call stack gracefully ignores the page and returns. * -- wli diff --git a/mm/memory.c b/mm/memory.c index 8461e2dd91d7..e9ef599498b5 100644 --- a/mm/memory.c +++ b/mm/memory.c @@ -114,6 +114,7 @@ static void free_pte_range(struct mmu_gather *tlb, pmd_t *pmd) { struct page *page = pmd_page(*pmd); pmd_clear(pmd); + pte_lock_deinit(page); pte_free_tlb(tlb, page); dec_page_state(nr_page_table_pages); tlb->mm->nr_ptes--; @@ -294,10 +295,12 @@ int __pte_alloc(struct mm_struct *mm, pmd_t *pmd, unsigned long address) if (!new) return -ENOMEM; + pte_lock_init(new); spin_lock(&mm->page_table_lock); - if (pmd_present(*pmd)) /* Another has populated it */ + if (pmd_present(*pmd)) { /* Another has populated it */ + pte_lock_deinit(new); pte_free(new); - else { + } else { mm->nr_ptes++; inc_page_state(nr_page_table_pages); pmd_populate(mm, pmd, new); @@ -432,7 +435,7 @@ again: if (!dst_pte) return -ENOMEM; src_pte = pte_offset_map_nested(src_pmd, addr); - src_ptl = &src_mm->page_table_lock; + src_ptl = pte_lockptr(src_mm, src_pmd); spin_lock(src_ptl); do { @@ -1194,15 +1197,16 @@ EXPORT_SYMBOL(remap_pfn_range); * (but do_wp_page is only called after already making such a check; * and do_anonymous_page and do_no_page can safely check later on). */ -static inline int pte_unmap_same(struct mm_struct *mm, +static inline int pte_unmap_same(struct mm_struct *mm, pmd_t *pmd, pte_t *page_table, pte_t orig_pte) { int same = 1; #if defined(CONFIG_SMP) || defined(CONFIG_PREEMPT) if (sizeof(pte_t) > sizeof(unsigned long)) { - spin_lock(&mm->page_table_lock); + spinlock_t *ptl = pte_lockptr(mm, pmd); + spin_lock(ptl); same = pte_same(*page_table, orig_pte); - spin_unlock(&mm->page_table_lock); + spin_unlock(ptl); } #endif pte_unmap(page_table); @@ -1655,7 +1659,7 @@ static int do_swap_page(struct mm_struct *mm, struct vm_area_struct *vma, pte_t pte; int ret = VM_FAULT_MINOR; - if (!pte_unmap_same(mm, page_table, orig_pte)) + if (!pte_unmap_same(mm, pmd, page_table, orig_pte)) goto out; entry = pte_to_swp_entry(orig_pte); @@ -1773,7 +1777,7 @@ static int do_anonymous_page(struct mm_struct *mm, struct vm_area_struct *vma, page_cache_get(page); entry = mk_pte(page, vma->vm_page_prot); - ptl = &mm->page_table_lock; + ptl = pte_lockptr(mm, pmd); spin_lock(ptl); if (!pte_none(*page_table)) goto release; @@ -1934,7 +1938,7 @@ static int do_file_page(struct mm_struct *mm, struct vm_area_struct *vma, pgoff_t pgoff; int err; - if (!pte_unmap_same(mm, page_table, orig_pte)) + if (!pte_unmap_same(mm, pmd, page_table, orig_pte)) return VM_FAULT_MINOR; if (unlikely(!(vma->vm_flags & VM_NONLINEAR))) { @@ -1992,7 +1996,7 @@ static inline int handle_pte_fault(struct mm_struct *mm, pte, pmd, write_access, entry); } - ptl = &mm->page_table_lock; + ptl = pte_lockptr(mm, pmd); spin_lock(ptl); if (unlikely(!pte_same(*pte, entry))) goto unlock; diff --git a/mm/mremap.c b/mm/mremap.c index 8de77b632a20..b535438c363c 100644 --- a/mm/mremap.c +++ b/mm/mremap.c @@ -72,7 +72,7 @@ static void move_ptes(struct vm_area_struct *vma, pmd_t *old_pmd, struct address_space *mapping = NULL; struct mm_struct *mm = vma->vm_mm; pte_t *old_pte, *new_pte, pte; - spinlock_t *old_ptl; + spinlock_t *old_ptl, *new_ptl; if (vma->vm_file) { /* @@ -88,8 +88,15 @@ static void move_ptes(struct vm_area_struct *vma, pmd_t *old_pmd, new_vma->vm_truncate_count = 0; } + /* + * We don't have to worry about the ordering of src and dst + * pte locks because exclusive mmap_sem prevents deadlock. + */ old_pte = pte_offset_map_lock(mm, old_pmd, old_addr, &old_ptl); new_pte = pte_offset_map_nested(new_pmd, new_addr); + new_ptl = pte_lockptr(mm, new_pmd); + if (new_ptl != old_ptl) + spin_lock(new_ptl); for (; old_addr < old_end; old_pte++, old_addr += PAGE_SIZE, new_pte++, new_addr += PAGE_SIZE) { @@ -101,6 +108,8 @@ static void move_ptes(struct vm_area_struct *vma, pmd_t *old_pmd, set_pte_at(mm, new_addr, new_pte, pte); } + if (new_ptl != old_ptl) + spin_unlock(new_ptl); pte_unmap_nested(new_pte - 1); pte_unmap_unlock(old_pte - 1, old_ptl); if (mapping) diff --git a/mm/page_alloc.c b/mm/page_alloc.c index 0541288ebf4b..a2995a5d012c 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -154,7 +154,7 @@ static void prep_compound_page(struct page *page, unsigned long order) struct page *p = page + i; SetPageCompound(p); - p->private = (unsigned long)page; + set_page_private(p, (unsigned long)page); } } @@ -174,7 +174,7 @@ static void destroy_compound_page(struct page *page, unsigned long order) if (!PageCompound(p)) bad_page(__FUNCTION__, page); - if (p->private != (unsigned long)page) + if (page_private(p) != (unsigned long)page) bad_page(__FUNCTION__, page); ClearPageCompound(p); } @@ -187,18 +187,18 @@ static void destroy_compound_page(struct page *page, unsigned long order) * So, we don't need atomic page->flags operations here. */ static inline unsigned long page_order(struct page *page) { - return page->private; + return page_private(page); } static inline void set_page_order(struct page *page, int order) { - page->private = order; + set_page_private(page, order); __SetPagePrivate(page); } static inline void rmv_page_order(struct page *page) { __ClearPagePrivate(page); - page->private = 0; + set_page_private(page, 0); } /* @@ -238,7 +238,7 @@ __find_combined_index(unsigned long page_idx, unsigned int order) * (a) the buddy is free && * (b) the buddy is on the buddy system && * (c) a page and its buddy have the same order. - * for recording page's order, we use page->private and PG_private. + * for recording page's order, we use page_private(page) and PG_private. * */ static inline int page_is_buddy(struct page *page, int order) @@ -264,7 +264,7 @@ static inline int page_is_buddy(struct page *page, int order) * parts of the VM system. * At each level, we keep a list of pages, which are heads of continuous * free pages of length of (1 << order) and marked with PG_Private.Page's - * order is recorded in page->private field. + * order is recorded in page_private(page) field. * So when we are allocating or freeing one, we can derive the state of the * other. That is, if we allocate a small block, and both were * free, the remainder of the region must be split into blocks. @@ -463,7 +463,7 @@ static void prep_new_page(struct page *page, int order) page->flags &= ~(1 << PG_uptodate | 1 << PG_error | 1 << PG_referenced | 1 << PG_arch_1 | 1 << PG_checked | 1 << PG_mappedtodisk); - page->private = 0; + set_page_private(page, 0); set_page_refs(page, order); kernel_map_pages(page, 1 << order, 1); } diff --git a/mm/page_io.c b/mm/page_io.c index 330e00d6db00..bb2b0d53889c 100644 --- a/mm/page_io.c +++ b/mm/page_io.c @@ -91,7 +91,8 @@ int swap_writepage(struct page *page, struct writeback_control *wbc) unlock_page(page); goto out; } - bio = get_swap_bio(GFP_NOIO, page->private, page, end_swap_bio_write); + bio = get_swap_bio(GFP_NOIO, page_private(page), page, + end_swap_bio_write); if (bio == NULL) { set_page_dirty(page); unlock_page(page); @@ -115,7 +116,8 @@ int swap_readpage(struct file *file, struct page *page) BUG_ON(!PageLocked(page)); ClearPageUptodate(page); - bio = get_swap_bio(GFP_KERNEL, page->private, page, end_swap_bio_read); + bio = get_swap_bio(GFP_KERNEL, page_private(page), page, + end_swap_bio_read); if (bio == NULL) { unlock_page(page); ret = -ENOMEM; diff --git a/mm/rmap.c b/mm/rmap.c index a84bdfe582c0..a33e779d1bd8 100644 --- a/mm/rmap.c +++ b/mm/rmap.c @@ -274,7 +274,7 @@ pte_t *page_check_address(struct page *page, struct mm_struct *mm, return NULL; } - ptl = &mm->page_table_lock; + ptl = pte_lockptr(mm, pmd); spin_lock(ptl); if (pte_present(*pte) && page_to_pfn(page) == pte_pfn(*pte)) { *ptlp = ptl; @@ -550,7 +550,7 @@ static int try_to_unmap_one(struct page *page, struct vm_area_struct *vma) update_hiwater_rss(mm); if (PageAnon(page)) { - swp_entry_t entry = { .val = page->private }; + swp_entry_t entry = { .val = page_private(page) }; /* * Store the swap location in the pte. * See handle_pte_fault() ... diff --git a/mm/shmem.c b/mm/shmem.c index 37777f4c11f8..dc25565a61e9 100644 --- a/mm/shmem.c +++ b/mm/shmem.c @@ -71,9 +71,6 @@ /* Pretend that each entry is of this size in directory's i_size */ #define BOGO_DIRENT_SIZE 20 -/* Keep swapped page count in private field of indirect struct page */ -#define nr_swapped private - /* Flag allocation requirements to shmem_getpage and shmem_swp_alloc */ enum sgp_type { SGP_QUICK, /* don't try more than file page cache lookup */ @@ -324,8 +321,10 @@ static void shmem_swp_set(struct shmem_inode_info *info, swp_entry_t *entry, uns entry->val = value; info->swapped += incdec; - if ((unsigned long)(entry - info->i_direct) >= SHMEM_NR_DIRECT) - kmap_atomic_to_page(entry)->nr_swapped += incdec; + if ((unsigned long)(entry - info->i_direct) >= SHMEM_NR_DIRECT) { + struct page *page = kmap_atomic_to_page(entry); + set_page_private(page, page_private(page) + incdec); + } } /* @@ -368,9 +367,8 @@ static swp_entry_t *shmem_swp_alloc(struct shmem_inode_info *info, unsigned long spin_unlock(&info->lock); page = shmem_dir_alloc(mapping_gfp_mask(inode->i_mapping) | __GFP_ZERO); - if (page) { - page->nr_swapped = 0; - } + if (page) + set_page_private(page, 0); spin_lock(&info->lock); if (!page) { @@ -561,7 +559,7 @@ static void shmem_truncate(struct inode *inode) diroff = 0; } subdir = dir[diroff]; - if (subdir && subdir->nr_swapped) { + if (subdir && page_private(subdir)) { size = limit - idx; if (size > ENTRIES_PER_PAGE) size = ENTRIES_PER_PAGE; @@ -572,10 +570,10 @@ static void shmem_truncate(struct inode *inode) nr_swaps_freed += freed; if (offset) spin_lock(&info->lock); - subdir->nr_swapped -= freed; + set_page_private(subdir, page_private(subdir) - freed); if (offset) spin_unlock(&info->lock); - BUG_ON(subdir->nr_swapped > offset); + BUG_ON(page_private(subdir) > offset); } if (offset) offset = 0; @@ -743,7 +741,7 @@ static int shmem_unuse_inode(struct shmem_inode_info *info, swp_entry_t entry, s dir = shmem_dir_map(subdir); } subdir = *dir; - if (subdir && subdir->nr_swapped) { + if (subdir && page_private(subdir)) { ptr = shmem_swp_map(subdir); size = limit - idx; if (size > ENTRIES_PER_PAGE) diff --git a/mm/swap.c b/mm/swap.c index 21d15f99805c..b89512877ec2 100644 --- a/mm/swap.c +++ b/mm/swap.c @@ -39,7 +39,7 @@ int page_cluster; void put_page(struct page *page) { if (unlikely(PageCompound(page))) { - page = (struct page *)page->private; + page = (struct page *)page_private(page); if (put_page_testzero(page)) { void (*dtor)(struct page *page); diff --git a/mm/swap_state.c b/mm/swap_state.c index 132164f7d0a7..cafc1edcbeba 100644 --- a/mm/swap_state.c +++ b/mm/swap_state.c @@ -83,7 +83,7 @@ static int __add_to_swap_cache(struct page *page, swp_entry_t entry, page_cache_get(page); SetPageLocked(page); SetPageSwapCache(page); - page->private = entry.val; + set_page_private(page, entry.val); total_swapcache_pages++; pagecache_acct(1); } @@ -126,8 +126,8 @@ void __delete_from_swap_cache(struct page *page) BUG_ON(PageWriteback(page)); BUG_ON(PagePrivate(page)); - radix_tree_delete(&swapper_space.page_tree, page->private); - page->private = 0; + radix_tree_delete(&swapper_space.page_tree, page_private(page)); + set_page_private(page, 0); ClearPageSwapCache(page); total_swapcache_pages--; pagecache_acct(-1); @@ -197,7 +197,7 @@ void delete_from_swap_cache(struct page *page) { swp_entry_t entry; - entry.val = page->private; + entry.val = page_private(page); write_lock_irq(&swapper_space.tree_lock); __delete_from_swap_cache(page); diff --git a/mm/swapfile.c b/mm/swapfile.c index 510f0039b000..8970c0b74194 100644 --- a/mm/swapfile.c +++ b/mm/swapfile.c @@ -61,7 +61,7 @@ void swap_unplug_io_fn(struct backing_dev_info *unused_bdi, struct page *page) swp_entry_t entry; down_read(&swap_unplug_sem); - entry.val = page->private; + entry.val = page_private(page); if (PageSwapCache(page)) { struct block_device *bdev = swap_info[swp_type(entry)].bdev; struct backing_dev_info *bdi; @@ -69,8 +69,8 @@ void swap_unplug_io_fn(struct backing_dev_info *unused_bdi, struct page *page) /* * If the page is removed from swapcache from under us (with a * racy try_to_unuse/swapoff) we need an additional reference - * count to avoid reading garbage from page->private above. If - * the WARN_ON triggers during a swapoff it maybe the race + * count to avoid reading garbage from page_private(page) above. + * If the WARN_ON triggers during a swapoff it maybe the race * condition and it's harmless. However if it triggers without * swapoff it signals a problem. */ @@ -294,7 +294,7 @@ static inline int page_swapcount(struct page *page) struct swap_info_struct *p; swp_entry_t entry; - entry.val = page->private; + entry.val = page_private(page); p = swap_info_get(entry); if (p) { /* Subtract the 1 for the swap cache itself */ @@ -339,7 +339,7 @@ int remove_exclusive_swap_page(struct page *page) if (page_count(page) != 2) /* 2: us + cache */ return 0; - entry.val = page->private; + entry.val = page_private(page); p = swap_info_get(entry); if (!p) return 0; @@ -1042,7 +1042,7 @@ int page_queue_congested(struct page *page) BUG_ON(!PageLocked(page)); /* It pins the swap_info_struct */ if (PageSwapCache(page)) { - swp_entry_t entry = { .val = page->private }; + swp_entry_t entry = { .val = page_private(page) }; struct swap_info_struct *sis; sis = get_swap_info_struct(swp_type(entry)); diff --git a/mm/vmscan.c b/mm/vmscan.c index 41d1064aabfb..135bf8ca96ee 100644 --- a/mm/vmscan.c +++ b/mm/vmscan.c @@ -521,7 +521,7 @@ static int shrink_list(struct list_head *page_list, struct scan_control *sc) #ifdef CONFIG_SWAP if (PageSwapCache(page)) { - swp_entry_t swap = { .val = page->private }; + swp_entry_t swap = { .val = page_private(page) }; __delete_from_swap_cache(page); write_unlock_irq(&mapping->tree_lock); swap_free(swap); -- cgit v1.2.3 From f412ac08c9861b4791af0145934c22f1458686da Mon Sep 17 00:00:00 2001 From: Hugh Dickins Date: Sat, 29 Oct 2005 18:16:41 -0700 Subject: [PATCH] mm: fix rss and mmlist locking A couple of oddities were guarded by page_table_lock, no longer properly guarded when that is split. The mm_counters of file_rss and anon_rss: make those an atomic_t, or an atomic64_t if the architecture supports it, in such a case. Definitions by courtesy of Christoph Lameter: who spent considerable effort on more scalable ways of counting, but found insufficient benefit in practice. And adding an mm with swap to the mmlist for swapoff: the list is well- guarded by its own lock, but the list_empty check now has to be repeated inside it. Signed-off-by: Hugh Dickins Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/sched.h | 42 ++++++++++++++++++++++++++++++++++++++---- mm/memory.c | 4 +++- mm/rmap.c | 3 ++- 3 files changed, 43 insertions(+), 6 deletions(-) (limited to 'include/linux') diff --git a/include/linux/sched.h b/include/linux/sched.h index 292cb57ce38f..1c30bc308ef1 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -249,13 +249,47 @@ arch_get_unmapped_area_topdown(struct file *filp, unsigned long addr, extern void arch_unmap_area(struct mm_struct *, unsigned long); extern void arch_unmap_area_topdown(struct mm_struct *, unsigned long); +#if NR_CPUS >= CONFIG_SPLIT_PTLOCK_CPUS +/* + * The mm counters are not protected by its page_table_lock, + * so must be incremented atomically. + */ +#ifdef ATOMIC64_INIT +#define set_mm_counter(mm, member, value) atomic64_set(&(mm)->_##member, value) +#define get_mm_counter(mm, member) ((unsigned long)atomic64_read(&(mm)->_##member)) +#define add_mm_counter(mm, member, value) atomic64_add(value, &(mm)->_##member) +#define inc_mm_counter(mm, member) atomic64_inc(&(mm)->_##member) +#define dec_mm_counter(mm, member) atomic64_dec(&(mm)->_##member) +typedef atomic64_t mm_counter_t; +#else /* !ATOMIC64_INIT */ +/* + * The counters wrap back to 0 at 2^32 * PAGE_SIZE, + * that is, at 16TB if using 4kB page size. + */ +#define set_mm_counter(mm, member, value) atomic_set(&(mm)->_##member, value) +#define get_mm_counter(mm, member) ((unsigned long)atomic_read(&(mm)->_##member)) +#define add_mm_counter(mm, member, value) atomic_add(value, &(mm)->_##member) +#define inc_mm_counter(mm, member) atomic_inc(&(mm)->_##member) +#define dec_mm_counter(mm, member) atomic_dec(&(mm)->_##member) +typedef atomic_t mm_counter_t; +#endif /* !ATOMIC64_INIT */ + +#else /* NR_CPUS < CONFIG_SPLIT_PTLOCK_CPUS */ +/* + * The mm counters are protected by its page_table_lock, + * so can be incremented directly. + */ #define set_mm_counter(mm, member, value) (mm)->_##member = (value) #define get_mm_counter(mm, member) ((mm)->_##member) #define add_mm_counter(mm, member, value) (mm)->_##member += (value) #define inc_mm_counter(mm, member) (mm)->_##member++ #define dec_mm_counter(mm, member) (mm)->_##member-- -#define get_mm_rss(mm) ((mm)->_file_rss + (mm)->_anon_rss) +typedef unsigned long mm_counter_t; + +#endif /* NR_CPUS < CONFIG_SPLIT_PTLOCK_CPUS */ +#define get_mm_rss(mm) \ + (get_mm_counter(mm, file_rss) + get_mm_counter(mm, anon_rss)) #define update_hiwater_rss(mm) do { \ unsigned long _rss = get_mm_rss(mm); \ if ((mm)->hiwater_rss < _rss) \ @@ -266,8 +300,6 @@ extern void arch_unmap_area_topdown(struct mm_struct *, unsigned long); (mm)->hiwater_vm = (mm)->total_vm; \ } while (0) -typedef unsigned long mm_counter_t; - struct mm_struct { struct vm_area_struct * mmap; /* list of VMAs */ struct rb_root mm_rb; @@ -291,7 +323,9 @@ struct mm_struct { * by mmlist_lock */ - /* Special counters protected by the page_table_lock */ + /* Special counters, in some configurations protected by the + * page_table_lock, in other configurations by being atomic. + */ mm_counter_t _file_rss; mm_counter_t _anon_rss; diff --git a/mm/memory.c b/mm/memory.c index e9ef599498b5..d68421dd64ef 100644 --- a/mm/memory.c +++ b/mm/memory.c @@ -372,7 +372,9 @@ copy_one_pte(struct mm_struct *dst_mm, struct mm_struct *src_mm, /* make sure dst_mm is on swapoff's mmlist. */ if (unlikely(list_empty(&dst_mm->mmlist))) { spin_lock(&mmlist_lock); - list_add(&dst_mm->mmlist, &src_mm->mmlist); + if (list_empty(&dst_mm->mmlist)) + list_add(&dst_mm->mmlist, + &src_mm->mmlist); spin_unlock(&mmlist_lock); } } diff --git a/mm/rmap.c b/mm/rmap.c index a33e779d1bd8..a7427bbf57e4 100644 --- a/mm/rmap.c +++ b/mm/rmap.c @@ -559,7 +559,8 @@ static int try_to_unmap_one(struct page *page, struct vm_area_struct *vma) swap_duplicate(entry); if (list_empty(&mm->mmlist)) { spin_lock(&mmlist_lock); - list_add(&mm->mmlist, &init_mm.mmlist); + if (list_empty(&mm->mmlist)) + list_add(&mm->mmlist, &init_mm.mmlist); spin_unlock(&mmlist_lock); } set_pte_at(mm, address, pte, swp_entry_to_pte(entry)); -- cgit v1.2.3 From b8072f099b7829a6ff3eba618e1d079a81f753f8 Mon Sep 17 00:00:00 2001 From: Hugh Dickins Date: Sat, 29 Oct 2005 18:16:41 -0700 Subject: [PATCH] mm: update comments to pte lock Updated several references to page_table_lock in common code comments. Signed-off-by: Hugh Dickins Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/asm-generic/pgtable.h | 2 +- include/linux/mempolicy.h | 3 +-- mm/filemap.c | 6 +++--- mm/rmap.c | 10 +++++----- mm/swap_state.c | 3 +-- 5 files changed, 11 insertions(+), 13 deletions(-) (limited to 'include/linux') diff --git a/include/asm-generic/pgtable.h b/include/asm-generic/pgtable.h index ff28c8b31f58..7dca30a26c53 100644 --- a/include/asm-generic/pgtable.h +++ b/include/asm-generic/pgtable.h @@ -8,7 +8,7 @@ * - update the page tables * - inform the TLB about the new one * - * We hold the mm semaphore for reading and vma->vm_mm->page_table_lock. + * We hold the mm semaphore for reading, and the pte lock. * * Note: the old pte is known to not be writable, so we don't need to * worry about dirty bits etc getting lost. diff --git a/include/linux/mempolicy.h b/include/linux/mempolicy.h index 38e60a099399..7af8cb836e78 100644 --- a/include/linux/mempolicy.h +++ b/include/linux/mempolicy.h @@ -47,8 +47,7 @@ struct vm_area_struct; * Locking policy for interlave: * In process context there is no locking because only the process accesses * its own state. All vma manipulation is somewhat protected by a down_read on - * mmap_sem. For allocating in the interleave policy the page_table_lock - * must be also aquired to protect il_next. + * mmap_sem. * * Freeing policy: * When policy is MPOL_BIND v.zonelist is kmalloc'ed and must be kfree'd. diff --git a/mm/filemap.c b/mm/filemap.c index f560b41c8f61..036599d1177e 100644 --- a/mm/filemap.c +++ b/mm/filemap.c @@ -66,7 +66,7 @@ generic_file_direct_IO(int rw, struct kiocb *iocb, const struct iovec *iov, * * ->mmap_sem * ->i_mmap_lock - * ->page_table_lock (various places, mainly in mmap.c) + * ->page_table_lock or pte_lock (various, mainly in memory.c) * ->mapping->tree_lock (arch-dependent flush_dcache_mmap_lock) * * ->mmap_sem @@ -86,9 +86,9 @@ generic_file_direct_IO(int rw, struct kiocb *iocb, const struct iovec *iov, * ->anon_vma.lock (vma_adjust) * * ->anon_vma.lock - * ->page_table_lock (anon_vma_prepare and various) + * ->page_table_lock or pte_lock (anon_vma_prepare and various) * - * ->page_table_lock + * ->page_table_lock or pte_lock * ->swap_lock (try_to_unmap_one) * ->private_lock (try_to_unmap_one) * ->tree_lock (try_to_unmap_one) diff --git a/mm/rmap.c b/mm/rmap.c index a7427bbf57e4..914d04b98bee 100644 --- a/mm/rmap.c +++ b/mm/rmap.c @@ -32,7 +32,7 @@ * page->flags PG_locked (lock_page) * mapping->i_mmap_lock * anon_vma->lock - * mm->page_table_lock + * mm->page_table_lock or pte_lock * zone->lru_lock (in mark_page_accessed) * swap_lock (in swap_duplicate, swap_info_get) * mmlist_lock (in mmput, drain_mmlist and others) @@ -244,7 +244,7 @@ unsigned long page_address_in_vma(struct page *page, struct vm_area_struct *vma) /* * Check that @page is mapped at @address into @mm. * - * On success returns with mapped pte and locked mm->page_table_lock. + * On success returns with pte mapped and locked. */ pte_t *page_check_address(struct page *page, struct mm_struct *mm, unsigned long address, spinlock_t **ptlp) @@ -445,7 +445,7 @@ int page_referenced(struct page *page, int is_locked, int ignore_token) * @vma: the vm area in which the mapping is added * @address: the user virtual address mapped * - * The caller needs to hold the mm->page_table_lock. + * The caller needs to hold the pte lock. */ void page_add_anon_rmap(struct page *page, struct vm_area_struct *vma, unsigned long address) @@ -468,7 +468,7 @@ void page_add_anon_rmap(struct page *page, * page_add_file_rmap - add pte mapping to a file page * @page: the page to add the mapping to * - * The caller needs to hold the mm->page_table_lock. + * The caller needs to hold the pte lock. */ void page_add_file_rmap(struct page *page) { @@ -483,7 +483,7 @@ void page_add_file_rmap(struct page *page) * page_remove_rmap - take down pte mapping from a page * @page: page to remove mapping from * - * Caller needs to hold the mm->page_table_lock. + * The caller needs to hold the pte lock. */ void page_remove_rmap(struct page *page) { diff --git a/mm/swap_state.c b/mm/swap_state.c index cafc1edcbeba..dfd9a46755b8 100644 --- a/mm/swap_state.c +++ b/mm/swap_state.c @@ -259,8 +259,7 @@ static inline void free_swap_cache(struct page *page) /* * Perform a free_page(), also freeing any swap cache associated with - * this page if it is the last user of the page. Can not do a lock_page, - * as we are holding the page_table_lock spinlock. + * this page if it is the last user of the page. */ void free_page_and_swap_cache(struct page *page) { -- cgit v1.2.3 From 4ca644d970bf2542623228a4624af356d20ca267 Mon Sep 17 00:00:00 2001 From: Dave Hansen Date: Sat, 29 Oct 2005 18:16:51 -0700 Subject: [PATCH] memory hotplug prep: __section_nr helper A little helper that we use in the hotplug code. Signed-off-by: Dave Hansen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/mmzone.h | 1 + mm/sparse.c | 25 +++++++++++++++++++++++++ 2 files changed, 26 insertions(+) (limited to 'include/linux') diff --git a/include/linux/mmzone.h b/include/linux/mmzone.h index 7519eb4191e7..4674145bb63d 100644 --- a/include/linux/mmzone.h +++ b/include/linux/mmzone.h @@ -509,6 +509,7 @@ static inline struct mem_section *__nr_to_section(unsigned long nr) return NULL; return &mem_section[SECTION_NR_TO_ROOT(nr)][nr & SECTION_ROOT_MASK]; } +extern int __section_nr(struct mem_section* ms); /* * We use the lower bits of the mem_map pointer to store diff --git a/mm/sparse.c b/mm/sparse.c index 347249a4917a..0d3bd4bf3aaa 100644 --- a/mm/sparse.c +++ b/mm/sparse.c @@ -72,6 +72,31 @@ static inline int sparse_index_init(unsigned long section_nr, int nid) } #endif +/* + * Although written for the SPARSEMEM_EXTREME case, this happens + * to also work for the flat array case becase + * NR_SECTION_ROOTS==NR_MEM_SECTIONS. + */ +int __section_nr(struct mem_section* ms) +{ + unsigned long root_nr; + struct mem_section* root; + + for (root_nr = 0; + root_nr < NR_MEM_SECTIONS; + root_nr += SECTIONS_PER_ROOT) { + root = __nr_to_section(root_nr); + + if (!root) + continue; + + if ((ms >= root) && (ms < (root + SECTIONS_PER_ROOT))) + break; + } + + return (root_nr * SECTIONS_PER_ROOT) + (ms - root); +} + /* Record a memory area against a node. */ void memory_present(int nid, unsigned long start, unsigned long end) { -- cgit v1.2.3 From 208d54e5513c0c02d85af0990901354c74364d5c Mon Sep 17 00:00:00 2001 From: Dave Hansen Date: Sat, 29 Oct 2005 18:16:52 -0700 Subject: [PATCH] memory hotplug locking: node_size_lock pgdat->node_size_lock is basically only neeeded in one place in the normal code: show_mem(), which is the arch-specific sysrq-m printing function. Strictly speaking, the architectures not doing memory hotplug do no need this locking in show_mem(). However, they are all included for completeness. This should also make any future consolidation of all of the implementations a little more straightforward. This lock is also held in the sparsemem code during a memory removal, as sections are invalidated. This is the place there pfn_valid() is made false for a memory area that's being removed. The lock is only required when doing pfn_valid() operations on memory which the user does not already have a reference on the page, such as in show_mem(). Signed-off-by: Dave Hansen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/alpha/mm/numa.c | 3 +++ arch/i386/mm/pgtable.c | 3 +++ arch/ia64/mm/discontig.c | 7 ++++++- arch/m32r/mm/init.c | 9 ++++++++- arch/parisc/mm/init.c | 3 +++ arch/ppc64/mm/init.c | 6 ++++++ include/linux/memory_hotplug.h | 34 ++++++++++++++++++++++++++++++++++ include/linux/mmzone.h | 12 ++++++++++++ mm/page_alloc.c | 1 + 9 files changed, 76 insertions(+), 2 deletions(-) create mode 100644 include/linux/memory_hotplug.h (limited to 'include/linux') diff --git a/arch/alpha/mm/numa.c b/arch/alpha/mm/numa.c index c7481d59b6df..6d5251254f68 100644 --- a/arch/alpha/mm/numa.c +++ b/arch/alpha/mm/numa.c @@ -371,6 +371,8 @@ show_mem(void) show_free_areas(); printk("Free swap: %6ldkB\n", nr_swap_pages<<(PAGE_SHIFT-10)); for_each_online_node(nid) { + unsigned long flags; + pgdat_resize_lock(NODE_DATA(nid), &flags); i = node_spanned_pages(nid); while (i-- > 0) { struct page *page = nid_page_nr(nid, i); @@ -384,6 +386,7 @@ show_mem(void) else shared += page_count(page) - 1; } + pgdat_resize_unlock(NODE_DATA(nid), &flags); } printk("%ld pages of RAM\n",total); printk("%ld free pages\n",free); diff --git a/arch/i386/mm/pgtable.c b/arch/i386/mm/pgtable.c index 39c099f15b5f..9db3242103be 100644 --- a/arch/i386/mm/pgtable.c +++ b/arch/i386/mm/pgtable.c @@ -31,11 +31,13 @@ void show_mem(void) pg_data_t *pgdat; unsigned long i; struct page_state ps; + unsigned long flags; printk(KERN_INFO "Mem-info:\n"); show_free_areas(); printk(KERN_INFO "Free swap: %6ldkB\n", nr_swap_pages<<(PAGE_SHIFT-10)); for_each_pgdat(pgdat) { + pgdat_resize_lock(pgdat, &flags); for (i = 0; i < pgdat->node_spanned_pages; ++i) { page = pgdat_page_nr(pgdat, i); total++; @@ -48,6 +50,7 @@ void show_mem(void) else if (page_count(page)) shared += page_count(page) - 1; } + pgdat_resize_unlock(pgdat, &flags); } printk(KERN_INFO "%d pages of RAM\n", total); printk(KERN_INFO "%d pages of HIGHMEM\n", highmem); diff --git a/arch/ia64/mm/discontig.c b/arch/ia64/mm/discontig.c index a3788fb84809..a88cdb7232f8 100644 --- a/arch/ia64/mm/discontig.c +++ b/arch/ia64/mm/discontig.c @@ -555,9 +555,13 @@ void show_mem(void) show_free_areas(); printk("Free swap: %6ldkB\n", nr_swap_pages<<(PAGE_SHIFT-10)); for_each_pgdat(pgdat) { - unsigned long present = pgdat->node_present_pages; + unsigned long present; + unsigned long flags; int shared = 0, cached = 0, reserved = 0; + printk("Node ID: %d\n", pgdat->node_id); + pgdat_resize_lock(pgdat, &flags); + present = pgdat->node_present_pages; for(i = 0; i < pgdat->node_spanned_pages; i++) { struct page *page; if (pfn_valid(pgdat->node_start_pfn + i)) @@ -571,6 +575,7 @@ void show_mem(void) else if (page_count(page)) shared += page_count(page)-1; } + pgdat_resize_unlock(pgdat, &flags); total_present += present; total_reserved += reserved; total_cached += cached; diff --git a/arch/m32r/mm/init.c b/arch/m32r/mm/init.c index d9a40b1fe8ba..6facf15b04f3 100644 --- a/arch/m32r/mm/init.c +++ b/arch/m32r/mm/init.c @@ -48,6 +48,8 @@ void show_mem(void) show_free_areas(); printk("Free swap: %6ldkB\n",nr_swap_pages<<(PAGE_SHIFT-10)); for_each_pgdat(pgdat) { + unsigned long flags; + pgdat_resize_lock(pgdat, &flags); for (i = 0; i < pgdat->node_spanned_pages; ++i) { page = pgdat_page_nr(pgdat, i); total++; @@ -60,6 +62,7 @@ void show_mem(void) else if (page_count(page)) shared += page_count(page) - 1; } + pgdat_resize_unlock(pgdat, &flags); } printk("%d pages of RAM\n", total); printk("%d pages of HIGHMEM\n",highmem); @@ -150,10 +153,14 @@ int __init reservedpages_count(void) int reservedpages, nid, i; reservedpages = 0; - for_each_online_node(nid) + for_each_online_node(nid) { + unsigned long flags; + pgdat_resize_lock(NODE_DATA(nid), &flags); for (i = 0 ; i < MAX_LOW_PFN(nid) - START_PFN(nid) ; i++) if (PageReserved(nid_page_nr(nid, i))) reservedpages++; + pgdat_resize_unlock(NODE_DATA(nid), &flags); + } return reservedpages; } diff --git a/arch/parisc/mm/init.c b/arch/parisc/mm/init.c index 2886ad70db48..29b998e430e6 100644 --- a/arch/parisc/mm/init.c +++ b/arch/parisc/mm/init.c @@ -505,7 +505,9 @@ void show_mem(void) for (j = node_start_pfn(i); j < node_end_pfn(i); j++) { struct page *p; + unsigned long flags; + pgdat_resize_lock(NODE_DATA(i), &flags); p = nid_page_nr(i, j) - node_start_pfn(i); total++; @@ -517,6 +519,7 @@ void show_mem(void) free++; else shared += page_count(p) - 1; + pgdat_resize_unlock(NODE_DATA(i), &flags); } } #endif diff --git a/arch/ppc64/mm/init.c b/arch/ppc64/mm/init.c index a45584b3440c..975b26de34d6 100644 --- a/arch/ppc64/mm/init.c +++ b/arch/ppc64/mm/init.c @@ -104,6 +104,8 @@ void show_mem(void) show_free_areas(); printk("Free swap: %6ldkB\n", nr_swap_pages<<(PAGE_SHIFT-10)); for_each_pgdat(pgdat) { + unsigned long flags; + pgdat_resize_lock(pgdat, &flags); for (i = 0; i < pgdat->node_spanned_pages; i++) { page = pgdat_page_nr(pgdat, i); total++; @@ -114,6 +116,7 @@ void show_mem(void) else if (page_count(page)) shared += page_count(page) - 1; } + pgdat_resize_unlock(pgdat, &flags); } printk("%ld pages of RAM\n", total); printk("%ld reserved pages\n", reserved); @@ -647,11 +650,14 @@ void __init mem_init(void) #endif for_each_pgdat(pgdat) { + unsigned long flags; + pgdat_resize_lock(pgdat, &flags); for (i = 0; i < pgdat->node_spanned_pages; i++) { page = pgdat_page_nr(pgdat, i); if (PageReserved(page)) reservedpages++; } + pgdat_resize_unlock(pgdat, &flags); } codesize = (unsigned long)&_etext - (unsigned long)&_stext; diff --git a/include/linux/memory_hotplug.h b/include/linux/memory_hotplug.h new file mode 100644 index 000000000000..e8103be9d528 --- /dev/null +++ b/include/linux/memory_hotplug.h @@ -0,0 +1,34 @@ +#ifndef __LINUX_MEMORY_HOTPLUG_H +#define __LINUX_MEMORY_HOTPLUG_H + +#include +#include + +#ifdef CONFIG_MEMORY_HOTPLUG +/* + * pgdat resizing functions + */ +static inline +void pgdat_resize_lock(struct pglist_data *pgdat, unsigned long *flags) +{ + spin_lock_irqsave(&pgdat->node_size_lock, *flags); +} +static inline +void pgdat_resize_unlock(struct pglist_data *pgdat, unsigned long *flags) +{ + spin_lock_irqrestore(&pgdat->node_size_lock, *flags); +} +static inline +void pgdat_resize_init(struct pglist_data *pgdat) +{ + spin_lock_init(&pgdat->node_size_lock); +} +#else /* ! CONFIG_MEMORY_HOTPLUG */ +/* + * Stub functions for when hotplug is off + */ +static inline void pgdat_resize_lock(struct pglist_data *p, unsigned long *f) {} +static inline void pgdat_resize_unlock(struct pglist_data *p, unsigned long *f) {} +static inline void pgdat_resize_init(struct pglist_data *pgdat) {} +#endif +#endif /* __LINUX_MEMORY_HOTPLUG_H */ diff --git a/include/linux/mmzone.h b/include/linux/mmzone.h index 4674145bb63d..e050d68963a1 100644 --- a/include/linux/mmzone.h +++ b/include/linux/mmzone.h @@ -273,6 +273,16 @@ typedef struct pglist_data { struct page *node_mem_map; #endif struct bootmem_data *bdata; +#ifdef CONFIG_MEMORY_HOTPLUG + /* + * Must be held any time you expect node_start_pfn, node_present_pages + * or node_spanned_pages stay constant. Holding this will also + * guarantee that any pfn_valid() stays that way. + * + * Nests above zone->lock and zone->size_seqlock. + */ + spinlock_t node_size_lock; +#endif unsigned long node_start_pfn; unsigned long node_present_pages; /* total number of physical pages */ unsigned long node_spanned_pages; /* total size of physical page @@ -293,6 +303,8 @@ typedef struct pglist_data { #endif #define nid_page_nr(nid, pagenr) pgdat_page_nr(NODE_DATA(nid),(pagenr)) +#include + extern struct pglist_data *pgdat_list; void __get_zone_counts(unsigned long *active, unsigned long *inactive, diff --git a/mm/page_alloc.c b/mm/page_alloc.c index a51ef94eec33..32fad6d23200 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -1958,6 +1958,7 @@ static void __init free_area_init_core(struct pglist_data *pgdat, int nid = pgdat->node_id; unsigned long zone_start_pfn = pgdat->node_start_pfn; + pgdat_resize_init(pgdat); pgdat->nr_zones = 0; init_waitqueue_head(&pgdat->kswapd_wait); pgdat->kswapd_max_order = 0; -- cgit v1.2.3 From bdc8cb984576ab5b550c8b24c6fa111a873503e3 Mon Sep 17 00:00:00 2001 From: Dave Hansen Date: Sat, 29 Oct 2005 18:16:53 -0700 Subject: [PATCH] memory hotplug locking: zone span seqlock See the "fixup bad_range()" patch for more information, but this actually creates a the lock to protect things making assumptions about a zone's size staying constant at runtime. Signed-off-by: Dave Hansen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/memory_hotplug.h | 39 +++++++++++++++++++++++++++++++++++++-- include/linux/mmzone.h | 15 +++++++++++++++ mm/page_alloc.c | 19 ++++++++++++++----- 3 files changed, 66 insertions(+), 7 deletions(-) (limited to 'include/linux') diff --git a/include/linux/memory_hotplug.h b/include/linux/memory_hotplug.h index e8103be9d528..4b08bc947578 100644 --- a/include/linux/memory_hotplug.h +++ b/include/linux/memory_hotplug.h @@ -16,13 +16,36 @@ void pgdat_resize_lock(struct pglist_data *pgdat, unsigned long *flags) static inline void pgdat_resize_unlock(struct pglist_data *pgdat, unsigned long *flags) { - spin_lock_irqrestore(&pgdat->node_size_lock, *flags); + spin_unlock_irqrestore(&pgdat->node_size_lock, *flags); } static inline void pgdat_resize_init(struct pglist_data *pgdat) { spin_lock_init(&pgdat->node_size_lock); } +/* + * Zone resizing functions + */ +static inline unsigned zone_span_seqbegin(struct zone *zone) +{ + return read_seqbegin(&zone->span_seqlock); +} +static inline int zone_span_seqretry(struct zone *zone, unsigned iv) +{ + return read_seqretry(&zone->span_seqlock, iv); +} +static inline void zone_span_writelock(struct zone *zone) +{ + write_seqlock(&zone->span_seqlock); +} +static inline void zone_span_writeunlock(struct zone *zone) +{ + write_sequnlock(&zone->span_seqlock); +} +static inline void zone_seqlock_init(struct zone *zone) +{ + seqlock_init(&zone->span_seqlock); +} #else /* ! CONFIG_MEMORY_HOTPLUG */ /* * Stub functions for when hotplug is off @@ -30,5 +53,17 @@ void pgdat_resize_init(struct pglist_data *pgdat) static inline void pgdat_resize_lock(struct pglist_data *p, unsigned long *f) {} static inline void pgdat_resize_unlock(struct pglist_data *p, unsigned long *f) {} static inline void pgdat_resize_init(struct pglist_data *pgdat) {} -#endif + +static inline unsigned zone_span_seqbegin(struct zone *zone) +{ + return 0; +} +static inline int zone_span_seqretry(struct zone *zone, unsigned iv) +{ + return 0; +} +static inline void zone_span_writelock(struct zone *zone) {} +static inline void zone_span_writeunlock(struct zone *zone) {} +static inline void zone_seqlock_init(struct zone *zone) {} +#endif /* ! CONFIG_MEMORY_HOTPLUG */ #endif /* __LINUX_MEMORY_HOTPLUG_H */ diff --git a/include/linux/mmzone.h b/include/linux/mmzone.h index e050d68963a1..f5fa3082fd6a 100644 --- a/include/linux/mmzone.h +++ b/include/linux/mmzone.h @@ -12,6 +12,7 @@ #include #include #include +#include #include /* Free memory management - zoned buddy allocator. */ @@ -137,6 +138,10 @@ struct zone { * free areas of different sizes */ spinlock_t lock; +#ifdef CONFIG_MEMORY_HOTPLUG + /* see spanned/present_pages for more description */ + seqlock_t span_seqlock; +#endif struct free_area free_area[MAX_ORDER]; @@ -220,6 +225,16 @@ struct zone { /* zone_start_pfn == zone_start_paddr >> PAGE_SHIFT */ unsigned long zone_start_pfn; + /* + * zone_start_pfn, spanned_pages and present_pages are all + * protected by span_seqlock. It is a seqlock because it has + * to be read outside of zone->lock, and it is done in the main + * allocator path. But, it is written quite infrequently. + * + * The lock is declared along with zone->lock because it is + * frequently read in proximity to zone->lock. It's good to + * give them a chance of being in the same cacheline. + */ unsigned long spanned_pages; /* total size, including holes */ unsigned long present_pages; /* amount of memory (excluding holes) */ diff --git a/mm/page_alloc.c b/mm/page_alloc.c index 32fad6d23200..817635f2ab62 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -33,6 +33,7 @@ #include #include #include +#include #include #include @@ -80,12 +81,19 @@ unsigned long __initdata nr_all_pages; static int page_outside_zone_boundaries(struct zone *zone, struct page *page) { - if (page_to_pfn(page) >= zone->zone_start_pfn + zone->spanned_pages) - return 1; - if (page_to_pfn(page) < zone->zone_start_pfn) - return 1; + int ret = 0; + unsigned seq; + unsigned long pfn = page_to_pfn(page); - return 0; + do { + seq = zone_span_seqbegin(zone); + if (pfn >= zone->zone_start_pfn + zone->spanned_pages) + ret = 1; + else if (pfn < zone->zone_start_pfn) + ret = 1; + } while (zone_span_seqretry(zone, seq)); + + return ret; } static int page_is_consistent(struct zone *zone, struct page *page) @@ -1980,6 +1988,7 @@ static void __init free_area_init_core(struct pglist_data *pgdat, zone->name = zone_names[j]; spin_lock_init(&zone->lock); spin_lock_init(&zone->lru_lock); + zone_seqlock_init(zone); zone->zone_pgdat = pgdat; zone->free_pages = 0; -- cgit v1.2.3 From 3947be1969a9ce455ec30f60ef51efb10e4323d1 Mon Sep 17 00:00:00 2001 From: Dave Hansen Date: Sat, 29 Oct 2005 18:16:54 -0700 Subject: [PATCH] memory hotplug: sysfs and add/remove functions This adds generic memory add/remove and supporting functions for memory hotplug into a new file as well as a memory hotplug kernel config option. Individual architecture patches will follow. For now, disable memory hotplug when swsusp is enabled. There's a lot of churn there right now. We'll fix it up properly once it calms down. Signed-off-by: Matt Tolentino Signed-off-by: Dave Hansen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/base/Makefile | 1 + drivers/base/init.c | 2 + drivers/base/memory.c | 455 +++++++++++++++++++++++++++++++++++++++++ include/linux/memory.h | 94 +++++++++ include/linux/memory_hotplug.h | 35 ++++ include/linux/mm.h | 1 + mm/Kconfig | 8 + mm/Makefile | 2 +- mm/memory_hotplug.c | 178 ++++++++++++++++ mm/page_alloc.c | 4 +- 10 files changed, 777 insertions(+), 3 deletions(-) create mode 100644 drivers/base/memory.c create mode 100644 include/linux/memory.h create mode 100644 mm/memory_hotplug.c (limited to 'include/linux') diff --git a/drivers/base/Makefile b/drivers/base/Makefile index 66d9c4643fc1..f12898d53078 100644 --- a/drivers/base/Makefile +++ b/drivers/base/Makefile @@ -7,6 +7,7 @@ obj-y := core.o sys.o bus.o dd.o \ obj-y += power/ obj-$(CONFIG_FW_LOADER) += firmware_class.o obj-$(CONFIG_NUMA) += node.o +obj-$(CONFIG_MEMORY_HOTPLUG) += memory.o ifeq ($(CONFIG_DEBUG_DRIVER),y) EXTRA_CFLAGS += -DDEBUG diff --git a/drivers/base/init.c b/drivers/base/init.c index 84e604e25c4f..c648914b9cde 100644 --- a/drivers/base/init.c +++ b/drivers/base/init.c @@ -9,6 +9,7 @@ #include #include +#include #include "base.h" @@ -33,5 +34,6 @@ void __init driver_init(void) platform_bus_init(); system_bus_init(); cpu_dev_init(); + memory_dev_init(); attribute_container_init(); } diff --git a/drivers/base/memory.c b/drivers/base/memory.c new file mode 100644 index 000000000000..785cb6e6b91c --- /dev/null +++ b/drivers/base/memory.c @@ -0,0 +1,455 @@ +/* + * drivers/base/memory.c - basic Memory class support + * + * Written by Matt Tolentino + * Dave Hansen + * + * This file provides the necessary infrastructure to represent + * a SPARSEMEM-memory-model system's physical memory in /sysfs. + * All arch-independent code that assumes MEMORY_HOTPLUG requires + * SPARSEMEM should be contained here, or in mm/memory_hotplug.c. + */ + +#include +#include +#include +#include /* capable() */ +#include +#include +#include +#include +#include +#include +#include +#include + +#define MEMORY_CLASS_NAME "memory" + +static struct sysdev_class memory_sysdev_class = { + set_kset_name(MEMORY_CLASS_NAME), +}; +EXPORT_SYMBOL(memory_sysdev_class); + +static char *memory_hotplug_name(struct kset *kset, struct kobject *kobj) +{ + return MEMORY_CLASS_NAME; +} + +static int memory_hotplug(struct kset *kset, struct kobject *kobj, char **envp, + int num_envp, char *buffer, int buffer_size) +{ + int retval = 0; + + return retval; +} + +static struct kset_hotplug_ops memory_hotplug_ops = { + .name = memory_hotplug_name, + .hotplug = memory_hotplug, +}; + +static struct notifier_block *memory_chain; + +static int register_memory_notifier(struct notifier_block *nb) +{ + return notifier_chain_register(&memory_chain, nb); +} + +static void unregister_memory_notifier(struct notifier_block *nb) +{ + notifier_chain_unregister(&memory_chain, nb); +} + +/* + * register_memory - Setup a sysfs device for a memory block + */ +static int +register_memory(struct memory_block *memory, struct mem_section *section, + struct node *root) +{ + int error; + + memory->sysdev.cls = &memory_sysdev_class; + memory->sysdev.id = __section_nr(section); + + error = sysdev_register(&memory->sysdev); + + if (root && !error) + error = sysfs_create_link(&root->sysdev.kobj, + &memory->sysdev.kobj, + kobject_name(&memory->sysdev.kobj)); + + return error; +} + +static void +unregister_memory(struct memory_block *memory, struct mem_section *section, + struct node *root) +{ + BUG_ON(memory->sysdev.cls != &memory_sysdev_class); + BUG_ON(memory->sysdev.id != __section_nr(section)); + + sysdev_unregister(&memory->sysdev); + if (root) + sysfs_remove_link(&root->sysdev.kobj, + kobject_name(&memory->sysdev.kobj)); +} + +/* + * use this as the physical section index that this memsection + * uses. + */ + +static ssize_t show_mem_phys_index(struct sys_device *dev, char *buf) +{ + struct memory_block *mem = + container_of(dev, struct memory_block, sysdev); + return sprintf(buf, "%08lx\n", mem->phys_index); +} + +/* + * online, offline, going offline, etc. + */ +static ssize_t show_mem_state(struct sys_device *dev, char *buf) +{ + struct memory_block *mem = + container_of(dev, struct memory_block, sysdev); + ssize_t len = 0; + + /* + * We can probably put these states in a nice little array + * so that they're not open-coded + */ + switch (mem->state) { + case MEM_ONLINE: + len = sprintf(buf, "online\n"); + break; + case MEM_OFFLINE: + len = sprintf(buf, "offline\n"); + break; + case MEM_GOING_OFFLINE: + len = sprintf(buf, "going-offline\n"); + break; + default: + len = sprintf(buf, "ERROR-UNKNOWN-%ld\n", + mem->state); + WARN_ON(1); + break; + } + + return len; +} + +static inline int memory_notify(unsigned long val, void *v) +{ + return notifier_call_chain(&memory_chain, val, v); +} + +/* + * MEMORY_HOTPLUG depends on SPARSEMEM in mm/Kconfig, so it is + * OK to have direct references to sparsemem variables in here. + */ +static int +memory_block_action(struct memory_block *mem, unsigned long action) +{ + int i; + unsigned long psection; + unsigned long start_pfn, start_paddr; + struct page *first_page; + int ret; + int old_state = mem->state; + + psection = mem->phys_index; + first_page = pfn_to_page(psection << PFN_SECTION_SHIFT); + + /* + * The probe routines leave the pages reserved, just + * as the bootmem code does. Make sure they're still + * that way. + */ + if (action == MEM_ONLINE) { + for (i = 0; i < PAGES_PER_SECTION; i++) { + if (PageReserved(first_page+i)) + continue; + + printk(KERN_WARNING "section number %ld page number %d " + "not reserved, was it already online? \n", + psection, i); + return -EBUSY; + } + } + + switch (action) { + case MEM_ONLINE: + start_pfn = page_to_pfn(first_page); + ret = online_pages(start_pfn, PAGES_PER_SECTION); + break; + case MEM_OFFLINE: + mem->state = MEM_GOING_OFFLINE; + memory_notify(MEM_GOING_OFFLINE, NULL); + start_paddr = page_to_pfn(first_page) << PAGE_SHIFT; + ret = remove_memory(start_paddr, + PAGES_PER_SECTION << PAGE_SHIFT); + if (ret) { + mem->state = old_state; + break; + } + memory_notify(MEM_MAPPING_INVALID, NULL); + break; + default: + printk(KERN_WARNING "%s(%p, %ld) unknown action: %ld\n", + __FUNCTION__, mem, action, action); + WARN_ON(1); + ret = -EINVAL; + } + /* + * For now, only notify on successful memory operations + */ + if (!ret) + memory_notify(action, NULL); + + return ret; +} + +static int memory_block_change_state(struct memory_block *mem, + unsigned long to_state, unsigned long from_state_req) +{ + int ret = 0; + down(&mem->state_sem); + + if (mem->state != from_state_req) { + ret = -EINVAL; + goto out; + } + + ret = memory_block_action(mem, to_state); + if (!ret) + mem->state = to_state; + +out: + up(&mem->state_sem); + return ret; +} + +static ssize_t +store_mem_state(struct sys_device *dev, const char *buf, size_t count) +{ + struct memory_block *mem; + unsigned int phys_section_nr; + int ret = -EINVAL; + + mem = container_of(dev, struct memory_block, sysdev); + phys_section_nr = mem->phys_index; + + if (!valid_section_nr(phys_section_nr)) + goto out; + + if (!strncmp(buf, "online", min((int)count, 6))) + ret = memory_block_change_state(mem, MEM_ONLINE, MEM_OFFLINE); + else if(!strncmp(buf, "offline", min((int)count, 7))) + ret = memory_block_change_state(mem, MEM_OFFLINE, MEM_ONLINE); +out: + if (ret) + return ret; + return count; +} + +/* + * phys_device is a bad name for this. What I really want + * is a way to differentiate between memory ranges that + * are part of physical devices that constitute + * a complete removable unit or fru. + * i.e. do these ranges belong to the same physical device, + * s.t. if I offline all of these sections I can then + * remove the physical device? + */ +static ssize_t show_phys_device(struct sys_device *dev, char *buf) +{ + struct memory_block *mem = + container_of(dev, struct memory_block, sysdev); + return sprintf(buf, "%d\n", mem->phys_device); +} + +static SYSDEV_ATTR(phys_index, 0444, show_mem_phys_index, NULL); +static SYSDEV_ATTR(state, 0644, show_mem_state, store_mem_state); +static SYSDEV_ATTR(phys_device, 0444, show_phys_device, NULL); + +#define mem_create_simple_file(mem, attr_name) \ + sysdev_create_file(&mem->sysdev, &attr_##attr_name) +#define mem_remove_simple_file(mem, attr_name) \ + sysdev_remove_file(&mem->sysdev, &attr_##attr_name) + +/* + * Block size attribute stuff + */ +static ssize_t +print_block_size(struct class *class, char *buf) +{ + return sprintf(buf, "%lx\n", (unsigned long)PAGES_PER_SECTION * PAGE_SIZE); +} + +static CLASS_ATTR(block_size_bytes, 0444, print_block_size, NULL); + +static int block_size_init(void) +{ + sysfs_create_file(&memory_sysdev_class.kset.kobj, + &class_attr_block_size_bytes.attr); + return 0; +} + +/* + * Some architectures will have custom drivers to do this, and + * will not need to do it from userspace. The fake hot-add code + * as well as ppc64 will do all of their discovery in userspace + * and will require this interface. + */ +#ifdef CONFIG_ARCH_MEMORY_PROBE +static ssize_t +memory_probe_store(struct class *class, const char __user *buf, size_t count) +{ + u64 phys_addr; + int ret; + + phys_addr = simple_strtoull(buf, NULL, 0); + + ret = add_memory(phys_addr, PAGES_PER_SECTION << PAGE_SHIFT); + + if (ret) + count = ret; + + return count; +} +static CLASS_ATTR(probe, 0700, NULL, memory_probe_store); + +static int memory_probe_init(void) +{ + sysfs_create_file(&memory_sysdev_class.kset.kobj, + &class_attr_probe.attr); + return 0; +} +#else +#define memory_probe_init(...) do {} while (0) +#endif + +/* + * Note that phys_device is optional. It is here to allow for + * differentiation between which *physical* devices each + * section belongs to... + */ + +static int add_memory_block(unsigned long node_id, struct mem_section *section, + unsigned long state, int phys_device) +{ + size_t size = sizeof(struct memory_block); + struct memory_block *mem = kmalloc(size, GFP_KERNEL); + int ret = 0; + + if (!mem) + return -ENOMEM; + + memset(mem, 0, size); + + mem->phys_index = __section_nr(section); + mem->state = state; + init_MUTEX(&mem->state_sem); + mem->phys_device = phys_device; + + ret = register_memory(mem, section, NULL); + if (!ret) + ret = mem_create_simple_file(mem, phys_index); + if (!ret) + ret = mem_create_simple_file(mem, state); + if (!ret) + ret = mem_create_simple_file(mem, phys_device); + + return ret; +} + +/* + * For now, we have a linear search to go find the appropriate + * memory_block corresponding to a particular phys_index. If + * this gets to be a real problem, we can always use a radix + * tree or something here. + * + * This could be made generic for all sysdev classes. + */ +static struct memory_block *find_memory_block(struct mem_section *section) +{ + struct kobject *kobj; + struct sys_device *sysdev; + struct memory_block *mem; + char name[sizeof(MEMORY_CLASS_NAME) + 9 + 1]; + + /* + * This only works because we know that section == sysdev->id + * slightly redundant with sysdev_register() + */ + sprintf(&name[0], "%s%d", MEMORY_CLASS_NAME, __section_nr(section)); + + kobj = kset_find_obj(&memory_sysdev_class.kset, name); + if (!kobj) + return NULL; + + sysdev = container_of(kobj, struct sys_device, kobj); + mem = container_of(sysdev, struct memory_block, sysdev); + + return mem; +} + +int remove_memory_block(unsigned long node_id, struct mem_section *section, + int phys_device) +{ + struct memory_block *mem; + + mem = find_memory_block(section); + mem_remove_simple_file(mem, phys_index); + mem_remove_simple_file(mem, state); + mem_remove_simple_file(mem, phys_device); + unregister_memory(mem, section, NULL); + + return 0; +} + +/* + * need an interface for the VM to add new memory regions, + * but without onlining it. + */ +int register_new_memory(struct mem_section *section) +{ + return add_memory_block(0, section, MEM_OFFLINE, 0); +} + +int unregister_memory_section(struct mem_section *section) +{ + if (!valid_section(section)) + return -EINVAL; + + return remove_memory_block(0, section, 0); +} + +/* + * Initialize the sysfs support for memory devices... + */ +int __init memory_dev_init(void) +{ + unsigned int i; + int ret; + + memory_sysdev_class.kset.hotplug_ops = &memory_hotplug_ops; + ret = sysdev_class_register(&memory_sysdev_class); + + /* + * Create entries for memory sections that were found + * during boot and have been initialized + */ + for (i = 0; i < NR_MEM_SECTIONS; i++) { + if (!valid_section_nr(i)) + continue; + add_memory_block(0, __nr_to_section(i), MEM_ONLINE, 0); + } + + memory_probe_init(); + block_size_init(); + + return ret; +} diff --git a/include/linux/memory.h b/include/linux/memory.h new file mode 100644 index 000000000000..0def328ab5cf --- /dev/null +++ b/include/linux/memory.h @@ -0,0 +1,94 @@ +/* + * include/linux/memory.h - generic memory definition + * + * This is mainly for topological representation. We define the + * basic "struct memory_block" here, which can be embedded in per-arch + * definitions or NUMA information. + * + * Basic handling of the devices is done in drivers/base/memory.c + * and system devices are handled in drivers/base/sys.c. + * + * Memory block are exported via sysfs in the class/memory/devices/ + * directory. + * + */ +#ifndef _LINUX_MEMORY_H_ +#define _LINUX_MEMORY_H_ + +#include +#include +#include + +#include + +struct memory_block { + unsigned long phys_index; + unsigned long state; + /* + * This serializes all state change requests. It isn't + * held during creation because the control files are + * created long after the critical areas during + * initialization. + */ + struct semaphore state_sem; + int phys_device; /* to which fru does this belong? */ + void *hw; /* optional pointer to fw/hw data */ + int (*phys_callback)(struct memory_block *); + struct sys_device sysdev; +}; + +/* These states are exposed to userspace as text strings in sysfs */ +#define MEM_ONLINE (1<<0) /* exposed to userspace */ +#define MEM_GOING_OFFLINE (1<<1) /* exposed to userspace */ +#define MEM_OFFLINE (1<<2) /* exposed to userspace */ + +/* + * All of these states are currently kernel-internal for notifying + * kernel components and architectures. + * + * For MEM_MAPPING_INVALID, all notifier chains with priority >0 + * are called before pfn_to_page() becomes invalid. The priority=0 + * entry is reserved for the function that actually makes + * pfn_to_page() stop working. Any notifiers that want to be called + * after that should have priority <0. + */ +#define MEM_MAPPING_INVALID (1<<3) + +#ifndef CONFIG_MEMORY_HOTPLUG +static inline int memory_dev_init(void) +{ + return 0; +} +static inline int register_memory_notifier(struct notifier_block *nb) +{ + return 0; +} +static inline void unregister_memory_notifier(struct notifier_block *nb) +{ +} +#else +extern int register_memory(struct memory_block *, struct mem_section *section, struct node *); +extern int register_new_memory(struct mem_section *); +extern int unregister_memory_section(struct mem_section *); +extern int memory_dev_init(void); +extern int register_memory_notifier(struct notifier_block *nb); +extern void unregister_memory_notifier(struct notifier_block *nb); + +#define CONFIG_MEM_BLOCK_SIZE (PAGES_PER_SECTION< #include +#include +#include #ifdef CONFIG_MEMORY_HOTPLUG /* @@ -46,6 +48,19 @@ static inline void zone_seqlock_init(struct zone *zone) { seqlock_init(&zone->span_seqlock); } +extern int zone_grow_free_lists(struct zone *zone, unsigned long new_nr_pages); +extern int zone_grow_waitqueues(struct zone *zone, unsigned long nr_pages); +extern int add_one_highpage(struct page *page, int pfn, int bad_ppro); +/* need some defines for these for archs that don't support it */ +extern void online_page(struct page *page); +/* VM interface that may be used by firmware interface */ +extern int add_memory(u64 start, u64 size); +extern int remove_memory(u64 start, u64 size); +extern int online_pages(unsigned long, unsigned long); + +/* reasonably generic interface to expand the physical pages in a zone */ +extern int __add_pages(struct zone *zone, unsigned long start_pfn, + unsigned long nr_pages); #else /* ! CONFIG_MEMORY_HOTPLUG */ /* * Stub functions for when hotplug is off @@ -65,5 +80,25 @@ static inline int zone_span_seqretry(struct zone *zone, unsigned iv) static inline void zone_span_writelock(struct zone *zone) {} static inline void zone_span_writeunlock(struct zone *zone) {} static inline void zone_seqlock_init(struct zone *zone) {} + +static inline int mhp_notimplemented(const char *func) +{ + printk(KERN_WARNING "%s() called, with CONFIG_MEMORY_HOTPLUG disabled\n", func); + dump_stack(); + return -ENOSYS; +} + +static inline int __add_pages(struct zone *zone, unsigned long start_pfn, + unsigned long nr_pages) +{ + return mhp_notimplemented(__FUNCTION__); +} #endif /* ! CONFIG_MEMORY_HOTPLUG */ +static inline int __remove_pages(struct zone *zone, unsigned long start_pfn, + unsigned long nr_pages) +{ + printk(KERN_WARNING "%s() called, not yet supported\n", __FUNCTION__); + dump_stack(); + return -ENOSYS; +} #endif /* __LINUX_MEMORY_HOTPLUG_H */ diff --git a/include/linux/mm.h b/include/linux/mm.h index 8a514eca40d5..5c1fb0a2e806 100644 --- a/include/linux/mm.h +++ b/include/linux/mm.h @@ -840,6 +840,7 @@ extern void free_area_init_node(int nid, pg_data_t *pgdat, unsigned long * zones_size, unsigned long zone_start_pfn, unsigned long *zholes_size); extern void memmap_init_zone(unsigned long, int, unsigned long, unsigned long); +extern void setup_per_zone_pages_min(void); extern void mem_init(void); extern void show_mem(void); extern void si_meminfo(struct sysinfo * val); diff --git a/mm/Kconfig b/mm/Kconfig index f35a550ba4b9..1a4473fcb2ca 100644 --- a/mm/Kconfig +++ b/mm/Kconfig @@ -112,6 +112,14 @@ config SPARSEMEM_EXTREME def_bool y depends on SPARSEMEM && !SPARSEMEM_STATIC +# eventually, we can have this option just 'select SPARSEMEM' +config MEMORY_HOTPLUG + bool "Allow for memory hot-add" + depends on SPARSEMEM && HOTPLUG && !SOFTWARE_SUSPEND + +comment "Memory hotplug is currently incompatible with Software Suspend" + depends on SPARSEMEM && HOTPLUG && SOFTWARE_SUSPEND + # Heavily threaded applications may benefit from splitting the mm-wide # page_table_lock, so that faults on different parts of the user address # space can be handled with less contention: split it at this NR_CPUS. diff --git a/mm/Makefile b/mm/Makefile index 4cd69e3ce421..2fa6d2ca9f28 100644 --- a/mm/Makefile +++ b/mm/Makefile @@ -18,5 +18,5 @@ obj-$(CONFIG_NUMA) += mempolicy.o obj-$(CONFIG_SPARSEMEM) += sparse.o obj-$(CONFIG_SHMEM) += shmem.o obj-$(CONFIG_TINY_SHMEM) += tiny-shmem.o - +obj-$(CONFIG_MEMORY_HOTPLUG) += memory_hotplug.o obj-$(CONFIG_FS_XIP) += filemap_xip.o diff --git a/mm/memory_hotplug.c b/mm/memory_hotplug.c new file mode 100644 index 000000000000..855e0fc928b3 --- /dev/null +++ b/mm/memory_hotplug.c @@ -0,0 +1,178 @@ +/* + * linux/mm/memory_hotplug.c + * + * Copyright (C) + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include + +static struct page *__kmalloc_section_memmap(unsigned long nr_pages) +{ + struct page *page, *ret; + unsigned long memmap_size = sizeof(struct page) * nr_pages; + + page = alloc_pages(GFP_KERNEL, get_order(memmap_size)); + if (page) + goto got_map_page; + + ret = vmalloc(memmap_size); + if (ret) + goto got_map_ptr; + + return NULL; +got_map_page: + ret = (struct page *)pfn_to_kaddr(page_to_pfn(page)); +got_map_ptr: + memset(ret, 0, memmap_size); + + return ret; +} + +extern void zonetable_add(struct zone *zone, int nid, int zid, unsigned long pfn, + unsigned long size); +static void __add_zone(struct zone *zone, unsigned long phys_start_pfn) +{ + struct pglist_data *pgdat = zone->zone_pgdat; + int nr_pages = PAGES_PER_SECTION; + int nid = pgdat->node_id; + int zone_type; + + zone_type = zone - pgdat->node_zones; + memmap_init_zone(nr_pages, nid, zone_type, phys_start_pfn); + zonetable_add(zone, nid, zone_type, phys_start_pfn, nr_pages); +} + +extern int sparse_add_one_section(struct zone *, unsigned long, + struct page *mem_map); +static int __add_section(struct zone *zone, unsigned long phys_start_pfn) +{ + struct pglist_data *pgdat = zone->zone_pgdat; + int nr_pages = PAGES_PER_SECTION; + struct page *memmap; + int ret; + + /* + * This can potentially allocate memory, and does its own + * internal locking. + */ + sparse_index_init(pfn_to_section_nr(phys_start_pfn), pgdat->node_id); + + pgdat_resize_lock(pgdat, &flags); + memmap = __kmalloc_section_memmap(nr_pages); + ret = sparse_add_one_section(zone, phys_start_pfn, memmap); + pgdat_resize_unlock(pgdat, &flags); + + if (ret <= 0) { + /* the mem_map didn't get used */ + if (memmap >= (struct page *)VMALLOC_START && + memmap < (struct page *)VMALLOC_END) + vfree(memmap); + else + free_pages((unsigned long)memmap, + get_order(sizeof(struct page) * nr_pages)); + } + + if (ret < 0) + return ret; + + __add_zone(zone, phys_start_pfn); + return register_new_memory(__pfn_to_section(phys_start_pfn)); +} + +/* + * Reasonably generic function for adding memory. It is + * expected that archs that support memory hotplug will + * call this function after deciding the zone to which to + * add the new pages. + */ +int __add_pages(struct zone *zone, unsigned long phys_start_pfn, + unsigned long nr_pages) +{ + unsigned long i; + int err = 0; + + for (i = 0; i < nr_pages; i += PAGES_PER_SECTION) { + err = __add_section(zone, phys_start_pfn + i); + + if (err) + break; + } + + return err; +} + +static void grow_zone_span(struct zone *zone, + unsigned long start_pfn, unsigned long end_pfn) +{ + unsigned long old_zone_end_pfn; + + zone_span_writelock(zone); + + old_zone_end_pfn = zone->zone_start_pfn + zone->spanned_pages; + if (start_pfn < zone->zone_start_pfn) + zone->zone_start_pfn = start_pfn; + + if (end_pfn > old_zone_end_pfn) + zone->spanned_pages = end_pfn - zone->zone_start_pfn; + + zone_span_writeunlock(zone); +} + +static void grow_pgdat_span(struct pglist_data *pgdat, + unsigned long start_pfn, unsigned long end_pfn) +{ + unsigned long old_pgdat_end_pfn = + pgdat->node_start_pfn + pgdat->node_spanned_pages; + + if (start_pfn < pgdat->node_start_pfn) + pgdat->node_start_pfn = start_pfn; + + if (end_pfn > old_pgdat_end_pfn) + pgdat->node_spanned_pages = end_pfn - pgdat->node_spanned_pages; +} + +int online_pages(unsigned long pfn, unsigned long nr_pages) +{ + unsigned long i; + unsigned long flags; + unsigned long onlined_pages = 0; + struct zone *zone; + + /* + * This doesn't need a lock to do pfn_to_page(). + * The section can't be removed here because of the + * memory_block->state_sem. + */ + zone = page_zone(pfn_to_page(pfn)); + pgdat_resize_lock(zone->zone_pgdat, &flags); + grow_zone_span(zone, pfn, pfn + nr_pages); + grow_pgdat_span(zone->zone_pgdat, pfn, pfn + nr_pages); + pgdat_resize_unlock(zone->zone_pgdat, &flags); + + for (i = 0; i < nr_pages; i++) { + struct page *page = pfn_to_page(pfn + i); + online_page(page); + onlined_pages++; + } + zone->present_pages += onlined_pages; + + return 0; +} diff --git a/mm/page_alloc.c b/mm/page_alloc.c index 817635f2ab62..183abf39b445 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -1686,7 +1686,7 @@ static void __init calculate_zone_totalpages(struct pglist_data *pgdat, * up by free_all_bootmem() once the early boot process is * done. Non-atomic initialization, single-pass. */ -void __init memmap_init_zone(unsigned long size, int nid, unsigned long zone, +void __devinit memmap_init_zone(unsigned long size, int nid, unsigned long zone, unsigned long start_pfn) { struct page *page; @@ -2407,7 +2407,7 @@ static void setup_per_zone_lowmem_reserve(void) * that the pages_{min,low,high} values for each zone are set correctly * with respect to min_free_kbytes. */ -static void setup_per_zone_pages_min(void) +void setup_per_zone_pages_min(void) { unsigned long pages_min = min_free_kbytes >> (PAGE_SHIFT - 10); unsigned long lowmem_pages = 0; -- cgit v1.2.3 From a7dac447bb9cef27d4d29cdf63e2d7809c50b1f4 Mon Sep 17 00:00:00 2001 From: Jeff Garzik Date: Sun, 30 Oct 2005 04:44:42 -0500 Subject: [libata] change ata_qc_complete() to take error mask as second arg The second argument to ata_qc_complete() was being used for two purposes: communicate the ATA Status register to the completion function, and indicate an error. On legacy PCI IDE hardware, the latter is often implicit in the former. On more modern hardware, the driver often completely emulated a Status register value, passing ATA_ERR as an indication that something went wrong. Now that previous code changes have eliminated the need to use drv_stat arg to communicate the ATA Status register value, we can convert it to a mask of possible error classes. This will lead to more flexible error handling in the future. --- drivers/scsi/ahci.c | 4 ++-- drivers/scsi/libata-core.c | 31 +++++++++++++++--------------- drivers/scsi/libata-scsi.c | 46 +++++++++++++++++++++++++++------------------ drivers/scsi/libata.h | 2 +- drivers/scsi/pdc_adma.c | 13 ++++++++----- drivers/scsi/sata_mv.c | 11 ++++++----- drivers/scsi/sata_promise.c | 16 +++++++--------- drivers/scsi/sata_qstor.c | 7 ++++--- drivers/scsi/sata_sil24.c | 12 +++++++----- drivers/scsi/sata_sx4.c | 10 +++++----- include/linux/libata.h | 28 +++++++++++++++++++++++++-- 11 files changed, 109 insertions(+), 71 deletions(-) (limited to 'include/linux') diff --git a/drivers/scsi/ahci.c b/drivers/scsi/ahci.c index 03829aedfd39..5efb3c50aa8a 100644 --- a/drivers/scsi/ahci.c +++ b/drivers/scsi/ahci.c @@ -600,7 +600,7 @@ static void ahci_eng_timeout(struct ata_port *ap) * not being called from the SCSI EH. */ qc->scsidone = scsi_finish_command; - ata_qc_complete(qc, ATA_ERR); + ata_qc_complete(qc, AC_ERR_OTHER); } spin_unlock_irqrestore(&host_set->lock, flags); @@ -629,7 +629,7 @@ static inline int ahci_host_intr(struct ata_port *ap, struct ata_queued_cmd *qc) if (status & PORT_IRQ_FATAL) { ahci_intr_error(ap, status); if (qc) - ata_qc_complete(qc, ATA_ERR); + ata_qc_complete(qc, AC_ERR_OTHER); } return 1; diff --git a/drivers/scsi/libata-core.c b/drivers/scsi/libata-core.c index 771bc7d376bc..cc089f1fb114 100644 --- a/drivers/scsi/libata-core.c +++ b/drivers/scsi/libata-core.c @@ -2663,7 +2663,7 @@ static int ata_sg_setup(struct ata_queued_cmd *qc) * None. (grabs host lock) */ -void ata_poll_qc_complete(struct ata_queued_cmd *qc, u8 drv_stat) +void ata_poll_qc_complete(struct ata_queued_cmd *qc, unsigned int err_mask) { struct ata_port *ap = qc->ap; unsigned long flags; @@ -2671,7 +2671,7 @@ void ata_poll_qc_complete(struct ata_queued_cmd *qc, u8 drv_stat) spin_lock_irqsave(&ap->host_set->lock, flags); ap->flags &= ~ATA_FLAG_NOINTR; ata_irq_on(ap); - ata_qc_complete(qc, drv_stat); + ata_qc_complete(qc, err_mask); spin_unlock_irqrestore(&ap->host_set->lock, flags); } @@ -2768,7 +2768,7 @@ static int ata_pio_complete (struct ata_port *ap) ap->hsm_task_state = HSM_ST_IDLE; - ata_poll_qc_complete(qc, drv_stat); + ata_poll_qc_complete(qc, 0); /* another command may start at this point */ @@ -3136,18 +3136,15 @@ static void ata_pio_block(struct ata_port *ap) static void ata_pio_error(struct ata_port *ap) { struct ata_queued_cmd *qc; - u8 drv_stat; + + printk(KERN_WARNING "ata%u: PIO error\n", ap->id); qc = ata_qc_from_tag(ap, ap->active_tag); assert(qc != NULL); - drv_stat = ata_chk_status(ap); - printk(KERN_WARNING "ata%u: PIO error, drv_stat 0x%x\n", - ap->id, drv_stat); - ap->hsm_task_state = HSM_ST_IDLE; - ata_poll_qc_complete(qc, drv_stat | ATA_ERR); + ata_poll_qc_complete(qc, AC_ERR_ATA_BUS); } static void ata_pio_task(void *_data) @@ -3270,7 +3267,7 @@ static void ata_qc_timeout(struct ata_queued_cmd *qc) ap->id, qc->tf.command, drv_stat, host_stat); /* complete taskfile transaction */ - ata_qc_complete(qc, drv_stat); + ata_qc_complete(qc, ac_err_mask(drv_stat)); break; } @@ -3375,7 +3372,7 @@ struct ata_queued_cmd *ata_qc_new_init(struct ata_port *ap, return qc; } -int ata_qc_complete_noop(struct ata_queued_cmd *qc, u8 drv_stat) +int ata_qc_complete_noop(struct ata_queued_cmd *qc, unsigned int err_mask) { return 0; } @@ -3434,7 +3431,7 @@ void ata_qc_free(struct ata_queued_cmd *qc) * spin_lock_irqsave(host_set lock) */ -void ata_qc_complete(struct ata_queued_cmd *qc, u8 drv_stat) +void ata_qc_complete(struct ata_queued_cmd *qc, unsigned int err_mask) { int rc; @@ -3451,7 +3448,7 @@ void ata_qc_complete(struct ata_queued_cmd *qc, u8 drv_stat) qc->flags &= ~ATA_QCFLAG_ACTIVE; /* call completion callback */ - rc = qc->complete_fn(qc, drv_stat); + rc = qc->complete_fn(qc, err_mask); /* if callback indicates not to complete command (non-zero), * return immediately @@ -3889,7 +3886,7 @@ inline unsigned int ata_host_intr (struct ata_port *ap, ap->ops->irq_clear(ap); /* complete taskfile transaction */ - ata_qc_complete(qc, status); + ata_qc_complete(qc, ac_err_mask(status)); break; default: @@ -3984,7 +3981,7 @@ static void atapi_packet_task(void *_data) /* sleep-wait for BSY to clear */ DPRINTK("busy wait\n"); if (ata_busy_sleep(ap, ATA_TMOUT_CDB_QUICK, ATA_TMOUT_CDB)) - goto err_out; + goto err_out_status; /* make sure DRQ is set */ status = ata_chk_status(ap); @@ -4021,8 +4018,10 @@ static void atapi_packet_task(void *_data) return; +err_out_status: + status = ata_chk_status(ap); err_out: - ata_poll_qc_complete(qc, ATA_ERR); + ata_poll_qc_complete(qc, __ac_err_mask(status)); } diff --git a/drivers/scsi/libata-scsi.c b/drivers/scsi/libata-scsi.c index 89a04b1a5a0e..1e3792f86fcf 100644 --- a/drivers/scsi/libata-scsi.c +++ b/drivers/scsi/libata-scsi.c @@ -560,7 +560,7 @@ void ata_gen_ata_desc_sense(struct ata_queued_cmd *qc) * Use ata_to_sense_error() to map status register bits * onto sense key, asc & ascq. */ - if (unlikely(tf->command & (ATA_BUSY | ATA_DF | ATA_ERR | ATA_DRQ))) { + if (tf->command & (ATA_BUSY | ATA_DF | ATA_ERR | ATA_DRQ)) { ata_to_sense_error(qc->ap->id, tf->command, tf->feature, &sb[1], &sb[2], &sb[3]); sb[1] &= 0x0f; @@ -635,7 +635,7 @@ void ata_gen_fixed_sense(struct ata_queued_cmd *qc) * Use ata_to_sense_error() to map status register bits * onto sense key, asc & ascq. */ - if (unlikely(tf->command & (ATA_BUSY | ATA_DF | ATA_ERR | ATA_DRQ))) { + if (tf->command & (ATA_BUSY | ATA_DF | ATA_ERR | ATA_DRQ)) { ata_to_sense_error(qc->ap->id, tf->command, tf->feature, &sb[2], &sb[12], &sb[13]); sb[2] &= 0x0f; @@ -644,7 +644,11 @@ void ata_gen_fixed_sense(struct ata_queued_cmd *qc) sb[0] = 0x70; sb[7] = 0x0a; - if (tf->flags & ATA_TFLAG_LBA && !(tf->flags & ATA_TFLAG_LBA48)) { + if (tf->flags & ATA_TFLAG_LBA48) { + /* TODO: find solution for LBA48 descriptors */ + } + + else if (tf->flags & ATA_TFLAG_LBA) { /* A small (28b) LBA will fit in the 32b info field */ sb[0] |= 0x80; /* set valid bit */ sb[3] = tf->device & 0x0f; @@ -652,6 +656,10 @@ void ata_gen_fixed_sense(struct ata_queued_cmd *qc) sb[5] = tf->lbam; sb[6] = tf->lbal; } + + else { + /* TODO: C/H/S */ + } } /** @@ -1199,10 +1207,12 @@ nothing_to_do: return 1; } -static int ata_scsi_qc_complete(struct ata_queued_cmd *qc, u8 drv_stat) +static int ata_scsi_qc_complete(struct ata_queued_cmd *qc, + unsigned int err_mask) { struct scsi_cmnd *cmd = qc->scsicmd; - int need_sense = drv_stat & (ATA_ERR | ATA_BUSY | ATA_DRQ); + u8 *cdb = cmd->cmnd; + int need_sense = (err_mask != 0); /* For ATA pass thru (SAT) commands, generate a sense block if * user mandated it or if there's an error. Note that if we @@ -1211,8 +1221,8 @@ static int ata_scsi_qc_complete(struct ata_queued_cmd *qc, u8 drv_stat) * whether the command completed successfully or not. If there * was no error, SK, ASC and ASCQ will all be zero. */ - if (((cmd->cmnd[0] == ATA_16) || (cmd->cmnd[0] == ATA_12)) && - ((cmd->cmnd[2] & 0x20) || need_sense)) { + if (((cdb[0] == ATA_16) || (cdb[0] == ATA_12)) && + ((cdb[2] & 0x20) || need_sense)) { ata_gen_ata_desc_sense(qc); } else { if (!need_sense) { @@ -1995,21 +2005,13 @@ void atapi_request_sense(struct ata_port *ap, struct ata_device *dev, DPRINTK("EXIT\n"); } -static int atapi_qc_complete(struct ata_queued_cmd *qc, u8 drv_stat) +static int atapi_qc_complete(struct ata_queued_cmd *qc, unsigned int err_mask) { struct scsi_cmnd *cmd = qc->scsicmd; - VPRINTK("ENTER, drv_stat == 0x%x\n", drv_stat); - - if (unlikely(drv_stat & (ATA_BUSY | ATA_DRQ))) - /* FIXME: not quite right; we don't want the - * translation of taskfile registers into - * a sense descriptors, since that's only - * correct for ATA, not ATAPI - */ - ata_gen_ata_desc_sense(qc); + VPRINTK("ENTER, err_mask 0x%X\n", err_mask); - else if (unlikely(drv_stat & ATA_ERR)) { + if (unlikely(err_mask & AC_ERR_DEV)) { DPRINTK("request check condition\n"); /* FIXME: command completion with check condition @@ -2026,6 +2028,14 @@ static int atapi_qc_complete(struct ata_queued_cmd *qc, u8 drv_stat) return 1; } + else if (unlikely(err_mask)) + /* FIXME: not quite right; we don't want the + * translation of taskfile registers into + * a sense descriptors, since that's only + * correct for ATA, not ATAPI + */ + ata_gen_ata_desc_sense(qc); + else { u8 *scsicmd = cmd->cmnd; diff --git a/drivers/scsi/libata.h b/drivers/scsi/libata.h index 65c264b91136..10ecd9e15e4f 100644 --- a/drivers/scsi/libata.h +++ b/drivers/scsi/libata.h @@ -39,7 +39,7 @@ struct ata_scsi_args { /* libata-core.c */ extern int atapi_enabled; -extern int ata_qc_complete_noop(struct ata_queued_cmd *qc, u8 drv_stat); +extern int ata_qc_complete_noop(struct ata_queued_cmd *qc, unsigned int err_mask); extern struct ata_queued_cmd *ata_qc_new_init(struct ata_port *ap, struct ata_device *dev); extern void ata_rwcmd_protocol(struct ata_queued_cmd *qc); diff --git a/drivers/scsi/pdc_adma.c b/drivers/scsi/pdc_adma.c index af99feb9d237..74b3b2c629b6 100644 --- a/drivers/scsi/pdc_adma.c +++ b/drivers/scsi/pdc_adma.c @@ -451,7 +451,7 @@ static inline unsigned int adma_intr_pkt(struct ata_host_set *host_set) struct adma_port_priv *pp; struct ata_queued_cmd *qc; void __iomem *chan = ADMA_REGS(mmio_base, port_no); - u8 drv_stat = 0, status = readb(chan + ADMA_STATUS); + u8 status = readb(chan + ADMA_STATUS); if (status == 0) continue; @@ -464,11 +464,14 @@ static inline unsigned int adma_intr_pkt(struct ata_host_set *host_set) continue; qc = ata_qc_from_tag(ap, ap->active_tag); if (qc && (!(qc->tf.ctl & ATA_NIEN))) { + unsigned int err_mask = 0; + if ((status & (aPERR | aPSD | aUIRQ))) - drv_stat = ATA_ERR; + err_mask = AC_ERR_OTHER; else if (pp->pkt[0] != cDONE) - drv_stat = ATA_ERR; - ata_qc_complete(qc, drv_stat); + err_mask = AC_ERR_OTHER; + + ata_qc_complete(qc, err_mask); } } return handled; @@ -498,7 +501,7 @@ static inline unsigned int adma_intr_mmio(struct ata_host_set *host_set) /* complete taskfile transaction */ pp->state = adma_state_idle; - ata_qc_complete(qc, status); + ata_qc_complete(qc, ac_err_mask(status)); handled = 1; } } diff --git a/drivers/scsi/sata_mv.c b/drivers/scsi/sata_mv.c index dcef5fe8600b..936d1ce5575f 100644 --- a/drivers/scsi/sata_mv.c +++ b/drivers/scsi/sata_mv.c @@ -1065,6 +1065,7 @@ static void mv_host_intr(struct ata_host_set *host_set, u32 relevant, struct ata_queued_cmd *qc; u32 hc_irq_cause; int shift, port, port0, hard_port, handled; + unsigned int err_mask; u8 ata_status = 0; if (hc == 0) { @@ -1100,15 +1101,15 @@ static void mv_host_intr(struct ata_host_set *host_set, u32 relevant, handled++; } + err_mask = ac_err_mask(ata_status); + shift = port << 1; /* (port * 2) */ if (port >= MV_PORTS_PER_HC) { shift++; /* skip bit 8 in the HC Main IRQ reg */ } if ((PORT0_ERR << shift) & relevant) { mv_err_intr(ap); - /* OR in ATA_ERR to ensure libata knows we took one */ - ata_status = readb((void __iomem *) - ap->ioaddr.status_addr) | ATA_ERR; + err_mask |= AC_ERR_OTHER; handled++; } @@ -1118,7 +1119,7 @@ static void mv_host_intr(struct ata_host_set *host_set, u32 relevant, VPRINTK("port %u IRQ found for qc, " "ata_status 0x%x\n", port,ata_status); /* mark qc status appropriately */ - ata_qc_complete(qc, ata_status); + ata_qc_complete(qc, err_mask); } } } @@ -1294,7 +1295,7 @@ static void mv_eng_timeout(struct ata_port *ap) */ spin_lock_irqsave(&ap->host_set->lock, flags); qc->scsidone = scsi_finish_command; - ata_qc_complete(qc, ATA_ERR); + ata_qc_complete(qc, AC_ERR_OTHER); spin_unlock_irqrestore(&ap->host_set->lock, flags); } } diff --git a/drivers/scsi/sata_promise.c b/drivers/scsi/sata_promise.c index 63911f16b6ec..8f41702275db 100644 --- a/drivers/scsi/sata_promise.c +++ b/drivers/scsi/sata_promise.c @@ -399,7 +399,8 @@ static void pdc_eng_timeout(struct ata_port *ap) case ATA_PROT_DMA: case ATA_PROT_NODATA: printk(KERN_ERR "ata%u: command timeout\n", ap->id); - ata_qc_complete(qc, ata_wait_idle(ap) | ATA_ERR); + drv_stat = ata_wait_idle(ap); + ata_qc_complete(qc, __ac_err_mask(drv_stat)); break; default: @@ -408,7 +409,7 @@ static void pdc_eng_timeout(struct ata_port *ap) printk(KERN_ERR "ata%u: unknown timeout, cmd 0x%x stat 0x%x\n", ap->id, qc->tf.command, drv_stat); - ata_qc_complete(qc, drv_stat); + ata_qc_complete(qc, ac_err_mask(drv_stat)); break; } @@ -420,24 +421,21 @@ out: static inline unsigned int pdc_host_intr( struct ata_port *ap, struct ata_queued_cmd *qc) { - u8 status; - unsigned int handled = 0, have_err = 0; + unsigned int handled = 0, err_mask = 0; u32 tmp; void __iomem *mmio = (void __iomem *) ap->ioaddr.cmd_addr + PDC_GLOBAL_CTL; tmp = readl(mmio); if (tmp & PDC_ERR_MASK) { - have_err = 1; + err_mask = AC_ERR_DEV; pdc_reset_port(ap); } switch (qc->tf.protocol) { case ATA_PROT_DMA: case ATA_PROT_NODATA: - status = ata_wait_idle(ap); - if (have_err) - status |= ATA_ERR; - ata_qc_complete(qc, status); + err_mask |= ac_err_mask(ata_wait_idle(ap)); + ata_qc_complete(qc, err_mask); handled = 1; break; diff --git a/drivers/scsi/sata_qstor.c b/drivers/scsi/sata_qstor.c index 1aaf3304d397..d95a02fa7afb 100644 --- a/drivers/scsi/sata_qstor.c +++ b/drivers/scsi/sata_qstor.c @@ -400,11 +400,12 @@ static inline unsigned int qs_intr_pkt(struct ata_host_set *host_set) qc = ata_qc_from_tag(ap, ap->active_tag); if (qc && (!(qc->tf.ctl & ATA_NIEN))) { switch (sHST) { - case 0: /* sucessful CPB */ + case 0: /* successful CPB */ case 3: /* device error */ pp->state = qs_state_idle; qs_enter_reg_mode(qc->ap); - ata_qc_complete(qc, sDST); + ata_qc_complete(qc, + ac_err_mask(sDST)); break; default: break; @@ -441,7 +442,7 @@ static inline unsigned int qs_intr_mmio(struct ata_host_set *host_set) /* complete taskfile transaction */ pp->state = qs_state_idle; - ata_qc_complete(qc, status); + ata_qc_complete(qc, ac_err_mask(status)); handled = 1; } } diff --git a/drivers/scsi/sata_sil24.c b/drivers/scsi/sata_sil24.c index e18a1e2bb65e..4afe2b15b803 100644 --- a/drivers/scsi/sata_sil24.c +++ b/drivers/scsi/sata_sil24.c @@ -498,7 +498,7 @@ static void sil24_eng_timeout(struct ata_port *ap) qc = ata_qc_from_tag(ap, ap->active_tag); if (!qc) { - printk(KERN_ERR "ata%u: BUG: tiemout without command\n", + printk(KERN_ERR "ata%u: BUG: timeout without command\n", ap->id); return; } @@ -512,7 +512,7 @@ static void sil24_eng_timeout(struct ata_port *ap) */ printk(KERN_ERR "ata%u: command timeout\n", ap->id); qc->scsidone = scsi_finish_command; - ata_qc_complete(qc, ATA_ERR); + ata_qc_complete(qc, AC_ERR_OTHER); sil24_reset_controller(ap); } @@ -523,6 +523,7 @@ static void sil24_error_intr(struct ata_port *ap, u32 slot_stat) struct sil24_port_priv *pp = ap->private_data; void __iomem *port = (void __iomem *)ap->ioaddr.cmd_addr; u32 irq_stat, cmd_err, sstatus, serror; + unsigned int err_mask; irq_stat = readl(port + PORT_IRQ_STAT); writel(irq_stat, port + PORT_IRQ_STAT); /* clear irq */ @@ -550,17 +551,18 @@ static void sil24_error_intr(struct ata_port *ap, u32 slot_stat) * Device is reporting error, tf registers are valid. */ sil24_update_tf(ap); + err_mask = ac_err_mask(pp->tf.command); } else { /* * Other errors. libata currently doesn't have any * mechanism to report these errors. Just turn on * ATA_ERR. */ - pp->tf.command = ATA_ERR; + err_mask = AC_ERR_OTHER; } if (qc) - ata_qc_complete(qc, pp->tf.command); + ata_qc_complete(qc, err_mask); sil24_reset_controller(ap); } @@ -585,7 +587,7 @@ static inline void sil24_host_intr(struct ata_port *ap) sil24_update_tf(ap); if (qc) - ata_qc_complete(qc, pp->tf.command); + ata_qc_complete(qc, ac_err_mask(pp->tf.command)); } else sil24_error_intr(ap, slot_stat); } diff --git a/drivers/scsi/sata_sx4.c b/drivers/scsi/sata_sx4.c index af08f4f650c1..d9a8baff0d4d 100644 --- a/drivers/scsi/sata_sx4.c +++ b/drivers/scsi/sata_sx4.c @@ -718,7 +718,7 @@ static inline unsigned int pdc20621_host_intr( struct ata_port *ap, VPRINTK("ata%u: read hdma, 0x%x 0x%x\n", ap->id, readl(mmio + 0x104), readl(mmio + PDC_HDMA_CTLSTAT)); /* get drive status; clear intr; complete txn */ - ata_qc_complete(qc, ata_wait_idle(ap)); + ata_qc_complete(qc, ac_err_mask(ata_wait_idle(ap))); pdc20621_pop_hdma(qc); } @@ -756,7 +756,7 @@ static inline unsigned int pdc20621_host_intr( struct ata_port *ap, VPRINTK("ata%u: write ata, 0x%x 0x%x\n", ap->id, readl(mmio + 0x104), readl(mmio + PDC_HDMA_CTLSTAT)); /* get drive status; clear intr; complete txn */ - ata_qc_complete(qc, ata_wait_idle(ap)); + ata_qc_complete(qc, ac_err_mask(ata_wait_idle(ap))); pdc20621_pop_hdma(qc); } handled = 1; @@ -766,7 +766,7 @@ static inline unsigned int pdc20621_host_intr( struct ata_port *ap, status = ata_busy_wait(ap, ATA_BUSY | ATA_DRQ, 1000); DPRINTK("BUS_NODATA (drv_stat 0x%X)\n", status); - ata_qc_complete(qc, status); + ata_qc_complete(qc, ac_err_mask(status)); handled = 1; } else { @@ -881,7 +881,7 @@ static void pdc_eng_timeout(struct ata_port *ap) case ATA_PROT_DMA: case ATA_PROT_NODATA: printk(KERN_ERR "ata%u: command timeout\n", ap->id); - ata_qc_complete(qc, ata_wait_idle(ap) | ATA_ERR); + ata_qc_complete(qc, __ac_err_mask(ata_wait_idle(ap))); break; default: @@ -890,7 +890,7 @@ static void pdc_eng_timeout(struct ata_port *ap) printk(KERN_ERR "ata%u: unknown timeout, cmd 0x%x stat 0x%x\n", ap->id, qc->tf.command, drv_stat); - ata_qc_complete(qc, drv_stat); + ata_qc_complete(qc, ac_err_mask(drv_stat)); break; } diff --git a/include/linux/libata.h b/include/linux/libata.h index a4cce9936a80..0ba3af7a1236 100644 --- a/include/linux/libata.h +++ b/include/linux/libata.h @@ -172,6 +172,13 @@ enum hsm_task_states { HSM_ST_ERR, }; +enum ata_completion_errors { + AC_ERR_OTHER = (1 << 0), + AC_ERR_DEV = (1 << 1), + AC_ERR_ATA_BUS = (1 << 2), + AC_ERR_HOST_BUS = (1 << 3), +}; + /* forward declarations */ struct scsi_device; struct ata_port_operations; @@ -179,7 +186,7 @@ struct ata_port; struct ata_queued_cmd; /* typedefs */ -typedef int (*ata_qc_cb_t) (struct ata_queued_cmd *qc, u8 drv_stat); +typedef int (*ata_qc_cb_t) (struct ata_queued_cmd *qc, unsigned int err_mask); struct ata_ioports { unsigned long cmd_addr; @@ -453,7 +460,7 @@ extern void ata_bmdma_start (struct ata_queued_cmd *qc); extern void ata_bmdma_stop(struct ata_queued_cmd *qc); extern u8 ata_bmdma_status(struct ata_port *ap); extern void ata_bmdma_irq_clear(struct ata_port *ap); -extern void ata_qc_complete(struct ata_queued_cmd *qc, u8 drv_stat); +extern void ata_qc_complete(struct ata_queued_cmd *qc, unsigned int err_mask); extern void ata_eng_timeout(struct ata_port *ap); extern void ata_scsi_simulate(u16 *id, struct scsi_cmnd *cmd, void (*done)(struct scsi_cmnd *)); @@ -716,4 +723,21 @@ static inline int ata_try_flush_cache(const struct ata_device *dev) ata_id_has_flush_ext(dev->id); } +static inline unsigned int ac_err_mask(u8 status) +{ + if (status & ATA_BUSY) + return AC_ERR_ATA_BUS; + if (status & (ATA_ERR | ATA_DF)) + return AC_ERR_DEV; + return 0; +} + +static inline unsigned int __ac_err_mask(u8 status) +{ + unsigned int mask = ac_err_mask(status); + if (mask == 0) + return AC_ERR_OTHER; + return mask; +} + #endif /* __LINUX_LIBATA_H__ */ -- cgit v1.2.3 From 0d078f6f96809c95c69b99d6605a502b0ac63d3d Mon Sep 17 00:00:00 2001 From: Brian Gerst Date: Sun, 30 Oct 2005 14:59:20 -0800 Subject: [PATCH] CONFIG_IA32 Add CONFIG_X86_32 for i386. This allows selecting options that only apply to 32-bit systems. (X86 && !X86_64) becomes X86_32 (X86 || X86_64) becomes X86 Signed-off-by: Brian Gerst Cc: Sam Ravnborg Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/i386/Kconfig | 6 +++++- drivers/char/Kconfig | 4 ++-- drivers/char/agp/Kconfig | 16 ++++++++-------- drivers/char/hangcheck-timer.c | 2 +- drivers/crypto/Kconfig | 2 +- drivers/firmware/Kconfig | 2 +- drivers/input/misc/Kconfig | 2 +- drivers/pci/hotplug/pciehp_pci.c | 2 +- drivers/pcmcia/rsrc_nonstatic.c | 2 +- drivers/video/Kconfig | 6 +++--- drivers/video/console/Kconfig | 2 +- fs/Kconfig | 2 +- fs/Kconfig.binfmt | 2 +- include/linux/dmi.h | 2 +- lib/Kconfig.debug | 2 +- 15 files changed, 29 insertions(+), 25 deletions(-) (limited to 'include/linux') diff --git a/arch/i386/Kconfig b/arch/i386/Kconfig index d2703cda61ea..2e3e8dbcb79f 100644 --- a/arch/i386/Kconfig +++ b/arch/i386/Kconfig @@ -5,7 +5,7 @@ mainmenu "Linux Kernel Configuration" -config X86 +config X86_32 bool default y help @@ -18,6 +18,10 @@ config SEMAPHORE_SLEEPERS bool default y +config X86 + bool + default y + config MMU bool default y diff --git a/drivers/char/Kconfig b/drivers/char/Kconfig index c29365d5b524..7902c3630681 100644 --- a/drivers/char/Kconfig +++ b/drivers/char/Kconfig @@ -661,7 +661,7 @@ config HW_RANDOM config NVRAM tristate "/dev/nvram support" - depends on ATARI || X86 || X86_64 || ARM || GENERIC_NVRAM + depends on ATARI || X86 || ARM || GENERIC_NVRAM ---help--- If you say Y here and create a character special file /dev/nvram with major number 10 and minor number 144 using mknod ("man mknod"), @@ -985,7 +985,7 @@ config MAX_RAW_DEVS config HANGCHECK_TIMER tristate "Hangcheck timer" - depends on X86_64 || X86 || IA64 || PPC64 || ARCH_S390 + depends on X86 || IA64 || PPC64 || ARCH_S390 help The hangcheck-timer module detects when the system has gone out to lunch past a certain margin. It can reboot the system diff --git a/drivers/char/agp/Kconfig b/drivers/char/agp/Kconfig index 7f8c1b53b754..486ed8a11b59 100644 --- a/drivers/char/agp/Kconfig +++ b/drivers/char/agp/Kconfig @@ -27,7 +27,7 @@ config AGP config AGP_ALI tristate "ALI chipset support" - depends on AGP && X86 && !X86_64 + depends on AGP && X86_32 ---help--- This option gives you AGP support for the GLX component of XFree86 4.x on the following ALi chipsets. The supported chipsets @@ -45,7 +45,7 @@ config AGP_ALI config AGP_ATI tristate "ATI chipset support" - depends on AGP && X86 && !X86_64 + depends on AGP && X86_32 ---help--- This option gives you AGP support for the GLX component of XFree86 4.x on the ATI RadeonIGP family of chipsets. @@ -55,7 +55,7 @@ config AGP_ATI config AGP_AMD tristate "AMD Irongate, 761, and 762 chipset support" - depends on AGP && X86 && !X86_64 + depends on AGP && X86_32 help This option gives you AGP support for the GLX component of XFree86 4.x on AMD Irongate, 761, and 762 chipsets. @@ -91,7 +91,7 @@ config AGP_INTEL config AGP_NVIDIA tristate "NVIDIA nForce/nForce2 chipset support" - depends on AGP && X86 && !X86_64 + depends on AGP && X86_32 help This option gives you AGP support for the GLX component of XFree86 4.x on the following NVIDIA chipsets. The supported chipsets @@ -99,7 +99,7 @@ config AGP_NVIDIA config AGP_SIS tristate "SiS chipset support" - depends on AGP && X86 && !X86_64 + depends on AGP && X86_32 help This option gives you AGP support for the GLX component of XFree86 4.x on Silicon Integrated Systems [SiS] chipsets. @@ -111,14 +111,14 @@ config AGP_SIS config AGP_SWORKS tristate "Serverworks LE/HE chipset support" - depends on AGP && X86 && !X86_64 + depends on AGP && X86_32 help Say Y here to support the Serverworks AGP card. See for product descriptions and images. config AGP_VIA tristate "VIA chipset support" - depends on AGP && X86 && !X86_64 + depends on AGP && X86_32 help This option gives you AGP support for the GLX component of XFree86 4.x on VIA MVP3/Apollo Pro chipsets. @@ -154,7 +154,7 @@ config AGP_UNINORTH config AGP_EFFICEON tristate "Transmeta Efficeon support" - depends on AGP && X86 && !X86_64 + depends on AGP && X86_32 help This option gives you AGP support for the Transmeta Efficeon series processors with integrated northbridges. diff --git a/drivers/char/hangcheck-timer.c b/drivers/char/hangcheck-timer.c index a54bc93353af..66e53dd450ff 100644 --- a/drivers/char/hangcheck-timer.c +++ b/drivers/char/hangcheck-timer.c @@ -117,7 +117,7 @@ __setup("hcheck_reboot", hangcheck_parse_reboot); __setup("hcheck_dump_tasks", hangcheck_parse_dump_tasks); #endif /* not MODULE */ -#if defined(CONFIG_X86) || defined(CONFIG_X86_64) +#if defined(CONFIG_X86) # define HAVE_MONOTONIC # define TIMER_FREQ 1000000000ULL #elif defined(CONFIG_ARCH_S390) diff --git a/drivers/crypto/Kconfig b/drivers/crypto/Kconfig index 094835cce321..4263935443cc 100644 --- a/drivers/crypto/Kconfig +++ b/drivers/crypto/Kconfig @@ -2,7 +2,7 @@ menu "Hardware crypto devices" config CRYPTO_DEV_PADLOCK tristate "Support for VIA PadLock ACE" - depends on CRYPTO && X86 && !X86_64 + depends on CRYPTO && X86_32 help Some VIA processors come with an integrated crypto engine (so called VIA PadLock ACE, Advanced Cryptography Engine) diff --git a/drivers/firmware/Kconfig b/drivers/firmware/Kconfig index 327b58e64875..9925de291e31 100644 --- a/drivers/firmware/Kconfig +++ b/drivers/firmware/Kconfig @@ -70,7 +70,7 @@ config DELL_RBU config DCDBAS tristate "Dell Systems Management Base Driver" - depends on X86 || X86_64 + depends on X86 default m help The Dell Systems Management Base Driver provides a sysfs interface diff --git a/drivers/input/misc/Kconfig b/drivers/input/misc/Kconfig index bb934e6d9636..b3eaac1b35b6 100644 --- a/drivers/input/misc/Kconfig +++ b/drivers/input/misc/Kconfig @@ -14,7 +14,7 @@ if INPUT_MISC config INPUT_PCSPKR tristate "PC Speaker support" - depends on ALPHA || X86 || X86_64 || MIPS || PPC_PREP || PPC_CHRP || PPC_PSERIES + depends on ALPHA || X86 || MIPS || PPC_PREP || PPC_CHRP || PPC_PSERIES help Say Y here if you want the standard PC Speaker to be used for bells and whistles. diff --git a/drivers/pci/hotplug/pciehp_pci.c b/drivers/pci/hotplug/pciehp_pci.c index 33b539b34f7e..ff17d8e07e94 100644 --- a/drivers/pci/hotplug/pciehp_pci.c +++ b/drivers/pci/hotplug/pciehp_pci.c @@ -113,7 +113,7 @@ int pciehp_unconfigure_device(struct pci_func* func) */ int pciehp_set_irq (u8 bus_num, u8 dev_num, u8 int_pin, u8 irq_num) { -#if defined(CONFIG_X86) && !defined(CONFIG_X86_IO_APIC) && !defined(CONFIG_X86_64) +#if defined(CONFIG_X86_32) && !defined(CONFIG_X86_IO_APIC) int rc; u16 temp_word; struct pci_dev fakedev; diff --git a/drivers/pcmcia/rsrc_nonstatic.c b/drivers/pcmcia/rsrc_nonstatic.c index fc87e7e2b6b8..00960a379b9c 100644 --- a/drivers/pcmcia/rsrc_nonstatic.c +++ b/drivers/pcmcia/rsrc_nonstatic.c @@ -779,7 +779,7 @@ static int nonstatic_autoadd_resources(struct pcmcia_socket *s) if (!s->cb_dev || !s->cb_dev->bus) return -ENODEV; -#if defined(CONFIG_X86) || defined(CONFIG_X86_64) +#if defined(CONFIG_X86) /* If this is the root bus, the risk of hitting * some strange system devices which aren't protected * by either ACPI resource tables or properly requested diff --git a/drivers/video/Kconfig b/drivers/video/Kconfig index 7e297947a2b2..7192b770bfb6 100644 --- a/drivers/video/Kconfig +++ b/drivers/video/Kconfig @@ -494,7 +494,7 @@ config FB_TGA config FB_VESA bool "VESA VGA graphics support" - depends on (FB = y) && (X86 || X86_64) + depends on (FB = y) && X86 select FB_CFB_FILLRECT select FB_CFB_COPYAREA select FB_CFB_IMAGEBLIT @@ -712,7 +712,7 @@ config FB_RIVA_DEBUG config FB_I810 tristate "Intel 810/815 support (EXPERIMENTAL)" - depends on FB && EXPERIMENTAL && PCI && X86 && !X86_64 + depends on FB && EXPERIMENTAL && PCI && X86_32 select AGP select AGP_INTEL select FB_MODE_HELPERS @@ -761,7 +761,7 @@ config FB_I810_I2C config FB_INTEL tristate "Intel 830M/845G/852GM/855GM/865G support (EXPERIMENTAL)" - depends on FB && EXPERIMENTAL && PCI && X86 && !X86_64 + depends on FB && EXPERIMENTAL && PCI && X86_32 select AGP select AGP_INTEL select FB_MODE_HELPERS diff --git a/drivers/video/console/Kconfig b/drivers/video/console/Kconfig index 7e731691e2a9..6a9ae2b3d1ab 100644 --- a/drivers/video/console/Kconfig +++ b/drivers/video/console/Kconfig @@ -28,7 +28,7 @@ config VGA_CONSOLE config VIDEO_SELECT bool "Video mode selection support" - depends on (X86 || X86_64) && VGA_CONSOLE + depends on X86 && VGA_CONSOLE ---help--- This enables support for text mode selection on kernel startup. If you want to take advantage of some high-resolution text mode your diff --git a/fs/Kconfig b/fs/Kconfig index 48f5422cb19a..01a295232f75 100644 --- a/fs/Kconfig +++ b/fs/Kconfig @@ -810,7 +810,7 @@ config TMPFS config HUGETLBFS bool "HugeTLB file system support" - depends X86 || IA64 || PPC64 || SPARC64 || SUPERH || X86_64 || BROKEN + depends X86 || IA64 || PPC64 || SPARC64 || SUPERH || BROKEN config HUGETLB_PAGE def_bool HUGETLBFS diff --git a/fs/Kconfig.binfmt b/fs/Kconfig.binfmt index 434c19d076ac..175b2e8177c1 100644 --- a/fs/Kconfig.binfmt +++ b/fs/Kconfig.binfmt @@ -57,7 +57,7 @@ config BINFMT_SHARED_FLAT config BINFMT_AOUT tristate "Kernel support for a.out and ECOFF binaries" - depends on (X86 && !X86_64) || ALPHA || ARM || M68K || SPARC32 + depends on X86_32 || ALPHA || ARM || M68K || SPARC32 ---help--- A.out (Assembler.OUTput) is a set of formats for libraries and executables used in the earliest versions of UNIX. Linux used diff --git a/include/linux/dmi.h b/include/linux/dmi.h index a415f1d93e9a..05f4132622fc 100644 --- a/include/linux/dmi.h +++ b/include/linux/dmi.h @@ -60,7 +60,7 @@ struct dmi_device { void *device_data; /* Type specific data */ }; -#if defined(CONFIG_X86) && !defined(CONFIG_X86_64) +#if defined(CONFIG_X86_32) extern int dmi_check_system(struct dmi_system_id *list); extern char * dmi_get_system_info(int field); diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug index 016e89a44ac8..807535e7a1e7 100644 --- a/lib/Kconfig.debug +++ b/lib/Kconfig.debug @@ -128,7 +128,7 @@ config DEBUG_HIGHMEM config DEBUG_BUGVERBOSE bool "Verbose BUG() reporting (adds 70K)" if DEBUG_KERNEL && EMBEDDED depends on BUG - depends on ARM || ARM26 || M32R || M68K || SPARC32 || SPARC64 || (X86 && !X86_64) || FRV + depends on ARM || ARM26 || M32R || M68K || SPARC32 || SPARC64 || X86_32 || FRV default !EMBEDDED help Say Y here to make BUG() panics output the file name and line number -- cgit v1.2.3 From d381d8a9a08cac9824096213069159be17fd2e2f Mon Sep 17 00:00:00 2001 From: James Morris Date: Sun, 30 Oct 2005 14:59:22 -0800 Subject: [PATCH] SELinux: canonicalize getxattr() This patch allows SELinux to canonicalize the value returned from getxattr() via the security_inode_getsecurity() hook, which is called after the fs level getxattr() function. The purpose of this is to allow the in-core security context for an inode to override the on-disk value. This could happen in cases such as upgrading a system to a different labeling form (e.g. standard SELinux to MLS) without needing to do a full relabel of the filesystem. In such cases, we want getxattr() to return the canonical security context that the kernel is using rather than what is stored on disk. The implementation hooks into the inode_getsecurity(), adding another parameter to indicate the result of the preceding fs-level getxattr() call, so that SELinux knows whether to compare a value obtained from disk with the kernel value. We also now allow getxattr() to work for mountpoint labeled filesystems (i.e. mount with option context=foo_t), as we are able to return the kernel value to the user. Signed-off-by: James Morris Signed-off-by: Stephen Smalley Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/xattr.c | 14 +++++++++----- include/linux/security.h | 11 +++++++---- security/dummy.c | 2 +- security/selinux/hooks.c | 46 ++++++++++++++++++++++++++++++++-------------- 4 files changed, 49 insertions(+), 24 deletions(-) (limited to 'include/linux') diff --git a/fs/xattr.c b/fs/xattr.c index 3f9c64bea151..f6e00c0e114f 100644 --- a/fs/xattr.c +++ b/fs/xattr.c @@ -143,7 +143,7 @@ getxattr(struct dentry *d, char __user *name, void __user *value, size_t size) if (size) { if (size > XATTR_SIZE_MAX) size = XATTR_SIZE_MAX; - kvalue = kmalloc(size, GFP_KERNEL); + kvalue = kzalloc(size, GFP_KERNEL); if (!kvalue) return -ENOMEM; } @@ -154,11 +154,15 @@ getxattr(struct dentry *d, char __user *name, void __user *value, size_t size) error = -EOPNOTSUPP; if (d->d_inode->i_op && d->d_inode->i_op->getxattr) error = d->d_inode->i_op->getxattr(d, kname, kvalue, size); - else if (!strncmp(kname, XATTR_SECURITY_PREFIX, - sizeof XATTR_SECURITY_PREFIX - 1)) { + + if (!strncmp(kname, XATTR_SECURITY_PREFIX, + sizeof XATTR_SECURITY_PREFIX - 1)) { const char *suffix = kname + sizeof XATTR_SECURITY_PREFIX - 1; - error = security_inode_getsecurity(d->d_inode, suffix, kvalue, - size); + int rv = security_inode_getsecurity(d->d_inode, suffix, kvalue, + size, error); + /* Security module active: overwrite error value */ + if (rv != -EOPNOTSUPP) + error = rv; } if (error > 0) { if (size && copy_to_user(value, kvalue, error)) diff --git a/include/linux/security.h b/include/linux/security.h index dac956ed98f0..607ee209ea3b 100644 --- a/include/linux/security.h +++ b/include/linux/security.h @@ -385,6 +385,9 @@ struct swap_info_struct; * NULL to request the size of the buffer required. @size indicates * the size of @buffer in bytes. Note that @name is the remainder * of the attribute name after the security. prefix has been removed. + * @err is the return value from the preceding fs getxattr call, + * and can be used by the security module to determine whether it + * should try and canonicalize the attribute value. * Return number of bytes used/required on success. * @inode_setsecurity: * Set the security label associated with @name for @inode from the @@ -1091,7 +1094,7 @@ struct security_operations { int (*inode_getxattr) (struct dentry *dentry, char *name); int (*inode_listxattr) (struct dentry *dentry); int (*inode_removexattr) (struct dentry *dentry, char *name); - int (*inode_getsecurity)(struct inode *inode, const char *name, void *buffer, size_t size); + int (*inode_getsecurity)(struct inode *inode, const char *name, void *buffer, size_t size, int err); int (*inode_setsecurity)(struct inode *inode, const char *name, const void *value, size_t size, int flags); int (*inode_listsecurity)(struct inode *inode, char *buffer, size_t buffer_size); @@ -1580,11 +1583,11 @@ static inline int security_inode_removexattr (struct dentry *dentry, char *name) return security_ops->inode_removexattr (dentry, name); } -static inline int security_inode_getsecurity(struct inode *inode, const char *name, void *buffer, size_t size) +static inline int security_inode_getsecurity(struct inode *inode, const char *name, void *buffer, size_t size, int err) { if (unlikely (IS_PRIVATE (inode))) return 0; - return security_ops->inode_getsecurity(inode, name, buffer, size); + return security_ops->inode_getsecurity(inode, name, buffer, size, err); } static inline int security_inode_setsecurity(struct inode *inode, const char *name, const void *value, size_t size, int flags) @@ -2222,7 +2225,7 @@ static inline int security_inode_removexattr (struct dentry *dentry, char *name) return cap_inode_removexattr(dentry, name); } -static inline int security_inode_getsecurity(struct inode *inode, const char *name, void *buffer, size_t size) +static inline int security_inode_getsecurity(struct inode *inode, const char *name, void *buffer, size_t size, int err) { return -EOPNOTSUPP; } diff --git a/security/dummy.c b/security/dummy.c index 3d34f3de7e82..2a0337a52d32 100644 --- a/security/dummy.c +++ b/security/dummy.c @@ -377,7 +377,7 @@ static int dummy_inode_removexattr (struct dentry *dentry, char *name) return 0; } -static int dummy_inode_getsecurity(struct inode *inode, const char *name, void *buffer, size_t size) +static int dummy_inode_getsecurity(struct inode *inode, const char *name, void *buffer, size_t size, int err) { return -EOPNOTSUPP; } diff --git a/security/selinux/hooks.c b/security/selinux/hooks.c index 8cd33b2cd865..d9ec85292e1c 100644 --- a/security/selinux/hooks.c +++ b/security/selinux/hooks.c @@ -2208,9 +2208,6 @@ static int selinux_inode_getxattr (struct dentry *dentry, char *name) struct inode *inode = dentry->d_inode; struct superblock_security_struct *sbsec = inode->i_sb->s_security; - if (sbsec->behavior == SECURITY_FS_USE_MNTPOINT) - return -EOPNOTSUPP; - return dentry_has_perm(current, NULL, dentry, FILE__GETATTR); } @@ -2241,33 +2238,54 @@ static int selinux_inode_removexattr (struct dentry *dentry, char *name) return -EACCES; } -static int selinux_inode_getsecurity(struct inode *inode, const char *name, void *buffer, size_t size) +/* + * Copy the in-core inode security context value to the user. If the + * getxattr() prior to this succeeded, check to see if we need to + * canonicalize the value to be finally returned to the user. + * + * Permission check is handled by selinux_inode_getxattr hook. + */ +static int selinux_inode_getsecurity(struct inode *inode, const char *name, void *buffer, size_t size, int err) { struct inode_security_struct *isec = inode->i_security; char *context; unsigned len; int rc; - /* Permission check handled by selinux_inode_getxattr hook.*/ - - if (strcmp(name, XATTR_SELINUX_SUFFIX)) - return -EOPNOTSUPP; + if (strcmp(name, XATTR_SELINUX_SUFFIX)) { + rc = -EOPNOTSUPP; + goto out; + } rc = security_sid_to_context(isec->sid, &context, &len); if (rc) - return rc; + goto out; + /* Probe for required buffer size */ if (!buffer || !size) { - kfree(context); - return len; + rc = len; + goto out_free; } + if (size < len) { - kfree(context); - return -ERANGE; + rc = -ERANGE; + goto out_free; + } + + if (err > 0) { + if ((len == err) && !(memcmp(context, buffer, len))) { + /* Don't need to canonicalize value */ + rc = err; + goto out_free; + } + memset(buffer, 0, size); } memcpy(buffer, context, len); + rc = len; +out_free: kfree(context); - return len; +out: + return rc; } static int selinux_inode_setsecurity(struct inode *inode, const char *name, -- cgit v1.2.3 From bda98685b855f71f7e2fc5378aa3cdfb24a9db65 Mon Sep 17 00:00:00 2001 From: Ingo Molnar Date: Sun, 30 Oct 2005 14:59:44 -0800 Subject: [PATCH] x86: inline spin_unlock if !CONFIG_DEBUG_SPINLOCK and !CONFIG_PREEMPT Signed-off-by: Ingo Molnar Signed-off-by: Miklos Szeredi Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/spinlock.h | 31 +++++++++++++++++++++++++------ 1 file changed, 25 insertions(+), 6 deletions(-) (limited to 'include/linux') diff --git a/include/linux/spinlock.h b/include/linux/spinlock.h index cdc99a27840d..0e9682c9def5 100644 --- a/include/linux/spinlock.h +++ b/include/linux/spinlock.h @@ -171,23 +171,42 @@ extern int __lockfunc generic__raw_read_trylock(raw_rwlock_t *lock); #define write_lock_irq(lock) _write_lock_irq(lock) #define write_lock_bh(lock) _write_lock_bh(lock) -#define spin_unlock(lock) _spin_unlock(lock) -#define write_unlock(lock) _write_unlock(lock) -#define read_unlock(lock) _read_unlock(lock) +/* + * We inline the unlock functions in the nondebug case: + */ +#if defined(CONFIG_DEBUG_SPINLOCK) || defined(CONFIG_PREEMPT) || !defined(CONFIG_SMP) +# define spin_unlock(lock) _spin_unlock(lock) +# define read_unlock(lock) _read_unlock(lock) +# define write_unlock(lock) _write_unlock(lock) +#else +# define spin_unlock(lock) __raw_spin_unlock(&(lock)->raw_lock) +# define read_unlock(lock) __raw_read_unlock(&(lock)->raw_lock) +# define write_unlock(lock) __raw_write_unlock(&(lock)->raw_lock) +#endif + +#if defined(CONFIG_DEBUG_SPINLOCK) || defined(CONFIG_PREEMPT) || !defined(CONFIG_SMP) +# define spin_unlock_irq(lock) _spin_unlock_irq(lock) +# define read_unlock_irq(lock) _read_unlock_irq(lock) +# define write_unlock_irq(lock) _write_unlock_irq(lock) +#else +# define spin_unlock_irq(lock) \ + do { __raw_spin_unlock(&(lock)->raw_lock); local_irq_enable(); } while (0) +# define read_unlock_irq(lock) \ + do { __raw_read_unlock(&(lock)->raw_lock); local_irq_enable(); } while (0) +# define write_unlock_irq(lock) \ + do { __raw_write_unlock(&(lock)->raw_lock); local_irq_enable(); } while (0) +#endif #define spin_unlock_irqrestore(lock, flags) \ _spin_unlock_irqrestore(lock, flags) -#define spin_unlock_irq(lock) _spin_unlock_irq(lock) #define spin_unlock_bh(lock) _spin_unlock_bh(lock) #define read_unlock_irqrestore(lock, flags) \ _read_unlock_irqrestore(lock, flags) -#define read_unlock_irq(lock) _read_unlock_irq(lock) #define read_unlock_bh(lock) _read_unlock_bh(lock) #define write_unlock_irqrestore(lock, flags) \ _write_unlock_irqrestore(lock, flags) -#define write_unlock_irq(lock) _write_unlock_irq(lock) #define write_unlock_bh(lock) _write_unlock_bh(lock) #define spin_trylock_bh(lock) __cond_lock(_spin_trylock_bh(lock)) -- cgit v1.2.3 From ad74557a49d1dea428fb0ad60e75a5aa37610e1d Mon Sep 17 00:00:00 2001 From: Ashok Raj Date: Sun, 30 Oct 2005 14:59:49 -0800 Subject: [PATCH] introduce get_cpu_sysdev() to retrieve a sysfs entry for a cpu. Some modules creating sysfs entries under /sys/devices/system/cpu/cpuX/ need to know the parent sysfs entry to make devices under them. This will just return the sysfs entry for a given cpu. sysfs entries showing under each cpu sysfs can be easily created if such entries can be created by registering a sysfs driver for cpuclass. The issue is when the entry is created the CPU may not be online, hence we would need to defer the creation until the online notification comes. Current users: cache entries for Intel CPU's and cpufreq subsystem. Signed-off-by: Ashok Raj Signed-off-by: Venkatesh Pallipadi Cc: Dave Jones Cc: Zwane Mwaikambo Cc: Greg KH Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/base/cpu.c | 16 ++++++++++++++-- include/linux/cpu.h | 1 + 2 files changed, 15 insertions(+), 2 deletions(-) (limited to 'include/linux') diff --git a/drivers/base/cpu.c b/drivers/base/cpu.c index 081c927b1ed8..a95844790f7b 100644 --- a/drivers/base/cpu.c +++ b/drivers/base/cpu.c @@ -16,6 +16,8 @@ struct sysdev_class cpu_sysdev_class = { }; EXPORT_SYMBOL(cpu_sysdev_class); +static struct sys_device *cpu_sys_devices[NR_CPUS]; + #ifdef CONFIG_HOTPLUG_CPU int __attribute__((weak)) smp_prepare_cpu (int cpu) { @@ -64,6 +66,7 @@ static void __devinit register_cpu_control(struct cpu *cpu) } void unregister_cpu(struct cpu *cpu, struct node *root) { + int logical_cpu = cpu->sysdev.id; if (root) sysfs_remove_link(&root->sysdev.kobj, @@ -71,7 +74,7 @@ void unregister_cpu(struct cpu *cpu, struct node *root) sysdev_remove_file(&cpu->sysdev, &attr_online); sysdev_unregister(&cpu->sysdev); - + cpu_sys_devices[logical_cpu] = NULL; return; } #else /* ... !CONFIG_HOTPLUG_CPU */ @@ -103,10 +106,19 @@ int __devinit register_cpu(struct cpu *cpu, int num, struct node *root) kobject_name(&cpu->sysdev.kobj)); if (!error && !cpu->no_control) register_cpu_control(cpu); + if (!error) + cpu_sys_devices[num] = &cpu->sysdev; return error; } - +struct sys_device *get_cpu_sysdev(int cpu) +{ + if (cpu < NR_CPUS) + return cpu_sys_devices[cpu]; + else + return NULL; +} +EXPORT_SYMBOL_GPL(get_cpu_sysdev); int __init cpu_dev_init(void) { diff --git a/include/linux/cpu.h b/include/linux/cpu.h index 86980c68234a..1f7b2c097503 100644 --- a/include/linux/cpu.h +++ b/include/linux/cpu.h @@ -32,6 +32,7 @@ struct cpu { }; extern int register_cpu(struct cpu *, int, struct node *); +extern struct sys_device *get_cpu_sysdev(int cpu); #ifdef CONFIG_HOTPLUG_CPU extern void unregister_cpu(struct cpu *, struct node *); #endif -- cgit v1.2.3 From 25761b6eb7b33823bcfff6bfe2a015badcd76fb8 Mon Sep 17 00:00:00 2001 From: "Rafael J. Wysocki" Date: Sun, 30 Oct 2005 14:59:56 -0800 Subject: [PATCH] swsusp: move snapshot functionality to separate file The following patch moves the functionality of swsusp related to creating and handling the snapshot of memory to a separate file, snapshot.c This should enable us to untangle the code in the future and eventually to implement some parts of swsusp.c in the user space. The patch does not change the code. Signed-off-by: Rafael J. Wysocki Signed-off-by: Pavel Machek Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/suspend.h | 6 + kernel/power/Makefile | 2 +- kernel/power/power.h | 17 ++ kernel/power/snapshot.c | 464 ++++++++++++++++++++++++++++++++++++++++++++++++ kernel/power/swsusp.c | 440 +-------------------------------------------- 5 files changed, 492 insertions(+), 437 deletions(-) create mode 100644 kernel/power/snapshot.c (limited to 'include/linux') diff --git a/include/linux/suspend.h b/include/linux/suspend.h index ba448c760168..380915e9563d 100644 --- a/include/linux/suspend.h +++ b/include/linux/suspend.h @@ -74,4 +74,10 @@ void __restore_processor_state(struct saved_context *ctxt); extern unsigned long get_usable_page(gfp_t gfp_mask); extern void free_eaten_memory(void); +/* + * XXX: We try to keep some more pages free so that I/O operations succeed + * without paging. Might this be more? + */ +#define PAGES_FOR_IO 512 + #endif /* _LINUX_SWSUSP_H */ diff --git a/kernel/power/Makefile b/kernel/power/Makefile index 2f438d0eaa13..c71eb4579c07 100644 --- a/kernel/power/Makefile +++ b/kernel/power/Makefile @@ -4,7 +4,7 @@ EXTRA_CFLAGS += -DDEBUG endif obj-y := main.o process.o console.o pm.o -obj-$(CONFIG_SOFTWARE_SUSPEND) += swsusp.o disk.o +obj-$(CONFIG_SOFTWARE_SUSPEND) += swsusp.o disk.o snapshot.o obj-$(CONFIG_SUSPEND_SMP) += smp.o diff --git a/kernel/power/power.h b/kernel/power/power.h index 6748de23e83c..e54dd8435de7 100644 --- a/kernel/power/power.h +++ b/kernel/power/power.h @@ -53,3 +53,20 @@ extern void thaw_processes(void); extern int pm_prepare_console(void); extern void pm_restore_console(void); + + +/* References to section boundaries */ +extern const void __nosave_begin, __nosave_end; + +extern unsigned int nr_copy_pages; +extern suspend_pagedir_t *pagedir_nosave; +extern suspend_pagedir_t *pagedir_save; + +extern asmlinkage int swsusp_arch_suspend(void); +extern asmlinkage int swsusp_arch_resume(void); + +extern int restore_highmem(void); +extern void free_pagedir(struct pbe *pblist); +extern struct pbe * alloc_pagedir(unsigned nr_pages); +extern void create_pbe_list(struct pbe *pblist, unsigned nr_pages); +extern int enough_swap(void); diff --git a/kernel/power/snapshot.c b/kernel/power/snapshot.c new file mode 100644 index 000000000000..0f0a7f306b0d --- /dev/null +++ b/kernel/power/snapshot.c @@ -0,0 +1,464 @@ +/* + * linux/kernel/power/swsusp.c + * + * This file is to realize architecture-independent + * machine suspend feature using pretty near only high-level routines + * + * Copyright (C) 1998-2005 Pavel Machek + * + * This file is released under the GPLv2, and is based on swsusp.c. + * + */ + + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include +#include +#include +#include +#include + +#include +#include +#include + +#include "power.h" + + + + +#ifdef CONFIG_HIGHMEM +struct highmem_page { + char *data; + struct page *page; + struct highmem_page *next; +}; + +static struct highmem_page *highmem_copy; + +static int save_highmem_zone(struct zone *zone) +{ + unsigned long zone_pfn; + mark_free_pages(zone); + for (zone_pfn = 0; zone_pfn < zone->spanned_pages; ++zone_pfn) { + struct page *page; + struct highmem_page *save; + void *kaddr; + unsigned long pfn = zone_pfn + zone->zone_start_pfn; + + if (!(pfn%1000)) + printk("."); + if (!pfn_valid(pfn)) + continue; + page = pfn_to_page(pfn); + /* + * This condition results from rvmalloc() sans vmalloc_32() + * and architectural memory reservations. This should be + * corrected eventually when the cases giving rise to this + * are better understood. + */ + if (PageReserved(page)) { + printk("highmem reserved page?!\n"); + continue; + } + BUG_ON(PageNosave(page)); + if (PageNosaveFree(page)) + continue; + save = kmalloc(sizeof(struct highmem_page), GFP_ATOMIC); + if (!save) + return -ENOMEM; + save->next = highmem_copy; + save->page = page; + save->data = (void *) get_zeroed_page(GFP_ATOMIC); + if (!save->data) { + kfree(save); + return -ENOMEM; + } + kaddr = kmap_atomic(page, KM_USER0); + memcpy(save->data, kaddr, PAGE_SIZE); + kunmap_atomic(kaddr, KM_USER0); + highmem_copy = save; + } + return 0; +} +#endif /* CONFIG_HIGHMEM */ + + +static int save_highmem(void) +{ +#ifdef CONFIG_HIGHMEM + struct zone *zone; + int res = 0; + + pr_debug("swsusp: Saving Highmem\n"); + for_each_zone (zone) { + if (is_highmem(zone)) + res = save_highmem_zone(zone); + if (res) + return res; + } +#endif + return 0; +} + +int restore_highmem(void) +{ +#ifdef CONFIG_HIGHMEM + printk("swsusp: Restoring Highmem\n"); + while (highmem_copy) { + struct highmem_page *save = highmem_copy; + void *kaddr; + highmem_copy = save->next; + + kaddr = kmap_atomic(save->page, KM_USER0); + memcpy(kaddr, save->data, PAGE_SIZE); + kunmap_atomic(kaddr, KM_USER0); + free_page((long) save->data); + kfree(save); + } +#endif + return 0; +} + + +static int pfn_is_nosave(unsigned long pfn) +{ + unsigned long nosave_begin_pfn = __pa(&__nosave_begin) >> PAGE_SHIFT; + unsigned long nosave_end_pfn = PAGE_ALIGN(__pa(&__nosave_end)) >> PAGE_SHIFT; + return (pfn >= nosave_begin_pfn) && (pfn < nosave_end_pfn); +} + +/** + * saveable - Determine whether a page should be cloned or not. + * @pfn: The page + * + * We save a page if it's Reserved, and not in the range of pages + * statically defined as 'unsaveable', or if it isn't reserved, and + * isn't part of a free chunk of pages. + */ + +static int saveable(struct zone * zone, unsigned long * zone_pfn) +{ + unsigned long pfn = *zone_pfn + zone->zone_start_pfn; + struct page * page; + + if (!pfn_valid(pfn)) + return 0; + + page = pfn_to_page(pfn); + BUG_ON(PageReserved(page) && PageNosave(page)); + if (PageNosave(page)) + return 0; + if (PageReserved(page) && pfn_is_nosave(pfn)) { + pr_debug("[nosave pfn 0x%lx]", pfn); + return 0; + } + if (PageNosaveFree(page)) + return 0; + + return 1; +} + +static void count_data_pages(void) +{ + struct zone *zone; + unsigned long zone_pfn; + + nr_copy_pages = 0; + + for_each_zone (zone) { + if (is_highmem(zone)) + continue; + mark_free_pages(zone); + for (zone_pfn = 0; zone_pfn < zone->spanned_pages; ++zone_pfn) + nr_copy_pages += saveable(zone, &zone_pfn); + } +} + +static void copy_data_pages(void) +{ + struct zone *zone; + unsigned long zone_pfn; + struct pbe *pbe = pagedir_nosave, *p; + + pr_debug("copy_data_pages(): pages to copy: %d\n", nr_copy_pages); + for_each_zone (zone) { + if (is_highmem(zone)) + continue; + mark_free_pages(zone); + /* This is necessary for swsusp_free() */ + for_each_pb_page (p, pagedir_nosave) + SetPageNosaveFree(virt_to_page(p)); + for_each_pbe(p, pagedir_nosave) + SetPageNosaveFree(virt_to_page(p->address)); + for (zone_pfn = 0; zone_pfn < zone->spanned_pages; ++zone_pfn) { + if (saveable(zone, &zone_pfn)) { + struct page * page; + page = pfn_to_page(zone_pfn + zone->zone_start_pfn); + BUG_ON(!pbe); + pbe->orig_address = (unsigned long)page_address(page); + /* copy_page is not usable for copying task structs. */ + memcpy((void *)pbe->address, (void *)pbe->orig_address, PAGE_SIZE); + pbe = pbe->next; + } + } + } + BUG_ON(pbe); +} + + +/** + * free_pagedir - free pages allocated with alloc_pagedir() + */ + +void free_pagedir(struct pbe *pblist) +{ + struct pbe *pbe; + + while (pblist) { + pbe = (pblist + PB_PAGE_SKIP)->next; + ClearPageNosave(virt_to_page(pblist)); + ClearPageNosaveFree(virt_to_page(pblist)); + free_page((unsigned long)pblist); + pblist = pbe; + } +} + +/** + * fill_pb_page - Create a list of PBEs on a given memory page + */ + +static inline void fill_pb_page(struct pbe *pbpage) +{ + struct pbe *p; + + p = pbpage; + pbpage += PB_PAGE_SKIP; + do + p->next = p + 1; + while (++p < pbpage); +} + +/** + * create_pbe_list - Create a list of PBEs on top of a given chain + * of memory pages allocated with alloc_pagedir() + */ + +void create_pbe_list(struct pbe *pblist, unsigned nr_pages) +{ + struct pbe *pbpage, *p; + unsigned num = PBES_PER_PAGE; + + for_each_pb_page (pbpage, pblist) { + if (num >= nr_pages) + break; + + fill_pb_page(pbpage); + num += PBES_PER_PAGE; + } + if (pbpage) { + for (num -= PBES_PER_PAGE - 1, p = pbpage; num < nr_pages; p++, num++) + p->next = p + 1; + p->next = NULL; + } + pr_debug("create_pbe_list(): initialized %d PBEs\n", num); +} + +static void *alloc_image_page(void) +{ + void *res = (void *)get_zeroed_page(GFP_ATOMIC | __GFP_COLD); + if (res) { + SetPageNosave(virt_to_page(res)); + SetPageNosaveFree(virt_to_page(res)); + } + return res; +} + +/** + * alloc_pagedir - Allocate the page directory. + * + * First, determine exactly how many pages we need and + * allocate them. + * + * We arrange the pages in a chain: each page is an array of PBES_PER_PAGE + * struct pbe elements (pbes) and the last element in the page points + * to the next page. + * + * On each page we set up a list of struct_pbe elements. + */ + +struct pbe * alloc_pagedir(unsigned nr_pages) +{ + unsigned num; + struct pbe *pblist, *pbe; + + if (!nr_pages) + return NULL; + + pr_debug("alloc_pagedir(): nr_pages = %d\n", nr_pages); + pblist = (struct pbe *)alloc_image_page(); + /* FIXME: rewrite this ugly loop */ + for (pbe = pblist, num = PBES_PER_PAGE; pbe && num < nr_pages; + pbe = pbe->next, num += PBES_PER_PAGE) { + pbe += PB_PAGE_SKIP; + pbe->next = (struct pbe *)alloc_image_page(); + } + if (!pbe) { /* get_zeroed_page() failed */ + free_pagedir(pblist); + pblist = NULL; + } + return pblist; +} + +/** + * Free pages we allocated for suspend. Suspend pages are alocated + * before atomic copy, so we need to free them after resume. + */ + +void swsusp_free(void) +{ + struct zone *zone; + unsigned long zone_pfn; + + for_each_zone(zone) { + for (zone_pfn = 0; zone_pfn < zone->spanned_pages; ++zone_pfn) + if (pfn_valid(zone_pfn + zone->zone_start_pfn)) { + struct page * page; + page = pfn_to_page(zone_pfn + zone->zone_start_pfn); + if (PageNosave(page) && PageNosaveFree(page)) { + ClearPageNosave(page); + ClearPageNosaveFree(page); + free_page((long) page_address(page)); + } + } + } +} + + +/** + * enough_free_mem - Make sure we enough free memory to snapshot. + * + * Returns TRUE or FALSE after checking the number of available + * free pages. + */ + +static int enough_free_mem(void) +{ + pr_debug("swsusp: available memory: %u pages\n", nr_free_pages()); + return nr_free_pages() > (nr_copy_pages + PAGES_FOR_IO + + nr_copy_pages/PBES_PER_PAGE + !!(nr_copy_pages%PBES_PER_PAGE)); +} + + +static int swsusp_alloc(void) +{ + struct pbe * p; + + pagedir_nosave = NULL; + + if (MAX_PBES < nr_copy_pages / PBES_PER_PAGE + + !!(nr_copy_pages % PBES_PER_PAGE)) + return -ENOSPC; + + if (!(pagedir_save = alloc_pagedir(nr_copy_pages))) { + printk(KERN_ERR "suspend: Allocating pagedir failed.\n"); + return -ENOMEM; + } + create_pbe_list(pagedir_save, nr_copy_pages); + pagedir_nosave = pagedir_save; + + for_each_pbe (p, pagedir_save) { + p->address = (unsigned long)alloc_image_page(); + if (!p->address) { + printk(KERN_ERR "suspend: Allocating image pages failed.\n"); + swsusp_free(); + return -ENOMEM; + } + } + + return 0; +} + +static int suspend_prepare_image(void) +{ + int error; + + pr_debug("swsusp: critical section: \n"); + if (save_highmem()) { + printk(KERN_CRIT "swsusp: Not enough free pages for highmem\n"); + restore_highmem(); + return -ENOMEM; + } + + drain_local_pages(); + count_data_pages(); + printk("swsusp: Need to copy %u pages\n", nr_copy_pages); + + pr_debug("swsusp: pages needed: %u + %lu + %u, free: %u\n", + nr_copy_pages, + nr_copy_pages/PBES_PER_PAGE + !!(nr_copy_pages%PBES_PER_PAGE), + PAGES_FOR_IO, nr_free_pages()); + + if (!enough_free_mem()) { + printk(KERN_ERR "swsusp: Not enough free memory\n"); + return -ENOMEM; + } + + if (!enough_swap()) { + printk(KERN_ERR "swsusp: Not enough free swap\n"); + return -ENOSPC; + } + + error = swsusp_alloc(); + if (error) + return error; + + /* During allocating of suspend pagedir, new cold pages may appear. + * Kill them. + */ + drain_local_pages(); + copy_data_pages(); + + /* + * End of critical section. From now on, we can write to memory, + * but we should not touch disk. This specially means we must _not_ + * touch swap space! Except we must write out our image of course. + */ + + printk("swsusp: critical section/: done (%d pages copied)\n", nr_copy_pages ); + return 0; +} + + +asmlinkage int swsusp_save(void) +{ + return suspend_prepare_image(); +} diff --git a/kernel/power/swsusp.c b/kernel/power/swsusp.c index ae46506e2137..fc50b5d2dd26 100644 --- a/kernel/power/swsusp.c +++ b/kernel/power/swsusp.c @@ -5,7 +5,7 @@ * machine suspend feature using pretty near only high-level routines * * Copyright (C) 1998-2001 Gabor Kuti - * Copyright (C) 1998,2001-2004 Pavel Machek + * Copyright (C) 1998,2001-2005 Pavel Machek * * This file is released under the GPLv2. * @@ -84,16 +84,10 @@ #define MAXKEY 32 #define MAXIV 32 -/* References to section boundaries */ -extern const void __nosave_begin, __nosave_end; - -/* Variables to be preserved over suspend */ -static int nr_copy_pages_check; - extern char resume_file[]; /* Local variables that should not be affected by save */ -static unsigned int nr_copy_pages __nosavedata = 0; +unsigned int nr_copy_pages __nosavedata = 0; /* Suspend pagedir is allocated before final copy, therefore it must be freed after resume @@ -109,7 +103,7 @@ static unsigned int nr_copy_pages __nosavedata = 0; MMU hardware. */ suspend_pagedir_t *pagedir_nosave __nosavedata = NULL; -static suspend_pagedir_t *pagedir_save; +suspend_pagedir_t *pagedir_save; #define SWSUSP_SIG "S1SUSPEND" @@ -123,12 +117,6 @@ static struct swsusp_header { static struct swsusp_info swsusp_info; -/* - * XXX: We try to keep some more pages free so that I/O operations succeed - * without paging. Might this be more? - */ -#define PAGES_FOR_IO 512 - /* * Saving part... */ @@ -552,335 +540,6 @@ static int write_suspend_image(void) goto Done; } - -#ifdef CONFIG_HIGHMEM -struct highmem_page { - char *data; - struct page *page; - struct highmem_page *next; -}; - -static struct highmem_page *highmem_copy; - -static int save_highmem_zone(struct zone *zone) -{ - unsigned long zone_pfn; - mark_free_pages(zone); - for (zone_pfn = 0; zone_pfn < zone->spanned_pages; ++zone_pfn) { - struct page *page; - struct highmem_page *save; - void *kaddr; - unsigned long pfn = zone_pfn + zone->zone_start_pfn; - - if (!(pfn%1000)) - printk("."); - if (!pfn_valid(pfn)) - continue; - page = pfn_to_page(pfn); - /* - * PageReserved results from rvmalloc() sans vmalloc_32() - * and architectural memory reservations. - * - * rvmalloc should not cause this, because all implementations - * appear to always be using vmalloc_32 on architectures with - * highmem. This is a good thing, because we would like to save - * rvmalloc pages. - * - * It appears to be triggered by pages which do not point to - * valid memory (see arch/i386/mm/init.c:one_highpage_init(), - * which sets PageReserved if the page does not point to valid - * RAM. - * - * XXX: must remove usage of PageReserved! - */ - if (PageReserved(page)) - continue; - BUG_ON(PageNosave(page)); - if (PageNosaveFree(page)) - continue; - save = kmalloc(sizeof(struct highmem_page), GFP_ATOMIC); - if (!save) - return -ENOMEM; - save->next = highmem_copy; - save->page = page; - save->data = (void *) get_zeroed_page(GFP_ATOMIC); - if (!save->data) { - kfree(save); - return -ENOMEM; - } - kaddr = kmap_atomic(page, KM_USER0); - memcpy(save->data, kaddr, PAGE_SIZE); - kunmap_atomic(kaddr, KM_USER0); - highmem_copy = save; - } - return 0; -} -#endif /* CONFIG_HIGHMEM */ - - -static int save_highmem(void) -{ -#ifdef CONFIG_HIGHMEM - struct zone *zone; - int res = 0; - - pr_debug("swsusp: Saving Highmem\n"); - for_each_zone (zone) { - if (is_highmem(zone)) - res = save_highmem_zone(zone); - if (res) - return res; - } -#endif - return 0; -} - -static int restore_highmem(void) -{ -#ifdef CONFIG_HIGHMEM - printk("swsusp: Restoring Highmem\n"); - while (highmem_copy) { - struct highmem_page *save = highmem_copy; - void *kaddr; - highmem_copy = save->next; - - kaddr = kmap_atomic(save->page, KM_USER0); - memcpy(kaddr, save->data, PAGE_SIZE); - kunmap_atomic(kaddr, KM_USER0); - free_page((long) save->data); - kfree(save); - } -#endif - return 0; -} - - -static int pfn_is_nosave(unsigned long pfn) -{ - unsigned long nosave_begin_pfn = __pa(&__nosave_begin) >> PAGE_SHIFT; - unsigned long nosave_end_pfn = PAGE_ALIGN(__pa(&__nosave_end)) >> PAGE_SHIFT; - return (pfn >= nosave_begin_pfn) && (pfn < nosave_end_pfn); -} - -/** - * saveable - Determine whether a page should be cloned or not. - * @pfn: The page - * - * We save a page if it's Reserved, and not in the range of pages - * statically defined as 'unsaveable', or if it isn't reserved, and - * isn't part of a free chunk of pages. - */ - -static int saveable(struct zone * zone, unsigned long * zone_pfn) -{ - unsigned long pfn = *zone_pfn + zone->zone_start_pfn; - struct page * page; - - if (!pfn_valid(pfn)) - return 0; - - page = pfn_to_page(pfn); - if (PageNosave(page)) - return 0; - if (pfn_is_nosave(pfn)) { - pr_debug("[nosave pfn 0x%lx]", pfn); - return 0; - } - if (PageNosaveFree(page)) - return 0; - - return 1; -} - -static void count_data_pages(void) -{ - struct zone *zone; - unsigned long zone_pfn; - - nr_copy_pages = 0; - - for_each_zone (zone) { - if (is_highmem(zone)) - continue; - mark_free_pages(zone); - for (zone_pfn = 0; zone_pfn < zone->spanned_pages; ++zone_pfn) - nr_copy_pages += saveable(zone, &zone_pfn); - } -} - -static void copy_data_pages(void) -{ - struct zone *zone; - unsigned long zone_pfn; - struct pbe *pbe = pagedir_nosave, *p; - - pr_debug("copy_data_pages(): pages to copy: %d\n", nr_copy_pages); - for_each_zone (zone) { - if (is_highmem(zone)) - continue; - mark_free_pages(zone); - /* This is necessary for swsusp_free() */ - for_each_pb_page (p, pagedir_nosave) - SetPageNosaveFree(virt_to_page(p)); - for_each_pbe(p, pagedir_nosave) - SetPageNosaveFree(virt_to_page(p->address)); - for (zone_pfn = 0; zone_pfn < zone->spanned_pages; ++zone_pfn) { - if (saveable(zone, &zone_pfn)) { - struct page * page; - page = pfn_to_page(zone_pfn + zone->zone_start_pfn); - BUG_ON(!pbe); - pbe->orig_address = (unsigned long)page_address(page); - /* copy_page is not usable for copying task structs. */ - memcpy((void *)pbe->address, (void *)pbe->orig_address, PAGE_SIZE); - pbe = pbe->next; - } - } - } - BUG_ON(pbe); -} - - -/** - * free_pagedir - free pages allocated with alloc_pagedir() - */ - -static inline void free_pagedir(struct pbe *pblist) -{ - struct pbe *pbe; - - while (pblist) { - pbe = (pblist + PB_PAGE_SKIP)->next; - ClearPageNosave(virt_to_page(pblist)); - ClearPageNosaveFree(virt_to_page(pblist)); - free_page((unsigned long)pblist); - pblist = pbe; - } -} - -/** - * fill_pb_page - Create a list of PBEs on a given memory page - */ - -static inline void fill_pb_page(struct pbe *pbpage) -{ - struct pbe *p; - - p = pbpage; - pbpage += PB_PAGE_SKIP; - do - p->next = p + 1; - while (++p < pbpage); -} - -/** - * create_pbe_list - Create a list of PBEs on top of a given chain - * of memory pages allocated with alloc_pagedir() - */ - -static void create_pbe_list(struct pbe *pblist, unsigned nr_pages) -{ - struct pbe *pbpage, *p; - unsigned num = PBES_PER_PAGE; - - for_each_pb_page (pbpage, pblist) { - if (num >= nr_pages) - break; - - fill_pb_page(pbpage); - num += PBES_PER_PAGE; - } - if (pbpage) { - for (num -= PBES_PER_PAGE - 1, p = pbpage; num < nr_pages; p++, num++) - p->next = p + 1; - p->next = NULL; - } - pr_debug("create_pbe_list(): initialized %d PBEs\n", num); -} - -static void *alloc_image_page(void) -{ - void *res = (void *)get_zeroed_page(GFP_ATOMIC | __GFP_COLD); - if (res) { - SetPageNosave(virt_to_page(res)); - SetPageNosaveFree(virt_to_page(res)); - } - return res; -} - -/** - * alloc_pagedir - Allocate the page directory. - * - * First, determine exactly how many pages we need and - * allocate them. - * - * We arrange the pages in a chain: each page is an array of PBES_PER_PAGE - * struct pbe elements (pbes) and the last element in the page points - * to the next page. - * - * On each page we set up a list of struct_pbe elements. - */ - -static struct pbe * alloc_pagedir(unsigned nr_pages) -{ - unsigned num; - struct pbe *pblist, *pbe; - - if (!nr_pages) - return NULL; - - pr_debug("alloc_pagedir(): nr_pages = %d\n", nr_pages); - pblist = (struct pbe *)alloc_image_page(); - for (pbe = pblist, num = PBES_PER_PAGE; pbe && num < nr_pages; - pbe = pbe->next, num += PBES_PER_PAGE) { - pbe += PB_PAGE_SKIP; - pbe->next = (struct pbe *)alloc_image_page(); - } - if (!pbe) { /* get_zeroed_page() failed */ - free_pagedir(pblist); - pblist = NULL; - } - return pblist; -} - -/** - * Free pages we allocated for suspend. Suspend pages are alocated - * before atomic copy, so we need to free them after resume. - */ - -void swsusp_free(void) -{ - struct zone *zone; - unsigned long zone_pfn; - - for_each_zone(zone) { - for (zone_pfn = 0; zone_pfn < zone->spanned_pages; ++zone_pfn) - if (pfn_valid(zone_pfn + zone->zone_start_pfn)) { - struct page * page; - page = pfn_to_page(zone_pfn + zone->zone_start_pfn); - if (PageNosave(page) && PageNosaveFree(page)) { - ClearPageNosave(page); - ClearPageNosaveFree(page); - free_page((long) page_address(page)); - } - } - } -} - -/** - * enough_free_mem - Make sure we enough free memory to snapshot. - * - * Returns TRUE or FALSE after checking the number of available - * free pages. - */ - -static int enough_free_mem(void) -{ - pr_debug("swsusp: available memory: %u pages\n", nr_free_pages()); - return nr_free_pages() > (nr_copy_pages + PAGES_FOR_IO + - nr_copy_pages/PBES_PER_PAGE + !!(nr_copy_pages%PBES_PER_PAGE)); -} - - /** * enough_swap - Make sure we have enough swap to save the image. * @@ -891,7 +550,7 @@ static int enough_free_mem(void) * We should only consider resume_device. */ -static int enough_swap(void) +int enough_swap(void) { struct sysinfo i; @@ -901,88 +560,6 @@ static int enough_swap(void) nr_copy_pages/PBES_PER_PAGE + !!(nr_copy_pages%PBES_PER_PAGE)); } -static int swsusp_alloc(void) -{ - struct pbe * p; - - pagedir_nosave = NULL; - - if (!(pagedir_save = alloc_pagedir(nr_copy_pages))) { - printk(KERN_ERR "suspend: Allocating pagedir failed.\n"); - return -ENOMEM; - } - create_pbe_list(pagedir_save, nr_copy_pages); - pagedir_nosave = pagedir_save; - - for_each_pbe (p, pagedir_save) { - p->address = (unsigned long)alloc_image_page(); - if (!p->address) { - printk(KERN_ERR "suspend: Allocating image pages failed.\n"); - swsusp_free(); - return -ENOMEM; - } - } - - return 0; -} - -static int suspend_prepare_image(void) -{ - int error; - - pr_debug("swsusp: critical section: \n"); - if (save_highmem()) { - printk(KERN_CRIT "swsusp: Not enough free pages for highmem\n"); - restore_highmem(); - return -ENOMEM; - } - - drain_local_pages(); - count_data_pages(); - printk("swsusp: Need to copy %u pages\n", nr_copy_pages); - nr_copy_pages_check = nr_copy_pages; - - pr_debug("swsusp: pages needed: %u + %lu + %u, free: %u\n", - nr_copy_pages, - nr_copy_pages/PBES_PER_PAGE + !!(nr_copy_pages%PBES_PER_PAGE), - PAGES_FOR_IO, nr_free_pages()); - - if (!enough_free_mem()) { - printk(KERN_ERR "swsusp: Not enough free memory\n"); - return -ENOMEM; - } - - if (MAX_PBES < nr_copy_pages / PBES_PER_PAGE + - !!(nr_copy_pages % PBES_PER_PAGE)) { - printk(KERN_ERR "swsusp: Too many image pages\n"); - return -ENOSPC; - } - - if (!enough_swap()) { - printk(KERN_ERR "swsusp: Not enough free swap\n"); - return -ENOSPC; - } - - error = swsusp_alloc(); - if (error) - return error; - - /* During allocating of suspend pagedir, new cold pages may appear. - * Kill them. - */ - drain_local_pages(); - copy_data_pages(); - - /* - * End of critical section. From now on, we can write to memory, - * but we should not touch disk. This specially means we must _not_ - * touch swap space! Except we must write out our image of course. - */ - - printk("swsusp: critical section/: done (%d pages copied)\n", nr_copy_pages ); - return 0; -} - /* It is important _NOT_ to umount filesystems at this point. We want * them synced (in case something goes wrong) but we DO not want to mark @@ -1002,14 +579,6 @@ int swsusp_write(void) } -extern asmlinkage int swsusp_arch_suspend(void); -extern asmlinkage int swsusp_arch_resume(void); - - -asmlinkage int swsusp_save(void) -{ - return suspend_prepare_image(); -} int swsusp_suspend(void) { @@ -1041,7 +610,6 @@ int swsusp_suspend(void) printk(KERN_ERR "Error %d suspending\n", error); /* Restore control flow magically appears here */ restore_processor_state(); - BUG_ON (nr_copy_pages_check != nr_copy_pages); restore_highmem(); device_power_up(); local_irq_enable(); -- cgit v1.2.3 From 2c1b4a5ca48831595979a850f40ced8e7da026f8 Mon Sep 17 00:00:00 2001 From: "Rafael J. Wysocki" Date: Sun, 30 Oct 2005 14:59:58 -0800 Subject: [PATCH] swsusp: rework memory freeing on resume The following patch makes swsusp use the PG_nosave and PG_nosave_free flags to mark pages that should be freed in case of an error during resume. This allows us to simplify the code and to use swsusp_free() in all of the swsusp's resume error paths, which makes them actually work. Signed-off-by: Rafael J. Wysocki Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/x86_64/kernel/suspend.c | 84 ++++++++------------------------- include/linux/suspend.h | 3 +- kernel/power/disk.c | 14 +++--- kernel/power/power.h | 2 +- kernel/power/snapshot.c | 2 +- kernel/power/swsusp.c | 110 ++++++++++++++----------------------------- 6 files changed, 65 insertions(+), 150 deletions(-) (limited to 'include/linux') diff --git a/arch/x86_64/kernel/suspend.c b/arch/x86_64/kernel/suspend.c index 02516823f514..fd2bef780882 100644 --- a/arch/x86_64/kernel/suspend.c +++ b/arch/x86_64/kernel/suspend.c @@ -147,57 +147,7 @@ extern int restore_image(void); pgd_t *temp_level4_pgt; -static void **pages; - -static inline void *__add_page(void) -{ - void **c; - - c = (void **)get_usable_page(GFP_ATOMIC); - if (c) { - *c = pages; - pages = c; - } - return c; -} - -static inline void *__next_page(void) -{ - void **c; - - c = pages; - if (c) { - pages = *c; - *c = NULL; - } - return c; -} - -/* - * Try to allocate as many usable pages as needed and daisy chain them. - * If one allocation fails, free the pages allocated so far - */ -static int alloc_usable_pages(unsigned long n) -{ - void *p; - - pages = NULL; - do - if (!__add_page()) - break; - while (--n); - if (n) { - p = __next_page(); - while (p) { - free_page((unsigned long)p); - p = __next_page(); - } - return -ENOMEM; - } - return 0; -} - -static void res_phys_pud_init(pud_t *pud, unsigned long address, unsigned long end) +static int res_phys_pud_init(pud_t *pud, unsigned long address, unsigned long end) { long i, j; @@ -211,7 +161,9 @@ static void res_phys_pud_init(pud_t *pud, unsigned long address, unsigned long e if (paddr >= end) break; - pmd = (pmd_t *)__next_page(); + pmd = (pmd_t *)get_safe_page(GFP_ATOMIC); + if (!pmd) + return -ENOMEM; set_pud(pud, __pud(__pa(pmd) | _KERNPG_TABLE)); for (j = 0; j < PTRS_PER_PMD; pmd++, j++, paddr += PMD_SIZE) { unsigned long pe; @@ -223,13 +175,17 @@ static void res_phys_pud_init(pud_t *pud, unsigned long address, unsigned long e set_pmd(pmd, __pmd(pe)); } } + return 0; } -static void set_up_temporary_mappings(void) +static int set_up_temporary_mappings(void) { unsigned long start, end, next; + int error; - temp_level4_pgt = (pgd_t *)__next_page(); + temp_level4_pgt = (pgd_t *)get_safe_page(GFP_ATOMIC); + if (!temp_level4_pgt) + return -ENOMEM; /* It is safe to reuse the original kernel mapping */ set_pgd(temp_level4_pgt + pgd_index(__START_KERNEL_map), @@ -240,29 +196,27 @@ static void set_up_temporary_mappings(void) end = (unsigned long)pfn_to_kaddr(end_pfn); for (; start < end; start = next) { - pud_t *pud = (pud_t *)__next_page(); + pud_t *pud = (pud_t *)get_safe_page(GFP_ATOMIC); + if (!pud) + return -ENOMEM; next = start + PGDIR_SIZE; if (next > end) next = end; - res_phys_pud_init(pud, __pa(start), __pa(next)); + if ((error = res_phys_pud_init(pud, __pa(start), __pa(next)))) + return error; set_pgd(temp_level4_pgt + pgd_index(start), mk_kernel_pgd(__pa(pud))); } + return 0; } int swsusp_arch_resume(void) { - unsigned long n; + int error; - n = ((end_pfn << PAGE_SHIFT) + PUD_SIZE - 1) >> PUD_SHIFT; - n += (n + PTRS_PER_PUD - 1) / PTRS_PER_PUD + 1; - pr_debug("swsusp_arch_resume(): pages needed = %lu\n", n); - if (alloc_usable_pages(n)) { - free_eaten_memory(); - return -ENOMEM; - } /* We have got enough memory and from now on we cannot recover */ - set_up_temporary_mappings(); + if ((error = set_up_temporary_mappings())) + return error; restore_image(); return 0; } diff --git a/include/linux/suspend.h b/include/linux/suspend.h index 380915e9563d..a61c04f804b2 100644 --- a/include/linux/suspend.h +++ b/include/linux/suspend.h @@ -71,8 +71,7 @@ void restore_processor_state(void); struct saved_context; void __save_processor_state(struct saved_context *ctxt); void __restore_processor_state(struct saved_context *ctxt); -extern unsigned long get_usable_page(gfp_t gfp_mask); -extern void free_eaten_memory(void); +unsigned long get_safe_page(gfp_t gfp_mask); /* * XXX: We try to keep some more pages free so that I/O operations succeed diff --git a/kernel/power/disk.c b/kernel/power/disk.c index 761956e813f5..44ef5e799df0 100644 --- a/kernel/power/disk.c +++ b/kernel/power/disk.c @@ -30,7 +30,6 @@ extern int swsusp_check(void); extern int swsusp_read(void); extern void swsusp_close(void); extern int swsusp_resume(void); -extern int swsusp_free(void); static int noresume = 0; @@ -252,14 +251,17 @@ static int software_resume(void) pr_debug("PM: Reading swsusp image.\n"); - if ((error = swsusp_read())) - goto Cleanup; + if ((error = swsusp_read())) { + swsusp_free(); + goto Thaw; + } pr_debug("PM: Preparing devices for restore.\n"); if ((error = device_suspend(PMSG_FREEZE))) { printk("Some devices failed to suspend\n"); - goto Free; + swsusp_free(); + goto Thaw; } mb(); @@ -268,9 +270,7 @@ static int software_resume(void) swsusp_resume(); pr_debug("PM: Restore failed, recovering.n"); device_resume(); - Free: - swsusp_free(); - Cleanup: + Thaw: unprepare_processes(); Done: /* For success case, the suspend path will release the lock */ diff --git a/kernel/power/power.h b/kernel/power/power.h index 28afcb090149..d4fd96a135ab 100644 --- a/kernel/power/power.h +++ b/kernel/power/power.h @@ -66,7 +66,7 @@ extern asmlinkage int swsusp_arch_suspend(void); extern asmlinkage int swsusp_arch_resume(void); extern int restore_highmem(void); -extern void free_pagedir(struct pbe *pblist); extern struct pbe * alloc_pagedir(unsigned nr_pages); extern void create_pbe_list(struct pbe *pblist, unsigned nr_pages); +extern void swsusp_free(void); extern int enough_swap(unsigned nr_pages); diff --git a/kernel/power/snapshot.c b/kernel/power/snapshot.c index 03916cf3ff02..84e686bdb40b 100644 --- a/kernel/power/snapshot.c +++ b/kernel/power/snapshot.c @@ -240,7 +240,7 @@ static void copy_data_pages(struct pbe *pblist) * free_pagedir - free pages allocated with alloc_pagedir() */ -void free_pagedir(struct pbe *pblist) +static void free_pagedir(struct pbe *pblist) { struct pbe *pbe; diff --git a/kernel/power/swsusp.c b/kernel/power/swsusp.c index f6abfdb0a02a..50667f4f3a2b 100644 --- a/kernel/power/swsusp.c +++ b/kernel/power/swsusp.c @@ -629,6 +629,11 @@ int swsusp_resume(void) * execution continues at place where swsusp_arch_suspend was called */ BUG_ON(!error); + /* The only reason why swsusp_arch_resume() can fail is memory being + * very tight, so we have to free it as soon as we can to avoid + * subsequent failures + */ + swsusp_free(); restore_processor_state(); restore_highmem(); touch_softlockup_watchdog(); @@ -644,54 +649,28 @@ int swsusp_resume(void) * * We don't know which pages are usable until we allocate them. * - * Allocated but unusable (ie eaten) memory pages are linked together - * to create a list, so that we can free them easily - * - * We could have used a type other than (void *) - * for this purpose, but ... + * Allocated but unusable (ie eaten) memory pages are marked so that + * swsusp_free() can release them */ -static void **eaten_memory = NULL; -static inline void eat_page(void *page) -{ - void **c; - - c = eaten_memory; - eaten_memory = page; - *eaten_memory = c; -} - -unsigned long get_usable_page(gfp_t gfp_mask) +unsigned long get_safe_page(gfp_t gfp_mask) { unsigned long m; - m = get_zeroed_page(gfp_mask); - while (!PageNosaveFree(virt_to_page(m))) { - eat_page((void *)m); + do { m = get_zeroed_page(gfp_mask); - if (!m) - break; + if (m && PageNosaveFree(virt_to_page(m))) + /* This is for swsusp_free() */ + SetPageNosave(virt_to_page(m)); + } while (m && PageNosaveFree(virt_to_page(m))); + if (m) { + /* This is for swsusp_free() */ + SetPageNosave(virt_to_page(m)); + SetPageNosaveFree(virt_to_page(m)); } return m; } -void free_eaten_memory(void) -{ - unsigned long m; - void **c; - int i = 0; - - c = eaten_memory; - while (c) { - m = (unsigned long)c; - c = *c; - free_page(m); - i++; - } - eaten_memory = NULL; - pr_debug("swsusp: %d unused pages freed\n", i); -} - /** * check_pagedir - We ensure here that pages that the PBEs point to * won't collide with pages where we're going to restore from the loaded @@ -709,7 +688,7 @@ static int check_pagedir(struct pbe *pblist) p->address = 0UL; for_each_pbe (p, pblist) { - p->address = get_usable_page(GFP_ATOMIC); + p->address = get_safe_page(GFP_ATOMIC); if (!p->address) return -ENOMEM; } @@ -728,7 +707,7 @@ static struct pbe * swsusp_pagedir_relocate(struct pbe *pblist) unsigned long zone_pfn; struct pbe *pbpage, *tail, *p; void *m; - int rel = 0, error = 0; + int rel = 0; if (!pblist) /* a sanity check */ return NULL; @@ -736,41 +715,37 @@ static struct pbe * swsusp_pagedir_relocate(struct pbe *pblist) pr_debug("swsusp: Relocating pagedir (%lu pages to check)\n", swsusp_info.pagedir_pages); - /* Set page flags */ + /* Clear page flags */ for_each_zone (zone) { for (zone_pfn = 0; zone_pfn < zone->spanned_pages; ++zone_pfn) - SetPageNosaveFree(pfn_to_page(zone_pfn + + if (pfn_valid(zone_pfn + zone->zone_start_pfn)) + ClearPageNosaveFree(pfn_to_page(zone_pfn + zone->zone_start_pfn)); } - /* Clear orig addresses */ + /* Mark orig addresses */ for_each_pbe (p, pblist) - ClearPageNosaveFree(virt_to_page(p->orig_address)); + SetPageNosaveFree(virt_to_page(p->orig_address)); tail = pblist + PB_PAGE_SKIP; /* Relocate colliding pages */ for_each_pb_page (pbpage, pblist) { - if (!PageNosaveFree(virt_to_page((unsigned long)pbpage))) { - m = (void *)get_usable_page(GFP_ATOMIC | __GFP_COLD); - if (!m) { - error = -ENOMEM; - break; - } + if (PageNosaveFree(virt_to_page((unsigned long)pbpage))) { + m = (void *)get_safe_page(GFP_ATOMIC | __GFP_COLD); + if (!m) + return NULL; memcpy(m, (void *)pbpage, PAGE_SIZE); if (pbpage == pblist) pblist = (struct pbe *)m; else tail->next = (struct pbe *)m; - - eat_page((void *)pbpage); pbpage = (struct pbe *)m; /* We have to link the PBEs again */ - for (p = pbpage; p < pbpage + PB_PAGE_SKIP; p++) if (p->next) /* needed to save the end */ p->next = p + 1; @@ -780,15 +755,13 @@ static struct pbe * swsusp_pagedir_relocate(struct pbe *pblist) tail = pbpage + PB_PAGE_SKIP; } - if (error) { - printk("\nswsusp: Out of memory\n\n"); - free_pagedir(pblist); - free_eaten_memory(); - pblist = NULL; - /* Is this even worth handling? It should never ever happen, and we - have just lost user's state, anyway... */ - } else - printk("swsusp: Relocated %d pages\n", rel); + /* This is for swsusp_free() */ + for_each_pb_page (pbpage, pblist) { + SetPageNosave(virt_to_page(pbpage)); + SetPageNosaveFree(virt_to_page(pbpage)); + } + + printk("swsusp: Relocated %d pages\n", rel); return pblist; } @@ -1006,9 +979,7 @@ static int read_pagedir(struct pbe *pblist) break; } - if (error) - free_pagedir(pblist); - else + if (!error) BUG_ON(i != swsusp_info.pagedir_pages); return error; @@ -1051,15 +1022,6 @@ static int read_suspend_image(void) if (!error) error = data_read(pagedir_nosave); - if (error) { /* We fail cleanly */ - free_eaten_memory(); - for_each_pbe (p, pagedir_nosave) - if (p->address) { - free_page(p->address); - p->address = 0UL; - } - free_pagedir(pagedir_nosave); - } return error; } -- cgit v1.2.3 From eb9289eb20df6b54214c45ac7c6bf5179a149026 Mon Sep 17 00:00:00 2001 From: Shaohua Li Date: Sun, 30 Oct 2005 15:00:01 -0800 Subject: [PATCH] introduce .valid callback for pm_ops Add pm_ops.valid callback, so only the available pm states show in /sys/power/state. And this also makes an earlier states error report at enter_state before we do actual suspend/resume. Signed-off-by: Shaohua Li Acked-by: Pavel Machek Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/acpi/sleep/main.c | 8 ++++++++ include/linux/pm.h | 1 + kernel/power/main.c | 5 ++++- 3 files changed, 13 insertions(+), 1 deletion(-) (limited to 'include/linux') diff --git a/drivers/acpi/sleep/main.c b/drivers/acpi/sleep/main.c index aee50b453265..930427fc0c4b 100644 --- a/drivers/acpi/sleep/main.c +++ b/drivers/acpi/sleep/main.c @@ -158,7 +158,15 @@ int acpi_suspend(u32 acpi_state) return -EINVAL; } +static int acpi_pm_state_valid(suspend_state_t pm_state) +{ + u32 acpi_state = acpi_suspend_states[pm_state]; + + return sleep_states[acpi_state]; +} + static struct pm_ops acpi_pm_ops = { + .valid = acpi_pm_state_valid, .prepare = acpi_pm_prepare, .enter = acpi_pm_enter, .finish = acpi_pm_finish, diff --git a/include/linux/pm.h b/include/linux/pm.h index c61d5de837ef..1514098d156d 100644 --- a/include/linux/pm.h +++ b/include/linux/pm.h @@ -170,6 +170,7 @@ typedef int __bitwise suspend_disk_method_t; struct pm_ops { suspend_disk_method_t pm_disk_mode; + int (*valid)(suspend_state_t state); int (*prepare)(suspend_state_t state); int (*enter)(suspend_state_t state); int (*finish)(suspend_state_t state); diff --git a/kernel/power/main.c b/kernel/power/main.c index 22bdc93cc038..18d7d693fbba 100644 --- a/kernel/power/main.c +++ b/kernel/power/main.c @@ -167,6 +167,8 @@ static int enter_state(suspend_state_t state) { int error; + if (pm_ops->valid && !pm_ops->valid(state)) + return -ENODEV; if (down_trylock(&pm_sem)) return -EBUSY; @@ -236,7 +238,8 @@ static ssize_t state_show(struct subsystem * subsys, char * buf) char * s = buf; for (i = 0; i < PM_SUSPEND_MAX; i++) { - if (pm_states[i]) + if (pm_states[i] && pm_ops && (!pm_ops->valid + ||(pm_ops->valid && pm_ops->valid(i)))) s += sprintf(s,"%s ",pm_states[i]); } s += sprintf(s,"\n"); -- cgit v1.2.3 From aaa4059bc2dca7fa816624a28db1958c3a22df9b Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Sun, 30 Oct 2005 15:00:16 -0800 Subject: [PATCH] ext3: Fix unmapped buffers in transaction's lists Fix the problem (BUG 4964) with unmapped buffers in transaction's t_sync_data list. The problem is we need to call filesystem's own invalidatepage() from block_write_full_page(). block_write_full_page() must call filesystem's invalidatepage(). Otherwise following nasty race can happen: proc 1 proc 2 ------ ------ - write some new data to 'offset' => bh gets to the transactions data list - starts truncate => i_size set to new size - mpage_writepages() - ext3_ordered_writepage() to 'offset' - block_write_full_page() - page->index > end_index+1 - block_invalidatepage() - discard_buffer() - clear_buffer_mapped() - commit triggers and finds unmapped buffer - BOOM! Signed-off-by: Jan Kara Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/buffer.c | 11 ++++++++++- include/linux/buffer_head.h | 1 + mm/truncate.c | 11 +---------- 3 files changed, 12 insertions(+), 11 deletions(-) (limited to 'include/linux') diff --git a/fs/buffer.c b/fs/buffer.c index 2066e4cb700c..75cac9ada026 100644 --- a/fs/buffer.c +++ b/fs/buffer.c @@ -1637,6 +1637,15 @@ out: } EXPORT_SYMBOL(block_invalidatepage); +int do_invalidatepage(struct page *page, unsigned long offset) +{ + int (*invalidatepage)(struct page *, unsigned long); + invalidatepage = page->mapping->a_ops->invalidatepage; + if (invalidatepage == NULL) + invalidatepage = block_invalidatepage; + return (*invalidatepage)(page, offset); +} + /* * We attach and possibly dirty the buffers atomically wrt * __set_page_dirty_buffers() via private_lock. try_to_free_buffers @@ -2696,7 +2705,7 @@ int block_write_full_page(struct page *page, get_block_t *get_block, * they may have been added in ext3_writepage(). Make them * freeable here, so the page does not leak. */ - block_invalidatepage(page, 0); + do_invalidatepage(page, 0); unlock_page(page); return 0; /* don't care */ } diff --git a/include/linux/buffer_head.h b/include/linux/buffer_head.h index c937d6e65502..1db061bb6b08 100644 --- a/include/linux/buffer_head.h +++ b/include/linux/buffer_head.h @@ -190,6 +190,7 @@ extern int buffer_heads_over_limit; */ int try_to_release_page(struct page * page, gfp_t gfp_mask); int block_invalidatepage(struct page *page, unsigned long offset); +int do_invalidatepage(struct page *page, unsigned long offset); int block_write_full_page(struct page *page, get_block_t *get_block, struct writeback_control *wbc); int block_read_full_page(struct page*, get_block_t*); diff --git a/mm/truncate.c b/mm/truncate.c index 60c8764bfac2..29c18f68dc35 100644 --- a/mm/truncate.c +++ b/mm/truncate.c @@ -13,18 +13,9 @@ #include #include #include /* grr. try_to_release_page, - block_invalidatepage */ + do_invalidatepage */ -static int do_invalidatepage(struct page *page, unsigned long offset) -{ - int (*invalidatepage)(struct page *, unsigned long); - invalidatepage = page->mapping->a_ops->invalidatepage; - if (invalidatepage == NULL) - invalidatepage = block_invalidatepage; - return (*invalidatepage)(page, offset); -} - static inline void truncate_partial_page(struct page *page, unsigned partial) { memclear_highpage_flush(page, partial, PAGE_CACHE_SIZE-partial); -- cgit v1.2.3 From a8db2db1e6a8d323d87a67c5391d48fe2b97faf5 Mon Sep 17 00:00:00 2001 From: Oleg Nesterov Date: Sun, 30 Oct 2005 15:01:38 -0800 Subject: [PATCH] introduce setup_timer() helper Every user of init_timer() also needs to initialize ->function and ->data fields. This patch adds a simple setup_timer() helper for that. The schedule_timeout() is patched as an example of usage. Signed-off-by: Oleg Nesterov Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/x86_64/kernel/i8259.c | 8 ++++---- include/linux/timer.h | 9 +++++++++ kernel/timer.c | 8 ++------ 3 files changed, 15 insertions(+), 10 deletions(-) (limited to 'include/linux') diff --git a/arch/x86_64/kernel/i8259.c b/arch/x86_64/kernel/i8259.c index b2a238b5a17e..c6c9791d77c1 100644 --- a/arch/x86_64/kernel/i8259.c +++ b/arch/x86_64/kernel/i8259.c @@ -494,7 +494,7 @@ void invalidate_interrupt7(void); void thermal_interrupt(void); void i8254_timer_resume(void); -static void setup_timer(void) +static void setup_timer_hardware(void) { outb_p(0x34,0x43); /* binary, mode 2, LSB/MSB, ch 0 */ udelay(10); @@ -505,13 +505,13 @@ static void setup_timer(void) static int timer_resume(struct sys_device *dev) { - setup_timer(); + setup_timer_hardware(); return 0; } void i8254_timer_resume(void) { - setup_timer(); + setup_timer_hardware(); } static struct sysdev_class timer_sysclass = { @@ -594,7 +594,7 @@ void __init init_IRQ(void) * Set the clock to HZ Hz, we already have a valid * vector now: */ - setup_timer(); + setup_timer_hardware(); if (!acpi_ioapic) setup_irq(2, &irq2); diff --git a/include/linux/timer.h b/include/linux/timer.h index 3340f3bd135d..ddd5bbe1fc8e 100644 --- a/include/linux/timer.h +++ b/include/linux/timer.h @@ -38,6 +38,15 @@ extern struct timer_base_s __init_timer_base; void fastcall init_timer(struct timer_list * timer); +static inline void setup_timer(struct timer_list * timer, + void (*function)(unsigned long), + unsigned long data) +{ + timer->function = function; + timer->data = data; + init_timer(timer); +} + /*** * timer_pending - is a timer pending? * @timer: the timer in question diff --git a/kernel/timer.c b/kernel/timer.c index 6a2e5f8dc725..6ed1a826e5ce 100644 --- a/kernel/timer.c +++ b/kernel/timer.c @@ -1146,12 +1146,8 @@ fastcall signed long __sched schedule_timeout(signed long timeout) expire = timeout + jiffies; - init_timer(&timer); - timer.expires = expire; - timer.data = (unsigned long) current; - timer.function = process_timeout; - - add_timer(&timer); + setup_timer(&timer, process_timeout, (unsigned long)current); + __mod_timer(&timer, expire); schedule(); del_singleshot_timer_sync(&timer); -- cgit v1.2.3 From 61e1a9ea4b425eb8c3b4965c35fe953bd881728f Mon Sep 17 00:00:00 2001 From: Alan Stern Date: Sun, 30 Oct 2005 15:01:40 -0800 Subject: [PATCH] Add kthread_stop_sem() Enhance the kthread API by adding kthread_stop_sem, for use in stopping threads that spend their idle time waiting on a semaphore. Signed-off-by: Alan Stern Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/kthread.h | 12 ++++++++++++ kernel/kthread.c | 13 +++++++++++-- 2 files changed, 23 insertions(+), 2 deletions(-) (limited to 'include/linux') diff --git a/include/linux/kthread.h b/include/linux/kthread.h index 3fa786448db3..ebdd41fd1082 100644 --- a/include/linux/kthread.h +++ b/include/linux/kthread.h @@ -69,6 +69,18 @@ void kthread_bind(struct task_struct *k, unsigned int cpu); * was never called. */ int kthread_stop(struct task_struct *k); +/** + * kthread_stop_sem: stop a thread created by kthread_create(). + * @k: thread created by kthread_create(). + * @s: semaphore that @k waits on while idle. + * + * Does essentially the same thing as kthread_stop() above, but wakes + * @k by calling up(@s). + * + * Returns the result of threadfn(), or -EINTR if wake_up_process() + * was never called. */ +int kthread_stop_sem(struct task_struct *k, struct semaphore *s); + /** * kthread_should_stop: should this kthread return now? * diff --git a/kernel/kthread.c b/kernel/kthread.c index f50f174e92da..e75950a1092c 100644 --- a/kernel/kthread.c +++ b/kernel/kthread.c @@ -164,6 +164,12 @@ void kthread_bind(struct task_struct *k, unsigned int cpu) EXPORT_SYMBOL(kthread_bind); int kthread_stop(struct task_struct *k) +{ + return kthread_stop_sem(k, NULL); +} +EXPORT_SYMBOL(kthread_stop); + +int kthread_stop_sem(struct task_struct *k, struct semaphore *s) { int ret; @@ -178,7 +184,10 @@ int kthread_stop(struct task_struct *k) /* Now set kthread_should_stop() to true, and wake it up. */ kthread_stop_info.k = k; - wake_up_process(k); + if (s) + up(s); + else + wake_up_process(k); put_task_struct(k); /* Once it dies, reset stop ptr, gather result and we're done. */ @@ -189,7 +198,7 @@ int kthread_stop(struct task_struct *k) return ret; } -EXPORT_SYMBOL(kthread_stop); +EXPORT_SYMBOL(kthread_stop_sem); static __init int helper_init(void) { -- cgit v1.2.3 From 1bb34a412750291e4e5e9f1d0fe7ae1b7e976098 Mon Sep 17 00:00:00 2001 From: john stultz Date: Sun, 30 Oct 2005 15:01:42 -0800 Subject: [PATCH] NTP shift_right cleanup Create a macro shift_right() that avoids the numerous ugly conditionals in the NTP code that look like: if(a < 0) b = -(-a >> shift); else b = a >> shift; Replacing it with: b = shift_right(a, shift); This should have zero effect on the logic, however it should probably have a bit of testing just to be sure. Also replace open-coded min/max with the macros. Signed-off-by : John Stultz Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/timex.h | 7 +++++++ kernel/time.c | 25 ++++++------------------ kernel/timer.c | 53 +++++++++++---------------------------------------- 3 files changed, 24 insertions(+), 61 deletions(-) (limited to 'include/linux') diff --git a/include/linux/timex.h b/include/linux/timex.h index 7e050a2cc35b..04a4a8cb4ed3 100644 --- a/include/linux/timex.h +++ b/include/linux/timex.h @@ -282,6 +282,13 @@ static inline int ntp_synced(void) return !(time_status & STA_UNSYNC); } +/* Required to safely shift negative values */ +#define shift_right(x, s) ({ \ + __typeof__(x) __x = (x); \ + __typeof__(s) __s = (s); \ + __x < 0 ? -(-__x >> __s) : __x >> __s; \ +}) + #ifdef CONFIG_TIME_INTERPOLATION diff --git a/kernel/time.c b/kernel/time.c index a3c2100470e1..245d595a13cb 100644 --- a/kernel/time.c +++ b/kernel/time.c @@ -338,30 +338,20 @@ int do_adjtimex(struct timex *txc) if (mtemp >= MINSEC) { ltemp = (time_offset / mtemp) << (SHIFT_USEC - SHIFT_UPDATE); - if (ltemp < 0) - time_freq -= -ltemp >> SHIFT_KH; - else - time_freq += ltemp >> SHIFT_KH; + time_freq += shift_right(ltemp, SHIFT_KH); } else /* calibration interval too short (p. 12) */ result = TIME_ERROR; } else { /* PLL mode */ if (mtemp < MAXSEC) { ltemp *= mtemp; - if (ltemp < 0) - time_freq -= -ltemp >> (time_constant + - time_constant + - SHIFT_KF - SHIFT_USEC); - else - time_freq += ltemp >> (time_constant + + time_freq += shift_right(ltemp,(time_constant + time_constant + - SHIFT_KF - SHIFT_USEC); + SHIFT_KF - SHIFT_USEC)); } else /* calibration interval too long (p. 12) */ result = TIME_ERROR; } - if (time_freq > time_tolerance) - time_freq = time_tolerance; - else if (time_freq < -time_tolerance) - time_freq = -time_tolerance; + time_freq = min(time_freq, time_tolerance); + time_freq = max(time_freq, -time_tolerance); } /* STA_PLL || STA_PPSTIME */ } /* txc->modes & ADJ_OFFSET */ if (txc->modes & ADJ_TICK) { @@ -384,10 +374,7 @@ leave: if ((time_status & (STA_UNSYNC|STA_CLOCKERR)) != 0 if ((txc->modes & ADJ_OFFSET_SINGLESHOT) == ADJ_OFFSET_SINGLESHOT) txc->offset = save_adjust; else { - if (time_offset < 0) - txc->offset = -(-time_offset >> SHIFT_UPDATE); - else - txc->offset = time_offset >> SHIFT_UPDATE; + txc->offset = shift_right(time_offset, SHIFT_UPDATE); } txc->freq = time_freq + pps_freq; txc->maxerror = time_maxerror; diff --git a/kernel/timer.c b/kernel/timer.c index 6ed1a826e5ce..6b94adb45b03 100644 --- a/kernel/timer.c +++ b/kernel/timer.c @@ -703,23 +703,13 @@ static void second_overflow(void) * the adjustment over not more than the number of * seconds between updates. */ - if (time_offset < 0) { - ltemp = -time_offset; - if (!(time_status & STA_FLL)) - ltemp >>= SHIFT_KG + time_constant; - if (ltemp > (MAXPHASE / MINSEC) << SHIFT_UPDATE) - ltemp = (MAXPHASE / MINSEC) << SHIFT_UPDATE; - time_offset += ltemp; - time_adj = -ltemp << (SHIFT_SCALE - SHIFT_HZ - SHIFT_UPDATE); - } else { ltemp = time_offset; if (!(time_status & STA_FLL)) - ltemp >>= SHIFT_KG + time_constant; - if (ltemp > (MAXPHASE / MINSEC) << SHIFT_UPDATE) - ltemp = (MAXPHASE / MINSEC) << SHIFT_UPDATE; + ltemp = shift_right(ltemp, SHIFT_KG + time_constant); + ltemp = min(ltemp, (MAXPHASE / MINSEC) << SHIFT_UPDATE); + ltemp = max(ltemp, -(MAXPHASE / MINSEC) << SHIFT_UPDATE); time_offset -= ltemp; time_adj = ltemp << (SHIFT_SCALE - SHIFT_HZ - SHIFT_UPDATE); - } /* * Compute the frequency estimate and additional phase @@ -736,39 +726,25 @@ static void second_overflow(void) STA_PPSWANDER | STA_PPSERROR); } ltemp = time_freq + pps_freq; - if (ltemp < 0) - time_adj -= -ltemp >> - (SHIFT_USEC + SHIFT_HZ - SHIFT_SCALE); - else - time_adj += ltemp >> - (SHIFT_USEC + SHIFT_HZ - SHIFT_SCALE); + time_adj += shift_right(ltemp,(SHIFT_USEC + SHIFT_HZ - SHIFT_SCALE)); #if HZ == 100 /* Compensate for (HZ==100) != (1 << SHIFT_HZ). * Add 25% and 3.125% to get 128.125; => only 0.125% error (p. 14) */ - if (time_adj < 0) - time_adj -= (-time_adj >> 2) + (-time_adj >> 5); - else - time_adj += (time_adj >> 2) + (time_adj >> 5); + time_adj += shift_right(time_adj, 2) + shift_right(time_adj, 5); #endif #if HZ == 250 /* Compensate for (HZ==250) != (1 << SHIFT_HZ). * Add 1.5625% and 0.78125% to get 255.85938; => only 0.05% error (p. 14) */ - if (time_adj < 0) - time_adj -= (-time_adj >> 6) + (-time_adj >> 7); - else - time_adj += (time_adj >> 6) + (time_adj >> 7); + time_adj += shift_right(time_adj, 6) + shift_right(time_adj, 7); #endif #if HZ == 1000 /* Compensate for (HZ==1000) != (1 << SHIFT_HZ). * Add 1.5625% and 0.78125% to get 1023.4375; => only 0.05% error (p. 14) */ - if (time_adj < 0) - time_adj -= (-time_adj >> 6) + (-time_adj >> 7); - else - time_adj += (time_adj >> 6) + (time_adj >> 7); + time_adj += shift_right(time_adj, 6) + shift_right(time_adj, 7); #endif } @@ -787,10 +763,8 @@ static void update_wall_time_one_tick(void) * Limit the amount of the step to be in the range * -tickadj .. +tickadj */ - if (time_adjust > tickadj) - time_adjust_step = tickadj; - else if (time_adjust < -tickadj) - time_adjust_step = -tickadj; + time_adjust_step = min(time_adjust_step, (long)tickadj); + time_adjust_step = max(time_adjust_step, (long)-tickadj); /* Reduce by this step the amount of time left */ time_adjust -= time_adjust_step; @@ -801,13 +775,8 @@ static void update_wall_time_one_tick(void) * advance the tick more. */ time_phase += time_adj; - if (time_phase <= -FINENSEC) { - long ltemp = -time_phase >> (SHIFT_SCALE - 10); - time_phase += ltemp << (SHIFT_SCALE - 10); - delta_nsec -= ltemp; - } - else if (time_phase >= FINENSEC) { - long ltemp = time_phase >> (SHIFT_SCALE - 10); + if ((time_phase >= FINENSEC) || (time_phase <= -FINENSEC)) { + long ltemp = shift_right(time_phase, (SHIFT_SCALE - 10)); time_phase -= ltemp << (SHIFT_SCALE - 10); delta_nsec += ltemp; } -- cgit v1.2.3 From dfc4f94d2ff95fc92127d3e512c1df7cab274fb8 Mon Sep 17 00:00:00 2001 From: Andrew Morton Date: Sun, 30 Oct 2005 15:02:03 -0800 Subject: [PATCH] remove timer debug field Remove timer_list.magic and associated debugging code. I originally added this when a spinlock was added to timer_list - this meant that an all-zeroes timer became illegal and init_timer() was required. That spinlock isn't even there any more, although timer.base must now be initialised. I'll keep this debugging code in -mm. Signed-off-by: Alexey Dobriyan Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/s390/kernel/vtime.c | 18 ++---------------- include/linux/timer.h | 5 ----- kernel/timer.c | 35 ----------------------------------- 3 files changed, 2 insertions(+), 56 deletions(-) (limited to 'include/linux') diff --git a/arch/s390/kernel/vtime.c b/arch/s390/kernel/vtime.c index fa0726507b3d..22a895ecb7a4 100644 --- a/arch/s390/kernel/vtime.c +++ b/arch/s390/kernel/vtime.c @@ -24,7 +24,6 @@ #include #include -#define VTIMER_MAGIC (TIMER_MAGIC + 1) static ext_int_info_t ext_int_info_timer; DEFINE_PER_CPU(struct vtimer_queue, virt_cpu_timer); @@ -277,20 +276,12 @@ static void do_cpu_timer_interrupt(struct pt_regs *regs, __u16 error_code) void init_virt_timer(struct vtimer_list *timer) { - timer->magic = VTIMER_MAGIC; timer->function = NULL; INIT_LIST_HEAD(&timer->entry); spin_lock_init(&timer->lock); } EXPORT_SYMBOL(init_virt_timer); -static inline int check_vtimer(struct vtimer_list *timer) -{ - if (timer->magic != VTIMER_MAGIC) - return -EINVAL; - return 0; -} - static inline int vtimer_pending(struct vtimer_list *timer) { return (!list_empty(&timer->entry)); @@ -346,7 +337,7 @@ static void internal_add_vtimer(struct vtimer_list *timer) static inline int prepare_vtimer(struct vtimer_list *timer) { - if (check_vtimer(timer) || !timer->function) { + if (!timer->function) { printk("add_virt_timer: uninitialized timer\n"); return -EINVAL; } @@ -414,7 +405,7 @@ int mod_virt_timer(struct vtimer_list *timer, __u64 expires) unsigned long flags; int cpu; - if (check_vtimer(timer) || !timer->function) { + if (!timer->function) { printk("mod_virt_timer: uninitialized timer\n"); return -EINVAL; } @@ -481,11 +472,6 @@ int del_virt_timer(struct vtimer_list *timer) unsigned long flags; struct vtimer_queue *vt_list; - if (check_vtimer(timer)) { - printk("del_virt_timer: timer not initialized\n"); - return -EINVAL; - } - /* check if timer is pending */ if (!vtimer_pending(timer)) return 0; diff --git a/include/linux/timer.h b/include/linux/timer.h index ddd5bbe1fc8e..b1dc583bb4d4 100644 --- a/include/linux/timer.h +++ b/include/linux/timer.h @@ -12,16 +12,12 @@ struct timer_list { struct list_head entry; unsigned long expires; - unsigned long magic; - void (*function)(unsigned long); unsigned long data; struct timer_base_s *base; }; -#define TIMER_MAGIC 0x4b87ad6e - extern struct timer_base_s __init_timer_base; #define TIMER_INITIALIZER(_function, _expires, _data) { \ @@ -29,7 +25,6 @@ extern struct timer_base_s __init_timer_base; .expires = (_expires), \ .data = (_data), \ .base = &__init_timer_base, \ - .magic = TIMER_MAGIC, \ } #define DEFINE_TIMER(_name, _function, _expires, _data) \ diff --git a/kernel/timer.c b/kernel/timer.c index cc18857601e2..562d53eabb46 100644 --- a/kernel/timer.c +++ b/kernel/timer.c @@ -91,30 +91,6 @@ static inline void set_running_timer(tvec_base_t *base, #endif } -static void check_timer_failed(struct timer_list *timer) -{ - static int whine_count; - if (whine_count < 16) { - whine_count++; - printk("Uninitialised timer!\n"); - printk("This is just a warning. Your computer is OK\n"); - printk("function=0x%p, data=0x%lx\n", - timer->function, timer->data); - dump_stack(); - } - /* - * Now fix it up - */ - timer->magic = TIMER_MAGIC; -} - -static inline void check_timer(struct timer_list *timer) -{ - if (timer->magic != TIMER_MAGIC) - check_timer_failed(timer); -} - - static void internal_add_timer(tvec_base_t *base, struct timer_list *timer) { unsigned long expires = timer->expires; @@ -177,7 +153,6 @@ void fastcall init_timer(struct timer_list *timer) { timer->entry.next = NULL; timer->base = &per_cpu(tvec_bases, raw_smp_processor_id()).t_base; - timer->magic = TIMER_MAGIC; } EXPORT_SYMBOL(init_timer); @@ -230,7 +205,6 @@ int __mod_timer(struct timer_list *timer, unsigned long expires) int ret = 0; BUG_ON(!timer->function); - check_timer(timer); base = lock_timer_base(timer, &flags); @@ -283,9 +257,6 @@ void add_timer_on(struct timer_list *timer, int cpu) unsigned long flags; BUG_ON(timer_pending(timer) || !timer->function); - - check_timer(timer); - spin_lock_irqsave(&base->t_base.lock, flags); timer->base = &base->t_base; internal_add_timer(base, timer); @@ -316,8 +287,6 @@ int mod_timer(struct timer_list *timer, unsigned long expires) { BUG_ON(!timer->function); - check_timer(timer); - /* * This is a common optimization triggered by the * networking code - if the timer is re-modified @@ -348,8 +317,6 @@ int del_timer(struct timer_list *timer) unsigned long flags; int ret = 0; - check_timer(timer); - if (timer_pending(timer)) { base = lock_timer_base(timer, &flags); if (timer_pending(timer)) { @@ -412,8 +379,6 @@ out: */ int del_timer_sync(struct timer_list *timer) { - check_timer(timer); - for (;;) { int ret = try_to_del_timer_sync(timer); if (ret >= 0) -- cgit v1.2.3 From 2f51201662b28dbf8c15fb7eb972bc51c6cc3fa5 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Sun, 30 Oct 2005 15:02:16 -0800 Subject: [PATCH] reduce sizeof(struct file) Now that RCU applied on 'struct file' seems stable, we can place f_rcuhead in a memory location that is not anymore used at call_rcu(&f->f_rcuhead, file_free_rcu) time, to reduce the size of this critical kernel object. The trick I used is to move f_rcuhead and f_list in an union called f_u The callers are changed so that f_rcuhead becomes f_u.fu_rcuhead and f_list becomes f_u.f_list Signed-off-by: Eric Dumazet Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/char/tty_io.c | 2 +- fs/dquot.c | 2 +- fs/file_table.c | 14 +++++++------- fs/proc/generic.c | 2 +- fs/super.c | 2 +- include/linux/fs.h | 10 ++++++++-- security/selinux/hooks.c | 2 +- security/selinux/selinuxfs.c | 2 +- 8 files changed, 21 insertions(+), 15 deletions(-) (limited to 'include/linux') diff --git a/drivers/char/tty_io.c b/drivers/char/tty_io.c index f5649a337743..c586bfa852ee 100644 --- a/drivers/char/tty_io.c +++ b/drivers/char/tty_io.c @@ -809,7 +809,7 @@ static void do_tty_hangup(void *data) check_tty_count(tty, "do_tty_hangup"); file_list_lock(); /* This breaks for file handles being sent over AF_UNIX sockets ? */ - list_for_each_entry(filp, &tty->tty_files, f_list) { + list_for_each_entry(filp, &tty->tty_files, f_u.fu_list) { if (filp->f_op->write == redirected_tty_write) cons_filp = filp; if (filp->f_op->write != tty_write) diff --git a/fs/dquot.c b/fs/dquot.c index 05f3327d64a3..ea7644227a65 100644 --- a/fs/dquot.c +++ b/fs/dquot.c @@ -662,7 +662,7 @@ static void add_dquot_ref(struct super_block *sb, int type) restart: file_list_lock(); list_for_each(p, &sb->s_files) { - struct file *filp = list_entry(p, struct file, f_list); + struct file *filp = list_entry(p, struct file, f_u.fu_list); struct inode *inode = filp->f_dentry->d_inode; if (filp->f_mode & FMODE_WRITE && dqinit_needed(inode, type)) { struct dentry *dentry = dget(filp->f_dentry); diff --git a/fs/file_table.c b/fs/file_table.c index 86ec8ae985b4..4dc205546547 100644 --- a/fs/file_table.c +++ b/fs/file_table.c @@ -56,13 +56,13 @@ void filp_dtor(void * objp, struct kmem_cache_s *cachep, unsigned long dflags) static inline void file_free_rcu(struct rcu_head *head) { - struct file *f = container_of(head, struct file, f_rcuhead); + struct file *f = container_of(head, struct file, f_u.fu_rcuhead); kmem_cache_free(filp_cachep, f); } static inline void file_free(struct file *f) { - call_rcu(&f->f_rcuhead, file_free_rcu); + call_rcu(&f->f_u.fu_rcuhead, file_free_rcu); } /* Find an unused file structure and return a pointer to it. @@ -95,7 +95,7 @@ struct file *get_empty_filp(void) f->f_gid = current->fsgid; rwlock_init(&f->f_owner.lock); /* f->f_version: 0 */ - INIT_LIST_HEAD(&f->f_list); + INIT_LIST_HEAD(&f->f_u.fu_list); return f; over: @@ -225,15 +225,15 @@ void file_move(struct file *file, struct list_head *list) if (!list) return; file_list_lock(); - list_move(&file->f_list, list); + list_move(&file->f_u.fu_list, list); file_list_unlock(); } void file_kill(struct file *file) { - if (!list_empty(&file->f_list)) { + if (!list_empty(&file->f_u.fu_list)) { file_list_lock(); - list_del_init(&file->f_list); + list_del_init(&file->f_u.fu_list); file_list_unlock(); } } @@ -245,7 +245,7 @@ int fs_may_remount_ro(struct super_block *sb) /* Check that no files are currently opened for writing. */ file_list_lock(); list_for_each(p, &sb->s_files) { - struct file *file = list_entry(p, struct file, f_list); + struct file *file = list_entry(p, struct file, f_u.fu_list); struct inode *inode = file->f_dentry->d_inode; /* File with pending delete? */ diff --git a/fs/proc/generic.c b/fs/proc/generic.c index 8a8c34461d48..b638fb500743 100644 --- a/fs/proc/generic.c +++ b/fs/proc/generic.c @@ -533,7 +533,7 @@ static void proc_kill_inodes(struct proc_dir_entry *de) */ file_list_lock(); list_for_each(p, &sb->s_files) { - struct file * filp = list_entry(p, struct file, f_list); + struct file * filp = list_entry(p, struct file, f_u.fu_list); struct dentry * dentry = filp->f_dentry; struct inode * inode; struct file_operations *fops; diff --git a/fs/super.c b/fs/super.c index 6e57ee252e14..f60155ec7780 100644 --- a/fs/super.c +++ b/fs/super.c @@ -513,7 +513,7 @@ static void mark_files_ro(struct super_block *sb) struct file *f; file_list_lock(); - list_for_each_entry(f, &sb->s_files, f_list) { + list_for_each_entry(f, &sb->s_files, f_u.fu_list) { if (S_ISREG(f->f_dentry->d_inode->i_mode) && file_count(f)) f->f_mode &= ~FMODE_WRITE; } diff --git a/include/linux/fs.h b/include/linux/fs.h index f83d997c5582..6d6226732c93 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -574,7 +574,14 @@ struct file_ra_state { #define RA_FLAG_INCACHE 0x02 /* file is already in cache */ struct file { - struct list_head f_list; + /* + * fu_list becomes invalid after file_free is called and queued via + * fu_rcuhead for RCU freeing + */ + union { + struct list_head fu_list; + struct rcu_head fu_rcuhead; + } f_u; struct dentry *f_dentry; struct vfsmount *f_vfsmnt; struct file_operations *f_op; @@ -598,7 +605,6 @@ struct file { spinlock_t f_ep_lock; #endif /* #ifdef CONFIG_EPOLL */ struct address_space *f_mapping; - struct rcu_head f_rcuhead; }; extern spinlock_t files_lock; #define file_list_lock() spin_lock(&files_lock); diff --git a/security/selinux/hooks.c b/security/selinux/hooks.c index d50d7a782cf5..bb62838be496 100644 --- a/security/selinux/hooks.c +++ b/security/selinux/hooks.c @@ -1609,7 +1609,7 @@ static inline void flush_unauthorized_files(struct files_struct * files) if (tty) { file_list_lock(); - file = list_entry(tty->tty_files.next, typeof(*file), f_list); + file = list_entry(tty->tty_files.next, typeof(*file), f_u.fu_list); if (file) { /* Revalidate access to controlling tty. Use inode_has_perm on the tty inode directly rather diff --git a/security/selinux/selinuxfs.c b/security/selinux/selinuxfs.c index 7a387fec2a0c..fdc382389720 100644 --- a/security/selinux/selinuxfs.c +++ b/security/selinux/selinuxfs.c @@ -914,7 +914,7 @@ static void sel_remove_bools(struct dentry *de) file_list_lock(); list_for_each(p, &sb->s_files) { - struct file * filp = list_entry(p, struct file, f_list); + struct file * filp = list_entry(p, struct file, f_u.fu_list); struct dentry * dentry = filp->f_dentry; if (dentry->d_parent != de) { -- cgit v1.2.3 From 19a4fcb531659f2f7d18b5d04cee039176e9540d Mon Sep 17 00:00:00 2001 From: Oleg Nesterov Date: Sun, 30 Oct 2005 15:02:17 -0800 Subject: [PATCH] kill sigqueue->lock This lock is used in sigqueue_free(), but it is always equal to current->sighand->siglock, so we don't need to keep it in the struct sigqueue. Signed-off-by: Oleg Nesterov Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/signal.h | 1 - kernel/signal.c | 10 ++++------ 2 files changed, 4 insertions(+), 7 deletions(-) (limited to 'include/linux') diff --git a/include/linux/signal.h b/include/linux/signal.h index 7be18b5e2fb4..5dd5f02c5c5f 100644 --- a/include/linux/signal.h +++ b/include/linux/signal.h @@ -25,7 +25,6 @@ struct sigqueue { struct list_head list; - spinlock_t *lock; int flags; siginfo_t info; struct user_struct *user; diff --git a/kernel/signal.c b/kernel/signal.c index 6904bbbfe116..8d6e64dfa5c6 100644 --- a/kernel/signal.c +++ b/kernel/signal.c @@ -277,7 +277,6 @@ static struct sigqueue *__sigqueue_alloc(struct task_struct *t, gfp_t flags, } else { INIT_LIST_HEAD(&q->list); q->flags = 0; - q->lock = NULL; q->user = get_uid(t->user); } return(q); @@ -1371,11 +1370,12 @@ void sigqueue_free(struct sigqueue *q) * pending queue. */ if (unlikely(!list_empty(&q->list))) { - read_lock(&tasklist_lock); - spin_lock_irqsave(q->lock, flags); + spinlock_t *lock = ¤t->sighand->siglock; + read_lock(&tasklist_lock); + spin_lock_irqsave(lock, flags); if (!list_empty(&q->list)) list_del_init(&q->list); - spin_unlock_irqrestore(q->lock, flags); + spin_unlock_irqrestore(lock, flags); read_unlock(&tasklist_lock); } q->flags &= ~SIGQUEUE_PREALLOC; @@ -1414,7 +1414,6 @@ send_sigqueue(int sig, struct sigqueue *q, struct task_struct *p) goto out; } - q->lock = &p->sighand->siglock; list_add_tail(&q->list, &p->pending.list); sigaddset(&p->pending.signal, sig); if (!sigismember(&p->blocked, sig)) @@ -1462,7 +1461,6 @@ send_group_sigqueue(int sig, struct sigqueue *q, struct task_struct *p) * We always use the shared queue for process-wide signals, * to avoid several races. */ - q->lock = &p->sighand->siglock; list_add_tail(&q->list, &p->signal->shared_pending.list); sigaddset(&p->signal->shared_pending.signal, sig); -- cgit v1.2.3 From dfb7dac3af623a68262536437af008ed6aba4d88 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Sun, 30 Oct 2005 15:02:22 -0800 Subject: [PATCH] unify sys_ptrace prototype Make sure we always return, as all syscalls should. Also move the common prototype to Signed-off-by: Christoph Hellwig Signed-off-by: Miklos Szeredi Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/arm/kernel/ptrace.c | 2 +- arch/arm26/kernel/ptrace.c | 2 +- arch/frv/kernel/ptrace.c | 2 +- arch/h8300/kernel/ptrace.c | 2 +- arch/i386/kernel/ptrace.c | 2 +- arch/ia64/ia32/sys_ia32.c | 1 + arch/m32r/kernel/ptrace.c | 2 +- arch/m68k/kernel/ptrace.c | 2 +- arch/m68knommu/kernel/ptrace.c | 2 +- arch/mips/kernel/ptrace.c | 2 +- arch/ppc/kernel/ptrace.c | 2 +- arch/ppc64/kernel/ptrace.c | 2 +- arch/sh/kernel/ptrace.c | 2 +- arch/sh64/kernel/ptrace.c | 2 +- arch/um/include/sysdep-i386/syscalls.h | 1 - arch/v850/kernel/ptrace.c | 2 +- arch/xtensa/kernel/ptrace.c | 2 +- include/asm-arm/unistd.h | 1 - include/asm-arm26/unistd.h | 1 - include/asm-cris/unistd.h | 1 - include/asm-h8300/unistd.h | 1 - include/asm-i386/unistd.h | 1 - include/asm-ia64/unistd.h | 2 -- include/asm-m32r/unistd.h | 1 - include/asm-m68k/unistd.h | 1 - include/asm-m68knommu/unistd.h | 1 - include/asm-mips/unistd.h | 1 - include/asm-ppc/unistd.h | 1 - include/asm-ppc64/unistd.h | 1 - include/asm-s390/unistd.h | 1 - include/asm-sh/unistd.h | 1 - include/asm-v850/unistd.h | 1 - include/asm-x86_64/unistd.h | 2 -- include/linux/syscalls.h | 1 + 34 files changed, 17 insertions(+), 34 deletions(-) (limited to 'include/linux') diff --git a/arch/arm/kernel/ptrace.c b/arch/arm/kernel/ptrace.c index cd99b83f14c2..9bd8609a2926 100644 --- a/arch/arm/kernel/ptrace.c +++ b/arch/arm/kernel/ptrace.c @@ -782,7 +782,7 @@ static int do_ptrace(int request, struct task_struct *child, long addr, long dat return ret; } -asmlinkage int sys_ptrace(long request, long pid, long addr, long data) +asmlinkage long sys_ptrace(long request, long pid, long addr, long data) { struct task_struct *child; int ret; diff --git a/arch/arm26/kernel/ptrace.c b/arch/arm26/kernel/ptrace.c index 8a52124de0e1..cf7e977d18c8 100644 --- a/arch/arm26/kernel/ptrace.c +++ b/arch/arm26/kernel/ptrace.c @@ -665,7 +665,7 @@ static int do_ptrace(int request, struct task_struct *child, long addr, long dat return ret; } -asmlinkage int sys_ptrace(long request, long pid, long addr, long data) +asmlinkage long sys_ptrace(long request, long pid, long addr, long data) { struct task_struct *child; int ret; diff --git a/arch/frv/kernel/ptrace.c b/arch/frv/kernel/ptrace.c index cbe03cba9f02..cb335a14a315 100644 --- a/arch/frv/kernel/ptrace.c +++ b/arch/frv/kernel/ptrace.c @@ -106,7 +106,7 @@ void ptrace_enable(struct task_struct *child) child->thread.frame0->__status |= REG__STATUS_STEP; } -asmlinkage int sys_ptrace(long request, long pid, long addr, long data) +asmlinkage long sys_ptrace(long request, long pid, long addr, long data) { struct task_struct *child; unsigned long tmp; diff --git a/arch/h8300/kernel/ptrace.c b/arch/h8300/kernel/ptrace.c index 05c15e869777..a569fe4aa284 100644 --- a/arch/h8300/kernel/ptrace.c +++ b/arch/h8300/kernel/ptrace.c @@ -57,7 +57,7 @@ void ptrace_disable(struct task_struct *child) h8300_disable_trace(child); } -asmlinkage int sys_ptrace(long request, long pid, long addr, long data) +asmlinkage long sys_ptrace(long request, long pid, long addr, long data) { struct task_struct *child; int ret; diff --git a/arch/i386/kernel/ptrace.c b/arch/i386/kernel/ptrace.c index 7b6368bf8974..efd11f09c996 100644 --- a/arch/i386/kernel/ptrace.c +++ b/arch/i386/kernel/ptrace.c @@ -354,7 +354,7 @@ ptrace_set_thread_area(struct task_struct *child, return 0; } -asmlinkage int sys_ptrace(long request, long pid, long addr, long data) +asmlinkage long sys_ptrace(long request, long pid, long addr, long data) { struct task_struct *child; struct user * dummy = NULL; diff --git a/arch/ia64/ia32/sys_ia32.c b/arch/ia64/ia32/sys_ia32.c index 3fa67ecebc83..dc282710421a 100644 --- a/arch/ia64/ia32/sys_ia32.c +++ b/arch/ia64/ia32/sys_ia32.c @@ -36,6 +36,7 @@ #include #include #include +#include #include #include #include diff --git a/arch/m32r/kernel/ptrace.c b/arch/m32r/kernel/ptrace.c index 124f7c1b775e..078d2a0e71c2 100644 --- a/arch/m32r/kernel/ptrace.c +++ b/arch/m32r/kernel/ptrace.c @@ -756,7 +756,7 @@ do_ptrace(long request, struct task_struct *child, long addr, long data) return ret; } -asmlinkage int sys_ptrace(long request, long pid, long addr, long data) +asmlinkage long sys_ptrace(long request, long pid, long addr, long data) { struct task_struct *child; int ret; diff --git a/arch/m68k/kernel/ptrace.c b/arch/m68k/kernel/ptrace.c index 8ed1b01a6a87..f7f1d2e5b90b 100644 --- a/arch/m68k/kernel/ptrace.c +++ b/arch/m68k/kernel/ptrace.c @@ -121,7 +121,7 @@ void ptrace_disable(struct task_struct *child) child->thread.work.syscall_trace = 0; } -asmlinkage int sys_ptrace(long request, long pid, long addr, long data) +asmlinkage long sys_ptrace(long request, long pid, long addr, long data) { struct task_struct *child; unsigned long tmp; diff --git a/arch/m68knommu/kernel/ptrace.c b/arch/m68knommu/kernel/ptrace.c index 9724e1cd82e5..621d7b91ccfe 100644 --- a/arch/m68knommu/kernel/ptrace.c +++ b/arch/m68knommu/kernel/ptrace.c @@ -101,7 +101,7 @@ void ptrace_disable(struct task_struct *child) put_reg(child, PT_SR, tmp); } -asmlinkage int sys_ptrace(long request, long pid, long addr, long data) +asmlinkage long sys_ptrace(long request, long pid, long addr, long data) { struct task_struct *child; int ret; diff --git a/arch/mips/kernel/ptrace.c b/arch/mips/kernel/ptrace.c index fcceab8f2e00..f1b0f3e1f95b 100644 --- a/arch/mips/kernel/ptrace.c +++ b/arch/mips/kernel/ptrace.c @@ -174,7 +174,7 @@ int ptrace_setfpregs (struct task_struct *child, __u32 __user *data) return 0; } -asmlinkage int sys_ptrace(long request, long pid, long addr, long data) +asmlinkage long sys_ptrace(long request, long pid, long addr, long data) { struct task_struct *child; int ret; diff --git a/arch/ppc/kernel/ptrace.c b/arch/ppc/kernel/ptrace.c index e7aee4108dea..e2744b6879da 100644 --- a/arch/ppc/kernel/ptrace.c +++ b/arch/ppc/kernel/ptrace.c @@ -240,7 +240,7 @@ void ptrace_disable(struct task_struct *child) clear_single_step(child); } -int sys_ptrace(long request, long pid, long addr, long data) +long sys_ptrace(long request, long pid, long addr, long data) { struct task_struct *child; int ret = -EPERM; diff --git a/arch/ppc64/kernel/ptrace.c b/arch/ppc64/kernel/ptrace.c index b1c044ca5756..b33073c31724 100644 --- a/arch/ppc64/kernel/ptrace.c +++ b/arch/ppc64/kernel/ptrace.c @@ -53,7 +53,7 @@ void ptrace_disable(struct task_struct *child) clear_single_step(child); } -int sys_ptrace(long request, long pid, long addr, long data) +long sys_ptrace(long request, long pid, long addr, long data) { struct task_struct *child; int ret = -EPERM; diff --git a/arch/sh/kernel/ptrace.c b/arch/sh/kernel/ptrace.c index b28919b65682..1fbe5a428e31 100644 --- a/arch/sh/kernel/ptrace.c +++ b/arch/sh/kernel/ptrace.c @@ -80,7 +80,7 @@ void ptrace_disable(struct task_struct *child) /* nothing to do.. */ } -asmlinkage int sys_ptrace(long request, long pid, long addr, long data) +asmlinkage long sys_ptrace(long request, long pid, long addr, long data) { struct task_struct *child; struct user * dummy = NULL; diff --git a/arch/sh64/kernel/ptrace.c b/arch/sh64/kernel/ptrace.c index fd2000956dae..71f2eec00b99 100644 --- a/arch/sh64/kernel/ptrace.c +++ b/arch/sh64/kernel/ptrace.c @@ -121,7 +121,7 @@ put_fpu_long(struct task_struct *task, unsigned long addr, unsigned long data) return 0; } -asmlinkage int sys_ptrace(long request, long pid, long addr, long data) +asmlinkage long sys_ptrace(long request, long pid, long addr, long data) { struct task_struct *child; extern void poke_real_address_q(unsigned long long addr, unsigned long long data); diff --git a/arch/um/include/sysdep-i386/syscalls.h b/arch/um/include/sysdep-i386/syscalls.h index a0d5b74d3731..57bd79efbee3 100644 --- a/arch/um/include/sysdep-i386/syscalls.h +++ b/arch/um/include/sysdep-i386/syscalls.h @@ -11,7 +11,6 @@ typedef long syscall_handler_t(struct pt_regs); /* Not declared on x86, incompatible declarations on x86_64, so these have * to go here rather than in sys_call_table.c */ -extern syscall_handler_t sys_ptrace; extern syscall_handler_t sys_rt_sigaction; extern syscall_handler_t old_mmap_i386; diff --git a/arch/v850/kernel/ptrace.c b/arch/v850/kernel/ptrace.c index 4726b87f5e5a..d6077ff47d22 100644 --- a/arch/v850/kernel/ptrace.c +++ b/arch/v850/kernel/ptrace.c @@ -113,7 +113,7 @@ static int set_single_step (struct task_struct *t, int val) return 1; } -int sys_ptrace(long request, long pid, long addr, long data) +long sys_ptrace(long request, long pid, long addr, long data) { struct task_struct *child; int rval; diff --git a/arch/xtensa/kernel/ptrace.c b/arch/xtensa/kernel/ptrace.c index 2659efdd4e99..14460743de07 100644 --- a/arch/xtensa/kernel/ptrace.c +++ b/arch/xtensa/kernel/ptrace.c @@ -45,7 +45,7 @@ void ptrace_disable(struct task_struct *child) /* Nothing to do.. */ } -int sys_ptrace(long request, long pid, long addr, long data) +long sys_ptrace(long request, long pid, long addr, long data) { struct task_struct *child; int ret = -EPERM; diff --git a/include/asm-arm/unistd.h b/include/asm-arm/unistd.h index c49df635a80f..d626e70faded 100644 --- a/include/asm-arm/unistd.h +++ b/include/asm-arm/unistd.h @@ -544,7 +544,6 @@ asmlinkage int sys_clone(unsigned long clone_flags, unsigned long newsp, asmlinkage int sys_fork(struct pt_regs *regs); asmlinkage int sys_vfork(struct pt_regs *regs); asmlinkage int sys_pipe(unsigned long *fildes); -asmlinkage int sys_ptrace(long request, long pid, long addr, long data); struct sigaction; asmlinkage long sys_rt_sigaction(int sig, const struct sigaction __user *act, diff --git a/include/asm-arm26/unistd.h b/include/asm-arm26/unistd.h index dfa0b0c30aa3..be4c2fb9c049 100644 --- a/include/asm-arm26/unistd.h +++ b/include/asm-arm26/unistd.h @@ -480,7 +480,6 @@ asmlinkage int sys_clone(unsigned long clone_flags, unsigned long newsp, asmlinkage int sys_fork(struct pt_regs *regs); asmlinkage int sys_vfork(struct pt_regs *regs); asmlinkage int sys_pipe(unsigned long *fildes); -asmlinkage int sys_ptrace(long request, long pid, long addr, long data); struct sigaction; asmlinkage long sys_rt_sigaction(int sig, const struct sigaction __user *act, diff --git a/include/asm-cris/unistd.h b/include/asm-cris/unistd.h index 28232ad2ff34..156a34bfc583 100644 --- a/include/asm-cris/unistd.h +++ b/include/asm-cris/unistd.h @@ -367,7 +367,6 @@ asmlinkage int sys_fork(long r10, long r11, long r12, long r13, asmlinkage int sys_vfork(long r10, long r11, long r12, long r13, long mof, long srp, struct pt_regs *regs); asmlinkage int sys_pipe(unsigned long __user *fildes); -asmlinkage int sys_ptrace(long request, long pid, long addr, long data); struct sigaction; asmlinkage long sys_rt_sigaction(int sig, const struct sigaction __user *act, diff --git a/include/asm-h8300/unistd.h b/include/asm-h8300/unistd.h index 56a6401886fa..56a4a5686c88 100644 --- a/include/asm-h8300/unistd.h +++ b/include/asm-h8300/unistd.h @@ -528,7 +528,6 @@ asmlinkage long sys_mmap2(unsigned long addr, unsigned long len, asmlinkage int sys_execve(char *name, char **argv, char **envp, int dummy, ...); asmlinkage int sys_pipe(unsigned long *fildes); -asmlinkage int sys_ptrace(long request, long pid, long addr, long data); struct sigaction; asmlinkage long sys_rt_sigaction(int sig, const struct sigaction __user *act, diff --git a/include/asm-i386/unistd.h b/include/asm-i386/unistd.h index fbaf90a3968c..0f92e78dfea1 100644 --- a/include/asm-i386/unistd.h +++ b/include/asm-i386/unistd.h @@ -448,7 +448,6 @@ asmlinkage int sys_clone(struct pt_regs regs); asmlinkage int sys_fork(struct pt_regs regs); asmlinkage int sys_vfork(struct pt_regs regs); asmlinkage int sys_pipe(unsigned long __user *fildes); -asmlinkage int sys_ptrace(long request, long pid, long addr, long data); asmlinkage long sys_iopl(unsigned long unused); struct sigaction; asmlinkage long sys_rt_sigaction(int sig, diff --git a/include/asm-ia64/unistd.h b/include/asm-ia64/unistd.h index 3a0c69524656..6d96a67439be 100644 --- a/include/asm-ia64/unistd.h +++ b/include/asm-ia64/unistd.h @@ -383,8 +383,6 @@ struct sigaction; long sys_execve(char __user *filename, char __user * __user *argv, char __user * __user *envp, struct pt_regs *regs); asmlinkage long sys_pipe(void); -asmlinkage long sys_ptrace(long request, pid_t pid, - unsigned long addr, unsigned long data); asmlinkage long sys_rt_sigaction(int sig, const struct sigaction __user *act, struct sigaction __user *oact, diff --git a/include/asm-m32r/unistd.h b/include/asm-m32r/unistd.h index 8552d8f45ab1..ac399e1f7bc0 100644 --- a/include/asm-m32r/unistd.h +++ b/include/asm-m32r/unistd.h @@ -452,7 +452,6 @@ asmlinkage int sys_clone(struct pt_regs regs); asmlinkage int sys_fork(struct pt_regs regs); asmlinkage int sys_vfork(struct pt_regs regs); asmlinkage int sys_pipe(unsigned long __user *fildes); -asmlinkage int sys_ptrace(long request, long pid, long addr, long data); struct sigaction; asmlinkage long sys_rt_sigaction(int sig, const struct sigaction __user *act, diff --git a/include/asm-m68k/unistd.h b/include/asm-m68k/unistd.h index cbabde4f8a45..c2554bcd1747 100644 --- a/include/asm-m68k/unistd.h +++ b/include/asm-m68k/unistd.h @@ -444,7 +444,6 @@ asmlinkage long sys_mmap2( unsigned long fd, unsigned long pgoff); asmlinkage int sys_execve(char *name, char **argv, char **envp); asmlinkage int sys_pipe(unsigned long *fildes); -asmlinkage int sys_ptrace(long request, long pid, long addr, long data); struct pt_regs; struct sigaction; asmlinkage long sys_rt_sigaction(int sig, diff --git a/include/asm-m68knommu/unistd.h b/include/asm-m68knommu/unistd.h index 84b6fa14459f..5373988a7e51 100644 --- a/include/asm-m68knommu/unistd.h +++ b/include/asm-m68knommu/unistd.h @@ -504,7 +504,6 @@ asmlinkage long sys_mmap2(unsigned long addr, unsigned long len, unsigned long fd, unsigned long pgoff); asmlinkage int sys_execve(char *name, char **argv, char **envp); asmlinkage int sys_pipe(unsigned long *fildes); -asmlinkage int sys_ptrace(long request, long pid, long addr, long data); struct pt_regs; int sys_request_irq(unsigned int, irqreturn_t (*)(int, void *, struct pt_regs *), diff --git a/include/asm-mips/unistd.h b/include/asm-mips/unistd.h index c9eaf4c104de..89ea8b60e945 100644 --- a/include/asm-mips/unistd.h +++ b/include/asm-mips/unistd.h @@ -1177,7 +1177,6 @@ asmlinkage long sys_mmap2( unsigned long fd, unsigned long pgoff); asmlinkage int sys_execve(nabi_no_regargs struct pt_regs regs); asmlinkage int sys_pipe(nabi_no_regargs struct pt_regs regs); -asmlinkage int sys_ptrace(long request, long pid, long addr, long data); struct sigaction; asmlinkage long sys_rt_sigaction(int sig, const struct sigaction __user *act, diff --git a/include/asm-ppc/unistd.h b/include/asm-ppc/unistd.h index 3173ab3d2eb9..404c143e643d 100644 --- a/include/asm-ppc/unistd.h +++ b/include/asm-ppc/unistd.h @@ -469,7 +469,6 @@ int sys_fork(int p1, int p2, int p3, int p4, int p5, int p6, int sys_vfork(int p1, int p2, int p3, int p4, int p5, int p6, struct pt_regs *regs); int sys_pipe(int __user *fildes); -int sys_ptrace(long request, long pid, long addr, long data); struct sigaction; long sys_rt_sigaction(int sig, const struct sigaction __user *act, diff --git a/include/asm-ppc64/unistd.h b/include/asm-ppc64/unistd.h index 977bc980c1af..6b5fcbadbc5b 100644 --- a/include/asm-ppc64/unistd.h +++ b/include/asm-ppc64/unistd.h @@ -467,7 +467,6 @@ int sys_vfork(unsigned long p1, unsigned long p2, unsigned long p3, unsigned long p4, unsigned long p5, unsigned long p6, struct pt_regs *regs); int sys_pipe(int __user *fildes); -int sys_ptrace(long request, long pid, long addr, long data); struct sigaction; long sys_rt_sigaction(int sig, const struct sigaction __user *act, struct sigaction __user *oact, size_t sigsetsize); diff --git a/include/asm-s390/unistd.h b/include/asm-s390/unistd.h index 221e965da924..f97d92691f17 100644 --- a/include/asm-s390/unistd.h +++ b/include/asm-s390/unistd.h @@ -590,7 +590,6 @@ asmlinkage long sys_clone(struct pt_regs regs); asmlinkage long sys_fork(struct pt_regs regs); asmlinkage long sys_vfork(struct pt_regs regs); asmlinkage long sys_pipe(unsigned long __user *fildes); -asmlinkage long sys_ptrace(long request, long pid, long addr, long data); struct sigaction; asmlinkage long sys_rt_sigaction(int sig, const struct sigaction __user *act, diff --git a/include/asm-sh/unistd.h b/include/asm-sh/unistd.h index ea89e8f223ea..f2c8e14d1fd9 100644 --- a/include/asm-sh/unistd.h +++ b/include/asm-sh/unistd.h @@ -503,7 +503,6 @@ asmlinkage int sys_vfork(unsigned long r4, unsigned long r5, asmlinkage int sys_pipe(unsigned long r4, unsigned long r5, unsigned long r6, unsigned long r7, struct pt_regs regs); -asmlinkage int sys_ptrace(long request, long pid, long addr, long data); asmlinkage ssize_t sys_pread_wrapper(unsigned int fd, char *buf, size_t count, long dummy, loff_t pos); asmlinkage ssize_t sys_pwrite_wrapper(unsigned int fd, const char *buf, diff --git a/include/asm-v850/unistd.h b/include/asm-v850/unistd.h index 3b552096c0e8..5a86f8e976ec 100644 --- a/include/asm-v850/unistd.h +++ b/include/asm-v850/unistd.h @@ -452,7 +452,6 @@ unsigned long sys_mmap2(unsigned long addr, size_t len, struct pt_regs; int sys_execve (char *name, char **argv, char **envp, struct pt_regs *regs); int sys_pipe (int *fildes); -int sys_ptrace(long request, long pid, long addr, long data); struct sigaction; asmlinkage long sys_rt_sigaction(int sig, const struct sigaction __user *act, diff --git a/include/asm-x86_64/unistd.h b/include/asm-x86_64/unistd.h index 11ba931cf82f..3c494b65d33a 100644 --- a/include/asm-x86_64/unistd.h +++ b/include/asm-x86_64/unistd.h @@ -780,8 +780,6 @@ asmlinkage long sys_pipe(int *fildes); #include #include -asmlinkage long sys_ptrace(long request, long pid, - unsigned long addr, long data); asmlinkage long sys_iopl(unsigned int level, struct pt_regs *regs); asmlinkage long sys_ioperm(unsigned long from, unsigned long num, int turn_on); struct sigaction; diff --git a/include/linux/syscalls.h b/include/linux/syscalls.h index a6f03e473737..c7007b1db91d 100644 --- a/include/linux/syscalls.h +++ b/include/linux/syscalls.h @@ -491,6 +491,7 @@ asmlinkage long sys_nfsservctl(int cmd, asmlinkage long sys_syslog(int type, char __user *buf, int len); asmlinkage long sys_uselib(const char __user *library); asmlinkage long sys_ni_syscall(void); +asmlinkage long sys_ptrace(long request, long pid, long addr, long data); asmlinkage long sys_add_key(const char __user *_type, const char __user *_description, -- cgit v1.2.3 From 15d2bace5ec907530a3d0e0cf4bb1bd29f3ad7b7 Mon Sep 17 00:00:00 2001 From: Andrew Morton Date: Sun, 30 Oct 2005 15:02:24 -0800 Subject: [PATCH] add_timer() of a pending timer is illegal In the recent timer rework we lost the check for an add_timer() of an already-pending timer. That check was useful for networking, so put it back. Cc: "David S. Miller" Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/timer.h | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'include/linux') diff --git a/include/linux/timer.h b/include/linux/timer.h index b1dc583bb4d4..72f3a7781106 100644 --- a/include/linux/timer.h +++ b/include/linux/timer.h @@ -78,8 +78,9 @@ extern unsigned long next_timer_interrupt(void); * Timers with an ->expired field in the past will be executed in the next * timer tick. */ -static inline void add_timer(struct timer_list * timer) +static inline void add_timer(struct timer_list *timer) { + BUG_ON(timer_pending(timer)); __mod_timer(timer, timer->expires); } -- cgit v1.2.3 From 053199edf54f685e7dea765b60d4d5e9070dadec Mon Sep 17 00:00:00 2001 From: Paul Jackson Date: Sun, 30 Oct 2005 15:02:30 -0800 Subject: [PATCH] cpusets: dual semaphore locking overhaul Overhaul cpuset locking. Replace single semaphore with two semaphores. The suggestion to use two locks was made by Roman Zippel. Both locks are global. Code that wants to modify cpusets must first acquire the exclusive manage_sem, which allows them read-only access to cpusets, and holds off other would-be modifiers. Before making actual changes, the second semaphore, callback_sem must be acquired as well. Code that needs only to query cpusets must acquire callback_sem, which is also a global exclusive lock. The earlier problems with double tripping are avoided, because it is allowed for holders of manage_sem to nest the second callback_sem lock, and only callback_sem is needed by code called from within __alloc_pages(), where the double tripping had been possible. This is not quite the same as a normal read/write semaphore, because obtaining read-only access with intent to change must hold off other such attempts, while allowing read-only access w/o such intention. Changing cpusets involves several related checks and changes, which must be done while allowing read-only queries (to avoid the double trip), but while ensuring nothing changes (holding off other would be modifiers.) This overhaul of cpuset locking also makes careful use of task_lock() to guard access to the task->cpuset pointer, closing a couple of race conditions noticed while reading this code (thanks, Roman). I've never seen these races fail in any use or test. See further the comments in the code. Signed-off-by: Paul Jackson Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/sched.h | 2 +- kernel/cpuset.c | 418 +++++++++++++++++++++++++++++++++----------------- 2 files changed, 282 insertions(+), 138 deletions(-) (limited to 'include/linux') diff --git a/include/linux/sched.h b/include/linux/sched.h index 1c30bc308ef1..b2d2dc14f0b9 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -1211,7 +1211,7 @@ extern void unhash_process(struct task_struct *p); /* * Protects ->fs, ->files, ->mm, ->ptrace, ->group_info, ->comm, keyring * subscriptions and synchronises with wait4(). Also used in procfs. Also - * pins the final release of task.io_context. + * pins the final release of task.io_context. Also protects ->cpuset. * * Nests both inside and outside of read_lock(&tasklist_lock). * It must not be nested with write_lock_irq(&tasklist_lock), diff --git a/kernel/cpuset.c b/kernel/cpuset.c index cd54dba2be18..7491352276b2 100644 --- a/kernel/cpuset.c +++ b/kernel/cpuset.c @@ -60,6 +60,9 @@ struct cpuset { cpumask_t cpus_allowed; /* CPUs allowed to tasks in cpuset */ nodemask_t mems_allowed; /* Memory Nodes allowed to tasks */ + /* + * Count is atomic so can incr (fork) or decr (exit) without a lock. + */ atomic_t count; /* count tasks using this cpuset */ /* @@ -142,44 +145,91 @@ static struct vfsmount *cpuset_mount; static struct super_block *cpuset_sb = NULL; /* - * cpuset_sem should be held by anyone who is depending on the children - * or sibling lists of any cpuset, or performing non-atomic operations - * on the flags or *_allowed values of a cpuset, such as raising the - * CS_REMOVED flag bit iff it is not already raised, or reading and - * conditionally modifying the *_allowed values. One kernel global - * cpuset semaphore should be sufficient - these things don't change - * that much. - * - * The code that modifies cpusets holds cpuset_sem across the entire - * operation, from cpuset_common_file_write() down, single threading - * all cpuset modifications (except for counter manipulations from - * fork and exit) across the system. This presumes that cpuset - * modifications are rare - better kept simple and safe, even if slow. - * - * The code that reads cpusets, such as in cpuset_common_file_read() - * and below, only holds cpuset_sem across small pieces of code, such - * as when reading out possibly multi-word cpumasks and nodemasks, as - * the risks are less, and the desire for performance a little greater. - * The proc_cpuset_show() routine needs to hold cpuset_sem to insure - * that no cs->dentry is NULL, as it walks up the cpuset tree to root. - * - * The hooks from fork and exit, cpuset_fork() and cpuset_exit(), don't - * (usually) grab cpuset_sem. These are the two most performance - * critical pieces of code here. The exception occurs on exit(), - * when a task in a notify_on_release cpuset exits. Then cpuset_sem + * We have two global cpuset semaphores below. They can nest. + * It is ok to first take manage_sem, then nest callback_sem. We also + * require taking task_lock() when dereferencing a tasks cpuset pointer. + * See "The task_lock() exception", at the end of this comment. + * + * A task must hold both semaphores to modify cpusets. If a task + * holds manage_sem, then it blocks others wanting that semaphore, + * ensuring that it is the only task able to also acquire callback_sem + * and be able to modify cpusets. It can perform various checks on + * the cpuset structure first, knowing nothing will change. It can + * also allocate memory while just holding manage_sem. While it is + * performing these checks, various callback routines can briefly + * acquire callback_sem to query cpusets. Once it is ready to make + * the changes, it takes callback_sem, blocking everyone else. + * + * Calls to the kernel memory allocator can not be made while holding + * callback_sem, as that would risk double tripping on callback_sem + * from one of the callbacks into the cpuset code from within + * __alloc_pages(). + * + * If a task is only holding callback_sem, then it has read-only + * access to cpusets. + * + * The task_struct fields mems_allowed and mems_generation may only + * be accessed in the context of that task, so require no locks. + * + * Any task can increment and decrement the count field without lock. + * So in general, code holding manage_sem or callback_sem can't rely + * on the count field not changing. However, if the count goes to + * zero, then only attach_task(), which holds both semaphores, can + * increment it again. Because a count of zero means that no tasks + * are currently attached, therefore there is no way a task attached + * to that cpuset can fork (the other way to increment the count). + * So code holding manage_sem or callback_sem can safely assume that + * if the count is zero, it will stay zero. Similarly, if a task + * holds manage_sem or callback_sem on a cpuset with zero count, it + * knows that the cpuset won't be removed, as cpuset_rmdir() needs + * both of those semaphores. + * + * A possible optimization to improve parallelism would be to make + * callback_sem a R/W semaphore (rwsem), allowing the callback routines + * to proceed in parallel, with read access, until the holder of + * manage_sem needed to take this rwsem for exclusive write access + * and modify some cpusets. + * + * The cpuset_common_file_write handler for operations that modify + * the cpuset hierarchy holds manage_sem across the entire operation, + * single threading all such cpuset modifications across the system. + * + * The cpuset_common_file_read() handlers only hold callback_sem across + * small pieces of code, such as when reading out possibly multi-word + * cpumasks and nodemasks. + * + * The fork and exit callbacks cpuset_fork() and cpuset_exit(), don't + * (usually) take either semaphore. These are the two most performance + * critical pieces of code here. The exception occurs on cpuset_exit(), + * when a task in a notify_on_release cpuset exits. Then manage_sem * is taken, and if the cpuset count is zero, a usermode call made * to /sbin/cpuset_release_agent with the name of the cpuset (path * relative to the root of cpuset file system) as the argument. * - * A cpuset can only be deleted if both its 'count' of using tasks is - * zero, and its list of 'children' cpusets is empty. Since all tasks - * in the system use _some_ cpuset, and since there is always at least - * one task in the system (init, pid == 1), therefore, top_cpuset - * always has either children cpusets and/or using tasks. So no need - * for any special hack to ensure that top_cpuset cannot be deleted. + * A cpuset can only be deleted if both its 'count' of using tasks + * is zero, and its list of 'children' cpusets is empty. Since all + * tasks in the system use _some_ cpuset, and since there is always at + * least one task in the system (init, pid == 1), therefore, top_cpuset + * always has either children cpusets and/or using tasks. So we don't + * need a special hack to ensure that top_cpuset cannot be deleted. + * + * The above "Tale of Two Semaphores" would be complete, but for: + * + * The task_lock() exception + * + * The need for this exception arises from the action of attach_task(), + * which overwrites one tasks cpuset pointer with another. It does + * so using both semaphores, however there are several performance + * critical places that need to reference task->cpuset without the + * expense of grabbing a system global semaphore. Therefore except as + * noted below, when dereferencing or, as in attach_task(), modifying + * a tasks cpuset pointer we use task_lock(), which acts on a spinlock + * (task->alloc_lock) already in the task_struct routinely used for + * such matters. */ -static DECLARE_MUTEX(cpuset_sem); +static DECLARE_MUTEX(manage_sem); +static DECLARE_MUTEX(callback_sem); /* * A couple of forward declarations required, due to cyclic reference loop: @@ -354,7 +404,7 @@ static inline struct cftype *__d_cft(struct dentry *dentry) } /* - * Call with cpuset_sem held. Writes path of cpuset into buf. + * Call with manage_sem held. Writes path of cpuset into buf. * Returns 0 on success, -errno on error. */ @@ -406,10 +456,11 @@ static int cpuset_path(const struct cpuset *cs, char *buf, int buflen) * status of the /sbin/cpuset_release_agent task, so no sense holding * our caller up for that. * - * The simple act of forking that task might require more memory, - * which might need cpuset_sem. So this routine must be called while - * cpuset_sem is not held, to avoid a possible deadlock. See also - * comments for check_for_release(), below. + * When we had only one cpuset semaphore, we had to call this + * without holding it, to avoid deadlock when call_usermodehelper() + * allocated memory. With two locks, we could now call this while + * holding manage_sem, but we still don't, so as to minimize + * the time manage_sem is held. */ static void cpuset_release_agent(const char *pathbuf) @@ -441,15 +492,15 @@ static void cpuset_release_agent(const char *pathbuf) * cs is notify_on_release() and now both the user count is zero and * the list of children is empty, prepare cpuset path in a kmalloc'd * buffer, to be returned via ppathbuf, so that the caller can invoke - * cpuset_release_agent() with it later on, once cpuset_sem is dropped. - * Call here with cpuset_sem held. + * cpuset_release_agent() with it later on, once manage_sem is dropped. + * Call here with manage_sem held. * * This check_for_release() routine is responsible for kmalloc'ing * pathbuf. The above cpuset_release_agent() is responsible for * kfree'ing pathbuf. The caller of these routines is responsible * for providing a pathbuf pointer, initialized to NULL, then - * calling check_for_release() with cpuset_sem held and the address - * of the pathbuf pointer, then dropping cpuset_sem, then calling + * calling check_for_release() with manage_sem held and the address + * of the pathbuf pointer, then dropping manage_sem, then calling * cpuset_release_agent() with pathbuf, as set by check_for_release(). */ @@ -480,7 +531,7 @@ static void check_for_release(struct cpuset *cs, char **ppathbuf) * One way or another, we guarantee to return some non-empty subset * of cpu_online_map. * - * Call with cpuset_sem held. + * Call with callback_sem held. */ static void guarantee_online_cpus(const struct cpuset *cs, cpumask_t *pmask) @@ -504,7 +555,7 @@ static void guarantee_online_cpus(const struct cpuset *cs, cpumask_t *pmask) * One way or another, we guarantee to return some non-empty subset * of node_online_map. * - * Call with cpuset_sem held. + * Call with callback_sem held. */ static void guarantee_online_mems(const struct cpuset *cs, nodemask_t *pmask) @@ -519,31 +570,44 @@ static void guarantee_online_mems(const struct cpuset *cs, nodemask_t *pmask) } /* - * Refresh current tasks mems_allowed and mems_generation from - * current tasks cpuset. Call with cpuset_sem held. - * - * Be sure to call refresh_mems() on any cpuset operation which - * (1) holds cpuset_sem, and (2) might possibly alloc memory. - * Call after obtaining cpuset_sem lock, before any possible - * allocation. Otherwise one risks trying to allocate memory - * while the task cpuset_mems_generation is not the same as - * the mems_generation in its cpuset, which would deadlock on - * cpuset_sem in cpuset_update_current_mems_allowed(). - * - * Since we hold cpuset_sem, once refresh_mems() is called, the - * test (current->cpuset_mems_generation != cs->mems_generation) - * in cpuset_update_current_mems_allowed() will remain false, - * until we drop cpuset_sem. Anyone else who would change our - * cpusets mems_generation needs to lock cpuset_sem first. + * Refresh current tasks mems_allowed and mems_generation from current + * tasks cpuset. + * + * Call without callback_sem or task_lock() held. May be called with + * or without manage_sem held. Will acquire task_lock() and might + * acquire callback_sem during call. + * + * The task_lock() is required to dereference current->cpuset safely. + * Without it, we could pick up the pointer value of current->cpuset + * in one instruction, and then attach_task could give us a different + * cpuset, and then the cpuset we had could be removed and freed, + * and then on our next instruction, we could dereference a no longer + * valid cpuset pointer to get its mems_generation field. + * + * This routine is needed to update the per-task mems_allowed data, + * within the tasks context, when it is trying to allocate memory + * (in various mm/mempolicy.c routines) and notices that some other + * task has been modifying its cpuset. */ static void refresh_mems(void) { - struct cpuset *cs = current->cpuset; + int my_cpusets_mem_gen; + + task_lock(current); + my_cpusets_mem_gen = current->cpuset->mems_generation; + task_unlock(current); - if (current->cpuset_mems_generation != cs->mems_generation) { + if (current->cpuset_mems_generation != my_cpusets_mem_gen) { + struct cpuset *cs; + + down(&callback_sem); + task_lock(current); + cs = current->cpuset; guarantee_online_mems(cs, ¤t->mems_allowed); current->cpuset_mems_generation = cs->mems_generation; + task_unlock(current); + up(&callback_sem); } } @@ -552,7 +616,7 @@ static void refresh_mems(void) * * One cpuset is a subset of another if all its allowed CPUs and * Memory Nodes are a subset of the other, and its exclusive flags - * are only set if the other's are set. + * are only set if the other's are set. Call holding manage_sem. */ static int is_cpuset_subset(const struct cpuset *p, const struct cpuset *q) @@ -570,7 +634,7 @@ static int is_cpuset_subset(const struct cpuset *p, const struct cpuset *q) * If we replaced the flag and mask values of the current cpuset * (cur) with those values in the trial cpuset (trial), would * our various subset and exclusive rules still be valid? Presumes - * cpuset_sem held. + * manage_sem held. * * 'cur' is the address of an actual, in-use cpuset. Operations * such as list traversal that depend on the actual address of the @@ -624,7 +688,7 @@ static int validate_change(const struct cpuset *cur, const struct cpuset *trial) * exclusive child cpusets * Build these two partitions by calling partition_sched_domains * - * Call with cpuset_sem held. May nest a call to the + * Call with manage_sem held. May nest a call to the * lock_cpu_hotplug()/unlock_cpu_hotplug() pair. */ @@ -669,6 +733,10 @@ static void update_cpu_domains(struct cpuset *cur) unlock_cpu_hotplug(); } +/* + * Call with manage_sem held. May take callback_sem during call. + */ + static int update_cpumask(struct cpuset *cs, char *buf) { struct cpuset trialcs; @@ -685,12 +753,18 @@ static int update_cpumask(struct cpuset *cs, char *buf) if (retval < 0) return retval; cpus_unchanged = cpus_equal(cs->cpus_allowed, trialcs.cpus_allowed); + down(&callback_sem); cs->cpus_allowed = trialcs.cpus_allowed; + up(&callback_sem); if (is_cpu_exclusive(cs) && !cpus_unchanged) update_cpu_domains(cs); return 0; } +/* + * Call with manage_sem held. May take callback_sem during call. + */ + static int update_nodemask(struct cpuset *cs, char *buf) { struct cpuset trialcs; @@ -705,9 +779,11 @@ static int update_nodemask(struct cpuset *cs, char *buf) return -ENOSPC; retval = validate_change(cs, &trialcs); if (retval == 0) { + down(&callback_sem); cs->mems_allowed = trialcs.mems_allowed; atomic_inc(&cpuset_mems_generation); cs->mems_generation = atomic_read(&cpuset_mems_generation); + up(&callback_sem); } return retval; } @@ -718,6 +794,8 @@ static int update_nodemask(struct cpuset *cs, char *buf) * CS_NOTIFY_ON_RELEASE) * cs: the cpuset to update * buf: the buffer where we read the 0 or 1 + * + * Call with manage_sem held. */ static int update_flag(cpuset_flagbits_t bit, struct cpuset *cs, char *buf) @@ -739,16 +817,27 @@ static int update_flag(cpuset_flagbits_t bit, struct cpuset *cs, char *buf) return err; cpu_exclusive_changed = (is_cpu_exclusive(cs) != is_cpu_exclusive(&trialcs)); + down(&callback_sem); if (turning_on) set_bit(bit, &cs->flags); else clear_bit(bit, &cs->flags); + up(&callback_sem); if (cpu_exclusive_changed) update_cpu_domains(cs); return 0; } +/* + * Attack task specified by pid in 'pidbuf' to cpuset 'cs', possibly + * writing the path of the old cpuset in 'ppathbuf' if it needs to be + * notified on release. + * + * Call holding manage_sem. May take callback_sem and task_lock of + * the task 'pid' during call. + */ + static int attach_task(struct cpuset *cs, char *pidbuf, char **ppathbuf) { pid_t pid; @@ -765,7 +854,7 @@ static int attach_task(struct cpuset *cs, char *pidbuf, char **ppathbuf) read_lock(&tasklist_lock); tsk = find_task_by_pid(pid); - if (!tsk) { + if (!tsk || tsk->flags & PF_EXITING) { read_unlock(&tasklist_lock); return -ESRCH; } @@ -783,10 +872,13 @@ static int attach_task(struct cpuset *cs, char *pidbuf, char **ppathbuf) get_task_struct(tsk); } + down(&callback_sem); + task_lock(tsk); oldcs = tsk->cpuset; if (!oldcs) { task_unlock(tsk); + up(&callback_sem); put_task_struct(tsk); return -ESRCH; } @@ -797,6 +889,7 @@ static int attach_task(struct cpuset *cs, char *pidbuf, char **ppathbuf) guarantee_online_cpus(cs, &cpus); set_cpus_allowed(tsk, cpus); + up(&callback_sem); put_task_struct(tsk); if (atomic_dec_and_test(&oldcs->count)) check_for_release(oldcs, ppathbuf); @@ -840,7 +933,7 @@ static ssize_t cpuset_common_file_write(struct file *file, const char __user *us } buffer[nbytes] = 0; /* nul-terminate */ - down(&cpuset_sem); + down(&manage_sem); if (is_removed(cs)) { retval = -ENODEV; @@ -874,7 +967,7 @@ static ssize_t cpuset_common_file_write(struct file *file, const char __user *us if (retval == 0) retval = nbytes; out2: - up(&cpuset_sem); + up(&manage_sem); cpuset_release_agent(pathbuf); out1: kfree(buffer); @@ -914,9 +1007,9 @@ static int cpuset_sprintf_cpulist(char *page, struct cpuset *cs) { cpumask_t mask; - down(&cpuset_sem); + down(&callback_sem); mask = cs->cpus_allowed; - up(&cpuset_sem); + up(&callback_sem); return cpulist_scnprintf(page, PAGE_SIZE, mask); } @@ -925,9 +1018,9 @@ static int cpuset_sprintf_memlist(char *page, struct cpuset *cs) { nodemask_t mask; - down(&cpuset_sem); + down(&callback_sem); mask = cs->mems_allowed; - up(&cpuset_sem); + up(&callback_sem); return nodelist_scnprintf(page, PAGE_SIZE, mask); } @@ -1135,7 +1228,9 @@ struct ctr_struct { /* * Load into 'pidarray' up to 'npids' of the tasks using cpuset 'cs'. - * Return actual number of pids loaded. + * Return actual number of pids loaded. No need to task_lock(p) + * when reading out p->cpuset, as we don't really care if it changes + * on the next cycle, and we are not going to try to dereference it. */ static inline int pid_array_load(pid_t *pidarray, int npids, struct cpuset *cs) { @@ -1177,6 +1272,12 @@ static int pid_array_to_buf(char *buf, int sz, pid_t *a, int npids) return cnt; } +/* + * Handle an open on 'tasks' file. Prepare a buffer listing the + * process id's of tasks currently attached to the cpuset being opened. + * + * Does not require any specific cpuset semaphores, and does not take any. + */ static int cpuset_tasks_open(struct inode *unused, struct file *file) { struct cpuset *cs = __d_cs(file->f_dentry->d_parent); @@ -1324,7 +1425,7 @@ static long cpuset_create(struct cpuset *parent, const char *name, int mode) if (!cs) return -ENOMEM; - down(&cpuset_sem); + down(&manage_sem); refresh_mems(); cs->flags = 0; if (notify_on_release(parent)) @@ -1339,25 +1440,27 @@ static long cpuset_create(struct cpuset *parent, const char *name, int mode) cs->parent = parent; + down(&callback_sem); list_add(&cs->sibling, &cs->parent->children); + up(&callback_sem); err = cpuset_create_dir(cs, name, mode); if (err < 0) goto err; /* - * Release cpuset_sem before cpuset_populate_dir() because it + * Release manage_sem before cpuset_populate_dir() because it * will down() this new directory's i_sem and if we race with * another mkdir, we might deadlock. */ - up(&cpuset_sem); + up(&manage_sem); err = cpuset_populate_dir(cs->dentry); /* If err < 0, we have a half-filled directory - oh well ;) */ return 0; err: list_del(&cs->sibling); - up(&cpuset_sem); + up(&manage_sem); kfree(cs); return err; } @@ -1379,30 +1482,32 @@ static int cpuset_rmdir(struct inode *unused_dir, struct dentry *dentry) /* the vfs holds both inode->i_sem already */ - down(&cpuset_sem); + down(&manage_sem); refresh_mems(); if (atomic_read(&cs->count) > 0) { - up(&cpuset_sem); + up(&manage_sem); return -EBUSY; } if (!list_empty(&cs->children)) { - up(&cpuset_sem); + up(&manage_sem); return -EBUSY; } parent = cs->parent; + down(&callback_sem); set_bit(CS_REMOVED, &cs->flags); if (is_cpu_exclusive(cs)) update_cpu_domains(cs); list_del(&cs->sibling); /* delete my sibling from parent->children */ - if (list_empty(&parent->children)) - check_for_release(parent, &pathbuf); spin_lock(&cs->dentry->d_lock); d = dget(cs->dentry); cs->dentry = NULL; spin_unlock(&d->d_lock); cpuset_d_remove_dir(d); dput(d); - up(&cpuset_sem); + up(&callback_sem); + if (list_empty(&parent->children)) + check_for_release(parent, &pathbuf); + up(&manage_sem); cpuset_release_agent(pathbuf); return 0; } @@ -1462,16 +1567,26 @@ void __init cpuset_init_smp(void) * cpuset_fork - attach newly forked task to its parents cpuset. * @tsk: pointer to task_struct of forking parent process. * - * Description: By default, on fork, a task inherits its - * parent's cpuset. The pointer to the shared cpuset is - * automatically copied in fork.c by dup_task_struct(). - * This cpuset_fork() routine need only increment the usage - * counter in that cpuset. + * Description: A task inherits its parent's cpuset at fork(). + * + * A pointer to the shared cpuset was automatically copied in fork.c + * by dup_task_struct(). However, we ignore that copy, since it was + * not made under the protection of task_lock(), so might no longer be + * a valid cpuset pointer. attach_task() might have already changed + * current->cpuset, allowing the previously referenced cpuset to + * be removed and freed. Instead, we task_lock(current) and copy + * its present value of current->cpuset for our freshly forked child. + * + * At the point that cpuset_fork() is called, 'current' is the parent + * task, and the passed argument 'child' points to the child task. **/ -void cpuset_fork(struct task_struct *tsk) +void cpuset_fork(struct task_struct *child) { - atomic_inc(&tsk->cpuset->count); + task_lock(current); + child->cpuset = current->cpuset; + atomic_inc(&child->cpuset->count); + task_unlock(current); } /** @@ -1480,35 +1595,42 @@ void cpuset_fork(struct task_struct *tsk) * * Description: Detach cpuset from @tsk and release it. * - * Note that cpusets marked notify_on_release force every task - * in them to take the global cpuset_sem semaphore when exiting. - * This could impact scaling on very large systems. Be reluctant - * to use notify_on_release cpusets where very high task exit - * scaling is required on large systems. - * - * Don't even think about derefencing 'cs' after the cpuset use - * count goes to zero, except inside a critical section guarded - * by the cpuset_sem semaphore. If you don't hold cpuset_sem, - * then a zero cpuset use count is a license to any other task to - * nuke the cpuset immediately. + * Note that cpusets marked notify_on_release force every task in + * them to take the global manage_sem semaphore when exiting. + * This could impact scaling on very large systems. Be reluctant to + * use notify_on_release cpusets where very high task exit scaling + * is required on large systems. + * + * Don't even think about derefencing 'cs' after the cpuset use count + * goes to zero, except inside a critical section guarded by manage_sem + * or callback_sem. Otherwise a zero cpuset use count is a license to + * any other task to nuke the cpuset immediately, via cpuset_rmdir(). + * + * This routine has to take manage_sem, not callback_sem, because + * it is holding that semaphore while calling check_for_release(), + * which calls kmalloc(), so can't be called holding callback__sem(). + * + * We don't need to task_lock() this reference to tsk->cpuset, + * because tsk is already marked PF_EXITING, so attach_task() won't + * mess with it. **/ void cpuset_exit(struct task_struct *tsk) { struct cpuset *cs; - task_lock(tsk); + BUG_ON(!(tsk->flags & PF_EXITING)); + cs = tsk->cpuset; tsk->cpuset = NULL; - task_unlock(tsk); if (notify_on_release(cs)) { char *pathbuf = NULL; - down(&cpuset_sem); + down(&manage_sem); if (atomic_dec_and_test(&cs->count)) check_for_release(cs, &pathbuf); - up(&cpuset_sem); + up(&manage_sem); cpuset_release_agent(pathbuf); } else { atomic_dec(&cs->count); @@ -1529,11 +1651,11 @@ cpumask_t cpuset_cpus_allowed(const struct task_struct *tsk) { cpumask_t mask; - down(&cpuset_sem); + down(&callback_sem); task_lock((struct task_struct *)tsk); guarantee_online_cpus(tsk->cpuset, &mask); task_unlock((struct task_struct *)tsk); - up(&cpuset_sem); + up(&callback_sem); return mask; } @@ -1549,19 +1671,28 @@ void cpuset_init_current_mems_allowed(void) * If the current tasks cpusets mems_allowed changed behind our backs, * update current->mems_allowed and mems_generation to the new value. * Do not call this routine if in_interrupt(). + * + * Call without callback_sem or task_lock() held. May be called + * with or without manage_sem held. Unless exiting, it will acquire + * task_lock(). Also might acquire callback_sem during call to + * refresh_mems(). */ void cpuset_update_current_mems_allowed(void) { - struct cpuset *cs = current->cpuset; + struct cpuset *cs; + int need_to_refresh = 0; + task_lock(current); + cs = current->cpuset; if (!cs) - return; /* task is exiting */ - if (current->cpuset_mems_generation != cs->mems_generation) { - down(&cpuset_sem); + goto done; + if (current->cpuset_mems_generation != cs->mems_generation) + need_to_refresh = 1; +done: + task_unlock(current); + if (need_to_refresh) refresh_mems(); - up(&cpuset_sem); - } } /** @@ -1595,7 +1726,7 @@ int cpuset_zonelist_valid_mems_allowed(struct zonelist *zl) /* * nearest_exclusive_ancestor() - Returns the nearest mem_exclusive - * ancestor to the specified cpuset. Call while holding cpuset_sem. + * ancestor to the specified cpuset. Call holding callback_sem. * If no ancestor is mem_exclusive (an unusual configuration), then * returns the root cpuset. */ @@ -1622,12 +1753,12 @@ static const struct cpuset *nearest_exclusive_ancestor(const struct cpuset *cs) * GFP_KERNEL allocations are not so marked, so can escape to the * nearest mem_exclusive ancestor cpuset. * - * Scanning up parent cpusets requires cpuset_sem. The __alloc_pages() + * Scanning up parent cpusets requires callback_sem. The __alloc_pages() * routine only calls here with __GFP_HARDWALL bit _not_ set if * it's a GFP_KERNEL allocation, and all nodes in the current tasks * mems_allowed came up empty on the first pass over the zonelist. * So only GFP_KERNEL allocations, if all nodes in the cpuset are - * short of memory, might require taking the cpuset_sem semaphore. + * short of memory, might require taking the callback_sem semaphore. * * The first loop over the zonelist in mm/page_alloc.c:__alloc_pages() * calls here with __GFP_HARDWALL always set in gfp_mask, enforcing @@ -1659,14 +1790,16 @@ int cpuset_zone_allowed(struct zone *z, gfp_t gfp_mask) return 0; /* Not hardwall and node outside mems_allowed: scan up cpusets */ - down(&cpuset_sem); - cs = current->cpuset; - if (!cs) - goto done; /* current task exiting */ - cs = nearest_exclusive_ancestor(cs); + down(&callback_sem); + + if (current->flags & PF_EXITING) /* Let dying task have memory */ + return 1; + task_lock(current); + cs = nearest_exclusive_ancestor(current->cpuset); + task_unlock(current); + allowed = node_isset(node, cs->mems_allowed); -done: - up(&cpuset_sem); + up(&callback_sem); return allowed; } @@ -1679,7 +1812,7 @@ done: * determine if task @p's memory usage might impact the memory * available to the current task. * - * Acquires cpuset_sem - not suitable for calling from a fast path. + * Acquires callback_sem - not suitable for calling from a fast path. **/ int cpuset_excl_nodes_overlap(const struct task_struct *p) @@ -1687,18 +1820,27 @@ int cpuset_excl_nodes_overlap(const struct task_struct *p) const struct cpuset *cs1, *cs2; /* my and p's cpuset ancestors */ int overlap = 0; /* do cpusets overlap? */ - down(&cpuset_sem); - cs1 = current->cpuset; - if (!cs1) - goto done; /* current task exiting */ - cs2 = p->cpuset; - if (!cs2) - goto done; /* task p is exiting */ - cs1 = nearest_exclusive_ancestor(cs1); - cs2 = nearest_exclusive_ancestor(cs2); + down(&callback_sem); + + task_lock(current); + if (current->flags & PF_EXITING) { + task_unlock(current); + goto done; + } + cs1 = nearest_exclusive_ancestor(current->cpuset); + task_unlock(current); + + task_lock((struct task_struct *)p); + if (p->flags & PF_EXITING) { + task_unlock((struct task_struct *)p); + goto done; + } + cs2 = nearest_exclusive_ancestor(p->cpuset); + task_unlock((struct task_struct *)p); + overlap = nodes_intersects(cs1->mems_allowed, cs2->mems_allowed); done: - up(&cpuset_sem); + up(&callback_sem); return overlap; } @@ -1707,6 +1849,10 @@ done: * proc_cpuset_show() * - Print tasks cpuset path into seq_file. * - Used for /proc//cpuset. + * - No need to task_lock(tsk) on this tsk->cpuset reference, as it + * doesn't really matter if tsk->cpuset changes after we read it, + * and we take manage_sem, keeping attach_task() from changing it + * anyway. */ static int proc_cpuset_show(struct seq_file *m, void *v) @@ -1721,10 +1867,8 @@ static int proc_cpuset_show(struct seq_file *m, void *v) return -ENOMEM; tsk = m->private; - down(&cpuset_sem); - task_lock(tsk); + down(&manage_sem); cs = tsk->cpuset; - task_unlock(tsk); if (!cs) { retval = -EINVAL; goto out; @@ -1736,7 +1880,7 @@ static int proc_cpuset_show(struct seq_file *m, void *v) seq_puts(m, buf); seq_putc(m, '\n'); out: - up(&cpuset_sem); + up(&manage_sem); kfree(buf); return retval; } -- cgit v1.2.3 From fb5eeeee44edb248b4837416966f19731f497f79 Mon Sep 17 00:00:00 2001 From: Paul Jackson Date: Sun, 30 Oct 2005 15:02:33 -0800 Subject: [PATCH] cpusets: bitmap and mask remap operators In the forthcoming task migration support, a key calculation will be mapping cpu and node numbers from the old set to the new set while preserving cpuset-relative offset. For example, if a task and its pages on nodes 8-11 are being migrated to nodes 24-27, then pages on node 9 (the 2nd node in the old set) should be moved to node 25 (the 2nd node in the new set.) As with other bitmap operations, the proper way to code this is to provide the underlying calculation in lib/bitmap.c, and then to provide the usual cpumask and nodemask wrappers. This patch provides that. These operations are termed 'remap' operations. Both remapping a single bit and a set of bits is supported. Signed-off-by: Paul Jackson Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/bitmap.h | 6 ++ include/linux/cpumask.h | 20 ++++++ include/linux/nodemask.h | 20 ++++++ lib/bitmap.c | 166 +++++++++++++++++++++++++++++++++++++++++++++++ 4 files changed, 212 insertions(+) (limited to 'include/linux') diff --git a/include/linux/bitmap.h b/include/linux/bitmap.h index 86dd5502b05c..7d8ff97b3e92 100644 --- a/include/linux/bitmap.h +++ b/include/linux/bitmap.h @@ -40,6 +40,8 @@ * bitmap_weight(src, nbits) Hamming Weight: number set bits * bitmap_shift_right(dst, src, n, nbits) *dst = *src >> n * bitmap_shift_left(dst, src, n, nbits) *dst = *src << n + * bitmap_remap(dst, src, old, new, nbits) *dst = map(old, new)(src) + * bitmap_bitremap(oldbit, old, new, nbits) newbit = map(old, new)(oldbit) * bitmap_scnprintf(buf, len, src, nbits) Print bitmap src to buf * bitmap_parse(ubuf, ulen, dst, nbits) Parse bitmap dst from user buf * bitmap_scnlistprintf(buf, len, src, nbits) Print bitmap src as list to buf @@ -104,6 +106,10 @@ extern int bitmap_scnlistprintf(char *buf, unsigned int len, const unsigned long *src, int nbits); extern int bitmap_parselist(const char *buf, unsigned long *maskp, int nmaskbits); +extern void bitmap_remap(unsigned long *dst, const unsigned long *src, + const unsigned long *old, const unsigned long *new, int bits); +extern int bitmap_bitremap(int oldbit, + const unsigned long *old, const unsigned long *new, int bits); extern int bitmap_find_free_region(unsigned long *bitmap, int bits, int order); extern void bitmap_release_region(unsigned long *bitmap, int pos, int order); extern int bitmap_allocate_region(unsigned long *bitmap, int pos, int order); diff --git a/include/linux/cpumask.h b/include/linux/cpumask.h index 9bdba8169b41..13e9f4a3ab26 100644 --- a/include/linux/cpumask.h +++ b/include/linux/cpumask.h @@ -12,6 +12,8 @@ * see bitmap_scnprintf() and bitmap_parse() in lib/bitmap.c. * For details of cpulist_scnprintf() and cpulist_parse(), see * bitmap_scnlistprintf() and bitmap_parselist(), also in bitmap.c. + * For details of cpu_remap(), see bitmap_bitremap in lib/bitmap.c + * For details of cpus_remap(), see bitmap_remap in lib/bitmap.c. * * The available cpumask operations are: * @@ -50,6 +52,8 @@ * int cpumask_parse(ubuf, ulen, mask) Parse ascii string as cpumask * int cpulist_scnprintf(buf, len, mask) Format cpumask as list for printing * int cpulist_parse(buf, map) Parse ascii string as cpulist + * int cpu_remap(oldbit, old, new) newbit = map(old, new)(oldbit) + * int cpus_remap(dst, src, old, new) *dst = map(old, new)(src) * * for_each_cpu_mask(cpu, mask) for-loop cpu over mask * @@ -294,6 +298,22 @@ static inline int __cpulist_parse(const char *buf, cpumask_t *dstp, int nbits) return bitmap_parselist(buf, dstp->bits, nbits); } +#define cpu_remap(oldbit, old, new) \ + __cpu_remap((oldbit), &(old), &(new), NR_CPUS) +static inline int __cpu_remap(int oldbit, + const cpumask_t *oldp, const cpumask_t *newp, int nbits) +{ + return bitmap_bitremap(oldbit, oldp->bits, newp->bits, nbits); +} + +#define cpus_remap(dst, src, old, new) \ + __cpus_remap(&(dst), &(src), &(old), &(new), NR_CPUS) +static inline void __cpus_remap(cpumask_t *dstp, const cpumask_t *srcp, + const cpumask_t *oldp, const cpumask_t *newp, int nbits) +{ + bitmap_remap(dstp->bits, srcp->bits, oldp->bits, newp->bits, nbits); +} + #if NR_CPUS > 1 #define for_each_cpu_mask(cpu, mask) \ for ((cpu) = first_cpu(mask); \ diff --git a/include/linux/nodemask.h b/include/linux/nodemask.h index e96fe9062500..4726ef7ba8e8 100644 --- a/include/linux/nodemask.h +++ b/include/linux/nodemask.h @@ -12,6 +12,8 @@ * see bitmap_scnprintf() and bitmap_parse() in lib/bitmap.c. * For details of nodelist_scnprintf() and nodelist_parse(), see * bitmap_scnlistprintf() and bitmap_parselist(), also in bitmap.c. + * For details of node_remap(), see bitmap_bitremap in lib/bitmap.c. + * For details of nodes_remap(), see bitmap_remap in lib/bitmap.c. * * The available nodemask operations are: * @@ -52,6 +54,8 @@ * int nodemask_parse(ubuf, ulen, mask) Parse ascii string as nodemask * int nodelist_scnprintf(buf, len, mask) Format nodemask as list for printing * int nodelist_parse(buf, map) Parse ascii string as nodelist + * int node_remap(oldbit, old, new) newbit = map(old, new)(oldbit) + * int nodes_remap(dst, src, old, new) *dst = map(old, new)(dst) * * for_each_node_mask(node, mask) for-loop node over mask * @@ -307,6 +311,22 @@ static inline int __nodelist_parse(const char *buf, nodemask_t *dstp, int nbits) return bitmap_parselist(buf, dstp->bits, nbits); } +#define node_remap(oldbit, old, new) \ + __node_remap((oldbit), &(old), &(new), MAX_NUMNODES) +static inline int __node_remap(int oldbit, + const nodemask_t *oldp, const nodemask_t *newp, int nbits) +{ + return bitmap_bitremap(oldbit, oldp->bits, newp->bits, nbits); +} + +#define nodes_remap(dst, src, old, new) \ + __nodes_remap(&(dst), &(src), &(old), &(new), MAX_NUMNODES) +static inline void __nodes_remap(nodemask_t *dstp, const nodemask_t *srcp, + const nodemask_t *oldp, const nodemask_t *newp, int nbits) +{ + bitmap_remap(dstp->bits, srcp->bits, oldp->bits, newp->bits, nbits); +} + #if MAX_NUMNODES > 1 #define for_each_node_mask(node, mask) \ for ((node) = first_node(mask); \ diff --git a/lib/bitmap.c b/lib/bitmap.c index fb9371fdd44a..23d3b1147fe9 100644 --- a/lib/bitmap.c +++ b/lib/bitmap.c @@ -511,6 +511,172 @@ int bitmap_parselist(const char *bp, unsigned long *maskp, int nmaskbits) } EXPORT_SYMBOL(bitmap_parselist); +/* + * bitmap_pos_to_ord(buf, pos, bits) + * @buf: pointer to a bitmap + * @pos: a bit position in @buf (0 <= @pos < @bits) + * @bits: number of valid bit positions in @buf + * + * Map the bit at position @pos in @buf (of length @bits) to the + * ordinal of which set bit it is. If it is not set or if @pos + * is not a valid bit position, map to zero (0). + * + * If for example, just bits 4 through 7 are set in @buf, then @pos + * values 4 through 7 will get mapped to 0 through 3, respectively, + * and other @pos values will get mapped to 0. When @pos value 7 + * gets mapped to (returns) @ord value 3 in this example, that means + * that bit 7 is the 3rd (starting with 0th) set bit in @buf. + * + * The bit positions 0 through @bits are valid positions in @buf. + */ +static int bitmap_pos_to_ord(const unsigned long *buf, int pos, int bits) +{ + int ord = 0; + + if (pos >= 0 && pos < bits) { + int i; + + for (i = find_first_bit(buf, bits); + i < pos; + i = find_next_bit(buf, bits, i + 1)) + ord++; + if (i > pos) + ord = 0; + } + return ord; +} + +/** + * bitmap_ord_to_pos(buf, ord, bits) + * @buf: pointer to bitmap + * @ord: ordinal bit position (n-th set bit, n >= 0) + * @bits: number of valid bit positions in @buf + * + * Map the ordinal offset of bit @ord in @buf to its position in @buf. + * If @ord is not the ordinal offset of a set bit in @buf, map to zero (0). + * + * If for example, just bits 4 through 7 are set in @buf, then @ord + * values 0 through 3 will get mapped to 4 through 7, respectively, + * and all other @ord valuds will get mapped to 0. When @ord value 3 + * gets mapped to (returns) @pos value 7 in this example, that means + * that the 3rd set bit (starting with 0th) is at position 7 in @buf. + * + * The bit positions 0 through @bits are valid positions in @buf. + */ +static int bitmap_ord_to_pos(const unsigned long *buf, int ord, int bits) +{ + int pos = 0; + + if (ord >= 0 && ord < bits) { + int i; + + for (i = find_first_bit(buf, bits); + i < bits && ord > 0; + i = find_next_bit(buf, bits, i + 1)) + ord--; + if (i < bits && ord == 0) + pos = i; + } + + return pos; +} + +/** + * bitmap_remap - Apply map defined by a pair of bitmaps to another bitmap + * @src: subset to be remapped + * @dst: remapped result + * @old: defines domain of map + * @new: defines range of map + * @bits: number of bits in each of these bitmaps + * + * Let @old and @new define a mapping of bit positions, such that + * whatever position is held by the n-th set bit in @old is mapped + * to the n-th set bit in @new. In the more general case, allowing + * for the possibility that the weight 'w' of @new is less than the + * weight of @old, map the position of the n-th set bit in @old to + * the position of the m-th set bit in @new, where m == n % w. + * + * If either of the @old and @new bitmaps are empty, or if@src and @dst + * point to the same location, then this routine does nothing. + * + * The positions of unset bits in @old are mapped to the position of + * the first set bit in @new. + * + * Apply the above specified mapping to @src, placing the result in + * @dst, clearing any bits previously set in @dst. + * + * The resulting value of @dst will have either the same weight as + * @src, or less weight in the general case that the mapping wasn't + * injective due to the weight of @new being less than that of @old. + * The resulting value of @dst will never have greater weight than + * that of @src, except perhaps in the case that one of the above + * conditions was not met and this routine just returned. + * + * For example, lets say that @old has bits 4 through 7 set, and + * @new has bits 12 through 15 set. This defines the mapping of bit + * position 4 to 12, 5 to 13, 6 to 14 and 7 to 15, and of all other + * bit positions to 12 (the first set bit in @new. So if say @src + * comes into this routine with bits 1, 5 and 7 set, then @dst should + * leave with bits 12, 13 and 15 set. + */ +void bitmap_remap(unsigned long *dst, const unsigned long *src, + const unsigned long *old, const unsigned long *new, + int bits) +{ + int s; + + if (bitmap_weight(old, bits) == 0) + return; + if (bitmap_weight(new, bits) == 0) + return; + if (dst == src) /* following doesn't handle inplace remaps */ + return; + + bitmap_zero(dst, bits); + for (s = find_first_bit(src, bits); + s < bits; + s = find_next_bit(src, bits, s + 1)) { + int x = bitmap_pos_to_ord(old, s, bits); + int y = bitmap_ord_to_pos(new, x, bits); + set_bit(y, dst); + } +} +EXPORT_SYMBOL(bitmap_remap); + +/** + * bitmap_bitremap - Apply map defined by a pair of bitmaps to a single bit + * @oldbit - bit position to be mapped + * @old: defines domain of map + * @new: defines range of map + * @bits: number of bits in each of these bitmaps + * + * Let @old and @new define a mapping of bit positions, such that + * whatever position is held by the n-th set bit in @old is mapped + * to the n-th set bit in @new. In the more general case, allowing + * for the possibility that the weight 'w' of @new is less than the + * weight of @old, map the position of the n-th set bit in @old to + * the position of the m-th set bit in @new, where m == n % w. + * + * The positions of unset bits in @old are mapped to the position of + * the first set bit in @new. + * + * Apply the above specified mapping to bit position @oldbit, returning + * the new bit position. + * + * For example, lets say that @old has bits 4 through 7 set, and + * @new has bits 12 through 15 set. This defines the mapping of bit + * position 4 to 12, 5 to 13, 6 to 14 and 7 to 15, and of all other + * bit positions to 12 (the first set bit in @new. So if say @oldbit + * is 5, then this routine returns 13. + */ +int bitmap_bitremap(int oldbit, const unsigned long *old, + const unsigned long *new, int bits) +{ + int x = bitmap_pos_to_ord(old, oldbit, bits); + return bitmap_ord_to_pos(new, x, bits); +} +EXPORT_SYMBOL(bitmap_bitremap); + /** * bitmap_find_free_region - find a contiguous aligned mem region * @bitmap: an array of unsigned longs corresponding to the bitmap -- cgit v1.2.3 From 68860ec10bcc07ab4f89f9d940e3b77ae5ca13b3 Mon Sep 17 00:00:00 2001 From: Paul Jackson Date: Sun, 30 Oct 2005 15:02:36 -0800 Subject: [PATCH] cpusets: automatic numa mempolicy rebinding This patch automatically updates a tasks NUMA mempolicy when its cpuset memory placement changes. It does so within the context of the task, without any need to support low level external mempolicy manipulation. If a system is not using cpusets, or if running on a system with just the root (all-encompassing) cpuset, then this remap is a no-op. Only when a task is moved between cpusets, or a cpusets memory placement is changed does the following apply. Otherwise, the main routine below, rebind_policy() is not even called. When mixing cpusets, scheduler affinity, and NUMA mempolicies, the essential role of cpusets is to place jobs (several related tasks) on a set of CPUs and Memory Nodes, the essential role of sched_setaffinity is to manage a jobs processor placement within its allowed cpuset, and the essential role of NUMA mempolicy (mbind, set_mempolicy) is to manage a jobs memory placement within its allowed cpuset. However, CPU affinity and NUMA memory placement are managed within the kernel using absolute system wide numbering, not cpuset relative numbering. This is ok until a job is migrated to a different cpuset, or what's the same, a jobs cpuset is moved to different CPUs and Memory Nodes. Then the CPU affinity and NUMA memory placement of the tasks in the job need to be updated, to preserve their cpuset-relative position. This can be done for CPU affinity using sched_setaffinity() from user code, as one task can modify anothers CPU affinity. This cannot be done from an external task for NUMA memory placement, as that can only be modified in the context of the task using it. However, it easy enough to remap a tasks NUMA mempolicy automatically when a task is migrated, using the existing cpuset mechanism to trigger a refresh of a tasks memory placement after its cpuset has changed. All that is needed is the old and new nodemask, and notice to the task that it needs to rebind its mempolicy. The tasks mems_allowed has the old mask, the tasks cpuset has the new mask, and the existing cpuset_update_current_mems_allowed() mechanism provides the notice. The bitmap/cpumask/nodemask remap operators provide the cpuset relative calculations. This patch leaves open a couple of issues: 1) Updating vma and shmfs/tmpfs/hugetlbfs memory policies: These mempolicies may reference nodes outside of those allowed to the current task by its cpuset. Tasks are migrated as part of jobs, which reside on what might be several cpusets in a subtree. When such a job is migrated, all NUMA memory policy references to nodes within that cpuset subtree should be translated, and references to any nodes outside that subtree should be left untouched. A future patch will provide the cpuset mechanism needed to mark such subtrees. With that patch, we will be able to correctly migrate these other memory policies across a job migration. 2) Updating cpuset, affinity and memory policies in user space: This is harder. Any placement state stored in user space using system-wide numbering will be invalidated across a migration. More work will be required to provide user code with a migration-safe means to manage its cpuset relative placement, while preserving the current API's that pass system wide numbers, not cpuset relative numbers across the kernel-user boundary. Signed-off-by: Paul Jackson Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/mempolicy.h | 6 +++++ kernel/cpuset.c | 4 +++ mm/mempolicy.c | 64 +++++++++++++++++++++++++++++++++++++++++++++++ 3 files changed, 74 insertions(+) (limited to 'include/linux') diff --git a/include/linux/mempolicy.h b/include/linux/mempolicy.h index 7af8cb836e78..8b67cf837ca9 100644 --- a/include/linux/mempolicy.h +++ b/include/linux/mempolicy.h @@ -154,6 +154,7 @@ struct mempolicy *get_vma_policy(struct task_struct *task, extern void numa_default_policy(void); extern void numa_policy_init(void); +extern void numa_policy_rebind(const nodemask_t *old, const nodemask_t *new); extern struct mempolicy default_policy; #else @@ -226,6 +227,11 @@ static inline void numa_default_policy(void) { } +static inline void numa_policy_rebind(const nodemask_t *old, + const nodemask_t *new) +{ +} + #endif /* CONFIG_NUMA */ #endif /* __KERNEL__ */ diff --git a/kernel/cpuset.c b/kernel/cpuset.c index 6633f3fb6417..5a737ed9dac7 100644 --- a/kernel/cpuset.c +++ b/kernel/cpuset.c @@ -32,6 +32,7 @@ #include #include #include +#include #include #include #include @@ -600,6 +601,7 @@ static void refresh_mems(void) if (current->cpuset_mems_generation != my_cpusets_mem_gen) { struct cpuset *cs; + nodemask_t oldmem = current->mems_allowed; down(&callback_sem); task_lock(current); @@ -608,6 +610,8 @@ static void refresh_mems(void) current->cpuset_mems_generation = cs->mems_generation; task_unlock(current); up(&callback_sem); + if (!nodes_equal(oldmem, current->mems_allowed)) + numa_policy_rebind(&oldmem, ¤t->mems_allowed); } } diff --git a/mm/mempolicy.c b/mm/mempolicy.c index 2076b1542b8a..5abc57c2b8bd 100644 --- a/mm/mempolicy.c +++ b/mm/mempolicy.c @@ -457,6 +457,7 @@ long do_get_mempolicy(int *policy, nodemask_t *nmask, struct vm_area_struct *vma = NULL; struct mempolicy *pol = current->mempolicy; + cpuset_update_current_mems_allowed(); if (flags & ~(unsigned long)(MPOL_F_NODE|MPOL_F_ADDR)) return -EINVAL; if (flags & MPOL_F_ADDR) { @@ -1206,3 +1207,66 @@ void numa_default_policy(void) { do_set_mempolicy(MPOL_DEFAULT, NULL); } + +/* Migrate a policy to a different set of nodes */ +static void rebind_policy(struct mempolicy *pol, const nodemask_t *old, + const nodemask_t *new) +{ + nodemask_t tmp; + + if (!pol) + return; + + switch (pol->policy) { + case MPOL_DEFAULT: + break; + case MPOL_INTERLEAVE: + nodes_remap(tmp, pol->v.nodes, *old, *new); + pol->v.nodes = tmp; + current->il_next = node_remap(current->il_next, *old, *new); + break; + case MPOL_PREFERRED: + pol->v.preferred_node = node_remap(pol->v.preferred_node, + *old, *new); + break; + case MPOL_BIND: { + nodemask_t nodes; + struct zone **z; + struct zonelist *zonelist; + + nodes_clear(nodes); + for (z = pol->v.zonelist->zones; *z; z++) + node_set((*z)->zone_pgdat->node_id, nodes); + nodes_remap(tmp, nodes, *old, *new); + nodes = tmp; + + zonelist = bind_zonelist(&nodes); + + /* If no mem, then zonelist is NULL and we keep old zonelist. + * If that old zonelist has no remaining mems_allowed nodes, + * then zonelist_policy() will "FALL THROUGH" to MPOL_DEFAULT. + */ + + if (zonelist) { + /* Good - got mem - substitute new zonelist */ + kfree(pol->v.zonelist); + pol->v.zonelist = zonelist; + } + break; + } + default: + BUG(); + break; + } +} + +/* + * Someone moved this task to different nodes. Fixup mempolicies. + * + * TODO - fixup current->mm->vma and shmfs/tmpfs/hugetlbfs policies as well, + * once we have a cpuset mechanism to mark which cpuset subtree is migrating. + */ +void numa_policy_rebind(const nodemask_t *old, const nodemask_t *new) +{ + rebind_policy(current->mempolicy, old, new); +} -- cgit v1.2.3 From 29db9190634067c5a328ee5fcc2890251b836b4b Mon Sep 17 00:00:00 2001 From: David Howells Date: Sun, 30 Oct 2005 15:02:44 -0800 Subject: [PATCH] Keys: Add LSM hooks for key management [try #3] The attached patch adds LSM hooks for key management facilities. The notable changes are: (1) The key struct now supports a security pointer for the use of security modules. This will permit key labelling and restrictions on which programs may access a key. (2) Security modules get a chance to note (or abort) the allocation of a key. (3) The key permission checking can now be enhanced by the security modules; the permissions check consults LSM if all other checks bear out. (4) The key permissions checking functions now return an error code rather than a boolean value. (5) An extra permission has been added to govern the modification of attributes (UID, GID, permissions). Note that there isn't an LSM hook specifically for each keyctl() operation, but rather the permissions hook allows control of individual operations based on the permission request bits. Key management access control through LSM is enabled by automatically if both CONFIG_KEYS and CONFIG_SECURITY are enabled. This should be applied on top of the patch ensubjected: [PATCH] Keys: Possessor permissions should be additive Signed-Off-By: David Howells Signed-off-by: Chris Wright Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- Documentation/keys.txt | 22 +++++++------ include/linux/key-ui.h | 3 +- include/linux/key.h | 13 +++++--- include/linux/security.h | 73 ++++++++++++++++++++++++++++++++++++++++++++ security/dummy.c | 23 ++++++++++++++ security/keys/key.c | 56 +++++++++++++++++++++++---------- security/keys/keyctl.c | 13 +++++--- security/keys/keyring.c | 21 ++++++++----- security/keys/permission.c | 7 ++++- security/keys/process_keys.c | 9 +++--- 10 files changed, 191 insertions(+), 49 deletions(-) (limited to 'include/linux') diff --git a/Documentation/keys.txt b/Documentation/keys.txt index 4afe03a58c5b..31154882000a 100644 --- a/Documentation/keys.txt +++ b/Documentation/keys.txt @@ -196,7 +196,7 @@ KEY ACCESS PERMISSIONS Keys have an owner user ID, a group access ID, and a permissions mask. The mask has up to eight bits each for possessor, user, group and other access. Only -five of each set of eight bits are defined. These permissions granted are: +six of each set of eight bits are defined. These permissions granted are: (*) View @@ -224,6 +224,10 @@ five of each set of eight bits are defined. These permissions granted are: keyring to a key, a process must have Write permission on the keyring and Link permission on the key. + (*) Set Attribute + + This permits a key's UID, GID and permissions mask to be changed. + For changing the ownership, group ID or permissions mask, being the owner of the key or having the sysadmin capability is sufficient. @@ -242,15 +246,15 @@ about the status of the key service: this way: SERIAL FLAGS USAGE EXPY PERM UID GID TYPE DESCRIPTION: SUMMARY - 00000001 I----- 39 perm 1f1f0000 0 0 keyring _uid_ses.0: 1/4 - 00000002 I----- 2 perm 1f1f0000 0 0 keyring _uid.0: empty - 00000007 I----- 1 perm 1f1f0000 0 0 keyring _pid.1: empty - 0000018d I----- 1 perm 1f1f0000 0 0 keyring _pid.412: empty - 000004d2 I--Q-- 1 perm 1f1f0000 32 -1 keyring _uid.32: 1/4 - 000004d3 I--Q-- 3 perm 1f1f0000 32 -1 keyring _uid_ses.32: empty + 00000001 I----- 39 perm 1f3f0000 0 0 keyring _uid_ses.0: 1/4 + 00000002 I----- 2 perm 1f3f0000 0 0 keyring _uid.0: empty + 00000007 I----- 1 perm 1f3f0000 0 0 keyring _pid.1: empty + 0000018d I----- 1 perm 1f3f0000 0 0 keyring _pid.412: empty + 000004d2 I--Q-- 1 perm 1f3f0000 32 -1 keyring _uid.32: 1/4 + 000004d3 I--Q-- 3 perm 1f3f0000 32 -1 keyring _uid_ses.32: empty 00000892 I--QU- 1 perm 1f000000 0 0 user metal:copper: 0 - 00000893 I--Q-N 1 35s 1f1f0000 0 0 user metal:silver: 0 - 00000894 I--Q-- 1 10h 001f0000 0 0 user metal:gold: 0 + 00000893 I--Q-N 1 35s 1f3f0000 0 0 user metal:silver: 0 + 00000894 I--Q-- 1 10h 003f0000 0 0 user metal:gold: 0 The flags are: diff --git a/include/linux/key-ui.h b/include/linux/key-ui.h index 7a2e332067c3..e8b8a7a5c496 100644 --- a/include/linux/key-ui.h +++ b/include/linux/key-ui.h @@ -24,7 +24,8 @@ extern spinlock_t key_serial_lock; #define KEY_WRITE 0x04 /* require permission to update / modify */ #define KEY_SEARCH 0x08 /* require permission to search (keyring) or find (key) */ #define KEY_LINK 0x10 /* require permission to link */ -#define KEY_ALL 0x1f /* all the above permissions */ +#define KEY_SETATTR 0x20 /* require permission to change attributes */ +#define KEY_ALL 0x3f /* all the above permissions */ /* * the keyring payload contains a list of the keys to which the keyring is diff --git a/include/linux/key.h b/include/linux/key.h index f1efa016dbf3..53513a3be53b 100644 --- a/include/linux/key.h +++ b/include/linux/key.h @@ -40,28 +40,32 @@ struct key; #define KEY_POS_WRITE 0x04000000 /* possessor can update key payload / add link to keyring */ #define KEY_POS_SEARCH 0x08000000 /* possessor can find a key in search / search a keyring */ #define KEY_POS_LINK 0x10000000 /* possessor can create a link to a key/keyring */ -#define KEY_POS_ALL 0x1f000000 +#define KEY_POS_SETATTR 0x20000000 /* possessor can set key attributes */ +#define KEY_POS_ALL 0x3f000000 #define KEY_USR_VIEW 0x00010000 /* user permissions... */ #define KEY_USR_READ 0x00020000 #define KEY_USR_WRITE 0x00040000 #define KEY_USR_SEARCH 0x00080000 #define KEY_USR_LINK 0x00100000 -#define KEY_USR_ALL 0x001f0000 +#define KEY_USR_SETATTR 0x00200000 +#define KEY_USR_ALL 0x003f0000 #define KEY_GRP_VIEW 0x00000100 /* group permissions... */ #define KEY_GRP_READ 0x00000200 #define KEY_GRP_WRITE 0x00000400 #define KEY_GRP_SEARCH 0x00000800 #define KEY_GRP_LINK 0x00001000 -#define KEY_GRP_ALL 0x00001f00 +#define KEY_GRP_SETATTR 0x00002000 +#define KEY_GRP_ALL 0x00003f00 #define KEY_OTH_VIEW 0x00000001 /* third party permissions... */ #define KEY_OTH_READ 0x00000002 #define KEY_OTH_WRITE 0x00000004 #define KEY_OTH_SEARCH 0x00000008 #define KEY_OTH_LINK 0x00000010 -#define KEY_OTH_ALL 0x0000001f +#define KEY_OTH_SETATTR 0x00000020 +#define KEY_OTH_ALL 0x0000003f struct seq_file; struct user_struct; @@ -119,6 +123,7 @@ struct key { struct key_type *type; /* type of key */ struct rw_semaphore sem; /* change vs change sem */ struct key_user *user; /* owner of this key */ + void *security; /* security data for this key */ time_t expiry; /* time at which key expires (or 0) */ uid_t uid; gid_t gid; diff --git a/include/linux/security.h b/include/linux/security.h index 607ee209ea3b..f7e0ae018712 100644 --- a/include/linux/security.h +++ b/include/linux/security.h @@ -30,6 +30,7 @@ #include #include #include +#include struct ctl_table; @@ -788,6 +789,27 @@ struct swap_info_struct; * @sk_free_security: * Deallocate security structure. * + * Security hooks affecting all Key Management operations + * + * @key_alloc: + * Permit allocation of a key and assign security data. Note that key does + * not have a serial number assigned at this point. + * @key points to the key. + * Return 0 if permission is granted, -ve error otherwise. + * @key_free: + * Notification of destruction; free security data. + * @key points to the key. + * No return value. + * @key_permission: + * See whether a specific operational right is granted to a process on a + * key. + * @key_ref refers to the key (key pointer + possession attribute bit). + * @context points to the process to provide the context against which to + * evaluate the security data on the key. + * @perm describes the combination of permissions required of this key. + * Return 1 if permission granted, 0 if permission denied and -ve it the + * normal permissions model should be effected. + * * Security hooks affecting all System V IPC operations. * * @ipc_permission: @@ -1216,6 +1238,17 @@ struct security_operations { int (*sk_alloc_security) (struct sock *sk, int family, gfp_t priority); void (*sk_free_security) (struct sock *sk); #endif /* CONFIG_SECURITY_NETWORK */ + + /* key management security hooks */ +#ifdef CONFIG_KEYS + int (*key_alloc)(struct key *key); + void (*key_free)(struct key *key); + int (*key_permission)(key_ref_t key_ref, + struct task_struct *context, + key_perm_t perm); + +#endif /* CONFIG_KEYS */ + }; /* global variables */ @@ -2764,5 +2797,45 @@ static inline void security_sk_free(struct sock *sk) } #endif /* CONFIG_SECURITY_NETWORK */ +#ifdef CONFIG_KEYS +#ifdef CONFIG_SECURITY +static inline int security_key_alloc(struct key *key) +{ + return security_ops->key_alloc(key); +} + +static inline void security_key_free(struct key *key) +{ + security_ops->key_free(key); +} + +static inline int security_key_permission(key_ref_t key_ref, + struct task_struct *context, + key_perm_t perm) +{ + return security_ops->key_permission(key_ref, context, perm); +} + +#else + +static inline int security_key_alloc(struct key *key) +{ + return 0; +} + +static inline void security_key_free(struct key *key) +{ +} + +static inline int security_key_permission(key_ref_t key_ref, + struct task_struct *context, + key_perm_t perm) +{ + return 0; +} + +#endif +#endif /* CONFIG_KEYS */ + #endif /* ! __LINUX_SECURITY_H */ diff --git a/security/dummy.c b/security/dummy.c index 2a0337a52d32..3ca5f2b828a0 100644 --- a/security/dummy.c +++ b/security/dummy.c @@ -803,6 +803,23 @@ static int dummy_setprocattr(struct task_struct *p, char *name, void *value, siz return -EINVAL; } +#ifdef CONFIG_KEYS +static inline int dummy_key_alloc(struct key *key) +{ + return 0; +} + +static inline void dummy_key_free(struct key *key) +{ +} + +static inline int dummy_key_permission(key_ref_t key_ref, + struct task_struct *context, + key_perm_t perm) +{ + return 0; +} +#endif /* CONFIG_KEYS */ struct security_operations dummy_security_ops; @@ -954,5 +971,11 @@ void security_fixup_ops (struct security_operations *ops) set_to_dummy_if_null(ops, sk_alloc_security); set_to_dummy_if_null(ops, sk_free_security); #endif /* CONFIG_SECURITY_NETWORK */ +#ifdef CONFIG_KEYS + set_to_dummy_if_null(ops, key_alloc); + set_to_dummy_if_null(ops, key_free); + set_to_dummy_if_null(ops, key_permission); +#endif /* CONFIG_KEYS */ + } diff --git a/security/keys/key.c b/security/keys/key.c index 2182be9e9309..ccde17aff616 100644 --- a/security/keys/key.c +++ b/security/keys/key.c @@ -1,6 +1,6 @@ /* key.c: basic authentication token and access key management * - * Copyright (C) 2004-5 Red Hat, Inc. All Rights Reserved. + * Copyright (C) 2004 Red Hat, Inc. All Rights Reserved. * Written by David Howells (dhowells@redhat.com) * * This program is free software; you can redistribute it and/or @@ -13,6 +13,7 @@ #include #include #include +#include #include #include #include "internal.h" @@ -253,6 +254,7 @@ struct key *key_alloc(struct key_type *type, const char *desc, struct key_user *user = NULL; struct key *key; size_t desclen, quotalen; + int ret; key = ERR_PTR(-EINVAL); if (!desc || !*desc) @@ -305,6 +307,7 @@ struct key *key_alloc(struct key_type *type, const char *desc, key->flags = 0; key->expiry = 0; key->payload.data = NULL; + key->security = NULL; if (!not_in_quota) key->flags |= 1 << KEY_FLAG_IN_QUOTA; @@ -315,16 +318,34 @@ struct key *key_alloc(struct key_type *type, const char *desc, key->magic = KEY_DEBUG_MAGIC; #endif + /* let the security module know about the key */ + ret = security_key_alloc(key); + if (ret < 0) + goto security_error; + /* publish the key by giving it a serial number */ atomic_inc(&user->nkeys); key_alloc_serial(key); - error: +error: return key; - no_memory_3: +security_error: + kfree(key->description); + kmem_cache_free(key_jar, key); + if (!not_in_quota) { + spin_lock(&user->lock); + user->qnkeys--; + user->qnbytes -= quotalen; + spin_unlock(&user->lock); + } + key_user_put(user); + key = ERR_PTR(ret); + goto error; + +no_memory_3: kmem_cache_free(key_jar, key); - no_memory_2: +no_memory_2: if (!not_in_quota) { spin_lock(&user->lock); user->qnkeys--; @@ -332,11 +353,11 @@ struct key *key_alloc(struct key_type *type, const char *desc, spin_unlock(&user->lock); } key_user_put(user); - no_memory_1: +no_memory_1: key = ERR_PTR(-ENOMEM); goto error; - no_quota: +no_quota: spin_unlock(&user->lock); key_user_put(user); key = ERR_PTR(-EDQUOT); @@ -556,6 +577,8 @@ static void key_cleanup(void *data) key_check(key); + security_key_free(key); + /* deal with the user's key tracking and quota */ if (test_bit(KEY_FLAG_IN_QUOTA, &key->flags)) { spin_lock(&key->user->lock); @@ -700,8 +723,8 @@ static inline key_ref_t __key_update(key_ref_t key_ref, int ret; /* need write permission on the key to update it */ - ret = -EACCES; - if (!key_permission(key_ref, KEY_WRITE)) + ret = key_permission(key_ref, KEY_WRITE); + if (ret < 0) goto error; ret = -EEXIST; @@ -711,7 +734,6 @@ static inline key_ref_t __key_update(key_ref_t key_ref, down_write(&key->sem); ret = key->type->update(key, payload, plen); - if (ret == 0) /* updating a negative key instantiates it */ clear_bit(KEY_FLAG_NEGATIVE, &key->flags); @@ -768,9 +790,11 @@ key_ref_t key_create_or_update(key_ref_t keyring_ref, /* if we're going to allocate a new key, we're going to have * to modify the keyring */ - key_ref = ERR_PTR(-EACCES); - if (!key_permission(keyring_ref, KEY_WRITE)) + ret = key_permission(keyring_ref, KEY_WRITE); + if (ret < 0) { + key_ref = ERR_PTR(ret); goto error_3; + } /* search for an existing key of the same type and description in the * destination keyring @@ -780,8 +804,8 @@ key_ref_t key_create_or_update(key_ref_t keyring_ref, goto found_matching_key; /* decide on the permissions we want */ - perm = KEY_POS_VIEW | KEY_POS_SEARCH | KEY_POS_LINK; - perm |= KEY_USR_VIEW | KEY_USR_SEARCH | KEY_USR_LINK; + perm = KEY_POS_VIEW | KEY_POS_SEARCH | KEY_POS_LINK | KEY_POS_SETATTR; + perm |= KEY_USR_VIEW | KEY_USR_SEARCH | KEY_USR_LINK | KEY_USR_SETATTR; if (ktype->read) perm |= KEY_POS_READ | KEY_USR_READ; @@ -840,16 +864,16 @@ int key_update(key_ref_t key_ref, const void *payload, size_t plen) key_check(key); /* the key must be writable */ - ret = -EACCES; - if (!key_permission(key_ref, KEY_WRITE)) + ret = key_permission(key_ref, KEY_WRITE); + if (ret < 0) goto error; /* attempt to update it if supported */ ret = -EOPNOTSUPP; if (key->type->update) { down_write(&key->sem); - ret = key->type->update(key, payload, plen); + ret = key->type->update(key, payload, plen); if (ret == 0) /* updating a negative key instantiates it */ clear_bit(KEY_FLAG_NEGATIVE, &key->flags); diff --git a/security/keys/keyctl.c b/security/keys/keyctl.c index 4c670ee6acf9..b7a468fabdf9 100644 --- a/security/keys/keyctl.c +++ b/security/keys/keyctl.c @@ -624,8 +624,8 @@ long keyctl_keyring_search(key_serial_t ringid, /* link the resulting key to the destination keyring if we can */ if (dest_ref) { - ret = -EACCES; - if (!key_permission(key_ref, KEY_LINK)) + ret = key_permission(key_ref, KEY_LINK); + if (ret < 0) goto error6; ret = key_link(key_ref_to_ptr(dest_ref), key_ref_to_ptr(key_ref)); @@ -676,8 +676,11 @@ long keyctl_read_key(key_serial_t keyid, char __user *buffer, size_t buflen) key = key_ref_to_ptr(key_ref); /* see if we can read it directly */ - if (key_permission(key_ref, KEY_READ)) + ret = key_permission(key_ref, KEY_READ); + if (ret == 0) goto can_read_key; + if (ret != -EACCES) + goto error; /* we can't; see if it's searchable from this process's keyrings * - we automatically take account of the fact that it may be @@ -726,7 +729,7 @@ long keyctl_chown_key(key_serial_t id, uid_t uid, gid_t gid) if (uid == (uid_t) -1 && gid == (gid_t) -1) goto error; - key_ref = lookup_user_key(NULL, id, 1, 1, 0); + key_ref = lookup_user_key(NULL, id, 1, 1, KEY_SETATTR); if (IS_ERR(key_ref)) { ret = PTR_ERR(key_ref); goto error; @@ -786,7 +789,7 @@ long keyctl_setperm_key(key_serial_t id, key_perm_t perm) if (perm & ~(KEY_POS_ALL | KEY_USR_ALL | KEY_GRP_ALL | KEY_OTH_ALL)) goto error; - key_ref = lookup_user_key(NULL, id, 1, 1, 0); + key_ref = lookup_user_key(NULL, id, 1, 1, KEY_SETATTR); if (IS_ERR(key_ref)) { ret = PTR_ERR(key_ref); goto error; diff --git a/security/keys/keyring.c b/security/keys/keyring.c index 0639396dd441..e1cc4dd79012 100644 --- a/security/keys/keyring.c +++ b/security/keys/keyring.c @@ -13,6 +13,7 @@ #include #include #include +#include #include #include #include @@ -309,7 +310,9 @@ struct key *keyring_alloc(const char *description, uid_t uid, gid_t gid, int ret; keyring = key_alloc(&key_type_keyring, description, - uid, gid, KEY_POS_ALL | KEY_USR_ALL, not_in_quota); + uid, gid, + (KEY_POS_ALL & ~KEY_POS_SETATTR) | KEY_USR_ALL, + not_in_quota); if (!IS_ERR(keyring)) { ret = key_instantiate_and_link(keyring, NULL, 0, dest, NULL); @@ -359,9 +362,11 @@ key_ref_t keyring_search_aux(key_ref_t keyring_ref, key_check(keyring); /* top keyring must have search permission to begin the search */ - key_ref = ERR_PTR(-EACCES); - if (!key_task_permission(keyring_ref, context, KEY_SEARCH)) + err = key_task_permission(keyring_ref, context, KEY_SEARCH); + if (err < 0) { + key_ref = ERR_PTR(err); goto error; + } key_ref = ERR_PTR(-ENOTDIR); if (keyring->type != &key_type_keyring) @@ -402,8 +407,8 @@ descend: continue; /* key must have search permissions */ - if (!key_task_permission(make_key_ref(key, possessed), - context, KEY_SEARCH)) + if (key_task_permission(make_key_ref(key, possessed), + context, KEY_SEARCH) < 0) continue; /* we set a different error code if we find a negative key */ @@ -430,7 +435,7 @@ ascend: continue; if (!key_task_permission(make_key_ref(key, possessed), - context, KEY_SEARCH)) + context, KEY_SEARCH) < 0) continue; /* stack the current position */ @@ -521,7 +526,7 @@ key_ref_t __keyring_search_one(key_ref_t keyring_ref, (!key->type->match || key->type->match(key, description)) && key_permission(make_key_ref(key, possessed), - perm) && + perm) < 0 && !test_bit(KEY_FLAG_REVOKED, &key->flags) ) goto found; @@ -617,7 +622,7 @@ struct key *find_keyring_by_name(const char *name, key_serial_t bound) continue; if (!key_permission(make_key_ref(keyring, 0), - KEY_SEARCH)) + KEY_SEARCH) < 0) continue; /* found a potential candidate, but we still need to diff --git a/security/keys/permission.c b/security/keys/permission.c index 03db073ba45c..e7f579c0eaf5 100644 --- a/security/keys/permission.c +++ b/security/keys/permission.c @@ -10,6 +10,7 @@ */ #include +#include #include "internal.h" /*****************************************************************************/ @@ -63,7 +64,11 @@ use_these_perms: kperm = kperm & perm & KEY_ALL; - return kperm == perm; + if (kperm != perm) + return -EACCES; + + /* let LSM be the final arbiter */ + return security_key_permission(key_ref, context, perm); } /* end key_task_permission() */ diff --git a/security/keys/process_keys.c b/security/keys/process_keys.c index d42d2158ce13..566b1cc0118a 100644 --- a/security/keys/process_keys.c +++ b/security/keys/process_keys.c @@ -39,7 +39,7 @@ struct key root_user_keyring = { .type = &key_type_keyring, .user = &root_key_user, .sem = __RWSEM_INITIALIZER(root_user_keyring.sem), - .perm = KEY_POS_ALL | KEY_USR_ALL, + .perm = (KEY_POS_ALL & ~KEY_POS_SETATTR) | KEY_USR_ALL, .flags = 1 << KEY_FLAG_INSTANTIATED, .description = "_uid.0", #ifdef KEY_DEBUGGING @@ -54,7 +54,7 @@ struct key root_session_keyring = { .type = &key_type_keyring, .user = &root_key_user, .sem = __RWSEM_INITIALIZER(root_session_keyring.sem), - .perm = KEY_POS_ALL | KEY_USR_ALL, + .perm = (KEY_POS_ALL & ~KEY_POS_SETATTR) | KEY_USR_ALL, .flags = 1 << KEY_FLAG_INSTANTIATED, .description = "_uid_ses.0", #ifdef KEY_DEBUGGING @@ -666,9 +666,8 @@ key_ref_t lookup_user_key(struct task_struct *context, key_serial_t id, goto invalid_key; /* check the permissions */ - ret = -EACCES; - - if (!key_task_permission(key_ref, context, perm)) + ret = key_task_permission(key_ref, context, perm); + if (ret < 0) goto invalid_key; error: -- cgit v1.2.3 From 6ea05db06f32adac65b4c4e95ef998839149de12 Mon Sep 17 00:00:00 2001 From: Miklos Szeredi Date: Sun, 30 Oct 2005 15:02:52 -0800 Subject: [PATCH] fuse: remove unused define Setting ctime is implicit in all setattr cases, so the FATTR_CTIME definition is unnecessary. It is used by neither the kernel nor by userspace. Signed-off-by: Miklos Szeredi Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/fuse.h | 1 - 1 file changed, 1 deletion(-) (limited to 'include/linux') diff --git a/include/linux/fuse.h b/include/linux/fuse.h index acbeb96a3353..f98854c2abd7 100644 --- a/include/linux/fuse.h +++ b/include/linux/fuse.h @@ -61,7 +61,6 @@ struct fuse_kstatfs { #define FATTR_SIZE (1 << 3) #define FATTR_ATIME (1 << 4) #define FATTR_MTIME (1 << 5) -#define FATTR_CTIME (1 << 6) /** * Flags returned by the OPEN request -- cgit v1.2.3 From 52303e8b5f8aa234865d40d76ea16b0ff4b27022 Mon Sep 17 00:00:00 2001 From: Pavel Roskin Date: Sun, 30 Oct 2005 15:03:03 -0800 Subject: [PATCH] modules: fix sparse warning for every MODULE_PARM sparse complains about every MODULE_PARM used in a module: warning: symbol '__parm_foo' was not declared. Should it be static? The fix is to split declaration and initialization. While MODULE_PARM is obsolete, it's not something sparse should report. Signed-off-by: Pavel Roskin Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/module.h | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) (limited to 'include/linux') diff --git a/include/linux/module.h b/include/linux/module.h index f05372b7fe77..84d75f3a8aca 100644 --- a/include/linux/module.h +++ b/include/linux/module.h @@ -554,7 +554,9 @@ static inline void MODULE_PARM_(void) { } #ifdef MODULE /* DEPRECATED: Do not use. */ #define MODULE_PARM(var,type) \ -struct obsolete_modparm __parm_##var __attribute__((section("__obsparm"))) = \ +extern struct obsolete_modparm __parm_##var \ +__attribute__((section("__obsparm"))); \ +struct obsolete_modparm __parm_##var = \ { __stringify(var), type, &MODULE_PARM_ }; \ __MODULE_PARM_TYPE(var, type); #else -- cgit v1.2.3 From c0398ee6c2fe3c1743ecc74a6378b67ba0880a4f Mon Sep 17 00:00:00 2001 From: Nikita Danilov Date: Sun, 30 Oct 2005 15:03:10 -0800 Subject: [PATCH] include/linux/kernel.h:BUILD_BUG_ON(): fix a comment Fix comment describing BUILD_BUG_ON: BUG_ON is not an assertion (unfortunately). Signed-off-by: Nikita Danilov Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/kernel.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include/linux') diff --git a/include/linux/kernel.h b/include/linux/kernel.h index 4367ce4db52a..f1925ccc9fe1 100644 --- a/include/linux/kernel.h +++ b/include/linux/kernel.h @@ -307,7 +307,7 @@ struct sysinfo { char _f[20-2*sizeof(long)-sizeof(int)]; /* Padding: libc5 uses this.. */ }; -/* Force a compilation error if condition is false */ +/* Force a compilation error if condition is true */ #define BUILD_BUG_ON(condition) ((void)sizeof(char[1 - 2*!!(condition)])) #ifdef CONFIG_SYSCTL -- cgit v1.2.3 From a241ec65aeac3d69a08a7b153cccbdb7ea35063f Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Sun, 30 Oct 2005 15:03:12 -0800 Subject: [PATCH] RCU torture-testing kernel module This patch is a rewrite of the one submitted on October 1st, using modules (http://marc.theaimsgroup.com/?l=linux-kernel&m=112819093522998&w=2). This rewrite adds a tristate CONFIG_RCU_TORTURE_TEST, which enables an intense torture test of the RCU infratructure. This is needed due to the continued changes to the RCU infrastructure to accommodate dynamic ticks, CPU hotplug, realtime, and so on. Most of the code is in a separate file that is compiled only if the CONFIG variable is set. Documentation on how to run the test and interpret the output is also included. This code has been tested on i386 and ppc64, and an earlier version of the code has received extensive testing on a number of architectures as part of the PREEMPT_RT patchset. Signed-off-by: "Paul E. McKenney" Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- Documentation/RCU/torture.txt | 122 +++++++++++ include/linux/rcupdate.h | 1 + kernel/Makefile | 1 + kernel/rcupdate.c | 10 + kernel/rcutorture.c | 492 ++++++++++++++++++++++++++++++++++++++++++ lib/Kconfig.debug | 21 ++ mm/mmap.c | 2 +- 7 files changed, 648 insertions(+), 1 deletion(-) create mode 100644 Documentation/RCU/torture.txt create mode 100644 kernel/rcutorture.c (limited to 'include/linux') diff --git a/Documentation/RCU/torture.txt b/Documentation/RCU/torture.txt new file mode 100644 index 000000000000..e4c38152f7f7 --- /dev/null +++ b/Documentation/RCU/torture.txt @@ -0,0 +1,122 @@ +RCU Torture Test Operation + + +CONFIG_RCU_TORTURE_TEST + +The CONFIG_RCU_TORTURE_TEST config option is available for all RCU +implementations. It creates an rcutorture kernel module that can +be loaded to run a torture test. The test periodically outputs +status messages via printk(), which can be examined via the dmesg +command (perhaps grepping for "rcutorture"). The test is started +when the module is loaded, and stops when the module is unloaded. + +However, actually setting this config option to "y" results in the system +running the test immediately upon boot, and ending only when the system +is taken down. Normally, one will instead want to build the system +with CONFIG_RCU_TORTURE_TEST=m and to use modprobe and rmmod to control +the test, perhaps using a script similar to the one shown at the end of +this document. Note that you will need CONFIG_MODULE_UNLOAD in order +to be able to end the test. + + +MODULE PARAMETERS + +This module has the following parameters: + +nreaders This is the number of RCU reading threads supported. + The default is twice the number of CPUs. Why twice? + To properly exercise RCU implementations with preemptible + read-side critical sections. + +stat_interval The number of seconds between output of torture + statistics (via printk()). Regardless of the interval, + statistics are printed when the module is unloaded. + Setting the interval to zero causes the statistics to + be printed -only- when the module is unloaded, and this + is the default. + +verbose Enable debug printk()s. Default is disabled. + + +OUTPUT + +The statistics output is as follows: + + rcutorture: --- Start of test: nreaders=16 stat_interval=0 verbose=0 + rcutorture: rtc: 0000000000000000 ver: 1916 tfle: 0 rta: 1916 rtaf: 0 rtf: 1915 + rcutorture: Reader Pipe: 1466408 9747 0 0 0 0 0 0 0 0 0 + rcutorture: Reader Batch: 1464477 11678 0 0 0 0 0 0 0 0 + rcutorture: Free-Block Circulation: 1915 1915 1915 1915 1915 1915 1915 1915 1915 1915 0 + rcutorture: --- End of test + +The command "dmesg | grep rcutorture:" will extract this information on +most systems. On more esoteric configurations, it may be necessary to +use other commands to access the output of the printk()s used by +the RCU torture test. The printk()s use KERN_ALERT, so they should +be evident. ;-) + +The entries are as follows: + +o "ggp": The number of counter flips (or batches) since boot. + +o "rtc": The hexadecimal address of the structure currently visible + to readers. + +o "ver": The number of times since boot that the rcutw writer task + has changed the structure visible to readers. + +o "tfle": If non-zero, indicates that the "torture freelist" + containing structure to be placed into the "rtc" area is empty. + This condition is important, since it can fool you into thinking + that RCU is working when it is not. :-/ + +o "rta": Number of structures allocated from the torture freelist. + +o "rtaf": Number of allocations from the torture freelist that have + failed due to the list being empty. + +o "rtf": Number of frees into the torture freelist. + +o "Reader Pipe": Histogram of "ages" of structures seen by readers. + If any entries past the first two are non-zero, RCU is broken. + And rcutorture prints the error flag string "!!!" to make sure + you notice. The age of a newly allocated structure is zero, + it becomes one when removed from reader visibility, and is + incremented once per grace period subsequently -- and is freed + after passing through (RCU_TORTURE_PIPE_LEN-2) grace periods. + + The output displayed above was taken from a correctly working + RCU. If you want to see what it looks like when broken, break + it yourself. ;-) + +o "Reader Batch": Another histogram of "ages" of structures seen + by readers, but in terms of counter flips (or batches) rather + than in terms of grace periods. The legal number of non-zero + entries is again two. The reason for this separate view is + that it is easier to get the third entry to show up in the + "Reader Batch" list than in the "Reader Pipe" list. + +o "Free-Block Circulation": Shows the number of torture structures + that have reached a given point in the pipeline. The first element + should closely correspond to the number of structures allocated, + the second to the number that have been removed from reader view, + and all but the last remaining to the corresponding number of + passes through a grace period. The last entry should be zero, + as it is only incremented if a torture structure's counter + somehow gets incremented farther than it should. + + +USAGE + +The following script may be used to torture RCU: + + #!/bin/sh + + modprobe rcutorture + sleep 100 + rmmod rcutorture + dmesg | grep rcutorture: + +The output can be manually inspected for the error flag of "!!!". +One could of course create a more elaborate script that automatically +checked for such errors. diff --git a/include/linux/rcupdate.h b/include/linux/rcupdate.h index 70191a5a148f..cce25591eec2 100644 --- a/include/linux/rcupdate.h +++ b/include/linux/rcupdate.h @@ -275,6 +275,7 @@ static inline int rcu_pending(int cpu) extern void rcu_init(void); extern void rcu_check_callbacks(int cpu, int user); extern void rcu_restart_cpu(int cpu); +extern long rcu_batches_completed(void); /* Exported interfaces */ extern void FASTCALL(call_rcu(struct rcu_head *head, diff --git a/kernel/Makefile b/kernel/Makefile index 980b5e454441..4f5a1453093a 100644 --- a/kernel/Makefile +++ b/kernel/Makefile @@ -31,6 +31,7 @@ obj-$(CONFIG_DETECT_SOFTLOCKUP) += softlockup.o obj-$(CONFIG_GENERIC_HARDIRQS) += irq/ obj-$(CONFIG_CRASH_DUMP) += crash_dump.o obj-$(CONFIG_SECCOMP) += seccomp.o +obj-$(CONFIG_RCU_TORTURE_TEST) += rcutorture.o ifneq ($(CONFIG_SCHED_NO_NO_OMIT_FRAME_POINTER),y) # According to Alan Modra , the -fno-omit-frame-pointer is diff --git a/kernel/rcupdate.c b/kernel/rcupdate.c index 2559d4b8f23f..c4d159a21e04 100644 --- a/kernel/rcupdate.c +++ b/kernel/rcupdate.c @@ -153,6 +153,15 @@ void fastcall call_rcu_bh(struct rcu_head *head, local_irq_restore(flags); } +/* + * Return the number of RCU batches processed thus far. Useful + * for debug and statistics. + */ +long rcu_batches_completed(void) +{ + return rcu_ctrlblk.completed; +} + /* * Invoke the completed RCU callbacks. They are expected to be in * a per-cpu list. @@ -501,6 +510,7 @@ void synchronize_kernel(void) } module_param(maxbatch, int, 0); +EXPORT_SYMBOL_GPL(rcu_batches_completed); EXPORT_SYMBOL(call_rcu); /* WARNING: GPL-only in April 2006. */ EXPORT_SYMBOL(call_rcu_bh); /* WARNING: GPL-only in April 2006. */ EXPORT_SYMBOL_GPL(synchronize_rcu); diff --git a/kernel/rcutorture.c b/kernel/rcutorture.c new file mode 100644 index 000000000000..9b58f1eff3ca --- /dev/null +++ b/kernel/rcutorture.c @@ -0,0 +1,492 @@ +/* + * Read-Copy Update /proc-based torture test facility + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. + * + * Copyright (C) IBM Corporation, 2005 + * + * Authors: Paul E. McKenney + * + * See also: Documentation/RCU/torture.txt + */ +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +MODULE_LICENSE("GPL"); + +static int nreaders = -1; /* # reader threads, defaults to 4*ncpus */ +static int stat_interval = 0; /* Interval between stats, in seconds. */ + /* Defaults to "only at end of test". */ +static int verbose = 0; /* Print more debug info. */ + +MODULE_PARM(nreaders, "i"); +MODULE_PARM_DESC(nreaders, "Number of RCU reader threads"); +MODULE_PARM(stat_interval, "i"); +MODULE_PARM_DESC(stat_interval, "Number of seconds between stats printk()s"); +MODULE_PARM(verbose, "i"); +MODULE_PARM_DESC(verbose, "Enable verbose debugging printk()s"); +#define TORTURE_FLAG "rcutorture: " +#define PRINTK_STRING(s) \ + do { printk(KERN_ALERT TORTURE_FLAG s "\n"); } while (0) +#define VERBOSE_PRINTK_STRING(s) \ + do { if (verbose) printk(KERN_ALERT TORTURE_FLAG s "\n"); } while (0) +#define VERBOSE_PRINTK_ERRSTRING(s) \ + do { if (verbose) printk(KERN_ALERT TORTURE_FLAG "!!! " s "\n"); } while (0) + +static char printk_buf[4096]; + +static int nrealreaders; +static struct task_struct *writer_task; +static struct task_struct **reader_tasks; +static struct task_struct *stats_task; + +#define RCU_TORTURE_PIPE_LEN 10 + +struct rcu_torture { + struct rcu_head rtort_rcu; + int rtort_pipe_count; + struct list_head rtort_free; +}; + +static int fullstop = 0; /* stop generating callbacks at test end. */ +static LIST_HEAD(rcu_torture_freelist); +static struct rcu_torture *rcu_torture_current = NULL; +static long rcu_torture_current_version = 0; +static struct rcu_torture rcu_tortures[10 * RCU_TORTURE_PIPE_LEN]; +static DEFINE_SPINLOCK(rcu_torture_lock); +static DEFINE_PER_CPU(long [RCU_TORTURE_PIPE_LEN + 1], rcu_torture_count) = + { 0 }; +static DEFINE_PER_CPU(long [RCU_TORTURE_PIPE_LEN + 1], rcu_torture_batch) = + { 0 }; +static atomic_t rcu_torture_wcount[RCU_TORTURE_PIPE_LEN + 1]; +atomic_t n_rcu_torture_alloc; +atomic_t n_rcu_torture_alloc_fail; +atomic_t n_rcu_torture_free; + +/* + * Allocate an element from the rcu_tortures pool. + */ +struct rcu_torture * +rcu_torture_alloc(void) +{ + struct list_head *p; + + spin_lock(&rcu_torture_lock); + if (list_empty(&rcu_torture_freelist)) { + atomic_inc(&n_rcu_torture_alloc_fail); + spin_unlock(&rcu_torture_lock); + return NULL; + } + atomic_inc(&n_rcu_torture_alloc); + p = rcu_torture_freelist.next; + list_del_init(p); + spin_unlock(&rcu_torture_lock); + return container_of(p, struct rcu_torture, rtort_free); +} + +/* + * Free an element to the rcu_tortures pool. + */ +static void +rcu_torture_free(struct rcu_torture *p) +{ + atomic_inc(&n_rcu_torture_free); + spin_lock(&rcu_torture_lock); + list_add_tail(&p->rtort_free, &rcu_torture_freelist); + spin_unlock(&rcu_torture_lock); +} + +static void +rcu_torture_cb(struct rcu_head *p) +{ + int i; + struct rcu_torture *rp = container_of(p, struct rcu_torture, rtort_rcu); + + if (fullstop) { + /* Test is ending, just drop callbacks on the floor. */ + /* The next initialization will pick up the pieces. */ + return; + } + i = rp->rtort_pipe_count; + if (i > RCU_TORTURE_PIPE_LEN) + i = RCU_TORTURE_PIPE_LEN; + atomic_inc(&rcu_torture_wcount[i]); + if (++rp->rtort_pipe_count >= RCU_TORTURE_PIPE_LEN) + rcu_torture_free(rp); + else + call_rcu(p, rcu_torture_cb); +} + +struct rcu_random_state { + unsigned long rrs_state; + unsigned long rrs_count; +}; + +#define RCU_RANDOM_MULT 39916801 /* prime */ +#define RCU_RANDOM_ADD 479001701 /* prime */ +#define RCU_RANDOM_REFRESH 10000 + +#define DEFINE_RCU_RANDOM(name) struct rcu_random_state name = { 0, 0 } + +/* + * Crude but fast random-number generator. Uses a linear congruential + * generator, with occasional help from get_random_bytes(). + */ +static long +rcu_random(struct rcu_random_state *rrsp) +{ + long refresh; + + if (--rrsp->rrs_count < 0) { + get_random_bytes(&refresh, sizeof(refresh)); + rrsp->rrs_state += refresh; + rrsp->rrs_count = RCU_RANDOM_REFRESH; + } + rrsp->rrs_state = rrsp->rrs_state * RCU_RANDOM_MULT + RCU_RANDOM_ADD; + return swahw32(rrsp->rrs_state); +} + +/* + * RCU torture writer kthread. Repeatedly substitutes a new structure + * for that pointed to by rcu_torture_current, freeing the old structure + * after a series of grace periods (the "pipeline"). + */ +static int +rcu_torture_writer(void *arg) +{ + int i; + long oldbatch = rcu_batches_completed(); + struct rcu_torture *rp; + struct rcu_torture *old_rp; + static DEFINE_RCU_RANDOM(rand); + + VERBOSE_PRINTK_STRING("rcu_torture_writer task started"); + do { + schedule_timeout_uninterruptible(1); + if (rcu_batches_completed() == oldbatch) + continue; + if ((rp = rcu_torture_alloc()) == NULL) + continue; + rp->rtort_pipe_count = 0; + udelay(rcu_random(&rand) & 0x3ff); + old_rp = rcu_torture_current; + rcu_assign_pointer(rcu_torture_current, rp); + smp_wmb(); + if (old_rp != NULL) { + i = old_rp->rtort_pipe_count; + if (i > RCU_TORTURE_PIPE_LEN) + i = RCU_TORTURE_PIPE_LEN; + atomic_inc(&rcu_torture_wcount[i]); + old_rp->rtort_pipe_count++; + call_rcu(&old_rp->rtort_rcu, rcu_torture_cb); + } + rcu_torture_current_version++; + oldbatch = rcu_batches_completed(); + } while (!kthread_should_stop() && !fullstop); + VERBOSE_PRINTK_STRING("rcu_torture_writer task stopping"); + while (!kthread_should_stop()) + schedule_timeout_uninterruptible(1); + return 0; +} + +/* + * RCU torture reader kthread. Repeatedly dereferences rcu_torture_current, + * incrementing the corresponding element of the pipeline array. The + * counter in the element should never be greater than 1, otherwise, the + * RCU implementation is broken. + */ +static int +rcu_torture_reader(void *arg) +{ + int completed; + DEFINE_RCU_RANDOM(rand); + struct rcu_torture *p; + int pipe_count; + + VERBOSE_PRINTK_STRING("rcu_torture_reader task started"); + do { + rcu_read_lock(); + completed = rcu_batches_completed(); + p = rcu_dereference(rcu_torture_current); + if (p == NULL) { + /* Wait for rcu_torture_writer to get underway */ + rcu_read_unlock(); + schedule_timeout_interruptible(HZ); + continue; + } + udelay(rcu_random(&rand) & 0x7f); + preempt_disable(); + pipe_count = p->rtort_pipe_count; + if (pipe_count > RCU_TORTURE_PIPE_LEN) { + /* Should not happen, but... */ + pipe_count = RCU_TORTURE_PIPE_LEN; + } + ++__get_cpu_var(rcu_torture_count)[pipe_count]; + completed = rcu_batches_completed() - completed; + if (completed > RCU_TORTURE_PIPE_LEN) { + /* Should not happen, but... */ + completed = RCU_TORTURE_PIPE_LEN; + } + ++__get_cpu_var(rcu_torture_batch)[completed]; + preempt_enable(); + rcu_read_unlock(); + schedule(); + } while (!kthread_should_stop() && !fullstop); + VERBOSE_PRINTK_STRING("rcu_torture_reader task stopping"); + while (!kthread_should_stop()) + schedule_timeout_uninterruptible(1); + return 0; +} + +/* + * Create an RCU-torture statistics message in the specified buffer. + */ +static int +rcu_torture_printk(char *page) +{ + int cnt = 0; + int cpu; + int i; + long pipesummary[RCU_TORTURE_PIPE_LEN + 1] = { 0 }; + long batchsummary[RCU_TORTURE_PIPE_LEN + 1] = { 0 }; + + for_each_cpu(cpu) { + for (i = 0; i < RCU_TORTURE_PIPE_LEN + 1; i++) { + pipesummary[i] += per_cpu(rcu_torture_count, cpu)[i]; + batchsummary[i] += per_cpu(rcu_torture_batch, cpu)[i]; + } + } + for (i = RCU_TORTURE_PIPE_LEN - 1; i >= 0; i--) { + if (pipesummary[i] != 0) + break; + } + cnt += sprintf(&page[cnt], "rcutorture: "); + cnt += sprintf(&page[cnt], + "rtc: %p ver: %ld tfle: %d rta: %d rtaf: %d rtf: %d", + rcu_torture_current, + rcu_torture_current_version, + list_empty(&rcu_torture_freelist), + atomic_read(&n_rcu_torture_alloc), + atomic_read(&n_rcu_torture_alloc_fail), + atomic_read(&n_rcu_torture_free)); + cnt += sprintf(&page[cnt], "\nrcutorture: "); + if (i > 1) + cnt += sprintf(&page[cnt], "!!! "); + cnt += sprintf(&page[cnt], "Reader Pipe: "); + for (i = 0; i < RCU_TORTURE_PIPE_LEN + 1; i++) + cnt += sprintf(&page[cnt], " %ld", pipesummary[i]); + cnt += sprintf(&page[cnt], "\nrcutorture: "); + cnt += sprintf(&page[cnt], "Reader Batch: "); + for (i = 0; i < RCU_TORTURE_PIPE_LEN; i++) + cnt += sprintf(&page[cnt], " %ld", batchsummary[i]); + cnt += sprintf(&page[cnt], "\nrcutorture: "); + cnt += sprintf(&page[cnt], "Free-Block Circulation: "); + for (i = 0; i < RCU_TORTURE_PIPE_LEN + 1; i++) { + cnt += sprintf(&page[cnt], " %d", + atomic_read(&rcu_torture_wcount[i])); + } + cnt += sprintf(&page[cnt], "\n"); + return cnt; +} + +/* + * Print torture statistics. Caller must ensure that there is only + * one call to this function at a given time!!! This is normally + * accomplished by relying on the module system to only have one copy + * of the module loaded, and then by giving the rcu_torture_stats + * kthread full control (or the init/cleanup functions when rcu_torture_stats + * thread is not running). + */ +static void +rcu_torture_stats_print(void) +{ + int cnt; + + cnt = rcu_torture_printk(printk_buf); + printk(KERN_ALERT "%s", printk_buf); +} + +/* + * Periodically prints torture statistics, if periodic statistics printing + * was specified via the stat_interval module parameter. + * + * No need to worry about fullstop here, since this one doesn't reference + * volatile state or register callbacks. + */ +static int +rcu_torture_stats(void *arg) +{ + VERBOSE_PRINTK_STRING("rcu_torture_stats task started"); + do { + schedule_timeout_interruptible(stat_interval * HZ); + rcu_torture_stats_print(); + } while (!kthread_should_stop()); + VERBOSE_PRINTK_STRING("rcu_torture_stats task stopping"); + return 0; +} + +static void +rcu_torture_cleanup(void) +{ + int i; + + fullstop = 1; + if (writer_task != NULL) { + VERBOSE_PRINTK_STRING("Stopping rcu_torture_writer task"); + kthread_stop(writer_task); + } + writer_task = NULL; + + if (reader_tasks != NULL) { + for (i = 0; i < nrealreaders; i++) { + if (reader_tasks[i] != NULL) { + VERBOSE_PRINTK_STRING( + "Stopping rcu_torture_reader task"); + kthread_stop(reader_tasks[i]); + } + reader_tasks[i] = NULL; + } + kfree(reader_tasks); + reader_tasks = NULL; + } + rcu_torture_current = NULL; + + if (stats_task != NULL) { + VERBOSE_PRINTK_STRING("Stopping rcu_torture_stats task"); + kthread_stop(stats_task); + } + stats_task = NULL; + + /* Wait for all RCU callbacks to fire. */ + + for (i = 0; i < RCU_TORTURE_PIPE_LEN; i++) + synchronize_rcu(); + rcu_torture_stats_print(); /* -After- the stats thread is stopped! */ + PRINTK_STRING("--- End of test"); +} + +static int +rcu_torture_init(void) +{ + int i; + int cpu; + int firsterr = 0; + + /* Process args and tell the world that the torturer is on the job. */ + + if (nreaders >= 0) + nrealreaders = nreaders; + else + nrealreaders = 2 * num_online_cpus(); + printk(KERN_ALERT TORTURE_FLAG + "--- Start of test: nreaders=%d stat_interval=%d verbose=%d\n", + nrealreaders, stat_interval, verbose); + fullstop = 0; + + /* Set up the freelist. */ + + INIT_LIST_HEAD(&rcu_torture_freelist); + for (i = 0; i < sizeof(rcu_tortures) / sizeof(rcu_tortures[0]); i++) { + list_add_tail(&rcu_tortures[i].rtort_free, + &rcu_torture_freelist); + } + + /* Initialize the statistics so that each run gets its own numbers. */ + + rcu_torture_current = NULL; + rcu_torture_current_version = 0; + atomic_set(&n_rcu_torture_alloc, 0); + atomic_set(&n_rcu_torture_alloc_fail, 0); + atomic_set(&n_rcu_torture_free, 0); + for (i = 0; i < RCU_TORTURE_PIPE_LEN + 1; i++) + atomic_set(&rcu_torture_wcount[i], 0); + for_each_cpu(cpu) { + for (i = 0; i < RCU_TORTURE_PIPE_LEN + 1; i++) { + per_cpu(rcu_torture_count, cpu)[i] = 0; + per_cpu(rcu_torture_batch, cpu)[i] = 0; + } + } + + /* Start up the kthreads. */ + + VERBOSE_PRINTK_STRING("Creating rcu_torture_writer task"); + writer_task = kthread_run(rcu_torture_writer, NULL, + "rcu_torture_writer"); + if (IS_ERR(writer_task)) { + firsterr = PTR_ERR(writer_task); + VERBOSE_PRINTK_ERRSTRING("Failed to create writer"); + writer_task = NULL; + goto unwind; + } + reader_tasks = kmalloc(nrealreaders * sizeof(reader_tasks[0]), + GFP_KERNEL); + if (reader_tasks == NULL) { + VERBOSE_PRINTK_ERRSTRING("out of memory"); + firsterr = -ENOMEM; + goto unwind; + } + for (i = 0; i < nrealreaders; i++) { + VERBOSE_PRINTK_STRING("Creating rcu_torture_reader task"); + reader_tasks[i] = kthread_run(rcu_torture_reader, NULL, + "rcu_torture_reader"); + if (IS_ERR(reader_tasks[i])) { + firsterr = PTR_ERR(reader_tasks[i]); + VERBOSE_PRINTK_ERRSTRING("Failed to create reader"); + reader_tasks[i] = NULL; + goto unwind; + } + } + if (stat_interval > 0) { + VERBOSE_PRINTK_STRING("Creating rcu_torture_stats task"); + stats_task = kthread_run(rcu_torture_stats, NULL, + "rcu_torture_stats"); + if (IS_ERR(stats_task)) { + firsterr = PTR_ERR(stats_task); + VERBOSE_PRINTK_ERRSTRING("Failed to create stats"); + stats_task = NULL; + goto unwind; + } + } + return 0; + +unwind: + rcu_torture_cleanup(); + return firsterr; +} + +module_init(rcu_torture_init); +module_exit(rcu_torture_cleanup); diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug index f9306b14d2a6..156822e3cc79 100644 --- a/lib/Kconfig.debug +++ b/lib/Kconfig.debug @@ -168,6 +168,14 @@ config DEBUG_FS If unsure, say N. +config DEBUG_VM + bool "Debug VM" + depends on DEBUG_KERNEL + help + Enable this to debug the virtual-memory system. + + If unsure, say N. + config FRAME_POINTER bool "Compile the kernel with frame pointers" depends on DEBUG_KERNEL && (X86 || CRIS || M68K || M68KNOMMU || FRV || UML) @@ -178,3 +186,16 @@ config FRAME_POINTER some architectures or if you use external debuggers. If you don't debug the kernel, you can say N. +config RCU_TORTURE_TEST + tristate "torture tests for RCU" + depends on DEBUG_KERNEL + default n + help + This option provides a kernel module that runs torture tests + on the RCU infrastructure. The kernel module may be built + after the fact on the running kernel to be tested, if desired. + + Say Y here if you want RCU torture tests to start automatically + at boot time (you probably don't). + Say M if you want the RCU torture tests to build as a module. + Say N if you are unsure. diff --git a/mm/mmap.c b/mm/mmap.c index 5ecc2cf3e1d7..320dda1778c3 100644 --- a/mm/mmap.c +++ b/mm/mmap.c @@ -1840,7 +1840,7 @@ asmlinkage long sys_munmap(unsigned long addr, size_t len) static inline void verify_mm_writelocked(struct mm_struct *mm) { -#ifdef CONFIG_DEBUG_KERNEL +#ifdef CONFIG_DEBUG_VM if (unlikely(down_read_trylock(&mm->mmap_sem))) { WARN_ON(1); up_read(&mm->mmap_sem); -- cgit v1.2.3 From 4098f9918e068e51fed1727f6ba80efcec372378 Mon Sep 17 00:00:00 2001 From: Paul Jackson Date: Sun, 30 Oct 2005 15:03:21 -0800 Subject: [PATCH] sched: hardcode non-smp set_cpus_allowed Simplify the UP (1 CPU) implementatin of set_cpus_allowed. The one CPU is hardcoded to be cpu 0 - so just test for that bit, and avoid having to pick up the cpu_online_map. Also, unexport cpu_online_map: it was only needed for set_cpus_allowed(). Signed-off-by: Paul Jackson Acked-by: Ingo Molnar Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/sched.h | 2 +- kernel/sched.c | 1 - 2 files changed, 1 insertion(+), 2 deletions(-) (limited to 'include/linux') diff --git a/include/linux/sched.h b/include/linux/sched.h index b2d2dc14f0b9..41285a0e7258 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -940,7 +940,7 @@ extern int set_cpus_allowed(task_t *p, cpumask_t new_mask); #else static inline int set_cpus_allowed(task_t *p, cpumask_t new_mask) { - if (!cpus_intersects(new_mask, cpu_online_map)) + if (!cpu_isset(0, new_mask)) return -EINVAL; return 0; } diff --git a/kernel/sched.c b/kernel/sched.c index 4f26c544d02c..340dd238c16d 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -3877,7 +3877,6 @@ EXPORT_SYMBOL(cpu_present_map); #ifndef CONFIG_SMP cpumask_t cpu_online_map = CPU_MASK_ALL; -EXPORT_SYMBOL_GPL(cpu_online_map); cpumask_t cpu_possible_map = CPU_MASK_ALL; #endif -- cgit v1.2.3 From 621d31219d9a788bda924a0613048053f3f5f211 Mon Sep 17 00:00:00 2001 From: Oleg Nesterov Date: Sun, 30 Oct 2005 15:03:45 -0800 Subject: [PATCH] cleanup the usage of SEND_SIG_xxx constants This patch simplifies some checks for magic siginfo values. It should not change the behaviour in any way. Signed-off-by: Oleg Nesterov Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/sched.h | 5 +++++ kernel/signal.c | 18 +++++++----------- security/selinux/hooks.c | 3 +-- 3 files changed, 13 insertions(+), 13 deletions(-) (limited to 'include/linux') diff --git a/include/linux/sched.h b/include/linux/sched.h index 41285a0e7258..03b68a7b4b82 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -1084,6 +1084,11 @@ extern int do_sigaltstack(const stack_t __user *, stack_t __user *, unsigned lon #define SEND_SIG_PRIV ((struct siginfo *) 1) #define SEND_SIG_FORCED ((struct siginfo *) 2) +static inline int is_si_special(const struct siginfo *info) +{ + return info <= SEND_SIG_FORCED; +} + /* True if we are on the alternate signal stack. */ static inline int on_sig_stack(unsigned long sp) diff --git a/kernel/signal.c b/kernel/signal.c index 1f7b2aaa4a39..27533b90c347 100644 --- a/kernel/signal.c +++ b/kernel/signal.c @@ -651,9 +651,7 @@ static int check_kill_permission(int sig, struct siginfo *info, if (!valid_signal(sig)) return error; error = -EPERM; - if ((info == SEND_SIG_NOINFO || - (info != SEND_SIG_PRIV && info != SEND_SIG_FORCED - && SI_FROMUSER(info))) + if ((info == SEND_SIG_NOINFO || (!is_si_special(info) && SI_FROMUSER(info))) && ((sig != SIGCONT) || (current->signal->session != t->signal->session)) && (current->euid ^ t->suid) && (current->euid ^ t->uid) @@ -802,7 +800,7 @@ static int send_signal(int sig, struct siginfo *info, struct task_struct *t, pass on the info struct. */ q = __sigqueue_alloc(t, GFP_ATOMIC, (sig < SIGRTMIN && - (info < SEND_SIG_FORCED || + (is_si_special(info) || info->si_code >= 0))); if (q) { list_add_tail(&q->list, &signals->list); @@ -825,16 +823,14 @@ static int send_signal(int sig, struct siginfo *info, struct task_struct *t, copy_siginfo(&q->info, info); break; } - } else { - if (sig >= SIGRTMIN - && info != SEND_SIG_NOINFO && info != SEND_SIG_PRIV - && info->si_code != SI_USER) + } else if (!is_si_special(info)) { + if (sig >= SIGRTMIN && info->si_code != SI_USER) /* * Queue overflow, abort. We may abort if the signal was rt * and sent by user using something other than kill(). */ return -EAGAIN; - if ((info > SEND_SIG_PRIV) && (info->si_code == SI_TIMER)) + if (info->si_code == SI_TIMER) /* * Set up a return to indicate that we dropped * the signal. @@ -860,7 +856,7 @@ specific_send_sig_info(int sig, struct siginfo *info, struct task_struct *t) BUG(); assert_spin_locked(&t->sighand->siglock); - if ((info > SEND_SIG_FORCED) && (info->si_code == SI_TIMER)) + if (!is_si_special(info) && (info->si_code == SI_TIMER)) /* * Set up a return to indicate that we dropped the signal. */ @@ -1052,7 +1048,7 @@ __group_send_sig_info(int sig, struct siginfo *info, struct task_struct *p) assert_spin_locked(&p->sighand->siglock); handle_stop_signal(sig, p); - if ((info > SEND_SIG_FORCED) && (info->si_code == SI_TIMER)) + if (!is_si_special(info) && (info->si_code == SI_TIMER)) /* * Set up a return to indicate that we dropped the signal. */ diff --git a/security/selinux/hooks.c b/security/selinux/hooks.c index 295ac472faf1..45c41490d521 100644 --- a/security/selinux/hooks.c +++ b/security/selinux/hooks.c @@ -2713,8 +2713,7 @@ static int selinux_task_kill(struct task_struct *p, struct siginfo *info, int si if (rc) return rc; - if (info != SEND_SIG_NOINFO && (info == SEND_SIG_PRIV || - info == SEND_SIG_FORCED || SI_FROMKERNEL(info))) + if (info != SEND_SIG_NOINFO && (is_si_special(info) || SI_FROMKERNEL(info))) return 0; if (!sig) -- cgit v1.2.3 From 4e57b6817880946a3a78d5d8cad1ace363f7e449 Mon Sep 17 00:00:00 2001 From: Tim Schmielau Date: Sun, 30 Oct 2005 15:03:48 -0800 Subject: [PATCH] fix missing includes I recently picked up my older work to remove unnecessary #includes of sched.h, starting from a patch by Dave Jones to not include sched.h from module.h. This reduces the number of indirect includes of sched.h by ~300. Another ~400 pointless direct includes can be removed after this disentangling (patch to follow later). However, quite a few indirect includes need to be fixed up for this. In order to feed the patches through -mm with as little disturbance as possible, I've split out the fixes I accumulated up to now (complete for i386 and x86_64, more archs to follow later) and post them before the real patch. This way this large part of the patch is kept simple with only adding #includes, and all hunks are independent of each other. So if any hunk rejects or gets in the way of other patches, just drop it. My scripts will pick it up again in the next round. Signed-off-by: Tim Schmielau Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- Documentation/firmware_class/firmware_sample_driver.c | 1 + Documentation/firmware_class/firmware_sample_firmware_class.c | 2 ++ arch/arm/common/amba.c | 2 ++ arch/arm/common/scoop.c | 3 +++ arch/arm/kernel/arthur.c | 1 + arch/arm/mach-imx/generic.c | 2 ++ arch/arm/mach-integrator/clock.c | 1 + arch/arm/mach-integrator/integrator_ap.c | 1 + arch/arm/mach-integrator/lm.c | 1 + arch/arm/mach-iop3xx/iq31244-pci.c | 2 ++ arch/arm/mach-iop3xx/iq80321-pci.c | 2 ++ arch/arm/mach-iop3xx/iq80331-pci.c | 2 ++ arch/arm/mach-iop3xx/iq80332-pci.c | 2 ++ arch/arm/mach-pxa/generic.c | 1 + arch/arm/mach-sa1100/generic.c | 1 + arch/arm/mach-versatile/clock.c | 1 + arch/arm/plat-omap/clock.c | 1 + arch/cris/arch-v10/drivers/axisflashmap.c | 1 + arch/cris/arch-v32/drivers/axisflashmap.c | 1 + arch/cris/kernel/time.c | 1 + arch/i386/kernel/cpu/cpufreq/acpi-cpufreq.c | 1 + arch/i386/kernel/cpu/cpufreq/p4-clockmod.c | 1 + arch/i386/kernel/cpu/cpufreq/powernow-k8.c | 1 + arch/i386/kernel/cpu/cpufreq/speedstep-centrino.c | 1 + arch/i386/kernel/cpu/intel_cacheinfo.c | 1 + arch/ia64/kernel/cyclone.c | 1 + arch/m32r/lib/csum_partial_copy.c | 2 +- arch/mips/sgi-ip27/ip27-berr.c | 1 + arch/ppc/syslib/of_device.c | 2 ++ arch/ppc64/kernel/hvcserver.c | 2 ++ arch/ppc64/kernel/of_device.c | 2 ++ arch/ppc64/lib/locks.c | 2 ++ arch/sh/drivers/dma/dma-sysfs.c | 1 + arch/sh/kernel/cpufreq.c | 1 + arch/xtensa/kernel/platform.c | 1 + drivers/acpi/processor_idle.c | 1 + drivers/base/class.c | 1 + drivers/base/platform.c | 1 + drivers/base/sys.c | 1 + drivers/block/cciss_scsi.c | 10 +++++++--- drivers/block/paride/paride.c | 1 + drivers/block/paride/pg.c | 2 ++ drivers/block/paride/pt.c | 1 + drivers/char/agp/ali-agp.c | 1 + drivers/char/agp/amd64-agp.c | 1 + drivers/char/agp/ati-agp.c | 2 ++ drivers/char/agp/i460-agp.c | 2 ++ drivers/char/agp/isoch.c | 1 + drivers/char/agp/sworks-agp.c | 2 ++ drivers/char/drm/drm_sysfs.c | 2 ++ drivers/char/mwave/3780i.c | 2 ++ drivers/char/watchdog/cpu5wdt.c | 1 + drivers/char/watchdog/mixcomwd.c | 2 ++ drivers/char/watchdog/pcwd.c | 2 +- drivers/char/watchdog/sc520_wdt.c | 1 + drivers/char/watchdog/softdog.c | 2 ++ drivers/infiniband/core/cache.c | 1 + drivers/infiniband/core/sa_query.c | 1 + drivers/infiniband/hw/mthca/mthca_av.c | 2 ++ drivers/infiniband/hw/mthca/mthca_mad.c | 3 +++ drivers/infiniband/hw/mthca/mthca_mcg.c | 2 ++ drivers/infiniband/hw/mthca/mthca_profile.c | 2 ++ drivers/infiniband/hw/mthca/mthca_qp.c | 2 ++ drivers/infiniband/hw/mthca/mthca_reset.c | 1 + drivers/infiniband/hw/mthca/mthca_uar.c | 2 ++ drivers/input/gameport/gameport.c | 1 + drivers/input/joystick/a3d.c | 1 + drivers/input/joystick/adi.c | 1 + drivers/input/joystick/analog.c | 1 + drivers/input/joystick/cobra.c | 1 + drivers/input/joystick/gf2k.c | 1 + drivers/input/joystick/grip.c | 1 + drivers/input/joystick/grip_mp.c | 1 + drivers/input/joystick/guillemot.c | 1 + drivers/input/joystick/interact.c | 1 + drivers/input/joystick/joydump.c | 1 + drivers/input/joystick/sidewinder.c | 1 + drivers/input/joystick/tmdc.c | 1 + drivers/input/serio/hp_sdc_mlc.c | 1 + drivers/isdn/capi/capifs.c | 1 + drivers/macintosh/macio_asic.c | 2 ++ drivers/mca/mca-device.c | 1 + drivers/media/common/ir-common.c | 1 + drivers/media/dvb/dvb-core/dvb_ca_en50221.c | 1 + drivers/media/dvb/frontends/bcm3510.c | 3 +++ drivers/media/dvb/frontends/dib3000mb.c | 2 ++ drivers/media/dvb/frontends/dib3000mc.c | 2 ++ drivers/media/dvb/frontends/dvb_dummy_fe.c | 2 ++ drivers/media/dvb/frontends/lgdt330x.c | 2 ++ drivers/media/dvb/frontends/mt312.c | 2 ++ drivers/media/dvb/frontends/mt352.c | 2 ++ drivers/media/dvb/frontends/nxt2002.c | 2 ++ drivers/media/dvb/frontends/or51132.c | 2 ++ drivers/media/dvb/frontends/or51211.c | 2 ++ drivers/media/dvb/frontends/s5h1420.c | 2 ++ drivers/media/dvb/frontends/sp8870.c | 2 ++ drivers/media/dvb/frontends/sp887x.c | 2 ++ drivers/media/dvb/frontends/stv0297.c | 2 ++ drivers/media/dvb/frontends/stv0299.c | 1 + drivers/media/dvb/frontends/tda1004x.c | 4 ++++ drivers/media/dvb/frontends/tda8083.c | 1 + drivers/media/radio/miropcm20-rds.c | 1 + drivers/message/i2o/device.c | 2 ++ drivers/message/i2o/driver.c | 3 +++ drivers/message/i2o/exec-osm.c | 4 ++++ drivers/message/i2o/iop.c | 1 + drivers/mtd/chips/jedec.c | 1 + drivers/mtd/devices/lart.c | 1 + drivers/mtd/devices/phram.c | 1 + drivers/mtd/maps/bast-flash.c | 1 + drivers/mtd/maps/ceiva.c | 1 + drivers/mtd/maps/dc21285.c | 1 + drivers/mtd/maps/dilnetpc.c | 5 ++++- drivers/mtd/maps/epxa10db-flash.c | 5 ++++- drivers/mtd/maps/fortunet.c | 5 ++++- drivers/mtd/maps/ixp2000.c | 6 ++++-- drivers/mtd/maps/ixp4xx.c | 7 +++++-- drivers/mtd/maps/lubbock-flash.c | 3 +++ drivers/mtd/maps/mainstone-flash.c | 3 +++ drivers/mtd/maps/omap-toto-flash.c | 2 +- drivers/mtd/maps/omap_nor.c | 2 ++ drivers/mtd/maps/pci.c | 1 + drivers/mtd/maps/plat-ram.c | 1 + drivers/mtd/maps/tqm8xxl.c | 4 +++- drivers/mtd/mtdblock.c | 1 + drivers/mtd/mtdchar.c | 1 + drivers/mtd/mtdconcat.c | 2 +- drivers/mtd/nand/s3c2410.c | 1 + drivers/pci/hotplug/cpcihp_generic.c | 1 + drivers/pci/hotplug/cpcihp_zt5550.c | 1 + drivers/pci/hotplug/fakephp.c | 2 ++ drivers/pci/hotplug/pciehprm_nonacpi.c | 3 +++ drivers/pci/hotplug/rpadlpar_core.c | 3 +++ drivers/pci/hotplug/rpaphp_pci.c | 4 +++- drivers/pci/hotplug/rpaphp_slot.c | 3 +++ drivers/pci/hotplug/shpchp.h | 2 ++ drivers/pci/hotplug/shpchprm_nonacpi.c | 2 ++ drivers/pci/pci-driver.c | 2 ++ drivers/pci/pci.c | 1 + drivers/pci/pcie/portdrv_core.c | 2 ++ drivers/pci/pcie/portdrv_pci.c | 1 + drivers/pci/rom.c | 1 + drivers/pnp/manager.c | 2 ++ drivers/pnp/pnpbios/rsparser.c | 2 ++ drivers/s390/cio/cmf.c | 3 +++ drivers/s390/cio/device.c | 1 + drivers/s390/cio/device_fsm.c | 2 ++ drivers/scsi/scsi_transport_fc.c | 1 + drivers/scsi/scsi_transport_iscsi.c | 3 +++ drivers/scsi/sym53c8xx_2/sym_hipd.c | 3 +++ drivers/scsi/sym53c8xx_2/sym_hipd.h | 2 ++ drivers/sh/superhyway/superhyway.c | 2 ++ drivers/usb/host/ohci-omap.c | 3 +++ drivers/usb/host/ohci-pci.c | 2 ++ drivers/usb/host/ohci-pxa27x.c | 1 + drivers/w1/w1_family.c | 1 + drivers/zorro/zorro-sysfs.c | 1 + drivers/zorro/zorro.c | 2 ++ fs/filesystems.c | 1 + fs/jffs2/background.c | 1 + fs/jffs2/wbuf.c | 2 ++ include/linux/cpufreq.h | 1 + include/linux/gameport.h | 1 + include/linux/i2c.h | 1 + include/linux/i2o.h | 8 ++++++-- include/linux/kobj_map.h | 2 ++ include/linux/mtd/map.h | 3 +++ include/linux/serial.h | 1 + include/linux/textsearch.h | 1 + include/pcmcia/ss.h | 1 + include/scsi/scsi_cmnd.h | 1 + include/scsi/scsi_transport_fc.h | 1 + kernel/kallsyms.c | 1 + kernel/kprobes.c | 1 + kernel/params.c | 1 + lib/kobject.c | 1 + lib/smp_processor_id.c | 1 + lib/sort.c | 1 + lib/vsprintf.c | 1 + sound/oss/ac97_codec.c | 1 + 180 files changed, 299 insertions(+), 18 deletions(-) (limited to 'include/linux') diff --git a/Documentation/firmware_class/firmware_sample_driver.c b/Documentation/firmware_class/firmware_sample_driver.c index 4bef8c25172c..d3ad2c24490a 100644 --- a/Documentation/firmware_class/firmware_sample_driver.c +++ b/Documentation/firmware_class/firmware_sample_driver.c @@ -13,6 +13,7 @@ #include #include #include +#include #include "linux/firmware.h" diff --git a/Documentation/firmware_class/firmware_sample_firmware_class.c b/Documentation/firmware_class/firmware_sample_firmware_class.c index 09eab2f1b373..57b956aecbc5 100644 --- a/Documentation/firmware_class/firmware_sample_firmware_class.c +++ b/Documentation/firmware_class/firmware_sample_firmware_class.c @@ -14,6 +14,8 @@ #include #include #include +#include +#include #include diff --git a/arch/arm/common/amba.c b/arch/arm/common/amba.c index c6beb751f2a9..e1013112c354 100644 --- a/arch/arm/common/amba.c +++ b/arch/arm/common/amba.c @@ -10,6 +10,8 @@ #include #include #include +#include +#include #include #include diff --git a/arch/arm/common/scoop.c b/arch/arm/common/scoop.c index e8356b76d7c6..4af0cf5f3bfb 100644 --- a/arch/arm/common/scoop.c +++ b/arch/arm/common/scoop.c @@ -12,6 +12,9 @@ */ #include +#include +#include + #include #include diff --git a/arch/arm/kernel/arthur.c b/arch/arm/kernel/arthur.c index a418dad6692c..0ee2e9819631 100644 --- a/arch/arm/kernel/arthur.c +++ b/arch/arm/kernel/arthur.c @@ -18,6 +18,7 @@ #include #include #include +#include #include diff --git a/arch/arm/mach-imx/generic.c b/arch/arm/mach-imx/generic.c index cb14b0682cef..837d7f0bda4c 100644 --- a/arch/arm/mach-imx/generic.c +++ b/arch/arm/mach-imx/generic.c @@ -26,6 +26,8 @@ #include #include #include +#include + #include #include #include diff --git a/arch/arm/mach-integrator/clock.c b/arch/arm/mach-integrator/clock.c index 56200594db3c..73c360685cad 100644 --- a/arch/arm/mach-integrator/clock.c +++ b/arch/arm/mach-integrator/clock.c @@ -13,6 +13,7 @@ #include #include #include +#include #include #include diff --git a/arch/arm/mach-integrator/integrator_ap.c b/arch/arm/mach-integrator/integrator_ap.c index f368b85f0447..764ceb49470a 100644 --- a/arch/arm/mach-integrator/integrator_ap.c +++ b/arch/arm/mach-integrator/integrator_ap.c @@ -30,6 +30,7 @@ #include #include #include +#include /* HZ */ #include #include #include diff --git a/arch/arm/mach-integrator/lm.c b/arch/arm/mach-integrator/lm.c index c5f19d160598..5b41e3a724e1 100644 --- a/arch/arm/mach-integrator/lm.c +++ b/arch/arm/mach-integrator/lm.c @@ -10,6 +10,7 @@ #include #include #include +#include #include diff --git a/arch/arm/mach-iop3xx/iq31244-pci.c b/arch/arm/mach-iop3xx/iq31244-pci.c index f997daa800bf..c6a973ba8fc6 100644 --- a/arch/arm/mach-iop3xx/iq31244-pci.c +++ b/arch/arm/mach-iop3xx/iq31244-pci.c @@ -14,6 +14,8 @@ #include #include #include +#include +#include #include #include diff --git a/arch/arm/mach-iop3xx/iq80321-pci.c b/arch/arm/mach-iop3xx/iq80321-pci.c index 79fea3d20b66..802f6d091b75 100644 --- a/arch/arm/mach-iop3xx/iq80321-pci.c +++ b/arch/arm/mach-iop3xx/iq80321-pci.c @@ -14,6 +14,8 @@ #include #include #include +#include +#include #include #include diff --git a/arch/arm/mach-iop3xx/iq80331-pci.c b/arch/arm/mach-iop3xx/iq80331-pci.c index f37a0e26b466..654e450a1311 100644 --- a/arch/arm/mach-iop3xx/iq80331-pci.c +++ b/arch/arm/mach-iop3xx/iq80331-pci.c @@ -13,6 +13,8 @@ #include #include #include +#include +#include #include #include diff --git a/arch/arm/mach-iop3xx/iq80332-pci.c b/arch/arm/mach-iop3xx/iq80332-pci.c index b9807aa2aade..65951ffe4631 100644 --- a/arch/arm/mach-iop3xx/iq80332-pci.c +++ b/arch/arm/mach-iop3xx/iq80332-pci.c @@ -13,6 +13,8 @@ #include #include #include +#include +#include #include #include diff --git a/arch/arm/mach-pxa/generic.c b/arch/arm/mach-pxa/generic.c index 3248bc9b9495..9c0289333301 100644 --- a/arch/arm/mach-pxa/generic.c +++ b/arch/arm/mach-pxa/generic.c @@ -23,6 +23,7 @@ #include #include #include +#include #include #include diff --git a/arch/arm/mach-sa1100/generic.c b/arch/arm/mach-sa1100/generic.c index f94b0fbcdcc8..83eba8b54816 100644 --- a/arch/arm/mach-sa1100/generic.c +++ b/arch/arm/mach-sa1100/generic.c @@ -17,6 +17,7 @@ #include #include #include +#include /* just for sched_clock() - funny that */ #include #include diff --git a/arch/arm/mach-versatile/clock.c b/arch/arm/mach-versatile/clock.c index 48025c2b9987..b96a2ea15d41 100644 --- a/arch/arm/mach-versatile/clock.c +++ b/arch/arm/mach-versatile/clock.c @@ -13,6 +13,7 @@ #include #include #include +#include #include #include diff --git a/arch/arm/plat-omap/clock.c b/arch/arm/plat-omap/clock.c index 52a58b2da288..a020fe16428f 100644 --- a/arch/arm/plat-omap/clock.c +++ b/arch/arm/plat-omap/clock.c @@ -13,6 +13,7 @@ #include #include #include +#include #include #include diff --git a/arch/cris/arch-v10/drivers/axisflashmap.c b/arch/cris/arch-v10/drivers/axisflashmap.c index 11ab3836aac6..56b038c8d482 100644 --- a/arch/cris/arch-v10/drivers/axisflashmap.c +++ b/arch/cris/arch-v10/drivers/axisflashmap.c @@ -140,6 +140,7 @@ #include #include #include +#include #include #include diff --git a/arch/cris/arch-v32/drivers/axisflashmap.c b/arch/cris/arch-v32/drivers/axisflashmap.c index 78ed52b1cdac..b679f983b90a 100644 --- a/arch/cris/arch-v32/drivers/axisflashmap.c +++ b/arch/cris/arch-v32/drivers/axisflashmap.c @@ -20,6 +20,7 @@ #include #include #include +#include #include #include diff --git a/arch/cris/kernel/time.c b/arch/cris/kernel/time.c index b863815de78d..66ba8898db07 100644 --- a/arch/cris/kernel/time.c +++ b/arch/cris/kernel/time.c @@ -31,6 +31,7 @@ #include #include #include +#include /* just for sched_clock() - funny that */ int have_rtc; /* used to remember if we have an RTC or not */; diff --git a/arch/i386/kernel/cpu/cpufreq/acpi-cpufreq.c b/arch/i386/kernel/cpu/cpufreq/acpi-cpufreq.c index 822c8ce9d1f1..caa9f7711343 100644 --- a/arch/i386/kernel/cpu/cpufreq/acpi-cpufreq.c +++ b/arch/i386/kernel/cpu/cpufreq/acpi-cpufreq.c @@ -32,6 +32,7 @@ #include #include #include +#include /* current */ #include #include #include diff --git a/arch/i386/kernel/cpu/cpufreq/p4-clockmod.c b/arch/i386/kernel/cpu/cpufreq/p4-clockmod.c index aa622d52c6e5..270f2188d68b 100644 --- a/arch/i386/kernel/cpu/cpufreq/p4-clockmod.c +++ b/arch/i386/kernel/cpu/cpufreq/p4-clockmod.c @@ -28,6 +28,7 @@ #include #include #include +#include /* current / set_cpus_allowed() */ #include #include diff --git a/arch/i386/kernel/cpu/cpufreq/powernow-k8.c b/arch/i386/kernel/cpu/cpufreq/powernow-k8.c index 58ca98fdc2ca..2d5c9adba0cd 100644 --- a/arch/i386/kernel/cpu/cpufreq/powernow-k8.c +++ b/arch/i386/kernel/cpu/cpufreq/powernow-k8.c @@ -32,6 +32,7 @@ #include #include #include +#include /* for current / set_cpus_allowed() */ #include #include diff --git a/arch/i386/kernel/cpu/cpufreq/speedstep-centrino.c b/arch/i386/kernel/cpu/cpufreq/speedstep-centrino.c index c397b6220430..1465974256c9 100644 --- a/arch/i386/kernel/cpu/cpufreq/speedstep-centrino.c +++ b/arch/i386/kernel/cpu/cpufreq/speedstep-centrino.c @@ -22,6 +22,7 @@ #include #include #include +#include /* current */ #include #include diff --git a/arch/i386/kernel/cpu/intel_cacheinfo.c b/arch/i386/kernel/cpu/intel_cacheinfo.c index f0839334881c..4dc42a189ae5 100644 --- a/arch/i386/kernel/cpu/intel_cacheinfo.c +++ b/arch/i386/kernel/cpu/intel_cacheinfo.c @@ -11,6 +11,7 @@ #include #include #include +#include #include #include diff --git a/arch/ia64/kernel/cyclone.c b/arch/ia64/kernel/cyclone.c index 768c7e46957c..6ade3790ce07 100644 --- a/arch/ia64/kernel/cyclone.c +++ b/arch/ia64/kernel/cyclone.c @@ -2,6 +2,7 @@ #include #include #include +#include #include /* IBM Summit (EXA) Cyclone counter code*/ diff --git a/arch/m32r/lib/csum_partial_copy.c b/arch/m32r/lib/csum_partial_copy.c index ddb16a83a8ce..3d5f06145854 100644 --- a/arch/m32r/lib/csum_partial_copy.c +++ b/arch/m32r/lib/csum_partial_copy.c @@ -18,10 +18,10 @@ #include #include +#include #include #include -#include #include /* diff --git a/arch/mips/sgi-ip27/ip27-berr.c b/arch/mips/sgi-ip27/ip27-berr.c index e1829a5d3b19..07631a97670b 100644 --- a/arch/mips/sgi-ip27/ip27-berr.c +++ b/arch/mips/sgi-ip27/ip27-berr.c @@ -10,6 +10,7 @@ #include #include #include +#include /* for SIGBUS */ #include #include diff --git a/arch/ppc/syslib/of_device.c b/arch/ppc/syslib/of_device.c index 93c7231ea709..85b821251635 100644 --- a/arch/ppc/syslib/of_device.c +++ b/arch/ppc/syslib/of_device.c @@ -4,6 +4,8 @@ #include #include #include +#include + #include #include diff --git a/arch/ppc64/kernel/hvcserver.c b/arch/ppc64/kernel/hvcserver.c index bde8f42da854..4d584172055a 100644 --- a/arch/ppc64/kernel/hvcserver.c +++ b/arch/ppc64/kernel/hvcserver.c @@ -22,6 +22,8 @@ #include #include #include +#include + #include #include #include diff --git a/arch/ppc64/kernel/of_device.c b/arch/ppc64/kernel/of_device.c index 9f200f0f2ad5..3aabfd0d3dda 100644 --- a/arch/ppc64/kernel/of_device.c +++ b/arch/ppc64/kernel/of_device.c @@ -4,6 +4,8 @@ #include #include #include +#include + #include #include diff --git a/arch/ppc64/lib/locks.c b/arch/ppc64/lib/locks.c index 033643ab69e0..d622c1d58e4e 100644 --- a/arch/ppc64/lib/locks.c +++ b/arch/ppc64/lib/locks.c @@ -17,6 +17,8 @@ #include #include #include +#include + #include #include diff --git a/arch/sh/drivers/dma/dma-sysfs.c b/arch/sh/drivers/dma/dma-sysfs.c index 71a6d4e7809f..6e3b58bd8795 100644 --- a/arch/sh/drivers/dma/dma-sysfs.c +++ b/arch/sh/drivers/dma/dma-sysfs.c @@ -13,6 +13,7 @@ #include #include #include +#include #include static struct sysdev_class dma_sysclass = { diff --git a/arch/sh/kernel/cpufreq.c b/arch/sh/kernel/cpufreq.c index e0b384bef55f..47abf6e49dfb 100644 --- a/arch/sh/kernel/cpufreq.c +++ b/arch/sh/kernel/cpufreq.c @@ -20,6 +20,7 @@ #include #include #include +#include /* set_cpus_allowed() */ #include #include diff --git a/arch/xtensa/kernel/platform.c b/arch/xtensa/kernel/platform.c index 03674daabc66..a17930747f20 100644 --- a/arch/xtensa/kernel/platform.c +++ b/arch/xtensa/kernel/platform.c @@ -18,6 +18,7 @@ #include #include #include +#include /* HZ */ #define _F(r,f,a,b) \ r __platform_##f a b; \ diff --git a/drivers/acpi/processor_idle.c b/drivers/acpi/processor_idle.c index 26a3a4016115..161db4acfb91 100644 --- a/drivers/acpi/processor_idle.c +++ b/drivers/acpi/processor_idle.c @@ -37,6 +37,7 @@ #include #include #include +#include /* need_resched() */ #include #include diff --git a/drivers/base/class.c b/drivers/base/class.c index c3e569730afe..db65fd0babe9 100644 --- a/drivers/base/class.c +++ b/drivers/base/class.c @@ -17,6 +17,7 @@ #include #include #include +#include #include "base.h" #define to_class_attr(_attr) container_of(_attr, struct class_attribute, attr) diff --git a/drivers/base/platform.c b/drivers/base/platform.c index 75ce8711bca5..08d9cc99c7de 100644 --- a/drivers/base/platform.c +++ b/drivers/base/platform.c @@ -16,6 +16,7 @@ #include #include #include +#include #include "base.h" diff --git a/drivers/base/sys.c b/drivers/base/sys.c index 3431eb6004c3..66ed8f2fece5 100644 --- a/drivers/base/sys.c +++ b/drivers/base/sys.c @@ -21,6 +21,7 @@ #include #include #include +#include extern struct subsystem devices_subsys; diff --git a/drivers/block/cciss_scsi.c b/drivers/block/cciss_scsi.c index e183a3ef7839..ec27976a57da 100644 --- a/drivers/block/cciss_scsi.c +++ b/drivers/block/cciss_scsi.c @@ -28,13 +28,17 @@ through the array controller. Note in particular, neither physical nor logical disks are presented through the scsi layer. */ +#include +#include +#include +#include + +#include + #include #include #include #include -#include -#include -#include #include "cciss_scsi.h" diff --git a/drivers/block/paride/paride.c b/drivers/block/paride/paride.c index 1fef136c0e41..ce94aa11f6a7 100644 --- a/drivers/block/paride/paride.c +++ b/drivers/block/paride/paride.c @@ -29,6 +29,7 @@ #include #include #include +#include /* TASK_* */ #ifdef CONFIG_PARPORT_MODULE #define CONFIG_PARPORT diff --git a/drivers/block/paride/pg.c b/drivers/block/paride/pg.c index 82f2d6d2eeef..6f5df0fad703 100644 --- a/drivers/block/paride/pg.c +++ b/drivers/block/paride/pg.c @@ -162,6 +162,8 @@ enum {D_PRT, D_PRO, D_UNI, D_MOD, D_SLV, D_DLY}; #include #include #include +#include /* current, TASK_* */ +#include #include diff --git a/drivers/block/paride/pt.c b/drivers/block/paride/pt.c index 686c95573452..715ae5dc88fb 100644 --- a/drivers/block/paride/pt.c +++ b/drivers/block/paride/pt.c @@ -146,6 +146,7 @@ static int (*drives[4])[6] = {&drive0, &drive1, &drive2, &drive3}; #include #include #include +#include /* current, TASK_*, schedule_timeout() */ #include diff --git a/drivers/char/agp/ali-agp.c b/drivers/char/agp/ali-agp.c index 9c9c9c2247ce..b02fc2267159 100644 --- a/drivers/char/agp/ali-agp.c +++ b/drivers/char/agp/ali-agp.c @@ -7,6 +7,7 @@ #include #include #include +#include /* PAGE_SIZE */ #include "agp.h" #define ALI_AGPCTRL 0xb8 diff --git a/drivers/char/agp/amd64-agp.c b/drivers/char/agp/amd64-agp.c index 0a7624a9b1c1..0e6c3a31d344 100644 --- a/drivers/char/agp/amd64-agp.c +++ b/drivers/char/agp/amd64-agp.c @@ -13,6 +13,7 @@ #include #include #include +#include /* PAGE_SIZE */ #include "agp.h" /* Will need to be increased if AMD64 ever goes >8-way. */ diff --git a/drivers/char/agp/ati-agp.c b/drivers/char/agp/ati-agp.c index e572ced9100a..0b6e72642d6e 100644 --- a/drivers/char/agp/ati-agp.c +++ b/drivers/char/agp/ati-agp.c @@ -6,6 +6,8 @@ #include #include #include +#include +#include #include #include #include "agp.h" diff --git a/drivers/char/agp/i460-agp.c b/drivers/char/agp/i460-agp.c index 94943298c03e..a2d9e5e48bbe 100644 --- a/drivers/char/agp/i460-agp.c +++ b/drivers/char/agp/i460-agp.c @@ -10,6 +10,8 @@ #include #include #include +#include +#include #include #include "agp.h" diff --git a/drivers/char/agp/isoch.c b/drivers/char/agp/isoch.c index c9ac731504f2..40083241804e 100644 --- a/drivers/char/agp/isoch.c +++ b/drivers/char/agp/isoch.c @@ -6,6 +6,7 @@ #include #include #include +#include #include "agp.h" diff --git a/drivers/char/agp/sworks-agp.c b/drivers/char/agp/sworks-agp.c index a9fb12c20eb7..71ea59a1dbeb 100644 --- a/drivers/char/agp/sworks-agp.c +++ b/drivers/char/agp/sworks-agp.c @@ -5,6 +5,8 @@ #include #include #include +#include +#include #include #include "agp.h" diff --git a/drivers/char/drm/drm_sysfs.c b/drivers/char/drm/drm_sysfs.c index 475cc5e555e1..6d3449761914 100644 --- a/drivers/char/drm/drm_sysfs.c +++ b/drivers/char/drm/drm_sysfs.c @@ -15,6 +15,8 @@ #include #include #include +#include +#include #include "drm_core.h" #include "drmP.h" diff --git a/drivers/char/mwave/3780i.c b/drivers/char/mwave/3780i.c index 613aed9e1840..d1fe05e83882 100644 --- a/drivers/char/mwave/3780i.c +++ b/drivers/char/mwave/3780i.c @@ -53,6 +53,8 @@ #include #include #include +#include /* cond_resched() */ + #include #include #include diff --git a/drivers/char/watchdog/cpu5wdt.c b/drivers/char/watchdog/cpu5wdt.c index 2865dac0a813..e75045fe2641 100644 --- a/drivers/char/watchdog/cpu5wdt.c +++ b/drivers/char/watchdog/cpu5wdt.c @@ -28,6 +28,7 @@ #include #include #include +#include #include #include diff --git a/drivers/char/watchdog/mixcomwd.c b/drivers/char/watchdog/mixcomwd.c index 7fc2188386d9..d8dede575402 100644 --- a/drivers/char/watchdog/mixcomwd.c +++ b/drivers/char/watchdog/mixcomwd.c @@ -45,6 +45,8 @@ #include #include #include +#include +#include #include #include diff --git a/drivers/char/watchdog/pcwd.c b/drivers/char/watchdog/pcwd.c index 427ad51b7a35..37c9e13ad3ac 100644 --- a/drivers/char/watchdog/pcwd.c +++ b/drivers/char/watchdog/pcwd.c @@ -66,7 +66,7 @@ #include #include #include - +#include /* TASK_INTERRUPTIBLE, set_current_state() and friends */ #include #include diff --git a/drivers/char/watchdog/sc520_wdt.c b/drivers/char/watchdog/sc520_wdt.c index 72501be79b0c..4ee9974ad8cb 100644 --- a/drivers/char/watchdog/sc520_wdt.c +++ b/drivers/char/watchdog/sc520_wdt.c @@ -63,6 +63,7 @@ #include #include #include +#include #include #include diff --git a/drivers/char/watchdog/softdog.c b/drivers/char/watchdog/softdog.c index 20e5eb8667f2..a91edaf3a350 100644 --- a/drivers/char/watchdog/softdog.c +++ b/drivers/char/watchdog/softdog.c @@ -47,6 +47,8 @@ #include #include #include +#include + #include #define PFX "SoftDog: " diff --git a/drivers/infiniband/core/cache.c b/drivers/infiniband/core/cache.c index f014e639088c..c57a3871184c 100644 --- a/drivers/infiniband/core/cache.c +++ b/drivers/infiniband/core/cache.c @@ -38,6 +38,7 @@ #include #include #include +#include /* INIT_WORK, schedule_work(), flush_scheduled_work() */ #include diff --git a/drivers/infiniband/core/sa_query.c b/drivers/infiniband/core/sa_query.c index 89ce9dc210d4..acda7d63d6fe 100644 --- a/drivers/infiniband/core/sa_query.c +++ b/drivers/infiniband/core/sa_query.c @@ -43,6 +43,7 @@ #include #include #include +#include #include #include diff --git a/drivers/infiniband/hw/mthca/mthca_av.c b/drivers/infiniband/hw/mthca/mthca_av.c index 889e85096736..22fdc446f25c 100644 --- a/drivers/infiniband/hw/mthca/mthca_av.c +++ b/drivers/infiniband/hw/mthca/mthca_av.c @@ -34,6 +34,8 @@ */ #include +#include +#include #include #include diff --git a/drivers/infiniband/hw/mthca/mthca_mad.c b/drivers/infiniband/hw/mthca/mthca_mad.c index 8561b297a19b..1229c604c6e0 100644 --- a/drivers/infiniband/hw/mthca/mthca_mad.c +++ b/drivers/infiniband/hw/mthca/mthca_mad.c @@ -34,6 +34,9 @@ * $Id: mthca_mad.c 1349 2004-12-16 21:09:43Z roland $ */ +#include +#include + #include #include #include diff --git a/drivers/infiniband/hw/mthca/mthca_mcg.c b/drivers/infiniband/hw/mthca/mthca_mcg.c index b47ea7daf088..2fc449da418d 100644 --- a/drivers/infiniband/hw/mthca/mthca_mcg.c +++ b/drivers/infiniband/hw/mthca/mthca_mcg.c @@ -33,6 +33,8 @@ */ #include +#include +#include #include "mthca_dev.h" #include "mthca_cmd.h" diff --git a/drivers/infiniband/hw/mthca/mthca_profile.c b/drivers/infiniband/hw/mthca/mthca_profile.c index 0576056b34f4..bd1338682074 100644 --- a/drivers/infiniband/hw/mthca/mthca_profile.c +++ b/drivers/infiniband/hw/mthca/mthca_profile.c @@ -35,6 +35,8 @@ #include #include +#include +#include #include "mthca_profile.h" diff --git a/drivers/infiniband/hw/mthca/mthca_qp.c b/drivers/infiniband/hw/mthca/mthca_qp.c index 62ff091505da..7c9afde5ace5 100644 --- a/drivers/infiniband/hw/mthca/mthca_qp.c +++ b/drivers/infiniband/hw/mthca/mthca_qp.c @@ -36,6 +36,8 @@ */ #include +#include +#include #include #include diff --git a/drivers/infiniband/hw/mthca/mthca_reset.c b/drivers/infiniband/hw/mthca/mthca_reset.c index 4f995391dd1d..df5e494a9d38 100644 --- a/drivers/infiniband/hw/mthca/mthca_reset.c +++ b/drivers/infiniband/hw/mthca/mthca_reset.c @@ -37,6 +37,7 @@ #include #include #include +#include #include "mthca_dev.h" #include "mthca_cmd.h" diff --git a/drivers/infiniband/hw/mthca/mthca_uar.c b/drivers/infiniband/hw/mthca/mthca_uar.c index 1c8791ded6ff..8e9219842be4 100644 --- a/drivers/infiniband/hw/mthca/mthca_uar.c +++ b/drivers/infiniband/hw/mthca/mthca_uar.c @@ -32,6 +32,8 @@ * $Id$ */ +#include /* PAGE_SHIFT */ + #include "mthca_dev.h" #include "mthca_memfree.h" diff --git a/drivers/input/gameport/gameport.c b/drivers/input/gameport/gameport.c index ab09cf4093e3..0506934244f0 100644 --- a/drivers/input/gameport/gameport.c +++ b/drivers/input/gameport/gameport.c @@ -21,6 +21,7 @@ #include #include #include +#include /* HZ */ /*#include */ diff --git a/drivers/input/joystick/a3d.c b/drivers/input/joystick/a3d.c index bf65430181fa..4571ea3a4b92 100644 --- a/drivers/input/joystick/a3d.c +++ b/drivers/input/joystick/a3d.c @@ -34,6 +34,7 @@ #include #include #include +#include #define DRIVER_DESC "FP-Gaming Assasin 3D joystick driver" diff --git a/drivers/input/joystick/adi.c b/drivers/input/joystick/adi.c index 9d95459f4bcb..704bf70f1db7 100644 --- a/drivers/input/joystick/adi.c +++ b/drivers/input/joystick/adi.c @@ -34,6 +34,7 @@ #include #include #include +#include #define DRIVER_DESC "Logitech ADI joystick family driver" diff --git a/drivers/input/joystick/analog.c b/drivers/input/joystick/analog.c index c75ac6eb1ffb..3121961e3e7c 100644 --- a/drivers/input/joystick/analog.c +++ b/drivers/input/joystick/analog.c @@ -38,6 +38,7 @@ #include #include #include +#include #include #define DRIVER_DESC "Analog joystick and gamepad driver" diff --git a/drivers/input/joystick/cobra.c b/drivers/input/joystick/cobra.c index 9a3dfc724a41..1909f7ef340c 100644 --- a/drivers/input/joystick/cobra.c +++ b/drivers/input/joystick/cobra.c @@ -34,6 +34,7 @@ #include #include #include +#include #define DRIVER_DESC "Creative Labs Blaster GamePad Cobra driver" diff --git a/drivers/input/joystick/gf2k.c b/drivers/input/joystick/gf2k.c index e151f8c5bcb9..8a3ad455eb38 100644 --- a/drivers/input/joystick/gf2k.c +++ b/drivers/input/joystick/gf2k.c @@ -35,6 +35,7 @@ #include #include #include +#include #define DRIVER_DESC "Genius Flight 2000 joystick driver" diff --git a/drivers/input/joystick/grip.c b/drivers/input/joystick/grip.c index e206bb56e53c..a936e7aedb10 100644 --- a/drivers/input/joystick/grip.c +++ b/drivers/input/joystick/grip.c @@ -34,6 +34,7 @@ #include #include #include +#include #define DRIVER_DESC "Gravis GrIP protocol joystick driver" diff --git a/drivers/input/joystick/grip_mp.c b/drivers/input/joystick/grip_mp.c index a0ba93ccac72..51a912222e85 100644 --- a/drivers/input/joystick/grip_mp.c +++ b/drivers/input/joystick/grip_mp.c @@ -19,6 +19,7 @@ #include #include #include +#include #define DRIVER_DESC "Gravis Grip Multiport driver" diff --git a/drivers/input/joystick/guillemot.c b/drivers/input/joystick/guillemot.c index c528473c09d8..6e2c721c26ba 100644 --- a/drivers/input/joystick/guillemot.c +++ b/drivers/input/joystick/guillemot.c @@ -35,6 +35,7 @@ #include #include #include +#include #define DRIVER_DESC "Guillemot Digital joystick driver" diff --git a/drivers/input/joystick/interact.c b/drivers/input/joystick/interact.c index 8511ee7bb263..c4ed01758226 100644 --- a/drivers/input/joystick/interact.c +++ b/drivers/input/joystick/interact.c @@ -38,6 +38,7 @@ #include #include #include +#include #define DRIVER_DESC "InterAct digital joystick driver" diff --git a/drivers/input/joystick/joydump.c b/drivers/input/joystick/joydump.c index 4234ccaf9146..88ec5a918f2e 100644 --- a/drivers/input/joystick/joydump.c +++ b/drivers/input/joystick/joydump.c @@ -34,6 +34,7 @@ #include #include #include +#include #define DRIVER_DESC "Gameport data dumper module" diff --git a/drivers/input/joystick/sidewinder.c b/drivers/input/joystick/sidewinder.c index eaaad45cc750..78dd163cd702 100644 --- a/drivers/input/joystick/sidewinder.c +++ b/drivers/input/joystick/sidewinder.c @@ -33,6 +33,7 @@ #include #include #include +#include #define DRIVER_DESC "Microsoft SideWinder joystick family driver" diff --git a/drivers/input/joystick/tmdc.c b/drivers/input/joystick/tmdc.c index 3a7d1bb46472..60e2aac7d06e 100644 --- a/drivers/input/joystick/tmdc.c +++ b/drivers/input/joystick/tmdc.c @@ -38,6 +38,7 @@ #include #include #include +#include #define DRIVER_DESC "ThrustMaster DirectConnect joystick driver" diff --git a/drivers/input/serio/hp_sdc_mlc.c b/drivers/input/serio/hp_sdc_mlc.c index e3c44ffae674..1c9426fd5205 100644 --- a/drivers/input/serio/hp_sdc_mlc.c +++ b/drivers/input/serio/hp_sdc_mlc.c @@ -40,6 +40,7 @@ #include #include #include +#include #define PREFIX "HP SDC MLC: " diff --git a/drivers/isdn/capi/capifs.c b/drivers/isdn/capi/capifs.c index 3abd7fc6e5ef..7b564c0dd996 100644 --- a/drivers/isdn/capi/capifs.c +++ b/drivers/isdn/capi/capifs.c @@ -15,6 +15,7 @@ #include #include #include +#include /* current */ MODULE_DESCRIPTION("CAPI4Linux: /dev/capi/ filesystem"); MODULE_AUTHOR("Carsten Paeth"); diff --git a/drivers/macintosh/macio_asic.c b/drivers/macintosh/macio_asic.c index 1ee003346923..c34c96d18907 100644 --- a/drivers/macintosh/macio_asic.c +++ b/drivers/macintosh/macio_asic.c @@ -17,6 +17,8 @@ #include #include #include +#include + #include #include #include diff --git a/drivers/mca/mca-device.c b/drivers/mca/mca-device.c index 76d430aa243f..e7adf89fae41 100644 --- a/drivers/mca/mca-device.c +++ b/drivers/mca/mca-device.c @@ -29,6 +29,7 @@ #include #include #include +#include /** * mca_device_read_stored_pos - read POS register from stored data diff --git a/drivers/media/common/ir-common.c b/drivers/media/common/ir-common.c index 06f4d4686a6c..31fccb4f05d6 100644 --- a/drivers/media/common/ir-common.c +++ b/drivers/media/common/ir-common.c @@ -22,6 +22,7 @@ #include #include +#include #include /* -------------------------------------------------------------------------- */ diff --git a/drivers/media/dvb/dvb-core/dvb_ca_en50221.c b/drivers/media/dvb/dvb-core/dvb_ca_en50221.c index 88757e2634e5..2aa767f9bd7d 100644 --- a/drivers/media/dvb/dvb-core/dvb_ca_en50221.c +++ b/drivers/media/dvb/dvb-core/dvb_ca_en50221.c @@ -36,6 +36,7 @@ #include #include #include +#include #include "dvb_ca_en50221.h" #include "dvb_ringbuffer.h" diff --git a/drivers/media/dvb/frontends/bcm3510.c b/drivers/media/dvb/frontends/bcm3510.c index f5fdc5c3e605..f6d4ee78bdd4 100644 --- a/drivers/media/dvb/frontends/bcm3510.c +++ b/drivers/media/dvb/frontends/bcm3510.c @@ -36,6 +36,9 @@ #include #include #include +#include +#include +#include #include "dvb_frontend.h" #include "bcm3510.h" diff --git a/drivers/media/dvb/frontends/dib3000mb.c b/drivers/media/dvb/frontends/dib3000mb.c index 21433e1831e7..6b0553608610 100644 --- a/drivers/media/dvb/frontends/dib3000mb.c +++ b/drivers/media/dvb/frontends/dib3000mb.c @@ -27,6 +27,8 @@ #include #include #include +#include +#include #include "dib3000-common.h" #include "dib3000mb_priv.h" diff --git a/drivers/media/dvb/frontends/dib3000mc.c b/drivers/media/dvb/frontends/dib3000mc.c index 441de665fec3..c024fad17337 100644 --- a/drivers/media/dvb/frontends/dib3000mc.c +++ b/drivers/media/dvb/frontends/dib3000mc.c @@ -26,6 +26,8 @@ #include #include #include +#include +#include #include "dib3000-common.h" #include "dib3000mc_priv.h" diff --git a/drivers/media/dvb/frontends/dvb_dummy_fe.c b/drivers/media/dvb/frontends/dvb_dummy_fe.c index cff93b9d8ab2..794be520d590 100644 --- a/drivers/media/dvb/frontends/dvb_dummy_fe.c +++ b/drivers/media/dvb/frontends/dvb_dummy_fe.c @@ -22,6 +22,8 @@ #include #include #include +#include +#include #include "dvb_frontend.h" #include "dvb_dummy_fe.h" diff --git a/drivers/media/dvb/frontends/lgdt330x.c b/drivers/media/dvb/frontends/lgdt330x.c index 7142b9c51dd2..8dde72bd1046 100644 --- a/drivers/media/dvb/frontends/lgdt330x.c +++ b/drivers/media/dvb/frontends/lgdt330x.c @@ -37,6 +37,8 @@ #include #include #include +#include +#include #include #include "dvb_frontend.h" diff --git a/drivers/media/dvb/frontends/mt312.c b/drivers/media/dvb/frontends/mt312.c index e455aecd76b2..e38454901dd1 100644 --- a/drivers/media/dvb/frontends/mt312.c +++ b/drivers/media/dvb/frontends/mt312.c @@ -29,6 +29,8 @@ #include #include #include +#include +#include #include "dvb_frontend.h" #include "mt312_priv.h" diff --git a/drivers/media/dvb/frontends/mt352.c b/drivers/media/dvb/frontends/mt352.c index cc1bc0edd65e..f0c610f2c2df 100644 --- a/drivers/media/dvb/frontends/mt352.c +++ b/drivers/media/dvb/frontends/mt352.c @@ -35,6 +35,8 @@ #include #include #include +#include +#include #include "dvb_frontend.h" #include "mt352_priv.h" diff --git a/drivers/media/dvb/frontends/nxt2002.c b/drivers/media/dvb/frontends/nxt2002.c index 35a1d60f1927..30786b1911bd 100644 --- a/drivers/media/dvb/frontends/nxt2002.c +++ b/drivers/media/dvb/frontends/nxt2002.c @@ -32,6 +32,8 @@ #include #include #include +#include +#include #include "dvb_frontend.h" #include "nxt2002.h" diff --git a/drivers/media/dvb/frontends/or51132.c b/drivers/media/dvb/frontends/or51132.c index b6d0eecc59eb..817b044c7fd1 100644 --- a/drivers/media/dvb/frontends/or51132.c +++ b/drivers/media/dvb/frontends/or51132.c @@ -36,6 +36,8 @@ #include #include #include +#include +#include #include #include "dvb_frontend.h" diff --git a/drivers/media/dvb/frontends/or51211.c b/drivers/media/dvb/frontends/or51211.c index ad56a9958404..8a9db23dd1b7 100644 --- a/drivers/media/dvb/frontends/or51211.c +++ b/drivers/media/dvb/frontends/or51211.c @@ -34,6 +34,8 @@ #include #include #include +#include +#include #include #include "dvb_frontend.h" diff --git a/drivers/media/dvb/frontends/s5h1420.c b/drivers/media/dvb/frontends/s5h1420.c index c7fe27fd530c..f265418e3261 100644 --- a/drivers/media/dvb/frontends/s5h1420.c +++ b/drivers/media/dvb/frontends/s5h1420.c @@ -26,6 +26,8 @@ Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. #include #include #include +#include +#include #include "dvb_frontend.h" #include "s5h1420.h" diff --git a/drivers/media/dvb/frontends/sp8870.c b/drivers/media/dvb/frontends/sp8870.c index 764a95a2e212..1c6b2e9264bc 100644 --- a/drivers/media/dvb/frontends/sp8870.c +++ b/drivers/media/dvb/frontends/sp8870.c @@ -32,6 +32,8 @@ #include #include #include +#include +#include #include "dvb_frontend.h" #include "sp8870.h" diff --git a/drivers/media/dvb/frontends/sp887x.c b/drivers/media/dvb/frontends/sp887x.c index d868a6927a16..73384e75625e 100644 --- a/drivers/media/dvb/frontends/sp887x.c +++ b/drivers/media/dvb/frontends/sp887x.c @@ -14,6 +14,8 @@ #include #include #include +#include +#include #include "dvb_frontend.h" #include "sp887x.h" diff --git a/drivers/media/dvb/frontends/stv0297.c b/drivers/media/dvb/frontends/stv0297.c index 8d09afd7545d..6122ba754bc5 100644 --- a/drivers/media/dvb/frontends/stv0297.c +++ b/drivers/media/dvb/frontends/stv0297.c @@ -24,6 +24,8 @@ #include #include #include +#include +#include #include "dvb_frontend.h" #include "stv0297.h" diff --git a/drivers/media/dvb/frontends/stv0299.c b/drivers/media/dvb/frontends/stv0299.c index 2d62931f20b5..889d9257215d 100644 --- a/drivers/media/dvb/frontends/stv0299.c +++ b/drivers/media/dvb/frontends/stv0299.c @@ -48,6 +48,7 @@ #include #include #include +#include #include #include "dvb_frontend.h" diff --git a/drivers/media/dvb/frontends/tda1004x.c b/drivers/media/dvb/frontends/tda1004x.c index 74cea9f8d721..3529c618f828 100644 --- a/drivers/media/dvb/frontends/tda1004x.c +++ b/drivers/media/dvb/frontends/tda1004x.c @@ -32,6 +32,10 @@ #include #include #include +#include +#include +#include + #include "dvb_frontend.h" #include "tda1004x.h" diff --git a/drivers/media/dvb/frontends/tda8083.c b/drivers/media/dvb/frontends/tda8083.c index 168e013d23bd..c05cf1861051 100644 --- a/drivers/media/dvb/frontends/tda8083.c +++ b/drivers/media/dvb/frontends/tda8083.c @@ -30,6 +30,7 @@ #include #include #include +#include #include "dvb_frontend.h" #include "tda8083.h" diff --git a/drivers/media/radio/miropcm20-rds.c b/drivers/media/radio/miropcm20-rds.c index df79d5e0aaed..e09214082e01 100644 --- a/drivers/media/radio/miropcm20-rds.c +++ b/drivers/media/radio/miropcm20-rds.c @@ -14,6 +14,7 @@ #include #include #include +#include /* current, TASK_*, schedule_timeout() */ #include #include #include "miropcm20-rds-core.h" diff --git a/drivers/message/i2o/device.c b/drivers/message/i2o/device.c index d9879965eb50..8eb50cdb8ae1 100644 --- a/drivers/message/i2o/device.c +++ b/drivers/message/i2o/device.c @@ -16,6 +16,8 @@ #include #include #include +#include +#include #include "core.h" /** diff --git a/drivers/message/i2o/driver.c b/drivers/message/i2o/driver.c index 0079a4be0af2..0fb9c4e2ad4c 100644 --- a/drivers/message/i2o/driver.c +++ b/drivers/message/i2o/driver.c @@ -17,6 +17,9 @@ #include #include #include +#include +#include +#include #include "core.h" #define OSM_NAME "i2o" diff --git a/drivers/message/i2o/exec-osm.c b/drivers/message/i2o/exec-osm.c index bda2c62648ba..b675b4ebbebd 100644 --- a/drivers/message/i2o/exec-osm.c +++ b/drivers/message/i2o/exec-osm.c @@ -30,6 +30,10 @@ #include #include #include +#include +#include +#include +#include /* HZ */ #include "core.h" #define OSM_NAME "exec-osm" diff --git a/drivers/message/i2o/iop.c b/drivers/message/i2o/iop.c index 361da8d1d5e7..61b837de4b6a 100644 --- a/drivers/message/i2o/iop.c +++ b/drivers/message/i2o/iop.c @@ -28,6 +28,7 @@ #include #include #include +#include #include "core.h" #define OSM_NAME "i2o" diff --git a/drivers/mtd/chips/jedec.c b/drivers/mtd/chips/jedec.c index 62d235a9a4e2..4f6778f3ee3e 100644 --- a/drivers/mtd/chips/jedec.c +++ b/drivers/mtd/chips/jedec.c @@ -17,6 +17,7 @@ #include #include #include +#include #include #include #include diff --git a/drivers/mtd/devices/lart.c b/drivers/mtd/devices/lart.c index dfd335e4a2a8..df987a53ed9c 100644 --- a/drivers/mtd/devices/lart.c +++ b/drivers/mtd/devices/lart.c @@ -44,6 +44,7 @@ #include #include #include +#include #include #ifdef HAVE_PARTITIONS #include diff --git a/drivers/mtd/devices/phram.c b/drivers/mtd/devices/phram.c index a423a382095a..765c0179c8df 100644 --- a/drivers/mtd/devices/phram.c +++ b/drivers/mtd/devices/phram.c @@ -22,6 +22,7 @@ #include #include #include +#include #include #define ERROR(fmt, args...) printk(KERN_ERR "phram: " fmt , ## args) diff --git a/drivers/mtd/maps/bast-flash.c b/drivers/mtd/maps/bast-flash.c index 0ba0ff7d43b9..63104c73ca3c 100644 --- a/drivers/mtd/maps/bast-flash.c +++ b/drivers/mtd/maps/bast-flash.c @@ -33,6 +33,7 @@ #include #include #include +#include #include #include diff --git a/drivers/mtd/maps/ceiva.c b/drivers/mtd/maps/ceiva.c index da8584a662f4..c68b31dc7e6d 100644 --- a/drivers/mtd/maps/ceiva.c +++ b/drivers/mtd/maps/ceiva.c @@ -20,6 +20,7 @@ #include #include #include +#include #include #include diff --git a/drivers/mtd/maps/dc21285.c b/drivers/mtd/maps/dc21285.c index 938c41f2f056..e5b74169fde6 100644 --- a/drivers/mtd/maps/dc21285.c +++ b/drivers/mtd/maps/dc21285.c @@ -13,6 +13,7 @@ #include #include #include +#include #include #include diff --git a/drivers/mtd/maps/dilnetpc.c b/drivers/mtd/maps/dilnetpc.c index 0bc79c93a584..f99519692cb7 100644 --- a/drivers/mtd/maps/dilnetpc.c +++ b/drivers/mtd/maps/dilnetpc.c @@ -30,12 +30,15 @@ #include #include #include -#include +#include + #include #include #include #include +#include + /* ** The DIL/NetPC keeps its BIOS in two distinct flash blocks. ** Destroying any of these blocks transforms the DNPC into diff --git a/drivers/mtd/maps/epxa10db-flash.c b/drivers/mtd/maps/epxa10db-flash.c index ab6dbe2b8cce..1df6188926b3 100644 --- a/drivers/mtd/maps/epxa10db-flash.c +++ b/drivers/mtd/maps/epxa10db-flash.c @@ -27,12 +27,15 @@ #include #include #include -#include +#include + #include #include #include +#include #include + #ifdef CONFIG_EPXA10DB #define BOARD_NAME "EPXA10DB" #else diff --git a/drivers/mtd/maps/fortunet.c b/drivers/mtd/maps/fortunet.c index 068bb6a54520..00f7bbe5479e 100644 --- a/drivers/mtd/maps/fortunet.c +++ b/drivers/mtd/maps/fortunet.c @@ -7,11 +7,14 @@ #include #include #include -#include +#include + #include #include #include +#include + #define MAX_NUM_REGIONS 4 #define MAX_NUM_PARTITIONS 8 diff --git a/drivers/mtd/maps/ixp2000.c b/drivers/mtd/maps/ixp2000.c index a9f86c7fbd52..1e5d6e1d05f3 100644 --- a/drivers/mtd/maps/ixp2000.c +++ b/drivers/mtd/maps/ixp2000.c @@ -22,11 +22,13 @@ #include #include #include +#include +#include +#include + #include #include #include -#include -#include #include #include diff --git a/drivers/mtd/maps/ixp4xx.c b/drivers/mtd/maps/ixp4xx.c index 3fcc32884074..da316e543237 100644 --- a/drivers/mtd/maps/ixp4xx.c +++ b/drivers/mtd/maps/ixp4xx.c @@ -20,11 +20,14 @@ #include #include #include +#include +#include +#include + #include #include #include -#include -#include + #include #include diff --git a/drivers/mtd/maps/lubbock-flash.c b/drivers/mtd/maps/lubbock-flash.c index 1298de475c9a..2337e0c46750 100644 --- a/drivers/mtd/maps/lubbock-flash.c +++ b/drivers/mtd/maps/lubbock-flash.c @@ -15,10 +15,13 @@ #include #include #include +#include + #include #include #include #include + #include #include #include diff --git a/drivers/mtd/maps/mainstone-flash.c b/drivers/mtd/maps/mainstone-flash.c index 87e93fa60588..da0f8a692628 100644 --- a/drivers/mtd/maps/mainstone-flash.c +++ b/drivers/mtd/maps/mainstone-flash.c @@ -16,9 +16,12 @@ #include #include #include +#include + #include #include #include + #include #include #include diff --git a/drivers/mtd/maps/omap-toto-flash.c b/drivers/mtd/maps/omap-toto-flash.c index 496109071cb1..da36e8dddd17 100644 --- a/drivers/mtd/maps/omap-toto-flash.c +++ b/drivers/mtd/maps/omap-toto-flash.c @@ -12,9 +12,9 @@ #include #include #include - #include #include +#include #include #include diff --git a/drivers/mtd/maps/omap_nor.c b/drivers/mtd/maps/omap_nor.c index b17bca657daf..fa84566245a7 100644 --- a/drivers/mtd/maps/omap_nor.c +++ b/drivers/mtd/maps/omap_nor.c @@ -36,6 +36,8 @@ #include #include #include +#include + #include #include #include diff --git a/drivers/mtd/maps/pci.c b/drivers/mtd/maps/pci.c index 18dbd3af1eaa..d9c64e99ee32 100644 --- a/drivers/mtd/maps/pci.c +++ b/drivers/mtd/maps/pci.c @@ -17,6 +17,7 @@ #include #include #include +#include #include #include diff --git a/drivers/mtd/maps/plat-ram.c b/drivers/mtd/maps/plat-ram.c index 118b04544cad..a0577ea00c3c 100644 --- a/drivers/mtd/maps/plat-ram.c +++ b/drivers/mtd/maps/plat-ram.c @@ -30,6 +30,7 @@ #include #include #include +#include #include #include diff --git a/drivers/mtd/maps/tqm8xxl.c b/drivers/mtd/maps/tqm8xxl.c index 995e9991cb8d..4e28b977f224 100644 --- a/drivers/mtd/maps/tqm8xxl.c +++ b/drivers/mtd/maps/tqm8xxl.c @@ -27,12 +27,14 @@ #include #include #include -#include +#include #include #include #include +#include + #define FLASH_ADDR 0x40000000 #define FLASH_SIZE 0x00800000 #define FLASH_BANK_MAX 4 diff --git a/drivers/mtd/mtdblock.c b/drivers/mtd/mtdblock.c index b7c32c242bc7..400dd9c89883 100644 --- a/drivers/mtd/mtdblock.c +++ b/drivers/mtd/mtdblock.c @@ -15,6 +15,7 @@ #include #include #include +#include /* TASK_* */ #include #include diff --git a/drivers/mtd/mtdchar.c b/drivers/mtd/mtdchar.c index c534fd5d95cb..16df1e4fb0e9 100644 --- a/drivers/mtd/mtdchar.c +++ b/drivers/mtd/mtdchar.c @@ -13,6 +13,7 @@ #include #include #include +#include /* TASK_* */ #include #include diff --git a/drivers/mtd/mtdconcat.c b/drivers/mtd/mtdconcat.c index 8f66d093c80d..f3e65af33a9c 100644 --- a/drivers/mtd/mtdconcat.c +++ b/drivers/mtd/mtdconcat.c @@ -14,7 +14,7 @@ #include #include #include - +#include /* TASK_* */ #include #include diff --git a/drivers/mtd/nand/s3c2410.c b/drivers/mtd/nand/s3c2410.c index b47ebcb31e0f..b58ba236a9eb 100644 --- a/drivers/mtd/nand/s3c2410.c +++ b/drivers/mtd/nand/s3c2410.c @@ -51,6 +51,7 @@ #include #include #include +#include #include #include diff --git a/drivers/pci/hotplug/cpcihp_generic.c b/drivers/pci/hotplug/cpcihp_generic.c index a62a4345b466..2d4639d6841f 100644 --- a/drivers/pci/hotplug/cpcihp_generic.c +++ b/drivers/pci/hotplug/cpcihp_generic.c @@ -39,6 +39,7 @@ #include #include #include +#include #include "cpci_hotplug.h" #define DRIVER_VERSION "0.1" diff --git a/drivers/pci/hotplug/cpcihp_zt5550.c b/drivers/pci/hotplug/cpcihp_zt5550.c index 790abadd816c..f7cb00da38df 100644 --- a/drivers/pci/hotplug/cpcihp_zt5550.c +++ b/drivers/pci/hotplug/cpcihp_zt5550.c @@ -36,6 +36,7 @@ #include #include #include +#include /* SA_SHIRQ */ #include "cpci_hotplug.h" #include "cpcihp_zt5550.h" diff --git a/drivers/pci/hotplug/fakephp.c b/drivers/pci/hotplug/fakephp.c index 8e47fa66e25e..060d74775d7b 100644 --- a/drivers/pci/hotplug/fakephp.c +++ b/drivers/pci/hotplug/fakephp.c @@ -37,6 +37,8 @@ #include #include #include +#include +#include #include "pci_hotplug.h" #include "../pci.h" diff --git a/drivers/pci/hotplug/pciehprm_nonacpi.c b/drivers/pci/hotplug/pciehprm_nonacpi.c index 3622965f8961..33b2c69a0829 100644 --- a/drivers/pci/hotplug/pciehprm_nonacpi.c +++ b/drivers/pci/hotplug/pciehprm_nonacpi.c @@ -33,10 +33,13 @@ #include #include #include +#include + #include #ifdef CONFIG_IA64 #include #endif + #include "pciehp.h" #include "pciehprm.h" #include "pciehprm_nonacpi.h" diff --git a/drivers/pci/hotplug/rpadlpar_core.c b/drivers/pci/hotplug/rpadlpar_core.c index ad1017da8656..fcb66b9a0e28 100644 --- a/drivers/pci/hotplug/rpadlpar_core.c +++ b/drivers/pci/hotplug/rpadlpar_core.c @@ -16,10 +16,13 @@ */ #include #include +#include + #include #include #include #include + #include "../pci.h" #include "rpaphp.h" #include "rpadlpar.h" diff --git a/drivers/pci/hotplug/rpaphp_pci.c b/drivers/pci/hotplug/rpaphp_pci.c index 46c157d26a2f..f7c12d7dfcfc 100644 --- a/drivers/pci/hotplug/rpaphp_pci.c +++ b/drivers/pci/hotplug/rpaphp_pci.c @@ -23,11 +23,13 @@ * */ #include +#include + #include #include #include -#include "../pci.h" /* for pci_add_new_bus */ +#include "../pci.h" /* for pci_add_new_bus */ #include "rpaphp.h" static struct pci_bus *find_bus_among_children(struct pci_bus *bus, diff --git a/drivers/pci/hotplug/rpaphp_slot.c b/drivers/pci/hotplug/rpaphp_slot.c index 0e8815495083..daa89ae57123 100644 --- a/drivers/pci/hotplug/rpaphp_slot.c +++ b/drivers/pci/hotplug/rpaphp_slot.c @@ -27,6 +27,9 @@ #include #include #include +#include +#include + #include #include "rpaphp.h" diff --git a/drivers/pci/hotplug/shpchp.h b/drivers/pci/hotplug/shpchp.h index abe2cf411e68..08ad26a0cae7 100644 --- a/drivers/pci/hotplug/shpchp.h +++ b/drivers/pci/hotplug/shpchp.h @@ -32,6 +32,8 @@ #include #include #include +#include /* signal_pending(), struct timer_list */ + #include "pci_hotplug.h" #if !defined(MODULE) diff --git a/drivers/pci/hotplug/shpchprm_nonacpi.c b/drivers/pci/hotplug/shpchprm_nonacpi.c index d70fe5408417..c6b40998eeb3 100644 --- a/drivers/pci/hotplug/shpchprm_nonacpi.c +++ b/drivers/pci/hotplug/shpchprm_nonacpi.c @@ -32,6 +32,8 @@ #include #include #include +#include + #include "shpchp.h" int shpchprm_get_physical_slot_number(struct controller *ctrl, u32 *sun, u8 busnum, u8 devnum) diff --git a/drivers/pci/pci-driver.c b/drivers/pci/pci-driver.c index 8972e6a3aac0..ae986e590b48 100644 --- a/drivers/pci/pci-driver.c +++ b/drivers/pci/pci-driver.c @@ -8,6 +8,8 @@ #include #include #include +#include +#include #include "pci.h" /* diff --git a/drivers/pci/pci.c b/drivers/pci/pci.c index 61b855c99e39..e74d75843047 100644 --- a/drivers/pci/pci.c +++ b/drivers/pci/pci.c @@ -15,6 +15,7 @@ #include #include #include +#include #include /* isa_dma_bridge_buggy */ #include "pci.h" diff --git a/drivers/pci/pcie/portdrv_core.c b/drivers/pci/pcie/portdrv_core.c index 14f05d22bb70..467a4ceccf10 100644 --- a/drivers/pci/pcie/portdrv_core.c +++ b/drivers/pci/pcie/portdrv_core.c @@ -11,6 +11,8 @@ #include #include #include +#include +#include #include #include "portdrv.h" diff --git a/drivers/pci/pcie/portdrv_pci.c b/drivers/pci/pcie/portdrv_pci.c index 3c565ce7f77b..02260141dc81 100644 --- a/drivers/pci/pcie/portdrv_pci.c +++ b/drivers/pci/pcie/portdrv_pci.c @@ -12,6 +12,7 @@ #include #include #include +#include #include #include "portdrv.h" diff --git a/drivers/pci/rom.c b/drivers/pci/rom.c index 49bd21702314..598a115cd00e 100644 --- a/drivers/pci/rom.c +++ b/drivers/pci/rom.c @@ -9,6 +9,7 @@ #include #include #include +#include #include "pci.h" diff --git a/drivers/pnp/manager.c b/drivers/pnp/manager.c index 94442ffd4aed..cbb2749db178 100644 --- a/drivers/pnp/manager.c +++ b/drivers/pnp/manager.c @@ -12,6 +12,8 @@ #include #include #include +#include +#include #include "base.h" DECLARE_MUTEX(pnp_res_mutex); diff --git a/drivers/pnp/pnpbios/rsparser.c b/drivers/pnp/pnpbios/rsparser.c index b0ca65b68645..5e38cd7335f7 100644 --- a/drivers/pnp/pnpbios/rsparser.c +++ b/drivers/pnp/pnpbios/rsparser.c @@ -7,6 +7,8 @@ #include #include #include +#include +#include #ifdef CONFIG_PCI #include diff --git a/drivers/s390/cio/cmf.c b/drivers/s390/cio/cmf.c index 8cc4f1a940dc..c05b069c2996 100644 --- a/drivers/s390/cio/cmf.c +++ b/drivers/s390/cio/cmf.c @@ -30,10 +30,13 @@ #include #include #include +#include +#include /* get_clock() */ #include #include #include +#include #include "cio.h" #include "css.h" diff --git a/drivers/s390/cio/device.c b/drivers/s390/cio/device.c index 1c2659766c09..811c9d150637 100644 --- a/drivers/s390/cio/device.c +++ b/drivers/s390/cio/device.c @@ -22,6 +22,7 @@ #include #include +#include /* HZ */ #include "cio.h" #include "css.h" diff --git a/drivers/s390/cio/device_fsm.c b/drivers/s390/cio/device_fsm.c index fbe4202a3f6f..c1c89f4fd4e3 100644 --- a/drivers/s390/cio/device_fsm.c +++ b/drivers/s390/cio/device_fsm.c @@ -11,6 +11,8 @@ #include #include #include +#include +#include #include #include diff --git a/drivers/scsi/scsi_transport_fc.c b/drivers/scsi/scsi_transport_fc.c index 771e97ef136e..b856e140e65f 100644 --- a/drivers/scsi/scsi_transport_fc.c +++ b/drivers/scsi/scsi_transport_fc.c @@ -26,6 +26,7 @@ */ #include #include +#include /* workqueue stuff, HZ */ #include #include #include diff --git a/drivers/scsi/scsi_transport_iscsi.c b/drivers/scsi/scsi_transport_iscsi.c index 8bb8222ea589..d2caa35059d9 100644 --- a/drivers/scsi/scsi_transport_iscsi.c +++ b/drivers/scsi/scsi_transport_iscsi.c @@ -19,6 +19,9 @@ * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ #include +#include +#include + #include #include #include diff --git a/drivers/scsi/sym53c8xx_2/sym_hipd.c b/drivers/scsi/sym53c8xx_2/sym_hipd.c index e753ba27dc59..a1a58e1d5ad3 100644 --- a/drivers/scsi/sym53c8xx_2/sym_hipd.c +++ b/drivers/scsi/sym53c8xx_2/sym_hipd.c @@ -37,6 +37,9 @@ * along with this program; if not, write to the Free Software * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */ + +#include + #include "sym_glue.h" #include "sym_nvram.h" diff --git a/drivers/scsi/sym53c8xx_2/sym_hipd.h b/drivers/scsi/sym53c8xx_2/sym_hipd.h index 3131a6bf7ab7..3a264a408216 100644 --- a/drivers/scsi/sym53c8xx_2/sym_hipd.h +++ b/drivers/scsi/sym53c8xx_2/sym_hipd.h @@ -37,6 +37,8 @@ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */ +#include + #ifndef SYM_HIPD_H #define SYM_HIPD_H diff --git a/drivers/sh/superhyway/superhyway.c b/drivers/sh/superhyway/superhyway.c index f056276b08a1..28757cb9d246 100644 --- a/drivers/sh/superhyway/superhyway.c +++ b/drivers/sh/superhyway/superhyway.c @@ -16,6 +16,8 @@ #include #include #include +#include +#include static int superhyway_devices; diff --git a/drivers/usb/host/ohci-omap.c b/drivers/usb/host/ohci-omap.c index 45efeed1fcc3..49815ec4b842 100644 --- a/drivers/usb/host/ohci-omap.c +++ b/drivers/usb/host/ohci-omap.c @@ -14,6 +14,9 @@ * This file is licenced under the GPL. */ +#include /* SA_INTERRUPT */ +#include + #include #include #include diff --git a/drivers/usb/host/ohci-pci.c b/drivers/usb/host/ohci-pci.c index bf1d5ab4aa3a..7ce1d9ef0289 100644 --- a/drivers/usb/host/ohci-pci.c +++ b/drivers/usb/host/ohci-pci.c @@ -14,6 +14,8 @@ * This file is licenced under the GPL. */ +#include + #ifdef CONFIG_PPC_PMAC #include #include diff --git a/drivers/usb/host/ohci-pxa27x.c b/drivers/usb/host/ohci-pxa27x.c index d287dcccd415..f4a4aeda40b7 100644 --- a/drivers/usb/host/ohci-pxa27x.c +++ b/drivers/usb/host/ohci-pxa27x.c @@ -20,6 +20,7 @@ */ #include +#include #include #include #include diff --git a/drivers/w1/w1_family.c b/drivers/w1/w1_family.c index 88c517a4c178..9e293e139a0e 100644 --- a/drivers/w1/w1_family.c +++ b/drivers/w1/w1_family.c @@ -21,6 +21,7 @@ #include #include +#include /* schedule_timeout() */ #include #include "w1_family.h" diff --git a/drivers/zorro/zorro-sysfs.c b/drivers/zorro/zorro-sysfs.c index 04ca8840acf1..87c29d7b6c17 100644 --- a/drivers/zorro/zorro-sysfs.c +++ b/drivers/zorro/zorro-sysfs.c @@ -14,6 +14,7 @@ #include #include #include +#include #include "zorro.h" diff --git a/drivers/zorro/zorro.c b/drivers/zorro/zorro.c index d3c05dfe20d2..0f2b40605b06 100644 --- a/drivers/zorro/zorro.c +++ b/drivers/zorro/zorro.c @@ -16,6 +16,8 @@ #include #include #include +#include + #include #include diff --git a/fs/filesystems.c b/fs/filesystems.c index 44082bfdfec9..9f1072836c8e 100644 --- a/fs/filesystems.c +++ b/fs/filesystems.c @@ -12,6 +12,7 @@ #include #include #include +#include /* for 'current' */ #include /* diff --git a/fs/jffs2/background.c b/fs/jffs2/background.c index 0f224384f176..8210ac16a368 100644 --- a/fs/jffs2/background.c +++ b/fs/jffs2/background.c @@ -15,6 +15,7 @@ #include #include #include +#include #include "nodelist.h" diff --git a/fs/jffs2/wbuf.c b/fs/jffs2/wbuf.c index 996d922e503e..316133c626b7 100644 --- a/fs/jffs2/wbuf.c +++ b/fs/jffs2/wbuf.c @@ -18,6 +18,8 @@ #include #include #include +#include + #include "nodelist.h" /* For testing write failures */ diff --git a/include/linux/cpufreq.h b/include/linux/cpufreq.h index ff7f80f48df1..d068176b7ad7 100644 --- a/include/linux/cpufreq.h +++ b/include/linux/cpufreq.h @@ -23,6 +23,7 @@ #include #include #include +#include #define CPUFREQ_NAME_LEN 16 diff --git a/include/linux/gameport.h b/include/linux/gameport.h index cd623eccdbea..2401dea2b867 100644 --- a/include/linux/gameport.h +++ b/include/linux/gameport.h @@ -12,6 +12,7 @@ #include #include #include +#include struct gameport { diff --git a/include/linux/i2c.h b/include/linux/i2c.h index f88577ca3b3a..5e19a7ba69b2 100644 --- a/include/linux/i2c.h +++ b/include/linux/i2c.h @@ -31,6 +31,7 @@ #include #include #include /* for struct device */ +#include /* for completion */ #include /* --- For i2c-isa ---------------------------------------------------- */ diff --git a/include/linux/i2o.h b/include/linux/i2o.h index 92300325dbcd..d79c8a4bc4f8 100644 --- a/include/linux/i2o.h +++ b/include/linux/i2o.h @@ -25,10 +25,14 @@ /* How many different OSM's are we allowing */ #define I2O_MAX_DRIVERS 8 -#include -#include /* Needed for MUTEX init macros */ #include #include +#include +#include +#include /* work_struct */ + +#include +#include /* Needed for MUTEX init macros */ /* message queue empty */ #define I2O_QUEUE_EMPTY 0xffffffff diff --git a/include/linux/kobj_map.h b/include/linux/kobj_map.h index b6cc10bf8dfc..cbe7d8008042 100644 --- a/include/linux/kobj_map.h +++ b/include/linux/kobj_map.h @@ -1,5 +1,7 @@ #ifdef __KERNEL__ +#include + typedef struct kobject *kobj_probe_t(dev_t, int *, void *); struct kobj_map; diff --git a/include/linux/mtd/map.h b/include/linux/mtd/map.h index 142963f01d29..fc28841f3409 100644 --- a/include/linux/mtd/map.h +++ b/include/linux/mtd/map.h @@ -8,7 +8,10 @@ #include #include #include +#include + #include + #include #include #include diff --git a/include/linux/serial.h b/include/linux/serial.h index 12cd9cf65e8f..33fc8cb8ddfb 100644 --- a/include/linux/serial.h +++ b/include/linux/serial.h @@ -11,6 +11,7 @@ #define _LINUX_SERIAL_H #ifdef __KERNEL__ +#include #include /* diff --git a/include/linux/textsearch.h b/include/linux/textsearch.h index fc5bb4e91a58..7dac8f04d28e 100644 --- a/include/linux/textsearch.h +++ b/include/linux/textsearch.h @@ -8,6 +8,7 @@ #include #include #include +#include struct ts_config; diff --git a/include/pcmcia/ss.h b/include/pcmcia/ss.h index c8592c7e8eaa..e788bbc5657d 100644 --- a/include/pcmcia/ss.h +++ b/include/pcmcia/ss.h @@ -17,6 +17,7 @@ #include #include +#include /* task_struct, completion */ #include #include diff --git a/include/scsi/scsi_cmnd.h b/include/scsi/scsi_cmnd.h index e6b61fab66dd..7529f4388bb4 100644 --- a/include/scsi/scsi_cmnd.h +++ b/include/scsi/scsi_cmnd.h @@ -4,6 +4,7 @@ #include #include #include +#include struct request; struct scatterlist; diff --git a/include/scsi/scsi_transport_fc.h b/include/scsi/scsi_transport_fc.h index b0d445437372..c04405bead2d 100644 --- a/include/scsi/scsi_transport_fc.h +++ b/include/scsi/scsi_transport_fc.h @@ -28,6 +28,7 @@ #define SCSI_TRANSPORT_FC_H #include +#include struct scsi_transport_template; diff --git a/kernel/kallsyms.c b/kernel/kallsyms.c index 13bcec151b57..39277dd6bf90 100644 --- a/kernel/kallsyms.c +++ b/kernel/kallsyms.c @@ -18,6 +18,7 @@ #include #include #include +#include /* for cond_resched */ #include #include diff --git a/kernel/kprobes.c b/kernel/kprobes.c index f3ea492ab44d..ce4915dd683a 100644 --- a/kernel/kprobes.c +++ b/kernel/kprobes.c @@ -35,6 +35,7 @@ #include #include #include +#include #include #include #include diff --git a/kernel/params.c b/kernel/params.c index 1a8614bac5d5..47ba69547945 100644 --- a/kernel/params.c +++ b/kernel/params.c @@ -23,6 +23,7 @@ #include #include #include +#include #if 0 #define DEBUGP printk diff --git a/lib/kobject.c b/lib/kobject.c index 253d3004ace9..a181abed89f6 100644 --- a/lib/kobject.c +++ b/lib/kobject.c @@ -14,6 +14,7 @@ #include #include #include +#include /** * populate_dir - populate directory with attributes. diff --git a/lib/smp_processor_id.c b/lib/smp_processor_id.c index 42c08ef828c5..eddc9b3d3876 100644 --- a/lib/smp_processor_id.c +++ b/lib/smp_processor_id.c @@ -5,6 +5,7 @@ */ #include #include +#include unsigned int debug_smp_processor_id(void) { diff --git a/lib/sort.c b/lib/sort.c index ddc4d35df289..5f3b51ffa1dc 100644 --- a/lib/sort.c +++ b/lib/sort.c @@ -7,6 +7,7 @@ #include #include #include +#include static void u32_swap(void *a, void *b, int size) { diff --git a/lib/vsprintf.c b/lib/vsprintf.c index e4e9031dd9c3..b07db5ca3f66 100644 --- a/lib/vsprintf.c +++ b/lib/vsprintf.c @@ -23,6 +23,7 @@ #include #include +#include /* for PAGE_SIZE */ #include /** diff --git a/sound/oss/ac97_codec.c b/sound/oss/ac97_codec.c index 3ecef4689f1b..fd25aca25120 100644 --- a/sound/oss/ac97_codec.c +++ b/sound/oss/ac97_codec.c @@ -55,6 +55,7 @@ #include #include #include +#include #define CODEC_ID_BUFSZ 14 -- cgit v1.2.3 From 451cbaa1c328082832a8fbcc427cd4416c602c5a Mon Sep 17 00:00:00 2001 From: OGAWA Hirofumi Date: Sun, 30 Oct 2005 15:03:49 -0800 Subject: [PATCH] fat: cleanup and optimization of checksum Signed-off-by: OGAWA Hirofumi Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/fat/dir.c | 10 ++-------- fs/vfat/namei.c | 3 +-- include/linux/msdos_fs.h | 11 +++++++++++ 3 files changed, 14 insertions(+), 10 deletions(-) (limited to 'include/linux') diff --git a/fs/fat/dir.c b/fs/fat/dir.c index 895049b2ac9c..204d8614406f 100644 --- a/fs/fat/dir.c +++ b/fs/fat/dir.c @@ -263,7 +263,6 @@ parse_record: unsigned char id; unsigned char slot; unsigned char slots; - unsigned char sum; unsigned char alias_checksum; if (!unicode) { @@ -317,9 +316,7 @@ parse_long: goto parse_long; if (IS_FREE(de->name) || (de->attr & ATTR_VOLUME)) continue; - for (sum = 0, i = 0; i < 11; i++) - sum = (((sum&1)<<7)|((sum&0xfe)>>1)) + de->name[i]; - if (sum != alias_checksum) + if (fat_checksum(de->name) != alias_checksum) nr_slots = 0; } @@ -479,7 +476,6 @@ GetNew: unsigned char id; unsigned char slot; unsigned char slots; - unsigned char sum; unsigned char alias_checksum; if (!unicode) { @@ -534,9 +530,7 @@ ParseLong: goto ParseLong; if (IS_FREE(de->name) || (de->attr & ATTR_VOLUME)) goto RecEnd; - for (sum = 0, i = 0; i < 11; i++) - sum = (((sum&1)<<7)|((sum&0xfe)>>1)) + de->name[i]; - if (sum != alias_checksum) + if (fat_checksum(de->name) != alias_checksum) long_slots = 0; } diff --git a/fs/vfat/namei.c b/fs/vfat/namei.c index 1c6f6b57ef1c..467346b123d9 100644 --- a/fs/vfat/namei.c +++ b/fs/vfat/namei.c @@ -621,8 +621,7 @@ static int vfat_build_slots(struct inode *dir, const unsigned char *name, } /* build the entry of long file name */ - for (cksum = i = 0; i < 11; i++) - cksum = (((cksum&1)<<7)|((cksum&0xfe)>>1)) + msdos_name[i]; + cksum = fat_checksum(msdos_name); *nr_slots = usize / 13; for (ps = slots, i = *nr_slots; i > 0; i--, ps++) { diff --git a/include/linux/msdos_fs.h b/include/linux/msdos_fs.h index 9a3d27257984..941da5c016a0 100644 --- a/include/linux/msdos_fs.h +++ b/include/linux/msdos_fs.h @@ -282,6 +282,17 @@ static inline u8 fat_attr(struct inode *inode) MSDOS_I(inode)->i_attrs; } +static inline unsigned char fat_checksum(const __u8 *name) +{ + unsigned char s = name[0]; + s = (s<<7) + (s>>1) + name[1]; s = (s<<7) + (s>>1) + name[2]; + s = (s<<7) + (s>>1) + name[3]; s = (s<<7) + (s>>1) + name[4]; + s = (s<<7) + (s>>1) + name[5]; s = (s<<7) + (s>>1) + name[6]; + s = (s<<7) + (s>>1) + name[7]; s = (s<<7) + (s>>1) + name[8]; + s = (s<<7) + (s>>1) + name[9]; s = (s<<7) + (s>>1) + name[10]; + return s; +} + static inline sector_t fat_clus_to_blknr(struct msdos_sb_info *sbi, int clus) { return ((sector_t)clus - FAT_START_ENT) * sbi->sec_per_clus -- cgit v1.2.3