From 2b0996c7646293c71c1297c00e07e54cee9bec5c Mon Sep 17 00:00:00 2001 From: Rosen Penev Date: Fri, 6 Sep 2024 12:53:59 -0700 Subject: wifi: ath9k: remove ath9k_platform_data MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Completely unused here in favor of Device Tree based setup. The DT code in here should currently match what is available with platform files. Any such lapse can always be added. Signed-off-by: Rosen Penev Acked-by: Toke Høiland-Jørgensen Signed-off-by: Kalle Valo Link: https://patch.msgid.link/20240906195359.6982-4-rosenp@gmail.com --- include/linux/ath9k_platform.h | 51 ------------------------------------------ 1 file changed, 51 deletions(-) delete mode 100644 include/linux/ath9k_platform.h (limited to 'include/linux') diff --git a/include/linux/ath9k_platform.h b/include/linux/ath9k_platform.h deleted file mode 100644 index 76860a461ed2..000000000000 --- a/include/linux/ath9k_platform.h +++ /dev/null @@ -1,51 +0,0 @@ -/* - * Copyright (c) 2008 Atheros Communications Inc. - * Copyright (c) 2009 Gabor Juhos - * Copyright (c) 2009 Imre Kaloz - * - * Permission to use, copy, modify, and/or distribute this software for any - * purpose with or without fee is hereby granted, provided that the above - * copyright notice and this permission notice appear in all copies. - * - * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES - * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF - * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR - * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES - * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN - * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF - * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE. - */ - -#ifndef _LINUX_ATH9K_PLATFORM_H -#define _LINUX_ATH9K_PLATFORM_H - -#define ATH9K_PLAT_EEP_MAX_WORDS 2048 - -struct ath9k_platform_data { - const char *eeprom_name; - - u16 eeprom_data[ATH9K_PLAT_EEP_MAX_WORDS]; - u8 *macaddr; - - int led_pin; - u32 gpio_mask; - u32 gpio_val; - - u32 bt_active_pin; - u32 bt_priority_pin; - u32 wlan_active_pin; - - bool endian_check; - bool is_clk_25mhz; - bool tx_gain_buffalo; - bool disable_2ghz; - bool disable_5ghz; - bool led_active_high; - - int (*get_mac_revision)(void); - int (*external_reset)(void); - - bool use_eeprom; -}; - -#endif /* _LINUX_ATH9K_PLATFORM_H */ -- cgit v1.2.3 From 277b339c4ba5c3d51f86892efd7bfbb012318ed8 Mon Sep 17 00:00:00 2001 From: "Russell King (Oracle)" Date: Tue, 1 Oct 2024 17:04:10 +0100 Subject: net: pcs: xpcs: move PCS reset to .pcs_pre_config() Move the PCS reset to .pcs_pre_config() rather than at creation time, which means we call the reset function with the interface that we're actually going to be using to talk to the downstream device. Reviewed-by: Vladimir Oltean Tested-by: Vladimir Oltean # sja1105 Signed-off-by: Russell King (Oracle) Tested-by: for them? Link: https://patch.msgid.link/E1svfMA-005ZI3-Va@rmk-PC.armlinux.org.uk Signed-off-by: Jakub Kicinski --- drivers/net/pcs/pcs-xpcs.c | 39 +++++++++++++++++++++++++++++---------- include/linux/pcs/pcs-xpcs.h | 1 + 2 files changed, 30 insertions(+), 10 deletions(-) (limited to 'include/linux') diff --git a/drivers/net/pcs/pcs-xpcs.c b/drivers/net/pcs/pcs-xpcs.c index 82463f9d50c8..7c6c40ddf722 100644 --- a/drivers/net/pcs/pcs-xpcs.c +++ b/drivers/net/pcs/pcs-xpcs.c @@ -659,6 +659,30 @@ int xpcs_config_eee(struct dw_xpcs *xpcs, int mult_fact_100ns, int enable) } EXPORT_SYMBOL_GPL(xpcs_config_eee); +static void xpcs_pre_config(struct phylink_pcs *pcs, phy_interface_t interface) +{ + struct dw_xpcs *xpcs = phylink_pcs_to_xpcs(pcs); + const struct dw_xpcs_compat *compat; + int ret; + + if (!xpcs->need_reset) + return; + + compat = xpcs_find_compat(xpcs->desc, interface); + if (!compat) { + dev_err(&xpcs->mdiodev->dev, "unsupported interface %s\n", + phy_modes(interface)); + return; + } + + ret = xpcs_soft_reset(xpcs, compat); + if (ret) + dev_err(&xpcs->mdiodev->dev, "soft reset failed: %pe\n", + ERR_PTR(ret)); + + xpcs->need_reset = false; +} + static int xpcs_config_aneg_c37_sgmii(struct dw_xpcs *xpcs, unsigned int neg_mode) { @@ -1365,6 +1389,7 @@ static const struct dw_xpcs_desc xpcs_desc_list[] = { static const struct phylink_pcs_ops xpcs_phylink_ops = { .pcs_validate = xpcs_validate, + .pcs_pre_config = xpcs_pre_config, .pcs_config = xpcs_config, .pcs_get_state = xpcs_get_state, .pcs_an_restart = xpcs_an_restart, @@ -1460,18 +1485,12 @@ static int xpcs_init_id(struct dw_xpcs *xpcs) static int xpcs_init_iface(struct dw_xpcs *xpcs, phy_interface_t interface) { - const struct dw_xpcs_compat *compat; - - compat = xpcs_find_compat(xpcs->desc, interface); - if (!compat) - return -EINVAL; - - if (xpcs->info.pma == WX_TXGBE_XPCS_PMA_10G_ID) { + if (xpcs->info.pma == WX_TXGBE_XPCS_PMA_10G_ID) xpcs->pcs.poll = false; - return 0; - } + else + xpcs->need_reset = true; - return xpcs_soft_reset(xpcs, compat); + return 0; } static struct dw_xpcs *xpcs_create(struct mdio_device *mdiodev, diff --git a/include/linux/pcs/pcs-xpcs.h b/include/linux/pcs/pcs-xpcs.h index b4a4eb6c8866..fd75d0605bb6 100644 --- a/include/linux/pcs/pcs-xpcs.h +++ b/include/linux/pcs/pcs-xpcs.h @@ -61,6 +61,7 @@ struct dw_xpcs { struct clk_bulk_data clks[DW_XPCS_NUM_CLKS]; struct phylink_pcs pcs; phy_interface_t interface; + bool need_reset; }; int xpcs_get_an_mode(struct dw_xpcs *xpcs, phy_interface_t interface); -- cgit v1.2.3 From bedea1539acb808c870ad85b16f8d8f0bef5e0bb Mon Sep 17 00:00:00 2001 From: "Russell King (Oracle)" Date: Tue, 1 Oct 2024 17:04:26 +0100 Subject: net: pcs: xpcs: add xpcs_destroy_pcs() and xpcs_create_pcs_mdiodev() Provide xpcs create/destroy functions that return and take a phylink_pcs pointer instead of an xpcs pointer. This will be used by drivers that have been converted to use phylink_pcs pointers internally, rather than dw_xpcs pointers. As xpcs_create_mdiodev() no longer makes use of its interface argument, pass PHY_INTERFACE_MODE_NA into xpcs_create_mdiodev() until it is removed later in the series. Reviewed-by: Vladimir Oltean Signed-off-by: Russell King (Oracle) Link: https://patch.msgid.link/E1svfMQ-005ZIL-Bi@rmk-PC.armlinux.org.uk Signed-off-by: Jakub Kicinski --- drivers/net/pcs/pcs-xpcs.c | 18 ++++++++++++++++++ include/linux/pcs/pcs-xpcs.h | 3 +++ 2 files changed, 21 insertions(+) (limited to 'include/linux') diff --git a/drivers/net/pcs/pcs-xpcs.c b/drivers/net/pcs/pcs-xpcs.c index 8765b01c0b5d..9b61f97222b9 100644 --- a/drivers/net/pcs/pcs-xpcs.c +++ b/drivers/net/pcs/pcs-xpcs.c @@ -1550,6 +1550,18 @@ struct dw_xpcs *xpcs_create_mdiodev(struct mii_bus *bus, int addr, } EXPORT_SYMBOL_GPL(xpcs_create_mdiodev); +struct phylink_pcs *xpcs_create_pcs_mdiodev(struct mii_bus *bus, int addr) +{ + struct dw_xpcs *xpcs; + + xpcs = xpcs_create_mdiodev(bus, addr, PHY_INTERFACE_MODE_NA); + if (IS_ERR(xpcs)) + return ERR_CAST(xpcs); + + return &xpcs->pcs; +} +EXPORT_SYMBOL_GPL(xpcs_create_pcs_mdiodev); + /** * xpcs_create_fwnode() - Create a DW xPCS instance from @fwnode * @fwnode: fwnode handle poining to the DW XPCS device @@ -1599,5 +1611,11 @@ void xpcs_destroy(struct dw_xpcs *xpcs) } EXPORT_SYMBOL_GPL(xpcs_destroy); +void xpcs_destroy_pcs(struct phylink_pcs *pcs) +{ + xpcs_destroy(phylink_pcs_to_xpcs(pcs)); +} +EXPORT_SYMBOL_GPL(xpcs_destroy_pcs); + MODULE_DESCRIPTION("Synopsys DesignWare XPCS library"); MODULE_LICENSE("GPL v2"); diff --git a/include/linux/pcs/pcs-xpcs.h b/include/linux/pcs/pcs-xpcs.h index fd75d0605bb6..a4e2243ce647 100644 --- a/include/linux/pcs/pcs-xpcs.h +++ b/include/linux/pcs/pcs-xpcs.h @@ -78,4 +78,7 @@ struct dw_xpcs *xpcs_create_fwnode(struct fwnode_handle *fwnode, phy_interface_t interface); void xpcs_destroy(struct dw_xpcs *xpcs); +struct phylink_pcs *xpcs_create_pcs_mdiodev(struct mii_bus *bus, int addr); +void xpcs_destroy_pcs(struct phylink_pcs *pcs); + #endif /* __LINUX_PCS_XPCS_H */ -- cgit v1.2.3 From bf5a61645bb2d51be53da5b26f948045b571147c Mon Sep 17 00:00:00 2001 From: "Russell King (Oracle)" Date: Tue, 1 Oct 2024 17:04:51 +0100 Subject: net: pcs: xpcs: drop interface argument from xpcs_create*() The XPCS sub-driver no longer uses the "interface" argument to the xpcs_create_mdiodev() and xpcs_create_fwnode() functions. Remove this now unnecessary argument, updating the stmmac driver appropriately. Reviewed-by: Vladimir Oltean Signed-off-by: Russell King (Oracle) Link: https://patch.msgid.link/E1svfMp-005ZIp-UX@rmk-PC.armlinux.org.uk Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/stmicro/stmmac/stmmac_mdio.c | 7 +++---- drivers/net/pcs/pcs-xpcs.c | 10 +++------- include/linux/pcs/pcs-xpcs.h | 6 ++---- 3 files changed, 8 insertions(+), 15 deletions(-) (limited to 'include/linux') diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_mdio.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_mdio.c index 03f90676b3ad..0c7d81ddd440 100644 --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_mdio.c +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_mdio.c @@ -500,23 +500,22 @@ int stmmac_pcs_setup(struct net_device *ndev) struct fwnode_handle *devnode, *pcsnode; struct dw_xpcs *xpcs = NULL; struct stmmac_priv *priv; - int addr, mode, ret; + int addr, ret; priv = netdev_priv(ndev); - mode = priv->plat->phy_interface; devnode = priv->plat->port_node; if (priv->plat->pcs_init) { ret = priv->plat->pcs_init(priv); } else if (fwnode_property_present(devnode, "pcs-handle")) { pcsnode = fwnode_find_reference(devnode, "pcs-handle", 0); - xpcs = xpcs_create_fwnode(pcsnode, mode); + xpcs = xpcs_create_fwnode(pcsnode); fwnode_handle_put(pcsnode); ret = PTR_ERR_OR_ZERO(xpcs); } else if (priv->plat->mdio_bus_data && priv->plat->mdio_bus_data->pcs_mask) { addr = ffs(priv->plat->mdio_bus_data->pcs_mask) - 1; - xpcs = xpcs_create_mdiodev(priv->mii, addr, mode); + xpcs = xpcs_create_mdiodev(priv->mii, addr); ret = PTR_ERR_OR_ZERO(xpcs); } else { return 0; diff --git a/drivers/net/pcs/pcs-xpcs.c b/drivers/net/pcs/pcs-xpcs.c index 9b61f97222b9..f25e7afdfdf5 100644 --- a/drivers/net/pcs/pcs-xpcs.c +++ b/drivers/net/pcs/pcs-xpcs.c @@ -1520,14 +1520,12 @@ out_free_data: * xpcs_create_mdiodev() - create a DW xPCS instance with the MDIO @addr * @bus: pointer to the MDIO-bus descriptor for the device to be looked at * @addr: device MDIO-bus ID - * @interface: requested PHY interface * * Return: a pointer to the DW XPCS handle if successful, otherwise -ENODEV if * the PCS device couldn't be found on the bus and other negative errno related * to the data allocation and MDIO-bus communications. */ -struct dw_xpcs *xpcs_create_mdiodev(struct mii_bus *bus, int addr, - phy_interface_t interface) +struct dw_xpcs *xpcs_create_mdiodev(struct mii_bus *bus, int addr) { struct mdio_device *mdiodev; struct dw_xpcs *xpcs; @@ -1554,7 +1552,7 @@ struct phylink_pcs *xpcs_create_pcs_mdiodev(struct mii_bus *bus, int addr) { struct dw_xpcs *xpcs; - xpcs = xpcs_create_mdiodev(bus, addr, PHY_INTERFACE_MODE_NA); + xpcs = xpcs_create_mdiodev(bus, addr); if (IS_ERR(xpcs)) return ERR_CAST(xpcs); @@ -1565,7 +1563,6 @@ EXPORT_SYMBOL_GPL(xpcs_create_pcs_mdiodev); /** * xpcs_create_fwnode() - Create a DW xPCS instance from @fwnode * @fwnode: fwnode handle poining to the DW XPCS device - * @interface: requested PHY interface * * Return: a pointer to the DW XPCS handle if successful, otherwise -ENODEV if * the fwnode device is unavailable or the PCS device couldn't be found on the @@ -1573,8 +1570,7 @@ EXPORT_SYMBOL_GPL(xpcs_create_pcs_mdiodev); * other negative errno related to the data allocations and MDIO-bus * communications. */ -struct dw_xpcs *xpcs_create_fwnode(struct fwnode_handle *fwnode, - phy_interface_t interface) +struct dw_xpcs *xpcs_create_fwnode(struct fwnode_handle *fwnode) { struct mdio_device *mdiodev; struct dw_xpcs *xpcs; diff --git a/include/linux/pcs/pcs-xpcs.h b/include/linux/pcs/pcs-xpcs.h index a4e2243ce647..758daabb76c7 100644 --- a/include/linux/pcs/pcs-xpcs.h +++ b/include/linux/pcs/pcs-xpcs.h @@ -72,10 +72,8 @@ int xpcs_do_config(struct dw_xpcs *xpcs, phy_interface_t interface, void xpcs_get_interfaces(struct dw_xpcs *xpcs, unsigned long *interfaces); int xpcs_config_eee(struct dw_xpcs *xpcs, int mult_fact_100ns, int enable); -struct dw_xpcs *xpcs_create_mdiodev(struct mii_bus *bus, int addr, - phy_interface_t interface); -struct dw_xpcs *xpcs_create_fwnode(struct fwnode_handle *fwnode, - phy_interface_t interface); +struct dw_xpcs *xpcs_create_mdiodev(struct mii_bus *bus, int addr); +struct dw_xpcs *xpcs_create_fwnode(struct fwnode_handle *fwnode); void xpcs_destroy(struct dw_xpcs *xpcs); struct phylink_pcs *xpcs_create_pcs_mdiodev(struct mii_bus *bus, int addr); -- cgit v1.2.3 From faefc9730d073e88a490e37b00e3e196b823abcb Mon Sep 17 00:00:00 2001 From: "Russell King (Oracle)" Date: Tue, 1 Oct 2024 17:04:57 +0100 Subject: net: pcs: xpcs: make xpcs_do_config() and xpcs_link_up() internal As nothing outside pcs-xpcs.c calls neither xpcs_do_config() nor xpcs_link_up(), remove their exports and prototypes. Reviewed-by: Vladimir Oltean Signed-off-by: Russell King (Oracle) Link: https://patch.msgid.link/E1svfMv-005ZIv-2M@rmk-PC.armlinux.org.uk Signed-off-by: Jakub Kicinski --- drivers/net/pcs/pcs-xpcs.c | 11 +++++------ include/linux/pcs/pcs-xpcs.h | 4 ---- 2 files changed, 5 insertions(+), 10 deletions(-) (limited to 'include/linux') diff --git a/drivers/net/pcs/pcs-xpcs.c b/drivers/net/pcs/pcs-xpcs.c index f25e7afdfdf5..0a01c552f591 100644 --- a/drivers/net/pcs/pcs-xpcs.c +++ b/drivers/net/pcs/pcs-xpcs.c @@ -851,8 +851,9 @@ static int xpcs_config_2500basex(struct dw_xpcs *xpcs) return xpcs_write(xpcs, MDIO_MMD_VEND2, DW_VR_MII_MMD_CTRL, ret); } -int xpcs_do_config(struct dw_xpcs *xpcs, phy_interface_t interface, - const unsigned long *advertising, unsigned int neg_mode) +static int xpcs_do_config(struct dw_xpcs *xpcs, phy_interface_t interface, + const unsigned long *advertising, + unsigned int neg_mode) { const struct dw_xpcs_compat *compat; int ret; @@ -905,7 +906,6 @@ int xpcs_do_config(struct dw_xpcs *xpcs, phy_interface_t interface, return 0; } -EXPORT_SYMBOL_GPL(xpcs_do_config); static int xpcs_config(struct phylink_pcs *pcs, unsigned int neg_mode, phy_interface_t interface, @@ -1207,8 +1207,8 @@ static void xpcs_link_up_1000basex(struct dw_xpcs *xpcs, unsigned int neg_mode, pr_err("%s: xpcs_write returned %pe\n", __func__, ERR_PTR(ret)); } -void xpcs_link_up(struct phylink_pcs *pcs, unsigned int neg_mode, - phy_interface_t interface, int speed, int duplex) +static void xpcs_link_up(struct phylink_pcs *pcs, unsigned int neg_mode, + phy_interface_t interface, int speed, int duplex) { struct dw_xpcs *xpcs = phylink_pcs_to_xpcs(pcs); @@ -1219,7 +1219,6 @@ void xpcs_link_up(struct phylink_pcs *pcs, unsigned int neg_mode, if (interface == PHY_INTERFACE_MODE_1000BASEX) return xpcs_link_up_1000basex(xpcs, neg_mode, speed, duplex); } -EXPORT_SYMBOL_GPL(xpcs_link_up); static void xpcs_an_restart(struct phylink_pcs *pcs) { diff --git a/include/linux/pcs/pcs-xpcs.h b/include/linux/pcs/pcs-xpcs.h index 758daabb76c7..abda475111d1 100644 --- a/include/linux/pcs/pcs-xpcs.h +++ b/include/linux/pcs/pcs-xpcs.h @@ -65,10 +65,6 @@ struct dw_xpcs { }; int xpcs_get_an_mode(struct dw_xpcs *xpcs, phy_interface_t interface); -void xpcs_link_up(struct phylink_pcs *pcs, unsigned int neg_mode, - phy_interface_t interface, int speed, int duplex); -int xpcs_do_config(struct dw_xpcs *xpcs, phy_interface_t interface, - const unsigned long *advertising, unsigned int neg_mode); void xpcs_get_interfaces(struct dw_xpcs *xpcs, unsigned long *interfaces); int xpcs_config_eee(struct dw_xpcs *xpcs, int mult_fact_100ns, int enable); -- cgit v1.2.3 From 816ad8f1e498fe5e9e992da137316302219f2137 Mon Sep 17 00:00:00 2001 From: Vladimir Oltean Date: Wed, 2 Oct 2024 14:51:52 -0700 Subject: lib: packing: remove kernel-doc from header file It is not necessary to have the kernel-doc duplicated both in the header and in the implementation. It is better to have it near the implementation of the function, since in C, a function can have N declarations, but only one definition. Signed-off-by: Vladimir Oltean Signed-off-by: Jacob Keller Reviewed-by: Przemek Kitszel Reviewed-by: Vladimir Oltean Link: https://patch.msgid.link/20241002-packing-kunit-tests-and-split-pack-unpack-v2-3-8373e551eae3@intel.com Signed-off-by: Jakub Kicinski --- include/linux/packing.h | 26 -------------------------- 1 file changed, 26 deletions(-) (limited to 'include/linux') diff --git a/include/linux/packing.h b/include/linux/packing.h index 8d6571feb95d..69baefebcd02 100644 --- a/include/linux/packing.h +++ b/include/linux/packing.h @@ -17,32 +17,6 @@ enum packing_op { UNPACK, }; -/** - * packing - Convert numbers (currently u64) between a packed and an unpacked - * format. Unpacked means laid out in memory in the CPU's native - * understanding of integers, while packed means anything else that - * requires translation. - * - * @pbuf: Pointer to a buffer holding the packed value. - * @uval: Pointer to an u64 holding the unpacked value. - * @startbit: The index (in logical notation, compensated for quirks) where - * the packed value starts within pbuf. Must be larger than, or - * equal to, endbit. - * @endbit: The index (in logical notation, compensated for quirks) where - * the packed value ends within pbuf. Must be smaller than, or equal - * to, startbit. - * @op: If PACK, then uval will be treated as const pointer and copied (packed) - * into pbuf, between startbit and endbit. - * If UNPACK, then pbuf will be treated as const pointer and the logical - * value between startbit and endbit will be copied (unpacked) to uval. - * @quirks: A bit mask of QUIRK_LITTLE_ENDIAN, QUIRK_LSW32_IS_FIRST and - * QUIRK_MSB_ON_THE_RIGHT. - * - * Return: 0 on success, EINVAL or ERANGE if called incorrectly. Assuming - * correct usage, return code may be discarded. - * If op is PACK, pbuf is modified. - * If op is UNPACK, uval is modified. - */ int packing(void *pbuf, u64 *uval, int startbit, int endbit, size_t pbuflen, enum packing_op op, u8 quirks); -- cgit v1.2.3 From 7263f64e16d90ded44ce211381b2def83db32fd9 Mon Sep 17 00:00:00 2001 From: Vladimir Oltean Date: Wed, 2 Oct 2024 14:51:53 -0700 Subject: lib: packing: add pack() and unpack() wrappers over packing() Geert Uytterhoeven described packing() as "really bad API" because of not being able to enforce const correctness. The same function is used both when "pbuf" is input and "uval" is output, as in the other way around. Create 2 wrapper functions where const correctness can be ensured. Do ugly type casts inside, to be able to reuse packing() as currently implemented - which will _not_ modify the input argument. Also, take the opportunity to change the type of startbit and endbit to size_t - an unsigned type - in these new function prototypes. When int, an extra check for negative values is necessary. Hopefully, when packing() goes away completely, that check can be dropped. My concern is that code which does rely on the conditional directionality of packing() is harder to refactor without blowing up in size. So it may take a while to completely eliminate packing(). But let's make alternatives available for those who do not need that. Link: https://lore.kernel.org/netdev/20210223112003.2223332-1-geert+renesas@glider.be/ Signed-off-by: Vladimir Oltean Signed-off-by: Jacob Keller Reviewed-by: Przemek Kitszel Reviewed-by: Vladimir Oltean Link: https://patch.msgid.link/20241002-packing-kunit-tests-and-split-pack-unpack-v2-4-8373e551eae3@intel.com Signed-off-by: Jakub Kicinski --- include/linux/packing.h | 6 ++++++ lib/packing.c | 54 +++++++++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 60 insertions(+) (limited to 'include/linux') diff --git a/include/linux/packing.h b/include/linux/packing.h index 69baefebcd02..ea25cb93cc70 100644 --- a/include/linux/packing.h +++ b/include/linux/packing.h @@ -20,4 +20,10 @@ enum packing_op { int packing(void *pbuf, u64 *uval, int startbit, int endbit, size_t pbuflen, enum packing_op op, u8 quirks); +int pack(void *pbuf, const u64 *uval, size_t startbit, size_t endbit, + size_t pbuflen, u8 quirks); + +int unpack(const void *pbuf, u64 *uval, size_t startbit, size_t endbit, + size_t pbuflen, u8 quirks); + #endif diff --git a/lib/packing.c b/lib/packing.c index 435236a914fe..2922db8a528c 100644 --- a/lib/packing.c +++ b/lib/packing.c @@ -90,6 +90,8 @@ static int calculate_box_addr(int box, size_t len, u8 quirks) * @quirks: A bit mask of QUIRK_LITTLE_ENDIAN, QUIRK_LSW32_IS_FIRST and * QUIRK_MSB_ON_THE_RIGHT. * + * Note: this is deprecated, prefer to use pack() or unpack() in new code. + * * Return: 0 on success, EINVAL or ERANGE if called incorrectly. Assuming * correct usage, return code may be discarded. * If op is PACK, pbuf is modified. @@ -216,4 +218,56 @@ int packing(void *pbuf, u64 *uval, int startbit, int endbit, size_t pbuflen, } EXPORT_SYMBOL(packing); +/** + * pack - Pack u64 number into bitfield of buffer. + * + * @pbuf: Pointer to a buffer holding the packed value. + * @uval: Pointer to an u64 holding the unpacked value. + * @startbit: The index (in logical notation, compensated for quirks) where + * the packed value starts within pbuf. Must be larger than, or + * equal to, endbit. + * @endbit: The index (in logical notation, compensated for quirks) where + * the packed value ends within pbuf. Must be smaller than, or equal + * to, startbit. + * @pbuflen: The length in bytes of the packed buffer pointed to by @pbuf. + * @quirks: A bit mask of QUIRK_LITTLE_ENDIAN, QUIRK_LSW32_IS_FIRST and + * QUIRK_MSB_ON_THE_RIGHT. + * + * Return: 0 on success, EINVAL or ERANGE if called incorrectly. Assuming + * correct usage, return code may be discarded. The @pbuf memory will + * be modified on success. + */ +int pack(void *pbuf, const u64 *uval, size_t startbit, size_t endbit, + size_t pbuflen, u8 quirks) +{ + return packing(pbuf, (u64 *)uval, startbit, endbit, pbuflen, PACK, quirks); +} +EXPORT_SYMBOL(pack); + +/** + * unpack - Unpack u64 number from packed buffer. + * + * @pbuf: Pointer to a buffer holding the packed value. + * @uval: Pointer to an u64 holding the unpacked value. + * @startbit: The index (in logical notation, compensated for quirks) where + * the packed value starts within pbuf. Must be larger than, or + * equal to, endbit. + * @endbit: The index (in logical notation, compensated for quirks) where + * the packed value ends within pbuf. Must be smaller than, or equal + * to, startbit. + * @pbuflen: The length in bytes of the packed buffer pointed to by @pbuf. + * @quirks: A bit mask of QUIRK_LITTLE_ENDIAN, QUIRK_LSW32_IS_FIRST and + * QUIRK_MSB_ON_THE_RIGHT. + * + * Return: 0 on success, EINVAL or ERANGE if called incorrectly. Assuming + * correct usage, return code may be discarded. The @uval will be + * modified on success. + */ +int unpack(const void *pbuf, u64 *uval, size_t startbit, size_t endbit, + size_t pbuflen, u8 quirks) +{ + return packing((void *)pbuf, uval, startbit, endbit, pbuflen, UNPACK, quirks); +} +EXPORT_SYMBOL(unpack); + MODULE_DESCRIPTION("Generic bitfield packing and unpacking"); -- cgit v1.2.3 From 28aec9ca29f05dabb835293acc6e202bf51b8092 Mon Sep 17 00:00:00 2001 From: Vladimir Oltean Date: Wed, 2 Oct 2024 14:51:54 -0700 Subject: lib: packing: duplicate pack() and unpack() implementations packing() is now used in some hot paths, and it would be good to get rid of some ifs and buts that depend on "op", to speed things up a little bit. With the main implementations now taking size_t endbit, we no longer have to check for negative values. Update the local integer variables to also be size_t to match. Signed-off-by: Vladimir Oltean Signed-off-by: Jacob Keller Reviewed-by: Przemek Kitszel Reviewed-by: Vladimir Oltean Link: https://patch.msgid.link/20241002-packing-kunit-tests-and-split-pack-unpack-v2-5-8373e551eae3@intel.com Signed-off-by: Jakub Kicinski --- include/linux/packing.h | 4 +- lib/packing.c | 221 ++++++++++++++++++++++++++++++------------------ 2 files changed, 143 insertions(+), 82 deletions(-) (limited to 'include/linux') diff --git a/include/linux/packing.h b/include/linux/packing.h index ea25cb93cc70..5d36dcd06f60 100644 --- a/include/linux/packing.h +++ b/include/linux/packing.h @@ -20,8 +20,8 @@ enum packing_op { int packing(void *pbuf, u64 *uval, int startbit, int endbit, size_t pbuflen, enum packing_op op, u8 quirks); -int pack(void *pbuf, const u64 *uval, size_t startbit, size_t endbit, - size_t pbuflen, u8 quirks); +int pack(void *pbuf, u64 uval, size_t startbit, size_t endbit, size_t pbuflen, + u8 quirks); int unpack(const void *pbuf, u64 *uval, size_t startbit, size_t endbit, size_t pbuflen, u8 quirks); diff --git a/lib/packing.c b/lib/packing.c index 2922db8a528c..500184530d07 100644 --- a/lib/packing.c +++ b/lib/packing.c @@ -9,11 +9,11 @@ #include #include -static void adjust_for_msb_right_quirk(u64 *to_write, int *box_start_bit, - int *box_end_bit, u8 *box_mask) +static void adjust_for_msb_right_quirk(u64 *to_write, size_t *box_start_bit, + size_t *box_end_bit, u8 *box_mask) { - int box_bit_width = *box_start_bit - *box_end_bit + 1; - int new_box_start_bit, new_box_end_bit; + size_t box_bit_width = *box_start_bit - *box_end_bit + 1; + size_t new_box_start_bit, new_box_end_bit; *to_write >>= *box_end_bit; *to_write = bitrev8(*to_write) >> (8 - box_bit_width); @@ -48,7 +48,7 @@ static void adjust_for_msb_right_quirk(u64 *to_write, int *box_start_bit, * * Return: the physical offset into the buffer corresponding to the logical box. */ -static int calculate_box_addr(int box, size_t len, u8 quirks) +static size_t calculate_box_addr(size_t box, size_t len, u8 quirks) { size_t offset_of_group, offset_in_group, this_group = box / 4; size_t group_size; @@ -69,13 +69,10 @@ static int calculate_box_addr(int box, size_t len, u8 quirks) } /** - * packing - Convert numbers (currently u64) between a packed and an unpacked - * format. Unpacked means laid out in memory in the CPU's native - * understanding of integers, while packed means anything else that - * requires translation. + * pack - Pack u64 number into bitfield of buffer. * * @pbuf: Pointer to a buffer holding the packed value. - * @uval: Pointer to an u64 holding the unpacked value. + * @uval: CPU-readable unpacked value to pack. * @startbit: The index (in logical notation, compensated for quirks) where * the packed value starts within pbuf. Must be larger than, or * equal to, endbit. @@ -83,58 +80,45 @@ static int calculate_box_addr(int box, size_t len, u8 quirks) * the packed value ends within pbuf. Must be smaller than, or equal * to, startbit. * @pbuflen: The length in bytes of the packed buffer pointed to by @pbuf. - * @op: If PACK, then uval will be treated as const pointer and copied (packed) - * into pbuf, between startbit and endbit. - * If UNPACK, then pbuf will be treated as const pointer and the logical - * value between startbit and endbit will be copied (unpacked) to uval. * @quirks: A bit mask of QUIRK_LITTLE_ENDIAN, QUIRK_LSW32_IS_FIRST and * QUIRK_MSB_ON_THE_RIGHT. * - * Note: this is deprecated, prefer to use pack() or unpack() in new code. - * * Return: 0 on success, EINVAL or ERANGE if called incorrectly. Assuming - * correct usage, return code may be discarded. - * If op is PACK, pbuf is modified. - * If op is UNPACK, uval is modified. + * correct usage, return code may be discarded. The @pbuf memory will + * be modified on success. */ -int packing(void *pbuf, u64 *uval, int startbit, int endbit, size_t pbuflen, - enum packing_op op, u8 quirks) +int pack(void *pbuf, u64 uval, size_t startbit, size_t endbit, size_t pbuflen, + u8 quirks) { - /* Number of bits for storing "uval" - * also width of the field to access in the pbuf - */ - u64 value_width; /* Logical byte indices corresponding to the * start and end of the field. */ int plogical_first_u8, plogical_last_u8, box; + /* width of the field to access in the pbuf */ + u64 value_width; /* startbit is expected to be larger than endbit, and both are * expected to be within the logically addressable range of the buffer. */ - if (unlikely(startbit < endbit || startbit >= 8 * pbuflen || endbit < 0)) + if (unlikely(startbit < endbit || startbit >= 8 * pbuflen)) /* Invalid function call */ return -EINVAL; value_width = startbit - endbit + 1; - if (value_width > 64) + if (unlikely(value_width > 64)) return -ERANGE; /* Check if "uval" fits in "value_width" bits. * If value_width is 64, the check will fail, but any * 64-bit uval will surely fit. */ - if (op == PACK && value_width < 64 && (*uval >= (1ull << value_width))) + if (unlikely(value_width < 64 && uval >= (1ull << value_width))) /* Cannot store "uval" inside "value_width" bits. * Truncating "uval" is most certainly not desirable, * so simply erroring out is appropriate. */ return -ERANGE; - /* Initialize parameter */ - if (op == UNPACK) - *uval = 0; - /* Iterate through an idealistic view of the pbuf as an u64 with * no quirks, u8 by u8 (aligned at u8 boundaries), from high to low * logical bit significance. "box" denotes the current logical u8. @@ -144,11 +128,12 @@ int packing(void *pbuf, u64 *uval, int startbit, int endbit, size_t pbuflen, for (box = plogical_first_u8; box >= plogical_last_u8; box--) { /* Bit indices into the currently accessed 8-bit box */ - int box_start_bit, box_end_bit, box_addr; + size_t box_start_bit, box_end_bit, box_addr; u8 box_mask; /* Corresponding bits from the unpacked u64 parameter */ - int proj_start_bit, proj_end_bit; + size_t proj_start_bit, proj_end_bit; u64 proj_mask; + u64 pval; /* This u8 may need to be accessed in its entirety * (from bit 7 to bit 0), or not, depending on the @@ -182,44 +167,25 @@ int packing(void *pbuf, u64 *uval, int startbit, int endbit, size_t pbuflen, */ box_addr = calculate_box_addr(box, pbuflen, quirks); - if (op == UNPACK) { - u64 pval; - - /* Read from pbuf, write to uval */ - pval = ((u8 *)pbuf)[box_addr] & box_mask; - if (quirks & QUIRK_MSB_ON_THE_RIGHT) - adjust_for_msb_right_quirk(&pval, - &box_start_bit, - &box_end_bit, - &box_mask); - - pval >>= box_end_bit; - pval <<= proj_end_bit; - *uval &= ~proj_mask; - *uval |= pval; - } else { - u64 pval; - - /* Write to pbuf, read from uval */ - pval = (*uval) & proj_mask; - pval >>= proj_end_bit; - if (quirks & QUIRK_MSB_ON_THE_RIGHT) - adjust_for_msb_right_quirk(&pval, - &box_start_bit, - &box_end_bit, - &box_mask); - - pval <<= box_end_bit; - ((u8 *)pbuf)[box_addr] &= ~box_mask; - ((u8 *)pbuf)[box_addr] |= pval; - } + /* Write to pbuf, read from uval */ + pval = uval & proj_mask; + pval >>= proj_end_bit; + if (quirks & QUIRK_MSB_ON_THE_RIGHT) + adjust_for_msb_right_quirk(&pval, + &box_start_bit, + &box_end_bit, + &box_mask); + + pval <<= box_end_bit; + ((u8 *)pbuf)[box_addr] &= ~box_mask; + ((u8 *)pbuf)[box_addr] |= pval; } return 0; } -EXPORT_SYMBOL(packing); +EXPORT_SYMBOL(pack); /** - * pack - Pack u64 number into bitfield of buffer. + * unpack - Unpack u64 number from packed buffer. * * @pbuf: Pointer to a buffer holding the packed value. * @uval: Pointer to an u64 holding the unpacked value. @@ -234,18 +200,103 @@ EXPORT_SYMBOL(packing); * QUIRK_MSB_ON_THE_RIGHT. * * Return: 0 on success, EINVAL or ERANGE if called incorrectly. Assuming - * correct usage, return code may be discarded. The @pbuf memory will - * be modified on success. + * correct usage, return code may be discarded. The @uval will be + * modified on success. */ -int pack(void *pbuf, const u64 *uval, size_t startbit, size_t endbit, - size_t pbuflen, u8 quirks) +int unpack(const void *pbuf, u64 *uval, size_t startbit, size_t endbit, + size_t pbuflen, u8 quirks) { - return packing(pbuf, (u64 *)uval, startbit, endbit, pbuflen, PACK, quirks); + /* Logical byte indices corresponding to the + * start and end of the field. + */ + int plogical_first_u8, plogical_last_u8, box; + /* width of the field to access in the pbuf */ + u64 value_width; + + /* startbit is expected to be larger than endbit, and both are + * expected to be within the logically addressable range of the buffer. + */ + if (unlikely(startbit < endbit || startbit >= 8 * pbuflen)) + /* Invalid function call */ + return -EINVAL; + + value_width = startbit - endbit + 1; + if (unlikely(value_width > 64)) + return -ERANGE; + + /* Initialize parameter */ + *uval = 0; + + /* Iterate through an idealistic view of the pbuf as an u64 with + * no quirks, u8 by u8 (aligned at u8 boundaries), from high to low + * logical bit significance. "box" denotes the current logical u8. + */ + plogical_first_u8 = startbit / 8; + plogical_last_u8 = endbit / 8; + + for (box = plogical_first_u8; box >= plogical_last_u8; box--) { + /* Bit indices into the currently accessed 8-bit box */ + size_t box_start_bit, box_end_bit, box_addr; + u8 box_mask; + /* Corresponding bits from the unpacked u64 parameter */ + size_t proj_start_bit, proj_end_bit; + u64 proj_mask; + u64 pval; + + /* This u8 may need to be accessed in its entirety + * (from bit 7 to bit 0), or not, depending on the + * input arguments startbit and endbit. + */ + if (box == plogical_first_u8) + box_start_bit = startbit % 8; + else + box_start_bit = 7; + if (box == plogical_last_u8) + box_end_bit = endbit % 8; + else + box_end_bit = 0; + + /* We have determined the box bit start and end. + * Now we calculate where this (masked) u8 box would fit + * in the unpacked (CPU-readable) u64 - the u8 box's + * projection onto the unpacked u64. Though the + * box is u8, the projection is u64 because it may fall + * anywhere within the unpacked u64. + */ + proj_start_bit = ((box * 8) + box_start_bit) - endbit; + proj_end_bit = ((box * 8) + box_end_bit) - endbit; + proj_mask = GENMASK_ULL(proj_start_bit, proj_end_bit); + box_mask = GENMASK_ULL(box_start_bit, box_end_bit); + + /* Determine the offset of the u8 box inside the pbuf, + * adjusted for quirks. The adjusted box_addr will be used for + * effective addressing inside the pbuf (so it's not + * logical any longer). + */ + box_addr = calculate_box_addr(box, pbuflen, quirks); + + /* Read from pbuf, write to uval */ + pval = ((u8 *)pbuf)[box_addr] & box_mask; + if (quirks & QUIRK_MSB_ON_THE_RIGHT) + adjust_for_msb_right_quirk(&pval, + &box_start_bit, + &box_end_bit, + &box_mask); + + pval >>= box_end_bit; + pval <<= proj_end_bit; + *uval &= ~proj_mask; + *uval |= pval; + } + return 0; } -EXPORT_SYMBOL(pack); +EXPORT_SYMBOL(unpack); /** - * unpack - Unpack u64 number from packed buffer. + * packing - Convert numbers (currently u64) between a packed and an unpacked + * format. Unpacked means laid out in memory in the CPU's native + * understanding of integers, while packed means anything else that + * requires translation. * * @pbuf: Pointer to a buffer holding the packed value. * @uval: Pointer to an u64 holding the unpacked value. @@ -256,18 +307,28 @@ EXPORT_SYMBOL(pack); * the packed value ends within pbuf. Must be smaller than, or equal * to, startbit. * @pbuflen: The length in bytes of the packed buffer pointed to by @pbuf. + * @op: If PACK, then uval will be treated as const pointer and copied (packed) + * into pbuf, between startbit and endbit. + * If UNPACK, then pbuf will be treated as const pointer and the logical + * value between startbit and endbit will be copied (unpacked) to uval. * @quirks: A bit mask of QUIRK_LITTLE_ENDIAN, QUIRK_LSW32_IS_FIRST and * QUIRK_MSB_ON_THE_RIGHT. * + * Note: this is deprecated, prefer to use pack() or unpack() in new code. + * * Return: 0 on success, EINVAL or ERANGE if called incorrectly. Assuming - * correct usage, return code may be discarded. The @uval will be - * modified on success. + * correct usage, return code may be discarded. + * If op is PACK, pbuf is modified. + * If op is UNPACK, uval is modified. */ -int unpack(const void *pbuf, u64 *uval, size_t startbit, size_t endbit, - size_t pbuflen, u8 quirks) +int packing(void *pbuf, u64 *uval, int startbit, int endbit, size_t pbuflen, + enum packing_op op, u8 quirks) { - return packing((void *)pbuf, uval, startbit, endbit, pbuflen, UNPACK, quirks); + if (op == PACK) + return pack(pbuf, *uval, startbit, endbit, pbuflen, quirks); + + return unpack(pbuf, uval, startbit, endbit, pbuflen, quirks); } -EXPORT_SYMBOL(unpack); +EXPORT_SYMBOL(packing); MODULE_DESCRIPTION("Generic bitfield packing and unpacking"); -- cgit v1.2.3 From b63c755cb65d43c8aba987c4f6b57c77c6f123f2 Mon Sep 17 00:00:00 2001 From: "Dr. David Alan Gilbert" Date: Mon, 30 Sep 2024 14:29:53 +0100 Subject: appletalk: Remove deadcode alloc_ltalkdev in net/appletalk/dev.c is dead since commit 00f3696f7555 ("net: appletalk: remove cops support") Removing it (and it's helper) leaves dev.c and if_ltalk.h empty; remove them and the Makefile entry. tun.c was including that if_ltalk.h but actually wanted the uapi version for LTALK_ALEN, fix up the path. Signed-off-by: Dr. David Alan Gilbert Reviewed-by: Willem de Bruijn Acked-by: Jason Wang Signed-off-by: David S. Miller --- drivers/net/tun.c | 2 +- include/linux/if_ltalk.h | 8 -------- net/appletalk/Makefile | 2 +- net/appletalk/dev.c | 46 ---------------------------------------------- 4 files changed, 2 insertions(+), 56 deletions(-) delete mode 100644 include/linux/if_ltalk.h delete mode 100644 net/appletalk/dev.c (limited to 'include/linux') diff --git a/drivers/net/tun.c b/drivers/net/tun.c index 9a0f6eb32016..d7a865ef370b 100644 --- a/drivers/net/tun.c +++ b/drivers/net/tun.c @@ -71,7 +71,7 @@ #include #include #include -#include +#include #include #include #include diff --git a/include/linux/if_ltalk.h b/include/linux/if_ltalk.h deleted file mode 100644 index 4cc1c0b77870..000000000000 --- a/include/linux/if_ltalk.h +++ /dev/null @@ -1,8 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -#ifndef __LINUX_LTALK_H -#define __LINUX_LTALK_H - -#include - -extern struct net_device *alloc_ltalkdev(int sizeof_priv); -#endif diff --git a/net/appletalk/Makefile b/net/appletalk/Makefile index 33164d972d37..152312a15180 100644 --- a/net/appletalk/Makefile +++ b/net/appletalk/Makefile @@ -5,6 +5,6 @@ obj-$(CONFIG_ATALK) += appletalk.o -appletalk-y := aarp.o ddp.o dev.o +appletalk-y := aarp.o ddp.o appletalk-$(CONFIG_PROC_FS) += atalk_proc.o appletalk-$(CONFIG_SYSCTL) += sysctl_net_atalk.o diff --git a/net/appletalk/dev.c b/net/appletalk/dev.c deleted file mode 100644 index 284c8e585533..000000000000 --- a/net/appletalk/dev.c +++ /dev/null @@ -1,46 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0 -/* - * Moved here from drivers/net/net_init.c, which is: - * Written 1993,1994,1995 by Donald Becker. - */ - -#include -#include -#include -#include -#include - -static void ltalk_setup(struct net_device *dev) -{ - /* Fill in the fields of the device structure with localtalk-generic values. */ - - dev->type = ARPHRD_LOCALTLK; - dev->hard_header_len = LTALK_HLEN; - dev->mtu = LTALK_MTU; - dev->addr_len = LTALK_ALEN; - dev->tx_queue_len = 10; - - dev->broadcast[0] = 0xFF; - - dev->flags = IFF_BROADCAST|IFF_MULTICAST|IFF_NOARP; -} - -/** - * alloc_ltalkdev - Allocates and sets up an localtalk device - * @sizeof_priv: Size of additional driver-private structure to be allocated - * for this localtalk device - * - * Fill in the fields of the device structure with localtalk-generic - * values. Basically does everything except registering the device. - * - * Constructs a new net device, complete with a private data area of - * size @sizeof_priv. A 32-byte (not bit) alignment is enforced for - * this private data area. - */ - -struct net_device *alloc_ltalkdev(int sizeof_priv) -{ - return alloc_netdev(sizeof_priv, "lt%d", NET_NAME_UNKNOWN, - ltalk_setup); -} -EXPORT_SYMBOL(alloc_ltalkdev); -- cgit v1.2.3 From 5acd957a986c167d357e617a887630203be29ea4 Mon Sep 17 00:00:00 2001 From: Cosmin Ratiu Date: Tue, 1 Oct 2024 13:37:04 +0300 Subject: net/mlx5: hw counters: Make fc_stats & fc_pool private The mlx5_fc_stats and mlx5_fc_pool structs are only used from fs_counters.c. As such, make them private there. mlx5_fc_pool is not used or referenced at all outside fs_counters. mlx5_fc_stats is referenced from mlx5_core_dev, so instead of having it as a direct member (which requires exporting it from fs_counters), store a pointer to it, allocate it on init and clear it on destroy. One caveat is that a simple container_of to get from a 'work' struct to the outermost mlx5_core_dev struct directly no longer works, so an extra pointer had to be added to mlx5_fc_stats back to the parent dev. Signed-off-by: Cosmin Ratiu Signed-off-by: Tariq Toukan Link: https://patch.msgid.link/20241001103709.58127-2-tariqt@nvidia.com Signed-off-by: Jakub Kicinski --- .../net/ethernet/mellanox/mlx5/core/fs_counters.c | 79 ++++++++++++++++------ include/linux/mlx5/driver.h | 33 +-------- 2 files changed, 60 insertions(+), 52 deletions(-) (limited to 'include/linux') diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_counters.c b/drivers/net/ethernet/mellanox/mlx5/core/fs_counters.c index 0c26d707eed2..7d6174d0f260 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/fs_counters.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_counters.c @@ -69,6 +69,36 @@ struct mlx5_fc { struct mlx5_fc_cache cache ____cacheline_aligned_in_smp; }; +struct mlx5_fc_pool { + struct mlx5_core_dev *dev; + struct mutex pool_lock; /* protects pool lists */ + struct list_head fully_used; + struct list_head partially_used; + struct list_head unused; + int available_fcs; + int used_fcs; + int threshold; +}; + +struct mlx5_fc_stats { + spinlock_t counters_idr_lock; /* protects counters_idr */ + struct idr counters_idr; + struct list_head counters; + struct llist_head addlist; + struct llist_head dellist; + + struct workqueue_struct *wq; + struct delayed_work work; + unsigned long next_query; + unsigned long sampling_interval; /* jiffies */ + u32 *bulk_query_out; + int bulk_query_len; + size_t num_counters; + bool bulk_query_alloc_failed; + unsigned long next_bulk_query_alloc; + struct mlx5_fc_pool fc_pool; +}; + static void mlx5_fc_pool_init(struct mlx5_fc_pool *fc_pool, struct mlx5_core_dev *dev); static void mlx5_fc_pool_cleanup(struct mlx5_fc_pool *fc_pool); static struct mlx5_fc *mlx5_fc_pool_acquire_counter(struct mlx5_fc_pool *fc_pool); @@ -109,7 +139,7 @@ static void mlx5_fc_pool_release_counter(struct mlx5_fc_pool *fc_pool, struct ml static struct list_head *mlx5_fc_counters_lookup_next(struct mlx5_core_dev *dev, u32 id) { - struct mlx5_fc_stats *fc_stats = &dev->priv.fc_stats; + struct mlx5_fc_stats *fc_stats = dev->priv.fc_stats; unsigned long next_id = (unsigned long)id + 1; struct mlx5_fc *counter; unsigned long tmp; @@ -137,7 +167,7 @@ static void mlx5_fc_stats_insert(struct mlx5_core_dev *dev, static void mlx5_fc_stats_remove(struct mlx5_core_dev *dev, struct mlx5_fc *counter) { - struct mlx5_fc_stats *fc_stats = &dev->priv.fc_stats; + struct mlx5_fc_stats *fc_stats = dev->priv.fc_stats; list_del(&counter->list); @@ -178,7 +208,7 @@ static void mlx5_fc_stats_query_counter_range(struct mlx5_core_dev *dev, struct mlx5_fc *first, u32 last_id) { - struct mlx5_fc_stats *fc_stats = &dev->priv.fc_stats; + struct mlx5_fc_stats *fc_stats = dev->priv.fc_stats; bool query_more_counters = (first->id <= last_id); int cur_bulk_len = fc_stats->bulk_query_len; u32 *data = fc_stats->bulk_query_out; @@ -225,7 +255,7 @@ static void mlx5_fc_free(struct mlx5_core_dev *dev, struct mlx5_fc *counter) static void mlx5_fc_release(struct mlx5_core_dev *dev, struct mlx5_fc *counter) { - struct mlx5_fc_stats *fc_stats = &dev->priv.fc_stats; + struct mlx5_fc_stats *fc_stats = dev->priv.fc_stats; if (counter->bulk) mlx5_fc_pool_release_counter(&fc_stats->fc_pool, counter); @@ -235,7 +265,7 @@ static void mlx5_fc_release(struct mlx5_core_dev *dev, struct mlx5_fc *counter) static void mlx5_fc_stats_bulk_query_size_increase(struct mlx5_core_dev *dev) { - struct mlx5_fc_stats *fc_stats = &dev->priv.fc_stats; + struct mlx5_fc_stats *fc_stats = dev->priv.fc_stats; int max_bulk_len = get_max_bulk_query_len(dev); unsigned long now = jiffies; u32 *bulk_query_out_tmp; @@ -270,9 +300,9 @@ static void mlx5_fc_stats_bulk_query_size_increase(struct mlx5_core_dev *dev) static void mlx5_fc_stats_work(struct work_struct *work) { - struct mlx5_core_dev *dev = container_of(work, struct mlx5_core_dev, - priv.fc_stats.work.work); - struct mlx5_fc_stats *fc_stats = &dev->priv.fc_stats; + struct mlx5_fc_stats *fc_stats = container_of(work, struct mlx5_fc_stats, + work.work); + struct mlx5_core_dev *dev = fc_stats->fc_pool.dev; /* Take dellist first to ensure that counters cannot be deleted before * they are inserted. */ @@ -334,7 +364,7 @@ static struct mlx5_fc *mlx5_fc_single_alloc(struct mlx5_core_dev *dev) static struct mlx5_fc *mlx5_fc_acquire(struct mlx5_core_dev *dev, bool aging) { - struct mlx5_fc_stats *fc_stats = &dev->priv.fc_stats; + struct mlx5_fc_stats *fc_stats = dev->priv.fc_stats; struct mlx5_fc *counter; if (aging && MLX5_CAP_GEN(dev, flow_counter_bulk_alloc) != 0) { @@ -349,7 +379,7 @@ static struct mlx5_fc *mlx5_fc_acquire(struct mlx5_core_dev *dev, bool aging) struct mlx5_fc *mlx5_fc_create_ex(struct mlx5_core_dev *dev, bool aging) { struct mlx5_fc *counter = mlx5_fc_acquire(dev, aging); - struct mlx5_fc_stats *fc_stats = &dev->priv.fc_stats; + struct mlx5_fc_stats *fc_stats = dev->priv.fc_stats; int err; if (IS_ERR(counter)) @@ -389,7 +419,7 @@ err_out_alloc: struct mlx5_fc *mlx5_fc_create(struct mlx5_core_dev *dev, bool aging) { struct mlx5_fc *counter = mlx5_fc_create_ex(dev, aging); - struct mlx5_fc_stats *fc_stats = &dev->priv.fc_stats; + struct mlx5_fc_stats *fc_stats = dev->priv.fc_stats; if (aging) mod_delayed_work(fc_stats->wq, &fc_stats->work, 0); @@ -405,7 +435,7 @@ EXPORT_SYMBOL(mlx5_fc_id); void mlx5_fc_destroy(struct mlx5_core_dev *dev, struct mlx5_fc *counter) { - struct mlx5_fc_stats *fc_stats = &dev->priv.fc_stats; + struct mlx5_fc_stats *fc_stats = dev->priv.fc_stats; if (!counter) return; @@ -422,10 +452,14 @@ EXPORT_SYMBOL(mlx5_fc_destroy); int mlx5_init_fc_stats(struct mlx5_core_dev *dev) { - struct mlx5_fc_stats *fc_stats = &dev->priv.fc_stats; + struct mlx5_fc_stats *fc_stats; int init_bulk_len; int init_out_len; + fc_stats = kzalloc(sizeof(*fc_stats), GFP_KERNEL); + if (!fc_stats) + return -ENOMEM; + spin_lock_init(&fc_stats->counters_idr_lock); idr_init(&fc_stats->counters_idr); INIT_LIST_HEAD(&fc_stats->counters); @@ -436,7 +470,7 @@ int mlx5_init_fc_stats(struct mlx5_core_dev *dev) init_out_len = mlx5_cmd_fc_get_bulk_query_out_len(init_bulk_len); fc_stats->bulk_query_out = kzalloc(init_out_len, GFP_KERNEL); if (!fc_stats->bulk_query_out) - return -ENOMEM; + goto err_bulk; fc_stats->bulk_query_len = init_bulk_len; fc_stats->wq = create_singlethread_workqueue("mlx5_fc"); @@ -447,23 +481,27 @@ int mlx5_init_fc_stats(struct mlx5_core_dev *dev) INIT_DELAYED_WORK(&fc_stats->work, mlx5_fc_stats_work); mlx5_fc_pool_init(&fc_stats->fc_pool, dev); + dev->priv.fc_stats = fc_stats; + return 0; err_wq_create: kfree(fc_stats->bulk_query_out); +err_bulk: + kfree(fc_stats); return -ENOMEM; } void mlx5_cleanup_fc_stats(struct mlx5_core_dev *dev) { - struct mlx5_fc_stats *fc_stats = &dev->priv.fc_stats; + struct mlx5_fc_stats *fc_stats = dev->priv.fc_stats; struct llist_node *tmplist; struct mlx5_fc *counter; struct mlx5_fc *tmp; - cancel_delayed_work_sync(&dev->priv.fc_stats.work); - destroy_workqueue(dev->priv.fc_stats.wq); - dev->priv.fc_stats.wq = NULL; + cancel_delayed_work_sync(&fc_stats->work); + destroy_workqueue(fc_stats->wq); + fc_stats->wq = NULL; tmplist = llist_del_all(&fc_stats->addlist); llist_for_each_entry_safe(counter, tmp, tmplist, addlist) @@ -475,6 +513,7 @@ void mlx5_cleanup_fc_stats(struct mlx5_core_dev *dev) mlx5_fc_pool_cleanup(&fc_stats->fc_pool); idr_destroy(&fc_stats->counters_idr); kfree(fc_stats->bulk_query_out); + kfree(fc_stats); } int mlx5_fc_query(struct mlx5_core_dev *dev, struct mlx5_fc *counter, @@ -518,7 +557,7 @@ void mlx5_fc_queue_stats_work(struct mlx5_core_dev *dev, struct delayed_work *dwork, unsigned long delay) { - struct mlx5_fc_stats *fc_stats = &dev->priv.fc_stats; + struct mlx5_fc_stats *fc_stats = dev->priv.fc_stats; queue_delayed_work(fc_stats->wq, dwork, delay); } @@ -526,7 +565,7 @@ void mlx5_fc_queue_stats_work(struct mlx5_core_dev *dev, void mlx5_fc_update_sampling_interval(struct mlx5_core_dev *dev, unsigned long interval) { - struct mlx5_fc_stats *fc_stats = &dev->priv.fc_stats; + struct mlx5_fc_stats *fc_stats = dev->priv.fc_stats; fc_stats->sampling_interval = min_t(unsigned long, interval, fc_stats->sampling_interval); diff --git a/include/linux/mlx5/driver.h b/include/linux/mlx5/driver.h index e23c692a34c7..fc7e6153b73d 100644 --- a/include/linux/mlx5/driver.h +++ b/include/linux/mlx5/driver.h @@ -45,7 +45,6 @@ #include #include #include -#include #include #include #include @@ -474,36 +473,6 @@ struct mlx5_core_sriov { u16 max_ec_vfs; }; -struct mlx5_fc_pool { - struct mlx5_core_dev *dev; - struct mutex pool_lock; /* protects pool lists */ - struct list_head fully_used; - struct list_head partially_used; - struct list_head unused; - int available_fcs; - int used_fcs; - int threshold; -}; - -struct mlx5_fc_stats { - spinlock_t counters_idr_lock; /* protects counters_idr */ - struct idr counters_idr; - struct list_head counters; - struct llist_head addlist; - struct llist_head dellist; - - struct workqueue_struct *wq; - struct delayed_work work; - unsigned long next_query; - unsigned long sampling_interval; /* jiffies */ - u32 *bulk_query_out; - int bulk_query_len; - size_t num_counters; - bool bulk_query_alloc_failed; - unsigned long next_bulk_query_alloc; - struct mlx5_fc_pool fc_pool; -}; - struct mlx5_events; struct mlx5_mpfs; struct mlx5_eswitch; @@ -630,7 +599,7 @@ struct mlx5_priv { struct mlx5_devcom_comp_dev *hca_devcom_comp; struct mlx5_fw_reset *fw_reset; struct mlx5_core_roce roce; - struct mlx5_fc_stats fc_stats; + struct mlx5_fc_stats *fc_stats; struct mlx5_rl_table rl_table; struct mlx5_ft_pool *ft_pool; -- cgit v1.2.3 From d1c9cffe4b01f4d8bc52169139a5fedd48908abc Mon Sep 17 00:00:00 2001 From: Cosmin Ratiu Date: Tue, 1 Oct 2024 13:37:09 +0300 Subject: net/mlx5: hw counters: Remove mlx5_fc_create_ex It no longer serves any purpose and is identical to mlx5_fc_create upon which it was originally based of. Signed-off-by: Cosmin Ratiu Signed-off-by: Tariq Toukan Link: https://patch.msgid.link/20241001103709.58127-7-tariqt@nvidia.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c | 2 +- drivers/net/ethernet/mellanox/mlx5/core/fs_counters.c | 7 +------ include/linux/mlx5/fs.h | 3 --- 3 files changed, 2 insertions(+), 10 deletions(-) (limited to 'include/linux') diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c b/drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c index dcfccaaa8d91..4877a9d86807 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c @@ -1026,7 +1026,7 @@ mlx5_tc_ct_counter_create(struct mlx5_tc_ct_priv *ct_priv) return ERR_PTR(-ENOMEM); counter->is_shared = false; - counter->counter = mlx5_fc_create_ex(ct_priv->dev, true); + counter->counter = mlx5_fc_create(ct_priv->dev, true); if (IS_ERR(counter->counter)) { ct_dbg("Failed to create counter for ct entry"); ret = PTR_ERR(counter->counter); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_counters.c b/drivers/net/ethernet/mellanox/mlx5/core/fs_counters.c index 0b80c33cba5f..62d0c689796b 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/fs_counters.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_counters.c @@ -275,7 +275,7 @@ static struct mlx5_fc *mlx5_fc_acquire(struct mlx5_core_dev *dev, bool aging) return mlx5_fc_single_alloc(dev); } -struct mlx5_fc *mlx5_fc_create_ex(struct mlx5_core_dev *dev, bool aging) +struct mlx5_fc *mlx5_fc_create(struct mlx5_core_dev *dev, bool aging) { struct mlx5_fc *counter = mlx5_fc_acquire(dev, aging); struct mlx5_fc_stats *fc_stats = dev->priv.fc_stats; @@ -304,11 +304,6 @@ err_out_alloc: mlx5_fc_release(dev, counter); return ERR_PTR(err); } - -struct mlx5_fc *mlx5_fc_create(struct mlx5_core_dev *dev, bool aging) -{ - return mlx5_fc_create_ex(dev, aging); -} EXPORT_SYMBOL(mlx5_fc_create); u32 mlx5_fc_id(struct mlx5_fc *counter) diff --git a/include/linux/mlx5/fs.h b/include/linux/mlx5/fs.h index b744e554f014..438db888bde0 100644 --- a/include/linux/mlx5/fs.h +++ b/include/linux/mlx5/fs.h @@ -298,9 +298,6 @@ int mlx5_modify_rule_destination(struct mlx5_flow_handle *handler, struct mlx5_fc *mlx5_fc_create(struct mlx5_core_dev *dev, bool aging); -/* As mlx5_fc_create() but doesn't queue stats refresh thread. */ -struct mlx5_fc *mlx5_fc_create_ex(struct mlx5_core_dev *dev, bool aging); - void mlx5_fc_destroy(struct mlx5_core_dev *dev, struct mlx5_fc *counter); u64 mlx5_fc_query_lastuse(struct mlx5_fc *counter); void mlx5_fc_query_cached(struct mlx5_fc *counter, -- cgit v1.2.3 From f858cc9eed5b05cbe38d7ffd2787c21e3718eb7d Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Thu, 3 Oct 2024 12:12:18 +0000 Subject: net: add IFLA_MAX_PACING_OFFLOAD_HORIZON device attribute Some network devices have the ability to offload EDT (Earliest Departure Time) which is the model used for TCP pacing and FQ packet scheduler. Some of them implement the timing wheel mechanism described in https://saeed.github.io/files/carousel-sigcomm17.pdf with an associated 'timing wheel horizon'. This patch adds dev->max_pacing_offload_horizon expressing this timing wheel horizon in nsec units. This is a read-only attribute. Unless a driver sets it, dev->max_pacing_offload_horizon is zero. v2: addressed Jakub feedback ( https://lore.kernel.org/netdev/20240930152304.472767-2-edumazet@google.com/T/#mf6294d714c41cc459962154cc2580ce3c9693663 ) v3: added yaml doc (also per Jakub feedback) Signed-off-by: Eric Dumazet Reviewed-by: Willem de Bruijn Link: https://patch.msgid.link/20241003121219.2396589-2-edumazet@google.com Signed-off-by: Jakub Kicinski --- Documentation/netlink/specs/rt_link.yaml | 4 ++++ Documentation/networking/net_cachelines/net_device.rst | 1 + include/linux/netdevice.h | 4 ++++ include/uapi/linux/if_link.h | 1 + net/core/rtnetlink.c | 4 ++++ tools/include/uapi/linux/if_link.h | 1 + 6 files changed, 15 insertions(+) (limited to 'include/linux') diff --git a/Documentation/netlink/specs/rt_link.yaml b/Documentation/netlink/specs/rt_link.yaml index 0c4d5d40cae9..d7131a1afadf 100644 --- a/Documentation/netlink/specs/rt_link.yaml +++ b/Documentation/netlink/specs/rt_link.yaml @@ -1137,6 +1137,10 @@ attribute-sets: name: dpll-pin type: nest nested-attributes: link-dpll-pin-attrs + - + name: max-pacing-offload-horizon + type: uint + doc: EDT offload horizon supported by the device (in nsec). - name: af-spec-attrs attributes: diff --git a/Documentation/networking/net_cachelines/net_device.rst b/Documentation/networking/net_cachelines/net_device.rst index 22b07c814f4a..49f03cb78c6e 100644 --- a/Documentation/networking/net_cachelines/net_device.rst +++ b/Documentation/networking/net_cachelines/net_device.rst @@ -183,3 +183,4 @@ struct_devlink_port* devlink_port struct_dpll_pin* dpll_pin struct hlist_head page_pools struct dim_irq_moder* irq_moder +u64 max_pacing_offload_horizon diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index 4d20c776a4ff..49a7e7db0883 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -2009,6 +2009,8 @@ enum netdev_reg_state { * @dpll_pin: Pointer to the SyncE source pin of a DPLL subsystem, * where the clock is recovered. * + * @max_pacing_offload_horizon: max EDT offload horizon in nsec. + * * FIXME: cleanup struct net_device such that network protocol info * moves out. */ @@ -2399,6 +2401,8 @@ struct net_device { /** @irq_moder: dim parameters used if IS_ENABLED(CONFIG_DIMLIB). */ struct dim_irq_moder *irq_moder; + u64 max_pacing_offload_horizon; + u8 priv[] ____cacheline_aligned __counted_by(priv_len); } ____cacheline_aligned; diff --git a/include/uapi/linux/if_link.h b/include/uapi/linux/if_link.h index 6dc258993b17..506ba9c80e83 100644 --- a/include/uapi/linux/if_link.h +++ b/include/uapi/linux/if_link.h @@ -377,6 +377,7 @@ enum { IFLA_GSO_IPV4_MAX_SIZE, IFLA_GRO_IPV4_MAX_SIZE, IFLA_DPLL_PIN, + IFLA_MAX_PACING_OFFLOAD_HORIZON, __IFLA_MAX }; diff --git a/net/core/rtnetlink.c b/net/core/rtnetlink.c index f0a520987085..682d8d3127db 100644 --- a/net/core/rtnetlink.c +++ b/net/core/rtnetlink.c @@ -1118,6 +1118,7 @@ static noinline size_t if_nlmsg_size(const struct net_device *dev, + nla_total_size(MAX_ADDR_LEN) /* IFLA_PERM_ADDRESS */ + rtnl_devlink_port_size(dev) + rtnl_dpll_pin_size(dev) + + nla_total_size(8) /* IFLA_MAX_PACING_OFFLOAD_HORIZON */ + 0; } @@ -1867,6 +1868,8 @@ static int rtnl_fill_ifinfo(struct sk_buff *skb, READ_ONCE(dev->tso_max_size)) || nla_put_u32(skb, IFLA_TSO_MAX_SEGS, READ_ONCE(dev->tso_max_segs)) || + nla_put_uint(skb, IFLA_MAX_PACING_OFFLOAD_HORIZON, + READ_ONCE(dev->max_pacing_offload_horizon)) || #ifdef CONFIG_RPS nla_put_u32(skb, IFLA_NUM_RX_QUEUES, READ_ONCE(dev->num_rx_queues)) || @@ -1975,6 +1978,7 @@ nla_put_failure: } static const struct nla_policy ifla_policy[IFLA_MAX+1] = { + [IFLA_UNSPEC] = { .strict_start_type = IFLA_DPLL_PIN }, [IFLA_IFNAME] = { .type = NLA_STRING, .len = IFNAMSIZ-1 }, [IFLA_ADDRESS] = { .type = NLA_BINARY, .len = MAX_ADDR_LEN }, [IFLA_BROADCAST] = { .type = NLA_BINARY, .len = MAX_ADDR_LEN }, diff --git a/tools/include/uapi/linux/if_link.h b/tools/include/uapi/linux/if_link.h index f0d71b2a3f1e..96ec2b01e725 100644 --- a/tools/include/uapi/linux/if_link.h +++ b/tools/include/uapi/linux/if_link.h @@ -377,6 +377,7 @@ enum { IFLA_GSO_IPV4_MAX_SIZE, IFLA_GRO_IPV4_MAX_SIZE, IFLA_DPLL_PIN, + IFLA_MAX_PACING_OFFLOAD_HORIZON, __IFLA_MAX }; -- cgit v1.2.3 From a3f5f4c2f9b6bc2aa6f5a3e8e23b7519e4f2e3e3 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Fri, 4 Oct 2024 13:47:20 +0000 Subject: ipv4: remove fib_info_devhash[] Upcoming per-netns RTNL conversion needs to get rid of shared hash tables. fib_info_devhash[] is one of them. It is unclear why we used a hash table, because a single hlist_head per net device was cheaper and scalable. Signed-off-by: Eric Dumazet Reviewed-by: Kuniyuki Iwashima Reviewed-by: David Ahern Link: https://patch.msgid.link/20241004134720.579244-5-edumazet@google.com Signed-off-by: Jakub Kicinski --- .../networking/net_cachelines/net_device.rst | 1 + include/linux/netdevice.h | 3 ++ net/ipv4/fib_semantics.c | 35 ++++++++++------------ 3 files changed, 19 insertions(+), 20 deletions(-) (limited to 'include/linux') diff --git a/Documentation/networking/net_cachelines/net_device.rst b/Documentation/networking/net_cachelines/net_device.rst index 49f03cb78c6e..556711c4d3cf 100644 --- a/Documentation/networking/net_cachelines/net_device.rst +++ b/Documentation/networking/net_cachelines/net_device.rst @@ -83,6 +83,7 @@ unsigned_int allmulti bool uc_promisc unsigned_char nested_level struct_in_device* ip_ptr read_mostly read_mostly __in_dev_get +struct hlist_head fib_nh_head struct_inet6_dev* ip6_ptr read_mostly read_mostly __in6_dev_get struct_vlan_info* vlan_info struct_dsa_port* dsa_ptr diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index 49a7e7db0883..3baf8e539b6f 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -2211,6 +2211,9 @@ struct net_device { /* Protocol-specific pointers */ struct in_device __rcu *ip_ptr; + /** @fib_nh_head: nexthops associated with this netdev */ + struct hlist_head fib_nh_head; + #if IS_ENABLED(CONFIG_VLAN_8021Q) struct vlan_info __rcu *vlan_info; #endif diff --git a/net/ipv4/fib_semantics.c b/net/ipv4/fib_semantics.c index ece779bfb8f6..d2cee5c314f5 100644 --- a/net/ipv4/fib_semantics.c +++ b/net/ipv4/fib_semantics.c @@ -56,10 +56,6 @@ static unsigned int fib_info_hash_size; static unsigned int fib_info_hash_bits; static unsigned int fib_info_cnt; -#define DEVINDEX_HASHBITS 8 -#define DEVINDEX_HASHSIZE (1U << DEVINDEX_HASHBITS) -static struct hlist_head fib_info_devhash[DEVINDEX_HASHSIZE]; - /* for_nexthops and change_nexthops only used when nexthop object * is not set in a fib_info. The logic within can reference fib_nh. */ @@ -319,12 +315,9 @@ static inline int nh_comp(struct fib_info *fi, struct fib_info *ofi) return 0; } -static struct hlist_head * -fib_info_devhash_bucket(const struct net_device *dev) +static struct hlist_head *fib_nh_head(struct net_device *dev) { - u32 val = net_hash_mix(dev_net(dev)) ^ dev->ifindex; - - return &fib_info_devhash[hash_32(val, DEVINDEX_HASHBITS)]; + return &dev->fib_nh_head; } static unsigned int fib_info_hashfn_1(int init_val, u8 protocol, u8 scope, @@ -435,11 +428,11 @@ int ip_fib_check_default(__be32 gw, struct net_device *dev) struct hlist_head *head; struct fib_nh *nh; - head = fib_info_devhash_bucket(dev); + head = fib_nh_head(dev); hlist_for_each_entry_rcu(nh, head, nh_hash) { - if (nh->fib_nh_dev == dev && - nh->fib_nh_gw4 == gw && + DEBUG_NET_WARN_ON_ONCE(nh->fib_nh_dev != dev); + if (nh->fib_nh_gw4 == gw && !(nh->fib_nh_flags & RTNH_F_DEAD)) { return 0; } @@ -1595,7 +1588,7 @@ link_it: if (!nexthop_nh->fib_nh_dev) continue; - head = fib_info_devhash_bucket(nexthop_nh->fib_nh_dev); + head = fib_nh_head(nexthop_nh->fib_nh_dev); hlist_add_head_rcu(&nexthop_nh->nh_hash, head); } endfor_nexthops(fi) } @@ -1948,12 +1941,12 @@ void fib_nhc_update_mtu(struct fib_nh_common *nhc, u32 new, u32 orig) void fib_sync_mtu(struct net_device *dev, u32 orig_mtu) { - struct hlist_head *head = fib_info_devhash_bucket(dev); + struct hlist_head *head = fib_nh_head(dev); struct fib_nh *nh; hlist_for_each_entry(nh, head, nh_hash) { - if (nh->fib_nh_dev == dev) - fib_nhc_update_mtu(&nh->nh_common, dev->mtu, orig_mtu); + DEBUG_NET_WARN_ON_ONCE(nh->fib_nh_dev != dev); + fib_nhc_update_mtu(&nh->nh_common, dev->mtu, orig_mtu); } } @@ -1967,7 +1960,7 @@ void fib_sync_mtu(struct net_device *dev, u32 orig_mtu) */ int fib_sync_down_dev(struct net_device *dev, unsigned long event, bool force) { - struct hlist_head *head = fib_info_devhash_bucket(dev); + struct hlist_head *head = fib_nh_head(dev); struct fib_info *prev_fi = NULL; int scope = RT_SCOPE_NOWHERE; struct fib_nh *nh; @@ -1981,7 +1974,8 @@ int fib_sync_down_dev(struct net_device *dev, unsigned long event, bool force) int dead; BUG_ON(!fi->fib_nhs); - if (nh->fib_nh_dev != dev || fi == prev_fi) + DEBUG_NET_WARN_ON_ONCE(nh->fib_nh_dev != dev); + if (fi == prev_fi) continue; prev_fi = fi; dead = 0; @@ -2131,7 +2125,7 @@ int fib_sync_up(struct net_device *dev, unsigned char nh_flags) } prev_fi = NULL; - head = fib_info_devhash_bucket(dev); + head = fib_nh_head(dev); ret = 0; hlist_for_each_entry(nh, head, nh_hash) { @@ -2139,7 +2133,8 @@ int fib_sync_up(struct net_device *dev, unsigned char nh_flags) int alive; BUG_ON(!fi->fib_nhs); - if (nh->fib_nh_dev != dev || fi == prev_fi) + DEBUG_NET_WARN_ON_ONCE(nh->fib_nh_dev != dev); + if (fi == prev_fi) continue; prev_fi = fi; -- cgit v1.2.3 From 20a4da20e0bd0297bf1ce083362e78b6702f991e Mon Sep 17 00:00:00 2001 From: Oleksij Rempel Date: Fri, 4 Oct 2024 11:01:00 +0200 Subject: net: phy: Add support for PHY timing-role configuration via device tree Introduce support for configuring the master/slave role of PHYs based on the `timing-role` property in the device tree. While this functionality is necessary for Single Pair Ethernet (SPE) PHYs (1000/100/10Base-T1) where hardware strap pins may be unavailable or incorrectly set, it works for any PHY type. Signed-off-by: Oleksij Rempel Reviewed-by: Russell King (Oracle) Reviewed-by: Florian Fainelli Reviewed-by: Divya Koppera Signed-off-by: Paolo Abeni --- drivers/net/phy/phy-core.c | 33 +++++++++++++++++++++++++++++++++ drivers/net/phy/phy_device.c | 3 +++ include/linux/phy.h | 1 + 3 files changed, 37 insertions(+) (limited to 'include/linux') diff --git a/drivers/net/phy/phy-core.c b/drivers/net/phy/phy-core.c index 1f98b6a96c15..4e8db12d6092 100644 --- a/drivers/net/phy/phy-core.c +++ b/drivers/net/phy/phy-core.c @@ -412,6 +412,39 @@ void of_set_phy_eee_broken(struct phy_device *phydev) phydev->eee_broken_modes = broken; } +/** + * of_set_phy_timing_role - Set the master/slave mode of the PHY + * + * @phydev: The phy_device struct + * + * Set master/slave configuration of the PHY based on the device tree. + */ +void of_set_phy_timing_role(struct phy_device *phydev) +{ + struct device_node *node = phydev->mdio.dev.of_node; + const char *master; + + if (!IS_ENABLED(CONFIG_OF_MDIO)) + return; + + if (!node) + return; + + if (of_property_read_string(node, "timing-role", &master)) + return; + + if (strcmp(master, "forced-master") == 0) + phydev->master_slave_set = MASTER_SLAVE_CFG_MASTER_FORCE; + else if (strcmp(master, "forced-slave") == 0) + phydev->master_slave_set = MASTER_SLAVE_CFG_SLAVE_FORCE; + else if (strcmp(master, "preferred-master") == 0) + phydev->master_slave_set = MASTER_SLAVE_CFG_MASTER_PREFERRED; + else if (strcmp(master, "preferred-slave") == 0) + phydev->master_slave_set = MASTER_SLAVE_CFG_SLAVE_PREFERRED; + else + phydev_warn(phydev, "Unknown master-slave mode %s\n", master); +} + /** * phy_resolve_aneg_pause - Determine pause autoneg results * diff --git a/drivers/net/phy/phy_device.c b/drivers/net/phy/phy_device.c index 560e338b307a..4ccf504a8b2c 100644 --- a/drivers/net/phy/phy_device.c +++ b/drivers/net/phy/phy_device.c @@ -3608,6 +3608,9 @@ static int phy_probe(struct device *dev) */ of_set_phy_eee_broken(phydev); + /* Get master/slave strap overrides */ + of_set_phy_timing_role(phydev); + /* The Pause Frame bits indicate that the PHY can support passing * pause frames. During autonegotiation, the PHYs will determine if * they should allow pause frames to pass. The MAC driver should then diff --git a/include/linux/phy.h b/include/linux/phy.h index a98bc91a0cde..ff762a3d8270 100644 --- a/include/linux/phy.h +++ b/include/linux/phy.h @@ -1260,6 +1260,7 @@ size_t phy_speeds(unsigned int *speeds, size_t size, unsigned long *mask); void of_set_phy_supported(struct phy_device *phydev); void of_set_phy_eee_broken(struct phy_device *phydev); +void of_set_phy_timing_role(struct phy_device *phydev); int phy_speed_down_core(struct phy_device *phydev); /** -- cgit v1.2.3 From ec763c234d7f60c5bce0fa2611ba79f5be1af76b Mon Sep 17 00:00:00 2001 From: Kuniyuki Iwashima Date: Fri, 4 Oct 2024 15:10:28 -0700 Subject: Revert "rtnetlink: add guard for RTNL" This reverts commit 464eb03c4a7cfb32cb3324249193cf6bb5b35152. Once we have a per-netns RTNL, we won't use guard(rtnl). Also, there's no users for now. $ grep -rnI "guard(rtnl" || true $ Suggested-by: Eric Dumazet Link: https://lore.kernel.org/netdev/CANn89i+KoYzUH+VPLdGmLABYf5y4TW0hrM4UAeQQJ9AREty0iw@mail.gmail.com/ Signed-off-by: Kuniyuki Iwashima Reviewed-by: Eric Dumazet Signed-off-by: Paolo Abeni --- include/linux/rtnetlink.h | 3 --- 1 file changed, 3 deletions(-) (limited to 'include/linux') diff --git a/include/linux/rtnetlink.h b/include/linux/rtnetlink.h index a7da7dfc06a2..cdfc897f1e3c 100644 --- a/include/linux/rtnetlink.h +++ b/include/linux/rtnetlink.h @@ -7,7 +7,6 @@ #include #include #include -#include #include extern int rtnetlink_send(struct sk_buff *skb, struct net *net, u32 pid, u32 group, int echo); @@ -47,8 +46,6 @@ extern int rtnl_is_locked(void); extern int rtnl_lock_killable(void); extern bool refcount_dec_and_rtnl_lock(refcount_t *r); -DEFINE_LOCK_GUARD_0(rtnl, rtnl_lock(), rtnl_unlock()) - extern wait_queue_head_t netdev_unregistering_wq; extern atomic_t dev_unreg_count; extern struct rw_semaphore pernet_ops_rwsem; -- cgit v1.2.3 From 76aed95319da25d6884dff01d5f0149e4b542f96 Mon Sep 17 00:00:00 2001 From: Kuniyuki Iwashima Date: Fri, 4 Oct 2024 15:10:29 -0700 Subject: rtnetlink: Add per-netns RTNL. The goal is to break RTNL down into per-netns mutex. This patch adds per-netns mutex and its helper functions, rtnl_net_lock() and rtnl_net_unlock(). rtnl_net_lock() acquires the global RTNL and per-netns RTNL mutex, and rtnl_net_unlock() releases them. We will replace 800+ rtnl_lock() with rtnl_net_lock() and finally removes rtnl_lock() in rtnl_net_lock(). When we need to nest per-netns RTNL mutex, we will use __rtnl_net_lock(), and its locking order is defined by rtnl_net_lock_cmp_fn() as follows: 1. init_net is first 2. netns address ascending order Note that the conversion will be done under CONFIG_DEBUG_NET_SMALL_RTNL with LOCKDEP so that we can carefully add the extra mutex without slowing down RTNL operations during conversion. Signed-off-by: Kuniyuki Iwashima Reviewed-by: Eric Dumazet Signed-off-by: Paolo Abeni --- include/linux/rtnetlink.h | 21 ++++++++++++++++ include/net/net_namespace.h | 4 ++++ net/Kconfig.debug | 15 ++++++++++++ net/core/net_namespace.c | 6 +++++ net/core/rtnetlink.c | 58 +++++++++++++++++++++++++++++++++++++++++++++ 5 files changed, 104 insertions(+) (limited to 'include/linux') diff --git a/include/linux/rtnetlink.h b/include/linux/rtnetlink.h index cdfc897f1e3c..edd840a49989 100644 --- a/include/linux/rtnetlink.h +++ b/include/linux/rtnetlink.h @@ -92,6 +92,27 @@ static inline bool lockdep_rtnl_is_held(void) #define rcu_replace_pointer_rtnl(rp, p) \ rcu_replace_pointer(rp, p, lockdep_rtnl_is_held()) +#ifdef CONFIG_DEBUG_NET_SMALL_RTNL +void __rtnl_net_lock(struct net *net); +void __rtnl_net_unlock(struct net *net); +void rtnl_net_lock(struct net *net); +void rtnl_net_unlock(struct net *net); +int rtnl_net_lock_cmp_fn(const struct lockdep_map *a, const struct lockdep_map *b); +#else +static inline void __rtnl_net_lock(struct net *net) {} +static inline void __rtnl_net_unlock(struct net *net) {} + +static inline void rtnl_net_lock(struct net *net) +{ + rtnl_lock(); +} + +static inline void rtnl_net_unlock(struct net *net) +{ + rtnl_unlock(); +} +#endif + static inline struct netdev_queue *dev_ingress_queue(struct net_device *dev) { return rtnl_dereference(dev->ingress_queue); diff --git a/include/net/net_namespace.h b/include/net/net_namespace.h index e67b483cc8bb..873c0f9fdac6 100644 --- a/include/net/net_namespace.h +++ b/include/net/net_namespace.h @@ -188,6 +188,10 @@ struct net { #if IS_ENABLED(CONFIG_SMC) struct netns_smc smc; #endif +#ifdef CONFIG_DEBUG_NET_SMALL_RTNL + /* Move to a better place when the config guard is removed. */ + struct mutex rtnl_mutex; +#endif } __randomize_layout; #include diff --git a/net/Kconfig.debug b/net/Kconfig.debug index 5e3fffe707dd..277fab8c4d77 100644 --- a/net/Kconfig.debug +++ b/net/Kconfig.debug @@ -24,3 +24,18 @@ config DEBUG_NET help Enable extra sanity checks in networking. This is mostly used by fuzzers, but is safe to select. + +config DEBUG_NET_SMALL_RTNL + bool "Add extra per-netns mutex inside RTNL" + depends on DEBUG_KERNEL && NET && LOCK_DEBUGGING_SUPPORT + select PROVE_LOCKING + default n + help + rtnl_lock() is being replaced with rtnl_net_lock() that + acquires the global RTNL and a small per-netns RTNL mutex. + + During the conversion, rtnl_net_lock() just adds an extra + mutex in every RTNL scope and slows down the operations. + + Once the conversion completes, rtnl_lock() will be removed + and rtnetlink will gain per-netns scalability. diff --git a/net/core/net_namespace.c b/net/core/net_namespace.c index e39479f1c9a4..105e3cd26763 100644 --- a/net/core/net_namespace.c +++ b/net/core/net_namespace.c @@ -334,6 +334,12 @@ static __net_init void preinit_net(struct net *net, struct user_namespace *user_ idr_init(&net->netns_ids); spin_lock_init(&net->nsid_lock); mutex_init(&net->ipv4.ra_mutex); + +#ifdef CONFIG_DEBUG_NET_SMALL_RTNL + mutex_init(&net->rtnl_mutex); + lock_set_cmp_fn(&net->rtnl_mutex, rtnl_net_lock_cmp_fn, NULL); +#endif + preinit_net_sysctl(net); } diff --git a/net/core/rtnetlink.c b/net/core/rtnetlink.c index 682d8d3127db..46567fa54e42 100644 --- a/net/core/rtnetlink.c +++ b/net/core/rtnetlink.c @@ -179,6 +179,64 @@ bool lockdep_rtnl_is_held(void) EXPORT_SYMBOL(lockdep_rtnl_is_held); #endif /* #ifdef CONFIG_PROVE_LOCKING */ +#ifdef CONFIG_DEBUG_NET_SMALL_RTNL +void __rtnl_net_lock(struct net *net) +{ + ASSERT_RTNL(); + + mutex_lock(&net->rtnl_mutex); +} +EXPORT_SYMBOL(__rtnl_net_lock); + +void __rtnl_net_unlock(struct net *net) +{ + ASSERT_RTNL(); + + mutex_unlock(&net->rtnl_mutex); +} +EXPORT_SYMBOL(__rtnl_net_unlock); + +void rtnl_net_lock(struct net *net) +{ + rtnl_lock(); + __rtnl_net_lock(net); +} +EXPORT_SYMBOL(rtnl_net_lock); + +void rtnl_net_unlock(struct net *net) +{ + __rtnl_net_unlock(net); + rtnl_unlock(); +} +EXPORT_SYMBOL(rtnl_net_unlock); + +static int rtnl_net_cmp_locks(const struct net *net_a, const struct net *net_b) +{ + if (net_eq(net_a, net_b)) + return 0; + + /* always init_net first */ + if (net_eq(net_a, &init_net)) + return -1; + + if (net_eq(net_b, &init_net)) + return 1; + + /* otherwise lock in ascending order */ + return net_a < net_b ? -1 : 1; +} + +int rtnl_net_lock_cmp_fn(const struct lockdep_map *a, const struct lockdep_map *b) +{ + const struct net *net_a, *net_b; + + net_a = container_of(a, struct net, rtnl_mutex.dep_map); + net_b = container_of(b, struct net, rtnl_mutex.dep_map); + + return rtnl_net_cmp_locks(net_a, net_b); +} +#endif + static struct rtnl_link __rcu *__rcu *rtnl_msg_handlers[RTNL_FAMILY_MAX + 1]; static inline int rtm_msgindex(int msgtype) -- cgit v1.2.3 From 844e5e7e656d3a7a904fd5607f8491d6fd01db8e Mon Sep 17 00:00:00 2001 From: Kuniyuki Iwashima Date: Fri, 4 Oct 2024 15:10:30 -0700 Subject: rtnetlink: Add assertion helpers for per-netns RTNL. Once an RTNL scope is converted with rtnl_net_lock(), we will replace RTNL helper functions inside the scope with the following per-netns alternatives: ASSERT_RTNL() -> ASSERT_RTNL_NET(net) rcu_dereference_rtnl(p) -> rcu_dereference_rtnl_net(net, p) Note that the per-netns helpers are equivalent to the conventional helpers unless CONFIG_DEBUG_NET_SMALL_RTNL is enabled. Signed-off-by: Kuniyuki Iwashima Reviewed-by: Eric Dumazet Signed-off-by: Paolo Abeni --- include/linux/rtnetlink.h | 45 +++++++++++++++++++++++++++++++++++++++++---- net/core/rtnetlink.c | 12 ++++++++++++ 2 files changed, 53 insertions(+), 4 deletions(-) (limited to 'include/linux') diff --git a/include/linux/rtnetlink.h b/include/linux/rtnetlink.h index edd840a49989..8468a4ce8510 100644 --- a/include/linux/rtnetlink.h +++ b/include/linux/rtnetlink.h @@ -51,6 +51,10 @@ extern atomic_t dev_unreg_count; extern struct rw_semaphore pernet_ops_rwsem; extern struct rw_semaphore net_rwsem; +#define ASSERT_RTNL() \ + WARN_ONCE(!rtnl_is_locked(), \ + "RTNL: assertion failed at %s (%d)\n", __FILE__, __LINE__) + #ifdef CONFIG_PROVE_LOCKING extern bool lockdep_rtnl_is_held(void); #else @@ -98,6 +102,22 @@ void __rtnl_net_unlock(struct net *net); void rtnl_net_lock(struct net *net); void rtnl_net_unlock(struct net *net); int rtnl_net_lock_cmp_fn(const struct lockdep_map *a, const struct lockdep_map *b); + +bool rtnl_net_is_locked(struct net *net); + +#define ASSERT_RTNL_NET(net) \ + WARN_ONCE(!rtnl_net_is_locked(net), \ + "RTNL_NET: assertion failed at %s (%d)\n", \ + __FILE__, __LINE__) + +bool lockdep_rtnl_net_is_held(struct net *net); + +#define rcu_dereference_rtnl_net(net, p) \ + rcu_dereference_check(p, lockdep_rtnl_net_is_held(net)) +#define rtnl_net_dereference(net, p) \ + rcu_dereference_protected(p, lockdep_rtnl_net_is_held(net)) +#define rcu_replace_pointer_rtnl_net(net, rp, p) \ + rcu_replace_pointer(rp, p, lockdep_rtnl_net_is_held(net)) #else static inline void __rtnl_net_lock(struct net *net) {} static inline void __rtnl_net_unlock(struct net *net) {} @@ -111,6 +131,27 @@ static inline void rtnl_net_unlock(struct net *net) { rtnl_unlock(); } + +static inline void ASSERT_RTNL_NET(struct net *net) +{ + ASSERT_RTNL(); +} + +static inline void *rcu_dereference_rtnl_net(struct net *net, void *p) +{ + return rcu_dereference_rtnl(p); +} + +static inline void *rtnl_net_dereference(struct net *net, void *p) +{ + return rtnl_dereference(p); +} + +static inline void *rcu_replace_pointer_rtnl_net(struct net *net, + void *rp, void *p) +{ + return rcu_replace_pointer_rtnl(rp, p); +} #endif static inline struct netdev_queue *dev_ingress_queue(struct net_device *dev) @@ -140,10 +181,6 @@ void rtnetlink_init(void); void __rtnl_unlock(void); void rtnl_kfree_skbs(struct sk_buff *head, struct sk_buff *tail); -#define ASSERT_RTNL() \ - WARN_ONCE(!rtnl_is_locked(), \ - "RTNL: assertion failed at %s (%d)\n", __FILE__, __LINE__) - extern int ndo_dflt_fdb_dump(struct sk_buff *skb, struct netlink_callback *cb, struct net_device *dev, diff --git a/net/core/rtnetlink.c b/net/core/rtnetlink.c index 46567fa54e42..6d68247aea70 100644 --- a/net/core/rtnetlink.c +++ b/net/core/rtnetlink.c @@ -235,6 +235,18 @@ int rtnl_net_lock_cmp_fn(const struct lockdep_map *a, const struct lockdep_map * return rtnl_net_cmp_locks(net_a, net_b); } + +bool rtnl_net_is_locked(struct net *net) +{ + return rtnl_is_locked() && mutex_is_locked(&net->rtnl_mutex); +} +EXPORT_SYMBOL(rtnl_net_is_locked); + +bool lockdep_rtnl_net_is_held(struct net *net) +{ + return lockdep_rtnl_is_held() && lockdep_is_held(&net->rtnl_mutex); +} +EXPORT_SYMBOL(lockdep_rtnl_net_is_held); #endif static struct rtnl_link __rcu *__rcu *rtnl_msg_handlers[RTNL_FAMILY_MAX + 1]; -- cgit v1.2.3 From 836265d31631e28000fc8917ce697fc687a58724 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Mon, 7 Oct 2024 21:35:25 +0200 Subject: wifi: remove iw_public_data from struct net_device Given the previous patches, we no longer need the struct iw_public_data etc., it's only used by the old Intel drivers (and ps3_gelic creates it but then doesn't use it). Remove all of that, including the pointer in struct net_device. Link: https://patch.msgid.link/20241007213525.8b2d52b60531.I6a27aaf30bded9a0977f07f47fba2bd31a3b3330@changeid Signed-off-by: Johannes Berg --- Documentation/networking/net_cachelines/net_device.rst | 1 - drivers/net/ethernet/toshiba/ps3_gelic_wireless.c | 1 - drivers/net/ethernet/toshiba/ps3_gelic_wireless.h | 1 - drivers/net/wireless/intel/ipw2x00/ipw2100.c | 2 -- drivers/net/wireless/intel/ipw2x00/ipw2100.h | 2 -- drivers/net/wireless/intel/ipw2x00/ipw2200.c | 1 - drivers/net/wireless/intel/ipw2x00/ipw2200.h | 2 -- drivers/net/wireless/intel/ipw2x00/libipw_spy.c | 7 ++++--- include/linux/netdevice.h | 2 -- include/net/iw_handler.h | 18 ------------------ 10 files changed, 4 insertions(+), 33 deletions(-) (limited to 'include/linux') diff --git a/Documentation/networking/net_cachelines/net_device.rst b/Documentation/networking/net_cachelines/net_device.rst index 22b07c814f4a..b7d1d4c7a1ab 100644 --- a/Documentation/networking/net_cachelines/net_device.rst +++ b/Documentation/networking/net_cachelines/net_device.rst @@ -51,7 +51,6 @@ struct_net_device_core_stats* core_stats atomic_t carrier_up_count atomic_t carrier_down_count struct_iw_handler_def* wireless_handlers -struct_iw_public_data* wireless_data struct_ethtool_ops* ethtool_ops struct_l3mdev_ops* l3mdev_ops struct_ndisc_ops* ndisc_ops diff --git a/drivers/net/ethernet/toshiba/ps3_gelic_wireless.c b/drivers/net/ethernet/toshiba/ps3_gelic_wireless.c index 44488c153ea2..4fbe4b7cd12a 100644 --- a/drivers/net/ethernet/toshiba/ps3_gelic_wireless.c +++ b/drivers/net/ethernet/toshiba/ps3_gelic_wireless.c @@ -2566,7 +2566,6 @@ static void gelic_wl_setup_netdev_ops(struct net_device *netdev) netdev->ethtool_ops = &gelic_wl_ethtool_ops; netdev->netdev_ops = &gelic_wl_netdevice_ops; - netdev->wireless_data = &wl->wireless_data; netdev->wireless_handlers = &gelic_wl_wext_handler_def; } diff --git a/drivers/net/ethernet/toshiba/ps3_gelic_wireless.h b/drivers/net/ethernet/toshiba/ps3_gelic_wireless.h index 1f203d1ae8db..dbabf538e10a 100644 --- a/drivers/net/ethernet/toshiba/ps3_gelic_wireless.h +++ b/drivers/net/ethernet/toshiba/ps3_gelic_wireless.h @@ -276,7 +276,6 @@ struct gelic_wl_info { u8 active_bssid[ETH_ALEN]; /* associated bssid */ unsigned int essid_len; - struct iw_public_data wireless_data; struct iw_statistics iwstat; }; diff --git a/drivers/net/wireless/intel/ipw2x00/ipw2100.c b/drivers/net/wireless/intel/ipw2x00/ipw2100.c index a89e06c1b8ee..15bf35f2de2a 100644 --- a/drivers/net/wireless/intel/ipw2x00/ipw2100.c +++ b/drivers/net/wireless/intel/ipw2x00/ipw2100.c @@ -6022,8 +6022,6 @@ static struct net_device *ipw2100_alloc_device(struct pci_dev *pci_dev, dev->netdev_ops = &ipw2100_netdev_ops; dev->ethtool_ops = &ipw2100_ethtool_ops; dev->wireless_handlers = &ipw2100_wx_handler_def; - priv->wireless_data.libipw = priv->ieee; - dev->wireless_data = &priv->wireless_data; dev->watchdog_timeo = 3 * HZ; dev->irq = 0; dev->min_mtu = 68; diff --git a/drivers/net/wireless/intel/ipw2x00/ipw2100.h b/drivers/net/wireless/intel/ipw2x00/ipw2100.h index b34085ade3aa..6f81f509b9cb 100644 --- a/drivers/net/wireless/intel/ipw2x00/ipw2100.h +++ b/drivers/net/wireless/intel/ipw2x00/ipw2100.h @@ -554,8 +554,6 @@ struct ipw2100_priv { struct net_device *net_dev; struct iw_statistics wstats; - struct iw_public_data wireless_data; - struct tasklet_struct irq_tasklet; struct delayed_work reset_work; diff --git a/drivers/net/wireless/intel/ipw2x00/ipw2200.c b/drivers/net/wireless/intel/ipw2x00/ipw2200.c index c0e9d2109e34..be1d971b3d32 100644 --- a/drivers/net/wireless/intel/ipw2x00/ipw2200.c +++ b/drivers/net/wireless/intel/ipw2x00/ipw2200.c @@ -11645,7 +11645,6 @@ static int ipw_pci_probe(struct pci_dev *pdev, net_dev->netdev_ops = &ipw_netdev_ops; priv->ieee->spy_enabled = true; - net_dev->wireless_data = &priv->wireless_data; net_dev->wireless_handlers = &ipw_wx_handler_def; net_dev->ethtool_ops = &ipw_ethtool_ops; diff --git a/drivers/net/wireless/intel/ipw2x00/ipw2200.h b/drivers/net/wireless/intel/ipw2x00/ipw2200.h index 46f119123b49..d3db54e6d37c 100644 --- a/drivers/net/wireless/intel/ipw2x00/ipw2200.h +++ b/drivers/net/wireless/intel/ipw2x00/ipw2200.h @@ -1274,8 +1274,6 @@ struct ipw_priv { struct iw_statistics wstats; - struct iw_public_data wireless_data; - int user_requested_scan; u8 direct_scan_ssid[IW_ESSID_MAX_SIZE]; u8 direct_scan_ssid_len; diff --git a/drivers/net/wireless/intel/ipw2x00/libipw_spy.c b/drivers/net/wireless/intel/ipw2x00/libipw_spy.c index 979aeb10aeeb..ba876e92f7f6 100644 --- a/drivers/net/wireless/intel/ipw2x00/libipw_spy.c +++ b/drivers/net/wireless/intel/ipw2x00/libipw_spy.c @@ -18,9 +18,10 @@ static struct iw_spy_data *get_spydata(struct net_device *dev) { - if (dev->wireless_data && dev->wireless_data->libipw && - dev->wireless_data->libipw->spy_enabled) - return &dev->wireless_data->libipw->spy_data; + struct libipw_device *ieee = netdev_priv(dev); + + if (ieee->spy_enabled) + return &ieee->spy_data; return NULL; } diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index e87b5e488325..f0abed1e6e45 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -1773,7 +1773,6 @@ enum netdev_reg_state { * @wireless_handlers: List of functions to handle Wireless Extensions, * instead of ioctl, * see for details. - * @wireless_data: Instance data managed by the core of wireless extensions * * @netdev_ops: Includes several pointers to callbacks, * if one wants to override the ndo_*() functions @@ -2150,7 +2149,6 @@ struct net_device { #ifdef CONFIG_WIRELESS_EXT const struct iw_handler_def *wireless_handlers; - struct iw_public_data *wireless_data; #endif const struct ethtool_ops *ethtool_ops; #ifdef CONFIG_NET_L3_MASTER_DEV diff --git a/include/net/iw_handler.h b/include/net/iw_handler.h index a7b502958d27..fc44fcca1d5c 100644 --- a/include/net/iw_handler.h +++ b/include/net/iw_handler.h @@ -404,24 +404,6 @@ struct iw_spy_data { u_char spy_thr_under[IW_MAX_SPY]; }; -/* --------------------- DEVICE WIRELESS DATA --------------------- */ -/* - * This is all the wireless data specific to a device instance that - * is managed by the core of Wireless Extensions or the 802.11 layer. - * We only keep pointer to those structures, so that a driver is free - * to share them between instances. - * This structure should be initialised before registering the device. - * Access to this data follow the same rules as any other struct net_device - * data (i.e. valid as long as struct net_device exist, same locking rules). - */ -/* Forward declaration */ -struct libipw_device; -/* The struct */ -struct iw_public_data { - /* Legacy structure managed by the ipw2x00-specific IEEE 802.11 layer */ - struct libipw_device * libipw; -}; - /**************************** PROTOTYPES ****************************/ /* * Functions part of the Wireless Extensions (defined in net/wireless/wext-core.c). -- cgit v1.2.3 From 9e1a98aac11b757e2634304bbe6bed5f5a58e0ee Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Mon, 7 Oct 2024 21:50:25 +0200 Subject: wifi: wext: merge adjacent CONFIG_COMPAT ifdef blocks Simplify this, and also add a comment at the #endif. Link: https://patch.msgid.link/20241007215025.5ecdad1e02ed.I54efa895efc496e06ba41e1c39c9df9e23b0171f@changeid Signed-off-by: Johannes Berg --- include/linux/wireless.h | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) (limited to 'include/linux') diff --git a/include/linux/wireless.h b/include/linux/wireless.h index e6e34d74dda0..03e5d3fe226d 100644 --- a/include/linux/wireless.h +++ b/include/linux/wireless.h @@ -21,8 +21,7 @@ struct compat_iw_point { __u16 length; __u16 flags; }; -#endif -#ifdef CONFIG_COMPAT + struct __compat_iw_event { __u16 len; /* Real length of this stuff */ __u16 cmd; /* Wireless IOCTL */ @@ -49,5 +48,5 @@ struct __compat_iw_event { #define IW_EV_COMPAT_POINT_LEN \ (IW_EV_COMPAT_LCP_LEN + sizeof(struct compat_iw_point) - \ IW_EV_COMPAT_POINT_OFF) -#endif +#endif /* CONFIG_COMPAT */ #endif /* _LINUX_WIRELESS_H */ -- cgit v1.2.3 From f042365a26b07006064c2cfa2293e16cad243b0d Mon Sep 17 00:00:00 2001 From: "Russell King (Oracle)" Date: Fri, 4 Oct 2024 11:20:56 +0100 Subject: net: pcs: xpcs: provide a helper to get the phylink pcs given xpcs Provide a helper to provide the pointer to the phylink_pcs struct given a valid xpcs pointer. This will be necessary when we make struct dw_xpcs private to pcs-xpcs.c Signed-off-by: Russell King (Oracle) Signed-off-by: David S. Miller --- drivers/net/ethernet/stmicro/stmmac/dwmac-intel.c | 2 +- drivers/net/pcs/pcs-xpcs.c | 6 ++++++ include/linux/pcs/pcs-xpcs.h | 1 + 3 files changed, 8 insertions(+), 1 deletion(-) (limited to 'include/linux') diff --git a/drivers/net/ethernet/stmicro/stmmac/dwmac-intel.c b/drivers/net/ethernet/stmicro/stmmac/dwmac-intel.c index 83ad7c7935e3..48acba5eb178 100644 --- a/drivers/net/ethernet/stmicro/stmmac/dwmac-intel.c +++ b/drivers/net/ethernet/stmicro/stmmac/dwmac-intel.c @@ -451,7 +451,7 @@ static struct phylink_pcs *intel_mgbe_select_pcs(struct stmmac_priv *priv, * should always be an XPCS. The original code would always * return this if present. */ - return &priv->hw->xpcs->pcs; + return xpcs_to_phylink_pcs(priv->hw->xpcs); } static int intel_mgbe_common_data(struct pci_dev *pdev, diff --git a/drivers/net/pcs/pcs-xpcs.c b/drivers/net/pcs/pcs-xpcs.c index 8bde87ab971f..a7f6d56183a7 100644 --- a/drivers/net/pcs/pcs-xpcs.c +++ b/drivers/net/pcs/pcs-xpcs.c @@ -132,6 +132,12 @@ xpcs_find_compat(struct dw_xpcs *xpcs, phy_interface_t interface) return NULL; } +struct phylink_pcs *xpcs_to_phylink_pcs(struct dw_xpcs *xpcs) +{ + return &xpcs->pcs; +} +EXPORT_SYMBOL_GPL(xpcs_to_phylink_pcs); + int xpcs_get_an_mode(struct dw_xpcs *xpcs, phy_interface_t interface) { const struct dw_xpcs_compat *compat; diff --git a/include/linux/pcs/pcs-xpcs.h b/include/linux/pcs/pcs-xpcs.h index abda475111d1..868515f3cc88 100644 --- a/include/linux/pcs/pcs-xpcs.h +++ b/include/linux/pcs/pcs-xpcs.h @@ -64,6 +64,7 @@ struct dw_xpcs { bool need_reset; }; +struct phylink_pcs *xpcs_to_phylink_pcs(struct dw_xpcs *xpcs); int xpcs_get_an_mode(struct dw_xpcs *xpcs, phy_interface_t interface); void xpcs_get_interfaces(struct dw_xpcs *xpcs, unsigned long *interfaces); int xpcs_config_eee(struct dw_xpcs *xpcs, int mult_fact_100ns, -- cgit v1.2.3 From accd5f5cd2e1b0ed6805a7c24765648d213561e5 Mon Sep 17 00:00:00 2001 From: "Russell King (Oracle)" Date: Fri, 4 Oct 2024 11:21:01 +0100 Subject: net: pcs: xpcs: move definition of struct dw_xpcs to private header There should be no reason for anything outside the XPCS code to know the contents of struct dw_xpcs - this is a private structure to XPCS. Move the definition to the private pcs-xpcs.h header, leaving a declaration in the global pcs/pcs-xpcs.h Signed-off-by: Russell King (Oracle) Signed-off-by: David S. Miller --- drivers/net/pcs/pcs-xpcs.h | 18 ++++++++++++++++++ include/linux/pcs/pcs-xpcs.h | 18 +----------------- 2 files changed, 19 insertions(+), 17 deletions(-) (limited to 'include/linux') diff --git a/drivers/net/pcs/pcs-xpcs.h b/drivers/net/pcs/pcs-xpcs.h index fa05adfae220..1b546eae8280 100644 --- a/drivers/net/pcs/pcs-xpcs.h +++ b/drivers/net/pcs/pcs-xpcs.h @@ -123,6 +123,24 @@ #define DW_XPCS_INFO_DECLARE(_name, _pcs, _pma) \ static const struct dw_xpcs_info _name = { .pcs = _pcs, .pma = _pma } +struct dw_xpcs_desc; + +enum dw_xpcs_clock { + DW_XPCS_CORE_CLK, + DW_XPCS_PAD_CLK, + DW_XPCS_NUM_CLKS, +}; + +struct dw_xpcs { + struct dw_xpcs_info info; + const struct dw_xpcs_desc *desc; + struct mdio_device *mdiodev; + struct clk_bulk_data clks[DW_XPCS_NUM_CLKS]; + struct phylink_pcs pcs; + phy_interface_t interface; + bool need_reset; +}; + int xpcs_read(struct dw_xpcs *xpcs, int dev, u32 reg); int xpcs_write(struct dw_xpcs *xpcs, int dev, u32 reg, u16 val); int xpcs_read_vpcs(struct dw_xpcs *xpcs, int reg); diff --git a/include/linux/pcs/pcs-xpcs.h b/include/linux/pcs/pcs-xpcs.h index 868515f3cc88..b5b5d17998b8 100644 --- a/include/linux/pcs/pcs-xpcs.h +++ b/include/linux/pcs/pcs-xpcs.h @@ -21,8 +21,6 @@ #define DW_AN_C37_1000BASEX 4 #define DW_10GBASER 5 -struct dw_xpcs_desc; - enum dw_xpcs_pcs_id { DW_XPCS_ID_NATIVE = 0, NXP_SJA1105_XPCS_ID = 0x00000010, @@ -48,21 +46,7 @@ struct dw_xpcs_info { u32 pma; }; -enum dw_xpcs_clock { - DW_XPCS_CORE_CLK, - DW_XPCS_PAD_CLK, - DW_XPCS_NUM_CLKS, -}; - -struct dw_xpcs { - struct dw_xpcs_info info; - const struct dw_xpcs_desc *desc; - struct mdio_device *mdiodev; - struct clk_bulk_data clks[DW_XPCS_NUM_CLKS]; - struct phylink_pcs pcs; - phy_interface_t interface; - bool need_reset; -}; +struct dw_xpcs; struct phylink_pcs *xpcs_to_phylink_pcs(struct dw_xpcs *xpcs); int xpcs_get_an_mode(struct dw_xpcs *xpcs, phy_interface_t interface); -- cgit v1.2.3 From 87173021f1583ee37f4801fcde354729da8db3dc Mon Sep 17 00:00:00 2001 From: Kuniyuki Iwashima Date: Tue, 8 Oct 2024 10:29:03 -0700 Subject: ipv4: Link IPv4 address to per-netns hash table. As a prep for per-netns RTNL conversion, we want to namespacify the IPv4 address hash table and the GC work. Let's allocate the per-netns IPv4 address hash table to net->ipv4.inet_addr_lst and link IPv4 addresses into it. The actual users will be converted later. Note that the IPv6 address hash table is already namespacified. Reviewed-by: Eric Dumazet Signed-off-by: Kuniyuki Iwashima Link: https://patch.msgid.link/20241008172906.1326-2-kuniyu@amazon.com Signed-off-by: Jakub Kicinski --- include/linux/inetdevice.h | 1 + include/net/netns/ipv4.h | 1 + net/ipv4/devinet.c | 22 +++++++++++++++++++--- 3 files changed, 21 insertions(+), 3 deletions(-) (limited to 'include/linux') diff --git a/include/linux/inetdevice.h b/include/linux/inetdevice.h index cb5280e6cc21..d0c2bf67a9b0 100644 --- a/include/linux/inetdevice.h +++ b/include/linux/inetdevice.h @@ -142,6 +142,7 @@ static inline void ipv4_devconf_setall(struct in_device *in_dev) struct in_ifaddr { struct hlist_node hash; + struct hlist_node addr_lst; struct in_ifaddr __rcu *ifa_next; struct in_device *ifa_dev; struct rcu_head rcu_head; diff --git a/include/net/netns/ipv4.h b/include/net/netns/ipv4.h index 276f622f3516..29eba2eaaa26 100644 --- a/include/net/netns/ipv4.h +++ b/include/net/netns/ipv4.h @@ -270,5 +270,6 @@ struct netns_ipv4 { atomic_t rt_genid; siphash_key_t ip_id_key; + struct hlist_head *inet_addr_lst; }; #endif diff --git a/net/ipv4/devinet.c b/net/ipv4/devinet.c index ab76744383cf..059807a627a6 100644 --- a/net/ipv4/devinet.c +++ b/net/ipv4/devinet.c @@ -134,11 +134,13 @@ static void inet_hash_insert(struct net *net, struct in_ifaddr *ifa) ASSERT_RTNL(); hlist_add_head_rcu(&ifa->hash, &inet_addr_lst[hash]); + hlist_add_head_rcu(&ifa->addr_lst, &net->ipv4.inet_addr_lst[hash]); } static void inet_hash_remove(struct in_ifaddr *ifa) { ASSERT_RTNL(); + hlist_del_init_rcu(&ifa->addr_lst); hlist_del_init_rcu(&ifa->hash); } @@ -228,6 +230,7 @@ static struct in_ifaddr *inet_alloc_ifa(struct in_device *in_dev) ifa->ifa_dev = in_dev; INIT_HLIST_NODE(&ifa->hash); + INIT_HLIST_NODE(&ifa->addr_lst); return ifa; } @@ -2663,14 +2666,21 @@ static struct ctl_table ctl_forward_entry[] = { static __net_init int devinet_init_net(struct net *net) { - int err; - struct ipv4_devconf *all, *dflt; #ifdef CONFIG_SYSCTL - struct ctl_table *tbl; struct ctl_table_header *forw_hdr; + struct ctl_table *tbl; #endif + struct ipv4_devconf *all, *dflt; + int err; + int i; err = -ENOMEM; + net->ipv4.inet_addr_lst = kmalloc_array(IN4_ADDR_HSIZE, + sizeof(struct hlist_head), + GFP_KERNEL); + if (!net->ipv4.inet_addr_lst) + goto err_alloc_hash; + all = kmemdup(&ipv4_devconf, sizeof(ipv4_devconf), GFP_KERNEL); if (!all) goto err_alloc_all; @@ -2731,6 +2741,9 @@ static __net_init int devinet_init_net(struct net *net) net->ipv4.forw_hdr = forw_hdr; #endif + for (i = 0; i < IN4_ADDR_HSIZE; i++) + INIT_HLIST_HEAD(&net->ipv4.inet_addr_lst[i]); + net->ipv4.devconf_all = all; net->ipv4.devconf_dflt = dflt; return 0; @@ -2748,6 +2761,8 @@ err_alloc_ctl: err_alloc_dflt: kfree(all); err_alloc_all: + kfree(net->ipv4.inet_addr_lst); +err_alloc_hash: return err; } @@ -2766,6 +2781,7 @@ static __net_exit void devinet_exit_net(struct net *net) #endif kfree(net->ipv4.devconf_dflt); kfree(net->ipv4.devconf_all); + kfree(net->ipv4.inet_addr_lst); } static __net_initdata struct pernet_operations devinet_ops = { -- cgit v1.2.3 From 99ee348e6a41cf24b334a1bb7cde87239e8e2d95 Mon Sep 17 00:00:00 2001 From: Kuniyuki Iwashima Date: Tue, 8 Oct 2024 10:29:06 -0700 Subject: ipv4: Retire global IPv4 hash table inet_addr_lst. No one uses inet_addr_lst anymore, so let's remove it. Reviewed-by: Eric Dumazet Signed-off-by: Kuniyuki Iwashima Link: https://patch.msgid.link/20241008172906.1326-5-kuniyu@amazon.com Signed-off-by: Jakub Kicinski --- include/linux/inetdevice.h | 1 - net/ipv4/devinet.c | 10 ---------- 2 files changed, 11 deletions(-) (limited to 'include/linux') diff --git a/include/linux/inetdevice.h b/include/linux/inetdevice.h index d0c2bf67a9b0..d9c690c8c80b 100644 --- a/include/linux/inetdevice.h +++ b/include/linux/inetdevice.h @@ -141,7 +141,6 @@ static inline void ipv4_devconf_setall(struct in_device *in_dev) ARP_EVICT_NOCARRIER) struct in_ifaddr { - struct hlist_node hash; struct hlist_node addr_lst; struct in_ifaddr __rcu *ifa_next; struct in_device *ifa_dev; diff --git a/net/ipv4/devinet.c b/net/ipv4/devinet.c index ac245944e89e..7c156f85b7d2 100644 --- a/net/ipv4/devinet.c +++ b/net/ipv4/devinet.c @@ -119,8 +119,6 @@ struct inet_fill_args { #define IN4_ADDR_HSIZE_SHIFT 8 #define IN4_ADDR_HSIZE (1U << IN4_ADDR_HSIZE_SHIFT) -static struct hlist_head inet_addr_lst[IN4_ADDR_HSIZE]; - static u32 inet_addr_hash(const struct net *net, __be32 addr) { u32 val = (__force u32) addr ^ net_hash_mix(net); @@ -133,7 +131,6 @@ static void inet_hash_insert(struct net *net, struct in_ifaddr *ifa) u32 hash = inet_addr_hash(net, ifa->ifa_local); ASSERT_RTNL(); - hlist_add_head_rcu(&ifa->hash, &inet_addr_lst[hash]); hlist_add_head_rcu(&ifa->addr_lst, &net->ipv4.inet_addr_lst[hash]); } @@ -141,7 +138,6 @@ static void inet_hash_remove(struct in_ifaddr *ifa) { ASSERT_RTNL(); hlist_del_init_rcu(&ifa->addr_lst); - hlist_del_init_rcu(&ifa->hash); } /** @@ -228,7 +224,6 @@ static struct in_ifaddr *inet_alloc_ifa(struct in_device *in_dev) in_dev_hold(in_dev); ifa->ifa_dev = in_dev; - INIT_HLIST_NODE(&ifa->hash); INIT_HLIST_NODE(&ifa->addr_lst); return ifa; @@ -2804,11 +2799,6 @@ static struct rtnl_af_ops inet_af_ops __read_mostly = { void __init devinet_init(void) { - int i; - - for (i = 0; i < IN4_ADDR_HSIZE; i++) - INIT_HLIST_HEAD(&inet_addr_lst[i]); - register_pernet_subsys(&devinet_ops); register_netdevice_notifier(&ip_netdev_notifier); -- cgit v1.2.3 From 016f426a14f09faa8bdb68b063c2947edf3108a1 Mon Sep 17 00:00:00 2001 From: Cosmin Ratiu Date: Tue, 8 Oct 2024 21:32:09 +0300 Subject: net/mlx5: qos: Flesh out element_attributes in mlx5_ifc.h This is used for multiple purposes, depending on the scheduling element created. There are a few helper struct defined a long time ago, but they are not easy to find in the file and they are about to get new members. This commit cleans up this area a bit by: - moving the helper structs closer to where they are relevant. - defining a helper union to include all of them to help discoverability. - making use of it everywhere element_attributes is used. - using a consistent 'attr' name. Signed-off-by: Cosmin Ratiu Signed-off-by: Tariq Toukan Signed-off-by: Paolo Abeni --- drivers/net/ethernet/mellanox/mlx5/core/esw/qos.c | 18 +++--- include/linux/mlx5/mlx5_ifc.h | 67 +++++++++++++---------- 2 files changed, 45 insertions(+), 40 deletions(-) (limited to 'include/linux') diff --git a/drivers/net/ethernet/mellanox/mlx5/core/esw/qos.c b/drivers/net/ethernet/mellanox/mlx5/core/esw/qos.c index 02a3563f51ad..7154eeff4fd4 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/esw/qos.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/esw/qos.c @@ -339,7 +339,7 @@ static int esw_qos_vport_create_sched_element(struct mlx5_eswitch *esw, struct mlx5_esw_rate_group *group = vport->qos.group; struct mlx5_core_dev *dev = esw->dev; u32 parent_tsar_ix; - void *vport_elem; + void *attr; int err; if (!esw_qos_element_type_supported(dev, SCHEDULING_CONTEXT_ELEMENT_TYPE_VPORT)) @@ -348,8 +348,8 @@ static int esw_qos_vport_create_sched_element(struct mlx5_eswitch *esw, parent_tsar_ix = group ? group->tsar_ix : esw->qos.root_tsar_ix; MLX5_SET(scheduling_context, sched_ctx, element_type, SCHEDULING_CONTEXT_ELEMENT_TYPE_VPORT); - vport_elem = MLX5_ADDR_OF(scheduling_context, sched_ctx, element_attributes); - MLX5_SET(vport_element, vport_elem, vport_number, vport->vport); + attr = MLX5_ADDR_OF(scheduling_context, sched_ctx, element_attributes); + MLX5_SET(vport_element, attr, vport_number, vport->vport); MLX5_SET(scheduling_context, sched_ctx, parent_element_id, parent_tsar_ix); MLX5_SET(scheduling_context, sched_ctx, max_average_bw, max_rate); MLX5_SET(scheduling_context, sched_ctx, bw_share, bw_share); @@ -443,8 +443,8 @@ __esw_qos_create_rate_group(struct mlx5_eswitch *esw, struct netlink_ext_ack *ex { u32 tsar_ctx[MLX5_ST_SZ_DW(scheduling_context)] = {}; struct mlx5_esw_rate_group *group; - __be32 *attr; u32 divider; + void *attr; int err; group = kzalloc(sizeof(*group), GFP_KERNEL); @@ -453,12 +453,10 @@ __esw_qos_create_rate_group(struct mlx5_eswitch *esw, struct netlink_ext_ack *ex MLX5_SET(scheduling_context, tsar_ctx, element_type, SCHEDULING_CONTEXT_ELEMENT_TYPE_TSAR); - - attr = MLX5_ADDR_OF(scheduling_context, tsar_ctx, element_attributes); - *attr = cpu_to_be32(TSAR_ELEMENT_TSAR_TYPE_DWRR << 16); - MLX5_SET(scheduling_context, tsar_ctx, parent_element_id, esw->qos.root_tsar_ix); + attr = MLX5_ADDR_OF(scheduling_context, tsar_ctx, element_attributes); + MLX5_SET(tsar_element, attr, tsar_type, TSAR_ELEMENT_TSAR_TYPE_DWRR); err = mlx5_create_scheduling_element_cmd(esw->dev, SCHEDULING_HIERARCHY_E_SWITCH, tsar_ctx, @@ -559,7 +557,7 @@ static int esw_qos_create(struct mlx5_eswitch *esw, struct netlink_ext_ack *exta { u32 tsar_ctx[MLX5_ST_SZ_DW(scheduling_context)] = {}; struct mlx5_core_dev *dev = esw->dev; - __be32 *attr; + void *attr; int err; if (!MLX5_CAP_GEN(dev, qos) || !MLX5_CAP_QOS(dev, esw_scheduling)) @@ -573,7 +571,7 @@ static int esw_qos_create(struct mlx5_eswitch *esw, struct netlink_ext_ack *exta SCHEDULING_CONTEXT_ELEMENT_TYPE_TSAR); attr = MLX5_ADDR_OF(scheduling_context, tsar_ctx, element_attributes); - *attr = cpu_to_be32(TSAR_ELEMENT_TSAR_TYPE_DWRR << 16); + MLX5_SET(tsar_element, attr, tsar_type, TSAR_ELEMENT_TSAR_TYPE_DWRR); err = mlx5_create_scheduling_element_cmd(dev, SCHEDULING_HIERARCHY_E_SWITCH, diff --git a/include/linux/mlx5/mlx5_ifc.h b/include/linux/mlx5/mlx5_ifc.h index 96d369112bfa..c79ba6197673 100644 --- a/include/linux/mlx5/mlx5_ifc.h +++ b/include/linux/mlx5/mlx5_ifc.h @@ -4105,11 +4105,47 @@ enum { ELEMENT_TYPE_CAP_MASK_QUEUE_GROUP = 1 << 4, }; +enum { + TSAR_ELEMENT_TSAR_TYPE_DWRR = 0x0, + TSAR_ELEMENT_TSAR_TYPE_ROUND_ROBIN = 0x1, + TSAR_ELEMENT_TSAR_TYPE_ETS = 0x2, +}; + +enum { + TSAR_TYPE_CAP_MASK_DWRR = 1 << 0, + TSAR_TYPE_CAP_MASK_ROUND_ROBIN = 1 << 1, + TSAR_TYPE_CAP_MASK_ETS = 1 << 2, +}; + +struct mlx5_ifc_tsar_element_bits { + u8 reserved_at_0[0x8]; + u8 tsar_type[0x8]; + u8 reserved_at_10[0x10]; +}; + +struct mlx5_ifc_vport_element_bits { + u8 reserved_at_0[0x10]; + u8 vport_number[0x10]; +}; + +struct mlx5_ifc_vport_tc_element_bits { + u8 traffic_class[0x4]; + u8 reserved_at_4[0xc]; + u8 vport_number[0x10]; +}; + +union mlx5_ifc_element_attributes_bits { + struct mlx5_ifc_tsar_element_bits tsar; + struct mlx5_ifc_vport_element_bits vport; + struct mlx5_ifc_vport_tc_element_bits vport_tc; + u8 reserved_at_0[0x20]; +}; + struct mlx5_ifc_scheduling_context_bits { u8 element_type[0x8]; u8 reserved_at_8[0x18]; - u8 element_attributes[0x20]; + union mlx5_ifc_element_attributes_bits element_attributes; u8 parent_element_id[0x20]; @@ -4798,35 +4834,6 @@ struct mlx5_ifc_register_loopback_control_bits { u8 reserved_at_20[0x60]; }; -struct mlx5_ifc_vport_tc_element_bits { - u8 traffic_class[0x4]; - u8 reserved_at_4[0xc]; - u8 vport_number[0x10]; -}; - -struct mlx5_ifc_vport_element_bits { - u8 reserved_at_0[0x10]; - u8 vport_number[0x10]; -}; - -enum { - TSAR_ELEMENT_TSAR_TYPE_DWRR = 0x0, - TSAR_ELEMENT_TSAR_TYPE_ROUND_ROBIN = 0x1, - TSAR_ELEMENT_TSAR_TYPE_ETS = 0x2, -}; - -enum { - TSAR_TYPE_CAP_MASK_DWRR = 1 << 0, - TSAR_TYPE_CAP_MASK_ROUND_ROBIN = 1 << 1, - TSAR_TYPE_CAP_MASK_ETS = 1 << 2, -}; - -struct mlx5_ifc_tsar_element_bits { - u8 reserved_at_0[0x8]; - u8 tsar_type[0x8]; - u8 reserved_at_10[0x10]; -}; - enum { MLX5_TEARDOWN_HCA_OUT_FORCE_STATE_SUCCESS = 0x0, MLX5_TEARDOWN_HCA_OUT_FORCE_STATE_FAIL = 0x1, -- cgit v1.2.3 From 13d68a16430312fc21990f48326366eb73891202 Mon Sep 17 00:00:00 2001 From: Paolo Abeni Date: Wed, 9 Oct 2024 10:09:47 +0200 Subject: genetlink: extend info user-storage to match NL cb ctx This allows a more uniform implementation of non-dump and dump operations, and will be used later in the series to avoid some per-operation allocation. Additionally rename the NL_ASSERT_DUMP_CTX_FITS macro, to fit a more extended usage. Suggested-by: Jakub Kicinski Reviewed-by: Jakub Kicinski Reviewed-by: Jiri Pirko Signed-off-by: Paolo Abeni Link: https://patch.msgid.link/1130cc2896626b84587a2a5f96a5c6829638f4da.1728460186.git.pabeni@redhat.com Signed-off-by: Jakub Kicinski --- drivers/net/vxlan/vxlan_mdb.c | 2 +- include/linux/netlink.h | 5 +++-- include/net/genetlink.h | 8 ++++++-- net/core/netdev-genl.c | 2 +- net/core/rtnetlink.c | 2 +- net/devlink/devl_internal.h | 2 +- net/ethtool/rss.c | 2 +- net/netfilter/nf_conntrack_netlink.c | 2 +- net/netlink/genetlink.c | 4 ++-- 9 files changed, 17 insertions(+), 12 deletions(-) (limited to 'include/linux') diff --git a/drivers/net/vxlan/vxlan_mdb.c b/drivers/net/vxlan/vxlan_mdb.c index 60eb95a06d55..ebed05a2804c 100644 --- a/drivers/net/vxlan/vxlan_mdb.c +++ b/drivers/net/vxlan/vxlan_mdb.c @@ -284,7 +284,7 @@ int vxlan_mdb_dump(struct net_device *dev, struct sk_buff *skb, ASSERT_RTNL(); - NL_ASSERT_DUMP_CTX_FITS(struct vxlan_mdb_dump_ctx); + NL_ASSERT_CTX_FITS(struct vxlan_mdb_dump_ctx); nlh = nlmsg_put(skb, NETLINK_CB(cb->skb).portid, cb->nlh->nlmsg_seq, RTM_NEWMDB, sizeof(*bpm), diff --git a/include/linux/netlink.h b/include/linux/netlink.h index b332c2048c75..a3ca198a3a9e 100644 --- a/include/linux/netlink.h +++ b/include/linux/netlink.h @@ -34,6 +34,7 @@ struct netlink_skb_parms { #define NETLINK_CB(skb) (*(struct netlink_skb_parms*)&((skb)->cb)) #define NETLINK_CREDS(skb) (&NETLINK_CB((skb)).creds) +#define NETLINK_CTX_SIZE 48 void netlink_table_grab(void); @@ -293,7 +294,7 @@ struct netlink_callback { int flags; bool strict_check; union { - u8 ctx[48]; + u8 ctx[NETLINK_CTX_SIZE]; /* args is deprecated. Cast a struct over ctx instead * for proper type safety. @@ -302,7 +303,7 @@ struct netlink_callback { }; }; -#define NL_ASSERT_DUMP_CTX_FITS(type_name) \ +#define NL_ASSERT_CTX_FITS(type_name) \ BUILD_BUG_ON(sizeof(type_name) > \ sizeof_field(struct netlink_callback, ctx)) diff --git a/include/net/genetlink.h b/include/net/genetlink.h index 9ab49bfeae78..9d3726e8f90e 100644 --- a/include/net/genetlink.h +++ b/include/net/genetlink.h @@ -124,7 +124,8 @@ struct genl_family { * @genlhdr: generic netlink message header * @attrs: netlink attributes * @_net: network namespace - * @user_ptr: user pointers + * @ctx: storage space for the use by the family + * @user_ptr: user pointers (deprecated, use ctx instead) * @extack: extended ACK report struct */ struct genl_info { @@ -135,7 +136,10 @@ struct genl_info { struct genlmsghdr * genlhdr; struct nlattr ** attrs; possible_net_t _net; - void * user_ptr[2]; + union { + u8 ctx[NETLINK_CTX_SIZE]; + void * user_ptr[2]; + }; struct netlink_ext_ack *extack; }; diff --git a/net/core/netdev-genl.c b/net/core/netdev-genl.c index 1cb954f2d39e..358cba248796 100644 --- a/net/core/netdev-genl.c +++ b/net/core/netdev-genl.c @@ -24,7 +24,7 @@ struct netdev_nl_dump_ctx { static struct netdev_nl_dump_ctx *netdev_dump_ctx(struct netlink_callback *cb) { - NL_ASSERT_DUMP_CTX_FITS(struct netdev_nl_dump_ctx); + NL_ASSERT_CTX_FITS(struct netdev_nl_dump_ctx); return (struct netdev_nl_dump_ctx *)cb->ctx; } diff --git a/net/core/rtnetlink.c b/net/core/rtnetlink.c index 6d68247aea70..a9b81b7d9746 100644 --- a/net/core/rtnetlink.c +++ b/net/core/rtnetlink.c @@ -6243,7 +6243,7 @@ static int rtnl_mdb_dump(struct sk_buff *skb, struct netlink_callback *cb) int idx, s_idx; int err; - NL_ASSERT_DUMP_CTX_FITS(struct rtnl_mdb_dump_ctx); + NL_ASSERT_CTX_FITS(struct rtnl_mdb_dump_ctx); if (cb->strict_check) { err = rtnl_mdb_valid_dump_req(cb->nlh, cb->extack); diff --git a/net/devlink/devl_internal.h b/net/devlink/devl_internal.h index c7a8e13f917c..a9f064ab9ed9 100644 --- a/net/devlink/devl_internal.h +++ b/net/devlink/devl_internal.h @@ -166,7 +166,7 @@ int devlink_nl_dumpit(struct sk_buff *msg, struct netlink_callback *cb, static inline struct devlink_nl_dump_state * devlink_dump_state(struct netlink_callback *cb) { - NL_ASSERT_DUMP_CTX_FITS(struct devlink_nl_dump_state); + NL_ASSERT_CTX_FITS(struct devlink_nl_dump_state); return (struct devlink_nl_dump_state *)cb->ctx; } diff --git a/net/ethtool/rss.c b/net/ethtool/rss.c index e07386275e14..7cb106b590ab 100644 --- a/net/ethtool/rss.c +++ b/net/ethtool/rss.c @@ -224,7 +224,7 @@ struct rss_nl_dump_ctx { static struct rss_nl_dump_ctx *rss_dump_ctx(struct netlink_callback *cb) { - NL_ASSERT_DUMP_CTX_FITS(struct rss_nl_dump_ctx); + NL_ASSERT_CTX_FITS(struct rss_nl_dump_ctx); return (struct rss_nl_dump_ctx *)cb->ctx; } diff --git a/net/netfilter/nf_conntrack_netlink.c b/net/netfilter/nf_conntrack_netlink.c index 6a1239433830..36168f8b6efa 100644 --- a/net/netfilter/nf_conntrack_netlink.c +++ b/net/netfilter/nf_conntrack_netlink.c @@ -3870,7 +3870,7 @@ static int __init ctnetlink_init(void) { int ret; - NL_ASSERT_DUMP_CTX_FITS(struct ctnetlink_list_dump_ctx); + NL_ASSERT_CTX_FITS(struct ctnetlink_list_dump_ctx); ret = nfnetlink_subsys_register(&ctnl_subsys); if (ret < 0) { diff --git a/net/netlink/genetlink.c b/net/netlink/genetlink.c index feb54c63a116..29387b605f3e 100644 --- a/net/netlink/genetlink.c +++ b/net/netlink/genetlink.c @@ -997,7 +997,7 @@ static int genl_start(struct netlink_callback *cb) info->info.attrs = attrs; genl_info_net_set(&info->info, sock_net(cb->skb->sk)); info->info.extack = cb->extack; - memset(&info->info.user_ptr, 0, sizeof(info->info.user_ptr)); + memset(&info->info.ctx, 0, sizeof(info->info.ctx)); cb->data = info; if (ops->start) { @@ -1104,7 +1104,7 @@ static int genl_family_rcv_msg_doit(const struct genl_family *family, info.attrs = attrbuf; info.extack = extack; genl_info_net_set(&info, net); - memset(&info.user_ptr, 0, sizeof(info.user_ptr)); + memset(&info.ctx, 0, sizeof(info.ctx)); if (ops->pre_doit) { err = ops->pre_doit(ops, skb, &info); -- cgit v1.2.3 From 4b623f9f0f59652ea71fcb27d60b4c3b65126dbb Mon Sep 17 00:00:00 2001 From: Paolo Abeni Date: Wed, 9 Oct 2024 10:09:49 +0200 Subject: net-shapers: implement NL get operation Introduce the basic infrastructure to implement the net-shaper core functionality. Each network devices carries a net-shaper cache, the NL get() operation fetches the data from such cache. The cache is initially empty, will be fill by the set()/group() operation implemented later and is destroyed at device cleanup time. The net_shaper_fill_handle(), net_shaper_ctx_init(), and net_shaper_generic_pre() implementations handle generic index type attributes, despite the current caller always pass a constant value to avoid more noise in later patches using them with different attributes. Reviewed-by: Jakub Kicinski Reviewed-by: Jiri Pirko Signed-off-by: Paolo Abeni Link: https://patch.msgid.link/ddd10fd645a9367803ad02fca4a5664ea5ace170.1728460186.git.pabeni@redhat.com Signed-off-by: Jakub Kicinski --- Documentation/networking/kapi.rst | 3 + include/linux/netdevice.h | 21 +++ include/net/net_shaper.h | 120 ++++++++++++++ net/core/dev.c | 6 + net/core/dev.h | 6 + net/shaper/shaper.c | 335 +++++++++++++++++++++++++++++++++++++- 6 files changed, 484 insertions(+), 7 deletions(-) create mode 100644 include/net/net_shaper.h (limited to 'include/linux') diff --git a/Documentation/networking/kapi.rst b/Documentation/networking/kapi.rst index ea55f462cefa..98682b9a13ee 100644 --- a/Documentation/networking/kapi.rst +++ b/Documentation/networking/kapi.rst @@ -104,6 +104,9 @@ Driver Support .. kernel-doc:: include/linux/netdevice.h :internal: +.. kernel-doc:: include/net/net_shaper.h + :internal: + PHY Support ----------- diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index 3baf8e539b6f..e6b93d01e631 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -1603,6 +1603,14 @@ struct net_device_ops { int (*ndo_hwtstamp_set)(struct net_device *dev, struct kernel_hwtstamp_config *kernel_config, struct netlink_ext_ack *extack); + +#if IS_ENABLED(CONFIG_NET_SHAPER) + /** + * @net_shaper_ops: Device shaping offload operations + * see include/net/net_shapers.h + */ + const struct net_shaper_ops *net_shaper_ops; +#endif }; /** @@ -2406,6 +2414,19 @@ struct net_device { u64 max_pacing_offload_horizon; + /** + * @lock: protects @net_shaper_hierarchy, feel free to use for other + * netdev-scope protection. Ordering: take after rtnl_lock. + */ + struct mutex lock; + +#if IS_ENABLED(CONFIG_NET_SHAPER) + /** + * @net_shaper_hierarchy: data tracking the current shaper status + * see include/net/net_shapers.h + */ + struct net_shaper_hierarchy *net_shaper_hierarchy; +#endif u8 priv[] ____cacheline_aligned __counted_by(priv_len); } ____cacheline_aligned; diff --git a/include/net/net_shaper.h b/include/net/net_shaper.h new file mode 100644 index 000000000000..5c3f49b52fe9 --- /dev/null +++ b/include/net/net_shaper.h @@ -0,0 +1,120 @@ +/* SPDX-License-Identifier: GPL-2.0-or-later */ + +#ifndef _NET_SHAPER_H_ +#define _NET_SHAPER_H_ + +#include + +#include + +struct net_device; +struct devlink; +struct netlink_ext_ack; + +enum net_shaper_binding_type { + NET_SHAPER_BINDING_TYPE_NETDEV, + /* NET_SHAPER_BINDING_TYPE_DEVLINK_PORT */ +}; + +struct net_shaper_binding { + enum net_shaper_binding_type type; + union { + struct net_device *netdev; + struct devlink *devlink; + }; +}; + +struct net_shaper_handle { + enum net_shaper_scope scope; + u32 id; +}; + +/** + * struct net_shaper - represents a shaping node on the NIC H/W + * zeroed field are considered not set. + * @parent: Unique identifier for the shaper parent, usually implied + * @handle: Unique identifier for this shaper + * @metric: Specify if the rate limits refers to PPS or BPS + * @bw_min: Minimum guaranteed rate for this shaper + * @bw_max: Maximum peak rate allowed for this shaper + * @burst: Maximum burst for the peek rate of this shaper + * @priority: Scheduling priority for this shaper + * @weight: Scheduling weight for this shaper + */ +struct net_shaper { + struct net_shaper_handle parent; + struct net_shaper_handle handle; + enum net_shaper_metric metric; + u64 bw_min; + u64 bw_max; + u64 burst; + u32 priority; + u32 weight; + + /* private: */ + u32 leaves; /* accounted only for NODE scope */ + struct rcu_head rcu; +}; + +/** + * struct net_shaper_ops - Operations on device H/W shapers + * + * The operations applies to either net_device and devlink objects. + * The initial shaping configuration at device initialization is empty: + * does not constraint the rate in any way. + * The network core keeps track of the applied user-configuration in + * the net_device or devlink structure. + * The operations are serialized via a per device lock. + * + * Device not supporting any kind of nesting should not provide the + * group operation. + * + * Each shaper is uniquely identified within the device with a 'handle' + * comprising the shaper scope and a scope-specific id. + */ +struct net_shaper_ops { + /** + * @group: create the specified shapers scheduling group + * + * Nest the @leaves shapers identified under the * @node shaper. + * All the shapers belong to the device specified by @binding. + * The @leaves arrays size is specified by @leaves_count. + * Create either the @leaves and the @node shaper; or if they already + * exists, links them together in the desired way. + * @leaves scope must be NET_SHAPER_SCOPE_QUEUE. + */ + int (*group)(struct net_shaper_binding *binding, int leaves_count, + const struct net_shaper *leaves, + const struct net_shaper *node, + struct netlink_ext_ack *extack); + + /** + * @set: Updates the specified shaper + * + * Updates or creates the @shaper on the device specified by @binding. + */ + int (*set)(struct net_shaper_binding *binding, + const struct net_shaper *shaper, + struct netlink_ext_ack *extack); + + /** + * @delete: Removes the specified shaper + * + * Removes the shaper configuration as identified by the given @handle + * on the device specified by @binding, restoring the default behavior. + */ + int (*delete)(struct net_shaper_binding *binding, + const struct net_shaper_handle *handle, + struct netlink_ext_ack *extack); + + /** + * @capabilities: get the shaper features supported by the device + * + * Fills the bitmask @cap with the supported capabilities for the + * specified @scope and device specified by @binding. + */ + void (*capabilities)(struct net_shaper_binding *binding, + enum net_shaper_scope scope, unsigned long *cap); +}; + +#endif diff --git a/net/core/dev.c b/net/core/dev.c index ea5fbcd133ae..6e727c49a6f7 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -11147,6 +11147,8 @@ struct net_device *alloc_netdev_mqs(int sizeof_priv, const char *name, hash_init(dev->qdisc_hash); #endif + mutex_init(&dev->lock); + dev->priv_flags = IFF_XMIT_DST_RELEASE | IFF_XMIT_DST_RELEASE_PERM; setup(dev); @@ -11217,6 +11219,8 @@ void free_netdev(struct net_device *dev) return; } + mutex_destroy(&dev->lock); + kfree(dev->ethtool); netif_free_tx_queues(dev); netif_free_rx_queues(dev); @@ -11426,6 +11430,8 @@ void unregister_netdevice_many_notify(struct list_head *head, mutex_destroy(&dev->ethtool->rss_lock); + net_shaper_flush_netdev(dev); + if (skb) rtmsg_ifinfo_send(skb, dev, GFP_KERNEL, portid, nlh); diff --git a/net/core/dev.h b/net/core/dev.h index 5654325c5b71..13c558874af3 100644 --- a/net/core/dev.h +++ b/net/core/dev.h @@ -35,6 +35,12 @@ void dev_addr_flush(struct net_device *dev); int dev_addr_init(struct net_device *dev); void dev_addr_check(struct net_device *dev); +#if IS_ENABLED(CONFIG_NET_SHAPER) +void net_shaper_flush_netdev(struct net_device *dev); +#else +static inline void net_shaper_flush_netdev(struct net_device *dev) {} +#endif + /* sysctls not referred to from outside net/core/ */ extern int netdev_unregister_timeout_secs; extern int weight_p; diff --git a/net/shaper/shaper.c b/net/shaper/shaper.c index a1b20888f502..22daf7dde999 100644 --- a/net/shaper/shaper.c +++ b/net/shaper/shaper.c @@ -1,30 +1,333 @@ // SPDX-License-Identifier: GPL-2.0-or-later +#include +#include +#include #include +#include +#include #include +#include +#include +#include #include "shaper_nl_gen.h" +#include "../core/dev.h" + +#define NET_SHAPER_SCOPE_SHIFT 26 +#define NET_SHAPER_ID_MASK GENMASK(NET_SHAPER_SCOPE_SHIFT - 1, 0) +#define NET_SHAPER_SCOPE_MASK GENMASK(31, NET_SHAPER_SCOPE_SHIFT) + +#define NET_SHAPER_ID_UNSPEC NET_SHAPER_ID_MASK + +struct net_shaper_hierarchy { + struct xarray shapers; +}; + +struct net_shaper_nl_ctx { + struct net_shaper_binding binding; + netdevice_tracker dev_tracker; + unsigned long start_index; +}; + +static struct net_shaper_binding *net_shaper_binding_from_ctx(void *ctx) +{ + return &((struct net_shaper_nl_ctx *)ctx)->binding; +} + +static struct net_shaper_hierarchy * +net_shaper_hierarchy(struct net_shaper_binding *binding) +{ + /* Pairs with WRITE_ONCE() in net_shaper_hierarchy_setup. */ + if (binding->type == NET_SHAPER_BINDING_TYPE_NETDEV) + return READ_ONCE(binding->netdev->net_shaper_hierarchy); + + /* No other type supported yet. */ + return NULL; +} + +static int net_shaper_fill_binding(struct sk_buff *msg, + const struct net_shaper_binding *binding, + u32 type) +{ + /* Should never happen, as currently only NETDEV is supported. */ + if (WARN_ON_ONCE(binding->type != NET_SHAPER_BINDING_TYPE_NETDEV)) + return -EINVAL; + + if (nla_put_u32(msg, type, binding->netdev->ifindex)) + return -EMSGSIZE; + + return 0; +} + +static int net_shaper_fill_handle(struct sk_buff *msg, + const struct net_shaper_handle *handle, + u32 type) +{ + struct nlattr *handle_attr; + + if (handle->scope == NET_SHAPER_SCOPE_UNSPEC) + return 0; + + handle_attr = nla_nest_start(msg, type); + if (!handle_attr) + return -EMSGSIZE; + + if (nla_put_u32(msg, NET_SHAPER_A_HANDLE_SCOPE, handle->scope) || + (handle->scope >= NET_SHAPER_SCOPE_QUEUE && + nla_put_u32(msg, NET_SHAPER_A_HANDLE_ID, handle->id))) + goto handle_nest_cancel; + + nla_nest_end(msg, handle_attr); + return 0; + +handle_nest_cancel: + nla_nest_cancel(msg, handle_attr); + return -EMSGSIZE; +} + +static int +net_shaper_fill_one(struct sk_buff *msg, + const struct net_shaper_binding *binding, + const struct net_shaper *shaper, + const struct genl_info *info) +{ + void *hdr; + + hdr = genlmsg_iput(msg, info); + if (!hdr) + return -EMSGSIZE; + + if (net_shaper_fill_binding(msg, binding, NET_SHAPER_A_IFINDEX) || + net_shaper_fill_handle(msg, &shaper->parent, + NET_SHAPER_A_PARENT) || + net_shaper_fill_handle(msg, &shaper->handle, + NET_SHAPER_A_HANDLE) || + ((shaper->bw_min || shaper->bw_max || shaper->burst) && + nla_put_u32(msg, NET_SHAPER_A_METRIC, shaper->metric)) || + (shaper->bw_min && + nla_put_uint(msg, NET_SHAPER_A_BW_MIN, shaper->bw_min)) || + (shaper->bw_max && + nla_put_uint(msg, NET_SHAPER_A_BW_MAX, shaper->bw_max)) || + (shaper->burst && + nla_put_uint(msg, NET_SHAPER_A_BURST, shaper->burst)) || + (shaper->priority && + nla_put_u32(msg, NET_SHAPER_A_PRIORITY, shaper->priority)) || + (shaper->weight && + nla_put_u32(msg, NET_SHAPER_A_WEIGHT, shaper->weight))) + goto nla_put_failure; + + genlmsg_end(msg, hdr); + + return 0; + +nla_put_failure: + genlmsg_cancel(msg, hdr); + return -EMSGSIZE; +} + +/* Initialize the context fetching the relevant device and + * acquiring a reference to it. + */ +static int net_shaper_ctx_setup(const struct genl_info *info, int type, + struct net_shaper_nl_ctx *ctx) +{ + struct net *ns = genl_info_net(info); + struct net_device *dev; + int ifindex; + + if (GENL_REQ_ATTR_CHECK(info, type)) + return -EINVAL; + + ifindex = nla_get_u32(info->attrs[type]); + dev = netdev_get_by_index(ns, ifindex, &ctx->dev_tracker, GFP_KERNEL); + if (!dev) { + NL_SET_BAD_ATTR(info->extack, info->attrs[type]); + return -ENOENT; + } + + if (!dev->netdev_ops->net_shaper_ops) { + NL_SET_BAD_ATTR(info->extack, info->attrs[type]); + netdev_put(dev, &ctx->dev_tracker); + return -EOPNOTSUPP; + } + + ctx->binding.type = NET_SHAPER_BINDING_TYPE_NETDEV; + ctx->binding.netdev = dev; + return 0; +} + +static void net_shaper_ctx_cleanup(struct net_shaper_nl_ctx *ctx) +{ + if (ctx->binding.type == NET_SHAPER_BINDING_TYPE_NETDEV) + netdev_put(ctx->binding.netdev, &ctx->dev_tracker); +} + +static u32 net_shaper_handle_to_index(const struct net_shaper_handle *handle) +{ + return FIELD_PREP(NET_SHAPER_SCOPE_MASK, handle->scope) | + FIELD_PREP(NET_SHAPER_ID_MASK, handle->id); +} + +static struct net_shaper * +net_shaper_lookup(struct net_shaper_binding *binding, + const struct net_shaper_handle *handle) +{ + struct net_shaper_hierarchy *hierarchy = net_shaper_hierarchy(binding); + u32 index = net_shaper_handle_to_index(handle); + + return hierarchy ? xa_load(&hierarchy->shapers, index) : NULL; +} + +static int net_shaper_parse_handle(const struct nlattr *attr, + const struct genl_info *info, + struct net_shaper_handle *handle) +{ + struct nlattr *tb[NET_SHAPER_A_HANDLE_MAX + 1]; + struct nlattr *id_attr; + u32 id = 0; + int ret; + + ret = nla_parse_nested(tb, NET_SHAPER_A_HANDLE_MAX, attr, + net_shaper_handle_nl_policy, info->extack); + if (ret < 0) + return ret; + + if (NL_REQ_ATTR_CHECK(info->extack, attr, tb, + NET_SHAPER_A_HANDLE_SCOPE)) + return -EINVAL; + + handle->scope = nla_get_u32(tb[NET_SHAPER_A_HANDLE_SCOPE]); + + /* The default id for NODE scope shapers is an invalid one + * to help the 'group' operation discriminate between new + * NODE shaper creation (ID_UNSPEC) and reuse of existing + * shaper (any other value). + */ + id_attr = tb[NET_SHAPER_A_HANDLE_ID]; + if (id_attr) + id = nla_get_u32(id_attr); + else if (handle->scope == NET_SHAPER_SCOPE_NODE) + id = NET_SHAPER_ID_UNSPEC; + + handle->id = id; + return 0; +} + +static int net_shaper_generic_pre(struct genl_info *info, int type) +{ + struct net_shaper_nl_ctx *ctx = (struct net_shaper_nl_ctx *)info->ctx; + + BUILD_BUG_ON(sizeof(*ctx) > sizeof(info->ctx)); + + return net_shaper_ctx_setup(info, type, ctx); +} + int net_shaper_nl_pre_doit(const struct genl_split_ops *ops, struct sk_buff *skb, struct genl_info *info) { - return -EOPNOTSUPP; + return net_shaper_generic_pre(info, NET_SHAPER_A_IFINDEX); +} + +static void net_shaper_generic_post(struct genl_info *info) +{ + net_shaper_ctx_cleanup((struct net_shaper_nl_ctx *)info->ctx); } void net_shaper_nl_post_doit(const struct genl_split_ops *ops, struct sk_buff *skb, struct genl_info *info) { + net_shaper_generic_post(info); +} + +int net_shaper_nl_pre_dumpit(struct netlink_callback *cb) +{ + struct net_shaper_nl_ctx *ctx = (struct net_shaper_nl_ctx *)cb->ctx; + const struct genl_info *info = genl_info_dump(cb); + + return net_shaper_ctx_setup(info, NET_SHAPER_A_IFINDEX, ctx); +} + +int net_shaper_nl_post_dumpit(struct netlink_callback *cb) +{ + net_shaper_ctx_cleanup((struct net_shaper_nl_ctx *)cb->ctx); + return 0; } int net_shaper_nl_get_doit(struct sk_buff *skb, struct genl_info *info) { - return -EOPNOTSUPP; + struct net_shaper_binding *binding; + struct net_shaper_handle handle; + struct net_shaper *shaper; + struct sk_buff *msg; + int ret; + + if (GENL_REQ_ATTR_CHECK(info, NET_SHAPER_A_HANDLE)) + return -EINVAL; + + binding = net_shaper_binding_from_ctx(info->ctx); + ret = net_shaper_parse_handle(info->attrs[NET_SHAPER_A_HANDLE], info, + &handle); + if (ret < 0) + return ret; + + msg = nlmsg_new(NLMSG_DEFAULT_SIZE, GFP_KERNEL); + if (!msg) + return -ENOMEM; + + rcu_read_lock(); + shaper = net_shaper_lookup(binding, &handle); + if (!shaper) { + NL_SET_BAD_ATTR(info->extack, + info->attrs[NET_SHAPER_A_HANDLE]); + rcu_read_unlock(); + ret = -ENOENT; + goto free_msg; + } + + ret = net_shaper_fill_one(msg, binding, shaper, info); + rcu_read_unlock(); + if (ret) + goto free_msg; + + ret = genlmsg_reply(msg, info); + if (ret) + goto free_msg; + + return 0; + +free_msg: + nlmsg_free(msg); + return ret; } int net_shaper_nl_get_dumpit(struct sk_buff *skb, struct netlink_callback *cb) { - return -EOPNOTSUPP; + struct net_shaper_nl_ctx *ctx = (struct net_shaper_nl_ctx *)cb->ctx; + const struct genl_info *info = genl_info_dump(cb); + struct net_shaper_hierarchy *hierarchy; + struct net_shaper_binding *binding; + struct net_shaper *shaper; + int ret = 0; + + /* Don't error out dumps performed before any set operation. */ + binding = net_shaper_binding_from_ctx(ctx); + hierarchy = net_shaper_hierarchy(binding); + if (!hierarchy) + return 0; + + rcu_read_lock(); + for (; (shaper = xa_find(&hierarchy->shapers, &ctx->start_index, + U32_MAX, XA_PRESENT)); ctx->start_index++) { + ret = net_shaper_fill_one(skb, binding, shaper, info); + if (ret) + break; + } + rcu_read_unlock(); + + return ret; } int net_shaper_nl_set_doit(struct sk_buff *skb, struct genl_info *info) @@ -37,14 +340,32 @@ int net_shaper_nl_delete_doit(struct sk_buff *skb, struct genl_info *info) return -EOPNOTSUPP; } -int net_shaper_nl_pre_dumpit(struct netlink_callback *cb) +static void net_shaper_flush(struct net_shaper_binding *binding) { - return -EOPNOTSUPP; + struct net_shaper_hierarchy *hierarchy = net_shaper_hierarchy(binding); + struct net_shaper *cur; + unsigned long index; + + if (!hierarchy) + return; + + xa_lock(&hierarchy->shapers); + xa_for_each(&hierarchy->shapers, index, cur) { + __xa_erase(&hierarchy->shapers, index); + kfree(cur); + } + xa_unlock(&hierarchy->shapers); + kfree(hierarchy); } -int net_shaper_nl_post_dumpit(struct netlink_callback *cb) +void net_shaper_flush_netdev(struct net_device *dev) { - return -EOPNOTSUPP; + struct net_shaper_binding binding = { + .type = NET_SHAPER_BINDING_TYPE_NETDEV, + .netdev = dev, + }; + + net_shaper_flush(&binding); } static int __init shaper_init(void) -- cgit v1.2.3 From 608a5c05c39b75fa2539ce9e521d289c5a5326f7 Mon Sep 17 00:00:00 2001 From: Wenjun Wu Date: Wed, 9 Oct 2024 10:09:58 +0200 Subject: virtchnl: support queue rate limit and quanta size configuration This patch adds new virtchnl opcodes and structures for rate limit and quanta size configuration, which include: 1. VIRTCHNL_OP_CONFIG_QUEUE_BW, to configure max bandwidth for each VF per queue. 2. VIRTCHNL_OP_CONFIG_QUANTA, to configure quanta size per queue. 3. VIRTCHNL_OP_GET_QOS_CAPS, VF queries current QoS configuration, such as enabled TCs, arbiter type, up2tc and bandwidth of VSI node. The configuration is previously set by DCB and PF, and now is the potential QoS capability of VF. VF can take it as reference to configure queue TC mapping. Reviewed-by: Jiri Pirko Signed-off-by: Wenjun Wu Signed-off-by: Paolo Abeni Link: https://patch.msgid.link/839002f7bd6f63b985a060a51b079f6e6dbbe237.1728460186.git.pabeni@redhat.com Signed-off-by: Jakub Kicinski --- include/linux/avf/virtchnl.h | 119 +++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 119 insertions(+) (limited to 'include/linux') diff --git a/include/linux/avf/virtchnl.h b/include/linux/avf/virtchnl.h index f41395264dca..223e433c39fe 100644 --- a/include/linux/avf/virtchnl.h +++ b/include/linux/avf/virtchnl.h @@ -89,6 +89,9 @@ enum virtchnl_rx_hsplit { VIRTCHNL_RX_HSPLIT_SPLIT_SCTP = 8, }; +enum virtchnl_bw_limit_type { + VIRTCHNL_BW_SHAPER = 0, +}; /* END GENERIC DEFINES */ /* Opcodes for VF-PF communication. These are placed in the v_opcode field @@ -151,6 +154,11 @@ enum virtchnl_ops { VIRTCHNL_OP_DISABLE_VLAN_STRIPPING_V2 = 55, VIRTCHNL_OP_ENABLE_VLAN_INSERTION_V2 = 56, VIRTCHNL_OP_DISABLE_VLAN_INSERTION_V2 = 57, + /* opcode 57 - 65 are reserved */ + VIRTCHNL_OP_GET_QOS_CAPS = 66, + /* opcode 68 through 111 are reserved */ + VIRTCHNL_OP_CONFIG_QUEUE_BW = 112, + VIRTCHNL_OP_CONFIG_QUANTA = 113, VIRTCHNL_OP_MAX, }; @@ -261,6 +269,7 @@ VIRTCHNL_CHECK_STRUCT_LEN(16, virtchnl_vsi_resource); #define VIRTCHNL_VF_OFFLOAD_RX_FLEX_DESC BIT(26) #define VIRTCHNL_VF_OFFLOAD_ADV_RSS_PF BIT(27) #define VIRTCHNL_VF_OFFLOAD_FDIR_PF BIT(28) +#define VIRTCHNL_VF_OFFLOAD_QOS BIT(29) #define VF_BASE_MODE_OFFLOADS (VIRTCHNL_VF_OFFLOAD_L2 | \ VIRTCHNL_VF_OFFLOAD_VLAN | \ @@ -1416,6 +1425,85 @@ struct virtchnl_fdir_del { VIRTCHNL_CHECK_STRUCT_LEN(12, virtchnl_fdir_del); +struct virtchnl_shaper_bw { + /* Unit is Kbps */ + u32 committed; + u32 peak; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(8, virtchnl_shaper_bw); + +/* VIRTCHNL_OP_GET_QOS_CAPS + * VF sends this message to get its QoS Caps, such as + * TC number, Arbiter and Bandwidth. + */ +struct virtchnl_qos_cap_elem { + u8 tc_num; + u8 tc_prio; +#define VIRTCHNL_ABITER_STRICT 0 +#define VIRTCHNL_ABITER_ETS 2 + u8 arbiter; +#define VIRTCHNL_STRICT_WEIGHT 1 + u8 weight; + enum virtchnl_bw_limit_type type; + union { + struct virtchnl_shaper_bw shaper; + u8 pad2[32]; + }; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(40, virtchnl_qos_cap_elem); + +struct virtchnl_qos_cap_list { + u16 vsi_id; + u16 num_elem; + struct virtchnl_qos_cap_elem cap[]; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(4, virtchnl_qos_cap_list); +#define virtchnl_qos_cap_list_LEGACY_SIZEOF 44 + +/* VIRTCHNL_OP_CONFIG_QUEUE_BW */ +struct virtchnl_queue_bw { + u16 queue_id; + u8 tc; + u8 pad; + struct virtchnl_shaper_bw shaper; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(12, virtchnl_queue_bw); + +struct virtchnl_queues_bw_cfg { + u16 vsi_id; + u16 num_queues; + struct virtchnl_queue_bw cfg[]; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(4, virtchnl_queues_bw_cfg); +#define virtchnl_queues_bw_cfg_LEGACY_SIZEOF 16 + +enum virtchnl_queue_type { + VIRTCHNL_QUEUE_TYPE_TX = 0, + VIRTCHNL_QUEUE_TYPE_RX = 1, +}; + +/* structure to specify a chunk of contiguous queues */ +struct virtchnl_queue_chunk { + /* see enum virtchnl_queue_type */ + s32 type; + u16 start_queue_id; + u16 num_queues; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(8, virtchnl_queue_chunk); + +struct virtchnl_quanta_cfg { + u16 quanta_size; + struct virtchnl_queue_chunk queue_select; +}; + +VIRTCHNL_CHECK_STRUCT_LEN(12, virtchnl_quanta_cfg); + #define __vss_byone(p, member, count, old) \ (struct_size(p, member, count) + (old - 1 - struct_size(p, member, 0))) @@ -1438,6 +1526,8 @@ VIRTCHNL_CHECK_STRUCT_LEN(12, virtchnl_fdir_del); __vss(virtchnl_vlan_filter_list_v2, __vss_byelem, p, m, c), \ __vss(virtchnl_tc_info, __vss_byelem, p, m, c), \ __vss(virtchnl_rdma_qvlist_info, __vss_byelem, p, m, c), \ + __vss(virtchnl_qos_cap_list, __vss_byelem, p, m, c), \ + __vss(virtchnl_queues_bw_cfg, __vss_byelem, p, m, c), \ __vss(virtchnl_rss_key, __vss_byone, p, m, c), \ __vss(virtchnl_rss_lut, __vss_byone, p, m, c)) @@ -1637,6 +1727,35 @@ virtchnl_vc_validate_vf_msg(struct virtchnl_version_info *ver, u32 v_opcode, case VIRTCHNL_OP_DISABLE_VLAN_INSERTION_V2: valid_len = sizeof(struct virtchnl_vlan_setting); break; + case VIRTCHNL_OP_GET_QOS_CAPS: + break; + case VIRTCHNL_OP_CONFIG_QUEUE_BW: + valid_len = virtchnl_queues_bw_cfg_LEGACY_SIZEOF; + if (msglen >= valid_len) { + struct virtchnl_queues_bw_cfg *q_bw = + (struct virtchnl_queues_bw_cfg *)msg; + + valid_len = virtchnl_struct_size(q_bw, cfg, + q_bw->num_queues); + if (q_bw->num_queues == 0) { + err_msg_format = true; + break; + } + } + break; + case VIRTCHNL_OP_CONFIG_QUANTA: + valid_len = sizeof(struct virtchnl_quanta_cfg); + if (msglen >= valid_len) { + struct virtchnl_quanta_cfg *q_quanta = + (struct virtchnl_quanta_cfg *)msg; + + if (q_quanta->quanta_size == 0 || + q_quanta->queue_select.num_queues == 0) { + err_msg_format = true; + break; + } + } + break; /* These are always errors coming from the VF. */ case VIRTCHNL_OP_EVENT: case VIRTCHNL_OP_UNKNOWN: -- cgit v1.2.3 From 454bbde8f0d465e93e5a3a4003ac6c7e62fa4473 Mon Sep 17 00:00:00 2001 From: Menglong Dong Date: Wed, 9 Oct 2024 10:28:19 +0800 Subject: net: skb: add pskb_network_may_pull_reason() helper Introduce the function pskb_network_may_pull_reason() and make pskb_network_may_pull() a simple inline call to it. The drop reasons of it just come from pskb_may_pull_reason. Signed-off-by: Menglong Dong Reviewed-by: Simon Horman Signed-off-by: David S. Miller --- include/linux/skbuff.h | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) (limited to 'include/linux') diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h index 39f1d16f3628..48f1e0fa2a13 100644 --- a/include/linux/skbuff.h +++ b/include/linux/skbuff.h @@ -3130,9 +3130,15 @@ static inline int skb_inner_network_offset(const struct sk_buff *skb) return skb_inner_network_header(skb) - skb->data; } +static inline enum skb_drop_reason +pskb_network_may_pull_reason(struct sk_buff *skb, unsigned int len) +{ + return pskb_may_pull_reason(skb, skb_network_offset(skb) + len); +} + static inline int pskb_network_may_pull(struct sk_buff *skb, unsigned int len) { - return pskb_may_pull(skb, skb_network_offset(skb) + len); + return pskb_network_may_pull_reason(skb, len) == SKB_NOT_DROPPED_YET; } /* -- cgit v1.2.3 From 78e2baf3d96edd21c6f26d8afc0e68d02ec2c51c Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Thu, 10 Oct 2024 17:48:13 +0000 Subject: net: add TIME_WAIT logic to sk_to_full_sk() TCP will soon attach TIME_WAIT sockets to some ACK and RST. Make sure sk_to_full_sk() detects this and does not return a non full socket. v3: also changed sk_const_to_full_sk() Signed-off-by: Eric Dumazet Reviewed-by: Kuniyuki Iwashima Reviewed-by: Martin KaFai Lau Reviewed-by: Brian Vazquez Link: https://patch.msgid.link/20241010174817.1543642-2-edumazet@google.com Signed-off-by: Jakub Kicinski --- include/linux/bpf-cgroup.h | 2 +- include/net/inet_sock.h | 8 ++++++-- net/core/filter.c | 6 +----- 3 files changed, 8 insertions(+), 8 deletions(-) (limited to 'include/linux') diff --git a/include/linux/bpf-cgroup.h b/include/linux/bpf-cgroup.h index ce91d9b2acb9..f0f219271daf 100644 --- a/include/linux/bpf-cgroup.h +++ b/include/linux/bpf-cgroup.h @@ -209,7 +209,7 @@ static inline bool cgroup_bpf_sock_enabled(struct sock *sk, int __ret = 0; \ if (cgroup_bpf_enabled(CGROUP_INET_EGRESS) && sk) { \ typeof(sk) __sk = sk_to_full_sk(sk); \ - if (sk_fullsock(__sk) && __sk == skb_to_full_sk(skb) && \ + if (__sk && __sk == skb_to_full_sk(skb) && \ cgroup_bpf_sock_enabled(__sk, CGROUP_INET_EGRESS)) \ __ret = __cgroup_bpf_run_filter_skb(__sk, skb, \ CGROUP_INET_EGRESS); \ diff --git a/include/net/inet_sock.h b/include/net/inet_sock.h index f01dd273bea6..56d8bc5593d3 100644 --- a/include/net/inet_sock.h +++ b/include/net/inet_sock.h @@ -321,8 +321,10 @@ static inline unsigned long inet_cmsg_flags(const struct inet_sock *inet) static inline struct sock *sk_to_full_sk(struct sock *sk) { #ifdef CONFIG_INET - if (sk && sk->sk_state == TCP_NEW_SYN_RECV) + if (sk && READ_ONCE(sk->sk_state) == TCP_NEW_SYN_RECV) sk = inet_reqsk(sk)->rsk_listener; + if (sk && READ_ONCE(sk->sk_state) == TCP_TIME_WAIT) + sk = NULL; #endif return sk; } @@ -331,8 +333,10 @@ static inline struct sock *sk_to_full_sk(struct sock *sk) static inline const struct sock *sk_const_to_full_sk(const struct sock *sk) { #ifdef CONFIG_INET - if (sk && sk->sk_state == TCP_NEW_SYN_RECV) + if (sk && READ_ONCE(sk->sk_state) == TCP_NEW_SYN_RECV) sk = ((const struct request_sock *)sk)->rsk_listener; + if (sk && READ_ONCE(sk->sk_state) == TCP_TIME_WAIT) + sk = NULL; #endif return sk; } diff --git a/net/core/filter.c b/net/core/filter.c index bd0d08bf76bb..202c1d386e19 100644 --- a/net/core/filter.c +++ b/net/core/filter.c @@ -6778,8 +6778,6 @@ __bpf_sk_lookup(struct sk_buff *skb, struct bpf_sock_tuple *tuple, u32 len, /* sk_to_full_sk() may return (sk)->rsk_listener, so make sure the original sk * sock refcnt is decremented to prevent a request_sock leak. */ - if (!sk_fullsock(sk2)) - sk2 = NULL; if (sk2 != sk) { sock_gen_put(sk); /* Ensure there is no need to bump sk2 refcnt */ @@ -6826,8 +6824,6 @@ bpf_sk_lookup(struct sk_buff *skb, struct bpf_sock_tuple *tuple, u32 len, /* sk_to_full_sk() may return (sk)->rsk_listener, so make sure the original sk * sock refcnt is decremented to prevent a request_sock leak. */ - if (!sk_fullsock(sk2)) - sk2 = NULL; if (sk2 != sk) { sock_gen_put(sk); /* Ensure there is no need to bump sk2 refcnt */ @@ -7276,7 +7272,7 @@ BPF_CALL_1(bpf_get_listener_sock, struct sock *, sk) { sk = sk_to_full_sk(sk); - if (sk->sk_state == TCP_LISTEN && sock_flag(sk, SOCK_RCU_FREE)) + if (sk && sk->sk_state == TCP_LISTEN && sock_flag(sk, SOCK_RCU_FREE)) return (unsigned long)sk; return (unsigned long)NULL; -- cgit v1.2.3 From f15e3b3ddb9fab1c1731b6154e2cd6573fb54c4d Mon Sep 17 00:00:00 2001 From: Joe Damato Date: Fri, 11 Oct 2024 18:44:56 +0000 Subject: net: napi: Make napi_defer_hard_irqs per-NAPI Add defer_hard_irqs to napi_struct in preparation for per-NAPI settings. The existing sysfs parameter is respected; writes to sysfs will write to all NAPI structs for the device and the net_device defer_hard_irq field. Reads from sysfs show the net_device field. The ability to set defer_hard_irqs on specific NAPI instances will be added in a later commit, via netdev-genl. Signed-off-by: Joe Damato Reviewed-by: Eric Dumazet Reviewed-by: Jakub Kicinski Link: https://patch.msgid.link/20241011184527.16393-2-jdamato@fastly.com Signed-off-by: Jakub Kicinski --- .../networking/net_cachelines/net_device.rst | 1 + include/linux/netdevice.h | 3 +- net/core/dev.c | 10 +++--- net/core/dev.h | 36 ++++++++++++++++++++++ net/core/net-sysfs.c | 2 +- 5 files changed, 45 insertions(+), 7 deletions(-) (limited to 'include/linux') diff --git a/Documentation/networking/net_cachelines/net_device.rst b/Documentation/networking/net_cachelines/net_device.rst index 1b018ac35e9a..5a7388b2ab6f 100644 --- a/Documentation/networking/net_cachelines/net_device.rst +++ b/Documentation/networking/net_cachelines/net_device.rst @@ -186,4 +186,5 @@ struct dpll_pin* dpll_pin struct hlist_head page_pools struct dim_irq_moder* irq_moder u64 max_pacing_offload_horizon +u32 napi_defer_hard_irqs =================================== =========================== =================== =================== =================================================================================== diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index e6b93d01e631..2e7bc23660ec 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -373,6 +373,7 @@ struct napi_struct { unsigned int napi_id; struct hrtimer timer; struct task_struct *thread; + u32 defer_hard_irqs; /* control-path-only fields follow */ struct list_head dev_list; struct hlist_node napi_hash_node; @@ -2085,7 +2086,6 @@ struct net_device { unsigned int real_num_rx_queues; struct netdev_rx_queue *_rx; unsigned long gro_flush_timeout; - u32 napi_defer_hard_irqs; unsigned int gro_max_size; unsigned int gro_ipv4_max_size; rx_handler_func_t __rcu *rx_handler; @@ -2413,6 +2413,7 @@ struct net_device { struct dim_irq_moder *irq_moder; u64 max_pacing_offload_horizon; + u32 napi_defer_hard_irqs; /** * @lock: protects @net_shaper_hierarchy, feel free to use for other diff --git a/net/core/dev.c b/net/core/dev.c index b590eefce3b4..fbaa9eabf77f 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -6233,7 +6233,7 @@ bool napi_complete_done(struct napi_struct *n, int work_done) if (work_done) { if (n->gro_bitmask) timeout = READ_ONCE(n->dev->gro_flush_timeout); - n->defer_hard_irqs_count = READ_ONCE(n->dev->napi_defer_hard_irqs); + n->defer_hard_irqs_count = napi_get_defer_hard_irqs(n); } if (n->defer_hard_irqs_count > 0) { n->defer_hard_irqs_count--; @@ -6371,7 +6371,7 @@ static void busy_poll_stop(struct napi_struct *napi, void *have_poll_lock, bpf_net_ctx = bpf_net_ctx_set(&__bpf_net_ctx); if (flags & NAPI_F_PREFER_BUSY_POLL) { - napi->defer_hard_irqs_count = READ_ONCE(napi->dev->napi_defer_hard_irqs); + napi->defer_hard_irqs_count = napi_get_defer_hard_irqs(napi); timeout = READ_ONCE(napi->dev->gro_flush_timeout); if (napi->defer_hard_irqs_count && timeout) { hrtimer_start(&napi->timer, ns_to_ktime(timeout), HRTIMER_MODE_REL_PINNED); @@ -6653,6 +6653,7 @@ void netif_napi_add_weight(struct net_device *dev, struct napi_struct *napi, INIT_HLIST_NODE(&napi->napi_hash_node); hrtimer_init(&napi->timer, CLOCK_MONOTONIC, HRTIMER_MODE_REL_PINNED); napi->timer.function = napi_watchdog; + napi_set_defer_hard_irqs(napi, READ_ONCE(dev->napi_defer_hard_irqs)); init_gro_hash(napi); napi->skb = NULL; INIT_LIST_HEAD(&napi->rx_list); @@ -11059,7 +11060,7 @@ void netdev_sw_irq_coalesce_default_on(struct net_device *dev) if (!IS_ENABLED(CONFIG_PREEMPT_RT)) { dev->gro_flush_timeout = 20000; - dev->napi_defer_hard_irqs = 1; + netdev_set_defer_hard_irqs(dev, 1); } } EXPORT_SYMBOL_GPL(netdev_sw_irq_coalesce_default_on); @@ -12003,7 +12004,6 @@ static void __init net_dev_struct_check(void) CACHELINE_ASSERT_GROUP_MEMBER(struct net_device, net_device_read_rx, real_num_rx_queues); CACHELINE_ASSERT_GROUP_MEMBER(struct net_device, net_device_read_rx, _rx); CACHELINE_ASSERT_GROUP_MEMBER(struct net_device, net_device_read_rx, gro_flush_timeout); - CACHELINE_ASSERT_GROUP_MEMBER(struct net_device, net_device_read_rx, napi_defer_hard_irqs); CACHELINE_ASSERT_GROUP_MEMBER(struct net_device, net_device_read_rx, gro_max_size); CACHELINE_ASSERT_GROUP_MEMBER(struct net_device, net_device_read_rx, gro_ipv4_max_size); CACHELINE_ASSERT_GROUP_MEMBER(struct net_device, net_device_read_rx, rx_handler); @@ -12015,7 +12015,7 @@ static void __init net_dev_struct_check(void) #ifdef CONFIG_NET_XGRESS CACHELINE_ASSERT_GROUP_MEMBER(struct net_device, net_device_read_rx, tcx_ingress); #endif - CACHELINE_ASSERT_GROUP_SIZE(struct net_device, net_device_read_rx, 104); + CACHELINE_ASSERT_GROUP_SIZE(struct net_device, net_device_read_rx, 100); } /* diff --git a/net/core/dev.h b/net/core/dev.h index d3ea92949ff3..0716b1048261 100644 --- a/net/core/dev.h +++ b/net/core/dev.h @@ -148,6 +148,42 @@ static inline void netif_set_gro_ipv4_max_size(struct net_device *dev, WRITE_ONCE(dev->gro_ipv4_max_size, size); } +/** + * napi_get_defer_hard_irqs - get the NAPI's defer_hard_irqs + * @n: napi struct to get the defer_hard_irqs field from + * + * Return: the per-NAPI value of the defar_hard_irqs field. + */ +static inline u32 napi_get_defer_hard_irqs(const struct napi_struct *n) +{ + return READ_ONCE(n->defer_hard_irqs); +} + +/** + * napi_set_defer_hard_irqs - set the defer_hard_irqs for a napi + * @n: napi_struct to set the defer_hard_irqs field + * @defer: the value the field should be set to + */ +static inline void napi_set_defer_hard_irqs(struct napi_struct *n, u32 defer) +{ + WRITE_ONCE(n->defer_hard_irqs, defer); +} + +/** + * netdev_set_defer_hard_irqs - set defer_hard_irqs for all NAPIs of a netdev + * @netdev: the net_device for which all NAPIs will have defer_hard_irqs set + * @defer: the defer_hard_irqs value to set + */ +static inline void netdev_set_defer_hard_irqs(struct net_device *netdev, + u32 defer) +{ + struct napi_struct *napi; + + WRITE_ONCE(netdev->napi_defer_hard_irqs, defer); + list_for_each_entry(napi, &netdev->napi_list, dev_list) + napi_set_defer_hard_irqs(napi, defer); +} + int rps_cpumask_housekeeping(struct cpumask *mask); #if defined(CONFIG_DEBUG_NET) && defined(CONFIG_BPF_SYSCALL) diff --git a/net/core/net-sysfs.c b/net/core/net-sysfs.c index 05cf5347f25e..25125f356a15 100644 --- a/net/core/net-sysfs.c +++ b/net/core/net-sysfs.c @@ -429,7 +429,7 @@ static int change_napi_defer_hard_irqs(struct net_device *dev, unsigned long val if (val > S32_MAX) return -ERANGE; - WRITE_ONCE(dev->napi_defer_hard_irqs, val); + netdev_set_defer_hard_irqs(dev, (u32)val); return 0; } -- cgit v1.2.3 From acb8d4ed5661d05f794ef2ce34fd11e699e9ca32 Mon Sep 17 00:00:00 2001 From: Joe Damato Date: Fri, 11 Oct 2024 18:44:58 +0000 Subject: net: napi: Make gro_flush_timeout per-NAPI Allow per-NAPI gro_flush_timeout setting. The existing sysfs parameter is respected; writes to sysfs will write to all NAPI structs for the device and the net_device gro_flush_timeout field. Reads from sysfs will read from the net_device field. The ability to set gro_flush_timeout on specific NAPI instances will be added in a later commit, via netdev-genl. Signed-off-by: Joe Damato Reviewed-by: Eric Dumazet Reviewed-by: Jakub Kicinski Link: https://patch.msgid.link/20241011184527.16393-4-jdamato@fastly.com Signed-off-by: Jakub Kicinski --- .../networking/net_cachelines/net_device.rst | 1 + include/linux/netdevice.h | 3 +- net/core/dev.c | 12 +++---- net/core/dev.h | 40 ++++++++++++++++++++++ net/core/net-sysfs.c | 2 +- 5 files changed, 50 insertions(+), 8 deletions(-) (limited to 'include/linux') diff --git a/Documentation/networking/net_cachelines/net_device.rst b/Documentation/networking/net_cachelines/net_device.rst index 5a7388b2ab6f..67910ea49160 100644 --- a/Documentation/networking/net_cachelines/net_device.rst +++ b/Documentation/networking/net_cachelines/net_device.rst @@ -186,5 +186,6 @@ struct dpll_pin* dpll_pin struct hlist_head page_pools struct dim_irq_moder* irq_moder u64 max_pacing_offload_horizon +unsigned_long gro_flush_timeout u32 napi_defer_hard_irqs =================================== =========================== =================== =================== =================================================================================== diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index 2e7bc23660ec..93241d4de437 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -373,6 +373,7 @@ struct napi_struct { unsigned int napi_id; struct hrtimer timer; struct task_struct *thread; + unsigned long gro_flush_timeout; u32 defer_hard_irqs; /* control-path-only fields follow */ struct list_head dev_list; @@ -2085,7 +2086,6 @@ struct net_device { int ifindex; unsigned int real_num_rx_queues; struct netdev_rx_queue *_rx; - unsigned long gro_flush_timeout; unsigned int gro_max_size; unsigned int gro_ipv4_max_size; rx_handler_func_t __rcu *rx_handler; @@ -2413,6 +2413,7 @@ struct net_device { struct dim_irq_moder *irq_moder; u64 max_pacing_offload_horizon; + unsigned long gro_flush_timeout; u32 napi_defer_hard_irqs; /** diff --git a/net/core/dev.c b/net/core/dev.c index fbaa9eabf77f..e21ace3551d5 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -6232,12 +6232,12 @@ bool napi_complete_done(struct napi_struct *n, int work_done) if (work_done) { if (n->gro_bitmask) - timeout = READ_ONCE(n->dev->gro_flush_timeout); + timeout = napi_get_gro_flush_timeout(n); n->defer_hard_irqs_count = napi_get_defer_hard_irqs(n); } if (n->defer_hard_irqs_count > 0) { n->defer_hard_irqs_count--; - timeout = READ_ONCE(n->dev->gro_flush_timeout); + timeout = napi_get_gro_flush_timeout(n); if (timeout) ret = false; } @@ -6372,7 +6372,7 @@ static void busy_poll_stop(struct napi_struct *napi, void *have_poll_lock, if (flags & NAPI_F_PREFER_BUSY_POLL) { napi->defer_hard_irqs_count = napi_get_defer_hard_irqs(napi); - timeout = READ_ONCE(napi->dev->gro_flush_timeout); + timeout = napi_get_gro_flush_timeout(napi); if (napi->defer_hard_irqs_count && timeout) { hrtimer_start(&napi->timer, ns_to_ktime(timeout), HRTIMER_MODE_REL_PINNED); skip_schedule = true; @@ -6654,6 +6654,7 @@ void netif_napi_add_weight(struct net_device *dev, struct napi_struct *napi, hrtimer_init(&napi->timer, CLOCK_MONOTONIC, HRTIMER_MODE_REL_PINNED); napi->timer.function = napi_watchdog; napi_set_defer_hard_irqs(napi, READ_ONCE(dev->napi_defer_hard_irqs)); + napi_set_gro_flush_timeout(napi, READ_ONCE(dev->gro_flush_timeout)); init_gro_hash(napi); napi->skb = NULL; INIT_LIST_HEAD(&napi->rx_list); @@ -11059,7 +11060,7 @@ void netdev_sw_irq_coalesce_default_on(struct net_device *dev) WARN_ON(dev->reg_state == NETREG_REGISTERED); if (!IS_ENABLED(CONFIG_PREEMPT_RT)) { - dev->gro_flush_timeout = 20000; + netdev_set_gro_flush_timeout(dev, 20000); netdev_set_defer_hard_irqs(dev, 1); } } @@ -12003,7 +12004,6 @@ static void __init net_dev_struct_check(void) CACHELINE_ASSERT_GROUP_MEMBER(struct net_device, net_device_read_rx, ifindex); CACHELINE_ASSERT_GROUP_MEMBER(struct net_device, net_device_read_rx, real_num_rx_queues); CACHELINE_ASSERT_GROUP_MEMBER(struct net_device, net_device_read_rx, _rx); - CACHELINE_ASSERT_GROUP_MEMBER(struct net_device, net_device_read_rx, gro_flush_timeout); CACHELINE_ASSERT_GROUP_MEMBER(struct net_device, net_device_read_rx, gro_max_size); CACHELINE_ASSERT_GROUP_MEMBER(struct net_device, net_device_read_rx, gro_ipv4_max_size); CACHELINE_ASSERT_GROUP_MEMBER(struct net_device, net_device_read_rx, rx_handler); @@ -12015,7 +12015,7 @@ static void __init net_dev_struct_check(void) #ifdef CONFIG_NET_XGRESS CACHELINE_ASSERT_GROUP_MEMBER(struct net_device, net_device_read_rx, tcx_ingress); #endif - CACHELINE_ASSERT_GROUP_SIZE(struct net_device, net_device_read_rx, 100); + CACHELINE_ASSERT_GROUP_SIZE(struct net_device, net_device_read_rx, 92); } /* diff --git a/net/core/dev.h b/net/core/dev.h index 0716b1048261..7d0aab7e3ef1 100644 --- a/net/core/dev.h +++ b/net/core/dev.h @@ -184,6 +184,46 @@ static inline void netdev_set_defer_hard_irqs(struct net_device *netdev, napi_set_defer_hard_irqs(napi, defer); } +/** + * napi_get_gro_flush_timeout - get the gro_flush_timeout + * @n: napi struct to get the gro_flush_timeout from + * + * Return: the per-NAPI value of the gro_flush_timeout field. + */ +static inline unsigned long +napi_get_gro_flush_timeout(const struct napi_struct *n) +{ + return READ_ONCE(n->gro_flush_timeout); +} + +/** + * napi_set_gro_flush_timeout - set the gro_flush_timeout for a napi + * @n: napi struct to set the gro_flush_timeout + * @timeout: timeout value to set + * + * napi_set_gro_flush_timeout sets the per-NAPI gro_flush_timeout + */ +static inline void napi_set_gro_flush_timeout(struct napi_struct *n, + unsigned long timeout) +{ + WRITE_ONCE(n->gro_flush_timeout, timeout); +} + +/** + * netdev_set_gro_flush_timeout - set gro_flush_timeout of a netdev's NAPIs + * @netdev: the net_device for which all NAPIs will have gro_flush_timeout set + * @timeout: the timeout value to set + */ +static inline void netdev_set_gro_flush_timeout(struct net_device *netdev, + unsigned long timeout) +{ + struct napi_struct *napi; + + WRITE_ONCE(netdev->gro_flush_timeout, timeout); + list_for_each_entry(napi, &netdev->napi_list, dev_list) + napi_set_gro_flush_timeout(napi, timeout); +} + int rps_cpumask_housekeeping(struct cpumask *mask); #if defined(CONFIG_DEBUG_NET) && defined(CONFIG_BPF_SYSCALL) diff --git a/net/core/net-sysfs.c b/net/core/net-sysfs.c index 25125f356a15..2d9afc6e2161 100644 --- a/net/core/net-sysfs.c +++ b/net/core/net-sysfs.c @@ -409,7 +409,7 @@ NETDEVICE_SHOW_RW(tx_queue_len, fmt_dec); static int change_gro_flush_timeout(struct net_device *dev, unsigned long val) { - WRITE_ONCE(dev->gro_flush_timeout, val); + netdev_set_gro_flush_timeout(dev, val); return 0; } -- cgit v1.2.3 From 86e25f40aa1e9e54e081e55016f65b5c92523989 Mon Sep 17 00:00:00 2001 From: Joe Damato Date: Fri, 11 Oct 2024 18:45:00 +0000 Subject: net: napi: Add napi_config Add a persistent NAPI config area for NAPI configuration to the core. Drivers opt-in to setting the persistent config for a NAPI by passing an index when calling netif_napi_add_config. napi_config is allocated in alloc_netdev_mqs, freed in free_netdev (after the NAPIs are deleted). Drivers which call netif_napi_add_config will have persistent per-NAPI settings: NAPI IDs, gro_flush_timeout, and defer_hard_irq settings. Per-NAPI settings are saved in napi_disable and restored in napi_enable. Co-developed-by: Martin Karsten Signed-off-by: Martin Karsten Signed-off-by: Joe Damato Reviewed-by: Jakub Kicinski Reviewed-by: Eric Dumazet Link: https://patch.msgid.link/20241011184527.16393-6-jdamato@fastly.com Signed-off-by: Jakub Kicinski --- .../networking/net_cachelines/net_device.rst | 1 + include/linux/netdevice.h | 36 +++++++++- net/core/dev.c | 80 ++++++++++++++++++++-- net/core/dev.h | 12 ++++ 4 files changed, 119 insertions(+), 10 deletions(-) (limited to 'include/linux') diff --git a/Documentation/networking/net_cachelines/net_device.rst b/Documentation/networking/net_cachelines/net_device.rst index 67910ea49160..db6192b2bb50 100644 --- a/Documentation/networking/net_cachelines/net_device.rst +++ b/Documentation/networking/net_cachelines/net_device.rst @@ -186,6 +186,7 @@ struct dpll_pin* dpll_pin struct hlist_head page_pools struct dim_irq_moder* irq_moder u64 max_pacing_offload_horizon +struct_napi_config* napi_config unsigned_long gro_flush_timeout u32 napi_defer_hard_irqs =================================== =========================== =================== =================== =================================================================================== diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index 93241d4de437..8feaca12655e 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -342,6 +342,15 @@ struct gro_list { */ #define GRO_HASH_BUCKETS 8 +/* + * Structure for per-NAPI config + */ +struct napi_config { + u64 gro_flush_timeout; + u32 defer_hard_irqs; + unsigned int napi_id; +}; + /* * Structure for NAPI scheduling similar to tasklet but with weighting */ @@ -379,6 +388,8 @@ struct napi_struct { struct list_head dev_list; struct hlist_node napi_hash_node; int irq; + int index; + struct napi_config *config; }; enum { @@ -1868,9 +1879,6 @@ enum netdev_reg_state { * allocated at register_netdev() time * @real_num_rx_queues: Number of RX queues currently active in device * @xdp_prog: XDP sockets filter program pointer - * @gro_flush_timeout: timeout for GRO layer in NAPI - * @napi_defer_hard_irqs: If not zero, provides a counter that would - * allow to avoid NIC hard IRQ, on busy queues. * * @rx_handler: handler for received packets * @rx_handler_data: XXX: need comments on this one @@ -2020,6 +2028,11 @@ enum netdev_reg_state { * where the clock is recovered. * * @max_pacing_offload_horizon: max EDT offload horizon in nsec. + * @napi_config: An array of napi_config structures containing per-NAPI + * settings. + * @gro_flush_timeout: timeout for GRO layer in NAPI + * @napi_defer_hard_irqs: If not zero, provides a counter that would + * allow to avoid NIC hard IRQ, on busy queues. * * FIXME: cleanup struct net_device such that network protocol info * moves out. @@ -2413,6 +2426,7 @@ struct net_device { struct dim_irq_moder *irq_moder; u64 max_pacing_offload_horizon; + struct napi_config *napi_config; unsigned long gro_flush_timeout; u32 napi_defer_hard_irqs; @@ -2678,6 +2692,22 @@ netif_napi_add_tx_weight(struct net_device *dev, netif_napi_add_weight(dev, napi, poll, weight); } +/** + * netif_napi_add_config - initialize a NAPI context with persistent config + * @dev: network device + * @napi: NAPI context + * @poll: polling function + * @index: the NAPI index + */ +static inline void +netif_napi_add_config(struct net_device *dev, struct napi_struct *napi, + int (*poll)(struct napi_struct *, int), int index) +{ + napi->index = index; + napi->config = &dev->napi_config[index]; + netif_napi_add_weight(dev, napi, poll, NAPI_POLL_WEIGHT); +} + /** * netif_napi_add_tx() - initialize a NAPI context to be used for Tx only * @dev: network device diff --git a/net/core/dev.c b/net/core/dev.c index e21ace3551d5..c682173a7642 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -6505,6 +6505,23 @@ EXPORT_SYMBOL(napi_busy_loop); #endif /* CONFIG_NET_RX_BUSY_POLL */ +static void __napi_hash_add_with_id(struct napi_struct *napi, + unsigned int napi_id) +{ + napi->napi_id = napi_id; + hlist_add_head_rcu(&napi->napi_hash_node, + &napi_hash[napi->napi_id % HASH_SIZE(napi_hash)]); +} + +static void napi_hash_add_with_id(struct napi_struct *napi, + unsigned int napi_id) +{ + spin_lock(&napi_hash_lock); + WARN_ON_ONCE(napi_by_id(napi_id)); + __napi_hash_add_with_id(napi, napi_id); + spin_unlock(&napi_hash_lock); +} + static void napi_hash_add(struct napi_struct *napi) { if (test_bit(NAPI_STATE_NO_BUSY_POLL, &napi->state)) @@ -6517,10 +6534,8 @@ static void napi_hash_add(struct napi_struct *napi) if (unlikely(++napi_gen_id < MIN_NAPI_ID)) napi_gen_id = MIN_NAPI_ID; } while (napi_by_id(napi_gen_id)); - napi->napi_id = napi_gen_id; - hlist_add_head_rcu(&napi->napi_hash_node, - &napi_hash[napi->napi_id % HASH_SIZE(napi_hash)]); + __napi_hash_add_with_id(napi, napi_gen_id); spin_unlock(&napi_hash_lock); } @@ -6643,6 +6658,28 @@ void netif_queue_set_napi(struct net_device *dev, unsigned int queue_index, } EXPORT_SYMBOL(netif_queue_set_napi); +static void napi_restore_config(struct napi_struct *n) +{ + n->defer_hard_irqs = n->config->defer_hard_irqs; + n->gro_flush_timeout = n->config->gro_flush_timeout; + /* a NAPI ID might be stored in the config, if so use it. if not, use + * napi_hash_add to generate one for us. It will be saved to the config + * in napi_disable. + */ + if (n->config->napi_id) + napi_hash_add_with_id(n, n->config->napi_id); + else + napi_hash_add(n); +} + +static void napi_save_config(struct napi_struct *n) +{ + n->config->defer_hard_irqs = n->defer_hard_irqs; + n->config->gro_flush_timeout = n->gro_flush_timeout; + n->config->napi_id = n->napi_id; + napi_hash_del(n); +} + void netif_napi_add_weight(struct net_device *dev, struct napi_struct *napi, int (*poll)(struct napi_struct *, int), int weight) { @@ -6653,8 +6690,6 @@ void netif_napi_add_weight(struct net_device *dev, struct napi_struct *napi, INIT_HLIST_NODE(&napi->napi_hash_node); hrtimer_init(&napi->timer, CLOCK_MONOTONIC, HRTIMER_MODE_REL_PINNED); napi->timer.function = napi_watchdog; - napi_set_defer_hard_irqs(napi, READ_ONCE(dev->napi_defer_hard_irqs)); - napi_set_gro_flush_timeout(napi, READ_ONCE(dev->gro_flush_timeout)); init_gro_hash(napi); napi->skb = NULL; INIT_LIST_HEAD(&napi->rx_list); @@ -6672,7 +6707,13 @@ void netif_napi_add_weight(struct net_device *dev, struct napi_struct *napi, set_bit(NAPI_STATE_SCHED, &napi->state); set_bit(NAPI_STATE_NPSVC, &napi->state); list_add_rcu(&napi->dev_list, &dev->napi_list); - napi_hash_add(napi); + + /* default settings from sysfs are applied to all NAPIs. any per-NAPI + * configuration will be loaded in napi_enable + */ + napi_set_defer_hard_irqs(napi, READ_ONCE(dev->napi_defer_hard_irqs)); + napi_set_gro_flush_timeout(napi, READ_ONCE(dev->gro_flush_timeout)); + napi_get_frags_check(napi); /* Create kthread for this napi if dev->threaded is set. * Clear dev->threaded if kthread creation failed so that @@ -6704,6 +6745,11 @@ void napi_disable(struct napi_struct *n) hrtimer_cancel(&n->timer); + if (n->config) + napi_save_config(n); + else + napi_hash_del(n); + clear_bit(NAPI_STATE_DISABLE, &n->state); } EXPORT_SYMBOL(napi_disable); @@ -6719,6 +6765,11 @@ void napi_enable(struct napi_struct *n) { unsigned long new, val = READ_ONCE(n->state); + if (n->config) + napi_restore_config(n); + else + napi_hash_add(n); + do { BUG_ON(!test_bit(NAPI_STATE_SCHED, &val)); @@ -6748,7 +6799,11 @@ void __netif_napi_del(struct napi_struct *napi) if (!test_and_clear_bit(NAPI_STATE_LISTED, &napi->state)) return; - napi_hash_del(napi); + if (napi->config) { + napi->index = -1; + napi->config = NULL; + } + list_del_rcu(&napi->dev_list); napi_free_frags(napi); @@ -11085,6 +11140,8 @@ struct net_device *alloc_netdev_mqs(int sizeof_priv, const char *name, unsigned int txqs, unsigned int rxqs) { struct net_device *dev; + size_t napi_config_sz; + unsigned int maxqs; BUG_ON(strlen(name) >= sizeof(dev->name)); @@ -11098,6 +11155,8 @@ struct net_device *alloc_netdev_mqs(int sizeof_priv, const char *name, return NULL; } + maxqs = max(txqs, rxqs); + dev = kvzalloc(struct_size(dev, priv, sizeof_priv), GFP_KERNEL_ACCOUNT | __GFP_RETRY_MAYFAIL); if (!dev) @@ -11174,6 +11233,11 @@ struct net_device *alloc_netdev_mqs(int sizeof_priv, const char *name, if (!dev->ethtool) goto free_all; + napi_config_sz = array_size(maxqs, sizeof(*dev->napi_config)); + dev->napi_config = kvzalloc(napi_config_sz, GFP_KERNEL_ACCOUNT); + if (!dev->napi_config) + goto free_all; + strscpy(dev->name, name); dev->name_assign_type = name_assign_type; dev->group = INIT_NETDEV_GROUP; @@ -11237,6 +11301,8 @@ void free_netdev(struct net_device *dev) list_for_each_entry_safe(p, n, &dev->napi_list, dev_list) netif_napi_del(p); + kvfree(dev->napi_config); + ref_tracker_dir_exit(&dev->refcnt_tracker); #ifdef CONFIG_PCPU_DEV_REFCNT free_percpu(dev->pcpu_refcnt); diff --git a/net/core/dev.h b/net/core/dev.h index 7d0aab7e3ef1..7881bced70a9 100644 --- a/net/core/dev.h +++ b/net/core/dev.h @@ -177,11 +177,17 @@ static inline void napi_set_defer_hard_irqs(struct napi_struct *n, u32 defer) static inline void netdev_set_defer_hard_irqs(struct net_device *netdev, u32 defer) { + unsigned int count = max(netdev->num_rx_queues, + netdev->num_tx_queues); struct napi_struct *napi; + int i; WRITE_ONCE(netdev->napi_defer_hard_irqs, defer); list_for_each_entry(napi, &netdev->napi_list, dev_list) napi_set_defer_hard_irqs(napi, defer); + + for (i = 0; i < count; i++) + netdev->napi_config[i].defer_hard_irqs = defer; } /** @@ -217,11 +223,17 @@ static inline void napi_set_gro_flush_timeout(struct napi_struct *n, static inline void netdev_set_gro_flush_timeout(struct net_device *netdev, unsigned long timeout) { + unsigned int count = max(netdev->num_rx_queues, + netdev->num_tx_queues); struct napi_struct *napi; + int i; WRITE_ONCE(netdev->gro_flush_timeout, timeout); list_for_each_entry(napi, &netdev->napi_list, dev_list) napi_set_gro_flush_timeout(napi, timeout); + + for (i = 0; i < count; i++) + netdev->napi_config[i].gro_flush_timeout = timeout; } int rps_cpumask_housekeeping(struct cpumask *mask); -- cgit v1.2.3 From a274465cc3bef2dfd9c9ea5100848dda0a8641e1 Mon Sep 17 00:00:00 2001 From: Daniel Golle Date: Thu, 10 Oct 2024 13:54:19 +0100 Subject: net: phy: support 'active-high' property for PHY LEDs In addition to 'active-low' and 'inactive-high-impedance' also support 'active-high' property for PHY LED pin configuration. As only either 'active-high' or 'active-low' can be set at the same time, WARN and return an error in case both are set. Signed-off-by: Daniel Golle Reviewed-by: Andrew Lunn Link: https://patch.msgid.link/91598487773d768f254d5faf06cf65b13e972f0e.1728558223.git.daniel@makrotopia.org Signed-off-by: Paolo Abeni --- drivers/net/phy/phy_device.c | 6 ++++++ include/linux/phy.h | 5 +++-- 2 files changed, 9 insertions(+), 2 deletions(-) (limited to 'include/linux') diff --git a/drivers/net/phy/phy_device.c b/drivers/net/phy/phy_device.c index 94da999b6fde..563497a3274c 100644 --- a/drivers/net/phy/phy_device.c +++ b/drivers/net/phy/phy_device.c @@ -3358,11 +3358,17 @@ static int of_phy_led(struct phy_device *phydev, if (index > U8_MAX) return -EINVAL; + if (of_property_read_bool(led, "active-high")) + set_bit(PHY_LED_ACTIVE_HIGH, &modes); if (of_property_read_bool(led, "active-low")) set_bit(PHY_LED_ACTIVE_LOW, &modes); if (of_property_read_bool(led, "inactive-high-impedance")) set_bit(PHY_LED_INACTIVE_HIGH_IMPEDANCE, &modes); + if (WARN_ON(modes & BIT(PHY_LED_ACTIVE_LOW) && + modes & BIT(PHY_LED_ACTIVE_HIGH))) + return -EINVAL; + if (modes) { /* Return error if asked to set polarity modes but not supported */ if (!phydev->drv->led_polarity_set) diff --git a/include/linux/phy.h b/include/linux/phy.h index ff762a3d8270..bf0eb4e5d35c 100644 --- a/include/linux/phy.h +++ b/include/linux/phy.h @@ -877,8 +877,9 @@ struct phy_plca_status { /* Modes for PHY LED configuration */ enum phy_led_modes { - PHY_LED_ACTIVE_LOW = 0, - PHY_LED_INACTIVE_HIGH_IMPEDANCE = 1, + PHY_LED_ACTIVE_HIGH = 0, + PHY_LED_ACTIVE_LOW = 1, + PHY_LED_INACTIVE_HIGH_IMPEDANCE = 2, /* keep it last */ __PHY_LED_MODES_NUM, -- cgit v1.2.3 From d811ac148f0afd2f3f7e1cd7f54de8da973ec5e3 Mon Sep 17 00:00:00 2001 From: Paolo Abeni Date: Tue, 15 Oct 2024 15:56:35 +0200 Subject: virtchnl: fix m68k build. The kernel test robot reported a build failure on m68k in the intel driver due to the recent shapers-related changes. The mentioned arch has funny alignment properties, let's be explicit about the binary layout expectation introducing a padding field. Fixes: 608a5c05c39b ("virtchnl: support queue rate limit and quanta size configuration") Reported-by: kernel test robot Closes: https://lore.kernel.org/oe-kbuild-all/202410131710.71Wt6LKO-lkp@intel.com/ Reviewed-by: Alexander Lobakin Reviewed-by: Paul Menzel Reviewed-by: Jacob Keller Link: https://patch.msgid.link/e45d1c9f17356d431b03b419f60b8b763d2ff768.1729000481.git.pabeni@redhat.com Signed-off-by: Paolo Abeni --- include/linux/avf/virtchnl.h | 1 + 1 file changed, 1 insertion(+) (limited to 'include/linux') diff --git a/include/linux/avf/virtchnl.h b/include/linux/avf/virtchnl.h index 223e433c39fe..13a11f3c09b8 100644 --- a/include/linux/avf/virtchnl.h +++ b/include/linux/avf/virtchnl.h @@ -1499,6 +1499,7 @@ VIRTCHNL_CHECK_STRUCT_LEN(8, virtchnl_queue_chunk); struct virtchnl_quanta_cfg { u16 quanta_size; + u16 pad; struct virtchnl_queue_chunk queue_select; }; -- cgit v1.2.3 From e44ef3f66c5472c2cbc6957c684d7279c26b0db1 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Fri, 18 Oct 2024 05:21:08 +0000 Subject: netpoll: remove ndo_netpoll_setup() second argument npinfo is not used in any of the ndo_netpoll_setup() methods. Signed-off-by: Eric Dumazet Reviewed-by: Simon Horman Link: https://patch.msgid.link/20241018052108.2610827-1-edumazet@google.com Signed-off-by: Paolo Abeni --- drivers/net/bonding/bond_main.c | 2 +- drivers/net/macvlan.c | 2 +- drivers/net/team/team_core.c | 3 +-- include/linux/netdevice.h | 3 +-- net/8021q/vlan_dev.c | 2 +- net/bridge/br_device.c | 2 +- net/core/netpoll.c | 2 +- net/dsa/user.c | 3 +-- 8 files changed, 8 insertions(+), 11 deletions(-) (limited to 'include/linux') diff --git a/drivers/net/bonding/bond_main.c b/drivers/net/bonding/bond_main.c index b1bffd8e9a95..3928287f5865 100644 --- a/drivers/net/bonding/bond_main.c +++ b/drivers/net/bonding/bond_main.c @@ -1476,7 +1476,7 @@ static void bond_netpoll_cleanup(struct net_device *bond_dev) slave_disable_netpoll(slave); } -static int bond_netpoll_setup(struct net_device *dev, struct netpoll_info *ni) +static int bond_netpoll_setup(struct net_device *dev) { struct bonding *bond = netdev_priv(dev); struct list_head *iter; diff --git a/drivers/net/macvlan.c b/drivers/net/macvlan.c index cf18e66de142..edbd5afcec41 100644 --- a/drivers/net/macvlan.c +++ b/drivers/net/macvlan.c @@ -1117,7 +1117,7 @@ static void macvlan_dev_poll_controller(struct net_device *dev) return; } -static int macvlan_dev_netpoll_setup(struct net_device *dev, struct netpoll_info *npinfo) +static int macvlan_dev_netpoll_setup(struct net_device *dev) { struct macvlan_dev *vlan = netdev_priv(dev); struct net_device *real_dev = vlan->lowerdev; diff --git a/drivers/net/team/team_core.c b/drivers/net/team/team_core.c index 18191d5a8bd4..a1b27b69f010 100644 --- a/drivers/net/team/team_core.c +++ b/drivers/net/team/team_core.c @@ -1946,8 +1946,7 @@ static void team_netpoll_cleanup(struct net_device *dev) mutex_unlock(&team->lock); } -static int team_netpoll_setup(struct net_device *dev, - struct netpoll_info *npifo) +static int team_netpoll_setup(struct net_device *dev) { struct team *team = netdev_priv(dev); struct team_port *port; diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index 8feaca12655e..86a0b7eb9461 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -1425,8 +1425,7 @@ struct net_device_ops { __be16 proto, u16 vid); #ifdef CONFIG_NET_POLL_CONTROLLER void (*ndo_poll_controller)(struct net_device *dev); - int (*ndo_netpoll_setup)(struct net_device *dev, - struct netpoll_info *info); + int (*ndo_netpoll_setup)(struct net_device *dev); void (*ndo_netpoll_cleanup)(struct net_device *dev); #endif int (*ndo_set_vf_mac)(struct net_device *dev, diff --git a/net/8021q/vlan_dev.c b/net/8021q/vlan_dev.c index 458040e8a0e0..91d134961357 100644 --- a/net/8021q/vlan_dev.c +++ b/net/8021q/vlan_dev.c @@ -725,7 +725,7 @@ static void vlan_dev_poll_controller(struct net_device *dev) return; } -static int vlan_dev_netpoll_setup(struct net_device *dev, struct netpoll_info *npinfo) +static int vlan_dev_netpoll_setup(struct net_device *dev) { struct vlan_dev_priv *vlan = vlan_dev_priv(dev); struct net_device *real_dev = vlan->real_dev; diff --git a/net/bridge/br_device.c b/net/bridge/br_device.c index 26b79feb385d..0ab4613aa07a 100644 --- a/net/bridge/br_device.c +++ b/net/bridge/br_device.c @@ -328,7 +328,7 @@ int br_netpoll_enable(struct net_bridge_port *p) return __br_netpoll_enable(p); } -static int br_netpoll_setup(struct net_device *dev, struct netpoll_info *ni) +static int br_netpoll_setup(struct net_device *dev) { struct net_bridge *br = netdev_priv(dev); struct net_bridge_port *p; diff --git a/net/core/netpoll.c b/net/core/netpoll.c index aa49b92e9194..94b7f07a952f 100644 --- a/net/core/netpoll.c +++ b/net/core/netpoll.c @@ -641,7 +641,7 @@ int __netpoll_setup(struct netpoll *np, struct net_device *ndev) ops = ndev->netdev_ops; if (ops->ndo_netpoll_setup) { - err = ops->ndo_netpoll_setup(ndev, npinfo); + err = ops->ndo_netpoll_setup(ndev); if (err) goto free_npinfo; } diff --git a/net/dsa/user.c b/net/dsa/user.c index 64f660d2334b..91a1fa5f8ab0 100644 --- a/net/dsa/user.c +++ b/net/dsa/user.c @@ -1308,8 +1308,7 @@ static int dsa_user_set_pauseparam(struct net_device *dev, } #ifdef CONFIG_NET_POLL_CONTROLLER -static int dsa_user_netpoll_setup(struct net_device *dev, - struct netpoll_info *ni) +static int dsa_user_netpoll_setup(struct net_device *dev) { struct net_device *conduit = dsa_user_to_conduit(dev); struct dsa_user_priv *p = netdev_priv(dev); -- cgit v1.2.3 From 7cfc1b1fa8673fe386304194d4cc2c8fe555bbf9 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Fri, 18 Oct 2024 05:23:10 +0000 Subject: net: netdev_tx_sent_queue() small optimization Change smp_mb() imediately following a set_bit() with smp_mb__after_atomic(). Signed-off-by: Eric Dumazet Link: https://patch.msgid.link/20241018052310.2612084-1-edumazet@google.com Signed-off-by: Paolo Abeni --- include/linux/netdevice.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include/linux') diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index 86a0b7eb9461..bbd30f3c5d29 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -3517,7 +3517,7 @@ static inline void netdev_tx_sent_queue(struct netdev_queue *dev_queue, * because in netdev_tx_completed_queue we update the dql_completed * before checking the XOFF flag. */ - smp_mb(); + smp_mb__after_atomic(); /* check again in case another CPU has just made room avail */ if (unlikely(dql_avail(&dev_queue->dql) >= 0)) -- cgit v1.2.3 From 8dc6d81c6b2acc434b00c4585f0594739031c4e4 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Tue, 22 Oct 2024 15:18:34 +0200 Subject: debugfs: add small file operations for most files As struct file_operations is really big, but (most) debugfs files only use simple_open, read, write and perhaps seek, and don't need anything else, this wastes a lot of space for NULL pointers. Add a struct debugfs_short_fops and some bookkeeping code in debugfs so that users can use that with debugfs_create_file() using _Generic to figure out which function to use. Converting mac80211 to use it where possible saves quite a bit of space: 1010127 205064 1220 1216411 128f9b net/mac80211/mac80211.ko (before) 981199 205064 1220 1187483 121e9b net/mac80211/mac80211.ko (after) ------- -28928 = ~28KiB With a marginal space cost in debugfs: 8701 550 16 9267 2433 fs/debugfs/inode.o (before) 25233 325 32 25590 63f6 fs/debugfs/file.o (before) 8914 558 16 9488 2510 fs/debugfs/inode.o (after) 25380 325 32 25737 6489 fs/debugfs/file.o (after) --------------- +360 +8 (All on x86-64) A simple spatch suggests there are more than 300 instances, not even counting the ones hidden in macros like in mac80211, that could be trivially converted, for additional savings of about 240 bytes for each. Reviewed-by: Greg Kroah-Hartman Link: https://patch.msgid.link/20241022151838.26f9925fb959.Ia80b55e934bbfc45ce0df42a3233d34b35508046@changeid Signed-off-by: Johannes Berg --- fs/debugfs/file.c | 100 +++++++++++++++++++++++++++++++++--------------- fs/debugfs/inode.c | 63 +++++++++++++----------------- fs/debugfs/internal.h | 6 +++ include/linux/debugfs.h | 62 ++++++++++++++++++++++++++++-- 4 files changed, 160 insertions(+), 71 deletions(-) (limited to 'include/linux') diff --git a/fs/debugfs/file.c b/fs/debugfs/file.c index 67299e8b734e..47dc96dfe386 100644 --- a/fs/debugfs/file.c +++ b/fs/debugfs/file.c @@ -100,8 +100,16 @@ int debugfs_file_get(struct dentry *dentry) if (!fsd) return -ENOMEM; - fsd->real_fops = (void *)((unsigned long)d_fsd & - ~DEBUGFS_FSDATA_IS_REAL_FOPS_BIT); + if ((unsigned long)d_fsd & DEBUGFS_FSDATA_IS_SHORT_FOPS_BIT) { + fsd->real_fops = NULL; + fsd->short_fops = (void *)((unsigned long)d_fsd & + ~(DEBUGFS_FSDATA_IS_REAL_FOPS_BIT | + DEBUGFS_FSDATA_IS_SHORT_FOPS_BIT)); + } else { + fsd->real_fops = (void *)((unsigned long)d_fsd & + ~DEBUGFS_FSDATA_IS_REAL_FOPS_BIT); + fsd->short_fops = NULL; + } refcount_set(&fsd->active_users, 1); init_completion(&fsd->active_users_drained); INIT_LIST_HEAD(&fsd->cancellations); @@ -241,9 +249,10 @@ static int debugfs_locked_down(struct inode *inode, { if ((inode->i_mode & 07777 & ~0444) == 0 && !(filp->f_mode & FMODE_WRITE) && - !real_fops->unlocked_ioctl && - !real_fops->compat_ioctl && - !real_fops->mmap) + (!real_fops || + (!real_fops->unlocked_ioctl && + !real_fops->compat_ioctl && + !real_fops->mmap))) return 0; if (security_locked_down(LOCKDOWN_DEBUGFS)) @@ -316,19 +325,38 @@ static ret_type full_proxy_ ## name(proto) \ return r; \ } -FULL_PROXY_FUNC(llseek, loff_t, filp, - PROTO(struct file *filp, loff_t offset, int whence), - ARGS(filp, offset, whence)); +#define FULL_PROXY_FUNC_BOTH(name, ret_type, filp, proto, args) \ +static ret_type full_proxy_ ## name(proto) \ +{ \ + struct dentry *dentry = F_DENTRY(filp); \ + struct debugfs_fsdata *fsd; \ + ret_type r; \ + \ + r = debugfs_file_get(dentry); \ + if (unlikely(r)) \ + return r; \ + fsd = dentry->d_fsdata; \ + if (fsd->real_fops) \ + r = fsd->real_fops->name(args); \ + else \ + r = fsd->short_fops->name(args); \ + debugfs_file_put(dentry); \ + return r; \ +} + +FULL_PROXY_FUNC_BOTH(llseek, loff_t, filp, + PROTO(struct file *filp, loff_t offset, int whence), + ARGS(filp, offset, whence)); -FULL_PROXY_FUNC(read, ssize_t, filp, - PROTO(struct file *filp, char __user *buf, size_t size, - loff_t *ppos), - ARGS(filp, buf, size, ppos)); +FULL_PROXY_FUNC_BOTH(read, ssize_t, filp, + PROTO(struct file *filp, char __user *buf, size_t size, + loff_t *ppos), + ARGS(filp, buf, size, ppos)); -FULL_PROXY_FUNC(write, ssize_t, filp, - PROTO(struct file *filp, const char __user *buf, size_t size, - loff_t *ppos), - ARGS(filp, buf, size, ppos)); +FULL_PROXY_FUNC_BOTH(write, ssize_t, filp, + PROTO(struct file *filp, const char __user *buf, + size_t size, loff_t *ppos), + ARGS(filp, buf, size, ppos)); FULL_PROXY_FUNC(unlocked_ioctl, long, filp, PROTO(struct file *filp, unsigned int cmd, unsigned long arg), @@ -363,7 +391,7 @@ static int full_proxy_release(struct inode *inode, struct file *filp) * not to leak any resources. Releasers must not assume that * ->i_private is still being meaningful here. */ - if (real_fops->release) + if (real_fops && real_fops->release) r = real_fops->release(inode, filp); replace_fops(filp, d_inode(dentry)->i_fop); @@ -373,39 +401,48 @@ static int full_proxy_release(struct inode *inode, struct file *filp) } static void __full_proxy_fops_init(struct file_operations *proxy_fops, - const struct file_operations *real_fops) + struct debugfs_fsdata *fsd) { proxy_fops->release = full_proxy_release; - if (real_fops->llseek) + + if ((fsd->real_fops && fsd->real_fops->llseek) || + (fsd->short_fops && fsd->short_fops->llseek)) proxy_fops->llseek = full_proxy_llseek; - if (real_fops->read) + + if ((fsd->real_fops && fsd->real_fops->read) || + (fsd->short_fops && fsd->short_fops->read)) proxy_fops->read = full_proxy_read; - if (real_fops->write) + + if ((fsd->real_fops && fsd->real_fops->write) || + (fsd->short_fops && fsd->short_fops->write)) proxy_fops->write = full_proxy_write; - if (real_fops->poll) + + if (fsd->real_fops && fsd->real_fops->poll) proxy_fops->poll = full_proxy_poll; - if (real_fops->unlocked_ioctl) + + if (fsd->real_fops && fsd->real_fops->unlocked_ioctl) proxy_fops->unlocked_ioctl = full_proxy_unlocked_ioctl; } static int full_proxy_open(struct inode *inode, struct file *filp) { struct dentry *dentry = F_DENTRY(filp); - const struct file_operations *real_fops = NULL; + const struct file_operations *real_fops; struct file_operations *proxy_fops = NULL; + struct debugfs_fsdata *fsd; int r; r = debugfs_file_get(dentry); if (r) return r == -EIO ? -ENOENT : r; - real_fops = debugfs_real_fops(filp); - + fsd = dentry->d_fsdata; + real_fops = fsd->real_fops; r = debugfs_locked_down(inode, filp, real_fops); if (r) goto out; - if (!fops_get(real_fops)) { + if (real_fops && !fops_get(real_fops)) { #ifdef CONFIG_MODULES if (real_fops->owner && real_fops->owner->state == MODULE_STATE_GOING) { @@ -426,11 +463,14 @@ static int full_proxy_open(struct inode *inode, struct file *filp) r = -ENOMEM; goto free_proxy; } - __full_proxy_fops_init(proxy_fops, real_fops); + __full_proxy_fops_init(proxy_fops, fsd); replace_fops(filp, proxy_fops); - if (real_fops->open) { - r = real_fops->open(inode, filp); + if (!real_fops || real_fops->open) { + if (real_fops) + r = real_fops->open(inode, filp); + else + r = simple_open(inode, filp); if (r) { replace_fops(filp, d_inode(dentry)->i_fop); goto free_proxy; diff --git a/fs/debugfs/inode.c b/fs/debugfs/inode.c index 66d9b3b4c588..38a9c7eb97e6 100644 --- a/fs/debugfs/inode.c +++ b/fs/debugfs/inode.c @@ -412,7 +412,7 @@ static struct dentry *end_creating(struct dentry *dentry) static struct dentry *__debugfs_create_file(const char *name, umode_t mode, struct dentry *parent, void *data, const struct file_operations *proxy_fops, - const struct file_operations *real_fops) + const void *real_fops) { struct dentry *dentry; struct inode *inode; @@ -450,49 +450,38 @@ static struct dentry *__debugfs_create_file(const char *name, umode_t mode, return end_creating(dentry); } -/** - * debugfs_create_file - create a file in the debugfs filesystem - * @name: a pointer to a string containing the name of the file to create. - * @mode: the permission that the file should have. - * @parent: a pointer to the parent dentry for this file. This should be a - * directory dentry if set. If this parameter is NULL, then the - * file will be created in the root of the debugfs filesystem. - * @data: a pointer to something that the caller will want to get to later - * on. The inode.i_private pointer will point to this value on - * the open() call. - * @fops: a pointer to a struct file_operations that should be used for - * this file. - * - * This is the basic "create a file" function for debugfs. It allows for a - * wide range of flexibility in creating a file, or a directory (if you want - * to create a directory, the debugfs_create_dir() function is - * recommended to be used instead.) - * - * This function will return a pointer to a dentry if it succeeds. This - * pointer must be passed to the debugfs_remove() function when the file is - * to be removed (no automatic cleanup happens if your module is unloaded, - * you are responsible here.) If an error occurs, ERR_PTR(-ERROR) will be - * returned. - * - * If debugfs is not enabled in the kernel, the value -%ENODEV will be - * returned. - * - * NOTE: it's expected that most callers should _ignore_ the errors returned - * by this function. Other debugfs functions handle the fact that the "dentry" - * passed to them could be an error and they don't crash in that case. - * Drivers should generally work fine even if debugfs fails to init anyway. - */ -struct dentry *debugfs_create_file(const char *name, umode_t mode, - struct dentry *parent, void *data, - const struct file_operations *fops) +struct dentry *debugfs_create_file_full(const char *name, umode_t mode, + struct dentry *parent, void *data, + const struct file_operations *fops) { + if (WARN_ON((unsigned long)fops & + (DEBUGFS_FSDATA_IS_SHORT_FOPS_BIT | + DEBUGFS_FSDATA_IS_REAL_FOPS_BIT))) + return ERR_PTR(-EINVAL); return __debugfs_create_file(name, mode, parent, data, fops ? &debugfs_full_proxy_file_operations : &debugfs_noop_file_operations, fops); } -EXPORT_SYMBOL_GPL(debugfs_create_file); +EXPORT_SYMBOL_GPL(debugfs_create_file_full); + +struct dentry *debugfs_create_file_short(const char *name, umode_t mode, + struct dentry *parent, void *data, + const struct debugfs_short_fops *fops) +{ + if (WARN_ON((unsigned long)fops & + (DEBUGFS_FSDATA_IS_SHORT_FOPS_BIT | + DEBUGFS_FSDATA_IS_REAL_FOPS_BIT))) + return ERR_PTR(-EINVAL); + + return __debugfs_create_file(name, mode, parent, data, + fops ? &debugfs_full_proxy_file_operations : + &debugfs_noop_file_operations, + (const void *)((unsigned long)fops | + DEBUGFS_FSDATA_IS_SHORT_FOPS_BIT)); +} +EXPORT_SYMBOL_GPL(debugfs_create_file_short); /** * debugfs_create_file_unsafe - create a file in the debugfs filesystem diff --git a/fs/debugfs/internal.h b/fs/debugfs/internal.h index dae80c2a469e..a3edfa4f0d8e 100644 --- a/fs/debugfs/internal.h +++ b/fs/debugfs/internal.h @@ -18,6 +18,7 @@ extern const struct file_operations debugfs_full_proxy_file_operations; struct debugfs_fsdata { const struct file_operations *real_fops; + const struct debugfs_short_fops *short_fops; union { /* automount_fn is used when real_fops is NULL */ debugfs_automount_t automount; @@ -39,6 +40,11 @@ struct debugfs_fsdata { * pointer gets its lowest bit set. */ #define DEBUGFS_FSDATA_IS_REAL_FOPS_BIT BIT(0) +/* + * A dentry's ->d_fsdata, when pointing to real fops, is with + * short fops instead of full fops. + */ +#define DEBUGFS_FSDATA_IS_SHORT_FOPS_BIT BIT(1) /* Access BITS */ #define DEBUGFS_ALLOW_API BIT(0) diff --git a/include/linux/debugfs.h b/include/linux/debugfs.h index 0928a6c8ae1e..59444b495d49 100644 --- a/include/linux/debugfs.h +++ b/include/linux/debugfs.h @@ -71,9 +71,63 @@ typedef struct vfsmount *(*debugfs_automount_t)(struct dentry *, void *); struct dentry *debugfs_lookup(const char *name, struct dentry *parent); -struct dentry *debugfs_create_file(const char *name, umode_t mode, - struct dentry *parent, void *data, - const struct file_operations *fops); +struct debugfs_short_fops { + ssize_t (*read)(struct file *, char __user *, size_t, loff_t *); + ssize_t (*write)(struct file *, const char __user *, size_t, loff_t *); + loff_t (*llseek) (struct file *, loff_t, int); +}; + +struct dentry *debugfs_create_file_full(const char *name, umode_t mode, + struct dentry *parent, void *data, + const struct file_operations *fops); +struct dentry *debugfs_create_file_short(const char *name, umode_t mode, + struct dentry *parent, void *data, + const struct debugfs_short_fops *fops); + +/** + * debugfs_create_file - create a file in the debugfs filesystem + * @name: a pointer to a string containing the name of the file to create. + * @mode: the permission that the file should have. + * @parent: a pointer to the parent dentry for this file. This should be a + * directory dentry if set. If this parameter is NULL, then the + * file will be created in the root of the debugfs filesystem. + * @data: a pointer to something that the caller will want to get to later + * on. The inode.i_private pointer will point to this value on + * the open() call. + * @fops: a pointer to a struct file_operations or struct debugfs_short_fops that + * should be used for this file. + * + * This is the basic "create a file" function for debugfs. It allows for a + * wide range of flexibility in creating a file, or a directory (if you want + * to create a directory, the debugfs_create_dir() function is + * recommended to be used instead.) + * + * This function will return a pointer to a dentry if it succeeds. This + * pointer must be passed to the debugfs_remove() function when the file is + * to be removed (no automatic cleanup happens if your module is unloaded, + * you are responsible here.) If an error occurs, ERR_PTR(-ERROR) will be + * returned. + * + * If debugfs is not enabled in the kernel, the value -%ENODEV will be + * returned. + * + * If fops points to a struct debugfs_short_fops, then simple_open() will be + * used for the open, and only read/write/llseek are supported and are proxied, + * so no module reference or release are needed. + * + * NOTE: it's expected that most callers should _ignore_ the errors returned + * by this function. Other debugfs functions handle the fact that the "dentry" + * passed to them could be an error and they don't crash in that case. + * Drivers should generally work fine even if debugfs fails to init anyway. + */ +#define debugfs_create_file(name, mode, parent, data, fops) \ + _Generic(fops, \ + const struct file_operations *: debugfs_create_file_full, \ + const struct debugfs_short_fops *: debugfs_create_file_short, \ + struct file_operations *: debugfs_create_file_full, \ + struct debugfs_short_fops *: debugfs_create_file_short) \ + (name, mode, parent, data, fops) + struct dentry *debugfs_create_file_unsafe(const char *name, umode_t mode, struct dentry *parent, void *data, const struct file_operations *fops); @@ -207,7 +261,7 @@ static inline struct dentry *debugfs_lookup(const char *name, static inline struct dentry *debugfs_create_file(const char *name, umode_t mode, struct dentry *parent, void *data, - const struct file_operations *fops) + const void *fops) { return ERR_PTR(-ENODEV); } -- cgit v1.2.3 From 9cb7e40d388d6c0e4677809c6b2950bc67fd8830 Mon Sep 17 00:00:00 2001 From: Kuniyuki Iwashima Date: Mon, 21 Oct 2024 11:32:28 -0700 Subject: rtnetlink: Make per-netns RTNL dereference helpers to macro. MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit When CONFIG_DEBUG_NET_SMALL_RTNL is off, rtnl_net_dereference() is the static inline wrapper of rtnl_dereference() returning a plain (void *) pointer to make sure net is always evaluated as requested in [0]. But, it makes sparse complain [1] when the pointer has __rcu annotation: net/ipv4/devinet.c:674:47: sparse: warning: incorrect type in argument 2 (different address spaces) net/ipv4/devinet.c:674:47: sparse: expected void *p net/ipv4/devinet.c:674:47: sparse: got struct in_ifaddr [noderef] __rcu * Also, if we evaluate net as (void *) in a macro, then the compiler in turn fails to build due to -Werror=unused-value. #define rtnl_net_dereference(net, p) \ ({ \ (void *)net; \ rtnl_dereference(p); \ }) net/ipv4/devinet.c: In function ‘inet_rtm_deladdr’: ./include/linux/rtnetlink.h:154:17: error: statement with no effect [-Werror=unused-value] 154 | (void *)net; \ net/ipv4/devinet.c:674:21: note: in expansion of macro ‘rtnl_net_dereference’ 674 | (ifa = rtnl_net_dereference(net, *ifap)) != NULL; | ^~~~~~~~~~~~~~~~~~~~ Let's go back to the original simplest macro. Note that checkpatch complains about this approach, but it's one-shot and less noisy than the other two. WARNING: Argument 'net' is not used in function-like macro #76: FILE: include/linux/rtnetlink.h:142: +#define rtnl_net_dereference(net, p) \ + rtnl_dereference(p) Fixes: 844e5e7e656d ("rtnetlink: Add assertion helpers for per-netns RTNL.") Link: https://lore.kernel.org/netdev/20241004132145.7fd208e9@kernel.org/ [0] Reported-by: kernel test robot Closes: https://lore.kernel.org/oe-kbuild-all/202410200325.SaEJmyZS-lkp@intel.com/ [1] Signed-off-by: Kuniyuki Iwashima Reviewed-by: Simon Horman Signed-off-by: Paolo Abeni --- include/linux/rtnetlink.h | 21 ++++++--------------- 1 file changed, 6 insertions(+), 15 deletions(-) (limited to 'include/linux') diff --git a/include/linux/rtnetlink.h b/include/linux/rtnetlink.h index 8468a4ce8510..0e62918de63b 100644 --- a/include/linux/rtnetlink.h +++ b/include/linux/rtnetlink.h @@ -137,21 +137,12 @@ static inline void ASSERT_RTNL_NET(struct net *net) ASSERT_RTNL(); } -static inline void *rcu_dereference_rtnl_net(struct net *net, void *p) -{ - return rcu_dereference_rtnl(p); -} - -static inline void *rtnl_net_dereference(struct net *net, void *p) -{ - return rtnl_dereference(p); -} - -static inline void *rcu_replace_pointer_rtnl_net(struct net *net, - void *rp, void *p) -{ - return rcu_replace_pointer_rtnl(rp, p); -} +#define rcu_dereference_rtnl_net(net, p) \ + rcu_dereference_rtnl(p) +#define rtnl_net_dereference(net, p) \ + rtnl_dereference(p) +#define rcu_replace_pointer_rtnl_net(net, rp, p) \ + rcu_replace_pointer_rtnl(rp, p) #endif static inline struct netdev_queue *dev_ingress_queue(struct net_device *dev) -- cgit v1.2.3 From d4b483208b2606add41a22bdd3c8cd6d36009319 Mon Sep 17 00:00:00 2001 From: Kuniyuki Iwashima Date: Mon, 21 Oct 2024 11:32:33 -0700 Subject: ipv4: Use per-netns RTNL helpers in inet_rtm_newaddr(). inet_rtm_to_ifa() and find_matching_ifa() are called under rtnl_net_lock(). __in_dev_get_rtnl() and in_dev_for_each_ifa_rtnl() there can use per-netns RTNL helpers. Let's define and use __in_dev_get_rtnl_net() and in_dev_for_each_ifa_rtnl_net(). Signed-off-by: Kuniyuki Iwashima Reviewed-by: Eric Dumazet Signed-off-by: Paolo Abeni --- include/linux/inetdevice.h | 9 +++++++++ net/ipv4/devinet.c | 8 ++++---- 2 files changed, 13 insertions(+), 4 deletions(-) (limited to 'include/linux') diff --git a/include/linux/inetdevice.h b/include/linux/inetdevice.h index d9c690c8c80b..5730ba6b1cfa 100644 --- a/include/linux/inetdevice.h +++ b/include/linux/inetdevice.h @@ -226,6 +226,10 @@ static __inline__ bool bad_mask(__be32 mask, __be32 addr) for (ifa = rtnl_dereference((in_dev)->ifa_list); ifa; \ ifa = rtnl_dereference(ifa->ifa_next)) +#define in_dev_for_each_ifa_rtnl_net(net, ifa, in_dev) \ + for (ifa = rtnl_net_dereference(net, (in_dev)->ifa_list); ifa; \ + ifa = rtnl_net_dereference(net, ifa->ifa_next)) + #define in_dev_for_each_ifa_rcu(ifa, in_dev) \ for (ifa = rcu_dereference((in_dev)->ifa_list); ifa; \ ifa = rcu_dereference(ifa->ifa_next)) @@ -252,6 +256,11 @@ static inline struct in_device *__in_dev_get_rtnl(const struct net_device *dev) return rtnl_dereference(dev->ip_ptr); } +static inline struct in_device *__in_dev_get_rtnl_net(const struct net_device *dev) +{ + return rtnl_net_dereference(dev_net(dev), dev->ip_ptr); +} + /* called with rcu_read_lock or rtnl held */ static inline bool ip_ignore_linkdown(const struct net_device *dev) { diff --git a/net/ipv4/devinet.c b/net/ipv4/devinet.c index 7f24bc38981b..e14e35c22054 100644 --- a/net/ipv4/devinet.c +++ b/net/ipv4/devinet.c @@ -901,7 +901,7 @@ static struct in_ifaddr *inet_rtm_to_ifa(struct net *net, struct nlmsghdr *nlh, goto errout; } - in_dev = __in_dev_get_rtnl(dev); + in_dev = __in_dev_get_rtnl_net(dev); err = -ENOBUFS; if (!in_dev) goto errout; @@ -948,12 +948,12 @@ errout: return ERR_PTR(err); } -static struct in_ifaddr *find_matching_ifa(struct in_ifaddr *ifa) +static struct in_ifaddr *find_matching_ifa(struct net *net, struct in_ifaddr *ifa) { struct in_device *in_dev = ifa->ifa_dev; struct in_ifaddr *ifa1; - in_dev_for_each_ifa_rtnl(ifa1, in_dev) { + in_dev_for_each_ifa_rtnl_net(net, ifa1, in_dev) { if (ifa1->ifa_mask == ifa->ifa_mask && inet_ifa_match(ifa1->ifa_address, ifa) && ifa1->ifa_local == ifa->ifa_local) @@ -989,7 +989,7 @@ static int inet_rtm_newaddr(struct sk_buff *skb, struct nlmsghdr *nlh, goto unlock; } - ifa_existing = find_matching_ifa(ifa); + ifa_existing = find_matching_ifa(net, ifa); if (!ifa_existing) { /* It would be best to check for !NLM_F_CREATE here but * userspace already relies on not having to provide this. -- cgit v1.2.3 From d1c81818aa227b37d65b40f9883109c5256b9bfb Mon Sep 17 00:00:00 2001 From: Kuniyuki Iwashima Date: Mon, 21 Oct 2024 11:32:36 -0700 Subject: rtnetlink: Define rtnl_net_trylock(). We will need the per-netns version of rtnl_trylock(). rtnl_net_trylock() calls __rtnl_net_lock() only when rtnl_trylock() successfully holds RTNL. When RTNL is removed, we will use mutex_trylock() for per-netns RTNL. Signed-off-by: Kuniyuki Iwashima Reviewed-by: Eric Dumazet Signed-off-by: Paolo Abeni --- include/linux/rtnetlink.h | 6 ++++++ net/core/rtnetlink.c | 11 +++++++++++ 2 files changed, 17 insertions(+) (limited to 'include/linux') diff --git a/include/linux/rtnetlink.h b/include/linux/rtnetlink.h index 0e62918de63b..14b88f551920 100644 --- a/include/linux/rtnetlink.h +++ b/include/linux/rtnetlink.h @@ -101,6 +101,7 @@ void __rtnl_net_lock(struct net *net); void __rtnl_net_unlock(struct net *net); void rtnl_net_lock(struct net *net); void rtnl_net_unlock(struct net *net); +int rtnl_net_trylock(struct net *net); int rtnl_net_lock_cmp_fn(const struct lockdep_map *a, const struct lockdep_map *b); bool rtnl_net_is_locked(struct net *net); @@ -132,6 +133,11 @@ static inline void rtnl_net_unlock(struct net *net) rtnl_unlock(); } +static inline int rtnl_net_trylock(struct net *net) +{ + return rtnl_trylock(); +} + static inline void ASSERT_RTNL_NET(struct net *net) { ASSERT_RTNL(); diff --git a/net/core/rtnetlink.c b/net/core/rtnetlink.c index 194a81e5f608..dda8230fdfd4 100644 --- a/net/core/rtnetlink.c +++ b/net/core/rtnetlink.c @@ -210,6 +210,17 @@ void rtnl_net_unlock(struct net *net) } EXPORT_SYMBOL(rtnl_net_unlock); +int rtnl_net_trylock(struct net *net) +{ + int ret = rtnl_trylock(); + + if (ret) + __rtnl_net_lock(net); + + return ret; +} +EXPORT_SYMBOL(rtnl_net_trylock); + static int rtnl_net_cmp_locks(const struct net *net_a, const struct net *net_b) { if (net_eq(net_a, net_b)) -- cgit v1.2.3 From f247fd22e9f27d919d81861e4dcc438e0b6d179b Mon Sep 17 00:00:00 2001 From: Sven Schnelle Date: Wed, 23 Oct 2024 08:56:00 +0200 Subject: s390/time: Add clocksource id to TOD clock To allow specifying the clock source in the upcoming PtP driver, add a clocksource ID to the s390 TOD clock. Acked-by: Heiko Carstens Acked-by: Richard Cochran Signed-off-by: Sven Schnelle Link: https://patch.msgid.link/20241023065601.449586-2-svens@linux.ibm.com Signed-off-by: Jakub Kicinski --- arch/s390/kernel/time.c | 1 + include/linux/clocksource_ids.h | 1 + 2 files changed, 2 insertions(+) (limited to 'include/linux') diff --git a/arch/s390/kernel/time.c b/arch/s390/kernel/time.c index b713effe0579..4214901c3ab0 100644 --- a/arch/s390/kernel/time.c +++ b/arch/s390/kernel/time.c @@ -255,6 +255,7 @@ static struct clocksource clocksource_tod = { .shift = 24, .flags = CLOCK_SOURCE_IS_CONTINUOUS, .vdso_clock_mode = VDSO_CLOCKMODE_TOD, + .id = CSID_S390_TOD, }; struct clocksource * __init clocksource_default_clock(void) diff --git a/include/linux/clocksource_ids.h b/include/linux/clocksource_ids.h index 2bb4d8c2f1b0..c4ef4ae2eded 100644 --- a/include/linux/clocksource_ids.h +++ b/include/linux/clocksource_ids.h @@ -6,6 +6,7 @@ enum clocksource_ids { CSID_GENERIC = 0, CSID_ARM_ARCH_COUNTER, + CSID_S390_TOD, CSID_X86_TSC_EARLY, CSID_X86_TSC, CSID_X86_KVM_CLK, -- cgit v1.2.3 From a911bad094b010e276f072fe9a599b66e59ed5fe Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Tue, 29 Oct 2024 19:14:25 +0000 Subject: dql: annotate data-races around dql->last_obj_cnt dql->last_obj_cnt is read/written from different contexts, without any lock synchronization. Use READ_ONCE()/WRITE_ONCE() to avoid load/store tearing. Signed-off-by: Eric Dumazet Reviewed-by: Joe Damato Link: https://patch.msgid.link/20241029191425.2519085-1-edumazet@google.com Signed-off-by: Jakub Kicinski --- include/linux/dynamic_queue_limits.h | 2 +- lib/dynamic_queue_limits.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) (limited to 'include/linux') diff --git a/include/linux/dynamic_queue_limits.h b/include/linux/dynamic_queue_limits.h index 281298e77a15..808b1a5102e7 100644 --- a/include/linux/dynamic_queue_limits.h +++ b/include/linux/dynamic_queue_limits.h @@ -127,7 +127,7 @@ static inline void dql_queued(struct dql *dql, unsigned int count) if (WARN_ON_ONCE(count > DQL_MAX_OBJECT)) return; - dql->last_obj_cnt = count; + WRITE_ONCE(dql->last_obj_cnt, count); /* We want to force a write first, so that cpu do not attempt * to get cache line containing last_obj_cnt, num_queued, adj_limit diff --git a/lib/dynamic_queue_limits.c b/lib/dynamic_queue_limits.c index e49deddd3de9..c1b7638a594a 100644 --- a/lib/dynamic_queue_limits.c +++ b/lib/dynamic_queue_limits.c @@ -179,7 +179,7 @@ void dql_completed(struct dql *dql, unsigned int count) dql->adj_limit = limit + completed; dql->prev_ovlimit = ovlimit; - dql->prev_last_obj_cnt = dql->last_obj_cnt; + dql->prev_last_obj_cnt = READ_ONCE(dql->last_obj_cnt); dql->num_completed = completed; dql->prev_num_queued = num_queued; -- cgit v1.2.3 From a1afb959add1fad43cb337448c244ed70bac3109 Mon Sep 17 00:00:00 2001 From: Jiri Pirko Date: Wed, 30 Oct 2024 09:11:56 +0100 Subject: dpll: add clock quality level attribute and op In order to allow driver expose quality level of the clock it is running, introduce a new netlink attr with enum to carry it to the userspace. Also, introduce an op the dpll netlink code calls into the driver to obtain the value. Signed-off-by: Jiri Pirko Link: https://patch.msgid.link/20241030081157.966604-2-jiri@resnulli.us Signed-off-by: Jakub Kicinski --- Documentation/netlink/specs/dpll.yaml | 41 +++++++++++++++++++++++++++++++++++ drivers/dpll/dpll_netlink.c | 24 ++++++++++++++++++++ include/linux/dpll.h | 4 ++++ include/uapi/linux/dpll.h | 24 ++++++++++++++++++++ 4 files changed, 93 insertions(+) (limited to 'include/linux') diff --git a/Documentation/netlink/specs/dpll.yaml b/Documentation/netlink/specs/dpll.yaml index f2894ca35de8..8feefeae5376 100644 --- a/Documentation/netlink/specs/dpll.yaml +++ b/Documentation/netlink/specs/dpll.yaml @@ -85,6 +85,36 @@ definitions: This may happen for example if dpll device was previously locked on an input pin of type PIN_TYPE_SYNCE_ETH_PORT. render-max: true + - + type: enum + name: clock-quality-level + doc: | + level of quality of a clock device. This mainly applies when + the dpll lock-status is DPLL_LOCK_STATUS_HOLDOVER. + The current list is defined according to the table 11-7 contained + in ITU-T G.8264/Y.1364 document. One may extend this list freely + by other ITU-T defined clock qualities, or different ones defined + by another standardization body (for those, please use + different prefix). + entries: + - + name: itu-opt1-prc + value: 1 + - + name: itu-opt1-ssu-a + - + name: itu-opt1-ssu-b + - + name: itu-opt1-eec1 + - + name: itu-opt1-prtc + - + name: itu-opt1-eprtc + - + name: itu-opt1-eeec + - + name: itu-opt1-eprc + render-max: true - type: const name: temp-divider @@ -252,6 +282,17 @@ attribute-sets: name: lock-status-error type: u32 enum: lock-status-error + - + name: clock-quality-level + type: u32 + enum: clock-quality-level + multi-attr: true + doc: | + Level of quality of a clock device. This mainly applies when + the dpll lock-status is DPLL_LOCK_STATUS_HOLDOVER. This could + be put to message multiple times to indicate possible parallel + quality levels (e.g. one specified by ITU option 1 and another + one specified by option 2). - name: pin enum-name: dpll_a_pin diff --git a/drivers/dpll/dpll_netlink.c b/drivers/dpll/dpll_netlink.c index fc0280dcddd1..c130f87147fa 100644 --- a/drivers/dpll/dpll_netlink.c +++ b/drivers/dpll/dpll_netlink.c @@ -169,6 +169,27 @@ dpll_msg_add_temp(struct sk_buff *msg, struct dpll_device *dpll, return 0; } +static int +dpll_msg_add_clock_quality_level(struct sk_buff *msg, struct dpll_device *dpll, + struct netlink_ext_ack *extack) +{ + const struct dpll_device_ops *ops = dpll_device_ops(dpll); + DECLARE_BITMAP(qls, DPLL_CLOCK_QUALITY_LEVEL_MAX) = { 0 }; + enum dpll_clock_quality_level ql; + int ret; + + if (!ops->clock_quality_level_get) + return 0; + ret = ops->clock_quality_level_get(dpll, dpll_priv(dpll), qls, extack); + if (ret) + return ret; + for_each_set_bit(ql, qls, DPLL_CLOCK_QUALITY_LEVEL_MAX) + if (nla_put_u32(msg, DPLL_A_CLOCK_QUALITY_LEVEL, ql)) + return -EMSGSIZE; + + return 0; +} + static int dpll_msg_add_pin_prio(struct sk_buff *msg, struct dpll_pin *pin, struct dpll_pin_ref *ref, @@ -557,6 +578,9 @@ dpll_device_get_one(struct dpll_device *dpll, struct sk_buff *msg, if (ret) return ret; ret = dpll_msg_add_lock_status(msg, dpll, extack); + if (ret) + return ret; + ret = dpll_msg_add_clock_quality_level(msg, dpll, extack); if (ret) return ret; ret = dpll_msg_add_mode(msg, dpll, extack); diff --git a/include/linux/dpll.h b/include/linux/dpll.h index 81f7b623d0ba..5e4f9ab1cf75 100644 --- a/include/linux/dpll.h +++ b/include/linux/dpll.h @@ -26,6 +26,10 @@ struct dpll_device_ops { struct netlink_ext_ack *extack); int (*temp_get)(const struct dpll_device *dpll, void *dpll_priv, s32 *temp, struct netlink_ext_ack *extack); + int (*clock_quality_level_get)(const struct dpll_device *dpll, + void *dpll_priv, + unsigned long *qls, + struct netlink_ext_ack *extack); }; struct dpll_pin_ops { diff --git a/include/uapi/linux/dpll.h b/include/uapi/linux/dpll.h index b0654ade7b7e..2b7ec2da4bcc 100644 --- a/include/uapi/linux/dpll.h +++ b/include/uapi/linux/dpll.h @@ -79,6 +79,29 @@ enum dpll_lock_status_error { DPLL_LOCK_STATUS_ERROR_MAX = (__DPLL_LOCK_STATUS_ERROR_MAX - 1) }; +/** + * enum dpll_clock_quality_level - level of quality of a clock device. This + * mainly applies when the dpll lock-status is DPLL_LOCK_STATUS_HOLDOVER. The + * current list is defined according to the table 11-7 contained in ITU-T + * G.8264/Y.1364 document. One may extend this list freely by other ITU-T + * defined clock qualities, or different ones defined by another + * standardization body (for those, please use different prefix). + */ +enum dpll_clock_quality_level { + DPLL_CLOCK_QUALITY_LEVEL_ITU_OPT1_PRC = 1, + DPLL_CLOCK_QUALITY_LEVEL_ITU_OPT1_SSU_A, + DPLL_CLOCK_QUALITY_LEVEL_ITU_OPT1_SSU_B, + DPLL_CLOCK_QUALITY_LEVEL_ITU_OPT1_EEC1, + DPLL_CLOCK_QUALITY_LEVEL_ITU_OPT1_PRTC, + DPLL_CLOCK_QUALITY_LEVEL_ITU_OPT1_EPRTC, + DPLL_CLOCK_QUALITY_LEVEL_ITU_OPT1_EEEC, + DPLL_CLOCK_QUALITY_LEVEL_ITU_OPT1_EPRC, + + /* private: */ + __DPLL_CLOCK_QUALITY_LEVEL_MAX, + DPLL_CLOCK_QUALITY_LEVEL_MAX = (__DPLL_CLOCK_QUALITY_LEVEL_MAX - 1) +}; + #define DPLL_TEMP_DIVIDER 1000 /** @@ -180,6 +203,7 @@ enum dpll_a { DPLL_A_TEMP, DPLL_A_TYPE, DPLL_A_LOCK_STATUS_ERROR, + DPLL_A_CLOCK_QUALITY_LEVEL, __DPLL_A_MAX, DPLL_A_MAX = (__DPLL_A_MAX - 1) -- cgit v1.2.3 From a8f80673ca0daaad990882529a5b4dc5114071e7 Mon Sep 17 00:00:00 2001 From: Yafang Shao Date: Thu, 24 Oct 2024 17:37:41 +0800 Subject: compiler_types: Add noinline_for_tracing annotation Kernel functions that are not inlined can be easily hooked with BPF for tracing. However, functions intended for tracing may still be inlined unexpectedly. For example, in our case, after upgrading the compiler from GCC 9 to GCC 11, the tcp_drop_reason() function was inlined, which broke our monitoring tools. To prevent this, we need to ensure that the function remains non-inlined. The noinline_for_tracing annotation is introduced as a general solution for preventing inlining of kernel functions that need to be traced. This approach avoids the need for adding individual noinline comments to each function and provides a more consistent way to maintain traceability. Link: https://lore.kernel.org/netdev/CANn89iKvr44ipuRYFaPTpzwz=B_+pgA94jsggQ946mjwreV6Aw@mail.gmail.com/ Suggested-by: Eric Dumazet Signed-off-by: Yafang Shao Link: https://patch.msgid.link/20241024093742.87681-2-laoar.shao@gmail.com Signed-off-by: Jakub Kicinski --- include/linux/compiler_types.h | 6 ++++++ 1 file changed, 6 insertions(+) (limited to 'include/linux') diff --git a/include/linux/compiler_types.h b/include/linux/compiler_types.h index 1a957ea2f4fe..0c8b9601e603 100644 --- a/include/linux/compiler_types.h +++ b/include/linux/compiler_types.h @@ -265,6 +265,12 @@ struct ftrace_likely_data { */ #define noinline_for_stack noinline +/* + * Use noinline_for_tracing for functions that should not be inlined. + * For tracing reasons. + */ +#define noinline_for_tracing noinline + /* * Sanitizer helper attributes: Because using __always_inline and * __no_sanitize_* conflict, provide helper attributes that will either expand -- cgit v1.2.3 From 3bd9b9abdf1563a22041b7255baea6d449902f1a Mon Sep 17 00:00:00 2001 From: "Gustavo A. R. Silva" Date: Tue, 29 Oct 2024 15:58:47 -0600 Subject: net: ethtool: Avoid thousands of -Wflex-array-member-not-at-end warnings -Wflex-array-member-not-at-end was introduced in GCC-14, and we are getting ready to enable it, globally. Change the type of the middle struct member currently causing trouble from `struct ethtool_link_settings` to `struct ethtool_link_settings_hdr`. Additionally, update the type of some variables in various functions that don't access the flexible-array member, changing them to the newly created `struct ethtool_link_settings_hdr`. These changes are needed because the type of the conflicting middle members changed. So, those instances that expect the type to be `struct ethtool_link_settings` should be adjusted to the newly created type `struct ethtool_link_settings_hdr`. Also, adjust variable declarations to follow the reverse xmas tree convention. Fix 3338 of the following -Wflex-array-member-not-at-end warnings: include/linux/ethtool.h:214:38: warning: structure containing a flexible array member is not at the end of another structure [-Wflex-array-member-not-at-end] Signed-off-by: Gustavo A. R. Silva Link: https://patch.msgid.link/0bc2809fe2a6c11dd4c8a9a10d9bd65cccdb559b.1730238285.git.gustavoars@kernel.org Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c | 6 +++--- drivers/net/ethernet/chelsio/cxgb4/cxgb4_ethtool.c | 4 ++-- drivers/net/ethernet/chelsio/cxgb4vf/cxgb4vf_main.c | 2 +- drivers/net/ethernet/cisco/enic/enic_ethtool.c | 2 +- drivers/net/ethernet/qlogic/qede/qede_ethtool.c | 4 ++-- include/linux/ethtool.h | 2 +- net/ethtool/ioctl.c | 2 +- net/ethtool/linkinfo.c | 8 ++++---- net/ethtool/linkmodes.c | 18 +++++++++++------- 9 files changed, 26 insertions(+), 22 deletions(-) (limited to 'include/linux') diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c b/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c index f71cc8188b4e..e0ebe69110bf 100644 --- a/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c @@ -2781,7 +2781,7 @@ u32 bnxt_fw_to_ethtool_speed(u16 fw_link_speed) static void bnxt_get_default_speeds(struct ethtool_link_ksettings *lk_ksettings, struct bnxt_link_info *link_info) { - struct ethtool_link_settings *base = &lk_ksettings->base; + struct ethtool_link_settings_hdr *base = &lk_ksettings->base; if (link_info->link_state == BNXT_LINK_STATE_UP) { base->speed = bnxt_fw_to_ethtool_speed(link_info->link_speed); @@ -2800,7 +2800,7 @@ static void bnxt_get_default_speeds(struct ethtool_link_ksettings *lk_ksettings, static int bnxt_get_link_ksettings(struct net_device *dev, struct ethtool_link_ksettings *lk_ksettings) { - struct ethtool_link_settings *base = &lk_ksettings->base; + struct ethtool_link_settings_hdr *base = &lk_ksettings->base; enum ethtool_link_mode_bit_indices link_mode; struct bnxt *bp = netdev_priv(dev); struct bnxt_link_info *link_info; @@ -3023,9 +3023,9 @@ u16 bnxt_get_fw_auto_link_speeds(const unsigned long *mode) static int bnxt_set_link_ksettings(struct net_device *dev, const struct ethtool_link_ksettings *lk_ksettings) { + const struct ethtool_link_settings_hdr *base = &lk_ksettings->base; struct bnxt *bp = netdev_priv(dev); struct bnxt_link_info *link_info = &bp->link_info; - const struct ethtool_link_settings *base = &lk_ksettings->base; bool set_pause = false; u32 speed, lanes = 0; int rc = 0; diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_ethtool.c b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_ethtool.c index 7f3f5afa864f..45d28a65347e 100644 --- a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_ethtool.c +++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_ethtool.c @@ -662,8 +662,8 @@ static unsigned int lmm_to_fw_caps(const unsigned long *link_mode_mask) static int get_link_ksettings(struct net_device *dev, struct ethtool_link_ksettings *link_ksettings) { + struct ethtool_link_settings_hdr *base = &link_ksettings->base; struct port_info *pi = netdev_priv(dev); - struct ethtool_link_settings *base = &link_ksettings->base; /* For the nonce, the Firmware doesn't send up Port State changes * when the Virtual Interface attached to the Port is down. So @@ -717,9 +717,9 @@ static int get_link_ksettings(struct net_device *dev, static int set_link_ksettings(struct net_device *dev, const struct ethtool_link_ksettings *link_ksettings) { + const struct ethtool_link_settings_hdr *base = &link_ksettings->base; struct port_info *pi = netdev_priv(dev); struct link_config *lc = &pi->link_cfg; - const struct ethtool_link_settings *base = &link_ksettings->base; struct link_config old_lc; unsigned int fw_caps; int ret = 0; diff --git a/drivers/net/ethernet/chelsio/cxgb4vf/cxgb4vf_main.c b/drivers/net/ethernet/chelsio/cxgb4vf/cxgb4vf_main.c index 2fbe0f059a0b..61d08547e3f9 100644 --- a/drivers/net/ethernet/chelsio/cxgb4vf/cxgb4vf_main.c +++ b/drivers/net/ethernet/chelsio/cxgb4vf/cxgb4vf_main.c @@ -1436,8 +1436,8 @@ static void fw_caps_to_lmm(enum fw_port_type port_type, static int cxgb4vf_get_link_ksettings(struct net_device *dev, struct ethtool_link_ksettings *link_ksettings) { + struct ethtool_link_settings_hdr *base = &link_ksettings->base; struct port_info *pi = netdev_priv(dev); - struct ethtool_link_settings *base = &link_ksettings->base; /* For the nonce, the Firmware doesn't send up Port State changes * when the Virtual Interface attached to the Port is down. So diff --git a/drivers/net/ethernet/cisco/enic/enic_ethtool.c b/drivers/net/ethernet/cisco/enic/enic_ethtool.c index f7986f2b6a17..4fe85780a950 100644 --- a/drivers/net/ethernet/cisco/enic/enic_ethtool.c +++ b/drivers/net/ethernet/cisco/enic/enic_ethtool.c @@ -129,8 +129,8 @@ static void enic_intr_coal_set_rx(struct enic *enic, u32 timer) static int enic_get_ksettings(struct net_device *netdev, struct ethtool_link_ksettings *ecmd) { + struct ethtool_link_settings_hdr *base = &ecmd->base; struct enic *enic = netdev_priv(netdev); - struct ethtool_link_settings *base = &ecmd->base; ethtool_link_ksettings_add_link_mode(ecmd, supported, 10000baseT_Full); diff --git a/drivers/net/ethernet/qlogic/qede/qede_ethtool.c b/drivers/net/ethernet/qlogic/qede/qede_ethtool.c index e50e1df0a433..c553da16d4b1 100644 --- a/drivers/net/ethernet/qlogic/qede/qede_ethtool.c +++ b/drivers/net/ethernet/qlogic/qede/qede_ethtool.c @@ -504,7 +504,7 @@ static int qede_get_link_ksettings(struct net_device *dev, struct ethtool_link_ksettings *cmd) { typeof(cmd->link_modes) *link_modes = &cmd->link_modes; - struct ethtool_link_settings *base = &cmd->base; + struct ethtool_link_settings_hdr *base = &cmd->base; struct qede_dev *edev = netdev_priv(dev); struct qed_link_output current_link; @@ -537,7 +537,7 @@ static int qede_get_link_ksettings(struct net_device *dev, static int qede_set_link_ksettings(struct net_device *dev, const struct ethtool_link_ksettings *cmd) { - const struct ethtool_link_settings *base = &cmd->base; + const struct ethtool_link_settings_hdr *base = &cmd->base; const struct ethtool_forced_speed_map *map; struct qede_dev *edev = netdev_priv(dev); struct qed_link_output current_link; diff --git a/include/linux/ethtool.h b/include/linux/ethtool.h index 12f6dc567598..1199e308c8dd 100644 --- a/include/linux/ethtool.h +++ b/include/linux/ethtool.h @@ -211,7 +211,7 @@ void ethtool_rxfh_context_lost(struct net_device *dev, u32 context_id); * fields, but they are allowed to overwrite them (will be ignored). */ struct ethtool_link_ksettings { - struct ethtool_link_settings base; + struct ethtool_link_settings_hdr base; struct { __ETHTOOL_DECLARE_LINK_MODE_MASK(supported); __ETHTOOL_DECLARE_LINK_MODE_MASK(advertising); diff --git a/net/ethtool/ioctl.c b/net/ethtool/ioctl.c index 5cc131cdb1bc..7da94e26ced6 100644 --- a/net/ethtool/ioctl.c +++ b/net/ethtool/ioctl.c @@ -425,7 +425,7 @@ convert_link_ksettings_to_legacy_settings( /* layout of the struct passed from/to userland */ struct ethtool_link_usettings { - struct ethtool_link_settings base; + struct ethtool_link_settings_hdr base; struct { __u32 supported[__ETHTOOL_LINK_MODE_MASK_NU32]; __u32 advertising[__ETHTOOL_LINK_MODE_MASK_NU32]; diff --git a/net/ethtool/linkinfo.c b/net/ethtool/linkinfo.c index 30b8ce275159..2d5bc57160be 100644 --- a/net/ethtool/linkinfo.c +++ b/net/ethtool/linkinfo.c @@ -8,9 +8,9 @@ struct linkinfo_req_info { }; struct linkinfo_reply_data { - struct ethnl_reply_data base; - struct ethtool_link_ksettings ksettings; - struct ethtool_link_settings *lsettings; + struct ethnl_reply_data base; + struct ethtool_link_ksettings ksettings; + struct ethtool_link_settings_hdr *lsettings; }; #define LINKINFO_REPDATA(__reply_base) \ @@ -98,7 +98,7 @@ static int ethnl_set_linkinfo(struct ethnl_req_info *req_info, struct genl_info *info) { struct ethtool_link_ksettings ksettings = {}; - struct ethtool_link_settings *lsettings; + struct ethtool_link_settings_hdr *lsettings; struct net_device *dev = req_info->dev; struct nlattr **tb = info->attrs; bool mod = false; diff --git a/net/ethtool/linkmodes.c b/net/ethtool/linkmodes.c index 259cd9ef1f2a..17e49cf89f03 100644 --- a/net/ethtool/linkmodes.c +++ b/net/ethtool/linkmodes.c @@ -11,10 +11,10 @@ struct linkmodes_req_info { }; struct linkmodes_reply_data { - struct ethnl_reply_data base; - struct ethtool_link_ksettings ksettings; - struct ethtool_link_settings *lsettings; - bool peer_empty; + struct ethnl_reply_data base; + struct ethtool_link_ksettings ksettings; + struct ethtool_link_settings_hdr *lsettings; + bool peer_empty; }; #define LINKMODES_REPDATA(__reply_base) \ @@ -62,10 +62,12 @@ static int linkmodes_reply_size(const struct ethnl_req_info *req_base, { const struct linkmodes_reply_data *data = LINKMODES_REPDATA(reply_base); const struct ethtool_link_ksettings *ksettings = &data->ksettings; - const struct ethtool_link_settings *lsettings = &ksettings->base; bool compact = req_base->flags & ETHTOOL_FLAG_COMPACT_BITSETS; + const struct ethtool_link_settings_hdr *lsettings; int len, ret; + lsettings = &ksettings->base; + len = nla_total_size(sizeof(u8)) /* LINKMODES_AUTONEG */ + nla_total_size(sizeof(u32)) /* LINKMODES_SPEED */ + nla_total_size(sizeof(u32)) /* LINKMODES_LANES */ @@ -103,10 +105,12 @@ static int linkmodes_fill_reply(struct sk_buff *skb, { const struct linkmodes_reply_data *data = LINKMODES_REPDATA(reply_base); const struct ethtool_link_ksettings *ksettings = &data->ksettings; - const struct ethtool_link_settings *lsettings = &ksettings->base; bool compact = req_base->flags & ETHTOOL_FLAG_COMPACT_BITSETS; + const struct ethtool_link_settings_hdr *lsettings; int ret; + lsettings = &ksettings->base; + if (nla_put_u8(skb, ETHTOOL_A_LINKMODES_AUTONEG, lsettings->autoneg)) return -EMSGSIZE; @@ -237,7 +241,7 @@ static int ethnl_update_linkmodes(struct genl_info *info, struct nlattr **tb, struct ethtool_link_ksettings *ksettings, bool *mod, const struct net_device *dev) { - struct ethtool_link_settings *lsettings = &ksettings->base; + struct ethtool_link_settings_hdr *lsettings = &ksettings->base; bool req_speed, req_lanes, req_duplex; const struct nlattr *master_slave_cfg, *lanes_cfg; int ret; -- cgit v1.2.3 From a865276872ec4f129f8a582634be82dcc275dc2a Mon Sep 17 00:00:00 2001 From: Caleb Sander Mateos Date: Wed, 30 Oct 2024 18:23:25 -0600 Subject: dim: make dim_calc_stats() inputs const pointers Make the start and end arguments to dim_calc_stats() const pointers to clarify that the function does not modify their values. Signed-off-by: Caleb Sander Mateos Reviewed-by: Vladimir Oltean Reviewed-by: Florian Fainelli Reviewed-by: Arthur Kiyanovski Link: https://patch.msgid.link/20241031002326.3426181-1-csander@purestorage.com Signed-off-by: Jakub Kicinski --- include/linux/dim.h | 3 ++- lib/dim/dim.c | 3 ++- 2 files changed, 4 insertions(+), 2 deletions(-) (limited to 'include/linux') diff --git a/include/linux/dim.h b/include/linux/dim.h index 1b581ff25a15..84579a50ae7f 100644 --- a/include/linux/dim.h +++ b/include/linux/dim.h @@ -351,7 +351,8 @@ void dim_park_tired(struct dim *dim); * Takes into consideration counter wrap-around. * Returned boolean indicates whether curr_stats are reliable. */ -bool dim_calc_stats(struct dim_sample *start, struct dim_sample *end, +bool dim_calc_stats(const struct dim_sample *start, + const struct dim_sample *end, struct dim_stats *curr_stats); /** diff --git a/lib/dim/dim.c b/lib/dim/dim.c index 83b65ac74d73..97c3d084ebf0 100644 --- a/lib/dim/dim.c +++ b/lib/dim/dim.c @@ -54,7 +54,8 @@ void dim_park_tired(struct dim *dim) } EXPORT_SYMBOL(dim_park_tired); -bool dim_calc_stats(struct dim_sample *start, struct dim_sample *end, +bool dim_calc_stats(const struct dim_sample *start, + const struct dim_sample *end, struct dim_stats *curr_stats) { /* u32 holds up to 71 minutes, should be enough */ -- cgit v1.2.3 From 61bf0009a7657d394d942c8ee961b9ea5f2168fe Mon Sep 17 00:00:00 2001 From: Caleb Sander Mateos Date: Wed, 30 Oct 2024 18:23:26 -0600 Subject: dim: pass dim_sample to net_dim() by reference net_dim() is currently passed a struct dim_sample argument by value. struct dim_sample is 24 bytes. Since this is greater 16 bytes, x86-64 passes it on the stack. All callers have already initialized dim_sample on the stack, so passing it by value requires pushing a duplicated copy to the stack. Either witing to the stack and immediately reading it, or perhaps dereferencing addresses relative to the stack pointer in a chain of push instructions, seems to perform quite poorly. In a heavy TCP workload, mlx5e_handle_rx_dim() consumes 3% of CPU time, 94% of which is attributed to the first push instruction to copy dim_sample on the stack for the call to net_dim(): // Call ktime_get() 0.26 |4ead2: call 4ead7 // Pass the address of struct dim in %rdi |4ead7: lea 0x3d0(%rbx),%rdi // Set dim_sample.pkt_ctr |4eade: mov %r13d,0x8(%rsp) // Set dim_sample.byte_ctr |4eae3: mov %r12d,0xc(%rsp) // Set dim_sample.event_ctr 0.15 |4eae8: mov %bp,0x10(%rsp) // Duplicate dim_sample on the stack 94.16 |4eaed: push 0x10(%rsp) 2.79 |4eaf1: push 0x10(%rsp) 0.07 |4eaf5: push %rax // Call net_dim() 0.21 |4eaf6: call 4eafb To allow the caller to reuse the struct dim_sample already on the stack, pass the struct dim_sample by reference to net_dim(). Signed-off-by: Caleb Sander Mateos Reviewed-by: Vladimir Oltean Reviewed-by: Shannon Nelson Reviewed-by: Florian Fainelli Reviewed-by: Arthur Kiyanovski Reviewed-by: Louis Peens Link: https://patch.msgid.link/20241031002326.3426181-2-csander@purestorage.com Signed-off-by: Jakub Kicinski --- Documentation/networking/net_dim.rst | 2 +- drivers/net/ethernet/amazon/ena/ena_netdev.c | 2 +- drivers/net/ethernet/broadcom/bcmsysport.c | 2 +- drivers/net/ethernet/broadcom/bnxt/bnxt.c | 4 ++-- drivers/net/ethernet/broadcom/genet/bcmgenet.c | 2 +- drivers/net/ethernet/freescale/enetc/enetc.c | 2 +- drivers/net/ethernet/hisilicon/hns3/hns3_enet.c | 4 ++-- drivers/net/ethernet/intel/ice/ice_txrx.c | 4 ++-- drivers/net/ethernet/intel/idpf/idpf_txrx.c | 4 ++-- drivers/net/ethernet/marvell/octeontx2/nic/otx2_txrx.c | 2 +- drivers/net/ethernet/mediatek/mtk_eth_soc.c | 4 ++-- drivers/net/ethernet/mellanox/mlx5/core/en_txrx.c | 4 ++-- drivers/net/ethernet/netronome/nfp/nfd3/dp.c | 4 ++-- drivers/net/ethernet/netronome/nfp/nfdk/dp.c | 4 ++-- drivers/net/ethernet/pensando/ionic/ionic_txrx.c | 2 +- drivers/net/virtio_net.c | 2 +- drivers/soc/fsl/dpio/dpio-service.c | 2 +- include/linux/dim.h | 2 +- lib/dim/net_dim.c | 10 +++++----- 19 files changed, 31 insertions(+), 31 deletions(-) (limited to 'include/linux') diff --git a/Documentation/networking/net_dim.rst b/Documentation/networking/net_dim.rst index 8908fd7b0a8d..4377998e6826 100644 --- a/Documentation/networking/net_dim.rst +++ b/Documentation/networking/net_dim.rst @@ -156,7 +156,7 @@ usage is not complete but it should make the outline of the usage clear. my_entity->bytes, &dim_sample); /* Call net DIM */ - net_dim(&my_entity->dim, dim_sample); + net_dim(&my_entity->dim, &dim_sample); ... } diff --git a/drivers/net/ethernet/amazon/ena/ena_netdev.c b/drivers/net/ethernet/amazon/ena/ena_netdev.c index 96df20854eb9..63c8a2328142 100644 --- a/drivers/net/ethernet/amazon/ena/ena_netdev.c +++ b/drivers/net/ethernet/amazon/ena/ena_netdev.c @@ -1383,7 +1383,7 @@ static void ena_adjust_adaptive_rx_intr_moderation(struct ena_napi *ena_napi) rx_ring->rx_stats.bytes, &dim_sample); - net_dim(&ena_napi->dim, dim_sample); + net_dim(&ena_napi->dim, &dim_sample); rx_ring->per_napi_packets = 0; } diff --git a/drivers/net/ethernet/broadcom/bcmsysport.c b/drivers/net/ethernet/broadcom/bcmsysport.c index caff6e87a488..031e9e0cca53 100644 --- a/drivers/net/ethernet/broadcom/bcmsysport.c +++ b/drivers/net/ethernet/broadcom/bcmsysport.c @@ -1029,7 +1029,7 @@ static int bcm_sysport_poll(struct napi_struct *napi, int budget) if (priv->dim.use_dim) { dim_update_sample(priv->dim.event_ctr, priv->dim.packets, priv->dim.bytes, &dim_sample); - net_dim(&priv->dim.dim, dim_sample); + net_dim(&priv->dim.dim, &dim_sample); } return work_done; diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.c b/drivers/net/ethernet/broadcom/bnxt/bnxt.c index 6dd6541d8619..ca42b81133d7 100644 --- a/drivers/net/ethernet/broadcom/bnxt/bnxt.c +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.c @@ -3102,7 +3102,7 @@ static int bnxt_poll(struct napi_struct *napi, int budget) cpr->rx_packets, cpr->rx_bytes, &dim_sample); - net_dim(&cpr->dim, dim_sample); + net_dim(&cpr->dim, &dim_sample); } return work_done; } @@ -3233,7 +3233,7 @@ poll_done: cpr_rx->rx_packets, cpr_rx->rx_bytes, &dim_sample); - net_dim(&cpr->dim, dim_sample); + net_dim(&cpr->dim, &dim_sample); } return work_done; } diff --git a/drivers/net/ethernet/broadcom/genet/bcmgenet.c b/drivers/net/ethernet/broadcom/genet/bcmgenet.c index 10966ab15373..53a949eb9180 100644 --- a/drivers/net/ethernet/broadcom/genet/bcmgenet.c +++ b/drivers/net/ethernet/broadcom/genet/bcmgenet.c @@ -2405,7 +2405,7 @@ static int bcmgenet_rx_poll(struct napi_struct *napi, int budget) if (ring->dim.use_dim) { dim_update_sample(ring->dim.event_ctr, ring->dim.packets, ring->dim.bytes, &dim_sample); - net_dim(&ring->dim.dim, dim_sample); + net_dim(&ring->dim.dim, &dim_sample); } return work_done; diff --git a/drivers/net/ethernet/freescale/enetc/enetc.c b/drivers/net/ethernet/freescale/enetc/enetc.c index c09370eab319..05dedea6185a 100644 --- a/drivers/net/ethernet/freescale/enetc/enetc.c +++ b/drivers/net/ethernet/freescale/enetc/enetc.c @@ -718,7 +718,7 @@ static void enetc_rx_net_dim(struct enetc_int_vector *v) v->rx_ring.stats.packets, v->rx_ring.stats.bytes, &dim_sample); - net_dim(&v->rx_dim, dim_sample); + net_dim(&v->rx_dim, &dim_sample); } static int enetc_bd_ready_count(struct enetc_bdr *tx_ring, int ci) diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c index b09f0cca34dc..fbfd3ee5648f 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c +++ b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c @@ -4478,7 +4478,7 @@ static void hns3_update_rx_int_coalesce(struct hns3_enet_tqp_vector *tqp_vector) dim_update_sample(tqp_vector->event_cnt, rx_group->total_packets, rx_group->total_bytes, &sample); - net_dim(&rx_group->dim, sample); + net_dim(&rx_group->dim, &sample); } static void hns3_update_tx_int_coalesce(struct hns3_enet_tqp_vector *tqp_vector) @@ -4491,7 +4491,7 @@ static void hns3_update_tx_int_coalesce(struct hns3_enet_tqp_vector *tqp_vector) dim_update_sample(tqp_vector->event_cnt, tx_group->total_packets, tx_group->total_bytes, &sample); - net_dim(&tx_group->dim, sample); + net_dim(&tx_group->dim, &sample); } static int hns3_nic_common_poll(struct napi_struct *napi, int budget) diff --git a/drivers/net/ethernet/intel/ice/ice_txrx.c b/drivers/net/ethernet/intel/ice/ice_txrx.c index 8208055d6e7f..5d2d7736fd5f 100644 --- a/drivers/net/ethernet/intel/ice/ice_txrx.c +++ b/drivers/net/ethernet/intel/ice/ice_txrx.c @@ -1352,14 +1352,14 @@ static void ice_net_dim(struct ice_q_vector *q_vector) struct dim_sample dim_sample; __ice_update_sample(q_vector, tx, &dim_sample, true); - net_dim(&tx->dim, dim_sample); + net_dim(&tx->dim, &dim_sample); } if (ITR_IS_DYNAMIC(rx)) { struct dim_sample dim_sample; __ice_update_sample(q_vector, rx, &dim_sample, false); - net_dim(&rx->dim, dim_sample); + net_dim(&rx->dim, &dim_sample); } } diff --git a/drivers/net/ethernet/intel/idpf/idpf_txrx.c b/drivers/net/ethernet/intel/idpf/idpf_txrx.c index d4e6f0e10487..da2a5becf62f 100644 --- a/drivers/net/ethernet/intel/idpf/idpf_txrx.c +++ b/drivers/net/ethernet/intel/idpf/idpf_txrx.c @@ -3679,7 +3679,7 @@ static void idpf_net_dim(struct idpf_q_vector *q_vector) idpf_update_dim_sample(q_vector, &dim_sample, &q_vector->tx_dim, packets, bytes); - net_dim(&q_vector->tx_dim, dim_sample); + net_dim(&q_vector->tx_dim, &dim_sample); check_rx_itr: if (!IDPF_ITR_IS_DYNAMIC(q_vector->rx_intr_mode)) @@ -3698,7 +3698,7 @@ check_rx_itr: idpf_update_dim_sample(q_vector, &dim_sample, &q_vector->rx_dim, packets, bytes); - net_dim(&q_vector->rx_dim, dim_sample); + net_dim(&q_vector->rx_dim, &dim_sample); } /** diff --git a/drivers/net/ethernet/marvell/octeontx2/nic/otx2_txrx.c b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_txrx.c index 933e18ba2fb2..7aaf32e9aa95 100644 --- a/drivers/net/ethernet/marvell/octeontx2/nic/otx2_txrx.c +++ b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_txrx.c @@ -527,7 +527,7 @@ static void otx2_adjust_adaptive_coalese(struct otx2_nic *pfvf, struct otx2_cq_p rx_frames + tx_frames, rx_bytes + tx_bytes, &dim_sample); - net_dim(&cq_poll->dim, dim_sample); + net_dim(&cq_poll->dim, &dim_sample); } int otx2_napi_handler(struct napi_struct *napi, int budget) diff --git a/drivers/net/ethernet/mediatek/mtk_eth_soc.c b/drivers/net/ethernet/mediatek/mtk_eth_soc.c index f01ceee5f02d..53485142938c 100644 --- a/drivers/net/ethernet/mediatek/mtk_eth_soc.c +++ b/drivers/net/ethernet/mediatek/mtk_eth_soc.c @@ -2227,7 +2227,7 @@ rx_done: eth->rx_bytes += bytes; dim_update_sample(eth->rx_events, eth->rx_packets, eth->rx_bytes, &dim_sample); - net_dim(ð->rx_dim, dim_sample); + net_dim(ð->rx_dim, &dim_sample); if (xdp_flush) xdp_do_flush(); @@ -2377,7 +2377,7 @@ static int mtk_poll_tx(struct mtk_eth *eth, int budget) dim_update_sample(eth->tx_events, eth->tx_packets, eth->tx_bytes, &dim_sample); - net_dim(ð->tx_dim, dim_sample); + net_dim(ð->tx_dim, &dim_sample); if (mtk_queue_stopped(eth) && (atomic_read(&ring->free_count) > ring->thresh)) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_txrx.c b/drivers/net/ethernet/mellanox/mlx5/core/en_txrx.c index 5873fde65c2e..417098f0b2bb 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_txrx.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_txrx.c @@ -55,7 +55,7 @@ static void mlx5e_handle_tx_dim(struct mlx5e_txqsq *sq) return; dim_update_sample(sq->cq.event_ctr, stats->packets, stats->bytes, &dim_sample); - net_dim(sq->dim, dim_sample); + net_dim(sq->dim, &dim_sample); } static void mlx5e_handle_rx_dim(struct mlx5e_rq *rq) @@ -67,7 +67,7 @@ static void mlx5e_handle_rx_dim(struct mlx5e_rq *rq) return; dim_update_sample(rq->cq.event_ctr, stats->packets, stats->bytes, &dim_sample); - net_dim(rq->dim, dim_sample); + net_dim(rq->dim, &dim_sample); } void mlx5e_trigger_irq(struct mlx5e_icosq *sq) diff --git a/drivers/net/ethernet/netronome/nfp/nfd3/dp.c b/drivers/net/ethernet/netronome/nfp/nfd3/dp.c index d215efc6cad0..f1c6c47564b1 100644 --- a/drivers/net/ethernet/netronome/nfp/nfd3/dp.c +++ b/drivers/net/ethernet/netronome/nfp/nfd3/dp.c @@ -1179,7 +1179,7 @@ int nfp_nfd3_poll(struct napi_struct *napi, int budget) } while (u64_stats_fetch_retry(&r_vec->rx_sync, start)); dim_update_sample(r_vec->event_ctr, pkts, bytes, &dim_sample); - net_dim(&r_vec->rx_dim, dim_sample); + net_dim(&r_vec->rx_dim, &dim_sample); } if (r_vec->nfp_net->tx_coalesce_adapt_on && r_vec->tx_ring) { @@ -1194,7 +1194,7 @@ int nfp_nfd3_poll(struct napi_struct *napi, int budget) } while (u64_stats_fetch_retry(&r_vec->tx_sync, start)); dim_update_sample(r_vec->event_ctr, pkts, bytes, &dim_sample); - net_dim(&r_vec->tx_dim, dim_sample); + net_dim(&r_vec->tx_dim, &dim_sample); } return pkts_polled; diff --git a/drivers/net/ethernet/netronome/nfp/nfdk/dp.c b/drivers/net/ethernet/netronome/nfp/nfdk/dp.c index dae5af7d1845..ebeb6ab4465c 100644 --- a/drivers/net/ethernet/netronome/nfp/nfdk/dp.c +++ b/drivers/net/ethernet/netronome/nfp/nfdk/dp.c @@ -1289,7 +1289,7 @@ int nfp_nfdk_poll(struct napi_struct *napi, int budget) } while (u64_stats_fetch_retry(&r_vec->rx_sync, start)); dim_update_sample(r_vec->event_ctr, pkts, bytes, &dim_sample); - net_dim(&r_vec->rx_dim, dim_sample); + net_dim(&r_vec->rx_dim, &dim_sample); } if (r_vec->nfp_net->tx_coalesce_adapt_on && r_vec->tx_ring) { @@ -1304,7 +1304,7 @@ int nfp_nfdk_poll(struct napi_struct *napi, int budget) } while (u64_stats_fetch_retry(&r_vec->tx_sync, start)); dim_update_sample(r_vec->event_ctr, pkts, bytes, &dim_sample); - net_dim(&r_vec->tx_dim, dim_sample); + net_dim(&r_vec->tx_dim, &dim_sample); } return pkts_polled; diff --git a/drivers/net/ethernet/pensando/ionic/ionic_txrx.c b/drivers/net/ethernet/pensando/ionic/ionic_txrx.c index 0eeda7e502db..2ac59564ded1 100644 --- a/drivers/net/ethernet/pensando/ionic/ionic_txrx.c +++ b/drivers/net/ethernet/pensando/ionic/ionic_txrx.c @@ -928,7 +928,7 @@ static void ionic_dim_update(struct ionic_qcq *qcq, int napi_mode) dim_update_sample(qcq->cq.bound_intr->rearm_count, pkts, bytes, &dim_sample); - net_dim(&qcq->dim, dim_sample); + net_dim(&qcq->dim, &dim_sample); } int ionic_tx_napi(struct napi_struct *napi, int budget) diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c index 792e9eadbfc3..869586c17ffd 100644 --- a/drivers/net/virtio_net.c +++ b/drivers/net/virtio_net.c @@ -2804,7 +2804,7 @@ static void virtnet_rx_dim_update(struct virtnet_info *vi, struct receive_queue u64_stats_read(&rq->stats.bytes), &cur_sample); - net_dim(&rq->dim, cur_sample); + net_dim(&rq->dim, &cur_sample); rq->packets_in_napi = 0; } diff --git a/drivers/soc/fsl/dpio/dpio-service.c b/drivers/soc/fsl/dpio/dpio-service.c index b811446e0fa5..0b60ed16297c 100644 --- a/drivers/soc/fsl/dpio/dpio-service.c +++ b/drivers/soc/fsl/dpio/dpio-service.c @@ -891,7 +891,7 @@ void dpaa2_io_update_net_dim(struct dpaa2_io *d, __u64 frames, __u64 bytes) d->frames += frames; dim_update_sample(d->event_ctr, d->frames, d->bytes, &dim_sample); - net_dim(&d->rx_dim, dim_sample); + net_dim(&d->rx_dim, &dim_sample); spin_unlock(&d->dim_lock); } diff --git a/include/linux/dim.h b/include/linux/dim.h index 84579a50ae7f..06543fd40fcc 100644 --- a/include/linux/dim.h +++ b/include/linux/dim.h @@ -425,7 +425,7 @@ struct dim_cq_moder net_dim_get_def_tx_moderation(u8 cq_period_mode); * This is the main logic of the algorithm, where data is processed in order * to decide on next required action. */ -void net_dim(struct dim *dim, struct dim_sample end_sample); +void net_dim(struct dim *dim, const struct dim_sample *end_sample); /* RDMA DIM */ diff --git a/lib/dim/net_dim.c b/lib/dim/net_dim.c index d7e7028e9b19..d6aa09a979b3 100644 --- a/lib/dim/net_dim.c +++ b/lib/dim/net_dim.c @@ -347,7 +347,7 @@ static bool net_dim_decision(struct dim_stats *curr_stats, struct dim *dim) return dim->profile_ix != prev_ix; } -void net_dim(struct dim *dim, struct dim_sample end_sample) +void net_dim(struct dim *dim, const struct dim_sample *end_sample) { struct dim_stats curr_stats; u16 nevents; @@ -355,11 +355,11 @@ void net_dim(struct dim *dim, struct dim_sample end_sample) switch (dim->state) { case DIM_MEASURE_IN_PROGRESS: nevents = BIT_GAP(BITS_PER_TYPE(u16), - end_sample.event_ctr, + end_sample->event_ctr, dim->start_sample.event_ctr); if (nevents < DIM_NEVENTS) break; - if (!dim_calc_stats(&dim->start_sample, &end_sample, &curr_stats)) + if (!dim_calc_stats(&dim->start_sample, end_sample, &curr_stats)) break; if (net_dim_decision(&curr_stats, dim)) { dim->state = DIM_APPLY_NEW_PROFILE; @@ -368,8 +368,8 @@ void net_dim(struct dim *dim, struct dim_sample end_sample) } fallthrough; case DIM_START_MEASURE: - dim_update_sample(end_sample.event_ctr, end_sample.pkt_ctr, - end_sample.byte_ctr, &dim->start_sample); + dim_update_sample(end_sample->event_ctr, end_sample->pkt_ctr, + end_sample->byte_ctr, &dim->start_sample); dim->state = DIM_MEASURE_IN_PROGRESS; break; case DIM_APPLY_NEW_PROFILE: -- cgit v1.2.3 From fe5ba6bf91b3e30118c59fe51048cda101de6542 Mon Sep 17 00:00:00 2001 From: Wei Fang Date: Wed, 30 Oct 2024 17:39:15 +0800 Subject: net: enetc: add initial netc-blk-ctrl driver support The netc-blk-ctrl driver is used to configure Integrated Endpoint Register Block (IERB) and Privileged Register Block (PRB) of NETC. For i.MX platforms, it is also used to configure the NETCMIX block. The IERB contains registers that are used for pre-boot initialization, debug, and non-customer configuration. The PRB controls global reset and global error handling for NETC. The NETCMIX block is mainly used to set MII protocol and PCS protocol of the links, it also contains settings for some other functions. Note the IERB configuration registers can only be written after being unlocked by PRB, otherwise, all write operations are inhibited. A warm reset is performed when the IERB is unlocked, and it results in an FLR to all NETC devices. Therefore, all NETC device drivers must be probed or initialized after the warm reset is finished. Signed-off-by: Wei Fang Reviewed-by: Frank Li Signed-off-by: David S. Miller --- drivers/net/ethernet/freescale/enetc/Kconfig | 14 + drivers/net/ethernet/freescale/enetc/Makefile | 3 + .../net/ethernet/freescale/enetc/netc_blk_ctrl.c | 445 +++++++++++++++++++++ include/linux/fsl/netc_global.h | 19 + 4 files changed, 481 insertions(+) create mode 100644 drivers/net/ethernet/freescale/enetc/netc_blk_ctrl.c create mode 100644 include/linux/fsl/netc_global.h (limited to 'include/linux') diff --git a/drivers/net/ethernet/freescale/enetc/Kconfig b/drivers/net/ethernet/freescale/enetc/Kconfig index 4d75e6807e92..51d80ea959d4 100644 --- a/drivers/net/ethernet/freescale/enetc/Kconfig +++ b/drivers/net/ethernet/freescale/enetc/Kconfig @@ -75,3 +75,17 @@ config FSL_ENETC_QOS enable/disable from user space via Qos commands(tc). In the kernel side, it can be loaded by Qos driver. Currently, it is only support taprio(802.1Qbv) and Credit Based Shaper(802.1Qbu). + +config NXP_NETC_BLK_CTRL + tristate "NETC blocks control driver" + help + This driver configures Integrated Endpoint Register Block (IERB) and + Privileged Register Block (PRB) of NETC. For i.MX platforms, it also + includes the configuration of NETCMIX block. + The IERB contains registers that are used for pre-boot initialization, + debug, and non-customer configuration. The PRB controls global reset + and global error handling for NETC. The NETCMIX block is mainly used + to set MII protocol and PCS protocol of the links, it also contains + settings for some other functions. + + If compiled as module (M), the module name is nxp-netc-blk-ctrl. diff --git a/drivers/net/ethernet/freescale/enetc/Makefile b/drivers/net/ethernet/freescale/enetc/Makefile index b13cbbabb2ea..737c32f83ea5 100644 --- a/drivers/net/ethernet/freescale/enetc/Makefile +++ b/drivers/net/ethernet/freescale/enetc/Makefile @@ -19,3 +19,6 @@ fsl-enetc-mdio-y := enetc_pci_mdio.o enetc_mdio.o obj-$(CONFIG_FSL_ENETC_PTP_CLOCK) += fsl-enetc-ptp.o fsl-enetc-ptp-y := enetc_ptp.o + +obj-$(CONFIG_NXP_NETC_BLK_CTRL) += nxp-netc-blk-ctrl.o +nxp-netc-blk-ctrl-y := netc_blk_ctrl.o diff --git a/drivers/net/ethernet/freescale/enetc/netc_blk_ctrl.c b/drivers/net/ethernet/freescale/enetc/netc_blk_ctrl.c new file mode 100644 index 000000000000..bcb8eefeb93c --- /dev/null +++ b/drivers/net/ethernet/freescale/enetc/netc_blk_ctrl.c @@ -0,0 +1,445 @@ +// SPDX-License-Identifier: (GPL-2.0+ OR BSD-3-Clause) +/* + * NXP NETC Blocks Control Driver + * + * Copyright 2024 NXP + * + * This driver is used for pre-initialization of NETC, such as PCS and MII + * protocols, LDID, warm reset, etc. Therefore, all NETC device drivers can + * only be probed after the netc-blk-crtl driver has completed initialization. + * In addition, when the system enters suspend mode, IERB, PRB, and NETCMIX + * will be powered off, except for WOL. Therefore, when the system resumes, + * these blocks need to be reinitialized. + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +/* NETCMIX registers */ +#define IMX95_CFG_LINK_IO_VAR 0x0 +#define IO_VAR_16FF_16G_SERDES 0x1 +#define IO_VAR(port, var) (((var) & 0xf) << ((port) << 2)) + +#define IMX95_CFG_LINK_MII_PROT 0x4 +#define CFG_LINK_MII_PORT_0 GENMASK(3, 0) +#define CFG_LINK_MII_PORT_1 GENMASK(7, 4) +#define MII_PROT_MII 0x0 +#define MII_PROT_RMII 0x1 +#define MII_PROT_RGMII 0x2 +#define MII_PROT_SERIAL 0x3 +#define MII_PROT(port, prot) (((prot) & 0xf) << ((port) << 2)) + +#define IMX95_CFG_LINK_PCS_PROT(a) (0x8 + (a) * 4) +#define PCS_PROT_1G_SGMII BIT(0) +#define PCS_PROT_2500M_SGMII BIT(1) +#define PCS_PROT_XFI BIT(3) +#define PCS_PROT_SFI BIT(4) +#define PCS_PROT_10G_SXGMII BIT(6) + +/* NETC privileged register block register */ +#define PRB_NETCRR 0x100 +#define NETCRR_SR BIT(0) +#define NETCRR_LOCK BIT(1) + +#define PRB_NETCSR 0x104 +#define NETCSR_ERROR BIT(0) +#define NETCSR_STATE BIT(1) + +/* NETC integrated endpoint register block register */ +#define IERB_EMDIOFAUXR 0x344 +#define IERB_T0FAUXR 0x444 +#define IERB_EFAUXR(a) (0x3044 + 0x100 * (a)) +#define IERB_VFAUXR(a) (0x4004 + 0x40 * (a)) +#define FAUXR_LDID GENMASK(3, 0) + +/* Platform information */ +#define IMX95_ENETC0_BUS_DEVFN 0x0 +#define IMX95_ENETC1_BUS_DEVFN 0x40 +#define IMX95_ENETC2_BUS_DEVFN 0x80 + +/* Flags for different platforms */ +#define NETC_HAS_NETCMIX BIT(0) + +struct netc_devinfo { + u32 flags; + int (*netcmix_init)(struct platform_device *pdev); + int (*ierb_init)(struct platform_device *pdev); +}; + +struct netc_blk_ctrl { + void __iomem *prb; + void __iomem *ierb; + void __iomem *netcmix; + + const struct netc_devinfo *devinfo; + struct platform_device *pdev; + struct dentry *debugfs_root; +}; + +static void netc_reg_write(void __iomem *base, u32 offset, u32 val) +{ + netc_write(base + offset, val); +} + +static u32 netc_reg_read(void __iomem *base, u32 offset) +{ + return netc_read(base + offset); +} + +static int netc_of_pci_get_bus_devfn(struct device_node *np) +{ + u32 reg[5]; + int error; + + error = of_property_read_u32_array(np, "reg", reg, ARRAY_SIZE(reg)); + if (error) + return error; + + return (reg[0] >> 8) & 0xffff; +} + +static int netc_get_link_mii_protocol(phy_interface_t interface) +{ + switch (interface) { + case PHY_INTERFACE_MODE_MII: + return MII_PROT_MII; + case PHY_INTERFACE_MODE_RMII: + return MII_PROT_RMII; + case PHY_INTERFACE_MODE_RGMII: + case PHY_INTERFACE_MODE_RGMII_ID: + case PHY_INTERFACE_MODE_RGMII_RXID: + case PHY_INTERFACE_MODE_RGMII_TXID: + return MII_PROT_RGMII; + case PHY_INTERFACE_MODE_SGMII: + case PHY_INTERFACE_MODE_2500BASEX: + case PHY_INTERFACE_MODE_10GBASER: + case PHY_INTERFACE_MODE_XGMII: + case PHY_INTERFACE_MODE_USXGMII: + return MII_PROT_SERIAL; + default: + return -EINVAL; + } +} + +static int imx95_netcmix_init(struct platform_device *pdev) +{ + struct netc_blk_ctrl *priv = platform_get_drvdata(pdev); + struct device_node *np = pdev->dev.of_node; + phy_interface_t interface; + int bus_devfn, mii_proto; + u32 val; + int err; + + /* Default setting of MII protocol */ + val = MII_PROT(0, MII_PROT_RGMII) | MII_PROT(1, MII_PROT_RGMII) | + MII_PROT(2, MII_PROT_SERIAL); + + /* Update the link MII protocol through parsing phy-mode */ + for_each_available_child_of_node_scoped(np, child) { + for_each_available_child_of_node_scoped(child, gchild) { + if (!of_device_is_compatible(gchild, "pci1131,e101")) + continue; + + bus_devfn = netc_of_pci_get_bus_devfn(gchild); + if (bus_devfn < 0) + return -EINVAL; + + if (bus_devfn == IMX95_ENETC2_BUS_DEVFN) + continue; + + err = of_get_phy_mode(gchild, &interface); + if (err) + continue; + + mii_proto = netc_get_link_mii_protocol(interface); + if (mii_proto < 0) + return -EINVAL; + + switch (bus_devfn) { + case IMX95_ENETC0_BUS_DEVFN: + val = u32_replace_bits(val, mii_proto, + CFG_LINK_MII_PORT_0); + break; + case IMX95_ENETC1_BUS_DEVFN: + val = u32_replace_bits(val, mii_proto, + CFG_LINK_MII_PORT_1); + break; + default: + return -EINVAL; + } + } + } + + /* Configure Link I/O variant */ + netc_reg_write(priv->netcmix, IMX95_CFG_LINK_IO_VAR, + IO_VAR(2, IO_VAR_16FF_16G_SERDES)); + /* Configure Link 2 PCS protocol */ + netc_reg_write(priv->netcmix, IMX95_CFG_LINK_PCS_PROT(2), + PCS_PROT_10G_SXGMII); + netc_reg_write(priv->netcmix, IMX95_CFG_LINK_MII_PROT, val); + + return 0; +} + +static bool netc_ierb_is_locked(struct netc_blk_ctrl *priv) +{ + return !!(netc_reg_read(priv->prb, PRB_NETCRR) & NETCRR_LOCK); +} + +static int netc_lock_ierb(struct netc_blk_ctrl *priv) +{ + u32 val; + + netc_reg_write(priv->prb, PRB_NETCRR, NETCRR_LOCK); + + return read_poll_timeout(netc_reg_read, val, !(val & NETCSR_STATE), + 100, 2000, false, priv->prb, PRB_NETCSR); +} + +static int netc_unlock_ierb_with_warm_reset(struct netc_blk_ctrl *priv) +{ + u32 val; + + netc_reg_write(priv->prb, PRB_NETCRR, 0); + + return read_poll_timeout(netc_reg_read, val, !(val & NETCRR_LOCK), + 1000, 100000, true, priv->prb, PRB_NETCRR); +} + +static int imx95_ierb_init(struct platform_device *pdev) +{ + struct netc_blk_ctrl *priv = platform_get_drvdata(pdev); + + /* EMDIO : No MSI-X intterupt */ + netc_reg_write(priv->ierb, IERB_EMDIOFAUXR, 0); + /* ENETC0 PF */ + netc_reg_write(priv->ierb, IERB_EFAUXR(0), 0); + /* ENETC0 VF0 */ + netc_reg_write(priv->ierb, IERB_VFAUXR(0), 1); + /* ENETC0 VF1 */ + netc_reg_write(priv->ierb, IERB_VFAUXR(1), 2); + /* ENETC1 PF */ + netc_reg_write(priv->ierb, IERB_EFAUXR(1), 3); + /* ENETC1 VF0 */ + netc_reg_write(priv->ierb, IERB_VFAUXR(2), 5); + /* ENETC1 VF1 */ + netc_reg_write(priv->ierb, IERB_VFAUXR(3), 6); + /* ENETC2 PF */ + netc_reg_write(priv->ierb, IERB_EFAUXR(2), 4); + /* ENETC2 VF0 */ + netc_reg_write(priv->ierb, IERB_VFAUXR(4), 5); + /* ENETC2 VF1 */ + netc_reg_write(priv->ierb, IERB_VFAUXR(5), 6); + /* NETC TIMER */ + netc_reg_write(priv->ierb, IERB_T0FAUXR, 7); + + return 0; +} + +static int netc_ierb_init(struct platform_device *pdev) +{ + struct netc_blk_ctrl *priv = platform_get_drvdata(pdev); + const struct netc_devinfo *devinfo = priv->devinfo; + int err; + + if (netc_ierb_is_locked(priv)) { + err = netc_unlock_ierb_with_warm_reset(priv); + if (err) { + dev_err(&pdev->dev, "Unlock IERB failed.\n"); + return err; + } + } + + if (devinfo->ierb_init) { + err = devinfo->ierb_init(pdev); + if (err) + return err; + } + + err = netc_lock_ierb(priv); + if (err) { + dev_err(&pdev->dev, "Lock IERB failed.\n"); + return err; + } + + return 0; +} + +#if IS_ENABLED(CONFIG_DEBUG_FS) +static int netc_prb_show(struct seq_file *s, void *data) +{ + struct netc_blk_ctrl *priv = s->private; + u32 val; + + val = netc_reg_read(priv->prb, PRB_NETCRR); + seq_printf(s, "[PRB NETCRR] Lock:%d SR:%d\n", + (val & NETCRR_LOCK) ? 1 : 0, + (val & NETCRR_SR) ? 1 : 0); + + val = netc_reg_read(priv->prb, PRB_NETCSR); + seq_printf(s, "[PRB NETCSR] State:%d Error:%d\n", + (val & NETCSR_STATE) ? 1 : 0, + (val & NETCSR_ERROR) ? 1 : 0); + + return 0; +} +DEFINE_SHOW_ATTRIBUTE(netc_prb); + +static void netc_blk_ctrl_create_debugfs(struct netc_blk_ctrl *priv) +{ + struct dentry *root; + + root = debugfs_create_dir("netc_blk_ctrl", NULL); + if (IS_ERR(root)) + return; + + priv->debugfs_root = root; + + debugfs_create_file("prb", 0444, root, priv, &netc_prb_fops); +} + +static void netc_blk_ctrl_remove_debugfs(struct netc_blk_ctrl *priv) +{ + debugfs_remove_recursive(priv->debugfs_root); + priv->debugfs_root = NULL; +} + +#else + +static void netc_blk_ctrl_create_debugfs(struct netc_blk_ctrl *priv) +{ +} + +static void netc_blk_ctrl_remove_debugfs(struct netc_blk_ctrl *priv) +{ +} +#endif + +static int netc_prb_check_error(struct netc_blk_ctrl *priv) +{ + if (netc_reg_read(priv->prb, PRB_NETCSR) & NETCSR_ERROR) + return -1; + + return 0; +} + +static const struct netc_devinfo imx95_devinfo = { + .flags = NETC_HAS_NETCMIX, + .netcmix_init = imx95_netcmix_init, + .ierb_init = imx95_ierb_init, +}; + +static const struct of_device_id netc_blk_ctrl_match[] = { + { .compatible = "nxp,imx95-netc-blk-ctrl", .data = &imx95_devinfo }, + {}, +}; +MODULE_DEVICE_TABLE(of, netc_blk_ctrl_match); + +static int netc_blk_ctrl_probe(struct platform_device *pdev) +{ + struct device_node *node = pdev->dev.of_node; + const struct netc_devinfo *devinfo; + struct device *dev = &pdev->dev; + const struct of_device_id *id; + struct netc_blk_ctrl *priv; + struct clk *ipg_clk; + void __iomem *regs; + int err; + + priv = devm_kzalloc(dev, sizeof(*priv), GFP_KERNEL); + if (!priv) + return -ENOMEM; + + priv->pdev = pdev; + ipg_clk = devm_clk_get_optional_enabled(dev, "ipg"); + if (IS_ERR(ipg_clk)) + return dev_err_probe(dev, PTR_ERR(ipg_clk), + "Set ipg clock failed\n"); + + id = of_match_device(netc_blk_ctrl_match, dev); + if (!id) + return dev_err_probe(dev, -EINVAL, "Cannot match device\n"); + + devinfo = (struct netc_devinfo *)id->data; + if (!devinfo) + return dev_err_probe(dev, -EINVAL, "No device information\n"); + + priv->devinfo = devinfo; + regs = devm_platform_ioremap_resource_byname(pdev, "ierb"); + if (IS_ERR(regs)) + return dev_err_probe(dev, PTR_ERR(regs), + "Missing IERB resource\n"); + + priv->ierb = regs; + regs = devm_platform_ioremap_resource_byname(pdev, "prb"); + if (IS_ERR(regs)) + return dev_err_probe(dev, PTR_ERR(regs), + "Missing PRB resource\n"); + + priv->prb = regs; + if (devinfo->flags & NETC_HAS_NETCMIX) { + regs = devm_platform_ioremap_resource_byname(pdev, "netcmix"); + if (IS_ERR(regs)) + return dev_err_probe(dev, PTR_ERR(regs), + "Missing NETCMIX resource\n"); + priv->netcmix = regs; + } + + platform_set_drvdata(pdev, priv); + if (devinfo->netcmix_init) { + err = devinfo->netcmix_init(pdev); + if (err) + return dev_err_probe(dev, err, + "Initializing NETCMIX failed\n"); + } + + err = netc_ierb_init(pdev); + if (err) + return dev_err_probe(dev, err, "Initializing IERB failed\n"); + + if (netc_prb_check_error(priv) < 0) + dev_warn(dev, "The current IERB configuration is invalid\n"); + + netc_blk_ctrl_create_debugfs(priv); + + err = of_platform_populate(node, NULL, NULL, dev); + if (err) { + netc_blk_ctrl_remove_debugfs(priv); + return dev_err_probe(dev, err, "of_platform_populate failed\n"); + } + + return 0; +} + +static void netc_blk_ctrl_remove(struct platform_device *pdev) +{ + struct netc_blk_ctrl *priv = platform_get_drvdata(pdev); + + of_platform_depopulate(&pdev->dev); + netc_blk_ctrl_remove_debugfs(priv); +} + +static struct platform_driver netc_blk_ctrl_driver = { + .driver = { + .name = "nxp-netc-blk-ctrl", + .of_match_table = netc_blk_ctrl_match, + }, + .probe = netc_blk_ctrl_probe, + .remove = netc_blk_ctrl_remove, +}; + +module_platform_driver(netc_blk_ctrl_driver); + +MODULE_DESCRIPTION("NXP NETC Blocks Control Driver"); +MODULE_LICENSE("Dual BSD/GPL"); diff --git a/include/linux/fsl/netc_global.h b/include/linux/fsl/netc_global.h new file mode 100644 index 000000000000..fdecca8c90f0 --- /dev/null +++ b/include/linux/fsl/netc_global.h @@ -0,0 +1,19 @@ +/* SPDX-License-Identifier: (GPL-2.0+ OR BSD-3-Clause) */ +/* Copyright 2024 NXP + */ +#ifndef __NETC_GLOBAL_H +#define __NETC_GLOBAL_H + +#include + +static inline u32 netc_read(void __iomem *reg) +{ + return ioread32(reg); +} + +static inline void netc_write(void __iomem *reg, u32 val) +{ + iowrite32(val, reg); +} + +#endif -- cgit v1.2.3 From 0a2cdeeae9ddc14d752173be6af98bc9fb45c6ad Mon Sep 17 00:00:00 2001 From: Menglong Dong Date: Mon, 4 Nov 2024 15:00:41 +0800 Subject: net: tcp: replace the document for "lsndtime" in tcp_sock Commit d5fed5addb2b ("tcp: reorganize tcp_sock fast path variables") moved the fields around and misplaced the documentation for "lsndtime". So, let's replace it in the proper place. Signed-off-by: Menglong Dong Reviewed-by: Eric Dumazet Link: https://patch.msgid.link/20241104070041.64302-1-dongml2@chinatelecom.cn Signed-off-by: Jakub Kicinski --- include/linux/tcp.h | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) (limited to 'include/linux') diff --git a/include/linux/tcp.h b/include/linux/tcp.h index 6a5e08b937b3..f88daaa76d83 100644 --- a/include/linux/tcp.h +++ b/include/linux/tcp.h @@ -200,7 +200,6 @@ struct tcp_sock { /* TX read-mostly hotpath cache lines */ __cacheline_group_begin(tcp_sock_read_tx); - /* timestamp of last sent data packet (for restart window) */ u32 max_window; /* Maximal window ever seen from peer */ u32 rcv_ssthresh; /* Current window clamp */ u32 reordering; /* Packet reordering metric. */ @@ -263,7 +262,7 @@ struct tcp_sock { u32 chrono_stat[3]; /* Time in jiffies for chrono_stat stats */ u32 write_seq; /* Tail(+1) of data held in tcp send buffer */ u32 pushed_seq; /* Last pushed seq, required to talk to windows */ - u32 lsndtime; + u32 lsndtime; /* timestamp of last sent data packet (for restart window) */ u32 mdev_us; /* medium deviation */ u32 rtt_seq; /* sequence number to update rttvar */ u64 tcp_wstamp_ns; /* departure time for next sent data packet */ -- cgit v1.2.3 From 1e4033b53db48cc77c436feac9c6d7d63db87b87 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Tue, 5 Nov 2024 17:43:57 +0000 Subject: net: skb_reset_mac_len() must check if mac_header was set Recent discussions show that skb_reset_mac_len() should be more careful. We expect the MAC header being set. If not, clear skb->mac_len and fire a warning for CONFIG_DEBUG_NET=y builds. If after investigations we find that not having a MAC header was okay, we can remove the warning. Signed-off-by: Eric Dumazet Link: https://lore.kernel.org/netdev/CANn89iJZGH+yEfJxfPWa3Hm7jxb-aeY2Up4HufmLMnVuQXt38A@mail.gmail.com/T/ Cc: En-Wei Wu Reviewed-by: Joe Damato Reviewed-by: Xuan Zhuo Link: https://patch.msgid.link/20241105174403.850330-2-edumazet@google.com Signed-off-by: Jakub Kicinski --- include/linux/skbuff.h | 17 +++++++++++------ 1 file changed, 11 insertions(+), 6 deletions(-) (limited to 'include/linux') diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h index 48f1e0fa2a13..5d8fefa71cac 100644 --- a/include/linux/skbuff.h +++ b/include/linux/skbuff.h @@ -2909,9 +2909,19 @@ static inline void skb_reset_inner_headers(struct sk_buff *skb) skb->inner_transport_header = skb->transport_header; } +static inline int skb_mac_header_was_set(const struct sk_buff *skb) +{ + return skb->mac_header != (typeof(skb->mac_header))~0U; +} + static inline void skb_reset_mac_len(struct sk_buff *skb) { - skb->mac_len = skb->network_header - skb->mac_header; + if (!skb_mac_header_was_set(skb)) { + DEBUG_NET_WARN_ON_ONCE(1); + skb->mac_len = 0; + } else { + skb->mac_len = skb->network_header - skb->mac_header; + } } static inline unsigned char *skb_inner_transport_header(const struct sk_buff @@ -3014,11 +3024,6 @@ static inline void skb_set_network_header(struct sk_buff *skb, const int offset) skb->network_header += offset; } -static inline int skb_mac_header_was_set(const struct sk_buff *skb) -{ - return skb->mac_header != (typeof(skb->mac_header))~0U; -} - static inline unsigned char *skb_mac_header(const struct sk_buff *skb) { DEBUG_NET_WARN_ON_ONCE(!skb_mac_header_was_set(skb)); -- cgit v1.2.3 From cfe8394e06f22847ecae0312bdd0b633b471b3f0 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Tue, 5 Nov 2024 17:43:58 +0000 Subject: net: add debug check in skb_reset_inner_transport_header() Make sure (skb->data - skb->head) can fit in skb->inner_transport_header This needs CONFIG_DEBUG_NET=y. Signed-off-by: Eric Dumazet Reviewed-by: Joe Damato Reviewed-by: Xuan Zhuo Link: https://patch.msgid.link/20241105174403.850330-3-edumazet@google.com Signed-off-by: Jakub Kicinski --- include/linux/skbuff.h | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) (limited to 'include/linux') diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h index 5d8fefa71cac..75795568803c 100644 --- a/include/linux/skbuff.h +++ b/include/linux/skbuff.h @@ -2937,7 +2937,10 @@ static inline int skb_inner_transport_offset(const struct sk_buff *skb) static inline void skb_reset_inner_transport_header(struct sk_buff *skb) { - skb->inner_transport_header = skb->data - skb->head; + long offset = skb->data - skb->head; + + DEBUG_NET_WARN_ON_ONCE(offset != (typeof(skb->inner_transport_header))offset); + skb->inner_transport_header = offset; } static inline void skb_set_inner_transport_header(struct sk_buff *skb, -- cgit v1.2.3 From 1732e4bedb3e29986da6ea315e8bf8caf74e1e38 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Tue, 5 Nov 2024 17:43:59 +0000 Subject: net: add debug check in skb_reset_inner_network_header() Make sure (skb->data - skb->head) can fit in skb->inner_network_header This needs CONFIG_DEBUG_NET=y. Signed-off-by: Eric Dumazet Reviewed-by: Joe Damato Reviewed-by: Xuan Zhuo Link: https://patch.msgid.link/20241105174403.850330-4-edumazet@google.com Signed-off-by: Jakub Kicinski --- include/linux/skbuff.h | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) (limited to 'include/linux') diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h index 75795568803c..8dafd1295a6e 100644 --- a/include/linux/skbuff.h +++ b/include/linux/skbuff.h @@ -2957,7 +2957,10 @@ static inline unsigned char *skb_inner_network_header(const struct sk_buff *skb) static inline void skb_reset_inner_network_header(struct sk_buff *skb) { - skb->inner_network_header = skb->data - skb->head; + long offset = skb->data - skb->head; + + DEBUG_NET_WARN_ON_ONCE(offset != (typeof(skb->inner_network_header))offset); + skb->inner_network_header = offset; } static inline void skb_set_inner_network_header(struct sk_buff *skb, -- cgit v1.2.3 From 78a0cb2f45dc9327f26956c242f7f4b450efff49 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Tue, 5 Nov 2024 17:44:00 +0000 Subject: net: add debug check in skb_reset_inner_mac_header() Make sure (skb->data - skb->head) can fit in skb->inner_mac_header This needs CONFIG_DEBUG_NET=y. Signed-off-by: Eric Dumazet Reviewed-by: Joe Damato Reviewed-by: Xuan Zhuo Link: https://patch.msgid.link/20241105174403.850330-5-edumazet@google.com Signed-off-by: Jakub Kicinski --- include/linux/skbuff.h | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) (limited to 'include/linux') diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h index 8dafd1295a6e..32a7d8a65b9f 100644 --- a/include/linux/skbuff.h +++ b/include/linux/skbuff.h @@ -2982,7 +2982,10 @@ static inline unsigned char *skb_inner_mac_header(const struct sk_buff *skb) static inline void skb_reset_inner_mac_header(struct sk_buff *skb) { - skb->inner_mac_header = skb->data - skb->head; + long offset = skb->data - skb->head; + + DEBUG_NET_WARN_ON_ONCE(offset != (typeof(skb->inner_mac_header))offset); + skb->inner_mac_header = offset; } static inline void skb_set_inner_mac_header(struct sk_buff *skb, -- cgit v1.2.3 From ae50ea52bdd77eabd8f3c4630c1b03c3373f61a5 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Tue, 5 Nov 2024 17:44:01 +0000 Subject: net: add debug check in skb_reset_transport_header() Make sure (skb->data - skb->head) can fit in skb->transport_header This needs CONFIG_DEBUG_NET=y. Signed-off-by: Eric Dumazet Reviewed-by: Joe Damato Reviewed-by: Xuan Zhuo Link: https://patch.msgid.link/20241105174403.850330-6-edumazet@google.com Signed-off-by: Jakub Kicinski --- include/linux/skbuff.h | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) (limited to 'include/linux') diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h index 32a7d8a65b9f..f76524844e7e 100644 --- a/include/linux/skbuff.h +++ b/include/linux/skbuff.h @@ -3007,7 +3007,10 @@ static inline unsigned char *skb_transport_header(const struct sk_buff *skb) static inline void skb_reset_transport_header(struct sk_buff *skb) { - skb->transport_header = skb->data - skb->head; + long offset = skb->data - skb->head; + + DEBUG_NET_WARN_ON_ONCE(offset != (typeof(skb->transport_header))offset); + skb->transport_header = offset; } static inline void skb_set_transport_header(struct sk_buff *skb, -- cgit v1.2.3 From 305ae87dafc1a9f35374a783119d9e6001d1b8e0 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Tue, 5 Nov 2024 17:44:02 +0000 Subject: net: add debug check in skb_reset_network_header() Make sure (skb->data - skb->head) can fit in skb->network_header This needs CONFIG_DEBUG_NET=y. Signed-off-by: Eric Dumazet Reviewed-by: Joe Damato Reviewed-by: Xuan Zhuo Link: https://patch.msgid.link/20241105174403.850330-7-edumazet@google.com Signed-off-by: Jakub Kicinski --- include/linux/skbuff.h | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) (limited to 'include/linux') diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h index f76524844e7e..f1e49a32880d 100644 --- a/include/linux/skbuff.h +++ b/include/linux/skbuff.h @@ -3027,7 +3027,10 @@ static inline unsigned char *skb_network_header(const struct sk_buff *skb) static inline void skb_reset_network_header(struct sk_buff *skb) { - skb->network_header = skb->data - skb->head; + long offset = skb->data - skb->head; + + DEBUG_NET_WARN_ON_ONCE(offset != (typeof(skb->network_header))offset); + skb->network_header = offset; } static inline void skb_set_network_header(struct sk_buff *skb, const int offset) -- cgit v1.2.3 From 3b6167e9bfc9eae4edad065c166c667f9a8e693a Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Tue, 5 Nov 2024 17:44:03 +0000 Subject: net: add debug check in skb_reset_mac_header() Make sure (skb->data - skb->head) can fit in skb->mac_header This needs CONFIG_DEBUG_NET=y. Signed-off-by: Eric Dumazet Reviewed-by: Joe Damato Reviewed-by: Xuan Zhuo Link: https://patch.msgid.link/20241105174403.850330-8-edumazet@google.com Signed-off-by: Jakub Kicinski --- include/linux/skbuff.h | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) (limited to 'include/linux') diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h index f1e49a32880d..e5095d75abba 100644 --- a/include/linux/skbuff.h +++ b/include/linux/skbuff.h @@ -3063,7 +3063,10 @@ static inline void skb_unset_mac_header(struct sk_buff *skb) static inline void skb_reset_mac_header(struct sk_buff *skb) { - skb->mac_header = skb->data - skb->head; + long offset = skb->data - skb->head; + + DEBUG_NET_WARN_ON_ONCE(offset != (typeof(skb->mac_header))offset); + skb->mac_header = offset; } static inline void skb_set_mac_header(struct sk_buff *skb, const int offset) -- cgit v1.2.3 From 495e7c8e9601245738a6ab53c992cc5aa0b13cb4 Mon Sep 17 00:00:00 2001 From: Jinjian Song Date: Mon, 4 Nov 2024 17:44:34 +0800 Subject: wwan: core: Add WWAN ADB and MIPC port type Add new WWAN ports that connect to the device's ADB protocol interface and MTK MIPC diagnostic interface. Signed-off-by: Jinjian Song Reviewed-by: Sergey Ryazanov Signed-off-by: Paolo Abeni --- drivers/net/wwan/wwan_core.c | 8 ++++++++ include/linux/wwan.h | 4 ++++ 2 files changed, 12 insertions(+) (limited to 'include/linux') diff --git a/drivers/net/wwan/wwan_core.c b/drivers/net/wwan/wwan_core.c index 8dc88cc65781..a51e2755991a 100644 --- a/drivers/net/wwan/wwan_core.c +++ b/drivers/net/wwan/wwan_core.c @@ -334,6 +334,14 @@ static const struct { .name = "FASTBOOT", .devsuf = "fastboot", }, + [WWAN_PORT_ADB] = { + .name = "ADB", + .devsuf = "adb", + }, + [WWAN_PORT_MIPC] = { + .name = "MIPC", + .devsuf = "mipc", + }, }; static ssize_t type_show(struct device *dev, struct device_attribute *attr, diff --git a/include/linux/wwan.h b/include/linux/wwan.h index 170fdee6339c..79c781875c09 100644 --- a/include/linux/wwan.h +++ b/include/linux/wwan.h @@ -17,6 +17,8 @@ * @WWAN_PORT_FIREHOSE: XML based command protocol * @WWAN_PORT_XMMRPC: Control protocol for Intel XMM modems * @WWAN_PORT_FASTBOOT: Fastboot protocol control + * @WWAN_PORT_ADB: ADB protocol control + * @WWAN_PORT_MIPC: MTK MIPC diagnostic interface * * @WWAN_PORT_MAX: Highest supported port types * @WWAN_PORT_UNKNOWN: Special value to indicate an unknown port type @@ -30,6 +32,8 @@ enum wwan_port_type { WWAN_PORT_FIREHOSE, WWAN_PORT_XMMRPC, WWAN_PORT_FASTBOOT, + WWAN_PORT_ADB, + WWAN_PORT_MIPC, /* Add new port types above this line */ -- cgit v1.2.3 From 9c477088b60dd23f3120b2b01b14d85047a033da Mon Sep 17 00:00:00 2001 From: Heiner Kallweit Date: Wed, 6 Nov 2024 21:19:59 +0100 Subject: net: phy: make genphy_c45_write_eee_adv() static genphy_c45_write_eee_adv() isn't used outside phy-c45.c, so make it static. Signed-off-by: Heiner Kallweit Reviewed-by: Andrew Lunn Reviewed-by: Florian Fainelli Link: https://patch.msgid.link/d23bd784-44e6-4a15-af3a-b37379156521@gmail.com Signed-off-by: Jakub Kicinski --- drivers/net/phy/phy-c45.c | 3 ++- include/linux/phy.h | 1 - 2 files changed, 2 insertions(+), 2 deletions(-) (limited to 'include/linux') diff --git a/drivers/net/phy/phy-c45.c b/drivers/net/phy/phy-c45.c index 8c4dd94c7495..66be5832ba51 100644 --- a/drivers/net/phy/phy-c45.c +++ b/drivers/net/phy/phy-c45.c @@ -680,7 +680,8 @@ EXPORT_SYMBOL_GPL(genphy_c45_read_mdix); * @phydev: target phy_device struct * @adv: the linkmode advertisement settings */ -int genphy_c45_write_eee_adv(struct phy_device *phydev, unsigned long *adv) +static int genphy_c45_write_eee_adv(struct phy_device *phydev, + unsigned long *adv) { int val, changed = 0; diff --git a/include/linux/phy.h b/include/linux/phy.h index bf0eb4e5d35c..bb09a5a128e1 100644 --- a/include/linux/phy.h +++ b/include/linux/phy.h @@ -1953,7 +1953,6 @@ int genphy_c45_ethtool_get_eee(struct phy_device *phydev, struct ethtool_keee *data); int genphy_c45_ethtool_set_eee(struct phy_device *phydev, struct ethtool_keee *data); -int genphy_c45_write_eee_adv(struct phy_device *phydev, unsigned long *adv); int genphy_c45_an_config_eee_aneg(struct phy_device *phydev); int genphy_c45_read_eee_adv(struct phy_device *phydev, unsigned long *adv); -- cgit v1.2.3 From db73835f54fc57bc267d43836905588f24ddac85 Mon Sep 17 00:00:00 2001 From: Heiner Kallweit Date: Wed, 6 Nov 2024 21:25:12 +0100 Subject: net: phy: remove genphy_config_eee_advert bcm_config_lre_aneg() doesn't use genphy_config_eee_advert() any longer. As this was the only user, we can remove genphy_config_eee_advert() now. Signed-off-by: Heiner Kallweit Reviewed-by: Andrew Lunn Reviewed-by: Florian Fainelli Link: https://patch.msgid.link/37da7f3e-b883-4c07-9881-b8c0516822b7@gmail.com Signed-off-by: Jakub Kicinski --- drivers/net/phy/phy_device.c | 23 ----------------------- include/linux/phy.h | 1 - 2 files changed, 24 deletions(-) (limited to 'include/linux') diff --git a/drivers/net/phy/phy_device.c b/drivers/net/phy/phy_device.c index 563497a3274c..bc24c9f2786b 100644 --- a/drivers/net/phy/phy_device.c +++ b/drivers/net/phy/phy_device.c @@ -2239,29 +2239,6 @@ static int genphy_c37_config_advert(struct phy_device *phydev) adv); } -/** - * genphy_config_eee_advert - disable unwanted eee mode advertisement - * @phydev: target phy_device struct - * - * Description: Writes MDIO_AN_EEE_ADV after disabling unsupported energy - * efficent ethernet modes. Returns 0 if the PHY's advertisement hasn't - * changed, and 1 if it has changed. - */ -int genphy_config_eee_advert(struct phy_device *phydev) -{ - int err; - - /* Nothing to disable */ - if (!phydev->eee_broken_modes) - return 0; - - err = phy_modify_mmd_changed(phydev, MDIO_MMD_AN, MDIO_AN_EEE_ADV, - phydev->eee_broken_modes, 0); - /* If the call failed, we assume that EEE is not supported */ - return err < 0 ? 0 : err; -} -EXPORT_SYMBOL(genphy_config_eee_advert); - /** * genphy_setup_forced - configures/forces speed/duplex from @phydev * @phydev: target phy_device struct diff --git a/include/linux/phy.h b/include/linux/phy.h index bb09a5a128e1..1e4127c495c0 100644 --- a/include/linux/phy.h +++ b/include/linux/phy.h @@ -1885,7 +1885,6 @@ int genphy_read_abilities(struct phy_device *phydev); int genphy_setup_forced(struct phy_device *phydev); int genphy_restart_aneg(struct phy_device *phydev); int genphy_check_and_restart_aneg(struct phy_device *phydev, bool restart); -int genphy_config_eee_advert(struct phy_device *phydev); int __genphy_config_aneg(struct phy_device *phydev, bool changed); int genphy_aneg_done(struct phy_device *phydev); int genphy_update_link(struct phy_device *phydev); -- cgit v1.2.3 From f7f52738637f4361c108cad36e23ee98959a9006 Mon Sep 17 00:00:00 2001 From: Gilad Naaman Date: Thu, 7 Nov 2024 16:04:43 +0000 Subject: neighbour: Create netdev->neighbour association Create a mapping between a netdev and its neighoburs, allowing for much cheaper flushes. Signed-off-by: Gilad Naaman Reviewed-by: Eric Dumazet Reviewed-by: Kuniyuki Iwashima Link: https://patch.msgid.link/20241107160444.2913124-7-gnaaman@drivenets.com Signed-off-by: Jakub Kicinski --- .../networking/net_cachelines/net_device.rst | 1 + include/linux/netdevice.h | 7 ++ include/net/neighbour.h | 9 +- include/net/neighbour_tables.h | 12 +++ net/core/neighbour.c | 96 +++++++++++++--------- 5 files changed, 80 insertions(+), 45 deletions(-) create mode 100644 include/net/neighbour_tables.h (limited to 'include/linux') diff --git a/Documentation/networking/net_cachelines/net_device.rst b/Documentation/networking/net_cachelines/net_device.rst index ade50d4e67cf..15e31ece675f 100644 --- a/Documentation/networking/net_cachelines/net_device.rst +++ b/Documentation/networking/net_cachelines/net_device.rst @@ -188,4 +188,5 @@ u64 max_pacing_offload_horizon struct_napi_config* napi_config unsigned_long gro_flush_timeout u32 napi_defer_hard_irqs +struct hlist_head neighbours[2] =================================== =========================== =================== =================== =================================================================================== diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index 3c552b648b27..df4483598628 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -52,6 +52,7 @@ #include #include #include +#include struct netpoll_info; struct device; @@ -2032,6 +2033,9 @@ enum netdev_reg_state { * @napi_defer_hard_irqs: If not zero, provides a counter that would * allow to avoid NIC hard IRQ, on busy queues. * + * @neighbours: List heads pointing to this device's neighbours' + * dev_list, one per address-family. + * * FIXME: cleanup struct net_device such that network protocol info * moves out. */ @@ -2440,6 +2444,9 @@ struct net_device { */ struct net_shaper_hierarchy *net_shaper_hierarchy; #endif + + struct hlist_head neighbours[NEIGH_NR_TABLES]; + u8 priv[] ____cacheline_aligned __counted_by(priv_len); } ____cacheline_aligned; diff --git a/include/net/neighbour.h b/include/net/neighbour.h index 40aac1e24c68..9a832cab5b1d 100644 --- a/include/net/neighbour.h +++ b/include/net/neighbour.h @@ -29,6 +29,7 @@ #include #include #include +#include /* * NUD stands for "neighbor unreachability detection" @@ -136,6 +137,7 @@ struct neigh_statistics { struct neighbour { struct hlist_node hash; + struct hlist_node dev_list; struct neigh_table *tbl; struct neigh_parms *parms; unsigned long confirmed; @@ -236,13 +238,6 @@ struct neigh_table { struct pneigh_entry **phash_buckets; }; -enum { - NEIGH_ARP_TABLE = 0, - NEIGH_ND_TABLE = 1, - NEIGH_NR_TABLES, - NEIGH_LINK_TABLE = NEIGH_NR_TABLES /* Pseudo table for neigh_xmit */ -}; - static inline int neigh_parms_family(struct neigh_parms *p) { return p->tbl->family; diff --git a/include/net/neighbour_tables.h b/include/net/neighbour_tables.h new file mode 100644 index 000000000000..bcffbe8f7601 --- /dev/null +++ b/include/net/neighbour_tables.h @@ -0,0 +1,12 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef _NET_NEIGHBOUR_TABLES_H +#define _NET_NEIGHBOUR_TABLES_H + +enum { + NEIGH_ARP_TABLE = 0, + NEIGH_ND_TABLE = 1, + NEIGH_NR_TABLES, + NEIGH_LINK_TABLE = NEIGH_NR_TABLES /* Pseudo table for neigh_xmit */ +}; + +#endif diff --git a/net/core/neighbour.c b/net/core/neighbour.c index 59f359c7b5e3..5e572f6eaf2c 100644 --- a/net/core/neighbour.c +++ b/net/core/neighbour.c @@ -60,6 +60,25 @@ static int pneigh_ifdown_and_unlock(struct neigh_table *tbl, static const struct seq_operations neigh_stat_seq_ops; #endif +static struct hlist_head *neigh_get_dev_table(struct net_device *dev, int family) +{ + int i; + + switch (family) { + default: + DEBUG_NET_WARN_ON_ONCE(1); + fallthrough; /* to avoid panic by null-ptr-deref */ + case AF_INET: + i = NEIGH_ARP_TABLE; + break; + case AF_INET6: + i = NEIGH_ND_TABLE; + break; + } + + return &dev->neighbours[i]; +} + /* Neighbour hash table buckets are protected with rwlock tbl->lock. @@ -211,6 +230,7 @@ bool neigh_remove_one(struct neighbour *n) write_lock(&n->lock); if (refcount_read(&n->refcnt) == 1) { hlist_del_rcu(&n->hash); + hlist_del_rcu(&n->dev_list); neigh_mark_dead(n); retval = true; } @@ -351,48 +371,42 @@ static void pneigh_queue_purge(struct sk_buff_head *list, struct net *net, static void neigh_flush_dev(struct neigh_table *tbl, struct net_device *dev, bool skip_perm) { - int i; - struct neigh_hash_table *nht; - - nht = rcu_dereference_protected(tbl->nht, - lockdep_is_held(&tbl->lock)); + struct hlist_head *dev_head; + struct hlist_node *tmp; + struct neighbour *n; - for (i = 0; i < (1 << nht->hash_shift); i++) { - struct hlist_node *tmp; - struct neighbour *n; + dev_head = neigh_get_dev_table(dev, tbl->family); - neigh_for_each_in_bucket_safe(n, tmp, &nht->hash_heads[i]) { - if (dev && n->dev != dev) - continue; - if (skip_perm && n->nud_state & NUD_PERMANENT) - continue; + hlist_for_each_entry_safe(n, tmp, dev_head, dev_list) { + if (skip_perm && n->nud_state & NUD_PERMANENT) + continue; - hlist_del_rcu(&n->hash); - write_lock(&n->lock); - neigh_del_timer(n); - neigh_mark_dead(n); - if (refcount_read(&n->refcnt) != 1) { - /* The most unpleasant situation. - We must destroy neighbour entry, - but someone still uses it. - - The destroy will be delayed until - the last user releases us, but - we must kill timers etc. and move - it to safe state. - */ - __skb_queue_purge(&n->arp_queue); - n->arp_queue_len_bytes = 0; - WRITE_ONCE(n->output, neigh_blackhole); - if (n->nud_state & NUD_VALID) - n->nud_state = NUD_NOARP; - else - n->nud_state = NUD_NONE; - neigh_dbg(2, "neigh %p is stray\n", n); - } - write_unlock(&n->lock); - neigh_cleanup_and_release(n); + hlist_del_rcu(&n->hash); + hlist_del_rcu(&n->dev_list); + write_lock(&n->lock); + neigh_del_timer(n); + neigh_mark_dead(n); + if (refcount_read(&n->refcnt) != 1) { + /* The most unpleasant situation. + * We must destroy neighbour entry, + * but someone still uses it. + * + * The destroy will be delayed until + * the last user releases us, but + * we must kill timers etc. and move + * it to safe state. + */ + __skb_queue_purge(&n->arp_queue); + n->arp_queue_len_bytes = 0; + WRITE_ONCE(n->output, neigh_blackhole); + if (n->nud_state & NUD_VALID) + n->nud_state = NUD_NOARP; + else + n->nud_state = NUD_NONE; + neigh_dbg(2, "neigh %p is stray\n", n); } + write_unlock(&n->lock); + neigh_cleanup_and_release(n); } } @@ -655,6 +669,10 @@ ___neigh_create(struct neigh_table *tbl, const void *pkey, if (want_ref) neigh_hold(n); hlist_add_head_rcu(&n->hash, &nht->hash_heads[hash_val]); + + hlist_add_head_rcu(&n->dev_list, + neigh_get_dev_table(dev, tbl->family)); + write_unlock_bh(&tbl->lock); neigh_dbg(2, "neigh %p is created\n", n); rc = n; @@ -935,6 +953,7 @@ static void neigh_periodic_work(struct work_struct *work) !time_in_range_open(jiffies, n->used, n->used + NEIGH_VAR(n->parms, GC_STALETIME)))) { hlist_del_rcu(&n->hash); + hlist_del_rcu(&n->dev_list); neigh_mark_dead(n); write_unlock(&n->lock); neigh_cleanup_and_release(n); @@ -3054,6 +3073,7 @@ void __neigh_for_each_release(struct neigh_table *tbl, release = cb(n); if (release) { hlist_del_rcu(&n->hash); + hlist_del_rcu(&n->dev_list); neigh_mark_dead(n); } write_unlock(&n->lock); -- cgit v1.2.3 From 65941f10caf2c04781a7defa4ec0ab119dbd235a Mon Sep 17 00:00:00 2001 From: Yunsheng Lin Date: Mon, 28 Oct 2024 19:53:37 +0800 Subject: mm: move the page fragment allocator from page_alloc into its own file Inspired by [1], move the page fragment allocator from page_alloc into its own c file and header file, as we are about to make more change for it to replace another page_frag implementation in sock.c As this patchset is going to replace 'struct page_frag' with 'struct page_frag_cache' in sched.h, including page_frag_cache.h in sched.h has a compiler error caused by interdependence between mm_types.h and mm.h for asm-offsets.c, see [2]. So avoid the compiler error by moving 'struct page_frag_cache' to mm_types_task.h as suggested by Alexander, see [3]. 1. https://lore.kernel.org/all/20230411160902.4134381-3-dhowells@redhat.com/ 2. https://lore.kernel.org/all/15623dac-9358-4597-b3ee-3694a5956920@gmail.com/ 3. https://lore.kernel.org/all/CAKgT0UdH1yD=LSCXFJ=YM_aiA4OomD-2wXykO42bizaWMt_HOA@mail.gmail.com/ CC: David Howells CC: Linux-MM Signed-off-by: Yunsheng Lin Acked-by: Andrew Morton Reviewed-by: Alexander Duyck Link: https://patch.msgid.link/20241028115343.3405838-3-linyunsheng@huawei.com Signed-off-by: Jakub Kicinski --- include/linux/gfp.h | 22 ---- include/linux/mm_types.h | 18 --- include/linux/mm_types_task.h | 18 +++ include/linux/page_frag_cache.h | 31 +++++ include/linux/skbuff.h | 1 + mm/Makefile | 1 + mm/page_alloc.c | 136 ------------------- mm/page_frag_cache.c | 145 +++++++++++++++++++++ .../selftests/mm/page_frag/page_frag_test.c | 2 +- 9 files changed, 197 insertions(+), 177 deletions(-) create mode 100644 include/linux/page_frag_cache.h create mode 100644 mm/page_frag_cache.c (limited to 'include/linux') diff --git a/include/linux/gfp.h b/include/linux/gfp.h index a951de920e20..a0a6d25f883f 100644 --- a/include/linux/gfp.h +++ b/include/linux/gfp.h @@ -371,28 +371,6 @@ __meminit void *alloc_pages_exact_nid_noprof(int nid, size_t size, gfp_t gfp_mas extern void __free_pages(struct page *page, unsigned int order); extern void free_pages(unsigned long addr, unsigned int order); -struct page_frag_cache; -void page_frag_cache_drain(struct page_frag_cache *nc); -extern void __page_frag_cache_drain(struct page *page, unsigned int count); -void *__page_frag_alloc_align(struct page_frag_cache *nc, unsigned int fragsz, - gfp_t gfp_mask, unsigned int align_mask); - -static inline void *page_frag_alloc_align(struct page_frag_cache *nc, - unsigned int fragsz, gfp_t gfp_mask, - unsigned int align) -{ - WARN_ON_ONCE(!is_power_of_2(align)); - return __page_frag_alloc_align(nc, fragsz, gfp_mask, -align); -} - -static inline void *page_frag_alloc(struct page_frag_cache *nc, - unsigned int fragsz, gfp_t gfp_mask) -{ - return __page_frag_alloc_align(nc, fragsz, gfp_mask, ~0u); -} - -extern void page_frag_free(void *addr); - #define __free_page(page) __free_pages((page), 0) #define free_page(addr) free_pages((addr), 0) diff --git a/include/linux/mm_types.h b/include/linux/mm_types.h index 6e3bdf8e38bc..92314ef2d978 100644 --- a/include/linux/mm_types.h +++ b/include/linux/mm_types.h @@ -521,9 +521,6 @@ static_assert(sizeof(struct ptdesc) <= sizeof(struct page)); */ #define STRUCT_PAGE_MAX_SHIFT (order_base_2(sizeof(struct page))) -#define PAGE_FRAG_CACHE_MAX_SIZE __ALIGN_MASK(32768, ~PAGE_MASK) -#define PAGE_FRAG_CACHE_MAX_ORDER get_order(PAGE_FRAG_CACHE_MAX_SIZE) - /* * page_private can be used on tail pages. However, PagePrivate is only * checked by the VM on the head page. So page_private on the tail pages @@ -542,21 +539,6 @@ static inline void *folio_get_private(struct folio *folio) return folio->private; } -struct page_frag_cache { - void * va; -#if (PAGE_SIZE < PAGE_FRAG_CACHE_MAX_SIZE) - __u16 offset; - __u16 size; -#else - __u32 offset; -#endif - /* we maintain a pagecount bias, so that we dont dirty cache line - * containing page->_refcount every time we allocate a fragment. - */ - unsigned int pagecnt_bias; - bool pfmemalloc; -}; - typedef unsigned long vm_flags_t; /* diff --git a/include/linux/mm_types_task.h b/include/linux/mm_types_task.h index bff5706b76e1..0ac6daebdd5c 100644 --- a/include/linux/mm_types_task.h +++ b/include/linux/mm_types_task.h @@ -8,6 +8,7 @@ * (These are defined separately to decouple sched.h from mm_types.h as much as possible.) */ +#include #include #include @@ -43,6 +44,23 @@ struct page_frag { #endif }; +#define PAGE_FRAG_CACHE_MAX_SIZE __ALIGN_MASK(32768, ~PAGE_MASK) +#define PAGE_FRAG_CACHE_MAX_ORDER get_order(PAGE_FRAG_CACHE_MAX_SIZE) +struct page_frag_cache { + void *va; +#if (PAGE_SIZE < PAGE_FRAG_CACHE_MAX_SIZE) + __u16 offset; + __u16 size; +#else + __u32 offset; +#endif + /* we maintain a pagecount bias, so that we dont dirty cache line + * containing page->_refcount every time we allocate a fragment. + */ + unsigned int pagecnt_bias; + bool pfmemalloc; +}; + /* Track pages that require TLB flushes */ struct tlbflush_unmap_batch { #ifdef CONFIG_ARCH_WANT_BATCHED_UNMAP_TLB_FLUSH diff --git a/include/linux/page_frag_cache.h b/include/linux/page_frag_cache.h new file mode 100644 index 000000000000..67ac8626ed9b --- /dev/null +++ b/include/linux/page_frag_cache.h @@ -0,0 +1,31 @@ +/* SPDX-License-Identifier: GPL-2.0 */ + +#ifndef _LINUX_PAGE_FRAG_CACHE_H +#define _LINUX_PAGE_FRAG_CACHE_H + +#include +#include +#include + +void page_frag_cache_drain(struct page_frag_cache *nc); +void __page_frag_cache_drain(struct page *page, unsigned int count); +void *__page_frag_alloc_align(struct page_frag_cache *nc, unsigned int fragsz, + gfp_t gfp_mask, unsigned int align_mask); + +static inline void *page_frag_alloc_align(struct page_frag_cache *nc, + unsigned int fragsz, gfp_t gfp_mask, + unsigned int align) +{ + WARN_ON_ONCE(!is_power_of_2(align)); + return __page_frag_alloc_align(nc, fragsz, gfp_mask, -align); +} + +static inline void *page_frag_alloc(struct page_frag_cache *nc, + unsigned int fragsz, gfp_t gfp_mask) +{ + return __page_frag_alloc_align(nc, fragsz, gfp_mask, ~0u); +} + +void page_frag_free(void *addr); + +#endif diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h index e5095d75abba..60535c706851 100644 --- a/include/linux/skbuff.h +++ b/include/linux/skbuff.h @@ -31,6 +31,7 @@ #include #include #include +#include #include #if IS_ENABLED(CONFIG_NF_CONNTRACK) #include diff --git a/mm/Makefile b/mm/Makefile index d5639b036166..dba52bb0da8a 100644 --- a/mm/Makefile +++ b/mm/Makefile @@ -65,6 +65,7 @@ page-alloc-$(CONFIG_SHUFFLE_PAGE_ALLOCATOR) += shuffle.o memory-hotplug-$(CONFIG_MEMORY_HOTPLUG) += memory_hotplug.o obj-y += page-alloc.o +obj-y += page_frag_cache.o obj-y += init-mm.o obj-y += memblock.o obj-y += $(memory-hotplug-y) diff --git a/mm/page_alloc.c b/mm/page_alloc.c index 94a2ffe28008..34b3eb74630a 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -4836,142 +4836,6 @@ void free_pages(unsigned long addr, unsigned int order) EXPORT_SYMBOL(free_pages); -/* - * Page Fragment: - * An arbitrary-length arbitrary-offset area of memory which resides - * within a 0 or higher order page. Multiple fragments within that page - * are individually refcounted, in the page's reference counter. - * - * The page_frag functions below provide a simple allocation framework for - * page fragments. This is used by the network stack and network device - * drivers to provide a backing region of memory for use as either an - * sk_buff->head, or to be used in the "frags" portion of skb_shared_info. - */ -static struct page *__page_frag_cache_refill(struct page_frag_cache *nc, - gfp_t gfp_mask) -{ - struct page *page = NULL; - gfp_t gfp = gfp_mask; - -#if (PAGE_SIZE < PAGE_FRAG_CACHE_MAX_SIZE) - gfp_mask = (gfp_mask & ~__GFP_DIRECT_RECLAIM) | __GFP_COMP | - __GFP_NOWARN | __GFP_NORETRY | __GFP_NOMEMALLOC; - page = alloc_pages_node(NUMA_NO_NODE, gfp_mask, - PAGE_FRAG_CACHE_MAX_ORDER); - nc->size = page ? PAGE_FRAG_CACHE_MAX_SIZE : PAGE_SIZE; -#endif - if (unlikely(!page)) - page = alloc_pages_node(NUMA_NO_NODE, gfp, 0); - - nc->va = page ? page_address(page) : NULL; - - return page; -} - -void page_frag_cache_drain(struct page_frag_cache *nc) -{ - if (!nc->va) - return; - - __page_frag_cache_drain(virt_to_head_page(nc->va), nc->pagecnt_bias); - nc->va = NULL; -} -EXPORT_SYMBOL(page_frag_cache_drain); - -void __page_frag_cache_drain(struct page *page, unsigned int count) -{ - VM_BUG_ON_PAGE(page_ref_count(page) == 0, page); - - if (page_ref_sub_and_test(page, count)) - free_unref_page(page, compound_order(page)); -} -EXPORT_SYMBOL(__page_frag_cache_drain); - -void *__page_frag_alloc_align(struct page_frag_cache *nc, - unsigned int fragsz, gfp_t gfp_mask, - unsigned int align_mask) -{ - unsigned int size = PAGE_SIZE; - struct page *page; - int offset; - - if (unlikely(!nc->va)) { -refill: - page = __page_frag_cache_refill(nc, gfp_mask); - if (!page) - return NULL; - -#if (PAGE_SIZE < PAGE_FRAG_CACHE_MAX_SIZE) - /* if size can vary use size else just use PAGE_SIZE */ - size = nc->size; -#endif - /* Even if we own the page, we do not use atomic_set(). - * This would break get_page_unless_zero() users. - */ - page_ref_add(page, PAGE_FRAG_CACHE_MAX_SIZE); - - /* reset page count bias and offset to start of new frag */ - nc->pfmemalloc = page_is_pfmemalloc(page); - nc->pagecnt_bias = PAGE_FRAG_CACHE_MAX_SIZE + 1; - nc->offset = size; - } - - offset = nc->offset - fragsz; - if (unlikely(offset < 0)) { - page = virt_to_page(nc->va); - - if (!page_ref_sub_and_test(page, nc->pagecnt_bias)) - goto refill; - - if (unlikely(nc->pfmemalloc)) { - free_unref_page(page, compound_order(page)); - goto refill; - } - -#if (PAGE_SIZE < PAGE_FRAG_CACHE_MAX_SIZE) - /* if size can vary use size else just use PAGE_SIZE */ - size = nc->size; -#endif - /* OK, page count is 0, we can safely set it */ - set_page_count(page, PAGE_FRAG_CACHE_MAX_SIZE + 1); - - /* reset page count bias and offset to start of new frag */ - nc->pagecnt_bias = PAGE_FRAG_CACHE_MAX_SIZE + 1; - offset = size - fragsz; - if (unlikely(offset < 0)) { - /* - * The caller is trying to allocate a fragment - * with fragsz > PAGE_SIZE but the cache isn't big - * enough to satisfy the request, this may - * happen in low memory conditions. - * We don't release the cache page because - * it could make memory pressure worse - * so we simply return NULL here. - */ - return NULL; - } - } - - nc->pagecnt_bias--; - offset &= align_mask; - nc->offset = offset; - - return nc->va + offset; -} -EXPORT_SYMBOL(__page_frag_alloc_align); - -/* - * Frees a page fragment allocated out of either a compound or order 0 page. - */ -void page_frag_free(void *addr) -{ - struct page *page = virt_to_head_page(addr); - - if (unlikely(put_page_testzero(page))) - free_unref_page(page, compound_order(page)); -} -EXPORT_SYMBOL(page_frag_free); - static void *make_alloc_exact(unsigned long addr, unsigned int order, size_t size) { diff --git a/mm/page_frag_cache.c b/mm/page_frag_cache.c new file mode 100644 index 000000000000..609a485cd02a --- /dev/null +++ b/mm/page_frag_cache.c @@ -0,0 +1,145 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* Page fragment allocator + * + * Page Fragment: + * An arbitrary-length arbitrary-offset area of memory which resides within a + * 0 or higher order page. Multiple fragments within that page are + * individually refcounted, in the page's reference counter. + * + * The page_frag functions provide a simple allocation framework for page + * fragments. This is used by the network stack and network device drivers to + * provide a backing region of memory for use as either an sk_buff->head, or to + * be used in the "frags" portion of skb_shared_info. + */ + +#include +#include +#include +#include +#include +#include "internal.h" + +static struct page *__page_frag_cache_refill(struct page_frag_cache *nc, + gfp_t gfp_mask) +{ + struct page *page = NULL; + gfp_t gfp = gfp_mask; + +#if (PAGE_SIZE < PAGE_FRAG_CACHE_MAX_SIZE) + gfp_mask = (gfp_mask & ~__GFP_DIRECT_RECLAIM) | __GFP_COMP | + __GFP_NOWARN | __GFP_NORETRY | __GFP_NOMEMALLOC; + page = alloc_pages_node(NUMA_NO_NODE, gfp_mask, + PAGE_FRAG_CACHE_MAX_ORDER); + nc->size = page ? PAGE_FRAG_CACHE_MAX_SIZE : PAGE_SIZE; +#endif + if (unlikely(!page)) + page = alloc_pages_node(NUMA_NO_NODE, gfp, 0); + + nc->va = page ? page_address(page) : NULL; + + return page; +} + +void page_frag_cache_drain(struct page_frag_cache *nc) +{ + if (!nc->va) + return; + + __page_frag_cache_drain(virt_to_head_page(nc->va), nc->pagecnt_bias); + nc->va = NULL; +} +EXPORT_SYMBOL(page_frag_cache_drain); + +void __page_frag_cache_drain(struct page *page, unsigned int count) +{ + VM_BUG_ON_PAGE(page_ref_count(page) == 0, page); + + if (page_ref_sub_and_test(page, count)) + free_unref_page(page, compound_order(page)); +} +EXPORT_SYMBOL(__page_frag_cache_drain); + +void *__page_frag_alloc_align(struct page_frag_cache *nc, + unsigned int fragsz, gfp_t gfp_mask, + unsigned int align_mask) +{ + unsigned int size = PAGE_SIZE; + struct page *page; + int offset; + + if (unlikely(!nc->va)) { +refill: + page = __page_frag_cache_refill(nc, gfp_mask); + if (!page) + return NULL; + +#if (PAGE_SIZE < PAGE_FRAG_CACHE_MAX_SIZE) + /* if size can vary use size else just use PAGE_SIZE */ + size = nc->size; +#endif + /* Even if we own the page, we do not use atomic_set(). + * This would break get_page_unless_zero() users. + */ + page_ref_add(page, PAGE_FRAG_CACHE_MAX_SIZE); + + /* reset page count bias and offset to start of new frag */ + nc->pfmemalloc = page_is_pfmemalloc(page); + nc->pagecnt_bias = PAGE_FRAG_CACHE_MAX_SIZE + 1; + nc->offset = size; + } + + offset = nc->offset - fragsz; + if (unlikely(offset < 0)) { + page = virt_to_page(nc->va); + + if (!page_ref_sub_and_test(page, nc->pagecnt_bias)) + goto refill; + + if (unlikely(nc->pfmemalloc)) { + free_unref_page(page, compound_order(page)); + goto refill; + } + +#if (PAGE_SIZE < PAGE_FRAG_CACHE_MAX_SIZE) + /* if size can vary use size else just use PAGE_SIZE */ + size = nc->size; +#endif + /* OK, page count is 0, we can safely set it */ + set_page_count(page, PAGE_FRAG_CACHE_MAX_SIZE + 1); + + /* reset page count bias and offset to start of new frag */ + nc->pagecnt_bias = PAGE_FRAG_CACHE_MAX_SIZE + 1; + offset = size - fragsz; + if (unlikely(offset < 0)) { + /* + * The caller is trying to allocate a fragment + * with fragsz > PAGE_SIZE but the cache isn't big + * enough to satisfy the request, this may + * happen in low memory conditions. + * We don't release the cache page because + * it could make memory pressure worse + * so we simply return NULL here. + */ + return NULL; + } + } + + nc->pagecnt_bias--; + offset &= align_mask; + nc->offset = offset; + + return nc->va + offset; +} +EXPORT_SYMBOL(__page_frag_alloc_align); + +/* + * Frees a page fragment allocated out of either a compound or order 0 page. + */ +void page_frag_free(void *addr) +{ + struct page *page = virt_to_head_page(addr); + + if (unlikely(put_page_testzero(page))) + free_unref_page(page, compound_order(page)); +} +EXPORT_SYMBOL(page_frag_free); diff --git a/tools/testing/selftests/mm/page_frag/page_frag_test.c b/tools/testing/selftests/mm/page_frag/page_frag_test.c index 912d97b99107..13c44133e009 100644 --- a/tools/testing/selftests/mm/page_frag/page_frag_test.c +++ b/tools/testing/selftests/mm/page_frag/page_frag_test.c @@ -6,12 +6,12 @@ * Copyright (C) 2024 Yunsheng Lin */ -#include #include #include #include #include #include +#include #define TEST_FAILED_PREFIX "page_frag_test failed: " -- cgit v1.2.3 From 3d18dfe69ce46f106af327736d2261d7e3ee81c0 Mon Sep 17 00:00:00 2001 From: Yunsheng Lin Date: Mon, 28 Oct 2024 19:53:39 +0800 Subject: mm: page_frag: avoid caller accessing 'page_frag_cache' directly Use appropriate frag_page API instead of caller accessing 'page_frag_cache' directly. CC: Andrew Morton CC: Linux-MM Signed-off-by: Yunsheng Lin Reviewed-by: Alexander Duyck Acked-by: Chuck Lever Link: https://patch.msgid.link/20241028115343.3405838-5-linyunsheng@huawei.com Signed-off-by: Jakub Kicinski --- drivers/vhost/net.c | 2 +- include/linux/page_frag_cache.h | 10 ++++++++++ net/core/skbuff.c | 6 +++--- net/rxrpc/conn_object.c | 4 +--- net/rxrpc/local_object.c | 4 +--- net/sunrpc/svcsock.c | 6 ++---- tools/testing/selftests/mm/page_frag/page_frag_test.c | 2 +- 7 files changed, 19 insertions(+), 15 deletions(-) (limited to 'include/linux') diff --git a/drivers/vhost/net.c b/drivers/vhost/net.c index f16279351db5..9ad37c012189 100644 --- a/drivers/vhost/net.c +++ b/drivers/vhost/net.c @@ -1325,7 +1325,7 @@ static int vhost_net_open(struct inode *inode, struct file *f) vqs[VHOST_NET_VQ_RX]); f->private_data = n; - n->pf_cache.va = NULL; + page_frag_cache_init(&n->pf_cache); return 0; } diff --git a/include/linux/page_frag_cache.h b/include/linux/page_frag_cache.h index 67ac8626ed9b..0a52f7a179c8 100644 --- a/include/linux/page_frag_cache.h +++ b/include/linux/page_frag_cache.h @@ -7,6 +7,16 @@ #include #include +static inline void page_frag_cache_init(struct page_frag_cache *nc) +{ + nc->va = NULL; +} + +static inline bool page_frag_cache_is_pfmemalloc(struct page_frag_cache *nc) +{ + return !!nc->pfmemalloc; +} + void page_frag_cache_drain(struct page_frag_cache *nc); void __page_frag_cache_drain(struct page *page, unsigned int count); void *__page_frag_alloc_align(struct page_frag_cache *nc, unsigned int fragsz, diff --git a/net/core/skbuff.c b/net/core/skbuff.c index 00afeb90c23a..6841e61a6bd0 100644 --- a/net/core/skbuff.c +++ b/net/core/skbuff.c @@ -753,14 +753,14 @@ struct sk_buff *__netdev_alloc_skb(struct net_device *dev, unsigned int len, if (in_hardirq() || irqs_disabled()) { nc = this_cpu_ptr(&netdev_alloc_cache); data = page_frag_alloc(nc, len, gfp_mask); - pfmemalloc = nc->pfmemalloc; + pfmemalloc = page_frag_cache_is_pfmemalloc(nc); } else { local_bh_disable(); local_lock_nested_bh(&napi_alloc_cache.bh_lock); nc = this_cpu_ptr(&napi_alloc_cache.page); data = page_frag_alloc(nc, len, gfp_mask); - pfmemalloc = nc->pfmemalloc; + pfmemalloc = page_frag_cache_is_pfmemalloc(nc); local_unlock_nested_bh(&napi_alloc_cache.bh_lock); local_bh_enable(); @@ -850,7 +850,7 @@ struct sk_buff *napi_alloc_skb(struct napi_struct *napi, unsigned int len) len = SKB_HEAD_ALIGN(len); data = page_frag_alloc(&nc->page, len, gfp_mask); - pfmemalloc = nc->page.pfmemalloc; + pfmemalloc = page_frag_cache_is_pfmemalloc(&nc->page); } local_unlock_nested_bh(&napi_alloc_cache.bh_lock); diff --git a/net/rxrpc/conn_object.c b/net/rxrpc/conn_object.c index 1539d315afe7..694c4df7a1a3 100644 --- a/net/rxrpc/conn_object.c +++ b/net/rxrpc/conn_object.c @@ -337,9 +337,7 @@ static void rxrpc_clean_up_connection(struct work_struct *work) */ rxrpc_purge_queue(&conn->rx_queue); - if (conn->tx_data_alloc.va) - __page_frag_cache_drain(virt_to_page(conn->tx_data_alloc.va), - conn->tx_data_alloc.pagecnt_bias); + page_frag_cache_drain(&conn->tx_data_alloc); call_rcu(&conn->rcu, rxrpc_rcu_free_connection); } diff --git a/net/rxrpc/local_object.c b/net/rxrpc/local_object.c index f9623ace2201..2792d2304605 100644 --- a/net/rxrpc/local_object.c +++ b/net/rxrpc/local_object.c @@ -452,9 +452,7 @@ void rxrpc_destroy_local(struct rxrpc_local *local) #endif rxrpc_purge_queue(&local->rx_queue); rxrpc_purge_client_connections(local); - if (local->tx_alloc.va) - __page_frag_cache_drain(virt_to_page(local->tx_alloc.va), - local->tx_alloc.pagecnt_bias); + page_frag_cache_drain(&local->tx_alloc); } /* diff --git a/net/sunrpc/svcsock.c b/net/sunrpc/svcsock.c index 825ec5357691..b785425c3315 100644 --- a/net/sunrpc/svcsock.c +++ b/net/sunrpc/svcsock.c @@ -1608,7 +1608,6 @@ static void svc_tcp_sock_detach(struct svc_xprt *xprt) static void svc_sock_free(struct svc_xprt *xprt) { struct svc_sock *svsk = container_of(xprt, struct svc_sock, sk_xprt); - struct page_frag_cache *pfc = &svsk->sk_frag_cache; struct socket *sock = svsk->sk_sock; trace_svcsock_free(svsk, sock); @@ -1618,8 +1617,7 @@ static void svc_sock_free(struct svc_xprt *xprt) sockfd_put(sock); else sock_release(sock); - if (pfc->va) - __page_frag_cache_drain(virt_to_head_page(pfc->va), - pfc->pagecnt_bias); + + page_frag_cache_drain(&svsk->sk_frag_cache); kfree(svsk); } diff --git a/tools/testing/selftests/mm/page_frag/page_frag_test.c b/tools/testing/selftests/mm/page_frag/page_frag_test.c index 13c44133e009..e806c1866e36 100644 --- a/tools/testing/selftests/mm/page_frag/page_frag_test.c +++ b/tools/testing/selftests/mm/page_frag/page_frag_test.c @@ -126,7 +126,7 @@ static int __init page_frag_test_init(void) u64 duration; int ret; - test_nc.va = NULL; + page_frag_cache_init(&test_nc); atomic_set(&nthreads, 2); init_completion(&wait); -- cgit v1.2.3 From 0c3ce2f50261cd2f654d931eeb933c370a3a7d7a Mon Sep 17 00:00:00 2001 From: Yunsheng Lin Date: Mon, 28 Oct 2024 19:53:41 +0800 Subject: mm: page_frag: reuse existing space for 'size' and 'pfmemalloc' Currently there is one 'struct page_frag' for every 'struct sock' and 'struct task_struct', we are about to replace the 'struct page_frag' with 'struct page_frag_cache' for them. Before begin the replacing, we need to ensure the size of 'struct page_frag_cache' is not bigger than the size of 'struct page_frag', as there may be tens of thousands of 'struct sock' and 'struct task_struct' instances in the system. By or'ing the page order & pfmemalloc with lower bits of 'va' instead of using 'u16' or 'u32' for page size and 'u8' for pfmemalloc, we are able to avoid 3 or 5 bytes space waste. And page address & pfmemalloc & order is unchanged for the same page in the same 'page_frag_cache' instance, it makes sense to fit them together. After this patch, the size of 'struct page_frag_cache' should be the same as the size of 'struct page_frag'. CC: Andrew Morton CC: Linux-MM Signed-off-by: Yunsheng Lin Reviewed-by: Alexander Duyck Link: https://patch.msgid.link/20241028115343.3405838-7-linyunsheng@huawei.com Signed-off-by: Jakub Kicinski --- include/linux/mm_types_task.h | 19 ++++++----- include/linux/page_frag_cache.h | 24 ++++++++++++-- mm/page_frag_cache.c | 70 ++++++++++++++++++++++++++++------------- 3 files changed, 81 insertions(+), 32 deletions(-) (limited to 'include/linux') diff --git a/include/linux/mm_types_task.h b/include/linux/mm_types_task.h index 0ac6daebdd5c..a82aa80c0ba4 100644 --- a/include/linux/mm_types_task.h +++ b/include/linux/mm_types_task.h @@ -47,18 +47,21 @@ struct page_frag { #define PAGE_FRAG_CACHE_MAX_SIZE __ALIGN_MASK(32768, ~PAGE_MASK) #define PAGE_FRAG_CACHE_MAX_ORDER get_order(PAGE_FRAG_CACHE_MAX_SIZE) struct page_frag_cache { - void *va; -#if (PAGE_SIZE < PAGE_FRAG_CACHE_MAX_SIZE) + /* encoded_page consists of the virtual address, pfmemalloc bit and + * order of a page. + */ + unsigned long encoded_page; + + /* we maintain a pagecount bias, so that we dont dirty cache line + * containing page->_refcount every time we allocate a fragment. + */ +#if (PAGE_SIZE < PAGE_FRAG_CACHE_MAX_SIZE) && (BITS_PER_LONG <= 32) __u16 offset; - __u16 size; + __u16 pagecnt_bias; #else __u32 offset; + __u32 pagecnt_bias; #endif - /* we maintain a pagecount bias, so that we dont dirty cache line - * containing page->_refcount every time we allocate a fragment. - */ - unsigned int pagecnt_bias; - bool pfmemalloc; }; /* Track pages that require TLB flushes */ diff --git a/include/linux/page_frag_cache.h b/include/linux/page_frag_cache.h index 0a52f7a179c8..41a91df82631 100644 --- a/include/linux/page_frag_cache.h +++ b/include/linux/page_frag_cache.h @@ -3,18 +3,38 @@ #ifndef _LINUX_PAGE_FRAG_CACHE_H #define _LINUX_PAGE_FRAG_CACHE_H +#include #include #include #include +#if (PAGE_SIZE < PAGE_FRAG_CACHE_MAX_SIZE) +/* Use a full byte here to enable assembler optimization as the shift + * operation is usually expecting a byte. + */ +#define PAGE_FRAG_CACHE_ORDER_MASK GENMASK(7, 0) +#else +/* Compiler should be able to figure out we don't read things as any value + * ANDed with 0 is 0. + */ +#define PAGE_FRAG_CACHE_ORDER_MASK 0 +#endif + +#define PAGE_FRAG_CACHE_PFMEMALLOC_BIT (PAGE_FRAG_CACHE_ORDER_MASK + 1) + +static inline bool encoded_page_decode_pfmemalloc(unsigned long encoded_page) +{ + return !!(encoded_page & PAGE_FRAG_CACHE_PFMEMALLOC_BIT); +} + static inline void page_frag_cache_init(struct page_frag_cache *nc) { - nc->va = NULL; + nc->encoded_page = 0; } static inline bool page_frag_cache_is_pfmemalloc(struct page_frag_cache *nc) { - return !!nc->pfmemalloc; + return encoded_page_decode_pfmemalloc(nc->encoded_page); } void page_frag_cache_drain(struct page_frag_cache *nc); diff --git a/mm/page_frag_cache.c b/mm/page_frag_cache.c index 4c8e04379cb3..a36fd09bf275 100644 --- a/mm/page_frag_cache.c +++ b/mm/page_frag_cache.c @@ -12,6 +12,7 @@ * be used in the "frags" portion of skb_shared_info. */ +#include #include #include #include @@ -19,9 +20,36 @@ #include #include "internal.h" +static unsigned long encoded_page_create(struct page *page, unsigned int order, + bool pfmemalloc) +{ + BUILD_BUG_ON(PAGE_FRAG_CACHE_MAX_ORDER > PAGE_FRAG_CACHE_ORDER_MASK); + BUILD_BUG_ON(PAGE_FRAG_CACHE_PFMEMALLOC_BIT >= PAGE_SIZE); + + return (unsigned long)page_address(page) | + (order & PAGE_FRAG_CACHE_ORDER_MASK) | + ((unsigned long)pfmemalloc * PAGE_FRAG_CACHE_PFMEMALLOC_BIT); +} + +static unsigned long encoded_page_decode_order(unsigned long encoded_page) +{ + return encoded_page & PAGE_FRAG_CACHE_ORDER_MASK; +} + +static void *encoded_page_decode_virt(unsigned long encoded_page) +{ + return (void *)(encoded_page & PAGE_MASK); +} + +static struct page *encoded_page_decode_page(unsigned long encoded_page) +{ + return virt_to_page((void *)encoded_page); +} + static struct page *__page_frag_cache_refill(struct page_frag_cache *nc, gfp_t gfp_mask) { + unsigned long order = PAGE_FRAG_CACHE_MAX_ORDER; struct page *page = NULL; gfp_t gfp = gfp_mask; @@ -30,23 +58,26 @@ static struct page *__page_frag_cache_refill(struct page_frag_cache *nc, __GFP_NOWARN | __GFP_NORETRY | __GFP_NOMEMALLOC; page = alloc_pages_node(NUMA_NO_NODE, gfp_mask, PAGE_FRAG_CACHE_MAX_ORDER); - nc->size = page ? PAGE_FRAG_CACHE_MAX_SIZE : PAGE_SIZE; #endif - if (unlikely(!page)) + if (unlikely(!page)) { page = alloc_pages_node(NUMA_NO_NODE, gfp, 0); + order = 0; + } - nc->va = page ? page_address(page) : NULL; + nc->encoded_page = page ? + encoded_page_create(page, order, page_is_pfmemalloc(page)) : 0; return page; } void page_frag_cache_drain(struct page_frag_cache *nc) { - if (!nc->va) + if (!nc->encoded_page) return; - __page_frag_cache_drain(virt_to_head_page(nc->va), nc->pagecnt_bias); - nc->va = NULL; + __page_frag_cache_drain(encoded_page_decode_page(nc->encoded_page), + nc->pagecnt_bias); + nc->encoded_page = 0; } EXPORT_SYMBOL(page_frag_cache_drain); @@ -63,35 +94,29 @@ void *__page_frag_alloc_align(struct page_frag_cache *nc, unsigned int fragsz, gfp_t gfp_mask, unsigned int align_mask) { -#if (PAGE_SIZE < PAGE_FRAG_CACHE_MAX_SIZE) - unsigned int size = nc->size; -#else - unsigned int size = PAGE_SIZE; -#endif - unsigned int offset; + unsigned long encoded_page = nc->encoded_page; + unsigned int size, offset; struct page *page; - if (unlikely(!nc->va)) { + if (unlikely(!encoded_page)) { refill: page = __page_frag_cache_refill(nc, gfp_mask); if (!page) return NULL; -#if (PAGE_SIZE < PAGE_FRAG_CACHE_MAX_SIZE) - /* if size can vary use size else just use PAGE_SIZE */ - size = nc->size; -#endif + encoded_page = nc->encoded_page; + /* Even if we own the page, we do not use atomic_set(). * This would break get_page_unless_zero() users. */ page_ref_add(page, PAGE_FRAG_CACHE_MAX_SIZE); /* reset page count bias and offset to start of new frag */ - nc->pfmemalloc = page_is_pfmemalloc(page); nc->pagecnt_bias = PAGE_FRAG_CACHE_MAX_SIZE + 1; nc->offset = 0; } + size = PAGE_SIZE << encoded_page_decode_order(encoded_page); offset = __ALIGN_KERNEL_MASK(nc->offset, ~align_mask); if (unlikely(offset + fragsz > size)) { if (unlikely(fragsz > PAGE_SIZE)) { @@ -107,13 +132,14 @@ refill: return NULL; } - page = virt_to_page(nc->va); + page = encoded_page_decode_page(encoded_page); if (!page_ref_sub_and_test(page, nc->pagecnt_bias)) goto refill; - if (unlikely(nc->pfmemalloc)) { - free_unref_page(page, compound_order(page)); + if (unlikely(encoded_page_decode_pfmemalloc(encoded_page))) { + free_unref_page(page, + encoded_page_decode_order(encoded_page)); goto refill; } @@ -128,7 +154,7 @@ refill: nc->pagecnt_bias--; nc->offset = offset + fragsz; - return nc->va + offset; + return encoded_page_decode_virt(encoded_page) + offset; } EXPORT_SYMBOL(__page_frag_alloc_align); -- cgit v1.2.3 From 5dc51ec86df6e2214d8398079c1e31736593ab53 Mon Sep 17 00:00:00 2001 From: Martin Karsten Date: Sat, 9 Nov 2024 05:02:31 +0000 Subject: net: Add napi_struct parameter irq_suspend_timeout Add a per-NAPI IRQ suspension parameter, which can be get/set with netdev-genl. This patch doesn't change any behavior but prepares the code for other changes in the following commits which use irq_suspend_timeout as a timeout for IRQ suspension. Signed-off-by: Martin Karsten Co-developed-by: Joe Damato Signed-off-by: Joe Damato Tested-by: Joe Damato Tested-by: Martin Karsten Acked-by: Stanislav Fomichev Reviewed-by: Sridhar Samudrala Link: https://patch.msgid.link/20241109050245.191288-2-jdamato@fastly.com Signed-off-by: Jakub Kicinski --- Documentation/netlink/specs/netdev.yaml | 7 +++++++ include/linux/netdevice.h | 2 ++ include/uapi/linux/netdev.h | 1 + net/core/dev.c | 2 ++ net/core/dev.h | 25 +++++++++++++++++++++++++ net/core/netdev-genl-gen.c | 5 +++-- net/core/netdev-genl.c | 12 ++++++++++++ tools/include/uapi/linux/netdev.h | 1 + 8 files changed, 53 insertions(+), 2 deletions(-) (limited to 'include/linux') diff --git a/Documentation/netlink/specs/netdev.yaml b/Documentation/netlink/specs/netdev.yaml index f9cb97d6106c..cbb544bd6c84 100644 --- a/Documentation/netlink/specs/netdev.yaml +++ b/Documentation/netlink/specs/netdev.yaml @@ -263,6 +263,11 @@ attribute-sets: the end of a NAPI cycle. This may add receive latency in exchange for reducing the number of frames processed by the network stack. type: uint + - + name: irq-suspend-timeout + doc: The timeout, in nanoseconds, of how long to suspend irq + processing, if event polling finds events + type: uint - name: queue attributes: @@ -653,6 +658,7 @@ operations: - pid - defer-hard-irqs - gro-flush-timeout + - irq-suspend-timeout dump: request: attributes: @@ -704,6 +710,7 @@ operations: - id - defer-hard-irqs - gro-flush-timeout + - irq-suspend-timeout kernel-family: headers: [ "linux/list.h"] diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index df4483598628..0aae346d919e 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -348,6 +348,7 @@ struct gro_list { */ struct napi_config { u64 gro_flush_timeout; + u64 irq_suspend_timeout; u32 defer_hard_irqs; unsigned int napi_id; }; @@ -384,6 +385,7 @@ struct napi_struct { struct hrtimer timer; struct task_struct *thread; unsigned long gro_flush_timeout; + unsigned long irq_suspend_timeout; u32 defer_hard_irqs; /* control-path-only fields follow */ struct list_head dev_list; diff --git a/include/uapi/linux/netdev.h b/include/uapi/linux/netdev.h index e3ebb49f60d2..e4be227d3ad6 100644 --- a/include/uapi/linux/netdev.h +++ b/include/uapi/linux/netdev.h @@ -124,6 +124,7 @@ enum { NETDEV_A_NAPI_PID, NETDEV_A_NAPI_DEFER_HARD_IRQS, NETDEV_A_NAPI_GRO_FLUSH_TIMEOUT, + NETDEV_A_NAPI_IRQ_SUSPEND_TIMEOUT, __NETDEV_A_NAPI_MAX, NETDEV_A_NAPI_MAX = (__NETDEV_A_NAPI_MAX - 1) diff --git a/net/core/dev.c b/net/core/dev.c index 6a31152e4606..4d910872963f 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -6666,6 +6666,7 @@ static void napi_restore_config(struct napi_struct *n) { n->defer_hard_irqs = n->config->defer_hard_irqs; n->gro_flush_timeout = n->config->gro_flush_timeout; + n->irq_suspend_timeout = n->config->irq_suspend_timeout; /* a NAPI ID might be stored in the config, if so use it. if not, use * napi_hash_add to generate one for us. It will be saved to the config * in napi_disable. @@ -6680,6 +6681,7 @@ static void napi_save_config(struct napi_struct *n) { n->config->defer_hard_irqs = n->defer_hard_irqs; n->config->gro_flush_timeout = n->gro_flush_timeout; + n->config->irq_suspend_timeout = n->irq_suspend_timeout; n->config->napi_id = n->napi_id; napi_hash_del(n); } diff --git a/net/core/dev.h b/net/core/dev.h index 7881bced70a9..d043dee25a68 100644 --- a/net/core/dev.h +++ b/net/core/dev.h @@ -236,6 +236,31 @@ static inline void netdev_set_gro_flush_timeout(struct net_device *netdev, netdev->napi_config[i].gro_flush_timeout = timeout; } +/** + * napi_get_irq_suspend_timeout - get the irq_suspend_timeout + * @n: napi struct to get the irq_suspend_timeout from + * + * Return: the per-NAPI value of the irq_suspend_timeout field. + */ +static inline unsigned long +napi_get_irq_suspend_timeout(const struct napi_struct *n) +{ + return READ_ONCE(n->irq_suspend_timeout); +} + +/** + * napi_set_irq_suspend_timeout - set the irq_suspend_timeout for a napi + * @n: napi struct to set the irq_suspend_timeout + * @timeout: timeout value to set + * + * napi_set_irq_suspend_timeout sets the per-NAPI irq_suspend_timeout + */ +static inline void napi_set_irq_suspend_timeout(struct napi_struct *n, + unsigned long timeout) +{ + WRITE_ONCE(n->irq_suspend_timeout, timeout); +} + int rps_cpumask_housekeeping(struct cpumask *mask); #if defined(CONFIG_DEBUG_NET) && defined(CONFIG_BPF_SYSCALL) diff --git a/net/core/netdev-genl-gen.c b/net/core/netdev-genl-gen.c index 21de7e10be16..a89cbd8d87c3 100644 --- a/net/core/netdev-genl-gen.c +++ b/net/core/netdev-genl-gen.c @@ -92,10 +92,11 @@ static const struct nla_policy netdev_bind_rx_nl_policy[NETDEV_A_DMABUF_FD + 1] }; /* NETDEV_CMD_NAPI_SET - do */ -static const struct nla_policy netdev_napi_set_nl_policy[NETDEV_A_NAPI_GRO_FLUSH_TIMEOUT + 1] = { +static const struct nla_policy netdev_napi_set_nl_policy[NETDEV_A_NAPI_IRQ_SUSPEND_TIMEOUT + 1] = { [NETDEV_A_NAPI_ID] = { .type = NLA_U32, }, [NETDEV_A_NAPI_DEFER_HARD_IRQS] = NLA_POLICY_FULL_RANGE(NLA_U32, &netdev_a_napi_defer_hard_irqs_range), [NETDEV_A_NAPI_GRO_FLUSH_TIMEOUT] = { .type = NLA_UINT, }, + [NETDEV_A_NAPI_IRQ_SUSPEND_TIMEOUT] = { .type = NLA_UINT, }, }; /* Ops table for netdev */ @@ -186,7 +187,7 @@ static const struct genl_split_ops netdev_nl_ops[] = { .cmd = NETDEV_CMD_NAPI_SET, .doit = netdev_nl_napi_set_doit, .policy = netdev_napi_set_nl_policy, - .maxattr = NETDEV_A_NAPI_GRO_FLUSH_TIMEOUT, + .maxattr = NETDEV_A_NAPI_IRQ_SUSPEND_TIMEOUT, .flags = GENL_ADMIN_PERM | GENL_CMD_CAP_DO, }, }; diff --git a/net/core/netdev-genl.c b/net/core/netdev-genl.c index b49c3b4e5fbe..765ce7c9d73b 100644 --- a/net/core/netdev-genl.c +++ b/net/core/netdev-genl.c @@ -161,6 +161,7 @@ static int netdev_nl_napi_fill_one(struct sk_buff *rsp, struct napi_struct *napi, const struct genl_info *info) { + unsigned long irq_suspend_timeout; unsigned long gro_flush_timeout; u32 napi_defer_hard_irqs; void *hdr; @@ -196,6 +197,11 @@ netdev_nl_napi_fill_one(struct sk_buff *rsp, struct napi_struct *napi, napi_defer_hard_irqs)) goto nla_put_failure; + irq_suspend_timeout = napi_get_irq_suspend_timeout(napi); + if (nla_put_uint(rsp, NETDEV_A_NAPI_IRQ_SUSPEND_TIMEOUT, + irq_suspend_timeout)) + goto nla_put_failure; + gro_flush_timeout = napi_get_gro_flush_timeout(napi); if (nla_put_uint(rsp, NETDEV_A_NAPI_GRO_FLUSH_TIMEOUT, gro_flush_timeout)) @@ -306,6 +312,7 @@ int netdev_nl_napi_get_dumpit(struct sk_buff *skb, struct netlink_callback *cb) static int netdev_nl_napi_set_config(struct napi_struct *napi, struct genl_info *info) { + u64 irq_suspend_timeout = 0; u64 gro_flush_timeout = 0; u32 defer = 0; @@ -314,6 +321,11 @@ netdev_nl_napi_set_config(struct napi_struct *napi, struct genl_info *info) napi_set_defer_hard_irqs(napi, defer); } + if (info->attrs[NETDEV_A_NAPI_IRQ_SUSPEND_TIMEOUT]) { + irq_suspend_timeout = nla_get_uint(info->attrs[NETDEV_A_NAPI_IRQ_SUSPEND_TIMEOUT]); + napi_set_irq_suspend_timeout(napi, irq_suspend_timeout); + } + if (info->attrs[NETDEV_A_NAPI_GRO_FLUSH_TIMEOUT]) { gro_flush_timeout = nla_get_uint(info->attrs[NETDEV_A_NAPI_GRO_FLUSH_TIMEOUT]); napi_set_gro_flush_timeout(napi, gro_flush_timeout); diff --git a/tools/include/uapi/linux/netdev.h b/tools/include/uapi/linux/netdev.h index e3ebb49f60d2..e4be227d3ad6 100644 --- a/tools/include/uapi/linux/netdev.h +++ b/tools/include/uapi/linux/netdev.h @@ -124,6 +124,7 @@ enum { NETDEV_A_NAPI_PID, NETDEV_A_NAPI_DEFER_HARD_IRQS, NETDEV_A_NAPI_GRO_FLUSH_TIMEOUT, + NETDEV_A_NAPI_IRQ_SUSPEND_TIMEOUT, __NETDEV_A_NAPI_MAX, NETDEV_A_NAPI_MAX = (__NETDEV_A_NAPI_MAX - 1) -- cgit v1.2.3 From 12079a59ce52e72a342c49cfacf0281213fd6f32 Mon Sep 17 00:00:00 2001 From: Breno Leitao Date: Thu, 7 Nov 2024 08:11:44 -0800 Subject: net: Implement fault injection forcing skb reallocation Introduce a fault injection mechanism to force skb reallocation. The primary goal is to catch bugs related to pointer invalidation after potential skb reallocation. The fault injection mechanism aims to identify scenarios where callers retain pointers to various headers in the skb but fail to reload these pointers after calling a function that may reallocate the data. This type of bug can lead to memory corruption or crashes if the old, now-invalid pointers are used. By forcing reallocation through fault injection, we can stress-test code paths and ensure proper pointer management after potential skb reallocations. Add a hook for fault injection in the following functions: * pskb_trim_rcsum() * pskb_may_pull_reason() * pskb_trim() As the other fault injection mechanism, protect it under a debug Kconfig called CONFIG_FAIL_SKB_REALLOC. This patch was *heavily* inspired by Jakub's proposal from: https://lore.kernel.org/all/20240719174140.47a868e6@kernel.org/ CC: Akinobu Mita Suggested-by: Jakub Kicinski Signed-off-by: Breno Leitao Reviewed-by: Akinobu Mita Acked-by: Paolo Abeni Acked-by: Guillaume Nault Link: https://patch.msgid.link/20241107-fault_v6-v6-1-1b82cb6ecacd@debian.org Signed-off-by: Paolo Abeni --- Documentation/admin-guide/kernel-parameters.txt | 1 + Documentation/fault-injection/fault-injection.rst | 40 ++++++++ include/linux/skbuff.h | 9 ++ lib/Kconfig.debug | 10 ++ net/core/Makefile | 1 + net/core/skb_fault_injection.c | 106 ++++++++++++++++++++++ 6 files changed, 167 insertions(+) create mode 100644 net/core/skb_fault_injection.c (limited to 'include/linux') diff --git a/Documentation/admin-guide/kernel-parameters.txt b/Documentation/admin-guide/kernel-parameters.txt index 1518343bbe22..2fb830453dcc 100644 --- a/Documentation/admin-guide/kernel-parameters.txt +++ b/Documentation/admin-guide/kernel-parameters.txt @@ -1546,6 +1546,7 @@ failslab= fail_usercopy= fail_page_alloc= + fail_skb_realloc= fail_make_request=[KNL] General fault injection mechanism. Format: ,,, diff --git a/Documentation/fault-injection/fault-injection.rst b/Documentation/fault-injection/fault-injection.rst index 8b8aeea71c68..1c14ba08fbfc 100644 --- a/Documentation/fault-injection/fault-injection.rst +++ b/Documentation/fault-injection/fault-injection.rst @@ -45,6 +45,32 @@ Available fault injection capabilities ALLOW_ERROR_INJECTION() macro, by setting debugfs entries under /sys/kernel/debug/fail_function. No boot option supported. +- fail_skb_realloc + + inject skb (socket buffer) reallocation events into the network path. The + primary goal is to identify and prevent issues related to pointer + mismanagement in the network subsystem. By forcing skb reallocation at + strategic points, this feature creates scenarios where existing pointers to + skb headers become invalid. + + When the fault is injected and the reallocation is triggered, cached pointers + to skb headers and data no longer reference valid memory locations. This + deliberate invalidation helps expose code paths where proper pointer updating + is neglected after a reallocation event. + + By creating these controlled fault scenarios, the system can catch instances + where stale pointers are used, potentially leading to memory corruption or + system instability. + + To select the interface to act on, write the network name to + /sys/kernel/debug/fail_skb_realloc/devname. + If this field is left empty (which is the default value), skb reallocation + will be forced on all network interfaces. + + The effectiveness of this fault detection is enhanced when KASAN is + enabled, as it helps identify invalid memory references and use-after-free + (UAF) issues. + - NVMe fault injection inject NVMe status code and retry flag on devices permitted by setting @@ -216,6 +242,19 @@ configuration of fault-injection capabilities. use a negative errno, you better use 'printf' instead of 'echo', e.g.: $ printf %#x -12 > retval +- /sys/kernel/debug/fail_skb_realloc/devname: + + Specifies the network interface on which to force SKB reallocation. If + left empty, SKB reallocation will be applied to all network interfaces. + + Example usage:: + + # Force skb reallocation on eth0 + echo "eth0" > /sys/kernel/debug/fail_skb_realloc/devname + + # Clear the selection and force skb reallocation on all interfaces + echo "" > /sys/kernel/debug/fail_skb_realloc/devname + Boot option ^^^^^^^^^^^ @@ -227,6 +266,7 @@ use the boot option:: fail_usercopy= fail_make_request= fail_futex= + fail_skb_realloc= mmc_core.fail_request=,,, proc entries diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h index 60535c706851..58009fa66102 100644 --- a/include/linux/skbuff.h +++ b/include/linux/skbuff.h @@ -2682,6 +2682,12 @@ static inline void skb_assert_len(struct sk_buff *skb) #endif /* CONFIG_DEBUG_NET */ } +#if defined(CONFIG_FAIL_SKB_REALLOC) +void skb_might_realloc(struct sk_buff *skb); +#else +static inline void skb_might_realloc(struct sk_buff *skb) {} +#endif + /* * Add data to an sk_buff */ @@ -2782,6 +2788,7 @@ static inline enum skb_drop_reason pskb_may_pull_reason(struct sk_buff *skb, unsigned int len) { DEBUG_NET_WARN_ON_ONCE(len > INT_MAX); + skb_might_realloc(skb); if (likely(len <= skb_headlen(skb))) return SKB_NOT_DROPPED_YET; @@ -3240,6 +3247,7 @@ static inline int __pskb_trim(struct sk_buff *skb, unsigned int len) static inline int pskb_trim(struct sk_buff *skb, unsigned int len) { + skb_might_realloc(skb); return (len < skb->len) ? __pskb_trim(skb, len) : 0; } @@ -3994,6 +4002,7 @@ int pskb_trim_rcsum_slow(struct sk_buff *skb, unsigned int len); static inline int pskb_trim_rcsum(struct sk_buff *skb, unsigned int len) { + skb_might_realloc(skb); if (likely(len >= skb->len)) return 0; return pskb_trim_rcsum_slow(skb, len); diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug index 7312ae7c3cc5..67b669d2e70e 100644 --- a/lib/Kconfig.debug +++ b/lib/Kconfig.debug @@ -2115,6 +2115,16 @@ config FAIL_SUNRPC Provide fault-injection capability for SunRPC and its consumers. +config FAIL_SKB_REALLOC + bool "Fault-injection capability forcing skb to reallocate" + depends on FAULT_INJECTION_DEBUG_FS + help + Provide fault-injection capability that forces the skb to be + reallocated, catching possible invalid pointers to the skb. + + For more information, check + Documentation/dev-tools/fault-injection/fault-injection.rst + config FAULT_INJECTION_CONFIGFS bool "Configfs interface for fault-injection capabilities" depends on FAULT_INJECTION diff --git a/net/core/Makefile b/net/core/Makefile index 5a72a87ee0f1..d9326600e289 100644 --- a/net/core/Makefile +++ b/net/core/Makefile @@ -46,3 +46,4 @@ obj-$(CONFIG_OF) += of_net.o obj-$(CONFIG_NET_TEST) += net_test.o obj-$(CONFIG_NET_DEVMEM) += devmem.o obj-$(CONFIG_DEBUG_NET_SMALL_RTNL) += rtnl_net_debug.o +obj-$(CONFIG_FAIL_SKB_REALLOC) += skb_fault_injection.o diff --git a/net/core/skb_fault_injection.c b/net/core/skb_fault_injection.c new file mode 100644 index 000000000000..4235db6bdfad --- /dev/null +++ b/net/core/skb_fault_injection.c @@ -0,0 +1,106 @@ +// SPDX-License-Identifier: GPL-2.0-only + +#include +#include +#include +#include + +static struct { + struct fault_attr attr; + char devname[IFNAMSIZ]; + bool filtered; +} skb_realloc = { + .attr = FAULT_ATTR_INITIALIZER, + .filtered = false, +}; + +static bool should_fail_net_realloc_skb(struct sk_buff *skb) +{ + struct net_device *net = skb->dev; + + if (skb_realloc.filtered && + strncmp(net->name, skb_realloc.devname, IFNAMSIZ)) + /* device name filter set, but names do not match */ + return false; + + if (!should_fail(&skb_realloc.attr, 1)) + return false; + + return true; +} +ALLOW_ERROR_INJECTION(should_fail_net_realloc_skb, TRUE); + +void skb_might_realloc(struct sk_buff *skb) +{ + if (!should_fail_net_realloc_skb(skb)) + return; + + pskb_expand_head(skb, 0, 0, GFP_ATOMIC); +} +EXPORT_SYMBOL(skb_might_realloc); + +static int __init fail_skb_realloc_setup(char *str) +{ + return setup_fault_attr(&skb_realloc.attr, str); +} +__setup("fail_skb_realloc=", fail_skb_realloc_setup); + +static void reset_settings(void) +{ + skb_realloc.filtered = false; + memset(&skb_realloc.devname, 0, IFNAMSIZ); +} + +static ssize_t devname_write(struct file *file, const char __user *buffer, + size_t count, loff_t *ppos) +{ + ssize_t ret; + + reset_settings(); + ret = simple_write_to_buffer(&skb_realloc.devname, IFNAMSIZ, + ppos, buffer, count); + if (ret < 0) + return ret; + + skb_realloc.devname[IFNAMSIZ - 1] = '\0'; + /* Remove a possible \n at the end of devname */ + strim(skb_realloc.devname); + + if (strnlen(skb_realloc.devname, IFNAMSIZ)) + skb_realloc.filtered = true; + + return count; +} + +static ssize_t devname_read(struct file *file, + char __user *buffer, + size_t size, loff_t *ppos) +{ + if (!skb_realloc.filtered) + return 0; + + return simple_read_from_buffer(buffer, size, ppos, &skb_realloc.devname, + strlen(skb_realloc.devname)); +} + +static const struct file_operations devname_ops = { + .write = devname_write, + .read = devname_read, +}; + +static int __init fail_skb_realloc_debugfs(void) +{ + umode_t mode = S_IFREG | 0600; + struct dentry *dir; + + dir = fault_create_debugfs_attr("fail_skb_realloc", NULL, + &skb_realloc.attr); + if (IS_ERR(dir)) + return PTR_ERR(dir); + + debugfs_create_file("devname", mode, dir, NULL, &devname_ops); + + return 0; +} + +late_initcall(fail_skb_realloc_debugfs); -- cgit v1.2.3 From 406c5548c661df0bff6bb6ee79bf9d49faf23e31 Mon Sep 17 00:00:00 2001 From: MeiChia Chiu Date: Tue, 12 Nov 2024 16:38:46 +0800 Subject: wifi: mac80211: Support EHT 1024 aggregation size in TX Support EHT 1024 aggregation size in TX The 1024 agg size for RX is supported but not for TX. This patch adds this support and refactors common parsing logics for addbaext in both process_addba_resp and process_addba_req into a function. Reviewed-by: Shayne Chen Reviewed-by: Money Wang Co-developed-by: Peter Chiu Signed-off-by: Peter Chiu Signed-off-by: MeiChia Chiu Link: https://patch.msgid.link/20241112083846.32063-1-MeiChia.Chiu@mediatek.com [pass elems/len instead of mgmt/len/is_req] Signed-off-by: Johannes Berg --- include/linux/ieee80211.h | 2 + net/mac80211/agg-rx.c | 94 ++++++++++++++++++++++++++++------------------ net/mac80211/agg-tx.c | 31 ++++++++++----- net/mac80211/ht.c | 2 +- net/mac80211/ieee80211_i.h | 9 ++++- 5 files changed, 90 insertions(+), 48 deletions(-) (limited to 'include/linux') diff --git a/include/linux/ieee80211.h b/include/linux/ieee80211.h index 456bca45ff05..05dedc45505c 100644 --- a/include/linux/ieee80211.h +++ b/include/linux/ieee80211.h @@ -1445,6 +1445,8 @@ struct ieee80211_mgmt { __le16 status; __le16 capab; __le16 timeout; + /* followed by BA Extension */ + u8 variable[]; } __packed addba_resp; struct{ u8 action_code; diff --git a/net/mac80211/agg-rx.c b/net/mac80211/agg-rx.c index fe7eab4b681b..f3fbe5a4395e 100644 --- a/net/mac80211/agg-rx.c +++ b/net/mac80211/agg-rx.c @@ -170,28 +170,63 @@ static void sta_rx_agg_reorder_timer_expired(struct timer_list *t) rcu_read_unlock(); } -static void ieee80211_add_addbaext(struct ieee80211_sub_if_data *sdata, - struct sk_buff *skb, - const struct ieee80211_addba_ext_ie *req, - u16 buf_size) +void ieee80211_add_addbaext(struct sk_buff *skb, + const u8 req_addba_ext_data, + u16 buf_size) { - struct ieee80211_addba_ext_ie *resp; + struct ieee80211_addba_ext_ie *addba_ext; u8 *pos; pos = skb_put_zero(skb, 2 + sizeof(struct ieee80211_addba_ext_ie)); *pos++ = WLAN_EID_ADDBA_EXT; *pos++ = sizeof(struct ieee80211_addba_ext_ie); - resp = (struct ieee80211_addba_ext_ie *)pos; - resp->data = req->data & IEEE80211_ADDBA_EXT_NO_FRAG; + addba_ext = (struct ieee80211_addba_ext_ie *)pos; - resp->data |= u8_encode_bits(buf_size >> IEEE80211_ADDBA_EXT_BUF_SIZE_SHIFT, - IEEE80211_ADDBA_EXT_BUF_SIZE_MASK); + addba_ext->data = IEEE80211_ADDBA_EXT_NO_FRAG; + if (req_addba_ext_data) + addba_ext->data &= req_addba_ext_data; + + addba_ext->data |= + u8_encode_bits(buf_size >> IEEE80211_ADDBA_EXT_BUF_SIZE_SHIFT, + IEEE80211_ADDBA_EXT_BUF_SIZE_MASK); +} + +u8 ieee80211_retrieve_addba_ext_data(struct sta_info *sta, + const void *elem_data, ssize_t elem_len, + u16 *buf_size) +{ + struct ieee802_11_elems *elems; + u8 buf_size_1k, data = 0; + + if (!sta->sta.deflink.he_cap.has_he) + return 0; + + if (elem_len <= 0) + return 0; + + elems = ieee802_11_parse_elems(elem_data, elem_len, true, NULL); + + if (elems && !elems->parse_error && elems->addba_ext_ie) { + data = elems->addba_ext_ie->data; + + if (!sta->sta.deflink.eht_cap.has_eht || !buf_size) + goto free; + + buf_size_1k = u8_get_bits(elems->addba_ext_ie->data, + IEEE80211_ADDBA_EXT_BUF_SIZE_MASK); + *buf_size |= (u16)buf_size_1k << + IEEE80211_ADDBA_EXT_BUF_SIZE_SHIFT; + } +free: + kfree(elems); + + return data; } static void ieee80211_send_addba_resp(struct sta_info *sta, u8 *da, u16 tid, u8 dialog_token, u16 status, u16 policy, u16 buf_size, u16 timeout, - const struct ieee80211_addba_ext_ie *addbaext) + const u8 req_addba_ext_data) { struct ieee80211_sub_if_data *sdata = sta->sdata; struct ieee80211_local *local = sdata->local; @@ -223,8 +258,8 @@ static void ieee80211_send_addba_resp(struct sta_info *sta, u8 *da, u16 tid, mgmt->u.action.u.addba_resp.timeout = cpu_to_le16(timeout); mgmt->u.action.u.addba_resp.status = cpu_to_le16(status); - if (sta->sta.deflink.he_cap.has_he && addbaext) - ieee80211_add_addbaext(sdata, skb, addbaext, buf_size); + if (sta->sta.deflink.he_cap.has_he) + ieee80211_add_addbaext(skb, req_addba_ext_data, buf_size); ieee80211_tx_skb(sdata, skb); } @@ -233,7 +268,7 @@ void __ieee80211_start_rx_ba_session(struct sta_info *sta, u8 dialog_token, u16 timeout, u16 start_seq_num, u16 ba_policy, u16 tid, u16 buf_size, bool tx, bool auto_seq, - const struct ieee80211_addba_ext_ie *addbaext) + const u8 addba_ext_data) { struct ieee80211_local *local = sta->sdata->local; struct tid_ampdu_rx *tid_agg_rx; @@ -419,7 +454,7 @@ end: if (tx) ieee80211_send_addba_resp(sta, sta->sta.addr, tid, dialog_token, status, 1, buf_size, - timeout, addbaext); + timeout, addba_ext_data); } void ieee80211_process_addba_request(struct ieee80211_local *local, @@ -428,9 +463,7 @@ void ieee80211_process_addba_request(struct ieee80211_local *local, size_t len) { u16 capab, tid, timeout, ba_policy, buf_size, start_seq_num; - struct ieee802_11_elems *elems = NULL; - u8 dialog_token; - int ies_len; + u8 dialog_token, addba_ext_data; /* extract session parameters from addba request frame */ dialog_token = mgmt->u.action.u.addba_req.dialog_token; @@ -443,28 +476,17 @@ void ieee80211_process_addba_request(struct ieee80211_local *local, tid = (capab & IEEE80211_ADDBA_PARAM_TID_MASK) >> 2; buf_size = (capab & IEEE80211_ADDBA_PARAM_BUF_SIZE_MASK) >> 6; - ies_len = len - offsetof(struct ieee80211_mgmt, - u.action.u.addba_req.variable); - if (ies_len) { - elems = ieee802_11_parse_elems(mgmt->u.action.u.addba_req.variable, - ies_len, true, NULL); - if (!elems || elems->parse_error) - goto free; - } - - if (sta->sta.deflink.eht_cap.has_eht && elems && elems->addba_ext_ie) { - u8 buf_size_1k = u8_get_bits(elems->addba_ext_ie->data, - IEEE80211_ADDBA_EXT_BUF_SIZE_MASK); - - buf_size |= buf_size_1k << IEEE80211_ADDBA_EXT_BUF_SIZE_SHIFT; - } + addba_ext_data = + ieee80211_retrieve_addba_ext_data(sta, + mgmt->u.action.u.addba_req.variable, + len - + offsetof(typeof(*mgmt), + u.action.u.addba_req.variable), + &buf_size); __ieee80211_start_rx_ba_session(sta, dialog_token, timeout, start_seq_num, ba_policy, tid, - buf_size, true, false, - elems ? elems->addba_ext_ie : NULL); -free: - kfree(elems); + buf_size, true, false, addba_ext_data); } void ieee80211_manage_rx_ba_offl(struct ieee80211_vif *vif, diff --git a/net/mac80211/agg-tx.c b/net/mac80211/agg-tx.c index 04cb45cfb310..61f2cac37728 100644 --- a/net/mac80211/agg-tx.c +++ b/net/mac80211/agg-tx.c @@ -58,23 +58,24 @@ * complete. */ -static void ieee80211_send_addba_request(struct ieee80211_sub_if_data *sdata, - const u8 *da, u16 tid, +static void ieee80211_send_addba_request(struct sta_info *sta, u16 tid, u8 dialog_token, u16 start_seq_num, u16 agg_size, u16 timeout) { + struct ieee80211_sub_if_data *sdata = sta->sdata; struct ieee80211_local *local = sdata->local; struct sk_buff *skb; struct ieee80211_mgmt *mgmt; u16 capab; - skb = dev_alloc_skb(sizeof(*mgmt) + local->hw.extra_tx_headroom); - + skb = dev_alloc_skb(sizeof(*mgmt) + + 2 + sizeof(struct ieee80211_addba_ext_ie) + + local->hw.extra_tx_headroom); if (!skb) return; skb_reserve(skb, local->hw.extra_tx_headroom); - mgmt = ieee80211_mgmt_ba(skb, da, sdata); + mgmt = ieee80211_mgmt_ba(skb, sta->sta.addr, sdata); skb_put(skb, 1 + sizeof(mgmt->u.action.u.addba_req)); @@ -93,6 +94,9 @@ static void ieee80211_send_addba_request(struct ieee80211_sub_if_data *sdata, mgmt->u.action.u.addba_req.start_seq_num = cpu_to_le16(start_seq_num << 4); + if (sta->sta.deflink.he_cap.has_he) + ieee80211_add_addbaext(skb, 0, agg_size); + ieee80211_tx_skb_tid(sdata, skb, tid, -1); } @@ -460,8 +464,11 @@ static void ieee80211_send_addba_with_timeout(struct sta_info *sta, sta->ampdu_mlme.addba_req_num[tid]++; spin_unlock_bh(&sta->lock); - if (sta->sta.deflink.he_cap.has_he) { + if (sta->sta.deflink.eht_cap.has_eht) { buf_size = local->hw.max_tx_aggregation_subframes; + } else if (sta->sta.deflink.he_cap.has_he) { + buf_size = min_t(u16, local->hw.max_tx_aggregation_subframes, + IEEE80211_MAX_AMPDU_BUF_HE); } else { /* * We really should use what the driver told us it will @@ -473,9 +480,8 @@ static void ieee80211_send_addba_with_timeout(struct sta_info *sta, } /* send AddBA request */ - ieee80211_send_addba_request(sdata, sta->sta.addr, tid, - tid_tx->dialog_token, tid_tx->ssn, - buf_size, tid_tx->timeout); + ieee80211_send_addba_request(sta, tid, tid_tx->dialog_token, + tid_tx->ssn, buf_size, tid_tx->timeout); WARN_ON(test_and_set_bit(HT_AGG_STATE_SENT_ADDBA, &tid_tx->state)); } @@ -970,6 +976,13 @@ void ieee80211_process_addba_resp(struct ieee80211_local *local, amsdu = capab & IEEE80211_ADDBA_PARAM_AMSDU_MASK; tid = u16_get_bits(capab, IEEE80211_ADDBA_PARAM_TID_MASK); buf_size = u16_get_bits(capab, IEEE80211_ADDBA_PARAM_BUF_SIZE_MASK); + + ieee80211_retrieve_addba_ext_data(sta, + mgmt->u.action.u.addba_resp.variable, + len - offsetof(typeof(*mgmt), + u.action.u.addba_resp.variable), + &buf_size); + buf_size = min(buf_size, local->hw.max_tx_aggregation_subframes); txq = sta->sta.txq[tid]; diff --git a/net/mac80211/ht.c b/net/mac80211/ht.c index 1c2b7dd8976a..32390d8a9d75 100644 --- a/net/mac80211/ht.c +++ b/net/mac80211/ht.c @@ -379,7 +379,7 @@ void ieee80211_ba_session_work(struct wiphy *wiphy, struct wiphy_work *work) sta->ampdu_mlme.tid_rx_manage_offl)) __ieee80211_start_rx_ba_session(sta, 0, 0, 0, 1, tid, IEEE80211_MAX_AMPDU_BUF_HT, - false, true, NULL); + false, true, 0); if (test_and_clear_bit(tid + IEEE80211_NUM_TIDS, sta->ampdu_mlme.tid_rx_manage_offl)) diff --git a/net/mac80211/ieee80211_i.h b/net/mac80211/ieee80211_i.h index 7dcb46120abc..752297bcde76 100644 --- a/net/mac80211/ieee80211_i.h +++ b/net/mac80211/ieee80211_i.h @@ -2111,14 +2111,19 @@ int ieee80211_send_smps_action(struct ieee80211_sub_if_data *sdata, const u8 *bssid, int link_id); bool ieee80211_smps_is_restrictive(enum ieee80211_smps_mode smps_mode_old, enum ieee80211_smps_mode smps_mode_new); - +void ieee80211_add_addbaext(struct sk_buff *skb, + const u8 req_addba_ext_data, + u16 buf_size); +u8 ieee80211_retrieve_addba_ext_data(struct sta_info *sta, + const void *elem_data, ssize_t elem_len, + u16 *buf_size); void __ieee80211_stop_rx_ba_session(struct sta_info *sta, u16 tid, u16 initiator, u16 reason, bool stop); void __ieee80211_start_rx_ba_session(struct sta_info *sta, u8 dialog_token, u16 timeout, u16 start_seq_num, u16 ba_policy, u16 tid, u16 buf_size, bool tx, bool auto_seq, - const struct ieee80211_addba_ext_ie *addbaext); + const u8 addba_ext_data); void ieee80211_sta_tear_down_BA_sessions(struct sta_info *sta, enum ieee80211_agg_stop_reason reason); void ieee80211_process_delba(struct ieee80211_sub_if_data *sdata, -- cgit v1.2.3 From 16220cb315a0a10d2a003d9af534988686e675ea Mon Sep 17 00:00:00 2001 From: Tristram Ha Date: Fri, 8 Nov 2024 17:57:05 -0800 Subject: net: dsa: microchip: Add LAN9646 switch support to KSZ DSA driver LAN9646 switch is a 6-port switch with functions like KSZ9897. It has 4 internal PHYs and 1 SGMII port. The chip id read from hardware is same as KSZ9477, so software driver needs to create a new chip id and group allowable functions under its chip data structure to differentiate the product. Signed-off-by: Tristram Ha Link: https://patch.msgid.link/20241109015705.82685-3-Tristram.Ha@microchip.com Signed-off-by: Jakub Kicinski --- drivers/net/dsa/microchip/ksz9477.c | 4 +++ drivers/net/dsa/microchip/ksz9477_i2c.c | 14 +++++++- drivers/net/dsa/microchip/ksz_common.c | 50 +++++++++++++++++++++++++++-- drivers/net/dsa/microchip/ksz_common.h | 1 + drivers/net/dsa/microchip/ksz_spi.c | 7 ++++ include/linux/platform_data/microchip-ksz.h | 1 + 6 files changed, 74 insertions(+), 3 deletions(-) (limited to 'include/linux') diff --git a/drivers/net/dsa/microchip/ksz9477.c b/drivers/net/dsa/microchip/ksz9477.c index 0ba658a72d8f..d16817e0476f 100644 --- a/drivers/net/dsa/microchip/ksz9477.c +++ b/drivers/net/dsa/microchip/ksz9477.c @@ -1131,6 +1131,10 @@ void ksz9477_config_cpu_port(struct dsa_switch *ds) if (i == dev->cpu_port) continue; ksz_port_stp_state_set(ds, i, BR_STATE_DISABLED); + + /* Power down the internal PHY if port is unused. */ + if (dsa_is_unused_port(ds, i) && dev->info->internal_phy[i]) + ksz_pwrite16(dev, i, 0x100, BMCR_PDOWN); } } diff --git a/drivers/net/dsa/microchip/ksz9477_i2c.c b/drivers/net/dsa/microchip/ksz9477_i2c.c index 7d7560f23a73..1c6d7fc16772 100644 --- a/drivers/net/dsa/microchip/ksz9477_i2c.c +++ b/drivers/net/dsa/microchip/ksz9477_i2c.c @@ -2,7 +2,7 @@ /* * Microchip KSZ9477 series register access through I2C * - * Copyright (C) 2018-2019 Microchip Technology Inc. + * Copyright (C) 2018-2024 Microchip Technology Inc. */ #include @@ -16,6 +16,8 @@ KSZ_REGMAP_TABLE(ksz9477, not_used, 16, 0, 0); static int ksz9477_i2c_probe(struct i2c_client *i2c) { + const struct ksz_chip_data *chip; + struct device *ddev = &i2c->dev; struct regmap_config rc; struct ksz_device *dev; int i, ret; @@ -24,6 +26,12 @@ static int ksz9477_i2c_probe(struct i2c_client *i2c) if (!dev) return -ENOMEM; + chip = device_get_match_data(ddev); + if (!chip) + return -EINVAL; + + /* Save chip id to do special initialization when probing. */ + dev->chip_id = chip->chip_id; for (i = 0; i < __KSZ_NUM_REGMAPS; i++) { rc = ksz9477_regmap_config[i]; rc.lock_arg = &dev->regmap_mutex; @@ -111,6 +119,10 @@ static const struct of_device_id ksz9477_dt_ids[] = { .compatible = "microchip,ksz9567", .data = &ksz_switch_chips[KSZ9567] }, + { + .compatible = "microchip,lan9646", + .data = &ksz_switch_chips[LAN9646] + }, {}, }; MODULE_DEVICE_TABLE(of, ksz9477_dt_ids); diff --git a/drivers/net/dsa/microchip/ksz_common.c b/drivers/net/dsa/microchip/ksz_common.c index d57782cdda59..920443ee8ffd 100644 --- a/drivers/net/dsa/microchip/ksz_common.c +++ b/drivers/net/dsa/microchip/ksz_common.c @@ -1908,6 +1908,41 @@ const struct ksz_chip_data ksz_switch_chips[] = { .internal_phy = {true, true, true, true, false, false, true, true}, }, + + [LAN9646] = { + .chip_id = LAN9646_CHIP_ID, + .dev_name = "LAN9646", + .num_vlans = 4096, + .num_alus = 4096, + .num_statics = 16, + .cpu_ports = 0x7F, /* can be configured as cpu port */ + .port_cnt = 7, /* total physical port count */ + .port_nirqs = 4, + .num_tx_queues = 4, + .num_ipms = 8, + .ops = &ksz9477_dev_ops, + .phylink_mac_ops = &ksz9477_phylink_mac_ops, + .phy_errata_9477 = true, + .mib_names = ksz9477_mib_names, + .mib_cnt = ARRAY_SIZE(ksz9477_mib_names), + .reg_mib_cnt = MIB_COUNTER_NUM, + .regs = ksz9477_regs, + .masks = ksz9477_masks, + .shifts = ksz9477_shifts, + .xmii_ctrl0 = ksz9477_xmii_ctrl0, + .xmii_ctrl1 = ksz9477_xmii_ctrl1, + .supports_mii = {false, false, false, false, + false, true, true}, + .supports_rmii = {false, false, false, false, + false, true, true}, + .supports_rgmii = {false, false, false, false, + false, true, true}, + .internal_phy = {true, true, true, true, + true, false, false}, + .gbit_capable = {true, true, true, true, true, true, true}, + .wr_table = &ksz9477_register_set, + .rd_table = &ksz9477_register_set, + }, }; EXPORT_SYMBOL_GPL(ksz_switch_chips); @@ -2970,6 +3005,7 @@ static u32 ksz_get_phy_flags(struct dsa_switch *ds, int port) case KSZ9896_CHIP_ID: /* KSZ9896C Errata DS80000757A Module 3 */ case KSZ9897_CHIP_ID: + case LAN9646_CHIP_ID: /* KSZ9897R Errata DS80000758C Module 4 */ /* Energy Efficient Ethernet (EEE) feature select must be manually disabled * The EEE feature is enabled by default, but it is not fully @@ -3230,6 +3266,7 @@ static void ksz_port_teardown(struct dsa_switch *ds, int port) case KSZ9893_CHIP_ID: case KSZ9896_CHIP_ID: case KSZ9897_CHIP_ID: + case LAN9646_CHIP_ID: if (dsa_is_user_port(ds, port)) ksz9477_port_acl_free(dev, port); } @@ -3286,7 +3323,8 @@ static enum dsa_tag_protocol ksz_get_tag_protocol(struct dsa_switch *ds, dev->chip_id == KSZ9477_CHIP_ID || dev->chip_id == KSZ9896_CHIP_ID || dev->chip_id == KSZ9897_CHIP_ID || - dev->chip_id == KSZ9567_CHIP_ID) + dev->chip_id == KSZ9567_CHIP_ID || + dev->chip_id == LAN9646_CHIP_ID) proto = DSA_TAG_PROTO_KSZ9477; if (is_lan937x(dev)) @@ -3405,6 +3443,7 @@ static int ksz_max_mtu(struct dsa_switch *ds, int port) case LAN9372_CHIP_ID: case LAN9373_CHIP_ID: case LAN9374_CHIP_ID: + case LAN9646_CHIP_ID: return KSZ9477_MAX_FRAME_SIZE - VLAN_ETH_HLEN - ETH_FCS_LEN; } @@ -3427,6 +3466,7 @@ static int ksz_validate_eee(struct dsa_switch *ds, int port) case KSZ9893_CHIP_ID: case KSZ9896_CHIP_ID: case KSZ9897_CHIP_ID: + case LAN9646_CHIP_ID: return 0; } @@ -3779,7 +3819,10 @@ static int ksz_switch_detect(struct ksz_device *dev) case LAN9372_CHIP_ID: case LAN9373_CHIP_ID: case LAN9374_CHIP_ID: - dev->chip_id = id32; + + /* LAN9646 does not have its own chip id. */ + if (dev->chip_id != LAN9646_CHIP_ID) + dev->chip_id = id32; break; case KSZ9893_CHIP_ID: ret = ksz_read8(dev, REG_CHIP_ID4, @@ -3818,6 +3861,7 @@ static int ksz_cls_flower_add(struct dsa_switch *ds, int port, case KSZ9893_CHIP_ID: case KSZ9896_CHIP_ID: case KSZ9897_CHIP_ID: + case LAN9646_CHIP_ID: return ksz9477_cls_flower_add(ds, port, cls, ingress); } @@ -3838,6 +3882,7 @@ static int ksz_cls_flower_del(struct dsa_switch *ds, int port, case KSZ9893_CHIP_ID: case KSZ9896_CHIP_ID: case KSZ9897_CHIP_ID: + case LAN9646_CHIP_ID: return ksz9477_cls_flower_del(ds, port, cls, ingress); } @@ -4925,6 +4970,7 @@ static int ksz_parse_drive_strength(struct ksz_device *dev) case KSZ9893_CHIP_ID: case KSZ9896_CHIP_ID: case KSZ9897_CHIP_ID: + case LAN9646_CHIP_ID: return ksz9477_drive_strength_write(dev, of_props, ARRAY_SIZE(of_props)); default: diff --git a/drivers/net/dsa/microchip/ksz_common.h b/drivers/net/dsa/microchip/ksz_common.h index bbb548af201e..b3bb75ca0796 100644 --- a/drivers/net/dsa/microchip/ksz_common.h +++ b/drivers/net/dsa/microchip/ksz_common.h @@ -236,6 +236,7 @@ enum ksz_model { LAN9372, LAN9373, LAN9374, + LAN9646, }; enum ksz_regs { diff --git a/drivers/net/dsa/microchip/ksz_spi.c b/drivers/net/dsa/microchip/ksz_spi.c index 1c6652f2b9fe..108a958dc356 100644 --- a/drivers/net/dsa/microchip/ksz_spi.c +++ b/drivers/net/dsa/microchip/ksz_spi.c @@ -54,6 +54,8 @@ static int ksz_spi_probe(struct spi_device *spi) if (!chip) return -EINVAL; + /* Save chip id to do special initialization when probing. */ + dev->chip_id = chip->chip_id; if (chip->chip_id == KSZ88X3_CHIP_ID) regmap_config = ksz8863_regmap_config; else if (chip->chip_id == KSZ8795_CHIP_ID || @@ -203,6 +205,10 @@ static const struct of_device_id ksz_dt_ids[] = { .compatible = "microchip,lan9374", .data = &ksz_switch_chips[LAN9374] }, + { + .compatible = "microchip,lan9646", + .data = &ksz_switch_chips[LAN9646] + }, {}, }; MODULE_DEVICE_TABLE(of, ksz_dt_ids); @@ -228,6 +234,7 @@ static const struct spi_device_id ksz_spi_ids[] = { { "lan9372" }, { "lan9373" }, { "lan9374" }, + { "lan9646" }, { }, }; MODULE_DEVICE_TABLE(spi, ksz_spi_ids); diff --git a/include/linux/platform_data/microchip-ksz.h b/include/linux/platform_data/microchip-ksz.h index 2ee1a679e592..0e0e8fe6975f 100644 --- a/include/linux/platform_data/microchip-ksz.h +++ b/include/linux/platform_data/microchip-ksz.h @@ -42,6 +42,7 @@ enum ksz_chip_id { LAN9372_CHIP_ID = 0x00937200, LAN9373_CHIP_ID = 0x00937300, LAN9374_CHIP_ID = 0x00937400, + LAN9646_CHIP_ID = 0x00964600, }; struct ksz_platform_data { -- cgit v1.2.3 From 721aa69e708b7432af83c4bb00a30e2b7c27da28 Mon Sep 17 00:00:00 2001 From: Heiner Kallweit Date: Fri, 8 Nov 2024 07:54:47 +0100 Subject: net: phy: convert eee_broken_modes to a linkmode bitmap eee_broken_modes has a eee_cap1 register layout currently. This doen't allow to flag e.g. 2.5Gbps or 5Gbps BaseT EEE as broken. To overcome this limitation switch eee_broken_modes to a linkmode bitmap. Signed-off-by: Heiner Kallweit Link: https://patch.msgid.link/dfe0c9ff-84b0-4328-86d7-e917ebc084a1@gmail.com Signed-off-by: Jakub Kicinski --- drivers/net/phy/micrel.c | 2 +- drivers/net/phy/phy-c45.c | 12 +++++------- drivers/net/phy/phy-core.c | 21 +++++++++------------ include/linux/phy.h | 9 ++++----- 4 files changed, 19 insertions(+), 25 deletions(-) (limited to 'include/linux') diff --git a/drivers/net/phy/micrel.c b/drivers/net/phy/micrel.c index 43c82a87bc3a..3ef508840674 100644 --- a/drivers/net/phy/micrel.c +++ b/drivers/net/phy/micrel.c @@ -2004,7 +2004,7 @@ static int ksz9477_config_init(struct phy_device *phydev) * in this switch shall be regarded as broken. */ if (phydev->dev_flags & MICREL_NO_EEE) - phydev->eee_broken_modes = -1; + linkmode_fill(phydev->eee_broken_modes); return kszphy_config_init(phydev); } diff --git a/drivers/net/phy/phy-c45.c b/drivers/net/phy/phy-c45.c index 7b0ffa3515aa..da5c15310d8c 100644 --- a/drivers/net/phy/phy-c45.c +++ b/drivers/net/phy/phy-c45.c @@ -683,15 +683,13 @@ EXPORT_SYMBOL_GPL(genphy_c45_read_mdix); static int genphy_c45_write_eee_adv(struct phy_device *phydev, unsigned long *adv) { + __ETHTOOL_DECLARE_LINK_MODE_MASK(tmp); int val, changed = 0; - if (linkmode_intersects(phydev->supported_eee, PHY_EEE_CAP1_FEATURES)) { - val = linkmode_to_mii_eee_cap1_t(adv); + linkmode_andnot(tmp, adv, phydev->eee_broken_modes); - /* In eee_broken_modes are stored MDIO_AN_EEE_ADV specific raw - * register values. - */ - val &= ~phydev->eee_broken_modes; + if (linkmode_intersects(phydev->supported_eee, PHY_EEE_CAP1_FEATURES)) { + val = linkmode_to_mii_eee_cap1_t(tmp); /* IEEE 802.3-2018 45.2.7.13 EEE advertisement 1 * (Register 7.60) @@ -709,7 +707,7 @@ static int genphy_c45_write_eee_adv(struct phy_device *phydev, } if (linkmode_intersects(phydev->supported_eee, PHY_EEE_CAP2_FEATURES)) { - val = linkmode_to_mii_eee_cap2_t(adv); + val = linkmode_to_mii_eee_cap2_t(tmp); /* IEEE 802.3-2022 45.2.7.16 EEE advertisement 2 * (Register 7.62) diff --git a/drivers/net/phy/phy-core.c b/drivers/net/phy/phy-core.c index 4e8db12d6092..6bf3ec985f3d 100644 --- a/drivers/net/phy/phy-core.c +++ b/drivers/net/phy/phy-core.c @@ -388,28 +388,25 @@ void of_set_phy_supported(struct phy_device *phydev) void of_set_phy_eee_broken(struct phy_device *phydev) { struct device_node *node = phydev->mdio.dev.of_node; - u32 broken = 0; + unsigned long *modes = phydev->eee_broken_modes; - if (!IS_ENABLED(CONFIG_OF_MDIO)) + if (!IS_ENABLED(CONFIG_OF_MDIO) || !node) return; - if (!node) - return; + linkmode_zero(modes); if (of_property_read_bool(node, "eee-broken-100tx")) - broken |= MDIO_EEE_100TX; + linkmode_set_bit(ETHTOOL_LINK_MODE_100baseT_Full_BIT, modes); if (of_property_read_bool(node, "eee-broken-1000t")) - broken |= MDIO_EEE_1000T; + linkmode_set_bit(ETHTOOL_LINK_MODE_1000baseT_Full_BIT, modes); if (of_property_read_bool(node, "eee-broken-10gt")) - broken |= MDIO_EEE_10GT; + linkmode_set_bit(ETHTOOL_LINK_MODE_10000baseT_Full_BIT, modes); if (of_property_read_bool(node, "eee-broken-1000kx")) - broken |= MDIO_EEE_1000KX; + linkmode_set_bit(ETHTOOL_LINK_MODE_1000baseKX_Full_BIT, modes); if (of_property_read_bool(node, "eee-broken-10gkx4")) - broken |= MDIO_EEE_10GKX4; + linkmode_set_bit(ETHTOOL_LINK_MODE_10000baseKX4_Full_BIT, modes); if (of_property_read_bool(node, "eee-broken-10gkr")) - broken |= MDIO_EEE_10GKR; - - phydev->eee_broken_modes = broken; + linkmode_set_bit(ETHTOOL_LINK_MODE_10000baseKR_Full_BIT, modes); } /** diff --git a/include/linux/phy.h b/include/linux/phy.h index 1e4127c495c0..a6622f873d03 100644 --- a/include/linux/phy.h +++ b/include/linux/phy.h @@ -721,16 +721,15 @@ struct phy_device { /* used for eee validation and configuration*/ __ETHTOOL_DECLARE_LINK_MODE_MASK(supported_eee); __ETHTOOL_DECLARE_LINK_MODE_MASK(advertising_eee); + /* Energy efficient ethernet modes which should be prohibited */ + __ETHTOOL_DECLARE_LINK_MODE_MASK(eee_broken_modes); bool eee_enabled; + bool enable_tx_lpi; + struct eee_config eee_cfg; /* Host supported PHY interface types. Should be ignored if empty. */ DECLARE_PHY_INTERFACE_MASK(host_interfaces); - /* Energy efficient ethernet modes which should be prohibited */ - u32 eee_broken_modes; - bool enable_tx_lpi; - struct eee_config eee_cfg; - #ifdef CONFIG_LED_TRIGGER_PHY struct phy_led_trigger *phy_led_triggers; unsigned int phy_num_led_triggers; -- cgit v1.2.3 From ed623fb8e38e2a241da12864778ec9c9cf930c65 Mon Sep 17 00:00:00 2001 From: Heiner Kallweit Date: Fri, 8 Nov 2024 08:07:10 +0100 Subject: net: phy: add phy_set_eee_broken Add an accessor for eee_broken_modes, so that drivers don't have to deal with phylib internals. Signed-off-by: Heiner Kallweit Link: https://patch.msgid.link/0f8ee279-d40d-4489-a3b0-d993472d744a@gmail.com Signed-off-by: Jakub Kicinski --- include/linux/phy.h | 10 ++++++++++ 1 file changed, 10 insertions(+) (limited to 'include/linux') diff --git a/include/linux/phy.h b/include/linux/phy.h index a6622f873d03..b8346db42727 100644 --- a/include/linux/phy.h +++ b/include/linux/phy.h @@ -1263,6 +1263,16 @@ void of_set_phy_eee_broken(struct phy_device *phydev); void of_set_phy_timing_role(struct phy_device *phydev); int phy_speed_down_core(struct phy_device *phydev); +/** + * phy_set_eee_broken - Mark an EEE mode as broken so that it isn't advertised. + * @phydev: The phy_device struct + * @link_mode: The broken EEE mode + */ +static inline void phy_set_eee_broken(struct phy_device *phydev, u32 link_mode) +{ + linkmode_set_bit(link_mode, phydev->eee_broken_modes); +} + /** * phy_is_started - Convenience function to check whether PHY is started * @phydev: The phy_device struct -- cgit v1.2.3 From e7cb7cf43afb96a4fd8f68c2dfb9b2fbdd444654 Mon Sep 17 00:00:00 2001 From: Alistair Francis Date: Tue, 12 Nov 2024 20:54:29 +1000 Subject: include: mdio: Remove mdio45_ethtool_gset() mdio45_ethtool_gset() is never called, so let's remove it. Signed-off-by: Alistair Francis Link: https://patch.msgid.link/20241112105430.438491-1-alistair@alistair23.me Signed-off-by: Jakub Kicinski --- include/linux/mdio.h | 16 ---------------- 1 file changed, 16 deletions(-) (limited to 'include/linux') diff --git a/include/linux/mdio.h b/include/linux/mdio.h index efeca5bd7600..c63f43645d50 100644 --- a/include/linux/mdio.h +++ b/include/linux/mdio.h @@ -173,22 +173,6 @@ mdio45_ethtool_ksettings_get_npage(const struct mdio_if_info *mdio, struct ethtool_link_ksettings *cmd, u32 npage_adv, u32 npage_lpa); -/** - * mdio45_ethtool_gset - get settings for ETHTOOL_GSET - * @mdio: MDIO interface - * @ecmd: Ethtool request structure - * - * Since the CSRs for auto-negotiation using next pages are not fully - * standardised, this function does not attempt to decode them. Use - * mdio45_ethtool_gset_npage() to specify advertisement bits from next - * pages. - */ -static inline void mdio45_ethtool_gset(const struct mdio_if_info *mdio, - struct ethtool_cmd *ecmd) -{ - mdio45_ethtool_gset_npage(mdio, ecmd, 0, 0); -} - /** * mdio45_ethtool_ksettings_get - get settings for ETHTOOL_GLINKSETTINGS * @mdio: MDIO interface -- cgit v1.2.3 From 575092a7f0ce4c6d5907cd908d28c0f69690a136 Mon Sep 17 00:00:00 2001 From: Alistair Francis Date: Tue, 12 Nov 2024 20:54:30 +1000 Subject: mdio: Remove mdio45_ethtool_gset_npage() The mdio45_ethtool_gset_npage() function isn't called, so let's remove it. Signed-off-by: Alistair Francis Link: https://patch.msgid.link/20241112105430.438491-2-alistair@alistair23.me Signed-off-by: Jakub Kicinski --- drivers/net/mdio.c | 172 --------------------------------------------------- include/linux/mdio.h | 3 - 2 files changed, 175 deletions(-) (limited to 'include/linux') diff --git a/drivers/net/mdio.c b/drivers/net/mdio.c index e08c90ac0c6e..f67a4d4005e7 100644 --- a/drivers/net/mdio.c +++ b/drivers/net/mdio.c @@ -166,178 +166,6 @@ static u32 mdio45_get_an(const struct mdio_if_info *mdio, u16 addr) return result; } -/** - * mdio45_ethtool_gset_npage - get settings for ETHTOOL_GSET - * @mdio: MDIO interface - * @ecmd: Ethtool request structure - * @npage_adv: Modes currently advertised on next pages - * @npage_lpa: Modes advertised by link partner on next pages - * - * The @ecmd parameter is expected to have been cleared before calling - * mdio45_ethtool_gset_npage(). - * - * Since the CSRs for auto-negotiation using next pages are not fully - * standardised, this function does not attempt to decode them. The - * caller must pass them in. - */ -void mdio45_ethtool_gset_npage(const struct mdio_if_info *mdio, - struct ethtool_cmd *ecmd, - u32 npage_adv, u32 npage_lpa) -{ - int reg; - u32 speed; - - BUILD_BUG_ON(MDIO_SUPPORTS_C22 != ETH_MDIO_SUPPORTS_C22); - BUILD_BUG_ON(MDIO_SUPPORTS_C45 != ETH_MDIO_SUPPORTS_C45); - - ecmd->transceiver = XCVR_INTERNAL; - ecmd->phy_address = mdio->prtad; - ecmd->mdio_support = - mdio->mode_support & (MDIO_SUPPORTS_C45 | MDIO_SUPPORTS_C22); - - reg = mdio->mdio_read(mdio->dev, mdio->prtad, MDIO_MMD_PMAPMD, - MDIO_CTRL2); - switch (reg & MDIO_PMA_CTRL2_TYPE) { - case MDIO_PMA_CTRL2_10GBT: - case MDIO_PMA_CTRL2_1000BT: - case MDIO_PMA_CTRL2_100BTX: - case MDIO_PMA_CTRL2_10BT: - ecmd->port = PORT_TP; - ecmd->supported = SUPPORTED_TP; - reg = mdio->mdio_read(mdio->dev, mdio->prtad, MDIO_MMD_PMAPMD, - MDIO_SPEED); - if (reg & MDIO_SPEED_10G) - ecmd->supported |= SUPPORTED_10000baseT_Full; - if (reg & MDIO_PMA_SPEED_1000) - ecmd->supported |= (SUPPORTED_1000baseT_Full | - SUPPORTED_1000baseT_Half); - if (reg & MDIO_PMA_SPEED_100) - ecmd->supported |= (SUPPORTED_100baseT_Full | - SUPPORTED_100baseT_Half); - if (reg & MDIO_PMA_SPEED_10) - ecmd->supported |= (SUPPORTED_10baseT_Full | - SUPPORTED_10baseT_Half); - ecmd->advertising = ADVERTISED_TP; - break; - - case MDIO_PMA_CTRL2_10GBCX4: - ecmd->port = PORT_OTHER; - ecmd->supported = 0; - ecmd->advertising = 0; - break; - - case MDIO_PMA_CTRL2_10GBKX4: - case MDIO_PMA_CTRL2_10GBKR: - case MDIO_PMA_CTRL2_1000BKX: - ecmd->port = PORT_OTHER; - ecmd->supported = SUPPORTED_Backplane; - reg = mdio->mdio_read(mdio->dev, mdio->prtad, MDIO_MMD_PMAPMD, - MDIO_PMA_EXTABLE); - if (reg & MDIO_PMA_EXTABLE_10GBKX4) - ecmd->supported |= SUPPORTED_10000baseKX4_Full; - if (reg & MDIO_PMA_EXTABLE_10GBKR) - ecmd->supported |= SUPPORTED_10000baseKR_Full; - if (reg & MDIO_PMA_EXTABLE_1000BKX) - ecmd->supported |= SUPPORTED_1000baseKX_Full; - reg = mdio->mdio_read(mdio->dev, mdio->prtad, MDIO_MMD_PMAPMD, - MDIO_PMA_10GBR_FECABLE); - if (reg & MDIO_PMA_10GBR_FECABLE_ABLE) - ecmd->supported |= SUPPORTED_10000baseR_FEC; - ecmd->advertising = ADVERTISED_Backplane; - break; - - /* All the other defined modes are flavours of optical */ - default: - ecmd->port = PORT_FIBRE; - ecmd->supported = SUPPORTED_FIBRE; - ecmd->advertising = ADVERTISED_FIBRE; - break; - } - - if (mdio->mmds & MDIO_DEVS_AN) { - ecmd->supported |= SUPPORTED_Autoneg; - reg = mdio->mdio_read(mdio->dev, mdio->prtad, MDIO_MMD_AN, - MDIO_CTRL1); - if (reg & MDIO_AN_CTRL1_ENABLE) { - ecmd->autoneg = AUTONEG_ENABLE; - ecmd->advertising |= - ADVERTISED_Autoneg | - mdio45_get_an(mdio, MDIO_AN_ADVERTISE) | - npage_adv; - } else { - ecmd->autoneg = AUTONEG_DISABLE; - } - } else { - ecmd->autoneg = AUTONEG_DISABLE; - } - - if (ecmd->autoneg) { - u32 modes = 0; - int an_stat = mdio->mdio_read(mdio->dev, mdio->prtad, - MDIO_MMD_AN, MDIO_STAT1); - - /* If AN is complete and successful, report best common - * mode, otherwise report best advertised mode. */ - if (an_stat & MDIO_AN_STAT1_COMPLETE) { - ecmd->lp_advertising = - mdio45_get_an(mdio, MDIO_AN_LPA) | npage_lpa; - if (an_stat & MDIO_AN_STAT1_LPABLE) - ecmd->lp_advertising |= ADVERTISED_Autoneg; - modes = ecmd->advertising & ecmd->lp_advertising; - } - if ((modes & ~ADVERTISED_Autoneg) == 0) - modes = ecmd->advertising; - - if (modes & (ADVERTISED_10000baseT_Full | - ADVERTISED_10000baseKX4_Full | - ADVERTISED_10000baseKR_Full)) { - speed = SPEED_10000; - ecmd->duplex = DUPLEX_FULL; - } else if (modes & (ADVERTISED_1000baseT_Full | - ADVERTISED_1000baseT_Half | - ADVERTISED_1000baseKX_Full)) { - speed = SPEED_1000; - ecmd->duplex = !(modes & ADVERTISED_1000baseT_Half); - } else if (modes & (ADVERTISED_100baseT_Full | - ADVERTISED_100baseT_Half)) { - speed = SPEED_100; - ecmd->duplex = !!(modes & ADVERTISED_100baseT_Full); - } else { - speed = SPEED_10; - ecmd->duplex = !!(modes & ADVERTISED_10baseT_Full); - } - } else { - /* Report forced settings */ - reg = mdio->mdio_read(mdio->dev, mdio->prtad, MDIO_MMD_PMAPMD, - MDIO_CTRL1); - speed = (((reg & MDIO_PMA_CTRL1_SPEED1000) ? 100 : 1) - * ((reg & MDIO_PMA_CTRL1_SPEED100) ? 100 : 10)); - ecmd->duplex = (reg & MDIO_CTRL1_FULLDPLX || - speed == SPEED_10000); - } - - ethtool_cmd_speed_set(ecmd, speed); - - /* 10GBASE-T MDI/MDI-X */ - if (ecmd->port == PORT_TP - && (ethtool_cmd_speed(ecmd) == SPEED_10000)) { - switch (mdio->mdio_read(mdio->dev, mdio->prtad, MDIO_MMD_PMAPMD, - MDIO_PMA_10GBT_SWAPPOL)) { - case MDIO_PMA_10GBT_SWAPPOL_ABNX | MDIO_PMA_10GBT_SWAPPOL_CDNX: - ecmd->eth_tp_mdix = ETH_TP_MDI; - break; - case 0: - ecmd->eth_tp_mdix = ETH_TP_MDI_X; - break; - default: - /* It's complicated... */ - ecmd->eth_tp_mdix = ETH_TP_MDI_INVALID; - break; - } - } -} -EXPORT_SYMBOL(mdio45_ethtool_gset_npage); - /** * mdio45_ethtool_ksettings_get_npage - get settings for ETHTOOL_GLINKSETTINGS * @mdio: MDIO interface diff --git a/include/linux/mdio.h b/include/linux/mdio.h index c63f43645d50..3c3deac57894 100644 --- a/include/linux/mdio.h +++ b/include/linux/mdio.h @@ -165,9 +165,6 @@ extern int mdio_set_flag(const struct mdio_if_info *mdio, bool sense); extern int mdio45_links_ok(const struct mdio_if_info *mdio, u32 mmds); extern int mdio45_nway_restart(const struct mdio_if_info *mdio); -extern void mdio45_ethtool_gset_npage(const struct mdio_if_info *mdio, - struct ethtool_cmd *ecmd, - u32 npage_adv, u32 npage_lpa); extern void mdio45_ethtool_ksettings_get_npage(const struct mdio_if_info *mdio, struct ethtool_link_ksettings *cmd, -- cgit v1.2.3 From 9e43ad7a1edef268acac603e1975c8f50a20d02f Mon Sep 17 00:00:00 2001 From: Edward Cree Date: Wed, 13 Nov 2024 12:13:09 +0000 Subject: net: ethtool: only allow set_rxnfc with rss + ring_cookie if driver opts in Ethtool ntuple filters with FLOW_RSS were originally defined as adding the base queue ID (ring_cookie) to the value from the indirection table, so that the same table could distribute over more than one set of queues when used by different filters. However, some drivers / hardware ignore the ring_cookie, and simply use the indirection table entries as queue IDs directly. Thus, for drivers which have not opted in by setting ethtool_ops.cap_rss_rxnfc_adds to declare that they support the original (addition) semantics, reject in ethtool_set_rxnfc any filter which combines FLOW_RSS and a nonzero ring. (For a ring_cookie of zero, both behaviours are equivalent.) Set the cap bit in sfc, as it is known to support this feature. Signed-off-by: Edward Cree Reviewed-by: Martin Habets Link: https://patch.msgid.link/cc3da0844083b0e301a33092a6299e4042b65221.1731499022.git.ecree.xilinx@gmail.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/sfc/ef100_ethtool.c | 1 + drivers/net/ethernet/sfc/ethtool.c | 1 + include/linux/ethtool.h | 4 ++++ net/ethtool/ioctl.c | 5 +++++ 4 files changed, 11 insertions(+) (limited to 'include/linux') diff --git a/drivers/net/ethernet/sfc/ef100_ethtool.c b/drivers/net/ethernet/sfc/ef100_ethtool.c index 5c2551369812..6c3b74000d3b 100644 --- a/drivers/net/ethernet/sfc/ef100_ethtool.c +++ b/drivers/net/ethernet/sfc/ef100_ethtool.c @@ -59,6 +59,7 @@ const struct ethtool_ops ef100_ethtool_ops = { .get_rxfh_indir_size = efx_ethtool_get_rxfh_indir_size, .get_rxfh_key_size = efx_ethtool_get_rxfh_key_size, .rxfh_per_ctx_key = true, + .cap_rss_rxnfc_adds = true, .rxfh_priv_size = sizeof(struct efx_rss_context_priv), .get_rxfh = efx_ethtool_get_rxfh, .set_rxfh = efx_ethtool_set_rxfh, diff --git a/drivers/net/ethernet/sfc/ethtool.c b/drivers/net/ethernet/sfc/ethtool.c index bb1930818beb..83d715544f7f 100644 --- a/drivers/net/ethernet/sfc/ethtool.c +++ b/drivers/net/ethernet/sfc/ethtool.c @@ -263,6 +263,7 @@ const struct ethtool_ops efx_ethtool_ops = { .get_rxfh_indir_size = efx_ethtool_get_rxfh_indir_size, .get_rxfh_key_size = efx_ethtool_get_rxfh_key_size, .rxfh_per_ctx_key = true, + .cap_rss_rxnfc_adds = true, .rxfh_priv_size = sizeof(struct efx_rss_context_priv), .get_rxfh = efx_ethtool_get_rxfh, .set_rxfh = efx_ethtool_set_rxfh, diff --git a/include/linux/ethtool.h b/include/linux/ethtool.h index 1199e308c8dd..299280c94d07 100644 --- a/include/linux/ethtool.h +++ b/include/linux/ethtool.h @@ -734,6 +734,9 @@ struct kernel_ethtool_ts_info { * @rxfh_per_ctx_key: device supports setting different RSS key for each * additional context. Netlink API should report hfunc, key, and input_xfrm * for every context, not just context 0. + * @cap_rss_rxnfc_adds: device supports nonzero ring_cookie in filters with + * %FLOW_RSS flag; the queue ID from the filter is added to the value from + * the indirection table to determine the delivery queue. * @rxfh_indir_space: max size of RSS indirection tables, if indirection table * size as returned by @get_rxfh_indir_size may change during lifetime * of the device. Leave as 0 if the table size is constant. @@ -956,6 +959,7 @@ struct ethtool_ops { u32 cap_rss_ctx_supported:1; u32 cap_rss_sym_xor_supported:1; u32 rxfh_per_ctx_key:1; + u32 cap_rss_rxnfc_adds:1; u32 rxfh_indir_space; u16 rxfh_key_space; u16 rxfh_priv_size; diff --git a/net/ethtool/ioctl.c b/net/ethtool/ioctl.c index 7da94e26ced6..d86399bcf223 100644 --- a/net/ethtool/ioctl.c +++ b/net/ethtool/ioctl.c @@ -992,6 +992,11 @@ static noinline_for_stack int ethtool_set_rxnfc(struct net_device *dev, if (rc) return rc; + /* Nonzero ring with RSS only makes sense if NIC adds them together */ + if (info.flow_type & FLOW_RSS && !ops->cap_rss_rxnfc_adds && + ethtool_get_flow_spec_ring(info.fs.ring_cookie)) + return -EINVAL; + if (ops->get_rxfh) { struct ethtool_rxfh_param rxfh = {}; -- cgit v1.2.3 From 41ffcd95015f18178ddc53fed919c2842d52fe38 Mon Sep 17 00:00:00 2001 From: "Russell King (Oracle)" Date: Thu, 14 Nov 2024 10:33:27 +0000 Subject: net: phy: fix phylib's dual eee_enabled phylib has two eee_enabled members. Some parts of the code are using phydev->eee_enabled, other parts are using phydev->eee_cfg.eee_enabled. This leads to incorrect behaviour as their state goes out of sync. ethtool --show-eee shows incorrect information, and --set-eee sometimes doesn't take effect. Fix this by only having one eee_enabled member - that in eee_cfg. Fixes: 49168d1980e2 ("net: phy: Add phy_support_eee() indicating MAC support EEE") Signed-off-by: Russell King (Oracle) Reviewed-by: Heiner Kallweit Link: https://patch.msgid.link/E1tBXAF-00341F-EQ@rmk-PC.armlinux.org.uk Signed-off-by: Jakub Kicinski --- drivers/net/phy/phy-c45.c | 4 +--- drivers/net/phy/phy_device.c | 4 ++-- include/linux/phy.h | 2 -- 3 files changed, 3 insertions(+), 7 deletions(-) (limited to 'include/linux') diff --git a/drivers/net/phy/phy-c45.c b/drivers/net/phy/phy-c45.c index 5695935fdce9..ac987e5e82dc 100644 --- a/drivers/net/phy/phy-c45.c +++ b/drivers/net/phy/phy-c45.c @@ -942,7 +942,7 @@ EXPORT_SYMBOL_GPL(genphy_c45_read_eee_abilities); */ int genphy_c45_an_config_eee_aneg(struct phy_device *phydev) { - if (!phydev->eee_enabled) { + if (!phydev->eee_cfg.eee_enabled) { __ETHTOOL_DECLARE_LINK_MODE_MASK(adv) = {}; return genphy_c45_write_eee_adv(phydev, adv); @@ -1575,8 +1575,6 @@ int genphy_c45_ethtool_set_eee(struct phy_device *phydev, linkmode_copy(phydev->advertising_eee, adv); } - phydev->eee_enabled = data->eee_enabled; - ret = genphy_c45_an_config_eee_aneg(phydev); if (ret > 0) { ret = phy_restart_aneg(phydev); diff --git a/drivers/net/phy/phy_device.c b/drivers/net/phy/phy_device.c index 499797646580..5dfa2aa53c90 100644 --- a/drivers/net/phy/phy_device.c +++ b/drivers/net/phy/phy_device.c @@ -3595,12 +3595,12 @@ static int phy_probe(struct device *dev) /* There is no "enabled" flag. If PHY is advertising, assume it is * kind of enabled. */ - phydev->eee_enabled = !linkmode_empty(phydev->advertising_eee); + phydev->eee_cfg.eee_enabled = !linkmode_empty(phydev->advertising_eee); /* Some PHYs may advertise, by default, not support EEE modes. So, * we need to clean them. */ - if (phydev->eee_enabled) + if (phydev->eee_cfg.eee_enabled) linkmode_and(phydev->advertising_eee, phydev->supported_eee, phydev->advertising_eee); diff --git a/include/linux/phy.h b/include/linux/phy.h index a98bc91a0cde..44890cdf40a2 100644 --- a/include/linux/phy.h +++ b/include/linux/phy.h @@ -601,7 +601,6 @@ struct macsec_ops; * @adv_old: Saved advertised while power saving for WoL * @supported_eee: supported PHY EEE linkmodes * @advertising_eee: Currently advertised EEE linkmodes - * @eee_enabled: Flag indicating whether the EEE feature is enabled * @enable_tx_lpi: When True, MAC should transmit LPI to PHY * @eee_cfg: User configuration of EEE * @lp_advertising: Current link partner advertised linkmodes @@ -721,7 +720,6 @@ struct phy_device { /* used for eee validation and configuration*/ __ETHTOOL_DECLARE_LINK_MODE_MASK(supported_eee); __ETHTOOL_DECLARE_LINK_MODE_MASK(advertising_eee); - bool eee_enabled; /* Host supported PHY interface types. Should be ignored if empty. */ DECLARE_PHY_INTERFACE_MASK(host_interfaces); -- cgit v1.2.3 From 221a9c1df790fa711d65daf5ba05d0addc279153 Mon Sep 17 00:00:00 2001 From: Breno Leitao Date: Thu, 14 Nov 2024 03:00:11 -0800 Subject: net: netpoll: Individualize the skb pool The current implementation of the netpoll system uses a global skb pool, which can lead to inefficient memory usage and waste when targets are disabled or no longer in use. This can result in a significant amount of memory being unnecessarily allocated and retained, potentially causing performance issues and limiting the availability of resources for other system components. Modify the netpoll system to assign a skb pool to each target instead of using a global one. This approach allows for more fine-grained control over memory allocation and deallocation, ensuring that resources are only allocated and retained as needed. Signed-off-by: Breno Leitao Link: https://patch.msgid.link/20241114-skb_buffers_v2-v3-1-9be9f52a8b69@debian.org Signed-off-by: Jakub Kicinski --- include/linux/netpoll.h | 1 + net/core/netpoll.c | 31 +++++++++++++------------------ 2 files changed, 14 insertions(+), 18 deletions(-) (limited to 'include/linux') diff --git a/include/linux/netpoll.h b/include/linux/netpoll.h index cd4e28db0cbd..77635b885c18 100644 --- a/include/linux/netpoll.h +++ b/include/linux/netpoll.h @@ -32,6 +32,7 @@ struct netpoll { bool ipv6; u16 local_port, remote_port; u8 remote_mac[ETH_ALEN]; + struct sk_buff_head skb_pool; }; struct netpoll_info { diff --git a/net/core/netpoll.c b/net/core/netpoll.c index 94b7f07a952f..719c9aae845f 100644 --- a/net/core/netpoll.c +++ b/net/core/netpoll.c @@ -45,9 +45,6 @@ #define MAX_UDP_CHUNK 1460 #define MAX_SKBS 32 - -static struct sk_buff_head skb_pool; - #define USEC_PER_POLL 50 #define MAX_SKB_SIZE \ @@ -234,20 +231,23 @@ void netpoll_poll_enable(struct net_device *dev) up(&ni->dev_lock); } -static void refill_skbs(void) +static void refill_skbs(struct netpoll *np) { + struct sk_buff_head *skb_pool; struct sk_buff *skb; unsigned long flags; - spin_lock_irqsave(&skb_pool.lock, flags); - while (skb_pool.qlen < MAX_SKBS) { + skb_pool = &np->skb_pool; + + spin_lock_irqsave(&skb_pool->lock, flags); + while (skb_pool->qlen < MAX_SKBS) { skb = alloc_skb(MAX_SKB_SIZE, GFP_ATOMIC); if (!skb) break; - __skb_queue_tail(&skb_pool, skb); + __skb_queue_tail(skb_pool, skb); } - spin_unlock_irqrestore(&skb_pool.lock, flags); + spin_unlock_irqrestore(&skb_pool->lock, flags); } static void zap_completion_queue(void) @@ -284,12 +284,12 @@ static struct sk_buff *find_skb(struct netpoll *np, int len, int reserve) struct sk_buff *skb; zap_completion_queue(); - refill_skbs(); + refill_skbs(np); repeat: skb = alloc_skb(len, GFP_ATOMIC); if (!skb) - skb = skb_dequeue(&skb_pool); + skb = skb_dequeue(&np->skb_pool); if (!skb) { if (++count < 10) { @@ -673,6 +673,8 @@ int netpoll_setup(struct netpoll *np) struct in_device *in_dev; int err; + skb_queue_head_init(&np->skb_pool); + rtnl_lock(); if (np->dev_name[0]) { struct net *net = current->nsproxy->net_ns; @@ -773,7 +775,7 @@ put_noaddr: } /* fill up the skb queue */ - refill_skbs(); + refill_skbs(np); err = __netpoll_setup(np, ndev); if (err) @@ -792,13 +794,6 @@ unlock: } EXPORT_SYMBOL(netpoll_setup); -static int __init netpoll_init(void) -{ - skb_queue_head_init(&skb_pool); - return 0; -} -core_initcall(netpoll_init); - static void rcu_cleanup_netpoll_info(struct rcu_head *rcu_head) { struct netpoll_info *npinfo = -- cgit v1.2.3 From 4b42fbc6bd8f73d9ded535d8c61ccaa837ff3bd4 Mon Sep 17 00:00:00 2001 From: Petr Machata Date: Thu, 14 Nov 2024 15:09:53 +0100 Subject: ndo_fdb_add: Add a parameter to report whether notification was sent Currently when FDB entries are added to or deleted from a VXLAN netdevice, the VXLAN driver emits one notification, including the VXLAN-specific attributes. The core however always sends a notification as well, a generic one. Thus two notifications are unnecessarily sent for these operations. A similar situation comes up with bridge driver, which also emits notifications on its own: # ip link add name vx type vxlan id 1000 dstport 4789 # bridge monitor fdb & [1] 1981693 # bridge fdb add de:ad:be:ef:13:37 dev vx self dst 192.0.2.1 de:ad:be:ef:13:37 dev vx dst 192.0.2.1 self permanent de:ad:be:ef:13:37 dev vx self permanent In order to prevent this duplicity, add a paremeter to ndo_fdb_add, bool *notified. The flag is primed to false, and if the callee sends a notification on its own, it sets it to true, thus informing the core that it should not generate another notification. Signed-off-by: Petr Machata Reviewed-by: Amit Cohen Reviewed-by: Nikolay Aleksandrov Link: https://patch.msgid.link/cbf6ae8195e85cbf922f8058ce4eba770f3b71ed.1731589511.git.petrm@nvidia.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/intel/i40e/i40e_main.c | 3 ++- drivers/net/ethernet/intel/ice/ice_main.c | 4 +++- drivers/net/ethernet/intel/igb/igb_main.c | 2 +- drivers/net/ethernet/intel/ixgbe/ixgbe_main.c | 2 +- drivers/net/ethernet/mscc/ocelot_net.c | 2 +- drivers/net/ethernet/qlogic/qlcnic/qlcnic_main.c | 2 +- drivers/net/macvlan.c | 2 +- drivers/net/vxlan/vxlan_core.c | 5 ++++- include/linux/netdevice.h | 5 ++++- net/bridge/br_fdb.c | 12 +++++++----- net/bridge/br_private.h | 2 +- net/core/rtnetlink.c | 9 ++++++--- 12 files changed, 32 insertions(+), 18 deletions(-) (limited to 'include/linux') diff --git a/drivers/net/ethernet/intel/i40e/i40e_main.c b/drivers/net/ethernet/intel/i40e/i40e_main.c index 55fb362eb508..ab5febf83ec3 100644 --- a/drivers/net/ethernet/intel/i40e/i40e_main.c +++ b/drivers/net/ethernet/intel/i40e/i40e_main.c @@ -13095,12 +13095,13 @@ static int i40e_get_phys_port_id(struct net_device *netdev, * @addr: the MAC address entry being added * @vid: VLAN ID * @flags: instructions from stack about fdb operation + * @notified: whether notification was emitted * @extack: netlink extended ack, unused currently */ static int i40e_ndo_fdb_add(struct ndmsg *ndm, struct nlattr *tb[], struct net_device *dev, const unsigned char *addr, u16 vid, - u16 flags, + u16 flags, bool *notified, struct netlink_ext_ack *extack) { struct i40e_netdev_priv *np = netdev_priv(dev); diff --git a/drivers/net/ethernet/intel/ice/ice_main.c b/drivers/net/ethernet/intel/ice/ice_main.c index a6f586f9bfd1..c875036f654b 100644 --- a/drivers/net/ethernet/intel/ice/ice_main.c +++ b/drivers/net/ethernet/intel/ice/ice_main.c @@ -6125,12 +6125,14 @@ ice_set_tx_maxrate(struct net_device *netdev, int queue_index, u32 maxrate) * @addr: the MAC address entry being added * @vid: VLAN ID * @flags: instructions from stack about fdb operation + * @notified: whether notification was emitted * @extack: netlink extended ack */ static int ice_fdb_add(struct ndmsg *ndm, struct nlattr __always_unused *tb[], struct net_device *dev, const unsigned char *addr, u16 vid, - u16 flags, struct netlink_ext_ack __always_unused *extack) + u16 flags, bool *notified, + struct netlink_ext_ack __always_unused *extack) { int err; diff --git a/drivers/net/ethernet/intel/igb/igb_main.c b/drivers/net/ethernet/intel/igb/igb_main.c index f1d088168723..f0528bd13184 100644 --- a/drivers/net/ethernet/intel/igb/igb_main.c +++ b/drivers/net/ethernet/intel/igb/igb_main.c @@ -2486,7 +2486,7 @@ static int igb_set_features(struct net_device *netdev, static int igb_ndo_fdb_add(struct ndmsg *ndm, struct nlattr *tb[], struct net_device *dev, const unsigned char *addr, u16 vid, - u16 flags, + u16 flags, bool *notified, struct netlink_ext_ack *extack) { /* guarantee we can provide a unique filter for the unicast address */ diff --git a/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c b/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c index 8b8404d8c946..adc9392463ce 100644 --- a/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c +++ b/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c @@ -9954,7 +9954,7 @@ static int ixgbe_set_features(struct net_device *netdev, static int ixgbe_ndo_fdb_add(struct ndmsg *ndm, struct nlattr *tb[], struct net_device *dev, const unsigned char *addr, u16 vid, - u16 flags, + u16 flags, bool *notified, struct netlink_ext_ack *extack) { /* guarantee we can provide a unique filter for the unicast address */ diff --git a/drivers/net/ethernet/mscc/ocelot_net.c b/drivers/net/ethernet/mscc/ocelot_net.c index 7c9540a71725..4f15ba2c5525 100644 --- a/drivers/net/ethernet/mscc/ocelot_net.c +++ b/drivers/net/ethernet/mscc/ocelot_net.c @@ -730,7 +730,7 @@ static void ocelot_get_stats64(struct net_device *dev, static int ocelot_port_fdb_add(struct ndmsg *ndm, struct nlattr *tb[], struct net_device *dev, const unsigned char *addr, - u16 vid, u16 flags, + u16 vid, u16 flags, bool *notified, struct netlink_ext_ack *extack) { struct ocelot_port_private *priv = netdev_priv(dev); diff --git a/drivers/net/ethernet/qlogic/qlcnic/qlcnic_main.c b/drivers/net/ethernet/qlogic/qlcnic/qlcnic_main.c index b3588a1ebc25..2484cebd97d4 100644 --- a/drivers/net/ethernet/qlogic/qlcnic/qlcnic_main.c +++ b/drivers/net/ethernet/qlogic/qlcnic/qlcnic_main.c @@ -394,7 +394,7 @@ static int qlcnic_fdb_del(struct ndmsg *ndm, struct nlattr *tb[], static int qlcnic_fdb_add(struct ndmsg *ndm, struct nlattr *tb[], struct net_device *netdev, const unsigned char *addr, u16 vid, u16 flags, - struct netlink_ext_ack *extack) + bool *notified, struct netlink_ext_ack *extack) { struct qlcnic_adapter *adapter = netdev_priv(netdev); int err = 0; diff --git a/drivers/net/macvlan.c b/drivers/net/macvlan.c index edbd5afcec41..dfb462e63248 100644 --- a/drivers/net/macvlan.c +++ b/drivers/net/macvlan.c @@ -1024,7 +1024,7 @@ static int macvlan_vlan_rx_kill_vid(struct net_device *dev, static int macvlan_fdb_add(struct ndmsg *ndm, struct nlattr *tb[], struct net_device *dev, const unsigned char *addr, u16 vid, - u16 flags, + u16 flags, bool *notified, struct netlink_ext_ack *extack) { struct macvlan_dev *vlan = netdev_priv(dev); diff --git a/drivers/net/vxlan/vxlan_core.c b/drivers/net/vxlan/vxlan_core.c index 42b07bc2b107..22f17c5c7549 100644 --- a/drivers/net/vxlan/vxlan_core.c +++ b/drivers/net/vxlan/vxlan_core.c @@ -1241,7 +1241,7 @@ static int vxlan_fdb_parse(struct nlattr *tb[], struct vxlan_dev *vxlan, static int vxlan_fdb_add(struct ndmsg *ndm, struct nlattr *tb[], struct net_device *dev, const unsigned char *addr, u16 vid, u16 flags, - struct netlink_ext_ack *extack) + bool *notified, struct netlink_ext_ack *extack) { struct vxlan_dev *vxlan = netdev_priv(dev); /* struct net *net = dev_net(vxlan->dev); */ @@ -1277,6 +1277,9 @@ static int vxlan_fdb_add(struct ndmsg *ndm, struct nlattr *tb[], nhid, true, extack); spin_unlock_bh(&vxlan->hash_lock[hash_index]); + if (!err) + *notified = true; + return err; } diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index 0aae346d919e..6a7fd191e1ee 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -1248,8 +1248,10 @@ struct netdev_net_notifier { * int (*ndo_fdb_add)(struct ndmsg *ndm, struct nlattr *tb[], * struct net_device *dev, * const unsigned char *addr, u16 vid, u16 flags, - * struct netlink_ext_ack *extack); + * bool *notified, struct netlink_ext_ack *extack); * Adds an FDB entry to dev for addr. + * Callee shall set *notified to true if it sent any appropriate + * notification(s). Otherwise core will send a generic one. * int (*ndo_fdb_del)(struct ndmsg *ndm, struct nlattr *tb[], * struct net_device *dev, * const unsigned char *addr, u16 vid) @@ -1525,6 +1527,7 @@ struct net_device_ops { const unsigned char *addr, u16 vid, u16 flags, + bool *notified, struct netlink_ext_ack *extack); int (*ndo_fdb_del)(struct ndmsg *ndm, struct nlattr *tb[], diff --git a/net/bridge/br_fdb.c b/net/bridge/br_fdb.c index 77f110035df1..5f29958f3ddd 100644 --- a/net/bridge/br_fdb.c +++ b/net/bridge/br_fdb.c @@ -1152,7 +1152,7 @@ static int fdb_add_entry(struct net_bridge *br, struct net_bridge_port *source, static int __br_fdb_add(struct ndmsg *ndm, struct net_bridge *br, struct net_bridge_port *p, const unsigned char *addr, u16 nlh_flags, u16 vid, struct nlattr *nfea_tb[], - struct netlink_ext_ack *extack) + bool *notified, struct netlink_ext_ack *extack) { int err = 0; @@ -1183,6 +1183,8 @@ static int __br_fdb_add(struct ndmsg *ndm, struct net_bridge *br, spin_unlock_bh(&br->hash_lock); } + if (!err) + *notified = true; return err; } @@ -1195,7 +1197,7 @@ static const struct nla_policy br_nda_fdb_pol[NFEA_MAX + 1] = { int br_fdb_add(struct ndmsg *ndm, struct nlattr *tb[], struct net_device *dev, const unsigned char *addr, u16 vid, u16 nlh_flags, - struct netlink_ext_ack *extack) + bool *notified, struct netlink_ext_ack *extack) { struct nlattr *nfea_tb[NFEA_MAX + 1], *attr; struct net_bridge_vlan_group *vg; @@ -1258,10 +1260,10 @@ int br_fdb_add(struct ndmsg *ndm, struct nlattr *tb[], /* VID was specified, so use it. */ err = __br_fdb_add(ndm, br, p, addr, nlh_flags, vid, nfea_tb, - extack); + notified, extack); } else { err = __br_fdb_add(ndm, br, p, addr, nlh_flags, 0, nfea_tb, - extack); + notified, extack); if (err || !vg || !vg->num_vlans) goto out; @@ -1273,7 +1275,7 @@ int br_fdb_add(struct ndmsg *ndm, struct nlattr *tb[], if (!br_vlan_should_use(v)) continue; err = __br_fdb_add(ndm, br, p, addr, nlh_flags, v->vid, - nfea_tb, extack); + nfea_tb, notified, extack); if (err) goto out; } diff --git a/net/bridge/br_private.h b/net/bridge/br_private.h index 041f6e571a20..ebfc59049ec1 100644 --- a/net/bridge/br_private.h +++ b/net/bridge/br_private.h @@ -858,7 +858,7 @@ int br_fdb_delete_bulk(struct nlmsghdr *nlh, struct net_device *dev, struct netlink_ext_ack *extack); int br_fdb_add(struct ndmsg *nlh, struct nlattr *tb[], struct net_device *dev, const unsigned char *addr, u16 vid, u16 nlh_flags, - struct netlink_ext_ack *extack); + bool *notified, struct netlink_ext_ack *extack); int br_fdb_dump(struct sk_buff *skb, struct netlink_callback *cb, struct net_device *dev, struct net_device *fdev, int *idx); int br_fdb_get(struct sk_buff *skb, struct nlattr *tb[], struct net_device *dev, diff --git a/net/core/rtnetlink.c b/net/core/rtnetlink.c index 327fa4957929..f31b2436cde5 100644 --- a/net/core/rtnetlink.c +++ b/net/core/rtnetlink.c @@ -4578,9 +4578,10 @@ static int rtnl_fdb_add(struct sk_buff *skb, struct nlmsghdr *nlh, netif_is_bridge_port(dev)) { struct net_device *br_dev = netdev_master_upper_dev_get(dev); const struct net_device_ops *ops = br_dev->netdev_ops; + bool notified = false; err = ops->ndo_fdb_add(ndm, tb, dev, addr, vid, - nlh->nlmsg_flags, extack); + nlh->nlmsg_flags, ¬ified, extack); if (err) goto out; else @@ -4589,16 +4590,18 @@ static int rtnl_fdb_add(struct sk_buff *skb, struct nlmsghdr *nlh, /* Embedded bridge, macvlan, and any other device support */ if ((ndm->ndm_flags & NTF_SELF)) { + bool notified = false; + if (dev->netdev_ops->ndo_fdb_add) err = dev->netdev_ops->ndo_fdb_add(ndm, tb, dev, addr, vid, nlh->nlmsg_flags, - extack); + ¬ified, extack); else err = ndo_dflt_fdb_add(ndm, tb, dev, addr, vid, nlh->nlmsg_flags); - if (!err) { + if (!err && !notified) { rtnl_fdb_notify(dev, addr, vid, RTM_NEWNEIGH, ndm->ndm_state); ndm->ndm_flags &= ~NTF_SELF; -- cgit v1.2.3 From 42575ad5aab932273475d1ec3e7881cb5a05420e Mon Sep 17 00:00:00 2001 From: Petr Machata Date: Thu, 14 Nov 2024 15:09:54 +0100 Subject: ndo_fdb_del: Add a parameter to report whether notification was sent In a similar fashion to ndo_fdb_add, which was covered in the previous patch, add the bool *notified argument to ndo_fdb_del. Callees that send a notification on their own set the flag to true. Signed-off-by: Petr Machata Reviewed-by: Amit Cohen Reviewed-by: Nikolay Aleksandrov Link: https://patch.msgid.link/06b1acf4953ef0a5ed153ef1f32d7292044f2be6.1731589511.git.petrm@nvidia.com Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/intel/ice/ice_main.c | 4 +++- drivers/net/ethernet/mscc/ocelot_net.c | 2 +- drivers/net/ethernet/qlogic/qlcnic/qlcnic_main.c | 2 +- drivers/net/macvlan.c | 2 +- drivers/net/vxlan/vxlan_core.c | 5 ++++- include/linux/netdevice.h | 9 +++++++-- net/bridge/br_fdb.c | 15 ++++++++------- net/bridge/br_private.h | 2 +- net/core/rtnetlink.c | 11 ++++++++--- 9 files changed, 34 insertions(+), 18 deletions(-) (limited to 'include/linux') diff --git a/drivers/net/ethernet/intel/ice/ice_main.c b/drivers/net/ethernet/intel/ice/ice_main.c index c875036f654b..b79848fe2a9e 100644 --- a/drivers/net/ethernet/intel/ice/ice_main.c +++ b/drivers/net/ethernet/intel/ice/ice_main.c @@ -6166,12 +6166,14 @@ ice_fdb_add(struct ndmsg *ndm, struct nlattr __always_unused *tb[], * @dev: the net device pointer * @addr: the MAC address entry being added * @vid: VLAN ID + * @notified: whether notification was emitted * @extack: netlink extended ack */ static int ice_fdb_del(struct ndmsg *ndm, __always_unused struct nlattr *tb[], struct net_device *dev, const unsigned char *addr, - __always_unused u16 vid, struct netlink_ext_ack *extack) + __always_unused u16 vid, bool *notified, + struct netlink_ext_ack *extack) { int err; diff --git a/drivers/net/ethernet/mscc/ocelot_net.c b/drivers/net/ethernet/mscc/ocelot_net.c index 4f15ba2c5525..558e03301aa8 100644 --- a/drivers/net/ethernet/mscc/ocelot_net.c +++ b/drivers/net/ethernet/mscc/ocelot_net.c @@ -744,7 +744,7 @@ static int ocelot_port_fdb_add(struct ndmsg *ndm, struct nlattr *tb[], static int ocelot_port_fdb_del(struct ndmsg *ndm, struct nlattr *tb[], struct net_device *dev, const unsigned char *addr, u16 vid, - struct netlink_ext_ack *extack) + bool *notified, struct netlink_ext_ack *extack) { struct ocelot_port_private *priv = netdev_priv(dev); struct ocelot_port *ocelot_port = &priv->port; diff --git a/drivers/net/ethernet/qlogic/qlcnic/qlcnic_main.c b/drivers/net/ethernet/qlogic/qlcnic/qlcnic_main.c index 2484cebd97d4..eb69121df726 100644 --- a/drivers/net/ethernet/qlogic/qlcnic/qlcnic_main.c +++ b/drivers/net/ethernet/qlogic/qlcnic/qlcnic_main.c @@ -367,7 +367,7 @@ static int qlcnic_set_mac(struct net_device *netdev, void *p) static int qlcnic_fdb_del(struct ndmsg *ndm, struct nlattr *tb[], struct net_device *netdev, - const unsigned char *addr, u16 vid, + const unsigned char *addr, u16 vid, bool *notified, struct netlink_ext_ack *extack) { struct qlcnic_adapter *adapter = netdev_priv(netdev); diff --git a/drivers/net/macvlan.c b/drivers/net/macvlan.c index dfb462e63248..fed4fe2a4748 100644 --- a/drivers/net/macvlan.c +++ b/drivers/net/macvlan.c @@ -1049,7 +1049,7 @@ static int macvlan_fdb_add(struct ndmsg *ndm, struct nlattr *tb[], static int macvlan_fdb_del(struct ndmsg *ndm, struct nlattr *tb[], struct net_device *dev, - const unsigned char *addr, u16 vid, + const unsigned char *addr, u16 vid, bool *notified, struct netlink_ext_ack *extack) { struct macvlan_dev *vlan = netdev_priv(dev); diff --git a/drivers/net/vxlan/vxlan_core.c b/drivers/net/vxlan/vxlan_core.c index 22f17c5c7549..9ea63059d52d 100644 --- a/drivers/net/vxlan/vxlan_core.c +++ b/drivers/net/vxlan/vxlan_core.c @@ -1319,7 +1319,7 @@ out: /* Delete entry (via netlink) */ static int vxlan_fdb_delete(struct ndmsg *ndm, struct nlattr *tb[], struct net_device *dev, - const unsigned char *addr, u16 vid, + const unsigned char *addr, u16 vid, bool *notified, struct netlink_ext_ack *extack) { struct vxlan_dev *vxlan = netdev_priv(dev); @@ -1341,6 +1341,9 @@ static int vxlan_fdb_delete(struct ndmsg *ndm, struct nlattr *tb[], true); spin_unlock_bh(&vxlan->hash_lock[hash_index]); + if (!err) + *notified = true; + return err; } diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index 6a7fd191e1ee..ecc686409161 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -1254,8 +1254,11 @@ struct netdev_net_notifier { * notification(s). Otherwise core will send a generic one. * int (*ndo_fdb_del)(struct ndmsg *ndm, struct nlattr *tb[], * struct net_device *dev, - * const unsigned char *addr, u16 vid) + * const unsigned char *addr, u16 vid + * bool *notified, struct netlink_ext_ack *extack); * Deletes the FDB entry from dev corresponding to addr. + * Callee shall set *notified to true if it sent any appropriate + * notification(s). Otherwise core will send a generic one. * int (*ndo_fdb_del_bulk)(struct nlmsghdr *nlh, struct net_device *dev, * struct netlink_ext_ack *extack); * int (*ndo_fdb_dump)(struct sk_buff *skb, struct netlink_callback *cb, @@ -1533,7 +1536,9 @@ struct net_device_ops { struct nlattr *tb[], struct net_device *dev, const unsigned char *addr, - u16 vid, struct netlink_ext_ack *extack); + u16 vid, + bool *notified, + struct netlink_ext_ack *extack); int (*ndo_fdb_del_bulk)(struct nlmsghdr *nlh, struct net_device *dev, struct netlink_ext_ack *extack); diff --git a/net/bridge/br_fdb.c b/net/bridge/br_fdb.c index 5f29958f3ddd..82bac2426631 100644 --- a/net/bridge/br_fdb.c +++ b/net/bridge/br_fdb.c @@ -1287,7 +1287,7 @@ out: static int fdb_delete_by_addr_and_port(struct net_bridge *br, const struct net_bridge_port *p, - const u8 *addr, u16 vlan) + const u8 *addr, u16 vlan, bool *notified) { struct net_bridge_fdb_entry *fdb; @@ -1296,18 +1296,19 @@ static int fdb_delete_by_addr_and_port(struct net_bridge *br, return -ENOENT; fdb_delete(br, fdb, true); + *notified = true; return 0; } static int __br_fdb_delete(struct net_bridge *br, const struct net_bridge_port *p, - const unsigned char *addr, u16 vid) + const unsigned char *addr, u16 vid, bool *notified) { int err; spin_lock_bh(&br->hash_lock); - err = fdb_delete_by_addr_and_port(br, p, addr, vid); + err = fdb_delete_by_addr_and_port(br, p, addr, vid, notified); spin_unlock_bh(&br->hash_lock); return err; @@ -1316,7 +1317,7 @@ static int __br_fdb_delete(struct net_bridge *br, /* Remove neighbor entry with RTM_DELNEIGH */ int br_fdb_delete(struct ndmsg *ndm, struct nlattr *tb[], struct net_device *dev, - const unsigned char *addr, u16 vid, + const unsigned char *addr, u16 vid, bool *notified, struct netlink_ext_ack *extack) { struct net_bridge_vlan_group *vg; @@ -1339,19 +1340,19 @@ int br_fdb_delete(struct ndmsg *ndm, struct nlattr *tb[], } if (vid) { - err = __br_fdb_delete(br, p, addr, vid); + err = __br_fdb_delete(br, p, addr, vid, notified); } else { struct net_bridge_vlan *v; err = -ENOENT; - err &= __br_fdb_delete(br, p, addr, 0); + err &= __br_fdb_delete(br, p, addr, 0, notified); if (!vg || !vg->num_vlans) return err; list_for_each_entry(v, &vg->vlan_list, vlist) { if (!br_vlan_should_use(v)) continue; - err &= __br_fdb_delete(br, p, addr, v->vid); + err &= __br_fdb_delete(br, p, addr, v->vid, notified); } } diff --git a/net/bridge/br_private.h b/net/bridge/br_private.h index ebfc59049ec1..9853cfbb9d14 100644 --- a/net/bridge/br_private.h +++ b/net/bridge/br_private.h @@ -853,7 +853,7 @@ void br_fdb_update(struct net_bridge *br, struct net_bridge_port *source, int br_fdb_delete(struct ndmsg *ndm, struct nlattr *tb[], struct net_device *dev, const unsigned char *addr, u16 vid, - struct netlink_ext_ack *extack); + bool *notified, struct netlink_ext_ack *extack); int br_fdb_delete_bulk(struct nlmsghdr *nlh, struct net_device *dev, struct netlink_ext_ack *extack); int br_fdb_add(struct ndmsg *nlh, struct nlattr *tb[], struct net_device *dev, diff --git a/net/core/rtnetlink.c b/net/core/rtnetlink.c index f31b2436cde5..dd142f444659 100644 --- a/net/core/rtnetlink.c +++ b/net/core/rtnetlink.c @@ -4701,11 +4701,13 @@ static int rtnl_fdb_del(struct sk_buff *skb, struct nlmsghdr *nlh, if ((!ndm->ndm_flags || ndm->ndm_flags & NTF_MASTER) && netif_is_bridge_port(dev)) { struct net_device *br_dev = netdev_master_upper_dev_get(dev); + bool notified = false; ops = br_dev->netdev_ops; if (!del_bulk) { if (ops->ndo_fdb_del) - err = ops->ndo_fdb_del(ndm, tb, dev, addr, vid, extack); + err = ops->ndo_fdb_del(ndm, tb, dev, addr, vid, + ¬ified, extack); } else { if (ops->ndo_fdb_del_bulk) err = ops->ndo_fdb_del_bulk(nlh, dev, extack); @@ -4719,10 +4721,13 @@ static int rtnl_fdb_del(struct sk_buff *skb, struct nlmsghdr *nlh, /* Embedded bridge, macvlan, and any other device support */ if (ndm->ndm_flags & NTF_SELF) { + bool notified = false; + ops = dev->netdev_ops; if (!del_bulk) { if (ops->ndo_fdb_del) - err = ops->ndo_fdb_del(ndm, tb, dev, addr, vid, extack); + err = ops->ndo_fdb_del(ndm, tb, dev, addr, vid, + ¬ified, extack); else err = ndo_dflt_fdb_del(ndm, tb, dev, addr, vid); } else { @@ -4733,7 +4738,7 @@ static int rtnl_fdb_del(struct sk_buff *skb, struct nlmsghdr *nlh, } if (!err) { - if (!del_bulk) + if (!del_bulk && !notified) rtnl_fdb_notify(dev, addr, vid, RTM_DELNEIGH, ndm->ndm_state); ndm->ndm_flags &= ~NTF_SELF; -- cgit v1.2.3 From 3ef66af31feaf5ff5dd73e63b1327789822ed476 Mon Sep 17 00:00:00 2001 From: Xuan Zhuo Date: Tue, 12 Nov 2024 09:29:20 +0800 Subject: virtio_ring: introduce add api for premapped Two APIs are introduced to submit premapped per-buffers. int virtqueue_add_inbuf_premapped(struct virtqueue *vq, struct scatterlist *sg, unsigned int num, void *data, void *ctx, gfp_t gfp); int virtqueue_add_outbuf_premapped(struct virtqueue *vq, struct scatterlist *sg, unsigned int num, void *data, gfp_t gfp); Signed-off-by: Xuan Zhuo Acked-by: Jason Wang Link: https://patch.msgid.link/20241112012928.102478-6-xuanzhuo@linux.alibaba.com Signed-off-by: Jakub Kicinski --- drivers/virtio/virtio_ring.c | 48 ++++++++++++++++++++++++++++++++++++++++++++ include/linux/virtio.h | 11 ++++++++++ 2 files changed, 59 insertions(+) (limited to 'include/linux') diff --git a/drivers/virtio/virtio_ring.c b/drivers/virtio/virtio_ring.c index fefa85a5e6b6..0842d27886e5 100644 --- a/drivers/virtio/virtio_ring.c +++ b/drivers/virtio/virtio_ring.c @@ -2276,6 +2276,29 @@ int virtqueue_add_outbuf(struct virtqueue *vq, } EXPORT_SYMBOL_GPL(virtqueue_add_outbuf); +/** + * virtqueue_add_outbuf_premapped - expose output buffers to other end + * @vq: the struct virtqueue we're talking about. + * @sg: scatterlist (must be well-formed and terminated!) + * @num: the number of entries in @sg readable by other side + * @data: the token identifying the buffer. + * @gfp: how to do memory allocations (if necessary). + * + * Caller must ensure we don't call this with other virtqueue operations + * at the same time (except where noted). + * + * Return: + * Returns zero or a negative error (ie. ENOSPC, ENOMEM, EIO). + */ +int virtqueue_add_outbuf_premapped(struct virtqueue *vq, + struct scatterlist *sg, unsigned int num, + void *data, + gfp_t gfp) +{ + return virtqueue_add(vq, &sg, num, 1, 0, data, NULL, true, gfp); +} +EXPORT_SYMBOL_GPL(virtqueue_add_outbuf_premapped); + /** * virtqueue_add_inbuf - expose input buffers to other end * @vq: the struct virtqueue we're talking about. @@ -2322,6 +2345,31 @@ int virtqueue_add_inbuf_ctx(struct virtqueue *vq, } EXPORT_SYMBOL_GPL(virtqueue_add_inbuf_ctx); +/** + * virtqueue_add_inbuf_premapped - expose input buffers to other end + * @vq: the struct virtqueue we're talking about. + * @sg: scatterlist (must be well-formed and terminated!) + * @num: the number of entries in @sg writable by other side + * @data: the token identifying the buffer. + * @ctx: extra context for the token + * @gfp: how to do memory allocations (if necessary). + * + * Caller must ensure we don't call this with other virtqueue operations + * at the same time (except where noted). + * + * Return: + * Returns zero or a negative error (ie. ENOSPC, ENOMEM, EIO). + */ +int virtqueue_add_inbuf_premapped(struct virtqueue *vq, + struct scatterlist *sg, unsigned int num, + void *data, + void *ctx, + gfp_t gfp) +{ + return virtqueue_add(vq, &sg, num, 0, 1, data, ctx, true, gfp); +} +EXPORT_SYMBOL_GPL(virtqueue_add_inbuf_premapped); + /** * virtqueue_dma_dev - get the dma dev * @_vq: the struct virtqueue we're talking about. diff --git a/include/linux/virtio.h b/include/linux/virtio.h index 306137a15d07..13b3f55abca3 100644 --- a/include/linux/virtio.h +++ b/include/linux/virtio.h @@ -56,6 +56,17 @@ int virtqueue_add_inbuf_ctx(struct virtqueue *vq, void *ctx, gfp_t gfp); +int virtqueue_add_inbuf_premapped(struct virtqueue *vq, + struct scatterlist *sg, unsigned int num, + void *data, + void *ctx, + gfp_t gfp); + +int virtqueue_add_outbuf_premapped(struct virtqueue *vq, + struct scatterlist *sg, unsigned int num, + void *data, + gfp_t gfp); + int virtqueue_add_sgs(struct virtqueue *vq, struct scatterlist *sgs[], unsigned int out_sgs, -- cgit v1.2.3 From 880ebcbe06636c42cfb328039581e177c6cd6ba5 Mon Sep 17 00:00:00 2001 From: Xuan Zhuo Date: Tue, 12 Nov 2024 09:29:22 +0800 Subject: virtio_ring: remove API virtqueue_set_dma_premapped Now, this API is useless. remove it. Signed-off-by: Xuan Zhuo Acked-by: Jason Wang Link: https://patch.msgid.link/20241112012928.102478-8-xuanzhuo@linux.alibaba.com Signed-off-by: Jakub Kicinski --- drivers/net/virtio_net.c | 13 ------------ drivers/virtio/virtio_ring.c | 48 -------------------------------------------- include/linux/virtio.h | 2 -- 3 files changed, 63 deletions(-) (limited to 'include/linux') diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c index 8aca4a3fc7e8..183ad5e6bef0 100644 --- a/drivers/net/virtio_net.c +++ b/drivers/net/virtio_net.c @@ -6168,15 +6168,6 @@ err_ctrl: return -ENOMEM; } -static void virtnet_rq_set_premapped(struct virtnet_info *vi) -{ - int i; - - for (i = 0; i < vi->max_queue_pairs; i++) - /* error should never happen */ - BUG_ON(virtqueue_set_dma_premapped(vi->rq[i].vq)); -} - static int init_vqs(struct virtnet_info *vi) { int ret; @@ -6190,10 +6181,6 @@ static int init_vqs(struct virtnet_info *vi) if (ret) goto err_free; - /* disable for big mode */ - if (!vi->big_packets || vi->mergeable_rx_bufs) - virtnet_rq_set_premapped(vi); - cpus_read_lock(); virtnet_set_affinity(vi); cpus_read_unlock(); diff --git a/drivers/virtio/virtio_ring.c b/drivers/virtio/virtio_ring.c index 0842d27886e5..8167be01b400 100644 --- a/drivers/virtio/virtio_ring.c +++ b/drivers/virtio/virtio_ring.c @@ -180,9 +180,6 @@ struct vring_virtqueue { /* Host publishes avail event idx */ bool event; - /* Do DMA mapping by driver */ - bool premapped; - /* Head of free buffer list. */ unsigned int free_head; /* Number we've added since last sync. */ @@ -2098,7 +2095,6 @@ static struct virtqueue *vring_create_virtqueue_packed( vq->packed_ring = true; vq->dma_dev = dma_dev; vq->use_dma_api = vring_use_dma_api(vdev); - vq->premapped = false; vq->indirect = virtio_has_feature(vdev, VIRTIO_RING_F_INDIRECT_DESC) && !context; @@ -2691,7 +2687,6 @@ static struct virtqueue *__vring_new_virtqueue(unsigned int index, #endif vq->dma_dev = dma_dev; vq->use_dma_api = vring_use_dma_api(vdev); - vq->premapped = false; vq->indirect = virtio_has_feature(vdev, VIRTIO_RING_F_INDIRECT_DESC) && !context; @@ -2818,49 +2813,6 @@ int virtqueue_resize(struct virtqueue *_vq, u32 num, } EXPORT_SYMBOL_GPL(virtqueue_resize); -/** - * virtqueue_set_dma_premapped - set the vring premapped mode - * @_vq: the struct virtqueue we're talking about. - * - * Enable the premapped mode of the vq. - * - * The vring in premapped mode does not do dma internally, so the driver must - * do dma mapping in advance. The driver must pass the dma_address through - * dma_address of scatterlist. When the driver got a used buffer from - * the vring, it has to unmap the dma address. - * - * This function must be called immediately after creating the vq, or after vq - * reset, and before adding any buffers to it. - * - * Caller must ensure we don't call this with other virtqueue operations - * at the same time (except where noted). - * - * Returns zero or a negative error. - * 0: success. - * -EINVAL: too late to enable premapped mode, the vq already contains buffers. - */ -int virtqueue_set_dma_premapped(struct virtqueue *_vq) -{ - struct vring_virtqueue *vq = to_vvq(_vq); - u32 num; - - START_USE(vq); - - num = vq->packed_ring ? vq->packed.vring.num : vq->split.vring.num; - - if (num != vq->vq.num_free) { - END_USE(vq); - return -EINVAL; - } - - vq->premapped = true; - - END_USE(vq); - - return 0; -} -EXPORT_SYMBOL_GPL(virtqueue_set_dma_premapped); - /** * virtqueue_reset - detach and recycle all unused buffers * @_vq: the struct virtqueue we're talking about. diff --git a/include/linux/virtio.h b/include/linux/virtio.h index 13b3f55abca3..338e0f5efb4b 100644 --- a/include/linux/virtio.h +++ b/include/linux/virtio.h @@ -93,8 +93,6 @@ bool virtqueue_enable_cb(struct virtqueue *vq); unsigned virtqueue_enable_cb_prepare(struct virtqueue *vq); -int virtqueue_set_dma_premapped(struct virtqueue *_vq); - bool virtqueue_poll(struct virtqueue *vq, unsigned); bool virtqueue_enable_cb_delayed(struct virtqueue *vq); -- cgit v1.2.3 From dab78a1745ab3c6001e1e4d50a9d09efef8e260d Mon Sep 17 00:00:00 2001 From: Philo Lu Date: Thu, 14 Nov 2024 18:52:05 +0800 Subject: net/udp: Add 4-tuple hash list basis Add a new hash list, hash4, in udp table. It will be used to implement 4-tuple hash for connected udp sockets. This patch adds the hlist to table, and implements helpers and the initialization. 4-tuple hash is implemented in the following patch. hash4 uses hlist_nulls to avoid moving wrongly onto another hlist due to concurrent rehash, because rehash() can happen with lookup(). Co-developed-by: Cambda Zhu Signed-off-by: Cambda Zhu Co-developed-by: Fred Chen Signed-off-by: Fred Chen Co-developed-by: Yubing Qiu Signed-off-by: Yubing Qiu Signed-off-by: Philo Lu Acked-by: Willem de Bruijn Acked-by: Paolo Abeni Signed-off-by: David S. Miller --- include/linux/udp.h | 11 +++++++ include/net/udp.h | 85 +++++++++++++++++++++++++++++++++++++++++++++++++++-- net/ipv4/udp.c | 6 ++-- 3 files changed, 97 insertions(+), 5 deletions(-) (limited to 'include/linux') diff --git a/include/linux/udp.h b/include/linux/udp.h index 3eb3f2b9a2a0..0807e21cfec9 100644 --- a/include/linux/udp.h +++ b/include/linux/udp.h @@ -56,6 +56,12 @@ struct udp_sock { int pending; /* Any pending frames ? */ __u8 encap_type; /* Is this an Encapsulation socket? */ +#if !IS_ENABLED(CONFIG_BASE_SMALL) + /* For UDP 4-tuple hash */ + __u16 udp_lrpa_hash; + struct hlist_nulls_node udp_lrpa_node; +#endif + /* * Following member retains the information to create a UDP header * when the socket is uncorked. @@ -206,6 +212,11 @@ static inline void udp_allow_gso(struct sock *sk) #define udp_portaddr_for_each_entry_rcu(__sk, list) \ hlist_for_each_entry_rcu(__sk, list, __sk_common.skc_portaddr_node) +#if !IS_ENABLED(CONFIG_BASE_SMALL) +#define udp_lrpa_for_each_entry_rcu(__up, node, list) \ + hlist_nulls_for_each_entry_rcu(__up, node, list, udp_lrpa_node) +#endif + #define IS_UDPLITE(__sk) (__sk->sk_protocol == IPPROTO_UDPLITE) #endif /* _LINUX_UDP_H */ diff --git a/include/net/udp.h b/include/net/udp.h index 62a7207e65f2..edb669967130 100644 --- a/include/net/udp.h +++ b/include/net/udp.h @@ -50,14 +50,21 @@ struct udp_skb_cb { #define UDP_SKB_CB(__skb) ((struct udp_skb_cb *)((__skb)->cb)) /** - * struct udp_hslot - UDP hash slot used by udp_table.hash + * struct udp_hslot - UDP hash slot used by udp_table.hash/hash4 * * @head: head of list of sockets + * @nulls_head: head of list of sockets, only used by hash4 * @count: number of sockets in 'head' list * @lock: spinlock protecting changes to head/count */ struct udp_hslot { - struct hlist_head head; + union { + struct hlist_head head; + /* hash4 uses hlist_nulls to avoid moving wrongly onto another + * hlist, because rehash() can happen with lookup(). + */ + struct hlist_nulls_head nulls_head; + }; int count; spinlock_t lock; } __aligned(2 * sizeof(long)); @@ -82,12 +89,17 @@ struct udp_hslot_main { * * @hash: hash table, sockets are hashed on (local port) * @hash2: hash table, sockets are hashed on (local port, local address) + * @hash4: hash table, connected sockets are hashed on + * (local port, local address, remote port, remote address) * @mask: number of slots in hash tables, minus 1 * @log: log2(number of slots in hash table) */ struct udp_table { struct udp_hslot *hash; struct udp_hslot_main *hash2; +#if !IS_ENABLED(CONFIG_BASE_SMALL) + struct udp_hslot *hash4; +#endif unsigned int mask; unsigned int log; }; @@ -114,13 +126,80 @@ static inline struct udp_hslot *udp_hashslot2(struct udp_table *table, static inline void udp_table_hash4_init(struct udp_table *table) { } + +static inline struct udp_hslot *udp_hashslot4(struct udp_table *table, + unsigned int hash) +{ + BUILD_BUG(); + return NULL; +} + +static inline bool udp_hashed4(const struct sock *sk) +{ + return false; +} + +static inline unsigned int udp_hash4_slot_size(void) +{ + return 0; +} + +static inline bool udp_has_hash4(const struct udp_hslot *hslot2) +{ + return false; +} + +static inline void udp_hash4_inc(struct udp_hslot *hslot2) +{ +} + +static inline void udp_hash4_dec(struct udp_hslot *hslot2) +{ +} #else /* !CONFIG_BASE_SMALL */ /* Must be called with table->hash2 initialized */ static inline void udp_table_hash4_init(struct udp_table *table) { - for (int i = 0; i <= table->mask; i++) + table->hash4 = (void *)(table->hash2 + (table->mask + 1)); + for (int i = 0; i <= table->mask; i++) { table->hash2[i].hash4_cnt = 0; + + INIT_HLIST_NULLS_HEAD(&table->hash4[i].nulls_head, i); + table->hash4[i].count = 0; + spin_lock_init(&table->hash4[i].lock); + } +} + +static inline struct udp_hslot *udp_hashslot4(struct udp_table *table, + unsigned int hash) +{ + return &table->hash4[hash & table->mask]; +} + +static inline bool udp_hashed4(const struct sock *sk) +{ + return !hlist_nulls_unhashed(&udp_sk(sk)->udp_lrpa_node); +} + +static inline unsigned int udp_hash4_slot_size(void) +{ + return sizeof(struct udp_hslot); +} + +static inline bool udp_has_hash4(const struct udp_hslot *hslot2) +{ + return UDP_HSLOT_MAIN(hslot2)->hash4_cnt; +} + +static inline void udp_hash4_inc(struct udp_hslot *hslot2) +{ + UDP_HSLOT_MAIN(hslot2)->hash4_cnt++; +} + +static inline void udp_hash4_dec(struct udp_hslot *hslot2) +{ + UDP_HSLOT_MAIN(hslot2)->hash4_cnt--; } #endif /* CONFIG_BASE_SMALL */ diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c index 2fdac5fae2a8..0bc0881d6569 100644 --- a/net/ipv4/udp.c +++ b/net/ipv4/udp.c @@ -3427,7 +3427,8 @@ void __init udp_table_init(struct udp_table *table, const char *name) { unsigned int i, slot_size; - slot_size = sizeof(struct udp_hslot) + sizeof(struct udp_hslot_main); + slot_size = sizeof(struct udp_hslot) + sizeof(struct udp_hslot_main) + + udp_hash4_slot_size(); table->hash = alloc_large_system_hash(name, slot_size, uhash_entries, @@ -3482,7 +3483,8 @@ static struct udp_table __net_init *udp_pernet_table_alloc(unsigned int hash_ent if (!udptable) goto out; - slot_size = sizeof(struct udp_hslot) + sizeof(struct udp_hslot_main); + slot_size = sizeof(struct udp_hslot) + sizeof(struct udp_hslot_main) + + udp_hash4_slot_size(); udptable->hash = vmalloc_huge(hash_entries * slot_size, GFP_KERNEL_ACCOUNT); if (!udptable->hash) -- cgit v1.2.3 From 1cfb5e57886aa69a992ff0ebd32e4651eb0fc995 Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Fri, 15 Nov 2024 12:43:03 -0800 Subject: Revert "net: ethtool: Avoid thousands of -Wflex-array-member-not-at-end warnings" This reverts commit 3bd9b9abdf1563a22041b7255baea6d449902f1a. We cannot use the new tagged struct group because it throws C++ errors even under "extern C". Signed-off-by: Kees Cook Link: https://patch.msgid.link/20241115204308.3821419-1-kees@kernel.org Signed-off-by: Jakub Kicinski --- drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c | 6 +++--- drivers/net/ethernet/chelsio/cxgb4/cxgb4_ethtool.c | 4 ++-- drivers/net/ethernet/chelsio/cxgb4vf/cxgb4vf_main.c | 2 +- drivers/net/ethernet/cisco/enic/enic_ethtool.c | 2 +- drivers/net/ethernet/qlogic/qede/qede_ethtool.c | 4 ++-- include/linux/ethtool.h | 2 +- net/ethtool/ioctl.c | 2 +- net/ethtool/linkinfo.c | 8 ++++---- net/ethtool/linkmodes.c | 18 +++++++----------- 9 files changed, 22 insertions(+), 26 deletions(-) (limited to 'include/linux') diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c b/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c index cfd2c65b1c90..061a40b1974b 100644 --- a/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c @@ -2780,7 +2780,7 @@ u32 bnxt_fw_to_ethtool_speed(u16 fw_link_speed) static void bnxt_get_default_speeds(struct ethtool_link_ksettings *lk_ksettings, struct bnxt_link_info *link_info) { - struct ethtool_link_settings_hdr *base = &lk_ksettings->base; + struct ethtool_link_settings *base = &lk_ksettings->base; if (link_info->link_state == BNXT_LINK_STATE_UP) { base->speed = bnxt_fw_to_ethtool_speed(link_info->link_speed); @@ -2799,7 +2799,7 @@ static void bnxt_get_default_speeds(struct ethtool_link_ksettings *lk_ksettings, static int bnxt_get_link_ksettings(struct net_device *dev, struct ethtool_link_ksettings *lk_ksettings) { - struct ethtool_link_settings_hdr *base = &lk_ksettings->base; + struct ethtool_link_settings *base = &lk_ksettings->base; enum ethtool_link_mode_bit_indices link_mode; struct bnxt *bp = netdev_priv(dev); struct bnxt_link_info *link_info; @@ -3022,9 +3022,9 @@ u16 bnxt_get_fw_auto_link_speeds(const unsigned long *mode) static int bnxt_set_link_ksettings(struct net_device *dev, const struct ethtool_link_ksettings *lk_ksettings) { - const struct ethtool_link_settings_hdr *base = &lk_ksettings->base; struct bnxt *bp = netdev_priv(dev); struct bnxt_link_info *link_info = &bp->link_info; + const struct ethtool_link_settings *base = &lk_ksettings->base; bool set_pause = false; u32 speed, lanes = 0; int rc = 0; diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_ethtool.c b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_ethtool.c index 45d28a65347e..7f3f5afa864f 100644 --- a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_ethtool.c +++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_ethtool.c @@ -662,8 +662,8 @@ static unsigned int lmm_to_fw_caps(const unsigned long *link_mode_mask) static int get_link_ksettings(struct net_device *dev, struct ethtool_link_ksettings *link_ksettings) { - struct ethtool_link_settings_hdr *base = &link_ksettings->base; struct port_info *pi = netdev_priv(dev); + struct ethtool_link_settings *base = &link_ksettings->base; /* For the nonce, the Firmware doesn't send up Port State changes * when the Virtual Interface attached to the Port is down. So @@ -717,9 +717,9 @@ static int get_link_ksettings(struct net_device *dev, static int set_link_ksettings(struct net_device *dev, const struct ethtool_link_ksettings *link_ksettings) { - const struct ethtool_link_settings_hdr *base = &link_ksettings->base; struct port_info *pi = netdev_priv(dev); struct link_config *lc = &pi->link_cfg; + const struct ethtool_link_settings *base = &link_ksettings->base; struct link_config old_lc; unsigned int fw_caps; int ret = 0; diff --git a/drivers/net/ethernet/chelsio/cxgb4vf/cxgb4vf_main.c b/drivers/net/ethernet/chelsio/cxgb4vf/cxgb4vf_main.c index 61d08547e3f9..2fbe0f059a0b 100644 --- a/drivers/net/ethernet/chelsio/cxgb4vf/cxgb4vf_main.c +++ b/drivers/net/ethernet/chelsio/cxgb4vf/cxgb4vf_main.c @@ -1436,8 +1436,8 @@ static void fw_caps_to_lmm(enum fw_port_type port_type, static int cxgb4vf_get_link_ksettings(struct net_device *dev, struct ethtool_link_ksettings *link_ksettings) { - struct ethtool_link_settings_hdr *base = &link_ksettings->base; struct port_info *pi = netdev_priv(dev); + struct ethtool_link_settings *base = &link_ksettings->base; /* For the nonce, the Firmware doesn't send up Port State changes * when the Virtual Interface attached to the Port is down. So diff --git a/drivers/net/ethernet/cisco/enic/enic_ethtool.c b/drivers/net/ethernet/cisco/enic/enic_ethtool.c index 95b071153fed..d607b4f0542c 100644 --- a/drivers/net/ethernet/cisco/enic/enic_ethtool.c +++ b/drivers/net/ethernet/cisco/enic/enic_ethtool.c @@ -129,8 +129,8 @@ static void enic_intr_coal_set_rx(struct enic *enic, u32 timer) static int enic_get_ksettings(struct net_device *netdev, struct ethtool_link_ksettings *ecmd) { - struct ethtool_link_settings_hdr *base = &ecmd->base; struct enic *enic = netdev_priv(netdev); + struct ethtool_link_settings *base = &ecmd->base; ethtool_link_ksettings_add_link_mode(ecmd, supported, 10000baseT_Full); diff --git a/drivers/net/ethernet/qlogic/qede/qede_ethtool.c b/drivers/net/ethernet/qlogic/qede/qede_ethtool.c index c553da16d4b1..e50e1df0a433 100644 --- a/drivers/net/ethernet/qlogic/qede/qede_ethtool.c +++ b/drivers/net/ethernet/qlogic/qede/qede_ethtool.c @@ -504,7 +504,7 @@ static int qede_get_link_ksettings(struct net_device *dev, struct ethtool_link_ksettings *cmd) { typeof(cmd->link_modes) *link_modes = &cmd->link_modes; - struct ethtool_link_settings_hdr *base = &cmd->base; + struct ethtool_link_settings *base = &cmd->base; struct qede_dev *edev = netdev_priv(dev); struct qed_link_output current_link; @@ -537,7 +537,7 @@ static int qede_get_link_ksettings(struct net_device *dev, static int qede_set_link_ksettings(struct net_device *dev, const struct ethtool_link_ksettings *cmd) { - const struct ethtool_link_settings_hdr *base = &cmd->base; + const struct ethtool_link_settings *base = &cmd->base; const struct ethtool_forced_speed_map *map; struct qede_dev *edev = netdev_priv(dev); struct qed_link_output current_link; diff --git a/include/linux/ethtool.h b/include/linux/ethtool.h index 299280c94d07..b8b935b52603 100644 --- a/include/linux/ethtool.h +++ b/include/linux/ethtool.h @@ -211,7 +211,7 @@ void ethtool_rxfh_context_lost(struct net_device *dev, u32 context_id); * fields, but they are allowed to overwrite them (will be ignored). */ struct ethtool_link_ksettings { - struct ethtool_link_settings_hdr base; + struct ethtool_link_settings base; struct { __ETHTOOL_DECLARE_LINK_MODE_MASK(supported); __ETHTOOL_DECLARE_LINK_MODE_MASK(advertising); diff --git a/net/ethtool/ioctl.c b/net/ethtool/ioctl.c index d86399bcf223..61df8ce44379 100644 --- a/net/ethtool/ioctl.c +++ b/net/ethtool/ioctl.c @@ -425,7 +425,7 @@ convert_link_ksettings_to_legacy_settings( /* layout of the struct passed from/to userland */ struct ethtool_link_usettings { - struct ethtool_link_settings_hdr base; + struct ethtool_link_settings base; struct { __u32 supported[__ETHTOOL_LINK_MODE_MASK_NU32]; __u32 advertising[__ETHTOOL_LINK_MODE_MASK_NU32]; diff --git a/net/ethtool/linkinfo.c b/net/ethtool/linkinfo.c index 2d5bc57160be..30b8ce275159 100644 --- a/net/ethtool/linkinfo.c +++ b/net/ethtool/linkinfo.c @@ -8,9 +8,9 @@ struct linkinfo_req_info { }; struct linkinfo_reply_data { - struct ethnl_reply_data base; - struct ethtool_link_ksettings ksettings; - struct ethtool_link_settings_hdr *lsettings; + struct ethnl_reply_data base; + struct ethtool_link_ksettings ksettings; + struct ethtool_link_settings *lsettings; }; #define LINKINFO_REPDATA(__reply_base) \ @@ -98,7 +98,7 @@ static int ethnl_set_linkinfo(struct ethnl_req_info *req_info, struct genl_info *info) { struct ethtool_link_ksettings ksettings = {}; - struct ethtool_link_settings_hdr *lsettings; + struct ethtool_link_settings *lsettings; struct net_device *dev = req_info->dev; struct nlattr **tb = info->attrs; bool mod = false; diff --git a/net/ethtool/linkmodes.c b/net/ethtool/linkmodes.c index 17e49cf89f03..259cd9ef1f2a 100644 --- a/net/ethtool/linkmodes.c +++ b/net/ethtool/linkmodes.c @@ -11,10 +11,10 @@ struct linkmodes_req_info { }; struct linkmodes_reply_data { - struct ethnl_reply_data base; - struct ethtool_link_ksettings ksettings; - struct ethtool_link_settings_hdr *lsettings; - bool peer_empty; + struct ethnl_reply_data base; + struct ethtool_link_ksettings ksettings; + struct ethtool_link_settings *lsettings; + bool peer_empty; }; #define LINKMODES_REPDATA(__reply_base) \ @@ -62,12 +62,10 @@ static int linkmodes_reply_size(const struct ethnl_req_info *req_base, { const struct linkmodes_reply_data *data = LINKMODES_REPDATA(reply_base); const struct ethtool_link_ksettings *ksettings = &data->ksettings; + const struct ethtool_link_settings *lsettings = &ksettings->base; bool compact = req_base->flags & ETHTOOL_FLAG_COMPACT_BITSETS; - const struct ethtool_link_settings_hdr *lsettings; int len, ret; - lsettings = &ksettings->base; - len = nla_total_size(sizeof(u8)) /* LINKMODES_AUTONEG */ + nla_total_size(sizeof(u32)) /* LINKMODES_SPEED */ + nla_total_size(sizeof(u32)) /* LINKMODES_LANES */ @@ -105,12 +103,10 @@ static int linkmodes_fill_reply(struct sk_buff *skb, { const struct linkmodes_reply_data *data = LINKMODES_REPDATA(reply_base); const struct ethtool_link_ksettings *ksettings = &data->ksettings; + const struct ethtool_link_settings *lsettings = &ksettings->base; bool compact = req_base->flags & ETHTOOL_FLAG_COMPACT_BITSETS; - const struct ethtool_link_settings_hdr *lsettings; int ret; - lsettings = &ksettings->base; - if (nla_put_u8(skb, ETHTOOL_A_LINKMODES_AUTONEG, lsettings->autoneg)) return -EMSGSIZE; @@ -241,7 +237,7 @@ static int ethnl_update_linkmodes(struct genl_info *info, struct nlattr **tb, struct ethtool_link_ksettings *ksettings, bool *mod, const struct net_device *dev) { - struct ethtool_link_settings_hdr *lsettings = &ksettings->base; + struct ethtool_link_settings *lsettings = &ksettings->base; bool req_speed, req_lanes, req_duplex; const struct nlattr *master_slave_cfg, *lanes_cfg; int ret; -- cgit v1.2.3 From a57d5a72f8dec7db8a79d0016fb0a3bdecc82b56 Mon Sep 17 00:00:00 2001 From: Breno Leitao Date: Mon, 18 Nov 2024 03:15:18 -0800 Subject: netpoll: Use rcu_access_pointer() in netpoll_poll_lock The ndev->npinfo pointer in netpoll_poll_lock() is RCU-protected but is being accessed directly for a NULL check. While no RCU read lock is held in this context, we should still use proper RCU primitives for consistency and correctness. Replace the direct NULL check with rcu_access_pointer(), which is the appropriate primitive when only checking for NULL without dereferencing the pointer. This function provides the necessary ordering guarantees without requiring RCU read-side protection. Fixes: bea3348eef27 ("[NET]: Make NAPI polling independent of struct net_device objects.") Signed-off-by: Breno Leitao Reviewed-by: Michal Kubiak Link: https://patch.msgid.link/20241118-netpoll_rcu-v1-2-a1888dcb4a02@debian.org Signed-off-by: Jakub Kicinski --- include/linux/netpoll.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include/linux') diff --git a/include/linux/netpoll.h b/include/linux/netpoll.h index cd4e28db0cbd..959a4daacea1 100644 --- a/include/linux/netpoll.h +++ b/include/linux/netpoll.h @@ -72,7 +72,7 @@ static inline void *netpoll_poll_lock(struct napi_struct *napi) { struct net_device *dev = napi->dev; - if (dev && dev->npinfo) { + if (dev && rcu_access_pointer(dev->npinfo)) { int owner = smp_processor_id(); while (cmpxchg(&napi->poll_owner, -1, owner) != -1) -- cgit v1.2.3