From 29a1f599c0cc37004f92ba455d1ccda3db0b6a94 Mon Sep 17 00:00:00 2001 From: Baolin Wang Date: Thu, 14 Dec 2017 13:31:43 +0800 Subject: rtc: Add tracepoints for RTC system It will be more helpful to add some tracepoints to track RTC actions when debugging RTC driver. Below sample is that we set/read the RTC time, then set 2 alarms, so we can see the trace logs: set/read RTC time: kworker/0:1-67 [000] 21.814245: rtc_set_time: UTC (1510301580) (0) kworker/0:1-67 [000] 21.814312: rtc_read_time: UTC (1510301580) (0) set the first alarm timer: kworker/0:1-67 [000] 21.829238: rtc_timer_enqueue: RTC timer:(ffffffc15eb49bc8) expires:1510301700000000000 period:0 kworker/0:1-67 [000] 22.018279: rtc_set_alarm: UTC (1510301700) (0) set the second alarm timer: kworker/0:1-67 [000] 22.230284: rtc_timer_enqueue: RTC timer:(ffffff80088e6430) expires:1510301820000000000 period:0 the first alarm timer was expired: kworker/0:1-67 [000] 145.155584: rtc_timer_dequeue: RTC timer:(ffffffc15eb49bc8) expires:1510301700000000000 period:0 kworker/0:1-67 [000] 145.155593: rtc_timer_fired: RTC timer:(ffffffc15eb49bc8) expires:1510301700000000000 period:0 kworker/0:1-67 [000] 145.172504: rtc_set_alarm: UTC (1510301820) (0) the second alarm timer was expired: kworker/0:1-67 [000] 269.102353: rtc_timer_dequeue: RTC timer:(ffffff80088e6430) expires:1510301820000000000 period:0 kworker/0:1-67 [000] 269.102360: rtc_timer_fired: RTC timer:(ffffff80088e6430) expires:1510301820000000000 period:0 disable alarm irq: kworker/0:1-67 [000] 269.102469: rtc_alarm_irq_enable: disable RTC alarm IRQ (0) Signed-off-by: Baolin Wang Signed-off-by: Alexandre Belloni --- include/trace/events/rtc.h | 206 +++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 206 insertions(+) create mode 100644 include/trace/events/rtc.h (limited to 'include/trace') diff --git a/include/trace/events/rtc.h b/include/trace/events/rtc.h new file mode 100644 index 000000000000..621333f1c890 --- /dev/null +++ b/include/trace/events/rtc.h @@ -0,0 +1,206 @@ +#undef TRACE_SYSTEM +#define TRACE_SYSTEM rtc + +#if !defined(_TRACE_RTC_H) || defined(TRACE_HEADER_MULTI_READ) +#define _TRACE_RTC_H + +#include +#include + +DECLARE_EVENT_CLASS(rtc_time_alarm_class, + + TP_PROTO(time64_t secs, int err), + + TP_ARGS(secs, err), + + TP_STRUCT__entry( + __field(time64_t, secs) + __field(int, err) + ), + + TP_fast_assign( + __entry->secs = secs; + __entry->err = err; + ), + + TP_printk("UTC (%lld) (%d)", + __entry->secs, __entry->err + ) +); + +DEFINE_EVENT(rtc_time_alarm_class, rtc_set_time, + + TP_PROTO(time64_t secs, int err), + + TP_ARGS(secs, err) +); + +DEFINE_EVENT(rtc_time_alarm_class, rtc_read_time, + + TP_PROTO(time64_t secs, int err), + + TP_ARGS(secs, err) +); + +DEFINE_EVENT(rtc_time_alarm_class, rtc_set_alarm, + + TP_PROTO(time64_t secs, int err), + + TP_ARGS(secs, err) +); + +DEFINE_EVENT(rtc_time_alarm_class, rtc_read_alarm, + + TP_PROTO(time64_t secs, int err), + + TP_ARGS(secs, err) +); + +TRACE_EVENT(rtc_irq_set_freq, + + TP_PROTO(int freq, int err), + + TP_ARGS(freq, err), + + TP_STRUCT__entry( + __field(int, freq) + __field(int, err) + ), + + TP_fast_assign( + __entry->freq = freq; + __entry->err = err; + ), + + TP_printk("set RTC periodic IRQ frequency:%u (%d)", + __entry->freq, __entry->err + ) +); + +TRACE_EVENT(rtc_irq_set_state, + + TP_PROTO(int enabled, int err), + + TP_ARGS(enabled, err), + + TP_STRUCT__entry( + __field(int, enabled) + __field(int, err) + ), + + TP_fast_assign( + __entry->enabled = enabled; + __entry->err = err; + ), + + TP_printk("%s RTC 2^N Hz periodic IRQs (%d)", + __entry->enabled ? "enable" : "disable", + __entry->err + ) +); + +TRACE_EVENT(rtc_alarm_irq_enable, + + TP_PROTO(unsigned int enabled, int err), + + TP_ARGS(enabled, err), + + TP_STRUCT__entry( + __field(unsigned int, enabled) + __field(int, err) + ), + + TP_fast_assign( + __entry->enabled = enabled; + __entry->err = err; + ), + + TP_printk("%s RTC alarm IRQ (%d)", + __entry->enabled ? "enable" : "disable", + __entry->err + ) +); + +DECLARE_EVENT_CLASS(rtc_offset_class, + + TP_PROTO(long offset, int err), + + TP_ARGS(offset, err), + + TP_STRUCT__entry( + __field(long, offset) + __field(int, err) + ), + + TP_fast_assign( + __entry->offset = offset; + __entry->err = err; + ), + + TP_printk("RTC offset: %ld (%d)", + __entry->offset, __entry->err + ) +); + +DEFINE_EVENT(rtc_offset_class, rtc_set_offset, + + TP_PROTO(long offset, int err), + + TP_ARGS(offset, err) +); + +DEFINE_EVENT(rtc_offset_class, rtc_read_offset, + + TP_PROTO(long offset, int err), + + TP_ARGS(offset, err) +); + +DECLARE_EVENT_CLASS(rtc_timer_class, + + TP_PROTO(struct rtc_timer *timer), + + TP_ARGS(timer), + + TP_STRUCT__entry( + __field(struct rtc_timer *, timer) + __field(ktime_t, expires) + __field(ktime_t, period) + ), + + TP_fast_assign( + __entry->timer = timer; + __entry->expires = timer->node.expires; + __entry->period = timer->period; + ), + + TP_printk("RTC timer:(%p) expires:%lld period:%lld", + __entry->timer, __entry->expires, __entry->period + ) +); + +DEFINE_EVENT(rtc_timer_class, rtc_timer_enqueue, + + TP_PROTO(struct rtc_timer *timer), + + TP_ARGS(timer) +); + +DEFINE_EVENT(rtc_timer_class, rtc_timer_dequeue, + + TP_PROTO(struct rtc_timer *timer), + + TP_ARGS(timer) +); + +DEFINE_EVENT(rtc_timer_class, rtc_timer_fired, + + TP_PROTO(struct rtc_timer *timer), + + TP_ARGS(timer) +); + +#endif /* _TRACE_RTC_H */ + +/* This part must be outside protection */ +#include -- cgit v1.2.3 From ccf0f32acd436b9e554303fd571f1bbf5f49d8e2 Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Sun, 18 Feb 2018 20:53:23 -0500 Subject: ext4: add tracepoints for shutdown and file system errors Signed-off-by: Theodore Ts'o --- fs/ext4/ioctl.c | 1 + fs/ext4/super.c | 4 ++++ include/trace/events/ext4.h | 43 +++++++++++++++++++++++++++++++++++++++++++ 3 files changed, 48 insertions(+) (limited to 'include/trace') diff --git a/fs/ext4/ioctl.c b/fs/ext4/ioctl.c index 7e99ad02f1ba..4d1b1575f8ac 100644 --- a/fs/ext4/ioctl.c +++ b/fs/ext4/ioctl.c @@ -481,6 +481,7 @@ static int ext4_shutdown(struct super_block *sb, unsigned long arg) return 0; ext4_msg(sb, KERN_ALERT, "shut down requested (%d)", flags); + trace_ext4_shutdown(sb, flags); switch (flags) { case EXT4_GOING_FLAGS_DEFAULT: diff --git a/fs/ext4/super.c b/fs/ext4/super.c index 39bf464c35f1..756f515b762d 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -448,6 +448,7 @@ void __ext4_error(struct super_block *sb, const char *function, if (unlikely(ext4_forced_shutdown(EXT4_SB(sb)))) return; + trace_ext4_error(sb, function, line); if (ext4_error_ratelimit(sb)) { va_start(args, fmt); vaf.fmt = fmt; @@ -472,6 +473,7 @@ void __ext4_error_inode(struct inode *inode, const char *function, if (unlikely(ext4_forced_shutdown(EXT4_SB(inode->i_sb)))) return; + trace_ext4_error(inode->i_sb, function, line); es->s_last_error_ino = cpu_to_le32(inode->i_ino); es->s_last_error_block = cpu_to_le64(block); if (ext4_error_ratelimit(inode->i_sb)) { @@ -507,6 +509,7 @@ void __ext4_error_file(struct file *file, const char *function, if (unlikely(ext4_forced_shutdown(EXT4_SB(inode->i_sb)))) return; + trace_ext4_error(inode->i_sb, function, line); es = EXT4_SB(inode->i_sb)->s_es; es->s_last_error_ino = cpu_to_le32(inode->i_ino); if (ext4_error_ratelimit(inode->i_sb)) { @@ -719,6 +722,7 @@ __acquires(bitlock) if (unlikely(ext4_forced_shutdown(EXT4_SB(sb)))) return; + trace_ext4_error(sb, function, line); es->s_last_error_ino = cpu_to_le32(ino); es->s_last_error_block = cpu_to_le64(block); __save_error_info(sb, function, line); diff --git a/include/trace/events/ext4.h b/include/trace/events/ext4.h index 4d0e3af4e561..0e31eb136c57 100644 --- a/include/trace/events/ext4.h +++ b/include/trace/events/ext4.h @@ -2585,6 +2585,49 @@ DEFINE_GETFSMAP_EVENT(ext4_getfsmap_low_key); DEFINE_GETFSMAP_EVENT(ext4_getfsmap_high_key); DEFINE_GETFSMAP_EVENT(ext4_getfsmap_mapping); +TRACE_EVENT(ext4_shutdown, + TP_PROTO(struct super_block *sb, unsigned long flags), + + TP_ARGS(sb, flags), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( unsigned, flags ) + ), + + TP_fast_assign( + __entry->dev = sb->s_dev; + __entry->flags = flags; + ), + + TP_printk("dev %d,%d flags %u", + MAJOR(__entry->dev), MINOR(__entry->dev), + __entry->flags) +); + +TRACE_EVENT(ext4_error, + TP_PROTO(struct super_block *sb, const char *function, + unsigned int line), + + TP_ARGS(sb, function, line), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( const char *, function ) + __field( unsigned, line ) + ), + + TP_fast_assign( + __entry->dev = sb->s_dev; + __entry->function = function; + __entry->line = line; + ), + + TP_printk("dev %d,%d function %s line %u", + MAJOR(__entry->dev), MINOR(__entry->dev), + __entry->function, __entry->line) +); + #endif /* _TRACE_EXT4_H */ /* This part must be outside protection */ -- cgit v1.2.3 From 9a414201ae7ea089699a0cbd36533345ca17233b Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Wed, 31 Jan 2018 19:23:24 -0800 Subject: rcu: Add more tracing of expedited grace periods This commit adds more tracing of expedited grace periods to enable improved debugging of slowdowns. Signed-off-by: Paul E. McKenney --- include/trace/events/rcu.h | 3 +++ kernel/rcu/rcu.h | 8 +++++++- kernel/rcu/tree_exp.h | 12 ++++++++++++ 3 files changed, 22 insertions(+), 1 deletion(-) (limited to 'include/trace') diff --git a/include/trace/events/rcu.h b/include/trace/events/rcu.h index 0b50fda80db0..e56a618f2a59 100644 --- a/include/trace/events/rcu.h +++ b/include/trace/events/rcu.h @@ -179,6 +179,9 @@ TRACE_EVENT(rcu_grace_period_init, * * "snap": Captured snapshot of expedited grace period sequence number. * "start": Started a real expedited grace period. + * "reset": Started resetting the tree + * "select": Started selecting the CPUs to wait on. + * "startwait": Started waiting on selected CPUs. * "end": Ended a real expedited grace period. * "endwake": Woke piggybackers up. * "done": Someone else did the expedited grace period for us. diff --git a/kernel/rcu/rcu.h b/kernel/rcu/rcu.h index 5d13f651cf08..507a0802c717 100644 --- a/kernel/rcu/rcu.h +++ b/kernel/rcu/rcu.h @@ -77,12 +77,18 @@ static inline void rcu_seq_start(unsigned long *sp) WARN_ON_ONCE(rcu_seq_state(*sp) != 1); } +/* Compute the end-of-grace-period value for the specified sequence number. */ +static inline unsigned long rcu_seq_endval(unsigned long *sp) +{ + return (*sp | RCU_SEQ_STATE_MASK) + 1; +} + /* Adjust sequence number for end of update-side operation. */ static inline void rcu_seq_end(unsigned long *sp) { smp_mb(); /* Ensure update-side operation before counter increment. */ WARN_ON_ONCE(!rcu_seq_state(*sp)); - WRITE_ONCE(*sp, (*sp | RCU_SEQ_STATE_MASK) + 1); + WRITE_ONCE(*sp, rcu_seq_endval(sp)); } /* Take a snapshot of the update side's sequence number. */ diff --git a/kernel/rcu/tree_exp.h b/kernel/rcu/tree_exp.h index 46d61b597731..70ad12abde36 100644 --- a/kernel/rcu/tree_exp.h +++ b/kernel/rcu/tree_exp.h @@ -28,6 +28,15 @@ static void rcu_exp_gp_seq_start(struct rcu_state *rsp) rcu_seq_start(&rsp->expedited_sequence); } +/* + * Return then value that expedited-grace-period counter will have + * at the end of the current grace period. + */ +static unsigned long rcu_exp_gp_seq_endval(struct rcu_state *rsp) +{ + return rcu_seq_endval(&rsp->expedited_sequence); +} + /* * Record the end of an expedited grace period. */ @@ -366,7 +375,9 @@ static void sync_rcu_exp_select_cpus(struct rcu_state *rsp, int ret; struct rcu_node *rnp; + trace_rcu_exp_grace_period(rsp->name, rcu_exp_gp_seq_endval(rsp), TPS("reset")); sync_exp_reset_tree(rsp); + trace_rcu_exp_grace_period(rsp->name, rcu_exp_gp_seq_endval(rsp), TPS("select")); rcu_for_each_leaf_node(rsp, rnp) { raw_spin_lock_irqsave_rcu_node(rnp, flags); @@ -443,6 +454,7 @@ static void synchronize_sched_expedited_wait(struct rcu_state *rsp) struct rcu_node *rnp_root = rcu_get_root(rsp); int ret; + trace_rcu_exp_grace_period(rsp->name, rcu_exp_gp_seq_endval(rsp), TPS("startwait")); jiffies_stall = rcu_jiffies_till_stall_check(); jiffies_start = jiffies; -- cgit v1.2.3 From 7f5d42d05155523a4c42c2c5170f2a368217aed5 Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Wed, 31 Jan 2018 19:40:22 -0800 Subject: rcu: Trace expedited GP delays due to transitioning CPUs If a CPU is transitioning to or from offline state, an expedited grace period may undergo a timed wait. This timed wait can unduly delay grace periods, so this commit adds a trace statement to make it visible. Signed-off-by: Paul E. McKenney --- include/trace/events/rcu.h | 1 + kernel/rcu/tree_exp.h | 3 ++- 2 files changed, 3 insertions(+), 1 deletion(-) (limited to 'include/trace') diff --git a/include/trace/events/rcu.h b/include/trace/events/rcu.h index e56a618f2a59..d8c33298c153 100644 --- a/include/trace/events/rcu.h +++ b/include/trace/events/rcu.h @@ -181,6 +181,7 @@ TRACE_EVENT(rcu_grace_period_init, * "start": Started a real expedited grace period. * "reset": Started resetting the tree * "select": Started selecting the CPUs to wait on. + * "selectofl": Selected CPU partially offline. * "startwait": Started waiting on selected CPUs. * "end": Ended a real expedited grace period. * "endwake": Woke piggybackers up. diff --git a/kernel/rcu/tree_exp.h b/kernel/rcu/tree_exp.h index 70ad12abde36..fecb6b6ab452 100644 --- a/kernel/rcu/tree_exp.h +++ b/kernel/rcu/tree_exp.h @@ -32,7 +32,7 @@ static void rcu_exp_gp_seq_start(struct rcu_state *rsp) * Return then value that expedited-grace-period counter will have * at the end of the current grace period. */ -static unsigned long rcu_exp_gp_seq_endval(struct rcu_state *rsp) +static __maybe_unused unsigned long rcu_exp_gp_seq_endval(struct rcu_state *rsp) { return rcu_seq_endval(&rsp->expedited_sequence); } @@ -428,6 +428,7 @@ retry_ipi: (rnp->expmask & mask)) { /* Online, so delay for a bit and try again. */ raw_spin_unlock_irqrestore_rcu_node(rnp, flags); + trace_rcu_exp_grace_period(rsp->name, rcu_exp_gp_seq_endval(rsp), TPS("selectofl")); schedule_timeout_uninterruptible(1); goto retry_ipi; } -- cgit v1.2.3 From 5f171577b4f35b44795a73bde8cf2c49b4073925 Mon Sep 17 00:00:00 2001 From: James Hogan Date: Tue, 24 Oct 2017 16:52:32 +0100 Subject: Drop a bunch of metag references Now that arch/metag/ has been removed, drop a bunch of metag references in various codes across the whole tree: - VM_GROWSUP and __VM_ARCH_SPECIFIC_1. - MT_METAG_* ELF note types. - METAG Kconfig dependencies (FRAME_POINTER) and ranges (MAX_STACK_SIZE_MB). - metag cases in tools (checkstack.pl, recordmcount.c, perf). Signed-off-by: James Hogan Acked-by: Steven Rostedt (VMware) Acked-by: Peter Zijlstra (Intel) Reviewed-by: Guenter Roeck Cc: Ingo Molnar Cc: Arnaldo Carvalho de Melo Cc: Alexander Shishkin Cc: Jiri Olsa Cc: Namhyung Kim Cc: linux-mm@kvack.org Cc: linux-metag@vger.kernel.org --- include/linux/cpuhotplug.h | 1 - include/linux/mm.h | 2 -- include/trace/events/mmflags.h | 2 +- include/uapi/linux/elf.h | 3 --- lib/Kconfig.debug | 2 +- mm/Kconfig | 7 +++---- scripts/checkstack.pl | 4 ---- scripts/recordmcount.c | 20 -------------------- tools/perf/perf-sys.h | 4 ---- 9 files changed, 5 insertions(+), 40 deletions(-) (limited to 'include/trace') diff --git a/include/linux/cpuhotplug.h b/include/linux/cpuhotplug.h index 5172ad0daa7c..c7a950681f3a 100644 --- a/include/linux/cpuhotplug.h +++ b/include/linux/cpuhotplug.h @@ -108,7 +108,6 @@ enum cpuhp_state { CPUHP_AP_PERF_X86_CQM_STARTING, CPUHP_AP_PERF_X86_CSTATE_STARTING, CPUHP_AP_PERF_XTENSA_STARTING, - CPUHP_AP_PERF_METAG_STARTING, CPUHP_AP_MIPS_OP_LOONGSON3_STARTING, CPUHP_AP_ARM_SDEI_STARTING, CPUHP_AP_ARM_VFP_STARTING, diff --git a/include/linux/mm.h b/include/linux/mm.h index ad06d42adb1a..ccac10682ce5 100644 --- a/include/linux/mm.h +++ b/include/linux/mm.h @@ -241,8 +241,6 @@ extern unsigned int kobjsize(const void *objp); # define VM_SAO VM_ARCH_1 /* Strong Access Ordering (powerpc) */ #elif defined(CONFIG_PARISC) # define VM_GROWSUP VM_ARCH_1 -#elif defined(CONFIG_METAG) -# define VM_GROWSUP VM_ARCH_1 #elif defined(CONFIG_IA64) # define VM_GROWSUP VM_ARCH_1 #elif !defined(CONFIG_MMU) diff --git a/include/trace/events/mmflags.h b/include/trace/events/mmflags.h index dbe1bb058c09..a81cffb76d89 100644 --- a/include/trace/events/mmflags.h +++ b/include/trace/events/mmflags.h @@ -115,7 +115,7 @@ IF_HAVE_PG_IDLE(PG_idle, "idle" ) #define __VM_ARCH_SPECIFIC_1 {VM_PAT, "pat" } #elif defined(CONFIG_PPC) #define __VM_ARCH_SPECIFIC_1 {VM_SAO, "sao" } -#elif defined(CONFIG_PARISC) || defined(CONFIG_METAG) || defined(CONFIG_IA64) +#elif defined(CONFIG_PARISC) || defined(CONFIG_IA64) #define __VM_ARCH_SPECIFIC_1 {VM_GROWSUP, "growsup" } #elif !defined(CONFIG_MMU) #define __VM_ARCH_SPECIFIC_1 {VM_MAPPED_COPY,"mappedcopy" } diff --git a/include/uapi/linux/elf.h b/include/uapi/linux/elf.h index 3bf73fb58045..e2535d6dcec7 100644 --- a/include/uapi/linux/elf.h +++ b/include/uapi/linux/elf.h @@ -420,9 +420,6 @@ typedef struct elf64_shdr { #define NT_ARM_HW_WATCH 0x403 /* ARM hardware watchpoint registers */ #define NT_ARM_SYSTEM_CALL 0x404 /* ARM system call number */ #define NT_ARM_SVE 0x405 /* ARM Scalable Vector Extension registers */ -#define NT_METAG_CBUF 0x500 /* Metag catch buffer registers */ -#define NT_METAG_RPIPE 0x501 /* Metag read pipeline state */ -#define NT_METAG_TLS 0x502 /* Metag TLS pointer */ #define NT_ARC_V2 0x600 /* ARCv2 accumulator/extra registers */ /* Note header in a PT_NOTE section */ diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug index 6088408ef26c..d1c523e408e9 100644 --- a/lib/Kconfig.debug +++ b/lib/Kconfig.debug @@ -356,7 +356,7 @@ config FRAME_POINTER bool "Compile the kernel with frame pointers" depends on DEBUG_KERNEL && \ (CRIS || M68K || FRV || UML || \ - SUPERH || BLACKFIN || MN10300 || METAG) || \ + SUPERH || BLACKFIN || MN10300) || \ ARCH_WANT_FRAME_POINTERS default y if (DEBUG_INFO && UML) || ARCH_WANT_FRAME_POINTERS help diff --git a/mm/Kconfig b/mm/Kconfig index c782e8fb7235..abefa573bcd8 100644 --- a/mm/Kconfig +++ b/mm/Kconfig @@ -627,15 +627,14 @@ config GENERIC_EARLY_IOREMAP config MAX_STACK_SIZE_MB int "Maximum user stack size for 32-bit processes (MB)" default 80 - range 8 256 if METAG range 8 2048 depends on STACK_GROWSUP && (!64BIT || COMPAT) help This is the maximum stack size in Megabytes in the VM layout of 32-bit user processes when the stack grows upwards (currently only on parisc - and metag arch). The stack will be located at the highest memory - address minus the given value, unless the RLIMIT_STACK hard limit is - changed to a smaller value in which case that is used. + arch). The stack will be located at the highest memory address minus + the given value, unless the RLIMIT_STACK hard limit is changed to a + smaller value in which case that is used. A sane initial value is 80 MB. diff --git a/scripts/checkstack.pl b/scripts/checkstack.pl index cb993801e4b2..eeb9ac8dbcfb 100755 --- a/scripts/checkstack.pl +++ b/scripts/checkstack.pl @@ -64,10 +64,6 @@ my (@stack, $re, $dre, $x, $xs, $funcre); # 2b6c: 4e56 fb70 linkw %fp,#-1168 # 1df770: defc ffe4 addaw #-28,%sp $re = qr/.*(?:linkw %fp,|addaw )#-([0-9]{1,4})(?:,%sp)?$/o; - } elsif ($arch eq 'metag') { - #400026fc: 40 00 00 82 ADD A0StP,A0StP,#0x8 - $re = qr/.*ADD.*A0StP,A0StP,\#(0x$x{1,8})/o; - $funcre = qr/^$x* <[^\$](.*)>:$/; } elsif ($arch eq 'mips64') { #8800402c: 67bdfff0 daddiu sp,sp,-16 $re = qr/.*daddiu.*sp,sp,-(([0-9]{2}|[3-9])[0-9]{2})/o; diff --git a/scripts/recordmcount.c b/scripts/recordmcount.c index 16e086dcc567..8c9691c3329e 100644 --- a/scripts/recordmcount.c +++ b/scripts/recordmcount.c @@ -33,20 +33,6 @@ #include #include -/* - * glibc synced up and added the metag number but didn't add the relocations. - * Work around this in a crude manner for now. - */ -#ifndef EM_METAG -#define EM_METAG 174 -#endif -#ifndef R_METAG_ADDR32 -#define R_METAG_ADDR32 2 -#endif -#ifndef R_METAG_NONE -#define R_METAG_NONE 3 -#endif - #ifndef EM_AARCH64 #define EM_AARCH64 183 #define R_AARCH64_NONE 0 @@ -538,12 +524,6 @@ do_file(char const *const fname) gpfx = '_'; break; case EM_IA_64: reltype = R_IA64_IMM64; gpfx = '_'; break; - case EM_METAG: reltype = R_METAG_ADDR32; - altmcount = "_mcount_wrapper"; - rel_type_nop = R_METAG_NONE; - /* We happen to have the same requirement as MIPS */ - is_fake_mcount32 = MIPS32_is_fake_mcount; - break; case EM_MIPS: /* reltype: e_class */ gpfx = '_'; break; case EM_PPC: reltype = R_PPC_ADDR32; gpfx = '_'; break; case EM_PPC64: reltype = R_PPC64_ADDR64; gpfx = '_'; break; diff --git a/tools/perf/perf-sys.h b/tools/perf/perf-sys.h index 36673f98d66b..3eb7a39169f6 100644 --- a/tools/perf/perf-sys.h +++ b/tools/perf/perf-sys.h @@ -46,10 +46,6 @@ #define CPUINFO_PROC {"Processor"} #endif -#ifdef __metag__ -#define CPUINFO_PROC {"CPU"} -#endif - #ifdef __xtensa__ #define CPUINFO_PROC {"core ID"} #endif -- cgit v1.2.3 From 393da91819e35af538ef97c7c6a04899e2fbfe0e Mon Sep 17 00:00:00 2001 From: Liu Bo Date: Fri, 5 Jan 2018 12:51:16 -0700 Subject: Btrfs: add tracepoint for em's EEXIST case This is adding a tracepoint 'btrfs_handle_em_exist' to help debug the subtle bugs around merge_extent_mapping. Signed-off-by: Liu Bo Reviewed-by: Josef Bacik Signed-off-by: David Sterba --- fs/btrfs/extent_map.c | 3 +++ include/trace/events/btrfs.h | 35 +++++++++++++++++++++++++++++++++++ 2 files changed, 38 insertions(+) (limited to 'include/trace') diff --git a/fs/btrfs/extent_map.c b/fs/btrfs/extent_map.c index c80dea7c69af..b8ead8dc2ebe 100644 --- a/fs/btrfs/extent_map.c +++ b/fs/btrfs/extent_map.c @@ -551,6 +551,9 @@ int btrfs_add_extent_mapping(struct extent_map_tree *em_tree, ret = 0; existing = search_extent_mapping(em_tree, start, len); + + trace_btrfs_handle_em_exist(existing, em, start, len); + /* * existing will always be non-NULL, since there must be * extent causing the -EEXIST. diff --git a/include/trace/events/btrfs.h b/include/trace/events/btrfs.h index c3ac5ec86519..486771e3f4cb 100644 --- a/include/trace/events/btrfs.h +++ b/include/trace/events/btrfs.h @@ -248,6 +248,41 @@ TRACE_EVENT_CONDITION(btrfs_get_extent, __entry->refs, __entry->compress_type) ); +TRACE_EVENT(btrfs_handle_em_exist, + + TP_PROTO(const struct extent_map *existing, const struct extent_map *map, u64 start, u64 len), + + TP_ARGS(existing, map, start, len), + + TP_STRUCT__entry( + __field( u64, e_start ) + __field( u64, e_len ) + __field( u64, map_start ) + __field( u64, map_len ) + __field( u64, start ) + __field( u64, len ) + ), + + TP_fast_assign( + __entry->e_start = existing->start; + __entry->e_len = existing->len; + __entry->map_start = map->start; + __entry->map_len = map->len; + __entry->start = start; + __entry->len = len; + ), + + TP_printk("start=%llu len=%llu " + "existing(start=%llu len=%llu) " + "em(start=%llu len=%llu)", + (unsigned long long)__entry->start, + (unsigned long long)__entry->len, + (unsigned long long)__entry->e_start, + (unsigned long long)__entry->e_len, + (unsigned long long)__entry->map_start, + (unsigned long long)__entry->map_len) +); + /* file extent item */ DECLARE_EVENT_CLASS(btrfs__file_extent_item_regular, -- cgit v1.2.3 From 827efed6a66ef8a1c071400b5952fee4a5ffedf9 Mon Sep 17 00:00:00 2001 From: David Howells Date: Tue, 27 Mar 2018 23:02:47 +0100 Subject: rxrpc: Trace resend Add a tracepoint to trace packet resend events and to dump the Tx annotation buffer for added illumination. Signed-off-by: David Howells --- include/trace/events/rxrpc.h | 24 ++++++++++++++++++++++++ net/rxrpc/call_event.c | 1 + 2 files changed, 25 insertions(+) (limited to 'include/trace') diff --git a/include/trace/events/rxrpc.h b/include/trace/events/rxrpc.h index 36cb50c111a6..41979f907575 100644 --- a/include/trace/events/rxrpc.h +++ b/include/trace/events/rxrpc.h @@ -420,6 +420,7 @@ rxrpc_rtt_rx_traces; rxrpc_timer_traces; rxrpc_propose_ack_traces; rxrpc_propose_ack_outcomes; +rxrpc_congest_modes; rxrpc_congest_changes; /* @@ -1229,6 +1230,29 @@ TRACE_EVENT(rxrpc_connect_call, __entry->call_id) ); +TRACE_EVENT(rxrpc_resend, + TP_PROTO(struct rxrpc_call *call, int ix), + + TP_ARGS(call, ix), + + TP_STRUCT__entry( + __field(struct rxrpc_call *, call ) + __field(int, ix ) + __array(u8, anno, 64 ) + ), + + TP_fast_assign( + __entry->call = call; + __entry->ix = ix; + memcpy(__entry->anno, call->rxtx_annotations, 64); + ), + + TP_printk("c=%p ix=%u a=%64phN", + __entry->call, + __entry->ix, + __entry->anno) + ); + #endif /* _TRACE_RXRPC_H */ /* This part must be outside protection */ diff --git a/net/rxrpc/call_event.c b/net/rxrpc/call_event.c index ad2ab1103189..6a62e42e1d8d 100644 --- a/net/rxrpc/call_event.c +++ b/net/rxrpc/call_event.c @@ -195,6 +195,7 @@ static void rxrpc_resend(struct rxrpc_call *call, unsigned long now_j) * the packets in the Tx buffer we're going to resend and what the new * resend timeout will be. */ + trace_rxrpc_resend(call, (cursor + 1) & RXRPC_RXTX_BUFF_MASK); oldest = now; for (seq = cursor + 1; before_eq(seq, top); seq++) { ix = seq & RXRPC_RXTX_BUFF_MASK; -- cgit v1.2.3 From a25e21f0bcd25673b91b97b9805db33350feec0f Mon Sep 17 00:00:00 2001 From: David Howells Date: Tue, 27 Mar 2018 23:03:00 +0100 Subject: rxrpc, afs: Use debug_ids rather than pointers in traces In rxrpc and afs, use the debug_ids that are monotonically allocated to various objects as they're allocated rather than pointers as kernel pointers are now hashed making them less useful. Further, the debug ids aren't reused anywhere nearly as quickly. In addition, allow kernel services that use rxrpc, such as afs, to take numbers from the rxrpc counter, assign them to their own call struct and pass them in to rxrpc for both client and service calls so that the trace lines for each will have the same ID tag. Signed-off-by: David Howells --- fs/afs/internal.h | 1 + fs/afs/rxrpc.c | 12 ++-- include/net/af_rxrpc.h | 11 ++- include/trace/events/afs.h | 69 +++++++++---------- include/trace/events/rxrpc.h | 155 ++++++++++++++++++++++--------------------- net/rxrpc/af_rxrpc.c | 7 +- net/rxrpc/ar-internal.h | 8 +-- net/rxrpc/call_accept.c | 18 +++-- net/rxrpc/call_object.c | 15 +++-- net/rxrpc/conn_event.c | 3 +- net/rxrpc/input.c | 6 +- net/rxrpc/sendmsg.c | 3 +- 12 files changed, 163 insertions(+), 145 deletions(-) (limited to 'include/trace') diff --git a/fs/afs/internal.h b/fs/afs/internal.h index f38d6a561a84..72217170b155 100644 --- a/fs/afs/internal.h +++ b/fs/afs/internal.h @@ -118,6 +118,7 @@ struct afs_call { bool ret_reply0; /* T if should return reply[0] on success */ bool upgrade; /* T to request service upgrade */ u16 service_id; /* Actual service ID (after upgrade) */ + unsigned int debug_id; /* Trace ID */ u32 operation_ID; /* operation ID for an incoming call */ u32 count; /* count for use in unmarshalling */ __be32 tmp; /* place to extract temporary data */ diff --git a/fs/afs/rxrpc.c b/fs/afs/rxrpc.c index e1126659f043..b819900916e6 100644 --- a/fs/afs/rxrpc.c +++ b/fs/afs/rxrpc.c @@ -131,6 +131,7 @@ static struct afs_call *afs_alloc_call(struct afs_net *net, call->type = type; call->net = net; + call->debug_id = atomic_inc_return(&rxrpc_debug_id); atomic_set(&call->usage, 1); INIT_WORK(&call->async_work, afs_process_async_call); init_waitqueue_head(&call->waitq); @@ -169,11 +170,12 @@ void afs_put_call(struct afs_call *call) afs_put_server(call->net, call->cm_server); afs_put_cb_interest(call->net, call->cbi); kfree(call->request); - kfree(call); - o = atomic_dec_return(&net->nr_outstanding_calls); trace_afs_call(call, afs_call_trace_free, 0, o, __builtin_return_address(0)); + kfree(call); + + o = atomic_dec_return(&net->nr_outstanding_calls); if (o == 0) wake_up_atomic_t(&net->nr_outstanding_calls); } @@ -378,7 +380,8 @@ long afs_make_call(struct afs_addr_cursor *ac, struct afs_call *call, (async ? afs_wake_up_async_call : afs_wake_up_call_waiter), - call->upgrade); + call->upgrade, + call->debug_id); if (IS_ERR(rxcall)) { ret = PTR_ERR(rxcall); goto error_kill_call; @@ -727,7 +730,8 @@ void afs_charge_preallocation(struct work_struct *work) afs_wake_up_async_call, afs_rx_attach, (unsigned long)call, - GFP_KERNEL) < 0) + GFP_KERNEL, + call->debug_id) < 0) break; call = NULL; } diff --git a/include/net/af_rxrpc.h b/include/net/af_rxrpc.h index 2b3a6eec4570..8ae8ee004258 100644 --- a/include/net/af_rxrpc.h +++ b/include/net/af_rxrpc.h @@ -31,6 +31,11 @@ enum rxrpc_call_completion { NR__RXRPC_CALL_COMPLETIONS }; +/* + * Debug ID counter for tracing. + */ +extern atomic_t rxrpc_debug_id; + typedef void (*rxrpc_notify_rx_t)(struct sock *, struct rxrpc_call *, unsigned long); typedef void (*rxrpc_notify_end_tx_t)(struct sock *, struct rxrpc_call *, @@ -50,7 +55,8 @@ struct rxrpc_call *rxrpc_kernel_begin_call(struct socket *, s64, gfp_t, rxrpc_notify_rx_t, - bool); + bool, + unsigned int); int rxrpc_kernel_send_data(struct socket *, struct rxrpc_call *, struct msghdr *, size_t, rxrpc_notify_end_tx_t); @@ -63,7 +69,8 @@ void rxrpc_kernel_get_peer(struct socket *, struct rxrpc_call *, struct sockaddr_rxrpc *); u64 rxrpc_kernel_get_rtt(struct socket *, struct rxrpc_call *); int rxrpc_kernel_charge_accept(struct socket *, rxrpc_notify_rx_t, - rxrpc_user_attach_call_t, unsigned long, gfp_t); + rxrpc_user_attach_call_t, unsigned long, gfp_t, + unsigned int); void rxrpc_kernel_set_tx_length(struct socket *, struct rxrpc_call *, s64); int rxrpc_kernel_retry_call(struct socket *, struct rxrpc_call *, struct sockaddr_rxrpc *, struct key *); diff --git a/include/trace/events/afs.h b/include/trace/events/afs.h index 6b59c63a8e51..63815f66b274 100644 --- a/include/trace/events/afs.h +++ b/include/trace/events/afs.h @@ -133,8 +133,7 @@ TRACE_EVENT(afs_recv_data, TP_ARGS(call, count, offset, want_more, ret), TP_STRUCT__entry( - __field(struct rxrpc_call *, rxcall ) - __field(struct afs_call *, call ) + __field(unsigned int, call ) __field(enum afs_call_state, state ) __field(unsigned int, count ) __field(unsigned int, offset ) @@ -144,8 +143,7 @@ TRACE_EVENT(afs_recv_data, ), TP_fast_assign( - __entry->rxcall = call->rxcall; - __entry->call = call; + __entry->call = call->debug_id; __entry->state = call->state; __entry->unmarshall = call->unmarshall; __entry->count = count; @@ -154,8 +152,7 @@ TRACE_EVENT(afs_recv_data, __entry->ret = ret; ), - TP_printk("c=%p ac=%p s=%u u=%u %u/%u wm=%u ret=%d", - __entry->rxcall, + TP_printk("c=%08x s=%u u=%u %u/%u wm=%u ret=%d", __entry->call, __entry->state, __entry->unmarshall, __entry->offset, __entry->count, @@ -168,21 +165,18 @@ TRACE_EVENT(afs_notify_call, TP_ARGS(rxcall, call), TP_STRUCT__entry( - __field(struct rxrpc_call *, rxcall ) - __field(struct afs_call *, call ) + __field(unsigned int, call ) __field(enum afs_call_state, state ) __field(unsigned short, unmarshall ) ), TP_fast_assign( - __entry->rxcall = rxcall; - __entry->call = call; + __entry->call = call->debug_id; __entry->state = call->state; __entry->unmarshall = call->unmarshall; ), - TP_printk("c=%p ac=%p s=%u u=%u", - __entry->rxcall, + TP_printk("c=%08x s=%u u=%u", __entry->call, __entry->state, __entry->unmarshall) ); @@ -193,21 +187,18 @@ TRACE_EVENT(afs_cb_call, TP_ARGS(call), TP_STRUCT__entry( - __field(struct rxrpc_call *, rxcall ) - __field(struct afs_call *, call ) + __field(unsigned int, call ) __field(const char *, name ) __field(u32, op ) ), TP_fast_assign( - __entry->rxcall = call->rxcall; - __entry->call = call; + __entry->call = call->debug_id; __entry->name = call->type->name; __entry->op = call->operation_ID; ), - TP_printk("c=%p ac=%p %s o=%u", - __entry->rxcall, + TP_printk("c=%08x %s o=%u", __entry->call, __entry->name, __entry->op) @@ -220,7 +211,7 @@ TRACE_EVENT(afs_call, TP_ARGS(call, op, usage, outstanding, where), TP_STRUCT__entry( - __field(struct afs_call *, call ) + __field(unsigned int, call ) __field(int, op ) __field(int, usage ) __field(int, outstanding ) @@ -228,14 +219,14 @@ TRACE_EVENT(afs_call, ), TP_fast_assign( - __entry->call = call; + __entry->call = call->debug_id; __entry->op = op; __entry->usage = usage; __entry->outstanding = outstanding; __entry->where = where; ), - TP_printk("c=%p %s u=%d o=%d sp=%pSR", + TP_printk("c=%08x %s u=%d o=%d sp=%pSR", __entry->call, __print_symbolic(__entry->op, afs_call_traces), __entry->usage, @@ -249,13 +240,13 @@ TRACE_EVENT(afs_make_fs_call, TP_ARGS(call, fid), TP_STRUCT__entry( - __field(struct afs_call *, call ) + __field(unsigned int, call ) __field(enum afs_fs_operation, op ) __field_struct(struct afs_fid, fid ) ), TP_fast_assign( - __entry->call = call; + __entry->call = call->debug_id; __entry->op = call->operation_ID; if (fid) { __entry->fid = *fid; @@ -266,7 +257,7 @@ TRACE_EVENT(afs_make_fs_call, } ), - TP_printk("c=%p %06x:%06x:%06x %s", + TP_printk("c=%08x %06x:%06x:%06x %s", __entry->call, __entry->fid.vid, __entry->fid.vnode, @@ -280,16 +271,16 @@ TRACE_EVENT(afs_make_vl_call, TP_ARGS(call), TP_STRUCT__entry( - __field(struct afs_call *, call ) + __field(unsigned int, call ) __field(enum afs_vl_operation, op ) ), TP_fast_assign( - __entry->call = call; + __entry->call = call->debug_id; __entry->op = call->operation_ID; ), - TP_printk("c=%p %s", + TP_printk("c=%08x %s", __entry->call, __print_symbolic(__entry->op, afs_vl_operations)) ); @@ -300,20 +291,20 @@ TRACE_EVENT(afs_call_done, TP_ARGS(call), TP_STRUCT__entry( - __field(struct afs_call *, call ) + __field(unsigned int, call ) __field(struct rxrpc_call *, rx_call ) __field(int, ret ) __field(u32, abort_code ) ), TP_fast_assign( - __entry->call = call; + __entry->call = call->debug_id; __entry->rx_call = call->rxcall; __entry->ret = call->error; __entry->abort_code = call->abort_code; ), - TP_printk(" c=%p ret=%d ab=%d [%p]", + TP_printk(" c=%08x ret=%d ab=%d [%p]", __entry->call, __entry->ret, __entry->abort_code, @@ -327,7 +318,7 @@ TRACE_EVENT(afs_send_pages, TP_ARGS(call, msg, first, last, offset), TP_STRUCT__entry( - __field(struct afs_call *, call ) + __field(unsigned int, call ) __field(pgoff_t, first ) __field(pgoff_t, last ) __field(unsigned int, nr ) @@ -337,7 +328,7 @@ TRACE_EVENT(afs_send_pages, ), TP_fast_assign( - __entry->call = call; + __entry->call = call->debug_id; __entry->first = first; __entry->last = last; __entry->nr = msg->msg_iter.nr_segs; @@ -346,7 +337,7 @@ TRACE_EVENT(afs_send_pages, __entry->flags = msg->msg_flags; ), - TP_printk(" c=%p %lx-%lx-%lx b=%x o=%x f=%x", + TP_printk(" c=%08x %lx-%lx-%lx b=%x o=%x f=%x", __entry->call, __entry->first, __entry->first + __entry->nr - 1, __entry->last, __entry->bytes, __entry->offset, @@ -360,7 +351,7 @@ TRACE_EVENT(afs_sent_pages, TP_ARGS(call, first, last, cursor, ret), TP_STRUCT__entry( - __field(struct afs_call *, call ) + __field(unsigned int, call ) __field(pgoff_t, first ) __field(pgoff_t, last ) __field(pgoff_t, cursor ) @@ -368,14 +359,14 @@ TRACE_EVENT(afs_sent_pages, ), TP_fast_assign( - __entry->call = call; + __entry->call = call->debug_id; __entry->first = first; __entry->last = last; __entry->cursor = cursor; __entry->ret = ret; ), - TP_printk(" c=%p %lx-%lx c=%lx r=%d", + TP_printk(" c=%08x %lx-%lx c=%lx r=%d", __entry->call, __entry->first, __entry->last, __entry->cursor, __entry->ret) @@ -450,7 +441,7 @@ TRACE_EVENT(afs_call_state, TP_ARGS(call, from, to, ret, remote_abort), TP_STRUCT__entry( - __field(struct afs_call *, call ) + __field(unsigned int, call ) __field(enum afs_call_state, from ) __field(enum afs_call_state, to ) __field(int, ret ) @@ -458,14 +449,14 @@ TRACE_EVENT(afs_call_state, ), TP_fast_assign( - __entry->call = call; + __entry->call = call->debug_id; __entry->from = from; __entry->to = to; __entry->ret = ret; __entry->abort = remote_abort; ), - TP_printk("c=%p %u->%u r=%d ab=%d", + TP_printk("c=%08x %u->%u r=%d ab=%d", __entry->call, __entry->from, __entry->to, __entry->ret, __entry->abort) diff --git a/include/trace/events/rxrpc.h b/include/trace/events/rxrpc.h index 41979f907575..4d2c2d35c5cb 100644 --- a/include/trace/events/rxrpc.h +++ b/include/trace/events/rxrpc.h @@ -439,20 +439,20 @@ TRACE_EVENT(rxrpc_conn, TP_ARGS(conn, op, usage, where), TP_STRUCT__entry( - __field(struct rxrpc_connection *, conn ) - __field(int, op ) - __field(int, usage ) - __field(const void *, where ) + __field(unsigned int, conn ) + __field(int, op ) + __field(int, usage ) + __field(const void *, where ) ), TP_fast_assign( - __entry->conn = conn; + __entry->conn = conn->debug_id; __entry->op = op; __entry->usage = usage; __entry->where = where; ), - TP_printk("C=%p %s u=%d sp=%pSR", + TP_printk("C=%08x %s u=%d sp=%pSR", __entry->conn, __print_symbolic(__entry->op, rxrpc_conn_traces), __entry->usage, @@ -466,7 +466,7 @@ TRACE_EVENT(rxrpc_client, TP_ARGS(conn, channel, op), TP_STRUCT__entry( - __field(struct rxrpc_connection *, conn ) + __field(unsigned int, conn ) __field(u32, cid ) __field(int, channel ) __field(int, usage ) @@ -475,7 +475,7 @@ TRACE_EVENT(rxrpc_client, ), TP_fast_assign( - __entry->conn = conn; + __entry->conn = conn->debug_id; __entry->channel = channel; __entry->usage = atomic_read(&conn->usage); __entry->op = op; @@ -483,7 +483,7 @@ TRACE_EVENT(rxrpc_client, __entry->cs = conn->cache_state; ), - TP_printk("C=%p h=%2d %s %s i=%08x u=%d", + TP_printk("C=%08x h=%2d %s %s i=%08x u=%d", __entry->conn, __entry->channel, __print_symbolic(__entry->op, rxrpc_client_traces), @@ -499,7 +499,7 @@ TRACE_EVENT(rxrpc_call, TP_ARGS(call, op, usage, where, aux), TP_STRUCT__entry( - __field(struct rxrpc_call *, call ) + __field(unsigned int, call ) __field(int, op ) __field(int, usage ) __field(const void *, where ) @@ -507,14 +507,14 @@ TRACE_EVENT(rxrpc_call, ), TP_fast_assign( - __entry->call = call; + __entry->call = call->debug_id; __entry->op = op; __entry->usage = usage; __entry->where = where; __entry->aux = aux; ), - TP_printk("c=%p %s u=%d sp=%pSR a=%p", + TP_printk("c=%08x %s u=%d sp=%pSR a=%p", __entry->call, __print_symbolic(__entry->op, rxrpc_call_traces), __entry->usage, @@ -593,12 +593,13 @@ TRACE_EVENT(rxrpc_rx_done, ); TRACE_EVENT(rxrpc_abort, - TP_PROTO(const char *why, u32 cid, u32 call_id, rxrpc_seq_t seq, - int abort_code, int error), + TP_PROTO(unsigned int call_nr, const char *why, u32 cid, u32 call_id, + rxrpc_seq_t seq, int abort_code, int error), - TP_ARGS(why, cid, call_id, seq, abort_code, error), + TP_ARGS(call_nr, why, cid, call_id, seq, abort_code, error), TP_STRUCT__entry( + __field(unsigned int, call_nr ) __array(char, why, 4 ) __field(u32, cid ) __field(u32, call_id ) @@ -609,6 +610,7 @@ TRACE_EVENT(rxrpc_abort, TP_fast_assign( memcpy(__entry->why, why, 4); + __entry->call_nr = call_nr; __entry->cid = cid; __entry->call_id = call_id; __entry->abort_code = abort_code; @@ -616,7 +618,8 @@ TRACE_EVENT(rxrpc_abort, __entry->seq = seq; ), - TP_printk("%08x:%08x s=%u a=%d e=%d %s", + TP_printk("c=%08x %08x:%08x s=%u a=%d e=%d %s", + __entry->call_nr, __entry->cid, __entry->call_id, __entry->seq, __entry->abort_code, __entry->error, __entry->why) ); @@ -627,7 +630,7 @@ TRACE_EVENT(rxrpc_transmit, TP_ARGS(call, why), TP_STRUCT__entry( - __field(struct rxrpc_call *, call ) + __field(unsigned int, call ) __field(enum rxrpc_transmit_trace, why ) __field(rxrpc_seq_t, tx_hard_ack ) __field(rxrpc_seq_t, tx_top ) @@ -635,14 +638,14 @@ TRACE_EVENT(rxrpc_transmit, ), TP_fast_assign( - __entry->call = call; + __entry->call = call->debug_id; __entry->why = why; __entry->tx_hard_ack = call->tx_hard_ack; __entry->tx_top = call->tx_top; __entry->tx_winsize = call->tx_winsize; ), - TP_printk("c=%p %s f=%08x n=%u/%u", + TP_printk("c=%08x %s f=%08x n=%u/%u", __entry->call, __print_symbolic(__entry->why, rxrpc_transmit_traces), __entry->tx_hard_ack + 1, @@ -657,7 +660,7 @@ TRACE_EVENT(rxrpc_rx_data, TP_ARGS(call, seq, serial, flags, anno), TP_STRUCT__entry( - __field(struct rxrpc_call *, call ) + __field(unsigned int, call ) __field(rxrpc_seq_t, seq ) __field(rxrpc_serial_t, serial ) __field(u8, flags ) @@ -665,14 +668,14 @@ TRACE_EVENT(rxrpc_rx_data, ), TP_fast_assign( - __entry->call = call; + __entry->call = call->debug_id; __entry->seq = seq; __entry->serial = serial; __entry->flags = flags; __entry->anno = anno; ), - TP_printk("c=%p DATA %08x q=%08x fl=%02x a=%02x", + TP_printk("c=%08x DATA %08x q=%08x fl=%02x a=%02x", __entry->call, __entry->serial, __entry->seq, @@ -688,7 +691,7 @@ TRACE_EVENT(rxrpc_rx_ack, TP_ARGS(call, serial, ack_serial, first, prev, reason, n_acks), TP_STRUCT__entry( - __field(struct rxrpc_call *, call ) + __field(unsigned int, call ) __field(rxrpc_serial_t, serial ) __field(rxrpc_serial_t, ack_serial ) __field(rxrpc_seq_t, first ) @@ -698,7 +701,7 @@ TRACE_EVENT(rxrpc_rx_ack, ), TP_fast_assign( - __entry->call = call; + __entry->call = call->debug_id; __entry->serial = serial; __entry->ack_serial = ack_serial; __entry->first = first; @@ -707,7 +710,7 @@ TRACE_EVENT(rxrpc_rx_ack, __entry->n_acks = n_acks; ), - TP_printk("c=%p %08x %s r=%08x f=%08x p=%08x n=%u", + TP_printk("c=%08x %08x %s r=%08x f=%08x p=%08x n=%u", __entry->call, __entry->serial, __print_symbolic(__entry->reason, rxrpc_ack_names), @@ -724,18 +727,18 @@ TRACE_EVENT(rxrpc_rx_abort, TP_ARGS(call, serial, abort_code), TP_STRUCT__entry( - __field(struct rxrpc_call *, call ) + __field(unsigned int, call ) __field(rxrpc_serial_t, serial ) __field(u32, abort_code ) ), TP_fast_assign( - __entry->call = call; + __entry->call = call->debug_id; __entry->serial = serial; __entry->abort_code = abort_code; ), - TP_printk("c=%p ABORT %08x ac=%d", + TP_printk("c=%08x ABORT %08x ac=%d", __entry->call, __entry->serial, __entry->abort_code) @@ -748,20 +751,20 @@ TRACE_EVENT(rxrpc_rx_rwind_change, TP_ARGS(call, serial, rwind, wake), TP_STRUCT__entry( - __field(struct rxrpc_call *, call ) + __field(unsigned int, call ) __field(rxrpc_serial_t, serial ) __field(u32, rwind ) __field(bool, wake ) ), TP_fast_assign( - __entry->call = call; + __entry->call = call->debug_id; __entry->serial = serial; __entry->rwind = rwind; __entry->wake = wake; ), - TP_printk("c=%p %08x rw=%u%s", + TP_printk("c=%08x %08x rw=%u%s", __entry->call, __entry->serial, __entry->rwind, @@ -775,7 +778,7 @@ TRACE_EVENT(rxrpc_tx_data, TP_ARGS(call, seq, serial, flags, retrans, lose), TP_STRUCT__entry( - __field(struct rxrpc_call *, call ) + __field(unsigned int, call ) __field(rxrpc_seq_t, seq ) __field(rxrpc_serial_t, serial ) __field(u8, flags ) @@ -784,7 +787,7 @@ TRACE_EVENT(rxrpc_tx_data, ), TP_fast_assign( - __entry->call = call; + __entry->call = call->debug_id; __entry->seq = seq; __entry->serial = serial; __entry->flags = flags; @@ -792,7 +795,7 @@ TRACE_EVENT(rxrpc_tx_data, __entry->lose = lose; ), - TP_printk("c=%p DATA %08x q=%08x fl=%02x%s%s", + TP_printk("c=%08x DATA %08x q=%08x fl=%02x%s%s", __entry->call, __entry->serial, __entry->seq, @@ -809,7 +812,7 @@ TRACE_EVENT(rxrpc_tx_ack, TP_ARGS(call, serial, ack_first, ack_serial, reason, n_acks), TP_STRUCT__entry( - __field(struct rxrpc_call *, call ) + __field(unsigned int, call ) __field(rxrpc_serial_t, serial ) __field(rxrpc_seq_t, ack_first ) __field(rxrpc_serial_t, ack_serial ) @@ -818,7 +821,7 @@ TRACE_EVENT(rxrpc_tx_ack, ), TP_fast_assign( - __entry->call = call; + __entry->call = call ? call->debug_id : 0; __entry->serial = serial; __entry->ack_first = ack_first; __entry->ack_serial = ack_serial; @@ -826,7 +829,7 @@ TRACE_EVENT(rxrpc_tx_ack, __entry->n_acks = n_acks; ), - TP_printk(" c=%p ACK %08x %s f=%08x r=%08x n=%u", + TP_printk(" c=%08x ACK %08x %s f=%08x r=%08x n=%u", __entry->call, __entry->serial, __print_symbolic(__entry->reason, rxrpc_ack_names), @@ -842,7 +845,7 @@ TRACE_EVENT(rxrpc_receive, TP_ARGS(call, why, serial, seq), TP_STRUCT__entry( - __field(struct rxrpc_call *, call ) + __field(unsigned int, call ) __field(enum rxrpc_receive_trace, why ) __field(rxrpc_serial_t, serial ) __field(rxrpc_seq_t, seq ) @@ -851,7 +854,7 @@ TRACE_EVENT(rxrpc_receive, ), TP_fast_assign( - __entry->call = call; + __entry->call = call->debug_id; __entry->why = why; __entry->serial = serial; __entry->seq = seq; @@ -859,7 +862,7 @@ TRACE_EVENT(rxrpc_receive, __entry->top = call->rx_top; ), - TP_printk("c=%p %s r=%08x q=%08x w=%08x-%08x", + TP_printk("c=%08x %s r=%08x q=%08x w=%08x-%08x", __entry->call, __print_symbolic(__entry->why, rxrpc_receive_traces), __entry->serial, @@ -876,7 +879,7 @@ TRACE_EVENT(rxrpc_recvmsg, TP_ARGS(call, why, seq, offset, len, ret), TP_STRUCT__entry( - __field(struct rxrpc_call *, call ) + __field(unsigned int, call ) __field(enum rxrpc_recvmsg_trace, why ) __field(rxrpc_seq_t, seq ) __field(unsigned int, offset ) @@ -885,7 +888,7 @@ TRACE_EVENT(rxrpc_recvmsg, ), TP_fast_assign( - __entry->call = call; + __entry->call = call->debug_id; __entry->why = why; __entry->seq = seq; __entry->offset = offset; @@ -893,7 +896,7 @@ TRACE_EVENT(rxrpc_recvmsg, __entry->ret = ret; ), - TP_printk("c=%p %s q=%08x o=%u l=%u ret=%d", + TP_printk("c=%08x %s q=%08x o=%u l=%u ret=%d", __entry->call, __print_symbolic(__entry->why, rxrpc_recvmsg_traces), __entry->seq, @@ -909,18 +912,18 @@ TRACE_EVENT(rxrpc_rtt_tx, TP_ARGS(call, why, send_serial), TP_STRUCT__entry( - __field(struct rxrpc_call *, call ) + __field(unsigned int, call ) __field(enum rxrpc_rtt_tx_trace, why ) __field(rxrpc_serial_t, send_serial ) ), TP_fast_assign( - __entry->call = call; + __entry->call = call->debug_id; __entry->why = why; __entry->send_serial = send_serial; ), - TP_printk("c=%p %s sr=%08x", + TP_printk("c=%08x %s sr=%08x", __entry->call, __print_symbolic(__entry->why, rxrpc_rtt_tx_traces), __entry->send_serial) @@ -934,7 +937,7 @@ TRACE_EVENT(rxrpc_rtt_rx, TP_ARGS(call, why, send_serial, resp_serial, rtt, nr, avg), TP_STRUCT__entry( - __field(struct rxrpc_call *, call ) + __field(unsigned int, call ) __field(enum rxrpc_rtt_rx_trace, why ) __field(u8, nr ) __field(rxrpc_serial_t, send_serial ) @@ -944,7 +947,7 @@ TRACE_EVENT(rxrpc_rtt_rx, ), TP_fast_assign( - __entry->call = call; + __entry->call = call->debug_id; __entry->why = why; __entry->send_serial = send_serial; __entry->resp_serial = resp_serial; @@ -953,7 +956,7 @@ TRACE_EVENT(rxrpc_rtt_rx, __entry->avg = avg; ), - TP_printk("c=%p %s sr=%08x rr=%08x rtt=%lld nr=%u avg=%lld", + TP_printk("c=%08x %s sr=%08x rr=%08x rtt=%lld nr=%u avg=%lld", __entry->call, __print_symbolic(__entry->why, rxrpc_rtt_rx_traces), __entry->send_serial, @@ -970,7 +973,7 @@ TRACE_EVENT(rxrpc_timer, TP_ARGS(call, why, now), TP_STRUCT__entry( - __field(struct rxrpc_call *, call ) + __field(unsigned int, call ) __field(enum rxrpc_timer_trace, why ) __field(long, now ) __field(long, ack_at ) @@ -984,7 +987,7 @@ TRACE_EVENT(rxrpc_timer, ), TP_fast_assign( - __entry->call = call; + __entry->call = call->debug_id; __entry->why = why; __entry->now = now; __entry->ack_at = call->ack_at; @@ -996,7 +999,7 @@ TRACE_EVENT(rxrpc_timer, __entry->timer = call->timer.expires; ), - TP_printk("c=%p %s a=%ld la=%ld r=%ld xr=%ld xq=%ld xt=%ld t=%ld", + TP_printk("c=%08x %s a=%ld la=%ld r=%ld xr=%ld xq=%ld xt=%ld t=%ld", __entry->call, __print_symbolic(__entry->why, rxrpc_timer_traces), __entry->ack_at - __entry->now, @@ -1039,7 +1042,7 @@ TRACE_EVENT(rxrpc_propose_ack, outcome), TP_STRUCT__entry( - __field(struct rxrpc_call *, call ) + __field(unsigned int, call ) __field(enum rxrpc_propose_ack_trace, why ) __field(rxrpc_serial_t, serial ) __field(u8, ack_reason ) @@ -1049,7 +1052,7 @@ TRACE_EVENT(rxrpc_propose_ack, ), TP_fast_assign( - __entry->call = call; + __entry->call = call->debug_id; __entry->why = why; __entry->serial = serial; __entry->ack_reason = ack_reason; @@ -1058,7 +1061,7 @@ TRACE_EVENT(rxrpc_propose_ack, __entry->outcome = outcome; ), - TP_printk("c=%p %s %s r=%08x i=%u b=%u%s", + TP_printk("c=%08x %s %s r=%08x i=%u b=%u%s", __entry->call, __print_symbolic(__entry->why, rxrpc_propose_ack_traces), __print_symbolic(__entry->ack_reason, rxrpc_ack_names), @@ -1075,20 +1078,20 @@ TRACE_EVENT(rxrpc_retransmit, TP_ARGS(call, seq, annotation, expiry), TP_STRUCT__entry( - __field(struct rxrpc_call *, call ) + __field(unsigned int, call ) __field(rxrpc_seq_t, seq ) __field(u8, annotation ) __field(s64, expiry ) ), TP_fast_assign( - __entry->call = call; + __entry->call = call->debug_id; __entry->seq = seq; __entry->annotation = annotation; __entry->expiry = expiry; ), - TP_printk("c=%p q=%x a=%02x xp=%lld", + TP_printk("c=%08x q=%x a=%02x xp=%lld", __entry->call, __entry->seq, __entry->annotation, @@ -1102,7 +1105,7 @@ TRACE_EVENT(rxrpc_congest, TP_ARGS(call, summary, ack_serial, change), TP_STRUCT__entry( - __field(struct rxrpc_call *, call ) + __field(unsigned int, call ) __field(enum rxrpc_congest_change, change ) __field(rxrpc_seq_t, hard_ack ) __field(rxrpc_seq_t, top ) @@ -1112,7 +1115,7 @@ TRACE_EVENT(rxrpc_congest, ), TP_fast_assign( - __entry->call = call; + __entry->call = call->debug_id; __entry->change = change; __entry->hard_ack = call->tx_hard_ack; __entry->top = call->tx_top; @@ -1121,7 +1124,7 @@ TRACE_EVENT(rxrpc_congest, memcpy(&__entry->sum, summary, sizeof(__entry->sum)); ), - TP_printk("c=%p r=%08x %s q=%08x %s cw=%u ss=%u nr=%u,%u nw=%u,%u r=%u b=%u u=%u d=%u l=%x%s%s%s", + TP_printk("c=%08x r=%08x %s q=%08x %s cw=%u ss=%u nr=%u,%u nw=%u,%u r=%u b=%u u=%u d=%u l=%x%s%s%s", __entry->call, __entry->ack_serial, __print_symbolic(__entry->sum.ack_reason, rxrpc_ack_names), @@ -1146,16 +1149,16 @@ TRACE_EVENT(rxrpc_disconnect_call, TP_ARGS(call), TP_STRUCT__entry( - __field(struct rxrpc_call *, call ) + __field(unsigned int, call ) __field(u32, abort_code ) ), TP_fast_assign( - __entry->call = call; + __entry->call = call->debug_id; __entry->abort_code = call->abort_code; ), - TP_printk("c=%p ab=%08x", + TP_printk("c=%08x ab=%08x", __entry->call, __entry->abort_code) ); @@ -1166,16 +1169,16 @@ TRACE_EVENT(rxrpc_improper_term, TP_ARGS(call), TP_STRUCT__entry( - __field(struct rxrpc_call *, call ) + __field(unsigned int, call ) __field(u32, abort_code ) ), TP_fast_assign( - __entry->call = call; + __entry->call = call->debug_id; __entry->abort_code = call->abort_code; ), - TP_printk("c=%p ab=%08x", + TP_printk("c=%08x ab=%08x", __entry->call, __entry->abort_code) ); @@ -1187,18 +1190,18 @@ TRACE_EVENT(rxrpc_rx_eproto, TP_ARGS(call, serial, why), TP_STRUCT__entry( - __field(struct rxrpc_call *, call ) + __field(unsigned int, call ) __field(rxrpc_serial_t, serial ) __field(const char *, why ) ), TP_fast_assign( - __entry->call = call; + __entry->call = call->debug_id; __entry->serial = serial; __entry->why = why; ), - TP_printk("c=%p EPROTO %08x %s", + TP_printk("c=%08x EPROTO %08x %s", __entry->call, __entry->serial, __entry->why) @@ -1210,20 +1213,20 @@ TRACE_EVENT(rxrpc_connect_call, TP_ARGS(call), TP_STRUCT__entry( - __field(struct rxrpc_call *, call ) + __field(unsigned int, call ) __field(unsigned long, user_call_ID ) __field(u32, cid ) __field(u32, call_id ) ), TP_fast_assign( - __entry->call = call; + __entry->call = call->debug_id; __entry->user_call_ID = call->user_call_ID; __entry->cid = call->cid; __entry->call_id = call->call_id; ), - TP_printk("c=%p u=%p %08x:%08x", + TP_printk("c=%08x u=%p %08x:%08x", __entry->call, (void *)__entry->user_call_ID, __entry->cid, @@ -1236,18 +1239,18 @@ TRACE_EVENT(rxrpc_resend, TP_ARGS(call, ix), TP_STRUCT__entry( - __field(struct rxrpc_call *, call ) + __field(unsigned int, call ) __field(int, ix ) __array(u8, anno, 64 ) ), TP_fast_assign( - __entry->call = call; + __entry->call = call->debug_id; __entry->ix = ix; memcpy(__entry->anno, call->rxtx_annotations, 64); ), - TP_printk("c=%p ix=%u a=%64phN", + TP_printk("c=%08x ix=%u a=%64phN", __entry->call, __entry->ix, __entry->anno) diff --git a/net/rxrpc/af_rxrpc.c b/net/rxrpc/af_rxrpc.c index 9e1c2c6b6a67..ec5ec68be1aa 100644 --- a/net/rxrpc/af_rxrpc.c +++ b/net/rxrpc/af_rxrpc.c @@ -40,6 +40,7 @@ static const struct proto_ops rxrpc_rpc_ops; /* current debugging ID */ atomic_t rxrpc_debug_id; +EXPORT_SYMBOL(rxrpc_debug_id); /* count of skbs currently in use */ atomic_t rxrpc_n_tx_skbs, rxrpc_n_rx_skbs; @@ -267,6 +268,7 @@ static int rxrpc_listen(struct socket *sock, int backlog) * @gfp: The allocation constraints * @notify_rx: Where to send notifications instead of socket queue * @upgrade: Request service upgrade for call + * @debug_id: The debug ID for tracing to be assigned to the call * * Allow a kernel service to begin a call on the nominated socket. This just * sets up all the internal tracking structures and allocates connection and @@ -282,7 +284,8 @@ struct rxrpc_call *rxrpc_kernel_begin_call(struct socket *sock, s64 tx_total_len, gfp_t gfp, rxrpc_notify_rx_t notify_rx, - bool upgrade) + bool upgrade, + unsigned int debug_id) { struct rxrpc_conn_parameters cp; struct rxrpc_call_params p; @@ -314,7 +317,7 @@ struct rxrpc_call *rxrpc_kernel_begin_call(struct socket *sock, cp.exclusive = false; cp.upgrade = upgrade; cp.service_id = srx->srx_service; - call = rxrpc_new_client_call(rx, &cp, srx, &p, gfp); + call = rxrpc_new_client_call(rx, &cp, srx, &p, gfp, debug_id); /* The socket has been unlocked. */ if (!IS_ERR(call)) { call->notify_rx = notify_rx; diff --git a/net/rxrpc/ar-internal.h b/net/rxrpc/ar-internal.h index 416688381eb7..9c9817ddafc5 100644 --- a/net/rxrpc/ar-internal.h +++ b/net/rxrpc/ar-internal.h @@ -691,7 +691,6 @@ struct rxrpc_send_params { * af_rxrpc.c */ extern atomic_t rxrpc_n_tx_skbs, rxrpc_n_rx_skbs; -extern atomic_t rxrpc_debug_id; extern struct workqueue_struct *rxrpc_workqueue; /* @@ -732,11 +731,12 @@ extern unsigned int rxrpc_max_call_lifetime; extern struct kmem_cache *rxrpc_call_jar; struct rxrpc_call *rxrpc_find_call_by_user_ID(struct rxrpc_sock *, unsigned long); -struct rxrpc_call *rxrpc_alloc_call(struct rxrpc_sock *, gfp_t); +struct rxrpc_call *rxrpc_alloc_call(struct rxrpc_sock *, gfp_t, unsigned int); struct rxrpc_call *rxrpc_new_client_call(struct rxrpc_sock *, struct rxrpc_conn_parameters *, struct sockaddr_rxrpc *, - struct rxrpc_call_params *, gfp_t); + struct rxrpc_call_params *, gfp_t, + unsigned int); int rxrpc_retry_client_call(struct rxrpc_sock *, struct rxrpc_call *, struct rxrpc_conn_parameters *, @@ -822,7 +822,7 @@ static inline bool __rxrpc_abort_call(const char *why, struct rxrpc_call *call, rxrpc_seq_t seq, u32 abort_code, int error) { - trace_rxrpc_abort(why, call->cid, call->call_id, seq, + trace_rxrpc_abort(call->debug_id, why, call->cid, call->call_id, seq, abort_code, error); return __rxrpc_set_call_completion(call, RXRPC_CALL_LOCALLY_ABORTED, abort_code, error); diff --git a/net/rxrpc/call_accept.c b/net/rxrpc/call_accept.c index 3028298ca561..92ebd1d7e0bb 100644 --- a/net/rxrpc/call_accept.c +++ b/net/rxrpc/call_accept.c @@ -34,7 +34,8 @@ static int rxrpc_service_prealloc_one(struct rxrpc_sock *rx, struct rxrpc_backlog *b, rxrpc_notify_rx_t notify_rx, rxrpc_user_attach_call_t user_attach_call, - unsigned long user_call_ID, gfp_t gfp) + unsigned long user_call_ID, gfp_t gfp, + unsigned int debug_id) { const void *here = __builtin_return_address(0); struct rxrpc_call *call; @@ -94,7 +95,7 @@ static int rxrpc_service_prealloc_one(struct rxrpc_sock *rx, /* Now it gets complicated, because calls get registered with the * socket here, particularly if a user ID is preassigned by the user. */ - call = rxrpc_alloc_call(rx, gfp); + call = rxrpc_alloc_call(rx, gfp, debug_id); if (!call) return -ENOMEM; call->flags |= (1 << RXRPC_CALL_IS_SERVICE); @@ -174,7 +175,8 @@ int rxrpc_service_prealloc(struct rxrpc_sock *rx, gfp_t gfp) if (rx->discard_new_call) return 0; - while (rxrpc_service_prealloc_one(rx, b, NULL, NULL, 0, gfp) == 0) + while (rxrpc_service_prealloc_one(rx, b, NULL, NULL, 0, gfp, + atomic_inc_return(&rxrpc_debug_id)) == 0) ; return 0; @@ -347,7 +349,7 @@ struct rxrpc_call *rxrpc_new_incoming_call(struct rxrpc_local *local, service_id == rx->second_service)) goto found_service; - trace_rxrpc_abort("INV", sp->hdr.cid, sp->hdr.callNumber, sp->hdr.seq, + trace_rxrpc_abort(0, "INV", sp->hdr.cid, sp->hdr.callNumber, sp->hdr.seq, RX_INVALID_OPERATION, EOPNOTSUPP); skb->mark = RXRPC_SKB_MARK_LOCAL_ABORT; skb->priority = RX_INVALID_OPERATION; @@ -358,7 +360,7 @@ found_service: spin_lock(&rx->incoming_lock); if (rx->sk.sk_state == RXRPC_SERVER_LISTEN_DISABLED || rx->sk.sk_state == RXRPC_CLOSE) { - trace_rxrpc_abort("CLS", sp->hdr.cid, sp->hdr.callNumber, + trace_rxrpc_abort(0, "CLS", sp->hdr.cid, sp->hdr.callNumber, sp->hdr.seq, RX_INVALID_OPERATION, ESHUTDOWN); skb->mark = RXRPC_SKB_MARK_LOCAL_ABORT; skb->priority = RX_INVALID_OPERATION; @@ -635,6 +637,7 @@ out_discard: * @user_attach_call: Func to attach call to user_call_ID * @user_call_ID: The tag to attach to the preallocated call * @gfp: The allocation conditions. + * @debug_id: The tracing debug ID. * * Charge up the socket with preallocated calls, each with a user ID. A * function should be provided to effect the attachment from the user's side. @@ -645,7 +648,8 @@ out_discard: int rxrpc_kernel_charge_accept(struct socket *sock, rxrpc_notify_rx_t notify_rx, rxrpc_user_attach_call_t user_attach_call, - unsigned long user_call_ID, gfp_t gfp) + unsigned long user_call_ID, gfp_t gfp, + unsigned int debug_id) { struct rxrpc_sock *rx = rxrpc_sk(sock->sk); struct rxrpc_backlog *b = rx->backlog; @@ -655,6 +659,6 @@ int rxrpc_kernel_charge_accept(struct socket *sock, return rxrpc_service_prealloc_one(rx, b, notify_rx, user_attach_call, user_call_ID, - gfp); + gfp, debug_id); } EXPORT_SYMBOL(rxrpc_kernel_charge_accept); diff --git a/net/rxrpc/call_object.c b/net/rxrpc/call_object.c index 0b2db38dd32d..147657dfe757 100644 --- a/net/rxrpc/call_object.c +++ b/net/rxrpc/call_object.c @@ -99,7 +99,8 @@ found_extant_call: /* * allocate a new call */ -struct rxrpc_call *rxrpc_alloc_call(struct rxrpc_sock *rx, gfp_t gfp) +struct rxrpc_call *rxrpc_alloc_call(struct rxrpc_sock *rx, gfp_t gfp, + unsigned int debug_id) { struct rxrpc_call *call; @@ -138,7 +139,7 @@ struct rxrpc_call *rxrpc_alloc_call(struct rxrpc_sock *rx, gfp_t gfp) spin_lock_init(&call->notify_lock); rwlock_init(&call->state_lock); atomic_set(&call->usage, 1); - call->debug_id = atomic_inc_return(&rxrpc_debug_id); + call->debug_id = debug_id; call->tx_total_len = -1; call->next_rx_timo = 20 * HZ; call->next_req_timo = 1 * HZ; @@ -166,14 +167,15 @@ nomem: */ static struct rxrpc_call *rxrpc_alloc_client_call(struct rxrpc_sock *rx, struct sockaddr_rxrpc *srx, - gfp_t gfp) + gfp_t gfp, + unsigned int debug_id) { struct rxrpc_call *call; ktime_t now; _enter(""); - call = rxrpc_alloc_call(rx, gfp); + call = rxrpc_alloc_call(rx, gfp, debug_id); if (!call) return ERR_PTR(-ENOMEM); call->state = RXRPC_CALL_CLIENT_AWAIT_CONN; @@ -214,7 +216,8 @@ struct rxrpc_call *rxrpc_new_client_call(struct rxrpc_sock *rx, struct rxrpc_conn_parameters *cp, struct sockaddr_rxrpc *srx, struct rxrpc_call_params *p, - gfp_t gfp) + gfp_t gfp, + unsigned int debug_id) __releases(&rx->sk.sk_lock.slock) { struct rxrpc_call *call, *xcall; @@ -225,7 +228,7 @@ struct rxrpc_call *rxrpc_new_client_call(struct rxrpc_sock *rx, _enter("%p,%lx", rx, p->user_call_ID); - call = rxrpc_alloc_client_call(rx, srx, gfp); + call = rxrpc_alloc_client_call(rx, srx, gfp, debug_id); if (IS_ERR(call)) { release_sock(&rx->sk); _leave(" = %ld", PTR_ERR(call)); diff --git a/net/rxrpc/conn_event.c b/net/rxrpc/conn_event.c index b1dfae107431..d2ec3fd593e8 100644 --- a/net/rxrpc/conn_event.c +++ b/net/rxrpc/conn_event.c @@ -160,7 +160,8 @@ static void rxrpc_abort_calls(struct rxrpc_connection *conn, lockdep_is_held(&conn->channel_lock)); if (call) { if (compl == RXRPC_CALL_LOCALLY_ABORTED) - trace_rxrpc_abort("CON", call->cid, + trace_rxrpc_abort(call->debug_id, + "CON", call->cid, call->call_id, 0, abort_code, error); if (rxrpc_set_call_completion(call, compl, diff --git a/net/rxrpc/input.c b/net/rxrpc/input.c index 6fc61400337f..2a868fdab0ae 100644 --- a/net/rxrpc/input.c +++ b/net/rxrpc/input.c @@ -1307,21 +1307,21 @@ out_unlock: wrong_security: rcu_read_unlock(); - trace_rxrpc_abort("SEC", sp->hdr.cid, sp->hdr.callNumber, sp->hdr.seq, + trace_rxrpc_abort(0, "SEC", sp->hdr.cid, sp->hdr.callNumber, sp->hdr.seq, RXKADINCONSISTENCY, EBADMSG); skb->priority = RXKADINCONSISTENCY; goto post_abort; reupgrade: rcu_read_unlock(); - trace_rxrpc_abort("UPG", sp->hdr.cid, sp->hdr.callNumber, sp->hdr.seq, + trace_rxrpc_abort(0, "UPG", sp->hdr.cid, sp->hdr.callNumber, sp->hdr.seq, RX_PROTOCOL_ERROR, EBADMSG); goto protocol_error; bad_message_unlock: rcu_read_unlock(); bad_message: - trace_rxrpc_abort("BAD", sp->hdr.cid, sp->hdr.callNumber, sp->hdr.seq, + trace_rxrpc_abort(0, "BAD", sp->hdr.cid, sp->hdr.callNumber, sp->hdr.seq, RX_PROTOCOL_ERROR, EBADMSG); protocol_error: skb->priority = RX_PROTOCOL_ERROR; diff --git a/net/rxrpc/sendmsg.c b/net/rxrpc/sendmsg.c index 09f2a3e05221..8503f279b467 100644 --- a/net/rxrpc/sendmsg.c +++ b/net/rxrpc/sendmsg.c @@ -579,7 +579,8 @@ rxrpc_new_client_call_for_sendmsg(struct rxrpc_sock *rx, struct msghdr *msg, cp.exclusive = rx->exclusive | p->exclusive; cp.upgrade = p->upgrade; cp.service_id = srx->srx_service; - call = rxrpc_new_client_call(rx, &cp, srx, &p->call, GFP_KERNEL); + call = rxrpc_new_client_call(rx, &cp, srx, &p->call, GFP_KERNEL, + atomic_inc_return(&rxrpc_debug_id)); /* The socket is now unlocked */ _leave(" = %p\n", call); -- cgit v1.2.3 From 1bae5d229532b4e8dfd5728cb3b8373bc9eec9eb Mon Sep 17 00:00:00 2001 From: David Howells Date: Tue, 27 Mar 2018 23:08:20 +0100 Subject: rxrpc: Trace call completion Add a tracepoint to track rxrpc calls moving into the completed state and to log the completion type and the recorded error value and abort code. Signed-off-by: David Howells --- include/trace/events/rxrpc.h | 33 +++++++++++++++++++++++++++++++++ net/rxrpc/ar-internal.h | 1 + 2 files changed, 34 insertions(+) (limited to 'include/trace') diff --git a/include/trace/events/rxrpc.h b/include/trace/events/rxrpc.h index 4d2c2d35c5cb..2ea788f6f95d 100644 --- a/include/trace/events/rxrpc.h +++ b/include/trace/events/rxrpc.h @@ -400,6 +400,13 @@ enum rxrpc_congest_change { EM(RXRPC_ACK_IDLE, "IDL") \ E_(RXRPC_ACK__INVALID, "-?-") +#define rxrpc_completions \ + EM(RXRPC_CALL_SUCCEEDED, "Succeeded") \ + EM(RXRPC_CALL_REMOTELY_ABORTED, "RemoteAbort") \ + EM(RXRPC_CALL_LOCALLY_ABORTED, "LocalAbort") \ + EM(RXRPC_CALL_LOCAL_ERROR, "LocalError") \ + E_(RXRPC_CALL_NETWORK_ERROR, "NetError") + /* * Export enum symbols via userspace. */ @@ -624,6 +631,32 @@ TRACE_EVENT(rxrpc_abort, __entry->abort_code, __entry->error, __entry->why) ); +TRACE_EVENT(rxrpc_call_complete, + TP_PROTO(struct rxrpc_call *call), + + TP_ARGS(call), + + TP_STRUCT__entry( + __field(unsigned int, call ) + __field(enum rxrpc_call_completion, compl ) + __field(int, error ) + __field(u32, abort_code ) + ), + + TP_fast_assign( + __entry->call = call->debug_id; + __entry->compl = call->completion; + __entry->error = call->error; + __entry->abort_code = call->abort_code; + ), + + TP_printk("c=%08x %s r=%d ac=%d", + __entry->call, + __print_symbolic(__entry->compl, rxrpc_completions), + __entry->error, + __entry->abort_code) + ); + TRACE_EVENT(rxrpc_transmit, TP_PROTO(struct rxrpc_call *call, enum rxrpc_transmit_trace why), diff --git a/net/rxrpc/ar-internal.h b/net/rxrpc/ar-internal.h index 9c9817ddafc5..21cf164b6d85 100644 --- a/net/rxrpc/ar-internal.h +++ b/net/rxrpc/ar-internal.h @@ -778,6 +778,7 @@ static inline bool __rxrpc_set_call_completion(struct rxrpc_call *call, call->error = error; call->completion = compl, call->state = RXRPC_CALL_COMPLETE; + trace_rxrpc_call_complete(call); wake_up(&call->waitq); return true; } -- cgit v1.2.3 From c105547501016897194358b11451608a8d5f9a02 Mon Sep 17 00:00:00 2001 From: Alexei Starovoitov Date: Wed, 28 Mar 2018 12:05:32 -0700 Subject: treewide: remove large struct-pass-by-value from tracepoint arguments - fix trace_hfi1_ctxt_info() to pass large struct by reference instead of by value - convert 'type array[]' tracepoint arguments into 'type *array', since compiler will warn that sizeof('type array[]') == sizeof('type *array') and later should be used instead The CAST_TO_U64 macro in the later patch will enforce that tracepoint arguments can only be integers, pointers, or less than 8 byte structures. Larger structures should be passed by reference. Signed-off-by: Alexei Starovoitov Signed-off-by: Daniel Borkmann --- drivers/infiniband/hw/hfi1/file_ops.c | 2 +- drivers/infiniband/hw/hfi1/trace_ctxts.h | 12 ++++++------ include/trace/events/f2fs.h | 2 +- net/wireless/trace.h | 2 +- sound/firewire/amdtp-stream-trace.h | 2 +- 5 files changed, 10 insertions(+), 10 deletions(-) (limited to 'include/trace') diff --git a/drivers/infiniband/hw/hfi1/file_ops.c b/drivers/infiniband/hw/hfi1/file_ops.c index 41fafebe3b0d..da4aa1a95b11 100644 --- a/drivers/infiniband/hw/hfi1/file_ops.c +++ b/drivers/infiniband/hw/hfi1/file_ops.c @@ -1153,7 +1153,7 @@ static int get_ctxt_info(struct hfi1_filedata *fd, unsigned long arg, u32 len) cinfo.sdma_ring_size = fd->cq->nentries; cinfo.rcvegr_size = uctxt->egrbufs.rcvtid_size; - trace_hfi1_ctxt_info(uctxt->dd, uctxt->ctxt, fd->subctxt, cinfo); + trace_hfi1_ctxt_info(uctxt->dd, uctxt->ctxt, fd->subctxt, &cinfo); if (copy_to_user((void __user *)arg, &cinfo, len)) return -EFAULT; diff --git a/drivers/infiniband/hw/hfi1/trace_ctxts.h b/drivers/infiniband/hw/hfi1/trace_ctxts.h index 4eb4cc798035..e00c8a7d559c 100644 --- a/drivers/infiniband/hw/hfi1/trace_ctxts.h +++ b/drivers/infiniband/hw/hfi1/trace_ctxts.h @@ -106,7 +106,7 @@ TRACE_EVENT(hfi1_uctxtdata, TRACE_EVENT(hfi1_ctxt_info, TP_PROTO(struct hfi1_devdata *dd, unsigned int ctxt, unsigned int subctxt, - struct hfi1_ctxt_info cinfo), + struct hfi1_ctxt_info *cinfo), TP_ARGS(dd, ctxt, subctxt, cinfo), TP_STRUCT__entry(DD_DEV_ENTRY(dd) __field(unsigned int, ctxt) @@ -120,11 +120,11 @@ TRACE_EVENT(hfi1_ctxt_info, TP_fast_assign(DD_DEV_ASSIGN(dd); __entry->ctxt = ctxt; __entry->subctxt = subctxt; - __entry->egrtids = cinfo.egrtids; - __entry->rcvhdrq_cnt = cinfo.rcvhdrq_cnt; - __entry->rcvhdrq_size = cinfo.rcvhdrq_entsize; - __entry->sdma_ring_size = cinfo.sdma_ring_size; - __entry->rcvegr_size = cinfo.rcvegr_size; + __entry->egrtids = cinfo->egrtids; + __entry->rcvhdrq_cnt = cinfo->rcvhdrq_cnt; + __entry->rcvhdrq_size = cinfo->rcvhdrq_entsize; + __entry->sdma_ring_size = cinfo->sdma_ring_size; + __entry->rcvegr_size = cinfo->rcvegr_size; ), TP_printk("[%s] ctxt %u:%u " CINFO_FMT, __get_str(dev), diff --git a/include/trace/events/f2fs.h b/include/trace/events/f2fs.h index 06c87f9f720c..795698925d20 100644 --- a/include/trace/events/f2fs.h +++ b/include/trace/events/f2fs.h @@ -491,7 +491,7 @@ DEFINE_EVENT(f2fs__truncate_node, f2fs_truncate_node, TRACE_EVENT(f2fs_truncate_partial_nodes, - TP_PROTO(struct inode *inode, nid_t nid[], int depth, int err), + TP_PROTO(struct inode *inode, nid_t *nid, int depth, int err), TP_ARGS(inode, nid, depth, err), diff --git a/net/wireless/trace.h b/net/wireless/trace.h index 5152938b358d..018c81fa72fb 100644 --- a/net/wireless/trace.h +++ b/net/wireless/trace.h @@ -3137,7 +3137,7 @@ TRACE_EVENT(rdev_start_radar_detection, TRACE_EVENT(rdev_set_mcast_rate, TP_PROTO(struct wiphy *wiphy, struct net_device *netdev, - int mcast_rate[NUM_NL80211_BANDS]), + int *mcast_rate), TP_ARGS(wiphy, netdev, mcast_rate), TP_STRUCT__entry( WIPHY_ENTRY diff --git a/sound/firewire/amdtp-stream-trace.h b/sound/firewire/amdtp-stream-trace.h index ea0d486652c8..54cdd4ffa9ce 100644 --- a/sound/firewire/amdtp-stream-trace.h +++ b/sound/firewire/amdtp-stream-trace.h @@ -14,7 +14,7 @@ #include TRACE_EVENT(in_packet, - TP_PROTO(const struct amdtp_stream *s, u32 cycles, u32 cip_header[2], unsigned int payload_length, unsigned int index), + TP_PROTO(const struct amdtp_stream *s, u32 cycles, u32 *cip_header, unsigned int payload_length, unsigned int index), TP_ARGS(s, cycles, cip_header, payload_length, index), TP_STRUCT__entry( __field(unsigned int, second) -- cgit v1.2.3 From c4f6699dfcb8558d138fe838f741b2c10f416cf9 Mon Sep 17 00:00:00 2001 From: Alexei Starovoitov Date: Wed, 28 Mar 2018 12:05:37 -0700 Subject: bpf: introduce BPF_RAW_TRACEPOINT Introduce BPF_PROG_TYPE_RAW_TRACEPOINT bpf program type to access kernel internal arguments of the tracepoints in their raw form. >From bpf program point of view the access to the arguments look like: struct bpf_raw_tracepoint_args { __u64 args[0]; }; int bpf_prog(struct bpf_raw_tracepoint_args *ctx) { // program can read args[N] where N depends on tracepoint // and statically verified at program load+attach time } kprobe+bpf infrastructure allows programs access function arguments. This feature allows programs access raw tracepoint arguments. Similar to proposed 'dynamic ftrace events' there are no abi guarantees to what the tracepoints arguments are and what their meaning is. The program needs to type cast args properly and use bpf_probe_read() helper to access struct fields when argument is a pointer. For every tracepoint __bpf_trace_##call function is prepared. In assembler it looks like: (gdb) disassemble __bpf_trace_xdp_exception Dump of assembler code for function __bpf_trace_xdp_exception: 0xffffffff81132080 <+0>: mov %ecx,%ecx 0xffffffff81132082 <+2>: jmpq 0xffffffff811231f0 where TRACE_EVENT(xdp_exception, TP_PROTO(const struct net_device *dev, const struct bpf_prog *xdp, u32 act), The above assembler snippet is casting 32-bit 'act' field into 'u64' to pass into bpf_trace_run3(), while 'dev' and 'xdp' args are passed as-is. All of ~500 of __bpf_trace_*() functions are only 5-10 byte long and in total this approach adds 7k bytes to .text. This approach gives the lowest possible overhead while calling trace_xdp_exception() from kernel C code and transitioning into bpf land. Since tracepoint+bpf are used at speeds of 1M+ events per second this is valuable optimization. The new BPF_RAW_TRACEPOINT_OPEN sys_bpf command is introduced that returns anon_inode FD of 'bpf-raw-tracepoint' object. The user space looks like: // load bpf prog with BPF_PROG_TYPE_RAW_TRACEPOINT type prog_fd = bpf_prog_load(...); // receive anon_inode fd for given bpf_raw_tracepoint with prog attached raw_tp_fd = bpf_raw_tracepoint_open("xdp_exception", prog_fd); Ctrl-C of tracing daemon or cmdline tool that uses this feature will automatically detach bpf program, unload it and unregister tracepoint probe. On the kernel side the __bpf_raw_tp_map section of pointers to tracepoint definition and to __bpf_trace_*() probe function is used to find a tracepoint with "xdp_exception" name and corresponding __bpf_trace_xdp_exception() probe function which are passed to tracepoint_probe_register() to connect probe with tracepoint. Addition of bpf_raw_tracepoint doesn't interfere with ftrace and perf tracepoint mechanisms. perf_event_open() can be used in parallel on the same tracepoint. Multiple bpf_raw_tracepoint_open("xdp_exception", prog_fd) are permitted. Each with its own bpf program. The kernel will execute all tracepoint probes and all attached bpf programs. In the future bpf_raw_tracepoints can be extended with query/introspection logic. __bpf_raw_tp_map section logic was contributed by Steven Rostedt Signed-off-by: Alexei Starovoitov Signed-off-by: Steven Rostedt (VMware) Acked-by: Steven Rostedt (VMware) Signed-off-by: Daniel Borkmann --- include/asm-generic/vmlinux.lds.h | 10 +++ include/linux/bpf_types.h | 1 + include/linux/trace_events.h | 42 +++++++++ include/linux/tracepoint-defs.h | 6 ++ include/trace/bpf_probe.h | 92 +++++++++++++++++++ include/trace/define_trace.h | 1 + include/uapi/linux/bpf.h | 11 +++ kernel/bpf/syscall.c | 78 ++++++++++++++++ kernel/trace/bpf_trace.c | 183 ++++++++++++++++++++++++++++++++++++++ 9 files changed, 424 insertions(+) create mode 100644 include/trace/bpf_probe.h (limited to 'include/trace') diff --git a/include/asm-generic/vmlinux.lds.h b/include/asm-generic/vmlinux.lds.h index 1ab0e520d6fc..8add3493a202 100644 --- a/include/asm-generic/vmlinux.lds.h +++ b/include/asm-generic/vmlinux.lds.h @@ -178,6 +178,15 @@ #define TRACE_SYSCALLS() #endif +#ifdef CONFIG_BPF_EVENTS +#define BPF_RAW_TP() STRUCT_ALIGN(); \ + VMLINUX_SYMBOL(__start__bpf_raw_tp) = .; \ + KEEP(*(__bpf_raw_tp_map)) \ + VMLINUX_SYMBOL(__stop__bpf_raw_tp) = .; +#else +#define BPF_RAW_TP() +#endif + #ifdef CONFIG_SERIAL_EARLYCON #define EARLYCON_TABLE() STRUCT_ALIGN(); \ VMLINUX_SYMBOL(__earlycon_table) = .; \ @@ -249,6 +258,7 @@ LIKELY_PROFILE() \ BRANCH_PROFILE() \ TRACE_PRINTKS() \ + BPF_RAW_TP() \ TRACEPOINT_STR() /* diff --git a/include/linux/bpf_types.h b/include/linux/bpf_types.h index 5e2e8a49fb21..6d7243bfb0ff 100644 --- a/include/linux/bpf_types.h +++ b/include/linux/bpf_types.h @@ -19,6 +19,7 @@ BPF_PROG_TYPE(BPF_PROG_TYPE_SK_MSG, sk_msg) BPF_PROG_TYPE(BPF_PROG_TYPE_KPROBE, kprobe) BPF_PROG_TYPE(BPF_PROG_TYPE_TRACEPOINT, tracepoint) BPF_PROG_TYPE(BPF_PROG_TYPE_PERF_EVENT, perf_event) +BPF_PROG_TYPE(BPF_PROG_TYPE_RAW_TRACEPOINT, raw_tracepoint) #endif #ifdef CONFIG_CGROUP_BPF BPF_PROG_TYPE(BPF_PROG_TYPE_CGROUP_DEVICE, cg_dev) diff --git a/include/linux/trace_events.h b/include/linux/trace_events.h index 8a1442c4e513..b0357cd198b0 100644 --- a/include/linux/trace_events.h +++ b/include/linux/trace_events.h @@ -468,6 +468,9 @@ unsigned int trace_call_bpf(struct trace_event_call *call, void *ctx); int perf_event_attach_bpf_prog(struct perf_event *event, struct bpf_prog *prog); void perf_event_detach_bpf_prog(struct perf_event *event); int perf_event_query_prog_array(struct perf_event *event, void __user *info); +int bpf_probe_register(struct bpf_raw_event_map *btp, struct bpf_prog *prog); +int bpf_probe_unregister(struct bpf_raw_event_map *btp, struct bpf_prog *prog); +struct bpf_raw_event_map *bpf_find_raw_tracepoint(const char *name); #else static inline unsigned int trace_call_bpf(struct trace_event_call *call, void *ctx) { @@ -487,6 +490,18 @@ perf_event_query_prog_array(struct perf_event *event, void __user *info) { return -EOPNOTSUPP; } +static inline int bpf_probe_register(struct bpf_raw_event_map *btp, struct bpf_prog *p) +{ + return -EOPNOTSUPP; +} +static inline int bpf_probe_unregister(struct bpf_raw_event_map *btp, struct bpf_prog *p) +{ + return -EOPNOTSUPP; +} +static inline struct bpf_raw_event_map *bpf_find_raw_tracepoint(const char *name) +{ + return NULL; +} #endif enum { @@ -546,6 +561,33 @@ extern void ftrace_profile_free_filter(struct perf_event *event); void perf_trace_buf_update(void *record, u16 type); void *perf_trace_buf_alloc(int size, struct pt_regs **regs, int *rctxp); +void bpf_trace_run1(struct bpf_prog *prog, u64 arg1); +void bpf_trace_run2(struct bpf_prog *prog, u64 arg1, u64 arg2); +void bpf_trace_run3(struct bpf_prog *prog, u64 arg1, u64 arg2, + u64 arg3); +void bpf_trace_run4(struct bpf_prog *prog, u64 arg1, u64 arg2, + u64 arg3, u64 arg4); +void bpf_trace_run5(struct bpf_prog *prog, u64 arg1, u64 arg2, + u64 arg3, u64 arg4, u64 arg5); +void bpf_trace_run6(struct bpf_prog *prog, u64 arg1, u64 arg2, + u64 arg3, u64 arg4, u64 arg5, u64 arg6); +void bpf_trace_run7(struct bpf_prog *prog, u64 arg1, u64 arg2, + u64 arg3, u64 arg4, u64 arg5, u64 arg6, u64 arg7); +void bpf_trace_run8(struct bpf_prog *prog, u64 arg1, u64 arg2, + u64 arg3, u64 arg4, u64 arg5, u64 arg6, u64 arg7, + u64 arg8); +void bpf_trace_run9(struct bpf_prog *prog, u64 arg1, u64 arg2, + u64 arg3, u64 arg4, u64 arg5, u64 arg6, u64 arg7, + u64 arg8, u64 arg9); +void bpf_trace_run10(struct bpf_prog *prog, u64 arg1, u64 arg2, + u64 arg3, u64 arg4, u64 arg5, u64 arg6, u64 arg7, + u64 arg8, u64 arg9, u64 arg10); +void bpf_trace_run11(struct bpf_prog *prog, u64 arg1, u64 arg2, + u64 arg3, u64 arg4, u64 arg5, u64 arg6, u64 arg7, + u64 arg8, u64 arg9, u64 arg10, u64 arg11); +void bpf_trace_run12(struct bpf_prog *prog, u64 arg1, u64 arg2, + u64 arg3, u64 arg4, u64 arg5, u64 arg6, u64 arg7, + u64 arg8, u64 arg9, u64 arg10, u64 arg11, u64 arg12); void perf_trace_run_bpf_submit(void *raw_data, int size, int rctx, struct trace_event_call *call, u64 count, struct pt_regs *regs, struct hlist_head *head, diff --git a/include/linux/tracepoint-defs.h b/include/linux/tracepoint-defs.h index 64ed7064f1fa..22c5a46e9693 100644 --- a/include/linux/tracepoint-defs.h +++ b/include/linux/tracepoint-defs.h @@ -35,4 +35,10 @@ struct tracepoint { struct tracepoint_func __rcu *funcs; }; +struct bpf_raw_event_map { + struct tracepoint *tp; + void *bpf_func; + u32 num_args; +} __aligned(32); + #endif diff --git a/include/trace/bpf_probe.h b/include/trace/bpf_probe.h new file mode 100644 index 000000000000..505dae0bed80 --- /dev/null +++ b/include/trace/bpf_probe.h @@ -0,0 +1,92 @@ +/* SPDX-License-Identifier: GPL-2.0 */ + +#undef TRACE_SYSTEM_VAR + +#ifdef CONFIG_BPF_EVENTS + +#undef __entry +#define __entry entry + +#undef __get_dynamic_array +#define __get_dynamic_array(field) \ + ((void *)__entry + (__entry->__data_loc_##field & 0xffff)) + +#undef __get_dynamic_array_len +#define __get_dynamic_array_len(field) \ + ((__entry->__data_loc_##field >> 16) & 0xffff) + +#undef __get_str +#define __get_str(field) ((char *)__get_dynamic_array(field)) + +#undef __get_bitmask +#define __get_bitmask(field) (char *)__get_dynamic_array(field) + +#undef __perf_count +#define __perf_count(c) (c) + +#undef __perf_task +#define __perf_task(t) (t) + +/* cast any integer, pointer, or small struct to u64 */ +#define UINTTYPE(size) \ + __typeof__(__builtin_choose_expr(size == 1, (u8)1, \ + __builtin_choose_expr(size == 2, (u16)2, \ + __builtin_choose_expr(size == 4, (u32)3, \ + __builtin_choose_expr(size == 8, (u64)4, \ + (void)5))))) +#define __CAST_TO_U64(x) ({ \ + typeof(x) __src = (x); \ + UINTTYPE(sizeof(x)) __dst; \ + memcpy(&__dst, &__src, sizeof(__dst)); \ + (u64)__dst; }) + +#define __CAST1(a,...) __CAST_TO_U64(a) +#define __CAST2(a,...) __CAST_TO_U64(a), __CAST1(__VA_ARGS__) +#define __CAST3(a,...) __CAST_TO_U64(a), __CAST2(__VA_ARGS__) +#define __CAST4(a,...) __CAST_TO_U64(a), __CAST3(__VA_ARGS__) +#define __CAST5(a,...) __CAST_TO_U64(a), __CAST4(__VA_ARGS__) +#define __CAST6(a,...) __CAST_TO_U64(a), __CAST5(__VA_ARGS__) +#define __CAST7(a,...) __CAST_TO_U64(a), __CAST6(__VA_ARGS__) +#define __CAST8(a,...) __CAST_TO_U64(a), __CAST7(__VA_ARGS__) +#define __CAST9(a,...) __CAST_TO_U64(a), __CAST8(__VA_ARGS__) +#define __CAST10(a,...) __CAST_TO_U64(a), __CAST9(__VA_ARGS__) +#define __CAST11(a,...) __CAST_TO_U64(a), __CAST10(__VA_ARGS__) +#define __CAST12(a,...) __CAST_TO_U64(a), __CAST11(__VA_ARGS__) +/* tracepoints with more than 12 arguments will hit build error */ +#define CAST_TO_U64(...) CONCATENATE(__CAST, COUNT_ARGS(__VA_ARGS__))(__VA_ARGS__) + +#undef DECLARE_EVENT_CLASS +#define DECLARE_EVENT_CLASS(call, proto, args, tstruct, assign, print) \ +static notrace void \ +__bpf_trace_##call(void *__data, proto) \ +{ \ + struct bpf_prog *prog = __data; \ + CONCATENATE(bpf_trace_run, COUNT_ARGS(args))(prog, CAST_TO_U64(args)); \ +} + +/* + * This part is compiled out, it is only here as a build time check + * to make sure that if the tracepoint handling changes, the + * bpf probe will fail to compile unless it too is updated. + */ +#undef DEFINE_EVENT +#define DEFINE_EVENT(template, call, proto, args) \ +static inline void bpf_test_probe_##call(void) \ +{ \ + check_trace_callback_type_##call(__bpf_trace_##template); \ +} \ +static struct bpf_raw_event_map __used \ + __attribute__((section("__bpf_raw_tp_map"))) \ +__bpf_trace_tp_map_##call = { \ + .tp = &__tracepoint_##call, \ + .bpf_func = (void *)__bpf_trace_##template, \ + .num_args = COUNT_ARGS(args), \ +}; + + +#undef DEFINE_EVENT_PRINT +#define DEFINE_EVENT_PRINT(template, name, proto, args, print) \ + DEFINE_EVENT(template, name, PARAMS(proto), PARAMS(args)) + +#include TRACE_INCLUDE(TRACE_INCLUDE_FILE) +#endif /* CONFIG_BPF_EVENTS */ diff --git a/include/trace/define_trace.h b/include/trace/define_trace.h index d9e3d4aa3f6e..cb30c5532144 100644 --- a/include/trace/define_trace.h +++ b/include/trace/define_trace.h @@ -95,6 +95,7 @@ #ifdef TRACEPOINTS_ENABLED #include #include +#include #endif #undef TRACE_EVENT diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h index 18b7c510c511..1878201c2d77 100644 --- a/include/uapi/linux/bpf.h +++ b/include/uapi/linux/bpf.h @@ -94,6 +94,7 @@ enum bpf_cmd { BPF_MAP_GET_FD_BY_ID, BPF_OBJ_GET_INFO_BY_FD, BPF_PROG_QUERY, + BPF_RAW_TRACEPOINT_OPEN, }; enum bpf_map_type { @@ -134,6 +135,7 @@ enum bpf_prog_type { BPF_PROG_TYPE_SK_SKB, BPF_PROG_TYPE_CGROUP_DEVICE, BPF_PROG_TYPE_SK_MSG, + BPF_PROG_TYPE_RAW_TRACEPOINT, }; enum bpf_attach_type { @@ -344,6 +346,11 @@ union bpf_attr { __aligned_u64 prog_ids; __u32 prog_cnt; } query; + + struct { + __u64 name; + __u32 prog_fd; + } raw_tracepoint; } __attribute__((aligned(8))); /* BPF helper function descriptions: @@ -1152,4 +1159,8 @@ struct bpf_cgroup_dev_ctx { __u32 minor; }; +struct bpf_raw_tracepoint_args { + __u64 args[0]; +}; + #endif /* _UAPI__LINUX_BPF_H__ */ diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c index 77d45bd9f507..95ca2523fa6e 100644 --- a/kernel/bpf/syscall.c +++ b/kernel/bpf/syscall.c @@ -1315,6 +1315,81 @@ static int bpf_obj_get(const union bpf_attr *attr) attr->file_flags); } +struct bpf_raw_tracepoint { + struct bpf_raw_event_map *btp; + struct bpf_prog *prog; +}; + +static int bpf_raw_tracepoint_release(struct inode *inode, struct file *filp) +{ + struct bpf_raw_tracepoint *raw_tp = filp->private_data; + + if (raw_tp->prog) { + bpf_probe_unregister(raw_tp->btp, raw_tp->prog); + bpf_prog_put(raw_tp->prog); + } + kfree(raw_tp); + return 0; +} + +static const struct file_operations bpf_raw_tp_fops = { + .release = bpf_raw_tracepoint_release, + .read = bpf_dummy_read, + .write = bpf_dummy_write, +}; + +#define BPF_RAW_TRACEPOINT_OPEN_LAST_FIELD raw_tracepoint.prog_fd + +static int bpf_raw_tracepoint_open(const union bpf_attr *attr) +{ + struct bpf_raw_tracepoint *raw_tp; + struct bpf_raw_event_map *btp; + struct bpf_prog *prog; + char tp_name[128]; + int tp_fd, err; + + if (strncpy_from_user(tp_name, u64_to_user_ptr(attr->raw_tracepoint.name), + sizeof(tp_name) - 1) < 0) + return -EFAULT; + tp_name[sizeof(tp_name) - 1] = 0; + + btp = bpf_find_raw_tracepoint(tp_name); + if (!btp) + return -ENOENT; + + raw_tp = kzalloc(sizeof(*raw_tp), GFP_USER); + if (!raw_tp) + return -ENOMEM; + raw_tp->btp = btp; + + prog = bpf_prog_get_type(attr->raw_tracepoint.prog_fd, + BPF_PROG_TYPE_RAW_TRACEPOINT); + if (IS_ERR(prog)) { + err = PTR_ERR(prog); + goto out_free_tp; + } + + err = bpf_probe_register(raw_tp->btp, prog); + if (err) + goto out_put_prog; + + raw_tp->prog = prog; + tp_fd = anon_inode_getfd("bpf-raw-tracepoint", &bpf_raw_tp_fops, raw_tp, + O_CLOEXEC); + if (tp_fd < 0) { + bpf_probe_unregister(raw_tp->btp, prog); + err = tp_fd; + goto out_put_prog; + } + return tp_fd; + +out_put_prog: + bpf_prog_put(prog); +out_free_tp: + kfree(raw_tp); + return err; +} + #ifdef CONFIG_CGROUP_BPF #define BPF_PROG_ATTACH_LAST_FIELD attach_flags @@ -1925,6 +2000,9 @@ SYSCALL_DEFINE3(bpf, int, cmd, union bpf_attr __user *, uattr, unsigned int, siz case BPF_OBJ_GET_INFO_BY_FD: err = bpf_obj_get_info_by_fd(&attr, uattr); break; + case BPF_RAW_TRACEPOINT_OPEN: + err = bpf_raw_tracepoint_open(&attr); + break; default: err = -EINVAL; break; diff --git a/kernel/trace/bpf_trace.c b/kernel/trace/bpf_trace.c index 7f9691c86b6e..463e72d18c4c 100644 --- a/kernel/trace/bpf_trace.c +++ b/kernel/trace/bpf_trace.c @@ -735,6 +735,86 @@ static const struct bpf_func_proto *pe_prog_func_proto(enum bpf_func_id func_id) } } +/* + * bpf_raw_tp_regs are separate from bpf_pt_regs used from skb/xdp + * to avoid potential recursive reuse issue when/if tracepoints are added + * inside bpf_*_event_output and/or bpf_get_stack_id + */ +static DEFINE_PER_CPU(struct pt_regs, bpf_raw_tp_regs); +BPF_CALL_5(bpf_perf_event_output_raw_tp, struct bpf_raw_tracepoint_args *, args, + struct bpf_map *, map, u64, flags, void *, data, u64, size) +{ + struct pt_regs *regs = this_cpu_ptr(&bpf_raw_tp_regs); + + perf_fetch_caller_regs(regs); + return ____bpf_perf_event_output(regs, map, flags, data, size); +} + +static const struct bpf_func_proto bpf_perf_event_output_proto_raw_tp = { + .func = bpf_perf_event_output_raw_tp, + .gpl_only = true, + .ret_type = RET_INTEGER, + .arg1_type = ARG_PTR_TO_CTX, + .arg2_type = ARG_CONST_MAP_PTR, + .arg3_type = ARG_ANYTHING, + .arg4_type = ARG_PTR_TO_MEM, + .arg5_type = ARG_CONST_SIZE_OR_ZERO, +}; + +BPF_CALL_3(bpf_get_stackid_raw_tp, struct bpf_raw_tracepoint_args *, args, + struct bpf_map *, map, u64, flags) +{ + struct pt_regs *regs = this_cpu_ptr(&bpf_raw_tp_regs); + + perf_fetch_caller_regs(regs); + /* similar to bpf_perf_event_output_tp, but pt_regs fetched differently */ + return bpf_get_stackid((unsigned long) regs, (unsigned long) map, + flags, 0, 0); +} + +static const struct bpf_func_proto bpf_get_stackid_proto_raw_tp = { + .func = bpf_get_stackid_raw_tp, + .gpl_only = true, + .ret_type = RET_INTEGER, + .arg1_type = ARG_PTR_TO_CTX, + .arg2_type = ARG_CONST_MAP_PTR, + .arg3_type = ARG_ANYTHING, +}; + +static const struct bpf_func_proto *raw_tp_prog_func_proto(enum bpf_func_id func_id) +{ + switch (func_id) { + case BPF_FUNC_perf_event_output: + return &bpf_perf_event_output_proto_raw_tp; + case BPF_FUNC_get_stackid: + return &bpf_get_stackid_proto_raw_tp; + default: + return tracing_func_proto(func_id); + } +} + +static bool raw_tp_prog_is_valid_access(int off, int size, + enum bpf_access_type type, + struct bpf_insn_access_aux *info) +{ + /* largest tracepoint in the kernel has 12 args */ + if (off < 0 || off >= sizeof(__u64) * 12) + return false; + if (type != BPF_READ) + return false; + if (off % size != 0) + return false; + return true; +} + +const struct bpf_verifier_ops raw_tracepoint_verifier_ops = { + .get_func_proto = raw_tp_prog_func_proto, + .is_valid_access = raw_tp_prog_is_valid_access, +}; + +const struct bpf_prog_ops raw_tracepoint_prog_ops = { +}; + static bool pe_prog_is_valid_access(int off, int size, enum bpf_access_type type, struct bpf_insn_access_aux *info) { @@ -908,3 +988,106 @@ int perf_event_query_prog_array(struct perf_event *event, void __user *info) return ret; } + +extern struct bpf_raw_event_map __start__bpf_raw_tp[]; +extern struct bpf_raw_event_map __stop__bpf_raw_tp[]; + +struct bpf_raw_event_map *bpf_find_raw_tracepoint(const char *name) +{ + struct bpf_raw_event_map *btp = __start__bpf_raw_tp; + + for (; btp < __stop__bpf_raw_tp; btp++) { + if (!strcmp(btp->tp->name, name)) + return btp; + } + return NULL; +} + +static __always_inline +void __bpf_trace_run(struct bpf_prog *prog, u64 *args) +{ + rcu_read_lock(); + preempt_disable(); + (void) BPF_PROG_RUN(prog, args); + preempt_enable(); + rcu_read_unlock(); +} + +#define UNPACK(...) __VA_ARGS__ +#define REPEAT_1(FN, DL, X, ...) FN(X) +#define REPEAT_2(FN, DL, X, ...) FN(X) UNPACK DL REPEAT_1(FN, DL, __VA_ARGS__) +#define REPEAT_3(FN, DL, X, ...) FN(X) UNPACK DL REPEAT_2(FN, DL, __VA_ARGS__) +#define REPEAT_4(FN, DL, X, ...) FN(X) UNPACK DL REPEAT_3(FN, DL, __VA_ARGS__) +#define REPEAT_5(FN, DL, X, ...) FN(X) UNPACK DL REPEAT_4(FN, DL, __VA_ARGS__) +#define REPEAT_6(FN, DL, X, ...) FN(X) UNPACK DL REPEAT_5(FN, DL, __VA_ARGS__) +#define REPEAT_7(FN, DL, X, ...) FN(X) UNPACK DL REPEAT_6(FN, DL, __VA_ARGS__) +#define REPEAT_8(FN, DL, X, ...) FN(X) UNPACK DL REPEAT_7(FN, DL, __VA_ARGS__) +#define REPEAT_9(FN, DL, X, ...) FN(X) UNPACK DL REPEAT_8(FN, DL, __VA_ARGS__) +#define REPEAT_10(FN, DL, X, ...) FN(X) UNPACK DL REPEAT_9(FN, DL, __VA_ARGS__) +#define REPEAT_11(FN, DL, X, ...) FN(X) UNPACK DL REPEAT_10(FN, DL, __VA_ARGS__) +#define REPEAT_12(FN, DL, X, ...) FN(X) UNPACK DL REPEAT_11(FN, DL, __VA_ARGS__) +#define REPEAT(X, FN, DL, ...) REPEAT_##X(FN, DL, __VA_ARGS__) + +#define SARG(X) u64 arg##X +#define COPY(X) args[X] = arg##X + +#define __DL_COM (,) +#define __DL_SEM (;) + +#define __SEQ_0_11 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11 + +#define BPF_TRACE_DEFN_x(x) \ + void bpf_trace_run##x(struct bpf_prog *prog, \ + REPEAT(x, SARG, __DL_COM, __SEQ_0_11)) \ + { \ + u64 args[x]; \ + REPEAT(x, COPY, __DL_SEM, __SEQ_0_11); \ + __bpf_trace_run(prog, args); \ + } \ + EXPORT_SYMBOL_GPL(bpf_trace_run##x) +BPF_TRACE_DEFN_x(1); +BPF_TRACE_DEFN_x(2); +BPF_TRACE_DEFN_x(3); +BPF_TRACE_DEFN_x(4); +BPF_TRACE_DEFN_x(5); +BPF_TRACE_DEFN_x(6); +BPF_TRACE_DEFN_x(7); +BPF_TRACE_DEFN_x(8); +BPF_TRACE_DEFN_x(9); +BPF_TRACE_DEFN_x(10); +BPF_TRACE_DEFN_x(11); +BPF_TRACE_DEFN_x(12); + +static int __bpf_probe_register(struct bpf_raw_event_map *btp, struct bpf_prog *prog) +{ + struct tracepoint *tp = btp->tp; + + /* + * check that program doesn't access arguments beyond what's + * available in this tracepoint + */ + if (prog->aux->max_ctx_offset > btp->num_args * sizeof(u64)) + return -EINVAL; + + return tracepoint_probe_register(tp, (void *)btp->bpf_func, prog); +} + +int bpf_probe_register(struct bpf_raw_event_map *btp, struct bpf_prog *prog) +{ + int err; + + mutex_lock(&bpf_event_mutex); + err = __bpf_probe_register(btp, prog); + mutex_unlock(&bpf_event_mutex); + return err; +} + +int bpf_probe_unregister(struct bpf_raw_event_map *btp, struct bpf_prog *prog) +{ + int err; + + mutex_lock(&bpf_event_mutex); + err = tracepoint_probe_unregister(btp->tp, (void *)btp->bpf_func, prog); + mutex_unlock(&bpf_event_mutex); + return err; +} -- cgit v1.2.3 From 09d2bf595db4b4075ea721acd61e180d6bb18f88 Mon Sep 17 00:00:00 2001 From: David Howells Date: Fri, 30 Mar 2018 21:05:28 +0100 Subject: rxrpc: Add a tracepoint to track rxrpc_local refcounting Add a tracepoint to track reference counting on the rxrpc_local struct. Signed-off-by: David Howells --- include/trace/events/rxrpc.h | 43 +++++++++++++++++++++++++++++ net/rxrpc/ar-internal.h | 27 +++--------------- net/rxrpc/call_accept.c | 3 +- net/rxrpc/local_object.c | 65 ++++++++++++++++++++++++++++++++++++++++++-- 4 files changed, 111 insertions(+), 27 deletions(-) (limited to 'include/trace') diff --git a/include/trace/events/rxrpc.h b/include/trace/events/rxrpc.h index 2ea788f6f95d..0410dfeb79c6 100644 --- a/include/trace/events/rxrpc.h +++ b/include/trace/events/rxrpc.h @@ -42,6 +42,14 @@ enum rxrpc_skb_trace { rxrpc_skb_tx_seen, }; +enum rxrpc_local_trace { + rxrpc_local_got, + rxrpc_local_new, + rxrpc_local_processing, + rxrpc_local_put, + rxrpc_local_queued, +}; + enum rxrpc_conn_trace { rxrpc_conn_got, rxrpc_conn_new_client, @@ -215,6 +223,13 @@ enum rxrpc_congest_change { EM(rxrpc_skb_tx_rotated, "Tx ROT") \ E_(rxrpc_skb_tx_seen, "Tx SEE") +#define rxrpc_local_traces \ + EM(rxrpc_local_got, "GOT") \ + EM(rxrpc_local_new, "NEW") \ + EM(rxrpc_local_processing, "PRO") \ + EM(rxrpc_local_put, "PUT") \ + E_(rxrpc_local_queued, "QUE") + #define rxrpc_conn_traces \ EM(rxrpc_conn_got, "GOT") \ EM(rxrpc_conn_new_client, "NWc") \ @@ -416,6 +431,7 @@ enum rxrpc_congest_change { #define E_(a, b) TRACE_DEFINE_ENUM(a); rxrpc_skb_traces; +rxrpc_local_traces; rxrpc_conn_traces; rxrpc_client_traces; rxrpc_call_traces; @@ -439,6 +455,33 @@ rxrpc_congest_changes; #define EM(a, b) { a, b }, #define E_(a, b) { a, b } +TRACE_EVENT(rxrpc_local, + TP_PROTO(struct rxrpc_local *local, enum rxrpc_local_trace op, + int usage, const void *where), + + TP_ARGS(local, op, usage, where), + + TP_STRUCT__entry( + __field(unsigned int, local ) + __field(int, op ) + __field(int, usage ) + __field(const void *, where ) + ), + + TP_fast_assign( + __entry->local = local->debug_id; + __entry->op = op; + __entry->usage = usage; + __entry->where = where; + ), + + TP_printk("L=%08x %s u=%d sp=%pSR", + __entry->local, + __print_symbolic(__entry->op, rxrpc_local_traces), + __entry->usage, + __entry->where) + ); + TRACE_EVENT(rxrpc_conn, TP_PROTO(struct rxrpc_connection *conn, enum rxrpc_conn_trace op, int usage, const void *where), diff --git a/net/rxrpc/ar-internal.h b/net/rxrpc/ar-internal.h index 2a2b0fdfb157..cc51d3eb0548 100644 --- a/net/rxrpc/ar-internal.h +++ b/net/rxrpc/ar-internal.h @@ -981,31 +981,12 @@ extern void rxrpc_process_local_events(struct rxrpc_local *); * local_object.c */ struct rxrpc_local *rxrpc_lookup_local(struct net *, const struct sockaddr_rxrpc *); -void __rxrpc_put_local(struct rxrpc_local *); +struct rxrpc_local *rxrpc_get_local(struct rxrpc_local *); +struct rxrpc_local *rxrpc_get_local_maybe(struct rxrpc_local *); +void rxrpc_put_local(struct rxrpc_local *); +void rxrpc_queue_local(struct rxrpc_local *); void rxrpc_destroy_all_locals(struct rxrpc_net *); -static inline void rxrpc_get_local(struct rxrpc_local *local) -{ - atomic_inc(&local->usage); -} - -static inline -struct rxrpc_local *rxrpc_get_local_maybe(struct rxrpc_local *local) -{ - return atomic_inc_not_zero(&local->usage) ? local : NULL; -} - -static inline void rxrpc_put_local(struct rxrpc_local *local) -{ - if (local && atomic_dec_and_test(&local->usage)) - __rxrpc_put_local(local); -} - -static inline void rxrpc_queue_local(struct rxrpc_local *local) -{ - rxrpc_queue_work(&local->processor); -} - /* * misc.c */ diff --git a/net/rxrpc/call_accept.c b/net/rxrpc/call_accept.c index 493545033e42..5a9b1d916124 100644 --- a/net/rxrpc/call_accept.c +++ b/net/rxrpc/call_accept.c @@ -296,8 +296,7 @@ static struct rxrpc_call *rxrpc_alloc_incoming_call(struct rxrpc_sock *rx, b->conn_backlog[conn_tail] = NULL; smp_store_release(&b->conn_backlog_tail, (conn_tail + 1) & (RXRPC_BACKLOG_MAX - 1)); - rxrpc_get_local(local); - conn->params.local = local; + conn->params.local = rxrpc_get_local(local); conn->params.peer = peer; rxrpc_see_connection(conn); rxrpc_new_incoming_connection(rx, conn, skb); diff --git a/net/rxrpc/local_object.c b/net/rxrpc/local_object.c index 38b99db30e54..8b54e9531d52 100644 --- a/net/rxrpc/local_object.c +++ b/net/rxrpc/local_object.c @@ -95,6 +95,7 @@ static struct rxrpc_local *rxrpc_alloc_local(struct rxrpc_net *rxnet, local->debug_id = atomic_inc_return(&rxrpc_debug_id); memcpy(&local->srx, srx, sizeof(*srx)); local->srx.srx_service = 0; + trace_rxrpc_local(local, rxrpc_local_new, 1, NULL); } _leave(" = %p", local); @@ -256,15 +257,74 @@ addr_in_use: return ERR_PTR(-EADDRINUSE); } +/* + * Get a ref on a local endpoint. + */ +struct rxrpc_local *rxrpc_get_local(struct rxrpc_local *local) +{ + const void *here = __builtin_return_address(0); + int n; + + n = atomic_inc_return(&local->usage); + trace_rxrpc_local(local, rxrpc_local_got, n, here); + return local; +} + +/* + * Get a ref on a local endpoint unless its usage has already reached 0. + */ +struct rxrpc_local *rxrpc_get_local_maybe(struct rxrpc_local *local) +{ + const void *here = __builtin_return_address(0); + + if (local) { + int n = __atomic_add_unless(&local->usage, 1, 0); + if (n > 0) + trace_rxrpc_local(local, rxrpc_local_got, n + 1, here); + else + local = NULL; + } + return local; +} + +/* + * Queue a local endpoint. + */ +void rxrpc_queue_local(struct rxrpc_local *local) +{ + const void *here = __builtin_return_address(0); + + if (rxrpc_queue_work(&local->processor)) + trace_rxrpc_local(local, rxrpc_local_queued, + atomic_read(&local->usage), here); +} + /* * A local endpoint reached its end of life. */ -void __rxrpc_put_local(struct rxrpc_local *local) +static void __rxrpc_put_local(struct rxrpc_local *local) { _enter("%d", local->debug_id); rxrpc_queue_work(&local->processor); } +/* + * Drop a ref on a local endpoint. + */ +void rxrpc_put_local(struct rxrpc_local *local) +{ + const void *here = __builtin_return_address(0); + int n; + + if (local) { + n = atomic_dec_return(&local->usage); + trace_rxrpc_local(local, rxrpc_local_put, n, here); + + if (n == 0) + __rxrpc_put_local(local); + } +} + /* * Destroy a local endpoint's socket and then hand the record to RCU to dispose * of. @@ -322,7 +382,8 @@ static void rxrpc_local_processor(struct work_struct *work) container_of(work, struct rxrpc_local, processor); bool again; - _enter("%d", local->debug_id); + trace_rxrpc_local(local, rxrpc_local_processing, + atomic_read(&local->usage), NULL); do { again = false; -- cgit v1.2.3 From 1159d4b496f57d5b8ee27c8b90b9d01c332e2e11 Mon Sep 17 00:00:00 2001 From: David Howells Date: Fri, 30 Mar 2018 21:05:38 +0100 Subject: rxrpc: Add a tracepoint to track rxrpc_peer refcounting Add a tracepoint to track reference counting on the rxrpc_peer struct. Signed-off-by: David Howells --- include/trace/events/rxrpc.h | 42 ++++++++++++++++++++++++++++ net/rxrpc/ar-internal.h | 23 +++------------- net/rxrpc/peer_event.c | 2 +- net/rxrpc/peer_object.c | 65 ++++++++++++++++++++++++++++++++++++++++++-- 4 files changed, 110 insertions(+), 22 deletions(-) (limited to 'include/trace') diff --git a/include/trace/events/rxrpc.h b/include/trace/events/rxrpc.h index 0410dfeb79c6..9e96c2fe2793 100644 --- a/include/trace/events/rxrpc.h +++ b/include/trace/events/rxrpc.h @@ -50,6 +50,14 @@ enum rxrpc_local_trace { rxrpc_local_queued, }; +enum rxrpc_peer_trace { + rxrpc_peer_got, + rxrpc_peer_new, + rxrpc_peer_processing, + rxrpc_peer_put, + rxrpc_peer_queued_error, +}; + enum rxrpc_conn_trace { rxrpc_conn_got, rxrpc_conn_new_client, @@ -230,6 +238,13 @@ enum rxrpc_congest_change { EM(rxrpc_local_put, "PUT") \ E_(rxrpc_local_queued, "QUE") +#define rxrpc_peer_traces \ + EM(rxrpc_peer_got, "GOT") \ + EM(rxrpc_peer_new, "NEW") \ + EM(rxrpc_peer_processing, "PRO") \ + EM(rxrpc_peer_put, "PUT") \ + E_(rxrpc_peer_queued_error, "QER") + #define rxrpc_conn_traces \ EM(rxrpc_conn_got, "GOT") \ EM(rxrpc_conn_new_client, "NWc") \ @@ -482,6 +497,33 @@ TRACE_EVENT(rxrpc_local, __entry->where) ); +TRACE_EVENT(rxrpc_peer, + TP_PROTO(struct rxrpc_peer *peer, enum rxrpc_peer_trace op, + int usage, const void *where), + + TP_ARGS(peer, op, usage, where), + + TP_STRUCT__entry( + __field(unsigned int, peer ) + __field(int, op ) + __field(int, usage ) + __field(const void *, where ) + ), + + TP_fast_assign( + __entry->peer = peer->debug_id; + __entry->op = op; + __entry->usage = usage; + __entry->where = where; + ), + + TP_printk("P=%08x %s u=%d sp=%pSR", + __entry->peer, + __print_symbolic(__entry->op, rxrpc_peer_traces), + __entry->usage, + __entry->where) + ); + TRACE_EVENT(rxrpc_conn, TP_PROTO(struct rxrpc_connection *conn, enum rxrpc_conn_trace op, int usage, const void *where), diff --git a/net/rxrpc/ar-internal.h b/net/rxrpc/ar-internal.h index d40d54b78567..c46583bc255d 100644 --- a/net/rxrpc/ar-internal.h +++ b/net/rxrpc/ar-internal.h @@ -1041,25 +1041,10 @@ struct rxrpc_peer *rxrpc_lookup_peer(struct rxrpc_local *, struct rxrpc_peer *rxrpc_alloc_peer(struct rxrpc_local *, gfp_t); struct rxrpc_peer *rxrpc_lookup_incoming_peer(struct rxrpc_local *, struct rxrpc_peer *); - -static inline struct rxrpc_peer *rxrpc_get_peer(struct rxrpc_peer *peer) -{ - atomic_inc(&peer->usage); - return peer; -} - -static inline -struct rxrpc_peer *rxrpc_get_peer_maybe(struct rxrpc_peer *peer) -{ - return atomic_inc_not_zero(&peer->usage) ? peer : NULL; -} - -extern void __rxrpc_put_peer(struct rxrpc_peer *peer); -static inline void rxrpc_put_peer(struct rxrpc_peer *peer) -{ - if (peer && atomic_dec_and_test(&peer->usage)) - __rxrpc_put_peer(peer); -} +struct rxrpc_peer *rxrpc_get_peer(struct rxrpc_peer *); +struct rxrpc_peer *rxrpc_get_peer_maybe(struct rxrpc_peer *); +void rxrpc_put_peer(struct rxrpc_peer *); +void __rxrpc_queue_peer_error(struct rxrpc_peer *); /* * proc.c diff --git a/net/rxrpc/peer_event.c b/net/rxrpc/peer_event.c index d01eb9a06448..78c2f95d1f22 100644 --- a/net/rxrpc/peer_event.c +++ b/net/rxrpc/peer_event.c @@ -192,7 +192,7 @@ void rxrpc_error_report(struct sock *sk) rxrpc_free_skb(skb, rxrpc_skb_rx_freed); /* The ref we obtained is passed off to the work item */ - rxrpc_queue_work(&peer->error_distributor); + __rxrpc_queue_peer_error(peer); _leave(""); } diff --git a/net/rxrpc/peer_object.c b/net/rxrpc/peer_object.c index 94a6dbfcf129..a4a750aea1e5 100644 --- a/net/rxrpc/peer_object.c +++ b/net/rxrpc/peer_object.c @@ -386,9 +386,54 @@ struct rxrpc_peer *rxrpc_lookup_peer(struct rxrpc_local *local, } /* - * Discard a ref on a remote peer record. + * Get a ref on a peer record. */ -void __rxrpc_put_peer(struct rxrpc_peer *peer) +struct rxrpc_peer *rxrpc_get_peer(struct rxrpc_peer *peer) +{ + const void *here = __builtin_return_address(0); + int n; + + n = atomic_inc_return(&peer->usage); + trace_rxrpc_peer(peer, rxrpc_peer_got, n, here); + return peer; +} + +/* + * Get a ref on a peer record unless its usage has already reached 0. + */ +struct rxrpc_peer *rxrpc_get_peer_maybe(struct rxrpc_peer *peer) +{ + const void *here = __builtin_return_address(0); + + if (peer) { + int n = __atomic_add_unless(&peer->usage, 1, 0); + if (n > 0) + trace_rxrpc_peer(peer, rxrpc_peer_got, n + 1, here); + else + peer = NULL; + } + return peer; +} + +/* + * Queue a peer record. This passes the caller's ref to the workqueue. + */ +void __rxrpc_queue_peer_error(struct rxrpc_peer *peer) +{ + const void *here = __builtin_return_address(0); + int n; + + n = atomic_read(&peer->usage); + if (rxrpc_queue_work(&peer->error_distributor)) + trace_rxrpc_peer(peer, rxrpc_peer_queued_error, n, here); + else + rxrpc_put_peer(peer); +} + +/* + * Discard a peer record. + */ +static void __rxrpc_put_peer(struct rxrpc_peer *peer) { struct rxrpc_net *rxnet = peer->local->rxnet; @@ -402,6 +447,22 @@ void __rxrpc_put_peer(struct rxrpc_peer *peer) kfree_rcu(peer, rcu); } +/* + * Drop a ref on a peer record. + */ +void rxrpc_put_peer(struct rxrpc_peer *peer) +{ + const void *here = __builtin_return_address(0); + int n; + + if (peer) { + n = atomic_dec_return(&peer->usage); + trace_rxrpc_peer(peer, rxrpc_peer_put, n, here); + if (n == 0) + __rxrpc_put_peer(peer); + } +} + /** * rxrpc_kernel_get_peer - Get the peer address of a call * @sock: The socket on which the call is in progress. -- cgit v1.2.3 From 64ee4e751a1c43b155afe2c1c07212893836f36d Mon Sep 17 00:00:00 2001 From: Qu Wenruo Date: Tue, 12 Dec 2017 15:34:27 +0800 Subject: btrfs: qgroup: Update trace events to use new separate rsv types Signed-off-by: Qu Wenruo Signed-off-by: David Sterba --- fs/btrfs/qgroup.c | 36 +++++++++++++++++++++--------------- include/trace/events/btrfs.h | 17 ++++++++++++----- 2 files changed, 33 insertions(+), 20 deletions(-) (limited to 'include/trace') diff --git a/fs/btrfs/qgroup.c b/fs/btrfs/qgroup.c index 87672d03c8ac..8ec103deb361 100644 --- a/fs/btrfs/qgroup.c +++ b/fs/btrfs/qgroup.c @@ -75,15 +75,19 @@ static const char *qgroup_rsv_type_str(enum btrfs_qgroup_rsv_type type) } #endif -static void qgroup_rsv_add(struct btrfs_qgroup *qgroup, u64 num_bytes, +static void qgroup_rsv_add(struct btrfs_fs_info *fs_info, + struct btrfs_qgroup *qgroup, u64 num_bytes, enum btrfs_qgroup_rsv_type type) { + trace_qgroup_update_reserve(fs_info, qgroup, num_bytes, type); qgroup->rsv.values[type] += num_bytes; } -static void qgroup_rsv_release(struct btrfs_qgroup *qgroup, u64 num_bytes, +static void qgroup_rsv_release(struct btrfs_fs_info *fs_info, + struct btrfs_qgroup *qgroup, u64 num_bytes, enum btrfs_qgroup_rsv_type type) { + trace_qgroup_update_reserve(fs_info, qgroup, -(s64)num_bytes, type); if (qgroup->rsv.values[type] >= num_bytes) { qgroup->rsv.values[type] -= num_bytes; return; @@ -97,22 +101,24 @@ static void qgroup_rsv_release(struct btrfs_qgroup *qgroup, u64 num_bytes, qgroup->rsv.values[type] = 0; } -static void qgroup_rsv_add_by_qgroup(struct btrfs_qgroup *dest, - struct btrfs_qgroup *src) +static void qgroup_rsv_add_by_qgroup(struct btrfs_fs_info *fs_info, + struct btrfs_qgroup *dest, + struct btrfs_qgroup *src) { int i; for (i = 0; i < BTRFS_QGROUP_RSV_LAST; i++) - qgroup_rsv_add(dest, src->rsv.values[i], i); + qgroup_rsv_add(fs_info, dest, src->rsv.values[i], i); } -static void qgroup_rsv_release_by_qgroup(struct btrfs_qgroup *dest, +static void qgroup_rsv_release_by_qgroup(struct btrfs_fs_info *fs_info, + struct btrfs_qgroup *dest, struct btrfs_qgroup *src) { int i; for (i = 0; i < BTRFS_QGROUP_RSV_LAST; i++) - qgroup_rsv_release(dest, src->rsv.values[i], i); + qgroup_rsv_release(fs_info, dest, src->rsv.values[i], i); } static void btrfs_qgroup_update_old_refcnt(struct btrfs_qgroup *qg, u64 seq, @@ -1114,9 +1120,9 @@ static int __qgroup_excl_accounting(struct btrfs_fs_info *fs_info, qgroup->excl_cmpr += sign * num_bytes; if (sign > 0) - qgroup_rsv_add_by_qgroup(qgroup, src); + qgroup_rsv_add_by_qgroup(fs_info, qgroup, src); else - qgroup_rsv_release_by_qgroup(qgroup, src); + qgroup_rsv_release_by_qgroup(fs_info, qgroup, src); qgroup_dirty(fs_info, qgroup); @@ -1137,9 +1143,9 @@ static int __qgroup_excl_accounting(struct btrfs_fs_info *fs_info, WARN_ON(sign < 0 && qgroup->excl < num_bytes); qgroup->excl += sign * num_bytes; if (sign > 0) - qgroup_rsv_add_by_qgroup(qgroup, src); + qgroup_rsv_add_by_qgroup(fs_info, qgroup, src); else - qgroup_rsv_release_by_qgroup(qgroup, src); + qgroup_rsv_release_by_qgroup(fs_info, qgroup, src); qgroup->excl_cmpr += sign * num_bytes; qgroup_dirty(fs_info, qgroup); @@ -2495,8 +2501,8 @@ retry: qg = unode_aux_to_qgroup(unode); - trace_qgroup_update_reserve(fs_info, qg, num_bytes); - qgroup_rsv_add(qg, num_bytes, type); + trace_qgroup_update_reserve(fs_info, qg, num_bytes, type); + qgroup_rsv_add(fs_info, qg, num_bytes, type); } out: @@ -2542,8 +2548,8 @@ void btrfs_qgroup_free_refroot(struct btrfs_fs_info *fs_info, qg = unode_aux_to_qgroup(unode); - trace_qgroup_update_reserve(fs_info, qg, -(s64)num_bytes); - qgroup_rsv_release(qg, num_bytes, type); + trace_qgroup_update_reserve(fs_info, qg, -(s64)num_bytes, type); + qgroup_rsv_release(fs_info, qg, num_bytes, type); list_for_each_entry(glist, &qg->groups, next_group) { ret = ulist_add(fs_info->qgroup_ulist, diff --git a/include/trace/events/btrfs.h b/include/trace/events/btrfs.h index 486771e3f4cb..54b9af822a3a 100644 --- a/include/trace/events/btrfs.h +++ b/include/trace/events/btrfs.h @@ -71,6 +71,11 @@ TRACE_DEFINE_ENUM(COMMIT_TRANS); { BTRFS_FILE_EXTENT_REG, "REG" }, \ { BTRFS_FILE_EXTENT_PREALLOC, "PREALLOC"}) +#define show_qgroup_rsv_type(type) \ + __print_symbolic(type, \ + { BTRFS_QGROUP_RSV_DATA, "DATA" }, \ + { BTRFS_QGROUP_RSV_META, "META" }) + #define BTRFS_GROUP_FLAGS \ { BTRFS_BLOCK_GROUP_DATA, "DATA"}, \ { BTRFS_BLOCK_GROUP_SYSTEM, "SYSTEM"}, \ @@ -1633,24 +1638,26 @@ TRACE_EVENT(qgroup_update_counters, TRACE_EVENT(qgroup_update_reserve, TP_PROTO(struct btrfs_fs_info *fs_info, struct btrfs_qgroup *qgroup, - s64 diff), + s64 diff, int type), - TP_ARGS(fs_info, qgroup, diff), + TP_ARGS(fs_info, qgroup, diff, type), TP_STRUCT__entry_btrfs( __field( u64, qgid ) __field( u64, cur_reserved ) __field( s64, diff ) + __field( int, type ) ), TP_fast_assign_btrfs(fs_info, __entry->qgid = qgroup->qgroupid; - __entry->cur_reserved = qgroup->reserved; + __entry->cur_reserved = qgroup->rsv.values[type]; __entry->diff = diff; ), - TP_printk_btrfs("qgid=%llu cur_reserved=%llu diff=%lld", - __entry->qgid, __entry->cur_reserved, __entry->diff) + TP_printk_btrfs("qgid=%llu type=%s cur_reserved=%llu diff=%lld", + __entry->qgid, show_qgroup_rsv_type(__entry->type), + __entry->cur_reserved, __entry->diff) ); TRACE_EVENT(qgroup_meta_reserve, -- cgit v1.2.3 From 733e03a0b26a463d75aa86083c9fab856571e7fc Mon Sep 17 00:00:00 2001 From: Qu Wenruo Date: Tue, 12 Dec 2017 15:34:29 +0800 Subject: btrfs: qgroup: Split meta rsv type into meta_prealloc and meta_pertrans Btrfs uses 2 different methods to reseve metadata qgroup space. 1) Reserve at btrfs_start_transaction() time This is quite straightforward, caller will use the trans handler allocated to modify b-trees. In this case, reserved metadata should be kept until qgroup numbers are updated. 2) Reserve by using block_rsv first, and later btrfs_join_transaction() This is more complicated, caller will reserve space using block_rsv first, and then later call btrfs_join_transaction() to get a trans handle. In this case, before we modify trees, the reserved space can be modified on demand, and after btrfs_join_transaction(), such reserved space should also be kept until qgroup numbers are updated. Since these two types behave differently, split the original "META" reservation type into 2 sub-types: META_PERTRANS: For above case 1) META_PREALLOC: For reservations that happened before btrfs_join_transaction() of case 2) NOTE: This patch will only convert existing qgroup meta reservation callers according to its situation, not ensuring all callers are at correct timing. Such fix will be added in later patches. Signed-off-by: Qu Wenruo [ update comments ] Signed-off-by: David Sterba --- fs/btrfs/extent-tree.c | 8 ++--- fs/btrfs/qgroup.c | 22 +++++++------- fs/btrfs/qgroup.h | 69 ++++++++++++++++++++++++++++++++++++++++---- fs/btrfs/transaction.c | 8 ++--- include/trace/events/btrfs.h | 5 ++-- 5 files changed, 87 insertions(+), 25 deletions(-) (limited to 'include/trace') diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 0b1f01dd02de..020c1a1a6526 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -5977,7 +5977,7 @@ int btrfs_subvolume_reserve_metadata(struct btrfs_root *root, if (test_bit(BTRFS_FS_QUOTA_ENABLED, &fs_info->flags)) { /* One for parent inode, two for dir entries */ num_bytes = 3 * fs_info->nodesize; - ret = btrfs_qgroup_reserve_meta(root, num_bytes, true); + ret = btrfs_qgroup_reserve_meta_prealloc(root, num_bytes, true); if (ret) return ret; } else { @@ -5996,7 +5996,7 @@ int btrfs_subvolume_reserve_metadata(struct btrfs_root *root, ret = btrfs_block_rsv_migrate(global_rsv, rsv, num_bytes, 1); if (ret && *qgroup_reserved) - btrfs_qgroup_free_meta(root, *qgroup_reserved); + btrfs_qgroup_free_meta_prealloc(root, *qgroup_reserved); return ret; } @@ -6072,7 +6072,7 @@ int btrfs_delalloc_reserve_metadata(struct btrfs_inode *inode, u64 num_bytes) spin_unlock(&inode->lock); if (test_bit(BTRFS_FS_QUOTA_ENABLED, &fs_info->flags)) { - ret = btrfs_qgroup_reserve_meta(root, + ret = btrfs_qgroup_reserve_meta_prealloc(root, nr_extents * fs_info->nodesize, true); if (ret) goto out_fail; @@ -6080,7 +6080,7 @@ int btrfs_delalloc_reserve_metadata(struct btrfs_inode *inode, u64 num_bytes) ret = btrfs_inode_rsv_refill(inode, flush); if (unlikely(ret)) { - btrfs_qgroup_free_meta(root, + btrfs_qgroup_free_meta_prealloc(root, nr_extents * fs_info->nodesize); goto out_fail; } diff --git a/fs/btrfs/qgroup.c b/fs/btrfs/qgroup.c index c0deebfecd93..8831eaa14204 100644 --- a/fs/btrfs/qgroup.c +++ b/fs/btrfs/qgroup.c @@ -69,8 +69,10 @@ static const char *qgroup_rsv_type_str(enum btrfs_qgroup_rsv_type type) { if (type == BTRFS_QGROUP_RSV_DATA) return "data"; - if (type == BTRFS_QGROUP_RSV_META) - return "meta"; + if (type == BTRFS_QGROUP_RSV_META_PERTRANS) + return "meta_pertrans"; + if (type == BTRFS_QGROUP_RSV_META_PREALLOC) + return "meta_prealloc"; return NULL; } #endif @@ -3065,8 +3067,8 @@ int btrfs_qgroup_release_data(struct inode *inode, u64 start, u64 len) return __btrfs_qgroup_release_data(inode, NULL, start, len, 0); } -int btrfs_qgroup_reserve_meta(struct btrfs_root *root, int num_bytes, - bool enforce) +int __btrfs_qgroup_reserve_meta(struct btrfs_root *root, int num_bytes, + enum btrfs_qgroup_rsv_type type, bool enforce) { struct btrfs_fs_info *fs_info = root->fs_info; int ret; @@ -3077,14 +3079,14 @@ int btrfs_qgroup_reserve_meta(struct btrfs_root *root, int num_bytes, BUG_ON(num_bytes != round_down(num_bytes, fs_info->nodesize)); trace_qgroup_meta_reserve(root, (s64)num_bytes); - ret = qgroup_reserve(root, num_bytes, enforce, BTRFS_QGROUP_RSV_META); + ret = qgroup_reserve(root, num_bytes, enforce, type); if (ret < 0) return ret; atomic64_add(num_bytes, &root->qgroup_meta_rsv); return ret; } -void btrfs_qgroup_free_meta_all(struct btrfs_root *root) +void btrfs_qgroup_free_meta_all_pertrans(struct btrfs_root *root) { struct btrfs_fs_info *fs_info = root->fs_info; u64 reserved; @@ -3098,10 +3100,11 @@ void btrfs_qgroup_free_meta_all(struct btrfs_root *root) return; trace_qgroup_meta_reserve(root, -(s64)reserved); btrfs_qgroup_free_refroot(fs_info, root->objectid, reserved, - BTRFS_QGROUP_RSV_META); + BTRFS_QGROUP_RSV_META_PERTRANS); } -void btrfs_qgroup_free_meta(struct btrfs_root *root, int num_bytes) +void __btrfs_qgroup_free_meta(struct btrfs_root *root, int num_bytes, + enum btrfs_qgroup_rsv_type type) { struct btrfs_fs_info *fs_info = root->fs_info; @@ -3113,8 +3116,7 @@ void btrfs_qgroup_free_meta(struct btrfs_root *root, int num_bytes) WARN_ON(atomic64_read(&root->qgroup_meta_rsv) < num_bytes); atomic64_sub(num_bytes, &root->qgroup_meta_rsv); trace_qgroup_meta_reserve(root, -(s64)num_bytes); - btrfs_qgroup_free_refroot(fs_info, root->objectid, num_bytes, - BTRFS_QGROUP_RSV_META); + btrfs_qgroup_free_refroot(fs_info, root->objectid, num_bytes, type); } /* diff --git a/fs/btrfs/qgroup.h b/fs/btrfs/qgroup.h index 279e71a21695..987a5a49deb8 100644 --- a/fs/btrfs/qgroup.h +++ b/fs/btrfs/qgroup.h @@ -61,9 +61,31 @@ struct btrfs_qgroup_extent_record { struct ulist *old_roots; }; +/* + * Qgroup reservation types: + * + * DATA: + * space reserved for data + * + * META_PERTRANS: + * Space reserved for metadata (per-transaction) + * Due to the fact that qgroup data is only updated at transaction commit + * time, reserved space for metadata must be kept until transaction + * commits. + * Any metadata reserved that are used in btrfs_start_transaction() should + * be of this type. + * + * META_PREALLOC: + * There are cases where metadata space is reserved before starting + * transaction, and then btrfs_join_transaction() to get a trans handle. + * Any metadata reserved for such usage should be of this type. + * And after join_transaction() part (or all) of such reservation should + * be converted into META_PERTRANS. + */ enum btrfs_qgroup_rsv_type { BTRFS_QGROUP_RSV_DATA = 0, - BTRFS_QGROUP_RSV_META, + BTRFS_QGROUP_RSV_META_PERTRANS, + BTRFS_QGROUP_RSV_META_PREALLOC, BTRFS_QGROUP_RSV_LAST, }; @@ -269,9 +291,46 @@ int btrfs_qgroup_release_data(struct inode *inode, u64 start, u64 len); int btrfs_qgroup_free_data(struct inode *inode, struct extent_changeset *reserved, u64 start, u64 len); -int btrfs_qgroup_reserve_meta(struct btrfs_root *root, int num_bytes, - bool enforce); -void btrfs_qgroup_free_meta_all(struct btrfs_root *root); -void btrfs_qgroup_free_meta(struct btrfs_root *root, int num_bytes); +int __btrfs_qgroup_reserve_meta(struct btrfs_root *root, int num_bytes, + enum btrfs_qgroup_rsv_type type, bool enforce); +/* Reserve metadata space for pertrans and prealloc type */ +static inline int btrfs_qgroup_reserve_meta_pertrans(struct btrfs_root *root, + int num_bytes, bool enforce) +{ + return __btrfs_qgroup_reserve_meta(root, num_bytes, + BTRFS_QGROUP_RSV_META_PERTRANS, enforce); +} +static inline int btrfs_qgroup_reserve_meta_prealloc(struct btrfs_root *root, + int num_bytes, bool enforce) +{ + return __btrfs_qgroup_reserve_meta(root, num_bytes, + BTRFS_QGROUP_RSV_META_PREALLOC, enforce); +} + +void __btrfs_qgroup_free_meta(struct btrfs_root *root, int num_bytes, + enum btrfs_qgroup_rsv_type type); + +/* Free per-transaction meta reservation for error handling */ +static inline void btrfs_qgroup_free_meta_pertrans(struct btrfs_root *root, + int num_bytes) +{ + __btrfs_qgroup_free_meta(root, num_bytes, + BTRFS_QGROUP_RSV_META_PERTRANS); +} + +/* Pre-allocated meta reservation can be freed at need */ +static inline void btrfs_qgroup_free_meta_prealloc(struct btrfs_root *root, + int num_bytes) +{ + __btrfs_qgroup_free_meta(root, num_bytes, + BTRFS_QGROUP_RSV_META_PREALLOC); +} + +/* + * Per-transaction meta reservation should be all freed at transaction commit + * time + */ +void btrfs_qgroup_free_meta_all_pertrans(struct btrfs_root *root); + void btrfs_qgroup_check_reserved_leak(struct inode *inode); #endif /* __BTRFS_QGROUP__ */ diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index 15f6541303bc..5c4cf0f9146b 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -498,8 +498,8 @@ start_transaction(struct btrfs_root *root, unsigned int num_items, */ if (num_items && root != fs_info->chunk_root) { qgroup_reserved = num_items * fs_info->nodesize; - ret = btrfs_qgroup_reserve_meta(root, qgroup_reserved, - enforce_qgroups); + ret = btrfs_qgroup_reserve_meta_pertrans(root, qgroup_reserved, + enforce_qgroups); if (ret) return ERR_PTR(ret); @@ -596,7 +596,7 @@ alloc_fail: btrfs_block_rsv_release(fs_info, &fs_info->trans_block_rsv, num_bytes); reserve_fail: - btrfs_qgroup_free_meta(root, qgroup_reserved); + btrfs_qgroup_free_meta_pertrans(root, qgroup_reserved); return ERR_PTR(ret); } @@ -1284,7 +1284,7 @@ static noinline int commit_fs_roots(struct btrfs_trans_handle *trans) spin_lock(&fs_info->fs_roots_radix_lock); if (err) break; - btrfs_qgroup_free_meta_all(root); + btrfs_qgroup_free_meta_all_pertrans(root); } } spin_unlock(&fs_info->fs_roots_radix_lock); diff --git a/include/trace/events/btrfs.h b/include/trace/events/btrfs.h index 54b9af822a3a..eee778ba1414 100644 --- a/include/trace/events/btrfs.h +++ b/include/trace/events/btrfs.h @@ -73,8 +73,9 @@ TRACE_DEFINE_ENUM(COMMIT_TRANS); #define show_qgroup_rsv_type(type) \ __print_symbolic(type, \ - { BTRFS_QGROUP_RSV_DATA, "DATA" }, \ - { BTRFS_QGROUP_RSV_META, "META" }) + { BTRFS_QGROUP_RSV_DATA, "DATA" }, \ + { BTRFS_QGROUP_RSV_META_PERTRANS, "META_PERTRANS" }, \ + { BTRFS_QGROUP_RSV_META_PREALLOC, "META_PREALLOC" }) #define BTRFS_GROUP_FLAGS \ { BTRFS_BLOCK_GROUP_DATA, "DATA"}, \ -- cgit v1.2.3 From 4ee0d8832c2ecd08fd4ccbaa55484e6a500f2f34 Mon Sep 17 00:00:00 2001 From: Qu Wenruo Date: Tue, 12 Dec 2017 15:34:35 +0800 Subject: btrfs: qgroup: Update trace events for metadata reservation Now trace_qgroup_meta_reserve() will have extra type parameter. And introduce two new trace events: 1) trace_qgroup_meta_free_all_pertrans() For btrfs_qgroup_free_meta_all_pertrans() 2) trace_qgroup_meta_convert() For btrfs_qgroup_convert_reserved_meta() Signed-off-by: Qu Wenruo Signed-off-by: David Sterba --- fs/btrfs/qgroup.c | 7 +++--- include/trace/events/btrfs.h | 55 ++++++++++++++++++++++++++++++++++++++++++-- 2 files changed, 57 insertions(+), 5 deletions(-) (limited to 'include/trace') diff --git a/fs/btrfs/qgroup.c b/fs/btrfs/qgroup.c index 836819c34c95..92e2c9f15951 100644 --- a/fs/btrfs/qgroup.c +++ b/fs/btrfs/qgroup.c @@ -3138,7 +3138,7 @@ int __btrfs_qgroup_reserve_meta(struct btrfs_root *root, int num_bytes, return 0; BUG_ON(num_bytes != round_down(num_bytes, fs_info->nodesize)); - trace_qgroup_meta_reserve(root, (s64)num_bytes); + trace_qgroup_meta_reserve(root, type, (s64)num_bytes); ret = qgroup_reserve(root, num_bytes, enforce, type); if (ret < 0) return ret; @@ -3163,7 +3163,7 @@ void btrfs_qgroup_free_meta_all_pertrans(struct btrfs_root *root) return; /* TODO: Update trace point to handle such free */ - trace_qgroup_meta_reserve(root, 0); + trace_qgroup_meta_free_all_pertrans(root); /* Special value -1 means to free all reserved space */ btrfs_qgroup_free_refroot(fs_info, root->objectid, (u64)-1, BTRFS_QGROUP_RSV_META_PERTRANS); @@ -3185,7 +3185,7 @@ void __btrfs_qgroup_free_meta(struct btrfs_root *root, int num_bytes, */ num_bytes = sub_root_meta_rsv(root, num_bytes, type); BUG_ON(num_bytes != round_down(num_bytes, fs_info->nodesize)); - trace_qgroup_meta_reserve(root, -(s64)num_bytes); + trace_qgroup_meta_reserve(root, type, -(s64)num_bytes); btrfs_qgroup_free_refroot(fs_info, root->objectid, num_bytes, type); } @@ -3245,6 +3245,7 @@ void btrfs_qgroup_convert_reserved_meta(struct btrfs_root *root, int num_bytes) /* Same as btrfs_qgroup_free_meta_prealloc() */ num_bytes = sub_root_meta_rsv(root, num_bytes, BTRFS_QGROUP_RSV_META_PREALLOC); + trace_qgroup_meta_convert(root, num_bytes); qgroup_convert_meta(fs_info, root->objectid, num_bytes); } diff --git a/include/trace/events/btrfs.h b/include/trace/events/btrfs.h index eee778ba1414..965c650a5273 100644 --- a/include/trace/events/btrfs.h +++ b/include/trace/events/btrfs.h @@ -1663,6 +1663,28 @@ TRACE_EVENT(qgroup_update_reserve, TRACE_EVENT(qgroup_meta_reserve, + TP_PROTO(struct btrfs_root *root, s64 diff, int type), + + TP_ARGS(root, diff, type), + + TP_STRUCT__entry_btrfs( + __field( u64, refroot ) + __field( s64, diff ) + __field( int, type ) + ), + + TP_fast_assign_btrfs(root->fs_info, + __entry->refroot = root->objectid; + __entry->diff = diff; + ), + + TP_printk_btrfs("refroot=%llu(%s) type=%s diff=%lld", + show_root_type(__entry->refroot), + show_qgroup_rsv_type(__entry->type), __entry->diff) +); + +TRACE_EVENT(qgroup_meta_convert, + TP_PROTO(struct btrfs_root *root, s64 diff), TP_ARGS(root, diff), @@ -1670,6 +1692,7 @@ TRACE_EVENT(qgroup_meta_reserve, TP_STRUCT__entry_btrfs( __field( u64, refroot ) __field( s64, diff ) + __field( int, type ) ), TP_fast_assign_btrfs(root->fs_info, @@ -1677,8 +1700,36 @@ TRACE_EVENT(qgroup_meta_reserve, __entry->diff = diff; ), - TP_printk_btrfs("refroot=%llu(%s) diff=%lld", - show_root_type(__entry->refroot), __entry->diff) + TP_printk_btrfs("refroot=%llu(%s) type=%s->%s diff=%lld", + show_root_type(__entry->refroot), + show_qgroup_rsv_type(BTRFS_QGROUP_RSV_META_PREALLOC), + show_qgroup_rsv_type(BTRFS_QGROUP_RSV_META_PERTRANS), + __entry->diff) +); + +TRACE_EVENT(qgroup_meta_free_all_pertrans, + + TP_PROTO(struct btrfs_root *root), + + TP_ARGS(root), + + TP_STRUCT__entry_btrfs( + __field( u64, refroot ) + __field( s64, diff ) + __field( int, type ) + ), + + TP_fast_assign_btrfs(root->fs_info, + __entry->refroot = root->objectid; + spin_lock(&root->qgroup_meta_rsv_lock); + __entry->diff = -(s64)root->qgroup_meta_rsv_pertrans; + spin_unlock(&root->qgroup_meta_rsv_lock); + __entry->type = BTRFS_QGROUP_RSV_META_PERTRANS; + ), + + TP_printk_btrfs("refroot=%llu(%s) type=%s diff=%lld", + show_root_type(__entry->refroot), + show_qgroup_rsv_type(__entry->type), __entry->diff) ); DECLARE_EVENT_CLASS(btrfs__prelim_ref, -- cgit v1.2.3 From 03edb90f57a5f7edaa13db3174e269331ef9a527 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Tue, 27 Mar 2018 10:49:53 -0400 Subject: sunrpc: Update show_svc_xprt_flags() to include recently added flags XPT_KILL_TEMP was added by commit 546125d16142 ("sunrpc: don't call sleeping functions from the notifier block callbacks"), and XPT_CONG_CTRL was added by commit 362142b25843 ("sunrpc: flag transports as having congestion control") . Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- include/trace/events/sunrpc.h | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) (limited to 'include/trace') diff --git a/include/trace/events/sunrpc.h b/include/trace/events/sunrpc.h index 970c91a83173..a23144471b6b 100644 --- a/include/trace/events/sunrpc.h +++ b/include/trace/events/sunrpc.h @@ -591,7 +591,9 @@ DEFINE_EVENT(svc_rqst_status, svc_send, { (1UL << XPT_OLD), "XPT_OLD"}, \ { (1UL << XPT_LISTENER), "XPT_LISTENER"}, \ { (1UL << XPT_CACHE_AUTH), "XPT_CACHE_AUTH"}, \ - { (1UL << XPT_LOCAL), "XPT_LOCAL"}) + { (1UL << XPT_LOCAL), "XPT_LOCAL"}, \ + { (1UL << XPT_KILL_TEMP), "XPT_KILL_TEMP"}, \ + { (1UL << XPT_CONG_CTRL), "XPT_CONG_CTRL"}) TRACE_EVENT(svc_xprt_do_enqueue, TP_PROTO(struct svc_xprt *xprt, struct svc_rqst *rqst), -- cgit v1.2.3 From caa3e106dc623eb41542e6221abecf9956e8a0e6 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Tue, 27 Mar 2018 10:50:07 -0400 Subject: sunrpc: Move trace_svc_xprt_dequeue() Reduce the amount of noise generated by trace_svc_xprt_dequeue by moving it to the end of svc_get_next_xprt. This generates exactly one trace event when a ready xprt is found, rather than spurious events when there is no work to do. The empty events contain no information that can't be obtained simply by tracing function calls to svc_xprt_dequeue. A small additional benefit is simplification of the svc_xprt_event trace class, which no longer has to handle the case when the @xprt parameter is NULL. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- include/trace/events/sunrpc.h | 16 +++++----------- net/sunrpc/svc_xprt.c | 5 +---- 2 files changed, 6 insertions(+), 15 deletions(-) (limited to 'include/trace') diff --git a/include/trace/events/sunrpc.h b/include/trace/events/sunrpc.h index a23144471b6b..9bba3070f873 100644 --- a/include/trace/events/sunrpc.h +++ b/include/trace/events/sunrpc.h @@ -634,24 +634,18 @@ DECLARE_EVENT_CLASS(svc_xprt_event, TP_STRUCT__entry( __field(struct svc_xprt *, xprt) __field(unsigned long, flags) - __dynamic_array(unsigned char, addr, xprt != NULL ? - xprt->xpt_remotelen : 0) + __dynamic_array(unsigned char, addr, xprt->xpt_remotelen) ), TP_fast_assign( __entry->xprt = xprt; - if (xprt) { - memcpy(__get_dynamic_array(addr), - &xprt->xpt_remote, - xprt->xpt_remotelen); - __entry->flags = xprt->xpt_flags; - } else - __entry->flags = 0; + __entry->flags = xprt->xpt_flags; + memcpy(__get_dynamic_array(addr), + &xprt->xpt_remote, xprt->xpt_remotelen); ), TP_printk("xprt=0x%p addr=%pIScp flags=%s", __entry->xprt, - __get_dynamic_array_len(addr) != 0 ? - (struct sockaddr *)__get_dynamic_array(addr) : NULL, + (struct sockaddr *)__get_dynamic_array(addr), show_svc_xprt_flags(__entry->flags)) ); diff --git a/net/sunrpc/svc_xprt.c b/net/sunrpc/svc_xprt.c index 4e3b4c596bae..71f47187b4ec 100644 --- a/net/sunrpc/svc_xprt.c +++ b/net/sunrpc/svc_xprt.c @@ -454,13 +454,9 @@ static struct svc_xprt *svc_xprt_dequeue(struct svc_pool *pool) struct svc_xprt, xpt_ready); list_del_init(&xprt->xpt_ready); svc_xprt_get(xprt); - - dprintk("svc: transport %p dequeued, inuse=%d\n", - xprt, kref_read(&xprt->xpt_ref)); } spin_unlock_bh(&pool->sp_lock); out: - trace_svc_xprt_dequeue(xprt); return xprt; } @@ -734,6 +730,7 @@ out_found: rqstp->rq_chandle.thread_wait = 5*HZ; else rqstp->rq_chandle.thread_wait = 1*HZ; + trace_svc_xprt_dequeue(rqstp->rq_xprt); return rqstp->rq_xprt; } -- cgit v1.2.3 From 41f306d0c287e0cc04054135f9f4ceb003ad6795 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Tue, 27 Mar 2018 10:50:42 -0400 Subject: sunrpc: Simplify trace_svc_recv There doesn't seem to be a lot of value in calling trace_svc_recv in the failing case. 1. There are two very common cases: one is the transport is not ready, and the other is shutdown. Neither is terribly interesting. 2. The trace record for the failing case contains nothing but the status code. Therefore the trace point call site in the error exit is removed. Since the trace point is now recording a length instead of a status, rename the status field and remove the case that records a zero XID. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- include/trace/events/sunrpc.h | 14 +++++++------- net/sunrpc/svc_xprt.c | 1 - 2 files changed, 7 insertions(+), 8 deletions(-) (limited to 'include/trace') diff --git a/include/trace/events/sunrpc.h b/include/trace/events/sunrpc.h index 9bba3070f873..5849bfb3ece2 100644 --- a/include/trace/events/sunrpc.h +++ b/include/trace/events/sunrpc.h @@ -485,28 +485,28 @@ TRACE_EVENT(xs_tcp_data_recv, { (1UL << RQ_BUSY), "RQ_BUSY"}) TRACE_EVENT(svc_recv, - TP_PROTO(struct svc_rqst *rqst, int status), + TP_PROTO(struct svc_rqst *rqst, int len), - TP_ARGS(rqst, status), + TP_ARGS(rqst, len), TP_STRUCT__entry( __field(u32, xid) - __field(int, status) + __field(int, len) __field(unsigned long, flags) __dynamic_array(unsigned char, addr, rqst->rq_addrlen) ), TP_fast_assign( - __entry->xid = status > 0 ? be32_to_cpu(rqst->rq_xid) : 0; - __entry->status = status; + __entry->xid = be32_to_cpu(rqst->rq_xid); + __entry->len = len; __entry->flags = rqst->rq_flags; memcpy(__get_dynamic_array(addr), &rqst->rq_addr, rqst->rq_addrlen); ), - TP_printk("addr=%pIScp xid=0x%08x status=%d flags=%s", + TP_printk("addr=%pIScp xid=0x%08x len=%d flags=%s", (struct sockaddr *)__get_dynamic_array(addr), - __entry->xid, __entry->status, + __entry->xid, __entry->len, show_rqstp_flags(__entry->flags)) ); diff --git a/net/sunrpc/svc_xprt.c b/net/sunrpc/svc_xprt.c index 5fe150c78d0a..47384d0b1673 100644 --- a/net/sunrpc/svc_xprt.c +++ b/net/sunrpc/svc_xprt.c @@ -848,7 +848,6 @@ out_release: rqstp->rq_res.len = 0; svc_xprt_release(rqstp); out: - trace_svc_recv(rqstp, err); return err; } EXPORT_SYMBOL_GPL(svc_recv); -- cgit v1.2.3 From ece200ddd54b9ce840cfee554fb812560c545c7d Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Tue, 27 Mar 2018 10:51:00 -0400 Subject: sunrpc: Save remote presentation address in svc_xprt for trace events TP_printk defines a format string that is passed to user space for converting raw trace event records to something human-readable. My user space's printf (Oracle Linux 7), however, does not have a %pI format specifier. The result is that what is supposed to be an IP address in the output of "trace-cmd report" is just a string that says the field couldn't be displayed. To fix this, adopt the same approach as the client: maintain a pre- formated presentation address for occasions when %pI is not available. The location of the trace_svc_send trace point is adjusted so that rqst->rq_xprt is not NULL when the trace event is recorded. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- include/linux/sunrpc/svc_xprt.h | 4 ++ include/trace/events/sunrpc.h | 89 ++++++++++++-------------------- net/sunrpc/svc_xprt.c | 3 +- net/sunrpc/svcsock.c | 1 + net/sunrpc/xprtrdma/svc_rdma_transport.c | 4 +- 5 files changed, 43 insertions(+), 58 deletions(-) (limited to 'include/trace') diff --git a/include/linux/sunrpc/svc_xprt.h b/include/linux/sunrpc/svc_xprt.h index 19475acb68ea..c3d72066d4b1 100644 --- a/include/linux/sunrpc/svc_xprt.h +++ b/include/linux/sunrpc/svc_xprt.h @@ -83,6 +83,7 @@ struct svc_xprt { size_t xpt_locallen; /* length of address */ struct sockaddr_storage xpt_remote; /* remote peer's address */ size_t xpt_remotelen; /* length of address */ + char xpt_remotebuf[INET6_ADDRSTRLEN + 10]; struct rpc_wait_queue xpt_bc_pending; /* backchannel wait queue */ struct list_head xpt_users; /* callbacks on free */ @@ -152,7 +153,10 @@ static inline void svc_xprt_set_remote(struct svc_xprt *xprt, { memcpy(&xprt->xpt_remote, sa, salen); xprt->xpt_remotelen = salen; + snprintf(xprt->xpt_remotebuf, sizeof(xprt->xpt_remotebuf) - 1, + "%pISpc", sa); } + static inline unsigned short svc_addr_port(const struct sockaddr *sa) { const struct sockaddr_in *sin = (const struct sockaddr_in *)sa; diff --git a/include/trace/events/sunrpc.h b/include/trace/events/sunrpc.h index 5849bfb3ece2..1ec8c4c45766 100644 --- a/include/trace/events/sunrpc.h +++ b/include/trace/events/sunrpc.h @@ -493,20 +493,18 @@ TRACE_EVENT(svc_recv, __field(u32, xid) __field(int, len) __field(unsigned long, flags) - __dynamic_array(unsigned char, addr, rqst->rq_addrlen) + __string(addr, rqst->rq_xprt->xpt_remotebuf) ), TP_fast_assign( __entry->xid = be32_to_cpu(rqst->rq_xid); __entry->len = len; __entry->flags = rqst->rq_flags; - memcpy(__get_dynamic_array(addr), - &rqst->rq_addr, rqst->rq_addrlen); + __assign_str(addr, rqst->rq_xprt->xpt_remotebuf); ), - TP_printk("addr=%pIScp xid=0x%08x len=%d flags=%s", - (struct sockaddr *)__get_dynamic_array(addr), - __entry->xid, __entry->len, + TP_printk("addr=%s xid=0x%08x len=%d flags=%s", + __get_str(addr), __entry->xid, __entry->len, show_rqstp_flags(__entry->flags)) ); @@ -519,20 +517,18 @@ DECLARE_EVENT_CLASS(svc_rqst_event, TP_STRUCT__entry( __field(u32, xid) __field(unsigned long, flags) - __dynamic_array(unsigned char, addr, rqst->rq_addrlen) + __string(addr, rqst->rq_xprt->xpt_remotebuf) ), TP_fast_assign( __entry->xid = be32_to_cpu(rqst->rq_xid); __entry->flags = rqst->rq_flags; - memcpy(__get_dynamic_array(addr), - &rqst->rq_addr, rqst->rq_addrlen); + __assign_str(addr, rqst->rq_xprt->xpt_remotebuf); ), - TP_printk("addr=%pIScp rq_xid=0x%08x flags=%s", - (struct sockaddr *)__get_dynamic_array(addr), - __entry->xid, - show_rqstp_flags(__entry->flags)) + TP_printk("addr=%s xid=0x%08x flags=%s", + __get_str(addr), __entry->xid, + show_rqstp_flags(__entry->flags)) ); DEFINE_EVENT(svc_rqst_event, svc_defer, @@ -553,21 +549,19 @@ DECLARE_EVENT_CLASS(svc_rqst_status, __field(u32, xid) __field(int, status) __field(unsigned long, flags) - __dynamic_array(unsigned char, addr, rqst->rq_addrlen) + __string(addr, rqst->rq_xprt->xpt_remotebuf) ), TP_fast_assign( __entry->xid = be32_to_cpu(rqst->rq_xid); __entry->status = status; __entry->flags = rqst->rq_flags; - memcpy(__get_dynamic_array(addr), - &rqst->rq_addr, rqst->rq_addrlen); + __assign_str(addr, rqst->rq_xprt->xpt_remotebuf); ), - TP_printk("addr=%pIScp rq_xid=0x%08x status=%d flags=%s", - (struct sockaddr *)__get_dynamic_array(addr), - __entry->xid, - __entry->status, show_rqstp_flags(__entry->flags)) + TP_printk("addr=%s xid=0x%08x status=%d flags=%s", + __get_str(addr), __entry->xid, + __entry->status, show_rqstp_flags(__entry->flags)) ); DEFINE_EVENT(svc_rqst_status, svc_process, @@ -604,26 +598,19 @@ TRACE_EVENT(svc_xprt_do_enqueue, __field(struct svc_xprt *, xprt) __field(int, pid) __field(unsigned long, flags) - __dynamic_array(unsigned char, addr, xprt != NULL ? - xprt->xpt_remotelen : 0) + __string(addr, xprt->xpt_remotebuf) ), TP_fast_assign( __entry->xprt = xprt; __entry->pid = rqst? rqst->rq_task->pid : 0; - if (xprt) { - memcpy(__get_dynamic_array(addr), - &xprt->xpt_remote, - xprt->xpt_remotelen); - __entry->flags = xprt->xpt_flags; - } else - __entry->flags = 0; + __entry->flags = xprt->xpt_flags; + __assign_str(addr, xprt->xpt_remotebuf); ), - TP_printk("xprt=0x%p addr=%pIScp pid=%d flags=%s", __entry->xprt, - __get_dynamic_array_len(addr) != 0 ? - (struct sockaddr *)__get_dynamic_array(addr) : NULL, - __entry->pid, show_svc_xprt_flags(__entry->flags)) + TP_printk("xprt=%p addr=%s pid=%d flags=%s", + __entry->xprt, __get_str(addr), + __entry->pid, show_svc_xprt_flags(__entry->flags)) ); DECLARE_EVENT_CLASS(svc_xprt_event, @@ -634,19 +621,18 @@ DECLARE_EVENT_CLASS(svc_xprt_event, TP_STRUCT__entry( __field(struct svc_xprt *, xprt) __field(unsigned long, flags) - __dynamic_array(unsigned char, addr, xprt->xpt_remotelen) + __string(addr, xprt->xpt_remotebuf) ), TP_fast_assign( __entry->xprt = xprt; __entry->flags = xprt->xpt_flags; - memcpy(__get_dynamic_array(addr), - &xprt->xpt_remote, xprt->xpt_remotelen); + __assign_str(addr, xprt->xpt_remotebuf); ), - TP_printk("xprt=0x%p addr=%pIScp flags=%s", __entry->xprt, - (struct sockaddr *)__get_dynamic_array(addr), - show_svc_xprt_flags(__entry->flags)) + TP_printk("xprt=%p addr=%s flags=%s", + __entry->xprt, __get_str(addr), + show_svc_xprt_flags(__entry->flags)) ); DEFINE_EVENT(svc_xprt_event, svc_xprt_dequeue, @@ -682,25 +668,18 @@ TRACE_EVENT(svc_handle_xprt, __field(struct svc_xprt *, xprt) __field(int, len) __field(unsigned long, flags) - __dynamic_array(unsigned char, addr, xprt != NULL ? - xprt->xpt_remotelen : 0) + __string(addr, xprt->xpt_remotebuf) ), TP_fast_assign( __entry->xprt = xprt; __entry->len = len; - if (xprt) { - memcpy(__get_dynamic_array(addr), - &xprt->xpt_remote, - xprt->xpt_remotelen); - __entry->flags = xprt->xpt_flags; - } else - __entry->flags = 0; + __entry->flags = xprt->xpt_flags; + __assign_str(addr, xprt->xpt_remotebuf); ), - TP_printk("xprt=0x%p addr=%pIScp len=%d flags=%s", __entry->xprt, - __get_dynamic_array_len(addr) != 0 ? - (struct sockaddr *)__get_dynamic_array(addr) : NULL, + TP_printk("xprt=%p addr=%s len=%d flags=%s", + __entry->xprt, __get_str(addr), __entry->len, show_svc_xprt_flags(__entry->flags)) ); @@ -712,18 +691,16 @@ DECLARE_EVENT_CLASS(svc_deferred_event, TP_STRUCT__entry( __field(u32, xid) - __dynamic_array(unsigned char, addr, dr->addrlen) + __string(addr, dr->xprt->xpt_remotebuf) ), TP_fast_assign( __entry->xid = be32_to_cpu(*(__be32 *)(dr->args + (dr->xprt_hlen>>2))); - memcpy(__get_dynamic_array(addr), &dr->addr, dr->addrlen); + __assign_str(addr, dr->xprt->xpt_remotebuf); ), - TP_printk("addr=%pIScp xid=0x%08x", - (struct sockaddr *)__get_dynamic_array(addr), - __entry->xid) + TP_printk("addr=%s xid=0x%08x", __get_str(addr), __entry->xid) ); DEFINE_EVENT(svc_deferred_event, svc_drop_deferred, diff --git a/net/sunrpc/svc_xprt.c b/net/sunrpc/svc_xprt.c index 47384d0b1673..f745754a55ea 100644 --- a/net/sunrpc/svc_xprt.c +++ b/net/sunrpc/svc_xprt.c @@ -173,6 +173,7 @@ void svc_xprt_init(struct net *net, struct svc_xprt_class *xcl, set_bit(XPT_BUSY, &xprt->xpt_flags); rpc_init_wait_queue(&xprt->xpt_bc_pending, "xpt_bc_pending"); xprt->xpt_net = get_net(net); + strcpy(xprt->xpt_remotebuf, "uninitialized"); } EXPORT_SYMBOL_GPL(svc_xprt_init); @@ -894,12 +895,12 @@ int svc_send(struct svc_rqst *rqstp) len = xprt->xpt_ops->xpo_sendto(rqstp); mutex_unlock(&xprt->xpt_mutex); rpc_wake_up(&xprt->xpt_bc_pending); + trace_svc_send(rqstp, len); svc_xprt_release(rqstp); if (len == -ECONNREFUSED || len == -ENOTCONN || len == -EAGAIN) len = 0; out: - trace_svc_send(rqstp, len); return len; } diff --git a/net/sunrpc/svcsock.c b/net/sunrpc/svcsock.c index 9b6703588e35..4ca1d92b531a 100644 --- a/net/sunrpc/svcsock.c +++ b/net/sunrpc/svcsock.c @@ -1310,6 +1310,7 @@ static void svc_tcp_init(struct svc_sock *svsk, struct svc_serv *serv) set_bit(XPT_CONG_CTRL, &svsk->sk_xprt.xpt_flags); if (sk->sk_state == TCP_LISTEN) { dprintk("setting up TCP socket for listening\n"); + strcpy(svsk->sk_xprt.xpt_remotebuf, "listener"); set_bit(XPT_LISTENER, &svsk->sk_xprt.xpt_flags); sk->sk_data_ready = svc_tcp_listen_data_ready; set_bit(XPT_CONN, &svsk->sk_xprt.xpt_flags); diff --git a/net/sunrpc/xprtrdma/svc_rdma_transport.c b/net/sunrpc/xprtrdma/svc_rdma_transport.c index 17da06d6b8e5..96cc8f6597d3 100644 --- a/net/sunrpc/xprtrdma/svc_rdma_transport.c +++ b/net/sunrpc/xprtrdma/svc_rdma_transport.c @@ -401,8 +401,10 @@ static struct svcxprt_rdma *rdma_create_xprt(struct svc_serv *serv, */ set_bit(XPT_CONG_CTRL, &cma_xprt->sc_xprt.xpt_flags); - if (listener) + if (listener) { + strcpy(cma_xprt->sc_xprt.xpt_remotebuf, "listener"); set_bit(XPT_LISTENER, &cma_xprt->sc_xprt.xpt_flags); + } return cma_xprt; } -- cgit v1.2.3 From 0b9547bf6b94317b3f8e2496dc2b44cb6e599b01 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Tue, 27 Mar 2018 10:51:22 -0400 Subject: sunrpc: Re-purpose trace_svc_process Currently, trace_svc_process has two call sites: 1. Just after a call to svc_send. svc_send already invokes trace_svc_send with the same arguments just before returning 2. Just before a call to svc_drop. svc_drop already invokes trace_svc_drop with the same arguments just after it is called Therefore trace_svc_process does not provide any additional information not already provided by these other trace points. However, it would be useful to record the incoming RPC procedure. So reuse trace_svc_process for this purpose. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- include/trace/events/sunrpc.h | 30 ++++++++++++++++++++++++++---- net/sunrpc/svc.c | 9 +++------ 2 files changed, 29 insertions(+), 10 deletions(-) (limited to 'include/trace') diff --git a/include/trace/events/sunrpc.h b/include/trace/events/sunrpc.h index 1ec8c4c45766..5a8157c04900 100644 --- a/include/trace/events/sunrpc.h +++ b/include/trace/events/sunrpc.h @@ -508,6 +508,32 @@ TRACE_EVENT(svc_recv, show_rqstp_flags(__entry->flags)) ); +TRACE_EVENT(svc_process, + TP_PROTO(const struct svc_rqst *rqst, const char *name), + + TP_ARGS(rqst, name), + + TP_STRUCT__entry( + __field(u32, xid) + __field(u32, vers) + __field(u32, proc) + __string(service, name) + __string(addr, rqst->rq_xprt->xpt_remotebuf) + ), + + TP_fast_assign( + __entry->xid = be32_to_cpu(rqst->rq_xid); + __entry->vers = rqst->rq_vers; + __entry->proc = rqst->rq_proc; + __assign_str(service, name); + __assign_str(addr, rqst->rq_xprt->xpt_remotebuf); + ), + + TP_printk("addr=%s xid=0x%08x service=%s vers=%u proc=%u", + __get_str(addr), __entry->xid, + __get_str(service), __entry->vers, __entry->proc) +); + DECLARE_EVENT_CLASS(svc_rqst_event, TP_PROTO(struct svc_rqst *rqst), @@ -564,10 +590,6 @@ DECLARE_EVENT_CLASS(svc_rqst_status, __entry->status, show_rqstp_flags(__entry->flags)) ); -DEFINE_EVENT(svc_rqst_status, svc_process, - TP_PROTO(struct svc_rqst *rqst, int status), - TP_ARGS(rqst, status)); - DEFINE_EVENT(svc_rqst_status, svc_send, TP_PROTO(struct svc_rqst *rqst, int status), TP_ARGS(rqst, status)); diff --git a/net/sunrpc/svc.c b/net/sunrpc/svc.c index 387cc4add6f6..f19987f5d3b5 100644 --- a/net/sunrpc/svc.c +++ b/net/sunrpc/svc.c @@ -1255,6 +1255,7 @@ svc_process_common(struct svc_rqst *rqstp, struct kvec *argv, struct kvec *resv) /* Syntactic check complete */ serv->sv_stats->rpccnt++; + trace_svc_process(rqstp, progp->pg_name); /* Build the reply header. */ statp = resv->iov_base +resv->iov_len; @@ -1431,14 +1432,10 @@ svc_process(struct svc_rqst *rqstp) } /* Returns 1 for send, 0 for drop */ - if (likely(svc_process_common(rqstp, argv, resv))) { - int ret = svc_send(rqstp); + if (likely(svc_process_common(rqstp, argv, resv))) + return svc_send(rqstp); - trace_svc_process(rqstp, ret); - return ret; - } out_drop: - trace_svc_process(rqstp, 0); svc_drop(rqstp); return 0; } -- cgit v1.2.3 From aaba72cd4e793fbf1c04e06dee3d2c3710339678 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Tue, 27 Mar 2018 10:51:39 -0400 Subject: sunrpc: Report per-RPC execution stats Introduce a mechanism to report the server-side execution latency of each RPC. The goal is to enable user space to filter the trace record for latency outliers, build histograms, etc. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- include/linux/sunrpc/svc.h | 1 + include/trace/events/sunrpc.h | 21 +++++++++++++++++++++ net/sunrpc/svc_xprt.c | 3 ++- 3 files changed, 24 insertions(+), 1 deletion(-) (limited to 'include/trace') diff --git a/include/linux/sunrpc/svc.h b/include/linux/sunrpc/svc.h index 786ae2255f05..3bd7504066e1 100644 --- a/include/linux/sunrpc/svc.h +++ b/include/linux/sunrpc/svc.h @@ -283,6 +283,7 @@ struct svc_rqst { int rq_reserved; /* space on socket outq * reserved for this request */ + ktime_t rq_stime; /* start time */ struct cache_req rq_chandle; /* handle passed to caches for * request delaying diff --git a/include/trace/events/sunrpc.h b/include/trace/events/sunrpc.h index 5a8157c04900..1b383f71ccd7 100644 --- a/include/trace/events/sunrpc.h +++ b/include/trace/events/sunrpc.h @@ -705,6 +705,27 @@ TRACE_EVENT(svc_handle_xprt, __entry->len, show_svc_xprt_flags(__entry->flags)) ); +TRACE_EVENT(svc_stats_latency, + TP_PROTO(const struct svc_rqst *rqst), + + TP_ARGS(rqst), + + TP_STRUCT__entry( + __field(u32, xid) + __field(unsigned long, execute) + __string(addr, rqst->rq_xprt->xpt_remotebuf) + ), + + TP_fast_assign( + __entry->xid = be32_to_cpu(rqst->rq_xid); + __entry->execute = ktime_to_us(ktime_sub(ktime_get(), + rqst->rq_stime)); + __assign_str(addr, rqst->rq_xprt->xpt_remotebuf); + ), + + TP_printk("addr=%s xid=0x%08x execute-us=%lu", + __get_str(addr), __entry->xid, __entry->execute) +); DECLARE_EVENT_CLASS(svc_deferred_event, TP_PROTO(struct svc_deferred_req *dr), diff --git a/net/sunrpc/svc_xprt.c b/net/sunrpc/svc_xprt.c index f745754a55ea..a7425da14f5b 100644 --- a/net/sunrpc/svc_xprt.c +++ b/net/sunrpc/svc_xprt.c @@ -782,7 +782,7 @@ static int svc_handle_xprt(struct svc_rqst *rqstp, struct svc_xprt *xprt) len = svc_deferred_recv(rqstp); else len = xprt->xpt_ops->xpo_recvfrom(rqstp); - dprintk("svc: got len=%d\n", len); + rqstp->rq_stime = ktime_get(); rqstp->rq_reserved = serv->sv_max_mesg; atomic_add(rqstp->rq_reserved, &xprt->xpt_reserved); } @@ -888,6 +888,7 @@ int svc_send(struct svc_rqst *rqstp) /* Grab mutex to serialize outgoing data. */ mutex_lock(&xprt->xpt_mutex); + trace_svc_stats_latency(rqstp); if (test_bit(XPT_DEAD, &xprt->xpt_flags) || test_bit(XPT_CLOSE, &xprt->xpt_flags)) len = -ENOTCONN; -- cgit v1.2.3 From 55f5088c22cc83dbc64394abfbf76cd1ff5e7cd0 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Tue, 27 Mar 2018 10:52:27 -0400 Subject: svc: Report xprt dequeue latency Record the time between when a rqstp is enqueued on a transport and when it is dequeued. This includes how long the rqstp waits on the queue and how long it takes the kernel scheduler to wake a nfsd thread to service it. The svc_xprt_dequeue trace point is altered to include the number of microseconds between xprt_enqueue and xprt_dequeue. Signed-off-by: Chuck Lever Signed-off-by: J. Bruce Fields --- include/linux/sunrpc/svc.h | 1 + include/trace/events/sunrpc.h | 30 ++++++++++++++++++++++++++---- net/sunrpc/svc_xprt.c | 4 ++-- 3 files changed, 29 insertions(+), 6 deletions(-) (limited to 'include/trace') diff --git a/include/linux/sunrpc/svc.h b/include/linux/sunrpc/svc.h index 3bd7504066e1..dc4c009deec1 100644 --- a/include/linux/sunrpc/svc.h +++ b/include/linux/sunrpc/svc.h @@ -272,6 +272,7 @@ struct svc_rqst { #define RQ_BUSY (6) /* request is busy */ #define RQ_DATA (7) /* request has data */ unsigned long rq_flags; /* flags field */ + ktime_t rq_qtime; /* enqueue time */ void * rq_argp; /* decoded arguments */ void * rq_resp; /* xdr'd results */ diff --git a/include/trace/events/sunrpc.h b/include/trace/events/sunrpc.h index 1b383f71ccd7..922cb8968fb2 100644 --- a/include/trace/events/sunrpc.h +++ b/include/trace/events/sunrpc.h @@ -657,14 +657,36 @@ DECLARE_EVENT_CLASS(svc_xprt_event, show_svc_xprt_flags(__entry->flags)) ); -DEFINE_EVENT(svc_xprt_event, svc_xprt_dequeue, - TP_PROTO(struct svc_xprt *xprt), - TP_ARGS(xprt)); - DEFINE_EVENT(svc_xprt_event, svc_xprt_no_write_space, TP_PROTO(struct svc_xprt *xprt), TP_ARGS(xprt)); +TRACE_EVENT(svc_xprt_dequeue, + TP_PROTO(struct svc_rqst *rqst), + + TP_ARGS(rqst), + + TP_STRUCT__entry( + __field(struct svc_xprt *, xprt) + __field(unsigned long, flags) + __field(unsigned long, wakeup) + __string(addr, rqst->rq_xprt->xpt_remotebuf) + ), + + TP_fast_assign( + __entry->xprt = rqst->rq_xprt; + __entry->flags = rqst->rq_xprt->xpt_flags; + __entry->wakeup = ktime_to_us(ktime_sub(ktime_get(), + rqst->rq_qtime)); + __assign_str(addr, rqst->rq_xprt->xpt_remotebuf); + ), + + TP_printk("xprt=%p addr=%s flags=%s wakeup-us=%lu", + __entry->xprt, __get_str(addr), + show_svc_xprt_flags(__entry->flags), + __entry->wakeup) +); + TRACE_EVENT(svc_wake_up, TP_PROTO(int pid), diff --git a/net/sunrpc/svc_xprt.c b/net/sunrpc/svc_xprt.c index a7425da14f5b..5185efb9027b 100644 --- a/net/sunrpc/svc_xprt.c +++ b/net/sunrpc/svc_xprt.c @@ -409,6 +409,7 @@ void svc_xprt_do_enqueue(struct svc_xprt *xprt) if (test_and_set_bit(RQ_BUSY, &rqstp->rq_flags)) continue; atomic_long_inc(&pool->sp_stats.threads_woken); + rqstp->rq_qtime = ktime_get(); wake_up_process(rqstp->rq_task); goto out_unlock; } @@ -530,7 +531,6 @@ void svc_wake_up(struct svc_serv *serv) if (test_bit(RQ_BUSY, &rqstp->rq_flags)) continue; rcu_read_unlock(); - dprintk("svc: daemon %p woken up.\n", rqstp); wake_up_process(rqstp->rq_task); trace_svc_wake_up(rqstp->rq_task->pid); return; @@ -726,7 +726,7 @@ out_found: rqstp->rq_chandle.thread_wait = 5*HZ; else rqstp->rq_chandle.thread_wait = 1*HZ; - trace_svc_xprt_dequeue(rqstp->rq_xprt); + trace_svc_xprt_dequeue(rqstp); return rqstp->rq_xprt; } -- cgit v1.2.3 From a18feb55769b705a44c4107786c4045eae2e87b6 Mon Sep 17 00:00:00 2001 From: David Howells Date: Wed, 4 Apr 2018 13:41:27 +0100 Subject: fscache: Add tracepoints Add some tracepoints to fscache: (*) fscache_cookie - Tracks a cookie's usage count. (*) fscache_netfs - Logs registration of a network filesystem, including the pointer to the cookie allocated. (*) fscache_acquire - Logs cookie acquisition. (*) fscache_relinquish - Logs cookie relinquishment. (*) fscache_enable - Logs enablement of a cookie. (*) fscache_disable - Logs disablement of a cookie. (*) fscache_osm - Tracks execution of states in the object state machine. and cachefiles: (*) cachefiles_ref - Tracks a cachefiles object's usage count. (*) cachefiles_lookup - Logs result of lookup_one_len(). (*) cachefiles_mkdir - Logs result of vfs_mkdir(). (*) cachefiles_create - Logs result of vfs_create(). (*) cachefiles_unlink - Logs calls to vfs_unlink(). (*) cachefiles_rename - Logs calls to vfs_rename(). (*) cachefiles_mark_active - Logs an object becoming active. (*) cachefiles_wait_active - Logs a wait for an old object to be destroyed. (*) cachefiles_mark_inactive - Logs an object becoming inactive. (*) cachefiles_mark_buried - Logs the burial of an object. Signed-off-by: David Howells --- fs/cachefiles/interface.c | 18 ++- fs/cachefiles/internal.h | 2 + fs/cachefiles/main.c | 1 + fs/cachefiles/namei.c | 42 +++-- fs/fscache/cookie.c | 46 ++++-- fs/fscache/internal.h | 16 +- fs/fscache/main.c | 1 + fs/fscache/netfs.c | 3 +- fs/fscache/object.c | 36 +++-- include/linux/fscache-cache.h | 18 ++- include/trace/events/cachefiles.h | 325 ++++++++++++++++++++++++++++++++++++++ include/trace/events/fscache.h | 277 ++++++++++++++++++++++++++++++++ 12 files changed, 731 insertions(+), 54 deletions(-) create mode 100644 include/trace/events/cachefiles.h create mode 100644 include/trace/events/fscache.h (limited to 'include/trace') diff --git a/fs/cachefiles/interface.c b/fs/cachefiles/interface.c index e7f16a77a22a..405ebc3932c2 100644 --- a/fs/cachefiles/interface.c +++ b/fs/cachefiles/interface.c @@ -177,10 +177,12 @@ static void cachefiles_lookup_complete(struct fscache_object *_object) * increment the usage count on an inode object (may fail if unmounting) */ static -struct fscache_object *cachefiles_grab_object(struct fscache_object *_object) +struct fscache_object *cachefiles_grab_object(struct fscache_object *_object, + enum fscache_obj_ref_trace why) { struct cachefiles_object *object = container_of(_object, struct cachefiles_object, fscache); + int u; _enter("{OBJ%x,%d}", _object->debug_id, atomic_read(&object->usage)); @@ -188,7 +190,9 @@ struct fscache_object *cachefiles_grab_object(struct fscache_object *_object) ASSERT((atomic_read(&object->usage) & 0xffff0000) != 0x6b6b0000); #endif - atomic_inc(&object->usage); + u = atomic_inc_return(&object->usage); + trace_cachefiles_ref(object, _object->cookie, + (enum cachefiles_obj_ref_trace)why, u); return &object->fscache; } @@ -309,10 +313,12 @@ static void cachefiles_drop_object(struct fscache_object *_object) /* * dispose of a reference to an object */ -static void cachefiles_put_object(struct fscache_object *_object) +static void cachefiles_put_object(struct fscache_object *_object, + enum fscache_obj_ref_trace why) { struct cachefiles_object *object; struct fscache_cache *cache; + int u; ASSERT(_object); @@ -328,7 +334,11 @@ static void cachefiles_put_object(struct fscache_object *_object) ASSERTIFCMP(object->fscache.parent, object->fscache.parent->n_children, >, 0); - if (atomic_dec_and_test(&object->usage)) { + u = atomic_dec_return(&object->usage); + trace_cachefiles_ref(object, _object->cookie, + (enum cachefiles_obj_ref_trace)why, u); + ASSERTCMP(u, !=, -1); + if (u == 0) { _debug("- kill object OBJ%x", object->fscache.debug_id); ASSERT(!test_bit(CACHEFILES_OBJECT_ACTIVE, &object->flags)); diff --git a/fs/cachefiles/internal.h b/fs/cachefiles/internal.h index bb3a02ca9da4..d2f6f996e65a 100644 --- a/fs/cachefiles/internal.h +++ b/fs/cachefiles/internal.h @@ -124,6 +124,8 @@ struct cachefiles_xattr { uint8_t data[]; }; +#include + /* * note change of state for daemon */ diff --git a/fs/cachefiles/main.c b/fs/cachefiles/main.c index 711f13d8c2de..f54d3f5b2e40 100644 --- a/fs/cachefiles/main.c +++ b/fs/cachefiles/main.c @@ -22,6 +22,7 @@ #include #include #include +#define CREATE_TRACE_POINTS #include "internal.h" unsigned cachefiles_debug; diff --git a/fs/cachefiles/namei.c b/fs/cachefiles/namei.c index 3978b324cbca..5fc214256316 100644 --- a/fs/cachefiles/namei.c +++ b/fs/cachefiles/namei.c @@ -120,6 +120,7 @@ static void cachefiles_mark_object_buried(struct cachefiles_cache *cache, } write_unlock(&cache->active_lock); + trace_cachefiles_mark_buried(NULL, dentry, why); _leave(" [no owner]"); return; @@ -130,6 +131,8 @@ found_dentry: object->fscache.state->name, dentry); + trace_cachefiles_mark_buried(object, dentry, why); + if (fscache_object_is_live(&object->fscache)) { pr_err("\n"); pr_err("Error: Can't preemptively bury live object\n"); @@ -158,13 +161,15 @@ static int cachefiles_mark_object_active(struct cachefiles_cache *cache, try_again: write_lock(&cache->active_lock); + dentry = object->dentry; + trace_cachefiles_mark_active(object, dentry); + if (test_and_set_bit(CACHEFILES_OBJECT_ACTIVE, &object->flags)) { pr_err("Error: Object already active\n"); cachefiles_printk_object(object, NULL); BUG(); } - dentry = object->dentry; _p = &cache->active_nodes.rb_node; while (*_p) { _parent = *_p; @@ -191,6 +196,8 @@ try_again: /* an old object from a previous incarnation is hogging the slot - we * need to wait for it to be destroyed */ wait_for_old_object: + trace_cachefiles_wait_active(object, dentry, xobject); + if (fscache_object_is_live(&xobject->fscache)) { pr_err("\n"); pr_err("Error: Unexpected object collision\n"); @@ -248,12 +255,12 @@ wait_for_old_object: ASSERT(!test_bit(CACHEFILES_OBJECT_ACTIVE, &xobject->flags)); - cache->cache.ops->put_object(&xobject->fscache); + cache->cache.ops->put_object(&xobject->fscache, cachefiles_obj_put_wait_retry); goto try_again; requeue: clear_bit(CACHEFILES_OBJECT_ACTIVE, &object->flags); - cache->cache.ops->put_object(&xobject->fscache); + cache->cache.ops->put_object(&xobject->fscache, cachefiles_obj_put_wait_timeo); _leave(" = -ETIMEDOUT"); return -ETIMEDOUT; } @@ -265,6 +272,11 @@ void cachefiles_mark_object_inactive(struct cachefiles_cache *cache, struct cachefiles_object *object, blkcnt_t i_blocks) { + struct dentry *dentry = object->dentry; + struct inode *inode = d_backing_inode(dentry); + + trace_cachefiles_mark_inactive(object, dentry, inode); + write_lock(&cache->active_lock); rb_erase(&object->active_node, &cache->active_nodes); clear_bit(CACHEFILES_OBJECT_ACTIVE, &object->flags); @@ -288,6 +300,7 @@ void cachefiles_mark_object_inactive(struct cachefiles_cache *cache, * - unlocks the directory mutex */ static int cachefiles_bury_object(struct cachefiles_cache *cache, + struct cachefiles_object *object, struct dentry *dir, struct dentry *rep, bool preemptive, @@ -312,6 +325,7 @@ static int cachefiles_bury_object(struct cachefiles_cache *cache, if (ret < 0) { cachefiles_io_error(cache, "Unlink security error"); } else { + trace_cachefiles_unlink(object, rep, why); ret = vfs_unlink(d_inode(dir), rep, NULL); if (preemptive) @@ -413,6 +427,7 @@ try_again: if (ret < 0) { cachefiles_io_error(cache, "Rename security error %d", ret); } else { + trace_cachefiles_rename(object, rep, grave, why); ret = vfs_rename(d_inode(dir), rep, d_inode(cache->graveyard), grave, NULL, 0); if (ret != 0 && ret != -ENOMEM) @@ -458,7 +473,7 @@ int cachefiles_delete_object(struct cachefiles_cache *cache, /* we need to check that our parent is _still_ our parent - it * may have been renamed */ if (dir == object->dentry->d_parent) { - ret = cachefiles_bury_object(cache, dir, + ret = cachefiles_bury_object(cache, object, dir, object->dentry, false, FSCACHE_OBJECT_WAS_RETIRED); } else { @@ -486,6 +501,7 @@ int cachefiles_walk_to_object(struct cachefiles_object *parent, { struct cachefiles_cache *cache; struct dentry *dir, *next = NULL; + struct inode *inode; struct path path; unsigned long start; const char *name; @@ -529,13 +545,17 @@ lookup_again: start = jiffies; next = lookup_one_len(name, dir, nlen); cachefiles_hist(cachefiles_lookup_histogram, start); - if (IS_ERR(next)) + if (IS_ERR(next)) { + trace_cachefiles_lookup(object, next, NULL); goto lookup_error; + } - _debug("next -> %p %s", next, d_backing_inode(next) ? "positive" : "negative"); + inode = d_backing_inode(next); + trace_cachefiles_lookup(object, next, inode); + _debug("next -> %p %s", next, inode ? "positive" : "negative"); if (!key) - object->new = !d_backing_inode(next); + object->new = !inode; /* if this element of the path doesn't exist, then the lookup phase * failed, and we can release any readers in the certain knowledge that @@ -558,6 +578,8 @@ lookup_again: start = jiffies; ret = vfs_mkdir(d_inode(dir), next, 0); cachefiles_hist(cachefiles_mkdir_histogram, start); + if (!key) + trace_cachefiles_mkdir(object, next, ret); if (ret < 0) goto create_error; @@ -587,6 +609,7 @@ lookup_again: start = jiffies; ret = vfs_create(d_inode(dir), next, S_IFREG, true); cachefiles_hist(cachefiles_create_histogram, start); + trace_cachefiles_create(object, next, ret); if (ret < 0) goto create_error; @@ -629,7 +652,8 @@ lookup_again: * mutex) */ object->dentry = NULL; - ret = cachefiles_bury_object(cache, dir, next, true, + ret = cachefiles_bury_object(cache, object, dir, next, + true, FSCACHE_OBJECT_IS_STALE); dput(next); next = NULL; @@ -955,7 +979,7 @@ int cachefiles_cull(struct cachefiles_cache *cache, struct dentry *dir, /* actually remove the victim (drops the dir mutex) */ _debug("bury"); - ret = cachefiles_bury_object(cache, dir, victim, false, + ret = cachefiles_bury_object(cache, NULL, dir, victim, false, FSCACHE_OBJECT_WAS_CULLED); if (ret < 0) goto error; diff --git a/fs/fscache/cookie.c b/fs/fscache/cookie.c index 98d22f495cd8..20bc3341f113 100644 --- a/fs/fscache/cookie.c +++ b/fs/fscache/cookie.c @@ -101,7 +101,7 @@ struct fscache_cookie *__fscache_acquire_cookie( */ atomic_set(&cookie->n_active, 1); - atomic_inc(&parent->usage); + fscache_cookie_get(parent, fscache_cookie_get_acquire_parent); atomic_inc(&parent->n_children); cookie->def = def; @@ -125,6 +125,8 @@ struct fscache_cookie *__fscache_acquire_cookie( break; } + trace_fscache_acquire(cookie); + if (enable) { /* if the object is an index then we need do nothing more here * - we create indices on disk when we need them as an index @@ -134,7 +136,8 @@ struct fscache_cookie *__fscache_acquire_cookie( set_bit(FSCACHE_COOKIE_ENABLED, &cookie->flags); } else { atomic_dec(&parent->n_children); - __fscache_cookie_put(cookie); + fscache_cookie_put(cookie, + fscache_cookie_put_acquire_nobufs); fscache_stat(&fscache_n_acquires_nobufs); _leave(" = NULL"); return NULL; @@ -159,6 +162,8 @@ void __fscache_enable_cookie(struct fscache_cookie *cookie, { _enter("%p", cookie); + trace_fscache_enable(cookie); + wait_on_bit_lock(&cookie->flags, FSCACHE_COOKIE_ENABLEMENT_LOCK, TASK_UNINTERRUPTIBLE); @@ -318,7 +323,7 @@ static int fscache_alloc_object(struct fscache_cache *cache, * attached to the cookie */ if (fscache_attach_object(cookie, object) < 0) { fscache_stat(&fscache_n_cop_put_object); - cache->ops->put_object(object); + cache->ops->put_object(object, fscache_obj_put_attach_fail); fscache_stat_d(&fscache_n_cop_put_object); } @@ -338,7 +343,7 @@ object_already_extant: error_put: fscache_stat(&fscache_n_cop_put_object); - cache->ops->put_object(object); + cache->ops->put_object(object, fscache_obj_put_alloc_fail); fscache_stat_d(&fscache_n_cop_put_object); error: _leave(" = %d", ret); @@ -398,7 +403,7 @@ static int fscache_attach_object(struct fscache_cookie *cookie, /* attach to the cookie */ object->cookie = cookie; - atomic_inc(&cookie->usage); + fscache_cookie_get(cookie, fscache_cookie_get_attach_object); hlist_add_head(&object->cookie_link, &cookie->backing_objects); fscache_objlist_add(object); @@ -516,6 +521,8 @@ void __fscache_disable_cookie(struct fscache_cookie *cookie, bool invalidate) _enter("%p,%u", cookie, invalidate); + trace_fscache_disable(cookie); + ASSERTCMP(atomic_read(&cookie->n_active), >, 0); if (atomic_read(&cookie->n_children) != 0) { @@ -601,6 +608,8 @@ void __fscache_relinquish_cookie(struct fscache_cookie *cookie, bool retire) cookie, cookie->def->name, cookie->netfs_data, atomic_read(&cookie->n_active), retire); + trace_fscache_relinquish(cookie, retire); + /* No further netfs-accessing operations on this cookie permitted */ if (test_and_set_bit(FSCACHE_COOKIE_RELINQUISHED, &cookie->flags)) BUG(); @@ -620,35 +629,38 @@ void __fscache_relinquish_cookie(struct fscache_cookie *cookie, bool retire) /* Dispose of the netfs's link to the cookie */ ASSERTCMP(atomic_read(&cookie->usage), >, 0); - fscache_cookie_put(cookie); + fscache_cookie_put(cookie, fscache_cookie_put_relinquish); _leave(""); } EXPORT_SYMBOL(__fscache_relinquish_cookie); /* - * destroy a cookie + * Drop a reference to a cookie. */ -void __fscache_cookie_put(struct fscache_cookie *cookie) +void fscache_cookie_put(struct fscache_cookie *cookie, + enum fscache_cookie_trace where) { struct fscache_cookie *parent; + int usage; _enter("%p", cookie); - for (;;) { - _debug("FREE COOKIE %p", cookie); + do { + usage = atomic_dec_return(&cookie->usage); + trace_fscache_cookie(cookie, where, usage); + + if (usage > 0) + return; + BUG_ON(usage < 0); + parent = cookie->parent; BUG_ON(!hlist_empty(&cookie->backing_objects)); kmem_cache_free(fscache_cookie_jar, cookie); - if (!parent) - break; - cookie = parent; - BUG_ON(atomic_read(&cookie->usage) <= 0); - if (!atomic_dec_and_test(&cookie->usage)) - break; - } + where = fscache_cookie_put_parent; + } while (cookie); _leave(""); } diff --git a/fs/fscache/internal.h b/fs/fscache/internal.h index 0ff4b49a0037..c27e2db3004e 100644 --- a/fs/fscache/internal.h +++ b/fs/fscache/internal.h @@ -29,6 +29,7 @@ #define pr_fmt(fmt) "FS-Cache: " fmt #include +#include #include #define FSCACHE_MIN_THREADS 4 @@ -49,7 +50,8 @@ extern struct fscache_cache *fscache_select_cache_for_object( extern struct kmem_cache *fscache_cookie_jar; extern void fscache_cookie_init_once(void *); -extern void __fscache_cookie_put(struct fscache_cookie *); +extern void fscache_cookie_put(struct fscache_cookie *, + enum fscache_cookie_trace); /* * fsdef.c @@ -311,14 +313,12 @@ static inline void fscache_raise_event(struct fscache_object *object, fscache_enqueue_object(object); } -/* - * drop a reference to a cookie - */ -static inline void fscache_cookie_put(struct fscache_cookie *cookie) +static inline void fscache_cookie_get(struct fscache_cookie *cookie, + enum fscache_cookie_trace where) { - BUG_ON(atomic_read(&cookie->usage) <= 0); - if (atomic_dec_and_test(&cookie->usage)) - __fscache_cookie_put(cookie); + int usage = atomic_inc_return(&cookie->usage); + + trace_fscache_cookie(cookie, where, usage); } /* diff --git a/fs/fscache/main.c b/fs/fscache/main.c index 249968dcbf5c..7dce110bf17d 100644 --- a/fs/fscache/main.c +++ b/fs/fscache/main.c @@ -16,6 +16,7 @@ #include #include #include +#define CREATE_TRACE_POINTS #include "internal.h" MODULE_DESCRIPTION("FS Cache Manager"); diff --git a/fs/fscache/netfs.c b/fs/fscache/netfs.c index a8aa00be4444..c816600d1dde 100644 --- a/fs/fscache/netfs.c +++ b/fs/fscache/netfs.c @@ -60,7 +60,7 @@ int __fscache_register_netfs(struct fscache_netfs *netfs) goto already_registered; } - atomic_inc(&cookie->parent->usage); + fscache_cookie_get(cookie->parent, fscache_cookie_get_register_netfs); atomic_inc(&cookie->parent->n_children); netfs->primary_index = cookie; @@ -68,6 +68,7 @@ int __fscache_register_netfs(struct fscache_netfs *netfs) ret = 0; pr_notice("Netfs '%s' registered for caching\n", netfs->name); + trace_fscache_netfs(netfs); already_registered: up_write(&fscache_addremove_sem); diff --git a/fs/fscache/object.c b/fs/fscache/object.c index 7a182c87f378..99afe64352a5 100644 --- a/fs/fscache/object.c +++ b/fs/fscache/object.c @@ -138,8 +138,10 @@ static const struct fscache_transition fscache_osm_run_oob[] = { { 0, NULL } }; -static int fscache_get_object(struct fscache_object *); -static void fscache_put_object(struct fscache_object *); +static int fscache_get_object(struct fscache_object *, + enum fscache_obj_ref_trace); +static void fscache_put_object(struct fscache_object *, + enum fscache_obj_ref_trace); static bool fscache_enqueue_dependents(struct fscache_object *, int); static void fscache_dequeue_object(struct fscache_object *); @@ -170,6 +172,7 @@ static void fscache_object_sm_dispatcher(struct fscache_object *object) const struct fscache_transition *t; const struct fscache_state *state, *new_state; unsigned long events, event_mask; + bool oob; int event = -1; ASSERT(object != NULL); @@ -188,6 +191,7 @@ restart_masked: if (events & object->oob_event_mask) { _debug("{OBJ%x} oob %lx", object->debug_id, events & object->oob_event_mask); + oob = true; for (t = object->oob_table; t->events; t++) { if (events & t->events) { state = t->transit_to; @@ -199,6 +203,7 @@ restart_masked: } } } + oob = false; /* Wait states are just transition tables */ if (!state->work) { @@ -207,6 +212,8 @@ restart_masked: if (events & t->events) { new_state = t->transit_to; event = fls(events & t->events) - 1; + trace_fscache_osm(object, state, + true, false, event); clear_bit(event, &object->events); _debug("{OBJ%x} ev %d: %s -> %s", object->debug_id, event, @@ -226,6 +233,7 @@ restart_masked: execute_work_state: _debug("{OBJ%x} exec %s", object->debug_id, state->name); + trace_fscache_osm(object, state, false, oob, event); new_state = state->work(object, event); event = -1; if (new_state == NO_TRANSIT) { @@ -279,7 +287,7 @@ static void fscache_object_work_func(struct work_struct *work) start = jiffies; fscache_object_sm_dispatcher(object); fscache_hist(fscache_objs_histogram, start); - fscache_put_object(object); + fscache_put_object(object, fscache_obj_put_work); } /** @@ -397,7 +405,7 @@ static const struct fscache_state *fscache_initialise_object(struct fscache_obje fscache_stat(&fscache_n_cop_grab_object); success = false; if (fscache_object_is_live(parent) && - object->cache->ops->grab_object(object)) { + object->cache->ops->grab_object(object, fscache_obj_get_add_to_deps)) { list_add(&object->dep_link, &parent->dependents); success = true; } @@ -745,7 +753,7 @@ static const struct fscache_state *fscache_drop_object(struct fscache_object *ob } /* this just shifts the object release to the work processor */ - fscache_put_object(object); + fscache_put_object(object, fscache_obj_put_drop_obj); fscache_stat(&fscache_n_object_dead); _leave(""); @@ -755,12 +763,13 @@ static const struct fscache_state *fscache_drop_object(struct fscache_object *ob /* * get a ref on an object */ -static int fscache_get_object(struct fscache_object *object) +static int fscache_get_object(struct fscache_object *object, + enum fscache_obj_ref_trace why) { int ret; fscache_stat(&fscache_n_cop_grab_object); - ret = object->cache->ops->grab_object(object) ? 0 : -EAGAIN; + ret = object->cache->ops->grab_object(object, why) ? 0 : -EAGAIN; fscache_stat_d(&fscache_n_cop_grab_object); return ret; } @@ -768,10 +777,11 @@ static int fscache_get_object(struct fscache_object *object) /* * Discard a ref on an object */ -static void fscache_put_object(struct fscache_object *object) +static void fscache_put_object(struct fscache_object *object, + enum fscache_obj_ref_trace why) { fscache_stat(&fscache_n_cop_put_object); - object->cache->ops->put_object(object); + object->cache->ops->put_object(object, why); fscache_stat_d(&fscache_n_cop_put_object); } @@ -786,7 +796,7 @@ void fscache_object_destroy(struct fscache_object *object) fscache_objlist_remove(object); /* We can get rid of the cookie now */ - fscache_cookie_put(object->cookie); + fscache_cookie_put(object->cookie, fscache_cookie_put_object); object->cookie = NULL; } EXPORT_SYMBOL(fscache_object_destroy); @@ -798,7 +808,7 @@ void fscache_enqueue_object(struct fscache_object *object) { _enter("{OBJ%x}", object->debug_id); - if (fscache_get_object(object) >= 0) { + if (fscache_get_object(object, fscache_obj_get_queue) >= 0) { wait_queue_head_t *cong_wq = &get_cpu_var(fscache_object_cong_wait); @@ -806,7 +816,7 @@ void fscache_enqueue_object(struct fscache_object *object) if (fscache_object_congested()) wake_up(cong_wq); } else - fscache_put_object(object); + fscache_put_object(object, fscache_obj_put_queue); put_cpu_var(fscache_object_cong_wait); } @@ -866,7 +876,7 @@ static bool fscache_enqueue_dependents(struct fscache_object *object, int event) list_del_init(&dep->dep_link); fscache_raise_event(dep, event); - fscache_put_object(dep); + fscache_put_object(dep, fscache_obj_put_enq_dep); if (!list_empty(&object->dependents) && need_resched()) { ret = false; diff --git a/include/linux/fscache-cache.h b/include/linux/fscache-cache.h index b19fa8592fc2..fbe102f37074 100644 --- a/include/linux/fscache-cache.h +++ b/include/linux/fscache-cache.h @@ -29,6 +29,18 @@ struct fscache_cache_ops; struct fscache_object; struct fscache_operation; +enum fscache_obj_ref_trace { + fscache_obj_get_add_to_deps, + fscache_obj_get_queue, + fscache_obj_put_alloc_fail, + fscache_obj_put_attach_fail, + fscache_obj_put_drop_obj, + fscache_obj_put_enq_dep, + fscache_obj_put_queue, + fscache_obj_put_work, + fscache_obj_ref__nr_traces +}; + /* * cache tag definition */ @@ -231,7 +243,8 @@ struct fscache_cache_ops { void (*lookup_complete)(struct fscache_object *object); /* increment the usage count on this object (may fail if unmounting) */ - struct fscache_object *(*grab_object)(struct fscache_object *object); + struct fscache_object *(*grab_object)(struct fscache_object *object, + enum fscache_obj_ref_trace why); /* pin an object in the cache */ int (*pin_object)(struct fscache_object *object); @@ -254,7 +267,8 @@ struct fscache_cache_ops { void (*drop_object)(struct fscache_object *object); /* dispose of a reference to an object */ - void (*put_object)(struct fscache_object *object); + void (*put_object)(struct fscache_object *object, + enum fscache_obj_ref_trace why); /* sync a cache */ void (*sync_cache)(struct fscache_cache *cache); diff --git a/include/trace/events/cachefiles.h b/include/trace/events/cachefiles.h new file mode 100644 index 000000000000..aa86e7dba511 --- /dev/null +++ b/include/trace/events/cachefiles.h @@ -0,0 +1,325 @@ +/* CacheFiles tracepoints + * + * Copyright (C) 2016 Red Hat, Inc. All Rights Reserved. + * Written by David Howells (dhowells@redhat.com) + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public Licence + * as published by the Free Software Foundation; either version + * 2 of the Licence, or (at your option) any later version. + */ +#undef TRACE_SYSTEM +#define TRACE_SYSTEM cachefiles + +#if !defined(_TRACE_CACHEFILES_H) || defined(TRACE_HEADER_MULTI_READ) +#define _TRACE_CACHEFILES_H + +#include + +/* + * Define enums for tracing information. + */ +#ifndef __CACHEFILES_DECLARE_TRACE_ENUMS_ONCE_ONLY +#define __CACHEFILES_DECLARE_TRACE_ENUMS_ONCE_ONLY + +enum cachefiles_obj_ref_trace { + cachefiles_obj_put_wait_retry = fscache_obj_ref__nr_traces, + cachefiles_obj_put_wait_timeo, + cachefiles_obj_ref__nr_traces +}; + +#endif + +/* + * Define enum -> string mappings for display. + */ +#define cachefiles_obj_kill_traces \ + EM(FSCACHE_OBJECT_IS_STALE, "stale") \ + EM(FSCACHE_OBJECT_NO_SPACE, "no_space") \ + EM(FSCACHE_OBJECT_WAS_RETIRED, "was_retired") \ + E_(FSCACHE_OBJECT_WAS_CULLED, "was_culled") + +#define cachefiles_obj_ref_traces \ + EM(fscache_obj_get_add_to_deps, "GET add_to_deps") \ + EM(fscache_obj_get_queue, "GET queue") \ + EM(fscache_obj_put_alloc_fail, "PUT alloc_fail") \ + EM(fscache_obj_put_attach_fail, "PUT attach_fail") \ + EM(fscache_obj_put_drop_obj, "PUT drop_obj") \ + EM(fscache_obj_put_enq_dep, "PUT enq_dep") \ + EM(fscache_obj_put_queue, "PUT queue") \ + EM(fscache_obj_put_work, "PUT work") \ + EM(cachefiles_obj_put_wait_retry, "PUT wait_retry") \ + E_(cachefiles_obj_put_wait_timeo, "PUT wait_timeo") + +/* + * Export enum symbols via userspace. + */ +#undef EM +#undef E_ +#define EM(a, b) TRACE_DEFINE_ENUM(a); +#define E_(a, b) TRACE_DEFINE_ENUM(a); + +cachefiles_obj_kill_traces; +cachefiles_obj_ref_traces; + +/* + * Now redefine the EM() and E_() macros to map the enums to the strings that + * will be printed in the output. + */ +#undef EM +#undef E_ +#define EM(a, b) { a, b }, +#define E_(a, b) { a, b } + + +TRACE_EVENT(cachefiles_ref, + TP_PROTO(struct cachefiles_object *obj, + struct fscache_cookie *cookie, + enum cachefiles_obj_ref_trace why, + int usage), + + TP_ARGS(obj, cookie, why, usage), + + /* Note that obj may be NULL */ + TP_STRUCT__entry( + __field(struct cachefiles_object *, obj ) + __field(struct fscache_cookie *, cookie ) + __field(enum cachefiles_obj_ref_trace, why ) + __field(int, usage ) + ), + + TP_fast_assign( + __entry->obj = obj; + __entry->cookie = cookie; + __entry->usage = usage; + __entry->why = why; + ), + + TP_printk("c=%p o=%p u=%d %s", + __entry->cookie, __entry->obj, __entry->usage, + __print_symbolic(__entry->why, cachefiles_obj_ref_traces)) + ); + +TRACE_EVENT(cachefiles_lookup, + TP_PROTO(struct cachefiles_object *obj, + struct dentry *de, + struct inode *inode), + + TP_ARGS(obj, de, inode), + + TP_STRUCT__entry( + __field(struct cachefiles_object *, obj ) + __field(struct dentry *, de ) + __field(struct inode *, inode ) + ), + + TP_fast_assign( + __entry->obj = obj; + __entry->de = de; + __entry->inode = inode; + ), + + TP_printk("o=%p d=%p i=%p", + __entry->obj, __entry->de, __entry->inode) + ); + +TRACE_EVENT(cachefiles_mkdir, + TP_PROTO(struct cachefiles_object *obj, + struct dentry *de, int ret), + + TP_ARGS(obj, de, ret), + + TP_STRUCT__entry( + __field(struct cachefiles_object *, obj ) + __field(struct dentry *, de ) + __field(int, ret ) + ), + + TP_fast_assign( + __entry->obj = obj; + __entry->de = de; + __entry->ret = ret; + ), + + TP_printk("o=%p d=%p r=%u", + __entry->obj, __entry->de, __entry->ret) + ); + +TRACE_EVENT(cachefiles_create, + TP_PROTO(struct cachefiles_object *obj, + struct dentry *de, int ret), + + TP_ARGS(obj, de, ret), + + TP_STRUCT__entry( + __field(struct cachefiles_object *, obj ) + __field(struct dentry *, de ) + __field(int, ret ) + ), + + TP_fast_assign( + __entry->obj = obj; + __entry->de = de; + __entry->ret = ret; + ), + + TP_printk("o=%p d=%p r=%u", + __entry->obj, __entry->de, __entry->ret) + ); + +TRACE_EVENT(cachefiles_unlink, + TP_PROTO(struct cachefiles_object *obj, + struct dentry *de, + enum fscache_why_object_killed why), + + TP_ARGS(obj, de, why), + + /* Note that obj may be NULL */ + TP_STRUCT__entry( + __field(struct cachefiles_object *, obj ) + __field(struct dentry *, de ) + __field(enum fscache_why_object_killed, why ) + ), + + TP_fast_assign( + __entry->obj = obj; + __entry->de = de; + __entry->why = why; + ), + + TP_printk("o=%p d=%p w=%s", + __entry->obj, __entry->de, + __print_symbolic(__entry->why, cachefiles_obj_kill_traces)) + ); + +TRACE_EVENT(cachefiles_rename, + TP_PROTO(struct cachefiles_object *obj, + struct dentry *de, + struct dentry *to, + enum fscache_why_object_killed why), + + TP_ARGS(obj, de, to, why), + + /* Note that obj may be NULL */ + TP_STRUCT__entry( + __field(struct cachefiles_object *, obj ) + __field(struct dentry *, de ) + __field(struct dentry *, to ) + __field(enum fscache_why_object_killed, why ) + ), + + TP_fast_assign( + __entry->obj = obj; + __entry->de = de; + __entry->to = to; + __entry->why = why; + ), + + TP_printk("o=%p d=%p t=%p w=%s", + __entry->obj, __entry->de, __entry->to, + __print_symbolic(__entry->why, cachefiles_obj_kill_traces)) + ); + +TRACE_EVENT(cachefiles_mark_active, + TP_PROTO(struct cachefiles_object *obj, + struct dentry *de), + + TP_ARGS(obj, de), + + /* Note that obj may be NULL */ + TP_STRUCT__entry( + __field(struct cachefiles_object *, obj ) + __field(struct dentry *, de ) + ), + + TP_fast_assign( + __entry->obj = obj; + __entry->de = de; + ), + + TP_printk("o=%p d=%p", + __entry->obj, __entry->de) + ); + +TRACE_EVENT(cachefiles_wait_active, + TP_PROTO(struct cachefiles_object *obj, + struct dentry *de, + struct cachefiles_object *xobj), + + TP_ARGS(obj, de, xobj), + + /* Note that obj may be NULL */ + TP_STRUCT__entry( + __field(struct cachefiles_object *, obj ) + __field(struct dentry *, de ) + __field(struct cachefiles_object *, xobj ) + __field(u16, flags ) + __field(u16, fsc_flags ) + ), + + TP_fast_assign( + __entry->obj = obj; + __entry->de = de; + __entry->xobj = xobj; + __entry->flags = xobj->flags; + __entry->fsc_flags = xobj->fscache.flags; + ), + + TP_printk("o=%p d=%p wo=%p wf=%x wff=%x", + __entry->obj, __entry->de, __entry->xobj, + __entry->flags, __entry->fsc_flags) + ); + +TRACE_EVENT(cachefiles_mark_inactive, + TP_PROTO(struct cachefiles_object *obj, + struct dentry *de, + struct inode *inode), + + TP_ARGS(obj, de, inode), + + /* Note that obj may be NULL */ + TP_STRUCT__entry( + __field(struct cachefiles_object *, obj ) + __field(struct dentry *, de ) + __field(struct inode *, inode ) + ), + + TP_fast_assign( + __entry->obj = obj; + __entry->de = de; + __entry->inode = inode; + ), + + TP_printk("o=%p d=%p i=%p", + __entry->obj, __entry->de, __entry->inode) + ); + +TRACE_EVENT(cachefiles_mark_buried, + TP_PROTO(struct cachefiles_object *obj, + struct dentry *de, + enum fscache_why_object_killed why), + + TP_ARGS(obj, de, why), + + /* Note that obj may be NULL */ + TP_STRUCT__entry( + __field(struct cachefiles_object *, obj ) + __field(struct dentry *, de ) + __field(enum fscache_why_object_killed, why ) + ), + + TP_fast_assign( + __entry->obj = obj; + __entry->de = de; + __entry->why = why; + ), + + TP_printk("o=%p d=%p w=%s", + __entry->obj, __entry->de, + __print_symbolic(__entry->why, cachefiles_obj_kill_traces)) + ); + +#endif /* _TRACE_CACHEFILES_H */ + +/* This part must be outside protection */ +#include diff --git a/include/trace/events/fscache.h b/include/trace/events/fscache.h new file mode 100644 index 000000000000..99f5355d6281 --- /dev/null +++ b/include/trace/events/fscache.h @@ -0,0 +1,277 @@ +/* FS-Cache tracepoints + * + * Copyright (C) 2016 Red Hat, Inc. All Rights Reserved. + * Written by David Howells (dhowells@redhat.com) + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public Licence + * as published by the Free Software Foundation; either version + * 2 of the Licence, or (at your option) any later version. + */ +#undef TRACE_SYSTEM +#define TRACE_SYSTEM fscache + +#if !defined(_TRACE_FSCACHE_H) || defined(TRACE_HEADER_MULTI_READ) +#define _TRACE_FSCACHE_H + +#include +#include + +/* + * Define enums for tracing information. + */ +#ifndef __FSCACHE_DECLARE_TRACE_ENUMS_ONCE_ONLY +#define __FSCACHE_DECLARE_TRACE_ENUMS_ONCE_ONLY + +enum fscache_cookie_trace { + fscache_cookie_get_acquire_parent, + fscache_cookie_get_attach_object, + fscache_cookie_get_register_netfs, + fscache_cookie_put_acquire_nobufs, + fscache_cookie_put_relinquish, + fscache_cookie_put_object, + fscache_cookie_put_parent, +}; + +#endif + +/* + * Declare tracing information enums and their string mappings for display. + */ +#define fscache_cookie_traces \ + EM(fscache_cookie_get_acquire_parent, "GET prn") \ + EM(fscache_cookie_get_attach_object, "GET obj") \ + EM(fscache_cookie_get_register_netfs, "GET net") \ + EM(fscache_cookie_put_acquire_nobufs, "PUT nbf") \ + EM(fscache_cookie_put_relinquish, "PUT rlq") \ + EM(fscache_cookie_put_object, "PUT obj") \ + E_(fscache_cookie_put_parent, "PUT prn") + +/* + * Export enum symbols via userspace. + */ +#undef EM +#undef E_ +#define EM(a, b) TRACE_DEFINE_ENUM(a); +#define E_(a, b) TRACE_DEFINE_ENUM(a); + +fscache_cookie_traces; + +/* + * Now redefine the EM() and E_() macros to map the enums to the strings that + * will be printed in the output. + */ +#undef EM +#undef E_ +#define EM(a, b) { a, b }, +#define E_(a, b) { a, b } + + +TRACE_EVENT(fscache_cookie, + TP_PROTO(struct fscache_cookie *cookie, + enum fscache_cookie_trace where, + int usage), + + TP_ARGS(cookie, where, usage), + + TP_STRUCT__entry( + __field(struct fscache_cookie *, cookie ) + __field(struct fscache_cookie *, parent ) + __field(enum fscache_cookie_trace, where ) + __field(int, usage ) + __field(int, n_children ) + __field(int, n_active ) + __field(u8, flags ) + ), + + TP_fast_assign( + __entry->cookie = cookie; + __entry->parent = cookie->parent; + __entry->where = where; + __entry->usage = usage; + __entry->n_children = atomic_read(&cookie->n_children); + __entry->n_active = atomic_read(&cookie->n_active); + __entry->flags = cookie->flags; + ), + + TP_printk("%s c=%p u=%d p=%p Nc=%d Na=%d f=%02x", + __print_symbolic(__entry->where, fscache_cookie_traces), + __entry->cookie, __entry->usage, + __entry->parent, __entry->n_children, __entry->n_active, + __entry->flags) + ); + +TRACE_EVENT(fscache_netfs, + TP_PROTO(struct fscache_netfs *netfs), + + TP_ARGS(netfs), + + TP_STRUCT__entry( + __field(struct fscache_cookie *, cookie ) + __array(char, name, 8 ) + ), + + TP_fast_assign( + __entry->cookie = netfs->primary_index; + strncpy(__entry->name, netfs->name, 8); + __entry->name[7] = 0; + ), + + TP_printk("c=%p n=%s", + __entry->cookie, __entry->name) + ); + +TRACE_EVENT(fscache_acquire, + TP_PROTO(struct fscache_cookie *cookie), + + TP_ARGS(cookie), + + TP_STRUCT__entry( + __field(struct fscache_cookie *, cookie ) + __field(struct fscache_cookie *, parent ) + __array(char, name, 8 ) + __field(int, p_usage ) + __field(int, p_n_children ) + __field(u8, p_flags ) + ), + + TP_fast_assign( + __entry->cookie = cookie; + __entry->parent = cookie->parent; + __entry->p_usage = atomic_read(&cookie->parent->usage); + __entry->p_n_children = atomic_read(&cookie->parent->n_children); + __entry->p_flags = cookie->parent->flags; + memcpy(__entry->name, cookie->def->name, 8); + __entry->name[7] = 0; + ), + + TP_printk("c=%p p=%p pu=%d pc=%d pf=%02x n=%s", + __entry->cookie, __entry->parent, __entry->p_usage, + __entry->p_n_children, __entry->p_flags, __entry->name) + ); + +TRACE_EVENT(fscache_relinquish, + TP_PROTO(struct fscache_cookie *cookie, bool retire), + + TP_ARGS(cookie, retire), + + TP_STRUCT__entry( + __field(struct fscache_cookie *, cookie ) + __field(struct fscache_cookie *, parent ) + __field(int, usage ) + __field(int, n_children ) + __field(int, n_active ) + __field(u8, flags ) + __field(bool, retire ) + ), + + TP_fast_assign( + __entry->cookie = cookie; + __entry->parent = cookie->parent; + __entry->usage = atomic_read(&cookie->usage); + __entry->n_children = atomic_read(&cookie->n_children); + __entry->n_active = atomic_read(&cookie->n_active); + __entry->flags = cookie->flags; + __entry->retire = retire; + ), + + TP_printk("c=%p u=%d p=%p Nc=%d Na=%d f=%02x r=%u", + __entry->cookie, __entry->usage, + __entry->parent, __entry->n_children, __entry->n_active, + __entry->flags, __entry->retire) + ); + +TRACE_EVENT(fscache_enable, + TP_PROTO(struct fscache_cookie *cookie), + + TP_ARGS(cookie), + + TP_STRUCT__entry( + __field(struct fscache_cookie *, cookie ) + __field(int, usage ) + __field(int, n_children ) + __field(int, n_active ) + __field(u8, flags ) + ), + + TP_fast_assign( + __entry->cookie = cookie; + __entry->usage = atomic_read(&cookie->usage); + __entry->n_children = atomic_read(&cookie->n_children); + __entry->n_active = atomic_read(&cookie->n_active); + __entry->flags = cookie->flags; + ), + + TP_printk("c=%p u=%d Nc=%d Na=%d f=%02x", + __entry->cookie, __entry->usage, + __entry->n_children, __entry->n_active, __entry->flags) + ); + +TRACE_EVENT(fscache_disable, + TP_PROTO(struct fscache_cookie *cookie), + + TP_ARGS(cookie), + + TP_STRUCT__entry( + __field(struct fscache_cookie *, cookie ) + __field(int, usage ) + __field(int, n_children ) + __field(int, n_active ) + __field(u8, flags ) + ), + + TP_fast_assign( + __entry->cookie = cookie; + __entry->usage = atomic_read(&cookie->usage); + __entry->n_children = atomic_read(&cookie->n_children); + __entry->n_active = atomic_read(&cookie->n_active); + __entry->flags = cookie->flags; + ), + + TP_printk("c=%p u=%d Nc=%d Na=%d f=%02x", + __entry->cookie, __entry->usage, + __entry->n_children, __entry->n_active, __entry->flags) + ); + +TRACE_EVENT(fscache_osm, + TP_PROTO(struct fscache_object *object, + const struct fscache_state *state, + bool wait, bool oob, s8 event_num), + + TP_ARGS(object, state, wait, oob, event_num), + + TP_STRUCT__entry( + __field(struct fscache_cookie *, cookie ) + __field(struct fscache_object *, object ) + __array(char, state, 8 ) + __field(bool, wait ) + __field(bool, oob ) + __field(s8, event_num ) + ), + + TP_fast_assign( + __entry->cookie = object->cookie; + __entry->object = object; + __entry->wait = wait; + __entry->oob = oob; + __entry->event_num = event_num; + memcpy(__entry->state, state->short_name, 8); + ), + + TP_printk("c=%p o=%p %s %s%sev=%d", + __entry->cookie, + __entry->object, + __entry->state, + __print_symbolic(__entry->wait, + { true, "WAIT" }, + { false, "WORK" }), + __print_symbolic(__entry->oob, + { true, " OOB " }, + { false, " " }), + __entry->event_num) + ); + +#endif /* _TRACE_FSCACHE_H */ + +/* This part must be outside protection */ +#include -- cgit v1.2.3 From 08c2e3d087840cd1e7141b62d92f3dc897147984 Mon Sep 17 00:00:00 2001 From: David Howells Date: Wed, 4 Apr 2018 13:41:27 +0100 Subject: fscache: Add more tracepoints Add more tracepoints to fscache, including: (*) fscache_page - Tracks netfs pages known to fscache. (*) fscache_check_page - Tracks the netfs querying whether a page is pending storage. (*) fscache_wake_cookie - Tracks cookies being woken up after a page completes/aborts storage in the cache. (*) fscache_op - Tracks operations being initialised. (*) fscache_wrote_page - Tracks return of the backend write_page op. (*) fscache_gang_lookup - Tracks lookup of pages to be stored in the write operation. Signed-off-by: David Howells --- fs/fscache/cookie.c | 3 +- fs/fscache/object.c | 3 +- fs/fscache/operation.c | 26 ++++- fs/fscache/page.c | 51 ++++++++- include/linux/fscache-cache.h | 3 +- include/trace/events/fscache.h | 252 +++++++++++++++++++++++++++++++++++++++++ 6 files changed, 330 insertions(+), 8 deletions(-) (limited to 'include/trace') diff --git a/fs/fscache/cookie.c b/fs/fscache/cookie.c index 20bc3341f113..ea1f80daaff4 100644 --- a/fs/fscache/cookie.c +++ b/fs/fscache/cookie.c @@ -691,10 +691,11 @@ int __fscache_check_consistency(struct fscache_cookie *cookie) if (!op) return -ENOMEM; - fscache_operation_init(op, NULL, NULL, NULL); + fscache_operation_init(cookie, op, NULL, NULL, NULL); op->flags = FSCACHE_OP_MYTHREAD | (1 << FSCACHE_OP_WAITING) | (1 << FSCACHE_OP_UNUSE_COOKIE); + trace_fscache_page_op(cookie, NULL, op, fscache_page_op_check_consistency); spin_lock(&cookie->lock); diff --git a/fs/fscache/object.c b/fs/fscache/object.c index 99afe64352a5..7c0ddb7ae29a 100644 --- a/fs/fscache/object.c +++ b/fs/fscache/object.c @@ -982,11 +982,12 @@ static const struct fscache_state *_fscache_invalidate_object(struct fscache_obj if (!op) goto nomem; - fscache_operation_init(op, object->cache->ops->invalidate_object, + fscache_operation_init(cookie, op, object->cache->ops->invalidate_object, NULL, NULL); op->flags = FSCACHE_OP_ASYNC | (1 << FSCACHE_OP_EXCLUSIVE) | (1 << FSCACHE_OP_UNUSE_COOKIE); + trace_fscache_page_op(cookie, NULL, op, fscache_page_op_invalidate); spin_lock(&cookie->lock); if (fscache_submit_exclusive_op(object, op) < 0) diff --git a/fs/fscache/operation.c b/fs/fscache/operation.c index de67745e1cd7..7a071e1e952d 100644 --- a/fs/fscache/operation.c +++ b/fs/fscache/operation.c @@ -32,7 +32,8 @@ static void fscache_operation_dummy_cancel(struct fscache_operation *op) * Do basic initialisation of an operation. The caller must still set flags, * object and processor if needed. */ -void fscache_operation_init(struct fscache_operation *op, +void fscache_operation_init(struct fscache_cookie *cookie, + struct fscache_operation *op, fscache_operation_processor_t processor, fscache_operation_cancel_t cancel, fscache_operation_release_t release) @@ -46,6 +47,7 @@ void fscache_operation_init(struct fscache_operation *op, op->release = release; INIT_LIST_HEAD(&op->pend_link); fscache_stat(&fscache_n_op_initialised); + trace_fscache_op(cookie, op, fscache_op_init); } EXPORT_SYMBOL(fscache_operation_init); @@ -59,6 +61,8 @@ EXPORT_SYMBOL(fscache_operation_init); */ void fscache_enqueue_operation(struct fscache_operation *op) { + struct fscache_cookie *cookie = op->object->cookie; + _enter("{OBJ%x OP%x,%u}", op->object->debug_id, op->debug_id, atomic_read(&op->usage)); @@ -71,12 +75,14 @@ void fscache_enqueue_operation(struct fscache_operation *op) fscache_stat(&fscache_n_op_enqueue); switch (op->flags & FSCACHE_OP_TYPE) { case FSCACHE_OP_ASYNC: + trace_fscache_op(cookie, op, fscache_op_enqueue_async); _debug("queue async"); atomic_inc(&op->usage); if (!queue_work(fscache_op_wq, &op->work)) fscache_put_operation(op); break; case FSCACHE_OP_MYTHREAD: + trace_fscache_op(cookie, op, fscache_op_enqueue_mythread); _debug("queue for caller's attention"); break; default: @@ -101,6 +107,8 @@ static void fscache_run_op(struct fscache_object *object, wake_up_bit(&op->flags, FSCACHE_OP_WAITING); if (op->processor) fscache_enqueue_operation(op); + else + trace_fscache_op(object->cookie, op, fscache_op_run); fscache_stat(&fscache_n_op_run); } @@ -155,6 +163,8 @@ int fscache_submit_exclusive_op(struct fscache_object *object, _enter("{OBJ%x OP%x},", object->debug_id, op->debug_id); + trace_fscache_op(object->cookie, op, fscache_op_submit_ex); + ASSERTCMP(op->state, ==, FSCACHE_OP_ST_INITIALISED); ASSERTCMP(atomic_read(&op->usage), >, 0); @@ -240,6 +250,8 @@ int fscache_submit_op(struct fscache_object *object, _enter("{OBJ%x OP%x},{%u}", object->debug_id, op->debug_id, atomic_read(&op->usage)); + trace_fscache_op(object->cookie, op, fscache_op_submit); + ASSERTCMP(op->state, ==, FSCACHE_OP_ST_INITIALISED); ASSERTCMP(atomic_read(&op->usage), >, 0); @@ -357,6 +369,8 @@ int fscache_cancel_op(struct fscache_operation *op, _enter("OBJ%x OP%x}", op->object->debug_id, op->debug_id); + trace_fscache_op(object->cookie, op, fscache_op_cancel); + ASSERTCMP(op->state, >=, FSCACHE_OP_ST_PENDING); ASSERTCMP(op->state, !=, FSCACHE_OP_ST_CANCELLED); ASSERTCMP(atomic_read(&op->usage), >, 0); @@ -419,6 +433,8 @@ void fscache_cancel_all_ops(struct fscache_object *object) fscache_stat(&fscache_n_op_cancelled); list_del_init(&op->pend_link); + trace_fscache_op(object->cookie, op, fscache_op_cancel_all); + ASSERTCMP(op->state, ==, FSCACHE_OP_ST_PENDING); op->cancel(op); op->state = FSCACHE_OP_ST_CANCELLED; @@ -454,9 +470,11 @@ void fscache_op_complete(struct fscache_operation *op, bool cancelled) spin_lock(&object->lock); if (!cancelled) { + trace_fscache_op(object->cookie, op, fscache_op_completed); op->state = FSCACHE_OP_ST_COMPLETE; } else { op->cancel(op); + trace_fscache_op(object->cookie, op, fscache_op_cancelled); op->state = FSCACHE_OP_ST_CANCELLED; } @@ -488,6 +506,8 @@ void fscache_put_operation(struct fscache_operation *op) if (!atomic_dec_and_test(&op->usage)) return; + trace_fscache_op(op->object->cookie, op, fscache_op_put); + _debug("PUT OP"); ASSERTIFCMP(op->state != FSCACHE_OP_ST_INITIALISED && op->state != FSCACHE_OP_ST_COMPLETE, @@ -563,6 +583,8 @@ void fscache_operation_gc(struct work_struct *work) spin_unlock(&cache->op_gc_list_lock); object = op->object; + trace_fscache_op(object->cookie, op, fscache_op_gc); + spin_lock(&object->lock); _debug("GC DEFERRED REL OBJ%x OP%x", @@ -601,6 +623,8 @@ void fscache_op_work_func(struct work_struct *work) _enter("{OBJ%x OP%x,%d}", op->object->debug_id, op->debug_id, atomic_read(&op->usage)); + trace_fscache_op(op->object->cookie, op, fscache_op_work); + ASSERT(op->processor != NULL); start = jiffies; op->processor(op); diff --git a/fs/fscache/page.c b/fs/fscache/page.c index b9f18bf4227d..810b33aced1c 100644 --- a/fs/fscache/page.c +++ b/fs/fscache/page.c @@ -27,6 +27,7 @@ bool __fscache_check_page_write(struct fscache_cookie *cookie, struct page *page rcu_read_lock(); val = radix_tree_lookup(&cookie->stores, page->index); rcu_read_unlock(); + trace_fscache_check_page(cookie, page, val, 0); return val != NULL; } @@ -39,6 +40,8 @@ void __fscache_wait_on_page_write(struct fscache_cookie *cookie, struct page *pa { wait_queue_head_t *wq = bit_waitqueue(&cookie->flags, 0); + trace_fscache_page(cookie, page, fscache_page_write_wait); + wait_event(*wq, !__fscache_check_page_write(cookie, page)); } EXPORT_SYMBOL(__fscache_wait_on_page_write); @@ -69,6 +72,8 @@ bool __fscache_maybe_release_page(struct fscache_cookie *cookie, _enter("%p,%p,%x", cookie, page, gfp); + trace_fscache_page(cookie, page, fscache_page_maybe_release); + try_again: rcu_read_lock(); val = radix_tree_lookup(&cookie->stores, page->index); @@ -101,6 +106,7 @@ try_again: } xpage = radix_tree_delete(&cookie->stores, page->index); + trace_fscache_page(cookie, page, fscache_page_radix_delete); spin_unlock(&cookie->stores_lock); if (xpage) { @@ -112,6 +118,7 @@ try_again: } wake_up_bit(&cookie->flags, 0); + trace_fscache_wake_cookie(cookie); if (xpage) put_page(xpage); __fscache_uncache_page(cookie, page); @@ -144,7 +151,7 @@ static void fscache_end_page_write(struct fscache_object *object, struct page *page) { struct fscache_cookie *cookie; - struct page *xpage = NULL; + struct page *xpage = NULL, *val; spin_lock(&object->lock); cookie = object->cookie; @@ -154,13 +161,24 @@ static void fscache_end_page_write(struct fscache_object *object, spin_lock(&cookie->stores_lock); radix_tree_tag_clear(&cookie->stores, page->index, FSCACHE_COOKIE_STORING_TAG); + trace_fscache_page(cookie, page, fscache_page_radix_clear_store); if (!radix_tree_tag_get(&cookie->stores, page->index, FSCACHE_COOKIE_PENDING_TAG)) { fscache_stat(&fscache_n_store_radix_deletes); xpage = radix_tree_delete(&cookie->stores, page->index); + trace_fscache_page(cookie, page, fscache_page_radix_delete); + trace_fscache_page(cookie, page, fscache_page_write_end); + + val = radix_tree_lookup(&cookie->stores, page->index); + trace_fscache_check_page(cookie, page, val, 1); + } else { + trace_fscache_page(cookie, page, fscache_page_write_end_pend); } spin_unlock(&cookie->stores_lock); wake_up_bit(&cookie->flags, 0); + trace_fscache_wake_cookie(cookie); + } else { + trace_fscache_page(cookie, page, fscache_page_write_end_noc); } spin_unlock(&object->lock); if (xpage) @@ -215,7 +233,8 @@ int __fscache_attr_changed(struct fscache_cookie *cookie) return -ENOMEM; } - fscache_operation_init(op, fscache_attr_changed_op, NULL, NULL); + fscache_operation_init(cookie, op, fscache_attr_changed_op, NULL, NULL); + trace_fscache_page_op(cookie, NULL, op, fscache_page_op_attr_changed); op->flags = FSCACHE_OP_ASYNC | (1 << FSCACHE_OP_EXCLUSIVE) | (1 << FSCACHE_OP_UNUSE_COOKIE); @@ -299,7 +318,7 @@ static struct fscache_retrieval *fscache_alloc_retrieval( return NULL; } - fscache_operation_init(&op->op, NULL, + fscache_operation_init(cookie, &op->op, NULL, fscache_do_cancel_retrieval, fscache_release_retrieval_op); op->op.flags = FSCACHE_OP_MYTHREAD | @@ -370,6 +389,7 @@ int fscache_wait_for_operation_activation(struct fscache_object *object, fscache_stat(stat_op_waits); if (wait_on_bit(&op->flags, FSCACHE_OP_WAITING, TASK_INTERRUPTIBLE) != 0) { + trace_fscache_op(object->cookie, op, fscache_op_signal); ret = fscache_cancel_op(op, false); if (ret == 0) return -ERESTARTSYS; @@ -391,6 +411,7 @@ check_if_dead: if (unlikely(fscache_object_is_dying(object) || fscache_cache_is_broken(object))) { enum fscache_operation_state state = op->state; + trace_fscache_op(object->cookie, op, fscache_op_signal); fscache_cancel_op(op, true); if (stat_object_dead) fscache_stat(stat_object_dead); @@ -445,6 +466,7 @@ int __fscache_read_or_alloc_page(struct fscache_cookie *cookie, return -ENOMEM; } atomic_set(&op->n_pages, 1); + trace_fscache_page_op(cookie, page, &op->op, fscache_page_op_retr_one); spin_lock(&cookie->lock); @@ -573,6 +595,7 @@ int __fscache_read_or_alloc_pages(struct fscache_cookie *cookie, if (!op) return -ENOMEM; atomic_set(&op->n_pages, *nr_pages); + trace_fscache_page_op(cookie, NULL, &op->op, fscache_page_op_retr_multi); spin_lock(&cookie->lock); @@ -684,6 +707,7 @@ int __fscache_alloc_page(struct fscache_cookie *cookie, if (!op) return -ENOMEM; atomic_set(&op->n_pages, 1); + trace_fscache_page_op(cookie, page, &op->op, fscache_page_op_alloc_one); spin_lock(&cookie->lock); @@ -813,9 +837,11 @@ again: fscache_stat(&fscache_n_store_calls); /* find a page to store */ + results[0] = NULL; page = NULL; n = radix_tree_gang_lookup_tag(&cookie->stores, results, 0, 1, FSCACHE_COOKIE_PENDING_TAG); + trace_fscache_gang_lookup(cookie, &op->op, results, n, op->store_limit); if (n != 1) goto superseded; page = results[0]; @@ -825,6 +851,7 @@ again: FSCACHE_COOKIE_STORING_TAG); radix_tree_tag_clear(&cookie->stores, page->index, FSCACHE_COOKIE_PENDING_TAG); + trace_fscache_page(cookie, page, fscache_page_radix_pend2store); spin_unlock(&cookie->stores_lock); spin_unlock(&object->lock); @@ -836,6 +863,7 @@ again: fscache_stat(&fscache_n_cop_write_page); ret = object->cache->ops->write_page(op, page); fscache_stat_d(&fscache_n_cop_write_page); + trace_fscache_wrote_page(cookie, page, &op->op, ret); fscache_end_page_write(object, page); if (ret < 0) { fscache_abort_object(object); @@ -849,6 +877,7 @@ again: discard_page: fscache_stat(&fscache_n_store_pages_over_limit); + trace_fscache_wrote_page(cookie, page, &op->op, -ENOBUFS); fscache_end_page_write(object, page); goto again; @@ -887,6 +916,8 @@ void fscache_invalidate_writes(struct fscache_cookie *cookie) for (i = n - 1; i >= 0; i--) { page = results[i]; radix_tree_delete(&cookie->stores, page->index); + trace_fscache_page(cookie, page, fscache_page_radix_delete); + trace_fscache_page(cookie, page, fscache_page_inval); } spin_unlock(&cookie->stores_lock); @@ -896,6 +927,7 @@ void fscache_invalidate_writes(struct fscache_cookie *cookie) } wake_up_bit(&cookie->flags, 0); + trace_fscache_wake_cookie(cookie); _leave(""); } @@ -954,7 +986,7 @@ int __fscache_write_page(struct fscache_cookie *cookie, if (!op) goto nomem; - fscache_operation_init(&op->op, fscache_write_op, NULL, + fscache_operation_init(cookie, &op->op, fscache_write_op, NULL, fscache_release_write_op); op->op.flags = FSCACHE_OP_ASYNC | (1 << FSCACHE_OP_WAITING) | @@ -964,6 +996,8 @@ int __fscache_write_page(struct fscache_cookie *cookie, if (ret < 0) goto nomem_free; + trace_fscache_page_op(cookie, page, &op->op, fscache_page_op_write_one); + ret = -ENOBUFS; spin_lock(&cookie->lock); @@ -975,6 +1009,8 @@ int __fscache_write_page(struct fscache_cookie *cookie, if (test_bit(FSCACHE_IOERROR, &object->cache->flags)) goto nobufs; + trace_fscache_page(cookie, page, fscache_page_write); + /* add the page to the pending-storage radix tree on the backing * object */ spin_lock(&object->lock); @@ -990,8 +1026,10 @@ int __fscache_write_page(struct fscache_cookie *cookie, goto nobufs_unlock_obj; } + trace_fscache_page(cookie, page, fscache_page_radix_insert); radix_tree_tag_set(&cookie->stores, page->index, FSCACHE_COOKIE_PENDING_TAG); + trace_fscache_page(cookie, page, fscache_page_radix_set_pend); get_page(page); /* we only want one writer at a time, but we do need to queue new @@ -1034,6 +1072,7 @@ already_pending: submit_failed: spin_lock(&cookie->stores_lock); radix_tree_delete(&cookie->stores, page->index); + trace_fscache_page(cookie, page, fscache_page_radix_delete); spin_unlock(&cookie->stores_lock); wake_cookie = __fscache_unuse_cookie(cookie); put_page(page); @@ -1080,6 +1119,8 @@ void __fscache_uncache_page(struct fscache_cookie *cookie, struct page *page) if (!PageFsCache(page)) goto done; + trace_fscache_page(cookie, page, fscache_page_uncache); + /* get the object */ spin_lock(&cookie->lock); @@ -1128,6 +1169,8 @@ void fscache_mark_page_cached(struct fscache_retrieval *op, struct page *page) atomic_inc(&fscache_n_marks); #endif + trace_fscache_page(cookie, page, fscache_page_cached); + _debug("- mark %p{%lx}", page, page->index); if (TestSetPageFsCache(page)) { static bool once_only; diff --git a/include/linux/fscache-cache.h b/include/linux/fscache-cache.h index fbe102f37074..3e764fd38d9f 100644 --- a/include/linux/fscache-cache.h +++ b/include/linux/fscache-cache.h @@ -135,7 +135,8 @@ extern void fscache_op_work_func(struct work_struct *work); extern void fscache_enqueue_operation(struct fscache_operation *); extern void fscache_op_complete(struct fscache_operation *, bool); extern void fscache_put_operation(struct fscache_operation *); -extern void fscache_operation_init(struct fscache_operation *, +extern void fscache_operation_init(struct fscache_cookie *, + struct fscache_operation *, fscache_operation_processor_t, fscache_operation_cancel_t, fscache_operation_release_t); diff --git a/include/trace/events/fscache.h b/include/trace/events/fscache.h index 99f5355d6281..82c060fe6635 100644 --- a/include/trace/events/fscache.h +++ b/include/trace/events/fscache.h @@ -33,6 +33,53 @@ enum fscache_cookie_trace { fscache_cookie_put_parent, }; +enum fscache_page_trace { + fscache_page_cached, + fscache_page_inval, + fscache_page_maybe_release, + fscache_page_radix_clear_store, + fscache_page_radix_delete, + fscache_page_radix_insert, + fscache_page_radix_pend2store, + fscache_page_radix_set_pend, + fscache_page_uncache, + fscache_page_write, + fscache_page_write_end, + fscache_page_write_end_pend, + fscache_page_write_end_noc, + fscache_page_write_wait, + fscache_page_trace__nr +}; + +enum fscache_op_trace { + fscache_op_cancel, + fscache_op_cancel_all, + fscache_op_cancelled, + fscache_op_completed, + fscache_op_enqueue_async, + fscache_op_enqueue_mythread, + fscache_op_gc, + fscache_op_init, + fscache_op_put, + fscache_op_run, + fscache_op_signal, + fscache_op_submit, + fscache_op_submit_ex, + fscache_op_work, + fscache_op_trace__nr +}; + +enum fscache_page_op_trace { + fscache_page_op_alloc_one, + fscache_page_op_attr_changed, + fscache_page_op_check_consistency, + fscache_page_op_invalidate, + fscache_page_op_retr_multi, + fscache_page_op_retr_one, + fscache_page_op_write_one, + fscache_page_op_trace__nr +}; + #endif /* @@ -47,6 +94,47 @@ enum fscache_cookie_trace { EM(fscache_cookie_put_object, "PUT obj") \ E_(fscache_cookie_put_parent, "PUT prn") +#define fscache_page_traces \ + EM(fscache_page_cached, "Cached ") \ + EM(fscache_page_inval, "InvalPg") \ + EM(fscache_page_maybe_release, "MayRels") \ + EM(fscache_page_uncache, "Uncache") \ + EM(fscache_page_radix_clear_store, "RxCStr ") \ + EM(fscache_page_radix_delete, "RxDel ") \ + EM(fscache_page_radix_insert, "RxIns ") \ + EM(fscache_page_radix_pend2store, "RxP2S ") \ + EM(fscache_page_radix_set_pend, "RxSPend ") \ + EM(fscache_page_write, "WritePg") \ + EM(fscache_page_write_end, "EndPgWr") \ + EM(fscache_page_write_end_pend, "EndPgWP") \ + EM(fscache_page_write_end_noc, "EndPgNC") \ + E_(fscache_page_write_wait, "WtOnWrt") + +#define fscache_op_traces \ + EM(fscache_op_cancel, "Cancel1") \ + EM(fscache_op_cancel_all, "CancelA") \ + EM(fscache_op_cancelled, "Canclld") \ + EM(fscache_op_completed, "Complet") \ + EM(fscache_op_enqueue_async, "EnqAsyn") \ + EM(fscache_op_enqueue_mythread, "EnqMyTh") \ + EM(fscache_op_gc, "GC ") \ + EM(fscache_op_init, "Init ") \ + EM(fscache_op_put, "Put ") \ + EM(fscache_op_run, "Run ") \ + EM(fscache_op_signal, "Signal ") \ + EM(fscache_op_submit, "Submit ") \ + EM(fscache_op_submit_ex, "SubmitX") \ + E_(fscache_op_work, "Work ") + +#define fscache_page_op_traces \ + EM(fscache_page_op_alloc_one, "Alloc1 ") \ + EM(fscache_page_op_attr_changed, "AttrChg") \ + EM(fscache_page_op_check_consistency, "CheckCn") \ + EM(fscache_page_op_invalidate, "Inval ") \ + EM(fscache_page_op_retr_multi, "RetrMul") \ + EM(fscache_page_op_retr_one, "Retr1 ") \ + E_(fscache_page_op_write_one, "Write1 ") + /* * Export enum symbols via userspace. */ @@ -271,6 +359,170 @@ TRACE_EVENT(fscache_osm, __entry->event_num) ); +TRACE_EVENT(fscache_page, + TP_PROTO(struct fscache_cookie *cookie, struct page *page, + enum fscache_page_trace why), + + TP_ARGS(cookie, page, why), + + TP_STRUCT__entry( + __field(struct fscache_cookie *, cookie ) + __field(pgoff_t, page ) + __field(enum fscache_page_trace, why ) + ), + + TP_fast_assign( + __entry->cookie = cookie; + __entry->page = page->index; + __entry->why = why; + ), + + TP_printk("c=%p %s pg=%lx", + __entry->cookie, + __print_symbolic(__entry->why, fscache_page_traces), + __entry->page) + ); + +TRACE_EVENT(fscache_check_page, + TP_PROTO(struct fscache_cookie *cookie, struct page *page, + void *val, int n), + + TP_ARGS(cookie, page, val, n), + + TP_STRUCT__entry( + __field(struct fscache_cookie *, cookie ) + __field(void *, page ) + __field(void *, val ) + __field(int, n ) + ), + + TP_fast_assign( + __entry->cookie = cookie; + __entry->page = page; + __entry->val = val; + __entry->n = n; + ), + + TP_printk("c=%p pg=%p val=%p n=%d", + __entry->cookie, __entry->page, __entry->val, __entry->n) + ); + +TRACE_EVENT(fscache_wake_cookie, + TP_PROTO(struct fscache_cookie *cookie), + + TP_ARGS(cookie), + + TP_STRUCT__entry( + __field(struct fscache_cookie *, cookie ) + ), + + TP_fast_assign( + __entry->cookie = cookie; + ), + + TP_printk("c=%p", __entry->cookie) + ); + +TRACE_EVENT(fscache_op, + TP_PROTO(struct fscache_cookie *cookie, struct fscache_operation *op, + enum fscache_op_trace why), + + TP_ARGS(cookie, op, why), + + TP_STRUCT__entry( + __field(struct fscache_cookie *, cookie ) + __field(struct fscache_operation *, op ) + __field(enum fscache_op_trace, why ) + ), + + TP_fast_assign( + __entry->cookie = cookie; + __entry->op = op; + __entry->why = why; + ), + + TP_printk("c=%p op=%p %s", + __entry->cookie, __entry->op, + __print_symbolic(__entry->why, fscache_op_traces)) + ); + +TRACE_EVENT(fscache_page_op, + TP_PROTO(struct fscache_cookie *cookie, struct page *page, + struct fscache_operation *op, enum fscache_page_op_trace what), + + TP_ARGS(cookie, page, op, what), + + TP_STRUCT__entry( + __field(struct fscache_cookie *, cookie ) + __field(pgoff_t, page ) + __field(struct fscache_operation *, op ) + __field(enum fscache_page_op_trace, what ) + ), + + TP_fast_assign( + __entry->cookie = cookie; + __entry->page = page ? page->index : 0; + __entry->op = op; + __entry->what = what; + ), + + TP_printk("c=%p %s pg=%lx op=%p", + __entry->cookie, + __print_symbolic(__entry->what, fscache_page_op_traces), + __entry->page, __entry->op) + ); + +TRACE_EVENT(fscache_wrote_page, + TP_PROTO(struct fscache_cookie *cookie, struct page *page, + struct fscache_operation *op, int ret), + + TP_ARGS(cookie, page, op, ret), + + TP_STRUCT__entry( + __field(struct fscache_cookie *, cookie ) + __field(pgoff_t, page ) + __field(struct fscache_operation *, op ) + __field(int, ret ) + ), + + TP_fast_assign( + __entry->cookie = cookie; + __entry->page = page->index; + __entry->op = op; + __entry->ret = ret; + ), + + TP_printk("c=%p pg=%lx op=%p ret=%d", + __entry->cookie, __entry->page, __entry->op, __entry->ret) + ); + +TRACE_EVENT(fscache_gang_lookup, + TP_PROTO(struct fscache_cookie *cookie, struct fscache_operation *op, + void **results, int n, pgoff_t store_limit), + + TP_ARGS(cookie, op, results, n, store_limit), + + TP_STRUCT__entry( + __field(struct fscache_cookie *, cookie ) + __field(struct fscache_operation *, op ) + __field(pgoff_t, results0 ) + __field(int, n ) + __field(pgoff_t, store_limit ) + ), + + TP_fast_assign( + __entry->cookie = cookie; + __entry->op = op; + __entry->results0 = results[0] ? ((struct page *)results[0])->index : (pgoff_t)-1; + __entry->n = n; + __entry->store_limit = store_limit; + ), + + TP_printk("c=%p op=%p r0=%lx n=%d sl=%lx", + __entry->cookie, __entry->op, __entry->results0, __entry->n, + __entry->store_limit) + ); + #endif /* _TRACE_FSCACHE_H */ /* This part must be outside protection */ -- cgit v1.2.3 From 310253514bbf179c5f82e20a7a4bbf07abc7f5ad Mon Sep 17 00:00:00 2001 From: Anshuman Khandual Date: Thu, 5 Apr 2018 16:22:08 -0700 Subject: mm/migrate: rename migration reason MR_CMA to MR_CONTIG_RANGE alloc_contig_range() initiates compaction and eventual migration for the purpose of either CMA or HugeTLB allocations. At present, the reason code remains the same MR_CMA for either of these cases. Let's make it MR_CONTIG_RANGE which will appropriately reflect the reason code in both these cases. Link: http://lkml.kernel.org/r/20180202091518.18798-1-khandual@linux.vnet.ibm.com Signed-off-by: Anshuman Khandual Acked-by: Michal Hocko Reviewed-by: Andrew Morton Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/powerpc/mm/mmu_context_iommu.c | 2 +- include/linux/migrate.h | 2 +- include/trace/events/migrate.h | 2 +- mm/page_alloc.c | 2 +- 4 files changed, 4 insertions(+), 4 deletions(-) (limited to 'include/trace') diff --git a/arch/powerpc/mm/mmu_context_iommu.c b/arch/powerpc/mm/mmu_context_iommu.c index e0a2d8e806ed..9a8a084e4aba 100644 --- a/arch/powerpc/mm/mmu_context_iommu.c +++ b/arch/powerpc/mm/mmu_context_iommu.c @@ -112,7 +112,7 @@ static int mm_iommu_move_page_from_cma(struct page *page) put_page(page); /* Drop the gup reference */ ret = migrate_pages(&cma_migrate_pages, new_iommu_non_cma_page, - NULL, 0, MIGRATE_SYNC, MR_CMA); + NULL, 0, MIGRATE_SYNC, MR_CONTIG_RANGE); if (ret) { if (!list_empty(&cma_migrate_pages)) putback_movable_pages(&cma_migrate_pages); diff --git a/include/linux/migrate.h b/include/linux/migrate.h index a2246cf670ba..ab45f8a0d288 100644 --- a/include/linux/migrate.h +++ b/include/linux/migrate.h @@ -25,7 +25,7 @@ enum migrate_reason { MR_SYSCALL, /* also applies to cpusets */ MR_MEMPOLICY_MBIND, MR_NUMA_MISPLACED, - MR_CMA, + MR_CONTIG_RANGE, MR_TYPES }; diff --git a/include/trace/events/migrate.h b/include/trace/events/migrate.h index bcf4daccd6be..711372845945 100644 --- a/include/trace/events/migrate.h +++ b/include/trace/events/migrate.h @@ -20,7 +20,7 @@ EM( MR_SYSCALL, "syscall_or_cpuset") \ EM( MR_MEMPOLICY_MBIND, "mempolicy_mbind") \ EM( MR_NUMA_MISPLACED, "numa_misplaced") \ - EMe(MR_CMA, "cma") + EMe(MR_CONTIG_RANGE, "contig_range") /* * First define the enums in the above macros to be exported to userspace diff --git a/mm/page_alloc.c b/mm/page_alloc.c index 4ea018263210..531d6acb0106 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -7591,7 +7591,7 @@ static int __alloc_contig_migrate_range(struct compact_control *cc, cc->nr_migratepages -= nr_reclaimed; ret = migrate_pages(&cc->migratepages, alloc_migrate_target, - NULL, 0, cc->mode, MR_CMA); + NULL, 0, cc->mode, MR_CONTIG_RANGE); } if (ret < 0) { putback_movable_pages(&cc->migratepages); -- cgit v1.2.3 From 5ecd9d403ad081ed2de7b118c1e96124d4e0ba6c Mon Sep 17 00:00:00 2001 From: David Rientjes Date: Thu, 5 Apr 2018 16:25:16 -0700 Subject: mm, page_alloc: wakeup kcompactd even if kswapd cannot free more memory Kswapd will not wakeup if per-zone watermarks are not failing or if too many previous attempts at background reclaim have failed. This can be true if there is a lot of free memory available. For high- order allocations, kswapd is responsible for waking up kcompactd for background compaction. If the zone is not below its watermarks or reclaim has recently failed (lots of free memory, nothing left to reclaim), kcompactd does not get woken up. When __GFP_DIRECT_RECLAIM is not allowed, allow kcompactd to still be woken up even if kswapd will not reclaim. This allows high-order allocations, such as thp, to still trigger background compaction even when the zone has an abundance of free memory. Link: http://lkml.kernel.org/r/alpine.DEB.2.20.1803111659420.209721@chino.kir.corp.google.com Signed-off-by: David Rientjes Acked-by: Vlastimil Babka Cc: Mel Gorman Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- .../trace/postprocess/trace-vmscan-postprocess.pl | 4 +-- include/linux/mmzone.h | 3 +- include/trace/events/vmscan.h | 17 +++++++----- mm/page_alloc.c | 14 ++++++---- mm/vmscan.c | 32 ++++++++++++++++------ 5 files changed, 45 insertions(+), 25 deletions(-) (limited to 'include/trace') diff --git a/Documentation/trace/postprocess/trace-vmscan-postprocess.pl b/Documentation/trace/postprocess/trace-vmscan-postprocess.pl index ba976805853a..66bfd8396877 100644 --- a/Documentation/trace/postprocess/trace-vmscan-postprocess.pl +++ b/Documentation/trace/postprocess/trace-vmscan-postprocess.pl @@ -111,7 +111,7 @@ my $regex_direct_begin_default = 'order=([0-9]*) may_writepage=([0-9]*) gfp_flag my $regex_direct_end_default = 'nr_reclaimed=([0-9]*)'; my $regex_kswapd_wake_default = 'nid=([0-9]*) order=([0-9]*)'; my $regex_kswapd_sleep_default = 'nid=([0-9]*)'; -my $regex_wakeup_kswapd_default = 'nid=([0-9]*) zid=([0-9]*) order=([0-9]*)'; +my $regex_wakeup_kswapd_default = 'nid=([0-9]*) zid=([0-9]*) order=([0-9]*) gfp_flags=([A-Z_|]*)'; my $regex_lru_isolate_default = 'isolate_mode=([0-9]*) classzone_idx=([0-9]*) order=([0-9]*) nr_requested=([0-9]*) nr_scanned=([0-9]*) nr_skipped=([0-9]*) nr_taken=([0-9]*) lru=([a-z_]*)'; my $regex_lru_shrink_inactive_default = 'nid=([0-9]*) nr_scanned=([0-9]*) nr_reclaimed=([0-9]*) nr_dirty=([0-9]*) nr_writeback=([0-9]*) nr_congested=([0-9]*) nr_immediate=([0-9]*) nr_activate=([0-9]*) nr_ref_keep=([0-9]*) nr_unmap_fail=([0-9]*) priority=([0-9]*) flags=([A-Z_|]*)'; my $regex_lru_shrink_active_default = 'lru=([A-Z_]*) nr_scanned=([0-9]*) nr_rotated=([0-9]*) priority=([0-9]*)'; @@ -201,7 +201,7 @@ $regex_kswapd_sleep = generate_traceevent_regex( $regex_wakeup_kswapd = generate_traceevent_regex( "vmscan/mm_vmscan_wakeup_kswapd", $regex_wakeup_kswapd_default, - "nid", "zid", "order"); + "nid", "zid", "order", "gfp_flags"); $regex_lru_isolate = generate_traceevent_regex( "vmscan/mm_vmscan_lru_isolate", $regex_lru_isolate_default, diff --git a/include/linux/mmzone.h b/include/linux/mmzone.h index 5d935411d3c4..f11ae29005f1 100644 --- a/include/linux/mmzone.h +++ b/include/linux/mmzone.h @@ -776,7 +776,8 @@ static inline bool is_dev_zone(const struct zone *zone) #include void build_all_zonelists(pg_data_t *pgdat); -void wakeup_kswapd(struct zone *zone, int order, enum zone_type classzone_idx); +void wakeup_kswapd(struct zone *zone, gfp_t gfp_mask, int order, + enum zone_type classzone_idx); bool __zone_watermark_ok(struct zone *z, unsigned int order, unsigned long mark, int classzone_idx, unsigned int alloc_flags, long free_pages); diff --git a/include/trace/events/vmscan.h b/include/trace/events/vmscan.h index e0b8b9173e1c..6570c5b45ba1 100644 --- a/include/trace/events/vmscan.h +++ b/include/trace/events/vmscan.h @@ -78,26 +78,29 @@ TRACE_EVENT(mm_vmscan_kswapd_wake, TRACE_EVENT(mm_vmscan_wakeup_kswapd, - TP_PROTO(int nid, int zid, int order), + TP_PROTO(int nid, int zid, int order, gfp_t gfp_flags), - TP_ARGS(nid, zid, order), + TP_ARGS(nid, zid, order, gfp_flags), TP_STRUCT__entry( - __field( int, nid ) - __field( int, zid ) - __field( int, order ) + __field( int, nid ) + __field( int, zid ) + __field( int, order ) + __field( gfp_t, gfp_flags ) ), TP_fast_assign( __entry->nid = nid; __entry->zid = zid; __entry->order = order; + __entry->gfp_flags = gfp_flags; ), - TP_printk("nid=%d zid=%d order=%d", + TP_printk("nid=%d zid=%d order=%d gfp_flags=%s", __entry->nid, __entry->zid, - __entry->order) + __entry->order, + show_gfp_flags(__entry->gfp_flags)) ); DECLARE_EVENT_CLASS(mm_vmscan_direct_reclaim_begin_template, diff --git a/mm/page_alloc.c b/mm/page_alloc.c index f6005b7c3446..02c1a60d7937 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -3805,16 +3805,18 @@ retry: return page; } -static void wake_all_kswapds(unsigned int order, const struct alloc_context *ac) +static void wake_all_kswapds(unsigned int order, gfp_t gfp_mask, + const struct alloc_context *ac) { struct zoneref *z; struct zone *zone; pg_data_t *last_pgdat = NULL; + enum zone_type high_zoneidx = ac->high_zoneidx; - for_each_zone_zonelist_nodemask(zone, z, ac->zonelist, - ac->high_zoneidx, ac->nodemask) { + for_each_zone_zonelist_nodemask(zone, z, ac->zonelist, high_zoneidx, + ac->nodemask) { if (last_pgdat != zone->zone_pgdat) - wakeup_kswapd(zone, order, ac->high_zoneidx); + wakeup_kswapd(zone, gfp_mask, order, high_zoneidx); last_pgdat = zone->zone_pgdat; } } @@ -4093,7 +4095,7 @@ retry_cpuset: goto nopage; if (gfp_mask & __GFP_KSWAPD_RECLAIM) - wake_all_kswapds(order, ac); + wake_all_kswapds(order, gfp_mask, ac); /* * The adjusted alloc_flags might result in immediate success, so try @@ -4151,7 +4153,7 @@ retry_cpuset: retry: /* Ensure kswapd doesn't accidentally go to sleep as long as we loop */ if (gfp_mask & __GFP_KSWAPD_RECLAIM) - wake_all_kswapds(order, ac); + wake_all_kswapds(order, gfp_mask, ac); reserve_flags = __gfp_pfmemalloc_flags(gfp_mask); if (reserve_flags) diff --git a/mm/vmscan.c b/mm/vmscan.c index 976be140a8ce..4390a8d5be41 100644 --- a/mm/vmscan.c +++ b/mm/vmscan.c @@ -3539,16 +3539,21 @@ kswapd_try_sleep: } /* - * A zone is low on free memory, so wake its kswapd task to service it. + * A zone is low on free memory or too fragmented for high-order memory. If + * kswapd should reclaim (direct reclaim is deferred), wake it up for the zone's + * pgdat. It will wake up kcompactd after reclaiming memory. If kswapd reclaim + * has failed or is not needed, still wake up kcompactd if only compaction is + * needed. */ -void wakeup_kswapd(struct zone *zone, int order, enum zone_type classzone_idx) +void wakeup_kswapd(struct zone *zone, gfp_t gfp_flags, int order, + enum zone_type classzone_idx) { pg_data_t *pgdat; if (!managed_zone(zone)) return; - if (!cpuset_zone_allowed(zone, GFP_KERNEL | __GFP_HARDWALL)) + if (!cpuset_zone_allowed(zone, gfp_flags)) return; pgdat = zone->zone_pgdat; pgdat->kswapd_classzone_idx = kswapd_classzone_idx(pgdat, @@ -3557,14 +3562,23 @@ void wakeup_kswapd(struct zone *zone, int order, enum zone_type classzone_idx) if (!waitqueue_active(&pgdat->kswapd_wait)) return; - /* Hopeless node, leave it to direct reclaim */ - if (pgdat->kswapd_failures >= MAX_RECLAIM_RETRIES) - return; - - if (pgdat_balanced(pgdat, order, classzone_idx)) + /* Hopeless node, leave it to direct reclaim if possible */ + if (pgdat->kswapd_failures >= MAX_RECLAIM_RETRIES || + pgdat_balanced(pgdat, order, classzone_idx)) { + /* + * There may be plenty of free memory available, but it's too + * fragmented for high-order allocations. Wake up kcompactd + * and rely on compaction_suitable() to determine if it's + * needed. If it fails, it will defer subsequent attempts to + * ratelimit its work. + */ + if (!(gfp_flags & __GFP_DIRECT_RECLAIM)) + wakeup_kcompactd(pgdat, order, classzone_idx); return; + } - trace_mm_vmscan_wakeup_kswapd(pgdat->node_id, classzone_idx, order); + trace_mm_vmscan_wakeup_kswapd(pgdat->node_id, classzone_idx, order, + gfp_flags); wake_up_interruptible(&pgdat->kswapd_wait); } -- cgit v1.2.3 From 4ee7c60de83ac01fa4c33c55937357601631e8ad Mon Sep 17 00:00:00 2001 From: "Steven Rostedt (VMware)" Date: Fri, 23 Mar 2018 10:18:03 -0400 Subject: init, tracing: Add initcall trace events Being able to trace the start and stop of initcalls is useful to see where the timings are an issue. There is already an "initcall_debug" parameter, but that can cause a large overhead itself, as the printing of the information may take longer than the initcall functions. Adding in a start and finish trace event around the initcall functions, as well as a trace event that records the level of the initcalls, one can get a much finer measurement of the times and interactions of the initcalls themselves, as trace events are much lighter than printk()s. Suggested-by: Abderrahmane Benbachir Signed-off-by: Steven Rostedt (VMware) --- include/trace/events/initcall.h | 66 +++++++++++++++++++++++++++++++++++++++++ init/main.c | 7 +++++ 2 files changed, 73 insertions(+) create mode 100644 include/trace/events/initcall.h (limited to 'include/trace') diff --git a/include/trace/events/initcall.h b/include/trace/events/initcall.h new file mode 100644 index 000000000000..8d6cf10d27c9 --- /dev/null +++ b/include/trace/events/initcall.h @@ -0,0 +1,66 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#undef TRACE_SYSTEM +#define TRACE_SYSTEM initcall + +#if !defined(_TRACE_INITCALL_H) || defined(TRACE_HEADER_MULTI_READ) +#define _TRACE_INITCALL_H + +#include + +TRACE_EVENT(initcall_level, + + TP_PROTO(const char *level), + + TP_ARGS(level), + + TP_STRUCT__entry( + __string(level, level) + ), + + TP_fast_assign( + __assign_str(level, level); + ), + + TP_printk("level=%s", __get_str(level)) +); + +TRACE_EVENT(initcall_start, + + TP_PROTO(initcall_t func), + + TP_ARGS(func), + + TP_STRUCT__entry( + __field(initcall_t, func) + ), + + TP_fast_assign( + __entry->func = func; + ), + + TP_printk("func=%pS", __entry->func) +); + +TRACE_EVENT(initcall_finish, + + TP_PROTO(initcall_t func, int ret), + + TP_ARGS(func, ret), + + TP_STRUCT__entry( + __field(initcall_t, func) + __field(int, ret) + ), + + TP_fast_assign( + __entry->func = func; + __entry->ret = ret; + ), + + TP_printk("func=%pS ret=%d", __entry->func, __entry->ret) +); + +#endif /* if !defined(_TRACE_GPIO_H) || defined(TRACE_HEADER_MULTI_READ) */ + +/* This part must be outside protection */ +#include diff --git a/init/main.c b/init/main.c index 0ebdd5f15be8..2af8f2bb5ca8 100644 --- a/init/main.c +++ b/init/main.c @@ -97,6 +97,9 @@ #include #include +#define CREATE_TRACE_POINTS +#include + static int kernel_init(void *); extern void init_IRQ(void); @@ -827,10 +830,12 @@ int __init_or_module do_one_initcall(initcall_t fn) if (initcall_blacklisted(fn)) return -EPERM; + trace_initcall_start(fn); if (initcall_debug) ret = do_one_initcall_debug(fn); else ret = fn(); + trace_initcall_finish(fn, ret); msgbuf[0] = 0; @@ -895,6 +900,7 @@ static void __init do_initcall_level(int level) level, level, NULL, &repair_env_string); + trace_initcall_level(initcall_level_names[level]); for (fn = initcall_levels[level]; fn < initcall_levels[level+1]; fn++) do_one_initcall(*fn); } @@ -929,6 +935,7 @@ static void __init do_pre_smp_initcalls(void) { initcall_t *fn; + trace_initcall_level("early"); for (fn = __initcall_start; fn < __initcall0_start; fn++) do_one_initcall(*fn); } -- cgit v1.2.3 From ec0328e46d6e5d0f17372eb90ab8e333c2ac7ca9 Mon Sep 17 00:00:00 2001 From: David Howells Date: Wed, 4 Apr 2018 13:41:28 +0100 Subject: fscache: Maintain a catalogue of allocated cookies Maintain a catalogue of allocated cookies so that cookie collisions can be handled properly. For the moment, this just involves printing a warning and returning a NULL cookie to the caller of fscache_acquire_cookie(), but in future it might make sense to wait for the old cookie to finish being cleaned up. This requires the cookie key to be stored attached to the cookie so that we still have the key available if the netfs relinquishes the cookie. This is done by an earlier patch. The catalogue also renders redundant fscache_netfs_list (used for checking for duplicates), so that can be removed. Signed-off-by: David Howells Acked-by: Anna Schumaker Tested-by: Steve Dickson --- fs/fscache/cookie.c | 294 +++++++++++++++++++++++++++++++++-------- fs/fscache/internal.h | 7 + fs/fscache/netfs.c | 81 +++--------- include/linux/fscache.h | 8 +- include/trace/events/fscache.h | 8 ++ 5 files changed, 279 insertions(+), 119 deletions(-) (limited to 'include/trace') diff --git a/fs/fscache/cookie.c b/fs/fscache/cookie.c index 8ca9a932d225..7dc55b93a830 100644 --- a/fs/fscache/cookie.c +++ b/fs/fscache/cookie.c @@ -21,6 +21,9 @@ struct kmem_cache *fscache_cookie_jar; static atomic_t fscache_object_debug_id = ATOMIC_INIT(0); +#define fscache_cookie_hash_shift 15 +static struct hlist_bl_head fscache_cookie_hash[1 << fscache_cookie_hash_shift]; + static int fscache_acquire_non_index_cookie(struct fscache_cookie *cookie, loff_t object_size); static int fscache_alloc_object(struct fscache_cache *cache, @@ -28,6 +31,44 @@ static int fscache_alloc_object(struct fscache_cache *cache, static int fscache_attach_object(struct fscache_cookie *cookie, struct fscache_object *object); +static void fscache_print_cookie(struct fscache_cookie *cookie, char prefix) +{ + struct hlist_node *object; + const u8 *k; + unsigned loop; + + pr_err("%c-cookie c=%p [p=%p fl=%lx nc=%u na=%u]\n", + prefix, cookie, cookie->parent, cookie->flags, + atomic_read(&cookie->n_children), + atomic_read(&cookie->n_active)); + pr_err("%c-cookie d=%p n=%p\n", + prefix, cookie->def, cookie->netfs_data); + + object = READ_ONCE(cookie->backing_objects.first); + if (object) + pr_err("%c-cookie o=%p\n", + prefix, hlist_entry(object, struct fscache_object, cookie_link)); + + pr_err("%c-key=[%u] '", prefix, cookie->key_len); + k = (cookie->key_len <= sizeof(cookie->inline_key)) ? + cookie->inline_key : cookie->key; + for (loop = 0; loop < cookie->key_len; loop++) + pr_cont("%02x", k[loop]); + pr_cont("'\n"); +} + +void fscache_free_cookie(struct fscache_cookie *cookie) +{ + if (cookie) { + BUG_ON(!hlist_empty(&cookie->backing_objects)); + if (cookie->aux_len > sizeof(cookie->inline_aux)) + kfree(cookie->aux); + if (cookie->key_len > sizeof(cookie->inline_key)) + kfree(cookie->key); + kmem_cache_free(fscache_cookie_jar, cookie); + } +} + /* * initialise an cookie jar slab element prior to any use */ @@ -41,6 +82,170 @@ void fscache_cookie_init_once(void *_cookie) INIT_HLIST_HEAD(&cookie->backing_objects); } +/* + * Set the index key in a cookie. The cookie struct has space for a 12-byte + * key plus length and hash, but if that's not big enough, it's instead a + * pointer to a buffer containing 3 bytes of hash, 1 byte of length and then + * the key data. + */ +static int fscache_set_key(struct fscache_cookie *cookie, + const void *index_key, size_t index_key_len) +{ + unsigned long long h; + u32 *buf; + int i; + + cookie->key_len = index_key_len; + + if (index_key_len > sizeof(cookie->inline_key)) { + buf = kzalloc(index_key_len, GFP_KERNEL); + if (!buf) + return -ENOMEM; + cookie->key = buf; + } else { + buf = (u32 *)cookie->inline_key; + buf[0] = 0; + buf[1] = 0; + buf[2] = 0; + } + + memcpy(buf, index_key, index_key_len); + + /* Calculate a hash and combine this with the length in the first word + * or first half word + */ + h = (unsigned long)cookie->parent; + h += index_key_len + cookie->type; + for (i = 0; i < (index_key_len + sizeof(u32) - 1) / sizeof(u32); i++) + h += buf[i]; + + cookie->key_hash = h ^ (h >> 32); + return 0; +} + +static long fscache_compare_cookie(const struct fscache_cookie *a, + const struct fscache_cookie *b) +{ + const void *ka, *kb; + + if (a->key_hash != b->key_hash) + return (long)a->key_hash - (long)b->key_hash; + if (a->parent != b->parent) + return (long)a->parent - (long)b->parent; + if (a->key_len != b->key_len) + return (long)a->key_len - (long)b->key_len; + if (a->type != b->type) + return (long)a->type - (long)b->type; + + if (a->key_len <= sizeof(a->inline_key)) { + ka = &a->inline_key; + kb = &b->inline_key; + } else { + ka = a->key; + kb = b->key; + } + return memcmp(ka, kb, a->key_len); +} + +/* + * Allocate a cookie. + */ +struct fscache_cookie *fscache_alloc_cookie( + struct fscache_cookie *parent, + const struct fscache_cookie_def *def, + const void *index_key, size_t index_key_len, + const void *aux_data, size_t aux_data_len, + void *netfs_data, + loff_t object_size) +{ + struct fscache_cookie *cookie; + + /* allocate and initialise a cookie */ + cookie = kmem_cache_alloc(fscache_cookie_jar, GFP_KERNEL); + if (!cookie) + return NULL; + + cookie->key_len = index_key_len; + cookie->aux_len = aux_data_len; + + if (fscache_set_key(cookie, index_key, index_key_len) < 0) + goto nomem; + + if (cookie->aux_len <= sizeof(cookie->inline_aux)) { + memcpy(cookie->inline_aux, aux_data, cookie->aux_len); + } else { + cookie->aux = kmemdup(aux_data, cookie->aux_len, GFP_KERNEL); + if (!cookie->aux) + goto nomem; + } + + atomic_set(&cookie->usage, 1); + atomic_set(&cookie->n_children, 0); + + /* We keep the active count elevated until relinquishment to prevent an + * attempt to wake up every time the object operations queue quiesces. + */ + atomic_set(&cookie->n_active, 1); + + cookie->def = def; + cookie->parent = parent; + cookie->netfs_data = netfs_data; + cookie->flags = (1 << FSCACHE_COOKIE_NO_DATA_YET); + cookie->type = def->type; + + /* radix tree insertion won't use the preallocation pool unless it's + * told it may not wait */ + INIT_RADIX_TREE(&cookie->stores, GFP_NOFS & ~__GFP_DIRECT_RECLAIM); + return cookie; + +nomem: + fscache_free_cookie(cookie); + return NULL; +} + +/* + * Attempt to insert the new cookie into the hash. If there's a collision, we + * return the old cookie if it's not in use and an error otherwise. + */ +struct fscache_cookie *fscache_hash_cookie(struct fscache_cookie *candidate) +{ + struct fscache_cookie *cursor; + struct hlist_bl_head *h; + struct hlist_bl_node *p; + unsigned int bucket; + + bucket = candidate->key_hash & (ARRAY_SIZE(fscache_cookie_hash) - 1); + h = &fscache_cookie_hash[bucket]; + + hlist_bl_lock(h); + hlist_bl_for_each_entry(cursor, p, h, hash_link) { + if (fscache_compare_cookie(candidate, cursor) == 0) + goto collision; + } + + __set_bit(FSCACHE_COOKIE_ACQUIRED, &candidate->flags); + fscache_cookie_get(candidate->parent, fscache_cookie_get_acquire_parent); + atomic_inc(&candidate->parent->n_children); + hlist_bl_add_head(&candidate->hash_link, h); + hlist_bl_unlock(h); + return candidate; + +collision: + if (test_and_set_bit(FSCACHE_COOKIE_ACQUIRED, &cursor->flags)) { + trace_fscache_cookie(cursor, fscache_cookie_collision, + atomic_read(&cursor->usage)); + pr_err("Duplicate cookie detected\n"); + fscache_print_cookie(cursor, 'O'); + fscache_print_cookie(candidate, 'N'); + hlist_bl_unlock(h); + return NULL; + } + + fscache_cookie_get(cursor, fscache_cookie_get_reacquire); + hlist_bl_unlock(h); + return cursor; +} + /* * request a cookie to represent an object (index, datafile, xattr, etc) * - parent specifies the parent object @@ -65,7 +270,7 @@ struct fscache_cookie *__fscache_acquire_cookie( loff_t object_size, bool enable) { - struct fscache_cookie *cookie; + struct fscache_cookie *candidate, *cookie; BUG_ON(!def); @@ -95,53 +300,24 @@ struct fscache_cookie *__fscache_acquire_cookie( BUG_ON(def->type == FSCACHE_COOKIE_TYPE_INDEX && parent->type != FSCACHE_COOKIE_TYPE_INDEX); - /* allocate and initialise a cookie */ - cookie = kmem_cache_alloc(fscache_cookie_jar, GFP_KERNEL); - if (!cookie) { + candidate = fscache_alloc_cookie(parent, def, + index_key, index_key_len, + aux_data, aux_data_len, + netfs_data, object_size); + if (!candidate) { fscache_stat(&fscache_n_acquires_oom); _leave(" [ENOMEM]"); return NULL; } - cookie->key_len = index_key_len; - cookie->aux_len = aux_data_len; - - if (cookie->key_len <= sizeof(cookie->inline_key)) { - memcpy(cookie->inline_key, index_key, cookie->key_len); - } else { - cookie->key = kmemdup(index_key, cookie->key_len, GFP_KERNEL); - if (!cookie->key) - goto nomem; - } - - if (cookie->aux_len <= sizeof(cookie->inline_aux)) { - memcpy(cookie->inline_aux, aux_data, cookie->aux_len); - } else { - cookie->aux = kmemdup(aux_data, cookie->aux_len, GFP_KERNEL); - if (!cookie->aux) - goto nomem; + cookie = fscache_hash_cookie(candidate); + if (!cookie) { + trace_fscache_cookie(candidate, fscache_cookie_discard, 1); + goto out; } - atomic_set(&cookie->usage, 1); - atomic_set(&cookie->n_children, 0); - - /* We keep the active count elevated until relinquishment to prevent an - * attempt to wake up every time the object operations queue quiesces. - */ - atomic_set(&cookie->n_active, 1); - - fscache_cookie_get(parent, fscache_cookie_get_acquire_parent); - atomic_inc(&parent->n_children); - - cookie->def = def; - cookie->parent = parent; - cookie->netfs_data = netfs_data; - cookie->flags = (1 << FSCACHE_COOKIE_NO_DATA_YET); - cookie->type = def->type; - - /* radix tree insertion won't use the preallocation pool unless it's - * told it may not wait */ - INIT_RADIX_TREE(&cookie->stores, GFP_NOFS & ~__GFP_DIRECT_RECLAIM); + if (cookie == candidate) + candidate = NULL; switch (cookie->type) { case FSCACHE_COOKIE_TYPE_INDEX: @@ -178,16 +354,10 @@ struct fscache_cookie *__fscache_acquire_cookie( } fscache_stat(&fscache_n_acquires_ok); - _leave(" = %p", cookie); - return cookie; -nomem: - if (cookie->aux_len > sizeof(cookie->inline_aux)) - kfree(cookie->aux); - if (cookie->key_len > sizeof(cookie->inline_key)) - kfree(cookie->key); - kmem_cache_free(fscache_cookie_jar, cookie); - return NULL; +out: + fscache_free_cookie(candidate); + return cookie; } EXPORT_SYMBOL(__fscache_acquire_cookie); @@ -678,6 +848,22 @@ void __fscache_relinquish_cookie(struct fscache_cookie *cookie, } EXPORT_SYMBOL(__fscache_relinquish_cookie); +/* + * Remove a cookie from the hash table. + */ +static void fscache_unhash_cookie(struct fscache_cookie *cookie) +{ + struct hlist_bl_head *h; + unsigned int bucket; + + bucket = cookie->key_hash & (ARRAY_SIZE(fscache_cookie_hash) - 1); + h = &fscache_cookie_hash[bucket]; + + hlist_bl_lock(h); + hlist_bl_del(&cookie->hash_link); + hlist_bl_unlock(h); +} + /* * Drop a reference to a cookie. */ @@ -698,12 +884,8 @@ void fscache_cookie_put(struct fscache_cookie *cookie, BUG_ON(usage < 0); parent = cookie->parent; - BUG_ON(!hlist_empty(&cookie->backing_objects)); - if (cookie->aux_len > sizeof(cookie->inline_aux)) - kfree(cookie->aux); - if (cookie->key_len > sizeof(cookie->inline_key)) - kfree(cookie->key); - kmem_cache_free(fscache_cookie_jar, cookie); + fscache_unhash_cookie(cookie); + fscache_free_cookie(cookie); cookie = parent; where = fscache_cookie_put_parent; diff --git a/fs/fscache/internal.h b/fs/fscache/internal.h index 5f905a499306..500650f938fe 100644 --- a/fs/fscache/internal.h +++ b/fs/fscache/internal.h @@ -49,7 +49,14 @@ extern struct fscache_cache *fscache_select_cache_for_object( */ extern struct kmem_cache *fscache_cookie_jar; +extern void fscache_free_cookie(struct fscache_cookie *); extern void fscache_cookie_init_once(void *); +extern struct fscache_cookie *fscache_alloc_cookie(struct fscache_cookie *, + const struct fscache_cookie_def *, + const void *, size_t, + const void *, size_t, + void *, loff_t); +extern struct fscache_cookie *fscache_hash_cookie(struct fscache_cookie *); extern void fscache_cookie_put(struct fscache_cookie *, enum fscache_cookie_trace); diff --git a/fs/fscache/netfs.c b/fs/fscache/netfs.c index a5998dfab7e7..c2f605483cc5 100644 --- a/fs/fscache/netfs.c +++ b/fs/fscache/netfs.c @@ -14,85 +14,51 @@ #include #include "internal.h" -static LIST_HEAD(fscache_netfs_list); - /* * register a network filesystem for caching */ int __fscache_register_netfs(struct fscache_netfs *netfs) { - struct fscache_netfs *ptr; - struct fscache_cookie *cookie; - int ret; + struct fscache_cookie *candidate, *cookie; _enter("{%s}", netfs->name); - INIT_LIST_HEAD(&netfs->link); - /* allocate a cookie for the primary index */ - cookie = kmem_cache_zalloc(fscache_cookie_jar, GFP_KERNEL); - - if (!cookie) { + candidate = fscache_alloc_cookie(&fscache_fsdef_index, + &fscache_fsdef_netfs_def, + netfs->name, strlen(netfs->name), + &netfs->version, sizeof(netfs->version), + netfs, 0); + if (!candidate) { _leave(" = -ENOMEM"); return -ENOMEM; } - cookie->key_len = strlen(netfs->name); - if (cookie->key_len <= sizeof(cookie->inline_key)) { - memcpy(cookie->inline_key, netfs->name, strlen(netfs->name)); - } else { - ret = -ENOMEM; - cookie->key = kmemdup(netfs->name, cookie->key_len, GFP_KERNEL); - if (!cookie->key) - goto nomem; - } - - cookie->aux_len = sizeof(netfs->version); - memcpy(cookie->inline_aux, &netfs->version, cookie->aux_len); - - /* initialise the primary index cookie */ - atomic_set(&cookie->usage, 1); - atomic_set(&cookie->n_children, 0); - atomic_set(&cookie->n_active, 1); - - cookie->def = &fscache_fsdef_netfs_def; - cookie->parent = &fscache_fsdef_index; - cookie->netfs_data = netfs; - cookie->flags = 1 << FSCACHE_COOKIE_ENABLED; - cookie->type = FSCACHE_COOKIE_TYPE_INDEX; - - spin_lock_init(&cookie->lock); - spin_lock_init(&cookie->stores_lock); - INIT_HLIST_HEAD(&cookie->backing_objects); + candidate->flags = 1 << FSCACHE_COOKIE_ENABLED; /* check the netfs type is not already present */ - down_write(&fscache_addremove_sem); - - ret = -EEXIST; - list_for_each_entry(ptr, &fscache_netfs_list, link) { - if (strcmp(ptr->name, netfs->name) == 0) - goto already_registered; + cookie = fscache_hash_cookie(candidate); + if (!cookie) + goto already_registered; + if (cookie != candidate) { + trace_fscache_cookie(candidate, fscache_cookie_discard, 1); + fscache_free_cookie(candidate); } fscache_cookie_get(cookie->parent, fscache_cookie_get_register_netfs); atomic_inc(&cookie->parent->n_children); netfs->primary_index = cookie; - list_add(&netfs->link, &fscache_netfs_list); - ret = 0; pr_notice("Netfs '%s' registered for caching\n", netfs->name); trace_fscache_netfs(netfs); + _leave(" = 0"); + return 0; already_registered: - up_write(&fscache_addremove_sem); - -nomem: - if (ret < 0) - kmem_cache_free(fscache_cookie_jar, cookie); - - _leave(" = %d", ret); - return ret; + fscache_cookie_put(candidate, fscache_cookie_put_dup_netfs); + _leave(" = -EEXIST"); + return -EEXIST; } EXPORT_SYMBOL(__fscache_register_netfs); @@ -104,15 +70,8 @@ void __fscache_unregister_netfs(struct fscache_netfs *netfs) { _enter("{%s.%u}", netfs->name, netfs->version); - down_write(&fscache_addremove_sem); - - list_del(&netfs->link); fscache_relinquish_cookie(netfs->primary_index, NULL, false); - - up_write(&fscache_addremove_sem); - - pr_notice("Netfs '%s' unregistered from caching\n", - netfs->name); + pr_notice("Netfs '%s' unregistered from caching\n", netfs->name); _leave(""); } diff --git a/include/linux/fscache.h b/include/linux/fscache.h index eb38f39cf832..84b90a79d75a 100644 --- a/include/linux/fscache.h +++ b/include/linux/fscache.h @@ -22,6 +22,7 @@ #include #include #include +#include #if defined(CONFIG_FSCACHE) || defined(CONFIG_FSCACHE_MODULE) #define fscache_available() (1) @@ -124,7 +125,6 @@ struct fscache_netfs { uint32_t version; /* indexing version */ const char *name; /* filesystem name */ struct fscache_cookie *primary_index; - struct list_head link; /* internal link */ }; /* @@ -143,6 +143,7 @@ struct fscache_cookie { struct hlist_head backing_objects; /* object(s) backing this file/index */ const struct fscache_cookie_def *def; /* definition */ struct fscache_cookie *parent; /* parent of this entry */ + struct hlist_bl_node hash_link; /* Link in hash table */ void *netfs_data; /* back pointer to netfs */ struct radix_tree_root stores; /* pages to be stored on this cookie */ #define FSCACHE_COOKIE_PENDING_TAG 0 /* pages tag: pending write to cache */ @@ -156,11 +157,14 @@ struct fscache_cookie { #define FSCACHE_COOKIE_RELINQUISHED 4 /* T if cookie has been relinquished */ #define FSCACHE_COOKIE_ENABLED 5 /* T if cookie is enabled */ #define FSCACHE_COOKIE_ENABLEMENT_LOCK 6 /* T if cookie is being en/disabled */ -#define FSCACHE_COOKIE_AUX_UPDATED 7 /* T if the auxiliary data was updated */ +#define FSCACHE_COOKIE_AUX_UPDATED 8 /* T if the auxiliary data was updated */ +#define FSCACHE_COOKIE_ACQUIRED 9 /* T if cookie is in use */ +#define FSCACHE_COOKIE_RELINQUISHING 10 /* T if cookie is being relinquished */ u8 type; /* Type of object */ u8 key_len; /* Length of index key */ u8 aux_len; /* Length of auxiliary data */ + u32 key_hash; /* Hash of parent, type, key, len */ union { void *key; /* Index key */ u8 inline_key[16]; /* - If the key is short enough */ diff --git a/include/trace/events/fscache.h b/include/trace/events/fscache.h index 82c060fe6635..686cfe997ed2 100644 --- a/include/trace/events/fscache.h +++ b/include/trace/events/fscache.h @@ -24,10 +24,14 @@ #define __FSCACHE_DECLARE_TRACE_ENUMS_ONCE_ONLY enum fscache_cookie_trace { + fscache_cookie_collision, + fscache_cookie_discard, fscache_cookie_get_acquire_parent, fscache_cookie_get_attach_object, + fscache_cookie_get_reacquire, fscache_cookie_get_register_netfs, fscache_cookie_put_acquire_nobufs, + fscache_cookie_put_dup_netfs, fscache_cookie_put_relinquish, fscache_cookie_put_object, fscache_cookie_put_parent, @@ -86,10 +90,14 @@ enum fscache_page_op_trace { * Declare tracing information enums and their string mappings for display. */ #define fscache_cookie_traces \ + EM(fscache_cookie_collision, "*COLLISION*") \ + EM(fscache_cookie_discard, "DISCARD") \ EM(fscache_cookie_get_acquire_parent, "GET prn") \ EM(fscache_cookie_get_attach_object, "GET obj") \ + EM(fscache_cookie_get_reacquire, "GET raq") \ EM(fscache_cookie_get_register_netfs, "GET net") \ EM(fscache_cookie_put_acquire_nobufs, "PUT nbf") \ + EM(fscache_cookie_put_dup_netfs, "PUT dnt") \ EM(fscache_cookie_put_relinquish, "PUT rlq") \ EM(fscache_cookie_put_object, "PUT obj") \ E_(fscache_cookie_put_parent, "PUT prn") -- cgit v1.2.3 From 5cf9dd55a0ec26428f2824aadd16bfa305a5b603 Mon Sep 17 00:00:00 2001 From: David Howells Date: Mon, 9 Apr 2018 21:12:31 +0100 Subject: afs: Prospectively look up extra files when doing a single lookup When afs_lookup() is called, prospectively look up the next 50 uncached fids also from that same directory and cache the results, rather than just looking up the one file requested. This allows us to use the FS.InlineBulkStatus RPC op to increase efficiency by fetching up to 50 file statuses at a time. Signed-off-by: David Howells --- fs/afs/afs.h | 21 ++-- fs/afs/afs_fs.h | 2 + fs/afs/callback.c | 8 +- fs/afs/cmservice.c | 12 +- fs/afs/dir.c | 288 +++++++++++++++++++++++++++++++++++++++------ fs/afs/fsclient.c | 272 +++++++++++++++++++++++++++++++++++++++++- fs/afs/internal.h | 10 +- include/trace/events/afs.h | 2 + 8 files changed, 552 insertions(+), 63 deletions(-) (limited to 'include/trace') diff --git a/fs/afs/afs.h b/fs/afs/afs.h index b94d0edc2b78..a670bca6d3ba 100644 --- a/fs/afs/afs.h +++ b/fs/afs/afs.h @@ -67,10 +67,14 @@ typedef enum { } afs_callback_type_t; struct afs_callback { - struct afs_fid fid; /* file identifier */ - unsigned version; /* callback version */ - unsigned expiry; /* time at which expires */ - afs_callback_type_t type; /* type of callback */ + unsigned version; /* Callback version */ + unsigned expiry; /* Time at which expires */ + afs_callback_type_t type; /* Type of callback */ +}; + +struct afs_callback_break { + struct afs_fid fid; /* File identifier */ + struct afs_callback cb; /* Callback details */ }; #define AFSCBMAX 50 /* maximum callbacks transferred per bulk op */ @@ -123,21 +127,22 @@ typedef u32 afs_access_t; * AFS file status information */ struct afs_file_status { + u64 size; /* file size */ + afs_dataversion_t data_version; /* current data version */ + time_t mtime_client; /* last time client changed data */ + time_t mtime_server; /* last time server changed data */ unsigned if_version; /* interface version */ #define AFS_FSTATUS_VERSION 1 + unsigned abort_code; /* Abort if bulk-fetching this failed */ afs_file_type_t type; /* file type */ unsigned nlink; /* link count */ - u64 size; /* file size */ - afs_dataversion_t data_version; /* current data version */ u32 author; /* author ID */ kuid_t owner; /* owner ID */ kgid_t group; /* group ID */ afs_access_t caller_access; /* access rights for authenticated caller */ afs_access_t anon_access; /* access rights for unauthenticated caller */ umode_t mode; /* UNIX mode */ - time_t mtime_client; /* last time client changed data */ - time_t mtime_server; /* last time server changed data */ s32 lock_count; /* file lock count (0=UNLK -1=WRLCK +ve=#RDLCK */ }; diff --git a/fs/afs/afs_fs.h b/fs/afs/afs_fs.h index d47b6d01e4c0..ddfa88a7a9c0 100644 --- a/fs/afs/afs_fs.h +++ b/fs/afs/afs_fs.h @@ -31,10 +31,12 @@ enum AFS_FS_Operations { FSGETVOLUMEINFO = 148, /* AFS Get information about a volume */ FSGETVOLUMESTATUS = 149, /* AFS Get volume status information */ FSGETROOTVOLUME = 151, /* AFS Get root volume name */ + FSBULKSTATUS = 155, /* AFS Fetch multiple file statuses */ FSSETLOCK = 156, /* AFS Request a file lock */ FSEXTENDLOCK = 157, /* AFS Extend a file lock */ FSRELEASELOCK = 158, /* AFS Release a file lock */ FSLOOKUP = 161, /* AFS lookup file in directory */ + FSINLINEBULKSTATUS = 65536, /* AFS Fetch multiple file statuses with inline errors */ FSFETCHDATA64 = 65537, /* AFS Fetch file data */ FSSTOREDATA64 = 65538, /* AFS Store file data */ FSGIVEUPALLCALLBACKS = 65539, /* AFS Give up all outstanding callbacks on a server */ diff --git a/fs/afs/callback.c b/fs/afs/callback.c index bf082c719645..6049ca837498 100644 --- a/fs/afs/callback.c +++ b/fs/afs/callback.c @@ -187,7 +187,7 @@ static void afs_break_one_callback(struct afs_server *server, * allow the fileserver to break callback promises */ void afs_break_callbacks(struct afs_server *server, size_t count, - struct afs_callback callbacks[]) + struct afs_callback_break *callbacks) { _enter("%p,%zu,", server, count); @@ -199,9 +199,9 @@ void afs_break_callbacks(struct afs_server *server, size_t count, callbacks->fid.vid, callbacks->fid.vnode, callbacks->fid.unique, - callbacks->version, - callbacks->expiry, - callbacks->type + callbacks->cb.version, + callbacks->cb.expiry, + callbacks->cb.type ); afs_break_one_callback(server, &callbacks->fid); } diff --git a/fs/afs/cmservice.c b/fs/afs/cmservice.c index 83ff283979a4..fa07f83e9f29 100644 --- a/fs/afs/cmservice.c +++ b/fs/afs/cmservice.c @@ -178,8 +178,8 @@ static void SRXAFSCB_CallBack(struct work_struct *work) */ static int afs_deliver_cb_callback(struct afs_call *call) { + struct afs_callback_break *cb; struct sockaddr_rxrpc srx; - struct afs_callback *cb; struct afs_server *server; __be32 *bp; int ret, loop; @@ -218,7 +218,7 @@ static int afs_deliver_cb_callback(struct afs_call *call) _debug("unmarshall FID array"); call->request = kcalloc(call->count, - sizeof(struct afs_callback), + sizeof(struct afs_callback_break), GFP_KERNEL); if (!call->request) return -ENOMEM; @@ -229,7 +229,7 @@ static int afs_deliver_cb_callback(struct afs_call *call) cb->fid.vid = ntohl(*bp++); cb->fid.vnode = ntohl(*bp++); cb->fid.unique = ntohl(*bp++); - cb->type = AFSCM_CB_UNTYPED; + cb->cb.type = AFSCM_CB_UNTYPED; } call->offset = 0; @@ -260,9 +260,9 @@ static int afs_deliver_cb_callback(struct afs_call *call) cb = call->request; bp = call->buffer; for (loop = call->count2; loop > 0; loop--, cb++) { - cb->version = ntohl(*bp++); - cb->expiry = ntohl(*bp++); - cb->type = ntohl(*bp++); + cb->cb.version = ntohl(*bp++); + cb->cb.expiry = ntohl(*bp++); + cb->cb.type = ntohl(*bp++); } call->offset = 0; diff --git a/fs/afs/dir.c b/fs/afs/dir.c index ba2b458b36d1..27c5231e89e7 100644 --- a/fs/afs/dir.c +++ b/fs/afs/dir.c @@ -29,8 +29,10 @@ static int afs_readdir(struct file *file, struct dir_context *ctx); static int afs_d_revalidate(struct dentry *dentry, unsigned int flags); static int afs_d_delete(const struct dentry *dentry); static void afs_d_release(struct dentry *dentry); -static int afs_lookup_filldir(struct dir_context *ctx, const char *name, int nlen, +static int afs_lookup_one_filldir(struct dir_context *ctx, const char *name, int nlen, loff_t fpos, u64 ino, unsigned dtype); +static int afs_lookup_filldir(struct dir_context *ctx, const char *name, int nlen, + loff_t fpos, u64 ino, unsigned dtype); static int afs_create(struct inode *dir, struct dentry *dentry, umode_t mode, bool excl); static int afs_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode); @@ -134,11 +136,22 @@ struct afs_dir_page { union afs_dir_block blocks[PAGE_SIZE / sizeof(union afs_dir_block)]; }; +struct afs_lookup_one_cookie { + struct dir_context ctx; + struct qstr name; + bool found; + struct afs_fid fid; +}; + struct afs_lookup_cookie { - struct dir_context ctx; - struct afs_fid fid; - struct qstr name; - int found; + struct dir_context ctx; + struct qstr name; + bool found; + bool one_only; + unsigned short nr_fids; + struct afs_file_status *statuses; + struct afs_callback *callbacks; + struct afs_fid fids[50]; }; /* @@ -330,7 +343,8 @@ static int afs_dir_iterate_block(struct dir_context *ctx, /* found the next entry */ if (!dir_emit(ctx, dire->u.name, nlen, ntohl(dire->u.vnode), - ctx->actor == afs_lookup_filldir ? + (ctx->actor == afs_lookup_filldir || + ctx->actor == afs_lookup_one_filldir)? ntohl(dire->u.unique) : DT_UNKNOWN)) { _leave(" = 0 [full]"); return 0; @@ -414,15 +428,15 @@ static int afs_readdir(struct file *file, struct dir_context *ctx) } /* - * search the directory for a name + * Search the directory for a single name * - if afs_dir_iterate_block() spots this function, it'll pass the FID * uniquifier through dtype */ -static int afs_lookup_filldir(struct dir_context *ctx, const char *name, - int nlen, loff_t fpos, u64 ino, unsigned dtype) +static int afs_lookup_one_filldir(struct dir_context *ctx, const char *name, + int nlen, loff_t fpos, u64 ino, unsigned dtype) { - struct afs_lookup_cookie *cookie = - container_of(ctx, struct afs_lookup_cookie, ctx); + struct afs_lookup_one_cookie *cookie = + container_of(ctx, struct afs_lookup_one_cookie, ctx); _enter("{%s,%u},%s,%u,,%llu,%u", cookie->name.name, cookie->name.len, name, nlen, @@ -447,15 +461,15 @@ static int afs_lookup_filldir(struct dir_context *ctx, const char *name, } /* - * do a lookup in a directory + * Do a lookup of a single name in a directory * - just returns the FID the dentry name maps to if found */ -static int afs_do_lookup(struct inode *dir, struct dentry *dentry, - struct afs_fid *fid, struct key *key) +static int afs_do_lookup_one(struct inode *dir, struct dentry *dentry, + struct afs_fid *fid, struct key *key) { struct afs_super_info *as = dir->i_sb->s_fs_info; - struct afs_lookup_cookie cookie = { - .ctx.actor = afs_lookup_filldir, + struct afs_lookup_one_cookie cookie = { + .ctx.actor = afs_lookup_one_filldir, .name = dentry->d_name, .fid.vid = as->volume->vid }; @@ -481,6 +495,212 @@ static int afs_do_lookup(struct inode *dir, struct dentry *dentry, return 0; } +/* + * search the directory for a name + * - if afs_dir_iterate_block() spots this function, it'll pass the FID + * uniquifier through dtype + */ +static int afs_lookup_filldir(struct dir_context *ctx, const char *name, + int nlen, loff_t fpos, u64 ino, unsigned dtype) +{ + struct afs_lookup_cookie *cookie = + container_of(ctx, struct afs_lookup_cookie, ctx); + int ret; + + _enter("{%s,%u},%s,%u,,%llu,%u", + cookie->name.name, cookie->name.len, name, nlen, + (unsigned long long) ino, dtype); + + /* insanity checks first */ + BUILD_BUG_ON(sizeof(union afs_dir_block) != 2048); + BUILD_BUG_ON(sizeof(union afs_dirent) != 32); + + if (cookie->found) { + if (cookie->nr_fids < 50) { + cookie->fids[cookie->nr_fids].vnode = ino; + cookie->fids[cookie->nr_fids].unique = dtype; + cookie->nr_fids++; + } + } else if (cookie->name.len == nlen && + memcmp(cookie->name.name, name, nlen) == 0) { + cookie->fids[0].vnode = ino; + cookie->fids[0].unique = dtype; + cookie->found = 1; + if (cookie->one_only) + return -1; + } + + ret = cookie->nr_fids >= 50 ? -1 : 0; + _leave(" = %d", ret); + return ret; +} + +/* + * Do a lookup in a directory. We make use of bulk lookup to query a slew of + * files in one go and create inodes for them. The inode of the file we were + * asked for is returned. + */ +static struct inode *afs_do_lookup(struct inode *dir, struct dentry *dentry, + struct key *key) +{ + struct afs_lookup_cookie *cookie; + struct afs_cb_interest *cbi = NULL; + struct afs_super_info *as = dir->i_sb->s_fs_info; + struct afs_iget_data data; + struct afs_fs_cursor fc; + struct afs_vnode *dvnode = AFS_FS_I(dir); + struct inode *inode = NULL; + int ret, i; + + _enter("{%lu},%p{%pd},", dir->i_ino, dentry, dentry); + + cookie = kzalloc(sizeof(struct afs_lookup_cookie), GFP_KERNEL); + if (!cookie) + return ERR_PTR(-ENOMEM); + + cookie->ctx.actor = afs_lookup_filldir; + cookie->name = dentry->d_name; + cookie->nr_fids = 1; /* slot 0 is saved for the fid we actually want */ + + read_seqlock_excl(&dvnode->cb_lock); + if (dvnode->cb_interest && + dvnode->cb_interest->server && + test_bit(AFS_SERVER_FL_NO_IBULK, &dvnode->cb_interest->server->flags)) + cookie->one_only = true; + read_sequnlock_excl(&dvnode->cb_lock); + + for (i = 0; i < 50; i++) + cookie->fids[i].vid = as->volume->vid; + + /* search the directory */ + ret = afs_dir_iterate(dir, &cookie->ctx, key); + if (ret < 0) { + inode = ERR_PTR(ret); + goto out; + } + + inode = ERR_PTR(-ENOENT); + if (!cookie->found) + goto out; + + /* Check to see if we already have an inode for the primary fid. */ + data.volume = dvnode->volume; + data.fid = cookie->fids[0]; + inode = ilookup5(dir->i_sb, cookie->fids[0].vnode, afs_iget5_test, &data); + if (inode) + goto out; + + /* Need space for examining all the selected files */ + inode = ERR_PTR(-ENOMEM); + cookie->statuses = kcalloc(cookie->nr_fids, sizeof(struct afs_file_status), + GFP_KERNEL); + if (!cookie->statuses) + goto out; + + cookie->callbacks = kcalloc(cookie->nr_fids, sizeof(struct afs_callback), + GFP_KERNEL); + if (!cookie->callbacks) + goto out_s; + + /* Try FS.InlineBulkStatus first. Abort codes for the individual + * lookups contained therein are stored in the reply without aborting + * the whole operation. + */ + if (cookie->one_only) + goto no_inline_bulk_status; + + inode = ERR_PTR(-ERESTARTSYS); + if (afs_begin_vnode_operation(&fc, dvnode, key)) { + while (afs_select_fileserver(&fc)) { + if (test_bit(AFS_SERVER_FL_NO_IBULK, + &fc.cbi->server->flags)) { + fc.ac.abort_code = RX_INVALID_OPERATION; + fc.ac.error = -ECONNABORTED; + break; + } + afs_fs_inline_bulk_status(&fc, + afs_v2net(dvnode), + cookie->fids, + cookie->statuses, + cookie->callbacks, + cookie->nr_fids, NULL); + } + + if (fc.ac.error == 0) + cbi = afs_get_cb_interest(fc.cbi); + if (fc.ac.abort_code == RX_INVALID_OPERATION) + set_bit(AFS_SERVER_FL_NO_IBULK, &fc.cbi->server->flags); + inode = ERR_PTR(afs_end_vnode_operation(&fc)); + } + + if (!IS_ERR(inode)) + goto success; + if (fc.ac.abort_code != RX_INVALID_OPERATION) + goto out_c; + +no_inline_bulk_status: + /* We could try FS.BulkStatus next, but this aborts the entire op if + * any of the lookups fails - so, for the moment, revert to + * FS.FetchStatus for just the primary fid. + */ + cookie->nr_fids = 1; + inode = ERR_PTR(-ERESTARTSYS); + if (afs_begin_vnode_operation(&fc, dvnode, key)) { + while (afs_select_fileserver(&fc)) { + afs_fs_fetch_status(&fc, + afs_v2net(dvnode), + cookie->fids, + cookie->statuses, + cookie->callbacks, + NULL); + } + + if (fc.ac.error == 0) + cbi = afs_get_cb_interest(fc.cbi); + inode = ERR_PTR(afs_end_vnode_operation(&fc)); + } + + if (IS_ERR(inode)) + goto out_c; + + for (i = 0; i < cookie->nr_fids; i++) + cookie->statuses[i].abort_code = 0; + +success: + /* Turn all the files into inodes and save the first one - which is the + * one we actually want. + */ + if (cookie->statuses[0].abort_code != 0) + inode = ERR_PTR(afs_abort_to_error(cookie->statuses[0].abort_code)); + + for (i = 0; i < cookie->nr_fids; i++) { + struct inode *ti; + + if (cookie->statuses[i].abort_code != 0) + continue; + + ti = afs_iget(dir->i_sb, key, &cookie->fids[i], + &cookie->statuses[i], + &cookie->callbacks[i], + cbi); + if (i == 0) { + inode = ti; + } else { + if (!IS_ERR(ti)) + iput(ti); + } + } + +out_c: + afs_put_cb_interest(afs_v2net(dvnode), cbi); + kfree(cookie->callbacks); +out_s: + kfree(cookie->statuses); +out: + kfree(cookie); + return inode; +} + /* * Probe to see if a cell may exist. This prevents positive dentries from * being created unnecessarily. @@ -516,8 +736,7 @@ static int afs_probe_cell_name(struct dentry *dentry) * Try to auto mount the mountpoint with pseudo directory, if the autocell * operation is setted. */ -static struct inode *afs_try_auto_mntpt(struct dentry *dentry, - struct inode *dir, struct afs_fid *fid) +static struct inode *afs_try_auto_mntpt(struct dentry *dentry, struct inode *dir) { struct afs_vnode *vnode = AFS_FS_I(dir); struct inode *inode; @@ -539,7 +758,6 @@ static struct inode *afs_try_auto_mntpt(struct dentry *dentry, goto out; } - *fid = AFS_FS_I(inode)->fid; _leave("= %p", inode); return inode; @@ -554,16 +772,13 @@ out: static struct dentry *afs_lookup(struct inode *dir, struct dentry *dentry, unsigned int flags) { - struct afs_vnode *vnode; - struct afs_fid fid; + struct afs_vnode *dvnode = AFS_FS_I(dir); struct inode *inode; struct key *key; int ret; - vnode = AFS_FS_I(dir); - _enter("{%x:%u},%p{%pd},", - vnode->fid.vid, vnode->fid.vnode, dentry, dentry); + dvnode->fid.vid, dvnode->fid.vnode, dentry, dentry); ASSERTCMP(d_inode(dentry), ==, NULL); @@ -572,28 +787,29 @@ static struct dentry *afs_lookup(struct inode *dir, struct dentry *dentry, return ERR_PTR(-ENAMETOOLONG); } - if (test_bit(AFS_VNODE_DELETED, &vnode->flags)) { + if (test_bit(AFS_VNODE_DELETED, &dvnode->flags)) { _leave(" = -ESTALE"); return ERR_PTR(-ESTALE); } - key = afs_request_key(vnode->volume->cell); + key = afs_request_key(dvnode->volume->cell); if (IS_ERR(key)) { _leave(" = %ld [key]", PTR_ERR(key)); return ERR_CAST(key); } - ret = afs_validate(vnode, key); + ret = afs_validate(dvnode, key); if (ret < 0) { key_put(key); _leave(" = %d [val]", ret); return ERR_PTR(ret); } - ret = afs_do_lookup(dir, dentry, &fid, key); - if (ret < 0) { + inode = afs_do_lookup(dir, dentry, key); + if (IS_ERR(inode)) { + ret = PTR_ERR(inode); if (ret == -ENOENT) { - inode = afs_try_auto_mntpt(dentry, dir, &fid); + inode = afs_try_auto_mntpt(dentry, dir); if (!IS_ERR(inode)) { key_put(key); goto success; @@ -611,10 +827,9 @@ static struct dentry *afs_lookup(struct inode *dir, struct dentry *dentry, _leave(" = %d [do]", ret); return ERR_PTR(ret); } - dentry->d_fsdata = (void *)(unsigned long) vnode->status.data_version; + dentry->d_fsdata = (void *)(unsigned long)dvnode->status.data_version; /* instantiate the dentry */ - inode = afs_iget(dir->i_sb, key, &fid, NULL, NULL, NULL); key_put(key); if (IS_ERR(inode)) { _leave(" = %ld", PTR_ERR(inode)); @@ -623,9 +838,7 @@ static struct dentry *afs_lookup(struct inode *dir, struct dentry *dentry, success: d_add(dentry, inode); - _leave(" = 0 { vn=%u u=%u } -> { ino=%lu v=%u }", - fid.vnode, - fid.unique, + _leave(" = 0 { ino=%lu v=%u }", d_inode(dentry)->i_ino, d_inode(dentry)->i_generation); @@ -639,7 +852,6 @@ static struct dentry *afs_dynroot_lookup(struct inode *dir, struct dentry *dentr unsigned int flags) { struct afs_vnode *vnode; - struct afs_fid fid; struct inode *inode; int ret; @@ -654,7 +866,7 @@ static struct dentry *afs_dynroot_lookup(struct inode *dir, struct dentry *dentr return ERR_PTR(-ENAMETOOLONG); } - inode = afs_try_auto_mntpt(dentry, dir, &fid); + inode = afs_try_auto_mntpt(dentry, dir); if (IS_ERR(inode)) { ret = PTR_ERR(inode); if (ret == -ENOENT) { @@ -736,7 +948,7 @@ static int afs_d_revalidate(struct dentry *dentry, unsigned int flags) _debug("dir modified"); /* search the directory for this vnode */ - ret = afs_do_lookup(&dir->vfs_inode, dentry, &fid, key); + ret = afs_do_lookup_one(&dir->vfs_inode, dentry, &fid, key); switch (ret) { case 0: /* the filename maps to something */ diff --git a/fs/afs/fsclient.c b/fs/afs/fsclient.c index 88ec38c2d83c..75554ee98d02 100644 --- a/fs/afs/fsclient.c +++ b/fs/afs/fsclient.c @@ -94,7 +94,7 @@ static void xdr_decode_AFSFetchStatus(const __be32 **_bp, data_version |= (u64) ntohl(*bp++) << 32; EXTRACT(status->lock_count); size |= (u64) ntohl(*bp++) << 32; - bp++; /* spare 4 */ + EXTRACT(status->abort_code); /* spare 4 */ *_bp = bp; if (size != status->size) { @@ -274,7 +274,7 @@ static void xdr_decode_AFSFetchVolumeStatus(const __be32 **_bp, /* * deliver reply data to an FS.FetchStatus */ -static int afs_deliver_fs_fetch_status(struct afs_call *call) +static int afs_deliver_fs_fetch_status_vnode(struct afs_call *call) { struct afs_vnode *vnode = call->reply[0]; const __be32 *bp; @@ -300,10 +300,10 @@ static int afs_deliver_fs_fetch_status(struct afs_call *call) /* * FS.FetchStatus operation type */ -static const struct afs_call_type afs_RXFSFetchStatus = { - .name = "FS.FetchStatus", +static const struct afs_call_type afs_RXFSFetchStatus_vnode = { + .name = "FS.FetchStatus(vnode)", .op = afs_FS_FetchStatus, - .deliver = afs_deliver_fs_fetch_status, + .deliver = afs_deliver_fs_fetch_status_vnode, .destructor = afs_flat_call_destructor, }; @@ -320,7 +320,8 @@ int afs_fs_fetch_file_status(struct afs_fs_cursor *fc, struct afs_volsync *volsy _enter(",%x,{%x:%u},,", key_serial(fc->key), vnode->fid.vid, vnode->fid.vnode); - call = afs_alloc_flat_call(net, &afs_RXFSFetchStatus, 16, (21 + 3 + 6) * 4); + call = afs_alloc_flat_call(net, &afs_RXFSFetchStatus_vnode, + 16, (21 + 3 + 6) * 4); if (!call) { fc->ac.error = -ENOMEM; return -ENOMEM; @@ -1947,3 +1948,262 @@ int afs_fs_get_capabilities(struct afs_net *net, trace_afs_make_fs_call(call, NULL); return afs_make_call(ac, call, GFP_NOFS, false); } + +/* + * Deliver reply data to an FS.FetchStatus with no vnode. + */ +static int afs_deliver_fs_fetch_status(struct afs_call *call) +{ + struct afs_file_status *status = call->reply[1]; + struct afs_callback *callback = call->reply[2]; + struct afs_volsync *volsync = call->reply[3]; + struct afs_vnode *vnode = call->reply[0]; + const __be32 *bp; + int ret; + + ret = afs_transfer_reply(call); + if (ret < 0) + return ret; + + _enter("{%x:%u}", vnode->fid.vid, vnode->fid.vnode); + + /* unmarshall the reply once we've received all of it */ + bp = call->buffer; + xdr_decode_AFSFetchStatus(&bp, status, vnode, NULL); + callback[call->count].version = ntohl(bp[0]); + callback[call->count].expiry = ntohl(bp[1]); + callback[call->count].type = ntohl(bp[2]); + if (vnode) + xdr_decode_AFSCallBack(call, vnode, &bp); + else + bp += 3; + if (volsync) + xdr_decode_AFSVolSync(&bp, volsync); + + _leave(" = 0 [done]"); + return 0; +} + +/* + * FS.FetchStatus operation type + */ +static const struct afs_call_type afs_RXFSFetchStatus = { + .name = "FS.FetchStatus", + .op = afs_FS_FetchStatus, + .deliver = afs_deliver_fs_fetch_status, + .destructor = afs_flat_call_destructor, +}; + +/* + * Fetch the status information for a fid without needing a vnode handle. + */ +int afs_fs_fetch_status(struct afs_fs_cursor *fc, + struct afs_net *net, + struct afs_fid *fid, + struct afs_file_status *status, + struct afs_callback *callback, + struct afs_volsync *volsync) +{ + struct afs_call *call; + __be32 *bp; + + _enter(",%x,{%x:%u},,", + key_serial(fc->key), fid->vid, fid->vnode); + + call = afs_alloc_flat_call(net, &afs_RXFSFetchStatus, 16, (21 + 3 + 6) * 4); + if (!call) { + fc->ac.error = -ENOMEM; + return -ENOMEM; + } + + call->key = fc->key; + call->reply[0] = NULL; /* vnode for fid[0] */ + call->reply[1] = status; + call->reply[2] = callback; + call->reply[3] = volsync; + + /* marshall the parameters */ + bp = call->request; + bp[0] = htonl(FSFETCHSTATUS); + bp[1] = htonl(fid->vid); + bp[2] = htonl(fid->vnode); + bp[3] = htonl(fid->unique); + + call->cb_break = fc->cb_break; + afs_use_fs_server(call, fc->cbi); + trace_afs_make_fs_call(call, fid); + return afs_make_call(&fc->ac, call, GFP_NOFS, false); +} + +/* + * Deliver reply data to an FS.InlineBulkStatus call + */ +static int afs_deliver_fs_inline_bulk_status(struct afs_call *call) +{ + struct afs_file_status *statuses; + struct afs_callback *callbacks; + struct afs_vnode *vnode = call->reply[0]; + const __be32 *bp; + u32 tmp; + int ret; + + _enter("{%u}", call->unmarshall); + + switch (call->unmarshall) { + case 0: + call->offset = 0; + call->unmarshall++; + + /* Extract the file status count and array in two steps */ + case 1: + _debug("extract status count"); + ret = afs_extract_data(call, &call->tmp, 4, true); + if (ret < 0) + return ret; + + tmp = ntohl(call->tmp); + _debug("status count: %u/%u", tmp, call->count2); + if (tmp != call->count2) + return -EBADMSG; + + call->count = 0; + call->unmarshall++; + more_counts: + call->offset = 0; + + case 2: + _debug("extract status array %u", call->count); + ret = afs_extract_data(call, call->buffer, 21 * 4, true); + if (ret < 0) + return ret; + + bp = call->buffer; + statuses = call->reply[1]; + xdr_decode_AFSFetchStatus(&bp, &statuses[call->count], + call->count == 0 ? vnode : NULL, + NULL); + + call->count++; + if (call->count < call->count2) + goto more_counts; + + call->count = 0; + call->unmarshall++; + call->offset = 0; + + /* Extract the callback count and array in two steps */ + case 3: + _debug("extract CB count"); + ret = afs_extract_data(call, &call->tmp, 4, true); + if (ret < 0) + return ret; + + tmp = ntohl(call->tmp); + _debug("CB count: %u", tmp); + if (tmp != call->count2) + return -EBADMSG; + call->count = 0; + call->unmarshall++; + more_cbs: + call->offset = 0; + + case 4: + _debug("extract CB array"); + ret = afs_extract_data(call, call->buffer, 3 * 4, true); + if (ret < 0) + return ret; + + _debug("unmarshall CB array"); + bp = call->buffer; + callbacks = call->reply[2]; + callbacks[call->count].version = ntohl(bp[0]); + callbacks[call->count].expiry = ntohl(bp[1]); + callbacks[call->count].type = ntohl(bp[2]); + statuses = call->reply[1]; + if (call->count == 0 && vnode && statuses[0].abort_code == 0) + xdr_decode_AFSCallBack(call, vnode, &bp); + call->count++; + if (call->count < call->count2) + goto more_cbs; + + call->offset = 0; + call->unmarshall++; + + case 5: + ret = afs_extract_data(call, call->buffer, 6 * 4, false); + if (ret < 0) + return ret; + + bp = call->buffer; + if (call->reply[3]) + xdr_decode_AFSVolSync(&bp, call->reply[3]); + + call->offset = 0; + call->unmarshall++; + + case 6: + break; + } + + _leave(" = 0 [done]"); + return 0; +} + +/* + * FS.InlineBulkStatus operation type + */ +static const struct afs_call_type afs_RXFSInlineBulkStatus = { + .name = "FS.InlineBulkStatus", + .op = afs_FS_InlineBulkStatus, + .deliver = afs_deliver_fs_inline_bulk_status, + .destructor = afs_flat_call_destructor, +}; + +/* + * Fetch the status information for up to 50 files + */ +int afs_fs_inline_bulk_status(struct afs_fs_cursor *fc, + struct afs_net *net, + struct afs_fid *fids, + struct afs_file_status *statuses, + struct afs_callback *callbacks, + unsigned int nr_fids, + struct afs_volsync *volsync) +{ + struct afs_call *call; + __be32 *bp; + int i; + + _enter(",%x,{%x:%u},%u", + key_serial(fc->key), fids[0].vid, fids[1].vnode, nr_fids); + + call = afs_alloc_flat_call(net, &afs_RXFSInlineBulkStatus, + (2 + nr_fids * 3) * 4, + 21 * 4); + if (!call) { + fc->ac.error = -ENOMEM; + return -ENOMEM; + } + + call->key = fc->key; + call->reply[0] = NULL; /* vnode for fid[0] */ + call->reply[1] = statuses; + call->reply[2] = callbacks; + call->reply[3] = volsync; + call->count2 = nr_fids; + + /* marshall the parameters */ + bp = call->request; + *bp++ = htonl(FSINLINEBULKSTATUS); + *bp++ = htonl(nr_fids); + for (i = 0; i < nr_fids; i++) { + *bp++ = htonl(fids[i].vid); + *bp++ = htonl(fids[i].vnode); + *bp++ = htonl(fids[i].unique); + } + + call->cb_break = fc->cb_break; + afs_use_fs_server(call, fc->cbi); + trace_afs_make_fs_call(call, &fids[0]); + return afs_make_call(&fc->ac, call, GFP_NOFS, false); +} diff --git a/fs/afs/internal.h b/fs/afs/internal.h index 135192b7dc04..55b07e818400 100644 --- a/fs/afs/internal.h +++ b/fs/afs/internal.h @@ -363,6 +363,7 @@ struct afs_server { #define AFS_SERVER_FL_UPDATING 4 #define AFS_SERVER_FL_PROBED 5 /* The fileserver has been probed */ #define AFS_SERVER_FL_PROBING 6 /* Fileserver is being probed */ +#define AFS_SERVER_FL_NO_IBULK 7 /* Fileserver doesn't support FS.InlineBulkStatus */ atomic_t usage; u32 addr_version; /* Address list version */ @@ -611,7 +612,7 @@ extern struct fscache_cookie_def afs_vnode_cache_index_def; */ extern void afs_init_callback_state(struct afs_server *); extern void afs_break_callback(struct afs_vnode *); -extern void afs_break_callbacks(struct afs_server *, size_t,struct afs_callback[]); +extern void afs_break_callbacks(struct afs_server *, size_t, struct afs_callback_break*); extern int afs_register_server_cb_interest(struct afs_vnode *, struct afs_server_entry *); extern void afs_put_cb_interest(struct afs_net *, struct afs_cb_interest *); @@ -702,6 +703,13 @@ extern int afs_fs_give_up_all_callbacks(struct afs_net *, struct afs_server *, struct afs_addr_cursor *, struct key *); extern int afs_fs_get_capabilities(struct afs_net *, struct afs_server *, struct afs_addr_cursor *, struct key *); +extern int afs_fs_inline_bulk_status(struct afs_fs_cursor *, struct afs_net *, + struct afs_fid *, struct afs_file_status *, + struct afs_callback *, unsigned int, + struct afs_volsync *); +extern int afs_fs_fetch_status(struct afs_fs_cursor *, struct afs_net *, + struct afs_fid *, struct afs_file_status *, + struct afs_callback *, struct afs_volsync *); /* * inode.c diff --git a/include/trace/events/afs.h b/include/trace/events/afs.h index 63815f66b274..0419b7e1e968 100644 --- a/include/trace/events/afs.h +++ b/include/trace/events/afs.h @@ -49,6 +49,7 @@ enum afs_fs_operation { afs_FS_ExtendLock = 157, /* AFS Extend a file lock */ afs_FS_ReleaseLock = 158, /* AFS Release a file lock */ afs_FS_Lookup = 161, /* AFS lookup file in directory */ + afs_FS_InlineBulkStatus = 65536, /* AFS Fetch multiple file statuses with errors */ afs_FS_FetchData64 = 65537, /* AFS Fetch file data */ afs_FS_StoreData64 = 65538, /* AFS Store file data */ afs_FS_GiveUpAllCallBacks = 65539, /* AFS Give up all our callbacks on a server */ @@ -93,6 +94,7 @@ enum afs_vl_operation { EM(afs_FS_ExtendLock, "FS.ExtendLock") \ EM(afs_FS_ReleaseLock, "FS.ReleaseLock") \ EM(afs_FS_Lookup, "FS.Lookup") \ + EM(afs_FS_InlineBulkStatus, "FS.InlineBulkStatus") \ EM(afs_FS_FetchData64, "FS.FetchData64") \ EM(afs_FS_StoreData64, "FS.StoreData64") \ EM(afs_FS_GiveUpAllCallBacks, "FS.GiveUpAllCallBacks") \ -- cgit v1.2.3 From 63a4681ff39cb63314b8ff41319e70fb0e606ed2 Mon Sep 17 00:00:00 2001 From: David Howells Date: Fri, 6 Apr 2018 14:17:25 +0100 Subject: afs: Locally edit directory data for mkdir/create/unlink/... Locally edit the contents of an AFS directory upon a successful inode operation that modifies that directory (such as mkdir, create and unlink) so that we can avoid the current practice of re-downloading the directory after each change. This is viable provided that the directory version number we get back from the modifying RPC op is exactly incremented by 1 from what we had previously. The data in the directory contents is in a defined format that we have to parse locally to perform lookups and readdir, so modifying isn't a problem. If the edit fails, we just clear the VALID flag on the directory and it will be reloaded next time it is needed. Signed-off-by: David Howells --- fs/afs/Makefile | 1 + fs/afs/dir.c | 83 +++++++- fs/afs/dir_edit.c | 505 +++++++++++++++++++++++++++++++++++++++++++++ fs/afs/fsclient.c | 35 ++-- fs/afs/inode.c | 7 +- fs/afs/internal.h | 19 +- fs/afs/proc.c | 4 + include/trace/events/afs.h | 90 ++++++++ 8 files changed, 715 insertions(+), 29 deletions(-) create mode 100644 fs/afs/dir_edit.c (limited to 'include/trace') diff --git a/fs/afs/Makefile b/fs/afs/Makefile index 6a055423c192..532acae25453 100644 --- a/fs/afs/Makefile +++ b/fs/afs/Makefile @@ -12,6 +12,7 @@ kafs-objs := \ cell.o \ cmservice.o \ dir.o \ + dir_edit.o \ dynroot.o \ file.o \ flock.o \ diff --git a/fs/afs/dir.c b/fs/afs/dir.c index f078ae63d870..43bb3b23a879 100644 --- a/fs/afs/dir.c +++ b/fs/afs/dir.c @@ -130,17 +130,26 @@ static bool afs_dir_check_page(struct afs_vnode *dvnode, struct page *page, qty /= sizeof(union afs_xdr_dir_block); /* check them */ - dbuf = page_address(page); + dbuf = kmap(page); for (tmp = 0; tmp < qty; tmp++) { if (dbuf->blocks[tmp].hdr.magic != AFS_DIR_MAGIC) { printk("kAFS: %s(%lx): bad magic %d/%d is %04hx\n", __func__, dvnode->vfs_inode.i_ino, tmp, qty, ntohs(dbuf->blocks[tmp].hdr.magic)); trace_afs_dir_check_failed(dvnode, off, i_size); + kunmap(page); goto error; } + + /* Make sure each block is NUL terminated so we can reasonably + * use string functions on it. The filenames in the page + * *should* be NUL-terminated anyway. + */ + ((u8 *)&dbuf->blocks[tmp])[AFS_DIR_BLOCK_SIZE - 1] = 0; } + kunmap(page); + checked: afs_stat_v(dvnode, n_read_dir); return true; @@ -1114,6 +1123,7 @@ static int afs_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode) struct afs_vnode *dvnode = AFS_FS_I(dir); struct afs_fid newfid; struct key *key; + u64 data_version = dvnode->status.data_version; int ret; mode |= S_IFDIR; @@ -1131,7 +1141,7 @@ static int afs_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode) if (afs_begin_vnode_operation(&fc, dvnode, key)) { while (afs_select_fileserver(&fc)) { fc.cb_break = dvnode->cb_break + dvnode->cb_s_break; - afs_fs_create(&fc, dentry->d_name.name, mode, + afs_fs_create(&fc, dentry->d_name.name, mode, data_version, &newfid, &newstatus, &newcb); } @@ -1145,6 +1155,11 @@ static int afs_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode) goto error_key; } + if (ret == 0 && + test_bit(AFS_VNODE_DIR_VALID, &dvnode->flags)) + afs_edit_dir_add(dvnode, &dentry->d_name, &newfid, + afs_edit_dir_for_create); + key_put(key); _leave(" = 0"); return 0; @@ -1168,6 +1183,7 @@ static void afs_dir_remove_subdir(struct dentry *dentry) clear_nlink(&vnode->vfs_inode); set_bit(AFS_VNODE_DELETED, &vnode->flags); clear_bit(AFS_VNODE_CB_PROMISED, &vnode->flags); + clear_bit(AFS_VNODE_DIR_VALID, &vnode->flags); } } @@ -1179,6 +1195,7 @@ static int afs_rmdir(struct inode *dir, struct dentry *dentry) struct afs_fs_cursor fc; struct afs_vnode *dvnode = AFS_FS_I(dir); struct key *key; + u64 data_version = dvnode->status.data_version; int ret; _enter("{%x:%u},{%pd}", @@ -1194,13 +1211,18 @@ static int afs_rmdir(struct inode *dir, struct dentry *dentry) if (afs_begin_vnode_operation(&fc, dvnode, key)) { while (afs_select_fileserver(&fc)) { fc.cb_break = dvnode->cb_break + dvnode->cb_s_break; - afs_fs_remove(&fc, dentry->d_name.name, true); + afs_fs_remove(&fc, dentry->d_name.name, true, + data_version); } afs_vnode_commit_status(&fc, dvnode, fc.cb_break); ret = afs_end_vnode_operation(&fc); - if (ret == 0) + if (ret == 0) { afs_dir_remove_subdir(dentry); + if (test_bit(AFS_VNODE_DIR_VALID, &dvnode->flags)) + afs_edit_dir_remove(dvnode, &dentry->d_name, + afs_edit_dir_for_rmdir); + } } key_put(key); @@ -1265,6 +1287,7 @@ static int afs_unlink(struct inode *dir, struct dentry *dentry) struct afs_vnode *dvnode = AFS_FS_I(dir), *vnode; struct key *key; unsigned long d_version = (unsigned long)dentry->d_fsdata; + u64 data_version = dvnode->status.data_version; int ret; _enter("{%x:%u},{%pd}", @@ -1291,7 +1314,8 @@ static int afs_unlink(struct inode *dir, struct dentry *dentry) if (afs_begin_vnode_operation(&fc, dvnode, key)) { while (afs_select_fileserver(&fc)) { fc.cb_break = dvnode->cb_break + dvnode->cb_s_break; - afs_fs_remove(&fc, dentry->d_name.name, false); + afs_fs_remove(&fc, dentry->d_name.name, false, + data_version); } afs_vnode_commit_status(&fc, dvnode, fc.cb_break); @@ -1300,6 +1324,10 @@ static int afs_unlink(struct inode *dir, struct dentry *dentry) ret = afs_dir_remove_link( dentry, key, d_version, (unsigned long)dvnode->status.data_version); + if (ret == 0 && + test_bit(AFS_VNODE_DIR_VALID, &dvnode->flags)) + afs_edit_dir_remove(dvnode, &dentry->d_name, + afs_edit_dir_for_unlink); } error_key: @@ -1321,6 +1349,7 @@ static int afs_create(struct inode *dir, struct dentry *dentry, umode_t mode, struct afs_vnode *dvnode = AFS_FS_I(dir); struct afs_fid newfid; struct key *key; + u64 data_version = dvnode->status.data_version; int ret; mode |= S_IFREG; @@ -1342,7 +1371,7 @@ static int afs_create(struct inode *dir, struct dentry *dentry, umode_t mode, if (afs_begin_vnode_operation(&fc, dvnode, key)) { while (afs_select_fileserver(&fc)) { fc.cb_break = dvnode->cb_break + dvnode->cb_s_break; - afs_fs_create(&fc, dentry->d_name.name, mode, + afs_fs_create(&fc, dentry->d_name.name, mode, data_version, &newfid, &newstatus, &newcb); } @@ -1356,6 +1385,10 @@ static int afs_create(struct inode *dir, struct dentry *dentry, umode_t mode, goto error_key; } + if (test_bit(AFS_VNODE_DIR_VALID, &dvnode->flags)) + afs_edit_dir_add(dvnode, &dentry->d_name, &newfid, + afs_edit_dir_for_create); + key_put(key); _leave(" = 0"); return 0; @@ -1377,10 +1410,12 @@ static int afs_link(struct dentry *from, struct inode *dir, struct afs_fs_cursor fc; struct afs_vnode *dvnode, *vnode; struct key *key; + u64 data_version; int ret; vnode = AFS_FS_I(d_inode(from)); dvnode = AFS_FS_I(dir); + data_version = dvnode->status.data_version; _enter("{%x:%u},{%x:%u},{%pd}", vnode->fid.vid, vnode->fid.vnode, @@ -1407,7 +1442,7 @@ static int afs_link(struct dentry *from, struct inode *dir, while (afs_select_fileserver(&fc)) { fc.cb_break = dvnode->cb_break + dvnode->cb_s_break; fc.cb_break_2 = vnode->cb_break + vnode->cb_s_break; - afs_fs_link(&fc, vnode, dentry->d_name.name); + afs_fs_link(&fc, vnode, dentry->d_name.name, data_version); } afs_vnode_commit_status(&fc, dvnode, fc.cb_break); @@ -1423,6 +1458,10 @@ static int afs_link(struct dentry *from, struct inode *dir, goto error_key; } + if (test_bit(AFS_VNODE_DIR_VALID, &dvnode->flags)) + afs_edit_dir_add(dvnode, &dentry->d_name, &vnode->fid, + afs_edit_dir_for_link); + key_put(key); _leave(" = 0"); return 0; @@ -1446,6 +1485,7 @@ static int afs_symlink(struct inode *dir, struct dentry *dentry, struct afs_vnode *dvnode = AFS_FS_I(dir); struct afs_fid newfid; struct key *key; + u64 data_version = dvnode->status.data_version; int ret; _enter("{%x:%u},{%pd},%s", @@ -1470,7 +1510,8 @@ static int afs_symlink(struct inode *dir, struct dentry *dentry, if (afs_begin_vnode_operation(&fc, dvnode, key)) { while (afs_select_fileserver(&fc)) { fc.cb_break = dvnode->cb_break + dvnode->cb_s_break; - afs_fs_symlink(&fc, dentry->d_name.name, content, + afs_fs_symlink(&fc, dentry->d_name.name, + content, data_version, &newfid, &newstatus); } @@ -1484,6 +1525,10 @@ static int afs_symlink(struct inode *dir, struct dentry *dentry, goto error_key; } + if (test_bit(AFS_VNODE_DIR_VALID, &dvnode->flags)) + afs_edit_dir_add(dvnode, &dentry->d_name, &newfid, + afs_edit_dir_for_symlink); + key_put(key); _leave(" = 0"); return 0; @@ -1506,6 +1551,8 @@ static int afs_rename(struct inode *old_dir, struct dentry *old_dentry, struct afs_fs_cursor fc; struct afs_vnode *orig_dvnode, *new_dvnode, *vnode; struct key *key; + u64 orig_data_version, new_data_version; + bool new_negative = d_is_negative(new_dentry); int ret; if (flags) @@ -1514,6 +1561,8 @@ static int afs_rename(struct inode *old_dir, struct dentry *old_dentry, vnode = AFS_FS_I(d_inode(old_dentry)); orig_dvnode = AFS_FS_I(old_dir); new_dvnode = AFS_FS_I(new_dir); + orig_data_version = orig_dvnode->status.data_version; + new_data_version = new_dvnode->status.data_version; _enter("{%x:%u},{%x:%u},{%x:%u},{%pd}", orig_dvnode->fid.vid, orig_dvnode->fid.vnode, @@ -1539,7 +1588,8 @@ static int afs_rename(struct inode *old_dir, struct dentry *old_dentry, fc.cb_break = orig_dvnode->cb_break + orig_dvnode->cb_s_break; fc.cb_break_2 = new_dvnode->cb_break + new_dvnode->cb_s_break; afs_fs_rename(&fc, old_dentry->d_name.name, - new_dvnode, new_dentry->d_name.name); + new_dvnode, new_dentry->d_name.name, + orig_data_version, new_data_version); } afs_vnode_commit_status(&fc, orig_dvnode, fc.cb_break); @@ -1551,6 +1601,21 @@ static int afs_rename(struct inode *old_dir, struct dentry *old_dentry, goto error_key; } + if (ret == 0) { + if (test_bit(AFS_VNODE_DIR_VALID, &orig_dvnode->flags)) + afs_edit_dir_remove(orig_dvnode, &old_dentry->d_name, + afs_edit_dir_for_rename); + + if (!new_negative && + test_bit(AFS_VNODE_DIR_VALID, &new_dvnode->flags)) + afs_edit_dir_remove(new_dvnode, &new_dentry->d_name, + afs_edit_dir_for_rename); + + if (test_bit(AFS_VNODE_DIR_VALID, &new_dvnode->flags)) + afs_edit_dir_add(new_dvnode, &new_dentry->d_name, + &vnode->fid, afs_edit_dir_for_rename); + } + error_key: key_put(key); error: diff --git a/fs/afs/dir_edit.c b/fs/afs/dir_edit.c new file mode 100644 index 000000000000..8b400f5aead5 --- /dev/null +++ b/fs/afs/dir_edit.c @@ -0,0 +1,505 @@ +/* AFS filesystem directory editing + * + * Copyright (C) 2018 Red Hat, Inc. All Rights Reserved. + * Written by David Howells (dhowells@redhat.com) + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public Licence + * as published by the Free Software Foundation; either version + * 2 of the Licence, or (at your option) any later version. + */ + +#include +#include +#include +#include +#include +#include "internal.h" +#include "xdr_fs.h" + +/* + * Find a number of contiguous clear bits in a directory block bitmask. + * + * There are 64 slots, which means we can load the entire bitmap into a + * variable. The first bit doesn't count as it corresponds to the block header + * slot. nr_slots is between 1 and 9. + */ +static int afs_find_contig_bits(union afs_xdr_dir_block *block, unsigned int nr_slots) +{ + u64 bitmap; + u32 mask; + int bit, n; + + bitmap = (u64)block->hdr.bitmap[0] << 0 * 8; + bitmap |= (u64)block->hdr.bitmap[1] << 1 * 8; + bitmap |= (u64)block->hdr.bitmap[2] << 2 * 8; + bitmap |= (u64)block->hdr.bitmap[3] << 3 * 8; + bitmap |= (u64)block->hdr.bitmap[4] << 4 * 8; + bitmap |= (u64)block->hdr.bitmap[5] << 5 * 8; + bitmap |= (u64)block->hdr.bitmap[6] << 6 * 8; + bitmap |= (u64)block->hdr.bitmap[7] << 7 * 8; + bitmap >>= 1; /* The first entry is metadata */ + bit = 1; + mask = (1 << nr_slots) - 1; + + do { + if (sizeof(unsigned long) == 8) + n = ffz(bitmap); + else + n = ((u32)bitmap) != 0 ? + ffz((u32)bitmap) : + ffz((u32)(bitmap >> 32)) + 32; + bitmap >>= n; + bit += n; + + if ((bitmap & mask) == 0) { + if (bit > 64 - nr_slots) + return -1; + return bit; + } + + n = __ffs(bitmap); + bitmap >>= n; + bit += n; + } while (bitmap); + + return -1; +} + +/* + * Set a number of contiguous bits in the directory block bitmap. + */ +static void afs_set_contig_bits(union afs_xdr_dir_block *block, + int bit, unsigned int nr_slots) +{ + u64 mask, before, after; + + mask = (1 << nr_slots) - 1; + mask <<= bit; + + before = *(u64 *)block->hdr.bitmap; + + block->hdr.bitmap[0] |= (u8)(mask >> 0 * 8); + block->hdr.bitmap[1] |= (u8)(mask >> 1 * 8); + block->hdr.bitmap[2] |= (u8)(mask >> 2 * 8); + block->hdr.bitmap[3] |= (u8)(mask >> 3 * 8); + block->hdr.bitmap[4] |= (u8)(mask >> 4 * 8); + block->hdr.bitmap[5] |= (u8)(mask >> 5 * 8); + block->hdr.bitmap[6] |= (u8)(mask >> 6 * 8); + block->hdr.bitmap[7] |= (u8)(mask >> 7 * 8); + + after = *(u64 *)block->hdr.bitmap; +} + +/* + * Clear a number of contiguous bits in the directory block bitmap. + */ +static void afs_clear_contig_bits(union afs_xdr_dir_block *block, + int bit, unsigned int nr_slots) +{ + u64 mask, before, after; + + mask = (1 << nr_slots) - 1; + mask <<= bit; + + before = *(u64 *)block->hdr.bitmap; + + block->hdr.bitmap[0] &= ~(u8)(mask >> 0 * 8); + block->hdr.bitmap[1] &= ~(u8)(mask >> 1 * 8); + block->hdr.bitmap[2] &= ~(u8)(mask >> 2 * 8); + block->hdr.bitmap[3] &= ~(u8)(mask >> 3 * 8); + block->hdr.bitmap[4] &= ~(u8)(mask >> 4 * 8); + block->hdr.bitmap[5] &= ~(u8)(mask >> 5 * 8); + block->hdr.bitmap[6] &= ~(u8)(mask >> 6 * 8); + block->hdr.bitmap[7] &= ~(u8)(mask >> 7 * 8); + + after = *(u64 *)block->hdr.bitmap; +} + +/* + * Scan a directory block looking for a dirent of the right name. + */ +static int afs_dir_scan_block(union afs_xdr_dir_block *block, struct qstr *name, + unsigned int blocknum) +{ + union afs_xdr_dirent *de; + u64 bitmap; + int d, len, n; + + _enter(""); + + bitmap = (u64)block->hdr.bitmap[0] << 0 * 8; + bitmap |= (u64)block->hdr.bitmap[1] << 1 * 8; + bitmap |= (u64)block->hdr.bitmap[2] << 2 * 8; + bitmap |= (u64)block->hdr.bitmap[3] << 3 * 8; + bitmap |= (u64)block->hdr.bitmap[4] << 4 * 8; + bitmap |= (u64)block->hdr.bitmap[5] << 5 * 8; + bitmap |= (u64)block->hdr.bitmap[6] << 6 * 8; + bitmap |= (u64)block->hdr.bitmap[7] << 7 * 8; + + for (d = (blocknum == 0 ? AFS_DIR_RESV_BLOCKS0 : AFS_DIR_RESV_BLOCKS); + d < AFS_DIR_SLOTS_PER_BLOCK; + d++) { + if (!((bitmap >> d) & 1)) + continue; + de = &block->dirents[d]; + if (de->u.valid != 1) + continue; + + /* The block was NUL-terminated by afs_dir_check_page(). */ + len = strlen(de->u.name); + if (len == name->len && + memcmp(de->u.name, name->name, name->len) == 0) + return d; + + n = round_up(12 + len + 1 + 4, AFS_DIR_DIRENT_SIZE); + n /= AFS_DIR_DIRENT_SIZE; + d += n - 1; + } + + return -1; +} + +/* + * Initialise a new directory block. Note that block 0 is special and contains + * some extra metadata. + */ +static void afs_edit_init_block(union afs_xdr_dir_block *meta, + union afs_xdr_dir_block *block, int block_num) +{ + memset(block, 0, sizeof(*block)); + block->hdr.npages = htons(1); + block->hdr.magic = AFS_DIR_MAGIC; + block->hdr.bitmap[0] = 1; + + if (block_num == 0) { + block->hdr.bitmap[0] = 0xff; + block->hdr.bitmap[1] = 0x1f; + memset(block->meta.alloc_ctrs, + AFS_DIR_SLOTS_PER_BLOCK, + sizeof(block->meta.alloc_ctrs)); + meta->meta.alloc_ctrs[0] = + AFS_DIR_SLOTS_PER_BLOCK - AFS_DIR_RESV_BLOCKS0; + } + + if (block_num < AFS_DIR_BLOCKS_WITH_CTR) + meta->meta.alloc_ctrs[block_num] = + AFS_DIR_SLOTS_PER_BLOCK - AFS_DIR_RESV_BLOCKS; +} + +/* + * Edit a directory's file data to add a new directory entry. Doing this after + * create, mkdir, symlink, link or rename if the data version number is + * incremented by exactly one avoids the need to re-download the entire + * directory contents. + * + * The caller must hold the inode locked. + */ +void afs_edit_dir_add(struct afs_vnode *vnode, + struct qstr *name, struct afs_fid *new_fid, + enum afs_edit_dir_reason why) +{ + union afs_xdr_dir_block *meta, *block; + struct afs_xdr_dir_page *meta_page, *dir_page; + union afs_xdr_dirent *de; + struct page *page0, *page; + unsigned int need_slots, nr_blocks, b; + pgoff_t index; + loff_t i_size; + gfp_t gfp; + int slot; + + _enter(",,{%d,%s},", name->len, name->name); + + i_size = i_size_read(&vnode->vfs_inode); + if (i_size > AFS_DIR_BLOCK_SIZE * AFS_DIR_MAX_BLOCKS || + (i_size & (AFS_DIR_BLOCK_SIZE - 1))) { + clear_bit(AFS_VNODE_DIR_VALID, &vnode->flags); + return; + } + + gfp = vnode->vfs_inode.i_mapping->gfp_mask; + page0 = find_or_create_page(vnode->vfs_inode.i_mapping, 0, gfp); + if (!page0) { + clear_bit(AFS_VNODE_DIR_VALID, &vnode->flags); + _leave(" [fgp]"); + return; + } + + /* Work out how many slots we're going to need. */ + need_slots = round_up(12 + name->len + 1 + 4, AFS_DIR_DIRENT_SIZE); + need_slots /= AFS_DIR_DIRENT_SIZE; + + meta_page = kmap(page0); + meta = &meta_page->blocks[0]; + if (i_size == 0) + goto new_directory; + nr_blocks = i_size / AFS_DIR_BLOCK_SIZE; + + /* Find a block that has sufficient slots available. Each VM page + * contains two or more directory blocks. + */ + for (b = 0; b < nr_blocks + 1; b++) { + /* If the directory extended into a new page, then we need to + * tack a new page on the end. + */ + index = b / AFS_DIR_BLOCKS_PER_PAGE; + if (index == 0) { + page = page0; + dir_page = meta_page; + } else { + if (nr_blocks >= AFS_DIR_MAX_BLOCKS) + goto error; + gfp = vnode->vfs_inode.i_mapping->gfp_mask; + page = find_or_create_page(vnode->vfs_inode.i_mapping, + index, gfp); + if (!page) + goto error; + if (!PagePrivate(page)) { + set_page_private(page, 1); + SetPagePrivate(page); + } + dir_page = kmap(page); + } + + /* Abandon the edit if we got a callback break. */ + if (!test_bit(AFS_VNODE_DIR_VALID, &vnode->flags)) + goto invalidated; + + block = &dir_page->blocks[b % AFS_DIR_BLOCKS_PER_PAGE]; + + _debug("block %u: %2u %3u %u", + b, + (b < AFS_DIR_BLOCKS_WITH_CTR) ? meta->meta.alloc_ctrs[b] : 99, + ntohs(block->hdr.npages), + ntohs(block->hdr.magic)); + + /* Initialise the block if necessary. */ + if (b == nr_blocks) { + _debug("init %u", b); + afs_edit_init_block(meta, block, b); + i_size_write(&vnode->vfs_inode, (b + 1) * AFS_DIR_BLOCK_SIZE); + } + + /* Only lower dir pages have a counter in the header. */ + if (b >= AFS_DIR_BLOCKS_WITH_CTR || + meta->meta.alloc_ctrs[b] >= need_slots) { + /* We need to try and find one or more consecutive + * slots to hold the entry. + */ + slot = afs_find_contig_bits(block, need_slots); + if (slot >= 0) { + _debug("slot %u", slot); + goto found_space; + } + } + + if (page != page0) { + unlock_page(page); + kunmap(page); + put_page(page); + } + } + + /* There are no spare slots of sufficient size, yet the operation + * succeeded. Download the directory again. + */ + trace_afs_edit_dir(vnode, why, afs_edit_dir_create_nospc, 0, 0, 0, 0, name->name); + clear_bit(AFS_VNODE_DIR_VALID, &vnode->flags); + goto out_unmap; + +new_directory: + afs_edit_init_block(meta, meta, 0); + i_size = AFS_DIR_BLOCK_SIZE; + i_size_write(&vnode->vfs_inode, i_size); + slot = AFS_DIR_RESV_BLOCKS0; + page = page0; + block = meta; + nr_blocks = 1; + b = 0; + +found_space: + /* Set the dirent slot. */ + trace_afs_edit_dir(vnode, why, afs_edit_dir_create, b, slot, + new_fid->vnode, new_fid->unique, name->name); + de = &block->dirents[slot]; + de->u.valid = 1; + de->u.unused[0] = 0; + de->u.hash_next = 0; // TODO: Really need to maintain this + de->u.vnode = htonl(new_fid->vnode); + de->u.unique = htonl(new_fid->unique); + memcpy(de->u.name, name->name, name->len + 1); + de->u.name[name->len] = 0; + + /* Adjust the bitmap. */ + afs_set_contig_bits(block, slot, need_slots); + if (page != page0) { + unlock_page(page); + kunmap(page); + put_page(page); + } + + /* Adjust the allocation counter. */ + if (b < AFS_DIR_BLOCKS_WITH_CTR) + meta->meta.alloc_ctrs[b] -= need_slots; + + inode_inc_iversion_raw(&vnode->vfs_inode); + afs_stat_v(vnode, n_dir_cr); + _debug("Insert %s in %u[%u]", name->name, b, slot); + +out_unmap: + unlock_page(page0); + kunmap(page0); + put_page(page0); + _leave(""); + return; + +invalidated: + trace_afs_edit_dir(vnode, why, afs_edit_dir_create_inval, 0, 0, 0, 0, name->name); + clear_bit(AFS_VNODE_DIR_VALID, &vnode->flags); + if (page != page0) { + kunmap(page); + put_page(page); + } + goto out_unmap; + +error: + trace_afs_edit_dir(vnode, why, afs_edit_dir_create_error, 0, 0, 0, 0, name->name); + clear_bit(AFS_VNODE_DIR_VALID, &vnode->flags); + goto out_unmap; +} + +/* + * Edit a directory's file data to remove a new directory entry. Doing this + * after unlink, rmdir or rename if the data version number is incremented by + * exactly one avoids the need to re-download the entire directory contents. + * + * The caller must hold the inode locked. + */ +void afs_edit_dir_remove(struct afs_vnode *vnode, + struct qstr *name, enum afs_edit_dir_reason why) +{ + struct afs_xdr_dir_page *meta_page, *dir_page; + union afs_xdr_dir_block *meta, *block; + union afs_xdr_dirent *de; + struct page *page0, *page; + unsigned int need_slots, nr_blocks, b; + pgoff_t index; + loff_t i_size; + int slot; + + _enter(",,{%d,%s},", name->len, name->name); + + i_size = i_size_read(&vnode->vfs_inode); + if (i_size < AFS_DIR_BLOCK_SIZE || + i_size > AFS_DIR_BLOCK_SIZE * AFS_DIR_MAX_BLOCKS || + (i_size & (AFS_DIR_BLOCK_SIZE - 1))) { + clear_bit(AFS_VNODE_DIR_VALID, &vnode->flags); + return; + } + nr_blocks = i_size / AFS_DIR_BLOCK_SIZE; + + page0 = find_lock_page(vnode->vfs_inode.i_mapping, 0); + if (!page0) { + clear_bit(AFS_VNODE_DIR_VALID, &vnode->flags); + _leave(" [fgp]"); + return; + } + + /* Work out how many slots we're going to discard. */ + need_slots = round_up(12 + name->len + 1 + 4, AFS_DIR_DIRENT_SIZE); + need_slots /= AFS_DIR_DIRENT_SIZE; + + meta_page = kmap(page0); + meta = &meta_page->blocks[0]; + + /* Find a page that has sufficient slots available. Each VM page + * contains two or more directory blocks. + */ + for (b = 0; b < nr_blocks; b++) { + index = b / AFS_DIR_BLOCKS_PER_PAGE; + if (index != 0) { + page = find_lock_page(vnode->vfs_inode.i_mapping, index); + if (!page) + goto error; + dir_page = kmap(page); + } else { + page = page0; + dir_page = meta_page; + } + + /* Abandon the edit if we got a callback break. */ + if (!test_bit(AFS_VNODE_DIR_VALID, &vnode->flags)) + goto invalidated; + + block = &dir_page->blocks[b % AFS_DIR_BLOCKS_PER_PAGE]; + + if (b > AFS_DIR_BLOCKS_WITH_CTR || + meta->meta.alloc_ctrs[b] <= AFS_DIR_SLOTS_PER_BLOCK - 1 - need_slots) { + slot = afs_dir_scan_block(block, name, b); + if (slot >= 0) + goto found_dirent; + } + + if (page != page0) { + unlock_page(page); + kunmap(page); + put_page(page); + } + } + + /* Didn't find the dirent to clobber. Download the directory again. */ + trace_afs_edit_dir(vnode, why, afs_edit_dir_delete_noent, + 0, 0, 0, 0, name->name); + clear_bit(AFS_VNODE_DIR_VALID, &vnode->flags); + goto out_unmap; + +found_dirent: + de = &block->dirents[slot]; + + trace_afs_edit_dir(vnode, why, afs_edit_dir_delete, b, slot, + ntohl(de->u.vnode), ntohl(de->u.unique), + name->name); + + memset(de, 0, sizeof(*de) * need_slots); + + /* Adjust the bitmap. */ + afs_clear_contig_bits(block, slot, need_slots); + if (page != page0) { + unlock_page(page); + kunmap(page); + put_page(page); + } + + /* Adjust the allocation counter. */ + if (b < AFS_DIR_BLOCKS_WITH_CTR) + meta->meta.alloc_ctrs[b] += need_slots; + + inode_set_iversion_raw(&vnode->vfs_inode, vnode->status.data_version); + afs_stat_v(vnode, n_dir_rm); + _debug("Remove %s from %u[%u]", name->name, b, slot); + +out_unmap: + unlock_page(page0); + kunmap(page0); + put_page(page0); + _leave(""); + return; + +invalidated: + trace_afs_edit_dir(vnode, why, afs_edit_dir_delete_inval, + 0, 0, 0, 0, name->name); + clear_bit(AFS_VNODE_DIR_VALID, &vnode->flags); + if (page != page0) { + unlock_page(page); + kunmap(page); + put_page(page); + } + goto out_unmap; + +error: + trace_afs_edit_dir(vnode, why, afs_edit_dir_delete_error, + 0, 0, 0, 0, name->name); + clear_bit(AFS_VNODE_DIR_VALID, &vnode->flags); + goto out_unmap; +} diff --git a/fs/afs/fsclient.c b/fs/afs/fsclient.c index b66ff0dc8a5a..20d6304a0d3e 100644 --- a/fs/afs/fsclient.c +++ b/fs/afs/fsclient.c @@ -107,6 +107,13 @@ void afs_update_inode_from_status(struct afs_vnode *vnode, } else { set_bit(AFS_VNODE_ZAP_DATA, &vnode->flags); } + } else if (vnode->status.type == AFS_FTYPE_DIR) { + /* Expected directory change is handled elsewhere so + * that we can locally edit the directory and save on a + * download. + */ + if (test_bit(AFS_VNODE_DIR_VALID, &vnode->flags)) + flags &= ~AFS_VNODE_DATA_CHANGED; } } @@ -190,10 +197,7 @@ static int xdr_decode_AFSFetchStatus(const __be32 **_bp, size = (u64)ntohl(xdr->size_lo); size |= (u64)ntohl(xdr->size_hi) << 32; - if (size != status->size) { - status->size = size; - flags |= AFS_VNODE_DATA_CHANGED; - } + status->size = size; data_version = (u64)ntohl(xdr->data_version_lo); data_version |= (u64)ntohl(xdr->data_version_hi) << 32; @@ -736,6 +740,7 @@ static const struct afs_call_type afs_RXFSMakeDir = { int afs_fs_create(struct afs_fs_cursor *fc, const char *name, umode_t mode, + u64 current_data_version, struct afs_fid *newfid, struct afs_file_status *newstatus, struct afs_callback *newcb) @@ -763,7 +768,7 @@ int afs_fs_create(struct afs_fs_cursor *fc, call->reply[1] = newfid; call->reply[2] = newstatus; call->reply[3] = newcb; - call->expected_version = vnode->status.data_version; + call->expected_version = current_data_version + 1; /* marshall the parameters */ bp = call->request; @@ -836,7 +841,8 @@ static const struct afs_call_type afs_RXFSRemoveDir = { /* * remove a file or directory */ -int afs_fs_remove(struct afs_fs_cursor *fc, const char *name, bool isdir) +int afs_fs_remove(struct afs_fs_cursor *fc, const char *name, bool isdir, + u64 current_data_version) { struct afs_vnode *vnode = fc->vnode; struct afs_call *call; @@ -858,7 +864,7 @@ int afs_fs_remove(struct afs_fs_cursor *fc, const char *name, bool isdir) call->key = fc->key; call->reply[0] = vnode; - call->expected_version = vnode->status.data_version; + call->expected_version = current_data_version + 1; /* marshall the parameters */ bp = call->request; @@ -920,7 +926,7 @@ static const struct afs_call_type afs_RXFSLink = { * make a hard link */ int afs_fs_link(struct afs_fs_cursor *fc, struct afs_vnode *vnode, - const char *name) + const char *name, u64 current_data_version) { struct afs_vnode *dvnode = fc->vnode; struct afs_call *call; @@ -941,7 +947,7 @@ int afs_fs_link(struct afs_fs_cursor *fc, struct afs_vnode *vnode, call->key = fc->key; call->reply[0] = dvnode; call->reply[1] = vnode; - call->expected_version = vnode->status.data_version; + call->expected_version = current_data_version + 1; /* marshall the parameters */ bp = call->request; @@ -1009,6 +1015,7 @@ static const struct afs_call_type afs_RXFSSymlink = { int afs_fs_symlink(struct afs_fs_cursor *fc, const char *name, const char *contents, + u64 current_data_version, struct afs_fid *newfid, struct afs_file_status *newstatus) { @@ -1037,7 +1044,7 @@ int afs_fs_symlink(struct afs_fs_cursor *fc, call->reply[0] = vnode; call->reply[1] = newfid; call->reply[2] = newstatus; - call->expected_version = vnode->status.data_version; + call->expected_version = current_data_version + 1; /* marshall the parameters */ bp = call->request; @@ -1117,7 +1124,9 @@ static const struct afs_call_type afs_RXFSRename = { int afs_fs_rename(struct afs_fs_cursor *fc, const char *orig_name, struct afs_vnode *new_dvnode, - const char *new_name) + const char *new_name, + u64 current_orig_data_version, + u64 current_new_data_version) { struct afs_vnode *orig_dvnode = fc->vnode; struct afs_call *call; @@ -1145,8 +1154,8 @@ int afs_fs_rename(struct afs_fs_cursor *fc, call->key = fc->key; call->reply[0] = orig_dvnode; call->reply[1] = new_dvnode; - call->expected_version = orig_dvnode->status.data_version; - call->expected_version_2 = new_dvnode->status.data_version; + call->expected_version = current_orig_data_version + 1; + call->expected_version_2 = current_new_data_version + 1; /* marshall the parameters */ bp = call->request; diff --git a/fs/afs/inode.c b/fs/afs/inode.c index 69bcfb82dd69..3808dcbbac6d 100644 --- a/fs/afs/inode.c +++ b/fs/afs/inode.c @@ -395,8 +395,11 @@ int afs_validate(struct afs_vnode *vnode, struct key *key) if (test_bit(AFS_VNODE_CB_PROMISED, &vnode->flags)) { if (vnode->cb_s_break != vnode->cb_interest->server->cb_s_break) { vnode->cb_s_break = vnode->cb_interest->server->cb_s_break; - } else if (test_bit(AFS_VNODE_DIR_VALID, &vnode->flags) && - !test_bit(AFS_VNODE_ZAP_DATA, &vnode->flags) && + } else if (vnode->status.type == AFS_FTYPE_DIR && + test_bit(AFS_VNODE_DIR_VALID, &vnode->flags) && + vnode->cb_expires_at - 10 > now) { + valid = true; + } else if (!test_bit(AFS_VNODE_ZAP_DATA, &vnode->flags) && vnode->cb_expires_at - 10 > now) { valid = true; } diff --git a/fs/afs/internal.h b/fs/afs/internal.h index 86f3066d9ab0..703ddb8c5d57 100644 --- a/fs/afs/internal.h +++ b/fs/afs/internal.h @@ -271,6 +271,8 @@ struct afs_net { atomic_t n_inval; /* Number of invalidations by the server */ atomic_t n_relpg; /* Number of invalidations by releasepage */ atomic_t n_read_dir; /* Number of directory pages read */ + atomic_t n_dir_cr; /* Number of directory entry creation edits */ + atomic_t n_dir_rm; /* Number of directory entry removal edits */ }; extern const char afs_init_sysname[]; @@ -679,6 +681,13 @@ extern const struct dentry_operations afs_fs_dentry_operations; extern void afs_d_release(struct dentry *); +/* + * dir_edit.c + */ +extern void afs_edit_dir_add(struct afs_vnode *, struct qstr *, struct afs_fid *, + enum afs_edit_dir_reason); +extern void afs_edit_dir_remove(struct afs_vnode *, struct qstr *, enum afs_edit_dir_reason); + /* * dynroot.c */ @@ -725,14 +734,14 @@ extern void afs_update_inode_from_status(struct afs_vnode *, struct afs_file_sta extern int afs_fs_fetch_file_status(struct afs_fs_cursor *, struct afs_volsync *, bool); extern int afs_fs_give_up_callbacks(struct afs_net *, struct afs_server *); extern int afs_fs_fetch_data(struct afs_fs_cursor *, struct afs_read *); -extern int afs_fs_create(struct afs_fs_cursor *, const char *, umode_t, +extern int afs_fs_create(struct afs_fs_cursor *, const char *, umode_t, u64, struct afs_fid *, struct afs_file_status *, struct afs_callback *); -extern int afs_fs_remove(struct afs_fs_cursor *, const char *, bool); -extern int afs_fs_link(struct afs_fs_cursor *, struct afs_vnode *, const char *); -extern int afs_fs_symlink(struct afs_fs_cursor *, const char *, const char *, +extern int afs_fs_remove(struct afs_fs_cursor *, const char *, bool, u64); +extern int afs_fs_link(struct afs_fs_cursor *, struct afs_vnode *, const char *, u64); +extern int afs_fs_symlink(struct afs_fs_cursor *, const char *, const char *, u64, struct afs_fid *, struct afs_file_status *); extern int afs_fs_rename(struct afs_fs_cursor *, const char *, - struct afs_vnode *, const char *); + struct afs_vnode *, const char *, u64, u64); extern int afs_fs_store_data(struct afs_fs_cursor *, struct address_space *, pgoff_t, pgoff_t, unsigned, unsigned); extern int afs_fs_setattr(struct afs_fs_cursor *, struct iattr *); diff --git a/fs/afs/proc.c b/fs/afs/proc.c index 6f5a000f44a7..3212bce0d4fb 100644 --- a/fs/afs/proc.c +++ b/fs/afs/proc.c @@ -918,6 +918,10 @@ static int afs_proc_stats_show(struct seq_file *m, void *v) seq_printf(m, "dir-data: rdpg=%u\n", atomic_read(&net->n_read_dir)); + + seq_printf(m, "dir-edit: cr=%u rm=%u\n", + atomic_read(&net->n_dir_cr), + atomic_read(&net->n_dir_rm)); return 0; } diff --git a/include/trace/events/afs.h b/include/trace/events/afs.h index 0419b7e1e968..f46dee0f5ced 100644 --- a/include/trace/events/afs.h +++ b/include/trace/events/afs.h @@ -63,6 +63,27 @@ enum afs_vl_operation { afs_VL_GetCapabilities = 65537, /* AFS Get VL server capabilities */ }; +enum afs_edit_dir_op { + afs_edit_dir_create, + afs_edit_dir_create_error, + afs_edit_dir_create_inval, + afs_edit_dir_create_nospc, + afs_edit_dir_delete, + afs_edit_dir_delete_error, + afs_edit_dir_delete_inval, + afs_edit_dir_delete_noent, +}; + +enum afs_edit_dir_reason { + afs_edit_dir_for_create, + afs_edit_dir_for_link, + afs_edit_dir_for_mkdir, + afs_edit_dir_for_rename, + afs_edit_dir_for_rmdir, + afs_edit_dir_for_symlink, + afs_edit_dir_for_unlink, +}; + #endif /* end __AFS_DECLARE_TRACE_ENUMS_ONCE_ONLY */ /* @@ -106,6 +127,25 @@ enum afs_vl_operation { EM(afs_YFSVL_GetEndpoints, "YFSVL.GetEndpoints") \ E_(afs_VL_GetCapabilities, "VL.GetCapabilities") +#define afs_edit_dir_ops \ + EM(afs_edit_dir_create, "create") \ + EM(afs_edit_dir_create_error, "c_fail") \ + EM(afs_edit_dir_create_inval, "c_invl") \ + EM(afs_edit_dir_create_nospc, "c_nspc") \ + EM(afs_edit_dir_delete, "delete") \ + EM(afs_edit_dir_delete_error, "d_err ") \ + EM(afs_edit_dir_delete_inval, "d_invl") \ + E_(afs_edit_dir_delete_noent, "d_nent") + +#define afs_edit_dir_reasons \ + EM(afs_edit_dir_for_create, "Create") \ + EM(afs_edit_dir_for_link, "Link ") \ + EM(afs_edit_dir_for_mkdir, "MkDir ") \ + EM(afs_edit_dir_for_rename, "Rename") \ + EM(afs_edit_dir_for_rmdir, "RmDir ") \ + EM(afs_edit_dir_for_symlink, "Symlnk") \ + E_(afs_edit_dir_for_unlink, "Unlink") + /* * Export enum symbols via userspace. @@ -118,6 +158,8 @@ enum afs_vl_operation { afs_call_traces; afs_fs_operations; afs_vl_operations; +afs_edit_dir_ops; +afs_edit_dir_reasons; /* * Now redefine the EM() and E_() macros to map the enums to the strings that @@ -464,6 +506,54 @@ TRACE_EVENT(afs_call_state, __entry->ret, __entry->abort) ); +TRACE_EVENT(afs_edit_dir, + TP_PROTO(struct afs_vnode *dvnode, + enum afs_edit_dir_reason why, + enum afs_edit_dir_op op, + unsigned int block, + unsigned int slot, + unsigned int f_vnode, + unsigned int f_unique, + const char *name), + + TP_ARGS(dvnode, why, op, block, slot, f_vnode, f_unique, name), + + TP_STRUCT__entry( + __field(unsigned int, vnode ) + __field(unsigned int, unique ) + __field(enum afs_edit_dir_reason, why ) + __field(enum afs_edit_dir_op, op ) + __field(unsigned int, block ) + __field(unsigned short, slot ) + __field(unsigned int, f_vnode ) + __field(unsigned int, f_unique ) + __array(char, name, 18 ) + ), + + TP_fast_assign( + int __len = strlen(name); + __len = min(__len, 17); + __entry->vnode = dvnode->fid.vnode; + __entry->unique = dvnode->fid.unique; + __entry->why = why; + __entry->op = op; + __entry->block = block; + __entry->slot = slot; + __entry->f_vnode = f_vnode; + __entry->f_unique = f_unique; + memcpy(__entry->name, name, __len); + __entry->name[__len] = 0; + ), + + TP_printk("d=%x:%x %s %s %u[%u] f=%x:%x %s", + __entry->vnode, __entry->unique, + __print_symbolic(__entry->why, afs_edit_dir_reasons), + __print_symbolic(__entry->op, afs_edit_dir_ops), + __entry->block, __entry->slot, + __entry->f_vnode, __entry->f_unique, + __entry->name) + ); + #endif /* _TRACE_AFS_H */ /* This part must be outside protection */ -- cgit v1.2.3 From 5f702c8e124f967146a735a19f0b00a2469487d1 Mon Sep 17 00:00:00 2001 From: David Howells Date: Fri, 6 Apr 2018 14:17:25 +0100 Subject: afs: Trace protocol errors Trace protocol errors detected in afs. Signed-off-by: David Howells --- fs/afs/cmservice.c | 4 +-- fs/afs/fsclient.c | 68 ++++++++++++++++++++++++---------------------- fs/afs/inode.c | 2 +- fs/afs/internal.h | 1 + fs/afs/rxrpc.c | 9 ++++++ fs/afs/vlclient.c | 20 +++++++------- include/trace/events/afs.h | 21 ++++++++++++++ 7 files changed, 79 insertions(+), 46 deletions(-) (limited to 'include/trace') diff --git a/fs/afs/cmservice.c b/fs/afs/cmservice.c index fa07f83e9f29..357de908df3a 100644 --- a/fs/afs/cmservice.c +++ b/fs/afs/cmservice.c @@ -201,7 +201,7 @@ static int afs_deliver_cb_callback(struct afs_call *call) call->count = ntohl(call->tmp); _debug("FID count: %u", call->count); if (call->count > AFSCBMAX) - return -EBADMSG; + return afs_protocol_error(call, -EBADMSG); call->buffer = kmalloc(call->count * 3 * 4, GFP_KERNEL); if (!call->buffer) @@ -245,7 +245,7 @@ static int afs_deliver_cb_callback(struct afs_call *call) call->count2 = ntohl(call->tmp); _debug("CB count: %u", call->count2); if (call->count2 != call->count && call->count2 != 0) - return -EBADMSG; + return afs_protocol_error(call, -EBADMSG); call->offset = 0; call->unmarshall++; diff --git a/fs/afs/fsclient.c b/fs/afs/fsclient.c index 20d6304a0d3e..efacdb7c1dee 100644 --- a/fs/afs/fsclient.c +++ b/fs/afs/fsclient.c @@ -126,7 +126,8 @@ void afs_update_inode_from_status(struct afs_vnode *vnode, /* * decode an AFSFetchStatus block */ -static int xdr_decode_AFSFetchStatus(const __be32 **_bp, +static int xdr_decode_AFSFetchStatus(struct afs_call *call, + const __be32 **_bp, struct afs_file_status *status, struct afs_vnode *vnode, const afs_dataversion_t *expected_version, @@ -167,6 +168,7 @@ static int xdr_decode_AFSFetchStatus(const __be32 **_bp, case AFS_FTYPE_INVALID: if (abort_code != 0) { status->abort_code = abort_code; + ret = 0; goto out; } /* Fall through */ @@ -229,7 +231,7 @@ out: bad: xdr_dump_bad(*_bp); - ret = -EBADMSG; + ret = afs_protocol_error(call, -EBADMSG); goto out; } @@ -372,9 +374,9 @@ static int afs_deliver_fs_fetch_status_vnode(struct afs_call *call) /* unmarshall the reply once we've received all of it */ bp = call->buffer; - if (xdr_decode_AFSFetchStatus(&bp, &vnode->status, vnode, + if (xdr_decode_AFSFetchStatus(call, &bp, &vnode->status, vnode, &call->expected_version, NULL) < 0) - return -EBADMSG; + return afs_protocol_error(call, -EBADMSG); xdr_decode_AFSCallBack(call, vnode, &bp); if (call->reply[1]) xdr_decode_AFSVolSync(&bp, call->reply[1]); @@ -553,9 +555,9 @@ static int afs_deliver_fs_fetch_data(struct afs_call *call) return ret; bp = call->buffer; - if (xdr_decode_AFSFetchStatus(&bp, &vnode->status, vnode, + if (xdr_decode_AFSFetchStatus(call, &bp, &vnode->status, vnode, &vnode->status.data_version, req) < 0) - return -EBADMSG; + return afs_protocol_error(call, -EBADMSG); xdr_decode_AFSCallBack(call, vnode, &bp); if (call->reply[1]) xdr_decode_AFSVolSync(&bp, call->reply[1]); @@ -706,10 +708,10 @@ static int afs_deliver_fs_create_vnode(struct afs_call *call) /* unmarshall the reply once we've received all of it */ bp = call->buffer; xdr_decode_AFSFid(&bp, call->reply[1]); - if (xdr_decode_AFSFetchStatus(&bp, call->reply[2], NULL, NULL, NULL) < 0 || - xdr_decode_AFSFetchStatus(&bp, &vnode->status, vnode, + if (xdr_decode_AFSFetchStatus(call, &bp, call->reply[2], NULL, NULL, NULL) < 0 || + xdr_decode_AFSFetchStatus(call, &bp, &vnode->status, vnode, &call->expected_version, NULL) < 0) - return -EBADMSG; + return afs_protocol_error(call, -EBADMSG); xdr_decode_AFSCallBack_raw(&bp, call->reply[3]); /* xdr_decode_AFSVolSync(&bp, call->reply[X]); */ @@ -812,9 +814,9 @@ static int afs_deliver_fs_remove(struct afs_call *call) /* unmarshall the reply once we've received all of it */ bp = call->buffer; - if (xdr_decode_AFSFetchStatus(&bp, &vnode->status, vnode, + if (xdr_decode_AFSFetchStatus(call, &bp, &vnode->status, vnode, &call->expected_version, NULL) < 0) - return -EBADMSG; + return afs_protocol_error(call, -EBADMSG); /* xdr_decode_AFSVolSync(&bp, call->reply[X]); */ _leave(" = 0 [done]"); @@ -902,10 +904,10 @@ static int afs_deliver_fs_link(struct afs_call *call) /* unmarshall the reply once we've received all of it */ bp = call->buffer; - if (xdr_decode_AFSFetchStatus(&bp, &vnode->status, vnode, NULL, NULL) < 0 || - xdr_decode_AFSFetchStatus(&bp, &dvnode->status, dvnode, + if (xdr_decode_AFSFetchStatus(call, &bp, &vnode->status, vnode, NULL, NULL) < 0 || + xdr_decode_AFSFetchStatus(call, &bp, &dvnode->status, dvnode, &call->expected_version, NULL) < 0) - return -EBADMSG; + return afs_protocol_error(call, -EBADMSG); /* xdr_decode_AFSVolSync(&bp, call->reply[X]); */ _leave(" = 0 [done]"); @@ -989,10 +991,10 @@ static int afs_deliver_fs_symlink(struct afs_call *call) /* unmarshall the reply once we've received all of it */ bp = call->buffer; xdr_decode_AFSFid(&bp, call->reply[1]); - if (xdr_decode_AFSFetchStatus(&bp, call->reply[2], NULL, NULL, NULL) || - xdr_decode_AFSFetchStatus(&bp, &vnode->status, vnode, + if (xdr_decode_AFSFetchStatus(call, &bp, call->reply[2], NULL, NULL, NULL) || + xdr_decode_AFSFetchStatus(call, &bp, &vnode->status, vnode, &call->expected_version, NULL) < 0) - return -EBADMSG; + return afs_protocol_error(call, -EBADMSG); /* xdr_decode_AFSVolSync(&bp, call->reply[X]); */ _leave(" = 0 [done]"); @@ -1095,13 +1097,13 @@ static int afs_deliver_fs_rename(struct afs_call *call) /* unmarshall the reply once we've received all of it */ bp = call->buffer; - if (xdr_decode_AFSFetchStatus(&bp, &orig_dvnode->status, orig_dvnode, + if (xdr_decode_AFSFetchStatus(call, &bp, &orig_dvnode->status, orig_dvnode, &call->expected_version, NULL) < 0) - return -EBADMSG; + return afs_protocol_error(call, -EBADMSG); if (new_dvnode != orig_dvnode && - xdr_decode_AFSFetchStatus(&bp, &new_dvnode->status, new_dvnode, + xdr_decode_AFSFetchStatus(call, &bp, &new_dvnode->status, new_dvnode, &call->expected_version_2, NULL) < 0) - return -EBADMSG; + return afs_protocol_error(call, -EBADMSG); /* xdr_decode_AFSVolSync(&bp, call->reply[X]); */ _leave(" = 0 [done]"); @@ -1204,9 +1206,9 @@ static int afs_deliver_fs_store_data(struct afs_call *call) /* unmarshall the reply once we've received all of it */ bp = call->buffer; - if (xdr_decode_AFSFetchStatus(&bp, &vnode->status, vnode, + if (xdr_decode_AFSFetchStatus(call, &bp, &vnode->status, vnode, &call->expected_version, NULL) < 0) - return -EBADMSG; + return afs_protocol_error(call, -EBADMSG); /* xdr_decode_AFSVolSync(&bp, call->reply[X]); */ afs_pages_written_back(vnode, call); @@ -1380,9 +1382,9 @@ static int afs_deliver_fs_store_status(struct afs_call *call) /* unmarshall the reply once we've received all of it */ bp = call->buffer; - if (xdr_decode_AFSFetchStatus(&bp, &vnode->status, vnode, + if (xdr_decode_AFSFetchStatus(call, &bp, &vnode->status, vnode, &call->expected_version, NULL) < 0) - return -EBADMSG; + return afs_protocol_error(call, -EBADMSG); /* xdr_decode_AFSVolSync(&bp, call->reply[X]); */ _leave(" = 0 [done]"); @@ -1585,7 +1587,7 @@ static int afs_deliver_fs_get_volume_status(struct afs_call *call) call->count = ntohl(call->tmp); _debug("volname length: %u", call->count); if (call->count >= AFSNAMEMAX) - return -EBADMSG; + return afs_protocol_error(call, -EBADMSG); call->offset = 0; call->unmarshall++; @@ -1632,7 +1634,7 @@ static int afs_deliver_fs_get_volume_status(struct afs_call *call) call->count = ntohl(call->tmp); _debug("offline msg length: %u", call->count); if (call->count >= AFSNAMEMAX) - return -EBADMSG; + return afs_protocol_error(call, -EBADMSG); call->offset = 0; call->unmarshall++; @@ -1679,7 +1681,7 @@ static int afs_deliver_fs_get_volume_status(struct afs_call *call) call->count = ntohl(call->tmp); _debug("motd length: %u", call->count); if (call->count >= AFSNAMEMAX) - return -EBADMSG; + return afs_protocol_error(call, -EBADMSG); call->offset = 0; call->unmarshall++; @@ -2082,7 +2084,7 @@ static int afs_deliver_fs_fetch_status(struct afs_call *call) /* unmarshall the reply once we've received all of it */ bp = call->buffer; - xdr_decode_AFSFetchStatus(&bp, status, vnode, + xdr_decode_AFSFetchStatus(call, &bp, status, vnode, &call->expected_version, NULL); callback[call->count].version = ntohl(bp[0]); callback[call->count].expiry = ntohl(bp[1]); @@ -2179,7 +2181,7 @@ static int afs_deliver_fs_inline_bulk_status(struct afs_call *call) tmp = ntohl(call->tmp); _debug("status count: %u/%u", tmp, call->count2); if (tmp != call->count2) - return -EBADMSG; + return afs_protocol_error(call, -EBADMSG); call->count = 0; call->unmarshall++; @@ -2194,10 +2196,10 @@ static int afs_deliver_fs_inline_bulk_status(struct afs_call *call) bp = call->buffer; statuses = call->reply[1]; - if (xdr_decode_AFSFetchStatus(&bp, &statuses[call->count], + if (xdr_decode_AFSFetchStatus(call, &bp, &statuses[call->count], call->count == 0 ? vnode : NULL, NULL, NULL) < 0) - return -EBADMSG; + return afs_protocol_error(call, -EBADMSG); call->count++; if (call->count < call->count2) @@ -2217,7 +2219,7 @@ static int afs_deliver_fs_inline_bulk_status(struct afs_call *call) tmp = ntohl(call->tmp); _debug("CB count: %u", tmp); if (tmp != call->count2) - return -EBADMSG; + return afs_protocol_error(call, -EBADMSG); call->count = 0; call->unmarshall++; more_cbs: diff --git a/fs/afs/inode.c b/fs/afs/inode.c index 3808dcbbac6d..06194cfe9724 100644 --- a/fs/afs/inode.c +++ b/fs/afs/inode.c @@ -82,7 +82,7 @@ static int afs_inode_init_from_status(struct afs_vnode *vnode, struct key *key) default: printk("kAFS: AFS vnode with undefined type\n"); read_sequnlock_excl(&vnode->cb_lock); - return -EBADMSG; + return afs_protocol_error(NULL, -EBADMSG); } inode->i_blocks = 0; diff --git a/fs/afs/internal.h b/fs/afs/internal.h index 703ddb8c5d57..6ae023cbf00e 100644 --- a/fs/afs/internal.h +++ b/fs/afs/internal.h @@ -875,6 +875,7 @@ extern void afs_flat_call_destructor(struct afs_call *); extern void afs_send_empty_reply(struct afs_call *); extern void afs_send_simple_reply(struct afs_call *, const void *, size_t); extern int afs_extract_data(struct afs_call *, void *, size_t, bool); +extern int afs_protocol_error(struct afs_call *, int); static inline int afs_transfer_reply(struct afs_call *call) { diff --git a/fs/afs/rxrpc.c b/fs/afs/rxrpc.c index f7ae54b6a393..5c6263972ec9 100644 --- a/fs/afs/rxrpc.c +++ b/fs/afs/rxrpc.c @@ -926,3 +926,12 @@ int afs_extract_data(struct afs_call *call, void *buf, size_t count, afs_set_call_complete(call, ret, remote_abort); return ret; } + +/* + * Log protocol error production. + */ +noinline int afs_protocol_error(struct afs_call *call, int error) +{ + trace_afs_protocol_error(call, error, __builtin_return_address(0)); + return error; +} diff --git a/fs/afs/vlclient.c b/fs/afs/vlclient.c index f9d89795e41b..1ed7e2fd2f35 100644 --- a/fs/afs/vlclient.c +++ b/fs/afs/vlclient.c @@ -450,7 +450,7 @@ again: call->count2 = ntohl(*bp); /* Type or next count */ if (call->count > YFS_MAXENDPOINTS) - return -EBADMSG; + return afs_protocol_error(call, -EBADMSG); alist = afs_alloc_addrlist(call->count, FS_SERVICE, AFS_FS_PORT); if (!alist) @@ -474,7 +474,7 @@ again: size = sizeof(__be32) * (1 + 4 + 1); break; default: - return -EBADMSG; + return afs_protocol_error(call, -EBADMSG); } size += sizeof(__be32); @@ -487,18 +487,18 @@ again: switch (call->count2) { case YFS_ENDPOINT_IPV4: if (ntohl(bp[0]) != sizeof(__be32) * 2) - return -EBADMSG; + return afs_protocol_error(call, -EBADMSG); afs_merge_fs_addr4(alist, bp[1], ntohl(bp[2])); bp += 3; break; case YFS_ENDPOINT_IPV6: if (ntohl(bp[0]) != sizeof(__be32) * 5) - return -EBADMSG; + return afs_protocol_error(call, -EBADMSG); afs_merge_fs_addr6(alist, bp + 1, ntohl(bp[5])); bp += 6; break; default: - return -EBADMSG; + return afs_protocol_error(call, -EBADMSG); } /* Got either the type of the next entry or the count of @@ -517,7 +517,7 @@ again: if (!call->count) goto end; if (call->count > YFS_MAXENDPOINTS) - return -EBADMSG; + return afs_protocol_error(call, -EBADMSG); call->unmarshall = 3; @@ -545,7 +545,7 @@ again: size = sizeof(__be32) * (1 + 4 + 1); break; default: - return -EBADMSG; + return afs_protocol_error(call, -EBADMSG); } if (call->count > 1) @@ -558,16 +558,16 @@ again: switch (call->count2) { case YFS_ENDPOINT_IPV4: if (ntohl(bp[0]) != sizeof(__be32) * 2) - return -EBADMSG; + return afs_protocol_error(call, -EBADMSG); bp += 3; break; case YFS_ENDPOINT_IPV6: if (ntohl(bp[0]) != sizeof(__be32) * 5) - return -EBADMSG; + return afs_protocol_error(call, -EBADMSG); bp += 6; break; default: - return -EBADMSG; + return afs_protocol_error(call, -EBADMSG); } /* Got either the type of the next entry or the count of diff --git a/include/trace/events/afs.h b/include/trace/events/afs.h index f46dee0f5ced..f0820554caa9 100644 --- a/include/trace/events/afs.h +++ b/include/trace/events/afs.h @@ -554,6 +554,27 @@ TRACE_EVENT(afs_edit_dir, __entry->name) ); +TRACE_EVENT(afs_protocol_error, + TP_PROTO(struct afs_call *call, int error, const void *where), + + TP_ARGS(call, error, where), + + TP_STRUCT__entry( + __field(unsigned int, call ) + __field(int, error ) + __field(const void *, where ) + ), + + TP_fast_assign( + __entry->call = call ? call->debug_id : 0; + __entry->error = error; + __entry->where = where; + ), + + TP_printk("c=%08x r=%d sp=%pSR", + __entry->call, __entry->error, __entry->where) + ); + #endif /* _TRACE_AFS_H */ /* This part must be outside protection */ -- cgit v1.2.3 From e671edb9428c8a61662aaf8c39f5edced7cc45c7 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Fri, 16 Mar 2018 10:33:44 -0400 Subject: sunrpc: Simplify synopsis of some trace points Clean up: struct rpc_task carries a pointer to a struct rpc_clnt, and in fact task->tk_client is always what is passed into trace points that are already passing @task. Signed-off-by: Chuck Lever Signed-off-by: Anna Schumaker --- include/trace/events/sunrpc.h | 40 +++++++++++++++++++++------------------- net/sunrpc/clnt.c | 2 +- net/sunrpc/sched.c | 10 +++++----- 3 files changed, 27 insertions(+), 25 deletions(-) (limited to 'include/trace') diff --git a/include/trace/events/sunrpc.h b/include/trace/events/sunrpc.h index 970c91a83173..53bbaeac08dd 100644 --- a/include/trace/events/sunrpc.h +++ b/include/trace/events/sunrpc.h @@ -50,9 +50,9 @@ DEFINE_EVENT(rpc_task_status, rpc_bind_status, ); TRACE_EVENT(rpc_connect_status, - TP_PROTO(struct rpc_task *task, int status), + TP_PROTO(const struct rpc_task *task), - TP_ARGS(task, status), + TP_ARGS(task), TP_STRUCT__entry( __field(unsigned int, task_id) @@ -63,7 +63,7 @@ TRACE_EVENT(rpc_connect_status, TP_fast_assign( __entry->task_id = task->tk_pid; __entry->client_id = task->tk_client->cl_clid; - __entry->status = status; + __entry->status = task->tk_status; ), TP_printk("task:%u@%u status=%d", @@ -103,9 +103,9 @@ TRACE_EVENT(rpc_request, DECLARE_EVENT_CLASS(rpc_task_running, - TP_PROTO(const struct rpc_clnt *clnt, const struct rpc_task *task, const void *action), + TP_PROTO(const struct rpc_task *task, const void *action), - TP_ARGS(clnt, task, action), + TP_ARGS(task, action), TP_STRUCT__entry( __field(unsigned int, task_id) @@ -117,7 +117,8 @@ DECLARE_EVENT_CLASS(rpc_task_running, ), TP_fast_assign( - __entry->client_id = clnt ? clnt->cl_clid : -1; + __entry->client_id = task->tk_client ? + task->tk_client->cl_clid : -1; __entry->task_id = task->tk_pid; __entry->action = action; __entry->runstate = task->tk_runstate; @@ -136,33 +137,33 @@ DECLARE_EVENT_CLASS(rpc_task_running, DEFINE_EVENT(rpc_task_running, rpc_task_begin, - TP_PROTO(const struct rpc_clnt *clnt, const struct rpc_task *task, const void *action), + TP_PROTO(const struct rpc_task *task, const void *action), - TP_ARGS(clnt, task, action) + TP_ARGS(task, action) ); DEFINE_EVENT(rpc_task_running, rpc_task_run_action, - TP_PROTO(const struct rpc_clnt *clnt, const struct rpc_task *task, const void *action), + TP_PROTO(const struct rpc_task *task, const void *action), - TP_ARGS(clnt, task, action) + TP_ARGS(task, action) ); DEFINE_EVENT(rpc_task_running, rpc_task_complete, - TP_PROTO(const struct rpc_clnt *clnt, const struct rpc_task *task, const void *action), + TP_PROTO(const struct rpc_task *task, const void *action), - TP_ARGS(clnt, task, action) + TP_ARGS(task, action) ); DECLARE_EVENT_CLASS(rpc_task_queued, - TP_PROTO(const struct rpc_clnt *clnt, const struct rpc_task *task, const struct rpc_wait_queue *q), + TP_PROTO(const struct rpc_task *task, const struct rpc_wait_queue *q), - TP_ARGS(clnt, task, q), + TP_ARGS(task, q), TP_STRUCT__entry( __field(unsigned int, task_id) @@ -175,7 +176,8 @@ DECLARE_EVENT_CLASS(rpc_task_queued, ), TP_fast_assign( - __entry->client_id = clnt ? clnt->cl_clid : -1; + __entry->client_id = task->tk_client ? + task->tk_client->cl_clid : -1; __entry->task_id = task->tk_pid; __entry->timeout = task->tk_timeout; __entry->runstate = task->tk_runstate; @@ -196,17 +198,17 @@ DECLARE_EVENT_CLASS(rpc_task_queued, DEFINE_EVENT(rpc_task_queued, rpc_task_sleep, - TP_PROTO(const struct rpc_clnt *clnt, const struct rpc_task *task, const struct rpc_wait_queue *q), + TP_PROTO(const struct rpc_task *task, const struct rpc_wait_queue *q), - TP_ARGS(clnt, task, q) + TP_ARGS(task, q) ); DEFINE_EVENT(rpc_task_queued, rpc_task_wakeup, - TP_PROTO(const struct rpc_clnt *clnt, const struct rpc_task *task, const struct rpc_wait_queue *q), + TP_PROTO(const struct rpc_task *task, const struct rpc_wait_queue *q), - TP_ARGS(clnt, task, q) + TP_ARGS(task, q) ); diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c index 6e432ecd7f99..079f06fe2c5d 100644 --- a/net/sunrpc/clnt.c +++ b/net/sunrpc/clnt.c @@ -1887,7 +1887,7 @@ call_connect_status(struct rpc_task *task) dprint_status(task); - trace_rpc_connect_status(task, status); + trace_rpc_connect_status(task); task->tk_status = 0; switch (status) { case -ECONNREFUSED: diff --git a/net/sunrpc/sched.c b/net/sunrpc/sched.c index d9db2eab3a8d..3fe5d60ab0e2 100644 --- a/net/sunrpc/sched.c +++ b/net/sunrpc/sched.c @@ -276,7 +276,7 @@ static void rpc_set_active(struct rpc_task *task) { rpc_task_set_debuginfo(task); set_bit(RPC_TASK_ACTIVE, &task->tk_runstate); - trace_rpc_task_begin(task->tk_client, task, NULL); + trace_rpc_task_begin(task, NULL); } /* @@ -291,7 +291,7 @@ static int rpc_complete_task(struct rpc_task *task) unsigned long flags; int ret; - trace_rpc_task_complete(task->tk_client, task, NULL); + trace_rpc_task_complete(task, NULL); spin_lock_irqsave(&wq->lock, flags); clear_bit(RPC_TASK_ACTIVE, &task->tk_runstate); @@ -358,7 +358,7 @@ static void __rpc_sleep_on_priority(struct rpc_wait_queue *q, dprintk("RPC: %5u sleep_on(queue \"%s\" time %lu)\n", task->tk_pid, rpc_qname(q), jiffies); - trace_rpc_task_sleep(task->tk_client, task, q); + trace_rpc_task_sleep(task, q); __rpc_add_wait_queue(q, task, queue_priority); @@ -428,7 +428,7 @@ static void __rpc_do_wake_up_task_on_wq(struct workqueue_struct *wq, return; } - trace_rpc_task_wakeup(task->tk_client, task, queue); + trace_rpc_task_wakeup(task, queue); __rpc_remove_wait_queue(queue, task); @@ -780,7 +780,7 @@ static void __rpc_execute(struct rpc_task *task) } if (!do_action) break; - trace_rpc_task_run_action(task->tk_client, task, do_action); + trace_rpc_task_run_action(task, do_action); do_action(task); /* -- cgit v1.2.3 From 40bf7eb304b5659991ed932c0cd5bee6a7c88f4f Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Fri, 16 Mar 2018 10:33:49 -0400 Subject: sunrpc: Add static trace point to report RPC latency stats Introduce a low-overhead mechanism to report information about latencies of individual RPCs. The goal is to enable user space to filter the trace record for latency outliers, or build histograms, etc. Signed-off-by: Chuck Lever Signed-off-by: Anna Schumaker --- include/trace/events/sunrpc.h | 45 +++++++++++++++++++++++++++++++++++++++++++ net/sunrpc/stats.c | 16 ++++++++++----- 2 files changed, 56 insertions(+), 5 deletions(-) (limited to 'include/trace') diff --git a/include/trace/events/sunrpc.h b/include/trace/events/sunrpc.h index 53bbaeac08dd..fda9e726f1cc 100644 --- a/include/trace/events/sunrpc.h +++ b/include/trace/events/sunrpc.h @@ -212,6 +212,51 @@ DEFINE_EVENT(rpc_task_queued, rpc_task_wakeup, ); +TRACE_EVENT(rpc_stats_latency, + + TP_PROTO( + const struct rpc_task *task, + ktime_t backlog, + ktime_t rtt, + ktime_t execute + ), + + TP_ARGS(task, backlog, rtt, execute), + + TP_STRUCT__entry( + __field(u32, xid) + __field(int, version) + __string(progname, task->tk_client->cl_program->name) + __string(procname, rpc_proc_name(task)) + __field(unsigned long, backlog) + __field(unsigned long, rtt) + __field(unsigned long, execute) + __string(addr, + task->tk_xprt->address_strings[RPC_DISPLAY_ADDR]) + __string(port, + task->tk_xprt->address_strings[RPC_DISPLAY_PORT]) + ), + + TP_fast_assign( + __entry->xid = be32_to_cpu(task->tk_rqstp->rq_xid); + __entry->version = task->tk_client->cl_vers; + __assign_str(progname, task->tk_client->cl_program->name) + __assign_str(procname, rpc_proc_name(task)) + __entry->backlog = ktime_to_us(backlog); + __entry->rtt = ktime_to_us(rtt); + __entry->execute = ktime_to_us(execute); + __assign_str(addr, + task->tk_xprt->address_strings[RPC_DISPLAY_ADDR]); + __assign_str(port, + task->tk_xprt->address_strings[RPC_DISPLAY_PORT]); + ), + + TP_printk("peer=[%s]:%s xid=0x%08x %sv%d %s backlog=%lu rtt=%lu execute=%lu", + __get_str(addr), __get_str(port), __entry->xid, + __get_str(progname), __entry->version, __get_str(procname), + __entry->backlog, __entry->rtt, __entry->execute) +); + /* * First define the enums in the below macros to be exported to userspace * via TRACE_DEFINE_ENUM(). diff --git a/net/sunrpc/stats.c b/net/sunrpc/stats.c index 1e671333c3d5..f68aa46c9dd7 100644 --- a/net/sunrpc/stats.c +++ b/net/sunrpc/stats.c @@ -24,6 +24,8 @@ #include #include +#include + #include "netns.h" #define RPCDBG_FACILITY RPCDBG_MISC @@ -148,7 +150,7 @@ void rpc_count_iostats_metrics(const struct rpc_task *task, struct rpc_iostats *op_metrics) { struct rpc_rqst *req = task->tk_rqstp; - ktime_t delta, now; + ktime_t backlog, execute, now; if (!op_metrics || !req) return; @@ -164,16 +166,20 @@ void rpc_count_iostats_metrics(const struct rpc_task *task, op_metrics->om_bytes_sent += req->rq_xmit_bytes_sent; op_metrics->om_bytes_recv += req->rq_reply_bytes_recvd; + backlog = 0; if (ktime_to_ns(req->rq_xtime)) { - delta = ktime_sub(req->rq_xtime, task->tk_start); - op_metrics->om_queue = ktime_add(op_metrics->om_queue, delta); + backlog = ktime_sub(req->rq_xtime, task->tk_start); + op_metrics->om_queue = ktime_add(op_metrics->om_queue, backlog); } + op_metrics->om_rtt = ktime_add(op_metrics->om_rtt, req->rq_rtt); - delta = ktime_sub(now, task->tk_start); - op_metrics->om_execute = ktime_add(op_metrics->om_execute, delta); + execute = ktime_sub(now, task->tk_start); + op_metrics->om_execute = ktime_add(op_metrics->om_execute, execute); spin_unlock(&op_metrics->om_lock); + + trace_rpc_stats_latency(req->rq_task, backlog, req->rq_rtt, execute); } EXPORT_SYMBOL_GPL(rpc_count_iostats_metrics); -- cgit v1.2.3 From a25a4cb3af177a4cf5621ffbf4fa89ae60c6d4d7 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Fri, 16 Mar 2018 10:33:55 -0400 Subject: sunrpc: Add static trace point to report result of RPC ping This information can help track down local misconfiguration issues as well as network partitions and unresponsive servers. There are several ways to send a ping, and with transport multi- plexing, the exact rpc_xprt that is used is sometimes not known by the upper layer. The rpc_xprt pointer passed to the trace point call also has to be RCU-safe. I found a spot inside the client FSM where an rpc_xprt pointer is always available and safe to use. Suggested-by: Bill Baker Signed-off-by: Chuck Lever Signed-off-by: Anna Schumaker --- include/trace/events/sunrpc.h | 21 +++++++++++++++++++++ net/sunrpc/clnt.c | 6 ++++++ 2 files changed, 27 insertions(+) (limited to 'include/trace') diff --git a/include/trace/events/sunrpc.h b/include/trace/events/sunrpc.h index fda9e726f1cc..76887d60f0c0 100644 --- a/include/trace/events/sunrpc.h +++ b/include/trace/events/sunrpc.h @@ -453,6 +453,27 @@ DEFINE_EVENT(rpc_xprt_event, xprt_complete_rqst, TP_PROTO(struct rpc_xprt *xprt, __be32 xid, int status), TP_ARGS(xprt, xid, status)); +TRACE_EVENT(xprt_ping, + TP_PROTO(const struct rpc_xprt *xprt, int status), + + TP_ARGS(xprt, status), + + TP_STRUCT__entry( + __field(int, status) + __string(addr, xprt->address_strings[RPC_DISPLAY_ADDR]) + __string(port, xprt->address_strings[RPC_DISPLAY_PORT]) + ), + + TP_fast_assign( + __entry->status = status; + __assign_str(addr, xprt->address_strings[RPC_DISPLAY_ADDR]); + __assign_str(port, xprt->address_strings[RPC_DISPLAY_PORT]); + ), + + TP_printk("peer=[%s]:%s status=%d", + __get_str(addr), __get_str(port), __entry->status) +); + TRACE_EVENT(xs_tcp_data_ready, TP_PROTO(struct rpc_xprt *xprt, int err, unsigned int total), diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c index 079f06fe2c5d..166f8c1680d1 100644 --- a/net/sunrpc/clnt.c +++ b/net/sunrpc/clnt.c @@ -2014,6 +2014,9 @@ call_transmit_status(struct rpc_task *task) case -EPERM: if (RPC_IS_SOFTCONN(task)) { xprt_end_transmit(task); + if (!task->tk_msg.rpc_proc->p_proc) + trace_xprt_ping(task->tk_xprt, + task->tk_status); rpc_exit(task, task->tk_status); break; } @@ -2112,6 +2115,9 @@ call_status(struct rpc_task *task) struct rpc_rqst *req = task->tk_rqstp; int status; + if (!task->tk_msg.rpc_proc->p_proc) + trace_xprt_ping(task->tk_xprt, task->tk_status); + if (req->rq_reply_bytes_recvd > 0 && !req->rq_bytes_sent) task->tk_status = req->rq_reply_bytes_recvd; -- cgit v1.2.3 From d51d1e64500fcb48fc6a18c77c965b8f48a175f2 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Tue, 10 Apr 2018 16:28:07 -0700 Subject: mm, vmscan, tracing: use pointer to reclaim_stat struct in trace event The trace event trace_mm_vmscan_lru_shrink_inactive() currently has 12 parameters! Seven of them are from the reclaim_stat structure. This structure is currently local to mm/vmscan.c. By moving it to the global vmstat.h header, we can also reference it from the vmscan tracepoints. In moving it, it brings down the overhead of passing so many arguments to the trace event. In the future, we may limit the number of arguments that a trace event may pass (ideally just 6, but more realistically it may be 8). Before this patch, the code to call the trace event is this: 0f 83 aa fe ff ff jae ffffffff811e6261 48 8b 45 a0 mov -0x60(%rbp),%rax 45 8b 64 24 20 mov 0x20(%r12),%r12d 44 8b 6d d4 mov -0x2c(%rbp),%r13d 8b 4d d0 mov -0x30(%rbp),%ecx 44 8b 75 cc mov -0x34(%rbp),%r14d 44 8b 7d c8 mov -0x38(%rbp),%r15d 48 89 45 90 mov %rax,-0x70(%rbp) 8b 83 b8 fe ff ff mov -0x148(%rbx),%eax 8b 55 c0 mov -0x40(%rbp),%edx 8b 7d c4 mov -0x3c(%rbp),%edi 8b 75 b8 mov -0x48(%rbp),%esi 89 45 80 mov %eax,-0x80(%rbp) 65 ff 05 e4 f7 e2 7e incl %gs:0x7ee2f7e4(%rip) # 15bd0 <__preempt_count> 48 8b 05 75 5b 13 01 mov 0x1135b75(%rip),%rax # ffffffff8231bf68 <__tracepoint_mm_vmscan_lru_shrink_inactive+0x28> 48 85 c0 test %rax,%rax 74 72 je ffffffff811e646a 48 89 c3 mov %rax,%rbx 4c 8b 10 mov (%rax),%r10 89 f8 mov %edi,%eax 48 89 85 68 ff ff ff mov %rax,-0x98(%rbp) 89 f0 mov %esi,%eax 48 89 85 60 ff ff ff mov %rax,-0xa0(%rbp) 89 c8 mov %ecx,%eax 48 89 85 78 ff ff ff mov %rax,-0x88(%rbp) 89 d0 mov %edx,%eax 48 89 85 70 ff ff ff mov %rax,-0x90(%rbp) 8b 45 8c mov -0x74(%rbp),%eax 48 8b 7b 08 mov 0x8(%rbx),%rdi 48 83 c3 18 add $0x18,%rbx 50 push %rax 41 54 push %r12 41 55 push %r13 ff b5 78 ff ff ff pushq -0x88(%rbp) 41 56 push %r14 41 57 push %r15 ff b5 70 ff ff ff pushq -0x90(%rbp) 4c 8b 8d 68 ff ff ff mov -0x98(%rbp),%r9 4c 8b 85 60 ff ff ff mov -0xa0(%rbp),%r8 48 8b 4d 98 mov -0x68(%rbp),%rcx 48 8b 55 90 mov -0x70(%rbp),%rdx 8b 75 80 mov -0x80(%rbp),%esi 41 ff d2 callq *%r10 After the patch: 0f 83 a8 fe ff ff jae ffffffff811e626d 8b 9b b8 fe ff ff mov -0x148(%rbx),%ebx 45 8b 64 24 20 mov 0x20(%r12),%r12d 4c 8b 6d a0 mov -0x60(%rbp),%r13 65 ff 05 f5 f7 e2 7e incl %gs:0x7ee2f7f5(%rip) # 15bd0 <__preempt_count> 4c 8b 35 86 5b 13 01 mov 0x1135b86(%rip),%r14 # ffffffff8231bf68 <__tracepoint_mm_vmscan_lru_shrink_inactive+0x28> 4d 85 f6 test %r14,%r14 74 2a je ffffffff811e6411 49 8b 06 mov (%r14),%rax 8b 4d 8c mov -0x74(%rbp),%ecx 49 8b 7e 08 mov 0x8(%r14),%rdi 49 83 c6 18 add $0x18,%r14 4c 89 ea mov %r13,%rdx 45 89 e1 mov %r12d,%r9d 4c 8d 45 b8 lea -0x48(%rbp),%r8 89 de mov %ebx,%esi 51 push %rcx 48 8b 4d 98 mov -0x68(%rbp),%rcx ff d0 callq *%rax Link: http://lkml.kernel.org/r/2559d7cb-ec60-1200-2362-04fa34fd02bb@fb.com Link: http://lkml.kernel.org/r/20180322121003.4177af15@gandalf.local.home Signed-off-by: Steven Rostedt (VMware) Reported-by: Alexei Starovoitov Acked-by: David Rientjes Acked-by: Michal Hocko Cc: Mel Gorman Cc: Vlastimil Babka Cc: Andrey Ryabinin Cc: Alexei Starovoitov Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/vmstat.h | 11 +++++++++++ include/trace/events/vmscan.h | 24 +++++++++--------------- mm/vmscan.c | 18 +----------------- 3 files changed, 21 insertions(+), 32 deletions(-) (limited to 'include/trace') diff --git a/include/linux/vmstat.h b/include/linux/vmstat.h index a4c2317d8b9f..f25cef84b41d 100644 --- a/include/linux/vmstat.h +++ b/include/linux/vmstat.h @@ -20,6 +20,17 @@ extern int sysctl_vm_numa_stat_handler(struct ctl_table *table, int write, void __user *buffer, size_t *length, loff_t *ppos); #endif +struct reclaim_stat { + unsigned nr_dirty; + unsigned nr_unqueued_dirty; + unsigned nr_congested; + unsigned nr_writeback; + unsigned nr_immediate; + unsigned nr_activate; + unsigned nr_ref_keep; + unsigned nr_unmap_fail; +}; + #ifdef CONFIG_VM_EVENT_COUNTERS /* * Light weight per cpu counter implementation. diff --git a/include/trace/events/vmscan.h b/include/trace/events/vmscan.h index 6570c5b45ba1..a1cb91342231 100644 --- a/include/trace/events/vmscan.h +++ b/include/trace/events/vmscan.h @@ -346,15 +346,9 @@ TRACE_EVENT(mm_vmscan_lru_shrink_inactive, TP_PROTO(int nid, unsigned long nr_scanned, unsigned long nr_reclaimed, - unsigned long nr_dirty, unsigned long nr_writeback, - unsigned long nr_congested, unsigned long nr_immediate, - unsigned long nr_activate, unsigned long nr_ref_keep, - unsigned long nr_unmap_fail, - int priority, int file), + struct reclaim_stat *stat, int priority, int file), - TP_ARGS(nid, nr_scanned, nr_reclaimed, nr_dirty, nr_writeback, - nr_congested, nr_immediate, nr_activate, nr_ref_keep, - nr_unmap_fail, priority, file), + TP_ARGS(nid, nr_scanned, nr_reclaimed, stat, priority, file), TP_STRUCT__entry( __field(int, nid) @@ -375,13 +369,13 @@ TRACE_EVENT(mm_vmscan_lru_shrink_inactive, __entry->nid = nid; __entry->nr_scanned = nr_scanned; __entry->nr_reclaimed = nr_reclaimed; - __entry->nr_dirty = nr_dirty; - __entry->nr_writeback = nr_writeback; - __entry->nr_congested = nr_congested; - __entry->nr_immediate = nr_immediate; - __entry->nr_activate = nr_activate; - __entry->nr_ref_keep = nr_ref_keep; - __entry->nr_unmap_fail = nr_unmap_fail; + __entry->nr_dirty = stat->nr_dirty; + __entry->nr_writeback = stat->nr_writeback; + __entry->nr_congested = stat->nr_congested; + __entry->nr_immediate = stat->nr_immediate; + __entry->nr_activate = stat->nr_activate; + __entry->nr_ref_keep = stat->nr_ref_keep; + __entry->nr_unmap_fail = stat->nr_unmap_fail; __entry->priority = priority; __entry->reclaim_flags = trace_shrink_flags(file); ), diff --git a/mm/vmscan.c b/mm/vmscan.c index e411385b304a..a1d7ba0136fe 100644 --- a/mm/vmscan.c +++ b/mm/vmscan.c @@ -902,17 +902,6 @@ static void page_check_dirty_writeback(struct page *page, mapping->a_ops->is_dirty_writeback(page, dirty, writeback); } -struct reclaim_stat { - unsigned nr_dirty; - unsigned nr_unqueued_dirty; - unsigned nr_congested; - unsigned nr_writeback; - unsigned nr_immediate; - unsigned nr_activate; - unsigned nr_ref_keep; - unsigned nr_unmap_fail; -}; - /* * shrink_page_list() returns the number of reclaimed pages */ @@ -1823,12 +1812,7 @@ shrink_inactive_list(unsigned long nr_to_scan, struct lruvec *lruvec, sc->nr.file_taken += nr_taken; trace_mm_vmscan_lru_shrink_inactive(pgdat->node_id, - nr_scanned, nr_reclaimed, - stat.nr_dirty, stat.nr_writeback, - stat.nr_congested, stat.nr_immediate, - stat.nr_activate, stat.nr_ref_keep, - stat.nr_unmap_fail, - sc->priority, file); + nr_scanned, nr_reclaimed, &stat, sc->priority, file); return nr_reclaimed; } -- cgit v1.2.3 From 6667e6d91c88a788d5ab1b1e999d245473e84a90 Mon Sep 17 00:00:00 2001 From: Ohad Sharabi Date: Wed, 28 Mar 2018 12:42:18 +0300 Subject: scsi: ufs: add trace event for ufs upiu Add UFS Protocol Information Units(upiu) trace events for ufs driver, used to trace various ufs transaction types- command, task-management and device management. The trace-point format is generic and can be easily adapted to trace other upius if needed. Currently tracing ufs transaction of type 'device management', which this patch introduce, cannot be obtained from any other trace. Device management transactions are used for communication with the device such as reading and writing descriptor or attributes etc. Signed-off-by: Ohad Sharabi Reviewed-by: Stanislav Nijnikov Reviewed-by: Bart Van Assche Signed-off-by: Martin K. Petersen --- drivers/scsi/ufs/ufshcd.c | 40 ++++++++++++++++++++++++++++++++++++++++ include/trace/events/ufs.h | 27 +++++++++++++++++++++++++++ 2 files changed, 67 insertions(+) (limited to 'include/trace') diff --git a/drivers/scsi/ufs/ufshcd.c b/drivers/scsi/ufs/ufshcd.c index c5b1bf1cadcb..00e79057f870 100644 --- a/drivers/scsi/ufs/ufshcd.c +++ b/drivers/scsi/ufs/ufshcd.c @@ -276,6 +276,35 @@ static inline void ufshcd_remove_non_printable(char *val) *val = ' '; } +static void ufshcd_add_cmd_upiu_trace(struct ufs_hba *hba, unsigned int tag, + const char *str) +{ + struct utp_upiu_req *rq = hba->lrb[tag].ucd_req_ptr; + + trace_ufshcd_upiu(dev_name(hba->dev), str, &rq->header, &rq->sc.cdb); +} + +static void ufshcd_add_query_upiu_trace(struct ufs_hba *hba, unsigned int tag, + const char *str) +{ + struct utp_upiu_req *rq = hba->lrb[tag].ucd_req_ptr; + + trace_ufshcd_upiu(dev_name(hba->dev), str, &rq->header, &rq->qr); +} + +static void ufshcd_add_tm_upiu_trace(struct ufs_hba *hba, unsigned int tag, + const char *str) +{ + struct utp_task_req_desc *descp; + struct utp_upiu_task_req *task_req; + int off = (int)tag - hba->nutrs; + + descp = &hba->utmrdl_base_addr[off]; + task_req = (struct utp_upiu_task_req *)descp->task_req_upiu; + trace_ufshcd_upiu(dev_name(hba->dev), str, &task_req->header, + &task_req->input_param1); +} + static void ufshcd_add_command_trace(struct ufs_hba *hba, unsigned int tag, const char *str) { @@ -285,6 +314,9 @@ static void ufshcd_add_command_trace(struct ufs_hba *hba, struct ufshcd_lrb *lrbp; int transfer_len = -1; + /* trace UPIU also */ + ufshcd_add_cmd_upiu_trace(hba, tag, str); + if (!trace_ufshcd_command_enabled()) return; @@ -2550,6 +2582,7 @@ static int ufshcd_exec_dev_cmd(struct ufs_hba *hba, hba->dev_cmd.complete = &wait; + ufshcd_add_query_upiu_trace(hba, tag, "query_send"); /* Make sure descriptors are ready before ringing the doorbell */ wmb(); spin_lock_irqsave(hba->host->host_lock, flags); @@ -2559,6 +2592,9 @@ static int ufshcd_exec_dev_cmd(struct ufs_hba *hba, err = ufshcd_wait_for_dev_cmd(hba, lrbp, timeout); + ufshcd_add_query_upiu_trace(hba, tag, + err ? "query_complete_err" : "query_complete"); + out_put_tag: ufshcd_put_dev_cmd_tag(hba, tag); wake_up(&hba->dev_cmd.tag_wq); @@ -5443,11 +5479,14 @@ static int ufshcd_issue_tm_cmd(struct ufs_hba *hba, int lun_id, int task_id, spin_unlock_irqrestore(host->host_lock, flags); + ufshcd_add_tm_upiu_trace(hba, task_tag, "tm_send"); + /* wait until the task management command is completed */ err = wait_event_timeout(hba->tm_wq, test_bit(free_slot, &hba->tm_condition), msecs_to_jiffies(TM_CMD_TIMEOUT)); if (!err) { + ufshcd_add_tm_upiu_trace(hba, task_tag, "tm_complete_err"); dev_err(hba->dev, "%s: task management cmd 0x%.2x timed-out\n", __func__, tm_function); if (ufshcd_clear_tm_cmd(hba, free_slot)) @@ -5456,6 +5495,7 @@ static int ufshcd_issue_tm_cmd(struct ufs_hba *hba, int lun_id, int task_id, err = -ETIMEDOUT; } else { err = ufshcd_task_req_compl(hba, free_slot, tm_response); + ufshcd_add_tm_upiu_trace(hba, task_tag, "tm_complete"); } clear_bit(free_slot, &hba->tm_condition); diff --git a/include/trace/events/ufs.h b/include/trace/events/ufs.h index bf6f82673492..f8260e5c79ad 100644 --- a/include/trace/events/ufs.h +++ b/include/trace/events/ufs.h @@ -257,6 +257,33 @@ TRACE_EVENT(ufshcd_command, ) ); +TRACE_EVENT(ufshcd_upiu, + TP_PROTO(const char *dev_name, const char *str, void *hdr, void *tsf), + + TP_ARGS(dev_name, str, hdr, tsf), + + TP_STRUCT__entry( + __string(dev_name, dev_name) + __string(str, str) + __array(unsigned char, hdr, 12) + __array(unsigned char, tsf, 16) + ), + + TP_fast_assign( + __assign_str(dev_name, dev_name); + __assign_str(str, str); + memcpy(__entry->hdr, hdr, sizeof(__entry->hdr)); + memcpy(__entry->tsf, tsf, sizeof(__entry->tsf)); + ), + + TP_printk( + "%s: %s: HDR:%s, CDB:%s", + __get_str(str), __get_str(dev_name), + __print_hex(__entry->hdr, sizeof(__entry->hdr)), + __print_hex(__entry->tsf, sizeof(__entry->tsf)) + ) +); + #endif /* if !defined(_TRACE_UFS_H) || defined(TRACE_HEADER_MULTI_READ) */ /* This part must be outside protection */ -- cgit v1.2.3 From 1cdae042fc63dd987f39ffb3258e54fdac8b9852 Mon Sep 17 00:00:00 2001 From: Ahbong Chang Date: Mon, 16 Apr 2018 10:36:25 +0800 Subject: tracing: Add missing forward declaration Without this forward declaration compile may fail if this header is included only for registering other probe event without struct pool_workqueue. Link: http://lkml.kernel.org/r/20180416023626.139915-1-cwahbong@google.com Reviewed-by: Todd Poynor Signed-off-by: Ahbong Chang Signed-off-by: Steven Rostedt (VMware) --- include/trace/events/workqueue.h | 2 ++ 1 file changed, 2 insertions(+) (limited to 'include/trace') diff --git a/include/trace/events/workqueue.h b/include/trace/events/workqueue.h index 2f057a494d93..9a761bc6a251 100644 --- a/include/trace/events/workqueue.h +++ b/include/trace/events/workqueue.h @@ -25,6 +25,8 @@ DECLARE_EVENT_CLASS(workqueue_work, TP_printk("work struct %p", __entry->work) ); +struct pool_workqueue; + /** * workqueue_queue_work - called when a work gets queued * @req_cpu: the requested cpu -- cgit v1.2.3