From c41b20e721ea4f6f20f66a66e7f0c3c97a2ca9c2 Mon Sep 17 00:00:00 2001 From: Adam Buchbinder Date: Fri, 11 Dec 2009 16:35:39 -0500 Subject: Fix misspellings of "truly" in comments. Some comments misspell "truly"; this fixes them. No code changes. Signed-off-by: Adam Buchbinder Signed-off-by: Jiri Kosina --- kernel/trace/ring_buffer.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/ring_buffer.c b/kernel/trace/ring_buffer.c index 8c1b2d290718..9ab578f1bb65 100644 --- a/kernel/trace/ring_buffer.c +++ b/kernel/trace/ring_buffer.c @@ -2541,7 +2541,7 @@ EXPORT_SYMBOL_GPL(ring_buffer_record_disable); * @buffer: The ring buffer to enable writes * * Note, multiple disables will need the same number of enables - * to truely enable the writing (much like preempt_disable). + * to truly enable the writing (much like preempt_disable). */ void ring_buffer_record_enable(struct ring_buffer *buffer) { @@ -2577,7 +2577,7 @@ EXPORT_SYMBOL_GPL(ring_buffer_record_disable_cpu); * @cpu: The CPU to enable. * * Note, multiple disables will need the same number of enables - * to truely enable the writing (much like preempt_disable). + * to truly enable the writing (much like preempt_disable). */ void ring_buffer_record_enable_cpu(struct ring_buffer *buffer, int cpu) { -- cgit v1.2.3 From 1537a3638cbf741d3826c1002026cce487a6bee0 Mon Sep 17 00:00:00 2001 From: Daniel Mack Date: Fri, 29 Jan 2010 15:57:49 +0800 Subject: tree-wide: fix 'lenght' typo in comments and code Some misspelled occurences of 'octet' and some comments were also fixed as I was on it. Signed-off-by: Daniel Mack Cc: Jiri Kosina Cc: Joe Perches Cc: Junio C Hamano Signed-off-by: Jiri Kosina --- kernel/trace/trace.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace.h b/kernel/trace/trace.h index 4df6a77eb196..e4b32c8aa85f 100644 --- a/kernel/trace/trace.h +++ b/kernel/trace/trace.h @@ -549,7 +549,7 @@ static inline int ftrace_trace_task(struct task_struct *task) * struct trace_parser - servers for reading the user input separated by spaces * @cont: set if the input is not complete - no final space char was found * @buffer: holds the parsed user input - * @idx: user input lenght + * @idx: user input length * @size: buffer size */ struct trace_parser { -- cgit v1.2.3 From ae1f30384baef4056438d81b305a6a5199b0d16c Mon Sep 17 00:00:00 2001 From: Frederic Weisbecker Date: Sun, 28 Feb 2010 19:42:38 +0100 Subject: tracing: Include irqflags headers from trace clock trace_clock.c includes spinlock.h, which ends up including asm/system.h, which in turn includes linux/irqflags.h in x86. So the definition of raw_local_irq_save is luckily covered there, but this is not the case in parisc: tip/kernel/trace/trace_clock.c:86: error: implicit declaration of function 'raw_local_irq_save' tip/kernel/trace/trace_clock.c:112: error: implicit declaration of function 'raw_local_irq_restore' We need to include linux/irqflags.h directly from trace_clock.c to avoid such build error. Signed-off-by: Frederic Weisbecker Cc: Steven Rostedt Cc: Robert Richter Cc: Peter Zijlstra Signed-off-by: Ingo Molnar --- kernel/trace/trace_clock.c | 1 + 1 file changed, 1 insertion(+) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_clock.c b/kernel/trace/trace_clock.c index 84a3a7ba072a..6fbfb8f417b9 100644 --- a/kernel/trace/trace_clock.c +++ b/kernel/trace/trace_clock.c @@ -13,6 +13,7 @@ * Tracer plugins will chose a default from these clocks. */ #include +#include #include #include #include -- cgit v1.2.3 From ac91d85456372a90af5b85eb6620fd2efb1e431b Mon Sep 17 00:00:00 2001 From: Lai Jiangshan Date: Tue, 2 Mar 2010 17:54:50 +0800 Subject: tracing: Fix warning in s_next of trace file ops This warning in s_next() can be triggered by lseek(): [] ? s_next+0x77/0x80 [] warn_slowpath_common+0x81/0xa0 [] ? s_next+0x77/0x80 [] warn_slowpath_null+0x1a/0x20 [] s_next+0x77/0x80 [] traverse+0x117/0x200 [] seq_lseek+0xa3/0x120 [] ? seq_lseek+0x0/0x120 [] vfs_llseek+0x41/0x50 [] sys_llseek+0x66/0xa0 [] sysenter_do_call+0x12/0x26 The iterator "leftover" variable is zeroed in the opening of the trace file. But lseek can call s_start() which will call s_next() without reseting the "leftover" variable back to zero, which might trigger the WARN_ON_ONCE(iter->leftover) that is in s_next(). Cc: stable@kernel.org Signed-off-by: Lai Jiangshan LKML-Reference: <4B8CE06A.9090207@cn.fujitsu.com> Signed-off-by: Steven Rostedt --- kernel/trace/trace.c | 1 + 1 file changed, 1 insertion(+) (limited to 'kernel/trace') diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index 032c57ca6502..5edf410bc540 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -1703,6 +1703,7 @@ static void *s_start(struct seq_file *m, loff_t *pos) ftrace_enable_cpu(); + iter->leftover = 0; for (p = iter; p && l < *pos; p = s_next(m, p, &l)) ; -- cgit v1.2.3 From 8d53dd546f36073e0d29b0cfc24c665db301e3e7 Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Wed, 3 Mar 2010 17:50:18 -0800 Subject: rcu, ftrace: Fix RCU lockdep splat in ftrace_perf_buf_prepare() Change the pair of rcu_dereference() calls in ftrace_perf_buf_prepare() to rcu_dereference_sched(). Signed-off-by: Paul E. McKenney Acked-by: Frederic Weisbecker Cc: Steven Rostedt Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: mathieu.desnoyers@polymtl.ca Cc: josh@joshtriplett.org Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: Valdis.Kletnieks@vt.edu Cc: dhowells@redhat.com Cc: Frederic Weisbecker LKML-Reference: <1267667418-32233-3-git-send-email-paulmck@linux.vnet.ibm.com> Signed-off-by: Ingo Molnar --- kernel/trace/trace_event_profile.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_event_profile.c b/kernel/trace/trace_event_profile.c index f0d693005075..c1cc3ab633de 100644 --- a/kernel/trace/trace_event_profile.c +++ b/kernel/trace/trace_event_profile.c @@ -138,9 +138,9 @@ __kprobes void *ftrace_perf_buf_prepare(int size, unsigned short type, cpu = smp_processor_id(); if (in_nmi()) - trace_buf = rcu_dereference(perf_trace_buf_nmi); + trace_buf = rcu_dereference_sched(perf_trace_buf_nmi); else - trace_buf = rcu_dereference(perf_trace_buf); + trace_buf = rcu_dereference_sched(perf_trace_buf); if (!trace_buf) goto err; -- cgit v1.2.3 From 801c29fd1fdeb84f60241beb445ff5db154450ae Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Fri, 5 Mar 2010 20:02:19 -0500 Subject: function-graph: Fix unused reference to ftrace_set_func() The declaration of ftrace_set_func() is at the start of the ftrace.c file and wrapped with a #ifdef CONFIG_FUNCTION_GRAPH condition. If function graph tracing is enabled but CONFIG_DYNAMIC_FTRACE is not, a warning about that function being declared static and unused is given. This really should have been placed within the CONFIG_FUNCTION_GRAPH condition that uses ftrace_set_func(). Moving the declaration down fixes the warning and makes the code cleaner. Reported-by: Peter Zijlstra Signed-off-by: Steven Rostedt --- kernel/trace/ftrace.c | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index d996353473fd..d0407c9f368c 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -85,10 +85,6 @@ ftrace_func_t ftrace_trace_function __read_mostly = ftrace_stub; ftrace_func_t __ftrace_trace_function __read_mostly = ftrace_stub; ftrace_func_t ftrace_pid_function __read_mostly = ftrace_stub; -#ifdef CONFIG_FUNCTION_GRAPH_TRACER -static int ftrace_set_func(unsigned long *array, int *idx, char *buffer); -#endif - static void ftrace_list_func(unsigned long ip, unsigned long parent_ip) { struct ftrace_ops *op = ftrace_list; @@ -2300,6 +2296,8 @@ __setup("ftrace_filter=", set_ftrace_filter); #ifdef CONFIG_FUNCTION_GRAPH_TRACER static char ftrace_graph_buf[FTRACE_FILTER_SIZE] __initdata; +static int ftrace_set_func(unsigned long *array, int *idx, char *buffer); + static int __init set_graph_function(char *str) { strlcpy(ftrace_graph_buf, str, FTRACE_FILTER_SIZE); -- cgit v1.2.3 From a094fe04c751698a18c3a0d376a3bdb117f1e0d8 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Fri, 5 Mar 2010 20:08:58 -0500 Subject: function-graph: Use comment notation for func names of dangling '}' When a '}' does not have a matching function start, the name is printed within parenthesis. But this makes it confusing between ending '}' and function starts. This patch makes the function name appear in C comment notation. Old view: 3) 1.281 us | } (might_fault) 3) 3.620 us | } (filldir) 3) 5.251 us | } (call_filldir) 3) | call_filldir() { 3) | filldir() { New view: 3) 1.281 us | } /* might_fault */ 3) 3.620 us | } /* filldir */ 3) 5.251 us | } /* call_filldir */ 3) | call_filldir() { 3) | filldir() { Requested-by: Ingo Molnar Signed-off-by: Steven Rostedt --- kernel/trace/trace_functions_graph.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_functions_graph.c b/kernel/trace/trace_functions_graph.c index e998a824e9db..7b1f24618d97 100644 --- a/kernel/trace/trace_functions_graph.c +++ b/kernel/trace/trace_functions_graph.c @@ -920,7 +920,7 @@ print_graph_return(struct ftrace_graph_ret *trace, struct trace_seq *s, if (!ret) return TRACE_TYPE_PARTIAL_LINE; } else { - ret = trace_seq_printf(s, "} (%ps)\n", (void *)trace->func); + ret = trace_seq_printf(s, "} /* %ps */\n", (void *)trace->func); if (!ret) return TRACE_TYPE_PARTIAL_LINE; } -- cgit v1.2.3 From 1acaa1b2d9b5904c9cce06122990a2d71046ce16 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Fri, 5 Mar 2010 18:23:50 -0300 Subject: tracing: Update the comm field in the right variable in update_max_tr The latency output showed: # | task: -3 (uid:0 nice:0 policy:1 rt_prio:99) The comm is missing in the "task:" and it looks like a minus 3 is the output. The correct display should be: # | task: migration/0-3 (uid:0 nice:0 policy:1 rt_prio:99) The problem is that the comm is being stored in the wrong data structure. The max_tr.data[cpu] is what stores the comm, not the tr->data[cpu]. Before this patch the max_tr.data[cpu]->comm was zeroed and the /debug/trace ended up showing just the '-' sign followed by the pid. Also remove a needless initialization of max_data. Signed-off-by: Arnaldo Carvalho de Melo LKML-Reference: <1267824230-23861-1-git-send-email-acme@infradead.org> Signed-off-by: Steven Rostedt --- kernel/trace/trace.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index 032c57ca6502..6efd5cb3c252 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -592,7 +592,7 @@ static void __update_max_tr(struct trace_array *tr, struct task_struct *tsk, int cpu) { struct trace_array_cpu *data = tr->data[cpu]; - struct trace_array_cpu *max_data = tr->data[cpu]; + struct trace_array_cpu *max_data; max_tr.cpu = cpu; max_tr.time_start = data->preempt_timestamp; @@ -602,7 +602,7 @@ __update_max_tr(struct trace_array *tr, struct task_struct *tsk, int cpu) max_data->critical_start = data->critical_start; max_data->critical_end = data->critical_end; - memcpy(data->comm, tsk->comm, TASK_COMM_LEN); + memcpy(max_data->comm, tsk->comm, TASK_COMM_LEN); max_data->pid = tsk->pid; max_data->uid = task_uid(tsk); max_data->nice = tsk->static_prio - 20 - MAX_RT_PRIO; -- cgit v1.2.3 From 0e95017355dcf43031da6d0e360a748717e56df1 Mon Sep 17 00:00:00 2001 From: Tim Bird Date: Thu, 25 Feb 2010 15:36:43 -0800 Subject: function-graph: Add tracing_thresh support to function_graph tracer Add support for tracing_thresh to the function_graph tracer. This version of this feature isolates the checks into new entry and return functions, to avoid adding more conditional code into the main function_graph paths. When the tracing_thresh is set and the function graph tracer is enabled, only the functions that took longer than the time in microseconds that was set in tracing_thresh are recorded. To do this efficiently, only the function exits are recorded: [tracing]# echo 100 > tracing_thresh [tracing]# echo function_graph > current_tracer [tracing]# cat trace # tracer: function_graph # # CPU DURATION FUNCTION CALLS # | | | | | | | 1) ! 119.214 us | } /* smp_apic_timer_interrupt */ 1) <========== | 0) ! 101.527 us | } /* __rcu_process_callbacks */ 0) ! 126.461 us | } /* rcu_process_callbacks */ 0) ! 145.111 us | } /* __do_softirq */ 0) ! 149.667 us | } /* do_softirq */ 0) ! 168.817 us | } /* irq_exit */ 0) ! 248.254 us | } /* smp_apic_timer_interrupt */ Also, add support for specifying tracing_thresh on the kernel command line. When used like so: "tracing_thresh=200 ftrace=function_graph" this can be used to analyse system startup. It is important to disable tracing soon after boot, in order to avoid losing the trace data. Acked-by: Frederic Weisbecker Signed-off-by: Tim Bird LKML-Reference: <4B87098B.4040308@am.sony.com> Signed-off-by: Steven Rostedt --- kernel/trace/trace.c | 20 ++++++++++++++++++-- kernel/trace/trace.h | 3 ++- kernel/trace/trace_functions_graph.c | 25 +++++++++++++++++++++++-- 3 files changed, 43 insertions(+), 5 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index 6efd5cb3c252..ababedb4e87f 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -374,6 +374,21 @@ static int __init set_buf_size(char *str) } __setup("trace_buf_size=", set_buf_size); +static int __init set_tracing_thresh(char *str) +{ + unsigned long threshhold; + int ret; + + if (!str) + return 0; + ret = strict_strtoul(str, 0, &threshhold); + if (ret < 0) + return 0; + tracing_thresh = threshhold * 1000; + return 1; +} +__setup("tracing_thresh=", set_tracing_thresh); + unsigned long nsecs_to_usecs(unsigned long nsecs) { return nsecs / 1000; @@ -579,9 +594,10 @@ static ssize_t trace_seq_to_buffer(struct trace_seq *s, void *buf, size_t cnt) static arch_spinlock_t ftrace_max_lock = (arch_spinlock_t)__ARCH_SPIN_LOCK_UNLOCKED; +unsigned long __read_mostly tracing_thresh; + #ifdef CONFIG_TRACER_MAX_TRACE unsigned long __read_mostly tracing_max_latency; -unsigned long __read_mostly tracing_thresh; /* * Copy the new maximum trace into the separate maximum-trace @@ -4248,10 +4264,10 @@ static __init int tracer_init_debugfs(void) #ifdef CONFIG_TRACER_MAX_TRACE trace_create_file("tracing_max_latency", 0644, d_tracer, &tracing_max_latency, &tracing_max_lat_fops); +#endif trace_create_file("tracing_thresh", 0644, d_tracer, &tracing_thresh, &tracing_max_lat_fops); -#endif trace_create_file("README", 0444, d_tracer, NULL, &tracing_readme_fops); diff --git a/kernel/trace/trace.h b/kernel/trace/trace.h index fd05bcaf91b0..1bc8cd1431d7 100644 --- a/kernel/trace/trace.h +++ b/kernel/trace/trace.h @@ -396,9 +396,10 @@ extern int process_new_ksym_entry(char *ksymname, int op, unsigned long addr); extern unsigned long nsecs_to_usecs(unsigned long nsecs); +extern unsigned long tracing_thresh; + #ifdef CONFIG_TRACER_MAX_TRACE extern unsigned long tracing_max_latency; -extern unsigned long tracing_thresh; void update_max_tr(struct trace_array *tr, struct task_struct *tsk, int cpu); void update_max_tr_single(struct trace_array *tr, diff --git a/kernel/trace/trace_functions_graph.c b/kernel/trace/trace_functions_graph.c index 7b1f24618d97..e9df04b60267 100644 --- a/kernel/trace/trace_functions_graph.c +++ b/kernel/trace/trace_functions_graph.c @@ -237,6 +237,14 @@ int trace_graph_entry(struct ftrace_graph_ent *trace) return ret; } +int trace_graph_thresh_entry(struct ftrace_graph_ent *trace) +{ + if (tracing_thresh) + return 1; + else + return trace_graph_entry(trace); +} + static void __trace_graph_return(struct trace_array *tr, struct ftrace_graph_ret *trace, unsigned long flags, @@ -290,13 +298,26 @@ void set_graph_array(struct trace_array *tr) smp_mb(); } +void trace_graph_thresh_return(struct ftrace_graph_ret *trace) +{ + if (tracing_thresh && + (trace->rettime - trace->calltime < tracing_thresh)) + return; + else + trace_graph_return(trace); +} + static int graph_trace_init(struct trace_array *tr) { int ret; set_graph_array(tr); - ret = register_ftrace_graph(&trace_graph_return, - &trace_graph_entry); + if (tracing_thresh) + ret = register_ftrace_graph(&trace_graph_thresh_return, + &trace_graph_thresh_entry); + else + ret = register_ftrace_graph(&trace_graph_return, + &trace_graph_entry); if (ret) return ret; tracing_start_cmdline_record(); -- cgit v1.2.3 From c530665c31c0140b74ca7689e7f836177796e5bd Mon Sep 17 00:00:00 2001 From: Frederic Weisbecker Date: Wed, 3 Mar 2010 07:16:16 +0100 Subject: perf: Take a hot regs snapshot for trace events We are taking a wrong regs snapshot when a trace event triggers. Either we use get_irq_regs(), which gives us the interrupted registers if we are in an interrupt, or we use task_pt_regs() which gives us the state before we entered the kernel, assuming we are lucky enough to be no kernel thread, in which case task_pt_regs() returns the initial set of regs when the kernel thread was started. What we want is different. We need a hot snapshot of the regs, so that we can get the instruction pointer to record in the sample, the frame pointer for the callchain, and some other things. Let's use the new perf_fetch_caller_regs() for that. Comparison with perf record -e lock: -R -a -f -g Before: perf [kernel] [k] __do_softirq | --- __do_softirq | |--55.16%-- __open | --44.84%-- __write_nocancel After: perf [kernel] [k] perf_tp_event | --- perf_tp_event | |--41.07%-- lock_acquire | | | |--39.36%-- _raw_spin_lock | | | | | |--7.81%-- hrtimer_interrupt | | | smp_apic_timer_interrupt | | | apic_timer_interrupt The old case was producing unreliable callchains. Now having right frame and instruction pointers, we have the trace we want. Also syscalls and kprobe events already have the right regs, let's use them instead of wasting a retrieval. v2: Follow the rename perf_save_regs() -> perf_fetch_caller_regs() Signed-off-by: Frederic Weisbecker Cc: Ingo Molnar Cc: Thomas Gleixner Cc: H. Peter Anvin Cc: Peter Zijlstra Cc: Paul Mackerras Cc: Steven Rostedt Cc: Arnaldo Carvalho de Melo Cc: Masami Hiramatsu Cc: Jason Baron Cc: Archs --- kernel/trace/trace_event_profile.c | 3 ++- kernel/trace/trace_kprobe.c | 5 +++-- kernel/trace/trace_syscalls.c | 4 ++-- 3 files changed, 7 insertions(+), 5 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_event_profile.c b/kernel/trace/trace_event_profile.c index f0d693005075..e66d21e15a0f 100644 --- a/kernel/trace/trace_event_profile.c +++ b/kernel/trace/trace_event_profile.c @@ -2,13 +2,14 @@ * trace event based perf counter profiling * * Copyright (C) 2009 Red Hat Inc, Peter Zijlstra - * + * Copyright (C) 2009-2010 Frederic Weisbecker */ #include #include #include "trace.h" +DEFINE_PER_CPU(struct pt_regs, perf_trace_regs); static char *perf_trace_buf; static char *perf_trace_buf_nmi; diff --git a/kernel/trace/trace_kprobe.c b/kernel/trace/trace_kprobe.c index 505c92273b1a..f7a20a8bfb31 100644 --- a/kernel/trace/trace_kprobe.c +++ b/kernel/trace/trace_kprobe.c @@ -1240,7 +1240,7 @@ static __kprobes void kprobe_profile_func(struct kprobe *kp, for (i = 0; i < tp->nr_args; i++) entry->args[i] = call_fetch(&tp->args[i].fetch, regs); - ftrace_perf_buf_submit(entry, size, rctx, entry->ip, 1, irq_flags); + ftrace_perf_buf_submit(entry, size, rctx, entry->ip, 1, irq_flags, regs); } /* Kretprobe profile handler */ @@ -1271,7 +1271,8 @@ static __kprobes void kretprobe_profile_func(struct kretprobe_instance *ri, for (i = 0; i < tp->nr_args; i++) entry->args[i] = call_fetch(&tp->args[i].fetch, regs); - ftrace_perf_buf_submit(entry, size, rctx, entry->ret_ip, 1, irq_flags); + ftrace_perf_buf_submit(entry, size, rctx, entry->ret_ip, 1, + irq_flags, regs); } static int probe_profile_enable(struct ftrace_event_call *call) diff --git a/kernel/trace/trace_syscalls.c b/kernel/trace/trace_syscalls.c index cba47d7935cc..7e6e84fb7b6c 100644 --- a/kernel/trace/trace_syscalls.c +++ b/kernel/trace/trace_syscalls.c @@ -467,7 +467,7 @@ static void prof_syscall_enter(struct pt_regs *regs, long id) rec->nr = syscall_nr; syscall_get_arguments(current, regs, 0, sys_data->nb_args, (unsigned long *)&rec->args); - ftrace_perf_buf_submit(rec, size, rctx, 0, 1, flags); + ftrace_perf_buf_submit(rec, size, rctx, 0, 1, flags, regs); } int prof_sysenter_enable(struct ftrace_event_call *call) @@ -542,7 +542,7 @@ static void prof_syscall_exit(struct pt_regs *regs, long ret) rec->nr = syscall_nr; rec->ret = syscall_get_return_value(current, regs); - ftrace_perf_buf_submit(rec, size, rctx, 0, 1, flags); + ftrace_perf_buf_submit(rec, size, rctx, 0, 1, flags, regs); } int prof_sysexit_enable(struct ftrace_event_call *call) -- cgit v1.2.3 From 97d5a22005f38057b4bc0d95f81cd26510268794 Mon Sep 17 00:00:00 2001 From: Frederic Weisbecker Date: Fri, 5 Mar 2010 05:35:37 +0100 Subject: perf: Drop the obsolete profile naming for trace events Drop the obsolete "profile" naming used by perf for trace events. Perf can now do more than simple events counting, so generalize the API naming. Signed-off-by: Frederic Weisbecker Cc: Ingo Molnar Cc: Peter Zijlstra Cc: Paul Mackerras Cc: Steven Rostedt Cc: Masami Hiramatsu Cc: Jason Baron --- kernel/trace/Makefile | 2 +- kernel/trace/trace_event_perf.c | 165 +++++++++++++++++++++++++++++++++++++ kernel/trace/trace_event_profile.c | 165 ------------------------------------- kernel/trace/trace_events.c | 2 +- kernel/trace/trace_kprobe.c | 28 +++---- kernel/trace/trace_syscalls.c | 72 ++++++++-------- 6 files changed, 217 insertions(+), 217 deletions(-) create mode 100644 kernel/trace/trace_event_perf.c delete mode 100644 kernel/trace/trace_event_profile.c (limited to 'kernel/trace') diff --git a/kernel/trace/Makefile b/kernel/trace/Makefile index d00c6fe23f54..78edc6490038 100644 --- a/kernel/trace/Makefile +++ b/kernel/trace/Makefile @@ -52,7 +52,7 @@ obj-$(CONFIG_EVENT_TRACING) += trace_events.o obj-$(CONFIG_EVENT_TRACING) += trace_export.o obj-$(CONFIG_FTRACE_SYSCALLS) += trace_syscalls.o ifeq ($(CONFIG_PERF_EVENTS),y) -obj-$(CONFIG_EVENT_TRACING) += trace_event_profile.o +obj-$(CONFIG_EVENT_TRACING) += trace_event_perf.o endif obj-$(CONFIG_EVENT_TRACING) += trace_events_filter.o obj-$(CONFIG_KPROBE_EVENT) += trace_kprobe.o diff --git a/kernel/trace/trace_event_perf.c b/kernel/trace/trace_event_perf.c new file mode 100644 index 000000000000..f315b12a41d8 --- /dev/null +++ b/kernel/trace/trace_event_perf.c @@ -0,0 +1,165 @@ +/* + * trace event based perf event profiling/tracing + * + * Copyright (C) 2009 Red Hat Inc, Peter Zijlstra + * Copyright (C) 2009-2010 Frederic Weisbecker + */ + +#include +#include +#include "trace.h" + +DEFINE_PER_CPU(struct pt_regs, perf_trace_regs); + +static char *perf_trace_buf; +static char *perf_trace_buf_nmi; + +typedef typeof(char [PERF_MAX_TRACE_SIZE]) perf_trace_t ; + +/* Count the events in use (per event id, not per instance) */ +static int total_ref_count; + +static int perf_trace_event_enable(struct ftrace_event_call *event) +{ + char *buf; + int ret = -ENOMEM; + + if (event->perf_refcount++ > 0) + return 0; + + if (!total_ref_count) { + buf = (char *)alloc_percpu(perf_trace_t); + if (!buf) + goto fail_buf; + + rcu_assign_pointer(perf_trace_buf, buf); + + buf = (char *)alloc_percpu(perf_trace_t); + if (!buf) + goto fail_buf_nmi; + + rcu_assign_pointer(perf_trace_buf_nmi, buf); + } + + ret = event->perf_event_enable(event); + if (!ret) { + total_ref_count++; + return 0; + } + +fail_buf_nmi: + if (!total_ref_count) { + free_percpu(perf_trace_buf_nmi); + free_percpu(perf_trace_buf); + perf_trace_buf_nmi = NULL; + perf_trace_buf = NULL; + } +fail_buf: + event->perf_refcount--; + + return ret; +} + +int perf_trace_enable(int event_id) +{ + struct ftrace_event_call *event; + int ret = -EINVAL; + + mutex_lock(&event_mutex); + list_for_each_entry(event, &ftrace_events, list) { + if (event->id == event_id && event->perf_event_enable && + try_module_get(event->mod)) { + ret = perf_trace_event_enable(event); + break; + } + } + mutex_unlock(&event_mutex); + + return ret; +} + +static void perf_trace_event_disable(struct ftrace_event_call *event) +{ + char *buf, *nmi_buf; + + if (--event->perf_refcount > 0) + return; + + event->perf_event_disable(event); + + if (!--total_ref_count) { + buf = perf_trace_buf; + rcu_assign_pointer(perf_trace_buf, NULL); + + nmi_buf = perf_trace_buf_nmi; + rcu_assign_pointer(perf_trace_buf_nmi, NULL); + + /* + * Ensure every events in profiling have finished before + * releasing the buffers + */ + synchronize_sched(); + + free_percpu(buf); + free_percpu(nmi_buf); + } +} + +void perf_trace_disable(int event_id) +{ + struct ftrace_event_call *event; + + mutex_lock(&event_mutex); + list_for_each_entry(event, &ftrace_events, list) { + if (event->id == event_id) { + perf_trace_event_disable(event); + module_put(event->mod); + break; + } + } + mutex_unlock(&event_mutex); +} + +__kprobes void *perf_trace_buf_prepare(int size, unsigned short type, + int *rctxp, unsigned long *irq_flags) +{ + struct trace_entry *entry; + char *trace_buf, *raw_data; + int pc, cpu; + + pc = preempt_count(); + + /* Protect the per cpu buffer, begin the rcu read side */ + local_irq_save(*irq_flags); + + *rctxp = perf_swevent_get_recursion_context(); + if (*rctxp < 0) + goto err_recursion; + + cpu = smp_processor_id(); + + if (in_nmi()) + trace_buf = rcu_dereference(perf_trace_buf_nmi); + else + trace_buf = rcu_dereference(perf_trace_buf); + + if (!trace_buf) + goto err; + + raw_data = per_cpu_ptr(trace_buf, cpu); + + /* zero the dead bytes from align to not leak stack to user */ + *(u64 *)(&raw_data[size - sizeof(u64)]) = 0ULL; + + entry = (struct trace_entry *)raw_data; + tracing_generic_entry_update(entry, *irq_flags, pc); + entry->type = type; + + return raw_data; +err: + perf_swevent_put_recursion_context(*rctxp); +err_recursion: + local_irq_restore(*irq_flags); + return NULL; +} +EXPORT_SYMBOL_GPL(perf_trace_buf_prepare); diff --git a/kernel/trace/trace_event_profile.c b/kernel/trace/trace_event_profile.c deleted file mode 100644 index e66d21e15a0f..000000000000 --- a/kernel/trace/trace_event_profile.c +++ /dev/null @@ -1,165 +0,0 @@ -/* - * trace event based perf counter profiling - * - * Copyright (C) 2009 Red Hat Inc, Peter Zijlstra - * Copyright (C) 2009-2010 Frederic Weisbecker - */ - -#include -#include -#include "trace.h" - -DEFINE_PER_CPU(struct pt_regs, perf_trace_regs); - -static char *perf_trace_buf; -static char *perf_trace_buf_nmi; - -typedef typeof(char [FTRACE_MAX_PROFILE_SIZE]) perf_trace_t ; - -/* Count the events in use (per event id, not per instance) */ -static int total_profile_count; - -static int ftrace_profile_enable_event(struct ftrace_event_call *event) -{ - char *buf; - int ret = -ENOMEM; - - if (event->profile_count++ > 0) - return 0; - - if (!total_profile_count) { - buf = (char *)alloc_percpu(perf_trace_t); - if (!buf) - goto fail_buf; - - rcu_assign_pointer(perf_trace_buf, buf); - - buf = (char *)alloc_percpu(perf_trace_t); - if (!buf) - goto fail_buf_nmi; - - rcu_assign_pointer(perf_trace_buf_nmi, buf); - } - - ret = event->profile_enable(event); - if (!ret) { - total_profile_count++; - return 0; - } - -fail_buf_nmi: - if (!total_profile_count) { - free_percpu(perf_trace_buf_nmi); - free_percpu(perf_trace_buf); - perf_trace_buf_nmi = NULL; - perf_trace_buf = NULL; - } -fail_buf: - event->profile_count--; - - return ret; -} - -int ftrace_profile_enable(int event_id) -{ - struct ftrace_event_call *event; - int ret = -EINVAL; - - mutex_lock(&event_mutex); - list_for_each_entry(event, &ftrace_events, list) { - if (event->id == event_id && event->profile_enable && - try_module_get(event->mod)) { - ret = ftrace_profile_enable_event(event); - break; - } - } - mutex_unlock(&event_mutex); - - return ret; -} - -static void ftrace_profile_disable_event(struct ftrace_event_call *event) -{ - char *buf, *nmi_buf; - - if (--event->profile_count > 0) - return; - - event->profile_disable(event); - - if (!--total_profile_count) { - buf = perf_trace_buf; - rcu_assign_pointer(perf_trace_buf, NULL); - - nmi_buf = perf_trace_buf_nmi; - rcu_assign_pointer(perf_trace_buf_nmi, NULL); - - /* - * Ensure every events in profiling have finished before - * releasing the buffers - */ - synchronize_sched(); - - free_percpu(buf); - free_percpu(nmi_buf); - } -} - -void ftrace_profile_disable(int event_id) -{ - struct ftrace_event_call *event; - - mutex_lock(&event_mutex); - list_for_each_entry(event, &ftrace_events, list) { - if (event->id == event_id) { - ftrace_profile_disable_event(event); - module_put(event->mod); - break; - } - } - mutex_unlock(&event_mutex); -} - -__kprobes void *ftrace_perf_buf_prepare(int size, unsigned short type, - int *rctxp, unsigned long *irq_flags) -{ - struct trace_entry *entry; - char *trace_buf, *raw_data; - int pc, cpu; - - pc = preempt_count(); - - /* Protect the per cpu buffer, begin the rcu read side */ - local_irq_save(*irq_flags); - - *rctxp = perf_swevent_get_recursion_context(); - if (*rctxp < 0) - goto err_recursion; - - cpu = smp_processor_id(); - - if (in_nmi()) - trace_buf = rcu_dereference(perf_trace_buf_nmi); - else - trace_buf = rcu_dereference(perf_trace_buf); - - if (!trace_buf) - goto err; - - raw_data = per_cpu_ptr(trace_buf, cpu); - - /* zero the dead bytes from align to not leak stack to user */ - *(u64 *)(&raw_data[size - sizeof(u64)]) = 0ULL; - - entry = (struct trace_entry *)raw_data; - tracing_generic_entry_update(entry, *irq_flags, pc); - entry->type = type; - - return raw_data; -err: - perf_swevent_put_recursion_context(*rctxp); -err_recursion: - local_irq_restore(*irq_flags); - return NULL; -} -EXPORT_SYMBOL_GPL(ftrace_perf_buf_prepare); diff --git a/kernel/trace/trace_events.c b/kernel/trace/trace_events.c index 3f972ad98d04..beab8bf2f310 100644 --- a/kernel/trace/trace_events.c +++ b/kernel/trace/trace_events.c @@ -938,7 +938,7 @@ event_create_dir(struct ftrace_event_call *call, struct dentry *d_events, trace_create_file("enable", 0644, call->dir, call, enable); - if (call->id && call->profile_enable) + if (call->id && call->perf_event_enable) trace_create_file("id", 0444, call->dir, call, id); diff --git a/kernel/trace/trace_kprobe.c b/kernel/trace/trace_kprobe.c index f7a20a8bfb31..1251e367bae9 100644 --- a/kernel/trace/trace_kprobe.c +++ b/kernel/trace/trace_kprobe.c @@ -1214,7 +1214,7 @@ static int set_print_fmt(struct trace_probe *tp) #ifdef CONFIG_PERF_EVENTS /* Kprobe profile handler */ -static __kprobes void kprobe_profile_func(struct kprobe *kp, +static __kprobes void kprobe_perf_func(struct kprobe *kp, struct pt_regs *regs) { struct trace_probe *tp = container_of(kp, struct trace_probe, rp.kp); @@ -1227,11 +1227,11 @@ static __kprobes void kprobe_profile_func(struct kprobe *kp, __size = SIZEOF_KPROBE_TRACE_ENTRY(tp->nr_args); size = ALIGN(__size + sizeof(u32), sizeof(u64)); size -= sizeof(u32); - if (WARN_ONCE(size > FTRACE_MAX_PROFILE_SIZE, + if (WARN_ONCE(size > PERF_MAX_TRACE_SIZE, "profile buffer not large enough")) return; - entry = ftrace_perf_buf_prepare(size, call->id, &rctx, &irq_flags); + entry = perf_trace_buf_prepare(size, call->id, &rctx, &irq_flags); if (!entry) return; @@ -1240,11 +1240,11 @@ static __kprobes void kprobe_profile_func(struct kprobe *kp, for (i = 0; i < tp->nr_args; i++) entry->args[i] = call_fetch(&tp->args[i].fetch, regs); - ftrace_perf_buf_submit(entry, size, rctx, entry->ip, 1, irq_flags, regs); + perf_trace_buf_submit(entry, size, rctx, entry->ip, 1, irq_flags, regs); } /* Kretprobe profile handler */ -static __kprobes void kretprobe_profile_func(struct kretprobe_instance *ri, +static __kprobes void kretprobe_perf_func(struct kretprobe_instance *ri, struct pt_regs *regs) { struct trace_probe *tp = container_of(ri->rp, struct trace_probe, rp); @@ -1257,11 +1257,11 @@ static __kprobes void kretprobe_profile_func(struct kretprobe_instance *ri, __size = SIZEOF_KRETPROBE_TRACE_ENTRY(tp->nr_args); size = ALIGN(__size + sizeof(u32), sizeof(u64)); size -= sizeof(u32); - if (WARN_ONCE(size > FTRACE_MAX_PROFILE_SIZE, + if (WARN_ONCE(size > PERF_MAX_TRACE_SIZE, "profile buffer not large enough")) return; - entry = ftrace_perf_buf_prepare(size, call->id, &rctx, &irq_flags); + entry = perf_trace_buf_prepare(size, call->id, &rctx, &irq_flags); if (!entry) return; @@ -1271,11 +1271,11 @@ static __kprobes void kretprobe_profile_func(struct kretprobe_instance *ri, for (i = 0; i < tp->nr_args; i++) entry->args[i] = call_fetch(&tp->args[i].fetch, regs); - ftrace_perf_buf_submit(entry, size, rctx, entry->ret_ip, 1, + perf_trace_buf_submit(entry, size, rctx, entry->ret_ip, 1, irq_flags, regs); } -static int probe_profile_enable(struct ftrace_event_call *call) +static int probe_perf_enable(struct ftrace_event_call *call) { struct trace_probe *tp = (struct trace_probe *)call->data; @@ -1287,7 +1287,7 @@ static int probe_profile_enable(struct ftrace_event_call *call) return enable_kprobe(&tp->rp.kp); } -static void probe_profile_disable(struct ftrace_event_call *call) +static void probe_perf_disable(struct ftrace_event_call *call) { struct trace_probe *tp = (struct trace_probe *)call->data; @@ -1312,7 +1312,7 @@ int kprobe_dispatcher(struct kprobe *kp, struct pt_regs *regs) kprobe_trace_func(kp, regs); #ifdef CONFIG_PERF_EVENTS if (tp->flags & TP_FLAG_PROFILE) - kprobe_profile_func(kp, regs); + kprobe_perf_func(kp, regs); #endif return 0; /* We don't tweek kernel, so just return 0 */ } @@ -1326,7 +1326,7 @@ int kretprobe_dispatcher(struct kretprobe_instance *ri, struct pt_regs *regs) kretprobe_trace_func(ri, regs); #ifdef CONFIG_PERF_EVENTS if (tp->flags & TP_FLAG_PROFILE) - kretprobe_profile_func(ri, regs); + kretprobe_perf_func(ri, regs); #endif return 0; /* We don't tweek kernel, so just return 0 */ } @@ -1359,8 +1359,8 @@ static int register_probe_event(struct trace_probe *tp) call->unregfunc = probe_event_disable; #ifdef CONFIG_PERF_EVENTS - call->profile_enable = probe_profile_enable; - call->profile_disable = probe_profile_disable; + call->perf_event_enable = probe_perf_enable; + call->perf_event_disable = probe_perf_disable; #endif call->data = tp; ret = trace_add_event_call(call); diff --git a/kernel/trace/trace_syscalls.c b/kernel/trace/trace_syscalls.c index 7e6e84fb7b6c..33c2a5b769dc 100644 --- a/kernel/trace/trace_syscalls.c +++ b/kernel/trace/trace_syscalls.c @@ -428,12 +428,12 @@ core_initcall(init_ftrace_syscalls); #ifdef CONFIG_PERF_EVENTS -static DECLARE_BITMAP(enabled_prof_enter_syscalls, NR_syscalls); -static DECLARE_BITMAP(enabled_prof_exit_syscalls, NR_syscalls); -static int sys_prof_refcount_enter; -static int sys_prof_refcount_exit; +static DECLARE_BITMAP(enabled_perf_enter_syscalls, NR_syscalls); +static DECLARE_BITMAP(enabled_perf_exit_syscalls, NR_syscalls); +static int sys_perf_refcount_enter; +static int sys_perf_refcount_exit; -static void prof_syscall_enter(struct pt_regs *regs, long id) +static void perf_syscall_enter(struct pt_regs *regs, long id) { struct syscall_metadata *sys_data; struct syscall_trace_enter *rec; @@ -443,7 +443,7 @@ static void prof_syscall_enter(struct pt_regs *regs, long id) int size; syscall_nr = syscall_get_nr(current, regs); - if (!test_bit(syscall_nr, enabled_prof_enter_syscalls)) + if (!test_bit(syscall_nr, enabled_perf_enter_syscalls)) return; sys_data = syscall_nr_to_meta(syscall_nr); @@ -455,11 +455,11 @@ static void prof_syscall_enter(struct pt_regs *regs, long id) size = ALIGN(size + sizeof(u32), sizeof(u64)); size -= sizeof(u32); - if (WARN_ONCE(size > FTRACE_MAX_PROFILE_SIZE, - "profile buffer not large enough")) + if (WARN_ONCE(size > PERF_MAX_TRACE_SIZE, + "perf buffer not large enough")) return; - rec = (struct syscall_trace_enter *)ftrace_perf_buf_prepare(size, + rec = (struct syscall_trace_enter *)perf_trace_buf_prepare(size, sys_data->enter_event->id, &rctx, &flags); if (!rec) return; @@ -467,10 +467,10 @@ static void prof_syscall_enter(struct pt_regs *regs, long id) rec->nr = syscall_nr; syscall_get_arguments(current, regs, 0, sys_data->nb_args, (unsigned long *)&rec->args); - ftrace_perf_buf_submit(rec, size, rctx, 0, 1, flags, regs); + perf_trace_buf_submit(rec, size, rctx, 0, 1, flags, regs); } -int prof_sysenter_enable(struct ftrace_event_call *call) +int perf_sysenter_enable(struct ftrace_event_call *call) { int ret = 0; int num; @@ -478,34 +478,34 @@ int prof_sysenter_enable(struct ftrace_event_call *call) num = ((struct syscall_metadata *)call->data)->syscall_nr; mutex_lock(&syscall_trace_lock); - if (!sys_prof_refcount_enter) - ret = register_trace_sys_enter(prof_syscall_enter); + if (!sys_perf_refcount_enter) + ret = register_trace_sys_enter(perf_syscall_enter); if (ret) { pr_info("event trace: Could not activate" "syscall entry trace point"); } else { - set_bit(num, enabled_prof_enter_syscalls); - sys_prof_refcount_enter++; + set_bit(num, enabled_perf_enter_syscalls); + sys_perf_refcount_enter++; } mutex_unlock(&syscall_trace_lock); return ret; } -void prof_sysenter_disable(struct ftrace_event_call *call) +void perf_sysenter_disable(struct ftrace_event_call *call) { int num; num = ((struct syscall_metadata *)call->data)->syscall_nr; mutex_lock(&syscall_trace_lock); - sys_prof_refcount_enter--; - clear_bit(num, enabled_prof_enter_syscalls); - if (!sys_prof_refcount_enter) - unregister_trace_sys_enter(prof_syscall_enter); + sys_perf_refcount_enter--; + clear_bit(num, enabled_perf_enter_syscalls); + if (!sys_perf_refcount_enter) + unregister_trace_sys_enter(perf_syscall_enter); mutex_unlock(&syscall_trace_lock); } -static void prof_syscall_exit(struct pt_regs *regs, long ret) +static void perf_syscall_exit(struct pt_regs *regs, long ret) { struct syscall_metadata *sys_data; struct syscall_trace_exit *rec; @@ -515,7 +515,7 @@ static void prof_syscall_exit(struct pt_regs *regs, long ret) int size; syscall_nr = syscall_get_nr(current, regs); - if (!test_bit(syscall_nr, enabled_prof_exit_syscalls)) + if (!test_bit(syscall_nr, enabled_perf_exit_syscalls)) return; sys_data = syscall_nr_to_meta(syscall_nr); @@ -530,11 +530,11 @@ static void prof_syscall_exit(struct pt_regs *regs, long ret) * Impossible, but be paranoid with the future * How to put this check outside runtime? */ - if (WARN_ONCE(size > FTRACE_MAX_PROFILE_SIZE, - "exit event has grown above profile buffer size")) + if (WARN_ONCE(size > PERF_MAX_TRACE_SIZE, + "exit event has grown above perf buffer size")) return; - rec = (struct syscall_trace_exit *)ftrace_perf_buf_prepare(size, + rec = (struct syscall_trace_exit *)perf_trace_buf_prepare(size, sys_data->exit_event->id, &rctx, &flags); if (!rec) return; @@ -542,10 +542,10 @@ static void prof_syscall_exit(struct pt_regs *regs, long ret) rec->nr = syscall_nr; rec->ret = syscall_get_return_value(current, regs); - ftrace_perf_buf_submit(rec, size, rctx, 0, 1, flags, regs); + perf_trace_buf_submit(rec, size, rctx, 0, 1, flags, regs); } -int prof_sysexit_enable(struct ftrace_event_call *call) +int perf_sysexit_enable(struct ftrace_event_call *call) { int ret = 0; int num; @@ -553,30 +553,30 @@ int prof_sysexit_enable(struct ftrace_event_call *call) num = ((struct syscall_metadata *)call->data)->syscall_nr; mutex_lock(&syscall_trace_lock); - if (!sys_prof_refcount_exit) - ret = register_trace_sys_exit(prof_syscall_exit); + if (!sys_perf_refcount_exit) + ret = register_trace_sys_exit(perf_syscall_exit); if (ret) { pr_info("event trace: Could not activate" "syscall exit trace point"); } else { - set_bit(num, enabled_prof_exit_syscalls); - sys_prof_refcount_exit++; + set_bit(num, enabled_perf_exit_syscalls); + sys_perf_refcount_exit++; } mutex_unlock(&syscall_trace_lock); return ret; } -void prof_sysexit_disable(struct ftrace_event_call *call) +void perf_sysexit_disable(struct ftrace_event_call *call) { int num; num = ((struct syscall_metadata *)call->data)->syscall_nr; mutex_lock(&syscall_trace_lock); - sys_prof_refcount_exit--; - clear_bit(num, enabled_prof_exit_syscalls); - if (!sys_prof_refcount_exit) - unregister_trace_sys_exit(prof_syscall_exit); + sys_perf_refcount_exit--; + clear_bit(num, enabled_perf_exit_syscalls); + if (!sys_perf_refcount_exit) + unregister_trace_sys_exit(perf_syscall_exit); mutex_unlock(&syscall_trace_lock); } -- cgit v1.2.3 From 3f379b03fbfddd20536389a85c6456f8233d1f8d Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Fri, 5 Mar 2010 15:03:25 -0800 Subject: ftrace: Replace read_barrier_depends() with rcu_dereference_raw() Replace the calls to read_barrier_depends() in ftrace_list_func() with rcu_dereference_raw() to improve readability. The reason that we use rcu_dereference_raw() here is that removed entries are never freed, instead they are simply leaked. This is one of a very few cases where use of rcu_dereference_raw() is the long-term right answer. And I don't yet know of any others. ;-) Signed-off-by: Paul E. McKenney Acked-by: Steven Rostedt Cc: Frederic Weisbecker Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: mathieu.desnoyers@polymtl.ca Cc: josh@joshtriplett.org Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: Valdis.Kletnieks@vt.edu Cc: dhowells@redhat.com LKML-Reference: <1267830207-9474-1-git-send-email-paulmck@linux.vnet.ibm.com> Signed-off-by: Ingo Molnar --- kernel/trace/ftrace.c | 22 +++++++++++++--------- 1 file changed, 13 insertions(+), 9 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index 83783579378f..8c5adc0e5db3 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -27,6 +27,7 @@ #include #include #include +#include #include @@ -88,18 +89,22 @@ ftrace_func_t ftrace_pid_function __read_mostly = ftrace_stub; static int ftrace_set_func(unsigned long *array, int *idx, char *buffer); #endif +/* + * Traverse the ftrace_list, invoking all entries. The reason that we + * can use rcu_dereference_raw() is that elements removed from this list + * are simply leaked, so there is no need to interact with a grace-period + * mechanism. The rcu_dereference_raw() calls are needed to handle + * concurrent insertions into the ftrace_list. + * + * Silly Alpha and silly pointer-speculation compiler optimizations! + */ static void ftrace_list_func(unsigned long ip, unsigned long parent_ip) { - struct ftrace_ops *op = ftrace_list; - - /* in case someone actually ports this to alpha! */ - read_barrier_depends(); + struct ftrace_ops *op = rcu_dereference_raw(ftrace_list); /*see above*/ while (op != &ftrace_list_end) { - /* silly alpha */ - read_barrier_depends(); op->func(ip, parent_ip); - op = op->next; + op = rcu_dereference_raw(op->next); /*see above*/ }; } @@ -154,8 +159,7 @@ static int __register_ftrace_function(struct ftrace_ops *ops) * the ops->next pointer is valid before another CPU sees * the ops pointer included into the ftrace_list. */ - smp_wmb(); - ftrace_list = ops; + rcu_assign_pointer(ftrace_list, ops); if (ftrace_enabled) { ftrace_func_t func; -- cgit v1.2.3 From 639fe4b12f92b54c9c3b38c82cdafaa38cfd3e63 Mon Sep 17 00:00:00 2001 From: Xiao Guangrong Date: Thu, 11 Mar 2010 15:30:35 +0800 Subject: perf: export perf_trace_regs and perf_arch_fetch_caller_regs Export perf_trace_regs and perf_arch_fetch_caller_regs since module will use these. Signed-off-by: Xiao Guangrong [ use EXPORT_PER_CPU_SYMBOL_GPL() ] Signed-off-by: Peter Zijlstra LKML-Reference: <4B989C1B.2090407@cn.fujitsu.com> Signed-off-by: Ingo Molnar --- kernel/trace/trace_event_perf.c | 1 + 1 file changed, 1 insertion(+) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_event_perf.c b/kernel/trace/trace_event_perf.c index f315b12a41d8..0709e4f75114 100644 --- a/kernel/trace/trace_event_perf.c +++ b/kernel/trace/trace_event_perf.c @@ -10,6 +10,7 @@ #include "trace.h" DEFINE_PER_CPU(struct pt_regs, perf_trace_regs); +EXPORT_PER_CPU_SYMBOL_GPL(perf_trace_regs); static char *perf_trace_buf; static char *perf_trace_buf_nmi; -- cgit v1.2.3 From 52fbe9cde7fdb5c6fac196d7ebd2d92d05ef3cd4 Mon Sep 17 00:00:00 2001 From: Lai Jiangshan Date: Mon, 8 Mar 2010 14:50:43 +0800 Subject: ring-buffer: Move disabled check into preempt disable section The ring buffer resizing and resetting relies on a schedule RCU action. The buffers are disabled, a synchronize_sched() is called and then the resize or reset takes place. But this only works if the disabling of the buffers are within the preempt disabled section, otherwise a window exists that the buffers can be written to while a reset or resize takes place. Cc: stable@kernel.org Reported-by: Li Zefan Signed-off-by: Lai Jiangshan LKML-Reference: <4B949E43.2010906@cn.fujitsu.com> Signed-off-by: Steven Rostedt --- kernel/trace/ring_buffer.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/ring_buffer.c b/kernel/trace/ring_buffer.c index 8c1b2d290718..54191d6ed195 100644 --- a/kernel/trace/ring_buffer.c +++ b/kernel/trace/ring_buffer.c @@ -2232,12 +2232,12 @@ ring_buffer_lock_reserve(struct ring_buffer *buffer, unsigned long length) if (ring_buffer_flags != RB_BUFFERS_ON) return NULL; - if (atomic_read(&buffer->record_disabled)) - return NULL; - /* If we are tracing schedule, we don't want to recurse */ resched = ftrace_preempt_disable(); + if (atomic_read(&buffer->record_disabled)) + goto out_nocheck; + if (trace_recursive_lock()) goto out_nocheck; @@ -2469,11 +2469,11 @@ int ring_buffer_write(struct ring_buffer *buffer, if (ring_buffer_flags != RB_BUFFERS_ON) return -EBUSY; - if (atomic_read(&buffer->record_disabled)) - return -EBUSY; - resched = ftrace_preempt_disable(); + if (atomic_read(&buffer->record_disabled)) + goto out; + cpu = raw_smp_processor_id(); if (!cpumask_test_cpu(cpu, buffer->cpumask)) -- cgit v1.2.3 From ea14eb714041d40fcc5180b5a586034503650149 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Fri, 12 Mar 2010 19:41:23 -0500 Subject: function-graph: Init curr_ret_stack with ret_stack If the graph tracer is active, and a task is forked but the allocating of the processes graph stack fails, it can cause crash later on. This is due to the temporary stack being NULL, but the curr_ret_stack variable is copied from the parent. If it is not -1, then in ftrace_graph_probe_sched_switch() the following: for (index = next->curr_ret_stack; index >= 0; index--) next->ret_stack[index].calltime += timestamp; Will cause a kernel OOPS. Found with Li Zefan's ftrace_stress_test. Cc: stable@kernel.org Signed-off-by: Steven Rostedt --- kernel/trace/ftrace.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel/trace') diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index d4d1238b096b..bb53edbb5c8c 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -3349,6 +3349,7 @@ void ftrace_graph_init_task(struct task_struct *t) { /* Make sure we do not use the parent ret_stack */ t->ret_stack = NULL; + t->curr_ret_stack = -1; if (ftrace_graph_active) { struct ftrace_ret_stack *ret_stack; @@ -3358,7 +3359,6 @@ void ftrace_graph_init_task(struct task_struct *t) GFP_KERNEL); if (!ret_stack) return; - t->curr_ret_stack = -1; atomic_set(&t->tracing_graph_pause, 0); atomic_set(&t->trace_overrun, 0); t->ftrace_timestamp = 0; -- cgit v1.2.3 From 283740c619d211e34572cc93c8cdba92ccbdb9cc Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Fri, 12 Mar 2010 19:48:41 -0500 Subject: tracing: Use same local variable when resetting the ring buffer In the ftrace code that resets the ring buffer it references the buffer with a local variable, but then uses the tr->buffer as the parameter to reset. If the wakeup tracer is running, which can switch the tr->buffer with the max saved buffer, this can break the requirement of disabling the buffer before the reset. buffer = tr->buffer; ring_buffer_record_disable(buffer); synchronize_sched(); __tracing_reset(tr->buffer, cpu); If the tr->buffer is swapped, then the reset is not happening to the buffer that was disabled. This will cause the ring buffer to fail. Found with Li Zefan's ftrace_stress_test. Cc: stable@kernel.org Reported-by: Lai Jiangshan Signed-off-by: Steven Rostedt --- kernel/trace/trace.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index 6af8d7bc953b..60de37bd0f75 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -840,10 +840,10 @@ out: mutex_unlock(&trace_types_lock); } -static void __tracing_reset(struct trace_array *tr, int cpu) +static void __tracing_reset(struct ring_buffer *buffer, int cpu) { ftrace_disable_cpu(); - ring_buffer_reset_cpu(tr->buffer, cpu); + ring_buffer_reset_cpu(buffer, cpu); ftrace_enable_cpu(); } @@ -855,7 +855,7 @@ void tracing_reset(struct trace_array *tr, int cpu) /* Make sure all commits have finished */ synchronize_sched(); - __tracing_reset(tr, cpu); + __tracing_reset(buffer, cpu); ring_buffer_record_enable(buffer); } @@ -873,7 +873,7 @@ void tracing_reset_online_cpus(struct trace_array *tr) tr->time_start = ftrace_now(tr->cpu); for_each_online_cpu(cpu) - __tracing_reset(tr, cpu); + __tracing_reset(buffer, cpu); ring_buffer_record_enable(buffer); } -- cgit v1.2.3 From a2f8071428ed9a0f06865f417c962421c9a6b488 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Fri, 12 Mar 2010 19:56:00 -0500 Subject: tracing: Disable buffer switching when starting or stopping trace When the trace iterator is read, tracing_start() and tracing_stop() is called to stop tracing while the iterator is processing the trace output. These functions disable both the standard buffer and the max latency buffer. But if the wakeup tracer is running, it can switch these buffers between the two disables: buffer = global_trace.buffer; if (buffer) ring_buffer_record_disable(buffer); <<<--------- swap happens here buffer = max_tr.buffer; if (buffer) ring_buffer_record_disable(buffer); What happens is that we disabled the same buffer twice. On tracing_start() we can enable the same buffer twice. All ring_buffer_record_disable() must be matched with a ring_buffer_record_enable() or the buffer can be disable permanently, or enable prematurely, and cause a bug where a reset happens while a trace is commiting. This patch protects these two by taking the ftrace_max_lock to prevent a switch from occurring. Found with Li Zefan's ftrace_stress_test. Cc: stable@kernel.org Reported-by: Lai Jiangshan Signed-off-by: Steven Rostedt --- kernel/trace/trace.c | 9 +++++++++ 1 file changed, 9 insertions(+) (limited to 'kernel/trace') diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index 60de37bd0f75..484337d33959 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -950,6 +950,8 @@ void tracing_start(void) goto out; } + /* Prevent the buffers from switching */ + arch_spin_lock(&ftrace_max_lock); buffer = global_trace.buffer; if (buffer) @@ -959,6 +961,8 @@ void tracing_start(void) if (buffer) ring_buffer_record_enable(buffer); + arch_spin_unlock(&ftrace_max_lock); + ftrace_start(); out: spin_unlock_irqrestore(&tracing_start_lock, flags); @@ -980,6 +984,9 @@ void tracing_stop(void) if (trace_stop_count++) goto out; + /* Prevent the buffers from switching */ + arch_spin_lock(&ftrace_max_lock); + buffer = global_trace.buffer; if (buffer) ring_buffer_record_disable(buffer); @@ -988,6 +995,8 @@ void tracing_stop(void) if (buffer) ring_buffer_record_disable(buffer); + arch_spin_unlock(&ftrace_max_lock); + out: spin_unlock_irqrestore(&tracing_start_lock, flags); } -- cgit v1.2.3 From b6345879ccbd9b92864fbd7eb8ac48acdb4d6b15 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Fri, 12 Mar 2010 20:03:30 -0500 Subject: tracing: Do not record user stack trace from NMI context A bug was found with Li Zefan's ftrace_stress_test that caused applications to segfault during the test. Placing a tracing_off() in the segfault code, and examining several traces, I found that the following was always the case. The lock tracer was enabled (lockdep being required) and userstack was enabled. Testing this out, I just enabled the two, but that was not good enough. I needed to run something else that could trigger it. Running a load like hackbench did not work, but executing a new program would. The following would trigger the segfault within seconds: # echo 1 > /debug/tracing/options/userstacktrace # echo 1 > /debug/tracing/events/lock/enable # while :; do ls > /dev/null ; done Enabling the function graph tracer and looking at what was happening I finally noticed that all cashes happened just after an NMI. 1) | copy_user_handle_tail() { 1) | bad_area_nosemaphore() { 1) | __bad_area_nosemaphore() { 1) | no_context() { 1) | fixup_exception() { 1) 0.319 us | search_exception_tables(); 1) 0.873 us | } [...] 1) 0.314 us | __rcu_read_unlock(); 1) 0.325 us | native_apic_mem_write(); 1) 0.943 us | } 1) 0.304 us | rcu_nmi_exit(); [...] 1) 0.479 us | find_vma(); 1) | bad_area() { 1) | __bad_area() { After capturing several traces of failures, all of them happened after an NMI. Curious about this, I added a trace_printk() to the NMI handler to read the regs->ip to see where the NMI happened. In which I found out it was here: ffffffff8135b660 : ffffffff8135b660: 48 83 ec 78 sub $0x78,%rsp ffffffff8135b664: e8 97 01 00 00 callq ffffffff8135b800 What was happening is that the NMI would happen at the place that a page fault occurred. It would call rcu_read_lock() which was traced by the lock events, and the user_stack_trace would run. This would trigger a page fault inside the NMI. I do not see where the CR2 register is saved or restored in NMI handling. This means that it would corrupt the page fault handling that the NMI interrupted. The reason the while loop of ls helped trigger the bug, was that each execution of ls would cause lots of pages to be faulted in, and increase the chances of the race happening. The simple solution is to not allow user stack traces in NMI context. After this patch, I ran the above "ls" test for a couple of hours without any issues. Without this patch, the bug would trigger in less than a minute. Cc: stable@kernel.org Reported-by: Li Zefan Signed-off-by: Steven Rostedt --- kernel/trace/trace.c | 7 +++++++ 1 file changed, 7 insertions(+) (limited to 'kernel/trace') diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index 484337d33959..e52683f7c3b2 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -1284,6 +1284,13 @@ ftrace_trace_userstack(struct ring_buffer *buffer, unsigned long flags, int pc) if (!(trace_flags & TRACE_ITER_USERSTACKTRACE)) return; + /* + * NMIs can not handle page faults, even with fix ups. + * The save user stack can (and often does) fault. + */ + if (unlikely(in_nmi())) + return; + event = trace_buffer_lock_reserve(buffer, TRACE_USER_STACK, sizeof(*entry), flags, pc); if (!event) -- cgit v1.2.3 From dcd5c1662db59a6b82942f47fb6ac9dd63f6d3dd Mon Sep 17 00:00:00 2001 From: Frederic Weisbecker Date: Tue, 16 Mar 2010 01:05:02 +0100 Subject: perf: Fix unexported generic perf_arch_fetch_caller_regs perf_arch_fetch_caller_regs() is exported for the overriden x86 version, but not for the generic weak version. As a general rule, weak functions should not have their symbol exported in the same file they are defined. So let's export it on trace_event_perf.c as it is used by trace events only. This fixes: ERROR: ".perf_arch_fetch_caller_regs" [fs/xfs/xfs.ko] undefined! ERROR: ".perf_arch_fetch_caller_regs" [arch/powerpc/platforms/cell/spufs/spufs.ko] undefined! -v2: And also only build it if trace events are enabled. -v3: Fix changelog mistake Reported-by: Stephen Rothwell Signed-off-by: Frederic Weisbecker Cc: Peter Zijlstra Cc: Xiao Guangrong Cc: Paul Mackerras LKML-Reference: <1268697902-9518-1-git-send-regression-fweisbec@gmail.com> Signed-off-by: Ingo Molnar --- kernel/trace/trace_event_perf.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_event_perf.c b/kernel/trace/trace_event_perf.c index 0709e4f75114..7d79a10c3cde 100644 --- a/kernel/trace/trace_event_perf.c +++ b/kernel/trace/trace_event_perf.c @@ -12,6 +12,8 @@ DEFINE_PER_CPU(struct pt_regs, perf_trace_regs); EXPORT_PER_CPU_SYMBOL_GPL(perf_trace_regs); +EXPORT_SYMBOL_GPL(perf_arch_fetch_caller_regs); + static char *perf_trace_buf; static char *perf_trace_buf_nmi; -- cgit v1.2.3 From 2271048d1b3b0aabf83d25b29c20646dcabedc05 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Thu, 18 Mar 2010 17:54:19 -0400 Subject: ring-buffer: Do 8 byte alignment for 64 bit that can not handle 4 byte align The ring buffer uses 4 byte alignment while recording events into the buffer, even on 64bit machines. This saves space when there are lots of events being recorded at 4 byte boundaries. The ring buffer has a zero copy method to write into the buffer, with the reserving of space and then committing it. This may cause problems when writing an 8 byte word into a 4 byte alignment (not 8). For x86 and PPC this is not an issue, but on some architectures this would cause an out-of-alignment exception. This patch uses CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS to determine if it is OK to use 4 byte alignments on 64 bit machines. If it is not, it forces the ring buffer event header to be 8 bytes and not 4, and will align the length of the data to be 8 byte aligned. This keeps the data payload at 8 byte alignments and will allow these machines to run without issue. The trick to this is that the header can be either 4 bytes or 8 bytes depending on the length of the data payload. The 4 byte header has a length field that supports up to 112 bytes. If the length of the data is more than 112, the length field is set to zero, and the actual length is stored in the next 4 bytes after the header. When CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS is not set, the code forces zero in the 4 byte header forcing the length to be stored in the 4 byte array, even with a small data load. It also forces the length of the data load to be 8 byte aligned. The combination of these two guarantee that the data is always at 8 byte alignment. Tested-by: Frederic Weisbecker (on sparc64) Reported-by: Frederic Weisbecker Acked-by: David S. Miller Signed-off-by: Steven Rostedt --- kernel/trace/ring_buffer.c | 14 +++++++++++--- 1 file changed, 11 insertions(+), 3 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/ring_buffer.c b/kernel/trace/ring_buffer.c index 05a9f83b8819..d1187ef20caf 100644 --- a/kernel/trace/ring_buffer.c +++ b/kernel/trace/ring_buffer.c @@ -207,6 +207,14 @@ EXPORT_SYMBOL_GPL(tracing_is_on); #define RB_MAX_SMALL_DATA (RB_ALIGNMENT * RINGBUF_TYPE_DATA_TYPE_LEN_MAX) #define RB_EVNT_MIN_SIZE 8U /* two 32bit words */ +#if !defined(CONFIG_64BIT) || defined(CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS) +# define RB_FORCE_8BYTE_ALIGNMENT 0 +# define RB_ARCH_ALIGNMENT RB_ALIGNMENT +#else +# define RB_FORCE_8BYTE_ALIGNMENT 1 +# define RB_ARCH_ALIGNMENT 8U +#endif + /* define RINGBUF_TYPE_DATA for 'case RINGBUF_TYPE_DATA:' */ #define RINGBUF_TYPE_DATA 0 ... RINGBUF_TYPE_DATA_TYPE_LEN_MAX @@ -1547,7 +1555,7 @@ rb_update_event(struct ring_buffer_event *event, case 0: length -= RB_EVNT_HDR_SIZE; - if (length > RB_MAX_SMALL_DATA) + if (length > RB_MAX_SMALL_DATA || RB_FORCE_8BYTE_ALIGNMENT) event->array[0] = length; else event->type_len = DIV_ROUND_UP(length, RB_ALIGNMENT); @@ -1722,11 +1730,11 @@ static unsigned rb_calculate_event_length(unsigned length) if (!length) length = 1; - if (length > RB_MAX_SMALL_DATA) + if (length > RB_MAX_SMALL_DATA || RB_FORCE_8BYTE_ALIGNMENT) length += sizeof(event.array[0]); length += RB_EVNT_HDR_SIZE; - length = ALIGN(length, RB_ALIGNMENT); + length = ALIGN(length, RB_ARCH_ALIGNMENT); return length; } -- cgit v1.2.3