From 60a7ecf42661f2b22168751298592da6ee210c9e Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Wed, 5 Nov 2008 16:05:44 -0500 Subject: ftrace: add quick function trace stop Impact: quick start and stop of function tracer This patch adds a way to disable the function tracer quickly without the need to run kstop_machine. It adds a new variable called function_trace_stop which will stop the calls to functions from mcount when set. This is just an on/off switch and does not handle recursion like preempt_disable(). It's main purpose is to help other tracers/debuggers start and stop tracing fuctions without the need to call kstop_machine. The config option HAVE_FUNCTION_TRACE_MCOUNT_TEST is added for archs that implement the testing of the function_trace_stop in the mcount arch dependent code. Otherwise, the test is done in the C code. x86 is the only arch at the moment that supports this. Signed-off-by: Steven Rostedt Signed-off-by: Ingo Molnar --- kernel/trace/ftrace.c | 47 +++++++++++++++++++++++++++++++++++++---------- 1 file changed, 37 insertions(+), 10 deletions(-) (limited to 'kernel/trace/ftrace.c') diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index 4a39d24568c..896c71f0f4c 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -47,6 +47,9 @@ int ftrace_enabled __read_mostly; static int last_ftrace_enabled; +/* Quick disabling of function tracer. */ +int function_trace_stop; + /* * ftrace_disabled is set when an anomaly is discovered. * ftrace_disabled is much stronger than ftrace_enabled. @@ -63,6 +66,7 @@ static struct ftrace_ops ftrace_list_end __read_mostly = static struct ftrace_ops *ftrace_list __read_mostly = &ftrace_list_end; ftrace_func_t ftrace_trace_function __read_mostly = ftrace_stub; +ftrace_func_t __ftrace_trace_function __read_mostly = ftrace_stub; static void ftrace_list_func(unsigned long ip, unsigned long parent_ip) { @@ -88,8 +92,23 @@ static void ftrace_list_func(unsigned long ip, unsigned long parent_ip) void clear_ftrace_function(void) { ftrace_trace_function = ftrace_stub; + __ftrace_trace_function = ftrace_stub; } +#ifndef CONFIG_HAVE_FUNCTION_TRACE_MCOUNT_TEST +/* + * For those archs that do not test ftrace_trace_stop in their + * mcount call site, we need to do it from C. + */ +static void ftrace_test_stop_func(unsigned long ip, unsigned long parent_ip) +{ + if (function_trace_stop) + return; + + __ftrace_trace_function(ip, parent_ip); +} +#endif + static int __register_ftrace_function(struct ftrace_ops *ops) { /* should not be called from interrupt context */ @@ -110,10 +129,18 @@ static int __register_ftrace_function(struct ftrace_ops *ops) * For one func, simply call it directly. * For more than one func, call the chain. */ +#ifdef CONFIG_HAVE_FUNCTION_TRACE_MCOUNT_TEST if (ops->next == &ftrace_list_end) ftrace_trace_function = ops->func; else ftrace_trace_function = ftrace_list_func; +#else + if (ops->next == &ftrace_list_end) + __ftrace_trace_function = ops->func; + else + __ftrace_trace_function = ftrace_list_func; + ftrace_trace_function = ftrace_test_stop_func; +#endif } spin_unlock(&ftrace_lock); @@ -526,7 +553,7 @@ static void ftrace_run_update_code(int command) } static ftrace_func_t saved_ftrace_func; -static int ftrace_start; +static int ftrace_start_up; static DEFINE_MUTEX(ftrace_start_lock); static void ftrace_startup(void) @@ -537,8 +564,8 @@ static void ftrace_startup(void) return; mutex_lock(&ftrace_start_lock); - ftrace_start++; - if (ftrace_start == 1) + ftrace_start_up++; + if (ftrace_start_up == 1) command |= FTRACE_ENABLE_CALLS; if (saved_ftrace_func != ftrace_trace_function) { @@ -562,8 +589,8 @@ static void ftrace_shutdown(void) return; mutex_lock(&ftrace_start_lock); - ftrace_start--; - if (!ftrace_start) + ftrace_start_up--; + if (!ftrace_start_up) command |= FTRACE_DISABLE_CALLS; if (saved_ftrace_func != ftrace_trace_function) { @@ -589,8 +616,8 @@ static void ftrace_startup_sysctl(void) mutex_lock(&ftrace_start_lock); /* Force update next time */ saved_ftrace_func = NULL; - /* ftrace_start is true if we want ftrace running */ - if (ftrace_start) + /* ftrace_start_up is true if we want ftrace running */ + if (ftrace_start_up) command |= FTRACE_ENABLE_CALLS; ftrace_run_update_code(command); @@ -605,8 +632,8 @@ static void ftrace_shutdown_sysctl(void) return; mutex_lock(&ftrace_start_lock); - /* ftrace_start is true if ftrace is running */ - if (ftrace_start) + /* ftrace_start_up is true if ftrace is running */ + if (ftrace_start_up) command |= FTRACE_DISABLE_CALLS; ftrace_run_update_code(command); @@ -1186,7 +1213,7 @@ ftrace_regex_release(struct inode *inode, struct file *file, int enable) mutex_lock(&ftrace_sysctl_lock); mutex_lock(&ftrace_start_lock); - if (iter->filtered && ftrace_start && ftrace_enabled) + if (iter->filtered && ftrace_start_up && ftrace_enabled) ftrace_run_update_code(FTRACE_ENABLE_CALLS); mutex_unlock(&ftrace_start_lock); mutex_unlock(&ftrace_sysctl_lock); -- cgit v1.2.3-70-g09d2 From 0183fb1c94b74862b073590fc52c56b7364b7bad Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Fri, 7 Nov 2008 22:36:02 -0500 Subject: ftrace: fix set_ftrace_filter Impact: fix of output of set_ftrace_filter Commit ftrace: do not show freed records in available_filter_functions Removed a bit too much from the set_ftrace_filter code, where we now see all functions in the set_ftrace_filter file even when we set a filter. Signed-off-by: Steven Rostedt Signed-off-by: Ingo Molnar --- kernel/trace/ftrace.c | 3 +++ 1 file changed, 3 insertions(+) (limited to 'kernel/trace/ftrace.c') diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index 896c71f0f4c..4d2e751bfb1 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -765,6 +765,9 @@ t_next(struct seq_file *m, void *v, loff_t *pos) ((iter->flags & FTRACE_ITER_FAILURES) && !(rec->flags & FTRACE_FL_FAILED)) || + ((iter->flags & FTRACE_ITER_FILTER) && + !(rec->flags & FTRACE_FL_FILTER)) || + ((iter->flags & FTRACE_ITER_NOTRACE) && !(rec->flags & FTRACE_FL_NOTRACE))) { rec = NULL; -- cgit v1.2.3-70-g09d2 From 15e6cb3673ea6277999642802406a764b49391b0 Mon Sep 17 00:00:00 2001 From: Frederic Weisbecker Date: Tue, 11 Nov 2008 07:14:25 +0100 Subject: tracing: add a tracer to catch execution time of kernel functions Impact: add new tracing plugin which can trace full (entry+exit) function calls This tracer uses the low level function return ftrace plugin to measure the execution time of the kernel functions. The first field is the caller of the function, the second is the measured function, and the last one is the execution time in nanoseconds. - v3: - HAVE_FUNCTION_RET_TRACER have been added. Each arch that support ftrace return should enable it. - ftrace_return_stub becomes ftrace_stub. - CONFIG_FUNCTION_RET_TRACER depends now on CONFIG_FUNCTION_TRACER - Return traces printing can be used for other tracers on trace.c - Adapt to the new tracing API (no more ctrl_update callback) - Correct the check of "disabled" during insertion. - Minor changes... Signed-off-by: Frederic Weisbecker Cc: Steven Rostedt Signed-off-by: Ingo Molnar --- kernel/trace/Kconfig | 14 ++++++ kernel/trace/Makefile | 1 + kernel/trace/ftrace.c | 16 +++++++ kernel/trace/trace.c | 65 +++++++++++++++++++++++---- kernel/trace/trace.h | 35 +++++++++++++++ kernel/trace/trace_functions_return.c | 82 +++++++++++++++++++++++++++++++++++ 6 files changed, 205 insertions(+), 8 deletions(-) create mode 100644 kernel/trace/trace_functions_return.c (limited to 'kernel/trace/ftrace.c') diff --git a/kernel/trace/Kconfig b/kernel/trace/Kconfig index fc4febc3334..d986216c832 100644 --- a/kernel/trace/Kconfig +++ b/kernel/trace/Kconfig @@ -9,6 +9,9 @@ config NOP_TRACER config HAVE_FUNCTION_TRACER bool +config HAVE_FUNCTION_RET_TRACER + bool + config HAVE_FUNCTION_TRACE_MCOUNT_TEST bool help @@ -54,6 +57,17 @@ config FUNCTION_TRACER (the bootup default), then the overhead of the instructions is very small and not measurable even in micro-benchmarks. +config FUNCTION_RET_TRACER + bool "Kernel Function return Tracer" + depends on !DYNAMIC_FTRACE + depends on HAVE_FUNCTION_RET_TRACER + depends on FUNCTION_TRACER + help + Enable the kernel to trace a function at its return. + It's first purpose is to trace the duration of functions. + This is done by setting the current return address on the thread + info structure of the current task. + config IRQSOFF_TRACER bool "Interrupts-off Latency Tracer" default n diff --git a/kernel/trace/Makefile b/kernel/trace/Makefile index c8228b1a49e..3e1f361bbc1 100644 --- a/kernel/trace/Makefile +++ b/kernel/trace/Makefile @@ -24,5 +24,6 @@ obj-$(CONFIG_NOP_TRACER) += trace_nop.o obj-$(CONFIG_STACK_TRACER) += trace_stack.o obj-$(CONFIG_MMIOTRACE) += trace_mmiotrace.o obj-$(CONFIG_BOOT_TRACER) += trace_boot.o +obj-$(CONFIG_FUNCTION_RET_TRACER) += trace_functions_return.o libftrace-y := ftrace.o diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index 4d2e751bfb1..f03fe74ecd6 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -1484,3 +1484,19 @@ ftrace_enable_sysctl(struct ctl_table *table, int write, return ret; } +#ifdef CONFIG_FUNCTION_RET_TRACER +trace_function_return_t ftrace_function_return = + (trace_function_return_t)ftrace_stub; +void register_ftrace_return(trace_function_return_t func) +{ + ftrace_function_return = func; +} + +void unregister_ftrace_return(void) +{ + ftrace_function_return = (trace_function_return_t)ftrace_stub; +} +#endif + + + diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index 216bbe7547a..a3f7ae9cd8e 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -244,13 +244,6 @@ unsigned long nsecs_to_usecs(unsigned long nsecs) return nsecs / 1000; } -/* - * TRACE_ITER_SYM_MASK masks the options in trace_flags that - * control the output of kernel symbols. - */ -#define TRACE_ITER_SYM_MASK \ - (TRACE_ITER_PRINT_PARENT|TRACE_ITER_SYM_OFFSET|TRACE_ITER_SYM_ADDR) - /* These must match the bit postions in trace_iterator_flags */ static const char *trace_options[] = { "print-parent", @@ -810,6 +803,35 @@ trace_function(struct trace_array *tr, struct trace_array_cpu *data, ring_buffer_unlock_commit(tr->buffer, event, irq_flags); } +#ifdef CONFIG_FUNCTION_RET_TRACER +static void __trace_function_return(struct trace_array *tr, + struct trace_array_cpu *data, + struct ftrace_retfunc *trace, + unsigned long flags, + int pc) +{ + struct ring_buffer_event *event; + struct ftrace_ret_entry *entry; + unsigned long irq_flags; + + if (unlikely(local_read(&__get_cpu_var(ftrace_cpu_disabled)))) + return; + + event = ring_buffer_lock_reserve(global_trace.buffer, sizeof(*entry), + &irq_flags); + if (!event) + return; + entry = ring_buffer_event_data(event); + tracing_generic_entry_update(&entry->ent, flags, pc); + entry->ent.type = TRACE_FN_RET; + entry->ip = trace->func; + entry->parent_ip = trace->ret; + entry->rettime = trace->rettime; + entry->calltime = trace->calltime; + ring_buffer_unlock_commit(global_trace.buffer, event, irq_flags); +} +#endif + void ftrace(struct trace_array *tr, struct trace_array_cpu *data, unsigned long ip, unsigned long parent_ip, unsigned long flags, @@ -1038,6 +1060,29 @@ function_trace_call(unsigned long ip, unsigned long parent_ip) raw_local_irq_restore(flags); } +#ifdef CONFIG_FUNCTION_RET_TRACER +void trace_function_return(struct ftrace_retfunc *trace) +{ + struct trace_array *tr = &global_trace; + struct trace_array_cpu *data; + unsigned long flags; + long disabled; + int cpu; + int pc; + + raw_local_irq_save(flags); + cpu = raw_smp_processor_id(); + data = tr->data[cpu]; + disabled = atomic_inc_return(&data->disabled); + if (likely(disabled == 1)) { + pc = preempt_count(); + __trace_function_return(tr, data, trace, flags, pc); + } + atomic_dec(&data->disabled); + raw_local_irq_restore(flags); +} +#endif /* CONFIG_FUNCTION_RET_TRACER */ + static struct ftrace_ops trace_ops __read_mostly = { .func = function_trace_call, @@ -1285,7 +1330,7 @@ seq_print_sym_offset(struct trace_seq *s, const char *fmt, # define IP_FMT "%016lx" #endif -static int +int seq_print_ip_sym(struct trace_seq *s, unsigned long ip, unsigned long sym_flags) { int ret; @@ -1738,6 +1783,10 @@ static enum print_line_t print_trace_fmt(struct trace_iterator *iter) trace_seq_print_cont(s, iter); break; } + case TRACE_FN_RET: { + return print_return_function(iter); + break; + } } return TRACE_TYPE_HANDLED; } diff --git a/kernel/trace/trace.h b/kernel/trace/trace.h index 978145088fb..e40ce0c1469 100644 --- a/kernel/trace/trace.h +++ b/kernel/trace/trace.h @@ -22,6 +22,7 @@ enum trace_type { TRACE_MMIO_RW, TRACE_MMIO_MAP, TRACE_BOOT, + TRACE_FN_RET, __TRACE_LAST_TYPE }; @@ -48,6 +49,15 @@ struct ftrace_entry { unsigned long ip; unsigned long parent_ip; }; + +/* Function return entry */ +struct ftrace_ret_entry { + struct trace_entry ent; + unsigned long ip; + unsigned long parent_ip; + unsigned long long calltime; + unsigned long long rettime; +}; extern struct tracer boot_tracer; /* @@ -218,6 +228,7 @@ extern void __ftrace_bad_type(void); IF_ASSIGN(var, ent, struct trace_mmiotrace_map, \ TRACE_MMIO_MAP); \ IF_ASSIGN(var, ent, struct trace_boot, TRACE_BOOT); \ + IF_ASSIGN(var, ent, struct ftrace_ret_entry, TRACE_FN_RET); \ __ftrace_bad_type(); \ } while (0) @@ -321,6 +332,8 @@ void trace_function(struct trace_array *tr, unsigned long ip, unsigned long parent_ip, unsigned long flags, int pc); +void +trace_function_return(struct ftrace_retfunc *trace); void tracing_start_cmdline_record(void); void tracing_stop_cmdline_record(void); @@ -393,6 +406,10 @@ extern void *head_page(struct trace_array_cpu *data); extern int trace_seq_printf(struct trace_seq *s, const char *fmt, ...); extern void trace_seq_print_cont(struct trace_seq *s, struct trace_iterator *iter); + +extern int +seq_print_ip_sym(struct trace_seq *s, unsigned long ip, + unsigned long sym_flags); extern ssize_t trace_seq_to_user(struct trace_seq *s, char __user *ubuf, size_t cnt); extern long ns2usecs(cycle_t nsec); @@ -400,6 +417,17 @@ extern int trace_vprintk(unsigned long ip, const char *fmt, va_list args); extern unsigned long trace_flags; +/* Standard output formatting function used for function return traces */ +#ifdef CONFIG_FUNCTION_RET_TRACER +extern enum print_line_t print_return_function(struct trace_iterator *iter); +#else +static inline enum print_line_t +print_return_function(struct trace_iterator *iter) +{ + return TRACE_TYPE_UNHANDLED; +} +#endif + /* * trace_iterator_flags is an enumeration that defines bit * positions into trace_flags that controls the output. @@ -422,6 +450,13 @@ enum trace_iterator_flags { TRACE_ITER_PREEMPTONLY = 0x800, }; +/* + * TRACE_ITER_SYM_MASK masks the options in trace_flags that + * control the output of kernel symbols. + */ +#define TRACE_ITER_SYM_MASK \ + (TRACE_ITER_PRINT_PARENT|TRACE_ITER_SYM_OFFSET|TRACE_ITER_SYM_ADDR) + extern struct tracer nop_trace; /** diff --git a/kernel/trace/trace_functions_return.c b/kernel/trace/trace_functions_return.c new file mode 100644 index 00000000000..7680b21537d --- /dev/null +++ b/kernel/trace/trace_functions_return.c @@ -0,0 +1,82 @@ +/* + * + * Function return tracer. + * Copyright (c) 2008 Frederic Weisbecker + * Mostly borrowed from function tracer which + * is Copyright (c) Steven Rostedt + * + */ +#include +#include +#include +#include + +#include "trace.h" + + +static void start_return_trace(struct trace_array *tr) +{ + register_ftrace_return(&trace_function_return); +} + +static void stop_return_trace(struct trace_array *tr) +{ + unregister_ftrace_return(); +} + +static void return_trace_init(struct trace_array *tr) +{ + int cpu; + for_each_online_cpu(cpu) + tracing_reset(tr, cpu); + + start_return_trace(tr); +} + +static void return_trace_reset(struct trace_array *tr) +{ + stop_return_trace(tr); +} + + +enum print_line_t +print_return_function(struct trace_iterator *iter) +{ + struct trace_seq *s = &iter->seq; + struct trace_entry *entry = iter->ent; + struct ftrace_ret_entry *field; + int ret; + + if (entry->type == TRACE_FN_RET) { + trace_assign_type(field, entry); + ret = trace_seq_printf(s, "%pF -> ", (void *)field->parent_ip); + if (!ret) + return TRACE_TYPE_PARTIAL_LINE; + ret = seq_print_ip_sym(s, field->ip, + trace_flags & TRACE_ITER_SYM_MASK); + if (!ret) + return TRACE_TYPE_PARTIAL_LINE; + ret = trace_seq_printf(s, " (%llu ns)\n", + field->rettime - field->calltime); + if (!ret) + return TRACE_TYPE_PARTIAL_LINE; + else + return TRACE_TYPE_HANDLED; + } + return TRACE_TYPE_UNHANDLED; +} + +static struct tracer return_trace __read_mostly = +{ + .name = "return", + .init = return_trace_init, + .reset = return_trace_reset, + .print_line = print_return_function +}; + +static __init int init_return_trace(void) +{ + return register_tracer(&return_trace); +} + +device_initcall(init_return_trace); -- cgit v1.2.3-70-g09d2 From b3535c6390f27d04273e4eee0bc687f171fbf5f4 Mon Sep 17 00:00:00 2001 From: walimis Date: Fri, 14 Nov 2008 00:21:02 +0800 Subject: ftrace: remove unnecessary if condition of __unregister_ftrace_function Because it has goto out before ftrace_list == &ftrace_list_end, that's to say, we never meet this condition. Signed-off-by: walimis Acked-by: Steven Rostedt Signed-off-by: Ingo Molnar --- kernel/trace/ftrace.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) (limited to 'kernel/trace/ftrace.c') diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index beb21a51e1e..54cb9a7d15e 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -179,8 +179,7 @@ static int __unregister_ftrace_function(struct ftrace_ops *ops) if (ftrace_enabled) { /* If we only have one func left, then call that directly */ - if (ftrace_list == &ftrace_list_end || - ftrace_list->next == &ftrace_list_end) + if (ftrace_list->next == &ftrace_list_end) ftrace_trace_function = ftrace_list->func; } -- cgit v1.2.3-70-g09d2 From f3c7ac40a99f4044b843e6e2c4f46ab2d354c563 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Fri, 14 Nov 2008 16:21:19 -0800 Subject: ftrace: remove condition from ftrace_record_ip Impact: let module functions be recorded when dyn ftrace not enabled When dynamic ftrace had a daemon and a hash to record the locations of mcount callers at run time, the recording needed to stop when ftrace was disabled. But now that the recording is done at compile time and the ftrace_record_ip is only called at boot up and when a module is loaded, we no longer need to check if ftrace_enabled is set. In fact, this breaks module load if it is not set because we skip over module functions. Signed-off-by: Steven Rostedt Signed-off-by: Ingo Molnar --- kernel/trace/ftrace.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel/trace/ftrace.c') diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index 54cb9a7d15e..3160254f6c7 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -334,7 +334,7 @@ ftrace_record_ip(unsigned long ip) { struct dyn_ftrace *rec; - if (!ftrace_enabled || ftrace_disabled) + if (ftrace_disabled) return NULL; rec = ftrace_alloc_dyn_node(ip); -- cgit v1.2.3-70-g09d2 From b17e8a37a13d0e87165054714434534bb7e69f2d Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Fri, 14 Nov 2008 16:21:19 -0800 Subject: ftrace: disable ftrace on anomalies in trace start and stop Impact: robust feature to disable ftrace on start or stop tracing on error Currently only the initial conversion to nops will disable ftrace on an anomaly. But if an anomaly happens on start or stopping of the tracer, it will silently fail. This patch adds a check there too, to disable ftrace and warn if the conversion fails. Signed-off-by: Steven Rostedt Signed-off-by: Ingo Molnar --- kernel/trace/ftrace.c | 81 ++++++++++++++++++++++++++++----------------------- 1 file changed, 44 insertions(+), 37 deletions(-) (limited to 'kernel/trace/ftrace.c') diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index 3160254f6c7..d5bd21f3952 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -348,6 +348,47 @@ ftrace_record_ip(unsigned long ip) return rec; } +static void print_ip_ins(const char *fmt, unsigned char *p) +{ + int i; + + printk(KERN_CONT "%s", fmt); + + for (i = 0; i < MCOUNT_INSN_SIZE; i++) + printk(KERN_CONT "%s%02x", i ? ":" : "", p[i]); +} + +static void ftrace_bug(int failed, unsigned long ip, + unsigned char *expected, + unsigned char *replace) +{ + switch (failed) { + case -EFAULT: + FTRACE_WARN_ON_ONCE(1); + pr_info("ftrace faulted on modifying "); + print_ip_sym(ip); + break; + case -EINVAL: + FTRACE_WARN_ON_ONCE(1); + pr_info("ftrace failed to modify "); + print_ip_sym(ip); + print_ip_ins(" expected: ", expected); + print_ip_ins(" actual: ", (unsigned char *)ip); + print_ip_ins(" replace: ", replace); + printk(KERN_CONT "\n"); + break; + case -EPERM: + FTRACE_WARN_ON_ONCE(1); + pr_info("ftrace faulted on writing "); + print_ip_sym(ip); + break; + default: + FTRACE_WARN_ON_ONCE(1); + pr_info("ftrace faulted on unknown error "); + print_ip_sym(ip); + } +} + #define FTRACE_ADDR ((long)(ftrace_caller)) static int @@ -465,22 +506,13 @@ static void ftrace_replace_code(int enable) if ((system_state == SYSTEM_BOOTING) || !core_kernel_text(rec->ip)) { ftrace_free_rec(rec); - } + } else + ftrace_bug(failed, rec->ip, old, new); } } } } -static void print_ip_ins(const char *fmt, unsigned char *p) -{ - int i; - - printk(KERN_CONT "%s", fmt); - - for (i = 0; i < MCOUNT_INSN_SIZE; i++) - printk(KERN_CONT "%s%02x", i ? ":" : "", p[i]); -} - static int ftrace_code_disable(struct dyn_ftrace *rec) { @@ -495,32 +527,7 @@ ftrace_code_disable(struct dyn_ftrace *rec) ret = ftrace_modify_code(ip, call, nop); if (ret) { - switch (ret) { - case -EFAULT: - FTRACE_WARN_ON_ONCE(1); - pr_info("ftrace faulted on modifying "); - print_ip_sym(ip); - break; - case -EINVAL: - FTRACE_WARN_ON_ONCE(1); - pr_info("ftrace failed to modify "); - print_ip_sym(ip); - print_ip_ins(" expected: ", call); - print_ip_ins(" actual: ", (unsigned char *)ip); - print_ip_ins(" replace: ", nop); - printk(KERN_CONT "\n"); - break; - case -EPERM: - FTRACE_WARN_ON_ONCE(1); - pr_info("ftrace faulted on writing "); - print_ip_sym(ip); - break; - default: - FTRACE_WARN_ON_ONCE(1); - pr_info("ftrace faulted on unknown error "); - print_ip_sym(ip); - } - + ftrace_bug(ret, ip, call, nop); rec->flags |= FTRACE_FL_FAILED; return 0; } -- cgit v1.2.3-70-g09d2 From 918c115410c6cc57033835b6a401e57697f9ea4f Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Fri, 14 Nov 2008 16:21:19 -0800 Subject: ftrace: do not process freed records Impact: keep from converting freed records When the tracer is started or stopped, it converts all code pointed to by the saved records into callers to ftrace or nops. When modules are unloaded, their records are freed, but they still exist within the record pages. This patch changes the code to skip over freed records. Signed-off-by: Steven Rostedt Signed-off-by: Ingo Molnar --- kernel/trace/ftrace.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) (limited to 'kernel/trace/ftrace.c') diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index d5bd21f3952..3940c71ac2a 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -488,8 +488,12 @@ static void ftrace_replace_code(int enable) for (i = 0; i < pg->index; i++) { rec = &pg->records[i]; - /* don't modify code that has already faulted */ - if (rec->flags & FTRACE_FL_FAILED) + /* + * Skip over free records and records that have + * failed. + */ + if (rec->flags & FTRACE_FL_FREE || + rec->flags & FTRACE_FL_FAILED) continue; /* ignore updates to this record's mcount site */ -- cgit v1.2.3-70-g09d2 From 31e889098a80ceb3e9e3c555d522b2686a6663c6 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Fri, 14 Nov 2008 16:21:19 -0800 Subject: ftrace: pass module struct to arch dynamic ftrace functions Impact: allow archs more flexibility on dynamic ftrace implementations Dynamic ftrace has largly been developed on x86. Since x86 does not have the same limitations as other architectures, the ftrace interaction between the generic code and the architecture specific code was not flexible enough to handle some of the issues that other architectures have. Most notably, module trampolines. Due to the limited branch distance that archs make in calling kernel core code from modules, the module load code must create a trampoline to jump to what will make the larger jump into core kernel code. The problem arises when this happens to a call to mcount. Ftrace checks all code before modifying it and makes sure the current code is what it expects. Right now, there is not enough information to handle modifying module trampolines. This patch changes the API between generic dynamic ftrace code and the arch dependent code. There is now two functions for modifying code: ftrace_make_nop(mod, rec, addr) - convert the code at rec->ip into a nop, where the original text is calling addr. (mod is the module struct if called by module init) ftrace_make_caller(rec, addr) - convert the code rec->ip that should be a nop into a caller to addr. The record "rec" now has a new field called "arch" where the architecture can add any special attributes to each call site record. Signed-off-by: Steven Rostedt Signed-off-by: Ingo Molnar --- arch/x86/include/asm/ftrace.h | 8 ++++++ arch/x86/kernel/ftrace.c | 29 +++++++++++++++++--- include/linux/ftrace.h | 53 +++++++++++++++++++++++++++--------- kernel/module.c | 2 +- kernel/trace/ftrace.c | 62 +++++++++++++++++-------------------------- 5 files changed, 100 insertions(+), 54 deletions(-) (limited to 'kernel/trace/ftrace.c') diff --git a/arch/x86/include/asm/ftrace.h b/arch/x86/include/asm/ftrace.h index 9b6a1fa19e7..2bb43b433e0 100644 --- a/arch/x86/include/asm/ftrace.h +++ b/arch/x86/include/asm/ftrace.h @@ -17,6 +17,14 @@ static inline unsigned long ftrace_call_adjust(unsigned long addr) */ return addr - 1; } + +#ifdef CONFIG_DYNAMIC_FTRACE + +struct dyn_arch_ftrace { + /* No extra data needed for x86 */ +}; + +#endif /* CONFIG_DYNAMIC_FTRACE */ #endif /* __ASSEMBLY__ */ #endif /* CONFIG_FUNCTION_TRACER */ diff --git a/arch/x86/kernel/ftrace.c b/arch/x86/kernel/ftrace.c index fe832738e1e..762222ad138 100644 --- a/arch/x86/kernel/ftrace.c +++ b/arch/x86/kernel/ftrace.c @@ -166,7 +166,7 @@ static int ftrace_calc_offset(long ip, long addr) return (int)(addr - ip); } -unsigned char *ftrace_call_replace(unsigned long ip, unsigned long addr) +static unsigned char *ftrace_call_replace(unsigned long ip, unsigned long addr) { static union ftrace_code_union calc; @@ -311,12 +311,12 @@ do_ftrace_mod_code(unsigned long ip, void *new_code) static unsigned char ftrace_nop[MCOUNT_INSN_SIZE]; -unsigned char *ftrace_nop_replace(void) +static unsigned char *ftrace_nop_replace(void) { return ftrace_nop; } -int +static int ftrace_modify_code(unsigned long ip, unsigned char *old_code, unsigned char *new_code) { @@ -349,6 +349,29 @@ ftrace_modify_code(unsigned long ip, unsigned char *old_code, return 0; } +int ftrace_make_nop(struct module *mod, + struct dyn_ftrace *rec, unsigned long addr) +{ + unsigned char *new, *old; + unsigned long ip = rec->ip; + + old = ftrace_call_replace(ip, addr); + new = ftrace_nop_replace(); + + return ftrace_modify_code(rec->ip, old, new); +} + +int ftrace_make_call(struct dyn_ftrace *rec, unsigned long addr) +{ + unsigned char *new, *old; + unsigned long ip = rec->ip; + + old = ftrace_nop_replace(); + new = ftrace_call_replace(ip, addr); + + return ftrace_modify_code(rec->ip, old, new); +} + int ftrace_update_ftrace_func(ftrace_func_t func) { unsigned long ip = (unsigned long)(&ftrace_call); diff --git a/include/linux/ftrace.h b/include/linux/ftrace.h index 4fbc4a8b86a..166a2070ef6 100644 --- a/include/linux/ftrace.h +++ b/include/linux/ftrace.h @@ -74,6 +74,9 @@ static inline void ftrace_start(void) { } #endif /* CONFIG_FUNCTION_TRACER */ #ifdef CONFIG_DYNAMIC_FTRACE +/* asm/ftrace.h must be defined for archs supporting dynamic ftrace */ +#include + enum { FTRACE_FL_FREE = (1 << 0), FTRACE_FL_FAILED = (1 << 1), @@ -88,6 +91,7 @@ struct dyn_ftrace { struct list_head list; unsigned long ip; /* address of mcount call-site */ unsigned long flags; + struct dyn_arch_ftrace arch; }; int ftrace_force_update(void); @@ -95,22 +99,40 @@ void ftrace_set_filter(unsigned char *buf, int len, int reset); /* defined in arch */ extern int ftrace_ip_converted(unsigned long ip); -extern unsigned char *ftrace_nop_replace(void); -extern unsigned char *ftrace_call_replace(unsigned long ip, unsigned long addr); extern int ftrace_dyn_arch_init(void *data); extern int ftrace_update_ftrace_func(ftrace_func_t func); extern void ftrace_caller(void); extern void ftrace_call(void); extern void mcount_call(void); -/* May be defined in arch */ -extern int ftrace_arch_read_dyn_info(char *buf, int size); +/** + * ftrace_make_nop - convert code into top + * @mod: module structure if called by module load initialization + * @rec: the mcount call site record + * @addr: the address that the call site should be calling + * + * This is a very sensitive operation and great care needs + * to be taken by the arch. The operation should carefully + * read the location, check to see if what is read is indeed + * what we expect it to be, and then on success of the compare, + * it should write to the location. + * + * The code segment at @rec->ip should be a caller to @addr + * + * Return must be: + * 0 on success + * -EFAULT on error reading the location + * -EINVAL on a failed compare of the contents + * -EPERM on error writing to the location + * Any other value will be considered a failure. + */ +extern int ftrace_make_nop(struct module *mod, + struct dyn_ftrace *rec, unsigned long addr); /** - * ftrace_modify_code - modify code segment - * @ip: the address of the code segment - * @old_code: the contents of what is expected to be there - * @new_code: the code to patch in + * ftrace_make_call - convert a nop call site into a call to addr + * @rec: the mcount call site record + * @addr: the address that the call site should call * * This is a very sensitive operation and great care needs * to be taken by the arch. The operation should carefully @@ -118,6 +140,8 @@ extern int ftrace_arch_read_dyn_info(char *buf, int size); * what we expect it to be, and then on success of the compare, * it should write to the location. * + * The code segment at @rec->ip should be a nop + * * Return must be: * 0 on success * -EFAULT on error reading the location @@ -125,8 +149,11 @@ extern int ftrace_arch_read_dyn_info(char *buf, int size); * -EPERM on error writing to the location * Any other value will be considered a failure. */ -extern int ftrace_modify_code(unsigned long ip, unsigned char *old_code, - unsigned char *new_code); +extern int ftrace_make_call(struct dyn_ftrace *rec, unsigned long addr); + + +/* May be defined in arch */ +extern int ftrace_arch_read_dyn_info(char *buf, int size); extern int skip_trace(unsigned long ip); @@ -259,11 +286,13 @@ static inline void ftrace_dump(void) { } #ifdef CONFIG_FTRACE_MCOUNT_RECORD extern void ftrace_init(void); -extern void ftrace_init_module(unsigned long *start, unsigned long *end); +extern void ftrace_init_module(struct module *mod, + unsigned long *start, unsigned long *end); #else static inline void ftrace_init(void) { } static inline void -ftrace_init_module(unsigned long *start, unsigned long *end) { } +ftrace_init_module(struct module *mod, + unsigned long *start, unsigned long *end) { } #endif diff --git a/kernel/module.c b/kernel/module.c index 1f4cc00e0c2..69791274e89 100644 --- a/kernel/module.c +++ b/kernel/module.c @@ -2201,7 +2201,7 @@ static noinline struct module *load_module(void __user *umod, /* sechdrs[0].sh_size is always zero */ mseg = section_objs(hdr, sechdrs, secstrings, "__mcount_loc", sizeof(*mseg), &num_mcount); - ftrace_init_module(mseg, mseg + num_mcount); + ftrace_init_module(mod, mseg, mseg + num_mcount); err = module_finalize(hdr, sechdrs, mod); if (err < 0) diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index 3940c71ac2a..e9a5fbfce08 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -358,9 +358,7 @@ static void print_ip_ins(const char *fmt, unsigned char *p) printk(KERN_CONT "%s%02x", i ? ":" : "", p[i]); } -static void ftrace_bug(int failed, unsigned long ip, - unsigned char *expected, - unsigned char *replace) +static void ftrace_bug(int failed, unsigned long ip) { switch (failed) { case -EFAULT: @@ -372,9 +370,7 @@ static void ftrace_bug(int failed, unsigned long ip, FTRACE_WARN_ON_ONCE(1); pr_info("ftrace failed to modify "); print_ip_sym(ip); - print_ip_ins(" expected: ", expected); print_ip_ins(" actual: ", (unsigned char *)ip); - print_ip_ins(" replace: ", replace); printk(KERN_CONT "\n"); break; case -EPERM: @@ -392,8 +388,7 @@ static void ftrace_bug(int failed, unsigned long ip, #define FTRACE_ADDR ((long)(ftrace_caller)) static int -__ftrace_replace_code(struct dyn_ftrace *rec, - unsigned char *old, unsigned char *new, int enable) +__ftrace_replace_code(struct dyn_ftrace *rec, int enable) { unsigned long ip, fl; @@ -435,12 +430,10 @@ __ftrace_replace_code(struct dyn_ftrace *rec, * otherwise enable it! */ if (fl & FTRACE_FL_ENABLED) { - /* swap new and old */ - new = old; - old = ftrace_call_replace(ip, FTRACE_ADDR); + enable = 0; rec->flags &= ~FTRACE_FL_ENABLED; } else { - new = ftrace_call_replace(ip, FTRACE_ADDR); + enable = 1; rec->flags |= FTRACE_FL_ENABLED; } } else { @@ -453,10 +446,7 @@ __ftrace_replace_code(struct dyn_ftrace *rec, fl = rec->flags & (FTRACE_FL_NOTRACE | FTRACE_FL_ENABLED); if (fl == FTRACE_FL_NOTRACE) return 0; - - new = ftrace_call_replace(ip, FTRACE_ADDR); - } else - old = ftrace_call_replace(ip, FTRACE_ADDR); + } if (enable) { if (rec->flags & FTRACE_FL_ENABLED) @@ -469,21 +459,18 @@ __ftrace_replace_code(struct dyn_ftrace *rec, } } - return ftrace_modify_code(ip, old, new); + if (enable) + return ftrace_make_call(rec, FTRACE_ADDR); + else + return ftrace_make_nop(NULL, rec, FTRACE_ADDR); } static void ftrace_replace_code(int enable) { int i, failed; - unsigned char *new = NULL, *old = NULL; struct dyn_ftrace *rec; struct ftrace_page *pg; - if (enable) - old = ftrace_nop_replace(); - else - new = ftrace_nop_replace(); - for (pg = ftrace_pages_start; pg; pg = pg->next) { for (i = 0; i < pg->index; i++) { rec = &pg->records[i]; @@ -504,34 +491,30 @@ static void ftrace_replace_code(int enable) unfreeze_record(rec); } - failed = __ftrace_replace_code(rec, old, new, enable); + failed = __ftrace_replace_code(rec, enable); if (failed && (rec->flags & FTRACE_FL_CONVERTED)) { rec->flags |= FTRACE_FL_FAILED; if ((system_state == SYSTEM_BOOTING) || !core_kernel_text(rec->ip)) { ftrace_free_rec(rec); } else - ftrace_bug(failed, rec->ip, old, new); + ftrace_bug(failed, rec->ip); } } } } static int -ftrace_code_disable(struct dyn_ftrace *rec) +ftrace_code_disable(struct module *mod, struct dyn_ftrace *rec) { unsigned long ip; - unsigned char *nop, *call; int ret; ip = rec->ip; - nop = ftrace_nop_replace(); - call = ftrace_call_replace(ip, mcount_addr); - - ret = ftrace_modify_code(ip, call, nop); + ret = ftrace_make_nop(mod, rec, mcount_addr); if (ret) { - ftrace_bug(ret, ip, call, nop); + ftrace_bug(ret, ip); rec->flags |= FTRACE_FL_FAILED; return 0; } @@ -650,7 +633,7 @@ static cycle_t ftrace_update_time; static unsigned long ftrace_update_cnt; unsigned long ftrace_update_tot_cnt; -static int ftrace_update_code(void) +static int ftrace_update_code(struct module *mod) { struct dyn_ftrace *p, *t; cycle_t start, stop; @@ -667,7 +650,7 @@ static int ftrace_update_code(void) list_del_init(&p->list); /* convert record (i.e, patch mcount-call with NOP) */ - if (ftrace_code_disable(p)) { + if (ftrace_code_disable(mod, p)) { p->flags |= FTRACE_FL_CONVERTED; ftrace_update_cnt++; } else @@ -1309,7 +1292,8 @@ static __init int ftrace_init_debugfs(void) fs_initcall(ftrace_init_debugfs); -static int ftrace_convert_nops(unsigned long *start, +static int ftrace_convert_nops(struct module *mod, + unsigned long *start, unsigned long *end) { unsigned long *p; @@ -1325,18 +1309,19 @@ static int ftrace_convert_nops(unsigned long *start, /* disable interrupts to prevent kstop machine */ local_irq_save(flags); - ftrace_update_code(); + ftrace_update_code(mod); local_irq_restore(flags); mutex_unlock(&ftrace_start_lock); return 0; } -void ftrace_init_module(unsigned long *start, unsigned long *end) +void ftrace_init_module(struct module *mod, + unsigned long *start, unsigned long *end) { if (ftrace_disabled || start == end) return; - ftrace_convert_nops(start, end); + ftrace_convert_nops(mod, start, end); } extern unsigned long __start_mcount_loc[]; @@ -1366,7 +1351,8 @@ void __init ftrace_init(void) last_ftrace_enabled = ftrace_enabled = 1; - ret = ftrace_convert_nops(__start_mcount_loc, + ret = ftrace_convert_nops(NULL, + __start_mcount_loc, __stop_mcount_loc); return; -- cgit v1.2.3-70-g09d2 From 20e5227e9f55ae1969934821ccbf581563785bbe Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Fri, 14 Nov 2008 16:21:19 -0800 Subject: ftrace: allow NULL pointers in mcount_loc Impact: make ftrace_convert_nops() more permissive Due to the way different architecture linkers combine the data sections of the mcount_loc (the section that lists all the locations that call mcount), there may be zeros added in that section. This is usually due to strange alignments that the linker performs, that pads in zeros. This patch makes the conversion code to nops skip any pointer in the mcount_loc section that is NULL. Signed-off-by: Steven Rostedt Signed-off-by: Ingo Molnar --- kernel/trace/ftrace.c | 8 ++++++++ 1 file changed, 8 insertions(+) (limited to 'kernel/trace/ftrace.c') diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index e9a5fbfce08..cc4219135dc 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -1304,6 +1304,14 @@ static int ftrace_convert_nops(struct module *mod, p = start; while (p < end) { addr = ftrace_call_adjust(*p++); + /* + * Some architecture linkers will pad between + * the different mcount_loc sections of different + * object files to satisfy alignments. + * Skip any NULL pointers. + */ + if (!addr) + continue; ftrace_record_ip(addr); } -- cgit v1.2.3-70-g09d2 From 982c350b9ec4b3564d67f3627a274ae61bbc7e95 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Sat, 15 Nov 2008 16:31:41 -0500 Subject: ftrace: fix dyn ftrace filter Impact: correct implementation of dyn ftrace filter The old decisions made by the filter algorithm was complex and incorrect. This lead to inconsistent enabling or disabling of functions when the filter was used. This patch simplifies that code and in doing so, corrects the usage of the filters. Signed-off-by: Steven Rostedt Signed-off-by: Ingo Molnar --- kernel/trace/ftrace.c | 83 ++++++++++++++++++++++----------------------------- 1 file changed, 36 insertions(+), 47 deletions(-) (limited to 'kernel/trace/ftrace.c') diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index cc4219135dc..b9f2e22faf2 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -394,72 +394,62 @@ __ftrace_replace_code(struct dyn_ftrace *rec, int enable) ip = rec->ip; - if (ftrace_filtered && enable) { + /* + * If this record is not to be traced and + * it is not enabled then do nothing. + * + * If this record is not to be traced and + * it is enabled then disabled it. + * + */ + if (rec->flags & FTRACE_FL_NOTRACE) { + if (rec->flags & FTRACE_FL_ENABLED) + rec->flags &= ~FTRACE_FL_ENABLED; + else + return 0; + + } else if (ftrace_filtered && enable) { /* - * If filtering is on: - * - * If this record is set to be filtered and - * is enabled then do nothing. - * - * If this record is set to be filtered and - * it is not enabled, enable it. - * - * If this record is not set to be filtered - * and it is not enabled do nothing. - * - * If this record is set not to trace then - * do nothing. - * - * If this record is set not to trace and - * it is enabled then disable it. - * - * If this record is not set to be filtered and - * it is enabled, disable it. + * Filtering is on: */ - fl = rec->flags & (FTRACE_FL_FILTER | FTRACE_FL_NOTRACE | - FTRACE_FL_ENABLED); + fl = rec->flags & (FTRACE_FL_FILTER | FTRACE_FL_ENABLED); - if ((fl == (FTRACE_FL_FILTER | FTRACE_FL_ENABLED)) || - (fl == (FTRACE_FL_FILTER | FTRACE_FL_NOTRACE)) || - !fl || (fl == FTRACE_FL_NOTRACE)) + /* Record is filtered and enabled, do nothing */ + if (fl == (FTRACE_FL_FILTER | FTRACE_FL_ENABLED)) return 0; - /* - * If it is enabled disable it, - * otherwise enable it! - */ - if (fl & FTRACE_FL_ENABLED) { - enable = 0; + /* Record is not filtered and is not enabled do nothing */ + if (!fl) + return 0; + + /* Record is not filtered but enabled, disable it */ + if (fl == FTRACE_FL_ENABLED) rec->flags &= ~FTRACE_FL_ENABLED; - } else { - enable = 1; + else + /* Otherwise record is filtered but not enabled, enable it */ rec->flags |= FTRACE_FL_ENABLED; - } } else { + /* Disable or not filtered */ if (enable) { - /* - * If this record is set not to trace and is - * not enabled, do nothing. - */ - fl = rec->flags & (FTRACE_FL_NOTRACE | FTRACE_FL_ENABLED); - if (fl == FTRACE_FL_NOTRACE) - return 0; - } - - if (enable) { + /* if record is enabled, do nothing */ if (rec->flags & FTRACE_FL_ENABLED) return 0; + rec->flags |= FTRACE_FL_ENABLED; + } else { + + /* if record is not enabled do nothing */ if (!(rec->flags & FTRACE_FL_ENABLED)) return 0; + rec->flags &= ~FTRACE_FL_ENABLED; } } - if (enable) + if (rec->flags & FTRACE_FL_ENABLED) return ftrace_make_call(rec, FTRACE_ADDR); else return ftrace_make_nop(NULL, rec, FTRACE_ADDR); @@ -554,8 +544,7 @@ static void ftrace_startup(void) mutex_lock(&ftrace_start_lock); ftrace_start_up++; - if (ftrace_start_up == 1) - command |= FTRACE_ENABLE_CALLS; + command |= FTRACE_ENABLE_CALLS; if (saved_ftrace_func != ftrace_trace_function) { saved_ftrace_func = ftrace_trace_function; -- cgit v1.2.3-70-g09d2 From ee02a2e5c88ca2e4d6921f08d037b46d5bf82641 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Sat, 15 Nov 2008 16:31:41 -0500 Subject: ftrace: make filtered functions effective on setting Impact: set filtered functions at time the filter is set It can be confusing when the set_filter_functions is set (or cleared) and the functions being recorded by the dynamic tracer does not match. This patch causes the code to be updated if the function tracer is enabled and the filter is changed. Signed-off-by: Steven Rostedt Signed-off-by: Ingo Molnar --- kernel/trace/ftrace.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel/trace/ftrace.c') diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index b9f2e22faf2..b42ec1de546 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -1194,7 +1194,7 @@ ftrace_regex_release(struct inode *inode, struct file *file, int enable) mutex_lock(&ftrace_sysctl_lock); mutex_lock(&ftrace_start_lock); - if (iter->filtered && ftrace_start_up && ftrace_enabled) + if (ftrace_start_up && ftrace_enabled) ftrace_run_update_code(FTRACE_ENABLE_CALLS); mutex_unlock(&ftrace_start_lock); mutex_unlock(&ftrace_sysctl_lock); -- cgit v1.2.3-70-g09d2 From e7d3737ea1b102030f44e96c97754101e41515f0 Mon Sep 17 00:00:00 2001 From: Frederic Weisbecker Date: Sun, 16 Nov 2008 06:02:06 +0100 Subject: tracing/function-return-tracer: support for dynamic ftrace on function return tracer This patch adds the support for dynamic tracing on the function return tracer. The whole difference with normal dynamic function tracing is that we don't need to hook on a particular callback. The only pro that we want is to nop or set dynamically the calls to ftrace_caller (which is ftrace_return_caller here). Some security checks ensure that we are not trying to launch dynamic tracing for return tracing while normal function tracing is already running. An example of trace with getnstimeofday set as a filter: ktime_get_ts+0x22/0x50 -> getnstimeofday (2283 ns) ktime_get_ts+0x22/0x50 -> getnstimeofday (1396 ns) ktime_get_ts+0x22/0x50 -> getnstimeofday (1382 ns) ktime_get_ts+0x22/0x50 -> getnstimeofday (1825 ns) ktime_get_ts+0x22/0x50 -> getnstimeofday (1426 ns) ktime_get_ts+0x22/0x50 -> getnstimeofday (1464 ns) ktime_get_ts+0x22/0x50 -> getnstimeofday (1524 ns) ktime_get_ts+0x22/0x50 -> getnstimeofday (1382 ns) ktime_get_ts+0x22/0x50 -> getnstimeofday (1382 ns) ktime_get_ts+0x22/0x50 -> getnstimeofday (1434 ns) ktime_get_ts+0x22/0x50 -> getnstimeofday (1464 ns) ktime_get_ts+0x22/0x50 -> getnstimeofday (1502 ns) ktime_get_ts+0x22/0x50 -> getnstimeofday (1404 ns) ktime_get_ts+0x22/0x50 -> getnstimeofday (1397 ns) ktime_get_ts+0x22/0x50 -> getnstimeofday (1051 ns) ktime_get_ts+0x22/0x50 -> getnstimeofday (1314 ns) ktime_get_ts+0x22/0x50 -> getnstimeofday (1344 ns) ktime_get_ts+0x22/0x50 -> getnstimeofday (1163 ns) ktime_get_ts+0x22/0x50 -> getnstimeofday (1390 ns) ktime_get_ts+0x22/0x50 -> getnstimeofday (1374 ns) Signed-off-by: Frederic Weisbecker Signed-off-by: Ingo Molnar --- arch/x86/kernel/entry_32.S | 18 ++- arch/x86/kernel/ftrace.c | 258 +++++++++++++++++----------------- include/linux/ftrace.h | 16 ++- kernel/trace/Kconfig | 1 - kernel/trace/ftrace.c | 58 +++++++- kernel/trace/trace_functions_return.c | 15 +- 6 files changed, 211 insertions(+), 155 deletions(-) (limited to 'kernel/trace/ftrace.c') diff --git a/arch/x86/kernel/entry_32.S b/arch/x86/kernel/entry_32.S index f9762114983..74defe21ba4 100644 --- a/arch/x86/kernel/entry_32.S +++ b/arch/x86/kernel/entry_32.S @@ -1190,7 +1190,7 @@ ENTRY(mcount) jnz trace #ifdef CONFIG_FUNCTION_RET_TRACER cmpl $ftrace_stub, ftrace_function_return - jnz trace_return + jnz ftrace_return_caller #endif .globl ftrace_stub ftrace_stub: @@ -1211,9 +1211,15 @@ trace: popl %ecx popl %eax jmp ftrace_stub +END(mcount) +#endif /* CONFIG_DYNAMIC_FTRACE */ +#endif /* CONFIG_FUNCTION_TRACER */ #ifdef CONFIG_FUNCTION_RET_TRACER -trace_return: +ENTRY(ftrace_return_caller) + cmpl $0, function_trace_stop + jne ftrace_stub + pushl %eax pushl %ecx pushl %edx @@ -1223,7 +1229,8 @@ trace_return: popl %edx popl %ecx popl %eax - jmp ftrace_stub + ret +END(ftrace_return_caller) .globl return_to_handler return_to_handler: @@ -1237,10 +1244,7 @@ return_to_handler: popl %ecx popl %eax ret -#endif /* CONFIG_FUNCTION_RET_TRACER */ -END(mcount) -#endif /* CONFIG_DYNAMIC_FTRACE */ -#endif /* CONFIG_FUNCTION_TRACER */ +#endif .section .rodata,"a" #include "syscall_table_32.S" diff --git a/arch/x86/kernel/ftrace.c b/arch/x86/kernel/ftrace.c index d98b5a8ecf4..924153edd97 100644 --- a/arch/x86/kernel/ftrace.c +++ b/arch/x86/kernel/ftrace.c @@ -24,134 +24,6 @@ #include - -#ifdef CONFIG_FUNCTION_RET_TRACER - -/* - * These functions are picked from those used on - * this page for dynamic ftrace. They have been - * simplified to ignore all traces in NMI context. - */ -static atomic_t in_nmi; - -void ftrace_nmi_enter(void) -{ - atomic_inc(&in_nmi); -} - -void ftrace_nmi_exit(void) -{ - atomic_dec(&in_nmi); -} - -/* Add a function return address to the trace stack on thread info.*/ -static int push_return_trace(unsigned long ret, unsigned long long time, - unsigned long func) -{ - int index; - struct thread_info *ti = current_thread_info(); - - /* The return trace stack is full */ - if (ti->curr_ret_stack == FTRACE_RET_STACK_SIZE - 1) - return -EBUSY; - - index = ++ti->curr_ret_stack; - barrier(); - ti->ret_stack[index].ret = ret; - ti->ret_stack[index].func = func; - ti->ret_stack[index].calltime = time; - - return 0; -} - -/* Retrieve a function return address to the trace stack on thread info.*/ -static void pop_return_trace(unsigned long *ret, unsigned long long *time, - unsigned long *func) -{ - int index; - - struct thread_info *ti = current_thread_info(); - index = ti->curr_ret_stack; - *ret = ti->ret_stack[index].ret; - *func = ti->ret_stack[index].func; - *time = ti->ret_stack[index].calltime; - ti->curr_ret_stack--; -} - -/* - * Send the trace to the ring-buffer. - * @return the original return address. - */ -unsigned long ftrace_return_to_handler(void) -{ - struct ftrace_retfunc trace; - pop_return_trace(&trace.ret, &trace.calltime, &trace.func); - trace.rettime = cpu_clock(raw_smp_processor_id()); - ftrace_function_return(&trace); - - return trace.ret; -} - -/* - * Hook the return address and push it in the stack of return addrs - * in current thread info. - */ -void prepare_ftrace_return(unsigned long *parent, unsigned long self_addr) -{ - unsigned long old; - unsigned long long calltime; - int faulted; - unsigned long return_hooker = (unsigned long) - &return_to_handler; - - /* Nmi's are currently unsupported */ - if (atomic_read(&in_nmi)) - return; - - /* - * Protect against fault, even if it shouldn't - * happen. This tool is too much intrusive to - * ignore such a protection. - */ - asm volatile( - "1: movl (%[parent_old]), %[old]\n" - "2: movl %[return_hooker], (%[parent_replaced])\n" - " movl $0, %[faulted]\n" - - ".section .fixup, \"ax\"\n" - "3: movl $1, %[faulted]\n" - ".previous\n" - - ".section __ex_table, \"a\"\n" - " .long 1b, 3b\n" - " .long 2b, 3b\n" - ".previous\n" - - : [parent_replaced] "=r" (parent), [old] "=r" (old), - [faulted] "=r" (faulted) - : [parent_old] "0" (parent), [return_hooker] "r" (return_hooker) - : "memory" - ); - - if (WARN_ON(faulted)) { - unregister_ftrace_return(); - return; - } - - if (WARN_ON(!__kernel_text_address(old))) { - unregister_ftrace_return(); - *parent = old; - return; - } - - calltime = cpu_clock(raw_smp_processor_id()); - - if (push_return_trace(old, calltime, self_addr) == -EBUSY) - *parent = old; -} - -#endif - #ifdef CONFIG_DYNAMIC_FTRACE union ftrace_code_union { @@ -450,3 +322,133 @@ int __init ftrace_dyn_arch_init(void *data) return 0; } #endif + +#ifdef CONFIG_FUNCTION_RET_TRACER + +#ifndef CONFIG_DYNAMIC_FTRACE + +/* + * These functions are picked from those used on + * this page for dynamic ftrace. They have been + * simplified to ignore all traces in NMI context. + */ +static atomic_t in_nmi; + +void ftrace_nmi_enter(void) +{ + atomic_inc(&in_nmi); +} + +void ftrace_nmi_exit(void) +{ + atomic_dec(&in_nmi); +} +#endif /* !CONFIG_DYNAMIC_FTRACE */ + +/* Add a function return address to the trace stack on thread info.*/ +static int push_return_trace(unsigned long ret, unsigned long long time, + unsigned long func) +{ + int index; + struct thread_info *ti = current_thread_info(); + + /* The return trace stack is full */ + if (ti->curr_ret_stack == FTRACE_RET_STACK_SIZE - 1) + return -EBUSY; + + index = ++ti->curr_ret_stack; + barrier(); + ti->ret_stack[index].ret = ret; + ti->ret_stack[index].func = func; + ti->ret_stack[index].calltime = time; + + return 0; +} + +/* Retrieve a function return address to the trace stack on thread info.*/ +static void pop_return_trace(unsigned long *ret, unsigned long long *time, + unsigned long *func) +{ + int index; + + struct thread_info *ti = current_thread_info(); + index = ti->curr_ret_stack; + *ret = ti->ret_stack[index].ret; + *func = ti->ret_stack[index].func; + *time = ti->ret_stack[index].calltime; + ti->curr_ret_stack--; +} + +/* + * Send the trace to the ring-buffer. + * @return the original return address. + */ +unsigned long ftrace_return_to_handler(void) +{ + struct ftrace_retfunc trace; + pop_return_trace(&trace.ret, &trace.calltime, &trace.func); + trace.rettime = cpu_clock(raw_smp_processor_id()); + ftrace_function_return(&trace); + + return trace.ret; +} + +/* + * Hook the return address and push it in the stack of return addrs + * in current thread info. + */ +void prepare_ftrace_return(unsigned long *parent, unsigned long self_addr) +{ + unsigned long old; + unsigned long long calltime; + int faulted; + unsigned long return_hooker = (unsigned long) + &return_to_handler; + + /* Nmi's are currently unsupported */ + if (atomic_read(&in_nmi)) + return; + + /* + * Protect against fault, even if it shouldn't + * happen. This tool is too much intrusive to + * ignore such a protection. + */ + asm volatile( + "1: movl (%[parent_old]), %[old]\n" + "2: movl %[return_hooker], (%[parent_replaced])\n" + " movl $0, %[faulted]\n" + + ".section .fixup, \"ax\"\n" + "3: movl $1, %[faulted]\n" + ".previous\n" + + ".section __ex_table, \"a\"\n" + " .long 1b, 3b\n" + " .long 2b, 3b\n" + ".previous\n" + + : [parent_replaced] "=r" (parent), [old] "=r" (old), + [faulted] "=r" (faulted) + : [parent_old] "0" (parent), [return_hooker] "r" (return_hooker) + : "memory" + ); + + if (WARN_ON(faulted)) { + unregister_ftrace_return(); + return; + } + + if (WARN_ON(!__kernel_text_address(old))) { + unregister_ftrace_return(); + *parent = old; + return; + } + + calltime = cpu_clock(raw_smp_processor_id()); + + if (push_return_trace(old, calltime, self_addr) == -EBUSY) + *parent = old; +} + +#endif /* CONFIG_FUNCTION_RET_TRACER */ diff --git a/include/linux/ftrace.h b/include/linux/ftrace.h index 166a2070ef6..f1af1aab00e 100644 --- a/include/linux/ftrace.h +++ b/include/linux/ftrace.h @@ -25,6 +25,17 @@ struct ftrace_ops { extern int function_trace_stop; +/* + * Type of the current tracing. + */ +enum ftrace_tracing_type_t { + FTRACE_TYPE_ENTER = 0, /* Hook the call of the function */ + FTRACE_TYPE_RETURN, /* Hook the return of the function */ +}; + +/* Current tracing type, default is FTRACE_TYPE_ENTER */ +extern enum ftrace_tracing_type_t ftrace_tracing_type; + /** * ftrace_stop - stop function tracer. * @@ -104,6 +115,9 @@ extern int ftrace_update_ftrace_func(ftrace_func_t func); extern void ftrace_caller(void); extern void ftrace_call(void); extern void mcount_call(void); +#ifdef CONFIG_FUNCTION_RET_TRACER +extern void ftrace_return_caller(void); +#endif /** * ftrace_make_nop - convert code into top @@ -310,7 +324,7 @@ struct ftrace_retfunc { /* Type of a callback handler of tracing return function */ typedef void (*trace_function_return_t)(struct ftrace_retfunc *); -extern void register_ftrace_return(trace_function_return_t func); +extern int register_ftrace_return(trace_function_return_t func); /* The current handler in use */ extern trace_function_return_t ftrace_function_return; extern void unregister_ftrace_return(void); diff --git a/kernel/trace/Kconfig b/kernel/trace/Kconfig index 9c89526b6b7..b8378fad29a 100644 --- a/kernel/trace/Kconfig +++ b/kernel/trace/Kconfig @@ -59,7 +59,6 @@ config FUNCTION_TRACER config FUNCTION_RET_TRACER bool "Kernel Function return Tracer" - depends on !DYNAMIC_FTRACE depends on HAVE_FUNCTION_RET_TRACER depends on FUNCTION_TRACER help diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index b42ec1de546..2f78a45aac1 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -50,6 +50,9 @@ static int last_ftrace_enabled; /* Quick disabling of function tracer. */ int function_trace_stop; +/* By default, current tracing type is normal tracing. */ +enum ftrace_tracing_type_t ftrace_tracing_type = FTRACE_TYPE_ENTER; + /* * ftrace_disabled is set when an anomaly is discovered. * ftrace_disabled is much stronger than ftrace_enabled. @@ -385,12 +388,21 @@ static void ftrace_bug(int failed, unsigned long ip) } } -#define FTRACE_ADDR ((long)(ftrace_caller)) static int __ftrace_replace_code(struct dyn_ftrace *rec, int enable) { unsigned long ip, fl; + unsigned long ftrace_addr; + +#ifdef CONFIG_FUNCTION_RET_TRACER + if (ftrace_tracing_type == FTRACE_TYPE_ENTER) + ftrace_addr = (unsigned long)ftrace_caller; + else + ftrace_addr = (unsigned long)ftrace_return_caller; +#else + ftrace_addr = (unsigned long)ftrace_caller; +#endif ip = rec->ip; @@ -450,9 +462,9 @@ __ftrace_replace_code(struct dyn_ftrace *rec, int enable) } if (rec->flags & FTRACE_FL_ENABLED) - return ftrace_make_call(rec, FTRACE_ADDR); + return ftrace_make_call(rec, ftrace_addr); else - return ftrace_make_nop(NULL, rec, FTRACE_ADDR); + return ftrace_make_nop(NULL, rec, ftrace_addr); } static void ftrace_replace_code(int enable) @@ -1405,10 +1417,17 @@ int register_ftrace_function(struct ftrace_ops *ops) return -1; mutex_lock(&ftrace_sysctl_lock); + + if (ftrace_tracing_type == FTRACE_TYPE_RETURN) { + ret = -EBUSY; + goto out; + } + ret = __register_ftrace_function(ops); ftrace_startup(); - mutex_unlock(&ftrace_sysctl_lock); +out: + mutex_unlock(&ftrace_sysctl_lock); return ret; } @@ -1474,16 +1493,45 @@ ftrace_enable_sysctl(struct ctl_table *table, int write, } #ifdef CONFIG_FUNCTION_RET_TRACER + +/* The callback that hooks the return of a function */ trace_function_return_t ftrace_function_return = (trace_function_return_t)ftrace_stub; -void register_ftrace_return(trace_function_return_t func) + +int register_ftrace_return(trace_function_return_t func) { + int ret = 0; + + mutex_lock(&ftrace_sysctl_lock); + + /* + * Don't launch return tracing if normal function + * tracing is already running. + */ + if (ftrace_trace_function != ftrace_stub) { + ret = -EBUSY; + goto out; + } + + ftrace_tracing_type = FTRACE_TYPE_RETURN; ftrace_function_return = func; + ftrace_startup(); + +out: + mutex_unlock(&ftrace_sysctl_lock); + return ret; } void unregister_ftrace_return(void) { + mutex_lock(&ftrace_sysctl_lock); + ftrace_function_return = (trace_function_return_t)ftrace_stub; + ftrace_shutdown(); + /* Restore normal tracing type */ + ftrace_tracing_type = FTRACE_TYPE_ENTER; + + mutex_unlock(&ftrace_sysctl_lock); } #endif diff --git a/kernel/trace/trace_functions_return.c b/kernel/trace/trace_functions_return.c index 61185f756a1..a68564af022 100644 --- a/kernel/trace/trace_functions_return.c +++ b/kernel/trace/trace_functions_return.c @@ -14,29 +14,18 @@ #include "trace.h" -static void start_return_trace(struct trace_array *tr) -{ - register_ftrace_return(&trace_function_return); -} - -static void stop_return_trace(struct trace_array *tr) -{ - unregister_ftrace_return(); -} - static int return_trace_init(struct trace_array *tr) { int cpu; for_each_online_cpu(cpu) tracing_reset(tr, cpu); - start_return_trace(tr); - return 0; + return register_ftrace_return(&trace_function_return); } static void return_trace_reset(struct trace_array *tr) { - stop_return_trace(tr); + unregister_ftrace_return(); } -- cgit v1.2.3-70-g09d2 From f201ae2356c74bcae130b2177b3dca903ea98071 Mon Sep 17 00:00:00 2001 From: Frederic Weisbecker Date: Sun, 23 Nov 2008 06:22:56 +0100 Subject: tracing/function-return-tracer: store return stack into task_struct and allocate it dynamically Impact: use deeper function tracing depth safely Some tests showed that function return tracing needed a more deeper depth of function calls. But it could be unsafe to store these return addresses to the stack. So these arrays will now be allocated dynamically into task_struct of current only when the tracer is activated. Typical scheme when tracer is activated: - allocate a return stack for each task in global list. - fork: allocate the return stack for the newly created task - exit: free return stack of current - idle init: same as fork I chose a default depth of 50. I don't have overruns anymore. Signed-off-by: Frederic Weisbecker Signed-off-by: Ingo Molnar --- arch/x86/include/asm/ftrace.h | 1 - arch/x86/include/asm/thread_info.h | 29 ------------ arch/x86/kernel/ftrace.c | 29 ++++++------ include/linux/ftrace.h | 5 ++ include/linux/sched.h | 23 +++++---- kernel/exit.c | 5 +- kernel/fork.c | 4 ++ kernel/sched.c | 3 ++ kernel/trace/ftrace.c | 96 +++++++++++++++++++++++++++++++++++++- 9 files changed, 137 insertions(+), 58 deletions(-) (limited to 'kernel/trace/ftrace.c') diff --git a/arch/x86/include/asm/ftrace.h b/arch/x86/include/asm/ftrace.h index 2bb43b433e0..754a3e082f9 100644 --- a/arch/x86/include/asm/ftrace.h +++ b/arch/x86/include/asm/ftrace.h @@ -29,7 +29,6 @@ struct dyn_arch_ftrace { #endif /* CONFIG_FUNCTION_TRACER */ #ifdef CONFIG_FUNCTION_RET_TRACER -#define FTRACE_RET_STACK_SIZE 20 #ifndef __ASSEMBLY__ diff --git a/arch/x86/include/asm/thread_info.h b/arch/x86/include/asm/thread_info.h index e90e81ef6ab..0921b4018c1 100644 --- a/arch/x86/include/asm/thread_info.h +++ b/arch/x86/include/asm/thread_info.h @@ -40,36 +40,8 @@ struct thread_info { */ __u8 supervisor_stack[0]; #endif - -#ifdef CONFIG_FUNCTION_RET_TRACER - /* Index of current stored adress in ret_stack */ - int curr_ret_stack; - /* Stack of return addresses for return function tracing */ - struct ftrace_ret_stack ret_stack[FTRACE_RET_STACK_SIZE]; - /* - * Number of functions that haven't been traced - * because of depth overrun. - */ - atomic_t trace_overrun; -#endif }; -#ifdef CONFIG_FUNCTION_RET_TRACER -#define INIT_THREAD_INFO(tsk) \ -{ \ - .task = &tsk, \ - .exec_domain = &default_exec_domain, \ - .flags = 0, \ - .cpu = 0, \ - .preempt_count = 1, \ - .addr_limit = KERNEL_DS, \ - .restart_block = { \ - .fn = do_no_restart_syscall, \ - }, \ - .curr_ret_stack = -1,\ - .trace_overrun = ATOMIC_INIT(0) \ -} -#else #define INIT_THREAD_INFO(tsk) \ { \ .task = &tsk, \ @@ -82,7 +54,6 @@ struct thread_info { .fn = do_no_restart_syscall, \ }, \ } -#endif #define init_thread_info (init_thread_union.thread_info) #define init_stack (init_thread_union.stack) diff --git a/arch/x86/kernel/ftrace.c b/arch/x86/kernel/ftrace.c index 356bb1eb6e9..bb137f7297e 100644 --- a/arch/x86/kernel/ftrace.c +++ b/arch/x86/kernel/ftrace.c @@ -350,19 +350,21 @@ static int push_return_trace(unsigned long ret, unsigned long long time, unsigned long func) { int index; - struct thread_info *ti = current_thread_info(); + + if (!current->ret_stack) + return -EBUSY; /* The return trace stack is full */ - if (ti->curr_ret_stack == FTRACE_RET_STACK_SIZE - 1) { - atomic_inc(&ti->trace_overrun); + if (current->curr_ret_stack == FTRACE_RETFUNC_DEPTH - 1) { + atomic_inc(¤t->trace_overrun); return -EBUSY; } - index = ++ti->curr_ret_stack; + index = ++current->curr_ret_stack; barrier(); - ti->ret_stack[index].ret = ret; - ti->ret_stack[index].func = func; - ti->ret_stack[index].calltime = time; + current->ret_stack[index].ret = ret; + current->ret_stack[index].func = func; + current->ret_stack[index].calltime = time; return 0; } @@ -373,13 +375,12 @@ static void pop_return_trace(unsigned long *ret, unsigned long long *time, { int index; - struct thread_info *ti = current_thread_info(); - index = ti->curr_ret_stack; - *ret = ti->ret_stack[index].ret; - *func = ti->ret_stack[index].func; - *time = ti->ret_stack[index].calltime; - *overrun = atomic_read(&ti->trace_overrun); - ti->curr_ret_stack--; + index = current->curr_ret_stack; + *ret = current->ret_stack[index].ret; + *func = current->ret_stack[index].func; + *time = current->ret_stack[index].calltime; + *overrun = atomic_read(¤t->trace_overrun); + current->curr_ret_stack--; } /* diff --git a/include/linux/ftrace.h b/include/linux/ftrace.h index f7ba4ea5e12..2ba259b2def 100644 --- a/include/linux/ftrace.h +++ b/include/linux/ftrace.h @@ -323,6 +323,8 @@ struct ftrace_retfunc { }; #ifdef CONFIG_FUNCTION_RET_TRACER +#define FTRACE_RETFUNC_DEPTH 50 +#define FTRACE_RETSTACK_ALLOC_SIZE 32 /* Type of a callback handler of tracing return function */ typedef void (*trace_function_return_t)(struct ftrace_retfunc *); @@ -330,6 +332,9 @@ extern int register_ftrace_return(trace_function_return_t func); /* The current handler in use */ extern trace_function_return_t ftrace_function_return; extern void unregister_ftrace_return(void); + +extern void ftrace_retfunc_init_task(struct task_struct *t); +extern void ftrace_retfunc_exit_task(struct task_struct *t); #endif #endif /* _LINUX_FTRACE_H */ diff --git a/include/linux/sched.h b/include/linux/sched.h index c8e0db46420..bee1e93c95a 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -1352,6 +1352,17 @@ struct task_struct { unsigned long default_timer_slack_ns; struct list_head *scm_work_list; +#ifdef CONFIG_FUNCTION_RET_TRACER + /* Index of current stored adress in ret_stack */ + int curr_ret_stack; + /* Stack of return addresses for return function tracing */ + struct ftrace_ret_stack *ret_stack; + /* + * Number of functions that haven't been traced + * because of depth overrun. + */ + atomic_t trace_overrun; +#endif }; /* @@ -2006,18 +2017,6 @@ static inline void setup_thread_stack(struct task_struct *p, struct task_struct { *task_thread_info(p) = *task_thread_info(org); task_thread_info(p)->task = p; - -#ifdef CONFIG_FUNCTION_RET_TRACER - /* - * When fork() creates a child process, this function is called. - * But the child task may not inherit the return adresses traced - * by the return function tracer because it will directly execute - * in userspace and will not return to kernel functions its parent - * used. - */ - task_thread_info(p)->curr_ret_stack = -1; - atomic_set(&task_thread_info(p)->trace_overrun, 0); -#endif } static inline unsigned long *end_of_stack(struct task_struct *p) diff --git a/kernel/exit.c b/kernel/exit.c index 35c8ec2ba03..b9d446329da 100644 --- a/kernel/exit.c +++ b/kernel/exit.c @@ -47,6 +47,7 @@ #include #include #include +#include #include #include @@ -1127,7 +1128,9 @@ NORET_TYPE void do_exit(long code) preempt_disable(); /* causes final put_task_struct in finish_task_switch(). */ tsk->state = TASK_DEAD; - +#ifdef CONFIG_FUNCTION_RET_TRACER + ftrace_retfunc_exit_task(tsk); +#endif schedule(); BUG(); /* Avoid "noreturn function does return". */ diff --git a/kernel/fork.c b/kernel/fork.c index ac62f43ee43..d1eb30e69cc 100644 --- a/kernel/fork.c +++ b/kernel/fork.c @@ -47,6 +47,7 @@ #include #include #include +#include #include #include #include @@ -1269,6 +1270,9 @@ static struct task_struct *copy_process(unsigned long clone_flags, total_forks++; spin_unlock(¤t->sighand->siglock); write_unlock_irq(&tasklist_lock); +#ifdef CONFIG_FUNCTION_RET_TRACER + ftrace_retfunc_init_task(p); +#endif proc_fork_connector(p); cgroup_post_fork(p); return p; diff --git a/kernel/sched.c b/kernel/sched.c index 4de56108c86..fb17205950d 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -5901,6 +5901,9 @@ void __cpuinit init_idle(struct task_struct *idle, int cpu) * The idle tasks have their own, simple scheduling class: */ idle->sched_class = &idle_sched_class; +#ifdef CONFIG_FUNCTION_RET_TRACER + ftrace_retfunc_init_task(idle); +#endif } /* diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index f212da48668..90d99fb02ae 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -1498,10 +1498,77 @@ ftrace_enable_sysctl(struct ctl_table *table, int write, #ifdef CONFIG_FUNCTION_RET_TRACER +static atomic_t ftrace_retfunc_active; + /* The callback that hooks the return of a function */ trace_function_return_t ftrace_function_return = (trace_function_return_t)ftrace_stub; + +/* Try to assign a return stack array on FTRACE_RETSTACK_ALLOC_SIZE tasks. */ +static int alloc_retstack_tasklist(struct ftrace_ret_stack **ret_stack_list) +{ + int i; + int ret = 0; + unsigned long flags; + int start = 0, end = FTRACE_RETSTACK_ALLOC_SIZE; + struct task_struct *g, *t; + + for (i = 0; i < FTRACE_RETSTACK_ALLOC_SIZE; i++) { + ret_stack_list[i] = kmalloc(FTRACE_RETFUNC_DEPTH + * sizeof(struct ftrace_ret_stack), + GFP_KERNEL); + if (!ret_stack_list[i]) { + start = 0; + end = i; + ret = -ENOMEM; + goto free; + } + } + + read_lock_irqsave(&tasklist_lock, flags); + do_each_thread(g, t) { + if (start == end) { + ret = -EAGAIN; + goto unlock; + } + + if (t->ret_stack == NULL) { + t->ret_stack = ret_stack_list[start++]; + t->curr_ret_stack = -1; + atomic_set(&t->trace_overrun, 0); + } + } while_each_thread(g, t); + +unlock: + read_unlock_irqrestore(&tasklist_lock, flags); +free: + for (i = start; i < end; i++) + kfree(ret_stack_list[i]); + return ret; +} + +/* Allocate a return stack for each task */ +static int start_return_tracing(void) +{ + struct ftrace_ret_stack **ret_stack_list; + int ret; + + ret_stack_list = kmalloc(FTRACE_RETSTACK_ALLOC_SIZE * + sizeof(struct ftrace_ret_stack *), + GFP_KERNEL); + + if (!ret_stack_list) + return -ENOMEM; + + do { + ret = alloc_retstack_tasklist(ret_stack_list); + } while (ret == -EAGAIN); + + kfree(ret_stack_list); + return ret; +} + int register_ftrace_return(trace_function_return_t func) { int ret = 0; @@ -1516,7 +1583,12 @@ int register_ftrace_return(trace_function_return_t func) ret = -EBUSY; goto out; } - + atomic_inc(&ftrace_retfunc_active); + ret = start_return_tracing(); + if (ret) { + atomic_dec(&ftrace_retfunc_active); + goto out; + } ftrace_tracing_type = FTRACE_TYPE_RETURN; ftrace_function_return = func; ftrace_startup(); @@ -1530,6 +1602,7 @@ void unregister_ftrace_return(void) { mutex_lock(&ftrace_sysctl_lock); + atomic_dec(&ftrace_retfunc_active); ftrace_function_return = (trace_function_return_t)ftrace_stub; ftrace_shutdown(); /* Restore normal tracing type */ @@ -1537,6 +1610,27 @@ void unregister_ftrace_return(void) mutex_unlock(&ftrace_sysctl_lock); } + +/* Allocate a return stack for newly created task */ +void ftrace_retfunc_init_task(struct task_struct *t) +{ + if (atomic_read(&ftrace_retfunc_active)) { + t->ret_stack = kmalloc(FTRACE_RETFUNC_DEPTH + * sizeof(struct ftrace_ret_stack), + GFP_KERNEL); + if (!t->ret_stack) + return; + t->curr_ret_stack = -1; + atomic_set(&t->trace_overrun, 0); + } else + t->ret_stack = NULL; +} + +void ftrace_retfunc_exit_task(struct task_struct *t) +{ + kfree(t->ret_stack); + t->ret_stack = NULL; +} #endif -- cgit v1.2.3-70-g09d2 From eae849ca034c7f1015f0a6f17421ebc737f0a069 Mon Sep 17 00:00:00 2001 From: Frederic Weisbecker Date: Sun, 23 Nov 2008 17:33:12 +0100 Subject: tracing/function-return-tracer: don't trace kfree while it frees the return stack Impact: fix a crash While I killed the cat process, I got sometimes the following (but rare) crash: [ 65.689027] Pid: 2969, comm: cat Not tainted (2.6.28-rc6-tip #83) AMILO Li 2727 [ 65.689027] EIP: 0060:[<00000000>] EFLAGS: 00010082 CPU: 1 [ 65.689027] EIP is at 0x0 [ 65.689027] EAX: 00000000 EBX: f66cd780 ECX: c019a64a EDX: f66cd780 [ 65.689027] ESI: 00000286 EDI: f66cd780 EBP: f630be2c ESP: f630be24 [ 65.689027] DS: 007b ES: 007b FS: 00d8 GS: 0000 SS: 0068 [ 65.689027] Process cat (pid: 2969, ti=f630a000 task=f66cd780 task.ti=f630a000) [ 65.689027] Stack: [ 65.689027] 00000012 f630bd54 f630be7c c012c853 00000000 c0133cc9 f66cda54 f630be5c [ 65.689027] f630be68 f66cda54 f66cd88c f66cd878 f7070000 00000001 f630be90 c0135dbc [ 65.689027] f614a614 f630be68 f630be68 f65ba200 00000002 f630bf10 f630be90 c012cad6 [ 65.689027] Call Trace: [ 65.689027] [] ? do_exit+0x603/0x850 [ 65.689027] [] ? next_signal+0x9/0x40 [ 65.689027] [] ? dequeue_signal+0x8c/0x180 [ 65.689027] [] ? do_group_exit+0x36/0x90 [ 65.689027] [] ? get_signal_to_deliver+0x20c/0x390 [ 65.689027] [] ? do_notify_resume+0x99/0x8b0 [ 65.689027] [] ? tty_ldisc_deref+0x5a/0x80 [ 65.689027] [] ? trace_hardirqs_on+0xb/0x10 [ 65.689027] [] ? tty_ldisc_deref+0x5a/0x80 [ 65.689027] [] ? n_tty_write+0x0/0x340 [ 65.689027] [] ? redirected_tty_write+0x82/0x90 [ 65.689027] [] ? vfs_write+0x99/0xd0 [ 65.689027] [] ? redirected_tty_write+0x0/0x90 [ 65.689027] [] ? sys_write+0x42/0x70 [ 65.689027] [] ? work_notifysig+0x13/0x19 [ 65.689027] Code: Bad EIP value. [ 65.689027] EIP: [<00000000>] 0x0 SS:ESP 0068:f630be24 This is because on do_exit(), kfree is called to free the return addresses stack but kfree is traced and stored its return address in this stack. This patch fixes it. Signed-off-by: Frederic Weisbecker Signed-off-by: Ingo Molnar --- kernel/trace/ftrace.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) (limited to 'kernel/trace/ftrace.c') diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index 90d99fb02ae..53042f118f2 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -1628,8 +1628,13 @@ void ftrace_retfunc_init_task(struct task_struct *t) void ftrace_retfunc_exit_task(struct task_struct *t) { - kfree(t->ret_stack); + struct ftrace_ret_stack *ret_stack = t->ret_stack; + t->ret_stack = NULL; + /* NULL must become visible to IRQs before we free it: */ + barrier(); + + kfree(ret_stack); } #endif -- cgit v1.2.3-70-g09d2 From fb52607afcd0629776f1dc9e657647ceae81dd50 Mon Sep 17 00:00:00 2001 From: Frederic Weisbecker Date: Tue, 25 Nov 2008 21:07:04 +0100 Subject: tracing/function-return-tracer: change the name into function-graph-tracer Impact: cleanup This patch changes the name of the "return function tracer" into function-graph-tracer which is a more suitable name for a tracing which makes one able to retrieve the ordered call stack during the code flow. Signed-off-by: Frederic Weisbecker Acked-by: Steven Rostedt Signed-off-by: Ingo Molnar --- arch/x86/Kconfig | 2 +- arch/x86/include/asm/ftrace.h | 4 +- arch/x86/kernel/Makefile | 4 +- arch/x86/kernel/entry_32.S | 12 ++--- arch/x86/kernel/ftrace.c | 12 ++--- include/linux/ftrace.h | 24 ++++----- include/linux/ftrace_irq.h | 2 +- include/linux/sched.h | 2 +- kernel/Makefile | 2 +- kernel/fork.c | 4 +- kernel/sched.c | 2 +- kernel/trace/Kconfig | 19 ++++--- kernel/trace/Makefile | 2 +- kernel/trace/ftrace.c | 26 +++++----- kernel/trace/trace.c | 18 +++---- kernel/trace/trace.h | 12 ++--- kernel/trace/trace_functions_graph.c | 98 ++++++++++++++++++++++++++++++++++++ 17 files changed, 173 insertions(+), 72 deletions(-) create mode 100644 kernel/trace/trace_functions_graph.c (limited to 'kernel/trace/ftrace.c') diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig index e49a4fd718f..0842b112768 100644 --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -29,7 +29,7 @@ config X86 select HAVE_FTRACE_MCOUNT_RECORD select HAVE_DYNAMIC_FTRACE select HAVE_FUNCTION_TRACER - select HAVE_FUNCTION_RET_TRACER if X86_32 + select HAVE_FUNCTION_GRAPH_TRACER if X86_32 select HAVE_FUNCTION_TRACE_MCOUNT_TEST select HAVE_KVM if ((X86_32 && !X86_VOYAGER && !X86_VISWS && !X86_NUMAQ) || X86_64) select HAVE_ARCH_KGDB if !X86_VOYAGER diff --git a/arch/x86/include/asm/ftrace.h b/arch/x86/include/asm/ftrace.h index 754a3e082f9..7e61b4ceb9a 100644 --- a/arch/x86/include/asm/ftrace.h +++ b/arch/x86/include/asm/ftrace.h @@ -28,7 +28,7 @@ struct dyn_arch_ftrace { #endif /* __ASSEMBLY__ */ #endif /* CONFIG_FUNCTION_TRACER */ -#ifdef CONFIG_FUNCTION_RET_TRACER +#ifdef CONFIG_FUNCTION_GRAPH_TRACER #ifndef __ASSEMBLY__ @@ -51,6 +51,6 @@ struct ftrace_ret_stack { extern void return_to_handler(void); #endif /* __ASSEMBLY__ */ -#endif /* CONFIG_FUNCTION_RET_TRACER */ +#endif /* CONFIG_FUNCTION_GRAPH_TRACER */ #endif /* _ASM_X86_FTRACE_H */ diff --git a/arch/x86/kernel/Makefile b/arch/x86/kernel/Makefile index af2bc36ca1c..64939a0c398 100644 --- a/arch/x86/kernel/Makefile +++ b/arch/x86/kernel/Makefile @@ -14,7 +14,7 @@ CFLAGS_REMOVE_paravirt-spinlocks.o = -pg CFLAGS_REMOVE_ftrace.o = -pg endif -ifdef CONFIG_FUNCTION_RET_TRACER +ifdef CONFIG_FUNCTION_GRAPH_TRACER # Don't trace __switch_to() but let it for function tracer CFLAGS_REMOVE_process_32.o = -pg endif @@ -70,7 +70,7 @@ obj-$(CONFIG_X86_LOCAL_APIC) += apic.o nmi.o obj-$(CONFIG_X86_IO_APIC) += io_apic.o obj-$(CONFIG_X86_REBOOTFIXUPS) += reboot_fixups_32.o obj-$(CONFIG_DYNAMIC_FTRACE) += ftrace.o -obj-$(CONFIG_FUNCTION_RET_TRACER) += ftrace.o +obj-$(CONFIG_FUNCTION_GRAPH_TRACER) += ftrace.o obj-$(CONFIG_KEXEC) += machine_kexec_$(BITS).o obj-$(CONFIG_KEXEC) += relocate_kernel_$(BITS).o crash.o obj-$(CONFIG_CRASH_DUMP) += crash_dump_$(BITS).o diff --git a/arch/x86/kernel/entry_32.S b/arch/x86/kernel/entry_32.S index 74defe21ba4..2b1f0f081a6 100644 --- a/arch/x86/kernel/entry_32.S +++ b/arch/x86/kernel/entry_32.S @@ -1188,9 +1188,9 @@ ENTRY(mcount) cmpl $ftrace_stub, ftrace_trace_function jnz trace -#ifdef CONFIG_FUNCTION_RET_TRACER - cmpl $ftrace_stub, ftrace_function_return - jnz ftrace_return_caller +#ifdef CONFIG_FUNCTION_GRAPH_TRACER + cmpl $ftrace_stub, ftrace_graph_function + jnz ftrace_graph_caller #endif .globl ftrace_stub ftrace_stub: @@ -1215,8 +1215,8 @@ END(mcount) #endif /* CONFIG_DYNAMIC_FTRACE */ #endif /* CONFIG_FUNCTION_TRACER */ -#ifdef CONFIG_FUNCTION_RET_TRACER -ENTRY(ftrace_return_caller) +#ifdef CONFIG_FUNCTION_GRAPH_TRACER +ENTRY(ftrace_graph_caller) cmpl $0, function_trace_stop jne ftrace_stub @@ -1230,7 +1230,7 @@ ENTRY(ftrace_return_caller) popl %ecx popl %eax ret -END(ftrace_return_caller) +END(ftrace_graph_caller) .globl return_to_handler return_to_handler: diff --git a/arch/x86/kernel/ftrace.c b/arch/x86/kernel/ftrace.c index bb137f7297e..3595a4c14ab 100644 --- a/arch/x86/kernel/ftrace.c +++ b/arch/x86/kernel/ftrace.c @@ -323,7 +323,7 @@ int __init ftrace_dyn_arch_init(void *data) } #endif -#ifdef CONFIG_FUNCTION_RET_TRACER +#ifdef CONFIG_FUNCTION_GRAPH_TRACER #ifndef CONFIG_DYNAMIC_FTRACE @@ -389,11 +389,11 @@ static void pop_return_trace(unsigned long *ret, unsigned long long *time, */ unsigned long ftrace_return_to_handler(void) { - struct ftrace_retfunc trace; + struct ftrace_graph_ret trace; pop_return_trace(&trace.ret, &trace.calltime, &trace.func, &trace.overrun); trace.rettime = cpu_clock(raw_smp_processor_id()); - ftrace_function_return(&trace); + ftrace_graph_function(&trace); return trace.ret; } @@ -440,12 +440,12 @@ void prepare_ftrace_return(unsigned long *parent, unsigned long self_addr) ); if (WARN_ON(faulted)) { - unregister_ftrace_return(); + unregister_ftrace_graph(); return; } if (WARN_ON(!__kernel_text_address(old))) { - unregister_ftrace_return(); + unregister_ftrace_graph(); *parent = old; return; } @@ -456,4 +456,4 @@ void prepare_ftrace_return(unsigned long *parent, unsigned long self_addr) *parent = old; } -#endif /* CONFIG_FUNCTION_RET_TRACER */ +#endif /* CONFIG_FUNCTION_GRAPH_TRACER */ diff --git a/include/linux/ftrace.h b/include/linux/ftrace.h index 7854d87b97b..b4ac734ad8d 100644 --- a/include/linux/ftrace.h +++ b/include/linux/ftrace.h @@ -115,8 +115,8 @@ extern int ftrace_update_ftrace_func(ftrace_func_t func); extern void ftrace_caller(void); extern void ftrace_call(void); extern void mcount_call(void); -#ifdef CONFIG_FUNCTION_RET_TRACER -extern void ftrace_return_caller(void); +#ifdef CONFIG_FUNCTION_GRAPH_TRACER +extern void ftrace_graph_caller(void); #endif /** @@ -315,7 +315,7 @@ ftrace_init_module(struct module *mod, /* * Structure that defines a return function trace. */ -struct ftrace_retfunc { +struct ftrace_graph_ret { unsigned long ret; /* Return address */ unsigned long func; /* Current function */ unsigned long long calltime; @@ -324,22 +324,22 @@ struct ftrace_retfunc { unsigned long overrun; }; -#ifdef CONFIG_FUNCTION_RET_TRACER +#ifdef CONFIG_FUNCTION_GRAPH_TRACER #define FTRACE_RETFUNC_DEPTH 50 #define FTRACE_RETSTACK_ALLOC_SIZE 32 /* Type of a callback handler of tracing return function */ -typedef void (*trace_function_return_t)(struct ftrace_retfunc *); +typedef void (*trace_function_graph_t)(struct ftrace_graph_ret *); -extern int register_ftrace_return(trace_function_return_t func); +extern int register_ftrace_graph(trace_function_graph_t func); /* The current handler in use */ -extern trace_function_return_t ftrace_function_return; -extern void unregister_ftrace_return(void); +extern trace_function_graph_t ftrace_graph_function; +extern void unregister_ftrace_graph(void); -extern void ftrace_retfunc_init_task(struct task_struct *t); -extern void ftrace_retfunc_exit_task(struct task_struct *t); +extern void ftrace_graph_init_task(struct task_struct *t); +extern void ftrace_graph_exit_task(struct task_struct *t); #else -static inline void ftrace_retfunc_init_task(struct task_struct *t) { } -static inline void ftrace_retfunc_exit_task(struct task_struct *t) { } +static inline void ftrace_graph_init_task(struct task_struct *t) { } +static inline void ftrace_graph_exit_task(struct task_struct *t) { } #endif #endif /* _LINUX_FTRACE_H */ diff --git a/include/linux/ftrace_irq.h b/include/linux/ftrace_irq.h index 0b4df55d7a7..366a054d0b0 100644 --- a/include/linux/ftrace_irq.h +++ b/include/linux/ftrace_irq.h @@ -2,7 +2,7 @@ #define _LINUX_FTRACE_IRQ_H -#if defined(CONFIG_DYNAMIC_FTRACE) || defined(CONFIG_FUNCTION_RET_TRACER) +#if defined(CONFIG_DYNAMIC_FTRACE) || defined(CONFIG_FUNCTION_GRAPH_TRACER) extern void ftrace_nmi_enter(void); extern void ftrace_nmi_exit(void); #else diff --git a/include/linux/sched.h b/include/linux/sched.h index d02a0ca70ee..7ad48f2a275 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -1365,7 +1365,7 @@ struct task_struct { unsigned long default_timer_slack_ns; struct list_head *scm_work_list; -#ifdef CONFIG_FUNCTION_RET_TRACER +#ifdef CONFIG_FUNCTION_GRAPH_TRACER /* Index of current stored adress in ret_stack */ int curr_ret_stack; /* Stack of return addresses for return function tracing */ diff --git a/kernel/Makefile b/kernel/Makefile index 03a45e7e87b..703cf3b7389 100644 --- a/kernel/Makefile +++ b/kernel/Makefile @@ -21,7 +21,7 @@ CFLAGS_REMOVE_cgroup-debug.o = -pg CFLAGS_REMOVE_sched_clock.o = -pg CFLAGS_REMOVE_sched.o = -pg endif -ifdef CONFIG_FUNCTION_RET_TRACER +ifdef CONFIG_FUNCTION_GRAPH_TRACER CFLAGS_REMOVE_extable.o = -pg # For __kernel_text_address() CFLAGS_REMOVE_module.o = -pg # For __module_text_address() endif diff --git a/kernel/fork.c b/kernel/fork.c index d6e1a3205f6..5f82a999c03 100644 --- a/kernel/fork.c +++ b/kernel/fork.c @@ -140,7 +140,7 @@ void free_task(struct task_struct *tsk) prop_local_destroy_single(&tsk->dirties); free_thread_info(tsk->stack); rt_mutex_debug_task_free(tsk); - ftrace_retfunc_exit_task(tsk); + ftrace_graph_exit_task(tsk); free_task_struct(tsk); } EXPORT_SYMBOL(free_task); @@ -1271,7 +1271,7 @@ static struct task_struct *copy_process(unsigned long clone_flags, total_forks++; spin_unlock(¤t->sighand->siglock); write_unlock_irq(&tasklist_lock); - ftrace_retfunc_init_task(p); + ftrace_graph_init_task(p); proc_fork_connector(p); cgroup_post_fork(p); return p; diff --git a/kernel/sched.c b/kernel/sched.c index 388d9db044a..52490bf6b88 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -5901,7 +5901,7 @@ void __cpuinit init_idle(struct task_struct *idle, int cpu) * The idle tasks have their own, simple scheduling class: */ idle->sched_class = &idle_sched_class; - ftrace_retfunc_init_task(idle); + ftrace_graph_init_task(idle); } /* diff --git a/kernel/trace/Kconfig b/kernel/trace/Kconfig index 620feadff67..eb9b901e077 100644 --- a/kernel/trace/Kconfig +++ b/kernel/trace/Kconfig @@ -12,7 +12,7 @@ config NOP_TRACER config HAVE_FUNCTION_TRACER bool -config HAVE_FUNCTION_RET_TRACER +config HAVE_FUNCTION_GRAPH_TRACER bool config HAVE_FUNCTION_TRACE_MCOUNT_TEST @@ -63,15 +63,18 @@ config FUNCTION_TRACER (the bootup default), then the overhead of the instructions is very small and not measurable even in micro-benchmarks. -config FUNCTION_RET_TRACER - bool "Kernel Function return Tracer" - depends on HAVE_FUNCTION_RET_TRACER +config FUNCTION_GRAPH_TRACER + bool "Kernel Function Graph Tracer" + depends on HAVE_FUNCTION_GRAPH_TRACER depends on FUNCTION_TRACER help - Enable the kernel to trace a function at its return. - It's first purpose is to trace the duration of functions. - This is done by setting the current return address on the thread - info structure of the current task. + Enable the kernel to trace a function at both its return + and its entry. + It's first purpose is to trace the duration of functions and + draw a call graph for each thread with some informations like + the return value. + This is done by setting the current return address on the current + task structure into a stack of calls. config IRQSOFF_TRACER bool "Interrupts-off Latency Tracer" diff --git a/kernel/trace/Makefile b/kernel/trace/Makefile index cef4bcb4e82..08c5fe6ddc0 100644 --- a/kernel/trace/Makefile +++ b/kernel/trace/Makefile @@ -29,7 +29,7 @@ obj-$(CONFIG_NOP_TRACER) += trace_nop.o obj-$(CONFIG_STACK_TRACER) += trace_stack.o obj-$(CONFIG_MMIOTRACE) += trace_mmiotrace.o obj-$(CONFIG_BOOT_TRACER) += trace_boot.o -obj-$(CONFIG_FUNCTION_RET_TRACER) += trace_functions_return.o +obj-$(CONFIG_FUNCTION_GRAPH_TRACER) += trace_functions_graph.o obj-$(CONFIG_TRACE_BRANCH_PROFILING) += trace_branch.o obj-$(CONFIG_BTS_TRACER) += trace_bts.o diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index 53042f118f2..9e19976af72 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -395,11 +395,11 @@ __ftrace_replace_code(struct dyn_ftrace *rec, int enable) unsigned long ip, fl; unsigned long ftrace_addr; -#ifdef CONFIG_FUNCTION_RET_TRACER +#ifdef CONFIG_FUNCTION_GRAPH_TRACER if (ftrace_tracing_type == FTRACE_TYPE_ENTER) ftrace_addr = (unsigned long)ftrace_caller; else - ftrace_addr = (unsigned long)ftrace_return_caller; + ftrace_addr = (unsigned long)ftrace_graph_caller; #else ftrace_addr = (unsigned long)ftrace_caller; #endif @@ -1496,13 +1496,13 @@ ftrace_enable_sysctl(struct ctl_table *table, int write, return ret; } -#ifdef CONFIG_FUNCTION_RET_TRACER +#ifdef CONFIG_FUNCTION_GRAPH_TRACER static atomic_t ftrace_retfunc_active; /* The callback that hooks the return of a function */ -trace_function_return_t ftrace_function_return = - (trace_function_return_t)ftrace_stub; +trace_function_graph_t ftrace_graph_function = + (trace_function_graph_t)ftrace_stub; /* Try to assign a return stack array on FTRACE_RETSTACK_ALLOC_SIZE tasks. */ @@ -1549,7 +1549,7 @@ free: } /* Allocate a return stack for each task */ -static int start_return_tracing(void) +static int start_graph_tracing(void) { struct ftrace_ret_stack **ret_stack_list; int ret; @@ -1569,7 +1569,7 @@ static int start_return_tracing(void) return ret; } -int register_ftrace_return(trace_function_return_t func) +int register_ftrace_graph(trace_function_graph_t func) { int ret = 0; @@ -1584,13 +1584,13 @@ int register_ftrace_return(trace_function_return_t func) goto out; } atomic_inc(&ftrace_retfunc_active); - ret = start_return_tracing(); + ret = start_graph_tracing(); if (ret) { atomic_dec(&ftrace_retfunc_active); goto out; } ftrace_tracing_type = FTRACE_TYPE_RETURN; - ftrace_function_return = func; + ftrace_graph_function = func; ftrace_startup(); out: @@ -1598,12 +1598,12 @@ out: return ret; } -void unregister_ftrace_return(void) +void unregister_ftrace_graph(void) { mutex_lock(&ftrace_sysctl_lock); atomic_dec(&ftrace_retfunc_active); - ftrace_function_return = (trace_function_return_t)ftrace_stub; + ftrace_graph_function = (trace_function_graph_t)ftrace_stub; ftrace_shutdown(); /* Restore normal tracing type */ ftrace_tracing_type = FTRACE_TYPE_ENTER; @@ -1612,7 +1612,7 @@ void unregister_ftrace_return(void) } /* Allocate a return stack for newly created task */ -void ftrace_retfunc_init_task(struct task_struct *t) +void ftrace_graph_init_task(struct task_struct *t) { if (atomic_read(&ftrace_retfunc_active)) { t->ret_stack = kmalloc(FTRACE_RETFUNC_DEPTH @@ -1626,7 +1626,7 @@ void ftrace_retfunc_init_task(struct task_struct *t) t->ret_stack = NULL; } -void ftrace_retfunc_exit_task(struct task_struct *t) +void ftrace_graph_exit_task(struct task_struct *t) { struct ftrace_ret_stack *ret_stack = t->ret_stack; diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index 8df8fdd69c9..f21ab2c68fd 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -878,15 +878,15 @@ trace_function(struct trace_array *tr, struct trace_array_cpu *data, ring_buffer_unlock_commit(tr->buffer, event, irq_flags); } -#ifdef CONFIG_FUNCTION_RET_TRACER -static void __trace_function_return(struct trace_array *tr, +#ifdef CONFIG_FUNCTION_GRAPH_TRACER +static void __trace_function_graph(struct trace_array *tr, struct trace_array_cpu *data, - struct ftrace_retfunc *trace, + struct ftrace_graph_ret *trace, unsigned long flags, int pc) { struct ring_buffer_event *event; - struct ftrace_ret_entry *entry; + struct ftrace_graph_entry *entry; unsigned long irq_flags; if (unlikely(local_read(&__get_cpu_var(ftrace_cpu_disabled)))) @@ -1177,8 +1177,8 @@ function_trace_call(unsigned long ip, unsigned long parent_ip) local_irq_restore(flags); } -#ifdef CONFIG_FUNCTION_RET_TRACER -void trace_function_return(struct ftrace_retfunc *trace) +#ifdef CONFIG_FUNCTION_GRAPH_TRACER +void trace_function_graph(struct ftrace_graph_ret *trace) { struct trace_array *tr = &global_trace; struct trace_array_cpu *data; @@ -1193,12 +1193,12 @@ void trace_function_return(struct ftrace_retfunc *trace) disabled = atomic_inc_return(&data->disabled); if (likely(disabled == 1)) { pc = preempt_count(); - __trace_function_return(tr, data, trace, flags, pc); + __trace_function_graph(tr, data, trace, flags, pc); } atomic_dec(&data->disabled); raw_local_irq_restore(flags); } -#endif /* CONFIG_FUNCTION_RET_TRACER */ +#endif /* CONFIG_FUNCTION_GRAPH_TRACER */ static struct ftrace_ops trace_ops __read_mostly = { @@ -2001,7 +2001,7 @@ static enum print_line_t print_trace_fmt(struct trace_iterator *iter) break; } case TRACE_FN_RET: { - return print_return_function(iter); + return print_graph_function(iter); break; } case TRACE_BRANCH: { diff --git a/kernel/trace/trace.h b/kernel/trace/trace.h index 3abd645e8af..72b5ef86876 100644 --- a/kernel/trace/trace.h +++ b/kernel/trace/trace.h @@ -57,7 +57,7 @@ struct ftrace_entry { }; /* Function return entry */ -struct ftrace_ret_entry { +struct ftrace_graph_entry { struct trace_entry ent; unsigned long ip; unsigned long parent_ip; @@ -264,7 +264,7 @@ extern void __ftrace_bad_type(void); IF_ASSIGN(var, ent, struct trace_boot_call, TRACE_BOOT_CALL);\ IF_ASSIGN(var, ent, struct trace_boot_ret, TRACE_BOOT_RET);\ IF_ASSIGN(var, ent, struct trace_branch, TRACE_BRANCH); \ - IF_ASSIGN(var, ent, struct ftrace_ret_entry, TRACE_FN_RET);\ + IF_ASSIGN(var, ent, struct ftrace_graph_entry, TRACE_FN_RET);\ IF_ASSIGN(var, ent, struct bts_entry, TRACE_BTS);\ __ftrace_bad_type(); \ } while (0) @@ -398,7 +398,7 @@ void trace_function(struct trace_array *tr, unsigned long parent_ip, unsigned long flags, int pc); void -trace_function_return(struct ftrace_retfunc *trace); +trace_function_graph(struct ftrace_graph_ret *trace); void trace_bts(struct trace_array *tr, unsigned long from, @@ -489,11 +489,11 @@ extern int trace_vprintk(unsigned long ip, const char *fmt, va_list args); extern unsigned long trace_flags; /* Standard output formatting function used for function return traces */ -#ifdef CONFIG_FUNCTION_RET_TRACER -extern enum print_line_t print_return_function(struct trace_iterator *iter); +#ifdef CONFIG_FUNCTION_GRAPH_TRACER +extern enum print_line_t print_graph_function(struct trace_iterator *iter); #else static inline enum print_line_t -print_return_function(struct trace_iterator *iter) +print_graph_function(struct trace_iterator *iter) { return TRACE_TYPE_UNHANDLED; } diff --git a/kernel/trace/trace_functions_graph.c b/kernel/trace/trace_functions_graph.c new file mode 100644 index 00000000000..f5bad4624d2 --- /dev/null +++ b/kernel/trace/trace_functions_graph.c @@ -0,0 +1,98 @@ +/* + * + * Function graph tracer. + * Copyright (c) 2008 Frederic Weisbecker + * Mostly borrowed from function tracer which + * is Copyright (c) Steven Rostedt + * + */ +#include +#include +#include +#include + +#include "trace.h" + + +#define TRACE_GRAPH_PRINT_OVERRUN 0x1 +static struct tracer_opt trace_opts[] = { + /* Display overruns or not */ + { TRACER_OPT(overrun, TRACE_GRAPH_PRINT_OVERRUN) }, + { } /* Empty entry */ +}; + +static struct tracer_flags tracer_flags = { + .val = 0, /* Don't display overruns by default */ + .opts = trace_opts +}; + + +static int graph_trace_init(struct trace_array *tr) +{ + int cpu; + for_each_online_cpu(cpu) + tracing_reset(tr, cpu); + + return register_ftrace_graph(&trace_function_graph); +} + +static void graph_trace_reset(struct trace_array *tr) +{ + unregister_ftrace_graph(); +} + + +enum print_line_t +print_graph_function(struct trace_iterator *iter) +{ + struct trace_seq *s = &iter->seq; + struct trace_entry *entry = iter->ent; + struct ftrace_graph_entry *field; + int ret; + + if (entry->type == TRACE_FN_RET) { + trace_assign_type(field, entry); + ret = trace_seq_printf(s, "%pF -> ", (void *)field->parent_ip); + if (!ret) + return TRACE_TYPE_PARTIAL_LINE; + + ret = seq_print_ip_sym(s, field->ip, + trace_flags & TRACE_ITER_SYM_MASK); + if (!ret) + return TRACE_TYPE_PARTIAL_LINE; + + ret = trace_seq_printf(s, " (%llu ns)", + field->rettime - field->calltime); + if (!ret) + return TRACE_TYPE_PARTIAL_LINE; + + if (tracer_flags.val & TRACE_GRAPH_PRINT_OVERRUN) { + ret = trace_seq_printf(s, " (Overruns: %lu)", + field->overrun); + if (!ret) + return TRACE_TYPE_PARTIAL_LINE; + } + + ret = trace_seq_printf(s, "\n"); + if (!ret) + return TRACE_TYPE_PARTIAL_LINE; + + return TRACE_TYPE_HANDLED; + } + return TRACE_TYPE_UNHANDLED; +} + +static struct tracer graph_trace __read_mostly = { + .name = "function-graph", + .init = graph_trace_init, + .reset = graph_trace_reset, + .print_line = print_graph_function, + .flags = &tracer_flags, +}; + +static __init int init_graph_trace(void) +{ + return register_tracer(&graph_trace); +} + +device_initcall(init_graph_trace); -- cgit v1.2.3-70-g09d2 From 287b6e68ca7209caec40b2f44f837c580a413bae Mon Sep 17 00:00:00 2001 From: Frederic Weisbecker Date: Wed, 26 Nov 2008 00:57:25 +0100 Subject: tracing/function-return-tracer: set a more human readable output Impact: feature This patch sets a C-like output for the function graph tracing. For this aim, we now call two handler for each function: one on the entry and one other on return. This way we can draw a well-ordered call stack. The pid of the previous trace is loosely stored to be compared against the one of the current trace to see if there were a context switch. Without this little feature, the call tree would seem broken at some locations. We could use the sched_tracer to capture these sched_events but this way of processing is much more simpler. 2 spaces have been chosen for indentation to fit the screen while deep calls. The time of execution in nanosecs is printed just after closed braces, it seems more easy this way to find the corresponding function. If the time was printed as a first column, it would be not so easy to find the corresponding function if it is called on a deep depth. I plan to output the return value but on 32 bits CPU, the return value can be 32 or 64, and its difficult to guess on which case we are. I don't know what would be the better solution on X86-32: only print eax (low-part) or even edx (high-part). Actually it's thee same problem when a function return a 8 bits value, the high part of eax could contain junk values... Here is an example of trace: sys_read() { fget_light() { } 526 vfs_read() { rw_verify_area() { security_file_permission() { cap_file_permission() { } 519 } 1564 } 2640 do_sync_read() { pipe_read() { __might_sleep() { } 511 pipe_wait() { prepare_to_wait() { } 760 deactivate_task() { dequeue_task() { dequeue_task_fair() { dequeue_entity() { update_curr() { update_min_vruntime() { } 504 } 1587 clear_buddies() { } 512 add_cfs_task_weight() { } 519 update_min_vruntime() { } 511 } 5602 dequeue_entity() { update_curr() { update_min_vruntime() { } 496 } 1631 clear_buddies() { } 496 update_min_vruntime() { } 527 } 4580 hrtick_update() { hrtick_start_fair() { } 488 } 1489 } 13700 } 14949 } 16016 msecs_to_jiffies() { } 496 put_prev_task_fair() { } 504 pick_next_task_fair() { } 489 pick_next_task_rt() { } 496 pick_next_task_fair() { } 489 pick_next_task_idle() { } 489 ------------8<---------- thread 4 ------------8<---------- finish_task_switch() { } 1203 do_softirq() { __do_softirq() { __local_bh_disable() { } 669 rcu_process_callbacks() { __rcu_process_callbacks() { cpu_quiet() { rcu_start_batch() { } 503 } 1647 } 3128 __rcu_process_callbacks() { } 542 } 5362 _local_bh_enable() { } 587 } 8880 } 9986 kthread_should_stop() { } 669 deactivate_task() { dequeue_task() { dequeue_task_fair() { dequeue_entity() { update_curr() { calc_delta_mine() { } 511 update_min_vruntime() { } 511 } 2813 Signed-off-by: Frederic Weisbecker Acked-by: Steven Rostedt Signed-off-by: Ingo Molnar --- arch/x86/kernel/ftrace.c | 32 +++++++---- include/linux/ftrace.h | 25 ++++++-- kernel/trace/ftrace.c | 30 +++++----- kernel/trace/trace.c | 67 ++++++++++++++++++---- kernel/trace/trace.h | 28 +++++---- kernel/trace/trace_functions_graph.c | 104 ++++++++++++++++++++++++++-------- kernel/trace/trace_functions_return.c | 98 -------------------------------- 7 files changed, 208 insertions(+), 176 deletions(-) delete mode 100644 kernel/trace/trace_functions_return.c (limited to 'kernel/trace/ftrace.c') diff --git a/arch/x86/kernel/ftrace.c b/arch/x86/kernel/ftrace.c index 3595a4c14ab..26b2d92d48b 100644 --- a/arch/x86/kernel/ftrace.c +++ b/arch/x86/kernel/ftrace.c @@ -347,7 +347,7 @@ void ftrace_nmi_exit(void) /* Add a function return address to the trace stack on thread info.*/ static int push_return_trace(unsigned long ret, unsigned long long time, - unsigned long func) + unsigned long func, int *depth) { int index; @@ -365,21 +365,22 @@ static int push_return_trace(unsigned long ret, unsigned long long time, current->ret_stack[index].ret = ret; current->ret_stack[index].func = func; current->ret_stack[index].calltime = time; + *depth = index; return 0; } /* Retrieve a function return address to the trace stack on thread info.*/ -static void pop_return_trace(unsigned long *ret, unsigned long long *time, - unsigned long *func, unsigned long *overrun) +static void pop_return_trace(struct ftrace_graph_ret *trace, unsigned long *ret) { int index; index = current->curr_ret_stack; *ret = current->ret_stack[index].ret; - *func = current->ret_stack[index].func; - *time = current->ret_stack[index].calltime; - *overrun = atomic_read(¤t->trace_overrun); + trace->func = current->ret_stack[index].func; + trace->calltime = current->ret_stack[index].calltime; + trace->overrun = atomic_read(¤t->trace_overrun); + trace->depth = index; current->curr_ret_stack--; } @@ -390,12 +391,13 @@ static void pop_return_trace(unsigned long *ret, unsigned long long *time, unsigned long ftrace_return_to_handler(void) { struct ftrace_graph_ret trace; - pop_return_trace(&trace.ret, &trace.calltime, &trace.func, - &trace.overrun); + unsigned long ret; + + pop_return_trace(&trace, &ret); trace.rettime = cpu_clock(raw_smp_processor_id()); - ftrace_graph_function(&trace); + ftrace_graph_return(&trace); - return trace.ret; + return ret; } /* @@ -407,6 +409,7 @@ void prepare_ftrace_return(unsigned long *parent, unsigned long self_addr) unsigned long old; unsigned long long calltime; int faulted; + struct ftrace_graph_ent trace; unsigned long return_hooker = (unsigned long) &return_to_handler; @@ -452,8 +455,15 @@ void prepare_ftrace_return(unsigned long *parent, unsigned long self_addr) calltime = cpu_clock(raw_smp_processor_id()); - if (push_return_trace(old, calltime, self_addr) == -EBUSY) + if (push_return_trace(old, calltime, + self_addr, &trace.depth) == -EBUSY) { *parent = old; + return; + } + + trace.func = self_addr; + ftrace_graph_entry(&trace); + } #endif /* CONFIG_FUNCTION_GRAPH_TRACER */ diff --git a/include/linux/ftrace.h b/include/linux/ftrace.h index b4ac734ad8d..fc2d5498719 100644 --- a/include/linux/ftrace.h +++ b/include/linux/ftrace.h @@ -312,27 +312,40 @@ ftrace_init_module(struct module *mod, #endif +/* + * Structure that defines an entry function trace. + */ +struct ftrace_graph_ent { + unsigned long func; /* Current function */ + int depth; +}; + /* * Structure that defines a return function trace. */ struct ftrace_graph_ret { - unsigned long ret; /* Return address */ unsigned long func; /* Current function */ unsigned long long calltime; unsigned long long rettime; /* Number of functions that overran the depth limit for current task */ unsigned long overrun; + int depth; }; #ifdef CONFIG_FUNCTION_GRAPH_TRACER #define FTRACE_RETFUNC_DEPTH 50 #define FTRACE_RETSTACK_ALLOC_SIZE 32 -/* Type of a callback handler of tracing return function */ -typedef void (*trace_function_graph_t)(struct ftrace_graph_ret *); +/* Type of the callback handlers for tracing function graph*/ +typedef void (*trace_func_graph_ret_t)(struct ftrace_graph_ret *); /* return */ +typedef void (*trace_func_graph_ent_t)(struct ftrace_graph_ent *); /* entry */ + +extern int register_ftrace_graph(trace_func_graph_ret_t retfunc, + trace_func_graph_ent_t entryfunc); + +/* The current handlers in use */ +extern trace_func_graph_ret_t ftrace_graph_return; +extern trace_func_graph_ent_t ftrace_graph_entry; -extern int register_ftrace_graph(trace_function_graph_t func); -/* The current handler in use */ -extern trace_function_graph_t ftrace_graph_function; extern void unregister_ftrace_graph(void); extern void ftrace_graph_init_task(struct task_struct *t); diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index 9e19976af72..7e2d3b91692 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -1498,12 +1498,13 @@ ftrace_enable_sysctl(struct ctl_table *table, int write, #ifdef CONFIG_FUNCTION_GRAPH_TRACER -static atomic_t ftrace_retfunc_active; - -/* The callback that hooks the return of a function */ -trace_function_graph_t ftrace_graph_function = - (trace_function_graph_t)ftrace_stub; +static atomic_t ftrace_graph_active; +/* The callbacks that hook a function */ +trace_func_graph_ret_t ftrace_graph_return = + (trace_func_graph_ret_t)ftrace_stub; +trace_func_graph_ent_t ftrace_graph_entry = + (trace_func_graph_ent_t)ftrace_stub; /* Try to assign a return stack array on FTRACE_RETSTACK_ALLOC_SIZE tasks. */ static int alloc_retstack_tasklist(struct ftrace_ret_stack **ret_stack_list) @@ -1569,7 +1570,8 @@ static int start_graph_tracing(void) return ret; } -int register_ftrace_graph(trace_function_graph_t func) +int register_ftrace_graph(trace_func_graph_ret_t retfunc, + trace_func_graph_ent_t entryfunc) { int ret = 0; @@ -1583,14 +1585,15 @@ int register_ftrace_graph(trace_function_graph_t func) ret = -EBUSY; goto out; } - atomic_inc(&ftrace_retfunc_active); + atomic_inc(&ftrace_graph_active); ret = start_graph_tracing(); if (ret) { - atomic_dec(&ftrace_retfunc_active); + atomic_dec(&ftrace_graph_active); goto out; } ftrace_tracing_type = FTRACE_TYPE_RETURN; - ftrace_graph_function = func; + ftrace_graph_return = retfunc; + ftrace_graph_entry = entryfunc; ftrace_startup(); out: @@ -1602,8 +1605,9 @@ void unregister_ftrace_graph(void) { mutex_lock(&ftrace_sysctl_lock); - atomic_dec(&ftrace_retfunc_active); - ftrace_graph_function = (trace_function_graph_t)ftrace_stub; + atomic_dec(&ftrace_graph_active); + ftrace_graph_return = (trace_func_graph_ret_t)ftrace_stub; + ftrace_graph_entry = (trace_func_graph_ent_t)ftrace_stub; ftrace_shutdown(); /* Restore normal tracing type */ ftrace_tracing_type = FTRACE_TYPE_ENTER; @@ -1614,7 +1618,7 @@ void unregister_ftrace_graph(void) /* Allocate a return stack for newly created task */ void ftrace_graph_init_task(struct task_struct *t) { - if (atomic_read(&ftrace_retfunc_active)) { + if (atomic_read(&ftrace_graph_active)) { t->ret_stack = kmalloc(FTRACE_RETFUNC_DEPTH * sizeof(struct ftrace_ret_stack), GFP_KERNEL); @@ -1638,5 +1642,3 @@ void ftrace_graph_exit_task(struct task_struct *t) } #endif - - diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index f21ab2c68fd..9d5f7c94f25 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -879,14 +879,38 @@ trace_function(struct trace_array *tr, struct trace_array_cpu *data, } #ifdef CONFIG_FUNCTION_GRAPH_TRACER -static void __trace_function_graph(struct trace_array *tr, +static void __trace_graph_entry(struct trace_array *tr, + struct trace_array_cpu *data, + struct ftrace_graph_ent *trace, + unsigned long flags, + int pc) +{ + struct ring_buffer_event *event; + struct ftrace_graph_ent_entry *entry; + unsigned long irq_flags; + + if (unlikely(local_read(&__get_cpu_var(ftrace_cpu_disabled)))) + return; + + event = ring_buffer_lock_reserve(global_trace.buffer, sizeof(*entry), + &irq_flags); + if (!event) + return; + entry = ring_buffer_event_data(event); + tracing_generic_entry_update(&entry->ent, flags, pc); + entry->ent.type = TRACE_GRAPH_ENT; + entry->graph_ent = *trace; + ring_buffer_unlock_commit(global_trace.buffer, event, irq_flags); +} + +static void __trace_graph_return(struct trace_array *tr, struct trace_array_cpu *data, struct ftrace_graph_ret *trace, unsigned long flags, int pc) { struct ring_buffer_event *event; - struct ftrace_graph_entry *entry; + struct ftrace_graph_ret_entry *entry; unsigned long irq_flags; if (unlikely(local_read(&__get_cpu_var(ftrace_cpu_disabled)))) @@ -898,12 +922,8 @@ static void __trace_function_graph(struct trace_array *tr, return; entry = ring_buffer_event_data(event); tracing_generic_entry_update(&entry->ent, flags, pc); - entry->ent.type = TRACE_FN_RET; - entry->ip = trace->func; - entry->parent_ip = trace->ret; - entry->rettime = trace->rettime; - entry->calltime = trace->calltime; - entry->overrun = trace->overrun; + entry->ent.type = TRACE_GRAPH_RET; + entry->ret = *trace; ring_buffer_unlock_commit(global_trace.buffer, event, irq_flags); } #endif @@ -1178,7 +1198,7 @@ function_trace_call(unsigned long ip, unsigned long parent_ip) } #ifdef CONFIG_FUNCTION_GRAPH_TRACER -void trace_function_graph(struct ftrace_graph_ret *trace) +void trace_graph_entry(struct ftrace_graph_ent *trace) { struct trace_array *tr = &global_trace; struct trace_array_cpu *data; @@ -1193,7 +1213,28 @@ void trace_function_graph(struct ftrace_graph_ret *trace) disabled = atomic_inc_return(&data->disabled); if (likely(disabled == 1)) { pc = preempt_count(); - __trace_function_graph(tr, data, trace, flags, pc); + __trace_graph_entry(tr, data, trace, flags, pc); + } + atomic_dec(&data->disabled); + raw_local_irq_restore(flags); +} + +void trace_graph_return(struct ftrace_graph_ret *trace) +{ + struct trace_array *tr = &global_trace; + struct trace_array_cpu *data; + unsigned long flags; + long disabled; + int cpu; + int pc; + + raw_local_irq_save(flags); + cpu = raw_smp_processor_id(); + data = tr->data[cpu]; + disabled = atomic_inc_return(&data->disabled); + if (likely(disabled == 1)) { + pc = preempt_count(); + __trace_graph_return(tr, data, trace, flags, pc); } atomic_dec(&data->disabled); raw_local_irq_restore(flags); @@ -2000,9 +2041,11 @@ static enum print_line_t print_trace_fmt(struct trace_iterator *iter) trace_seq_print_cont(s, iter); break; } - case TRACE_FN_RET: { + case TRACE_GRAPH_RET: { + return print_graph_function(iter); + } + case TRACE_GRAPH_ENT: { return print_graph_function(iter); - break; } case TRACE_BRANCH: { struct trace_branch *field; diff --git a/kernel/trace/trace.h b/kernel/trace/trace.h index 72b5ef86876..ffe1bb1eb62 100644 --- a/kernel/trace/trace.h +++ b/kernel/trace/trace.h @@ -25,7 +25,8 @@ enum trace_type { TRACE_BRANCH, TRACE_BOOT_CALL, TRACE_BOOT_RET, - TRACE_FN_RET, + TRACE_GRAPH_RET, + TRACE_GRAPH_ENT, TRACE_USER_STACK, TRACE_BTS, @@ -56,14 +57,16 @@ struct ftrace_entry { unsigned long parent_ip; }; +/* Function call entry */ +struct ftrace_graph_ent_entry { + struct trace_entry ent; + struct ftrace_graph_ent graph_ent; +}; + /* Function return entry */ -struct ftrace_graph_entry { - struct trace_entry ent; - unsigned long ip; - unsigned long parent_ip; - unsigned long long calltime; - unsigned long long rettime; - unsigned long overrun; +struct ftrace_graph_ret_entry { + struct trace_entry ent; + struct ftrace_graph_ret ret; }; extern struct tracer boot_tracer; @@ -264,7 +267,10 @@ extern void __ftrace_bad_type(void); IF_ASSIGN(var, ent, struct trace_boot_call, TRACE_BOOT_CALL);\ IF_ASSIGN(var, ent, struct trace_boot_ret, TRACE_BOOT_RET);\ IF_ASSIGN(var, ent, struct trace_branch, TRACE_BRANCH); \ - IF_ASSIGN(var, ent, struct ftrace_graph_entry, TRACE_FN_RET);\ + IF_ASSIGN(var, ent, struct ftrace_graph_ent_entry, \ + TRACE_GRAPH_ENT); \ + IF_ASSIGN(var, ent, struct ftrace_graph_ret_entry, \ + TRACE_GRAPH_RET); \ IF_ASSIGN(var, ent, struct bts_entry, TRACE_BTS);\ __ftrace_bad_type(); \ } while (0) @@ -397,9 +403,9 @@ void trace_function(struct trace_array *tr, unsigned long ip, unsigned long parent_ip, unsigned long flags, int pc); -void -trace_function_graph(struct ftrace_graph_ret *trace); +void trace_graph_return(struct ftrace_graph_ret *trace); +void trace_graph_entry(struct ftrace_graph_ent *trace); void trace_bts(struct trace_array *tr, unsigned long from, unsigned long to); diff --git a/kernel/trace/trace_functions_graph.c b/kernel/trace/trace_functions_graph.c index f5bad4624d2..b6f0cc2a00c 100644 --- a/kernel/trace/trace_functions_graph.c +++ b/kernel/trace/trace_functions_graph.c @@ -13,6 +13,7 @@ #include "trace.h" +#define TRACE_GRAPH_INDENT 2 #define TRACE_GRAPH_PRINT_OVERRUN 0x1 static struct tracer_opt trace_opts[] = { @@ -26,6 +27,8 @@ static struct tracer_flags tracer_flags = { .opts = trace_opts }; +/* pid on the last trace processed */ +static pid_t last_pid = -1; static int graph_trace_init(struct trace_array *tr) { @@ -33,7 +36,8 @@ static int graph_trace_init(struct trace_array *tr) for_each_online_cpu(cpu) tracing_reset(tr, cpu); - return register_ftrace_graph(&trace_function_graph); + return register_ftrace_graph(&trace_graph_return, + &trace_graph_entry); } static void graph_trace_reset(struct trace_array *tr) @@ -41,45 +45,97 @@ static void graph_trace_reset(struct trace_array *tr) unregister_ftrace_graph(); } +/* If the pid changed since the last trace, output this event */ +static int verif_pid(struct trace_seq *s, pid_t pid) +{ + if (last_pid != -1 && last_pid == pid) + return 1; -enum print_line_t -print_graph_function(struct trace_iterator *iter) + last_pid = pid; + return trace_seq_printf(s, "\n------------8<---------- thread %d" + " ------------8<----------\n\n", + pid); +} + +static enum print_line_t +print_graph_entry(struct ftrace_graph_ent *call, struct trace_seq *s, + struct trace_entry *ent) { - struct trace_seq *s = &iter->seq; - struct trace_entry *entry = iter->ent; - struct ftrace_graph_entry *field; + int i; int ret; - if (entry->type == TRACE_FN_RET) { - trace_assign_type(field, entry); - ret = trace_seq_printf(s, "%pF -> ", (void *)field->parent_ip); - if (!ret) - return TRACE_TYPE_PARTIAL_LINE; + if (!verif_pid(s, ent->pid)) + return TRACE_TYPE_PARTIAL_LINE; - ret = seq_print_ip_sym(s, field->ip, - trace_flags & TRACE_ITER_SYM_MASK); + for (i = 0; i < call->depth * TRACE_GRAPH_INDENT; i++) { + ret = trace_seq_printf(s, " "); if (!ret) return TRACE_TYPE_PARTIAL_LINE; + } + + ret = seq_print_ip_sym(s, call->func, 0); + if (!ret) + return TRACE_TYPE_PARTIAL_LINE; + + ret = trace_seq_printf(s, "() {\n"); + if (!ret) + return TRACE_TYPE_PARTIAL_LINE; + return TRACE_TYPE_HANDLED; +} + +static enum print_line_t +print_graph_return(struct ftrace_graph_ret *trace, struct trace_seq *s, + struct trace_entry *ent) +{ + int i; + int ret; + + if (!verif_pid(s, ent->pid)) + return TRACE_TYPE_PARTIAL_LINE; - ret = trace_seq_printf(s, " (%llu ns)", - field->rettime - field->calltime); + for (i = 0; i < trace->depth * TRACE_GRAPH_INDENT; i++) { + ret = trace_seq_printf(s, " "); if (!ret) return TRACE_TYPE_PARTIAL_LINE; + } + + ret = trace_seq_printf(s, "} "); + if (!ret) + return TRACE_TYPE_PARTIAL_LINE; - if (tracer_flags.val & TRACE_GRAPH_PRINT_OVERRUN) { - ret = trace_seq_printf(s, " (Overruns: %lu)", - field->overrun); - if (!ret) - return TRACE_TYPE_PARTIAL_LINE; - } + ret = trace_seq_printf(s, "%llu\n", trace->rettime - trace->calltime); + if (!ret) + return TRACE_TYPE_PARTIAL_LINE; - ret = trace_seq_printf(s, "\n"); + if (tracer_flags.val & TRACE_GRAPH_PRINT_OVERRUN) { + ret = trace_seq_printf(s, " (Overruns: %lu)\n", + trace->overrun); if (!ret) return TRACE_TYPE_PARTIAL_LINE; + } + return TRACE_TYPE_HANDLED; +} + +enum print_line_t +print_graph_function(struct trace_iterator *iter) +{ + struct trace_seq *s = &iter->seq; + struct trace_entry *entry = iter->ent; - return TRACE_TYPE_HANDLED; + switch (entry->type) { + case TRACE_GRAPH_ENT: { + struct ftrace_graph_ent_entry *field; + trace_assign_type(field, entry); + return print_graph_entry(&field->graph_ent, s, entry); + } + case TRACE_GRAPH_RET: { + struct ftrace_graph_ret_entry *field; + trace_assign_type(field, entry); + return print_graph_return(&field->ret, s, entry); + } + default: + return TRACE_TYPE_UNHANDLED; } - return TRACE_TYPE_UNHANDLED; } static struct tracer graph_trace __read_mostly = { diff --git a/kernel/trace/trace_functions_return.c b/kernel/trace/trace_functions_return.c deleted file mode 100644 index e00d64509c9..00000000000 --- a/kernel/trace/trace_functions_return.c +++ /dev/null @@ -1,98 +0,0 @@ -/* - * - * Function return tracer. - * Copyright (c) 2008 Frederic Weisbecker - * Mostly borrowed from function tracer which - * is Copyright (c) Steven Rostedt - * - */ -#include -#include -#include -#include - -#include "trace.h" - - -#define TRACE_RETURN_PRINT_OVERRUN 0x1 -static struct tracer_opt trace_opts[] = { - /* Display overruns or not */ - { TRACER_OPT(overrun, TRACE_RETURN_PRINT_OVERRUN) }, - { } /* Empty entry */ -}; - -static struct tracer_flags tracer_flags = { - .val = 0, /* Don't display overruns by default */ - .opts = trace_opts -}; - - -static int return_trace_init(struct trace_array *tr) -{ - int cpu; - for_each_online_cpu(cpu) - tracing_reset(tr, cpu); - - return register_ftrace_return(&trace_function_return); -} - -static void return_trace_reset(struct trace_array *tr) -{ - unregister_ftrace_return(); -} - - -enum print_line_t -print_return_function(struct trace_iterator *iter) -{ - struct trace_seq *s = &iter->seq; - struct trace_entry *entry = iter->ent; - struct ftrace_ret_entry *field; - int ret; - - if (entry->type == TRACE_FN_RET) { - trace_assign_type(field, entry); - ret = trace_seq_printf(s, "%pF -> ", (void *)field->parent_ip); - if (!ret) - return TRACE_TYPE_PARTIAL_LINE; - - ret = seq_print_ip_sym(s, field->ip, - trace_flags & TRACE_ITER_SYM_MASK); - if (!ret) - return TRACE_TYPE_PARTIAL_LINE; - - ret = trace_seq_printf(s, " (%llu ns)", - field->rettime - field->calltime); - if (!ret) - return TRACE_TYPE_PARTIAL_LINE; - - if (tracer_flags.val & TRACE_RETURN_PRINT_OVERRUN) { - ret = trace_seq_printf(s, " (Overruns: %lu)", - field->overrun); - if (!ret) - return TRACE_TYPE_PARTIAL_LINE; - } - - ret = trace_seq_printf(s, "\n"); - if (!ret) - return TRACE_TYPE_PARTIAL_LINE; - - return TRACE_TYPE_HANDLED; - } - return TRACE_TYPE_UNHANDLED; -} - -static struct tracer return_trace __read_mostly = { - .name = "return", - .init = return_trace_init, - .reset = return_trace_reset, - .print_line = print_return_function, - .flags = &tracer_flags, -}; - -static __init int init_return_trace(void) -{ - return register_tracer(&return_trace); -} - -device_initcall(init_return_trace); -- cgit v1.2.3-70-g09d2 From df4fc31558dd2a3a30292ddb3a64c2a5befcec73 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Wed, 26 Nov 2008 00:16:23 -0500 Subject: ftrace: add function tracing to single thread Impact: feature to function trace a single thread This patch adds the ability to function trace a single thread. The file: /debugfs/tracing/set_ftrace_pid contains the pid to trace. Valid pids are any positive integer. Writing any negative number to this file will disable the pid tracing and the function tracer will go back to tracing all of threads. This feature works with both static and dynamic function tracing. Signed-off-by: Steven Rostedt Signed-off-by: Ingo Molnar --- Documentation/ftrace.txt | 79 ++++++++++++++++++ kernel/trace/ftrace.c | 209 +++++++++++++++++++++++++++++++++++++++++------ 2 files changed, 262 insertions(+), 26 deletions(-) (limited to 'kernel/trace/ftrace.c') diff --git a/Documentation/ftrace.txt b/Documentation/ftrace.txt index 35a78bc6651..de05042f11b 100644 --- a/Documentation/ftrace.txt +++ b/Documentation/ftrace.txt @@ -127,6 +127,8 @@ of ftrace. Here is a list of some of the key files: be traced. If a function exists in both set_ftrace_filter and set_ftrace_notrace, the function will _not_ be traced. + set_ftrace_pid: Have the function tracer only trace a single thread. + available_filter_functions: This lists the functions that ftrace has processed and can trace. These are the function names that you can pass to "set_ftrace_filter" or @@ -1073,6 +1075,83 @@ For simple one time traces, the above is sufficent. For anything else, a search through /proc/mounts may be needed to find where the debugfs file-system is mounted. + +Single thread tracing +--------------------- + +By writing into /debug/tracing/set_ftrace_pid you can trace a +single thread. For example: + +# cat /debug/tracing/set_ftrace_pid +no pid +# echo 3111 > /debug/tracing/set_ftrace_pid +# cat /debug/tracing/set_ftrace_pid +3111 +# echo function > /debug/tracing/current_tracer +# cat /debug/tracing/trace | head + # tracer: function + # + # TASK-PID CPU# TIMESTAMP FUNCTION + # | | | | | + yum-updatesd-3111 [003] 1637.254676: finish_task_switch <-thread_return + yum-updatesd-3111 [003] 1637.254681: hrtimer_cancel <-schedule_hrtimeout_range + yum-updatesd-3111 [003] 1637.254682: hrtimer_try_to_cancel <-hrtimer_cancel + yum-updatesd-3111 [003] 1637.254683: lock_hrtimer_base <-hrtimer_try_to_cancel + yum-updatesd-3111 [003] 1637.254685: fget_light <-do_sys_poll + yum-updatesd-3111 [003] 1637.254686: pipe_poll <-do_sys_poll +# echo -1 > /debug/tracing/set_ftrace_pid +# cat /debug/tracing/trace |head + # tracer: function + # + # TASK-PID CPU# TIMESTAMP FUNCTION + # | | | | | + ##### CPU 3 buffer started #### + yum-updatesd-3111 [003] 1701.957688: free_poll_entry <-poll_freewait + yum-updatesd-3111 [003] 1701.957689: remove_wait_queue <-free_poll_entry + yum-updatesd-3111 [003] 1701.957691: fput <-free_poll_entry + yum-updatesd-3111 [003] 1701.957692: audit_syscall_exit <-sysret_audit + yum-updatesd-3111 [003] 1701.957693: path_put <-audit_syscall_exit + +If you want to trace a function when executing, you could use +something like this simple program: + +#include +#include +#include +#include +#include +#include + +int main (int argc, char **argv) +{ + if (argc < 1) + exit(-1); + + if (fork() > 0) { + int fd, ffd; + char line[64]; + int s; + + ffd = open("/debug/tracing/current_tracer", O_WRONLY); + if (ffd < 0) + exit(-1); + write(ffd, "nop", 3); + + fd = open("/debug/tracing/set_ftrace_pid", O_WRONLY); + s = sprintf(line, "%d\n", getpid()); + write(fd, line, s); + + write(ffd, "function", 8); + + close(fd); + close(ffd); + + execvp(argv[1], argv+1); + } + + return 0; +} + dynamic ftrace -------------- diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index 7e2d3b91692..00d98c65fad 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -47,6 +47,9 @@ int ftrace_enabled __read_mostly; static int last_ftrace_enabled; +/* ftrace_pid_trace >= 0 will only trace threads with this pid */ +static int ftrace_pid_trace = -1; + /* Quick disabling of function tracer. */ int function_trace_stop; @@ -61,6 +64,7 @@ static int ftrace_disabled __read_mostly; static DEFINE_SPINLOCK(ftrace_lock); static DEFINE_MUTEX(ftrace_sysctl_lock); +static DEFINE_MUTEX(ftrace_start_lock); static struct ftrace_ops ftrace_list_end __read_mostly = { @@ -70,6 +74,7 @@ static struct ftrace_ops ftrace_list_end __read_mostly = static struct ftrace_ops *ftrace_list __read_mostly = &ftrace_list_end; 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; static void ftrace_list_func(unsigned long ip, unsigned long parent_ip) { @@ -86,6 +91,21 @@ static void ftrace_list_func(unsigned long ip, unsigned long parent_ip) }; } +static void ftrace_pid_func(unsigned long ip, unsigned long parent_ip) +{ + if (current->pid != ftrace_pid_trace) + return; + + ftrace_pid_function(ip, parent_ip); +} + +static void set_ftrace_pid_function(ftrace_func_t func) +{ + /* do not set ftrace_pid_function to itself! */ + if (func != ftrace_pid_func) + ftrace_pid_function = func; +} + /** * clear_ftrace_function - reset the ftrace function * @@ -96,6 +116,7 @@ void clear_ftrace_function(void) { ftrace_trace_function = ftrace_stub; __ftrace_trace_function = ftrace_stub; + ftrace_pid_function = ftrace_stub; } #ifndef CONFIG_HAVE_FUNCTION_TRACE_MCOUNT_TEST @@ -128,20 +149,26 @@ static int __register_ftrace_function(struct ftrace_ops *ops) ftrace_list = ops; if (ftrace_enabled) { + ftrace_func_t func; + + if (ops->next == &ftrace_list_end) + func = ops->func; + else + func = ftrace_list_func; + + if (ftrace_pid_trace >= 0) { + set_ftrace_pid_function(func); + func = ftrace_pid_func; + } + /* * For one func, simply call it directly. * For more than one func, call the chain. */ #ifdef CONFIG_HAVE_FUNCTION_TRACE_MCOUNT_TEST - if (ops->next == &ftrace_list_end) - ftrace_trace_function = ops->func; - else - ftrace_trace_function = ftrace_list_func; + ftrace_trace_function = func; #else - if (ops->next == &ftrace_list_end) - __ftrace_trace_function = ops->func; - else - __ftrace_trace_function = ftrace_list_func; + __ftrace_trace_function = func; ftrace_trace_function = ftrace_test_stop_func; #endif } @@ -182,8 +209,19 @@ static int __unregister_ftrace_function(struct ftrace_ops *ops) if (ftrace_enabled) { /* If we only have one func left, then call that directly */ - if (ftrace_list->next == &ftrace_list_end) - ftrace_trace_function = ftrace_list->func; + if (ftrace_list->next == &ftrace_list_end) { + ftrace_func_t func = ftrace_list->func; + + if (ftrace_pid_trace >= 0) { + set_ftrace_pid_function(func); + func = ftrace_pid_func; + } +#ifdef CONFIG_HAVE_FUNCTION_TRACE_MCOUNT_TEST + ftrace_trace_function = func; +#else + __ftrace_trace_function = func; +#endif + } } out: @@ -192,6 +230,38 @@ static int __unregister_ftrace_function(struct ftrace_ops *ops) return ret; } +static void ftrace_update_pid_func(void) +{ + ftrace_func_t func; + + /* should not be called from interrupt context */ + spin_lock(&ftrace_lock); + + if (ftrace_trace_function == ftrace_stub) + goto out; + + func = ftrace_trace_function; + + if (ftrace_pid_trace >= 0) { + set_ftrace_pid_function(func); + func = ftrace_pid_func; + } else { + if (func != ftrace_pid_func) + goto out; + + set_ftrace_pid_function(func); + } + +#ifdef CONFIG_HAVE_FUNCTION_TRACE_MCOUNT_TEST + ftrace_trace_function = func; +#else + __ftrace_trace_function = func; +#endif + + out: + spin_unlock(&ftrace_lock); +} + #ifdef CONFIG_DYNAMIC_FTRACE #ifndef CONFIG_FTRACE_MCOUNT_RECORD # error Dynamic ftrace depends on MCOUNT_RECORD @@ -545,7 +615,19 @@ static void ftrace_run_update_code(int command) static ftrace_func_t saved_ftrace_func; static int ftrace_start_up; -static DEFINE_MUTEX(ftrace_start_lock); + +static void ftrace_startup_enable(int command) +{ + if (saved_ftrace_func != ftrace_trace_function) { + saved_ftrace_func = ftrace_trace_function; + command |= FTRACE_UPDATE_TRACE_FUNC; + } + + if (!command || !ftrace_enabled) + return; + + ftrace_run_update_code(command); +} static void ftrace_startup(void) { @@ -558,16 +640,8 @@ static void ftrace_startup(void) ftrace_start_up++; command |= FTRACE_ENABLE_CALLS; - if (saved_ftrace_func != ftrace_trace_function) { - saved_ftrace_func = ftrace_trace_function; - command |= FTRACE_UPDATE_TRACE_FUNC; - } - - if (!command || !ftrace_enabled) - goto out; + ftrace_startup_enable(command); - ftrace_run_update_code(command); - out: mutex_unlock(&ftrace_start_lock); } @@ -1262,13 +1336,10 @@ static struct file_operations ftrace_notrace_fops = { .release = ftrace_notrace_release, }; -static __init int ftrace_init_debugfs(void) +static __init int ftrace_init_dyn_debugfs(struct dentry *d_tracer) { - struct dentry *d_tracer; struct dentry *entry; - d_tracer = tracing_init_dentry(); - entry = debugfs_create_file("available_filter_functions", 0444, d_tracer, NULL, &ftrace_avail_fops); if (!entry) @@ -1295,8 +1366,6 @@ static __init int ftrace_init_debugfs(void) return 0; } -fs_initcall(ftrace_init_debugfs); - static int ftrace_convert_nops(struct module *mod, unsigned long *start, unsigned long *end) @@ -1382,12 +1451,100 @@ static int __init ftrace_nodyn_init(void) } device_initcall(ftrace_nodyn_init); +static inline int ftrace_init_dyn_debugfs(struct dentry *d_tracer) { return 0; } +static inline void ftrace_startup_enable(int command) { } # define ftrace_startup() do { } while (0) # define ftrace_shutdown() do { } while (0) # define ftrace_startup_sysctl() do { } while (0) # define ftrace_shutdown_sysctl() do { } while (0) #endif /* CONFIG_DYNAMIC_FTRACE */ +static ssize_t +ftrace_pid_read(struct file *file, char __user *ubuf, + size_t cnt, loff_t *ppos) +{ + char buf[64]; + int r; + + if (ftrace_pid_trace >= 0) + r = sprintf(buf, "%u\n", ftrace_pid_trace); + else + r = sprintf(buf, "no pid\n"); + + return simple_read_from_buffer(ubuf, cnt, ppos, buf, r); +} + +static ssize_t +ftrace_pid_write(struct file *filp, const char __user *ubuf, + size_t cnt, loff_t *ppos) +{ + char buf[64]; + long val; + int ret; + + if (cnt >= sizeof(buf)) + return -EINVAL; + + if (copy_from_user(&buf, ubuf, cnt)) + return -EFAULT; + + buf[cnt] = 0; + + ret = strict_strtol(buf, 10, &val); + if (ret < 0) + return ret; + + mutex_lock(&ftrace_start_lock); + if (ret < 0) { + /* disable pid tracing */ + if (ftrace_pid_trace < 0) + goto out; + ftrace_pid_trace = -1; + + } else { + + if (ftrace_pid_trace == val) + goto out; + + ftrace_pid_trace = val; + } + + /* update the function call */ + ftrace_update_pid_func(); + ftrace_startup_enable(0); + + out: + mutex_unlock(&ftrace_start_lock); + + return cnt; +} + +static struct file_operations ftrace_pid_fops = { + .read = ftrace_pid_read, + .write = ftrace_pid_write, +}; + +static __init int ftrace_init_debugfs(void) +{ + struct dentry *d_tracer; + struct dentry *entry; + + d_tracer = tracing_init_dentry(); + if (!d_tracer) + return 0; + + ftrace_init_dyn_debugfs(d_tracer); + + entry = debugfs_create_file("set_ftrace_pid", 0644, d_tracer, + NULL, &ftrace_pid_fops); + if (!entry) + pr_warning("Could not create debugfs " + "'set_ftrace_pid' entry\n"); + return 0; +} + +fs_initcall(ftrace_init_debugfs); + /** * ftrace_kill - kill ftrace * -- cgit v1.2.3-70-g09d2 From 5a45cfe1c64862e8cd3b0d79d7c4ba71c3118915 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Wed, 26 Nov 2008 00:16:24 -0500 Subject: ftrace: use code patching for ftrace graph tracer Impact: more efficient code for ftrace graph tracer This patch uses the dynamic patching, when available, to patch the function graph code into the kernel. This patch will ease the way for letting both function tracing and function graph tracing run together. Signed-off-by: Steven Rostedt Signed-off-by: Ingo Molnar --- arch/x86/kernel/entry_32.S | 5 +++++ arch/x86/kernel/ftrace.c | 48 ++++++++++++++++++++++++++++++++++++++++++++-- include/linux/ftrace.h | 5 +++++ kernel/trace/ftrace.c | 35 ++++++++++++++++----------------- 4 files changed, 72 insertions(+), 21 deletions(-) (limited to 'kernel/trace/ftrace.c') diff --git a/arch/x86/kernel/entry_32.S b/arch/x86/kernel/entry_32.S index 7def9fd5c1e..958af86186c 100644 --- a/arch/x86/kernel/entry_32.S +++ b/arch/x86/kernel/entry_32.S @@ -1174,6 +1174,11 @@ ftrace_call: popl %edx popl %ecx popl %eax +#ifdef CONFIG_FUNCTION_GRAPH_TRACER +.globl ftrace_graph_call +ftrace_graph_call: + jmp ftrace_stub +#endif .globl ftrace_stub ftrace_stub: diff --git a/arch/x86/kernel/ftrace.c b/arch/x86/kernel/ftrace.c index 26b2d92d48b..7ef914e6a2f 100644 --- a/arch/x86/kernel/ftrace.c +++ b/arch/x86/kernel/ftrace.c @@ -111,7 +111,6 @@ static void ftrace_mod_code(void) */ mod_code_status = probe_kernel_write(mod_code_ip, mod_code_newcode, MCOUNT_INSN_SIZE); - } void ftrace_nmi_enter(void) @@ -325,7 +324,51 @@ int __init ftrace_dyn_arch_init(void *data) #ifdef CONFIG_FUNCTION_GRAPH_TRACER -#ifndef CONFIG_DYNAMIC_FTRACE +#ifdef CONFIG_DYNAMIC_FTRACE +extern void ftrace_graph_call(void); + +static int ftrace_mod_jmp(unsigned long ip, + int old_offset, int new_offset) +{ + unsigned char code[MCOUNT_INSN_SIZE]; + + if (probe_kernel_read(code, (void *)ip, MCOUNT_INSN_SIZE)) + return -EFAULT; + + if (code[0] != 0xe9 || old_offset != *(int *)(&code[1])) + return -EINVAL; + + *(int *)(&code[1]) = new_offset; + + if (do_ftrace_mod_code(ip, &code)) + return -EPERM; + + return 0; +} + +int ftrace_enable_ftrace_graph_caller(void) +{ + unsigned long ip = (unsigned long)(&ftrace_graph_call); + int old_offset, new_offset; + + old_offset = (unsigned long)(&ftrace_stub) - (ip + MCOUNT_INSN_SIZE); + new_offset = (unsigned long)(&ftrace_graph_caller) - (ip + MCOUNT_INSN_SIZE); + + return ftrace_mod_jmp(ip, old_offset, new_offset); +} + +int ftrace_disable_ftrace_graph_caller(void) +{ + unsigned long ip = (unsigned long)(&ftrace_graph_call); + int old_offset, new_offset; + + old_offset = (unsigned long)(&ftrace_graph_caller) - (ip + MCOUNT_INSN_SIZE); + new_offset = (unsigned long)(&ftrace_stub) - (ip + MCOUNT_INSN_SIZE); + + return ftrace_mod_jmp(ip, old_offset, new_offset); +} + +#else /* CONFIG_DYNAMIC_FTRACE */ /* * These functions are picked from those used on @@ -343,6 +386,7 @@ void ftrace_nmi_exit(void) { atomic_dec(&in_nmi); } + #endif /* !CONFIG_DYNAMIC_FTRACE */ /* Add a function return address to the trace stack on thread info.*/ diff --git a/include/linux/ftrace.h b/include/linux/ftrace.h index fc2d5498719..f9792c0d73f 100644 --- a/include/linux/ftrace.h +++ b/include/linux/ftrace.h @@ -117,6 +117,11 @@ extern void ftrace_call(void); extern void mcount_call(void); #ifdef CONFIG_FUNCTION_GRAPH_TRACER extern void ftrace_graph_caller(void); +extern int ftrace_enable_ftrace_graph_caller(void); +extern int ftrace_disable_ftrace_graph_caller(void); +#else +static inline int ftrace_enable_ftrace_graph_caller(void) { return 0; } +static inline int ftrace_disable_ftrace_graph_caller(void) { return 0; } #endif /** diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index 00d98c65fad..5f7c8642d58 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -281,6 +281,8 @@ enum { FTRACE_UPDATE_TRACE_FUNC = (1 << 2), FTRACE_ENABLE_MCOUNT = (1 << 3), FTRACE_DISABLE_MCOUNT = (1 << 4), + FTRACE_START_FUNC_RET = (1 << 5), + FTRACE_STOP_FUNC_RET = (1 << 6), }; static int ftrace_filtered; @@ -465,14 +467,7 @@ __ftrace_replace_code(struct dyn_ftrace *rec, int enable) unsigned long ip, fl; unsigned long ftrace_addr; -#ifdef CONFIG_FUNCTION_GRAPH_TRACER - if (ftrace_tracing_type == FTRACE_TYPE_ENTER) - ftrace_addr = (unsigned long)ftrace_caller; - else - ftrace_addr = (unsigned long)ftrace_graph_caller; -#else ftrace_addr = (unsigned long)ftrace_caller; -#endif ip = rec->ip; @@ -605,6 +600,11 @@ static int __ftrace_modify_code(void *data) if (*command & FTRACE_UPDATE_TRACE_FUNC) ftrace_update_ftrace_func(ftrace_trace_function); + if (*command & FTRACE_START_FUNC_RET) + ftrace_enable_ftrace_graph_caller(); + else if (*command & FTRACE_STOP_FUNC_RET) + ftrace_disable_ftrace_graph_caller(); + return 0; } @@ -629,10 +629,8 @@ static void ftrace_startup_enable(int command) ftrace_run_update_code(command); } -static void ftrace_startup(void) +static void ftrace_startup(int command) { - int command = 0; - if (unlikely(ftrace_disabled)) return; @@ -645,10 +643,8 @@ static void ftrace_startup(void) mutex_unlock(&ftrace_start_lock); } -static void ftrace_shutdown(void) +static void ftrace_shutdown(int command) { - int command = 0; - if (unlikely(ftrace_disabled)) return; @@ -1453,8 +1449,9 @@ device_initcall(ftrace_nodyn_init); static inline int ftrace_init_dyn_debugfs(struct dentry *d_tracer) { return 0; } static inline void ftrace_startup_enable(int command) { } -# define ftrace_startup() do { } while (0) -# define ftrace_shutdown() do { } while (0) +/* Keep as macros so we do not need to define the commands */ +# define ftrace_startup(command) do { } while (0) +# define ftrace_shutdown(command) do { } while (0) # define ftrace_startup_sysctl() do { } while (0) # define ftrace_shutdown_sysctl() do { } while (0) #endif /* CONFIG_DYNAMIC_FTRACE */ @@ -1585,7 +1582,7 @@ int register_ftrace_function(struct ftrace_ops *ops) } ret = __register_ftrace_function(ops); - ftrace_startup(); + ftrace_startup(0); out: mutex_unlock(&ftrace_sysctl_lock); @@ -1604,7 +1601,7 @@ int unregister_ftrace_function(struct ftrace_ops *ops) mutex_lock(&ftrace_sysctl_lock); ret = __unregister_ftrace_function(ops); - ftrace_shutdown(); + ftrace_shutdown(0); mutex_unlock(&ftrace_sysctl_lock); return ret; @@ -1751,7 +1748,7 @@ int register_ftrace_graph(trace_func_graph_ret_t retfunc, ftrace_tracing_type = FTRACE_TYPE_RETURN; ftrace_graph_return = retfunc; ftrace_graph_entry = entryfunc; - ftrace_startup(); + ftrace_startup(FTRACE_START_FUNC_RET); out: mutex_unlock(&ftrace_sysctl_lock); @@ -1765,7 +1762,7 @@ void unregister_ftrace_graph(void) atomic_dec(&ftrace_graph_active); ftrace_graph_return = (trace_func_graph_ret_t)ftrace_stub; ftrace_graph_entry = (trace_func_graph_ent_t)ftrace_stub; - ftrace_shutdown(); + ftrace_shutdown(FTRACE_STOP_FUNC_RET); /* Restore normal tracing type */ ftrace_tracing_type = FTRACE_TYPE_ENTER; -- cgit v1.2.3-70-g09d2 From e53a6319cca69111c1643dc9f18f4465d7f1cbf0 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Wed, 26 Nov 2008 00:16:25 -0500 Subject: ftrace: let function tracing and function return run together Impact: feature This patch enables function tracing and function return to run together. I've tested this by enabling the stack tracer and return tracer, where both the function entry and function return are used together with dynamic ftrace. Signed-off-by: Steven Rostedt Signed-off-by: Ingo Molnar --- kernel/trace/ftrace.c | 22 ++-------------------- 1 file changed, 2 insertions(+), 20 deletions(-) (limited to 'kernel/trace/ftrace.c') diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index 5f7c8642d58..cbf8b09f63a 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -53,9 +53,6 @@ static int ftrace_pid_trace = -1; /* Quick disabling of function tracer. */ int function_trace_stop; -/* By default, current tracing type is normal tracing. */ -enum ftrace_tracing_type_t ftrace_tracing_type = FTRACE_TYPE_ENTER; - /* * ftrace_disabled is set when an anomaly is discovered. * ftrace_disabled is much stronger than ftrace_enabled. @@ -1576,15 +1573,9 @@ int register_ftrace_function(struct ftrace_ops *ops) mutex_lock(&ftrace_sysctl_lock); - if (ftrace_tracing_type == FTRACE_TYPE_RETURN) { - ret = -EBUSY; - goto out; - } - ret = __register_ftrace_function(ops); ftrace_startup(0); -out: mutex_unlock(&ftrace_sysctl_lock); return ret; } @@ -1731,23 +1722,16 @@ int register_ftrace_graph(trace_func_graph_ret_t retfunc, mutex_lock(&ftrace_sysctl_lock); - /* - * Don't launch return tracing if normal function - * tracing is already running. - */ - if (ftrace_trace_function != ftrace_stub) { - ret = -EBUSY; - goto out; - } atomic_inc(&ftrace_graph_active); ret = start_graph_tracing(); if (ret) { atomic_dec(&ftrace_graph_active); goto out; } - ftrace_tracing_type = FTRACE_TYPE_RETURN; + ftrace_graph_return = retfunc; ftrace_graph_entry = entryfunc; + ftrace_startup(FTRACE_START_FUNC_RET); out: @@ -1763,8 +1747,6 @@ void unregister_ftrace_graph(void) ftrace_graph_return = (trace_func_graph_ret_t)ftrace_stub; ftrace_graph_entry = (trace_func_graph_ent_t)ftrace_stub; ftrace_shutdown(FTRACE_STOP_FUNC_RET); - /* Restore normal tracing type */ - ftrace_tracing_type = FTRACE_TYPE_ENTER; mutex_unlock(&ftrace_sysctl_lock); } -- cgit v1.2.3-70-g09d2 From 50cdaf08a8ec1d7f43987705da7aff7cf949708f Mon Sep 17 00:00:00 2001 From: Liming Wang Date: Fri, 28 Nov 2008 12:13:21 +0800 Subject: ftrace: improve seq_operation of ftrace Impact: make ftrace position computing more sane First remove useless ->pos field. Then we needn't check seq_printf in .show like other place. Signed-off-by: Liming Wang Reviewed-by: Bruce Ashfield Signed-off-by: Ingo Molnar --- kernel/trace/ftrace.c | 27 ++++++++++----------------- 1 file changed, 10 insertions(+), 17 deletions(-) (limited to 'kernel/trace/ftrace.c') diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index cbf8b09f63a..08b536a2614 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -786,7 +786,6 @@ enum { #define FTRACE_BUFF_MAX (KSYM_SYMBOL_LEN+4) /* room for wildcards */ struct ftrace_iterator { - loff_t pos; struct ftrace_page *pg; unsigned idx; unsigned flags; @@ -811,6 +810,8 @@ t_next(struct seq_file *m, void *v, loff_t *pos) iter->pg = iter->pg->next; iter->idx = 0; goto retry; + } else { + iter->idx = -1; } } else { rec = &iter->pg->records[iter->idx++]; @@ -833,8 +834,6 @@ t_next(struct seq_file *m, void *v, loff_t *pos) } spin_unlock(&ftrace_lock); - iter->pos = *pos; - return rec; } @@ -842,13 +841,15 @@ static void *t_start(struct seq_file *m, loff_t *pos) { struct ftrace_iterator *iter = m->private; void *p = NULL; - loff_t l = -1; - if (*pos > iter->pos) - *pos = iter->pos; + if (*pos > 0) { + if (iter->idx < 0) + return p; + (*pos)--; + iter->idx--; + } - l = *pos; - p = t_next(m, p, &l); + p = t_next(m, p, pos); return p; } @@ -859,21 +860,15 @@ static void t_stop(struct seq_file *m, void *p) static int t_show(struct seq_file *m, void *v) { - struct ftrace_iterator *iter = m->private; struct dyn_ftrace *rec = v; char str[KSYM_SYMBOL_LEN]; - int ret = 0; if (!rec) return 0; kallsyms_lookup(rec->ip, NULL, NULL, NULL, str); - ret = seq_printf(m, "%s\n", str); - if (ret < 0) { - iter->pos--; - iter->idx--; - } + seq_printf(m, "%s\n", str); return 0; } @@ -899,7 +894,6 @@ ftrace_avail_open(struct inode *inode, struct file *file) return -ENOMEM; iter->pg = ftrace_pages_start; - iter->pos = 0; ret = seq_open(file, &show_ftrace_seq_ops); if (!ret) { @@ -986,7 +980,6 @@ ftrace_regex_open(struct inode *inode, struct file *file, int enable) if (file->f_mode & FMODE_READ) { iter->pg = ftrace_pages_start; - iter->pos = 0; iter->flags = enable ? FTRACE_ITER_FILTER : FTRACE_ITER_NOTRACE; -- cgit v1.2.3-70-g09d2 From 66eafebc1086014709dc38f52ddcb3d67d9b346c Mon Sep 17 00:00:00 2001 From: Liming Wang Date: Tue, 2 Dec 2008 10:33:08 +0800 Subject: function trace: fix a bug of single thread function trace Impact: fix "no output from tracer" bug caused by ftrace_update_pid_func() When disabling single thread function trace using "echo -1 > set_ftrace_pid", the normal function trace has to restore to original function, otherwise the normal function trace will not work well. Without this commit, something like below: $ ps |grep 850 850 root 2556 S -/bin/sh $ echo 850 > /debug/tracing/set_ftrace_pid $ echo function > /debug/tracing/current_tracer $ echo 1 > /debug/tracing/tracing_enabled $ sleep 1 $ echo 0 > /debug/tracing/tracing_enabled $ cat /debug/tracing/trace_pipe |wc -l 59704 $ echo -1 > /debug/tracing/set_ftrace_pid $ echo 1 > /debug/tracing/tracing_enabled $ sleep 1 $ echo 0 > /debug/tracing/tracing_enabled $ more /debug/tracing/trace_pipe <====== nothing output now! it should output trace record. Signed-off-by: Liming Wang Signed-off-by: Ingo Molnar --- kernel/trace/ftrace.c | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-) (limited to 'kernel/trace/ftrace.c') diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index 08b536a2614..6d89ab46c6e 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -243,10 +243,8 @@ static void ftrace_update_pid_func(void) set_ftrace_pid_function(func); func = ftrace_pid_func; } else { - if (func != ftrace_pid_func) - goto out; - - set_ftrace_pid_function(func); + if (func == ftrace_pid_func) + func = ftrace_pid_function; } #ifdef CONFIG_HAVE_FUNCTION_TRACE_MCOUNT_TEST -- cgit v1.2.3-70-g09d2 From 48d68b20d00865035b8b65e69af343d0f53fac9d Mon Sep 17 00:00:00 2001 From: Frederic Weisbecker Date: Tue, 2 Dec 2008 00:20:39 +0100 Subject: tracing/function-graph-tracer: support for x86-64 Impact: extend and enable the function graph tracer to 64-bit x86 This patch implements the support for function graph tracer under x86-64. Both static and dynamic tracing are supported. This causes some small CPP conditional asm on arch/x86/kernel/ftrace.c I wanted to use probe_kernel_read/write to make the return address saving/patching code more generic but it causes tracing recursion. That would be perhaps useful to implement a notrace version of these function for other archs ports. Note that arch/x86/process_64.c is not traced, as in X86-32. I first thought __switch_to() was responsible of crashes during tracing because I believed current task were changed inside but that's actually not the case (actually yes, but not the "current" pointer). So I will have to investigate to find the functions that harm here, to enable tracing of the other functions inside (but there is no issue at this time, while process_64.c stays out of -pg flags). A little possible race condition is fixed inside this patch too. When the tracer allocate a return stack dynamically, the current depth is not initialized before but after. An interrupt could occur at this time and, after seeing that the return stack is allocated, the tracer could try to trace it with a random uninitialized depth. It's a prevention, even if I hadn't problems with it. Signed-off-by: Frederic Weisbecker Cc: Steven Rostedt Cc: Tim Bird Signed-off-by: Ingo Molnar --- arch/x86/Kconfig | 2 +- arch/x86/kernel/Makefile | 1 + arch/x86/kernel/entry_64.S | 74 ++++++++++++++++++++++++++++++++++++++++++++++ arch/x86/kernel/ftrace.c | 11 ++++++- kernel/trace/ftrace.c | 4 ++- 5 files changed, 89 insertions(+), 3 deletions(-) (limited to 'kernel/trace/ftrace.c') diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig index 0842b112768..45c86fb9413 100644 --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -29,7 +29,7 @@ config X86 select HAVE_FTRACE_MCOUNT_RECORD select HAVE_DYNAMIC_FTRACE select HAVE_FUNCTION_TRACER - select HAVE_FUNCTION_GRAPH_TRACER if X86_32 + select HAVE_FUNCTION_GRAPH_TRACER select HAVE_FUNCTION_TRACE_MCOUNT_TEST select HAVE_KVM if ((X86_32 && !X86_VOYAGER && !X86_VISWS && !X86_NUMAQ) || X86_64) select HAVE_ARCH_KGDB if !X86_VOYAGER diff --git a/arch/x86/kernel/Makefile b/arch/x86/kernel/Makefile index 64939a0c398..d274425fb07 100644 --- a/arch/x86/kernel/Makefile +++ b/arch/x86/kernel/Makefile @@ -17,6 +17,7 @@ endif ifdef CONFIG_FUNCTION_GRAPH_TRACER # Don't trace __switch_to() but let it for function tracer CFLAGS_REMOVE_process_32.o = -pg +CFLAGS_REMOVE_process_64.o = -pg endif # diff --git a/arch/x86/kernel/entry_64.S b/arch/x86/kernel/entry_64.S index 08aa6b10933..2aa0526ac30 100644 --- a/arch/x86/kernel/entry_64.S +++ b/arch/x86/kernel/entry_64.S @@ -98,6 +98,12 @@ ftrace_call: movq (%rsp), %rax addq $0x38, %rsp +#ifdef CONFIG_FUNCTION_GRAPH_TRACER +.globl ftrace_graph_call +ftrace_graph_call: + jmp ftrace_stub +#endif + .globl ftrace_stub ftrace_stub: retq @@ -110,6 +116,12 @@ ENTRY(mcount) cmpq $ftrace_stub, ftrace_trace_function jnz trace + +#ifdef CONFIG_FUNCTION_GRAPH_TRACER + cmpq $ftrace_stub, ftrace_graph_return + jnz ftrace_graph_caller +#endif + .globl ftrace_stub ftrace_stub: retq @@ -145,6 +157,68 @@ END(mcount) #endif /* CONFIG_DYNAMIC_FTRACE */ #endif /* CONFIG_FUNCTION_TRACER */ +#ifdef CONFIG_FUNCTION_GRAPH_TRACER +ENTRY(ftrace_graph_caller) + cmpl $0, function_trace_stop + jne ftrace_stub + + subq $0x38, %rsp + movq %rax, (%rsp) + movq %rcx, 8(%rsp) + movq %rdx, 16(%rsp) + movq %rsi, 24(%rsp) + movq %rdi, 32(%rsp) + movq %r8, 40(%rsp) + movq %r9, 48(%rsp) + + leaq 8(%rbp), %rdi + movq 0x38(%rsp), %rsi + + call prepare_ftrace_return + + movq 48(%rsp), %r9 + movq 40(%rsp), %r8 + movq 32(%rsp), %rdi + movq 24(%rsp), %rsi + movq 16(%rsp), %rdx + movq 8(%rsp), %rcx + movq (%rsp), %rax + addq $0x38, %rsp + retq +END(ftrace_graph_caller) + + +.globl return_to_handler +return_to_handler: + subq $80, %rsp + + movq %rax, (%rsp) + movq %rcx, 8(%rsp) + movq %rdx, 16(%rsp) + movq %rsi, 24(%rsp) + movq %rdi, 32(%rsp) + movq %r8, 40(%rsp) + movq %r9, 48(%rsp) + movq %r10, 56(%rsp) + movq %r11, 64(%rsp) + + call ftrace_return_to_handler + + movq %rax, 72(%rsp) + movq 64(%rsp), %r11 + movq 56(%rsp), %r10 + movq 48(%rsp), %r9 + movq 40(%rsp), %r8 + movq 32(%rsp), %rdi + movq 24(%rsp), %rsi + movq 16(%rsp), %rdx + movq 8(%rsp), %rcx + movq (%rsp), %rax + addq $72, %rsp + retq +#endif + + #ifndef CONFIG_PREEMPT #define retint_kernel retint_restore_args #endif diff --git a/arch/x86/kernel/ftrace.c b/arch/x86/kernel/ftrace.c index 7ef914e6a2f..58832478b94 100644 --- a/arch/x86/kernel/ftrace.c +++ b/arch/x86/kernel/ftrace.c @@ -467,8 +467,13 @@ void prepare_ftrace_return(unsigned long *parent, unsigned long self_addr) * ignore such a protection. */ asm volatile( +#ifdef CONFIG_X86_64 + "1: movq (%[parent_old]), %[old]\n" + "2: movq %[return_hooker], (%[parent_replaced])\n" +#else "1: movl (%[parent_old]), %[old]\n" "2: movl %[return_hooker], (%[parent_replaced])\n" +#endif " movl $0, %[faulted]\n" ".section .fixup, \"ax\"\n" @@ -476,8 +481,13 @@ void prepare_ftrace_return(unsigned long *parent, unsigned long self_addr) ".previous\n" ".section __ex_table, \"a\"\n" +#ifdef CONFIG_X86_64 + " .quad 1b, 3b\n" + " .quad 2b, 3b\n" +#else " .long 1b, 3b\n" " .long 2b, 3b\n" +#endif ".previous\n" : [parent_replaced] "=r" (parent), [old] "=r" (old), @@ -509,5 +519,4 @@ void prepare_ftrace_return(unsigned long *parent, unsigned long self_addr) ftrace_graph_entry(&trace); } - #endif /* CONFIG_FUNCTION_GRAPH_TRACER */ diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index 08b536a2614..f7249962752 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -1673,8 +1673,10 @@ static int alloc_retstack_tasklist(struct ftrace_ret_stack **ret_stack_list) } if (t->ret_stack == NULL) { - t->ret_stack = ret_stack_list[start++]; t->curr_ret_stack = -1; + /* Make sure IRQs see the -1 first: */ + barrier(); + t->ret_stack = ret_stack_list[start++]; atomic_set(&t->trace_overrun, 0); } } while_each_thread(g, t); -- cgit v1.2.3-70-g09d2 From 14a866c567e040ccf6240d68b083dd1dbbde63e6 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Tue, 2 Dec 2008 23:50:02 -0500 Subject: ftrace: add ftrace_graph_stop() Impact: new ftrace_graph_stop function While developing more features of function graph, I hit a bug that caused the WARN_ON to trigger in the prepare_ftrace_return function. Well, it was hard for me to find out that was happening because the bug would not print, it would just cause a hard lockup or reboot. The reason is that it is not safe to call printk from this function. Looking further, I also found that it calls unregister_ftrace_graph, which grabs a mutex and calls kstop machine. This would definitely lock the box up if it were to trigger. This patch adds a fast and safe ftrace_graph_stop() which will stop the function tracer. Then it is safe to call the WARN ON. Signed-off-by: Steven Rostedt Signed-off-by: Ingo Molnar --- arch/x86/kernel/ftrace.c | 10 ++++++---- include/linux/ftrace.h | 2 ++ kernel/trace/ftrace.c | 5 +++++ 3 files changed, 13 insertions(+), 4 deletions(-) (limited to 'kernel/trace/ftrace.c') diff --git a/arch/x86/kernel/ftrace.c b/arch/x86/kernel/ftrace.c index 1a5b8f8cb3c..adba8e9a427 100644 --- a/arch/x86/kernel/ftrace.c +++ b/arch/x86/kernel/ftrace.c @@ -484,14 +484,16 @@ void prepare_ftrace_return(unsigned long *parent, unsigned long self_addr) : "memory" ); - if (WARN_ON(faulted)) { - unregister_ftrace_graph(); + if (unlikely(faulted)) { + ftrace_graph_stop(); + WARN_ON(1); return; } - if (WARN_ON(!__kernel_text_address(old))) { - unregister_ftrace_graph(); + if (unlikely(!__kernel_text_address(old))) { + ftrace_graph_stop(); *parent = old; + WARN_ON(1); return; } diff --git a/include/linux/ftrace.h b/include/linux/ftrace.h index afba918c623..58ca1c3a3f4 100644 --- a/include/linux/ftrace.h +++ b/include/linux/ftrace.h @@ -376,6 +376,8 @@ typedef void (*trace_func_graph_ent_t)(struct ftrace_graph_ent *); /* entry */ extern int register_ftrace_graph(trace_func_graph_ret_t retfunc, trace_func_graph_ent_t entryfunc); +extern void ftrace_graph_stop(void); + /* The current handlers in use */ extern trace_func_graph_ret_t ftrace_graph_return; extern trace_func_graph_ent_t ftrace_graph_entry; diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index 2e78628443e..a44af05ae2d 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -1769,5 +1769,10 @@ void ftrace_graph_exit_task(struct task_struct *t) kfree(ret_stack); } + +void ftrace_graph_stop(void) +{ + ftrace_stop(); +} #endif -- cgit v1.2.3-70-g09d2 From e49dc19c6a19ea112fcb94b7c62ec62cdd5c08aa Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Tue, 2 Dec 2008 23:50:05 -0500 Subject: ftrace: function graph return for function entry Impact: feature, let entry function decide to trace or not This patch lets the graph tracer entry function decide if the tracing should be done at the end as well. This requires all function graph entry functions return 1 if it should trace, or 0 if the return should not be traced. Signed-off-by: Steven Rostedt Signed-off-by: Ingo Molnar --- arch/x86/kernel/entry_32.S | 3 +++ arch/x86/kernel/entry_64.S | 3 +++ arch/x86/kernel/ftrace.c | 7 ++++++- include/linux/ftrace.h | 2 +- kernel/trace/ftrace.c | 10 +++++++--- kernel/trace/trace.c | 4 +++- kernel/trace/trace.h | 2 +- 7 files changed, 24 insertions(+), 7 deletions(-) (limited to 'kernel/trace/ftrace.c') diff --git a/arch/x86/kernel/entry_32.S b/arch/x86/kernel/entry_32.S index 826682abed1..43ceb3f454b 100644 --- a/arch/x86/kernel/entry_32.S +++ b/arch/x86/kernel/entry_32.S @@ -1196,6 +1196,9 @@ ENTRY(mcount) #ifdef CONFIG_FUNCTION_GRAPH_TRACER cmpl $ftrace_stub, ftrace_graph_return jnz ftrace_graph_caller + + cmpl $ftrace_graph_entry_stub, ftrace_graph_entry + jnz ftrace_graph_caller #endif .globl ftrace_stub ftrace_stub: diff --git a/arch/x86/kernel/entry_64.S b/arch/x86/kernel/entry_64.S index 9060ba6497e..54e0bbdccb9 100644 --- a/arch/x86/kernel/entry_64.S +++ b/arch/x86/kernel/entry_64.S @@ -120,6 +120,9 @@ ENTRY(mcount) #ifdef CONFIG_FUNCTION_GRAPH_TRACER cmpq $ftrace_stub, ftrace_graph_return jnz ftrace_graph_caller + + cmpq $ftrace_graph_entry_stub, ftrace_graph_entry + jnz ftrace_graph_caller #endif .globl ftrace_stub diff --git a/arch/x86/kernel/ftrace.c b/arch/x86/kernel/ftrace.c index adba8e9a427..d278ad2ebda 100644 --- a/arch/x86/kernel/ftrace.c +++ b/arch/x86/kernel/ftrace.c @@ -425,6 +425,7 @@ static void pop_return_trace(struct ftrace_graph_ret *trace, unsigned long *ret) trace->calltime = current->ret_stack[index].calltime; trace->overrun = atomic_read(¤t->trace_overrun); trace->depth = index; + barrier(); current->curr_ret_stack--; } @@ -506,7 +507,11 @@ void prepare_ftrace_return(unsigned long *parent, unsigned long self_addr) } trace.func = self_addr; - ftrace_graph_entry(&trace); + /* Only trace if the calling function expects to */ + if (!ftrace_graph_entry(&trace)) { + current->curr_ret_stack--; + *parent = old; + } } #endif /* CONFIG_FUNCTION_GRAPH_TRACER */ diff --git a/include/linux/ftrace.h b/include/linux/ftrace.h index 58ca1c3a3f4..469ceb3e85b 100644 --- a/include/linux/ftrace.h +++ b/include/linux/ftrace.h @@ -371,7 +371,7 @@ struct ftrace_graph_ret { #define FTRACE_RETSTACK_ALLOC_SIZE 32 /* Type of the callback handlers for tracing function graph*/ typedef void (*trace_func_graph_ret_t)(struct ftrace_graph_ret *); /* return */ -typedef void (*trace_func_graph_ent_t)(struct ftrace_graph_ent *); /* entry */ +typedef int (*trace_func_graph_ent_t)(struct ftrace_graph_ent *); /* entry */ extern int register_ftrace_graph(trace_func_graph_ret_t retfunc, trace_func_graph_ent_t entryfunc); diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index a44af05ae2d..65b9e863056 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -1636,11 +1636,15 @@ ftrace_enable_sysctl(struct ctl_table *table, int write, static atomic_t ftrace_graph_active; +int ftrace_graph_entry_stub(struct ftrace_graph_ent *trace) +{ + return 0; +} + /* The callbacks that hook a function */ trace_func_graph_ret_t ftrace_graph_return = (trace_func_graph_ret_t)ftrace_stub; -trace_func_graph_ent_t ftrace_graph_entry = - (trace_func_graph_ent_t)ftrace_stub; +trace_func_graph_ent_t ftrace_graph_entry = ftrace_graph_entry_stub; /* Try to assign a return stack array on FTRACE_RETSTACK_ALLOC_SIZE tasks. */ static int alloc_retstack_tasklist(struct ftrace_ret_stack **ret_stack_list) @@ -1738,7 +1742,7 @@ void unregister_ftrace_graph(void) atomic_dec(&ftrace_graph_active); ftrace_graph_return = (trace_func_graph_ret_t)ftrace_stub; - ftrace_graph_entry = (trace_func_graph_ent_t)ftrace_stub; + ftrace_graph_entry = ftrace_graph_entry_stub; ftrace_shutdown(FTRACE_STOP_FUNC_RET); mutex_unlock(&ftrace_sysctl_lock); diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index 380de630ebc..8b6409a62b5 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -1200,7 +1200,7 @@ function_trace_call(unsigned long ip, unsigned long parent_ip) } #ifdef CONFIG_FUNCTION_GRAPH_TRACER -void trace_graph_entry(struct ftrace_graph_ent *trace) +int trace_graph_entry(struct ftrace_graph_ent *trace) { struct trace_array *tr = &global_trace; struct trace_array_cpu *data; @@ -1219,6 +1219,8 @@ void trace_graph_entry(struct ftrace_graph_ent *trace) } atomic_dec(&data->disabled); local_irq_restore(flags); + + return 1; } void trace_graph_return(struct ftrace_graph_ret *trace) diff --git a/kernel/trace/trace.h b/kernel/trace/trace.h index f96f4e787ff..0565ae9a221 100644 --- a/kernel/trace/trace.h +++ b/kernel/trace/trace.h @@ -412,7 +412,7 @@ void trace_function(struct trace_array *tr, unsigned long flags, int pc); void trace_graph_return(struct ftrace_graph_ret *trace); -void trace_graph_entry(struct ftrace_graph_ent *trace); +int trace_graph_entry(struct ftrace_graph_ent *trace); void trace_bts(struct trace_array *tr, unsigned long from, unsigned long to); -- cgit v1.2.3-70-g09d2 From ea4e2bc4d9f7370e57a343ccb5e7c0ad3222ec3c Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Wed, 3 Dec 2008 15:36:57 -0500 Subject: ftrace: graph of a single function This patch adds the file: /debugfs/tracing/set_graph_function which can be used along with the function graph tracer. When this file is empty, the function graph tracer will act as usual. When the file has a function in it, the function graph tracer will only trace that function. For example: # echo blk_unplug > /debugfs/tracing/set_graph_function # cat /debugfs/tracing/trace [...] ------------------------------------------ | 2) make-19003 => kjournald-2219 ------------------------------------------ 2) | blk_unplug() { 2) | dm_unplug_all() { 2) | dm_get_table() { 2) 1.381 us | _read_lock(); 2) 0.911 us | dm_table_get(); 2) 1. 76 us | _read_unlock(); 2) + 12.912 us | } 2) | dm_table_unplug_all() { 2) | blk_unplug() { 2) 0.778 us | generic_unplug_device(); 2) 2.409 us | } 2) 5.992 us | } 2) 0.813 us | dm_table_put(); 2) + 29. 90 us | } 2) + 34.532 us | } You can add up to 32 functions into this file. Currently we limit it to 32, but this may change with later improvements. To add another function, use the append '>>': # echo sys_read >> /debugfs/tracing/set_graph_function # cat /debugfs/tracing/set_graph_function blk_unplug sys_read Using the '>' will clear out the function and write anew: # echo sys_write > /debug/tracing/set_graph_function # cat /debug/tracing/set_graph_function sys_write Note, if you have function graph running while doing this, the small time between clearing it and updating it will cause the graph to record all functions. This should not be an issue because after it sets the filter, only those functions will be recorded from then on. If you need to only record a particular function then set this file first before starting the function graph tracer. In the future this side effect may be corrected. The set_graph_function file is similar to the set_ftrace_filter but it does not take wild cards nor does it allow for more than one function to be set with a single write. There is no technical reason why this is the case, I just do not have the time yet to implement that. Note, dynamic ftrace must be enabled for this to appear because it uses the dynamic ftrace records to match the name to the mcount call sites. Signed-off-by: Steven Rostedt Signed-off-by: Ingo Molnar --- include/linux/ftrace.h | 46 ++++++++++ include/linux/sched.h | 4 + kernel/trace/ftrace.c | 227 +++++++++++++++++++++++++++++++++++++++++++++++++ kernel/trace/trace.c | 8 ++ kernel/trace/trace.h | 30 ++++++- 5 files changed, 314 insertions(+), 1 deletion(-) (limited to 'kernel/trace/ftrace.c') diff --git a/include/linux/ftrace.h b/include/linux/ftrace.h index 469ceb3e85b..b295d3106bf 100644 --- a/include/linux/ftrace.h +++ b/include/linux/ftrace.h @@ -7,6 +7,7 @@ #include #include #include +#include #ifdef CONFIG_FUNCTION_TRACER @@ -391,4 +392,49 @@ static inline void ftrace_graph_init_task(struct task_struct *t) { } static inline void ftrace_graph_exit_task(struct task_struct *t) { } #endif +#ifdef CONFIG_TRACING +#include + +/* flags for current->trace */ +enum { + TSK_TRACE_FL_TRACE_BIT = 0, + TSK_TRACE_FL_GRAPH_BIT = 1, +}; +enum { + TSK_TRACE_FL_TRACE = 1 << TSK_TRACE_FL_TRACE_BIT, + TSK_TRACE_FL_GRAPH = 1 << TSK_TRACE_FL_GRAPH_BIT, +}; + +static inline void set_tsk_trace_trace(struct task_struct *tsk) +{ + set_bit(TSK_TRACE_FL_TRACE_BIT, &tsk->trace); +} + +static inline void clear_tsk_trace_trace(struct task_struct *tsk) +{ + clear_bit(TSK_TRACE_FL_TRACE_BIT, &tsk->trace); +} + +static inline int test_tsk_trace_trace(struct task_struct *tsk) +{ + return tsk->trace & TSK_TRACE_FL_TRACE; +} + +static inline void set_tsk_trace_graph(struct task_struct *tsk) +{ + set_bit(TSK_TRACE_FL_GRAPH_BIT, &tsk->trace); +} + +static inline void clear_tsk_trace_graph(struct task_struct *tsk) +{ + clear_bit(TSK_TRACE_FL_GRAPH_BIT, &tsk->trace); +} + +static inline int test_tsk_trace_graph(struct task_struct *tsk) +{ + return tsk->trace & TSK_TRACE_FL_GRAPH; +} + +#endif /* CONFIG_TRACING */ + #endif /* _LINUX_FTRACE_H */ diff --git a/include/linux/sched.h b/include/linux/sched.h index 2d0a93c3122..4c152e0acc9 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -1380,6 +1380,10 @@ struct task_struct { */ atomic_t trace_overrun; #endif +#ifdef CONFIG_TRACING + /* state flags for use by tracers */ + unsigned long trace; +#endif }; /* diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index 65b9e863056..b17a30350f0 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -1320,6 +1320,224 @@ static struct file_operations ftrace_notrace_fops = { .release = ftrace_notrace_release, }; +#ifdef CONFIG_FUNCTION_GRAPH_TRACER + +static DEFINE_MUTEX(graph_lock); + +int ftrace_graph_count; +unsigned long ftrace_graph_funcs[FTRACE_GRAPH_MAX_FUNCS] __read_mostly; + +static void * +g_next(struct seq_file *m, void *v, loff_t *pos) +{ + unsigned long *array = m->private; + int index = *pos; + + (*pos)++; + + if (index >= ftrace_graph_count) + return NULL; + + return &array[index]; +} + +static void *g_start(struct seq_file *m, loff_t *pos) +{ + void *p = NULL; + + mutex_lock(&graph_lock); + + p = g_next(m, p, pos); + + return p; +} + +static void g_stop(struct seq_file *m, void *p) +{ + mutex_unlock(&graph_lock); +} + +static int g_show(struct seq_file *m, void *v) +{ + unsigned long *ptr = v; + char str[KSYM_SYMBOL_LEN]; + + if (!ptr) + return 0; + + kallsyms_lookup(*ptr, NULL, NULL, NULL, str); + + seq_printf(m, "%s\n", str); + + return 0; +} + +static struct seq_operations ftrace_graph_seq_ops = { + .start = g_start, + .next = g_next, + .stop = g_stop, + .show = g_show, +}; + +static int +ftrace_graph_open(struct inode *inode, struct file *file) +{ + int ret = 0; + + if (unlikely(ftrace_disabled)) + return -ENODEV; + + mutex_lock(&graph_lock); + if ((file->f_mode & FMODE_WRITE) && + !(file->f_flags & O_APPEND)) { + ftrace_graph_count = 0; + memset(ftrace_graph_funcs, 0, sizeof(ftrace_graph_funcs)); + } + + if (file->f_mode & FMODE_READ) { + ret = seq_open(file, &ftrace_graph_seq_ops); + if (!ret) { + struct seq_file *m = file->private_data; + m->private = ftrace_graph_funcs; + } + } else + file->private_data = ftrace_graph_funcs; + mutex_unlock(&graph_lock); + + return ret; +} + +static ssize_t +ftrace_graph_read(struct file *file, char __user *ubuf, + size_t cnt, loff_t *ppos) +{ + if (file->f_mode & FMODE_READ) + return seq_read(file, ubuf, cnt, ppos); + else + return -EPERM; +} + +static int +ftrace_set_func(unsigned long *array, int idx, char *buffer) +{ + char str[KSYM_SYMBOL_LEN]; + struct dyn_ftrace *rec; + struct ftrace_page *pg; + int found = 0; + int i; + + if (ftrace_disabled) + return -ENODEV; + + /* should not be called from interrupt context */ + spin_lock(&ftrace_lock); + + for (pg = ftrace_pages_start; pg; pg = pg->next) { + for (i = 0; i < pg->index; i++) { + rec = &pg->records[i]; + + if (rec->flags & (FTRACE_FL_FAILED | FTRACE_FL_FREE)) + continue; + + kallsyms_lookup(rec->ip, NULL, NULL, NULL, str); + if (strcmp(str, buffer) == 0) { + found = 1; + array[idx] = rec->ip; + break; + } + } + } + spin_unlock(&ftrace_lock); + + return found ? 0 : -EINVAL; +} + +static ssize_t +ftrace_graph_write(struct file *file, const char __user *ubuf, + size_t cnt, loff_t *ppos) +{ + unsigned char buffer[FTRACE_BUFF_MAX+1]; + unsigned long *array; + size_t read = 0; + ssize_t ret; + int index = 0; + char ch; + + if (!cnt || cnt < 0) + return 0; + + mutex_lock(&graph_lock); + + if (ftrace_graph_count >= FTRACE_GRAPH_MAX_FUNCS) { + ret = -EBUSY; + goto out; + } + + if (file->f_mode & FMODE_READ) { + struct seq_file *m = file->private_data; + array = m->private; + } else + array = file->private_data; + + ret = get_user(ch, ubuf++); + if (ret) + goto out; + read++; + cnt--; + + /* skip white space */ + while (cnt && isspace(ch)) { + ret = get_user(ch, ubuf++); + if (ret) + goto out; + read++; + cnt--; + } + + if (isspace(ch)) { + *ppos += read; + ret = read; + goto out; + } + + while (cnt && !isspace(ch)) { + if (index < FTRACE_BUFF_MAX) + buffer[index++] = ch; + else { + ret = -EINVAL; + goto out; + } + ret = get_user(ch, ubuf++); + if (ret) + goto out; + read++; + cnt--; + } + buffer[index] = 0; + + /* we allow only one at a time */ + ret = ftrace_set_func(array, ftrace_graph_count, buffer); + if (ret) + goto out; + + ftrace_graph_count++; + + file->f_pos += read; + + ret = read; + out: + mutex_unlock(&graph_lock); + + return ret; +} + +static const struct file_operations ftrace_graph_fops = { + .open = ftrace_graph_open, + .read = ftrace_graph_read, + .write = ftrace_graph_write, +}; +#endif /* CONFIG_FUNCTION_GRAPH_TRACER */ + static __init int ftrace_init_dyn_debugfs(struct dentry *d_tracer) { struct dentry *entry; @@ -1347,6 +1565,15 @@ static __init int ftrace_init_dyn_debugfs(struct dentry *d_tracer) pr_warning("Could not create debugfs " "'set_ftrace_notrace' entry\n"); +#ifdef CONFIG_FUNCTION_GRAPH_TRACER + entry = debugfs_create_file("set_graph_function", 0444, d_tracer, + NULL, + &ftrace_graph_fops); + if (!entry) + pr_warning("Could not create debugfs " + "'set_graph_function' entry\n"); +#endif /* CONFIG_FUNCTION_GRAPH_TRACER */ + return 0; } diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index 8b6409a62b5..710b39acd81 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -1209,6 +1209,9 @@ int trace_graph_entry(struct ftrace_graph_ent *trace) int cpu; int pc; + if (!ftrace_graph_addr(trace->func)) + return 0; + local_irq_save(flags); cpu = raw_smp_processor_id(); data = tr->data[cpu]; @@ -1217,6 +1220,9 @@ int trace_graph_entry(struct ftrace_graph_ent *trace) pc = preempt_count(); __trace_graph_entry(tr, data, trace, flags, pc); } + /* Only do the atomic if it is not already set */ + if (!test_tsk_trace_graph(current)) + set_tsk_trace_graph(current); atomic_dec(&data->disabled); local_irq_restore(flags); @@ -1240,6 +1246,8 @@ void trace_graph_return(struct ftrace_graph_ret *trace) pc = preempt_count(); __trace_graph_return(tr, data, trace, flags, pc); } + if (!trace->depth) + clear_tsk_trace_graph(current); atomic_dec(&data->disabled); local_irq_restore(flags); } diff --git a/kernel/trace/trace.h b/kernel/trace/trace.h index 0565ae9a221..41f026bfc9e 100644 --- a/kernel/trace/trace.h +++ b/kernel/trace/trace.h @@ -505,13 +505,41 @@ extern unsigned long trace_flags; /* Standard output formatting function used for function return traces */ #ifdef CONFIG_FUNCTION_GRAPH_TRACER extern enum print_line_t print_graph_function(struct trace_iterator *iter); + +#ifdef CONFIG_DYNAMIC_FTRACE +/* TODO: make this variable */ +#define FTRACE_GRAPH_MAX_FUNCS 32 +extern int ftrace_graph_count; +extern unsigned long ftrace_graph_funcs[FTRACE_GRAPH_MAX_FUNCS]; + +static inline int ftrace_graph_addr(unsigned long addr) +{ + int i; + + if (!ftrace_graph_count || test_tsk_trace_graph(current)) + return 1; + + for (i = 0; i < ftrace_graph_count; i++) { + if (addr == ftrace_graph_funcs[i]) + return 1; + } + + return 0; +} #else +static inline int ftrace_trace_addr(unsigned long addr) +{ + return 1 +} +#endif /* CONFIG_DYNAMIC_FTRACE */ + +#else /* CONFIG_FUNCTION_GRAPH_TRACER */ static inline enum print_line_t print_graph_function(struct trace_iterator *iter) { return TRACE_TYPE_UNHANDLED; } -#endif +#endif /* CONFIG_FUNCTION_GRAPH_TRACER */ /* * trace_iterator_flags is an enumeration that defines bit -- cgit v1.2.3-70-g09d2 From 0ef8cde56ab92ab3f65221246dc1622c6b5068b3 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Wed, 3 Dec 2008 15:36:58 -0500 Subject: ftrace: use task struct trace flag to filter on pid Impact: clean up Use the new task struct trace flags to determine if a process should be traced or not. Note: this moves the searching of the pid to the slow path of setting the pid field. This needs to be converted to the pid name space. Signed-off-by: Steven Rostedt Signed-off-by: Ingo Molnar --- kernel/trace/ftrace.c | 28 +++++++++++++++++++++++++--- 1 file changed, 25 insertions(+), 3 deletions(-) (limited to 'kernel/trace/ftrace.c') diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index b17a30350f0..c5049f54a27 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -47,7 +47,7 @@ int ftrace_enabled __read_mostly; static int last_ftrace_enabled; -/* ftrace_pid_trace >= 0 will only trace threads with this pid */ +/* set when tracing only a pid */ static int ftrace_pid_trace = -1; /* Quick disabling of function tracer. */ @@ -90,7 +90,7 @@ static void ftrace_list_func(unsigned long ip, unsigned long parent_ip) static void ftrace_pid_func(unsigned long ip, unsigned long parent_ip) { - if (current->pid != ftrace_pid_trace) + if (!test_tsk_trace_trace(current)) return; ftrace_pid_function(ip, parent_ip); @@ -1714,11 +1714,33 @@ ftrace_pid_write(struct file *filp, const char __user *ubuf, ftrace_pid_trace = -1; } else { + struct task_struct *p; + int found = 0; if (ftrace_pid_trace == val) goto out; - ftrace_pid_trace = val; + /* + * Find the task that matches this pid. + * TODO: use pid namespaces instead. + */ + rcu_read_lock(); + for_each_process(p) { + if (p->pid == val) { + found = 1; + set_tsk_trace_trace(p); + } else if (test_tsk_trace_trace(p)) + clear_tsk_trace_trace(p); + } + rcu_read_unlock(); + + if (found) + ftrace_pid_trace = val; + else { + if (ftrace_pid_trace < 0) + goto out; + ftrace_pid_trace = -1; + } } /* update the function call */ -- cgit v1.2.3-70-g09d2 From 804a685162a7080386714166776f57255a75238e Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Wed, 3 Dec 2008 15:36:59 -0500 Subject: ftrace: trace single pid for function graph tracer Impact: New feature This patch makes the changes to set_ftrace_pid apply to the function graph tracer. # echo $$ > /debugfs/tracing/set_ftrace_pid # echo function_graph > /debugfs/tracing/current_tracer Will cause only the current task to be traced. Note, the trace flags are also inherited by child processes, so the children of the shell will also be traced. Signed-off-by: Steven Rostedt Signed-off-by: Ingo Molnar --- kernel/trace/ftrace.c | 2 +- kernel/trace/trace.c | 3 +++ kernel/trace/trace.h | 10 ++++++++++ 3 files changed, 14 insertions(+), 1 deletion(-) (limited to 'kernel/trace/ftrace.c') diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index c5049f54a27..57592a9dd63 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -48,7 +48,7 @@ int ftrace_enabled __read_mostly; static int last_ftrace_enabled; /* set when tracing only a pid */ -static int ftrace_pid_trace = -1; +int ftrace_pid_trace = -1; /* Quick disabling of function tracer. */ int function_trace_stop; diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index 710b39acd81..1bd9574404e 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -1209,6 +1209,9 @@ int trace_graph_entry(struct ftrace_graph_ent *trace) int cpu; int pc; + if (!ftrace_trace_task(current)) + return 0; + if (!ftrace_graph_addr(trace->func)) return 0; diff --git a/kernel/trace/trace.h b/kernel/trace/trace.h index 41f026bfc9e..95fff37ed97 100644 --- a/kernel/trace/trace.h +++ b/kernel/trace/trace.h @@ -541,6 +541,16 @@ print_graph_function(struct trace_iterator *iter) } #endif /* CONFIG_FUNCTION_GRAPH_TRACER */ +extern int ftrace_pid_trace; + +static inline int ftrace_trace_task(struct task_struct *task) +{ + if (ftrace_pid_trace < 0) + return 1; + + return test_tsk_trace_trace(task); +} + /* * trace_iterator_flags is an enumeration that defines bit * positions into trace_flags that controls the output. -- cgit v1.2.3-70-g09d2 From 978f3a45d9499c7a447ca7615455cefb63d44165 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Thu, 4 Dec 2008 00:26:40 -0500 Subject: ftrace: use struct pid Impact: clean up, extend PID filtering to PID namespaces Eric Biederman suggested using the struct pid for filtering on pids in the kernel. This patch is based off of a demonstration of an implementation that Eric sent me in an email. Signed-off-by: Steven Rostedt Signed-off-by: Ingo Molnar --- kernel/trace/ftrace.c | 76 +++++++++++++++++++++++++++++---------------------- kernel/trace/trace.h | 4 +-- 2 files changed, 46 insertions(+), 34 deletions(-) (limited to 'kernel/trace/ftrace.c') diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index 57592a9dd63..10b1d7c1b1d 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -48,7 +48,7 @@ int ftrace_enabled __read_mostly; static int last_ftrace_enabled; /* set when tracing only a pid */ -int ftrace_pid_trace = -1; +struct pid *ftrace_pid_trace; /* Quick disabling of function tracer. */ int function_trace_stop; @@ -153,7 +153,7 @@ static int __register_ftrace_function(struct ftrace_ops *ops) else func = ftrace_list_func; - if (ftrace_pid_trace >= 0) { + if (ftrace_pid_trace) { set_ftrace_pid_function(func); func = ftrace_pid_func; } @@ -209,7 +209,7 @@ static int __unregister_ftrace_function(struct ftrace_ops *ops) if (ftrace_list->next == &ftrace_list_end) { ftrace_func_t func = ftrace_list->func; - if (ftrace_pid_trace >= 0) { + if (ftrace_pid_trace) { set_ftrace_pid_function(func); func = ftrace_pid_func; } @@ -239,7 +239,7 @@ static void ftrace_update_pid_func(void) func = ftrace_trace_function; - if (ftrace_pid_trace >= 0) { + if (ftrace_pid_trace) { set_ftrace_pid_function(func); func = ftrace_pid_func; } else { @@ -1678,18 +1678,40 @@ ftrace_pid_read(struct file *file, char __user *ubuf, char buf[64]; int r; - if (ftrace_pid_trace >= 0) - r = sprintf(buf, "%u\n", ftrace_pid_trace); + if (ftrace_pid_trace) + r = sprintf(buf, "%u\n", pid_nr(ftrace_pid_trace)); else r = sprintf(buf, "no pid\n"); return simple_read_from_buffer(ubuf, cnt, ppos, buf, r); } +static void clear_ftrace_pid_task(struct pid **pid) +{ + struct task_struct *p; + + do_each_pid_task(*pid, PIDTYPE_PID, p) { + clear_tsk_trace_trace(p); + } while_each_pid_task(*pid, PIDTYPE_PID, p); + put_pid(*pid); + + *pid = NULL; +} + +static void set_ftrace_pid_task(struct pid *pid) +{ + struct task_struct *p; + + do_each_pid_task(pid, PIDTYPE_PID, p) { + set_tsk_trace_trace(p); + } while_each_pid_task(pid, PIDTYPE_PID, p); +} + static ssize_t ftrace_pid_write(struct file *filp, const char __user *ubuf, size_t cnt, loff_t *ppos) { + struct pid *pid; char buf[64]; long val; int ret; @@ -1707,40 +1729,30 @@ ftrace_pid_write(struct file *filp, const char __user *ubuf, return ret; mutex_lock(&ftrace_start_lock); - if (ret < 0) { + if (val < 0) { /* disable pid tracing */ - if (ftrace_pid_trace < 0) + if (!ftrace_pid_trace) goto out; - ftrace_pid_trace = -1; + + clear_ftrace_pid_task(&ftrace_pid_trace); } else { - struct task_struct *p; - int found = 0; + pid = find_get_pid(val); - if (ftrace_pid_trace == val) + if (pid == ftrace_pid_trace) { + put_pid(pid); goto out; - - /* - * Find the task that matches this pid. - * TODO: use pid namespaces instead. - */ - rcu_read_lock(); - for_each_process(p) { - if (p->pid == val) { - found = 1; - set_tsk_trace_trace(p); - } else if (test_tsk_trace_trace(p)) - clear_tsk_trace_trace(p); } - rcu_read_unlock(); - if (found) - ftrace_pid_trace = val; - else { - if (ftrace_pid_trace < 0) - goto out; - ftrace_pid_trace = -1; - } + if (ftrace_pid_trace) + clear_ftrace_pid_task(&ftrace_pid_trace); + + if (!pid) + goto out; + + ftrace_pid_trace = pid; + + set_ftrace_pid_task(ftrace_pid_trace); } /* update the function call */ diff --git a/kernel/trace/trace.h b/kernel/trace/trace.h index 95fff37ed97..8b81b4d727b 100644 --- a/kernel/trace/trace.h +++ b/kernel/trace/trace.h @@ -541,11 +541,11 @@ print_graph_function(struct trace_iterator *iter) } #endif /* CONFIG_FUNCTION_GRAPH_TRACER */ -extern int ftrace_pid_trace; +extern struct pid *ftrace_pid_trace; static inline int ftrace_trace_task(struct task_struct *task) { - if (ftrace_pid_trace < 0) + if (ftrace_pid_trace) return 1; return test_tsk_trace_trace(task); -- cgit v1.2.3-70-g09d2 From e32d89569128e76bdf84867be0928902ca9f7555 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Thu, 4 Dec 2008 00:26:41 -0500 Subject: ftrace: add ability to only trace swapper tasks Impact: new feature This patch lets the swapper tasks of all CPUS be filtered by the set_ftrace_pid file. If '0' is echoed into this file, then all the idle tasks (aka swapper) is flagged to be traced. This affects all CPU idle tasks. Signed-off-by: Steven Rostedt Signed-off-by: Ingo Molnar --- kernel/trace/ftrace.c | 74 +++++++++++++++++++++++++++++++++++++++++++-------- 1 file changed, 63 insertions(+), 11 deletions(-) (limited to 'kernel/trace/ftrace.c') diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index 10b1d7c1b1d..eb57dc1ea09 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -49,6 +49,7 @@ static int last_ftrace_enabled; /* set when tracing only a pid */ struct pid *ftrace_pid_trace; +static struct pid * const ftrace_swapper_pid = (struct pid *)1; /* Quick disabling of function tracer. */ int function_trace_stop; @@ -1678,7 +1679,9 @@ ftrace_pid_read(struct file *file, char __user *ubuf, char buf[64]; int r; - if (ftrace_pid_trace) + if (ftrace_pid_trace == ftrace_swapper_pid) + r = sprintf(buf, "swapper tasks\n"); + else if (ftrace_pid_trace) r = sprintf(buf, "%u\n", pid_nr(ftrace_pid_trace)); else r = sprintf(buf, "no pid\n"); @@ -1686,19 +1689,43 @@ ftrace_pid_read(struct file *file, char __user *ubuf, return simple_read_from_buffer(ubuf, cnt, ppos, buf, r); } -static void clear_ftrace_pid_task(struct pid **pid) +static void clear_ftrace_swapper(void) { struct task_struct *p; + int cpu; - do_each_pid_task(*pid, PIDTYPE_PID, p) { + get_online_cpus(); + for_each_online_cpu(cpu) { + p = idle_task(cpu); clear_tsk_trace_trace(p); - } while_each_pid_task(*pid, PIDTYPE_PID, p); - put_pid(*pid); + } + put_online_cpus(); +} - *pid = NULL; +static void set_ftrace_swapper(void) +{ + struct task_struct *p; + int cpu; + + get_online_cpus(); + for_each_online_cpu(cpu) { + p = idle_task(cpu); + set_tsk_trace_trace(p); + } + put_online_cpus(); } -static void set_ftrace_pid_task(struct pid *pid) +static void clear_ftrace_pid(struct pid *pid) +{ + struct task_struct *p; + + do_each_pid_task(pid, PIDTYPE_PID, p) { + clear_tsk_trace_trace(p); + } while_each_pid_task(pid, PIDTYPE_PID, p); + put_pid(pid); +} + +static void set_ftrace_pid(struct pid *pid) { struct task_struct *p; @@ -1707,6 +1734,24 @@ static void set_ftrace_pid_task(struct pid *pid) } while_each_pid_task(pid, PIDTYPE_PID, p); } +static void clear_ftrace_pid_task(struct pid **pid) +{ + if (*pid == ftrace_swapper_pid) + clear_ftrace_swapper(); + else + clear_ftrace_pid(*pid); + + *pid = NULL; +} + +static void set_ftrace_pid_task(struct pid *pid) +{ + if (pid == ftrace_swapper_pid) + set_ftrace_swapper(); + else + set_ftrace_pid(pid); +} + static ssize_t ftrace_pid_write(struct file *filp, const char __user *ubuf, size_t cnt, loff_t *ppos) @@ -1737,11 +1782,18 @@ ftrace_pid_write(struct file *filp, const char __user *ubuf, clear_ftrace_pid_task(&ftrace_pid_trace); } else { - pid = find_get_pid(val); + /* swapper task is special */ + if (!val) { + pid = ftrace_swapper_pid; + if (pid == ftrace_pid_trace) + goto out; + } else { + pid = find_get_pid(val); - if (pid == ftrace_pid_trace) { - put_pid(pid); - goto out; + if (pid == ftrace_pid_trace) { + put_pid(pid); + goto out; + } } if (ftrace_pid_trace) -- cgit v1.2.3-70-g09d2 From faec2ec505d397e9426754722b6e80d519c4938f Mon Sep 17 00:00:00 2001 From: Liming Wang Date: Thu, 4 Dec 2008 14:24:49 +0800 Subject: ftrace: avoid duplicated function when writing set_graph_function Impact: fix a bug in function filter setting when writing function to set_graph_function, we should check whether it has existed in set_graph_function to avoid duplicating. Signed-off-by: Liming Wang Signed-off-by: Ingo Molnar --- kernel/trace/ftrace.c | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-) (limited to 'kernel/trace/ftrace.c') diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index eb57dc1ea09..d2b15653816 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -1425,7 +1425,7 @@ ftrace_set_func(unsigned long *array, int idx, char *buffer) struct dyn_ftrace *rec; struct ftrace_page *pg; int found = 0; - int i; + int i, j; if (ftrace_disabled) return -ENODEV; @@ -1443,7 +1443,13 @@ ftrace_set_func(unsigned long *array, int idx, char *buffer) kallsyms_lookup(rec->ip, NULL, NULL, NULL, str); if (strcmp(str, buffer) == 0) { found = 1; - array[idx] = rec->ip; + for (j = 0; j < idx; j++) + if (array[j] == rec->ip) { + found = 0; + break; + } + if (found) + array[idx] = rec->ip; break; } } -- cgit v1.2.3-70-g09d2 From 21bbecdaaef3a6acc19905ab88c0587817318870 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Thu, 4 Dec 2008 23:30:56 -0500 Subject: ftrace: use init_struct_pid as swapper pid Impact: clean up Using (struct pid *)-1 as the pointer for ftrace_swapper_pid is a little confusing for others. This patch uses the address of the actual init pid structure instead. This change is only for clarity. It does not affect the code itself. Hopefully soon the swapper tasks will all have their own pid structure and then we can clean up the code a bit more. Signed-off-by: Steven Rostedt Signed-off-by: Ingo Molnar --- kernel/trace/ftrace.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel/trace/ftrace.c') diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index d2b15653816..2971fe48f55 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -49,7 +49,7 @@ static int last_ftrace_enabled; /* set when tracing only a pid */ struct pid *ftrace_pid_trace; -static struct pid * const ftrace_swapper_pid = (struct pid *)1; +static struct pid * const ftrace_swapper_pid = &init_struct_pid; /* Quick disabling of function tracer. */ int function_trace_stop; -- cgit v1.2.3-70-g09d2 From 380c4b1411ccd6885f92b2c8ceb08433a720f44e Mon Sep 17 00:00:00 2001 From: Frederic Weisbecker Date: Sat, 6 Dec 2008 03:43:41 +0100 Subject: tracing/function-graph-tracer: append the tracing_graph_flag Impact: Provide a way to pause the function graph tracer As suggested by Steven Rostedt, the previous patch that prevented from spinlock function tracing shouldn't use the raw_spinlock to fix it. It's much better to follow lockdep with normal spinlock, so this patch adds a new flag for each task to make the function graph tracer able to be paused. We also can send an ftrace_printk whithout worrying of the irrelevant traced spinlock during insertion. Signed-off-by: Frederic Weisbecker Signed-off-by: Ingo Molnar --- arch/x86/kernel/ftrace.c | 5 ++++- include/linux/ftrace.h | 13 +++++++++++++ include/linux/sched.h | 2 ++ kernel/trace/ftrace.c | 2 ++ kernel/trace/trace.c | 18 +++++------------- 5 files changed, 26 insertions(+), 14 deletions(-) (limited to 'kernel/trace/ftrace.c') diff --git a/arch/x86/kernel/ftrace.c b/arch/x86/kernel/ftrace.c index f98c4076a17..1b43086b097 100644 --- a/arch/x86/kernel/ftrace.c +++ b/arch/x86/kernel/ftrace.c @@ -476,7 +476,10 @@ void prepare_ftrace_return(unsigned long *parent, unsigned long self_addr) &return_to_handler; /* Nmi's are currently unsupported */ - if (atomic_read(&in_nmi)) + if (unlikely(atomic_read(&in_nmi))) + return; + + if (unlikely(atomic_read(¤t->tracing_graph_pause))) return; /* diff --git a/include/linux/ftrace.h b/include/linux/ftrace.h index 449fa8e9e34..11cac81eed0 100644 --- a/include/linux/ftrace.h +++ b/include/linux/ftrace.h @@ -401,6 +401,16 @@ static inline int task_curr_ret_stack(struct task_struct *t) { return t->curr_ret_stack; } + +static inline void pause_graph_tracing(void) +{ + atomic_inc(¤t->tracing_graph_pause); +} + +static inline void unpause_graph_tracing(void) +{ + atomic_dec(¤t->tracing_graph_pause); +} #else #define __notrace_funcgraph @@ -412,6 +422,9 @@ static inline int task_curr_ret_stack(struct task_struct *tsk) { return -1; } + +static inline void pause_graph_tracing(void) { } +static inline void unpause_graph_tracing(void) { } #endif #ifdef CONFIG_TRACING diff --git a/include/linux/sched.h b/include/linux/sched.h index 4c152e0acc9..4b81fc5f773 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -1379,6 +1379,8 @@ struct task_struct { * because of depth overrun. */ atomic_t trace_overrun; + /* Pause for the tracing */ + atomic_t tracing_graph_pause; #endif #ifdef CONFIG_TRACING /* state flags for use by tracers */ diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index 2971fe48f55..a12f80efcea 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -1998,6 +1998,7 @@ static int alloc_retstack_tasklist(struct ftrace_ret_stack **ret_stack_list) /* Make sure IRQs see the -1 first: */ barrier(); t->ret_stack = ret_stack_list[start++]; + atomic_set(&t->tracing_graph_pause, 0); atomic_set(&t->trace_overrun, 0); } } while_each_thread(g, t); @@ -2077,6 +2078,7 @@ void ftrace_graph_init_task(struct task_struct *t) if (!t->ret_stack) return; t->curr_ret_stack = -1; + atomic_set(&t->tracing_graph_pause, 0); atomic_set(&t->trace_overrun, 0); } else t->ret_stack = NULL; diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index 33549537f30..0b8659bd5ad 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -3590,14 +3590,7 @@ static __init int tracer_init_debugfs(void) int trace_vprintk(unsigned long ip, int depth, const char *fmt, va_list args) { - /* - * Raw Spinlock because a normal spinlock would be traced here - * and append an irrelevant couple spin_lock_irqsave/ - * spin_unlock_irqrestore traced by ftrace around this - * TRACE_PRINTK trace. - */ - static raw_spinlock_t trace_buf_lock = - (raw_spinlock_t)__RAW_SPIN_LOCK_UNLOCKED; + static DEFINE_SPINLOCK(trace_buf_lock); static char trace_buf[TRACE_BUF_SIZE]; struct ring_buffer_event *event; @@ -3618,8 +3611,8 @@ int trace_vprintk(unsigned long ip, int depth, const char *fmt, va_list args) if (unlikely(atomic_read(&data->disabled))) goto out; - local_irq_save(flags); - __raw_spin_lock(&trace_buf_lock); + pause_graph_tracing(); + spin_lock_irqsave(&trace_buf_lock, irq_flags); len = vsnprintf(trace_buf, TRACE_BUF_SIZE, fmt, args); len = min(len, TRACE_BUF_SIZE-1); @@ -3640,9 +3633,8 @@ int trace_vprintk(unsigned long ip, int depth, const char *fmt, va_list args) ring_buffer_unlock_commit(tr->buffer, event, irq_flags); out_unlock: - __raw_spin_unlock(&trace_buf_lock); - local_irq_restore(flags); - + spin_unlock_irqrestore(&trace_buf_lock, irq_flags); + unpause_graph_tracing(); out: preempt_enable_notrace(); -- cgit v1.2.3-70-g09d2 From ea3a6d6d60b2504c573fe3415f6617e8310c0236 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Wed, 17 Dec 2008 15:05:36 -0500 Subject: ftrace: add not to regex on filtering functions Impact: enhancement Ingo Molnar has asked about a way to remove items from the filter lists. Currently, you can only add or replace items. The way items are added to the list is through opening one of the list files (set_ftrace_filter or set_ftrace_notrace) via append. If the file is opened for truncate, the list is cleared. echo spin_lock > /debug/tracing/set_ftrace_filter The above will replace the list with only spin_lock echo spin_lock >> /debug/tracing/set_ftrace_filter The above will add spin_lock to the list. Now this patch adds: echo '!spin_lock' >> /debug/tracing/set_ftrace_filter This will remove spin_lock from the list. The limited glob features of these lists also can be notted. echo '!spin_*' >> /debug/tracing/set_ftrace_filter This will remove all functions that start with 'spin_' Note: echo '!spin_*' > /debug/tracing/set_ftrace_filter will simply clear out the list (notice the '>' instead of '>>') Signed-off-by: Steven Rostedt Signed-off-by: Ingo Molnar --- kernel/trace/ftrace.c | 15 +++++++++++++-- 1 file changed, 13 insertions(+), 2 deletions(-) (limited to 'kernel/trace/ftrace.c') diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index a12f80efcea..2f32969c09d 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -1047,6 +1047,13 @@ ftrace_match(unsigned char *buff, int len, int enable) int type = MATCH_FULL; unsigned long flag = enable ? FTRACE_FL_FILTER : FTRACE_FL_NOTRACE; unsigned i, match = 0, search_len = 0; + int not = 0; + + if (buff[0] == '!') { + not = 1; + buff++; + len--; + } for (i = 0; i < len; i++) { if (buff[i] == '*') { @@ -1100,8 +1107,12 @@ ftrace_match(unsigned char *buff, int len, int enable) matched = 1; break; } - if (matched) - rec->flags |= flag; + if (matched) { + if (not) + rec->flags &= ~flag; + else + rec->flags |= flag; + } } pg = pg->next; } -- cgit v1.2.3-70-g09d2