From dd17c8f72993f9461e9c19250e3f155d6d99df22 Mon Sep 17 00:00:00 2001 From: Rusty Russell Date: Thu, 29 Oct 2009 22:34:15 +0900 Subject: percpu: remove per_cpu__ prefix. Now that the return from alloc_percpu is compatible with the address of per-cpu vars, it makes sense to hand around the address of per-cpu variables. To make this sane, we remove the per_cpu__ prefix we used created to stop people accidentally using these vars directly. Now we have sparse, we can use that (next patch). tj: * Updated to convert stuff which were missed by or added after the original patch. * Kill per_cpu_var() macro. Signed-off-by: Rusty Russell Signed-off-by: Tejun Heo Reviewed-by: Christoph Lameter --- kernel/trace/trace.c | 6 +++--- kernel/trace/trace_functions_graph.c | 4 ++-- 2 files changed, 5 insertions(+), 5 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index 85a5ed70b5b2..b808177af816 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -91,12 +91,12 @@ DEFINE_PER_CPU(int, ftrace_cpu_disabled); static inline void ftrace_disable_cpu(void) { preempt_disable(); - __this_cpu_inc(per_cpu_var(ftrace_cpu_disabled)); + __this_cpu_inc(ftrace_cpu_disabled); } static inline void ftrace_enable_cpu(void) { - __this_cpu_dec(per_cpu_var(ftrace_cpu_disabled)); + __this_cpu_dec(ftrace_cpu_disabled); preempt_enable(); } @@ -1085,7 +1085,7 @@ trace_function(struct trace_array *tr, struct ftrace_entry *entry; /* If we are reading the ring buffer, don't trace */ - if (unlikely(__this_cpu_read(per_cpu_var(ftrace_cpu_disabled)))) + if (unlikely(__this_cpu_read(ftrace_cpu_disabled))) return; event = trace_buffer_lock_reserve(buffer, TRACE_FN, sizeof(*entry), diff --git a/kernel/trace/trace_functions_graph.c b/kernel/trace/trace_functions_graph.c index 90a6daa10962..8614e3241ff8 100644 --- a/kernel/trace/trace_functions_graph.c +++ b/kernel/trace/trace_functions_graph.c @@ -176,7 +176,7 @@ static int __trace_graph_entry(struct trace_array *tr, struct ring_buffer *buffer = tr->buffer; struct ftrace_graph_ent_entry *entry; - if (unlikely(__this_cpu_read(per_cpu_var(ftrace_cpu_disabled)))) + if (unlikely(__this_cpu_read(ftrace_cpu_disabled))) return 0; event = trace_buffer_lock_reserve(buffer, TRACE_GRAPH_ENT, @@ -240,7 +240,7 @@ static void __trace_graph_return(struct trace_array *tr, struct ring_buffer *buffer = tr->buffer; struct ftrace_graph_ret_entry *entry; - if (unlikely(__this_cpu_read(per_cpu_var(ftrace_cpu_disabled)))) + if (unlikely(__this_cpu_read(ftrace_cpu_disabled))) return; event = trace_buffer_lock_reserve(buffer, TRACE_GRAPH_RET, -- cgit v1.2.3 From 07b139c8c81b97bbe55c68daf0cbeca8b1c609ca Mon Sep 17 00:00:00 2001 From: Li Zefan Date: Mon, 21 Dec 2009 14:27:35 +0800 Subject: perf events: Remove CONFIG_EVENT_PROFILE Quoted from Ingo: | This reminds me - i think we should eliminate CONFIG_EVENT_PROFILE - | it's an unnecessary Kconfig complication. If both PERF_EVENTS and | EVENT_TRACING is enabled we should expose generic tracepoints. | | Nor is it limited to event 'profiling', so it has become a misnomer as | well. Signed-off-by: Li Zefan Cc: Frederic Weisbecker Cc: Steven Rostedt Cc: Peter Zijlstra Cc: Paul Mackerras LKML-Reference: <4B2F1557.2050705@cn.fujitsu.com> Signed-off-by: Ingo Molnar --- kernel/trace/Makefile | 4 +++- kernel/trace/trace_events_filter.c | 4 ++-- kernel/trace/trace_kprobe.c | 14 +++++++------- kernel/trace/trace_syscalls.c | 5 ++--- 4 files changed, 14 insertions(+), 13 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/Makefile b/kernel/trace/Makefile index cd9ecd89ec77..d00c6fe23f54 100644 --- a/kernel/trace/Makefile +++ b/kernel/trace/Makefile @@ -51,7 +51,9 @@ endif obj-$(CONFIG_EVENT_TRACING) += trace_events.o obj-$(CONFIG_EVENT_TRACING) += trace_export.o obj-$(CONFIG_FTRACE_SYSCALLS) += trace_syscalls.o -obj-$(CONFIG_EVENT_PROFILE) += trace_event_profile.o +ifeq ($(CONFIG_PERF_EVENTS),y) +obj-$(CONFIG_EVENT_TRACING) += trace_event_profile.o +endif obj-$(CONFIG_EVENT_TRACING) += trace_events_filter.o obj-$(CONFIG_KPROBE_EVENT) += trace_kprobe.o obj-$(CONFIG_KSYM_TRACER) += trace_ksym.o diff --git a/kernel/trace/trace_events_filter.c b/kernel/trace/trace_events_filter.c index 50504cb228de..74563d7e102e 100644 --- a/kernel/trace/trace_events_filter.c +++ b/kernel/trace/trace_events_filter.c @@ -1360,7 +1360,7 @@ out_unlock: return err; } -#ifdef CONFIG_EVENT_PROFILE +#ifdef CONFIG_PERF_EVENTS void ftrace_profile_free_filter(struct perf_event *event) { @@ -1428,5 +1428,5 @@ out_unlock: return err; } -#endif /* CONFIG_EVENT_PROFILE */ +#endif /* CONFIG_PERF_EVENTS */ diff --git a/kernel/trace/trace_kprobe.c b/kernel/trace/trace_kprobe.c index 375f81a568dc..75d75dec226a 100644 --- a/kernel/trace/trace_kprobe.c +++ b/kernel/trace/trace_kprobe.c @@ -1249,7 +1249,7 @@ static int kretprobe_event_show_format(struct ftrace_event_call *call, ", REC->" FIELD_STRING_RETIP); } -#ifdef CONFIG_EVENT_PROFILE +#ifdef CONFIG_PERF_EVENTS /* Kprobe profile handler */ static __kprobes int kprobe_profile_func(struct kprobe *kp, @@ -1407,7 +1407,7 @@ static void probe_profile_disable(struct ftrace_event_call *call) disable_kprobe(&tp->rp.kp); } } -#endif /* CONFIG_EVENT_PROFILE */ +#endif /* CONFIG_PERF_EVENTS */ static __kprobes @@ -1417,10 +1417,10 @@ int kprobe_dispatcher(struct kprobe *kp, struct pt_regs *regs) if (tp->flags & TP_FLAG_TRACE) kprobe_trace_func(kp, regs); -#ifdef CONFIG_EVENT_PROFILE +#ifdef CONFIG_PERF_EVENTS if (tp->flags & TP_FLAG_PROFILE) kprobe_profile_func(kp, regs); -#endif /* CONFIG_EVENT_PROFILE */ +#endif return 0; /* We don't tweek kernel, so just return 0 */ } @@ -1431,10 +1431,10 @@ int kretprobe_dispatcher(struct kretprobe_instance *ri, struct pt_regs *regs) if (tp->flags & TP_FLAG_TRACE) kretprobe_trace_func(ri, regs); -#ifdef CONFIG_EVENT_PROFILE +#ifdef CONFIG_PERF_EVENTS if (tp->flags & TP_FLAG_PROFILE) kretprobe_profile_func(ri, regs); -#endif /* CONFIG_EVENT_PROFILE */ +#endif return 0; /* We don't tweek kernel, so just return 0 */ } @@ -1463,7 +1463,7 @@ static int register_probe_event(struct trace_probe *tp) call->regfunc = probe_event_enable; call->unregfunc = probe_event_disable; -#ifdef CONFIG_EVENT_PROFILE +#ifdef CONFIG_PERF_EVENTS call->profile_enable = probe_profile_enable; call->profile_disable = probe_profile_disable; #endif diff --git a/kernel/trace/trace_syscalls.c b/kernel/trace/trace_syscalls.c index 75289f372dd2..f694f66d75b0 100644 --- a/kernel/trace/trace_syscalls.c +++ b/kernel/trace/trace_syscalls.c @@ -421,7 +421,7 @@ int __init init_ftrace_syscalls(void) } core_initcall(init_ftrace_syscalls); -#ifdef CONFIG_EVENT_PROFILE +#ifdef CONFIG_PERF_EVENTS static DECLARE_BITMAP(enabled_prof_enter_syscalls, NR_syscalls); static DECLARE_BITMAP(enabled_prof_exit_syscalls, NR_syscalls); @@ -626,6 +626,5 @@ void prof_sysexit_disable(struct ftrace_event_call *call) mutex_unlock(&syscall_trace_lock); } -#endif - +#endif /* CONFIG_PERF_EVENTS */ -- cgit v1.2.3 From 79615760f380ec86cd58204744e774c33fab9211 Mon Sep 17 00:00:00 2001 From: Christoph Lameter Date: Tue, 5 Jan 2010 15:34:50 +0900 Subject: local_t: Move local.h include to ringbuffer.c and ring_buffer_benchmark.c ringbuffer*.c are the last users of local.h. Remove the include from modules.h and add it to ringbuffer files. Signed-off-by: Christoph Lameter Signed-off-by: Tejun Heo --- kernel/trace/ring_buffer.c | 1 + kernel/trace/ring_buffer_benchmark.c | 1 + 2 files changed, 2 insertions(+) (limited to 'kernel/trace') diff --git a/kernel/trace/ring_buffer.c b/kernel/trace/ring_buffer.c index 2326b04c95c4..eb6c8988c31a 100644 --- a/kernel/trace/ring_buffer.c +++ b/kernel/trace/ring_buffer.c @@ -20,6 +20,7 @@ #include #include +#include #include "trace.h" /* diff --git a/kernel/trace/ring_buffer_benchmark.c b/kernel/trace/ring_buffer_benchmark.c index b2477caf09c2..df74c7982255 100644 --- a/kernel/trace/ring_buffer_benchmark.c +++ b/kernel/trace/ring_buffer_benchmark.c @@ -8,6 +8,7 @@ #include #include #include +#include struct rb_page { u64 ts; -- cgit v1.2.3 From 809826a389040e0ad9d646b587bccc0e34691afd Mon Sep 17 00:00:00 2001 From: Lai Jiangshan Date: Tue, 15 Dec 2009 15:39:34 +0800 Subject: tracing: Have __dynamic_array() define a field This is part of a patch set that removes the show_format method in the ftrace event macros. This patch set requires that all fields are added to the ftrace_event_call->fields. This patch changes __dynamic_array() to call trace_define_field() to include fields that use __dynamic_array(). Signed-off-by: Lai Jiangshan LKML-Reference: <4B273D36.8090100@cn.fujitsu.com> Signed-off-by: Steven Rostedt --- kernel/trace/trace_export.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_export.c b/kernel/trace/trace_export.c index d4fa5dc1ee4e..9978a4f40090 100644 --- a/kernel/trace/trace_export.c +++ b/kernel/trace/trace_export.c @@ -175,7 +175,12 @@ ftrace_format_##name(struct ftrace_event_call *unused, \ return ret; #undef __dynamic_array -#define __dynamic_array(type, item) +#define __dynamic_array(type, item) \ + ret = trace_define_field(event_call, #type, #item, \ + offsetof(typeof(field), item), \ + 0, is_signed_type(type), FILTER_OTHER);\ + if (ret) \ + return ret; #undef FTRACE_ENTRY #define FTRACE_ENTRY(name, struct_name, id, tstruct, print) \ -- cgit v1.2.3 From 509e760cd91c831983097ae174cb6c0b8c6c8e6b Mon Sep 17 00:00:00 2001 From: Lai Jiangshan Date: Tue, 15 Dec 2009 15:39:42 +0800 Subject: tracing: Add print_fmt field This is part of a patch set that removes the show_format method in the ftrace event macros. The print_fmt field is added to hold the string that shows the print_fmt in the event format files. This patch only adds the field but it is currently not used. Later patches will use this field to enable us to remove the show_format field and function. Signed-off-by: Lai Jiangshan LKML-Reference: <4B273D3E.2000704@cn.fujitsu.com> Signed-off-by: Steven Rostedt --- kernel/trace/trace_export.c | 7 +++++++ 1 file changed, 7 insertions(+) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_export.c b/kernel/trace/trace_export.c index 9978a4f40090..95d14b640a66 100644 --- a/kernel/trace/trace_export.c +++ b/kernel/trace/trace_export.c @@ -203,6 +203,9 @@ static int ftrace_raw_init_event(struct ftrace_event_call *call) return 0; } +#undef __entry +#define __entry REC + #undef __field #define __field(type, item) @@ -218,6 +221,9 @@ static int ftrace_raw_init_event(struct ftrace_event_call *call) #undef __dynamic_array #define __dynamic_array(type, item) +#undef F_printk +#define F_printk(fmt, args...) #fmt ", " __stringify(args) + #undef FTRACE_ENTRY #define FTRACE_ENTRY(call, struct_name, type, tstruct, print) \ \ @@ -228,6 +234,7 @@ __attribute__((section("_ftrace_events"))) event_##call = { \ .id = type, \ .system = __stringify(TRACE_SYSTEM), \ .raw_init = ftrace_raw_init_event, \ + .print_fmt = print, \ .show_format = ftrace_format_##call, \ .define_fields = ftrace_define_fields_##call, \ }; \ -- cgit v1.2.3 From 50307a45f8515f6244e3b08e6b19824b9fbfe293 Mon Sep 17 00:00:00 2001 From: Lai Jiangshan Date: Tue, 15 Dec 2009 15:39:45 +0800 Subject: tracing/syscalls: Init print_fmt for syscall events This is part of a patch set that removes the show_format method in the ftrace event macros. Add the print_fmt initialization to the syscall events. The print_fmt is still not used, but will be in the follow up patches. Signed-off-by: Lai Jiangshan LKML-Reference: <4B273D41.609@cn.fujitsu.com> Signed-off-by: Steven Rostedt --- kernel/trace/trace_syscalls.c | 68 ++++++++++++++++++++++++++++++++++++++++++- 1 file changed, 67 insertions(+), 1 deletion(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_syscalls.c b/kernel/trace/trace_syscalls.c index 75289f372dd2..1352b0a36fac 100644 --- a/kernel/trace/trace_syscalls.c +++ b/kernel/trace/trace_syscalls.c @@ -191,6 +191,67 @@ int syscall_enter_format(struct ftrace_event_call *call, struct trace_seq *s) return trace_seq_putc(s, '\n'); } +static +int __set_enter_print_fmt(struct syscall_metadata *entry, char *buf, int len) +{ + int i; + int pos = 0; + + /* When len=0, we just calculate the needed length */ +#define LEN_OR_ZERO (len ? len - pos : 0) + + pos += snprintf(buf + pos, LEN_OR_ZERO, "\""); + for (i = 0; i < entry->nb_args; i++) { + pos += snprintf(buf + pos, LEN_OR_ZERO, "%s: 0x%%0%zulx%s", + entry->args[i], sizeof(unsigned long), + i == entry->nb_args - 1 ? "" : ", "); + } + pos += snprintf(buf + pos, LEN_OR_ZERO, "\""); + + for (i = 0; i < entry->nb_args; i++) { + pos += snprintf(buf + pos, LEN_OR_ZERO, + ", ((unsigned long)(REC->%s))", entry->args[i]); + } + +#undef LEN_OR_ZERO + + /* return the length of print_fmt */ + return pos; +} + +static int set_syscall_print_fmt(struct ftrace_event_call *call) +{ + char *print_fmt; + int len; + struct syscall_metadata *entry = call->data; + + if (entry->enter_event != call) { + call->print_fmt = "\"0x%lx\", REC->ret"; + return 0; + } + + /* First: called with 0 length to calculate the needed length */ + len = __set_enter_print_fmt(entry, NULL, 0); + + print_fmt = kmalloc(len + 1, GFP_KERNEL); + if (!print_fmt) + return -ENOMEM; + + /* Second: actually write the @print_fmt */ + __set_enter_print_fmt(entry, print_fmt, len + 1); + call->print_fmt = print_fmt; + + return 0; +} + +static void free_syscall_print_fmt(struct ftrace_event_call *call) +{ + struct syscall_metadata *entry = call->data; + + if (entry->enter_event == call) + kfree(call->print_fmt); +} + int syscall_exit_format(struct ftrace_event_call *call, struct trace_seq *s) { int ret; @@ -386,9 +447,14 @@ int init_syscall_trace(struct ftrace_event_call *call) { int id; + if (set_syscall_print_fmt(call) < 0) + return -ENOMEM; + id = register_ftrace_event(call->event); - if (!id) + if (!id) { + free_syscall_print_fmt(call); return -ENODEV; + } call->id = id; INIT_LIST_HEAD(&call->fields); return 0; -- cgit v1.2.3 From a342a0280b981c130e32dbb94dbd3a57959c4d04 Mon Sep 17 00:00:00 2001 From: Lai Jiangshan Date: Tue, 15 Dec 2009 15:39:49 +0800 Subject: tracing/kprobes: Init print_fmt for kprobe events This is part of a patch set that removes the show_format method in the ftrace event macros. Add the print_fmt initialization to the kprobe events. The print_fmt is still not used, but will be in the follow up patches. Signed-off-by: Lai Jiangshan LKML-Reference: <4B273D45.3080100@cn.fujitsu.com> Acked-by: Masami Hiramatsu Signed-off-by: Steven Rostedt --- kernel/trace/trace_kprobe.c | 64 ++++++++++++++++++++++++++++++++++++++++++++- 1 file changed, 63 insertions(+), 1 deletion(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_kprobe.c b/kernel/trace/trace_kprobe.c index 6ea90c0e2c96..147491dccead 100644 --- a/kernel/trace/trace_kprobe.c +++ b/kernel/trace/trace_kprobe.c @@ -1250,6 +1250,62 @@ static int kretprobe_event_show_format(struct ftrace_event_call *call, ", REC->" FIELD_STRING_RETIP); } +static int __set_print_fmt(struct trace_probe *tp, char *buf, int len) +{ + int i; + int pos = 0; + + const char *fmt, *arg; + + if (!probe_is_return(tp)) { + fmt = "(%lx)"; + arg = "REC->" FIELD_STRING_IP; + } else { + fmt = "(%lx <- %lx)"; + arg = "REC->" FIELD_STRING_FUNC ", REC->" FIELD_STRING_RETIP; + } + + /* When len=0, we just calculate the needed length */ +#define LEN_OR_ZERO (len ? len - pos : 0) + + pos += snprintf(buf + pos, LEN_OR_ZERO, "\"%s", fmt); + + for (i = 0; i < tp->nr_args; i++) { + pos += snprintf(buf + pos, LEN_OR_ZERO, " %s=%%lx", + tp->args[i].name); + } + + pos += snprintf(buf + pos, LEN_OR_ZERO, "\", %s", arg); + + for (i = 0; i < tp->nr_args; i++) { + pos += snprintf(buf + pos, LEN_OR_ZERO, ", REC->%s", + tp->args[i].name); + } + +#undef LEN_OR_ZERO + + /* return the length of print_fmt */ + return pos; +} + +static int set_print_fmt(struct trace_probe *tp) +{ + int len; + char *print_fmt; + + /* First: called with 0 length to calculate the needed length */ + len = __set_print_fmt(tp, NULL, 0); + print_fmt = kmalloc(len + 1, GFP_KERNEL); + if (!print_fmt) + return -ENOMEM; + + /* Second: actually write the @print_fmt */ + __set_print_fmt(tp, print_fmt, len + 1); + tp->call.print_fmt = print_fmt; + + return 0; +} + #ifdef CONFIG_EVENT_PROFILE /* Kprobe profile handler */ @@ -1456,10 +1512,14 @@ static int register_probe_event(struct trace_probe *tp) call->show_format = kprobe_event_show_format; call->define_fields = kprobe_event_define_fields; } + if (set_print_fmt(tp) < 0) + return -ENOMEM; call->event = &tp->event; call->id = register_ftrace_event(&tp->event); - if (!call->id) + if (!call->id) { + kfree(call->print_fmt); return -ENODEV; + } call->enabled = 0; call->regfunc = probe_event_enable; call->unregfunc = probe_event_disable; @@ -1472,6 +1532,7 @@ static int register_probe_event(struct trace_probe *tp) ret = trace_add_event_call(call); if (ret) { pr_info("Failed to register kprobe event: %s\n", call->name); + kfree(call->print_fmt); unregister_ftrace_event(&tp->event); } return ret; @@ -1481,6 +1542,7 @@ static void unregister_probe_event(struct trace_probe *tp) { /* tp->event is unregistered in trace_remove_event_call() */ trace_remove_event_call(&tp->call); + kfree(tp->call.print_fmt); } /* Make a debugfs interface for controling probe points */ -- cgit v1.2.3 From c7ef3a9004201bca90626db246a19dadd2c29c9b Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Mon, 28 Dec 2009 21:13:59 -0500 Subject: tracing: Have syscall tracing call its own init function In the clean up of having all events call one specific function, the syscall event init was changed to call this helper function. With the new print_fmt updates, the syscalls need to do special initializations. This patch converts the syscall events to call its own init function again. Cc: Lai Jiangshan Cc: Li Zefan Signed-off-by: Steven Rostedt --- kernel/trace/trace_syscalls.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_syscalls.c b/kernel/trace/trace_syscalls.c index 1352b0a36fac..a78e86349ecb 100644 --- a/kernel/trace/trace_syscalls.c +++ b/kernel/trace/trace_syscalls.c @@ -450,14 +450,14 @@ int init_syscall_trace(struct ftrace_event_call *call) if (set_syscall_print_fmt(call) < 0) return -ENOMEM; - id = register_ftrace_event(call->event); - if (!id) { + id = trace_event_raw_init(call); + + if (id < 0) { free_syscall_print_fmt(call); - return -ENODEV; + return id; } - call->id = id; - INIT_LIST_HEAD(&call->fields); - return 0; + + return id; } int __init init_ftrace_syscalls(void) -- cgit v1.2.3 From 5a65e956220efc2421e21ee56d6153fd5c533a95 Mon Sep 17 00:00:00 2001 From: Lai Jiangshan Date: Tue, 15 Dec 2009 15:39:53 +0800 Subject: tracing: Use defined fields and print_fmt to print formats The calls ftrace_format_##call() and ftrace_define_fields_##call() are almost duplicate in functionality. With the addition of the print_fmt in previous patches, these two functions can be merged into one. The trace_define_field() defines the fields and links them into the struct ftrace_event_call. The previous patches introduced the print_fmt field and this can now be used with the trace_define_field() to create the event format file fields and print_fmt field. The struct ftrace_event_call->fields are used to print the fields The struct ftrace_event_call->print_fmt is used to print the "print fmt: XXXXXXXXXXX" line. Signed-off-by: Lai Jiangshan LKML-Reference: <4B273D49.5000006@cn.fujitsu.com> Signed-off-by: Steven Rostedt --- kernel/trace/trace_events.c | 65 ++++++++++++++++++++++++++++----------------- 1 file changed, 41 insertions(+), 24 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_events.c b/kernel/trace/trace_events.c index 189b09baf4fb..250ec865d5f5 100644 --- a/kernel/trace/trace_events.c +++ b/kernel/trace/trace_events.c @@ -528,33 +528,16 @@ extern char *__bad_type_size(void); #type, "common_" #name, offsetof(typeof(field), name), \ sizeof(field.name), is_signed_type(type) -static int trace_write_header(struct trace_seq *s) -{ - struct trace_entry field; - - /* struct trace_entry */ - return trace_seq_printf(s, - "\tfield:%s %s;\toffset:%zu;\tsize:%zu;\tsigned:%u;\n" - "\tfield:%s %s;\toffset:%zu;\tsize:%zu;\tsigned:%u;\n" - "\tfield:%s %s;\toffset:%zu;\tsize:%zu;\tsigned:%u;\n" - "\tfield:%s %s;\toffset:%zu;\tsize:%zu;\tsigned:%u;\n" - "\tfield:%s %s;\toffset:%zu;\tsize:%zu;\tsigned:%u;\n" - "\n", - FIELD(unsigned short, type), - FIELD(unsigned char, flags), - FIELD(unsigned char, preempt_count), - FIELD(int, pid), - FIELD(int, lock_depth)); -} - static ssize_t event_format_read(struct file *filp, char __user *ubuf, size_t cnt, loff_t *ppos) { struct ftrace_event_call *call = filp->private_data; + struct ftrace_event_field *field; struct trace_seq *s; + int common_field_count = 5; char *buf; - int r; + int r = 0; if (*ppos) return 0; @@ -565,14 +548,48 @@ event_format_read(struct file *filp, char __user *ubuf, size_t cnt, trace_seq_init(s); - /* If any of the first writes fail, so will the show_format. */ - trace_seq_printf(s, "name: %s\n", call->name); trace_seq_printf(s, "ID: %d\n", call->id); trace_seq_printf(s, "format:\n"); - trace_write_header(s); - r = call->show_format(call, s); + list_for_each_entry_reverse(field, &call->fields, link) { + /* + * Smartly shows the array type(except dynamic array). + * Normal: + * field:TYPE VAR + * If TYPE := TYPE[LEN], it is shown: + * field:TYPE VAR[LEN] + */ + const char *array_descriptor = strchr(field->type, '['); + + if (!strncmp(field->type, "__data_loc", 10)) + array_descriptor = NULL; + + if (!array_descriptor) { + r = trace_seq_printf(s, "\tfield:%s %s;\toffset:%u;" + "\tsize:%u;\tsigned:%d;\n", + field->type, field->name, field->offset, + field->size, !!field->is_signed); + } else { + r = trace_seq_printf(s, "\tfield:%.*s %s%s;\toffset:%u;" + "\tsize:%u;\tsigned:%d;\n", + (int)(array_descriptor - field->type), + field->type, field->name, + array_descriptor, field->offset, + field->size, !!field->is_signed); + } + + if (--common_field_count == 0) + r = trace_seq_printf(s, "\n"); + + if (!r) + break; + } + + if (r) + r = trace_seq_printf(s, "\nprint fmt: %s\n", + call->print_fmt); + if (!r) { /* * ug! The format output is bigger than a PAGE!! -- cgit v1.2.3 From 0fa0edaf32b9a78b9854f1da98d4511a501089b0 Mon Sep 17 00:00:00 2001 From: Lai Jiangshan Date: Tue, 15 Dec 2009 15:39:57 +0800 Subject: tracing: Remove show_format and related macros from TRACE_EVENT The previous patches added the use of print_fmt string and changes the trace_define_field() function to also create the fields and format output for the event format files. text data bss dec hex filename 5857201 1355780 9336808 16549789 fc879d vmlinux 5884589 1351684 9337896 16574169 fce6d9 vmlinux-orig The above shows the size of the vmlinux after this patch set compared to the vmlinux-orig which is before the patch set. This saves us 27k on text, 1k on bss and adds just 4k of data. The total savings of 24k in size. Signed-off-by: Lai Jiangshan LKML-Reference: <4B273D4D.40604@cn.fujitsu.com> Acked-by: Masami Hiramatsu Signed-off-by: Steven Rostedt --- kernel/trace/trace_events.c | 12 ------- kernel/trace/trace_export.c | 73 ---------------------------------------- kernel/trace/trace_kprobe.c | 78 ------------------------------------------- kernel/trace/trace_syscalls.c | 66 ------------------------------------ 4 files changed, 229 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_events.c b/kernel/trace/trace_events.c index 250ec865d5f5..c2a3077b7353 100644 --- a/kernel/trace/trace_events.c +++ b/kernel/trace/trace_events.c @@ -520,14 +520,6 @@ out: return ret; } -extern char *__bad_type_size(void); - -#undef FIELD -#define FIELD(type, name) \ - sizeof(type) != sizeof(field.name) ? __bad_type_size() : \ - #type, "common_" #name, offsetof(typeof(field), name), \ - sizeof(field.name), is_signed_type(type) - static ssize_t event_format_read(struct file *filp, char __user *ubuf, size_t cnt, loff_t *ppos) @@ -965,10 +957,6 @@ event_create_dir(struct ftrace_event_call *call, struct dentry *d_events, filter); } - /* A trace may not want to export its format */ - if (!call->show_format) - return 0; - trace_create_file("format", 0444, call->dir, call, format); diff --git a/kernel/trace/trace_export.c b/kernel/trace/trace_export.c index 95d14b640a66..e091f64ba6ce 100644 --- a/kernel/trace/trace_export.c +++ b/kernel/trace/trace_export.c @@ -62,78 +62,6 @@ static void __always_unused ____ftrace_check_##name(void) \ #include "trace_entries.h" - -#undef __field -#define __field(type, item) \ - ret = trace_seq_printf(s, "\tfield:" #type " " #item ";\t" \ - "offset:%zu;\tsize:%zu;\tsigned:%u;\n", \ - offsetof(typeof(field), item), \ - sizeof(field.item), is_signed_type(type)); \ - if (!ret) \ - return 0; - -#undef __field_desc -#define __field_desc(type, container, item) \ - ret = trace_seq_printf(s, "\tfield:" #type " " #item ";\t" \ - "offset:%zu;\tsize:%zu;\tsigned:%u;\n", \ - offsetof(typeof(field), container.item), \ - sizeof(field.container.item), \ - is_signed_type(type)); \ - if (!ret) \ - return 0; - -#undef __array -#define __array(type, item, len) \ - ret = trace_seq_printf(s, "\tfield:" #type " " #item "[" #len "];\t" \ - "offset:%zu;\tsize:%zu;\tsigned:%u;\n", \ - offsetof(typeof(field), item), \ - sizeof(field.item), is_signed_type(type)); \ - if (!ret) \ - return 0; - -#undef __array_desc -#define __array_desc(type, container, item, len) \ - ret = trace_seq_printf(s, "\tfield:" #type " " #item "[" #len "];\t" \ - "offset:%zu;\tsize:%zu;\tsigned:%u;\n", \ - offsetof(typeof(field), container.item), \ - sizeof(field.container.item), \ - is_signed_type(type)); \ - if (!ret) \ - return 0; - -#undef __dynamic_array -#define __dynamic_array(type, item) \ - ret = trace_seq_printf(s, "\tfield:" #type " " #item ";\t" \ - "offset:%zu;\tsize:0;\tsigned:%u;\n", \ - offsetof(typeof(field), item), \ - is_signed_type(type)); \ - if (!ret) \ - return 0; - -#undef F_printk -#define F_printk(fmt, args...) "%s, %s\n", #fmt, __stringify(args) - -#undef __entry -#define __entry REC - -#undef FTRACE_ENTRY -#define FTRACE_ENTRY(name, struct_name, id, tstruct, print) \ -static int \ -ftrace_format_##name(struct ftrace_event_call *unused, \ - struct trace_seq *s) \ -{ \ - struct struct_name field __attribute__((unused)); \ - int ret = 0; \ - \ - tstruct; \ - \ - trace_seq_printf(s, "\nprint fmt: " print); \ - \ - return ret; \ -} - -#include "trace_entries.h" - #undef __field #define __field(type, item) \ ret = trace_define_field(event_call, #type, #item, \ @@ -235,7 +163,6 @@ __attribute__((section("_ftrace_events"))) event_##call = { \ .system = __stringify(TRACE_SYSTEM), \ .raw_init = ftrace_raw_init_event, \ .print_fmt = print, \ - .show_format = ftrace_format_##call, \ .define_fields = ftrace_define_fields_##call, \ }; \ diff --git a/kernel/trace/trace_kprobe.c b/kernel/trace/trace_kprobe.c index 147491dccead..c99029916c76 100644 --- a/kernel/trace/trace_kprobe.c +++ b/kernel/trace/trace_kprobe.c @@ -1174,82 +1174,6 @@ static int kretprobe_event_define_fields(struct ftrace_event_call *event_call) return 0; } -static int __probe_event_show_format(struct trace_seq *s, - struct trace_probe *tp, const char *fmt, - const char *arg) -{ - int i; - - /* Show format */ - if (!trace_seq_printf(s, "\nprint fmt: \"%s", fmt)) - return 0; - - for (i = 0; i < tp->nr_args; i++) - if (!trace_seq_printf(s, " %s=%%lx", tp->args[i].name)) - return 0; - - if (!trace_seq_printf(s, "\", %s", arg)) - return 0; - - for (i = 0; i < tp->nr_args; i++) - if (!trace_seq_printf(s, ", REC->%s", tp->args[i].name)) - return 0; - - return trace_seq_puts(s, "\n"); -} - -#undef SHOW_FIELD -#define SHOW_FIELD(type, item, name) \ - do { \ - ret = trace_seq_printf(s, "\tfield:" #type " %s;\t" \ - "offset:%u;\tsize:%u;\tsigned:%d;\n", name,\ - (unsigned int)offsetof(typeof(field), item),\ - (unsigned int)sizeof(type), \ - is_signed_type(type)); \ - if (!ret) \ - return 0; \ - } while (0) - -static int kprobe_event_show_format(struct ftrace_event_call *call, - struct trace_seq *s) -{ - struct kprobe_trace_entry field __attribute__((unused)); - int ret, i; - struct trace_probe *tp = (struct trace_probe *)call->data; - - SHOW_FIELD(unsigned long, ip, FIELD_STRING_IP); - SHOW_FIELD(int, nargs, FIELD_STRING_NARGS); - - /* Show fields */ - for (i = 0; i < tp->nr_args; i++) - SHOW_FIELD(unsigned long, args[i], tp->args[i].name); - trace_seq_puts(s, "\n"); - - return __probe_event_show_format(s, tp, "(%lx)", - "REC->" FIELD_STRING_IP); -} - -static int kretprobe_event_show_format(struct ftrace_event_call *call, - struct trace_seq *s) -{ - struct kretprobe_trace_entry field __attribute__((unused)); - int ret, i; - struct trace_probe *tp = (struct trace_probe *)call->data; - - SHOW_FIELD(unsigned long, func, FIELD_STRING_FUNC); - SHOW_FIELD(unsigned long, ret_ip, FIELD_STRING_RETIP); - SHOW_FIELD(int, nargs, FIELD_STRING_NARGS); - - /* Show fields */ - for (i = 0; i < tp->nr_args; i++) - SHOW_FIELD(unsigned long, args[i], tp->args[i].name); - trace_seq_puts(s, "\n"); - - return __probe_event_show_format(s, tp, "(%lx <- %lx)", - "REC->" FIELD_STRING_FUNC - ", REC->" FIELD_STRING_RETIP); -} - static int __set_print_fmt(struct trace_probe *tp, char *buf, int len) { int i; @@ -1504,12 +1428,10 @@ static int register_probe_event(struct trace_probe *tp) if (probe_is_return(tp)) { tp->event.trace = print_kretprobe_event; call->raw_init = probe_event_raw_init; - call->show_format = kretprobe_event_show_format; call->define_fields = kretprobe_event_define_fields; } else { tp->event.trace = print_kprobe_event; call->raw_init = probe_event_raw_init; - call->show_format = kprobe_event_show_format; call->define_fields = kprobe_event_define_fields; } if (set_print_fmt(tp) < 0) diff --git a/kernel/trace/trace_syscalls.c b/kernel/trace/trace_syscalls.c index a78e86349ecb..49cea70fbf6d 100644 --- a/kernel/trace/trace_syscalls.c +++ b/kernel/trace/trace_syscalls.c @@ -143,54 +143,6 @@ extern char *__bad_type_size(void); #type, #name, offsetof(typeof(trace), name), \ sizeof(trace.name), is_signed_type(type) -int syscall_enter_format(struct ftrace_event_call *call, struct trace_seq *s) -{ - int i; - int ret; - struct syscall_metadata *entry = call->data; - struct syscall_trace_enter trace; - int offset = offsetof(struct syscall_trace_enter, args); - - ret = trace_seq_printf(s, "\tfield:%s %s;\toffset:%zu;\tsize:%zu;" - "\tsigned:%u;\n", - SYSCALL_FIELD(int, nr)); - if (!ret) - return 0; - - for (i = 0; i < entry->nb_args; i++) { - ret = trace_seq_printf(s, "\tfield:%s %s;", entry->types[i], - entry->args[i]); - if (!ret) - return 0; - ret = trace_seq_printf(s, "\toffset:%d;\tsize:%zu;" - "\tsigned:%u;\n", offset, - sizeof(unsigned long), - is_signed_type(unsigned long)); - if (!ret) - return 0; - offset += sizeof(unsigned long); - } - - trace_seq_puts(s, "\nprint fmt: \""); - for (i = 0; i < entry->nb_args; i++) { - ret = trace_seq_printf(s, "%s: 0x%%0%zulx%s", entry->args[i], - sizeof(unsigned long), - i == entry->nb_args - 1 ? "" : ", "); - if (!ret) - return 0; - } - trace_seq_putc(s, '"'); - - for (i = 0; i < entry->nb_args; i++) { - ret = trace_seq_printf(s, ", ((unsigned long)(REC->%s))", - entry->args[i]); - if (!ret) - return 0; - } - - return trace_seq_putc(s, '\n'); -} - static int __set_enter_print_fmt(struct syscall_metadata *entry, char *buf, int len) { @@ -252,24 +204,6 @@ static void free_syscall_print_fmt(struct ftrace_event_call *call) kfree(call->print_fmt); } -int syscall_exit_format(struct ftrace_event_call *call, struct trace_seq *s) -{ - int ret; - struct syscall_trace_exit trace; - - ret = trace_seq_printf(s, - "\tfield:%s %s;\toffset:%zu;\tsize:%zu;" - "\tsigned:%u;\n" - "\tfield:%s %s;\toffset:%zu;\tsize:%zu;" - "\tsigned:%u;\n", - SYSCALL_FIELD(int, nr), - SYSCALL_FIELD(long, ret)); - if (!ret) - return 0; - - return trace_seq_printf(s, "\nprint fmt: \"0x%%lx\", REC->ret\n"); -} - int syscall_enter_define_fields(struct ftrace_event_call *call) { struct syscall_trace_enter trace; -- cgit v1.2.3 From 7e53bd42d14c75192b99674c40fcc359392da59d Mon Sep 17 00:00:00 2001 From: Lai Jiangshan Date: Wed, 6 Jan 2010 20:08:50 +0800 Subject: tracing: Consolidate protection of reader access to the ring buffer At the beginning, access to the ring buffer was fully serialized by trace_types_lock. Patch d7350c3f4569 gives more freedom to readers, and patch b04cc6b1f6 adds code to protect trace_pipe and cpu#/trace_pipe. But actually it is not enough, ring buffer readers are not always read-only, they may consume data. This patch makes accesses to trace, trace_pipe, trace_pipe_raw cpu#/trace, cpu#/trace_pipe and cpu#/trace_pipe_raw serialized. And removes tracing_reader_cpumask which is used to protect trace_pipe. Details: Ring buffer serializes readers, but it is low level protection. The validity of the events (which returns by ring_buffer_peek() ..etc) are not protected by ring buffer. The content of events may become garbage if we allow another process to consume these events concurrently: A) the page of the consumed events may become a normal page (not reader page) in ring buffer, and this page will be rewritten by the events producer. B) The page of the consumed events may become a page for splice_read, and this page will be returned to system. This patch adds trace_access_lock() and trace_access_unlock() primitives. These primitives allow multi process access to different cpu ring buffers concurrently. These primitives don't distinguish read-only and read-consume access. Multi read-only access is also serialized. And we don't use these primitives when we open files, we only use them when we read files. Signed-off-by: Lai Jiangshan LKML-Reference: <4B447D52.1050602@cn.fujitsu.com> Signed-off-by: Steven Rostedt --- kernel/trace/trace.c | 136 ++++++++++++++++++++++++++++++++++++--------------- 1 file changed, 97 insertions(+), 39 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index 0df1b0f2cb9e..abdd333a0825 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -32,6 +32,7 @@ #include #include #include +#include #include #include #include @@ -102,9 +103,6 @@ static inline void ftrace_enable_cpu(void) static cpumask_var_t __read_mostly tracing_buffer_mask; -/* Define which cpu buffers are currently read in trace_pipe */ -static cpumask_var_t tracing_reader_cpumask; - #define for_each_tracing_cpu(cpu) \ for_each_cpu(cpu, tracing_buffer_mask) @@ -243,12 +241,91 @@ static struct tracer *current_trace __read_mostly; /* * trace_types_lock is used to protect the trace_types list. - * This lock is also used to keep user access serialized. - * Accesses from userspace will grab this lock while userspace - * activities happen inside the kernel. */ static DEFINE_MUTEX(trace_types_lock); +/* + * serialize the access of the ring buffer + * + * ring buffer serializes readers, but it is low level protection. + * The validity of the events (which returns by ring_buffer_peek() ..etc) + * are not protected by ring buffer. + * + * The content of events may become garbage if we allow other process consumes + * these events concurrently: + * A) the page of the consumed events may become a normal page + * (not reader page) in ring buffer, and this page will be rewrited + * by events producer. + * B) The page of the consumed events may become a page for splice_read, + * and this page will be returned to system. + * + * These primitives allow multi process access to different cpu ring buffer + * concurrently. + * + * These primitives don't distinguish read-only and read-consume access. + * Multi read-only access are also serialized. + */ + +#ifdef CONFIG_SMP +static DECLARE_RWSEM(all_cpu_access_lock); +static DEFINE_PER_CPU(struct mutex, cpu_access_lock); + +static inline void trace_access_lock(int cpu) +{ + if (cpu == TRACE_PIPE_ALL_CPU) { + /* gain it for accessing the whole ring buffer. */ + down_write(&all_cpu_access_lock); + } else { + /* gain it for accessing a cpu ring buffer. */ + + /* Firstly block other trace_access_lock(TRACE_PIPE_ALL_CPU). */ + down_read(&all_cpu_access_lock); + + /* Secondly block other access to this @cpu ring buffer. */ + mutex_lock(&per_cpu(cpu_access_lock, cpu)); + } +} + +static inline void trace_access_unlock(int cpu) +{ + if (cpu == TRACE_PIPE_ALL_CPU) { + up_write(&all_cpu_access_lock); + } else { + mutex_unlock(&per_cpu(cpu_access_lock, cpu)); + up_read(&all_cpu_access_lock); + } +} + +static inline void trace_access_lock_init(void) +{ + int cpu; + + for_each_possible_cpu(cpu) + mutex_init(&per_cpu(cpu_access_lock, cpu)); +} + +#else + +static DEFINE_MUTEX(access_lock); + +static inline void trace_access_lock(int cpu) +{ + (void)cpu; + mutex_lock(&access_lock); +} + +static inline void trace_access_unlock(int cpu) +{ + (void)cpu; + mutex_unlock(&access_lock); +} + +static inline void trace_access_lock_init(void) +{ +} + +#endif + /* trace_wait is a waitqueue for tasks blocked on trace_poll */ static DECLARE_WAIT_QUEUE_HEAD(trace_wait); @@ -1580,12 +1657,6 @@ static void tracing_iter_reset(struct trace_iterator *iter, int cpu) } /* - * No necessary locking here. The worst thing which can - * happen is loosing events consumed at the same time - * by a trace_pipe reader. - * Other than that, we don't risk to crash the ring buffer - * because it serializes the readers. - * * The current tracer is copied to avoid a global locking * all around. */ @@ -1640,12 +1711,16 @@ static void *s_start(struct seq_file *m, loff_t *pos) } trace_event_read_lock(); + trace_access_lock(cpu_file); return p; } static void s_stop(struct seq_file *m, void *p) { + struct trace_iterator *iter = m->private; + atomic_dec(&trace_record_cmdline_disabled); + trace_access_unlock(iter->cpu_file); trace_event_read_unlock(); } @@ -2836,22 +2911,6 @@ static int tracing_open_pipe(struct inode *inode, struct file *filp) mutex_lock(&trace_types_lock); - /* We only allow one reader per cpu */ - if (cpu_file == TRACE_PIPE_ALL_CPU) { - if (!cpumask_empty(tracing_reader_cpumask)) { - ret = -EBUSY; - goto out; - } - cpumask_setall(tracing_reader_cpumask); - } else { - if (!cpumask_test_cpu(cpu_file, tracing_reader_cpumask)) - cpumask_set_cpu(cpu_file, tracing_reader_cpumask); - else { - ret = -EBUSY; - goto out; - } - } - /* create a buffer to store the information to pass to userspace */ iter = kzalloc(sizeof(*iter), GFP_KERNEL); if (!iter) { @@ -2907,12 +2966,6 @@ static int tracing_release_pipe(struct inode *inode, struct file *file) mutex_lock(&trace_types_lock); - if (iter->cpu_file == TRACE_PIPE_ALL_CPU) - cpumask_clear(tracing_reader_cpumask); - else - cpumask_clear_cpu(iter->cpu_file, tracing_reader_cpumask); - - if (iter->trace->pipe_close) iter->trace->pipe_close(iter); @@ -3074,6 +3127,7 @@ waitagain: iter->pos = -1; trace_event_read_lock(); + trace_access_lock(iter->cpu_file); while (find_next_entry_inc(iter) != NULL) { enum print_line_t ret; int len = iter->seq.len; @@ -3090,6 +3144,7 @@ waitagain: if (iter->seq.len >= cnt) break; } + trace_access_unlock(iter->cpu_file); trace_event_read_unlock(); /* Now copy what we have to the user */ @@ -3215,6 +3270,7 @@ static ssize_t tracing_splice_read_pipe(struct file *filp, } trace_event_read_lock(); + trace_access_lock(iter->cpu_file); /* Fill as many pages as possible. */ for (i = 0, rem = len; i < PIPE_BUFFERS && rem; i++) { @@ -3238,6 +3294,7 @@ static ssize_t tracing_splice_read_pipe(struct file *filp, trace_seq_init(&iter->seq); } + trace_access_unlock(iter->cpu_file); trace_event_read_unlock(); mutex_unlock(&iter->mutex); @@ -3539,10 +3596,12 @@ tracing_buffers_read(struct file *filp, char __user *ubuf, info->read = 0; + trace_access_lock(info->cpu); ret = ring_buffer_read_page(info->tr->buffer, &info->spare, count, info->cpu, 0); + trace_access_unlock(info->cpu); if (ret < 0) return 0; @@ -3670,6 +3729,7 @@ tracing_buffers_splice_read(struct file *file, loff_t *ppos, len &= PAGE_MASK; } + trace_access_lock(info->cpu); entries = ring_buffer_entries_cpu(info->tr->buffer, info->cpu); for (i = 0; i < PIPE_BUFFERS && len && entries; i++, len -= PAGE_SIZE) { @@ -3717,6 +3777,7 @@ tracing_buffers_splice_read(struct file *file, loff_t *ppos, entries = ring_buffer_entries_cpu(info->tr->buffer, info->cpu); } + trace_access_unlock(info->cpu); spd.nr_pages = i; /* did we read anything? */ @@ -4153,6 +4214,8 @@ static __init int tracer_init_debugfs(void) struct dentry *d_tracer; int cpu; + trace_access_lock_init(); + d_tracer = tracing_init_dentry(); trace_create_file("tracing_enabled", 0644, d_tracer, @@ -4387,9 +4450,6 @@ __init static int tracer_alloc_buffers(void) if (!alloc_cpumask_var(&tracing_cpumask, GFP_KERNEL)) goto out_free_buffer_mask; - if (!zalloc_cpumask_var(&tracing_reader_cpumask, GFP_KERNEL)) - goto out_free_tracing_cpumask; - /* To save memory, keep the ring buffer size to its minimum */ if (ring_buffer_expanded) ring_buf_size = trace_buf_size; @@ -4447,8 +4507,6 @@ __init static int tracer_alloc_buffers(void) return 0; out_free_cpumask: - free_cpumask_var(tracing_reader_cpumask); -out_free_tracing_cpumask: free_cpumask_var(tracing_cpumask); out_free_buffer_mask: free_cpumask_var(tracing_buffer_mask); -- cgit v1.2.3 From d931369b74b3d6f2044f595af6f3dd074f65d9cf Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Wed, 6 Jan 2010 17:27:11 -0500 Subject: tracing: Add stack dump to trace_printk if stacktrace option is set If the ftrace stacktrace option is set, then add the stack dumps to trace_printk. Requested-by: Peter Zijlstra Signed-off-by: Steven Rostedt --- kernel/trace/trace.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index abdd333a0825..5314c90bbc83 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -1392,8 +1392,10 @@ int trace_vbprintk(unsigned long ip, const char *fmt, va_list args) entry->fmt = fmt; memcpy(entry->buf, trace_buf, sizeof(u32) * len); - if (!filter_check_discard(call, entry, buffer, event)) + if (!filter_check_discard(call, entry, buffer, event)) { ring_buffer_unlock_commit(buffer, event); + ftrace_trace_stack(buffer, flags, 6, pc); + } out_unlock: arch_spin_unlock(&trace_buf_lock); @@ -1466,8 +1468,10 @@ int trace_array_vprintk(struct trace_array *tr, memcpy(&entry->buf, trace_buf, len); entry->buf[len] = '\0'; - if (!filter_check_discard(call, entry, buffer, event)) + if (!filter_check_discard(call, entry, buffer, event)) { ring_buffer_unlock_commit(buffer, event); + ftrace_trace_stack(buffer, irq_flags, 6, pc); + } out_unlock: arch_spin_unlock(&trace_buf_lock); -- cgit v1.2.3 From 14640106f243a3b29944d7198569090fa6546f2d Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Tue, 5 Jan 2010 17:46:48 -0500 Subject: tracing/kprobe: Drop function argument access syntax Drop function argument access syntax, because the function arguments depend on not only architecture but also compile-options and function API. And now, we have perf-probe for finding register/memory assigned to each argument. Signed-off-by: Masami Hiramatsu Cc: Frederic Weisbecker Cc: Arnaldo Carvalho de Melo Cc: systemtap Cc: DLE Cc: Frederic Weisbecker Cc: Steven Rostedt Cc: Roland McGrath Cc: Oleg Nesterov Cc: Mahesh Salgaonkar Cc: Benjamin Herrenschmidt Cc: Michael Neuling Cc: linuxppc-dev@ozlabs.org LKML-Reference: <20100105224648.19431.52309.stgit@dhcp-100-2-132.bos.redhat.com> Signed-off-by: Ingo Molnar --- kernel/trace/trace_kprobe.c | 18 +----------------- 1 file changed, 1 insertion(+), 17 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_kprobe.c b/kernel/trace/trace_kprobe.c index 47f54ab57b68..7ac728ded964 100644 --- a/kernel/trace/trace_kprobe.c +++ b/kernel/trace/trace_kprobe.c @@ -91,11 +91,6 @@ static __kprobes unsigned long fetch_memory(struct pt_regs *regs, void *addr) return retval; } -static __kprobes unsigned long fetch_argument(struct pt_regs *regs, void *num) -{ - return regs_get_argument_nth(regs, (unsigned int)((unsigned long)num)); -} - static __kprobes unsigned long fetch_retvalue(struct pt_regs *regs, void *dummy) { @@ -231,9 +226,7 @@ static int probe_arg_string(char *buf, size_t n, struct fetch_func *ff) { int ret = -EINVAL; - if (ff->func == fetch_argument) - ret = snprintf(buf, n, "$arg%lu", (unsigned long)ff->data); - else if (ff->func == fetch_register) { + if (ff->func == fetch_register) { const char *name; name = regs_query_register_name((unsigned int)((long)ff->data)); ret = snprintf(buf, n, "%%%s", name); @@ -489,14 +482,6 @@ static int parse_probe_vars(char *arg, struct fetch_func *ff, int is_return) } } else ret = -EINVAL; - } else if (strncmp(arg, "arg", 3) == 0 && isdigit(arg[3])) { - ret = strict_strtoul(arg + 3, 10, ¶m); - if (ret || param > PARAM_MAX_ARGS) - ret = -EINVAL; - else { - ff->func = fetch_argument; - ff->data = (void *)param; - } } else ret = -EINVAL; return ret; @@ -611,7 +596,6 @@ static int create_trace_probe(int argc, char **argv) * - Add kprobe: p[:[GRP/]EVENT] KSYM[+OFFS]|KADDR [FETCHARGS] * - Add kretprobe: r[:[GRP/]EVENT] KSYM[+0] [FETCHARGS] * Fetch args: - * $argN : fetch Nth of function argument. (N:0-) * $retval : fetch return value * $stack : fetch stack address * $stackN : fetch Nth of stack (N:0-) -- cgit v1.2.3 From 24a53652e361321b09df5040711e69387344ce09 Mon Sep 17 00:00:00 2001 From: Frederic Weisbecker Date: Thu, 14 Jan 2010 08:23:15 +0100 Subject: tracing: Drop the tr check from the graph tracing path Each time we save a function entry from the function graph tracer, we check if the trace array is set, which is wasteful because it is set anyway before we start the tracer. All we need is to ensure we have good read and write orderings. When we set the trace array, we just need to guarantee it to be visible before starting tracing. Signed-off-by: Frederic Weisbecker Acked-by: Steven Rostedt Cc: Lai Jiangshan Cc: Paul E. McKenney LKML-Reference: <1263453795-7496-1-git-send-regression-fweisbec@gmail.com> Signed-off-by: Ingo Molnar --- kernel/trace/trace_functions_graph.c | 19 ++++++++++--------- 1 file changed, 10 insertions(+), 9 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_functions_graph.c b/kernel/trace/trace_functions_graph.c index b1342c5d37cf..f2252296607c 100644 --- a/kernel/trace/trace_functions_graph.c +++ b/kernel/trace/trace_functions_graph.c @@ -212,9 +212,6 @@ int trace_graph_entry(struct ftrace_graph_ent *trace) int cpu; int pc; - if (unlikely(!tr)) - return 0; - if (!ftrace_trace_task(current)) return 0; @@ -287,11 +284,20 @@ void trace_graph_return(struct ftrace_graph_ret *trace) local_irq_restore(flags); } +void set_graph_array(struct trace_array *tr) +{ + graph_array = tr; + + /* Make graph_array visible before we start tracing */ + + smp_mb(); +} + static int graph_trace_init(struct trace_array *tr) { int ret; - graph_array = tr; + set_graph_array(tr); ret = register_ftrace_graph(&trace_graph_return, &trace_graph_entry); if (ret) @@ -301,11 +307,6 @@ static int graph_trace_init(struct trace_array *tr) return 0; } -void set_graph_array(struct trace_array *tr) -{ - graph_array = tr; -} - static void graph_trace_reset(struct trace_array *tr) { tracing_stop_cmdline_record(); -- cgit v1.2.3 From 231e36f4d2e63dd770db80b9f5113310c2bcfcfd Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Thu, 14 Jan 2010 00:12:12 -0500 Subject: tracing/kprobe: Update kprobe tracing self test for new syntax Update kprobe tracing self test for new syntax (it supports deleting individual probes, and drops $argN support) and behavior change (new probes are disabled in default). This selftest includes the following checks: - Adding function-entry probe and return probe with arguments. - Enabling these probes. - Deleting it individually. Signed-off-by: Masami Hiramatsu Cc: systemtap Cc: DLE Cc: Steven Rostedt Cc: Frederic Weisbecker LKML-Reference: <20100114051211.7814.29436.stgit@localhost6.localdomain6> Signed-off-by: Ingo Molnar --- kernel/trace/trace_kprobe.c | 55 ++++++++++++++++++++++++++++++++++++++------- 1 file changed, 47 insertions(+), 8 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_kprobe.c b/kernel/trace/trace_kprobe.c index 7ac728ded964..d6266cad6953 100644 --- a/kernel/trace/trace_kprobe.c +++ b/kernel/trace/trace_kprobe.c @@ -1507,28 +1507,67 @@ static int kprobe_trace_selftest_target(int a1, int a2, int a3, static __init int kprobe_trace_self_tests_init(void) { - int ret; + int ret, warn = 0; int (*target)(int, int, int, int, int, int); + struct trace_probe *tp; target = kprobe_trace_selftest_target; pr_info("Testing kprobe tracing: "); ret = command_trace_probe("p:testprobe kprobe_trace_selftest_target " - "$arg1 $arg2 $arg3 $arg4 $stack $stack0"); - if (WARN_ON_ONCE(ret)) - pr_warning("error enabling function entry\n"); + "$stack $stack0 +0($stack)"); + if (WARN_ON_ONCE(ret)) { + pr_warning("error on probing function entry.\n"); + warn++; + } else { + /* Enable trace point */ + tp = find_probe_event("testprobe", KPROBE_EVENT_SYSTEM); + if (WARN_ON_ONCE(tp == NULL)) { + pr_warning("error on getting new probe.\n"); + warn++; + } else + probe_event_enable(&tp->call); + } ret = command_trace_probe("r:testprobe2 kprobe_trace_selftest_target " "$retval"); - if (WARN_ON_ONCE(ret)) - pr_warning("error enabling function return\n"); + if (WARN_ON_ONCE(ret)) { + pr_warning("error on probing function return.\n"); + warn++; + } else { + /* Enable trace point */ + tp = find_probe_event("testprobe2", KPROBE_EVENT_SYSTEM); + if (WARN_ON_ONCE(tp == NULL)) { + pr_warning("error on getting new probe.\n"); + warn++; + } else + probe_event_enable(&tp->call); + } + + if (warn) + goto end; ret = target(1, 2, 3, 4, 5, 6); - cleanup_all_probes(); + ret = command_trace_probe("-:testprobe"); + if (WARN_ON_ONCE(ret)) { + pr_warning("error on deleting a probe.\n"); + warn++; + } + + ret = command_trace_probe("-:testprobe2"); + if (WARN_ON_ONCE(ret)) { + pr_warning("error on deleting a probe.\n"); + warn++; + } - pr_cont("OK\n"); +end: + cleanup_all_probes(); + if (warn) + pr_cont("NG: Some tests are failed. Please check them.\n"); + else + pr_cont("OK\n"); return 0; } -- cgit v1.2.3 From 74bf4076f2ed79b5510440b72a561823a8852ec0 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Mon, 25 Jan 2010 15:11:53 -0500 Subject: tracing: Prevent kernel oops with corrupted buffer If the contents of the ftrace ring buffer gets corrupted and the trace file is read, it could create a kernel oops (usualy just killing the user task thread). This is caused by the checking of the pid in the buffer. If the pid is negative, it still references the cmdline cache array, which could point to an invalid address. The simple fix is to test for negative PIDs. Signed-off-by: Steven Rostedt --- kernel/trace/trace.c | 5 +++++ 1 file changed, 5 insertions(+) (limited to 'kernel/trace') diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index 0df1b0f2cb9e..eac6875cb990 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -951,6 +951,11 @@ void trace_find_cmdline(int pid, char comm[]) return; } + if (WARN_ON_ONCE(pid < 0)) { + strcpy(comm, ""); + return; + } + if (pid > PID_MAX_DEFAULT) { strcpy(comm, "<...>"); return; -- cgit v1.2.3 From 492a74f4210e15f4701422e2e1c4cd3c1e45ddae Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Mon, 25 Jan 2010 15:17:47 -0500 Subject: ring-buffer: Check if ring buffer iterator has stale data Usually reads of the ring buffer is performed by a single task. There are two types of reads from the ring buffer. One is a consuming read which will consume the entry that was read and the next read will be the entry that follows. The other is an iterator that will let the user read the contents of the ring buffer without modifying it. When an iterator is allocated, writes to the ring buffer are disabled to protect the iterator. The problem exists when consuming reads happen while an iterator is allocated. Specifically, the kind of read that swaps out an entire page (used by splice) and replaces it with a new read. If the iterator is on the page that is swapped out, then the next read may read from this swapped out page and return garbage. This patch adds a check when reading the iterator to make sure that the iterator contents are still valid. If a consuming read has taken place, the iterator is reset. Signed-off-by: Steven Rostedt --- kernel/trace/ring_buffer.c | 13 +++++++++++++ 1 file changed, 13 insertions(+) (limited to 'kernel/trace') diff --git a/kernel/trace/ring_buffer.c b/kernel/trace/ring_buffer.c index edefe3b2801b..503b630e0bda 100644 --- a/kernel/trace/ring_buffer.c +++ b/kernel/trace/ring_buffer.c @@ -464,6 +464,8 @@ struct ring_buffer_iter { struct ring_buffer_per_cpu *cpu_buffer; unsigned long head; struct buffer_page *head_page; + struct buffer_page *cache_reader_page; + unsigned long cache_read; u64 read_stamp; }; @@ -2716,6 +2718,8 @@ static void rb_iter_reset(struct ring_buffer_iter *iter) iter->read_stamp = cpu_buffer->read_stamp; else iter->read_stamp = iter->head_page->page->time_stamp; + iter->cache_reader_page = cpu_buffer->reader_page; + iter->cache_read = cpu_buffer->read; } /** @@ -3066,6 +3070,15 @@ rb_iter_peek(struct ring_buffer_iter *iter, u64 *ts) cpu_buffer = iter->cpu_buffer; buffer = cpu_buffer->buffer; + /* + * Check if someone performed a consuming read to + * the buffer. A consuming read invalidates the iterator + * and we need to reset the iterator in this case. + */ + if (unlikely(iter->cache_read != cpu_buffer->read || + iter->cache_reader_page != cpu_buffer->reader_page)) + rb_iter_reset(iter); + again: /* * We repeat when a timestamp is encountered. -- cgit v1.2.3 From 3c05d7482777f15e71bb4cb1ba78dee2800dfec6 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Tue, 26 Jan 2010 16:14:08 -0500 Subject: ring-buffer: Check for end of page in iterator If the iterator comes to an empty page for some reason, or if the page is emptied by a consuming read. The iterator code currently does not check if the iterator is pass the contents, and may return a false entry. This patch adds a check to the ring buffer iterator to test if the current page has been completely read and sets the iterator to the next page if necessary. Signed-off-by: Steven Rostedt --- kernel/trace/ring_buffer.c | 11 ++++++++--- 1 file changed, 8 insertions(+), 3 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/ring_buffer.c b/kernel/trace/ring_buffer.c index 503b630e0bda..8c1b2d290718 100644 --- a/kernel/trace/ring_buffer.c +++ b/kernel/trace/ring_buffer.c @@ -3064,9 +3064,6 @@ rb_iter_peek(struct ring_buffer_iter *iter, u64 *ts) struct ring_buffer_event *event; int nr_loops = 0; - if (ring_buffer_iter_empty(iter)) - return NULL; - cpu_buffer = iter->cpu_buffer; buffer = cpu_buffer->buffer; @@ -3080,6 +3077,9 @@ rb_iter_peek(struct ring_buffer_iter *iter, u64 *ts) rb_iter_reset(iter); again: + if (ring_buffer_iter_empty(iter)) + return NULL; + /* * We repeat when a timestamp is encountered. * We can get multiple timestamps by nested interrupts or also @@ -3094,6 +3094,11 @@ rb_iter_peek(struct ring_buffer_iter *iter, u64 *ts) if (rb_per_cpu_empty(cpu_buffer)) return NULL; + if (iter->head >= local_read(&iter->head_page->page->commit)) { + rb_inc_iter(iter); + goto again; + } + event = rb_iter_head_event(iter); switch (event->type_len) { -- cgit v1.2.3 From 03688970347bfea32823953a7ce5886d1713205f Mon Sep 17 00:00:00 2001 From: Mike Frysinger Date: Fri, 22 Jan 2010 08:12:47 -0500 Subject: tracing/documentation: Cover new frame pointer semantics Update the graph tracer examples to cover the new frame pointer semantics (in terms of passing it along). Move the HAVE_FUNCTION_GRAPH_FP_TEST docs out of the Kconfig, into the right place, and expand on the details. Signed-off-by: Mike Frysinger LKML-Reference: <1264165967-18938-1-git-send-email-vapier@gentoo.org> Signed-off-by: Steven Rostedt --- kernel/trace/Kconfig | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/Kconfig b/kernel/trace/Kconfig index 6c22d8a2f289..60e2ce0181ee 100644 --- a/kernel/trace/Kconfig +++ b/kernel/trace/Kconfig @@ -27,9 +27,7 @@ config HAVE_FUNCTION_GRAPH_TRACER config HAVE_FUNCTION_GRAPH_FP_TEST bool help - An arch may pass in a unique value (frame pointer) to both the - entering and exiting of a function. On exit, the value is compared - and if it does not match, then it will panic the kernel. + See Documentation/trace/ftrace-design.txt config HAVE_FUNCTION_TRACE_MCOUNT_TEST bool -- cgit v1.2.3 From ea2c68a08fedb5053ba312d661e47df9f4d72411 Mon Sep 17 00:00:00 2001 From: Lai Jiangshan Date: Wed, 13 Jan 2010 19:38:30 +0800 Subject: tracing: Simplify test for function_graph tracing start point In the function graph tracer, a calling function is to be traced only when it is enabled through the set_graph_function file, or when it is nested in an enabled function. Current code uses TSK_TRACE_FL_GRAPH to test whether it is nested or not. Looking at the code, we can get this: (trace->depth > 0) <==> (TSK_TRACE_FL_GRAPH is set) trace->depth is more explicit to tell that it is nested. So we use trace->depth directly and simplify the code. No functionality is changed. TSK_TRACE_FL_GRAPH is not removed yet, it is left for future usage. Signed-off-by: Lai Jiangshan Cc: Ingo Molnar Cc: Steven Rostedt LKML-Reference: <4B4DB0B6.7040607@cn.fujitsu.com> Signed-off-by: Frederic Weisbecker --- kernel/trace/trace.h | 2 +- kernel/trace/trace_functions_graph.c | 8 ++------ 2 files changed, 3 insertions(+), 7 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace.h b/kernel/trace/trace.h index 4df6a77eb196..ce077fbbf552 100644 --- a/kernel/trace/trace.h +++ b/kernel/trace/trace.h @@ -504,7 +504,7 @@ static inline int ftrace_graph_addr(unsigned long addr) { int i; - if (!ftrace_graph_count || test_tsk_trace_graph(current)) + if (!ftrace_graph_count) return 1; for (i = 0; i < ftrace_graph_count; i++) { diff --git a/kernel/trace/trace_functions_graph.c b/kernel/trace/trace_functions_graph.c index f2252296607c..616b135c9eb9 100644 --- a/kernel/trace/trace_functions_graph.c +++ b/kernel/trace/trace_functions_graph.c @@ -215,7 +215,8 @@ int trace_graph_entry(struct ftrace_graph_ent *trace) if (!ftrace_trace_task(current)) return 0; - if (!ftrace_graph_addr(trace->func)) + /* trace it when it is-nested-in or is a function enabled. */ + if (!(trace->depth || ftrace_graph_addr(trace->func))) return 0; local_irq_save(flags); @@ -228,9 +229,6 @@ int trace_graph_entry(struct ftrace_graph_ent *trace) } else { ret = 0; } - /* 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); @@ -278,8 +276,6 @@ void trace_graph_return(struct ftrace_graph_ret *trace) pc = preempt_count(); __trace_graph_return(tr, trace, flags, pc); } - if (!trace->depth) - clear_tsk_trace_graph(current); atomic_dec(&data->disabled); local_irq_restore(flags); } -- cgit v1.2.3 From 430ad5a600a83956749307b13257c464c3826b55 Mon Sep 17 00:00:00 2001 From: Xiao Guangrong Date: Thu, 28 Jan 2010 09:32:29 +0800 Subject: perf: Factorize trace events raw sample buffer operations Introduce ftrace_perf_buf_prepare() and ftrace_perf_buf_submit() to gather the common code that operates on raw events sampling buffer. This cleans up redundant code between regular trace events, syscall events and kprobe events. Changelog v1->v2: - Rename function name as per Masami and Frederic's suggestion - Add __kprobes for ftrace_perf_buf_prepare() and make ftrace_perf_buf_submit() inline as per Masami's suggestion - Export ftrace_perf_buf_prepare since modules will use it Signed-off-by: Xiao Guangrong Acked-by: Masami Hiramatsu Cc: Ingo Molnar Cc: Steven Rostedt Cc: Paul Mackerras Cc: Jason Baron Cc: Peter Zijlstra LKML-Reference: <4B60E92D.9000808@cn.fujitsu.com> Signed-off-by: Frederic Weisbecker --- kernel/trace/trace_event_profile.c | 52 ++++++++++++++++++++--- kernel/trace/trace_kprobe.c | 86 +++++--------------------------------- kernel/trace/trace_syscalls.c | 71 +++++-------------------------- 3 files changed, 67 insertions(+), 142 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_event_profile.c b/kernel/trace/trace_event_profile.c index 9e25573242cf..f0d693005075 100644 --- a/kernel/trace/trace_event_profile.c +++ b/kernel/trace/trace_event_profile.c @@ -6,14 +6,12 @@ */ #include +#include #include "trace.h" -char *perf_trace_buf; -EXPORT_SYMBOL_GPL(perf_trace_buf); - -char *perf_trace_buf_nmi; -EXPORT_SYMBOL_GPL(perf_trace_buf_nmi); +static char *perf_trace_buf; +static char *perf_trace_buf_nmi; typedef typeof(char [FTRACE_MAX_PROFILE_SIZE]) perf_trace_t ; @@ -120,3 +118,47 @@ void ftrace_profile_disable(int event_id) } mutex_unlock(&event_mutex); } + +__kprobes void *ftrace_perf_buf_prepare(int size, unsigned short type, + int *rctxp, unsigned long *irq_flags) +{ + struct trace_entry *entry; + char *trace_buf, *raw_data; + int pc, cpu; + + pc = preempt_count(); + + /* Protect the per cpu buffer, begin the rcu read side */ + local_irq_save(*irq_flags); + + *rctxp = perf_swevent_get_recursion_context(); + if (*rctxp < 0) + goto err_recursion; + + cpu = smp_processor_id(); + + if (in_nmi()) + trace_buf = rcu_dereference(perf_trace_buf_nmi); + else + trace_buf = rcu_dereference(perf_trace_buf); + + if (!trace_buf) + goto err; + + raw_data = per_cpu_ptr(trace_buf, cpu); + + /* zero the dead bytes from align to not leak stack to user */ + *(u64 *)(&raw_data[size - sizeof(u64)]) = 0ULL; + + entry = (struct trace_entry *)raw_data; + tracing_generic_entry_update(entry, *irq_flags, pc); + entry->type = type; + + return raw_data; +err: + perf_swevent_put_recursion_context(*rctxp); +err_recursion: + local_irq_restore(*irq_flags); + return NULL; +} +EXPORT_SYMBOL_GPL(ftrace_perf_buf_prepare); diff --git a/kernel/trace/trace_kprobe.c b/kernel/trace/trace_kprobe.c index d6266cad6953..2e28ee36646f 100644 --- a/kernel/trace/trace_kprobe.c +++ b/kernel/trace/trace_kprobe.c @@ -1243,14 +1243,10 @@ static __kprobes int kprobe_profile_func(struct kprobe *kp, struct trace_probe *tp = container_of(kp, struct trace_probe, rp.kp); struct ftrace_event_call *call = &tp->call; struct kprobe_trace_entry *entry; - struct trace_entry *ent; - int size, __size, i, pc, __cpu; + int size, __size, i; unsigned long irq_flags; - char *trace_buf; - char *raw_data; int rctx; - pc = preempt_count(); __size = SIZEOF_KPROBE_TRACE_ENTRY(tp->nr_args); size = ALIGN(__size + sizeof(u32), sizeof(u64)); size -= sizeof(u32); @@ -1258,45 +1254,16 @@ static __kprobes int kprobe_profile_func(struct kprobe *kp, "profile buffer not large enough")) return 0; - /* - * Protect the non nmi buffer - * This also protects the rcu read side - */ - local_irq_save(irq_flags); - - rctx = perf_swevent_get_recursion_context(); - if (rctx < 0) - goto end_recursion; - - __cpu = smp_processor_id(); - - if (in_nmi()) - trace_buf = rcu_dereference(perf_trace_buf_nmi); - else - trace_buf = rcu_dereference(perf_trace_buf); - - if (!trace_buf) - goto end; - - raw_data = per_cpu_ptr(trace_buf, __cpu); - - /* Zero dead bytes from alignment to avoid buffer leak to userspace */ - *(u64 *)(&raw_data[size - sizeof(u64)]) = 0ULL; - entry = (struct kprobe_trace_entry *)raw_data; - ent = &entry->ent; + entry = ftrace_perf_buf_prepare(size, call->id, &rctx, &irq_flags); + if (!entry) + return 0; - tracing_generic_entry_update(ent, irq_flags, pc); - ent->type = call->id; entry->nargs = tp->nr_args; entry->ip = (unsigned long)kp->addr; for (i = 0; i < tp->nr_args; i++) entry->args[i] = call_fetch(&tp->args[i].fetch, regs); - perf_tp_event(call->id, entry->ip, 1, entry, size); -end: - perf_swevent_put_recursion_context(rctx); -end_recursion: - local_irq_restore(irq_flags); + ftrace_perf_buf_submit(entry, size, rctx, entry->ip, 1, irq_flags); return 0; } @@ -1308,14 +1275,10 @@ static __kprobes int kretprobe_profile_func(struct kretprobe_instance *ri, struct trace_probe *tp = container_of(ri->rp, struct trace_probe, rp); struct ftrace_event_call *call = &tp->call; struct kretprobe_trace_entry *entry; - struct trace_entry *ent; - int size, __size, i, pc, __cpu; + int size, __size, i; unsigned long irq_flags; - char *trace_buf; - char *raw_data; int rctx; - pc = preempt_count(); __size = SIZEOF_KRETPROBE_TRACE_ENTRY(tp->nr_args); size = ALIGN(__size + sizeof(u32), sizeof(u64)); size -= sizeof(u32); @@ -1323,46 +1286,17 @@ static __kprobes int kretprobe_profile_func(struct kretprobe_instance *ri, "profile buffer not large enough")) return 0; - /* - * Protect the non nmi buffer - * This also protects the rcu read side - */ - local_irq_save(irq_flags); - - rctx = perf_swevent_get_recursion_context(); - if (rctx < 0) - goto end_recursion; - - __cpu = smp_processor_id(); - - if (in_nmi()) - trace_buf = rcu_dereference(perf_trace_buf_nmi); - else - trace_buf = rcu_dereference(perf_trace_buf); - - if (!trace_buf) - goto end; - - raw_data = per_cpu_ptr(trace_buf, __cpu); - - /* Zero dead bytes from alignment to avoid buffer leak to userspace */ - *(u64 *)(&raw_data[size - sizeof(u64)]) = 0ULL; - entry = (struct kretprobe_trace_entry *)raw_data; - ent = &entry->ent; + entry = ftrace_perf_buf_prepare(size, call->id, &rctx, &irq_flags); + if (!entry) + return 0; - tracing_generic_entry_update(ent, irq_flags, pc); - ent->type = call->id; entry->nargs = tp->nr_args; entry->func = (unsigned long)tp->rp.kp.addr; entry->ret_ip = (unsigned long)ri->ret_addr; for (i = 0; i < tp->nr_args; i++) entry->args[i] = call_fetch(&tp->args[i].fetch, regs); - perf_tp_event(call->id, entry->ret_ip, 1, entry, size); -end: - perf_swevent_put_recursion_context(rctx); -end_recursion: - local_irq_restore(irq_flags); + ftrace_perf_buf_submit(entry, size, rctx, entry->ret_ip, 1, irq_flags); return 0; } diff --git a/kernel/trace/trace_syscalls.c b/kernel/trace/trace_syscalls.c index f694f66d75b0..4e332b9e449c 100644 --- a/kernel/trace/trace_syscalls.c +++ b/kernel/trace/trace_syscalls.c @@ -433,12 +433,9 @@ static void prof_syscall_enter(struct pt_regs *regs, long id) struct syscall_metadata *sys_data; struct syscall_trace_enter *rec; unsigned long flags; - char *trace_buf; - char *raw_data; int syscall_nr; int rctx; int size; - int cpu; syscall_nr = syscall_get_nr(current, regs); if (!test_bit(syscall_nr, enabled_prof_enter_syscalls)) @@ -457,37 +454,15 @@ static void prof_syscall_enter(struct pt_regs *regs, long id) "profile buffer not large enough")) return; - /* Protect the per cpu buffer, begin the rcu read side */ - local_irq_save(flags); - - rctx = perf_swevent_get_recursion_context(); - if (rctx < 0) - goto end_recursion; - - cpu = smp_processor_id(); - - trace_buf = rcu_dereference(perf_trace_buf); - - if (!trace_buf) - goto end; - - raw_data = per_cpu_ptr(trace_buf, cpu); - - /* zero the dead bytes from align to not leak stack to user */ - *(u64 *)(&raw_data[size - sizeof(u64)]) = 0ULL; + rec = (struct syscall_trace_enter *)ftrace_perf_buf_prepare(size, + sys_data->enter_event->id, &rctx, &flags); + if (!rec) + return; - rec = (struct syscall_trace_enter *) raw_data; - tracing_generic_entry_update(&rec->ent, 0, 0); - rec->ent.type = sys_data->enter_event->id; rec->nr = syscall_nr; syscall_get_arguments(current, regs, 0, sys_data->nb_args, (unsigned long *)&rec->args); - perf_tp_event(sys_data->enter_event->id, 0, 1, rec, size); - -end: - perf_swevent_put_recursion_context(rctx); -end_recursion: - local_irq_restore(flags); + ftrace_perf_buf_submit(rec, size, rctx, 0, 1, flags); } int prof_sysenter_enable(struct ftrace_event_call *call) @@ -531,11 +506,8 @@ static void prof_syscall_exit(struct pt_regs *regs, long ret) struct syscall_trace_exit *rec; unsigned long flags; int syscall_nr; - char *trace_buf; - char *raw_data; int rctx; int size; - int cpu; syscall_nr = syscall_get_nr(current, regs); if (!test_bit(syscall_nr, enabled_prof_exit_syscalls)) @@ -557,38 +529,15 @@ static void prof_syscall_exit(struct pt_regs *regs, long ret) "exit event has grown above profile buffer size")) return; - /* Protect the per cpu buffer, begin the rcu read side */ - local_irq_save(flags); - - rctx = perf_swevent_get_recursion_context(); - if (rctx < 0) - goto end_recursion; - - cpu = smp_processor_id(); - - trace_buf = rcu_dereference(perf_trace_buf); - - if (!trace_buf) - goto end; - - raw_data = per_cpu_ptr(trace_buf, cpu); - - /* zero the dead bytes from align to not leak stack to user */ - *(u64 *)(&raw_data[size - sizeof(u64)]) = 0ULL; - - rec = (struct syscall_trace_exit *)raw_data; + rec = (struct syscall_trace_exit *)ftrace_perf_buf_prepare(size, + sys_data->exit_event->id, &rctx, &flags); + if (!rec) + return; - tracing_generic_entry_update(&rec->ent, 0, 0); - rec->ent.type = sys_data->exit_event->id; rec->nr = syscall_nr; rec->ret = syscall_get_return_value(current, regs); - perf_tp_event(sys_data->exit_event->id, 0, 1, rec, size); - -end: - perf_swevent_put_recursion_context(rctx); -end_recursion: - local_irq_restore(flags); + ftrace_perf_buf_submit(rec, size, rctx, 0, 1, flags); } int prof_sysexit_enable(struct ftrace_event_call *call) -- cgit v1.2.3 From 1e12a4a7a3a78bc9c3aaf3486dde3b8ab1cdf465 Mon Sep 17 00:00:00 2001 From: Xiao Guangrong Date: Thu, 28 Jan 2010 09:34:27 +0800 Subject: tracing/kprobe: Cleanup unused return value of tracing functions The return values of the kprobe's tracing functions are meaningless, lets remove these. Signed-off-by: Xiao Guangrong Acked-by: Masami Hiramatsu Cc: Steven Rostedt Cc: Ingo Molnar Cc: Paul Mackerras Cc: Jason Baron Cc: Peter Zijlstra LKML-Reference: <4B60E9A3.2040505@cn.fujitsu.com> [fweisbec@gmail: whitespace fixes, drop useless void returns in end of functions] Signed-off-by: Frederic Weisbecker --- kernel/trace/trace_kprobe.c | 27 ++++++++++----------------- 1 file changed, 10 insertions(+), 17 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_kprobe.c b/kernel/trace/trace_kprobe.c index 2e28ee36646f..6178abf3637e 100644 --- a/kernel/trace/trace_kprobe.c +++ b/kernel/trace/trace_kprobe.c @@ -942,7 +942,7 @@ static const struct file_operations kprobe_profile_ops = { }; /* Kprobe handler */ -static __kprobes int kprobe_trace_func(struct kprobe *kp, struct pt_regs *regs) +static __kprobes void kprobe_trace_func(struct kprobe *kp, struct pt_regs *regs) { struct trace_probe *tp = container_of(kp, struct trace_probe, rp.kp); struct kprobe_trace_entry *entry; @@ -962,7 +962,7 @@ static __kprobes int kprobe_trace_func(struct kprobe *kp, struct pt_regs *regs) event = trace_current_buffer_lock_reserve(&buffer, call->id, size, irq_flags, pc); if (!event) - return 0; + return; entry = ring_buffer_event_data(event); entry->nargs = tp->nr_args; @@ -972,11 +972,10 @@ static __kprobes int kprobe_trace_func(struct kprobe *kp, struct pt_regs *regs) if (!filter_current_check_discard(buffer, call, entry, event)) trace_nowake_buffer_unlock_commit(buffer, event, irq_flags, pc); - return 0; } /* Kretprobe handler */ -static __kprobes int kretprobe_trace_func(struct kretprobe_instance *ri, +static __kprobes void kretprobe_trace_func(struct kretprobe_instance *ri, struct pt_regs *regs) { struct trace_probe *tp = container_of(ri->rp, struct trace_probe, rp); @@ -995,7 +994,7 @@ static __kprobes int kretprobe_trace_func(struct kretprobe_instance *ri, event = trace_current_buffer_lock_reserve(&buffer, call->id, size, irq_flags, pc); if (!event) - return 0; + return; entry = ring_buffer_event_data(event); entry->nargs = tp->nr_args; @@ -1006,8 +1005,6 @@ static __kprobes int kretprobe_trace_func(struct kretprobe_instance *ri, if (!filter_current_check_discard(buffer, call, entry, event)) trace_nowake_buffer_unlock_commit(buffer, event, irq_flags, pc); - - return 0; } /* Event entry printers */ @@ -1237,7 +1234,7 @@ static int kretprobe_event_show_format(struct ftrace_event_call *call, #ifdef CONFIG_PERF_EVENTS /* Kprobe profile handler */ -static __kprobes int kprobe_profile_func(struct kprobe *kp, +static __kprobes void kprobe_profile_func(struct kprobe *kp, struct pt_regs *regs) { struct trace_probe *tp = container_of(kp, struct trace_probe, rp.kp); @@ -1252,11 +1249,11 @@ static __kprobes int kprobe_profile_func(struct kprobe *kp, size -= sizeof(u32); if (WARN_ONCE(size > FTRACE_MAX_PROFILE_SIZE, "profile buffer not large enough")) - return 0; + return; entry = ftrace_perf_buf_prepare(size, call->id, &rctx, &irq_flags); if (!entry) - return 0; + return; entry->nargs = tp->nr_args; entry->ip = (unsigned long)kp->addr; @@ -1264,12 +1261,10 @@ static __kprobes int kprobe_profile_func(struct kprobe *kp, entry->args[i] = call_fetch(&tp->args[i].fetch, regs); ftrace_perf_buf_submit(entry, size, rctx, entry->ip, 1, irq_flags); - - return 0; } /* Kretprobe profile handler */ -static __kprobes int kretprobe_profile_func(struct kretprobe_instance *ri, +static __kprobes void kretprobe_profile_func(struct kretprobe_instance *ri, struct pt_regs *regs) { struct trace_probe *tp = container_of(ri->rp, struct trace_probe, rp); @@ -1284,11 +1279,11 @@ static __kprobes int kretprobe_profile_func(struct kretprobe_instance *ri, size -= sizeof(u32); if (WARN_ONCE(size > FTRACE_MAX_PROFILE_SIZE, "profile buffer not large enough")) - return 0; + return; entry = ftrace_perf_buf_prepare(size, call->id, &rctx, &irq_flags); if (!entry) - return 0; + return; entry->nargs = tp->nr_args; entry->func = (unsigned long)tp->rp.kp.addr; @@ -1297,8 +1292,6 @@ static __kprobes int kretprobe_profile_func(struct kretprobe_instance *ri, entry->args[i] = call_fetch(&tp->args[i].fetch, regs); ftrace_perf_buf_submit(entry, size, rctx, entry->ret_ip, 1, irq_flags); - - return 0; } static int probe_profile_enable(struct ftrace_event_call *call) -- cgit v1.2.3 From 4f48f8b7fd18c44f8478174f9925cc3c059c6ce4 Mon Sep 17 00:00:00 2001 From: Lai Jiangshan Date: Tue, 2 Feb 2010 15:32:09 +0800 Subject: tracing: Fix circular dead lock in stack trace When we cat /tracing/stack_trace, we may cause circular lock: sys_read() t_start() arch_spin_lock(&max_stack_lock); t_show() seq_printf(), vsnprintf() .... /* they are all trace-able, when they are traced, max_stack_lock may be required again. */ The following script can trigger this circular dead lock very easy: #!/bin/bash echo 1 > /proc/sys/kernel/stack_tracer_enabled mount -t debugfs xxx /mnt > /dev/null 2>&1 ( # make check_stack() zealous to require max_stack_lock for ((; ;)) { echo 1 > /mnt/tracing/stack_max_size } ) & for ((; ;)) { cat /mnt/tracing/stack_trace > /dev/null } To fix this bug, we increase the percpu trace_active before require the lock. Reported-by: Li Zefan Signed-off-by: Lai Jiangshan LKML-Reference: <4B67D4F9.9080905@cn.fujitsu.com> Signed-off-by: Steven Rostedt --- kernel/trace/trace_stack.c | 24 ++++++++++++++++++++++++ 1 file changed, 24 insertions(+) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_stack.c b/kernel/trace/trace_stack.c index 678a5120ee30..f4bc9b27de5f 100644 --- a/kernel/trace/trace_stack.c +++ b/kernel/trace/trace_stack.c @@ -157,6 +157,7 @@ stack_max_size_write(struct file *filp, const char __user *ubuf, unsigned long val, flags; char buf[64]; int ret; + int cpu; if (count >= sizeof(buf)) return -EINVAL; @@ -171,9 +172,20 @@ stack_max_size_write(struct file *filp, const char __user *ubuf, return ret; local_irq_save(flags); + + /* + * In case we trace inside arch_spin_lock() or after (NMI), + * we will cause circular lock, so we also need to increase + * the percpu trace_active here. + */ + cpu = smp_processor_id(); + per_cpu(trace_active, cpu)++; + arch_spin_lock(&max_stack_lock); *ptr = val; arch_spin_unlock(&max_stack_lock); + + per_cpu(trace_active, cpu)--; local_irq_restore(flags); return count; @@ -206,7 +218,13 @@ t_next(struct seq_file *m, void *v, loff_t *pos) static void *t_start(struct seq_file *m, loff_t *pos) { + int cpu; + local_irq_disable(); + + cpu = smp_processor_id(); + per_cpu(trace_active, cpu)++; + arch_spin_lock(&max_stack_lock); if (*pos == 0) @@ -217,7 +235,13 @@ static void *t_start(struct seq_file *m, loff_t *pos) static void t_stop(struct seq_file *m, void *p) { + int cpu; + arch_spin_unlock(&max_stack_lock); + + cpu = smp_processor_id(); + per_cpu(trace_active, cpu)--; + local_irq_enable(); } -- cgit v1.2.3 From 2cfa19780d61740f65790c5bae363b759d7c96fa Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Tue, 2 Feb 2010 16:49:11 -0500 Subject: ftrace/alternatives: Introducing *_text_reserved functions Introducing *_text_reserved functions for checking the text address range is partially reserved or not. This patch provides checking routines for x86 smp alternatives and dynamic ftrace. Since both functions modify fixed pieces of kernel text, they should reserve and protect those from other dynamic text modifier, like kprobes. This will also be extended when introducing other subsystems which modify fixed pieces of kernel text. Dynamic text modifiers should avoid those. Signed-off-by: Masami Hiramatsu Cc: systemtap Cc: DLE Cc: Steven Rostedt Cc: przemyslaw@pawelczyk.it Cc: Frederic Weisbecker Cc: Ananth N Mavinakayanahalli Cc: Jim Keniston Cc: Mathieu Desnoyers Cc: Jason Baron LKML-Reference: <20100202214911.4694.16587.stgit@dhcp-100-2-132.bos.redhat.com> Signed-off-by: Ingo Molnar --- kernel/trace/ftrace.c | 15 +++++++++++++++ 1 file changed, 15 insertions(+) (limited to 'kernel/trace') diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index 1e6640f80454..3d90661a5f40 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -1025,6 +1025,21 @@ static void ftrace_bug(int failed, unsigned long ip) } +/* Return 1 if the address range is reserved for ftrace */ +int ftrace_text_reserved(void *start, void *end) +{ + struct dyn_ftrace *rec; + struct ftrace_page *pg; + + do_for_each_ftrace_rec(pg, rec) { + if (rec->ip <= (unsigned long)end && + rec->ip + MCOUNT_INSN_SIZE > (unsigned long)start) + return 1; + } while_for_each_ftrace_rec(); + return 0; +} + + static int __ftrace_replace_code(struct dyn_ftrace *rec, int enable) { -- cgit v1.2.3 From f24bb999d2b9f2950e5cac5b69bffedf73c24ea4 Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Tue, 2 Feb 2010 16:49:25 -0500 Subject: ftrace: Remove record freezing Remove record freezing. Because kprobes never puts probe on ftrace's mcount call anymore, it doesn't need ftrace to check whether kprobes on it. Signed-off-by: Masami Hiramatsu Cc: systemtap Cc: DLE Cc: Steven Rostedt Cc: przemyslaw@pawelczyk.it Cc: Frederic Weisbecker LKML-Reference: <20100202214925.4694.73469.stgit@dhcp-100-2-132.bos.redhat.com> Signed-off-by: Ingo Molnar --- kernel/trace/ftrace.c | 39 --------------------------------------- 1 file changed, 39 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index 3d90661a5f40..1904797f4a8a 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -22,7 +22,6 @@ #include #include #include -#include #include #include #include @@ -898,36 +897,6 @@ static struct dyn_ftrace *ftrace_free_records; } \ } -#ifdef CONFIG_KPROBES - -static int frozen_record_count; - -static inline void freeze_record(struct dyn_ftrace *rec) -{ - if (!(rec->flags & FTRACE_FL_FROZEN)) { - rec->flags |= FTRACE_FL_FROZEN; - frozen_record_count++; - } -} - -static inline void unfreeze_record(struct dyn_ftrace *rec) -{ - if (rec->flags & FTRACE_FL_FROZEN) { - rec->flags &= ~FTRACE_FL_FROZEN; - frozen_record_count--; - } -} - -static inline int record_frozen(struct dyn_ftrace *rec) -{ - return rec->flags & FTRACE_FL_FROZEN; -} -#else -# define freeze_record(rec) ({ 0; }) -# define unfreeze_record(rec) ({ 0; }) -# define record_frozen(rec) ({ 0; }) -#endif /* CONFIG_KPROBES */ - static void ftrace_free_rec(struct dyn_ftrace *rec) { rec->freelist = ftrace_free_records; @@ -1091,14 +1060,6 @@ static void ftrace_replace_code(int enable) !(rec->flags & FTRACE_FL_CONVERTED)) continue; - /* ignore updates to this record's mcount site */ - if (get_kprobe((void *)rec->ip)) { - freeze_record(rec); - continue; - } else { - unfreeze_record(rec); - } - failed = __ftrace_replace_code(rec, enable); if (failed) { rec->flags |= FTRACE_FL_FAILED; -- cgit v1.2.3 From c41b20e721ea4f6f20f66a66e7f0c3c97a2ca9c2 Mon Sep 17 00:00:00 2001 From: Adam Buchbinder Date: Fri, 11 Dec 2009 16:35:39 -0500 Subject: Fix misspellings of "truly" in comments. Some comments misspell "truly"; this fixes them. No code changes. Signed-off-by: Adam Buchbinder Signed-off-by: Jiri Kosina --- kernel/trace/ring_buffer.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/ring_buffer.c b/kernel/trace/ring_buffer.c index 8c1b2d290718..9ab578f1bb65 100644 --- a/kernel/trace/ring_buffer.c +++ b/kernel/trace/ring_buffer.c @@ -2541,7 +2541,7 @@ EXPORT_SYMBOL_GPL(ring_buffer_record_disable); * @buffer: The ring buffer to enable writes * * Note, multiple disables will need the same number of enables - * to truely enable the writing (much like preempt_disable). + * to truly enable the writing (much like preempt_disable). */ void ring_buffer_record_enable(struct ring_buffer *buffer) { @@ -2577,7 +2577,7 @@ EXPORT_SYMBOL_GPL(ring_buffer_record_disable_cpu); * @cpu: The CPU to enable. * * Note, multiple disables will need the same number of enables - * to truely enable the writing (much like preempt_disable). + * to truly enable the writing (much like preempt_disable). */ void ring_buffer_record_enable_cpu(struct ring_buffer *buffer, int cpu) { -- cgit v1.2.3 From 1537a3638cbf741d3826c1002026cce487a6bee0 Mon Sep 17 00:00:00 2001 From: Daniel Mack Date: Fri, 29 Jan 2010 15:57:49 +0800 Subject: tree-wide: fix 'lenght' typo in comments and code Some misspelled occurences of 'octet' and some comments were also fixed as I was on it. Signed-off-by: Daniel Mack Cc: Jiri Kosina Cc: Joe Perches Cc: Junio C Hamano Signed-off-by: Jiri Kosina --- kernel/trace/trace.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace.h b/kernel/trace/trace.h index 4df6a77eb196..e4b32c8aa85f 100644 --- a/kernel/trace/trace.h +++ b/kernel/trace/trace.h @@ -549,7 +549,7 @@ static inline int ftrace_trace_task(struct task_struct *task) * struct trace_parser - servers for reading the user input separated by spaces * @cont: set if the input is not complete - no final space char was found * @buffer: holds the parsed user input - * @idx: user input lenght + * @idx: user input length * @size: buffer size */ struct trace_parser { -- cgit v1.2.3 From ede55c9d78101fef0d8e620940a5163f14b02f29 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Wed, 27 Jan 2010 11:25:54 -0500 Subject: tracing: Add correct/incorrect to sort keys for branch annotation output The branch annotation is a bit difficult to see the worst offenders because it only sorts by percentage: correct incorrect % Function File Line ------- --------- - -------- ---- ---- 0 163 100 qdisc_restart sch_generic.c 179 0 163 100 pfifo_fast_dequeue sch_generic.c 447 0 4 100 pskb_trim_rcsum skbuff.h 1689 0 4 100 llc_rcv llc_input.c 170 0 18 100 psmouse_interrupt psmouse-base.c 304 0 3 100 atkbd_interrupt atkbd.c 389 0 5 100 usb_alloc_dev usb.c 437 0 11 100 vsscanf vsprintf.c 1897 0 2 100 IS_ERR err.h 34 0 23 100 __rmqueue_fallback page_alloc.c 865 0 4 100 probe_wakeup_sched_switch trace_sched_wakeup.c 142 0 3 100 move_masked_irq migration.c 11 Adding the incorrect and correct values as sort keys makes this file a bit more informative: correct incorrect % Function File Line ------- --------- - -------- ---- ---- 0 366541 100 audit_syscall_entry auditsc.c 1637 0 366538 100 audit_syscall_exit auditsc.c 1685 0 115839 100 sched_info_switch sched_stats.h 269 0 74567 100 sched_info_queued sched_stats.h 222 0 66578 100 sched_info_dequeued sched_stats.h 177 0 15113 100 trace_workqueue_insertion workqueue.h 38 0 15107 100 trace_workqueue_execution workqueue.h 45 0 3622 100 syscall_trace_leave ptrace.c 1772 0 2750 100 sched_move_task sched.c 10100 0 2750 100 sched_move_task sched.c 10110 0 1815 100 pre_schedule_rt sched_rt.c 1462 0 837 100 audit_alloc auditsc.c 879 0 814 100 tcp_mss_split_point tcp_output.c 1302 Signed-off-by: Steven Rostedt --- kernel/trace/trace_branch.c | 19 +++++++++++++++++-- 1 file changed, 17 insertions(+), 2 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_branch.c b/kernel/trace/trace_branch.c index 4a194f08f88c..b9bc4d470177 100644 --- a/kernel/trace/trace_branch.c +++ b/kernel/trace/trace_branch.c @@ -307,8 +307,23 @@ static int annotated_branch_stat_cmp(void *p1, void *p2) return -1; if (percent_a > percent_b) return 1; - else - return 0; + + if (a->incorrect < b->incorrect) + return -1; + if (a->incorrect > b->incorrect) + return 1; + + /* + * Since the above shows worse (incorrect) cases + * first, we continue that by showing best (correct) + * cases last. + */ + if (a->correct > b->correct) + return -1; + if (a->correct < b->correct) + return 1; + + return 0; } static struct tracer_stat annotated_branch_stats = { -- cgit v1.2.3 From c7c6b1fe9f942c1a30585ec2210a09dfff238506 Mon Sep 17 00:00:00 2001 From: Li Zefan Date: Wed, 10 Feb 2010 15:43:04 +0800 Subject: ftrace: Allow to remove a single function from function graph filter I don't see why we can only clear all functions from the filter. After patching: # echo sys_open > set_graph_function # echo sys_close >> set_graph_function # cat set_graph_function sys_open sys_close # echo '!sys_close' >> set_graph_function # cat set_graph_function sys_open Signed-off-by: Li Zefan LKML-Reference: <4B726388.2000408@cn.fujitsu.com> Signed-off-by: Steven Rostedt --- kernel/trace/ftrace.c | 51 +++++++++++++++++++++++++++++++-------------------- kernel/trace/trace.h | 3 ++- 2 files changed, 33 insertions(+), 21 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index 7968762c8167..43bec993c864 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -2426,6 +2426,7 @@ static const struct file_operations ftrace_notrace_fops = { static DEFINE_MUTEX(graph_lock); int ftrace_graph_count; +int ftrace_graph_filter_enabled; unsigned long ftrace_graph_funcs[FTRACE_GRAPH_MAX_FUNCS] __read_mostly; static void * @@ -2448,7 +2449,7 @@ static void *g_start(struct seq_file *m, loff_t *pos) mutex_lock(&graph_lock); /* Nothing, tell g_show to print all functions are enabled */ - if (!ftrace_graph_count && !*pos) + if (!ftrace_graph_filter_enabled && !*pos) return (void *)1; return __g_next(m, pos); @@ -2494,6 +2495,7 @@ ftrace_graph_open(struct inode *inode, struct file *file) mutex_lock(&graph_lock); if ((file->f_mode & FMODE_WRITE) && (file->f_flags & O_TRUNC)) { + ftrace_graph_filter_enabled = 0; ftrace_graph_count = 0; memset(ftrace_graph_funcs, 0, sizeof(ftrace_graph_funcs)); } @@ -2519,7 +2521,7 @@ ftrace_set_func(unsigned long *array, int *idx, char *buffer) struct dyn_ftrace *rec; struct ftrace_page *pg; int search_len; - int found = 0; + int fail = 1; int type, not; char *search; bool exists; @@ -2530,37 +2532,51 @@ ftrace_set_func(unsigned long *array, int *idx, char *buffer) /* decode regex */ type = filter_parse_regex(buffer, strlen(buffer), &search, ¬); - if (not) - return -EINVAL; + if (!not && *idx >= FTRACE_GRAPH_MAX_FUNCS) + return -EBUSY; search_len = strlen(search); mutex_lock(&ftrace_lock); do_for_each_ftrace_rec(pg, rec) { - if (*idx >= FTRACE_GRAPH_MAX_FUNCS) - break; - if (rec->flags & (FTRACE_FL_FAILED | FTRACE_FL_FREE)) continue; if (ftrace_match_record(rec, search, search_len, type)) { - /* ensure it is not already in the array */ + /* if it is in the array */ exists = false; - for (i = 0; i < *idx; i++) + for (i = 0; i < *idx; i++) { if (array[i] == rec->ip) { exists = true; break; } - if (!exists) - array[(*idx)++] = rec->ip; - found = 1; + } + + if (!not) { + fail = 0; + if (!exists) { + array[(*idx)++] = rec->ip; + if (*idx >= FTRACE_GRAPH_MAX_FUNCS) + goto out; + } + } else { + if (exists) { + array[i] = array[--(*idx)]; + array[*idx] = 0; + fail = 0; + } + } } } while_for_each_ftrace_rec(); - +out: mutex_unlock(&ftrace_lock); - return found ? 0 : -EINVAL; + if (fail) + return -EINVAL; + + ftrace_graph_filter_enabled = 1; + return 0; } static ssize_t @@ -2570,16 +2586,11 @@ ftrace_graph_write(struct file *file, const char __user *ubuf, struct trace_parser parser; ssize_t read, ret; - if (!cnt || cnt < 0) + if (!cnt) return 0; mutex_lock(&graph_lock); - if (ftrace_graph_count >= FTRACE_GRAPH_MAX_FUNCS) { - ret = -EBUSY; - goto out_unlock; - } - if (trace_parser_get_init(&parser, FTRACE_BUFF_MAX)) { ret = -ENOMEM; goto out_unlock; diff --git a/kernel/trace/trace.h b/kernel/trace/trace.h index ce077fbbf552..b477fce41edf 100644 --- a/kernel/trace/trace.h +++ b/kernel/trace/trace.h @@ -497,6 +497,7 @@ trace_print_graph_duration(unsigned long long duration, struct trace_seq *s); #ifdef CONFIG_DYNAMIC_FTRACE /* TODO: make this variable */ #define FTRACE_GRAPH_MAX_FUNCS 32 +extern int ftrace_graph_filter_enabled; extern int ftrace_graph_count; extern unsigned long ftrace_graph_funcs[FTRACE_GRAPH_MAX_FUNCS]; @@ -504,7 +505,7 @@ static inline int ftrace_graph_addr(unsigned long addr) { int i; - if (!ftrace_graph_count) + if (!ftrace_graph_filter_enabled) return 1; for (i = 0; i < ftrace_graph_count; i++) { -- cgit v1.2.3 From a9bb18f36c8056f0712fb28c52c0f85d98438dfb Mon Sep 17 00:00:00 2001 From: Heiko Carstens Date: Wed, 10 Feb 2010 17:23:47 +0100 Subject: tracing/kprobes: Fix probe parsing Trying to add a probe like: echo p:myprobe 0x10000 > /sys/kernel/debug/tracing/kprobe_events will fail since the wrong pointer is passed to strict_strtoul when trying to convert the address to an unsigned long. Signed-off-by: Heiko Carstens Acked-by: Masami Hiramatsu Cc: Frederic Weisbecker Cc: Steven Rostedt LKML-Reference: <20100210162346.GA6933@osiris.boeblingen.de.ibm.com> Signed-off-by: Ingo Molnar --- kernel/trace/trace_kprobe.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_kprobe.c b/kernel/trace/trace_kprobe.c index 6ea90c0e2c96..50b1b8239806 100644 --- a/kernel/trace/trace_kprobe.c +++ b/kernel/trace/trace_kprobe.c @@ -689,7 +689,7 @@ static int create_trace_probe(int argc, char **argv) return -EINVAL; } /* an address specified */ - ret = strict_strtoul(&argv[0][2], 0, (unsigned long *)&addr); + ret = strict_strtoul(&argv[1][0], 0, (unsigned long *)&addr); if (ret) { pr_info("Failed to parse address.\n"); return ret; -- cgit v1.2.3 From e7b8e675d9c71b868b66f62f725a948047514719 Mon Sep 17 00:00:00 2001 From: Mike Frysinger Date: Tue, 26 Jan 2010 04:40:03 -0500 Subject: tracing: Unify arch_syscall_addr() implementations Most implementations of arch_syscall_addr() are the same, so create a default version in common code and move the one piece that differs (the syscall table) to asm/syscall.h. New arch ports don't have to waste time copying & pasting this simple function. The s390/sparc versions need to be different, so document why. Signed-off-by: Mike Frysinger Acked-by: David S. Miller Acked-by: Paul Mundt Acked-by: Heiko Carstens Cc: Steven Rostedt LKML-Reference: <1264498803-17278-1-git-send-email-vapier@gentoo.org> Signed-off-by: Frederic Weisbecker --- kernel/trace/trace_syscalls.c | 5 +++++ 1 file changed, 5 insertions(+) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_syscalls.c b/kernel/trace/trace_syscalls.c index 49cea70fbf6d..ecf00782b46c 100644 --- a/kernel/trace/trace_syscalls.c +++ b/kernel/trace/trace_syscalls.c @@ -394,6 +394,11 @@ int init_syscall_trace(struct ftrace_event_call *call) return id; } +unsigned long __init arch_syscall_addr(int nr) +{ + return (unsigned long)sys_call_table[nr]; +} + int __init init_ftrace_syscalls(void) { struct syscall_metadata *meta; -- cgit v1.2.3 From f850c30c8b426ba1688cb63b1a3e534eed03a138 Mon Sep 17 00:00:00 2001 From: Heiko Carstens Date: Wed, 10 Feb 2010 17:25:17 +0100 Subject: tracing/kprobes: Make Kconfig dependencies generic KPROBES_EVENT actually depends on the regs and stack access API (b1cf540f) and not on x86. So introduce a new config option which architectures can select if they have the API implemented and switch x86. Signed-off-by: Heiko Carstens Acked-by: Masami Hiramatsu Cc: Ingo Molnar Cc: Martin Schwidefsky LKML-Reference: <20100210162517.GB6933@osiris.boeblingen.de.ibm.com> Signed-off-by: Frederic Weisbecker --- kernel/trace/Kconfig | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel/trace') diff --git a/kernel/trace/Kconfig b/kernel/trace/Kconfig index 6c22d8a2f289..40fef552f012 100644 --- a/kernel/trace/Kconfig +++ b/kernel/trace/Kconfig @@ -451,7 +451,7 @@ config BLK_DEV_IO_TRACE config KPROBE_EVENT depends on KPROBES - depends on X86 + depends on HAVE_REGS_AND_STACK_ACCESS_API bool "Enable kprobes-based dynamic events" select TRACING default y -- cgit v1.2.3 From 86c38a31aa7f2dd6e74a262710bf8ebf7455acc5 Mon Sep 17 00:00:00 2001 From: Jeff Mahoney Date: Wed, 24 Feb 2010 13:59:23 -0500 Subject: tracing: Fix ftrace_event_call alignment for use with gcc 4.5 GCC 4.5 introduces behavior that forces the alignment of structures to use the largest possible value. The default value is 32 bytes, so if some structures are defined with a 4-byte alignment and others aren't declared with an alignment constraint at all - it will align at 32-bytes. For things like the ftrace events, this results in a non-standard array. When initializing the ftrace subsystem, we traverse the _ftrace_events section and call the initialization callback for each event. When the structures are misaligned, we could be treating another part of the structure (or the zeroed out space between them) as a function pointer. This patch forces the alignment for all the ftrace_event_call structures to 4 bytes. Without this patch, the kernel fails to boot very early when built with gcc 4.5. It's trivial to check the alignment of the members of the array, so it might be worthwhile to add something to the build system to do that automatically. Unfortunately, that only covers this case. I've asked one of the gcc developers about adding a warning when this condition is seen. Cc: stable@kernel.org Signed-off-by: Jeff Mahoney LKML-Reference: <4B85770B.6010901@suse.com> Signed-off-by: Steven Rostedt --- kernel/trace/trace.h | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace.h b/kernel/trace/trace.h index b477fce41edf..fd05bcaf91b0 100644 --- a/kernel/trace/trace.h +++ b/kernel/trace/trace.h @@ -792,7 +792,8 @@ extern const char *__stop___trace_bprintk_fmt[]; #undef FTRACE_ENTRY #define FTRACE_ENTRY(call, struct_name, id, tstruct, print) \ - extern struct ftrace_event_call event_##call; + extern struct ftrace_event_call \ + __attribute__((__aligned__(4))) event_##call; #undef FTRACE_ENTRY_DUP #define FTRACE_ENTRY_DUP(call, struct_name, id, tstruct, print) \ FTRACE_ENTRY(call, struct_name, id, PARAMS(tstruct), PARAMS(print)) -- cgit v1.2.3 From 1ab83a89411556c4d5914dcf0d5da190178ae7db Mon Sep 17 00:00:00 2001 From: Li Zefan Date: Wed, 24 Feb 2010 15:28:14 +0800 Subject: tracing: Remove CONFIG_TRACE_POWER from kernel config The power tracer has been converted to power trace events. Acked-by: Frederic Weisbecker Signed-off-by: Li Zefan LKML-Reference: <4B84D50E.4070806@cn.fujitsu.com> Signed-off-by: Steven Rostedt --- kernel/trace/Kconfig | 9 --------- 1 file changed, 9 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/Kconfig b/kernel/trace/Kconfig index 6c22d8a2f289..ca2d3a8778b6 100644 --- a/kernel/trace/Kconfig +++ b/kernel/trace/Kconfig @@ -330,15 +330,6 @@ config BRANCH_TRACER Say N if unsure. -config POWER_TRACER - bool "Trace power consumption behavior" - depends on X86 - select GENERIC_TRACER - help - This tracer helps developers to analyze and optimize the kernel's - power management decisions, specifically the C-state and P-state - behavior. - config KSYM_TRACER bool "Trace read and write access on kernel memory locations" depends on HAVE_HW_BREAKPOINT -- cgit v1.2.3 From 6574658b3bc7c408581629de5efb809f125cce8c Mon Sep 17 00:00:00 2001 From: Wenji Huang Date: Wed, 24 Feb 2010 15:40:22 +0800 Subject: tracing: Fix typo in prof_sysexit_enable() Signed-off-by: Wenji Huang LKML-Reference: <1266997226-6833-1-git-send-email-wenji.huang@oracle.com> Signed-off-by: Steven Rostedt --- kernel/trace/trace_syscalls.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_syscalls.c b/kernel/trace/trace_syscalls.c index 49cea70fbf6d..8cdda95da81a 100644 --- a/kernel/trace/trace_syscalls.c +++ b/kernel/trace/trace_syscalls.c @@ -603,7 +603,7 @@ int prof_sysexit_enable(struct ftrace_event_call *call) ret = register_trace_sys_exit(prof_syscall_exit); if (ret) { pr_info("event trace: Could not activate" - "syscall entry trace point"); + "syscall exit trace point"); } else { set_bit(num, enabled_prof_exit_syscalls); sys_prof_refcount_exit++; -- cgit v1.2.3 From a5efd925115cbc1f90195dca9a25f7b8daa10c37 Mon Sep 17 00:00:00 2001 From: Wenji Huang Date: Wed, 24 Feb 2010 15:40:23 +0800 Subject: tracing: Fix typo of info text in trace_kprobe.c Signed-off-by: Wenji Huang LKML-Reference: <1266997226-6833-2-git-send-email-wenji.huang@oracle.com> Signed-off-by: Steven Rostedt --- kernel/trace/trace_kprobe.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_kprobe.c b/kernel/trace/trace_kprobe.c index c99029916c76..8d4bd16d6f70 100644 --- a/kernel/trace/trace_kprobe.c +++ b/kernel/trace/trace_kprobe.c @@ -651,12 +651,12 @@ static int create_trace_probe(int argc, char **argv) event = strchr(group, '/') + 1; event[-1] = '\0'; if (strlen(group) == 0) { - pr_info("Group name is not specifiled\n"); + pr_info("Group name is not specified\n"); return -EINVAL; } } if (strlen(event) == 0) { - pr_info("Event name is not specifiled\n"); + pr_info("Event name is not specified\n"); return -EINVAL; } } -- cgit v1.2.3 From c85f3a91f84d5a85f179c2504bb7a39370c82b41 Mon Sep 17 00:00:00 2001 From: Wenji Huang Date: Wed, 24 Feb 2010 15:40:24 +0800 Subject: tracing: Remove unnecessary variable in print_graph_return The "cpu" variable is declared at the start of the function and also within a branch, with the exact same initialization. Remove the local variable of the same name in the branch. Signed-off-by: Wenji Huang LKML-Reference: <1266997226-6833-3-git-send-email-wenji.huang@oracle.com> Signed-off-by: Steven Rostedt --- kernel/trace/trace_functions_graph.c | 1 - 1 file changed, 1 deletion(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_functions_graph.c b/kernel/trace/trace_functions_graph.c index 616b135c9eb9..112561df2a0a 100644 --- a/kernel/trace/trace_functions_graph.c +++ b/kernel/trace/trace_functions_graph.c @@ -855,7 +855,6 @@ print_graph_return(struct ftrace_graph_ret *trace, struct trace_seq *s, int i; if (data) { - int cpu = iter->cpu; int *depth = &(per_cpu_ptr(data->cpu_data, cpu)->depth); /* -- cgit v1.2.3 From 7b60997f73865b019e595720185c85285ca3df9a Mon Sep 17 00:00:00 2001 From: Wenji Huang Date: Wed, 24 Feb 2010 15:40:26 +0800 Subject: tracing: Simplify memory recycle of trace_define_field Discard freeing field->type since it is not necessary. Reviewed-by: Li Zefan Signed-off-by: Wenji Huang LKML-Reference: <1266997226-6833-5-git-send-email-wenji.huang@oracle.com> Signed-off-by: Steven Rostedt --- kernel/trace/trace_events.c | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_events.c b/kernel/trace/trace_events.c index c2a3077b7353..3f972ad98d04 100644 --- a/kernel/trace/trace_events.c +++ b/kernel/trace/trace_events.c @@ -60,10 +60,8 @@ int trace_define_field(struct ftrace_event_call *call, const char *type, return 0; err: - if (field) { + if (field) kfree(field->name); - kfree(field->type); - } kfree(field); return -ENOMEM; -- cgit v1.2.3 From f1c7f517a5dc23bce07efa5ed55e2c074ed9d4ba Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Fri, 26 Feb 2010 17:08:16 -0500 Subject: ftrace: Add function names to dangling } in function graph tracer The function graph tracer is currently the most invasive tracer in the ftrace family. It can easily overflow the buffer even with 10megs per CPU. This means that events can often be lost. On start up, or after events are lost, if the function return is recorded but the function enter was lost, all we get to see is the exiting '}'. Here is how a typical trace output starts: [tracing] cat trace # tracer: function_graph # # CPU DURATION FUNCTION CALLS # | | | | | | | 0) + 91.897 us | } 0) ! 567.961 us | } 0) <========== | 0) ! 579.083 us | _raw_spin_lock_irqsave(); 0) 4.694 us | _raw_spin_unlock_irqrestore(); 0) ! 594.862 us | } 0) ! 603.361 us | } 0) ! 613.574 us | } 0) ! 623.554 us | } 0) 3.653 us | fget_light(); 0) | sock_poll() { There are a series of '}' with no matching "func() {". There's no information to what functions these ending brackets belong to. This patch adds a stack on the per cpu structure used in outputting the function graph tracer to keep track of what function was outputted. Then on a function exit event, it checks the depth to see if the function exit has a matching entry event. If it does, then it only prints the '}', otherwise it adds the function name after the '}'. This allows function exit events to show what function they belong to at trace output startup, when the entry was lost due to ring buffer overflow, or even after a new task is scheduled in. Here is what the above trace will look like after this patch: [tracing] cat trace # tracer: function_graph # # CPU DURATION FUNCTION CALLS # | | | | | | | 0) + 91.897 us | } (irq_exit) 0) ! 567.961 us | } (smp_apic_timer_interrupt) 0) <========== | 0) ! 579.083 us | _raw_spin_lock_irqsave(); 0) 4.694 us | _raw_spin_unlock_irqrestore(); 0) ! 594.862 us | } (add_wait_queue) 0) ! 603.361 us | } (__pollwait) 0) ! 613.574 us | } (tcp_poll) 0) ! 623.554 us | } (sock_poll) 0) 3.653 us | fget_light(); 0) | sock_poll() { Signed-off-by: Steven Rostedt --- kernel/trace/trace_functions_graph.c | 52 +++++++++++++++++++++++++++++------- 1 file changed, 43 insertions(+), 9 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_functions_graph.c b/kernel/trace/trace_functions_graph.c index 112561df2a0a..e998a824e9db 100644 --- a/kernel/trace/trace_functions_graph.c +++ b/kernel/trace/trace_functions_graph.c @@ -18,6 +18,7 @@ struct fgraph_cpu_data { pid_t last_pid; int depth; int ignore; + unsigned long enter_funcs[FTRACE_RETFUNC_DEPTH]; }; struct fgraph_data { @@ -670,15 +671,21 @@ print_graph_entry_leaf(struct trace_iterator *iter, duration = graph_ret->rettime - graph_ret->calltime; if (data) { + struct fgraph_cpu_data *cpu_data; int cpu = iter->cpu; - int *depth = &(per_cpu_ptr(data->cpu_data, cpu)->depth); + + cpu_data = per_cpu_ptr(data->cpu_data, cpu); /* * Comments display at + 1 to depth. Since * this is a leaf function, keep the comments * equal to this depth. */ - *depth = call->depth - 1; + cpu_data->depth = call->depth - 1; + + /* No need to keep this function around for this depth */ + if (call->depth < FTRACE_RETFUNC_DEPTH) + cpu_data->enter_funcs[call->depth] = 0; } /* Overhead */ @@ -718,10 +725,15 @@ print_graph_entry_nested(struct trace_iterator *iter, int i; if (data) { + struct fgraph_cpu_data *cpu_data; int cpu = iter->cpu; - int *depth = &(per_cpu_ptr(data->cpu_data, cpu)->depth); - *depth = call->depth; + cpu_data = per_cpu_ptr(data->cpu_data, cpu); + cpu_data->depth = call->depth; + + /* Save this function pointer to see if the exit matches */ + if (call->depth < FTRACE_RETFUNC_DEPTH) + cpu_data->enter_funcs[call->depth] = call->func; } /* No overhead */ @@ -851,18 +863,28 @@ print_graph_return(struct ftrace_graph_ret *trace, struct trace_seq *s, struct fgraph_data *data = iter->private; pid_t pid = ent->pid; int cpu = iter->cpu; + int func_match = 1; int ret; int i; if (data) { - int *depth = &(per_cpu_ptr(data->cpu_data, cpu)->depth); + struct fgraph_cpu_data *cpu_data; + int cpu = iter->cpu; + + cpu_data = per_cpu_ptr(data->cpu_data, cpu); /* * Comments display at + 1 to depth. This is the * return from a function, we now want the comments * to display at the same level of the bracket. */ - *depth = trace->depth - 1; + cpu_data->depth = trace->depth - 1; + + if (trace->depth < FTRACE_RETFUNC_DEPTH) { + if (cpu_data->enter_funcs[trace->depth] != trace->func) + func_match = 0; + cpu_data->enter_funcs[trace->depth] = 0; + } } if (print_graph_prologue(iter, s, 0, 0)) @@ -887,9 +909,21 @@ print_graph_return(struct ftrace_graph_ret *trace, struct trace_seq *s, return TRACE_TYPE_PARTIAL_LINE; } - ret = trace_seq_printf(s, "}\n"); - if (!ret) - return TRACE_TYPE_PARTIAL_LINE; + /* + * If the return function does not have a matching entry, + * then the entry was lost. Instead of just printing + * the '}' and letting the user guess what function this + * belongs to, write out the function name. + */ + if (func_match) { + ret = trace_seq_printf(s, "}\n"); + if (!ret) + return TRACE_TYPE_PARTIAL_LINE; + } else { + ret = trace_seq_printf(s, "} (%ps)\n", (void *)trace->func); + if (!ret) + return TRACE_TYPE_PARTIAL_LINE; + } /* Overrun */ if (tracer_flags.val & TRACE_GRAPH_PRINT_OVERRUN) { -- cgit v1.2.3 From ae1f30384baef4056438d81b305a6a5199b0d16c Mon Sep 17 00:00:00 2001 From: Frederic Weisbecker Date: Sun, 28 Feb 2010 19:42:38 +0100 Subject: tracing: Include irqflags headers from trace clock trace_clock.c includes spinlock.h, which ends up including asm/system.h, which in turn includes linux/irqflags.h in x86. So the definition of raw_local_irq_save is luckily covered there, but this is not the case in parisc: tip/kernel/trace/trace_clock.c:86: error: implicit declaration of function 'raw_local_irq_save' tip/kernel/trace/trace_clock.c:112: error: implicit declaration of function 'raw_local_irq_restore' We need to include linux/irqflags.h directly from trace_clock.c to avoid such build error. Signed-off-by: Frederic Weisbecker Cc: Steven Rostedt Cc: Robert Richter Cc: Peter Zijlstra Signed-off-by: Ingo Molnar --- kernel/trace/trace_clock.c | 1 + 1 file changed, 1 insertion(+) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_clock.c b/kernel/trace/trace_clock.c index 84a3a7ba072a..6fbfb8f417b9 100644 --- a/kernel/trace/trace_clock.c +++ b/kernel/trace/trace_clock.c @@ -13,6 +13,7 @@ * Tracer plugins will chose a default from these clocks. */ #include +#include #include #include #include -- cgit v1.2.3 From 9a8c28c8311e30ba97499447d5a11662f5aea094 Mon Sep 17 00:00:00 2001 From: Dmitry Monakhov Date: Sat, 27 Feb 2010 20:35:13 +0300 Subject: blktrace: perform cleanup after setup error Currently even if BLKTRACESETUP ioctl has failed user must call BLKTRACETEARDOWN to be shure what all staff was cleaned, which is contr-intuitive. Let's setup ioctl make necessery cleanup by it self. Signed-off-by: Dmitry Monakhov Signed-off-by: Jens Axboe --- kernel/trace/blktrace.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/blktrace.c b/kernel/trace/blktrace.c index d9d6206e0b14..07f945a99430 100644 --- a/kernel/trace/blktrace.c +++ b/kernel/trace/blktrace.c @@ -540,9 +540,10 @@ int blk_trace_setup(struct request_queue *q, char *name, dev_t dev, if (ret) return ret; - if (copy_to_user(arg, &buts, sizeof(buts))) + if (copy_to_user(arg, &buts, sizeof(buts))) { + blk_trace_remove(q); return -EFAULT; - + } return 0; } EXPORT_SYMBOL_GPL(blk_trace_setup); -- cgit v1.2.3 From ac91d85456372a90af5b85eb6620fd2efb1e431b Mon Sep 17 00:00:00 2001 From: Lai Jiangshan Date: Tue, 2 Mar 2010 17:54:50 +0800 Subject: tracing: Fix warning in s_next of trace file ops This warning in s_next() can be triggered by lseek(): [] ? s_next+0x77/0x80 [] warn_slowpath_common+0x81/0xa0 [] ? s_next+0x77/0x80 [] warn_slowpath_null+0x1a/0x20 [] s_next+0x77/0x80 [] traverse+0x117/0x200 [] seq_lseek+0xa3/0x120 [] ? seq_lseek+0x0/0x120 [] vfs_llseek+0x41/0x50 [] sys_llseek+0x66/0xa0 [] sysenter_do_call+0x12/0x26 The iterator "leftover" variable is zeroed in the opening of the trace file. But lseek can call s_start() which will call s_next() without reseting the "leftover" variable back to zero, which might trigger the WARN_ON_ONCE(iter->leftover) that is in s_next(). Cc: stable@kernel.org Signed-off-by: Lai Jiangshan LKML-Reference: <4B8CE06A.9090207@cn.fujitsu.com> Signed-off-by: Steven Rostedt --- kernel/trace/trace.c | 1 + 1 file changed, 1 insertion(+) (limited to 'kernel/trace') diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index 032c57ca6502..5edf410bc540 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -1703,6 +1703,7 @@ static void *s_start(struct seq_file *m, loff_t *pos) ftrace_enable_cpu(); + iter->leftover = 0; for (p = iter; p && l < *pos; p = s_next(m, p, &l)) ; -- cgit v1.2.3 From 8d53dd546f36073e0d29b0cfc24c665db301e3e7 Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Wed, 3 Mar 2010 17:50:18 -0800 Subject: rcu, ftrace: Fix RCU lockdep splat in ftrace_perf_buf_prepare() Change the pair of rcu_dereference() calls in ftrace_perf_buf_prepare() to rcu_dereference_sched(). Signed-off-by: Paul E. McKenney Acked-by: Frederic Weisbecker Cc: Steven Rostedt Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: mathieu.desnoyers@polymtl.ca Cc: josh@joshtriplett.org Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: Valdis.Kletnieks@vt.edu Cc: dhowells@redhat.com Cc: Frederic Weisbecker LKML-Reference: <1267667418-32233-3-git-send-email-paulmck@linux.vnet.ibm.com> Signed-off-by: Ingo Molnar --- kernel/trace/trace_event_profile.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_event_profile.c b/kernel/trace/trace_event_profile.c index f0d693005075..c1cc3ab633de 100644 --- a/kernel/trace/trace_event_profile.c +++ b/kernel/trace/trace_event_profile.c @@ -138,9 +138,9 @@ __kprobes void *ftrace_perf_buf_prepare(int size, unsigned short type, cpu = smp_processor_id(); if (in_nmi()) - trace_buf = rcu_dereference(perf_trace_buf_nmi); + trace_buf = rcu_dereference_sched(perf_trace_buf_nmi); else - trace_buf = rcu_dereference(perf_trace_buf); + trace_buf = rcu_dereference_sched(perf_trace_buf); if (!trace_buf) goto err; -- cgit v1.2.3 From 801c29fd1fdeb84f60241beb445ff5db154450ae Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Fri, 5 Mar 2010 20:02:19 -0500 Subject: function-graph: Fix unused reference to ftrace_set_func() The declaration of ftrace_set_func() is at the start of the ftrace.c file and wrapped with a #ifdef CONFIG_FUNCTION_GRAPH condition. If function graph tracing is enabled but CONFIG_DYNAMIC_FTRACE is not, a warning about that function being declared static and unused is given. This really should have been placed within the CONFIG_FUNCTION_GRAPH condition that uses ftrace_set_func(). Moving the declaration down fixes the warning and makes the code cleaner. Reported-by: Peter Zijlstra Signed-off-by: Steven Rostedt --- kernel/trace/ftrace.c | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index d996353473fd..d0407c9f368c 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -85,10 +85,6 @@ ftrace_func_t ftrace_trace_function __read_mostly = ftrace_stub; ftrace_func_t __ftrace_trace_function __read_mostly = ftrace_stub; ftrace_func_t ftrace_pid_function __read_mostly = ftrace_stub; -#ifdef CONFIG_FUNCTION_GRAPH_TRACER -static int ftrace_set_func(unsigned long *array, int *idx, char *buffer); -#endif - static void ftrace_list_func(unsigned long ip, unsigned long parent_ip) { struct ftrace_ops *op = ftrace_list; @@ -2300,6 +2296,8 @@ __setup("ftrace_filter=", set_ftrace_filter); #ifdef CONFIG_FUNCTION_GRAPH_TRACER static char ftrace_graph_buf[FTRACE_FILTER_SIZE] __initdata; +static int ftrace_set_func(unsigned long *array, int *idx, char *buffer); + static int __init set_graph_function(char *str) { strlcpy(ftrace_graph_buf, str, FTRACE_FILTER_SIZE); -- cgit v1.2.3 From a094fe04c751698a18c3a0d376a3bdb117f1e0d8 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Fri, 5 Mar 2010 20:08:58 -0500 Subject: function-graph: Use comment notation for func names of dangling '}' When a '}' does not have a matching function start, the name is printed within parenthesis. But this makes it confusing between ending '}' and function starts. This patch makes the function name appear in C comment notation. Old view: 3) 1.281 us | } (might_fault) 3) 3.620 us | } (filldir) 3) 5.251 us | } (call_filldir) 3) | call_filldir() { 3) | filldir() { New view: 3) 1.281 us | } /* might_fault */ 3) 3.620 us | } /* filldir */ 3) 5.251 us | } /* call_filldir */ 3) | call_filldir() { 3) | filldir() { Requested-by: Ingo Molnar Signed-off-by: Steven Rostedt --- kernel/trace/trace_functions_graph.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_functions_graph.c b/kernel/trace/trace_functions_graph.c index e998a824e9db..7b1f24618d97 100644 --- a/kernel/trace/trace_functions_graph.c +++ b/kernel/trace/trace_functions_graph.c @@ -920,7 +920,7 @@ print_graph_return(struct ftrace_graph_ret *trace, struct trace_seq *s, if (!ret) return TRACE_TYPE_PARTIAL_LINE; } else { - ret = trace_seq_printf(s, "} (%ps)\n", (void *)trace->func); + ret = trace_seq_printf(s, "} /* %ps */\n", (void *)trace->func); if (!ret) return TRACE_TYPE_PARTIAL_LINE; } -- cgit v1.2.3 From 1acaa1b2d9b5904c9cce06122990a2d71046ce16 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Fri, 5 Mar 2010 18:23:50 -0300 Subject: tracing: Update the comm field in the right variable in update_max_tr The latency output showed: # | task: -3 (uid:0 nice:0 policy:1 rt_prio:99) The comm is missing in the "task:" and it looks like a minus 3 is the output. The correct display should be: # | task: migration/0-3 (uid:0 nice:0 policy:1 rt_prio:99) The problem is that the comm is being stored in the wrong data structure. The max_tr.data[cpu] is what stores the comm, not the tr->data[cpu]. Before this patch the max_tr.data[cpu]->comm was zeroed and the /debug/trace ended up showing just the '-' sign followed by the pid. Also remove a needless initialization of max_data. Signed-off-by: Arnaldo Carvalho de Melo LKML-Reference: <1267824230-23861-1-git-send-email-acme@infradead.org> Signed-off-by: Steven Rostedt --- kernel/trace/trace.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index 032c57ca6502..6efd5cb3c252 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -592,7 +592,7 @@ static void __update_max_tr(struct trace_array *tr, struct task_struct *tsk, int cpu) { struct trace_array_cpu *data = tr->data[cpu]; - struct trace_array_cpu *max_data = tr->data[cpu]; + struct trace_array_cpu *max_data; max_tr.cpu = cpu; max_tr.time_start = data->preempt_timestamp; @@ -602,7 +602,7 @@ __update_max_tr(struct trace_array *tr, struct task_struct *tsk, int cpu) max_data->critical_start = data->critical_start; max_data->critical_end = data->critical_end; - memcpy(data->comm, tsk->comm, TASK_COMM_LEN); + memcpy(max_data->comm, tsk->comm, TASK_COMM_LEN); max_data->pid = tsk->pid; max_data->uid = task_uid(tsk); max_data->nice = tsk->static_prio - 20 - MAX_RT_PRIO; -- cgit v1.2.3 From 0e95017355dcf43031da6d0e360a748717e56df1 Mon Sep 17 00:00:00 2001 From: Tim Bird Date: Thu, 25 Feb 2010 15:36:43 -0800 Subject: function-graph: Add tracing_thresh support to function_graph tracer Add support for tracing_thresh to the function_graph tracer. This version of this feature isolates the checks into new entry and return functions, to avoid adding more conditional code into the main function_graph paths. When the tracing_thresh is set and the function graph tracer is enabled, only the functions that took longer than the time in microseconds that was set in tracing_thresh are recorded. To do this efficiently, only the function exits are recorded: [tracing]# echo 100 > tracing_thresh [tracing]# echo function_graph > current_tracer [tracing]# cat trace # tracer: function_graph # # CPU DURATION FUNCTION CALLS # | | | | | | | 1) ! 119.214 us | } /* smp_apic_timer_interrupt */ 1) <========== | 0) ! 101.527 us | } /* __rcu_process_callbacks */ 0) ! 126.461 us | } /* rcu_process_callbacks */ 0) ! 145.111 us | } /* __do_softirq */ 0) ! 149.667 us | } /* do_softirq */ 0) ! 168.817 us | } /* irq_exit */ 0) ! 248.254 us | } /* smp_apic_timer_interrupt */ Also, add support for specifying tracing_thresh on the kernel command line. When used like so: "tracing_thresh=200 ftrace=function_graph" this can be used to analyse system startup. It is important to disable tracing soon after boot, in order to avoid losing the trace data. Acked-by: Frederic Weisbecker Signed-off-by: Tim Bird LKML-Reference: <4B87098B.4040308@am.sony.com> Signed-off-by: Steven Rostedt --- kernel/trace/trace.c | 20 ++++++++++++++++++-- kernel/trace/trace.h | 3 ++- kernel/trace/trace_functions_graph.c | 25 +++++++++++++++++++++++-- 3 files changed, 43 insertions(+), 5 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index 6efd5cb3c252..ababedb4e87f 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -374,6 +374,21 @@ static int __init set_buf_size(char *str) } __setup("trace_buf_size=", set_buf_size); +static int __init set_tracing_thresh(char *str) +{ + unsigned long threshhold; + int ret; + + if (!str) + return 0; + ret = strict_strtoul(str, 0, &threshhold); + if (ret < 0) + return 0; + tracing_thresh = threshhold * 1000; + return 1; +} +__setup("tracing_thresh=", set_tracing_thresh); + unsigned long nsecs_to_usecs(unsigned long nsecs) { return nsecs / 1000; @@ -579,9 +594,10 @@ static ssize_t trace_seq_to_buffer(struct trace_seq *s, void *buf, size_t cnt) static arch_spinlock_t ftrace_max_lock = (arch_spinlock_t)__ARCH_SPIN_LOCK_UNLOCKED; +unsigned long __read_mostly tracing_thresh; + #ifdef CONFIG_TRACER_MAX_TRACE unsigned long __read_mostly tracing_max_latency; -unsigned long __read_mostly tracing_thresh; /* * Copy the new maximum trace into the separate maximum-trace @@ -4248,10 +4264,10 @@ static __init int tracer_init_debugfs(void) #ifdef CONFIG_TRACER_MAX_TRACE trace_create_file("tracing_max_latency", 0644, d_tracer, &tracing_max_latency, &tracing_max_lat_fops); +#endif trace_create_file("tracing_thresh", 0644, d_tracer, &tracing_thresh, &tracing_max_lat_fops); -#endif trace_create_file("README", 0444, d_tracer, NULL, &tracing_readme_fops); diff --git a/kernel/trace/trace.h b/kernel/trace/trace.h index fd05bcaf91b0..1bc8cd1431d7 100644 --- a/kernel/trace/trace.h +++ b/kernel/trace/trace.h @@ -396,9 +396,10 @@ extern int process_new_ksym_entry(char *ksymname, int op, unsigned long addr); extern unsigned long nsecs_to_usecs(unsigned long nsecs); +extern unsigned long tracing_thresh; + #ifdef CONFIG_TRACER_MAX_TRACE extern unsigned long tracing_max_latency; -extern unsigned long tracing_thresh; void update_max_tr(struct trace_array *tr, struct task_struct *tsk, int cpu); void update_max_tr_single(struct trace_array *tr, diff --git a/kernel/trace/trace_functions_graph.c b/kernel/trace/trace_functions_graph.c index 7b1f24618d97..e9df04b60267 100644 --- a/kernel/trace/trace_functions_graph.c +++ b/kernel/trace/trace_functions_graph.c @@ -237,6 +237,14 @@ int trace_graph_entry(struct ftrace_graph_ent *trace) return ret; } +int trace_graph_thresh_entry(struct ftrace_graph_ent *trace) +{ + if (tracing_thresh) + return 1; + else + return trace_graph_entry(trace); +} + static void __trace_graph_return(struct trace_array *tr, struct ftrace_graph_ret *trace, unsigned long flags, @@ -290,13 +298,26 @@ void set_graph_array(struct trace_array *tr) smp_mb(); } +void trace_graph_thresh_return(struct ftrace_graph_ret *trace) +{ + if (tracing_thresh && + (trace->rettime - trace->calltime < tracing_thresh)) + return; + else + trace_graph_return(trace); +} + static int graph_trace_init(struct trace_array *tr) { int ret; set_graph_array(tr); - ret = register_ftrace_graph(&trace_graph_return, - &trace_graph_entry); + if (tracing_thresh) + ret = register_ftrace_graph(&trace_graph_thresh_return, + &trace_graph_thresh_entry); + else + ret = register_ftrace_graph(&trace_graph_return, + &trace_graph_entry); if (ret) return ret; tracing_start_cmdline_record(); -- cgit v1.2.3 From c530665c31c0140b74ca7689e7f836177796e5bd Mon Sep 17 00:00:00 2001 From: Frederic Weisbecker Date: Wed, 3 Mar 2010 07:16:16 +0100 Subject: perf: Take a hot regs snapshot for trace events We are taking a wrong regs snapshot when a trace event triggers. Either we use get_irq_regs(), which gives us the interrupted registers if we are in an interrupt, or we use task_pt_regs() which gives us the state before we entered the kernel, assuming we are lucky enough to be no kernel thread, in which case task_pt_regs() returns the initial set of regs when the kernel thread was started. What we want is different. We need a hot snapshot of the regs, so that we can get the instruction pointer to record in the sample, the frame pointer for the callchain, and some other things. Let's use the new perf_fetch_caller_regs() for that. Comparison with perf record -e lock: -R -a -f -g Before: perf [kernel] [k] __do_softirq | --- __do_softirq | |--55.16%-- __open | --44.84%-- __write_nocancel After: perf [kernel] [k] perf_tp_event | --- perf_tp_event | |--41.07%-- lock_acquire | | | |--39.36%-- _raw_spin_lock | | | | | |--7.81%-- hrtimer_interrupt | | | smp_apic_timer_interrupt | | | apic_timer_interrupt The old case was producing unreliable callchains. Now having right frame and instruction pointers, we have the trace we want. Also syscalls and kprobe events already have the right regs, let's use them instead of wasting a retrieval. v2: Follow the rename perf_save_regs() -> perf_fetch_caller_regs() Signed-off-by: Frederic Weisbecker Cc: Ingo Molnar Cc: Thomas Gleixner Cc: H. Peter Anvin Cc: Peter Zijlstra Cc: Paul Mackerras Cc: Steven Rostedt Cc: Arnaldo Carvalho de Melo Cc: Masami Hiramatsu Cc: Jason Baron Cc: Archs --- kernel/trace/trace_event_profile.c | 3 ++- kernel/trace/trace_kprobe.c | 5 +++-- kernel/trace/trace_syscalls.c | 4 ++-- 3 files changed, 7 insertions(+), 5 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_event_profile.c b/kernel/trace/trace_event_profile.c index f0d693005075..e66d21e15a0f 100644 --- a/kernel/trace/trace_event_profile.c +++ b/kernel/trace/trace_event_profile.c @@ -2,13 +2,14 @@ * trace event based perf counter profiling * * Copyright (C) 2009 Red Hat Inc, Peter Zijlstra - * + * Copyright (C) 2009-2010 Frederic Weisbecker */ #include #include #include "trace.h" +DEFINE_PER_CPU(struct pt_regs, perf_trace_regs); static char *perf_trace_buf; static char *perf_trace_buf_nmi; diff --git a/kernel/trace/trace_kprobe.c b/kernel/trace/trace_kprobe.c index 505c92273b1a..f7a20a8bfb31 100644 --- a/kernel/trace/trace_kprobe.c +++ b/kernel/trace/trace_kprobe.c @@ -1240,7 +1240,7 @@ static __kprobes void kprobe_profile_func(struct kprobe *kp, for (i = 0; i < tp->nr_args; i++) entry->args[i] = call_fetch(&tp->args[i].fetch, regs); - ftrace_perf_buf_submit(entry, size, rctx, entry->ip, 1, irq_flags); + ftrace_perf_buf_submit(entry, size, rctx, entry->ip, 1, irq_flags, regs); } /* Kretprobe profile handler */ @@ -1271,7 +1271,8 @@ static __kprobes void kretprobe_profile_func(struct kretprobe_instance *ri, for (i = 0; i < tp->nr_args; i++) entry->args[i] = call_fetch(&tp->args[i].fetch, regs); - ftrace_perf_buf_submit(entry, size, rctx, entry->ret_ip, 1, irq_flags); + ftrace_perf_buf_submit(entry, size, rctx, entry->ret_ip, 1, + irq_flags, regs); } static int probe_profile_enable(struct ftrace_event_call *call) diff --git a/kernel/trace/trace_syscalls.c b/kernel/trace/trace_syscalls.c index cba47d7935cc..7e6e84fb7b6c 100644 --- a/kernel/trace/trace_syscalls.c +++ b/kernel/trace/trace_syscalls.c @@ -467,7 +467,7 @@ static void prof_syscall_enter(struct pt_regs *regs, long id) rec->nr = syscall_nr; syscall_get_arguments(current, regs, 0, sys_data->nb_args, (unsigned long *)&rec->args); - ftrace_perf_buf_submit(rec, size, rctx, 0, 1, flags); + ftrace_perf_buf_submit(rec, size, rctx, 0, 1, flags, regs); } int prof_sysenter_enable(struct ftrace_event_call *call) @@ -542,7 +542,7 @@ static void prof_syscall_exit(struct pt_regs *regs, long ret) rec->nr = syscall_nr; rec->ret = syscall_get_return_value(current, regs); - ftrace_perf_buf_submit(rec, size, rctx, 0, 1, flags); + ftrace_perf_buf_submit(rec, size, rctx, 0, 1, flags, regs); } int prof_sysexit_enable(struct ftrace_event_call *call) -- cgit v1.2.3 From 97d5a22005f38057b4bc0d95f81cd26510268794 Mon Sep 17 00:00:00 2001 From: Frederic Weisbecker Date: Fri, 5 Mar 2010 05:35:37 +0100 Subject: perf: Drop the obsolete profile naming for trace events Drop the obsolete "profile" naming used by perf for trace events. Perf can now do more than simple events counting, so generalize the API naming. Signed-off-by: Frederic Weisbecker Cc: Ingo Molnar Cc: Peter Zijlstra Cc: Paul Mackerras Cc: Steven Rostedt Cc: Masami Hiramatsu Cc: Jason Baron --- kernel/trace/Makefile | 2 +- kernel/trace/trace_event_perf.c | 165 +++++++++++++++++++++++++++++++++++++ kernel/trace/trace_event_profile.c | 165 ------------------------------------- kernel/trace/trace_events.c | 2 +- kernel/trace/trace_kprobe.c | 28 +++---- kernel/trace/trace_syscalls.c | 72 ++++++++-------- 6 files changed, 217 insertions(+), 217 deletions(-) create mode 100644 kernel/trace/trace_event_perf.c delete mode 100644 kernel/trace/trace_event_profile.c (limited to 'kernel/trace') diff --git a/kernel/trace/Makefile b/kernel/trace/Makefile index d00c6fe23f54..78edc6490038 100644 --- a/kernel/trace/Makefile +++ b/kernel/trace/Makefile @@ -52,7 +52,7 @@ obj-$(CONFIG_EVENT_TRACING) += trace_events.o obj-$(CONFIG_EVENT_TRACING) += trace_export.o obj-$(CONFIG_FTRACE_SYSCALLS) += trace_syscalls.o ifeq ($(CONFIG_PERF_EVENTS),y) -obj-$(CONFIG_EVENT_TRACING) += trace_event_profile.o +obj-$(CONFIG_EVENT_TRACING) += trace_event_perf.o endif obj-$(CONFIG_EVENT_TRACING) += trace_events_filter.o obj-$(CONFIG_KPROBE_EVENT) += trace_kprobe.o diff --git a/kernel/trace/trace_event_perf.c b/kernel/trace/trace_event_perf.c new file mode 100644 index 000000000000..f315b12a41d8 --- /dev/null +++ b/kernel/trace/trace_event_perf.c @@ -0,0 +1,165 @@ +/* + * trace event based perf event profiling/tracing + * + * Copyright (C) 2009 Red Hat Inc, Peter Zijlstra + * Copyright (C) 2009-2010 Frederic Weisbecker + */ + +#include +#include +#include "trace.h" + +DEFINE_PER_CPU(struct pt_regs, perf_trace_regs); + +static char *perf_trace_buf; +static char *perf_trace_buf_nmi; + +typedef typeof(char [PERF_MAX_TRACE_SIZE]) perf_trace_t ; + +/* Count the events in use (per event id, not per instance) */ +static int total_ref_count; + +static int perf_trace_event_enable(struct ftrace_event_call *event) +{ + char *buf; + int ret = -ENOMEM; + + if (event->perf_refcount++ > 0) + return 0; + + if (!total_ref_count) { + buf = (char *)alloc_percpu(perf_trace_t); + if (!buf) + goto fail_buf; + + rcu_assign_pointer(perf_trace_buf, buf); + + buf = (char *)alloc_percpu(perf_trace_t); + if (!buf) + goto fail_buf_nmi; + + rcu_assign_pointer(perf_trace_buf_nmi, buf); + } + + ret = event->perf_event_enable(event); + if (!ret) { + total_ref_count++; + return 0; + } + +fail_buf_nmi: + if (!total_ref_count) { + free_percpu(perf_trace_buf_nmi); + free_percpu(perf_trace_buf); + perf_trace_buf_nmi = NULL; + perf_trace_buf = NULL; + } +fail_buf: + event->perf_refcount--; + + return ret; +} + +int perf_trace_enable(int event_id) +{ + struct ftrace_event_call *event; + int ret = -EINVAL; + + mutex_lock(&event_mutex); + list_for_each_entry(event, &ftrace_events, list) { + if (event->id == event_id && event->perf_event_enable && + try_module_get(event->mod)) { + ret = perf_trace_event_enable(event); + break; + } + } + mutex_unlock(&event_mutex); + + return ret; +} + +static void perf_trace_event_disable(struct ftrace_event_call *event) +{ + char *buf, *nmi_buf; + + if (--event->perf_refcount > 0) + return; + + event->perf_event_disable(event); + + if (!--total_ref_count) { + buf = perf_trace_buf; + rcu_assign_pointer(perf_trace_buf, NULL); + + nmi_buf = perf_trace_buf_nmi; + rcu_assign_pointer(perf_trace_buf_nmi, NULL); + + /* + * Ensure every events in profiling have finished before + * releasing the buffers + */ + synchronize_sched(); + + free_percpu(buf); + free_percpu(nmi_buf); + } +} + +void perf_trace_disable(int event_id) +{ + struct ftrace_event_call *event; + + mutex_lock(&event_mutex); + list_for_each_entry(event, &ftrace_events, list) { + if (event->id == event_id) { + perf_trace_event_disable(event); + module_put(event->mod); + break; + } + } + mutex_unlock(&event_mutex); +} + +__kprobes void *perf_trace_buf_prepare(int size, unsigned short type, + int *rctxp, unsigned long *irq_flags) +{ + struct trace_entry *entry; + char *trace_buf, *raw_data; + int pc, cpu; + + pc = preempt_count(); + + /* Protect the per cpu buffer, begin the rcu read side */ + local_irq_save(*irq_flags); + + *rctxp = perf_swevent_get_recursion_context(); + if (*rctxp < 0) + goto err_recursion; + + cpu = smp_processor_id(); + + if (in_nmi()) + trace_buf = rcu_dereference(perf_trace_buf_nmi); + else + trace_buf = rcu_dereference(perf_trace_buf); + + if (!trace_buf) + goto err; + + raw_data = per_cpu_ptr(trace_buf, cpu); + + /* zero the dead bytes from align to not leak stack to user */ + *(u64 *)(&raw_data[size - sizeof(u64)]) = 0ULL; + + entry = (struct trace_entry *)raw_data; + tracing_generic_entry_update(entry, *irq_flags, pc); + entry->type = type; + + return raw_data; +err: + perf_swevent_put_recursion_context(*rctxp); +err_recursion: + local_irq_restore(*irq_flags); + return NULL; +} +EXPORT_SYMBOL_GPL(perf_trace_buf_prepare); diff --git a/kernel/trace/trace_event_profile.c b/kernel/trace/trace_event_profile.c deleted file mode 100644 index e66d21e15a0f..000000000000 --- a/kernel/trace/trace_event_profile.c +++ /dev/null @@ -1,165 +0,0 @@ -/* - * trace event based perf counter profiling - * - * Copyright (C) 2009 Red Hat Inc, Peter Zijlstra - * Copyright (C) 2009-2010 Frederic Weisbecker - */ - -#include -#include -#include "trace.h" - -DEFINE_PER_CPU(struct pt_regs, perf_trace_regs); - -static char *perf_trace_buf; -static char *perf_trace_buf_nmi; - -typedef typeof(char [FTRACE_MAX_PROFILE_SIZE]) perf_trace_t ; - -/* Count the events in use (per event id, not per instance) */ -static int total_profile_count; - -static int ftrace_profile_enable_event(struct ftrace_event_call *event) -{ - char *buf; - int ret = -ENOMEM; - - if (event->profile_count++ > 0) - return 0; - - if (!total_profile_count) { - buf = (char *)alloc_percpu(perf_trace_t); - if (!buf) - goto fail_buf; - - rcu_assign_pointer(perf_trace_buf, buf); - - buf = (char *)alloc_percpu(perf_trace_t); - if (!buf) - goto fail_buf_nmi; - - rcu_assign_pointer(perf_trace_buf_nmi, buf); - } - - ret = event->profile_enable(event); - if (!ret) { - total_profile_count++; - return 0; - } - -fail_buf_nmi: - if (!total_profile_count) { - free_percpu(perf_trace_buf_nmi); - free_percpu(perf_trace_buf); - perf_trace_buf_nmi = NULL; - perf_trace_buf = NULL; - } -fail_buf: - event->profile_count--; - - return ret; -} - -int ftrace_profile_enable(int event_id) -{ - struct ftrace_event_call *event; - int ret = -EINVAL; - - mutex_lock(&event_mutex); - list_for_each_entry(event, &ftrace_events, list) { - if (event->id == event_id && event->profile_enable && - try_module_get(event->mod)) { - ret = ftrace_profile_enable_event(event); - break; - } - } - mutex_unlock(&event_mutex); - - return ret; -} - -static void ftrace_profile_disable_event(struct ftrace_event_call *event) -{ - char *buf, *nmi_buf; - - if (--event->profile_count > 0) - return; - - event->profile_disable(event); - - if (!--total_profile_count) { - buf = perf_trace_buf; - rcu_assign_pointer(perf_trace_buf, NULL); - - nmi_buf = perf_trace_buf_nmi; - rcu_assign_pointer(perf_trace_buf_nmi, NULL); - - /* - * Ensure every events in profiling have finished before - * releasing the buffers - */ - synchronize_sched(); - - free_percpu(buf); - free_percpu(nmi_buf); - } -} - -void ftrace_profile_disable(int event_id) -{ - struct ftrace_event_call *event; - - mutex_lock(&event_mutex); - list_for_each_entry(event, &ftrace_events, list) { - if (event->id == event_id) { - ftrace_profile_disable_event(event); - module_put(event->mod); - break; - } - } - mutex_unlock(&event_mutex); -} - -__kprobes void *ftrace_perf_buf_prepare(int size, unsigned short type, - int *rctxp, unsigned long *irq_flags) -{ - struct trace_entry *entry; - char *trace_buf, *raw_data; - int pc, cpu; - - pc = preempt_count(); - - /* Protect the per cpu buffer, begin the rcu read side */ - local_irq_save(*irq_flags); - - *rctxp = perf_swevent_get_recursion_context(); - if (*rctxp < 0) - goto err_recursion; - - cpu = smp_processor_id(); - - if (in_nmi()) - trace_buf = rcu_dereference(perf_trace_buf_nmi); - else - trace_buf = rcu_dereference(perf_trace_buf); - - if (!trace_buf) - goto err; - - raw_data = per_cpu_ptr(trace_buf, cpu); - - /* zero the dead bytes from align to not leak stack to user */ - *(u64 *)(&raw_data[size - sizeof(u64)]) = 0ULL; - - entry = (struct trace_entry *)raw_data; - tracing_generic_entry_update(entry, *irq_flags, pc); - entry->type = type; - - return raw_data; -err: - perf_swevent_put_recursion_context(*rctxp); -err_recursion: - local_irq_restore(*irq_flags); - return NULL; -} -EXPORT_SYMBOL_GPL(ftrace_perf_buf_prepare); diff --git a/kernel/trace/trace_events.c b/kernel/trace/trace_events.c index 3f972ad98d04..beab8bf2f310 100644 --- a/kernel/trace/trace_events.c +++ b/kernel/trace/trace_events.c @@ -938,7 +938,7 @@ event_create_dir(struct ftrace_event_call *call, struct dentry *d_events, trace_create_file("enable", 0644, call->dir, call, enable); - if (call->id && call->profile_enable) + if (call->id && call->perf_event_enable) trace_create_file("id", 0444, call->dir, call, id); diff --git a/kernel/trace/trace_kprobe.c b/kernel/trace/trace_kprobe.c index f7a20a8bfb31..1251e367bae9 100644 --- a/kernel/trace/trace_kprobe.c +++ b/kernel/trace/trace_kprobe.c @@ -1214,7 +1214,7 @@ static int set_print_fmt(struct trace_probe *tp) #ifdef CONFIG_PERF_EVENTS /* Kprobe profile handler */ -static __kprobes void kprobe_profile_func(struct kprobe *kp, +static __kprobes void kprobe_perf_func(struct kprobe *kp, struct pt_regs *regs) { struct trace_probe *tp = container_of(kp, struct trace_probe, rp.kp); @@ -1227,11 +1227,11 @@ static __kprobes void kprobe_profile_func(struct kprobe *kp, __size = SIZEOF_KPROBE_TRACE_ENTRY(tp->nr_args); size = ALIGN(__size + sizeof(u32), sizeof(u64)); size -= sizeof(u32); - if (WARN_ONCE(size > FTRACE_MAX_PROFILE_SIZE, + if (WARN_ONCE(size > PERF_MAX_TRACE_SIZE, "profile buffer not large enough")) return; - entry = ftrace_perf_buf_prepare(size, call->id, &rctx, &irq_flags); + entry = perf_trace_buf_prepare(size, call->id, &rctx, &irq_flags); if (!entry) return; @@ -1240,11 +1240,11 @@ static __kprobes void kprobe_profile_func(struct kprobe *kp, for (i = 0; i < tp->nr_args; i++) entry->args[i] = call_fetch(&tp->args[i].fetch, regs); - ftrace_perf_buf_submit(entry, size, rctx, entry->ip, 1, irq_flags, regs); + perf_trace_buf_submit(entry, size, rctx, entry->ip, 1, irq_flags, regs); } /* Kretprobe profile handler */ -static __kprobes void kretprobe_profile_func(struct kretprobe_instance *ri, +static __kprobes void kretprobe_perf_func(struct kretprobe_instance *ri, struct pt_regs *regs) { struct trace_probe *tp = container_of(ri->rp, struct trace_probe, rp); @@ -1257,11 +1257,11 @@ static __kprobes void kretprobe_profile_func(struct kretprobe_instance *ri, __size = SIZEOF_KRETPROBE_TRACE_ENTRY(tp->nr_args); size = ALIGN(__size + sizeof(u32), sizeof(u64)); size -= sizeof(u32); - if (WARN_ONCE(size > FTRACE_MAX_PROFILE_SIZE, + if (WARN_ONCE(size > PERF_MAX_TRACE_SIZE, "profile buffer not large enough")) return; - entry = ftrace_perf_buf_prepare(size, call->id, &rctx, &irq_flags); + entry = perf_trace_buf_prepare(size, call->id, &rctx, &irq_flags); if (!entry) return; @@ -1271,11 +1271,11 @@ static __kprobes void kretprobe_profile_func(struct kretprobe_instance *ri, for (i = 0; i < tp->nr_args; i++) entry->args[i] = call_fetch(&tp->args[i].fetch, regs); - ftrace_perf_buf_submit(entry, size, rctx, entry->ret_ip, 1, + perf_trace_buf_submit(entry, size, rctx, entry->ret_ip, 1, irq_flags, regs); } -static int probe_profile_enable(struct ftrace_event_call *call) +static int probe_perf_enable(struct ftrace_event_call *call) { struct trace_probe *tp = (struct trace_probe *)call->data; @@ -1287,7 +1287,7 @@ static int probe_profile_enable(struct ftrace_event_call *call) return enable_kprobe(&tp->rp.kp); } -static void probe_profile_disable(struct ftrace_event_call *call) +static void probe_perf_disable(struct ftrace_event_call *call) { struct trace_probe *tp = (struct trace_probe *)call->data; @@ -1312,7 +1312,7 @@ int kprobe_dispatcher(struct kprobe *kp, struct pt_regs *regs) kprobe_trace_func(kp, regs); #ifdef CONFIG_PERF_EVENTS if (tp->flags & TP_FLAG_PROFILE) - kprobe_profile_func(kp, regs); + kprobe_perf_func(kp, regs); #endif return 0; /* We don't tweek kernel, so just return 0 */ } @@ -1326,7 +1326,7 @@ int kretprobe_dispatcher(struct kretprobe_instance *ri, struct pt_regs *regs) kretprobe_trace_func(ri, regs); #ifdef CONFIG_PERF_EVENTS if (tp->flags & TP_FLAG_PROFILE) - kretprobe_profile_func(ri, regs); + kretprobe_perf_func(ri, regs); #endif return 0; /* We don't tweek kernel, so just return 0 */ } @@ -1359,8 +1359,8 @@ static int register_probe_event(struct trace_probe *tp) call->unregfunc = probe_event_disable; #ifdef CONFIG_PERF_EVENTS - call->profile_enable = probe_profile_enable; - call->profile_disable = probe_profile_disable; + call->perf_event_enable = probe_perf_enable; + call->perf_event_disable = probe_perf_disable; #endif call->data = tp; ret = trace_add_event_call(call); diff --git a/kernel/trace/trace_syscalls.c b/kernel/trace/trace_syscalls.c index 7e6e84fb7b6c..33c2a5b769dc 100644 --- a/kernel/trace/trace_syscalls.c +++ b/kernel/trace/trace_syscalls.c @@ -428,12 +428,12 @@ core_initcall(init_ftrace_syscalls); #ifdef CONFIG_PERF_EVENTS -static DECLARE_BITMAP(enabled_prof_enter_syscalls, NR_syscalls); -static DECLARE_BITMAP(enabled_prof_exit_syscalls, NR_syscalls); -static int sys_prof_refcount_enter; -static int sys_prof_refcount_exit; +static DECLARE_BITMAP(enabled_perf_enter_syscalls, NR_syscalls); +static DECLARE_BITMAP(enabled_perf_exit_syscalls, NR_syscalls); +static int sys_perf_refcount_enter; +static int sys_perf_refcount_exit; -static void prof_syscall_enter(struct pt_regs *regs, long id) +static void perf_syscall_enter(struct pt_regs *regs, long id) { struct syscall_metadata *sys_data; struct syscall_trace_enter *rec; @@ -443,7 +443,7 @@ static void prof_syscall_enter(struct pt_regs *regs, long id) int size; syscall_nr = syscall_get_nr(current, regs); - if (!test_bit(syscall_nr, enabled_prof_enter_syscalls)) + if (!test_bit(syscall_nr, enabled_perf_enter_syscalls)) return; sys_data = syscall_nr_to_meta(syscall_nr); @@ -455,11 +455,11 @@ static void prof_syscall_enter(struct pt_regs *regs, long id) size = ALIGN(size + sizeof(u32), sizeof(u64)); size -= sizeof(u32); - if (WARN_ONCE(size > FTRACE_MAX_PROFILE_SIZE, - "profile buffer not large enough")) + if (WARN_ONCE(size > PERF_MAX_TRACE_SIZE, + "perf buffer not large enough")) return; - rec = (struct syscall_trace_enter *)ftrace_perf_buf_prepare(size, + rec = (struct syscall_trace_enter *)perf_trace_buf_prepare(size, sys_data->enter_event->id, &rctx, &flags); if (!rec) return; @@ -467,10 +467,10 @@ static void prof_syscall_enter(struct pt_regs *regs, long id) rec->nr = syscall_nr; syscall_get_arguments(current, regs, 0, sys_data->nb_args, (unsigned long *)&rec->args); - ftrace_perf_buf_submit(rec, size, rctx, 0, 1, flags, regs); + perf_trace_buf_submit(rec, size, rctx, 0, 1, flags, regs); } -int prof_sysenter_enable(struct ftrace_event_call *call) +int perf_sysenter_enable(struct ftrace_event_call *call) { int ret = 0; int num; @@ -478,34 +478,34 @@ int prof_sysenter_enable(struct ftrace_event_call *call) num = ((struct syscall_metadata *)call->data)->syscall_nr; mutex_lock(&syscall_trace_lock); - if (!sys_prof_refcount_enter) - ret = register_trace_sys_enter(prof_syscall_enter); + if (!sys_perf_refcount_enter) + ret = register_trace_sys_enter(perf_syscall_enter); if (ret) { pr_info("event trace: Could not activate" "syscall entry trace point"); } else { - set_bit(num, enabled_prof_enter_syscalls); - sys_prof_refcount_enter++; + set_bit(num, enabled_perf_enter_syscalls); + sys_perf_refcount_enter++; } mutex_unlock(&syscall_trace_lock); return ret; } -void prof_sysenter_disable(struct ftrace_event_call *call) +void perf_sysenter_disable(struct ftrace_event_call *call) { int num; num = ((struct syscall_metadata *)call->data)->syscall_nr; mutex_lock(&syscall_trace_lock); - sys_prof_refcount_enter--; - clear_bit(num, enabled_prof_enter_syscalls); - if (!sys_prof_refcount_enter) - unregister_trace_sys_enter(prof_syscall_enter); + sys_perf_refcount_enter--; + clear_bit(num, enabled_perf_enter_syscalls); + if (!sys_perf_refcount_enter) + unregister_trace_sys_enter(perf_syscall_enter); mutex_unlock(&syscall_trace_lock); } -static void prof_syscall_exit(struct pt_regs *regs, long ret) +static void perf_syscall_exit(struct pt_regs *regs, long ret) { struct syscall_metadata *sys_data; struct syscall_trace_exit *rec; @@ -515,7 +515,7 @@ static void prof_syscall_exit(struct pt_regs *regs, long ret) int size; syscall_nr = syscall_get_nr(current, regs); - if (!test_bit(syscall_nr, enabled_prof_exit_syscalls)) + if (!test_bit(syscall_nr, enabled_perf_exit_syscalls)) return; sys_data = syscall_nr_to_meta(syscall_nr); @@ -530,11 +530,11 @@ static void prof_syscall_exit(struct pt_regs *regs, long ret) * Impossible, but be paranoid with the future * How to put this check outside runtime? */ - if (WARN_ONCE(size > FTRACE_MAX_PROFILE_SIZE, - "exit event has grown above profile buffer size")) + if (WARN_ONCE(size > PERF_MAX_TRACE_SIZE, + "exit event has grown above perf buffer size")) return; - rec = (struct syscall_trace_exit *)ftrace_perf_buf_prepare(size, + rec = (struct syscall_trace_exit *)perf_trace_buf_prepare(size, sys_data->exit_event->id, &rctx, &flags); if (!rec) return; @@ -542,10 +542,10 @@ static void prof_syscall_exit(struct pt_regs *regs, long ret) rec->nr = syscall_nr; rec->ret = syscall_get_return_value(current, regs); - ftrace_perf_buf_submit(rec, size, rctx, 0, 1, flags, regs); + perf_trace_buf_submit(rec, size, rctx, 0, 1, flags, regs); } -int prof_sysexit_enable(struct ftrace_event_call *call) +int perf_sysexit_enable(struct ftrace_event_call *call) { int ret = 0; int num; @@ -553,30 +553,30 @@ int prof_sysexit_enable(struct ftrace_event_call *call) num = ((struct syscall_metadata *)call->data)->syscall_nr; mutex_lock(&syscall_trace_lock); - if (!sys_prof_refcount_exit) - ret = register_trace_sys_exit(prof_syscall_exit); + if (!sys_perf_refcount_exit) + ret = register_trace_sys_exit(perf_syscall_exit); if (ret) { pr_info("event trace: Could not activate" "syscall exit trace point"); } else { - set_bit(num, enabled_prof_exit_syscalls); - sys_prof_refcount_exit++; + set_bit(num, enabled_perf_exit_syscalls); + sys_perf_refcount_exit++; } mutex_unlock(&syscall_trace_lock); return ret; } -void prof_sysexit_disable(struct ftrace_event_call *call) +void perf_sysexit_disable(struct ftrace_event_call *call) { int num; num = ((struct syscall_metadata *)call->data)->syscall_nr; mutex_lock(&syscall_trace_lock); - sys_prof_refcount_exit--; - clear_bit(num, enabled_prof_exit_syscalls); - if (!sys_prof_refcount_exit) - unregister_trace_sys_exit(prof_syscall_exit); + sys_perf_refcount_exit--; + clear_bit(num, enabled_perf_exit_syscalls); + if (!sys_perf_refcount_exit) + unregister_trace_sys_exit(perf_syscall_exit); mutex_unlock(&syscall_trace_lock); } -- cgit v1.2.3 From 3f379b03fbfddd20536389a85c6456f8233d1f8d Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Fri, 5 Mar 2010 15:03:25 -0800 Subject: ftrace: Replace read_barrier_depends() with rcu_dereference_raw() Replace the calls to read_barrier_depends() in ftrace_list_func() with rcu_dereference_raw() to improve readability. The reason that we use rcu_dereference_raw() here is that removed entries are never freed, instead they are simply leaked. This is one of a very few cases where use of rcu_dereference_raw() is the long-term right answer. And I don't yet know of any others. ;-) Signed-off-by: Paul E. McKenney Acked-by: Steven Rostedt Cc: Frederic Weisbecker Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: mathieu.desnoyers@polymtl.ca Cc: josh@joshtriplett.org Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: Valdis.Kletnieks@vt.edu Cc: dhowells@redhat.com LKML-Reference: <1267830207-9474-1-git-send-email-paulmck@linux.vnet.ibm.com> Signed-off-by: Ingo Molnar --- kernel/trace/ftrace.c | 22 +++++++++++++--------- 1 file changed, 13 insertions(+), 9 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index 83783579378f..8c5adc0e5db3 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -27,6 +27,7 @@ #include #include #include +#include #include @@ -88,18 +89,22 @@ ftrace_func_t ftrace_pid_function __read_mostly = ftrace_stub; static int ftrace_set_func(unsigned long *array, int *idx, char *buffer); #endif +/* + * Traverse the ftrace_list, invoking all entries. The reason that we + * can use rcu_dereference_raw() is that elements removed from this list + * are simply leaked, so there is no need to interact with a grace-period + * mechanism. The rcu_dereference_raw() calls are needed to handle + * concurrent insertions into the ftrace_list. + * + * Silly Alpha and silly pointer-speculation compiler optimizations! + */ static void ftrace_list_func(unsigned long ip, unsigned long parent_ip) { - struct ftrace_ops *op = ftrace_list; - - /* in case someone actually ports this to alpha! */ - read_barrier_depends(); + struct ftrace_ops *op = rcu_dereference_raw(ftrace_list); /*see above*/ while (op != &ftrace_list_end) { - /* silly alpha */ - read_barrier_depends(); op->func(ip, parent_ip); - op = op->next; + op = rcu_dereference_raw(op->next); /*see above*/ }; } @@ -154,8 +159,7 @@ static int __register_ftrace_function(struct ftrace_ops *ops) * the ops->next pointer is valid before another CPU sees * the ops pointer included into the ftrace_list. */ - smp_wmb(); - ftrace_list = ops; + rcu_assign_pointer(ftrace_list, ops); if (ftrace_enabled) { ftrace_func_t func; -- cgit v1.2.3 From 639fe4b12f92b54c9c3b38c82cdafaa38cfd3e63 Mon Sep 17 00:00:00 2001 From: Xiao Guangrong Date: Thu, 11 Mar 2010 15:30:35 +0800 Subject: perf: export perf_trace_regs and perf_arch_fetch_caller_regs Export perf_trace_regs and perf_arch_fetch_caller_regs since module will use these. Signed-off-by: Xiao Guangrong [ use EXPORT_PER_CPU_SYMBOL_GPL() ] Signed-off-by: Peter Zijlstra LKML-Reference: <4B989C1B.2090407@cn.fujitsu.com> Signed-off-by: Ingo Molnar --- kernel/trace/trace_event_perf.c | 1 + 1 file changed, 1 insertion(+) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_event_perf.c b/kernel/trace/trace_event_perf.c index f315b12a41d8..0709e4f75114 100644 --- a/kernel/trace/trace_event_perf.c +++ b/kernel/trace/trace_event_perf.c @@ -10,6 +10,7 @@ #include "trace.h" DEFINE_PER_CPU(struct pt_regs, perf_trace_regs); +EXPORT_PER_CPU_SYMBOL_GPL(perf_trace_regs); static char *perf_trace_buf; static char *perf_trace_buf_nmi; -- cgit v1.2.3 From 52fbe9cde7fdb5c6fac196d7ebd2d92d05ef3cd4 Mon Sep 17 00:00:00 2001 From: Lai Jiangshan Date: Mon, 8 Mar 2010 14:50:43 +0800 Subject: ring-buffer: Move disabled check into preempt disable section The ring buffer resizing and resetting relies on a schedule RCU action. The buffers are disabled, a synchronize_sched() is called and then the resize or reset takes place. But this only works if the disabling of the buffers are within the preempt disabled section, otherwise a window exists that the buffers can be written to while a reset or resize takes place. Cc: stable@kernel.org Reported-by: Li Zefan Signed-off-by: Lai Jiangshan LKML-Reference: <4B949E43.2010906@cn.fujitsu.com> Signed-off-by: Steven Rostedt --- kernel/trace/ring_buffer.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/ring_buffer.c b/kernel/trace/ring_buffer.c index 8c1b2d290718..54191d6ed195 100644 --- a/kernel/trace/ring_buffer.c +++ b/kernel/trace/ring_buffer.c @@ -2232,12 +2232,12 @@ ring_buffer_lock_reserve(struct ring_buffer *buffer, unsigned long length) if (ring_buffer_flags != RB_BUFFERS_ON) return NULL; - if (atomic_read(&buffer->record_disabled)) - return NULL; - /* If we are tracing schedule, we don't want to recurse */ resched = ftrace_preempt_disable(); + if (atomic_read(&buffer->record_disabled)) + goto out_nocheck; + if (trace_recursive_lock()) goto out_nocheck; @@ -2469,11 +2469,11 @@ int ring_buffer_write(struct ring_buffer *buffer, if (ring_buffer_flags != RB_BUFFERS_ON) return -EBUSY; - if (atomic_read(&buffer->record_disabled)) - return -EBUSY; - resched = ftrace_preempt_disable(); + if (atomic_read(&buffer->record_disabled)) + goto out; + cpu = raw_smp_processor_id(); if (!cpumask_test_cpu(cpu, buffer->cpumask)) -- cgit v1.2.3 From ea14eb714041d40fcc5180b5a586034503650149 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Fri, 12 Mar 2010 19:41:23 -0500 Subject: function-graph: Init curr_ret_stack with ret_stack If the graph tracer is active, and a task is forked but the allocating of the processes graph stack fails, it can cause crash later on. This is due to the temporary stack being NULL, but the curr_ret_stack variable is copied from the parent. If it is not -1, then in ftrace_graph_probe_sched_switch() the following: for (index = next->curr_ret_stack; index >= 0; index--) next->ret_stack[index].calltime += timestamp; Will cause a kernel OOPS. Found with Li Zefan's ftrace_stress_test. Cc: stable@kernel.org Signed-off-by: Steven Rostedt --- kernel/trace/ftrace.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel/trace') diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index d4d1238b096b..bb53edbb5c8c 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -3349,6 +3349,7 @@ void ftrace_graph_init_task(struct task_struct *t) { /* Make sure we do not use the parent ret_stack */ t->ret_stack = NULL; + t->curr_ret_stack = -1; if (ftrace_graph_active) { struct ftrace_ret_stack *ret_stack; @@ -3358,7 +3359,6 @@ void ftrace_graph_init_task(struct task_struct *t) GFP_KERNEL); if (!ret_stack) return; - t->curr_ret_stack = -1; atomic_set(&t->tracing_graph_pause, 0); atomic_set(&t->trace_overrun, 0); t->ftrace_timestamp = 0; -- cgit v1.2.3 From 283740c619d211e34572cc93c8cdba92ccbdb9cc Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Fri, 12 Mar 2010 19:48:41 -0500 Subject: tracing: Use same local variable when resetting the ring buffer In the ftrace code that resets the ring buffer it references the buffer with a local variable, but then uses the tr->buffer as the parameter to reset. If the wakeup tracer is running, which can switch the tr->buffer with the max saved buffer, this can break the requirement of disabling the buffer before the reset. buffer = tr->buffer; ring_buffer_record_disable(buffer); synchronize_sched(); __tracing_reset(tr->buffer, cpu); If the tr->buffer is swapped, then the reset is not happening to the buffer that was disabled. This will cause the ring buffer to fail. Found with Li Zefan's ftrace_stress_test. Cc: stable@kernel.org Reported-by: Lai Jiangshan Signed-off-by: Steven Rostedt --- kernel/trace/trace.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index 6af8d7bc953b..60de37bd0f75 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -840,10 +840,10 @@ out: mutex_unlock(&trace_types_lock); } -static void __tracing_reset(struct trace_array *tr, int cpu) +static void __tracing_reset(struct ring_buffer *buffer, int cpu) { ftrace_disable_cpu(); - ring_buffer_reset_cpu(tr->buffer, cpu); + ring_buffer_reset_cpu(buffer, cpu); ftrace_enable_cpu(); } @@ -855,7 +855,7 @@ void tracing_reset(struct trace_array *tr, int cpu) /* Make sure all commits have finished */ synchronize_sched(); - __tracing_reset(tr, cpu); + __tracing_reset(buffer, cpu); ring_buffer_record_enable(buffer); } @@ -873,7 +873,7 @@ void tracing_reset_online_cpus(struct trace_array *tr) tr->time_start = ftrace_now(tr->cpu); for_each_online_cpu(cpu) - __tracing_reset(tr, cpu); + __tracing_reset(buffer, cpu); ring_buffer_record_enable(buffer); } -- cgit v1.2.3 From a2f8071428ed9a0f06865f417c962421c9a6b488 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Fri, 12 Mar 2010 19:56:00 -0500 Subject: tracing: Disable buffer switching when starting or stopping trace When the trace iterator is read, tracing_start() and tracing_stop() is called to stop tracing while the iterator is processing the trace output. These functions disable both the standard buffer and the max latency buffer. But if the wakeup tracer is running, it can switch these buffers between the two disables: buffer = global_trace.buffer; if (buffer) ring_buffer_record_disable(buffer); <<<--------- swap happens here buffer = max_tr.buffer; if (buffer) ring_buffer_record_disable(buffer); What happens is that we disabled the same buffer twice. On tracing_start() we can enable the same buffer twice. All ring_buffer_record_disable() must be matched with a ring_buffer_record_enable() or the buffer can be disable permanently, or enable prematurely, and cause a bug where a reset happens while a trace is commiting. This patch protects these two by taking the ftrace_max_lock to prevent a switch from occurring. Found with Li Zefan's ftrace_stress_test. Cc: stable@kernel.org Reported-by: Lai Jiangshan Signed-off-by: Steven Rostedt --- kernel/trace/trace.c | 9 +++++++++ 1 file changed, 9 insertions(+) (limited to 'kernel/trace') diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index 60de37bd0f75..484337d33959 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -950,6 +950,8 @@ void tracing_start(void) goto out; } + /* Prevent the buffers from switching */ + arch_spin_lock(&ftrace_max_lock); buffer = global_trace.buffer; if (buffer) @@ -959,6 +961,8 @@ void tracing_start(void) if (buffer) ring_buffer_record_enable(buffer); + arch_spin_unlock(&ftrace_max_lock); + ftrace_start(); out: spin_unlock_irqrestore(&tracing_start_lock, flags); @@ -980,6 +984,9 @@ void tracing_stop(void) if (trace_stop_count++) goto out; + /* Prevent the buffers from switching */ + arch_spin_lock(&ftrace_max_lock); + buffer = global_trace.buffer; if (buffer) ring_buffer_record_disable(buffer); @@ -988,6 +995,8 @@ void tracing_stop(void) if (buffer) ring_buffer_record_disable(buffer); + arch_spin_unlock(&ftrace_max_lock); + out: spin_unlock_irqrestore(&tracing_start_lock, flags); } -- cgit v1.2.3 From b6345879ccbd9b92864fbd7eb8ac48acdb4d6b15 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Fri, 12 Mar 2010 20:03:30 -0500 Subject: tracing: Do not record user stack trace from NMI context A bug was found with Li Zefan's ftrace_stress_test that caused applications to segfault during the test. Placing a tracing_off() in the segfault code, and examining several traces, I found that the following was always the case. The lock tracer was enabled (lockdep being required) and userstack was enabled. Testing this out, I just enabled the two, but that was not good enough. I needed to run something else that could trigger it. Running a load like hackbench did not work, but executing a new program would. The following would trigger the segfault within seconds: # echo 1 > /debug/tracing/options/userstacktrace # echo 1 > /debug/tracing/events/lock/enable # while :; do ls > /dev/null ; done Enabling the function graph tracer and looking at what was happening I finally noticed that all cashes happened just after an NMI. 1) | copy_user_handle_tail() { 1) | bad_area_nosemaphore() { 1) | __bad_area_nosemaphore() { 1) | no_context() { 1) | fixup_exception() { 1) 0.319 us | search_exception_tables(); 1) 0.873 us | } [...] 1) 0.314 us | __rcu_read_unlock(); 1) 0.325 us | native_apic_mem_write(); 1) 0.943 us | } 1) 0.304 us | rcu_nmi_exit(); [...] 1) 0.479 us | find_vma(); 1) | bad_area() { 1) | __bad_area() { After capturing several traces of failures, all of them happened after an NMI. Curious about this, I added a trace_printk() to the NMI handler to read the regs->ip to see where the NMI happened. In which I found out it was here: ffffffff8135b660 : ffffffff8135b660: 48 83 ec 78 sub $0x78,%rsp ffffffff8135b664: e8 97 01 00 00 callq ffffffff8135b800 What was happening is that the NMI would happen at the place that a page fault occurred. It would call rcu_read_lock() which was traced by the lock events, and the user_stack_trace would run. This would trigger a page fault inside the NMI. I do not see where the CR2 register is saved or restored in NMI handling. This means that it would corrupt the page fault handling that the NMI interrupted. The reason the while loop of ls helped trigger the bug, was that each execution of ls would cause lots of pages to be faulted in, and increase the chances of the race happening. The simple solution is to not allow user stack traces in NMI context. After this patch, I ran the above "ls" test for a couple of hours without any issues. Without this patch, the bug would trigger in less than a minute. Cc: stable@kernel.org Reported-by: Li Zefan Signed-off-by: Steven Rostedt --- kernel/trace/trace.c | 7 +++++++ 1 file changed, 7 insertions(+) (limited to 'kernel/trace') diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index 484337d33959..e52683f7c3b2 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -1284,6 +1284,13 @@ ftrace_trace_userstack(struct ring_buffer *buffer, unsigned long flags, int pc) if (!(trace_flags & TRACE_ITER_USERSTACKTRACE)) return; + /* + * NMIs can not handle page faults, even with fix ups. + * The save user stack can (and often does) fault. + */ + if (unlikely(in_nmi())) + return; + event = trace_buffer_lock_reserve(buffer, TRACE_USER_STACK, sizeof(*entry), flags, pc); if (!event) -- cgit v1.2.3 From dcd5c1662db59a6b82942f47fb6ac9dd63f6d3dd Mon Sep 17 00:00:00 2001 From: Frederic Weisbecker Date: Tue, 16 Mar 2010 01:05:02 +0100 Subject: perf: Fix unexported generic perf_arch_fetch_caller_regs perf_arch_fetch_caller_regs() is exported for the overriden x86 version, but not for the generic weak version. As a general rule, weak functions should not have their symbol exported in the same file they are defined. So let's export it on trace_event_perf.c as it is used by trace events only. This fixes: ERROR: ".perf_arch_fetch_caller_regs" [fs/xfs/xfs.ko] undefined! ERROR: ".perf_arch_fetch_caller_regs" [arch/powerpc/platforms/cell/spufs/spufs.ko] undefined! -v2: And also only build it if trace events are enabled. -v3: Fix changelog mistake Reported-by: Stephen Rothwell Signed-off-by: Frederic Weisbecker Cc: Peter Zijlstra Cc: Xiao Guangrong Cc: Paul Mackerras LKML-Reference: <1268697902-9518-1-git-send-regression-fweisbec@gmail.com> Signed-off-by: Ingo Molnar --- kernel/trace/trace_event_perf.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_event_perf.c b/kernel/trace/trace_event_perf.c index 0709e4f75114..7d79a10c3cde 100644 --- a/kernel/trace/trace_event_perf.c +++ b/kernel/trace/trace_event_perf.c @@ -12,6 +12,8 @@ DEFINE_PER_CPU(struct pt_regs, perf_trace_regs); EXPORT_PER_CPU_SYMBOL_GPL(perf_trace_regs); +EXPORT_SYMBOL_GPL(perf_arch_fetch_caller_regs); + static char *perf_trace_buf; static char *perf_trace_buf_nmi; -- cgit v1.2.3 From 2271048d1b3b0aabf83d25b29c20646dcabedc05 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Thu, 18 Mar 2010 17:54:19 -0400 Subject: ring-buffer: Do 8 byte alignment for 64 bit that can not handle 4 byte align The ring buffer uses 4 byte alignment while recording events into the buffer, even on 64bit machines. This saves space when there are lots of events being recorded at 4 byte boundaries. The ring buffer has a zero copy method to write into the buffer, with the reserving of space and then committing it. This may cause problems when writing an 8 byte word into a 4 byte alignment (not 8). For x86 and PPC this is not an issue, but on some architectures this would cause an out-of-alignment exception. This patch uses CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS to determine if it is OK to use 4 byte alignments on 64 bit machines. If it is not, it forces the ring buffer event header to be 8 bytes and not 4, and will align the length of the data to be 8 byte aligned. This keeps the data payload at 8 byte alignments and will allow these machines to run without issue. The trick to this is that the header can be either 4 bytes or 8 bytes depending on the length of the data payload. The 4 byte header has a length field that supports up to 112 bytes. If the length of the data is more than 112, the length field is set to zero, and the actual length is stored in the next 4 bytes after the header. When CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS is not set, the code forces zero in the 4 byte header forcing the length to be stored in the 4 byte array, even with a small data load. It also forces the length of the data load to be 8 byte aligned. The combination of these two guarantee that the data is always at 8 byte alignment. Tested-by: Frederic Weisbecker (on sparc64) Reported-by: Frederic Weisbecker Acked-by: David S. Miller Signed-off-by: Steven Rostedt --- kernel/trace/ring_buffer.c | 14 +++++++++++--- 1 file changed, 11 insertions(+), 3 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/ring_buffer.c b/kernel/trace/ring_buffer.c index 05a9f83b8819..d1187ef20caf 100644 --- a/kernel/trace/ring_buffer.c +++ b/kernel/trace/ring_buffer.c @@ -207,6 +207,14 @@ EXPORT_SYMBOL_GPL(tracing_is_on); #define RB_MAX_SMALL_DATA (RB_ALIGNMENT * RINGBUF_TYPE_DATA_TYPE_LEN_MAX) #define RB_EVNT_MIN_SIZE 8U /* two 32bit words */ +#if !defined(CONFIG_64BIT) || defined(CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS) +# define RB_FORCE_8BYTE_ALIGNMENT 0 +# define RB_ARCH_ALIGNMENT RB_ALIGNMENT +#else +# define RB_FORCE_8BYTE_ALIGNMENT 1 +# define RB_ARCH_ALIGNMENT 8U +#endif + /* define RINGBUF_TYPE_DATA for 'case RINGBUF_TYPE_DATA:' */ #define RINGBUF_TYPE_DATA 0 ... RINGBUF_TYPE_DATA_TYPE_LEN_MAX @@ -1547,7 +1555,7 @@ rb_update_event(struct ring_buffer_event *event, case 0: length -= RB_EVNT_HDR_SIZE; - if (length > RB_MAX_SMALL_DATA) + if (length > RB_MAX_SMALL_DATA || RB_FORCE_8BYTE_ALIGNMENT) event->array[0] = length; else event->type_len = DIV_ROUND_UP(length, RB_ALIGNMENT); @@ -1722,11 +1730,11 @@ static unsigned rb_calculate_event_length(unsigned length) if (!length) length = 1; - if (length > RB_MAX_SMALL_DATA) + if (length > RB_MAX_SMALL_DATA || RB_FORCE_8BYTE_ALIGNMENT) length += sizeof(event.array[0]); length += RB_EVNT_HDR_SIZE; - length = ALIGN(length, RB_ALIGNMENT); + length = ALIGN(length, RB_ARCH_ALIGNMENT); return length; } -- cgit v1.2.3 From faa4602e47690fb11221e00f9b9697c8dc0d4b19 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Thu, 25 Mar 2010 14:51:50 +0100 Subject: x86, perf, bts, mm: Delete the never used BTS-ptrace code Support for the PMU's BTS features has been upstreamed in v2.6.32, but we still have the old and disabled ptrace-BTS, as Linus noticed it not so long ago. It's buggy: TIF_DEBUGCTLMSR is trampling all over that MSR without regard for other uses (perf) and doesn't provide the flexibility needed for perf either. Its users are ptrace-block-step and ptrace-bts, since ptrace-bts was never used and ptrace-block-step can be implemented using a much simpler approach. So axe all 3000 lines of it. That includes the *locked_memory*() APIs in mm/mlock.c as well. Reported-by: Linus Torvalds Signed-off-by: Peter Zijlstra Cc: Roland McGrath Cc: Oleg Nesterov Cc: Markus Metzger Cc: Steven Rostedt Cc: Andrew Morton LKML-Reference: <20100325135413.938004390@chello.nl> Signed-off-by: Ingo Molnar --- kernel/trace/Kconfig | 11 -- kernel/trace/Makefile | 1 - kernel/trace/trace.h | 4 - kernel/trace/trace_entries.h | 12 -- kernel/trace/trace_hw_branches.c | 312 --------------------------------------- kernel/trace/trace_selftest.c | 57 ------- 6 files changed, 397 deletions(-) delete mode 100644 kernel/trace/trace_hw_branches.c (limited to 'kernel/trace') diff --git a/kernel/trace/Kconfig b/kernel/trace/Kconfig index 13e13d428cd3..8b1797c4545b 100644 --- a/kernel/trace/Kconfig +++ b/kernel/trace/Kconfig @@ -44,9 +44,6 @@ config HAVE_FTRACE_MCOUNT_RECORD help See Documentation/trace/ftrace-design.txt -config HAVE_HW_BRANCH_TRACER - bool - config HAVE_SYSCALL_TRACEPOINTS bool help @@ -374,14 +371,6 @@ config STACK_TRACER Say N if unsure. -config HW_BRANCH_TRACER - depends on HAVE_HW_BRANCH_TRACER - bool "Trace hw branches" - select GENERIC_TRACER - help - This tracer records all branches on the system in a circular - buffer, giving access to the last N branches for each cpu. - config KMEMTRACE bool "Trace SLAB allocations" select GENERIC_TRACER diff --git a/kernel/trace/Makefile b/kernel/trace/Makefile index 78edc6490038..ffb1a5b0550e 100644 --- a/kernel/trace/Makefile +++ b/kernel/trace/Makefile @@ -41,7 +41,6 @@ obj-$(CONFIG_MMIOTRACE) += trace_mmiotrace.o obj-$(CONFIG_BOOT_TRACER) += trace_boot.o obj-$(CONFIG_FUNCTION_GRAPH_TRACER) += trace_functions_graph.o obj-$(CONFIG_TRACE_BRANCH_PROFILING) += trace_branch.o -obj-$(CONFIG_HW_BRANCH_TRACER) += trace_hw_branches.o obj-$(CONFIG_KMEMTRACE) += kmemtrace.o obj-$(CONFIG_WORKQUEUE_TRACER) += trace_workqueue.o obj-$(CONFIG_BLK_DEV_IO_TRACE) += blktrace.o diff --git a/kernel/trace/trace.h b/kernel/trace/trace.h index 2825ef2c0b15..bec2c973ff0c 100644 --- a/kernel/trace/trace.h +++ b/kernel/trace/trace.h @@ -34,7 +34,6 @@ enum trace_type { TRACE_GRAPH_RET, TRACE_GRAPH_ENT, TRACE_USER_STACK, - TRACE_HW_BRANCHES, TRACE_KMEM_ALLOC, TRACE_KMEM_FREE, TRACE_BLK, @@ -229,7 +228,6 @@ extern void __ftrace_bad_type(void); TRACE_GRAPH_ENT); \ IF_ASSIGN(var, ent, struct ftrace_graph_ret_entry, \ TRACE_GRAPH_RET); \ - IF_ASSIGN(var, ent, struct hw_branch_entry, TRACE_HW_BRANCHES);\ IF_ASSIGN(var, ent, struct kmemtrace_alloc_entry, \ TRACE_KMEM_ALLOC); \ IF_ASSIGN(var, ent, struct kmemtrace_free_entry, \ @@ -467,8 +465,6 @@ extern int trace_selftest_startup_sysprof(struct tracer *trace, struct trace_array *tr); extern int trace_selftest_startup_branch(struct tracer *trace, struct trace_array *tr); -extern int trace_selftest_startup_hw_branches(struct tracer *trace, - struct trace_array *tr); extern int trace_selftest_startup_ksym(struct tracer *trace, struct trace_array *tr); #endif /* CONFIG_FTRACE_STARTUP_TEST */ diff --git a/kernel/trace/trace_entries.h b/kernel/trace/trace_entries.h index c16a08f399df..dc008c1240da 100644 --- a/kernel/trace/trace_entries.h +++ b/kernel/trace/trace_entries.h @@ -318,18 +318,6 @@ FTRACE_ENTRY(branch, trace_branch, __entry->func, __entry->file, __entry->correct) ); -FTRACE_ENTRY(hw_branch, hw_branch_entry, - - TRACE_HW_BRANCHES, - - F_STRUCT( - __field( u64, from ) - __field( u64, to ) - ), - - F_printk("from: %llx to: %llx", __entry->from, __entry->to) -); - FTRACE_ENTRY(kmem_alloc, kmemtrace_alloc_entry, TRACE_KMEM_ALLOC, diff --git a/kernel/trace/trace_hw_branches.c b/kernel/trace/trace_hw_branches.c deleted file mode 100644 index 7b97000745f5..000000000000 --- a/kernel/trace/trace_hw_branches.c +++ /dev/null @@ -1,312 +0,0 @@ -/* - * h/w branch tracer for x86 based on BTS - * - * Copyright (C) 2008-2009 Intel Corporation. - * Markus Metzger , 2008-2009 - */ -#include -#include -#include -#include -#include -#include -#include - -#include - -#include "trace_output.h" -#include "trace.h" - - -#define BTS_BUFFER_SIZE (1 << 13) - -static DEFINE_PER_CPU(struct bts_tracer *, hwb_tracer); -static DEFINE_PER_CPU(unsigned char[BTS_BUFFER_SIZE], hwb_buffer); - -#define this_tracer per_cpu(hwb_tracer, smp_processor_id()) - -static int trace_hw_branches_enabled __read_mostly; -static int trace_hw_branches_suspended __read_mostly; -static struct trace_array *hw_branch_trace __read_mostly; - - -static void bts_trace_init_cpu(int cpu) -{ - per_cpu(hwb_tracer, cpu) = - ds_request_bts_cpu(cpu, per_cpu(hwb_buffer, cpu), - BTS_BUFFER_SIZE, NULL, (size_t)-1, - BTS_KERNEL); - - if (IS_ERR(per_cpu(hwb_tracer, cpu))) - per_cpu(hwb_tracer, cpu) = NULL; -} - -static int bts_trace_init(struct trace_array *tr) -{ - int cpu; - - hw_branch_trace = tr; - trace_hw_branches_enabled = 0; - - get_online_cpus(); - for_each_online_cpu(cpu) { - bts_trace_init_cpu(cpu); - - if (likely(per_cpu(hwb_tracer, cpu))) - trace_hw_branches_enabled = 1; - } - trace_hw_branches_suspended = 0; - put_online_cpus(); - - /* If we could not enable tracing on a single cpu, we fail. */ - return trace_hw_branches_enabled ? 0 : -EOPNOTSUPP; -} - -static void bts_trace_reset(struct trace_array *tr) -{ - int cpu; - - get_online_cpus(); - for_each_online_cpu(cpu) { - if (likely(per_cpu(hwb_tracer, cpu))) { - ds_release_bts(per_cpu(hwb_tracer, cpu)); - per_cpu(hwb_tracer, cpu) = NULL; - } - } - trace_hw_branches_enabled = 0; - trace_hw_branches_suspended = 0; - put_online_cpus(); -} - -static void bts_trace_start(struct trace_array *tr) -{ - int cpu; - - get_online_cpus(); - for_each_online_cpu(cpu) - if (likely(per_cpu(hwb_tracer, cpu))) - ds_resume_bts(per_cpu(hwb_tracer, cpu)); - trace_hw_branches_suspended = 0; - put_online_cpus(); -} - -static void bts_trace_stop(struct trace_array *tr) -{ - int cpu; - - get_online_cpus(); - for_each_online_cpu(cpu) - if (likely(per_cpu(hwb_tracer, cpu))) - ds_suspend_bts(per_cpu(hwb_tracer, cpu)); - trace_hw_branches_suspended = 1; - put_online_cpus(); -} - -static int __cpuinit bts_hotcpu_handler(struct notifier_block *nfb, - unsigned long action, void *hcpu) -{ - int cpu = (long)hcpu; - - switch (action) { - case CPU_ONLINE: - case CPU_DOWN_FAILED: - /* The notification is sent with interrupts enabled. */ - if (trace_hw_branches_enabled) { - bts_trace_init_cpu(cpu); - - if (trace_hw_branches_suspended && - likely(per_cpu(hwb_tracer, cpu))) - ds_suspend_bts(per_cpu(hwb_tracer, cpu)); - } - break; - - case CPU_DOWN_PREPARE: - /* The notification is sent with interrupts enabled. */ - if (likely(per_cpu(hwb_tracer, cpu))) { - ds_release_bts(per_cpu(hwb_tracer, cpu)); - per_cpu(hwb_tracer, cpu) = NULL; - } - } - - return NOTIFY_DONE; -} - -static struct notifier_block bts_hotcpu_notifier __cpuinitdata = { - .notifier_call = bts_hotcpu_handler -}; - -static void bts_trace_print_header(struct seq_file *m) -{ - seq_puts(m, "# CPU# TO <- FROM\n"); -} - -static enum print_line_t bts_trace_print_line(struct trace_iterator *iter) -{ - unsigned long symflags = TRACE_ITER_SYM_OFFSET; - struct trace_entry *entry = iter->ent; - struct trace_seq *seq = &iter->seq; - struct hw_branch_entry *it; - - trace_assign_type(it, entry); - - if (entry->type == TRACE_HW_BRANCHES) { - if (trace_seq_printf(seq, "%4d ", iter->cpu) && - seq_print_ip_sym(seq, it->to, symflags) && - trace_seq_printf(seq, "\t <- ") && - seq_print_ip_sym(seq, it->from, symflags) && - trace_seq_printf(seq, "\n")) - return TRACE_TYPE_HANDLED; - return TRACE_TYPE_PARTIAL_LINE; - } - return TRACE_TYPE_UNHANDLED; -} - -void trace_hw_branch(u64 from, u64 to) -{ - struct ftrace_event_call *call = &event_hw_branch; - struct trace_array *tr = hw_branch_trace; - struct ring_buffer_event *event; - struct ring_buffer *buf; - struct hw_branch_entry *entry; - unsigned long irq1; - int cpu; - - if (unlikely(!tr)) - return; - - if (unlikely(!trace_hw_branches_enabled)) - return; - - local_irq_save(irq1); - cpu = raw_smp_processor_id(); - if (atomic_inc_return(&tr->data[cpu]->disabled) != 1) - goto out; - - buf = tr->buffer; - event = trace_buffer_lock_reserve(buf, TRACE_HW_BRANCHES, - sizeof(*entry), 0, 0); - if (!event) - goto out; - entry = ring_buffer_event_data(event); - tracing_generic_entry_update(&entry->ent, 0, from); - entry->ent.type = TRACE_HW_BRANCHES; - entry->from = from; - entry->to = to; - if (!filter_check_discard(call, entry, buf, event)) - trace_buffer_unlock_commit(buf, event, 0, 0); - - out: - atomic_dec(&tr->data[cpu]->disabled); - local_irq_restore(irq1); -} - -static void trace_bts_at(const struct bts_trace *trace, void *at) -{ - struct bts_struct bts; - int err = 0; - - WARN_ON_ONCE(!trace->read); - if (!trace->read) - return; - - err = trace->read(this_tracer, at, &bts); - if (err < 0) - return; - - switch (bts.qualifier) { - case BTS_BRANCH: - trace_hw_branch(bts.variant.lbr.from, bts.variant.lbr.to); - break; - } -} - -/* - * Collect the trace on the current cpu and write it into the ftrace buffer. - * - * pre: tracing must be suspended on the current cpu - */ -static void trace_bts_cpu(void *arg) -{ - struct trace_array *tr = (struct trace_array *)arg; - const struct bts_trace *trace; - unsigned char *at; - - if (unlikely(!tr)) - return; - - if (unlikely(atomic_read(&tr->data[raw_smp_processor_id()]->disabled))) - return; - - if (unlikely(!this_tracer)) - return; - - trace = ds_read_bts(this_tracer); - if (!trace) - return; - - for (at = trace->ds.top; (void *)at < trace->ds.end; - at += trace->ds.size) - trace_bts_at(trace, at); - - for (at = trace->ds.begin; (void *)at < trace->ds.top; - at += trace->ds.size) - trace_bts_at(trace, at); -} - -static void trace_bts_prepare(struct trace_iterator *iter) -{ - int cpu; - - get_online_cpus(); - for_each_online_cpu(cpu) - if (likely(per_cpu(hwb_tracer, cpu))) - ds_suspend_bts(per_cpu(hwb_tracer, cpu)); - /* - * We need to collect the trace on the respective cpu since ftrace - * implicitly adds the record for the current cpu. - * Once that is more flexible, we could collect the data from any cpu. - */ - on_each_cpu(trace_bts_cpu, iter->tr, 1); - - for_each_online_cpu(cpu) - if (likely(per_cpu(hwb_tracer, cpu))) - ds_resume_bts(per_cpu(hwb_tracer, cpu)); - put_online_cpus(); -} - -static void trace_bts_close(struct trace_iterator *iter) -{ - tracing_reset_online_cpus(iter->tr); -} - -void trace_hw_branch_oops(void) -{ - if (this_tracer) { - ds_suspend_bts_noirq(this_tracer); - trace_bts_cpu(hw_branch_trace); - ds_resume_bts_noirq(this_tracer); - } -} - -struct tracer bts_tracer __read_mostly = -{ - .name = "hw-branch-tracer", - .init = bts_trace_init, - .reset = bts_trace_reset, - .print_header = bts_trace_print_header, - .print_line = bts_trace_print_line, - .start = bts_trace_start, - .stop = bts_trace_stop, - .open = trace_bts_prepare, - .close = trace_bts_close, -#ifdef CONFIG_FTRACE_SELFTEST - .selftest = trace_selftest_startup_hw_branches, -#endif /* CONFIG_FTRACE_SELFTEST */ -}; - -__init static int init_bts_trace(void) -{ - register_hotcpu_notifier(&bts_hotcpu_notifier); - return register_tracer(&bts_tracer); -} -device_initcall(init_bts_trace); diff --git a/kernel/trace/trace_selftest.c b/kernel/trace/trace_selftest.c index 280fea470d67..a7084e7c0427 100644 --- a/kernel/trace/trace_selftest.c +++ b/kernel/trace/trace_selftest.c @@ -16,7 +16,6 @@ static inline int trace_valid_entry(struct trace_entry *entry) case TRACE_BRANCH: case TRACE_GRAPH_ENT: case TRACE_GRAPH_RET: - case TRACE_HW_BRANCHES: case TRACE_KSYM: return 1; } @@ -754,62 +753,6 @@ trace_selftest_startup_branch(struct tracer *trace, struct trace_array *tr) } #endif /* CONFIG_BRANCH_TRACER */ -#ifdef CONFIG_HW_BRANCH_TRACER -int -trace_selftest_startup_hw_branches(struct tracer *trace, - struct trace_array *tr) -{ - struct trace_iterator *iter; - struct tracer tracer; - unsigned long count; - int ret; - - if (!trace->open) { - printk(KERN_CONT "missing open function..."); - return -1; - } - - ret = tracer_init(trace, tr); - if (ret) { - warn_failed_init_tracer(trace, ret); - return ret; - } - - /* - * The hw-branch tracer needs to collect the trace from the various - * cpu trace buffers - before tracing is stopped. - */ - iter = kzalloc(sizeof(*iter), GFP_KERNEL); - if (!iter) - return -ENOMEM; - - memcpy(&tracer, trace, sizeof(tracer)); - - iter->trace = &tracer; - iter->tr = tr; - iter->pos = -1; - mutex_init(&iter->mutex); - - trace->open(iter); - - mutex_destroy(&iter->mutex); - kfree(iter); - - tracing_stop(); - - ret = trace_test_buffer(tr, &count); - trace->reset(tr); - tracing_start(); - - if (!ret && !count) { - printk(KERN_CONT "no entries found.."); - ret = -1; - } - - return ret; -} -#endif /* CONFIG_HW_BRANCH_TRACER */ - #ifdef CONFIG_KSYM_TRACER static int ksym_selftest_dummy; -- cgit v1.2.3 From e36673ec5126f15a8cddf6049aede7bdcf484c26 Mon Sep 17 00:00:00 2001 From: Li Zefan Date: Wed, 24 Mar 2010 10:57:37 +0800 Subject: tracing: Fix lockdep warning in global_clock() # echo 1 > events/enable # echo global > trace_clock ------------[ cut here ]------------ WARNING: at kernel/lockdep.c:3162 check_flags+0xb2/0x190() ... ---[ end trace 3f86734a89416623 ]--- possible reason: unannotated irqs-on. ... There's no reason to use the raw_local_irq_save() in trace_clock_global. The local_irq_save() version is fine, and does not cause the bug in lockdep. Acked-by: Peter Zijlstra Signed-off-by: Li Zefan LKML-Reference: <4BA97FA1.7030606@cn.fujitsu.com> Signed-off-by: Steven Rostedt --- kernel/trace/trace_clock.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_clock.c b/kernel/trace/trace_clock.c index 6fbfb8f417b9..9d589d8dcd1a 100644 --- a/kernel/trace/trace_clock.c +++ b/kernel/trace/trace_clock.c @@ -84,7 +84,7 @@ u64 notrace trace_clock_global(void) int this_cpu; u64 now; - raw_local_irq_save(flags); + local_irq_save(flags); this_cpu = raw_smp_processor_id(); now = cpu_clock(this_cpu); @@ -110,7 +110,7 @@ u64 notrace trace_clock_global(void) arch_spin_unlock(&trace_clock_struct.lock); out: - raw_local_irq_restore(flags); + local_irq_restore(flags); return now; } -- cgit v1.2.3 From 292f60c0c4ab44aa2d589ba03c12e64a3b3c5e38 Mon Sep 17 00:00:00 2001 From: Julia Lawall Date: Mon, 29 Mar 2010 17:37:02 +0200 Subject: ring-buffer: Add missing unlock In some error handling cases the lock is not unlocked. The return is converted to a goto, to share the unlock at the end of the function. A simplified version of the semantic patch that finds this problem is as follows: (http://coccinelle.lip6.fr/) // @r exists@ expression E1; identifier f; @@ f (...) { <+... * spin_lock_irq (E1,...); ... when != E1 * return ...; ...+> } // Signed-off-by: Julia Lawall LKML-Reference: Signed-off-by: Steven Rostedt --- kernel/trace/ring_buffer.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/ring_buffer.c b/kernel/trace/ring_buffer.c index d1187ef20caf..9a0f9bf6a37b 100644 --- a/kernel/trace/ring_buffer.c +++ b/kernel/trace/ring_buffer.c @@ -1209,18 +1209,19 @@ rb_remove_pages(struct ring_buffer_per_cpu *cpu_buffer, unsigned nr_pages) for (i = 0; i < nr_pages; i++) { if (RB_WARN_ON(cpu_buffer, list_empty(cpu_buffer->pages))) - return; + goto out; p = cpu_buffer->pages->next; bpage = list_entry(p, struct buffer_page, list); list_del_init(&bpage->list); free_buffer_page(bpage); } if (RB_WARN_ON(cpu_buffer, list_empty(cpu_buffer->pages))) - return; + goto out; rb_reset_cpu(cpu_buffer); rb_check_pages(cpu_buffer); +out: spin_unlock_irq(&cpu_buffer->reader_lock); } @@ -1237,7 +1238,7 @@ rb_insert_pages(struct ring_buffer_per_cpu *cpu_buffer, for (i = 0; i < nr_pages; i++) { if (RB_WARN_ON(cpu_buffer, list_empty(pages))) - return; + goto out; p = pages->next; bpage = list_entry(p, struct buffer_page, list); list_del_init(&bpage->list); @@ -1246,6 +1247,7 @@ rb_insert_pages(struct ring_buffer_per_cpu *cpu_buffer, rb_reset_cpu(cpu_buffer); rb_check_pages(cpu_buffer); +out: spin_unlock_irq(&cpu_buffer->reader_lock); } -- cgit v1.2.3 From 5a0e3ad6af8660be21ca98a971cd00f331318c05 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Wed, 24 Mar 2010 17:04:11 +0900 Subject: include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h percpu.h is included by sched.h and module.h and thus ends up being included when building most .c files. percpu.h includes slab.h which in turn includes gfp.h making everything defined by the two files universally available and complicating inclusion dependencies. percpu.h -> slab.h dependency is about to be removed. Prepare for this change by updating users of gfp and slab facilities include those headers directly instead of assuming availability. As this conversion needs to touch large number of source files, the following script is used as the basis of conversion. http://userweb.kernel.org/~tj/misc/slabh-sweep.py The script does the followings. * Scan files for gfp and slab usages and update includes such that only the necessary includes are there. ie. if only gfp is used, gfp.h, if slab is used, slab.h. * When the script inserts a new include, it looks at the include blocks and try to put the new include such that its order conforms to its surrounding. It's put in the include block which contains core kernel includes, in the same order that the rest are ordered - alphabetical, Christmas tree, rev-Xmas-tree or at the end if there doesn't seem to be any matching order. * If the script can't find a place to put a new include (mostly because the file doesn't have fitting include block), it prints out an error message indicating which .h file needs to be added to the file. The conversion was done in the following steps. 1. The initial automatic conversion of all .c files updated slightly over 4000 files, deleting around 700 includes and adding ~480 gfp.h and ~3000 slab.h inclusions. The script emitted errors for ~400 files. 2. Each error was manually checked. Some didn't need the inclusion, some needed manual addition while adding it to implementation .h or embedding .c file was more appropriate for others. This step added inclusions to around 150 files. 3. The script was run again and the output was compared to the edits from #2 to make sure no file was left behind. 4. Several build tests were done and a couple of problems were fixed. e.g. lib/decompress_*.c used malloc/free() wrappers around slab APIs requiring slab.h to be added manually. 5. The script was run on all .h files but without automatically editing them as sprinkling gfp.h and slab.h inclusions around .h files could easily lead to inclusion dependency hell. Most gfp.h inclusion directives were ignored as stuff from gfp.h was usually wildly available and often used in preprocessor macros. Each slab.h inclusion directive was examined and added manually as necessary. 6. percpu.h was updated not to include slab.h. 7. Build test were done on the following configurations and failures were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my distributed build env didn't work with gcov compiles) and a few more options had to be turned off depending on archs to make things build (like ipr on powerpc/64 which failed due to missing writeq). * x86 and x86_64 UP and SMP allmodconfig and a custom test config. * powerpc and powerpc64 SMP allmodconfig * sparc and sparc64 SMP allmodconfig * ia64 SMP allmodconfig * s390 SMP allmodconfig * alpha SMP allmodconfig * um on x86_64 SMP allmodconfig 8. percpu.h modifications were reverted so that it could be applied as a separate patch and serve as bisection point. Given the fact that I had only a couple of failures from tests on step 6, I'm fairly confident about the coverage of this conversion patch. If there is a breakage, it's likely to be something in one of the arch headers which should be easily discoverable easily on most builds of the specific arch. Signed-off-by: Tejun Heo Guess-its-ok-by: Christoph Lameter Cc: Ingo Molnar Cc: Lee Schermerhorn --- kernel/trace/blktrace.c | 1 + kernel/trace/ftrace.c | 1 + kernel/trace/power-traces.c | 1 - kernel/trace/ring_buffer.c | 1 + kernel/trace/trace.c | 2 +- kernel/trace/trace_events.c | 1 + kernel/trace/trace_events_filter.c | 1 + kernel/trace/trace_functions_graph.c | 1 + kernel/trace/trace_ksym.c | 1 + kernel/trace/trace_mmiotrace.c | 1 + kernel/trace/trace_selftest.c | 1 + kernel/trace/trace_stat.c | 1 + kernel/trace/trace_syscalls.c | 1 + kernel/trace/trace_workqueue.c | 1 + 14 files changed, 13 insertions(+), 2 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/blktrace.c b/kernel/trace/blktrace.c index 07f945a99430..b3bc91a3f510 100644 --- a/kernel/trace/blktrace.c +++ b/kernel/trace/blktrace.c @@ -21,6 +21,7 @@ #include #include #include +#include #include #include #include diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index d9062f5cc0c0..2404b59b3097 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -24,6 +24,7 @@ #include #include #include +#include #include #include #include diff --git a/kernel/trace/power-traces.c b/kernel/trace/power-traces.c index 9f4f565b01e6..a22582a06161 100644 --- a/kernel/trace/power-traces.c +++ b/kernel/trace/power-traces.c @@ -9,7 +9,6 @@ #include #include #include -#include #define CREATE_TRACE_POINTS #include diff --git a/kernel/trace/ring_buffer.c b/kernel/trace/ring_buffer.c index d1187ef20caf..2c839ca5e5ce 100644 --- a/kernel/trace/ring_buffer.c +++ b/kernel/trace/ring_buffer.c @@ -14,6 +14,7 @@ #include #include #include +#include #include #include #include diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index 3ec2ee6f6560..44f916a04065 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -33,10 +33,10 @@ #include #include #include +#include #include #include #include -#include #include #include "trace.h" diff --git a/kernel/trace/trace_events.c b/kernel/trace/trace_events.c index beab8bf2f310..c697c7043349 100644 --- a/kernel/trace/trace_events.c +++ b/kernel/trace/trace_events.c @@ -15,6 +15,7 @@ #include #include #include +#include #include #include diff --git a/kernel/trace/trace_events_filter.c b/kernel/trace/trace_events_filter.c index 4615f62a04f1..88c0b6dbd7fe 100644 --- a/kernel/trace/trace_events_filter.c +++ b/kernel/trace/trace_events_filter.c @@ -22,6 +22,7 @@ #include #include #include +#include #include "trace.h" #include "trace_output.h" diff --git a/kernel/trace/trace_functions_graph.c b/kernel/trace/trace_functions_graph.c index e6989d9b44da..9aed1a5cf553 100644 --- a/kernel/trace/trace_functions_graph.c +++ b/kernel/trace/trace_functions_graph.c @@ -9,6 +9,7 @@ #include #include #include +#include #include #include "trace.h" diff --git a/kernel/trace/trace_ksym.c b/kernel/trace/trace_ksym.c index 94103cdcf9d8..d59cd6879477 100644 --- a/kernel/trace/trace_ksym.c +++ b/kernel/trace/trace_ksym.c @@ -23,6 +23,7 @@ #include #include #include +#include #include #include "trace_output.h" diff --git a/kernel/trace/trace_mmiotrace.c b/kernel/trace/trace_mmiotrace.c index 0acd834659ed..017fa376505d 100644 --- a/kernel/trace/trace_mmiotrace.c +++ b/kernel/trace/trace_mmiotrace.c @@ -9,6 +9,7 @@ #include #include #include +#include #include #include diff --git a/kernel/trace/trace_selftest.c b/kernel/trace/trace_selftest.c index 280fea470d67..81003b4d617f 100644 --- a/kernel/trace/trace_selftest.c +++ b/kernel/trace/trace_selftest.c @@ -3,6 +3,7 @@ #include #include #include +#include static inline int trace_valid_entry(struct trace_entry *entry) { diff --git a/kernel/trace/trace_stat.c b/kernel/trace/trace_stat.c index a4bb239eb987..96cffb269e73 100644 --- a/kernel/trace/trace_stat.c +++ b/kernel/trace/trace_stat.c @@ -10,6 +10,7 @@ #include +#include #include #include #include "trace_stat.h" diff --git a/kernel/trace/trace_syscalls.c b/kernel/trace/trace_syscalls.c index 33c2a5b769dc..4d6d711717f2 100644 --- a/kernel/trace/trace_syscalls.c +++ b/kernel/trace/trace_syscalls.c @@ -1,5 +1,6 @@ #include #include +#include #include #include #include diff --git a/kernel/trace/trace_workqueue.c b/kernel/trace/trace_workqueue.c index 40cafb07dffd..cc2d2faa7d9e 100644 --- a/kernel/trace/trace_workqueue.c +++ b/kernel/trace/trace_workqueue.c @@ -9,6 +9,7 @@ #include #include #include +#include #include #include "trace_stat.h" #include "trace.h" -- cgit v1.2.3 From 66a8cb95ed04025664d1db4e952155ee1dccd048 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Wed, 31 Mar 2010 13:21:56 -0400 Subject: ring-buffer: Add place holder recording of dropped events Currently, when the ring buffer drops events, it does not record the fact that it did so. It does inform the writer that the event was dropped by returning a NULL event, but it does not put in any place holder where the event was dropped. This is not a trivial thing to add because the ring buffer mostly runs in overwrite (flight recorder) mode. That is, when the ring buffer is full, new data will overwrite old data. In a produce/consumer mode, where new data is simply dropped when the ring buffer is full, it is trivial to add the placeholder for dropped events. When there's more room to write new data, then a special event can be added to notify the reader about the dropped events. But in overwrite mode, any new write can overwrite events. A place holder can not be inserted into the ring buffer since there never may be room. A reader could also come in at anytime and miss the placeholder. Luckily, the way the ring buffer works, the read side can find out if events were lost or not, and how many events. Everytime a write takes place, if it overwrites the header page (the next read) it updates a "overrun" variable that keeps track of the number of lost events. When a reader swaps out a page from the ring buffer, it can record this number, perfom the swap, and then check to see if the number changed, and take the diff if it has, which would be the number of events dropped. This can be stored by the reader and returned to callers of the reader. Since the reader page swap will fail if the writer moved the head page since the time the reader page set up the swap, this gives room to record the overruns without worrying about races. If the reader sets up the pages, records the overrun, than performs the swap, if the swap succeeds, then the overrun variable has not been updated since the setup before the swap. For binary readers of the ring buffer, a flag is set in the header of each sub page (sub buffer) of the ring buffer. This flag is embedded in the size field of the data on the sub buffer, in the 31st bit (the size can be 32 or 64 bits depending on the architecture), but only 27 bits needs to be used for the actual size (less actually). We could add a new field in the sub buffer header to also record the number of events dropped since the last read, but this will change the format of the binary ring buffer a bit too much. Perhaps this change can be made if the information on the number of events dropped is considered important enough. Note, the notification of dropped events is only used by consuming reads or peeking at the ring buffer. Iterating over the ring buffer does not keep this information because the necessary data is only available when a page swap is made, and the iterator does not swap out pages. Cc: Robert Richter Cc: Andi Kleen Cc: Li Zefan Cc: Arnaldo Carvalho de Melo Cc: "Luis Claudio R. Goncalves" Cc: Frederic Weisbecker Signed-off-by: Steven Rostedt --- kernel/trace/ring_buffer.c | 72 +++++++++++++++++++++++++++++++++--- kernel/trace/ring_buffer_benchmark.c | 2 +- kernel/trace/trace.c | 4 +- kernel/trace/trace_functions_graph.c | 5 ++- kernel/trace/trace_selftest.c | 2 +- 5 files changed, 73 insertions(+), 12 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/ring_buffer.c b/kernel/trace/ring_buffer.c index d1187ef20caf..8295650444c5 100644 --- a/kernel/trace/ring_buffer.c +++ b/kernel/trace/ring_buffer.c @@ -318,6 +318,9 @@ EXPORT_SYMBOL_GPL(ring_buffer_event_data); #define TS_MASK ((1ULL << TS_SHIFT) - 1) #define TS_DELTA_TEST (~TS_MASK) +/* Flag when events were overwritten */ +#define RB_MISSED_EVENTS (1 << 31) + struct buffer_data_page { u64 time_stamp; /* page time stamp */ local_t commit; /* write committed index */ @@ -416,6 +419,12 @@ int ring_buffer_print_page_header(struct trace_seq *s) (unsigned int)sizeof(field.commit), (unsigned int)is_signed_type(long)); + ret = trace_seq_printf(s, "\tfield: int overwrite;\t" + "offset:%u;\tsize:%u;\tsigned:%u;\n", + (unsigned int)offsetof(typeof(field), commit), + 1, + (unsigned int)is_signed_type(long)); + ret = trace_seq_printf(s, "\tfield: char data;\t" "offset:%u;\tsize:%u;\tsigned:%u;\n", (unsigned int)offsetof(typeof(field), data), @@ -439,6 +448,8 @@ struct ring_buffer_per_cpu { struct buffer_page *tail_page; /* write to tail */ struct buffer_page *commit_page; /* committed pages */ struct buffer_page *reader_page; + unsigned long lost_events; + unsigned long last_overrun; local_t commit_overrun; local_t overrun; local_t entries; @@ -2835,6 +2846,7 @@ static struct buffer_page * rb_get_reader_page(struct ring_buffer_per_cpu *cpu_buffer) { struct buffer_page *reader = NULL; + unsigned long overwrite; unsigned long flags; int nr_loops = 0; int ret; @@ -2895,6 +2907,18 @@ rb_get_reader_page(struct ring_buffer_per_cpu *cpu_buffer) /* The reader page will be pointing to the new head */ rb_set_list_to_head(cpu_buffer, &cpu_buffer->reader_page->list); + /* + * We want to make sure we read the overruns after we set up our + * pointers to the next object. The writer side does a + * cmpxchg to cross pages which acts as the mb on the writer + * side. Note, the reader will constantly fail the swap + * while the writer is updating the pointers, so this + * guarantees that the overwrite recorded here is the one we + * want to compare with the last_overrun. + */ + smp_mb(); + overwrite = local_read(&(cpu_buffer->overrun)); + /* * Here's the tricky part. * @@ -2926,6 +2950,11 @@ rb_get_reader_page(struct ring_buffer_per_cpu *cpu_buffer) cpu_buffer->reader_page = reader; rb_reset_reader_page(cpu_buffer); + if (overwrite != cpu_buffer->last_overrun) { + cpu_buffer->lost_events = overwrite - cpu_buffer->last_overrun; + cpu_buffer->last_overrun = overwrite; + } + goto again; out: @@ -3002,8 +3031,14 @@ static void rb_advance_iter(struct ring_buffer_iter *iter) rb_advance_iter(iter); } +static int rb_lost_events(struct ring_buffer_per_cpu *cpu_buffer) +{ + return cpu_buffer->lost_events; +} + static struct ring_buffer_event * -rb_buffer_peek(struct ring_buffer_per_cpu *cpu_buffer, u64 *ts) +rb_buffer_peek(struct ring_buffer_per_cpu *cpu_buffer, u64 *ts, + unsigned long *lost_events) { struct ring_buffer_event *event; struct buffer_page *reader; @@ -3055,6 +3090,8 @@ rb_buffer_peek(struct ring_buffer_per_cpu *cpu_buffer, u64 *ts) ring_buffer_normalize_time_stamp(cpu_buffer->buffer, cpu_buffer->cpu, ts); } + if (lost_events) + *lost_events = rb_lost_events(cpu_buffer); return event; default: @@ -3165,12 +3202,14 @@ static inline int rb_ok_to_lock(void) * @buffer: The ring buffer to read * @cpu: The cpu to peak at * @ts: The timestamp counter of this event. + * @lost_events: a variable to store if events were lost (may be NULL) * * This will return the event that will be read next, but does * not consume the data. */ struct ring_buffer_event * -ring_buffer_peek(struct ring_buffer *buffer, int cpu, u64 *ts) +ring_buffer_peek(struct ring_buffer *buffer, int cpu, u64 *ts, + unsigned long *lost_events) { struct ring_buffer_per_cpu *cpu_buffer = buffer->buffers[cpu]; struct ring_buffer_event *event; @@ -3185,7 +3224,7 @@ ring_buffer_peek(struct ring_buffer *buffer, int cpu, u64 *ts) local_irq_save(flags); if (dolock) spin_lock(&cpu_buffer->reader_lock); - event = rb_buffer_peek(cpu_buffer, ts); + event = rb_buffer_peek(cpu_buffer, ts, lost_events); if (event && event->type_len == RINGBUF_TYPE_PADDING) rb_advance_reader(cpu_buffer); if (dolock) @@ -3227,13 +3266,17 @@ ring_buffer_iter_peek(struct ring_buffer_iter *iter, u64 *ts) /** * ring_buffer_consume - return an event and consume it * @buffer: The ring buffer to get the next event from + * @cpu: the cpu to read the buffer from + * @ts: a variable to store the timestamp (may be NULL) + * @lost_events: a variable to store if events were lost (may be NULL) * * Returns the next event in the ring buffer, and that event is consumed. * Meaning, that sequential reads will keep returning a different event, * and eventually empty the ring buffer if the producer is slower. */ struct ring_buffer_event * -ring_buffer_consume(struct ring_buffer *buffer, int cpu, u64 *ts) +ring_buffer_consume(struct ring_buffer *buffer, int cpu, u64 *ts, + unsigned long *lost_events) { struct ring_buffer_per_cpu *cpu_buffer; struct ring_buffer_event *event = NULL; @@ -3254,9 +3297,11 @@ ring_buffer_consume(struct ring_buffer *buffer, int cpu, u64 *ts) if (dolock) spin_lock(&cpu_buffer->reader_lock); - event = rb_buffer_peek(cpu_buffer, ts); - if (event) + event = rb_buffer_peek(cpu_buffer, ts, lost_events); + if (event) { + cpu_buffer->lost_events = 0; rb_advance_reader(cpu_buffer); + } if (dolock) spin_unlock(&cpu_buffer->reader_lock); @@ -3405,6 +3450,9 @@ rb_reset_cpu(struct ring_buffer_per_cpu *cpu_buffer) cpu_buffer->write_stamp = 0; cpu_buffer->read_stamp = 0; + cpu_buffer->lost_events = 0; + cpu_buffer->last_overrun = 0; + rb_head_page_activate(cpu_buffer); } @@ -3684,6 +3732,7 @@ int ring_buffer_read_page(struct ring_buffer *buffer, unsigned int commit; unsigned int read; u64 save_timestamp; + int missed_events = 0; int ret = -1; if (!cpumask_test_cpu(cpu, buffer->cpumask)) @@ -3716,6 +3765,10 @@ int ring_buffer_read_page(struct ring_buffer *buffer, read = reader->read; commit = rb_page_commit(reader); + /* Check if any events were dropped */ + if (cpu_buffer->lost_events) + missed_events = 1; + /* * If this page has been partially read or * if len is not big enough to read the rest of the page or @@ -3779,6 +3832,13 @@ int ring_buffer_read_page(struct ring_buffer *buffer, } ret = read; + cpu_buffer->lost_events = 0; + /* + * Set a flag in the commit field if we lost events + */ + if (missed_events) + local_add(RB_MISSED_EVENTS, &bpage->commit); + out_unlock: spin_unlock_irqrestore(&cpu_buffer->reader_lock, flags); diff --git a/kernel/trace/ring_buffer_benchmark.c b/kernel/trace/ring_buffer_benchmark.c index df74c7982255..dc56556b55a2 100644 --- a/kernel/trace/ring_buffer_benchmark.c +++ b/kernel/trace/ring_buffer_benchmark.c @@ -81,7 +81,7 @@ static enum event_status read_event(int cpu) int *entry; u64 ts; - event = ring_buffer_consume(buffer, cpu, &ts); + event = ring_buffer_consume(buffer, cpu, &ts, NULL); if (!event) return EVENT_DROPPED; diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index 3ec2ee6f6560..fabb0033a9be 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -1556,7 +1556,7 @@ peek_next_entry(struct trace_iterator *iter, int cpu, u64 *ts) if (buf_iter) event = ring_buffer_iter_peek(buf_iter, ts); else - event = ring_buffer_peek(iter->tr->buffer, cpu, ts); + event = ring_buffer_peek(iter->tr->buffer, cpu, ts, NULL); ftrace_enable_cpu(); @@ -1635,7 +1635,7 @@ static void trace_consume(struct trace_iterator *iter) { /* Don't allow ftrace to trace into the ring buffers */ ftrace_disable_cpu(); - ring_buffer_consume(iter->tr->buffer, iter->cpu, &iter->ts); + ring_buffer_consume(iter->tr->buffer, iter->cpu, &iter->ts, NULL); ftrace_enable_cpu(); } diff --git a/kernel/trace/trace_functions_graph.c b/kernel/trace/trace_functions_graph.c index e6989d9b44da..a7f75fb10aa4 100644 --- a/kernel/trace/trace_functions_graph.c +++ b/kernel/trace/trace_functions_graph.c @@ -489,9 +489,10 @@ get_return_for_leaf(struct trace_iterator *iter, * We need to consume the current entry to see * the next one. */ - ring_buffer_consume(iter->tr->buffer, iter->cpu, NULL); + ring_buffer_consume(iter->tr->buffer, iter->cpu, + NULL, NULL); event = ring_buffer_peek(iter->tr->buffer, iter->cpu, - NULL); + NULL, NULL); } if (!event) diff --git a/kernel/trace/trace_selftest.c b/kernel/trace/trace_selftest.c index 280fea470d67..e50180874c63 100644 --- a/kernel/trace/trace_selftest.c +++ b/kernel/trace/trace_selftest.c @@ -29,7 +29,7 @@ static int trace_test_buffer_cpu(struct trace_array *tr, int cpu) struct trace_entry *entry; unsigned int loops = 0; - while ((event = ring_buffer_consume(tr->buffer, cpu, NULL))) { + while ((event = ring_buffer_consume(tr->buffer, cpu, NULL, NULL))) { entry = ring_buffer_event_data(event); /* -- cgit v1.2.3 From bc21b478425ac73f66a5ec0b375a5e0d12d609ce Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Wed, 31 Mar 2010 19:49:26 -0400 Subject: tracing: Show the lost events in the trace_pipe output Now that the ring buffer can keep track of where events are lost. Use this information to the output of trace_pipe: hackbench-3588 [001] 1326.701660: lock_acquire: ffffffff816591e0 read rcu_read_lock hackbench-3588 [001] 1326.701661: lock_acquire: ffff88003f4091f0 &(&dentry->d_lock)->rlock hackbench-3588 [001] 1326.701664: lock_release: ffff88003f4091f0 &(&dentry->d_lock)->rlock CPU:1 [LOST 673 EVENTS] hackbench-3588 [001] 1326.702711: kmem_cache_free: call_site=ffffffff81102b85 ptr=ffff880026d96738 hackbench-3588 [001] 1326.702712: lock_release: ffff88003e1480a8 &mm->mmap_sem hackbench-3588 [001] 1326.702713: lock_acquire: ffff88003e1480a8 &mm->mmap_sem Even works with the function graph tracer: 2) ! 170.098 us | } 2) 4.036 us | rcu_irq_exit(); 2) 3.657 us | idle_cpu(); 2) ! 190.301 us | } CPU:2 [LOST 2196 EVENTS] 2) 0.853 us | } /* cancel_dirty_page */ 2) | remove_from_page_cache() { 2) 1.578 us | _raw_spin_lock_irq(); 2) | __remove_from_page_cache() { Note, it does not work with the iterator "trace" file, since it requires the use of consuming the page from the ring buffer to determine how many events were lost, which the iterator does not do. Signed-off-by: Steven Rostedt --- kernel/trace/trace.c | 30 ++++++++++++++++++++++-------- 1 file changed, 22 insertions(+), 8 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index fabb0033a9be..0498bebcbfd1 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -1545,7 +1545,8 @@ static void trace_iterator_increment(struct trace_iterator *iter) } static struct trace_entry * -peek_next_entry(struct trace_iterator *iter, int cpu, u64 *ts) +peek_next_entry(struct trace_iterator *iter, int cpu, u64 *ts, + unsigned long *lost_events) { struct ring_buffer_event *event; struct ring_buffer_iter *buf_iter = iter->buffer_iter[cpu]; @@ -1556,7 +1557,8 @@ peek_next_entry(struct trace_iterator *iter, int cpu, u64 *ts) if (buf_iter) event = ring_buffer_iter_peek(buf_iter, ts); else - event = ring_buffer_peek(iter->tr->buffer, cpu, ts, NULL); + event = ring_buffer_peek(iter->tr->buffer, cpu, ts, + lost_events); ftrace_enable_cpu(); @@ -1564,10 +1566,12 @@ peek_next_entry(struct trace_iterator *iter, int cpu, u64 *ts) } static struct trace_entry * -__find_next_entry(struct trace_iterator *iter, int *ent_cpu, u64 *ent_ts) +__find_next_entry(struct trace_iterator *iter, int *ent_cpu, + unsigned long *missing_events, u64 *ent_ts) { struct ring_buffer *buffer = iter->tr->buffer; struct trace_entry *ent, *next = NULL; + unsigned long lost_events, next_lost = 0; int cpu_file = iter->cpu_file; u64 next_ts = 0, ts; int next_cpu = -1; @@ -1580,7 +1584,7 @@ __find_next_entry(struct trace_iterator *iter, int *ent_cpu, u64 *ent_ts) if (cpu_file > TRACE_PIPE_ALL_CPU) { if (ring_buffer_empty_cpu(buffer, cpu_file)) return NULL; - ent = peek_next_entry(iter, cpu_file, ent_ts); + ent = peek_next_entry(iter, cpu_file, ent_ts, missing_events); if (ent_cpu) *ent_cpu = cpu_file; @@ -1592,7 +1596,7 @@ __find_next_entry(struct trace_iterator *iter, int *ent_cpu, u64 *ent_ts) if (ring_buffer_empty_cpu(buffer, cpu)) continue; - ent = peek_next_entry(iter, cpu, &ts); + ent = peek_next_entry(iter, cpu, &ts, &lost_events); /* * Pick the entry with the smallest timestamp: @@ -1601,6 +1605,7 @@ __find_next_entry(struct trace_iterator *iter, int *ent_cpu, u64 *ent_ts) next = ent; next_cpu = cpu; next_ts = ts; + next_lost = lost_events; } } @@ -1610,6 +1615,9 @@ __find_next_entry(struct trace_iterator *iter, int *ent_cpu, u64 *ent_ts) if (ent_ts) *ent_ts = next_ts; + if (missing_events) + *missing_events = next_lost; + return next; } @@ -1617,13 +1625,14 @@ __find_next_entry(struct trace_iterator *iter, int *ent_cpu, u64 *ent_ts) struct trace_entry *trace_find_next_entry(struct trace_iterator *iter, int *ent_cpu, u64 *ent_ts) { - return __find_next_entry(iter, ent_cpu, ent_ts); + return __find_next_entry(iter, ent_cpu, NULL, ent_ts); } /* Find the next real entry, and increment the iterator to the next entry */ static void *find_next_entry_inc(struct trace_iterator *iter) { - iter->ent = __find_next_entry(iter, &iter->cpu, &iter->ts); + iter->ent = __find_next_entry(iter, &iter->cpu, + &iter->lost_events, &iter->ts); if (iter->ent) trace_iterator_increment(iter); @@ -1635,7 +1644,8 @@ static void trace_consume(struct trace_iterator *iter) { /* Don't allow ftrace to trace into the ring buffers */ ftrace_disable_cpu(); - ring_buffer_consume(iter->tr->buffer, iter->cpu, &iter->ts, NULL); + ring_buffer_consume(iter->tr->buffer, iter->cpu, &iter->ts, + &iter->lost_events); ftrace_enable_cpu(); } @@ -2030,6 +2040,10 @@ static enum print_line_t print_trace_line(struct trace_iterator *iter) { enum print_line_t ret; + if (iter->lost_events) + trace_seq_printf(&iter->seq, "CPU:%d [LOST %lu EVENTS]\n", + iter->cpu, iter->lost_events); + if (iter->trace && iter->trace->print_line) { ret = iter->trace->print_line(iter); if (ret != TRACE_TYPE_UNHANDLED) -- cgit v1.2.3 From ff0ff84a0767df48d728c36510365344a7e7d582 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Wed, 31 Mar 2010 22:11:42 -0400 Subject: ring-buffer: Add lost event count to end of sub buffer Currently, binary readers of the ring buffer only know where events were lost, but not how many events were lost at that location. This information is available, but it would require adding another field to the sub buffer header to include it. But when a event can not fit at the end of a sub buffer, it is written to the next sub buffer. This means there is a good chance that the buffer may have room to hold this counter. If it does, write the counter at the end of the sub buffer and set another flag in the data size field that states that this information exists. Signed-off-by: Steven Rostedt --- kernel/trace/ring_buffer.c | 37 +++++++++++++++++++++++++++++++++---- 1 file changed, 33 insertions(+), 4 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/ring_buffer.c b/kernel/trace/ring_buffer.c index 8295650444c5..dc6d563a6d22 100644 --- a/kernel/trace/ring_buffer.c +++ b/kernel/trace/ring_buffer.c @@ -320,6 +320,8 @@ EXPORT_SYMBOL_GPL(ring_buffer_event_data); /* Flag when events were overwritten */ #define RB_MISSED_EVENTS (1 << 31) +/* Missed count stored at end */ +#define RB_MISSED_STORED (1 << 30) struct buffer_data_page { u64 time_stamp; /* page time stamp */ @@ -340,6 +342,7 @@ struct buffer_page { local_t write; /* index for next write */ unsigned read; /* index for next read */ local_t entries; /* entries on this page */ + unsigned long real_end; /* real end of data */ struct buffer_data_page *page; /* Actual data page */ }; @@ -1769,6 +1772,13 @@ rb_reset_tail(struct ring_buffer_per_cpu *cpu_buffer, event = __rb_page_index(tail_page, tail); kmemcheck_annotate_bitfield(event, bitfield); + /* + * Save the original length to the meta data. + * This will be used by the reader to add lost event + * counter. + */ + tail_page->real_end = tail; + /* * If this event is bigger than the minimum size, then * we need to be careful that we don't subtract the @@ -2888,6 +2898,7 @@ rb_get_reader_page(struct ring_buffer_per_cpu *cpu_buffer) local_set(&cpu_buffer->reader_page->write, 0); local_set(&cpu_buffer->reader_page->entries, 0); local_set(&cpu_buffer->reader_page->page->commit, 0); + cpu_buffer->reader_page->real_end = 0; spin: /* @@ -3728,11 +3739,11 @@ int ring_buffer_read_page(struct ring_buffer *buffer, struct ring_buffer_event *event; struct buffer_data_page *bpage; struct buffer_page *reader; + unsigned long missed_events; unsigned long flags; unsigned int commit; unsigned int read; u64 save_timestamp; - int missed_events = 0; int ret = -1; if (!cpumask_test_cpu(cpu, buffer->cpumask)) @@ -3766,8 +3777,7 @@ int ring_buffer_read_page(struct ring_buffer *buffer, commit = rb_page_commit(reader); /* Check if any events were dropped */ - if (cpu_buffer->lost_events) - missed_events = 1; + missed_events = cpu_buffer->lost_events; /* * If this page has been partially read or @@ -3829,6 +3839,14 @@ int ring_buffer_read_page(struct ring_buffer *buffer, local_set(&reader->entries, 0); reader->read = 0; *data_page = bpage; + + /* + * Use the real_end for the data size, + * This gives us a chance to store the lost events + * on the page. + */ + if (reader->real_end) + local_set(&bpage->commit, reader->real_end); } ret = read; @@ -3836,8 +3854,19 @@ int ring_buffer_read_page(struct ring_buffer *buffer, /* * Set a flag in the commit field if we lost events */ - if (missed_events) + if (missed_events) { + commit = local_read(&bpage->commit); + + /* If there is room at the end of the page to save the + * missed events, then record it there. + */ + if (BUF_PAGE_SIZE - commit >= sizeof(missed_events)) { + memcpy(&bpage->data[commit], &missed_events, + sizeof(missed_events)); + local_add(RB_MISSED_STORED, &bpage->commit); + } local_add(RB_MISSED_EVENTS, &bpage->commit); + } out_unlock: spin_unlock_irqrestore(&cpu_buffer->reader_lock, flags); -- cgit v1.2.3 From eb1e79611cc9bfe21978230e3521e77ea2d7874a Mon Sep 17 00:00:00 2001 From: Frederic Weisbecker Date: Tue, 23 Mar 2010 00:08:59 +0100 Subject: perf: Correctly align perf event tracing buffer The trace event buffer used by perf to record raw sample events is typed as an array of char and may then not be aligned to 8 by alloc_percpu(). But we need it to be aligned to 8 in sparc64 because we cast this buffer into a random structure type built by the TRACE_EVENT() macro to store the traces. So if a random 64 bits field is accessed inside, it may be not under an expected good alignment. Use an array of long instead to force the appropriate alignment, and perform a compile time check to ensure the size in byte of the buffer is a multiple of sizeof(long) so that its actual size doesn't get shrinked under us. This fixes unaligned accesses reported while using perf lock in sparc 64. Suggested-by: David Miller Suggested-by: Tejun Heo Signed-off-by: Frederic Weisbecker Cc: Peter Zijlstra Cc: Arnaldo Carvalho de Melo Cc: Paul Mackerras Cc: Ingo Molnar Cc: David Miller Cc: Steven Rostedt --- kernel/trace/trace_event_perf.c | 11 +++++++++-- 1 file changed, 9 insertions(+), 2 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_event_perf.c b/kernel/trace/trace_event_perf.c index 81f691eb3a30..0565bb42566f 100644 --- a/kernel/trace/trace_event_perf.c +++ b/kernel/trace/trace_event_perf.c @@ -17,7 +17,12 @@ EXPORT_SYMBOL_GPL(perf_arch_fetch_caller_regs); static char *perf_trace_buf; static char *perf_trace_buf_nmi; -typedef typeof(char [PERF_MAX_TRACE_SIZE]) perf_trace_t ; +/* + * Force it to be aligned to unsigned long to avoid misaligned accesses + * suprises + */ +typedef typeof(unsigned long [PERF_MAX_TRACE_SIZE / sizeof(unsigned long)]) + perf_trace_t; /* Count the events in use (per event id, not per instance) */ static int total_ref_count; @@ -130,6 +135,8 @@ __kprobes void *perf_trace_buf_prepare(int size, unsigned short type, char *trace_buf, *raw_data; int pc, cpu; + BUILD_BUG_ON(PERF_MAX_TRACE_SIZE % sizeof(unsigned long)); + pc = preempt_count(); /* Protect the per cpu buffer, begin the rcu read side */ @@ -152,7 +159,7 @@ __kprobes void *perf_trace_buf_prepare(int size, unsigned short type, raw_data = per_cpu_ptr(trace_buf, cpu); /* zero the dead bytes from align to not leak stack to user */ - *(u64 *)(&raw_data[size - sizeof(u64)]) = 0ULL; + memset(&raw_data[size - sizeof(u64)], 0, sizeof(u64)); entry = (struct trace_entry *)raw_data; tracing_generic_entry_update(entry, *irq_flags, pc); -- cgit v1.2.3 From aa27497c2fb4c7f57706099bd489e683e5cc3e3b Mon Sep 17 00:00:00 2001 From: Lai Jiangshan Date: Mon, 5 Apr 2010 17:11:05 +0800 Subject: tracing: Fix uninitialized variable of tracing/trace output Because a local variable is not initialized, I got these when I did 'cat tracing/trace'. (not trace_pipe): CPU:0 [LOST 18446744071579453134 EVENTS] ps-3099 [000] 560.770221: lock_acquire: ffff880030865010 &(&dentry->d_lock)->rlock CPU:0 [LOST 18446744071579453134 EVENTS] ps-3099 [000] 560.770221: lock_release: ffff880030865010 &(&dentry->d_lock)->rlock CPU:0 [LOST 18446612133255294080 EVENTS] ps-3099 [000] 560.770221: lock_acquire: ffff880030865010 &(&dentry->d_lock)->rlock CPU:0 [LOST 18446744071579453134 EVENTS] ps-3099 [000] 560.770222: lock_release: ffff880030865010 &(&dentry->d_lock)->rlock CPU:0 [LOST 18446744071579453134 EVENTS] ps-3099 [000] 560.770222: lock_release: ffffffff816cfb98 dcache_lock See peek_next_entry(), it does not set *lost_events when we 'cat tracing/trace' Signed-off-by: Lai Jiangshan LKML-Reference: <4BB9A929.2000303@cn.fujitsu.com> Signed-off-by: Steven Rostedt --- kernel/trace/trace.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index 0498bebcbfd1..b9be232352b8 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -1571,7 +1571,7 @@ __find_next_entry(struct trace_iterator *iter, int *ent_cpu, { struct ring_buffer *buffer = iter->tr->buffer; struct trace_entry *ent, *next = NULL; - unsigned long lost_events, next_lost = 0; + unsigned long lost_events = 0, next_lost = 0; int cpu_file = iter->cpu_file; u64 next_ts = 0, ts; int next_cpu = -1; -- cgit v1.2.3 From 93ccae7a2227466a0d071fe52c51319f2f34c365 Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Mon, 12 Apr 2010 13:17:08 -0400 Subject: tracing/kprobes: Support basic types on dynamic events Support basic types of integer (u8, u16, u32, u64, s8, s16, s32, s64) in kprobe tracer. With this patch, users can specify above basic types on each arguments after ':'. If omitted, the argument type is set as unsigned long (u32 or u64, arch-dependent). e.g. echo 'p account_system_time+0 hardirq_offset=%si:s32' > kprobe_events adds a probe recording hardirq_offset in signed-32bits value on the entry of account_system_time. Cc: Ingo Molnar Cc: Steven Rostedt Cc: Paul Mackerras Cc: Peter Zijlstra Cc: Mike Galbraith Cc: Frederic Weisbecker LKML-Reference: <20100412171708.3790.18599.stgit@localhost6.localdomain6> Signed-off-by: Masami Hiramatsu Signed-off-by: Arnaldo Carvalho de Melo --- kernel/trace/trace.h | 16 +- kernel/trace/trace_kprobe.c | 535 +++++++++++++++++++++++++++----------------- 2 files changed, 331 insertions(+), 220 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace.h b/kernel/trace/trace.h index bec2c973ff0c..3ebdb6bd2362 100644 --- a/kernel/trace/trace.h +++ b/kernel/trace/trace.h @@ -102,29 +102,17 @@ struct syscall_trace_exit { long ret; }; -struct kprobe_trace_entry { +struct kprobe_trace_entry_head { struct trace_entry ent; unsigned long ip; - int nargs; - unsigned long args[]; }; -#define SIZEOF_KPROBE_TRACE_ENTRY(n) \ - (offsetof(struct kprobe_trace_entry, args) + \ - (sizeof(unsigned long) * (n))) - -struct kretprobe_trace_entry { +struct kretprobe_trace_entry_head { struct trace_entry ent; unsigned long func; unsigned long ret_ip; - int nargs; - unsigned long args[]; }; -#define SIZEOF_KRETPROBE_TRACE_ENTRY(n) \ - (offsetof(struct kretprobe_trace_entry, args) + \ - (sizeof(unsigned long) * (n))) - /* * trace_flag_type is an enumeration that holds different * states when a trace occurs. These are: diff --git a/kernel/trace/trace_kprobe.c b/kernel/trace/trace_kprobe.c index 1251e367bae9..a7514326052b 100644 --- a/kernel/trace/trace_kprobe.c +++ b/kernel/trace/trace_kprobe.c @@ -29,6 +29,8 @@ #include #include #include +#include +#include #include "trace.h" #include "trace_output.h" @@ -40,7 +42,6 @@ /* Reserved field names */ #define FIELD_STRING_IP "__probe_ip" -#define FIELD_STRING_NARGS "__probe_nargs" #define FIELD_STRING_RETIP "__probe_ret_ip" #define FIELD_STRING_FUNC "__probe_func" @@ -52,56 +53,102 @@ const char *reserved_field_names[] = { "common_tgid", "common_lock_depth", FIELD_STRING_IP, - FIELD_STRING_NARGS, FIELD_STRING_RETIP, FIELD_STRING_FUNC, }; -struct fetch_func { - unsigned long (*func)(struct pt_regs *, void *); +/* Printing function type */ +typedef int (*print_type_func_t)(struct trace_seq *, const char *, void *); +#define PRINT_TYPE_FUNC_NAME(type) print_type_##type +#define PRINT_TYPE_FMT_NAME(type) print_type_format_##type + +/* Printing in basic type function template */ +#define DEFINE_BASIC_PRINT_TYPE_FUNC(type, fmt, cast) \ +static __kprobes int PRINT_TYPE_FUNC_NAME(type)(struct trace_seq *s, \ + const char *name, void *data)\ +{ \ + return trace_seq_printf(s, " %s=" fmt, name, (cast)*(type *)data);\ +} \ +static const char PRINT_TYPE_FMT_NAME(type)[] = fmt; + +DEFINE_BASIC_PRINT_TYPE_FUNC(u8, "%x", unsigned int) +DEFINE_BASIC_PRINT_TYPE_FUNC(u16, "%x", unsigned int) +DEFINE_BASIC_PRINT_TYPE_FUNC(u32, "%lx", unsigned long) +DEFINE_BASIC_PRINT_TYPE_FUNC(u64, "%llx", unsigned long long) +DEFINE_BASIC_PRINT_TYPE_FUNC(s8, "%d", int) +DEFINE_BASIC_PRINT_TYPE_FUNC(s16, "%d", int) +DEFINE_BASIC_PRINT_TYPE_FUNC(s32, "%ld", long) +DEFINE_BASIC_PRINT_TYPE_FUNC(s64, "%lld", long long) + +/* Data fetch function type */ +typedef void (*fetch_func_t)(struct pt_regs *, void *, void *); + +struct fetch_param { + fetch_func_t fn; void *data; }; -static __kprobes unsigned long call_fetch(struct fetch_func *f, - struct pt_regs *regs) +static __kprobes void call_fetch(struct fetch_param *fprm, + struct pt_regs *regs, void *dest) { - return f->func(regs, f->data); + return fprm->fn(regs, fprm->data, dest); } -/* fetch handlers */ -static __kprobes unsigned long fetch_register(struct pt_regs *regs, - void *offset) -{ - return regs_get_register(regs, (unsigned int)((unsigned long)offset)); +#define FETCH_FUNC_NAME(kind, type) fetch_##kind##_##type +/* + * Define macro for basic types - we don't need to define s* types, because + * we have to care only about bitwidth at recording time. + */ +#define DEFINE_BASIC_FETCH_FUNCS(kind) \ +DEFINE_FETCH_##kind(u8) \ +DEFINE_FETCH_##kind(u16) \ +DEFINE_FETCH_##kind(u32) \ +DEFINE_FETCH_##kind(u64) + +#define CHECK_BASIC_FETCH_FUNCS(kind, fn) \ + ((FETCH_FUNC_NAME(kind, u8) == fn) || \ + (FETCH_FUNC_NAME(kind, u16) == fn) || \ + (FETCH_FUNC_NAME(kind, u32) == fn) || \ + (FETCH_FUNC_NAME(kind, u64) == fn)) + +/* Data fetch function templates */ +#define DEFINE_FETCH_reg(type) \ +static __kprobes void FETCH_FUNC_NAME(reg, type)(struct pt_regs *regs, \ + void *offset, void *dest) \ +{ \ + *(type *)dest = (type)regs_get_register(regs, \ + (unsigned int)((unsigned long)offset)); \ } - -static __kprobes unsigned long fetch_stack(struct pt_regs *regs, - void *num) -{ - return regs_get_kernel_stack_nth(regs, - (unsigned int)((unsigned long)num)); +DEFINE_BASIC_FETCH_FUNCS(reg) + +#define DEFINE_FETCH_stack(type) \ +static __kprobes void FETCH_FUNC_NAME(stack, type)(struct pt_regs *regs,\ + void *offset, void *dest) \ +{ \ + *(type *)dest = (type)regs_get_kernel_stack_nth(regs, \ + (unsigned int)((unsigned long)offset)); \ } +DEFINE_BASIC_FETCH_FUNCS(stack) -static __kprobes unsigned long fetch_memory(struct pt_regs *regs, void *addr) -{ - unsigned long retval; - - if (probe_kernel_address(addr, retval)) - return 0; - return retval; +#define DEFINE_FETCH_retval(type) \ +static __kprobes void FETCH_FUNC_NAME(retval, type)(struct pt_regs *regs,\ + void *dummy, void *dest) \ +{ \ + *(type *)dest = (type)regs_return_value(regs); \ } - -static __kprobes unsigned long fetch_retvalue(struct pt_regs *regs, - void *dummy) -{ - return regs_return_value(regs); -} - -static __kprobes unsigned long fetch_stack_address(struct pt_regs *regs, - void *dummy) -{ - return kernel_stack_pointer(regs); +DEFINE_BASIC_FETCH_FUNCS(retval) + +#define DEFINE_FETCH_memory(type) \ +static __kprobes void FETCH_FUNC_NAME(memory, type)(struct pt_regs *regs,\ + void *addr, void *dest) \ +{ \ + type retval; \ + if (probe_kernel_address(addr, retval)) \ + *(type *)dest = 0; \ + else \ + *(type *)dest = retval; \ } +DEFINE_BASIC_FETCH_FUNCS(memory) /* Memory fetching by symbol */ struct symbol_cache { @@ -145,51 +192,126 @@ static struct symbol_cache *alloc_symbol_cache(const char *sym, long offset) return sc; } -static __kprobes unsigned long fetch_symbol(struct pt_regs *regs, void *data) -{ - struct symbol_cache *sc = data; - - if (sc->addr) - return fetch_memory(regs, (void *)sc->addr); - else - return 0; +#define DEFINE_FETCH_symbol(type) \ +static __kprobes void FETCH_FUNC_NAME(symbol, type)(struct pt_regs *regs,\ + void *data, void *dest) \ +{ \ + struct symbol_cache *sc = data; \ + if (sc->addr) \ + fetch_memory_##type(regs, (void *)sc->addr, dest); \ + else \ + *(type *)dest = 0; \ } +DEFINE_BASIC_FETCH_FUNCS(symbol) -/* Special indirect memory access interface */ -struct indirect_fetch_data { - struct fetch_func orig; +/* Dereference memory access function */ +struct deref_fetch_param { + struct fetch_param orig; long offset; }; -static __kprobes unsigned long fetch_indirect(struct pt_regs *regs, void *data) -{ - struct indirect_fetch_data *ind = data; - unsigned long addr; - - addr = call_fetch(&ind->orig, regs); - if (addr) { - addr += ind->offset; - return fetch_memory(regs, (void *)addr); - } else - return 0; +#define DEFINE_FETCH_deref(type) \ +static __kprobes void FETCH_FUNC_NAME(deref, type)(struct pt_regs *regs,\ + void *data, void *dest) \ +{ \ + struct deref_fetch_param *dprm = data; \ + unsigned long addr; \ + call_fetch(&dprm->orig, regs, &addr); \ + if (addr) { \ + addr += dprm->offset; \ + fetch_memory_##type(regs, (void *)addr, dest); \ + } else \ + *(type *)dest = 0; \ } +DEFINE_BASIC_FETCH_FUNCS(deref) -static __kprobes void free_indirect_fetch_data(struct indirect_fetch_data *data) +static __kprobes void free_deref_fetch_param(struct deref_fetch_param *data) { - if (data->orig.func == fetch_indirect) - free_indirect_fetch_data(data->orig.data); - else if (data->orig.func == fetch_symbol) + if (CHECK_BASIC_FETCH_FUNCS(deref, data->orig.fn)) + free_deref_fetch_param(data->orig.data); + else if (CHECK_BASIC_FETCH_FUNCS(symbol, data->orig.fn)) free_symbol_cache(data->orig.data); kfree(data); } +/* Default (unsigned long) fetch type */ +#define __DEFAULT_FETCH_TYPE(t) u##t +#define _DEFAULT_FETCH_TYPE(t) __DEFAULT_FETCH_TYPE(t) +#define DEFAULT_FETCH_TYPE _DEFAULT_FETCH_TYPE(BITS_PER_LONG) +#define DEFAULT_FETCH_TYPE_STR __stringify(DEFAULT_FETCH_TYPE) + +#define ASSIGN_FETCH_FUNC(kind, type) \ + .kind = FETCH_FUNC_NAME(kind, type) + +#define ASSIGN_FETCH_TYPE(ptype, ftype, sign) \ + {.name = #ptype, \ + .size = sizeof(ftype), \ + .is_signed = sign, \ + .print = PRINT_TYPE_FUNC_NAME(ptype), \ + .fmt = PRINT_TYPE_FMT_NAME(ptype), \ +ASSIGN_FETCH_FUNC(reg, ftype), \ +ASSIGN_FETCH_FUNC(stack, ftype), \ +ASSIGN_FETCH_FUNC(retval, ftype), \ +ASSIGN_FETCH_FUNC(memory, ftype), \ +ASSIGN_FETCH_FUNC(symbol, ftype), \ +ASSIGN_FETCH_FUNC(deref, ftype), \ + } + +/* Fetch type information table */ +static const struct fetch_type { + const char *name; /* Name of type */ + size_t size; /* Byte size of type */ + int is_signed; /* Signed flag */ + print_type_func_t print; /* Print functions */ + const char *fmt; /* Fromat string */ + /* Fetch functions */ + fetch_func_t reg; + fetch_func_t stack; + fetch_func_t retval; + fetch_func_t memory; + fetch_func_t symbol; + fetch_func_t deref; +} fetch_type_table[] = { + ASSIGN_FETCH_TYPE(u8, u8, 0), + ASSIGN_FETCH_TYPE(u16, u16, 0), + ASSIGN_FETCH_TYPE(u32, u32, 0), + ASSIGN_FETCH_TYPE(u64, u64, 0), + ASSIGN_FETCH_TYPE(s8, u8, 1), + ASSIGN_FETCH_TYPE(s16, u16, 1), + ASSIGN_FETCH_TYPE(s32, u32, 1), + ASSIGN_FETCH_TYPE(s64, u64, 1), +}; + +static const struct fetch_type *find_fetch_type(const char *type) +{ + int i; + + if (!type) + type = DEFAULT_FETCH_TYPE_STR; + + for (i = 0; i < ARRAY_SIZE(fetch_type_table); i++) + if (strcmp(type, fetch_type_table[i].name) == 0) + return &fetch_type_table[i]; + return NULL; +} + +/* Special function : only accept unsigned long */ +static __kprobes void fetch_stack_address(struct pt_regs *regs, + void *dummy, void *dest) +{ + *(unsigned long *)dest = kernel_stack_pointer(regs); +} + /** * Kprobe event core functions */ struct probe_arg { - struct fetch_func fetch; - const char *name; + struct fetch_param fetch; + unsigned int offset; /* Offset from argument entry */ + const char *name; /* Name of this argument */ + const char *comm; /* Command of this argument */ + const struct fetch_type *type; /* Type of this argument */ }; /* Flags for trace_probe */ @@ -204,6 +326,7 @@ struct trace_probe { const char *symbol; /* symbol name */ struct ftrace_event_call call; struct trace_event event; + ssize_t size; /* trace entry size */ unsigned int nr_args; struct probe_arg args[]; }; @@ -212,6 +335,7 @@ struct trace_probe { (offsetof(struct trace_probe, args) + \ (sizeof(struct probe_arg) * (n))) + static __kprobes int probe_is_return(struct trace_probe *tp) { return tp->rp.handler != NULL; @@ -222,49 +346,6 @@ static __kprobes const char *probe_symbol(struct trace_probe *tp) return tp->symbol ? tp->symbol : "unknown"; } -static int probe_arg_string(char *buf, size_t n, struct fetch_func *ff) -{ - int ret = -EINVAL; - - if (ff->func == fetch_register) { - const char *name; - name = regs_query_register_name((unsigned int)((long)ff->data)); - ret = snprintf(buf, n, "%%%s", name); - } else if (ff->func == fetch_stack) - ret = snprintf(buf, n, "$stack%lu", (unsigned long)ff->data); - else if (ff->func == fetch_memory) - ret = snprintf(buf, n, "@0x%p", ff->data); - else if (ff->func == fetch_symbol) { - struct symbol_cache *sc = ff->data; - if (sc->offset) - ret = snprintf(buf, n, "@%s%+ld", sc->symbol, - sc->offset); - else - ret = snprintf(buf, n, "@%s", sc->symbol); - } else if (ff->func == fetch_retvalue) - ret = snprintf(buf, n, "$retval"); - else if (ff->func == fetch_stack_address) - ret = snprintf(buf, n, "$stack"); - else if (ff->func == fetch_indirect) { - struct indirect_fetch_data *id = ff->data; - size_t l = 0; - ret = snprintf(buf, n, "%+ld(", id->offset); - if (ret >= n) - goto end; - l += ret; - ret = probe_arg_string(buf + l, n - l, &id->orig); - if (ret < 0) - goto end; - l += ret; - ret = snprintf(buf + l, n - l, ")"); - ret += l; - } -end: - if (ret >= n) - return -ENOSPC; - return ret; -} - static int register_probe_event(struct trace_probe *tp); static void unregister_probe_event(struct trace_probe *tp); @@ -347,11 +428,12 @@ error: static void free_probe_arg(struct probe_arg *arg) { - if (arg->fetch.func == fetch_symbol) + if (CHECK_BASIC_FETCH_FUNCS(deref, arg->fetch.fn)) + free_deref_fetch_param(arg->fetch.data); + else if (CHECK_BASIC_FETCH_FUNCS(symbol, arg->fetch.fn)) free_symbol_cache(arg->fetch.data); - else if (arg->fetch.func == fetch_indirect) - free_indirect_fetch_data(arg->fetch.data); kfree(arg->name); + kfree(arg->comm); } static void free_trace_probe(struct trace_probe *tp) @@ -457,28 +539,30 @@ static int split_symbol_offset(char *symbol, unsigned long *offset) #define PARAM_MAX_ARGS 16 #define PARAM_MAX_STACK (THREAD_SIZE / sizeof(unsigned long)) -static int parse_probe_vars(char *arg, struct fetch_func *ff, int is_return) +static int parse_probe_vars(char *arg, const struct fetch_type *t, + struct fetch_param *f, int is_return) { int ret = 0; unsigned long param; if (strcmp(arg, "retval") == 0) { - if (is_return) { - ff->func = fetch_retvalue; - ff->data = NULL; - } else + if (is_return) + f->fn = t->retval; + else ret = -EINVAL; } else if (strncmp(arg, "stack", 5) == 0) { if (arg[5] == '\0') { - ff->func = fetch_stack_address; - ff->data = NULL; + if (strcmp(t->name, DEFAULT_FETCH_TYPE_STR) == 0) + f->fn = fetch_stack_address; + else + ret = -EINVAL; } else if (isdigit(arg[5])) { ret = strict_strtoul(arg + 5, 10, ¶m); if (ret || param > PARAM_MAX_STACK) ret = -EINVAL; else { - ff->func = fetch_stack; - ff->data = (void *)param; + f->fn = t->stack; + f->data = (void *)param; } } else ret = -EINVAL; @@ -488,7 +572,8 @@ static int parse_probe_vars(char *arg, struct fetch_func *ff, int is_return) } /* Recursive argument parser */ -static int __parse_probe_arg(char *arg, struct fetch_func *ff, int is_return) +static int __parse_probe_arg(char *arg, const struct fetch_type *t, + struct fetch_param *f, int is_return) { int ret = 0; unsigned long param; @@ -497,13 +582,13 @@ static int __parse_probe_arg(char *arg, struct fetch_func *ff, int is_return) switch (arg[0]) { case '$': - ret = parse_probe_vars(arg + 1, ff, is_return); + ret = parse_probe_vars(arg + 1, t, f, is_return); break; case '%': /* named register */ ret = regs_query_register_offset(arg + 1); if (ret >= 0) { - ff->func = fetch_register; - ff->data = (void *)(unsigned long)ret; + f->fn = t->reg; + f->data = (void *)(unsigned long)ret; ret = 0; } break; @@ -512,26 +597,22 @@ static int __parse_probe_arg(char *arg, struct fetch_func *ff, int is_return) ret = strict_strtoul(arg + 1, 0, ¶m); if (ret) break; - ff->func = fetch_memory; - ff->data = (void *)param; + f->fn = t->memory; + f->data = (void *)param; } else { ret = split_symbol_offset(arg + 1, &offset); if (ret) break; - ff->data = alloc_symbol_cache(arg + 1, offset); - if (ff->data) - ff->func = fetch_symbol; - else - ret = -EINVAL; + f->data = alloc_symbol_cache(arg + 1, offset); + if (f->data) + f->fn = t->symbol; } break; - case '+': /* indirect memory */ + case '+': /* deref memory */ case '-': tmp = strchr(arg, '('); - if (!tmp) { - ret = -EINVAL; + if (!tmp) break; - } *tmp = '\0'; ret = strict_strtol(arg + 1, 0, &offset); if (ret) @@ -541,38 +622,58 @@ static int __parse_probe_arg(char *arg, struct fetch_func *ff, int is_return) arg = tmp + 1; tmp = strrchr(arg, ')'); if (tmp) { - struct indirect_fetch_data *id; + struct deref_fetch_param *dprm; + const struct fetch_type *t2 = find_fetch_type(NULL); *tmp = '\0'; - id = kzalloc(sizeof(struct indirect_fetch_data), - GFP_KERNEL); - if (!id) + dprm = kzalloc(sizeof(struct deref_fetch_param), + GFP_KERNEL); + if (!dprm) return -ENOMEM; - id->offset = offset; - ret = __parse_probe_arg(arg, &id->orig, is_return); + dprm->offset = offset; + ret = __parse_probe_arg(arg, t2, &dprm->orig, + is_return); if (ret) - kfree(id); + kfree(dprm); else { - ff->func = fetch_indirect; - ff->data = (void *)id; + f->fn = t->deref; + f->data = (void *)dprm; } - } else - ret = -EINVAL; + } break; - default: - /* TODO: support custom handler */ - ret = -EINVAL; } + if (!ret && !f->fn) + ret = -EINVAL; return ret; } /* String length checking wrapper */ -static int parse_probe_arg(char *arg, struct fetch_func *ff, int is_return) +static int parse_probe_arg(char *arg, struct trace_probe *tp, + struct probe_arg *parg, int is_return) { + const char *t; + if (strlen(arg) > MAX_ARGSTR_LEN) { pr_info("Argument is too long.: %s\n", arg); return -ENOSPC; } - return __parse_probe_arg(arg, ff, is_return); + parg->comm = kstrdup(arg, GFP_KERNEL); + if (!parg->comm) { + pr_info("Failed to allocate memory for command '%s'.\n", arg); + return -ENOMEM; + } + t = strchr(parg->comm, ':'); + if (t) { + arg[t - parg->comm] = '\0'; + t++; + } + parg->type = find_fetch_type(t); + if (!parg->type) { + pr_info("Unsupported type: %s\n", t); + return -EINVAL; + } + parg->offset = tp->size; + tp->size += parg->type->size; + return __parse_probe_arg(arg, parg->type, &parg->fetch, is_return); } /* Return 1 if name is reserved or already used by another argument */ @@ -602,15 +703,18 @@ static int create_trace_probe(int argc, char **argv) * @ADDR : fetch memory at ADDR (ADDR should be in kernel) * @SYM[+|-offs] : fetch memory at SYM +|- offs (SYM is a data symbol) * %REG : fetch register REG - * Indirect memory fetch: + * Dereferencing memory fetch: * +|-offs(ARG) : fetch memory at ARG +|- offs address. * Alias name of args: * NAME=FETCHARG : set NAME as alias of FETCHARG. + * Type of args: + * FETCHARG:TYPE : use TYPE instead of unsigned long. */ struct trace_probe *tp; int i, ret = 0; int is_return = 0, is_delete = 0; - char *symbol = NULL, *event = NULL, *arg = NULL, *group = NULL; + char *symbol = NULL, *event = NULL, *group = NULL; + char *arg, *tmp; unsigned long offset = 0; void *addr = NULL; char buf[MAX_EVENT_NAME_LEN]; @@ -723,13 +827,6 @@ static int create_trace_probe(int argc, char **argv) else arg = argv[i]; - if (conflict_field_name(argv[i], tp->args, i)) { - pr_info("Argument%d name '%s' conflicts with " - "another field.\n", i, argv[i]); - ret = -EINVAL; - goto error; - } - tp->args[i].name = kstrdup(argv[i], GFP_KERNEL); if (!tp->args[i].name) { pr_info("Failed to allocate argument%d name '%s'.\n", @@ -737,9 +834,19 @@ static int create_trace_probe(int argc, char **argv) ret = -ENOMEM; goto error; } + tmp = strchr(tp->args[i].name, ':'); + if (tmp) + *tmp = '_'; /* convert : to _ */ + + if (conflict_field_name(tp->args[i].name, tp->args, i)) { + pr_info("Argument%d name '%s' conflicts with " + "another field.\n", i, argv[i]); + ret = -EINVAL; + goto error; + } /* Parse fetch argument */ - ret = parse_probe_arg(arg, &tp->args[i].fetch, is_return); + ret = parse_probe_arg(arg, tp, &tp->args[i], is_return); if (ret) { pr_info("Parse error at argument%d. (%d)\n", i, ret); kfree(tp->args[i].name); @@ -794,8 +901,7 @@ static void probes_seq_stop(struct seq_file *m, void *v) static int probes_seq_show(struct seq_file *m, void *v) { struct trace_probe *tp = v; - int i, ret; - char buf[MAX_ARGSTR_LEN + 1]; + int i; seq_printf(m, "%c", probe_is_return(tp) ? 'r' : 'p'); seq_printf(m, ":%s/%s", tp->call.system, tp->call.name); @@ -807,15 +913,10 @@ static int probes_seq_show(struct seq_file *m, void *v) else seq_printf(m, " %s", probe_symbol(tp)); - for (i = 0; i < tp->nr_args; i++) { - ret = probe_arg_string(buf, MAX_ARGSTR_LEN, &tp->args[i].fetch); - if (ret < 0) { - pr_warning("Argument%d decoding error(%d).\n", i, ret); - return ret; - } - seq_printf(m, " %s=%s", tp->args[i].name, buf); - } + for (i = 0; i < tp->nr_args; i++) + seq_printf(m, " %s=%s", tp->args[i].name, tp->args[i].comm); seq_printf(m, "\n"); + return 0; } @@ -945,9 +1046,10 @@ static const struct file_operations kprobe_profile_ops = { static __kprobes void kprobe_trace_func(struct kprobe *kp, struct pt_regs *regs) { struct trace_probe *tp = container_of(kp, struct trace_probe, rp.kp); - struct kprobe_trace_entry *entry; + struct kprobe_trace_entry_head *entry; struct ring_buffer_event *event; struct ring_buffer *buffer; + u8 *data; int size, i, pc; unsigned long irq_flags; struct ftrace_event_call *call = &tp->call; @@ -957,7 +1059,7 @@ static __kprobes void kprobe_trace_func(struct kprobe *kp, struct pt_regs *regs) local_save_flags(irq_flags); pc = preempt_count(); - size = SIZEOF_KPROBE_TRACE_ENTRY(tp->nr_args); + size = sizeof(*entry) + tp->size; event = trace_current_buffer_lock_reserve(&buffer, call->id, size, irq_flags, pc); @@ -965,10 +1067,10 @@ static __kprobes void kprobe_trace_func(struct kprobe *kp, struct pt_regs *regs) return; entry = ring_buffer_event_data(event); - entry->nargs = tp->nr_args; entry->ip = (unsigned long)kp->addr; + data = (u8 *)&entry[1]; for (i = 0; i < tp->nr_args; i++) - entry->args[i] = call_fetch(&tp->args[i].fetch, regs); + call_fetch(&tp->args[i].fetch, regs, data + tp->args[i].offset); if (!filter_current_check_discard(buffer, call, entry, event)) trace_nowake_buffer_unlock_commit(buffer, event, irq_flags, pc); @@ -979,9 +1081,10 @@ static __kprobes void kretprobe_trace_func(struct kretprobe_instance *ri, struct pt_regs *regs) { struct trace_probe *tp = container_of(ri->rp, struct trace_probe, rp); - struct kretprobe_trace_entry *entry; + struct kretprobe_trace_entry_head *entry; struct ring_buffer_event *event; struct ring_buffer *buffer; + u8 *data; int size, i, pc; unsigned long irq_flags; struct ftrace_event_call *call = &tp->call; @@ -989,7 +1092,7 @@ static __kprobes void kretprobe_trace_func(struct kretprobe_instance *ri, local_save_flags(irq_flags); pc = preempt_count(); - size = SIZEOF_KRETPROBE_TRACE_ENTRY(tp->nr_args); + size = sizeof(*entry) + tp->size; event = trace_current_buffer_lock_reserve(&buffer, call->id, size, irq_flags, pc); @@ -997,11 +1100,11 @@ static __kprobes void kretprobe_trace_func(struct kretprobe_instance *ri, return; entry = ring_buffer_event_data(event); - entry->nargs = tp->nr_args; entry->func = (unsigned long)tp->rp.kp.addr; entry->ret_ip = (unsigned long)ri->ret_addr; + data = (u8 *)&entry[1]; for (i = 0; i < tp->nr_args; i++) - entry->args[i] = call_fetch(&tp->args[i].fetch, regs); + call_fetch(&tp->args[i].fetch, regs, data + tp->args[i].offset); if (!filter_current_check_discard(buffer, call, entry, event)) trace_nowake_buffer_unlock_commit(buffer, event, irq_flags, pc); @@ -1011,13 +1114,14 @@ static __kprobes void kretprobe_trace_func(struct kretprobe_instance *ri, enum print_line_t print_kprobe_event(struct trace_iterator *iter, int flags) { - struct kprobe_trace_entry *field; + struct kprobe_trace_entry_head *field; struct trace_seq *s = &iter->seq; struct trace_event *event; struct trace_probe *tp; + u8 *data; int i; - field = (struct kprobe_trace_entry *)iter->ent; + field = (struct kprobe_trace_entry_head *)iter->ent; event = ftrace_find_event(field->ent.type); tp = container_of(event, struct trace_probe, event); @@ -1030,9 +1134,10 @@ print_kprobe_event(struct trace_iterator *iter, int flags) if (!trace_seq_puts(s, ")")) goto partial; - for (i = 0; i < field->nargs; i++) - if (!trace_seq_printf(s, " %s=%lx", - tp->args[i].name, field->args[i])) + data = (u8 *)&field[1]; + for (i = 0; i < tp->nr_args; i++) + if (!tp->args[i].type->print(s, tp->args[i].name, + data + tp->args[i].offset)) goto partial; if (!trace_seq_puts(s, "\n")) @@ -1046,13 +1151,14 @@ partial: enum print_line_t print_kretprobe_event(struct trace_iterator *iter, int flags) { - struct kretprobe_trace_entry *field; + struct kretprobe_trace_entry_head *field; struct trace_seq *s = &iter->seq; struct trace_event *event; struct trace_probe *tp; + u8 *data; int i; - field = (struct kretprobe_trace_entry *)iter->ent; + field = (struct kretprobe_trace_entry_head *)iter->ent; event = ftrace_find_event(field->ent.type); tp = container_of(event, struct trace_probe, event); @@ -1071,9 +1177,10 @@ print_kretprobe_event(struct trace_iterator *iter, int flags) if (!trace_seq_puts(s, ")")) goto partial; - for (i = 0; i < field->nargs; i++) - if (!trace_seq_printf(s, " %s=%lx", - tp->args[i].name, field->args[i])) + data = (u8 *)&field[1]; + for (i = 0; i < tp->nr_args; i++) + if (!tp->args[i].type->print(s, tp->args[i].name, + data + tp->args[i].offset)) goto partial; if (!trace_seq_puts(s, "\n")) @@ -1129,29 +1236,43 @@ static int probe_event_raw_init(struct ftrace_event_call *event_call) static int kprobe_event_define_fields(struct ftrace_event_call *event_call) { int ret, i; - struct kprobe_trace_entry field; + struct kprobe_trace_entry_head field; struct trace_probe *tp = (struct trace_probe *)event_call->data; DEFINE_FIELD(unsigned long, ip, FIELD_STRING_IP, 0); - DEFINE_FIELD(int, nargs, FIELD_STRING_NARGS, 1); /* Set argument names as fields */ - for (i = 0; i < tp->nr_args; i++) - DEFINE_FIELD(unsigned long, args[i], tp->args[i].name, 0); + for (i = 0; i < tp->nr_args; i++) { + ret = trace_define_field(event_call, tp->args[i].type->name, + tp->args[i].name, + sizeof(field) + tp->args[i].offset, + tp->args[i].type->size, + tp->args[i].type->is_signed, + FILTER_OTHER); + if (ret) + return ret; + } return 0; } static int kretprobe_event_define_fields(struct ftrace_event_call *event_call) { int ret, i; - struct kretprobe_trace_entry field; + struct kretprobe_trace_entry_head field; struct trace_probe *tp = (struct trace_probe *)event_call->data; DEFINE_FIELD(unsigned long, func, FIELD_STRING_FUNC, 0); DEFINE_FIELD(unsigned long, ret_ip, FIELD_STRING_RETIP, 0); - DEFINE_FIELD(int, nargs, FIELD_STRING_NARGS, 1); /* Set argument names as fields */ - for (i = 0; i < tp->nr_args; i++) - DEFINE_FIELD(unsigned long, args[i], tp->args[i].name, 0); + for (i = 0; i < tp->nr_args; i++) { + ret = trace_define_field(event_call, tp->args[i].type->name, + tp->args[i].name, + sizeof(field) + tp->args[i].offset, + tp->args[i].type->size, + tp->args[i].type->is_signed, + FILTER_OTHER); + if (ret) + return ret; + } return 0; } @@ -1176,8 +1297,8 @@ static int __set_print_fmt(struct trace_probe *tp, char *buf, int len) pos += snprintf(buf + pos, LEN_OR_ZERO, "\"%s", fmt); for (i = 0; i < tp->nr_args; i++) { - pos += snprintf(buf + pos, LEN_OR_ZERO, " %s=%%lx", - tp->args[i].name); + pos += snprintf(buf + pos, LEN_OR_ZERO, " %s=%s", + tp->args[i].name, tp->args[i].type->fmt); } pos += snprintf(buf + pos, LEN_OR_ZERO, "\", %s", arg); @@ -1219,12 +1340,13 @@ static __kprobes void kprobe_perf_func(struct kprobe *kp, { struct trace_probe *tp = container_of(kp, struct trace_probe, rp.kp); struct ftrace_event_call *call = &tp->call; - struct kprobe_trace_entry *entry; + struct kprobe_trace_entry_head *entry; + u8 *data; int size, __size, i; unsigned long irq_flags; int rctx; - __size = SIZEOF_KPROBE_TRACE_ENTRY(tp->nr_args); + __size = sizeof(*entry) + tp->size; size = ALIGN(__size + sizeof(u32), sizeof(u64)); size -= sizeof(u32); if (WARN_ONCE(size > PERF_MAX_TRACE_SIZE, @@ -1235,10 +1357,10 @@ static __kprobes void kprobe_perf_func(struct kprobe *kp, if (!entry) return; - entry->nargs = tp->nr_args; entry->ip = (unsigned long)kp->addr; + data = (u8 *)&entry[1]; for (i = 0; i < tp->nr_args; i++) - entry->args[i] = call_fetch(&tp->args[i].fetch, regs); + call_fetch(&tp->args[i].fetch, regs, data + tp->args[i].offset); perf_trace_buf_submit(entry, size, rctx, entry->ip, 1, irq_flags, regs); } @@ -1249,12 +1371,13 @@ static __kprobes void kretprobe_perf_func(struct kretprobe_instance *ri, { struct trace_probe *tp = container_of(ri->rp, struct trace_probe, rp); struct ftrace_event_call *call = &tp->call; - struct kretprobe_trace_entry *entry; + struct kretprobe_trace_entry_head *entry; + u8 *data; int size, __size, i; unsigned long irq_flags; int rctx; - __size = SIZEOF_KRETPROBE_TRACE_ENTRY(tp->nr_args); + __size = sizeof(*entry) + tp->size; size = ALIGN(__size + sizeof(u32), sizeof(u64)); size -= sizeof(u32); if (WARN_ONCE(size > PERF_MAX_TRACE_SIZE, @@ -1265,11 +1388,11 @@ static __kprobes void kretprobe_perf_func(struct kretprobe_instance *ri, if (!entry) return; - entry->nargs = tp->nr_args; entry->func = (unsigned long)tp->rp.kp.addr; entry->ret_ip = (unsigned long)ri->ret_addr; + data = (u8 *)&entry[1]; for (i = 0; i < tp->nr_args; i++) - entry->args[i] = call_fetch(&tp->args[i].fetch, regs); + call_fetch(&tp->args[i].fetch, regs, data + tp->args[i].offset); perf_trace_buf_submit(entry, size, rctx, entry->ret_ip, 1, irq_flags, regs); -- cgit v1.2.3 From cecbca96da387428e220e307a9c945e37e2f4d9e Mon Sep 17 00:00:00 2001 From: Frederic Weisbecker Date: Sun, 18 Apr 2010 19:08:41 +0200 Subject: tracing: Dump either the oops's cpu source or all cpus buffers The ftrace_dump_on_oops kernel parameter, sysctl and sysrq let one dump every cpu buffers when an oops or panic happens. It's nice when you have few cpus but it may take ages if have many, plus you miss the real origin of the problem in all the cpu traces. Sometimes, all you need is to dump the cpu buffer that triggered the opps, most of the time it is our main interest. This patch modifies ftrace_dump_on_oops to handle this choice. The ftrace_dump_on_oops kernel parameter, when it comes alone, has the same behaviour than before. But ftrace_dump_on_oops=orig_cpu will only dump the buffer of the cpu that oops'ed. Similarly, sysctl kernel.ftrace_dump_on_oops=1 and echo 1 > /proc/sys/kernel/ftrace_dump_on_oops keep their previous behaviour. But setting 2 jumps into cpu origin dump mode. v2: Fix double setup v3: Fix spelling issues reported by Randy Dunlap v4: Also update __ftrace_dump in the selftests Signed-off-by: Frederic Weisbecker Acked-by: David S. Miller Acked-by: Steven Rostedt Cc: Ingo Molnar Cc: Thomas Gleixner Cc: Li Zefan Cc: Lai Jiangshan --- kernel/trace/trace.c | 51 +++++++++++++++++++++++++++++++++---------- kernel/trace/trace_selftest.c | 5 +++-- 2 files changed, 42 insertions(+), 14 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index bed83cab6da2..7b516c7ef9a0 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -117,9 +117,12 @@ static cpumask_var_t __read_mostly tracing_buffer_mask; * * It is default off, but you can enable it with either specifying * "ftrace_dump_on_oops" in the kernel command line, or setting - * /proc/sys/kernel/ftrace_dump_on_oops to true. + * /proc/sys/kernel/ftrace_dump_on_oops + * Set 1 if you want to dump buffers of all CPUs + * Set 2 if you want to dump the buffer of the CPU that triggered oops */ -int ftrace_dump_on_oops; + +enum ftrace_dump_mode ftrace_dump_on_oops; static int tracing_set_tracer(const char *buf); @@ -139,8 +142,17 @@ __setup("ftrace=", set_cmdline_ftrace); static int __init set_ftrace_dump_on_oops(char *str) { - ftrace_dump_on_oops = 1; - return 1; + if (*str++ != '=' || !*str) { + ftrace_dump_on_oops = DUMP_ALL; + return 1; + } + + if (!strcmp("orig_cpu", str)) { + ftrace_dump_on_oops = DUMP_ORIG; + return 1; + } + + return 0; } __setup("ftrace_dump_on_oops", set_ftrace_dump_on_oops); @@ -4338,7 +4350,7 @@ static int trace_panic_handler(struct notifier_block *this, unsigned long event, void *unused) { if (ftrace_dump_on_oops) - ftrace_dump(); + ftrace_dump(ftrace_dump_on_oops); return NOTIFY_OK; } @@ -4355,7 +4367,7 @@ static int trace_die_handler(struct notifier_block *self, switch (val) { case DIE_OOPS: if (ftrace_dump_on_oops) - ftrace_dump(); + ftrace_dump(ftrace_dump_on_oops); break; default: break; @@ -4396,7 +4408,8 @@ trace_printk_seq(struct trace_seq *s) trace_seq_init(s); } -static void __ftrace_dump(bool disable_tracing) +static void +__ftrace_dump(bool disable_tracing, enum ftrace_dump_mode oops_dump_mode) { static arch_spinlock_t ftrace_dump_lock = (arch_spinlock_t)__ARCH_SPIN_LOCK_UNLOCKED; @@ -4429,12 +4442,25 @@ static void __ftrace_dump(bool disable_tracing) /* don't look at user memory in panic mode */ trace_flags &= ~TRACE_ITER_SYM_USEROBJ; - printk(KERN_TRACE "Dumping ftrace buffer:\n"); - /* Simulate the iterator */ iter.tr = &global_trace; iter.trace = current_trace; - iter.cpu_file = TRACE_PIPE_ALL_CPU; + + switch (oops_dump_mode) { + case DUMP_ALL: + iter.cpu_file = TRACE_PIPE_ALL_CPU; + break; + case DUMP_ORIG: + iter.cpu_file = raw_smp_processor_id(); + break; + case DUMP_NONE: + goto out_enable; + default: + printk(KERN_TRACE "Bad dumping mode, switching to all CPUs dump\n"); + iter.cpu_file = TRACE_PIPE_ALL_CPU; + } + + printk(KERN_TRACE "Dumping ftrace buffer:\n"); /* * We need to stop all tracing on all CPUS to read the @@ -4473,6 +4499,7 @@ static void __ftrace_dump(bool disable_tracing) else printk(KERN_TRACE "---------------------------------\n"); + out_enable: /* Re-enable tracing if requested */ if (!disable_tracing) { trace_flags |= old_userobj; @@ -4489,9 +4516,9 @@ static void __ftrace_dump(bool disable_tracing) } /* By default: disable tracing after the dump */ -void ftrace_dump(void) +void ftrace_dump(enum ftrace_dump_mode oops_dump_mode) { - __ftrace_dump(true); + __ftrace_dump(true, oops_dump_mode); } __init static int tracer_alloc_buffers(void) diff --git a/kernel/trace/trace_selftest.c b/kernel/trace/trace_selftest.c index 9398034f814a..6a9d36ddfcf2 100644 --- a/kernel/trace/trace_selftest.c +++ b/kernel/trace/trace_selftest.c @@ -256,7 +256,8 @@ trace_selftest_startup_function(struct tracer *trace, struct trace_array *tr) /* Maximum number of functions to trace before diagnosing a hang */ #define GRAPH_MAX_FUNC_TEST 100000000 -static void __ftrace_dump(bool disable_tracing); +static void +__ftrace_dump(bool disable_tracing, enum ftrace_dump_mode oops_dump_mode); static unsigned int graph_hang_thresh; /* Wrap the real function entry probe to avoid possible hanging */ @@ -267,7 +268,7 @@ static int trace_graph_entry_watchdog(struct ftrace_graph_ent *trace) ftrace_graph_stop(); printk(KERN_WARNING "BUG: Function graph tracer hang!\n"); if (ftrace_dump_on_oops) - __ftrace_dump(false); + __ftrace_dump(false, DUMP_ALL); return 0; } -- cgit v1.2.3 From 9106b69382912ddc403a307b69bf894a6f3004e4 Mon Sep 17 00:00:00 2001 From: Jiri Olsa Date: Fri, 2 Apr 2010 19:01:20 +0200 Subject: tracing: Add ftrace events for graph tracer Add ftrace events for graph tracer, so the graph output could be shared with other tracers. Signed-off-by: Jiri Olsa LKML-Reference: <1270227683-14631-2-git-send-email-jolsa@redhat.com> Signed-off-by: Steven Rostedt --- kernel/trace/trace_functions_graph.c | 28 +++++++++++++++++++++++++++- 1 file changed, 27 insertions(+), 1 deletion(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_functions_graph.c b/kernel/trace/trace_functions_graph.c index 669b9c31861d..db9e06bb766e 100644 --- a/kernel/trace/trace_functions_graph.c +++ b/kernel/trace/trace_functions_graph.c @@ -40,7 +40,7 @@ struct fgraph_data { #define TRACE_GRAPH_PRINT_OVERHEAD 0x4 #define TRACE_GRAPH_PRINT_PROC 0x8 #define TRACE_GRAPH_PRINT_DURATION 0x10 -#define TRACE_GRAPH_PRINT_ABS_TIME 0X20 +#define TRACE_GRAPH_PRINT_ABS_TIME 0x20 static struct tracer_opt trace_opts[] = { /* Display overruns? (for self-debug purpose) */ @@ -1096,6 +1096,12 @@ print_graph_function(struct trace_iterator *iter) return TRACE_TYPE_HANDLED; } +static enum print_line_t +print_graph_function_event(struct trace_iterator *iter, int flags) +{ + return print_graph_function(iter); +} + static void print_lat_header(struct seq_file *s) { static const char spaces[] = " " /* 16 spaces */ @@ -1199,6 +1205,16 @@ static void graph_trace_close(struct trace_iterator *iter) } } +static struct trace_event graph_trace_entry_event = { + .type = TRACE_GRAPH_ENT, + .trace = print_graph_function_event, +}; + +static struct trace_event graph_trace_ret_event = { + .type = TRACE_GRAPH_RET, + .trace = print_graph_function_event, +}; + static struct tracer graph_trace __read_mostly = { .name = "function_graph", .open = graph_trace_open, @@ -1220,6 +1236,16 @@ static __init int init_graph_trace(void) { max_bytes_for_cpu = snprintf(NULL, 0, "%d", nr_cpu_ids - 1); + if (!register_ftrace_event(&graph_trace_entry_event)) { + pr_warning("Warning: could not register graph trace events\n"); + return 1; + } + + if (!register_ftrace_event(&graph_trace_ret_event)) { + pr_warning("Warning: could not register graph trace events\n"); + return 1; + } + return register_tracer(&graph_trace); } -- cgit v1.2.3 From d7a8d9e907cc294ec7a4a7046d1886375fbcc82e Mon Sep 17 00:00:00 2001 From: Jiri Olsa Date: Fri, 2 Apr 2010 19:01:21 +0200 Subject: tracing: Have graph flags passed in to ouput functions Let the function graph tracer have custom flags passed to its output functions. Signed-off-by: Jiri Olsa LKML-Reference: <1270227683-14631-3-git-send-email-jolsa@redhat.com> Signed-off-by: Steven Rostedt --- kernel/trace/trace.h | 6 +- kernel/trace/trace_functions_graph.c | 123 ++++++++++++++++++++--------------- 2 files changed, 73 insertions(+), 56 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace.h b/kernel/trace/trace.h index 2825ef2c0b15..970004c5fa79 100644 --- a/kernel/trace/trace.h +++ b/kernel/trace/trace.h @@ -491,7 +491,9 @@ extern int trace_clock_id; /* 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); +extern enum print_line_t +print_graph_function_flags(struct trace_iterator *iter, u32 flags); +extern void print_graph_headers_flags(struct seq_file *s, u32 flags); extern enum print_line_t trace_print_graph_duration(unsigned long long duration, struct trace_seq *s); @@ -524,7 +526,7 @@ static inline int ftrace_graph_addr(unsigned long addr) #endif /* CONFIG_DYNAMIC_FTRACE */ #else /* CONFIG_FUNCTION_GRAPH_TRACER */ static inline enum print_line_t -print_graph_function(struct trace_iterator *iter) +print_graph_function_flags(struct trace_iterator *iter, u32 flags) { return TRACE_TYPE_UNHANDLED; } diff --git a/kernel/trace/trace_functions_graph.c b/kernel/trace/trace_functions_graph.c index db9e06bb766e..de5f6518aba0 100644 --- a/kernel/trace/trace_functions_graph.c +++ b/kernel/trace/trace_functions_graph.c @@ -527,17 +527,18 @@ get_return_for_leaf(struct trace_iterator *iter, /* Signal a overhead of time execution to the output */ static int -print_graph_overhead(unsigned long long duration, struct trace_seq *s) +print_graph_overhead(unsigned long long duration, struct trace_seq *s, + u32 flags) { /* If duration disappear, we don't need anything */ - if (!(tracer_flags.val & TRACE_GRAPH_PRINT_DURATION)) + if (!(flags & TRACE_GRAPH_PRINT_DURATION)) return 1; /* Non nested entry or return */ if (duration == -1) return trace_seq_printf(s, " "); - if (tracer_flags.val & TRACE_GRAPH_PRINT_OVERHEAD) { + if (flags & TRACE_GRAPH_PRINT_OVERHEAD) { /* Duration exceeded 100 msecs */ if (duration > 100000ULL) return trace_seq_printf(s, "! "); @@ -563,7 +564,7 @@ static int print_graph_abs_time(u64 t, struct trace_seq *s) static enum print_line_t print_graph_irq(struct trace_iterator *iter, unsigned long addr, - enum trace_type type, int cpu, pid_t pid) + enum trace_type type, int cpu, pid_t pid, u32 flags) { int ret; struct trace_seq *s = &iter->seq; @@ -573,21 +574,21 @@ print_graph_irq(struct trace_iterator *iter, unsigned long addr, return TRACE_TYPE_UNHANDLED; /* Absolute time */ - if (tracer_flags.val & TRACE_GRAPH_PRINT_ABS_TIME) { + if (flags & TRACE_GRAPH_PRINT_ABS_TIME) { ret = print_graph_abs_time(iter->ts, s); if (!ret) return TRACE_TYPE_PARTIAL_LINE; } /* Cpu */ - if (tracer_flags.val & TRACE_GRAPH_PRINT_CPU) { + if (flags & TRACE_GRAPH_PRINT_CPU) { ret = print_graph_cpu(s, cpu); if (ret == TRACE_TYPE_PARTIAL_LINE) return TRACE_TYPE_PARTIAL_LINE; } /* Proc */ - if (tracer_flags.val & TRACE_GRAPH_PRINT_PROC) { + if (flags & TRACE_GRAPH_PRINT_PROC) { ret = print_graph_proc(s, pid); if (ret == TRACE_TYPE_PARTIAL_LINE) return TRACE_TYPE_PARTIAL_LINE; @@ -597,7 +598,7 @@ print_graph_irq(struct trace_iterator *iter, unsigned long addr, } /* No overhead */ - ret = print_graph_overhead(-1, s); + ret = print_graph_overhead(-1, s, flags); if (!ret) return TRACE_TYPE_PARTIAL_LINE; @@ -610,7 +611,7 @@ print_graph_irq(struct trace_iterator *iter, unsigned long addr, return TRACE_TYPE_PARTIAL_LINE; /* Don't close the duration column if haven't one */ - if (tracer_flags.val & TRACE_GRAPH_PRINT_DURATION) + if (flags & TRACE_GRAPH_PRINT_DURATION) trace_seq_printf(s, " |"); ret = trace_seq_printf(s, "\n"); @@ -680,7 +681,8 @@ print_graph_duration(unsigned long long duration, struct trace_seq *s) static enum print_line_t print_graph_entry_leaf(struct trace_iterator *iter, struct ftrace_graph_ent_entry *entry, - struct ftrace_graph_ret_entry *ret_entry, struct trace_seq *s) + struct ftrace_graph_ret_entry *ret_entry, + struct trace_seq *s, u32 flags) { struct fgraph_data *data = iter->private; struct ftrace_graph_ret *graph_ret; @@ -712,12 +714,12 @@ print_graph_entry_leaf(struct trace_iterator *iter, } /* Overhead */ - ret = print_graph_overhead(duration, s); + ret = print_graph_overhead(duration, s, flags); if (!ret) return TRACE_TYPE_PARTIAL_LINE; /* Duration */ - if (tracer_flags.val & TRACE_GRAPH_PRINT_DURATION) { + if (flags & TRACE_GRAPH_PRINT_DURATION) { ret = print_graph_duration(duration, s); if (ret == TRACE_TYPE_PARTIAL_LINE) return TRACE_TYPE_PARTIAL_LINE; @@ -740,7 +742,7 @@ print_graph_entry_leaf(struct trace_iterator *iter, static enum print_line_t print_graph_entry_nested(struct trace_iterator *iter, struct ftrace_graph_ent_entry *entry, - struct trace_seq *s, int cpu) + struct trace_seq *s, int cpu, u32 flags) { struct ftrace_graph_ent *call = &entry->graph_ent; struct fgraph_data *data = iter->private; @@ -760,12 +762,12 @@ print_graph_entry_nested(struct trace_iterator *iter, } /* No overhead */ - ret = print_graph_overhead(-1, s); + ret = print_graph_overhead(-1, s, flags); if (!ret) return TRACE_TYPE_PARTIAL_LINE; /* No time */ - if (tracer_flags.val & TRACE_GRAPH_PRINT_DURATION) { + if (flags & TRACE_GRAPH_PRINT_DURATION) { ret = trace_seq_printf(s, " | "); if (!ret) return TRACE_TYPE_PARTIAL_LINE; @@ -791,7 +793,7 @@ print_graph_entry_nested(struct trace_iterator *iter, static enum print_line_t print_graph_prologue(struct trace_iterator *iter, struct trace_seq *s, - int type, unsigned long addr) + int type, unsigned long addr, u32 flags) { struct fgraph_data *data = iter->private; struct trace_entry *ent = iter->ent; @@ -804,27 +806,27 @@ print_graph_prologue(struct trace_iterator *iter, struct trace_seq *s, if (type) { /* Interrupt */ - ret = print_graph_irq(iter, addr, type, cpu, ent->pid); + ret = print_graph_irq(iter, addr, type, cpu, ent->pid, flags); if (ret == TRACE_TYPE_PARTIAL_LINE) return TRACE_TYPE_PARTIAL_LINE; } /* Absolute time */ - if (tracer_flags.val & TRACE_GRAPH_PRINT_ABS_TIME) { + if (flags & TRACE_GRAPH_PRINT_ABS_TIME) { ret = print_graph_abs_time(iter->ts, s); if (!ret) return TRACE_TYPE_PARTIAL_LINE; } /* Cpu */ - if (tracer_flags.val & TRACE_GRAPH_PRINT_CPU) { + if (flags & TRACE_GRAPH_PRINT_CPU) { ret = print_graph_cpu(s, cpu); if (ret == TRACE_TYPE_PARTIAL_LINE) return TRACE_TYPE_PARTIAL_LINE; } /* Proc */ - if (tracer_flags.val & TRACE_GRAPH_PRINT_PROC) { + if (flags & TRACE_GRAPH_PRINT_PROC) { ret = print_graph_proc(s, ent->pid); if (ret == TRACE_TYPE_PARTIAL_LINE) return TRACE_TYPE_PARTIAL_LINE; @@ -846,7 +848,7 @@ print_graph_prologue(struct trace_iterator *iter, struct trace_seq *s, static enum print_line_t print_graph_entry(struct ftrace_graph_ent_entry *field, struct trace_seq *s, - struct trace_iterator *iter) + struct trace_iterator *iter, u32 flags) { struct fgraph_data *data = iter->private; struct ftrace_graph_ent *call = &field->graph_ent; @@ -854,14 +856,14 @@ print_graph_entry(struct ftrace_graph_ent_entry *field, struct trace_seq *s, static enum print_line_t ret; int cpu = iter->cpu; - if (print_graph_prologue(iter, s, TRACE_GRAPH_ENT, call->func)) + if (print_graph_prologue(iter, s, TRACE_GRAPH_ENT, call->func, flags)) return TRACE_TYPE_PARTIAL_LINE; leaf_ret = get_return_for_leaf(iter, field); if (leaf_ret) - ret = print_graph_entry_leaf(iter, field, leaf_ret, s); + ret = print_graph_entry_leaf(iter, field, leaf_ret, s, flags); else - ret = print_graph_entry_nested(iter, field, s, cpu); + ret = print_graph_entry_nested(iter, field, s, cpu, flags); if (data) { /* @@ -880,7 +882,8 @@ print_graph_entry(struct ftrace_graph_ent_entry *field, struct trace_seq *s, static enum print_line_t print_graph_return(struct ftrace_graph_ret *trace, struct trace_seq *s, - struct trace_entry *ent, struct trace_iterator *iter) + struct trace_entry *ent, struct trace_iterator *iter, + u32 flags) { unsigned long long duration = trace->rettime - trace->calltime; struct fgraph_data *data = iter->private; @@ -910,16 +913,16 @@ print_graph_return(struct ftrace_graph_ret *trace, struct trace_seq *s, } } - if (print_graph_prologue(iter, s, 0, 0)) + if (print_graph_prologue(iter, s, 0, 0, flags)) return TRACE_TYPE_PARTIAL_LINE; /* Overhead */ - ret = print_graph_overhead(duration, s); + ret = print_graph_overhead(duration, s, flags); if (!ret) return TRACE_TYPE_PARTIAL_LINE; /* Duration */ - if (tracer_flags.val & TRACE_GRAPH_PRINT_DURATION) { + if (flags & TRACE_GRAPH_PRINT_DURATION) { ret = print_graph_duration(duration, s); if (ret == TRACE_TYPE_PARTIAL_LINE) return TRACE_TYPE_PARTIAL_LINE; @@ -949,14 +952,15 @@ print_graph_return(struct ftrace_graph_ret *trace, struct trace_seq *s, } /* Overrun */ - if (tracer_flags.val & TRACE_GRAPH_PRINT_OVERRUN) { + if (flags & TRACE_GRAPH_PRINT_OVERRUN) { ret = trace_seq_printf(s, " (Overruns: %lu)\n", trace->overrun); if (!ret) return TRACE_TYPE_PARTIAL_LINE; } - ret = print_graph_irq(iter, trace->func, TRACE_GRAPH_RET, cpu, pid); + ret = print_graph_irq(iter, trace->func, TRACE_GRAPH_RET, + cpu, pid, flags); if (ret == TRACE_TYPE_PARTIAL_LINE) return TRACE_TYPE_PARTIAL_LINE; @@ -964,8 +968,8 @@ print_graph_return(struct ftrace_graph_ret *trace, struct trace_seq *s, } static enum print_line_t -print_graph_comment(struct trace_seq *s, struct trace_entry *ent, - struct trace_iterator *iter) +print_graph_comment(struct trace_seq *s, struct trace_entry *ent, + struct trace_iterator *iter, u32 flags) { unsigned long sym_flags = (trace_flags & TRACE_ITER_SYM_MASK); struct fgraph_data *data = iter->private; @@ -977,16 +981,16 @@ print_graph_comment(struct trace_seq *s, struct trace_entry *ent, if (data) depth = per_cpu_ptr(data->cpu_data, iter->cpu)->depth; - if (print_graph_prologue(iter, s, 0, 0)) + if (print_graph_prologue(iter, s, 0, 0, flags)) return TRACE_TYPE_PARTIAL_LINE; /* No overhead */ - ret = print_graph_overhead(-1, s); + ret = print_graph_overhead(-1, s, flags); if (!ret) return TRACE_TYPE_PARTIAL_LINE; /* No time */ - if (tracer_flags.val & TRACE_GRAPH_PRINT_DURATION) { + if (flags & TRACE_GRAPH_PRINT_DURATION) { ret = trace_seq_printf(s, " | "); if (!ret) return TRACE_TYPE_PARTIAL_LINE; @@ -1041,7 +1045,7 @@ print_graph_comment(struct trace_seq *s, struct trace_entry *ent, enum print_line_t -print_graph_function(struct trace_iterator *iter) +print_graph_function_flags(struct trace_iterator *iter, u32 flags) { struct ftrace_graph_ent_entry *field; struct fgraph_data *data = iter->private; @@ -1062,7 +1066,7 @@ print_graph_function(struct trace_iterator *iter) if (data && data->failed) { field = &data->ent; iter->cpu = data->cpu; - ret = print_graph_entry(field, s, iter); + ret = print_graph_entry(field, s, iter, flags); if (ret == TRACE_TYPE_HANDLED && iter->cpu != cpu) { per_cpu_ptr(data->cpu_data, iter->cpu)->ignore = 1; ret = TRACE_TYPE_NO_CONSUME; @@ -1082,38 +1086,44 @@ print_graph_function(struct trace_iterator *iter) struct ftrace_graph_ent_entry saved; trace_assign_type(field, entry); saved = *field; - return print_graph_entry(&saved, s, iter); + return print_graph_entry(&saved, s, iter, flags); } case TRACE_GRAPH_RET: { struct ftrace_graph_ret_entry *field; trace_assign_type(field, entry); - return print_graph_return(&field->ret, s, entry, iter); + return print_graph_return(&field->ret, s, entry, iter, flags); } default: - return print_graph_comment(s, entry, iter); + return print_graph_comment(s, entry, iter, flags); } return TRACE_TYPE_HANDLED; } +static enum print_line_t +print_graph_function(struct trace_iterator *iter) +{ + return print_graph_function_flags(iter, tracer_flags.val); +} + static enum print_line_t print_graph_function_event(struct trace_iterator *iter, int flags) { return print_graph_function(iter); } -static void print_lat_header(struct seq_file *s) +static void print_lat_header(struct seq_file *s, u32 flags) { static const char spaces[] = " " /* 16 spaces */ " " /* 4 spaces */ " "; /* 17 spaces */ int size = 0; - if (tracer_flags.val & TRACE_GRAPH_PRINT_ABS_TIME) + if (flags & TRACE_GRAPH_PRINT_ABS_TIME) size += 16; - if (tracer_flags.val & TRACE_GRAPH_PRINT_CPU) + if (flags & TRACE_GRAPH_PRINT_CPU) size += 4; - if (tracer_flags.val & TRACE_GRAPH_PRINT_PROC) + if (flags & TRACE_GRAPH_PRINT_PROC) size += 17; seq_printf(s, "#%.*s _-----=> irqs-off \n", size, spaces); @@ -1124,42 +1134,47 @@ static void print_lat_header(struct seq_file *s) seq_printf(s, "#%.*s|||| / \n", size, spaces); } -static void print_graph_headers(struct seq_file *s) +void print_graph_headers_flags(struct seq_file *s, u32 flags) { int lat = trace_flags & TRACE_ITER_LATENCY_FMT; if (lat) - print_lat_header(s); + print_lat_header(s, flags); /* 1st line */ seq_printf(s, "#"); - if (tracer_flags.val & TRACE_GRAPH_PRINT_ABS_TIME) + if (flags & TRACE_GRAPH_PRINT_ABS_TIME) seq_printf(s, " TIME "); - if (tracer_flags.val & TRACE_GRAPH_PRINT_CPU) + if (flags & TRACE_GRAPH_PRINT_CPU) seq_printf(s, " CPU"); - if (tracer_flags.val & TRACE_GRAPH_PRINT_PROC) + if (flags & TRACE_GRAPH_PRINT_PROC) seq_printf(s, " TASK/PID "); if (lat) seq_printf(s, "|||||"); - if (tracer_flags.val & TRACE_GRAPH_PRINT_DURATION) + if (flags & TRACE_GRAPH_PRINT_DURATION) seq_printf(s, " DURATION "); seq_printf(s, " FUNCTION CALLS\n"); /* 2nd line */ seq_printf(s, "#"); - if (tracer_flags.val & TRACE_GRAPH_PRINT_ABS_TIME) + if (flags & TRACE_GRAPH_PRINT_ABS_TIME) seq_printf(s, " | "); - if (tracer_flags.val & TRACE_GRAPH_PRINT_CPU) + if (flags & TRACE_GRAPH_PRINT_CPU) seq_printf(s, " | "); - if (tracer_flags.val & TRACE_GRAPH_PRINT_PROC) + if (flags & TRACE_GRAPH_PRINT_PROC) seq_printf(s, " | | "); if (lat) seq_printf(s, "|||||"); - if (tracer_flags.val & TRACE_GRAPH_PRINT_DURATION) + if (flags & TRACE_GRAPH_PRINT_DURATION) seq_printf(s, " | | "); seq_printf(s, " | | | |\n"); } +static void print_graph_headers(struct seq_file *s) +{ + print_graph_headers_flags(s, tracer_flags.val); +} + static void graph_trace_open(struct trace_iterator *iter) { /* pid and depth on the last trace processed */ -- cgit v1.2.3 From 62b915f1060996a8e1f69be50e3b8e9e43b710cb Mon Sep 17 00:00:00 2001 From: Jiri Olsa Date: Fri, 2 Apr 2010 19:01:22 +0200 Subject: tracing: Add graph output support for irqsoff tracer Add function graph output to irqsoff tracer. The graph output is enabled by setting new 'display-graph' trace option. Signed-off-by: Jiri Olsa LKML-Reference: <1270227683-14631-4-git-send-email-jolsa@redhat.com> Signed-off-by: Steven Rostedt --- kernel/trace/trace.c | 35 +++-- kernel/trace/trace.h | 21 +++ kernel/trace/trace_functions_graph.c | 15 +- kernel/trace/trace_irqsoff.c | 271 +++++++++++++++++++++++++++++++++-- 4 files changed, 313 insertions(+), 29 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index 7b516c7ef9a0..8b9ba41ec146 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -1808,7 +1808,7 @@ static void print_func_help_header(struct seq_file *m) } -static void +void print_trace_header(struct seq_file *m, struct trace_iterator *iter) { unsigned long sym_flags = (trace_flags & TRACE_ITER_SYM_MASK); @@ -2017,7 +2017,7 @@ static enum print_line_t print_bin_fmt(struct trace_iterator *iter) return event ? event->binary(iter, 0) : TRACE_TYPE_HANDLED; } -static int trace_empty(struct trace_iterator *iter) +int trace_empty(struct trace_iterator *iter) { int cpu; @@ -2084,6 +2084,23 @@ static enum print_line_t print_trace_line(struct trace_iterator *iter) return print_trace_fmt(iter); } +void trace_default_header(struct seq_file *m) +{ + struct trace_iterator *iter = m->private; + + if (iter->iter_flags & TRACE_FILE_LAT_FMT) { + /* print nothing if the buffers are empty */ + if (trace_empty(iter)) + return; + print_trace_header(m, iter); + if (!(trace_flags & TRACE_ITER_VERBOSE)) + print_lat_help_header(m); + } else { + if (!(trace_flags & TRACE_ITER_VERBOSE)) + print_func_help_header(m); + } +} + static int s_show(struct seq_file *m, void *v) { struct trace_iterator *iter = v; @@ -2096,17 +2113,9 @@ static int s_show(struct seq_file *m, void *v) } if (iter->trace && iter->trace->print_header) iter->trace->print_header(m); - else if (iter->iter_flags & TRACE_FILE_LAT_FMT) { - /* print nothing if the buffers are empty */ - if (trace_empty(iter)) - return 0; - print_trace_header(m, iter); - if (!(trace_flags & TRACE_ITER_VERBOSE)) - print_lat_help_header(m); - } else { - if (!(trace_flags & TRACE_ITER_VERBOSE)) - print_func_help_header(m); - } + else + trace_default_header(m); + } else if (iter->leftover) { /* * If we filled the seq_file buffer earlier, we diff --git a/kernel/trace/trace.h b/kernel/trace/trace.h index 970004c5fa79..911e9864e94a 100644 --- a/kernel/trace/trace.h +++ b/kernel/trace/trace.h @@ -378,6 +378,9 @@ void trace_function(struct trace_array *tr, unsigned long ip, unsigned long parent_ip, unsigned long flags, int pc); +void trace_default_header(struct seq_file *m); +void print_trace_header(struct seq_file *m, struct trace_iterator *iter); +int trace_empty(struct trace_iterator *iter); void trace_graph_return(struct ftrace_graph_ret *trace); int trace_graph_entry(struct ftrace_graph_ent *trace); @@ -491,11 +494,29 @@ extern int trace_clock_id; /* Standard output formatting function used for function return traces */ #ifdef CONFIG_FUNCTION_GRAPH_TRACER + +/* Flag options */ +#define TRACE_GRAPH_PRINT_OVERRUN 0x1 +#define TRACE_GRAPH_PRINT_CPU 0x2 +#define TRACE_GRAPH_PRINT_OVERHEAD 0x4 +#define TRACE_GRAPH_PRINT_PROC 0x8 +#define TRACE_GRAPH_PRINT_DURATION 0x10 +#define TRACE_GRAPH_PRINT_ABS_TIME 0x20 + extern enum print_line_t print_graph_function_flags(struct trace_iterator *iter, u32 flags); extern void print_graph_headers_flags(struct seq_file *s, u32 flags); extern enum print_line_t trace_print_graph_duration(unsigned long long duration, struct trace_seq *s); +extern void graph_trace_open(struct trace_iterator *iter); +extern void graph_trace_close(struct trace_iterator *iter); +extern int __trace_graph_entry(struct trace_array *tr, + struct ftrace_graph_ent *trace, + unsigned long flags, int pc); +extern void __trace_graph_return(struct trace_array *tr, + struct ftrace_graph_ret *trace, + unsigned long flags, int pc); + #ifdef CONFIG_DYNAMIC_FTRACE /* TODO: make this variable */ diff --git a/kernel/trace/trace_functions_graph.c b/kernel/trace/trace_functions_graph.c index de5f6518aba0..dd11c830eb84 100644 --- a/kernel/trace/trace_functions_graph.c +++ b/kernel/trace/trace_functions_graph.c @@ -179,7 +179,7 @@ unsigned long ftrace_return_to_handler(unsigned long frame_pointer) return ret; } -static int __trace_graph_entry(struct trace_array *tr, +int __trace_graph_entry(struct trace_array *tr, struct ftrace_graph_ent *trace, unsigned long flags, int pc) @@ -246,7 +246,7 @@ int trace_graph_thresh_entry(struct ftrace_graph_ent *trace) return trace_graph_entry(trace); } -static void __trace_graph_return(struct trace_array *tr, +void __trace_graph_return(struct trace_array *tr, struct ftrace_graph_ret *trace, unsigned long flags, int pc) @@ -1093,6 +1093,11 @@ print_graph_function_flags(struct trace_iterator *iter, u32 flags) trace_assign_type(field, entry); return print_graph_return(&field->ret, s, entry, iter, flags); } + case TRACE_STACK: + case TRACE_FN: + /* dont trace stack and functions as comments */ + return TRACE_TYPE_UNHANDLED; + default: return print_graph_comment(s, entry, iter, flags); } @@ -1170,12 +1175,12 @@ void print_graph_headers_flags(struct seq_file *s, u32 flags) seq_printf(s, " | | | |\n"); } -static void print_graph_headers(struct seq_file *s) +void print_graph_headers(struct seq_file *s) { print_graph_headers_flags(s, tracer_flags.val); } -static void graph_trace_open(struct trace_iterator *iter) +void graph_trace_open(struct trace_iterator *iter) { /* pid and depth on the last trace processed */ struct fgraph_data *data; @@ -1210,7 +1215,7 @@ static void graph_trace_open(struct trace_iterator *iter) pr_warning("function graph tracer: not enough memory\n"); } -static void graph_trace_close(struct trace_iterator *iter) +void graph_trace_close(struct trace_iterator *iter) { struct fgraph_data *data = iter->private; diff --git a/kernel/trace/trace_irqsoff.c b/kernel/trace/trace_irqsoff.c index 2974bc7538c7..6fd486e0cef4 100644 --- a/kernel/trace/trace_irqsoff.c +++ b/kernel/trace/trace_irqsoff.c @@ -34,6 +34,9 @@ static int trace_type __read_mostly; static int save_lat_flag; +static void stop_irqsoff_tracer(struct trace_array *tr, int graph); +static int start_irqsoff_tracer(struct trace_array *tr, int graph); + #ifdef CONFIG_PREEMPT_TRACER static inline int preempt_trace(void) @@ -55,6 +58,23 @@ irq_trace(void) # define irq_trace() (0) #endif +#define TRACE_DISPLAY_GRAPH 1 + +static struct tracer_opt trace_opts[] = { +#ifdef CONFIG_FUNCTION_GRAPH_TRACER + /* display latency trace as call graph */ + { TRACER_OPT(display-graph, TRACE_DISPLAY_GRAPH) }, +#endif + { } /* Empty entry */ +}; + +static struct tracer_flags tracer_flags = { + .val = 0, + .opts = trace_opts, +}; + +#define is_graph() (tracer_flags.val & TRACE_DISPLAY_GRAPH) + /* * Sequence count - we record it when starting a measurement and * skip the latency if the sequence has changed - some other section @@ -108,6 +128,202 @@ static struct ftrace_ops trace_ops __read_mostly = }; #endif /* CONFIG_FUNCTION_TRACER */ +#ifdef CONFIG_FUNCTION_GRAPH_TRACER +static int irqsoff_set_flag(u32 old_flags, u32 bit, int set) +{ + int cpu; + + if (!(bit & TRACE_DISPLAY_GRAPH)) + return -EINVAL; + + if (!(is_graph() ^ set)) + return 0; + + stop_irqsoff_tracer(irqsoff_trace, !set); + + for_each_possible_cpu(cpu) + per_cpu(tracing_cpu, cpu) = 0; + + tracing_max_latency = 0; + tracing_reset_online_cpus(irqsoff_trace); + + return start_irqsoff_tracer(irqsoff_trace, set); +} + +static int irqsoff_graph_entry(struct ftrace_graph_ent *trace) +{ + struct trace_array *tr = irqsoff_trace; + struct trace_array_cpu *data; + unsigned long flags; + long disabled; + int ret; + int cpu; + int pc; + + cpu = raw_smp_processor_id(); + if (likely(!per_cpu(tracing_cpu, cpu))) + return 0; + + local_save_flags(flags); + /* slight chance to get a false positive on tracing_cpu */ + if (!irqs_disabled_flags(flags)) + return 0; + + data = tr->data[cpu]; + disabled = atomic_inc_return(&data->disabled); + + if (likely(disabled == 1)) { + pc = preempt_count(); + ret = __trace_graph_entry(tr, trace, flags, pc); + } else + ret = 0; + + atomic_dec(&data->disabled); + return ret; +} + +static void irqsoff_graph_return(struct ftrace_graph_ret *trace) +{ + struct trace_array *tr = irqsoff_trace; + struct trace_array_cpu *data; + unsigned long flags; + long disabled; + int cpu; + int pc; + + cpu = raw_smp_processor_id(); + if (likely(!per_cpu(tracing_cpu, cpu))) + return; + + local_save_flags(flags); + /* slight chance to get a false positive on tracing_cpu */ + if (!irqs_disabled_flags(flags)) + return; + + data = tr->data[cpu]; + disabled = atomic_inc_return(&data->disabled); + + if (likely(disabled == 1)) { + pc = preempt_count(); + __trace_graph_return(tr, trace, flags, pc); + } + + atomic_dec(&data->disabled); +} + +static void irqsoff_trace_open(struct trace_iterator *iter) +{ + if (is_graph()) + graph_trace_open(iter); + +} + +static void irqsoff_trace_close(struct trace_iterator *iter) +{ + if (iter->private) + graph_trace_close(iter); +} + +#define GRAPH_TRACER_FLAGS (TRACE_GRAPH_PRINT_CPU | \ + TRACE_GRAPH_PRINT_PROC) + +static enum print_line_t irqsoff_print_line(struct trace_iterator *iter) +{ + u32 flags = GRAPH_TRACER_FLAGS; + + if (trace_flags & TRACE_ITER_LATENCY_FMT) + flags |= TRACE_GRAPH_PRINT_DURATION; + else + flags |= TRACE_GRAPH_PRINT_ABS_TIME; + + /* + * In graph mode call the graph tracer output function, + * otherwise go with the TRACE_FN event handler + */ + if (is_graph()) + return print_graph_function_flags(iter, flags); + + return TRACE_TYPE_UNHANDLED; +} + +static void irqsoff_print_header(struct seq_file *s) +{ + if (is_graph()) { + struct trace_iterator *iter = s->private; + u32 flags = GRAPH_TRACER_FLAGS; + + if (trace_flags & TRACE_ITER_LATENCY_FMT) { + /* print nothing if the buffers are empty */ + if (trace_empty(iter)) + return; + + print_trace_header(s, iter); + flags |= TRACE_GRAPH_PRINT_DURATION; + } else + flags |= TRACE_GRAPH_PRINT_ABS_TIME; + + print_graph_headers_flags(s, flags); + } else + trace_default_header(s); +} + +static void +trace_graph_function(struct trace_array *tr, + unsigned long ip, unsigned long flags, int pc) +{ + u64 time = trace_clock_local(); + struct ftrace_graph_ent ent = { + .func = ip, + .depth = 0, + }; + struct ftrace_graph_ret ret = { + .func = ip, + .depth = 0, + .calltime = time, + .rettime = time, + }; + + __trace_graph_entry(tr, &ent, flags, pc); + __trace_graph_return(tr, &ret, flags, pc); +} + +static void +__trace_function(struct trace_array *tr, + unsigned long ip, unsigned long parent_ip, + unsigned long flags, int pc) +{ + if (!is_graph()) + trace_function(tr, ip, parent_ip, flags, pc); + else { + trace_graph_function(tr, parent_ip, flags, pc); + trace_graph_function(tr, ip, flags, pc); + } +} + +#else +#define __trace_function trace_function + +static int irqsoff_set_flag(u32 old_flags, u32 bit, int set) +{ + return -EINVAL; +} + +static int irqsoff_graph_entry(struct ftrace_graph_ent *trace) +{ + return -1; +} + +static enum print_line_t irqsoff_print_line(struct trace_iterator *iter) +{ + return TRACE_TYPE_UNHANDLED; +} + +static void irqsoff_graph_return(struct ftrace_graph_ret *trace) { } +static void irqsoff_print_header(struct seq_file *s) { } +static void irqsoff_trace_open(struct trace_iterator *iter) { } +static void irqsoff_trace_close(struct trace_iterator *iter) { } +#endif /* CONFIG_FUNCTION_GRAPH_TRACER */ + /* * Should this new latency be reported/recorded? */ @@ -150,7 +366,7 @@ check_critical_timing(struct trace_array *tr, if (!report_latency(delta)) goto out_unlock; - trace_function(tr, CALLER_ADDR0, parent_ip, flags, pc); + __trace_function(tr, CALLER_ADDR0, parent_ip, flags, pc); /* Skip 5 functions to get to the irq/preempt enable function */ __trace_stack(tr, flags, 5, pc); @@ -172,7 +388,7 @@ out_unlock: out: data->critical_sequence = max_sequence; data->preempt_timestamp = ftrace_now(cpu); - trace_function(tr, CALLER_ADDR0, parent_ip, flags, pc); + __trace_function(tr, CALLER_ADDR0, parent_ip, flags, pc); } static inline void @@ -204,7 +420,7 @@ start_critical_timing(unsigned long ip, unsigned long parent_ip) local_save_flags(flags); - trace_function(tr, ip, parent_ip, flags, preempt_count()); + __trace_function(tr, ip, parent_ip, flags, preempt_count()); per_cpu(tracing_cpu, cpu) = 1; @@ -238,7 +454,7 @@ stop_critical_timing(unsigned long ip, unsigned long parent_ip) atomic_inc(&data->disabled); local_save_flags(flags); - trace_function(tr, ip, parent_ip, flags, preempt_count()); + __trace_function(tr, ip, parent_ip, flags, preempt_count()); check_critical_timing(tr, data, parent_ip ? : ip, cpu); data->critical_start = 0; atomic_dec(&data->disabled); @@ -347,19 +563,32 @@ void trace_preempt_off(unsigned long a0, unsigned long a1) } #endif /* CONFIG_PREEMPT_TRACER */ -static void start_irqsoff_tracer(struct trace_array *tr) +static int start_irqsoff_tracer(struct trace_array *tr, int graph) { - register_ftrace_function(&trace_ops); - if (tracing_is_enabled()) + int ret = 0; + + if (!graph) + ret = register_ftrace_function(&trace_ops); + else + ret = register_ftrace_graph(&irqsoff_graph_return, + &irqsoff_graph_entry); + + if (!ret && tracing_is_enabled()) tracer_enabled = 1; else tracer_enabled = 0; + + return ret; } -static void stop_irqsoff_tracer(struct trace_array *tr) +static void stop_irqsoff_tracer(struct trace_array *tr, int graph) { tracer_enabled = 0; - unregister_ftrace_function(&trace_ops); + + if (!graph) + unregister_ftrace_function(&trace_ops); + else + unregister_ftrace_graph(); } static void __irqsoff_tracer_init(struct trace_array *tr) @@ -372,12 +601,14 @@ static void __irqsoff_tracer_init(struct trace_array *tr) /* make sure that the tracer is visible */ smp_wmb(); tracing_reset_online_cpus(tr); - start_irqsoff_tracer(tr); + + if (start_irqsoff_tracer(tr, is_graph())) + printk(KERN_ERR "failed to start irqsoff tracer\n"); } static void irqsoff_tracer_reset(struct trace_array *tr) { - stop_irqsoff_tracer(tr); + stop_irqsoff_tracer(tr, is_graph()); if (!save_lat_flag) trace_flags &= ~TRACE_ITER_LATENCY_FMT; @@ -409,9 +640,15 @@ static struct tracer irqsoff_tracer __read_mostly = .start = irqsoff_tracer_start, .stop = irqsoff_tracer_stop, .print_max = 1, + .print_header = irqsoff_print_header, + .print_line = irqsoff_print_line, + .flags = &tracer_flags, + .set_flag = irqsoff_set_flag, #ifdef CONFIG_FTRACE_SELFTEST .selftest = trace_selftest_startup_irqsoff, #endif + .open = irqsoff_trace_open, + .close = irqsoff_trace_close, }; # define register_irqsoff(trace) register_tracer(&trace) #else @@ -435,9 +672,15 @@ static struct tracer preemptoff_tracer __read_mostly = .start = irqsoff_tracer_start, .stop = irqsoff_tracer_stop, .print_max = 1, + .print_header = irqsoff_print_header, + .print_line = irqsoff_print_line, + .flags = &tracer_flags, + .set_flag = irqsoff_set_flag, #ifdef CONFIG_FTRACE_SELFTEST .selftest = trace_selftest_startup_preemptoff, #endif + .open = irqsoff_trace_open, + .close = irqsoff_trace_close, }; # define register_preemptoff(trace) register_tracer(&trace) #else @@ -463,9 +706,15 @@ static struct tracer preemptirqsoff_tracer __read_mostly = .start = irqsoff_tracer_start, .stop = irqsoff_tracer_stop, .print_max = 1, + .print_header = irqsoff_print_header, + .print_line = irqsoff_print_line, + .flags = &tracer_flags, + .set_flag = irqsoff_set_flag, #ifdef CONFIG_FTRACE_SELFTEST .selftest = trace_selftest_startup_preemptirqsoff, #endif + .open = irqsoff_trace_open, + .close = irqsoff_trace_close, }; # define register_preemptirqsoff(trace) register_tracer(&trace) -- cgit v1.2.3 From 72c9ddfd4c5bf54ef03cfdf57026416cb678eeba Mon Sep 17 00:00:00 2001 From: David Miller Date: Tue, 20 Apr 2010 15:47:11 -0700 Subject: ring-buffer: Make non-consuming read less expensive with lots of cpus. When performing a non-consuming read, a synchronize_sched() is performed once for every cpu which is actively tracing. This is very expensive, and can make it take several seconds to open up the 'trace' file with lots of cpus. Only one synchronize_sched() call is actually necessary. What is desired is for all cpus to see the disabling state change. So we transform the existing sequence: for_each_cpu() { ring_buffer_read_start(); } where each ring_buffer_start() call performs a synchronize_sched(), into the following: for_each_cpu() { ring_buffer_read_prepare(); } ring_buffer_read_prepare_sync(); for_each_cpu() { ring_buffer_read_start(); } wherein only the single ring_buffer_read_prepare_sync() call needs to do the synchronize_sched(). The first phase, via ring_buffer_read_prepare(), allocates the 'iter' memory and increments ->record_disabled. In the second phase, ring_buffer_read_prepare_sync() makes sure this ->record_disabled state is visible fully to all cpus. And in the final third phase, the ring_buffer_read_start() calls reset the 'iter' objects allocated in the first phase since we now know that none of the cpus are adding trace entries any more. This makes openning the 'trace' file nearly instantaneous on a sparc64 Niagara2 box with 128 cpus tracing. Signed-off-by: David S. Miller LKML-Reference: <20100420.154711.11246950.davem@davemloft.net> Signed-off-by: Steven Rostedt --- kernel/trace/ring_buffer.c | 64 ++++++++++++++++++++++++++++++++++++++-------- kernel/trace/trace.c | 11 +++++--- 2 files changed, 62 insertions(+), 13 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/ring_buffer.c b/kernel/trace/ring_buffer.c index 5885cdfc41f3..2a090448ef6b 100644 --- a/kernel/trace/ring_buffer.c +++ b/kernel/trace/ring_buffer.c @@ -3332,23 +3332,30 @@ ring_buffer_consume(struct ring_buffer *buffer, int cpu, u64 *ts, EXPORT_SYMBOL_GPL(ring_buffer_consume); /** - * ring_buffer_read_start - start a non consuming read of the buffer + * ring_buffer_read_prepare - Prepare for a non consuming read of the buffer * @buffer: The ring buffer to read from * @cpu: The cpu buffer to iterate over * - * This starts up an iteration through the buffer. It also disables - * the recording to the buffer until the reading is finished. - * This prevents the reading from being corrupted. This is not - * a consuming read, so a producer is not expected. + * This performs the initial preparations necessary to iterate + * through the buffer. Memory is allocated, buffer recording + * is disabled, and the iterator pointer is returned to the caller. * - * Must be paired with ring_buffer_finish. + * Disabling buffer recordng prevents the reading from being + * corrupted. This is not a consuming read, so a producer is not + * expected. + * + * After a sequence of ring_buffer_read_prepare calls, the user is + * expected to make at least one call to ring_buffer_prepare_sync. + * Afterwards, ring_buffer_read_start is invoked to get things going + * for real. + * + * This overall must be paired with ring_buffer_finish. */ struct ring_buffer_iter * -ring_buffer_read_start(struct ring_buffer *buffer, int cpu) +ring_buffer_read_prepare(struct ring_buffer *buffer, int cpu) { struct ring_buffer_per_cpu *cpu_buffer; struct ring_buffer_iter *iter; - unsigned long flags; if (!cpumask_test_cpu(cpu, buffer->cpumask)) return NULL; @@ -3362,15 +3369,52 @@ ring_buffer_read_start(struct ring_buffer *buffer, int cpu) iter->cpu_buffer = cpu_buffer; atomic_inc(&cpu_buffer->record_disabled); + + return iter; +} +EXPORT_SYMBOL_GPL(ring_buffer_read_prepare); + +/** + * ring_buffer_read_prepare_sync - Synchronize a set of prepare calls + * + * All previously invoked ring_buffer_read_prepare calls to prepare + * iterators will be synchronized. Afterwards, read_buffer_read_start + * calls on those iterators are allowed. + */ +void +ring_buffer_read_prepare_sync(void) +{ synchronize_sched(); +} +EXPORT_SYMBOL_GPL(ring_buffer_read_prepare_sync); + +/** + * ring_buffer_read_start - start a non consuming read of the buffer + * @iter: The iterator returned by ring_buffer_read_prepare + * + * This finalizes the startup of an iteration through the buffer. + * The iterator comes from a call to ring_buffer_read_prepare and + * an intervening ring_buffer_read_prepare_sync must have been + * performed. + * + * Must be paired with ring_buffer_finish. + */ +void +ring_buffer_read_start(struct ring_buffer_iter *iter) +{ + struct ring_buffer_per_cpu *cpu_buffer; + unsigned long flags; + + if (!iter) + return; + + cpu_buffer = iter->cpu_buffer; spin_lock_irqsave(&cpu_buffer->reader_lock, flags); arch_spin_lock(&cpu_buffer->lock); rb_iter_reset(iter); arch_spin_unlock(&cpu_buffer->lock); spin_unlock_irqrestore(&cpu_buffer->reader_lock, flags); - - return iter; } EXPORT_SYMBOL_GPL(ring_buffer_read_start); diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index 8b9ba41ec146..756d7283318b 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -2201,15 +2201,20 @@ __tracing_open(struct inode *inode, struct file *file) if (iter->cpu_file == TRACE_PIPE_ALL_CPU) { for_each_tracing_cpu(cpu) { - iter->buffer_iter[cpu] = - ring_buffer_read_start(iter->tr->buffer, cpu); + ring_buffer_read_prepare(iter->tr->buffer, cpu); + } + ring_buffer_read_prepare_sync(); + for_each_tracing_cpu(cpu) { + ring_buffer_read_start(iter->buffer_iter[cpu]); tracing_iter_reset(iter, cpu); } } else { cpu = iter->cpu_file; iter->buffer_iter[cpu] = - ring_buffer_read_start(iter->tr->buffer, cpu); + ring_buffer_read_prepare(iter->tr->buffer, cpu); + ring_buffer_read_prepare_sync(); + ring_buffer_read_start(iter->buffer_iter[cpu]); tracing_iter_reset(iter, cpu); } -- cgit v1.2.3 From a838b2e634405fb89ddbf4fa9412acb33911911f Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Tue, 27 Apr 2010 13:26:58 -0400 Subject: ring-buffer: Make benchmark handle missed events With the addition of the "missed events" flags that is stored in the commit field of the ring buffer page, the ring_buffer_benchmark was not updated to handle this. If events are missed, then the missed events flag is set in the ring buffer page, the benchmark will count that flag as part of the size of the page and will hit the BUG() when it tries to read beyond the page. The solution is simply to have the ring buffer benchmark mask off the extra bits. Reported-by: Ingo Molnar Signed-off-by: Steven Rostedt --- kernel/trace/ring_buffer_benchmark.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'kernel/trace') diff --git a/kernel/trace/ring_buffer_benchmark.c b/kernel/trace/ring_buffer_benchmark.c index dc56556b55a2..302f8a614635 100644 --- a/kernel/trace/ring_buffer_benchmark.c +++ b/kernel/trace/ring_buffer_benchmark.c @@ -113,7 +113,8 @@ static enum event_status read_page(int cpu) ret = ring_buffer_read_page(buffer, &bpage, PAGE_SIZE, cpu, 1); if (ret >= 0) { rpage = bpage; - commit = local_read(&rpage->commit); + /* The commit may have missed event flags set, clear them */ + commit = local_read(&rpage->commit) & 0xfffff; for (i = 0; i < commit && !kill_test; i += inc) { if (i >= (PAGE_SIZE - offsetof(struct rb_page, data))) { -- cgit v1.2.3 From e330b3bcd83199dd63a819d8d12e40f9edae6c77 Mon Sep 17 00:00:00 2001 From: Chase Douglas Date: Mon, 26 Apr 2010 14:02:05 -0400 Subject: tracing: Show sample std dev in function profiling When combined with function graph tracing the ftrace function profiler also prints the average run time of functions. While this gives us some good information, it doesn't tell us anything about the variance of the run times of the function. This change prints out the s^2 sample standard deviation alongside the average. This change adds one entry to the profile record structure. This increases the memory footprint of the function profiler by 1/3 on a 32-bit system, and by 1/5 on a 64-bit system when function graphing is enabled, though the memory is only allocated when the profiler is turned on. During the profiling, one extra line of code adds the squared calltime to the new record entry, so this should not adversly affect performance. Note that the square of the sample standard deviation is printed because there is no sqrt implementation for unsigned long long in the kernel. Signed-off-by: Chase Douglas LKML-Reference: <1272304925-2436-1-git-send-email-chase.douglas@canonical.com> [ fixed comment about ns^2 -> us^2 conversion ] Signed-off-by: Steven Rostedt --- kernel/trace/ftrace.c | 24 +++++++++++++++++++++--- 1 file changed, 21 insertions(+), 3 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index 2404b59b3097..3bcb340d6f02 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -264,6 +264,7 @@ struct ftrace_profile { unsigned long counter; #ifdef CONFIG_FUNCTION_GRAPH_TRACER unsigned long long time; + unsigned long long time_squared; #endif }; @@ -366,9 +367,9 @@ static int function_stat_headers(struct seq_file *m) { #ifdef CONFIG_FUNCTION_GRAPH_TRACER seq_printf(m, " Function " - "Hit Time Avg\n" + "Hit Time Avg s^2\n" " -------- " - "--- ---- ---\n"); + "--- ---- --- ---\n"); #else seq_printf(m, " Function Hit\n" " -------- ---\n"); @@ -384,6 +385,7 @@ static int function_stat_show(struct seq_file *m, void *v) static DEFINE_MUTEX(mutex); static struct trace_seq s; unsigned long long avg; + unsigned long long stddev; #endif kallsyms_lookup(rec->ip, NULL, NULL, NULL, str); @@ -394,11 +396,25 @@ static int function_stat_show(struct seq_file *m, void *v) avg = rec->time; do_div(avg, rec->counter); + /* Sample standard deviation (s^2) */ + if (rec->counter <= 1) + stddev = 0; + else { + stddev = rec->time_squared - rec->counter * avg * avg; + /* + * Divide only 1000 for ns^2 -> us^2 conversion. + * trace_print_graph_duration will divide 1000 again. + */ + do_div(stddev, (rec->counter - 1) * 1000); + } + mutex_lock(&mutex); trace_seq_init(&s); trace_print_graph_duration(rec->time, &s); trace_seq_puts(&s, " "); trace_print_graph_duration(avg, &s); + trace_seq_puts(&s, " "); + trace_print_graph_duration(stddev, &s); trace_print_seq(m, &s); mutex_unlock(&mutex); #endif @@ -668,8 +684,10 @@ static void profile_graph_return(struct ftrace_graph_ret *trace) } rec = ftrace_find_profiled_func(stat, trace->func); - if (rec) + if (rec) { rec->time += calltime; + rec->time_squared += calltime * calltime; + } out: local_irq_restore(flags); -- cgit v1.2.3 From 37e44bc50d91df1fe7edcf6f02fe168c6d802e64 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Tue, 27 Apr 2010 21:04:24 -0400 Subject: tracing: Fix sleep time function profiling When sleep_time is off the function profiler ignores the time that a task is scheduled out. When the task is scheduled out a timestamp is taken. When the task is scheduled back in, the timestamp is compared to the current time and the saved calltimes are adjusted accordingly. But when stopping the function profiler, the sched switch hook that does this adjustment was stopped before shutting down the tracer. This allowed some tasks to not get their timestamps set when they scheduled out. When the function profiler started again, this would skew the times of the scheduler functions. This patch moves the stopping of the sched switch to after the function profiler is stopped. It also ignores zero set calltimes, which may happen on start up. Signed-off-by: Steven Rostedt --- kernel/trace/ftrace.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) (limited to 'kernel/trace') diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index 3bcb340d6f02..8c9c2934c45f 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -666,6 +666,10 @@ static void profile_graph_return(struct ftrace_graph_ret *trace) if (!stat->hash || !ftrace_profile_enabled) goto out; + /* If the calltime was zero'd ignore it */ + if (!trace->calltime) + goto out; + calltime = trace->rettime - trace->calltime; if (!(trace_flags & TRACE_ITER_GRAPH_TIME)) { @@ -3357,11 +3361,11 @@ void unregister_ftrace_graph(void) goto out; ftrace_graph_active--; - unregister_trace_sched_switch(ftrace_graph_probe_sched_switch); ftrace_graph_return = (trace_func_graph_ret_t)ftrace_stub; ftrace_graph_entry = ftrace_graph_entry_stub; ftrace_shutdown(FTRACE_STOP_FUNC_RET); unregister_pm_notifier(&ftrace_suspend_notifier); + unregister_trace_sched_switch(ftrace_graph_probe_sched_switch); out: mutex_unlock(&ftrace_lock); -- cgit v1.2.3 From 5a2e3995951176e1aaa63d17ae2e1d26ac99003d Mon Sep 17 00:00:00 2001 From: Kei Tokunaga Date: Thu, 1 Apr 2010 20:40:58 +0900 Subject: [SCSI] ftrace: add __print_hex() __print_hex() prints values in an array in hex (w/o '0x') (space separated) EX) 92 33 32 f3 ee 4d Signed-off-by: Li Zefan Signed-off-by: Tomohiro Kusumi Signed-off-by: Kei Tokunaga Acked-by: Steven Rostedt Signed-off-by: James Bottomley --- kernel/trace/trace_output.c | 15 +++++++++++++++ 1 file changed, 15 insertions(+) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_output.c b/kernel/trace/trace_output.c index 8e46b3323cdc..7c4a0ca650b5 100644 --- a/kernel/trace/trace_output.c +++ b/kernel/trace/trace_output.c @@ -355,6 +355,21 @@ ftrace_print_symbols_seq(struct trace_seq *p, unsigned long val, } EXPORT_SYMBOL(ftrace_print_symbols_seq); +const char * +ftrace_print_hex_seq(struct trace_seq *p, const unsigned char *buf, int buf_len) +{ + int i; + const char *ret = p->buffer + p->len; + + for (i = 0; i < buf_len; i++) + trace_seq_printf(p, "%s%2.2x", i == 0 ? "" : " ", buf[i]); + + trace_seq_putc(p, 0); + + return ret; +} +EXPORT_SYMBOL(ftrace_print_hex_seq); + #ifdef CONFIG_KRETPROBES static inline const char *kretprobed(const char *name) { -- cgit v1.2.3 From bf81623542332bc2cedf3db49cbb2edb724780d2 Mon Sep 17 00:00:00 2001 From: Kei Tokunaga Date: Thu, 1 Apr 2010 20:41:40 +0900 Subject: [SCSI] add scsi trace core functions and put trace points Signed-off-by: Xiao Guangrong Signed-off-by: Tomohiro Kusumi Signed-off-by: Kei Tokunaga Signed-off-by: James Bottomley --- kernel/trace/trace_output.c | 1 + 1 file changed, 1 insertion(+) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_output.c b/kernel/trace/trace_output.c index 7c4a0ca650b5..9cb5df5dc656 100644 --- a/kernel/trace/trace_output.c +++ b/kernel/trace/trace_output.c @@ -209,6 +209,7 @@ int trace_seq_putc(struct trace_seq *s, unsigned char c) return 1; } +EXPORT_SYMBOL(trace_seq_putc); int trace_seq_putmem(struct trace_seq *s, const void *mem, size_t len) { -- cgit v1.2.3 From feef47d0cb530e8419dfa0b48141b538b89b1b1a Mon Sep 17 00:00:00 2001 From: Frederic Weisbecker Date: Fri, 23 Apr 2010 05:59:55 +0200 Subject: hw-breakpoints: Get the number of available registers on boot dynamically The breakpoint generic layer assumes that archs always know in advance the static number of address registers available to host breakpoints through the HBP_NUM macro. However this is not true for every archs. For example Arm needs to get this information dynamically to handle the compatiblity between different versions. To solve this, this patch proposes to drop the static HBP_NUM macro and let the arch provide the number of available slots through a new hw_breakpoint_slots() function. For archs that have CONFIG_HAVE_MIXED_BREAKPOINTS_REGS selected, it will be called once as the number of registers fits for instruction and data breakpoints together. For the others it will be called first to get the number of instruction breakpoint registers and another time to get the data breakpoint registers, the targeted type is given as a parameter of hw_breakpoint_slots(). Reported-by: Will Deacon Signed-off-by: Frederic Weisbecker Acked-by: Paul Mundt Cc: Mahesh Salgaonkar Cc: K. Prasad Cc: Benjamin Herrenschmidt Cc: Paul Mackerras Cc: Jason Wessel Cc: Ingo Molnar --- kernel/trace/trace_ksym.c | 26 +++++++------------------- 1 file changed, 7 insertions(+), 19 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_ksym.c b/kernel/trace/trace_ksym.c index d59cd6879477..8eaf00749b65 100644 --- a/kernel/trace/trace_ksym.c +++ b/kernel/trace/trace_ksym.c @@ -34,12 +34,6 @@ #include -/* - * For now, let us restrict the no. of symbols traced simultaneously to number - * of available hardware breakpoint registers. - */ -#define KSYM_TRACER_MAX HBP_NUM - #define KSYM_TRACER_OP_LEN 3 /* rw- */ struct trace_ksym { @@ -53,7 +47,6 @@ struct trace_ksym { static struct trace_array *ksym_trace_array; -static unsigned int ksym_filter_entry_count; static unsigned int ksym_tracing_enabled; static HLIST_HEAD(ksym_filter_head); @@ -181,13 +174,6 @@ int process_new_ksym_entry(char *ksymname, int op, unsigned long addr) struct trace_ksym *entry; int ret = -ENOMEM; - if (ksym_filter_entry_count >= KSYM_TRACER_MAX) { - printk(KERN_ERR "ksym_tracer: Maximum limit:(%d) reached. No" - " new requests for tracing can be accepted now.\n", - KSYM_TRACER_MAX); - return -ENOSPC; - } - entry = kzalloc(sizeof(struct trace_ksym), GFP_KERNEL); if (!entry) return -ENOMEM; @@ -203,13 +189,17 @@ int process_new_ksym_entry(char *ksymname, int op, unsigned long addr) if (IS_ERR(entry->ksym_hbp)) { ret = PTR_ERR(entry->ksym_hbp); - printk(KERN_INFO "ksym_tracer request failed. Try again" - " later!!\n"); + if (ret == -ENOSPC) { + printk(KERN_ERR "ksym_tracer: Maximum limit reached." + " No new requests for tracing can be accepted now.\n"); + } else { + printk(KERN_INFO "ksym_tracer request failed. Try again" + " later!!\n"); + } goto err; } hlist_add_head_rcu(&(entry->ksym_hlist), &ksym_filter_head); - ksym_filter_entry_count++; return 0; @@ -265,7 +255,6 @@ static void __ksym_trace_reset(void) hlist_for_each_entry_safe(entry, node, node1, &ksym_filter_head, ksym_hlist) { unregister_wide_hw_breakpoint(entry->ksym_hbp); - ksym_filter_entry_count--; hlist_del_rcu(&(entry->ksym_hlist)); synchronize_rcu(); kfree(entry); @@ -338,7 +327,6 @@ static ssize_t ksym_trace_filter_write(struct file *file, goto out_unlock; } /* Error or "symbol:---" case: drop it */ - ksym_filter_entry_count--; hlist_del_rcu(&(entry->ksym_hlist)); synchronize_rcu(); kfree(entry); -- cgit v1.2.3 From 956097912c40a03bf22603a3be73503fd9ea9e44 Mon Sep 17 00:00:00 2001 From: Borislav Petkov Date: Sun, 2 May 2010 08:03:54 +0200 Subject: ring-buffer: Wrap open-coded WARN_ONCE Wrap open-coded WARN_ONCE functionality into the equivalent macro. Signed-off-by: Borislav Petkov LKML-Reference: <20100502060354.GA5281@liondog.tnic> Signed-off-by: Steven Rostedt --- kernel/trace/ring_buffer.c | 14 +++++--------- 1 file changed, 5 insertions(+), 9 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/ring_buffer.c b/kernel/trace/ring_buffer.c index 2a090448ef6b..7f6059c5aa94 100644 --- a/kernel/trace/ring_buffer.c +++ b/kernel/trace/ring_buffer.c @@ -2000,17 +2000,13 @@ rb_add_time_stamp(struct ring_buffer_per_cpu *cpu_buffer, u64 *ts, u64 *delta) { struct ring_buffer_event *event; - static int once; int ret; - if (unlikely(*delta > (1ULL << 59) && !once++)) { - printk(KERN_WARNING "Delta way too big! %llu" - " ts=%llu write stamp = %llu\n", - (unsigned long long)*delta, - (unsigned long long)*ts, - (unsigned long long)cpu_buffer->write_stamp); - WARN_ON(1); - } + WARN_ONCE(*delta > (1ULL << 59), + KERN_WARNING "Delta way too big! %llu ts=%llu write stamp = %llu\n", + (unsigned long long)*delta, + (unsigned long long)*ts, + (unsigned long long)cpu_buffer->write_stamp); /* * The delta is too big, we to add a -- cgit v1.2.3 From 668eb65f092902eb7dd526af73d4a7f025a94612 Mon Sep 17 00:00:00 2001 From: Thiago Farina Date: Sun, 24 Jan 2010 11:03:50 -0500 Subject: tracing: Fix "integer as NULL pointer" warning. kernel/trace/trace_output.c:256:24: warning: Using plain integer as NULL pointer Signed-off-by: Thiago Farina LKML-Reference: <1264349038-1766-3-git-send-email-tfransosi@gmail.com> Signed-off-by: Steven Rostedt --- kernel/trace/trace_output.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_output.c b/kernel/trace/trace_output.c index 8e46b3323cdc..2404c129a8c9 100644 --- a/kernel/trace/trace_output.c +++ b/kernel/trace/trace_output.c @@ -253,7 +253,7 @@ void *trace_seq_reserve(struct trace_seq *s, size_t len) void *ret; if (s->full) - return 0; + return NULL; if (len > ((PAGE_SIZE - 1) - s->len)) { s->full = 1; -- cgit v1.2.3 From d9f599e1e6d019968b35d2dc63074b9e8964fa69 Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Sat, 20 Mar 2010 17:39:11 +0300 Subject: perf: Fix check at end of event search The original code doesn't work because "call" is never NULL there. Signed-off-by: Dan Carpenter LKML-Reference: <20100320143911.GF5331@bicker> Signed-off-by: Steven Rostedt --- kernel/trace/trace_events_filter.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_events_filter.c b/kernel/trace/trace_events_filter.c index 88c0b6dbd7fe..58092d844a1f 100644 --- a/kernel/trace/trace_events_filter.c +++ b/kernel/trace/trace_events_filter.c @@ -1398,7 +1398,7 @@ int ftrace_profile_set_filter(struct perf_event *event, int event_id, } err = -EINVAL; - if (!call) + if (&call->list == &ftrace_events) goto out_unlock; err = -EEXIST; -- cgit v1.2.3 From 27a9da6538ee18046d7bff8e36a9f783542c54c3 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Tue, 4 May 2010 20:36:56 +0200 Subject: sched: Remove rq argument to the tracepoints struct rq isn't visible outside of sched.o so its near useless to expose the pointer, also there are no users of it, so remove it. Acked-by: Steven Rostedt Signed-off-by: Peter Zijlstra LKML-Reference: <1272997616.1642.207.camel@laptop> Signed-off-by: Ingo Molnar --- kernel/trace/ftrace.c | 3 +-- kernel/trace/trace_sched_switch.c | 5 ++--- kernel/trace/trace_sched_wakeup.c | 5 ++--- 3 files changed, 5 insertions(+), 8 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index 2404b59b3097..aa3a92b511e2 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -3212,8 +3212,7 @@ free: } static void -ftrace_graph_probe_sched_switch(struct rq *__rq, struct task_struct *prev, - struct task_struct *next) +ftrace_graph_probe_sched_switch(struct task_struct *prev, struct task_struct *next) { unsigned long long timestamp; int index; diff --git a/kernel/trace/trace_sched_switch.c b/kernel/trace/trace_sched_switch.c index 5fca0f51fde4..a55fccfede5d 100644 --- a/kernel/trace/trace_sched_switch.c +++ b/kernel/trace/trace_sched_switch.c @@ -50,8 +50,7 @@ tracing_sched_switch_trace(struct trace_array *tr, } static void -probe_sched_switch(struct rq *__rq, struct task_struct *prev, - struct task_struct *next) +probe_sched_switch(struct task_struct *prev, struct task_struct *next) { struct trace_array_cpu *data; unsigned long flags; @@ -109,7 +108,7 @@ tracing_sched_wakeup_trace(struct trace_array *tr, } static void -probe_sched_wakeup(struct rq *__rq, struct task_struct *wakee, int success) +probe_sched_wakeup(struct task_struct *wakee, int success) { struct trace_array_cpu *data; unsigned long flags; diff --git a/kernel/trace/trace_sched_wakeup.c b/kernel/trace/trace_sched_wakeup.c index 0271742abb8d..8052446ceeaa 100644 --- a/kernel/trace/trace_sched_wakeup.c +++ b/kernel/trace/trace_sched_wakeup.c @@ -107,8 +107,7 @@ static void probe_wakeup_migrate_task(struct task_struct *task, int cpu) } static void notrace -probe_wakeup_sched_switch(struct rq *rq, struct task_struct *prev, - struct task_struct *next) +probe_wakeup_sched_switch(struct task_struct *prev, struct task_struct *next) { struct trace_array_cpu *data; cycle_t T0, T1, delta; @@ -200,7 +199,7 @@ static void wakeup_reset(struct trace_array *tr) } static void -probe_wakeup(struct rq *rq, struct task_struct *p, int success) +probe_wakeup(struct task_struct *p, int success) { struct trace_array_cpu *data; int cpu = smp_processor_id(); -- cgit v1.2.3 From 8f0820183056ad26dabc0202115848a92f1143fc Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Tue, 20 Apr 2010 10:47:33 -0400 Subject: tracing: Create class struct for events This patch creates a ftrace_event_class struct that event structs point to. This class struct will be made to hold information to modify the events. Currently the class struct only holds the events system name. This patch slightly increases the size, but this change lays the ground work of other changes to make the footprint of tracepoints smaller. With 82 standard tracepoints, and 618 system call tracepoints (two tracepoints per syscall: enter and exit): text data bss dec hex filename 4913961 1088356 861512 6863829 68bbd5 vmlinux.orig 4914025 1088868 861512 6864405 68be15 vmlinux.class This patch also cleans up some stale comments in ftrace.h. v2: Fixed missing semi-colon in macro. Acked-by: Frederic Weisbecker Acked-by: Mathieu Desnoyers Acked-by: Masami Hiramatsu Signed-off-by: Steven Rostedt --- kernel/trace/trace_events.c | 20 ++++++++++---------- kernel/trace/trace_events_filter.c | 6 +++--- kernel/trace/trace_export.c | 6 +++++- kernel/trace/trace_kprobe.c | 12 ++++++------ kernel/trace/trace_syscalls.c | 4 ++++ 5 files changed, 28 insertions(+), 20 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_events.c b/kernel/trace/trace_events.c index c697c7043349..2f54b48d3632 100644 --- a/kernel/trace/trace_events.c +++ b/kernel/trace/trace_events.c @@ -176,10 +176,10 @@ static int __ftrace_set_clr_event(const char *match, const char *sub, if (match && strcmp(match, call->name) != 0 && - strcmp(match, call->system) != 0) + strcmp(match, call->class->system) != 0) continue; - if (sub && strcmp(sub, call->system) != 0) + if (sub && strcmp(sub, call->class->system) != 0) continue; if (event && strcmp(event, call->name) != 0) @@ -355,8 +355,8 @@ static int t_show(struct seq_file *m, void *v) { struct ftrace_event_call *call = v; - if (strcmp(call->system, TRACE_SYSTEM) != 0) - seq_printf(m, "%s:", call->system); + if (strcmp(call->class->system, TRACE_SYSTEM) != 0) + seq_printf(m, "%s:", call->class->system); seq_printf(m, "%s\n", call->name); return 0; @@ -453,7 +453,7 @@ system_enable_read(struct file *filp, char __user *ubuf, size_t cnt, if (!call->name || !call->regfunc) continue; - if (system && strcmp(call->system, system) != 0) + if (system && strcmp(call->class->system, system) != 0) continue; /* @@ -925,8 +925,8 @@ event_create_dir(struct ftrace_event_call *call, struct dentry *d_events, * If the trace point header did not define TRACE_SYSTEM * then the system would be called "TRACE_SYSTEM". */ - if (strcmp(call->system, TRACE_SYSTEM) != 0) - d_events = event_subsystem_dir(call->system, d_events); + if (strcmp(call->class->system, TRACE_SYSTEM) != 0) + d_events = event_subsystem_dir(call->class->system, d_events); call->dir = debugfs_create_dir(call->name, d_events); if (!call->dir) { @@ -1041,7 +1041,7 @@ static void __trace_remove_event_call(struct ftrace_event_call *call) list_del(&call->list); trace_destroy_fields(call); destroy_preds(call); - remove_subsystem_dir(call->system); + remove_subsystem_dir(call->class->system); } /* Remove an event_call */ @@ -1399,8 +1399,8 @@ static __init void event_trace_self_tests(void) * syscalls as we test. */ #ifndef CONFIG_EVENT_TRACE_TEST_SYSCALLS - if (call->system && - strcmp(call->system, "syscalls") == 0) + if (call->class->system && + strcmp(call->class->system, "syscalls") == 0) continue; #endif diff --git a/kernel/trace/trace_events_filter.c b/kernel/trace/trace_events_filter.c index 88c0b6dbd7fe..ca329603d0bf 100644 --- a/kernel/trace/trace_events_filter.c +++ b/kernel/trace/trace_events_filter.c @@ -628,7 +628,7 @@ static int init_subsystem_preds(struct event_subsystem *system) if (!call->define_fields) continue; - if (strcmp(call->system, system->name) != 0) + if (strcmp(call->class->system, system->name) != 0) continue; err = init_preds(call); @@ -647,7 +647,7 @@ static void filter_free_subsystem_preds(struct event_subsystem *system) if (!call->define_fields) continue; - if (strcmp(call->system, system->name) != 0) + if (strcmp(call->class->system, system->name) != 0) continue; filter_disable_preds(call); @@ -1252,7 +1252,7 @@ static int replace_system_preds(struct event_subsystem *system, if (!call->define_fields) continue; - if (strcmp(call->system, system->name) != 0) + if (strcmp(call->class->system, system->name) != 0) continue; /* try to see if the filter can be applied */ diff --git a/kernel/trace/trace_export.c b/kernel/trace/trace_export.c index e091f64ba6ce..7f16e2163817 100644 --- a/kernel/trace/trace_export.c +++ b/kernel/trace/trace_export.c @@ -18,6 +18,10 @@ #undef TRACE_SYSTEM #define TRACE_SYSTEM ftrace +struct ftrace_event_class event_class_ftrace = { + .system = __stringify(TRACE_SYSTEM), +}; + /* not needed for this file */ #undef __field_struct #define __field_struct(type, item) @@ -160,7 +164,7 @@ __attribute__((__aligned__(4))) \ __attribute__((section("_ftrace_events"))) event_##call = { \ .name = #call, \ .id = type, \ - .system = __stringify(TRACE_SYSTEM), \ + .class = &event_class_ftrace, \ .raw_init = ftrace_raw_init_event, \ .print_fmt = print, \ .define_fields = ftrace_define_fields_##call, \ diff --git a/kernel/trace/trace_kprobe.c b/kernel/trace/trace_kprobe.c index 1251e367bae9..eda220bf2065 100644 --- a/kernel/trace/trace_kprobe.c +++ b/kernel/trace/trace_kprobe.c @@ -332,8 +332,8 @@ static struct trace_probe *alloc_trace_probe(const char *group, goto error; } - tp->call.system = kstrdup(group, GFP_KERNEL); - if (!tp->call.system) + tp->call.class->system = kstrdup(group, GFP_KERNEL); + if (!tp->call.class->system) goto error; INIT_LIST_HEAD(&tp->list); @@ -361,7 +361,7 @@ static void free_trace_probe(struct trace_probe *tp) for (i = 0; i < tp->nr_args; i++) free_probe_arg(&tp->args[i]); - kfree(tp->call.system); + kfree(tp->call.class->system); kfree(tp->call.name); kfree(tp->symbol); kfree(tp); @@ -374,7 +374,7 @@ static struct trace_probe *find_probe_event(const char *event, list_for_each_entry(tp, &probe_list, list) if (strcmp(tp->call.name, event) == 0 && - strcmp(tp->call.system, group) == 0) + strcmp(tp->call.class->system, group) == 0) return tp; return NULL; } @@ -399,7 +399,7 @@ static int register_trace_probe(struct trace_probe *tp) mutex_lock(&probe_lock); /* register as an event */ - old_tp = find_probe_event(tp->call.name, tp->call.system); + old_tp = find_probe_event(tp->call.name, tp->call.class->system); if (old_tp) { /* delete old event */ unregister_trace_probe(old_tp); @@ -798,7 +798,7 @@ static int probes_seq_show(struct seq_file *m, void *v) char buf[MAX_ARGSTR_LEN + 1]; seq_printf(m, "%c", probe_is_return(tp) ? 'r' : 'p'); - seq_printf(m, ":%s/%s", tp->call.system, tp->call.name); + seq_printf(m, ":%s/%s", tp->call.class->system, tp->call.name); if (!tp->symbol) seq_printf(m, " 0x%p", tp->rp.kp.addr); diff --git a/kernel/trace/trace_syscalls.c b/kernel/trace/trace_syscalls.c index 4d6d711717f2..d036a74a64f3 100644 --- a/kernel/trace/trace_syscalls.c +++ b/kernel/trace/trace_syscalls.c @@ -15,6 +15,10 @@ static int sys_refcount_exit; static DECLARE_BITMAP(enabled_enter_syscalls, NR_syscalls); static DECLARE_BITMAP(enabled_exit_syscalls, NR_syscalls); +struct ftrace_event_class event_class_syscalls = { + .system = "syscalls" +}; + extern unsigned long __start_syscalls_metadata[]; extern unsigned long __stop_syscalls_metadata[]; -- cgit v1.2.3 From 38516ab59fbc5b3bb278cf5e1fe2867c70cff32e Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Tue, 20 Apr 2010 17:04:50 -0400 Subject: tracing: Let tracepoints have data passed to tracepoint callbacks This patch adds data to be passed to tracepoint callbacks. The created functions from DECLARE_TRACE() now need a mandatory data parameter. For example: DECLARE_TRACE(mytracepoint, int value, value) Will create the register function: int register_trace_mytracepoint((void(*)(void *data, int value))probe, void *data); As the first argument, all callbacks (probes) must take a (void *data) parameter. So a callback for the above tracepoint will look like: void myprobe(void *data, int value) { } The callback may choose to ignore the data parameter. This change allows callbacks to register a private data pointer along with the function probe. void mycallback(void *data, int value); register_trace_mytracepoint(mycallback, mydata); Then the mycallback() will receive the "mydata" as the first parameter before the args. A more detailed example: DECLARE_TRACE(mytracepoint, TP_PROTO(int status), TP_ARGS(status)); /* In the C file */ DEFINE_TRACE(mytracepoint, TP_PROTO(int status), TP_ARGS(status)); [...] trace_mytracepoint(status); /* In a file registering this tracepoint */ int my_callback(void *data, int status) { struct my_struct my_data = data; [...] } [...] my_data = kmalloc(sizeof(*my_data), GFP_KERNEL); init_my_data(my_data); register_trace_mytracepoint(my_callback, my_data); The same callback can also be registered to the same tracepoint as long as the data registered is different. Note, the data must also be used to unregister the callback: unregister_trace_mytracepoint(my_callback, my_data); Because of the data parameter, tracepoints declared this way can not have no args. That is: DECLARE_TRACE(mytracepoint, TP_PROTO(void), TP_ARGS()); will cause an error. If no arguments are needed, a new macro can be used instead: DECLARE_TRACE_NOARGS(mytracepoint); Since there are no arguments, the proto and args fields are left out. This is part of a series to make the tracepoint footprint smaller: text data bss dec hex filename 4913961 1088356 861512 6863829 68bbd5 vmlinux.orig 4914025 1088868 861512 6864405 68be15 vmlinux.class 4918492 1084612 861512 6864616 68bee8 vmlinux.tracepoint Again, this patch also increases the size of the kernel, but lays the ground work for decreasing it. v5: Fixed net/core/drop_monitor.c to handle these updates. v4: Moved the DECLARE_TRACE() DECLARE_TRACE_NOARGS out of the #ifdef CONFIG_TRACE_POINTS, since the two are the same in both cases. The __DECLARE_TRACE() is what changes. Thanks to Frederic Weisbecker for pointing this out. v3: Made all register_* functions require data to be passed and all callbacks to take a void * parameter as its first argument. This makes the calling functions comply with C standards. Also added more comments to the modifications of DECLARE_TRACE(). v2: Made the DECLARE_TRACE() have the ability to pass arguments and added a new DECLARE_TRACE_NOARGS() for tracepoints that do not need any arguments. Acked-by: Mathieu Desnoyers Acked-by: Masami Hiramatsu Acked-by: Frederic Weisbecker Cc: Neil Horman Cc: David S. Miller Signed-off-by: Steven Rostedt --- kernel/trace/blktrace.c | 125 +++++++++++++++++++++----------------- kernel/trace/ftrace.c | 7 ++- kernel/trace/kmemtrace.c | 42 +++++++------ kernel/trace/trace_sched_switch.c | 20 +++--- kernel/trace/trace_sched_wakeup.c | 28 +++++---- kernel/trace/trace_syscalls.c | 24 ++++---- kernel/trace/trace_workqueue.c | 26 ++++---- 7 files changed, 150 insertions(+), 122 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/blktrace.c b/kernel/trace/blktrace.c index b3bc91a3f510..19d93f83e3c3 100644 --- a/kernel/trace/blktrace.c +++ b/kernel/trace/blktrace.c @@ -675,28 +675,33 @@ static void blk_add_trace_rq(struct request_queue *q, struct request *rq, } } -static void blk_add_trace_rq_abort(struct request_queue *q, struct request *rq) +static void blk_add_trace_rq_abort(void *ignore, + struct request_queue *q, struct request *rq) { blk_add_trace_rq(q, rq, BLK_TA_ABORT); } -static void blk_add_trace_rq_insert(struct request_queue *q, struct request *rq) +static void blk_add_trace_rq_insert(void *ignore, + struct request_queue *q, struct request *rq) { blk_add_trace_rq(q, rq, BLK_TA_INSERT); } -static void blk_add_trace_rq_issue(struct request_queue *q, struct request *rq) +static void blk_add_trace_rq_issue(void *ignore, + struct request_queue *q, struct request *rq) { blk_add_trace_rq(q, rq, BLK_TA_ISSUE); } -static void blk_add_trace_rq_requeue(struct request_queue *q, +static void blk_add_trace_rq_requeue(void *ignore, + struct request_queue *q, struct request *rq) { blk_add_trace_rq(q, rq, BLK_TA_REQUEUE); } -static void blk_add_trace_rq_complete(struct request_queue *q, +static void blk_add_trace_rq_complete(void *ignore, + struct request_queue *q, struct request *rq) { blk_add_trace_rq(q, rq, BLK_TA_COMPLETE); @@ -724,34 +729,40 @@ static void blk_add_trace_bio(struct request_queue *q, struct bio *bio, !bio_flagged(bio, BIO_UPTODATE), 0, NULL); } -static void blk_add_trace_bio_bounce(struct request_queue *q, struct bio *bio) +static void blk_add_trace_bio_bounce(void *ignore, + struct request_queue *q, struct bio *bio) { blk_add_trace_bio(q, bio, BLK_TA_BOUNCE); } -static void blk_add_trace_bio_complete(struct request_queue *q, struct bio *bio) +static void blk_add_trace_bio_complete(void *ignore, + struct request_queue *q, struct bio *bio) { blk_add_trace_bio(q, bio, BLK_TA_COMPLETE); } -static void blk_add_trace_bio_backmerge(struct request_queue *q, +static void blk_add_trace_bio_backmerge(void *ignore, + struct request_queue *q, struct bio *bio) { blk_add_trace_bio(q, bio, BLK_TA_BACKMERGE); } -static void blk_add_trace_bio_frontmerge(struct request_queue *q, +static void blk_add_trace_bio_frontmerge(void *ignore, + struct request_queue *q, struct bio *bio) { blk_add_trace_bio(q, bio, BLK_TA_FRONTMERGE); } -static void blk_add_trace_bio_queue(struct request_queue *q, struct bio *bio) +static void blk_add_trace_bio_queue(void *ignore, + struct request_queue *q, struct bio *bio) { blk_add_trace_bio(q, bio, BLK_TA_QUEUE); } -static void blk_add_trace_getrq(struct request_queue *q, +static void blk_add_trace_getrq(void *ignore, + struct request_queue *q, struct bio *bio, int rw) { if (bio) @@ -765,7 +776,8 @@ static void blk_add_trace_getrq(struct request_queue *q, } -static void blk_add_trace_sleeprq(struct request_queue *q, +static void blk_add_trace_sleeprq(void *ignore, + struct request_queue *q, struct bio *bio, int rw) { if (bio) @@ -779,7 +791,7 @@ static void blk_add_trace_sleeprq(struct request_queue *q, } } -static void blk_add_trace_plug(struct request_queue *q) +static void blk_add_trace_plug(void *ignore, struct request_queue *q) { struct blk_trace *bt = q->blk_trace; @@ -787,7 +799,7 @@ static void blk_add_trace_plug(struct request_queue *q) __blk_add_trace(bt, 0, 0, 0, BLK_TA_PLUG, 0, 0, NULL); } -static void blk_add_trace_unplug_io(struct request_queue *q) +static void blk_add_trace_unplug_io(void *ignore, struct request_queue *q) { struct blk_trace *bt = q->blk_trace; @@ -800,7 +812,7 @@ static void blk_add_trace_unplug_io(struct request_queue *q) } } -static void blk_add_trace_unplug_timer(struct request_queue *q) +static void blk_add_trace_unplug_timer(void *ignore, struct request_queue *q) { struct blk_trace *bt = q->blk_trace; @@ -813,7 +825,8 @@ static void blk_add_trace_unplug_timer(struct request_queue *q) } } -static void blk_add_trace_split(struct request_queue *q, struct bio *bio, +static void blk_add_trace_split(void *ignore, + struct request_queue *q, struct bio *bio, unsigned int pdu) { struct blk_trace *bt = q->blk_trace; @@ -839,8 +852,9 @@ static void blk_add_trace_split(struct request_queue *q, struct bio *bio, * it spans a stripe (or similar). Add a trace for that action. * **/ -static void blk_add_trace_remap(struct request_queue *q, struct bio *bio, - dev_t dev, sector_t from) +static void blk_add_trace_remap(void *ignore, + struct request_queue *q, struct bio *bio, + dev_t dev, sector_t from) { struct blk_trace *bt = q->blk_trace; struct blk_io_trace_remap r; @@ -869,7 +883,8 @@ static void blk_add_trace_remap(struct request_queue *q, struct bio *bio, * Add a trace for that action. * **/ -static void blk_add_trace_rq_remap(struct request_queue *q, +static void blk_add_trace_rq_remap(void *ignore, + struct request_queue *q, struct request *rq, dev_t dev, sector_t from) { @@ -921,64 +936,64 @@ static void blk_register_tracepoints(void) { int ret; - ret = register_trace_block_rq_abort(blk_add_trace_rq_abort); + ret = register_trace_block_rq_abort(blk_add_trace_rq_abort, NULL); WARN_ON(ret); - ret = register_trace_block_rq_insert(blk_add_trace_rq_insert); + ret = register_trace_block_rq_insert(blk_add_trace_rq_insert, NULL); WARN_ON(ret); - ret = register_trace_block_rq_issue(blk_add_trace_rq_issue); + ret = register_trace_block_rq_issue(blk_add_trace_rq_issue, NULL); WARN_ON(ret); - ret = register_trace_block_rq_requeue(blk_add_trace_rq_requeue); + ret = register_trace_block_rq_requeue(blk_add_trace_rq_requeue, NULL); WARN_ON(ret); - ret = register_trace_block_rq_complete(blk_add_trace_rq_complete); + ret = register_trace_block_rq_complete(blk_add_trace_rq_complete, NULL); WARN_ON(ret); - ret = register_trace_block_bio_bounce(blk_add_trace_bio_bounce); + ret = register_trace_block_bio_bounce(blk_add_trace_bio_bounce, NULL); WARN_ON(ret); - ret = register_trace_block_bio_complete(blk_add_trace_bio_complete); + ret = register_trace_block_bio_complete(blk_add_trace_bio_complete, NULL); WARN_ON(ret); - ret = register_trace_block_bio_backmerge(blk_add_trace_bio_backmerge); + ret = register_trace_block_bio_backmerge(blk_add_trace_bio_backmerge, NULL); WARN_ON(ret); - ret = register_trace_block_bio_frontmerge(blk_add_trace_bio_frontmerge); + ret = register_trace_block_bio_frontmerge(blk_add_trace_bio_frontmerge, NULL); WARN_ON(ret); - ret = register_trace_block_bio_queue(blk_add_trace_bio_queue); + ret = register_trace_block_bio_queue(blk_add_trace_bio_queue, NULL); WARN_ON(ret); - ret = register_trace_block_getrq(blk_add_trace_getrq); + ret = register_trace_block_getrq(blk_add_trace_getrq, NULL); WARN_ON(ret); - ret = register_trace_block_sleeprq(blk_add_trace_sleeprq); + ret = register_trace_block_sleeprq(blk_add_trace_sleeprq, NULL); WARN_ON(ret); - ret = register_trace_block_plug(blk_add_trace_plug); + ret = register_trace_block_plug(blk_add_trace_plug, NULL); WARN_ON(ret); - ret = register_trace_block_unplug_timer(blk_add_trace_unplug_timer); + ret = register_trace_block_unplug_timer(blk_add_trace_unplug_timer, NULL); WARN_ON(ret); - ret = register_trace_block_unplug_io(blk_add_trace_unplug_io); + ret = register_trace_block_unplug_io(blk_add_trace_unplug_io, NULL); WARN_ON(ret); - ret = register_trace_block_split(blk_add_trace_split); + ret = register_trace_block_split(blk_add_trace_split, NULL); WARN_ON(ret); - ret = register_trace_block_remap(blk_add_trace_remap); + ret = register_trace_block_remap(blk_add_trace_remap, NULL); WARN_ON(ret); - ret = register_trace_block_rq_remap(blk_add_trace_rq_remap); + ret = register_trace_block_rq_remap(blk_add_trace_rq_remap, NULL); WARN_ON(ret); } static void blk_unregister_tracepoints(void) { - unregister_trace_block_rq_remap(blk_add_trace_rq_remap); - unregister_trace_block_remap(blk_add_trace_remap); - unregister_trace_block_split(blk_add_trace_split); - unregister_trace_block_unplug_io(blk_add_trace_unplug_io); - unregister_trace_block_unplug_timer(blk_add_trace_unplug_timer); - unregister_trace_block_plug(blk_add_trace_plug); - unregister_trace_block_sleeprq(blk_add_trace_sleeprq); - unregister_trace_block_getrq(blk_add_trace_getrq); - unregister_trace_block_bio_queue(blk_add_trace_bio_queue); - unregister_trace_block_bio_frontmerge(blk_add_trace_bio_frontmerge); - unregister_trace_block_bio_backmerge(blk_add_trace_bio_backmerge); - unregister_trace_block_bio_complete(blk_add_trace_bio_complete); - unregister_trace_block_bio_bounce(blk_add_trace_bio_bounce); - unregister_trace_block_rq_complete(blk_add_trace_rq_complete); - unregister_trace_block_rq_requeue(blk_add_trace_rq_requeue); - unregister_trace_block_rq_issue(blk_add_trace_rq_issue); - unregister_trace_block_rq_insert(blk_add_trace_rq_insert); - unregister_trace_block_rq_abort(blk_add_trace_rq_abort); + unregister_trace_block_rq_remap(blk_add_trace_rq_remap, NULL); + unregister_trace_block_remap(blk_add_trace_remap, NULL); + unregister_trace_block_split(blk_add_trace_split, NULL); + unregister_trace_block_unplug_io(blk_add_trace_unplug_io, NULL); + unregister_trace_block_unplug_timer(blk_add_trace_unplug_timer, NULL); + unregister_trace_block_plug(blk_add_trace_plug, NULL); + unregister_trace_block_sleeprq(blk_add_trace_sleeprq, NULL); + unregister_trace_block_getrq(blk_add_trace_getrq, NULL); + unregister_trace_block_bio_queue(blk_add_trace_bio_queue, NULL); + unregister_trace_block_bio_frontmerge(blk_add_trace_bio_frontmerge, NULL); + unregister_trace_block_bio_backmerge(blk_add_trace_bio_backmerge, NULL); + unregister_trace_block_bio_complete(blk_add_trace_bio_complete, NULL); + unregister_trace_block_bio_bounce(blk_add_trace_bio_bounce, NULL); + unregister_trace_block_rq_complete(blk_add_trace_rq_complete, NULL); + unregister_trace_block_rq_requeue(blk_add_trace_rq_requeue, NULL); + unregister_trace_block_rq_issue(blk_add_trace_rq_issue, NULL); + unregister_trace_block_rq_insert(blk_add_trace_rq_insert, NULL); + unregister_trace_block_rq_abort(blk_add_trace_rq_abort, NULL); tracepoint_synchronize_unregister(); } diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index 32837e19e3bd..6d2cb14f9449 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -3234,7 +3234,8 @@ free: } static void -ftrace_graph_probe_sched_switch(struct task_struct *prev, struct task_struct *next) +ftrace_graph_probe_sched_switch(void *ignore, + struct task_struct *prev, struct task_struct *next) { unsigned long long timestamp; int index; @@ -3288,7 +3289,7 @@ static int start_graph_tracing(void) } while (ret == -EAGAIN); if (!ret) { - ret = register_trace_sched_switch(ftrace_graph_probe_sched_switch); + ret = register_trace_sched_switch(ftrace_graph_probe_sched_switch, NULL); if (ret) pr_info("ftrace_graph: Couldn't activate tracepoint" " probe to kernel_sched_switch\n"); @@ -3364,7 +3365,7 @@ void unregister_ftrace_graph(void) ftrace_graph_entry = ftrace_graph_entry_stub; ftrace_shutdown(FTRACE_STOP_FUNC_RET); unregister_pm_notifier(&ftrace_suspend_notifier); - unregister_trace_sched_switch(ftrace_graph_probe_sched_switch); + unregister_trace_sched_switch(ftrace_graph_probe_sched_switch, NULL); out: mutex_unlock(&ftrace_lock); diff --git a/kernel/trace/kmemtrace.c b/kernel/trace/kmemtrace.c index a91da69f153a..7253d0c1c32f 100644 --- a/kernel/trace/kmemtrace.c +++ b/kernel/trace/kmemtrace.c @@ -95,7 +95,8 @@ static inline void kmemtrace_free(enum kmemtrace_type_id type_id, trace_wake_up(); } -static void kmemtrace_kmalloc(unsigned long call_site, +static void kmemtrace_kmalloc(void *ignore, + unsigned long call_site, const void *ptr, size_t bytes_req, size_t bytes_alloc, @@ -105,7 +106,8 @@ static void kmemtrace_kmalloc(unsigned long call_site, bytes_req, bytes_alloc, gfp_flags, -1); } -static void kmemtrace_kmem_cache_alloc(unsigned long call_site, +static void kmemtrace_kmem_cache_alloc(void *ignore, + unsigned long call_site, const void *ptr, size_t bytes_req, size_t bytes_alloc, @@ -115,7 +117,8 @@ static void kmemtrace_kmem_cache_alloc(unsigned long call_site, bytes_req, bytes_alloc, gfp_flags, -1); } -static void kmemtrace_kmalloc_node(unsigned long call_site, +static void kmemtrace_kmalloc_node(void *ignore, + unsigned long call_site, const void *ptr, size_t bytes_req, size_t bytes_alloc, @@ -126,7 +129,8 @@ static void kmemtrace_kmalloc_node(unsigned long call_site, bytes_req, bytes_alloc, gfp_flags, node); } -static void kmemtrace_kmem_cache_alloc_node(unsigned long call_site, +static void kmemtrace_kmem_cache_alloc_node(void *ignore, + unsigned long call_site, const void *ptr, size_t bytes_req, size_t bytes_alloc, @@ -137,12 +141,14 @@ static void kmemtrace_kmem_cache_alloc_node(unsigned long call_site, bytes_req, bytes_alloc, gfp_flags, node); } -static void kmemtrace_kfree(unsigned long call_site, const void *ptr) +static void +kmemtrace_kfree(void *ignore, unsigned long call_site, const void *ptr) { kmemtrace_free(KMEMTRACE_TYPE_KMALLOC, call_site, ptr); } -static void kmemtrace_kmem_cache_free(unsigned long call_site, const void *ptr) +static void kmemtrace_kmem_cache_free(void *ignore, + unsigned long call_site, const void *ptr) { kmemtrace_free(KMEMTRACE_TYPE_CACHE, call_site, ptr); } @@ -151,34 +157,34 @@ static int kmemtrace_start_probes(void) { int err; - err = register_trace_kmalloc(kmemtrace_kmalloc); + err = register_trace_kmalloc(kmemtrace_kmalloc, NULL); if (err) return err; - err = register_trace_kmem_cache_alloc(kmemtrace_kmem_cache_alloc); + err = register_trace_kmem_cache_alloc(kmemtrace_kmem_cache_alloc, NULL); if (err) return err; - err = register_trace_kmalloc_node(kmemtrace_kmalloc_node); + err = register_trace_kmalloc_node(kmemtrace_kmalloc_node, NULL); if (err) return err; - err = register_trace_kmem_cache_alloc_node(kmemtrace_kmem_cache_alloc_node); + err = register_trace_kmem_cache_alloc_node(kmemtrace_kmem_cache_alloc_node, NULL); if (err) return err; - err = register_trace_kfree(kmemtrace_kfree); + err = register_trace_kfree(kmemtrace_kfree, NULL); if (err) return err; - err = register_trace_kmem_cache_free(kmemtrace_kmem_cache_free); + err = register_trace_kmem_cache_free(kmemtrace_kmem_cache_free, NULL); return err; } static void kmemtrace_stop_probes(void) { - unregister_trace_kmalloc(kmemtrace_kmalloc); - unregister_trace_kmem_cache_alloc(kmemtrace_kmem_cache_alloc); - unregister_trace_kmalloc_node(kmemtrace_kmalloc_node); - unregister_trace_kmem_cache_alloc_node(kmemtrace_kmem_cache_alloc_node); - unregister_trace_kfree(kmemtrace_kfree); - unregister_trace_kmem_cache_free(kmemtrace_kmem_cache_free); + unregister_trace_kmalloc(kmemtrace_kmalloc, NULL); + unregister_trace_kmem_cache_alloc(kmemtrace_kmem_cache_alloc, NULL); + unregister_trace_kmalloc_node(kmemtrace_kmalloc_node, NULL); + unregister_trace_kmem_cache_alloc_node(kmemtrace_kmem_cache_alloc_node, NULL); + unregister_trace_kfree(kmemtrace_kfree, NULL); + unregister_trace_kmem_cache_free(kmemtrace_kmem_cache_free, NULL); } static int kmem_trace_init(struct trace_array *tr) diff --git a/kernel/trace/trace_sched_switch.c b/kernel/trace/trace_sched_switch.c index a55fccfede5d..8f758d070c43 100644 --- a/kernel/trace/trace_sched_switch.c +++ b/kernel/trace/trace_sched_switch.c @@ -50,7 +50,7 @@ tracing_sched_switch_trace(struct trace_array *tr, } static void -probe_sched_switch(struct task_struct *prev, struct task_struct *next) +probe_sched_switch(void *ignore, struct task_struct *prev, struct task_struct *next) { struct trace_array_cpu *data; unsigned long flags; @@ -108,7 +108,7 @@ tracing_sched_wakeup_trace(struct trace_array *tr, } static void -probe_sched_wakeup(struct task_struct *wakee, int success) +probe_sched_wakeup(void *ignore, struct task_struct *wakee, int success) { struct trace_array_cpu *data; unsigned long flags; @@ -138,21 +138,21 @@ static int tracing_sched_register(void) { int ret; - ret = register_trace_sched_wakeup(probe_sched_wakeup); + ret = register_trace_sched_wakeup(probe_sched_wakeup, NULL); if (ret) { pr_info("wakeup trace: Couldn't activate tracepoint" " probe to kernel_sched_wakeup\n"); return ret; } - ret = register_trace_sched_wakeup_new(probe_sched_wakeup); + ret = register_trace_sched_wakeup_new(probe_sched_wakeup, NULL); if (ret) { pr_info("wakeup trace: Couldn't activate tracepoint" " probe to kernel_sched_wakeup_new\n"); goto fail_deprobe; } - ret = register_trace_sched_switch(probe_sched_switch); + ret = register_trace_sched_switch(probe_sched_switch, NULL); if (ret) { pr_info("sched trace: Couldn't activate tracepoint" " probe to kernel_sched_switch\n"); @@ -161,17 +161,17 @@ static int tracing_sched_register(void) return ret; fail_deprobe_wake_new: - unregister_trace_sched_wakeup_new(probe_sched_wakeup); + unregister_trace_sched_wakeup_new(probe_sched_wakeup, NULL); fail_deprobe: - unregister_trace_sched_wakeup(probe_sched_wakeup); + unregister_trace_sched_wakeup(probe_sched_wakeup, NULL); return ret; } static void tracing_sched_unregister(void) { - unregister_trace_sched_switch(probe_sched_switch); - unregister_trace_sched_wakeup_new(probe_sched_wakeup); - unregister_trace_sched_wakeup(probe_sched_wakeup); + unregister_trace_sched_switch(probe_sched_switch, NULL); + unregister_trace_sched_wakeup_new(probe_sched_wakeup, NULL); + unregister_trace_sched_wakeup(probe_sched_wakeup, NULL); } static void tracing_start_sched_switch(void) diff --git a/kernel/trace/trace_sched_wakeup.c b/kernel/trace/trace_sched_wakeup.c index 8052446ceeaa..0e73bc2ef8c5 100644 --- a/kernel/trace/trace_sched_wakeup.c +++ b/kernel/trace/trace_sched_wakeup.c @@ -98,7 +98,8 @@ static int report_latency(cycle_t delta) return 1; } -static void probe_wakeup_migrate_task(struct task_struct *task, int cpu) +static void +probe_wakeup_migrate_task(void *ignore, struct task_struct *task, int cpu) { if (task != wakeup_task) return; @@ -107,7 +108,8 @@ static void probe_wakeup_migrate_task(struct task_struct *task, int cpu) } static void notrace -probe_wakeup_sched_switch(struct task_struct *prev, struct task_struct *next) +probe_wakeup_sched_switch(void *ignore, + struct task_struct *prev, struct task_struct *next) { struct trace_array_cpu *data; cycle_t T0, T1, delta; @@ -199,7 +201,7 @@ static void wakeup_reset(struct trace_array *tr) } static void -probe_wakeup(struct task_struct *p, int success) +probe_wakeup(void *ignore, struct task_struct *p, int success) { struct trace_array_cpu *data; int cpu = smp_processor_id(); @@ -263,28 +265,28 @@ static void start_wakeup_tracer(struct trace_array *tr) { int ret; - ret = register_trace_sched_wakeup(probe_wakeup); + ret = register_trace_sched_wakeup(probe_wakeup, NULL); if (ret) { pr_info("wakeup trace: Couldn't activate tracepoint" " probe to kernel_sched_wakeup\n"); return; } - ret = register_trace_sched_wakeup_new(probe_wakeup); + ret = register_trace_sched_wakeup_new(probe_wakeup, NULL); if (ret) { pr_info("wakeup trace: Couldn't activate tracepoint" " probe to kernel_sched_wakeup_new\n"); goto fail_deprobe; } - ret = register_trace_sched_switch(probe_wakeup_sched_switch); + ret = register_trace_sched_switch(probe_wakeup_sched_switch, NULL); if (ret) { pr_info("sched trace: Couldn't activate tracepoint" " probe to kernel_sched_switch\n"); goto fail_deprobe_wake_new; } - ret = register_trace_sched_migrate_task(probe_wakeup_migrate_task); + ret = register_trace_sched_migrate_task(probe_wakeup_migrate_task, NULL); if (ret) { pr_info("wakeup trace: Couldn't activate tracepoint" " probe to kernel_sched_migrate_task\n"); @@ -311,19 +313,19 @@ static void start_wakeup_tracer(struct trace_array *tr) return; fail_deprobe_wake_new: - unregister_trace_sched_wakeup_new(probe_wakeup); + unregister_trace_sched_wakeup_new(probe_wakeup, NULL); fail_deprobe: - unregister_trace_sched_wakeup(probe_wakeup); + unregister_trace_sched_wakeup(probe_wakeup, NULL); } static void stop_wakeup_tracer(struct trace_array *tr) { tracer_enabled = 0; unregister_ftrace_function(&trace_ops); - unregister_trace_sched_switch(probe_wakeup_sched_switch); - unregister_trace_sched_wakeup_new(probe_wakeup); - unregister_trace_sched_wakeup(probe_wakeup); - unregister_trace_sched_migrate_task(probe_wakeup_migrate_task); + unregister_trace_sched_switch(probe_wakeup_sched_switch, NULL); + unregister_trace_sched_wakeup_new(probe_wakeup, NULL); + unregister_trace_sched_wakeup(probe_wakeup, NULL); + unregister_trace_sched_migrate_task(probe_wakeup_migrate_task, NULL); } static int __wakeup_tracer_init(struct trace_array *tr) diff --git a/kernel/trace/trace_syscalls.c b/kernel/trace/trace_syscalls.c index d036a74a64f3..b8d30e7ecd05 100644 --- a/kernel/trace/trace_syscalls.c +++ b/kernel/trace/trace_syscalls.c @@ -247,7 +247,7 @@ int syscall_exit_define_fields(struct ftrace_event_call *call) return ret; } -void ftrace_syscall_enter(struct pt_regs *regs, long id) +void ftrace_syscall_enter(void *ignore, struct pt_regs *regs, long id) { struct syscall_trace_enter *entry; struct syscall_metadata *sys_data; @@ -282,7 +282,7 @@ void ftrace_syscall_enter(struct pt_regs *regs, long id) trace_current_buffer_unlock_commit(buffer, event, 0, 0); } -void ftrace_syscall_exit(struct pt_regs *regs, long ret) +void ftrace_syscall_exit(void *ignore, struct pt_regs *regs, long ret) { struct syscall_trace_exit *entry; struct syscall_metadata *sys_data; @@ -324,7 +324,7 @@ int reg_event_syscall_enter(struct ftrace_event_call *call) return -ENOSYS; mutex_lock(&syscall_trace_lock); if (!sys_refcount_enter) - ret = register_trace_sys_enter(ftrace_syscall_enter); + ret = register_trace_sys_enter(ftrace_syscall_enter, NULL); if (!ret) { set_bit(num, enabled_enter_syscalls); sys_refcount_enter++; @@ -344,7 +344,7 @@ void unreg_event_syscall_enter(struct ftrace_event_call *call) sys_refcount_enter--; clear_bit(num, enabled_enter_syscalls); if (!sys_refcount_enter) - unregister_trace_sys_enter(ftrace_syscall_enter); + unregister_trace_sys_enter(ftrace_syscall_enter, NULL); mutex_unlock(&syscall_trace_lock); } @@ -358,7 +358,7 @@ int reg_event_syscall_exit(struct ftrace_event_call *call) return -ENOSYS; mutex_lock(&syscall_trace_lock); if (!sys_refcount_exit) - ret = register_trace_sys_exit(ftrace_syscall_exit); + ret = register_trace_sys_exit(ftrace_syscall_exit, NULL); if (!ret) { set_bit(num, enabled_exit_syscalls); sys_refcount_exit++; @@ -378,7 +378,7 @@ void unreg_event_syscall_exit(struct ftrace_event_call *call) sys_refcount_exit--; clear_bit(num, enabled_exit_syscalls); if (!sys_refcount_exit) - unregister_trace_sys_exit(ftrace_syscall_exit); + unregister_trace_sys_exit(ftrace_syscall_exit, NULL); mutex_unlock(&syscall_trace_lock); } @@ -438,7 +438,7 @@ static DECLARE_BITMAP(enabled_perf_exit_syscalls, NR_syscalls); static int sys_perf_refcount_enter; static int sys_perf_refcount_exit; -static void perf_syscall_enter(struct pt_regs *regs, long id) +static void perf_syscall_enter(void *ignore, struct pt_regs *regs, long id) { struct syscall_metadata *sys_data; struct syscall_trace_enter *rec; @@ -484,7 +484,7 @@ int perf_sysenter_enable(struct ftrace_event_call *call) mutex_lock(&syscall_trace_lock); if (!sys_perf_refcount_enter) - ret = register_trace_sys_enter(perf_syscall_enter); + ret = register_trace_sys_enter(perf_syscall_enter, NULL); if (ret) { pr_info("event trace: Could not activate" "syscall entry trace point"); @@ -506,11 +506,11 @@ void perf_sysenter_disable(struct ftrace_event_call *call) sys_perf_refcount_enter--; clear_bit(num, enabled_perf_enter_syscalls); if (!sys_perf_refcount_enter) - unregister_trace_sys_enter(perf_syscall_enter); + unregister_trace_sys_enter(perf_syscall_enter, NULL); mutex_unlock(&syscall_trace_lock); } -static void perf_syscall_exit(struct pt_regs *regs, long ret) +static void perf_syscall_exit(void *ignore, struct pt_regs *regs, long ret) { struct syscall_metadata *sys_data; struct syscall_trace_exit *rec; @@ -559,7 +559,7 @@ int perf_sysexit_enable(struct ftrace_event_call *call) mutex_lock(&syscall_trace_lock); if (!sys_perf_refcount_exit) - ret = register_trace_sys_exit(perf_syscall_exit); + ret = register_trace_sys_exit(perf_syscall_exit, NULL); if (ret) { pr_info("event trace: Could not activate" "syscall exit trace point"); @@ -581,7 +581,7 @@ void perf_sysexit_disable(struct ftrace_event_call *call) sys_perf_refcount_exit--; clear_bit(num, enabled_perf_exit_syscalls); if (!sys_perf_refcount_exit) - unregister_trace_sys_exit(perf_syscall_exit); + unregister_trace_sys_exit(perf_syscall_exit, NULL); mutex_unlock(&syscall_trace_lock); } diff --git a/kernel/trace/trace_workqueue.c b/kernel/trace/trace_workqueue.c index cc2d2faa7d9e..a7cc3793baf6 100644 --- a/kernel/trace/trace_workqueue.c +++ b/kernel/trace/trace_workqueue.c @@ -49,7 +49,8 @@ static void cpu_workqueue_stat_free(struct kref *kref) /* Insertion of a work */ static void -probe_workqueue_insertion(struct task_struct *wq_thread, +probe_workqueue_insertion(void *ignore, + struct task_struct *wq_thread, struct work_struct *work) { int cpu = cpumask_first(&wq_thread->cpus_allowed); @@ -70,7 +71,8 @@ found: /* Execution of a work */ static void -probe_workqueue_execution(struct task_struct *wq_thread, +probe_workqueue_execution(void *ignore, + struct task_struct *wq_thread, struct work_struct *work) { int cpu = cpumask_first(&wq_thread->cpus_allowed); @@ -90,7 +92,8 @@ found: } /* Creation of a cpu workqueue thread */ -static void probe_workqueue_creation(struct task_struct *wq_thread, int cpu) +static void probe_workqueue_creation(void *ignore, + struct task_struct *wq_thread, int cpu) { struct cpu_workqueue_stats *cws; unsigned long flags; @@ -114,7 +117,8 @@ static void probe_workqueue_creation(struct task_struct *wq_thread, int cpu) } /* Destruction of a cpu workqueue thread */ -static void probe_workqueue_destruction(struct task_struct *wq_thread) +static void +probe_workqueue_destruction(void *ignore, struct task_struct *wq_thread) { /* Workqueue only execute on one cpu */ int cpu = cpumask_first(&wq_thread->cpus_allowed); @@ -259,19 +263,19 @@ int __init trace_workqueue_early_init(void) { int ret, cpu; - ret = register_trace_workqueue_insertion(probe_workqueue_insertion); + ret = register_trace_workqueue_insertion(probe_workqueue_insertion, NULL); if (ret) goto out; - ret = register_trace_workqueue_execution(probe_workqueue_execution); + ret = register_trace_workqueue_execution(probe_workqueue_execution, NULL); if (ret) goto no_insertion; - ret = register_trace_workqueue_creation(probe_workqueue_creation); + ret = register_trace_workqueue_creation(probe_workqueue_creation, NULL); if (ret) goto no_execution; - ret = register_trace_workqueue_destruction(probe_workqueue_destruction); + ret = register_trace_workqueue_destruction(probe_workqueue_destruction, NULL); if (ret) goto no_creation; @@ -283,11 +287,11 @@ int __init trace_workqueue_early_init(void) return 0; no_creation: - unregister_trace_workqueue_creation(probe_workqueue_creation); + unregister_trace_workqueue_creation(probe_workqueue_creation, NULL); no_execution: - unregister_trace_workqueue_execution(probe_workqueue_execution); + unregister_trace_workqueue_execution(probe_workqueue_execution, NULL); no_insertion: - unregister_trace_workqueue_insertion(probe_workqueue_insertion); + unregister_trace_workqueue_insertion(probe_workqueue_insertion, NULL); out: pr_warning("trace_workqueue: unable to trace workqueues\n"); -- cgit v1.2.3 From 2239291aeb0379fe47980b0e560e0eb9fd7e82ec Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Wed, 21 Apr 2010 12:27:06 -0400 Subject: tracing: Remove per event trace registering This patch removes the register functions of TRACE_EVENT() to enable and disable tracepoints. The registering of a event is now down directly in the trace_events.c file. The tracepoint_probe_register() is now called directly. The prototypes are no longer type checked, but this should not be an issue since the tracepoints are created automatically by the macros. If a prototype is incorrect in the TRACE_EVENT() macro, then other macros will catch it. The trace_event_class structure now holds the probes to be called by the callbacks. This removes needing to have each event have a separate pointer for the probe. To handle kprobes and syscalls, since they register probes in a different manner, a "reg" field is added to the ftrace_event_class structure. If the "reg" field is assigned, then it will be called for enabling and disabling of the probe for either ftrace or perf. To let the reg function know what is happening, a new enum (trace_reg) is created that has the type of control that is needed. With this new rework, the 82 kernel events and 618 syscall events has their footprint dramatically lowered: text data bss dec hex filename 4913961 1088356 861512 6863829 68bbd5 vmlinux.orig 4914025 1088868 861512 6864405 68be15 vmlinux.class 4918492 1084612 861512 6864616 68bee8 vmlinux.tracepoint 4900252 1057412 861512 6819176 680d68 vmlinux.regs The size went from 6863829 to 6819176, that's a total of 44K in savings. With tracepoints being continuously added, this is critical that the footprint becomes minimal. v5: Added #ifdef CONFIG_PERF_EVENTS around a reference to perf specific structure in trace_events.c. v4: Fixed trace self tests to check probe because regfunc no longer exists. v3: Updated to handle void *data in beginning of probe parameters. Also added the tracepoint: check_trace_callback_type_##call(). v2: Changed the callback probes to pass void * and typecast the value within the function. Acked-by: Mathieu Desnoyers Acked-by: Masami Hiramatsu Acked-by: Frederic Weisbecker Signed-off-by: Steven Rostedt --- kernel/trace/trace_event_perf.c | 15 ++++++++--- kernel/trace/trace_events.c | 32 ++++++++++++++++------- kernel/trace/trace_kprobe.c | 34 ++++++++++++++++++------- kernel/trace/trace_syscalls.c | 56 +++++++++++++++++++++++++++++++++++++++-- 4 files changed, 114 insertions(+), 23 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_event_perf.c b/kernel/trace/trace_event_perf.c index 0565bb42566f..196fe9d26773 100644 --- a/kernel/trace/trace_event_perf.c +++ b/kernel/trace/trace_event_perf.c @@ -49,7 +49,12 @@ static int perf_trace_event_enable(struct ftrace_event_call *event) rcu_assign_pointer(perf_trace_buf_nmi, buf); } - ret = event->perf_event_enable(event); + if (event->class->reg) + ret = event->class->reg(event, TRACE_REG_PERF_REGISTER); + else + ret = tracepoint_probe_register(event->name, + event->class->perf_probe, + event); if (!ret) { total_ref_count++; return 0; @@ -75,7 +80,8 @@ int perf_trace_enable(int event_id) mutex_lock(&event_mutex); list_for_each_entry(event, &ftrace_events, list) { - if (event->id == event_id && event->perf_event_enable && + if (event->id == event_id && + event->class && event->class->perf_probe && try_module_get(event->mod)) { ret = perf_trace_event_enable(event); break; @@ -93,7 +99,10 @@ static void perf_trace_event_disable(struct ftrace_event_call *event) if (--event->perf_refcount > 0) return; - event->perf_event_disable(event); + if (event->class->reg) + event->class->reg(event, TRACE_REG_PERF_UNREGISTER); + else + tracepoint_probe_unregister(event->name, event->class->perf_probe, event); if (!--total_ref_count) { buf = perf_trace_buf; diff --git a/kernel/trace/trace_events.c b/kernel/trace/trace_events.c index 2f54b48d3632..19d1eb0a7188 100644 --- a/kernel/trace/trace_events.c +++ b/kernel/trace/trace_events.c @@ -127,13 +127,23 @@ static int ftrace_event_enable_disable(struct ftrace_event_call *call, if (call->enabled) { call->enabled = 0; tracing_stop_cmdline_record(); - call->unregfunc(call); + if (call->class->reg) + call->class->reg(call, TRACE_REG_UNREGISTER); + else + tracepoint_probe_unregister(call->name, + call->class->probe, + call); } break; case 1: if (!call->enabled) { tracing_start_cmdline_record(); - ret = call->regfunc(call); + if (call->class->reg) + ret = call->class->reg(call, TRACE_REG_REGISTER); + else + ret = tracepoint_probe_register(call->name, + call->class->probe, + call); if (ret) { tracing_stop_cmdline_record(); pr_info("event trace: Could not enable event " @@ -171,7 +181,8 @@ static int __ftrace_set_clr_event(const char *match, const char *sub, mutex_lock(&event_mutex); list_for_each_entry(call, &ftrace_events, list) { - if (!call->name || !call->regfunc) + if (!call->name || !call->class || + (!call->class->probe && !call->class->reg)) continue; if (match && @@ -297,7 +308,7 @@ t_next(struct seq_file *m, void *v, loff_t *pos) * The ftrace subsystem is for showing formats only. * They can not be enabled or disabled via the event files. */ - if (call->regfunc) + if (call->class && (call->class->probe || call->class->reg)) return call; } @@ -450,7 +461,8 @@ system_enable_read(struct file *filp, char __user *ubuf, size_t cnt, mutex_lock(&event_mutex); list_for_each_entry(call, &ftrace_events, list) { - if (!call->name || !call->regfunc) + if (!call->name || !call->class || + (!call->class->probe && !call->class->reg)) continue; if (system && strcmp(call->class->system, system) != 0) @@ -935,13 +947,15 @@ event_create_dir(struct ftrace_event_call *call, struct dentry *d_events, return -1; } - if (call->regfunc) + if (call->class->probe || call->class->reg) trace_create_file("enable", 0644, call->dir, call, enable); - if (call->id && call->perf_event_enable) +#ifdef CONFIG_PERF_EVENTS + if (call->id && (call->class->perf_probe || call->class->reg)) trace_create_file("id", 0444, call->dir, call, id); +#endif if (call->define_fields) { ret = trace_define_common_fields(call); @@ -1388,8 +1402,8 @@ static __init void event_trace_self_tests(void) list_for_each_entry(call, &ftrace_events, list) { - /* Only test those that have a regfunc */ - if (!call->regfunc) + /* Only test those that have a probe */ + if (!call->class || !call->class->probe) continue; /* diff --git a/kernel/trace/trace_kprobe.c b/kernel/trace/trace_kprobe.c index eda220bf2065..f8af21a53f0c 100644 --- a/kernel/trace/trace_kprobe.c +++ b/kernel/trace/trace_kprobe.c @@ -202,6 +202,7 @@ struct trace_probe { unsigned long nhit; unsigned int flags; /* For TP_FLAG_* */ const char *symbol; /* symbol name */ + struct ftrace_event_class class; struct ftrace_event_call call; struct trace_event event; unsigned int nr_args; @@ -323,6 +324,7 @@ static struct trace_probe *alloc_trace_probe(const char *group, goto error; } + tp->call.class = &tp->class; tp->call.name = kstrdup(event, GFP_KERNEL); if (!tp->call.name) goto error; @@ -332,8 +334,8 @@ static struct trace_probe *alloc_trace_probe(const char *group, goto error; } - tp->call.class->system = kstrdup(group, GFP_KERNEL); - if (!tp->call.class->system) + tp->class.system = kstrdup(group, GFP_KERNEL); + if (!tp->class.system) goto error; INIT_LIST_HEAD(&tp->list); @@ -1302,6 +1304,26 @@ static void probe_perf_disable(struct ftrace_event_call *call) } #endif /* CONFIG_PERF_EVENTS */ +static __kprobes +int kprobe_register(struct ftrace_event_call *event, enum trace_reg type) +{ + switch (type) { + case TRACE_REG_REGISTER: + return probe_event_enable(event); + case TRACE_REG_UNREGISTER: + probe_event_disable(event); + return 0; + +#ifdef CONFIG_PERF_EVENTS + case TRACE_REG_PERF_REGISTER: + return probe_perf_enable(event); + case TRACE_REG_PERF_UNREGISTER: + probe_perf_disable(event); + return 0; +#endif + } + return 0; +} static __kprobes int kprobe_dispatcher(struct kprobe *kp, struct pt_regs *regs) @@ -1355,13 +1377,7 @@ static int register_probe_event(struct trace_probe *tp) return -ENODEV; } call->enabled = 0; - call->regfunc = probe_event_enable; - call->unregfunc = probe_event_disable; - -#ifdef CONFIG_PERF_EVENTS - call->perf_event_enable = probe_perf_enable; - call->perf_event_disable = probe_perf_disable; -#endif + call->class->reg = kprobe_register; call->data = tp; ret = trace_add_event_call(call); if (ret) { diff --git a/kernel/trace/trace_syscalls.c b/kernel/trace/trace_syscalls.c index b8d30e7ecd05..a21d366cae46 100644 --- a/kernel/trace/trace_syscalls.c +++ b/kernel/trace/trace_syscalls.c @@ -15,8 +15,19 @@ static int sys_refcount_exit; static DECLARE_BITMAP(enabled_enter_syscalls, NR_syscalls); static DECLARE_BITMAP(enabled_exit_syscalls, NR_syscalls); -struct ftrace_event_class event_class_syscalls = { - .system = "syscalls" +static int syscall_enter_register(struct ftrace_event_call *event, + enum trace_reg type); +static int syscall_exit_register(struct ftrace_event_call *event, + enum trace_reg type); + +struct ftrace_event_class event_class_syscall_enter = { + .system = "syscalls", + .reg = syscall_enter_register +}; + +struct ftrace_event_class event_class_syscall_exit = { + .system = "syscalls", + .reg = syscall_exit_register }; extern unsigned long __start_syscalls_metadata[]; @@ -587,3 +598,44 @@ void perf_sysexit_disable(struct ftrace_event_call *call) #endif /* CONFIG_PERF_EVENTS */ +static int syscall_enter_register(struct ftrace_event_call *event, + enum trace_reg type) +{ + switch (type) { + case TRACE_REG_REGISTER: + return reg_event_syscall_enter(event); + case TRACE_REG_UNREGISTER: + unreg_event_syscall_enter(event); + return 0; + +#ifdef CONFIG_PERF_EVENTS + case TRACE_REG_PERF_REGISTER: + return perf_sysenter_enable(event); + case TRACE_REG_PERF_UNREGISTER: + perf_sysenter_disable(event); + return 0; +#endif + } + return 0; +} + +static int syscall_exit_register(struct ftrace_event_call *event, + enum trace_reg type) +{ + switch (type) { + case TRACE_REG_REGISTER: + return reg_event_syscall_exit(event); + case TRACE_REG_UNREGISTER: + unreg_event_syscall_exit(event); + return 0; + +#ifdef CONFIG_PERF_EVENTS + case TRACE_REG_PERF_REGISTER: + return perf_sysexit_enable(event); + case TRACE_REG_PERF_UNREGISTER: + perf_sysexit_disable(event); + return 0; +#endif + } + return 0; +} -- cgit v1.2.3 From 2e33af029556cb8bd22bf4f86f42d540249177ea Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Thu, 22 Apr 2010 10:35:55 -0400 Subject: tracing: Move fields from event to class structure Move the defined fields from the event to the class structure. Since the fields of the event are defined by the class they belong to, it makes sense to have the class hold the information instead of the individual events. The events of the same class would just hold duplicate information. After this change the size of the kernel dropped another 3K: text data bss dec hex filename 4913961 1088356 861512 6863829 68bbd5 vmlinux.orig 4900252 1057412 861512 6819176 680d68 vmlinux.regs 4900375 1053380 861512 6815267 67fe23 vmlinux.fields Although the text increased, this was mainly due to the C files having to adapt to the change. This is a constant increase, where new tracepoints will not increase the Text. But the big drop is in the data size (as well as needed allocations to hold the fields). This will give even more savings as more tracepoints are created. Note, if just TRACE_EVENT()s are used and not DECLARE_EVENT_CLASS() with several DEFINE_EVENT()s, then the savings will be lost. But we are pushing developers to consolidate events with DEFINE_EVENT() so this should not be an issue. The kprobes define a unique class to every new event, but are dynamic so it should not be a issue. The syscalls however have a single class but the fields for the individual events are different. The syscalls use a metadata to define the fields. I moved the fields list from the event to the metadata and added a "get_fields()" function to the class. This function is used to find the fields. For normal events and kprobes, get_fields() just returns a pointer to the fields list_head in the class. For syscall events, it returns the fields list_head in the metadata for the event. v2: Fixed the syscall fields. The syscall metadata needs a list of fields for both enter and exit. Acked-by: Frederic Weisbecker Acked-by: Mathieu Desnoyers Acked-by: Masami Hiramatsu Cc: Tom Zanussi Cc: Peter Zijlstra Signed-off-by: Steven Rostedt --- kernel/trace/trace.h | 3 +++ kernel/trace/trace_events.c | 48 ++++++++++++++++++++++++++++---------- kernel/trace/trace_events_filter.c | 10 ++++---- kernel/trace/trace_export.c | 14 +++++------ kernel/trace/trace_kprobe.c | 8 +++---- kernel/trace/trace_syscalls.c | 31 ++++++++++++++++++++---- 6 files changed, 83 insertions(+), 31 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace.h b/kernel/trace/trace.h index 911e9864e94a..c88c563a59a5 100644 --- a/kernel/trace/trace.h +++ b/kernel/trace/trace.h @@ -794,6 +794,9 @@ extern void print_subsystem_event_filter(struct event_subsystem *system, struct trace_seq *s); extern int filter_assign_type(const char *type); +struct list_head * +trace_get_fields(struct ftrace_event_call *event_call); + static inline int filter_check_discard(struct ftrace_event_call *call, void *rec, struct ring_buffer *buffer, diff --git a/kernel/trace/trace_events.c b/kernel/trace/trace_events.c index 19d1eb0a7188..acc0f55742c3 100644 --- a/kernel/trace/trace_events.c +++ b/kernel/trace/trace_events.c @@ -29,11 +29,23 @@ DEFINE_MUTEX(event_mutex); LIST_HEAD(ftrace_events); +struct list_head * +trace_get_fields(struct ftrace_event_call *event_call) +{ + if (!event_call->class->get_fields) + return &event_call->class->fields; + return event_call->class->get_fields(event_call); +} + int trace_define_field(struct ftrace_event_call *call, const char *type, const char *name, int offset, int size, int is_signed, int filter_type) { struct ftrace_event_field *field; + struct list_head *head; + + if (WARN_ON(!call->class)) + return 0; field = kzalloc(sizeof(*field), GFP_KERNEL); if (!field) @@ -56,7 +68,8 @@ int trace_define_field(struct ftrace_event_call *call, const char *type, field->size = size; field->is_signed = is_signed; - list_add(&field->link, &call->fields); + head = trace_get_fields(call); + list_add(&field->link, head); return 0; @@ -94,8 +107,10 @@ static int trace_define_common_fields(struct ftrace_event_call *call) void trace_destroy_fields(struct ftrace_event_call *call) { struct ftrace_event_field *field, *next; + struct list_head *head; - list_for_each_entry_safe(field, next, &call->fields, link) { + head = trace_get_fields(call); + list_for_each_entry_safe(field, next, head, link) { list_del(&field->link); kfree(field->type); kfree(field->name); @@ -111,7 +126,6 @@ int trace_event_raw_init(struct ftrace_event_call *call) if (!id) return -ENODEV; call->id = id; - INIT_LIST_HEAD(&call->fields); return 0; } @@ -537,6 +551,7 @@ event_format_read(struct file *filp, char __user *ubuf, size_t cnt, { struct ftrace_event_call *call = filp->private_data; struct ftrace_event_field *field; + struct list_head *head; struct trace_seq *s; int common_field_count = 5; char *buf; @@ -555,7 +570,8 @@ event_format_read(struct file *filp, char __user *ubuf, size_t cnt, trace_seq_printf(s, "ID: %d\n", call->id); trace_seq_printf(s, "format:\n"); - list_for_each_entry_reverse(field, &call->fields, link) { + head = trace_get_fields(call); + list_for_each_entry_reverse(field, head, link) { /* * Smartly shows the array type(except dynamic array). * Normal: @@ -931,6 +947,7 @@ event_create_dir(struct ftrace_event_call *call, struct dentry *d_events, const struct file_operations *filter, const struct file_operations *format) { + struct list_head *head; int ret; /* @@ -957,14 +974,21 @@ event_create_dir(struct ftrace_event_call *call, struct dentry *d_events, id); #endif - if (call->define_fields) { - ret = trace_define_common_fields(call); - if (!ret) - ret = call->define_fields(call); - if (ret < 0) { - pr_warning("Could not initialize trace point" - " events/%s\n", call->name); - return ret; + if (call->class->define_fields) { + /* + * Other events may have the same class. Only update + * the fields if they are not already defined. + */ + head = trace_get_fields(call); + if (list_empty(head)) { + ret = trace_define_common_fields(call); + if (!ret) + ret = call->class->define_fields(call); + if (ret < 0) { + pr_warning("Could not initialize trace point" + " events/%s\n", call->name); + return ret; + } } trace_create_file("filter", 0644, call->dir, call, filter); diff --git a/kernel/trace/trace_events_filter.c b/kernel/trace/trace_events_filter.c index ca329603d0bf..961f99b74bdd 100644 --- a/kernel/trace/trace_events_filter.c +++ b/kernel/trace/trace_events_filter.c @@ -500,8 +500,10 @@ static struct ftrace_event_field * find_event_field(struct ftrace_event_call *call, char *name) { struct ftrace_event_field *field; + struct list_head *head; - list_for_each_entry(field, &call->fields, link) { + head = trace_get_fields(call); + list_for_each_entry(field, head, link) { if (!strcmp(field->name, name)) return field; } @@ -625,7 +627,7 @@ static int init_subsystem_preds(struct event_subsystem *system) int err; list_for_each_entry(call, &ftrace_events, list) { - if (!call->define_fields) + if (!call->class || !call->class->define_fields) continue; if (strcmp(call->class->system, system->name) != 0) @@ -644,7 +646,7 @@ static void filter_free_subsystem_preds(struct event_subsystem *system) struct ftrace_event_call *call; list_for_each_entry(call, &ftrace_events, list) { - if (!call->define_fields) + if (!call->class || !call->class->define_fields) continue; if (strcmp(call->class->system, system->name) != 0) @@ -1249,7 +1251,7 @@ static int replace_system_preds(struct event_subsystem *system, list_for_each_entry(call, &ftrace_events, list) { struct event_filter *filter = call->filter; - if (!call->define_fields) + if (!call->class || !call->class->define_fields) continue; if (strcmp(call->class->system, system->name) != 0) diff --git a/kernel/trace/trace_export.c b/kernel/trace/trace_export.c index 7f16e2163817..e700a0c1803f 100644 --- a/kernel/trace/trace_export.c +++ b/kernel/trace/trace_export.c @@ -18,10 +18,6 @@ #undef TRACE_SYSTEM #define TRACE_SYSTEM ftrace -struct ftrace_event_class event_class_ftrace = { - .system = __stringify(TRACE_SYSTEM), -}; - /* not needed for this file */ #undef __field_struct #define __field_struct(type, item) @@ -131,7 +127,7 @@ ftrace_define_fields_##name(struct ftrace_event_call *event_call) \ static int ftrace_raw_init_event(struct ftrace_event_call *call) { - INIT_LIST_HEAD(&call->fields); + INIT_LIST_HEAD(&call->class->fields); return 0; } @@ -159,15 +155,19 @@ static int ftrace_raw_init_event(struct ftrace_event_call *call) #undef FTRACE_ENTRY #define FTRACE_ENTRY(call, struct_name, type, tstruct, print) \ \ +struct ftrace_event_class event_class_ftrace_##call = { \ + .system = __stringify(TRACE_SYSTEM), \ + .define_fields = ftrace_define_fields_##call, \ +}; \ + \ struct ftrace_event_call __used \ __attribute__((__aligned__(4))) \ __attribute__((section("_ftrace_events"))) event_##call = { \ .name = #call, \ .id = type, \ - .class = &event_class_ftrace, \ + .class = &event_class_ftrace_##call, \ .raw_init = ftrace_raw_init_event, \ .print_fmt = print, \ - .define_fields = ftrace_define_fields_##call, \ }; \ #include "trace_entries.h" diff --git a/kernel/trace/trace_kprobe.c b/kernel/trace/trace_kprobe.c index f8af21a53f0c..b14bf745356f 100644 --- a/kernel/trace/trace_kprobe.c +++ b/kernel/trace/trace_kprobe.c @@ -1112,8 +1112,6 @@ static void probe_event_disable(struct ftrace_event_call *call) static int probe_event_raw_init(struct ftrace_event_call *event_call) { - INIT_LIST_HEAD(&event_call->fields); - return 0; } @@ -1362,11 +1360,13 @@ static int register_probe_event(struct trace_probe *tp) if (probe_is_return(tp)) { tp->event.trace = print_kretprobe_event; call->raw_init = probe_event_raw_init; - call->define_fields = kretprobe_event_define_fields; + INIT_LIST_HEAD(&call->class->fields); + call->class->define_fields = kretprobe_event_define_fields; } else { tp->event.trace = print_kprobe_event; call->raw_init = probe_event_raw_init; - call->define_fields = kprobe_event_define_fields; + INIT_LIST_HEAD(&call->class->fields); + call->class->define_fields = kprobe_event_define_fields; } if (set_print_fmt(tp) < 0) return -ENOMEM; diff --git a/kernel/trace/trace_syscalls.c b/kernel/trace/trace_syscalls.c index a21d366cae46..cceccf0d2e91 100644 --- a/kernel/trace/trace_syscalls.c +++ b/kernel/trace/trace_syscalls.c @@ -20,14 +20,37 @@ static int syscall_enter_register(struct ftrace_event_call *event, static int syscall_exit_register(struct ftrace_event_call *event, enum trace_reg type); +static int syscall_enter_define_fields(struct ftrace_event_call *call); +static int syscall_exit_define_fields(struct ftrace_event_call *call); + +static struct list_head * +syscall_get_enter_fields(struct ftrace_event_call *call) +{ + struct syscall_metadata *entry = call->data; + + return &entry->enter_fields; +} + +static struct list_head * +syscall_get_exit_fields(struct ftrace_event_call *call) +{ + struct syscall_metadata *entry = call->data; + + return &entry->exit_fields; +} + struct ftrace_event_class event_class_syscall_enter = { .system = "syscalls", - .reg = syscall_enter_register + .reg = syscall_enter_register, + .define_fields = syscall_enter_define_fields, + .get_fields = syscall_get_enter_fields, }; struct ftrace_event_class event_class_syscall_exit = { .system = "syscalls", - .reg = syscall_exit_register + .reg = syscall_exit_register, + .define_fields = syscall_exit_define_fields, + .get_fields = syscall_get_exit_fields, }; extern unsigned long __start_syscalls_metadata[]; @@ -220,7 +243,7 @@ static void free_syscall_print_fmt(struct ftrace_event_call *call) kfree(call->print_fmt); } -int syscall_enter_define_fields(struct ftrace_event_call *call) +static int syscall_enter_define_fields(struct ftrace_event_call *call) { struct syscall_trace_enter trace; struct syscall_metadata *meta = call->data; @@ -243,7 +266,7 @@ int syscall_enter_define_fields(struct ftrace_event_call *call) return ret; } -int syscall_exit_define_fields(struct ftrace_event_call *call) +static int syscall_exit_define_fields(struct ftrace_event_call *call) { struct syscall_trace_exit trace; int ret; -- cgit v1.2.3 From 0405ab80aa94afb13bf9ac4a6fc9f2923d4b9114 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Thu, 22 Apr 2010 11:46:44 -0400 Subject: tracing: Move raw_init from events to class The raw_init function pointer in the event is used to initialize various kinds of events. The type of initialization needed is usually classed to the kind of event it is. Two events with the same class will always have the same initialization function, so it makes sense to move this to the class structure. Perhaps even making a special system structure would work since the initialization is the same for all events within a system. But since there's no system structure (yet), this will just move it to the class. text data bss dec hex filename 4913961 1088356 861512 6863829 68bbd5 vmlinux.orig 4900375 1053380 861512 6815267 67fe23 vmlinux.fields 4900382 1048964 861512 6810858 67ecea vmlinux.init The text grew very slightly, but this is a constant growth that happened with the changing of the C files that call the init code. The bigger savings is the data which will be saved the more events share a class. Acked-by: Mathieu Desnoyers Acked-by: Masami Hiramatsu Acked-by: Frederic Weisbecker Signed-off-by: Steven Rostedt --- kernel/trace/trace_events.c | 12 ++++++------ kernel/trace/trace_export.c | 2 +- kernel/trace/trace_kprobe.c | 6 +++--- kernel/trace/trace_syscalls.c | 2 ++ 4 files changed, 12 insertions(+), 10 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_events.c b/kernel/trace/trace_events.c index acc0f55742c3..8e94255a45e7 100644 --- a/kernel/trace/trace_events.c +++ b/kernel/trace/trace_events.c @@ -1008,8 +1008,8 @@ static int __trace_add_event_call(struct ftrace_event_call *call) if (!call->name) return -EINVAL; - if (call->raw_init) { - ret = call->raw_init(call); + if (call->class->raw_init) { + ret = call->class->raw_init(call); if (ret < 0) { if (ret != -ENOSYS) pr_warning("Could not initialize trace " @@ -1170,8 +1170,8 @@ static void trace_module_add_events(struct module *mod) /* The linker may leave blanks */ if (!call->name) continue; - if (call->raw_init) { - ret = call->raw_init(call); + if (call->class->raw_init) { + ret = call->class->raw_init(call); if (ret < 0) { if (ret != -ENOSYS) pr_warning("Could not initialize trace " @@ -1324,8 +1324,8 @@ static __init int event_trace_init(void) /* The linker may leave blanks */ if (!call->name) continue; - if (call->raw_init) { - ret = call->raw_init(call); + if (call->class->raw_init) { + ret = call->class->raw_init(call); if (ret < 0) { if (ret != -ENOSYS) pr_warning("Could not initialize trace " diff --git a/kernel/trace/trace_export.c b/kernel/trace/trace_export.c index e700a0c1803f..e878d06c0ac0 100644 --- a/kernel/trace/trace_export.c +++ b/kernel/trace/trace_export.c @@ -158,6 +158,7 @@ static int ftrace_raw_init_event(struct ftrace_event_call *call) struct ftrace_event_class event_class_ftrace_##call = { \ .system = __stringify(TRACE_SYSTEM), \ .define_fields = ftrace_define_fields_##call, \ + .raw_init = ftrace_raw_init_event, \ }; \ \ struct ftrace_event_call __used \ @@ -166,7 +167,6 @@ __attribute__((section("_ftrace_events"))) event_##call = { \ .name = #call, \ .id = type, \ .class = &event_class_ftrace_##call, \ - .raw_init = ftrace_raw_init_event, \ .print_fmt = print, \ }; \ diff --git a/kernel/trace/trace_kprobe.c b/kernel/trace/trace_kprobe.c index b14bf745356f..428f4a52de69 100644 --- a/kernel/trace/trace_kprobe.c +++ b/kernel/trace/trace_kprobe.c @@ -1359,13 +1359,13 @@ static int register_probe_event(struct trace_probe *tp) /* Initialize ftrace_event_call */ if (probe_is_return(tp)) { tp->event.trace = print_kretprobe_event; - call->raw_init = probe_event_raw_init; INIT_LIST_HEAD(&call->class->fields); + call->class->raw_init = probe_event_raw_init; call->class->define_fields = kretprobe_event_define_fields; } else { - tp->event.trace = print_kprobe_event; - call->raw_init = probe_event_raw_init; INIT_LIST_HEAD(&call->class->fields); + tp->event.trace = print_kprobe_event; + call->class->raw_init = probe_event_raw_init; call->class->define_fields = kprobe_event_define_fields; } if (set_print_fmt(tp) < 0) diff --git a/kernel/trace/trace_syscalls.c b/kernel/trace/trace_syscalls.c index cceccf0d2e91..1c1b9bf3ccc6 100644 --- a/kernel/trace/trace_syscalls.c +++ b/kernel/trace/trace_syscalls.c @@ -44,6 +44,7 @@ struct ftrace_event_class event_class_syscall_enter = { .reg = syscall_enter_register, .define_fields = syscall_enter_define_fields, .get_fields = syscall_get_enter_fields, + .raw_init = init_syscall_trace, }; struct ftrace_event_class event_class_syscall_exit = { @@ -51,6 +52,7 @@ struct ftrace_event_class event_class_syscall_exit = { .reg = syscall_exit_register, .define_fields = syscall_exit_define_fields, .get_fields = syscall_get_exit_fields, + .raw_init = init_syscall_trace, }; extern unsigned long __start_syscalls_metadata[]; -- cgit v1.2.3 From a9a5776380208a3e48a92d0c763ee1a3b486fb73 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Thu, 22 Apr 2010 18:46:14 -0400 Subject: tracing: Allow events to share their print functions Multiple events may use the same method to print their data. Instead of having all events have a pointer to their print funtions, the trace_event structure now points to a trace_event_functions structure that will hold the way to print ouf the event. The event itself is now passed to the print function to let the print function know what kind of event it should print. This opens the door to consolidating the way several events print their output. text data bss dec hex filename 4913961 1088356 861512 6863829 68bbd5 vmlinux.orig 4900382 1048964 861512 6810858 67ecea vmlinux.init 4900446 1049028 861512 6810986 67ed6a vmlinux.preprint This change slightly increases the size but is needed for the next change. v3: Fix the branch tracer events to handle this change. v2: Fix the new function graph tracer event calls to handle this change. Acked-by: Mathieu Desnoyers Acked-by: Masami Hiramatsu Acked-by: Frederic Weisbecker Signed-off-by: Steven Rostedt --- kernel/trace/blktrace.c | 13 +++- kernel/trace/kmemtrace.c | 28 +++++-- kernel/trace/trace.c | 9 ++- kernel/trace/trace_branch.c | 8 +- kernel/trace/trace_functions_graph.c | 13 +++- kernel/trace/trace_kprobe.c | 22 ++++-- kernel/trace/trace_output.c | 137 +++++++++++++++++++++++------------ kernel/trace/trace_output.h | 2 +- kernel/trace/trace_syscalls.c | 6 +- 9 files changed, 160 insertions(+), 78 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/blktrace.c b/kernel/trace/blktrace.c index 19d93f83e3c3..36ea2b65dcdc 100644 --- a/kernel/trace/blktrace.c +++ b/kernel/trace/blktrace.c @@ -1336,7 +1336,7 @@ out: } static enum print_line_t blk_trace_event_print(struct trace_iterator *iter, - int flags) + int flags, struct trace_event *event) { return print_one_line(iter, false); } @@ -1358,7 +1358,8 @@ static int blk_trace_synthesize_old_trace(struct trace_iterator *iter) } static enum print_line_t -blk_trace_event_print_binary(struct trace_iterator *iter, int flags) +blk_trace_event_print_binary(struct trace_iterator *iter, int flags, + struct trace_event *event) { return blk_trace_synthesize_old_trace(iter) ? TRACE_TYPE_HANDLED : TRACE_TYPE_PARTIAL_LINE; @@ -1396,12 +1397,16 @@ static struct tracer blk_tracer __read_mostly = { .set_flag = blk_tracer_set_flag, }; -static struct trace_event trace_blk_event = { - .type = TRACE_BLK, +static struct trace_event_functions trace_blk_event_funcs = { .trace = blk_trace_event_print, .binary = blk_trace_event_print_binary, }; +static struct trace_event trace_blk_event = { + .type = TRACE_BLK, + .funcs = &trace_blk_event_funcs, +}; + static int __init init_blk_tracer(void) { if (!register_ftrace_event(&trace_blk_event)) { diff --git a/kernel/trace/kmemtrace.c b/kernel/trace/kmemtrace.c index 7253d0c1c32f..bbfc1bb1660b 100644 --- a/kernel/trace/kmemtrace.c +++ b/kernel/trace/kmemtrace.c @@ -243,7 +243,8 @@ struct kmemtrace_user_event_alloc { }; static enum print_line_t -kmemtrace_print_alloc(struct trace_iterator *iter, int flags) +kmemtrace_print_alloc(struct trace_iterator *iter, int flags, + struct trace_event *event) { struct trace_seq *s = &iter->seq; struct kmemtrace_alloc_entry *entry; @@ -263,7 +264,8 @@ kmemtrace_print_alloc(struct trace_iterator *iter, int flags) } static enum print_line_t -kmemtrace_print_free(struct trace_iterator *iter, int flags) +kmemtrace_print_free(struct trace_iterator *iter, int flags, + struct trace_event *event) { struct trace_seq *s = &iter->seq; struct kmemtrace_free_entry *entry; @@ -281,7 +283,8 @@ kmemtrace_print_free(struct trace_iterator *iter, int flags) } static enum print_line_t -kmemtrace_print_alloc_user(struct trace_iterator *iter, int flags) +kmemtrace_print_alloc_user(struct trace_iterator *iter, int flags, + struct trace_event *event) { struct trace_seq *s = &iter->seq; struct kmemtrace_alloc_entry *entry; @@ -315,7 +318,8 @@ kmemtrace_print_alloc_user(struct trace_iterator *iter, int flags) } static enum print_line_t -kmemtrace_print_free_user(struct trace_iterator *iter, int flags) +kmemtrace_print_free_user(struct trace_iterator *iter, int flags, + struct trace_event *event) { struct trace_seq *s = &iter->seq; struct kmemtrace_free_entry *entry; @@ -469,18 +473,26 @@ static enum print_line_t kmemtrace_print_line(struct trace_iterator *iter) } } -static struct trace_event kmem_trace_alloc = { - .type = TRACE_KMEM_ALLOC, +static struct trace_event_functions kmem_trace_alloc_funcs = { .trace = kmemtrace_print_alloc, .binary = kmemtrace_print_alloc_user, }; -static struct trace_event kmem_trace_free = { - .type = TRACE_KMEM_FREE, +static struct trace_event kmem_trace_alloc = { + .type = TRACE_KMEM_ALLOC, + .funcs = &kmem_trace_alloc_funcs, +}; + +static struct trace_event_functions kmem_trace_free_funcs = { .trace = kmemtrace_print_free, .binary = kmemtrace_print_free_user, }; +static struct trace_event kmem_trace_free = { + .type = TRACE_KMEM_FREE, + .funcs = &kmem_trace_free_funcs, +}; + static struct tracer kmem_tracer __read_mostly = { .name = "kmemtrace", .init = kmem_trace_init, diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index 756d7283318b..ba0ec81158b2 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -1936,7 +1936,7 @@ static enum print_line_t print_trace_fmt(struct trace_iterator *iter) } if (event) - return event->trace(iter, sym_flags); + return event->funcs->trace(iter, sym_flags, event); if (!trace_seq_printf(s, "Unknown type %d\n", entry->type)) goto partial; @@ -1962,7 +1962,7 @@ static enum print_line_t print_raw_fmt(struct trace_iterator *iter) event = ftrace_find_event(entry->type); if (event) - return event->raw(iter, 0); + return event->funcs->raw(iter, 0, event); if (!trace_seq_printf(s, "%d ?\n", entry->type)) goto partial; @@ -1989,7 +1989,7 @@ static enum print_line_t print_hex_fmt(struct trace_iterator *iter) event = ftrace_find_event(entry->type); if (event) { - enum print_line_t ret = event->hex(iter, 0); + enum print_line_t ret = event->funcs->hex(iter, 0, event); if (ret != TRACE_TYPE_HANDLED) return ret; } @@ -2014,7 +2014,8 @@ static enum print_line_t print_bin_fmt(struct trace_iterator *iter) } event = ftrace_find_event(entry->type); - return event ? event->binary(iter, 0) : TRACE_TYPE_HANDLED; + return event ? event->funcs->binary(iter, 0, event) : + TRACE_TYPE_HANDLED; } int trace_empty(struct trace_iterator *iter) diff --git a/kernel/trace/trace_branch.c b/kernel/trace/trace_branch.c index b9bc4d470177..8d3538b4ea5f 100644 --- a/kernel/trace/trace_branch.c +++ b/kernel/trace/trace_branch.c @@ -143,7 +143,7 @@ static void branch_trace_reset(struct trace_array *tr) } static enum print_line_t trace_branch_print(struct trace_iterator *iter, - int flags) + int flags, struct trace_event *event) { struct trace_branch *field; @@ -167,9 +167,13 @@ static void branch_print_header(struct seq_file *s) " |\n"); } +static struct trace_event_functions trace_branch_funcs = { + .trace = trace_branch_print, +}; + static struct trace_event trace_branch_event = { .type = TRACE_BRANCH, - .trace = trace_branch_print, + .funcs = &trace_branch_funcs, }; static struct tracer branch_trace __read_mostly = diff --git a/kernel/trace/trace_functions_graph.c b/kernel/trace/trace_functions_graph.c index dd11c830eb84..79f4bac99a94 100644 --- a/kernel/trace/trace_functions_graph.c +++ b/kernel/trace/trace_functions_graph.c @@ -1025,7 +1025,7 @@ print_graph_comment(struct trace_seq *s, struct trace_entry *ent, if (!event) return TRACE_TYPE_UNHANDLED; - ret = event->trace(iter, sym_flags); + ret = event->funcs->trace(iter, sym_flags, event); if (ret != TRACE_TYPE_HANDLED) return ret; } @@ -1112,7 +1112,8 @@ print_graph_function(struct trace_iterator *iter) } static enum print_line_t -print_graph_function_event(struct trace_iterator *iter, int flags) +print_graph_function_event(struct trace_iterator *iter, int flags, + struct trace_event *event) { return print_graph_function(iter); } @@ -1225,14 +1226,18 @@ void graph_trace_close(struct trace_iterator *iter) } } +static struct trace_event_functions graph_functions = { + .trace = print_graph_function_event, +}; + static struct trace_event graph_trace_entry_event = { .type = TRACE_GRAPH_ENT, - .trace = print_graph_function_event, + .funcs = &graph_functions, }; static struct trace_event graph_trace_ret_event = { .type = TRACE_GRAPH_RET, - .trace = print_graph_function_event, + .funcs = &graph_functions }; static struct tracer graph_trace __read_mostly = { diff --git a/kernel/trace/trace_kprobe.c b/kernel/trace/trace_kprobe.c index 428f4a52de69..b989ae229a20 100644 --- a/kernel/trace/trace_kprobe.c +++ b/kernel/trace/trace_kprobe.c @@ -1011,16 +1011,15 @@ static __kprobes void kretprobe_trace_func(struct kretprobe_instance *ri, /* Event entry printers */ enum print_line_t -print_kprobe_event(struct trace_iterator *iter, int flags) +print_kprobe_event(struct trace_iterator *iter, int flags, + struct trace_event *event) { struct kprobe_trace_entry *field; struct trace_seq *s = &iter->seq; - struct trace_event *event; struct trace_probe *tp; int i; field = (struct kprobe_trace_entry *)iter->ent; - event = ftrace_find_event(field->ent.type); tp = container_of(event, struct trace_probe, event); if (!trace_seq_printf(s, "%s: (", tp->call.name)) @@ -1046,16 +1045,15 @@ partial: } enum print_line_t -print_kretprobe_event(struct trace_iterator *iter, int flags) +print_kretprobe_event(struct trace_iterator *iter, int flags, + struct trace_event *event) { struct kretprobe_trace_entry *field; struct trace_seq *s = &iter->seq; - struct trace_event *event; struct trace_probe *tp; int i; field = (struct kretprobe_trace_entry *)iter->ent; - event = ftrace_find_event(field->ent.type); tp = container_of(event, struct trace_probe, event); if (!trace_seq_printf(s, "%s: (", tp->call.name)) @@ -1351,6 +1349,14 @@ int kretprobe_dispatcher(struct kretprobe_instance *ri, struct pt_regs *regs) return 0; /* We don't tweek kernel, so just return 0 */ } +static struct trace_event_functions kretprobe_funcs = { + .trace = print_kretprobe_event +}; + +static struct trace_event_functions kprobe_funcs = { + .trace = print_kprobe_event +}; + static int register_probe_event(struct trace_probe *tp) { struct ftrace_event_call *call = &tp->call; @@ -1358,13 +1364,13 @@ static int register_probe_event(struct trace_probe *tp) /* Initialize ftrace_event_call */ if (probe_is_return(tp)) { - tp->event.trace = print_kretprobe_event; + tp->event.funcs = &kretprobe_funcs; INIT_LIST_HEAD(&call->class->fields); call->class->raw_init = probe_event_raw_init; call->class->define_fields = kretprobe_event_define_fields; } else { INIT_LIST_HEAD(&call->class->fields); - tp->event.trace = print_kprobe_event; + tp->event.funcs = &kprobe_funcs; call->class->raw_init = probe_event_raw_init; call->class->define_fields = kprobe_event_define_fields; } diff --git a/kernel/trace/trace_output.c b/kernel/trace/trace_output.c index 2404c129a8c9..fc9d4dbb089e 100644 --- a/kernel/trace/trace_output.c +++ b/kernel/trace/trace_output.c @@ -726,6 +726,9 @@ int register_ftrace_event(struct trace_event *event) if (WARN_ON(!event)) goto out; + if (WARN_ON(!event->funcs)) + goto out; + INIT_LIST_HEAD(&event->list); if (!event->type) { @@ -758,14 +761,14 @@ int register_ftrace_event(struct trace_event *event) goto out; } - if (event->trace == NULL) - event->trace = trace_nop_print; - if (event->raw == NULL) - event->raw = trace_nop_print; - if (event->hex == NULL) - event->hex = trace_nop_print; - if (event->binary == NULL) - event->binary = trace_nop_print; + if (event->funcs->trace == NULL) + event->funcs->trace = trace_nop_print; + if (event->funcs->raw == NULL) + event->funcs->raw = trace_nop_print; + if (event->funcs->hex == NULL) + event->funcs->hex = trace_nop_print; + if (event->funcs->binary == NULL) + event->funcs->binary = trace_nop_print; key = event->type & (EVENT_HASHSIZE - 1); @@ -807,13 +810,15 @@ EXPORT_SYMBOL_GPL(unregister_ftrace_event); * Standard events */ -enum print_line_t trace_nop_print(struct trace_iterator *iter, int flags) +enum print_line_t trace_nop_print(struct trace_iterator *iter, int flags, + struct trace_event *event) { return TRACE_TYPE_HANDLED; } /* TRACE_FN */ -static enum print_line_t trace_fn_trace(struct trace_iterator *iter, int flags) +static enum print_line_t trace_fn_trace(struct trace_iterator *iter, int flags, + struct trace_event *event) { struct ftrace_entry *field; struct trace_seq *s = &iter->seq; @@ -840,7 +845,8 @@ static enum print_line_t trace_fn_trace(struct trace_iterator *iter, int flags) return TRACE_TYPE_PARTIAL_LINE; } -static enum print_line_t trace_fn_raw(struct trace_iterator *iter, int flags) +static enum print_line_t trace_fn_raw(struct trace_iterator *iter, int flags, + struct trace_event *event) { struct ftrace_entry *field; @@ -854,7 +860,8 @@ static enum print_line_t trace_fn_raw(struct trace_iterator *iter, int flags) return TRACE_TYPE_HANDLED; } -static enum print_line_t trace_fn_hex(struct trace_iterator *iter, int flags) +static enum print_line_t trace_fn_hex(struct trace_iterator *iter, int flags, + struct trace_event *event) { struct ftrace_entry *field; struct trace_seq *s = &iter->seq; @@ -867,7 +874,8 @@ static enum print_line_t trace_fn_hex(struct trace_iterator *iter, int flags) return TRACE_TYPE_HANDLED; } -static enum print_line_t trace_fn_bin(struct trace_iterator *iter, int flags) +static enum print_line_t trace_fn_bin(struct trace_iterator *iter, int flags, + struct trace_event *event) { struct ftrace_entry *field; struct trace_seq *s = &iter->seq; @@ -880,14 +888,18 @@ static enum print_line_t trace_fn_bin(struct trace_iterator *iter, int flags) return TRACE_TYPE_HANDLED; } -static struct trace_event trace_fn_event = { - .type = TRACE_FN, +static struct trace_event_functions trace_fn_funcs = { .trace = trace_fn_trace, .raw = trace_fn_raw, .hex = trace_fn_hex, .binary = trace_fn_bin, }; +static struct trace_event trace_fn_event = { + .type = TRACE_FN, + .funcs = &trace_fn_funcs, +}; + /* TRACE_CTX an TRACE_WAKE */ static enum print_line_t trace_ctxwake_print(struct trace_iterator *iter, char *delim) @@ -916,13 +928,14 @@ static enum print_line_t trace_ctxwake_print(struct trace_iterator *iter, return TRACE_TYPE_HANDLED; } -static enum print_line_t trace_ctx_print(struct trace_iterator *iter, int flags) +static enum print_line_t trace_ctx_print(struct trace_iterator *iter, int flags, + struct trace_event *event) { return trace_ctxwake_print(iter, "==>"); } static enum print_line_t trace_wake_print(struct trace_iterator *iter, - int flags) + int flags, struct trace_event *event) { return trace_ctxwake_print(iter, " +"); } @@ -950,12 +963,14 @@ static int trace_ctxwake_raw(struct trace_iterator *iter, char S) return TRACE_TYPE_HANDLED; } -static enum print_line_t trace_ctx_raw(struct trace_iterator *iter, int flags) +static enum print_line_t trace_ctx_raw(struct trace_iterator *iter, int flags, + struct trace_event *event) { return trace_ctxwake_raw(iter, 0); } -static enum print_line_t trace_wake_raw(struct trace_iterator *iter, int flags) +static enum print_line_t trace_wake_raw(struct trace_iterator *iter, int flags, + struct trace_event *event) { return trace_ctxwake_raw(iter, '+'); } @@ -984,18 +999,20 @@ static int trace_ctxwake_hex(struct trace_iterator *iter, char S) return TRACE_TYPE_HANDLED; } -static enum print_line_t trace_ctx_hex(struct trace_iterator *iter, int flags) +static enum print_line_t trace_ctx_hex(struct trace_iterator *iter, int flags, + struct trace_event *event) { return trace_ctxwake_hex(iter, 0); } -static enum print_line_t trace_wake_hex(struct trace_iterator *iter, int flags) +static enum print_line_t trace_wake_hex(struct trace_iterator *iter, int flags, + struct trace_event *event) { return trace_ctxwake_hex(iter, '+'); } static enum print_line_t trace_ctxwake_bin(struct trace_iterator *iter, - int flags) + int flags, struct trace_event *event) { struct ctx_switch_entry *field; struct trace_seq *s = &iter->seq; @@ -1012,25 +1029,33 @@ static enum print_line_t trace_ctxwake_bin(struct trace_iterator *iter, return TRACE_TYPE_HANDLED; } -static struct trace_event trace_ctx_event = { - .type = TRACE_CTX, +static struct trace_event_functions trace_ctx_funcs = { .trace = trace_ctx_print, .raw = trace_ctx_raw, .hex = trace_ctx_hex, .binary = trace_ctxwake_bin, }; -static struct trace_event trace_wake_event = { - .type = TRACE_WAKE, +static struct trace_event trace_ctx_event = { + .type = TRACE_CTX, + .funcs = &trace_ctx_funcs, +}; + +static struct trace_event_functions trace_wake_funcs = { .trace = trace_wake_print, .raw = trace_wake_raw, .hex = trace_wake_hex, .binary = trace_ctxwake_bin, }; +static struct trace_event trace_wake_event = { + .type = TRACE_WAKE, + .funcs = &trace_wake_funcs, +}; + /* TRACE_SPECIAL */ static enum print_line_t trace_special_print(struct trace_iterator *iter, - int flags) + int flags, struct trace_event *event) { struct special_entry *field; @@ -1046,7 +1071,7 @@ static enum print_line_t trace_special_print(struct trace_iterator *iter, } static enum print_line_t trace_special_hex(struct trace_iterator *iter, - int flags) + int flags, struct trace_event *event) { struct special_entry *field; struct trace_seq *s = &iter->seq; @@ -1061,7 +1086,7 @@ static enum print_line_t trace_special_hex(struct trace_iterator *iter, } static enum print_line_t trace_special_bin(struct trace_iterator *iter, - int flags) + int flags, struct trace_event *event) { struct special_entry *field; struct trace_seq *s = &iter->seq; @@ -1075,18 +1100,22 @@ static enum print_line_t trace_special_bin(struct trace_iterator *iter, return TRACE_TYPE_HANDLED; } -static struct trace_event trace_special_event = { - .type = TRACE_SPECIAL, +static struct trace_event_functions trace_special_funcs = { .trace = trace_special_print, .raw = trace_special_print, .hex = trace_special_hex, .binary = trace_special_bin, }; +static struct trace_event trace_special_event = { + .type = TRACE_SPECIAL, + .funcs = &trace_special_funcs, +}; + /* TRACE_STACK */ static enum print_line_t trace_stack_print(struct trace_iterator *iter, - int flags) + int flags, struct trace_event *event) { struct stack_entry *field; struct trace_seq *s = &iter->seq; @@ -1114,17 +1143,21 @@ static enum print_line_t trace_stack_print(struct trace_iterator *iter, return TRACE_TYPE_PARTIAL_LINE; } -static struct trace_event trace_stack_event = { - .type = TRACE_STACK, +static struct trace_event_functions trace_stack_funcs = { .trace = trace_stack_print, .raw = trace_special_print, .hex = trace_special_hex, .binary = trace_special_bin, }; +static struct trace_event trace_stack_event = { + .type = TRACE_STACK, + .funcs = &trace_stack_funcs, +}; + /* TRACE_USER_STACK */ static enum print_line_t trace_user_stack_print(struct trace_iterator *iter, - int flags) + int flags, struct trace_event *event) { struct userstack_entry *field; struct trace_seq *s = &iter->seq; @@ -1143,17 +1176,22 @@ static enum print_line_t trace_user_stack_print(struct trace_iterator *iter, return TRACE_TYPE_PARTIAL_LINE; } -static struct trace_event trace_user_stack_event = { - .type = TRACE_USER_STACK, +static struct trace_event_functions trace_user_stack_funcs = { .trace = trace_user_stack_print, .raw = trace_special_print, .hex = trace_special_hex, .binary = trace_special_bin, }; +static struct trace_event trace_user_stack_event = { + .type = TRACE_USER_STACK, + .funcs = &trace_user_stack_funcs, +}; + /* TRACE_BPRINT */ static enum print_line_t -trace_bprint_print(struct trace_iterator *iter, int flags) +trace_bprint_print(struct trace_iterator *iter, int flags, + struct trace_event *event) { struct trace_entry *entry = iter->ent; struct trace_seq *s = &iter->seq; @@ -1178,7 +1216,8 @@ trace_bprint_print(struct trace_iterator *iter, int flags) static enum print_line_t -trace_bprint_raw(struct trace_iterator *iter, int flags) +trace_bprint_raw(struct trace_iterator *iter, int flags, + struct trace_event *event) { struct bprint_entry *field; struct trace_seq *s = &iter->seq; @@ -1197,16 +1236,19 @@ trace_bprint_raw(struct trace_iterator *iter, int flags) return TRACE_TYPE_PARTIAL_LINE; } +static struct trace_event_functions trace_bprint_funcs = { + .trace = trace_bprint_print, + .raw = trace_bprint_raw, +}; static struct trace_event trace_bprint_event = { .type = TRACE_BPRINT, - .trace = trace_bprint_print, - .raw = trace_bprint_raw, + .funcs = &trace_bprint_funcs, }; /* TRACE_PRINT */ static enum print_line_t trace_print_print(struct trace_iterator *iter, - int flags) + int flags, struct trace_event *event) { struct print_entry *field; struct trace_seq *s = &iter->seq; @@ -1225,7 +1267,8 @@ static enum print_line_t trace_print_print(struct trace_iterator *iter, return TRACE_TYPE_PARTIAL_LINE; } -static enum print_line_t trace_print_raw(struct trace_iterator *iter, int flags) +static enum print_line_t trace_print_raw(struct trace_iterator *iter, int flags, + struct trace_event *event) { struct print_entry *field; @@ -1240,12 +1283,16 @@ static enum print_line_t trace_print_raw(struct trace_iterator *iter, int flags) return TRACE_TYPE_PARTIAL_LINE; } -static struct trace_event trace_print_event = { - .type = TRACE_PRINT, +static struct trace_event_functions trace_print_funcs = { .trace = trace_print_print, .raw = trace_print_raw, }; +static struct trace_event trace_print_event = { + .type = TRACE_PRINT, + .funcs = &trace_print_funcs, +}; + static struct trace_event *events[] __initdata = { &trace_fn_event, diff --git a/kernel/trace/trace_output.h b/kernel/trace/trace_output.h index 9d91c72ba38b..c038eba0492b 100644 --- a/kernel/trace/trace_output.h +++ b/kernel/trace/trace_output.h @@ -25,7 +25,7 @@ extern void trace_event_read_unlock(void); extern struct trace_event *ftrace_find_event(int type); extern enum print_line_t trace_nop_print(struct trace_iterator *iter, - int flags); + int flags, struct trace_event *event); extern int trace_print_lat_fmt(struct trace_seq *s, struct trace_entry *entry); diff --git a/kernel/trace/trace_syscalls.c b/kernel/trace/trace_syscalls.c index 1c1b9bf3ccc6..3751c81998cb 100644 --- a/kernel/trace/trace_syscalls.c +++ b/kernel/trace/trace_syscalls.c @@ -93,7 +93,8 @@ static struct syscall_metadata *syscall_nr_to_meta(int nr) } enum print_line_t -print_syscall_enter(struct trace_iterator *iter, int flags) +print_syscall_enter(struct trace_iterator *iter, int flags, + struct trace_event *event) { struct trace_seq *s = &iter->seq; struct trace_entry *ent = iter->ent; @@ -145,7 +146,8 @@ end: } enum print_line_t -print_syscall_exit(struct trace_iterator *iter, int flags) +print_syscall_exit(struct trace_iterator *iter, int flags, + struct trace_event *event) { struct trace_seq *s = &iter->seq; struct trace_entry *ent = iter->ent; -- cgit v1.2.3 From 80decc70afc57c87eee9d6b836aec2ecacba3457 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Fri, 23 Apr 2010 10:00:22 -0400 Subject: tracing: Move print functions into event class Currently, every event has its own trace_event structure. This is fine since the structure is needed anyway. But the print function structure (trace_event_functions) is now separate. Since the output of the trace event is done by the class (with the exception of events defined by DEFINE_EVENT_PRINT), it makes sense to have the class define the print functions that all events in the class can use. This makes a bigger deal with the syscall events since all syscall events use the same class. The savings here is another 30K. text data bss dec hex filename 4913961 1088356 861512 6863829 68bbd5 vmlinux.orig 4900382 1048964 861512 6810858 67ecea vmlinux.init 4900446 1049028 861512 6810986 67ed6a vmlinux.preprint 4895024 1023812 861512 6780348 6775bc vmlinux.print To accomplish this, and to let the class know what event is being printed, the event structure is embedded in the ftrace_event_call structure. This should not be an issues since the event structure was created for each event anyway. Acked-by: Mathieu Desnoyers Acked-by: Masami Hiramatsu Acked-by: Frederic Weisbecker Signed-off-by: Steven Rostedt --- kernel/trace/trace_events.c | 6 +++--- kernel/trace/trace_kprobe.c | 14 ++++++-------- kernel/trace/trace_syscalls.c | 8 ++++++++ 3 files changed, 17 insertions(+), 11 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_events.c b/kernel/trace/trace_events.c index 8e94255a45e7..aafe5bff8f59 100644 --- a/kernel/trace/trace_events.c +++ b/kernel/trace/trace_events.c @@ -122,7 +122,7 @@ int trace_event_raw_init(struct ftrace_event_call *call) { int id; - id = register_ftrace_event(call->event); + id = register_ftrace_event(&call->event); if (!id) return -ENODEV; call->id = id; @@ -1073,8 +1073,8 @@ static void remove_subsystem_dir(const char *name) static void __trace_remove_event_call(struct ftrace_event_call *call) { ftrace_event_enable_disable(call, 0); - if (call->event) - __unregister_ftrace_event(call->event); + if (call->event.funcs) + __unregister_ftrace_event(&call->event); debugfs_remove_recursive(call->dir); list_del(&call->list); trace_destroy_fields(call); diff --git a/kernel/trace/trace_kprobe.c b/kernel/trace/trace_kprobe.c index b989ae229a20..d8061c3e02c9 100644 --- a/kernel/trace/trace_kprobe.c +++ b/kernel/trace/trace_kprobe.c @@ -204,7 +204,6 @@ struct trace_probe { const char *symbol; /* symbol name */ struct ftrace_event_class class; struct ftrace_event_call call; - struct trace_event event; unsigned int nr_args; struct probe_arg args[]; }; @@ -1020,7 +1019,7 @@ print_kprobe_event(struct trace_iterator *iter, int flags, int i; field = (struct kprobe_trace_entry *)iter->ent; - tp = container_of(event, struct trace_probe, event); + tp = container_of(event, struct trace_probe, call.event); if (!trace_seq_printf(s, "%s: (", tp->call.name)) goto partial; @@ -1054,7 +1053,7 @@ print_kretprobe_event(struct trace_iterator *iter, int flags, int i; field = (struct kretprobe_trace_entry *)iter->ent; - tp = container_of(event, struct trace_probe, event); + tp = container_of(event, struct trace_probe, call.event); if (!trace_seq_printf(s, "%s: (", tp->call.name)) goto partial; @@ -1364,20 +1363,19 @@ static int register_probe_event(struct trace_probe *tp) /* Initialize ftrace_event_call */ if (probe_is_return(tp)) { - tp->event.funcs = &kretprobe_funcs; INIT_LIST_HEAD(&call->class->fields); + call->event.funcs = &kretprobe_funcs; call->class->raw_init = probe_event_raw_init; call->class->define_fields = kretprobe_event_define_fields; } else { INIT_LIST_HEAD(&call->class->fields); - tp->event.funcs = &kprobe_funcs; + call->event.funcs = &kprobe_funcs; call->class->raw_init = probe_event_raw_init; call->class->define_fields = kprobe_event_define_fields; } if (set_print_fmt(tp) < 0) return -ENOMEM; - call->event = &tp->event; - call->id = register_ftrace_event(&tp->event); + call->id = register_ftrace_event(&call->event); if (!call->id) { kfree(call->print_fmt); return -ENODEV; @@ -1389,7 +1387,7 @@ static int register_probe_event(struct trace_probe *tp) if (ret) { pr_info("Failed to register kprobe event: %s\n", call->name); kfree(call->print_fmt); - unregister_ftrace_event(&tp->event); + unregister_ftrace_event(&call->event); } return ret; } diff --git a/kernel/trace/trace_syscalls.c b/kernel/trace/trace_syscalls.c index 3751c81998cb..7c7cfe95a853 100644 --- a/kernel/trace/trace_syscalls.c +++ b/kernel/trace/trace_syscalls.c @@ -39,6 +39,14 @@ syscall_get_exit_fields(struct ftrace_event_call *call) return &entry->exit_fields; } +struct trace_event_functions enter_syscall_print_funcs = { + .trace = print_syscall_enter, +}; + +struct trace_event_functions exit_syscall_print_funcs = { + .trace = print_syscall_exit, +}; + struct ftrace_event_class event_class_syscall_enter = { .system = "syscalls", .reg = syscall_enter_register, -- cgit v1.2.3 From 32c0edaeaad74a7883e736ae0f3798784cfc2a80 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Fri, 23 Apr 2010 10:38:03 -0400 Subject: tracing: Remove duplicate id information in event structure Now that the trace_event structure is embedded in the ftrace_event_call structure, there is no need for the ftrace_event_call id field. The id field is the same as the trace_event type field. Removing the id and re-arranging the structure brings down the tracepoint footprint by another 5K. text data bss dec hex filename 4913961 1088356 861512 6863829 68bbd5 vmlinux.orig 4895024 1023812 861512 6780348 6775bc vmlinux.print 4894944 1018052 861512 6774508 675eec vmlinux.id Acked-by: Mathieu Desnoyers Acked-by: Masami Hiramatsu Acked-by: Frederic Weisbecker Signed-off-by: Steven Rostedt --- kernel/trace/trace_event_perf.c | 4 ++-- kernel/trace/trace_events.c | 7 +++---- kernel/trace/trace_events_filter.c | 2 +- kernel/trace/trace_export.c | 4 ++-- kernel/trace/trace_kprobe.c | 18 ++++++++++-------- kernel/trace/trace_syscalls.c | 14 ++++++++------ 6 files changed, 26 insertions(+), 23 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_event_perf.c b/kernel/trace/trace_event_perf.c index 196fe9d26773..0a47e8d6b491 100644 --- a/kernel/trace/trace_event_perf.c +++ b/kernel/trace/trace_event_perf.c @@ -80,7 +80,7 @@ int perf_trace_enable(int event_id) mutex_lock(&event_mutex); list_for_each_entry(event, &ftrace_events, list) { - if (event->id == event_id && + if (event->event.type == event_id && event->class && event->class->perf_probe && try_module_get(event->mod)) { ret = perf_trace_event_enable(event); @@ -128,7 +128,7 @@ void perf_trace_disable(int event_id) mutex_lock(&event_mutex); list_for_each_entry(event, &ftrace_events, list) { - if (event->id == event_id) { + if (event->event.type == event_id) { perf_trace_event_disable(event); module_put(event->mod); break; diff --git a/kernel/trace/trace_events.c b/kernel/trace/trace_events.c index aafe5bff8f59..8daaca5475b5 100644 --- a/kernel/trace/trace_events.c +++ b/kernel/trace/trace_events.c @@ -125,7 +125,6 @@ int trace_event_raw_init(struct ftrace_event_call *call) id = register_ftrace_event(&call->event); if (!id) return -ENODEV; - call->id = id; return 0; } @@ -567,7 +566,7 @@ event_format_read(struct file *filp, char __user *ubuf, size_t cnt, trace_seq_init(s); trace_seq_printf(s, "name: %s\n", call->name); - trace_seq_printf(s, "ID: %d\n", call->id); + trace_seq_printf(s, "ID: %d\n", call->event.type); trace_seq_printf(s, "format:\n"); head = trace_get_fields(call); @@ -641,7 +640,7 @@ event_id_read(struct file *filp, char __user *ubuf, size_t cnt, loff_t *ppos) return -ENOMEM; trace_seq_init(s); - trace_seq_printf(s, "%d\n", call->id); + trace_seq_printf(s, "%d\n", call->event.type); r = simple_read_from_buffer(ubuf, cnt, ppos, s->buffer, s->len); @@ -969,7 +968,7 @@ event_create_dir(struct ftrace_event_call *call, struct dentry *d_events, enable); #ifdef CONFIG_PERF_EVENTS - if (call->id && (call->class->perf_probe || call->class->reg)) + if (call->event.type && (call->class->perf_probe || call->class->reg)) trace_create_file("id", 0444, call->dir, call, id); #endif diff --git a/kernel/trace/trace_events_filter.c b/kernel/trace/trace_events_filter.c index 961f99b74bdd..2702d6bbf1ab 100644 --- a/kernel/trace/trace_events_filter.c +++ b/kernel/trace/trace_events_filter.c @@ -1395,7 +1395,7 @@ int ftrace_profile_set_filter(struct perf_event *event, int event_id, mutex_lock(&event_mutex); list_for_each_entry(call, &ftrace_events, list) { - if (call->id == event_id) + if (call->event.type == event_id) break; } diff --git a/kernel/trace/trace_export.c b/kernel/trace/trace_export.c index e878d06c0ac0..8536e2a65969 100644 --- a/kernel/trace/trace_export.c +++ b/kernel/trace/trace_export.c @@ -153,7 +153,7 @@ static int ftrace_raw_init_event(struct ftrace_event_call *call) #define F_printk(fmt, args...) #fmt ", " __stringify(args) #undef FTRACE_ENTRY -#define FTRACE_ENTRY(call, struct_name, type, tstruct, print) \ +#define FTRACE_ENTRY(call, struct_name, etype, tstruct, print) \ \ struct ftrace_event_class event_class_ftrace_##call = { \ .system = __stringify(TRACE_SYSTEM), \ @@ -165,7 +165,7 @@ struct ftrace_event_call __used \ __attribute__((__aligned__(4))) \ __attribute__((section("_ftrace_events"))) event_##call = { \ .name = #call, \ - .id = type, \ + .event.type = etype, \ .class = &event_class_ftrace_##call, \ .print_fmt = print, \ }; \ diff --git a/kernel/trace/trace_kprobe.c b/kernel/trace/trace_kprobe.c index d8061c3e02c9..934078bca3f9 100644 --- a/kernel/trace/trace_kprobe.c +++ b/kernel/trace/trace_kprobe.c @@ -960,8 +960,8 @@ static __kprobes void kprobe_trace_func(struct kprobe *kp, struct pt_regs *regs) size = SIZEOF_KPROBE_TRACE_ENTRY(tp->nr_args); - event = trace_current_buffer_lock_reserve(&buffer, call->id, size, - irq_flags, pc); + event = trace_current_buffer_lock_reserve(&buffer, call->event.type, + size, irq_flags, pc); if (!event) return; @@ -992,8 +992,8 @@ static __kprobes void kretprobe_trace_func(struct kretprobe_instance *ri, size = SIZEOF_KRETPROBE_TRACE_ENTRY(tp->nr_args); - event = trace_current_buffer_lock_reserve(&buffer, call->id, size, - irq_flags, pc); + event = trace_current_buffer_lock_reserve(&buffer, call->event.type, + size, irq_flags, pc); if (!event) return; @@ -1228,7 +1228,8 @@ static __kprobes void kprobe_perf_func(struct kprobe *kp, "profile buffer not large enough")) return; - entry = perf_trace_buf_prepare(size, call->id, &rctx, &irq_flags); + entry = perf_trace_buf_prepare(size, call->event.type, + &rctx, &irq_flags); if (!entry) return; @@ -1258,7 +1259,8 @@ static __kprobes void kretprobe_perf_func(struct kretprobe_instance *ri, "profile buffer not large enough")) return; - entry = perf_trace_buf_prepare(size, call->id, &rctx, &irq_flags); + entry = perf_trace_buf_prepare(size, call->event.type, + &rctx, &irq_flags); if (!entry) return; @@ -1375,8 +1377,8 @@ static int register_probe_event(struct trace_probe *tp) } if (set_print_fmt(tp) < 0) return -ENOMEM; - call->id = register_ftrace_event(&call->event); - if (!call->id) { + ret = register_ftrace_event(&call->event); + if (!ret) { kfree(call->print_fmt); return -ENODEV; } diff --git a/kernel/trace/trace_syscalls.c b/kernel/trace/trace_syscalls.c index 7c7cfe95a853..9d358301ae3e 100644 --- a/kernel/trace/trace_syscalls.c +++ b/kernel/trace/trace_syscalls.c @@ -117,7 +117,7 @@ print_syscall_enter(struct trace_iterator *iter, int flags, if (!entry) goto end; - if (entry->enter_event->id != ent->type) { + if (entry->enter_event->event.type != ent->type) { WARN_ON_ONCE(1); goto end; } @@ -173,7 +173,7 @@ print_syscall_exit(struct trace_iterator *iter, int flags, return TRACE_TYPE_HANDLED; } - if (entry->exit_event->id != ent->type) { + if (entry->exit_event->event.type != ent->type) { WARN_ON_ONCE(1); return TRACE_TYPE_UNHANDLED; } @@ -315,7 +315,7 @@ void ftrace_syscall_enter(void *ignore, struct pt_regs *regs, long id) size = sizeof(*entry) + sizeof(unsigned long) * sys_data->nb_args; event = trace_current_buffer_lock_reserve(&buffer, - sys_data->enter_event->id, size, 0, 0); + sys_data->enter_event->event.type, size, 0, 0); if (!event) return; @@ -347,7 +347,7 @@ void ftrace_syscall_exit(void *ignore, struct pt_regs *regs, long ret) return; event = trace_current_buffer_lock_reserve(&buffer, - sys_data->exit_event->id, sizeof(*entry), 0, 0); + sys_data->exit_event->event.type, sizeof(*entry), 0, 0); if (!event) return; @@ -511,7 +511,8 @@ static void perf_syscall_enter(void *ignore, struct pt_regs *regs, long id) return; rec = (struct syscall_trace_enter *)perf_trace_buf_prepare(size, - sys_data->enter_event->id, &rctx, &flags); + sys_data->enter_event->event.type, + &rctx, &flags); if (!rec) return; @@ -586,7 +587,8 @@ static void perf_syscall_exit(void *ignore, struct pt_regs *regs, long ret) return; rec = (struct syscall_trace_exit *)perf_trace_buf_prepare(size, - sys_data->exit_event->id, &rctx, &flags); + sys_data->exit_event->event.type, + &rctx, &flags); if (!rec) return; -- cgit v1.2.3 From 553552ce1796c32cf4e3d4f45cd5b537de91dd1d Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Fri, 23 Apr 2010 11:12:36 -0400 Subject: tracing: Combine event filter_active and enable into single flags field The filter_active and enable both use an int (4 bytes each) to set a single flag. We can save 4 bytes per event by combining the two into a single integer. text data bss dec hex filename 4913961 1088356 861512 6863829 68bbd5 vmlinux.orig 4894944 1018052 861512 6774508 675eec vmlinux.id 4894871 1012292 861512 6768675 674823 vmlinux.flags This gives us another 5K in savings. The modification of both the enable and filter fields are done under the event_mutex, so it is still safe to combine the two. Note: Although Mathieu gave his Acked-by, he would like it documented that the reads of flags are not protected by the mutex. The way the code works, these reads will not break anything, but will have a residual effect. Since this behavior is the same even before this patch, describing this situation is left to another patch, as this patch does not change the behavior, but just brought it to Mathieu's attention. v2: Updated the event trace self test to for this change. Acked-by: Mathieu Desnoyers Acked-by: Masami Hiramatsu Acked-by: Frederic Weisbecker Cc: Tom Zanussi Signed-off-by: Steven Rostedt --- kernel/trace/trace.h | 2 +- kernel/trace/trace_events.c | 16 ++++++++-------- kernel/trace/trace_events_filter.c | 10 +++++----- kernel/trace/trace_kprobe.c | 2 +- 4 files changed, 15 insertions(+), 15 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace.h b/kernel/trace/trace.h index c88c563a59a5..63562595f2b2 100644 --- a/kernel/trace/trace.h +++ b/kernel/trace/trace.h @@ -802,7 +802,7 @@ filter_check_discard(struct ftrace_event_call *call, void *rec, struct ring_buffer *buffer, struct ring_buffer_event *event) { - if (unlikely(call->filter_active) && + if (unlikely(call->flags & TRACE_EVENT_FL_FILTERED) && !filter_match_preds(call->filter, rec)) { ring_buffer_discard_commit(buffer, event); return 1; diff --git a/kernel/trace/trace_events.c b/kernel/trace/trace_events.c index 8daaca5475b5..53cffc0b0801 100644 --- a/kernel/trace/trace_events.c +++ b/kernel/trace/trace_events.c @@ -137,8 +137,8 @@ static int ftrace_event_enable_disable(struct ftrace_event_call *call, switch (enable) { case 0: - if (call->enabled) { - call->enabled = 0; + if (call->flags & TRACE_EVENT_FL_ENABLED) { + call->flags &= ~TRACE_EVENT_FL_ENABLED; tracing_stop_cmdline_record(); if (call->class->reg) call->class->reg(call, TRACE_REG_UNREGISTER); @@ -149,7 +149,7 @@ static int ftrace_event_enable_disable(struct ftrace_event_call *call, } break; case 1: - if (!call->enabled) { + if (!(call->flags & TRACE_EVENT_FL_ENABLED)) { tracing_start_cmdline_record(); if (call->class->reg) ret = call->class->reg(call, TRACE_REG_REGISTER); @@ -163,7 +163,7 @@ static int ftrace_event_enable_disable(struct ftrace_event_call *call, "%s\n", call->name); break; } - call->enabled = 1; + call->flags |= TRACE_EVENT_FL_ENABLED; } break; } @@ -352,7 +352,7 @@ s_next(struct seq_file *m, void *v, loff_t *pos) (*pos)++; list_for_each_entry_continue(call, &ftrace_events, list) { - if (call->enabled) + if (call->flags & TRACE_EVENT_FL_ENABLED) return call; } @@ -411,7 +411,7 @@ event_enable_read(struct file *filp, char __user *ubuf, size_t cnt, struct ftrace_event_call *call = filp->private_data; char *buf; - if (call->enabled) + if (call->flags & TRACE_EVENT_FL_ENABLED) buf = "1\n"; else buf = "0\n"; @@ -486,7 +486,7 @@ system_enable_read(struct file *filp, char __user *ubuf, size_t cnt, * or if all events or cleared, or if we have * a mixture. */ - set |= (1 << !!call->enabled); + set |= (1 << !!(call->flags & TRACE_EVENT_FL_ENABLED)); /* * If we have a mixture, no need to look further. @@ -1447,7 +1447,7 @@ static __init void event_trace_self_tests(void) * If an event is already enabled, someone is using * it and the self test should not be on. */ - if (call->enabled) { + if (call->flags & TRACE_EVENT_FL_ENABLED) { pr_warning("Enabled event during self test!\n"); WARN_ON_ONCE(1); continue; diff --git a/kernel/trace/trace_events_filter.c b/kernel/trace/trace_events_filter.c index 2702d6bbf1ab..239ea5d77d68 100644 --- a/kernel/trace/trace_events_filter.c +++ b/kernel/trace/trace_events_filter.c @@ -547,7 +547,7 @@ static void filter_disable_preds(struct ftrace_event_call *call) struct event_filter *filter = call->filter; int i; - call->filter_active = 0; + call->flags &= ~TRACE_EVENT_FL_FILTERED; filter->n_preds = 0; for (i = 0; i < MAX_FILTER_PRED; i++) @@ -574,7 +574,7 @@ void destroy_preds(struct ftrace_event_call *call) { __free_preds(call->filter); call->filter = NULL; - call->filter_active = 0; + call->flags &= ~TRACE_EVENT_FL_FILTERED; } static struct event_filter *__alloc_preds(void) @@ -613,7 +613,7 @@ static int init_preds(struct ftrace_event_call *call) if (call->filter) return 0; - call->filter_active = 0; + call->flags &= ~TRACE_EVENT_FL_FILTERED; call->filter = __alloc_preds(); if (IS_ERR(call->filter)) return PTR_ERR(call->filter); @@ -1268,7 +1268,7 @@ static int replace_system_preds(struct event_subsystem *system, if (err) filter_disable_preds(call); else { - call->filter_active = 1; + call->flags |= TRACE_EVENT_FL_FILTERED; replace_filter_string(filter, filter_string); } fail = false; @@ -1317,7 +1317,7 @@ int apply_event_filter(struct ftrace_event_call *call, char *filter_string) if (err) append_filter_err(ps, call->filter); else - call->filter_active = 1; + call->flags |= TRACE_EVENT_FL_FILTERED; out: filter_opstack_clear(ps); postfix_clear(ps); diff --git a/kernel/trace/trace_kprobe.c b/kernel/trace/trace_kprobe.c index 934078bca3f9..0e3ded64cdb7 100644 --- a/kernel/trace/trace_kprobe.c +++ b/kernel/trace/trace_kprobe.c @@ -1382,7 +1382,7 @@ static int register_probe_event(struct trace_probe *tp) kfree(call->print_fmt); return -ENODEV; } - call->enabled = 0; + call->flags = 0; call->class->reg = kprobe_register; call->data = tp; ret = trace_add_event_call(call); -- cgit v1.2.3 From e1f7992e015ae1373d66c8068d0a45e4111a0aed Mon Sep 17 00:00:00 2001 From: Li Zefan Date: Mon, 10 May 2010 11:23:00 +0800 Subject: tracing: Fix function declarations if !CONFIG_STACKTRACE ftrace_trace_stack() and frace_trace_userstacke() take a struct ring_buffer argument, not struct trace_array. Commit e77405ad("tracing: pass around ring buffer instead of tracer") made this change. Signed-off-by: Li Zefan LKML-Reference: <4BE77C14.5010806@cn.fujitsu.com> Signed-off-by: Steven Rostedt --- kernel/trace/trace.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace.h b/kernel/trace/trace.h index 63562595f2b2..40cd1718fb1b 100644 --- a/kernel/trace/trace.h +++ b/kernel/trace/trace.h @@ -419,12 +419,12 @@ void ftrace_trace_userstack(struct ring_buffer *buffer, unsigned long flags, void __trace_stack(struct trace_array *tr, unsigned long flags, int skip, int pc); #else -static inline void ftrace_trace_stack(struct trace_array *tr, +static inline void ftrace_trace_stack(struct ring_buffer *buffer, unsigned long flags, int skip, int pc) { } -static inline void ftrace_trace_userstack(struct trace_array *tr, +static inline void ftrace_trace_userstack(struct ring_buffer *buffer, unsigned long flags, int pc) { } -- cgit v1.2.3 From 4f41c013f553957765902fb01475972f0af3e8e7 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Tue, 18 May 2010 18:08:32 +0200 Subject: perf/ftrace: Optimize perf/tracepoint interaction for single events When we've got but a single event per tracepoint there is no reason to try and multiplex it so don't. Signed-off-by: Peter Zijlstra Tested-by: Ingo Molnar Cc: Steven Rostedt Cc: Frederic Weisbecker Cc: Mike Galbraith Cc: Paul Mackerras Cc: Arnaldo Carvalho de Melo LKML-Reference: Signed-off-by: Ingo Molnar --- kernel/trace/trace_event_perf.c | 11 +++++++---- kernel/trace/trace_kprobe.c | 4 ++-- kernel/trace/trace_syscalls.c | 6 ++++-- 3 files changed, 13 insertions(+), 8 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_event_perf.c b/kernel/trace/trace_event_perf.c index 0565bb42566f..89b780a7c522 100644 --- a/kernel/trace/trace_event_perf.c +++ b/kernel/trace/trace_event_perf.c @@ -27,13 +27,15 @@ typedef typeof(unsigned long [PERF_MAX_TRACE_SIZE / sizeof(unsigned long)]) /* Count the events in use (per event id, not per instance) */ static int total_ref_count; -static int perf_trace_event_enable(struct ftrace_event_call *event) +static int perf_trace_event_enable(struct ftrace_event_call *event, void *data) { char *buf; int ret = -ENOMEM; - if (event->perf_refcount++ > 0) + if (event->perf_refcount++ > 0) { + event->perf_data = NULL; return 0; + } if (!total_ref_count) { buf = (char *)alloc_percpu(perf_trace_t); @@ -51,6 +53,7 @@ static int perf_trace_event_enable(struct ftrace_event_call *event) ret = event->perf_event_enable(event); if (!ret) { + event->perf_data = data; total_ref_count++; return 0; } @@ -68,7 +71,7 @@ fail_buf: return ret; } -int perf_trace_enable(int event_id) +int perf_trace_enable(int event_id, void *data) { struct ftrace_event_call *event; int ret = -EINVAL; @@ -77,7 +80,7 @@ int perf_trace_enable(int event_id) list_for_each_entry(event, &ftrace_events, list) { if (event->id == event_id && event->perf_event_enable && try_module_get(event->mod)) { - ret = perf_trace_event_enable(event); + ret = perf_trace_event_enable(event, data); break; } } diff --git a/kernel/trace/trace_kprobe.c b/kernel/trace/trace_kprobe.c index a7514326052b..2d7bf4146be8 100644 --- a/kernel/trace/trace_kprobe.c +++ b/kernel/trace/trace_kprobe.c @@ -1362,7 +1362,7 @@ static __kprobes void kprobe_perf_func(struct kprobe *kp, for (i = 0; i < tp->nr_args; i++) call_fetch(&tp->args[i].fetch, regs, data + tp->args[i].offset); - perf_trace_buf_submit(entry, size, rctx, entry->ip, 1, irq_flags, regs); + perf_trace_buf_submit(entry, size, rctx, entry->ip, 1, irq_flags, regs, call->perf_data); } /* Kretprobe profile handler */ @@ -1395,7 +1395,7 @@ static __kprobes void kretprobe_perf_func(struct kretprobe_instance *ri, call_fetch(&tp->args[i].fetch, regs, data + tp->args[i].offset); perf_trace_buf_submit(entry, size, rctx, entry->ret_ip, 1, - irq_flags, regs); + irq_flags, regs, call->perf_data); } static int probe_perf_enable(struct ftrace_event_call *call) diff --git a/kernel/trace/trace_syscalls.c b/kernel/trace/trace_syscalls.c index 4d6d711717f2..9eff1a4b49b9 100644 --- a/kernel/trace/trace_syscalls.c +++ b/kernel/trace/trace_syscalls.c @@ -468,7 +468,8 @@ static void perf_syscall_enter(struct pt_regs *regs, long id) rec->nr = syscall_nr; syscall_get_arguments(current, regs, 0, sys_data->nb_args, (unsigned long *)&rec->args); - perf_trace_buf_submit(rec, size, rctx, 0, 1, flags, regs); + perf_trace_buf_submit(rec, size, rctx, 0, 1, flags, regs, + sys_data->enter_event->perf_data); } int perf_sysenter_enable(struct ftrace_event_call *call) @@ -543,7 +544,8 @@ static void perf_syscall_exit(struct pt_regs *regs, long ret) rec->nr = syscall_nr; rec->ret = syscall_get_return_value(current, regs); - perf_trace_buf_submit(rec, size, rctx, 0, 1, flags, regs); + perf_trace_buf_submit(rec, size, rctx, 0, 1, flags, regs, + sys_data->exit_event->perf_data); } int perf_sysexit_enable(struct ftrace_event_call *call) -- cgit v1.2.3 From b7e2ecef92d2e7785e6d76b41e5ba8bcbc45259d Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Wed, 19 May 2010 10:52:27 +0200 Subject: perf, trace: Optimize tracepoints by removing IRQ-disable from perf/tracepoint interaction Improves performance. Acked-by: Frederic Weisbecker Signed-off-by: Peter Zijlstra Cc: Arnaldo Carvalho de Melo Cc: Frederic Weisbecker Cc: Paul Mackerras Cc: Mike Galbraith Cc: Steven Rostedt LKML-Reference: <1274259525.5605.10352.camel@twins> Signed-off-by: Ingo Molnar --- kernel/trace/trace_event_perf.c | 73 ++++++++++++++++------------------------- kernel/trace/trace_kprobe.c | 10 +++--- kernel/trace/trace_syscalls.c | 10 +++--- 3 files changed, 37 insertions(+), 56 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_event_perf.c b/kernel/trace/trace_event_perf.c index 89b780a7c522..a1304f8c4440 100644 --- a/kernel/trace/trace_event_perf.c +++ b/kernel/trace/trace_event_perf.c @@ -9,13 +9,9 @@ #include #include "trace.h" -DEFINE_PER_CPU(struct pt_regs, perf_trace_regs); -EXPORT_PER_CPU_SYMBOL_GPL(perf_trace_regs); - EXPORT_SYMBOL_GPL(perf_arch_fetch_caller_regs); -static char *perf_trace_buf; -static char *perf_trace_buf_nmi; +static char *perf_trace_buf[4]; /* * Force it to be aligned to unsigned long to avoid misaligned accesses @@ -29,7 +25,6 @@ static int total_ref_count; static int perf_trace_event_enable(struct ftrace_event_call *event, void *data) { - char *buf; int ret = -ENOMEM; if (event->perf_refcount++ > 0) { @@ -38,17 +33,16 @@ static int perf_trace_event_enable(struct ftrace_event_call *event, void *data) } if (!total_ref_count) { - buf = (char *)alloc_percpu(perf_trace_t); - if (!buf) - goto fail_buf; - - rcu_assign_pointer(perf_trace_buf, buf); + char *buf; + int i; - buf = (char *)alloc_percpu(perf_trace_t); - if (!buf) - goto fail_buf_nmi; + for (i = 0; i < 4; i++) { + buf = (char *)alloc_percpu(perf_trace_t); + if (!buf) + goto fail_buf; - rcu_assign_pointer(perf_trace_buf_nmi, buf); + rcu_assign_pointer(perf_trace_buf[i], buf); + } } ret = event->perf_event_enable(event); @@ -58,14 +52,15 @@ static int perf_trace_event_enable(struct ftrace_event_call *event, void *data) return 0; } -fail_buf_nmi: +fail_buf: if (!total_ref_count) { - free_percpu(perf_trace_buf_nmi); - free_percpu(perf_trace_buf); - perf_trace_buf_nmi = NULL; - perf_trace_buf = NULL; + int i; + + for (i = 0; i < 4; i++) { + free_percpu(perf_trace_buf[i]); + perf_trace_buf[i] = NULL; + } } -fail_buf: event->perf_refcount--; return ret; @@ -91,19 +86,19 @@ int perf_trace_enable(int event_id, void *data) static void perf_trace_event_disable(struct ftrace_event_call *event) { - char *buf, *nmi_buf; - if (--event->perf_refcount > 0) return; event->perf_event_disable(event); if (!--total_ref_count) { - buf = perf_trace_buf; - rcu_assign_pointer(perf_trace_buf, NULL); + char *buf[4]; + int i; - nmi_buf = perf_trace_buf_nmi; - rcu_assign_pointer(perf_trace_buf_nmi, NULL); + for (i = 0; i < 4; i++) { + buf[i] = perf_trace_buf[i]; + rcu_assign_pointer(perf_trace_buf[i], NULL); + } /* * Ensure every events in profiling have finished before @@ -111,8 +106,8 @@ static void perf_trace_event_disable(struct ftrace_event_call *event) */ synchronize_sched(); - free_percpu(buf); - free_percpu(nmi_buf); + for (i = 0; i < 4; i++) + free_percpu(buf[i]); } } @@ -132,47 +127,37 @@ void perf_trace_disable(int event_id) } __kprobes void *perf_trace_buf_prepare(int size, unsigned short type, - int *rctxp, unsigned long *irq_flags) + struct pt_regs *regs, int *rctxp) { struct trace_entry *entry; char *trace_buf, *raw_data; - int pc, cpu; + int pc; BUILD_BUG_ON(PERF_MAX_TRACE_SIZE % sizeof(unsigned long)); pc = preempt_count(); - /* Protect the per cpu buffer, begin the rcu read side */ - local_irq_save(*irq_flags); - *rctxp = perf_swevent_get_recursion_context(); if (*rctxp < 0) goto err_recursion; - cpu = smp_processor_id(); - - if (in_nmi()) - trace_buf = rcu_dereference_sched(perf_trace_buf_nmi); - else - trace_buf = rcu_dereference_sched(perf_trace_buf); - + trace_buf = rcu_dereference_sched(perf_trace_buf[*rctxp]); if (!trace_buf) goto err; - raw_data = per_cpu_ptr(trace_buf, cpu); + raw_data = per_cpu_ptr(trace_buf, smp_processor_id()); /* zero the dead bytes from align to not leak stack to user */ memset(&raw_data[size - sizeof(u64)], 0, sizeof(u64)); entry = (struct trace_entry *)raw_data; - tracing_generic_entry_update(entry, *irq_flags, pc); + tracing_generic_entry_update(entry, regs->flags, pc); entry->type = type; return raw_data; err: perf_swevent_put_recursion_context(*rctxp); err_recursion: - local_irq_restore(*irq_flags); return NULL; } EXPORT_SYMBOL_GPL(perf_trace_buf_prepare); diff --git a/kernel/trace/trace_kprobe.c b/kernel/trace/trace_kprobe.c index 2d7bf4146be8..20c96de0aea0 100644 --- a/kernel/trace/trace_kprobe.c +++ b/kernel/trace/trace_kprobe.c @@ -1343,7 +1343,6 @@ static __kprobes void kprobe_perf_func(struct kprobe *kp, struct kprobe_trace_entry_head *entry; u8 *data; int size, __size, i; - unsigned long irq_flags; int rctx; __size = sizeof(*entry) + tp->size; @@ -1353,7 +1352,7 @@ static __kprobes void kprobe_perf_func(struct kprobe *kp, "profile buffer not large enough")) return; - entry = perf_trace_buf_prepare(size, call->id, &rctx, &irq_flags); + entry = perf_trace_buf_prepare(size, call->id, regs, &rctx); if (!entry) return; @@ -1362,7 +1361,7 @@ static __kprobes void kprobe_perf_func(struct kprobe *kp, for (i = 0; i < tp->nr_args; i++) call_fetch(&tp->args[i].fetch, regs, data + tp->args[i].offset); - perf_trace_buf_submit(entry, size, rctx, entry->ip, 1, irq_flags, regs, call->perf_data); + perf_trace_buf_submit(entry, size, rctx, entry->ip, 1, regs, call->perf_data); } /* Kretprobe profile handler */ @@ -1374,7 +1373,6 @@ static __kprobes void kretprobe_perf_func(struct kretprobe_instance *ri, struct kretprobe_trace_entry_head *entry; u8 *data; int size, __size, i; - unsigned long irq_flags; int rctx; __size = sizeof(*entry) + tp->size; @@ -1384,7 +1382,7 @@ static __kprobes void kretprobe_perf_func(struct kretprobe_instance *ri, "profile buffer not large enough")) return; - entry = perf_trace_buf_prepare(size, call->id, &rctx, &irq_flags); + entry = perf_trace_buf_prepare(size, call->id, regs, &rctx); if (!entry) return; @@ -1395,7 +1393,7 @@ static __kprobes void kretprobe_perf_func(struct kretprobe_instance *ri, call_fetch(&tp->args[i].fetch, regs, data + tp->args[i].offset); perf_trace_buf_submit(entry, size, rctx, entry->ret_ip, 1, - irq_flags, regs, call->perf_data); + regs, call->perf_data); } static int probe_perf_enable(struct ftrace_event_call *call) diff --git a/kernel/trace/trace_syscalls.c b/kernel/trace/trace_syscalls.c index 9eff1a4b49b9..a657cefbb137 100644 --- a/kernel/trace/trace_syscalls.c +++ b/kernel/trace/trace_syscalls.c @@ -438,7 +438,6 @@ static void perf_syscall_enter(struct pt_regs *regs, long id) { struct syscall_metadata *sys_data; struct syscall_trace_enter *rec; - unsigned long flags; int syscall_nr; int rctx; int size; @@ -461,14 +460,14 @@ static void perf_syscall_enter(struct pt_regs *regs, long id) return; rec = (struct syscall_trace_enter *)perf_trace_buf_prepare(size, - sys_data->enter_event->id, &rctx, &flags); + sys_data->enter_event->id, regs, &rctx); if (!rec) return; rec->nr = syscall_nr; syscall_get_arguments(current, regs, 0, sys_data->nb_args, (unsigned long *)&rec->args); - perf_trace_buf_submit(rec, size, rctx, 0, 1, flags, regs, + perf_trace_buf_submit(rec, size, rctx, 0, 1, regs, sys_data->enter_event->perf_data); } @@ -511,7 +510,6 @@ static void perf_syscall_exit(struct pt_regs *regs, long ret) { struct syscall_metadata *sys_data; struct syscall_trace_exit *rec; - unsigned long flags; int syscall_nr; int rctx; int size; @@ -537,14 +535,14 @@ static void perf_syscall_exit(struct pt_regs *regs, long ret) return; rec = (struct syscall_trace_exit *)perf_trace_buf_prepare(size, - sys_data->exit_event->id, &rctx, &flags); + sys_data->exit_event->id, regs, &rctx); if (!rec) return; rec->nr = syscall_nr; rec->ret = syscall_get_return_value(current, regs); - perf_trace_buf_submit(rec, size, rctx, 0, 1, flags, regs, + perf_trace_buf_submit(rec, size, rctx, 0, 1, regs, sys_data->exit_event->perf_data); } -- cgit v1.2.3 From 1c024eca51fdc965290acf342ae16a476c2189d0 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Wed, 19 May 2010 14:02:22 +0200 Subject: perf, trace: Optimize tracepoints by using per-tracepoint-per-cpu hlist to track events Avoid the swevent hash-table by using per-tracepoint hlists. Also, avoid conditionals on the fast path by ordering with probe unregister so that we should never get on the callback path without the data being there. Signed-off-by: Peter Zijlstra Cc: Arnaldo Carvalho de Melo Cc: Frederic Weisbecker Cc: Paul Mackerras Cc: Mike Galbraith Cc: Steven Rostedt LKML-Reference: <20100521090710.473188012@chello.nl> Signed-off-by: Ingo Molnar --- kernel/trace/trace_event_perf.c | 127 +++++++++++++++++++++------------------- kernel/trace/trace_kprobe.c | 9 ++- kernel/trace/trace_syscalls.c | 11 ++-- 3 files changed, 80 insertions(+), 67 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_event_perf.c b/kernel/trace/trace_event_perf.c index a1304f8c4440..39d5ea7b0653 100644 --- a/kernel/trace/trace_event_perf.c +++ b/kernel/trace/trace_event_perf.c @@ -23,14 +23,25 @@ typedef typeof(unsigned long [PERF_MAX_TRACE_SIZE / sizeof(unsigned long)]) /* Count the events in use (per event id, not per instance) */ static int total_ref_count; -static int perf_trace_event_enable(struct ftrace_event_call *event, void *data) +static int perf_trace_event_init(struct ftrace_event_call *tp_event, + struct perf_event *p_event) { + struct hlist_head *list; int ret = -ENOMEM; + int cpu; - if (event->perf_refcount++ > 0) { - event->perf_data = NULL; + p_event->tp_event = tp_event; + if (tp_event->perf_refcount++ > 0) return 0; - } + + list = alloc_percpu(struct hlist_head); + if (!list) + goto fail; + + for_each_possible_cpu(cpu) + INIT_HLIST_HEAD(per_cpu_ptr(list, cpu)); + + tp_event->perf_events = list; if (!total_ref_count) { char *buf; @@ -39,20 +50,20 @@ static int perf_trace_event_enable(struct ftrace_event_call *event, void *data) for (i = 0; i < 4; i++) { buf = (char *)alloc_percpu(perf_trace_t); if (!buf) - goto fail_buf; + goto fail; - rcu_assign_pointer(perf_trace_buf[i], buf); + perf_trace_buf[i] = buf; } } - ret = event->perf_event_enable(event); - if (!ret) { - event->perf_data = data; - total_ref_count++; - return 0; - } + ret = tp_event->perf_event_enable(tp_event); + if (ret) + goto fail; -fail_buf: + total_ref_count++; + return 0; + +fail: if (!total_ref_count) { int i; @@ -61,21 +72,26 @@ fail_buf: perf_trace_buf[i] = NULL; } } - event->perf_refcount--; + + if (!--tp_event->perf_refcount) { + free_percpu(tp_event->perf_events); + tp_event->perf_events = NULL; + } return ret; } -int perf_trace_enable(int event_id, void *data) +int perf_trace_init(struct perf_event *p_event) { - struct ftrace_event_call *event; + struct ftrace_event_call *tp_event; + int event_id = p_event->attr.config; int ret = -EINVAL; mutex_lock(&event_mutex); - list_for_each_entry(event, &ftrace_events, list) { - if (event->id == event_id && event->perf_event_enable && - try_module_get(event->mod)) { - ret = perf_trace_event_enable(event, data); + list_for_each_entry(tp_event, &ftrace_events, list) { + if (tp_event->id == event_id && tp_event->perf_event_enable && + try_module_get(tp_event->mod)) { + ret = perf_trace_event_init(tp_event, p_event); break; } } @@ -84,53 +100,52 @@ int perf_trace_enable(int event_id, void *data) return ret; } -static void perf_trace_event_disable(struct ftrace_event_call *event) +int perf_trace_enable(struct perf_event *p_event) { - if (--event->perf_refcount > 0) - return; + struct ftrace_event_call *tp_event = p_event->tp_event; + struct hlist_head *list; - event->perf_event_disable(event); + list = tp_event->perf_events; + if (WARN_ON_ONCE(!list)) + return -EINVAL; - if (!--total_ref_count) { - char *buf[4]; - int i; - - for (i = 0; i < 4; i++) { - buf[i] = perf_trace_buf[i]; - rcu_assign_pointer(perf_trace_buf[i], NULL); - } + list = per_cpu_ptr(list, smp_processor_id()); + hlist_add_head_rcu(&p_event->hlist_entry, list); - /* - * Ensure every events in profiling have finished before - * releasing the buffers - */ - synchronize_sched(); + return 0; +} - for (i = 0; i < 4; i++) - free_percpu(buf[i]); - } +void perf_trace_disable(struct perf_event *p_event) +{ + hlist_del_rcu(&p_event->hlist_entry); } -void perf_trace_disable(int event_id) +void perf_trace_destroy(struct perf_event *p_event) { - struct ftrace_event_call *event; + struct ftrace_event_call *tp_event = p_event->tp_event; + int i; - mutex_lock(&event_mutex); - list_for_each_entry(event, &ftrace_events, list) { - if (event->id == event_id) { - perf_trace_event_disable(event); - module_put(event->mod); - break; + if (--tp_event->perf_refcount > 0) + return; + + tp_event->perf_event_disable(tp_event); + + free_percpu(tp_event->perf_events); + tp_event->perf_events = NULL; + + if (!--total_ref_count) { + for (i = 0; i < 4; i++) { + free_percpu(perf_trace_buf[i]); + perf_trace_buf[i] = NULL; } } - mutex_unlock(&event_mutex); } __kprobes void *perf_trace_buf_prepare(int size, unsigned short type, struct pt_regs *regs, int *rctxp) { struct trace_entry *entry; - char *trace_buf, *raw_data; + char *raw_data; int pc; BUILD_BUG_ON(PERF_MAX_TRACE_SIZE % sizeof(unsigned long)); @@ -139,13 +154,9 @@ __kprobes void *perf_trace_buf_prepare(int size, unsigned short type, *rctxp = perf_swevent_get_recursion_context(); if (*rctxp < 0) - goto err_recursion; - - trace_buf = rcu_dereference_sched(perf_trace_buf[*rctxp]); - if (!trace_buf) - goto err; + return NULL; - raw_data = per_cpu_ptr(trace_buf, smp_processor_id()); + raw_data = per_cpu_ptr(perf_trace_buf[*rctxp], smp_processor_id()); /* zero the dead bytes from align to not leak stack to user */ memset(&raw_data[size - sizeof(u64)], 0, sizeof(u64)); @@ -155,9 +166,5 @@ __kprobes void *perf_trace_buf_prepare(int size, unsigned short type, entry->type = type; return raw_data; -err: - perf_swevent_put_recursion_context(*rctxp); -err_recursion: - return NULL; } EXPORT_SYMBOL_GPL(perf_trace_buf_prepare); diff --git a/kernel/trace/trace_kprobe.c b/kernel/trace/trace_kprobe.c index 20c96de0aea0..4681f60dac00 100644 --- a/kernel/trace/trace_kprobe.c +++ b/kernel/trace/trace_kprobe.c @@ -1341,6 +1341,7 @@ static __kprobes void kprobe_perf_func(struct kprobe *kp, struct trace_probe *tp = container_of(kp, struct trace_probe, rp.kp); struct ftrace_event_call *call = &tp->call; struct kprobe_trace_entry_head *entry; + struct hlist_head *head; u8 *data; int size, __size, i; int rctx; @@ -1361,7 +1362,8 @@ static __kprobes void kprobe_perf_func(struct kprobe *kp, for (i = 0; i < tp->nr_args; i++) call_fetch(&tp->args[i].fetch, regs, data + tp->args[i].offset); - perf_trace_buf_submit(entry, size, rctx, entry->ip, 1, regs, call->perf_data); + head = per_cpu_ptr(call->perf_events, smp_processor_id()); + perf_trace_buf_submit(entry, size, rctx, entry->ip, 1, regs, head); } /* Kretprobe profile handler */ @@ -1371,6 +1373,7 @@ static __kprobes void kretprobe_perf_func(struct kretprobe_instance *ri, struct trace_probe *tp = container_of(ri->rp, struct trace_probe, rp); struct ftrace_event_call *call = &tp->call; struct kretprobe_trace_entry_head *entry; + struct hlist_head *head; u8 *data; int size, __size, i; int rctx; @@ -1392,8 +1395,8 @@ static __kprobes void kretprobe_perf_func(struct kretprobe_instance *ri, for (i = 0; i < tp->nr_args; i++) call_fetch(&tp->args[i].fetch, regs, data + tp->args[i].offset); - perf_trace_buf_submit(entry, size, rctx, entry->ret_ip, 1, - regs, call->perf_data); + head = per_cpu_ptr(call->perf_events, smp_processor_id()); + perf_trace_buf_submit(entry, size, rctx, entry->ret_ip, 1, regs, head); } static int probe_perf_enable(struct ftrace_event_call *call) diff --git a/kernel/trace/trace_syscalls.c b/kernel/trace/trace_syscalls.c index a657cefbb137..eb769f270291 100644 --- a/kernel/trace/trace_syscalls.c +++ b/kernel/trace/trace_syscalls.c @@ -438,6 +438,7 @@ static void perf_syscall_enter(struct pt_regs *regs, long id) { struct syscall_metadata *sys_data; struct syscall_trace_enter *rec; + struct hlist_head *head; int syscall_nr; int rctx; int size; @@ -467,8 +468,9 @@ static void perf_syscall_enter(struct pt_regs *regs, long id) rec->nr = syscall_nr; syscall_get_arguments(current, regs, 0, sys_data->nb_args, (unsigned long *)&rec->args); - perf_trace_buf_submit(rec, size, rctx, 0, 1, regs, - sys_data->enter_event->perf_data); + + head = per_cpu_ptr(sys_data->enter_event->perf_events, smp_processor_id()); + perf_trace_buf_submit(rec, size, rctx, 0, 1, regs, head); } int perf_sysenter_enable(struct ftrace_event_call *call) @@ -510,6 +512,7 @@ static void perf_syscall_exit(struct pt_regs *regs, long ret) { struct syscall_metadata *sys_data; struct syscall_trace_exit *rec; + struct hlist_head *head; int syscall_nr; int rctx; int size; @@ -542,8 +545,8 @@ static void perf_syscall_exit(struct pt_regs *regs, long ret) rec->nr = syscall_nr; rec->ret = syscall_get_return_value(current, regs); - perf_trace_buf_submit(rec, size, rctx, 0, 1, regs, - sys_data->exit_event->perf_data); + head = per_cpu_ptr(sys_data->exit_event->perf_events, smp_processor_id()); + perf_trace_buf_submit(rec, size, rctx, 0, 1, regs, head); } int perf_sysexit_enable(struct ftrace_event_call *call) -- cgit v1.2.3 From 35f3d14dbbc58447c61e38a162ea10add6b31dc7 Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Thu, 20 May 2010 10:43:18 +0200 Subject: pipe: add support for shrinking and growing pipes This patch adds F_GETPIPE_SZ and F_SETPIPE_SZ fcntl() actions for growing and shrinking the size of a pipe and adjusts pipe.c and splice.c (and relay and network splice) usage to work with these larger (or smaller) pipes. Signed-off-by: Jens Axboe --- kernel/trace/trace.c | 60 +++++++++++++++++++++++++++++++--------------------- 1 file changed, 36 insertions(+), 24 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index 44f916a04065..7b155a0e6f31 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -3269,12 +3269,12 @@ static ssize_t tracing_splice_read_pipe(struct file *filp, size_t len, unsigned int flags) { - struct page *pages[PIPE_BUFFERS]; - struct partial_page partial[PIPE_BUFFERS]; + struct page *pages_def[PIPE_DEF_BUFFERS]; + struct partial_page partial_def[PIPE_DEF_BUFFERS]; struct trace_iterator *iter = filp->private_data; struct splice_pipe_desc spd = { - .pages = pages, - .partial = partial, + .pages = pages_def, + .partial = partial_def, .nr_pages = 0, /* This gets updated below. */ .flags = flags, .ops = &tracing_pipe_buf_ops, @@ -3285,6 +3285,9 @@ static ssize_t tracing_splice_read_pipe(struct file *filp, size_t rem; unsigned int i; + if (splice_grow_spd(pipe, &spd)) + return -ENOMEM; + /* copy the tracer to avoid using a global lock all around */ mutex_lock(&trace_types_lock); if (unlikely(old_tracer != current_trace && current_trace)) { @@ -3315,23 +3318,23 @@ static ssize_t tracing_splice_read_pipe(struct file *filp, trace_access_lock(iter->cpu_file); /* Fill as many pages as possible. */ - for (i = 0, rem = len; i < PIPE_BUFFERS && rem; i++) { - pages[i] = alloc_page(GFP_KERNEL); - if (!pages[i]) + for (i = 0, rem = len; i < pipe->buffers && rem; i++) { + spd.pages[i] = alloc_page(GFP_KERNEL); + if (!spd.pages[i]) break; rem = tracing_fill_pipe_page(rem, iter); /* Copy the data into the page, so we can start over. */ ret = trace_seq_to_buffer(&iter->seq, - page_address(pages[i]), + page_address(spd.pages[i]), iter->seq.len); if (ret < 0) { - __free_page(pages[i]); + __free_page(spd.pages[i]); break; } - partial[i].offset = 0; - partial[i].len = iter->seq.len; + spd.partial[i].offset = 0; + spd.partial[i].len = iter->seq.len; trace_seq_init(&iter->seq); } @@ -3342,12 +3345,14 @@ static ssize_t tracing_splice_read_pipe(struct file *filp, spd.nr_pages = i; - return splice_to_pipe(pipe, &spd); + ret = splice_to_pipe(pipe, &spd); +out: + splice_shrink_spd(pipe, &spd); + return ret; out_err: mutex_unlock(&iter->mutex); - - return ret; + goto out; } static ssize_t @@ -3746,11 +3751,11 @@ tracing_buffers_splice_read(struct file *file, loff_t *ppos, unsigned int flags) { struct ftrace_buffer_info *info = file->private_data; - struct partial_page partial[PIPE_BUFFERS]; - struct page *pages[PIPE_BUFFERS]; + struct partial_page partial_def[PIPE_DEF_BUFFERS]; + struct page *pages_def[PIPE_DEF_BUFFERS]; struct splice_pipe_desc spd = { - .pages = pages, - .partial = partial, + .pages = pages_def, + .partial = partial_def, .flags = flags, .ops = &buffer_pipe_buf_ops, .spd_release = buffer_spd_release, @@ -3759,22 +3764,28 @@ tracing_buffers_splice_read(struct file *file, loff_t *ppos, int entries, size, i; size_t ret; + if (splice_grow_spd(pipe, &spd)) + return -ENOMEM; + if (*ppos & (PAGE_SIZE - 1)) { WARN_ONCE(1, "Ftrace: previous read must page-align\n"); - return -EINVAL; + ret = -EINVAL; + goto out; } if (len & (PAGE_SIZE - 1)) { WARN_ONCE(1, "Ftrace: splice_read should page-align\n"); - if (len < PAGE_SIZE) - return -EINVAL; + if (len < PAGE_SIZE) { + ret = -EINVAL; + goto out; + } len &= PAGE_MASK; } trace_access_lock(info->cpu); entries = ring_buffer_entries_cpu(info->tr->buffer, info->cpu); - for (i = 0; i < PIPE_BUFFERS && len && entries; i++, len -= PAGE_SIZE) { + for (i = 0; i < pipe->buffers && len && entries; i++, len -= PAGE_SIZE) { struct page *page; int r; @@ -3829,11 +3840,12 @@ tracing_buffers_splice_read(struct file *file, loff_t *ppos, else ret = 0; /* TODO: block */ - return ret; + goto out; } ret = splice_to_pipe(pipe, &spd); - + splice_shrink_spd(pipe, &spd); +out: return ret; } -- cgit v1.2.3 From 87f44bbc246c5244c76a701f8eefba7788bce64a Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Tue, 25 May 2010 11:02:55 +0200 Subject: perf, trace: Fix !x86 build bug Patch b7e2ecef92 (perf, trace: Optimize tracepoints by removing IRQ-disable from perf/tracepoint interaction) made the unfortunate mistake of assuming the world is x86 only, correct this. The problem was that perf_fetch_caller_regs() did local_save_flags() into regs->flags, and I re-used that to remove another local_save_flags(), forgetting !x86 doesn't have regs->flags. Do the reverse, remove the local_save_flags() from perf_fetch_caller_regs() and let the ftrace site do the local_save_flags() instead. Signed-off-by: Peter Zijlstra Acked-by: Paul Mackerras Cc: acme@redhat.com Cc: efault@gmx.de Cc: fweisbec@gmail.com Cc: rostedt@goodmis.org LKML-Reference: <1274778175.5882.623.camel@twins> Signed-off-by: Ingo Molnar --- kernel/trace/trace_event_perf.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_event_perf.c b/kernel/trace/trace_event_perf.c index 26b8607a0abc..cb6f365016e4 100644 --- a/kernel/trace/trace_event_perf.c +++ b/kernel/trace/trace_event_perf.c @@ -157,6 +157,7 @@ __kprobes void *perf_trace_buf_prepare(int size, unsigned short type, struct pt_regs *regs, int *rctxp) { struct trace_entry *entry; + unsigned long flags; char *raw_data; int pc; @@ -174,7 +175,8 @@ __kprobes void *perf_trace_buf_prepare(int size, unsigned short type, memset(&raw_data[size - sizeof(u64)], 0, sizeof(u64)); entry = (struct trace_entry *)raw_data; - tracing_generic_entry_update(entry, regs->flags, pc); + local_save_flags(flags); + tracing_generic_entry_update(entry, flags, pc); entry->type = type; return raw_data; -- cgit v1.2.3 From b3230c8b44da5838cf396942d5c1ab19f8e8f720 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Fri, 21 May 2010 11:55:21 -0400 Subject: ring-buffer: Reset "real_end" when page is filled The code to store the "lost events" requires knowing the real end of the page. Since the 'commit' includes the padding at the end of a page a "real_end" variable was used to keep track of the end not including the padding. If events were lost, the reader can place the count of events in the padded area if there is enough room. The bug this patch fixes is that when we fill the page we do not reset the real_end variable, and if the writer had wrapped a few times, the real_end would be incorrect. This patch simply resets the real_end if the page was filled. Signed-off-by: Steven Rostedt --- kernel/trace/ring_buffer.c | 8 ++++++++ 1 file changed, 8 insertions(+) (limited to 'kernel/trace') diff --git a/kernel/trace/ring_buffer.c b/kernel/trace/ring_buffer.c index 7f6059c5aa94..b0702ff78218 100644 --- a/kernel/trace/ring_buffer.c +++ b/kernel/trace/ring_buffer.c @@ -1768,6 +1768,14 @@ rb_reset_tail(struct ring_buffer_per_cpu *cpu_buffer, * must fill the old tail_page with padding. */ if (tail >= BUF_PAGE_SIZE) { + /* + * If the page was filled, then we still need + * to update the real_end. Reset it to zero + * and the reader will ignore it. + */ + if (tail == BUF_PAGE_SIZE) + tail_page->real_end = 0; + local_sub(length, &tail_page->write); return; } -- cgit v1.2.3 From 2711ca237a084286ea1c2dcf82ab2aadab23a00d Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Fri, 21 May 2010 13:32:26 -0400 Subject: ring-buffer: Move zeroing out excess in page to ring buffer code Currently the trace splice code zeros out the excess bytes in the page before sending it off to userspace. This is to make sure userspace is not getting anything it should not be when reading the pages, because the excess data was never initialized to zero before writing (for perfomance reasons). But the splice code has no business in doing this work, it should be done by the ring buffer. With the latest changes for recording lost events, the splice code gets it wrong anyway. Move the zeroing out of excess bytes into the ring buffer code. Signed-off-by: Steven Rostedt --- kernel/trace/ring_buffer.c | 11 +++++++++-- kernel/trace/trace.c | 6 ------ 2 files changed, 9 insertions(+), 8 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/ring_buffer.c b/kernel/trace/ring_buffer.c index b0702ff78218..1da7b6ea8b85 100644 --- a/kernel/trace/ring_buffer.c +++ b/kernel/trace/ring_buffer.c @@ -3902,12 +3902,12 @@ int ring_buffer_read_page(struct ring_buffer *buffer, ret = read; cpu_buffer->lost_events = 0; + + commit = local_read(&bpage->commit); /* * Set a flag in the commit field if we lost events */ if (missed_events) { - commit = local_read(&bpage->commit); - /* If there is room at the end of the page to save the * missed events, then record it there. */ @@ -3915,10 +3915,17 @@ int ring_buffer_read_page(struct ring_buffer *buffer, memcpy(&bpage->data[commit], &missed_events, sizeof(missed_events)); local_add(RB_MISSED_STORED, &bpage->commit); + commit += sizeof(missed_events); } local_add(RB_MISSED_EVENTS, &bpage->commit); } + /* + * This page may be off to user land. Zero it out here. + */ + if (commit < BUF_PAGE_SIZE) + memset(&bpage->data[commit], 0, BUF_PAGE_SIZE - commit); + out_unlock: spin_unlock_irqrestore(&cpu_buffer->reader_lock, flags); diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index ba0ec81158b2..95d0b1a28f93 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -3661,7 +3661,6 @@ tracing_buffers_read(struct file *filp, char __user *ubuf, size_t count, loff_t *ppos) { struct ftrace_buffer_info *info = filp->private_data; - unsigned int pos; ssize_t ret; size_t size; @@ -3688,11 +3687,6 @@ tracing_buffers_read(struct file *filp, char __user *ubuf, if (ret < 0) return 0; - pos = ring_buffer_page_len(info->spare); - - if (pos < PAGE_SIZE) - memset(info->spare + pos, 0, PAGE_SIZE - pos); - read: size = PAGE_SIZE - info->read; if (size > count) -- cgit v1.2.3 From 3771f0771154675d4a0ca780be2411f3cc357208 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Fri, 21 May 2010 12:31:09 +0200 Subject: perf_events, trace: Fix probe unregister race tracepoint_probe_unregister() does not synchronize against the probe callbacks, so do that explicitly. This properly serializes the callbacks and the free of the data used therein. Also, use this_cpu_ptr() where possible. Acked-by: Frederic Weisbecker Signed-off-by: Peter Zijlstra LKML-Reference: <1274438476.1674.1702.camel@laptop> Signed-off-by: Ingo Molnar --- kernel/trace/trace_event_perf.c | 10 ++++++++-- kernel/trace/trace_kprobe.c | 4 ++-- kernel/trace/trace_syscalls.c | 4 ++-- 3 files changed, 12 insertions(+), 6 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_event_perf.c b/kernel/trace/trace_event_perf.c index cb6f365016e4..49c7abf2ba5c 100644 --- a/kernel/trace/trace_event_perf.c +++ b/kernel/trace/trace_event_perf.c @@ -116,7 +116,7 @@ int perf_trace_enable(struct perf_event *p_event) if (WARN_ON_ONCE(!list)) return -EINVAL; - list = per_cpu_ptr(list, smp_processor_id()); + list = this_cpu_ptr(list); hlist_add_head_rcu(&p_event->hlist_entry, list); return 0; @@ -142,6 +142,12 @@ void perf_trace_destroy(struct perf_event *p_event) tp_event->class->perf_probe, tp_event); + /* + * Ensure our callback won't be called anymore. See + * tracepoint_probe_unregister() and __DO_TRACE(). + */ + synchronize_sched(); + free_percpu(tp_event->perf_events); tp_event->perf_events = NULL; @@ -169,7 +175,7 @@ __kprobes void *perf_trace_buf_prepare(int size, unsigned short type, if (*rctxp < 0) return NULL; - raw_data = per_cpu_ptr(perf_trace_buf[*rctxp], smp_processor_id()); + raw_data = this_cpu_ptr(perf_trace_buf[*rctxp]); /* zero the dead bytes from align to not leak stack to user */ memset(&raw_data[size - sizeof(u64)], 0, sizeof(u64)); diff --git a/kernel/trace/trace_kprobe.c b/kernel/trace/trace_kprobe.c index faf7cefd15da..f52b5f50299d 100644 --- a/kernel/trace/trace_kprobe.c +++ b/kernel/trace/trace_kprobe.c @@ -1359,7 +1359,7 @@ static __kprobes void kprobe_perf_func(struct kprobe *kp, for (i = 0; i < tp->nr_args; i++) call_fetch(&tp->args[i].fetch, regs, data + tp->args[i].offset); - head = per_cpu_ptr(call->perf_events, smp_processor_id()); + head = this_cpu_ptr(call->perf_events); perf_trace_buf_submit(entry, size, rctx, entry->ip, 1, regs, head); } @@ -1392,7 +1392,7 @@ static __kprobes void kretprobe_perf_func(struct kretprobe_instance *ri, for (i = 0; i < tp->nr_args; i++) call_fetch(&tp->args[i].fetch, regs, data + tp->args[i].offset); - head = per_cpu_ptr(call->perf_events, smp_processor_id()); + head = this_cpu_ptr(call->perf_events); perf_trace_buf_submit(entry, size, rctx, entry->ret_ip, 1, regs, head); } diff --git a/kernel/trace/trace_syscalls.c b/kernel/trace/trace_syscalls.c index d2c859cec9ea..34e35804304b 100644 --- a/kernel/trace/trace_syscalls.c +++ b/kernel/trace/trace_syscalls.c @@ -519,7 +519,7 @@ static void perf_syscall_enter(void *ignore, struct pt_regs *regs, long id) syscall_get_arguments(current, regs, 0, sys_data->nb_args, (unsigned long *)&rec->args); - head = per_cpu_ptr(sys_data->enter_event->perf_events, smp_processor_id()); + head = this_cpu_ptr(sys_data->enter_event->perf_events); perf_trace_buf_submit(rec, size, rctx, 0, 1, regs, head); } @@ -595,7 +595,7 @@ static void perf_syscall_exit(void *ignore, struct pt_regs *regs, long ret) rec->nr = syscall_nr; rec->ret = syscall_get_return_value(current, regs); - head = per_cpu_ptr(sys_data->exit_event->perf_events, smp_processor_id()); + head = this_cpu_ptr(sys_data->exit_event->perf_events); perf_trace_buf_submit(rec, size, rctx, 0, 1, regs, head); } -- cgit v1.2.3 From 2e97942fe57864588774f173cf4cd7bb68968b76 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Fri, 21 May 2010 16:22:33 +0200 Subject: perf_events, trace: Fix perf_trace_destroy(), mutex went missing Steve spotted I forgot to do the destroy under event_mutex. Reported-by: Steven Rostedt Signed-off-by: Peter Zijlstra LKML-Reference: <1274451913.1674.1707.camel@laptop> Signed-off-by: Ingo Molnar --- kernel/trace/trace_event_perf.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_event_perf.c b/kernel/trace/trace_event_perf.c index 49c7abf2ba5c..e6f65887842c 100644 --- a/kernel/trace/trace_event_perf.c +++ b/kernel/trace/trace_event_perf.c @@ -132,8 +132,9 @@ void perf_trace_destroy(struct perf_event *p_event) struct ftrace_event_call *tp_event = p_event->tp_event; int i; + mutex_lock(&event_mutex); if (--tp_event->perf_refcount > 0) - return; + goto out; if (tp_event->class->reg) tp_event->class->reg(tp_event, TRACE_REG_PERF_UNREGISTER); @@ -157,6 +158,8 @@ void perf_trace_destroy(struct perf_event *p_event) perf_trace_buf[i] = NULL; } } +out: + mutex_unlock(&event_mutex); } __kprobes void *perf_trace_buf_prepare(int size, unsigned short type, -- cgit v1.2.3 From 546cf44a1b507c1cbb5cf42bbe6169780567f36f Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Sat, 29 May 2010 11:45:07 -0700 Subject: blktrace: Fix new kernel-doc warnings Fix blktrace.c kernel-doc warnings: Warning(kernel/trace/blktrace.c:858): No description found for parameter 'ignore' Warning(kernel/trace/blktrace.c:890): No description found for parameter 'ignore' Signed-off-by: Randy Dunlap Cc: Jens Axboe Cc: Steven Rostedt Cc: Frederic Weisbecker LKML-Reference: <20100529114507.c466fc1e.randy.dunlap@oracle.com> Signed-off-by: Ingo Molnar --- kernel/trace/blktrace.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'kernel/trace') diff --git a/kernel/trace/blktrace.c b/kernel/trace/blktrace.c index 36ea2b65dcdc..638711c17504 100644 --- a/kernel/trace/blktrace.c +++ b/kernel/trace/blktrace.c @@ -842,6 +842,7 @@ static void blk_add_trace_split(void *ignore, /** * blk_add_trace_remap - Add a trace for a remap operation + * @ignore: trace callback data parameter (not used) * @q: queue the io is for * @bio: the source bio * @dev: target device @@ -873,6 +874,7 @@ static void blk_add_trace_remap(void *ignore, /** * blk_add_trace_rq_remap - Add a trace for a request-remap operation + * @ignore: trace callback data parameter (not used) * @q: queue the io is for * @rq: the source request * @dev: target device -- cgit v1.2.3 From a8fb2608053547bc3152ea61a5ec7cdfce5d942c Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Thu, 10 Jun 2010 14:53:16 -0400 Subject: perf/tracing: Fix regression of perf losing kprobe events With the addition of the code to shrink the kernel tracepoint infrastructure, we lost kprobes being traced by perf. The reason is that I tested if the "tp_event->class->perf_probe" existed before enabling it. This prevents "ftrace only" events (like the function trace events) from being enabled by perf. Unfortunately, kprobe events do not use perf_probe. This causes kprobes to be missed by perf. To fix this, we add the test to see if "tp_event->class->reg" exists as well as perf_probe. Normal trace events have only "perf_probe" but no "reg" function, and kprobes and syscalls have the "reg" but no "perf_probe". The ftrace unique events do not have either, so this is a valid test. If a kprobe or syscall is not to be probed by perf, the "reg" function is called anyway, and will return a failure and prevent perf from probing it. Reported-by: Srikar Dronamraju Tested-by: Srikar Dronamraju Acked-by: Peter Zijlstra Signed-off-by: Steven Rostedt --- kernel/trace/trace_event_perf.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_event_perf.c b/kernel/trace/trace_event_perf.c index e6f65887842c..8a2b73f7c068 100644 --- a/kernel/trace/trace_event_perf.c +++ b/kernel/trace/trace_event_perf.c @@ -96,7 +96,9 @@ int perf_trace_init(struct perf_event *p_event) mutex_lock(&event_mutex); list_for_each_entry(tp_event, &ftrace_events, list) { if (tp_event->event.type == event_id && - tp_event->class && tp_event->class->perf_probe && + tp_event->class && + (tp_event->class->perf_probe || + tp_event->class->reg) && try_module_get(tp_event->mod)) { ret = perf_trace_event_init(tp_event, p_event); break; -- cgit v1.2.3