diff options
Diffstat (limited to 'tools')
55 files changed, 10371 insertions, 1552 deletions
diff --git a/tools/perf/.gitignore b/tools/perf/.gitignore index d69a759a104..0854f110bf7 100644 --- a/tools/perf/.gitignore +++ b/tools/perf/.gitignore @@ -10,6 +10,7 @@ perf-stat perf-top perf*.1 perf*.xml +perf*.html common-cmds.h tags TAGS diff --git a/tools/perf/Documentation/perf-record.txt b/tools/perf/Documentation/perf-record.txt index 6be696b0a2b..0ff23de9e45 100644 --- a/tools/perf/Documentation/perf-record.txt +++ b/tools/perf/Documentation/perf-record.txt @@ -91,6 +91,10 @@ OPTIONS --no-samples:: Don't sample. +-R:: +--raw-samples:: +Collect raw sample records from all opened counters (typically for tracepoint counters). + SEE ALSO -------- linkperf:perf-stat[1], linkperf:perf-list[1] diff --git a/tools/perf/Documentation/perf-report.txt b/tools/perf/Documentation/perf-report.txt index e72e9311078..59f0b846cd7 100644 --- a/tools/perf/Documentation/perf-report.txt +++ b/tools/perf/Documentation/perf-report.txt @@ -27,6 +27,9 @@ OPTIONS -n --show-nr-samples Show the number of samples for each symbol +-T +--threads + Show per-thread event counters -C:: --comms=:: Only consider symbols in these comms. CSV that understands @@ -48,6 +51,16 @@ OPTIONS all occurances of this separator in symbol names (and other output) with a '.' character, that thus it's the only non valid separator. +-g [type,min]:: +--call-graph:: + Display callchains using type and min percent threshold. + type can be either: + - flat: single column, linear exposure of callchains. + - graph: use a graph tree, displaying absolute overhead rates. + - fractal: like graph, but displays relative rates. Each branch of + the tree is considered as a new profiled object. + + Default: fractal,0.5. + SEE ALSO -------- linkperf:perf-stat[1] diff --git a/tools/perf/Documentation/perf-sched.txt b/tools/perf/Documentation/perf-sched.txt new file mode 100644 index 00000000000..1ce79198997 --- /dev/null +++ b/tools/perf/Documentation/perf-sched.txt @@ -0,0 +1,41 @@ +perf-sched(1) +============== + +NAME +---- +perf-sched - Tool to trace/measure scheduler properties (latencies) + +SYNOPSIS +-------- +[verse] +'perf sched' {record|latency|replay|trace} + +DESCRIPTION +----------- +There's four variants of perf sched: + + 'perf sched record <command>' to record the scheduling events + of an arbitrary workload. + + 'perf sched latency' to report the per task scheduling latencies + and other scheduling properties of the workload. + + 'perf sched trace' to see a detailed trace of the workload that + was recorded. + + 'perf sched replay' to simulate the workload that was recorded + via perf sched record. (this is done by starting up mockup threads + that mimic the workload based on the events in the trace. These + threads can then replay the timings (CPU runtime and sleep patterns) + of the workload as it occured when it was recorded - and can repeat + it a number of times, measuring its performance.) + +OPTIONS +------- +-D:: +--dump-raw-trace=:: + Display verbose dump of the sched data. + +SEE ALSO +-------- +linkperf:perf-record[1] diff --git a/tools/perf/Documentation/perf-timechart.txt b/tools/perf/Documentation/perf-timechart.txt new file mode 100644 index 00000000000..1c2ed3090cc --- /dev/null +++ b/tools/perf/Documentation/perf-timechart.txt @@ -0,0 +1,38 @@ +perf-timechart(1) +================= + +NAME +---- +perf-timechart - Tool to visualize total system behavior during a workload + +SYNOPSIS +-------- +[verse] +'perf timechart' {record} + +DESCRIPTION +----------- +There are two variants of perf timechart: + + 'perf timechart record <command>' to record the system level events + of an arbitrary workload. + + 'perf timechart' to turn a trace into a Scalable Vector Graphics file, + that can be viewed with popular SVG viewers such as 'Inkscape'. + +OPTIONS +------- +-o:: +--output=:: + Select the output file (default: output.svg) +-i:: +--input=:: + Select the input file (default: perf.data) +-w:: +--width=:: + Select the width of the SVG file (default: 1000) + + +SEE ALSO +-------- +linkperf:perf-record[1] diff --git a/tools/perf/Documentation/perf-trace.txt b/tools/perf/Documentation/perf-trace.txt new file mode 100644 index 00000000000..41ed75398ca --- /dev/null +++ b/tools/perf/Documentation/perf-trace.txt @@ -0,0 +1,25 @@ +perf-trace(1) +============== + +NAME +---- +perf-trace - Read perf.data (created by perf record) and display trace output + +SYNOPSIS +-------- +[verse] +'perf trace' [-i <file> | --input=file] symbol_name + +DESCRIPTION +----------- +This command reads the input file and displays the trace recorded. + +OPTIONS +------- +-D:: +--dump-raw-trace=:: + Display verbose dump of the trace data. + +SEE ALSO +-------- +linkperf:perf-record[1] diff --git a/tools/perf/Makefile b/tools/perf/Makefile index c045b4271e5..b5f1953b614 100644 --- a/tools/perf/Makefile +++ b/tools/perf/Makefile @@ -166,7 +166,35 @@ endif # CFLAGS and LDFLAGS are for the users to override from the command line. -CFLAGS = $(M64) -ggdb3 -Wall -Wextra -Wstrict-prototypes -Wmissing-declarations -Wmissing-prototypes -std=gnu99 -Wdeclaration-after-statement -Werror -O6 +# +# Include saner warnings here, which can catch bugs: +# + +EXTRA_WARNINGS := -Wcast-align +EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wformat +EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wformat-security +EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wformat-y2k +EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wshadow +EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Winit-self +EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wpacked +EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wredundant-decls +EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wstack-protector +EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wstrict-aliasing=3 +EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wswitch-default +EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wswitch-enum +EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wno-system-headers +EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wundef +EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wvolatile-register-var +EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wwrite-strings +EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wbad-function-cast +EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wmissing-declarations +EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wmissing-prototypes +EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wnested-externs +EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wold-style-definition +EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wstrict-prototypes +EXTRA_WARNINGS := $(EXTRA_WARNINGS) -Wdeclaration-after-statement + +CFLAGS = $(M64) -ggdb3 -Wall -Wextra -std=gnu99 -Werror -O6 -fstack-protector-all -D_FORTIFY_SOURCE=2 $(EXTRA_WARNINGS) LDFLAGS = -lpthread -lrt -lelf -lm ALL_CFLAGS = $(CFLAGS) ALL_LDFLAGS = $(LDFLAGS) @@ -290,7 +318,7 @@ export PERL_PATH LIB_FILE=libperf.a -LIB_H += ../../include/linux/perf_counter.h +LIB_H += ../../include/linux/perf_event.h LIB_H += ../../include/linux/rbtree.h LIB_H += ../../include/linux/list.h LIB_H += util/include/linux/list.h @@ -310,6 +338,7 @@ LIB_H += util/sigchain.h LIB_H += util/symbol.h LIB_H += util/module.h LIB_H += util/color.h +LIB_H += util/values.h LIB_OBJS += util/abspath.o LIB_OBJS += util/alias.o @@ -337,14 +366,25 @@ LIB_OBJS += util/color.o LIB_OBJS += util/pager.o LIB_OBJS += util/header.o LIB_OBJS += util/callchain.o +LIB_OBJS += util/values.o +LIB_OBJS += util/debug.o +LIB_OBJS += util/map.o +LIB_OBJS += util/thread.o +LIB_OBJS += util/trace-event-parse.o +LIB_OBJS += util/trace-event-read.o +LIB_OBJS += util/trace-event-info.o +LIB_OBJS += util/svghelper.o BUILTIN_OBJS += builtin-annotate.o BUILTIN_OBJS += builtin-help.o +BUILTIN_OBJS += builtin-sched.o BUILTIN_OBJS += builtin-list.o BUILTIN_OBJS += builtin-record.o BUILTIN_OBJS += builtin-report.o BUILTIN_OBJS += builtin-stat.o +BUILTIN_OBJS += builtin-timechart.o BUILTIN_OBJS += builtin-top.o +BUILTIN_OBJS += builtin-trace.o PERFLIBS = $(LIB_FILE) @@ -673,6 +713,12 @@ builtin-help.o: builtin-help.c common-cmds.h PERF-CFLAGS '-DPERF_MAN_PATH="$(mandir_SQ)"' \ '-DPERF_INFO_PATH="$(infodir_SQ)"' $< +builtin-timechart.o: builtin-timechart.c common-cmds.h PERF-CFLAGS + $(QUIET_CC)$(CC) -o $*.o -c $(ALL_CFLAGS) \ + '-DPERF_HTML_PATH="$(htmldir_SQ)"' \ + '-DPERF_MAN_PATH="$(mandir_SQ)"' \ + '-DPERF_INFO_PATH="$(infodir_SQ)"' $< + $(BUILT_INS): perf$X $(QUIET_BUILT_IN)$(RM) $@ && \ ln perf$X $@ 2>/dev/null || \ diff --git a/tools/perf/builtin-annotate.c b/tools/perf/builtin-annotate.c index 5e17de984dc..1ec74161581 100644 --- a/tools/perf/builtin-annotate.c +++ b/tools/perf/builtin-annotate.c @@ -17,16 +17,13 @@ #include "util/string.h" #include "perf.h" +#include "util/debug.h" #include "util/parse-options.h" #include "util/parse-events.h" - -#define SHOW_KERNEL 1 -#define SHOW_USER 2 -#define SHOW_HV 4 +#include "util/thread.h" static char const *input_name = "perf.data"; -static char *vmlinux = "vmlinux"; static char default_sort_order[] = "comm,symbol"; static char *sort_order = default_sort_order; @@ -35,13 +32,6 @@ static int force; static int input; static int show_mask = SHOW_KERNEL | SHOW_USER | SHOW_HV; -static int dump_trace = 0; -#define dprintf(x...) do { if (dump_trace) printf(x); } while (0) - -static int verbose; - -static int modules; - static int full_paths; static int print_line; @@ -49,39 +39,8 @@ static int print_line; static unsigned long page_size; static unsigned long mmap_window = 32; -struct ip_event { - struct perf_event_header header; - u64 ip; - u32 pid, tid; -}; - -struct mmap_event { - struct perf_event_header header; - u32 pid, tid; - u64 start; - u64 len; - u64 pgoff; - char filename[PATH_MAX]; -}; - -struct comm_event { - struct perf_event_header header; - u32 pid, tid; - char comm[16]; -}; - -struct fork_event { - struct perf_event_header header; - u32 pid, ppid; -}; - -typedef union event_union { - struct perf_event_header header; - struct ip_event ip; - struct mmap_event mmap; - struct comm_event comm; - struct fork_event fork; -} event_t; +static struct rb_root threads; +static struct thread *last_match; struct sym_ext { @@ -90,323 +49,6 @@ struct sym_ext { char *path; }; -static LIST_HEAD(dsos); -static struct dso *kernel_dso; -static struct dso *vdso; - - -static void dsos__add(struct dso *dso) -{ - list_add_tail(&dso->node, &dsos); -} - -static struct dso *dsos__find(const char *name) -{ - struct dso *pos; - - list_for_each_entry(pos, &dsos, node) - if (strcmp(pos->name, name) == 0) - return pos; - return NULL; -} - -static struct dso *dsos__findnew(const char *name) -{ - struct dso *dso = dsos__find(name); - int nr; - - if (dso) - return dso; - - dso = dso__new(name, 0); - if (!dso) - goto out_delete_dso; - - nr = dso__load(dso, NULL, verbose); - if (nr < 0) { - if (verbose) - fprintf(stderr, "Failed to open: %s\n", name); - goto out_delete_dso; - } - if (!nr && verbose) { - fprintf(stderr, - "No symbols found in: %s, maybe install a debug package?\n", - name); - } - - dsos__add(dso); - - return dso; - -out_delete_dso: - dso__delete(dso); - return NULL; -} - -static void dsos__fprintf(FILE *fp) -{ - struct dso *pos; - - list_for_each_entry(pos, &dsos, node) - dso__fprintf(pos, fp); -} - -static struct symbol *vdso__find_symbol(struct dso *dso, u64 ip) -{ - return dso__find_symbol(dso, ip); -} - -static int load_kernel(void) -{ - int err; - - kernel_dso = dso__new("[kernel]", 0); - if (!kernel_dso) - return -1; - - err = dso__load_kernel(kernel_dso, vmlinux, NULL, verbose, modules); - if (err <= 0) { - dso__delete(kernel_dso); - kernel_dso = NULL; - } else - dsos__add(kernel_dso); - - vdso = dso__new("[vdso]", 0); - if (!vdso) - return -1; - - vdso->find_symbol = vdso__find_symbol; - - dsos__add(vdso); - - return err; -} - -struct map { - struct list_head node; - u64 start; - u64 end; - u64 pgoff; - u64 (*map_ip)(struct map *, u64); - struct dso *dso; -}; - -static u64 map__map_ip(struct map *map, u64 ip) -{ - return ip - map->start + map->pgoff; -} - -static u64 vdso__map_ip(struct map *map __used, u64 ip) -{ - return ip; -} - -static struct map *map__new(struct mmap_event *event) -{ - struct map *self = malloc(sizeof(*self)); - - if (self != NULL) { - const char *filename = event->filename; - - self->start = event->start; - self->end = event->start + event->len; - self->pgoff = event->pgoff; - - self->dso = dsos__findnew(filename); - if (self->dso == NULL) - goto out_delete; - - if (self->dso == vdso) - self->map_ip = vdso__map_ip; - else - self->map_ip = map__map_ip; - } - return self; -out_delete: - free(self); - return NULL; -} - -static struct map *map__clone(struct map *self) -{ - struct map *map = malloc(sizeof(*self)); - - if (!map) - return NULL; - - memcpy(map, self, sizeof(*self)); - - return map; -} - -static int map__overlap(struct map *l, struct map *r) -{ - if (l->start > r->start) { - struct map *t = l; - l = r; - r = t; - } - - if (l->end > r->start) - return 1; - - return 0; -} - -static size_t map__fprintf(struct map *self, FILE *fp) -{ - return fprintf(fp, " %Lx-%Lx %Lx %s\n", - self->start, self->end, self->pgoff, self->dso->name); -} - - -struct thread { - struct rb_node rb_node; - struct list_head maps; - pid_t pid; - char *comm; -}; - -static struct thread *thread__new(pid_t pid) -{ - struct thread *self = malloc(sizeof(*self)); - - if (self != NULL) { - self->pid = pid; - self->comm = malloc(32); - if (self->comm) - snprintf(self->comm, 32, ":%d", self->pid); - INIT_LIST_HEAD(&self->maps); - } - - return self; -} - -static int thread__set_comm(struct thread *self, const char *comm) -{ - if (self->comm) - free(self->comm); - self->comm = strdup(comm); - return self->comm ? 0 : -ENOMEM; -} - -static size_t thread__fprintf(struct thread *self, FILE *fp) -{ - struct map *pos; - size_t ret = fprintf(fp, "Thread %d %s\n", self->pid, self->comm); - - list_for_each_entry(pos, &self->maps, node) - ret += map__fprintf(pos, fp); - - return ret; -} - - -static struct rb_root threads; -static struct thread *last_match; - -static struct thread *threads__findnew(pid_t pid) -{ - struct rb_node **p = &threads.rb_node; - struct rb_node *parent = NULL; - struct thread *th; - - /* - * Font-end cache - PID lookups come in blocks, - * so most of the time we dont have to look up - * the full rbtree: - */ - if (last_match && last_match->pid == pid) - return last_match; - - while (*p != NULL) { - parent = *p; - th = rb_entry(parent, struct thread, rb_node); - - if (th->pid == pid) { - last_match = th; - return th; - } - - if (pid < th->pid) - p = &(*p)->rb_left; - else - p = &(*p)->rb_right; - } - - th = thread__new(pid); - if (th != NULL) { - rb_link_node(&th->rb_node, parent, p); - rb_insert_color(&th->rb_node, &threads); - last_match = th; - } - - return th; -} - -static void thread__insert_map(struct thread *self, struct map *map) -{ - struct map *pos, *tmp; - - list_for_each_entry_safe(pos, tmp, &self->maps, node) { - if (map__overlap(pos, map)) { - list_del_init(&pos->node); - /* XXX leaks dsos */ - free(pos); - } - } - - list_add_tail(&map->node, &self->maps); -} - -static int thread__fork(struct thread *self, struct thread *parent) -{ - struct map *map; - - if (self->comm) - free(self->comm); - self->comm = strdup(parent->comm); - if (!self->comm) - return -ENOMEM; - - list_for_each_entry(map, &parent->maps, node) { - struct map *new = map__clone(map); - if (!new) - return -ENOMEM; - thread__insert_map(self, new); - } - - return 0; -} - -static struct map *thread__find_map(struct thread *self, u64 ip) -{ - struct map *pos; - - if (self == NULL) - return NULL; - - list_for_each_entry(pos, &self->maps, node) - if (ip >= pos->start && ip <= pos->end) - return pos; - - return NULL; -} - -static size_t threads__fprintf(FILE *fp) -{ - size_t ret = 0; - struct rb_node *nd; - - for (nd = rb_first(&threads); nd; nd = rb_next(nd)) { - struct thread *pos = rb_entry(nd, struct thread, rb_node); - - ret += thread__fprintf(pos, fp); - } - - return ret; -} - /* * histogram, sorted on item, collects counts */ @@ -433,7 +75,7 @@ struct hist_entry { struct sort_entry { struct list_head list; - char *header; + const char *header; int64_t (*cmp)(struct hist_entry *, struct hist_entry *); int64_t (*collapse)(struct hist_entry *, struct hist_entry *); @@ -577,7 +219,7 @@ static struct sort_entry sort_sym = { static int sort__need_collapse = 0; struct sort_dimension { - char *name; + const char *name; struct sort_entry *entry; int taken; }; @@ -830,17 +472,6 @@ static void output__resort(void) } } -static void register_idle_thread(void) -{ - struct thread *thread = threads__findnew(0); - - if (thread == NULL || - thread__set_comm(thread, "[idle]")) { - fprintf(stderr, "problem inserting idle task.\n"); - exit(-1); - } -} - static unsigned long total = 0, total_mmap = 0, total_comm = 0, @@ -853,18 +484,20 @@ process_sample_event(event_t *event, unsigned long offset, unsigned long head) char level; int show = 0; struct dso *dso = NULL; - struct thread *thread = threads__findnew(event->ip.pid); + struct thread *thread; u64 ip = event->ip.ip; struct map *map = NULL; - dprintf("%p [%p]: PERF_EVENT (IP, %d): %d: %p\n", + thread = threads__findnew(event->ip.pid, &threads, &last_match); + + dump_printf("%p [%p]: PERF_EVENT (IP, %d): %d: %p\n", (void *)(offset + head), (void *)(long)(event->header.size), event->header.misc, event->ip.pid, (void *)(long)ip); - dprintf(" ... thread: %s:%d\n", thread->comm, thread->pid); + dump_printf(" ... thread: %s:%d\n", thread->comm, thread->pid); if (thread == NULL) { fprintf(stderr, "problem processing %d event, skipping it.\n", @@ -872,15 +505,15 @@ process_sample_event(event_t *event, unsigned long offset, unsigned long head) return -1; } - if (event->header.misc & PERF_EVENT_MISC_KERNEL) { + if (event->header.misc & PERF_RECORD_MISC_KERNEL) { show = SHOW_KERNEL; level = 'k'; dso = kernel_dso; - dprintf(" ...... dso: %s\n", dso->name); + dump_printf(" ...... dso: %s\n", dso->name); - } else if (event->header.misc & PERF_EVENT_MISC_USER) { + } else if (event->header.misc & PERF_RECORD_MISC_USER) { show = SHOW_USER; level = '.'; @@ -899,12 +532,12 @@ process_sample_event(event_t *event, unsigned long offset, unsigned long head) if ((long long)ip < 0) dso = kernel_dso; } - dprintf(" ...... dso: %s\n", dso ? dso->name : "<not found>"); + dump_printf(" ...... dso: %s\n", dso ? dso->name : "<not found>"); } else { show = SHOW_HV; level = 'H'; - dprintf(" ...... dso: [hypervisor]\n"); + dump_printf(" ...... dso: [hypervisor]\n"); } if (show & show_mask) { @@ -927,10 +560,12 @@ process_sample_event(event_t *event, unsigned long offset, unsigned long head) static int process_mmap_event(event_t *event, unsigned long offset, unsigned long head) { - struct thread *thread = threads__findnew(event->mmap.pid); - struct map *map = map__new(&event->mmap); + struct thread *thread; + struct map *map = map__new(&event->mmap, NULL, 0); + + thread = threads__findnew(event->mmap.pid, &threads, &last_match); - dprintf("%p [%p]: PERF_EVENT_MMAP %d: [%p(%p) @ %p]: %s\n", + dump_printf("%p [%p]: PERF_RECORD_MMAP %d: [%p(%p) @ %p]: %s\n", (void *)(offset + head), (void *)(long)(event->header.size), event->mmap.pid, @@ -940,7 +575,7 @@ process_mmap_event(event_t *event, unsigned long offset, unsigned long head) event->mmap.filename); if (thread == NULL || map == NULL) { - dprintf("problem processing PERF_EVENT_MMAP, skipping event.\n"); + dump_printf("problem processing PERF_RECORD_MMAP, skipping event.\n"); |