aboutsummaryrefslogtreecommitdiff
path: root/tools
diff options
context:
space:
mode:
Diffstat (limited to 'tools')
-rw-r--r--tools/perf/CREDITS30
-rw-r--r--tools/perf/Documentation/perf-report.txt14
-rw-r--r--tools/perf/Documentation/perf-stat.txt6
-rw-r--r--tools/perf/Makefile6
-rw-r--r--tools/perf/builtin-annotate.c8
-rw-r--r--tools/perf/builtin-record.c127
-rw-r--r--tools/perf/builtin-report.c236
-rw-r--r--tools/perf/builtin-stat.c171
-rw-r--r--tools/perf/builtin-top.c11
-rw-r--r--tools/perf/perf.h25
-rw-r--r--tools/perf/util/callchain.c174
-rw-r--r--tools/perf/util/callchain.h33
-rw-r--r--tools/perf/util/header.c242
-rw-r--r--tools/perf/util/header.h37
-rw-r--r--tools/perf/util/help.c15
-rw-r--r--tools/perf/util/pager.c5
-rw-r--r--tools/perf/util/parse-events.c153
-rw-r--r--tools/perf/util/run-command.c95
-rw-r--r--tools/perf/util/run-command.h5
-rw-r--r--tools/perf/util/strbuf.c2
-rw-r--r--tools/perf/util/string.h2
-rw-r--r--tools/perf/util/strlist.c184
-rw-r--r--tools/perf/util/strlist.h32
-rw-r--r--tools/perf/util/symbol.c16
-rw-r--r--tools/perf/util/symbol.h5
-rw-r--r--tools/perf/util/types.h (renamed from tools/perf/types.h)0
-rw-r--r--tools/perf/util/util.h15
27 files changed, 1295 insertions, 354 deletions
diff --git a/tools/perf/CREDITS b/tools/perf/CREDITS
new file mode 100644
index 00000000000..c2ddcb3acbd
--- /dev/null
+++ b/tools/perf/CREDITS
@@ -0,0 +1,30 @@
+Most of the infrastructure that 'perf' uses here has been reused
+from the Git project, as of version:
+
+ 66996ec: Sync with 1.6.2.4
+
+Here is an (incomplete!) list of main contributors to those files
+in util/* and elsewhere:
+
+ Alex Riesen
+ Christian Couder
+ Dmitry Potapov
+ Jeff King
+ Johannes Schindelin
+ Johannes Sixt
+ Junio C Hamano
+ Linus Torvalds
+ Matthias Kestenholz
+ Michal Ostrowski
+ Miklos Vajna
+ Petr Baudis
+ Pierre Habouzit
+ René Scharfe
+ Samuel Tardieu
+ Shawn O. Pearce
+ Steffen Prohaska
+ Steve Haslam
+
+Thanks guys!
+
+The full history of the files can be found in the upstream Git commits.
diff --git a/tools/perf/Documentation/perf-report.txt b/tools/perf/Documentation/perf-report.txt
index 52d3fc6846a..8aa3f8c8870 100644
--- a/tools/perf/Documentation/perf-report.txt
+++ b/tools/perf/Documentation/perf-report.txt
@@ -13,13 +13,25 @@ SYNOPSIS
DESCRIPTION
-----------
This command displays the performance counter profile information recorded
-via perf report.
+via perf record.
OPTIONS
-------
-i::
--input=::
Input file name. (default: perf.data)
+-d::
+--dsos=::
+ Only consider symbols in these dsos. CSV that understands
+ file://filename entries.
+-C::
+--comms=::
+ Only consider symbols in these comms. CSV that understands
+ file://filename entries.
+-S::
+--symbols=::
+ Only consider these symbols. CSV that understands
+ file://filename entries.
SEE ALSO
--------
diff --git a/tools/perf/Documentation/perf-stat.txt b/tools/perf/Documentation/perf-stat.txt
index c368a72721d..0d74346d21a 100644
--- a/tools/perf/Documentation/perf-stat.txt
+++ b/tools/perf/Documentation/perf-stat.txt
@@ -8,8 +8,8 @@ perf-stat - Run a command and gather performance counter statistics
SYNOPSIS
--------
[verse]
-'perf stat' [-e <EVENT> | --event=EVENT] [-l] [-a] <command>
-'perf stat' [-e <EVENT> | --event=EVENT] [-l] [-a] -- <command> [<options>]
+'perf stat' [-e <EVENT> | --event=EVENT] [-S] [-a] <command>
+'perf stat' [-e <EVENT> | --event=EVENT] [-S] [-a] -- <command> [<options>]
DESCRIPTION
-----------
@@ -40,7 +40,7 @@ OPTIONS
-a::
system-wide collection
--l::
+-S::
scale counter values
EXAMPLES
diff --git a/tools/perf/Makefile b/tools/perf/Makefile
index 36d7eef4991..9c6d0ae3708 100644
--- a/tools/perf/Makefile
+++ b/tools/perf/Makefile
@@ -290,7 +290,7 @@ LIB_FILE=libperf.a
LIB_H += ../../include/linux/perf_counter.h
LIB_H += perf.h
-LIB_H += types.h
+LIB_H += util/types.h
LIB_H += util/list.h
LIB_H += util/rbtree.h
LIB_H += util/levenshtein.h
@@ -301,6 +301,7 @@ LIB_H += util/util.h
LIB_H += util/help.h
LIB_H += util/strbuf.h
LIB_H += util/string.h
+LIB_H += util/strlist.h
LIB_H += util/run-command.h
LIB_H += util/sigchain.h
LIB_H += util/symbol.h
@@ -322,12 +323,15 @@ LIB_OBJS += util/run-command.o
LIB_OBJS += util/quote.o
LIB_OBJS += util/strbuf.o
LIB_OBJS += util/string.o
+LIB_OBJS += util/strlist.o
LIB_OBJS += util/usage.o
LIB_OBJS += util/wrapper.o
LIB_OBJS += util/sigchain.o
LIB_OBJS += util/symbol.o
LIB_OBJS += util/color.o
LIB_OBJS += util/pager.o
+LIB_OBJS += util/header.o
+LIB_OBJS += util/callchain.o
BUILTIN_OBJS += builtin-annotate.o
BUILTIN_OBJS += builtin-help.o
diff --git a/tools/perf/builtin-annotate.c b/tools/perf/builtin-annotate.c
index 7e58e3ad150..722c0f54e54 100644
--- a/tools/perf/builtin-annotate.c
+++ b/tools/perf/builtin-annotate.c
@@ -855,7 +855,7 @@ static unsigned long total = 0,
total_unknown = 0;
static int
-process_overflow_event(event_t *event, unsigned long offset, unsigned long head)
+process_sample_event(event_t *event, unsigned long offset, unsigned long head)
{
char level;
int show = 0;
@@ -1013,10 +1013,10 @@ process_period_event(event_t *event, unsigned long offset, unsigned long head)
static int
process_event(event_t *event, unsigned long offset, unsigned long head)
{
- if (event->header.misc & PERF_EVENT_MISC_OVERFLOW)
- return process_overflow_event(event, offset, head);
-
switch (event->header.type) {
+ case PERF_EVENT_SAMPLE:
+ return process_sample_event(event, offset, head);
+
case PERF_EVENT_MMAP:
return process_mmap_event(event, offset, head);
diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c
index d7ebbd75754..d18546f37d7 100644
--- a/tools/perf/builtin-record.c
+++ b/tools/perf/builtin-record.c
@@ -14,6 +14,8 @@
#include "util/parse-events.h"
#include "util/string.h"
+#include "util/header.h"
+
#include <unistd.h>
#include <sched.h>
@@ -39,6 +41,8 @@ static int force = 0;
static int append_file = 0;
static int call_graph = 0;
static int verbose = 0;
+static int inherit_stat = 0;
+static int no_samples = 0;
static long samples;
static struct timeval last_read;
@@ -52,7 +56,8 @@ static int nr_poll;
static int nr_cpu;
static int file_new = 1;
-static struct perf_file_header file_header;
+
+struct perf_header *header;
struct mmap_event {
struct perf_event_header header;
@@ -306,12 +311,11 @@ static void pid_synthesize_mmap_samples(pid_t pid)
continue;
pbf += n + 3;
if (*pbf == 'x') { /* vm_exec */
- char *execname = strrchr(bf, ' ');
+ char *execname = strchr(bf, '/');
- if (execname == NULL || execname[1] != '/')
+ if (execname == NULL)
continue;
- execname += 1;
size = strlen(execname);
execname[size - 1] = '\0'; /* Remove \n */
memcpy(mmap_ev.filename, execname, size);
@@ -329,7 +333,7 @@ static void pid_synthesize_mmap_samples(pid_t pid)
fclose(fp);
}
-static void synthesize_samples(void)
+static void synthesize_all(void)
{
DIR *proc;
struct dirent dirent, *next;
@@ -353,10 +357,35 @@ static void synthesize_samples(void)
static int group_fd;
+static struct perf_header_attr *get_header_attr(struct perf_counter_attr *a, int nr)
+{
+ struct perf_header_attr *h_attr;
+
+ if (nr < header->attrs) {
+ h_attr = header->attr[nr];
+ } else {
+ h_attr = perf_header_attr__new(a);
+ perf_header__add_attr(header, h_attr);
+ }
+
+ return h_attr;
+}
+
static void create_counter(int counter, int cpu, pid_t pid)
{
struct perf_counter_attr *attr = attrs + counter;
- int track = 1;
+ struct perf_header_attr *h_attr;
+ int track = !counter; /* only the first counter needs these */
+ struct {
+ u64 count;
+ u64 time_enabled;
+ u64 time_running;
+ u64 id;
+ } read_data;
+
+ attr->read_format = PERF_FORMAT_TOTAL_TIME_ENABLED |
+ PERF_FORMAT_TOTAL_TIME_RUNNING |
+ PERF_FORMAT_ID;
attr->sample_type = PERF_SAMPLE_IP | PERF_SAMPLE_TID;
@@ -366,25 +395,20 @@ static void create_counter(int counter, int cpu, pid_t pid)
attr->sample_freq = freq;
}
+ if (no_samples)
+ attr->sample_freq = 0;
+
+ if (inherit_stat)
+ attr->inherit_stat = 1;
+
if (call_graph)
attr->sample_type |= PERF_SAMPLE_CALLCHAIN;
- if (file_new) {
- file_header.sample_type = attr->sample_type;
- } else {
- if (file_header.sample_type != attr->sample_type) {
- fprintf(stderr, "incompatible append\n");
- exit(-1);
- }
- }
-
attr->mmap = track;
attr->comm = track;
attr->inherit = (cpu < 0) && inherit;
attr->disabled = 1;
- track = 0; /* only the first counter needs these */
-
try_again:
fd[nr_cpu][counter] = sys_perf_counter_open(attr, pid, cpu, group_fd, 0);
@@ -415,6 +439,22 @@ try_again:
exit(-1);
}
+ h_attr = get_header_attr(attr, counter);
+
+ if (!file_new) {
+ if (memcmp(&h_attr->attr, attr, sizeof(*attr))) {
+ fprintf(stderr, "incompatible append\n");
+ exit(-1);
+ }
+ }
+
+ if (read(fd[nr_cpu][counter], &read_data, sizeof(read_data)) == -1) {
+ perror("Unable to read perf file descriptor\n");
+ exit(-1);
+ }
+
+ perf_header_attr__add_id(h_attr, read_data.id);
+
assert(fd[nr_cpu][counter] >= 0);
fcntl(fd[nr_cpu][counter], F_SETFL, O_NONBLOCK);
@@ -445,11 +485,6 @@ static void open_counters(int cpu, pid_t pid)
{
int counter;
- if (pid > 0) {
- pid_synthesize_comm_event(pid, 0);
- pid_synthesize_mmap_samples(pid);
- }
-
group_fd = -1;
for (counter = 0; counter < nr_counters; counter++)
create_counter(counter, cpu, pid);
@@ -459,17 +494,16 @@ static void open_counters(int cpu, pid_t pid)
static void atexit_header(void)
{
- file_header.data_size += bytes_written;
+ header->data_size += bytes_written;
- if (pwrite(output, &file_header, sizeof(file_header), 0) == -1)
- perror("failed to write on file headers");
+ perf_header__write(header, output);
}
static int __cmd_record(int argc, const char **argv)
{
int i, counter;
struct stat st;
- pid_t pid;
+ pid_t pid = 0;
int flags;
int ret;
@@ -500,22 +534,31 @@ static int __cmd_record(int argc, const char **argv)
exit(-1);
}
- if (!file_new) {
- if (read(output, &file_header, sizeof(file_header)) == -1) {
- perror("failed to read file headers");
- exit(-1);
- }
-
- lseek(output, file_header.data_size, SEEK_CUR);
- }
+ if (!file_new)
+ header = perf_header__read(output);
+ else
+ header = perf_header__new();
atexit(atexit_header);
if (!system_wide) {
- open_counters(-1, target_pid != -1 ? target_pid : getpid());
+ pid = target_pid;
+ if (pid == -1)
+ pid = getpid();
+
+ open_counters(-1, pid);
} else for (i = 0; i < nr_cpus; i++)
open_counters(i, target_pid);
+ if (file_new)
+ perf_header__write(header, output);
+
+ if (!system_wide) {
+ pid_synthesize_comm_event(pid, 0);
+ pid_synthesize_mmap_samples(pid);
+ } else
+ synthesize_all();
+
if (target_pid == -1 && argc) {
pid = fork();
if (pid < 0)
@@ -539,10 +582,7 @@ static int __cmd_record(int argc, const char **argv)
}
}
- if (system_wide)
- synthesize_samples();
-
- while (!done) {
+ for (;;) {
int hits = samples;
for (i = 0; i < nr_cpu; i++) {
@@ -550,8 +590,11 @@ static int __cmd_record(int argc, const char **argv)
mmap_read(&mmap_array[i][counter]);
}
- if (hits == samples)
+ if (hits == samples) {
+ if (done)
+ break;
ret = poll(event_array, nr_poll, 100);
+ }
}
/*
@@ -600,6 +643,10 @@ static const struct option options[] = {
"do call-graph (stack chain/backtrace) recording"),
OPT_BOOLEAN('v', "verbose", &verbose,
"be more verbose (show counter open errors, etc)"),
+ OPT_BOOLEAN('s', "stat", &inherit_stat,
+ "per thread counts"),
+ OPT_BOOLEAN('n', "no-samples", &no_samples,
+ "don't sample"),
OPT_END()
};
diff --git a/tools/perf/builtin-report.c b/tools/perf/builtin-report.c
index 5eb5566f0c9..135b7837e6b 100644
--- a/tools/perf/builtin-report.c
+++ b/tools/perf/builtin-report.c
@@ -15,8 +15,11 @@
#include "util/rbtree.h"
#include "util/symbol.h"
#include "util/string.h"
+#include "util/callchain.h"
+#include "util/strlist.h"
#include "perf.h"
+#include "util/header.h"
#include "util/parse-options.h"
#include "util/parse-events.h"
@@ -30,6 +33,8 @@ static char *vmlinux = NULL;
static char default_sort_order[] = "comm,dso";
static char *sort_order = default_sort_order;
+static char *dso_list_str, *comm_list_str, *sym_list_str;
+static struct strlist *dso_list, *comm_list, *sym_list;
static int input;
static int show_mask = SHOW_KERNEL | SHOW_USER | SHOW_HV;
@@ -51,6 +56,9 @@ static char *parent_pattern = default_parent_pattern;
static regex_t parent_regex;
static int exclude_other = 1;
+static int callchain;
+
+static u64 sample_type;
struct ip_event {
struct perf_event_header header;
@@ -59,11 +67,6 @@ struct ip_event {
unsigned char __more_data[];
};
-struct ip_callchain {
- u64 nr;
- u64 ips[0];
-};
-
struct mmap_event {
struct perf_event_header header;
u32 pid, tid;
@@ -97,6 +100,13 @@ struct lost_event {
u64 lost;
};
+struct read_event {
+ struct perf_event_header header;
+ u32 pid,tid;
+ u64 value;
+ u64 format[3];
+};
+
typedef union event_union {
struct perf_event_header header;
struct ip_event ip;
@@ -105,6 +115,7 @@ typedef union event_union {
struct fork_event fork;
struct period_event period;
struct lost_event lost;
+ struct read_event read;
} event_t;
static LIST_HEAD(dsos);
@@ -229,7 +240,7 @@ static u64 vdso__map_ip(struct map *map, u64 ip)
static inline int is_anon_memory(const char *filename)
{
- return strcmp(filename, "//anon") == 0;
+ return strcmp(filename, "//anon") == 0;
}
static struct map *map__new(struct mmap_event *event)
@@ -400,9 +411,27 @@ static void thread__insert_map(struct thread *self, struct map *map)
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);
+ if (verbose >= 2) {
+ printf("overlapping maps:\n");
+ map__fprintf(map, stdout);
+ map__fprintf(pos, stdout);
+ }
+
+ if (map->start <= pos->start && map->end > pos->start)
+ pos->start = map->end;
+
+ if (map->end >= pos->end && map->start < pos->end)
+ pos->end = map->start;
+
+ if (verbose >= 2) {
+ printf("after collision:\n");
+ map__fprintf(pos, stdout);
+ }
+
+ if (pos->start >= pos->end) {
+ list_del_init(&pos->node);
+ free(pos);
+ }
}
}
@@ -464,17 +493,19 @@ static size_t threads__fprintf(FILE *fp)
static struct rb_root hist;
struct hist_entry {
- struct rb_node rb_node;
-
- struct thread *thread;
- struct map *map;
- struct dso *dso;
- struct symbol *sym;
- struct symbol *parent;
- u64 ip;
- char level;
-
- u64 count;
+ struct rb_node rb_node;
+
+ struct thread *thread;
+ struct map *map;
+ struct dso *dso;
+ struct symbol *sym;
+ struct symbol *parent;
+ u64 ip;
+ char level;
+ struct callchain_node callchain;
+ struct rb_root sorted_chain;
+
+ u64 count;
};
/*
@@ -745,6 +776,48 @@ hist_entry__collapse(struct hist_entry *left, struct hist_entry *right)
}
static size_t
+callchain__fprintf(FILE *fp, struct callchain_node *self, u64 total_samples)
+{
+ struct callchain_list *chain;
+ size_t ret = 0;
+
+ if (!self)
+ return 0;
+
+ ret += callchain__fprintf(fp, self->parent, total_samples);
+
+
+ list_for_each_entry(chain, &self->val, list)
+ ret += fprintf(fp, " %p\n", (void *)chain->ip);
+
+ return ret;
+}
+
+static size_t
+hist_entry_callchain__fprintf(FILE *fp, struct hist_entry *self,
+ u64 total_samples)
+{
+ struct rb_node *rb_node;
+ struct callchain_node *chain;
+ size_t ret = 0;
+
+ rb_node = rb_first(&self->sorted_chain);
+ while (rb_node) {
+ double percent;
+
+ chain = rb_entry(rb_node, struct callchain_node, rb_node);
+ percent = chain->hit * 100.0 / total_samples;
+ ret += fprintf(fp, " %6.2f%%\n", percent);
+ ret += callchain__fprintf(fp, chain, total_samples);
+ ret += fprintf(fp, "\n");
+ rb_node = rb_next(rb_node);
+ }
+
+ return ret;
+}
+
+
+static size_t
hist_entry__fprintf(FILE *fp, struct hist_entry *self, u64 total_samples)
{
struct sort_entry *se;
@@ -784,6 +857,9 @@ hist_entry__fprintf(FILE *fp, struct hist_entry *self, u64 total_samples)
ret += fprintf(fp, "\n");
+ if (callchain)
+ hist_entry_callchain__fprintf(fp, self, total_samples);
+
return ret;
}
@@ -797,7 +873,7 @@ resolve_symbol(struct thread *thread, struct map **mapp,
{
struct dso *dso = dsop ? *dsop : NULL;
struct map *map = mapp ? *mapp : NULL;
- uint64_t ip = *ipp;
+ u64 ip = *ipp;
if (!thread)
return NULL;
@@ -814,7 +890,6 @@ resolve_symbol(struct thread *thread, struct map **mapp,
*mapp = map;
got_map:
ip = map->map_ip(map, ip);
- *ipp = ip;
dso = map->dso;
} else {
@@ -828,6 +903,8 @@ got_map:
dso = kernel_dso;
}
dprintf(" ...... dso: %s\n", dso ? dso->name : "<not found>");
+ dprintf(" ...... map: %Lx -> %Lx\n", *ipp, ip);
+ *ipp = ip;
if (dsop)
*dsop = dso;
@@ -867,6 +944,7 @@ hist_entry__add(struct thread *thread, struct map *map, struct dso *dso,
.level = level,
.count = count,
.parent = NULL,
+ .sorted_chain = RB_ROOT
};
int cmp;
@@ -909,6 +987,8 @@ hist_entry__add(struct thread *thread, struct map *map, struct dso *dso,
if (!cmp) {
he->count += count;
+ if (callchain)
+ append_chain(&he->callchain, chain);
return 0;
}
@@ -922,6 +1002,10 @@ hist_entry__add(struct thread *thread, struct map *map, struct dso *dso,
if (!he)
return -ENOMEM;
*he = entry;
+ if (callchain) {
+ callchain_init(&he->callchain);
+ append_chain(&he->callchain, chain);
+ }
rb_link_node(&he->rb_node, parent, p);
rb_insert_color(&he->rb_node, &hist);
@@ -998,6 +1082,9 @@ static void output__insert_entry(struct hist_entry *he)
struct rb_node *parent = NULL;
struct hist_entry *iter;
+ if (callchain)
+ sort_chain_to_rbtree(&he->sorted_chain, &he->callchain);
+
while (*p != NULL) {
parent = *p;
iter = rb_entry(parent, struct hist_entry, rb_node);
@@ -1115,7 +1202,7 @@ static int validate_chain(struct ip_callchain *chain, event_t *event)
}
static int
-process_overflow_event(event_t *event, unsigned long offset, unsigned long head)
+process_sample_event(event_t *event, unsigned long offset, unsigned long head)
{
char level;
int show = 0;
@@ -1127,12 +1214,12 @@ process_overflow_event(event_t *event, unsigned long offset, unsigned long head)
void *more_data = event->ip.__more_data;
struct ip_callchain *chain = NULL;
- if (event->header.type & PERF_SAMPLE_PERIOD) {
+ if (sample_type & PERF_SAMPLE_PERIOD) {
period = *(u64 *)more_data;
more_data += sizeof(u64);
}
- dprintf("%p [%p]: PERF_EVENT (IP, %d): %d: %p period: %Ld\n",
+ dprintf("%p [%p]: PERF_EVENT_SAMPLE (IP, %d): %d: %p period: %Ld\n",
(void *)(offset + head),
(void *)(long)(event->header.size),
event->header.misc,
@@ -1140,7 +1227,7 @@ process_overflow_event(event_t *event, unsigned long offset, unsigned long head)
(void *)(long)ip,
(long long)period);
- if (event->header.type & PERF_SAMPLE_CALLCHAIN) {
+ if (sample_type & PERF_SAMPLE_CALLCHAIN) {
int i;
chain = (void *)more_data;
@@ -1166,6 +1253,9 @@ process_overflow_event(event_t *event, unsigned long offset, unsigned long head)
return -1;
}
+ if (comm_list && !strlist__has_entry(comm_list, thread->comm))
+ return 0;
+
if (event->header.misc & PERF_EVENT_MISC_KERNEL) {
show = SHOW_KERNEL;
level = 'k';
@@ -1188,6 +1278,12 @@ process_overflow_event(event_t *event, unsigned long offset, unsigned long head)
if (show & show_mask) {
struct symbol *sym = resolve_symbol(thread, &map, &dso, &ip);
+ if (dso_list && dso && dso->name && !strlist__has_entry(dso_list, dso->name))
+ return 0;
+
+ if (sym_list && sym && !strlist__has_entry(sym_list, sym->name))
+ return 0;
+
if (hist_entry__add(thread, map, dso, sym, ip, chain, level, period)) {
eprintf("problem incrementing symbol count, skipping event\n");
return -1;
@@ -1328,14 +1424,27 @@ static void trace_event(event_t *event)
}
static int
+process_read_event(event_t *event, unsigned long offset, unsigned long head)
+{
+ dprintf("%p [%p]: PERF_EVENT_READ: %d %d %Lu\n",
+ (void *)(offset + head),
+ (void *)(long)(event->header.size),
+ event->read.pid,
+ event->read.tid,
+ event->read.value);
+
+ return 0;
+}
+
+static int
process_event(event_t *event, unsigned long offset, unsigned long head)
{
trace_event(event);
- if (event->header.misc & PERF_EVENT_MISC_OVERFLOW)
- return process_overflow_event(event, offset, head);
-
switch (event->header.type) {
+ case PERF_EVENT_SAMPLE:
+ return process_sample_event(event, offset, head);
+
case PERF_EVENT_MMAP:
return process_mmap_event(event, offset, head);
@@ -1351,6 +1460,9 @@ process_event(event_t *event, unsigned long offset, unsigned long head)
case PERF_EVENT_LOST:
return process_lost_event(event, offset, head);
+ case PERF_EVENT_READ:
+ return process_read_event(event, offset, head);
+
/*
* We dont process them right now but they are fine:
*/
@@ -1366,13 +1478,30 @@ process_event(event_t *event, unsigned long offset, unsigned long head)
return 0;
}
-static struct perf_file_header file_header;
+static struct perf_header *header;
+
+static u64 perf_header__sample_type(void)
+{
+ u64 sample_type = 0;
+ int i;
+
+ for (i = 0; i < header->attrs; i++) {
+ struct perf_header_attr *attr = header->attr[i];
+
+ if (!sample_type)
+ sample_type = attr->attr.sample_type;
+ else if (sample_type != attr->attr.sample_type)
+ die("non matching sample_type");
+ }
+
+ return sample_type;
+}
static int __cmd_report(void)
{
int ret, rc = EXIT_FAILURE;
unsigned long offset = 0;
- unsigned long head = sizeof(file_header);
+ unsigned long head, shift;
struct stat stat;
event_t *event;
uint32_t size;
@@ -1400,13 +1529,12 @@ static int __cmd_report(void)
exit(0);
}
- if (read(input, &file_header, sizeof(file_header)) == -1) {
- perror("failed to read file headers");
- exit(-1);
- }
+ header = perf_header__read(input);
+ head = header->data_offset;
- if (sort__has_parent &&
- !(file_header.sample_type & PERF_SAMPLE_CALLCHAIN)) {
+ sample_type = perf_header__sample_type();
+
+ if (sort__has_parent && !(sample_type & PERF_SAMPLE_CALLCHAIN)) {
fprintf(stderr, "selected --sort parent, but no callchain data\n");
exit(-1);
}
@@ -1426,6 +1554,11 @@ static int __cmd_report(void)
cwd = NULL;
cwdlen = 0;
}
+
+ shift = page_size * (head / page_size);
+ offset += shift;
+ head -= shift;
+
remap:
buf = (char *)mmap(NULL, page_size * mmap_window, PROT_READ,
MAP_SHARED, input, offset);
@@ -1442,9 +1575,10 @@ more:
size = 8;
if (head + event->header.size >= page_size * mmap_window) {
- unsigned long shift = page_size * (head / page_size);
int ret;
+ shift = page_size * (head / page_size);
+
ret = munmap(buf, page_size * mmap_window);
assert(ret == 0);
@@ -1482,7 +1616,7 @@ more:
head += size;
- if (offset + head >= sizeof(file_header) + file_header.data_size)
+ if (offset + head >= header->data_offset + header->data_size)
goto done;
if (offset + head < stat.st_size)
@@ -1536,6 +1670,13 @@ static const struct option options[] = {
"regex filter to identify parent, see: '--sort parent'"),
OPT_BOOLEAN('x', "exclude-other", &exclude_other,
"Only display entries with parent-match"),
+ OPT_BOOLEAN('c', "callchain", &callchain, "Display callchains"),
+ OPT_STRING('d', "dsos", &dso_list_str, "dso[,dso...]",
+ "only consider symbols in these dsos"),
+ OPT_STRING('C', "comms", &comm_list_str, "comm[,comm...]",
+ "only consider symbols in these comms"),
+ OPT_STRING('S', "symbols", &sym_list_str, "symbol[,symbol...]",
+ "only consider these symbols"),
OPT_END()
};
@@ -1554,6 +1695,19 @@ static void setup_sorting(void)
free(str);
}
+static void setup_list(struct strlist **list, const char *list_str,
+ const char *list_name)
+{
+ if (list_str) {
+ *list = strlist__new(true, list_str);
+ if (!*list) {
+ fprintf(stderr, "problems parsing %s list\n",
+ list_name);
+ exit(129);
+ }
+ }
+}
+
int cmd_report(int argc, const char **argv, const char *prefix)
{
symbol__init();
@@ -1575,6 +1729,10 @@ int cmd_report(int argc, const char **argv, const char *prefix)
if (argc)
usage_with_options(report_usage, options);
+ setup_list(&dso_list, dso_list_str, "dso");
+ setup_list(&comm_list, comm_list_str, "comm");
+ setup_list(&sym_list, sym_list_str, "symbol");
+
setup_pager();
return __cmd_report();
diff --git a/tools/perf/builtin-stat.c b/tools/perf/builtin-stat.c
index 6d3eeac1ea2..2e03524a1de 100644
--- a/tools/perf/builtin-stat.c
+++ b/tools/perf/builtin-stat.c
@@ -32,6 +32,7 @@
* Wu Fengguang <fengguang.wu@intel.com>
* Mike Galbraith <efault@gmx.de>
* Paul Mackerras <paulus@samba.org>
+ * Jaswinder Singh Rajput <jaswinder@kernel.org>
*
* Released under the GPL v2. (and only v2, not any later version)
*/
@@ -45,7 +46,7 @@
#include <sys/prctl.h>
#include <math.h>
-static struct perf_counter_attr default_attrs[MAX_COUNTERS] = {
+static struct perf_counter_attr default_attrs[] = {
{ .type = PERF_TYPE_SOFTWARE, .config = PERF_COUNT_SW_TASK_CLOCK },
{ .type = PERF_TYPE_SOFTWARE, .config = PERF_COUNT_SW_CONTEXT_SWITCHES},
@@ -59,42 +60,28 @@ static struct perf_counter_attr default_attrs[MAX_COUNTERS] = {
};
+#define MAX_RUN 100
+
static int system_wide = 0;
-static int inherit = 1;
static int verbose = 0;
-
-static int fd[MAX_NR_CPUS][MAX_COUNTERS];
-
-static int target_pid = -1;
static int nr_cpus = 0;
-static unsigned int page_size;
+static int run_idx = 0;
+static int run_count = 1;
+static int inherit = 1;
static int scale = 1;
+static int target_pid = -1;
+static int null_run = 0;
-static const unsigned int default_count[] = {
- 1000000,
- 1000000,
- 10000,
- 10000,
- 1000000,
- 10000,
-};
-
-#define MAX_RUN 100
-
-static int run_count = 1;
-static int run_idx = 0;
-
-static u64 event_res[MAX_RUN][MAX_COUNTERS][3];
-static u64 event_scaled[MAX_RUN][MAX_COUNTERS];
-
-//static u64 event_hist[MAX_RUN][MAX_COUNTERS][3];
-
+static int fd[MAX_NR_CPUS][MAX_COUNTERS];
static u64 runtime_nsecs[MAX_RUN];
static u64 walltime_nsecs[MAX_RUN];
static u64 runtime_cycles[MAX_RUN];
+static u64 event_res[MAX_RUN][MAX_COUNTERS][3];
+static u64 event_scaled[MAX_RUN][MAX_COUNTERS];
+
static u64 event_res_avg[MAX_COUNTERS][3];
static u64 event_res_noise[MAX_COUNTERS][3];
@@ -109,7 +96,10 @@ static u64 walltime_nsecs_noise;
static u64 runtime_cycles_avg;
static u64 runtime_cycles_noise;
-static void create_perf_stat_counter(int counter)
+#define ERR_PERF_OPEN \
+"Error: counter %d, sys_perf