[PATCH 1/2] perf tools: Add a per tracepoint counter attribute to get raw sample

[PATCH 1/2] perf tools: Add a per tracepoint counter attribute to get raw sample

[Frederic Weisbecker]

Add a new flag field while opening a tracepoint perf counter:

	-e tracepoint_subsystem:tracepoint_name:flags

This is intended to be generic although for now it only supports the
r[e[c[o[r[d]]]]] flag:

	./perf record -e workqueue:workqueue_insertion:record
	./perf record -e workqueue:workqueue_insertion:r

will have the same effect: enabling the raw samples record for the
given tracepoint counter.

In the future, we may want to support further flags, separated by
commas.

Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Mike Galbraith <efault@gmx.de>
---
 tools/perf/builtin-record.c    |    2 +-
 tools/perf/util/parse-events.c |   10 ++++++++++
 2 files changed, 11 insertions(+), 1 deletions(-)

diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c
index 5aeb632..3be0301 100644
--- a/tools/perf/builtin-record.c
+++ b/tools/perf/builtin-record.c
@@ -380,7 +380,7 @@ static void create_counter(int counter, int cpu, pid_t pid)
 				  PERF_FORMAT_TOTAL_TIME_RUNNING |
 				  PERF_FORMAT_ID;
 
-	attr->sample_type	= PERF_SAMPLE_IP | PERF_SAMPLE_TID;
+	attr->sample_type	|= PERF_SAMPLE_IP | PERF_SAMPLE_TID;
 
 	if (freq) {
 		attr->sample_type	|= PERF_SAMPLE_PERIOD;
diff --git a/tools/perf/util/parse-events.c b/tools/perf/util/parse-events.c
index a5d661b..4bab278 100644
--- a/tools/perf/util/parse-events.c
+++ b/tools/perf/util/parse-events.c
@@ -379,6 +379,7 @@ static int parse_tracepoint_event(const char **strp,
 				    struct perf_counter_attr *attr)
 {
 	const char *evt_name;
+	char *flags;
 	char sys_name[MAX_EVENT_LENGTH];
 	char id_buf[4];
 	int fd;
@@ -400,6 +401,15 @@ static int parse_tracepoint_event(const char **strp,
 	strncpy(sys_name, *strp, sys_length);
 	sys_name[sys_length] = '\0';
 	evt_name = evt_name + 1;
+
+	flags = strchr(evt_name, ':');
+	if (flags) {
+		*flags = '\0';
+		flags++;
+		if (!strncmp(flags, "record", strlen(flags)))
+			attr->sample_type |= PERF_SAMPLE_RAW;
+	}
+
 	evt_length = strlen(evt_name);
 	if (evt_length >= MAX_EVENT_LENGTH)
 		return 0;
-- 
1.6.2.3

[PATCH] perf tools: Factorize the thread code in a dedicated file

[Frederic Weisbecker]

Factorize the thread management code used by perf-annotate and
perf-report in dedicated source and header files.

Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Mike Galbraith <efault@gmx.de>
---
 tools/perf/Makefile           |    1 +
 tools/perf/builtin-annotate.c |  173 ++++-------------------------------
 tools/perf/builtin-report.c   |  205 ++++------------------------------------
 tools/perf/util/thread.c      |  143 ++++++++++++++++++++++++++++
 tools/perf/util/thread.h      |   19 ++++
 5 files changed, 202 insertions(+), 339 deletions(-)
 create mode 100644 tools/perf/util/thread.c
 create mode 100644 tools/perf/util/thread.h

diff --git a/tools/perf/Makefile b/tools/perf/Makefile
index 68218cf..0056405 100644
--- a/tools/perf/Makefile
+++ b/tools/perf/Makefile
@@ -341,6 +341,7 @@ 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
 
 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 543c452..dacf0da 100644
--- a/tools/perf/builtin-annotate.c
+++ b/tools/perf/builtin-annotate.c
@@ -20,6 +20,7 @@
 
 #include "util/parse-options.h"
 #include "util/parse-events.h"
+#include "util/thread.h"
 
 #define SHOW_KERNEL	1
 #define SHOW_USER	2
@@ -44,6 +45,9 @@ static int		print_line;
 static unsigned long	page_size;
 static unsigned long	mmap_window = 32;
 
+static struct rb_root	threads;
+static struct thread	*last_match;
+
 
 struct sym_ext {
 	struct rb_node	node;
@@ -51,154 +55,6 @@ struct sym_ext {
 	char		*path;
 };
 
-
-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
  */
@@ -624,7 +480,7 @@ static void output__resort(void)
 
 static void register_idle_thread(void)
 {
-	struct thread *thread = threads__findnew(0);
+	struct thread *thread = threads__findnew(0, &threads, last_match);
 
 	if (thread == NULL ||
 			thread__set_comm(thread, "[idle]")) {
@@ -645,10 +501,12 @@ 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;
 
+	thread = threads__findnew(event->ip.pid, &threads, last_match);
+
 	dprintf("%p [%p]: PERF_EVENT (IP, %d): %d: %p\n",
 		(void *)(offset + head),
 		(void *)(long)(event->header.size),
@@ -719,9 +577,11 @@ 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 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",
 		(void *)(offset + head),
 		(void *)(long)(event->header.size),
@@ -745,8 +605,9 @@ process_mmap_event(event_t *event, unsigned long offset, unsigned long head)
 static int
 process_comm_event(event_t *event, unsigned long offset, unsigned long head)
 {
-	struct thread *thread = threads__findnew(event->comm.pid);
+	struct thread *thread;
 
+	thread = threads__findnew(event->comm.pid, &threads, last_match);
 	dprintf("%p [%p]: PERF_EVENT_COMM: %s:%d\n",
 		(void *)(offset + head),
 		(void *)(long)(event->header.size),
@@ -765,9 +626,11 @@ process_comm_event(event_t *event, unsigned long offset, unsigned long head)
 static int
 process_fork_event(event_t *event, unsigned long offset, unsigned long head)
 {
-	struct thread *thread = threads__findnew(event->fork.pid);
-	struct thread *parent = threads__findnew(event->fork.ppid);
+	struct thread *thread;
+	struct thread *parent;
 
+	thread = threads__findnew(event->fork.pid, &threads, last_match);
+	parent = threads__findnew(event->fork.ppid, &threads, last_match);
 	dprintf("%p [%p]: PERF_EVENT_FORK: %d:%d\n",
 		(void *)(offset + head),
 		(void *)(long)(event->header.size),
@@ -1202,7 +1065,7 @@ more:
 		return 0;
 
 	if (verbose >= 3)
-		threads__fprintf(stdout);
+		threads__fprintf(stdout, &threads);
 
 	if (verbose >= 2)
 		dsos__fprintf(stdout);
diff --git a/tools/perf/builtin-report.c b/tools/perf/builtin-report.c
index 6321951..c91ea34 100644
--- a/tools/perf/builtin-report.c
+++ b/tools/perf/builtin-report.c
@@ -25,6 +25,8 @@
 #include "util/parse-options.h"
 #include "util/parse-events.h"
 
+#include "util/thread.h"
+
 #define SHOW_KERNEL	1
 #define SHOW_USER	2
 #define SHOW_HV		4
@@ -71,6 +73,9 @@ static char		__cwd[PATH_MAX];
 static char		*cwd = __cwd;
 static int		cwdlen;
 
+static struct rb_root	threads;
+static struct thread	*last_match;
+
 static
 struct callchain_param	callchain_param = {
 	.mode	= CHAIN_GRAPH_REL,
@@ -106,187 +111,10 @@ static int repsep_fprintf(FILE *fp, const char *fmt, ...)
 	return n;
 }
 
-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 unsigned int dsos__col_width,
 		    comms__col_width,
 		    threads__col_width;
 
-static int thread__set_comm(struct thread *self, const char *comm)
-{
-	if (self->comm)
-		free(self->comm);
-	self->comm = strdup(comm);
-	if (!self->comm)
-		return -ENOMEM;
-
-	if (!col_width_list_str && !field_sep &&
-	    (!comm_list || strlist__has_entry(comm_list, comm))) {
-		unsigned int slen = strlen(comm);
-		if (slen > comms__col_width) {
-			comms__col_width = slen;
-			threads__col_width = slen + 6;
-		}
-	}
-
-	return 0;
-}
-
-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)) {
-			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);
-			}
-		}
-	}
-
-	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
  */
@@ -1228,7 +1056,7 @@ print_entries:
 
 static void register_idle_thread(void)
 {
-	struct thread *thread = threads__findnew(0);
+	struct thread *thread = threads__findnew(0, &threads, last_match);
 
 	if (thread == NULL ||
 			thread__set_comm(thread, "[idle]")) {
@@ -1263,7 +1091,7 @@ 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;
 	u64 period = 1;
 	struct map *map = NULL;
@@ -1271,6 +1099,8 @@ process_sample_event(event_t *event, unsigned long offset, unsigned long head)
 	struct ip_callchain *chain = NULL;
 	int cpumode;
 
+	thread = threads__findnew(event->ip.pid, &threads, last_match);
+
 	if (sample_type & PERF_SAMPLE_PERIOD) {
 		period = *(u64 *)more_data;
 		more_data += sizeof(u64);
@@ -1360,9 +1190,11 @@ 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 thread *thread;
 	struct map *map = map__new(&event->mmap, cwd, cwdlen);
 
+	thread = threads__findnew(event->mmap.pid, &threads, last_match);
+
 	dprintf("%p [%p]: PERF_EVENT_MMAP %d/%d: [%p(%p) @ %p]: %s\n",
 		(void *)(offset + head),
 		(void *)(long)(event->header.size),
@@ -1387,7 +1219,9 @@ process_mmap_event(event_t *event, unsigned long offset, unsigned long head)
 static int
 process_comm_event(event_t *event, unsigned long offset, unsigned long head)
 {
-	struct thread *thread = threads__findnew(event->comm.pid);
+	struct thread *thread;
+
+	thread = threads__findnew(event->comm.pid, &threads, last_match);
 
 	dprintf("%p [%p]: PERF_EVENT_COMM: %s:%d\n",
 		(void *)(offset + head),
@@ -1407,8 +1241,11 @@ process_comm_event(event_t *event, unsigned long offset, unsigned long head)
 static int
 process_task_event(event_t *event, unsigned long offset, unsigned long head)
 {
-	struct thread *thread = threads__findnew(event->fork.pid);
-	struct thread *parent = threads__findnew(event->fork.ppid);
+	struct thread *thread;
+	struct thread *parent;
+
+	thread = threads__findnew(event->fork.pid, &threads, last_match);
+	parent = threads__findnew(event->fork.ppid, &threads, last_match);
 
 	dprintf("%p [%p]: PERF_EVENT_%s: (%d:%d):(%d:%d)\n",
 		(void *)(offset + head),
@@ -1749,7 +1586,7 @@ done:
 		return 0;
 
 	if (verbose >= 3)
-		threads__fprintf(stdout);
+		threads__fprintf(stdout, &threads);
 
 	if (verbose >= 2)
 		dsos__fprintf(stdout);
diff --git a/tools/perf/util/thread.c b/tools/perf/util/thread.c
new file mode 100644
index 0000000..f2dd40e
--- /dev/null
+++ b/tools/perf/util/thread.c
@@ -0,0 +1,143 @@
+#include "../perf.h"
+#include <stdlib.h>
+#include <stdio.h>
+#include <string.h>
+#include "thread.h"
+#include "util.h"
+
+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;
+}
+
+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;
+}
+
+struct thread *
+threads__findnew(pid_t pid, struct rb_root *threads, struct thread *last_match)
+{
+	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;
+}
+
+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);
+}
+
+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;
+}
+
+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;
+}
+
+size_t threads__fprintf(FILE *fp, struct rb_root *threads)
+{
+	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;
+}
diff --git a/tools/perf/util/thread.h b/tools/perf/util/thread.h
new file mode 100644
index 0000000..d5edcba
--- /dev/null
+++ b/tools/perf/util/thread.h
@@ -0,0 +1,19 @@
+#include <linux/rbtree.h>
+#include <linux/list.h>
+#include <unistd.h>
+#include "symbol.h"
+
+struct thread {
+	struct rb_node	 rb_node;
+	struct list_head maps;
+	pid_t		 pid;
+	char		 *comm;
+};
+
+int thread__set_comm(struct thread *self, const char *comm);
+struct thread *
+threads__findnew(pid_t pid, struct rb_root *threads, struct thread *last_match);
+void thread__insert_map(struct thread *self, struct map *map);
+int thread__fork(struct thread *self, struct thread *parent);
+struct map *thread__find_map(struct thread *self, u64 ip);
+size_t threads__fprintf(FILE *fp, struct rb_root *threads);
-- 
1.6.2.3

[PATCH v2] perf tools: Factorize the thread code in a dedicated file

[Frederic Weisbecker]

Factorize the thread management code used by perf-annotate and
perf-report in dedicated source and header files.

v2: pass last_match by address so that it can actually be modified.

Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Mike Galbraith <efault@gmx.de>
---
 tools/perf/Makefile           |    1 +
 tools/perf/builtin-annotate.c |  173 ++++-------------------------------
 tools/perf/builtin-report.c   |  205 ++++------------------------------------
 tools/perf/util/thread.c      |  143 ++++++++++++++++++++++++++++
 tools/perf/util/thread.h      |   19 ++++
 5 files changed, 202 insertions(+), 339 deletions(-)
 create mode 100644 tools/perf/util/thread.c
 create mode 100644 tools/perf/util/thread.h

diff --git a/tools/perf/Makefile b/tools/perf/Makefile
index 68218cf..0056405 100644
--- a/tools/perf/Makefile
+++ b/tools/perf/Makefile
@@ -341,6 +341,7 @@ 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
 
 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 543c452..3bedaa5 100644
--- a/tools/perf/builtin-annotate.c
+++ b/tools/perf/builtin-annotate.c
@@ -20,6 +20,7 @@
 
 #include "util/parse-options.h"
 #include "util/parse-events.h"
+#include "util/thread.h"
 
 #define SHOW_KERNEL	1
 #define SHOW_USER	2
@@ -44,6 +45,9 @@ static int		print_line;
 static unsigned long	page_size;
 static unsigned long	mmap_window = 32;
 
+static struct rb_root	threads;
+static struct thread	*last_match;
+
 
 struct sym_ext {
 	struct rb_node	node;
@@ -51,154 +55,6 @@ struct sym_ext {
 	char		*path;
 };
 
-
-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
  */
@@ -624,7 +480,7 @@ static void output__resort(void)
 
 static void register_idle_thread(void)
 {
-	struct thread *thread = threads__findnew(0);
+	struct thread *thread = threads__findnew(0, &threads, &last_match);
 
 	if (thread == NULL ||
 			thread__set_comm(thread, "[idle]")) {
@@ -645,10 +501,12 @@ 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;
 
+	thread = threads__findnew(event->ip.pid, &threads, &last_match);
+
 	dprintf("%p [%p]: PERF_EVENT (IP, %d): %d: %p\n",
 		(void *)(offset + head),
 		(void *)(long)(event->header.size),
@@ -719,9 +577,11 @@ 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 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",
 		(void *)(offset + head),
 		(void *)(long)(event->header.size),
@@ -745,8 +605,9 @@ process_mmap_event(event_t *event, unsigned long offset, unsigned long head)
 static int
 process_comm_event(event_t *event, unsigned long offset, unsigned long head)
 {
-	struct thread *thread = threads__findnew(event->comm.pid);
+	struct thread *thread;
 
+	thread = threads__findnew(event->comm.pid, &threads, &last_match);
 	dprintf("%p [%p]: PERF_EVENT_COMM: %s:%d\n",
 		(void *)(offset + head),
 		(void *)(long)(event->header.size),
@@ -765,9 +626,11 @@ process_comm_event(event_t *event, unsigned long offset, unsigned long head)
 static int
 process_fork_event(event_t *event, unsigned long offset, unsigned long head)
 {
-	struct thread *thread = threads__findnew(event->fork.pid);
-	struct thread *parent = threads__findnew(event->fork.ppid);
+	struct thread *thread;
+	struct thread *parent;
 
+	thread = threads__findnew(event->fork.pid, &threads, &last_match);
+	parent = threads__findnew(event->fork.ppid, &threads, &last_match);
 	dprintf("%p [%p]: PERF_EVENT_FORK: %d:%d\n",
 		(void *)(offset + head),
 		(void *)(long)(event->header.size),
@@ -1202,7 +1065,7 @@ more:
 		return 0;
 
 	if (verbose >= 3)
-		threads__fprintf(stdout);
+		threads__fprintf(stdout, &threads);
 
 	if (verbose >= 2)
 		dsos__fprintf(stdout);
diff --git a/tools/perf/builtin-report.c b/tools/perf/builtin-report.c
index 6321951..298f26b 100644
--- a/tools/perf/builtin-report.c
+++ b/tools/perf/builtin-report.c
@@ -25,6 +25,8 @@
 #include "util/parse-options.h"
 #include "util/parse-events.h"
 
+#include "util/thread.h"
+
 #define SHOW_KERNEL	1
 #define SHOW_USER	2
 #define SHOW_HV		4
@@ -71,6 +73,9 @@ static char		__cwd[PATH_MAX];
 static char		*cwd = __cwd;
 static int		cwdlen;
 
+static struct rb_root	threads;
+static struct thread	*last_match;
+
 static
 struct callchain_param	callchain_param = {
 	.mode	= CHAIN_GRAPH_REL,
@@ -106,187 +111,10 @@ static int repsep_fprintf(FILE *fp, const char *fmt, ...)
 	return n;
 }
 
-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 unsigned int dsos__col_width,
 		    comms__col_width,
 		    threads__col_width;
 
-static int thread__set_comm(struct thread *self, const char *comm)
-{
-	if (self->comm)
-		free(self->comm);
-	self->comm = strdup(comm);
-	if (!self->comm)
-		return -ENOMEM;
-
-	if (!col_width_list_str && !field_sep &&
-	    (!comm_list || strlist__has_entry(comm_list, comm))) {
-		unsigned int slen = strlen(comm);
-		if (slen > comms__col_width) {
-			comms__col_width = slen;
-			threads__col_width = slen + 6;
-		}
-	}
-
-	return 0;
-}
-
-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)) {
-			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);
-			}
-		}
-	}
-
-	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
  */
@@ -1228,7 +1056,7 @@ print_entries:
 
 static void register_idle_thread(void)
 {
-	struct thread *thread = threads__findnew(0);
+	struct thread *thread = threads__findnew(0, &threads, &last_match);
 
 	if (thread == NULL ||
 			thread__set_comm(thread, "[idle]")) {
@@ -1263,7 +1091,7 @@ 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;
 	u64 period = 1;
 	struct map *map = NULL;
@@ -1271,6 +1099,8 @@ process_sample_event(event_t *event, unsigned long offset, unsigned long head)
 	struct ip_callchain *chain = NULL;
 	int cpumode;
 
+	thread = threads__findnew(event->ip.pid, &threads, &last_match);
+
 	if (sample_type & PERF_SAMPLE_PERIOD) {
 		period = *(u64 *)more_data;
 		more_data += sizeof(u64);
@@ -1360,9 +1190,11 @@ 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 thread *thread;
 	struct map *map = map__new(&event->mmap, cwd, cwdlen);
 
+	thread = threads__findnew(event->mmap.pid, &threads, &last_match);
+
 	dprintf("%p [%p]: PERF_EVENT_MMAP %d/%d: [%p(%p) @ %p]: %s\n",
 		(void *)(offset + head),
 		(void *)(long)(event->header.size),
@@ -1387,7 +1219,9 @@ process_mmap_event(event_t *event, unsigned long offset, unsigned long head)
 static int
 process_comm_event(event_t *event, unsigned long offset, unsigned long head)
 {
-	struct thread *thread = threads__findnew(event->comm.pid);
+	struct thread *thread;
+
+	thread = threads__findnew(event->comm.pid, &threads, &last_match);
 
 	dprintf("%p [%p]: PERF_EVENT_COMM: %s:%d\n",
 		(void *)(offset + head),
@@ -1407,8 +1241,11 @@ process_comm_event(event_t *event, unsigned long offset, unsigned long head)
 static int
 process_task_event(event_t *event, unsigned long offset, unsigned long head)
 {
-	struct thread *thread = threads__findnew(event->fork.pid);
-	struct thread *parent = threads__findnew(event->fork.ppid);
+	struct thread *thread;
+	struct thread *parent;
+
+	thread = threads__findnew(event->fork.pid, &threads, &last_match);
+	parent = threads__findnew(event->fork.ppid, &threads, &last_match);
 
 	dprintf("%p [%p]: PERF_EVENT_%s: (%d:%d):(%d:%d)\n",
 		(void *)(offset + head),
@@ -1749,7 +1586,7 @@ done:
 		return 0;
 
 	if (verbose >= 3)
-		threads__fprintf(stdout);
+		threads__fprintf(stdout, &threads);
 
 	if (verbose >= 2)
 		dsos__fprintf(stdout);
diff --git a/tools/perf/util/thread.c b/tools/perf/util/thread.c
new file mode 100644
index 0000000..00c14b9
--- /dev/null
+++ b/tools/perf/util/thread.c
@@ -0,0 +1,143 @@
+#include "../perf.h"
+#include <stdlib.h>
+#include <stdio.h>
+#include <string.h>
+#include "thread.h"
+#include "util.h"
+
+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;
+}
+
+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;
+}
+
+struct thread *
+threads__findnew(pid_t pid, struct rb_root *threads, struct thread **last_match)
+{
+	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;
+}
+
+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);
+}
+
+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;
+}
+
+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;
+}
+
+size_t threads__fprintf(FILE *fp, struct rb_root *threads)
+{
+	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;
+}
diff --git a/tools/perf/util/thread.h b/tools/perf/util/thread.h
new file mode 100644
index 0000000..b1c6671
--- /dev/null
+++ b/tools/perf/util/thread.h
@@ -0,0 +1,19 @@
+#include <linux/rbtree.h>
+#include <linux/list.h>
+#include <unistd.h>
+#include "symbol.h"
+
+struct thread {
+	struct rb_node	 rb_node;
+	struct list_head maps;
+	pid_t		 pid;
+	char		 *comm;
+};
+
+int thread__set_comm(struct thread *self, const char *comm);
+struct thread *
+threads__findnew(pid_t pid, struct rb_root *threads, struct thread **last_match);
+void thread__insert_map(struct thread *self, struct map *map);
+int thread__fork(struct thread *self, struct thread *parent);
+struct map *thread__find_map(struct thread *self, u64 ip);
+size_t threads__fprintf(FILE *fp, struct rb_root *threads);
-- 
1.6.2.3

[tip:perfcounters/core] perf tools: Factorize the thread code in a dedicated file

[tip-bot for Frederic Weisbecker]

Commit-ID:  6baa0a5ae0954fb2486c480a20556a9f1aee0965
Gitweb:     http://git.kernel.org/tip/6baa0a5ae0954fb2486c480a20556a9f1aee0965
Author:     Frederic Weisbecker <fweisbec@gmail.com>
AuthorDate: Fri, 14 Aug 2009 12:21:53 +0200
Committer:  Ingo Molnar <mingo@elte.hu>
CommitDate: Sat, 15 Aug 2009 16:10:19 +0200

perf tools: Factorize the thread code in a dedicated file

Factorize the thread management code used by perf-annotate and
perf-report in dedicated source and header files.

v2: pass last_match by address so that it can actually be
modified.

Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Mike Galbraith <efault@gmx.de>
LKML-Reference: <1250245313-6995-1-git-send-email-fweisbec@gmail.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>


---
 tools/perf/Makefile           |    1 +
 tools/perf/builtin-annotate.c |  173 ++++-------------------------------
 tools/perf/builtin-report.c   |  205 ++++------------------------------------
 tools/perf/util/thread.c      |  143 ++++++++++++++++++++++++++++
 tools/perf/util/thread.h      |   19 ++++
 5 files changed, 202 insertions(+), 339 deletions(-)

diff --git a/tools/perf/Makefile b/tools/perf/Makefile
index 68218cf..0056405 100644
--- a/tools/perf/Makefile
+++ b/tools/perf/Makefile
@@ -341,6 +341,7 @@ 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
 
 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 543c452..3bedaa5 100644
--- a/tools/perf/builtin-annotate.c
+++ b/tools/perf/builtin-annotate.c
@@ -20,6 +20,7 @@
 
 #include "util/parse-options.h"
 #include "util/parse-events.h"
+#include "util/thread.h"
 
 #define SHOW_KERNEL	1
 #define SHOW_USER	2
@@ -44,6 +45,9 @@ static int		print_line;
 static unsigned long	page_size;
 static unsigned long	mmap_window = 32;
 
+static struct rb_root	threads;
+static struct thread	*last_match;
+
 
 struct sym_ext {
 	struct rb_node	node;
@@ -51,154 +55,6 @@ struct sym_ext {
 	char		*path;
 };
 
-
-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
  */
@@ -624,7 +480,7 @@ static void output__resort(void)
 
 static void register_idle_thread(void)
 {
-	struct thread *thread = threads__findnew(0);
+	struct thread *thread = threads__findnew(0, &threads, &last_match);
 
 	if (thread == NULL ||
 			thread__set_comm(thread, "[idle]")) {
@@ -645,10 +501,12 @@ 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;
 
+	thread = threads__findnew(event->ip.pid, &threads, &last_match);
+
 	dprintf("%p [%p]: PERF_EVENT (IP, %d): %d: %p\n",
 		(void *)(offset + head),
 		(void *)(long)(event->header.size),
@@ -719,9 +577,11 @@ 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 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",
 		(void *)(offset + head),
 		(void *)(long)(event->header.size),
@@ -745,8 +605,9 @@ process_mmap_event(event_t *event, unsigned long offset, unsigned long head)
 static int
 process_comm_event(event_t *event, unsigned long offset, unsigned long head)
 {
-	struct thread *thread = threads__findnew(event->comm.pid);
+	struct thread *thread;
 
+	thread = threads__findnew(event->comm.pid, &threads, &last_match);
 	dprintf("%p [%p]: PERF_EVENT_COMM: %s:%d\n",
 		(void *)(offset + head),
 		(void *)(long)(event->header.size),
@@ -765,9 +626,11 @@ process_comm_event(event_t *event, unsigned long offset, unsigned long head)
 static int
 process_fork_event(event_t *event, unsigned long offset, unsigned long head)
 {
-	struct thread *thread = threads__findnew(event->fork.pid);
-	struct thread *parent = threads__findnew(event->fork.ppid);
+	struct thread *thread;
+	struct thread *parent;
 
+	thread = threads__findnew(event->fork.pid, &threads, &last_match);
+	parent = threads__findnew(event->fork.ppid, &threads, &last_match);
 	dprintf("%p [%p]: PERF_EVENT_FORK: %d:%d\n",
 		(void *)(offset + head),
 		(void *)(long)(event->header.size),
@@ -1202,7 +1065,7 @@ more:
 		return 0;
 
 	if (verbose >= 3)
-		threads__fprintf(stdout);
+		threads__fprintf(stdout, &threads);
 
 	if (verbose >= 2)
 		dsos__fprintf(stdout);
diff --git a/tools/perf/builtin-report.c b/tools/perf/builtin-report.c
index 6321951..298f26b 100644
--- a/tools/perf/builtin-report.c
+++ b/tools/perf/builtin-report.c
@@ -25,6 +25,8 @@
 #include "util/parse-options.h"
 #include "util/parse-events.h"
 
+#include "util/thread.h"
+
 #define SHOW_KERNEL	1
 #define SHOW_USER	2
 #define SHOW_HV		4
@@ -71,6 +73,9 @@ static char		__cwd[PATH_MAX];
 static char		*cwd = __cwd;
 static int		cwdlen;
 
+static struct rb_root	threads;
+static struct thread	*last_match;
+
 static
 struct callchain_param	callchain_param = {
 	.mode	= CHAIN_GRAPH_REL,
@@ -106,187 +111,10 @@ static int repsep_fprintf(FILE *fp, const char *fmt, ...)
 	return n;
 }
 
-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 unsigned int dsos__col_width,
 		    comms__col_width,
 		    threads__col_width;
 
-static int thread__set_comm(struct thread *self, const char *comm)
-{
-	if (self->comm)
-		free(self->comm);
-	self->comm = strdup(comm);
-	if (!self->comm)
-		return -ENOMEM;
-
-	if (!col_width_list_str && !field_sep &&
-	    (!comm_list || strlist__has_entry(comm_list, comm))) {
-		unsigned int slen = strlen(comm);
-		if (slen > comms__col_width) {
-			comms__col_width = slen;
-			threads__col_width = slen + 6;
-		}
-	}
-
-	return 0;
-}
-
-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)) {
-			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);
-			}
-		}
-	}
-
-	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
  */
@@ -1228,7 +1056,7 @@ print_entries:
 
 static void register_idle_thread(void)
 {
-	struct thread *thread = threads__findnew(0);
+	struct thread *thread = threads__findnew(0, &threads, &last_match);
 
 	if (thread == NULL ||
 			thread__set_comm(thread, "[idle]")) {
@@ -1263,7 +1091,7 @@ 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;
 	u64 period = 1;
 	struct map *map = NULL;
@@ -1271,6 +1099,8 @@ process_sample_event(event_t *event, unsigned long offset, unsigned long head)
 	struct ip_callchain *chain = NULL;
 	int cpumode;
 
+	thread = threads__findnew(event->ip.pid, &threads, &last_match);
+
 	if (sample_type & PERF_SAMPLE_PERIOD) {
 		period = *(u64 *)more_data;
 		more_data += sizeof(u64);
@@ -1360,9 +1190,11 @@ 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 thread *thread;
 	struct map *map = map__new(&event->mmap, cwd, cwdlen);
 
+	thread = threads__findnew(event->mmap.pid, &threads, &last_match);
+
 	dprintf("%p [%p]: PERF_EVENT_MMAP %d/%d: [%p(%p) @ %p]: %s\n",
 		(void *)(offset + head),
 		(void *)(long)(event->header.size),
@@ -1387,7 +1219,9 @@ process_mmap_event(event_t *event, unsigned long offset, unsigned long head)
 static int
 process_comm_event(event_t *event, unsigned long offset, unsigned long head)
 {
-	struct thread *thread = threads__findnew(event->comm.pid);
+	struct thread *thread;
+
+	thread = threads__findnew(event->comm.pid, &threads, &last_match);
 
 	dprintf("%p [%p]: PERF_EVENT_COMM: %s:%d\n",
 		(void *)(offset + head),
@@ -1407,8 +1241,11 @@ process_comm_event(event_t *event, unsigned long offset, unsigned long head)
 static int
 process_task_event(event_t *event, unsigned long offset, unsigned long head)
 {
-	struct thread *thread = threads__findnew(event->fork.pid);
-	struct thread *parent = threads__findnew(event->fork.ppid);
+	struct thread *thread;
+	struct thread *parent;
+
+	thread = threads__findnew(event->fork.pid, &threads, &last_match);
+	parent = threads__findnew(event->fork.ppid, &threads, &last_match);
 
 	dprintf("%p [%p]: PERF_EVENT_%s: (%d:%d):(%d:%d)\n",
 		(void *)(offset + head),
@@ -1749,7 +1586,7 @@ done:
 		return 0;
 
 	if (verbose >= 3)
-		threads__fprintf(stdout);
+		threads__fprintf(stdout, &threads);
 
 	if (verbose >= 2)
 		dsos__fprintf(stdout);
diff --git a/tools/perf/util/thread.c b/tools/perf/util/thread.c
new file mode 100644
index 0000000..00c14b9
--- /dev/null
+++ b/tools/perf/util/thread.c
@@ -0,0 +1,143 @@
+#include "../perf.h"
+#include <stdlib.h>
+#include <stdio.h>
+#include <string.h>
+#include "thread.h"
+#include "util.h"
+
+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;
+}
+
+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;
+}
+
+struct thread *
+threads__findnew(pid_t pid, struct rb_root *threads, struct thread **last_match)
+{
+	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;
+}
+
+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);
+}
+
+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;
+}
+
+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;
+}
+
+size_t threads__fprintf(FILE *fp, struct rb_root *threads)
+{
+	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;
+}
diff --git a/tools/perf/util/thread.h b/tools/perf/util/thread.h
new file mode 100644
index 0000000..b1c6671
--- /dev/null
+++ b/tools/perf/util/thread.h
@@ -0,0 +1,19 @@
+#include <linux/rbtree.h>
+#include <linux/list.h>
+#include <unistd.h>
+#include "symbol.h"
+
+struct thread {
+	struct rb_node	 rb_node;
+	struct list_head maps;
+	pid_t		 pid;
+	char		 *comm;
+};
+
+int thread__set_comm(struct thread *self, const char *comm);
+struct thread *
+threads__findnew(pid_t pid, struct rb_root *threads, struct thread **last_match);
+void thread__insert_map(struct thread *self, struct map *map);
+int thread__fork(struct thread *self, struct thread *parent);
+struct map *thread__find_map(struct thread *self, u64 ip);
+size_t threads__fprintf(FILE *fp, struct rb_root *threads);

[PATCH 2/2] perf tools: Add a general option to enable raw sample records

[Frederic Weisbecker]

While we can enable the perf sample records per tracepoint counter,
we may also want to enable this option for every tracepoint
counters to open, so that we don't need to add a :record flag
for all of them.

Add the -R, --raw-samples options for this purpose.

Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Mike Galbraith <efault@gmx.de>
---
 tools/perf/builtin-record.c |    5 +++++
 1 files changed, 5 insertions(+), 0 deletions(-)

diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c
index 3be0301..e67c4fa 100644
--- a/tools/perf/builtin-record.c
+++ b/tools/perf/builtin-record.c
@@ -35,6 +35,7 @@ static int			output;
 static const char		*output_name			= "perf.data";
 static int			group				= 0;
 static unsigned int		realtime_prio			= 0;
+static int			raw_samples			= 0;
 static int			system_wide			= 0;
 static int			profile_cpu			= -1;
 static pid_t			target_pid			= -1;
@@ -400,6 +401,8 @@ static void create_counter(int counter, int cpu, pid_t pid)
 	if (call_graph)
 		attr->sample_type	|= PERF_SAMPLE_CALLCHAIN;
 
+	if (raw_samples)
+		attr->sample_type	|= PERF_SAMPLE_RAW;
 
 	attr->mmap		= track;
 	attr->comm		= track;
@@ -632,6 +635,8 @@ static const struct option options[] = {
 		    "record events on existing pid"),
 	OPT_INTEGER('r', "realtime", &realtime_prio,
 		    "collect data with this RT SCHED_FIFO priority"),
+	OPT_BOOLEAN('R', "raw-samples", &raw_samples,
+		    "collect raw sample records from all opened counters"),
 	OPT_BOOLEAN('a', "all-cpus", &system_wide,
 			    "system-wide collection from all CPUs"),
 	OPT_BOOLEAN('A', "append", &append_file,
-- 
1.6.2.3

Re: [PATCH 2/2] perf tools: Add a general option to enable raw sample records

[Frederic Weisbecker]

Sorry, please ignore this one, it has been resent by mistake :-(

On Fri, Aug 14, 2009 at 12:05:00PM +0200, Frederic Weisbecker wrote:
> While we can enable the perf sample records per tracepoint counter,
> we may also want to enable this option for every tracepoint
> counters to open, so that we don't need to add a :record flag
> for all of them.
> 
> Add the -R, --raw-samples options for this purpose.
> 
> Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
> Cc: Peter Zijlstra <peterz@infradead.org>
> Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
> Cc: Mike Galbraith <efault@gmx.de>
> ---
>  tools/perf/builtin-record.c |    5 +++++
>  1 files changed, 5 insertions(+), 0 deletions(-)
> 
> diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c
> index 3be0301..e67c4fa 100644
> --- a/tools/perf/builtin-record.c
> +++ b/tools/perf/builtin-record.c
> @@ -35,6 +35,7 @@ static int			output;
>  static const char		*output_name			= "perf.data";
>  static int			group				= 0;
>  static unsigned int		realtime_prio			= 0;
> +static int			raw_samples			= 0;
>  static int			system_wide			= 0;
>  static int			profile_cpu			= -1;
>  static pid_t			target_pid			= -1;
> @@ -400,6 +401,8 @@ static void create_counter(int counter, int cpu, pid_t pid)
>  	if (call_graph)
>  		attr->sample_type	|= PERF_SAMPLE_CALLCHAIN;
>  
> +	if (raw_samples)
> +		attr->sample_type	|= PERF_SAMPLE_RAW;
>  
>  	attr->mmap		= track;
>  	attr->comm		= track;
> @@ -632,6 +635,8 @@ static const struct option options[] = {
>  		    "record events on existing pid"),
>  	OPT_INTEGER('r', "realtime", &realtime_prio,
>  		    "collect data with this RT SCHED_FIFO priority"),
> +	OPT_BOOLEAN('R', "raw-samples", &raw_samples,
> +		    "collect raw sample records from all opened counters"),
>  	OPT_BOOLEAN('a', "all-cpus", &system_wide,
>  			    "system-wide collection from all CPUs"),
>  	OPT_BOOLEAN('A', "append", &append_file,
> -- 
> 1.6.2.3
> 

Re: [PATCH 1/2] perf tools: Add a per tracepoint counter attribute to get raw sample

[Frederic Weisbecker]

Sorry, please ignore this one, it has been resent by mistake :-(

On Fri, Aug 14, 2009 at 12:04:58PM +0200, Frederic Weisbecker wrote:
> Add a new flag field while opening a tracepoint perf counter:
> 
> 	-e tracepoint_subsystem:tracepoint_name:flags
> 
> This is intended to be generic although for now it only supports the
> r[e[c[o[r[d]]]]] flag:
> 
> 	./perf record -e workqueue:workqueue_insertion:record
> 	./perf record -e workqueue:workqueue_insertion:r
> 
> will have the same effect: enabling the raw samples record for the
> given tracepoint counter.
> 
> In the future, we may want to support further flags, separated by
> commas.
> 
> Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
> Cc: Peter Zijlstra <peterz@infradead.org>
> Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
> Cc: Mike Galbraith <efault@gmx.de>
> ---
>  tools/perf/builtin-record.c    |    2 +-
>  tools/perf/util/parse-events.c |   10 ++++++++++
>  2 files changed, 11 insertions(+), 1 deletions(-)
> 
> diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c
> index 5aeb632..3be0301 100644
> --- a/tools/perf/builtin-record.c
> +++ b/tools/perf/builtin-record.c
> @@ -380,7 +380,7 @@ static void create_counter(int counter, int cpu, pid_t pid)
>  				  PERF_FORMAT_TOTAL_TIME_RUNNING |
>  				  PERF_FORMAT_ID;
>  
> -	attr->sample_type	= PERF_SAMPLE_IP | PERF_SAMPLE_TID;
> +	attr->sample_type	|= PERF_SAMPLE_IP | PERF_SAMPLE_TID;
>  
>  	if (freq) {
>  		attr->sample_type	|= PERF_SAMPLE_PERIOD;
> diff --git a/tools/perf/util/parse-events.c b/tools/perf/util/parse-events.c
> index a5d661b..4bab278 100644
> --- a/tools/perf/util/parse-events.c
> +++ b/tools/perf/util/parse-events.c
> @@ -379,6 +379,7 @@ static int parse_tracepoint_event(const char **strp,
>  				    struct perf_counter_attr *attr)
>  {
>  	const char *evt_name;
> +	char *flags;
>  	char sys_name[MAX_EVENT_LENGTH];
>  	char id_buf[4];
>  	int fd;
> @@ -400,6 +401,15 @@ static int parse_tracepoint_event(const char **strp,
>  	strncpy(sys_name, *strp, sys_length);
>  	sys_name[sys_length] = '\0';
>  	evt_name = evt_name + 1;
> +
> +	flags = strchr(evt_name, ':');
> +	if (flags) {
> +		*flags = '\0';
> +		flags++;
> +		if (!strncmp(flags, "record", strlen(flags)))
> +			attr->sample_type |= PERF_SAMPLE_RAW;
> +	}
> +
>  	evt_length = strlen(evt_name);
>  	if (evt_length >= MAX_EVENT_LENGTH)
>  		return 0;
> -- 
> 1.6.2.3
> 

[PATCH 1/2] perf tools: Add a per tracepoint counter attribute to get raw sample

[Frederic Weisbecker]

Add a new flag field while opening a tracepoint perf counter:

	-e tracepoint_subsystem:tracepoint_name:flags

This is intended to be generic although for now it only supports the
r[e[c[o[r[d]]]]] flag:

	./perf record -e workqueue:workqueue_insertion:record
	./perf record -e workqueue:workqueue_insertion:r

will have the same effect: enabling the raw samples record for the
given tracepoint counter.

In the future, we may want to support further flags, separated by
commas.

Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Mike Galbraith <efault@gmx.de>
---
 tools/perf/builtin-record.c    |    2 +-
 tools/perf/util/parse-events.c |   10 ++++++++++
 2 files changed, 11 insertions(+), 1 deletions(-)

diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c
index 5aeb632..3be0301 100644
--- a/tools/perf/builtin-record.c
+++ b/tools/perf/builtin-record.c
@@ -380,7 +380,7 @@ static void create_counter(int counter, int cpu, pid_t pid)
 				  PERF_FORMAT_TOTAL_TIME_RUNNING |
 				  PERF_FORMAT_ID;
 
-	attr->sample_type	= PERF_SAMPLE_IP | PERF_SAMPLE_TID;
+	attr->sample_type	|= PERF_SAMPLE_IP | PERF_SAMPLE_TID;
 
 	if (freq) {
 		attr->sample_type	|= PERF_SAMPLE_PERIOD;
diff --git a/tools/perf/util/parse-events.c b/tools/perf/util/parse-events.c
index a5d661b..4bab278 100644
--- a/tools/perf/util/parse-events.c
+++ b/tools/perf/util/parse-events.c
@@ -379,6 +379,7 @@ static int parse_tracepoint_event(const char **strp,
 				    struct perf_counter_attr *attr)
 {
 	const char *evt_name;
+	char *flags;
 	char sys_name[MAX_EVENT_LENGTH];
 	char id_buf[4];
 	int fd;
@@ -400,6 +401,15 @@ static int parse_tracepoint_event(const char **strp,
 	strncpy(sys_name, *strp, sys_length);
 	sys_name[sys_length] = '\0';
 	evt_name = evt_name + 1;
+
+	flags = strchr(evt_name, ':');
+	if (flags) {
+		*flags = '\0';
+		flags++;
+		if (!strncmp(flags, "record", strlen(flags)))
+			attr->sample_type |= PERF_SAMPLE_RAW;
+	}
+
 	evt_length = strlen(evt_name);
 	if (evt_length >= MAX_EVENT_LENGTH)
 		return 0;
-- 
1.6.2.3

[PATCH 2/2] perf tools: Add a general option to enable raw sample records

[Frederic Weisbecker]

While we can enable the perf sample records per tracepoint counter,
we may also want to enable this option for every tracepoint
counters to open, so that we don't need to add a :record flag
for all of them.

Add the -R, --raw-samples options for this purpose.

Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Mike Galbraith <efault@gmx.de>
---
 tools/perf/builtin-record.c |    5 +++++
 1 files changed, 5 insertions(+), 0 deletions(-)

diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c
index 3be0301..e67c4fa 100644
--- a/tools/perf/builtin-record.c
+++ b/tools/perf/builtin-record.c
@@ -35,6 +35,7 @@ static int			output;
 static const char		*output_name			= "perf.data";
 static int			group				= 0;
 static unsigned int		realtime_prio			= 0;
+static int			raw_samples			= 0;
 static int			system_wide			= 0;
 static int			profile_cpu			= -1;
 static pid_t			target_pid			= -1;
@@ -400,6 +401,8 @@ static void create_counter(int counter, int cpu, pid_t pid)
 	if (call_graph)
 		attr->sample_type	|= PERF_SAMPLE_CALLCHAIN;
 
+	if (raw_samples)
+		attr->sample_type	|= PERF_SAMPLE_RAW;
 
 	attr->mmap		= track;
 	attr->comm		= track;
@@ -632,6 +635,8 @@ static const struct option options[] = {
 		    "record events on existing pid"),
 	OPT_INTEGER('r', "realtime", &realtime_prio,
 		    "collect data with this RT SCHED_FIFO priority"),
+	OPT_BOOLEAN('R', "raw-samples", &raw_samples,
+		    "collect raw sample records from all opened counters"),
 	OPT_BOOLEAN('a', "all-cpus", &system_wide,
 			    "system-wide collection from all CPUs"),
 	OPT_BOOLEAN('A', "append", &append_file,
-- 
1.6.2.3