[GIT PULL 0/3] perf/core fixes and improvements

[GIT PULL 0/3] perf/core fixes and improvements

[Arnaldo Carvalho de Melo]

Hi Ingo,

        Please consider pulling from:

git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux-2.6 perf/core

Regards,

- Arnaldo

Arnaldo Carvalho de Melo (3):
  perf session: Pass evsel in event_ops->sample()
  perf build-id: Add quirk to deal with perf.data file format breakage
  perf symbols: Look at .dynsym again if .symtab not found

 tools/perf/builtin-annotate.c                      |   18 ++-----
 tools/perf/builtin-diff.c                          |    1 +
 tools/perf/builtin-inject.c                        |   11 ++++-
 tools/perf/builtin-kmem.c                          |    1 +
 tools/perf/builtin-lock.c                          |    4 +-
 tools/perf/builtin-report.c                        |   19 ++-----
 tools/perf/builtin-sched.c                         |    1 +
 tools/perf/builtin-script.c                        |   15 ++----
 tools/perf/builtin-timechart.c                     |   11 ++++
 tools/perf/util/build-id.c                         |    1 +
 tools/perf/util/header.c                           |   57 +++++++++++++++++++-
 tools/perf/util/hist.h                             |    1 +
 .../perf/util/scripting-engines/trace-event-perl.c |    1 +
 .../util/scripting-engines/trace-event-python.c    |    1 +
 tools/perf/util/session.c                          |   25 ++++++++-
 tools/perf/util/session.h                          |    7 ++-
 tools/perf/util/symbol.c                           |   25 +++++----
 tools/perf/util/trace-event-scripting.c            |    1 +
 tools/perf/util/trace-event.h                      |    1 +
 19 files changed, 142 insertions(+), 59 deletions(-)

[PATCH 1/3] perf session: Pass evsel in event_ops->sample()

[Arnaldo Carvalho de Melo]

From: Arnaldo Carvalho de Melo <acme@redhat.com>

Resolving the sample->id to an evsel since the most advanced tools,
report and annotate, and the others will too when they evolve to
properly support multi-event perf.data files.

Good also because it does an extra validation, checking that the ID is
valid when present. When that is not the case, the overhead is just a
branch + function call (perf_evlist__id2evsel).

Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Tom Zanussi <tzanussi@gmail.com>
LKML-Reference: <new-submission>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/builtin-annotate.c                      |   18 +++-----------
 tools/perf/builtin-diff.c                          |    1 +
 tools/perf/builtin-inject.c                        |   11 ++++++++-
 tools/perf/builtin-kmem.c                          |    1 +
 tools/perf/builtin-lock.c                          |    4 ++-
 tools/perf/builtin-report.c                        |   19 +++------------
 tools/perf/builtin-sched.c                         |    1 +
 tools/perf/builtin-script.c                        |   15 +++--------
 tools/perf/builtin-timechart.c                     |   11 ++++++++
 tools/perf/util/build-id.c                         |    1 +
 tools/perf/util/hist.h                             |    1 +
 .../perf/util/scripting-engines/trace-event-perl.c |    1 +
 .../util/scripting-engines/trace-event-python.c    |    1 +
 tools/perf/util/session.c                          |   25 ++++++++++++++++++-
 tools/perf/util/session.h                          |    7 ++++-
 tools/perf/util/trace-event-scripting.c            |    1 +
 tools/perf/util/trace-event.h                      |    1 +
 17 files changed, 73 insertions(+), 46 deletions(-)

diff --git a/tools/perf/builtin-annotate.c b/tools/perf/builtin-annotate.c
index 695de4b..e18eb7e 100644
--- a/tools/perf/builtin-annotate.c
+++ b/tools/perf/builtin-annotate.c
@@ -42,9 +42,9 @@ static const char *sym_hist_filter;
 
 static int perf_evlist__add_sample(struct perf_evlist *evlist,
 				   struct perf_sample *sample,
+				   struct perf_evsel *evsel,
 				   struct addr_location *al)
 {
-	struct perf_evsel *evsel;
 	struct hist_entry *he;
 	int ret;
 
@@ -59,18 +59,6 @@ static int perf_evlist__add_sample(struct perf_evlist *evlist,
 		return 0;
 	}
 
-	evsel = perf_evlist__id2evsel(evlist, sample->id);
-	if (evsel == NULL) {
-		/*
-		 * FIXME: Propagate this back, but at least we're in a builtin,
-		 * where exit() is allowed. ;-)
-		 */
-		ui__warning("Invalid %s file, contains samples with id not in "
-			    "its header!\n", input_name);
-		exit_browser(0);
-		exit(1);
-	}
-
 	he = __hists__add_entry(&evsel->hists, al, NULL, 1);
 	if (he == NULL)
 		return -ENOMEM;
@@ -92,6 +80,7 @@ static int perf_evlist__add_sample(struct perf_evlist *evlist,
 
 static int process_sample_event(union perf_event *event,
 				struct perf_sample *sample,
+				struct perf_evsel *evsel,
 				struct perf_session *session)
 {
 	struct addr_location al;
@@ -103,7 +92,8 @@ static int process_sample_event(union perf_event *event,
 		return -1;
 	}
 
-	if (!al.filtered && perf_evlist__add_sample(session->evlist, sample, &al)) {
+	if (!al.filtered &&
+	    perf_evlist__add_sample(session->evlist, sample, evsel, &al)) {
 		pr_warning("problem incrementing symbol count, "
 			   "skipping event\n");
 		return -1;
diff --git a/tools/perf/builtin-diff.c b/tools/perf/builtin-diff.c
index 6b7d911..e821999 100644
--- a/tools/perf/builtin-diff.c
+++ b/tools/perf/builtin-diff.c
@@ -32,6 +32,7 @@ static int hists__add_entry(struct hists *self,
 
 static int diff__process_sample_event(union perf_event *event,
 				      struct perf_sample *sample,
+				      struct perf_evsel *evsel __used,
 				      struct perf_session *session)
 {
 	struct addr_location al;
diff --git a/tools/perf/builtin-inject.c b/tools/perf/builtin-inject.c
index e29f04e..8dfc12b 100644
--- a/tools/perf/builtin-inject.c
+++ b/tools/perf/builtin-inject.c
@@ -43,6 +43,14 @@ static int perf_event__repipe(union perf_event *event,
 	return perf_event__repipe_synth(event, session);
 }
 
+static int perf_event__repipe_sample(union perf_event *event,
+			      struct perf_sample *sample __used,
+			      struct perf_evsel *evsel __used,
+			      struct perf_session *session)
+{
+	return perf_event__repipe_synth(event, session);
+}
+
 static int perf_event__repipe_mmap(union perf_event *event,
 				   struct perf_sample *sample,
 				   struct perf_session *session)
@@ -124,6 +132,7 @@ static int dso__inject_build_id(struct dso *self, struct perf_session *session)
 
 static int perf_event__inject_buildid(union perf_event *event,
 				      struct perf_sample *sample,
+				      struct perf_evsel *evsel __used,
 				      struct perf_session *session)
 {
 	struct addr_location al;
@@ -164,7 +173,7 @@ repipe:
 }
 
 struct perf_event_ops inject_ops = {
-	.sample		= perf_event__repipe,
+	.sample		= perf_event__repipe_sample,
 	.mmap		= perf_event__repipe,
 	.comm		= perf_event__repipe,
 	.fork		= perf_event__repipe,
diff --git a/tools/perf/builtin-kmem.c b/tools/perf/builtin-kmem.c
index 7f618f4..225e963 100644
--- a/tools/perf/builtin-kmem.c
+++ b/tools/perf/builtin-kmem.c
@@ -305,6 +305,7 @@ static void process_raw_event(union perf_event *raw_event __used, void *data,
 
 static int process_sample_event(union perf_event *event,
 				struct perf_sample *sample,
+				struct perf_evsel *evsel __used,
 				struct perf_session *session)
 {
 	struct thread *thread = perf_session__findnew(session, event->ip.pid);
diff --git a/tools/perf/builtin-lock.c b/tools/perf/builtin-lock.c
index 7a2a79d..9ac05aa 100644
--- a/tools/perf/builtin-lock.c
+++ b/tools/perf/builtin-lock.c
@@ -845,7 +845,9 @@ static void dump_info(void)
 		die("Unknown type of information\n");
 }
 
-static int process_sample_event(union perf_event *event, struct perf_sample *sample,
+static int process_sample_event(union perf_event *event,
+				struct perf_sample *sample,
+				struct perf_evsel *evsel __used,
 				struct perf_session *s)
 {
 	struct thread *thread = perf_session__findnew(s, sample->tid);
diff --git a/tools/perf/builtin-report.c b/tools/perf/builtin-report.c
index b1b8200..498c6f7 100644
--- a/tools/perf/builtin-report.c
+++ b/tools/perf/builtin-report.c
@@ -50,12 +50,12 @@ static symbol_filter_t	annotate_init;
 
 static int perf_session__add_hist_entry(struct perf_session *session,
 					struct addr_location *al,
-					struct perf_sample *sample)
+					struct perf_sample *sample,
+					struct perf_evsel *evsel)
 {
 	struct symbol *parent = NULL;
 	int err = 0;
 	struct hist_entry *he;
-	struct perf_evsel *evsel;
 
 	if ((sort__has_parent || symbol_conf.use_callchain) && sample->callchain) {
 		err = perf_session__resolve_callchain(session, al->thread,
@@ -64,18 +64,6 @@ static int perf_session__add_hist_entry(struct perf_session *session,
 			return err;
 	}
 
-	evsel = perf_evlist__id2evsel(session->evlist, sample->id);
-	if (evsel == NULL) {
-		/*
-		 * FIXME: Propagate this back, but at least we're in a builtin,
-		 * where exit() is allowed. ;-)
-		 */
-		ui__warning("Invalid %s file, contains samples with id %" PRIu64 " not in "
-			    "its header!\n", input_name, sample->id);
-		exit_browser(0);
-		exit(1);
-	}
-
 	he = __hists__add_entry(&evsel->hists, al, parent, sample->period);
 	if (he == NULL)
 		return -ENOMEM;
@@ -113,6 +101,7 @@ out:
 
 static int process_sample_event(union perf_event *event,
 				struct perf_sample *sample,
+				struct perf_evsel *evsel,
 				struct perf_session *session)
 {
 	struct addr_location al;
@@ -127,7 +116,7 @@ static int process_sample_event(union perf_event *event,
 	if (al.filtered || (hide_unresolved && al.sym == NULL))
 		return 0;
 
-	if (perf_session__add_hist_entry(session, &al, sample)) {
+	if (perf_session__add_hist_entry(session, &al, sample, evsel)) {
 		pr_debug("problem incrementing symbol period, skipping event\n");
 		return -1;
 	}
diff --git a/tools/perf/builtin-sched.c b/tools/perf/builtin-sched.c
index a32f411..dcfe887 100644
--- a/tools/perf/builtin-sched.c
+++ b/tools/perf/builtin-sched.c
@@ -1603,6 +1603,7 @@ static void process_raw_event(union perf_event *raw_event __used,
 
 static int process_sample_event(union perf_event *event,
 				struct perf_sample *sample,
+				struct perf_evsel *evsel __used,
 				struct perf_session *session)
 {
 	struct thread *thread;
diff --git a/tools/perf/builtin-script.c b/tools/perf/builtin-script.c
index 9f5fc54..ac574ea 100644
--- a/tools/perf/builtin-script.c
+++ b/tools/perf/builtin-script.c
@@ -162,19 +162,11 @@ static void print_sample_start(struct perf_sample *sample,
 
 static void process_event(union perf_event *event __unused,
 			  struct perf_sample *sample,
+			  struct perf_evsel *evsel,
 			  struct perf_session *session,
 			  struct thread *thread)
 {
-	struct perf_event_attr *attr;
-	struct perf_evsel *evsel;
-
-	evsel = perf_evlist__id2evsel(session->evlist, sample->id);
-	if (evsel == NULL) {
-		pr_err("Invalid data. Contains samples with id not in "
-		       "its header!\n");
-		return;
-	}
-	attr = &evsel->attr;
+	struct perf_event_attr *attr = &evsel->attr;
 
 	if (output_fields[attr->type] == 0)
 		return;
@@ -244,6 +236,7 @@ static char const		*input_name = "perf.data";
 
 static int process_sample_event(union perf_event *event,
 				struct perf_sample *sample,
+				struct perf_evsel *evsel,
 				struct perf_session *session)
 {
 	struct thread *thread = perf_session__findnew(session, event->ip.pid);
@@ -264,7 +257,7 @@ static int process_sample_event(union perf_event *event,
 		last_timestamp = sample->time;
 		return 0;
 	}
-	scripting_ops->process_event(event, sample, session, thread);
+	scripting_ops->process_event(event, sample, evsel, session, thread);
 
 	session->hists.stats.total_period += sample->period;
 	return 0;
diff --git a/tools/perf/builtin-timechart.c b/tools/perf/builtin-timechart.c
index 67c0459..aa26f4d 100644
--- a/tools/perf/builtin-timechart.c
+++ b/tools/perf/builtin-timechart.c
@@ -488,6 +488,7 @@ static void sched_switch(int cpu, u64 timestamp, struct trace_entry *te)
 
 static int process_sample_event(union perf_event *event __used,
 				struct perf_sample *sample,
+				struct perf_evsel *evsel __used,
 				struct perf_session *session)
 {
 	struct trace_entry *te;
@@ -506,6 +507,16 @@ static int process_sample_event(union perf_event *event __used,
 		struct power_entry_old *peo;
 		peo = (void *)te;
 #endif
+		/*
+		 * FIXME: use evsel, its already mapped from id to perf_evsel,
+		 * remove perf_header__find_event infrastructure bits.
+		 * Mapping all these "power:cpu_idle" strings to the tracepoint
+		 * ID and then just comparing against evsel->attr.config.
+		 *
+		 * e.g.:
+		 *
+		 * if (evsel->attr.config == power_cpu_idle_id)
+		 */
 		event_str = perf_header__find_event(te->type);
 
 		if (!event_str)
diff --git a/tools/perf/util/build-id.c b/tools/perf/util/build-id.c
index 31f934a..a91cd99 100644
--- a/tools/perf/util/build-id.c
+++ b/tools/perf/util/build-id.c
@@ -16,6 +16,7 @@
 
 static int build_id__mark_dso_hit(union perf_event *event,
 				  struct perf_sample *sample __used,
+				  struct perf_evsel *evsel __used,
 				  struct perf_session *session)
 {
 	struct addr_location al;
diff --git a/tools/perf/util/hist.h b/tools/perf/util/hist.h
index cb6858a..3beb97c 100644
--- a/tools/perf/util/hist.h
+++ b/tools/perf/util/hist.h
@@ -29,6 +29,7 @@ struct events_stats {
 	u32 nr_events[PERF_RECORD_HEADER_MAX];
 	u32 nr_unknown_events;
 	u32 nr_invalid_chains;
+	u32 nr_unknown_id;
 };
 
 enum hist_column {
diff --git a/tools/perf/util/scripting-engines/trace-event-perl.c b/tools/perf/util/scripting-engines/trace-event-perl.c
index 6214272..74350ff 100644
--- a/tools/perf/util/scripting-engines/trace-event-perl.c
+++ b/tools/perf/util/scripting-engines/trace-event-perl.c
@@ -247,6 +247,7 @@ static inline struct event *find_cache_event(int type)
 
 static void perl_process_event(union perf_event *pevent __unused,
 			       struct perf_sample *sample,
+			       struct perf_evsel *evsel,
 			       struct perf_session *session __unused,
 			       struct thread *thread)
 {
diff --git a/tools/perf/util/scripting-engines/trace-event-python.c b/tools/perf/util/scripting-engines/trace-event-python.c
index 1b85d60..6ccf70e 100644
--- a/tools/perf/util/scripting-engines/trace-event-python.c
+++ b/tools/perf/util/scripting-engines/trace-event-python.c
@@ -206,6 +206,7 @@ static inline struct event *find_cache_event(int type)
 
 static void python_process_event(union perf_event *pevent __unused,
 				 struct perf_sample *sample,
+				 struct perf_evsel *evsel __unused,
 				 struct perf_session *session __unused,
 				 struct thread *thread)
 {
diff --git a/tools/perf/util/session.c b/tools/perf/util/session.c
index c68cf40..caa2245 100644
--- a/tools/perf/util/session.c
+++ b/tools/perf/util/session.c
@@ -280,6 +280,15 @@ static int process_event_synth_stub(union perf_event *event __used,
 	return 0;
 }
 
+static int process_event_sample_stub(union perf_event *event __used,
+				     struct perf_sample *sample __used,
+				     struct perf_evsel *evsel __used,
+				     struct perf_session *session __used)
+{
+	dump_printf(": unhandled!\n");
+	return 0;
+}
+
 static int process_event_stub(union perf_event *event __used,
 			      struct perf_sample *sample __used,
 			      struct perf_session *session __used)
@@ -303,7 +312,7 @@ static int process_finished_round(union perf_event *event,
 static void perf_event_ops__fill_defaults(struct perf_event_ops *handler)
 {
 	if (handler->sample == NULL)
-		handler->sample = process_event_stub;
+		handler->sample = process_event_sample_stub;
 	if (handler->mmap == NULL)
 		handler->mmap = process_event_stub;
 	if (handler->comm == NULL)
@@ -698,12 +707,19 @@ static int perf_session_deliver_event(struct perf_session *session,
 				      struct perf_event_ops *ops,
 				      u64 file_offset)
 {
+	struct perf_evsel *evsel;
+
 	dump_event(session, event, file_offset, sample);
 
 	switch (event->header.type) {
 	case PERF_RECORD_SAMPLE:
 		dump_sample(session, event, sample);
-		return ops->sample(event, sample, session);
+		evsel = perf_evlist__id2evsel(session->evlist, sample->id);
+		if (evsel == NULL) {
+			++session->hists.stats.nr_unknown_id;
+			return -1;
+		}
+		return ops->sample(event, sample, evsel, session);
 	case PERF_RECORD_MMAP:
 		return ops->mmap(event, sample, session);
 	case PERF_RECORD_COMM:
@@ -845,6 +861,11 @@ static void perf_session__warn_about_errors(const struct perf_session *session,
 			    session->hists.stats.nr_unknown_events);
 	}
 
+	if (session->hists.stats.nr_unknown_id != 0) {
+		ui__warning("%u samples with id not present in the header\n",
+			    session->hists.stats.nr_unknown_id);
+	}
+
  	if (session->hists.stats.nr_invalid_chains != 0) {
  		ui__warning("Found invalid callchains!\n\n"
  			    "%u out of %u events were discarded for this reason.\n\n"
diff --git a/tools/perf/util/session.h b/tools/perf/util/session.h
index 0b3c9af..1ac481f 100644
--- a/tools/perf/util/session.h
+++ b/tools/perf/util/session.h
@@ -55,8 +55,11 @@ struct perf_session {
 	char			filename[0];
 };
 
+struct perf_evsel;
 struct perf_event_ops;
 
+typedef int (*event_sample)(union perf_event *event, struct perf_sample *sample,
+			    struct perf_evsel *evsel, struct perf_session *session);
 typedef int (*event_op)(union perf_event *self, struct perf_sample *sample,
 			struct perf_session *session);
 typedef int (*event_synth_op)(union perf_event *self,
@@ -65,8 +68,8 @@ typedef int (*event_op2)(union perf_event *self, struct perf_session *session,
 			 struct perf_event_ops *ops);
 
 struct perf_event_ops {
-	event_op	sample,
-			mmap,
+	event_sample	sample;
+	event_op	mmap,
 			comm,
 			fork,
 			exit,
diff --git a/tools/perf/util/trace-event-scripting.c b/tools/perf/util/trace-event-scripting.c
index 66f4b78..c9dcbec 100644
--- a/tools/perf/util/trace-event-scripting.c
+++ b/tools/perf/util/trace-event-scripting.c
@@ -38,6 +38,7 @@ static int stop_script_unsupported(void)
 
 static void process_event_unsupported(union perf_event *event __unused,
 				      struct perf_sample *sample __unused,
+				      struct perf_evsel *evsel __unused,
 				      struct perf_session *session __unused,
 				      struct thread *thread __unused)
 {
diff --git a/tools/perf/util/trace-event.h b/tools/perf/util/trace-event.h
index b04da57..f674dda 100644
--- a/tools/perf/util/trace-event.h
+++ b/tools/perf/util/trace-event.h
@@ -280,6 +280,7 @@ struct scripting_ops {
 	int (*stop_script) (void);
 	void (*process_event) (union perf_event *event,
 			       struct perf_sample *sample,
+			       struct perf_evsel *evsel,
 			       struct perf_session *session,
 			       struct thread *thread);
 	int (*generate_script) (const char *outfile);
-- 
1.6.2.5

[PATCH 2/3] perf build-id: Add quirk to deal with perf.data file format breakage

[Arnaldo Carvalho de Melo]

From: Arnaldo Carvalho de Melo <acme@redhat.com>

The a1645ce1 changeset:

"perf: 'perf kvm' tool for monitoring guest performance from host"

Added a field to struct build_id_event that broke the file format.

Since the kernel build-id is the first entry, process the table using
the old format if the well known '[kernel.kallsyms]' string for the
kernel build-id has the first 4 characters chopped off (where the pid_t
sits).

Reported-by: Han Pingtian <phan@redhat.com>
Cc: Avi Kivity <avi@redhat.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Tom Zanussi <tzanussi@gmail.com>
Cc: Zhang Yanmin <yanmin_zhang@linux.intel.com>
Cc: stable@kernel.org
LKML-Reference: <new-submission>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/util/header.c |   57 +++++++++++++++++++++++++++++++++++++++++++++-
 1 files changed, 56 insertions(+), 1 deletions(-)

diff --git a/tools/perf/util/header.c b/tools/perf/util/header.c
index e5230c0..93862a8 100644
--- a/tools/perf/util/header.c
+++ b/tools/perf/util/header.c
@@ -695,13 +695,50 @@ out:
 	return err;
 }
 
+static int perf_header__read_build_ids_abi_quirk(struct perf_header *header,
+						 int input, u64 offset, u64 size)
+{
+	struct perf_session *session = container_of(header, struct perf_session, header);
+	struct {
+		struct perf_event_header   header;
+		u8			   build_id[ALIGN(BUILD_ID_SIZE, sizeof(u64))];
+		char			   filename[0];
+	} old_bev;
+	struct build_id_event bev;
+	char filename[PATH_MAX];
+	u64 limit = offset + size;
+
+	while (offset < limit) {
+		ssize_t len;
+
+		if (read(input, &old_bev, sizeof(old_bev)) != sizeof(old_bev))
+			return -1;
+
+		if (header->needs_swap)
+			perf_event_header__bswap(&old_bev.header);
+
+		len = old_bev.header.size - sizeof(old_bev);
+		if (read(input, filename, len) != len)
+			return -1;
+
+		bev.header = old_bev.header;
+		bev.pid	   = 0;
+		memcpy(bev.build_id, old_bev.build_id, sizeof(bev.build_id));
+		__event_process_build_id(&bev, filename, session);
+
+		offset += bev.header.size;
+	}
+
+	return 0;
+}
+
 static int perf_header__read_build_ids(struct perf_header *header,
 				       int input, u64 offset, u64 size)
 {
 	struct perf_session *session = container_of(header, struct perf_session, header);
 	struct build_id_event bev;
 	char filename[PATH_MAX];
-	u64 limit = offset + size;
+	u64 limit = offset + size, orig_offset = offset;
 	int err = -1;
 
 	while (offset < limit) {
@@ -716,6 +753,24 @@ static int perf_header__read_build_ids(struct perf_header *header,
 		len = bev.header.size - sizeof(bev);
 		if (read(input, filename, len) != len)
 			goto out;
+		/*
+		 * The a1645ce1 changeset:
+		 *
+		 * "perf: 'perf kvm' tool for monitoring guest performance from host"
+		 *
+		 * Added a field to struct build_id_event that broke the file
+		 * format.
+		 *
+		 * Since the kernel build-id is the first entry, process the
+		 * table using the old format if the well known
+		 * '[kernel.kallsyms]' string for the kernel build-id has the
+		 * first 4 characters chopped off (where the pid_t sits).
+		 */
+		if (memcmp(filename, "nel.kallsyms]", 13) == 0) {
+			if (lseek(input, orig_offset, SEEK_SET) == (off_t)-1)
+				return -1;
+			return perf_header__read_build_ids_abi_quirk(header, input, offset, size);
+		}
 
 		__event_process_build_id(&bev, filename, session);
 
-- 
1.6.2.5

[PATCH 3/3] perf symbols: Look at .dynsym again if .symtab not found

[Arnaldo Carvalho de Melo]

From: Arnaldo Carvalho de Melo <acme@redhat.com>

The original intent of the code was to repeat the search with
want_symtab = 0. But as the code stands now, we never hit the "default"
case of the switch statement. Which means we never repeat the search.

Tested-by: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
Reported-by: Arun Sharma <asharma@fb.com>
Reported-by: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
Cc: Dave Martin <dave.martin@linaro.org>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Tom Zanussi <tzanussi@gmail.com>
LKML-Reference: <new-submission>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/util/symbol.c |   25 +++++++++++++------------
 1 files changed, 13 insertions(+), 12 deletions(-)

diff --git a/tools/perf/util/symbol.c b/tools/perf/util/symbol.c
index 651dbfe..17df793 100644
--- a/tools/perf/util/symbol.c
+++ b/tools/perf/util/symbol.c
@@ -1486,7 +1486,9 @@ int dso__load(struct dso *self, struct map *map, symbol_filter_t filter)
 	 * On the first pass, only load images if they have a full symtab.
 	 * Failing that, do a second pass where we accept .dynsym also
 	 */
-	for (self->symtab_type = SYMTAB__BUILD_ID_CACHE, want_symtab = 1;
+	want_symtab = 1;
+restart:
+	for (self->symtab_type = SYMTAB__BUILD_ID_CACHE;
 	     self->symtab_type != SYMTAB__NOT_FOUND;
 	     self->symtab_type++) {
 		switch (self->symtab_type) {
@@ -1536,17 +1538,7 @@ int dso__load(struct dso *self, struct map *map, symbol_filter_t filter)
 			snprintf(name, size, "%s%s", symbol_conf.symfs,
 				 self->long_name);
 			break;
-
-		default:
-			/*
-			 * If we wanted a full symtab but no image had one,
-			 * relax our requirements and repeat the search.
-			 */
-			if (want_symtab) {
-				want_symtab = 0;
-				self->symtab_type = SYMTAB__BUILD_ID_CACHE;
-			} else
-				continue;
+		default:;
 		}
 
 		/* Name is now the name of the next image to try */
@@ -1573,6 +1565,15 @@ int dso__load(struct dso *self, struct map *map, symbol_filter_t filter)
 		}
 	}
 
+	/*
+	 * If we wanted a full symtab but no image had one,
+	 * relax our requirements and repeat the search.
+	 */
+	if (ret <= 0 && want_symtab) {
+		want_symtab = 0;
+		goto restart;
+	}
+
 	free(name);
 	if (ret < 0 && strstr(self->name, " (deleted)") != NULL)
 		return 0;
-- 
1.6.2.5

Re: [GIT PULL 0/3] perf/core fixes and improvements

[Ingo Molnar]

* Arnaldo Carvalho de Melo <acme@infradead.org> wrote:

> Hi Ingo,
> 
>         Please consider pulling from:
> 
> git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux-2.6 perf/core
> 
> Regards,
> 
> - Arnaldo
> 
> Arnaldo Carvalho de Melo (3):
>   perf session: Pass evsel in event_ops->sample()
>   perf build-id: Add quirk to deal with perf.data file format breakage
>   perf symbols: Look at .dynsym again if .symtab not found
> 
>  tools/perf/builtin-annotate.c                      |   18 ++-----
>  tools/perf/builtin-diff.c                          |    1 +
>  tools/perf/builtin-inject.c                        |   11 ++++-
>  tools/perf/builtin-kmem.c                          |    1 +
>  tools/perf/builtin-lock.c                          |    4 +-
>  tools/perf/builtin-report.c                        |   19 ++-----
>  tools/perf/builtin-sched.c                         |    1 +
>  tools/perf/builtin-script.c                        |   15 ++----
>  tools/perf/builtin-timechart.c                     |   11 ++++
>  tools/perf/util/build-id.c                         |    1 +
>  tools/perf/util/header.c                           |   57 +++++++++++++++++++-
>  tools/perf/util/hist.h                             |    1 +
>  .../perf/util/scripting-engines/trace-event-perl.c |    1 +
>  .../util/scripting-engines/trace-event-python.c    |    1 +
>  tools/perf/util/session.c                          |   25 ++++++++-
>  tools/perf/util/session.h                          |    7 ++-
>  tools/perf/util/symbol.c                           |   25 +++++----
>  tools/perf/util/trace-event-scripting.c            |    1 +
>  tools/perf/util/trace-event.h                      |    1 +
>  19 files changed, 142 insertions(+), 59 deletions(-)

Pulled, thanks a lot Arnaldo!

	Ingo

Re: [GIT PULL 0/3] perf/core fixes and improvements

[Arnaldo Carvalho de Melo]

Em Mon, Mar 19, 2012 at 08:47:25PM +0100, Ingo Molnar escreveu:
> Pulled, thanks Arnaldo!
> 
> I noticed that there's no help text for GTK support - how will a 
> user find that 'perf report --gtk' will do something nice?
> 
> Another detail is that if GTK support is not compiled in:
> 
> Makefile:515: GTK2 not found, disables GTK2 support. Please install gtk2-devel or libgtk2.0-dev
> 
> then the --gtk output is not very informative:
> 
>   earth5:~/tip/tools/perf> perf report --gtk
>   earth5:~/tip/tools/perf> 
> 
> :-)
> 
> A third detail, I do have gtk2-devel installed:
> 
>  Package gtk2-devel-2.24.10-1.fc17.x86_64 already installed and latest version

Yeah, as you said, the ball is rolling! ;-)

> yet I got the above message.
> 
> It is not easy to figure out *why* a feature test failed, 
> unfortunately. It would be nice if V=1 or something like that 
> would output the failure.

Yeah, I also was frustrated by V=1 not affecting the feature tests, and
did what you did, just didn't got around to fix it properly.
 
> I applied the hack below and ran the gtk testcase, which gave:
> 
> earth5:~/tip/tools/perf> . ./test.4273.sh
> In file included from /usr/lib64/glib-2.0/include/glibconfig.h:9:0,
>                  from /usr/include/glib-2.0/glib/gtypes.h:34,
>                  from /usr/include/glib-2.0/glib/galloca.h:34,
>                  from /usr/include/glib-2.0/glib.h:32,
>                  from /usr/include/glib-2.0/gobject/gbinding.h:30,
>                  from /usr/include/glib-2.0/glib-object.h:25,
>                  from /usr/include/glib-2.0/gio/gioenums.h:30,
>                  from /usr/include/glib-2.0/gio/giotypes.h:30,
>                  from /usr/include/glib-2.0/gio/gio.h:28,
>                  from /usr/include/gtk-2.0/gdk/gdkapplaunchcontext.h:30,
>                  from /usr/include/gtk-2.0/gdk/gdk.h:32,
>                  from /usr/include/gtk-2.0/gtk/gtk.h:32,
>                  from test.4273.c:2:
> /usr/include/glib-2.0/glib/gmacros.h:346:7: error: "_MSC_VER" is not defined [-Werror=undef]
> cc1: all warnings being treated as errors
> 
> Which allowed me to fix the feature test and the gtk.h include 
> file wrapper via the second patch below.
> 
> And then I was greeted by the GTK report window on 'perf report 
> --gtk' ;-)
> 
> Thanks,
> 
> 	Ingo
> 
> diff --git a/tools/perf/config/utilities.mak b/tools/perf/config/utilities.mak
> index 8046182..68cf795 100644
> --- a/tools/perf/config/utilities.mak
> +++ b/tools/perf/config/utilities.mak
> @@ -183,6 +183,8 @@ _gea_err  = $(if $(1),$(error Please set '$(1)' appropriately))
>  # Usage: option = $(call try-cc, source-to-build, cc-options)
>  try-cc = $(shell sh -c						  \
>  	'TMP="$(OUTPUT)$(TMPOUT).$$$$";				  \
> +	 echo "$(1)" > test.$$$$.c;					  \
> +	 echo $(CC) -x c test.$$$$.c $(2) -o "$$TMP" > test.$$$$.sh;		  \
>  	 echo "$(1)" |						  \
>  	 $(CC) -x c - $(2) -o "$$TMP" > /dev/null 2>&1 && echo y; \
>  	 rm -f "$$TMP"')
> 
> Signed-off-by: Ingo Molnar <mingo@elte.hu>
> 
> diff --git a/tools/perf/config/feature-tests.mak b/tools/perf/config/feature-tests.mak
> index d9084e0..ae8b471 100644
> --- a/tools/perf/config/feature-tests.mak
> +++ b/tools/perf/config/feature-tests.mak
> @@ -68,7 +68,9 @@ endif
>  ifndef NO_GTK2
>  define SOURCE_GTK2
>  #pragma GCC diagnostic ignored \"-Wstrict-prototypes\"
> +#pragma GCC diagnostic ignored \"-Wundef\"
>  #include <gtk/gtk.h>
> +#pragma GCC diagnostic error \"-Wundef\"
>  #pragma GCC diagnostic error \"-Wstrict-prototypes\"
>  
>  int main(int argc, char *argv[])
> diff --git a/tools/perf/util/gtk/gtk.h b/tools/perf/util/gtk/gtk.h
> index 75177ee..c7a941f 100644
> --- a/tools/perf/util/gtk/gtk.h
> +++ b/tools/perf/util/gtk/gtk.h
> @@ -2,7 +2,9 @@
>  #define _PERF_GTK_H_ 1
>  
>  #pragma GCC diagnostic ignored "-Wstrict-prototypes"
> +#pragma GCC diagnostic ignored "-Wundef"
>  #include <gtk/gtk.h>
> +#pragma GCC diagnostic error "-Wundef"
>  #pragma GCC diagnostic error "-Wstrict-prototypes"
>  
>  #endif /* _PERF_GTK_H_ */

Re: [GIT PULL 0/3] perf/core fixes and improvements

[Ingo Molnar]

* Ingo Molnar <mingo@kernel.org> wrote:

> I noticed that there's no help text for GTK support - how will a 
> user find that 'perf report --gtk' will do something nice?

Ignore this one - I forgot to do 'make install-doc'.

Thanks,

	Ingo

Re: [GIT PULL 0/3] perf/core fixes and improvements

[Ingo Molnar]

* Ingo Molnar <mingo@kernel.org> wrote:

> diff --git a/tools/perf/config/utilities.mak b/tools/perf/config/utilities.mak
> index 8046182..68cf795 100644
> --- a/tools/perf/config/utilities.mak
> +++ b/tools/perf/config/utilities.mak
> @@ -183,6 +183,8 @@ _gea_err  = $(if $(1),$(error Please set '$(1)' appropriately))
>  # Usage: option = $(call try-cc, source-to-build, cc-options)
>  try-cc = $(shell sh -c						  \
>  	'TMP="$(OUTPUT)$(TMPOUT).$$$$";				  \
> +	 echo "$(1)" > test.$$$$.c;					  \
> +	 echo $(CC) -x c test.$$$$.c $(2) -o "$$TMP" > test.$$$$.sh;		  \
>  	 echo "$(1)" |						  \
>  	 $(CC) -x c - $(2) -o "$$TMP" > /dev/null 2>&1 && echo y; \
>  	 rm -f "$$TMP"')

So, instead of this hack we'd like to output the build failure 
that gcc gives, when 'make V=1' is specified or so.

Thanks,

	Ingo

Re: [GIT PULL 0/3] perf/core fixes and improvements

[Ingo Molnar]

* Arnaldo Carvalho de Melo <acme@infradead.org> wrote:

> Hi Ingo,
> 
> 	Please consider pulling, now there are two pull request worth of
> changesets in my perf/core branch,
> 
> - Arnaldo
> 
> The following changes since commit 6db6127c4dad634ab98709b81e2f2770890b0d53:
> 
>   perf report: Treat an argument as a symbol filter (2012-03-16 16:44:36 -0300)
> 
> are available in the git repository at:
> 
>   git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux tags/perf-core-for-mingo
> 
> for you to fetch changes up to c31a94570552dcaa517c4f7a043ffd28835016be:
> 
>   perf report: Add a simple GTK2-based 'perf report' browser (2012-03-19 15:13:29 -0300)
> 
> ----------------------------------------------------------------
> Fixes for the last batch from Namhyung and the initial GTK report browser
> from Pekka.
> 
> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
> 
> ----------------------------------------------------------------
> Namhyung Kim (2):
>       perf ui browser: Clean lines inside of the input window
>       perf report: Document --symbol-filter option
> 
> Pekka Enberg (1):
>       perf report: Add a simple GTK2-based 'perf report' browser
> 
>  tools/perf/Documentation/perf-report.txt |    5 +
>  tools/perf/Makefile                      |   14 +++
>  tools/perf/builtin-report.c              |   19 ++-
>  tools/perf/config/feature-tests.mak      |   15 +++
>  tools/perf/util/cache.h                  |   12 ++
>  tools/perf/util/gtk/browser.c            |  189 ++++++++++++++++++++++++++++++
>  tools/perf/util/gtk/gtk.h                |    8 ++
>  tools/perf/util/hist.h                   |   17 +++
>  tools/perf/util/ui/util.c                |   10 +-
>  9 files changed, 282 insertions(+), 7 deletions(-)
>  create mode 100644 tools/perf/util/gtk/browser.c
>  create mode 100644 tools/perf/util/gtk/gtk.h

Pulled, thanks Arnaldo!

I noticed that there's no help text for GTK support - how will a 
user find that 'perf report --gtk' will do something nice?

Another detail is that if GTK support is not compiled in:

Makefile:515: GTK2 not found, disables GTK2 support. Please install gtk2-devel or libgtk2.0-dev

then the --gtk output is not very informative:

  earth5:~/tip/tools/perf> perf report --gtk
  earth5:~/tip/tools/perf> 

:-)

A third detail, I do have gtk2-devel installed:

 Package gtk2-devel-2.24.10-1.fc17.x86_64 already installed and latest version

yet I got the above message.

It is not easy to figure out *why* a feature test failed, 
unfortunately. It would be nice if V=1 or something like that 
would output the failure.

I applied the hack below and ran the gtk testcase, which gave:

earth5:~/tip/tools/perf> . ./test.4273.sh
In file included from /usr/lib64/glib-2.0/include/glibconfig.h:9:0,
                 from /usr/include/glib-2.0/glib/gtypes.h:34,
                 from /usr/include/glib-2.0/glib/galloca.h:34,
                 from /usr/include/glib-2.0/glib.h:32,
                 from /usr/include/glib-2.0/gobject/gbinding.h:30,
                 from /usr/include/glib-2.0/glib-object.h:25,
                 from /usr/include/glib-2.0/gio/gioenums.h:30,
                 from /usr/include/glib-2.0/gio/giotypes.h:30,
                 from /usr/include/glib-2.0/gio/gio.h:28,
                 from /usr/include/gtk-2.0/gdk/gdkapplaunchcontext.h:30,
                 from /usr/include/gtk-2.0/gdk/gdk.h:32,
                 from /usr/include/gtk-2.0/gtk/gtk.h:32,
                 from test.4273.c:2:
/usr/include/glib-2.0/glib/gmacros.h:346:7: error: "_MSC_VER" is not defined [-Werror=undef]
cc1: all warnings being treated as errors

Which allowed me to fix the feature test and the gtk.h include 
file wrapper via the second patch below.

And then I was greeted by the GTK report window on 'perf report 
--gtk' ;-)

Thanks,

	Ingo

diff --git a/tools/perf/config/utilities.mak b/tools/perf/config/utilities.mak
index 8046182..68cf795 100644
--- a/tools/perf/config/utilities.mak
+++ b/tools/perf/config/utilities.mak
@@ -183,6 +183,8 @@ _gea_err  = $(if $(1),$(error Please set '$(1)' appropriately))
 # Usage: option = $(call try-cc, source-to-build, cc-options)
 try-cc = $(shell sh -c						  \
 	'TMP="$(OUTPUT)$(TMPOUT).$$$$";				  \
+	 echo "$(1)" > test.$$$$.c;					  \
+	 echo $(CC) -x c test.$$$$.c $(2) -o "$$TMP" > test.$$$$.sh;		  \
 	 echo "$(1)" |						  \
 	 $(CC) -x c - $(2) -o "$$TMP" > /dev/null 2>&1 && echo y; \
 	 rm -f "$$TMP"')

Signed-off-by: Ingo Molnar <mingo@elte.hu>

diff --git a/tools/perf/config/feature-tests.mak b/tools/perf/config/feature-tests.mak
index d9084e0..ae8b471 100644
--- a/tools/perf/config/feature-tests.mak
+++ b/tools/perf/config/feature-tests.mak
@@ -68,7 +68,9 @@ endif
 ifndef NO_GTK2
 define SOURCE_GTK2
 #pragma GCC diagnostic ignored \"-Wstrict-prototypes\"
+#pragma GCC diagnostic ignored \"-Wundef\"
 #include <gtk/gtk.h>
+#pragma GCC diagnostic error \"-Wundef\"
 #pragma GCC diagnostic error \"-Wstrict-prototypes\"
 
 int main(int argc, char *argv[])
diff --git a/tools/perf/util/gtk/gtk.h b/tools/perf/util/gtk/gtk.h
index 75177ee..c7a941f 100644
--- a/tools/perf/util/gtk/gtk.h
+++ b/tools/perf/util/gtk/gtk.h
@@ -2,7 +2,9 @@
 #define _PERF_GTK_H_ 1
 
 #pragma GCC diagnostic ignored "-Wstrict-prototypes"
+#pragma GCC diagnostic ignored "-Wundef"
 #include <gtk/gtk.h>
+#pragma GCC diagnostic error "-Wundef"
 #pragma GCC diagnostic error "-Wstrict-prototypes"
 
 #endif /* _PERF_GTK_H_ */

[GIT PULL 0/3] perf/core fixes and improvements

[Arnaldo Carvalho de Melo]

Hi Ingo,

	Please consider pulling, now there are two pull request worth of
changesets in my perf/core branch,

- Arnaldo

The following changes since commit 6db6127c4dad634ab98709b81e2f2770890b0d53:

  perf report: Treat an argument as a symbol filter (2012-03-16 16:44:36 -0300)

are available in the git repository at:

  git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux tags/perf-core-for-mingo

for you to fetch changes up to c31a94570552dcaa517c4f7a043ffd28835016be:

  perf report: Add a simple GTK2-based 'perf report' browser (2012-03-19 15:13:29 -0300)

----------------------------------------------------------------
Fixes for the last batch from Namhyung and the initial GTK report browser
from Pekka.

Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

----------------------------------------------------------------
Namhyung Kim (2):
      perf ui browser: Clean lines inside of the input window
      perf report: Document --symbol-filter option

Pekka Enberg (1):
      perf report: Add a simple GTK2-based 'perf report' browser

 tools/perf/Documentation/perf-report.txt |    5 +
 tools/perf/Makefile                      |   14 +++
 tools/perf/builtin-report.c              |   19 ++-
 tools/perf/config/feature-tests.mak      |   15 +++
 tools/perf/util/cache.h                  |   12 ++
 tools/perf/util/gtk/browser.c            |  189 ++++++++++++++++++++++++++++++
 tools/perf/util/gtk/gtk.h                |    8 ++
 tools/perf/util/hist.h                   |   17 +++
 tools/perf/util/ui/util.c                |   10 +-
 9 files changed, 282 insertions(+), 7 deletions(-)
 create mode 100644 tools/perf/util/gtk/browser.c
 create mode 100644 tools/perf/util/gtk/gtk.h

Re: [GIT PULL 0/3] perf/core fixes and improvements

[Ingo Molnar]

* Arnaldo Carvalho de Melo <acme@infradead.org> wrote:

> The following changes since commit 09bda4432a8a4d4db2b2b94697abc8d732a9ff73:
> 
>   Merge branch 'tip/perf/core' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace into perf/core (2012-02-17 12:55:07 +0100)
> 
> are available in the git repository at:
> 
> 
>   git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux perf-core-for-mingo
> 
> for you to fetch changes up to 6b1bee9035d430c4b4f586df6df4b3f840e89b5b:
> 
>   perf tools: fix broken perf record -a mode (2012-02-21 15:05:43 -0200)
> 
> ----------------------------------------------------------------
> Important fix from Stephane for system wide monitoring, problem
> introduced recently in the pid list patches.
> 
> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
> 
> ----------------------------------------------------------------
> Danny Kukawka (1):
>       perf tools: Remove duplicated string.h includes
> 
> Stefan Hajnoczi (1):
>       perf tools: Allow expressions in __print_symbolic() fields
> 
> Stephane Eranian (1):
>       perf tools: fix broken perf record -a mode
> 
>  tools/perf/util/probe-event.c       |    1 -
>  tools/perf/util/thread_map.c        |    2 +-
>  tools/perf/util/trace-event-parse.c |   12 ++++++++++++
>  3 files changed, 13 insertions(+), 2 deletions(-)

Pulled, thanks a lot Arnaldo!

	Ingo

[GIT PULL 0/3] perf/core fixes and improvements

[Arnaldo Carvalho de Melo]

The following changes since commit 09bda4432a8a4d4db2b2b94697abc8d732a9ff73:

  Merge branch 'tip/perf/core' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace into perf/core (2012-02-17 12:55:07 +0100)

are available in the git repository at:


  git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux perf-core-for-mingo

for you to fetch changes up to 6b1bee9035d430c4b4f586df6df4b3f840e89b5b:

  perf tools: fix broken perf record -a mode (2012-02-21 15:05:43 -0200)

----------------------------------------------------------------
Important fix from Stephane for system wide monitoring, problem
introduced recently in the pid list patches.

Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

----------------------------------------------------------------
Danny Kukawka (1):
      perf tools: Remove duplicated string.h includes

Stefan Hajnoczi (1):
      perf tools: Allow expressions in __print_symbolic() fields

Stephane Eranian (1):
      perf tools: fix broken perf record -a mode

 tools/perf/util/probe-event.c       |    1 -
 tools/perf/util/thread_map.c        |    2 +-
 tools/perf/util/trace-event-parse.c |   12 ++++++++++++
 3 files changed, 13 insertions(+), 2 deletions(-)

Re: [GIT PULL 0/3] perf/core fixes and improvements

[Ingo Molnar]

* Arnaldo Carvalho de Melo <acme@infradead.org> wrote:

> Hi Ingo,
> 
>         Please consider pulling from:
> 
> git://github.com/acmel/linux.git perf/core
> 
> Regards,
> 
> - Arnaldo
> 
> Andrew Vagin (1):
>   perf tools: Add ability to synthesize event according to a sample
> 
> Robert Richter (2):
>   perf script: Fix mem leaks and NULL pointer checks around strdup()s
>   perf script: Implement option for system-wide profiling
> 
>  tools/perf/builtin-script.c |   63 +++++++++++++++++++++-------------
>  tools/perf/util/event.h     |    3 ++
>  tools/perf/util/evsel.c     |   79 +++++++++++++++++++++++++++++++++++++++++++
>  tools/perf/util/session.h   |    8 ++++
>  4 files changed, 129 insertions(+), 24 deletions(-)

Pulled, thanks a lot Arnaldo!

	Ingo

[GIT PULL 0/3] perf/core fixes and improvements

[Arnaldo Carvalho de Melo]

Hi Ingo,

        Please consider pulling from:

git://github.com/acmel/linux.git perf/core

Regards,

- Arnaldo

Andrew Vagin (1):
  perf tools: Add ability to synthesize event according to a sample

Robert Richter (2):
  perf script: Fix mem leaks and NULL pointer checks around strdup()s
  perf script: Implement option for system-wide profiling

 tools/perf/builtin-script.c |   63 +++++++++++++++++++++-------------
 tools/perf/util/event.h     |    3 ++
 tools/perf/util/evsel.c     |   79 +++++++++++++++++++++++++++++++++++++++++++
 tools/perf/util/session.h   |    8 ++++
 4 files changed, 129 insertions(+), 24 deletions(-)

-- 
1.7.8.rc0.35.gee6df

Re: [GIT PULL 0/3] perf/core fixes and improvements

[Ingo Molnar]

* Arnaldo Carvalho de Melo <acme@infradead.org> wrote:

> Hi Ingo,
> 
>         Please consider pulling from:
> 
> git://github.com/acmel/linux.git perf/core
> 
> Regards,
> 
> - Arnaldo
> 
> Arnaldo Carvalho de Melo (2):
>   perf hists browser: Warn about lost events
>   perf report: Add progress bar when processing time ordered events
> 
> Masami Hiramatsu (1):
>   perf tools: Fix a typo of command name as trace-cmd
> 
>  tools/perf/builtin-top.c            |   41 +++++++++++++++++++++--------
>  tools/perf/util/hist.h              |    1 +
>  tools/perf/util/session.c           |   38 ++++++++++++++++++++++----
>  tools/perf/util/session.h           |    1 +
>  tools/perf/util/top.h               |    1 -
>  tools/perf/util/trace-event-info.c  |    2 +-
>  tools/perf/util/ui/browser.c        |   21 ++++++++++++---
>  tools/perf/util/ui/browser.h        |    3 +-
>  tools/perf/util/ui/browsers/hists.c |   49 ++++++++++++++++++++++++++++++----
>  9 files changed, 127 insertions(+), 30 deletions(-)

Pulled, thanks a lot Arnaldo!

	Ingo

[GIT PULL 0/3] perf/core fixes and improvements

[Arnaldo Carvalho de Melo]

Hi Ingo,

        Please consider pulling from:

git://github.com/acmel/linux.git perf/core

Regards,

- Arnaldo

Arnaldo Carvalho de Melo (2):
  perf hists browser: Warn about lost events
  perf report: Add progress bar when processing time ordered events

Masami Hiramatsu (1):
  perf tools: Fix a typo of command name as trace-cmd

 tools/perf/builtin-top.c            |   41 +++++++++++++++++++++--------
 tools/perf/util/hist.h              |    1 +
 tools/perf/util/session.c           |   38 ++++++++++++++++++++++----
 tools/perf/util/session.h           |    1 +
 tools/perf/util/top.h               |    1 -
 tools/perf/util/trace-event-info.c  |    2 +-
 tools/perf/util/ui/browser.c        |   21 ++++++++++++---
 tools/perf/util/ui/browser.h        |    3 +-
 tools/perf/util/ui/browsers/hists.c |   49 ++++++++++++++++++++++++++++++----
 9 files changed, 127 insertions(+), 30 deletions(-)

Re: [GIT PULL 0/3] perf/core fixes and improvements

[Ingo Molnar]

* Arnaldo Carvalho de Melo <acme@infradead.org> wrote:

> Hi Ingo,
> 
>         Please consider pulling from:
> 
> git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux-2.6 perf/core
> 
> Regards,
> 
> - Arnaldo
> 
> Arnaldo Carvalho de Melo (2):
>   perf symbols: Improve kallsyms symbol end addr calculation
>   perf test: Look forward for symbol aliases
> 
> Franck Bui-Huu (1):
>   perf probe: Fix wrong warning in __show_one_line() if read(1) errors
>     happen
> 
>  tools/perf/builtin-test.c     |   23 ++++++++++++++---
>  tools/perf/util/event.c       |    3 +-
>  tools/perf/util/probe-event.c |    2 +-
>  tools/perf/util/symbol.c      |   56 ++++++++++++++++++++++++++++++----------
>  tools/perf/util/symbol.h      |    2 +-
>  5 files changed, 65 insertions(+), 21 deletions(-)

Pulled, thanks a lot Arnaldo!

	Ingo

[GIT PULL 0/3] perf/core fixes and improvements

[Arnaldo Carvalho de Melo]

Hi Ingo,

        Please consider pulling from:

git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux-2.6 perf/core

Regards,

- Arnaldo

Arnaldo Carvalho de Melo (2):
  perf symbols: Improve kallsyms symbol end addr calculation
  perf test: Look forward for symbol aliases

Franck Bui-Huu (1):
  perf probe: Fix wrong warning in __show_one_line() if read(1) errors
    happen

 tools/perf/builtin-test.c     |   23 ++++++++++++++---
 tools/perf/util/event.c       |    3 +-
 tools/perf/util/probe-event.c |    2 +-
 tools/perf/util/symbol.c      |   56 ++++++++++++++++++++++++++++++----------
 tools/perf/util/symbol.h      |    2 +-
 5 files changed, 65 insertions(+), 21 deletions(-)

[GIT PULL 0/3] perf/core fixes and improvements

[Arnaldo Carvalho de Melo]

Hi Ingo,

        Please pull from:

git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux-2.6 perf/core

Regards,

- Arnaldo

Arnaldo Carvalho de Melo (3):
  perf ui browser: Return the exit key in all browsers
  perf ui browser: Add routines to compactly specify exit keys
  perf ui browser: Abstract some more slang operations

 tools/perf/builtin-annotate.c          |   17 ++--
 tools/perf/util/ui/browser.c           |   93 +++++++++++--------
 tools/perf/util/ui/browser.h           |    9 ++-
 tools/perf/util/ui/browsers/annotate.c |   37 ++++----
 tools/perf/util/ui/browsers/hists.c    |  157 ++++++++++++++------------------
 tools/perf/util/ui/browsers/map.c      |   23 ++---
 tools/perf/util/ui/util.c              |    4 +-
 tools/perf/util/util.h                 |   13 ---
 8 files changed, 161 insertions(+), 192 deletions(-)

Re: [GIT PULL 0/3] perf/core fixes and improvements

[Ingo Molnar]

* Arnaldo Carvalho de Melo <acme@infradead.org> wrote:

> 
> Hi Ingo,
> 
>         Please pull from:
> 
> git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux-2.6 perf/core
> 
> Regards,
> 
> - Arnaldo
> 
> Dave Martin (1):
>   perf events: Fix mmap offset determination
> 
> Srikar Dronamraju (2):
>   perf: expose event__process function
>   tracing/kprobes: unregister_trace_probe needs to be called under mutex
> 
>  kernel/trace/trace_kprobe.c |    3 +++
>  tools/perf/builtin-top.c    |   20 --------------------
>  tools/perf/util/event.c     |   28 +++++++++++++++++++++-------
>  tools/perf/util/event.h     |    1 +
>  4 files changed, 25 insertions(+), 27 deletions(-)

Pulled, thanks a lot Arnaldo!

	Ingo

[GIT PULL 0/3] perf/core fixes and improvements

[Arnaldo Carvalho de Melo]

Hi Ingo,

        Please pull from:

git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux-2.6 perf/core

Regards,

- Arnaldo

Dave Martin (1):
  perf events: Fix mmap offset determination

Srikar Dronamraju (2):
  perf: expose event__process function
  tracing/kprobes: unregister_trace_probe needs to be called under mutex

 kernel/trace/trace_kprobe.c |    3 +++
 tools/perf/builtin-top.c    |   20 --------------------
 tools/perf/util/event.c     |   28 +++++++++++++++++++++-------
 tools/perf/util/event.h     |    1 +
 4 files changed, 25 insertions(+), 27 deletions(-)

Re: [GIT PULL 0/3] perf/core fixes and improvements

[Ingo Molnar]

* Arnaldo Carvalho de Melo <acme@infradead.org> wrote:

> Hi Ingo,
> 
>         Please pull from:
> 
> git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux-2.6 perf/core
> 
> Regards,
> 
> - Arnaldo
> 
> Arnaldo Carvalho de Melo (3):
>   perf session: Free the ref_reloc_sym memory at the right place
>   perf session: Invalidate last_match when removing threads from rb_tree
>   perf tools: Don't keep unreferenced maps when unmaps are detected
> 
>  tools/perf/util/hist.c    |    2 +
>  tools/perf/util/map.c     |   49 +++++++++++++++++++++++++++++---------------
>  tools/perf/util/map.h     |   10 ++++++++-
>  tools/perf/util/session.c |    8 +++++++
>  tools/perf/util/symbol.c  |   43 +++++++++++++++++++++++++++++++++++++++
>  tools/perf/util/symbol.h  |    2 +
>  6 files changed, 96 insertions(+), 18 deletions(-)

Pulled, thanks Arnaldo!

	Ingo

[GIT PULL 0/3] perf/core fixes and improvements

[Arnaldo Carvalho de Melo]

Hi Ingo,

        Please pull from:

git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux-2.6 perf/core

Regards,

- Arnaldo

Arnaldo Carvalho de Melo (3):
  perf session: Free the ref_reloc_sym memory at the right place
  perf session: Invalidate last_match when removing threads from rb_tree
  perf tools: Don't keep unreferenced maps when unmaps are detected

 tools/perf/util/hist.c    |    2 +
 tools/perf/util/map.c     |   49 +++++++++++++++++++++++++++++---------------
 tools/perf/util/map.h     |   10 ++++++++-
 tools/perf/util/session.c |    8 +++++++
 tools/perf/util/symbol.c  |   43 +++++++++++++++++++++++++++++++++++++++
 tools/perf/util/symbol.h  |    2 +
 6 files changed, 96 insertions(+), 18 deletions(-)

Re: [GIT PULL 0/3] perf/core fixes and improvements

[Ingo Molnar]

* Arnaldo Carvalho de Melo <acme@infradead.org> wrote:

> Hi Ingo,
> 
>         Please pull from:
> 
> git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux-2.6 perf/core
> 
> Regards,
> 
> - Arnaldo
> 
> Masami Hiramatsu (3):
>   perf probe: Fix error message if get_real_path() failed
>   perf probe: Support comp_dir to find an absolute source path
>   perf probe: Fix the logic of die_compare_name
> 
>  tools/perf/util/probe-event.c  |   69 ++++++++++++++++++++++++++++
>  tools/perf/util/probe-event.h  |    1 +
>  tools/perf/util/probe-finder.c |   96 ++++++++++++++--------------------------
>  3 files changed, 103 insertions(+), 63 deletions(-)

Pulled, thanks Arnaldo!

	Ingo

[GIT PULL 0/3] perf/core fixes and improvements

[Arnaldo Carvalho de Melo]

Hi Ingo,

        Please pull from:

git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux-2.6 perf/core

Regards,

- Arnaldo

Masami Hiramatsu (3):
  perf probe: Fix error message if get_real_path() failed
  perf probe: Support comp_dir to find an absolute source path
  perf probe: Fix the logic of die_compare_name

 tools/perf/util/probe-event.c  |   69 ++++++++++++++++++++++++++++
 tools/perf/util/probe-event.h  |    1 +
 tools/perf/util/probe-finder.c |   96 ++++++++++++++--------------------------
 3 files changed, 103 insertions(+), 63 deletions(-)