[GIT PULL 00/16] perf/core improvements and fixes

[GIT PULL 00/16] perf/core improvements and fixes

[Arnaldo Carvalho de Melo]

Hi Ingo,

        Please consider pulling from:

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

	This was rebased against your perf/core on tesla.

Regards,

- Arnaldo

Andi Kleen (2):
  perf tools: Make stat/record print fatal signals of the target program
  perf: Support setting the disassembler style

Arnaldo Carvalho de Melo (6):
  perf symbols: Stop using 'self' in map_groups__ methods
  perf buildid-list: Add option to show the running kernel build id
  perf buildid-list: Support showing the build id in an ELF file
  perf top browser: Fix up line width calculation
  perf top: Improve lost events warning
  perf report: Fix stdio event name header printing

Jim Cromie (5):
  perf stat: Add --log-fd <N> option to redirect stderr elsewhere
  perf stat: Fix +- nan% in --no-aggr runs
  perf stat: Suppress printing std-dev when its 0
  perf stat: Allow tab as cvs delimiter
  perf stat: Fix spelling in comment

Jiri Olsa (2):
  perf sched: Fix script command documentation
  perf symbols: Treat all memory maps without dso file as loaded

Neil Horman (1):
  perf script: Add drop monitor script

 tools/perf/Documentation/perf-annotate.txt         |    3 +
 tools/perf/Documentation/perf-buildid-list.txt     |    6 +
 tools/perf/Documentation/perf-report.txt           |    3 +
 tools/perf/Documentation/perf-sched.txt            |    6 +-
 tools/perf/Documentation/perf-stat.txt             |   11 ++-
 tools/perf/builtin-annotate.c                      |    2 +
 tools/perf/builtin-buildid-list.c                  |   52 ++++++++++-
 tools/perf/builtin-record.c                        |   19 ++++-
 tools/perf/builtin-report.c                        |    7 +-
 tools/perf/builtin-stat.c                          |   35 ++++++-
 tools/perf/builtin-top.c                           |   34 ++++----
 .../perf/scripts/python/bin/net_dropmonitor-record |    2 +
 .../perf/scripts/python/bin/net_dropmonitor-report |    4 +
 tools/perf/scripts/python/net_dropmonitor.py       |   72 ++++++++++++++
 tools/perf/util/annotate.c                         |    6 +-
 tools/perf/util/annotate.h                         |    2 +
 tools/perf/util/map.c                              |   98 +++++++++++---------
 tools/perf/util/map.h                              |   42 ++++----
 tools/perf/util/top.h                              |    3 +
 tools/perf/util/ui/browsers/top.c                  |   28 +++++-
 20 files changed, 333 insertions(+), 102 deletions(-)
 create mode 100755 tools/perf/scripts/python/bin/net_dropmonitor-record
 create mode 100755 tools/perf/scripts/python/bin/net_dropmonitor-report
 create mode 100755 tools/perf/scripts/python/net_dropmonitor.py

[PATCH 01/16] perf symbols: Stop using 'self' in map_groups__ methods

[Arnaldo Carvalho de Melo]

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

Stop using this python/OOP convention, doesn't really helps. Will do
more from time to time till we get it cleaned up in all of /perf.

Suggested-by: Thomas Gleixner <tglx@linutronix.de>
Cc: David Ahern <dsahern@gmail.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Link: http://lkml.kernel.org/n/tip-rl9e690y60vnuyng05yp1zd3@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/util/map.c |   74 ++++++++++++++++++++++++------------------------
 tools/perf/util/map.h |   42 ++++++++++++++--------------
 2 files changed, 58 insertions(+), 58 deletions(-)

diff --git a/tools/perf/util/map.c b/tools/perf/util/map.c
index a16ecab..aa2f9fd 100644
--- a/tools/perf/util/map.c
+++ b/tools/perf/util/map.c
@@ -220,55 +220,55 @@ u64 map__objdump_2ip(struct map *map, u64 addr)
 	return ip;
 }
 
-void map_groups__init(struct map_groups *self)
+void map_groups__init(struct map_groups *mg)
 {
 	int i;
 	for (i = 0; i < MAP__NR_TYPES; ++i) {
-		self->maps[i] = RB_ROOT;
-		INIT_LIST_HEAD(&self->removed_maps[i]);
+		mg->maps[i] = RB_ROOT;
+		INIT_LIST_HEAD(&mg->removed_maps[i]);
 	}
-	self->machine = NULL;
+	mg->machine = NULL;
 }
 
-static void maps__delete(struct rb_root *self)
+static void maps__delete(struct rb_root *maps)
 {
-	struct rb_node *next = rb_first(self);
+	struct rb_node *next = rb_first(maps);
 
 	while (next) {
 		struct map *pos = rb_entry(next, struct map, rb_node);
 
 		next = rb_next(&pos->rb_node);
-		rb_erase(&pos->rb_node, self);
+		rb_erase(&pos->rb_node, maps);
 		map__delete(pos);
 	}
 }
 
-static void maps__delete_removed(struct list_head *self)
+static void maps__delete_removed(struct list_head *maps)
 {
 	struct map *pos, *n;
 
-	list_for_each_entry_safe(pos, n, self, node) {
+	list_for_each_entry_safe(pos, n, maps, node) {
 		list_del(&pos->node);
 		map__delete(pos);
 	}
 }
 
-void map_groups__exit(struct map_groups *self)
+void map_groups__exit(struct map_groups *mg)
 {
 	int i;
 
 	for (i = 0; i < MAP__NR_TYPES; ++i) {
-		maps__delete(&self->maps[i]);
-		maps__delete_removed(&self->removed_maps[i]);
+		maps__delete(&mg->maps[i]);
+		maps__delete_removed(&mg->removed_maps[i]);
 	}
 }
 
-void map_groups__flush(struct map_groups *self)
+void map_groups__flush(struct map_groups *mg)
 {
 	int type;
 
 	for (type = 0; type < MAP__NR_TYPES; type++) {
-		struct rb_root *root = &self->maps[type];
+		struct rb_root *root = &mg->maps[type];
 		struct rb_node *next = rb_first(root);
 
 		while (next) {
@@ -280,17 +280,17 @@ void map_groups__flush(struct map_groups *self)
 			 * instance in some hist_entry instances, so
 			 * just move them to a separate list.
 			 */
-			list_add_tail(&pos->node, &self->removed_maps[pos->type]);
+			list_add_tail(&pos->node, &mg->removed_maps[pos->type]);
 		}
 	}
 }
 
-struct symbol *map_groups__find_symbol(struct map_groups *self,
+struct symbol *map_groups__find_symbol(struct map_groups *mg,
 				       enum map_type type, u64 addr,
 				       struct map **mapp,
 				       symbol_filter_t filter)
 {
-	struct map *map = map_groups__find(self, type, addr);
+	struct map *map = map_groups__find(mg, type, addr);
 
 	if (map != NULL) {
 		if (mapp != NULL)
@@ -301,7 +301,7 @@ struct symbol *map_groups__find_symbol(struct map_groups *self,
 	return NULL;
 }
 
-struct symbol *map_groups__find_symbol_by_name(struct map_groups *self,
+struct symbol *map_groups__find_symbol_by_name(struct map_groups *mg,
 					       enum map_type type,
 					       const char *name,
 					       struct map **mapp,
@@ -309,7 +309,7 @@ struct symbol *map_groups__find_symbol_by_name(struct map_groups *self,
 {
 	struct rb_node *nd;
 
-	for (nd = rb_first(&self->maps[type]); nd; nd = rb_next(nd)) {
+	for (nd = rb_first(&mg->maps[type]); nd; nd = rb_next(nd)) {
 		struct map *pos = rb_entry(nd, struct map, rb_node);
 		struct symbol *sym = map__find_symbol_by_name(pos, name, filter);
 
@@ -323,13 +323,13 @@ struct symbol *map_groups__find_symbol_by_name(struct map_groups *self,
 	return NULL;
 }
 
-size_t __map_groups__fprintf_maps(struct map_groups *self,
+size_t __map_groups__fprintf_maps(struct map_groups *mg,
 				  enum map_type type, int verbose, FILE *fp)
 {
 	size_t printed = fprintf(fp, "%s:\n", map_type__name[type]);
 	struct rb_node *nd;
 
-	for (nd = rb_first(&self->maps[type]); nd; nd = rb_next(nd)) {
+	for (nd = rb_first(&mg->maps[type]); nd; nd = rb_next(nd)) {
 		struct map *pos = rb_entry(nd, struct map, rb_node);
 		printed += fprintf(fp, "Map:");
 		printed += map__fprintf(pos, fp);
@@ -342,22 +342,22 @@ size_t __map_groups__fprintf_maps(struct map_groups *self,
 	return printed;
 }
 
-size_t map_groups__fprintf_maps(struct map_groups *self, int verbose, FILE *fp)
+size_t map_groups__fprintf_maps(struct map_groups *mg, int verbose, FILE *fp)
 {
 	size_t printed = 0, i;
 	for (i = 0; i < MAP__NR_TYPES; ++i)
-		printed += __map_groups__fprintf_maps(self, i, verbose, fp);
+		printed += __map_groups__fprintf_maps(mg, i, verbose, fp);
 	return printed;
 }
 
-static size_t __map_groups__fprintf_removed_maps(struct map_groups *self,
+static size_t __map_groups__fprintf_removed_maps(struct map_groups *mg,
 						 enum map_type type,
 						 int verbose, FILE *fp)
 {
 	struct map *pos;
 	size_t printed = 0;
 
-	list_for_each_entry(pos, &self->removed_maps[type], node) {
+	list_for_each_entry(pos, &mg->removed_maps[type], node) {
 		printed += fprintf(fp, "Map:");
 		printed += map__fprintf(pos, fp);
 		if (verbose > 1) {
@@ -368,26 +368,26 @@ static size_t __map_groups__fprintf_removed_maps(struct map_groups *self,
 	return printed;
 }
 
-static size_t map_groups__fprintf_removed_maps(struct map_groups *self,
+static size_t map_groups__fprintf_removed_maps(struct map_groups *mg,
 					       int verbose, FILE *fp)
 {
 	size_t printed = 0, i;
 	for (i = 0; i < MAP__NR_TYPES; ++i)
-		printed += __map_groups__fprintf_removed_maps(self, i, verbose, fp);
+		printed += __map_groups__fprintf_removed_maps(mg, i, verbose, fp);
 	return printed;
 }
 
-size_t map_groups__fprintf(struct map_groups *self, int verbose, FILE *fp)
+size_t map_groups__fprintf(struct map_groups *mg, int verbose, FILE *fp)
 {
-	size_t printed = map_groups__fprintf_maps(self, verbose, fp);
+	size_t printed = map_groups__fprintf_maps(mg, verbose, fp);
 	printed += fprintf(fp, "Removed maps:\n");
-	return printed + map_groups__fprintf_removed_maps(self, verbose, fp);
+	return printed + map_groups__fprintf_removed_maps(mg, verbose, fp);
 }
 
-int map_groups__fixup_overlappings(struct map_groups *self, struct map *map,
+int map_groups__fixup_overlappings(struct map_groups *mg, struct map *map,
 				   int verbose, FILE *fp)
 {
-	struct rb_root *root = &self->maps[map->type];
+	struct rb_root *root = &mg->maps[map->type];
 	struct rb_node *next = rb_first(root);
 	int err = 0;
 
@@ -418,7 +418,7 @@ int map_groups__fixup_overlappings(struct map_groups *self, struct map *map,
 			}
 
 			before->end = map->start - 1;
-			map_groups__insert(self, before);
+			map_groups__insert(mg, before);
 			if (verbose >= 2)
 				map__fprintf(before, fp);
 		}
@@ -432,7 +432,7 @@ int map_groups__fixup_overlappings(struct map_groups *self, struct map *map,
 			}
 
 			after->start = map->end + 1;
-			map_groups__insert(self, after);
+			map_groups__insert(mg, after);
 			if (verbose >= 2)
 				map__fprintf(after, fp);
 		}
@@ -441,7 +441,7 @@ move_map:
 		 * If we have references, just move them to a separate list.
 		 */
 		if (pos->referenced)
-			list_add_tail(&pos->node, &self->removed_maps[map->type]);
+			list_add_tail(&pos->node, &mg->removed_maps[map->type]);
 		else
 			map__delete(pos);
 
@@ -455,7 +455,7 @@ move_map:
 /*
  * XXX This should not really _copy_ te maps, but refcount them.
  */
-int map_groups__clone(struct map_groups *self,
+int map_groups__clone(struct map_groups *mg,
 		      struct map_groups *parent, enum map_type type)
 {
 	struct rb_node *nd;
@@ -464,7 +464,7 @@ int map_groups__clone(struct map_groups *self,
 		struct map *new = map__clone(map);
 		if (new == NULL)
 			return -ENOMEM;
-		map_groups__insert(self, new);
+		map_groups__insert(mg, new);
 	}
 	return 0;
 }
diff --git a/tools/perf/util/map.h b/tools/perf/util/map.h
index b397c03..890d855 100644
--- a/tools/perf/util/map.h
+++ b/tools/perf/util/map.h
@@ -123,17 +123,17 @@ void map__fixup_end(struct map *self);
 
 void map__reloc_vmlinux(struct map *self);
 
-size_t __map_groups__fprintf_maps(struct map_groups *self,
+size_t __map_groups__fprintf_maps(struct map_groups *mg,
 				  enum map_type type, int verbose, FILE *fp);
 void maps__insert(struct rb_root *maps, struct map *map);
-void maps__remove(struct rb_root *self, struct map *map);
+void maps__remove(struct rb_root *maps, struct map *map);
 struct map *maps__find(struct rb_root *maps, u64 addr);
-void map_groups__init(struct map_groups *self);
-void map_groups__exit(struct map_groups *self);
-int map_groups__clone(struct map_groups *self,
+void map_groups__init(struct map_groups *mg);
+void map_groups__exit(struct map_groups *mg);
+int map_groups__clone(struct map_groups *mg,
 		      struct map_groups *parent, enum map_type type);
-size_t map_groups__fprintf(struct map_groups *self, int verbose, FILE *fp);
-size_t map_groups__fprintf_maps(struct map_groups *self, int verbose, FILE *fp);
+size_t map_groups__fprintf(struct map_groups *mg, int verbose, FILE *fp);
+size_t map_groups__fprintf_maps(struct map_groups *mg, int verbose, FILE *fp);
 
 typedef void (*machine__process_t)(struct machine *self, void *data);
 
@@ -162,29 +162,29 @@ static inline bool machine__is_host(struct machine *self)
 	return self ? self->pid == HOST_KERNEL_ID : false;
 }
 
-static inline void map_groups__insert(struct map_groups *self, struct map *map)
+static inline void map_groups__insert(struct map_groups *mg, struct map *map)
 {
-	maps__insert(&self->maps[map->type], map);
-	map->groups = self;
+	maps__insert(&mg->maps[map->type], map);
+	map->groups = mg;
 }
 
-static inline void map_groups__remove(struct map_groups *self, struct map *map)
+static inline void map_groups__remove(struct map_groups *mg, struct map *map)
 {
-	maps__remove(&self->maps[map->type], map);
+	maps__remove(&mg->maps[map->type], map);
 }
 
-static inline struct map *map_groups__find(struct map_groups *self,
+static inline struct map *map_groups__find(struct map_groups *mg,
 					   enum map_type type, u64 addr)
 {
-	return maps__find(&self->maps[type], addr);
+	return maps__find(&mg->maps[type], addr);
 }
 
-struct symbol *map_groups__find_symbol(struct map_groups *self,
+struct symbol *map_groups__find_symbol(struct map_groups *mg,
 				       enum map_type type, u64 addr,
 				       struct map **mapp,
 				       symbol_filter_t filter);
 
-struct symbol *map_groups__find_symbol_by_name(struct map_groups *self,
+struct symbol *map_groups__find_symbol_by_name(struct map_groups *mg,
 					       enum map_type type,
 					       const char *name,
 					       struct map **mapp,
@@ -208,11 +208,11 @@ struct symbol *machine__find_kernel_function(struct machine *self, u64 addr,
 }
 
 static inline
-struct symbol *map_groups__find_function_by_name(struct map_groups *self,
+struct symbol *map_groups__find_function_by_name(struct map_groups *mg,
 						 const char *name, struct map **mapp,
 						 symbol_filter_t filter)
 {
-	return map_groups__find_symbol_by_name(self, MAP__FUNCTION, name, mapp, filter);
+	return map_groups__find_symbol_by_name(mg, MAP__FUNCTION, name, mapp, filter);
 }
 
 static inline
@@ -225,13 +225,13 @@ struct symbol *machine__find_kernel_function_by_name(struct machine *self,
 						 filter);
 }
 
-int map_groups__fixup_overlappings(struct map_groups *self, struct map *map,
+int map_groups__fixup_overlappings(struct map_groups *mg, struct map *map,
 				   int verbose, FILE *fp);
 
-struct map *map_groups__find_by_name(struct map_groups *self,
+struct map *map_groups__find_by_name(struct map_groups *mg,
 				     enum map_type type, const char *name);
 struct map *machine__new_module(struct machine *self, u64 start, const char *filename);
 
-void map_groups__flush(struct map_groups *self);
+void map_groups__flush(struct map_groups *mg);
 
 #endif /* __PERF_MAP_H */
-- 
1.6.2.5

[PATCH 02/16] perf script: Add drop monitor script

[Arnaldo Carvalho de Melo]

From: Neil Horman <nhorman@tuxdriver.com>

A while back I created the dropmonitor protocol, which allowed users to get
reports of dropped frames communicated to them via a netlink socket.

While useful, several people have now asked that I integrate the ability
to do drop monitoring with perf, so they don't have to run additional
tools.

This patch adds a drop monitor script to the perf suite, and provides
the same output that the netlink socket does.

Cc: Ingo Molnar <mingo@elte.hu>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1309801217-22450-1-git-send-email-nhorman@tuxdriver.com
Signed-off-by: Neil Horman <nhorman@tuxdriver.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 .../perf/scripts/python/bin/net_dropmonitor-record |    2 +
 .../perf/scripts/python/bin/net_dropmonitor-report |    4 +
 tools/perf/scripts/python/net_dropmonitor.py       |   72 ++++++++++++++++++++
 3 files changed, 78 insertions(+), 0 deletions(-)
 create mode 100755 tools/perf/scripts/python/bin/net_dropmonitor-record
 create mode 100755 tools/perf/scripts/python/bin/net_dropmonitor-report
 create mode 100755 tools/perf/scripts/python/net_dropmonitor.py

diff --git a/tools/perf/scripts/python/bin/net_dropmonitor-record b/tools/perf/scripts/python/bin/net_dropmonitor-record
new file mode 100755
index 0000000..423fb81
--- /dev/null
+++ b/tools/perf/scripts/python/bin/net_dropmonitor-record
@@ -0,0 +1,2 @@
+#!/bin/bash
+perf record -e skb:kfree_skb $@
diff --git a/tools/perf/scripts/python/bin/net_dropmonitor-report b/tools/perf/scripts/python/bin/net_dropmonitor-report
new file mode 100755
index 0000000..8d698f5
--- /dev/null
+++ b/tools/perf/scripts/python/bin/net_dropmonitor-report
@@ -0,0 +1,4 @@
+#!/bin/bash
+# description: display a table of dropped frames
+
+perf script -s "$PERF_EXEC_PATH"/scripts/python/net_dropmonitor.py $@
diff --git a/tools/perf/scripts/python/net_dropmonitor.py b/tools/perf/scripts/python/net_dropmonitor.py
new file mode 100755
index 0000000..a4ffc95
--- /dev/null
+++ b/tools/perf/scripts/python/net_dropmonitor.py
@@ -0,0 +1,72 @@
+# Monitor the system for dropped packets and proudce a report of drop locations and counts
+
+import os
+import sys
+
+sys.path.append(os.environ['PERF_EXEC_PATH'] + \
+		'/scripts/python/Perf-Trace-Util/lib/Perf/Trace')
+
+from perf_trace_context import *
+from Core import *
+from Util import *
+
+drop_log = {}
+kallsyms = []
+
+def get_kallsyms_table():
+	global kallsyms
+	try:
+		f = open("/proc/kallsyms", "r")
+		linecount = 0
+		for line in f:
+			linecount = linecount+1
+		f.seek(0)
+	except:
+		return
+
+
+	j = 0
+	for line in f:
+		loc = int(line.split()[0], 16)
+		name = line.split()[2]
+		j = j +1
+		if ((j % 100) == 0):
+			print "\r" + str(j) + "/" + str(linecount),
+		kallsyms.append({ 'loc': loc, 'name' : name})
+
+	print "\r" + str(j) + "/" + str(linecount)
+	kallsyms.sort()
+	return
+
+def get_sym(sloc):
+	loc = int(sloc)
+	for i in kallsyms:
+		if (i['loc'] >= loc):
+			return (i['name'], i['loc']-loc)
+	return (None, 0)
+
+def print_drop_table():
+	print "%25s %25s %25s" % ("LOCATION", "OFFSET", "COUNT")
+	for i in drop_log.keys():
+		(sym, off) = get_sym(i)
+		if sym == None:
+			sym = i
+		print "%25s %25s %25s" % (sym, off, drop_log[i])
+
+
+def trace_begin():
+	print "Starting trace (Ctrl-C to dump results)"
+
+def trace_end():
+	print "Gathering kallsyms data"
+	get_kallsyms_table()
+	print_drop_table()
+
+# called from perf, when it finds a correspoinding event
+def skb__kfree_skb(name, context, cpu, sec, nsec, pid, comm,
+			skbaddr, protocol, location):
+	slocation = str(location)
+	try:
+		drop_log[slocation] = drop_log[slocation] + 1
+	except:
+		drop_log[slocation] = 1
-- 
1.6.2.5

[PATCH 03/16] perf buildid-list: Add option to show the running kernel build id

[Arnaldo Carvalho de Melo]

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

[root@emilia ~]# perf buildid-list -k
07b0c016a2b30004e86132d0239945b1e88f5d75

Useful when diagnosing build id problems in debuginfo packages, etc.

Cc: David Ahern <dsahern@gmail.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Link: http://lkml.kernel.org/n/tip-po1bl7acn6e1hhne90opmvtl@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/Documentation/perf-buildid-list.txt |    5 ++++
 tools/perf/builtin-buildid-list.c              |   31 ++++++++++++++++++++++-
 2 files changed, 34 insertions(+), 2 deletions(-)

diff --git a/tools/perf/Documentation/perf-buildid-list.txt b/tools/perf/Documentation/perf-buildid-list.txt
index 5eaac6f..770a7f3 100644
--- a/tools/perf/Documentation/perf-buildid-list.txt
+++ b/tools/perf/Documentation/perf-buildid-list.txt
@@ -16,6 +16,8 @@ This command displays the buildids found in a perf.data file, so that other
 tools can be used to fetch packages with matching symbol tables for use by
 perf report.
 
+It can also be used to show the build id of the running kernel.
+
 OPTIONS
 -------
 -H::
@@ -27,6 +29,9 @@ OPTIONS
 -f::
 --force::
 	Don't do ownership validation.
+-k::
+--kernel::
+	Show running kernel build id.
 -v::
 --verbose::
 	Be more verbose.
diff --git a/tools/perf/builtin-buildid-list.c b/tools/perf/builtin-buildid-list.c
index 5af32ae..4102ead 100644
--- a/tools/perf/builtin-buildid-list.c
+++ b/tools/perf/builtin-buildid-list.c
@@ -1,7 +1,8 @@
 /*
  * builtin-buildid-list.c
  *
- * Builtin buildid-list command: list buildids in perf.data
+ * Builtin buildid-list command: list buildids in perf.data or in the running
+ * kernel.
  *
  * Copyright (C) 2009, Red Hat Inc.
  * Copyright (C) 2009, Arnaldo Carvalho de Melo <acme@redhat.com>
@@ -17,6 +18,7 @@
 
 static char const *input_name = "perf.data";
 static bool force;
+static bool show_kernel;
 static bool with_hits;
 
 static const char * const buildid_list_usage[] = {
@@ -29,12 +31,13 @@ static const struct option options[] = {
 	OPT_STRING('i', "input", &input_name, "file",
 		    "input file name"),
 	OPT_BOOLEAN('f', "force", &force, "don't complain, do it"),
+	OPT_BOOLEAN('k', "kernel", &show_kernel, "Show current kernel build id"),
 	OPT_INCR('v', "verbose", &verbose,
 		    "be more verbose"),
 	OPT_END()
 };
 
-static int __cmd_buildid_list(void)
+static int perf_session__list_build_ids(void)
 {
 	struct perf_session *session;
 
@@ -52,6 +55,30 @@ static int __cmd_buildid_list(void)
 	return 0;
 }
 
+static int sysfs__fprintf_build_id(FILE *fp)
+{
+	u8 kallsyms_build_id[BUILD_ID_SIZE];
+	char sbuild_id[BUILD_ID_SIZE * 2 + 1];
+
+	if (sysfs__read_build_id("/sys/kernel/notes", kallsyms_build_id,
+				 sizeof(kallsyms_build_id)) != 0)
+		return -1;
+
+	build_id__sprintf(kallsyms_build_id, sizeof(kallsyms_build_id),
+			  sbuild_id);
+	fprintf(fp, "%s\n", sbuild_id);
+	return 0;
+}
+
+static int __cmd_buildid_list(void)
+{
+
+	if (show_kernel)
+		return sysfs__fprintf_build_id(stdout);
+
+	return perf_session__list_build_ids();
+}
+
 int cmd_buildid_list(int argc, const char **argv, const char *prefix __used)
 {
 	argc = parse_options(argc, argv, options, buildid_list_usage, 0);
-- 
1.6.2.5

[PATCH 04/16] perf buildid-list: Support showing the build id in an ELF file

[Arnaldo Carvalho de Melo]

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

Try first reading the build id, validating that it is an ELF file, etc.
Cheap as libelf will bail out as soon as the magic number check fails.

Useful when investigating debuginfo packaging problems like this one:

[root@emilia ~]# perf buildid-list -i /usr/lib/debug/lib/modules/`uname -r`/vmlinux
77bb4ea591a602d455ace759a377c9adfff1aba3
[root@emilia ~]# perf buildid-list -k
07b0c016a2b30004e86132d0239945b1e88f5d75
[root@emilia ~]#

Cc: David Ahern <dsahern@gmail.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Link: http://lkml.kernel.org/n/tip-4elot9oxwa0rr0d90dshca3a@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/Documentation/perf-buildid-list.txt |    3 +-
 tools/perf/builtin-buildid-list.c              |   27 +++++++++++++++++++++--
 2 files changed, 26 insertions(+), 4 deletions(-)

diff --git a/tools/perf/Documentation/perf-buildid-list.txt b/tools/perf/Documentation/perf-buildid-list.txt
index 770a7f3..cc22325 100644
--- a/tools/perf/Documentation/perf-buildid-list.txt
+++ b/tools/perf/Documentation/perf-buildid-list.txt
@@ -16,7 +16,8 @@ This command displays the buildids found in a perf.data file, so that other
 tools can be used to fetch packages with matching symbol tables for use by
 perf report.
 
-It can also be used to show the build id of the running kernel.
+It can also be used to show the build id of the running kernel or in an ELF
+file using -i/--input.
 
 OPTIONS
 -------
diff --git a/tools/perf/builtin-buildid-list.c b/tools/perf/builtin-buildid-list.c
index 4102ead..cb690a6 100644
--- a/tools/perf/builtin-buildid-list.c
+++ b/tools/perf/builtin-buildid-list.c
@@ -1,8 +1,8 @@
 /*
  * builtin-buildid-list.c
  *
- * Builtin buildid-list command: list buildids in perf.data or in the running
- * kernel.
+ * Builtin buildid-list command: list buildids in perf.data, in the running
+ * kernel and in ELF files.
  *
  * Copyright (C) 2009, Red Hat Inc.
  * Copyright (C) 2009, Arnaldo Carvalho de Melo <acme@redhat.com>
@@ -16,6 +16,8 @@
 #include "util/session.h"
 #include "util/symbol.h"
 
+#include <libelf.h>
+
 static char const *input_name = "perf.data";
 static bool force;
 static bool show_kernel;
@@ -70,12 +72,31 @@ static int sysfs__fprintf_build_id(FILE *fp)
 	return 0;
 }
 
-static int __cmd_buildid_list(void)
+static int filename__fprintf_build_id(const char *name, FILE *fp)
 {
+	u8 build_id[BUILD_ID_SIZE];
+	char sbuild_id[BUILD_ID_SIZE * 2 + 1];
 
+	if (filename__read_build_id(name, build_id,
+				    sizeof(build_id)) != sizeof(build_id))
+		return 0;
+
+	build_id__sprintf(build_id, sizeof(build_id), sbuild_id);
+	return fprintf(fp, "%s\n", sbuild_id);
+}
+
+static int __cmd_buildid_list(void)
+{
 	if (show_kernel)
 		return sysfs__fprintf_build_id(stdout);
 
+	elf_version(EV_CURRENT);
+	/*
+ 	 * See if this is an ELF file first:
+ 	 */
+	if (filename__fprintf_build_id(input_name, stdout))
+		return 0;
+
 	return perf_session__list_build_ids();
 }
 
-- 
1.6.2.5

[PATCH 05/16] perf top browser: Fix up line width calculation

[Arnaldo Carvalho de Melo]

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

Fixing an artifact where the last 3 chars of a long DSO name would
remain on the screen sometimes.

Cc: David Ahern <dsahern@gmail.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Link: http://lkml.kernel.org/n/tip-dkiakcl3z69dh1bt9uegaktv@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/util/ui/browsers/top.c |    4 ++--
 1 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/tools/perf/util/ui/browsers/top.c b/tools/perf/util/ui/browsers/top.c
index 88403cf..9d93810 100644
--- a/tools/perf/util/ui/browsers/top.c
+++ b/tools/perf/util/ui/browsers/top.c
@@ -43,10 +43,10 @@ static void perf_top_browser__write(struct ui_browser *browser, void *entry, int
 
 	if (top->evlist->nr_entries == 1 || !top->display_weighted) {
 		slsmg_printf("%20.2f ", syme->weight);
-		width -= 24;
+		width -= 21;
 	} else {
 		slsmg_printf("%9.1f %10ld ", syme->weight, syme->snap_count);
-		width -= 23;
+		width -= 20;
 	}
 
 	slsmg_printf("%4.1f%%", pcnt);
-- 
1.6.2.5

[PATCH 06/16] perf top: Improve lost events warning

[Arnaldo Carvalho de Melo]

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

Now it warns everytime that new events are lost.

And the TUI also warns now.

Cc: David Ahern <dsahern@gmail.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Link: http://lkml.kernel.org/n/tip-w1n168yrvrppnq6887s4u0wx@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/builtin-top.c          |   34 +++++++++++++++++-----------------
 tools/perf/util/top.h             |    3 +++
 tools/perf/util/ui/browsers/top.c |   24 ++++++++++++++++++++++++
 3 files changed, 44 insertions(+), 17 deletions(-)

diff --git a/tools/perf/builtin-top.c b/tools/perf/builtin-top.c
index a43433f..23c4f71 100644
--- a/tools/perf/builtin-top.c
+++ b/tools/perf/builtin-top.c
@@ -250,7 +250,7 @@ static void __list_insert_active_sym(struct sym_entry *syme)
 	list_add(&syme->node, &top.active_symbols);
 }
 
-static void print_sym_table(struct perf_session *session)
+static void print_sym_table(void)
 {
 	char bf[160];
 	int printed = 0;
@@ -270,10 +270,11 @@ static void print_sym_table(struct perf_session *session)
 
 	printf("%-*.*s\n", win_width, win_width, graph_dotted_line);
 
-	if (session->hists.stats.total_lost != 0) {
+	if (top.total_lost_warned != top.session->hists.stats.total_lost) {
+		top.total_lost_warned = top.session->hists.stats.total_lost;
 		color_fprintf(stdout, PERF_COLOR_RED, "WARNING:");
 		printf(" LOST %" PRIu64 " events, Check IO/CPU overload\n",
-		       session->hists.stats.total_lost);
+		       top.total_lost_warned);
 	}
 
 	if (top.sym_filter_entry) {
@@ -474,7 +475,7 @@ static int key_mapped(int c)
 	return 0;
 }
 
-static void handle_keypress(struct perf_session *session, int c)
+static void handle_keypress(int c)
 {
 	if (!key_mapped(c)) {
 		struct pollfd stdin_poll = { .fd = 0, .events = POLLIN };
@@ -550,7 +551,7 @@ static void handle_keypress(struct perf_session *session, int c)
 		case 'Q':
 			printf("exiting.\n");
 			if (dump_symtab)
-				perf_session__fprintf_dsos(session, stderr);
+				perf_session__fprintf_dsos(top.session, stderr);
 			exit(0);
 		case 's':
 			prompt_symbol(&top.sym_filter_entry, "Enter details symbol");
@@ -602,7 +603,6 @@ static void *display_thread(void *arg __used)
 	struct pollfd stdin_poll = { .fd = 0, .events = POLLIN };
 	struct termios tc, save;
 	int delay_msecs, c;
-	struct perf_session *session = (struct perf_session *) arg;
 
 	tcgetattr(0, &save);
 	tc = save;
@@ -617,13 +617,13 @@ repeat:
 	getc(stdin);
 
 	do {
-		print_sym_table(session);
+		print_sym_table();
 	} while (!poll(&stdin_poll, 1, delay_msecs) == 1);
 
 	c = getc(stdin);
 	tcsetattr(0, TCSAFLUSH, &save);
 
-	handle_keypress(session, c);
+	handle_keypress(c);
 	goto repeat;
 
 	return NULL;
@@ -935,27 +935,27 @@ static int __cmd_top(void)
 	 * FIXME: perf_session__new should allow passing a O_MMAP, so that all this
 	 * mmap reading, etc is encapsulated in it. Use O_WRONLY for now.
 	 */
-	struct perf_session *session = perf_session__new(NULL, O_WRONLY, false, false, NULL);
-	if (session == NULL)
+	top.session = perf_session__new(NULL, O_WRONLY, false, false, NULL);
+	if (top.session == NULL)
 		return -ENOMEM;
 
 	if (top.target_tid != -1)
 		perf_event__synthesize_thread_map(top.evlist->threads,
-						  perf_event__process, session);
+						  perf_event__process, top.session);
 	else
-		perf_event__synthesize_threads(perf_event__process, session);
+		perf_event__synthesize_threads(perf_event__process, top.session);
 
 	start_counters(top.evlist);
-	session->evlist = top.evlist;
-	perf_session__update_sample_type(session);
+	top.session->evlist = top.evlist;
+	perf_session__update_sample_type(top.session);
 
 	/* Wait for a minimal set of events before starting the snapshot */
 	poll(top.evlist->pollfd, top.evlist->nr_fds, 100);
 
-	perf_session__mmap_read(session);
+	perf_session__mmap_read(top.session);
 
 	if (pthread_create(&thread, NULL, (use_browser > 0 ? display_thread_tui :
-							     display_thread), session)) {
+							     display_thread), NULL)) {
 		printf("Could not create display thread.\n");
 		exit(-1);
 	}
@@ -973,7 +973,7 @@ static int __cmd_top(void)
 	while (1) {
 		u64 hits = top.samples;
 
-		perf_session__mmap_read(session);
+		perf_session__mmap_read(top.session);
 
 		if (hits == top.samples)
 			ret = poll(top.evlist->pollfd, top.evlist->nr_fds, 100);
diff --git a/tools/perf/util/top.h b/tools/perf/util/top.h
index bfbf95b..b07b041 100644
--- a/tools/perf/util/top.h
+++ b/tools/perf/util/top.h
@@ -10,6 +10,7 @@
 
 struct perf_evlist;
 struct perf_evsel;
+struct perf_session;
 
 struct sym_entry {
 	struct rb_node		rb_node;
@@ -38,6 +39,7 @@ struct perf_top {
 	u64		   kernel_samples, us_samples;
 	u64		   exact_samples;
 	u64		   guest_us_samples, guest_kernel_samples;
+	u64		   total_lost_warned;
 	int		   print_entries, count_filter, delay_secs;
 	int		   display_weighted, freq, rb_entries;
 	pid_t		   target_pid, target_tid;
@@ -45,6 +47,7 @@ struct perf_top {
 	const char	   *cpu_list;
 	struct sym_entry   *sym_filter_entry;
 	struct perf_evsel  *sym_evsel;
+	struct perf_session *session;
 };
 
 size_t perf_top__header_snprintf(struct perf_top *top, char *bf, size_t size);
diff --git a/tools/perf/util/ui/browsers/top.c b/tools/perf/util/ui/browsers/top.c
index 9d93810..9b6b43b 100644
--- a/tools/perf/util/ui/browsers/top.c
+++ b/tools/perf/util/ui/browsers/top.c
@@ -11,10 +11,12 @@
 #include "../helpline.h"
 #include "../libslang.h"
 #include "../util.h"
+#include "../ui.h"
 #include "../../evlist.h"
 #include "../../hist.h"
 #include "../../sort.h"
 #include "../../symbol.h"
+#include "../../session.h"
 #include "../../top.h"
 
 struct perf_top_browser {
@@ -143,6 +145,25 @@ do_annotation:
 	symbol__tui_annotate(sym, syme->map, 0, top->delay_secs * 1000);
 }
 
+static void perf_top_browser__warn_lost(struct perf_top_browser *browser)
+{
+	struct perf_top *top = browser->b.priv;
+	char msg[128];
+	int len;
+
+	top->total_lost_warned = top->session->hists.stats.total_lost;
+	pthread_mutex_lock(&ui__lock);
+	ui_browser__set_color(&browser->b, HE_COLORSET_TOP);
+	len = snprintf(msg, sizeof(msg),
+		      " WARNING: LOST %" PRIu64 " events, Check IO/CPU overload",
+		      top->total_lost_warned);
+	if (len > browser->b.width)
+		len = browser->b.width;
+	SLsmg_gotorc(0, browser->b.width - len);
+	slsmg_write_nstring(msg, len);
+	pthread_mutex_unlock(&ui__lock);
+}
+
 static int perf_top_browser__run(struct perf_top_browser *browser)
 {
 	int key;
@@ -174,6 +195,9 @@ static int perf_top_browser__run(struct perf_top_browser *browser)
 			ui_browser__set_color(&browser->b, NEWT_COLORSET_ROOT);
 			SLsmg_gotorc(0, 0);
 			slsmg_write_nstring(title, browser->b.width);
+
+			if (top->total_lost_warned != top->session->hists.stats.total_lost)
+				perf_top_browser__warn_lost(browser);
 			break;
 		case 'a':
 		case NEWT_KEY_RIGHT:
-- 
1.6.2.5

[PATCH 07/16] perf stat: Add --log-fd <N> option to redirect stderr elsewhere

[Arnaldo Carvalho de Melo]

From: Jim Cromie <jim.cromie@gmail.com>

This perf stat option emulates valgrind's --log-fd option, allowing the
user to send perf results elsewhere, and leaving stderr for use by the
program under test.  This complements --output file option, and is
mutually exclusive with it.

   3>results  perf stat --log-fd 3          -- $cmd
   3>>results perf stat --log-fd 3 --append -- $cmd

The perl distro's make test.valgrind target uses valgrind's --log-fd
option, I've adapted it to invoke perf also, and tested this patch
there.

Link: http://lkml.kernel.org/r/1315437244-3788-2-git-send-email-jim.cromie@gmail.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Jim Cromie <jim.cromie@gmail.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/Documentation/perf-stat.txt |   11 ++++++++++-
 tools/perf/builtin-stat.c              |   14 ++++++++++++++
 2 files changed, 24 insertions(+), 1 deletions(-)

diff --git a/tools/perf/Documentation/perf-stat.txt b/tools/perf/Documentation/perf-stat.txt
index 08394c4..8966b9a 100644
--- a/tools/perf/Documentation/perf-stat.txt
+++ b/tools/perf/Documentation/perf-stat.txt
@@ -95,12 +95,21 @@ corresponding events, i.e., they always refer to events defined earlier on the c
 line.
 
 -o file::
--output file::
+--output file::
 Print the output into the designated file.
 
 --append::
 Append to the output file designated with the -o option. Ignored if -o is not specified.
 
+--log-fd::
+
+Log output to fd, instead of stderr.  Complementary to --output, and mutually exclusive
+with it.  --append may be used here.  Examples:
+     3>results  perf stat --log-fd 3          -- $cmd
+     3>>results perf stat --log-fd 3 --append -- $cmd
+
+
+
 EXAMPLES
 --------
 
diff --git a/tools/perf/builtin-stat.c b/tools/perf/builtin-stat.c
index bec64a9..a43c680 100644
--- a/tools/perf/builtin-stat.c
+++ b/tools/perf/builtin-stat.c
@@ -196,6 +196,7 @@ static bool			csv_output			= false;
 static bool			group				= false;
 static const char		*output_name			= NULL;
 static FILE			*output				= NULL;
+static int			output_fd;
 
 static volatile int done = 0;
 
@@ -1080,6 +1081,8 @@ static const struct option options[] = {
 	OPT_STRING('o', "output", &output_name, "file",
 		    "output file name"),
 	OPT_BOOLEAN(0, "append", &append_file, "append to the output file"),
+	OPT_INTEGER(0, "log-fd", &output_fd,
+		    "log output to fd, instead of stderr"),
 	OPT_END()
 };
 
@@ -1166,6 +1169,10 @@ int cmd_stat(int argc, const char **argv, const char *prefix __used)
 	if (output_name && strcmp(output_name, "-"))
 		output = NULL;
 
+	if (output_name && output_fd) {
+		fprintf(stderr, "cannot use both --output and --log-fd\n");
+		usage_with_options(stat_usage, options);
+	}
 	if (!output) {
 		struct timespec tm;
 		mode = append_file ? "a" : "w";
@@ -1177,6 +1184,13 @@ int cmd_stat(int argc, const char **argv, const char *prefix __used)
 		}
 		clock_gettime(CLOCK_REALTIME, &tm);
 		fprintf(output, "# started on %s\n", ctime(&tm.tv_sec));
+	} else if (output_fd != 2) {
+		mode = append_file ? "a" : "w";
+		output = fdopen(output_fd, mode);
+		if (!output) {
+			perror("Failed opening logfd");
+			return -errno;
+		}
 	}
 
 	if (csv_sep)
-- 
1.6.2.5

[PATCH] perf stat: Fix default logfd to use stderr

[Robert Richter]

On 29.09.11 18:48:01, Arnaldo Carvalho de Melo wrote:
> From: Jim Cromie <jim.cromie@gmail.com>
> 
> This perf stat option emulates valgrind's --log-fd option, allowing the
> user to send perf results elsewhere, and leaving stderr for use by the
> program under test.  This complements --output file option, and is
> mutually exclusive with it.
> 
>    3>results  perf stat --log-fd 3          -- $cmd
>    3>>results perf stat --log-fd 3 --append -- $cmd
> 
> The perl distro's make test.valgrind target uses valgrind's --log-fd
> option, I've adapted it to invoke perf also, and tested this patch
> there.
> 
> Link: http://lkml.kernel.org/r/1315437244-3788-2-git-send-email-jim.cromie@gmail.com
> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
> Signed-off-by: Jim Cromie <jim.cromie@gmail.com>
> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

With certain shell redirections this (56f3bae) fails with a log fd
setup failure. Fix below.

-Robert



>From fd83259163c8022f9264a119d5ef7df594702f3a Mon Sep 17 00:00:00 2001
From: Robert Richter <robert.richter@amd.com>
Date: Thu, 7 Jun 2012 17:41:52 +0200
Subject: [PATCH] perf stat: Fix default logfd to use stderr

When running perf-stat in certain shell environments with stdout
redirection there is a logging file descriptor setup failure:

 Failed opening logfd: Invalid argument

Fixing this by setting the default fd to the correct value of 2.

Cc: Jim Cromie <jim.cromie@gmail.com>
Cc: <stable@vger.kernel.org> # 3.2+
Signed-off-by: Robert Richter <robert.richter@amd.com>
---
 tools/perf/builtin-stat.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/tools/perf/builtin-stat.c b/tools/perf/builtin-stat.c
index 2625899..c47e7d3 100644
--- a/tools/perf/builtin-stat.c
+++ b/tools/perf/builtin-stat.c
@@ -195,7 +195,7 @@ static bool			csv_output			= false;
 static bool			group				= false;
 static const char		*output_name			= NULL;
 static FILE			*output				= NULL;
-static int			output_fd;
+static int			output_fd			=  2;
 
 static volatile int done = 0;
 
-- 
1.7.8.4



-- 
Advanced Micro Devices, Inc.
Operating System Research Center

Re: [PATCH] perf stat: Fix default logfd to use stderr

[Namhyung Kim]

Hi, Robert

On Mon, 18 Jun 2012 16:51:08 +0200, Robert Richter wrote:
> On 29.09.11 18:48:01, Arnaldo Carvalho de Melo wrote:
>> From: Jim Cromie <jim.cromie@gmail.com>
>> 
>> This perf stat option emulates valgrind's --log-fd option, allowing the
>> user to send perf results elsewhere, and leaving stderr for use by the
>> program under test.  This complements --output file option, and is
>> mutually exclusive with it.
>> 
>>    3>results  perf stat --log-fd 3          -- $cmd
>>    3>>results perf stat --log-fd 3 --append -- $cmd
>> 
>> The perl distro's make test.valgrind target uses valgrind's --log-fd
>> option, I've adapted it to invoke perf also, and tested this patch
>> there.
>> 
>> Link: http://lkml.kernel.org/r/1315437244-3788-2-git-send-email-jim.cromie@gmail.com
>> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
>> Signed-off-by: Jim Cromie <jim.cromie@gmail.com>
>> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
>
> With certain shell redirections this (56f3bae) fails with a log fd
> setup failure. Fix below.
>

It looks somewhat related to Stephane's patch. Can you we check it too?

http://lkml.indiana.edu/hypermail/linux/kernel/1205.1/04232.html

Thanks,
Namhyung

Re: [PATCH] perf stat: Fix default logfd to use stderr

[Stephane Eranian]

On Tue, Jun 19, 2012 at 4:13 AM, Namhyung Kim <namhyung@kernel.org> wrote:
> Hi, Robert
>
> On Mon, 18 Jun 2012 16:51:08 +0200, Robert Richter wrote:
>> On 29.09.11 18:48:01, Arnaldo Carvalho de Melo wrote:
>>> From: Jim Cromie <jim.cromie@gmail.com>
>>>
>>> This perf stat option emulates valgrind's --log-fd option, allowing the
>>> user to send perf results elsewhere, and leaving stderr for use by the
>>> program under test.  This complements --output file option, and is
>>> mutually exclusive with it.
>>>
>>>    3>results  perf stat --log-fd 3          -- $cmd
>>>    3>>results perf stat --log-fd 3 --append -- $cmd
>>>
>>> The perl distro's make test.valgrind target uses valgrind's --log-fd
>>> option, I've adapted it to invoke perf also, and tested this patch
>>> there.
>>>
>>> Link: http://lkml.kernel.org/r/1315437244-3788-2-git-send-email-jim.cromie@gmail.com
>>> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
>>> Signed-off-by: Jim Cromie <jim.cromie@gmail.com>
>>> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
>>
>> With certain shell redirections this (56f3bae) fails with a log fd
>> setup failure. Fix below.
>>
>
> It looks somewhat related to Stephane's patch. Can you we check it too?
>
> http://lkml.indiana.edu/hypermail/linux/kernel/1205.1/04232.html
>
Yes, my patch is needed and it was posted over a month ago now.....
Arnaldo, please apply my patch.

> Thanks,
> Namhyung

Re: [PATCH] perf stat: Fix default logfd to use stderr

[Robert Richter]

On 19.06.12 08:09:31, Stephane Eranian wrote:
> On Tue, Jun 19, 2012 at 4:13 AM, Namhyung Kim <namhyung@kernel.org> wrote:
> > Hi, Robert
> >
> > On Mon, 18 Jun 2012 16:51:08 +0200, Robert Richter wrote:
> >> On 29.09.11 18:48:01, Arnaldo Carvalho de Melo wrote:
> >>> From: Jim Cromie <jim.cromie@gmail.com>
> >>>
> >>> This perf stat option emulates valgrind's --log-fd option, allowing the
> >>> user to send perf results elsewhere, and leaving stderr for use by the
> >>> program under test.  This complements --output file option, and is
> >>> mutually exclusive with it.
> >>>
> >>>    3>results  perf stat --log-fd 3          -- $cmd
> >>>    3>>results perf stat --log-fd 3 --append -- $cmd
> >>>
> >>> The perl distro's make test.valgrind target uses valgrind's --log-fd
> >>> option, I've adapted it to invoke perf also, and tested this patch
> >>> there.
> >>>
> >>> Link: http://lkml.kernel.org/r/1315437244-3788-2-git-send-email-jim.cromie@gmail.com
> >>> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
> >>> Signed-off-by: Jim Cromie <jim.cromie@gmail.com>
> >>> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
> >>
> >> With certain shell redirections this (56f3bae) fails with a log fd
> >> setup failure. Fix below.
> >>
> >
> > It looks somewhat related to Stephane's patch. Can you we check it too?
> >
> > http://lkml.indiana.edu/hypermail/linux/kernel/1205.1/04232.html

I applied the patch to my tree and this works too. Both patches
basically avoid fdopen() if a log-fd option is not given.

After reviewing the code again I noticed my code breaks the -o option.
So Stephane's patch is fine for me too, except for the fact that
--log-fd 0 is ignored. Maybe we change this by initializing output_fd
with -1 and modify the checks of output_fd?

> Yes, my patch is needed and it was posted over a month ago now.....
> Arnaldo, please apply my patch.

It is in the pull request for Ingo.

-Robert

-- 
Advanced Micro Devices, Inc.
Operating System Research Center

Re: [PATCH] perf stat: Fix default logfd to use stderr

[Robert Richter]

On 19.06.12 08:09:31, Stephane Eranian wrote:
> On Tue, Jun 19, 2012 at 4:13 AM, Namhyung Kim <namhyung@kernel.org> wrote:
> > Hi, Robert
> >
> > On Mon, 18 Jun 2012 16:51:08 +0200, Robert Richter wrote:
> >> On 29.09.11 18:48:01, Arnaldo Carvalho de Melo wrote:
> >>> From: Jim Cromie <jim.cromie@gmail.com>
> >>>
> >>> This perf stat option emulates valgrind's --log-fd option, allowing the
> >>> user to send perf results elsewhere, and leaving stderr for use by the
> >>> program under test. �This complements --output file option, and is
> >>> mutually exclusive with it.
> >>>
> >>> � �3>results �perf stat --log-fd 3 � � � � �-- $cmd
> >>> � �3>>results perf stat --log-fd 3 --append -- $cmd
> >>>
> >>> The perl distro's make test.valgrind target uses valgrind's --log-fd
> >>> option, I've adapted it to invoke perf also, and tested this patch
> >>> there.
> >>>
> >>> Link: http://lkml.kernel.org/r/1315437244-3788-2-git-send-email-jim.cromie@gmail.com
> >>> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
> >>> Signed-off-by: Jim Cromie <jim.cromie@gmail.com>
> >>> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
> >>
> >> With certain shell redirections this (56f3bae) fails with a log fd
> >> setup failure. Fix below.
> >>
> >
> > It looks somewhat related to Stephane's patch. Can you we check it too?
> >
> > http://lkml.indiana.edu/hypermail/linux/kernel/1205.1/04232.html

I applied the patch to my tree and this works too. Both patches
basically avoid fdopen() if a log-fd option is not given.

After reviewing the code again I noticed my code breaks the -o option.
So Stephane's patch is fine for me too, except for the fact that
--log-fd 0 is ignored. Maybe we change this by initializing output_fd
with -1 and modify the checks of output_fd?

> Yes, my patch is needed and it was posted over a month ago now.....
> Arnaldo, please apply my patch.

It is in the pull request for Ingo.

-Robert

-- 
Advanced Micro Devices, Inc.
Operating System Research Center

Re: [PATCH] perf stat: Fix default logfd to use stderr

[Arnaldo Carvalho de Melo]

Em Tue, Jun 19, 2012 at 08:09:31AM +0200, Stephane Eranian escreveu:
> On Tue, Jun 19, 2012 at 4:13 AM, Namhyung Kim <namhyung@kernel.org> wrote:
> > On Mon, 18 Jun 2012 16:51:08 +0200, Robert Richter wrote:
> >> With certain shell redirections this (56f3bae) fails with a log fd
> >> setup failure. Fix below.

> > It looks somewhat related to Stephane's patch. Can you we check it too?

> > http://lkml.indiana.edu/hypermail/linux/kernel/1205.1/04232.html

> Yes, my patch is needed and it was posted over a month ago now.....
> Arnaldo, please apply my patch.

It was applied a week ago, was merged already by Ingo, should go to
Linus soon:

http://git.kernel.org/?p=linux/kernel/git/tip/tip.git;a=commit;h=fc3e4d077d5c7a7bc1ad5bc143895b4e070e5a8b

- Arnaldo

Re: [PATCH] perf stat: Fix default logfd to use stderr

[Stephane Eranian]

Ok, I see it now.
thanks.
the only one left at this point is the pipe mode meta-data patch
to add header in pipe mode.

On Tue, Jun 19, 2012 at 1:34 PM, Arnaldo Carvalho de Melo
<acme@infradead.org> wrote:
> Em Tue, Jun 19, 2012 at 08:09:31AM +0200, Stephane Eranian escreveu:
>> On Tue, Jun 19, 2012 at 4:13 AM, Namhyung Kim <namhyung@kernel.org> wrote:
>> > On Mon, 18 Jun 2012 16:51:08 +0200, Robert Richter wrote:
>> >> With certain shell redirections this (56f3bae) fails with a log fd
>> >> setup failure. Fix below.
>
>> > It looks somewhat related to Stephane's patch. Can you we check it too?
>
>> > http://lkml.indiana.edu/hypermail/linux/kernel/1205.1/04232.html
>
>> Yes, my patch is needed and it was posted over a month ago now.....
>> Arnaldo, please apply my patch.
>
> It was applied a week ago, was merged already by Ingo, should go to
> Linus soon:
>
> http://git.kernel.org/?p=linux/kernel/git/tip/tip.git;a=commit;h=fc3e4d077d5c7a7bc1ad5bc143895b4e070e5a8b
>
> - Arnaldo

[PATCH 08/16] perf stat: Fix +- nan% in --no-aggr runs

[Arnaldo Carvalho de Melo]

From: Jim Cromie <jim.cromie@gmail.com>

Without this patch, running:

$ sudo ./perf stat -r20 --no-aggr -a perl -e '$i++ for 1..100000'

I get computations like this:

CPU0             12.488247 task-clock                #    1.224 CPUs utilized            ( +-  -nan% )
CPU1             12.488909 task-clock                #    1.225 CPUs utilized            ( +-  -nan% )
CPU2             12.500221 task-clock                #    1.226 CPUs utilized            ( +-  -nan% )
CPU3             12.481713 task-clock                #    1.224 CPUs utilized            ( +-  -nan% )

but with patch, I get:

CPU0              8.233682 task-clock                #    0.754 CPUs utilized            ( +-  0.00% )
CPU1              8.226318 task-clock                #    0.754 CPUs utilized            ( +-  0.00% )
CPU2              8.210737 task-clock                #    0.752 CPUs utilized            ( +-  0.00% )
CPU3              8.201691 task-clock                #    0.751 CPUs utilized            ( +-  0.00% )

Note that without --no-aggr, I get non-0 statistics both before and after patch:

        231.986022 task-clock                #    4.030 CPUs utilized            ( +-  0.97% )
               212 context-switches          #    0.001 M/sec                    ( +- 12.07% )
                 9 CPU-migrations            #    0.000 M/sec                    ( +- 25.80% )
               466 page-faults               #    0.002 M/sec                    ( +-  3.23% )
       174,318,593 cycles                    #    0.751 GHz                      ( +-  1.06% )

I couldnt see anything wrong in the caller, so fixed it in
stddev_stats().  ISTM that 0.00 is better than nan, since perf stat was
passed -A (--no-aggr) so no standard deviation should be expected, and
nan is suggestive of a deeper error.

When running with --no-aggr, perhaps we should suppress the statistics
printing entirely, or do so when they are 0.00.

Link: http://lkml.kernel.org/r/1315437244-3788-3-git-send-email-jim.cromie@gmail.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Jim Cromie <jim.cromie@gmail.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/builtin-stat.c |    9 +++++++--
 1 files changed, 7 insertions(+), 2 deletions(-)

diff --git a/tools/perf/builtin-stat.c b/tools/perf/builtin-stat.c
index a43c680..af0d65b 100644
--- a/tools/perf/builtin-stat.c
+++ b/tools/perf/builtin-stat.c
@@ -254,8 +254,13 @@ static double avg_stats(struct stats *stats)
  */
 static double stddev_stats(struct stats *stats)
 {
-	double variance = stats->M2 / (stats->n - 1);
-	double variance_mean = variance / stats->n;
+	double variance, variance_mean;
+
+	if (!stats->n)
+		return 0.0;
+
+	variance = stats->M2 / (stats->n - 1);
+	variance_mean = variance / stats->n;
 
 	return sqrt(variance_mean);
 }
-- 
1.6.2.5

[PATCH 09/16] perf stat: Suppress printing std-dev when its 0

[Arnaldo Carvalho de Melo]

From: Jim Cromie <jim.cromie@gmail.com>

For pretty output only (preserve column for cvs output), dont print
std-deviation when its 0.00.  Do this based upon value, instead of
checking for --no-aggr, since the stats could conceivably be computed
over the runs on each CPU, and theres no reason to preclude that.

Link: http://lkml.kernel.org/r/1315437244-3788-4-git-send-email-jim.cromie@gmail.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Jim Cromie <jim.cromie@gmail.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/builtin-stat.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/tools/perf/builtin-stat.c b/tools/perf/builtin-stat.c
index af0d65b..076eda5 100644
--- a/tools/perf/builtin-stat.c
+++ b/tools/perf/builtin-stat.c
@@ -528,7 +528,7 @@ static void print_noise_pct(double total, double avg)
 
 	if (csv_output)
 		fprintf(output, "%s%.2f%%", csv_sep, pct);
-	else
+	else if (pct)
 		fprintf(output, "  ( +-%6.2f%% )", pct);
 }
 
-- 
1.6.2.5

[PATCH 10/16] perf stat: Allow tab as cvs delimiter

[Arnaldo Carvalho de Melo]

From: Jim Cromie <jim.cromie@gmail.com>

If option -x '\t' is given, convert '\t' to "\t".  This makes cvs
printing more flexible.

Link: http://lkml.kernel.org/r/1315437244-3788-5-git-send-email-jim.cromie@gmail.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Jim Cromie <jim.cromie@gmail.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/builtin-stat.c |    6 ++++--
 1 files changed, 4 insertions(+), 2 deletions(-)

diff --git a/tools/perf/builtin-stat.c b/tools/perf/builtin-stat.c
index 076eda5..0ffbd97 100644
--- a/tools/perf/builtin-stat.c
+++ b/tools/perf/builtin-stat.c
@@ -1198,9 +1198,11 @@ int cmd_stat(int argc, const char **argv, const char *prefix __used)
 		}
 	}
 
-	if (csv_sep)
+	if (csv_sep) {
 		csv_output = true;
-	else
+		if (!strcmp(csv_sep, "\\t"))
+			csv_sep = "\t";
+	} else
 		csv_sep = DEFAULT_SEPARATOR;
 
 	/*
-- 
1.6.2.5

Re: [PATCH 10/16] perf stat: Allow tab as cvs delimiter

[Paul Bolle]

On Thu, 2011-09-29 at 19:48 -0300, Arnaldo Carvalho de Melo wrote:
> From: Jim Cromie <jim.cromie@gmail.com>
> 
> If option -x '\t' is given, convert '\t' to "\t".  This makes cvs
> printing more flexible.

I found this message when I searched my mail for occurences of 'cvs'. Is
it still possible to change 'cvs' to 'csv' in the commit summary and
message? That would keep the result for grepping of either term (in
commit summaries or messages) correct.


Paul Bolle

[PATCH 11/16] perf stat: Fix spelling in comment

[Arnaldo Carvalho de Melo]

From: Jim Cromie <jim.cromie@gmail.com>

Link: http://lkml.kernel.org/r/1315437244-3788-6-git-send-email-jim.cromie@gmail.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Jim Cromie <jim.cromie@gmail.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/builtin-stat.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/tools/perf/builtin-stat.c b/tools/perf/builtin-stat.c
index 0ffbd97..b567319 100644
--- a/tools/perf/builtin-stat.c
+++ b/tools/perf/builtin-stat.c
@@ -1209,7 +1209,7 @@ int cmd_stat(int argc, const char **argv, const char *prefix __used)
 	 * let the spreadsheet do the pretty-printing
 	 */
 	if (csv_output) {
-		/* User explicitely passed -B? */
+		/* User explicitly passed -B? */
 		if (big_num_opt == 1) {
 			fprintf(stderr, "-B option not supported with -x\n");
 			usage_with_options(stat_usage, options);
-- 
1.6.2.5

[PATCH 12/16] perf tools: Make stat/record print fatal signals of the target program

[Arnaldo Carvalho de Melo]

From: Andi Kleen <ak@linux.intel.com>

When a program crashes under perf there is no message about it, unlike
when running it from bash. This can be confusing and lead to wrong
actions during debugging.

Print fatal signals in perf stat/record.

Thanks to Furat Afram for finding the problem originally

Link: http://lkml.kernel.org/r/1316122302-24306-1-git-send-email-andi@firstfloor.org
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Stephane Eranian <eranian@google.com>
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/builtin-record.c |   19 +++++++++++++++++--
 tools/perf/builtin-stat.c   |    2 ++
 2 files changed, 19 insertions(+), 2 deletions(-)

diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c
index 6b0519f..042117f 100644
--- a/tools/perf/builtin-record.c
+++ b/tools/perf/builtin-record.c
@@ -73,6 +73,7 @@ static off_t			post_processing_offset;
 
 static struct perf_session	*session;
 static const char		*cpu_list;
+static const char               *progname;
 
 static void advance_output(size_t size)
 {
@@ -137,17 +138,29 @@ static void mmap_read(struct perf_mmap *md)
 
 static volatile int done = 0;
 static volatile int signr = -1;
+static volatile int child_finished = 0;
 
 static void sig_handler(int sig)
 {
+	if (sig == SIGCHLD)
+		child_finished = 1;
+
 	done = 1;
 	signr = sig;
 }
 
 static void sig_atexit(void)
 {
-	if (child_pid > 0)
-		kill(child_pid, SIGTERM);
+	int status;
+
+	if (child_pid > 0) {
+		if (!child_finished)
+			kill(child_pid, SIGTERM);
+
+		wait(&status);
+		if (WIFSIGNALED(status))
+			psignal(WTERMSIG(status), progname);
+	}
 
 	if (signr == -1 || signr == SIGUSR1)
 		return;
@@ -445,6 +458,8 @@ static int __cmd_record(int argc, const char **argv)
 	char buf;
 	struct machine *machine;
 
+	progname = argv[0];
+
 	page_size = sysconf(_SC_PAGE_SIZE);
 
 	atexit(sig_atexit);
diff --git a/tools/perf/builtin-stat.c b/tools/perf/builtin-stat.c
index b567319..7ce65f5 100644
--- a/tools/perf/builtin-stat.c
+++ b/tools/perf/builtin-stat.c
@@ -495,6 +495,8 @@ static int run_perf_stat(int argc __used, const char **argv)
 	if (forks) {
 		close(go_pipe[1]);
 		wait(&status);
+		if (WIFSIGNALED(status))
+			psignal(WTERMSIG(status), argv[0]);
 	} else {
 		while(!done) sleep(1);
 	}
-- 
1.6.2.5

[PATCH 13/16] perf: Support setting the disassembler style

[Arnaldo Carvalho de Melo]

From: Andi Kleen <ak@linux.intel.com>

Add -M option to report/annotate to pass directly to objdump.  This
allows to use -M intel for intel style disassembler syntax, which is
useful for people who are very used to the Intel syntax.

Link: http://lkml.kernel.org/r/1316122302-24306-2-git-send-email-andi@firstfloor.org
[committer note: Add missing Documentation bits, fixup conflicts with 3e6a2a7]
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Stephane Eranian <eranian@google.com>
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/Documentation/perf-annotate.txt |    3 +++
 tools/perf/Documentation/perf-report.txt   |    3 +++
 tools/perf/builtin-annotate.c              |    2 ++
 tools/perf/builtin-report.c                |    2 ++
 tools/perf/util/annotate.c                 |    6 +++++-
 tools/perf/util/annotate.h                 |    2 ++
 6 files changed, 17 insertions(+), 1 deletions(-)

diff --git a/tools/perf/Documentation/perf-annotate.txt b/tools/perf/Documentation/perf-annotate.txt
index 98a31e3..0102d83 100644
--- a/tools/perf/Documentation/perf-annotate.txt
+++ b/tools/perf/Documentation/perf-annotate.txt
@@ -83,6 +83,9 @@ OPTIONS
 --symfs=<directory>::
         Look for files with symbols relative to this directory.
 
+-M::
+--disassembler-style=:: Set disassembler style for objdump.
+
 SEE ALSO
 --------
 linkperf:perf-record[1], linkperf:perf-report[1]
diff --git a/tools/perf/Documentation/perf-report.txt b/tools/perf/Documentation/perf-report.txt
index 04253c0..6349b6c 100644
--- a/tools/perf/Documentation/perf-report.txt
+++ b/tools/perf/Documentation/perf-report.txt
@@ -134,6 +134,9 @@ OPTIONS
 	CPUs are specified with -: 0-2. Default is to report samples on all
 	CPUs.
 
+-M::
+--disassembler-style=:: Set disassembler style for objdump.
+
 SEE ALSO
 --------
 linkperf:perf-stat[1]
diff --git a/tools/perf/builtin-annotate.c b/tools/perf/builtin-annotate.c
index c5be288..cf68819 100644
--- a/tools/perf/builtin-annotate.c
+++ b/tools/perf/builtin-annotate.c
@@ -273,6 +273,8 @@ static const struct option options[] = {
 		    "Interleave source code with assembly code (default)"),
 	OPT_BOOLEAN('0', "asm-raw", &symbol_conf.annotate_asm_raw,
 		    "Display raw encoding of assembly instructions (default)"),
+	OPT_STRING('M', "disassembler-style", &disassembler_style, "disassembler style",
+		   "Specify disassembler style (e.g. -M intel for intel syntax)"),
 	OPT_END()
 };
 
diff --git a/tools/perf/builtin-report.c b/tools/perf/builtin-report.c
index d7ff277..a0673ee 100644
--- a/tools/perf/builtin-report.c
+++ b/tools/perf/builtin-report.c
@@ -487,6 +487,8 @@ static const struct option options[] = {
 	OPT_STRING(0, "symfs", &symbol_conf.symfs, "directory",
 		    "Look for files with symbols relative to this directory"),
 	OPT_STRING('c', "cpu", &cpu_list, "cpu", "list of cpus to profile"),
+	OPT_STRING('M', "disassembler-style", &disassembler_style, "disassembler style",
+		   "Specify disassembler style (e.g. -M intel for intel syntax)"),
 	OPT_END()
 };
 
diff --git a/tools/perf/util/annotate.c b/tools/perf/util/annotate.c
index 01d36ba..bc8f477 100644
--- a/tools/perf/util/annotate.c
+++ b/tools/perf/util/annotate.c
@@ -16,6 +16,8 @@
 #include "annotate.h"
 #include <pthread.h>
 
+const char 	*disassembler_style;
+
 int symbol__annotate_init(struct map *map __used, struct symbol *sym)
 {
 	struct annotation *notes = symbol__annotation(sym);
@@ -323,9 +325,11 @@ fallback:
 		 dso, dso->long_name, sym, sym->name);
 
 	snprintf(command, sizeof(command),
-		 "objdump --start-address=0x%016" PRIx64
+		 "objdump %s%s --start-address=0x%016" PRIx64
 		 " --stop-address=0x%016" PRIx64
 		 " -d %s %s -C %s|grep -v %s|expand",
+		 disassembler_style ? "-M " : "",
+		 disassembler_style ? disassembler_style : "",
 		 map__rip_2objdump(map, sym->start),
 		 map__rip_2objdump(map, sym->end),
 		 symbol_conf.annotate_asm_raw ? "" : "--no-show-raw",
diff --git a/tools/perf/util/annotate.h b/tools/perf/util/annotate.h
index c2c2868..6ede128 100644
--- a/tools/perf/util/annotate.h
+++ b/tools/perf/util/annotate.h
@@ -100,4 +100,6 @@ int symbol__tui_annotate(struct symbol *sym, struct map *map, int evidx,
 			 int refresh);
 #endif
 
+extern const char	*disassembler_style;
+
 #endif	/* __PERF_ANNOTATE_H */
-- 
1.6.2.5

[PATCH 14/16] perf report: Fix stdio event name header printing

[Arnaldo Carvalho de Melo]

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

In the past we tried to avoid printing the name of the event when just
one event was found in the perf.data file, after some refactorings it
ended up not printing the event name if just one hist_entry was found in
one of the events.

Fix it by always printing the name of the event, even if just one is
found.

Reported-by: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/n/tip-kikr0c7ou55bd9caok8569rf@git.kernel.org
Cc: David Ahern <dsahern@gmail.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/builtin-report.c |    5 +----
 1 files changed, 1 insertions(+), 4 deletions(-)

diff --git a/tools/perf/builtin-report.c b/tools/perf/builtin-report.c
index a0673ee..3d58334 100644
--- a/tools/perf/builtin-report.c
+++ b/tools/perf/builtin-report.c
@@ -229,10 +229,7 @@ static int perf_evlist__tty_browse_hists(struct perf_evlist *evlist,
 
 	list_for_each_entry(pos, &evlist->entries, node) {
 		struct hists *hists = &pos->hists;
-		const char *evname = NULL;
-
-		if (rb_first(&hists->entries) != rb_last(&hists->entries))
-			evname = event_name(pos);
+		const char *evname = event_name(pos);
 
 		hists__fprintf_nr_sample_events(hists, evname, stdout);
 		hists__fprintf(hists, NULL, false, stdout);
-- 
1.6.2.5

[PATCH 15/16] perf sched: Fix script command documentation

[Arnaldo Carvalho de Melo]

From: Jiri Olsa <jolsa@redhat.com>

Fixed leftover from trace -> script rename.

Link: http://lkml.kernel.org/r/1317114995-4534-1-git-send-email-jolsa@redhat.com
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Jiri Olsa <jolsa@redhat.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/Documentation/perf-sched.txt |    6 +++---
 1 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/tools/perf/Documentation/perf-sched.txt b/tools/perf/Documentation/perf-sched.txt
index 46822d5..5b212b5 100644
--- a/tools/perf/Documentation/perf-sched.txt
+++ b/tools/perf/Documentation/perf-sched.txt
@@ -8,7 +8,7 @@ perf-sched - Tool to trace/measure scheduler properties (latencies)
 SYNOPSIS
 --------
 [verse]
-'perf sched' {record|latency|map|replay|trace}
+'perf sched' {record|latency|map|replay|script}
 
 DESCRIPTION
 -----------
@@ -20,8 +20,8 @@ There are five variants of perf sched:
   'perf sched latency' to report the per task scheduling latencies
   and other scheduling properties of the workload.
 
-  'perf sched trace' to see a detailed trace of the workload that
-  was recorded.
+  'perf sched script' to see a detailed trace of the workload that
+   was recorded (aliased to 'perf script' for now).
 
   'perf sched replay' to simulate the workload that was recorded
   via perf sched record. (this is done by starting up mockup threads
-- 
1.6.2.5

[PATCH 16/16] perf symbols: Treat all memory maps without dso file as loaded

[Arnaldo Carvalho de Melo]

From: Jiri Olsa <jolsa@redhat.com>

The stack/vdso/heap memory maps dont have any dso file.  Setting the
perf dso objects as 'loaded' for these maps, we avoid unnecessary
warnings like:

  "Failed to open [stack], continuing without symbols"

All map__find_* functions still return NULL when searching for symbols
in these maps.

Link: http://lkml.kernel.org/r/20110824131834.GA2007@jolsa.brq.redhat.com
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Jiri Olsa <jolsa@redhat.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/util/map.c |   24 ++++++++++++++++++------
 1 files changed, 18 insertions(+), 6 deletions(-)

diff --git a/tools/perf/util/map.c b/tools/perf/util/map.c
index aa2f9fd..9cf0d43 100644
--- a/tools/perf/util/map.c
+++ b/tools/perf/util/map.c
@@ -18,6 +18,13 @@ static inline int is_anon_memory(const char *filename)
 	return strcmp(filename, "//anon") == 0;
 }
 
+static inline int is_no_dso_memory(const char *filename)
+{
+	return !strcmp(filename, "[stack]") ||
+	       !strcmp(filename, "[vdso]")  ||
+	       !strcmp(filename, "[heap]");
+}
+
 void map__init(struct map *self, enum map_type type,
 	       u64 start, u64 end, u64 pgoff, struct dso *dso)
 {
@@ -42,9 +49,10 @@ struct map *map__new(struct list_head *dsos__list, u64 start, u64 len,
 	if (self != NULL) {
 		char newfilename[PATH_MAX];
 		struct dso *dso;
-		int anon;
+		int anon, no_dso;
 
 		anon = is_anon_memory(filename);
+		no_dso = is_no_dso_memory(filename);
 
 		if (anon) {
 			snprintf(newfilename, sizeof(newfilename), "/tmp/perf-%d.map", pid);
@@ -57,12 +65,16 @@ struct map *map__new(struct list_head *dsos__list, u64 start, u64 len,
 
 		map__init(self, type, start, start + len, pgoff, dso);
 
-		if (anon) {
-set_identity:
+		if (anon || no_dso) {
 			self->map_ip = self->unmap_ip = identity__map_ip;
-		} else if (strcmp(filename, "[vdso]") == 0) {
-			dso__set_loaded(dso, self->type);
-			goto set_identity;
+
+			/*
+			 * Set memory without DSO as loaded. All map__find_*
+			 * functions still return NULL, and we avoid the
+			 * unnecessary map__load warning.
+			 */
+			if (no_dso)
+				dso__set_loaded(dso, self->type);
 		}
 	}
 	return self;
-- 
1.6.2.5

Re: [GIT PULL 00/16] perf/core improvements and fixes

[Ingo Molnar]

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

> Hi Ingo,
> 
>         Please consider pulling from:
> 
> git://github.com/acmel/linux.git perf/core
> 
> 	This was rebased against your perf/core on tesla.
> 
> Regards,
> 
> - Arnaldo
> 
> Andi Kleen (2):
>   perf tools: Make stat/record print fatal signals of the target program
>   perf: Support setting the disassembler style
> 
> Arnaldo Carvalho de Melo (6):
>   perf symbols: Stop using 'self' in map_groups__ methods
>   perf buildid-list: Add option to show the running kernel build id
>   perf buildid-list: Support showing the build id in an ELF file
>   perf top browser: Fix up line width calculation
>   perf top: Improve lost events warning
>   perf report: Fix stdio event name header printing
> 
> Jim Cromie (5):
>   perf stat: Add --log-fd <N> option to redirect stderr elsewhere
>   perf stat: Fix +- nan% in --no-aggr runs
>   perf stat: Suppress printing std-dev when its 0
>   perf stat: Allow tab as cvs delimiter
>   perf stat: Fix spelling in comment
> 
> Jiri Olsa (2):
>   perf sched: Fix script command documentation
>   perf symbols: Treat all memory maps without dso file as loaded
> 
> Neil Horman (1):
>   perf script: Add drop monitor script
> 
>  tools/perf/Documentation/perf-annotate.txt         |    3 +
>  tools/perf/Documentation/perf-buildid-list.txt     |    6 +
>  tools/perf/Documentation/perf-report.txt           |    3 +
>  tools/perf/Documentation/perf-sched.txt            |    6 +-
>  tools/perf/Documentation/perf-stat.txt             |   11 ++-
>  tools/perf/builtin-annotate.c                      |    2 +
>  tools/perf/builtin-buildid-list.c                  |   52 ++++++++++-
>  tools/perf/builtin-record.c                        |   19 ++++-
>  tools/perf/builtin-report.c                        |    7 +-
>  tools/perf/builtin-stat.c                          |   35 ++++++-
>  tools/perf/builtin-top.c                           |   34 ++++----
>  .../perf/scripts/python/bin/net_dropmonitor-record |    2 +
>  .../perf/scripts/python/bin/net_dropmonitor-report |    4 +
>  tools/perf/scripts/python/net_dropmonitor.py       |   72 ++++++++++++++
>  tools/perf/util/annotate.c                         |    6 +-
>  tools/perf/util/annotate.h                         |    2 +
>  tools/perf/util/map.c                              |   98 +++++++++++---------
>  tools/perf/util/map.h                              |   42 ++++----
>  tools/perf/util/top.h                              |    3 +
>  tools/perf/util/ui/browsers/top.c                  |   28 +++++-
>  20 files changed, 333 insertions(+), 102 deletions(-)
>  create mode 100755 tools/perf/scripts/python/bin/net_dropmonitor-record
>  create mode 100755 tools/perf/scripts/python/bin/net_dropmonitor-report
>  create mode 100755 tools/perf/scripts/python/net_dropmonitor.py

Pulled, thanks a lot Arnaldo!

	Ingo