[GIT PULL 00/12] perf/core improvements and fixes

[GIT PULL 00/12] perf/core improvements and fixes

[Arnaldo Carvalho de Melo]

From: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>

Hi Ingo,

	Please consider pulling,

- Arnaldo

The following changes since commit 0ed1e0bee0b2c6b4cc6d7a63787739a9d3ac8aa8:

  Merge tag 'perf-core-for-mingo' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux into perf/core (2013-11-30 13:42:48 +0100)

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 f885037ef6400ce4a4c122a88845dea2c9bca256:

  perf trace: Honour -m option (2013-12-02 16:29:40 -0300)

----------------------------------------------------------------
perf/tools improvements and fixes:

. Honour -m option in 'trace', the tool was offering the option to
  set the mmap size, but wasn't using it when doing the actual mmap
  on the events file descriptors, fix from Jiri Olsa.

. Correct the message in feature-libnuma checking, swowing the right
  devel package names for various distros, from Dongsheng Yang.

. Polish 'readn' function and introduce its counterpart, 'writen', from
  Jiri Olsa.

. Start moving timechart state from global variables to a 'perf_tool' derived
  'timechart' struct.

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

----------------------------------------------------------------
Arnaldo Carvalho de Melo (4):
      perf timechart: Introduce tool struct
      perf timechart: Move all_data per_pid list to 'struct timechart'
      perf timechart: Move power_events list to 'struct timechart'
      perf timechart: Move wake_events list to 'struct timechart'

Dongsheng Yang (1):
      perf tools: Correct the message in feature-libnuma checking.

Jiri Olsa (7):
      perf record: Unify data output code into perf_record__write function
      perf tools: Use correct return type for readn function
      perf tools: Fine tune readn function
      perf tools: Add writen function
      perf tools: Add perf_data_file__write interface
      perf tools: Include test-stackprotector-all.c in test-all
      perf trace: Honour -m option

 tools/perf/builtin-record.c                 |  20 +-
 tools/perf/builtin-timechart.c              | 364 +++++++++++++++-------------
 tools/perf/builtin-trace.c                  |   2 +-
 tools/perf/config/Makefile                  |   2 +-
 tools/perf/config/feature-checks/test-all.c |   5 +
 tools/perf/util/data.c                      |   6 +
 tools/perf/util/data.h                      |  14 +-
 tools/perf/util/header.c                    |  18 +-
 tools/perf/util/session.c                   |   2 +-
 tools/perf/util/util.c                      |  32 ++-
 tools/perf/util/util.h                      |   3 +-
 11 files changed, 256 insertions(+), 212 deletions(-)

[PATCH 01/12] perf record: Unify data output code into perf_record__write function

[Arnaldo Carvalho de Melo]

From: Jiri Olsa <jolsa@redhat.com>

Unifying current 2 data output functions do_write_output and
write_output into single one perf_record__write.

Signed-off-by: Jiri Olsa <jolsa@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1385634619-8129-2-git-send-email-jolsa@redhat.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/builtin-record.c | 20 ++++++--------------
 1 file changed, 6 insertions(+), 14 deletions(-)

diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c
index 65615a8bc25e..d93e2eef0979 100644
--- a/tools/perf/builtin-record.c
+++ b/tools/perf/builtin-record.c
@@ -76,7 +76,7 @@ struct perf_record {
 	long			samples;
 };
 
-static int do_write_output(struct perf_record *rec, void *buf, size_t size)
+static int perf_record__write(struct perf_record *rec, void *buf, size_t size)
 {
 	struct perf_data_file *file = &rec->file;
 
@@ -97,21 +97,13 @@ static int do_write_output(struct perf_record *rec, void *buf, size_t size)
 	return 0;
 }
 
-static int write_output(struct perf_record *rec, void *buf, size_t size)
-{
-	return do_write_output(rec, buf, size);
-}
-
 static int process_synthesized_event(struct perf_tool *tool,
 				     union perf_event *event,
 				     struct perf_sample *sample __maybe_unused,
 				     struct machine *machine __maybe_unused)
 {
 	struct perf_record *rec = container_of(tool, struct perf_record, tool);
-	if (write_output(rec, event, event->header.size) < 0)
-		return -1;
-
-	return 0;
+	return perf_record__write(rec, event, event->header.size);
 }
 
 static int perf_record__mmap_read(struct perf_record *rec,
@@ -136,7 +128,7 @@ static int perf_record__mmap_read(struct perf_record *rec,
 		size = md->mask + 1 - (old & md->mask);
 		old += size;
 
-		if (write_output(rec, buf, size) < 0) {
+		if (perf_record__write(rec, buf, size) < 0) {
 			rc = -1;
 			goto out;
 		}
@@ -146,7 +138,7 @@ static int perf_record__mmap_read(struct perf_record *rec,
 	size = head - old;
 	old += size;
 
-	if (write_output(rec, buf, size) < 0) {
+	if (perf_record__write(rec, buf, size) < 0) {
 		rc = -1;
 		goto out;
 	}
@@ -335,8 +327,8 @@ static int perf_record__mmap_read_all(struct perf_record *rec)
 	}
 
 	if (perf_header__has_feat(&rec->session->header, HEADER_TRACING_DATA))
-		rc = write_output(rec, &finished_round_event,
-				  sizeof(finished_round_event));
+		rc = perf_record__write(rec, &finished_round_event,
+					sizeof(finished_round_event));
 
 out:
 	return rc;
-- 
1.8.1.4

[PATCH 02/12] perf tools: Use correct return type for readn function

[Arnaldo Carvalho de Melo]

From: Jiri Olsa <jolsa@redhat.com>

Changing readn function return type to ssize_t because read returns
ssize_t not int.

Changing callers holding variable types as well.

Signed-off-by: Jiri Olsa <jolsa@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1385634619-8129-3-git-send-email-jolsa@redhat.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/util/header.c  | 18 +++++++++---------
 tools/perf/util/session.c |  2 +-
 tools/perf/util/util.c    |  4 ++--
 tools/perf/util/util.h    |  2 +-
 4 files changed, 13 insertions(+), 13 deletions(-)

diff --git a/tools/perf/util/header.c b/tools/perf/util/header.c
index 1cd035708931..3e755f2bfe8f 100644
--- a/tools/perf/util/header.c
+++ b/tools/perf/util/header.c
@@ -1709,7 +1709,7 @@ static int process_nrcpus(struct perf_file_section *section __maybe_unused,
 			  struct perf_header *ph, int fd,
 			  void *data __maybe_unused)
 {
-	size_t ret;
+	ssize_t ret;
 	u32 nr;
 
 	ret = readn(fd, &nr, sizeof(nr));
@@ -1753,7 +1753,7 @@ static int process_total_mem(struct perf_file_section *section __maybe_unused,
 			     void *data __maybe_unused)
 {
 	uint64_t mem;
-	size_t ret;
+	ssize_t ret;
 
 	ret = readn(fd, &mem, sizeof(mem));
 	if (ret != sizeof(mem))
@@ -1822,7 +1822,7 @@ static int process_cmdline(struct perf_file_section *section __maybe_unused,
 			   struct perf_header *ph, int fd,
 			   void *data __maybe_unused)
 {
-	size_t ret;
+	ssize_t ret;
 	char *str;
 	u32 nr, i;
 	struct strbuf sb;
@@ -1858,7 +1858,7 @@ static int process_cpu_topology(struct perf_file_section *section __maybe_unused
 				struct perf_header *ph, int fd,
 				void *data __maybe_unused)
 {
-	size_t ret;
+	ssize_t ret;
 	u32 nr, i;
 	char *str;
 	struct strbuf sb;
@@ -1914,7 +1914,7 @@ static int process_numa_topology(struct perf_file_section *section __maybe_unuse
 				 struct perf_header *ph, int fd,
 				 void *data __maybe_unused)
 {
-	size_t ret;
+	ssize_t ret;
 	u32 nr, node, i;
 	char *str;
 	uint64_t mem_total, mem_free;
@@ -1974,7 +1974,7 @@ static int process_pmu_mappings(struct perf_file_section *section __maybe_unused
 				struct perf_header *ph, int fd,
 				void *data __maybe_unused)
 {
-	size_t ret;
+	ssize_t ret;
 	char *name;
 	u32 pmu_num;
 	u32 type;
@@ -2534,7 +2534,7 @@ static int check_magic_endian(u64 magic, uint64_t hdr_sz,
 int perf_file_header__read(struct perf_file_header *header,
 			   struct perf_header *ph, int fd)
 {
-	int ret;
+	ssize_t ret;
 
 	lseek(fd, 0, SEEK_SET);
 
@@ -2628,7 +2628,7 @@ static int perf_file_header__read_pipe(struct perf_pipe_file_header *header,
 				       struct perf_header *ph, int fd,
 				       bool repipe)
 {
-	int ret;
+	ssize_t ret;
 
 	ret = readn(fd, header, sizeof(*header));
 	if (ret <= 0)
@@ -2669,7 +2669,7 @@ static int read_attr(int fd, struct perf_header *ph,
 	struct perf_event_attr *attr = &f_attr->attr;
 	size_t sz, left;
 	size_t our_sz = sizeof(f_attr->attr);
-	int ret;
+	ssize_t ret;
 
 	memset(f_attr, 0, sizeof(*f_attr));
 
diff --git a/tools/perf/util/session.c b/tools/perf/util/session.c
index b0b15e213df5..4ce146bae552 100644
--- a/tools/perf/util/session.c
+++ b/tools/perf/util/session.c
@@ -1158,7 +1158,7 @@ static int __perf_session__process_pipe_events(struct perf_session *session,
 	void *buf = NULL;
 	int skip = 0;
 	u64 head;
-	int err;
+	ssize_t err;
 	void *p;
 
 	perf_tool__fill_defaults(tool);
diff --git a/tools/perf/util/util.c b/tools/perf/util/util.c
index 28a0a89c1f73..9440481e9092 100644
--- a/tools/perf/util/util.c
+++ b/tools/perf/util/util.c
@@ -151,12 +151,12 @@ unsigned long convert_unit(unsigned long value, char *unit)
 	return value;
 }
 
-int readn(int fd, void *buf, size_t n)
+ssize_t readn(int fd, void *buf, size_t n)
 {
 	void *buf_start = buf;
 
 	while (n) {
-		int ret = read(fd, buf, n);
+		ssize_t ret = read(fd, buf, n);
 
 		if (ret <= 0)
 			return ret;
diff --git a/tools/perf/util/util.h b/tools/perf/util/util.h
index c8f362daba87..9f6b928f6b2b 100644
--- a/tools/perf/util/util.h
+++ b/tools/perf/util/util.h
@@ -253,7 +253,7 @@ bool strlazymatch(const char *str, const char *pat);
 int strtailcmp(const char *s1, const char *s2);
 char *strxfrchar(char *s, char from, char to);
 unsigned long convert_unit(unsigned long value, char *unit);
-int readn(int fd, void *buf, size_t size);
+ssize_t readn(int fd, void *buf, size_t n);
 
 struct perf_event_attr;
 
-- 
1.8.1.4

[PATCH 03/12] perf tools: Fine tune readn function

[Arnaldo Carvalho de Melo]

From: Jiri Olsa <jolsa@redhat.com>

Added a 'left' variable to make the flow clearer, and added a debug
check for the return value - returning 'n' is more obvious.

Added small comment for readn.

Signed-off-by: Jiri Olsa <jolsa@redhat.com>
Original-patch-by: Ingo Molnar <mingo@kernel.org>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1385634619-8129-4-git-send-email-jolsa@redhat.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/util/util.c | 16 +++++++++++-----
 1 file changed, 11 insertions(+), 5 deletions(-)

diff --git a/tools/perf/util/util.c b/tools/perf/util/util.c
index 9440481e9092..6ea0b4ae9569 100644
--- a/tools/perf/util/util.c
+++ b/tools/perf/util/util.c
@@ -6,6 +6,7 @@
 #endif
 #include <stdio.h>
 #include <stdlib.h>
+#include <linux/kernel.h>
 
 /*
  * XXX We need to find a better place for these things...
@@ -151,21 +152,26 @@ unsigned long convert_unit(unsigned long value, char *unit)
 	return value;
 }
 
+/*
+ * Read exactly 'n' bytes or return an error.
+ */
 ssize_t readn(int fd, void *buf, size_t n)
 {
 	void *buf_start = buf;
+	size_t left = n;
 
-	while (n) {
-		ssize_t ret = read(fd, buf, n);
+	while (left) {
+		ssize_t ret = read(fd, buf, left);
 
 		if (ret <= 0)
 			return ret;
 
-		n -= ret;
-		buf += ret;
+		left -= ret;
+		buf  += ret;
 	}
 
-	return buf - buf_start;
+	BUG_ON((size_t)(buf - buf_start) != n);
+	return n;
 }
 
 size_t hex_width(u64 v)
-- 
1.8.1.4

[PATCH 04/12] perf tools: Add writen function

[Arnaldo Carvalho de Melo]

From: Jiri Olsa <jolsa@redhat.com>

Adding 'writen' function as a synchronous wrapper for write syscall with
following prototype:

  ssize_t writen(int fd, void *buf, size_t n)

Returns the number of bytes written on success or -1 in case of err.

Signed-off-by: Jiri Olsa <jolsa@redhat.com>
Requested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1385634619-8129-5-git-send-email-jolsa@redhat.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/util/util.c | 24 +++++++++++++++++++-----
 tools/perf/util/util.h |  1 +
 2 files changed, 20 insertions(+), 5 deletions(-)

diff --git a/tools/perf/util/util.c b/tools/perf/util/util.c
index 6ea0b4ae9569..b1d5376b9dd9 100644
--- a/tools/perf/util/util.c
+++ b/tools/perf/util/util.c
@@ -152,16 +152,14 @@ unsigned long convert_unit(unsigned long value, char *unit)
 	return value;
 }
 
-/*
- * Read exactly 'n' bytes or return an error.
- */
-ssize_t readn(int fd, void *buf, size_t n)
+static ssize_t ion(bool is_read, int fd, void *buf, size_t n)
 {
 	void *buf_start = buf;
 	size_t left = n;
 
 	while (left) {
-		ssize_t ret = read(fd, buf, left);
+		ssize_t ret = is_read ? read(fd, buf, left) :
+					write(fd, buf, left);
 
 		if (ret <= 0)
 			return ret;
@@ -174,6 +172,22 @@ ssize_t readn(int fd, void *buf, size_t n)
 	return n;
 }
 
+/*
+ * Read exactly 'n' bytes or return an error.
+ */
+ssize_t readn(int fd, void *buf, size_t n)
+{
+	return ion(true, fd, buf, n);
+}
+
+/*
+ * Write exactly 'n' bytes or return an error.
+ */
+ssize_t writen(int fd, void *buf, size_t n)
+{
+	return ion(false, fd, buf, n);
+}
+
 size_t hex_width(u64 v)
 {
 	size_t n = 1;
diff --git a/tools/perf/util/util.h b/tools/perf/util/util.h
index 9f6b928f6b2b..ce0f73d4d91f 100644
--- a/tools/perf/util/util.h
+++ b/tools/perf/util/util.h
@@ -254,6 +254,7 @@ int strtailcmp(const char *s1, const char *s2);
 char *strxfrchar(char *s, char from, char to);
 unsigned long convert_unit(unsigned long value, char *unit);
 ssize_t readn(int fd, void *buf, size_t n);
+ssize_t writen(int fd, void *buf, size_t n);
 
 struct perf_event_attr;
 
-- 
1.8.1.4

[PATCH 05/12] perf tools: Add perf_data_file__write interface

[Arnaldo Carvalho de Melo]

From: Jiri Olsa <jolsa@redhat.com>

Adding perf_data_file__write interface to centralize output to files.
The function prototype is:

  ssize_t perf_data_file__write(struct perf_data_file *file,
                                void *buf, size_t size);

Returns number of bytes written or -1 in case of error.

NOTE: Also indenting 'struct perf_data_file' members, no functional
      change done.

Signed-off-by: Jiri Olsa <jolsa@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1385634619-8129-6-git-send-email-jolsa@redhat.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/util/data.c |  6 ++++++
 tools/perf/util/data.h | 14 ++++++++------
 2 files changed, 14 insertions(+), 6 deletions(-)

diff --git a/tools/perf/util/data.c b/tools/perf/util/data.c
index 7d09faf85cf1..1fbcd8bdc11b 100644
--- a/tools/perf/util/data.c
+++ b/tools/perf/util/data.c
@@ -118,3 +118,9 @@ void perf_data_file__close(struct perf_data_file *file)
 {
 	close(file->fd);
 }
+
+ssize_t perf_data_file__write(struct perf_data_file *file,
+			      void *buf, size_t size)
+{
+	return writen(file->fd, buf, size);
+}
diff --git a/tools/perf/util/data.h b/tools/perf/util/data.h
index 8c2df80152a5..2b15d0c95c7f 100644
--- a/tools/perf/util/data.h
+++ b/tools/perf/util/data.h
@@ -9,12 +9,12 @@ enum perf_data_mode {
 };
 
 struct perf_data_file {
-	const char *path;
-	int fd;
-	bool is_pipe;
-	bool force;
-	unsigned long size;
-	enum perf_data_mode mode;
+	const char		*path;
+	int			 fd;
+	bool			 is_pipe;
+	bool			 force;
+	unsigned long		 size;
+	enum perf_data_mode	 mode;
 };
 
 static inline bool perf_data_file__is_read(struct perf_data_file *file)
@@ -44,5 +44,7 @@ static inline unsigned long perf_data_file__size(struct perf_data_file *file)
 
 int perf_data_file__open(struct perf_data_file *file);
 void perf_data_file__close(struct perf_data_file *file);
+ssize_t perf_data_file__write(struct perf_data_file *file,
+			      void *buf, size_t size);
 
 #endif /* __PERF_DATA_H */
-- 
1.8.1.4

[PATCH 06/12] perf timechart: Introduce tool struct

[Arnaldo Carvalho de Melo]

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

To avoid having all those global variables and to use the interface to
event processing that is based on passing a 'perf_tool' struct that
should be embedded in a per tool specific struct passed to all the
sample processing callbacks.

There are some more globals to move, next patches will do it.

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stanislav Fomichev <stfomichev@yandex-team.ru>
Cc: Stephane Eranian <eranian@google.com>
Link: http://lkml.kernel.org/n/tip-0iah65pq796ezbk5u1lzwy1k@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/builtin-timechart.c | 216 ++++++++++++++++++++++-------------------
 1 file changed, 114 insertions(+), 102 deletions(-)

diff --git a/tools/perf/builtin-timechart.c b/tools/perf/builtin-timechart.c
index 680632d7e26a..e2d62f1a96e4 100644
--- a/tools/perf/builtin-timechart.c
+++ b/tools/perf/builtin-timechart.c
@@ -41,19 +41,18 @@
 #define SUPPORT_OLD_POWER_EVENTS 1
 #define PWR_EVENT_EXIT -1
 
-static int proc_num = 15;
-
-static unsigned int	numcpus;
-static u64		min_freq;	/* Lowest CPU frequency seen */
-static u64		max_freq;	/* Highest CPU frequency seen */
-static u64		turbo_frequency;
-
-static u64		first_time, last_time;
-
-static bool		power_only;
-static bool		tasks_only;
-static bool		with_backtrace;
-
+struct timechart {
+	struct perf_tool	tool;
+	int			proc_num;
+	unsigned int		numcpus;
+	u64			min_freq,	/* Lowest CPU frequency seen */
+				max_freq,	/* Highest CPU frequency seen */
+				turbo_frequency,
+				first_time, last_time;
+	bool			power_only,
+				tasks_only,
+				with_backtrace;
+};
 
 struct per_pidcomm;
 struct cpu_sample;
@@ -326,7 +325,7 @@ static void c_state_end(int cpu, u64 timestamp)
 	power_events = pwr;
 }
 
-static void p_state_change(int cpu, u64 timestamp, u64 new_freq)
+static void p_state_change(struct timechart *tchart, int cpu, u64 timestamp, u64 new_freq)
 {
 	struct power_event *pwr;
 
@@ -345,21 +344,21 @@ static void p_state_change(int cpu, u64 timestamp, u64 new_freq)
 	pwr->next = power_events;
 
 	if (!pwr->start_time)
-		pwr->start_time = first_time;
+		pwr->start_time = tchart->first_time;
 
 	power_events = pwr;
 
 	cpus_pstate_state[cpu] = new_freq;
 	cpus_pstate_start_times[cpu] = timestamp;
 
-	if ((u64)new_freq > max_freq)
-		max_freq = new_freq;
+	if ((u64)new_freq > tchart->max_freq)
+		tchart->max_freq = new_freq;
 
-	if (new_freq < min_freq || min_freq == 0)
-		min_freq = new_freq;
+	if (new_freq < tchart->min_freq || tchart->min_freq == 0)
+		tchart->min_freq = new_freq;
 
-	if (new_freq == max_freq - 1000)
-			turbo_frequency = max_freq;
+	if (new_freq == tchart->max_freq - 1000)
+		tchart->turbo_frequency = tchart->max_freq;
 }
 
 static void sched_wakeup(int cpu, u64 timestamp, int waker, int wakee,
@@ -506,36 +505,40 @@ exit:
 	return p;
 }
 
-typedef int (*tracepoint_handler)(struct perf_evsel *evsel,
+typedef int (*tracepoint_handler)(struct timechart *tchart,
+				  struct perf_evsel *evsel,
 				  struct perf_sample *sample,
 				  const char *backtrace);
 
-static int process_sample_event(struct perf_tool *tool __maybe_unused,
+static int process_sample_event(struct perf_tool *tool,
 				union perf_event *event,
 				struct perf_sample *sample,
 				struct perf_evsel *evsel,
-				struct machine *machine __maybe_unused)
+				struct machine *machine)
 {
+	struct timechart *tchart = container_of(tool, struct timechart, tool);
+
 	if (evsel->attr.sample_type & PERF_SAMPLE_TIME) {
-		if (!first_time || first_time > sample->time)
-			first_time = sample->time;
-		if (last_time < sample->time)
-			last_time = sample->time;
+		if (!tchart->first_time || tchart->first_time > sample->time)
+			tchart->first_time = sample->time;
+		if (tchart->last_time < sample->time)
+			tchart->last_time = sample->time;
 	}
 
-	if (sample->cpu > numcpus)
-		numcpus = sample->cpu;
+	if (sample->cpu > tchart->numcpus)
+		tchart->numcpus = sample->cpu;
 
 	if (evsel->handler != NULL) {
 		tracepoint_handler f = evsel->handler;
-		return f(evsel, sample, cat_backtrace(event, sample, machine));
+		return f(tchart, evsel, sample, cat_backtrace(event, sample, machine));
 	}
 
 	return 0;
 }
 
 static int
-process_sample_cpu_idle(struct perf_evsel *evsel,
+process_sample_cpu_idle(struct timechart *tchart __maybe_unused,
+			struct perf_evsel *evsel,
 			struct perf_sample *sample,
 			const char *backtrace __maybe_unused)
 {
@@ -550,19 +553,21 @@ process_sample_cpu_idle(struct perf_evsel *evsel,
 }
 
 static int
-process_sample_cpu_frequency(struct perf_evsel *evsel,
+process_sample_cpu_frequency(struct timechart *tchart,
+			     struct perf_evsel *evsel,
 			     struct perf_sample *sample,
 			     const char *backtrace __maybe_unused)
 {
 	u32 state = perf_evsel__intval(evsel, sample, "state");
 	u32 cpu_id = perf_evsel__intval(evsel, sample, "cpu_id");
 
-	p_state_change(cpu_id, sample->time, state);
+	p_state_change(tchart, cpu_id, sample->time, state);
 	return 0;
 }
 
 static int
-process_sample_sched_wakeup(struct perf_evsel *evsel,
+process_sample_sched_wakeup(struct timechart *tchart __maybe_unused,
+			    struct perf_evsel *evsel,
 			    struct perf_sample *sample,
 			    const char *backtrace)
 {
@@ -575,7 +580,8 @@ process_sample_sched_wakeup(struct perf_evsel *evsel,
 }
 
 static int
-process_sample_sched_switch(struct perf_evsel *evsel,
+process_sample_sched_switch(struct timechart *tchart __maybe_unused,
+			    struct perf_evsel *evsel,
 			    struct perf_sample *sample,
 			    const char *backtrace)
 {
@@ -590,7 +596,8 @@ process_sample_sched_switch(struct perf_evsel *evsel,
 
 #ifdef SUPPORT_OLD_POWER_EVENTS
 static int
-process_sample_power_start(struct perf_evsel *evsel,
+process_sample_power_start(struct timechart *tchart __maybe_unused,
+			   struct perf_evsel *evsel,
 			   struct perf_sample *sample,
 			   const char *backtrace __maybe_unused)
 {
@@ -602,7 +609,8 @@ process_sample_power_start(struct perf_evsel *evsel,
 }
 
 static int
-process_sample_power_end(struct perf_evsel *evsel __maybe_unused,
+process_sample_power_end(struct timechart *tchart __maybe_unused,
+			 struct perf_evsel *evsel __maybe_unused,
 			 struct perf_sample *sample,
 			 const char *backtrace __maybe_unused)
 {
@@ -611,14 +619,15 @@ process_sample_power_end(struct perf_evsel *evsel __maybe_unused,
 }
 
 static int
-process_sample_power_frequency(struct perf_evsel *evsel,
+process_sample_power_frequency(struct timechart *tchart,
+			       struct perf_evsel *evsel,
 			       struct perf_sample *sample,
 			       const char *backtrace __maybe_unused)
 {
 	u64 cpu_id = perf_evsel__intval(evsel, sample, "cpu_id");
 	u64 value = perf_evsel__intval(evsel, sample, "value");
 
-	p_state_change(cpu_id, sample->time, value);
+	p_state_change(tchart, cpu_id, sample->time, value);
 	return 0;
 }
 #endif /* SUPPORT_OLD_POWER_EVENTS */
@@ -627,12 +636,12 @@ process_sample_power_frequency(struct perf_evsel *evsel,
  * After the last sample we need to wrap up the current C/P state
  * and close out each CPU for these.
  */
-static void end_sample_processing(void)
+static void end_sample_processing(struct timechart *tchart)
 {
 	u64 cpu;
 	struct power_event *pwr;
 
-	for (cpu = 0; cpu <= numcpus; cpu++) {
+	for (cpu = 0; cpu <= tchart->numcpus; cpu++) {
 		/* C state */
 #if 0
 		pwr = zalloc(sizeof(*pwr));
@@ -641,7 +650,7 @@ static void end_sample_processing(void)
 
 		pwr->state = cpus_cstate_state[cpu];
 		pwr->start_time = cpus_cstate_start_times[cpu];
-		pwr->end_time = last_time;
+		pwr->end_time = tchart->last_time;
 		pwr->cpu = cpu;
 		pwr->type = CSTATE;
 		pwr->next = power_events;
@@ -656,15 +665,15 @@ static void end_sample_processing(void)
 
 		pwr->state = cpus_pstate_state[cpu];
 		pwr->start_time = cpus_pstate_start_times[cpu];
-		pwr->end_time = last_time;
+		pwr->end_time = tchart->last_time;
 		pwr->cpu = cpu;
 		pwr->type = PSTATE;
 		pwr->next = power_events;
 
 		if (!pwr->start_time)
-			pwr->start_time = first_time;
+			pwr->start_time = tchart->first_time;
 		if (!pwr->state)
-			pwr->state = min_freq;
+			pwr->state = tchart->min_freq;
 		power_events = pwr;
 	}
 }
@@ -718,7 +727,7 @@ static void sort_pids(void)
 }
 
 
-static void draw_c_p_states(void)
+static void draw_c_p_states(struct timechart *tchart)
 {
 	struct power_event *pwr;
 	pwr = power_events;
@@ -736,7 +745,7 @@ static void draw_c_p_states(void)
 	while (pwr) {
 		if (pwr->type == PSTATE) {
 			if (!pwr->state)
-				pwr->state = min_freq;
+				pwr->state = tchart->min_freq;
 			svg_pstate(pwr->cpu, pwr->start_time, pwr->end_time, pwr->state);
 		}
 		pwr = pwr->next;
@@ -833,14 +842,14 @@ static void draw_cpu_usage(void)
 	}
 }
 
-static void draw_process_bars(void)
+static void draw_process_bars(struct timechart *tchart)
 {
 	struct per_pid *p;
 	struct per_pidcomm *c;
 	struct cpu_sample *sample;
 	int Y = 0;
 
-	Y = 2 * numcpus + 2;
+	Y = 2 * tchart->numcpus + 2;
 
 	p = all_data;
 	while (p) {
@@ -922,7 +931,7 @@ static int passes_filter(struct per_pid *p, struct per_pidcomm *c)
 	return 0;
 }
 
-static int determine_display_tasks_filtered(void)
+static int determine_display_tasks_filtered(struct timechart *tchart)
 {
 	struct per_pid *p;
 	struct per_pidcomm *c;
@@ -932,11 +941,11 @@ static int determine_display_tasks_filtered(void)
 	while (p) {
 		p->display = 0;
 		if (p->start_time == 1)
-			p->start_time = first_time;
+			p->start_time = tchart->first_time;
 
 		/* no exit marker, task kept running to the end */
 		if (p->end_time == 0)
-			p->end_time = last_time;
+			p->end_time = tchart->last_time;
 
 		c = p->all;
 
@@ -944,7 +953,7 @@ static int determine_display_tasks_filtered(void)
 			c->display = 0;
 
 			if (c->start_time == 1)
-				c->start_time = first_time;
+				c->start_time = tchart->first_time;
 
 			if (passes_filter(p, c)) {
 				c->display = 1;
@@ -953,7 +962,7 @@ static int determine_display_tasks_filtered(void)
 			}
 
 			if (c->end_time == 0)
-				c->end_time = last_time;
+				c->end_time = tchart->last_time;
 
 			c = c->next;
 		}
@@ -962,24 +971,24 @@ static int determine_display_tasks_filtered(void)
 	return count;
 }
 
-static int determine_display_tasks(u64 threshold)
+static int determine_display_tasks(struct timechart *tchart, u64 threshold)
 {
 	struct per_pid *p;
 	struct per_pidcomm *c;
 	int count = 0;
 
 	if (process_filter)
-		return determine_display_tasks_filtered();
+		return determine_display_tasks_filtered(tchart);
 
 	p = all_data;
 	while (p) {
 		p->display = 0;
 		if (p->start_time == 1)
-			p->start_time = first_time;
+			p->start_time = tchart->first_time;
 
 		/* no exit marker, task kept running to the end */
 		if (p->end_time == 0)
-			p->end_time = last_time;
+			p->end_time = tchart->last_time;
 		if (p->total_time >= threshold)
 			p->display = 1;
 
@@ -989,7 +998,7 @@ static int determine_display_tasks(u64 threshold)
 			c->display = 0;
 
 			if (c->start_time == 1)
-				c->start_time = first_time;
+				c->start_time = tchart->first_time;
 
 			if (c->total_time >= threshold) {
 				c->display = 1;
@@ -997,7 +1006,7 @@ static int determine_display_tasks(u64 threshold)
 			}
 
 			if (c->end_time == 0)
-				c->end_time = last_time;
+				c->end_time = tchart->last_time;
 
 			c = c->next;
 		}
@@ -1010,52 +1019,45 @@ static int determine_display_tasks(u64 threshold)
 
 #define TIME_THRESH 10000000
 
-static void write_svg_file(const char *filename)
+static void write_svg_file(struct timechart *tchart, const char *filename)
 {
 	u64 i;
 	int count;
 	int thresh = TIME_THRESH;
 
-	numcpus++;
+	tchart->numcpus++;
 
-	if (power_only)
-		proc_num = 0;
+	if (tchart->power_only)
+		tchart->proc_num = 0;
 
 	/* We'd like to show at least proc_num tasks;
 	 * be less picky if we have fewer */
 	do {
-		count = determine_display_tasks(thresh);
+		count = determine_display_tasks(tchart, thresh);
 		thresh /= 10;
-	} while (!process_filter && thresh && count < proc_num);
+	} while (!process_filter && thresh && count < tchart->proc_num);
 
-	open_svg(filename, numcpus, count, first_time, last_time);
+	open_svg(filename, tchart->numcpus, count, tchart->first_time, tchart->last_time);
 
 	svg_time_grid();
 	svg_legenda();
 
-	for (i = 0; i < numcpus; i++)
-		svg_cpu_box(i, max_freq, turbo_frequency);
+	for (i = 0; i < tchart->numcpus; i++)
+		svg_cpu_box(i, tchart->max_freq, tchart->turbo_frequency);
 
 	draw_cpu_usage();
-	if (proc_num)
-		draw_process_bars();
-	if (!tasks_only)
-		draw_c_p_states();
-	if (proc_num)
+	if (tchart->proc_num)
+		draw_process_bars(tchart);
+	if (!tchart->tasks_only)
+		draw_c_p_states(tchart);
+	if (tchart->proc_num)
 		draw_wakeups();
 
 	svg_close();
 }
 
-static int __cmd_timechart(const char *output_name)
+static int __cmd_timechart(struct timechart *tchart, const char *output_name)
 {
-	struct perf_tool perf_timechart = {
-		.comm		 = process_comm_event,
-		.fork		 = process_fork_event,
-		.exit		 = process_exit_event,
-		.sample		 = process_sample_event,
-		.ordered_samples = true,
-	};
 	const struct perf_evsel_str_handler power_tracepoints[] = {
 		{ "power:cpu_idle",		process_sample_cpu_idle },
 		{ "power:cpu_frequency",	process_sample_cpu_frequency },
@@ -1073,7 +1075,7 @@ static int __cmd_timechart(const char *output_name)
 	};
 
 	struct perf_session *session = perf_session__new(&file, false,
-							 &perf_timechart);
+							 &tchart->tool);
 	int ret = -EINVAL;
 
 	if (session == NULL)
@@ -1088,24 +1090,24 @@ static int __cmd_timechart(const char *output_name)
 		goto out_delete;
 	}
 
-	ret = perf_session__process_events(session, &perf_timechart);
+	ret = perf_session__process_events(session, &tchart->tool);
 	if (ret)
 		goto out_delete;
 
-	end_sample_processing();
+	end_sample_processing(tchart);
 
 	sort_pids();
 
-	write_svg_file(output_name);
+	write_svg_file(tchart, output_name);
 
 	pr_info("Written %2.1f seconds of trace to %s.\n",
-		(last_time - first_time) / 1000000000.0, output_name);
+		(tchart->last_time - tchart->first_time) / 1000000000.0, output_name);
 out_delete:
 	perf_session__delete(session);
 	return ret;
 }
 
-static int __cmd_record(int argc, const char **argv)
+static int timechart__record(struct timechart *tchart, int argc, const char **argv)
 {
 	unsigned int rec_argc, i, j;
 	const char **rec_argv;
@@ -1153,15 +1155,15 @@ static int __cmd_record(int argc, const char **argv)
 	}
 #endif
 
-	if (power_only)
+	if (tchart->power_only)
 		tasks_args_nr = 0;
 
-	if (tasks_only) {
+	if (tchart->tasks_only) {
 		power_args_nr = 0;
 		old_power_args_nr = 0;
 	}
 
-	if (!with_backtrace)
+	if (!tchart->with_backtrace)
 		backtrace_args_no = 0;
 
 	record_elems = common_args_nr + tasks_args_nr +
@@ -1207,20 +1209,30 @@ parse_process(const struct option *opt __maybe_unused, const char *arg,
 int cmd_timechart(int argc, const char **argv,
 		  const char *prefix __maybe_unused)
 {
+	struct timechart tchart = {
+		.tool = {
+			.comm		 = process_comm_event,
+			.fork		 = process_fork_event,
+			.exit		 = process_exit_event,
+			.sample		 = process_sample_event,
+			.ordered_samples = true,
+		},
+		.proc_num = 15,
+	};
 	const char *output_name = "output.svg";
 	const struct option timechart_options[] = {
 	OPT_STRING('i', "input", &input_name, "file", "input file name"),
 	OPT_STRING('o', "output", &output_name, "file", "output file name"),
 	OPT_INTEGER('w', "width", &svg_page_width, "page width"),
-	OPT_BOOLEAN('P', "power-only", &power_only, "output power data only"),
-	OPT_BOOLEAN('T', "tasks-only", &tasks_only,
+	OPT_BOOLEAN('P', "power-only", &tchart.power_only, "output power data only"),
+	OPT_BOOLEAN('T', "tasks-only", &tchart.tasks_only,
 		    "output processes data only"),
 	OPT_CALLBACK('p', "process", NULL, "process",
 		      "process selector. Pass a pid or process name.",
 		       parse_process),
 	OPT_STRING(0, "symfs", &symbol_conf.symfs, "directory",
 		    "Look for files with symbols relative to this directory"),
-	OPT_INTEGER('n', "proc-num", &proc_num,
+	OPT_INTEGER('n', "proc-num", &tchart.proc_num,
 		    "min. number of tasks to print"),
 	OPT_END()
 	};
@@ -1230,10 +1242,10 @@ int cmd_timechart(int argc, const char **argv,
 	};
 
 	const struct option record_options[] = {
-	OPT_BOOLEAN('P', "power-only", &power_only, "output power data only"),
-	OPT_BOOLEAN('T', "tasks-only", &tasks_only,
+	OPT_BOOLEAN('P', "power-only", &tchart.power_only, "output power data only"),
+	OPT_BOOLEAN('T', "tasks-only", &tchart.tasks_only,
 		    "output processes data only"),
-	OPT_BOOLEAN('g', "callchain", &with_backtrace, "record callchain"),
+	OPT_BOOLEAN('g', "callchain", &tchart.with_backtrace, "record callchain"),
 	OPT_END()
 	};
 	const char * const record_usage[] = {
@@ -1243,7 +1255,7 @@ int cmd_timechart(int argc, const char **argv,
 	argc = parse_options(argc, argv, timechart_options, timechart_usage,
 			PARSE_OPT_STOP_AT_NON_OPTION);
 
-	if (power_only && tasks_only) {
+	if (tchart.power_only && tchart.tasks_only) {
 		pr_err("-P and -T options cannot be used at the same time.\n");
 		return -1;
 	}
@@ -1254,16 +1266,16 @@ int cmd_timechart(int argc, const char **argv,
 		argc = parse_options(argc, argv, record_options, record_usage,
 				     PARSE_OPT_STOP_AT_NON_OPTION);
 
-		if (power_only && tasks_only) {
+		if (tchart.power_only && tchart.tasks_only) {
 			pr_err("-P and -T options cannot be used at the same time.\n");
 			return -1;
 		}
 
-		return __cmd_record(argc, argv);
+		return timechart__record(&tchart, argc, argv);
 	} else if (argc)
 		usage_with_options(timechart_usage, timechart_options);
 
 	setup_pager();
 
-	return __cmd_timechart(output_name);
+	return __cmd_timechart(&tchart, output_name);
 }
-- 
1.8.1.4

[PATCH 07/12] perf timechart: Move all_data per_pid list to 'struct timechart'

[Arnaldo Carvalho de Melo]

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

Removing another global variable.

This one tho would be better done by using the machine infrastructure,
searching for the 'struct thread' with a pid, then using thread->priv,
etc.

TODO list material for now.

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stanislav Fomichev <stfomichev@yandex-team.ru>
Cc: Stephane Eranian <eranian@google.com>
Link: http://lkml.kernel.org/n/tip-yyfpudgjvr6mev4bue9u72a2@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/builtin-timechart.c | 113 +++++++++++++++++++++--------------------
 1 file changed, 59 insertions(+), 54 deletions(-)

diff --git a/tools/perf/builtin-timechart.c b/tools/perf/builtin-timechart.c
index e2d62f1a96e4..0c955acc08a2 100644
--- a/tools/perf/builtin-timechart.c
+++ b/tools/perf/builtin-timechart.c
@@ -41,8 +41,11 @@
 #define SUPPORT_OLD_POWER_EVENTS 1
 #define PWR_EVENT_EXIT -1
 
+struct per_pid;
+
 struct timechart {
 	struct perf_tool	tool;
+	struct per_pid		*all_data;
 	int			proc_num;
 	unsigned int		numcpus;
 	u64			min_freq,	/* Lowest CPU frequency seen */
@@ -123,8 +126,6 @@ struct cpu_sample {
 	const char *backtrace;
 };
 
-static struct per_pid *all_data;
-
 #define CSTATE 1
 #define PSTATE 2
 
@@ -157,9 +158,9 @@ struct process_filter {
 static struct process_filter *process_filter;
 
 
-static struct per_pid *find_create_pid(int pid)
+static struct per_pid *find_create_pid(struct timechart *tchart, int pid)
 {
-	struct per_pid *cursor = all_data;
+	struct per_pid *cursor = tchart->all_data;
 
 	while (cursor) {
 		if (cursor->pid == pid)
@@ -169,16 +170,16 @@ static struct per_pid *find_create_pid(int pid)
 	cursor = zalloc(sizeof(*cursor));
 	assert(cursor != NULL);
 	cursor->pid = pid;
-	cursor->next = all_data;
-	all_data = cursor;
+	cursor->next = tchart->all_data;
+	tchart->all_data = cursor;
 	return cursor;
 }
 
-static void pid_set_comm(int pid, char *comm)
+static void pid_set_comm(struct timechart *tchart, int pid, char *comm)
 {
 	struct per_pid *p;
 	struct per_pidcomm *c;
-	p = find_create_pid(pid);
+	p = find_create_pid(tchart, pid);
 	c = p->all;
 	while (c) {
 		if (c->comm && strcmp(c->comm, comm) == 0) {
@@ -200,14 +201,14 @@ static void pid_set_comm(int pid, char *comm)
 	p->all = c;
 }
 
-static void pid_fork(int pid, int ppid, u64 timestamp)
+static void pid_fork(struct timechart *tchart, int pid, int ppid, u64 timestamp)
 {
 	struct per_pid *p, *pp;
-	p = find_create_pid(pid);
-	pp = find_create_pid(ppid);
+	p = find_create_pid(tchart, pid);
+	pp = find_create_pid(tchart, ppid);
 	p->ppid = ppid;
 	if (pp->current && pp->current->comm && !p->current)
-		pid_set_comm(pid, pp->current->comm);
+		pid_set_comm(tchart, pid, pp->current->comm);
 
 	p->start_time = timestamp;
 	if (p->current) {
@@ -216,24 +217,24 @@ static void pid_fork(int pid, int ppid, u64 timestamp)
 	}
 }
 
-static void pid_exit(int pid, u64 timestamp)
+static void pid_exit(struct timechart *tchart, int pid, u64 timestamp)
 {
 	struct per_pid *p;
-	p = find_create_pid(pid);
+	p = find_create_pid(tchart, pid);
 	p->end_time = timestamp;
 	if (p->current)
 		p->current->end_time = timestamp;
 }
 
-static void
-pid_put_sample(int pid, int type, unsigned int cpu, u64 start, u64 end,
-	       const char *backtrace)
+static void pid_put_sample(struct timechart *tchart, int pid, int type,
+			   unsigned int cpu, u64 start, u64 end,
+			   const char *backtrace)
 {
 	struct per_pid *p;
 	struct per_pidcomm *c;
 	struct cpu_sample *sample;
 
-	p = find_create_pid(pid);
+	p = find_create_pid(tchart, pid);
 	c = p->current;
 	if (!c) {
 		c = zalloc(sizeof(*c));
@@ -271,30 +272,33 @@ static int cpus_cstate_state[MAX_CPUS];
 static u64 cpus_pstate_start_times[MAX_CPUS];
 static u64 cpus_pstate_state[MAX_CPUS];
 
-static int process_comm_event(struct perf_tool *tool __maybe_unused,
+static int process_comm_event(struct perf_tool *tool,
 			      union perf_event *event,
 			      struct perf_sample *sample __maybe_unused,
 			      struct machine *machine __maybe_unused)
 {
-	pid_set_comm(event->comm.tid, event->comm.comm);
+	struct timechart *tchart = container_of(tool, struct timechart, tool);
+	pid_set_comm(tchart, event->comm.tid, event->comm.comm);
 	return 0;
 }
 
-static int process_fork_event(struct perf_tool *tool __maybe_unused,
+static int process_fork_event(struct perf_tool *tool,
 			      union perf_event *event,
 			      struct perf_sample *sample __maybe_unused,
 			      struct machine *machine __maybe_unused)
 {
-	pid_fork(event->fork.pid, event->fork.ppid, event->fork.time);
+	struct timechart *tchart = container_of(tool, struct timechart, tool);
+	pid_fork(tchart, event->fork.pid, event->fork.ppid, event->fork.time);
 	return 0;
 }
 
-static int process_exit_event(struct perf_tool *tool __maybe_unused,
+static int process_exit_event(struct perf_tool *tool,
 			      union perf_event *event,
 			      struct perf_sample *sample __maybe_unused,
 			      struct machine *machine __maybe_unused)
 {
-	pid_exit(event->fork.pid, event->fork.time);
+	struct timechart *tchart = container_of(tool, struct timechart, tool);
+	pid_exit(tchart, event->fork.pid, event->fork.time);
 	return 0;
 }
 
@@ -361,8 +365,8 @@ static void p_state_change(struct timechart *tchart, int cpu, u64 timestamp, u64
 		tchart->turbo_frequency = tchart->max_freq;
 }
 
-static void sched_wakeup(int cpu, u64 timestamp, int waker, int wakee,
-			 u8 flags, const char *backtrace)
+static void sched_wakeup(struct timechart *tchart, int cpu, u64 timestamp,
+			 int waker, int wakee, u8 flags, const char *backtrace)
 {
 	struct per_pid *p;
 	struct wake_event *we = zalloc(sizeof(*we));
@@ -380,36 +384,37 @@ static void sched_wakeup(int cpu, u64 timestamp, int waker, int wakee,
 	we->wakee = wakee;
 	we->next = wake_events;
 	wake_events = we;
-	p = find_create_pid(we->wakee);
+	p = find_create_pid(tchart, we->wakee);
 
 	if (p && p->current && p->current->state == TYPE_NONE) {
 		p->current->state_since = timestamp;
 		p->current->state = TYPE_WAITING;
 	}
 	if (p && p->current && p->current->state == TYPE_BLOCKED) {
-		pid_put_sample(p->pid, p->current->state, cpu,
+		pid_put_sample(tchart, p->pid, p->current->state, cpu,
 			       p->current->state_since, timestamp, NULL);
 		p->current->state_since = timestamp;
 		p->current->state = TYPE_WAITING;
 	}
 }
 
-static void sched_switch(int cpu, u64 timestamp, int prev_pid, int next_pid,
-			 u64 prev_state, const char *backtrace)
+static void sched_switch(struct timechart *tchart, int cpu, u64 timestamp,
+			 int prev_pid, int next_pid, u64 prev_state,
+			 const char *backtrace)
 {
 	struct per_pid *p = NULL, *prev_p;
 
-	prev_p = find_create_pid(prev_pid);
+	prev_p = find_create_pid(tchart, prev_pid);
 
-	p = find_create_pid(next_pid);
+	p = find_create_pid(tchart, next_pid);
 
 	if (prev_p->current && prev_p->current->state != TYPE_NONE)
-		pid_put_sample(prev_pid, TYPE_RUNNING, cpu,
+		pid_put_sample(tchart, prev_pid, TYPE_RUNNING, cpu,
 			       prev_p->current->state_since, timestamp,
 			       backtrace);
 	if (p && p->current) {
 		if (p->current->state != TYPE_NONE)
-			pid_put_sample(next_pid, p->current->state, cpu,
+			pid_put_sample(tchart, next_pid, p->current->state, cpu,
 				       p->current->state_since, timestamp,
 				       backtrace);
 
@@ -566,7 +571,7 @@ process_sample_cpu_frequency(struct timechart *tchart,
 }
 
 static int
-process_sample_sched_wakeup(struct timechart *tchart __maybe_unused,
+process_sample_sched_wakeup(struct timechart *tchart,
 			    struct perf_evsel *evsel,
 			    struct perf_sample *sample,
 			    const char *backtrace)
@@ -575,12 +580,12 @@ process_sample_sched_wakeup(struct timechart *tchart __maybe_unused,
 	int waker = perf_evsel__intval(evsel, sample, "common_pid");
 	int wakee = perf_evsel__intval(evsel, sample, "pid");
 
-	sched_wakeup(sample->cpu, sample->time, waker, wakee, flags, backtrace);
+	sched_wakeup(tchart, sample->cpu, sample->time, waker, wakee, flags, backtrace);
 	return 0;
 }
 
 static int
-process_sample_sched_switch(struct timechart *tchart __maybe_unused,
+process_sample_sched_switch(struct timechart *tchart,
 			    struct perf_evsel *evsel,
 			    struct perf_sample *sample,
 			    const char *backtrace)
@@ -589,8 +594,8 @@ process_sample_sched_switch(struct timechart *tchart __maybe_unused,
 	int next_pid = perf_evsel__intval(evsel, sample, "next_pid");
 	u64 prev_state = perf_evsel__intval(evsel, sample, "prev_state");
 
-	sched_switch(sample->cpu, sample->time, prev_pid, next_pid, prev_state,
-		     backtrace);
+	sched_switch(tchart, sample->cpu, sample->time, prev_pid, next_pid,
+		     prev_state, backtrace);
 	return 0;
 }
 
@@ -681,16 +686,16 @@ static void end_sample_processing(struct timechart *tchart)
 /*
  * Sort the pid datastructure
  */
-static void sort_pids(void)
+static void sort_pids(struct timechart *tchart)
 {
 	struct per_pid *new_list, *p, *cursor, *prev;
 	/* sort by ppid first, then by pid, lowest to highest */
 
 	new_list = NULL;
 
-	while (all_data) {
-		p = all_data;
-		all_data = p->next;
+	while (tchart->all_data) {
+		p = tchart->all_data;
+		tchart->all_data = p->next;
 		p->next = NULL;
 
 		if (new_list == NULL) {
@@ -723,7 +728,7 @@ static void sort_pids(void)
 				prev->next = p;
 		}
 	}
-	all_data = new_list;
+	tchart->all_data = new_list;
 }
 
 
@@ -752,7 +757,7 @@ static void draw_c_p_states(struct timechart *tchart)
 	}
 }
 
-static void draw_wakeups(void)
+static void draw_wakeups(struct timechart *tchart)
 {
 	struct wake_event *we;
 	struct per_pid *p;
@@ -764,7 +769,7 @@ static void draw_wakeups(void)
 		char *task_from = NULL, *task_to = NULL;
 
 		/* locate the column of the waker and wakee */
-		p = all_data;
+		p = tchart->all_data;
 		while (p) {
 			if (p->pid == we->waker || p->pid == we->wakee) {
 				c = p->all;
@@ -820,12 +825,12 @@ static void draw_wakeups(void)
 	}
 }
 
-static void draw_cpu_usage(void)
+static void draw_cpu_usage(struct timechart *tchart)
 {
 	struct per_pid *p;
 	struct per_pidcomm *c;
 	struct cpu_sample *sample;
-	p = all_data;
+	p = tchart->all_data;
 	while (p) {
 		c = p->all;
 		while (c) {
@@ -851,7 +856,7 @@ static void draw_process_bars(struct timechart *tchart)
 
 	Y = 2 * tchart->numcpus + 2;
 
-	p = all_data;
+	p = tchart->all_data;
 	while (p) {
 		c = p->all;
 		while (c) {
@@ -937,7 +942,7 @@ static int determine_display_tasks_filtered(struct timechart *tchart)
 	struct per_pidcomm *c;
 	int count = 0;
 
-	p = all_data;
+	p = tchart->all_data;
 	while (p) {
 		p->display = 0;
 		if (p->start_time == 1)
@@ -980,7 +985,7 @@ static int determine_display_tasks(struct timechart *tchart, u64 threshold)
 	if (process_filter)
 		return determine_display_tasks_filtered(tchart);
 
-	p = all_data;
+	p = tchart->all_data;
 	while (p) {
 		p->display = 0;
 		if (p->start_time == 1)
@@ -1045,13 +1050,13 @@ static void write_svg_file(struct timechart *tchart, const char *filename)
 	for (i = 0; i < tchart->numcpus; i++)
 		svg_cpu_box(i, tchart->max_freq, tchart->turbo_frequency);
 
-	draw_cpu_usage();
+	draw_cpu_usage(tchart);
 	if (tchart->proc_num)
 		draw_process_bars(tchart);
 	if (!tchart->tasks_only)
 		draw_c_p_states(tchart);
 	if (tchart->proc_num)
-		draw_wakeups();
+		draw_wakeups(tchart);
 
 	svg_close();
 }
@@ -1096,7 +1101,7 @@ static int __cmd_timechart(struct timechart *tchart, const char *output_name)
 
 	end_sample_processing(tchart);
 
-	sort_pids();
+	sort_pids(tchart);
 
 	write_svg_file(tchart, output_name);
 
-- 
1.8.1.4

[PATCH 08/12] perf timechart: Move power_events list to 'struct timechart'

[Arnaldo Carvalho de Melo]

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

Removing another global variable.

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stanislav Fomichev <stfomichev@yandex-team.ru>
Cc: Stephane Eranian <eranian@google.com>
Link: http://lkml.kernel.org/n/tip-2akef3p9caau56itf5mugd2b@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/builtin-timechart.c | 31 ++++++++++++++++---------------
 1 file changed, 16 insertions(+), 15 deletions(-)

diff --git a/tools/perf/builtin-timechart.c b/tools/perf/builtin-timechart.c
index 0c955acc08a2..8ee0ff1777a8 100644
--- a/tools/perf/builtin-timechart.c
+++ b/tools/perf/builtin-timechart.c
@@ -42,10 +42,12 @@
 #define PWR_EVENT_EXIT -1
 
 struct per_pid;
+struct power_event;
 
 struct timechart {
 	struct perf_tool	tool;
 	struct per_pid		*all_data;
+	struct power_event	*power_events;
 	int			proc_num;
 	unsigned int		numcpus;
 	u64			min_freq,	/* Lowest CPU frequency seen */
@@ -146,7 +148,6 @@ struct wake_event {
 	const char *backtrace;
 };
 
-static struct power_event    *power_events;
 static struct wake_event     *wake_events;
 
 struct process_filter {
@@ -312,7 +313,7 @@ static void c_state_start(int cpu, u64 timestamp, int state)
 	cpus_cstate_state[cpu] = state;
 }
 
-static void c_state_end(int cpu, u64 timestamp)
+static void c_state_end(struct timechart *tchart, int cpu, u64 timestamp)
 {
 	struct power_event *pwr = zalloc(sizeof(*pwr));
 
@@ -324,9 +325,9 @@ static void c_state_end(int cpu, u64 timestamp)
 	pwr->end_time = timestamp;
 	pwr->cpu = cpu;
 	pwr->type = CSTATE;
-	pwr->next = power_events;
+	pwr->next = tchart->power_events;
 
-	power_events = pwr;
+	tchart->power_events = pwr;
 }
 
 static void p_state_change(struct timechart *tchart, int cpu, u64 timestamp, u64 new_freq)
@@ -345,12 +346,12 @@ static void p_state_change(struct timechart *tchart, int cpu, u64 timestamp, u64
 	pwr->end_time = timestamp;
 	pwr->cpu = cpu;
 	pwr->type = PSTATE;
-	pwr->next = power_events;
+	pwr->next = tchart->power_events;
 
 	if (!pwr->start_time)
 		pwr->start_time = tchart->first_time;
 
-	power_events = pwr;
+	tchart->power_events = pwr;
 
 	cpus_pstate_state[cpu] = new_freq;
 	cpus_pstate_start_times[cpu] = timestamp;
@@ -551,7 +552,7 @@ process_sample_cpu_idle(struct timechart *tchart __maybe_unused,
 	u32 cpu_id = perf_evsel__intval(evsel, sample, "cpu_id");
 
 	if (state == (u32)PWR_EVENT_EXIT)
-		c_state_end(cpu_id, sample->time);
+		c_state_end(tchart, cpu_id, sample->time);
 	else
 		c_state_start(cpu_id, sample->time, state);
 	return 0;
@@ -614,12 +615,12 @@ process_sample_power_start(struct timechart *tchart __maybe_unused,
 }
 
 static int
-process_sample_power_end(struct timechart *tchart __maybe_unused,
+process_sample_power_end(struct timechart *tchart,
 			 struct perf_evsel *evsel __maybe_unused,
 			 struct perf_sample *sample,
 			 const char *backtrace __maybe_unused)
 {
-	c_state_end(sample->cpu, sample->time);
+	c_state_end(tchart, sample->cpu, sample->time);
 	return 0;
 }
 
@@ -658,9 +659,9 @@ static void end_sample_processing(struct timechart *tchart)
 		pwr->end_time = tchart->last_time;
 		pwr->cpu = cpu;
 		pwr->type = CSTATE;
-		pwr->next = power_events;
+		pwr->next = tchart->power_events;
 
-		power_events = pwr;
+		tchart->power_events = pwr;
 #endif
 		/* P state */
 
@@ -673,13 +674,13 @@ static void end_sample_processing(struct timechart *tchart)
 		pwr->end_time = tchart->last_time;
 		pwr->cpu = cpu;
 		pwr->type = PSTATE;
-		pwr->next = power_events;
+		pwr->next = tchart->power_events;
 
 		if (!pwr->start_time)
 			pwr->start_time = tchart->first_time;
 		if (!pwr->state)
 			pwr->state = tchart->min_freq;
-		power_events = pwr;
+		tchart->power_events = pwr;
 	}
 }
 
@@ -735,7 +736,7 @@ static void sort_pids(struct timechart *tchart)
 static void draw_c_p_states(struct timechart *tchart)
 {
 	struct power_event *pwr;
-	pwr = power_events;
+	pwr = tchart->power_events;
 
 	/*
 	 * two pass drawing so that the P state bars are on top of the C state blocks
@@ -746,7 +747,7 @@ static void draw_c_p_states(struct timechart *tchart)
 		pwr = pwr->next;
 	}
 
-	pwr = power_events;
+	pwr = tchart->power_events;
 	while (pwr) {
 		if (pwr->type == PSTATE) {
 			if (!pwr->state)
-- 
1.8.1.4

[PATCH 09/12] perf timechart: Move wake_events list to 'struct timechart'

[Arnaldo Carvalho de Melo]

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

Removing another global variable.

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stanislav Fomichev <stfomichev@yandex-team.ru>
Cc: Stephane Eranian <eranian@google.com>
Link: http://lkml.kernel.org/n/tip-14rpuci11l2s0o01yta87kxe@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/builtin-timechart.c | 10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)

diff --git a/tools/perf/builtin-timechart.c b/tools/perf/builtin-timechart.c
index 8ee0ff1777a8..0bda620a717d 100644
--- a/tools/perf/builtin-timechart.c
+++ b/tools/perf/builtin-timechart.c
@@ -43,11 +43,13 @@
 
 struct per_pid;
 struct power_event;
+struct wake_event;
 
 struct timechart {
 	struct perf_tool	tool;
 	struct per_pid		*all_data;
 	struct power_event	*power_events;
+	struct wake_event	*wake_events;
 	int			proc_num;
 	unsigned int		numcpus;
 	u64			min_freq,	/* Lowest CPU frequency seen */
@@ -148,8 +150,6 @@ struct wake_event {
 	const char *backtrace;
 };
 
-static struct wake_event     *wake_events;
-
 struct process_filter {
 	char			*name;
 	int			pid;
@@ -383,8 +383,8 @@ static void sched_wakeup(struct timechart *tchart, int cpu, u64 timestamp,
 		we->waker = -1;
 
 	we->wakee = wakee;
-	we->next = wake_events;
-	wake_events = we;
+	we->next = tchart->wake_events;
+	tchart->wake_events = we;
 	p = find_create_pid(tchart, we->wakee);
 
 	if (p && p->current && p->current->state == TYPE_NONE) {
@@ -764,7 +764,7 @@ static void draw_wakeups(struct timechart *tchart)
 	struct per_pid *p;
 	struct per_pidcomm *c;
 
-	we = wake_events;
+	we = tchart->wake_events;
 	while (we) {
 		int from = 0, to = 0;
 		char *task_from = NULL, *task_to = NULL;
-- 
1.8.1.4

[PATCH 10/12] perf tools: Correct the message in feature-libnuma checking.

[Arnaldo Carvalho de Melo]

From: Dongsheng Yang <yangds.fnst@cn.fujitsu.com>

The package required for numa is named numactl-devel in Fedora or RHEL,
and libnuma-devel in OpenSuSE, and libnuma-dev in Ubuntu.

This patch corrects the package name in warning message in
feature-libnuma checking.

Signed-off-by: Dongsheng Yang <yangds.fnst@cn.fujitsu.com>
Acked-by: Peter Hurley <peter@hurleysoftware.com>
Cc: Peter Hurley <peter@hurleysoftware.com>
Link: http://lkml.kernel.org/r/1385998008-6851-1-git-send-email-yangds.fnst@cn.fujitsu.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/config/Makefile | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/tools/perf/config/Makefile b/tools/perf/config/Makefile
index f7d11a811c74..36e66ac40abc 100644
--- a/tools/perf/config/Makefile
+++ b/tools/perf/config/Makefile
@@ -533,7 +533,7 @@ endif
 
 ifndef NO_LIBNUMA
   ifeq ($(feature-libnuma), 0)
-    msg := $(warning No numa.h found, disables 'perf bench numa mem' benchmark, please install numa-libs-devel or libnuma-dev);
+    msg := $(warning No numa.h found, disables 'perf bench numa mem' benchmark, please install numactl-devel/libnuma-devel/libnuma-dev);
     NO_LIBNUMA := 1
   else
     CFLAGS += -DHAVE_LIBNUMA_SUPPORT
-- 
1.8.1.4

[PATCH 11/12] perf tools: Include test-stackprotector-all.c in test-all

[Arnaldo Carvalho de Melo]

From: Jiri Olsa <jolsa@redhat.com>

We're missing test-stackprotector-all.c in test-all check, adding it.

Signed-off-by: Jiri Olsa <jolsa@redhat.com>
Cc: Corey Ashford <cjashfor@linux.vnet.ibm.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Steven Rostedt <rostedt@goodmis.org>
Link: http://lkml.kernel.org/r/1385638408-23519-3-git-send-email-jolsa@redhat.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/config/feature-checks/test-all.c | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/tools/perf/config/feature-checks/test-all.c b/tools/perf/config/feature-checks/test-all.c
index 59e7a705e146..9b8a544155bb 100644
--- a/tools/perf/config/feature-checks/test-all.c
+++ b/tools/perf/config/feature-checks/test-all.c
@@ -85,6 +85,10 @@
 # include "test-timerfd.c"
 #undef main
 
+#define main main_test_stackprotector_all
+# include "test-stackprotector-all.c"
+#undef main
+
 int main(int argc, char *argv[])
 {
 	main_test_libpython();
@@ -106,6 +110,7 @@ int main(int argc, char *argv[])
 	main_test_backtrace();
 	main_test_libnuma();
 	main_test_timerfd();
+	main_test_stackprotector_all();
 
 	return 0;
 }
-- 
1.8.1.4

[PATCH 12/12] perf trace: Honour -m option

[Arnaldo Carvalho de Melo]

From: Jiri Olsa <jolsa@redhat.com>

Currently trace command supports '-m' option, but does not honours its
value and keeps the default.

Changing the perf_evlist__mmap function call to use the '-m' configured
value.

Signed-off-by: Jiri Olsa <jolsa@redhat.com>
Acked-by: David Ahern <dsahern@gmail.com>
Cc: Corey Ashford <cjashfor@linux.vnet.ibm.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1385657842-8914-1-git-send-email-jolsa@redhat.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/builtin-trace.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/tools/perf/builtin-trace.c b/tools/perf/builtin-trace.c
index e9f345e2551a..9f2a242fa79c 100644
--- a/tools/perf/builtin-trace.c
+++ b/tools/perf/builtin-trace.c
@@ -1890,7 +1890,7 @@ static int trace__run(struct trace *trace, int argc, const char **argv)
 	if (err < 0)
 		goto out_error_open;
 
-	err = perf_evlist__mmap(evlist, UINT_MAX, false);
+	err = perf_evlist__mmap(evlist, trace->opts.mmap_pages, false);
 	if (err < 0) {
 		fprintf(trace->output, "Couldn't mmap the events: %s\n", strerror(errno));
 		goto out_close_evlist;
-- 
1.8.1.4

Re: [GIT PULL 00/12] perf/core improvements and fixes

[Ingo Molnar]

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

> From: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
> 
> Hi Ingo,
> 
> 	Please consider pulling,
> 
> - Arnaldo
> 
> The following changes since commit 0ed1e0bee0b2c6b4cc6d7a63787739a9d3ac8aa8:
> 
>   Merge tag 'perf-core-for-mingo' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux into perf/core (2013-11-30 13:42:48 +0100)
> 
> 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 f885037ef6400ce4a4c122a88845dea2c9bca256:
> 
>   perf trace: Honour -m option (2013-12-02 16:29:40 -0300)
> 
> ----------------------------------------------------------------
> perf/tools improvements and fixes:
> 
> . Honour -m option in 'trace', the tool was offering the option to
>   set the mmap size, but wasn't using it when doing the actual mmap
>   on the events file descriptors, fix from Jiri Olsa.
> 
> . Correct the message in feature-libnuma checking, swowing the right
>   devel package names for various distros, from Dongsheng Yang.
> 
> . Polish 'readn' function and introduce its counterpart, 'writen', from
>   Jiri Olsa.
> 
> . Start moving timechart state from global variables to a 'perf_tool' derived
>   'timechart' struct.
> 
> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
> 
> ----------------------------------------------------------------
> Arnaldo Carvalho de Melo (4):
>       perf timechart: Introduce tool struct
>       perf timechart: Move all_data per_pid list to 'struct timechart'
>       perf timechart: Move power_events list to 'struct timechart'
>       perf timechart: Move wake_events list to 'struct timechart'
> 
> Dongsheng Yang (1):
>       perf tools: Correct the message in feature-libnuma checking.
> 
> Jiri Olsa (7):
>       perf record: Unify data output code into perf_record__write function
>       perf tools: Use correct return type for readn function
>       perf tools: Fine tune readn function
>       perf tools: Add writen function
>       perf tools: Add perf_data_file__write interface
>       perf tools: Include test-stackprotector-all.c in test-all
>       perf trace: Honour -m option
> 
>  tools/perf/builtin-record.c                 |  20 +-
>  tools/perf/builtin-timechart.c              | 364 +++++++++++++++-------------
>  tools/perf/builtin-trace.c                  |   2 +-
>  tools/perf/config/Makefile                  |   2 +-
>  tools/perf/config/feature-checks/test-all.c |   5 +
>  tools/perf/util/data.c                      |   6 +
>  tools/perf/util/data.h                      |  14 +-
>  tools/perf/util/header.c                    |  18 +-
>  tools/perf/util/session.c                   |   2 +-
>  tools/perf/util/util.c                      |  32 ++-
>  tools/perf/util/util.h                      |   3 +-
>  11 files changed, 256 insertions(+), 212 deletions(-)

Pulled, thanks a lot Arnaldo!

	Ingo