[PATCH 01/18] perf probe: Add union member access support

[PATCH 01/18] perf probe: Add union member access support

[Arnaldo Carvalho de Melo]

From: Hyeoncheol Lee <hyc.lee@gmail.com>

Union members can be accessed with '.' or '->' like data structure
member access

Signed-off-by: Hyunchul Lee <hyc.lee@gmail.com>
Acked-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
Link: http://lkml.kernel.org/r/CANFS6baeuSBxPGQ8SUZWZErJ2bWs-Nojg+FSo138E1QK8bJJig@mail.gmail.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/util/probe-finder.c |   24 ++++++++++++++++--------
 1 files changed, 16 insertions(+), 8 deletions(-)

diff --git a/tools/perf/util/probe-finder.c b/tools/perf/util/probe-finder.c
index 526ba56..1daf5c1 100644
--- a/tools/perf/util/probe-finder.c
+++ b/tools/perf/util/probe-finder.c
@@ -525,8 +525,10 @@ static int convert_variable_fields(Dwarf_Die *vr_die, const char *varname,
 			return -ENOENT;
 		}
 		/* Verify it is a data structure  */
-		if (dwarf_tag(&type) != DW_TAG_structure_type) {
-			pr_warning("%s is not a data structure.\n", varname);
+		tag = dwarf_tag(&type);
+		if (tag != DW_TAG_structure_type && tag != DW_TAG_union_type) {
+			pr_warning("%s is not a data structure nor an union.\n",
+				   varname);
 			return -EINVAL;
 		}
 
@@ -539,8 +541,9 @@ static int convert_variable_fields(Dwarf_Die *vr_die, const char *varname,
 			*ref_ptr = ref;
 	} else {
 		/* Verify it is a data structure  */
-		if (tag != DW_TAG_structure_type) {
-			pr_warning("%s is not a data structure.\n", varname);
+		if (tag != DW_TAG_structure_type && tag != DW_TAG_union_type) {
+			pr_warning("%s is not a data structure nor an union.\n",
+				   varname);
 			return -EINVAL;
 		}
 		if (field->name[0] == '[') {
@@ -567,10 +570,15 @@ static int convert_variable_fields(Dwarf_Die *vr_die, const char *varname,
 	}
 
 	/* Get the offset of the field */
-	ret = die_get_data_member_location(die_mem, &offs);
-	if (ret < 0) {
-		pr_warning("Failed to get the offset of %s.\n", field->name);
-		return ret;
+	if (tag == DW_TAG_union_type) {
+		offs = 0;
+	} else {
+		ret = die_get_data_member_location(die_mem, &offs);
+		if (ret < 0) {
+			pr_warning("Failed to get the offset of %s.\n",
+				   field->name);
+			return ret;
+		}
 	}
 	ref->offset += (long)offs;
 
-- 
1.7.1

[PATCH 02/18] perf test: Fixup for the die() removal

[Arnaldo Carvalho de Melo]

From: Namhyung Kim <namhyung.kim@lge.com>

The commit 32c7f7383a09 ("perf test: Remove die() calls") replaced die()
call to pr_debug + return -1, but it should be pr_err otherwise it'll
not show up unless -v option is given.  Fix it.

Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1347415866-303-1-git-send-email-namhyung@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/builtin-test.c |    8 ++++----
 1 files changed, 4 insertions(+), 4 deletions(-)

diff --git a/tools/perf/builtin-test.c b/tools/perf/builtin-test.c
index d33143e..4aed155 100644
--- a/tools/perf/builtin-test.c
+++ b/tools/perf/builtin-test.c
@@ -1026,15 +1026,15 @@ static int __test__rdpmc(void)
 
 	fd = sys_perf_event_open(&attr, 0, -1, -1, 0);
 	if (fd < 0) {
-		pr_debug("Error: sys_perf_event_open() syscall returned "
-			 "with %d (%s)\n", fd, strerror(errno));
+		pr_err("Error: sys_perf_event_open() syscall returned "
+		       "with %d (%s)\n", fd, strerror(errno));
 		return -1;
 	}
 
 	addr = mmap(NULL, page_size, PROT_READ, MAP_SHARED, fd, 0);
 	if (addr == (void *)(-1)) {
-		pr_debug("Error: mmap() syscall returned with (%s)\n",
-			 strerror(errno));
+		pr_err("Error: mmap() syscall returned with (%s)\n",
+		       strerror(errno));
 		goto out_close;
 	}
 
-- 
1.7.1

[PATCH 03/18] perf sched: Fixup for the die() removal

[Arnaldo Carvalho de Melo]

From: Namhyung Kim <namhyung.kim@lge.com>

The commit a116e05dcf61 ("perf sched: Remove die() calls") replaced
die() call to pr_debug + return -1, but it should be pr_err otherwise
it'll not show up unless -v option is given.  Fix it.

Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1347415866-303-2-git-send-email-namhyung@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/builtin-sched.c |   14 +++++++-------
 1 files changed, 7 insertions(+), 7 deletions(-)

diff --git a/tools/perf/builtin-sched.c b/tools/perf/builtin-sched.c
index af305f5..9b9e32e 100644
--- a/tools/perf/builtin-sched.c
+++ b/tools/perf/builtin-sched.c
@@ -438,8 +438,8 @@ static int self_open_counters(void)
 	fd = sys_perf_event_open(&attr, 0, -1, -1, 0);
 
 	if (fd < 0)
-		pr_debug("Error: sys_perf_event_open() syscall returned"
-			 "with %d (%s)\n", fd, strerror(errno));
+		pr_err("Error: sys_perf_event_open() syscall returned "
+		       "with %d (%s)\n", fd, strerror(errno));
 	return fd;
 }
 
@@ -700,7 +700,7 @@ static int replay_switch_event(struct perf_sched *sched,
 		delta = 0;
 
 	if (delta < 0) {
-		pr_debug("hm, delta: %" PRIu64 " < 0 ?\n", delta);
+		pr_err("hm, delta: %" PRIu64 " < 0 ?\n", delta);
 		return -1;
 	}
 
@@ -990,7 +990,7 @@ static int latency_runtime_event(struct perf_sched *sched,
 			return -1;
 		atoms = thread_atoms_search(&sched->atom_root, thread, &sched->cmp_pid);
 		if (!atoms) {
-			pr_debug("in-event: Internal tree error");
+			pr_err("in-event: Internal tree error");
 			return -1;
 		}
 		if (add_sched_out_event(atoms, 'R', timestamp))
@@ -1024,7 +1024,7 @@ static int latency_wakeup_event(struct perf_sched *sched,
 			return -1;
 		atoms = thread_atoms_search(&sched->atom_root, wakee, &sched->cmp_pid);
 		if (!atoms) {
-			pr_debug("wakeup-event: Internal tree error");
+			pr_err("wakeup-event: Internal tree error");
 			return -1;
 		}
 		if (add_sched_out_event(atoms, 'S', timestamp))
@@ -1079,7 +1079,7 @@ static int latency_migrate_task_event(struct perf_sched *sched,
 		register_pid(sched, migrant->pid, migrant->comm);
 		atoms = thread_atoms_search(&sched->atom_root, migrant, &sched->cmp_pid);
 		if (!atoms) {
-			pr_debug("migration-event: Internal tree error");
+			pr_err("migration-event: Internal tree error");
 			return -1;
 		}
 		if (add_sched_out_event(atoms, 'R', timestamp))
@@ -1286,7 +1286,7 @@ static int map_switch_event(struct perf_sched *sched, struct perf_evsel *evsel,
 		delta = 0;
 
 	if (delta < 0) {
-		pr_debug("hm, delta: %" PRIu64 " < 0 ?\n", delta);
+		pr_err("hm, delta: %" PRIu64 " < 0 ?\n", delta);
 		return -1;
 	}
 
-- 
1.7.1

[PATCH 04/18] tools lib traceevent: Define _GNU_SOURCE in Makefile

[Arnaldo Carvalho de Melo]

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

For the reasons stated on:

  commit 0a84f00
  Author: David Daney <david.daney@cavium.com>

    perf tools: Fix broken build by defining _GNU_SOURCE in Makefile

Acked-by: Steven Rostedt <rostedt@goodmis.org>
Cc: David Ahern <dsahern@gmail.com>
Cc: David Daney <david.daney@cavium.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Namhyung Kim <namhyung@gmail.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Link: http://lkml.kernel.org/n/tip-e2nofbmj4uf0ykgsytxvt9pu@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/lib/traceevent/Makefile      |    2 +-
 tools/lib/traceevent/event-parse.c |    1 -
 2 files changed, 1 insertions(+), 2 deletions(-)

diff --git a/tools/lib/traceevent/Makefile b/tools/lib/traceevent/Makefile
index 14131cb..04d959f 100644
--- a/tools/lib/traceevent/Makefile
+++ b/tools/lib/traceevent/Makefile
@@ -129,7 +129,7 @@ CFLAGS ?= -g -Wall
 
 # Append required CFLAGS
 override CFLAGS += $(CONFIG_FLAGS) $(INCLUDES) $(PLUGIN_DIR_SQ)
-override CFLAGS += $(udis86-flags)
+override CFLAGS += $(udis86-flags) -D_GNU_SOURCE
 
 ifeq ($(VERBOSE),1)
   Q =
diff --git a/tools/lib/traceevent/event-parse.c b/tools/lib/traceevent/event-parse.c
index 2c54cdd..77ebeb8 100644
--- a/tools/lib/traceevent/event-parse.c
+++ b/tools/lib/traceevent/event-parse.c
@@ -24,7 +24,6 @@
  *  Frederic Weisbecker gave his permission to relicense the code to
  *  the Lesser General Public License.
  */
-#define _GNU_SOURCE
 #include <stdio.h>
 #include <stdlib.h>
 #include <string.h>
-- 
1.7.1

[PATCH 05/18] perf hists browser: Fix output for 100.00%

[Arnaldo Carvalho de Melo]

From: Namhyung Kim <namhyung.kim@lge.com>

Current hpp format functions assume that the output will fit to 6
character including % sign (XX.YY%) so used "%5.2f%%" as a format
string.  However it might be the case if collapsing resulted in a single
entry which has 100.00% (7 character) of period. In this case the output
will be shifted by 1 character.

Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1347431706-7839-1-git-send-email-namhyung@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/ui/browsers/hists.c |    2 +-
 tools/perf/ui/gtk/browser.c    |    2 +-
 tools/perf/ui/hist.c           |   28 ++++++++++++++--------------
 3 files changed, 16 insertions(+), 16 deletions(-)

diff --git a/tools/perf/ui/browsers/hists.c b/tools/perf/ui/browsers/hists.c
index 5a5739b..1178317 100644
--- a/tools/perf/ui/browsers/hists.c
+++ b/tools/perf/ui/browsers/hists.c
@@ -571,7 +571,7 @@ static int hist_browser__hpp_color_ ## _name(struct perf_hpp *hpp,	\
 {									\
 	double percent = 100.0 * he->_field / hpp->total_period;	\
 	*(double *)hpp->ptr = percent;					\
-	return scnprintf(hpp->buf, hpp->size, "%5.2f%%", percent);	\
+	return scnprintf(hpp->buf, hpp->size, "%6.2f%%", percent);	\
 }
 
 HPP__COLOR_FN(overhead, period)
diff --git a/tools/perf/ui/gtk/browser.c b/tools/perf/ui/gtk/browser.c
index 55acba6..7ff99ec 100644
--- a/tools/perf/ui/gtk/browser.c
+++ b/tools/perf/ui/gtk/browser.c
@@ -56,7 +56,7 @@ static int perf_gtk__hpp_color_ ## _name(struct perf_hpp *hpp,			\
 	markup = perf_gtk__get_percent_color(percent);				\
 	if (markup)								\
 		ret += scnprintf(hpp->buf, hpp->size, "%s", markup);		\
-	ret += scnprintf(hpp->buf + ret, hpp->size - ret, "%5.2f%%", percent); 	\
+	ret += scnprintf(hpp->buf + ret, hpp->size - ret, "%6.2f%%", percent); 	\
 	if (markup)								\
 		ret += scnprintf(hpp->buf + ret, hpp->size - ret, "</span>"); 	\
 										\
diff --git a/tools/perf/ui/hist.c b/tools/perf/ui/hist.c
index 407e855..e3f8cd4 100644
--- a/tools/perf/ui/hist.c
+++ b/tools/perf/ui/hist.c
@@ -33,13 +33,13 @@ static int hpp__color_overhead(struct perf_hpp *hpp, struct hist_entry *he)
 			percent = 0.0;
 	}
 
-	return percent_color_snprintf(hpp->buf, hpp->size, "  %5.2f%%", percent);
+	return percent_color_snprintf(hpp->buf, hpp->size, " %6.2f%%", percent);
 }
 
 static int hpp__entry_overhead(struct perf_hpp *hpp, struct hist_entry *he)
 {
 	double percent = 100.0 * he->period / hpp->total_period;
-	const char *fmt = symbol_conf.field_sep ? "%.2f" : "  %5.2f%%";
+	const char *fmt = symbol_conf.field_sep ? "%.2f" : " %6.2f%%";
 
 	if (hpp->ptr) {
 		struct hists *old_hists = hpp->ptr;
@@ -57,52 +57,52 @@ static int hpp__entry_overhead(struct perf_hpp *hpp, struct hist_entry *he)
 
 static int hpp__header_overhead_sys(struct perf_hpp *hpp)
 {
-	const char *fmt = symbol_conf.field_sep ? "%s" : "%6s";
+	const char *fmt = symbol_conf.field_sep ? "%s" : "%7s";
 
 	return scnprintf(hpp->buf, hpp->size, fmt, "sys");
 }
 
 static int hpp__width_overhead_sys(struct perf_hpp *hpp __maybe_unused)
 {
-	return 6;
+	return 7;
 }
 
 static int hpp__color_overhead_sys(struct perf_hpp *hpp, struct hist_entry *he)
 {
 	double percent = 100.0 * he->period_sys / hpp->total_period;
-	return percent_color_snprintf(hpp->buf, hpp->size, "%5.2f%%", percent);
+	return percent_color_snprintf(hpp->buf, hpp->size, "%6.2f%%", percent);
 }
 
 static int hpp__entry_overhead_sys(struct perf_hpp *hpp, struct hist_entry *he)
 {
 	double percent = 100.0 * he->period_sys / hpp->total_period;
-	const char *fmt = symbol_conf.field_sep ? "%.2f" : "%5.2f%%";
+	const char *fmt = symbol_conf.field_sep ? "%.2f" : "%6.2f%%";
 
 	return scnprintf(hpp->buf, hpp->size, fmt, percent);
 }
 
 static int hpp__header_overhead_us(struct perf_hpp *hpp)
 {
-	const char *fmt = symbol_conf.field_sep ? "%s" : "%6s";
+	const char *fmt = symbol_conf.field_sep ? "%s" : "%7s";
 
 	return scnprintf(hpp->buf, hpp->size, fmt, "user");
 }
 
 static int hpp__width_overhead_us(struct perf_hpp *hpp __maybe_unused)
 {
-	return 6;
+	return 7;
 }
 
 static int hpp__color_overhead_us(struct perf_hpp *hpp, struct hist_entry *he)
 {
 	double percent = 100.0 * he->period_us / hpp->total_period;
-	return percent_color_snprintf(hpp->buf, hpp->size, "%5.2f%%", percent);
+	return percent_color_snprintf(hpp->buf, hpp->size, "%6.2f%%", percent);
 }
 
 static int hpp__entry_overhead_us(struct perf_hpp *hpp, struct hist_entry *he)
 {
 	double percent = 100.0 * he->period_us / hpp->total_period;
-	const char *fmt = symbol_conf.field_sep ? "%.2f" : "%5.2f%%";
+	const char *fmt = symbol_conf.field_sep ? "%.2f" : "%6.2f%%";
 
 	return scnprintf(hpp->buf, hpp->size, fmt, percent);
 }
@@ -121,14 +121,14 @@ static int hpp__color_overhead_guest_sys(struct perf_hpp *hpp,
 					 struct hist_entry *he)
 {
 	double percent = 100.0 * he->period_guest_sys / hpp->total_period;
-	return percent_color_snprintf(hpp->buf, hpp->size, "  %5.2f%% ", percent);
+	return percent_color_snprintf(hpp->buf, hpp->size, " %6.2f%% ", percent);
 }
 
 static int hpp__entry_overhead_guest_sys(struct perf_hpp *hpp,
 					 struct hist_entry *he)
 {
 	double percent = 100.0 * he->period_guest_sys / hpp->total_period;
-	const char *fmt = symbol_conf.field_sep ? "%.2f" : "  %5.2f%% ";
+	const char *fmt = symbol_conf.field_sep ? "%.2f" : " %6.2f%% ";
 
 	return scnprintf(hpp->buf, hpp->size, fmt, percent);
 }
@@ -147,14 +147,14 @@ static int hpp__color_overhead_guest_us(struct perf_hpp *hpp,
 					struct hist_entry *he)
 {
 	double percent = 100.0 * he->period_guest_us / hpp->total_period;
-	return percent_color_snprintf(hpp->buf, hpp->size, "  %5.2f%% ", percent);
+	return percent_color_snprintf(hpp->buf, hpp->size, " %6.2f%% ", percent);
 }
 
 static int hpp__entry_overhead_guest_us(struct perf_hpp *hpp,
 					struct hist_entry *he)
 {
 	double percent = 100.0 * he->period_guest_us / hpp->total_period;
-	const char *fmt = symbol_conf.field_sep ? "%.2f" : "  %5.2f%% ";
+	const char *fmt = symbol_conf.field_sep ? "%.2f" : " %6.2f%% ";
 
 	return scnprintf(hpp->buf, hpp->size, fmt, percent);
 }
-- 
1.7.1

[PATCH 06/18] perf hists browser: Fix first column printing

[Arnaldo Carvalho de Melo]

From: Namhyung Kim <namhyung.kim@lge.com>

As a side effect of commit f5951d56a2ab ("perf hists browser: Use
perf_hpp__format functions") perf report TUI got a problem of not
refreshing the first character.

Since the previous patch restores the column width of "overhead" to 7
we can start at column 0 now.

Reported-by: Ingo Molnar <mingo@kernel.org>
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1347431706-7839-2-git-send-email-namhyung@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/ui/browsers/hists.c |    4 ++--
 1 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/tools/perf/ui/browsers/hists.c b/tools/perf/ui/browsers/hists.c
index 1178317..a21f40b 100644
--- a/tools/perf/ui/browsers/hists.c
+++ b/tools/perf/ui/browsers/hists.c
@@ -605,7 +605,7 @@ static int hist_browser__show_entry(struct hist_browser *browser,
 	char s[256];
 	double percent;
 	int i, printed = 0;
-	int width = browser->b.width - 1;
+	int width = browser->b.width;
 	char folded_sign = ' ';
 	bool current_entry = ui_browser__is_current_entry(&browser->b, row);
 	off_t row_offset = entry->row_offset;
@@ -627,7 +627,7 @@ static int hist_browser__show_entry(struct hist_browser *browser,
 			.total_period	= browser->hists->stats.total_period,
 		};
 
-		ui_browser__gotorc(&browser->b, row, 1);
+		ui_browser__gotorc(&browser->b, row, 0);
 
 		for (i = 0; i < PERF_HPP__MAX_INDEX; i++) {
 			if (!perf_hpp__format[i].cond)
-- 
1.7.1

[PATCH 07/18] perf tools: Add sort__has_sym

[Arnaldo Carvalho de Melo]

From: Namhyung Kim <namhyung.kim@lge.com>

The sort__has_sym variable is for checking whether the sort_list
includes 'symbol' as a sort key.  It will be used for later patch.

Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1347611729-16994-1-git-send-email-namhyung@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/util/sort.c |    5 +++++
 tools/perf/util/sort.h |    1 +
 2 files changed, 6 insertions(+), 0 deletions(-)

diff --git a/tools/perf/util/sort.c b/tools/perf/util/sort.c
index 0981bc7..b5b1b92 100644
--- a/tools/perf/util/sort.c
+++ b/tools/perf/util/sort.c
@@ -8,6 +8,7 @@ const char	default_sort_order[] = "comm,dso,symbol";
 const char	*sort_order = default_sort_order;
 int		sort__need_collapse = 0;
 int		sort__has_parent = 0;
+int		sort__has_sym = 0;
 int		sort__branch_mode = -1; /* -1 = means not set */
 
 enum sort_type	sort__first_dimension;
@@ -511,6 +512,10 @@ int sort_dimension__add(const char *tok)
 				return -EINVAL;
 			}
 			sort__has_parent = 1;
+		} else if (sd->entry == &sort_sym ||
+			   sd->entry == &sort_sym_from ||
+			   sd->entry == &sort_sym_to) {
+			sort__has_sym = 1;
 		}
 
 		if (sd->taken)
diff --git a/tools/perf/util/sort.h b/tools/perf/util/sort.h
index e459c98..12d6347 100644
--- a/tools/perf/util/sort.h
+++ b/tools/perf/util/sort.h
@@ -31,6 +31,7 @@ extern const char *parent_pattern;
 extern const char default_sort_order[];
 extern int sort__need_collapse;
 extern int sort__has_parent;
+extern int sort__has_sym;
 extern int sort__branch_mode;
 extern struct sort_entry sort_comm;
 extern struct sort_entry sort_dso;
-- 
1.7.1

[PATCH 08/18] perf report: Enable integrated annotation only if possible

[Arnaldo Carvalho de Melo]

From: Namhyung Kim <namhyung.kim@lge.com>

The integrated annotation feature is supported only in TUI mode.  Also
it should be enabled with 'symbol' sort key otherwise resulting hist
entry doesn't need to have same symbol as of a sample so that it can
fail on hist_entry__inc_addr_samples with -ERANGE.

You can easily see it when start perf report TUI without symbol* sort
key.  This patch fixes the problem.

Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1347611729-16994-2-git-send-email-namhyung@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/builtin-report.c |   12 ++++++------
 1 files changed, 6 insertions(+), 6 deletions(-)

diff --git a/tools/perf/builtin-report.c b/tools/perf/builtin-report.c
index 97b2e63..b6696dd 100644
--- a/tools/perf/builtin-report.c
+++ b/tools/perf/builtin-report.c
@@ -93,7 +93,7 @@ static int perf_report__add_branch_hist_entry(struct perf_tool *tool,
 			struct annotation *notes;
 			err = -ENOMEM;
 			bx = he->branch_info;
-			if (bx->from.sym && use_browser > 0) {
+			if (bx->from.sym && use_browser == 1 && sort__has_sym) {
 				notes = symbol__annotation(bx->from.sym);
 				if (!notes->src
 				    && symbol__alloc_hist(bx->from.sym) < 0)
@@ -107,7 +107,7 @@ static int perf_report__add_branch_hist_entry(struct perf_tool *tool,
 					goto out;
 			}
 
-			if (bx->to.sym && use_browser > 0) {
+			if (bx->to.sym && use_browser == 1 && sort__has_sym) {
 				notes = symbol__annotation(bx->to.sym);
 				if (!notes->src
 				    && symbol__alloc_hist(bx->to.sym) < 0)
@@ -162,7 +162,7 @@ static int perf_evsel__add_hist_entry(struct perf_evsel *evsel,
 	 * so we don't allocated the extra space needed because the stdio
 	 * code will not use it.
 	 */
-	if (he->ms.sym != NULL && use_browser > 0) {
+	if (he->ms.sym != NULL && use_browser == 1 && sort__has_sym) {
 		struct annotation *notes = symbol__annotation(he->ms.sym);
 
 		assert(evsel != NULL);
@@ -692,12 +692,14 @@ int cmd_report(int argc, const char **argv, const char *prefix __maybe_unused)
 	else
 		use_browser = 0;
 
+	setup_sorting(report_usage, options);
+
 	/*
 	 * Only in the newt browser we are doing integrated annotation,
 	 * so don't allocate extra space that won't be used in the stdio
 	 * implementation.
 	 */
-	if (use_browser > 0) {
+	if (use_browser == 1 && sort__has_sym) {
 		symbol_conf.priv_size = sizeof(struct annotation);
 		report.annotate_init  = symbol__annotate_init;
 		/*
@@ -720,8 +722,6 @@ int cmd_report(int argc, const char **argv, const char *prefix __maybe_unused)
 	if (symbol__init() < 0)
 		goto error;
 
-	setup_sorting(report_usage, options);
-
 	if (parent_pattern != default_parent_pattern) {
 		if (sort_dimension__add("parent") < 0)
 			goto error;
-- 
1.7.1

[PATCH 09/18] perf stat: Move stats related code to util/stat.c

[Arnaldo Carvalho de Melo]

From: Xiao Guangrong <xiaoguangrong@linux.vnet.ibm.com>

Then, the code can be shared between kvm events and perf stat.

Signed-off-by: Xiao Guangrong <xiaoguangrong@linux.vnet.ibm.com>
[ Dong Hao <haodong@linux.vnet.ibm.com>: rebase it on acme's git tree ]
Signed-off-by: Dong Hao <haodong@linux.vnet.ibm.com>
Cc: Avi Kivity <avi@redhat.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: kvm@vger.kernel.org
Cc: Marcelo Tosatti <mtosatti@redhat.com>
Cc: Runzhen Wang <runzhen@linux.vnet.ibm.com>
Cc: Xiao Guangrong <xiaoguangrong@linux.vnet.ibm.com
Link: http://lkml.kernel.org/r/1347870675-31495-3-git-send-email-haodong@linux.vnet.ibm.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/Makefile       |    1 +
 tools/perf/builtin-stat.c |   56 +------------------------------------------
 tools/perf/util/stat.c    |   57 +++++++++++++++++++++++++++++++++++++++++++++
 tools/perf/util/stat.h    |   16 ++++++++++++
 4 files changed, 76 insertions(+), 54 deletions(-)
 create mode 100644 tools/perf/util/stat.c
 create mode 100644 tools/perf/util/stat.h

diff --git a/tools/perf/Makefile b/tools/perf/Makefile
index 209774b..5077f8e 100644
--- a/tools/perf/Makefile
+++ b/tools/perf/Makefile
@@ -406,6 +406,7 @@ LIB_OBJS += $(OUTPUT)util/target.o
 LIB_OBJS += $(OUTPUT)util/rblist.o
 LIB_OBJS += $(OUTPUT)util/intlist.o
 LIB_OBJS += $(OUTPUT)util/vdso.o
+LIB_OBJS += $(OUTPUT)util/stat.o
 
 LIB_OBJS += $(OUTPUT)ui/helpline.o
 LIB_OBJS += $(OUTPUT)ui/hist.o
diff --git a/tools/perf/builtin-stat.c b/tools/perf/builtin-stat.c
index dab347d..3c43a35 100644
--- a/tools/perf/builtin-stat.c
+++ b/tools/perf/builtin-stat.c
@@ -51,13 +51,13 @@
 #include "util/evsel.h"
 #include "util/debug.h"
 #include "util/color.h"
+#include "util/stat.h"
 #include "util/header.h"
 #include "util/cpumap.h"
 #include "util/thread.h"
 #include "util/thread_map.h"
 
 #include <sys/prctl.h>
-#include <math.h>
 #include <locale.h>
 
 #define DEFAULT_SEPARATOR	" "
@@ -199,11 +199,6 @@ static int			output_fd;
 
 static volatile int done = 0;
 
-struct stats
-{
-	double n, mean, M2;
-};
-
 struct perf_stat {
 	struct stats	  res_stats[3];
 };
@@ -220,50 +215,6 @@ static void perf_evsel__free_stat_priv(struct perf_evsel *evsel)
 	evsel->priv = NULL;
 }
 
-static void update_stats(struct stats *stats, u64 val)
-{
-	double delta;
-
-	stats->n++;
-	delta = val - stats->mean;
-	stats->mean += delta / stats->n;
-	stats->M2 += delta*(val - stats->mean);
-}
-
-static double avg_stats(struct stats *stats)
-{
-	return stats->mean;
-}
-
-/*
- * http://en.wikipedia.org/wiki/Algorithms_for_calculating_variance
- *
- *       (\Sum n_i^2) - ((\Sum n_i)^2)/n
- * s^2 = -------------------------------
- *                  n - 1
- *
- * http://en.wikipedia.org/wiki/Stddev
- *
- * The std dev of the mean is related to the std dev by:
- *
- *             s
- * s_mean = -------
- *          sqrt(n)
- *
- */
-static double stddev_stats(struct stats *stats)
-{
-	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);
-}
-
 static struct stats runtime_nsecs_stats[MAX_NR_CPUS];
 static struct stats runtime_cycles_stats[MAX_NR_CPUS];
 static struct stats runtime_stalled_cycles_front_stats[MAX_NR_CPUS];
@@ -559,10 +510,7 @@ static int run_perf_stat(int argc __maybe_unused, const char **argv)
 
 static void print_noise_pct(double total, double avg)
 {
-	double pct = 0.0;
-
-	if (avg)
-		pct = 100.0*total/avg;
+	double pct = rel_stddev_stats(total, avg);
 
 	if (csv_output)
 		fprintf(output, "%s%.2f%%", csv_sep, pct);
diff --git a/tools/perf/util/stat.c b/tools/perf/util/stat.c
new file mode 100644
index 0000000..2374212
--- /dev/null
+++ b/tools/perf/util/stat.c
@@ -0,0 +1,57 @@
+#include <math.h>
+
+#include "stat.h"
+
+void update_stats(struct stats *stats, u64 val)
+{
+	double delta;
+
+	stats->n++;
+	delta = val - stats->mean;
+	stats->mean += delta / stats->n;
+	stats->M2 += delta*(val - stats->mean);
+}
+
+double avg_stats(struct stats *stats)
+{
+	return stats->mean;
+}
+
+/*
+ * http://en.wikipedia.org/wiki/Algorithms_for_calculating_variance
+ *
+ *       (\Sum n_i^2) - ((\Sum n_i)^2)/n
+ * s^2 = -------------------------------
+ *                  n - 1
+ *
+ * http://en.wikipedia.org/wiki/Stddev
+ *
+ * The std dev of the mean is related to the std dev by:
+ *
+ *             s
+ * s_mean = -------
+ *          sqrt(n)
+ *
+ */
+double stddev_stats(struct stats *stats)
+{
+	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);
+}
+
+double rel_stddev_stats(double stddev, double avg)
+{
+	double pct = 0.0;
+
+	if (avg)
+		pct = 100.0 * stddev/avg;
+
+	return pct;
+}
diff --git a/tools/perf/util/stat.h b/tools/perf/util/stat.h
new file mode 100644
index 0000000..588367c
--- /dev/null
+++ b/tools/perf/util/stat.h
@@ -0,0 +1,16 @@
+#ifndef __PERF_STATS_H
+#define __PERF_STATS_H
+
+#include "types.h"
+
+struct stats
+{
+	double n, mean, M2;
+};
+
+void update_stats(struct stats *stats, u64 val);
+double avg_stats(struct stats *stats);
+double stddev_stats(struct stats *stats);
+double rel_stddev_stats(double stddev, double avg);
+
+#endif
-- 
1.7.1

[PATCH 10/18] perf archive: Remove -f from the rm command

[Arnaldo Carvalho de Melo]

From: Irina Tirdea <irina.tirdea@intel.com>

In Android, rm does not support the -f parameter.

Remove -f from rm and make sure rm does not fail even if the files to be
removed are not found.

Signed-off-by: Irina Tirdea <irina.tirdea@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Namhyung Kim <namhyung.kim@lge.com>
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/1347574063-22521-4-git-send-email-irina.tirdea@intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/perf-archive.sh |    4 ++--
 1 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/tools/perf/perf-archive.sh b/tools/perf/perf-archive.sh
index 95b6f8b..da94179 100644
--- a/tools/perf/perf-archive.sh
+++ b/tools/perf/perf-archive.sh
@@ -24,7 +24,7 @@ NOBUILDID=0000000000000000000000000000000000000000
 perf buildid-list -i $PERF_DATA --with-hits | grep -v "^$NOBUILDID " > $BUILDIDS
 if [ ! -s $BUILDIDS ] ; then
 	echo "perf archive: no build-ids found"
-	rm -f $BUILDIDS
+	rm $BUILDIDS || true
 	exit 1
 fi
 
@@ -40,7 +40,7 @@ while read build_id ; do
 done
 
 tar cfj $PERF_DATA.tar.bz2 -C $PERF_BUILDID_DIR -T $MANIFEST
-rm -f $MANIFEST $BUILDIDS
+rm $MANIFEST $BUILDIDS || true
 echo -e "Now please run:\n"
 echo -e "$ tar xvf $PERF_DATA.tar.bz2 -C ~/.debug\n"
 echo "wherever you need to run 'perf report' on."
-- 
1.7.1

[PATCH 11/18] perf archive: Make 'f' the last parameter for tar

[Arnaldo Carvalho de Melo]

From: Irina Tirdea <irina.tirdea@intel.com>

On some systems, tar needs to specify the name of the archive immediately
after the -f parameter.

Change the order of the parameters so tar can run properly.

Signed-off-by: Irina Tirdea <irina.tirdea@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Namhyung Kim <namhyung.kim@lge.com>
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/1347574063-22521-5-git-send-email-irina.tirdea@intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/perf-archive.sh |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/tools/perf/perf-archive.sh b/tools/perf/perf-archive.sh
index da94179..e919306 100644
--- a/tools/perf/perf-archive.sh
+++ b/tools/perf/perf-archive.sh
@@ -39,7 +39,7 @@ while read build_id ; do
 	echo ${filename#$PERF_BUILDID_LINKDIR} >> $MANIFEST
 done
 
-tar cfj $PERF_DATA.tar.bz2 -C $PERF_BUILDID_DIR -T $MANIFEST
+tar cjf $PERF_DATA.tar.bz2 -C $PERF_BUILDID_DIR -T $MANIFEST
 rm $MANIFEST $BUILDIDS || true
 echo -e "Now please run:\n"
 echo -e "$ tar xvf $PERF_DATA.tar.bz2 -C ~/.debug\n"
-- 
1.7.1

[PATCH 12/18] perf symbols: Filter samples with unresolved symbol when "--symbols" option is used

[Arnaldo Carvalho de Melo]

From: Feng Tang <feng.tang@intel.com>

Report/top commands support to only handle specific symbols with
"--symbols" option, but current code will keep those samples whose
symbol can't be resolved, which should actually be filtered.

If we run following commands:
$perf record -a tree
$perf report --symbols intel_idle -n
the output will be:

Without the patch:
==================
    46.27%        156     sshd  [unknown]
    26.05%         48  swapper  [kernel.kallsyms]
    17.26%         38     tree  libc-2.12.1.so
     7.69%         17     tree  tree
     2.73%          6     tree  ld-2.12.1.so

With the patch:
===============
   100.00%         48  swapper  [kernel.kallsyms]

Signed-off-by: Feng Tang <feng.tang@intel.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: David Ahern <dsahern@gmail.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1347007349-3102-2-git-send-email-feng.tang@intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/util/event.c |    5 +++--
 1 files changed, 3 insertions(+), 2 deletions(-)

diff --git a/tools/perf/util/event.c b/tools/perf/util/event.c
index 8202f5c..6715b19 100644
--- a/tools/perf/util/event.c
+++ b/tools/perf/util/event.c
@@ -904,8 +904,9 @@ int perf_event__preprocess_sample(const union perf_event *event,
 		al->sym = map__find_symbol(al->map, al->addr, filter);
 	}
 
-	if (symbol_conf.sym_list && al->sym &&
-	    !strlist__has_entry(symbol_conf.sym_list, al->sym->name))
+	if (symbol_conf.sym_list &&
+		(!al->sym || !strlist__has_entry(symbol_conf.sym_list,
+						al->sym->name)))
 		goto out_filtered;
 
 	return 0;
-- 
1.7.1

[PATCH 13/18] perf scripts: Add --symbols option to handle specific symbols

[Arnaldo Carvalho de Melo]

From: Feng Tang <feng.tang@intel.com>

Since perf script no longer only handle the trace points, we can add the
symbol filter option so that scripts can handle specified samples.

Signed-off-by: Feng Tang <feng.tang@intel.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: David Ahern <dsahern@gmail.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1347007349-3102-3-git-send-email-feng.tang@intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/builtin-script.c |    3 +++
 1 files changed, 3 insertions(+), 0 deletions(-)

diff --git a/tools/perf/builtin-script.c b/tools/perf/builtin-script.c
index 6d98a83..76577e6 100644
--- a/tools/perf/builtin-script.c
+++ b/tools/perf/builtin-script.c
@@ -14,6 +14,7 @@
 #include "util/util.h"
 #include "util/evlist.h"
 #include "util/evsel.h"
+#include "util/sort.h"
 #include <linux/bitmap.h>
 
 static char const		*script_name;
@@ -1143,6 +1144,8 @@ static const struct option options[] = {
 		     parse_output_fields),
 	OPT_BOOLEAN('a', "all-cpus", &system_wide,
 		     "system-wide collection from all CPUs"),
+	OPT_STRING('S', "symbols", &symbol_conf.sym_list_str, "symbol[,symbol...]",
+		   "only consider these symbols"),
 	OPT_STRING('C', "cpu", &cpu_list, "cpu", "list of cpus to profile"),
 	OPT_STRING('c', "comms", &symbol_conf.comm_list_str, "comm[,comm...]",
 		   "only display events for these comms"),
-- 
1.7.1

[PATCH 14/18] perf scripts: Add event_analyzing_sample-record/report

[Arnaldo Carvalho de Melo]

From: Feng Tang <feng.tang@intel.com>

So that event_analyzing_sample.py can be shown by "perf script -l"

Signed-off-by: Feng Tang <feng.tang@intel.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: David Ahern <dsahern@gmail.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1347007349-3102-4-git-send-email-feng.tang@intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 .../python/bin/event_analyzing_sample-record       |    8 ++++++++
 .../python/bin/event_analyzing_sample-report       |    3 +++
 2 files changed, 11 insertions(+), 0 deletions(-)
 create mode 100644 tools/perf/scripts/python/bin/event_analyzing_sample-record
 create mode 100644 tools/perf/scripts/python/bin/event_analyzing_sample-report

diff --git a/tools/perf/scripts/python/bin/event_analyzing_sample-record b/tools/perf/scripts/python/bin/event_analyzing_sample-record
new file mode 100644
index 0000000..5ce652d
--- /dev/null
+++ b/tools/perf/scripts/python/bin/event_analyzing_sample-record
@@ -0,0 +1,8 @@
+#!/bin/bash
+
+#
+# event_analyzing_sample.py can cover all type of perf samples including
+# the tracepoints, so no special record requirements, just record what
+# you want to analyze.
+#
+perf record $@
diff --git a/tools/perf/scripts/python/bin/event_analyzing_sample-report b/tools/perf/scripts/python/bin/event_analyzing_sample-report
new file mode 100644
index 0000000..0941fc9
--- /dev/null
+++ b/tools/perf/scripts/python/bin/event_analyzing_sample-report
@@ -0,0 +1,3 @@
+#!/bin/bash
+# description: analyze all perf samples
+perf script $@ -s "$PERF_EXEC_PATH"/scripts/python/event_analyzing_sample.py
-- 
1.7.1

[PATCH 15/18] perf scripts: Export a find_scripts() function

[Arnaldo Carvalho de Melo]

From: Feng Tang <feng.tang@intel.com>

So that other perf commands/browser has a way to dig out the available
scripts info in system, this is a preparation for the script browser.

Signed-off-by: Feng Tang <feng.tang@intel.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: David Ahern <dsahern@gmail.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1347007349-3102-5-git-send-email-feng.tang@intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/builtin-script.c |   55 +++++++++++++++++++++++++++++++++++++++++++
 tools/perf/builtin.h        |    1 +
 2 files changed, 56 insertions(+), 0 deletions(-)

diff --git a/tools/perf/builtin-script.c b/tools/perf/builtin-script.c
index 76577e6..1be843a 100644
--- a/tools/perf/builtin-script.c
+++ b/tools/perf/builtin-script.c
@@ -1032,6 +1032,61 @@ static int list_available_scripts(const struct option *opt __maybe_unused,
 	exit(0);
 }
 
+/*
+ * Return -1 if none is found, otherwise the actual scripts number.
+ *
+ * Currently the only user of this function is the script browser, which
+ * will list all statically runnable scripts, select one, execute it and
+ * show the output in a perf browser.
+ */
+int find_scripts(char **scripts_array, char **scripts_path_array)
+{
+	struct dirent *script_next, *lang_next, script_dirent, lang_dirent;
+	char scripts_path[MAXPATHLEN];
+	DIR *scripts_dir, *lang_dir;
+	char lang_path[MAXPATHLEN];
+	char *temp;
+	int i = 0;
+
+	snprintf(scripts_path, MAXPATHLEN, "%s/scripts", perf_exec_path());
+
+	scripts_dir = opendir(scripts_path);
+	if (!scripts_dir)
+		return -1;
+
+	for_each_lang(scripts_path, scripts_dir, lang_dirent, lang_next) {
+		snprintf(lang_path, MAXPATHLEN, "%s/%s", scripts_path,
+			 lang_dirent.d_name);
+#ifdef NO_LIBPERL
+		if (strstr(lang_path, "perl"))
+			continue;
+#endif
+#ifdef NO_LIBPYTHON
+		if (strstr(lang_path, "python"))
+			continue;
+#endif
+
+		lang_dir = opendir(lang_path);
+		if (!lang_dir)
+			continue;
+
+		for_each_script(lang_path, lang_dir, script_dirent, script_next) {
+			/* Skip those real time scripts: xxxtop.p[yl] */
+			if (strstr(script_dirent.d_name, "top."))
+				continue;
+			sprintf(scripts_path_array[i], "%s/%s", lang_path,
+				script_dirent.d_name);
+			temp = strchr(script_dirent.d_name, '.');
+			snprintf(scripts_array[i],
+				(temp - script_dirent.d_name) + 1,
+				"%s", script_dirent.d_name);
+			i++;
+		}
+	}
+
+	return i;
+}
+
 static char *get_script_path(const char *script_root, const char *suffix)
 {
 	struct dirent *script_next, *lang_next, script_dirent, lang_dirent;
diff --git a/tools/perf/builtin.h b/tools/perf/builtin.h
index b382bd5..3ea74ed 100644
--- a/tools/perf/builtin.h
+++ b/tools/perf/builtin.h
@@ -36,4 +36,5 @@ extern int cmd_kvm(int argc, const char **argv, const char *prefix);
 extern int cmd_test(int argc, const char **argv, const char *prefix);
 extern int cmd_inject(int argc, const char **argv, const char *prefix);
 
+extern int find_scripts(char **scripts_array, char **scripts_path_array);
 #endif
-- 
1.7.1

[PATCH 16/18] perf report: Add missing perf_hpp__init for pipe-mode

[Arnaldo Carvalho de Melo]

From: Namhyung Kim <namhyung.kim@lge.com>

The perf_hpp__init() function was only called from setup_browser() so
that the pipe-mode missed the initialization thus didn't respond to
related options.  Fix it.

Reported-by: Robert Richter <robert.richter@amd.com>
Tested-by: Robert Richter <robert.richter@amd.com>
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Robert Richter <robert.richter@amd.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: linux-tip-commits@vger.kernel.org
Link: http://lkml.kernel.org/r/87txv28spl.fsf_-_@sejong.aot.lge.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/builtin-report.c |    4 +++-
 1 files changed, 3 insertions(+), 1 deletions(-)

diff --git a/tools/perf/builtin-report.c b/tools/perf/builtin-report.c
index b6696dd..1da243d 100644
--- a/tools/perf/builtin-report.c
+++ b/tools/perf/builtin-report.c
@@ -689,8 +689,10 @@ int cmd_report(int argc, const char **argv, const char *prefix __maybe_unused)
 
 	if (strcmp(report.input_name, "-") != 0)
 		setup_browser(true);
-	else
+	else {
 		use_browser = 0;
+		perf_hpp__init(false, false);
+	}
 
 	setup_sorting(report_usage, options);
 
-- 
1.7.1

[PATCH 17/18] perf/x86: Add cpumask for uncore pmu

[Arnaldo Carvalho de Melo]

From: Yan, Zheng <zheng.z.yan@intel.com>

This patch adds a cpumask file to the uncore pmu sysfs directory.  The
cpumask file contains one active cpu for every socket.

Signed-off-by: "Yan, Zheng" <zheng.z.yan@intel.com>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Acked-by: Ingo Molnar <mingo@kernel.org>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Stephane Eranian <eranian@google.com>
Cc: "Yan, Zheng" <zheng.z.yan@intel.com>
Link: http://lkml.kernel.org/r/1347263631-23175-2-git-send-email-zheng.z.yan@intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 arch/x86/kernel/cpu/perf_event_intel_uncore.c |   28 ++++++++++++++++++++++--
 arch/x86/kernel/cpu/perf_event_intel_uncore.h |    6 +++-
 2 files changed, 29 insertions(+), 5 deletions(-)

diff --git a/arch/x86/kernel/cpu/perf_event_intel_uncore.c b/arch/x86/kernel/cpu/perf_event_intel_uncore.c
index 0a55710..62ec3e6 100644
--- a/arch/x86/kernel/cpu/perf_event_intel_uncore.c
+++ b/arch/x86/kernel/cpu/perf_event_intel_uncore.c
@@ -2341,6 +2341,27 @@ int uncore_pmu_event_init(struct perf_event *event)
 	return ret;
 }
 
+static ssize_t uncore_get_attr_cpumask(struct device *dev,
+				struct device_attribute *attr, char *buf)
+{
+	int n = cpulist_scnprintf(buf, PAGE_SIZE - 2, &uncore_cpu_mask);
+
+	buf[n++] = '\n';
+	buf[n] = '\0';
+	return n;
+}
+
+static DEVICE_ATTR(cpumask, S_IRUGO, uncore_get_attr_cpumask, NULL);
+
+static struct attribute *uncore_pmu_attrs[] = {
+	&dev_attr_cpumask.attr,
+	NULL,
+};
+
+static struct attribute_group uncore_pmu_attr_group = {
+	.attrs = uncore_pmu_attrs,
+};
+
 static int __init uncore_pmu_register(struct intel_uncore_pmu *pmu)
 {
 	int ret;
@@ -2378,8 +2399,8 @@ static void __init uncore_type_exit(struct intel_uncore_type *type)
 		free_percpu(type->pmus[i].box);
 	kfree(type->pmus);
 	type->pmus = NULL;
-	kfree(type->attr_groups[1]);
-	type->attr_groups[1] = NULL;
+	kfree(type->events_group);
+	type->events_group = NULL;
 }
 
 static void __init uncore_types_exit(struct intel_uncore_type **types)
@@ -2431,9 +2452,10 @@ static int __init uncore_type_init(struct intel_uncore_type *type)
 		for (j = 0; j < i; j++)
 			attrs[j] = &type->event_descs[j].attr.attr;
 
-		type->attr_groups[1] = events_group;
+		type->events_group = events_group;
 	}
 
+	type->pmu_group = &uncore_pmu_attr_group;
 	type->pmus = pmus;
 	return 0;
 fail:
diff --git a/arch/x86/kernel/cpu/perf_event_intel_uncore.h b/arch/x86/kernel/cpu/perf_event_intel_uncore.h
index 5b81c18..e68a455 100644
--- a/arch/x86/kernel/cpu/perf_event_intel_uncore.h
+++ b/arch/x86/kernel/cpu/perf_event_intel_uncore.h
@@ -369,10 +369,12 @@ struct intel_uncore_type {
 	struct intel_uncore_pmu *pmus;
 	struct intel_uncore_ops *ops;
 	struct uncore_event_desc *event_descs;
-	const struct attribute_group *attr_groups[3];
+	const struct attribute_group *attr_groups[4];
 };
 
-#define format_group attr_groups[0]
+#define pmu_group attr_groups[0]
+#define format_group attr_groups[1]
+#define events_group attr_groups[2]
 
 struct intel_uncore_ops {
 	void (*init_box)(struct intel_uncore_box *);
-- 
1.7.1

[PATCH 18/18] perf stat: Check PMU cpumask file

[Arnaldo Carvalho de Melo]

From: Yan, Zheng <zheng.z.yan@intel.com>

If user doesn't explicitly specify CPU list, perf-stat only collects
events on CPUs listed in the PMU cpumask file.

Signed-off-by: "Yah, Zheng" <zheng.z.yan@intel.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Stephane Eranian <eranian@google.com>
Link: http://lkml.kernel.org/r/1347263631-23175-3-git-send-email-zheng.z.yan@intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/builtin-stat.c      |   30 ++++++++++++++++++++----------
 tools/perf/util/cpumap.c       |   22 +++++++++++++++-------
 tools/perf/util/cpumap.h       |    2 +-
 tools/perf/util/evsel.h        |    1 +
 tools/perf/util/parse-events.c |   18 ++++++++++++++----
 tools/perf/util/pmu.c          |   30 ++++++++++++++++++++++++++++++
 tools/perf/util/pmu.h          |    1 +
 7 files changed, 82 insertions(+), 22 deletions(-)

diff --git a/tools/perf/builtin-stat.c b/tools/perf/builtin-stat.c
index 3c43a35..e0f65fe 100644
--- a/tools/perf/builtin-stat.c
+++ b/tools/perf/builtin-stat.c
@@ -215,6 +215,16 @@ static void perf_evsel__free_stat_priv(struct perf_evsel *evsel)
 	evsel->priv = NULL;
 }
 
+static inline struct cpu_map *perf_evsel__cpus(struct perf_evsel *evsel)
+{
+	return (evsel->cpus && !target.cpu_list) ? evsel->cpus : evsel_list->cpus;
+}
+
+static inline int perf_evsel__nr_cpus(struct perf_evsel *evsel)
+{
+	return perf_evsel__cpus(evsel)->nr;
+}
+
 static struct stats runtime_nsecs_stats[MAX_NR_CPUS];
 static struct stats runtime_cycles_stats[MAX_NR_CPUS];
 static struct stats runtime_stalled_cycles_front_stats[MAX_NR_CPUS];
@@ -246,7 +256,7 @@ retry:
 		evsel->attr.exclude_guest = evsel->attr.exclude_host = 0;
 
 	if (perf_target__has_cpu(&target)) {
-		ret = perf_evsel__open_per_cpu(evsel, evsel_list->cpus);
+		ret = perf_evsel__open_per_cpu(evsel, perf_evsel__cpus(evsel));
 		if (ret)
 			goto check_ret;
 		return 0;
@@ -327,7 +337,7 @@ static int read_counter_aggr(struct perf_evsel *counter)
 	u64 *count = counter->counts->aggr.values;
 	int i;
 
-	if (__perf_evsel__read(counter, evsel_list->cpus->nr,
+	if (__perf_evsel__read(counter, perf_evsel__nr_cpus(counter),
 			       evsel_list->threads->nr, scale) < 0)
 		return -1;
 
@@ -356,7 +366,7 @@ static int read_counter(struct perf_evsel *counter)
 	u64 *count;
 	int cpu;
 
-	for (cpu = 0; cpu < evsel_list->cpus->nr; cpu++) {
+	for (cpu = 0; cpu < perf_evsel__nr_cpus(counter); cpu++) {
 		if (__perf_evsel__read_on_cpu(counter, cpu, 0, scale) < 0)
 			return -1;
 
@@ -495,12 +505,12 @@ static int run_perf_stat(int argc __maybe_unused, const char **argv)
 	if (no_aggr) {
 		list_for_each_entry(counter, &evsel_list->entries, node) {
 			read_counter(counter);
-			perf_evsel__close_fd(counter, evsel_list->cpus->nr, 1);
+			perf_evsel__close_fd(counter, perf_evsel__nr_cpus(counter), 1);
 		}
 	} else {
 		list_for_each_entry(counter, &evsel_list->entries, node) {
 			read_counter_aggr(counter);
-			perf_evsel__close_fd(counter, evsel_list->cpus->nr,
+			perf_evsel__close_fd(counter, perf_evsel__nr_cpus(counter),
 					     evsel_list->threads->nr);
 		}
 	}
@@ -538,7 +548,7 @@ static void nsec_printout(int cpu, struct perf_evsel *evsel, double avg)
 	if (no_aggr)
 		sprintf(cpustr, "CPU%*d%s",
 			csv_output ? 0 : -4,
-			evsel_list->cpus->map[cpu], csv_sep);
+			perf_evsel__cpus(evsel)->map[cpu], csv_sep);
 
 	fprintf(output, fmt, cpustr, msecs, csv_sep, perf_evsel__name(evsel));
 
@@ -750,7 +760,7 @@ static void abs_printout(int cpu, struct perf_evsel *evsel, double avg)
 	if (no_aggr)
 		sprintf(cpustr, "CPU%*d%s",
 			csv_output ? 0 : -4,
-			evsel_list->cpus->map[cpu], csv_sep);
+			perf_evsel__cpus(evsel)->map[cpu], csv_sep);
 	else
 		cpu = 0;
 
@@ -911,14 +921,14 @@ static void print_counter(struct perf_evsel *counter)
 	u64 ena, run, val;
 	int cpu;
 
-	for (cpu = 0; cpu < evsel_list->cpus->nr; cpu++) {
+	for (cpu = 0; cpu < perf_evsel__nr_cpus(counter); cpu++) {
 		val = counter->counts->cpu[cpu].val;
 		ena = counter->counts->cpu[cpu].ena;
 		run = counter->counts->cpu[cpu].run;
 		if (run == 0 || ena == 0) {
 			fprintf(output, "CPU%*d%s%*s%s%*s",
 				csv_output ? 0 : -4,
-				evsel_list->cpus->map[cpu], csv_sep,
+				perf_evsel__cpus(counter)->map[cpu], csv_sep,
 				csv_output ? 0 : 18,
 				counter->supported ? CNTR_NOT_COUNTED : CNTR_NOT_SUPPORTED,
 				csv_sep,
@@ -1217,7 +1227,7 @@ int cmd_stat(int argc, const char **argv, const char *prefix __maybe_unused)
 
 	list_for_each_entry(pos, &evsel_list->entries, node) {
 		if (perf_evsel__alloc_stat_priv(pos) < 0 ||
-		    perf_evsel__alloc_counts(pos, evsel_list->cpus->nr) < 0)
+		    perf_evsel__alloc_counts(pos, perf_evsel__nr_cpus(pos)) < 0)
 			goto out_free_fd;
 	}
 
diff --git a/tools/perf/util/cpumap.c b/tools/perf/util/cpumap.c
index adc72f0..2b32ffa 100644
--- a/tools/perf/util/cpumap.c
+++ b/tools/perf/util/cpumap.c
@@ -38,24 +38,19 @@ static struct cpu_map *cpu_map__trim_new(int nr_cpus, int *tmp_cpus)
 	return cpus;
 }
 
-static struct cpu_map *cpu_map__read_all_cpu_map(void)
+struct cpu_map *cpu_map__read(FILE *file)
 {
 	struct cpu_map *cpus = NULL;
-	FILE *onlnf;
 	int nr_cpus = 0;
 	int *tmp_cpus = NULL, *tmp;
 	int max_entries = 0;
 	int n, cpu, prev;
 	char sep;
 
-	onlnf = fopen("/sys/devices/system/cpu/online", "r");
-	if (!onlnf)
-		return cpu_map__default_new();
-
 	sep = 0;
 	prev = -1;
 	for (;;) {
-		n = fscanf(onlnf, "%u%c", &cpu, &sep);
+		n = fscanf(file, "%u%c", &cpu, &sep);
 		if (n <= 0)
 			break;
 		if (prev >= 0) {
@@ -95,6 +90,19 @@ static struct cpu_map *cpu_map__read_all_cpu_map(void)
 		cpus = cpu_map__default_new();
 out_free_tmp:
 	free(tmp_cpus);
+	return cpus;
+}
+
+static struct cpu_map *cpu_map__read_all_cpu_map(void)
+{
+	struct cpu_map *cpus = NULL;
+	FILE *onlnf;
+
+	onlnf = fopen("/sys/devices/system/cpu/online", "r");
+	if (!onlnf)
+		return cpu_map__default_new();
+
+	cpus = cpu_map__read(onlnf);
 	fclose(onlnf);
 	return cpus;
 }
diff --git a/tools/perf/util/cpumap.h b/tools/perf/util/cpumap.h
index c415185..17b5264 100644
--- a/tools/perf/util/cpumap.h
+++ b/tools/perf/util/cpumap.h
@@ -11,7 +11,7 @@ struct cpu_map {
 struct cpu_map *cpu_map__new(const char *cpu_list);
 struct cpu_map *cpu_map__dummy_new(void);
 void cpu_map__delete(struct cpu_map *map);
-
+struct cpu_map *cpu_map__read(FILE *file);
 size_t cpu_map__fprintf(struct cpu_map *map, FILE *fp);
 
 #endif /* __PERF_CPUMAP_H */
diff --git a/tools/perf/util/evsel.h b/tools/perf/util/evsel.h
index dc40fe3..93876ba 100644
--- a/tools/perf/util/evsel.h
+++ b/tools/perf/util/evsel.h
@@ -66,6 +66,7 @@ struct perf_evsel {
 		void		*func;
 		void		*data;
 	} handler;
+	struct cpu_map		*cpus;
 	unsigned int		sample_size;
 	bool 			supported;
 	/* parse modifier helper */
diff --git a/tools/perf/util/parse-events.c b/tools/perf/util/parse-events.c
index 44afcf4..bf5d033 100644
--- a/tools/perf/util/parse-events.c
+++ b/tools/perf/util/parse-events.c
@@ -239,8 +239,11 @@ const char *event_type(int type)
 	return "unknown";
 }
 
-static int add_event(struct list_head **_list, int *idx,
-		     struct perf_event_attr *attr, char *name)
+
+
+static int __add_event(struct list_head **_list, int *idx,
+		       struct perf_event_attr *attr,
+		       char *name, struct cpu_map *cpus)
 {
 	struct perf_evsel *evsel;
 	struct list_head *list = *_list;
@@ -260,6 +263,7 @@ static int add_event(struct list_head **_list, int *idx,
 		return -ENOMEM;
 	}
 
+	evsel->cpus = cpus;
 	if (name)
 		evsel->name = strdup(name);
 	list_add_tail(&evsel->node, list);
@@ -267,6 +271,12 @@ static int add_event(struct list_head **_list, int *idx,
 	return 0;
 }
 
+static int add_event(struct list_head **_list, int *idx,
+		     struct perf_event_attr *attr, char *name)
+{
+	return __add_event(_list, idx, attr, name, NULL);
+}
+
 static int parse_aliases(char *str, const char *names[][PERF_EVSEL__MAX_ALIASES], int size)
 {
 	int i, j;
@@ -607,8 +617,8 @@ int parse_events_add_pmu(struct list_head **list, int *idx,
 	if (perf_pmu__config(pmu, &attr, head_config))
 		return -EINVAL;
 
-	return add_event(list, idx, &attr,
-			 pmu_event_name(head_config));
+	return __add_event(list, idx, &attr, pmu_event_name(head_config),
+			   pmu->cpus);
 }
 
 int parse_events__modifier_group(struct list_head *list,
diff --git a/tools/perf/util/pmu.c b/tools/perf/util/pmu.c
index 6631d82..8a2229d 100644
--- a/tools/perf/util/pmu.c
+++ b/tools/perf/util/pmu.c
@@ -9,6 +9,7 @@
 #include "util.h"
 #include "pmu.h"
 #include "parse-events.h"
+#include "cpumap.h"
 
 #define EVENT_SOURCE_DEVICE_PATH "/bus/event_source/devices/"
 
@@ -253,6 +254,33 @@ static void pmu_read_sysfs(void)
 	closedir(dir);
 }
 
+static struct cpu_map *pmu_cpumask(char *name)
+{
+	struct stat st;
+	char path[PATH_MAX];
+	const char *sysfs;
+	FILE *file;
+	struct cpu_map *cpus;
+
+	sysfs = sysfs_find_mountpoint();
+	if (!sysfs)
+		return NULL;
+
+	snprintf(path, PATH_MAX,
+		 "%s/bus/event_source/devices/%s/cpumask", sysfs, name);
+
+	if (stat(path, &st) < 0)
+		return NULL;
+
+	file = fopen(path, "r");
+	if (!file)
+		return NULL;
+
+	cpus = cpu_map__read(file);
+	fclose(file);
+	return cpus;
+}
+
 static struct perf_pmu *pmu_lookup(char *name)
 {
 	struct perf_pmu *pmu;
@@ -275,6 +303,8 @@ static struct perf_pmu *pmu_lookup(char *name)
 	if (!pmu)
 		return NULL;
 
+	pmu->cpus = pmu_cpumask(name);
+
 	pmu_aliases(name, &aliases);
 
 	INIT_LIST_HEAD(&pmu->format);
diff --git a/tools/perf/util/pmu.h b/tools/perf/util/pmu.h
index 47f68d3..53c7794 100644
--- a/tools/perf/util/pmu.h
+++ b/tools/perf/util/pmu.h
@@ -28,6 +28,7 @@ struct perf_pmu__alias {
 struct perf_pmu {
 	char *name;
 	__u32 type;
+	struct cpu_map *cpus;
 	struct list_head format;
 	struct list_head aliases;
 	struct list_head list;
-- 
1.7.1