[RFC 0/6] perf stat: Metrics calculation fix

[RFC 0/6] perf stat: Metrics calculation fix

[Jiri Olsa]

hi,
I'm sending Andi's patch accompanied with changes requested by Ingo.
  http://marc.info/?l=linux-kernel&m=139748629929175&w=2

I added simple bit calculated index array, while exclude_* bits
might have more strict logic and ban some combinations, but I
wanted to keep it simple and did not find better solution.

thanks,
jirka


---
Andi Kleen (1):
      perf stat: Fix metrics calculation with event qualifiers

Jiri Olsa (5):
      perf tools: Add 'I' event modifier for exclude_idle bit
      perf stat: Change metrics context calculation
      perf stat: Add metrics support for exclude_hv
      perf stat: Add metrics support for exclude_(host|guest)
      perf stat: Add metrics support for exclude_idle

 tools/perf/Documentation/perf-list.txt |   1 +
 tools/perf/builtin-stat.c              | 144 +++++++++++++++++++++++++++++++++++++++++------------------------
 tools/perf/tests/parse-events.c        |  40 ++++++++++++++++++
 tools/perf/util/parse-events.c         |   8 +++-
 tools/perf/util/parse-events.l         |   2 +-
 5 files changed, 141 insertions(+), 54 deletions(-)

[PATCH 1/6] perf tools: Add 'I' event modifier for exclude_idle bit

[Jiri Olsa]

Adding 'I' event modifier to have complete set of modifiers
for perf_event_attr:exclude_* bits.

Any event specified with 'I' modifier will have the
perf_event_attr:exclude_idle bit set.

  $ perf record -e cycles:I -vv ls 2>&1 | grep exclude_idle
  exclude_hv          0    exclude_idle        1

Adding automated tests.

Link: http://lkml.kernel.org/n/tip-6dx0y47hkv7c35rmz9c1u8l2@git.kernel.org
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
---
 tools/perf/Documentation/perf-list.txt |  1 +
 tools/perf/tests/parse-events.c        | 40 ++++++++++++++++++++++++++++++++++
 tools/perf/util/parse-events.c         |  8 ++++++-
 tools/perf/util/parse-events.l         |  2 +-
 4 files changed, 49 insertions(+), 2 deletions(-)

diff --git a/tools/perf/Documentation/perf-list.txt b/tools/perf/Documentation/perf-list.txt
index 4692d277980b..bada8933fdd4 100644
--- a/tools/perf/Documentation/perf-list.txt
+++ b/tools/perf/Documentation/perf-list.txt
@@ -26,6 +26,7 @@ counted. The following modifiers exist:
  u - user-space counting
  k - kernel counting
  h - hypervisor counting
+ I - non idle counting
  G - guest counting (in KVM guests)
  H - host counting (not in KVM guests)
  p - precise level
diff --git a/tools/perf/tests/parse-events.c b/tools/perf/tests/parse-events.c
index ac243ebcb20a..3de744961739 100644
--- a/tools/perf/tests/parse-events.c
+++ b/tools/perf/tests/parse-events.c
@@ -295,6 +295,36 @@ static int test__checkevent_genhw_modifier(struct perf_evlist *evlist)
 	return test__checkevent_genhw(evlist);
 }
 
+static int test__checkevent_exclude_idle_modifier(struct perf_evlist *evlist)
+{
+	struct perf_evsel *evsel = perf_evlist__first(evlist);
+
+	TEST_ASSERT_VAL("wrong exclude idle", evsel->attr.exclude_idle);
+	TEST_ASSERT_VAL("wrong exclude guest", !evsel->attr.exclude_guest);
+	TEST_ASSERT_VAL("wrong exclude host", !evsel->attr.exclude_host);
+	TEST_ASSERT_VAL("wrong exclude_user", !evsel->attr.exclude_user);
+	TEST_ASSERT_VAL("wrong exclude_kernel", !evsel->attr.exclude_kernel);
+	TEST_ASSERT_VAL("wrong exclude_hv", !evsel->attr.exclude_hv);
+	TEST_ASSERT_VAL("wrong precise_ip", !evsel->attr.precise_ip);
+
+	return test__checkevent_symbolic_name(evlist);
+}
+
+static int test__checkevent_exclude_idle_modifier_1(struct perf_evlist *evlist)
+{
+	struct perf_evsel *evsel = perf_evlist__first(evlist);
+
+	TEST_ASSERT_VAL("wrong exclude idle", evsel->attr.exclude_idle);
+	TEST_ASSERT_VAL("wrong exclude guest", !evsel->attr.exclude_guest);
+	TEST_ASSERT_VAL("wrong exclude host", evsel->attr.exclude_host);
+	TEST_ASSERT_VAL("wrong exclude_user", evsel->attr.exclude_user);
+	TEST_ASSERT_VAL("wrong exclude_kernel", !evsel->attr.exclude_kernel);
+	TEST_ASSERT_VAL("wrong exclude_hv", evsel->attr.exclude_hv);
+	TEST_ASSERT_VAL("wrong precise_ip", !evsel->attr.precise_ip);
+
+	return test__checkevent_symbolic_name(evlist);
+}
+
 static int test__checkevent_breakpoint_modifier(struct perf_evlist *evlist)
 {
 	struct perf_evsel *evsel = perf_evlist__first(evlist);
@@ -1494,6 +1524,16 @@ static struct evlist_test test__events[] = {
 		.id    = 100,
 	},
 #endif
+	{
+		.name  = "instructions:I",
+		.check = test__checkevent_exclude_idle_modifier,
+		.id    = 45,
+	},
+	{
+		.name  = "instructions:kIG",
+		.check = test__checkevent_exclude_idle_modifier_1,
+		.id    = 46,
+	},
 };
 
 static struct evlist_test test__events_pmu[] = {
diff --git a/tools/perf/util/parse-events.c b/tools/perf/util/parse-events.c
index fe07573d5ed4..be0655388b38 100644
--- a/tools/perf/util/parse-events.c
+++ b/tools/perf/util/parse-events.c
@@ -709,6 +709,7 @@ struct event_modifier {
 	int eh;
 	int eH;
 	int eG;
+	int eI;
 	int precise;
 	int exclude_GH;
 	int sample_read;
@@ -723,6 +724,7 @@ static int get_event_modifier(struct event_modifier *mod, char *str,
 	int eh = evsel ? evsel->attr.exclude_hv : 0;
 	int eH = evsel ? evsel->attr.exclude_host : 0;
 	int eG = evsel ? evsel->attr.exclude_guest : 0;
+	int eI = evsel ? evsel->attr.exclude_idle : 0;
 	int precise = evsel ? evsel->attr.precise_ip : 0;
 	int sample_read = 0;
 	int pinned = evsel ? evsel->attr.pinned : 0;
@@ -753,6 +755,8 @@ static int get_event_modifier(struct event_modifier *mod, char *str,
 			if (!exclude_GH)
 				exclude_GH = eG = eH = 1;
 			eH = 0;
+		} else if (*str == 'I') {
+			eI = 1;
 		} else if (*str == 'p') {
 			precise++;
 			/* use of precise requires exclude_guest */
@@ -786,6 +790,7 @@ static int get_event_modifier(struct event_modifier *mod, char *str,
 	mod->eh = eh;
 	mod->eH = eH;
 	mod->eG = eG;
+	mod->eI = eI;
 	mod->precise = precise;
 	mod->exclude_GH = exclude_GH;
 	mod->sample_read = sample_read;
@@ -803,7 +808,7 @@ static int check_modifier(char *str)
 	char *p = str;
 
 	/* The sizeof includes 0 byte as well. */
-	if (strlen(str) > (sizeof("ukhGHpppSD") - 1))
+	if (strlen(str) > (sizeof("ukhGHpppSDI") - 1))
 		return -1;
 
 	while (*p) {
@@ -839,6 +844,7 @@ int parse_events__modifier_event(struct list_head *list, char *str, bool add)
 		evsel->attr.precise_ip     = mod.precise;
 		evsel->attr.exclude_host   = mod.eH;
 		evsel->attr.exclude_guest  = mod.eG;
+		evsel->attr.exclude_idle   = mod.eI;
 		evsel->exclude_GH          = mod.exclude_GH;
 		evsel->sample_read         = mod.sample_read;
 
diff --git a/tools/perf/util/parse-events.l b/tools/perf/util/parse-events.l
index 94eacb6c1ef7..8895cf3132ab 100644
--- a/tools/perf/util/parse-events.l
+++ b/tools/perf/util/parse-events.l
@@ -101,7 +101,7 @@ num_raw_hex	[a-fA-F0-9]+
 name		[a-zA-Z_*?][a-zA-Z0-9_*?]*
 name_minus	[a-zA-Z_*?][a-zA-Z0-9\-_*?]*
 /* If you add a modifier you need to update check_modifier() */
-modifier_event	[ukhpGHSD]+
+modifier_event	[ukhpGHSDI]+
 modifier_bp	[rwx]{1,3}
 
 %%
-- 
1.9.3

[PATCH 2/6] perf stat: Fix metrics calculation with event qualifiers

[Jiri Olsa]

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

Currently in perf IPC and other metrics cannot be directly shown
separately for both user and kernel in a single run. The problem was
that the metrics matching code did not check event qualifiers.

With this patch the following case works correctly.

% perf stat -e cycles:k,cycles:u,instructions:k,instructions:u true

 Performance counter stats for 'true':

           531,718      cycles:k
           203,895      cycles:u
           338,151      instructions:k            #    0.64  insns per cycle
           105,961      instructions:u            #    0.52  insns per cycle

       0.002989739 seconds time elapsed

Previously it would misreport the ratios because they were matching
the wrong value.

The patch is fairly big, but quite mechanic as it just
adds context indexes everywhere.

Reported-by: William Cohen <wcohen@redhat.com>
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Link: http://lkml.kernel.org/n/tip-pfvwqm9cppgn45m90punnmib@git.kernel.org
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
---
 tools/perf/builtin-stat.c | 129 +++++++++++++++++++++++++++-------------------
 1 file changed, 77 insertions(+), 52 deletions(-)

diff --git a/tools/perf/builtin-stat.c b/tools/perf/builtin-stat.c
index f7b8218785f6..0b163ca4034f 100644
--- a/tools/perf/builtin-stat.c
+++ b/tools/perf/builtin-stat.c
@@ -247,21 +247,35 @@ out_free:
 	return -1;
 }
 
+#define NUM_CTX 3
+
+enum { CTX_USER, CTX_KERNEL, CTX_ALL };
+
 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];
-static struct stats runtime_stalled_cycles_back_stats[MAX_NR_CPUS];
-static struct stats runtime_branches_stats[MAX_NR_CPUS];
-static struct stats runtime_cacherefs_stats[MAX_NR_CPUS];
-static struct stats runtime_l1_dcache_stats[MAX_NR_CPUS];
-static struct stats runtime_l1_icache_stats[MAX_NR_CPUS];
-static struct stats runtime_ll_cache_stats[MAX_NR_CPUS];
-static struct stats runtime_itlb_cache_stats[MAX_NR_CPUS];
-static struct stats runtime_dtlb_cache_stats[MAX_NR_CPUS];
-static struct stats runtime_cycles_in_tx_stats[MAX_NR_CPUS];
+static struct stats runtime_cycles_stats[NUM_CTX][MAX_NR_CPUS];
+static struct stats runtime_stalled_cycles_front_stats[NUM_CTX][MAX_NR_CPUS];
+static struct stats runtime_stalled_cycles_back_stats[NUM_CTX][MAX_NR_CPUS];
+static struct stats runtime_branches_stats[NUM_CTX][MAX_NR_CPUS];
+static struct stats runtime_cacherefs_stats[NUM_CTX][MAX_NR_CPUS];
+static struct stats runtime_l1_dcache_stats[NUM_CTX][MAX_NR_CPUS];
+static struct stats runtime_l1_icache_stats[NUM_CTX][MAX_NR_CPUS];
+static struct stats runtime_ll_cache_stats[NUM_CTX][MAX_NR_CPUS];
+static struct stats runtime_itlb_cache_stats[NUM_CTX][MAX_NR_CPUS];
+static struct stats runtime_dtlb_cache_stats[NUM_CTX][MAX_NR_CPUS];
+static struct stats runtime_cycles_in_tx_stats[NUM_CTX][MAX_NR_CPUS];
 static struct stats walltime_nsecs_stats;
-static struct stats runtime_transaction_stats[MAX_NR_CPUS];
-static struct stats runtime_elision_stats[MAX_NR_CPUS];
+static struct stats runtime_transaction_stats[NUM_CTX][MAX_NR_CPUS];
+static struct stats runtime_elision_stats[NUM_CTX][MAX_NR_CPUS];
+
+static int evsel_context(struct perf_evsel *evsel)
+{
+	if (evsel->attr.exclude_kernel)
+		return CTX_USER;
+	if (evsel->attr.exclude_user)
+		return CTX_KERNEL;
+	/* Handle hypervisor too? */
+	return CTX_ALL;
+}
 
 static void perf_stat__reset_stats(struct perf_evlist *evlist)
 {
@@ -356,37 +370,39 @@ static struct perf_evsel *nth_evsel(int n)
 static void update_shadow_stats(struct perf_evsel *counter, u64 *count,
 				int cpu)
 {
+	int ctx = evsel_context(counter);
+
 	if (perf_evsel__match(counter, SOFTWARE, SW_TASK_CLOCK))
 		update_stats(&runtime_nsecs_stats[cpu], count[0]);
 	else if (perf_evsel__match(counter, HARDWARE, HW_CPU_CYCLES))
-		update_stats(&runtime_cycles_stats[cpu], count[0]);
+		update_stats(&runtime_cycles_stats[ctx][cpu], count[0]);
 	else if (transaction_run &&
 		 perf_evsel__cmp(counter, nth_evsel(T_CYCLES_IN_TX)))
-		update_stats(&runtime_cycles_in_tx_stats[cpu], count[0]);
+		update_stats(&runtime_transaction_stats[ctx][cpu], count[0]);
 	else if (transaction_run &&
 		 perf_evsel__cmp(counter, nth_evsel(T_TRANSACTION_START)))
-		update_stats(&runtime_transaction_stats[cpu], count[0]);
+		update_stats(&runtime_transaction_stats[ctx][cpu], count[0]);
 	else if (transaction_run &&
 		 perf_evsel__cmp(counter, nth_evsel(T_ELISION_START)))
-		update_stats(&runtime_elision_stats[cpu], count[0]);
+		update_stats(&runtime_elision_stats[ctx][cpu], count[0]);
 	else if (perf_evsel__match(counter, HARDWARE, HW_STALLED_CYCLES_FRONTEND))
-		update_stats(&runtime_stalled_cycles_front_stats[cpu], count[0]);
+		update_stats(&runtime_stalled_cycles_front_stats[ctx][cpu], count[0]);
 	else if (perf_evsel__match(counter, HARDWARE, HW_STALLED_CYCLES_BACKEND))
-		update_stats(&runtime_stalled_cycles_back_stats[cpu], count[0]);
+		update_stats(&runtime_stalled_cycles_back_stats[ctx][cpu], count[0]);
 	else if (perf_evsel__match(counter, HARDWARE, HW_BRANCH_INSTRUCTIONS))
-		update_stats(&runtime_branches_stats[cpu], count[0]);
+		update_stats(&runtime_branches_stats[ctx][cpu], count[0]);
 	else if (perf_evsel__match(counter, HARDWARE, HW_CACHE_REFERENCES))
-		update_stats(&runtime_cacherefs_stats[cpu], count[0]);
+		update_stats(&runtime_cacherefs_stats[ctx][cpu], count[0]);
 	else if (perf_evsel__match(counter, HW_CACHE, HW_CACHE_L1D))
-		update_stats(&runtime_l1_dcache_stats[cpu], count[0]);
+		update_stats(&runtime_l1_dcache_stats[ctx][cpu], count[0]);
 	else if (perf_evsel__match(counter, HW_CACHE, HW_CACHE_L1I))
-		update_stats(&runtime_l1_icache_stats[cpu], count[0]);
+		update_stats(&runtime_ll_cache_stats[ctx][cpu], count[0]);
 	else if (perf_evsel__match(counter, HW_CACHE, HW_CACHE_LL))
-		update_stats(&runtime_ll_cache_stats[cpu], count[0]);
+		update_stats(&runtime_ll_cache_stats[ctx][cpu], count[0]);
 	else if (perf_evsel__match(counter, HW_CACHE, HW_CACHE_DTLB))
-		update_stats(&runtime_dtlb_cache_stats[cpu], count[0]);
+		update_stats(&runtime_dtlb_cache_stats[ctx][cpu], count[0]);
 	else if (perf_evsel__match(counter, HW_CACHE, HW_CACHE_ITLB))
-		update_stats(&runtime_itlb_cache_stats[cpu], count[0]);
+		update_stats(&runtime_itlb_cache_stats[ctx][cpu], count[0]);
 }
 
 static void zero_per_pkg(struct perf_evsel *counter)
@@ -908,8 +924,9 @@ static void print_stalled_cycles_frontend(int cpu,
 {
 	double total, ratio = 0.0;
 	const char *color;
+	int ctx = evsel_context(evsel);
 
-	total = avg_stats(&runtime_cycles_stats[cpu]);
+	total = avg_stats(&runtime_cycles_stats[ctx][cpu]);
 
 	if (total)
 		ratio = avg / total * 100.0;
@@ -927,8 +944,9 @@ static void print_stalled_cycles_backend(int cpu,
 {
 	double total, ratio = 0.0;
 	const char *color;
+	int ctx = evsel_context(evsel);
 
-	total = avg_stats(&runtime_cycles_stats[cpu]);
+	total = avg_stats(&runtime_cycles_stats[ctx][cpu]);
 
 	if (total)
 		ratio = avg / total * 100.0;
@@ -946,8 +964,9 @@ static void print_branch_misses(int cpu,
 {
 	double total, ratio = 0.0;
 	const char *color;
+	int ctx = evsel_context(evsel);
 
-	total = avg_stats(&runtime_branches_stats[cpu]);
+	total = avg_stats(&runtime_branches_stats[ctx][cpu]);
 
 	if (total)
 		ratio = avg / total * 100.0;
@@ -965,8 +984,9 @@ static void print_l1_dcache_misses(int cpu,
 {
 	double total, ratio = 0.0;
 	const char *color;
+	int ctx = evsel_context(evsel);
 
-	total = avg_stats(&runtime_l1_dcache_stats[cpu]);
+	total = avg_stats(&runtime_l1_dcache_stats[ctx][cpu]);
 
 	if (total)
 		ratio = avg / total * 100.0;
@@ -984,8 +1004,9 @@ static void print_l1_icache_misses(int cpu,
 {
 	double total, ratio = 0.0;
 	const char *color;
+	int ctx = evsel_context(evsel);
 
-	total = avg_stats(&runtime_l1_icache_stats[cpu]);
+	total = avg_stats(&runtime_l1_icache_stats[ctx][cpu]);
 
 	if (total)
 		ratio = avg / total * 100.0;
@@ -1003,8 +1024,9 @@ static void print_dtlb_cache_misses(int cpu,
 {
 	double total, ratio = 0.0;
 	const char *color;
+	int ctx = evsel_context(evsel);
 
-	total = avg_stats(&runtime_dtlb_cache_stats[cpu]);
+	total = avg_stats(&runtime_dtlb_cache_stats[ctx][cpu]);
 
 	if (total)
 		ratio = avg / total * 100.0;
@@ -1022,8 +1044,9 @@ static void print_itlb_cache_misses(int cpu,
 {
 	double total, ratio = 0.0;
 	const char *color;
+	int ctx = evsel_context(evsel);
 
-	total = avg_stats(&runtime_itlb_cache_stats[cpu]);
+	total = avg_stats(&runtime_itlb_cache_stats[ctx][cpu]);
 
 	if (total)
 		ratio = avg / total * 100.0;
@@ -1041,8 +1064,9 @@ static void print_ll_cache_misses(int cpu,
 {
 	double total, ratio = 0.0;
 	const char *color;
+	int ctx = evsel_context(evsel);
 
-	total = avg_stats(&runtime_ll_cache_stats[cpu]);
+	total = avg_stats(&runtime_ll_cache_stats[ctx][cpu]);
 
 	if (total)
 		ratio = avg / total * 100.0;
@@ -1060,6 +1084,7 @@ static void abs_printout(int id, int nr, struct perf_evsel *evsel, double avg)
 	double sc =  evsel->scale;
 	const char *fmt;
 	int cpu = cpu_map__id_to_cpu(id);
+	int ctx = evsel_context(evsel);
 
 	if (csv_output) {
 		fmt = sc != 1.0 ?  "%.2f%s" : "%.0f%s";
@@ -1091,15 +1116,15 @@ static void abs_printout(int id, int nr, struct perf_evsel *evsel, double avg)
 		return;
 
 	if (perf_evsel__match(evsel, HARDWARE, HW_INSTRUCTIONS)) {
-		total = avg_stats(&runtime_cycles_stats[cpu]);
+		total = avg_stats(&runtime_cycles_stats[ctx][cpu]);
 		if (total) {
 			ratio = avg / total;
 			fprintf(output, " #   %5.2f  insns per cycle        ", ratio);
 		} else {
 			fprintf(output, "                                   ");
 		}
-		total = avg_stats(&runtime_stalled_cycles_front_stats[cpu]);
-		total = max(total, avg_stats(&runtime_stalled_cycles_back_stats[cpu]));
+		total = avg_stats(&runtime_stalled_cycles_front_stats[ctx][cpu]);
+		total = max(total, avg_stats(&runtime_stalled_cycles_back_stats[ctx][cpu]));
 
 		if (total && avg) {
 			ratio = total / avg;
@@ -1110,46 +1135,46 @@ static void abs_printout(int id, int nr, struct perf_evsel *evsel, double avg)
 		}
 
 	} else if (perf_evsel__match(evsel, HARDWARE, HW_BRANCH_MISSES) &&
-			runtime_branches_stats[cpu].n != 0) {
+			runtime_branches_stats[ctx][cpu].n != 0) {
 		print_branch_misses(cpu, evsel, avg);
 	} else if (
 		evsel->attr.type == PERF_TYPE_HW_CACHE &&
 		evsel->attr.config ==  ( PERF_COUNT_HW_CACHE_L1D |
 					((PERF_COUNT_HW_CACHE_OP_READ) << 8) |
 					((PERF_COUNT_HW_CACHE_RESULT_MISS) << 16)) &&
-			runtime_l1_dcache_stats[cpu].n != 0) {
+			runtime_l1_dcache_stats[ctx][cpu].n != 0) {
 		print_l1_dcache_misses(cpu, evsel, avg);
 	} else if (
 		evsel->attr.type == PERF_TYPE_HW_CACHE &&
 		evsel->attr.config ==  ( PERF_COUNT_HW_CACHE_L1I |
 					((PERF_COUNT_HW_CACHE_OP_READ) << 8) |
 					((PERF_COUNT_HW_CACHE_RESULT_MISS) << 16)) &&
-			runtime_l1_icache_stats[cpu].n != 0) {
+			runtime_l1_icache_stats[ctx][cpu].n != 0) {
 		print_l1_icache_misses(cpu, evsel, avg);
 	} else if (
 		evsel->attr.type == PERF_TYPE_HW_CACHE &&
 		evsel->attr.config ==  ( PERF_COUNT_HW_CACHE_DTLB |
 					((PERF_COUNT_HW_CACHE_OP_READ) << 8) |
 					((PERF_COUNT_HW_CACHE_RESULT_MISS) << 16)) &&
-			runtime_dtlb_cache_stats[cpu].n != 0) {
+			runtime_dtlb_cache_stats[ctx][cpu].n != 0) {
 		print_dtlb_cache_misses(cpu, evsel, avg);
 	} else if (
 		evsel->attr.type == PERF_TYPE_HW_CACHE &&
 		evsel->attr.config ==  ( PERF_COUNT_HW_CACHE_ITLB |
 					((PERF_COUNT_HW_CACHE_OP_READ) << 8) |
 					((PERF_COUNT_HW_CACHE_RESULT_MISS) << 16)) &&
-			runtime_itlb_cache_stats[cpu].n != 0) {
+			runtime_itlb_cache_stats[ctx][cpu].n != 0) {
 		print_itlb_cache_misses(cpu, evsel, avg);
 	} else if (
 		evsel->attr.type == PERF_TYPE_HW_CACHE &&
 		evsel->attr.config ==  ( PERF_COUNT_HW_CACHE_LL |
 					((PERF_COUNT_HW_CACHE_OP_READ) << 8) |
 					((PERF_COUNT_HW_CACHE_RESULT_MISS) << 16)) &&
-			runtime_ll_cache_stats[cpu].n != 0) {
+			runtime_ll_cache_stats[ctx][cpu].n != 0) {
 		print_ll_cache_misses(cpu, evsel, avg);
 	} else if (perf_evsel__match(evsel, HARDWARE, HW_CACHE_MISSES) &&
-			runtime_cacherefs_stats[cpu].n != 0) {
-		total = avg_stats(&runtime_cacherefs_stats[cpu]);
+			runtime_cacherefs_stats[ctx][cpu].n != 0) {
+		total = avg_stats(&runtime_cacherefs_stats[ctx][cpu]);
 
 		if (total)
 			ratio = avg * 100 / total;
@@ -1171,15 +1196,15 @@ static void abs_printout(int id, int nr, struct perf_evsel *evsel, double avg)
 		}
 	} else if (transaction_run &&
 		   perf_evsel__cmp(evsel, nth_evsel(T_CYCLES_IN_TX))) {
-		total = avg_stats(&runtime_cycles_stats[cpu]);
+		total = avg_stats(&runtime_cycles_stats[ctx][cpu]);
 		if (total)
 			fprintf(output,
 				" #   %5.2f%% transactional cycles   ",
 				100.0 * (avg / total));
 	} else if (transaction_run &&
 		   perf_evsel__cmp(evsel, nth_evsel(T_CYCLES_IN_TX_CP))) {
-		total = avg_stats(&runtime_cycles_stats[cpu]);
-		total2 = avg_stats(&runtime_cycles_in_tx_stats[cpu]);
+		total = avg_stats(&runtime_cycles_stats[ctx][cpu]);
+		total2 = avg_stats(&runtime_cycles_in_tx_stats[ctx][cpu]);
 		if (total2 < avg)
 			total2 = avg;
 		if (total)
@@ -1189,8 +1214,8 @@ static void abs_printout(int id, int nr, struct perf_evsel *evsel, double avg)
 	} else if (transaction_run &&
 		   perf_evsel__cmp(evsel, nth_evsel(T_TRANSACTION_START)) &&
 		   avg > 0 &&
-		   runtime_cycles_in_tx_stats[cpu].n != 0) {
-		total = avg_stats(&runtime_cycles_in_tx_stats[cpu]);
+		   runtime_cycles_in_tx_stats[ctx][cpu].n != 0) {
+		total = avg_stats(&runtime_cycles_in_tx_stats[ctx][cpu]);
 
 		if (total)
 			ratio = total / avg;
@@ -1199,8 +1224,8 @@ static void abs_printout(int id, int nr, struct perf_evsel *evsel, double avg)
 	} else if (transaction_run &&
 		   perf_evsel__cmp(evsel, nth_evsel(T_ELISION_START)) &&
 		   avg > 0 &&
-		   runtime_cycles_in_tx_stats[cpu].n != 0) {
-		total = avg_stats(&runtime_cycles_in_tx_stats[cpu]);
+		   runtime_cycles_in_tx_stats[ctx][cpu].n != 0) {
+		total = avg_stats(&runtime_cycles_in_tx_stats[ctx][cpu]);
 
 		if (total)
 			ratio = total / avg;
-- 
1.9.3

[PATCH 3/6] perf stat: Change metrics context calculation

[Jiri Olsa]

Changing metrics context calculation to allow more than
2 types of context.

Following patches will add support for the rest of the
exclude_* bits so we need separate array element for
all context combinations.

Link: http://lkml.kernel.org/n/tip-6mj0tn8z60ozof19xd80y6pe@git.kernel.org
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
---
 tools/perf/builtin-stat.c | 17 +++++++++++------
 1 file changed, 11 insertions(+), 6 deletions(-)

diff --git a/tools/perf/builtin-stat.c b/tools/perf/builtin-stat.c
index 0b163ca4034f..0c748dda0692 100644
--- a/tools/perf/builtin-stat.c
+++ b/tools/perf/builtin-stat.c
@@ -247,9 +247,13 @@ out_free:
 	return -1;
 }
 
-#define NUM_CTX 3
+enum {
+	CTX_BIT_USER	= 1 << 0,
+	CTX_BIT_KERNEL	= 1 << 1,
+	CTX_BIT_MAX	= 1 << 2,
+};
 
-enum { CTX_USER, CTX_KERNEL, CTX_ALL };
+#define NUM_CTX CTX_BIT_MAX
 
 static struct stats runtime_nsecs_stats[MAX_NR_CPUS];
 static struct stats runtime_cycles_stats[NUM_CTX][MAX_NR_CPUS];
@@ -269,12 +273,13 @@ static struct stats runtime_elision_stats[NUM_CTX][MAX_NR_CPUS];
 
 static int evsel_context(struct perf_evsel *evsel)
 {
+	int ctx = 0;
+
 	if (evsel->attr.exclude_kernel)
-		return CTX_USER;
+		ctx |= CTX_BIT_KERNEL;
 	if (evsel->attr.exclude_user)
-		return CTX_KERNEL;
-	/* Handle hypervisor too? */
-	return CTX_ALL;
+		ctx |= CTX_BIT_USER;
+	return ctx;
 }
 
 static void perf_stat__reset_stats(struct perf_evlist *evlist)
-- 
1.9.3

[PATCH 4/6] perf stat: Add metrics support for exclude_hv

[Jiri Olsa]

Separating metrics values for exclude_hv bit.

Link: http://lkml.kernel.org/n/tip-6lbwnpn4c49s4k9l6o5qc873@git.kernel.org
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
---
 tools/perf/builtin-stat.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/tools/perf/builtin-stat.c b/tools/perf/builtin-stat.c
index 0c748dda0692..ab5786b8cfdb 100644
--- a/tools/perf/builtin-stat.c
+++ b/tools/perf/builtin-stat.c
@@ -250,7 +250,8 @@ out_free:
 enum {
 	CTX_BIT_USER	= 1 << 0,
 	CTX_BIT_KERNEL	= 1 << 1,
-	CTX_BIT_MAX	= 1 << 2,
+	CTX_BIT_HV	= 1 << 2,
+	CTX_BIT_MAX	= 1 << 3,
 };
 
 #define NUM_CTX CTX_BIT_MAX
@@ -279,6 +280,8 @@ static int evsel_context(struct perf_evsel *evsel)
 		ctx |= CTX_BIT_KERNEL;
 	if (evsel->attr.exclude_user)
 		ctx |= CTX_BIT_USER;
+	if (evsel->attr.exclude_hv)
+		ctx |= CTX_BIT_HV;
 	return ctx;
 }
 
-- 
1.9.3

[PATCH 5/6] perf stat: Add metrics support for exclude_(host|guest)

[Jiri Olsa]

Separating metrics values for guest and host,
so we get proper values.

Link: http://lkml.kernel.org/n/tip-mqd77j3sfke60h8s45c7rqe0@git.kernel.org
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
---
 tools/perf/builtin-stat.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)

diff --git a/tools/perf/builtin-stat.c b/tools/perf/builtin-stat.c
index ab5786b8cfdb..ab8d7e6c0436 100644
--- a/tools/perf/builtin-stat.c
+++ b/tools/perf/builtin-stat.c
@@ -251,7 +251,8 @@ enum {
 	CTX_BIT_USER	= 1 << 0,
 	CTX_BIT_KERNEL	= 1 << 1,
 	CTX_BIT_HV	= 1 << 2,
-	CTX_BIT_MAX	= 1 << 3,
+	CTX_BIT_HOST	= 1 << 3,
+	CTX_BIT_MAX	= 1 << 4,
 };
 
 #define NUM_CTX CTX_BIT_MAX
@@ -282,6 +283,9 @@ static int evsel_context(struct perf_evsel *evsel)
 		ctx |= CTX_BIT_USER;
 	if (evsel->attr.exclude_hv)
 		ctx |= CTX_BIT_HV;
+	if (evsel->attr.exclude_host)
+		ctx |= CTX_BIT_HOST;
+
 	return ctx;
 }
 
-- 
1.9.3

[PATCH 6/6] perf stat: Add metrics support for exclude_idle

[Jiri Olsa]

Separating metrics values for exclude_idle bit.

Link: http://lkml.kernel.org/n/tip-ce8su75qdqve2i7qxy2usx9k@git.kernel.org
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
---
 tools/perf/builtin-stat.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/tools/perf/builtin-stat.c b/tools/perf/builtin-stat.c
index ab8d7e6c0436..fec089f1c364 100644
--- a/tools/perf/builtin-stat.c
+++ b/tools/perf/builtin-stat.c
@@ -252,7 +252,8 @@ enum {
 	CTX_BIT_KERNEL	= 1 << 1,
 	CTX_BIT_HV	= 1 << 2,
 	CTX_BIT_HOST	= 1 << 3,
-	CTX_BIT_MAX	= 1 << 4,
+	CTX_BIT_IDLE	= 1 << 4,
+	CTX_BIT_MAX	= 1 << 5,
 };
 
 #define NUM_CTX CTX_BIT_MAX
@@ -285,6 +286,8 @@ static int evsel_context(struct perf_evsel *evsel)
 		ctx |= CTX_BIT_HV;
 	if (evsel->attr.exclude_host)
 		ctx |= CTX_BIT_HOST;
+	if (evsel->attr.exclude_idle)
+		ctx |= CTX_BIT_IDLE;
 
 	return ctx;
 }
-- 
1.9.3

Re: [RFC 0/6] perf stat: Metrics calculation fix

[Jiri Olsa]

On Tue, Apr 07, 2015 at 11:25:13PM +0200, Jiri Olsa wrote:
> hi,
> I'm sending Andi's patch accompanied with changes requested by Ingo.
>   http://marc.info/?l=linux-kernel&m=139748629929175&w=2
> 
> I added simple bit calculated index array, while exclude_* bits
> might have more strict logic and ban some combinations, but I
> wanted to keep it simple and did not find better solution.

available in here:

git://git.kernel.org/pub/scm/linux/kernel/git/jolsa/perf.git
perf/andi

> 
> thanks,
> jirka
> 
> 
> ---
> Andi Kleen (1):
>       perf stat: Fix metrics calculation with event qualifiers
> 
> Jiri Olsa (5):
>       perf tools: Add 'I' event modifier for exclude_idle bit
>       perf stat: Change metrics context calculation
>       perf stat: Add metrics support for exclude_hv
>       perf stat: Add metrics support for exclude_(host|guest)
>       perf stat: Add metrics support for exclude_idle
> 
>  tools/perf/Documentation/perf-list.txt |   1 +
>  tools/perf/builtin-stat.c              | 144 +++++++++++++++++++++++++++++++++++++++++------------------------
>  tools/perf/tests/parse-events.c        |  40 ++++++++++++++++++
>  tools/perf/util/parse-events.c         |   8 +++-
>  tools/perf/util/parse-events.l         |   2 +-
>  5 files changed, 141 insertions(+), 54 deletions(-)

Re: [PATCH 1/6] perf tools: Add 'I' event modifier for exclude_idle bit

[Arnaldo Carvalho de Melo]

Em Tue, Apr 07, 2015 at 11:25:14PM +0200, Jiri Olsa escreveu:
> Adding 'I' event modifier to have complete set of modifiers
> for perf_event_attr:exclude_* bits.
> 
> Any event specified with 'I' modifier will have the
> perf_event_attr:exclude_idle bit set.
> 
>   $ perf record -e cycles:I -vv ls 2>&1 | grep exclude_idle
>   exclude_hv          0    exclude_idle        1
> 
> Adding automated tests.

Applied this one, waiting for an Ack from Andi, as you asked for,

- Arnaldo

Re: [PATCH 2/6] perf stat: Fix metrics calculation with event qualifiers

[Namhyung Kim]

Hi Jiri and Andi,

On Tue, Apr 07, 2015 at 11:25:15PM +0200, Jiri Olsa wrote:
> From: Andi Kleen <ak@linux.intel.com>
> 
> Currently in perf IPC and other metrics cannot be directly shown
> separately for both user and kernel in a single run. The problem was
> that the metrics matching code did not check event qualifiers.
> 
> With this patch the following case works correctly.
> 
> % perf stat -e cycles:k,cycles:u,instructions:k,instructions:u true
> 
>  Performance counter stats for 'true':
> 
>            531,718      cycles:k
>            203,895      cycles:u
>            338,151      instructions:k            #    0.64  insns per cycle
>            105,961      instructions:u            #    0.52  insns per cycle
> 
>        0.002989739 seconds time elapsed
> 
> Previously it would misreport the ratios because they were matching
> the wrong value.

This patch reminds me of following change:


>From a8ae843bf7cff7b68dc39b6f088c43669d140d00 Mon Sep 17 00:00:00 2001
From: Namhyung Kim <namhyung@kernel.org>
Date: Wed, 8 Apr 2015 21:56:12 +0900
Subject: [RFC] perf stat: Add -U/--user option

The -U/--user option is to control event modifier for all events easily.
It has same effect as if adding 'u' modifier to every events.

An example follows:

  $ perf stat -U true

   Performance counter stats for 'true':

            0.722461      task-clock:uH (msec)      #    0.520 CPUs utilized
                   0      context-switches:uH       #    0.000 K/sec
                   0      cpu-migrations:uH         #    0.000 K/sec
                  41      page-faults:uH            #    0.057 M/sec
             145,341      cycles:uH                 #    0.201 GHz
             517,317      stalled-cycles-frontend:uH #  355.93% frontend cycles idle
             497,570      stalled-cycles-backend:uH #  342.35% backend  cycles idle
              98,846      instructions:uH           #    0.68  insns per cycle
                                                    #    5.23  stalled cycles per insn
              19,552      branches:uH               #   27.063 M/sec
       <not counted>      branch-misses:uH

         0.001389613 seconds time elapsed

Cc: Andi Kleen <ak@linux.intel.com>
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
---
 tools/perf/Documentation/perf-stat.txt |  5 ++++-
 tools/perf/builtin-stat.c              | 11 +++++++++++
 2 files changed, 15 insertions(+), 1 deletion(-)

diff --git a/tools/perf/Documentation/perf-stat.txt b/tools/perf/Documentation/perf-stat.txt
index 04e150d83e7d..54b0734028a6 100644
--- a/tools/perf/Documentation/perf-stat.txt
+++ b/tools/perf/Documentation/perf-stat.txt
@@ -151,9 +151,12 @@ filter out the startup phase of the program, which is often very different.
 
 -T::
 --transaction::
-
 Print statistics of transactional execution if supported.
 
+-U::
+--user::
+Collect events only in user mode execution
+
 EXAMPLES
 --------
 
diff --git a/tools/perf/builtin-stat.c b/tools/perf/builtin-stat.c
index fec089f1c364..d191af942cb4 100644
--- a/tools/perf/builtin-stat.c
+++ b/tools/perf/builtin-stat.c
@@ -141,6 +141,7 @@ static unsigned int		interval			= 0;
 static unsigned int		initial_delay			= 0;
 static unsigned int		unit_width			= 4; /* strlen("unit") */
 static bool			forever				= false;
+static bool			mod_user			= false;
 static struct timespec		ref_time;
 static struct cpu_map		*aggr_map;
 static int			(*aggr_get_id)(struct cpu_map *m, int cpu);
@@ -1809,6 +1810,7 @@ int cmd_stat(int argc, const char **argv, const char *prefix __maybe_unused)
 		     "aggregate counts per physical processor core", AGGR_CORE),
 	OPT_UINTEGER('D', "delay", &initial_delay,
 		     "ms to wait before starting measurement after program start"),
+	OPT_BOOLEAN('U', "user", &mod_user, "Count event only in user mode"),
 	OPT_END()
 	};
 	const char * const stat_usage[] = {
@@ -1913,6 +1915,15 @@ int cmd_stat(int argc, const char **argv, const char *prefix __maybe_unused)
 	if (add_default_attributes())
 		goto out;
 
+	if (mod_user) {
+		struct perf_evsel *evsel;
+
+		evlist__for_each(evsel_list, evsel) {
+			evsel->attr.exclude_kernel = 1;
+			evsel->attr.exclude_hv = 1;
+		}
+	}
+
 	target__validate(&target);
 
 	if (perf_evlist__create_maps(evsel_list, &target) < 0) {
-- 
2.3.5

[tip:perf/core] perf tools: Add 'I' event modifier for exclude_idle bit

[tip-bot for Jiri Olsa]

Commit-ID:  a1e12da4796a4ddd0e911687a290eb396d1c64bf
Gitweb:     http://git.kernel.org/tip/a1e12da4796a4ddd0e911687a290eb396d1c64bf
Author:     Jiri Olsa <jolsa@kernel.org>
AuthorDate: Tue, 7 Apr 2015 23:25:14 +0200
Committer:  Arnaldo Carvalho de Melo <acme@redhat.com>
CommitDate: Wed, 8 Apr 2015 11:00:16 -0300

perf tools: Add 'I' event modifier for exclude_idle bit

Adding 'I' event modifier to have complete set of modifiers for
perf_event_attr:exclude_* bits.

Any event specified with 'I' modifier will have the
perf_event_attr:exclude_idle bit set.

  $ perf record -e cycles:I -vv ls 2>&1 | grep exclude_idle
  exclude_hv          0    exclude_idle        1

Adding automated tests.

Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: David Ahern <dsahern@gmail.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: William Cohen <wcohen@redhat.com>
Link: http://lkml.kernel.org/r/1428441919-23099-2-git-send-email-jolsa@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/Documentation/perf-list.txt |  1 +
 tools/perf/tests/parse-events.c        | 40 ++++++++++++++++++++++++++++++++++
 tools/perf/util/parse-events.c         |  8 ++++++-
 tools/perf/util/parse-events.l         |  2 +-
 4 files changed, 49 insertions(+), 2 deletions(-)

diff --git a/tools/perf/Documentation/perf-list.txt b/tools/perf/Documentation/perf-list.txt
index 4692d27..bada893 100644
--- a/tools/perf/Documentation/perf-list.txt
+++ b/tools/perf/Documentation/perf-list.txt
@@ -26,6 +26,7 @@ counted. The following modifiers exist:
  u - user-space counting
  k - kernel counting
  h - hypervisor counting
+ I - non idle counting
  G - guest counting (in KVM guests)
  H - host counting (not in KVM guests)
  p - precise level
diff --git a/tools/perf/tests/parse-events.c b/tools/perf/tests/parse-events.c
index ac243eb..3de7449 100644
--- a/tools/perf/tests/parse-events.c
+++ b/tools/perf/tests/parse-events.c
@@ -295,6 +295,36 @@ static int test__checkevent_genhw_modifier(struct perf_evlist *evlist)
 	return test__checkevent_genhw(evlist);
 }
 
+static int test__checkevent_exclude_idle_modifier(struct perf_evlist *evlist)
+{
+	struct perf_evsel *evsel = perf_evlist__first(evlist);
+
+	TEST_ASSERT_VAL("wrong exclude idle", evsel->attr.exclude_idle);
+	TEST_ASSERT_VAL("wrong exclude guest", !evsel->attr.exclude_guest);
+	TEST_ASSERT_VAL("wrong exclude host", !evsel->attr.exclude_host);
+	TEST_ASSERT_VAL("wrong exclude_user", !evsel->attr.exclude_user);
+	TEST_ASSERT_VAL("wrong exclude_kernel", !evsel->attr.exclude_kernel);
+	TEST_ASSERT_VAL("wrong exclude_hv", !evsel->attr.exclude_hv);
+	TEST_ASSERT_VAL("wrong precise_ip", !evsel->attr.precise_ip);
+
+	return test__checkevent_symbolic_name(evlist);
+}
+
+static int test__checkevent_exclude_idle_modifier_1(struct perf_evlist *evlist)
+{
+	struct perf_evsel *evsel = perf_evlist__first(evlist);
+
+	TEST_ASSERT_VAL("wrong exclude idle", evsel->attr.exclude_idle);
+	TEST_ASSERT_VAL("wrong exclude guest", !evsel->attr.exclude_guest);
+	TEST_ASSERT_VAL("wrong exclude host", evsel->attr.exclude_host);
+	TEST_ASSERT_VAL("wrong exclude_user", evsel->attr.exclude_user);
+	TEST_ASSERT_VAL("wrong exclude_kernel", !evsel->attr.exclude_kernel);
+	TEST_ASSERT_VAL("wrong exclude_hv", evsel->attr.exclude_hv);
+	TEST_ASSERT_VAL("wrong precise_ip", !evsel->attr.precise_ip);
+
+	return test__checkevent_symbolic_name(evlist);
+}
+
 static int test__checkevent_breakpoint_modifier(struct perf_evlist *evlist)
 {
 	struct perf_evsel *evsel = perf_evlist__first(evlist);
@@ -1494,6 +1524,16 @@ static struct evlist_test test__events[] = {
 		.id    = 100,
 	},
 #endif
+	{
+		.name  = "instructions:I",
+		.check = test__checkevent_exclude_idle_modifier,
+		.id    = 45,
+	},
+	{
+		.name  = "instructions:kIG",
+		.check = test__checkevent_exclude_idle_modifier_1,
+		.id    = 46,
+	},
 };
 
 static struct evlist_test test__events_pmu[] = {
diff --git a/tools/perf/util/parse-events.c b/tools/perf/util/parse-events.c
index fe07573..be06553 100644
--- a/tools/perf/util/parse-events.c
+++ b/tools/perf/util/parse-events.c
@@ -709,6 +709,7 @@ struct event_modifier {
 	int eh;
 	int eH;
 	int eG;
+	int eI;
 	int precise;
 	int exclude_GH;
 	int sample_read;
@@ -723,6 +724,7 @@ static int get_event_modifier(struct event_modifier *mod, char *str,
 	int eh = evsel ? evsel->attr.exclude_hv : 0;
 	int eH = evsel ? evsel->attr.exclude_host : 0;
 	int eG = evsel ? evsel->attr.exclude_guest : 0;
+	int eI = evsel ? evsel->attr.exclude_idle : 0;
 	int precise = evsel ? evsel->attr.precise_ip : 0;
 	int sample_read = 0;
 	int pinned = evsel ? evsel->attr.pinned : 0;
@@ -753,6 +755,8 @@ static int get_event_modifier(struct event_modifier *mod, char *str,
 			if (!exclude_GH)
 				exclude_GH = eG = eH = 1;
 			eH = 0;
+		} else if (*str == 'I') {
+			eI = 1;
 		} else if (*str == 'p') {
 			precise++;
 			/* use of precise requires exclude_guest */
@@ -786,6 +790,7 @@ static int get_event_modifier(struct event_modifier *mod, char *str,
 	mod->eh = eh;
 	mod->eH = eH;
 	mod->eG = eG;
+	mod->eI = eI;
 	mod->precise = precise;
 	mod->exclude_GH = exclude_GH;
 	mod->sample_read = sample_read;
@@ -803,7 +808,7 @@ static int check_modifier(char *str)
 	char *p = str;
 
 	/* The sizeof includes 0 byte as well. */
-	if (strlen(str) > (sizeof("ukhGHpppSD") - 1))
+	if (strlen(str) > (sizeof("ukhGHpppSDI") - 1))
 		return -1;
 
 	while (*p) {
@@ -839,6 +844,7 @@ int parse_events__modifier_event(struct list_head *list, char *str, bool add)
 		evsel->attr.precise_ip     = mod.precise;
 		evsel->attr.exclude_host   = mod.eH;
 		evsel->attr.exclude_guest  = mod.eG;
+		evsel->attr.exclude_idle   = mod.eI;
 		evsel->exclude_GH          = mod.exclude_GH;
 		evsel->sample_read         = mod.sample_read;
 
diff --git a/tools/perf/util/parse-events.l b/tools/perf/util/parse-events.l
index 94eacb6..8895cf3 100644
--- a/tools/perf/util/parse-events.l
+++ b/tools/perf/util/parse-events.l
@@ -101,7 +101,7 @@ num_raw_hex	[a-fA-F0-9]+
 name		[a-zA-Z_*?][a-zA-Z0-9_*?]*
 name_minus	[a-zA-Z_*?][a-zA-Z0-9\-_*?]*
 /* If you add a modifier you need to update check_modifier() */
-modifier_event	[ukhpGHSD]+
+modifier_event	[ukhpGHSDI]+
 modifier_bp	[rwx]{1,3}
 
 %%

Re: [RFC 0/6] perf stat: Metrics calculation fix

[Jiri Olsa]

On Tue, Apr 07, 2015 at 11:25:13PM +0200, Jiri Olsa wrote:
> hi,
> I'm sending Andi's patch accompanied with changes requested by Ingo.
>   http://marc.info/?l=linux-kernel&m=139748629929175&w=2
> 
> I added simple bit calculated index array, while exclude_* bits
> might have more strict logic and ban some combinations, but I
> wanted to keep it simple and did not find better solution.

ping

jirka

> 
> thanks,
> jirka
> 
> 
> ---
> Andi Kleen (1):
>       perf stat: Fix metrics calculation with event qualifiers
> 
> Jiri Olsa (5):
>       perf tools: Add 'I' event modifier for exclude_idle bit
>       perf stat: Change metrics context calculation
>       perf stat: Add metrics support for exclude_hv
>       perf stat: Add metrics support for exclude_(host|guest)
>       perf stat: Add metrics support for exclude_idle
> 
>  tools/perf/Documentation/perf-list.txt |   1 +
>  tools/perf/builtin-stat.c              | 144 +++++++++++++++++++++++++++++++++++++++++------------------------
>  tools/perf/tests/parse-events.c        |  40 ++++++++++++++++++
>  tools/perf/util/parse-events.c         |   8 +++-
>  tools/perf/util/parse-events.l         |   2 +-
>  5 files changed, 141 insertions(+), 54 deletions(-)

Re: [RFC 0/6] perf stat: Metrics calculation fix

[Namhyung Kim]

Hi Jiri,

On Tue, Apr 07, 2015 at 11:25:13PM +0200, Jiri Olsa wrote:
> hi,
> I'm sending Andi's patch accompanied with changes requested by Ingo.
>   http://marc.info/?l=linux-kernel&m=139748629929175&w=2
> 
> I added simple bit calculated index array, while exclude_* bits
> might have more strict logic and ban some combinations, but I
> wanted to keep it simple and did not find better solution.

For the series,

Acked-by: Namhyung Kim <namhyung@kernel.org>

Thanks,
Namhyung

Re: [PATCH 2/6] perf stat: Fix metrics calculation with event qualifiers

[Arnaldo Carvalho de Melo]

Em Wed, Apr 08, 2015 at 10:28:33PM +0900, Namhyung Kim escreveu:
> Hi Jiri and Andi,
> On Tue, Apr 07, 2015 at 11:25:15PM +0200, Jiri Olsa wrote:
> > From: Andi Kleen <ak@linux.intel.com>
> > % perf stat -e cycles:k,cycles:u,instructions:k,instructions:u true
> > 
> >  Performance counter stats for 'true':
> > 
> >            531,718      cycles:k
> >            203,895      cycles:u
> >            338,151      instructions:k            #    0.64  insns per cycle
> >            105,961      instructions:u            #    0.52  insns per cycle
> > 
> >        0.002989739 seconds time elapsed
> > 
> > Previously it would misreport the ratios because they were matching
> > the wrong value.
> 
> This patch reminds me of following change:


Yeah, useful, but do we have one for the kernel, hypervisor, guest, etc?
Could we instead have a --cpumode u?k?g?H? combo, so that we could filter by
that?

Applying this series, with your Ack, thanks!

- Arnaldo
 
> >From a8ae843bf7cff7b68dc39b6f088c43669d140d00 Mon Sep 17 00:00:00 2001
> From: Namhyung Kim <namhyung@kernel.org>
> Date: Wed, 8 Apr 2015 21:56:12 +0900
> Subject: [RFC] perf stat: Add -U/--user option
> 
> The -U/--user option is to control event modifier for all events easily.
> It has same effect as if adding 'u' modifier to every events.
> 
> An example follows:
> 
>   $ perf stat -U true
> 
>    Performance counter stats for 'true':
> 
>             0.722461      task-clock:uH (msec)      #    0.520 CPUs utilized
>                    0      context-switches:uH       #    0.000 K/sec
>                    0      cpu-migrations:uH         #    0.000 K/sec
>                   41      page-faults:uH            #    0.057 M/sec
>              145,341      cycles:uH                 #    0.201 GHz
>              517,317      stalled-cycles-frontend:uH #  355.93% frontend cycles idle
>              497,570      stalled-cycles-backend:uH #  342.35% backend  cycles idle
>               98,846      instructions:uH           #    0.68  insns per cycle
>                                                     #    5.23  stalled cycles per insn
>               19,552      branches:uH               #   27.063 M/sec
>        <not counted>      branch-misses:uH
> 
>          0.001389613 seconds time elapsed
> 
> Cc: Andi Kleen <ak@linux.intel.com>
> Signed-off-by: Namhyung Kim <namhyung@kernel.org>
> ---
>  tools/perf/Documentation/perf-stat.txt |  5 ++++-
>  tools/perf/builtin-stat.c              | 11 +++++++++++
>  2 files changed, 15 insertions(+), 1 deletion(-)
> 
> diff --git a/tools/perf/Documentation/perf-stat.txt b/tools/perf/Documentation/perf-stat.txt
> index 04e150d83e7d..54b0734028a6 100644
> --- a/tools/perf/Documentation/perf-stat.txt
> +++ b/tools/perf/Documentation/perf-stat.txt
> @@ -151,9 +151,12 @@ filter out the startup phase of the program, which is often very different.
>  
>  -T::
>  --transaction::
> -
>  Print statistics of transactional execution if supported.
>  
> +-U::
> +--user::
> +Collect events only in user mode execution
> +
>  EXAMPLES
>  --------
>  
> diff --git a/tools/perf/builtin-stat.c b/tools/perf/builtin-stat.c
> index fec089f1c364..d191af942cb4 100644
> --- a/tools/perf/builtin-stat.c
> +++ b/tools/perf/builtin-stat.c
> @@ -141,6 +141,7 @@ static unsigned int		interval			= 0;
>  static unsigned int		initial_delay			= 0;
>  static unsigned int		unit_width			= 4; /* strlen("unit") */
>  static bool			forever				= false;
> +static bool			mod_user			= false;
>  static struct timespec		ref_time;
>  static struct cpu_map		*aggr_map;
>  static int			(*aggr_get_id)(struct cpu_map *m, int cpu);
> @@ -1809,6 +1810,7 @@ int cmd_stat(int argc, const char **argv, const char *prefix __maybe_unused)
>  		     "aggregate counts per physical processor core", AGGR_CORE),
>  	OPT_UINTEGER('D', "delay", &initial_delay,
>  		     "ms to wait before starting measurement after program start"),
> +	OPT_BOOLEAN('U', "user", &mod_user, "Count event only in user mode"),
>  	OPT_END()
>  	};
>  	const char * const stat_usage[] = {
> @@ -1913,6 +1915,15 @@ int cmd_stat(int argc, const char **argv, const char *prefix __maybe_unused)
>  	if (add_default_attributes())
>  		goto out;
>  
> +	if (mod_user) {
> +		struct perf_evsel *evsel;
> +
> +		evlist__for_each(evsel_list, evsel) {
> +			evsel->attr.exclude_kernel = 1;
> +			evsel->attr.exclude_hv = 1;
> +		}
> +	}
> +
>  	target__validate(&target);
>  
>  	if (perf_evlist__create_maps(evsel_list, &target) < 0) {
> -- 
> 2.3.5

[tip:perf/core] perf stat: Fix metrics calculation with event qualifiers

[tip-bot for Andi Kleen]

Commit-ID:  a0aa21e2281f89a197ad8e7bd8008f3b7b735dfc
Gitweb:     http://git.kernel.org/tip/a0aa21e2281f89a197ad8e7bd8008f3b7b735dfc
Author:     Andi Kleen <ak@linux.intel.com>
AuthorDate: Tue, 7 Apr 2015 23:25:15 +0200
Committer:  Arnaldo Carvalho de Melo <acme@redhat.com>
CommitDate: Wed, 29 Apr 2015 10:38:04 -0300

perf stat: Fix metrics calculation with event qualifiers

Currently in perf IPC and other metrics cannot be directly shown
separately for both user and kernel in a single run. The problem was
that the metrics matching code did not check event qualifiers.

With this patch the following case works correctly.

% perf stat -e cycles:k,cycles:u,instructions:k,instructions:u true

 Performance counter stats for 'true':

           531,718      cycles:k
           203,895      cycles:u
           338,151      instructions:k            #    0.64  insns per cycle
           105,961      instructions:u            #    0.52  insns per cycle

       0.002989739 seconds time elapsed

Previously it would misreport the ratios because they were matching the
wrong value.

The patch is fairly big, but quite mechanic as it just adds context
indexes everywhere.

Reported-by: William Cohen <wcohen@redhat.com>
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Acked-by: Namhyung Kim <namhyung@kernel.org>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: David Ahern <dsahern@gmail.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: William Cohen <wcohen@redhat.com>
Link: http://lkml.kernel.org/r/1428441919-23099-3-git-send-email-jolsa@kernel.org
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/builtin-stat.c | 129 +++++++++++++++++++++++++++-------------------
 1 file changed, 77 insertions(+), 52 deletions(-)

diff --git a/tools/perf/builtin-stat.c b/tools/perf/builtin-stat.c
index 3dbd8c5..52f4330 100644
--- a/tools/perf/builtin-stat.c
+++ b/tools/perf/builtin-stat.c
@@ -247,21 +247,35 @@ out_free:
 	return -1;
 }
 
+#define NUM_CTX 3
+
+enum { CTX_USER, CTX_KERNEL, CTX_ALL };
+
 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];
-static struct stats runtime_stalled_cycles_back_stats[MAX_NR_CPUS];
-static struct stats runtime_branches_stats[MAX_NR_CPUS];
-static struct stats runtime_cacherefs_stats[MAX_NR_CPUS];
-static struct stats runtime_l1_dcache_stats[MAX_NR_CPUS];
-static struct stats runtime_l1_icache_stats[MAX_NR_CPUS];
-static struct stats runtime_ll_cache_stats[MAX_NR_CPUS];
-static struct stats runtime_itlb_cache_stats[MAX_NR_CPUS];
-static struct stats runtime_dtlb_cache_stats[MAX_NR_CPUS];
-static struct stats runtime_cycles_in_tx_stats[MAX_NR_CPUS];
+static struct stats runtime_cycles_stats[NUM_CTX][MAX_NR_CPUS];
+static struct stats runtime_stalled_cycles_front_stats[NUM_CTX][MAX_NR_CPUS];
+static struct stats runtime_stalled_cycles_back_stats[NUM_CTX][MAX_NR_CPUS];
+static struct stats runtime_branches_stats[NUM_CTX][MAX_NR_CPUS];
+static struct stats runtime_cacherefs_stats[NUM_CTX][MAX_NR_CPUS];
+static struct stats runtime_l1_dcache_stats[NUM_CTX][MAX_NR_CPUS];
+static struct stats runtime_l1_icache_stats[NUM_CTX][MAX_NR_CPUS];
+static struct stats runtime_ll_cache_stats[NUM_CTX][MAX_NR_CPUS];
+static struct stats runtime_itlb_cache_stats[NUM_CTX][MAX_NR_CPUS];
+static struct stats runtime_dtlb_cache_stats[NUM_CTX][MAX_NR_CPUS];
+static struct stats runtime_cycles_in_tx_stats[NUM_CTX][MAX_NR_CPUS];
 static struct stats walltime_nsecs_stats;
-static struct stats runtime_transaction_stats[MAX_NR_CPUS];
-static struct stats runtime_elision_stats[MAX_NR_CPUS];
+static struct stats runtime_transaction_stats[NUM_CTX][MAX_NR_CPUS];
+static struct stats runtime_elision_stats[NUM_CTX][MAX_NR_CPUS];
+
+static int evsel_context(struct perf_evsel *evsel)
+{
+	if (evsel->attr.exclude_kernel)
+		return CTX_USER;
+	if (evsel->attr.exclude_user)
+		return CTX_KERNEL;
+	/* Handle hypervisor too? */
+	return CTX_ALL;
+}
 
 static void perf_stat__reset_stats(struct perf_evlist *evlist)
 {
@@ -356,37 +370,39 @@ static struct perf_evsel *nth_evsel(int n)
 static void update_shadow_stats(struct perf_evsel *counter, u64 *count,
 				int cpu)
 {
+	int ctx = evsel_context(counter);
+
 	if (perf_evsel__match(counter, SOFTWARE, SW_TASK_CLOCK))
 		update_stats(&runtime_nsecs_stats[cpu], count[0]);
 	else if (perf_evsel__match(counter, HARDWARE, HW_CPU_CYCLES))
-		update_stats(&runtime_cycles_stats[cpu], count[0]);
+		update_stats(&runtime_cycles_stats[ctx][cpu], count[0]);
 	else if (transaction_run &&
 		 perf_evsel__cmp(counter, nth_evsel(T_CYCLES_IN_TX)))
-		update_stats(&runtime_cycles_in_tx_stats[cpu], count[0]);
+		update_stats(&runtime_transaction_stats[ctx][cpu], count[0]);
 	else if (transaction_run &&
 		 perf_evsel__cmp(counter, nth_evsel(T_TRANSACTION_START)))
-		update_stats(&runtime_transaction_stats[cpu], count[0]);
+		update_stats(&runtime_transaction_stats[ctx][cpu], count[0]);
 	else if (transaction_run &&
 		 perf_evsel__cmp(counter, nth_evsel(T_ELISION_START)))
-		update_stats(&runtime_elision_stats[cpu], count[0]);
+		update_stats(&runtime_elision_stats[ctx][cpu], count[0]);
 	else if (perf_evsel__match(counter, HARDWARE, HW_STALLED_CYCLES_FRONTEND))
-		update_stats(&runtime_stalled_cycles_front_stats[cpu], count[0]);
+		update_stats(&runtime_stalled_cycles_front_stats[ctx][cpu], count[0]);
 	else if (perf_evsel__match(counter, HARDWARE, HW_STALLED_CYCLES_BACKEND))
-		update_stats(&runtime_stalled_cycles_back_stats[cpu], count[0]);
+		update_stats(&runtime_stalled_cycles_back_stats[ctx][cpu], count[0]);
 	else if (perf_evsel__match(counter, HARDWARE, HW_BRANCH_INSTRUCTIONS))
-		update_stats(&runtime_branches_stats[cpu], count[0]);
+		update_stats(&runtime_branches_stats[ctx][cpu], count[0]);
 	else if (perf_evsel__match(counter, HARDWARE, HW_CACHE_REFERENCES))
-		update_stats(&runtime_cacherefs_stats[cpu], count[0]);
+		update_stats(&runtime_cacherefs_stats[ctx][cpu], count[0]);
 	else if (perf_evsel__match(counter, HW_CACHE, HW_CACHE_L1D))
-		update_stats(&runtime_l1_dcache_stats[cpu], count[0]);
+		update_stats(&runtime_l1_dcache_stats[ctx][cpu], count[0]);
 	else if (perf_evsel__match(counter, HW_CACHE, HW_CACHE_L1I))
-		update_stats(&runtime_l1_icache_stats[cpu], count[0]);
+		update_stats(&runtime_ll_cache_stats[ctx][cpu], count[0]);
 	else if (perf_evsel__match(counter, HW_CACHE, HW_CACHE_LL))
-		update_stats(&runtime_ll_cache_stats[cpu], count[0]);
+		update_stats(&runtime_ll_cache_stats[ctx][cpu], count[0]);
 	else if (perf_evsel__match(counter, HW_CACHE, HW_CACHE_DTLB))
-		update_stats(&runtime_dtlb_cache_stats[cpu], count[0]);
+		update_stats(&runtime_dtlb_cache_stats[ctx][cpu], count[0]);
 	else if (perf_evsel__match(counter, HW_CACHE, HW_CACHE_ITLB))
-		update_stats(&runtime_itlb_cache_stats[cpu], count[0]);
+		update_stats(&runtime_itlb_cache_stats[ctx][cpu], count[0]);
 }
 
 static void zero_per_pkg(struct perf_evsel *counter)
@@ -908,8 +924,9 @@ static void print_stalled_cycles_frontend(int cpu,
 {
 	double total, ratio = 0.0;
 	const char *color;
+	int ctx = evsel_context(evsel);
 
-	total = avg_stats(&runtime_cycles_stats[cpu]);
+	total = avg_stats(&runtime_cycles_stats[ctx][cpu]);
 
 	if (total)
 		ratio = avg / total * 100.0;
@@ -927,8 +944,9 @@ static void print_stalled_cycles_backend(int cpu,
 {
 	double total, ratio = 0.0;
 	const char *color;
+	int ctx = evsel_context(evsel);
 
-	total = avg_stats(&runtime_cycles_stats[cpu]);
+	total = avg_stats(&runtime_cycles_stats[ctx][cpu]);
 
 	if (total)
 		ratio = avg / total * 100.0;
@@ -946,8 +964,9 @@ static void print_branch_misses(int cpu,
 {
 	double total, ratio = 0.0;
 	const char *color;
+	int ctx = evsel_context(evsel);
 
-	total = avg_stats(&runtime_branches_stats[cpu]);
+	total = avg_stats(&runtime_branches_stats[ctx][cpu]);
 
 	if (total)
 		ratio = avg / total * 100.0;
@@ -965,8 +984,9 @@ static void print_l1_dcache_misses(int cpu,
 {
 	double total, ratio = 0.0;
 	const char *color;
+	int ctx = evsel_context(evsel);
 
-	total = avg_stats(&runtime_l1_dcache_stats[cpu]);
+	total = avg_stats(&runtime_l1_dcache_stats[ctx][cpu]);
 
 	if (total)
 		ratio = avg / total * 100.0;
@@ -984,8 +1004,9 @@ static void print_l1_icache_misses(int cpu,
 {
 	double total, ratio = 0.0;
 	const char *color;
+	int ctx = evsel_context(evsel);
 
-	total = avg_stats(&runtime_l1_icache_stats[cpu]);
+	total = avg_stats(&runtime_l1_icache_stats[ctx][cpu]);
 
 	if (total)
 		ratio = avg / total * 100.0;
@@ -1003,8 +1024,9 @@ static void print_dtlb_cache_misses(int cpu,
 {
 	double total, ratio = 0.0;
 	const char *color;
+	int ctx = evsel_context(evsel);
 
-	total = avg_stats(&runtime_dtlb_cache_stats[cpu]);
+	total = avg_stats(&runtime_dtlb_cache_stats[ctx][cpu]);
 
 	if (total)
 		ratio = avg / total * 100.0;
@@ -1022,8 +1044,9 @@ static void print_itlb_cache_misses(int cpu,
 {
 	double total, ratio = 0.0;
 	const char *color;
+	int ctx = evsel_context(evsel);
 
-	total = avg_stats(&runtime_itlb_cache_stats[cpu]);
+	total = avg_stats(&runtime_itlb_cache_stats[ctx][cpu]);
 
 	if (total)
 		ratio = avg / total * 100.0;
@@ -1041,8 +1064,9 @@ static void print_ll_cache_misses(int cpu,
 {
 	double total, ratio = 0.0;
 	const char *color;
+	int ctx = evsel_context(evsel);
 
-	total = avg_stats(&runtime_ll_cache_stats[cpu]);
+	total = avg_stats(&runtime_ll_cache_stats[ctx][cpu]);
 
 	if (total)
 		ratio = avg / total * 100.0;
@@ -1060,6 +1084,7 @@ static void abs_printout(int id, int nr, struct perf_evsel *evsel, double avg)
 	double sc =  evsel->scale;
 	const char *fmt;
 	int cpu = cpu_map__id_to_cpu(id);
+	int ctx = evsel_context(evsel);
 
 	if (csv_output) {
 		fmt = sc != 1.0 ?  "%.2f%s" : "%.0f%s";
@@ -1091,15 +1116,15 @@ static void abs_printout(int id, int nr, struct perf_evsel *evsel, double avg)
 		return;
 
 	if (perf_evsel__match(evsel, HARDWARE, HW_INSTRUCTIONS)) {
-		total = avg_stats(&runtime_cycles_stats[cpu]);
+		total = avg_stats(&runtime_cycles_stats[ctx][cpu]);
 		if (total) {
 			ratio = avg / total;
 			fprintf(output, " #   %5.2f  insns per cycle        ", ratio);
 		} else {
 			fprintf(output, "                                   ");
 		}
-		total = avg_stats(&runtime_stalled_cycles_front_stats[cpu]);
-		total = max(total, avg_stats(&runtime_stalled_cycles_back_stats[cpu]));
+		total = avg_stats(&runtime_stalled_cycles_front_stats[ctx][cpu]);
+		total = max(total, avg_stats(&runtime_stalled_cycles_back_stats[ctx][cpu]));
 
 		if (total && avg) {
 			ratio = total / avg;
@@ -1110,46 +1135,46 @@ static void abs_printout(int id, int nr, struct perf_evsel *evsel, double avg)
 		}
 
 	} else if (perf_evsel__match(evsel, HARDWARE, HW_BRANCH_MISSES) &&
-			runtime_branches_stats[cpu].n != 0) {
+			runtime_branches_stats[ctx][cpu].n != 0) {
 		print_branch_misses(cpu, evsel, avg);
 	} else if (
 		evsel->attr.type == PERF_TYPE_HW_CACHE &&
 		evsel->attr.config ==  ( PERF_COUNT_HW_CACHE_L1D |
 					((PERF_COUNT_HW_CACHE_OP_READ) << 8) |
 					((PERF_COUNT_HW_CACHE_RESULT_MISS) << 16)) &&
-			runtime_l1_dcache_stats[cpu].n != 0) {
+			runtime_l1_dcache_stats[ctx][cpu].n != 0) {
 		print_l1_dcache_misses(cpu, evsel, avg);
 	} else if (
 		evsel->attr.type == PERF_TYPE_HW_CACHE &&
 		evsel->attr.config ==  ( PERF_COUNT_HW_CACHE_L1I |
 					((PERF_COUNT_HW_CACHE_OP_READ) << 8) |
 					((PERF_COUNT_HW_CACHE_RESULT_MISS) << 16)) &&
-			runtime_l1_icache_stats[cpu].n != 0) {
+			runtime_l1_icache_stats[ctx][cpu].n != 0) {
 		print_l1_icache_misses(cpu, evsel, avg);
 	} else if (
 		evsel->attr.type == PERF_TYPE_HW_CACHE &&
 		evsel->attr.config ==  ( PERF_COUNT_HW_CACHE_DTLB |
 					((PERF_COUNT_HW_CACHE_OP_READ) << 8) |
 					((PERF_COUNT_HW_CACHE_RESULT_MISS) << 16)) &&
-			runtime_dtlb_cache_stats[cpu].n != 0) {
+			runtime_dtlb_cache_stats[ctx][cpu].n != 0) {
 		print_dtlb_cache_misses(cpu, evsel, avg);
 	} else if (
 		evsel->attr.type == PERF_TYPE_HW_CACHE &&
 		evsel->attr.config ==  ( PERF_COUNT_HW_CACHE_ITLB |
 					((PERF_COUNT_HW_CACHE_OP_READ) << 8) |
 					((PERF_COUNT_HW_CACHE_RESULT_MISS) << 16)) &&
-			runtime_itlb_cache_stats[cpu].n != 0) {
+			runtime_itlb_cache_stats[ctx][cpu].n != 0) {
 		print_itlb_cache_misses(cpu, evsel, avg);
 	} else if (
 		evsel->attr.type == PERF_TYPE_HW_CACHE &&
 		evsel->attr.config ==  ( PERF_COUNT_HW_CACHE_LL |
 					((PERF_COUNT_HW_CACHE_OP_READ) << 8) |
 					((PERF_COUNT_HW_CACHE_RESULT_MISS) << 16)) &&
-			runtime_ll_cache_stats[cpu].n != 0) {
+			runtime_ll_cache_stats[ctx][cpu].n != 0) {
 		print_ll_cache_misses(cpu, evsel, avg);
 	} else if (perf_evsel__match(evsel, HARDWARE, HW_CACHE_MISSES) &&
-			runtime_cacherefs_stats[cpu].n != 0) {
-		total = avg_stats(&runtime_cacherefs_stats[cpu]);
+			runtime_cacherefs_stats[ctx][cpu].n != 0) {
+		total = avg_stats(&runtime_cacherefs_stats[ctx][cpu]);
 
 		if (total)
 			ratio = avg * 100 / total;
@@ -1171,15 +1196,15 @@ static void abs_printout(int id, int nr, struct perf_evsel *evsel, double avg)
 		}
 	} else if (transaction_run &&
 		   perf_evsel__cmp(evsel, nth_evsel(T_CYCLES_IN_TX))) {
-		total = avg_stats(&runtime_cycles_stats[cpu]);
+		total = avg_stats(&runtime_cycles_stats[ctx][cpu]);
 		if (total)
 			fprintf(output,
 				" #   %5.2f%% transactional cycles   ",
 				100.0 * (avg / total));
 	} else if (transaction_run &&
 		   perf_evsel__cmp(evsel, nth_evsel(T_CYCLES_IN_TX_CP))) {
-		total = avg_stats(&runtime_cycles_stats[cpu]);
-		total2 = avg_stats(&runtime_cycles_in_tx_stats[cpu]);
+		total = avg_stats(&runtime_cycles_stats[ctx][cpu]);
+		total2 = avg_stats(&runtime_cycles_in_tx_stats[ctx][cpu]);
 		if (total2 < avg)
 			total2 = avg;
 		if (total)
@@ -1189,8 +1214,8 @@ static void abs_printout(int id, int nr, struct perf_evsel *evsel, double avg)
 	} else if (transaction_run &&
 		   perf_evsel__cmp(evsel, nth_evsel(T_TRANSACTION_START)) &&
 		   avg > 0 &&
-		   runtime_cycles_in_tx_stats[cpu].n != 0) {
-		total = avg_stats(&runtime_cycles_in_tx_stats[cpu]);
+		   runtime_cycles_in_tx_stats[ctx][cpu].n != 0) {
+		total = avg_stats(&runtime_cycles_in_tx_stats[ctx][cpu]);
 
 		if (total)
 			ratio = total / avg;
@@ -1199,8 +1224,8 @@ static void abs_printout(int id, int nr, struct perf_evsel *evsel, double avg)
 	} else if (transaction_run &&
 		   perf_evsel__cmp(evsel, nth_evsel(T_ELISION_START)) &&
 		   avg > 0 &&
-		   runtime_cycles_in_tx_stats[cpu].n != 0) {
-		total = avg_stats(&runtime_cycles_in_tx_stats[cpu]);
+		   runtime_cycles_in_tx_stats[ctx][cpu].n != 0) {
+		total = avg_stats(&runtime_cycles_in_tx_stats[ctx][cpu]);
 
 		if (total)
 			ratio = total / avg;

[tip:perf/core] perf stat: Change metrics context calculation

[tip-bot for Jiri Olsa]

Commit-ID:  9f71b4f39d7c01cae9d80e938774ff6b0a9bfd80
Gitweb:     http://git.kernel.org/tip/9f71b4f39d7c01cae9d80e938774ff6b0a9bfd80
Author:     Jiri Olsa <jolsa@kernel.org>
AuthorDate: Tue, 7 Apr 2015 23:25:16 +0200
Committer:  Arnaldo Carvalho de Melo <acme@redhat.com>
CommitDate: Wed, 29 Apr 2015 10:38:05 -0300

perf stat: Change metrics context calculation

Changing metrics context calculation to allow more than 2 types of
context.

Following patches will add support for the rest of the exclude_* bits so
we need separate array element for all context combinations.

Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Acked-by: Namhyung Kim <namhyung@kernel.org>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: David Ahern <dsahern@gmail.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: William Cohen <wcohen@redhat.com>
Link: http://lkml.kernel.org/r/1428441919-23099-4-git-send-email-jolsa@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/builtin-stat.c | 17 +++++++++++------
 1 file changed, 11 insertions(+), 6 deletions(-)

diff --git a/tools/perf/builtin-stat.c b/tools/perf/builtin-stat.c
index 52f4330..cca100d 100644
--- a/tools/perf/builtin-stat.c
+++ b/tools/perf/builtin-stat.c
@@ -247,9 +247,13 @@ out_free:
 	return -1;
 }
 
-#define NUM_CTX 3
+enum {
+	CTX_BIT_USER	= 1 << 0,
+	CTX_BIT_KERNEL	= 1 << 1,
+	CTX_BIT_MAX	= 1 << 2,
+};
 
-enum { CTX_USER, CTX_KERNEL, CTX_ALL };
+#define NUM_CTX CTX_BIT_MAX
 
 static struct stats runtime_nsecs_stats[MAX_NR_CPUS];
 static struct stats runtime_cycles_stats[NUM_CTX][MAX_NR_CPUS];
@@ -269,12 +273,13 @@ static struct stats runtime_elision_stats[NUM_CTX][MAX_NR_CPUS];
 
 static int evsel_context(struct perf_evsel *evsel)
 {
+	int ctx = 0;
+
 	if (evsel->attr.exclude_kernel)
-		return CTX_USER;
+		ctx |= CTX_BIT_KERNEL;
 	if (evsel->attr.exclude_user)
-		return CTX_KERNEL;
-	/* Handle hypervisor too? */
-	return CTX_ALL;
+		ctx |= CTX_BIT_USER;
+	return ctx;
 }
 
 static void perf_stat__reset_stats(struct perf_evlist *evlist)

[tip:perf/core] perf stat: Add metrics support for exclude_hv

[tip-bot for Jiri Olsa]

Commit-ID:  afef2fbd1435050377f6487c43f2aa6edc37b30f
Gitweb:     http://git.kernel.org/tip/afef2fbd1435050377f6487c43f2aa6edc37b30f
Author:     Jiri Olsa <jolsa@kernel.org>
AuthorDate: Tue, 7 Apr 2015 23:25:17 +0200
Committer:  Arnaldo Carvalho de Melo <acme@redhat.com>
CommitDate: Wed, 29 Apr 2015 10:38:05 -0300

perf stat: Add metrics support for exclude_hv

Separating metrics values for exclude_hv bit.

Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Acked-by: Namhyung Kim <namhyung@kernel.org>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: David Ahern <dsahern@gmail.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: William Cohen <wcohen@redhat.com>
Link: http://lkml.kernel.org/r/1428441919-23099-5-git-send-email-jolsa@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/builtin-stat.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/tools/perf/builtin-stat.c b/tools/perf/builtin-stat.c
index cca100d..5a88a14 100644
--- a/tools/perf/builtin-stat.c
+++ b/tools/perf/builtin-stat.c
@@ -250,7 +250,8 @@ out_free:
 enum {
 	CTX_BIT_USER	= 1 << 0,
 	CTX_BIT_KERNEL	= 1 << 1,
-	CTX_BIT_MAX	= 1 << 2,
+	CTX_BIT_HV	= 1 << 2,
+	CTX_BIT_MAX	= 1 << 3,
 };
 
 #define NUM_CTX CTX_BIT_MAX
@@ -279,6 +280,8 @@ static int evsel_context(struct perf_evsel *evsel)
 		ctx |= CTX_BIT_KERNEL;
 	if (evsel->attr.exclude_user)
 		ctx |= CTX_BIT_USER;
+	if (evsel->attr.exclude_hv)
+		ctx |= CTX_BIT_HV;
 	return ctx;
 }
 

[tip:perf/core] perf stat: Add metrics support for exclude_( host|guest)

[tip-bot for Jiri Olsa]

Commit-ID:  a2270d38a53aa88c67a72978fc8717e26c7f27d9
Gitweb:     http://git.kernel.org/tip/a2270d38a53aa88c67a72978fc8717e26c7f27d9
Author:     Jiri Olsa <jolsa@kernel.org>
AuthorDate: Tue, 7 Apr 2015 23:25:18 +0200
Committer:  Arnaldo Carvalho de Melo <acme@redhat.com>
CommitDate: Wed, 29 Apr 2015 10:38:05 -0300

perf stat: Add metrics support for exclude_(host|guest)

Separating metrics values for guest and host, so we get proper values.

Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Acked-by: Namhyung Kim <namhyung@kernel.org>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: David Ahern <dsahern@gmail.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: William Cohen <wcohen@redhat.com>
Link: http://lkml.kernel.org/r/1428441919-23099-6-git-send-email-jolsa@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/builtin-stat.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)

diff --git a/tools/perf/builtin-stat.c b/tools/perf/builtin-stat.c
index 5a88a14..ea52508 100644
--- a/tools/perf/builtin-stat.c
+++ b/tools/perf/builtin-stat.c
@@ -251,7 +251,8 @@ enum {
 	CTX_BIT_USER	= 1 << 0,
 	CTX_BIT_KERNEL	= 1 << 1,
 	CTX_BIT_HV	= 1 << 2,
-	CTX_BIT_MAX	= 1 << 3,
+	CTX_BIT_HOST	= 1 << 3,
+	CTX_BIT_MAX	= 1 << 4,
 };
 
 #define NUM_CTX CTX_BIT_MAX
@@ -282,6 +283,9 @@ static int evsel_context(struct perf_evsel *evsel)
 		ctx |= CTX_BIT_USER;
 	if (evsel->attr.exclude_hv)
 		ctx |= CTX_BIT_HV;
+	if (evsel->attr.exclude_host)
+		ctx |= CTX_BIT_HOST;
+
 	return ctx;
 }
 

[tip:perf/core] perf stat: Add metrics support for exclude_idle

[tip-bot for Jiri Olsa]

Commit-ID:  c4fa0d9c1e6aa360cfa2c36f7836a89da24a1b7a
Gitweb:     http://git.kernel.org/tip/c4fa0d9c1e6aa360cfa2c36f7836a89da24a1b7a
Author:     Jiri Olsa <jolsa@kernel.org>
AuthorDate: Tue, 7 Apr 2015 23:25:19 +0200
Committer:  Arnaldo Carvalho de Melo <acme@redhat.com>
CommitDate: Wed, 29 Apr 2015 10:38:06 -0300

perf stat: Add metrics support for exclude_idle

Separating metrics values for exclude_idle bit.

Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Acked-by: Namhyung Kim <namhyung@kernel.org>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: David Ahern <dsahern@gmail.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: William Cohen <wcohen@redhat.com>
Link: http://lkml.kernel.org/r/1428441919-23099-7-git-send-email-jolsa@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/builtin-stat.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/tools/perf/builtin-stat.c b/tools/perf/builtin-stat.c
index ea52508..fd577f7 100644
--- a/tools/perf/builtin-stat.c
+++ b/tools/perf/builtin-stat.c
@@ -252,7 +252,8 @@ enum {
 	CTX_BIT_KERNEL	= 1 << 1,
 	CTX_BIT_HV	= 1 << 2,
 	CTX_BIT_HOST	= 1 << 3,
-	CTX_BIT_MAX	= 1 << 4,
+	CTX_BIT_IDLE	= 1 << 4,
+	CTX_BIT_MAX	= 1 << 5,
 };
 
 #define NUM_CTX CTX_BIT_MAX
@@ -285,6 +286,8 @@ static int evsel_context(struct perf_evsel *evsel)
 		ctx |= CTX_BIT_HV;
 	if (evsel->attr.exclude_host)
 		ctx |= CTX_BIT_HOST;
+	if (evsel->attr.exclude_idle)
+		ctx |= CTX_BIT_IDLE;
 
 	return ctx;
 }