Skip to contentSkip to conversation

perf stat: Fix per socket shadow aggregation

PATCH In review Ian Rogers 2 participants 4 messages Dec 4, 2021 – Dec 7, 2021 perf events

lore
Share & tools
lore

Ian Rogers

author reviewer
#0 lore raw

PATCH 1/1 perf stat: Fix per socket shadow aggregation

An uncore device may have a CPU mask that specifies one CPU per socket:

$ cat /sys/devices/uncore_imc_0/cpumask
0,18
The perf_stat_config aggr_map will map a CPU to the socket and other
aggregation values for it. Fix an error where the index into CPU mask
was being used as the index into the aggr_map. For the cpumask above the
indexes 0 and 1 are passed to aggr_map rather than the CPUs 0 and 18.
---
 tools/perf/util/stat-display.c | 7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)
tools/perf/util/stat-display.c +4 -3
index 588601000f3f..7cfad5cfec38 100644
@@ -516,7 +516,7 @@ static void printout(struct perf_stat_config *config, struct aggr_cpu_id id, int
 static void aggr_update_shadow(struct perf_stat_config *config,
 			       struct evlist *evlist)
 {
-	int cpu, s;
+	int idx, cpu, s;
 	struct aggr_cpu_id s2, id;
 	u64 val;
 	struct evsel *counter;
@@ -525,11 +525,12 @@ static void aggr_update_shadow(struct perf_stat_config *config,
 		id = config->aggr_map->map[s];
 		evlist__for_each_entry(evlist, counter) {
 			val = 0;
-			for (cpu = 0; cpu < evsel__nr_cpus(counter); cpu++) {
+			for (idx = 0; idx < evsel__nr_cpus(counter); idx++) {
+				cpu = perf_cpu_map__cpu(evsel__cpus(counter), idx);
 				s2 = config->aggr_get_id(config, evlist->core.cpus, cpu);
 				if (!cpu_map__compare_aggr_cpu_id(s2, id))
 					continue;
-				val += perf_counts(counter->counts, cpu, 0)->val;
+				val += perf_counts(counter->counts, idx, 0)->val;
 			}
 			perf_stat__update_shadow_stats(counter, val,
 					first_shadow_cpu(config, counter, id),
Signature · 1 line
2.34.1.400.ga245620fadb-goog
#1 lore raw
Quoted 27 lines · Ian Rogers

An uncore device may have a CPU mask that specifies one CPU per socket:

$ cat /sys/devices/uncore_imc_0/cpumask
0,18
The perf_stat_config aggr_map will map a CPU to the socket and other
aggregation values for it. Fix an error where the index into CPU mask
was being used as the index into the aggr_map. For the cpumask above the
indexes 0 and 1 are passed to aggr_map rather than the CPUs 0 and 18.
---
 tools/perf/util/stat-display.c | 7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)
tools/perf/util/stat-display.c +1 -1
index 588601000f3f..7cfad5cfec38 100644
@@ -516,7 +516,7 @@ static void printout(struct perf_stat_config *config, struct aggr_cpu_id id, int
 static void aggr_update_shadow(struct perf_stat_config *config,
 			       struct evlist *evlist)
 {
-	int cpu, s;
+	int idx, cpu, s;
 	struct aggr_cpu_id s2, id;
 	u64 val;
 	struct evsel *counter;

tools/perf/util/stat-display.c static void aggr_update_shadow(struct perf_stat_config *config, · line 530

Quoted 7 lines
tools/perf/util/stat-display.c +2 -1
@@ -525,11 +525,12 @@ static void aggr_update_shadow(struct perf_stat_config *config,
 		id = config->aggr_map->map[s];
 		evlist__for_each_entry(evlist, counter) {
 			val = 0;
-			for (cpu = 0; cpu < evsel__nr_cpus(counter); cpu++) {
+			for (idx = 0; idx < evsel__nr_cpus(counter); idx++) {
+				cpu = perf_cpu_map__cpu(evsel__cpus(counter), idx);
 				s2 = config->aggr_get_id(config, evlist->core.cpus, cpu);

Hi Ian,

This same pattern of looping over the CPUs and calling aggr_get_id() is used a couple of other times. For example in aggr_cb() and first_shadow_cpu(). Do you think these also need updating?

Or could we fix it in the aggr_get_id() functions so that they expect an index instead of CPU ID and do the conversion themselves? The callbacks do say "idx" rather than "cpu" so maybe there is still come confusion.

For example:

perf_stat__get_die_cached(struct perf_stat_config *config,
struct perf_cpu_map *map, int idx)

Quoted 8 lines
tools/perf/util/stat-display.c +1 -1
 				if (!cpu_map__compare_aggr_cpu_id(s2, id))
 					continue;
-				val += perf_counts(counter->counts, cpu, 0)->val;
+				val += perf_counts(counter->counts, idx, 0)->val;
 			}
 			perf_stat__update_shadow_stats(counter, val,
 					first_shadow_cpu(config, counter, id),
author reviewer
#2 lore raw
Quoted 41 lines · James Clark
Quoted 27 lines · Ian Rogers

An uncore device may have a CPU mask that specifies one CPU per socket:

$ cat /sys/devices/uncore_imc_0/cpumask
0,18
The perf_stat_config aggr_map will map a CPU to the socket and other
aggregation values for it. Fix an error where the index into CPU mask
was being used as the index into the aggr_map. For the cpumask above the
indexes 0 and 1 are passed to aggr_map rather than the CPUs 0 and 18.
---
 tools/perf/util/stat-display.c | 7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)
tools/perf/util/stat-display.c +1 -1
index 588601000f3f..7cfad5cfec38 100644
@@ -516,7 +516,7 @@ static void printout(struct perf_stat_config *config, struct aggr_cpu_id id, int
 static void aggr_update_shadow(struct perf_stat_config *config,
                             struct evlist *evlist)
 {
-     int cpu, s;
+     int idx, cpu, s;
      struct aggr_cpu_id s2, id;
      u64 val;
      struct evsel *counter;

tools/perf/util/stat-display.c static void aggr_update_shadow(struct perf_stat_config *config, · line 530

Quoted 7 lines
tools/perf/util/stat-display.c +2 -1
@@ -525,11 +525,12 @@ static void aggr_update_shadow(struct perf_stat_config *config,
              id = config->aggr_map->map[s];
              evlist__for_each_entry(evlist, counter) {
                      val = 0;
-                     for (cpu = 0; cpu < evsel__nr_cpus(counter); cpu++) {
+                     for (idx = 0; idx < evsel__nr_cpus(counter); idx++) {
+                             cpu = perf_cpu_map__cpu(evsel__cpus(counter), idx);
                              s2 = config->aggr_get_id(config, evlist->core.cpus, cpu);

Hi Ian,

Quoted 3 lines

This same pattern of looping over the CPUs and calling aggr_get_id() is used a couple of other times. For example in aggr_cb() and first_shadow_cpu(). Do you think these also need updating?

Thanks for the feedback James!

For first_shadow_cpu the index is translated to the the cpu via the cpu map here:
https://git.kernel.org/pub/scm/linux/kernel/git/acme/linux.git/tree/tools/perf/util/stat-display.c?h=perf/core#n343
so I think it is sound.

aggr_cb looks to have the same bug as the index is being passed as the cpu:
https://git.kernel.org/pub/scm/linux/kernel/git/acme/linux.git/tree/tools/perf/util/stat-display.c?h=perf/core#n643

Quoted 8 lines

Or could we fix it in the aggr_get_id() functions so that they expect an index instead of CPU ID and do the conversion themselves? The callbacks do say "idx" rather than "cpu" so maybe there is still come confusion.

For example:

perf_stat__get_die_cached(struct perf_stat_config *config,
struct perf_cpu_map *map, int idx)

Agreed on the naming confusion. I'm a fan of using single element structs to get type safety in code like this. I wonder here if a for_each_cpu on perf_cpu_map would clean this up best. I'll play with a v2 patch set that addresses this problem more widely.

Thanks,
Ian

Quoted 8 lines
Quoted 8 lines
tools/perf/util/stat-display.c +1 -1
                              if (!cpu_map__compare_aggr_cpu_id(s2, id))
                                      continue;
-                             val += perf_counts(counter->counts, cpu, 0)->val;
+                             val += perf_counts(counter->counts, idx, 0)->val;
                      }
                      perf_stat__update_shadow_stats(counter, val,
                                      first_shadow_cpu(config, counter, id),
author reviewer
#3 lore raw
Quoted 63 lines · Ian Rogers
Quoted 41 lines · James Clark
Quoted 27 lines · Ian Rogers

An uncore device may have a CPU mask that specifies one CPU per socket:

$ cat /sys/devices/uncore_imc_0/cpumask
0,18
The perf_stat_config aggr_map will map a CPU to the socket and other
aggregation values for it. Fix an error where the index into CPU mask
was being used as the index into the aggr_map. For the cpumask above the
indexes 0 and 1 are passed to aggr_map rather than the CPUs 0 and 18.
---
 tools/perf/util/stat-display.c | 7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)
tools/perf/util/stat-display.c +1 -1
index 588601000f3f..7cfad5cfec38 100644
@@ -516,7 +516,7 @@ static void printout(struct perf_stat_config *config, struct aggr_cpu_id id, int
 static void aggr_update_shadow(struct perf_stat_config *config,
                             struct evlist *evlist)
 {
-     int cpu, s;
+     int idx, cpu, s;
      struct aggr_cpu_id s2, id;
      u64 val;
      struct evsel *counter;

tools/perf/util/stat-display.c static void aggr_update_shadow(struct perf_stat_config *config, · line 530

Quoted 7 lines
tools/perf/util/stat-display.c +2 -1
@@ -525,11 +525,12 @@ static void aggr_update_shadow(struct perf_stat_config *config,
              id = config->aggr_map->map[s];
              evlist__for_each_entry(evlist, counter) {
                      val = 0;
-                     for (cpu = 0; cpu < evsel__nr_cpus(counter); cpu++) {
+                     for (idx = 0; idx < evsel__nr_cpus(counter); idx++) {
+                             cpu = perf_cpu_map__cpu(evsel__cpus(counter), idx);
                              s2 = config->aggr_get_id(config, evlist->core.cpus, cpu);

Hi Ian,

Quoted 3 lines

This same pattern of looping over the CPUs and calling aggr_get_id() is used a couple of other times. For example in aggr_cb() and first_shadow_cpu(). Do you think these also need updating?

Thanks for the feedback James!

For first_shadow_cpu the index is translated to the the cpu via the cpu map here:
https://git.kernel.org/pub/scm/linux/kernel/git/acme/linux.git/tree/tools/perf/util/stat-display.c?h=perf/core#n343
so I think it is sound.

aggr_cb looks to have the same bug as the index is being passed as the cpu:
https://git.kernel.org/pub/scm/linux/kernel/git/acme/linux.git/tree/tools/perf/util/stat-display.c?h=perf/core#n643

Quoted 6 lines

Or could we fix it in the aggr_get_id() functions so that they expect an index instead of CPU ID and do the conversion themselves? The callbacks do say "idx" rather than "cpu" so maybe there is still come confusion.

For example:

Quoted 7 lines
Quoted 2 lines

perf_stat__get_die_cached(struct perf_stat_config *config,
struct perf_cpu_map *map, int idx)

Agreed on the naming confusion. I'm a fan of using single element structs to get type safety in code like this. I wonder here if a for_each_cpu on perf_cpu_map would clean this up best. I'll play with a v2 patch set that addresses this problem more widely.

So the real issue is the cpu map for the evlist (populated with all CPUs) is being indexed from values from the uncore counter's cpu mask in aggr_update_shadow. Given the easy confusion of index and cpu, I'm pushing to eliminate the passing of the map and index in v2.

Thanks,
Ian

Quoted 11 lines

Thanks,
Ian

Quoted 8 lines
Quoted 8 lines
tools/perf/util/stat-display.c +1 -1
                              if (!cpu_map__compare_aggr_cpu_id(s2, id))
                                      continue;
-                             val += perf_counts(counter->counts, cpu, 0)->val;
+                             val += perf_counts(counter->counts, idx, 0)->val;
                      }
                      perf_stat__update_shadow_stats(counter, val,
                                      first_shadow_cpu(config, counter, id),