perf test: Fix LBR test by ignoring idle task
Namhyung Kim
maintainerPATCH 1/2 perf test: Fix LBR test by ignoring idle task
I found 'perf record LBR tests' failing due to empty branch stacks.
$ perf test -v LBR ... LBR system wide any branch test Lowering default frequency rate from 4000 to 1000. Please consider tweaking /proc/sys/kernel/perf_event_max_sample_rate. [ perf record: Woken up 8 times to write data ] [ perf record: Captured and wrote 3.142 MB /tmp/__perf_test.perf.data.dgSBl (3572 samples) ] LBR system wide any branch test: 3572 samples LBR system wide any branch test [Failed empty br stack ratio exceed 2%: 3%] LBR system wide any call test Lowering default frequency rate from 4000 to 1000. Please consider tweaking /proc/sys/kernel/perf_event_max_sample_rate. [ perf record: Woken up 8 times to write data ] [ perf record: Captured and wrote 3.337 MB /tmp/__perf_test.perf.data.dgSBl (3967 samples) ] LBR system wide any call test: 3967 samples LBR system wide any call test [Failed empty br stack ratio exceed 2%: 9%] ...
The failing cases were in system-wide mode and I realized that the samples were from the idle tasks (swapper). I suspect going to/from idle state may affect the LBR contents.
If we can skip empty branch stacks from the idle tasks, the failure should go away. I can see the following output in perf report -D.
$ perf report -D | grep -m5 -A3 'branch stack: nr:0' ... -- ... branch stack: nr:0 ... thread: swapper:0 ...... dso: /proc/kcore
--
... branch stack: nr:0
... thread: swapper:0
...... dso: /proc/kcore
--
... branch stack: nr:0
... thread: DefaultEventMan:10282
...... dso: /proc/kcore
--
... branch stack: nr:0
... thread: swapper:0
...... dso: /proc/kcore
--
... branch stack: nr:0
... thread: swapper:0
...... dso: /proc/kcore
$ perf report -D | grep -c 'branch stack: nr:0' 145
$ perf report -D | grep -A3 'branch stack: nr:0' | grep thread | grep -c swapper i36
$ perf report -D | grep -A3 'branch stack: nr:0' | grep thread | grep -cv swapper 9
--- tools/perf/tests/shell/record_lbr.sh | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
tools/perf/tests/shell/record_lbr.sh +1 -1
@@ -93,7 +93,7 @@ lbr_test() { return fi - zero_nr=$(echo "$out" | grep -c 'branch stack: nr:0' || true) + zero_nr=$(echo "$out" | grep -A3 'branch stack: nr:0' | grep thread | grep -cv swapper || true) r=$(($zero_nr * 100 / $bs_nr)) if [ $r -gt $threshold ]; then echo "$test [Failed empty br stack ratio exceed $threshold%: $r%]"
Signature · 1 line
2.49.0.1015.ga840276032-goog
Namhyung Kim
maintainerPATCH 2/2 perf test: Update sysfs path for core PMU caps
While cpu is a system device, it'd be better to use a path for event_source devices when it checks PMU capability.
--- tools/perf/tests/shell/record_lbr.sh | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)
tools/perf/tests/shell/record_lbr.sh +2 -1
@@ -4,7 +4,8 @@ set -e -if [ ! -f /sys/devices/cpu/caps/branches ] && [ ! -f /sys/devices/cpu_core/caps/branches ] +if [ ! -f /sys/bus/event_source/devices/cpu/caps/branches ] && + [ ! -f /sys/bus/event_source/devices/cpu_core/caps/branches ] then echo "Skip: only x86 CPUs support LBR" exit 2
Signature · 1 line
2.49.0.1015.ga840276032-goog
Re: [PATCH 2/2] perf test: Update sysfs path for core PMU caps
Quoted 5 lines · Namhyung Kim
While cpu is a system device, it'd be better to use a path for
event_sourcedevices when it checks PMU capability.Signed-off-by Namhyung Kim <[email protected]>
Thanks,
Ian
Quoted 21 lines
--- tools/perf/tests/shell/record_lbr.sh | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)tools/perf/tests/shell/record_lbr.sh +2
-2@@ -4,7 +4,8 @@ set -e -if [ ! -f /sys/devices/cpu/caps/branches ] && [ ! -f /sys/devices/cpu_core/caps/branches ] +if [ ! -f /sys/bus/event_source/devices/cpu/caps/branches ] && + [ ! -f /sys/bus/event_source/devices/cpu_core/caps/branches ] then echo "Skip: only x86 CPUs support LBR" exit 2 --2.49.0.1015.ga840276032-goog
Re: [PATCH 2/2] perf test: Update sysfs path for core PMU caps
Quoted 8 lines · Ian Rogers
Quoted 5 lines · Namhyung Kim
While cpu is a system device, it'd be better to use a path for
event_sourcedevices when it checks PMU capability.Signed-off-by Namhyung Kim <[email protected]>Reviewed-by Ian Rogers <[email protected]>
Thanks, applied to perf-tools-next,
- Arnaldo
Quoted 58 lines · Namhyung Kim
I found 'perf record LBR tests' failing due to empty branch stacks.
$ perf test -v LBR ... LBR system wide any branch test Lowering default frequency rate from 4000 to 1000. Please consider tweaking /proc/sys/kernel/perf_event_max_sample_rate. [ perf record: Woken up 8 times to write data ] [ perf record: Captured and wrote 3.142 MB /tmp/__perf_test.perf.data.dgSBl (3572 samples) ] LBR system wide any branch test: 3572 samples LBR system wide any branch test [Failed empty br stack ratio exceed 2%: 3%] LBR system wide any call test Lowering default frequency rate from 4000 to 1000. Please consider tweaking /proc/sys/kernel/perf_event_max_sample_rate. [ perf record: Woken up 8 times to write data ] [ perf record: Captured and wrote 3.337 MB /tmp/__perf_test.perf.data.dgSBl (3967 samples) ] LBR system wide any call test: 3967 samples LBR system wide any call test [Failed empty br stack ratio exceed 2%: 9%] ...The failing cases were in system-wide mode and I realized that the samples were from the idle tasks (swapper). I suspect going to/from idle state may affect the LBR contents.
If we can skip empty branch stacks from the idle tasks, the failure should go away. I can see the following output in perf report -D.
$ perf report -D | grep -m5 -A3 'branch stack: nr:0' ... -- ... branch stack: nr:0 ... thread: swapper:0 ...... dso: /proc/kcore--
... branch stack: nr:0
... thread: swapper:0
...... dso:/proc/kcore--
... branch stack: nr:0
... thread: DefaultEventMan:10282
...... dso:/proc/kcore--
... branch stack: nr:0
... thread: swapper:0
...... dso:/proc/kcore--
... branch stack: nr:0
... thread: swapper:0
...... dso:/proc/kcore$ perf report -D | grep -c 'branch stack: nr:0' 145
Quoted 7 lines
$ perf report -D | grep -A3 'branch stack: nr:0' | grep thread | grep -c swapper i36$ perf report -D | grep -A3 'branch stack: nr:0' | grep thread | grep -cv swapper 9Signed-off-by Namhyung Kim <[email protected]>
Thanks,
Ian
Quoted 20 lines
--- tools/perf/tests/shell/record_lbr.sh | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)tools/perf/tests/shell/record_lbr.sh +1
-2@@ -93,7 +93,7 @@ lbr_test() { return fi - zero_nr=$(echo "$out" | grep -c 'branch stack: nr:0' || true) + zero_nr=$(echo "$out" | grep -A3 'branch stack: nr:0' | grep thread | grep -cv swapper || true) r=$(($zero_nr * 100 / $bs_nr)) if [ $r -gt $threshold ]; then echo "$test [Failed empty br stack ratio exceed $threshold%: $r%]" --2.49.0.1015.ga840276032-goog
Quoted 2 lines · Ian Rogers
Quoted 1 line · Namhyung Kim
I found 'perf record LBR tests' failing due to empty branch stacks.
<SNIP>
Quoted 3 lines
Quoted 1 line
Signed-off-by Namhyung Kim <[email protected]>Reviewed-by Ian Rogers <[email protected]>
Thanks, applied to perf-tools-next,
- Arnaldo