lkml.org 
[lkml]   [2020]   [Feb]   [13]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
Patch in this message
/
SubjectRe: [PATCH v3] perf stat: Show percore counts in per CPU output
From
Date


On 2/13/2020 9:20 PM, Ravi Bangoria wrote:
> Hi Jin,
>
> On 2/13/20 12:45 PM, Jin Yao wrote:
>> With this patch, for example,
>>
>>   # perf stat -e cpu/event=cpu-cycles,percore/ -a -A
>> --percore-show-thread  -- sleep 1
>>
>>    Performance counter stats for 'system wide':
>>
>>   CPU0               2,453,061      cpu/event=cpu-cycles,percore/
>>   CPU1               1,823,921      cpu/event=cpu-cycles,percore/
>>   CPU2               1,383,166      cpu/event=cpu-cycles,percore/
>>   CPU3               1,102,652      cpu/event=cpu-cycles,percore/
>>   CPU4               2,453,061      cpu/event=cpu-cycles,percore/
>>   CPU5               1,823,921      cpu/event=cpu-cycles,percore/
>>   CPU6               1,383,166      cpu/event=cpu-cycles,percore/
>>   CPU7               1,102,652      cpu/event=cpu-cycles,percore/
>>
>> We can see counts are duplicated in CPU pairs
>> (CPU0/CPU4, CPU1/CPU5, CPU2/CPU6, CPU3/CPU7).
>>
>
> I was trying this patch and I am getting bit weird results when any cpu
> is offline. Ex,
>
>   $ lscpu | grep list
>   On-line CPU(s) list:             0-4,6,7
>   Off-line CPU(s) list:            5
>
>   $ sudo ./perf stat -e cpu/event=cpu-cycles,percore/ -a -A
> --percore-show-thread -vv -- sleep 1
>     ...
>   cpu/event=cpu-cycles,percore/: 0: 23746491 1001189836 1001189836
>   cpu/event=cpu-cycles,percore/: 1: 19802666 1001291299 1001291299
>   cpu/event=cpu-cycles,percore/: 2: 24211983 1001394318 1001394318
>   cpu/event=cpu-cycles,percore/: 3: 54051396 1001516816 1001516816
>   cpu/event=cpu-cycles,percore/: 4: 6378825 1001064048 1001064048
>   cpu/event=cpu-cycles,percore/: 5: 21299840 1001166297 1001166297
>   cpu/event=cpu-cycles,percore/: 6: 13075410 1001274535 1001274535
>    Performance counter stats for 'system wide':
>   CPU0              30,125,316      cpu/event=cpu-cycles,percore/
>   CPU1              19,802,666      cpu/event=cpu-cycles,percore/
>   CPU2              45,511,823      cpu/event=cpu-cycles,percore/
>   CPU3              67,126,806      cpu/event=cpu-cycles,percore/
>   CPU4              30,125,316      cpu/event=cpu-cycles,percore/
>   CPU7              67,126,806      cpu/event=cpu-cycles,percore/
>   CPU0              30,125,316      cpu/event=cpu-cycles,percore/
>          1.001918764 seconds time elapsed
>
> I see proper result without --percore-show-thread:
>
>   $ sudo ./perf stat -e cpu/event=cpu-cycles,percore/ -a -A -vv -- sleep 1
>     ...
>   cpu/event=cpu-cycles,percore/: 0: 11676414 1001190709 1001190709
>   cpu/event=cpu-cycles,percore/: 1: 39119617 1001291459 1001291459
>   cpu/event=cpu-cycles,percore/: 2: 41821512 1001391158 1001391158
>   cpu/event=cpu-cycles,percore/: 3: 46853730 1001492799 1001492799
>   cpu/event=cpu-cycles,percore/: 4: 14448274 1001095948 1001095948
>   cpu/event=cpu-cycles,percore/: 5: 42238217 1001191187 1001191187
>   cpu/event=cpu-cycles,percore/: 6: 33129641 1001292072 1001292072
>    Performance counter stats for 'system wide':
>   S0-D0-C0             26,124,688      cpu/event=cpu-cycles,percore/
>   S0-D0-C1             39,119,617      cpu/event=cpu-cycles,percore/
>   S0-D0-C2             84,059,729      cpu/event=cpu-cycles,percore/
>   S0-D0-C3             79,983,371      cpu/event=cpu-cycles,percore/
>          1.001961563 seconds time elapsed
>
> [...]
>

Thanks so much for reporting this issue!

It looks I should use the cpu idx in print_percore_thread. I can't use
the cpu value. I have a fix:

diff --git a/tools/perf/util/stat-display.c b/tools/perf/util/stat-display.c
index 7eb3643a97ae..d89cb0da90f8 100644
--- a/tools/perf/util/stat-display.c
+++ b/tools/perf/util/stat-display.c
@@ -1149,13 +1149,11 @@ static void print_footer(struct perf_stat_config
*config)
static void print_percore_thread(struct perf_stat_config *config,
struct evsel *counter, char *prefix)
{
- int cpu, s, s2, id;
+ int s, s2, id;
bool first = true;

for (int i = 0; i < perf_evsel__nr_cpus(counter); i++) {
- cpu = perf_cpu_map__cpu(evsel__cpus(counter), i);
- s2 = config->aggr_get_id(config, evsel__cpus(counter), cpu);
-
+ s2 = config->aggr_get_id(config, evsel__cpus(counter), i);
for (s = 0; s < config->aggr_map->nr; s++) {
id = config->aggr_map->map[s];
if (s2 == id)
@@ -1164,7 +1162,7 @@ static void print_percore_thread(struct
perf_stat_config *config,

print_counter_aggrdata(config, counter, s,
prefix, false,
- &first, cpu);
+ &first, i);
}
}

With this fix, my test log:
root@kbl:~# perf stat -e cpu/event=cpu-cycles,percore/ -a -A
--percore-show-thread -- sleep 1

Performance counter stats for 'system wide':

CPU0 386,355 cpu/event=cpu-cycles,percore/
CPU1 538,325 cpu/event=cpu-cycles,percore/
CPU2 900,263 cpu/event=cpu-cycles,percore/
CPU3 1,871,488 cpu/event=cpu-cycles,percore/
CPU4 386,355 cpu/event=cpu-cycles,percore/
CPU6 900,263 cpu/event=cpu-cycles,percore/
CPU7 1,871,488 cpu/event=cpu-cycles,percore/

1.001476492 seconds time elapsed

Once I online all CPUs, the result is:

root@kbl:~# perf stat -e cpu/event=cpu-cycles,percore/ -a -A
--percore-show-thread -- sleep 1

Performance counter stats for 'system wide':

CPU0 1,371,762 cpu/event=cpu-cycles,percore/
CPU1 827,386 cpu/event=cpu-cycles,percore/
CPU2 309,934 cpu/event=cpu-cycles,percore/
CPU3 5,043,596 cpu/event=cpu-cycles,percore/
CPU4 1,371,762 cpu/event=cpu-cycles,percore/
CPU5 827,386 cpu/event=cpu-cycles,percore/
CPU6 309,934 cpu/event=cpu-cycles,percore/
CPU7 5,043,596 cpu/event=cpu-cycles,percore/

1.001535000 seconds time elapsed

>> +--percore-show-thread::
>> +The event modifier "percore" has supported to sum up the event counts
>> +for all hardware threads in a core and show the counts per core.
>> +
>> +This option with event modifier "percore" enabled also sums up the event
>> +counts for all hardware threads in a core but show the sum counts per
>> +hardware thread. This is essentially a replacement for the any bit and
>> +convenient for posting process.
>
> s/posting process/post processing/ ? :)
>

OK, thanks!

Thanks
Jin Yao

> Ravi
>

\
 
 \ /
  Last update: 2020-02-13 16:11    [W:0.068 / U:2.496 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site