* Perf test 7 fails on s390 since 6.12.0
@ 2024-12-02 14:53 Thomas Richter
2024-12-03 14:37 ` James Clark
0 siblings, 1 reply; 4+ messages in thread
From: Thomas Richter @ 2024-12-02 14:53 UTC (permalink / raw)
To: Ian Rogers, linux-perf-use.; +Cc: Sumanth Korikkar, Heiko Carstens
I just downloaded the latest kernel tree 6.13.0rc1 and observed several
failures of the perf test suite.
This test case is the first one:
# ./perf test -F 7
Unrecognized literal '#system_tsc_freq' 7: Simple expression parser : FAILED!
#
# ./perf test -Fvv 7
--- start ---
Using CPUID IBM,9175,705,ME1,3.8,002f
parsing metric: 1+1
parsing metric: FOO+BAR
lookup(FOO): val 1.000000
lookup(BAR): val 2.000000
...
parsing metric: #num_cpus
literal: #num_cpus = 8.000000
parsing metric: #num_cores
literal: #num_cores = 4.000000
parsing metric: #num_dies
literal: #num_dies = 0.000000
parsing metric: #num_packages
literal: #num_packages = 1.000000
parsing metric: #system_tsc_freq
Unrecognized literal '#system_tsc_freq'literal: #system_tsc_freq = nan
FAILED tests/expr.c:252 #system_tsc_freq == 0
---- end ----
7: Simple expression parser : FAILED!
[root@b83lp65 perf]#
I bisectred this to
# git bisect good
609aa2667f67c1588d4e741110265f74b7713177 is the first bad commit
commit 609aa2667f67c1588d4e741110265f74b7713177
Author: Ian Rogers <irogers@google.com>
Date: Tue Oct 1 20:20:12 2024 -0700
perf tool_pmu: Switch to standard pmu functions and json descriptions
Use the regular PMU approaches with tool json events to reduce the
amount of special tool_pmu code - tool_pmu__config_terms and
tool_pmu__for_each_event_cb are removed. Some functions remain, like
tool_pmu__str_to_event, as conveniences to metricgroups. Add
tool_pmu__skip_event/tool_pmu__num_skip_events to handle the case that
tool json events shouldn't appear on certain architectures. This isn't
done in jevents.py due to complexity in the empty-pmu-events.c and
when all vendor json is built into the tool.
Signed-off-by: Ian Rogers <irogers@google.com>
Acked-by: Namhyung Kim <namhyung@kernel.org>
Link: https://lore.kernel.org/r/20241002032016.333748-10-irogers@google.com
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Note this test used to work before on s390.
It now also fails on the 6.13.0-rc1.
Any hints on how to fix this? Thanks.
--
Thomas Richter, Dept 3303, IBM s390 Linux Development, Boeblingen, Germany
--
IBM Deutschland Research & Development GmbH
Vorsitzender des Aufsichtsrats: Wolfgang Wendt
Geschäftsführung: David Faller
Sitz der Gesellschaft: Böblingen / Registergericht: Amtsgericht Stuttgart, HRB 243294
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: Perf test 7 fails on s390 since 6.12.0
2024-12-02 14:53 Perf test 7 fails on s390 since 6.12.0 Thomas Richter
@ 2024-12-03 14:37 ` James Clark
2024-12-03 18:10 ` Namhyung Kim
2024-12-04 12:36 ` Thomas Richter
0 siblings, 2 replies; 4+ messages in thread
From: James Clark @ 2024-12-03 14:37 UTC (permalink / raw)
To: Thomas Richter, Ian Rogers, linux-perf-use., Athira Rajeev,
Leo Yan
Cc: Sumanth Korikkar, Heiko Carstens
On 02/12/2024 2:53 pm, Thomas Richter wrote:
> I just downloaded the latest kernel tree 6.13.0rc1 and observed several
> failures of the perf test suite.
>
> This test case is the first one:
> # ./perf test -F 7
> Unrecognized literal '#system_tsc_freq' 7: Simple expression parser : FAILED!
> #
Hi Thomas,
Did you see this thread:
https://lore.kernel.org/linux-perf-users/0F805B2F-35CC-4E0C-BD2F-84552C4C528E@linux.vnet.ibm.com/
Is that related to your issue? I looked but I couldn't find a V2 from
Athira.
The same test was also mentioned on "[GIT PULL] perf-tools changes for
v6.13", but I can't tell if it's the same type of failure or not.
James
>
> # ./perf test -Fvv 7
> --- start ---
> Using CPUID IBM,9175,705,ME1,3.8,002f
> parsing metric: 1+1
> parsing metric: FOO+BAR
> lookup(FOO): val 1.000000
> lookup(BAR): val 2.000000
> ...
>
> parsing metric: #num_cpus
> literal: #num_cpus = 8.000000
> parsing metric: #num_cores
> literal: #num_cores = 4.000000
> parsing metric: #num_dies
> literal: #num_dies = 0.000000
> parsing metric: #num_packages
> literal: #num_packages = 1.000000
> parsing metric: #system_tsc_freq
> Unrecognized literal '#system_tsc_freq'literal: #system_tsc_freq = nan
> FAILED tests/expr.c:252 #system_tsc_freq == 0
> ---- end ----
> 7: Simple expression parser : FAILED!
> [root@b83lp65 perf]#
>
> I bisectred this to
>
> # git bisect good
> 609aa2667f67c1588d4e741110265f74b7713177 is the first bad commit
> commit 609aa2667f67c1588d4e741110265f74b7713177
> Author: Ian Rogers <irogers@google.com>
> Date: Tue Oct 1 20:20:12 2024 -0700
>
> perf tool_pmu: Switch to standard pmu functions and json descriptions
>
> Use the regular PMU approaches with tool json events to reduce the
> amount of special tool_pmu code - tool_pmu__config_terms and
> tool_pmu__for_each_event_cb are removed. Some functions remain, like
> tool_pmu__str_to_event, as conveniences to metricgroups. Add
> tool_pmu__skip_event/tool_pmu__num_skip_events to handle the case that
> tool json events shouldn't appear on certain architectures. This isn't
> done in jevents.py due to complexity in the empty-pmu-events.c and
> when all vendor json is built into the tool.
>
> Signed-off-by: Ian Rogers <irogers@google.com>
> Acked-by: Namhyung Kim <namhyung@kernel.org>
> Link: https://lore.kernel.org/r/20241002032016.333748-10-irogers@google.com
> Signed-off-by: Namhyung Kim <namhyung@kernel.org>
>
>
> Note this test used to work before on s390.
> It now also fails on the 6.13.0-rc1.
>
> Any hints on how to fix this? Thanks.
>
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: Perf test 7 fails on s390 since 6.12.0
2024-12-03 14:37 ` James Clark
@ 2024-12-03 18:10 ` Namhyung Kim
2024-12-04 12:36 ` Thomas Richter
1 sibling, 0 replies; 4+ messages in thread
From: Namhyung Kim @ 2024-12-03 18:10 UTC (permalink / raw)
To: James Clark
Cc: Thomas Richter, Ian Rogers, linux-perf-use., Athira Rajeev,
Leo Yan, Sumanth Korikkar, Heiko Carstens
Hello,
On Tue, Dec 03, 2024 at 02:37:00PM +0000, James Clark wrote:
>
>
> On 02/12/2024 2:53 pm, Thomas Richter wrote:
> > I just downloaded the latest kernel tree 6.13.0rc1 and observed several
> > failures of the perf test suite.
> >
> > This test case is the first one:
> > # ./perf test -F 7
> > Unrecognized literal '#system_tsc_freq' 7: Simple expression parser : FAILED!
> > #
>
> Hi Thomas,
>
> Did you see this thread: https://lore.kernel.org/linux-perf-users/0F805B2F-35CC-4E0C-BD2F-84552C4C528E@linux.vnet.ibm.com/
>
> Is that related to your issue? I looked but I couldn't find a V2 from
> Athira.
Oh, I missed this one. I'll take a look again and let Athira proceed.
>
> The same test was also mentioned on "[GIT PULL] perf-tools changes for
> v6.13", but I can't tell if it's the same type of failure or not.
I suspect it's the same issue.
Thanks,
Namhyung
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: Perf test 7 fails on s390 since 6.12.0
2024-12-03 14:37 ` James Clark
2024-12-03 18:10 ` Namhyung Kim
@ 2024-12-04 12:36 ` Thomas Richter
1 sibling, 0 replies; 4+ messages in thread
From: Thomas Richter @ 2024-12-04 12:36 UTC (permalink / raw)
To: James Clark, Ian Rogers, linux-perf-use., Athira Rajeev, Leo Yan
Cc: Sumanth Korikkar, Heiko Carstens
On 12/3/24 15:37, James Clark wrote:
>
>
> On 02/12/2024 2:53 pm, Thomas Richter wrote:
>> I just downloaded the latest kernel tree 6.13.0rc1 and observed several
>> failures of the perf test suite.
>>
>> This test case is the first one:
>> # ./perf test -F 7
>> Unrecognized literal '#system_tsc_freq' 7: Simple expression parser : FAILED!
>> #
>
> Hi Thomas,
>
> Did you see this thread: https://lore.kernel.org/linux-perf-users/0F805B2F-35CC-4E0C-BD2F-84552C4C528E@linux.vnet.ibm.com/
>
> Is that related to your issue? I looked but I couldn't find a V2 from Athira.
>
> The same test was also mentioned on "[GIT PULL] perf-tools changes for v6.13", but I can't tell if it's the same type of failure or not.
>
> James
>
>
Thanks James,
yes it is the same issue. I applied above mentioned fix and it works again:
# ./perf test -Fvv 7# ./perf test -Fv 7
--- start ---
Using CPUID IBM,3931,704,A01,3.7,002f
division by zero
syntax error
---- end ----
7: Simple expression parser : Ok
#
For s390 you have my
Tested-By: Thomas Richter <tmricht@linux.ibm.com>
--
Thomas Richter, Dept 3303, IBM s390 Linux Development, Boeblingen, Germany
--
IBM Deutschland Research & Development GmbH
Vorsitzender des Aufsichtsrats: Wolfgang Wendt
Geschäftsführung: David Faller
Sitz der Gesellschaft: Böblingen / Registergericht: Amtsgericht Stuttgart, HRB 243294
^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2024-12-04 12:36 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2024-12-02 14:53 Perf test 7 fails on s390 since 6.12.0 Thomas Richter
2024-12-03 14:37 ` James Clark
2024-12-03 18:10 ` Namhyung Kim
2024-12-04 12:36 ` Thomas Richter
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).