* [linus:master] [perf vendor events intel] 3c989481ee: perf-sanity-tests.x86_topdown.topdown_events.fail
@ 2026-11-05 1:10 kernel test robot
2026-11-05 3:57 ` Ian Rogers
0 siblings, 1 reply; 2+ messages in thread
From: kernel test robot @ 2026-11-05 1:10 UTC (permalink / raw)
To: Ian Rogers
Cc: oe-lkp, lkp, linux-kernel, Arnaldo Carvalho de Melo,
linux-perf-users, yi1.lai
Hello,
kernel test robot noticed "perf-sanity-tests.x86_topdown.topdown_events.fail" on:
commit: 3c989481ee3cc22ac5d34e73862a4016af770710 ("perf vendor events intel: Update emeraldrapids events from 1.24 to 1.25") https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git master
[test failed on linux-next/master 7649ba73f9e7e1e04d0b46213e39c79ae09dd55c]
in testcase: perf-sanity-tests
with following parameters:
perf_compiler: gcc
group: group-02
config: x86_64-rhel-9.4-bpf
compiler: gcc-14
test machine: 256 threads 4 sockets INTEL(R) XEON(R) PLATINUM 8592+ (Emerald Rapids) with 256G memory
(please refer to attached dmesg/kmsg for entire log/backtrace)
If you fix the issue in a separate patch/commit (i.e. not just a new version of the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <yi1.lai@intel.com>
| Closes: https://lore.kernel.org/oe-lkp/202611050801.67bb4af4-lkp@intel.com
2026-10-04 03:23:58 sudo ASAN_OPTIONS=fast_unwind_on_malloc=0 /usr/src/linux-perf-x86_64-rhel-9.4-bpf-3c989481ee3cc22ac5d34e73862a4016af770710/tools/perf/perf test 77 -v
77: x86 topdown :
77.1: topdown events : Running (4 active)
---- start ----
...
event syntax error: 'cpu/mem_trans_retired.load_latency/'
\___ Bad event or PMU ...
event syntax error: '..atency/'
\___ expected numeric value ...
event syntax error: '..atency/'
\___ expected numeric value ...
WARNING: events were regrouped to match PMUs evlist after sorting/fixing: '{slots,cpu/topdown-bad-spec/}'
---- end(-1) ----
77.1: topdown events : FAILED!
77.2: topdown sorting : Ok
77.3: topdown slots injection : Ok
77.4: topdown metrics_clear no extra slots : Ok
=== Test Summary ===
Passed main tests : 0
Passed subtests : 3
Skipped tests : 0
Failed tests : 1
List of failed tests:
77.1: topdown events
The kernel config and materials to reproduce are available at:
https://download.01.org/0day-ci/archive/20261105/202611050801.67bb4af4-lkp@intel.com
--
0-DAY CI Kernel Test Service
https://github.com/intel/lkp-tests/wiki
^ permalink raw reply [flat|nested] 2+ messages in thread
* Re: [linus:master] [perf vendor events intel] 3c989481ee: perf-sanity-tests.x86_topdown.topdown_events.fail
2026-11-05 1:10 [linus:master] [perf vendor events intel] 3c989481ee: perf-sanity-tests.x86_topdown.topdown_events.fail kernel test robot
@ 2026-11-05 3:57 ` Ian Rogers
0 siblings, 0 replies; 2+ messages in thread
From: Ian Rogers @ 2026-11-05 3:57 UTC (permalink / raw)
To: kernel test robot, Edward, Thomas Falcon, Dapeng Mi, Taylor, Perry
Cc: oe-lkp, lkp, linux-kernel, Arnaldo Carvalho de Melo, linux-perf-users
On Wed, Nov 4, 2026 at 5:10 PM kernel test robot <yi1.lai@intel.com> wrote:
>
>
>
> Hello,
>
> kernel test robot noticed "perf-sanity-tests.x86_topdown.topdown_events.fail" on:
>
> commit: 3c989481ee3cc22ac5d34e73862a4016af770710 ("perf vendor events intel: Update emeraldrapids events from 1.24 to 1.25") https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git master
>
> [test failed on linux-next/master 7649ba73f9e7e1e04d0b46213e39c79ae09dd55c]
>
> in testcase: perf-sanity-tests
> with following parameters:
>
> perf_compiler: gcc
> group: group-02
>
>
> config: x86_64-rhel-9.4-bpf
> compiler: gcc-14
> test machine: 256 threads 4 sockets INTEL(R) XEON(R) PLATINUM 8592+ (Emerald Rapids) with 256G memory
>
> (please refer to attached dmesg/kmsg for entire log/backtrace)
>
>
> If you fix the issue in a separate patch/commit (i.e. not just a new version of the same patch/commit), kindly add following tags
> | Reported-by: kernel test robot <yi1.lai@intel.com>
> | Closes: https://lore.kernel.org/oe-lkp/202611050801.67bb4af4-lkp@intel.com
>
>
> 2026-10-04 03:23:58 sudo ASAN_OPTIONS=fast_unwind_on_malloc=0 /usr/src/linux-perf-x86_64-rhel-9.4-bpf-3c989481ee3cc22ac5d34e73862a4016af770710/tools/perf/perf test 77 -v
> 77: x86 topdown :
> 77.1: topdown events : Running (4 active)
> ---- start ----
> ...
> event syntax error: 'cpu/mem_trans_retired.load_latency/'
> \___ Bad event or PMU ...
The blamed commit added the event:
https://web.git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/tools/perf/pmu-events/arch/x86/emeraldrapids/memory.json?id=3c989481ee3cc22ac5d34e73862a4016af770710
```
...
+ "BriefDescription": "Counts randomly selected loads when the
latency from first dispatch to completion is greater than any number
of cycles.",
+ "Counter": "1,2,3,4,5,6,7",
+ "Data_LA": "1",
+ "EventCode": "0xcd",
+ "EventName": "MEM_TRANS_RETIRED.LOAD_LATENCY",
+ "MSRIndex": "0x3F6",
+ "PublicDescription": "Counts randomly selected loads when the
latency from first dispatch to completion is greater than any number
of cycles. Reported latency may be longer than just the memory
latency.",
+ "SampleAfterValue": "100003",
+ "UMask": "0x1"
+ },
...
```
MSRIndex 0x3F6 is ldlat:
https://web.git.kernel.org/pub/scm/linux/kernel/git/perf/perf-tools-next.git/tree/tools/perf/pmu-events/jevents.py?h=perf-tools-next#n238
and looking at this in the pmu-events-string.c I see:
```
/* offset=3016743 */
"mem_trans_retired.load_latency\000memory\000Counts randomly selected
loads when the latency from first dispatch to completion is greater
than any number of cycles Supports address when
precise\000event=0xcd,period=100003,umask=1,ldlat=None\000\00000\000\000\000\000Counts
randomly selected loads when the latency from first dispatch to
completion is greater than any number of cycles. Reported latency may
be longer than just the memory latency Supports address when
precise\000"
```
Notice the `ldlat=None`.
In jevents.py we have:
https://web.git.kernel.org/pub/scm/linux/kernel/git/perf/perf-tools-next.git/tree/tools/perf/pmu-events/jevents.py?h=perf-tools-next#n331
```
...
msr = lookup_msr(jd.get('MSRIndex'))
msrval = jd.get('MSRValue')
...
if msr:
event += f',{msr}{msrval}'
...
```
So the condition is only testing the msr and not also the msrval,
which is why `None` is placed in the output code.
This output should be in the parse_events output as the failed to
parse None should be somewhere in there like:
```
$ perf stat -e cycles/period=None/ -a sleep 1
event syntax error: 'cycles/period=None/'
\___ Bad event or PMU
Unable to find PMU or event on a PMU of 'cycles'
event syntax error: '..les/period=None/'
\___ expected numeric value
Run 'perf list' for a list of valid events
Usage: perf stat [<options>] [<command>]
-e, --event <event> event selector. use 'perf list' to list
available events
```
Anyway, I think it is sensible not to specify a ldlat value when the
MSRValue isn't given. I'll send a patch to jevents.py adding this.
Thanks,
Ian
> event syntax error: '..atency/'
> \___ expected numeric value ...
> event syntax error: '..atency/'
> \___ expected numeric value ...
> WARNING: events were regrouped to match PMUs evlist after sorting/fixing: '{slots,cpu/topdown-bad-spec/}'
> ---- end(-1) ----
> 77.1: topdown events : FAILED!
> 77.2: topdown sorting : Ok
> 77.3: topdown slots injection : Ok
> 77.4: topdown metrics_clear no extra slots : Ok
>
> === Test Summary ===
> Passed main tests : 0
> Passed subtests : 3
> Skipped tests : 0
> Failed tests : 1
> List of failed tests:
> 77.1: topdown events
>
>
> The kernel config and materials to reproduce are available at:
> https://download.01.org/0day-ci/archive/20261105/202611050801.67bb4af4-lkp@intel.com
>
>
> --
> 0-DAY CI Kernel Test Service
> https://github.com/intel/lkp-tests/wiki
>
>
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2026-11-05 3:58 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2026-11-05 1:10 [linus:master] [perf vendor events intel] 3c989481ee: perf-sanity-tests.x86_topdown.topdown_events.fail kernel test robot
2026-11-05 3:57 ` Ian Rogers
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
all inboxes | Powered by JetHome®