linux-kernel - Re: [PATCH v2] perf vendor events arm64: Add FUJITSU-MONAKA pmu event

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <cb11fe5e-e6ab-4653-b925-4c3cb5273b43@linaro.org>
Date: Tue, 17 Dec 2024 09:58:07 +0000
From: James Clark <james.clark@...aro.org>
To: "Yoshihiro Furudera (Fujitsu)" <fj5100bi@...itsu.com>
Cc: 'John Garry' <john.g.garry@...cle.com>, 'Will Deacon' <will@...nel.org>,
 'Mike Leach' <mike.leach@...aro.org>, 'Leo Yan' <leo.yan@...ux.dev>,
 'Peter Zijlstra' <peterz@...radead.org>, 'Ingo Molnar' <mingo@...hat.com>,
 'Arnaldo Carvalho de Melo' <acme@...nel.org>,
 'Namhyung Kim' <namhyung@...nel.org>, 'Mark Rutland' <mark.rutland@....com>,
 'Alexander Shishkin' <alexander.shishkin@...ux.intel.com>,
 'Jiri Olsa' <jolsa@...nel.org>, 'Ian Rogers' <irogers@...gle.com>,
 'Adrian Hunter' <adrian.hunter@...el.com>,
 "'Liang, Kan'" <kan.liang@...ux.intel.com>,
 'Ilkka Koskinen' <ilkka@...amperecomputing.com>,
 'Xu Yang' <xu.yang_2@....com>, 'Jing Zhang' <renyu.zj@...ux.alibaba.com>,
 'Lucas Stach' <l.stach@...gutronix.de>,
 "'linux-arm-kernel@...ts.infradead.org'"
 <linux-arm-kernel@...ts.infradead.org>,
 "'linux-perf-users@...r.kernel.org'" <linux-perf-users@...r.kernel.org>,
 "'linux-kernel@...r.kernel.org'" <linux-kernel@...r.kernel.org>,
 "Akio Kakuno (Fujitsu)" <fj3333bs@...itsu.com>,
 "'nick.forrington@....com'" <nick.forrington@....com>
Subject: Re: [PATCH v2] perf vendor events arm64: Add FUJITSU-MONAKA pmu event



On 17/12/2024 6:35 am, Yoshihiro Furudera (Fujitsu) wrote:
> Hi, James
> 
>>> Also in the common files, we're using the description strings from the
>>> Arm ARM. I noticed that the ones from your spec are slightly different.
>>> This is ok for now, but if we add any new Arm cores that use the same
>>> events this description may get overwritten. For example
>>> ASE_SVE_INST_SPEC in the Arm ARM is currently "Operation speculatively
>> executed, Advanced SIMD or SVE".
>>>
>>> If you have any actual relevant details that are different from the
>>> common string, you should put them in .../arch/arm64/fujitsu/monaka.
>>> But if you are ok with the potential overwrite (which looks like it
>>> should be ok) then you can leave them as is.
>>
>> I think it's okay to overwrite it, but I'll check the specs.
> 
> Regarding this matter, the descriptions of the events added to the common files
> (common-and-microarch.json, recommended.json) will be modified to match Arm ARM.
> Then, we will write current descriptions as "BriefDescription" in .../arch/arm64/fujitsu/monaka.
> We will post this as a v3 patch later.
> 
>>
>>>
>>> [...]
>>>
>>>> diff --git
>>>> a/tools/perf/pmu-events/arch/arm64/fujitsu/monaka/general.json
>>>> b/tools/perf/pmu-events/arch/arm64/fujitsu/monaka/general.json
>>>> new file mode 100644
>>>> index 000000000000..80bf17fb8f4c
>>>> --- /dev/null
>>>> +++ b/tools/perf/pmu-events/arch/arm64/fujitsu/monaka/general.json
>>>> @@ -0,0 +1,8 @@
>>>> +[
>>>> +    {
>>>> +        "ArchStdEvent": "CPU_CYCLES"
>>>> +    },
>>>
>>> Other cores put CPU_CYCLES in bus.json. For user friendlyness I would
>>> put the common ones into the same existing groups. The same issue
>>> applies for some other common events.
>>>
>>
>> I will update as suggested.
> 
> MONAKA's event grouping is based on the grouping of "neoverse n1/n2".
> We found that the groups in "neoverse n1/n2" may differ from those in other cores.
> For example, "CPU_CYCLES" is classified in general.json in "neoverse n1/n2",
> but in bus.json in many other cores.
> 
> As a result, it is not possible to completely match the event groups of other cores.
> We will be checking the current grouping of events and modifying them as much as possible
> to match the other core, and will post this later as a v3 patch.
> 
> Best Regards,
> Yoshihiro Furudera

Hmmm ok, I didn't realize it was already inconsistent. Still it's better 
to try to not introduce a third variant though.