[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <b11cdd84-47c6-4cc4-b950-514ff6a3a777@linaro.org>
Date: Fri, 3 Oct 2025 12:26:55 +0100
From: James Clark <james.clark@...aro.org>
To: Ian Rogers <irogers@...gle.com>
Cc: 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>, Adrian Hunter <adrian.hunter@...el.com>,
Kan Liang <kan.liang@...ux.intel.com>, Xu Yang <xu.yang_2@....com>,
Thomas Falcon <thomas.falcon@...el.com>, Andi Kleen <ak@...ux.intel.com>,
linux-kernel@...r.kernel.org, linux-perf-users@...r.kernel.org,
Atish Patra <atishp@...osinc.com>, Beeman Strong <beeman@...osinc.com>,
Leo Yan <leo.yan@....com>, Vince Weaver <vincent.weaver@...ne.edu>
Subject: Re: [PATCH v6 00/28] Legacy hardware/cache events as json
On 03/10/2025 10:20 am, James Clark wrote:
>
>
> On 02/10/2025 9:10 pm, Ian Rogers wrote:
>> On Thu, Oct 2, 2025 at 10:58 AM Ian Rogers <irogers@...gle.com> wrote:
>>> From bisecting, this change came from commit 9eac5612da1c ("perf stat:
>>> Don't skip failing group events"):
>>> https://lore.kernel.org/lkml/20250825211204.2784695-3-
>>> irogers@...gle.com/
>>> Taking a look.
>>
>> I sent a fix:
>> https://lore.kernel.org/linux-perf-users/20251002200604.1792141-2-
>> irogers@...gle.com/
>> that makes it so that only if all events are unsupported that perf
>> stat exits - this is addressing a bunch of latent issues and fixes the
>> perf-tools-next regression. I don't think there is a difference
>> between v3 and v6 wrt this behavior, I think you were probably just
>> cherry-picking the v6 patches onto a newer tree. When those 2 patches
>> land I can rebase this series on them and drop the first patch of this
>> series.
>>
>> Thanks,
>> Ian
>
> That is true about the cherry picking. I could only apply V3 to perf-
> tools and V6 to perf-tools-next, that was just a guess without the hash
> in the cover letter.
>
> I did test both unpatched perf-tools and perf-tools-next to confirm that
> the original behavior was the same in those. Once I saw it was the same
> I assumed the difference was only in the V3 vs V6. But you're right it
> could be because of the base.
>
> I'll retest with the fix applied
Yeah the behavior seems good now after testing with that fix applied and
the first patch of this set dropped.
Will wait for the rebase and then do more testing.
Powered by blists - more mailing lists