Re: [PATCH v2 00/21] Update uncore topics, 3x event updates, 2x new archs.

[Ian Rogers]

On Thu, Apr 13, 2023 at 4:51 AM Arnaldo Carvalho de Melo
<acme@xxxxxxxxxx> wrote:
>
> Em Thu, Apr 13, 2023 at 08:48:35AM -0300, Arnaldo Carvalho de Melo escreveu:
> > Em Wed, Apr 12, 2023 at 10:09:50PM -0700, Ian Rogers escreveu:
> > > Update the uncore PMUs and topic classification as created by:
> > > https://github.com/intel/perfmon/pull/70
> > >
> > > Event updates stem from:
> > > https://github.com/intel/perfmon/pull/68
> > > impacting alderlake, icelakex and sapphirerapids.
> > >
> > > Grand Ridge and Sierra Forest events stem from:
> > > https://github.com/intel/perfmon/pull/69
> > >
> > > Changes generated by with PR70 in place:
> > > https://github.com/intel/perfmon/blob/main/scripts/create_perf_json.py
> > >
> > > v2. Adds improvements to uncore topics, uncore PMU name fixes and
> > >     fixes a trigraph issue from ??? being in the json.
> >
> > Som problem with your mailer? Missing some patches, I'll try b4 +
> > manual...
>
> And it is failing on the first patch:
>
> [acme@toolbox perf-tools-next]$ b4 am -P1 -ctsl --cc-trailers 20230413051011.3440849-1-irogers@xxxxxxxxxx
> Grabbing thread from lore.kernel.org/all/20230413051011.3440849-1-irogers%40google.com/t.mbox.gz
> Checking for newer revisions
> Grabbing search results from lore.kernel.org
> Analyzing 14 messages in the thread
> Checking attestation on all messages, may take a moment...
> ---
>   ✓ [PATCH v2 1/21] perf vendor events intel: Update alderlake to v1.20
>     ✓ Signed: DKIM/google.com
>     + Link: https://lore.kernel.org/r/20230413051011.3440849-2-irogers@xxxxxxxxxx
>     + Signed-off-by: Arnaldo Carvalho de Melo <acme@xxxxxxxxxx>
> ---
> Total patches: 1 (cherrypicked: 1)
> ---
> Cover: ./v2_20230412_irogers_update_uncore_topics_3x_event_updates_2x_new_archs.cover
>  Link: https://lore.kernel.org/r/20230413051011.3440849-1-irogers@xxxxxxxxxx
>  Base: not specified
>        git am ./v2_20230412_irogers_update_uncore_topics_3x_event_updates_2x_new_archs.mbx
> ⬢[acme@toolbox perf-tools-next]$        git am ./v2_20230412_irogers_update_uncore_topics_3x_event_updates_2x_new_archs.mbx
> Applying: perf vendor events intel: Update alderlake to v1.20
> error: patch failed: tools/perf/pmu-events/arch/x86/alderlake/other.json:162
> error: tools/perf/pmu-events/arch/x86/alderlake/other.json: patch does not apply
> error: patch failed: tools/perf/pmu-events/arch/x86/alderlake/pipeline.json:31
> error: tools/perf/pmu-events/arch/x86/alderlake/pipeline.json: patch does not apply
> error: patch failed: tools/perf/pmu-events/arch/x86/mapfile.csv:1
> error: tools/perf/pmu-events/arch/x86/mapfile.csv: patch does not apply
> Patch failed at 0001 perf vendor events intel: Update alderlake to v1.20
> hint: Use 'git am --show-current-patch=diff' to see the failed patch
> When you have resolved this problem, run "git am --continue".
> If you prefer to skip this patch, run "git am --skip" instead.
> To restore the original branch and stop patching, run "git am --abort".
> ⬢[acme@toolbox perf-tools-next]$ git log --oneline -10
> 4453deacf44615b2 (HEAD -> perf-tools-next, quaco/perf-tools-next, acme/tmp.perf-tools-next, acme/perf-tools-next) perf sched: Fix sched latency analysis incorrection when using 'sched:sched_wakeup'
> 9efe423e43878934 (acme/perf/core) perf pmu: Use perf_cpu_map__set_nr() in perf_pmu__cpus_match() to allow for refcnt checking
> b277851417e0149a libperf: Add a perf_cpu_map__set_nr() available as an internal function for tools/perf to use
> 1f94479edb4decdc libperf: Make perf_cpu_map__alloc() available as an internal function for tools/perf to use
> 7bb1d048bdfe1c7d perf cpumap: Use perf_cpu_map__nr(cpus) to access cpus->nr
> 4e8db2d7520f780f perf map: Add map__refcnt() accessor to use in the maps test
> 984abd349d0f76d4 perf scripts python intel-pt-events: Delete unused 'event_attr variable
> eab5051788f6111a perf top: Expand the range of multithreaded phase
> ed4da0d3de1e1781 tools headers: Remove s390 ptrace.h in check-headers.sh
> 57f14b5ae1a97537 perf pmu: zfree() expects a pointer to a pointer to zero it after freeing its contents


Sorry, my fault. These two were already picked:
24 hours perf vendor events: Update icelakex to v1.20Ian Rogers2-2/+2
24 hours perf vendor events: Update alderlake to v1.20Ian Rogers3-4/+7

but in v2 the filename/topics changed. Let me rebase on top of the v1
version and resend as v3.

Thanks,
Ian