Re: [PATCH 2/4] perf vendor events amd: Add Zen 6 uncore events

[Arnaldo Carvalho de Melo]

On Tue, Jan 13, 2026 at 12:41:42PM -0800, Ian Rogers wrote:
> On Tue, Jan 13, 2026 at 12:27 PM Arnaldo Carvalho de Melo
> <acme@xxxxxxxxxx> wrote:
> >
> > On Thu, Jan 08, 2026 at 01:22:15PM +0530, Sandipan Das wrote:
> > > Add uncore events taken from Section 1.6 "L3 Cache Performance Monitor
> > > Counters" and Section 2.2 "UMC Performance Monitor Events" of the
> > > Performance Monitor Counters for AMD Family 1Ah Model 50h-57h Processors
> > > document available at the link below.
> > >
> > > This constitutes events which capture L3 cache and UMC command activity.
> >
> >   LD      /tmp/build/perf-tools-next/perf-util-in.o
> >   AR      /tmp/build/perf-tools-next/libperf-util.a
> >   CC      /tmp/build/perf-tools-next/pmu-events/pmu-events.o
> > /tmp/build/perf-tools-next/pmu-events/pmu-events.c:30902:37: error: ‘pmu_events__amdzen6’ defined but not used [-Werror=unused-const-variable=]
> > 30902 | static const struct pmu_table_entry pmu_events__amdzen6[] = {
> >       |                                     ^~~~~~~~~~~~~~~~~~~
> > cc1: all warnings being treated as errors
> > make[3]: *** [pmu-events/Build:89: /tmp/build/perf-tools-next/pmu-events/pmu-events.o] Error 1
> > make[2]: *** [Makefile.perf:772: /tmp/build/perf-tools-next/pmu-events/pmu-events-in.o] Error 2
> > make[1]: *** [Makefile.perf:288: sub-make] Error 2
> > make: *** [Makefile:119: install-bin] Error 2
> > make: Leaving directory '/home/acme/git/perf-tools-next/tools/perf'
> > ⬢ [acme@toolbx perf-tools-next]$
> >
> > When building at this patch (2/4) it breaks, lemme see if at the end it
> > works, but even then this can't stand this way, as we break bisection...
> >
> > Then, on 3/4 we get:
> >
> >   CC      /tmp/build/perf-tools-next/pmu-events/pmu-events.o
> > /tmp/build/perf-tools-next/pmu-events/pmu-events.c:30992:37: error: ‘pmu_metrics__amdzen6’ defined but not used [-Werror=unused-const-variable=]
> > 30992 | static const struct pmu_table_entry pmu_metrics__amdzen6[] = {
> >       |                                     ^~~~~~~~~~~~~~~~~~~~
> > /tmp/build/perf-tools-next/pmu-events/pmu-events.c:30908:37: error: ‘pmu_events__amdzen6’ defined but not used [-Werror=unused-const-variable=]
> > 30908 | static const struct pmu_table_entry pmu_events__amdzen6[] = {
> >       |                                     ^~~~~~~~~~~~~~~~~~~
> > cc1: all warnings being treated as errors
> > make[3]: *** [pmu-events/Build:89: /tmp/build/perf-tools-next/pmu-events/pmu-events.o] Error 1
> > make[2]: *** [Makefile.perf:772: /tmp/build/perf-tools-next/pmu-events/pmu-events-in.o] Error 2
> > make[1]: *** [Makefile.perf:288: sub-make] Error 2
> > make: *** [Makefile:119: install-bin] Error 2
> > make: Leaving directory '/home/acme/git/perf-tools-next/tools/perf'
> > ⬢ [acme@toolbx perf-tools-next]$
> >
> > Finally, on 4/4 everything builds.
> >
> > Can you please take a look to check how we can keep all bisection happy?
> 
> So patch 4 adds the mapfile.csv change and that's the thing that
> creates the references to the zen6 tables that is failing when
> building with the earlier patches. If patch 4 is made patch 1 then the
> series builds at each step (1,2,3,4) for me.

Trying that now... Did the trick, thanks! Applied.

- Arnaldo