kernel.git/tools/perf/pmu-events, branch master

perf build: Prevent "argument list too long" error

2026-03-04T14:22:31Z

Due to a recent change, building perf may result in a build error when it is trying to "prune orphans". The file list passed to "rm" may exceed what the shell can handle. The build will then abort with an error like this: TEST [...]/arm64/build/linux-custom/tools/perf/pmu-events/metric_test.log make[5]: /bin/sh: Argument list too long make[5]: *** [pmu-events/Build:217: prune_orphans] Error 127 make[5]: *** Waiting for unfinished jobs.... make[4]: *** [Makefile.perf:773: [...]/tools/perf/pmu-events/pmu-events-in.o] Error 2 make[4]: *** Waiting for unfinished jobs.... make[3]: *** [Makefile.perf:289: sub-make] Error 2 Processing the arguments via "xargs", instead of passing the list of files directly to "rm" via the shell, prevents this issue. Fixes: 36a1b0061a584430 ("perf build: Reduce pmu-events related copying and mkdirs") Reviewed-by: Ian Rogers Signed-off-by: Markus Mayer Cc: Adrian Hunter Cc: Alexander Shishkin Cc: Ingo Molnar Cc: James Clark Cc: Jiri Olsa Cc: Mark Rutland Cc: Namhyung Kim Cc: Peter Zijlstra Signed-off-by: Arnaldo Carvalho de Melo

perf build: Reduce pmu-events related copying and mkdirs

2026-02-06T22:11:36Z

When building to an output directory the previous code would remove files and then copy the source files over. Each source file copy would have a rule to make its directory. All JSON for every architecture was considered a source file. This led to unnecessary copying as a file would be deleted and then the same file copied again, unnecessary directory making, and copying of files not used in the build. A side-effect would be a lot of build messages. This change makes it so that all computed output files are created and then compared to all files in the OUTPUT directory. By filtering out the files that would be copied, unnecessary files can be determined and then deleted - note, this is a phony target which would remake the pmu-events.c if always depended upon, and so the dependency is conditional on there being files to remove. This has some overhead as the $(OUTPUT)/pmu-events is "find" over rather than just "rm -fr", but the savings from unnecessary copying, etc. should make up for this new make overhead. The copy target just does copying but has a dependency on the directory it needs being built, avoiding repetitive mkdirs. The source files for copying only consider the JEVENTS_ARCH unless the JEVENTS_ARCH is all. The metric JSON is only generated if appropriate, rather than always being generated and jevents.py deciding whether or not to use the files. The mypy and pylint targets are fixed as variable names had changed but the rules not updated. The line count of a build with "make -C tools/perf O=/tmp/perf clean all" prior to this change was 2181 lines, after this change it is 1596 lines. This is a reduction of 585 lines or about 27%. The generated pmu-events.c for JEVENTS_ARCH "x86" and "all" were validated as being identical after this change. Signed-off-by: Ian Rogers Cc: Adrian Hunter Cc: Alexander Shishkin Cc: Ingo Molnar Cc: James Clark Cc: Jiri Olsa Cc: Leo Yan Cc: Namhyung Kim Cc: Peter Zijlstra Signed-off-by: Arnaldo Carvalho de Melo

perf jevents: Validate that all names given an Event

2026-01-28T18:18:46Z

Validate they exist in a JSON file from one directory found from one directory above the model's JSON directory. This avoids broken fallback encodings being created. Signed-off-by: Ian Rogers Tested-by: Thomas Falcon Cc: Adrian Hunter Cc: Alexander Shishkin Cc: Benjamin Gray Cc: Caleb Biggers Cc: Edward Baker Cc: Ingo Molnar Cc: James Clark Cc: Jing Zhang Cc: Jiri Olsa Cc: John Garry Cc: Leo Yan Cc: Namhyung Kim Cc: Perry Taylor Cc: Peter Zijlstra Cc: Sandipan Das Cc: Weilin Wang Cc: Xu Yang Signed-off-by: Arnaldo Carvalho de Melo

perf jevents: Add cycles breakdown metric for arm64/AMD/Intel

2026-01-28T18:18:46Z

Breakdown cycles to user, kernel and guest. Add a common_metrics.py file for such metrics. Signed-off-by: Ian Rogers Tested-by: Thomas Falcon Cc: Adrian Hunter Cc: Alexander Shishkin Cc: Benjamin Gray Cc: Caleb Biggers Cc: Edward Baker Cc: Ingo Molnar Cc: James Clark Cc: Jing Zhang Cc: Jiri Olsa Cc: John Garry Cc: Leo Yan Cc: Namhyung Kim Cc: Perry Taylor Cc: Peter Zijlstra Cc: Sandipan Das Cc: Weilin Wang Cc: Xu Yang Signed-off-by: Arnaldo Carvalho de Melo

perf jevents: Add mesh bandwidth saturation metric for Intel

2026-01-28T18:18:46Z

Memory bandwidth saturation from CBOX/CHA events present in broadwellde, broadwellx, cascadelakex, haswellx, icelakex, skylakex and snowridgex. Signed-off-by: Ian Rogers Tested-by: Thomas Falcon Cc: Adrian Hunter Cc: Alexander Shishkin Cc: Benjamin Gray Cc: Caleb Biggers Cc: Edward Baker Cc: Ingo Molnar Cc: James Clark Cc: Jing Zhang Cc: Jiri Olsa Cc: John Garry Cc: Leo Yan Cc: Namhyung Kim Cc: Perry Taylor Cc: Peter Zijlstra Cc: Sandipan Das Cc: Weilin Wang Cc: Xu Yang Signed-off-by: Arnaldo Carvalho de Melo

perf jevents: Add upi_bw metric for Intel

2026-01-28T18:18:46Z

Break down UPI read and write bandwidth using uncore_upi counters. Signed-off-by: Ian Rogers Tested-by: Thomas Falcon Cc: Adrian Hunter Cc: Alexander Shishkin Cc: Benjamin Gray Cc: Caleb Biggers Cc: Edward Baker Cc: Ingo Molnar Cc: James Clark Cc: Jing Zhang Cc: Jiri Olsa Cc: John Garry Cc: Leo Yan Cc: Namhyung Kim Cc: Perry Taylor Cc: Peter Zijlstra Cc: Sandipan Das Cc: Weilin Wang Cc: Xu Yang Signed-off-by: Arnaldo Carvalho de Melo

perf jevents: Add local/remote miss latency metrics for Intel

2026-01-28T18:18:46Z

Derive from CBOX/CHA occupancy and inserts the average latency as is provided in Intel's uncore performance monitoring reference. Signed-off-by: Ian Rogers Tested-by: Thomas Falcon Cc: Adrian Hunter Cc: Alexander Shishkin Cc: Benjamin Gray Cc: Caleb Biggers Cc: Edward Baker Cc: Ingo Molnar Cc: James Clark Cc: Jing Zhang Cc: Jiri Olsa Cc: John Garry Cc: Leo Yan Cc: Namhyung Kim Cc: Perry Taylor Cc: Peter Zijlstra Cc: Sandipan Das Cc: Weilin Wang Cc: Xu Yang Signed-off-by: Arnaldo Carvalho de Melo

perf jevents: Add C-State metrics from the PCU PMU for Intel

2026-01-28T18:18:46Z

Use occupancy events fixed in: https://lore.kernel.org/lkml/20240226201517.3540187-1-irogers@google.com/ Metrics are at the socket level referring to cores, not hyperthreads. Signed-off-by: Ian Rogers Tested-by: Thomas Falcon Cc: Adrian Hunter Cc: Alexander Shishkin Cc: Benjamin Gray Cc: Caleb Biggers Cc: Edward Baker Cc: Ingo Molnar Cc: James Clark Cc: Jing Zhang Cc: Jiri Olsa Cc: John Garry Cc: Leo Yan Cc: Namhyung Kim Cc: Perry Taylor Cc: Peter Zijlstra Cc: Sandipan Das Cc: Weilin Wang Cc: Xu Yang Signed-off-by: Arnaldo Carvalho de Melo

perf jevents: Add dir breakdown metrics for Intel

2026-01-28T18:18:46Z

Breakdown directory hit, misses and requests. The implementation uses the M2M and CHA PMUs present in server models broadwellde, broadwellx cascadelakex, emeraldrapids, icelakex, sapphirerapids and skylakex. Signed-off-by: Ian Rogers Tested-by: Thomas Falcon Cc: Adrian Hunter Cc: Alexander Shishkin Cc: Benjamin Gray Cc: Caleb Biggers Cc: Edward Baker Cc: Ingo Molnar Cc: James Clark Cc: Jing Zhang Cc: Jiri Olsa Cc: John Garry Cc: Leo Yan Cc: Namhyung Kim Cc: Perry Taylor Cc: Peter Zijlstra Cc: Sandipan Das Cc: Weilin Wang Cc: Xu Yang Signed-off-by: Arnaldo Carvalho de Melo

perf jevents: Add local/remote "mem" breakdown metrics for Intel

2026-01-28T18:18:46Z

Breakdown local and remote memory bandwidth, read and writes. The implementation uses the HA and CHA PMUs present in server models broadwellde, broadwellx cascadelakex, emeraldrapids, haswellx, icelakex, ivytown, sapphirerapids and skylakex. Signed-off-by: Ian Rogers Tested-by: Thomas Falcon Cc: Adrian Hunter Cc: Alexander Shishkin Cc: Benjamin Gray Cc: Caleb Biggers Cc: Edward Baker Cc: Ingo Molnar Cc: James Clark Cc: Jing Zhang Cc: Jiri Olsa Cc: John Garry Cc: Leo Yan Cc: Namhyung Kim Cc: Perry Taylor Cc: Peter Zijlstra Cc: Sandipan Das Cc: Weilin Wang Cc: Xu Yang Signed-off-by: Arnaldo Carvalho de Melo