kernel/git/next/linux-next.git/tools/perf/tests/switch-tracking.c, branch master

perf evlist: Add reference count checking

2026-06-30T17:17:11+00:00

Now the evlist is reference counted, add reference count checking so that gets and puts are paired and easy to debug. Reference count checking is documented here: https://perfwiki.github.io/main/reference-count-checking/ This large patch is adding accessors to evlist functions and switching to their use. There was some minor renaming as evlist__mmap is now an accessor to the mmap variable, and the original evlist__mmap is renamed to evlist__do_mmap. Signed-off-by: Ian Rogers Cc: Adrian Hunter Cc: Alice Rogers Cc: Dapeng Mi Cc: Ingo Molnar Cc: James Clark Cc: Leo Yan Cc: Namhyung Kim Cc: Peter Zijlstra Cc: Thomas Richter Signed-off-by: Arnaldo Carvalho de Melo Signed-off-by: Namhyung Kim

perf evlist: Add reference count

2026-06-30T17:17:10+00:00

This a no-op for most of the perf tool. The reference count is set to 1 at allocation, the put will see the 1, decrement it and perform the delete. The purpose for adding the reference count is for the python code. Prior to this change the python code would clone evlists, but this has issues if events are opened, etc. This change adds a reference count for the evlists and a later change will add it to evsels. The combination is needed for the python code to operate correctly (not hit asserts in the evsel clone), but the changes are broken apart for the sake of smaller patches. Assisted-by: Gemini:gemini-3.1-pro-preview Signed-off-by: Ian Rogers Cc: Adrian Hunter Cc: Alice Rogers Cc: Dapeng Mi Cc: Ingo Molnar Cc: James Clark Cc: Leo Yan Cc: Namhyung Kim Cc: Peter Zijlstra Cc: Thomas Richter Signed-off-by: Arnaldo Carvalho de Melo Signed-off-by: Namhyung Kim

perf evlist: Try to avoid computing evsel from sample

2026-05-20T19:39:40+00:00

As struct perf_sample now directly contains its own resolved evsel pointer, passing the evsel separately is redundant and clutters the interface. Remove the redundant evsel parameter from evlist-specific handlers and structures, ensuring the tool always directly accesses the evsel bound to the sample. This simplifies the API signatures and eliminates the risk of passing an inconsistent evsel. Signed-off-by: Ian Rogers Acked-by: Namhyung Kim Cc: Adrian Hunter Cc: Albert Ou Cc: Alexander Shishkin Cc: Alexandre Ghiti Cc: Andi Kleen Cc: Andrew Jones Cc: Anup Patel Cc: Athira Rajeev Cc: Blake Jones Cc: Chen Ni Cc: Chun-Tse Shao Cc: Dapeng Mi Cc: Derek Foreman Cc: Dmitriy Vyukov Cc: Dr. David Alan Gilbert Cc: Howard Chu Cc: Hrishikesh Suresh Cc: Ingo Molnar Cc: James Clark Cc: Jiri Olsa Cc: Krzysztof Łopatowski Cc: Leo Yan Cc: Palmer Dabbelt Cc: Paul Walmsley Cc: Peter Zijlstra Cc: Quan Zhou Cc: Ravi Bangoria Cc: Swapnil Sapkal Cc: Thomas Falcon Cc: Tianyou Li Cc: Yujie Liu Cc: tanze Signed-off-by: Arnaldo Carvalho de Melo

perf evsel: Refactor evsel tracepoint sample accessors perf_sample

2026-05-20T19:35:59+00:00

The evsel argument to evsel__intval, evsel__rawptr, and similar functions, is unnecessary as it can be read from the sample. Remove the evsel and rename the function to match that the data is coming from the sample. Signed-off-by: Ian Rogers Acked-by: Namhyung Kim Cc: Adrian Hunter Cc: Albert Ou Cc: Alexander Shishkin Cc: Alexandre Ghiti Cc: Andi Kleen Cc: Andrew Jones Cc: Anup Patel Cc: Athira Rajeev Cc: Blake Jones Cc: Chen Ni Cc: Chun-Tse Shao Cc: Dapeng Mi Cc: Derek Foreman Cc: Dmitriy Vyukov Cc: Dr. David Alan Gilbert Cc: Howard Chu Cc: Hrishikesh Suresh Cc: Ingo Molnar Cc: James Clark Cc: Jiri Olsa Cc: Krzysztof Łopatowski Cc: Leo Yan Cc: Palmer Dabbelt Cc: Paul Walmsley Cc: Peter Zijlstra Cc: Quan Zhou Cc: Ravi Bangoria Cc: Swapnil Sapkal Cc: Thomas Falcon Cc: Tianyou Li Cc: Yujie Liu Cc: tanze Signed-off-by: Arnaldo Carvalho de Melo

perf sample: Make sure perf_sample__init/exit are used

2026-04-06T06:12:15+00:00

The deferred stack trace code wasn't using perf_sample__init/exit. Add the deferred stack trace clean up to perf_sample__exit which requires proper NULL initialization in perf_sample__init. Make the perf_sample__exit robust to being called more than once by using zfree. Make the error paths in evsel__parse_sample exit the sample. Add a merged_callchain boolean to capture that callchain is allocated, deferred_callchain doen't suffice for this. Pack the struct variables to avoid padding bytes for this. Similiarly powerpc_vpadtl_sample wasn't using perf_sample__init/exit, use it for consistency and potential issues with uninitialized variables. Similarly guest_session__inject_events in builtin-inject wasn't using perf_sample_init/exit. The lifetime management for fetched events is somewhat complex there, but when an event is fetched the sample should be initialized and needs exiting on error. The sample may be left in place so that future injects have access to it. Signed-off-by: Ian Rogers Signed-off-by: Namhyung Kim

perf test: Switch cycles event to cpu-cycles

2025-10-15T14:59:12+00:00

Without a PMU perf matches an event against any PMU with the event. Unfortunately some PMU drivers advertise a "cycles" event which is typically just a core event. As tests assume a core event, switch to use "cpu-cycles" that avoids the overloaded "cycles" event on troublesome PMUs and is so far not overloaded. Note, on x86 this changes a legacy event into a sysfs one. Signed-off-by: Ian Rogers Tested-by: James Clark Signed-off-by: Namhyung Kim

perf thread_map: Remove uid options

2025-06-09T18:18:18+00:00

Now the target doesn't have a uid, it is handled through BPF filters, remove the uid options to thread_map creation. Tidy up the functions used in tests to avoid passing unused arguments. Signed-off-by: Ian Rogers Link: https://lore.kernel.org/r/20250604174545.2853620-11-irogers@google.com Signed-off-by: Namhyung Kim

perf tests switch-tracking: Fix timestamp comparison

2025-05-24T14:40:47+00:00

The test might fail on the Arm64 platform with the error: # perf test -vvv "Track with sched_switch" Missing sched_switch events # The issue is caused by incorrect handling of timestamp comparisons. The comparison result, a signed 64-bit value, was being directly cast to an int, leading to incorrect sorting for sched events. The case does not fail everytime, usually I can trigger the failure after run 20 ~ 30 times: # while true; do perf test "Track with sched_switch"; done 106: Track with sched_switch : Ok 106: Track with sched_switch : Ok 106: Track with sched_switch : Ok 106: Track with sched_switch : Ok 106: Track with sched_switch : Ok 106: Track with sched_switch : Ok 106: Track with sched_switch : Ok 106: Track with sched_switch : Ok 106: Track with sched_switch : Ok 106: Track with sched_switch : Ok 106: Track with sched_switch : Ok 106: Track with sched_switch : Ok 106: Track with sched_switch : Ok 106: Track with sched_switch : Ok 106: Track with sched_switch : FAILED! 106: Track with sched_switch : Ok 106: Track with sched_switch : Ok 106: Track with sched_switch : Ok 106: Track with sched_switch : Ok 106: Track with sched_switch : Ok 106: Track with sched_switch : Ok 106: Track with sched_switch : Ok 106: Track with sched_switch : Ok 106: Track with sched_switch : FAILED! 106: Track with sched_switch : Ok 106: Track with sched_switch : Ok I used cross compiler to build Perf tool on my host machine and tested on Debian / Juno board. Generally, I think this issue is not very specific to GCC versions. As both internal CI and my local env can reproduce the issue. My Host Build compiler: # aarch64-linux-gnu-gcc --version aarch64-linux-gnu-gcc (Ubuntu 13.3.0-6ubuntu2~24.04) 13.3.0 Juno Board: # lsb_release -a No LSB modules are available. Distributor ID: Debian Description: Debian GNU/Linux 12 (bookworm) Release: 12 Codename: bookworm Fix this by explicitly returning 0, 1, or -1 based on whether the result is zero, positive, or negative. Fixes: d44bc558297222d9 ("perf tests: Add a test for tracking with sched_switch") Reviewed-by: Ian Rogers Signed-off-by: Leo Yan Cc: Adrian Hunter Cc: James Clark Cc: Kan Liang Cc: Namhyung Kim Link: https://lore.kernel.org/r/20250331172759.115604-1-leo.yan@arm.com Signed-off-by: Arnaldo Carvalho de Melo

perf sample: Make user_regs and intr_regs optional

2025-02-13T04:06:11+00:00

The struct dump_regs contains 512 bytes of cache_regs, meaning the two values in perf_sample contribute 1088 bytes of its total 1384 bytes size. Initializing this much memory has a cost reported by Tavian Barnes as about 2.5% when running `perf script --itrace=i0`: https://lore.kernel.org/lkml/d841b97b3ad2ca8bcab07e4293375fb7c32dfce7.1736618095.git.tavianator@tavianator.com/ Adrian Hunter replied that the zero initialization was necessary and couldn't simply be removed. This patch aims to strike a middle ground of still zeroing the perf_sample, but removing 79% of its size by make user_regs and intr_regs optional pointers to zalloc-ed memory. To support the allocation accessors are created for user_regs and intr_regs. To support correct cleanup perf_sample__init and perf_sample__exit functions are created and added throughout the code base. Signed-off-by: Ian Rogers Link: https://lore.kernel.org/r/20250113194345.1537821-1-irogers@google.com Signed-off-by: Namhyung Kim

perf tests switch-tracking: Set this test to run exclusively

2024-12-13T21:30:48+00:00

This test was failing when run with the default 'perf test' mode, which is to run multiple regression tests in parallel. Since it checks system_wide mode, set it to run in exclusive mode. Reviewed-by: Ian Rogers Cc: Adrian Hunter Cc: James Clark Cc: Jiri Olsa Cc: Kan Liang Cc: Namhyung Kim Cc: Ravi Bangoria Link: https://lore.kernel.org/lkml/Z1yPYqYYs_isO1PJ@x1 Signed-off-by: Arnaldo Carvalho de Melo