Re: [PATCH V1 00/23] perf intel-pt: Better support for perf record --cpu
From: Ian Rogers
Date: Thu May 05 2022 - 21:07:10 EST
On Thu, May 5, 2022 at 9:56 AM Adrian Hunter <adrian.hunter@xxxxxxxxx> wrote:
>
> Hi
>
> Here are V1 patches to support capturing Intel PT sideband events such as
> mmap, task, context switch, text poke etc, on every CPU even when tracing
> selected user_requested_cpus. That is, when using the perf record -C or
> --cpu option.
>
> This is needed for:
> 1. text poke: a text poke on any CPU affects all CPUs
> 2. tracing user space: a user space process can migrate between CPUs so
> mmap events that happen on a different CPU can be needed to decode a
> user_requested_cpus CPU.
>
> For example:
>
> Trace on CPU 1:
>
> perf record --kcore -C 1 -e intel_pt// &
>
> Start a task on CPU 0:
>
> taskset 0x1 testprog &
>
> Migrate it to CPU 1:
>
> taskset -p 0x2 <testprog pid>
>
> Stop tracing:
>
> kill %1
>
> Prior to these changes there will be errors decoding testprog
> in userspace because the comm and mmap events for testprog will not
> have been captured.
>
> There is quite a bit of preparation:
>
> The first patch is a small Intel PT test for system-wide side band. The
> test fails before the patches are applied, passed afterwards.
>
> perf intel-pt: Add a test for system-wide side band [new in V1]
>
> The next 5 patches stop auxtrace mixing up mmap idx between evlist and
> evsel. That is going to matter when
> evlist->all_cpus != evlist->user_requested_cpus != evsel->cpus:
>
> libperf evsel: Factor out perf_evsel__ioctl() [now applied]
> libperf evsel: Add perf_evsel__enable_thread()
> perf evlist: Use libperf functions in evlist__enable_event_idx()
> perf auxtrace: Move evlist__enable_event_idx() to auxtrace.c
> perf auxtrace: Do not mix up mmap idx
>
> The next 6 patches stop attempts to auxtrace mmap when it is not an
> auxtrace event e.g. when mmapping the CPUs on which only sideband is
> captured:
>
> libperf evlist: Remove ->idx() per_cpu parameter
> libperf evlist: Move ->idx() into mmap_per_evsel()
> libperf evlist: Add evsel as a parameter to ->idx()
> perf auxtrace: Record whether an auxtrace mmap is needed
> perf auxctrace: Add mmap_needed to auxtrace_mmap_params
> perf auxtrace: Remove auxtrace_mmap_params__set_idx() per_cpu parameter
>
> The next 5 patches switch to setting up dummy event maps before adding the
> evsel so that the evsel is subject to map propagation, primarily to cause
> addition of the evsel's CPUs to all_cpus.
>
> perf evlist: Factor out evlist__dummy_event()
> perf evlist: Add evlist__add_system_wide_dummy()
> perf record: Use evlist__add_system_wide_dummy() in record__config_text_poke()
> perf intel-pt: Use evlist__add_system_wide_dummy() for switch tracking
> perf intel-pt: Track sideband system-wide when needed
>
> The remaining patches make more significant changes.
>
> First change from using user_requested_cpus to using all_cpus where necessary:
>
> perf tools: Allow all_cpus to be a superset of user_requested_cpus
>
> Secondly, mmap all per-thread and all per-cpu events:
>
> libperf evlist: Allow mixing per-thread and per-cpu mmaps
> libperf evlist: Check nr_mmaps is correct [new in V1]
>
> Stop using system_wide flag for uncore because it will not work anymore:
>
> perf stat: Add requires_cpu flag for uncore
> libperf evsel: Add comments for booleans [new in V1]
>
> Finally change map propagation so that system-wide events retain their cpus and
> (dummy) threads:
>
> perf tools: Allow system-wide events to keep their own CPUs
> perf tools: Allow system-wide events to keep their own threads
>
>
> Changes in V1:
>
> perf intel-pt: Add a test for system-wide side band
> New patch
>
> libperf evsel: Factor out perf_evsel__ioctl()
> Dropped because it has been applied.
>
> libperf evsel: Add perf_evsel__enable_thread()
> Rename variable i -> idx
>
> perf auxtrace: Do not mix up mmap idx
> Rename variable cpu to cpu_map_idx
>
> perf tools: Allow all_cpus to be a superset of user_requested_cpus
> Add Acked-by: Ian Rogers <irogers@xxxxxxxxxx>
>
> libperf evlist: Allow mixing per-thread and per-cpu mmaps
> Fix perf_evlist__nr_mmaps() calculation
>
> libperf evlist: Check nr_mmaps is correct
> New patch
>
> libperf evsel: Add comments for booleans
> New patch
>
> perf tools: Allow system-wide events to keep their own CPUs
> perf tools: Allow system-wide events to keep their own threads
>
>
> Adrian Hunter (23):
> perf intel-pt: Add a test for system-wide side band
> libperf evsel: Add perf_evsel__enable_thread()
> perf evlist: Use libperf functions in evlist__enable_event_idx()
> perf auxtrace: Move evlist__enable_event_idx() to auxtrace.c
> perf auxtrace: Do not mix up mmap idx
> libperf evlist: Remove ->idx() per_cpu parameter
> libperf evlist: Move ->idx() into mmap_per_evsel()
> libperf evlist: Add evsel as a parameter to ->idx()
> perf auxtrace: Record whether an auxtrace mmap is needed
> perf auxtrace: Add mmap_needed to auxtrace_mmap_params
> perf auxtrace: Remove auxtrace_mmap_params__set_idx() per_cpu parameter
> perf evlist: Factor out evlist__dummy_event()
> perf evlist: Add evlist__add_system_wide_dummy()
> perf record: Use evlist__add_system_wide_dummy() in record__config_text_poke()
> perf intel-pt: Use evlist__add_system_wide_dummy() for switch tracking
> perf intel-pt: Track sideband system-wide when needed
> perf tools: Allow all_cpus to be a superset of user_requested_cpus
> libperf evlist: Allow mixing per-thread and per-cpu mmaps
> libperf evlist: Check nr_mmaps is correct
> perf stat: Add requires_cpu flag for uncore
> libperf evsel: Add comments for booleans
> perf tools: Allow system-wide events to keep their own CPUs
> perf tools: Allow system-wide events to keep their own threads
>
> tools/lib/perf/evlist.c | 80 ++++++++++-------------
> tools/lib/perf/evsel.c | 10 +++
> tools/lib/perf/include/internal/evlist.h | 3 +-
> tools/lib/perf/include/internal/evsel.h | 9 +++
> tools/lib/perf/include/internal/mmap.h | 5 ++
> tools/lib/perf/include/perf/evsel.h | 1 +
> tools/perf/arch/arm/util/cs-etm.c | 1 +
> tools/perf/arch/arm64/util/arm-spe.c | 1 +
> tools/perf/arch/s390/util/auxtrace.c | 1 +
> tools/perf/arch/x86/util/intel-bts.c | 1 +
> tools/perf/arch/x86/util/intel-pt.c | 32 ++++------
> tools/perf/builtin-record.c | 39 +++++-------
> tools/perf/builtin-stat.c | 5 +-
> tools/perf/tests/shell/test_intel_pt.sh | 71 +++++++++++++++++++++
> tools/perf/util/auxtrace.c | 31 +++++++--
> tools/perf/util/auxtrace.h | 8 ++-
> tools/perf/util/evlist.c | 106 +++++++++++++++----------------
> tools/perf/util/evlist.h | 7 +-
> tools/perf/util/evsel.c | 1 +
> tools/perf/util/evsel.h | 1 +
> tools/perf/util/mmap.c | 4 +-
> tools/perf/util/parse-events.c | 2 +-
> 22 files changed, 259 insertions(+), 160 deletions(-)
> create mode 100755 tools/perf/tests/shell/test_intel_pt.sh
>
>
> Regards
> Adrian
Thanks Adrian, other than a minor build breakage, and some possible
function name and comment tweaks, I'm happy with the change. I did
'perf test' on my SkylakeX and things look good too. One last minor
thing is the test_intel_pt.sh doesn't test --per-thread mode, this
might be good to cover for CPU map propagation issues.
Thanks,
Ian