use_stdio was associated with struct perf_data and not perf_data_file
meaning there was implicit use of fd rather than fptr that may not be
safe. For example, in perf_data_file__write. Reorganize perf_data_file
to better abstract use_stdio, add kernel-doc and more consistently use
perf_data__ accessors so that use_stdio is better respected.
Signed-off-by: Ian Rogers <irogers@google.com>
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Instead of using zalloc(nr_entries * sizeof_entry) that is what calloc()
does.
In some places where linux/zalloc.h isn't needed, remove it, add when
needed and was getting it indirectly.
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
The index into the cpumap array and the number of entries within the
array can never be negative, so let's make them unsigned. This is
prompted by reports that gcc 13 with -O6 is giving a
alloc-size-larger-than errors. The change makes the cpumap changes and
then updates the declaration of index variables throughout perf and
libperf to be unsigned. The two things are hard to separate as
compiler warnings about mixing signed and unsigned types breaks the
build.
Reported-by: Chingbin Li <liqb365@163.com>
Closes: https://lore.kernel.org/lkml/20260212025127.841090-1-liqb365@163.com/
Tested-by: Chingbin Li <liqb365@163.com>
Signed-off-by: Ian Rogers <irogers@google.com>
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Frame pointer callchains are not supported on s390 and dwarf
callchains are only supported on software events.
Switch the default event from the hardware 'cycles' event to the
software 'cpu-clock' or 'task-clock' on s390 if callchains are
enabled. Move some of the target initialization earlier in builtin-top
and builtin-record, so it is ready for use by evlist__new_default.
If frame pointer callchains are requested on s390 show a
warning. Modify the '-g' option of `perf top` and `perf record` to
default to dwarf callchains on s390.
Signed-off-by: Ian Rogers <irogers@google.com>
Tested-by: Thomas Richter <tmricht@linux.ibm.com>
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
record_opts__parse_callchain is shared by builtin-record and
builtin-trace, it is declared in callchain.h. Move the declaration to
callchain.c for consistency with the header. In other cases make the
option callback a small static stub that then calls into callchain.c.
Make the no argument '-g' callchain option just a short-cut for
'--call-graph fp' so that there is consistency in how the arguments
are handled. This requires the const char* string to be strdup-ed in
__parse_callchain_report_opt. For consistency also make
parse_callchain_record use strdup and remove some unnecessary
casts. Also, be more explicit about the '-g' behavior if there is a
.perfconfig file setting.
Signed-off-by: Ian Rogers <irogers@google.com>
Tested-by: Thomas Richter <tmricht@linux.ibm.com>
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Since commit ee27476fa3 ("perf record: Skip don't fail for events
that don't open"), if a user does not have permission to access a PMU
event, perf reports:
perf record -e cs_etm// -C 3 -- ls
Error:
Failure to open event 'cs_etm//u' on PMU 'cs_etm' which will be removed.
No fallback found for 'cs_etm//u' for error 13
Error:
Failure to open event 'dummy:u' on PMU 'software' which will be removed.
No fallback found for 'dummy:u' for error 13
Error:
Failure to open any events for recording.
The log is not very helpful, as no clear indication of what "error 13"
means or how to address the issue.
This commit restores evsel__open_strerror() to generate a readable error
message and print it out:
perf record -e cs_etm// -C 3 -- ls
Error:
Failure to open event 'cs_etm//' on PMU 'cs_etm' which will be removed.
Access to performance monitoring and observability operations is limited.
Consider adjusting /proc/sys/kernel/perf_event_paranoid setting to open
access to performance monitoring and observability operations for processes
without CAP_PERFMON, CAP_SYS_PTRACE or CAP_SYS_ADMIN Linux capability.
More information can be found at 'Perf events and tool security' document:
https://www.kernel.org/doc/html/latest/admin-guide/perf-security.html
perf_event_paranoid setting is 1:
-1: Allow use of (almost) all events by all users
Ignore mlock limit after perf_event_mlock_kb without CAP_IPC_LOCK
>= 0: Disallow raw and ftrace function tracepoint access
>= 1: Disallow CPU event access
>= 2: Disallow kernel profiling
To make the adjusted perf_event_paranoid setting permanent preserve it
in /etc/sysctl.conf (e.g. kernel.perf_event_paranoid = <setting>)
Error:
Failure to open event 'dummy:u' on PMU 'software' which will be removed.
Access to performance monitoring and observability operations is limited.
Consider adjusting /proc/sys/kernel/perf_event_paranoid setting to open
access to performance monitoring and observability operations for processes
without CAP_PERFMON, CAP_SYS_PTRACE or CAP_SYS_ADMIN Linux capability.
More information can be found at 'Perf events and tool security' document:
https://www.kernel.org/doc/html/latest/admin-guide/perf-security.html
perf_event_paranoid setting is 1:
-1: Allow use of (almost) all events by all users
Ignore mlock limit after perf_event_mlock_kb without CAP_IPC_LOCK
>= 0: Disallow raw and ftrace function tracepoint access
>= 1: Disallow CPU event access
>= 2: Disallow kernel profiling
To make the adjusted perf_event_paranoid setting permanent preserve it
in /etc/sysctl.conf (e.g. kernel.perf_event_paranoid = <setting>)
Error:
Failure to open any events for recording.
Reviewed-by: Ian Rogers <irogers@google.com>
Signed-off-by: Leo Yan <leo.yan@arm.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: James Clark <james.clark@linaro.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
The NO_AUXTRACE build option was used when the __get_cpuid feature
test failed or if it was provided on the command line. The option no
longer avoids a dependency on a library and so having the option is
just adding complexity to the code base. Remove the option
CONFIG_AUXTRACE from Build files and HAVE_AUXTRACE_SUPPORT by assuming
it is always defined.
Signed-off-by: Ian Rogers <irogers@google.com>
Reviewed-by: James Clark <james.clark@linaro.org>
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Recent change on enabling --buildid-mmap by default brought an issue
with build-id handling. With build-ID in MMAP2 records, we don't need
to save the build-ID table in the header of a perf data file.
But the actual file contents still need to be cached in the debug
directory for annotation etc. Split the build-ID header processing and
caching and make sure perf record to save hit DSOs in the build-ID cache
by moving perf_session__cache_build_ids() to the end of the record__
finish_output().
Reviewed-by: Ian Rogers <irogers@google.com>
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
When using perf record with the `--overwrite` option, a segmentation fault
occurs if an event fails to open. For example:
perf record -e cycles-ct -F 1000 -a --overwrite
Error:
cycles-ct:H: PMU Hardware doesn't support sampling/overflow-interrupts. Try 'perf stat'
perf: Segmentation fault
#0 0x6466b6 in dump_stack debug.c:366
#1 0x646729 in sighandler_dump_stack debug.c:378
#2 0x453fd1 in sigsegv_handler builtin-record.c:722
#3 0x7f8454e65090 in __restore_rt libc-2.32.so[54090]
#4 0x6c5671 in __perf_event__synthesize_id_index synthetic-events.c:1862
#5 0x6c5ac0 in perf_event__synthesize_id_index synthetic-events.c:1943
#6 0x458090 in record__synthesize builtin-record.c:2075
#7 0x45a85a in __cmd_record builtin-record.c:2888
#8 0x45deb6 in cmd_record builtin-record.c:4374
#9 0x4e5e33 in run_builtin perf.c:349
#10 0x4e60bf in handle_internal_command perf.c:401
#11 0x4e6215 in run_argv perf.c:448
#12 0x4e653a in main perf.c:555
#13 0x7f8454e4fa72 in __libc_start_main libc-2.32.so[3ea72]
#14 0x43a3ee in _start ??:0
The --overwrite option implies --tail-synthesize, which collects non-sample
events reflecting the system status when recording finishes. However, when
evsel opening fails (e.g., unsupported event 'cycles-ct'), session->evlist
is not initialized and remains NULL. The code unconditionally calls
record__synthesize() in the error path, which iterates through the NULL
evlist pointer and causes a segfault.
To fix it, move the record__synthesize() call inside the error check block, so
it's only called when there was no error during recording, ensuring that evlist
is properly initialized.
Fixes: 4ea648aec0 ("perf record: Add --tail-synthesize option")
Signed-off-by: Shuai Xue <xueshuai@linux.alibaba.com>
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Rather than distributing the code doing similar things to
evlist__new_default, use the one implementation so that paranoia and
wildcard scanning can be optimized.
Signed-off-by: Ian Rogers <irogers@google.com>
Tested-by: James Clark <james.clark@linaro.org>
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Whilst for many tools it is an expected behavior that failure to open
a perf event is a failure, ARM decided to name PMU events the same as
legacy events and then failed to rename such events on a server uncore
SLC PMU. As perf's default behavior when no PMU is specified is to
open the event on all PMUs that advertise/"have" the event, this
yielded failures when trying to make the priority of legacy and
sysfs/json events uniform - something requested by RISC-V and ARM. A
legacy event user on ARM hardware may find their event opened on an
uncore PMU which for perf record will fail. Arnaldo suggested skipping
such events which this patch implements. Rather than have the skipping
conditional on running on ARM, the skipping is done on all
architectures as such a fundamental behavioral difference could lead
to problems with tools built/depending on perf.
An example of perf record failing to open events on x86 is:
```
$ perf record -e data_read,cycles,LLC-prefetch-read -a sleep 0.1
Error:
Failure to open event 'data_read' on PMU 'uncore_imc_free_running_0' which will be removed.
The sys_perf_event_open() syscall returned with 22 (Invalid argument) for event (data_read).
"dmesg | grep -i perf" may provide additional information.
Error:
Failure to open event 'data_read' on PMU 'uncore_imc_free_running_1' which will be removed.
The sys_perf_event_open() syscall returned with 22 (Invalid argument) for event (data_read).
"dmesg | grep -i perf" may provide additional information.
Error:
Failure to open event 'LLC-prefetch-read' on PMU 'cpu' which will be removed.
The LLC-prefetch-read event is not supported.
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 2.188 MB perf.data (87 samples) ]
$ perf report --stats
Aggregated stats:
TOTAL events: 17255
MMAP events: 284 ( 1.6%)
COMM events: 1961 (11.4%)
EXIT events: 1 ( 0.0%)
FORK events: 1960 (11.4%)
SAMPLE events: 87 ( 0.5%)
MMAP2 events: 12836 (74.4%)
KSYMBOL events: 83 ( 0.5%)
BPF_EVENT events: 36 ( 0.2%)
FINISHED_ROUND events: 2 ( 0.0%)
ID_INDEX events: 1 ( 0.0%)
THREAD_MAP events: 1 ( 0.0%)
CPU_MAP events: 1 ( 0.0%)
TIME_CONV events: 1 ( 0.0%)
FINISHED_INIT events: 1 ( 0.0%)
cycles stats:
SAMPLE events: 87
```
If all events fail to open then the perf record will fail:
```
$ perf record -e LLC-prefetch-read true
Error:
Failure to open event 'LLC-prefetch-read' on PMU 'cpu' which will be removed.
The LLC-prefetch-read event is not supported.
Error:
Failure to open any events for recording
```
As an evlist may have dummy events that open when all command line
events fail we ignore dummy events when detecting if at least some
events open. This still permits the dummy event on its own to be used
as a permission check:
```
$ perf record -e dummy true
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.046 MB perf.data ]
```
but allows failure when a dummy event is implicilty inserted or when
there are insufficient permissions to open it:
```
$ perf record -e LLC-prefetch-read -a true
Error:
Failure to open event 'LLC-prefetch-read' on PMU 'cpu' which will be removed.
The LLC-prefetch-read event is not supported.
Error:
Failure to open any events for recording
```
As the first parsed event in an evlist is marked as tracking, removing
this event can remove tracking from the evlist, removing mmap events
and breaking symbolization. To avoid this, if a tracking event is
removed then the next event has tracking added.
The issue with legacy events is that on RISC-V they want the driver to
not have mappings from legacy to non-legacy config encodings for each
vendor/model due to size, complexity and difficulty to update. It was
reported that on ARM Apple-M? CPUs the legacy mapping in the driver
was broken and the sysfs/json events should always take precedent,
however, it isn't clear this is still the case. It is the case that
without working around this issue a legacy event like cycles without a
PMU can encode differently than when specified with a PMU - the
non-PMU version favoring legacy encodings, the PMU one avoiding legacy
encodings. Legacy events are also case sensitive while sysfs/json
events are not.
The patch removes events and then adjusts the idx value for each
evsel. This is done so that the dense xyarrays used for file
descriptors, etc. don't contain broken entries.
On ARM it could be common following this change to see a lot of
warnings for the cycles event due to many ARM PMUs advertising the
cycles event (ARM inconsistently have events bus_cycles and then
cycles implying CPU cycles, they also sometimes have a cpu_cycles
event). As cycles is a popular event, avoid potentially spamming users
with error messages on ARM when there are multiple cycles events in
the evlist, the error is still shown when verbose is enabled.
Prior versions without adding the tracking data and not warning for
cycles on ARM was:
Suggested-by: Arnaldo Carvalho de Melo <acme@kernel.org>
Tested-by: James Clark <james.clark@linaro.org>
Tested-by: Leo Yan <leo.yan@arm.com>
Tested-by: Atish Patra <atishp@rivosinc.com>
Tested-by: Thomas Richter <tmricht@linux.ibm.com>
Signed-off-by: Ian Rogers <irogers@google.com>
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
For the sake of Intel topdown events commit 9eac5612da ("perf
stat: Don't skip failing group events") changed 'perf stat' error
handling making it so that more errors were fatal and didn't report
"<not supported>" events. The change outside of topdown events was
unintentional.
The notion of "fatal" error handling was introduced in commit
e0e6a6ca3a ("perf stat: Factor out open error handling") and
refined in commits like commit cb5ef60067 ("perf stat: Error out
unsupported group leader immediately") to be an approach for avoiding
later assertion failures in the code base.
This change fixes those issues and removes the notion of a fatal error
on an event. If all events fail to open then a fatal error occurs with
the previous fatal error message. This seems to best match the notion of
supported events and allowing some errors not to stop 'perf stat', while
allowing the truly fatal no event case to terminate the tool early.
The evsel->errored flag is only used in the stat code but always just
meaning !evsel->supported although there is a comment about it being
sticky. Force all evsels to be supported in evsel__init and then clear
this when evsel__open fails. When an event is tried the supported is
set to true again. This simplifies the notion of whether an evsel is
broken.
In the get_group_fd code, fail to get a group fd when the evsel isn't
supported. If the leader isn't supported then it is also expected that
there is no group_fd as the leader will have been skipped. Therefore
change the BUG_ON test to be on supported rather than skippable. This
corrects the assertion errors that were the reason for the previous
fatal error handling.
Fixes: 9eac5612da ("perf stat: Don't skip failing group events")
Reviewed-by: James Clark <james.clark@linaro.org>
Signed-off-by: Ian Rogers <irogers@google.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Chun-Tse Shao <ctshao@google.com>
Cc: Dapeng Mi <dapeng1.mi@linux.intel.com>
Cc: Howard Chu <howardchu95@gmail.com>
Cc: Ian Rogers <irogers@google.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kan Liang <kan.liang@linux.intel.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Falcon <thomas.falcon@intel.com>
Link: https://lore.kernel.org/r/20251002220727.1889799-2-irogers@google.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
auxtrace_mmap__read and auxtrace_mmap__read_snapshot end up calling
`evsel__env(NULL)` which returns the global perf_env variable for the
host. Their only call is in perf record. Rather than use the global
variable pass through the perf_env for `perf record`.
Signed-off-by: Ian Rogers <irogers@google.com>
Link: https://lore.kernel.org/r/20250724163302.596743-18-irogers@google.com
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
The original PERF_RECORD_COMPRESS is not 8-byte aligned, which can cause
asan runtime error:
# Build with asan
$ make -C tools/perf O=/tmp/perf DEBUG=1 EXTRA_CFLAGS="-O0 -g -fno-omit-frame-pointer -fsanitize=undefined"
# Test success with many asan runtime errors:
$ /tmp/perf/perf test "Zstd perf.data compression/decompression" -vv
83: Zstd perf.data compression/decompression:
...
util/session.c:1959:13: runtime error: member access within misaligned address 0x7f69e3f99653 for type 'union perf_event', which requires 13 byte alignment
0x7f69e3f99653: note: pointer points here
d0 3a 50 69 44 00 00 00 00 00 08 00 bb 07 00 00 00 00 00 00 44 00 00 00 00 00 00 00 ff 07 00 00
^
util/session.c:2163:22: runtime error: member access within misaligned address 0x7f69e3f99653 for type 'union perf_event', which requires 8 byte alignment
0x7f69e3f99653: note: pointer points here
d0 3a 50 69 44 00 00 00 00 00 08 00 bb 07 00 00 00 00 00 00 44 00 00 00 00 00 00 00 ff 07 00 00
^
...
Since there is no way to align compressed data in zstd compression, this
patch add a new event type `PERF_RECORD_COMPRESSED2`, which adds a field
`data_size` to specify the actual compressed data size.
The `header.size` contains the total record size, including the padding
at the end to make it 8-byte aligned.
Tested with `Zstd perf.data compression/decompression`
Signed-off-by: Chun-Tse Shao <ctshao@google.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Ben Gainey <ben.gainey@arm.com>
Cc: Christophe Leroy <christophe.leroy@csgroup.eu>
Cc: Ian Rogers <irogers@google.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: James Clark <james.clark@linaro.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kan Liang <kan.liang@linux.intel.com>
Cc: Leo Yan <leo.yan@arm.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Nick Terrell <terrelln@fb.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: https://lore.kernel.org/r/20250303183646.327510-1-ctshao@google.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Switch output to using a strbuf so the storage can be resized.
Add a maximum size argument to avoid too much output that may happen for
uncore events.
Rename as scnprintf is no longer used.
Committer testing:
With the patch applied:
root@number:~# perf probe -x ~/bin/perf evlist__format_evsels
Added new event:
probe_perf:evlist_format_evsels (on evlist__format_evsels in /home/acme/bin/perf)
You can now use it in all perf tools, such as:
perf record -e probe_perf:evlist_format_evsels -aR sleep 1
root@number:~# perf probe -l
probe_perf:evlist_format_evsels (on evlist__format_evsels@util/evlist.c in /home/acme/bin/perf)
root@number:~# perf trace -e probe_perf:*/max-stack=10/ perf record -e cycles,instructions,cache-misses /tmp/bla
Failed to collect 'cycles,instructions,cache-misses' for the '/tmp/bla' workload: Permission denied
0.000 perf/3893011 probe_perf:evlist_format_evsels(__probe_ip: 6183397)
evlist__format_evsels (/home/acme/bin/perf)
__cmd_record (/home/acme/bin/perf)
cmd_record (/home/acme/bin/perf)
run_builtin (/home/acme/bin/perf)
handle_internal_command (/home/acme/bin/perf)
run_argv (/home/acme/bin/perf)
main (/home/acme/bin/perf)
__libc_start_call_main (/usr/lib64/libc.so.6)
__libc_start_main@@GLIBC_2.34 (/usr/lib64/libc.so.6)
_start (/home/acme/bin/perf)
root@number:~#
Reviewed-by: Kan Liang <kan.liang@linux.intel.com>
Signed-off-by: Ian Rogers <irogers@google.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Dominique Martinet <asmadeus@codewreck.org>
Cc: Dr. David Alan Gilbert <linux@treblig.org>
Cc: Howard Chu <howardchu95@gmail.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: James Clark <james.clark@linaro.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Levi Yun <yeoreum.yun@arm.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Weilin Wang <weilin.wang@intel.com>
Link: https://lore.kernel.org/r/20250402201549.4090305-4-irogers@google.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Specify the threshold for dumping offcpu samples with --off-cpu-thresh,
the unit is milliseconds. Default value is 500ms.
Example:
perf record --off-cpu --off-cpu-thresh 824
The example above collects direct off-cpu samples where the off-cpu time
is longer than 824ms.
Committer testing:
After commenting out the end off-cpu dump to have just the ones that are
added right after the task is scheduled back, and using a threshould of
1000ms, we see some periods (the 5th column, just before "offcpu-time"
in the 'perf script' output) that are over 1000.000.000 nanoseconds:
root@number:~# perf record --off-cpu --off-cpu-thresh 10000
^C[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 3.902 MB perf.data (34335 samples) ]
root@number:~# perf script
<SNIP>
Isolated Web Co 59932 [028] 63839.594437: 1000049427 offcpu-time:
7fe63c7976c2 __syscall_cancel_arch_end+0x0 (/usr/lib64/libc.so.6)
7fe63c78c04c __futex_abstimed_wait_common+0x7c (/usr/lib64/libc.so.6)
7fe63c78e928 pthread_cond_timedwait@@GLIBC_2.3.2+0x178 (/usr/lib64/libc.so.6)
5599974a9fe7 mozilla::detail::ConditionVariableImpl::wait_for(mozilla::detail::MutexImpl&, mozilla::BaseTimeDuration<mozilla::TimeDurationValueCalculator> const&)+0xe7 (/usr/lib64/fir>
100000000 [unknown] ([unknown])
swapper 0 [025] 63839.594459: 195724 cycles:P: ffffffffac328270 read_tsc+0x0 ([kernel.kallsyms])
Isolated Web Co 59932 [010] 63839.594466: 1000055278 offcpu-time:
7fe63c7976c2 __syscall_cancel_arch_end+0x0 (/usr/lib64/libc.so.6)
7fe63c78ba24 __syscall_cancel+0x14 (/usr/lib64/libc.so.6)
7fe63c804c4e __poll+0x1e (/usr/lib64/libc.so.6)
7fe633b0d1b8 PollWrapper(_GPollFD*, unsigned int, int) [clone .lto_priv.0]+0xf8 (/usr/lib64/firefox/libxul.so)
10000002c [unknown] ([unknown])
swapper 0 [027] 63839.594475: 134433 cycles:P: ffffffffad4c45d9 irqentry_enter+0x19 ([kernel.kallsyms])
swapper 0 [028] 63839.594499: 215838 cycles:P: ffffffffac39199a switch_mm_irqs_off+0x10a ([kernel.kallsyms])
MediaPD~oder #1 1407676 [027] 63839.594514: 134433 cycles:P: 7f982ef5e69f dct_IV(int*, int, int*)+0x24f (/usr/lib64/libfdk-aac.so.2.0.0)
swapper 0 [024] 63839.594524: 267411 cycles:P: ffffffffad4c6ee6 poll_idle+0x56 ([kernel.kallsyms])
MediaSu~sor #75 1093827 [026] 63839.594555: 332652 cycles:P: 55be753ad030 moz_xmalloc+0x200 (/usr/lib64/firefox/firefox)
swapper 0 [027] 63839.594616: 160548 cycles:P: ffffffffad144840 menu_select+0x570 ([kernel.kallsyms])
Isolated Web Co 14019 [027] 63839.595120: 1000050178 offcpu-time:
7fc9537cc6c2 __syscall_cancel_arch_end+0x0 (/usr/lib64/libc.so.6)
7fc9537c104c __futex_abstimed_wait_common+0x7c (/usr/lib64/libc.so.6)
7fc9537c3928 pthread_cond_timedwait@@GLIBC_2.3.2+0x178 (/usr/lib64/libc.so.6)
7fc95372a3c8 pt_TimedWait+0xb8 (/usr/lib64/libnspr4.so)
7fc95372a8d8 PR_WaitCondVar+0x68 (/usr/lib64/libnspr4.so)
7fc94afb1f7c WatchdogMain(void*)+0xac (/usr/lib64/firefox/libxul.so)
7fc947498660 [unknown] ([unknown])
7fc9535fce88 [unknown] ([unknown])
7fc94b620e60 WatchdogManager::~WatchdogManager()+0x0 (/usr/lib64/firefox/libxul.so)
fff8548387f8b48 [unknown] ([unknown])
swapper 0 [003] 63839.595712: 212948 cycles:P: ffffffffacd5b865 acpi_os_read_port+0x55 ([kernel.kallsyms])
<SNIP>
Suggested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Suggested-by: Ian Rogers <irogers@google.com>
Suggested-by: Namhyung Kim <namhyung@kernel.org>
Reviewed-by: Ian Rogers <irogers@google.com>
Signed-off-by: Howard Chu <howardchu95@gmail.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Tested-by: Gautam Menghani <gautam@linux.ibm.com>
Tested-by: Ian Rogers <irogers@google.com>
Acked-by: Namhyung Kim <namhyung@kernel.org>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: James Clark <james.clark@linaro.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kan Liang <kan.liang@linux.intel.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: https://lore.kernel.org/r/20241108204137.2444151-2-howardchu95@gmail.com
Link: https://lore.kernel.org/r/20250501022809.449767-10-howardchu95@gmail.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Parse the off-cpu event using parse_event(), as bpf-output.
Call evlist__enable_evsel() on off-cpu event. This fixes the inability
to collect direct off-cpu samples on a workload, as reported by Arnaldo
Carvalho de Melo <acme@redhat.com>.
The reason being, workload sets enable_on_exec instead of calling
evlist__enable(), but off-cpu event does not attach to an executable and
execve won't be called, so the fds from perf_event_open() are not
enabled.
no-inherit should be set to 1, here's the reason:
We update the BPF perf_event map for direct off-cpu sample dumping (in
following patches), it executes as follows:
bpf_map_update_value()
bpf_fd_array_map_update_elem()
perf_event_fd_array_get_ptr()
perf_event_read_local()
In perf_event_read_local(), there is:
int perf_event_read_local(struct perf_event *event, u64 *value,
u64 *enabled, u64 *running)
{
...
/*
* It must not be an event with inherit set, we cannot read
* all child counters from atomic context.
*/
if (event->attr.inherit) {
ret = -EOPNOTSUPP;
goto out;
}
Which means no-inherit has to be true for updating the BPF perf_event
map.
Moreover, for bpf-output events, we primarily want a system-wide event
instead of a per-task event.
The reason is that in BPF's bpf_perf_event_output(), BPF uses the CPU
index to retrieve the perf_event file descriptor it outputs to.
Making a bpf-output event system-wide naturally satisfies this
requirement by mapping CPU appropriately.
Suggested-by: Namhyung Kim <namhyung@kernel.org>
Reviewed-by: Ian Rogers <irogers@google.com>
Signed-off-by: Howard Chu <howardchu95@gmail.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Tested-by: Gautam Menghani <gautam@linux.ibm.com>
Tested-by: Ian Rogers <irogers@google.com>
Acked-by: Namhyung Kim <namhyung@kernel.org>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: James Clark <james.clark@linaro.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kan Liang <kan.liang@linux.intel.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: https://lore.kernel.org/r/20241108204137.2444151-4-howardchu95@gmail.com
Link: https://lore.kernel.org/r/20250501022809.449767-3-howardchu95@gmail.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
The BPF sideband information is tracked using a separate thread and
evlist. But it's only useful for profiling kernel and we can skip it
when users profile their application only.
It seems it already fails to open the sideband event in that case.
Let's remove the noise in the verbose output anyway.
Reviewed-by: Ian Rogers <irogers@google.com>
Acked-by: Song Liu <song@kernel.org>
Link: https://lore.kernel.org/r/20250226203039.1099131-1-namhyung@kernel.org
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
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 <tavianator@tavianator.com> as about 2.5% when running `perf
script --itrace=i0`:
https://lore.kernel.org/lkml/d841b97b3ad2ca8bcab07e4293375fb7c32dfce7.1736618095.git.tavianator@tavianator.com/
Adrian Hunter <adrian.hunter@intel.com> 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 <irogers@google.com>
Link: https://lore.kernel.org/r/20250113194345.1537821-1-irogers@google.com
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
To allow BPF filters for unprivileged users it needs to pin the BPF
objects to BPF-fs first. Let's add a new option to pin and unpin the
objects easily. I'm not sure 'perf record' is a right place to do this
but I don't have a better idea right now.
$ sudo perf record --setup-filter pin
The above command would pin BPF program and maps for the filter when the
system has BPF-fs (usually at /sys/fs/bpf/). To unpin the objects,
users can run the following command (as root).
$ sudo perf record --setup-filter unpin
Committer testing:
root@number:~# perf record --setup-filter pin
root@number:~# ls -la /sys/fs/bpf/perf_filter/
total 0
drwxr-xr-x. 2 root root 0 Jul 31 10:43 .
drwxr-xr-t. 3 root root 0 Jul 31 10:43 ..
-rw-rw-rw-. 1 root root 0 Jul 31 10:43 dropped
-rw-rw-rw-. 1 root root 0 Jul 31 10:43 filters
-rwxrwxrwx. 1 root root 0 Jul 31 10:43 perf_sample_filter
-rw-rw-rw-. 1 root root 0 Jul 31 10:43 pid_hash
-rw-------. 1 root root 0 Jul 31 10:43 sample_f_rodata
root@number:~# ls -la /sys/fs/bpf/perf_filter/perf_sample_filter
-rwxrwxrwx. 1 root root 0 Jul 31 10:43 /sys/fs/bpf/perf_filter/perf_sample_filter
root@number:~#
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Ian Rogers <irogers@google.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: KP Singh <kpsingh@kernel.org>
Cc: Kan Liang <kan.liang@linux.intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Song Liu <song@kernel.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: bpf@vger.kernel.org
Link: https://lore.kernel.org/r/20240703223035.2024586-8-namhyung@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Modified the object of 'memset' from '&lost.lost' to '&lost' in
record__read_lost_samples. This allows 'memset' to access memory properly
without causing out-of-bounds problems.
The problems got from builtin-record.c are:
In file included from /usr/include/string.h:495,
from util/parse-events.h:13,
from builtin-record.c:14:
In function 'memset',
inlined from 'record__read_lost_samples' at
builtin-record.c:1958:6,
inlined from '__cmd_record.constprop' at builtin-record.c:2817:2:
/usr/include/x86_64-linux-gnu/bits/string_fortified.h:71:10: error:
'__builtin_memset' offset [17, 64] from the object at 'lost' is out
of the bounds of referenced subobject 'lost' with type
'struct perf_record_lost_samples' at offset 0 [-Werror=array-bounds]
71|return __builtin___memset_chk (__dest,__ch,__len,__bos0 (__dest));
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
The error arised when performing a memset operation on the 'lost' variable,
the bytes of 'sizeof(lost)' exceeds that of '&lost.lost', which are 64
and 16.
Fixes: 6c1785cd75 ("perf record: Ensure space for lost samples")
Signed-off-by: Haoze Xie <royenheart@gmail.com>
Signed-off-by: Yuan Tan <tanyuan@tinylab.org>
Link: https://lore.kernel.org/r/11e12f171b846577cac698cd3999db3d7f6c4d03.1720372317.git.royenheart@gmail.com
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
The session has a header in it which contains a perf env with
bpf_progs. The bpf_progs are accessed by the sideband thread and so
the sideband thread must be stopped before the session is deleted, to
avoid a use after free. This error was detected by AddressSanitizer
in the following:
==2054673==ERROR: AddressSanitizer: heap-use-after-free on address 0x61d000161e00 at pc 0x55769289de54 bp 0x7f9df36d4ab0 sp 0x7f9df36d4aa8
READ of size 8 at 0x61d000161e00 thread T1
#0 0x55769289de53 in __perf_env__insert_bpf_prog_info util/env.c:42
#1 0x55769289dbb1 in perf_env__insert_bpf_prog_info util/env.c:29
#2 0x557692bbae29 in perf_env__add_bpf_info util/bpf-event.c:483
#3 0x557692bbb01a in bpf_event__sb_cb util/bpf-event.c:512
#4 0x5576928b75f4 in perf_evlist__poll_thread util/sideband_evlist.c:68
#5 0x7f9df96a63eb in start_thread nptl/pthread_create.c:444
#6 0x7f9df9726a4b in clone3 ../sysdeps/unix/sysv/linux/x86_64/clone3.S:81
0x61d000161e00 is located 384 bytes inside of 2136-byte region [0x61d000161c80,0x61d0001624d8)
freed by thread T0 here:
#0 0x7f9dfa6d7288 in __interceptor_free libsanitizer/asan/asan_malloc_linux.cpp:52
#1 0x557692978d50 in perf_session__delete util/session.c:319
#2 0x557692673959 in __cmd_record tools/perf/builtin-record.c:2884
#3 0x55769267a9f0 in cmd_record tools/perf/builtin-record.c:4259
#4 0x55769286710c in run_builtin tools/perf/perf.c:349
#5 0x557692867678 in handle_internal_command tools/perf/perf.c:402
#6 0x557692867a40 in run_argv tools/perf/perf.c:446
#7 0x557692867fae in main tools/perf/perf.c:562
#8 0x7f9df96456c9 in __libc_start_call_main ../sysdeps/nptl/libc_start_call_main.h:58
Fixes: 657ee55319 ("perf evlist: Introduce side band thread")
Signed-off-by: Ian Rogers <irogers@google.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Athira Rajeev <atrajeev@linux.vnet.ibm.com>
Cc: Christian Brauner <brauner@kernel.org>
Cc: Disha Goel <disgoel@linux.ibm.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: James Clark <james.clark@arm.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kajol Jain <kjain@linux.ibm.com>
Cc: Kan Liang <kan.liang@linux.intel.com>
Cc: K Prateek Nayak <kprateek.nayak@amd.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Song Liu <songliubraving@fb.com>
Cc: Tim Chen <tim.c.chen@linux.intel.com>
Cc: Yicong Yang <yangyicong@hisilicon.com>
Link: https://lore.kernel.org/r/20240301074639.2260708-1-irogers@google.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Currently pipe mode doesn't set the file size and it results in a
misleading message of 0 data size at the end. Although it might miss
some accounting for pipe header or more, just displaying the data size
would reduce the possible confusion.
Before:
$ perf record -o- perf test -w noploop | perf report -i- -q --percent-limit=1
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.000 MB - ] <====== (here)
99.58% perf perf [.] noploop
After:
$ perf record -o- perf test -w noploop | perf report -i- -q --percent-limit=1
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.229 MB - ]
99.46% perf perf [.] noploop
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Reviewed-by: Ian Rogers <irogers@google.com>
Link: https://lore.kernel.org/r/20240112231340.779469-1-namhyung@kernel.org
