Multiple threads may be creating and destroying BFD objects in
situations like `perf top`.
Without appropriate initialization crashes may occur during libbfd's
cache management.
BFD's locks require recursive mutexes, add support for these.
Committer testing:
This happens only when building with 'make BUILD_NONDISTRO=1' and having
the binutils-devel package (or equivalent) installed, i.e. linking with
binutils devel files, an opt-in perf build.
Before:
root@x1:~# perf top
perf: Segmentation fault
-------- backtrace --------
<SNIP multiple failed attempts at printing a backtrace>
root@x1:~#
After this patch it works as before.
Closes: https://lore.kernel.org/lkml/aQt66zhfxSA80xwt@gentoo.org/
Fixes: 95931d9a59 ("perf libbfd: Move libbfd functionality to its own file")
Reported-by: Guilherme Amadio <amadio@gentoo.org>
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: Ingo Molnar <mingo@redhat.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
With commit f0d0f978f3 ("perf header: Don't write empty BPF/BTF
info"), the write_bpf_( prog_info() | btf() ) functions exit without
writing anything if env->bpf_prog.(infos| btfs)_cnt is zero.
process_bpf_( prog_info() | btf() ), however, still expect a "count"
value to exist in the data file. If btf information is empty, for
example, process_bpf_btf will read garbage or some other data as the
number of btf nodes in the data file. As a result, the data file will
not be processed correctly.
Instead, write the count to the data file and exit if it is zero.
Fixes: f0d0f978f3 ("perf header: Don't write empty BPF/BTF info")
Reviewed-by: Ian Rogers <irogers@google.com>
Signed-off-by: Thomas Falcon <thomas.falcon@intel.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: Jiri Olsa <jolsa@kernel.org>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
I started seeing this in recent Fedora 42 kernels:
root@x1:~# uname -a
Linux x1 6.17.4-200.fc42.x86_64 #1 SMP PREEMPT_DYNAMIC Sun Oct 19 18:47:49 UTC 2025 x86_64 GNU/Linux
root@x1:~#
root@x1:~# perf test 1
1: vmlinux symtab matches kallsyms : FAILED!
root@x1:~#
Related to:
root@x1:~# grep ' 1 ' /proc/kallsyms
ffffffffb098bc00 1 __pfx__RNCINvNtNtNtCsfwaGRd4cjqE_4core4iter8adapters3map12map_try_foldjNtCskFudTml27HW_12drm_panic_qr7VersionuINtNtNtBa_3ops12control_flow11ControlFlowB10_ENcB10_0NCINvNvNtNtNtB8_6traits8iterator8Iterator4find5checkB10_NCNvMB12_B10_13from_segments0E0E0B12_
ffffffffb098bc10 1 _RNCINvNtNtNtCsfwaGRd4cjqE_4core4iter8adapters3map12map_try_foldjNtCskFudTml27HW_12drm_panic_qr7VersionuINtNtNtBa_3ops12control_flow11ControlFlowB10_ENcB10_0NCINvNvNtNtNtB8_6traits8iterator8Iterator4find5checkB10_NCNvMB12_B10_13from_segments0E0E0B12_
root@x1:~#
That is found in:
root@x1:~# pahole --running_kernel_vmlinux
/usr/lib/debug/lib/modules/6.17.4-200.fc42.x86_64/vmlinux
root@x1:~#
root@x1:~# readelf -sW /usr/lib/debug/lib/modules/6.17.4-200.fc42.x86_64/vmlinux | grep __pfx__RNCINvNtNtNtCsfwaGRd4cjqE_4core4iter8adapters3map12map_try_foldjNtCskFudTml27HW_12drm_panic_qr7VersionuINtNtNtBa_3ops12control_flow11ControlFlowB10_ENcB10_0NCINvNvNtNtNtB8_6traits8iterator8Iterator4find5checkB10_NCNvMB12_B10_13from_segments0E0E0B12_
150649: ffffffff81f8bc00 16 FUNC LOCAL DEFAULT 1 __pfx__RNCINvNtNtNtCsfwaGRd4cjqE_4core4iter8adapters3map12map_try_foldjNtCskFudTml27HW_12drm_panic_qr7VersionuINtNtNtBa_3ops12control_flow11ControlFlowB10_ENcB10_0NCINvNvNtNtNtB8_6traits8iterator8Iterator4find5checkB10_NCNvMB12_B10_13from_segments0E0E0B12_
root@x1:~#
But was being filtered out when reading /proc/kallsyms, as the '1'
symbol type was not being handled, do it, there are just two of them at
this point.
Cc: Alex Gaynor <alex.gaynor@gmail.com>
Cc: Alice Ryhl <aliceryhl@google.com>
Cc: Andreas Hindborg <a.hindborg@kernel.org>
Cc: Benno Lossin <lossin@kernel.org>
Cc: Björn Roy Baron <bjorn3_gh@protonmail.com>
Cc: Boqun Feng <boqun.feng@gmail.com>
Cc: Danilo Krummrich <dakr@kernel.org>
Cc: Gary Guo <gary@garyguo.net>
Cc: Miguel Ojeda <ojeda@kernel.org>
Cc: Trevor Gross <tmgross@umich.edu>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Pull perf tools updates from Arnaldo Carvalho de Melo:
- Extended 'perf annotate' with DWARF type information
(--code-with-type) integration in the TUI, including a 'T'
hotkey to toggle it
- Enhanced 'perf bench mem' with new mmap() workloads and control
over page/chunk sizes
- Fix 'perf stat' error handling to correctly display unsupported
events
- Improved support for Clang cross-compilation
- Refactored LLVM and Capstone disasm for modularity
- Introduced the :X modifier to exclude an event from automatic
regrouping
- Adjusted KVM sampling defaults to use the "cycles" event to prevent
failures
- Added comprehensive support for decoding PowerPC Dispatch Trace Log
(DTL)
- Updated Arm SPE tracing logic for better analysis of memory and snoop
details
- Synchronized Intel PMU events and metrics with TMA 5.1 across
multiple processor generations
- Converted dependencies like libperl and libtracefs to be opt-in
- Handle more Rust symbols in kallsyms ('N', debugging)
- Improve the python binding to allow for python based tools to use
more of the libraries, add a 'ilist' utility to test those new
bindings
- Various 'perf test' fixes
- Kan Liang no longer a perf tools reviewer
* tag 'perf-tools-for-v6.18-1-2025-10-08' of git://git.kernel.org/pub/scm/linux/kernel/git/perf/perf-tools: (192 commits)
perf tools: Fix arm64 libjvmti build by generating unistd_64.h
perf tests: Don't retest sections in "Object code reading"
perf docs: Document building with Clang
perf build: Support build with clang
perf test coresight: Dismiss clang warning for unroll loop thread
perf test coresight: Dismiss clang warning for thread loop
perf test coresight: Dismiss clang warning for memcpy thread
perf build: Disable thread safety analysis for perl header
perf build: Correct CROSS_ARCH for clang
perf python: split Clang options when invoking Popen
tools build: Align warning options with perf
perf disasm: Remove unused evsel from 'struct annotate_args'
perf srcline: Fallback between addr2line implementations
perf disasm: Make ins__scnprintf() and ins__is_nop() static
perf dso: Clean up read_symbol() error handling
perf dso: Support BPF programs in dso__read_symbol()
perf dso: Move read_symbol() from llvm/capstone to dso
perf llvm: Reduce LLVM initialization
perf check: Add libLLVM feature
perf parse-events: Fix parsing of >30kb event strings
...
Metrics may generate many particularly uncore event references. The
resulting event string may then be >32kb. The parse events lex is
using "%option reject" which stores backtracking state in a buffer
sized at roughtly 30kb. If the event string is larger than this then a
buffer overflow and typically a crash happens.
The need for "%option reject" was for BPF events which were removed in
commit 3d6dfae889 ("perf parse-events: Remove BPF event
support"). As "%option reject" is both a memory and performance cost
let's remove it and fix the parsing case for event strings being over
~30kb.
Whilst cleaning up "%option reject" make the header files accurately
reflect functions used in the code and tidy up not requiring yywrap.
Measuring on the "PMU JSON event tests" a modest reduction of 0.41%
user time and 0.27% max resident size was observed. More importantly
this change fixes parsing large metrics and event strings.
Signed-off-by: Ian Rogers <irogers@google.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kan Liang <kan.liang@linux.intel.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
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>
After KVM supports PEBS for guest on Intel platforms
(https://lore.kernel.org/all/20220411101946.20262-1-likexu@tencent.com/),
host loses the capability to sample guest with PEBS since all PEBS related
MSRs are switched to guest value after vm-entry, like IA32_DS_AREA MSR is
switched to guest GVA at vm-entry. This would lead to "perf kvm record"
fails to sample guest on Intel platforms since "cycles:P" event is used to
sample guest by default as below case shows.
sudo perf kvm record -a
^C[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.787 MB perf.data.guest ]
So to ensure guest record can be sampled successfully, use "cycles"
instead of "cycles:P" to sample guest record by default on Intel
platforms. With this patch, the guest record can be sampled
successfully.
sudo perf kvm record -a
^C[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.783 MB perf.data.guest (23 samples) ]
Fixes: cf8e55fe50 ("KVM: x86/pmu: Expose CPUIDs feature bits PDCM, DS, DTES64")
Reported-by: Kevin Tian <kevin.tian@intel.com>
Signed-off-by: Dapeng Mi <dapeng1.mi@linux.intel.com>
Acked-by: Namhyung Kim <namhyung@kernel.org>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Ian Rogers <irogers@google.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Kan Liang <kan.liang@linux.intel.com>
Cc: Like Xu <likexu@tencent.com>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
When exploring building bpf_skel with GCC's BPF support, there was a
build failure because of bpf_core_field_exists vs the mem_hops bitfield:
```
In file included from util/bpf_skel/sample_filter.bpf.c:6:
util/bpf_skel/sample_filter.bpf.c: In function 'perf_get_sample':
tools/perf/libbpf/include/bpf/bpf_core_read.h:169:42: error: cannot take address of bit-field 'mem_hops'
169 | #define ___bpf_field_ref1(field) (&(field))
| ^
tools/perf/libbpf/include/bpf/bpf_helpers.h:222:29: note: in expansion of macro '___bpf_field_ref1'
222 | #define ___bpf_concat(a, b) a ## b
| ^
tools/perf/libbpf/include/bpf/bpf_helpers.h:225:29: note: in expansion of macro '___bpf_concat'
225 | #define ___bpf_apply(fn, n) ___bpf_concat(fn, n)
| ^~~~~~~~~~~~~
tools/perf/libbpf/include/bpf/bpf_core_read.h:173:9: note: in expansion of macro '___bpf_apply'
173 | ___bpf_apply(___bpf_field_ref, ___bpf_narg(args))(args)
| ^~~~~~~~~~~~
tools/perf/libbpf/include/bpf/bpf_core_read.h:188:39: note: in expansion of macro '___bpf_field_ref'
188 | __builtin_preserve_field_info(___bpf_field_ref(field), BPF_FIELD_EXISTS)
| ^~~~~~~~~~~~~~~~
util/bpf_skel/sample_filter.bpf.c:167:29: note: in expansion of macro 'bpf_core_field_exists'
167 | if (bpf_core_field_exists(data->mem_hops))
| ^~~~~~~~~~~~~~~~~~~~~
cc1: error: argument is not a field access
```
___bpf_field_ref1 was adapted for GCC in 12bbcf8e84
but the trick added for compatibility in 3a8b8fc317
isn't compatible with that as an address is used as an argument.
Workaround this by calling __builtin_preserve_field_info directly as the
bpf_core_field_exists macro does, but without the ___bpf_field_ref use.
Co-developed-by: Andrew Pinski <andrew.pinski@oss.qualcomm.com>
Signed-off-by: Andrew Pinski <andrew.pinski@oss.qualcomm.com>
Signed-off-by: Sam James <sam@gentoo.org>
Acked-by: Yonghong Song <yonghong.song@linux.dev>
Tested-by: Namhyung Kim <namhyung@kernel.org>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.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: Mark Rutland <mark.rutland@arm.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: https://gcc.gnu.org/PR121420
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Don't open evsels on all CPUs, open them just on the CPUs they
support. This avoids opening say an e-core event on a p-core and
getting a failure - achieve this by getting rid of the "all_cpu_map".
In install_pe functions don't use the cpu_map_idx as a CPU number,
translate the cpu_map_idx, which is a dense index into the cpu_map
skipping holes at the beginning, to a proper CPU number.
Before:
```
$ perf stat --bpf-counters -a -e cycles,instructions -- sleep 1
Performance counter stats for 'system wide':
<not supported> cpu_atom/cycles/
566,270,672 cpu_core/cycles/
<not supported> cpu_atom/instructions/
572,792,836 cpu_core/instructions/ # 1.01 insn per cycle
1.001595384 seconds time elapsed
```
After:
```
$ perf stat --bpf-counters -a -e cycles,instructions -- sleep 1
Performance counter stats for 'system wide':
443,299,201 cpu_atom/cycles/
1,233,919,737 cpu_core/cycles/
213,634,112 cpu_atom/instructions/ # 0.48 insn per cycle
2,758,965,527 cpu_core/instructions/ # 2.24 insn per cycle
1.001699485 seconds time elapsed
```
Fixes: 7fac83aaf2 ("perf stat: Introduce 'bperf' to share hardware PMCs with BPF")
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: bpf@vger.kernel.org
Cc: Gabriele Monaco <gmonaco@redhat.com>
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: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Song Liu <songliubraving@fb.com>
Cc: Tengda Wu <wutengda@huaweicloud.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Create samples from DTL entries for displaying in 'perf report'
and 'perf script'.
When the different PERF_RECORD_XX records are processed from perf
session, powerpc_vpadtl_process_event() will be invoked.
For each of the PERF_RECORD_XX record, compare the timestamp of perf
record with timestamp of top element in the auxtrace heap.
Process the auxtrace queue if the timestamp of element from heap is
lower than timestamp from entry in perf record.
Sometimes it could happen that one buffer is only partially processed.
if the timestamp of occurrence of another event is more than currently
processed element in the queue, it will move on to next perf record.
So keep track of position of buffer to continue processing next time.
Update the timestamp of the auxtrace heap with the timestamp of last
processed entry from the auxtrace buffer.
Generate perf sample for each entry in the dispatch trace log.
Fill in the sample details:
- sample ip is picked from srr0 field of dtl_entry
- sample cpu is picked from processor_id of dtl_entry
- sample id is from sample_id of powerpc_vpadtl
- cpumode is set to PERF_RECORD_MISC_KERNEL
- Additionally save the details in raw_data of sample.
This is to print the relevant fields in perf_sample__fprintf_synth()
when called from builtin-script
The sample is processed by calling perf_session__deliver_synth_event()
so that it gets included in perf report.
Sample Output:
./perf record -a -e sched:*,vpa_dtl/dtl_all/ -c 1000000000 sleep 1
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.300 MB perf.data ]
./perf report
# Samples: 321 of event 'vpa-dtl'
# Event count (approx.): 321
#
# Children Self Command Shared Object Symbol
# ........ ........ ....... ................. ..............................
#
100.00% 100.00% swapper [kernel.kallsyms] [k] plpar_hcall_norets_notrace
Reviewed-by: Adrian Hunter <adrian.hunter@intel.com>
Signed-off-by: Athira Rajeev <atrajeev@linux.ibm.com>
Tested-by: Tejas Manhas <tejas05@linux.ibm.com>
Tested-by: Venkat Rao Bagalkote <venkat88@linux.ibm.com>
Cc: Aboorva Devarajan <aboorvad@linux.ibm.com>
Cc: Aditya Bodkhe <Aditya.Bodkhe1@ibm.com>
Cc: Hari Bathini <hbathini@linux.vnet.ibm.com>
Cc: Ian Rogers <irogers@google.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Madhavan Srinivasan <maddy@linux.ibm.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Shrikanth Hegde <sshegde@linux.ibm.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
When the Dispatch Trace Log data is collected along with other events
like sched tracepoint events, it needs to be correlated and present
interleaved along with these events.
Perf events can be collected parallely across the CPUs. Hence it needs
to be ensured events/dtl entries are processed in timestamp order.
An auxtrace_queue is created for each CPU.
Data within each queue is in increasing order of timestamp. Each
auxtrace queue has a array/list of auxtrace buffers.
When processing the auxtrace buffer, the data is mmapp'ed.
All auxtrace queues is maintained in auxtrace heap.
Each queue has a queue number and a timestamp.
The queues are sorted/added to head based on the time stamp.
So always the lowest timestamp (entries to be processed first) is on top
of the heap.
The auxtrace queue needs to be allocated and heap needs to be populated
in the sorted order of timestamp.
The queue needs to be filled with data only once via
powerpc_vpadtl__update_queues() function.
powerpc_vpadtl__setup_queues() iterates through all the entries to
allocate and setup the auxtrace queue.
To add to auxtrace heap, it is required to fetch the timebase of first
entry for each of the queue.
The first entry in the queue for VPA DTL PMU has the boot timebase,
frequency details which are needed to get timestamp which is required to
correlate with other events.
The very next entry is the actual trace data that provides timestamp for
occurrence of DTL event.
Formula used to get the timestamp from dtl entry is:
((timbase from DTL entry - boot time) / frequency) * 1000000000
powerpc_vpadtl_decode() adds the boot time and frequency as part of
powerpc_vpadtl_queue structure so that it can be reused.
Each of the dtl_entry is of 48 bytes size. Sometimes it could happen
that one buffer is only partially processed (if the timestamp of
occurrence of another event is more than currently processed element in
queue, it will move on to next event).
In order to keep track of position of buffer, additional fields is added
to powerpc_vpadtl_queue structure.
Reviewed-by: Adrian Hunter <adrian.hunter@intel.com>
Signed-off-by: Athira Rajeev <atrajeev@linux.ibm.com>
Tested-by: Tejas Manhas <tejas05@linux.ibm.com>
Tested-by: Venkat Rao Bagalkote <venkat88@linux.ibm.com>
Cc: Aboorva Devarajan <aboorvad@linux.ibm.com>
Cc: Aditya Bodkhe <Aditya.Bodkhe1@ibm.com>
Cc: Hari Bathini <hbathini@linux.vnet.ibm.com>
Cc: Ian Rogers <irogers@google.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Madhavan Srinivasan <maddy@linux.ibm.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Shrikanth Hegde <sshegde@linux.ibm.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Dispatch Trace Log details are captured as-is in PERF_RECORD_AUXTRACE
records.
To present dtl entries as samples, create an event with name as
"vpa-dtl" and type PERF_TYPE_SYNTH.
Add perf_synth_id, "PERF_SYNTH_POWERPC_VPA_DTL" as config value for the
event.
Create a sample id to be a fixed offset from evsel id.
To present the relevant fields from the "struct dtl_entry", prepare the
entries as events of type PERF_TYPE_SYNTH.
By defining as PERF_TYPE_SYNTH type, samples can be printed as part of
perf_sample__fprintf_synth in builtin-script.c
From powerpc_vpadtl_process_auxtrace_info(), invoke
auxtrace_queues__process_index() function which will queue the auxtrace
buffers by invoke auxtrace_queues__add_event().
Reviewed-by: Adrian Hunter <adrian.hunter@intel.com>
Signed-off-by: Athira Rajeev <atrajeev@linux.ibm.com>
Tested-by: Tejas Manhas <tejas05@linux.ibm.com>
Tested-by: Venkat Rao Bagalkote <venkat88@linux.ibm.com>
Cc: Aboorva Devarajan <aboorvad@linux.ibm.com>
Cc: Aditya Bodkhe <Aditya.Bodkhe1@ibm.com>
Cc: Hari Bathini <hbathini@linux.vnet.ibm.com>
Cc: Ian Rogers <irogers@google.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Madhavan Srinivasan <maddy@linux.ibm.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Shrikanth Hegde <sshegde@linux.ibm.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Add VPA DTL PMU auxtrace process function for "perf report -D".
The auxtrace event processing functions are defined in file
"util/powerpc-vpadtl.c".
Data structures used includes "struct powerpc_vpadtl_queue", "struct
powerpc_vpadtl" to store the auxtrace buffers in queue. Different
PERF_RECORD_XXX are generated during recording.
PERF_RECORD_AUXTRACE_INFO is processed first since it is of type
perf_user_event_type and perf session event delivers
perf_session__process_user_event() first.
Define function powerpc_vpadtl_process_auxtrace_info() to handle the
processing of PERF_RECORD_AUXTRACE_INFO records.
In this function, initialize the aux buffer queues using
auxtrace_queues__init().
Setup the required infrastructure for aux data processing.
The data is collected per CPU and auxtrace_queue is created for each
CPU.
Define powerpc_vpadtl_process_event() function to process
PERF_RECORD_AUXTRACE records.
In this, add the event to queue using auxtrace_queues__add_event() and
process the buffer in powerpc_vpadtl_dump_event().
The first entry in the buffer with timebase as zero has boot timebase
and frequency.
Remaining data is of format for "struct powerpc_vpadtl_entry".
Define the translation for dispatch_reasons and preempt_reasons, report
this when dump trace is invoked via powerpc_vpadtl_dump()
Sample output:
./perf record -a -e sched:*,vpa_dtl/dtl_all/ -c 1000000000 sleep 1
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.300 MB perf.data ]
./perf report -D
0 0 0x39b10 [0x30]: PERF_RECORD_AUXTRACE size: 0x690 offset: 0 ref: 0 idx: 0 tid: -1 cpu: 0
.
. ... VPA DTL PMU data: size 1680 bytes, entries is 35
. 00000000: boot_tb: 21349649546353231, tb_freq: 512000000
. 00000030: dispatch_reason:decrementer interrupt, preempt_reason:H_CEDE, enqueue_to_dispatch_time:7064, ready_to_enqueue_time:187, waiting_to_ready_time:6611773
. 00000060: dispatch_reason:priv doorbell, preempt_reason:H_CEDE, enqueue_to_dispatch_time:146, ready_to_enqueue_time:0, waiting_to_ready_time:15359437
. 00000090: dispatch_reason:decrementer interrupt, preempt_reason:H_CEDE, enqueue_to_dispatch_time:4868, ready_to_enqueue_time:232, waiting_to_ready_time:5100709
. 000000c0: dispatch_reason:priv doorbell, preempt_reason:H_CEDE, enqueue_to_dispatch_time:179, ready_to_enqueue_time:0, waiting_to_ready_time:30714243
. 000000f0: dispatch_reason:priv doorbell, preempt_reason:H_CEDE, enqueue_to_dispatch_time:197, ready_to_enqueue_time:0, waiting_to_ready_time:15350648
. 00000120: dispatch_reason:priv doorbell, preempt_reason:H_CEDE, enqueue_to_dispatch_time:213, ready_to_enqueue_time:0, waiting_to_ready_time:15353446
. 00000150: dispatch_reason:priv doorbell, preempt_reason:H_CEDE, enqueue_to_dispatch_time:212, ready_to_enqueue_time:0, waiting_to_ready_time:15355126
. 00000180: dispatch_reason:decrementer interrupt, preempt_reason:H_CEDE, enqueue_to_dispatch_time:6368, ready_to_enqueue_time:164, waiting_to_ready_time:5104665
Reviewed-by: Adrian Hunter <adrian.hunter@intel.com>
Signed-off-by: Athira Rajeev <atrajeev@linux.ibm.com>
Tested-by: Tejas Manhas <tejas05@linux.ibm.com>
Tested-by: Venkat Rao Bagalkote <venkat88@linux.ibm.com>
Cc: Aboorva Devarajan <aboorvad@linux.ibm.com>
Cc: Aditya Bodkhe <Aditya.Bodkhe1@ibm.com>
Cc: Hari Bathini <hbathini@linux.vnet.ibm.com>
Cc: Ian Rogers <irogers@google.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Madhavan Srinivasan <maddy@linux.ibm.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Shrikanth Hegde <sshegde@linux.ibm.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
The powerpc PMU collecting Dispatch Trace Log (DTL) entries makes use of
AUX support in perf infrastructure.
The PMU driver has the functionality to collect trace entries in the aux
buffer.
On the tools side, this data is made available as PERF_RECORD_AUXTRACE
records.
This record is generated by "perf record" command.
To enable the creation of PERF_RECORD_AUXTRACE, add functions to
initialize auxtrace records ie "auxtrace_record__init()".
Fill in fields for other callbacks like info_priv_size, info_fill, free,
recording options etc.
Define auxtrace_type as PERF_AUXTRACE_VPA_DTL. Add header file to
define vpa dtl pmu specific details.
Reviewed-by: Adrian Hunter <adrian.hunter@intel.com>
Signed-off-by: Athira Rajeev <atrajeev@linux.ibm.com>
Tested-by: Tejas Manhas <tejas05@linux.ibm.com>
Tested-by: Venkat Rao Bagalkote <venkat88@linux.ibm.com>
Cc: Aboorva Devarajan <aboorvad@linux.ibm.com>
Cc: Aditya Bodkhe <Aditya.Bodkhe1@ibm.com>
Cc: Hari Bathini <hbathini@linux.vnet.ibm.com>
Cc: Ian Rogers <irogers@google.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Madhavan Srinivasan <maddy@linux.ibm.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Shrikanth Hegde <sshegde@linux.ibm.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Pull x86 instruction decoder update from Borislav Petkov:
- Add instruction decoding support for the XOP-prefixed instruction set
present on the AMD Bulldozer uarch
[ These instructions don't normally happen, but a X86_NATIVE_CPU build
on a bulldozer host can make the compiler then use these unusual
instruction encodings ]
* tag 'x86_misc_for_v6.18_rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
x86/insn: Add XOP prefix instructions decoder support
When building perf with LIBBPF_DYNAMIC=1 on a fedora system with
libbpf-devel 1.5 I it was breaking with:
util/bpf-event.c: In function ‘format_btf_variable’:
util/bpf-event.c:291:18: error: ‘const struct btf_dump_type_data_opts’ has no member named ‘emit_strings’
291 | .emit_strings = 1,
| ^~~~~~~~~~~~
util/bpf-event.c:291:33: error: initialized field overwritten [-Werror=override-init]
291 | .emit_strings = 1,
| ^
util/bpf-event.c:291:33: note: (near initialization for ‘opts.skip_names’)
Check the version before using that feature.
Reviewed-by: Ian Rogers <irogers@google.com>
Acked-by: Namhyung Kim <namhyung@kernel.org>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Andrii Nakryiko <andrii@kernel.org>
Cc: James Clark <james.clark@linaro.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kan Liang <kan.liang@linux.intel.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Pull hardening updates from Kees Cook:
"One notable addition is the creation of the 'transitional' keyword for
kconfig so CONFIG renaming can go more smoothly.
This has been a long-standing deficiency, and with the renaming of
CONFIG_CFI_CLANG to CONFIG_CFI (since GCC will soon have KCFI
support), this came up again.
The breadth of the diffstat is mainly this renaming.
- Clean up usage of TRAILING_OVERLAP() (Gustavo A. R. Silva)
- lkdtm: fortify: Fix potential NULL dereference on kmalloc failure
(Junjie Cao)
- Add str_assert_deassert() helper (Lad Prabhakar)
- gcc-plugins: Remove TODO_verify_il for GCC >= 16
- kconfig: Fix BrokenPipeError warnings in selftests
- kconfig: Add transitional symbol attribute for migration support
- kcfi: Rename CONFIG_CFI_CLANG to CONFIG_CFI"
* tag 'hardening-v6.18-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/kees/linux:
lib/string_choices: Add str_assert_deassert() helper
kcfi: Rename CONFIG_CFI_CLANG to CONFIG_CFI
kconfig: Add transitional symbol attribute for migration support
kconfig: Fix BrokenPipeError warnings in selftests
gcc-plugins: Remove TODO_verify_il for GCC >= 16
stddef: Introduce __TRAILING_OVERLAP()
stddef: Remove token-pasting in TRAILING_OVERLAP()
lkdtm: fortify: Fix potential NULL dereference on kmalloc failure
The kernel's CFI implementation uses the KCFI ABI specifically, and is
not strictly tied to a particular compiler. In preparation for GCC
supporting KCFI, rename CONFIG_CFI_CLANG to CONFIG_CFI (along with
associated options).
Use new "transitional" Kconfig option for old CONFIG_CFI_CLANG that will
enable CONFIG_CFI during olddefconfig.
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Reviewed-by: Nathan Chancellor <nathan@kernel.org>
Link: https://lore.kernel.org/r/20250923213422.1105654-3-kees@kernel.org
Signed-off-by: Kees Cook <kees@kernel.org>
The PMU name is appearing twice in:
```
$ perf stat -e uncore_imc_free_running/data_total/ -A true
Performance counter stats for 'system wide':
CPU0 1.57 MiB uncore_imc_free_running_0/uncore_imc_free_running,data_total/
CPU0 1.58 MiB uncore_imc_free_running_1/uncore_imc_free_running,data_total/
0.000892376 seconds time elapsed
```
Use the pmu_name_len_no_suffix to avoid this problem.
Committer testing:
After this patch:
root@x1:~# perf stat -e uncore_imc_free_running/data_total/ -A true
Performance counter stats for 'system wide':
CPU0 1.69 MiB uncore_imc_free_running_0/data_total/
CPU0 1.68 MiB uncore_imc_free_running_1/data_total/
0.002141605 seconds time elapsed
root@x1:~#
Fixes: 7d45f402d3 ("perf evlist: Make uniquifying counter names consistent")
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: Athira Rajeev <atrajeev@linux.ibm.com>
Cc: Chun-Tse Shao <ctshao@google.com>
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: Kan Liang <kan.liang@linux.intel.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
If a user specifies an AUX buffer larger than 2 GiB, the returned size
may exceed 0x80000000. Since the err variable is defined as a signed
32-bit integer, such a value overflows and becomes negative.
As a result, the perf record command reports an error:
0x146e8 [0x30]: failed to process type: 71 [Unknown error 183711232]
Change the type of the err variable to a signed 64-bit integer to
accommodate large buffer sizes correctly.
Fixes: d5652d865e ("perf session: Add ability to skip 4GiB or more")
Reported-by: Tamas Zsoldos <tamas.zsoldos@arm.com>
Signed-off-by: Leo Yan <leo.yan@arm.com>
Acked-by: Namhyung Kim <namhyung@kernel.org>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Ian Rogers <irogers@google.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Link: https://lore.kernel.org/r/20250808-perf_fix_big_buffer_size-v1-1-45f45444a9a4@arm.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Ian Rogers
Thomas Falcon
Arnaldo Carvalho de Melo
Linus Torvalds
Leo Yan
Zecheng Li
Dapeng Mi
Sam James
Suchit Karunakaran
Athira Rajeev
Markus Heidelberg
Kees Cook