Add the perf.data file offset and use perf_event__name() instead of raw
event type integers in the 'problem processing event, skipping it'
messages emitted by process_sample_event() callbacks across annotate,
c2c, diff, kmem, kvm, kwork, lock, report, script, and build-id.
This lets users cross-reference skipped events with 'perf report -D'
output. Also add explicit #include "util/event.h" and <inttypes.h>
where needed to avoid depending on transitive includes.
Reviewed-by: Ian Rogers <irogers@google.com>
Assisted-by: Claude:claude-opus-4.6
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Add the perf.data file offset to the out-of-bounds CPU debug messages
in process_sample_cpu_idle(), process_sample_cpu_frequency(),
process_sample_sched_wakeup(), and process_sample_sched_switch().
Reviewed-by: Ian Rogers <irogers@google.com>
Assisted-by: Claude:claude-opus-4.6
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Add the perf.data file offset to the CPU out-of-bounds and
machine__resolve failure messages emitted when samples are skipped in
process_sched_switch_event(), process_sched_runtime_event(), and
timehist_sched_change_event(). Also switch event type from raw integer
to perf_event__name() string for readability.
Reviewed-by: Ian Rogers <irogers@google.com>
Assisted-by: Claude:claude-opus-4.6
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Add 'at offset %#<hex>' to all warning and error messages in session.c
that fire when events are skipped or processing stops due to validation
failures. This lets users cross-reference with 'perf report -D' output
to inspect the surrounding records and understand the corruption context.
Covers messages in perf_session__process_event() (alignment, min size,
swap failure), perf_session__deliver_event() (no evsel, parse failure,
CPU clamping), machines__deliver_event() (NAMESPACES, TEXT_POKE,
null-terminated string checks for MMAP/MMAP2/COMM/CGROUP/KSYMBOL), and
perf_session__process_user_event() (THREAD_MAP, CPU_MAP, STAT_CONFIG,
BPF_METADATA, HEADER_BUILD_ID).
Reviewed-by: Ian Rogers <irogers@google.com>
Assisted-by: Claude:claude-opus-4.6
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Add a file_offset field to struct perf_sample so that event processing
callbacks can report the byte offset of the problematic event in
perf.data, letting users cross-reference with 'perf report -D' output.
Set sample.file_offset in perf_session__deliver_event(), which is the
common entry point for both file mode (mmap'd offset) and pipe mode
(running byte counter from __perf_session__process_pipe_events).
The assignment is placed after evsel__parse_sample(), which zeroes
the struct via memset.
Preserve file_offset through the deferred callchain delivery path by
storing it in struct deferred_event and restoring it after
evlist__parse_sample() in both evlist__deliver_deferred_callchain()
and session__flush_deferred_samples().
Subsequent patches will use this field in skip/stop warning messages.
Reviewed-by: Ian Rogers <irogers@google.com>
Assisted-by: Claude:claude-opus-4.6
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
'perf annotate' checks that evlist->nr_br_cntr has been incremented to
determine whether to show branch counter information.
However, this data is not populated until after the check when events
are processed.
Therefore, this counter will always be less than zero and the Branch
Count column is never shown. Do this check after events have been
processed and branch counter data is updated.
Reviewed-by: Dapeng Mi <dapeng1.mi@linux.intel.com>
Signed-off-by: Thomas Falcon <thomas.falcon@intel.com>
Acked-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: James Clark <james.clark@linaro.org>
Cc: Jiri Olsa <jolsa@kernel.org>
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>
When displaying branch counter (br_cntr) information, a "+" suffix
represents that event occurrences may have been lost due to branch
counter saturation. However, this indicator was missing in perf script.
Add it back.
Before:
# Branch counter abbr list:
# cpu_core/event=0xc4,umask=0x20/ppp = A
# cpu_core/instructions/ = B
# cpu_core/MEM_INST_RETIRED.ALL_LOADS/ = C
# cpu_core/MEM_LOAD_RETIRED.L2_MISS/ = D
# '-' No event occurs
# '+' Event occurrences may be lost due to branch counter saturated
...
datasym+190:
00005567f9951676 jz 0x5567f995162dr_cntr: BBBC # PRED 1 cycles [1]
...
After:
...
datasym+190:
00005567f9951676 jz 0x5567f995162dr_cntr: BBB+C # PRED 1 cycles [1]
...
Reviewed-by: Dapeng Mi <dapeng1.mi@linux.intel.com>
Signed-off-by: Dapeng Mi <dapeng1.mi@linux.intel.com>
Signed-off-by: Thomas Falcon <thomas.falcon@intel.com>
Acked-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: James Clark <james.clark@linaro.org>
Cc: Jiri Olsa <jolsa@kernel.org>
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>
Opening an SPE event shows a warning that doesn't concern the user:
$ perf record -e arm_spe
Unknown/empty format name: discard
Perf only wants to know if the discard bit is set for configuring the
event, not in response to anything the user has done. Fix it by adding
another helper that returns if a config bit exists without warning.
We should probably keep the warning in evsel__get_config_val() to avoid
having every caller having to do it, and most format bits should never
be missing.
Add a test for the new helper. Rename the parent test function to be
more generic rather than adding a new one as it requires a lot of
boilerplate.
Reviewed-by: Ian Rogers <irogers@google.com>
Reviewed-by: Leo Yan <leo.yan@arm.com>
Signed-off-by: James Clark <james.clark@linaro.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: John Garry <john.g.garry@oracle.com>
Cc: Leo Yan <leo.yan@linux.dev>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Mike Leach <mike.leach@linaro.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Will Deacon <will@kernel.org>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
perf_exe() passes len to readlink() and then unconditionally writes a
trailing NUL at buf[n]. If readlink() returns len, the write lands one
byte past the buffer.
Read at most len - 1 bytes and keep the existing NUL termination. Also
guard the fallback path for tiny buffers so copying "perf" cannot
overflow.
Reviewed-by: Ian Rogers <irogers@google.com>
Signed-off-by: Miguel Martín Gil <miguel.martin.gil.uni@gmail.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Add IOMMU Translation Lookaside Buffer (TLB) and interrupt cache metrics
to perf jevents for Intel platforms. This enhances I/O performance
observability, allowing fleet-wide monitoring of IOMMU overhead.
These metrics are supported on platforms that expose the required uncore
IIO IOMMU events (such as Emerald Rapids and Granite Rapids).
The Intel implementation dynamically detects event availability at
generation time.
It requires at least the TLB events to expose the metric group, while
the interrupt cache events are optional. This allows platforms like
Emerald Rapids, which lack IOMMU interrupt cache events, to still expose
the IOMMU TLB metrics.
The following metrics are added:
- iotlb_total_hit: Total IOTLB hits (4K, 2M, 1G pages).
- iotlb_total_miss: Total IOTLB misses.
- iotlb_miss_rate: IOTLB miss rate.
- iotlb_interrupt_cache_hit: Interrupt cache hits.
- iotlb_interrupt_cache_miss: Interrupt cache misses (calculated as
lookup - hit, clamped to zero).
- iotlb_interrupt_cache_lookup: Interrupt cache lookups.
- iotlb_interrupt_cache_miss_rate: Interrupt cache miss rate.
Tested:
# perf stat -M \
iotlb_total_hit,iotlb_total_miss,iotlb_miss_rate \
--per-socket --metric-only -a -j -- sleep 10
{"socket" : "S0", "counters" : 10,
"hits iotlb_total_hit" : "3579249.0",
"% iotlb_miss_rate" : "0.0",
"misses iotlb_total_miss" : "3.0"}
{"socket" : "S1", "counters" : 10,
"hits iotlb_total_hit" : "0.0",
"% iotlb_miss_rate" : "0.0",
"misses iotlb_total_miss" : "0.0"}
Reviewed-by: Ian Rogers <irogers@google.com>
Signed-off-by: Chun-Tse Shao <ctshao@google.com>
Assisted-by: Gemini:gemini-3.1-pro-preview
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Dapeng Mi <dapeng1.mi@intel.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: James Clark <james.clark@linaro.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Perry Taylor <perry.taylor@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Sandipan Das <sandipan.das@amd.com>
Link: https://lore.kernel.org/r/20260528234455.434027-3-ctshao@google.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Add IOMMU Translation Lookaside Buffer (TLB) and interrupt cache metrics
to perf jevents for AMD platforms. This enhances I/O performance
observability, allowing fleet-wide monitoring of IOMMU overhead.
These metrics are supported on Zen 2 and newer processors (Rome, Milan,
Genoa, Turin) and are implemented using the standard `amd_iommu` PMU
events. The implementation uses the existing `_zen_model` helper to
ensure these are only generated for Zen 2+. Note that the pde events on
AMD cover both 2M and 1G pages, so 1G pages are implicitly included in
the total hits/misses metrics (sum of pte and pde events).
The following metrics are added:
- iotlb_total_hit: Total IOTLB hits (4K, 2M, 1G pages).
- iotlb_total_miss: Total IOTLB misses.
- iotlb_miss_rate: IOTLB miss rate.
- iotlb_interrupt_cache_hit: Interrupt cache hits.
- iotlb_interrupt_cache_miss: Interrupt cache misses.
- iotlb_interrupt_cache_lookup: Interrupt cache lookups.
- iotlb_interrupt_cache_miss_rate: Interrupt cache miss rate.
Tested:
# perf stat -M \
iotlb_total_hit,iotlb_total_miss,iotlb_miss_rate \
--per-socket --metric-only -a -j -- sleep 10
{"socket" : "S0", "counters" : 10,
"hits iotlb_total_hit" : "3579249.0",
"% iotlb_miss_rate" : "0.0",
"misses iotlb_total_miss" : "3.0"}
{"socket" : "S1", "counters" : 10,
"hits iotlb_total_hit" : "0.0",
"% iotlb_miss_rate" : "0.0",
"misses iotlb_total_miss" : "0.0"}
Reviewed-by: Ian Rogers <irogers@google.com>
Reviewed-by: Sandipan Das <sandipan.das@amd.com>
Signed-off-by: Chun-Tse Shao <ctshao@google.com>
Assisted-by: Gemini:gemini-3.1-pro-preview
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Dapeng Mi <dapeng1.mi@intel.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: James Clark <james.clark@linaro.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Perry Taylor <perry.taylor@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
The updated events and metrics were published in:
efa15280e05cdba6c2cc
Reviewed-by: Dapeng Mi <dapeng1.mi@linux.intel.com>
Signed-off-by: Ian Rogers <irogers@google.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Andreas Färber <afaerber@suse.de>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: James Clark <james.clark@linaro.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Manivannan Sadhasivam <mani@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Falcon <thomas.falcon@intel.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
The updated events and metrics were published in:
0718b7855542fe96774f
Reviewed-by: Dapeng Mi <dapeng1.mi@linux.intel.com>
Signed-off-by: Ian Rogers <irogers@google.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Andreas Färber <afaerber@suse.de>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: James Clark <james.clark@linaro.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Manivannan Sadhasivam <mani@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Falcon <thomas.falcon@intel.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
The updated events and metrics were published in:
f5593317a6
Reviewed-by: Dapeng Mi <dapeng1.mi@linux.intel.com>
Signed-off-by: Ian Rogers <irogers@google.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Andreas Färber <afaerber@suse.de>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: James Clark <james.clark@linaro.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Manivannan Sadhasivam <mani@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Falcon <thomas.falcon@intel.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
The updated events and metrics were published in:
419a6600ad
Reviewed-by: Dapeng Mi <dapeng1.mi@linux.intel.com>
Signed-off-by: Ian Rogers <irogers@google.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Andreas Färber <afaerber@suse.de>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: James Clark <james.clark@linaro.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Manivannan Sadhasivam <mani@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Falcon <thomas.falcon@intel.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
The updated events and metrics were published in:
fae822a0f9
Reviewed-by: Dapeng Mi <dapeng1.mi@linux.intel.com>
Signed-off-by: Ian Rogers <irogers@google.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Andreas Färber <afaerber@suse.de>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: James Clark <james.clark@linaro.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Manivannan Sadhasivam <mani@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Falcon <thomas.falcon@intel.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
The updated events and metrics were published in:
5b93c8c750
Reviewed-by: Dapeng Mi <dapeng1.mi@linux.intel.com>
Signed-off-by: Ian Rogers <irogers@google.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Andreas Färber <afaerber@suse.de>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: James Clark <james.clark@linaro.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Manivannan Sadhasivam <mani@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Falcon <thomas.falcon@intel.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
The updated events and metrics were published in:
50159a7712
Reviewed-by: Dapeng Mi <dapeng1.mi@linux.intel.com>
Signed-off-by: Ian Rogers <irogers@google.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Andreas Färber <afaerber@suse.de>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: James Clark <james.clark@linaro.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Manivannan Sadhasivam <mani@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Falcon <thomas.falcon@intel.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
The updated events and metrics were published in:
526f1bf0ad
Reviewed-by: Dapeng Mi <dapeng1.mi@linux.intel.com>
Signed-off-by: Ian Rogers <irogers@google.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Andreas Färber <afaerber@suse.de>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: James Clark <james.clark@linaro.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Manivannan Sadhasivam <mani@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Falcon <thomas.falcon@intel.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
The updated events and metrics were published in:
6de6be144be7f5e6092c9c1a1baa516055edb3c3
Reviewed-by: Dapeng Mi <dapeng1.mi@linux.intel.com>
Signed-off-by: Ian Rogers <irogers@google.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Andreas Färber <afaerber@suse.de>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: James Clark <james.clark@linaro.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Manivannan Sadhasivam <mani@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Falcon <thomas.falcon@intel.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
The updated events were published in:
90c505bcd9
Reviewed-by: Dapeng Mi <dapeng1.mi@linux.intel.com>
Signed-off-by: Ian Rogers <irogers@google.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Andreas Färber <afaerber@suse.de>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: James Clark <james.clark@linaro.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Manivannan Sadhasivam <mani@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Falcon <thomas.falcon@intel.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
The updated events were published in:
e55693d19f25a1cd4847
Signed-off-by: Ian Rogers <irogers@google.com>
Reviewed-by: Dapeng Mi <dapeng1.mi@linux.intel.com>
Link: https://lore.kernel.org/r/20260529045155.311805-3-irogers@google.com
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Thomas Falcon <thomas.falcon@intel.com>
Cc: Arnaldo Carvalho de Melo <acme@kernel.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Manivannan Sadhasivam <mani@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: James Clark <james.clark@linaro.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Andreas Färber <afaerber@suse.de>
Cc: linux-kernel@vger.kernel.org
Cc: linux-perf-users@vger.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
The updated events were published in:
e55693d19f25a1cd4847
Reviewed-by: Dapeng Mi <dapeng1.mi@linux.intel.com>
Signed-off-by: Ian Rogers <irogers@google.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Andreas Färber <afaerber@suse.de>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: James Clark <james.clark@linaro.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Manivannan Sadhasivam <mani@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Falcon <thomas.falcon@intel.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Since the leafloop test program was moved into the main Perf binary as a
workload, it inherited the same compiler options as Perf. In this case
the -fstack-protector option broke the assumption that simple leaf
frames don't have a stack frame on Arm. This causes
test_arm_callgraph_fp.sh to pass even if the stack isn't augmented with
the link register, making the test useless.
Fix it by rewriting the leaf function in assembly seeing as it's so
simple. Adding -fno-stack-protector would also work, but wouldn't be
robust against other future compiler option additions.
The local variables and 'a' variable were never needed so remove them to
simplify.
Reviewed-by: Ian Rogers <irogers@google.com>
Assisted-by: GitHub-Copilot:GPT-5.5
Signed-off-by: James Clark <james.clark@linaro.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: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Add a shell test that verifies perf report handles truncated perf.data
files gracefully — exiting with an error code rather than crashing with
SIGSEGV or SIGABRT.
The test records a simple workload, then truncates the resulting
perf.data at four offsets that exercise different parsing stages:
8 bytes — file header magic only
64 bytes — partial file header (attr section incomplete)
256 bytes — into the first events (partial event headers)
75% size — mid-stream truncation (partial event data)
For each truncation, perf report is run and the exit code is checked:
- Exit code 0 (success) fails the test — a truncated file should
never parse without error.
- Crash signals are detected portably via kill -l, which maps the
signal number to a name on the running system. This handles
architectures where signal numbers differ (e.g. SIGBUS is 7 on
x86/ARM but 10 on MIPS/SPARC). Core-dump and fatal signals
(KILL, ILL, ABRT, BUS, FPE, SEGV, TRAP, SYS) fail the test.
- Higher exit codes (200+) are perf's own negative-errno returns
(e.g. -EINVAL = 234) and are expected.
This exercises the bounds checking, minimum-size validation, and error
propagation added by the preceding patches in this series.
Testing it:
root@number:~# perf test truncat
84: Test that perf report handles truncated perf.data gracefully (no crash, no segfault — clean error exit).: Ok
root@number:~# perf test -vv truncat
84: Test that perf report handles truncated perf.data gracefully (no crash, no segfault — clean error exit).:
--- start ---
test child forked, pid 62890
---- end(0) ----
84: Test that perf report handles truncated perf.data gracefully (no crash, no segfault — clean error exit).: Ok
root@number:~#
Changes in v2:
- Add SIGKILL to the list of fatal signals so OOM kills from
resource exhaustion bugs are detected (Reported-by: sashiko-bot@kernel.org)
Reviewed-by: Ian Rogers <irogers@google.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Assisted-by: Claude:claude-opus-4.6-1m
[ Fixed the SPDX on the line where 'perf test' expects the test description, reviewed by Ian Rogers ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
On native-endian files, events are read from MAP_SHARED memory.
Multiple reads of event->header.size can return different values
if the file is concurrently modified, allowing an attacker to
bypass bounds checks performed on an earlier read.
Snapshot header.size into a local variable at function entry using
READ_ONCE() to prevent compiler rematerialization, and use it for
all size-dependent arithmetic within the function. This ensures
every bounds calculation uses the same value that was validated
by the reader.
Reported-by: sashiko-bot@kernel.org # Running on a local machine
Reviewed-by: Ian Rogers <irogers@google.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Assisted-by: Claude:claude-opus-4.6-1m
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
work->cpu comes from sample->cpu which is (u32)-1 when
PERF_SAMPLE_CPU is absent. Stored as int, this becomes -1
which passes the signed BUG_ON(work->cpu >= MAX_NR_CPUS) but
causes an out-of-bounds access on cpus_runtime[-1].
Replace the BUG_ON in top_calc_total_runtime() with an unsigned
bounds check that skips entries with invalid CPU values, counting
them for a summary warning.
Guard the same index in profile_event_match() (bitmap OOB),
top_calc_idle_time(), top_calc_irq_runtime(), top_calc_cpu_usage(),
and top_calc_load_runtime(). Also guard against division by zero
in top_calc_cpu_usage() when no runtime was accumulated.
Reported-by: sashiko-bot@kernel.org # Running on a local machine
Reviewed-by: Ian Rogers <irogers@google.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Yang Jihong <yangjihong@bytedance.com>
Assisted-by: Claude:claude-opus-4.6-1m
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Several downstream consumers (timechart, kwork, sched) use fixed-size
arrays indexed by CPU. A crafted perf.data can supply arbitrary CPU
values that index past these arrays, causing out-of-bounds access.
Validate sample.cpu against min(nr_cpus_avail, MAX_NR_CPUS) in
perf_session__deliver_event() before any tool callback runs. The
cap at MAX_NR_CPUS protects fixed-size downstream arrays; the true
nr_cpus_avail is preserved in env for header parsing (e.g.
process_cpu_topology) which needs the real count.
Fall back to MAX_NR_CPUS when HEADER_NRCPUS is missing (truncated
files, pipe mode, pre-2017 perf).
Only validate when PERF_SAMPLE_CPU is set in sample_type — when
absent, evsel__parse_sample() leaves sample.cpu as (u32)-1, a
sentinel that downstream tools (script, inject) check to identify
events without CPU info. Clamping it to 0 would break those checks.
Inline evlist__parse_sample() into perf_session__deliver_event()
so the evsel lookup needed for sample_type checking reuses the same
evsel that parsed the sample, avoiding a second evlist__event2evsel()
call on every event.
For pipe-mode streams where HEADER_NRCPUS may arrive late or not at
all, the MAX_NR_CPUS fallback ensures the bounds check is still
effective against the fixed-size downstream arrays.
Reported-by: sashiko-bot@kernel.org # Running on a local machine
Reviewed-by: Ian Rogers <irogers@google.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Assisted-by: Claude:claude-opus-4.6-1m
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
On 32-bit systems, sizeof(struct decomp) + decomp_len can wrap
size_t when comp_mmap_len is large. The preceding patch validates
comp_mmap_len alignment but does not cap the upper bound, so two
additions can still overflow:
1. decomp_len += decomp_last_rem: on 32-bit, adding a u64 to
size_t silently truncates, producing a corrupted decomp_len
that would bypass the subsequent overflow check and result
in an undersized buffer allocation.
2. sizeof(struct decomp) + decomp_len: the final addition could
overflow on systems with small size_t.
Add explicit overflow checks before each addition as
defense-in-depth.
Reported-by: sashiko-bot@kernel.org # Running on a local machine
Reviewed-by: Ian Rogers <irogers@google.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Assisted-by: Claude:claude-opus-4.6-1m
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Add several hardening checks to the compressed event decompression
pipeline:
1. Guard against decomp_last_rem underflow: check that
decomp_last->head does not exceed decomp_last->size before
subtracting. A u64 underflow here would produce a huge
decomp_len, causing an oversized mmap allocation.
2. Validate comp_mmap_len from the HEADER_COMPRESSED feature
section: reject values that are not 4K-aligned or smaller than
4096. The downstream decompression path checks allocation
sizes against SIZE_MAX, which handles 32-bit safety.
3. Validate COMPRESSED event header size: reject events where
header.size is too small to contain the fixed struct fields,
preventing underflow in the payload size calculation.
4. Validate COMPRESSED2 event data_size: check that data_size
does not exceed the available payload (header.size minus the
fixed struct fields) for the newer compressed format.
5. Reject compressed events when the HEADER_COMPRESSED feature
is missing from the file header, which means no decompression
context was initialized.
Reported-by: sashiko-bot@kernel.org # Running on a local machine
Reviewed-by: Ian Rogers <irogers@google.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Assisted-by: Claude:claude-opus-4.6-1m
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
PERF_RECORD_COMPRESSED2 events carry a data_size field that must be
byte-swapped when reading cross-endian perf.data files. Without a
swap handler, reading COMPRESSED2 events on a different-endian machine
would misinterpret data_size as a garbage value, causing the
decompression path to read the wrong number of bytes.
The compressed payload itself is a raw byte stream and needs no
swapping.
Fixes: 208c0e1683 ("perf record: Add 8-byte aligned event type PERF_RECORD_COMPRESSED2")
Reported-by: sashiko-bot@kernel.org # Running on a local machine
Reviewed-by: Ian Rogers <irogers@google.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Chun-Tse Shao <ctshao@google.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Assisted-by: Claude:claude-opus-4.6-1m
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
do_read_bitmap() reads a u64 bit count from the file and passes it
to bitmap_zalloc() without checking it against the remaining section
size. A crafted perf.data could trigger a large allocation that would
only fail later when the per-element reads exceed section bounds.
Additionally, bitmap_zalloc() takes an int parameter, so a crafted
size with bits set above bit 31 (e.g. 0x100000040) would pass the
section bounds check but truncate when passed to bitmap_zalloc(),
allocating a much smaller buffer than the subsequent read loop
expects.
Reject size values that exceed INT_MAX, and check that the data
needed (BITS_TO_U64(size) u64 values) fits in the remaining section
before allocating. Switch from bitmap_zalloc() to calloc() of u64
units so the allocation size matches the u64 read/write granularity
and avoids unsigned long vs u64 mismatch on 32-bit architectures.
Fix do_write_bitmap() to use memcpy to read u64-sized chunks from
the unsigned long bitmap, preventing out-of-bounds reads on 32-bit
systems where sizeof(unsigned long) is 4 but the bitmap is stored
in u64 units.
Fix process_mem_topology() minimum section size: the check used
nr * 2 * sizeof(u64) per node, but do_read_bitmap() reads an
additional u64 for the bitmap size, so the minimum is 3 * sizeof(u64).
Fix memory leak in process_mem_topology() error paths: replace
free(nodes) with memory_node__delete_nodes() to free per-node
bitmaps allocated by do_read_bitmap().
Currently used by process_mem_topology() for HEADER_MEM_TOPOLOGY.
Fixes: a881fc5603 ("perf header: Sanity check HEADER_MEM_TOPOLOGY")
Closes: https://lore.kernel.org/linux-perf-users/20260414224622.2AE69C19425@smtp.kernel.org/
Closes: https://lore.kernel.org/linux-perf-users/20260410223242.DD76FC19421@smtp.kernel.org/
Reported-by: sashiko-bot@kernel.org # Running on a local machine
Reviewed-by: Ian Rogers <irogers@google.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Assisted-by: Claude:claude-opus-4.6-1m
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
read_event_desc() reads nre (event count), sz (attr size), and nr
(IDs per event) from the file and uses them to control allocations
and loops without validating them against the section size.
A crafted perf.data could trigger large allocations or many loop
iterations before __do_read() eventually rejects the reads.
Add bounds checks in read_event_desc():
- Reject sz smaller than PERF_ATTR_SIZE_VER0.
- Require at least one event (nre > 0).
- Check that nre events fit in the remaining section, using the
minimum per-event footprint of sz + sizeof(u32).
- Pre-swap attr->size to native byte order, then reject values
below PERF_ATTR_SIZE_VER0 or above sz before calling
perf_event__attr_swap() to prevent heap out-of-bounds access.
- Handle ABI0 (attr.size == 0): substitute PERF_ATTR_SIZE_VER0,
and on native-endian files write the value back so
free_event_desc() does not treat the zero as its end-of-array
sentinel (it iterates while attr.size != 0). The swap path
skips the write-back — perf_event__attr_swap() has its own
ABI0 fallback that sets VER0 after swapping.
- Check that nr IDs fit in the remaining section before allocating.
Fixes: b30b617292 ("perf tools: Fix a problem when opening old perf.data with different byte order")
Reported-by: sashiko-bot@kernel.org # Running on a local machine
Reviewed-by: Ian Rogers <irogers@google.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Assisted-by: Claude:claude-opus-4.6-1m
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Harden feature section parsing against crafted perf.data files:
1. perf_header__process_sections() reads the feature section table
and passes each section's offset and size directly to the
processing callbacks without validating them against the actual
file size. A crafted section size would make all downstream
bounds checks against ff->size ineffective since they compare
against the untrusted, inflated bound. Add an fstat() check
with S_ISREG() guard and verify that each section's offset +
size does not extend past EOF.
2. __do_read_buf() validates reads against ff->size (section size),
but __do_read_fd() had no such check, so a malformed perf.data
with an understated section size could cause reads past the end
of the current section into the next section's data. Add the
bounds check in __do_read(), the common caller of both helpers,
so it is enforced uniformly for both the fd and buf paths.
Track the section-relative offset in __do_read_fd() so the
check works for the fd path. Reject negative sizes which on
32-bit can occur when a u32 >= 0x80000000 is passed as ssize_t.
3. do_read_string() relied on file data being null-padded. Add
explicit null-termination (buf[len-1] = '\0') after reading
and validate length (>= 1, fits within section) before
allocating, so callers like process_cpu_topology() never
receive an unterminated string.
4. Initialize feat_fd.offset to 0 (section-relative) instead of
section->offset (file-absolute) so the bounds tracking is
consistent with __do_read()'s section-relative comparison.
Adjust process_build_id() to use lseek() for its file-absolute
offset needs since it cannot rely on ff->offset for that.
5. Propagate ff->size to perf_file_section__fprintf_info() so its
reads are also bounded.
Reported-by: sashiko-bot@kernel.org # Running on a local machine
Reviewed-by: Ian Rogers <irogers@google.com>
Cc: David Carrillo-Cisneros <davidcc@google.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Assisted-by: Claude:claude-opus-4.6-1m
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
perf_session__read_header() reads f_attr.ids.size from the perf.data
file and divides it by sizeof(u64) to compute nr_ids, which is
declared as int. No validation is performed on the value before it
is used to allocate arrays and drive a read loop.
On 32-bit architectures, a crafted f_attr.ids.size of 0x100000000
(4 GB) produces nr_ids = 0x20000000, but the allocation size
1 * 0x20000000 * 8 overflows size_t to 0, so zalloc(0) returns a
valid pointer. The subsequent loop writes 0x20000000 IDs into that
zero-length buffer, corrupting the heap.
On 64-bit, the u64-to-int truncation silently drops high bits,
processing fewer IDs than the file claims. While not exploitable,
this is a data integrity issue.
Add validation before using f_attr.ids:
- Cap nr_attrs (attrs.size / attr_size) to MAX_NR_ATTRS (1 << 16)
with overflow-safe u64 comparison before assigning to int
- Reject ids.size not aligned to sizeof(u64)
- Cap ids.size / sizeof(u64) to MAX_IDS_PER_ATTR (1 << 24) to
prevent int truncation and size_t overflow on 32-bit
- Reject ids sections that extend past the end of the file,
guarded by S_ISREG() so non-regular files (block devices,
pipes) are not falsely rejected
Also fix perf_header__getbuffer64() to set errno = EIO when
readn() returns 0 (EOF). Without this, the out_errno path in
perf_session__read_header() returns -errno which is 0 (success)
on truncated files, causing downstream NULL dereferences.
Reported-by: sashiko-bot@kernel.org # Running on a local machine
Reviewed-by: Ian Rogers <irogers@google.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Assisted-by: Claude:claude-opus-4.6-1m
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
perf_session__read_header() discards the return value from
perf_header__process_sections(), so any error from a feature
section processor (process_nrcpus, process_compressed, etc.)
is silently ignored and the session opens as if nothing went
wrong.
This defeats the validation added by subsequent commits in this
series: a crafted perf.data that fails a feature section check
would still be processed with partially-initialized state.
Check the return value and fail the session if any feature
section processor returns an error.
For truncated files (data.size == 0, i.e. recording was
interrupted before the header was finalized), skip feature
section processing entirely and clear the feature bitmap so
tools use their "feature not present" fallbacks instead of
accessing uninitialized env fields.
Change the feature processor stubs for optional libraries
(libtraceevent, libbpf) from returning -1 to returning 0,
so that perf.data files containing these features can still be
opened on builds without the optional library — the feature is
simply skipped rather than causing a fatal error.
Also propagate evlist__prepare_tracepoint_events() failure as
-ENOMEM, since the function can fail due to strdup() allocation
failure inside evsel__prepare_tracepoint_event().
Fixes: 1c0b04d12ae9 ("perf tools: Add perf_session__read_header function")
Reviewed-by: Ian Rogers <irogers@google.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Assisted-by: Claude:claude-opus-4.6-1m
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
perf_event_attr__fprintf() accessed all struct fields unconditionally,
but attrs from older perf.data files or BPF-captured syscall payloads
may have a smaller size than the current struct. Fields beyond the
recorded size contain uninitialized or zero-filled data.
Add size-guarded macros (PRINT_ATTRn, PRINT_ATTRn_bf) that compare
each field's offset against attr->size before accessing it.
Guard the bitfield block (disabled, inherit, ... defer_output) with
attr_size >= 48. These bitfields share a single __u64 at offset 40,
which is within PERF_ATTR_SIZE_VER0 for validated perf.data attrs,
but BPF-captured attrs from perf trace can have a smaller size when
the tracee passes a minimal struct to sys_perf_event_open.
Also fix the BPF trace path: when perf trace intercepts
sys_perf_event_open via BPF, the program copies PERF_ATTR_SIZE_VER0
bytes when the tracee passes size=0, but leaves the size field as 0.
Set attr->size to PERF_ATTR_SIZE_VER0 in the augmented syscall
handler so the bounds checks match the actual copied size.
Reported-by: sashiko-bot@kernel.org # Running on a local machine
Reviewed-by: Ian Rogers <irogers@google.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Assisted-by: Claude:claude-opus-4.6-1m
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
strdup(ev->unit) and strdup(ev->name) read until '\0' with no
guarantee the string is null-terminated within event->header.size.
The dump_trace fprintf path has the same problem with %s.
Validate before either path runs — same class of bug fixed for
MMAP/MMAP2/COMM/CGROUP by perf_event__check_nul().
Also harden the event_update swap handler to:
- Validate SCALE event size before swapping the double at
offset 24, which exceeds the 24-byte min_size.
- Validate CPUS event size before accessing the cpu_map
type/nr/long_size fields, which also start at the min_size
boundary.
- Swap CPUS variant fields (type, nr, long_size) so the
processing path sees native byte order.
Add validation in perf_event__process_event_update() for all
event update variants (UNIT, NAME, SCALE, CPUS) before
dump_trace or processing.
Validate CPUS nr against payload size for both PERF_CPU_MAP__CPUS
and PERF_CPU_MAP__MASK types on the fprintf (dump_trace) path:
- CPUS: check nr does not exceed available cpu entries
- MASK: check nr does not exceed available mask entries for
both mask32 (long_size == 4) and mask64 (long_size == 8)
layouts, with underflow guards on the offsetof subtraction
Fix a missing break before the default case in the CPUS
switch path.
Reported-by: sashiko-bot@kernel.org # Running on a local machine
Reviewed-by: Ian Rogers <irogers@google.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Assisted-by: Claude:claude-opus-4.6-1m
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
PERF_RECORD_BPF_METADATA has no entry in perf_event__swap_ops[], so its
nr_entries field is never byte-swapped when reading a cross-endian
perf.data file. Downstream processing in
perf_event__fprintf_bpf_metadata() loops over nr_entries, so a
foreign-endian value causes out-of-bounds reads.
Add a swap handler that byte-swaps nr_entries after validating that
header.size is large enough. The entries[] array contains only char
arrays (key/value strings), so no per-entry swap is needed — but ensure
NUL-termination on the writable cross-endian path.
Validate header.size, nr_entries, and string NUL-termination in the
common event delivery path so that native-endian files with malicious
values are also rejected. Snapshot nr_entries via READ_ONCE() before
validation — the event is on a MAP_SHARED mmap that could theoretically
change between the bounds check and the loop.
Changes in v2:
- Snapshot event->header.size via READ_ONCE() into a local variable
to prevent a double-fetch underflow in the max_entries calculation
(Reported-by: sashiko-bot@kernel.org)
- Write back clamped nr_entries to the event on the swap path,
consistent with NAMESPACES and STAT_CONFIG handlers — without
writeback the native path sees the inflated nr and skips the
event entirely (Reported-by: sashiko-bot@kernel.org)
Fixes: ab38e84ba9 ("perf record: collect BPF metadata from existing BPF programs")
Reported-by: sashiko-bot@kernel.org # Running on a local machine
Reviewed-by: Ian Rogers <irogers@google.com>
Cc: Blake Jones <blakejones@google.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Assisted-by: Claude:claude-opus-4.6-1m
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Fix four issues in PERF_RECORD_AUXTRACE_ERROR handling:
1. auxtrace_error_name() takes a signed int parameter, but e->type
is __u32. A crafted value like 0xFFFFFFFF converts to -1, passes
the bounds check, and causes a negative array index. Fix by
changing the parameter to unsigned int.
2. The msg field is printed via %s without a length bound. The
min_size table only guarantees fields up to msg (offset 48), so
a truncated event has zero msg bytes within the event boundary.
Compute the available msg length from header.size, cap at
sizeof(e->msg), and use %.*s.
3. fmt >= 2 adds machine_pid and vcpu fields after msg[64]. Older
files may have fmt >= 2 but an event size that doesn't include
these fields. Add a size check in the swap handler to downgrade
fmt before the conditional field access, and a matching size
guard in the fprintf path for native-endian events (which are
mmap'd read-only and can't be modified in place).
4. python_process_auxtrace_error() had the same issues: msg was
passed to tuple_set_string() unbounded, and machine_pid/vcpu
were accessed unconditionally without checking fmt or event
size. Apply the same bounds checks.
Reported-by: sashiko-bot@kernel.org # Running on a local machine
Reviewed-by: Ian Rogers <irogers@google.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Assisted-by: Claude:claude-opus-4.6-1m
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
cpu_map__from_range() computes nr_cpus as end_cpu - start_cpu + 1.
When a crafted perf.data has start_cpu > end_cpu, this wraps to a
huge value, causing perf_cpu_map__empty_new() to attempt a massive
allocation.
Return NULL when the range is inverted.
Also clamp any_cpu to boolean (0 or 1) since it is added to the
allocation count — a crafted value > 1 would inflate the map size.
Harden cpu_map__from_mask() to reject unsupported long_size values
(anything other than 4 or 8), preventing misinterpretation of the
mask data layout.
Snapshot mmap'd fields via READ_ONCE() into locals to prevent
TOCTOU re-reads — the data pointer references MAP_SHARED mmap'd
memory that could theoretically change between reads on a
FUSE-backed file:
- cpu_map__from_range(): snapshot start_cpu, end_cpu, any_cpu
- cpu_map__from_entries(): snapshot nr and each cpu[i] element
- cpu_map__from_mask(): snapshot long_size (before validation,
closing the check-then-read gap), mask_nr
- perf_record_cpu_map_data__read_one_mask(): add u16 long_size
parameter so callers pass the validated copy instead of
re-reading data->mask32_data.long_size from mmap'd memory
Reported-by: sashiko-bot@kernel.org # Running on a local machine
Reviewed-by: Ian Rogers <irogers@google.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Assisted-by: Claude:claude-opus-4.6-1m
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
perf_header__read_build_ids() swaps the event header fields for cross-endian
perf.data files but not bev.pid. This causes perf_session__findnew_machine()
to look up the wrong machine for guest VM build IDs, misattributing them.
Swap bev.pid alongside the header fields.
Also add a build_id_swap callback for stream-mode build ID events,
and validate NUL-termination of build_id.filename on the native-endian
delivery path (perf_session__process_user_event) — events with
unterminated filenames are skipped.
Harden perf_header__read_build_ids() against crafted perf.data files:
- Add overflow check on offset + size to prevent wrap past ULLONG_MAX.
- Reject bev.header.size == 0 which would loop forever.
- Reject bev.header.size > remaining section to prevent reading past
the section boundary.
- Guard memcmp(filename, "nel.kallsyms]", 13) with len >= 13 to avoid
reading uninitialized stack memory on short filenames.
- Force NUL-termination of filename before passing it to functions
like machine__findnew_dso() that use strlen/strcmp.
Reported-by: sashiko-bot@kernel.org # Running on a local machine
Reviewed-by: Ian Rogers <irogers@google.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Assisted-by: Claude:claude-opus-4.6-1m
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Several event types use an nr field to control iteration over
variable-length arrays. The swap handlers byte-swap and loop using
these fields without bounds checks, and the native processing path
trusts them as well.
Add bounds checks on both paths for:
- PERF_RECORD_THREAD_MAP: validate nr against payload, return -1
on the swap path. On the native path, reject with -EINVAL.
- PERF_RECORD_NAMESPACES: clamp nr on the swap path (safe because
each entry is indexed by type; missing entries just won't be
resolved). Skip the event on the native path.
- PERF_RECORD_CPU_MAP: clamp nr for CPUS and MASK sub-types on
the swap path. Add bounds checks for mask64 which previously
had no nr validation. Skip the event on the native path.
- PERF_RECORD_STAT_CONFIG: clamp nr on the swap path (safe because
each config entry is self-describing via its tag). Skip the
event on the native path.
The swap path (cross-endian, writable MAP_PRIVATE mapping) can
safely clamp by writing back to the event. The native path
(read-only MAP_SHARED mapping) must skip instead of clamping
because writing to the mmap'd event would segfault.
Also fix stat_config swap range: change size += 1 to
size += sizeof(event->stat_config.nr) for clarity. The old +1
happened to work because mem_bswap_64 processes 8-byte chunks,
but the intent is to include the 8-byte nr field in the swap
range.
Changes in v2:
- Document that PERF_RECORD_NAMESPACES max_nr includes trailing
sample_id space when sample_id_all is present — harmless on the
swap path because both per-element bswap_64 and swap_sample_id_all()
perform the same u64 byte swap (Reported-by: sashiko-bot@kernel.org)
Reported-by: sashiko-bot@kernel.org # Running on a local machine
Reviewed-by: Ian Rogers <irogers@google.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Assisted-by: Claude:claude-opus-4.6-1m
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Harden PERF_RECORD_HEADER_ATTR handling against crafted perf.data:
- Validate attr.size: must be >= PERF_ATTR_SIZE_VER0, a multiple
of sizeof(u64), and fit within the event payload.
- Copy only min(attr.size, sizeof(struct perf_event_attr)) bytes
into a local attr, zeroing the rest so legacy files don't leak
adjacent event data into new fields.
- Keep the original attr.size so perf_event__synthesize_attr()
uses it for both allocation and ID-array placement.
Fix perf_event__synthesize_attr() to use attr->size (not the
compiled sizeof) for event allocation and layout, so perf inject
correctly re-synthesizes attrs from files recorded by a different
perf version. Without this, the ID array destination pointer
(computed via perf_record_header_attr_id()) would be inconsistent
with the allocation when attr->size differs from sizeof.
Also fix the parse-no-sample-id-all test to set attr.size, which
is now validated, and improve error handling in read_attr() for
short reads and invalid attr sizes.
Handle ABI0 pipe/inject events where attr.size is 0: use a local
attr_size variable set to PERF_ATTR_SIZE_VER0 for both the bounded
copy and ID array position, instead of writing back to the event.
Native-endian files may be MAP_SHARED (read-only mmap), so writing
to the event buffer would SIGSEGV. The swap path handles ABI0 in
perf_event__attr_swap() which writes to the MAP_PRIVATE copy.
header.size alignment is now validated centrally in
perf_session__process_event() (see "Add minimum event size and
alignment validation").
Reported-by: sashiko-bot@kernel.org # Running on a local machine
Reviewed-by: Ian Rogers <irogers@google.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Assisted-by: Claude:claude-opus-4.6-1m
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
session->time_conv = event->time_conv copies sizeof(struct
perf_record_time_conv) bytes unconditionally, but older kernels
emit shorter TIME_CONV events without the time_cycles, time_mask,
cap_user_time_zero, and cap_user_time_short fields.
For a 32-byte event (the original format), this reads 24 bytes
past the event boundary into adjacent mmap'd data. The garbage
values end up in session->time_conv and can cause incorrect TSC
conversion if cap_user_time_zero happens to be non-zero.
Replace the struct assignment with a bounded memcpy capped at
event->header.size, zeroing the remainder so extended fields
default to off when absent.
Reported-by: sashiko-bot@kernel.org # Running on a local machine
Reviewed-by: Ian Rogers <irogers@google.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Assisted-by: Claude:claude-opus-4.6-1m
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Change swap callbacks from void to int return so handlers can
propagate errors. All 28 existing handlers are converted to
return 0 on success, -1 on error. Three new handlers (KSYMBOL,
BPF_EVENT, HEADER_FEATURE) are added returning int from the
start, with sample_id_all handling for the kernel event types.
event_swap() propagates the return to its callers (process_event
and peek_event), which skip events that fail to swap.
Add perf_event__check_nul() for null-termination enforcement
on the common event delivery path for MMAP, MMAP2, COMM,
CGROUP, and KSYMBOL events. Events with
unterminated strings are skipped — native-endian files are
mapped read-only, so writing a NUL byte in place would segfault.
Swap handler hardening:
- Use strnlen bounded by event size (instead of strlen) in
COMM/MMAP/MMAP2/CGROUP swap handlers, returning -1 on
unterminated strings.
- Bounds check text_poke old_len+new_len before computing the
sample_id offset, returning -1 on overflow. Use offsetof()
for the native-path check in machines__deliver_event() since
sizeof() includes struct padding past the flexible array.
- Fix PERF_RECORD_SWITCH sample_id_all: non-CPU_WIDE SWITCH
events have sample_id immediately after the 8-byte header,
not at sizeof(struct perf_record_switch) which is the
CPU_WIDE variant size.
- Fix perf_event__time_conv_swap(): decouple time_cycles and
time_mask into independent per-field event_contains() checks,
so each field is only swapped when the event is large enough
to contain it. The original code guarded both fields under
a single time_cycles check, which would swap time_mask on a
short event that contains time_cycles but not time_mask.
- Handle ABI0 (attr.size == 0) in perf_event__attr_swap()
by substituting PERF_ATTR_SIZE_VER0, so bswap_safe()
correctly swaps VER0 fields instead of skipping everything.
- peek_events: on swap failure, advance past the malformed
entry instead of aborting the loop.
Note: the nr-field bounds checks for namespaces, thread_map,
cpu_map, and stat_config arrays are added by a subsequent
patch ("perf session: Validate nr fields against event size
on both swap and common paths"). The HEADER_ATTR attr.size
validation is added by ("perf session: Validate HEADER_ATTR
attr.size before swapping").
By establishing the int-returning swap infrastructure first,
all subsequent hardening patches can use direct error returns
from day one — no poison values, no workarounds for void return.
Changes in v2:
- peek_events: abort instead of skip for AUXTRACE events on
validation failure — skipping only header.size would land
inside the raw trace payload, causing subsequent iterations
to misparse data as events (Reported-by: sashiko-bot@kernel.org)
Fixes: 9aa0bfa370 ("perf tools: Handle PERF_RECORD_KSYMBOL")
Fixes: 45178a928a ("perf tools: Handle PERF_RECORD_BPF_EVENT")
Fixes: e9def1b2e7 ("perf tools: Add feature header record to pipe-mode")
Reported-by: sashiko-bot@kernel.org # Running on a local machine
Reviewed-by: Ian Rogers <irogers@google.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Carrillo-Cisneros <davidcc@google.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Song Liu <songliubraving@fb.com>
Assisted-by: Claude:claude-opus-4.6-1m
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
swap_sample_id_all() calls BUG_ON(size % sizeof(u64)) which kills
perf on any event where the sample_id_all tail is not 8-byte aligned.
A crafted perf.data can trigger this trivially.
Replace BUG_ON with a bounds check: skip the swap if the data pointer
is past the end of the event, and only swap when there are bytes
remaining.
Note: the strlen calls in string-field swap handlers (comm,
mmap, mmap2, cgroup) are replaced with bounded strnlen by the
next patch in this series ("perf session: Add validated swap
infrastructure with null-termination checks").
Reported-by: sashiko-bot@kernel.org # Running on a local machine
Reviewed-by: Ian Rogers <irogers@google.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Assisted-by: Claude:claude-opus-4.6-1m
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
The kernel dynamically sizes PERF_RECORD_READ based on
attr.read_format: only the fields enabled by PERF_FORMAT_TOTAL_TIME_ENABLED,
PERF_FORMAT_TOTAL_TIME_RUNNING, PERF_FORMAT_ID, and PERF_FORMAT_LOST
are emitted, packed with no gaps.
perf_event__read_swap() unconditionally byte-swapped time_enabled,
time_running, and id at their fixed struct offsets, causing
out-of-bounds access on smaller events and swapping the wrong
bytes when not all format fields are present. It also swapped
sample_id_all at a fixed offset past the full struct, which is
wrong for shorter events.
Replace the individual field swaps with a single mem_bswap_64()
over the entire tail from value onward. Since every field after
pid/tid is u64 regardless of which combination is present, this
correctly handles any read_format combination and any trailing
sample_id_all fields.
Similarly, dump_read() accessed optional fields via fixed struct
offsets, displaying values from wrong positions when not all
format bits are set. Walk the packed u64 array sequentially
instead, with bounds checks against event->header.size.
Reviewed-by: Ian Rogers <irogers@google.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Assisted-by: Claude:claude-opus-4.6-1m
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
zstd_decompress_stream() has two bugs in its multi-iteration loop:
1. After each ZSTD_decompressStream() call, the code advances
output.dst by output.pos but doesn't reset output.pos to 0.
ZSTD interprets output.pos relative to output.dst, so the
next iteration writes at (dst + pos) + pos = dst + 2*pos,
skipping a gap and potentially writing out of bounds.
2. On ZSTD_decompressStream() error, the loop executes break
and returns output.pos (which is > 0 if some bytes were
decompressed before the error). The caller checks
!decomp_size and skips the error, silently accepting
truncated or corrupted data.
Fix both by removing the output buffer adjustment — ZSTD
correctly accumulates output.pos across calls without it.
Return 0 on decompression error so the caller detects it.
Add a no-progress guard to prevent infinite loops if the
output buffer fills before all input is consumed.
Note: the compressed event data_size is validated against
header.size by a subsequent patch in this series
("perf tools: Harden compressed event processing").
Reported-by: sashiko-bot@kernel.org # Running on a local machine
Reviewed-by: Ian Rogers <irogers@google.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Assisted-by: Claude:claude-opus-4.6-1m
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
The error fallback does memcpy(dst, src, src_size) intending to store
uncompressed data when compression fails, but this has three bugs:
1. dst has been advanced past the record header (and potentially
past earlier compressed records), so the copy writes to the
wrong offset in the output buffer.
2. src still points to the start of the input, not to the
remaining uncompressed data at src + input.pos. On a second
or later iteration, previously compressed data would be
duplicated.
3. No check that dst_size >= src_size — if the remaining output
space is smaller, this is an out-of-bounds write.
Replace with return -1 after resetting the ZSTD compression
context via ZSTD_initCStream(). The -1 propagates through
zstd_compress() -> record__pushfn() -> perf_mmap__push() to the
recording loop, which breaks out and terminates recording.
Add an out_child_no_flush label in __cmd_record() so the
mmap-read failure path skips the final record__mmap_read_all()
flush — retrying the same read that just failed would just fail
again, and the flush is only useful when the mmap data is intact
but the control path (auxtrace, switch_output) had an error.
Consolidate all error paths through a single 'reset' label to
ensure the compression context is always reset on failure —
including the output-buffer-full path, where a bare return
without resetting would leave stale stream state that corrupts
output if the caller retries.
Also guard against process_header() writing the event header
before the buffer-full check: add a sizeof(perf_event_header)
pre-check so the callback never writes past the output buffer.
Guard against ZSTD making no progress: if output.pos is zero
after ZSTD_compressStream(), calling process_header(record, 0)
would re-trigger header initialization, double-subtracting the
header size from dst_size and underflowing the unsigned counter.
Also fix two pre-existing issues in the same function:
- Add a dst_size guard before subtracting the record header
size: if the output buffer is nearly full, the unsigned
dst_size -= size underflows to a huge value, causing
ZSTD_compressStream to write past the buffer boundary.
- Check the ZSTD_initCStream() return value and log an error
if the context reset itself fails.
Reported-by: sashiko-bot@kernel.org # Running on a local machine
Reviewed-by: Ian Rogers <irogers@google.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Assisted-by: Claude:claude-opus-4.6-1m
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
