linux-stable.git/tools/perf, branch v6.1.142

perf record: Fix incorrect --user-regs comments

2025-06-27T10:07:15+00:00

[ Upstream commit a4a859eb6704a8aa46aa1cec5396c8d41383a26b ]

The comment of "--user-regs" option is not correct, fix it.

"on interrupt," -> "in user space,"

Fixes: 84c417422798c897 ("perf record: Support direct --user-regs arguments")
Reviewed-by: Ian Rogers 
Signed-off-by: Dapeng Mi 
Cc: Adrian Hunter 
Cc: Alexander Shishkin 
Cc: Andi Kleen 
Cc: Ingo Molnar 
Cc: Kan Liang 
Cc: Namhyung Kim 
Cc: Peter Zijlstra 
Link: https://lore.kernel.org/r/20250403060810.196028-1-dapeng1.mi@linux.intel.com
Signed-off-by: Arnaldo Carvalho de Melo 
Signed-off-by: Sasha Levin

perf tests switch-tracking: Fix timestamp comparison

2025-06-27T10:07:15+00:00

[ Upstream commit 628e124404b3db5e10e17228e680a2999018ab33 ]

The test might fail on the Arm64 platform with the error:

  # perf test -vvv "Track with sched_switch"
  Missing sched_switch events
  #

The issue is caused by incorrect handling of timestamp comparisons. The
comparison result, a signed 64-bit value, was being directly cast to an
int, leading to incorrect sorting for sched events.

The case does not fail everytime, usually I can trigger the failure
after run 20 ~ 30 times:

  # while true; do perf test "Track with sched_switch"; done
  106: Track with sched_switch                                         : Ok
  106: Track with sched_switch                                         : Ok
  106: Track with sched_switch                                         : Ok
  106: Track with sched_switch                                         : Ok
  106: Track with sched_switch                                         : Ok
  106: Track with sched_switch                                         : Ok
  106: Track with sched_switch                                         : Ok
  106: Track with sched_switch                                         : Ok
  106: Track with sched_switch                                         : Ok
  106: Track with sched_switch                                         : Ok
  106: Track with sched_switch                                         : Ok
  106: Track with sched_switch                                         : Ok
  106: Track with sched_switch                                         : Ok
  106: Track with sched_switch                                         : Ok
  106: Track with sched_switch                                         : FAILED!
  106: Track with sched_switch                                         : Ok
  106: Track with sched_switch                                         : Ok
  106: Track with sched_switch                                         : Ok
  106: Track with sched_switch                                         : Ok
  106: Track with sched_switch                                         : Ok
  106: Track with sched_switch                                         : Ok
  106: Track with sched_switch                                         : Ok
  106: Track with sched_switch                                         : Ok
  106: Track with sched_switch                                         : FAILED!
  106: Track with sched_switch                                         : Ok
  106: Track with sched_switch                                         : Ok

I used cross compiler to build Perf tool on my host machine and tested on
Debian / Juno board.  Generally, I think this issue is not very specific
to GCC versions.  As both internal CI and my local env can reproduce the
issue.

My Host Build compiler:

  # aarch64-linux-gnu-gcc --version
  aarch64-linux-gnu-gcc (Ubuntu 13.3.0-6ubuntu2~24.04) 13.3.0

Juno Board:

  # lsb_release -a
  No LSB modules are available.
  Distributor ID: Debian
  Description:    Debian GNU/Linux 12 (bookworm)
  Release:        12
  Codename:       bookworm

Fix this by explicitly returning 0, 1, or -1 based on whether the result
is zero, positive, or negative.

Fixes: d44bc558297222d9 ("perf tests: Add a test for tracking with sched_switch")
Reviewed-by: Ian Rogers 
Signed-off-by: Leo Yan 
Cc: Adrian Hunter 
Cc: James Clark 
Cc: Kan Liang 
Cc: Namhyung Kim 
Link: https://lore.kernel.org/r/20250331172759.115604-1-leo.yan@arm.com
Signed-off-by: Arnaldo Carvalho de Melo 
Signed-off-by: Sasha Levin

perf scripts python: exported-sql-viewer.py: Fix pattern matching with Python 3

2025-06-27T10:07:15+00:00

[ Upstream commit 17e548405a81665fd14cee960db7d093d1396400 ]

The script allows the user to enter patterns to find symbols.

The pattern matching characters are converted for use in SQL.

For PostgreSQL the conversion involves using the Python maketrans()
method which is slightly different in Python 3 compared with Python 2.

Fix to work in Python 3.

Fixes: beda0e725e5f06ac ("perf script python: Add Python3 support to exported-sql-viewer.py")
Signed-off-by: Adrian Hunter 
Cc: Alexander Shishkin 
Cc: Ian Rogers 
Cc: Jiri Olsa 
Cc: Kan Liang 
Cc: Namhyung Kim 
Cc: Tony Jones 
Link: https://lore.kernel.org/r/20250512093932.79854-4-adrian.hunter@intel.com
Signed-off-by: Arnaldo Carvalho de Melo 
Signed-off-by: Sasha Levin

perf intel-pt: Fix PEBS-via-PT data_src

2025-06-27T10:07:15+00:00

[ Upstream commit e00eac6b5b6d956f38d8880c44bf7fd9954063c3 ]

The Fixes commit did not add support for decoding PEBS-via-PT data_src.
Fix by adding support.

PEBS-via-PT is a feature of some E-core processors, starting with
processors based on Tremont microarchitecture. Because the kernel only
supports Intel PT features that are on all processors, there is no support
for PEBS-via-PT on hybrids.

Currently that leaves processors based on Tremont, Gracemont and Crestmont,
however there are no events on Tremont that produce data_src information,
and for Gracemont and Crestmont there are only:

	mem-loads	event=0xd0,umask=0x5,ldlat=3
	mem-stores	event=0xd0,umask=0x6

Affected processors include Alder Lake N (Gracemont), Sierra Forest
(Crestmont) and Grand Ridge (Crestmont).

Example:

 # perf record -d -e intel_pt/branch=0/ -e mem-loads/aux-output/pp uname

 Before:

  # perf.before script --itrace=o -Fdata_src
            0 |OP No|LVL N/A|SNP N/A|TLB N/A|LCK No|BLK  N/A
            0 |OP No|LVL N/A|SNP N/A|TLB N/A|LCK No|BLK  N/A

 After:

  # perf script --itrace=o -Fdata_src
  10268100142 |OP LOAD|LVL L1 hit|SNP None|TLB L1 or L2 hit|LCK No|BLK  N/A
  10450100442 |OP LOAD|LVL L2 hit|SNP None|TLB L2 miss|LCK No|BLK  N/A

Fixes: 975846eddf907297 ("perf intel-pt: Add memory information to synthesized PEBS sample")
Reviewed-by: Kan Liang 
Signed-off-by: Adrian Hunter 
Cc: Alexander Shishkin 
Cc: Ian Rogers 
Cc: Jiri Olsa 
Cc: Namhyung Kim 
Link: https://lore.kernel.org/r/20250512093932.79854-2-adrian.hunter@intel.com
Signed-off-by: Arnaldo Carvalho de Melo 
Signed-off-by: Sasha Levin

perf ui browser hists: Set actions->thread before calling do_zoom_thread()

2025-06-27T10:07:14+00:00

[ Upstream commit 1741189d843a1d5ef38538bc52a3760e2e46cb2e ]

In 7cecb7fe8388d5c3 ("perf hists: Move sort__has_comm into struct
perf_hpp_list") it assumes that act->thread is set prior to calling
do_zoom_thread().

This doesn't happen when we use ESC or the Left arrow key to Zoom out of
a specific thread, making this operation not to work and we get stuck
into the thread zoom.

In 6422184b087ff435 ("perf hists browser: Simplify zooming code using
pstack_peek()") it says no need to set actions->thread, and at that
point that was true, but in 7cecb7fe8388d5c3 a actions->thread == NULL
check was added before the zoom out of thread could kick in.

We can zoom out using the alternative 't' thread zoom toggle hotkey to
finally set actions->thread before calling do_zoom_thread() and zoom
out, but lets also fix the ESC/Zoom out of thread case.

Fixes: 7cecb7fe8388d5c3 ("perf hists: Move sort__has_comm into struct perf_hpp_list")
Reported-by: Ingo Molnar 
Tested-by: Ingo Molnar 
Cc: Adrian Hunter 
Cc: Ian Rogers 
Cc: James Clark 
Cc: Jiri Olsa 
Cc: Kan Liang 
Cc: Namhyung Kim 
Link: https://lore.kernel.org/r/Z_TYux5fUg2pW-pF@gmail.com
Signed-off-by: Arnaldo Carvalho de Melo 
Signed-off-by: Sasha Levin

perf build: Warn when libdebuginfod devel files are not available

2025-06-27T10:07:14+00:00

[ Upstream commit 4fce4b91fd1aabb326c46e237eb4b19ab72598f8 ]

While working on 'perf version --build-options' I noticed that:

  $ perf version --build-options
  perf version 6.15.rc1.g312a07a00d31
                   aio: [ on  ]  # HAVE_AIO_SUPPORT
                   bpf: [ on  ]  # HAVE_LIBBPF_SUPPORT
         bpf_skeletons: [ on  ]  # HAVE_BPF_SKEL
            debuginfod: [ OFF ]  # HAVE_DEBUGINFOD_SUPPORT
  

And looking at tools/perf/Makefile.config I also noticed that it is not
opt-in, meaning we will attempt to build with it in all normal cases.

So add the usual warning at build time to let the user know that
something recommended is missing, now we see:

  Makefile.config:563: No elfutils/debuginfod.h found, no debuginfo server support, please install elfutils-debuginfod-client-devel or equivalent

And after following the recommendation:

  $ perf check feature debuginfod
            debuginfod: [ on  ]  # HAVE_DEBUGINFOD_SUPPORT
  $ ldd ~/bin/perf | grep debuginfo
	libdebuginfod.so.1 => /lib64/libdebuginfod.so.1 (0x00007fee5cf5f000)
  $

With this feature on several perf tools will fetch what is needed and
not require all the contents of the debuginfo packages, for instance:

  # rpm -qa | grep kernel-debuginfo
  # pahole --running_kernel_vmlinux
  pahole: couldn't find a vmlinux that matches the running kernel
  HINT: Maybe you're inside a container or missing a debuginfo package?
  #
  # perf trace -e open* perf probe --vars icmp_rcv
      0.000 ( 0.005 ms): perf/97391 openat(dfd: CWD, filename: "/etc/ld.so.cache", flags: RDONLY|CLOEXEC) = 3
      0.014 ( 0.004 ms): perf/97391 openat(dfd: CWD, filename: "/lib64/libm.so.6", flags: RDONLY|CLOEXEC) = 3
  
  32130.100 ( 0.008 ms): perf/97391 openat(dfd: CWD, filename: "/root/.cache/debuginfod_client/aa3c82b4a13f9c0e0301bebb20fe958c4db6f362/debuginfo") = 3
  
  Available variables at icmp_rcv
        @
                struct sk_buff* skb
  
  #
  # pahole --running_kernel_vmlinux
  /root/.cache/debuginfod_client/aa3c82b4a13f9c0e0301bebb20fe958c4db6f362/debuginfo
  # file /root/.cache/debuginfod_client/aa3c82b4a13f9c0e0301bebb20fe958c4db6f362/debuginfo
  /root/.cache/debuginfod_client/aa3c82b4a13f9c0e0301bebb20fe958c4db6f362/debuginfo: ELF 64-bit LSB executable, x86-64, version 1 (SYSV), statically linked, BuildID[sha1]=aa3c82b4a13f9c0e0301bebb20fe958c4db6f362, with debug_info, not stripped
  # ls -la /root/.cache/debuginfod_client/aa3c82b4a13f9c0e0301bebb20fe958c4db6f362/debuginfo
  -r--------. 1 root root 475401512 Mar 27 21:00 /root/.cache/debuginfod_client/aa3c82b4a13f9c0e0301bebb20fe958c4db6f362/debuginfo
  #

Then, cached:

  # perf stat --null perf probe --vars icmp_rcv
  Available variables at icmp_rcv
        @
                struct sk_buff* skb

   Performance counter stats for 'perf probe --vars icmp_rcv':

       0.671389041 seconds time elapsed

       0.519176000 seconds user
       0.150860000 seconds sys

Fixes: c7a14fdcb3fa7736 ("perf build-ids: Fall back to debuginfod query if debuginfo not found")
Tested-by: Ingo Molnar 
Cc: Adrian Hunter 
Cc: Dmitriy Vyukov 
Cc: Howard Chu 
Cc: Ian Rogers 
Cc: Jiri Olsa 
Cc: Kan Liang 
Cc: Namhyung Kim 
Cc: Peter Zijlstra 
Cc: Frank Ch. Eigler 
Link: https://lore.kernel.org/r/Z_dkNDj9EPFwPqq1@gmail.com
[ Folded patch from Ingo to have the debian/ubuntu devel package added build warning message ]
Signed-off-by: Arnaldo Carvalho de Melo 
Signed-off-by: Sasha Levin

perf tools: annotate asm_pure_loop.S

2025-04-10T12:33:37+00:00

[ Upstream commit 9a352a90e88a041f4b26d359493e12a7f5ae1a6a ]

Annotate so it is built with non-executable stack.

Fixes: 8b97519711c3 ("perf test: Add asm pureloop test tool")
Signed-off-by: Marcus Meissner 
Reviewed-by: Leo Yan 
Link: https://lore.kernel.org/r/20250323085410.23751-1-meissner@suse.de
Signed-off-by: Namhyung Kim 
Signed-off-by: Sasha Levin

perf python: Check if there is space to copy all the event

2025-04-10T12:33:37+00:00

[ Upstream commit 89aaeaf84231157288035b366cb6300c1c6cac64 ]

The pyrf_event__new() method copies the event obtained from the perf
ring buffer to a structure that will then be turned into a python object
for further consumption, so it copies perf_event.header.size bytes to
its 'event' member:

  $ pahole -C pyrf_event /tmp/build/perf-tools-next/python/perf.cpython-312-x86_64-linux-gnu.so
  struct pyrf_event {
  	PyObject                   ob_base;              /*     0    16 */
  	struct evsel *             evsel;                /*    16     8 */
  	struct perf_sample         sample;               /*    24   312 */

  	/* XXX last struct has 7 bytes of padding, 2 holes */

  	/* --- cacheline 5 boundary (320 bytes) was 16 bytes ago --- */
  	union perf_event           event;                /*   336  4168 */

  	/* size: 4504, cachelines: 71, members: 4 */
  	/* member types with holes: 1, total: 2 */
  	/* paddings: 1, sum paddings: 7 */
  	/* last cacheline: 24 bytes */
  };

  $

It was doing so without checking if the event just obtained has more
than that space, fix it.

This isn't a proper, final solution, as we need to support larger
events, but for the time being we at least bounds check and document it.

Fixes: 877108e42b1b9ba6 ("perf tools: Initial python binding")
Signed-off-by: Arnaldo Carvalho de Melo 
Reviewed-by: Ian Rogers 
Link: https://lore.kernel.org/r/20250312203141.285263-7-acme@kernel.org
Signed-off-by: Namhyung Kim 
Signed-off-by: Sasha Levin

perf python: Don't keep a raw_data pointer to consumed ring buffer space

2025-04-10T12:33:37+00:00

[ Upstream commit f3fed3ae34d606819d87a63d970cc3092a5be7ab ]

When processing tracepoints the perf python binding was parsing the
event before calling perf_mmap__consume(&md->core) in
pyrf_evlist__read_on_cpu().

But part of this event parsing was to set the perf_sample->raw_data
pointer to the payload of the event, which then could be overwritten by
other event before tracepoint fields were asked for via event.prev_comm
in a python program, for instance.

This also happened with other fields, but strings were were problems
were surfacing, as there is UTF-8 validation for the potentially garbled
data.

This ended up showing up as (with some added debugging messages):

  ( field 'prev_comm' ret=0x7f7c31f65110, raw_size=68 )  ( field 'prev_pid' ret=0x7f7c23b1bed0, raw_size=68 )  ( field 'prev_prio' ret=0x7f7c239c0030, raw_size=68 )  ( field 'prev_state' ret=0x7f7c239c0250, raw_size=68 ) time 14771421785867 prev_comm= prev_pid=1919907691 prev_prio=796026219 prev_state=0x303a32313175 ==>
  ( XXX '��' len=16, raw_size=68)  ( field 'next_comm' ret=(nil), raw_size=68 ) Traceback (most recent call last):
   File "/home/acme/git/perf-tools-next/tools/perf/python/tracepoint.py", line 51, in 
     main()
   File "/home/acme/git/perf-tools-next/tools/perf/python/tracepoint.py", line 46, in main
     event.next_comm,
     ^^^^^^^^^^^^^^^
  AttributeError: 'perf.sample_event' object has no attribute 'next_comm'

When event.next_comm was asked for, the PyUnicode_FromString() python
API would fail and that tracepoint field wouldn't be available, stopping
the tools/perf/python/tracepoint.py test tool.

But, since we already do a copy of the whole event in pyrf_event__new,
just use it and while at it remove what was done in in e8968e654191390a
("perf python: Fix pyrf_evlist__read_on_cpu event consuming") because we
don't really need to wait for parsing the sample before declaring the
event as consumed.

This copy is questionable as is now, as it limits the maximum event +
sample_type and tracepoint payload to sizeof(union perf_event), this all
has been "working" because 'struct perf_event_mmap2', the largest entry
in 'union perf_event' is:

  $ pahole -C perf_event ~/bin/perf | grep mmap2
	struct perf_record_mmap2   mmap2;              /*     0  4168 */
  $

Fixes: bae57e3825a3dded ("perf python: Add support to resolve tracepoint fields")
Signed-off-by: Arnaldo Carvalho de Melo 
Reviewed-by: Ian Rogers 
Link: https://lore.kernel.org/r/20250312203141.285263-6-acme@kernel.org
Signed-off-by: Namhyung Kim 
Signed-off-by: Sasha Levin

perf python: Decrement the refcount of just created event on failure

2025-04-10T12:33:37+00:00

[ Upstream commit 3de5a2bf5b4847f7a59a184568f969f8fe05d57f ]

To avoid a leak if we have the python object but then something happens
and we need to return the operation, decrement the offset of the newly
created object.

Fixes: 377f698db12150a1 ("perf python: Add struct evsel into struct pyrf_event")
Signed-off-by: Arnaldo Carvalho de Melo 
Reviewed-by: Ian Rogers 
Link: https://lore.kernel.org/r/20250312203141.285263-5-acme@kernel.org
Signed-off-by: Namhyung Kim 
Signed-off-by: Sasha Levin