linux-stable.git/tools/perf/tests, branch linux-6.11.y

perf test attr: Add back missing topdown events

2024-12-05T12:53:46+00:00

[ Upstream commit 6bff76af9635411214ca44ea38fc2781e78064b6 ]

With the patch 0b6c5371c03c "Add missing topdown metrics events" eight
topdown metric events with numbers ranging from 0x8000 to 0x8700 were
added to the test since they were added as 'perf stat' default events.
Later the patch 951efb9976ce "Update no event/metric expectations" kept
only 4 of those events(0x8000-0x8300).

Currently, the topdown events with numbers 0x8400 to 0x8700 are missing
from the list of expected events resulting in a failure. Add back the
missing topdown events.

Fixes: 951efb9976ce ("perf test attr: Update no event/metric expectations")
Signed-off-by: Veronika Molnarova 
Tested-by: Ian Rogers 
Cc: mpetlan@redhat.com
Link: https://lore.kernel.org/r/20240311081611.7835-1-vmolnaro@redhat.com
Signed-off-by: Namhyung Kim 
Signed-off-by: Sasha Levin

perf test: Add test for Intel TPEBS counting mode

2024-12-05T12:53:45+00:00

[ Upstream commit b2738fda24543777a623a7d1cc2a9985ab81b448 ]

Intel TPEBS sampling mode is supported through perf record. The counting mode
code uses perf record to capture retire_latency value and use it in metric
calculation. This test checks the counting mode code on Intel platforms.

Committer testing:

  root@x1:~# perf test tpebs
  123: test Intel TPEBS counting mode                                  : Ok
  root@x1:~# set -o vi
  root@x1:~# perf test tpebs
  123: test Intel TPEBS counting mode                                  : Ok
  root@x1:~# perf test -v tpebs
  123: test Intel TPEBS counting mode                                  : Ok
  root@x1:~# perf test -vvv tpebs
  123: test Intel TPEBS counting mode:
  --- start ---
  test child forked, pid 16603
  Testing without --record-tpebs
  Testing with --record-tpebs
  ---- end(0) ----
  123: test Intel TPEBS counting mode                                  : Ok
  root@x1:~#

Reviewed-by: Namhyung Kim 
Signed-off-by: Weilin Wang 
Acked-by: Ian Rogers 
Tested-by: Arnaldo Carvalho de Melo 
Cc: Adrian Hunter 
Cc: Alexander Shishkin 
Cc: Caleb Biggers 
Cc: Ingo Molnar 
Cc: Jiri Olsa 
Cc: Kan Liang 
Cc: Perry Taylor 
Cc: Peter Zijlstra 
Cc: Samantha Alt 
Link: https://lore.kernel.org/r/20240720062102.444578-9-weilin.wang@intel.com
Signed-off-by: Arnaldo Carvalho de Melo 
Stable-dep-of: 057f8bfc6f70 ("perf stat: Uniquify event name improvements")
Signed-off-by: Sasha Levin

perf test pmu: Set uninitialized PMU alias to null

2024-09-02T18:57:46+00:00

Commit 3e0bf9fde2984469 ("perf pmu: Restore full PMU name wildcard
support") adds a test case "PMU cmdline match" that covers PMU name
wildcard support provided by function perf_pmu__match(). The test works
with a wide range of supported combinations of PMU name matching but
omits the case that if the perf_pmu__match() cannot match the PMU name
to the wildcard, it tries to match its alias. However, this variable is
not set up, causing the test case to fail when run with subprocesses or
to segfault if run as a single process.

  ./perf test -vv 9
    9: Sysfs PMU tests                                                 :
    9.1: Parsing with PMU format directory                             : Ok
    9.2: Parsing with PMU event                                        : Ok
    9.3: PMU event names                                               : Ok
    9.4: PMU name combining                                            : Ok
    9.5: PMU name comparison                                           : Ok
    9.6: PMU cmdline match                                             : FAILED!

  ./perf test -F 9
    9.1: Parsing with PMU format directory                             : Ok
    9.2: Parsing with PMU event                                        : Ok
    9.3: PMU event names                                               : Ok
    9.4: PMU name combining                                            : Ok
    9.5: PMU name comparison                                           : Ok
  Segmentation fault (core dumped)

Initialize the PMU alias to null for all tests of perf_pmu__match()
as this functionality is not being tested and the alias matching works
exactly the same as the matching of the PMU name.

  ./perf test -F 9
    9.1: Parsing with PMU format directory                             : Ok
    9.2: Parsing with PMU event                                        : Ok
    9.3: PMU event names                                               : Ok
    9.4: PMU name combining                                            : Ok
    9.5: PMU name comparison                                           : Ok
    9.6: PMU cmdline match                                             : Ok

Fixes: 3e0bf9fde2984469 ("perf pmu: Restore full PMU name wildcard support")
Signed-off-by: Veronika Molnarova 
Cc: james.clark@arm.com
Cc: mpetlan@redhat.com
Cc: rstoyano@redhat.com
Link: https://lore.kernel.org/r/20240808103749.9356-1-vmolnaro@redhat.com
Signed-off-by: Namhyung Kim

Merge tag 'kbuild-v6.11' of git://git.kernel.org/pub/scm/linux/kernel/git/masahiroy/linux-kbuild

2024-07-23T21:32:21+00:00

Pull Kbuild updates from Masahiro Yamada:

 - Remove tristate choice support from Kconfig

 - Stop using the PROVIDE() directive in the linker script

 - Reduce the number of links for the combination of CONFIG_KALLSYMS and
   CONFIG_DEBUG_INFO_BTF

 - Enable the warning for symbol reference to .exit.* sections by
   default

 - Fix warnings in RPM package builds

 - Improve scripts/make_fit.py to generate a FIT image with separate
   base DTB and overlays

 - Improve choice value calculation in Kconfig

 - Fix conditional prompt behavior in choice in Kconfig

 - Remove support for the uncommon EMAIL environment variable in Debian
   package builds

 - Remove support for the uncommon "name " form for the DEBEMAIL
   environment variable

 - Raise the minimum supported GNU Make version to 4.0

 - Remove stale code for the absolute kallsyms

 - Move header files commonly used for host programs to scripts/include/

 - Introduce the pacman-pkg target to generate a pacman package used in
   Arch Linux

 - Clean up Kconfig

* tag 'kbuild-v6.11' of git://git.kernel.org/pub/scm/linux/kernel/git/masahiroy/linux-kbuild: (65 commits)
  kbuild: doc: gcc to CC change
  kallsyms: change sym_entry::percpu_absolute to bool type
  kallsyms: unify seq and start_pos fields of struct sym_entry
  kallsyms: add more original symbol type/name in comment lines
  kallsyms: use \t instead of a tab in printf()
  kallsyms: avoid repeated calculation of array size for markers
  kbuild: add script and target to generate pacman package
  modpost: use generic macros for hash table implementation
  kbuild: move some helper headers from scripts/kconfig/ to scripts/include/
  Makefile: add comment to discourage tools/* addition for kernel builds
  kbuild: clean up scripts/remove-stale-files
  kconfig: recursive checks drop file/lineno
  kbuild: rpm-pkg: introduce a simple changelog section for kernel.spec
  kallsyms: get rid of code for absolute kallsyms
  kbuild: Create INSTALL_PATH directory if it does not exist
  kbuild: Abort make on install failures
  kconfig: remove 'e1' and 'e2' macros from expression deduplication
  kconfig: remove SYMBOL_CHOICEVAL flag
  kconfig: add const qualifiers to several function arguments
  kconfig: call expr_eliminate_yn() at least once in expr_eliminate_dups()
  ...

kallsyms: get rid of code for absolute kallsyms

2024-07-20T07:33:21+00:00

Commit cf8e8658100d ("arch: Remove Itanium (IA-64) architecture")
removed the last use of the absolute kallsyms.

Signed-off-by: Jann Horn 
Acked-by: Arnd Bergmann 
Link: https://lore.kernel.org/all/20240221202655.2423854-1-jannh@google.com/
[masahiroy@kernel.org: rebase the code and reword the commit description]
Signed-off-by: Masahiro Yamada

perf pmu: Restore full PMU name wildcard support

2024-06-28T03:28:01+00:00

Commit b2b9d3a3f021 ("perf pmu: Support wildcards on pmu name in dynamic
pmu events") gives the following example for wildcarding a subset of
PMUs:

  E.g., in a system with the following dynamic pmus:

        mypmu_0
        mypmu_1
        mypmu_2
        mypmu_4

  perf stat -e mypmu_[01]//

Since commit f91fa2ae6360 ("perf pmu: Refactor perf_pmu__match()"), only
"*" has been supported, removing the ability to subset PMUs, even though
parse-events.l still supports ? and [] characters.

Fix it by using fnmatch() when any glob character is detected and add a
test which covers that and other scenarios of
perf_pmu__match_ignoring_suffix().

Fixes: f91fa2ae6360 ("perf pmu: Refactor perf_pmu__match()")
Signed-off-by: James Clark 
Reviewed-by: Ian Rogers 
Cc: robin.murphy@arm.com
Signed-off-by: Namhyung Kim 
Link: https://lore.kernel.org/r/20240626145448.896746-2-james.clark@arm.com

perf test stat_bpf_counter.sh: Stabilize the test results

2024-06-26T18:10:58+00:00

The test has been failing for some time when two separate runs of
perf benchmarks are recorded for cycles events and their counts are
compared, while once the recording was done with option --bpf-counters
and once without it. It is expected that the count of the samples
should be within a certain range, firstly the difference was set to be
within 10%, which was then later raised to 20%. However, the test case
keeps failing on certain architectures as recording the provided
benchmark can produce completely different counts based on the
current load of the system.

Sampling two separate runs on intel-eaglestream-spr-13 of "perf stat
--no-big-num -e cycles -- perf bench sched messaging -g 1 -l 100 -t":

 Performance counter stats for 'perf bench sched messaging -g 1 -l 100 -t':

         396782898      cycles

       0.010051983 seconds time elapsed

       0.008664000 seconds user
       0.097058000 seconds sys

 Performance counter stats for 'perf bench sched messaging -g 1 -l 100 -t':

        1431133032      cycles

       0.021803714 seconds time elapsed

       0.023377000 seconds user
       0.349918000 seconds sys

, which is ranging from 400mil to 1400mil samples.

Instead of recording the cycles use instructions event, which provides
more stable values. At the same time change the tested workload to one
of the provided testing workloads by perf that is not based on a
scheduler, which can provide another dependency on the current load.

Sampling instructions event with the new workload provide much more
stable results on intel-eaglestream-spr-13 of "perf stat --no-big-num
-e instructions -- perf test -w brstack":

 Performance counter stats for 'perf test -w brstack':

          64584494      instructions

       0.009173945 seconds time elapsed

       0.007262000 seconds user
       0.002071000 seconds sys

 Performance counter stats for 'perf test -w brstack':

          64672669      instructions

       0.008888135 seconds time elapsed

       0.005018000 seconds user
       0.004018000 seconds sys

Signed-off-by: Veronika Molnarova 
Acked-by: Namhyung Kim 
Cc: mpetlan@redhat.com
Signed-off-by: Namhyung Kim 
Link: https://lore.kernel.org/r/20240625092001.10909-1-vmolnaro@redhat.com

perf test: Make tests its own library

2024-06-26T18:07:11+00:00

Make the tests code its own library. This is done to avoid compiling
code twice, once for the perf tool and once for the perf python
module.

Signed-off-by: Ian Rogers 
Reviewed-by: James Clark 
Cc: Suzuki K Poulose 
Cc: Kees Cook 
Cc: Palmer Dabbelt 
Cc: Albert Ou 
Cc: Nick Terrell 
Cc: Gary Guo 
Cc: Alex Gaynor 
Cc: Boqun Feng 
Cc: Wedson Almeida Filho 
Cc: Ze Gao 
Cc: Alice Ryhl 
Cc: Andrei Vagin 
Cc: Yicong Yang 
Cc: Jonathan Cameron 
Cc: Guo Ren 
Cc: Miguel Ojeda 
Cc: Will Deacon 
Cc: Mike Leach 
Cc: Leo Yan 
Cc: Oliver Upton 
Cc: John Garry 
Cc: Benno Lossin 
Cc: Björn Roy Baron 
Cc: Andreas Hindborg 
Cc: Paul Walmsley 
Signed-off-by: Namhyung Kim 
Link: https://lore.kernel.org/r/20240625214117.953777-5-irogers@google.com

perf test: Check output of the probe ... --funcs command

2024-06-25T18:06:19+00:00

Test "perf probe of function from different CU" only checks if the perf
command has failed and doesn't test the --funcs output. In the issue
reported in the previous commit, the garbage output of the --funcs
command was being ignored by the test when it could have been caught.

The script first makes use of --funcs option with the perf probe command
to check if the function "foo" exists in the testfile before adding a
probe to it in the next command. The output of probe...--funcs command
is redirected to stdout, therefore, add '| grep "foo"' to validate the
result.

Signed-off-by: Chaitanya S Prakash 
Reviewed-by: Masami Hiramatsu (Google) 
Cc: anshuman.khandual@arm.com
Cc: james.clark@arm.com
Signed-off-by: Namhyung Kim 
Link: https://lore.kernel.org/r/20240601125946.1741414-11-ChaitanyaS.Prakash@arm.com

perf test: Make test_arm_callgraph_fp.sh more robust

2024-06-24T21:42:59+00:00

The 2 second sleep can cause the test to fail on very slow network file
systems because Perf ends up being killed before it finishes starting
up.

Fix it by making the leafloop workload end after a fixed time like the
other workloads so there is no need to kill it after 2 seconds.

Also remove the 1 second start sampling delay because it is similarly
fragile. Instead, search through all samples for a matching one, rather
than just checking the first sample and hoping it's in the right place.

Fixes: cd6382d82752 ("perf test arm64: Test unwinding using fame-pointer (fp) mode")
Signed-off-by: James Clark 
Acked-by: Namhyung Kim 
Cc: German Gomez 
Cc: Spoorthy S 
Cc: Kajol Jain 
Signed-off-by: Namhyung Kim 
Link: https://lore.kernel.org/r/20240612140316.3006660-1-james.clark@arm.com