linux.git/tools/perf/tests/shell, branch v6.0-rc2

perf offcpu: Update offcpu test for child process

2022-08-11T20:57:45+00:00

Record off-cpu data with perf bench sched messaging workload and count
the number of offcpu-time events.  Also update the test script not to
run next tests if failed already and revise the error messages.

  $ sudo ./perf test offcpu -v
   88: perf record offcpu profiling tests                              :
  --- start ---
  test child forked, pid 344780
  Checking off-cpu privilege
  Basic off-cpu test
  Basic off-cpu test [Success]
  Child task off-cpu test
  Child task off-cpu test [Success]
  test child finished with 0
  ---- end ----
  perf record offcpu profiling tests: Ok

Signed-off-by: Namhyung Kim 
Cc: Blake Jones 
Cc: Hao Luo 
Cc: Ian Rogers 
Cc: Ingo Molnar 
Cc: Jiri Olsa 
Cc: Milian Wolff 
Cc: Peter Zijlstra 
Cc: Song Liu 
Cc: bpf@vger.kernel.org
Link: https://lore.kernel.org/r/20220811185456.194721-5-namhyung@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo

perf test: JSON format checking

2022-08-10T13:44:01+00:00

Add field checking tests for perf stat JSON output.

Sanity checks the expected number of fields are present, that the
expected keys are present and they have the correct values.

Committer notes:

Had to fix this:

  -               $(INSTALL) tests/shell/lib/*.sh '$(DESTDIR_SQ)$(perfexec_instdir_SQ)/tests/shell/lib' \
  +               $(INSTALL) tests/shell/lib/*.sh '$(DESTDIR_SQ)$(perfexec_instdir_SQ)/tests/shell/lib'; \

Committer testing:

  [root@quaco ~]# perf test json
   90: perf stat JSON output linter                                    : Ok
  [root@quaco ~]# set -o vi
  [root@quaco ~]# perf test -v json
   90: perf stat JSON output linter                                    :
  --- start ---
  test child forked, pid 560794
  Checking json output: no args [Success]
  Checking json output: system wide [Success]
  Checking json output: system wide Checking json output: system wide no aggregation [Success]
  Checking json output: interval [Success]
  Checking json output: event [Success]
  Checking json output: per core [Success]
  Checking json output: per thread [Success]
  Checking json output: per die [Success]
  Checking json output: per node [Success]
  Checking json output: per socket [Success]
  test child finished with 0
  ---- end ----
  perf stat JSON output linter: Ok
  [root@quaco ~]#

Signed-off-by: Claire Jensen 
Acked-by: Namhyung Kim 
Tested-by: Arnaldo Carvalho de Melo 
Cc: Alexander Shishkin 
Cc: Alyssa Ross 
Cc: Claire Jensen 
Cc: Florian Fischer 
Cc: Ingo Molnar 
Cc: James Clark 
Cc: Jiri Olsa 
Cc: Kan Liang 
Cc: Like Xu 
Cc: Mark Rutland 
Cc: Peter Zijlstra 
Cc: Sandipan Das 
Cc: Stephane Eranian 
Cc: Xing Zhengjun 
Link: https://lore.kernel.org/r/20220805200105.2020995-3-irogers@google.com
Signed-off-by: Ian Rogers 
Signed-off-by: Arnaldo Carvalho de Melo

perf test: Add ARM SPE system wide test

2022-08-01T17:46:30+00:00

In the past it had a problem not setting the pid/tid on the sample
correctly when system-wide mode is used.  Although it's fixed now it'd
be nice if we have a test case for it.

Reviewed-by: German Gomez 
Signed-off-by: Namhyung Kim 
Tested-by: Leo Yan 
Cc: German Gomez 
Cc: Ian Rogers 
Cc: Ingo Molnar 
Cc: James Clark 
Cc: Jiri Olsa 
Cc: John Garry 
Cc: Leo Yan 
Cc: Mike Leach 
Cc: Peter Zijlstra 
Cc: Will Deacon 
Link: https://lore.kernel.org/r/20220701230932.1000495-1-namhyung@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo

perf test: Fix test case 95 ("Check branch stack sampling") on s390 and use same event

2022-07-29T13:31:06+00:00

On linux-next tree 'perf test 95' ("Check branch stack sampling") was
added recently.

s390 does not support branch sampling at all and the test case fails
despite for checking branch support before hand.

The check for support of branching uses the software event named "dummy",
as seen in the line:

  perf record -b -o- -e dummy -B true > /dev/null 2>&1 || exit 2

However when the branch recording is actually done, a different event is
used, as seen in the line:

  perf record -o $TMPDIR/... --branch-filter any,save_type,u -- ...

The event is omitted and for "perf record" the default event is cycles,
which is not supported by s390 and this fails when executed on s390:

  # perf record --branch-filter any,save_type,u -- /tmp/__perf_test.program.iDSmQ/a.out
  Error:
  cycles: PMU Hardware or event type doesn't support branch stack sampling.
  #

Therefore fix this and use the same event cycles for testing support
and actually running the test.

Output before:

  # ./perf test -Fv 95
  95: Check branch stack sampling                                     :
  --- start ---
  Testing user branch stack sampling
  ---- end ----
  Check branch stack sampling: FAILED!
  #

Output after:

  # ./perf test -Fv 95
  95: Check branch stack sampling                                     :
  --- start ---
  ---- end ----
  Check branch stack sampling: Skip
  #

Fixes: b55878c90ab92a24 ("perf test: Add test for branch stack sampling")
Reviewed-by: James Clark 
Signed-off-by: Thomas Richter 
Acked-by: German Gomez 
Cc: German Gomez 
Cc: Heiko Carstens 
Cc: Sumanth Korikkar 
Cc: Sven Schnelle 
Cc: Vasily Gorbik 
Link: https://lore.kernel.org/r/20220727141439.712582-1-tmricht@linux.ibm.com
Signed-off-by: Arnaldo Carvalho de Melo

perf test: Fix test case 83 ('perf stat CSV output linter') on s390

2022-07-20T16:20:33+00:00

Perf test case 83: perf stat CSV output linter might fail
on s390.
The reason for this is the output of the command

 ./perf stat -x, -A -a --no-merge true

which depends on a .config file setting. When CONFIG_SCHED_TOPOLOGY
is set, the output of above perf command is

   CPU0,1.50,msec,cpu-clock,1502781,100.00,1.052,CPUs utilized

When CONFIG_SCHED_TOPOLOGY is *NOT* set the output of above perf
command is

   0.95,msec,cpu-clock,949800,100.00,1.060,CPUs utilized

Fix the test case to accept both output formats.

Output before:
 # perf test 83
 83: perf stat CSV output linter       : FAILED!
 #

Output after:
 # ./perf test 83
 83: perf stat CSV output linter       : Ok
 #

Fixes: ec906102e5b7d339 ("perf test: Fix "perf stat CSV output linter" test on s390")
Signed-off-by: Thomas Richter 
Acked-by: Ian Rogers 
Acked-by: Sumanth Korikkar 
Cc: Heiko Carstens 
Cc: Sven Schnelle 
Cc: Vasily Gorbik 
Link: https://lore.kernel.org/r/20220720123419.220953-1-tmricht@linux.ibm.com
Signed-off-by: Arnaldo Carvalho de Melo

perf test: Add test for branch stack sampling

2022-07-12T11:55:36+00:00

Add a self test for branch stack sampling, to check that we get the
expected branch types, and filters behave as expected.

Suggested-by: Anshuman Khandual 
Signed-off-by: German Gomez 
Acked-by: Namhyung Kim 
Cc: Alexander Shishkin 
Cc: Ian Rogers 
Cc: Mark Rutland 
Link: https://lore.kernel.org/r/20220705150511.473919-2-german.gomez@arm.com
Tested-by: Jiri Olsa 
Signed-off-by: Arnaldo Carvalho de Melo

perf test: Make all metrics test more tolerant

2022-07-12T11:54:47+00:00

Metric names are truncated so don't try to match all of one.

Allow AMX metrics to skip as floating point ones do.

Metrics for optane memory can also skip rather than fail.

Add a system wide check for uncore metrics.

Restructure code to avoid extensive nesting.

Some impetus for this in:

https://lore.kernel.org/lkml/d32376b5-5538-ff00-6620-e74ad4b4abf2@huawei.com/

Suggested-by: John Garry 
Signed-off-by: Ian Rogers 
Cc: Alexander Shishkin 
Cc: Andi Kleen 
Cc: Ingo Molnar 
Cc: Jiri Olsa 
Cc: John Garry 
Cc: Kan Liang 
Cc: Mark Rutland 
Cc: Namhyung Kim 
Cc: Peter Zijlstra 
Cc: Stephane Eranian 
Link: https://lore.kernel.org/r/20220707153449.202409-2-irogers@google.com
Signed-off-by: Arnaldo Carvalho de Melo

perf test: Record only user callchains on the "Check Arm64 callgraphs are complete in fp mode" test

2022-06-19T13:41:43+00:00

The testcase 'Check Arm64 callgraphs are complete in fp mode' wants to
see the following output:

    610 leaf
    62f parent
    648 main

However, without excluding kernel callchains, the output might look like:

	ffffc2ff40ef1b5c arch_local_irq_enable
	ffffc2ff419d032c __schedule
	ffffc2ff419d06c0 schedule
	ffffc2ff40e4da30 do_notify_resume
	ffffc2ff40e421b0 work_pending
	             610 leaf
	             62f parent
	             648 main

Adding '--user-callchains' leaves only the wanted symbols in the chain.

Fixes: cd6382d82752737e ("perf test arm64: Test unwinding using fame-pointer (fp) mode")
Suggested-by: German Gomez 
Reviewed-by: German Gomez 
Reviewed-by: Leo Yan 
Signed-off-by: Michael Petlan 
Cc: German Gomez 
Cc: Jiri Olsa 
Link: https://lore.kernel.org/r/20220614105207.26223-1-mpetlan@redhat.com
Signed-off-by: Arnaldo Carvalho de Melo

perf test: Fix "perf stat CSV output linter" test on s390

2022-06-19T13:41:43+00:00

perf test -F 83 ("perf stat CSV output linter") fails on s390.

Reason is the wrong number of fields for certain CPU core/die/socket
related output.

On x84_64 the output of command:

  # ./perf stat -x, -A -a --no-merge true
  CPU0,1.50,msec,cpu-clock,1502781,100.00,1.052,CPUs utilized
  CPU1,1.48,msec,cpu-clock,1476113,100.00,1.034,CPUs utilized
  ...

results in 8 fields with 7 comma separators.

On s390 the output of command:

  #  ./perf stat -x, -A -a --no-merge -- true
  0.95,msec,cpu-clock,949800,100.00,1.060,CPUs utilized
  ...

results in 7 fields with 6 comma separators. Therefore this tests
fails on s390. Similar issues exist for per-die and per-socket output
which is not supported on s390.

I have rewritten the python program to count commas in each output line
into a bash function to achieve the same result. I hope this makes it a
bit easier.

Output before:

  # ./perf test -F 83
  83: perf stat CSV output linter  :
  Checking CSV output: no args [Success]
  Checking CSV output: system wide [Success]
  Checking CSV output: system wide Checking CSV output: \
	  system wide no aggregation 6.92,msec,cpu-clock,\
	  6918131,100.00,6.972,CPUs utilized
  ...
  RuntimeError: wrong number of fields. expected 7 in \
	  6.92,msec,cpu-clock,6918131,100.00,6.972,CPUs utilized

  FAILED!
  #

Output after:

  # ./perf test -F 83
  83: perf stat CSV output linter             :
  Checking CSV output: no args [Success]
  Checking CSV output: system wide [Success]
  Checking CSV output: system wide Checking CSV output:\
	  system wide no aggregation [Success]
  Checking CSV output: interval [Success]
  Checking CSV output: event [Success]
  Checking CSV output: per core [Success]
  Checking CSV output: per thread [Success]
  Checking CSV output: per die [Success]
  Checking CSV output: per node [Success]
  Checking CSV output: per socket [Success]
  Ok
  #

Committer notes:

Continues to work on x86_64

  $ perf test lint
   89: perf stat CSV output linter                                     : Ok
  $ perf test -v lint
  Couldn't bump rlimit(MEMLOCK), failures may take place when creating BPF maps, etc
   89: perf stat CSV output linter                                     :
  --- start ---
  test child forked, pid 53133
  Checking CSV output: no args [Success]
  Checking CSV output: system wide [Skip] paranoid and not root
  Checking CSV output: system wide [Skip] paranoid and not root
  Checking CSV output: interval [Success]
  Checking CSV output: event [Success]
  Checking CSV output: per core [Skip] paranoid and not root
  Checking CSV output: per thread [Skip] paranoid and not root
  Checking CSV output: per die [Skip] paranoid and not root
  Checking CSV output: per node [Skip] paranoid and not root
  Checking CSV output: per socket [Skip] paranoid and not root
  test child finished with 0
  ---- end ----
  perf stat CSV output linter: Ok
  $

Signed-off-by: Thomas Richter 
Acked-by: Ian Rogers 
Tested-by: Arnaldo Carvalho de Melo 
Cc: Claire Jensen 
Cc: Heiko Carstens 
Cc: Sumanth Korikkar 
Cc: Sven Schnelle 
Cc: Vasily Gorbik 
Cc: linux390-list@tuxmaker.boeblingen.de.ibm.com
Link: https://lore.kernel.org/r/20220603113034.2009728-1-tmricht@linux.ibm.com
Signed-off-by: Arnaldo Carvalho de Melo

perf test arm-spe: Check if perf-record hangs when recording workload with forks

2022-06-03T19:18:35+00:00

Add shell test to check if perf-record hangs when recording an arm_spe
event with forks.

The test FAILS if the Kernel is not patched with Commit 961c391217 ("perf:
Always wake the parent event").

Unpatched Kernel:

  $ perf test -v 90
  90: Check Arm SPE doesn't hang when there are forks
  --- start ---
  test child forked, pid 14232
  Recording workload with fork
  Log lines = 90 /tmp/__perf_test.stderr.0Nu0U
  Log lines after 1 second = 90 /tmp/__perf_test.stderr.0Nu0U
  SPE hang test: FAIL
  test child finished with -1
  ---- end ----
  Check Arm SPE trace data in workload with forks: FAILED!

Patched Kernel:

  $ perf test -v 90
  90: Check Arm SPE doesn't hang when there are forks
  --- start ---
  test child forked, pid 2930
  Compiling test program...
  Recording workload...
  Log lines = 478 /tmp/__perf_test.log.026AI
  Log lines after 1 second = 557 /tmp/__perf_test.log.026AI
  SPE hang test: PASS
  Cleaning up files...
  test child finished with 0
  ---- end ----
  Check Arm SPE trace data in workload with forks: Ok

Reviewed-by: James Clark 
Reviewed-by: Leo Yan 
Signed-off-by: German Gomez 
Cc: Alexander Shishkin 
Cc: Jiri Olsa 
Cc: Mark Rutland 
Cc: Namhyung Kim 
Link: https://lore.kernel.org/r/20220228165655.3920-1-german.gomez@arm.com
Signed-off-by: Arnaldo Carvalho de Melo