[PATCH 0/2] perf test attr: Fix and re-enable the test

[PATCH 0/2] perf test attr: Fix and re-enable the test

[Michael Petlan]

Hello all,

I have noticed that a testcase formerly known as "Setup struct perf_event_attr",
later renamed to "Perf attribute expectations test" has been disabled by
accident during the move.

That also probably led to forgetting to update the struct size, which is now
fixed by the second patch.

Michael


Michael Petlan (2):
  perf test attr: Propagate the return value from the test to the
    wrapper
  perf test attr: Fix wrong size expectation for events

 tools/perf/tests/shell/attr.sh                   | 1 +
 tools/perf/tests/shell/attr/base-record          | 2 +-
 tools/perf/tests/shell/attr/base-stat            | 2 +-
 tools/perf/tests/shell/attr/system-wide-dummy    | 2 +-
 tools/perf/tests/shell/attr/test-record-dummy-C0 | 2 +-
 5 files changed, 5 insertions(+), 4 deletions(-)

-- 
2.52.0

[PATCH 1/2] perf test attr: Propagate the return value from the test to the wrapper

[Michael Petlan]

Currently it does not matter what the python script actually returns,
the test always passes, as the $err variable is always 0.

Fix that.

Fixes: 8519e4f44c2af72214dc029f0334be068466e71f perf test: Add a shell wrapper for "Setup struct perf_event_attr"

Signed-off-by: Michael Petlan <mpetlan@redhat.com>
---
 tools/perf/tests/shell/attr.sh | 1 +
 1 file changed, 1 insertion(+)

diff --git a/tools/perf/tests/shell/attr.sh b/tools/perf/tests/shell/attr.sh
index 5a4e43b2471d..307e1535795a 100755
--- a/tools/perf/tests/shell/attr.sh
+++ b/tools/perf/tests/shell/attr.sh
@@ -18,5 +18,6 @@ trap trap_cleanup EXIT TERM INT
 shelldir=$(dirname "$0")
 perf_path=$(which perf)
 python "${shelldir}"/lib/attr.py -d "${shelldir}"/attr -v -p "$perf_path"
+err=$?
 cleanup
 exit $err
-- 
2.52.0

[PATCH 2/2] perf test attr: Fix wrong size expectation for events

[Michael Petlan]

Although config4 has been added into struct perf_event_attr, the size
expectation for the struct hasn't been adjusted. This hasn't been
noticed, since the testcase's return value has been ignored after
rewriting the testcase to python until previous commit.

Fix that.

Fixes: 80cdf208117a36de82a30d210b3b8df0f193e75b ("tools headers UAPI: Sync linux/perf_event.h with the kernel sources")
Fixes: cbbfba4847b8a5299d36e002bf864b21bb83295d ("perf: Add perf_event_attr::config4")

Signed-off-by: Michael Petlan <mpetlan@redhat.com>
---
 tools/perf/tests/shell/attr/base-record          | 2 +-
 tools/perf/tests/shell/attr/base-stat            | 2 +-
 tools/perf/tests/shell/attr/system-wide-dummy    | 2 +-
 tools/perf/tests/shell/attr/test-record-dummy-C0 | 2 +-
 4 files changed, 4 insertions(+), 4 deletions(-)

diff --git a/tools/perf/tests/shell/attr/base-record b/tools/perf/tests/shell/attr/base-record
index b44e4e6e4443..dd76fbdb628f 100644
--- a/tools/perf/tests/shell/attr/base-record
+++ b/tools/perf/tests/shell/attr/base-record
@@ -5,7 +5,7 @@ group_fd=-1
 flags=0|8
 cpu=*
 type=0|1
-size=136
+size=144
 config=0|1
 sample_period=*
 sample_type=263
diff --git a/tools/perf/tests/shell/attr/base-stat b/tools/perf/tests/shell/attr/base-stat
index fccd8ec4d1b0..d769e69bb4f5 100644
--- a/tools/perf/tests/shell/attr/base-stat
+++ b/tools/perf/tests/shell/attr/base-stat
@@ -5,7 +5,7 @@ group_fd=-1
 flags=0|8
 cpu=*
 type=0
-size=136
+size=144
 config=0
 sample_period=0
 sample_type=65536
diff --git a/tools/perf/tests/shell/attr/system-wide-dummy b/tools/perf/tests/shell/attr/system-wide-dummy
index a1e1d6a263bf..9806de3d9c9e 100644
--- a/tools/perf/tests/shell/attr/system-wide-dummy
+++ b/tools/perf/tests/shell/attr/system-wide-dummy
@@ -7,7 +7,7 @@ cpu=*
 pid=-1
 flags=8
 type=1
-size=136
+size=144
 config=9
 sample_period=1
 # PERF_SAMPLE_IP | PERF_SAMPLE_TID | PERF_SAMPLE_TIME |
diff --git a/tools/perf/tests/shell/attr/test-record-dummy-C0 b/tools/perf/tests/shell/attr/test-record-dummy-C0
index 91499405fff4..23e00e7f8ace 100644
--- a/tools/perf/tests/shell/attr/test-record-dummy-C0
+++ b/tools/perf/tests/shell/attr/test-record-dummy-C0
@@ -10,7 +10,7 @@ cpu=0
 pid=-1
 flags=8
 type=1
-size=136
+size=144
 config=9
 sample_period=4000
 # PERF_SAMPLE_IP | PERF_SAMPLE_TID | PERF_SAMPLE_TIME |
-- 
2.52.0

Re: [PATCH 1/2] perf test attr: Propagate the return value from the test to the wrapper

[sashiko-bot]

> Currently it does not matter what the python script actually returns,
> the test always passes, as the $err variable is always 0.
> 
> Fix that.
> 
> Fixes: 8519e4f44c2af72214dc029f0334be068466e71f perf test: Add a shell wrapper for "Setup struct perf_event_attr"
> 
> Signed-off-by: Michael Petlan <mpetlan@redhat.com>

Sashiko has reviewed this patch and found no issues. It looks great!

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260924223027.799654-1-mpetlan@redhat.com?part=1

Re: [PATCH 2/2] perf test attr: Fix wrong size expectation for events

[sashiko-bot]

> Although config4 has been added into struct perf_event_attr, the size
> expectation for the struct hasn't been adjusted. This hasn't been
> noticed, since the testcase's return value has been ignored after
> rewriting the testcase to python until previous commit.
> 
> Fix that.
> 
> Fixes: 80cdf208117a36de82a30d210b3b8df0f193e75b ("tools headers UAPI: Sync linux/perf_event.h with the kernel sources")
> Fixes: cbbfba4847b8a5299d36e002bf864b21bb83295d ("perf: Add perf_event_attr::config4")
> 
> Signed-off-by: Michael Petlan <mpetlan@redhat.com>

Sashiko has reviewed this patch and found no issues. It looks great!

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260924223027.799654-1-mpetlan@redhat.com?part=2

Re: [PATCH 0/2] perf test attr: Fix and re-enable the test

[Namhyung Kim]

Hello,

On Fri, Sep 25, 2026 at 12:30:25AM +0200, Michael Petlan wrote:
> Hello all,
> 
> I have noticed that a testcase formerly known as "Setup struct perf_event_attr",
> later renamed to "Perf attribute expectations test" has been disabled by
> accident during the move.
> 
> That also probably led to forgetting to update the struct size, which is now
> fixed by the second patch.
> 
> Michael
> 
> 
> Michael Petlan (2):
>   perf test attr: Propagate the return value from the test to the
>     wrapper
>   perf test attr: Fix wrong size expectation for events

The "Fixes" tag format is a bit different, but otherwise

Acked-by: Namhyung Kim <namhyung@kernel.org>

Thanks,
Namhyung

Re: [PATCH 0/2] perf test attr: Fix and re-enable the test

[Arnaldo Carvalho de Melo]

On Mon, Oct 05, 2026 at 03:47:25PM -0700, Namhyung Kim wrote:
> On Fri, Sep 25, 2026 at 12:30:25AM +0200, Michael Petlan wrote:
> > I have noticed that a testcase formerly known as "Setup struct perf_event_attr",
> > later renamed to "Perf attribute expectations test" has been disabled by
> > accident during the move.

> > That also probably led to forgetting to update the struct size, which is now
> > fixed by the second patch.

> > Michael Petlan (2):
> >   perf test attr: Propagate the return value from the test to the
> >     wrapper
> >   perf test attr: Fix wrong size expectation for events
 
> The "Fixes" tag format is a bit different, but otherwise

I fixed the Fixes tags.
 
> Acked-by: Namhyung Kim <namhyung@kernel.org>

Thanks, applied both patches,

- Arnaldo

Re: [PATCH 2/2] perf test attr: Fix wrong size expectation for events

[Arnaldo Carvalho de Melo]

On Fri, Sep 25, 2026 at 12:30:27AM +0200, Michael Petlan wrote:
> Although config4 has been added into struct perf_event_attr, the size
> expectation for the struct hasn't been adjusted. This hasn't been
> noticed, since the testcase's return value has been ignored after
> rewriting the testcase to python until previous commit.
> 
> Fix that.
> 
> Fixes: 80cdf208117a36de82a30d210b3b8df0f193e75b ("tools headers UAPI: Sync linux/perf_event.h with the kernel sources")

> Fixes: cbbfba4847b8a5299d36e002bf864b21bb83295d ("perf: Add perf_event_attr::config4")

This cset changes a kernel header, kernel developers are not expected to
update tools/ living code, they are welcomed to do so, so it didn't
introduce a problem, so the cset that updates the tools/include/ is the
one that should be in the fixes.

I was going to do that when I noticed this:

 #define PERF_ATTR_SIZE_VER9                    144     /* add: config4 */
+#define PERF_ATTR_SIZE_VER10                   176     /* Add: sample_simd_{vec|pred}_reg_* */

in tools/include/uapi/linux/perf_event.h, i.e. with your patch it will
continue broken, as it is not taking into account this one:

commit eb89aef367e4701038869e1ddfa719463524e3e3
Author: Dapeng Mi <dapeng1.mi@linux.intel.com>
Date:   Mon Sep 14 09:40:43 2026 +0800

    perf headers: Sync perf_event.h/perf_regs.h with the kernel headers

    Sync the UAPI header changes of supporting SIMD/eGPRs/SSP sampling into
    corresponding tools UAPI headers.

So I'm updating your patch to have 176 instead of 144.

The test is still failing tho, for some other reason:

running '/home/acme/libexec/perf-core/tests/shell/attr/test-record-graph-default'
test limitation '!aarch64'
excluded architecture list ['aarch64']
running '/home/acme/libexec/perf-core/tests/shell/attr/test-record-graph-default-aarch64'
test limitation 'aarch64'
skipped [x86_64] '/home/acme/libexec/perf-core/tests/shell/attr/test-record-graph-default-aarch64'
running '/home/acme/libexec/perf-core/tests/shell/attr/test-record-graph-dwarf'
expected sample_type=45359, got 12583
FAILED '/home/acme/libexec/perf-core/tests/shell/attr/test-record-graph-dwarf' - match failure
---- end(-1) ----
 82: Perf attribute expectations test                                                                       : FAILED!

=== Test Summary ===
Passed main tests : 0
Passed subtests   : 0
Skipped tests     : 0
Failed tests      : 1
List of failed tests:
   82: Perf attribute expectations test
root@x2:~#

Investigating.

- Arnaldo
 
> Signed-off-by: Michael Petlan <mpetlan@redhat.com>
> ---
>  tools/perf/tests/shell/attr/base-record          | 2 +-
>  tools/perf/tests/shell/attr/base-stat            | 2 +-
>  tools/perf/tests/shell/attr/system-wide-dummy    | 2 +-
>  tools/perf/tests/shell/attr/test-record-dummy-C0 | 2 +-
>  4 files changed, 4 insertions(+), 4 deletions(-)
> 
> diff --git a/tools/perf/tests/shell/attr/base-record b/tools/perf/tests/shell/attr/base-record
> index b44e4e6e4443..dd76fbdb628f 100644
> --- a/tools/perf/tests/shell/attr/base-record
> +++ b/tools/perf/tests/shell/attr/base-record
> @@ -5,7 +5,7 @@ group_fd=-1
>  flags=0|8
>  cpu=*
>  type=0|1
> -size=136
> +size=144
>  config=0|1
>  sample_period=*
>  sample_type=263
> diff --git a/tools/perf/tests/shell/attr/base-stat b/tools/perf/tests/shell/attr/base-stat
> index fccd8ec4d1b0..d769e69bb4f5 100644
> --- a/tools/perf/tests/shell/attr/base-stat
> +++ b/tools/perf/tests/shell/attr/base-stat
> @@ -5,7 +5,7 @@ group_fd=-1
>  flags=0|8
>  cpu=*
>  type=0
> -size=136
> +size=144
>  config=0
>  sample_period=0
>  sample_type=65536
> diff --git a/tools/perf/tests/shell/attr/system-wide-dummy b/tools/perf/tests/shell/attr/system-wide-dummy
> index a1e1d6a263bf..9806de3d9c9e 100644
> --- a/tools/perf/tests/shell/attr/system-wide-dummy
> +++ b/tools/perf/tests/shell/attr/system-wide-dummy
> @@ -7,7 +7,7 @@ cpu=*
>  pid=-1
>  flags=8
>  type=1
> -size=136
> +size=144
>  config=9
>  sample_period=1
>  # PERF_SAMPLE_IP | PERF_SAMPLE_TID | PERF_SAMPLE_TIME |
> diff --git a/tools/perf/tests/shell/attr/test-record-dummy-C0 b/tools/perf/tests/shell/attr/test-record-dummy-C0
> index 91499405fff4..23e00e7f8ace 100644
> --- a/tools/perf/tests/shell/attr/test-record-dummy-C0
> +++ b/tools/perf/tests/shell/attr/test-record-dummy-C0
> @@ -10,7 +10,7 @@ cpu=0
>  pid=-1
>  flags=8
>  type=1
> -size=136
> +size=144
>  config=9
>  sample_period=4000
>  # PERF_SAMPLE_IP | PERF_SAMPLE_TID | PERF_SAMPLE_TIME |
> -- 
> 2.52.0
> 

Re: [PATCH 2/2] perf test attr: Fix wrong size expectation for events

[Arnaldo Carvalho de Melo]

On Tue, Oct 06, 2026 at 03:33:57PM +0200, Arnaldo Carvalho de Melo wrote:
> The test is still failing tho, for some other reason:
 
> running '/home/acme/libexec/perf-core/tests/shell/attr/test-record-graph-default'
> test limitation '!aarch64'
> excluded architecture list ['aarch64']
> running '/home/acme/libexec/perf-core/tests/shell/attr/test-record-graph-default-aarch64'
> test limitation 'aarch64'
> skipped [x86_64] '/home/acme/libexec/perf-core/tests/shell/attr/test-record-graph-default-aarch64'
> running '/home/acme/libexec/perf-core/tests/shell/attr/test-record-graph-dwarf'
> expected sample_type=45359, got 12583
> FAILED '/home/acme/libexec/perf-core/tests/shell/attr/test-record-graph-dwarf' - match failure
> ---- end(-1) ----
>  82: Perf attribute expectations test                                                                       : FAILED!

Some fixes for problems undetected because of the lack of error
propagation:

⬢ [acme@tb-x2-perf perf-tools-next]$ git log --oneline -5
027108ff3952e8ae (HEAD -> perf-tools-next) perf test attr: Fix legacy event encodings in group tests
12d947c0d8e7aac6 perf test attr: Fix record dwarf sample_type expectation
cec590f3de003895 perf test attr: Fix wrong size expectation for events
04a0a19ede01e771 perf test attr: Propagate the return value from the test to the wrapper
1dc462fc21490767 (perf-tools-next.korg/tmp.perf-tools-next, perf-tools-next.korg/perf-tools-next, five/perf-tools-next, acme.korg/tmp.perf-tools-next, acme.korg/perf-tools-next) perf test: Add a test for perf timechart --tui
⬢ [acme@tb-x2-perf perf-tools-next]$

Still working to see all gets fixed and that test passes again.

- Arnado