Hi Abhay,

On Thu, Oct 23, 2025 at 04:10:23PM +0530, Abhay Kandpal wrote:
> I tested on ppcle 10 with GCC 14.3.1 and GLIBC 2.40
> Below are new failures observed.

Thanks.
 
> == 794 tests, 6 stderr failures, 0 stdout failures, 1 stderrB failure, 0 
> stdoutB failures, 0 post failures ==
> gdbserver_tests/mcleak                   (stderr)
> gdbserver_tests/mcleak                   (stderrB)
> memcheck/tests/leak-delta                (stderr)

If leak-delta fails, then the mcleak failures are probably the same.
What does memcheck/tests/leak-delta.stderr.diff say?

> memcheck/tests/linux/rfcomm              (stderr)

This seems a long standing issue, it looks like and endian issue?
It is for testing an bluetooth connection.

> helgrind/tests/getaddrinfo               (stderr)
This one is "flaky" on some other arches too.

> none/tests/fdleak_cmsg_xml               (stderr)
> none/tests/socket_close_xml              (stderr)
> 
> 
> Comparing with older release below are new two
> none/tests/fdleak_cmsg_xml               (stderr)
> none/tests/socket_close_xml              (stderr)
> 
> The above two failure are regular and non-critical in nature. Hence they are 
> not blocker for this release.
> I tested on other power versions also and found similar result.
> 
> We will investigate the new regression after the release.

Thanks. I agree that these results seem non-critical.

It is a little surprising the xml variants fail, but apparently not
the non-xml output ones. I do note there is a special ppc64le
socket_close_xml.stderr.exp-ppc64le and
fdleak_cmsg_xml.stderr.exp-ppc64le expect files which are missing one
call frame. Maybe that is related?

Cheers,

Mark


_______________________________________________
Valgrind-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/valgrind-users

Reply via email to