Re: buildbot failure in OpenCSW Buildbot on ggrep-solaris10-i386

[Jim Meyering]

On Tue, Aug 10, 2021 at 10:11 AM Dagobert Michelsen <dam@opencsw.org> wrote:
> Hi Bruno,
>
> Am 10.08.2021 um 09:58 schrieb Bruno Haible <bruno@clisp.org>:
> >> On Mon, Aug 9, 2021 at 9:54 PM <buildbot@opencsw.org> wrote:
> >>> The Buildbot has detected a new failure on builder ggrep-solaris10-i386 
> >>> while building ggrep. Full details are available at:
> >>>    
> >>> https://buildfarm.opencsw.org/buildbot/builders/ggrep-solaris10-i386/builds/234
> >>>
> >>> Buildbot URL: https://buildfarm.opencsw.org/buildbot/
> >>>
> >>> Buildslave for this Build: unstable10x
> >>>
> >>> Build Reason: The SingleBranchScheduler scheduler named 
> >>> 'schedule-ggrep-solaris10-i386' triggered this build
> >>> Build Source Stamp: [branch master] 
> >>> 6aaef3b8a7f109ab8f39fd46ede01aebbd9f8205
> >>> Blamelist: Jim Meyering <meyering@fb.com>
> >>>
> >>> BUILD FAILED: failed shell_3
> >>
> >> FYI, that shows this failure:
> >>
> >> ../build-aux/test-driver: line 107: 21993 Segmentation Fault
> >> (core dumped) "$@" > $log_file 2>&1
> >> FAIL: test-nl_langinfo-mt
> >>
> >> Cc'ing to bug-gnulib.
> >
> > In your log file I see:
> >
> >  FAIL: test-nl_langinfo-mt
> >  =========================
> >
> >  FAIL test-nl_langinfo-mt (exit status: 139)
> >
> > So, it's crashing with a plain SIGSEGV.
> >
> > There was a known bug, reported at [1] and supposedly fixed through [2].
> > At this point, it would require debugging.
> >
> > Dago:
> >  1) Can you reproduce it with Solaris 11? Then it would have higher
> >     priority than on Solaris 10 only.
> >  2) Can you produce a gdb stack trace of the crash?
> >
> > Bruno
> >
> > [1] https://lists.gnu.org/archive/html/bug-gnulib/2020-01/msg00023.html
> > [2] https://lists.gnu.org/archive/html/bug-gnulib/2020-09/msg00097.html
>
> Looks like the issue is in lmalloc. Maybe the cause is already obvious,
> I’ll see that I can replicate the issue on Solaris 11 meanwhile:

Any luck? I hope to make the release today.