[Bug 1620547] New: Crash during perl-XML-LibXML rebuild

https://bugzilla.redhat.com/show_bug.cgi?id=1620547

            Bug ID: 1620547
           Summary: Crash during perl-XML-LibXML rebuild
           Product: Fedora
           Version: rawhide
         Component: perl
          Assignee: jplesnik(a)redhat.com
          Reporter: ppisar(a)redhat.com
        QA Contact: extras-qa(a)fedoraproject.org
                CC: alexl(a)redhat.com, caillon+fedoraproject(a)gmail.com,
                    iarnell(a)gmail.com, jplesnik(a)redhat.com, kasal(a)ucw.cz,
                    mbarnes(a)fastmail.com, mmaslano(a)redhat.com,
                    perl-devel(a)lists.fedoraproject.org, ppisar(a)redhat.com,
                    psabata(a)redhat.com, rhughes(a)redhat.com,
                    sandmann(a)redhat.com, tcallawa(a)redhat.com
   External Bug ID: CPAN 91800



+++ This bug was initially created as a clone of Bug #1615923 +++

Description of problem:
Rebuild of perl-XML-LibXML occasionally fails due to perl crashing in the
execution of the 90threads.t test:

t/80registryleak.t ................... ok
PmmREFCNT_dec: REFCNT decremented below 0 for 123124e0! at t/90threads.t line
137.
*** Error in `/usr/bin/perl': corrupted size vs. prev_size: 0x00000000123123f0
***
======= Backtrace: =========
/lib64/libc.so.6(+0x7ca04)[0xffffaffeca04]
/lib64/libc.so.6(+0x7d48c)[0xffffaffed48c]
/usr/lib64/perl5/vendor_perl/auto/threads/threads.so(+0x2d5c)[0xffffaff12d5c]
/usr/lib64/perl5/vendor_perl/auto/threads/threads.so(ithread_mg_free+0x34)[0xffffaff12ea4]
/usr/lib64/perl5/CORE/libperl.so(+0xbafa0)[0xffffb03fafa0]
/usr/lib64/perl5/CORE/libperl.so(Perl_sv_unmagic+0x20)[0xffffb03fb0d0]
/usr/lib64/perl5/vendor_perl/auto/threads/threads.so(+0x3910)[0xffffaff13910]
/usr/lib64/perl5/CORE/libperl.so(Perl_pp_entersub+0x500)[0xffffb03f1854]
/usr/lib64/perl5/CORE/libperl.so(Perl_call_sv+0x5c4)[0xffffb038b2e4]
/usr/lib64/perl5/CORE/libperl.so(+0xb9aec)[0xffffb03f9aec]
/usr/lib64/perl5/CORE/libperl.so(Perl_sv_clear+0x420)[0xffffb03fa34c]
/usr/lib64/perl5/CORE/libperl.so(Perl_sv_free2+0x5c)[0xffffb03fa934]
/usr/lib64/perl5/CORE/libperl.so(Perl_free_tmps+0x74)[0xffffb041ae0c]
/usr/lib64/perl5/CORE/libperl.so(Perl_pp_nextstate+0x7c)[0xffffb03eb010]
/usr/lib64/perl5/CORE/libperl.so(Perl_runops_standard+0x20)[0xffffb03ea9b4]
/usr/lib64/perl5/CORE/libperl.so(perl_run+0x100)[0xffffb038f8ac]
/usr/bin/perl[0x400f74]
/lib64/libc.so.6(__libc_start_main+0xf0)[0xffffaff915d4]
/usr/bin/perl[0x400fac]

[...]

Test Summary Report
-------------------
t/90threads.t                      (Wstat: 134 Tests: 7 Failed: 0)
  Non-zero wait status: 134
  Parse errors: Bad plan.  You planned 24 tests but ran 7.

[...]

--- Additional comment from Petr Pisar on 2018-08-23 07:06:09 GMT ---

The threads support is known to be broken and there is no fix in sight. I
recommend disabling these thread tests that are actually optional and run
explicitly when building RPM package in contrast to upstream the does not run
them (for the threads support immaturity).

----

Reproducer:

$ while (THREAD_TEST=1 prove -b t/90threads.t); do :;done

-- 
You are receiving this mail because:
You are on the CC list for the bug.