Bug#786592: libparallel-prefork-perl: autopkgtest failure: t/01-base.t

[gregor herrmann]

Control: forwarded -1 https://rt.cpan.org/Public/Bug/Display.html?id=113449
Control: tag -1 - unreproducible
Control: severity -1 important
Control: retitle -1 libparallel-prefork-perl: intermittent test failures

On Sat, 26 Nov 2016 11:46:59 +0100, Santiago Vila wrote:

> On Sat, 23 May 2015, gregor herrmann wrote:
> > I can't reproduce this locally, 3 subsequential adt-run's succeed.
> > Maybe some race?
> After getting one build failure a few days ago (attached) I now tried
> 100 more times and it failed none of those 100 times, so I guess you
> will have to try more than three times.

Thanks for the additional data point!
 
> I would forward this upstream and disable the test until we know why
> it fails, as this failure makes the package to FTBFS when it happens.

The question is which test to disable. What ci.d.n and you found are
failures in t/01-base.t; I also found occasional failures in
t/06-wait-all-children-with-timeout.t (that's in 0.18, only in git
and not uploaded yet). And Petr Písař found failures in
t/03-spareworkers.t : https://rt.cpan.org/Public/Bug/Display.html?id=113449

CPAN testers at http://matrix.cpantesters.org/?dist=Parallel-Prefork+0.18
have failures in t/07-wait-all-children-does-not-block.t (also new in
0.18), t/01-base.t, and t/06-wait-all-children-with-timeout.t .

When running the tests in a loop, I now get a failure in
t/06-wait-all-children-with-timeout.t after approximately 10 minutes.


t/06-wait-all-children-with-timeout.t .... 
1..4
ok 1 - should reap one worker.
not ok 2 - should reap one worker.

#   Failed test 'should reap one worker.'
#   at t/06-wait-all-children-with-timeout.t line 53.
#          got: '12'
#     expected: '10'
ok 3 - all workers reaped.
ok 4 - properly called on_child_reap callback
# Looks like you failed 1 test of 4.
Dubious, test returned 1 (wstat 256, 0x100)
Failed 1/4 subtests 


15 minutes later the same. strace log attached.

30 minutes later: no more failures.


10 minutes of running only t/01-base.t in a loop: it hangs:

t/01-base.t .. 
1..5
ok 1 - use Parallel::Prefork;
ok 2

and then dies:

ok 3
not ok 4

#   Failed test at t/01-base.t line 66.
#          got: '10'
#     expected: '20'
ok 5 - properly called on_child_reap callback
# Looks like you failed 1 test of 5.
Dubious, test returned 1 (wstat 256, 0x100)
Failed 1/5 subtests 

Test Summary Report
-------------------
t/01-base.t (Wstat: 256 Tests: 5 Failed: 1)
  Failed test:  4
  Non-zero exit status: 1
Files=1, Tests=5, 100 wallclock secs ( 0.03 usr  0.00 sys +  0.04 cusr  0.03 csys =  0.10 CPU)
Result: FAIL


Strace log attached as well.



Cheers,
gregor

-- 
 .''`.  https://info.comodo.priv.at/ - Debian Developer https://www.debian.org
 : :' : OpenPGP fingerprint D1E1 316E 93A7 60A8 104D  85FA BB3A 6801 8649 AA06
 `. `'  Member of VIBE!AT & SPI, fellow of the Free Software Foundation Europe
   `-   NP: Cat Stevens: Moonshadow
-------------- next part --------------
A non-text attachment was scrubbed...
Name: log.t_01.gz
Type: application/gzip
Size: 62310 bytes
Desc: not available
URL: <http://lists.alioth.debian.org/pipermail/pkg-perl-maintainers/attachments/20161126/1e335d48/attachment-0002.bin>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: log.t_06.gz
Type: application/gzip
Size: 315631 bytes
Desc: not available
URL: <http://lists.alioth.debian.org/pipermail/pkg-perl-maintainers/attachments/20161126/1e335d48/attachment-0003.bin>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 963 bytes
Desc: Digital Signature
URL: <http://lists.alioth.debian.org/pipermail/pkg-perl-maintainers/attachments/20161126/1e335d48/attachment-0001.sig>