Skip Menu |

This queue is for tickets about the BibTeX-Parser CPAN distribution.

Report information
The Basics
Id: 110710
Status: resolved
Priority: 0/
Queue: BibTeX-Parser
People
Owner: Nobody in particular
Requestors: SREZIC [...] cpan.org
Cc:
AdminCc:
Bug Information
Severity: (no value)
Broken in: 0.66
Fixed in: 0.67

History Show all quoted text
  Mon Dec 28 11:41:57 2015 SREZIC [...] cpan.org - Ticket created
Subject: Possible hash randomization problems in test suite
See http://matrix.cpantesters.org/?dist=BibTeX-Parser%200.66 for a fail/pass overview --- there are many fail reports and almost no pass reports for newer perls. It is not 100% clear, but maybe some hash randomization issue is hitting here. t/13_output.t passes on my smoke machine if I set PERL_HASH_SEED to a special value: env PERL_HASH_SEED=61 PERL_PERTURB_KEYS=NO perl5.23.6 -Mblib t/13_output.t For most other PERL_HASH_SEED values the test fails.
  Mon Dec 28 11:47:15 2015 SREZIC [...] cpan.org - Correspondence added
On 2015-12-28 11:41:57, SREZIC wrote: Show quoted text
> See http://matrix.cpantesters.org/?dist=BibTeX-Parser%200.66 for a > fail/pass overview --- there are many fail reports and almost no pass > reports for newer perls. It is not 100% clear, but maybe some hash > randomization issue is hitting here. > > t/13_output.t passes on my smoke machine if I set PERL_HASH_SEED to a > special value: > > env PERL_HASH_SEED=61 PERL_PERTURB_KEYS=NO perl5.23.6 -Mblib > t/13_output.t > > For most other PERL_HASH_SEED values the test fails.
Also, the tests seem to _always_ fail with perl 5.14.4 and 5.16.3. These perls don't have hash randomization implemented, so my guess is that the hash implementation is different here and always returns a different order than expected.
  Tue Dec 29 06:07:05 2015 SREZIC [...] cpan.org - Correspondence added
On 2015-12-28 11:47:15, SREZIC wrote: Show quoted text
> On 2015-12-28 11:41:57, SREZIC wrote:
> > See http://matrix.cpantesters.org/?dist=BibTeX-Parser%200.66 for a > > fail/pass overview --- there are many fail reports and almost no pass > > reports for newer perls. It is not 100% clear, but maybe some hash > > randomization issue is hitting here. > > > > t/13_output.t passes on my smoke machine if I set PERL_HASH_SEED to a > > special value: > > > > env PERL_HASH_SEED=61 PERL_PERTURB_KEYS=NO perl5.23.6 -Mblib > > t/13_output.t > > > > For most other PERL_HASH_SEED values the test fails.
> > Also, the tests seem to _always_ fail with perl 5.14.4 and 5.16.3. > These perls don't have hash randomization implemented, so my guess is > that the hash implementation is different here and always returns a > different order than expected.
0.67 looks fine: http://matrix.cpantesters.org/?dist=BibTeX-Parser%200.67 Resolving.
  Tue Dec 29 06:07:12 2015 SREZIC [...] cpan.org - Status changed from 'new' to 'resolved'
  Tue Dec 29 06:07:12 2015 SREZIC [...] cpan.org - Fixed in 0.67 added