Bug #93527 for PDL-Dims: Fails with Test::Pod::Coverage installed

Tue Mar 04 17:38:11 2014 ANDK [...] cpan.org - Ticket created

Subject:

Fails with Test::Pod::Coverage installed

As per subject. Sample fail report: http://www.cpantesters.org/cpan/report/39939610 HTH && Thanks, (Disclaimer: this was discovered by statistical analysis, I'm not speaking as a user)

Fri Mar 14 14:14:23 2014 FANTASMA [...] cpan.org - Correspondence added

On Tue Mar 04 17:38:11 2014, ANDK wrote: Show quoted text

> As per subject. Sample fail report: > > http://www.cpantesters.org/cpan/report/39939610 > > HTH && Thanks, > > (Disclaimer: this was discovered by statistical analysis, I'm not > speaking as a user)

Thanks for reporting, my problem is, I don't know what it means. There is no offending code line mentioned. It passes on my box, so I'm at a loss where to start.

Fri Mar 14 14:14:23 2014 The RT System itself - Status changed from 'new' to 'open'

Sat Mar 15 01:40:02 2014 ANDK [...] cpan.org - Correspondence added

RT-Send-CC:

SREZIC [...] cpan.org

CC'ing Slaven who sent the report. Next time this happens, you can find the email of the tester via www.cpantesters.org. There is a link labeled "Find a tester". The href points then to http://stats.cpantesters.org/cpanmail.html Having said that, I also sent you such a report: http://www.cpantesters.org/cpan/report/40171094 At the moment I'm at a loss here, it's not obvious where the bug is. I'll try to follow up asap. Are you saying that you cannot reproduce, even when you have Test::Pod::Coverage installed?

Sat Mar 15 05:37:27 2014 SREZIC [...] cpan.org - Correspondence added

On 2014-03-14 14:14:23, FANTASMA wrote: Show quoted text

> On Tue Mar 04 17:38:11 2014, ANDK wrote:

> > As per subject. Sample fail report: > > > > http://www.cpantesters.org/cpan/report/39939610 > > > > HTH && Thanks, > > > > (Disclaimer: this was discovered by statistical analysis, I'm not > > speaking as a user)

> Thanks for reporting, my problem is, I don't know what it means. There > is no offending code line mentioned. > > It passes on my box, so I'm at a loss where to start.

It looks like an invisible taint-related error. It can be reproduced using the following command: $ perl5.19.9 -T -Mblib -e 'require "./lib/PDL/Dims.pm"; warn "ok"' Nothing is printed out here, perl silently dies without an error message. Exit code is non-zero. The same command without -T shows the expected "ok" message (and some unrelated smartmatch-related warnings in bleadperl). Running this line with strace shows the invisible error message: write(2, "Insecure $ENV{PATH} while running with -T switch at /opt/perl-5.19.9/lib/site_perl/5.19.9/i686-linux/PDL/IO/Pic.pm line 190, <DATA> line 207.\nCompilation failed in require at (eval 9) line 21, <DATA> line 207.\nBEGIN failed--compilation aborted at (eval 9) line 21, <DATA> line 207.\nBEGIN failed--compilation aborted at ./lib/PDL/Dims.pm line 29, <DATA> line 207.\nCompilation failed in require at -e line 1, <DATA> line 207.\n", 424) = 424 PDL::IO::Pic seems to do some heavy initialization stuff even if it is just require'd or use'd, doing some system() calls. This seems to be forbidden in taint mode unless $ENV{PATH} is set to a sane value. The above oneliner works if something like $ENV{PATH}="/bin:/usr/bin"; is put in front of the perl code. The easiest solution is to remove the "-T" in the shebang of the test script. A better solution is maybe to talk to the PDL::IO::Pic author and ask him/her to move the initialization stuff to a later phase. Regards, Slaven

Mon Mar 17 09:46:05 2014 FANTASMA [...] cpan.org - Correspondence added

RT-Send-CC:

SREZIC [...] cpan.org

I removed the -T switch, it should pass now. Incidentially, I did not have Pod-Coverage installed. After installing, I could reproduce.

Mon Mar 17 09:46:05 2014 FANTASMA [...] cpan.org - Status changed from 'open' to 'resolved'

Mon Mar 17 09:46:06 2014 FANTASMA [...] cpan.org - Fixed in 0.005002 added