Skip Menu |

This queue is for tickets about the Sys-Syslog CPAN distribution.

Report information
The Basics
Id: 69040
Status: resolved
Priority: 0/
Queue: Sys-Syslog

People
Owner: Nobody in particular
Requestors: ssb [...] umn.edu
Cc:
AdminCc:

Bug Information
Severity: Critical
Broken in: 0.29
Fixed in: 0.32



Subject: random smashing of @_ in Sys::Syslog::syslog
When syslog is invoked with the result of the match, the string gets passed in and shows up in @_ but the array gets smashed to undef at line 345. (Stepping through it with perldb.) Slight changes to the code will make the problem go away - specifically passing $s directly instead of using the match and passing $1. I see this with perl-5.8.8 on a RedHat Linux: v5.8.8 built for i686-linux-thread-multi as well as on Gentoo/Sabayon with perl-5.12.2: This is perl 5, version 12, subversion 2 (v5.12.2) built for x86_64-linux-thread-multi (with 7 registered patches, see perl -V for more detail) Locally applied patches: 0001-gentoo_MakeMaker-RUNPATH.diff 0002-gentoo_config_over.diff 0003-gentoo_cpan_definstalldirs.diff 0004-gentoo_cpanplus_definstalldirs.diff 0005-gentoo_create-libperl-soname.diff 0006-gentoo_MakeMaker-delete_packlist.diff 0007-fixes_8d66b3f9_h2hp_fix.diff It doesn't really seem like your problem, but I'm hoping you can replicate it because getting to this point has taken more time than I should have put into it and I'm assuming without drilling deeper it is going to be hard to convince anyone else to look at it. If there is anything else I can do, please let me know.
Subject: Syslog-test
Download Syslog-test
application/octet-stream 172b

Message body not shown because it is not plain text.

Subject: Re: [rt.cpan.org #69040] random smashing of @_ in Sys::Syslog::syslog
Date: Thu, 23 Jun 2011 19:10:48 +0200
To: bug-Sys-Syslog [...] rt.cpan.org
From: Sébastien Aperghis-Tramoni <saper [...] cpan.org>
ssb@umn.edu wrote via RT: Show quoted text
> When syslog is invoked with the result of the match, the string gets > passed in and shows up in @_ but the array gets smashed to undef at > line > 345. (Stepping through it with perldb.) > > Slight changes to the code will make the problem go away - > specifically > passing $s directly instead of using the match and passing $1. > > I see this with perl-5.8.8 on a RedHat Linux: > v5.8.8 built for i686-linux-thread-multi > as well as on Gentoo/Sabayon with perl-5.12.2: > > This is perl 5, version 12, subversion 2 (v5.12.2) built for > x86_64-linux-thread-multi > (with 7 registered patches, see perl -V for more detail) > > Locally applied patches: > 0001-gentoo_MakeMaker-RUNPATH.diff > 0002-gentoo_config_over.diff > 0003-gentoo_cpan_definstalldirs.diff > 0004-gentoo_cpanplus_definstalldirs.diff > 0005-gentoo_create-libperl-soname.diff > 0006-gentoo_MakeMaker-delete_packlist.diff > 0007-fixes_8d66b3f9_h2hp_fix.diff > > > It doesn't really seem like your problem, but I'm hoping you can > replicate it because getting to this point has taken more time than I > should have put into it and I'm assuming without drilling deeper it is > going to be hard to convince anyone else to look at it. > > If there is anything else I can do, please let me know.
Thanks for pointing this bug. I'm not really sure to see when the smashing may occur (can't reproduce the bug here), but I will try nonetheless to change the code to avoid this. Sorry for the time you have spent in that code. -- Sébastien Aperghis-Tramoni Close the world, txEn eht nepO.
From: ssb [...] umn.edu
It is too bad that you can't replicate it. I should probably try to open it as a bug against perl itself since that's really where the problem is. I wish I could see how to replicate it outside using your module, but I'm afraid it would be very hard so I haven't put much effort into it...I'm tempted to pursue it more even if you can make your code not susceptible to it though since I'm convinced it will show up somewhere else eventually. Thanks for whatever you manage to do to get it to go away.