I was working on my recent problem by sprinkling print statements thruoghout the code for my module. It appears that the problem is in the destructor.
Guess what the destructor does? It closes my CGI::Session object... the one I was having problems with a bit back.
I was using the "Storable" serializer. When I changed it to "FreezeThaw" or used the default... the sigfault stopped.
I'd still like to know why wrapping the code in a sub made a difference...
Re:Storable
phillup on 2004-01-07T19:50:29
Yeah... I'm looking into it today to see if I can figure it out enough to make a bug report.
If it is the same problem I've been having every so often... it has been a real bugger to track down. Very intermittent... and testing from the command line is much different than running in a mod_perl environment.
If it is the Storable module... I need to find a test case that demonstrates the problem, without CGI::Session... I think.Re:Storable
phillup on 2004-01-07T21:40:13
This is wierd...
I tried for about four hours to reduce it to the Storable module. Couldn't do it.
So... I changed the storage mechanism from file based to db based, using MySQL and the Storable serializer.
No segfault.
It only segfaults if I use file based storage and the Storable serializer.
I'm not sure if the problem is in the session module... but it sure looks like it to me. I ended up entering a bug into rt.cpan.org for the CGI::Session module.
