develooper Front page | perl.perl5.porters | Postings from November 1999

Re: [ID 19991129.001] Perl aborts with panic: POPSTACK

Thread Previous
From:
Hugo
Date:
November 29, 1999 16:05
Subject:
Re: [ID 19991129.001] Perl aborts with panic: POPSTACK
Message ID:
199911300007.AAA02341@crypt.compulink.co.uk
In <E11sZdd-0000RM-00@marvin.casa>, Enrico Zini writes:
:I'm developing an administration suite written in Perl, and in just another
:cycle of run/stare at the error/patch, the error was 'panic: POPSTACK'; as the
:Guide suggests, I choose not to panic but to report this as a bug.
:
:I include uuencoded (because I don't know if perlbug is going to ask me about
:files to include as attachments), the minimum subset of code to reproduce the
:error; the code contained in the file has no license in it yet, but it is
:assumed under GPL.

I'm sorry, but 2725 lines of code across 18 files is _not_ the minimum
subset of code to reproduce the bug.

:I think the bug comes from me doing spaghetti object oriented programming
:(Ehi, I'm italian and this is early development code, what would you expect?),
:probably triggered by some error in the code.

As far as I can tell, the POPSTACK occurs when a) a fatal error occurs
and b) the $SIG{__DIE__} handler invoked itself suffers a compile error
during an eval (trying to invoke Connection::Reports::CGI->new, which
doesn't exist). There may be some other factors and a few more nested
evals involved.

I believe it should be possible to reduce the problem to at most 100
lines of code across two or three files, which would make it a lot
easier to attempt to debug; but as the author of the code I think you
are better placed than I to do that.

Hugo van der Sanden

Thread Previous


nntp.perl.org: Perl Programming lists via nntp and http.
Comments to Ask Bjørn Hansen at ask@perl.org | Group listing | About