Segfault in mod_perl 1 during global destruction due to Apache::Tablerefcount issues
am 24.03.2011 10:57:00 von demerphqWe recently ran into the problem documented here:
http://www.mail-archive.com/modperl@perl.apache.org/msg00947 .html
http://tech.groups.yahoo.com/group/modperl/messages/55410?th readed=1&m=e&var=1&tidx=1
We ended up finding the offending global storage that was holding on
the Apache::Table objects, and once they forced to undef during
cleanup the segfaults went away.
We tried the patch that Stas posted there, and it helped.
However I am wondering if there was any further work done on this?
Are we missing some patches? And yes, we are trying to get onto mp2.
But for some time we will still have mp1 in use.
Cheers,
yves
--
perl -Mre=debug -e "/just|another|perl|hacker/"