Stas Bekman <[EMAIL PROTECTED]> writes:
The whole issue is very sensitive, and it's usually win32 that has a
problem with it due to the nature it's memory allocator/deallocator
works, so we really need to add tests that break before trying to
change anything.
+1, especially to fleshing all this out for the various PerlInterpScope directives.
FWIW, although the patch fixes this particular test for me,
I still occasionally get segfaults in t/perl/ithreads.t during
a full t/TEST run (with or without the patch, it's been a problem
I've had for quite a while). So there's still plenty of work to do ...
Everybody gets it once in a while (at the moment it's excluded from the distro). I can't see how it can be fixed, since it happens totally randomly.
Coming up with some more tests to isolate the ithread/PerlInterpScope bugs would really help a lot, because t/SMOKE isn't able to reduce it
to under ~400 tests (mainly because the ithreads.t tests doesn't *always* fail for me, and it *never* seems to fail when it's run all by itself).
Right. Most likely it's caused by a PERL_(SET|GET)_CONTEXT issue, but the cause could be different too. Debugging threads is not fun. I had cases where t/SMOKE would get just a few tens of tests, but it was still not very helpful.
-- __________________________________________________________________ Stas Bekman JAm_pH ------> Just Another mod_perl Hacker http://stason.org/ mod_perl Guide ---> http://perl.apache.org mailto:[EMAIL PROTECTED] http://use.perl.org http://apacheweek.com http://modperlbook.org http://apache.org http://ticketmaster.com
--------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
