On Mon, Jun 28, 2010 at 02:02, Tim Bunce <[email protected]> wrote:
> On Mon, Jun 28, 2010 at 12:24:44AM -0600, Alex Hunsaker wrote:
>> On Sun, Jun 27, 2010 at 15:18, Nicholas Clark <[email protected]> wrote:
>>
>> > This list, obviously, isn't the right place to report bugs in mod_perl.
>>
>> Yeah, ill go post a report on that list (tomorrow) and point back to
>> this thread.  Course feel free to beat me to the punch :-).
>
> Given Nicholas's analysis (thanks Nick!) I'd guess that you could create
> a test case that doesn't involve NYTProf by doing something like:
>
>    END { eval "END { }" for 1..10 }

Yep, kind of interestingly it fails with the below instead of a segfault:
Not a CODE reference.
END failed--call queue aborted.

Mainly as it is passing in NULL to call_sv().

BTW, thanks you guys are awesome!

> p.s. A cross-check of this analysis would be to change the "if (1)" in
> _END to "if (0)" so PL_endav doesn't get altered at END time.

Yeah, if I do this i no longer get a segfault.

> p.p.s. I'll tweak NYTProf to pre-extend PL_endav enough to reduce the
> risk of this happening.

I thought we might need to check that sv is not NULL in
resolve_sub_to_cv().  As it still might get called with NULL.  But I
checked out r1319 and everything seems to be working...

Thanks!

-- 
You've received this message because you are subscribed to
the Devel::NYTProf Development User group.

Group hosted at:  http://groups.google.com/group/develnytprof-dev
Project hosted at:  http://perl-devel-nytprof.googlecode.com
CPAN distribution:  http://search.cpan.org/dist/Devel-NYTProf

To post, email:  [email protected]
To unsubscribe, email:  [email protected]

Reply via email to