Re: Recursive zval patch for review/comments

Just gave it a quick once over and I think it sounds good, I'll play  
with it more this weekend.  One thing I noticed (while looking at  
some fault issues you also mentioned) is that we may want to set APCG 
(copied_zvals) = NULL in RINIT just in case the last run didn't get  
cleaned up properly.

-shire

On Feb 2, 2007, at 2:10 PM, Gopal V wrote:

> Hi,
>
> I'm not quite sure about this patch. But this does bring back
> up apc speed to pre recursive-zval levels.
>
> http://t3.dotgnu.info/code/recursive-zval-mk2.patch
>
> And somehow seems to fix a relatively random, but recurring
> coredump (more on that on monday) with apc-cvs right now on
> centos+apache1+x86.
>
> I'm not quite sure of this patch - i.e is APCG(copied_zvals) allocated
> in every scenario which could have a recursive alloc, is my assumption
> about IS_ARRAY check before hitting copied_zvals right, etc.
>
> Therefore, I'd like a couple more eyeballs on it - just to check
> sanity, at least.
>
> Cheers,
> Gopal
> -- 
> If you have a difficult task, give it to a lazy person --
> they will find an easier way to do it.
>                             -- Hlade's Law
>
> -- 
> APC Development Mailing List (http://pecl.php.net/APC)
> To unsubscribe, visit: http://www.php.net/unsub.php
>

-Brian Shire
  [email protected]
  [email protected]