On 11/06/2012 12:06 PM, Nuno Lopes wrote:
> While I won't oppose this patch since it has no performance impact, I
> believe this is a bug in the compiler and not in the code below.
> The compiler should know about setjmp(). glibc, for example, marks that
> function with __attribute__((returns_twice)).
Yeah, I don't disagree, but I also don't see a problem with helping the
compiler a bit here and it fixes a leak which bit someone badly.
PHP CVS Mailing List (http://www.php.net/)
To unsubscribe, visit: http://www.php.net/unsub.php