On Fri, 10 Jun 2011 16:07:31 -0700, Kenneth Graunke <kenn...@whitecape.org> wrote: > Prior to this patch, it would happily continue compiling and even > optimizing after a failure was detected. This seems wasteful. > > More importantly, the "message length > 11" failure seems to choke the > instruction scheduler, making it somehow use an undefined value and > segmentation fault.
Odd to have it inside that loop -- if it needs to be in there (instead of just outside afterwards), that would indicate that that check needs to be all over the place in the visitor. Pretty sure it doesn't need to, though. mlen > 11 might have been an issue by overflowing a 16-entry-on-the-stack array in the scheduler
pgpnSGOwKujVH.pgp
Description: PGP signature
_______________________________________________ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
