On 24 June 2015 at 12:57, Michel Dänzer <mic...@daenzer.net> wrote: > On 24.06.2015 11:39, Dave Airlie wrote: >>> >>> Actually, I'm almost 100% certain that there are lots of other strict >>> aliasing violations in the Mesa code. That's why we've always disabled it. >>> >>> More generally, IMO it's unrealistic to rely on strict aliasing for >>> optimization, because very few people really understand it (I'm not one >>> of them). >> >> I personally think we should get past the, aliasing is hard, lets go >> shopping, > > I'm not saying that. I'm saying we'll keep getting it wrong, which will > cause subtle and very hard to debug problems. > Well my hope is gcc and clang got better, that said I know the kernel disables it as well, but they do a lot worse things.
text data bss dec hex filename 5489951 187400 25712 5703063 570597 lib/mesa_dri_drivers.so text data bss dec hex filename 5446885 187400 25712 5659997 565d5d lib/mesa_dri_drivers.so here is a i965 driver, 43k code size drop. Again fixing things may not help that. Dave. _______________________________________________ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
