On 5 November 2013 12:13, Chris Forbes <[email protected]> wrote: > Unfortunate :( > > Is the bad behavior from violating this restriction observable? >
That's unclear. Sometimes restrictions like this wind up in the hardware docs because a hardware bug was found, in which case yes, the bad behaviour is definitely observable. Other times restrictions like these arise out of deliberate simplifying assumptions that were made during hardware design, in which case it's possible that violating the restriction is benign, but it's really hard to tell for sure without a vast validation effort. The hardware docs rarely distinguish between these two possibilities. I don't really have a good guess as to what the bad behaviour is in this particular case. > > In any case, Reviewed-by: Chris Forbes <[email protected]> > > On Wed, Nov 6, 2013 at 7:43 AM, Eric Anholt <[email protected]> wrote: > > Paul Berry <[email protected]> writes: > > > >> Hardware docs say we can only use SIMD8 dispatch in this condition. > > > > Reviewed-by: Eric Anholt <[email protected]> > > > > _______________________________________________ > > mesa-dev mailing list > > [email protected] > > http://lists.freedesktop.org/mailman/listinfo/mesa-dev > > >
_______________________________________________ mesa-dev mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/mesa-dev
