I had a few minor comments but everything looks pretty good. Do with the comments what you will.
Reviewed-by: Jason Ekstrand <ja...@jlekstrand.net> On Mon, May 16, 2016 at 9:22 PM, Francisco Jerez <curroje...@riseup.net> wrote: > This series fixes the FS register spilling code to apply correct > channel masks to the spilled data regardless of the datatype and > execution controls of the generating instruction, which is especially > important for instructions manipulating data of type size other than > 32 bit (e.g. FP64) or instructions with non-trivial execution controls > (which will become somewhat more common in SIMD32 shaders). > > A bunch of FP64 piglit tests that spill registers are fixed by this > series, and it considerably improves the piglit test pass rate with > INTEL_DEBUG set to spill_fs to force the register allocator to spill > all registers -- On SKL I only need an additional workaround to get it > to survive a full spill_fs piglit run without hangs or regressions, > but because the problem doesn't seem to be compiler-related I plan to > address it separately after some additional investigation. > > You can find the same series in my Mesa tree: > > https://cgit.freedesktop.org/~currojerez/mesa/log/?h=i965-spilling-fixes > _______________________________________________ > mesa-dev mailing list > mesa-dev@lists.freedesktop.org > https://lists.freedesktop.org/mailman/listinfo/mesa-dev >
_______________________________________________ mesa-dev mailing list mesa-dev@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/mesa-dev
