projects / platform / upstream / mesa.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 0ab01f4) | patch
ir3: Only use per-wave pvtmem layout for compute
authorConnor Abbott <cwabbott0@gmail.com>
Thu, 20 May 2021 16:30:12 +0000 (18:30 +0200)
committerMarge Bot <eric+marge@anholt.net>
Fri, 21 May 2021 20:45:07 +0000 (20:45 +0000)
commit9350900fcd2096d370a7654eef37c3363a38ba3a
treed5358733f6f53c6941fb999d664fc6f99dac173etree | snapshot
parent0ab01f42156411637d2f3aeba48132c259ef4c02commit | diff
ir3: Only use per-wave pvtmem layout for compute

The blob seems to do this since a630, and it fixes
spec@glsl-1.30@execution@fs-large-local-array on a650.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10922>
src/freedreno/ir3/ir3_shader.c diff | blob | history
Domain: Graphics System / GL;