projects / platform / upstream / llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: b056d0b) | patch
[AMDGPU] Use VGPR classes in divergent build_vector patterns
authorJay Foad <jay.foad@amd.com>
Fri, 21 Oct 2022 15:42:43 +0000 (16:42 +0100)
committerJay Foad <jay.foad@amd.com>
Fri, 21 Oct 2022 15:57:23 +0000 (16:57 +0100)
commit31bd3e6e92e6249c2741219b24830abf2213bd22
treeea6a3f5057b3cb874788aec3df24c47fa1f64a4etree | snapshot
parentb056d0bb691134c01e426851ab8217e2cdb573e2commit | diff
[AMDGPU] Use VGPR classes in divergent build_vector patterns

This does not appear to affect codegen, but using SGPR classes for
operands of VALU instructions looked weird.

Differential Revision: https://reviews.llvm.org/D136459
llvm/lib/Target/AMDGPU/SIInstructions.td diff | blob | history
Domain: System / Toolchain;