diff options
| author | Jay Foad <jay.foad@amd.com> | 2020-09-25 16:07:27 +0100 |
|---|---|---|
| committer | Jay Foad <jay.foad@amd.com> | 2020-10-05 09:55:10 +0100 |
| commit | 16778b19f2c2756a9e0dd04636fb2c269f684917 (patch) | |
| tree | 9997a69f5a7970ed5390a5b8badeadb28d9b76f7 /clang/lib/Frontend/CompilerInvocation.cpp | |
| parent | 0d5989bb24934802a9e6fcca63848a57a91efcc8 (diff) | |
| download | llvm-16778b19f2c2756a9e0dd04636fb2c269f684917.zip llvm-16778b19f2c2756a9e0dd04636fb2c269f684917.tar.gz llvm-16778b19f2c2756a9e0dd04636fb2c269f684917.tar.bz2 | |
[AMDGPU] Make bfe patterns divergence-aware
This tends to increase code size but more importantly it reduces vgpr
usage, and could avoid costly readfirstlanes if the result needs to be
in an sgpr.
Differential Revision: https://reviews.llvm.org/D88580
Diffstat (limited to 'clang/lib/Frontend/CompilerInvocation.cpp')
0 files changed, 0 insertions, 0 deletions