diff options
| author | Jay Foad <jay.foad@amd.com> | 2022-03-01 16:07:43 +0000 |
|---|---|---|
| committer | Jay Foad <jay.foad@amd.com> | 2022-03-02 10:22:10 +0000 |
| commit | 8bed52c9eb8b871aaee3e082b71e60c5dbf2fc46 (patch) | |
| tree | c9f3cbb7899699094fec18bd0d018773a83c3409 /llvm/lib/Object/Archive.cpp | |
| parent | d813116c9deaa960ddcce5b4b161ea589d6e9a34 (diff) | |
| download | llvm-8bed52c9eb8b871aaee3e082b71e60c5dbf2fc46.zip llvm-8bed52c9eb8b871aaee3e082b71e60c5dbf2fc46.tar.gz llvm-8bed52c9eb8b871aaee3e082b71e60c5dbf2fc46.tar.bz2 | |
[AMDGPU] Make more use of madmk/fmamk instructions
In convertToThreeAddress handle VOP2 mac/fmac instructions with a
literal src0 operand, since these are prime candidates for
converting to madmk/fmamk.
Previously this would only happen if src0 (or src1) was a register
defined by a move-immediate instruction, but in many cases these
operands have already been folded because SIFoldOperands runs
before TwoAddressInstructionPass.
Differential Revision: https://reviews.llvm.org/D120736
Diffstat (limited to 'llvm/lib/Object/Archive.cpp')
0 files changed, 0 insertions, 0 deletions