Files
clang-p2996/llvm/test/CodeGen/AMDGPU/fp-classify.ll
Brox Chen fb0e7b5f16 [AMDGPU][True16][CodeGen] Implement sgpr folding in true16 (#128929)
We haven't implemented 16 bit SGPRs. Currently allow 32-bit SGPRs to be
folded into True16 bit instructions taking 16 bit values. Also use
sgpr_32 when Imm is copied to spgr_lo16 so it could be further folded.
This improves generated code quality.
2025-04-02 16:08:26 -04:00

28 KiB