mirror of
https://github.com/llvm/llvm-project.git
synced 2025-04-23 20:46:06 +00:00

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.