Brox Chen 4456f31a8f

[AMDGPU][True16] Use COPY instead of V_MOV for non-imm operand in movePackToVALU lowering (#185754) ...

We should use COPY here for no-imm operands to reduce the number of
generated mov in the isa. However, there is an issue in
https://github.com/llvm/llvm-project/pull/162389#discussion_r2430459341
here that blocked me from doing it.

With https://github.com/llvm/llvm-project/pull/185751 this should work
now

2026-03-12 17:12:45 -04:00

11 MiB

Raw Blame History

The file is too large to be shown. View Raw