Changpeng Fang bdf8e308b7 AMDGPU: Don't avoid clamp of bit shift in BFE pattern (#115372) ...

Enable pattern matching from "x<<32-y>>32-y" to "bfe x, 0, y" when we
know y is in [0,31].
This is the follow-up for the PR:
https://github.com/llvm/llvm-project/pull/114279 to fix the issue:
https://github.com/llvm/llvm-project/issues/114282

2024-11-07 14:15:33 -08:00

127 KiB

Raw Blame History

View Raw