[llvm] [VectorCombine] Fold bitcast(bitreverse.v8i8(bitcast(IntTy))) into bswap+bitreverse (PR #209037)

Simon Pilgrim via llvm-commits llvm-commits at lists.llvm.org
Wed Aug 5 01:50:47 PDT 2026


RKSimon wrote:

VectorCombine usually selects the version with the lowest instruction count, but its not as consistent as it should be - and we've hit a number of infinite loop issues due to it as well.

https://github.com/llvm/llvm-project/pull/209037


More information about the llvm-commits mailing list