[llvm] [SelectionDAG] Add support for the 3-way comparison intrinsics [US]CMP (PR #91871)

via llvm-commits llvm-commits at lists.llvm.org
Sun Jun 16 14:02:08 PDT 2024


Poseydon42 wrote:

Fixed the crash with AVX-512. `WidenVecRes_CMP` now unrolls the operation into a bunch of scalar operations and then rebuilds a vector whenever the widened operand and result have a different number of elements. There is probably a way to avoid doing this in every situation, but I think it's worth adding that later and in a separate PR.

https://github.com/llvm/llvm-project/pull/91871


More information about the llvm-commits mailing list