[PATCH] D68360: PR41162 Implement LKK remainder and divisibility algorithms

[Tim Gymnich via Phabricator via llvm-commits]

TG908 added a comment.

We've got some mixed results here:

  - `urem` seems to work quite well on X86. `srem` not so much.
  - for most vector operations on X86 and AArch64 the optimization seem beneficial
  - `urem` / `srem` on AArch64 come with quite some overhead for loading constants
- for RISV only the non vector tests changed
- for Power only the vector tests changed




CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D68360/new/

https://reviews.llvm.org/D68360