[PATCH] D147729: [X86] Add inst fixup for `unpckps` -> `unpckdq`/`shufps`.

[Noah Goldstein via Phabricator via llvm-commits]

goldstein.w.n created this revision.
Herald added subscribers: pengfei, hiraditya.
Herald added a project: All.
goldstein.w.n requested review of this revision.
Herald added a project: LLVM.
Herald added a subscriber: llvm-commits.

`unpckps` has the same performance as `unpckpd` (only port5) wereas
`unpckdq`/`shufps` both can run on p15 on some newer architectures.

`unpckdq` is preferable if the target has no bypass delay on shuffles
as it saves 1-byte of code size versus `shufps` and can me used as a
replacement for the micro-fused `rm` version.

If the target has bypass delay on shuffles then we use `shufps`.


Repository:
  rG LLVM Github Monorepo

https://reviews.llvm.org/D147729

Files:
  llvm/lib/Target/X86/X86FixupInstTuning.cpp
  llvm/test/CodeGen/X86/tuning-shuffle-unpckps-avx512.ll
  llvm/test/CodeGen/X86/tuning-shuffle-unpckps.ll

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D147729.511484.patch
Type: text/x-patch
Size: 68927 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20230406/8d1d498f/attachment-0001.bin>