[PATCH] D111523: [AMDGPU] Don't emit 24 bit mul for > 32 bit result.

[Abinav Puthan Purayil via Phabricator via llvm-commits]

abinavpp created this revision.
abinavpp added reviewers: arsenm, foad, rampitec.
Herald added subscribers: kerbowa, hiraditya, t-tye, tpr, dstuttard, yaxunl, nhaehnle, jvesely, kzhuravl.
abinavpp requested review of this revision.
Herald added subscribers: llvm-commits, wdng.
Herald added a project: LLVM.

AMDGPU 24 bit mul yields a 32 bit result. We should not replace a > 32
bit wide mul with a 24 bit mul if we can't guarantee that the 24 bit mul
will fit in the original width.


Repository:
  rG LLVM Github Monorepo

https://reviews.llvm.org/D111523

Files:
  llvm/lib/Target/AMDGPU/AMDGPUCodeGenPrepare.cpp
  llvm/test/CodeGen/AMDGPU/amdgpu-codegenprepare-mul24.ll

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D111523.378573.patch
Type: text/x-patch
Size: 7766 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20211011/73670523/attachment.bin>