[PATCH] D158077: AMDGPU: Fix verifier error on splatted opencl fmin/fmax and ldexp calls

[Matt Arsenault via Phabricator via llvm-commits]

arsenm created this revision.
arsenm added reviewers: AMDGPU, rampitec, kzhuravl, ronlieb.
Herald added subscribers: foad, kerbowa, hiraditya, Anastasia, tpr, dstuttard, yaxunl, jvesely.
Herald added a project: All.
arsenm requested review of this revision.
Herald added a subscriber: wdng.
Herald added a project: LLVM.

Apparently the spec has overloads for fmin/fmax and ldexp with one of
the operands as scalar. We need to broadcast the scalars to the vector
type.


https://reviews.llvm.org/D158077

Files:
  llvm/lib/Target/AMDGPU/AMDGPULibCalls.cpp
  llvm/test/CodeGen/AMDGPU/amdgpu-simplify-libcall-fmax-splat.ll
  llvm/test/CodeGen/AMDGPU/amdgpu-simplify-libcall-fmin-splat.ll
  llvm/test/CodeGen/AMDGPU/amdgpu-simplify-libcall-ldexp-splat.ll

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D158077.550728.patch
Type: text/x-patch
Size: 69979 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20230816/b27cf56a/attachment-0001.bin>