[PATCH] D154517: AMDGPU: Always use v_rcp_f16 and v_rsq_f16

[Matt Arsenault via Phabricator via llvm-commits]

arsenm created this revision.
arsenm added reviewers: AMDGPU, foad, rampitec, b-sumner, Pierre-vh, jhuber6.
Herald added subscribers: StephenFan, kerbowa, hiraditya, Anastasia, tpr, dstuttard, yaxunl, jvesely, kzhuravl.
Herald added a project: All.
arsenm requested review of this revision.
Herald added subscribers: wangpc, wdng.
Herald added a project: LLVM.

These inherited the fast math checks from f32, but the manual suggests
these should be accurate enough for unconditional use. The definition
of correctly rounded is 0.5ulp, but the manual says "0.51ulp". I've
been a bit nervous about changing this as the OpenCL conformance test
does not cover half. Brute force produces identical values compared to
a reference host implementation for all values.


https://reviews.llvm.org/D154517

Files:
  llvm/lib/Target/AMDGPU/AMDGPUCodeGenPrepare.cpp
  llvm/lib/Target/AMDGPU/AMDGPULegalizerInfo.cpp
  llvm/lib/Target/AMDGPU/SIISelLowering.cpp
  llvm/test/CodeGen/AMDGPU/GlobalISel/fdiv.f16.ll
  llvm/test/CodeGen/AMDGPU/GlobalISel/legalize-fdiv.mir
  llvm/test/CodeGen/AMDGPU/fdiv.f16.ll
  llvm/test/CodeGen/AMDGPU/fneg-combines.f16.ll

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D154517.537379.patch
Type: text/x-patch
Size: 65979 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20230705/56927919/attachment.bin>