[PATCH] D84779: [AMDGPU] Add amdgpu specific loop threshold metadata

[Tim Corringham via Phabricator via llvm-commits]

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

Added custom loop metadata amdgpu.loop.unroll.threshold metadata to allow the
initial target specific threshold value to be specified on a loop by loop basis.
The intention is to be able to to allow more nuanced hints, e.g. specifying a
low threshold value rather than using llvm.loop.unroll.disable.


Repository:
  rG LLVM Github Monorepo

https://reviews.llvm.org/D84779

Files:
  llvm/lib/Target/AMDGPU/AMDGPUTargetTransformInfo.cpp
  llvm/test/Transforms/LoopUnroll/AMDGPU/unroll-threshold.ll

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D84779.281302.patch
Type: text/x-patch
Size: 5663 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20200728/0e2f7b44/attachment.bin>