[all-commits] [llvm/llvm-project] 3c1273: [AMDGPU] Add amdgpu specific loop threshold metadata

[Tim Corringham via All-commits]

  Branch: refs/heads/master
  Home:   https://github.com/llvm/llvm-project
  Commit: 3c1273d7378e1d51abf9b43c764ccb9828e4a26a
      https://github.com/llvm/llvm-project/commit/3c1273d7378e1d51abf9b43c764ccb9828e4a26a
  Author: Tim Corringham <tcorring at amd.com>
  Date:   2020-10-22 (Thu, 22 Oct 2020)

  Changed paths:
    M llvm/lib/Target/AMDGPU/AMDGPUTargetTransformInfo.cpp
    A llvm/test/Transforms/LoopUnroll/AMDGPU/unroll-threshold.ll

  Log Message:
  -----------
  [AMDGPU] Add amdgpu specific loop threshold metadata

Add new loop metadata amdgpu.loop.unroll.threshold to allow the initial AMDGPU
specific unroll 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 to indicate that a loop may be unrolled if cheap enough
rather than using the all or nothing llvm.loop.unroll.disable metadata.

Differential Revision: https://reviews.llvm.org/D84779