[all-commits] [llvm/llvm-project] 44920e: [AMDGPU] Disable sub-dword scralar loads IR widening

[Stanislav Mekhanoshin via All-commits]

  Branch: refs/heads/master
  Home:   https://github.com/llvm/llvm-project
  Commit: 44920e8566525c9490347912937e7e6463f33723
      https://github.com/llvm/llvm-project/commit/44920e8566525c9490347912937e7e6463f33723
  Author: Stanislav Mekhanoshin <Stanislav.Mekhanoshin at amd.com>
  Date:   2020-04-10 (Fri, 10 Apr 2020)

  Changed paths:
    M llvm/lib/Target/AMDGPU/AMDGPUCodeGenPrepare.cpp
    A llvm/test/CodeGen/AMDGPU/vectorize-loads.ll
    M llvm/test/CodeGen/AMDGPU/widen_extending_scalar_loads.ll

  Log Message:
  -----------
  [AMDGPU] Disable sub-dword scralar loads IR widening

These will be widened in the DAG. In the meanwhile early
widening prevents otherwise possible vectorization of
such loads.

Differential Revision: https://reviews.llvm.org/D77835