[all-commits] [llvm/llvm-project] 1658b8: [AMDGPU] Avoid using s_cmpk when src0 is not register

[jayfoad via All-commits]

  Branch: refs/heads/master
  Home:   https://github.com/llvm/llvm-project
  Commit: 1658b8d7ddb65eb78e1304b009f1043ab6d9463f
      https://github.com/llvm/llvm-project/commit/1658b8d7ddb65eb78e1304b009f1043ab6d9463f
  Author: Jay Foad <jay.foad at amd.com>
  Date:   2020-07-14 (Tue, 14 Jul 2020)

  Changed paths:
    M llvm/lib/Target/AMDGPU/SIShrinkInstructions.cpp
    A llvm/test/CodeGen/AMDGPU/cmp_shrink.mir

  Log Message:
  -----------
  [AMDGPU] Avoid using s_cmpk when src0 is not register

The hardware spec require src0 of s_cmpk should be a register. So, we
should not optimize s_cmp to s_cmpk if src0 is not register.

Patch by Ruiling Song!