AMDGPU: Force skip over SMRD, VMEM and s_waitcnt instructions
authorRhys Perry <pendingchaos02@gmail.com>
Wed, 17 Apr 2019 16:31:52 +0000 (16:31 +0000)
committerRhys Perry <pendingchaos02@gmail.com>
Wed, 17 Apr 2019 16:31:52 +0000 (16:31 +0000)
commitc2814e12e7fd1d51caa3c14a7053dedadd99152c
treec86a0d3ef817a9392abab9b107c4b3cb2b774f4d
parent00f2ab1c2f014ffc790c59fc0e9fa1bc8647cba2
AMDGPU: Force skip over SMRD, VMEM and s_waitcnt instructions

Summary: This fixes a large Dawn of War 3 performance regression with RADV from Mesa 19.0 to master which was caused by creating less code in some branches.

Reviewers: arsen, nhaehnle

Reviewed By: nhaehnle

Subscribers: arsenm, kzhuravl, jvesely, wdng, nhaehnle, yaxunl, dstuttard, tpr, t-tye, llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D60824

llvm-svn: 358592
llvm/lib/Target/AMDGPU/SIInsertSkips.cpp
llvm/test/CodeGen/AMDGPU/collapse-endcf.ll
llvm/test/CodeGen/AMDGPU/skip-if-dead.ll
llvm/test/CodeGen/AMDGPU/valu-i1.ll
llvm/test/CodeGen/AMDGPU/wqm.ll