[PATCH] D82194: [AMDGPU] Enable compare operations to be selected by divergence

[Stanislav Mekhanoshin via Phabricator via llvm-commits]

rampitec added inline comments.


================
Comment at: llvm/lib/Target/AMDGPU/SIInstrInfo.cpp:643
+          MBB, MI, DL,
+          get(ST.isWave32() ? AMDGPU::S_CSELECT_B32 : AMDGPU::S_CSELECT_B64),
+          DestReg)
----------------
Given the check above wave32 should not even get here and shall be handles elsewhere.


================
Comment at: llvm/lib/Target/AMDGPU/SIInstrInfo.cpp:5300
+      bool IsSCC = CondReg == AMDGPU::SCC;
+      Register VCC = ST.isWave32() ? AMDGPU::VCC_LO : AMDGPU::VCC;
+      Register EXEC = ST.isWave32() ? AMDGPU::EXEC_LO : AMDGPU::EXEC;
----------------
RI.getVCC()


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D82194/new/

https://reviews.llvm.org/D82194