[PATCH] D41292: [AMDGPU] Fixed incorrect uniform branch condition

Matt Arsenault via Phabricator via llvm-commits llvm-commits at lists.llvm.org
Thu Jan 4 11:47:35 PST 2018

arsenm added inline comments.

Comment at: lib/Target/AMDGPU/AMDGPUISelDAGToDAG.cpp:1680
+    // it gets changed to S_CBRANCH_VCCNZ in SIFixSGPRCopies, then this S_AND
+    // gets inserted there.)
+    Cond = SDValue(CurDAG->getMachineNode(AMDGPU::S_AND_B64, SL, MVT::i1,
Can you add further explanation? Should it still be OK as long as the incoming condition can be seen to be a VCC producer / not a bit op?


More information about the llvm-commits mailing list