AMDGPU: Fix incorrect commute with sub when folding immediates
When a fold of an immediate into a sub/subrev required shrinking the
instruction, the wrong VOP2 opcode was used. This was using the VOP2
equivalent of the original instruction, not the commuted instruction
with the inverted opcode.
llvm-svn: 359883
diff --git a/llvm/lib/Target/AMDGPU/SIFoldOperands.cpp b/llvm/lib/Target/AMDGPU/SIFoldOperands.cpp
index 75d2012..8db2ee2 100644
--- a/llvm/lib/Target/AMDGPU/SIFoldOperands.cpp
+++ b/llvm/lib/Target/AMDGPU/SIFoldOperands.cpp
@@ -372,7 +372,10 @@
assert(MI->getOperand(1).isDef());
- int Op32 = AMDGPU::getVOPe32(Opc);
+ // Make sure to get the 32-bit version of the commuted opcode.
+ unsigned MaybeCommutedOpc = MI->getOpcode();
+ int Op32 = AMDGPU::getVOPe32(MaybeCommutedOpc);
+
FoldList.push_back(FoldCandidate(MI, CommuteOpNo, OpToFold, true,
Op32));
return true;