Skip to content

Commit 0fb9f68

Browse files
authored
[SelectionDAG] Use getRegisterType instead of getTypeToTransformTo in ComputePHILiveOutRegInfo. (#80773)
Since we used getNumRegisters right before this, I think this is the correct interface we should be using here. I'm experimenting with making i32 legal on RISC-V 64, but using i64 for the register type between basic blocks. This was one of the first issues I found trying to do that.
1 parent 2a4f715 commit 0fb9f68

File tree

1 file changed

+1
-1
lines changed

1 file changed

+1
-1
lines changed

llvm/lib/CodeGen/SelectionDAG/FunctionLoweringInfo.cpp

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -431,7 +431,7 @@ void FunctionLoweringInfo::ComputePHILiveOutRegInfo(const PHINode *PN) {
431431

432432
if (TLI->getNumRegisters(PN->getContext(), IntVT) != 1)
433433
return;
434-
IntVT = TLI->getTypeToTransformTo(PN->getContext(), IntVT);
434+
IntVT = TLI->getRegisterType(PN->getContext(), IntVT);
435435
unsigned BitWidth = IntVT.getSizeInBits();
436436

437437
auto It = ValueMap.find(PN);

0 commit comments

Comments
 (0)