Don't use getSetCCResultType for creating a vselect
The vselect mask isn't a setcc.
This breaks in the case when the result of getSetCCResultType
is larger than the vector operands
e.g. %tmp = select i1 %cmp <2 x i8> %a, <2 x i8> %b
when getSetCCResultType returns <2 x i32>, the assertion
that the (MaskTy.getSizeInBits() == Op1.getValueType().getSizeInBits())
is hit.
No test since I don't think I can hit this with any of the current
targets. The R600/SI implementation would break, since it returns a
vector of i1 for this, but it doesn't reach ExpandSELECT for other
reasons.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@190376 91177308-0d34-0410-b5e6-96231b3b80d8
diff --git a/lib/CodeGen/SelectionDAG/LegalizeVectorOps.cpp b/lib/CodeGen/SelectionDAG/LegalizeVectorOps.cpp
index 6e5fdca..fcb48b9 100644
--- a/lib/CodeGen/SelectionDAG/LegalizeVectorOps.cpp
+++ b/lib/CodeGen/SelectionDAG/LegalizeVectorOps.cpp
@@ -599,8 +599,7 @@
return DAG.UnrollVectorOp(Op.getNode());
// Generate a mask operand.
- EVT MaskTy = TLI.getSetCCResultType(*DAG.getContext(), VT);
- assert(MaskTy.isVector() && "Invalid CC type");
+ EVT MaskTy = VT.changeVectorElementTypeToInteger();
assert(MaskTy.getSizeInBits() == Op1.getValueType().getSizeInBits()
&& "Invalid mask size");