[X86] Fixes bug in build_vector v4x32 lowering

r222375 made some improvements to build_vector lowering of v4x32 and v4xf32 into an insertps, but it missed a case where:

1. A single extracted element is used twice.
2. The lower of the two non-zero indexes should be preserved, and the higher should be used for the dest mask.

This caused a crash, since the source value for the insertps ends-up uninitialized.

Differential Revision: http://reviews.llvm.org/D6377

llvm-svn: 222635

1 files changed, 6 insertions, 3 deletions

diff --git a/llvm/lib/Target/X86/X86ISelLowering.cpp b/llvm/lib/Target/X86/X86ISelLowering.cpp
index 90e88171dcc..0347a517444 100644
--- a/llvm/lib/Target/X86/X86ISelLowering.cpp
+++ b/llvm/lib/Target/X86/X86ISelLowering.cpp
@@ -5771,7 +5771,8 @@ static SDValue LowerBuildVectorv4x32(SDValue Op, SelectionDAG &DAG,
   // We only know how to deal with build_vector nodes where elements are either
   // zeroable or extract_vector_elt with constant index.
   SDValue FirstNonZero;
-  for (int i=0; i < 4; ++i) {
+  unsigned FirstNonZeroIdx;
+  for (unsigned i=0; i < 4; ++i) {
     if (Zeroable[i])
       continue;
     SDValue Elt = Op->getOperand(i);
@@ -5782,8 +5783,10 @@ static SDValue LowerBuildVectorv4x32(SDValue Op, SelectionDAG &DAG,
     MVT VT = Elt.getOperand(0).getSimpleValueType();
     if (!VT.is128BitVector())
       return SDValue();
-    if (!FirstNonZero.getNode())
+    if (!FirstNonZero.getNode()) {
       FirstNonZero = Elt;
+      FirstNonZeroIdx = i;
+    }
   }
 
   assert(FirstNonZero.getNode() && "Unexpected build vector of all zeros!");
@@ -5822,7 +5825,7 @@ static SDValue LowerBuildVectorv4x32(SDValue Op, SelectionDAG &DAG,
     return SDValue();
 
   SDValue V2 = Elt.getOperand(0);
-  if (Elt == FirstNonZero)
+  if (Elt == FirstNonZero && EltIdx == FirstNonZeroIdx)
     V1 = SDValue();
 
   bool CanFold = true;