AVX-512: Fixed masked load / store instruction selection for KNL.

Patterns were missing for KNL target for <8 x i32>, <8 x float> masked load/store.

This intrinsic comes with all legal types:
<8 x float> @llvm.masked.load.v8f32(<8 x float>* %addr, i32 align, <8 x i1> %mask, <8 x float> %passThru),
but still requires lowering, because VMASKMOVPS, VMASKMOVDQU32 work with 512-bit vectors only.

All data operands should be widened to 512-bit vector.
The mask operand should be widened to v16i1 with zeroes.

Differential Revision: http://reviews.llvm.org/D15265

llvm-svn: 254909
diff --git a/llvm/lib/CodeGen/SelectionDAG/LegalizeVectorOps.cpp b/llvm/lib/CodeGen/SelectionDAG/LegalizeVectorOps.cpp
index 1fb7b16..8295b2a 100644
--- a/llvm/lib/CodeGen/SelectionDAG/LegalizeVectorOps.cpp
+++ b/llvm/lib/CodeGen/SelectionDAG/LegalizeVectorOps.cpp
@@ -244,7 +244,7 @@
         Changed = true;
         return LegalizeOp(ExpandStore(Op));
       }
-  } else if (Op.getOpcode() == ISD::MSCATTER)
+  } else if (Op.getOpcode() == ISD::MSCATTER || Op.getOpcode() == ISD::MSTORE)
     HasVectorValue = true;
 
   for (SDNode::value_iterator J = Node->value_begin(), E = Node->value_end();
@@ -344,6 +344,9 @@
   case ISD::MSCATTER:
     QueryType = cast<MaskedScatterSDNode>(Node)->getValue().getValueType();
     break;
+  case ISD::MSTORE:
+    QueryType = cast<MaskedStoreSDNode>(Node)->getValue().getValueType();
+    break;
   }
 
   switch (TLI.getOperationAction(Node->getOpcode(), QueryType)) {