SDag: fix how initial loads are formed when splitting vector ops.

Later code expects the vector loads produced to be directly
concatenable, which means we shouldn't pad anything except the last load
produced with UNDEF.

llvm-svn: 293088

Author: TNorthover

llvm/lib/CodeGen/SelectionDAG/LegalizeVectorTypes.cpp

@@ -3442,7 +3442,10 @@ SDValue DAGTypeLegalizer::GenWidenVectorLoads(SmallVectorImpl<SDValue> &LdChain,
                       LD->getPointerInfo().getWithOffset(Offset),
                       MinAlign(Align, Increment), MMOFlags, AAInfo);
       LdChain.push_back(L.getValue(1));
-      if (L->getValueType(0).isVector()) {
+      if (L->getValueType(0).isVector() && NewVTWidth >= LdWidth) {
+        // Later code assumes the vector loads produced will be mergeable, so we
+        // must pad the final entry up to the previous width. Scalars are
+        // combined separately.
         SmallVector<SDValue, 16> Loads;
         Loads.push_back(L);
         unsigned size = L->getValueSizeInBits(0);

llvm/test/CodeGen/ARM/vector-load.ll

@@ -251,3 +251,13 @@ define <4 x i32> @zextload_v8i8tov8i32_fake_update(<4 x i8>** %ptr) {
         %zlA = zext <4 x i8> %lA to <4 x i32>
 	ret <4 x i32> %zlA
 }
+
+; CHECK-LABEL: test_silly_load:
+; CHECK: ldr {{r[0-9]+}}, [r0, #24]
+; CHECK: vld1.8 {d{{[0-9]+}}, d{{[0-9]+}}}, [r0:128]!
+; CHECK: vldr d{{[0-9]+}}, [r0]
+
+define void @test_silly_load(<28 x i8>* %addr) {
+  load volatile <28 x i8>, <28 x i8>* %addr
+  ret void
+}