Index: llvm/trunk/lib/CodeGen/SelectionDAG/DAGCombiner.cpp
===================================================================
--- llvm/trunk/lib/CodeGen/SelectionDAG/DAGCombiner.cpp
+++ llvm/trunk/lib/CodeGen/SelectionDAG/DAGCombiner.cpp
@@ -7401,7 +7401,7 @@
   if (TLI.getTypeAction(*DAG.getContext(), Data.getValueType()) !=
       TargetLowering::TypeSplitVector)
     return SDValue();
-  SDValue MaskLo, MaskHi, Lo, Hi;
+  SDValue MaskLo, MaskHi;
   std::tie(MaskLo, MaskHi) = SplitVSETCC(Mask.getNode(), DAG);
 
   EVT LoVT, HiVT;
@@ -7429,17 +7429,15 @@
                           Alignment, MSC->getAAInfo(), MSC->getRanges());
 
   SDValue OpsLo[] = { Chain, DataLo, MaskLo, BasePtr, IndexLo, Scale };
-  Lo = DAG.getMaskedScatter(DAG.getVTList(MVT::Other), DataLo.getValueType(),
-                            DL, OpsLo, MMO);
-
-  SDValue OpsHi[] = { Chain, DataHi, MaskHi, BasePtr, IndexHi, Scale };
-  Hi = DAG.getMaskedScatter(DAG.getVTList(MVT::Other), DataHi.getValueType(),
-                            DL, OpsHi, MMO);
-
-  AddToWorklist(Lo.getNode());
-  AddToWorklist(Hi.getNode());
+  SDValue Lo = DAG.getMaskedScatter(DAG.getVTList(MVT::Other),
+                                    DataLo.getValueType(), DL, OpsLo, MMO);
 
-  return DAG.getNode(ISD::TokenFactor, DL, MVT::Other, Lo, Hi);
+  // The order of the Scatter operation after split is well defined. The "Hi"
+  // part comes after the "Lo". So these two operations should be chained one
+  // after another.
+  SDValue OpsHi[] = { Lo, DataHi, MaskHi, BasePtr, IndexHi, Scale };
+  return DAG.getMaskedScatter(DAG.getVTList(MVT::Other), DataHi.getValueType(),
+                              DL, OpsHi, MMO);
 }
 
 SDValue DAGCombiner::visitMSTORE(SDNode *N) {
Index: llvm/trunk/test/CodeGen/X86/masked_gather_scatter.ll
===================================================================
--- llvm/trunk/test/CodeGen/X86/masked_gather_scatter.ll
+++ llvm/trunk/test/CodeGen/X86/masked_gather_scatter.ll
@@ -2857,11 +2857,11 @@
 ; KNL_64-LABEL: test_scatter_setcc_split:
 ; KNL_64:       # %bb.0:
 ; KNL_64-NEXT:    vextractf64x4 $1, %zmm0, %ymm4
-; KNL_64-NEXT:    vptestnmd %zmm1, %zmm1, %k1
-; KNL_64-NEXT:    vextracti64x4 $1, %zmm1, %ymm1
+; KNL_64-NEXT:    vextracti64x4 $1, %zmm1, %ymm5
+; KNL_64-NEXT:    vptestnmd %zmm5, %zmm5, %k1
 ; KNL_64-NEXT:    vptestnmd %zmm1, %zmm1, %k2
-; KNL_64-NEXT:    vscatterdpd %zmm3, (%rdi,%ymm4,8) {%k2}
-; KNL_64-NEXT:    vscatterdpd %zmm2, (%rdi,%ymm0,8) {%k1}
+; KNL_64-NEXT:    vscatterdpd %zmm2, (%rdi,%ymm0,8) {%k2}
+; KNL_64-NEXT:    vscatterdpd %zmm3, (%rdi,%ymm4,8) {%k1}
 ; KNL_64-NEXT:    vzeroupper
 ; KNL_64-NEXT:    retq
 ;
@@ -2877,11 +2877,11 @@
 ; KNL_32-NEXT:    vmovapd 72(%ebp), %zmm3
 ; KNL_32-NEXT:    movl 8(%ebp), %eax
 ; KNL_32-NEXT:    vextractf64x4 $1, %zmm0, %ymm4
-; KNL_32-NEXT:    vptestnmd %zmm1, %zmm1, %k1
-; KNL_32-NEXT:    vextracti64x4 $1, %zmm1, %ymm1
+; KNL_32-NEXT:    vextracti64x4 $1, %zmm1, %ymm5
+; KNL_32-NEXT:    vptestnmd %zmm5, %zmm5, %k1
 ; KNL_32-NEXT:    vptestnmd %zmm1, %zmm1, %k2
-; KNL_32-NEXT:    vscatterdpd %zmm3, (%eax,%ymm4,8) {%k2}
-; KNL_32-NEXT:    vscatterdpd %zmm2, (%eax,%ymm0,8) {%k1}
+; KNL_32-NEXT:    vscatterdpd %zmm2, (%eax,%ymm0,8) {%k2}
+; KNL_32-NEXT:    vscatterdpd %zmm3, (%eax,%ymm4,8) {%k1}
 ; KNL_32-NEXT:    movl %ebp, %esp
 ; KNL_32-NEXT:    popl %ebp
 ; KNL_32-NEXT:    .cfi_def_cfa %esp, 4
@@ -2891,11 +2891,11 @@
 ; SKX-LABEL: test_scatter_setcc_split:
 ; SKX:       # %bb.0:
 ; SKX-NEXT:    vextractf64x4 $1, %zmm0, %ymm4
-; SKX-NEXT:    vptestnmd %ymm1, %ymm1, %k1
-; SKX-NEXT:    vextracti64x4 $1, %zmm1, %ymm1
+; SKX-NEXT:    vextracti64x4 $1, %zmm1, %ymm5
+; SKX-NEXT:    vptestnmd %ymm5, %ymm5, %k1
 ; SKX-NEXT:    vptestnmd %ymm1, %ymm1, %k2
-; SKX-NEXT:    vscatterdpd %zmm3, (%rdi,%ymm4,8) {%k2}
-; SKX-NEXT:    vscatterdpd %zmm2, (%rdi,%ymm0,8) {%k1}
+; SKX-NEXT:    vscatterdpd %zmm2, (%rdi,%ymm0,8) {%k2}
+; SKX-NEXT:    vscatterdpd %zmm3, (%rdi,%ymm4,8) {%k1}
 ; SKX-NEXT:    vzeroupper
 ; SKX-NEXT:    retq
 ;
@@ -2911,11 +2911,11 @@
 ; SKX_32-NEXT:    vmovapd 72(%ebp), %zmm3
 ; SKX_32-NEXT:    movl 8(%ebp), %eax
 ; SKX_32-NEXT:    vextractf64x4 $1, %zmm0, %ymm4
-; SKX_32-NEXT:    vptestnmd %ymm1, %ymm1, %k1
-; SKX_32-NEXT:    vextracti64x4 $1, %zmm1, %ymm1
+; SKX_32-NEXT:    vextracti64x4 $1, %zmm1, %ymm5
+; SKX_32-NEXT:    vptestnmd %ymm5, %ymm5, %k1
 ; SKX_32-NEXT:    vptestnmd %ymm1, %ymm1, %k2
-; SKX_32-NEXT:    vscatterdpd %zmm3, (%eax,%ymm4,8) {%k2}
-; SKX_32-NEXT:    vscatterdpd %zmm2, (%eax,%ymm0,8) {%k1}
+; SKX_32-NEXT:    vscatterdpd %zmm2, (%eax,%ymm0,8) {%k2}
+; SKX_32-NEXT:    vscatterdpd %zmm3, (%eax,%ymm4,8) {%k1}
 ; SKX_32-NEXT:    movl %ebp, %esp
 ; SKX_32-NEXT:    popl %ebp
 ; SKX_32-NEXT:    .cfi_def_cfa %esp, 4