!fixup address latest comments, thanks!

Author: fhahn

llvm/lib/Transforms/Vectorize/VPlanTransforms.cpp

@@ -851,20 +851,8 @@ static void simplifyRecipe(VPRecipeBase &R, VPTypeAnalysis &TypeInfo) {
       return;
     }
 
-    if (Blend->isNormalized()) {
-      /// Simplify BLEND %a, %b, Not(%mask) -> BLEND %b, %a, %mask.
-      VPValue *NewMask;
-      if (Blend->getNumOperands() == 3 &&
-          match(Blend->getMask(1), m_Not(m_VPValue(NewMask)))) {
-        VPValue *Inc0 = Blend->getIncomingValue(0);
-        VPValue *Inc1 = Blend->getIncomingValue(1);
-        Blend->setOperand(0, Inc1);
-        Blend->setOperand(1, Inc0);
-        Blend->setOperand(2, NewMask);
-      }
-
+    if (Blend->isNormalized())
       return;
-    }
 
     // Normalize the blend so its first incoming value is used as the initial
     // value with the others blended into it.
@@ -899,6 +887,17 @@ static void simplifyRecipe(VPRecipeBase &R, VPTypeAnalysis &TypeInfo) {
     Blend->replaceAllUsesWith(NewBlend);
     Blend->eraseFromParent();
     recursivelyDeleteDeadRecipes(DeadMask);
+
+    /// Simplify BLEND %a, %b, Not(%mask) -> BLEND %b, %a, %mask.
+    VPValue *NewMask;
+    if (NewBlend->getNumOperands() == 3 &&
+        match(NewBlend->getMask(1), m_Not(m_VPValue(NewMask)))) {
+      VPValue *Inc0 = NewBlend->getIncomingValue(0);
+      VPValue *Inc1 = NewBlend->getIncomingValue(1);
+      NewBlend->setOperand(0, Inc1);
+      NewBlend->setOperand(1, Inc0);
+      NewBlend->setOperand(2, NewMask);
+    }
     return;
   }
 

llvm/test/Transforms/LoopVectorize/if-pred-stores.ll

@@ -317,9 +317,9 @@ define void @bug18724(i1 %cond, ptr %ptr, i1 %cond.2, i64 %v.1, i32 %v.2) {
 ; VEC-NEXT:    [[N_MOD_VF:%.*]] = urem i64 [[TMP4]], 2
 ; VEC-NEXT:    [[N_VEC:%.*]] = sub i64 [[TMP4]], [[N_MOD_VF]]
 ; VEC-NEXT:    [[IND_END:%.*]] = add i64 [[V_1]], [[N_VEC]]
-; VEC-NEXT:    [[TMP5:%.*]] = insertelement <2 x i32> zeroinitializer, i32 [[V_2:%.*]], i32 0
 ; VEC-NEXT:    [[BROADCAST_SPLATINSERT:%.*]] = insertelement <2 x i1> poison, i1 [[COND_2:%.*]], i64 0
 ; VEC-NEXT:    [[BROADCAST_SPLAT:%.*]] = shufflevector <2 x i1> [[BROADCAST_SPLATINSERT]], <2 x i1> poison, <2 x i32> zeroinitializer
+; VEC-NEXT:    [[TMP5:%.*]] = insertelement <2 x i32> zeroinitializer, i32 [[V_2:%.*]], i32 0
 ; VEC-NEXT:    br label [[VECTOR_BODY:%.*]]
 ; VEC:       vector.body:
 ; VEC-NEXT:    [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[PRED_STORE_CONTINUE2:%.*]] ]

llvm/test/Transforms/LoopVectorize/unused-blend-mask-for-first-operand.ll

@@ -168,8 +168,6 @@ define void @test_not_first_lane_only_wide_compare_incoming_order_swapped(ptr %A
 ; CHECK-NEXT:  entry:
 ; CHECK-NEXT:    br i1 false, label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
 ; CHECK:       vector.ph:
-; CHECK-NEXT:    [[BROADCAST_SPLATINSERT:%.*]] = insertelement <4 x i16> poison, i16 [[X]], i64 0
-; CHECK-NEXT:    [[BROADCAST_SPLAT:%.*]] = shufflevector <4 x i16> [[BROADCAST_SPLATINSERT]], <4 x i16> poison, <4 x i32> zeroinitializer
 ; CHECK-NEXT:    br label [[VECTOR_BODY:%.*]]
 ; CHECK:       vector.body:
 ; CHECK-NEXT:    [[INDEX:%.*]] = phi i32 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
@@ -178,8 +176,8 @@ define void @test_not_first_lane_only_wide_compare_incoming_order_swapped(ptr %A
 ; CHECK-NEXT:    [[TMP1:%.*]] = getelementptr inbounds i16, ptr [[A]], i16 [[TMP0]]
 ; CHECK-NEXT:    [[TMP2:%.*]] = getelementptr inbounds i16, ptr [[TMP1]], i32 0
 ; CHECK-NEXT:    [[WIDE_LOAD:%.*]] = load <4 x i16>, ptr [[TMP2]], align 2
-; CHECK-NEXT:    [[TMP3:%.*]] = icmp ult <4 x i16> [[WIDE_LOAD]], [[BROADCAST_SPLAT]]
-; CHECK-NEXT:    [[TMP4:%.*]] = extractelement <4 x i1> [[TMP3]], i32 0
+; CHECK-NEXT:    [[TMP3:%.*]] = extractelement <4 x i16> [[WIDE_LOAD]], i32 0
+; CHECK-NEXT:    [[TMP4:%.*]] = icmp ult i16 [[TMP3]], [[X]]
 ; CHECK-NEXT:    [[PREDPHI:%.*]] = select i1 [[TMP4]], ptr poison, ptr [[B]]
 ; CHECK-NEXT:    [[TMP12:%.*]] = load i16, ptr [[PREDPHI]], align 2
 ; CHECK-NEXT:    [[BROADCAST_SPLATINSERT3:%.*]] = insertelement <4 x i16> poison, i16 [[TMP12]], i64 0