!fixup address latest comments, thanks

Author: fhahn

llvm/lib/Transforms/Vectorize/LoopVectorize.cpp

@@ -2375,10 +2375,9 @@ void InnerLoopVectorizer::scalarizeInstruction(const Instruction *Instr,
   if (auto *II = dyn_cast<AssumeInst>(Cloned))
     AC->registerAssumption(II);
 
-  // End if-block.
-  VPRegionBlock *Parent = RepRecipe->getParent()->getParent();
   assert(
-      (Parent || !RepRecipe->getParent()->getPlan()->getVectorLoopRegion() ||
+      (RepRecipe->getParent()->getParent() ||
+       !RepRecipe->getParent()->getPlan()->getVectorLoopRegion() ||
        all_of(RepRecipe->operands(),
               [](VPValue *Op) { return Op->isDefinedOutsideLoopRegions(); })) &&
       "Expected a recipe is either within a region or all of its operands "

llvm/test/Transforms/LoopVectorize/AArch64/masked-call.ll

@@ -996,9 +996,9 @@ define void @test_widen_exp_v2(ptr noalias %p2, ptr noalias %p, i64 %n) #5 {
 ; TFA_INTERLEAVE-NEXT:    [[ACTIVE_LANE_MASK_ENTRY1:%.*]] = icmp ult i64 1, [[TMP0]]
 ; TFA_INTERLEAVE-NEXT:    br label %[[VECTOR_BODY:.*]]
 ; TFA_INTERLEAVE:       [[VECTOR_BODY]]:
-; TFA_INTERLEAVE-NEXT:    [[INDEX:%.*]] = phi i64 [ 0, %[[ENTRY]] ], [ [[TMP27:%.*]], %[[TMP19:.*]] ]
-; TFA_INTERLEAVE-NEXT:    [[ACTIVE_LANE_MASK:%.*]] = phi i1 [ [[ACTIVE_LANE_MASK_ENTRY]], %[[ENTRY]] ], [ [[ACTIVE_LANE_MASK_NEXT:%.*]], %[[TMP19]] ]
-; TFA_INTERLEAVE-NEXT:    [[ACTIVE_LANE_MASK2:%.*]] = phi i1 [ [[ACTIVE_LANE_MASK_ENTRY1]], %[[ENTRY]] ], [ [[ACTIVE_LANE_MASK_NEXT6:%.*]], %[[TMP19]] ]
+; TFA_INTERLEAVE-NEXT:    [[INDEX:%.*]] = phi i64 [ 0, %[[ENTRY]] ], [ [[TMP27:%.*]], %[[PRED_STORE_CONTINUE5:.*]] ]
+; TFA_INTERLEAVE-NEXT:    [[ACTIVE_LANE_MASK:%.*]] = phi i1 [ [[ACTIVE_LANE_MASK_ENTRY]], %[[ENTRY]] ], [ [[ACTIVE_LANE_MASK_NEXT:%.*]], %[[PRED_STORE_CONTINUE5]] ]
+; TFA_INTERLEAVE-NEXT:    [[ACTIVE_LANE_MASK2:%.*]] = phi i1 [ [[ACTIVE_LANE_MASK_ENTRY1]], %[[ENTRY]] ], [ [[ACTIVE_LANE_MASK_NEXT6:%.*]], %[[PRED_STORE_CONTINUE5]] ]
 ; TFA_INTERLEAVE-NEXT:    [[TMP4:%.*]] = load double, ptr [[P2]], align 8
 ; TFA_INTERLEAVE-NEXT:    [[TMP5:%.*]] = tail call double @llvm.exp.f64(double [[TMP4]]) #[[ATTR7:[0-9]+]]
 ; TFA_INTERLEAVE-NEXT:    [[TMP8:%.*]] = tail call double @llvm.exp.f64(double [[TMP4]]) #[[ATTR7]]
@@ -1016,11 +1016,11 @@ define void @test_widen_exp_v2(ptr noalias %p2, ptr noalias %p, i64 %n) #5 {
 ; TFA_INTERLEAVE-NEXT:    [[TMP15:%.*]] = xor i1 [[TMP13]], true
 ; TFA_INTERLEAVE-NEXT:    [[TMP16:%.*]] = xor i1 [[TMP14]], true
 ; TFA_INTERLEAVE-NEXT:    [[TMP17:%.*]] = or i1 [[TMP15]], [[TMP16]]
-; TFA_INTERLEAVE-NEXT:    br i1 [[TMP17]], label %[[BB18:.*]], label %[[TMP19]]
+; TFA_INTERLEAVE-NEXT:    br i1 [[TMP17]], label %[[BB18:.*]], label %[[PRED_STORE_CONTINUE5]]
 ; TFA_INTERLEAVE:       [[BB18]]:
 ; TFA_INTERLEAVE-NEXT:    store double [[SPEC_SELECT]], ptr [[P]], align 8
-; TFA_INTERLEAVE-NEXT:    br label %[[TMP19]]
-; TFA_INTERLEAVE:       [[TMP19]]:
+; TFA_INTERLEAVE-NEXT:    br label %[[PRED_STORE_CONTINUE5]]
+; TFA_INTERLEAVE:       [[PRED_STORE_CONTINUE5]]:
 ; TFA_INTERLEAVE-NEXT:    [[TMP27]] = add i64 [[INDEX]], 2
 ; TFA_INTERLEAVE-NEXT:    [[TMP11:%.*]] = add i64 [[INDEX]], 1
 ; TFA_INTERLEAVE-NEXT:    [[ACTIVE_LANE_MASK_NEXT]] = icmp ult i64 [[INDEX]], [[TMP3]]

llvm/test/Transforms/LoopVectorize/X86/small-size.ll

@@ -258,6 +258,8 @@ define void @example2(i32 %n, i32 %x) optsize {
 
 ; Loop has no primary induction as its integer IV has step -1 starting at
 ; unknown N, but can still be vectorized.
+; Note: Most scalar pointer induction GEPs could be sunk into the conditional
+;       blocks.
 define void @example3(i32 %n, ptr noalias nocapture %p, ptr noalias nocapture %q) optsize {
 ; CHECK-LABEL: @example3(
 ; CHECK-NEXT:    [[TMP1:%.*]] = icmp eq i32 [[N:%.*]], 0
@@ -445,6 +447,8 @@ define void @example23b(ptr noalias nocapture %src, ptr noalias nocapture %dst)
 }
 
 ; We CAN vectorize this example by folding the tail it entails.
+; Note: Most scalar pointer induction GEPs could be sunk into the conditional
+;       blocks.
 define void @example23c(ptr noalias nocapture %src, ptr noalias nocapture %dst) optsize {
 ; CHECK-LABEL: @example23c(
 ; CHECK-NEXT:    br i1 false, label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]

llvm/test/Transforms/LoopVectorize/pointer-induction.ll

@@ -3,7 +3,8 @@
 ; RUN: opt < %s -passes=loop-vectorize -force-vector-interleave=1 -force-vector-width=4 -lv-strided-pointer-ivs=true -S | FileCheck --check-prefixes=CHECK,STRIDED %s
 target datalayout = "e-m:e-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128"
 
-
+; Note: Most scalar pointer induction GEPs could be sunk into the conditional
+;       blocks.
 ; Function Attrs: nofree norecurse nounwind
 define void @a(ptr readnone %b) {
 ; CHECK-LABEL: @a(

llvm/test/Transforms/LoopVectorize/tail-folding-vectorization-factor-1.ll

@@ -82,6 +82,8 @@ for.body:
   br i1 %exitcond, label %for.cond.cleanup, label %for.body
 }
 
+; Note: Most scalar pointer induction GEPs could be sunk into the conditional
+;       blocks.
 define void @VF1-VPWidenCanonicalIVRecipeExe(ptr %ptr1) {
 ; CHECK-LABEL: @VF1-VPWidenCanonicalIVRecipeExe(
 ; CHECK-NEXT:  entry: