Index: llvm/lib/Transforms/Utils/ScalarEvolutionExpander.cpp
===================================================================
--- llvm/lib/Transforms/Utils/ScalarEvolutionExpander.cpp
+++ llvm/lib/Transforms/Utils/ScalarEvolutionExpander.cpp
@@ -2179,6 +2179,19 @@
 
   const T *S = cast<T>(WorkItem->S);
   int Cost = 0;
+
+  // We may have some zero operands, and we shouldn't really charge for those.
+  // So how many non-zero coeffients are there?
+  int NumTerms = llvm::count_if(S->operands(), [](const SCEV *Op) {
+                                return !Op->isZero();
+                                });
+  // Ignoring constant term (operand 0), how many of the coeffients are u> 1?
+  int NumNonZeroDegreeNonOneTerms =
+    llvm::count_if(S->operands(), [](const SCEV *Op) {
+                   auto *SConst = dyn_cast<SCEVConstant>(Op);
+                   return !SConst || SConst->getAPInt().ugt(1);
+                   });
+
   SmallVector<unsigned, 2> Opcodes;
   auto CastCost = [&](unsigned Opcode) {
     Opcodes.push_back(Opcode);
@@ -2186,9 +2199,10 @@
                                 S->getOperand(0)->getType(),
                                 TTI::CastContextHint::None, CostKind);
   };
-  auto ArithCost = [&](unsigned Opcode) {
+  auto ArithCost = [&](unsigned Opcode, unsigned NumRequired) {
     Opcodes.push_back(Opcode);
-    return TTI.getArithmeticInstrCost(Opcode, S->getType(), CostKind);
+    int Cost = TTI.getArithmeticInstrCost(Opcode, S->getType(), CostKind);
+    return Cost * NumRequired;
   };
 
   switch (S->getSCEVType()) {
@@ -2208,17 +2222,17 @@
     if (auto *SC = dyn_cast<SCEVConstant>(S->getOperand(1)))
       if (SC->getAPInt().isPowerOf2())
         Opcode = Instruction::LShr;
-    Cost = ArithCost(Opcode);
+    Cost = ArithCost(Opcode, 1);
     break;
   }
   case scAddExpr:
-    Cost = ArithCost(Instruction::Add);
+    Cost = ArithCost(Instruction::Add, NumTerms - 1);
     break;
   case scMulExpr:
     // TODO: this is a very pessimistic cost modelling for Mul,
     // because of Bin Pow algorithm actually used by the expander,
     // see SCEVExpander::visitMulExpr(), ExpandOpBinPowN().
-    Cost = ArithCost(Instruction::Mul);
+    Cost = ArithCost(Instruction::Mul, NumNonZeroDegreeNonOneTerms);
     break;
   case scSMaxExpr:
   case scUMaxExpr:
@@ -2236,31 +2250,16 @@
     break;
   }
   case scAddRecExpr: {
-    int AddCost = ArithCost(Instruction::Add);
-    int MulCost = ArithCost(Instruction::Mul);
-
-    // In this polynominal, we may have some zero operands, and we shouldn't
-    // really charge for those. So how many non-zero coeffients are there?
-    int NumTerms = llvm::count_if(S->operands(), [](const SCEV *Op) {
-                                    return !Op->isZero();
-                                  });
     assert(NumTerms >= 1 && "Polynominal should have at least one term.");
     assert(!(*std::prev(S->operands().end()))->isZero() &&
            "Last operand should not be zero");
 
     // Much like with normal add expr, the polynominal will require
     // one less addition than the number of it's terms.
-    Cost = AddCost * (NumTerms - 1);
-
-    // Ignoring constant term (operand 0), how many of the coeffients are u> 1?
-    int NumNonZeroDegreeNonOneTerms =
-        llvm::count_if(S->operands(), [](const SCEV *Op) {
-                         auto *SConst = dyn_cast<SCEVConstant>(Op);
-                         return !SConst || SConst->getAPInt().ugt(1);
-                       });
+    int AddCost = ArithCost(Instruction::Add, NumTerms - 1);
     // Here, *each* one of those will require a multiplication.
-    Cost += MulCost * NumNonZeroDegreeNonOneTerms;
-
+    int MulCost = ArithCost(Instruction::Mul, NumNonZeroDegreeNonOneTerms);
+    Cost = AddCost + MulCost;
     // What is the degree of this polynominal?
     int PolyDegree = S->getNumOperands() - 1;
     assert(PolyDegree >= 1 && "Should be at least affine.");
Index: llvm/test/Transforms/IndVarSimplify/2011-11-01-lftrptr.ll
===================================================================
--- llvm/test/Transforms/IndVarSimplify/2011-11-01-lftrptr.ll
+++ llvm/test/Transforms/IndVarSimplify/2011-11-01-lftrptr.ll
@@ -21,7 +21,7 @@
 ; PTR64:       loop:
 ; PTR64-NEXT:    [[P_01_US_US:%.*]] = phi i8* [ null, [[PREHEADER]] ], [ [[GEP:%.*]], [[LOOP]] ]
 ; PTR64-NEXT:    [[GEP]] = getelementptr inbounds i8, i8* [[P_01_US_US]], i64 1
-; PTR64-NEXT:    [[SNEXT:%.*]] = load i8, i8* [[GEP]]
+; PTR64-NEXT:    [[SNEXT:%.*]] = load i8, i8* [[GEP]], align 1
 ; PTR64-NEXT:    [[EXITCOND:%.*]] = icmp ne i8* [[GEP]], [[END]]
 ; PTR64-NEXT:    br i1 [[EXITCOND]], label [[LOOP]], label [[EXIT_LOOPEXIT:%.*]]
 ; PTR64:       exit.loopexit:
@@ -41,7 +41,7 @@
 ; PTR32:       loop:
 ; PTR32-NEXT:    [[P_01_US_US:%.*]] = phi i8* [ null, [[PREHEADER]] ], [ [[GEP:%.*]], [[LOOP]] ]
 ; PTR32-NEXT:    [[GEP]] = getelementptr inbounds i8, i8* [[P_01_US_US]], i64 1
-; PTR32-NEXT:    [[SNEXT:%.*]] = load i8, i8* [[GEP]]
+; PTR32-NEXT:    [[SNEXT:%.*]] = load i8, i8* [[GEP]], align 1
 ; PTR32-NEXT:    [[EXITCOND:%.*]] = icmp ne i8* [[GEP]], [[END]]
 ; PTR32-NEXT:    br i1 [[EXITCOND]], label [[LOOP]], label [[EXIT_LOOPEXIT:%.*]]
 ; PTR32:       exit.loopexit:
@@ -84,9 +84,9 @@
 ; PTR64:       loop:
 ; PTR64-NEXT:    [[P_01_US_US:%.*]] = phi i8* [ [[BUF]], [[PREHEADER]] ], [ [[GEP:%.*]], [[LOOP]] ]
 ; PTR64-NEXT:    [[GEP]] = getelementptr inbounds i8, i8* [[P_01_US_US]], i64 1
-; PTR64-NEXT:    [[SNEXT:%.*]] = load i8, i8* [[GEP]]
-; PTR64-NEXT:    [[EXITCOND:%.*]] = icmp ne i8* [[GEP]], [[END]]
-; PTR64-NEXT:    br i1 [[EXITCOND]], label [[LOOP]], label [[EXIT_LOOPEXIT:%.*]]
+; PTR64-NEXT:    [[SNEXT:%.*]] = load i8, i8* [[GEP]], align 1
+; PTR64-NEXT:    [[CMP:%.*]] = icmp ult i8* [[GEP]], [[END]]
+; PTR64-NEXT:    br i1 [[CMP]], label [[LOOP]], label [[EXIT_LOOPEXIT:%.*]]
 ; PTR64:       exit.loopexit:
 ; PTR64-NEXT:    [[SNEXT_LCSSA:%.*]] = phi i8 [ [[SNEXT]], [[LOOP]] ]
 ; PTR64-NEXT:    br label [[EXIT]]
@@ -104,9 +104,9 @@
 ; PTR32:       loop:
 ; PTR32-NEXT:    [[P_01_US_US:%.*]] = phi i8* [ [[BUF]], [[PREHEADER]] ], [ [[GEP:%.*]], [[LOOP]] ]
 ; PTR32-NEXT:    [[GEP]] = getelementptr inbounds i8, i8* [[P_01_US_US]], i64 1
-; PTR32-NEXT:    [[SNEXT:%.*]] = load i8, i8* [[GEP]]
-; PTR32-NEXT:    [[EXITCOND:%.*]] = icmp ne i8* [[GEP]], [[END]]
-; PTR32-NEXT:    br i1 [[EXITCOND]], label [[LOOP]], label [[EXIT_LOOPEXIT:%.*]]
+; PTR32-NEXT:    [[SNEXT:%.*]] = load i8, i8* [[GEP]], align 1
+; PTR32-NEXT:    [[CMP:%.*]] = icmp ult i8* [[GEP]], [[END]]
+; PTR32-NEXT:    br i1 [[CMP]], label [[LOOP]], label [[EXIT_LOOPEXIT:%.*]]
 ; PTR32:       exit.loopexit:
 ; PTR32-NEXT:    [[SNEXT_LCSSA:%.*]] = phi i8 [ [[SNEXT]], [[LOOP]] ]
 ; PTR32-NEXT:    br label [[EXIT]]
@@ -146,18 +146,15 @@
 ; PTR64-NEXT:    [[GUARD:%.*]] = icmp ult i32 0, [[CNT]]
 ; PTR64-NEXT:    br i1 [[GUARD]], label [[PREHEADER:%.*]], label [[EXIT:%.*]]
 ; PTR64:       preheader:
-; PTR64-NEXT:    [[TMP1:%.*]] = add i32 [[EI]], -1
-; PTR64-NEXT:    [[TMP2:%.*]] = sub i32 [[TMP1]], [[BI]]
-; PTR64-NEXT:    [[TMP3:%.*]] = zext i32 [[TMP2]] to i64
-; PTR64-NEXT:    [[TMP4:%.*]] = add nuw nsw i64 [[TMP3]], 1
-; PTR64-NEXT:    [[TMP5:%.*]] = inttoptr i64 [[TMP4]] to i8*
 ; PTR64-NEXT:    br label [[LOOP:%.*]]
 ; PTR64:       loop:
 ; PTR64-NEXT:    [[P_01_US_US:%.*]] = phi i8* [ null, [[PREHEADER]] ], [ [[GEP:%.*]], [[LOOP]] ]
+; PTR64-NEXT:    [[IV:%.*]] = phi i32 [ 0, [[PREHEADER]] ], [ [[IVNEXT:%.*]], [[LOOP]] ]
 ; PTR64-NEXT:    [[GEP]] = getelementptr inbounds i8, i8* [[P_01_US_US]], i64 1
-; PTR64-NEXT:    [[SNEXT:%.*]] = load i8, i8* [[GEP]]
-; PTR64-NEXT:    [[EXITCOND:%.*]] = icmp ne i8* [[GEP]], [[TMP5]]
-; PTR64-NEXT:    br i1 [[EXITCOND]], label [[LOOP]], label [[EXIT_LOOPEXIT:%.*]]
+; PTR64-NEXT:    [[SNEXT:%.*]] = load i8, i8* [[GEP]], align 1
+; PTR64-NEXT:    [[IVNEXT]] = add nuw i32 [[IV]], 1
+; PTR64-NEXT:    [[CMP:%.*]] = icmp ult i32 [[IVNEXT]], [[CNT]]
+; PTR64-NEXT:    br i1 [[CMP]], label [[LOOP]], label [[EXIT_LOOPEXIT:%.*]]
 ; PTR64:       exit.loopexit:
 ; PTR64-NEXT:    [[SNEXT_LCSSA:%.*]] = phi i8 [ [[SNEXT]], [[LOOP]] ]
 ; PTR64-NEXT:    br label [[EXIT]]
@@ -171,17 +168,18 @@
 ; PTR32-NEXT:    [[BI:%.*]] = ptrtoint i8* [[BUF:%.*]] to i32
 ; PTR32-NEXT:    [[EI:%.*]] = ptrtoint i8* [[END:%.*]] to i32
 ; PTR32-NEXT:    [[CNT:%.*]] = sub i32 [[EI]], [[BI]]
-; PTR32-NEXT:    [[CNT1:%.*]] = inttoptr i32 [[CNT]] to i8*
 ; PTR32-NEXT:    [[GUARD:%.*]] = icmp ult i32 0, [[CNT]]
 ; PTR32-NEXT:    br i1 [[GUARD]], label [[PREHEADER:%.*]], label [[EXIT:%.*]]
 ; PTR32:       preheader:
 ; PTR32-NEXT:    br label [[LOOP:%.*]]
 ; PTR32:       loop:
 ; PTR32-NEXT:    [[P_01_US_US:%.*]] = phi i8* [ null, [[PREHEADER]] ], [ [[GEP:%.*]], [[LOOP]] ]
+; PTR32-NEXT:    [[IV:%.*]] = phi i32 [ 0, [[PREHEADER]] ], [ [[IVNEXT:%.*]], [[LOOP]] ]
 ; PTR32-NEXT:    [[GEP]] = getelementptr inbounds i8, i8* [[P_01_US_US]], i64 1
-; PTR32-NEXT:    [[SNEXT:%.*]] = load i8, i8* [[GEP]]
-; PTR32-NEXT:    [[EXITCOND:%.*]] = icmp ne i8* [[GEP]], [[CNT1]]
-; PTR32-NEXT:    br i1 [[EXITCOND]], label [[LOOP]], label [[EXIT_LOOPEXIT:%.*]]
+; PTR32-NEXT:    [[SNEXT:%.*]] = load i8, i8* [[GEP]], align 1
+; PTR32-NEXT:    [[IVNEXT]] = add nuw i32 [[IV]], 1
+; PTR32-NEXT:    [[CMP:%.*]] = icmp ult i32 [[IVNEXT]], [[CNT]]
+; PTR32-NEXT:    br i1 [[CMP]], label [[LOOP]], label [[EXIT_LOOPEXIT:%.*]]
 ; PTR32:       exit.loopexit:
 ; PTR32-NEXT:    [[SNEXT_LCSSA:%.*]] = phi i8 [ [[SNEXT]], [[LOOP]] ]
 ; PTR32-NEXT:    br label [[EXIT]]
@@ -231,10 +229,10 @@
 ; PTR64-NEXT:    [[P_01_US_US:%.*]] = phi i8* [ [[BUF]], [[PREHEADER]] ], [ [[GEP:%.*]], [[LOOP]] ]
 ; PTR64-NEXT:    [[IV:%.*]] = phi i32 [ [[BI]], [[PREHEADER]] ], [ [[IVNEXT:%.*]], [[LOOP]] ]
 ; PTR64-NEXT:    [[GEP]] = getelementptr inbounds i8, i8* [[P_01_US_US]], i64 1
-; PTR64-NEXT:    [[SNEXT:%.*]] = load i8, i8* [[GEP]]
+; PTR64-NEXT:    [[SNEXT:%.*]] = load i8, i8* [[GEP]], align 1
 ; PTR64-NEXT:    [[IVNEXT]] = add nuw i32 [[IV]], 1
-; PTR64-NEXT:    [[EXITCOND:%.*]] = icmp ne i32 [[IVNEXT]], [[CNT]]
-; PTR64-NEXT:    br i1 [[EXITCOND]], label [[LOOP]], label [[EXIT_LOOPEXIT:%.*]]
+; PTR64-NEXT:    [[CMP:%.*]] = icmp ult i32 [[IVNEXT]], [[CNT]]
+; PTR64-NEXT:    br i1 [[CMP]], label [[LOOP]], label [[EXIT_LOOPEXIT:%.*]]
 ; PTR64:       exit.loopexit:
 ; PTR64-NEXT:    [[SNEXT_LCSSA:%.*]] = phi i8 [ [[SNEXT]], [[LOOP]] ]
 ; PTR64-NEXT:    br label [[EXIT]]
@@ -256,10 +254,10 @@
 ; PTR32-NEXT:    [[P_01_US_US:%.*]] = phi i8* [ [[BUF]], [[PREHEADER]] ], [ [[GEP:%.*]], [[LOOP]] ]
 ; PTR32-NEXT:    [[IV:%.*]] = phi i32 [ [[BI]], [[PREHEADER]] ], [ [[IVNEXT:%.*]], [[LOOP]] ]
 ; PTR32-NEXT:    [[GEP]] = getelementptr inbounds i8, i8* [[P_01_US_US]], i64 1
-; PTR32-NEXT:    [[SNEXT:%.*]] = load i8, i8* [[GEP]]
+; PTR32-NEXT:    [[SNEXT:%.*]] = load i8, i8* [[GEP]], align 1
 ; PTR32-NEXT:    [[IVNEXT]] = add nuw i32 [[IV]], 1
-; PTR32-NEXT:    [[EXITCOND:%.*]] = icmp ne i32 [[IVNEXT]], [[CNT]]
-; PTR32-NEXT:    br i1 [[EXITCOND]], label [[LOOP]], label [[EXIT_LOOPEXIT:%.*]]
+; PTR32-NEXT:    [[CMP:%.*]] = icmp ult i32 [[IVNEXT]], [[CNT]]
+; PTR32-NEXT:    br i1 [[CMP]], label [[LOOP]], label [[EXIT_LOOPEXIT:%.*]]
 ; PTR32:       exit.loopexit:
 ; PTR32-NEXT:    [[SNEXT_LCSSA:%.*]] = phi i8 [ [[SNEXT]], [[LOOP]] ]
 ; PTR32-NEXT:    br label [[EXIT]]
Index: llvm/test/Transforms/IndVarSimplify/X86/loop-invariant-conditions.ll
===================================================================
--- llvm/test/Transforms/IndVarSimplify/X86/loop-invariant-conditions.ll
+++ llvm/test/Transforms/IndVarSimplify/X86/loop-invariant-conditions.ll
@@ -193,7 +193,7 @@
 define void @test7(i64 %start, i64* %inc_ptr) {
 ; CHECK-LABEL: @test7(
 ; CHECK-NEXT:  entry:
-; CHECK-NEXT:    [[INC:%.*]] = load i64, i64* [[INC_PTR:%.*]], !range !0
+; CHECK-NEXT:    [[INC:%.*]] = load i64, i64* [[INC_PTR:%.*]], align 8, [[RNG0:!range !.*]]
 ; CHECK-NEXT:    [[OK:%.*]] = icmp sge i64 [[INC]], 0
 ; CHECK-NEXT:    br i1 [[OK]], label [[LOOP_PREHEADER:%.*]], label [[FOR_END:%.*]]
 ; CHECK:       loop.preheader:
@@ -311,15 +311,12 @@
 define void @test3_neg(i64 %start) {
 ; CHECK-LABEL: @test3_neg(
 ; CHECK-NEXT:  entry:
-; CHECK-NEXT:    [[TMP0:%.*]] = icmp sgt i64 [[START:%.*]], -1
-; CHECK-NEXT:    [[SMAX:%.*]] = select i1 [[TMP0]], i64 [[START]], i64 -1
-; CHECK-NEXT:    [[TMP1:%.*]] = add i64 [[SMAX]], 1
 ; CHECK-NEXT:    br label [[LOOP:%.*]]
 ; CHECK:       loop:
-; CHECK-NEXT:    [[INDVARS_IV:%.*]] = phi i64 [ [[START]], [[ENTRY:%.*]] ], [ [[INDVARS_IV_NEXT:%.*]], [[LOOP]] ]
-; CHECK-NEXT:    [[INDVARS_IV_NEXT]] = add i64 [[INDVARS_IV]], 1
-; CHECK-NEXT:    [[EXITCOND:%.*]] = icmp ne i64 [[INDVARS_IV_NEXT]], [[TMP1]]
-; CHECK-NEXT:    br i1 [[EXITCOND]], label [[LOOP]], label [[FOR_END:%.*]]
+; CHECK-NEXT:    [[INDVARS_IV:%.*]] = phi i64 [ [[START:%.*]], [[ENTRY:%.*]] ], [ [[INDVARS_IV_NEXT:%.*]], [[LOOP]] ]
+; CHECK-NEXT:    [[INDVARS_IV_NEXT]] = add nsw i64 [[INDVARS_IV]], 1
+; CHECK-NEXT:    [[CMP1:%.*]] = icmp slt i64 [[INDVARS_IV]], -1
+; CHECK-NEXT:    br i1 [[CMP1]], label [[LOOP]], label [[FOR_END:%.*]]
 ; CHECK:       for.end:
 ; CHECK-NEXT:    ret void
 ;
@@ -339,19 +336,16 @@
 define void @test4_neg(i64 %start) {
 ; CHECK-LABEL: @test4_neg(
 ; CHECK-NEXT:  entry:
-; CHECK-NEXT:    [[TMP0:%.*]] = icmp sgt i64 [[START:%.*]], 0
-; CHECK-NEXT:    [[SMAX:%.*]] = select i1 [[TMP0]], i64 [[START]], i64 0
-; CHECK-NEXT:    [[TMP1:%.*]] = add nuw i64 [[SMAX]], 1
 ; CHECK-NEXT:    br label [[LOOP:%.*]]
 ; CHECK:       loop:
-; CHECK-NEXT:    [[INDVARS_IV:%.*]] = phi i64 [ [[START]], [[ENTRY:%.*]] ], [ [[INDVARS_IV_NEXT:%.*]], [[BACKEDGE:%.*]] ]
-; CHECK-NEXT:    [[INDVARS_IV_NEXT]] = add i64 [[INDVARS_IV]], 1
+; CHECK-NEXT:    [[INDVARS_IV:%.*]] = phi i64 [ [[START:%.*]], [[ENTRY:%.*]] ], [ [[INDVARS_IV_NEXT:%.*]], [[BACKEDGE:%.*]] ]
+; CHECK-NEXT:    [[INDVARS_IV_NEXT]] = add nsw i64 [[INDVARS_IV]], 1
 ; CHECK-NEXT:    [[CMP:%.*]] = icmp eq i64 [[INDVARS_IV_NEXT]], 25
 ; CHECK-NEXT:    br i1 [[CMP]], label [[BACKEDGE]], label [[FOR_END:%.*]]
 ; CHECK:       backedge:
 ; CHECK-NEXT:    call void @foo()
-; CHECK-NEXT:    [[EXITCOND:%.*]] = icmp eq i64 [[INDVARS_IV_NEXT]], [[TMP1]]
-; CHECK-NEXT:    br i1 [[EXITCOND]], label [[FOR_END]], label [[LOOP]]
+; CHECK-NEXT:    [[CMP1:%.*]] = icmp sgt i64 [[INDVARS_IV]], -1
+; CHECK-NEXT:    br i1 [[CMP1]], label [[FOR_END]], label [[LOOP]]
 ; CHECK:       for.end:
 ; CHECK-NEXT:    ret void
 ;
@@ -405,7 +399,7 @@
 define void @test8(i64 %start, i64* %inc_ptr) {
 ; CHECK-LABEL: @test8(
 ; CHECK-NEXT:  entry:
-; CHECK-NEXT:    [[INC:%.*]] = load i64, i64* [[INC_PTR:%.*]], !range !1
+; CHECK-NEXT:    [[INC:%.*]] = load i64, i64* [[INC_PTR:%.*]], align 8, [[RNG1:!range !.*]]
 ; CHECK-NEXT:    [[OK:%.*]] = icmp sge i64 [[INC]], 0
 ; CHECK-NEXT:    br i1 [[OK]], label [[LOOP_PREHEADER:%.*]], label [[FOR_END:%.*]]
 ; CHECK:       loop.preheader:
@@ -525,7 +519,7 @@
 define void @test11(i64* %inc_ptr) {
 ; CHECK-LABEL: @test11(
 ; CHECK-NEXT:  entry:
-; CHECK-NEXT:    [[INC:%.*]] = load i64, i64* [[INC_PTR:%.*]], !range !0
+; CHECK-NEXT:    [[INC:%.*]] = load i64, i64* [[INC_PTR:%.*]], align 8, [[RNG0]]
 ; CHECK-NEXT:    [[NE_COND:%.*]] = icmp ne i64 [[INC]], 0
 ; CHECK-NEXT:    br i1 [[NE_COND]], label [[LOOP_PREHEADER:%.*]], label [[EXIT:%.*]]
 ; CHECK:       loop.preheader:
@@ -576,7 +570,7 @@
 define void @test12(i64* %inc_ptr) {
 ; CHECK-LABEL: @test12(
 ; CHECK-NEXT:  entry:
-; CHECK-NEXT:    [[INC:%.*]] = load i64, i64* [[INC_PTR:%.*]], !range !0
+; CHECK-NEXT:    [[INC:%.*]] = load i64, i64* [[INC_PTR:%.*]], align 8, [[RNG0]]
 ; CHECK-NEXT:    br label [[LOOP:%.*]]
 ; CHECK:       loop:
 ; CHECK-NEXT:    [[IV:%.*]] = phi i64 [ [[INC]], [[ENTRY:%.*]] ], [ [[IV_NEXT:%.*]], [[BACKEDGE:%.*]] ]
Index: llvm/test/Transforms/IndVarSimplify/elim-extend.ll
===================================================================
--- llvm/test/Transforms/IndVarSimplify/elim-extend.ll
+++ llvm/test/Transforms/IndVarSimplify/elim-extend.ll
@@ -63,7 +63,7 @@
 ; CHECK-NEXT:    br i1 [[PRECOND]], label [[LOOP_PREHEADER:%.*]], label [[RETURN:%.*]]
 ; CHECK:       loop.preheader:
 ; CHECK-NEXT:    [[TMP0:%.*]] = sext i32 [[INIT]] to i64
-; CHECK-NEXT:    [[WIDE_TRIP_COUNT:%.*]] = sext i32 [[LIMIT]] to i64
+; CHECK-NEXT:    [[TMP1:%.*]] = sext i32 [[LIMIT]] to i64
 ; CHECK-NEXT:    br label [[LOOP:%.*]]
 ; CHECK:       loop:
 ; CHECK-NEXT:    [[INDVARS_IV:%.*]] = phi i64 [ [[TMP0]], [[LOOP_PREHEADER]] ], [ [[INDVARS_IV_NEXT:%.*]], [[LOOP]] ]
@@ -74,8 +74,8 @@
 ; CHECK-NEXT:    store i8 0, i8* [[POSTADR]], align 1
 ; CHECK-NEXT:    [[POSTADRNSW:%.*]] = getelementptr i8, i8* [[BASE]], i64 [[INDVARS_IV_NEXT]]
 ; CHECK-NEXT:    store i8 0, i8* [[POSTADRNSW]], align 1
-; CHECK-NEXT:    [[EXITCOND:%.*]] = icmp ne i64 [[INDVARS_IV_NEXT]], [[WIDE_TRIP_COUNT]]
-; CHECK-NEXT:    br i1 [[EXITCOND]], label [[LOOP]], label [[EXIT:%.*]]
+; CHECK-NEXT:    [[COND:%.*]] = icmp sgt i64 [[TMP1]], [[INDVARS_IV_NEXT]]
+; CHECK-NEXT:    br i1 [[COND]], label [[LOOP]], label [[EXIT:%.*]]
 ; CHECK:       exit:
 ; CHECK-NEXT:    br label [[RETURN]]
 ; CHECK:       return:
@@ -140,8 +140,8 @@
 ; CHECK-NEXT:    store i8 0, i8* [[ADR2]], align 1
 ; CHECK-NEXT:    [[ADR3:%.*]] = getelementptr i8, i8* [[ADDRESS]], i64 [[INDVARS_IV_NEXT]]
 ; CHECK-NEXT:    store i8 0, i8* [[ADR3]], align 1
-; CHECK-NEXT:    [[EXITCOND:%.*]] = icmp ne i64 [[INDVARS_IV_NEXT]], [[TMP0]]
-; CHECK-NEXT:    br i1 [[EXITCOND]], label [[INNERLOOP]], label [[INNEREXIT:%.*]]
+; CHECK-NEXT:    [[INNERCMP:%.*]] = icmp sgt i64 [[TMP0]], [[INDVARS_IV_NEXT]]
+; CHECK-NEXT:    br i1 [[INNERCMP]], label [[INNERLOOP]], label [[INNEREXIT:%.*]]
 ; CHECK:       innerexit:
 ; CHECK-NEXT:    [[INNERCOUNT_LCSSA_WIDE:%.*]] = phi i64 [ [[INDVARS_IV_NEXT]], [[INNERLOOP]] ]
 ; CHECK-NEXT:    [[TMP4:%.*]] = trunc i64 [[INNERCOUNT_LCSSA_WIDE]] to i32
@@ -154,8 +154,8 @@
 ; CHECK-NEXT:    [[ADR5:%.*]] = getelementptr i8, i8* [[ADDRESS]], i64 [[OFS5]]
 ; CHECK-NEXT:    store i8 0, i8* [[ADR5]], align 1
 ; CHECK-NEXT:    [[INDVARS_IV_NEXT2]] = add nuw nsw i64 [[INDVARS_IV1]], 1
-; CHECK-NEXT:    [[EXITCOND4:%.*]] = icmp ne i64 [[INDVARS_IV_NEXT2]], [[WIDE_TRIP_COUNT]]
-; CHECK-NEXT:    br i1 [[EXITCOND4]], label [[OUTERLOOP]], label [[RETURN:%.*]]
+; CHECK-NEXT:    [[EXITCOND:%.*]] = icmp ne i64 [[INDVARS_IV_NEXT2]], [[WIDE_TRIP_COUNT]]
+; CHECK-NEXT:    br i1 [[EXITCOND]], label [[OUTERLOOP]], label [[RETURN:%.*]]
 ; CHECK:       return:
 ; CHECK-NEXT:    ret void
 ;
Index: llvm/test/Transforms/IndVarSimplify/lftr.ll
===================================================================
--- llvm/test/Transforms/IndVarSimplify/lftr.ll
+++ llvm/test/Transforms/IndVarSimplify/lftr.ll
@@ -14,7 +14,7 @@
 ; CHECK:       loop:
 ; CHECK-NEXT:    [[I:%.*]] = phi i32 [ 0, [[ENTRY:%.*]] ], [ [[I_NEXT:%.*]], [[LOOP]] ]
 ; CHECK-NEXT:    [[I_NEXT]] = add nuw nsw i32 [[I]], 1
-; CHECK-NEXT:    store i32 [[I]], i32* @A
+; CHECK-NEXT:    store i32 [[I]], i32* @A, align 4
 ; CHECK-NEXT:    [[EXITCOND:%.*]] = icmp ne i32 [[I_NEXT]], 1001
 ; CHECK-NEXT:    br i1 [[EXITCOND]], label [[LOOP]], label [[LOOPEXIT:%.*]]
 ; CHECK:       loopexit:
@@ -42,7 +42,7 @@
 ; CHECK:       loop:
 ; CHECK-NEXT:    [[I:%.*]] = phi i32 [ 1000, [[ENTRY:%.*]] ], [ [[I_NEXT:%.*]], [[LOOP]] ]
 ; CHECK-NEXT:    [[I_NEXT]] = sub nsw i32 [[I]], 1
-; CHECK-NEXT:    store i32 [[I]], i32* @A
+; CHECK-NEXT:    store i32 [[I]], i32* @A, align 4
 ; CHECK-NEXT:    [[C:%.*]] = icmp ugt i32 [[I]], 0
 ; CHECK-NEXT:    br i1 [[C]], label [[LOOP]], label [[LOOPEXIT:%.*]]
 ; CHECK:       loopexit:
@@ -73,7 +73,7 @@
 ; CHECK:       loop:
 ; CHECK-NEXT:    [[I:%.*]] = phi i32 [ 7, [[ENTRY:%.*]] ], [ [[I_NEXT:%.*]], [[LOOP]] ]
 ; CHECK-NEXT:    [[I_NEXT]] = add nuw nsw i32 [[I]], 1
-; CHECK-NEXT:    store i32 [[I]], i32* @A
+; CHECK-NEXT:    store i32 [[I]], i32* @A, align 4
 ; CHECK-NEXT:    [[EXITCOND:%.*]] = icmp ne i32 [[I_NEXT]], 33
 ; CHECK-NEXT:    br i1 [[EXITCOND]], label [[LOOP]], label [[LOOPEXIT:%.*]]
 ; CHECK:       loopexit:
@@ -103,7 +103,7 @@
 ; CHECK:       loop:
 ; CHECK-NEXT:    [[I:%.*]] = phi i32 [ 7, [[ENTRY:%.*]] ], [ [[I_NEXT:%.*]], [[LOOP]] ]
 ; CHECK-NEXT:    [[I_NEXT]] = add nuw nsw i32 [[I]], 1
-; CHECK-NEXT:    store i32 [[I]], i32* @A
+; CHECK-NEXT:    store i32 [[I]], i32* @A, align 4
 ; CHECK-NEXT:    [[EXITCOND:%.*]] = icmp ne i32 [[I_NEXT]], 33
 ; CHECK-NEXT:    br i1 [[EXITCOND]], label [[LOOP]], label [[LOOPEXIT:%.*]]
 ; CHECK:       loopexit:
@@ -132,7 +132,7 @@
 ; CHECK:       loop:
 ; CHECK-NEXT:    [[I:%.*]] = phi i32 [ 7, [[ENTRY:%.*]] ], [ [[I_NEXT:%.*]], [[LOOP]] ]
 ; CHECK-NEXT:    [[I_NEXT]] = add nuw nsw i32 [[I]], 1
-; CHECK-NEXT:    store i32 [[I]], i32* @A
+; CHECK-NEXT:    store i32 [[I]], i32* @A, align 4
 ; CHECK-NEXT:    [[EXITCOND:%.*]] = icmp ne i32 [[I_NEXT]], 33
 ; CHECK-NEXT:    br i1 [[EXITCOND]], label [[LOOP]], label [[LOOPEXIT:%.*]]
 ; CHECK:       loopexit:
@@ -160,12 +160,13 @@
 ; CHECK:       loop:
 ; CHECK-NEXT:    [[I:%.*]] = phi i32 [ 10, [[ENTRY:%.*]] ], [ [[I_NEXT:%.*]], [[LOOP]] ]
 ; CHECK-NEXT:    [[I_NEXT]] = call i32 @llvm.loop.decrement.reg.i32(i32 [[I]], i32 1)
-; CHECK-NEXT:    store i32 [[I]], i32* @A
+; CHECK-NEXT:    store i32 [[I]], i32* @A, align 4
 ; CHECK-NEXT:    [[I2:%.*]] = mul i32 [[I]], [[I]]
-; CHECK-NEXT:    [[EXITCOND:%.*]] = icmp sgt i32 [[I2]], 0
-; CHECK-NEXT:    br i1 [[EXITCOND]], label [[LOOP]], label [[LOOPEXIT:%.*]]
+; CHECK-NEXT:    [[C:%.*]] = icmp sgt i32 [[I2]], 0
+; CHECK-NEXT:    br i1 [[C]], label [[LOOP]], label [[LOOPEXIT:%.*]]
 ; CHECK:       loopexit:
 ; CHECK-NEXT:    ret i32 0
+;
 
 entry:
   br label %loop
@@ -235,7 +236,7 @@
 ; CHECK:       loop:
 ; CHECK-NEXT:    [[I1:%.*]] = phi i8 [ [[I1_INC:%.*]], [[LOOP]] ], [ 0, [[LOOP_PREHEADER]] ]
 ; CHECK-NEXT:    [[I1_INC]] = add nuw nsw i8 [[I1]], 1
-; CHECK-NEXT:    store volatile i8 0, i8* [[A:%.*]]
+; CHECK-NEXT:    store volatile i8 0, i8* [[A:%.*]], align 1
 ; CHECK-NEXT:    [[EXITCOND:%.*]] = icmp ne i8 [[I1_INC]], [[TMP2]]
 ; CHECK-NEXT:    br i1 [[EXITCOND]], label [[LOOP]], label [[EXIT_LOOPEXIT:%.*]]
 ; CHECK:       exit.loopexit:
@@ -335,7 +336,7 @@
 ; CHECK-NEXT:    br label [[FOR_BODY:%.*]]
 ; CHECK:       for.body:
 ; CHECK-NEXT:    [[INDVARS_IV:%.*]] = phi i32 [ [[INDVARS_IV_NEXT:%.*]], [[FOR_BODY]] ], [ 0, [[ENTRY:%.*]] ]
-; CHECK-NEXT:    call void @bar(i32 [[INDVARS_IV]]) #2
+; CHECK-NEXT:    call void @bar(i32 [[INDVARS_IV]]) [[ATTR2:#.*]]
 ; CHECK-NEXT:    [[INDVARS_IV_NEXT]] = add nuw nsw i32 [[INDVARS_IV]], 1
 ; CHECK-NEXT:    [[EXITCOND:%.*]] = icmp ne i32 [[INDVARS_IV_NEXT]], 512
 ; CHECK-NEXT:    br i1 [[EXITCOND]], label [[FOR_BODY]], label [[FOR_END:%.*]]
@@ -364,7 +365,7 @@
 ; CHECK-NEXT:    br label [[DO_BODY:%.*]]
 ; CHECK:       do.body:
 ; CHECK-NEXT:    [[INDVARS_IV:%.*]] = phi i32 [ [[INDVARS_IV_NEXT:%.*]], [[DO_BODY]] ], [ 0, [[ENTRY:%.*]] ]
-; CHECK-NEXT:    call void @bar(i32 [[INDVARS_IV]]) #2
+; CHECK-NEXT:    call void @bar(i32 [[INDVARS_IV]]) [[ATTR2]]
 ; CHECK-NEXT:    [[CMP:%.*]] = icmp eq i32 [[INDVARS_IV]], 255
 ; CHECK-NEXT:    [[INDVARS_IV_NEXT]] = add nuw nsw i32 [[INDVARS_IV]], 1
 ; CHECK-NEXT:    br i1 [[CMP]], label [[DO_END:%.*]], label [[DO_BODY]]
@@ -414,9 +415,9 @@
 ; CHECK-NEXT:    [[ADD3:%.*]] = fadd float [[TEMP2]], [[MUL]]
 ; CHECK-NEXT:    store float [[ADD3]], float* [[ARRAYIDX2]], align 4
 ; CHECK-NEXT:    [[INDVARS_IV_NEXT]] = add nuw nsw i64 [[INDVARS_IV]], 1
-; CHECK-NEXT:    [[LFTR_WIDEIV:%.*]] = trunc i64 [[INDVARS_IV_NEXT]] to i32
-; CHECK-NEXT:    [[EXITCOND:%.*]] = icmp ne i32 [[LFTR_WIDEIV]], [[SUB]]
-; CHECK-NEXT:    br i1 [[EXITCOND]], label [[FOR_BODY]], label [[FOR_END_LOOPEXIT:%.*]]
+; CHECK-NEXT:    [[TEMP3:%.*]] = trunc i64 [[INDVARS_IV_NEXT]] to i32
+; CHECK-NEXT:    [[CMP:%.*]] = icmp ult i32 [[TEMP3]], [[SUB]]
+; CHECK-NEXT:    br i1 [[CMP]], label [[FOR_BODY]], label [[FOR_END_LOOPEXIT:%.*]]
 ; CHECK:       for.end.loopexit:
 ; CHECK-NEXT:    br label [[FOR_END]]
 ; CHECK:       for.end:
Index: llvm/test/Transforms/IndVarSimplify/ult-sub-to-eq.ll
===================================================================
--- llvm/test/Transforms/IndVarSimplify/ult-sub-to-eq.ll
+++ llvm/test/Transforms/IndVarSimplify/ult-sub-to-eq.ll
@@ -12,7 +12,7 @@
 ; CHECK-NEXT:    [[CMP4:%.*]] = icmp eq i32 [[DATA_LEN]], [[SAMPLE]]
 ; CHECK-NEXT:    br i1 [[CMP4]], label [[FOR_END:%.*]], label [[FOR_BODY_PREHEADER:%.*]]
 ; CHECK:       for.body.preheader:
-; CHECK-NEXT:    [[WIDE_TRIP_COUNT:%.*]] = zext i32 [[SUB]] to i64
+; CHECK-NEXT:    [[ZEXT:%.*]] = zext i32 [[SUB]] to i64
 ; CHECK-NEXT:    br label [[FOR_BODY:%.*]]
 ; CHECK:       for.body:
 ; CHECK-NEXT:    [[INDVARS_IV:%.*]] = phi i64 [ [[INDVARS_IV_NEXT:%.*]], [[FOR_BODY]] ], [ 0, [[FOR_BODY_PREHEADER]] ]
@@ -27,8 +27,8 @@
 ; CHECK-NEXT:    [[ADD3:%.*]] = fadd float [[TMP2]], [[MUL]]
 ; CHECK-NEXT:    store float [[ADD3]], float* [[ARRAYIDX2]], align 4
 ; CHECK-NEXT:    [[INDVARS_IV_NEXT]] = add nuw nsw i64 [[INDVARS_IV]], 1
-; CHECK-NEXT:    [[EXITCOND:%.*]] = icmp ne i64 [[INDVARS_IV_NEXT]], [[WIDE_TRIP_COUNT]]
-; CHECK-NEXT:    br i1 [[EXITCOND]], label [[FOR_BODY]], label [[FOR_END_LOOPEXIT:%.*]]
+; CHECK-NEXT:    [[TMP3:%.*]] = icmp ult i64 [[INDVARS_IV_NEXT]], [[ZEXT]]
+; CHECK-NEXT:    br i1 [[TMP3]], label [[FOR_BODY]], label [[FOR_END_LOOPEXIT:%.*]]
 ; CHECK:       for.end.loopexit:
 ; CHECK-NEXT:    br label [[FOR_END]]
 ; CHECK:       for.end:
Index: llvm/test/Transforms/IndVarSimplify/widen-loop-comp.ll
===================================================================
--- llvm/test/Transforms/IndVarSimplify/widen-loop-comp.ll
+++ llvm/test/Transforms/IndVarSimplify/widen-loop-comp.ll
@@ -457,15 +457,13 @@
 ; CHECK-NEXT:    br i1 [[E]], label [[FOR_COND_PREHEADER:%.*]], label [[LEAVE:%.*]]
 ; CHECK:       for.cond.preheader:
 ; CHECK-NEXT:    [[TMP0:%.*]] = zext i32 [[INIT]] to i64
-; CHECK-NEXT:    [[TMP1:%.*]] = icmp sgt i32 [[INIT]], [[B:%.*]]
-; CHECK-NEXT:    [[SMAX:%.*]] = select i1 [[TMP1]], i32 [[INIT]], i32 [[B]]
-; CHECK-NEXT:    [[WIDE_TRIP_COUNT:%.*]] = zext i32 [[SMAX]] to i64
+; CHECK-NEXT:    [[TMP1:%.*]] = sext i32 [[B:%.*]] to i64
 ; CHECK-NEXT:    br label [[FOR_COND:%.*]]
 ; CHECK:       for.cond:
 ; CHECK-NEXT:    [[INDVARS_IV:%.*]] = phi i64 [ [[TMP0]], [[FOR_COND_PREHEADER]] ], [ [[INDVARS_IV_NEXT:%.*]], [[FOR_BODY:%.*]] ]
 ; CHECK-NEXT:    [[SUM_0:%.*]] = phi i32 [ [[ADD:%.*]], [[FOR_BODY]] ], [ 0, [[FOR_COND_PREHEADER]] ]
-; CHECK-NEXT:    [[EXITCOND:%.*]] = icmp ne i64 [[INDVARS_IV]], [[WIDE_TRIP_COUNT]]
-; CHECK-NEXT:    br i1 [[EXITCOND]], label [[FOR_BODY]], label [[FOR_END:%.*]]
+; CHECK-NEXT:    [[CMP:%.*]] = icmp slt i64 [[INDVARS_IV]], [[TMP1]]
+; CHECK-NEXT:    br i1 [[CMP]], label [[FOR_BODY]], label [[FOR_END:%.*]]
 ; CHECK:       for.body:
 ; CHECK-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds i32, i32* [[A:%.*]], i64 [[INDVARS_IV]]
 ; CHECK-NEXT:    [[TMP2:%.*]] = load i32, i32* [[ARRAYIDX]], align 4
Index: llvm/test/Transforms/LoopUnroll/runtime-small-upperbound.ll
===================================================================
--- llvm/test/Transforms/LoopUnroll/runtime-small-upperbound.ll
+++ llvm/test/Transforms/LoopUnroll/runtime-small-upperbound.ll
@@ -100,63 +100,15 @@
 ; UPPER-NEXT:    [[TMP0:%.*]] = icmp ult i32 [[X]], 17
 ; UPPER-NEXT:    br i1 [[TMP0]], label [[LOOP_PREHEADER:%.*]], label [[EXIT:%.*]]
 ; UPPER:       loop.preheader:
-; UPPER-NEXT:    [[TMP1:%.*]] = sub i32 16, [[X]]
-; UPPER-NEXT:    [[TMP2:%.*]] = lshr i32 [[TMP1]], 2
-; UPPER-NEXT:    [[TMP3:%.*]] = add nuw nsw i32 [[TMP2]], 1
-; UPPER-NEXT:    [[TMP4:%.*]] = urem i32 [[TMP2]], 6
-; UPPER-NEXT:    [[TMP5:%.*]] = add i32 [[TMP4]], 1
-; UPPER-NEXT:    [[XTRAITER:%.*]] = urem i32 [[TMP5]], 6
-; UPPER-NEXT:    [[LCMP_MOD:%.*]] = icmp ne i32 [[XTRAITER]], 0
-; UPPER-NEXT:    br i1 [[LCMP_MOD]], label [[LOOP_PROL_PREHEADER:%.*]], label [[LOOP_PROL_LOOPEXIT:%.*]]
-; UPPER:       loop.prol.preheader:
-; UPPER-NEXT:    br label [[LOOP_PROL:%.*]]
-; UPPER:       loop.prol:
-; UPPER-NEXT:    [[IV_PROL:%.*]] = phi i32 [ [[IV_NEXT_PROL:%.*]], [[LOOP_PROL]] ], [ [[X]], [[LOOP_PROL_PREHEADER]] ]
-; UPPER-NEXT:    [[PTR_PROL:%.*]] = phi i8* [ [[PTR_NEXT_PROL:%.*]], [[LOOP_PROL]] ], [ [[Y]], [[LOOP_PROL_PREHEADER]] ]
-; UPPER-NEXT:    [[PROL_ITER:%.*]] = phi i32 [ [[XTRAITER]], [[LOOP_PROL_PREHEADER]] ], [ [[PROL_ITER_SUB:%.*]], [[LOOP_PROL]] ]
-; UPPER-NEXT:    [[IV_NEXT_PROL]] = add nuw i32 [[IV_PROL]], 4
-; UPPER-NEXT:    [[PTR_NEXT_PROL]] = getelementptr inbounds i8, i8* [[PTR_PROL]], i32 1
-; UPPER-NEXT:    store i8 [[ARG:%.*]], i8* [[PTR_NEXT_PROL]], align 1
-; UPPER-NEXT:    [[TMP6:%.*]] = icmp ult i32 [[IV_NEXT_PROL]], 17
-; UPPER-NEXT:    [[PROL_ITER_SUB]] = sub i32 [[PROL_ITER]], 1
-; UPPER-NEXT:    [[PROL_ITER_CMP:%.*]] = icmp ne i32 [[PROL_ITER_SUB]], 0
-; UPPER-NEXT:    br i1 [[PROL_ITER_CMP]], label [[LOOP_PROL]], label [[LOOP_PROL_LOOPEXIT_UNR_LCSSA:%.*]], [[LOOP0:!llvm.loop !.*]]
-; UPPER:       loop.prol.loopexit.unr-lcssa:
-; UPPER-NEXT:    [[IV_UNR_PH:%.*]] = phi i32 [ [[IV_NEXT_PROL]], [[LOOP_PROL]] ]
-; UPPER-NEXT:    [[PTR_UNR_PH:%.*]] = phi i8* [ [[PTR_NEXT_PROL]], [[LOOP_PROL]] ]
-; UPPER-NEXT:    br label [[LOOP_PROL_LOOPEXIT]]
-; UPPER:       loop.prol.loopexit:
-; UPPER-NEXT:    [[IV_UNR:%.*]] = phi i32 [ [[X]], [[LOOP_PREHEADER]] ], [ [[IV_UNR_PH]], [[LOOP_PROL_LOOPEXIT_UNR_LCSSA]] ]
-; UPPER-NEXT:    [[PTR_UNR:%.*]] = phi i8* [ [[Y]], [[LOOP_PREHEADER]] ], [ [[PTR_UNR_PH]], [[LOOP_PROL_LOOPEXIT_UNR_LCSSA]] ]
-; UPPER-NEXT:    [[TMP7:%.*]] = icmp ult i32 [[TMP2]], 5
-; UPPER-NEXT:    br i1 [[TMP7]], label [[EXIT_LOOPEXIT:%.*]], label [[LOOP_PREHEADER_NEW:%.*]]
-; UPPER:       loop.preheader.new:
 ; UPPER-NEXT:    br label [[LOOP:%.*]]
 ; UPPER:       loop:
-; UPPER-NEXT:    [[IV:%.*]] = phi i32 [ [[IV_UNR]], [[LOOP_PREHEADER_NEW]] ], [ [[IV_NEXT_5:%.*]], [[LOOP]] ]
-; UPPER-NEXT:    [[PTR:%.*]] = phi i8* [ [[PTR_UNR]], [[LOOP_PREHEADER_NEW]] ], [ [[PTR_NEXT_5:%.*]], [[LOOP]] ]
-; UPPER-NEXT:    [[IV_NEXT:%.*]] = add nuw i32 [[IV]], 4
-; UPPER-NEXT:    [[PTR_NEXT:%.*]] = getelementptr inbounds i8, i8* [[PTR]], i32 1
-; UPPER-NEXT:    store i8 [[ARG]], i8* [[PTR_NEXT]], align 1
-; UPPER-NEXT:    [[IV_NEXT_1:%.*]] = add nuw i32 [[IV_NEXT]], 4
-; UPPER-NEXT:    [[PTR_NEXT_1:%.*]] = getelementptr inbounds i8, i8* [[PTR_NEXT]], i32 1
-; UPPER-NEXT:    store i8 [[ARG]], i8* [[PTR_NEXT_1]], align 1
-; UPPER-NEXT:    [[IV_NEXT_2:%.*]] = add nuw i32 [[IV_NEXT_1]], 4
-; UPPER-NEXT:    [[PTR_NEXT_2:%.*]] = getelementptr inbounds i8, i8* [[PTR_NEXT_1]], i32 1
-; UPPER-NEXT:    store i8 [[ARG]], i8* [[PTR_NEXT_2]], align 1
-; UPPER-NEXT:    [[IV_NEXT_3:%.*]] = add nuw i32 [[IV_NEXT_2]], 4
-; UPPER-NEXT:    [[PTR_NEXT_3:%.*]] = getelementptr inbounds i8, i8* [[PTR_NEXT_2]], i32 1
-; UPPER-NEXT:    store i8 [[ARG]], i8* [[PTR_NEXT_3]], align 1
-; UPPER-NEXT:    [[IV_NEXT_4:%.*]] = add nuw i32 [[IV_NEXT_3]], 4
-; UPPER-NEXT:    [[PTR_NEXT_4:%.*]] = getelementptr inbounds i8, i8* [[PTR_NEXT_3]], i32 1
-; UPPER-NEXT:    store i8 [[ARG]], i8* [[PTR_NEXT_4]], align 1
-; UPPER-NEXT:    [[IV_NEXT_5]] = add nuw i32 [[IV_NEXT_4]], 4
-; UPPER-NEXT:    [[PTR_NEXT_5]] = getelementptr inbounds i8, i8* [[PTR_NEXT_4]], i32 1
-; UPPER-NEXT:    store i8 [[ARG]], i8* [[PTR_NEXT_5]], align 1
-; UPPER-NEXT:    [[TMP8:%.*]] = icmp ult i32 [[IV_NEXT_5]], 17
-; UPPER-NEXT:    br i1 [[TMP8]], label [[LOOP]], label [[EXIT_LOOPEXIT_UNR_LCSSA:%.*]]
-; UPPER:       exit.loopexit.unr-lcssa:
-; UPPER-NEXT:    br label [[EXIT_LOOPEXIT]]
+; UPPER-NEXT:    [[IV:%.*]] = phi i32 [ [[IV_NEXT:%.*]], [[LOOP]] ], [ [[X]], [[LOOP_PREHEADER]] ]
+; UPPER-NEXT:    [[PTR:%.*]] = phi i8* [ [[PTR_NEXT:%.*]], [[LOOP]] ], [ [[Y]], [[LOOP_PREHEADER]] ]
+; UPPER-NEXT:    [[IV_NEXT]] = add nuw i32 [[IV]], 4
+; UPPER-NEXT:    [[PTR_NEXT]] = getelementptr inbounds i8, i8* [[PTR]], i32 1
+; UPPER-NEXT:    store i8 [[ARG:%.*]], i8* [[PTR_NEXT]], align 1
+; UPPER-NEXT:    [[TMP1:%.*]] = icmp ult i32 [[IV_NEXT]], 17
+; UPPER-NEXT:    br i1 [[TMP1]], label [[LOOP]], label [[EXIT_LOOPEXIT:%.*]]
 ; UPPER:       exit.loopexit:
 ; UPPER-NEXT:    br label [[EXIT]]
 ; UPPER:       exit: