diff --git a/test/Transforms/Mem2Reg/debug-alloca-vla-1.ll b/test/Transforms/Mem2Reg/debug-alloca-vla-1.ll
index 6a2419c03e4..283c27e7b29 100644
--- a/test/Transforms/Mem2Reg/debug-alloca-vla-1.ll
+++ b/test/Transforms/Mem2Reg/debug-alloca-vla-1.ll
@@ -1,3 +1,4 @@
+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
 ; RUN: opt < %s -mem2reg -S | FileCheck %s
 
 ; Testing conversion from dbg.declare to dbg.value when the variable is a VLA.
@@ -14,21 +15,25 @@ target triple = "x86_64-apple-macosx10.12.0"
 
 ; Function Attrs: nounwind ssp uwtable
 define void @scan() #0 !dbg !4 {
+; CHECK-LABEL: @scan(
+; CHECK-NEXT:  entry:
+; CHECK-NEXT:    br label [[FOR_COND:%.*]], !dbg [[DBG10:![0-9]+]]
+; CHECK:       for.cond:
+; CHECK-NEXT:    [[VLA1_0:%.*]] = phi i32 [ undef, [[ENTRY:%.*]] ], [ [[T0:%.*]], [[FOR_COND]] ]
+; CHECK-NEXT:    call void @llvm.dbg.value(metadata i32 [[VLA1_0]], metadata [[META11:![0-9]+]], metadata !DIExpression()), !dbg [[DBG19:![0-9]+]]
+; CHECK-NEXT:    [[T0]] = add i32 [[VLA1_0]], 1
+; CHECK-NEXT:    call void @llvm.dbg.value(metadata i32 [[T0]], metadata [[META11]], metadata !DIExpression()), !dbg [[DBG19]]
+; CHECK-NEXT:    br label [[FOR_COND]], !dbg [[DBG10]]
+;
 entry:
   %vla1 = alloca i32, i32 1, align 8
   call void @llvm.dbg.declare(metadata i32* %vla1, metadata !10, metadata !DIExpression()), !dbg !18
   br label %for.cond, !dbg !18
 
 for.cond:                                         ; preds = %for.cond, %entry
-; CHECK: %[[PHI:.*]] = phi i32 [ undef, %entry ], [ %t0, %for.cond ]
   %entryN = load i32, i32* %vla1, align 8, !dbg !18
-; CHECK: call void @llvm.dbg.value(metadata i32 %[[PHI]],
-; CHECK-SAME:                      metadata !DIExpression())
   %t0 = add i32 %entryN, 1
-; CHECK: %t0 = add i32 %[[PHI]], 1
-; CHECK: call void @llvm.dbg.value(metadata i32 %t0,
-; CHECK-SAME:                      metadata !DIExpression())
- store i32 %t0, i32* %vla1, align 8, !dbg !18
+  store i32 %t0, i32* %vla1, align 8, !dbg !18
   br label %for.cond, !dbg !18
 }
 
diff --git a/test/Transforms/Mem2Reg/debug-alloca-vla-2.ll b/test/Transforms/Mem2Reg/debug-alloca-vla-2.ll
index 0d7d2e34c81..8b2bc128b0f 100644
--- a/test/Transforms/Mem2Reg/debug-alloca-vla-2.ll
+++ b/test/Transforms/Mem2Reg/debug-alloca-vla-2.ll
@@ -1,3 +1,4 @@
+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
 ; RUN: opt < %s -mem2reg -S | FileCheck %s
 
 ; Testing conversion from dbg.declare to dbg.value when the variable is a VLA.
@@ -17,20 +18,24 @@ target triple = "x86_64-apple-macosx10.12.0"
 
 ; Function Attrs: nounwind ssp uwtable
 define void @scan(i32 %n) #0 !dbg !4 {
+; CHECK-LABEL: @scan(
+; CHECK-NEXT:  entry:
+; CHECK-NEXT:    br label [[FOR_COND:%.*]], !dbg [[DBG10:![0-9]+]]
+; CHECK:       for.cond:
+; CHECK-NEXT:    [[VLA1_0:%.*]] = phi i32 [ undef, [[ENTRY:%.*]] ], [ [[T0:%.*]], [[FOR_COND]] ]
+; CHECK-NEXT:    [[T0]] = add i32 [[VLA1_0]], 1
+; CHECK-NEXT:    call void @llvm.dbg.value(metadata i32 undef, metadata [[META11:![0-9]+]], metadata !DIExpression()), !dbg [[DBG19:![0-9]+]]
+; CHECK-NEXT:    br label [[FOR_COND]], !dbg [[DBG10]]
+;
 entry:
   %vla1 = alloca i32, i32 %n, align 8
   call void @llvm.dbg.declare(metadata i32* %vla1, metadata !10, metadata !DIExpression()), !dbg !18
   br label %for.cond, !dbg !18
 
 for.cond:                                         ; preds = %for.cond, %entry
-; CHECK: %[[PHI:.*]] = phi i32 [ undef, %entry ], [ %t0, %for.cond ]
   %entryN = load i32, i32* %vla1, align 8, !dbg !18
-; CHECK-NOT: call void @llvm.dbg.value
   %t0 = add i32 %entryN, 1
-; CHECK: %t0 = add i32 %[[PHI]], 1
-; CHECK: call void @llvm.dbg.value(metadata i32 undef,
-; CHECK-SAME:                      metadata !DIExpression())
- store i32 %t0, i32* %vla1, align 8, !dbg !18
+  store i32 %t0, i32* %vla1, align 8, !dbg !18
   br label %for.cond, !dbg !18
 }
 
diff --git a/test/Transforms/Mem2Reg/pr24179.ll b/test/Transforms/Mem2Reg/pr24179.ll
index 72a9e61938d..8df27c86a59 100644
--- a/test/Transforms/Mem2Reg/pr24179.ll
+++ b/test/Transforms/Mem2Reg/pr24179.ll
@@ -1,3 +1,4 @@
+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
 ; RUN: opt -mem2reg < %s -S | FileCheck %s
 ; RUN: opt -passes=mem2reg < %s -S | FileCheck %s
 
@@ -9,20 +10,28 @@ declare i1 @use(i32)
 ; and a PHI node to be created.
 define void @test1() {
 ; CHECK-LABEL: @test1(
- entry:
+; CHECK-NEXT:  entry:
+; CHECK-NEXT:    br label [[LOOP:%.*]]
+; CHECK:       loop:
+; CHECK-NEXT:    [[T_0:%.*]] = phi i32 [ undef, [[ENTRY:%.*]] ], [ [[N:%.*]], [[LOOP]] ]
+; CHECK-NEXT:    [[C:%.*]] = call i1 @use(i32 [[T_0]])
+; CHECK-NEXT:    [[N]] = call i32 @def(i32 7)
+; CHECK-NEXT:    br i1 [[C]], label [[LOOP]], label [[EXIT:%.*]]
+; CHECK:       exit:
+; CHECK-NEXT:    ret void
+;
+  entry:
   %t = alloca i32
   br label %loop
 
- loop:
+  loop:
   %v = load i32, i32* %t
   %c = call i1 @use(i32 %v)
-; CHECK: [[PHI:%.*]] = phi i32 [ undef, %entry ], [ %n, %loop ]
-; CHECK: call i1 @use(i32 [[PHI]])
   %n = call i32 @def(i32 7)
   store i32 %n, i32* %t
   br i1 %c, label %loop, label %exit
 
- exit:
+  exit:
   ret void
 }
 
@@ -30,16 +39,23 @@ define void @test1() {
 ; replaced with an undef
 define void @test2() {
 ; CHECK-LABEL: @test2(
- entry:
+; CHECK-NEXT:  entry:
+; CHECK-NEXT:    br label [[LOOP:%.*]]
+; CHECK:       loop:
+; CHECK-NEXT:    [[C:%.*]] = call i1 @use(i32 undef)
+; CHECK-NEXT:    br i1 [[C]], label [[LOOP]], label [[EXIT:%.*]]
+; CHECK:       exit:
+; CHECK-NEXT:    ret void
+;
+  entry:
   %t = alloca i32
   br label %loop
 
- loop:
+  loop:
   %v = load i32, i32* %t
   %c = call i1 @use(i32 %v)
-; CHECK: %c = call i1 @use(i32 undef)
   br i1 %c, label %loop, label %exit
 
- exit:
+  exit:
   ret void
 }
diff --git a/test/Transforms/Mem2Reg/undef-order.ll b/test/Transforms/Mem2Reg/undef-order.ll
index 09687518c87..7642456a9da 100644
--- a/test/Transforms/Mem2Reg/undef-order.ll
+++ b/test/Transforms/Mem2Reg/undef-order.ll
@@ -1,53 +1,97 @@
+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
 ;RUN: opt -mem2reg -S < %s | FileCheck %s
 
 declare i1 @cond()
 
 define i32 @foo() {
+; CHECK-LABEL: @foo(
+; CHECK-NEXT:  Entry:
+; CHECK-NEXT:    [[C1:%.*]] = call i1 @cond()
+; CHECK-NEXT:    br i1 [[C1]], label [[STORE1:%.*]], label [[STORE2:%.*]]
+; CHECK:       Block1:
+; CHECK-NEXT:    br label [[JOIN:%.*]]
+; CHECK:       Block2:
+; CHECK-NEXT:    br label [[JOIN]]
+; CHECK:       Block3:
+; CHECK-NEXT:    br label [[JOIN]]
+; CHECK:       Block4:
+; CHECK-NEXT:    br label [[JOIN]]
+; CHECK:       Block5:
+; CHECK-NEXT:    br label [[JOIN]]
+; CHECK:       Store1:
+; CHECK-NEXT:    br label [[JOIN]]
+; CHECK:       Block6:
+; CHECK-NEXT:    br label [[JOIN]]
+; CHECK:       Block7:
+; CHECK-NEXT:    br label [[JOIN]]
+; CHECK:       Block8:
+; CHECK-NEXT:    br label [[JOIN]]
+; CHECK:       Block9:
+; CHECK-NEXT:    br label [[JOIN]]
+; CHECK:       Block10:
+; CHECK-NEXT:    br label [[JOIN]]
+; CHECK:       Store2:
+; CHECK-NEXT:    br label [[JOIN]]
+; CHECK:       Block11:
+; CHECK-NEXT:    br label [[JOIN]]
+; CHECK:       Block12:
+; CHECK-NEXT:    br label [[JOIN]]
+; CHECK:       Block13:
+; CHECK-NEXT:    br label [[JOIN]]
+; CHECK:       Block14:
+; CHECK-NEXT:    br label [[JOIN]]
+; CHECK:       Block15:
+; CHECK-NEXT:    br label [[JOIN]]
+; CHECK:       Block16:
+; CHECK-NEXT:    br label [[JOIN]]
+; CHECK:       Join:
+; CHECK-NEXT:    [[VAL_0:%.*]] = phi i32 [ 1, [[STORE1]] ], [ 2, [[STORE2]] ], [ undef, [[BLOCK1:%.*]] ], [ undef, [[BLOCK2:%.*]] ], [ undef, [[BLOCK3:%.*]] ], [ undef, [[BLOCK4:%.*]] ], [ undef, [[BLOCK5:%.*]] ], [ undef, [[BLOCK6:%.*]] ], [ undef, [[BLOCK7:%.*]] ], [ undef, [[BLOCK8:%.*]] ], [ undef, [[BLOCK9:%.*]] ], [ undef, [[BLOCK10:%.*]] ], [ undef, [[BLOCK11:%.*]] ], [ undef, [[BLOCK12:%.*]] ], [ undef, [[BLOCK13:%.*]] ], [ undef, [[BLOCK14:%.*]] ], [ undef, [[BLOCK15:%.*]] ], [ undef, [[BLOCK16:%.*]] ]
+; CHECK-NEXT:    ret i32 [[VAL_0]]
+;
 Entry:
-    %val = alloca i32
-    %c1 = call i1 @cond()
-    br i1 %c1, label %Store1, label %Store2
+  %val = alloca i32
+  %c1 = call i1 @cond()
+  br i1 %c1, label %Store1, label %Store2
 Block1:
-    br label %Join
+  br label %Join
 Block2:
-    br label %Join
+  br label %Join
 Block3:
-    br label %Join
+  br label %Join
 Block4:
-    br label %Join
+  br label %Join
 Block5:
-    br label %Join
+  br label %Join
 Store1:
-    store i32 1, i32* %val
-    br label %Join
+  store i32 1, i32* %val
+  br label %Join
 Block6:
-    br label %Join
+  br label %Join
 Block7:
-    br label %Join
+  br label %Join
 Block8:
-    br label %Join
+  br label %Join
 Block9:
-    br label %Join
+  br label %Join
 Block10:
-    br label %Join
+  br label %Join
 Store2:
-    store i32 2, i32* %val
-    br label %Join
+  store i32 2, i32* %val
+  br label %Join
 Block11:
-    br label %Join
+  br label %Join
 Block12:
-    br label %Join
+  br label %Join
 Block13:
-    br label %Join
+  br label %Join
 Block14:
-    br label %Join
+  br label %Join
 Block15:
-    br label %Join
+  br label %Join
 Block16:
-    br label %Join
+  br label %Join
 Join:
 ; Phi inserted here should have operands appended deterministically
-; CHECK: %val.0 = phi i32 [ 1, %Store1 ], [ 2, %Store2 ], [ undef, %Block1 ], [ undef, %Block2 ], [ undef, %Block3 ], [ undef, %Block4 ], [ undef, %Block5 ], [ undef, %Block6 ], [ undef, %Block7 ], [ undef, %Block8 ], [ undef, %Block9 ], [ undef, %Block10 ], [ undef, %Block11 ], [ undef, %Block12 ], [ undef, %Block13 ], [ undef, %Block14 ], [ undef, %Block15 ], [ undef, %Block16 ]
-    %result = load i32, i32* %val
-    ret i32 %result
+  %result = load i32, i32* %val
+  ret i32 %result
 }
diff --git a/test/Transforms/SROA/phi-and-select.ll b/test/Transforms/SROA/phi-and-select.ll
index 8f04e3df13f..c7737149813 100644
--- a/test/Transforms/SROA/phi-and-select.ll
+++ b/test/Transforms/SROA/phi-and-select.ll
@@ -1,70 +1,83 @@
+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
 ; RUN: opt < %s -sroa -S | FileCheck %s
 target datalayout = "e-p:64:64:64-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:32:64-f32:32:32-f64:64:64-v64:64:64-v128:128:128-a0:0:64-n8:16:32:64"
 
 define i32 @test1() {
 ; CHECK-LABEL: @test1(
+; CHECK-NEXT:  entry:
+; CHECK-NEXT:    [[COND:%.*]] = icmp sle i32 0, 1
+; CHECK-NEXT:    br i1 [[COND]], label [[THEN:%.*]], label [[EXIT:%.*]]
+; CHECK:       then:
+; CHECK-NEXT:    br label [[EXIT]]
+; CHECK:       exit:
+; CHECK-NEXT:    [[PHI_SROA_SPECULATED:%.*]] = phi i32 [ 1, [[THEN]] ], [ 0, [[ENTRY:%.*]] ]
+; CHECK-NEXT:    ret i32 [[PHI_SROA_SPECULATED]]
+;
 entry:
-	%a = alloca [2 x i32]
-; CHECK-NOT: alloca
+  %a = alloca [2 x i32]
 
   %a0 = getelementptr [2 x i32], [2 x i32]* %a, i64 0, i32 0
   %a1 = getelementptr [2 x i32], [2 x i32]* %a, i64 0, i32 1
-	store i32 0, i32* %a0
-	store i32 1, i32* %a1
-	%v0 = load i32, i32* %a0
-	%v1 = load i32, i32* %a1
-; CHECK-NOT: store
-; CHECK-NOT: load
+  store i32 0, i32* %a0
+  store i32 1, i32* %a1
+  %v0 = load i32, i32* %a0
+  %v1 = load i32, i32* %a1
 
-	%cond = icmp sle i32 %v0, %v1
-	br i1 %cond, label %then, label %exit
+  %cond = icmp sle i32 %v0, %v1
+  br i1 %cond, label %then, label %exit
 
 then:
-	br label %exit
+  br label %exit
 
 exit:
-	%phi = phi i32* [ %a1, %then ], [ %a0, %entry ]
-; CHECK: phi i32 [ 1, %{{.*}} ], [ 0, %{{.*}} ]
+  %phi = phi i32* [ %a1, %then ], [ %a0, %entry ]
 
-	%result = load i32, i32* %phi
-	ret i32 %result
+  %result = load i32, i32* %phi
+  ret i32 %result
 }
 
 define i32 @test2() {
 ; CHECK-LABEL: @test2(
+; CHECK-NEXT:  entry:
+; CHECK-NEXT:    [[COND:%.*]] = icmp sle i32 0, 1
+; CHECK-NEXT:    [[RESULT_SROA_SPECULATED:%.*]] = select i1 [[COND]], i32 1, i32 0
+; CHECK-NEXT:    ret i32 [[RESULT_SROA_SPECULATED]]
+;
 entry:
-	%a = alloca [2 x i32]
-; CHECK-NOT: alloca
+  %a = alloca [2 x i32]
 
   %a0 = getelementptr [2 x i32], [2 x i32]* %a, i64 0, i32 0
   %a1 = getelementptr [2 x i32], [2 x i32]* %a, i64 0, i32 1
-	store i32 0, i32* %a0
-	store i32 1, i32* %a1
-	%v0 = load i32, i32* %a0
-	%v1 = load i32, i32* %a1
-; CHECK-NOT: store
-; CHECK-NOT: load
+  store i32 0, i32* %a0
+  store i32 1, i32* %a1
+  %v0 = load i32, i32* %a0
+  %v1 = load i32, i32* %a1
 
-	%cond = icmp sle i32 %v0, %v1
-	%select = select i1 %cond, i32* %a1, i32* %a0
-; CHECK: select i1 %{{.*}}, i32 1, i32 0
+  %cond = icmp sle i32 %v0, %v1
+  %select = select i1 %cond, i32* %a1, i32* %a0
 
-	%result = load i32, i32* %select
-	ret i32 %result
+  %result = load i32, i32* %select
+  ret i32 %result
 }
 
 ; If bitcast isn't considered a safe phi/select use, the alloca
 ; remains as an array.
 ; FIXME: Why isn't this identical to test2?
-
-; CHECK-LABEL: @test2_bitcast(
-; CHECK: alloca i32
-; CHECK-NEXT: alloca i32
-
-; CHECK: %select = select i1 %cond, i32* %a.sroa.3, i32* %a.sroa.0
-; CHECK-NEXT: %select.bc = bitcast i32* %select to float*
-; CHECK-NEXT: %result = load float, float* %select.bc, align 4
 define float @test2_bitcast() {
+; CHECK-LABEL: @test2_bitcast(
+; CHECK-NEXT:  entry:
+; CHECK-NEXT:    [[A_SROA_0:%.*]] = alloca i32, align 4
+; CHECK-NEXT:    [[A_SROA_3:%.*]] = alloca i32, align 4
+; CHECK-NEXT:    store i32 0, i32* [[A_SROA_0]], align 4
+; CHECK-NEXT:    store i32 1, i32* [[A_SROA_3]], align 4
+; CHECK-NEXT:    [[A_SROA_0_0_A_SROA_0_0_V0:%.*]] = load i32, i32* [[A_SROA_0]], align 4
+; CHECK-NEXT:    [[A_SROA_3_0_A_SROA_3_4_V1:%.*]] = load i32, i32* [[A_SROA_3]], align 4
+; CHECK-NEXT:    [[COND:%.*]] = icmp sle i32 [[A_SROA_0_0_A_SROA_0_0_V0]], [[A_SROA_3_0_A_SROA_3_4_V1]]
+; CHECK-NEXT:    [[SELECT:%.*]] = select i1 [[COND]], i32* [[A_SROA_3]], i32* [[A_SROA_0]]
+; CHECK-NEXT:    [[SELECT_BC:%.*]] = bitcast i32* [[SELECT]] to float*
+; CHECK-NEXT:    [[RESULT:%.*]] = load float, float* [[SELECT_BC]], align 4
+; CHECK-NEXT:    ret float [[RESULT]]
+;
 entry:
   %a = alloca [2 x i32]
   %a0 = getelementptr [2 x i32], [2 x i32]* %a, i64 0, i32 0
@@ -80,14 +93,21 @@ entry:
   ret float %result
 }
 
-; CHECK-LABEL: @test2_addrspacecast(
-; CHECK: alloca i32
-; CHECK-NEXT: alloca i32
-
-; CHECK: %select = select i1 %cond, i32* %a.sroa.3, i32* %a.sroa.0
-; CHECK-NEXT: %select.asc = addrspacecast i32* %select to i32 addrspace(1)*
-; CHECK-NEXT: load i32, i32 addrspace(1)* %select.asc, align 4
 define i32 @test2_addrspacecast() {
+; CHECK-LABEL: @test2_addrspacecast(
+; CHECK-NEXT:  entry:
+; CHECK-NEXT:    [[A_SROA_0:%.*]] = alloca i32, align 4
+; CHECK-NEXT:    [[A_SROA_3:%.*]] = alloca i32, align 4
+; CHECK-NEXT:    store i32 0, i32* [[A_SROA_0]], align 4
+; CHECK-NEXT:    store i32 1, i32* [[A_SROA_3]], align 4
+; CHECK-NEXT:    [[A_SROA_0_0_A_SROA_0_0_V0:%.*]] = load i32, i32* [[A_SROA_0]], align 4
+; CHECK-NEXT:    [[A_SROA_3_0_A_SROA_3_4_V1:%.*]] = load i32, i32* [[A_SROA_3]], align 4
+; CHECK-NEXT:    [[COND:%.*]] = icmp sle i32 [[A_SROA_0_0_A_SROA_0_0_V0]], [[A_SROA_3_0_A_SROA_3_4_V1]]
+; CHECK-NEXT:    [[SELECT:%.*]] = select i1 [[COND]], i32* [[A_SROA_3]], i32* [[A_SROA_0]]
+; CHECK-NEXT:    [[SELECT_ASC:%.*]] = addrspacecast i32* [[SELECT]] to i32 addrspace(1)*
+; CHECK-NEXT:    [[RESULT:%.*]] = load i32, i32 addrspace(1)* [[SELECT_ASC]], align 4
+; CHECK-NEXT:    ret i32 [[RESULT]]
+;
 entry:
   %a = alloca [2 x i32]
   %a0 = getelementptr [2 x i32], [2 x i32]* %a, i64 0, i32 0
@@ -105,9 +125,38 @@ entry:
 
 define i32 @test3(i32 %x) {
 ; CHECK-LABEL: @test3(
+; CHECK-NEXT:  entry:
+; CHECK-NEXT:    switch i32 [[X:%.*]], label [[BB0:%.*]] [
+; CHECK-NEXT:    i32 1, label [[BB1:%.*]]
+; CHECK-NEXT:    i32 2, label [[BB2:%.*]]
+; CHECK-NEXT:    i32 3, label [[BB3:%.*]]
+; CHECK-NEXT:    i32 4, label [[BB4:%.*]]
+; CHECK-NEXT:    i32 5, label [[BB5:%.*]]
+; CHECK-NEXT:    i32 6, label [[BB6:%.*]]
+; CHECK-NEXT:    i32 7, label [[BB7:%.*]]
+; CHECK-NEXT:    ]
+; CHECK:       bb0:
+; CHECK-NEXT:    br label [[EXIT:%.*]]
+; CHECK:       bb1:
+; CHECK-NEXT:    br label [[EXIT]]
+; CHECK:       bb2:
+; CHECK-NEXT:    br label [[EXIT]]
+; CHECK:       bb3:
+; CHECK-NEXT:    br label [[EXIT]]
+; CHECK:       bb4:
+; CHECK-NEXT:    br label [[EXIT]]
+; CHECK:       bb5:
+; CHECK-NEXT:    br label [[EXIT]]
+; CHECK:       bb6:
+; CHECK-NEXT:    br label [[EXIT]]
+; CHECK:       bb7:
+; CHECK-NEXT:    br label [[EXIT]]
+; CHECK:       exit:
+; CHECK-NEXT:    [[PHI_SROA_SPECULATED:%.*]] = phi i32 [ 1, [[BB0]] ], [ 0, [[BB1]] ], [ 0, [[BB2]] ], [ 1, [[BB3]] ], [ 1, [[BB4]] ], [ 0, [[BB5]] ], [ 0, [[BB6]] ], [ 1, [[BB7]] ]
+; CHECK-NEXT:    ret i32 [[PHI_SROA_SPECULATED]]
+;
 entry:
-	%a = alloca [2 x i32]
-; CHECK-NOT: alloca
+  %a = alloca [2 x i32]
 
   ; Note that we build redundant GEPs here to ensure that having different GEPs
   ; into the same alloca partation continues to work with PHI speculation. This
@@ -116,124 +165,128 @@ entry:
   %a0b = getelementptr [2 x i32], [2 x i32]* %a, i64 0, i32 0
   %a1 = getelementptr [2 x i32], [2 x i32]* %a, i64 0, i32 1
   %a1b = getelementptr [2 x i32], [2 x i32]* %a, i64 0, i32 1
-	store i32 0, i32* %a0
-	store i32 1, i32* %a1
-; CHECK-NOT: store
+  store i32 0, i32* %a0
+  store i32 1, i32* %a1
 
   switch i32 %x, label %bb0 [ i32 1, label %bb1
-                              i32 2, label %bb2
-                              i32 3, label %bb3
-                              i32 4, label %bb4
-                              i32 5, label %bb5
-                              i32 6, label %bb6
-                              i32 7, label %bb7 ]
+  i32 2, label %bb2
+  i32 3, label %bb3
+  i32 4, label %bb4
+  i32 5, label %bb5
+  i32 6, label %bb6
+  i32 7, label %bb7 ]
 
 bb0:
-	br label %exit
+  br label %exit
 bb1:
-	br label %exit
+  br label %exit
 bb2:
-	br label %exit
+  br label %exit
 bb3:
-	br label %exit
+  br label %exit
 bb4:
-	br label %exit
+  br label %exit
 bb5:
-	br label %exit
+  br label %exit
 bb6:
-	br label %exit
+  br label %exit
 bb7:
-	br label %exit
+  br label %exit
 
 exit:
-	%phi = phi i32* [ %a1, %bb0 ], [ %a0, %bb1 ], [ %a0, %bb2 ], [ %a1, %bb3 ],
-                  [ %a1b, %bb4 ], [ %a0b, %bb5 ], [ %a0b, %bb6 ], [ %a1b, %bb7 ]
-; CHECK: phi i32 [ 1, %{{.*}} ], [ 0, %{{.*}} ], [ 0, %{{.*}} ], [ 1, %{{.*}} ], [ 1, %{{.*}} ], [ 0, %{{.*}} ], [ 0, %{{.*}} ], [ 1, %{{.*}} ]
+  %phi = phi i32* [ %a1, %bb0 ], [ %a0, %bb1 ], [ %a0, %bb2 ], [ %a1, %bb3 ],
+  [ %a1b, %bb4 ], [ %a0b, %bb5 ], [ %a0b, %bb6 ], [ %a1b, %bb7 ]
 
-	%result = load i32, i32* %phi
-	ret i32 %result
+  %result = load i32, i32* %phi
+  ret i32 %result
 }
 
 define i32 @test4() {
 ; CHECK-LABEL: @test4(
+; CHECK-NEXT:  entry:
+; CHECK-NEXT:    ret i32 0
+;
 entry:
-	%a = alloca [2 x i32]
-; CHECK-NOT: alloca
+  %a = alloca [2 x i32]
 
   %a0 = getelementptr [2 x i32], [2 x i32]* %a, i64 0, i32 0
   %a1 = getelementptr [2 x i32], [2 x i32]* %a, i64 0, i32 1
-	store i32 0, i32* %a0
-	store i32 1, i32* %a1
-	%v0 = load i32, i32* %a0
-	%v1 = load i32, i32* %a1
-; CHECK-NOT: store
-; CHECK-NOT: load
+  store i32 0, i32* %a0
+  store i32 1, i32* %a1
+  %v0 = load i32, i32* %a0
+  %v1 = load i32, i32* %a1
 
-	%cond = icmp sle i32 %v0, %v1
-	%select = select i1 %cond, i32* %a0, i32* %a0
-; CHECK-NOT: select
+  %cond = icmp sle i32 %v0, %v1
+  %select = select i1 %cond, i32* %a0, i32* %a0
 
-	%result = load i32, i32* %select
-	ret i32 %result
-; CHECK: ret i32 0
+  %result = load i32, i32* %select
+  ret i32 %result
 }
 
 define i32 @test5(i32* %b) {
 ; CHECK-LABEL: @test5(
+; CHECK-NEXT:  entry:
+; CHECK-NEXT:    ret i32 1
+;
 entry:
-	%a = alloca [2 x i32]
-; CHECK-NOT: alloca
+  %a = alloca [2 x i32]
 
   %a1 = getelementptr [2 x i32], [2 x i32]* %a, i64 0, i32 1
-	store i32 1, i32* %a1
-; CHECK-NOT: store
+  store i32 1, i32* %a1
 
-	%select = select i1 true, i32* %a1, i32* %b
-; CHECK-NOT: select
+  %select = select i1 true, i32* %a1, i32* %b
 
-	%result = load i32, i32* %select
-; CHECK-NOT: load
+  %result = load i32, i32* %select
 
-	ret i32 %result
-; CHECK: ret i32 1
+  ret i32 %result
 }
 
 declare void @f(i32*, i32*)
 
 define i32 @test6(i32* %b) {
 ; CHECK-LABEL: @test6(
+; CHECK-NEXT:  entry:
+; CHECK-NEXT:    [[SELECT2:%.*]] = select i1 false, i32* undef, i32* [[B:%.*]]
+; CHECK-NEXT:    [[SELECT3:%.*]] = select i1 false, i32* undef, i32* [[B]]
+; CHECK-NEXT:    call void @f(i32* [[SELECT2]], i32* [[SELECT3]])
+; CHECK-NEXT:    ret i32 1
+;
 entry:
-	%a = alloca [2 x i32]
+  %a = alloca [2 x i32]
   %c = alloca i32
-; CHECK-NOT: alloca
 
   %a1 = getelementptr [2 x i32], [2 x i32]* %a, i64 0, i32 1
-	store i32 1, i32* %a1
+  store i32 1, i32* %a1
 
-	%select = select i1 true, i32* %a1, i32* %b
-	%select2 = select i1 false, i32* %a1, i32* %b
+  %select = select i1 true, i32* %a1, i32* %b
+  %select2 = select i1 false, i32* %a1, i32* %b
   %select3 = select i1 false, i32* %c, i32* %b
-; CHECK: %[[select2:.*]] = select i1 false, i32* undef, i32* %b
-; CHECK: %[[select3:.*]] = select i1 false, i32* undef, i32* %b
 
   ; Note, this would potentially escape the alloca pointer except for the
   ; constant folding of the select.
   call void @f(i32* %select2, i32* %select3)
-; CHECK: call void @f(i32* %[[select2]], i32* %[[select3]])
 
 
-	%result = load i32, i32* %select
-; CHECK-NOT: load
+  %result = load i32, i32* %select
 
   %dead = load i32, i32* %c
 
-	ret i32 %result
-; CHECK: ret i32 1
+  ret i32 %result
 }
 
 define i32 @test7() {
 ; CHECK-LABEL: @test7(
-; CHECK-NOT: alloca
+; CHECK-NEXT:  entry:
+; CHECK-NEXT:    br i1 undef, label [[GOOD:%.*]], label [[BAD:%.*]]
+; CHECK:       good:
+; CHECK-NEXT:    br label [[EXIT:%.*]]
+; CHECK:       bad:
+; CHECK-NEXT:    [[P_SROA_SPECULATE_LOAD_BAD:%.*]] = load i32, i32* undef, align 4
+; CHECK-NEXT:    br label [[EXIT]]
+; CHECK:       exit:
+; CHECK-NEXT:    [[P_SROA_SPECULATED:%.*]] = phi i32 [ 0, [[GOOD]] ], [ [[P_SROA_SPECULATE_LOAD_BAD]], [[BAD]] ]
+; CHECK-NEXT:    ret i32 [[P_SROA_SPECULATED]]
+;
 
 entry:
   %X = alloca i32
@@ -250,23 +303,27 @@ bad:
   br label %exit
 
 exit:
-	%P = phi i32* [ %Y1, %good ], [ %Y2, %bad ]
-; CHECK: %[[phi:.*]] = phi i32 [ 0, %good ],
+  %P = phi i32* [ %Y1, %good ], [ %Y2, %bad ]
   %Z2 = load i32, i32* %P
   ret i32 %Z2
-; CHECK: ret i32 %[[phi]]
 }
 
 define i32 @test8(i32 %b, i32* %ptr) {
 ; Ensure that we rewrite allocas to the used type when that use is hidden by
 ; a PHI that can be speculated.
 ; CHECK-LABEL: @test8(
-; CHECK-NOT: alloca
-; CHECK-NOT: load
-; CHECK: %[[value:.*]] = load i32, i32* %ptr
-; CHECK-NOT: load
-; CHECK: %[[result:.*]] = phi i32 [ undef, %else ], [ %[[value]], %then ]
-; CHECK-NEXT: ret i32 %[[result]]
+; CHECK-NEXT:  entry:
+; CHECK-NEXT:    [[TEST:%.*]] = icmp ne i32 [[B:%.*]], 0
+; CHECK-NEXT:    br i1 [[TEST]], label [[THEN:%.*]], label [[ELSE:%.*]]
+; CHECK:       then:
+; CHECK-NEXT:    [[PHI_SROA_SPECULATE_LOAD_THEN:%.*]] = load i32, i32* [[PTR:%.*]], align 4
+; CHECK-NEXT:    br label [[EXIT:%.*]]
+; CHECK:       else:
+; CHECK-NEXT:    br label [[EXIT]]
+; CHECK:       exit:
+; CHECK-NEXT:    [[PHI_SROA_SPECULATED:%.*]] = phi i32 [ undef, [[ELSE]] ], [ [[PHI_SROA_SPECULATE_LOAD_THEN]], [[THEN]] ]
+; CHECK-NEXT:    ret i32 [[PHI_SROA_SPECULATED]]
+;
 
 entry:
   %f = alloca float
@@ -289,12 +346,13 @@ exit:
 define i32 @test9(i32 %b, i32* %ptr) {
 ; Same as @test8 but for a select rather than a PHI node.
 ; CHECK-LABEL: @test9(
-; CHECK-NOT: alloca
-; CHECK-NOT: load
-; CHECK: %[[value:.*]] = load i32, i32* %ptr
-; CHECK-NOT: load
-; CHECK: %[[result:.*]] = select i1 %{{.*}}, i32 undef, i32 %[[value]]
-; CHECK-NEXT: ret i32 %[[result]]
+; CHECK-NEXT:  entry:
+; CHECK-NEXT:    store i32 0, i32* [[PTR:%.*]], align 4
+; CHECK-NEXT:    [[TEST:%.*]] = icmp ne i32 [[B:%.*]], 0
+; CHECK-NEXT:    [[LOADED_SROA_SPECULATE_LOAD_FALSE:%.*]] = load i32, i32* [[PTR]], align 4
+; CHECK-NEXT:    [[LOADED_SROA_SPECULATED:%.*]] = select i1 [[TEST]], i32 undef, i32 [[LOADED_SROA_SPECULATE_LOAD_FALSE]]
+; CHECK-NEXT:    ret i32 [[LOADED_SROA_SPECULATED]]
+;
 
 entry:
   %f = alloca float
@@ -311,12 +369,22 @@ define float @test10(i32 %b, float* %ptr) {
 ; rewriting due to the necessity of inserting bitcasts when speculating a PHI
 ; node.
 ; CHECK-LABEL: @test10(
-; CHECK: %[[alloca:.*]] = alloca
-; CHECK: %[[argvalue:.*]] = load float, float* %ptr
-; CHECK: %[[cast:.*]] = bitcast double* %[[alloca]] to float*
-; CHECK: %[[allocavalue:.*]] = load float, float* %[[cast]]
-; CHECK: %[[result:.*]] = phi float [ %[[allocavalue]], %else ], [ %[[argvalue]], %then ]
-; CHECK-NEXT: ret float %[[result]]
+; CHECK-NEXT:  entry:
+; CHECK-NEXT:    [[F:%.*]] = alloca double, align 8
+; CHECK-NEXT:    store double 0.000000e+00, double* [[F]], align 8
+; CHECK-NEXT:    [[TEST:%.*]] = icmp ne i32 [[B:%.*]], 0
+; CHECK-NEXT:    br i1 [[TEST]], label [[THEN:%.*]], label [[ELSE:%.*]]
+; CHECK:       then:
+; CHECK-NEXT:    [[PHI_SROA_SPECULATE_LOAD_THEN:%.*]] = load float, float* [[PTR:%.*]], align 4
+; CHECK-NEXT:    br label [[EXIT:%.*]]
+; CHECK:       else:
+; CHECK-NEXT:    [[F_0_F_0_BITCAST_SROA_CAST:%.*]] = bitcast double* [[F]] to float*
+; CHECK-NEXT:    [[F_0_PHI_SROA_SPECULATE_LOAD_ELSE:%.*]] = load float, float* [[F_0_F_0_BITCAST_SROA_CAST]], align 8
+; CHECK-NEXT:    br label [[EXIT]]
+; CHECK:       exit:
+; CHECK-NEXT:    [[PHI_SROA_SPECULATED:%.*]] = phi float [ [[F_0_PHI_SROA_SPECULATE_LOAD_ELSE]], [[ELSE]] ], [ [[PHI_SROA_SPECULATE_LOAD_THEN]], [[THEN]] ]
+; CHECK-NEXT:    ret float [[PHI_SROA_SPECULATED]]
+;
 
 entry:
   %f = alloca double
@@ -340,12 +408,17 @@ exit:
 define float @test11(i32 %b, float* %ptr) {
 ; Same as @test10 but for a select rather than a PHI node.
 ; CHECK-LABEL: @test11(
-; CHECK: %[[alloca:.*]] = alloca
-; CHECK: %[[cast:.*]] = bitcast double* %[[alloca]] to float*
-; CHECK: %[[allocavalue:.*]] = load float, float* %[[cast]]
-; CHECK: %[[argvalue:.*]] = load float, float* %ptr
-; CHECK: %[[result:.*]] = select i1 %{{.*}}, float %[[allocavalue]], float %[[argvalue]]
-; CHECK-NEXT: ret float %[[result]]
+; CHECK-NEXT:  entry:
+; CHECK-NEXT:    [[F:%.*]] = alloca double, align 8
+; CHECK-NEXT:    store double 0.000000e+00, double* [[F]], align 8
+; CHECK-NEXT:    store float 0.000000e+00, float* [[PTR:%.*]], align 4
+; CHECK-NEXT:    [[TEST:%.*]] = icmp ne i32 [[B:%.*]], 0
+; CHECK-NEXT:    [[F_0_F_0_BITCAST_SROA_CAST:%.*]] = bitcast double* [[F]] to float*
+; CHECK-NEXT:    [[F_0_LOADED_SROA_SPECULATE_LOAD_TRUE:%.*]] = load float, float* [[F_0_F_0_BITCAST_SROA_CAST]], align 8
+; CHECK-NEXT:    [[LOADED_SROA_SPECULATE_LOAD_FALSE:%.*]] = load float, float* [[PTR]], align 4
+; CHECK-NEXT:    [[LOADED_SROA_SPECULATED:%.*]] = select i1 [[TEST]], float [[F_0_LOADED_SROA_SPECULATE_LOAD_TRUE]], float [[LOADED_SROA_SPECULATE_LOAD_FALSE]]
+; CHECK-NEXT:    ret float [[LOADED_SROA_SPECULATED]]
+;
 
 entry:
   %f = alloca double
@@ -362,9 +435,9 @@ define i32 @test12(i32 %x, i32* %p) {
 ; Ensure we don't crash or fail to nuke dead selects of allocas if no load is
 ; never found.
 ; CHECK-LABEL: @test12(
-; CHECK-NOT: alloca
-; CHECK-NOT: select
-; CHECK: ret i32 %x
+; CHECK-NEXT:  entry:
+; CHECK-NEXT:    ret i32 [[X:%.*]]
+;
 
 entry:
   %a = alloca i32
@@ -378,9 +451,13 @@ define i32 @test13(i32 %x, i32* %p) {
 ; Ensure we don't crash or fail to nuke dead phis of allocas if no load is ever
 ; found.
 ; CHECK-LABEL: @test13(
-; CHECK-NOT: alloca
-; CHECK-NOT: phi
-; CHECK: ret i32 %x
+; CHECK-NEXT:  entry:
+; CHECK-NEXT:    br label [[LOOP:%.*]]
+; CHECK:       loop:
+; CHECK-NEXT:    br i1 undef, label [[LOOP]], label [[EXIT:%.*]]
+; CHECK:       exit:
+; CHECK-NEXT:    ret i32 [[X:%.*]]
+;
 
 entry:
   %a = alloca i32
@@ -401,13 +478,26 @@ define i32 @test14(i1 %b1, i1 %b2, i32* %ptr) {
 ; speculatable toward promotion but the other is not. That should block all of
 ; the speculation.
 ; CHECK-LABEL: @test14(
-; CHECK: alloca
-; CHECK: alloca
-; CHECK: select
-; CHECK: phi
-; CHECK: phi
-; CHECK: select
-; CHECK: ret i32
+; CHECK-NEXT:  entry:
+; CHECK-NEXT:    [[F:%.*]] = alloca i32, align 4
+; CHECK-NEXT:    [[G:%.*]] = alloca i32, align 4
+; CHECK-NEXT:    store i32 0, i32* [[F]], align 4
+; CHECK-NEXT:    store i32 0, i32* [[G]], align 4
+; CHECK-NEXT:    [[F_SELECT:%.*]] = select i1 [[B1:%.*]], i32* [[F]], i32* [[PTR:%.*]]
+; CHECK-NEXT:    br i1 [[B2:%.*]], label [[THEN:%.*]], label [[ELSE:%.*]]
+; CHECK:       then:
+; CHECK-NEXT:    br label [[EXIT:%.*]]
+; CHECK:       else:
+; CHECK-NEXT:    br label [[EXIT]]
+; CHECK:       exit:
+; CHECK-NEXT:    [[F_PHI:%.*]] = phi i32* [ [[F]], [[THEN]] ], [ [[F_SELECT]], [[ELSE]] ]
+; CHECK-NEXT:    [[G_PHI:%.*]] = phi i32* [ [[G]], [[THEN]] ], [ [[PTR]], [[ELSE]] ]
+; CHECK-NEXT:    [[F_LOADED:%.*]] = load i32, i32* [[F_PHI]], align 4
+; CHECK-NEXT:    [[G_SELECT:%.*]] = select i1 [[B1]], i32* [[G]], i32* [[G_PHI]]
+; CHECK-NEXT:    [[G_LOADED:%.*]] = load i32, i32* [[G_SELECT]], align 4
+; CHECK-NEXT:    [[RESULT:%.*]] = add i32 [[F_LOADED]], [[G_LOADED]]
+; CHECK-NEXT:    ret i32 [[RESULT]]
+;
 
 entry:
   %f = alloca i32
@@ -437,8 +527,16 @@ define i32 @PR13905() {
 ; Check a pattern where we have a chain of dead phi nodes to ensure they are
 ; deleted and promotion can proceed.
 ; CHECK-LABEL: @PR13905(
-; CHECK-NOT: alloca i32
-; CHECK: ret i32 undef
+; CHECK-NEXT:  entry:
+; CHECK-NEXT:    br i1 undef, label [[LOOP1:%.*]], label [[EXIT:%.*]]
+; CHECK:       loop1:
+; CHECK-NEXT:    br i1 undef, label [[LOOP1]], label [[LOOP2:%.*]]
+; CHECK:       loop2:
+; CHECK-NEXT:    br i1 undef, label [[LOOP1]], label [[EXIT]]
+; CHECK:       exit:
+; CHECK-NEXT:    [[PHI2:%.*]] = phi i32* [ undef, [[LOOP2]] ], [ null, [[ENTRY:%.*]] ]
+; CHECK-NEXT:    ret i32 undef
+;
 
 entry:
   %h = alloca i32
@@ -462,7 +560,13 @@ define i32 @PR13906() {
 ; PHI nodes or select nodes. This triggers subtly differently from the above
 ; cases because the PHI node is (recursively) alive, but the select is dead.
 ; CHECK-LABEL: @PR13906(
-; CHECK-NOT: alloca
+; CHECK-NEXT:  entry:
+; CHECK-NEXT:    br label [[FOR_COND:%.*]]
+; CHECK:       for.cond:
+; CHECK-NEXT:    br i1 undef, label [[IF_THEN:%.*]], label [[FOR_COND]]
+; CHECK:       if.then:
+; CHECK-NEXT:    br label [[FOR_COND]]
+;
 
 entry:
   %c = alloca i32
@@ -480,6 +584,15 @@ if.then:
 
 define i64 @PR14132(i1 %flag) {
 ; CHECK-LABEL: @PR14132(
+; CHECK-NEXT:  entry:
+; CHECK-NEXT:    br i1 [[FLAG:%.*]], label [[IF_THEN:%.*]], label [[IF_END:%.*]]
+; CHECK:       if.then:
+; CHECK-NEXT:    [[B_0_LOAD_EXT:%.*]] = zext i8 1 to i64
+; CHECK-NEXT:    br label [[IF_END]]
+; CHECK:       if.end:
+; CHECK-NEXT:    [[PTR_0_SROA_SPECULATED:%.*]] = phi i64 [ [[B_0_LOAD_EXT]], [[IF_THEN]] ], [ 0, [[ENTRY:%.*]] ]
+; CHECK-NEXT:    ret i64 [[PTR_0_SROA_SPECULATED]]
+;
 ; Here we form a PHI-node by promoting the pointer alloca first, and then in
 ; order to promote the other two allocas, we speculate the load of the
 ; now-phi-node-pointer. In doing so we end up loading a 64-bit value from an i8
@@ -491,7 +604,6 @@ entry:
   %a = alloca i64, align 8
   %b = alloca i8, align 8
   %ptr = alloca i64*, align 8
-; CHECK-NOT: alloca
 
   %ptr.cast = bitcast i64** %ptr to i8**
   store i64 0, i64* %a, align 8
@@ -502,21 +614,31 @@ entry:
 if.then:
   store i8* %b, i8** %ptr.cast, align 8
   br label %if.end
-; CHECK-NOT: store
-; CHECK: %[[ext:.*]] = zext i8 1 to i64
 
 if.end:
   %tmp = load i64*, i64** %ptr, align 8
   %result = load i64, i64* %tmp, align 8
-; CHECK-NOT: load
-; CHECK: %[[result:.*]] = phi i64 [ %[[ext]], %if.then ], [ 0, %entry ]
 
   ret i64 %result
-; CHECK-NEXT: ret i64 %[[result]]
 }
 
 define float @PR16687(i64 %x, i1 %flag) {
 ; CHECK-LABEL: @PR16687(
+; CHECK-NEXT:  entry:
+; CHECK-NEXT:    [[A_SROA_0_0_EXTRACT_TRUNC:%.*]] = trunc i64 [[X:%.*]] to i32
+; CHECK-NEXT:    [[A_SROA_2_0_EXTRACT_SHIFT:%.*]] = lshr i64 [[X]], 32
+; CHECK-NEXT:    [[A_SROA_2_0_EXTRACT_TRUNC:%.*]] = trunc i64 [[A_SROA_2_0_EXTRACT_SHIFT]] to i32
+; CHECK-NEXT:    br i1 [[FLAG:%.*]], label [[THEN:%.*]], label [[ELSE:%.*]]
+; CHECK:       then:
+; CHECK-NEXT:    [[TMP0:%.*]] = bitcast i32 [[A_SROA_0_0_EXTRACT_TRUNC]] to float
+; CHECK-NEXT:    br label [[END:%.*]]
+; CHECK:       else:
+; CHECK-NEXT:    [[TMP1:%.*]] = bitcast i32 [[A_SROA_2_0_EXTRACT_TRUNC]] to float
+; CHECK-NEXT:    br label [[END]]
+; CHECK:       end:
+; CHECK-NEXT:    [[A_PHI_F_SROA_SPECULATED:%.*]] = phi float [ [[TMP0]], [[THEN]] ], [ [[TMP1]], [[ELSE]] ]
+; CHECK-NEXT:    ret float [[A_PHI_F_SROA_SPECULATED]]
+;
 ; Check that even when we try to speculate the same phi twice (in two slices)
 ; on an otherwise promotable construct, we don't get ahead of ourselves and try
 ; to promote one of the slices prior to speculating it.
@@ -525,31 +647,21 @@ entry:
   %a = alloca i64, align 8
   store i64 %x, i64* %a
   br i1 %flag, label %then, label %else
-; CHECK-NOT: alloca
-; CHECK-NOT: store
-; CHECK: %[[lo:.*]] = trunc i64 %x to i32
-; CHECK: %[[shift:.*]] = lshr i64 %x, 32
-; CHECK: %[[hi:.*]] = trunc i64 %[[shift]] to i32
 
 then:
   %a.f = bitcast i64* %a to float*
   br label %end
-; CHECK: %[[lo_cast:.*]] = bitcast i32 %[[lo]] to float
 
 else:
   %a.raw = bitcast i64* %a to i8*
   %a.raw.4 = getelementptr i8, i8* %a.raw, i64 4
   %a.raw.4.f = bitcast i8* %a.raw.4 to float*
   br label %end
-; CHECK: %[[hi_cast:.*]] = bitcast i32 %[[hi]] to float
 
 end:
   %a.phi.f = phi float* [ %a.f, %then ], [ %a.raw.4.f, %else ]
   %f = load float, float* %a.phi.f
   ret float %f
-; CHECK: %[[phi:.*]] = phi float [ %[[lo_cast]], %then ], [ %[[hi_cast]], %else ]
-; CHECK-NOT: load
-; CHECK: ret float %[[phi]]
 }
 
 ; Verifies we fixed PR20425. We should be able to promote all alloca's to
@@ -560,9 +672,19 @@ end:
 ; %2 = phi(%0, %1) // == slice
 define float @simplify_phi_nodes_that_equal_slice(i1 %cond, float* %temp) {
 ; CHECK-LABEL: @simplify_phi_nodes_that_equal_slice(
+; CHECK-NEXT:  entry:
+; CHECK-NEXT:    br i1 [[COND:%.*]], label [[THEN:%.*]], label [[ELSE:%.*]]
+; CHECK:       then:
+; CHECK-NEXT:    br label [[MERGE:%.*]]
+; CHECK:       else:
+; CHECK-NEXT:    br label [[MERGE]]
+; CHECK:       merge:
+; CHECK-NEXT:    [[ARR_SROA_0_0:%.*]] = phi float [ 1.000000e+00, [[THEN]] ], [ 2.000000e+00, [[ELSE]] ]
+; CHECK-NEXT:    store float 0.000000e+00, float* [[TEMP:%.*]], align 4
+; CHECK-NEXT:    ret float [[ARR_SROA_0_0]]
+;
 entry:
   %arr = alloca [4 x float], align 4
-; CHECK-NOT: alloca
   br i1 %cond, label %then, label %else
 
 then:
@@ -590,9 +712,21 @@ merge:
 ; %3 = phi(%1, %2) // == slice
 define float @simplify_phi_nodes_that_equal_slice_2(i1 %cond, float* %temp) {
 ; CHECK-LABEL: @simplify_phi_nodes_that_equal_slice_2(
+; CHECK-NEXT:  entry:
+; CHECK-NEXT:    br i1 [[COND:%.*]], label [[THEN:%.*]], label [[ELSE:%.*]]
+; CHECK:       then:
+; CHECK-NEXT:    br label [[THEN2:%.*]]
+; CHECK:       then2:
+; CHECK-NEXT:    br label [[MERGE:%.*]]
+; CHECK:       else:
+; CHECK-NEXT:    br label [[MERGE]]
+; CHECK:       merge:
+; CHECK-NEXT:    [[ARR_SROA_0_0:%.*]] = phi float [ 2.000000e+00, [[THEN2]] ], [ 3.000000e+00, [[ELSE]] ]
+; CHECK-NEXT:    store float 0.000000e+00, float* [[TEMP:%.*]], align 4
+; CHECK-NEXT:    ret float [[ARR_SROA_0_0]]
+;
 entry:
   %arr = alloca [4 x float], align 4
-; CHECK-NOT: alloca
   br i1 %cond, label %then, label %else
 
 then:
@@ -626,9 +760,25 @@ merge:
 ; make sure we insert *after* the first non-PHI instruction.
 define void @PR20822() {
 ; CHECK-LABEL: @PR20822(
+; CHECK-NEXT:  entry:
+; CHECK-NEXT:    [[F_SROA_0:%.*]] = alloca i32, align 4
+; CHECK-NEXT:    br i1 undef, label [[IF_END:%.*]], label [[FOR_COND:%.*]]
+; CHECK:       for.cond:
+; CHECK-NEXT:    br label [[IF_END]]
+; CHECK:       if.end:
+; CHECK-NEXT:    [[TMP0:%.*]] = phi i32 [ undef, [[ENTRY:%.*]] ], [ undef, [[FOR_COND]] ]
+; CHECK-NEXT:    [[F_SROA_0_0_F2_SROA_CAST1:%.*]] = bitcast i32* [[F_SROA_0]] to %struct.S*
+; CHECK-NEXT:    br i1 undef, label [[IF_THEN5:%.*]], label [[IF_THEN2:%.*]]
+; CHECK:       if.then2:
+; CHECK-NEXT:    br label [[IF_THEN5]]
+; CHECK:       if.then5:
+; CHECK-NEXT:    [[F1:%.*]] = phi %struct.S* [ undef, [[IF_THEN2]] ], [ [[F_SROA_0_0_F2_SROA_CAST1]], [[IF_END]] ]
+; CHECK-NEXT:    [[DOTFCA_0_GEP:%.*]] = getelementptr inbounds [[STRUCT_S:%.*]], %struct.S* [[F1]], i32 0, i32 0
+; CHECK-NEXT:    store i32 undef, i32* [[DOTFCA_0_GEP]], align 4
+; CHECK-NEXT:    ret void
+;
 entry:
   %f = alloca %struct.S, align 4
-; CHECK: %[[alloca:.*]] = alloca
   br i1 undef, label %if.end, label %for.cond
 
 for.cond:                                         ; preds = %for.cond, %entry
@@ -636,8 +786,6 @@ for.cond:                                         ; preds = %for.cond, %entry
 
 if.end:                                           ; preds = %for.cond, %entry
   %f2 = phi %struct.S* [ %f, %entry ], [ %f, %for.cond ]
-; CHECK: phi i32
-; CHECK: %[[cast:.*]] = bitcast i32* %[[alloca]] to %struct.S*
   phi i32 [ undef, %entry ], [ undef, %for.cond ]
   br i1 undef, label %if.then5, label %if.then2
 
@@ -646,28 +794,45 @@ if.then2:                                         ; preds = %if.end
 
 if.then5:                                         ; preds = %if.then2, %if.end
   %f1 = phi %struct.S* [ undef, %if.then2 ], [ %f2, %if.end ]
-; CHECK: phi {{.*}} %[[cast]]
   store %struct.S undef, %struct.S* %f1, align 4
   ret void
 }
 
 define i32 @phi_align(i32* %z) {
 ; CHECK-LABEL: @phi_align(
+; CHECK-NEXT:  entry:
+; CHECK-NEXT:    [[A_SROA_0:%.*]] = alloca [7 x i8], align 1
+; CHECK-NEXT:    [[A_SROA_0_3_A1_SROA_IDX:%.*]] = getelementptr inbounds [7 x i8], [7 x i8]* [[A_SROA_0]], i64 0, i64 3
+; CHECK-NEXT:    [[A_SROA_0_3_A1_SROA_CAST:%.*]] = bitcast i8* [[A_SROA_0_3_A1_SROA_IDX]] to i32*
+; CHECK-NEXT:    [[A_SROA_0_0_A0_SROA_CAST:%.*]] = bitcast [7 x i8]* [[A_SROA_0]] to i32*
+; CHECK-NEXT:    store i32 0, i32* [[A_SROA_0_0_A0_SROA_CAST]], align 1
+; CHECK-NEXT:    [[A_SROA_0_3_A1_SROA_IDX7:%.*]] = getelementptr inbounds [7 x i8], [7 x i8]* [[A_SROA_0]], i64 0, i64 3
+; CHECK-NEXT:    [[A_SROA_0_3_A1_SROA_CAST8:%.*]] = bitcast i8* [[A_SROA_0_3_A1_SROA_IDX7]] to i32*
+; CHECK-NEXT:    store i32 1, i32* [[A_SROA_0_3_A1_SROA_CAST8]], align 1
+; CHECK-NEXT:    [[A_SROA_0_0_A0_SROA_CAST6:%.*]] = bitcast [7 x i8]* [[A_SROA_0]] to i32*
+; CHECK-NEXT:    [[A_SROA_0_0_A_SROA_0_1_V0:%.*]] = load i32, i32* [[A_SROA_0_0_A0_SROA_CAST6]], align 1
+; CHECK-NEXT:    [[A_SROA_0_3_A1_SROA_IDX9:%.*]] = getelementptr inbounds [7 x i8], [7 x i8]* [[A_SROA_0]], i64 0, i64 3
+; CHECK-NEXT:    [[A_SROA_0_3_A1_SROA_CAST10:%.*]] = bitcast i8* [[A_SROA_0_3_A1_SROA_IDX9]] to i32*
+; CHECK-NEXT:    [[A_SROA_0_3_A_SROA_0_4_V1:%.*]] = load i32, i32* [[A_SROA_0_3_A1_SROA_CAST10]], align 1
+; CHECK-NEXT:    [[COND:%.*]] = icmp sle i32 [[A_SROA_0_0_A_SROA_0_1_V0]], [[A_SROA_0_3_A_SROA_0_4_V1]]
+; CHECK-NEXT:    br i1 [[COND]], label [[THEN:%.*]], label [[EXIT:%.*]]
+; CHECK:       then:
+; CHECK-NEXT:    br label [[EXIT]]
+; CHECK:       exit:
+; CHECK-NEXT:    [[PHI:%.*]] = phi i32* [ [[A_SROA_0_3_A1_SROA_CAST]], [[THEN]] ], [ [[Z:%.*]], [[ENTRY:%.*]] ]
+; CHECK-NEXT:    [[RESULT:%.*]] = load i32, i32* [[PHI]], align 1
+; CHECK-NEXT:    ret i32 [[RESULT]]
+;
 entry:
   %a = alloca [8 x i8], align 8
-; CHECK: alloca [7 x i8]
 
   %a0x = getelementptr [8 x i8], [8 x i8]* %a, i64 0, i32 1
   %a0 = bitcast i8* %a0x to i32*
   %a1x = getelementptr [8 x i8], [8 x i8]* %a, i64 0, i32 4
   %a1 = bitcast i8* %a1x to i32*
-; CHECK: store i32 0, {{.*}}, align 1
   store i32 0, i32* %a0, align 1
-; CHECK: store i32 1, {{.*}}, align 1
   store i32 1, i32* %a1, align 4
-; CHECK: load {{.*}}, align 1
   %v0 = load i32, i32* %a0, align 1
-; CHECK: load {{.*}}, align 1
   %v1 = load i32, i32* %a1, align 4
   %cond = icmp sle i32 %v0, %v1
   br i1 %cond, label %then, label %exit
@@ -676,8 +841,6 @@ then:
   br label %exit
 
 exit:
-; CHECK: %phi = phi i32* [ {{.*}}, %then ], [ %z, %entry ]
-; CHECK-NEXT: %result = load i32, i32* %phi, align 1
   %phi = phi i32* [ %a1, %then ], [ %z, %entry ]
   %result = load i32, i32* %phi, align 4
   ret i32 %result
@@ -686,7 +849,13 @@ exit:
 ; Don't speculate a load based on an earlier volatile operation.
 define i8 @volatile_select(i8* %p, i1 %b) {
 ; CHECK-LABEL: @volatile_select(
-; CHECK: select i1 %b, i8* %p, i8* %p2
+; CHECK-NEXT:    [[P2:%.*]] = alloca i8, align 1
+; CHECK-NEXT:    store i8 0, i8* [[P2]], align 1
+; CHECK-NEXT:    store volatile i8 0, i8* [[P:%.*]], align 1
+; CHECK-NEXT:    [[PX:%.*]] = select i1 [[B:%.*]], i8* [[P]], i8* [[P2]]
+; CHECK-NEXT:    [[V2:%.*]] = load i8, i8* [[PX]], align 1
+; CHECK-NEXT:    ret i8 [[V2]]
+;
   %p2 = alloca i8
   store i8 0, i8* %p2
   store volatile i8 0, i8* %p
diff --git a/test/Transforms/SROA/pr37267.ll b/test/Transforms/SROA/pr37267.ll
index 4fcb1f29aac..479bf9f7b51 100644
--- a/test/Transforms/SROA/pr37267.ll
+++ b/test/Transforms/SROA/pr37267.ll
@@ -1,3 +1,4 @@
+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
 ; RUN: opt < %s -sroa -S | FileCheck %s
 target datalayout = "E-p:64:64:64-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f32:32:32-f64:64:64-v64:64:64-n32:64-S128"
 target triple = "sparcv9-sun-solaris"
@@ -6,9 +7,11 @@ target triple = "sparcv9-sun-solaris"
 ; Check that we don't crash on this test.
 
 define i16 @f1() {
-; CHECK-LABEL: @f1
-; CHECK: %[[retval:.*]] = add i16 2, 2
-; CHECK: ret i16 %[[retval]]
+; CHECK-LABEL: @f1(
+; CHECK-NEXT:  bb1:
+; CHECK-NEXT:    [[RC:%.*]] = add i16 2, 2
+; CHECK-NEXT:    ret i16 [[RC]]
+;
 
 bb1:
 ; This 12-byte alloca is split into partitions as [0,2), [2,4), [4,8), [8,10), [10, 12).
@@ -40,9 +43,18 @@ bb1:
 }
 
 define i16 @f2() {
-; CHECK-LABEL: @f2
-; CHECK: %[[retval:.*]] = add i16 2, undef
-; CHECK: ret i16 %[[retval]]
+; CHECK-LABEL: @f2(
+; CHECK-NEXT:  bb1:
+; CHECK-NEXT:    [[A_3_SROA_2_2_INSERT_EXT:%.*]] = zext i16 undef to i32
+; CHECK-NEXT:    [[A_3_SROA_2_2_INSERT_MASK:%.*]] = and i32 undef, -65536
+; CHECK-NEXT:    [[A_3_SROA_2_2_INSERT_INSERT:%.*]] = or i32 [[A_3_SROA_2_2_INSERT_MASK]], [[A_3_SROA_2_2_INSERT_EXT]]
+; CHECK-NEXT:    [[A_3_SROA_0_2_INSERT_EXT:%.*]] = zext i16 undef to i32
+; CHECK-NEXT:    [[A_3_SROA_0_2_INSERT_SHIFT:%.*]] = shl i32 [[A_3_SROA_0_2_INSERT_EXT]], 16
+; CHECK-NEXT:    [[A_3_SROA_0_2_INSERT_MASK:%.*]] = and i32 [[A_3_SROA_2_2_INSERT_INSERT]], 65535
+; CHECK-NEXT:    [[A_3_SROA_0_2_INSERT_INSERT:%.*]] = or i32 [[A_3_SROA_0_2_INSERT_MASK]], [[A_3_SROA_0_2_INSERT_SHIFT]]
+; CHECK-NEXT:    [[RC:%.*]] = add i16 2, undef
+; CHECK-NEXT:    ret i16 [[RC]]
+;
 
 bb1:
 ; This 12-byte alloca is split into partitions as [0,2), [2,4), [4,8), [8,10), [10, 12).