diff --git a/llvm/test/Transforms/GVN/PRE/load-pre-nonlocal.ll b/llvm/test/Transforms/GVN/PRE/load-pre-nonlocal.ll
--- a/llvm/test/Transforms/GVN/PRE/load-pre-nonlocal.ll
+++ b/llvm/test/Transforms/GVN/PRE/load-pre-nonlocal.ll
@@ -1,3 +1,4 @@
+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
 ; RUN: opt -S -o - -basic-aa -domtree -gvn %s | FileCheck %s
 
 target datalayout = "e-m:o-i64:64-f80:128-n8:16:32:64-S128"
@@ -10,13 +11,35 @@
 
 ; Check that GVN doesn't determine %2 is partially redundant.
 
-; CHECK-LABEL: define i32 @volatile_load
-; CHECK: for.body:
-; CHECK: %2 = load i32, i32*
-; CHECK: %3 = load volatile i32, i32*
-; CHECK: for.cond.for.end_crit_edge:
-
 define i32 @volatile_load(i32 %n) {
+; CHECK-LABEL: @volatile_load(
+; CHECK-NEXT:  entry:
+; CHECK-NEXT:    [[CMP6:%.*]] = icmp sgt i32 [[N:%.*]], 0
+; CHECK-NEXT:    br i1 [[CMP6]], label [[FOR_BODY_LR_PH:%.*]], label [[FOR_END:%.*]]
+; CHECK:       for.body.lr.ph:
+; CHECK-NEXT:    [[TMP0:%.*]] = load i32*, i32** @a2, align 8, !tbaa [[TBAA5:![0-9]+]]
+; CHECK-NEXT:    [[TMP1:%.*]] = load i32*, i32** @a, align 8, !tbaa [[TBAA5]]
+; CHECK-NEXT:    br label [[FOR_BODY:%.*]]
+; CHECK:       for.body:
+; CHECK-NEXT:    [[INDVARS_IV:%.*]] = phi i64 [ 0, [[FOR_BODY_LR_PH]] ], [ [[INDVARS_IV_NEXT:%.*]], [[FOR_BODY]] ]
+; CHECK-NEXT:    [[S_09:%.*]] = phi i32 [ 0, [[FOR_BODY_LR_PH]] ], [ [[ADD:%.*]], [[FOR_BODY]] ]
+; CHECK-NEXT:    [[P_08:%.*]] = phi i32* [ [[TMP0]], [[FOR_BODY_LR_PH]] ], [ [[INCDEC_PTR:%.*]], [[FOR_BODY]] ]
+; CHECK-NEXT:    [[TMP2:%.*]] = load i32, i32* [[P_08]], align 4, !tbaa [[TBAA9:![0-9]+]]
+; CHECK-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds i32, i32* [[TMP1]], i64 [[INDVARS_IV]]
+; CHECK-NEXT:    store i32 [[TMP2]], i32* [[ARRAYIDX]], align 4, !tbaa [[TBAA9]]
+; CHECK-NEXT:    [[TMP3:%.*]] = load volatile i32, i32* [[P_08]], align 4, !tbaa [[TBAA9]]
+; CHECK-NEXT:    [[ADD]] = add nsw i32 [[TMP3]], [[S_09]]
+; CHECK-NEXT:    [[INDVARS_IV_NEXT]] = add nuw nsw i64 [[INDVARS_IV]], 1
+; CHECK-NEXT:    [[INCDEC_PTR]] = getelementptr inbounds i32, i32* [[P_08]], i64 1
+; CHECK-NEXT:    [[LFTR_WIDEIV:%.*]] = trunc i64 [[INDVARS_IV_NEXT]] to i32
+; CHECK-NEXT:    [[EXITCOND:%.*]] = icmp ne i32 [[LFTR_WIDEIV]], [[N]]
+; CHECK-NEXT:    br i1 [[EXITCOND]], label [[FOR_BODY]], label [[FOR_COND_FOR_END_CRIT_EDGE:%.*]]
+; CHECK:       for.cond.for.end_crit_edge:
+; CHECK-NEXT:    br label [[FOR_END]]
+; CHECK:       for.end:
+; CHECK-NEXT:    [[S_0_LCSSA:%.*]] = phi i32 [ [[ADD]], [[FOR_COND_FOR_END_CRIT_EDGE]] ], [ 0, [[ENTRY:%.*]] ]
+; CHECK-NEXT:    ret i32 [[S_0_LCSSA]]
+;
 entry:
   %cmp6 = icmp sgt i32 %n, 0
   br i1 %cmp6, label %for.body.lr.ph, label %for.end
@@ -53,16 +76,26 @@
 ; %1 is partially redundant if %0 can be widened to a 64-bit load.
 ; But we should not widen %0 to 64-bit load.
 
-; CHECK-LABEL: define i32 @overaligned_load
-; CHECK: if.then:
-; CHECK-NOT:   %0 = load i64
-; CHECK-NOT:   [[LSHR:%[0-9]+]] = lshr i64 %0, 32, !dbg [[LSHR_LOC:![0-9]+]]
-; CHECK-NOT:   trunc i64 [[LSHR]] to i32
-; CHECK: if.end:
-; CHECK: %1 = load i32, i32*
-; CHECK-NOT: [[LSHR_LOC]] = !DILocation(line: 101, column: 1, scope: !{{.*}})
-
 define i32 @overaligned_load(i32 %a, i32* nocapture %b) !dbg !13 {
+; CHECK-LABEL: @overaligned_load(
+; CHECK-NEXT:  entry:
+; CHECK-NEXT:    [[CMP:%.*]] = icmp sgt i32 [[A:%.*]], 0, !dbg [[DBG14:![0-9]+]]
+; CHECK-NEXT:    br i1 [[CMP]], label [[IF_THEN:%.*]], label [[IF_ELSE:%.*]], !dbg [[DBG14]]
+; CHECK:       if.then:
+; CHECK-NEXT:    [[TMP0:%.*]] = load i32, i32* getelementptr inbounds ([[STRUCT_S1:%.*]], %struct.S1* @s1, i64 0, i32 0), align 8, !dbg [[DBG15:![0-9]+]], !tbaa [[TBAA9]]
+; CHECK-NEXT:    br label [[IF_END:%.*]], !dbg [[DBG15]]
+; CHECK:       if.else:
+; CHECK-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds i32, i32* [[B:%.*]], i64 2, !dbg [[DBG16:![0-9]+]]
+; CHECK-NEXT:    store i32 10, i32* [[ARRAYIDX]], align 4, !dbg [[DBG16]], !tbaa [[TBAA9]]
+; CHECK-NEXT:    br label [[IF_END]], !dbg [[DBG16]]
+; CHECK:       if.end:
+; CHECK-NEXT:    [[I_0:%.*]] = phi i32 [ [[TMP0]], [[IF_THEN]] ], [ 0, [[IF_ELSE]] ]
+; CHECK-NEXT:    [[P_0:%.*]] = phi i32* [ getelementptr inbounds ([[STRUCT_S1]], %struct.S1* @s1, i64 0, i32 0), [[IF_THEN]] ], [ [[B]], [[IF_ELSE]] ]
+; CHECK-NEXT:    [[ADD_PTR:%.*]] = getelementptr inbounds i32, i32* [[P_0]], i64 1, !dbg [[DBG17:![0-9]+]]
+; CHECK-NEXT:    [[TMP1:%.*]] = load i32, i32* [[ADD_PTR]], align 4, !dbg [[DBG17]], !tbaa [[TBAA9]]
+; CHECK-NEXT:    [[ADD1:%.*]] = add nsw i32 [[TMP1]], [[I_0]], !dbg [[DBG17]]
+; CHECK-NEXT:    ret i32 [[ADD1]], !dbg [[DBG17]]
+;
 entry:
   %cmp = icmp sgt i32 %a, 0, !dbg !14
   br i1 %cmp, label %if.then, label %if.else, !dbg !14
@@ -107,6 +140,6 @@
 !16 = !DILocation(line: 102, column: 1, scope: !13)
 !17 = !DILocation(line: 103, column: 1, scope: !13)
 !18 = distinct !DICompileUnit(language: DW_LANG_C99, producer: "clang",
-                             file: !12,
-                             isOptimized: true, flags: "-O2",
-                             splitDebugFilename: "abc.debug", emissionKind: 2)
+  file: !12,
+  isOptimized: true, flags: "-O2",
+  splitDebugFilename: "abc.debug", emissionKind: 2)
diff --git a/llvm/test/Transforms/GVN/big-endian.ll b/llvm/test/Transforms/GVN/big-endian.ll
--- a/llvm/test/Transforms/GVN/big-endian.ll
+++ b/llvm/test/Transforms/GVN/big-endian.ll
@@ -1,15 +1,29 @@
+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
 ; RUN: opt -gvn -S < %s | FileCheck %s
 
-target datalayout = "E-m:e-i64:64-n32:64"                                                                                         
-target triple = "powerpc64-unknown-linux-gnu"                                                                                     
+target datalayout = "E-m:e-i64:64-n32:64"
+target triple = "powerpc64-unknown-linux-gnu"
 
 ;; Make sure we use correct bit shift based on storage size for
 ;; loads reusing a load value.
 define i64 @test1({ i1, i8 }* %predA, { i1, i8 }* %predB) {
-; CHECK-LABEL: @test1
-; CHECK-NOT: [[V1:%.*]] = load i16, i16* %{{.*}}
-; CHECK-NOT: [[V2:%.*]] = lshr i16 [[V1]], 8
-; CHECK-NOT: trunc i16 [[V2]] to i1
+; CHECK-LABEL: @test1(
+; CHECK-NEXT:    [[VALUELOADA_FCA_0_GEP:%.*]] = getelementptr inbounds { i1, i8 }, { i1, i8 }* [[PREDA:%.*]], i64 0, i32 0
+; CHECK-NEXT:    [[VALUELOADA_FCA_0_LOAD:%.*]] = load i1, i1* [[VALUELOADA_FCA_0_GEP]], align 8
+; CHECK-NEXT:    [[VALUELOADB_FCA_0_GEP:%.*]] = getelementptr inbounds { i1, i8 }, { i1, i8 }* [[PREDB:%.*]], i64 0, i32 0
+; CHECK-NEXT:    [[VALUELOADB_FCA_0_LOAD:%.*]] = load i1, i1* [[VALUELOADB_FCA_0_GEP]], align 8
+; CHECK-NEXT:    [[ISTRUE:%.*]] = and i1 [[VALUELOADA_FCA_0_LOAD]], [[VALUELOADB_FCA_0_LOAD]]
+; CHECK-NEXT:    [[VALUELOADA_FCA_1_GEP:%.*]] = getelementptr inbounds { i1, i8 }, { i1, i8 }* [[PREDA]], i64 0, i32 1
+; CHECK-NEXT:    [[VALUELOADA_FCA_1_LOAD:%.*]] = load i8, i8* [[VALUELOADA_FCA_1_GEP]], align 1
+; CHECK-NEXT:    [[ISNOTNULLA:%.*]] = icmp ne i8 [[VALUELOADA_FCA_1_LOAD]], 0
+; CHECK-NEXT:    [[VALUELOADB_FCA_1_GEP:%.*]] = getelementptr inbounds { i1, i8 }, { i1, i8 }* [[PREDB]], i64 0, i32 1
+; CHECK-NEXT:    [[VALUELOADB_FCA_1_LOAD:%.*]] = load i8, i8* [[VALUELOADB_FCA_1_GEP]], align 1
+; CHECK-NEXT:    [[ISNOTNULLB:%.*]] = icmp ne i8 [[VALUELOADB_FCA_1_LOAD]], 0
+; CHECK-NEXT:    [[ISNOTNULL:%.*]] = and i1 [[ISNOTNULLA]], [[ISNOTNULLB]]
+; CHECK-NEXT:    [[ISTRUEANDNOTNULL:%.*]] = and i1 [[ISTRUE]], [[ISNOTNULL]]
+; CHECK-NEXT:    [[RET:%.*]] = zext i1 [[ISTRUEANDNOTNULL]] to i64
+; CHECK-NEXT:    ret i64 [[RET]]
+;
 
   %valueLoadA.fca.0.gep = getelementptr inbounds { i1, i8 }, { i1, i8 }* %predA, i64 0, i32 0
   %valueLoadA.fca.0.load = load i1, i1* %valueLoadA.fca.0.gep, align 8
@@ -30,8 +44,12 @@
 
 ;; And likewise for loads reusing a store value.
 define i1 @test2(i8 %V, i8* %P) {
-; CHECK-LABEL: @test2
-; CHECK-NOT: lshr
+; CHECK-LABEL: @test2(
+; CHECK-NEXT:    store i8 [[V:%.*]], i8* [[P:%.*]], align 1
+; CHECK-NEXT:    [[P2:%.*]] = bitcast i8* [[P]] to i1*
+; CHECK-NEXT:    [[TMP1:%.*]] = trunc i8 [[V]] to i1
+; CHECK-NEXT:    ret i1 [[TMP1]]
+;
   store i8 %V, i8* %P
   %P2 = bitcast i8* %P to i1*
   %A = load i1, i1* %P2
diff --git a/llvm/test/Transforms/NewGVN/big-endian.ll b/llvm/test/Transforms/NewGVN/big-endian.ll
--- a/llvm/test/Transforms/NewGVN/big-endian.ll
+++ b/llvm/test/Transforms/NewGVN/big-endian.ll
@@ -1,15 +1,29 @@
+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
 ; RUN: opt -newgvn -S < %s | FileCheck %s
 
-target datalayout = "E-m:e-i64:64-n32:64"                                                                                         
-target triple = "powerpc64-unknown-linux-gnu"                                                                                     
+target datalayout = "E-m:e-i64:64-n32:64"
+target triple = "powerpc64-unknown-linux-gnu"
 
 ;; Make sure we use correct bit shift based on storage size for
 ;; loads reusing a load value.
 define i64 @test1({ i1, i8 }* %predA, { i1, i8 }* %predB) {
-; CHECK-LABEL: @test1
-; CHECK-NOT: [[V1:%.*]] = load i16, i16* %{{.*}}
-; CHECK-NOT: [[V2:%.*]] = lshr i16 [[V1]], 8
-; CHECK-NOT: trunc i16 [[V2]] to i1
+; CHECK-LABEL: @test1(
+; CHECK-NEXT:    [[VALUELOADA_FCA_0_GEP:%.*]] = getelementptr inbounds { i1, i8 }, { i1, i8 }* [[PREDA:%.*]], i64 0, i32 0
+; CHECK-NEXT:    [[VALUELOADA_FCA_0_LOAD:%.*]] = load i1, i1* [[VALUELOADA_FCA_0_GEP]], align 8
+; CHECK-NEXT:    [[VALUELOADB_FCA_0_GEP:%.*]] = getelementptr inbounds { i1, i8 }, { i1, i8 }* [[PREDB:%.*]], i64 0, i32 0
+; CHECK-NEXT:    [[VALUELOADB_FCA_0_LOAD:%.*]] = load i1, i1* [[VALUELOADB_FCA_0_GEP]], align 8
+; CHECK-NEXT:    [[ISTRUE:%.*]] = and i1 [[VALUELOADA_FCA_0_LOAD]], [[VALUELOADB_FCA_0_LOAD]]
+; CHECK-NEXT:    [[VALUELOADA_FCA_1_GEP:%.*]] = getelementptr inbounds { i1, i8 }, { i1, i8 }* [[PREDA]], i64 0, i32 1
+; CHECK-NEXT:    [[VALUELOADA_FCA_1_LOAD:%.*]] = load i8, i8* [[VALUELOADA_FCA_1_GEP]], align 1
+; CHECK-NEXT:    [[ISNOTNULLA:%.*]] = icmp ne i8 [[VALUELOADA_FCA_1_LOAD]], 0
+; CHECK-NEXT:    [[VALUELOADB_FCA_1_GEP:%.*]] = getelementptr inbounds { i1, i8 }, { i1, i8 }* [[PREDB]], i64 0, i32 1
+; CHECK-NEXT:    [[VALUELOADB_FCA_1_LOAD:%.*]] = load i8, i8* [[VALUELOADB_FCA_1_GEP]], align 1
+; CHECK-NEXT:    [[ISNOTNULLB:%.*]] = icmp ne i8 [[VALUELOADB_FCA_1_LOAD]], 0
+; CHECK-NEXT:    [[ISNOTNULL:%.*]] = and i1 [[ISNOTNULLA]], [[ISNOTNULLB]]
+; CHECK-NEXT:    [[ISTRUEANDNOTNULL:%.*]] = and i1 [[ISTRUE]], [[ISNOTNULL]]
+; CHECK-NEXT:    [[RET:%.*]] = zext i1 [[ISTRUEANDNOTNULL]] to i64
+; CHECK-NEXT:    ret i64 [[RET]]
+;
 
   %valueLoadA.fca.0.gep = getelementptr inbounds { i1, i8 }, { i1, i8 }* %predA, i64 0, i32 0
   %valueLoadA.fca.0.load = load i1, i1* %valueLoadA.fca.0.gep, align 8
@@ -30,8 +44,12 @@
 
 ;; And likewise for loads reusing a store value.
 define i1 @test2(i8 %V, i8* %P) {
-; CHECK-LABEL: @test2
-; CHECK-NOT: lshr
+; CHECK-LABEL: @test2(
+; CHECK-NEXT:    store i8 [[V:%.*]], i8* [[P:%.*]], align 1
+; CHECK-NEXT:    [[P2:%.*]] = bitcast i8* [[P]] to i1*
+; CHECK-NEXT:    [[A:%.*]] = load i1, i1* [[P2]], align 1
+; CHECK-NEXT:    ret i1 [[A]]
+;
   store i8 %V, i8* %P
   %P2 = bitcast i8* %P to i1*
   %A = load i1, i1* %P2