Change memcpy/memset/memmove to have dest and source alignments.

This is a follow on from a similar LLVM commit: r253511.

Note, this was reviewed (and more details are in) http://lists.llvm.org/pipermail/llvm-commits/Week-of-Mon-20151109/312083.html

These intrinsics currently have an explicit alignment argument which is
required to be a constant integer.  It represents the alignment of the
source and dest, and so must be the minimum of those.

This change allows source and dest to each have their own alignments
by using the alignment attribute on their arguments.  The alignment
argument itself is removed.

The only code change to clang is hidden in CGBuilder.h which now passes
both dest and source alignment to IRBuilder, instead of taking the minimum of
dest and source alignments.

Reviewed by Hal Finkel.

llvm-svn: 253512

23 files changed, 116 insertions, 116 deletions

diff --git a/clang/test/CodeGen/2007-11-07-CopyAggregateAlign.c b/clang/test/CodeGen/2007-11-07-CopyAggregateAlign.c
index 08d97702e69..b39768c4c2b 100644
--- a/clang/test/CodeGen/2007-11-07-CopyAggregateAlign.c
+++ b/clang/test/CodeGen/2007-11-07-CopyAggregateAlign.c
@@ -2,6 +2,6 @@
 struct A { char s, t, u, v; short a; };
 // CHECK: %a = alloca %struct.A, align 2
 // CHECK: %b = alloca %struct.A, align 2
-// CHECK: call void @llvm.memcpy.p0i8.p0i8.{{.*}}, i32 2, i1 false)
+// CHECK: call void @llvm.memcpy.p0i8.p0i8.i{{[0-9]*}}(i8* align 2 %{{[0-9]*}}, i8* align 2 %{{[0-9]*}}, {{.*}}, i1 false)
 
 void q() { struct A a, b; a = b; }
diff --git a/clang/test/CodeGen/2007-11-07-ZeroAggregateAlign.c b/clang/test/CodeGen/2007-11-07-ZeroAggregateAlign.c
index b059607ba58..eda8b3c4a77 100644
--- a/clang/test/CodeGen/2007-11-07-ZeroAggregateAlign.c
+++ b/clang/test/CodeGen/2007-11-07-ZeroAggregateAlign.c
@@ -1,5 +1,5 @@
 // RUN: %clang_cc1 -emit-llvm %s -o - | FileCheck %s
 struct A { short s; short t; int i; };
 // CHECK: %a = alloca %struct.A, align 4
-// CHECK: call void @llvm.memset.p0i8.{{.*}}i32 4, i1 false)
+// // CHECK: call void @llvm.memset.p0i8.i{{[0-9]*}}(i8* align 4 %{{[0-9]*}}, {{.*}}, i1 false)
 void q() { struct A a = {0}; }
diff --git a/clang/test/CodeGen/arm-arguments.c b/clang/test/CodeGen/arm-arguments.c
index ec3e1734b0c..ea2ace66cbd 100644
--- a/clang/test/CodeGen/arm-arguments.c
+++ b/clang/test/CodeGen/arm-arguments.c
@@ -208,13 +208,13 @@ float32x4_t f35(int i, s35_with_align s1, s35_with_align s2) {
 // APCS-GNU: %[[a:.*]] = alloca %struct.s35, align 16
 // APCS-GNU: %[[b:.*]] = bitcast %struct.s35* %[[a]] to i8*
 // APCS-GNU: %[[c:.*]] = bitcast %struct.s35* %0 to i8*
-// APCS-GNU: call void @llvm.memcpy.p0i8.p0i8.i32(i8* %[[b]], i8* %[[c]]
+// APCS-GNU: call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 16 %[[b]], i8* align 4 %[[c]]
 // APCS-GNU: %[[d:.*]] = bitcast %struct.s35* %[[a]] to <4 x float>*
 // APCS-GNU: load <4 x float>, <4 x float>* %[[d]], align 16
 // AAPCS-LABEL: define arm_aapcscc <4 x float> @f35(i32 %i, %struct.s35* byval align 8, %struct.s35* byval align 8)
 // AAPCS: %[[a:.*]] = alloca %struct.s35, align 16
 // AAPCS: %[[b:.*]] = bitcast %struct.s35* %[[a]] to i8*
 // AAPCS: %[[c:.*]] = bitcast %struct.s35* %0 to i8*
-// AAPCS: call void @llvm.memcpy.p0i8.p0i8.i32(i8* %[[b]], i8* %[[c]]
+// AAPCS: call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 16 %[[b]], i8* align 8 %[[c]]
 // AAPCS: %[[d:.*]] = bitcast %struct.s35* %[[a]] to <4 x float>*
 // AAPCS: load <4 x float>, <4 x float>* %[[d]], align 16
diff --git a/clang/test/CodeGen/arm64-be-bitfield.c b/clang/test/CodeGen/arm64-be-bitfield.c
index 132239ab831..9a39d31d33a 100644
--- a/clang/test/CodeGen/arm64-be-bitfield.c
+++ b/clang/test/CodeGen/arm64-be-bitfield.c
@@ -9,7 +9,7 @@ signed callee_b0f(struct bt3 bp11) {
 // IR: callee_b0f(i64 [[ARG:%.*]])
 // IR: store i64 [[ARG]], i64* [[PTR:%.*]], align 8
 // IR: [[BITCAST:%.*]] = bitcast i64* [[PTR]] to i8*
-// IR: call void @llvm.memcpy.p0i8.p0i8.i64(i8* {{.*}}, i8* [[BITCAST]], i64 4
+// IR: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 4 {{.*}}, i8* align 8 [[BITCAST]], i64 4
 // ARM: asr x0, x0, #54
   return bp11.b2;
 }
diff --git a/clang/test/CodeGen/atomic-arm64.c b/clang/test/CodeGen/atomic-arm64.c
index 5cae3d13498..0e79846085c 100644
--- a/clang/test/CodeGen/atomic-arm64.c
+++ b/clang/test/CodeGen/atomic-arm64.c
@@ -65,7 +65,7 @@ void test3(pointer_pair_t pair) {
 // CHECK:      [[TEMP:%.*]] = alloca [[QUAD_T:%.*]], align 8
 // CHECK-NEXT: [[T0:%.*]] = bitcast [[QUAD_T]]* [[TEMP]] to i8*
 // CHECK-NEXT: [[T1:%.*]] = bitcast [[QUAD_T]]* {{%.*}} to i8*
-// CHECK-NEXT: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[T0]], i8* [[T1]], i64 32, i32 8, i1 false)
+// CHECK-NEXT: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 8 [[T0]], i8* align 8 [[T1]], i64 32, i1 false)
 // CHECK-NEXT: [[T0:%.*]] = bitcast [[QUAD_T]]* [[TEMP]] to i256*
 // CHECK-NEXT: [[T1:%.*]] = bitcast i256* [[T0]] to i8*
 // CHECK-NEXT: call void @__atomic_store(i64 32, i8* bitcast ([[QUAD_T]]* @a_pointer_quad to i8*), i8* [[T1]], i32 5)
diff --git a/clang/test/CodeGen/block-byref-aggr.c b/clang/test/CodeGen/block-byref-aggr.c
index 7d146a2d477..962f100638f 100644
--- a/clang/test/CodeGen/block-byref-aggr.c
+++ b/clang/test/CodeGen/block-byref-aggr.c
@@ -24,7 +24,7 @@ void test0() {
 // CHECK-NEXT: [[T1:%.*]] = getelementptr inbounds [[BYREF]], [[BYREF]]* [[T0]], i32 0, i32 4
 // CHECK-NEXT: [[T2:%.*]] = bitcast [[AGG]]* [[T1]] to i8*
 // CHECK-NEXT: [[T3:%.*]] = bitcast [[AGG]]* [[TEMP]] to i8*
-// CHECK-NEXT: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[T2]], i8* [[T3]], i64 4, i32 4, i1 false)
+// CHECK-NEXT: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 8 [[T2]], i8* align 4 [[T3]], i64 4, i1 false)
 //   Verify that there's nothing else significant in the function.
 // CHECK-NEXT: [[T0:%.*]] = bitcast [[BYREF]]* [[A]] to i8*
 // CHECK-NEXT: call void @_Block_object_dispose(i8* [[T0]], i32 8)
@@ -50,14 +50,14 @@ void test1() {
 // CHECK-NEXT: [[T1:%.*]] = getelementptr inbounds [[B_BYREF]], [[B_BYREF]]* [[T0]], i32 0, i32 4
 // CHECK-NEXT: [[T2:%.*]] = bitcast [[AGG]]* [[T1]] to i8*
 // CHECK-NEXT: [[T3:%.*]] = bitcast [[AGG]]* [[TEMP]] to i8*
-// CHECK-NEXT: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[T2]], i8* [[T3]], i64 4, i32 4, i1 false)
+// CHECK-NEXT: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 8 [[T2]], i8* align 4 [[T3]], i64 4, i1 false)
 //   Then for 'a':
 // CHECK-NEXT: [[A_FORWARDING:%.*]] = getelementptr inbounds [[A_BYREF]], [[A_BYREF]]* [[A]], i32 0, i32 1
 // CHECK-NEXT: [[T0:%.*]] = load [[A_BYREF]]*, [[A_BYREF]]** [[A_FORWARDING]]
 // CHECK-NEXT: [[T1:%.*]] = getelementptr inbounds [[A_BYREF]], [[A_BYREF]]* [[T0]], i32 0, i32 4
 // CHECK-NEXT: [[T2:%.*]] = bitcast [[AGG]]* [[T1]] to i8*
 // CHECK-NEXT: [[T3:%.*]] = bitcast [[AGG]]* [[TEMP]] to i8*
-// CHECK-NEXT: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[T2]], i8* [[T3]], i64 4, i32 4, i1 false)
+// CHECK-NEXT: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 8 [[T2]], i8* align 4 [[T3]], i64 4, i1 false)
 //   Verify that there's nothing else significant in the function.
 // CHECK-NEXT: [[T0:%.*]] = bitcast [[B_BYREF]]* [[B]] to i8*
 // CHECK-NEXT: call void @_Block_object_dispose(i8* [[T0]], i32 8)
diff --git a/clang/test/CodeGen/builtin-memfns.c b/clang/test/CodeGen/builtin-memfns.c
index 4a06160ccbc..d19963eef75 100644
--- a/clang/test/CodeGen/builtin-memfns.c
+++ b/clang/test/CodeGen/builtin-memfns.c
@@ -53,15 +53,15 @@ int test6(char *X) {
 // PR12094
 int test7(int *p) {
   struct snd_pcm_hw_params_t* hwparams;  // incomplete type.
-  
-  // CHECK: call void @llvm.memset{{.*}}256, i32 4, i1 false)
+
+  // CHECK: call void @llvm.memset{{.*}}(i8* align 4 %{{[0-9]*}}, {{.*}}, i{{[0-9]*}} 256, i1 false)
   __builtin_memset(p, 0, 256);  // Should be alignment = 4
 
-  // CHECK: call void @llvm.memset{{.*}}256, i32 1, i1 false)
+  // CHECK: call void @llvm.memset{{.*}}(i8* align 1 %{{[0-9]*}}, {{.*}}, i{{[0-9]*}} 256, i1 false)
   __builtin_memset((char*)p, 0, 256);  // Should be alignment = 1
 
   __builtin_memset(hwparams, 0, 256);  // No crash alignment = 1
-  // CHECK: call void @llvm.memset{{.*}}256, i32 1, i1 false)
+  // CHECK: call void @llvm.memset{{.*}}(i8* align 1 %{{[0-9]*}}, {{.*}}, i{{[0-9]*}} 256, i1 false)
 }
 
 // <rdar://problem/11314941>
@@ -73,13 +73,13 @@ struct PS {
 struct PS ps;
 void test8(int *arg) {
   // CHECK: @test8
-  // CHECK: call void @llvm.memcpy{{.*}} 16, i32 1, i1 false)
+  // CHECK: call void @llvm.memcpy{{.*}}(i8* align 4 {{.*}}, i8* align 1 {{.*}}, i{{[0-9]*}} 16, i1 false)
   __builtin_memcpy(arg, ps.modes, sizeof(struct PS));
 }
 
 __attribute((aligned(16))) int x[4], y[4];
 void test9() {
   // CHECK: @test9
-  // CHECK: call void @llvm.memcpy{{.*}} 16, i32 16, i1 false)
+  // CHECK: call void @llvm.memcpy{{.*}}(i8* align 16 {{.*}}, i8* align 16 {{.*}}, i{{[0-9]*}} 16, i1 false)
   __builtin_memcpy(x, y, sizeof(y));
 }
diff --git a/clang/test/CodeGen/c11atomics-ios.c b/clang/test/CodeGen/c11atomics-ios.c
index fb731dfd7a1..0e24ed228ca 100644
--- a/clang/test/CodeGen/c11atomics-ios.c
+++ b/clang/test/CodeGen/c11atomics-ios.c
@@ -132,7 +132,7 @@ void testStruct(_Atomic(S) *fp) {
 // CHECK-NEXT: [[T0:%.*]] = load [[S]]*, [[S]]** [[FP]]
 // CHECK-NEXT: [[T1:%.*]] = bitcast [[S]]* [[TMP0]] to i8*
 // CHECK-NEXT: [[T2:%.*]] = bitcast [[S]]* [[F]] to i8*
-// CHECK-NEXT: call void @llvm.memcpy.p0i8.p0i8.i32(i8* [[T1]], i8* [[T2]], i32 8, i32 2, i1 false)
+// CHECK-NEXT: call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 8 [[T1]], i8* align 2 [[T2]], i32 8, i1 false)
 // CHECK-NEXT: [[T3:%.*]] = bitcast [[S]]* [[TMP0]] to i64*
 // CHECK-NEXT: [[T4:%.*]] = load i64, i64* [[T3]], align 8
 // CHECK-NEXT: [[T5:%.*]] = bitcast [[S]]* [[T0]] to i64*
@@ -154,7 +154,7 @@ void testPromotedStruct(_Atomic(PS) *fp) {
 
 // CHECK-NEXT: [[P:%.*]] = load [[APS]]*, [[APS]]** [[FP]]
 // CHECK-NEXT: [[T0:%.*]] = bitcast [[APS]]* [[P]] to i8*
-// CHECK-NEXT: call void @llvm.memset.p0i8.i64(i8* [[T0]], i8 0, i64 8, i32 8, i1 false)
+// CHECK-NEXT: call void @llvm.memset.p0i8.i64(i8* align 8 [[T0]], i8 0, i64 8, i1 false)
 // CHECK-NEXT: [[T0:%.*]] = getelementptr inbounds [[APS]], [[APS]]* [[P]], i32 0, i32 0
 // CHECK-NEXT: [[T1:%.*]] = getelementptr inbounds [[PS]], [[PS]]* [[T0]], i32 0, i32 0
 // CHECK-NEXT: store i16 1, i16* [[T1]], align 8
@@ -165,7 +165,7 @@ void testPromotedStruct(_Atomic(PS) *fp) {
   __c11_atomic_init(fp, (PS){1,2,3});
 
 // CHECK-NEXT: [[T0:%.*]] = bitcast [[APS]]* [[X]] to i8*
-// CHECK-NEXT: call void @llvm.memset.p0i8.i32(i8* [[T0]], i8 0, i32 8, i32 8, i1 false)
+// CHECK-NEXT: call void @llvm.memset.p0i8.i32(i8* align 8 [[T0]], i8 0, i32 8, i1 false)
 // CHECK-NEXT: [[T0:%.*]] = getelementptr inbounds [[APS]], [[APS]]* [[X]], i32 0, i32 0
 // CHECK-NEXT: [[T1:%.*]] = getelementptr inbounds [[PS]], [[PS]]* [[T0]], i32 0, i32 0
 // CHECK-NEXT: store i16 1, i16* [[T1]], align 8
@@ -183,16 +183,16 @@ void testPromotedStruct(_Atomic(PS) *fp) {
 // CHECK-NEXT: [[T0:%.*]] = getelementptr inbounds [[APS]], [[APS]]* [[TMP0]], i32 0, i32 0
 // CHECK-NEXT: [[T1:%.*]] = bitcast [[PS]]* [[F]] to i8*
 // CHECK-NEXT: [[T2:%.*]] = bitcast [[PS]]* [[T0]] to i8*
-// CHECK-NEXT: call void @llvm.memcpy.p0i8.p0i8.i32(i8* [[T1]], i8* [[T2]], i32 6, i32 2, i1 false)
+// CHECK-NEXT: call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 2 [[T1]], i8* align 8 [[T2]], i32 6, i1 false)
   PS f = *fp;
 
 // CHECK-NEXT: [[T0:%.*]] = load [[APS]]*, [[APS]]** [[FP]]
 // CHECK-NEXT: [[T1:%.*]] = bitcast { %struct.PS, [2 x i8] }* [[TMP1]] to i8*
-// CHECK-NEXT: call void @llvm.memset.p0i8.i32(i8* [[T1]], i8 0, i32 8, i32 8, i1 false)
+// CHECK-NEXT: call void @llvm.memset.p0i8.i32(i8* align 8 [[T1]], i8 0, i32 8, i1 false)
 // CHECK-NEXT: [[T1:%.*]] = getelementptr inbounds [[APS]], [[APS]]* [[TMP1]], i32 0, i32 0
 // CHECK-NEXT: [[T2:%.*]] = bitcast [[PS]]* [[T1]] to i8*
 // CHECK-NEXT: [[T3:%.*]] = bitcast [[PS]]* [[F]] to i8*
-// CHECK-NEXT: call void @llvm.memcpy.p0i8.p0i8.i32(i8* [[T2]], i8* [[T3]], i32 6, i32 2, i1 false)
+// CHECK-NEXT: call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 8 [[T2]], i8* align 2 [[T3]], i32 6, i1 false)
 // CHECK-NEXT: [[T4:%.*]] = bitcast [[APS]]* [[TMP1]] to i64*
 // CHECK-NEXT: [[T5:%.*]] = load i64, i64* [[T4]], align 8
 // CHECK-NEXT: [[T6:%.*]] = bitcast [[APS]]* [[T0]] to i64*
@@ -215,7 +215,7 @@ PS test_promoted_load(_Atomic(PS) *addr) {
   // CHECK:   [[ATOMIC_RES_STRUCT:%.*]] = bitcast i64* [[ATOMIC_RES64]] to %struct.PS*
   // CHECK:   [[AGG_RESULT8:%.*]] = bitcast %struct.PS* %agg.result to i8*
   // CHECK:   [[ATOMIC_RES8:%.*]] = bitcast %struct.PS* [[ATOMIC_RES_STRUCT]] to i8*
-  // CHECK:   call void @llvm.memcpy.p0i8.p0i8.i32(i8* [[AGG_RESULT8]], i8* [[ATOMIC_RES8]], i32 6, i32 2, i1 false)
+  // CHECK:   call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 2 [[AGG_RESULT8]], i8* align 8 [[ATOMIC_RES8]], i32 6, i1 false)
 
   return __c11_atomic_load(addr, 5);
 }
@@ -232,11 +232,11 @@ void test_promoted_store(_Atomic(PS) *addr, PS *val) {
   // CHECK:   [[VAL:%.*]] = load %struct.PS*, %struct.PS** [[VAL_ARG]], align 4
   // CHECK:   [[NONATOMIC_TMP8:%.*]] = bitcast %struct.PS* [[NONATOMIC_TMP]] to i8*
   // CHECK:   [[VAL8:%.*]] = bitcast %struct.PS* [[VAL]] to i8*
-  // CHECK:   call void @llvm.memcpy.p0i8.p0i8.i32(i8* [[NONATOMIC_TMP8]], i8* [[VAL8]], i32 6, i32 2, i1 false)
+  // CHECK:   call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 2 [[NONATOMIC_TMP8]], i8* align 2 [[VAL8]], i32 6, i1 false)
   // CHECK:   [[ADDR64:%.*]] = bitcast { %struct.PS, [2 x i8] }* [[ADDR]] to i64*
   // CHECK:   [[ATOMIC_VAL8:%.*]] = bitcast { %struct.PS, [2 x i8] }* [[ATOMIC_VAL]] to i8*
   // CHECK:   [[NONATOMIC_TMP8:%.*]] = bitcast %struct.PS* [[NONATOMIC_TMP]] to i8*
-  // CHECK:   call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[ATOMIC_VAL8]], i8* [[NONATOMIC_TMP8]], i64 6, i32 2, i1 false)
+  // CHECK:   call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 8 [[ATOMIC_VAL8]], i8* align 2 [[NONATOMIC_TMP8]], i64 6, i1 false)
   // CHECK:   [[ATOMIC_VAL64:%.*]] = bitcast { %struct.PS, [2 x i8] }* [[ATOMIC_VAL]] to i64*
   // CHECK:   [[VAL64:%.*]] = load i64, i64* [[ATOMIC_VAL64]], align 8
   // CHECK:   store atomic i64 [[VAL64]], i64* [[ADDR64]] seq_cst, align 8
@@ -257,11 +257,11 @@ PS test_promoted_exchange(_Atomic(PS) *addr, PS *val) {
   // CHECK:   [[VAL:%.*]] = load %struct.PS*, %struct.PS** [[VAL_ARG]], align 4
   // CHECK:   [[NONATOMIC_TMP8:%.*]] = bitcast %struct.PS* [[NONATOMIC_TMP]] to i8*
   // CHECK:   [[VAL8:%.*]] = bitcast %struct.PS* [[VAL]] to i8*
-  // CHECK:   call void @llvm.memcpy.p0i8.p0i8.i32(i8* [[NONATOMIC_TMP8]], i8* [[VAL8]], i32 6, i32 2, i1 false)
+  // CHECK:   call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 2 [[NONATOMIC_TMP8]], i8* align 2 [[VAL8]], i32 6, i1 false)
   // CHECK:   [[ADDR64:%.*]] = bitcast { %struct.PS, [2 x i8] }* [[ADDR]] to i64*
   // CHECK:   [[ATOMIC_VAL8:%.*]] = bitcast { %struct.PS, [2 x i8] }* [[ATOMIC_VAL]] to i8*
   // CHECK:   [[NONATOMIC_TMP8:%.*]] = bitcast %struct.PS* [[NONATOMIC_TMP]] to i8*
-  // CHECK:   call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[ATOMIC_VAL8]], i8* [[NONATOMIC_TMP8]], i64 6, i32 2, i1 false)
+  // CHECK:   call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 8 [[ATOMIC_VAL8]], i8* align 2 [[NONATOMIC_TMP8]], i64 6, i1 false)
   // CHECK:   [[ATOMIC_VAL64:%.*]] = bitcast { %struct.PS, [2 x i8] }* [[ATOMIC_VAL]] to i64*
   // CHECK:   [[ATOMIC_RES64:%.*]] = bitcast { %struct.PS, [2 x i8] }* [[ATOMIC_RES]] to i64*
   // CHECK:   [[VAL64:%.*]] = load i64, i64* [[ATOMIC_VAL64]], align 8
@@ -270,7 +270,7 @@ PS test_promoted_exchange(_Atomic(PS) *addr, PS *val) {
   // CHECK:   [[ATOMIC_RES_STRUCT:%.*]] = bitcast i64* [[ATOMIC_RES64]] to %struct.PS*
   // CHECK:   [[AGG_RESULT8:%.*]] = bitcast %struct.PS* %agg.result to i8*
   // CHECK:   [[ATOMIC_RES8:%.*]] = bitcast %struct.PS* [[ATOMIC_RES_STRUCT]] to i8*
-  // CHECK:   call void @llvm.memcpy.p0i8.p0i8.i32(i8* [[AGG_RESULT8]], i8* [[ATOMIC_RES8]], i32 6, i32 2, i1 false)
+  // CHECK:   call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 2 [[AGG_RESULT8]], i8* align 8 [[ATOMIC_RES8]], i32 6, i1 false)
   return __c11_atomic_exchange(addr, *val, 5);
 }
 
@@ -291,15 +291,15 @@ _Bool test_promoted_cmpxchg(_Atomic(PS) *addr, PS *desired, PS *new) {
   // CHECK:   [[NEW:%.*]] = load %struct.PS*, %struct.PS** [[NEW_ARG]], align 4
   // CHECK:   [[NONATOMIC_TMP8:%.*]] = bitcast %struct.PS* [[NONATOMIC_TMP]] to i8*
   // CHECK:   [[NEW8:%.*]] = bitcast %struct.PS* [[NEW]] to i8*
-  // CHECK:   call void @llvm.memcpy.p0i8.p0i8.i32(i8* [[NONATOMIC_TMP8]], i8* [[NEW8]], i32 6, i32 2, i1 false)
+  // CHECK:   call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 2 [[NONATOMIC_TMP8]], i8* align 2 [[NEW8]], i32 6, i1 false)
   // CHECK:   [[ADDR64:%.*]] = bitcast { %struct.PS, [2 x i8] }* [[ADDR]] to i64*
   // CHECK:   [[ATOMIC_DESIRED8:%.*]] = bitcast { %struct.PS, [2 x i8] }* [[ATOMIC_DESIRED:%.*]] to i8*
   // CHECK:   [[DESIRED8:%.*]] = bitcast %struct.PS* [[DESIRED]] to i8*
-  // CHECK:   call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[ATOMIC_DESIRED8]], i8* [[DESIRED8]], i64 6, i32 2, i1 false)
+  // CHECK:   call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 8 [[ATOMIC_DESIRED8]], i8* align 2 [[DESIRED8]], i64 6, i1 false)
   // CHECK:   [[ATOMIC_DESIRED64:%.*]] = bitcast { %struct.PS, [2 x i8] }* [[ATOMIC_DESIRED:%.*]] to i64*
   // CHECK:   [[ATOMIC_NEW8:%.*]] = bitcast { %struct.PS, [2 x i8] }* [[ATOMIC_NEW]] to i8*
   // CHECK:   [[NONATOMIC_TMP8:%.*]] = bitcast %struct.PS* [[NONATOMIC_TMP]] to i8*
-  // CHECK:   call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[ATOMIC_NEW8]], i8* [[NONATOMIC_TMP8]], i64 6, i32 2, i1 false)
+  // CHECK:   call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 8 [[ATOMIC_NEW8]], i8* align 2 [[NONATOMIC_TMP8]], i64 6, i1 false)
   // CHECK:   [[ATOMIC_NEW64:%.*]] = bitcast { %struct.PS, [2 x i8] }* [[ATOMIC_NEW]] to i64*
   // CHECK:   [[ATOMIC_DESIRED_VAL64:%.*]] = load i64, i64* [[ATOMIC_DESIRED64]], align 8
   // CHECK:   [[ATOMIC_NEW_VAL64:%.*]] = load i64, i64* [[ATOMIC_NEW64]], align 8
diff --git a/clang/test/CodeGen/c11atomics.c b/clang/test/CodeGen/c11atomics.c
index ccb642174c9..3774ded7c42 100644
--- a/clang/test/CodeGen/c11atomics.c
+++ b/clang/test/CodeGen/c11atomics.c
@@ -282,7 +282,7 @@ void testStruct(_Atomic(S) *fp) {
 // CHECK-NEXT: [[T0:%.*]] = load [[S]]*, [[S]]** [[FP]]
 // CHECK-NEXT: [[T1:%.*]] = bitcast [[S]]* [[TMP0]] to i8*
 // CHECK-NEXT: [[T2:%.*]] = bitcast [[S]]* [[F]] to i8*
-// CHECK-NEXT: call void @llvm.memcpy.p0i8.p0i8.i32(i8* [[T1]], i8* [[T2]], i32 8, i32 2, i1 false)
+// CHECK-NEXT: call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 8 [[T1]], i8* align 2 [[T2]], i32 8, i1 false)
 // CHECK-NEXT: [[T3:%.*]] = bitcast [[S]]* [[T0]] to i8*
 // CHECK-NEXT: [[T4:%.*]] = bitcast [[S]]* [[TMP0]] to i8*
 // CHECK-NEXT: call arm_aapcscc void @__atomic_store(i32 8, i8* [[T3]], i8* [[T4]], i32 5)
@@ -307,7 +307,7 @@ void testPromotedStruct(_Atomic(PS) *fp) {
 
 // CHECK-NEXT: [[P:%.*]] = load [[APS]]*, [[APS]]** [[FP]]
 // CHECK-NEXT: [[T0:%.*]] = bitcast [[APS]]* [[P]] to i8*
-// CHECK-NEXT: call void @llvm.memset.p0i8.i64(i8* [[T0]], i8 0, i64 8, i32 8, i1 false)
+// CHECK-NEXT: call void @llvm.memset.p0i8.i64(i8* align 8 [[T0]], i8 0, i64 8, i1 false)
 // CHECK-NEXT: [[T0:%.*]] = getelementptr inbounds [[APS]], [[APS]]* [[P]], i32 0, i32 0
 // CHECK-NEXT: [[T1:%.*]] = getelementptr inbounds [[PS]], [[PS]]* [[T0]], i32 0, i32 0
 // CHECK-NEXT: store i16 1, i16* [[T1]], align 8
@@ -318,7 +318,7 @@ void testPromotedStruct(_Atomic(PS) *fp) {
   __c11_atomic_init(fp, (PS){1,2,3});
 
 // CHECK-NEXT: [[T0:%.*]] = bitcast [[APS]]* [[X]] to i8*
-// CHECK-NEXT: call void @llvm.memset.p0i8.i32(i8* [[T0]], i8 0, i32 8, i32 8, i1 false)
+// CHECK-NEXT: call void @llvm.memset.p0i8.i32(i8* align 8 [[T0]], i8 0, i32 8, i1 false)
 // CHECK-NEXT: [[T0:%.*]] = getelementptr inbounds [[APS]], [[APS]]* [[X]], i32 0, i32 0
 // CHECK-NEXT: [[T1:%.*]] = getelementptr inbounds [[PS]], [[PS]]* [[T0]], i32 0, i32 0
 // CHECK-NEXT: store i16 1, i16* [[T1]], align 8
@@ -335,16 +335,16 @@ void testPromotedStruct(_Atomic(PS) *fp) {
 // CHECK-NEXT: [[T0:%.*]] = getelementptr inbounds [[APS]], [[APS]]* [[TMP0]], i32 0, i32 0
 // CHECK-NEXT: [[T1:%.*]] = bitcast [[PS]]* [[F]] to i8*
 // CHECK-NEXT: [[T2:%.*]] = bitcast [[PS]]* [[T0]] to i8*
-// CHECK-NEXT: call void @llvm.memcpy.p0i8.p0i8.i32(i8* [[T1]], i8* [[T2]], i32 6, i32 2, i1 false)
+// CHECK-NEXT: call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 2 [[T1]], i8* align 8 [[T2]], i32 6, i1 false)
   PS f = *fp;
 
 // CHECK-NEXT: [[T0:%.*]] = load [[APS]]*, [[APS]]** [[FP]]
 // CHECK-NEXT: [[T1:%.*]] = bitcast { %struct.PS, [2 x i8] }* [[TMP1]] to i8*
-// CHECK-NEXT: call void @llvm.memset.p0i8.i32(i8* [[T1]], i8 0, i32 8, i32 8, i1 false)
+// CHECK-NEXT: call void @llvm.memset.p0i8.i32(i8* align 8 [[T1]], i8 0, i32 8, i1 false)
 // CHECK-NEXT: [[T1:%.*]] = getelementptr inbounds [[APS]], [[APS]]* [[TMP1]], i32 0, i32 0
 // CHECK-NEXT: [[T2:%.*]] = bitcast [[PS]]* [[T1]] to i8*
 // CHECK-NEXT: [[T3:%.*]] = bitcast [[PS]]* [[F]] to i8*
-// CHECK-NEXT: call void @llvm.memcpy.p0i8.p0i8.i32(i8* [[T2]], i8* [[T3]], i32 6, i32 2, i1 false)
+// CHECK-NEXT: call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 8 [[T2]], i8* align 2 [[T3]], i32 6, i1 false)
 // CHECK-NEXT: [[T4:%.*]] = bitcast [[APS]]* [[T0]] to i8*
 // CHECK-NEXT: [[T5:%.*]] = bitcast [[APS]]* [[TMP1]] to i8*
 // CHECK-NEXT: call arm_aapcscc void @__atomic_store(i32 8, i8* [[T4]], i8* [[T5]], i32 5)
@@ -357,7 +357,7 @@ void testPromotedStruct(_Atomic(PS) *fp) {
 // CHECK-NEXT: [[T0:%.*]] = getelementptr inbounds [[APS]], [[APS]]* [[TMP3]], i32 0, i32 0
 // CHECK-NEXT: [[T1:%.*]] = bitcast %struct.PS* [[TMP2]] to i8*
 // CHECK-NEXT: [[T2:%.*]] = bitcast %struct.PS* [[T0]] to i8*
-// CHECK-NEXT: call void @llvm.memcpy.p0i8.p0i8.i32(i8* [[T1]], i8* [[T2]], i32 6, i32 2, i1 false)
+// CHECK-NEXT: call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 2 [[T1]], i8* align 8 [[T2]], i32 6, i1 false)
 // CHECK-NEXT: [[T0:%.*]] = getelementptr inbounds %struct.PS, %struct.PS* [[TMP2]], i32 0, i32 0
 // CHECK-NEXT: [[T1:%.*]] = load i16, i16* [[T0]], align 2
 // CHECK-NEXT: [[T2:%.*]] = sext i16 [[T1]] to i32
@@ -381,7 +381,7 @@ PS test_promoted_load(_Atomic(PS) *addr) {
   // CHECK:   [[ATOMIC_RES_STRUCT:%.*]] = bitcast i64* [[ATOMIC_RES64]] to %struct.PS*
   // CHECK:   [[AGG_RESULT8:%.*]] = bitcast %struct.PS* %agg.result to i8*
   // CHECK:   [[ATOMIC_RES8:%.*]] = bitcast %struct.PS* [[ATOMIC_RES_STRUCT]] to i8*
-  // CHECK:   call void @llvm.memcpy.p0i8.p0i8.i32(i8* [[AGG_RESULT8]], i8* [[ATOMIC_RES8]], i32 6, i32 2, i1 false)
+  // CHECK:   call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 2 [[AGG_RESULT8]], i8* align 8 [[ATOMIC_RES8]], i32 6, i1 false)
 
   return __c11_atomic_load(addr, 5);
 }
@@ -398,11 +398,11 @@ void test_promoted_store(_Atomic(PS) *addr, PS *val) {
   // CHECK:   [[VAL:%.*]] = load %struct.PS*, %struct.PS** [[VAL_ARG]], align 4
   // CHECK:   [[NONATOMIC_TMP8:%.*]] = bitcast %struct.PS* [[NONATOMIC_TMP]] to i8*
   // CHECK:   [[VAL8:%.*]] = bitcast %struct.PS* [[VAL]] to i8*
-  // CHECK:   call void @llvm.memcpy.p0i8.p0i8.i32(i8* [[NONATOMIC_TMP8]], i8* [[VAL8]], i32 6, i32 2, i1 false)
+  // CHECK:   call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 2 [[NONATOMIC_TMP8]], i8* align 2 [[VAL8]], i32 6, i1 false)
   // CHECK:   [[ADDR64:%.*]] = bitcast { %struct.PS, [2 x i8] }* [[ADDR]] to i64*
   // CHECK:   [[ATOMIC_VAL8:%.*]] = bitcast { %struct.PS, [2 x i8] }* [[ATOMIC_VAL]] to i8*
   // CHECK:   [[NONATOMIC_TMP8:%.*]] = bitcast %struct.PS* [[NONATOMIC_TMP]] to i8*
-  // CHECK:   call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[ATOMIC_VAL8]], i8* [[NONATOMIC_TMP8]], i64 6, i32 2, i1 false)
+  // CHECK:   call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 8 [[ATOMIC_VAL8]], i8* align 2 [[NONATOMIC_TMP8]], i64 6, i1 false)
   // CHECK:   [[ATOMIC_VAL64:%.*]] = bitcast { %struct.PS, [2 x i8] }* [[ATOMIC_VAL]] to i64*
   // CHECK:   [[ADDR8:%.*]] = bitcast i64* [[ADDR64]] to i8*
   // CHECK:   [[VAL64:%.*]] = load i64, i64* [[ATOMIC_VAL64]], align 2
@@ -423,11 +423,11 @@ PS test_promoted_exchange(_Atomic(PS) *addr, PS *val) {
   // CHECK:   [[VAL:%.*]] = load %struct.PS*, %struct.PS** [[VAL_ARG]], align 4
   // CHECK:   [[NONATOMIC_TMP8:%.*]] = bitcast %struct.PS* [[NONATOMIC_TMP]] to i8*
   // CHECK:   [[VAL8:%.*]] = bitcast %struct.PS* [[VAL]] to i8*
-  // CHECK:   call void @llvm.memcpy.p0i8.p0i8.i32(i8* [[NONATOMIC_TMP8]], i8* [[VAL8]], i32 6, i32 2, i1 false)
+  // CHECK:   call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 2 [[NONATOMIC_TMP8]], i8* align 2 [[VAL8]], i32 6, i1 false)
   // CHECK:   [[ADDR64:%.*]] = bitcast { %struct.PS, [2 x i8] }* [[ADDR]] to i64*
   // CHECK:   [[ATOMIC_VAL8:%.*]] = bitcast { %struct.PS, [2 x i8] }* [[ATOMIC_VAL]] to i8*
   // CHECK:   [[NONATOMIC_TMP8:%.*]] = bitcast %struct.PS* [[NONATOMIC_TMP]] to i8*
-  // CHECK:   call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[ATOMIC_VAL8]], i8* [[NONATOMIC_TMP8]], i64 6, i32 2, i1 false)
+  // CHECK:   call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 8 [[ATOMIC_VAL8]], i8* align 2 [[NONATOMIC_TMP8]], i64 6, i1 false)
   // CHECK:   [[ATOMIC_VAL64:%.*]] = bitcast { %struct.PS, [2 x i8] }* [[ATOMIC_VAL]] to i64*
   // CHECK:   [[ATOMIC_RES64:%.*]] = bitcast { %struct.PS, [2 x i8] }* [[ATOMIC_RES]] to i64*
   // CHECK:   [[ADDR8:%.*]] = bitcast i64* [[ADDR64]] to i8*
@@ -437,7 +437,7 @@ PS test_promoted_exchange(_Atomic(PS) *addr, PS *val) {
   // CHECK:   [[ATOMIC_RES_STRUCT:%.*]] = bitcast i64* [[ATOMIC_RES64]] to %struct.PS*
   // CHECK:   [[AGG_RESULT8:%.*]] = bitcast %struct.PS* %agg.result to i8*
   // CHECK:   [[ATOMIC_RES8:%.*]] = bitcast %struct.PS* [[ATOMIC_RES_STRUCT]] to i8*
-  // CHECK:   call void @llvm.memcpy.p0i8.p0i8.i32(i8* [[AGG_RESULT8]], i8* [[ATOMIC_RES8]], i32 6, i32 2, i1 false)
+  // CHECK:   call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 2 [[AGG_RESULT8]], i8* align 8 [[ATOMIC_RES8]], i32 6, i1 false)
   return __c11_atomic_exchange(addr, *val, 5);
 }
 
@@ -457,15 +457,15 @@ _Bool test_promoted_cmpxchg(_Atomic(PS) *addr, PS *desired, PS *new) {
   // CHECK:   [[NEW:%.*]] = load %struct.PS*, %struct.PS** [[NEW_ARG]], align 4
   // CHECK:   [[NONATOMIC_TMP8:%.*]] = bitcast %struct.PS* [[NONATOMIC_TMP]] to i8*
   // CHECK:   [[NEW8:%.*]] = bitcast %struct.PS* [[NEW]] to i8*
-  // CHECK:   call void @llvm.memcpy.p0i8.p0i8.i32(i8* [[NONATOMIC_TMP8]], i8* [[NEW8]], i32 6, i32 2, i1 false)
+  // CHECK:   call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 2 [[NONATOMIC_TMP8]], i8* align 2 [[NEW8]], i32 6, i1 false)
   // CHECK:   [[ADDR64:%.*]] = bitcast { %struct.PS, [2 x i8] }* [[ADDR]] to i64*
   // CHECK:   [[ATOMIC_DESIRED8:%.*]] = bitcast { %struct.PS, [2 x i8] }* [[ATOMIC_DESIRED]] to i8*
   // CHECK:   [[DESIRED8:%.*]] = bitcast %struct.PS* [[DESIRED]]to i8*
-  // CHECK:   call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[ATOMIC_DESIRED8]], i8* [[DESIRED8]], i64 6, i32 2, i1 false)
+  // CHECK:   call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 8 [[ATOMIC_DESIRED8]], i8* align 2 [[DESIRED8]], i64 6, i1 false)
   // CHECK:   [[ATOMIC_DESIRED64:%.*]] = bitcast { %struct.PS, [2 x i8] }* [[ATOMIC_DESIRED]] to i64*
   // CHECK:   [[ATOMIC_NEW8:%.*]] = bitcast { %struct.PS, [2 x i8] }* [[ATOMIC_NEW]] to i8*
   // CHECK:   [[NONATOMIC_TMP8:%.*]] = bitcast %struct.PS* [[NONATOMIC_TMP]] to i8*
-  // CHECK:   call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[ATOMIC_NEW8]], i8* [[NONATOMIC_TMP8]], i64 6, i32 2, i1 false)
+  // CHECK:   call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 8 [[ATOMIC_NEW8]], i8* align 2 [[NONATOMIC_TMP8]], i64 6, i1 false)
   // CHECK:   [[ATOMIC_NEW64:%.*]] = bitcast { %struct.PS, [2 x i8] }* [[ATOMIC_NEW]] to i64*
   // CHECK:   [[ADDR8:%.*]] = bitcast i64* [[ADDR64]] to i8*
   // CHECK:   [[ATOMIC_DESIRED8:%.*]] = bitcast i64* [[ATOMIC_DESIRED64]] to i8*
diff --git a/clang/test/CodeGen/compound-literal.c b/clang/test/CodeGen/compound-literal.c
index 85138bfaf5e..a81a80027a1 100644
--- a/clang/test/CodeGen/compound-literal.c
+++ b/clang/test/CodeGen/compound-literal.c
@@ -28,7 +28,7 @@ void f() {
   // CHECK-NEXT: store i32 [[TMP]], i32* [[CY]]
   // CHECK-NEXT: [[SI8:%[a-zA-Z0-9.]+]] = bitcast [[STRUCT]]* [[S]] to i8*
   // CHECK-NEXT: [[COMPOUNDLITI8:%[a-zA-Z0-9.]+]] = bitcast [[STRUCT]]* [[COMPOUNDLIT]] to i8*
-  // CHECK-NEXT: call void @llvm.memcpy{{.*}}(i8* [[SI8]], i8* [[COMPOUNDLITI8]]
+  // CHECK-NEXT: call void @llvm.memcpy{{.*}}(i8* align 4 [[SI8]], i8* align 4 [[COMPOUNDLITI8]]
   s = (S){s.y,s.x};
   // CHECK-NEXT: ret void
 }
@@ -62,7 +62,7 @@ struct G g(int x, int y, int z) {
 
   // CHECK-NEXT: [[T0:%.*]] = bitcast i48* [[COERCE_TEMP]] to i8*
   // CHECK-NEXT: [[T1:%.*]] = bitcast [[G]]* [[RESULT]] to i8*
-  // CHECK-NEXT: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[T0]], i8* [[T1]], i64 6
+  // CHECK-NEXT: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 8 [[T0]], i8* align 2 [[T1]], i64 6
   // CHECK-NEXT: [[T0:%.*]] = load i48, i48* [[COERCE_TEMP]]
   // CHECK-NEXT: ret i48 [[T0]]
 }
diff --git a/clang/test/CodeGen/no-opt-volatile-memcpy.c b/clang/test/CodeGen/no-opt-volatile-memcpy.c
index bf98df39af1..2781475b630 100644
--- a/clang/test/CodeGen/no-opt-volatile-memcpy.c
+++ b/clang/test/CodeGen/no-opt-volatile-memcpy.c
@@ -18,10 +18,10 @@ void foo (void) {
 // CHECK: %[[LS:.*]] = alloca %struct.s, align 4
 // CHECK-NEXT: %[[ZERO:.*]] = bitcast %struct.s* %[[LS]] to i8*
 // CHECK-NEXT: %[[ONE:.*]] = bitcast %struct.s* %[[LS]] to i8*
-// CHECK-NEXT: call void @llvm.memcpy.{{.*}}(i8* %[[ZERO]], i8* %[[ONE]], i64 132, i32 4, i1 true)
-// CHECK-NEXT: call void @llvm.memcpy.{{.*}}(i8* getelementptr inbounds (%struct.s, %struct.s* @gs, i32 0, i32 0, i32 0), i8* getelementptr inbounds (%struct.s, %struct.s* @gs, i32 0, i32 0, i32 0), i64 132, i32 4, i1 true)
+// CHECK-NEXT: call void @llvm.memcpy.{{.*}}(i8* align 4 %[[ZERO]], i8* align 4 %[[ONE]], i64 132, i1 true)
+// CHECK-NEXT: call void @llvm.memcpy.{{.*}}(i8* align 4 getelementptr inbounds (%struct.s, %struct.s* @gs, i32 0, i32 0, i32 0), i8* align 4 getelementptr inbounds (%struct.s, %struct.s* @gs, i32 0, i32 0, i32 0), i64 132, i1 true)
 // CHECK-NEXT: %[[TWO:.*]] = bitcast %struct.s* %[[LS]] to i8*
-// CHECK-NEXT: call void @llvm.memcpy.{{.*}}(i8* %[[TWO]], i8* getelementptr inbounds (%struct.s, %struct.s* @gs, i32 0, i32 0, i32 0), i64 132, i32 4, i1 true)
+// CHECK-NEXT: call void @llvm.memcpy.{{.*}}(i8* align 4 %[[TWO]], i8* align 4 getelementptr inbounds (%struct.s, %struct.s* @gs, i32 0, i32 0, i32 0), i64 132, i1 true)
 
 
 struct s1 {
@@ -35,6 +35,6 @@ void fee (void) {
   s.y = gs;
 }
 // CHECK-LABEL: define void @fee()
-// CHECK: call void @llvm.memcpy.{{.*}}(i8* getelementptr inbounds (%struct.s1, %struct.s1* @s, i32 0, i32 0, i32 0, i32 0), i8* getelementptr inbounds (%struct.s1, %struct.s1* @s, i32 0, i32 0, i32 0, i32 0), i64 132, i32 4, i1 true)
-// CHECK-NEXT: call void @llvm.memcpy.{{.*}}(i8* getelementptr inbounds (%struct.s1, %struct.s1* @s, i32 0, i32 0, i32 0, i32 0), i8* getelementptr inbounds (%struct.s, %struct.s* @gs, i32 0, i32 0, i32 0), i64 132, i32 4, i1 true)
+// CHECK: call void @llvm.memcpy.{{.*}}(i8* align 4 getelementptr inbounds (%struct.s1, %struct.s1* @s, i32 0, i32 0, i32 0, i32 0), i8* align 4 getelementptr inbounds (%struct.s1, %struct.s1* @s, i32 0, i32 0, i32 0, i32 0), i64 132, i1 true)
+// CHECK-NEXT: call void @llvm.memcpy.{{.*}}(i8* align 4 getelementptr inbounds (%struct.s1, %struct.s1* @s, i32 0, i32 0, i32 0, i32 0), i8* align 4 getelementptr inbounds (%struct.s, %struct.s* @gs, i32 0, i32 0, i32 0), i64 132, i1 true)
 
diff --git a/clang/test/CodeGen/packed-nest-unpacked.c b/clang/test/CodeGen/packed-nest-unpacked.c
index e2bbd41a9da..5e5aba5cc5a 100644
--- a/clang/test/CodeGen/packed-nest-unpacked.c
+++ b/clang/test/CodeGen/packed-nest-unpacked.c
@@ -9,26 +9,26 @@ struct X foo(void);
 // <rdar://problem/10463337>
 struct X test1() {
   // CHECK: @test1
-  // CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* {{.*}}, i8* bitcast (%struct.X* getelementptr inbounds (%struct.Y, %struct.Y* @g, i32 0, i32 1) to i8*), i64 24, i32 1, i1 false)
+  // CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 4 {{.*}}, i8* align 1 bitcast (%struct.X* getelementptr inbounds (%struct.Y, %struct.Y* @g, i32 0, i32 1) to i8*), i64 24, i1 false)
   return g.y;
 }
 struct X test2() {
   // CHECK: @test2
-  // CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* {{.*}}, i8* bitcast (%struct.X* getelementptr inbounds (%struct.Y, %struct.Y* @g, i32 0, i32 1) to i8*), i64 24, i32 1, i1 false)
+  // CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 4 {{.*}}, i8* align 1 bitcast (%struct.X* getelementptr inbounds (%struct.Y, %struct.Y* @g, i32 0, i32 1) to i8*), i64 24, i1 false)
   struct X a = g.y;
   return a;
 }
 
 void test3(struct X a) {
   // CHECK: @test3
-  // CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* bitcast (%struct.X* getelementptr inbounds (%struct.Y, %struct.Y* @g, i32 0, i32 1) to i8*), i8* {{.*}}, i64 24, i32 1, i1 false)
+  // CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 1 bitcast (%struct.X* getelementptr inbounds (%struct.Y, %struct.Y* @g, i32 0, i32 1) to i8*), i8* align 8 {{.*}}, i64 24, i1 false)
   g.y = a;
 }
 
 // <rdar://problem/10530444>
 void test4() {
   // CHECK: @test4
-  // CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* {{.*}}, i8* bitcast (%struct.X* getelementptr inbounds (%struct.Y, %struct.Y* @g, i32 0, i32 1) to i8*), i64 24, i32 1, i1 false)
+  // CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 4 {{.*}}, i8* align 1 bitcast (%struct.X* getelementptr inbounds (%struct.Y, %struct.Y* @g, i32 0, i32 1) to i8*), i64 24, i1 false)
   f(g.y);
 }
 
@@ -42,7 +42,7 @@ int test5() {
 // <rdar://problem/11220251>
 void test6() {
   // CHECK: @test6
-  // CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* bitcast (%struct.X* getelementptr inbounds (%struct.Y, %struct.Y* @g, i32 0, i32 1) to i8*), i8* %{{.*}}, i64 24, i32 1, i1 false)
+  // CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 1 bitcast (%struct.X* getelementptr inbounds (%struct.Y, %struct.Y* @g, i32 0, i32 1) to i8*), i8* align 4 %{{.*}}, i64 24, i1 false)
   g.y = foo();
 }
 
diff --git a/clang/test/CodeGen/packed-structure.c b/clang/test/CodeGen/packed-structure.c
index 7d1183dc5ca..2885fdf107f 100644
--- a/clang/test/CodeGen/packed-structure.c
+++ b/clang/test/CodeGen/packed-structure.c
@@ -29,7 +29,7 @@ int s0_load_x(struct s0 *a) { return a->x; }
 // CHECK-FUNCTIONS: ret i32 [[s0_load_y]]
 int s0_load_y(struct s0 *a) { return a->y; }
 // CHECK-FUNCTIONS-LABEL: define void @s0_copy
-// CHECK-FUNCTIONS: call void @llvm.memcpy.p0i8.p0i8.i64(i8* {{.*}}, i8* {{.*}}, i64 8, i32 4, i1 false)
+// CHECK-FUNCTIONS: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 4 {{.*}}, i8* align 4 {{.*}}, i64 8, i1 false)
 void s0_copy(struct s0 *a, struct s0 *b) { *b = *a; }
 
 //
@@ -55,7 +55,7 @@ int s1_load_x(struct s1 *a) { return a->x; }
 // CHECK-FUNCTIONS: ret i32 [[s1_load_y]]
 int s1_load_y(struct s1 *a) { return a->y; }
 // CHECK-FUNCTIONS-LABEL: define void @s1_copy
-// CHECK-FUNCTIONS: call void @llvm.memcpy.p0i8.p0i8.i64(i8* {{.*}}, i8* {{.*}}, i64 8, i32 1, i1 false)
+// CHECK-FUNCTIONS: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 1 {{.*}}, i8* align 1 {{.*}}, i64 8, i1 false)
 void s1_copy(struct s1 *a, struct s1 *b) { *b = *a; }
 
 //
@@ -83,7 +83,7 @@ int s2_load_x(struct s2 *a) { return a->x; }
 // CHECK-FUNCTIONS: ret i32 [[s2_load_y]]
 int s2_load_y(struct s2 *a) { return a->y; }
 // CHECK-FUNCTIONS-LABEL: define void @s2_copy
-// CHECK-FUNCTIONS: call void @llvm.memcpy.p0i8.p0i8.i64(i8* {{.*}}, i8* {{.*}}, i64 8, i32 2, i1 false)
+// CHECK-FUNCTIONS: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 2 {{.*}}, i8* align 2 {{.*}}, i64 8, i1 false)
 void s2_copy(struct s2 *a, struct s2 *b) { *b = *a; }
 
 struct __attribute__((packed, aligned)) s3 {
diff --git a/clang/test/CodeGen/partial-reinitialization2.c b/clang/test/CodeGen/partial-reinitialization2.c
index c4f6567b36c..34da674cedf 100644
--- a/clang/test/CodeGen/partial-reinitialization2.c
+++ b/clang/test/CodeGen/partial-reinitialization2.c
@@ -15,7 +15,7 @@ union ULP3 { struct LP3 l3; };
 // CHECK-LABEL: test1
 void test1(void)
 {
-  // CHECK: call void @llvm.memcpy{{.*}}%struct.P1, %struct.P1* @g1{{.*}}i64 6, i32 {{[0-9]}}, i1 false)
+  // CHECK: call void @llvm.memcpy{{.*}}(i8* align 1 %{{.*}}, i8* align 1 {{.*}} (%struct.P1, %struct.P1* @g1{{.*}}i64 6, i1 false)
   // CHECK: store i8 120, i8* %
 
   struct LP1 l = { .p1 = g1, .p1.x[2] = 'x' };
@@ -24,7 +24,7 @@ void test1(void)
 // CHECK-LABEL: test2
 void test2(void)
 {
-  // CHECK: call void @llvm.memcpy{{.*}}%struct.P1, %struct.P1* @g1{{.*}}i64 6, i32 {{[0-9]}}, i1 false)
+  // CHECK: call void @llvm.memcpy{{.*}}(i8* align 1 %{{.*}}, i8* align 1 {{.*}} (%struct.P1, %struct.P1* @g1{{.*}}i64 6, i1 false)
   // CHECK: store i8 114, i8* %
 
   struct LP1 l = { .p1 = g1, .p1.x[1] = 'r' };
@@ -33,7 +33,7 @@ void test2(void)
 // CHECK-LABEL: test3
 void test3(void)
 {
-  // CHECK: call void @llvm.memcpy{{.*}}%struct.P2* @g2{{.*}}i64 12, i32 {{[0-9]}}, i1 false)
+  // CHECK: call void @llvm.memcpy{{.*}}(i8* align 4 %{{.*}}, i8* align 4 {{.*}} (%struct.P2* @g2{{.*}}i64 12, i1 false)
   // CHECK: store i32 10, i32* %
 
   struct LP2 l = { .p2 = g2, .p2.b = 10 };
@@ -66,7 +66,7 @@ void test4(void)
   // CHECK: [[CALL:%[a-z0-9]+]] = call {{.*}}@get123()
   // CHECK: store{{.*}}[[CALL]], {{.*}}[[TMP0:%[a-z0-9]+]]
   // CHECK: [[TMP1:%[a-z0-9]+]] = bitcast {{.*}}[[TMP0]]
-  // CHECK: call void @llvm.memcpy{{.*}}[[TMP1]], i64 12, i32 {{[0-9]}}, i1 false)
+  // CHECK: call void @llvm.memcpy{{.*}}(i8* align 4 %{{.*}}, i8* align 8 [[TMP1]], i64 12, i1 false)
   // CHECK: store i32 100, i32* %
 
   struct LUP2 { union UP2 up; } var = { get123(), .up.p2.a = 100 };
@@ -76,12 +76,12 @@ void test4(void)
 void test5(void)
 {
   // .l3 = g3
-  // CHECK: call void @llvm.memcpy{{.*}}%struct.LP3, %struct.LP3* @g3{{.*}}i64 12, i32 {{[0-9]}}, i1 false)
+  // CHECK: call void @llvm.memcpy{{.*}}(i8* align 1 %{{.*}}, i8* align 1 {{.*}} (%struct.LP3, %struct.LP3* @g3{{.*}}i64 12, i1 false)
 
   // .l3.p1 = { [0] = g1 } implicitly sets [1] to zero
-  // CHECK: call void @llvm.memcpy{{.*}}%struct.P1, %struct.P1* @g1{{.*}}i64 6, i32 {{[0-9]}}, i1 false)
+  // CHECK: call void @llvm.memcpy{{.*}}(i8* align 1 %{{.*}}, i8* align 1 {{.*}} (%struct.P1, %struct.P1* @g1{{.*}}i64 6, i1 false)
   // CHECK: getelementptr{{.*}}%struct.P1, %struct.P1*{{.*}}i64 1
-  // CHECK: call void @llvm.memset{{.*}}i8 0, i64 6, i32 {{[0-9]}}, i1 false)
+  // CHECK: call void @llvm.memset{{.*}}(i8* align 1 %{{.*}}, i8 0, i64 6, i1 false)
 
   // .l3.p1[1].x[1] = 'x'
   // CHECK: store i8 120, i8* %
@@ -98,7 +98,7 @@ void test6(void)
   // CHECK: [[CALL:%[a-z0-9]+]] = call {{.*}}@get235()
   // CHECK: store{{.*}}[[CALL]], {{.*}}[[TMP0:%[a-z0-9]+]]
   // CHECK: [[TMP1:%[a-z0-9]+]] = bitcast {{.*}}[[TMP0]]
-  // CHECK: call void @llvm.memcpy{{.*}}[[TMP1]], i64 12, i32 {{[0-9]}}, i1 false)
+  // CHECK: call void @llvm.memcpy{{.*}}(i8* align 4 %{{.*}}, i8* align 8 [[TMP1]], i64 12, i1 false)
 
   // CHECK: store i32 10, i32* %
 
diff --git a/clang/test/CodeGen/ppc-varargs-struct.c b/clang/test/CodeGen/ppc-varargs-struct.c
index a6ba13799db..90d30388ebc 100644
--- a/clang/test/CodeGen/ppc-varargs-struct.c
+++ b/clang/test/CodeGen/ppc-varargs-struct.c
@@ -49,7 +49,7 @@ void testva (int n, ...)
 // CHECK-PPC-NEXT:  [[AGGR:%[a-z0-9]+]] = load %struct.x*, %struct.x** [[VAARG_ADDR]]
 // CHECK-PPC-NEXT:  [[DEST:%[0-9]+]] = bitcast %struct.x* %t to i8*
 // CHECK-PPC-NEXT:  [[SRC:%.+]] = bitcast %struct.x* [[AGGR]] to i8*
-// CHECK-PPC-NEXT:  call void @llvm.memcpy.p0i8.p0i8.i32(i8* [[DEST]], i8* [[SRC]], i32 16, i32 8, i1 false)
+// CHECK-PPC-NEXT:  call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 8 [[DEST]], i8* align 8 [[SRC]], i32 16, i1 false)
 
   int v = va_arg (ap, int);
   
diff --git a/clang/test/CodeGen/ppc64-align-struct.c b/clang/test/CodeGen/ppc64-align-struct.c
index 6a04d0cd84f..6fba2d262a4 100644
--- a/clang/test/CodeGen/ppc64-align-struct.c
+++ b/clang/test/CodeGen/ppc64-align-struct.c
@@ -56,7 +56,7 @@ void test7 (int x, struct test7 y)
 // CHECK: [[T0:%.*]] = bitcast i8* %[[CUR]] to %struct.test1*
 // CHECK: [[DEST:%.*]] = bitcast %struct.test1* %y to i8*
 // CHECK: [[SRC:%.*]] = bitcast %struct.test1* [[T0]] to i8*
-// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[DEST]], i8* [[SRC]], i64 8, i32 4, i1 false)
+// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 4 [[DEST]], i8* align 8 [[SRC]], i64 8, i1 false)
 struct test1 test1va (int x, ...)
 {
   struct test1 y;
@@ -79,7 +79,7 @@ struct test1 test1va (int x, ...)
 // CHECK: [[T0:%.*]] = bitcast i8* %[[ALIGN]] to %struct.test2*
 // CHECK: [[DEST:%.*]] = bitcast %struct.test2* %y to i8*
 // CHECK: [[SRC:%.*]] = bitcast %struct.test2* [[T0]] to i8*
-// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[DEST]], i8* [[SRC]], i64 16, i32 16, i1 false)
+// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 16 [[DEST]], i8* align 16 [[SRC]], i64 16, i1 false)
 struct test2 test2va (int x, ...)
 {
   struct test2 y;
@@ -102,7 +102,7 @@ struct test2 test2va (int x, ...)
 // CHECK: [[T0:%.*]] = bitcast i8* %[[ALIGN]] to %struct.test3*
 // CHECK: [[DEST:%.*]] = bitcast %struct.test3* %y to i8*
 // CHECK: [[SRC:%.*]] = bitcast %struct.test3* [[T0]] to i8*
-// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[DEST]], i8* [[SRC]], i64 32, i32 16, i1 false)
+// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 32 [[DEST]], i8* align 16 [[SRC]], i64 32, i1 false)
 struct test3 test3va (int x, ...)
 {
   struct test3 y;
@@ -121,7 +121,7 @@ struct test3 test3va (int x, ...)
 // CHECK: [[T0:%.*]] = bitcast i8* %[[CUR]] to %struct.test4*
 // CHECK: [[DEST:%.*]] = bitcast %struct.test4* %y to i8*
 // CHECK: [[SRC:%.*]] = bitcast %struct.test4* [[T0]] to i8*
-// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[DEST]], i8* [[SRC]], i64 12, i32 4, i1 false)
+// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 4 [[DEST]], i8* align 8 [[SRC]], i64 12, i1 false)
 struct test4 test4va (int x, ...)
 {
   struct test4 y;
@@ -140,7 +140,7 @@ struct test4 test4va (int x, ...)
 // CHECK: [[T0:%.*]] = bitcast i8* %[[CUR]] to %struct.test_longdouble*
 // CHECK: [[DEST:%.*]] = bitcast %struct.test_longdouble* %y to i8*
 // CHECK: [[SRC:%.*]] = bitcast %struct.test_longdouble* [[T0]] to i8*
-// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[DEST]], i8* [[SRC]], i64 16, i32 8, i1 false)
+// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 16 [[DEST]], i8* align 8 [[SRC]], i64 16, i1 false)
 struct test_longdouble { long double x; };
 struct test_longdouble testva_longdouble (int x, ...)
 {
@@ -164,7 +164,7 @@ struct test_longdouble testva_longdouble (int x, ...)
 // CHECK: [[T0:%.*]] = bitcast i8* %[[ALIGN]] to %struct.test_vector*
 // CHECK: [[DEST:%.*]] = bitcast %struct.test_vector* %y to i8*
 // CHECK: [[SRC:%.*]] = bitcast %struct.test_vector* [[T0]] to i8*
-// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[DEST]], i8* [[SRC]], i64 16, i32 16, i1 false)
+// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 16 [[DEST]], i8* align 16 [[SRC]], i64 16, i1 false)
 struct test_vector { vector int x; };
 struct test_vector testva_vector (int x, ...)
 {
diff --git a/clang/test/CodeGen/ppc64le-aggregates.c b/clang/test/CodeGen/ppc64le-aggregates.c
index 3ad4b06c688..49e1ec51099 100644
--- a/clang/test/CodeGen/ppc64le-aggregates.c
+++ b/clang/test/CodeGen/ppc64le-aggregates.c
@@ -104,7 +104,7 @@ void call_f8(void) { global_f8 = func_f8(global_f8); }
 // CHECK-LABEL: @call_f9
 // CHECK: %[[TMP1:[^ ]+]] = alloca [5 x i64]
 // CHECK: %[[TMP2:[^ ]+]] = bitcast [5 x i64]* %[[TMP1]] to i8*
-// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* %[[TMP2]], i8* bitcast (%struct.f9* @global_f9 to i8*), i64 36, i32 4, i1 false)
+// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 8 %[[TMP2]], i8* align 4 bitcast (%struct.f9* @global_f9 to i8*), i64 36, i1 false)
 // CHECK: %[[TMP3:[^ ]+]] = load [5 x i64], [5 x i64]* %[[TMP1]]
 // CHECK: call void @func_f9(%struct.f9* sret %{{[^ ]+}}, [5 x i64] %[[TMP3]])
 struct f9 global_f9;
diff --git a/clang/test/CodeGen/tbaa-struct.cpp b/clang/test/CodeGen/tbaa-struct.cpp
index 2623c42c9a3..260e22a36d8 100644
--- a/clang/test/CodeGen/tbaa-struct.cpp
+++ b/clang/test/CodeGen/tbaa-struct.cpp
@@ -12,7 +12,7 @@ void copy(struct A *a, struct A *b) {
   *a = *b;
 }
 
-// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* %{{.*}}, i8* %{{.*}}, i64 16, i32 4, i1 false), !tbaa.struct [[TS:!.*]]
+// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 4 %{{.*}}, i8* align 4 %{{.*}}, i64 16, i1 false), !tbaa.struct [[TS:!.*]]
 
 struct B {
   char c1;
@@ -24,7 +24,7 @@ void copy2(struct B *a, struct B *b) {
   *a = *b;
 }
 
-// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* %{{.*}}, i8* %{{.*}}, i64 24, i32 4, i1 false), !tbaa.struct [[TS2:!.*]]
+// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 4 %{{.*}}, i8* align 4 %{{.*}}, i64 24, i1 false), !tbaa.struct [[TS2:!.*]]
 
 typedef _Complex int T2;
 typedef _Complex char T5;
@@ -36,7 +36,7 @@ void copy3 (T1 *a, T1 *b) {
   *a = *b;
 }
 
-// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* %{{.*}}, i8* %{{.*}}, i64 12, i32 4, i1 false), !tbaa.struct [[TS3:!.*]]
+// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 4 %{{.*}}, i8* align 4 %{{.*}}, i64 12, i1 false), !tbaa.struct [[TS3:!.*]]
 
 // Make sure that zero-length bitfield works.
 #define ATTR __attribute__ ((ms_struct))
@@ -49,7 +49,7 @@ struct five {
 void copy4(struct five *a, struct five *b) {
   *a = *b;
 }
-// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* %{{.*}}, i8* %{{.*}}, i64 3, i32 1, i1 false), !tbaa.struct [[TS4:!.*]]
+// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 1 %{{.*}}, i8* align 1 %{{.*}}, i64 3, i1 false), !tbaa.struct [[TS4:!.*]]
 
 struct six {
   char a;
@@ -60,7 +60,7 @@ struct six {
 void copy5(struct six *a, struct six *b) {
   *a = *b;
 }
-// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* %{{.*}}, i8* %{{.*}}, i64 6, i32 1, i1 false), !tbaa.struct [[TS5:!.*]]
+// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 1 %{{.*}}, i8* align 1 %{{.*}}, i64 6, i1 false), !tbaa.struct [[TS5:!.*]]
 
 // CHECK: [[TS]] = !{i64 0, i64 2, !{{.*}}, i64 4, i64 4, !{{.*}}, i64 8, i64 1, !{{.*}}, i64 12, i64 4, !{{.*}}}
 // CHECK: [[CHAR:!.*]] = !{!"omnipotent char", !{{.*}}}
diff --git a/clang/test/CodeGen/volatile.c b/clang/test/CodeGen/volatile.c
index 52915f6f4d5..6a23f0188cb 100644
--- a/clang/test/CodeGen/volatile.c
+++ b/clang/test/CodeGen/volatile.c
@@ -199,7 +199,7 @@ int main() {
 // CHECK: store volatile i32 {{.*}}, i32* @vtS
   (void)vF2;
   // From vF2 to a temporary
-// CHECK: call void @llvm.memcpy.{{.*}}(i8* %{{.*}}, i8* {{.*}} @vF2 {{.*}}, i1 true)
+// CHECK: call void @llvm.memcpy.{{.*}}(i8* align {{[0-9]*}} %{{[0-9]*}}, i8* align {{[0-9]*}} {{.*@vF2.*}}, i{{[0-9]*}} {{[0-9]*}}, i1 true)
   vF2 = vF2;
   // vF2 to itself
 // CHECK: call void @llvm.memcpy.{{.*}}(i8* {{.*@vF2.*}}, i8* {{.*@vF2.*}}, i1 true)
@@ -209,6 +209,6 @@ int main() {
 // CHECK: call void @llvm.memcpy.{{.*}}(i8* {{.*@vF2.*}}, i8* {{.*@vF2.*}}, i1 true)
   vF2 = (vF2, vF2);
   // vF2 to a temporary, then vF2 to itself
-// CHECK: call void @llvm.memcpy.{{.*}}(i8* %{{.*}}, i8* {{.*@vF2.*}}, i1 true)
+// CHECK: call void @llvm.memcpy.{{.*}}(i8* align {{[0-9]*}} %{{[0-9]*}}, i8* align {{[0-9]*}} {{.*@vF2.*}}, i{{[0-9]*}} {{[0-9]*}}, i1 true)
 // CHECK: call void @llvm.memcpy.{{.*}}(i8* {{.*@vF2.*}}, i8* {{.*@vF2.*}}, i1 true)
 }
diff --git a/clang/test/CodeGen/x86-atomic-long_double.c b/clang/test/CodeGen/x86-atomic-long_double.c
index 9857c67592f..130ff45c034 100644
--- a/clang/test/CodeGen/x86-atomic-long_double.c
+++ b/clang/test/CodeGen/x86-atomic-long_double.c
@@ -15,12 +15,12 @@ long double testinc(_Atomic long double *addr) {
   // CHECK: [[OLD_VALUE:%.+]] = phi x86_fp80 [ [[LD_VALUE]], %{{.+}} ], [ [[LD_VALUE:%.+]], %[[ATOMIC_OP]] ]
   // CHECK: [[INC_VALUE:%.+]] = fadd x86_fp80 [[OLD_VALUE]],
   // CHECK: [[OLD_VALUE_VOID_ADDR:%.+]] = bitcast x86_fp80* [[OLD_VALUE_ADDR:%.+]] to i8*
-  // CHECK: call void @llvm.memset.p0i8.i64(i8* [[OLD_VALUE_VOID_ADDR]], i8 0, i64 16, i32 16, i1 false)
+  // CHECK: call void @llvm.memset.p0i8.i64(i8* align 16 [[OLD_VALUE_VOID_ADDR]], i8 0, i64 16, i1 false)
   // CHECK: store x86_fp80 [[OLD_VALUE]], x86_fp80* [[OLD_VALUE_ADDR]], align 16
   // CHECK: [[OLD_INT_ADDR:%.+]] = bitcast x86_fp80* [[OLD_VALUE_ADDR]] to i128*
   // CHECK: [[OLD_INT:%.+]] = load i128, i128* [[OLD_INT_ADDR]], align 16
   // CHECK: [[NEW_VALUE_VOID_ADDR:%.+]] = bitcast x86_fp80* [[NEW_VALUE_ADDR:%.+]] to i8*
-  // CHECK: call void @llvm.memset.p0i8.i64(i8* [[NEW_VALUE_VOID_ADDR]], i8 0, i64 16, i32 16, i1 false)
+  // CHECK: call void @llvm.memset.p0i8.i64(i8* align 16 [[NEW_VALUE_VOID_ADDR]], i8 0, i64 16, i1 false)
   // CHECK: store x86_fp80 [[INC_VALUE]], x86_fp80* [[NEW_VALUE_ADDR]], align 16
   // CHECK: [[NEW_INT_ADDR:%.+]] = bitcast x86_fp80* [[NEW_VALUE_ADDR]] to i128*
   // CHECK: [[NEW_INT:%.+]] = load i128, i128* [[NEW_INT_ADDR]], align 16
@@ -46,10 +46,10 @@ long double testinc(_Atomic long double *addr) {
   // CHECK32: [[OLD_VALUE:%.+]] = phi x86_fp80 [ [[LD_VALUE]], %{{.+}} ], [ [[LD_VALUE:%.+]], %[[ATOMIC_OP]] ]
   // CHECK32: [[INC_VALUE:%.+]] = fadd x86_fp80 [[OLD_VALUE]],
   // CHECK32: [[OLD_VALUE_VOID_ADDR:%.+]] = bitcast x86_fp80* [[OLD_VALUE_ADDR:%.+]] to i8*
-  // CHECK32: call void @llvm.memset.p0i8.i64(i8* [[OLD_VALUE_VOID_ADDR]], i8 0, i64 12, i32 4, i1 false)
+  // CHECK32: call void @llvm.memset.p0i8.i64(i8* align 4 [[OLD_VALUE_VOID_ADDR]], i8 0, i64 12, i1 false)
   // CHECK32: store x86_fp80 [[OLD_VALUE]], x86_fp80* [[OLD_VALUE_ADDR]], align 4
   // CHECK32: [[DESIRED_VALUE_VOID_ADDR:%.+]] = bitcast x86_fp80* [[DESIRED_VALUE_ADDR:%.+]] to i8*
-  // CHECK32: call void @llvm.memset.p0i8.i64(i8* [[DESIRED_VALUE_VOID_ADDR]], i8 0, i64 12, i32 4, i1 false)
+  // CHECK32: call void @llvm.memset.p0i8.i64(i8* align 4 [[DESIRED_VALUE_VOID_ADDR]], i8 0, i64 12, i1 false)
   // CHECK32: store x86_fp80 [[INC_VALUE]], x86_fp80* [[DESIRED_VALUE_ADDR]], align 4
   // CHECK32: [[OBJ:%.+]] = bitcast x86_fp80* [[ADDR]] to i8*
   // CHECK32: [[EXPECTED:%.+]] = bitcast x86_fp80* [[OLD_VALUE_ADDR]] to i8*
@@ -77,12 +77,12 @@ long double testdec(_Atomic long double *addr) {
   // CHECK: [[OLD_VALUE:%.+]] = phi x86_fp80 [ [[ORIG_LD_VALUE]], %{{.+}} ], [ [[LD_VALUE:%.+]], %[[ATOMIC_OP]] ]
   // CHECK: [[DEC_VALUE:%.+]] = fadd x86_fp80 [[OLD_VALUE]],
   // CHECK: [[OLD_VALUE_VOID_ADDR:%.+]] = bitcast x86_fp80* [[OLD_VALUE_ADDR:%.+]] to i8*
-  // CHECK: call void @llvm.memset.p0i8.i64(i8* [[OLD_VALUE_VOID_ADDR]], i8 0, i64 16, i32 16, i1 false)
+  // CHECK: call void @llvm.memset.p0i8.i64(i8* align 16 [[OLD_VALUE_VOID_ADDR]], i8 0, i64 16, i1 false)
   // CHECK: store x86_fp80 [[OLD_VALUE]], x86_fp80* [[OLD_VALUE_ADDR]], align 16
   // CHECK: [[OLD_INT_ADDR:%.+]] = bitcast x86_fp80* [[OLD_VALUE_ADDR]] to i128*
   // CHECK: [[OLD_INT:%.+]] = load i128, i128* [[OLD_INT_ADDR]], align 16
   // CHECK: [[NEW_VALUE_VOID_ADDR:%.+]] = bitcast x86_fp80* [[NEW_VALUE_ADDR:%.+]] to i8*
-  // CHECK: call void @llvm.memset.p0i8.i64(i8* [[NEW_VALUE_VOID_ADDR]], i8 0, i64 16, i32 16, i1 false)
+  // CHECK: call void @llvm.memset.p0i8.i64(i8* align 16 [[NEW_VALUE_VOID_ADDR]], i8 0, i64 16, i1 false)
   // CHECK: store x86_fp80 [[DEC_VALUE]], x86_fp80* [[NEW_VALUE_ADDR]], align 16
   // CHECK: [[NEW_INT_ADDR:%.+]] = bitcast x86_fp80* [[NEW_VALUE_ADDR]] to i128*
   // CHECK: [[NEW_INT:%.+]] = load i128, i128* [[NEW_INT_ADDR]], align 16
@@ -108,10 +108,10 @@ long double testdec(_Atomic long double *addr) {
   // CHECK32: [[OLD_VALUE:%.+]] = phi x86_fp80 [ [[ORIG_LD_VALUE]], %{{.+}} ], [ [[LD_VALUE:%.+]], %[[ATOMIC_OP]] ]
   // CHECK32: [[DEC_VALUE:%.+]] = fadd x86_fp80 [[OLD_VALUE]],
   // CHECK32: [[OLD_VALUE_VOID_ADDR:%.+]] = bitcast x86_fp80* [[OLD_VALUE_ADDR:%.+]] to i8*
-  // CHECK32: call void @llvm.memset.p0i8.i64(i8* [[OLD_VALUE_VOID_ADDR]], i8 0, i64 12, i32 4, i1 false)
+  // CHECK32: call void @llvm.memset.p0i8.i64(i8* align 4 [[OLD_VALUE_VOID_ADDR]], i8 0, i64 12, i1 false)
   // CHECK32: store x86_fp80 [[OLD_VALUE]], x86_fp80* [[OLD_VALUE_ADDR]], align 4
   // CHECK32: [[DESIRED_VALUE_VOID_ADDR:%.+]] = bitcast x86_fp80* [[DESIRED_VALUE_ADDR:%.+]] to i8*
-  // CHECK32: call void @llvm.memset.p0i8.i64(i8* [[DESIRED_VALUE_VOID_ADDR]], i8 0, i64 12, i32 4, i1 false)
+  // CHECK32: call void @llvm.memset.p0i8.i64(i8* align 4 [[DESIRED_VALUE_VOID_ADDR]], i8 0, i64 12, i1 false)
   // CHECK32: store x86_fp80 [[DEC_VALUE]], x86_fp80* [[DESIRED_VALUE_ADDR]], align 4
   // CHECK32: [[OBJ:%.+]] = bitcast x86_fp80* [[ADDR]] to i8*
   // CHECK32: [[EXPECTED:%.+]] = bitcast x86_fp80* [[OLD_VALUE_ADDR]] to i8*
@@ -140,12 +140,12 @@ long double testcompassign(_Atomic long double *addr) {
   // CHECK: [[OLD_VALUE:%.+]] = phi x86_fp80 [ [[LD_VALUE]], %{{.+}} ], [ [[LD_VALUE:%.+]], %[[ATOMIC_OP]] ]
   // CHECK: [[SUB_VALUE:%.+]] = fsub x86_fp80 [[OLD_VALUE]],
   // CHECK: [[OLD_VALUE_VOID_ADDR:%.+]] = bitcast x86_fp80* [[OLD_VALUE_ADDR:%.+]] to i8*
-  // CHECK: call void @llvm.memset.p0i8.i64(i8* [[OLD_VALUE_VOID_ADDR]], i8 0, i64 16, i32 16, i1 false)
+  // CHECK: call void @llvm.memset.p0i8.i64(i8* align 16 [[OLD_VALUE_VOID_ADDR]], i8 0, i64 16, i1 false)
   // CHECK: store x86_fp80 [[OLD_VALUE]], x86_fp80* [[OLD_VALUE_ADDR]], align 16
   // CHECK: [[OLD_INT_ADDR:%.+]] = bitcast x86_fp80* [[OLD_VALUE_ADDR]] to i128*
   // CHECK: [[OLD_INT:%.+]] = load i128, i128* [[OLD_INT_ADDR]], align 16
   // CHECK: [[NEW_VALUE_VOID_ADDR:%.+]] = bitcast x86_fp80* [[NEW_VALUE_ADDR:%.+]] to i8*
-  // CHECK: call void @llvm.memset.p0i8.i64(i8* [[NEW_VALUE_VOID_ADDR]], i8 0, i64 16, i32 16, i1 false)
+  // CHECK: call void @llvm.memset.p0i8.i64(i8* align 16 [[NEW_VALUE_VOID_ADDR]], i8 0, i64 16, i1 false)
   // CHECK: store x86_fp80 [[SUB_VALUE]], x86_fp80* [[NEW_VALUE_ADDR]], align 16
   // CHECK: [[NEW_INT_ADDR:%.+]] = bitcast x86_fp80* [[NEW_VALUE_ADDR]] to i128*
   // CHECK: [[NEW_INT:%.+]] = load i128, i128* [[NEW_INT_ADDR]], align 16
@@ -177,10 +177,10 @@ long double testcompassign(_Atomic long double *addr) {
   // CHECK32: [[OLD_VALUE:%.+]] = phi x86_fp80 [ [[LD_VALUE]], %{{.+}} ], [ [[LD_VALUE:%.+]], %[[ATOMIC_OP]] ]
   // CHECK32: [[INC_VALUE:%.+]] = fsub x86_fp80 [[OLD_VALUE]],
   // CHECK32: [[OLD_VALUE_VOID_ADDR:%.+]] = bitcast x86_fp80* [[OLD_VALUE_ADDR:%.+]] to i8*
-  // CHECK32: call void @llvm.memset.p0i8.i64(i8* [[OLD_VALUE_VOID_ADDR]], i8 0, i64 12, i32 4, i1 false)
+  // CHECK32: call void @llvm.memset.p0i8.i64(i8* align 4 [[OLD_VALUE_VOID_ADDR]], i8 0, i64 12, i1 false)
   // CHECK32: store x86_fp80 [[OLD_VALUE]], x86_fp80* [[OLD_VALUE_ADDR]], align 4
   // CHECK32: [[DESIRED_VALUE_VOID_ADDR:%.+]] = bitcast x86_fp80* [[DESIRED_VALUE_ADDR:%.+]] to i8*
-  // CHECK32: call void @llvm.memset.p0i8.i64(i8* [[DESIRED_VALUE_VOID_ADDR]], i8 0, i64 12, i32 4, i1 false)
+  // CHECK32: call void @llvm.memset.p0i8.i64(i8* align 4 [[DESIRED_VALUE_VOID_ADDR]], i8 0, i64 12, i1 false)
   // CHECK32: store x86_fp80 [[INC_VALUE]], x86_fp80* [[DESIRED_VALUE_ADDR]], align 4
   // CHECK32: [[OBJ:%.+]] = bitcast x86_fp80* [[ADDR]] to i8*
   // CHECK32: [[EXPECTED:%.+]] = bitcast x86_fp80* [[OLD_VALUE_ADDR]] to i8*
@@ -203,7 +203,7 @@ long double testassign(_Atomic long double *addr) {
   // CHECK: store x86_fp80* %{{.+}}, x86_fp80** [[ADDR_ADDR:%.+]], align 8
   // CHECK: [[ADDR:%.+]] = load x86_fp80*, x86_fp80** [[ADDR_ADDR]], align 8
   // CHECK: [[STORE_TEMP_VOID_PTR:%.+]] = bitcast x86_fp80* [[STORE_TEMP_PTR:%.+]] to i8*
-  // CHECK: call void @llvm.memset.p0i8.i64(i8* [[STORE_TEMP_VOID_PTR]], i8 0, i64 16, i32 16, i1 false)
+  // CHECK: call void @llvm.memset.p0i8.i64(i8* align 16 [[STORE_TEMP_VOID_PTR]], i8 0, i64 16, i1 false)
   // CHECK: store x86_fp80 {{.+}}, x86_fp80* [[STORE_TEMP_PTR]], align 16
   // CHECK: [[STORE_TEMP_INT_PTR:%.+]] = bitcast x86_fp80* [[STORE_TEMP_PTR]] to i128*
   // CHECK: [[STORE_TEMP_INT:%.+]] = load i128, i128* [[STORE_TEMP_INT_PTR]], align 16
@@ -213,7 +213,7 @@ long double testassign(_Atomic long double *addr) {
   // CHECK32: store x86_fp80* %{{.+}}, x86_fp80** [[ADDR_ADDR:%.+]], align 4
   // CHECK32: [[ADDR:%.+]] = load x86_fp80*, x86_fp80** [[ADDR_ADDR]], align 4
   // CHECK32: [[STORE_TEMP_VOID_PTR:%.+]] = bitcast x86_fp80* [[STORE_TEMP_PTR:%.+]] to i8*
-  // CHECK32: call void @llvm.memset.p0i8.i64(i8* [[STORE_TEMP_VOID_PTR]], i8 0, i64 12, i32 4, i1 false)
+  // CHECK32: call void @llvm.memset.p0i8.i64(i8* align 4 [[STORE_TEMP_VOID_PTR]], i8 0, i64 12, i1 false)
   // CHECK32: store x86_fp80 {{.+}}, x86_fp80* [[STORE_TEMP_PTR]], align 4
   // CHECK32: [[ADDR_VOID:%.+]] = bitcast x86_fp80* [[ADDR]] to i8*
   // CHECK32: [[STORE_TEMP_VOID_PTR:%.+]] = bitcast x86_fp80* [[STORE_TEMP_PTR]] to i8*
@@ -250,12 +250,12 @@ long double test_volatile_inc(volatile _Atomic long double *addr) {
   // CHECK: [[OLD_VALUE:%.+]] = phi x86_fp80 [ [[LD_VALUE]], %{{.+}} ], [ [[LD_VALUE:%.+]], %[[ATOMIC_OP]] ]
   // CHECK: [[INC_VALUE:%.+]] = fadd x86_fp80 [[OLD_VALUE]],
   // CHECK: [[OLD_VALUE_VOID_ADDR:%.+]] = bitcast x86_fp80* [[OLD_VALUE_ADDR:%.+]] to i8*
-  // CHECK: call void @llvm.memset.p0i8.i64(i8* [[OLD_VALUE_VOID_ADDR]], i8 0, i64 16, i32 16, i1 false)
+  // CHECK: call void @llvm.memset.p0i8.i64(i8* align 16 [[OLD_VALUE_VOID_ADDR]], i8 0, i64 16, i1 false)
   // CHECK: store x86_fp80 [[OLD_VALUE]], x86_fp80* [[OLD_VALUE_ADDR]], align 16
   // CHECK: [[OLD_INT_ADDR:%.+]] = bitcast x86_fp80* [[OLD_VALUE_ADDR]] to i128*
   // CHECK: [[OLD_INT:%.+]] = load i128, i128* [[OLD_INT_ADDR]], align 16
   // CHECK: [[NEW_VALUE_VOID_ADDR:%.+]] = bitcast x86_fp80* [[NEW_VALUE_ADDR:%.+]] to i8*
-  // CHECK: call void @llvm.memset.p0i8.i64(i8* [[NEW_VALUE_VOID_ADDR]], i8 0, i64 16, i32 16, i1 false)
+  // CHECK: call void @llvm.memset.p0i8.i64(i8* align 16 [[NEW_VALUE_VOID_ADDR]], i8 0, i64 16, i1 false)
   // CHECK: store x86_fp80 [[INC_VALUE]], x86_fp80* [[NEW_VALUE_ADDR]], align 16
   // CHECK: [[NEW_INT_ADDR:%.+]] = bitcast x86_fp80* [[NEW_VALUE_ADDR]] to i128*
   // CHECK: [[NEW_INT:%.+]] = load i128, i128* [[NEW_INT_ADDR]], align 16
@@ -281,10 +281,10 @@ long double test_volatile_inc(volatile _Atomic long double *addr) {
   // CHECK32: [[OLD_VALUE:%.+]] = phi x86_fp80 [ [[LD_VALUE]], %{{.+}} ], [ [[LD_VALUE:%.+]], %[[ATOMIC_OP]] ]
   // CHECK32: [[INC_VALUE:%.+]] = fadd x86_fp80 [[OLD_VALUE]],
   // CHECK32: [[OLD_VALUE_VOID_ADDR:%.+]] = bitcast x86_fp80* [[OLD_VALUE_ADDR:%.+]] to i8*
-  // CHECK32: call void @llvm.memset.p0i8.i64(i8* [[OLD_VALUE_VOID_ADDR]], i8 0, i64 12, i32 4, i1 false)
+  // CHECK32: call void @llvm.memset.p0i8.i64(i8* align 4 [[OLD_VALUE_VOID_ADDR]], i8 0, i64 12, i1 false)
   // CHECK32: store x86_fp80 [[OLD_VALUE]], x86_fp80* [[OLD_VALUE_ADDR]], align 4
   // CHECK32: [[DESIRED_VALUE_VOID_ADDR:%.+]] = bitcast x86_fp80* [[DESIRED_VALUE_ADDR:%.+]] to i8*
-  // CHECK32: call void @llvm.memset.p0i8.i64(i8* [[DESIRED_VALUE_VOID_ADDR]], i8 0, i64 12, i32 4, i1 false)
+  // CHECK32: call void @llvm.memset.p0i8.i64(i8* align 4 [[DESIRED_VALUE_VOID_ADDR]], i8 0, i64 12, i1 false)
   // CHECK32: store x86_fp80 [[INC_VALUE]], x86_fp80* [[DESIRED_VALUE_ADDR]], align 4
   // CHECK32: [[OBJ:%.+]] = bitcast x86_fp80* [[ADDR]] to i8*
   // CHECK32: [[EXPECTED:%.+]] = bitcast x86_fp80* [[OLD_VALUE_ADDR]] to i8*
@@ -311,12 +311,12 @@ long double test_volatile_dec(volatile _Atomic long double *addr) {
   // CHECK: [[OLD_VALUE:%.+]] = phi x86_fp80 [ [[ORIG_LD_VALUE]], %{{.+}} ], [ [[LD_VALUE:%.+]], %[[ATOMIC_OP]] ]
   // CHECK: [[DEC_VALUE:%.+]] = fadd x86_fp80 [[OLD_VALUE]],
   // CHECK: [[OLD_VALUE_VOID_ADDR:%.+]] = bitcast x86_fp80* [[OLD_VALUE_ADDR:%.+]] to i8*
-  // CHECK: call void @llvm.memset.p0i8.i64(i8* [[OLD_VALUE_VOID_ADDR]], i8 0, i64 16, i32 16, i1 false)
+  // CHECK: call void @llvm.memset.p0i8.i64(i8* align 16 [[OLD_VALUE_VOID_ADDR]], i8 0, i64 16, i1 false)
   // CHECK: store x86_fp80 [[OLD_VALUE]], x86_fp80* [[OLD_VALUE_ADDR]], align 16
   // CHECK: [[OLD_INT_ADDR:%.+]] = bitcast x86_fp80* [[OLD_VALUE_ADDR]] to i128*
   // CHECK: [[OLD_INT:%.+]] = load i128, i128* [[OLD_INT_ADDR]], align 16
   // CHECK: [[NEW_VALUE_VOID_ADDR:%.+]] = bitcast x86_fp80* [[NEW_VALUE_ADDR:%.+]] to i8*
-  // CHECK: call void @llvm.memset.p0i8.i64(i8* [[NEW_VALUE_VOID_ADDR]], i8 0, i64 16, i32 16, i1 false)
+  // CHECK: call void @llvm.memset.p0i8.i64(i8* align 16 [[NEW_VALUE_VOID_ADDR]], i8 0, i64 16, i1 false)
   // CHECK: store x86_fp80 [[DEC_VALUE]], x86_fp80* [[NEW_VALUE_ADDR]], align 16
   // CHECK: [[NEW_INT_ADDR:%.+]] = bitcast x86_fp80* [[NEW_VALUE_ADDR]] to i128*
   // CHECK: [[NEW_INT:%.+]] = load i128, i128* [[NEW_INT_ADDR]], align 16
@@ -342,10 +342,10 @@ long double test_volatile_dec(volatile _Atomic long double *addr) {
   // CHECK32: [[OLD_VALUE:%.+]] = phi x86_fp80 [ [[ORIG_LD_VALUE]], %{{.+}} ], [ [[LD_VALUE:%.+]], %[[ATOMIC_OP]] ]
   // CHECK32: [[DEC_VALUE:%.+]] = fadd x86_fp80 [[OLD_VALUE]],
   // CHECK32: [[OLD_VALUE_VOID_ADDR:%.+]] = bitcast x86_fp80* [[OLD_VALUE_ADDR:%.+]] to i8*
-  // CHECK32: call void @llvm.memset.p0i8.i64(i8* [[OLD_VALUE_VOID_ADDR]], i8 0, i64 12, i32 4, i1 false)
+  // CHECK32: call void @llvm.memset.p0i8.i64(i8* align 4 [[OLD_VALUE_VOID_ADDR]], i8 0, i64 12, i1 false)
   // CHECK32: store x86_fp80 [[OLD_VALUE]], x86_fp80* [[OLD_VALUE_ADDR]], align 4
   // CHECK32: [[DESIRED_VALUE_VOID_ADDR:%.+]] = bitcast x86_fp80* [[DESIRED_VALUE_ADDR:%.+]] to i8*
-  // CHECK32: call void @llvm.memset.p0i8.i64(i8* [[DESIRED_VALUE_VOID_ADDR]], i8 0, i64 12, i32 4, i1 false)
+  // CHECK32: call void @llvm.memset.p0i8.i64(i8* align 4 [[DESIRED_VALUE_VOID_ADDR]], i8 0, i64 12, i1 false)
   // CHECK32: store x86_fp80 [[DEC_VALUE]], x86_fp80* [[DESIRED_VALUE_ADDR]], align 4
   // CHECK32: [[OBJ:%.+]] = bitcast x86_fp80* [[ADDR]] to i8*
   // CHECK32: [[EXPECTED:%.+]] = bitcast x86_fp80* [[OLD_VALUE_ADDR]] to i8*
@@ -373,12 +373,12 @@ long double test_volatile_compassign(volatile _Atomic long double *addr) {
   // CHECK: [[OLD_VALUE:%.+]] = phi x86_fp80 [ [[LD_VALUE]], %{{.+}} ], [ [[LD_VALUE:%.+]], %[[ATOMIC_OP]] ]
   // CHECK: [[SUB_VALUE:%.+]] = fsub x86_fp80 [[OLD_VALUE]],
   // CHECK: [[OLD_VALUE_VOID_ADDR:%.+]] = bitcast x86_fp80* [[OLD_VALUE_ADDR:%.+]] to i8*
-  // CHECK: call void @llvm.memset.p0i8.i64(i8* [[OLD_VALUE_VOID_ADDR]], i8 0, i64 16, i32 16, i1 false)
+  // CHECK: call void @llvm.memset.p0i8.i64(i8* align 16 [[OLD_VALUE_VOID_ADDR]], i8 0, i64 16, i1 false)
   // CHECK: store x86_fp80 [[OLD_VALUE]], x86_fp80* [[OLD_VALUE_ADDR]], align 16
   // CHECK: [[OLD_INT_ADDR:%.+]] = bitcast x86_fp80* [[OLD_VALUE_ADDR]] to i128*
   // CHECK: [[OLD_INT:%.+]] = load i128, i128* [[OLD_INT_ADDR]], align 16
   // CHECK: [[NEW_VALUE_VOID_ADDR:%.+]] = bitcast x86_fp80* [[NEW_VALUE_ADDR:%.+]] to i8*
-  // CHECK: call void @llvm.memset.p0i8.i64(i8* [[NEW_VALUE_VOID_ADDR]], i8 0, i64 16, i32 16, i1 false)
+  // CHECK: call void @llvm.memset.p0i8.i64(i8* align 16 [[NEW_VALUE_VOID_ADDR]], i8 0, i64 16, i1 false)
   // CHECK: store x86_fp80 [[SUB_VALUE]], x86_fp80* [[NEW_VALUE_ADDR]], align 16
   // CHECK: [[NEW_INT_ADDR:%.+]] = bitcast x86_fp80* [[NEW_VALUE_ADDR]] to i128*
   // CHECK: [[NEW_INT:%.+]] = load i128, i128* [[NEW_INT_ADDR]], align 16
@@ -409,10 +409,10 @@ long double test_volatile_compassign(volatile _Atomic long double *addr) {
   // CHECK32: [[OLD_VALUE:%.+]] = phi x86_fp80 [ [[LD_VALUE]], %{{.+}} ], [ [[LD_VALUE:%.+]], %[[ATOMIC_OP]] ]
   // CHECK32: [[INC_VALUE:%.+]] = fsub x86_fp80 [[OLD_VALUE]],
   // CHECK32: [[OLD_VALUE_VOID_ADDR:%.+]] = bitcast x86_fp80* [[OLD_VALUE_ADDR:%.+]] to i8*
-  // CHECK32: call void @llvm.memset.p0i8.i64(i8* [[OLD_VALUE_VOID_ADDR]], i8 0, i64 12, i32 4, i1 false)
+  // CHECK32: call void @llvm.memset.p0i8.i64(i8* align 4 [[OLD_VALUE_VOID_ADDR]], i8 0, i64 12, i1 false)
   // CHECK32: store x86_fp80 [[OLD_VALUE]], x86_fp80* [[OLD_VALUE_ADDR]], align 4
   // CHECK32: [[DESIRED_VALUE_VOID_ADDR:%.+]] = bitcast x86_fp80* [[DESIRED_VALUE_ADDR:%.+]] to i8*
-  // CHECK32: call void @llvm.memset.p0i8.i64(i8* [[DESIRED_VALUE_VOID_ADDR]], i8 0, i64 12, i32 4, i1 false)
+  // CHECK32: call void @llvm.memset.p0i8.i64(i8* align 4 [[DESIRED_VALUE_VOID_ADDR]], i8 0, i64 12, i1 false)
   // CHECK32: store x86_fp80 [[INC_VALUE]], x86_fp80* [[DESIRED_VALUE_ADDR]], align 4
   // CHECK32: [[OBJ:%.+]] = bitcast x86_fp80* [[ADDR]] to i8*
   // CHECK32: [[EXPECTED:%.+]] = bitcast x86_fp80* [[OLD_VALUE_ADDR]] to i8*
@@ -435,7 +435,7 @@ long double test_volatile_assign(volatile _Atomic long double *addr) {
   // CHECK: store x86_fp80* %{{.+}}, x86_fp80** [[ADDR_ADDR:%.+]], align 8
   // CHECK: [[ADDR:%.+]] = load x86_fp80*, x86_fp80** [[ADDR_ADDR]], align 8
   // CHECK: [[STORE_TEMP_VOID_PTR:%.+]] = bitcast x86_fp80* [[STORE_TEMP_PTR:%.+]] to i8*
-  // CHECK: call void @llvm.memset.p0i8.i64(i8* [[STORE_TEMP_VOID_PTR]], i8 0, i64 16, i32 16, i1 false)
+  // CHECK: call void @llvm.memset.p0i8.i64(i8* align 16 [[STORE_TEMP_VOID_PTR]], i8 0, i64 16, i1 false)
   // CHECK: store x86_fp80 {{.+}}, x86_fp80* [[STORE_TEMP_PTR]], align 16
   // CHECK: [[STORE_TEMP_INT_PTR:%.+]] = bitcast x86_fp80* [[STORE_TEMP_PTR]] to i128*
   // CHECK: [[STORE_TEMP_INT:%.+]] = load i128, i128* [[STORE_TEMP_INT_PTR]], align 16
@@ -445,7 +445,7 @@ long double test_volatile_assign(volatile _Atomic long double *addr) {
   // CHECK32: store x86_fp80* %{{.+}}, x86_fp80** [[ADDR_ADDR:%.+]], align 4
   // CHECK32: [[ADDR:%.+]] = load x86_fp80*, x86_fp80** [[ADDR_ADDR]], align 4
   // CHECK32: [[STORE_TEMP_VOID_PTR:%.+]] = bitcast x86_fp80* [[STORE_TEMP_PTR:%.+]] to i8*
-  // CHECK32: call void @llvm.memset.p0i8.i64(i8* [[STORE_TEMP_VOID_PTR]], i8 0, i64 12, i32 4, i1 false)
+  // CHECK32: call void @llvm.memset.p0i8.i64(i8* align 4 [[STORE_TEMP_VOID_PTR]], i8 0, i64 12, i1 false)
   // CHECK32: store x86_fp80 {{.+}}, x86_fp80* [[STORE_TEMP_PTR]], align 4
   // CHECK32: [[ADDR_VOID:%.+]] = bitcast x86_fp80* [[ADDR]] to i8*
   // CHECK32: [[STORE_TEMP_VOID_PTR:%.+]] = bitcast x86_fp80* [[STORE_TEMP_PTR]] to i8*
diff --git a/clang/test/CodeGen/x86_32-arguments-realign.c b/clang/test/CodeGen/x86_32-arguments-realign.c
index 768e1cc4690..b99523b7eee 100644
--- a/clang/test/CodeGen/x86_32-arguments-realign.c
+++ b/clang/test/CodeGen/x86_32-arguments-realign.c
@@ -2,7 +2,7 @@
 // RUN: FileCheck < %t %s
 
 // CHECK-LABEL: define void @f0(%struct.s0* byval align 4)
-// CHECK:   call void @llvm.memcpy.p0i8.p0i8.i32(i8* %{{.*}}, i8* %{{.*}}, i32 16, i32 4, i1 false)
+// CHECK:   call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 16 %{{.*}}, i8* align 4 %{{.*}}, i32 16, i1 false)
 // CHECK: }
 struct s0 { long double a; };
 void f0(struct s0 a0) {
diff --git a/clang/test/CodeGen/x86_64-arguments.c b/clang/test/CodeGen/x86_64-arguments.c
index bb9fba19026..fac316420c9 100644
--- a/clang/test/CodeGen/x86_64-arguments.c
+++ b/clang/test/CodeGen/x86_64-arguments.c
@@ -433,7 +433,7 @@ void test51(struct test51_s *s, __builtin_va_list argList) {
 // CHECK-NEXT: [[CASTED_VALUE_ADDR:%.*]] = bitcast i8* [[VALUE_ADDR]] to [[STRUCT_TEST51]]
 // CHECK-NEXT: [[CASTED_TMP_ADDR:%.*]] = bitcast [[STRUCT_TEST51]]* [[TMP_ADDR]] to i8*
 // CHECK-NEXT: [[RECASTED_VALUE_ADDR:%.*]] = bitcast [[STRUCT_TEST51]]* [[CASTED_VALUE_ADDR]] to i8*
-// CHECK-NEXT: call void @llvm.memcpy.p0i8.p0i8.i64(i8* [[CASTED_TMP_ADDR]], i8* [[RECASTED_VALUE_ADDR]], i64 16, i32 8, i1 false)
+// CHECK-NEXT: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 16 [[CASTED_TMP_ADDR]], i8* align 8 [[RECASTED_VALUE_ADDR]], i64 16, i1 false)
 // CHECK-NEXT: add i32 {{.*}}, 16
 // CHECK-NEXT: store i32 {{.*}}, i32* {{.*}}
 // CHECK-NEXT: br label
diff --git a/clang/test/CodeGen/xcore-abi.c b/clang/test/CodeGen/xcore-abi.c
index 2bac78d92ed..90306ce31a5 100644
--- a/clang/test/CodeGen/xcore-abi.c
+++ b/clang/test/CodeGen/xcore-abi.c
@@ -80,7 +80,7 @@ void testva (int n, ...) {
   // CHECK: store i8* [[IN]], i8** [[AP]]
   // CHECK: [[V1:%[a-z0-9]+]] = bitcast %struct.x* [[V:%[a-z0-9]+]] to i8*
   // CHECK: [[P1:%[a-z0-9]+]] = bitcast %struct.x* [[P]] to i8*
-  // CHECK: call void @llvm.memcpy.p0i8.p0i8.i32(i8* [[V1]], i8* [[P1]], i32 20, i32 4, i1 false)
+  // CHECK: call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 4 [[V1]], i8* align 4 [[P1]], i32 20, i1 false)
   // CHECK: [[V2:%[a-z0-9]+]] = bitcast %struct.x* [[V]] to i8*
   // CHECK: call void @f(i8* [[V2]])
 
@@ -93,7 +93,7 @@ void testva (int n, ...) {
   // CHECK: store i8* [[IN]], i8** [[AP]]
   // CHECK: [[V1:%[a-z0-9]+]] = bitcast [4 x i32]* [[V0:%[a-z0-9]+]] to i8*
   // CHECK: [[P1:%[a-z0-9]+]] = bitcast [4 x i32]* [[P]] to i8*
-  // CHECK: call void @llvm.memcpy.p0i8.p0i8.i32(i8* [[V1]], i8* [[P1]], i32 16, i32 4, i1 false)
+  // CHECK: call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 4 [[V1]], i8* align 4 [[P1]], i32 16, i1 false)
   // CHECK: [[V2:%[a-z0-9]+]] = getelementptr inbounds [4 x i32], [4 x i32]* [[V0]], i32 0, i32 0
   // CHECK: store i32* [[V2]], i32** [[V:%[a-z0-9]+]], align 4
   // CHECK: [[V3:%[a-z0-9]+]] = load i32*, i32** [[V]], align 4