Change memcpy/memove/memset to have dest and source alignment attributes.

Summary:
  This change is step three in the series of changes to remove alignment argument from
memcpy/memmove/memset in favour of alignment attributes. Steps:

Step 1) Remove alignment parameter and create alignment parameter attributes for
memcpy/memmove/memset. ( rL322965, rC322964, rL322963 )
Step 2) Expand the IRBuilder API to allow creation of memcpy/memmove with differing
source and dest alignments. ( rL323597 )
Step 3) Update Clang to use the new IRBuilder API.
Step 4) Update Polly to use the new IRBuilder API.
Step 5) Update LLVM passes that create memcpy/memmove calls to use the new IRBuilder API,
and those that use use MemIntrinsicInst::[get|set]Alignment() to use getDestAlignment()
and getSourceAlignment() instead.
Step 6) Remove the single-alignment IRBuilder API for memcpy/memmove, and the
MemIntrinsicInst::[get|set]Alignment() methods.

Reference
   http://lists.llvm.org/pipermail/llvm-dev/2015-August/089384.html
   http://lists.llvm.org/pipermail/llvm-commits/Week-of-Mon-20151109/312083.html

Reviewers: rjmccall

Subscribers: jyknight, nemanjai, nhaehnle, javed.absar, sbc100, aheejin, kbarton, fedor.sergeev, cfe-commits

Differential Revision: https://reviews.llvm.org/D41677

llvm-svn: 323617

14 files changed, 40 insertions, 40 deletions

diff --git a/clang/test/CodeGen/arm-arguments.c b/clang/test/CodeGen/arm-arguments.c
index 0c3167fe0c1..ef4e76054ff 100644
--- a/clang/test/CodeGen/arm-arguments.c
+++ b/clang/test/CodeGen/arm-arguments.c
@@ -215,6 +215,6 @@ float32x4_t f35(int i, s35_with_align s1, s35_with_align s2) {
 // 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* align 8 %[[b]], i8* align 8 %[[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 6ba67c2aa12..cee59b87099 100644
--- a/clang/test/CodeGen/arm64-be-bitfield.c
+++ b/clang/test/CodeGen/arm64-be-bitfield.c
@@ -7,6 +7,6 @@ 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* align 4 [[BITCAST]], i64 4
+// IR: call void @llvm.memcpy.p0i8.p0i8.i64(i8* {{.*}}, i8* align 8 [[BITCAST]], i64 4
   return bp11.b2;
 }
diff --git a/clang/test/CodeGen/block-byref-aggr.c b/clang/test/CodeGen/block-byref-aggr.c
index e984177fa96..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* align 4 [[T2]], i8* align 4 [[T3]], i64 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* align 4 [[T2]], i8* align 4 [[T3]], i64 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* align 4 [[T2]], i8* align 4 [[T3]], i64 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 46127f1c952..d93a5aadae2 100644
--- a/clang/test/CodeGen/builtin-memfns.c
+++ b/clang/test/CodeGen/builtin-memfns.c
@@ -73,7 +73,7 @@ struct PS {
 struct PS ps;
 void test8(int *arg) {
   // CHECK: @test8
-  // CHECK: call void @llvm.memcpy{{.*}} align 1 {{.*}} align 1 {{.*}} 16, i1 false)
+  // CHECK: call void @llvm.memcpy{{.*}} align 4 {{.*}} align 1 {{.*}} 16, i1 false)
   __builtin_memcpy(arg, ps.modes, sizeof(struct PS));
 }
 
diff --git a/clang/test/CodeGen/c11atomics-ios.c b/clang/test/CodeGen/c11atomics-ios.c
index 6e3170f1237..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* align 2 [[T1]], i8* align 2 [[T2]], i32 8, 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*
@@ -183,7 +183,7 @@ 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* align 2 [[T1]], i8* align 2 [[T2]], i32 6, 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]]
@@ -192,7 +192,7 @@ void testPromotedStruct(_Atomic(PS) *fp) {
 // 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* align 2 [[T2]], i8* align 2 [[T3]], i32 6, 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* align 2 [[AGG_RESULT8]], i8* align 2 [[ATOMIC_RES8]], i32 6, 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);
 }
@@ -236,7 +236,7 @@ void test_promoted_store(_Atomic(PS) *addr, PS *val) {
   // 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* align 2 [[ATOMIC_VAL8]], i8* align 2 [[NONATOMIC_TMP8]], i64 6, 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
@@ -261,7 +261,7 @@ PS test_promoted_exchange(_Atomic(PS) *addr, PS *val) {
   // 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* align 2 [[ATOMIC_VAL8]], i8* align 2 [[NONATOMIC_TMP8]], i64 6, 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* align 2 [[AGG_RESULT8]], i8* align 2 [[ATOMIC_RES8]], i32 6, 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);
 }
 
@@ -295,11 +295,11 @@ _Bool test_promoted_cmpxchg(_Atomic(PS) *addr, PS *desired, PS *new) {
   // 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* align 2 [[ATOMIC_DESIRED8]], i8* align 2 [[DESIRED8]], i64 6, 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* align 2 [[ATOMIC_NEW8]], i8* align 2 [[NONATOMIC_TMP8]], i64 6, 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 259a63c7934..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* align 2 [[T1]], i8* align 2 [[T2]], i32 8, 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)
@@ -335,7 +335,7 @@ 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* align 2 [[T1]], i8* align 2 [[T2]], i32 6, 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]]
@@ -344,7 +344,7 @@ void testPromotedStruct(_Atomic(PS) *fp) {
 // 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* align 2 [[T2]], i8* align 2 [[T3]], i32 6, 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* align 2 [[T1]], i8* align 2 [[T2]], i32 6, 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* align 2 [[AGG_RESULT8]], i8* align 2 [[ATOMIC_RES8]], i32 6, 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);
 }
@@ -402,7 +402,7 @@ void test_promoted_store(_Atomic(PS) *addr, PS *val) {
   // 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* align 2 [[ATOMIC_VAL8]], i8* align 2 [[NONATOMIC_TMP8]], i64 6, 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
@@ -427,7 +427,7 @@ PS test_promoted_exchange(_Atomic(PS) *addr, PS *val) {
   // 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* align 2 [[ATOMIC_VAL8]], i8* align 2 [[NONATOMIC_TMP8]], i64 6, 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* align 2 [[AGG_RESULT8]], i8* align 2 [[ATOMIC_RES8]], i32 6, 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);
 }
 
@@ -461,11 +461,11 @@ _Bool test_promoted_cmpxchg(_Atomic(PS) *addr, PS *desired, PS *new) {
   // 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* align 2 [[ATOMIC_DESIRED8]], i8* align 2 [[DESIRED8]], i64 6, 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* align 2 [[ATOMIC_NEW8]], i8* align 2 [[NONATOMIC_TMP8]], i64 6, 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/packed-nest-unpacked.c b/clang/test/CodeGen/packed-nest-unpacked.c
index ac657032ee3..4124a8ade86 100644
--- a/clang/test/CodeGen/packed-nest-unpacked.c
+++ b/clang/test/CodeGen/packed-nest-unpacked.c
@@ -42,7 +42,7 @@ int test5() {
 // <rdar://problem/11220251>
 void test6() {
   // CHECK: @test6
-  // 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 1 %{{.*}}, i64 24, 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/ppc64-align-struct.c b/clang/test/CodeGen/ppc64-align-struct.c
index 9152e067803..3cf3e0475b3 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* align 4 [[DEST]], i8* align 4 [[SRC]], i64 8, 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;
@@ -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* align 16 [[DEST]], i8* align 16 [[SRC]], i64 32, 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* align 4 [[DEST]], i8* align 4 [[SRC]], i64 12, 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* align 8 [[DEST]], i8* align 8 [[SRC]], i64 16, 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, ...)
 {
diff --git a/clang/test/CodeGen/ppc64-soft-float.c b/clang/test/CodeGen/ppc64-soft-float.c
index 7c298cf234e..84ac2d55b63 100644
--- a/clang/test/CodeGen/ppc64-soft-float.c
+++ b/clang/test/CodeGen/ppc64-soft-float.c
@@ -92,7 +92,7 @@ void call_f2(void) { global_f2 = func_f2(global_f2); }
 // CHECK-BE: %[[TMP0:[^ ]+]] = alloca %struct.f3, align 4
 // CHECK: %[[TMP1:[^ ]+]] = alloca [2 x i64]
 // CHECK: %[[TMP2:[^ ]+]] = bitcast [2 x i64]* %[[TMP1]] to i8*
-// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 4 %[[TMP2]], i8* align 4 bitcast (%struct.f3* @global_f3 to i8*), i64 12, i1 false)
+// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 8 %[[TMP2]], i8* align 4 bitcast (%struct.f3* @global_f3 to i8*), i64 12, i1 false)
 // CHECK: %[[TMP3:[^ ]+]] = load [2 x i64], [2 x i64]* %[[TMP1]]
 // CHECK-LE: call { i64, i64 } @func_f3([2 x i64] %[[TMP3]])
 // CHECK-BE: call void @func_f3(%struct.f3* sret %[[TMP0]], [2 x i64] %[[TMP3]])
@@ -111,7 +111,7 @@ void call_f4(void) { global_f4 = func_f4(global_f4); }
 // CHECK: %[[TMP0:[^ ]+]] = alloca %struct.f5, align 4
 // CHECK: %[[TMP1:[^ ]+]] = alloca [3 x i64]
 // CHECK: %[[TMP2:[^ ]+]] = bitcast [3 x i64]* %[[TMP1]] to i8*
-// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 4 %[[TMP2]], i8* align 4 bitcast (%struct.f5* @global_f5 to i8*), i64 20, i1 false)
+// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 8 %[[TMP2]], i8* align 4 bitcast (%struct.f5* @global_f5 to i8*), i64 20, i1 false)
 // CHECK: %[[TMP3:[^ ]+]] = load [3 x i64], [3 x i64]* %[[TMP1]]
 // CHECK: call void @func_f5(%struct.f5* sret %[[TMP0]], [3 x i64] %[[TMP3]])
 struct f5 global_f5;
@@ -128,7 +128,7 @@ void call_f6(void) { global_f6 = func_f6(global_f6); }
 // CHECK: %[[TMP0:[^ ]+]] = alloca %struct.f7, align 4
 // CHECK: %[[TMP1:[^ ]+]] = alloca [4 x i64], align 8
 // CHECK: %[[TMP2:[^ ]+]] = bitcast [4 x i64]* %[[TMP1]] to i8*
-// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 4 %[[TMP2]], i8* align 4 bitcast (%struct.f7* @global_f7 to i8*), i64 28, i1 false)
+// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 8 %[[TMP2]], i8* align 4 bitcast (%struct.f7* @global_f7 to i8*), i64 28, i1 false)
 // CHECK: %[[TMP3:[^ ]+]] = load [4 x i64], [4 x i64]* %[[TMP1]], align 8
 // CHECK: call void @func_f7(%struct.f7* sret %[[TMP0]], [4 x i64] %[[TMP3]])
 struct f7 global_f7;
@@ -144,7 +144,7 @@ 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* align 4 %[[TMP2]], i8* align 4 bitcast (%struct.f9* @global_f9 to i8*), i64 36, 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;
@@ -162,7 +162,7 @@ void call_fab(void) { global_fab = func_fab(global_fab); }
 // CHECK-BE: %[[TMPX:[^ ]+]] = alloca %struct.fabc, align 4
 // CHECK: %[[TMP0:[^ ]+]] = alloca [2 x i64], align 8
 // CHECK: %[[TMP2:[^ ]+]] = bitcast [2 x i64]* %[[TMP0]] to i8*
-// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 4 %[[TMP2]], i8* align 4 bitcast (%struct.fabc* @global_fabc to i8*), i64 12, i1 false)
+// CHECK: call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 8 %[[TMP2]], i8* align 4 bitcast (%struct.fabc* @global_fabc to i8*), i64 12, i1 false)
 // CHECK: %[[TMP3:[^ ]+]] = load [2 x i64], [2 x i64]* %[[TMP0]], align 8
 // CHECK-LE: %call = call { i64, i64 } @func_fabc([2 x i64] %[[TMP3]])
 // CHECK-BE: call void @func_fabc(%struct.fabc* sret %[[TMPX]], [2 x i64] %[[TMP3]])
diff --git a/clang/test/CodeGen/ppc64le-aggregates.c b/clang/test/CodeGen/ppc64le-aggregates.c
index 68f28d34b7c..a7780beec5f 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* align 4 %[[TMP2]], i8* align 4 bitcast (%struct.f9* @global_f9 to i8*), i64 36, 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/riscv32-abi.c b/clang/test/CodeGen/riscv32-abi.c
index d3997c2ae94..ec98e3d8db0 100644
--- a/clang/test/CodeGen/riscv32-abi.c
+++ b/clang/test/CodeGen/riscv32-abi.c
@@ -385,7 +385,7 @@ double f_va_3(char *fmt, ...) {
 // CHECK-NEXT:    [[TMP5:%.*]] = bitcast i8* [[ARGP_CUR4]] to %struct.tiny*
 // CHECK-NEXT:    [[TMP6:%.*]] = bitcast %struct.tiny* [[TS]] to i8*
 // CHECK-NEXT:    [[TMP7:%.*]] = bitcast %struct.tiny* [[TMP5]] to i8*
-// CHECK-NEXT:    call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 1 [[TMP6]], i8* align 1 [[TMP7]], i32 4, i1 false)
+// CHECK-NEXT:    call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 1 [[TMP6]], i8* align 4 [[TMP7]], i32 4, i1 false)
 // CHECK-NEXT:    [[ARGP_CUR6:%.*]] = load i8*, i8** [[VA]], align 4
 // CHECK-NEXT:    [[ARGP_NEXT7:%.*]] = getelementptr inbounds i8, i8* [[ARGP_CUR6]], i32 8
 // CHECK-NEXT:    store i8* [[ARGP_NEXT7]], i8** [[VA]], align 4
diff --git a/clang/test/CodeGen/riscv64-abi.c b/clang/test/CodeGen/riscv64-abi.c
index 084fc798017..7a0f065fb4c 100644
--- a/clang/test/CodeGen/riscv64-abi.c
+++ b/clang/test/CodeGen/riscv64-abi.c
@@ -375,7 +375,7 @@ long double f_va_3(char *fmt, ...) {
 // CHECK-NEXT:    [[TMP2:%.*]] = bitcast i8* [[ARGP_CUR2]] to %struct.tiny*
 // CHECK-NEXT:    [[TMP3:%.*]] = bitcast %struct.tiny* [[TS]] to i8*
 // CHECK-NEXT:    [[TMP4:%.*]] = bitcast %struct.tiny* [[TMP2]] to i8*
-// CHECK-NEXT:    call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 2 [[TMP3]], i8* align 2 [[TMP4]], i64 8, i1 false)
+// CHECK-NEXT:    call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 2 [[TMP3]], i8* align 8 [[TMP4]], i64 8, i1 false)
 // CHECK-NEXT:    [[ARGP_CUR4:%.*]] = load i8*, i8** [[VA]], align 8
 // CHECK-NEXT:    [[ARGP_NEXT5:%.*]] = getelementptr inbounds i8, i8* [[ARGP_CUR4]], i64 16
 // CHECK-NEXT:    store i8* [[ARGP_NEXT5]], i8** [[VA]], align 8
diff --git a/clang/test/CodeGen/x86_32-arguments-realign.c b/clang/test/CodeGen/x86_32-arguments-realign.c
index 32149feae37..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* align 4 %{{.*}}, i8* align 4 %{{.*}}, i32 16, 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 c550d3a37ef..548980b32ae 100644
--- a/clang/test/CodeGen/x86_64-arguments.c
+++ b/clang/test/CodeGen/x86_64-arguments.c
@@ -434,7 +434,7 @@ 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* align 8 [[CASTED_TMP_ADDR]], i8* align 8 [[RECASTED_VALUE_ADDR]], i64 16, 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