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

This reverts commit r253511. This likely broke the bots in http://lab.llvm.org:8011/builders/clang-ppc64-elf-linux2/builds/20202 http://bb.pgr.jp/builders/clang-3stage-i686-linux/builds/3787 git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@253543 91177308-0d34-0410-b5e6-96231b3b80d8
author: Pete Cooper <peter_cooper@apple.com> 2015-11-19 05:56:52 +0000
committer: Pete Cooper <peter_cooper@apple.com> 2015-11-19 05:56:52 +0000
commit: 6d024c616a2a4959f8dfe5c64d27f89b394cf042 (patch)
tree: 3010d09bb33fb9c353cea4ee911a8023854f9a1d /test/CodeGen
parent: 0ab4d52b0ad3b1d16238aed4c3f8d50fa8bf8b40 (diff)
124 files changed, 604 insertions, 604 deletions
diff --git a/test/CodeGen/AArch64/PBQP-csr.ll b/test/CodeGen/AArch64/PBQP-csr.ll
index 19729850ae9..16d7f8cb7a5 100644
--- a/test/CodeGen/AArch64/PBQP-csr.ll
+++ b/test/CodeGen/AArch64/PBQP-csr.ll
@@ -22,7 +22,7 @@ entry:
   %z.i60 = getelementptr inbounds %rs, %rs* %r, i64 0, i32 9, i32 2
   %na = getelementptr inbounds %rs, %rs* %r, i64 0, i32 0
   %0 = bitcast double* %x.i to i8*
-  call void @llvm.memset.p0i8.i64(i8* %0, i8 0, i64 72, i1 false)
+  call void @llvm.memset.p0i8.i64(i8* %0, i8 0, i64 72, i32 8, i1 false)
   %1 = load i32, i32* %na, align 4
   %cmp70 = icmp sgt i32 %1, 0
   br i1 %cmp70, label %for.body.lr.ph, label %for.end
@@ -87,5 +87,5 @@ for.end:                                          ; preds = %for.end.loopexit, %
 }
 
 ; Function Attrs: nounwind
-declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i1)
+declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i32, i1)
 
diff --git a/test/CodeGen/AArch64/aarch64-deferred-spilling.ll b/test/CodeGen/AArch64/aarch64-deferred-spilling.ll
index 0a26178c319..7accdced7d4 100644
--- a/test/CodeGen/AArch64/aarch64-deferred-spilling.ll
+++ b/test/CodeGen/AArch64/aarch64-deferred-spilling.ll
@@ -41,7 +41,7 @@ declare i32 @fprintf(%struct.__sFILE* nocapture, i8* nocapture readonly, ...)
 
 declare void @bar(i32)
 
-declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i1)
+declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i32, i1)
 
 define i32 @foo(%struct.DState* %s) {
 entry:
@@ -71,7 +71,7 @@ if.end.thread:                                    ; preds = %entry
   %save_zvec = getelementptr inbounds %struct.DState, %struct.DState* %s, i64 0, i32 57
   %save_zj = getelementptr inbounds %struct.DState, %struct.DState* %s, i64 0, i32 58
   %tmp1 = bitcast i32* %save_i to i8*
-  call void @llvm.memset.p0i8.i64(i8* %tmp1, i8 0, i64 108, i1 false)
+  call void @llvm.memset.p0i8.i64(i8* %tmp1, i8 0, i64 108, i32 4, i1 false)
   br label %sw.default
 
 if.end:                                           ; preds = %entry
diff --git a/test/CodeGen/AArch64/arm64-2012-05-07-MemcpyAlignBug.ll b/test/CodeGen/AArch64/arm64-2012-05-07-MemcpyAlignBug.ll
index 5b1216fc858..d59b0d00438 100644
--- a/test/CodeGen/AArch64/arm64-2012-05-07-MemcpyAlignBug.ll
+++ b/test/CodeGen/AArch64/arm64-2012-05-07-MemcpyAlignBug.ll
@@ -14,8 +14,8 @@
 ; CHECK-NEXT: str  [[VAL2]], [x0]
 
 define void @foo(i8* %a) {
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %a, i8* bitcast ([3 x i32]* @b to i8*), i64 12, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %a, i8* bitcast ([3 x i32]* @b to i8*), i64 12, i32 4, i1 false)
   ret void
 }
 
-declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i32, i1) nounwind
diff --git a/test/CodeGen/AArch64/arm64-abi-varargs.ll b/test/CodeGen/AArch64/arm64-abi-varargs.ll
index 73012d014a4..03414b56144 100644
--- a/test/CodeGen/AArch64/arm64-abi-varargs.ll
+++ b/test/CodeGen/AArch64/arm64-abi-varargs.ll
@@ -168,10 +168,10 @@ entry:
   %4 = bitcast i8* %ap.align to %struct.s41*
   %5 = bitcast %struct.s41* %vs to i8*
   %6 = bitcast %struct.s41* %4 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %5, i8* %6, i64 16, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %5, i8* %6, i64 16, i32 16, i1 false)
   ret void
 }
-declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i32, i1) nounwind
 
 define void @bar2(i32 %x, i128 %s41.coerce) nounwind {
 entry:
diff --git a/test/CodeGen/AArch64/arm64-abi_align.ll b/test/CodeGen/AArch64/arm64-abi_align.ll
index 4b9ae55cb1e..dc9884f12f5 100644
--- a/test/CodeGen/AArch64/arm64-abi_align.ll
+++ b/test/CodeGen/AArch64/arm64-abi_align.ll
@@ -301,14 +301,14 @@ entry:
   %tmp = alloca %struct.s42, align 4
   %tmp1 = alloca %struct.s42, align 4
   %0 = bitcast %struct.s42* %tmp to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 4 %0, i8* align 4 bitcast (%struct.s42* @g42 to i8*), i64 24, i1 false), !tbaa.struct !4
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %0, i8* bitcast (%struct.s42* @g42 to i8*), i64 24, i32 4, i1 false), !tbaa.struct !4
   %1 = bitcast %struct.s42* %tmp1 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 4 %1, i8* align 4 bitcast (%struct.s42* @g42_2 to i8*), i64 24, i1 false), !tbaa.struct !4
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %1, i8* bitcast (%struct.s42* @g42_2 to i8*), i64 24, i32 4, i1 false), !tbaa.struct !4
   %call = call i32 @f42(i32 3, %struct.s42* %tmp, %struct.s42* %tmp1) #5
   ret i32 %call
 }
 
-declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i1) #4
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i32, i1) #4
 
 declare i32 @f42_stack(i32 %i, i32 %i2, i32 %i3, i32 %i4, i32 %i5, i32 %i6,
                        i32 %i7, i32 %i8, i32 %i9, %struct.s42* nocapture %s1,
@@ -347,9 +347,9 @@ entry:
   %tmp = alloca %struct.s42, align 4
   %tmp1 = alloca %struct.s42, align 4
   %0 = bitcast %struct.s42* %tmp to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 4 %0, i8* align 4 bitcast (%struct.s42* @g42 to i8*), i64 24, i1 false), !tbaa.struct !4
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %0, i8* bitcast (%struct.s42* @g42 to i8*), i64 24, i32 4, i1 false), !tbaa.struct !4
   %1 = bitcast %struct.s42* %tmp1 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 4 %1, i8* align 4 bitcast (%struct.s42* @g42_2 to i8*), i64 24, i1 false), !tbaa.struct !4
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %1, i8* bitcast (%struct.s42* @g42_2 to i8*), i64 24, i32 4, i1 false), !tbaa.struct !4
   %call = call i32 @f42_stack(i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7,
                        i32 8, i32 9, %struct.s42* %tmp, %struct.s42* %tmp1) #5
   ret i32 %call
@@ -415,9 +415,9 @@ entry:
   %tmp = alloca %struct.s43, align 16
   %tmp1 = alloca %struct.s43, align 16
   %0 = bitcast %struct.s43* %tmp to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 16 %0, i8* align 16 bitcast (%struct.s43* @g43 to i8*), i64 32, i1 false), !tbaa.struct !4
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %0, i8* bitcast (%struct.s43* @g43 to i8*), i64 32, i32 16, i1 false), !tbaa.struct !4
   %1 = bitcast %struct.s43* %tmp1 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 16 %1, i8* align 16 bitcast (%struct.s43* @g43_2 to i8*), i64 32, i1 false), !tbaa.struct !4
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %1, i8* bitcast (%struct.s43* @g43_2 to i8*), i64 32, i32 16, i1 false), !tbaa.struct !4
   %call = call i32 @f43(i32 3, %struct.s43* %tmp, %struct.s43* %tmp1) #5
   ret i32 %call
 }
@@ -466,9 +466,9 @@ entry:
   %tmp = alloca %struct.s43, align 16
   %tmp1 = alloca %struct.s43, align 16
   %0 = bitcast %struct.s43* %tmp to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 16 %0, i8* align 16 bitcast (%struct.s43* @g43 to i8*), i64 32, i1 false), !tbaa.struct !4
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %0, i8* bitcast (%struct.s43* @g43 to i8*), i64 32, i32 16, i1 false), !tbaa.struct !4
   %1 = bitcast %struct.s43* %tmp1 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 16 %1, i8* align 16 bitcast (%struct.s43* @g43_2 to i8*), i64 32, i1 false), !tbaa.struct !4
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %1, i8* bitcast (%struct.s43* @g43_2 to i8*), i64 32, i32 16, i1 false), !tbaa.struct !4
   %call = call i32 @f43_stack(i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7,
                        i32 8, i32 9, %struct.s43* %tmp, %struct.s43* %tmp1) #5
   ret i32 %call
diff --git a/test/CodeGen/AArch64/arm64-fast-isel-intrinsic.ll b/test/CodeGen/AArch64/arm64-fast-isel-intrinsic.ll
index dfb0dc67ffa..bb2889eaf4b 100644
--- a/test/CodeGen/AArch64/arm64-fast-isel-intrinsic.ll
+++ b/test/CodeGen/AArch64/arm64-fast-isel-intrinsic.ll
@@ -11,11 +11,11 @@ define void @t1() {
 ; ARM64: movz x2, #0x50
 ; ARM64: uxtb w1, w9
 ; ARM64: bl _memset
-  call void @llvm.memset.p0i8.i64(i8* getelementptr inbounds ([80 x i8], [80 x i8]* @message, i32 0, i32 0), i8 0, i64 80, i1 false)
+  call void @llvm.memset.p0i8.i64(i8* getelementptr inbounds ([80 x i8], [80 x i8]* @message, i32 0, i32 0), i8 0, i64 80, i32 16, i1 false)
   ret void
 }
 
-declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i1)
+declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i32, i1)
 
 define void @t2() {
 ; ARM64-LABEL: t2
@@ -25,11 +25,11 @@ define void @t2() {
 ; ARM64: add x1, x8, _message@PAGEOFF
 ; ARM64: movz x2, #0x50
 ; ARM64: bl _memcpy
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* getelementptr inbounds ([80 x i8], [80 x i8]* @temp, i32 0, i32 0), i8* getelementptr inbounds ([80 x i8], [80 x i8]* @message, i32 0, i32 0), i64 80, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* getelementptr inbounds ([80 x i8], [80 x i8]* @temp, i32 0, i32 0), i8* getelementptr inbounds ([80 x i8], [80 x i8]* @message, i32 0, i32 0), i64 80, i32 16, i1 false)
   ret void
 }
 
-declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i1)
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i32, i1)
 
 define void @t3() {
 ; ARM64-LABEL: t3
@@ -39,11 +39,11 @@ define void @t3() {
 ; ARM64: add x1, x8, _message@PAGEOFF
 ; ARM64: movz x2, #0x14
 ; ARM64: bl _memmove
-  call void @llvm.memmove.p0i8.p0i8.i64(i8* getelementptr inbounds ([80 x i8], [80 x i8]* @temp, i32 0, i32 0), i8* getelementptr inbounds ([80 x i8], [80 x i8]* @message, i32 0, i32 0), i64 20, i1 false)
+  call void @llvm.memmove.p0i8.p0i8.i64(i8* getelementptr inbounds ([80 x i8], [80 x i8]* @temp, i32 0, i32 0), i8* getelementptr inbounds ([80 x i8], [80 x i8]* @message, i32 0, i32 0), i64 20, i32 16, i1 false)
   ret void
 }
 
-declare void @llvm.memmove.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i1)
+declare void @llvm.memmove.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i32, i1)
 
 define void @t4() {
 ; ARM64-LABEL: t4
@@ -58,7 +58,7 @@ define void @t4() {
 ; ARM64: ldrb w11, [x9, #16]
 ; ARM64: strb w11, [x8, #16]
 ; ARM64: ret
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* getelementptr inbounds ([80 x i8], [80 x i8]* @temp, i32 0, i32 0), i8* getelementptr inbounds ([80 x i8], [80 x i8]* @message, i32 0, i32 0), i64 17, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* getelementptr inbounds ([80 x i8], [80 x i8]* @temp, i32 0, i32 0), i8* getelementptr inbounds ([80 x i8], [80 x i8]* @message, i32 0, i32 0), i64 17, i32 16, i1 false)
   ret void
 }
 
@@ -75,7 +75,7 @@ define void @t5() {
 ; ARM64: ldrb w11, [x9, #16]
 ; ARM64: strb w11, [x8, #16]
 ; ARM64: ret
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* getelementptr inbounds ([80 x i8], [80 x i8]* @temp, i32 0, i32 0), i8* getelementptr inbounds ([80 x i8], [80 x i8]* @message, i32 0, i32 0), i64 17, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* getelementptr inbounds ([80 x i8], [80 x i8]* @temp, i32 0, i32 0), i8* getelementptr inbounds ([80 x i8], [80 x i8]* @message, i32 0, i32 0), i64 17, i32 8, i1 false)
   ret void
 }
 
@@ -92,7 +92,7 @@ define void @t6() {
 ; ARM64: ldrb w10, [x9, #8]
 ; ARM64: strb w10, [x8, #8]
 ; ARM64: ret
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 4 getelementptr inbounds ([80 x i8], [80 x i8]* @temp, i32 0, i32 0), i8* align 4 getelementptr inbounds ([80 x i8], [80 x i8]* @message, i32 0, i32 0), i64 9, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* getelementptr inbounds ([80 x i8], [80 x i8]* @temp, i32 0, i32 0), i8* getelementptr inbounds ([80 x i8], [80 x i8]* @message, i32 0, i32 0), i64 9, i32 4, i1 false)
   ret void
 }
 
@@ -111,7 +111,7 @@ define void @t7() {
 ; ARM64: ldrb w10, [x9, #6]
 ; ARM64: strb w10, [x8, #6]
 ; ARM64: ret
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 2 getelementptr inbounds ([80 x i8], [80 x i8]* @temp, i32 0, i32 0), i8* align 2 getelementptr inbounds ([80 x i8], [80 x i8]* @message, i32 0, i32 0), i64 7, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* getelementptr inbounds ([80 x i8], [80 x i8]* @temp, i32 0, i32 0), i8* getelementptr inbounds ([80 x i8], [80 x i8]* @message, i32 0, i32 0), i64 7, i32 2, i1 false)
   ret void
 }
 
@@ -130,7 +130,7 @@ define void @t8() {
 ; ARM64: ldrb w10, [x9, #3]
 ; ARM64: strb w10, [x8, #3]
 ; ARM64: ret
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 1 getelementptr inbounds ([80 x i8], [80 x i8]* @temp, i32 0, i32 0), i8* align 1 getelementptr inbounds ([80 x i8], [80 x i8]* @message, i32 0, i32 0), i64 4, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* getelementptr inbounds ([80 x i8], [80 x i8]* @temp, i32 0, i32 0), i8* getelementptr inbounds ([80 x i8], [80 x i8]* @message, i32 0, i32 0), i64 4, i32 1, i1 false)
   ret void
 }
 
@@ -143,6 +143,6 @@ define void @test_distant_memcpy(i8* %dst) {
 ; ARM64: strb [[BYTE]], [x0]
   %array = alloca i8, i32 8192
   %elem = getelementptr i8, i8* %array, i32 8000
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %dst, i8* %elem, i64 1, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %dst, i8* %elem, i64 1, i32 1, i1 false)
   ret void
 }
diff --git a/test/CodeGen/AArch64/arm64-memcpy-inline.ll b/test/CodeGen/AArch64/arm64-memcpy-inline.ll
index 3d2d9c3bc47..2f91ce25233 100644
--- a/test/CodeGen/AArch64/arm64-memcpy-inline.ll
+++ b/test/CodeGen/AArch64/arm64-memcpy-inline.ll
@@ -22,7 +22,7 @@ entry:
 ; CHECK: strh [[REG1]], [x[[BASEREG2]], #8]
 ; CHECK: ldr [[REG2:x[0-9]+]],
 ; CHECK: str [[REG2]],
-  call void @llvm.memcpy.p0i8.p0i8.i32(i8* getelementptr inbounds (%struct.x, %struct.x* @dst, i32 0, i32 0), i8* getelementptr inbounds (%struct.x, %struct.x* @src, i32 0, i32 0), i32 11, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* getelementptr inbounds (%struct.x, %struct.x* @dst, i32 0, i32 0), i8* getelementptr inbounds (%struct.x, %struct.x* @src, i32 0, i32 0), i32 11, i32 8, i1 false)
   ret i32 0
 }
 
@@ -33,7 +33,7 @@ entry:
 ; CHECK: stur [[DEST]], [x0, #15]
 ; CHECK: ldr [[DEST:q[0-9]+]], [x[[BASEREG]]]
 ; CHECK: str [[DEST]], [x0]
-  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %C, i8* getelementptr inbounds ([31 x i8], [31 x i8]* @.str1, i64 0, i64 0), i64 31, i1 false)
+  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %C, i8* getelementptr inbounds ([31 x i8], [31 x i8]* @.str1, i64 0, i64 0), i64 31, i32 1, i1 false)
   ret void
 }
 
@@ -45,7 +45,7 @@ entry:
 ; CHECK: str [[REG3]], [x0, #32]
 ; CHECK: ldp [[DEST1:q[0-9]+]], [[DEST2:q[0-9]+]], [x{{[0-9]+}}]
 ; CHECK: stp [[DEST1]], [[DEST2]], [x0]
-  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %C, i8* getelementptr inbounds ([36 x i8], [36 x i8]* @.str2, i64 0, i64 0), i64 36, i1 false)
+  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %C, i8* getelementptr inbounds ([36 x i8], [36 x i8]* @.str2, i64 0, i64 0), i64 36, i32 1, i1 false)
   ret void
 }
 
@@ -56,7 +56,7 @@ entry:
 ; CHECK: str [[REG4]], [x0, #16]
 ; CHECK: ldr [[DEST:q[0-9]+]], [x[[BASEREG]]]
 ; CHECK: str [[DEST]], [x0]
-  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %C, i8* getelementptr inbounds ([24 x i8], [24 x i8]* @.str3, i64 0, i64 0), i64 24, i1 false)
+  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %C, i8* getelementptr inbounds ([24 x i8], [24 x i8]* @.str3, i64 0, i64 0), i64 24, i32 1, i1 false)
   ret void
 }
 
@@ -67,7 +67,7 @@ entry:
 ; CHECK: strh [[REG5]], [x0, #16]
 ; CHECK: ldr [[REG6:q[0-9]+]], [x{{[0-9]+}}]
 ; CHECK: str [[REG6]], [x0]
-  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %C, i8* getelementptr inbounds ([18 x i8], [18 x i8]* @.str4, i64 0, i64 0), i64 18, i1 false)
+  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %C, i8* getelementptr inbounds ([18 x i8], [18 x i8]* @.str4, i64 0, i64 0), i64 18, i32 1, i1 false)
   ret void
 }
 
@@ -80,7 +80,7 @@ entry:
 ; CHECK: movz [[REG8:w[0-9]+]],
 ; CHECK: movk [[REG8]],
 ; CHECK: str [[REG8]], [x0]
-  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %C, i8* getelementptr inbounds ([7 x i8], [7 x i8]* @.str5, i64 0, i64 0), i64 7, i1 false)
+  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %C, i8* getelementptr inbounds ([7 x i8], [7 x i8]* @.str5, i64 0, i64 0), i64 7, i32 1, i1 false)
   ret void
 }
 
@@ -91,7 +91,7 @@ entry:
 ; CHECK: stur [[REG9]], [x{{[0-9]+}}, #6]
 ; CHECK: ldr
 ; CHECK: str
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* getelementptr inbounds ([512 x i8], [512 x i8]* @spool.splbuf, i64 0, i64 0), i8* getelementptr inbounds ([14 x i8], [14 x i8]* @.str6, i64 0, i64 0), i64 14, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* getelementptr inbounds ([512 x i8], [512 x i8]* @spool.splbuf, i64 0, i64 0), i8* getelementptr inbounds ([14 x i8], [14 x i8]* @.str6, i64 0, i64 0), i64 14, i32 1, i1 false)
   ret void
 }
 
@@ -104,9 +104,9 @@ entry:
 ; CHECK: str [[REG10]], [x0]
   %0 = bitcast %struct.Foo* %a to i8*
   %1 = bitcast %struct.Foo* %b to i8*
-  tail call void @llvm.memcpy.p0i8.p0i8.i32(i8* %0, i8* %1, i32 16, i1 false)
+  tail call void @llvm.memcpy.p0i8.p0i8.i32(i8* %0, i8* %1, i32 16, i32 4, i1 false)
   ret void
 }
 
-declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i1) nounwind
-declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i32, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i32, i1) nounwind
diff --git a/test/CodeGen/AArch64/arm64-memset-inline.ll b/test/CodeGen/AArch64/arm64-memset-inline.ll
index dd5b51d7aeb..56959ade043 100644
--- a/test/CodeGen/AArch64/arm64-memset-inline.ll
+++ b/test/CodeGen/AArch64/arm64-memset-inline.ll
@@ -5,7 +5,7 @@ entry:
 ; CHECK-LABEL: t1:
 ; CHECK: str wzr, [x0, #8]
 ; CHECK: str xzr, [x0]
-  call void @llvm.memset.p0i8.i64(i8* %c, i8 0, i64 12, i1 false)
+  call void @llvm.memset.p0i8.i64(i8* %c, i8 0, i64 12, i32 8, i1 false)
   ret void
 }
 
@@ -17,11 +17,11 @@ entry:
 ; CHECK: str xzr, [sp, #8]
   %buf = alloca [26 x i8], align 1
   %0 = getelementptr inbounds [26 x i8], [26 x i8]* %buf, i32 0, i32 0
-  call void @llvm.memset.p0i8.i32(i8* %0, i8 0, i32 26, i1 false)
+  call void @llvm.memset.p0i8.i32(i8* %0, i8 0, i32 26, i32 1, i1 false)
   call void @something(i8* %0) nounwind
   ret void
 }
 
 declare void @something(i8*) nounwind
-declare void @llvm.memset.p0i8.i32(i8* nocapture, i8, i32, i1) nounwind
-declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i1) nounwind
+declare void @llvm.memset.p0i8.i32(i8* nocapture, i8, i32, i32, i1) nounwind
+declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i32, i1) nounwind
diff --git a/test/CodeGen/AArch64/arm64-memset-to-bzero.ll b/test/CodeGen/AArch64/arm64-memset-to-bzero.ll
index fa1d6c8b774..29036caabf3 100644
--- a/test/CodeGen/AArch64/arm64-memset-to-bzero.ll
+++ b/test/CodeGen/AArch64/arm64-memset-to-bzero.ll
@@ -9,11 +9,11 @@
 ; CHECK: memset
 define void @fct1(i8* nocapture %ptr) {
 entry:
-  tail call void @llvm.memset.p0i8.i64(i8* %ptr, i8 0, i64 256, i1 false)
+  tail call void @llvm.memset.p0i8.i64(i8* %ptr, i8 0, i64 256, i32 1, i1 false)
   ret void
 }
 
-declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i1)
+declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i32, i1)
 
 ; CHECK: @fct2
 ; When the size is bigger than 256, change into bzero.
@@ -21,7 +21,7 @@ declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i1)
 ; CHECK-LINUX: memset
 define void @fct2(i8* nocapture %ptr) {
 entry:
-  tail call void @llvm.memset.p0i8.i64(i8* %ptr, i8 0, i64 257, i1 false)
+  tail call void @llvm.memset.p0i8.i64(i8* %ptr, i8 0, i64 257, i32 1, i1 false)
   ret void
 }
 
@@ -32,7 +32,7 @@ entry:
 define void @fct3(i8* nocapture %ptr, i32 %unknown) {
 entry:
   %conv = sext i32 %unknown to i64
-  tail call void @llvm.memset.p0i8.i64(i8* %ptr, i8 0, i64 %conv, i1 false)
+  tail call void @llvm.memset.p0i8.i64(i8* %ptr, i8 0, i64 %conv, i32 1, i1 false)
   ret void
 }
 
diff --git a/test/CodeGen/AArch64/arm64-misaligned-memcpy-inline.ll b/test/CodeGen/AArch64/arm64-misaligned-memcpy-inline.ll
index 7ecf214b4be..85572f2cf0f 100644
--- a/test/CodeGen/AArch64/arm64-misaligned-memcpy-inline.ll
+++ b/test/CodeGen/AArch64/arm64-misaligned-memcpy-inline.ll
@@ -7,8 +7,8 @@ define void @t0(i8* %out, i8* %in) {
 ; CHECK:         orr w2, wzr, #0x10
 ; CHECK-NEXT:    bl _memcpy
 entry:
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %out, i8* %in, i64 16, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %out, i8* %in, i64 16, i32 1, i1 false)
   ret void
 }
 
-declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture readonly, i64, i1)
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture readonly, i64, i32, i1)
diff --git a/test/CodeGen/AArch64/arm64-misched-basic-A53.ll b/test/CodeGen/AArch64/arm64-misched-basic-A53.ll
index 184441d9d05..6db21043f67 100644
--- a/test/CodeGen/AArch64/arm64-misched-basic-A53.ll
+++ b/test/CodeGen/AArch64/arm64-misched-basic-A53.ll
@@ -25,9 +25,9 @@ entry:
   %yy = alloca i32, align 4
   store i32 0, i32* %retval
   %0 = bitcast [8 x i32]* %x to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %0, i8* bitcast ([8 x i32]* @main.x to i8*), i64 32, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %0, i8* bitcast ([8 x i32]* @main.x to i8*), i64 32, i32 4, i1 false)
   %1 = bitcast [8 x i32]* %y to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %1, i8* bitcast ([8 x i32]* @main.y to i8*), i64 32, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %1, i8* bitcast ([8 x i32]* @main.y to i8*), i64 32, i32 4, i1 false)
   store i32 0, i32* %xx, align 4
   store i32 0, i32* %yy, align 4
   store i32 0, i32* %i, align 4
@@ -104,7 +104,7 @@ define <4 x float> @neon4xfloat(<4 x float> %A, <4 x float> %B) {
 }
 
 ; Function Attrs: nounwind
-declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture readonly, i64, i1) #1
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture readonly, i64, i32, i1) #1
 
 attributes #0 = { nounwind "less-precise-fpmad"="false" "no-frame-pointer-elim"="true" "no-frame-pointer-elim-non-leaf" "no-infs-fp-math"="false" "no-nans-fp-math"="false" "stack-protector-buffer-size"="8" "unsafe-fp-math"="false" "use-soft-float"="false" }
 attributes #1 = { nounwind }
diff --git a/test/CodeGen/AArch64/arm64-misched-basic-A57.ll b/test/CodeGen/AArch64/arm64-misched-basic-A57.ll
index 231c2cff154..fac5f8ad2e9 100644
--- a/test/CodeGen/AArch64/arm64-misched-basic-A57.ll
+++ b/test/CodeGen/AArch64/arm64-misched-basic-A57.ll
@@ -32,9 +32,9 @@ entry:
   %yy = alloca i32, align 4
   store i32 0, i32* %retval
   %0 = bitcast [8 x i32]* %x to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %0, i8* bitcast ([8 x i32]* @main.x to i8*), i64 32, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %0, i8* bitcast ([8 x i32]* @main.x to i8*), i64 32, i32 4, i1 false)
   %1 = bitcast [8 x i32]* %y to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %1, i8* bitcast ([8 x i32]* @main.y to i8*), i64 32, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %1, i8* bitcast ([8 x i32]* @main.y to i8*), i64 32, i32 4, i1 false)
   store i32 0, i32* %xx, align 4
   store i32 0, i32* %yy, align 4
   store i32 0, i32* %i, align 4
@@ -106,7 +106,7 @@ for.end:                                          ; preds = %for.cond
 
 
 ; Function Attrs: nounwind
-declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture readonly, i64, i1) #1
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture readonly, i64, i32, i1) #1
 
 attributes #0 = { nounwind "less-precise-fpmad"="false" "no-frame-pointer-elim"="true" "no-frame-pointer-elim-non-leaf" "no-infs-fp-math"="false" "no-nans-fp-math"="false" "stack-protector-buffer-size"="8" "unsafe-fp-math"="false" "use-soft-float"="false" }
 attributes #1 = { nounwind }
diff --git a/test/CodeGen/AArch64/arm64-stur.ll b/test/CodeGen/AArch64/arm64-stur.ll
index f8675cb6cda..5f4cb9f3d95 100644
--- a/test/CodeGen/AArch64/arm64-stur.ll
+++ b/test/CodeGen/AArch64/arm64-stur.ll
@@ -55,11 +55,11 @@ define void @foo(%struct.X* nocapture %p) nounwind optsize ssp {
 ; CHECK-NEXT: ret
   %B = getelementptr inbounds %struct.X, %struct.X* %p, i64 0, i32 1
   %val = bitcast i64* %B to i8*
-  call void @llvm.memset.p0i8.i64(i8* %val, i8 0, i64 16, i1 false)
+  call void @llvm.memset.p0i8.i64(i8* %val, i8 0, i64 16, i32 1, i1 false)
   ret void
 }
 
-declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i1) nounwind
+declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i32, i1) nounwind
 
 ; Unaligned 16b stores are split into 8b stores for performance.
 ; radar://15424193
diff --git a/test/CodeGen/AArch64/arm64-virtual_base.ll b/test/CodeGen/AArch64/arm64-virtual_base.ll
index be864b7cc26..703d81a8d4f 100644
--- a/test/CodeGen/AArch64/arm64-virtual_base.ll
+++ b/test/CodeGen/AArch64/arm64-virtual_base.ll
@@ -43,9 +43,9 @@ entry:
   %tmp14 = bitcast double* %arraydecay5.3.1 to i8*
   %arraydecay11.3.1 = getelementptr inbounds %struct.Bicubic_Patch_Struct, %struct.Bicubic_Patch_Struct* %Shape, i64 0, i32 12, i64 1, i64 3, i64 0
   %tmp15 = bitcast double* %arraydecay11.3.1 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %tmp14, i8* %tmp15, i64 24, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %tmp14, i8* %tmp15, i64 24, i32 1, i1 false)
   ret void
 }
 
 ; Function Attrs: nounwind
-declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i1)
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i32, i1)
diff --git a/test/CodeGen/AArch64/fast-isel-memcpy.ll b/test/CodeGen/AArch64/fast-isel-memcpy.ll
index f8e41cb7f3b..07595a954db 100644
--- a/test/CodeGen/AArch64/fast-isel-memcpy.ll
+++ b/test/CodeGen/AArch64/fast-isel-memcpy.ll
@@ -8,8 +8,8 @@
 define void @test(i64 %a, i8* %b) {
   %1 = and i64 %a, 9223372036854775807
   %2 = inttoptr i64 %1 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %2, i8* %b, i64 8, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %2, i8* %b, i64 8, i32 8, i1 false)
   ret void
 }
 
-declare void @llvm.memcpy.p0i8.p0i8.i64(i8*, i8*, i64, i1)
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8*, i8*, i64, i32, i1)
diff --git a/test/CodeGen/AArch64/func-argpassing.ll b/test/CodeGen/AArch64/func-argpassing.ll
index f4930130b82..9100ae39282 100644
--- a/test/CodeGen/AArch64/func-argpassing.ll
+++ b/test/CodeGen/AArch64/func-argpassing.ll
@@ -186,11 +186,11 @@ define void @check_i128_stackalign(i32 %val0, i32 %val1, i32 %val2, i32 %val3,
     ret void
 }
 
-declare void @llvm.memcpy.p0i8.p0i8.i32(i8*, i8*, i32, i1)
+declare void @llvm.memcpy.p0i8.p0i8.i32(i8*, i8*, i32, i32, i1)
 
 define i32 @test_extern() {
 ; CHECK-LABEL: test_extern:
-  call void @llvm.memcpy.p0i8.p0i8.i32(i8* undef, i8* undef, i32 undef, i1 0)
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* undef, i8* undef, i32 undef, i32 4, i1 0)
 ; CHECK: bl memcpy
   ret i32 0
 }
diff --git a/test/CodeGen/AArch64/memcpy-f128.ll b/test/CodeGen/AArch64/memcpy-f128.ll
index f61593a1bd2..76db2974ab4 100644
--- a/test/CodeGen/AArch64/memcpy-f128.ll
+++ b/test/CodeGen/AArch64/memcpy-f128.ll
@@ -12,8 +12,8 @@ define void @test1() {
 ; CHECK: str q0
 ; CHECK: ret
 entry:
-  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* undef, i8* bitcast (%structA* @stubA to i8*), i64 48, i1 false)
+  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* undef, i8* bitcast (%structA* @stubA to i8*), i64 48, i32 8, i1 false)
   ret void
 }
 
-declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture readonly, i64, i1)
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture readonly, i64, i32, i1)
diff --git a/test/CodeGen/AArch64/tailcall-mem-intrinsics.ll b/test/CodeGen/AArch64/tailcall-mem-intrinsics.ll
index c780d15b58d..b970fb12415 100644
--- a/test/CodeGen/AArch64/tailcall-mem-intrinsics.ll
+++ b/test/CodeGen/AArch64/tailcall-mem-intrinsics.ll
@@ -4,7 +4,7 @@
 ; CHECK: b memcpy
 define void @tail_memcpy(i8* nocapture %p, i8* nocapture readonly %q, i32 %n) #0 {
 entry:
-  tail call void @llvm.memcpy.p0i8.p0i8.i32(i8* %p, i8* %q, i32 %n, i1 false)
+  tail call void @llvm.memcpy.p0i8.p0i8.i32(i8* %p, i8* %q, i32 %n, i32 1, i1 false)
   ret void
 }
 
@@ -12,7 +12,7 @@ entry:
 ; CHECK: b memmove
 define void @tail_memmove(i8* nocapture %p, i8* nocapture readonly %q, i32 %n) #0 {
 entry:
-  tail call void @llvm.memmove.p0i8.p0i8.i32(i8* %p, i8* %q, i32 %n, i1 false)
+  tail call void @llvm.memmove.p0i8.p0i8.i32(i8* %p, i8* %q, i32 %n, i32 1, i1 false)
   ret void
 }
 
@@ -20,12 +20,12 @@ entry:
 ; CHECK: b memset
 define void @tail_memset(i8* nocapture %p, i8 %c, i32 %n) #0 {
 entry:
-  tail call void @llvm.memset.p0i8.i32(i8* %p, i8 %c, i32 %n, i1 false)
+  tail call void @llvm.memset.p0i8.i32(i8* %p, i8 %c, i32 %n, i32 1, i1 false)
   ret void
 }
 
-declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture readonly, i32, i1) #0
-declare void @llvm.memmove.p0i8.p0i8.i32(i8* nocapture, i8* nocapture readonly, i32, i1) #0
-declare void @llvm.memset.p0i8.i32(i8* nocapture, i8, i32, i1) #0
+declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture readonly, i32, i32, i1) #0
+declare void @llvm.memmove.p0i8.p0i8.i32(i8* nocapture, i8* nocapture readonly, i32, i32, i1) #0
+declare void @llvm.memset.p0i8.i32(i8* nocapture, i8, i32, i32, i1) #0
 
 attributes #0 = { nounwind }
diff --git a/test/CodeGen/AMDGPU/llvm.memcpy.ll b/test/CodeGen/AMDGPU/llvm.memcpy.ll
index 55b447ba07b..e491732cf9c 100644
--- a/test/CodeGen/AMDGPU/llvm.memcpy.ll
+++ b/test/CodeGen/AMDGPU/llvm.memcpy.ll
@@ -1,8 +1,8 @@
 ; RUN: llc -march=amdgcn -mcpu=SI -verify-machineinstrs < %s | FileCheck -check-prefix=SI -check-prefix=FUNC %s
 ; RUN: llc -march=amdgcn -mcpu=tonga -verify-machineinstrs < %s | FileCheck -check-prefix=SI -check-prefix=FUNC %s
 
-declare void @llvm.memcpy.p3i8.p3i8.i32(i8 addrspace(3)* nocapture, i8 addrspace(3)* nocapture, i32, i1) nounwind
-declare void @llvm.memcpy.p1i8.p1i8.i64(i8 addrspace(1)* nocapture, i8 addrspace(1)* nocapture, i64, i1) nounwind
+declare void @llvm.memcpy.p3i8.p3i8.i32(i8 addrspace(3)* nocapture, i8 addrspace(3)* nocapture, i32, i32, i1) nounwind
+declare void @llvm.memcpy.p1i8.p1i8.i64(i8 addrspace(1)* nocapture, i8 addrspace(1)* nocapture, i64, i32, i1) nounwind
 
 
 ; FUNC-LABEL: {{^}}test_small_memcpy_i64_lds_to_lds_align1:
@@ -82,7 +82,7 @@ declare void @llvm.memcpy.p1i8.p1i8.i64(i8 addrspace(1)* nocapture, i8 addrspace
 define void @test_small_memcpy_i64_lds_to_lds_align1(i64 addrspace(3)* noalias %out, i64 addrspace(3)* noalias %in) nounwind {
   %bcin = bitcast i64 addrspace(3)* %in to i8 addrspace(3)*
   %bcout = bitcast i64 addrspace(3)* %out to i8 addrspace(3)*
-  call void @llvm.memcpy.p3i8.p3i8.i32(i8 addrspace(3)* align 1 %bcout, i8 addrspace(3)* align 1 %bcin, i32 32, i1 false) nounwind
+  call void @llvm.memcpy.p3i8.p3i8.i32(i8 addrspace(3)* %bcout, i8 addrspace(3)* %bcin, i32 32, i32 1, i1 false) nounwind
   ret void
 }
 
@@ -127,7 +127,7 @@ define void @test_small_memcpy_i64_lds_to_lds_align1(i64 addrspace(3)* noalias %
 define void @test_small_memcpy_i64_lds_to_lds_align2(i64 addrspace(3)* noalias %out, i64 addrspace(3)* noalias %in) nounwind {
   %bcin = bitcast i64 addrspace(3)* %in to i8 addrspace(3)*
   %bcout = bitcast i64 addrspace(3)* %out to i8 addrspace(3)*
-  call void @llvm.memcpy.p3i8.p3i8.i32(i8 addrspace(3)* align 2 %bcout, i8 addrspace(3)* align 2 %bcin, i32 32, i1 false) nounwind
+  call void @llvm.memcpy.p3i8.p3i8.i32(i8 addrspace(3)* %bcout, i8 addrspace(3)* %bcin, i32 32, i32 2, i1 false) nounwind
   ret void
 }
 
@@ -163,7 +163,7 @@ define void @test_small_memcpy_i64_lds_to_lds_align2(i64 addrspace(3)* noalias %
 define void @test_small_memcpy_i64_lds_to_lds_align4(i64 addrspace(3)* noalias %out, i64 addrspace(3)* noalias %in) nounwind {
   %bcin = bitcast i64 addrspace(3)* %in to i8 addrspace(3)*
   %bcout = bitcast i64 addrspace(3)* %out to i8 addrspace(3)*
-  call void @llvm.memcpy.p3i8.p3i8.i32(i8 addrspace(3)* align 4 %bcout, i8 addrspace(3)* align 4 %bcin, i32 32, i1 false) nounwind
+  call void @llvm.memcpy.p3i8.p3i8.i32(i8 addrspace(3)* %bcout, i8 addrspace(3)* %bcin, i32 32, i32 4, i1 false) nounwind
   ret void
 }
 
@@ -201,7 +201,7 @@ define void @test_small_memcpy_i64_lds_to_lds_align4(i64 addrspace(3)* noalias %
 define void @test_small_memcpy_i64_lds_to_lds_align8(i64 addrspace(3)* noalias %out, i64 addrspace(3)* noalias %in) nounwind {
   %bcin = bitcast i64 addrspace(3)* %in to i8 addrspace(3)*
   %bcout = bitcast i64 addrspace(3)* %out to i8 addrspace(3)*
-  call void @llvm.memcpy.p3i8.p3i8.i32(i8 addrspace(3)* align 8 %bcout, i8 addrspace(3)* align 8 %bcin, i32 32, i1 false) nounwind
+  call void @llvm.memcpy.p3i8.p3i8.i32(i8 addrspace(3)* %bcout, i8 addrspace(3)* %bcin, i32 32, i32 8, i1 false) nounwind
   ret void
 }
 
@@ -278,7 +278,7 @@ define void @test_small_memcpy_i64_lds_to_lds_align8(i64 addrspace(3)* noalias %
 define void @test_small_memcpy_i64_global_to_global_align1(i64 addrspace(1)* noalias %out, i64 addrspace(1)* noalias %in) nounwind {
   %bcin = bitcast i64 addrspace(1)* %in to i8 addrspace(1)*
   %bcout = bitcast i64 addrspace(1)* %out to i8 addrspace(1)*
-  call void @llvm.memcpy.p1i8.p1i8.i64(i8 addrspace(1)* align 1 %bcout, i8 addrspace(1)* align 1 %bcin, i64 32, i1 false) nounwind
+  call void @llvm.memcpy.p1i8.p1i8.i64(i8 addrspace(1)* %bcout, i8 addrspace(1)* %bcin, i64 32, i32 1, i1 false) nounwind
   ret void
 }
 
@@ -321,7 +321,7 @@ define void @test_small_memcpy_i64_global_to_global_align1(i64 addrspace(1)* noa
 define void @test_small_memcpy_i64_global_to_global_align2(i64 addrspace(1)* noalias %out, i64 addrspace(1)* noalias %in) nounwind {
   %bcin = bitcast i64 addrspace(1)* %in to i8 addrspace(1)*
   %bcout = bitcast i64 addrspace(1)* %out to i8 addrspace(1)*
-  call void @llvm.memcpy.p1i8.p1i8.i64(i8 addrspace(1)* align 2 %bcout, i8 addrspace(1)* align 2 %bcin, i64 32, i1 false) nounwind
+  call void @llvm.memcpy.p1i8.p1i8.i64(i8 addrspace(1)* %bcout, i8 addrspace(1)* %bcin, i64 32, i32 2, i1 false) nounwind
   ret void
 }
 
@@ -334,7 +334,7 @@ define void @test_small_memcpy_i64_global_to_global_align2(i64 addrspace(1)* noa
 define void @test_small_memcpy_i64_global_to_global_align4(i64 addrspace(1)* noalias %out, i64 addrspace(1)* noalias %in) nounwind {
   %bcin = bitcast i64 addrspace(1)* %in to i8 addrspace(1)*
   %bcout = bitcast i64 addrspace(1)* %out to i8 addrspace(1)*
-  call void @llvm.memcpy.p1i8.p1i8.i64(i8 addrspace(1)* align 4 %bcout, i8 addrspace(1)* align 4 %bcin, i64 32, i1 false) nounwind
+  call void @llvm.memcpy.p1i8.p1i8.i64(i8 addrspace(1)* %bcout, i8 addrspace(1)* %bcin, i64 32, i32 4, i1 false) nounwind
   ret void
 }
 
@@ -347,7 +347,7 @@ define void @test_small_memcpy_i64_global_to_global_align4(i64 addrspace(1)* noa
 define void @test_small_memcpy_i64_global_to_global_align8(i64 addrspace(1)* noalias %out, i64 addrspace(1)* noalias %in) nounwind {
   %bcin = bitcast i64 addrspace(1)* %in to i8 addrspace(1)*
   %bcout = bitcast i64 addrspace(1)* %out to i8 addrspace(1)*
-  call void @llvm.memcpy.p1i8.p1i8.i64(i8 addrspace(1)* align 8 %bcout, i8 addrspace(1)* align 8 %bcin, i64 32, i1 false) nounwind
+  call void @llvm.memcpy.p1i8.p1i8.i64(i8 addrspace(1)* %bcout, i8 addrspace(1)* %bcin, i64 32, i32 8, i1 false) nounwind
   ret void
 }
 
@@ -360,6 +360,6 @@ define void @test_small_memcpy_i64_global_to_global_align8(i64 addrspace(1)* noa
 define void @test_small_memcpy_i64_global_to_global_align16(i64 addrspace(1)* noalias %out, i64 addrspace(1)* noalias %in) nounwind {
   %bcin = bitcast i64 addrspace(1)* %in to i8 addrspace(1)*
   %bcout = bitcast i64 addrspace(1)* %out to i8 addrspace(1)*
-  call void @llvm.memcpy.p1i8.p1i8.i64(i8 addrspace(1)* align 16 %bcout, i8 addrspace(1)* align 16 %bcin, i64 32, i1 false) nounwind
+  call void @llvm.memcpy.p1i8.p1i8.i64(i8 addrspace(1)* %bcout, i8 addrspace(1)* %bcin, i64 32, i32 16, i1 false) nounwind
   ret void
 }
diff --git a/test/CodeGen/ARM/2009-03-07-SpillerBug.ll b/test/CodeGen/ARM/2009-03-07-SpillerBug.ll
index 0f2f3fecb6f..567400318ee 100644
--- a/test/CodeGen/ARM/2009-03-07-SpillerBug.ll
+++ b/test/CodeGen/ARM/2009-03-07-SpillerBug.ll
@@ -59,7 +59,7 @@ bb3:                                              ; preds = %entry
   %34 = fadd double %31, 0.000000e+00
   %35 = fadd double %32, 0.000000e+00
   %36 = bitcast %struct.ggPoint3* %x to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i32(i8* null, i8* %36, i32 24, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* null, i8* %36, i32 24, i32 4, i1 false)
   store double %33, double* null, align 8
   br i1 false, label %_Z20ggRaySphereIntersectRK6ggRay3RK8ggSphereddRd.exit, label %bb5.i.i.i
 
@@ -76,4 +76,4 @@ bb7:                                              ; preds = %entry
   ret i32 0
 }
 
-declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i32, i1) nounwind
diff --git a/test/CodeGen/ARM/2011-03-10-DAGCombineCrash.ll b/test/CodeGen/ARM/2011-03-10-DAGCombineCrash.ll
index 73497b2112b..c447a1f25b6 100644
--- a/test/CodeGen/ARM/2011-03-10-DAGCombineCrash.ll
+++ b/test/CodeGen/ARM/2011-03-10-DAGCombineCrash.ll
@@ -16,7 +16,7 @@ bb:                                               ; preds = %entry
 
 bb1:                                              ; preds = %entry
   %0 = call %struct.ui* @vn_pp_to_ui(i32* undef) nounwind
-  call void @llvm.memset.p0i8.i32(i8* undef, i8 0, i32 40, i1 false)
+  call void @llvm.memset.p0i8.i32(i8* undef, i8 0, i32 40, i32 4, i1 false)
   %1 = getelementptr inbounds %struct.ui, %struct.ui* %0, i32 0, i32 0
   store %struct.mo* undef, %struct.mo** %1, align 4
   %2 = getelementptr inbounds %struct.ui, %struct.ui* %0, i32 0, i32 5
@@ -40,7 +40,7 @@ bb6:                                              ; preds = %bb3
 
 declare %struct.ui* @vn_pp_to_ui(i32*)
 
-declare void @llvm.memset.p0i8.i32(i8* nocapture, i8, i32, i1) nounwind
+declare void @llvm.memset.p0i8.i32(i8* nocapture, i8, i32, i32, i1) nounwind
 
 declare i32 @mo_create_nnm(%struct.mo*, i64, i32**)
 
diff --git a/test/CodeGen/ARM/2011-10-26-memset-inline.ll b/test/CodeGen/ARM/2011-10-26-memset-inline.ll
index d2d3116cbd0..5df439389cd 100644
--- a/test/CodeGen/ARM/2011-10-26-memset-inline.ll
+++ b/test/CodeGen/ARM/2011-10-26-memset-inline.ll
@@ -14,8 +14,8 @@ target triple = "thumbv7-apple-ios5.0.0"
 ; CHECK-UNALIGNED:    str
 define void @foo(i8* nocapture %c) nounwind optsize {
 entry:
-  call void @llvm.memset.p0i8.i64(i8* %c, i8 -1, i64 5, i1 false)
+  call void @llvm.memset.p0i8.i64(i8* %c, i8 -1, i64 5, i32 1, i1 false)
   ret void
 }
 
-declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i1) nounwind
+declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i32, i1) nounwind
diff --git a/test/CodeGen/ARM/2011-10-26-memset-with-neon.ll b/test/CodeGen/ARM/2011-10-26-memset-with-neon.ll
index 7024a653b6c..c8e08c22ab1 100644
--- a/test/CodeGen/ARM/2011-10-26-memset-with-neon.ll
+++ b/test/CodeGen/ARM/2011-10-26-memset-with-neon.ll
@@ -5,8 +5,8 @@
 ; CHECK: vst1.64
 define void @f_0_40(i8* nocapture %c) nounwind optsize {
 entry:
-  call void @llvm.memset.p0i8.i64(i8* align 16 %c, i8 0, i64 40, i1 false)
+  call void @llvm.memset.p0i8.i64(i8* %c, i8 0, i64 40, i32 16, i1 false)
   ret void
 }
 
-declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i1) nounwind
+declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i32, i1) nounwind
diff --git a/test/CodeGen/ARM/2012-04-24-SplitEHCriticalEdge.ll b/test/CodeGen/ARM/2012-04-24-SplitEHCriticalEdge.ll
index ef33b2f5018..ce0dcc70952 100644
--- a/test/CodeGen/ARM/2012-04-24-SplitEHCriticalEdge.ll
+++ b/test/CodeGen/ARM/2012-04-24-SplitEHCriticalEdge.ll
@@ -19,7 +19,7 @@ declare i32 @llvm.eh.typeid.for(i8*) nounwind readnone
 
 declare i8* @__cxa_begin_catch(i8*)
 
-declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i32, i1) nounwind
 
 declare void @__cxa_end_catch()
 
diff --git a/test/CodeGen/ARM/Windows/memset.ll b/test/CodeGen/ARM/Windows/memset.ll
index c9b22f47a15..500e25e259c 100644
--- a/test/CodeGen/ARM/Windows/memset.ll
+++ b/test/CodeGen/ARM/Windows/memset.ll
@@ -2,11 +2,11 @@
 
 @source = common global [512 x i8] zeroinitializer, align 4
 
-declare void @llvm.memset.p0i8.i32(i8* nocapture, i8, i32, i1) nounwind
+declare void @llvm.memset.p0i8.i32(i8* nocapture, i8, i32, i32, i1) nounwind
 
 define void @function() {
 entry:
-  call void @llvm.memset.p0i8.i32(i8* bitcast ([512 x i8]* @source to i8*), i8 0, i32 512, i1 false)
+  call void @llvm.memset.p0i8.i32(i8* bitcast ([512 x i8]* @source to i8*), i8 0, i32 512, i32 0, i1 false)
   unreachable
 }
 
diff --git a/test/CodeGen/ARM/Windows/no-aeabi.ll b/test/CodeGen/ARM/Windows/no-aeabi.ll
index 35cc7d5032d..3971b9ccf58 100644
--- a/test/CodeGen/ARM/Windows/no-aeabi.ll
+++ b/test/CodeGen/ARM/Windows/no-aeabi.ll
@@ -1,14 +1,14 @@
 ; RUN: llc -mtriple=thumbv7-windows-itanium -mcpu=cortex-a9 -o - %s | FileCheck %s
 
-declare void @llvm.memmove.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i1) nounwind
-declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i1) nounwind
+declare void @llvm.memmove.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i32, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i32, i1) nounwind
 
 @source = common global [512 x i8] zeroinitializer, align 4
 @target = common global [512 x i8] zeroinitializer, align 4
 
 define void @move() nounwind {
 entry:
-  call void @llvm.memmove.p0i8.p0i8.i32(i8* bitcast ([512 x i8]* @target to i8*), i8* bitcast ([512 x i8]* @source to i8*), i32 512, i1 false)
+  call void @llvm.memmove.p0i8.p0i8.i32(i8* bitcast ([512 x i8]* @target to i8*), i8* bitcast ([512 x i8]* @source to i8*), i32 512, i32 0, i1 false)
   unreachable
 }
 
@@ -16,7 +16,7 @@ entry:
 
 define void @copy() nounwind {
 entry:
-  call void @llvm.memcpy.p0i8.p0i8.i32(i8* bitcast ([512 x i8]* @target to i8*), i8* bitcast ([512 x i8]* @source to i8*), i32 512, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* bitcast ([512 x i8]* @target to i8*), i8* bitcast ([512 x i8]* @source to i8*), i32 512, i32 0, i1 false)
   unreachable
 }
 
diff --git a/test/CodeGen/ARM/crash-O0.ll b/test/CodeGen/ARM/crash-O0.ll
index 8f3b5f10cd6..f92af999be5 100644
--- a/test/CodeGen/ARM/crash-O0.ll
+++ b/test/CodeGen/ARM/crash-O0.ll
@@ -12,7 +12,7 @@ entry:
 }
 
 @.str523 = private constant [256 x i8] c"<Unknown>\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00", align 4 ; <[256 x i8]*> [#uses=1]
-declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i32, i1) nounwind
 
 ; This function uses the scavenger for an ADDri instruction.
 ; ARMBaseRegisterInfo::estimateRSStackSizeLimit must return a 255 limit.
@@ -21,8 +21,8 @@ entry:
   %letter = alloca i8                             ; <i8*> [#uses=0]
   %prodvers = alloca [256 x i8]                   ; <[256 x i8]*> [#uses=1]
   %buildver = alloca [256 x i8]                   ; <[256 x i8]*> [#uses=0]
-  call void @llvm.memcpy.p0i8.p0i8.i32(i8* undef, i8* getelementptr inbounds ([256 x i8], [256 x i8]* @.str523, i32 0, i32 0), i32 256, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* undef, i8* getelementptr inbounds ([256 x i8], [256 x i8]* @.str523, i32 0, i32 0), i32 256, i32 1, i1 false)
   %prodvers2 = bitcast [256 x i8]* %prodvers to i8* ; <i8*> [#uses=1]
-  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %prodvers2, i8* getelementptr inbounds ([256 x i8], [256 x i8]* @.str523, i32 0, i32 0), i32 256, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %prodvers2, i8* getelementptr inbounds ([256 x i8], [256 x i8]* @.str523, i32 0, i32 0), i32 256, i32 1, i1 false)
   unreachable
 }
diff --git a/test/CodeGen/ARM/debug-info-blocks.ll b/test/CodeGen/ARM/debug-info-blocks.ll
index 07fb4ef9ba9..2a8898651f0 100644
--- a/test/CodeGen/ARM/debug-info-blocks.ll
+++ b/test/CodeGen/ARM/debug-info-blocks.ll
@@ -25,7 +25,7 @@ declare i8* @objc_msgSend(i8*, i8*, ...)
 
 declare void @llvm.dbg.value(metadata, i64, metadata, metadata) nounwind readnone
 
-declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i32, i1) nounwind
 
 define hidden void @foobar_func_block_invoke_0(i8* %.block_descriptor, %0* %loadedMydata, [4 x i32] %bounds.coerce0, [4 x i32] %data.coerce0) ssp !dbg !23 {
   %1 = alloca %0*, align 4
@@ -67,7 +67,7 @@ define hidden void @foobar_func_block_invoke_0(i8* %.block_descriptor, %0* %load
   %24 = bitcast i8* %23 to %struct.CR*, !dbg !143
   %25 = bitcast %struct.CR* %24 to i8*, !dbg !143
   %26 = bitcast %struct.CR* %data to i8*, !dbg !143
-  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %25, i8* %26, i32 16, i1 false), !dbg !143
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %25, i8* %26, i32 16, i32 4, i1 false), !dbg !143
   %27 = getelementptr inbounds %2, %2* %6, i32 0, i32 6, !dbg !144
   %28 = load %3*, %3** %27, align 4, !dbg !144
   %29 = load i32, i32* @"OBJC_IVAR_$_MyWork._bounds", !dbg !144
@@ -76,7 +76,7 @@ define hidden void @foobar_func_block_invoke_0(i8* %.block_descriptor, %0* %load
   %32 = bitcast i8* %31 to %struct.CR*, !dbg !144
   %33 = bitcast %struct.CR* %32 to i8*, !dbg !144
   %34 = bitcast %struct.CR* %bounds to i8*, !dbg !144
-  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %33, i8* %34, i32 16, i1 false), !dbg !144
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %33, i8* %34, i32 16, i32 4, i1 false), !dbg !144
   %35 = getelementptr inbounds %2, %2* %6, i32 0, i32 6, !dbg !145
   %36 = load %3*, %3** %35, align 4, !dbg !145
   %37 = getelementptr inbounds %2, %2* %6, i32 0, i32 5, !dbg !145
diff --git a/test/CodeGen/ARM/dyn-stackalloc.ll b/test/CodeGen/ARM/dyn-stackalloc.ll
index 06d8f8dc179..5b963fd64de 100644
--- a/test/CodeGen/ARM/dyn-stackalloc.ll
+++ b/test/CodeGen/ARM/dyn-stackalloc.ll
@@ -51,7 +51,7 @@ define void @t2(%struct.comment* %vc, i8* %tag, i8* %contents) {
   %tmp9 = call i8* @strcpy(i8* %tmp6, i8* %tag)
   %tmp6.len = call i32 @strlen(i8* %tmp6)
   %tmp6.indexed = getelementptr i8, i8* %tmp6, i32 %tmp6.len
-  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %tmp6.indexed, i8* getelementptr inbounds ([2 x i8], [2 x i8]* @str215, i32 0, i32 0), i32 2, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %tmp6.indexed, i8* getelementptr inbounds ([2 x i8], [2 x i8]* @str215, i32 0, i32 0), i32 2, i32 1, i1 false)
   %tmp15 = call i8* @strcat(i8* %tmp6, i8* %contents)
   call fastcc void @comment_add(%struct.comment* %vc, i8* %tmp6)
   ret void
@@ -65,4 +65,4 @@ declare fastcc void @comment_add(%struct.comment*, i8*)
 
 declare i8* @strcpy(i8*, i8*)
 
-declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i32, i1) nounwind
diff --git a/test/CodeGen/ARM/fast-isel-intrinsic.ll b/test/CodeGen/ARM/fast-isel-intrinsic.ll
index 131023e7ab6..1c7ff687938 100644
--- a/test/CodeGen/ARM/fast-isel-intrinsic.ll
+++ b/test/CodeGen/ARM/fast-isel-intrinsic.ll
@@ -39,11 +39,11 @@ define void @t1() nounwind ssp {
 ; THUMB-LONG: movt r3, :upper16:L_memset$non_lazy_ptr
 ; THUMB-LONG: ldr r3, [r3]
 ; THUMB-LONG: blx r3
-  call void @llvm.memset.p0i8.i32(i8* align 4 getelementptr inbounds ([60 x i8], [60 x i8]* @message1, i32 0, i32 5), i8 64, i32 10, i1 false)
+  call void @llvm.memset.p0i8.i32(i8* getelementptr inbounds ([60 x i8], [60 x i8]* @message1, i32 0, i32 5), i8 64, i32 10, i32 4, i1 false)
   ret void
 }
 
-declare void @llvm.memset.p0i8.i32(i8* nocapture, i8, i32, i1) nounwind
+declare void @llvm.memset.p0i8.i32(i8* nocapture, i8, i32, i32, i1) nounwind
 
 define void @t2() nounwind ssp {
 ; ARM-LABEL: t2:
@@ -78,11 +78,11 @@ define void @t2() nounwind ssp {
 ; THUMB-LONG: movt r3, :upper16:L_memcpy$non_lazy_ptr
 ; THUMB-LONG: ldr r3, [r3]
 ; THUMB-LONG: blx r3
-  call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 4 getelementptr inbounds ([60 x i8], [60 x i8]* @temp, i32 0, i32 4), i8* align 4 getelementptr inbounds ([60 x i8], [60 x i8]* @temp, i32 0, i32 16), i32 17, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* getelementptr inbounds ([60 x i8], [60 x i8]* @temp, i32 0, i32 4), i8* getelementptr inbounds ([60 x i8], [60 x i8]* @temp, i32 0, i32 16), i32 17, i32 4, i1 false)
   ret void
 }
 
-declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i32, i1) nounwind
 
 define void @t3() nounwind ssp {
 ; ARM-LABEL: t3:
@@ -115,7 +115,7 @@ define void @t3() nounwind ssp {
 ; THUMB-LONG: movt r3, :upper16:L_memmove$non_lazy_ptr
 ; THUMB-LONG: ldr r3, [r3]
 ; THUMB-LONG: blx r3
-  call void @llvm.memmove.p0i8.p0i8.i32(i8* getelementptr inbounds ([60 x i8], [60 x i8]* @temp, i32 0, i32 4), i8* getelementptr inbounds ([60 x i8], [60 x i8]* @temp, i32 0, i32 16), i32 10, i1 false)
+  call void @llvm.memmove.p0i8.p0i8.i32(i8* getelementptr inbounds ([60 x i8], [60 x i8]* @temp, i32 0, i32 4), i8* getelementptr inbounds ([60 x i8], [60 x i8]* @temp, i32 0, i32 16), i32 10, i32 1, i1 false)
   ret void
 }
 
@@ -142,11 +142,11 @@ define void @t4() nounwind ssp {
 ; THUMB: ldrh r1, [r0, #24]
 ; THUMB: strh r1, [r0, #12]
 ; THUMB: bx lr
-  call void @llvm.memcpy.p0i8.p0i8.i32(i8* getelementptr inbounds ([60 x i8], [60 x i8]* @temp, i32 0, i32 4), i8* getelementptr inbounds ([60 x i8], [60 x i8]* @temp, i32 0, i32 16), i32 10, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* getelementptr inbounds ([60 x i8], [60 x i8]* @temp, i32 0, i32 4), i8* getelementptr inbounds ([60 x i8], [60 x i8]* @temp, i32 0, i32 16), i32 10, i32 4, i1 false)
   ret void
 }
 
-declare void @llvm.memmove.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i1) nounwind
+declare void @llvm.memmove.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i32, i1) nounwind
 
 define void @t5() nounwind ssp {
 ; ARM-LABEL: t5:
@@ -179,7 +179,7 @@ define void @t5() nounwind ssp {
 ; THUMB: ldrh r1, [r0, #24]
 ; THUMB: strh r1, [r0, #12]
 ; THUMB: bx lr
-  call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 2 getelementptr inbounds ([60 x i8], [60 x i8]* @temp, i32 0, i32 4), i8* align 2 getelementptr inbounds ([60 x i8], [60 x i8]* @temp, i32 0, i32 16), i32 10, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* getelementptr inbounds ([60 x i8], [60 x i8]* @temp, i32 0, i32 4), i8* getelementptr inbounds ([60 x i8], [60 x i8]* @temp, i32 0, i32 16), i32 10, i32 2, i1 false)
   ret void
 }
 
@@ -234,14 +234,14 @@ define void @t6() nounwind ssp {
 ; THUMB: ldrb r1, [r0, #25]
 ; THUMB: strb r1, [r0, #13]
 ; THUMB: bx lr
-  call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 1 getelementptr inbounds ([60 x i8], [60 x i8]* @temp, i32 0, i32 4), i8* align 1 getelementptr inbounds ([60 x i8], [60 x i8]* @temp, i32 0, i32 16), i32 10, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* getelementptr inbounds ([60 x i8], [60 x i8]* @temp, i32 0, i32 4), i8* getelementptr inbounds ([60 x i8], [60 x i8]* @temp, i32 0, i32 16), i32 10, i32 1, i1 false)
   ret void
 }
 
 ; rdar://13202135
 define void @t7() nounwind ssp {
 ; Just make sure this doesn't assert when we have an odd length and an alignment of 2.
-  call void @llvm.memcpy.p0i8.p0i8.i32(i8* getelementptr inbounds ([60 x i8], [60 x i8]* @temp, i32 0, i32 4), i8* getelementptr inbounds ([60 x i8], [60 x i8]* @temp, i32 0, i32 16), i32 3, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* getelementptr inbounds ([60 x i8], [60 x i8]* @temp, i32 0, i32 4), i8* getelementptr inbounds ([60 x i8], [60 x i8]* @temp, i32 0, i32 16), i32 3, i32 2, i1 false)
   ret void
 }
 
diff --git a/test/CodeGen/ARM/machine-cse-cmp.ll b/test/CodeGen/ARM/machine-cse-cmp.ll
index 10e56a346a2..611cba6ed1f 100644
--- a/test/CodeGen/ARM/machine-cse-cmp.ll
+++ b/test/CodeGen/ARM/machine-cse-cmp.ll
@@ -37,14 +37,14 @@ entry:
 for.body.lr.ph:                                   ; preds = %entry
   %1 = icmp sgt i32 %0, 1
   %smax = select i1 %1, i32 %0, i32 1
-  call void @llvm.memset.p0i8.i32(i8* getelementptr inbounds ([250 x i8], [250 x i8]* @bar, i32 0, i32 0), i8 0, i32 %smax, i1 false)
+  call void @llvm.memset.p0i8.i32(i8* getelementptr inbounds ([250 x i8], [250 x i8]* @bar, i32 0, i32 0), i8 0, i32 %smax, i32 1, i1 false)
   unreachable
 
 for.cond1.preheader:                              ; preds = %entry
   ret void
 }
 
-declare void @llvm.memset.p0i8.i32(i8* nocapture, i8, i32, i1) nounwind
+declare void @llvm.memset.p0i8.i32(i8* nocapture, i8, i32, i32, i1) nounwind
 
 ; rdar://12462006
 define i8* @f3(i8* %base, i32* nocapture %offset, i32 %size) nounwind {
diff --git a/test/CodeGen/ARM/memcpy-inline.ll b/test/CodeGen/ARM/memcpy-inline.ll
index a5c525c062f..d874884dcb3 100644
--- a/test/CodeGen/ARM/memcpy-inline.ll
+++ b/test/CodeGen/ARM/memcpy-inline.ll
@@ -23,7 +23,7 @@ entry:
 ; CHECK-T1: strb [[TREG1]],
 ; CHECK-T1: ldrh [[TREG2:r[0-9]]],
 ; CHECK-T1: strh [[TREG2]]
-  call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 8 getelementptr inbounds (%struct.x, %struct.x* @dst, i32 0, i32 0), i8* align 8 getelementptr inbounds (%struct.x, %struct.x* @src, i32 0, i32 0), i32 11, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* getelementptr inbounds (%struct.x, %struct.x* @dst, i32 0, i32 0), i8* getelementptr inbounds (%struct.x, %struct.x* @src, i32 0, i32 0), i32 11, i32 8, i1 false)
   ret i32 0
 }
 
@@ -36,7 +36,7 @@ entry:
 ; CHECK: adds r1, #15
 ; CHECK: vld1.8 {d{{[0-9]+}}, d{{[0-9]+}}}, [r1]
 ; CHECK: vst1.8 {d{{[0-9]+}}, d{{[0-9]+}}}, [r0]
-  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %C, i8* getelementptr inbounds ([31 x i8], [31 x i8]* @.str1, i64 0, i64 0), i64 31, i1 false)
+  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %C, i8* getelementptr inbounds ([31 x i8], [31 x i8]* @.str1, i64 0, i64 0), i64 31, i32 1, i1 false)
   ret void
 }
 
@@ -50,7 +50,7 @@ entry:
 ; CHECK: vst1.8 {d{{[0-9]+}}, d{{[0-9]+}}}, [r0]!
 ; CHECK: vld1.64 {d{{[0-9]+}}, d{{[0-9]+}}}, [r1]
 ; CHECK: vst1.8 {d{{[0-9]+}}, d{{[0-9]+}}}, [r0]
-  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %C, i8* getelementptr inbounds ([36 x i8], [36 x i8]* @.str2, i64 0, i64 0), i64 36, i1 false)
+  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %C, i8* getelementptr inbounds ([36 x i8], [36 x i8]* @.str2, i64 0, i64 0), i64 36, i32 1, i1 false)
   ret void
 }
 
@@ -61,7 +61,7 @@ entry:
 ; CHECK: vst1.8 {d{{[0-9]+}}, d{{[0-9]+}}}, [r0]!
 ; CHECK: vldr d{{[0-9]+}}, [r1]
 ; CHECK: vst1.8 {d{{[0-9]+}}}, [r0]
-  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %C, i8* getelementptr inbounds ([24 x i8], [24 x i8]* @.str3, i64 0, i64 0), i64 24, i1 false)
+  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %C, i8* getelementptr inbounds ([24 x i8], [24 x i8]* @.str3, i64 0, i64 0), i64 24, i32 1, i1 false)
   ret void
 }
 
@@ -71,7 +71,7 @@ entry:
 ; CHECK: vld1.64 {[[REG3:d[0-9]+]], [[REG4:d[0-9]+]]}, [r1]
 ; CHECK: vst1.8 {[[REG3]], [[REG4]]}, [r0]!
 ; CHECK: strh [[REG5:r[0-9]+]], [r0]
-  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %C, i8* getelementptr inbounds ([18 x i8], [18 x i8]* @.str4, i64 0, i64 0), i64 18, i1 false)
+  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %C, i8* getelementptr inbounds ([18 x i8], [18 x i8]* @.str4, i64 0, i64 0), i64 18, i32 1, i1 false)
   ret void
 }
 
@@ -90,7 +90,7 @@ entry:
 ; CHECK-T1: strb [[TREG3]],
 ; CHECK-T1: movs [[TREG4:r[0-9]]],
 ; CHECK-T1: strb [[TREG4]],
-  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %C, i8* getelementptr inbounds ([7 x i8], [7 x i8]* @.str5, i64 0, i64 0), i64 7, i1 false)
+  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %C, i8* getelementptr inbounds ([7 x i8], [7 x i8]* @.str5, i64 0, i64 0), i64 7, i32 1, i1 false)
   ret void
 }
 
@@ -108,7 +108,7 @@ entry:
 ; CHECK-T1: strh [[TREG5]],
 ; CHECK-T1: ldr [[TREG6:r[0-9]]],
 ; CHECK-T1: str [[TREG6]]
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* getelementptr inbounds ([512 x i8], [512 x i8]* @spool.splbuf, i64 0, i64 0), i8* getelementptr inbounds ([14 x i8], [14 x i8]* @.str6, i64 0, i64 0), i64 14, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* getelementptr inbounds ([512 x i8], [512 x i8]* @spool.splbuf, i64 0, i64 0), i8* getelementptr inbounds ([14 x i8], [14 x i8]* @.str6, i64 0, i64 0), i64 14, i32 1, i1 false)
   ret void
 }
 
@@ -124,9 +124,9 @@ entry:
 ; CHECK-T1: str
   %0 = bitcast %struct.Foo* %a to i8*
   %1 = bitcast %struct.Foo* %b to i8*
-  tail call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 4 %0, i8* align 4 %1, i32 16, i1 false)
+  tail call void @llvm.memcpy.p0i8.p0i8.i32(i8* %0, i8* %1, i32 16, i32 4, i1 false)
   ret void
 }
 
-declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i1) nounwind
-declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i32, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i32, i1) nounwind
diff --git a/test/CodeGen/ARM/memfunc.ll b/test/CodeGen/ARM/memfunc.ll
index 1edf53b19d3..66743f3e9d5 100644
--- a/test/CodeGen/ARM/memfunc.ll
+++ b/test/CodeGen/ARM/memfunc.ll
@@ -14,13 +14,13 @@ entry:
   ; CHECK-DARWIN: bl _memmove
   ; CHECK-EABI: bl __aeabi_memmove
   ; CHECK-GNUEABI: bl memmove
-  call void @llvm.memmove.p0i8.p0i8.i32(i8* %dest, i8* %src, i32 500, i1 false)
+  call void @llvm.memmove.p0i8.p0i8.i32(i8* %dest, i8* %src, i32 500, i32 0, i1 false)
 
   ; CHECK-IOS: bl _memcpy
   ; CHECK-DARWIN: bl _memcpy
   ; CHECK-EABI: bl __aeabi_memcpy
   ; CHECK-GNUEABI: bl memcpy
-  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dest, i8* %src, i32 500, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dest, i8* %src, i32 500, i32 0, i1 false)
 
   ; EABI memset swaps arguments
   ; CHECK-IOS: mov r1, #1
@@ -31,7 +31,7 @@ entry:
   ; CHECK-EABI: bl __aeabi_memset
   ; CHECK-GNUEABI: mov r1, #1
   ; CHECK-GNUEABI: bl memset
-  call void @llvm.memset.p0i8.i32(i8* %dest, i8 1, i32 500, i1 false)
+  call void @llvm.memset.p0i8.i32(i8* %dest, i8 1, i32 500, i32 0, i1 false)
 
   ; EABI uses memclr if value set to 0
   ; CHECK-IOS: mov r1, #0
@@ -40,7 +40,7 @@ entry:
   ; CHECK-DARWIN: bl _memset
   ; CHECK-EABI: bl __aeabi_memclr
   ; CHECK-GNUEABI: bl memset
-  call void @llvm.memset.p0i8.i32(i8* %dest, i8 0, i32 500, i1 false)
+  call void @llvm.memset.p0i8.i32(i8* %dest, i8 0, i32 500, i32 0, i1 false)
 
   ; EABI uses aligned function variants if possible
 
@@ -48,49 +48,49 @@ entry:
   ; CHECK-DARWIN: bl _memmove
   ; CHECK-EABI: bl __aeabi_memmove4
   ; CHECK-GNUEABI: bl memmove
-  call void @llvm.memmove.p0i8.p0i8.i32(i8* align 4 %dest, i8* align 4 %src, i32 500, i1 false)
+  call void @llvm.memmove.p0i8.p0i8.i32(i8* %dest, i8* %src, i32 500, i32 4, i1 false)
 
   ; CHECK-IOS: bl _memcpy
   ; CHECK-DARWIN: bl _memcpy
   ; CHECK-EABI: bl __aeabi_memcpy4
   ; CHECK-GNUEABI: bl memcpy
-  call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 4 %dest, i8* align 4 %src, i32 500, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dest, i8* %src, i32 500, i32 4, i1 false)
 
   ; CHECK-IOS: bl _memset
   ; CHECK-DARWIN: bl _memset
   ; CHECK-EABI: bl __aeabi_memset4
   ; CHECK-GNUEABI: bl memset
-  call void @llvm.memset.p0i8.i32(i8* align 4 %dest, i8 1, i32 500, i1 false)
+  call void @llvm.memset.p0i8.i32(i8* %dest, i8 1, i32 500, i32 4, i1 false)
 
   ; CHECK-IOS: bl _memset
   ; CHECK-DARWIN: bl _memset
   ; CHECK-EABI: bl __aeabi_memclr4
   ; CHECK-GNUEABI: bl memset
-  call void @llvm.memset.p0i8.i32(i8* align 4 %dest, i8 0, i32 500, i1 false)
+  call void @llvm.memset.p0i8.i32(i8* %dest, i8 0, i32 500, i32 4, i1 false)
 
   ; CHECK-IOS: bl _memmove
   ; CHECK-DARWIN: bl _memmove
   ; CHECK-EABI: bl __aeabi_memmove8
   ; CHECK-GNUEABI: bl memmove
-  call void @llvm.memmove.p0i8.p0i8.i32(i8* align 8 %dest, i8* align 8 %src, i32 500, i1 false)
+  call void @llvm.memmove.p0i8.p0i8.i32(i8* %dest, i8* %src, i32 500, i32 8, i1 false)
 
   ; CHECK-IOS: bl _memcpy
   ; CHECK-DARWIN: bl _memcpy
   ; CHECK-EABI: bl __aeabi_memcpy8
   ; CHECK-GNUEABI: bl memcpy
-  call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 8 %dest, i8* align 8 %src, i32 500, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dest, i8* %src, i32 500, i32 8, i1 false)
 
   ; CHECK-IOS: bl _memset
   ; CHECK-DARWIN: bl _memset
   ; CHECK-EABI: bl __aeabi_memset8
   ; CHECK-GNUEABI: bl memset
-  call void @llvm.memset.p0i8.i32(i8* align 8 %dest, i8 1, i32 500, i1 false)
+  call void @llvm.memset.p0i8.i32(i8* %dest, i8 1, i32 500, i32 8, i1 false)
 
   ; CHECK-IOS: bl _memset
   ; CHECK-DARWIN: bl _memset
   ; CHECK-EABI: bl __aeabi_memclr8
   ; CHECK-GNUEABI: bl memset
-  call void @llvm.memset.p0i8.i32(i8* align 8 %dest, i8 0, i32 500, i1 false)
+  call void @llvm.memset.p0i8.i32(i8* %dest, i8 0, i32 500, i32 8, i1 false)
 
   unreachable
 }
@@ -111,7 +111,7 @@ entry:
   ; CHECK-GNUEABI: bl memmove
   %arr0 = alloca [9 x i8], align 1
   %0 = bitcast [9 x i8]* %arr0 to i8*
-  call void @llvm.memmove.p0i8.p0i8.i32(i8* %dest, i8* %0, i32 %n, i1 false)
+  call void @llvm.memmove.p0i8.p0i8.i32(i8* %dest, i8* %0, i32 %n, i32 0, i1 false)
 
   ; CHECK: add r1, sp, #16
   ; CHECK-IOS: bl _memcpy
@@ -120,7 +120,7 @@ entry:
   ; CHECK-GNUEABI: bl memcpy
   %arr1 = alloca [9 x i8], align 1
   %1 = bitcast [9 x i8]* %arr1 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dest, i8* %1, i32 %n, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dest, i8* %1, i32 %n, i32 0, i1 false)
 
   ; CHECK-IOS: mov r0, sp
   ; CHECK-IOS: mov r1, #1
@@ -136,7 +136,7 @@ entry:
   ; CHECK-GNUEABI: bl memset
   %arr2 = alloca [9 x i8], align 1
   %2 = bitcast [9 x i8]* %arr2 to i8*
-  call void @llvm.memset.p0i8.i32(i8* %2, i8 1, i32 %n, i1 false)
+  call void @llvm.memset.p0i8.i32(i8* %2, i8 1, i32 %n, i32 0, i1 false)
 
   unreachable
 }
@@ -153,7 +153,7 @@ entry:
   ; CHECK-GNUEABI: bl memmove
   %arr0 = alloca [7 x i8], align 1
   %0 = bitcast [7 x i8]* %arr0 to i8*
-  call void @llvm.memmove.p0i8.p0i8.i32(i8* %dest, i8* %0, i32 %n, i1 false)
+  call void @llvm.memmove.p0i8.p0i8.i32(i8* %dest, i8* %0, i32 %n, i32 0, i1 false)
 
   ; CHECK: {{add(.w)? r1, sp, #10}}
   ; CHECK-IOS: bl _memcpy
@@ -162,7 +162,7 @@ entry:
   ; CHECK-GNUEABI: bl memcpy
   %arr1 = alloca [7 x i8], align 1
   %1 = bitcast [7 x i8]* %arr1 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dest, i8* %1, i32 %n, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dest, i8* %1, i32 %n, i32 0, i1 false)
 
   ; CHECK: {{add(.w)? r0, sp, #3}}
   ; CHECK-IOS: mov r1, #1
@@ -175,7 +175,7 @@ entry:
   ; CHECK-GNUEABI: bl memset
   %arr2 = alloca [7 x i8], align 1
   %2 = bitcast [7 x i8]* %arr2 to i8*
-  call void @llvm.memset.p0i8.i32(i8* %2, i8 1, i32 %n, i1 false)
+  call void @llvm.memset.p0i8.i32(i8* %2, i8 1, i32 %n, i32 0, i1 false)
 
   unreachable
 }
@@ -192,7 +192,7 @@ entry:
   ; CHECK-GNUEABI: bl memmove
   %arr0 = alloca [9 x i8], align 1
   %0 = getelementptr inbounds [9 x i8], [9 x i8]* %arr0, i32 0, i32 4
-  call void @llvm.memmove.p0i8.p0i8.i32(i8* %dest, i8* %0, i32 %n, i1 false)
+  call void @llvm.memmove.p0i8.p0i8.i32(i8* %dest, i8* %0, i32 %n, i32 0, i1 false)
 
   ; CHECK: {{add(.w)? r., sp, #(10|14)}}
   ; CHECK-IOS: bl _memcpy
@@ -201,7 +201,7 @@ entry:
   ; CHECK-GNUEABI: bl memcpy
   %arr1 = alloca [9 x i8], align 1
   %1 = getelementptr inbounds [9 x i8], [9 x i8]* %arr1, i32 0, i32 4
-  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dest, i8* %1, i32 %n, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dest, i8* %1, i32 %n, i32 0, i1 false)
 
   ; CHECK: {{add(.w)? r., sp, #(1|5)}}
   ; CHECK-IOS: mov r1, #1
@@ -214,7 +214,7 @@ entry:
   ; CHECK-GNUEABI: bl memset
   %arr2 = alloca [9 x i8], align 1
   %2 = getelementptr inbounds [9 x i8], [9 x i8]* %arr2, i32 0, i32 4
-  call void @llvm.memset.p0i8.i32(i8* %2, i8 1, i32 %n, i1 false)
+  call void @llvm.memset.p0i8.i32(i8* %2, i8 1, i32 %n, i32 0, i1 false)
 
   unreachable
 }
@@ -231,7 +231,7 @@ entry:
   ; CHECK-GNUEABI: bl memmove
   %arr0 = alloca [13 x i8], align 1
   %0 = getelementptr inbounds [13 x i8], [13 x i8]* %arr0, i32 0, i32 1
-  call void @llvm.memmove.p0i8.p0i8.i32(i8* %dest, i8* %0, i32 %n, i1 false)
+  call void @llvm.memmove.p0i8.p0i8.i32(i8* %dest, i8* %0, i32 %n, i32 0, i1 false)
 
   ; CHECK: {{add(.w)? r., sp, #(10|14)}}
   ; CHECK-IOS: bl _memcpy
@@ -240,7 +240,7 @@ entry:
   ; CHECK-GNUEABI: bl memcpy
   %arr1 = alloca [13 x i8], align 1
   %1 = getelementptr inbounds [13 x i8], [13 x i8]* %arr1, i32 0, i32 1
-  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dest, i8* %1, i32 %n, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dest, i8* %1, i32 %n, i32 0, i1 false)
 
   ; CHECK: {{add(.w)? r., sp, #(1|5)}}
   ; CHECK-IOS: mov r1, #1
@@ -253,7 +253,7 @@ entry:
   ; CHECK-GNUEABI: bl memset
   %arr2 = alloca [13 x i8], align 1
   %2 = getelementptr inbounds [13 x i8], [13 x i8]* %arr2, i32 0, i32 1
-  call void @llvm.memset.p0i8.i32(i8* %2, i8 1, i32 %n, i1 false)
+  call void @llvm.memset.p0i8.i32(i8* %2, i8 1, i32 %n, i32 0, i1 false)
 
   unreachable
 }
@@ -270,7 +270,7 @@ entry:
   ; CHECK-GNUEABI: bl memmove
   %arr0 = alloca [13 x i8], align 1
   %0 = getelementptr inbounds [13 x i8], [13 x i8]* %arr0, i32 0, i32 %i
-  call void @llvm.memmove.p0i8.p0i8.i32(i8* %dest, i8* %0, i32 %n, i1 false)
+  call void @llvm.memmove.p0i8.p0i8.i32(i8* %dest, i8* %0, i32 %n, i32 0, i1 false)
 
   ; CHECK: {{add(.w)? r., sp, #(10|14)}}
   ; CHECK-IOS: bl _memcpy
@@ -279,7 +279,7 @@ entry:
   ; CHECK-GNUEABI: bl memcpy
   %arr1 = alloca [13 x i8], align 1
   %1 = getelementptr inbounds [13 x i8], [13 x i8]* %arr1, i32 0, i32 %i
-  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dest, i8* %1, i32 %n, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dest, i8* %1, i32 %n, i32 0, i1 false)
 
   ; CHECK: {{add(.w)? r., sp, #(1|5)}}
   ; CHECK-IOS: mov r1, #1
@@ -292,7 +292,7 @@ entry:
   ; CHECK-GNUEABI: bl memset
   %arr2 = alloca [13 x i8], align 1
   %2 = getelementptr inbounds [13 x i8], [13 x i8]* %arr2, i32 0, i32 %i
-  call void @llvm.memset.p0i8.i32(i8* %2, i8 1, i32 %n, i1 false)
+  call void @llvm.memset.p0i8.i32(i8* %2, i8 1, i32 %n, i32 0, i1 false)
 
   unreachable
 }
@@ -309,7 +309,7 @@ entry:
   ; CHECK-GNUEABI: bl memmove
   %arr0 = alloca [13 x i8], align 1
   %0 = getelementptr [13 x i8], [13 x i8]* %arr0, i32 0, i32 4
-  call void @llvm.memmove.p0i8.p0i8.i32(i8* %dest, i8* %0, i32 %n, i1 false)
+  call void @llvm.memmove.p0i8.p0i8.i32(i8* %dest, i8* %0, i32 %n, i32 0, i1 false)
 
   ; CHECK: {{add(.w)? r., sp, #(10|14)}}
   ; CHECK-IOS: bl _memcpy
@@ -318,7 +318,7 @@ entry:
   ; CHECK-GNUEABI: bl memcpy
   %arr1 = alloca [13 x i8], align 1
   %1 = getelementptr [13 x i8], [13 x i8]* %arr1, i32 0, i32 4
-  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dest, i8* %1, i32 %n, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dest, i8* %1, i32 %n, i32 0, i1 false)
 
   ; CHECK: {{add(.w)? r., sp, #(1|5)}}
   ; CHECK-IOS: mov r1, #1
@@ -331,7 +331,7 @@ entry:
   ; CHECK-GNUEABI: bl memset
   %arr2 = alloca [13 x i8], align 1
   %2 = getelementptr [13 x i8], [13 x i8]* %arr2, i32 0, i32 4
-  call void @llvm.memset.p0i8.i32(i8* %2, i8 1, i32 %n, i1 false)
+  call void @llvm.memset.p0i8.i32(i8* %2, i8 1, i32 %n, i32 0, i1 false)
 
   unreachable
 }
@@ -348,7 +348,7 @@ entry:
   ; CHECK-GNUEABI: bl memmove
   %arr0 = alloca [13 x i8], align 1
   %0 = getelementptr inbounds [13 x i8], [13 x i8]* %arr0, i32 0, i32 16
-  call void @llvm.memmove.p0i8.p0i8.i32(i8* %dest, i8* %0, i32 %n, i1 false)
+  call void @llvm.memmove.p0i8.p0i8.i32(i8* %dest, i8* %0, i32 %n, i32 0, i1 false)
 
   ; CHECK: {{add(.w)? r., sp, #(10|14)}}
   ; CHECK-IOS: bl _memcpy
@@ -357,7 +357,7 @@ entry:
   ; CHECK-GNUEABI: bl memcpy
   %arr1 = alloca [13 x i8], align 1
   %1 = getelementptr inbounds [13 x i8], [13 x i8]* %arr1, i32 0, i32 16
-  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dest, i8* %1, i32 %n, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dest, i8* %1, i32 %n, i32 0, i1 false)
 
   ; CHECK: {{add(.w)? r., sp, #(1|5)}}
   ; CHECK-IOS: mov r1, #1
@@ -370,7 +370,7 @@ entry:
   ; CHECK-GNUEABI: bl memset
   %arr2 = alloca [13 x i8], align 1
   %2 = getelementptr inbounds [13 x i8], [13 x i8]* %arr2, i32 0, i32 16
-  call void @llvm.memset.p0i8.i32(i8* %2, i8 1, i32 %n, i1 false)
+  call void @llvm.memset.p0i8.i32(i8* %2, i8 1, i32 %n, i32 0, i1 false)
 
   unreachable
 }
@@ -386,13 +386,13 @@ entry:
 @arr7 = external global [7 x i8], align 1
 define void @f9(i8* %dest, i32 %n) {
 entry:
-  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dest, i8* getelementptr inbounds ([7 x i8], [7 x i8]* @arr1, i32 0, i32 0), i32 %n, i1 false)
-  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dest, i8* getelementptr inbounds ([8 x i8], [8 x i8]* @arr2, i32 0, i32 0), i32 %n, i1 false)
-  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dest, i8* getelementptr inbounds ([7 x i8], [7 x i8]* @arr3, i32 0, i32 0), i32 %n, i1 false)
-  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dest, i8* getelementptr inbounds ([8 x i8], [8 x i8]* @arr4, i32 0, i32 0), i32 %n, i1 false)
-  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dest, i8* getelementptr inbounds ([7 x i8], [7 x i8]* @arr5, i32 0, i32 0), i32 %n, i1 false)
-  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dest, i8* getelementptr inbounds ([7 x i8], [7 x i8]* @arr6, i32 0, i32 0), i32 %n, i1 false)
-  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dest, i8* getelementptr inbounds ([7 x i8], [7 x i8]* @arr7, i32 0, i32 0), i32 %n, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dest, i8* getelementptr inbounds ([7 x i8], [7 x i8]* @arr1, i32 0, i32 0), i32 %n, i32 1, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dest, i8* getelementptr inbounds ([8 x i8], [8 x i8]* @arr2, i32 0, i32 0), i32 %n, i32 1, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dest, i8* getelementptr inbounds ([7 x i8], [7 x i8]* @arr3, i32 0, i32 0), i32 %n, i32 1, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dest, i8* getelementptr inbounds ([8 x i8], [8 x i8]* @arr4, i32 0, i32 0), i32 %n, i32 1, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dest, i8* getelementptr inbounds ([7 x i8], [7 x i8]* @arr5, i32 0, i32 0), i32 %n, i32 1, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dest, i8* getelementptr inbounds ([7 x i8], [7 x i8]* @arr6, i32 0, i32 0), i32 %n, i32 1, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dest, i8* getelementptr inbounds ([7 x i8], [7 x i8]* @arr7, i32 0, i32 0), i32 %n, i32 1, i1 false)
 
   unreachable
 }
@@ -417,6 +417,6 @@ entry:
 ; CHECK: arr6:
 ; CHECK-NOT: arr7:
 
-declare void @llvm.memmove.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i1) nounwind
-declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i1) nounwind
-declare void @llvm.memset.p0i8.i32(i8* nocapture, i8, i32, i1) nounwind
+declare void @llvm.memmove.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i32, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i32, i1) nounwind
+declare void @llvm.memset.p0i8.i32(i8* nocapture, i8, i32, i32, i1) nounwind
diff --git a/test/CodeGen/ARM/memset-inline.ll b/test/CodeGen/ARM/memset-inline.ll
index 98334600ddb..f6f8d562350 100644
--- a/test/CodeGen/ARM/memset-inline.ll
+++ b/test/CodeGen/ARM/memset-inline.ll
@@ -6,7 +6,7 @@ entry:
 ; CHECK: movs r1, #0
 ; CHECK: strd r1, r1, [r0]
 ; CHECK: str r1, [r0, #8]
-  call void @llvm.memset.p0i8.i64(i8* align 8 %c, i8 0, i64 12, i1 false)
+  call void @llvm.memset.p0i8.i64(i8* %c, i8 0, i64 12, i32 8, i1 false)
   ret void
 }
 
@@ -19,11 +19,11 @@ entry:
 ; CHECK: vst1.64 {d{{[0-9]+}}, d{{[0-9]+}}}, [r0]
   %buf = alloca [26 x i8], align 1
   %0 = getelementptr inbounds [26 x i8], [26 x i8]* %buf, i32 0, i32 0
-  call void @llvm.memset.p0i8.i32(i8* align 1 %0, i8 0, i32 26, i1 false)
+  call void @llvm.memset.p0i8.i32(i8* %0, i8 0, i32 26, i32 1, i1 false)
   call void @something(i8* %0) nounwind
   ret void
 }
 
 declare void @something(i8*) nounwind
-declare void @llvm.memset.p0i8.i32(i8* nocapture, i8, i32, i1) nounwind
-declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i1) nounwind
+declare void @llvm.memset.p0i8.i32(i8* nocapture, i8, i32, i32, i1) nounwind
+declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i32, i1) nounwind
diff --git a/test/CodeGen/ARM/stack-protector-bmovpcb_call.ll b/test/CodeGen/ARM/stack-protector-bmovpcb_call.ll
index e2fe0cfaabb..2a7a82da8f6 100644
--- a/test/CodeGen/ARM/stack-protector-bmovpcb_call.ll
+++ b/test/CodeGen/ARM/stack-protector-bmovpcb_call.ll
@@ -15,13 +15,13 @@ define i32 @main() #0 {
 entry:
   %title = alloca [15 x i8], align 1
   %0 = getelementptr inbounds [15 x i8], [15 x i8]* %title, i32 0, i32 0
-  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %0, i8* getelementptr inbounds ([15 x i8], [15 x i8]* @main.title, i32 0, i32 0), i32 15, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %0, i8* getelementptr inbounds ([15 x i8], [15 x i8]* @main.title, i32 0, i32 0), i32 15, i32 1, i1 false)
   %call = call i32 (i8*, ...) @printf(i8* getelementptr inbounds ([3 x i8], [3 x i8]* @.str, i32 0, i32 0), i8* %0) #3
   ret i32 0
 }
 
 ; Function Attrs: nounwind
-declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture readonly, i32, i1) #1
+declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture readonly, i32, i32, i1) #1
 
 ; Function Attrs: nounwind optsize
 declare i32 @printf(i8* nocapture readonly, ...) #2
diff --git a/test/CodeGen/ARM/struct-byval-frame-index.ll b/test/CodeGen/ARM/struct-byval-frame-index.ll
index a2e2c32f50e..52f70fe1e0f 100644
--- a/test/CodeGen/ARM/struct-byval-frame-index.ll
+++ b/test/CodeGen/ARM/struct-byval-frame-index.ll
@@ -61,10 +61,10 @@ target triple = "armv7l-unknown-linux-gnueabihf"
 @brefframe = external global [4 x [4 x i8]], align 1
 
 ; Function Attrs: nounwind
-declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i1) #0
+declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i32, i1) #0
 
 ; Function Attrs: nounwind
-declare void @llvm.memset.p0i8.i32(i8* nocapture, i8, i32, i1) #0
+declare void @llvm.memset.p0i8.i32(i8* nocapture, i8, i32, i32, i1) #0
 
 ; Function Attrs: nounwind
 declare void @SetMotionVectorsMB(%structK* nocapture, i32) #1
@@ -123,10 +123,10 @@ for.cond210.preheader:                            ; preds = %if.then169
   unreachable
 
 if.end230:                                        ; preds = %if.end164
-  tail call void @llvm.memcpy.p0i8.p0i8.i32(i8* undef, i8* bitcast ([4 x i32]* @b8mode to i8*), i32 16, i1 false)
+  tail call void @llvm.memcpy.p0i8.p0i8.i32(i8* undef, i8* bitcast ([4 x i32]* @b8mode to i8*), i32 16, i32 4, i1 false)
   %b8pdir = getelementptr inbounds %structK, %structK* %2, i32 %1, i32 15
   %3 = bitcast [4 x i32]* %b8pdir to i8*
-  tail call void @llvm.memcpy.p0i8.p0i8.i32(i8* %3, i8* bitcast ([4 x i32]* @b8pdir to i8*), i32 16, i1 false)
+  tail call void @llvm.memcpy.p0i8.p0i8.i32(i8* %3, i8* bitcast ([4 x i32]* @b8pdir to i8*), i32 16, i32 4, i1 false)
   br i1 undef, label %if.end236, label %if.then233
 
 if.then233:                                       ; preds = %if.end230
diff --git a/test/CodeGen/BPF/byval.ll b/test/CodeGen/BPF/byval.ll
index 2a09cf1376d..25ba909d9cd 100644
--- a/test/CodeGen/BPF/byval.ll
+++ b/test/CodeGen/BPF/byval.ll
@@ -16,7 +16,7 @@ entry:
   store i32 3, i32* %arrayinit.element2, align 8
   %arrayinit.start = getelementptr inbounds %struct.S, %struct.S* %.compoundliteral, i64 0, i32 0, i64 3
   %scevgep4 = bitcast i32* %arrayinit.start to i8*
-  call void @llvm.memset.p0i8.i64(i8* %scevgep4, i8 0, i64 28, i1 false)
+  call void @llvm.memset.p0i8.i64(i8* %scevgep4, i8 0, i64 28, i32 4, i1 false)
   call void @foo(i32 %a, %struct.S* byval align 8 %.compoundliteral) #3
   ret void
 }
@@ -24,4 +24,4 @@ entry:
 declare void @foo(i32, %struct.S* byval align 8) #1
 
 ; Function Attrs: nounwind
-declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i1) #3
+declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i32, i1) #3
diff --git a/test/CodeGen/BPF/ex1.ll b/test/CodeGen/BPF/ex1.ll
index 4aa90ef13df..546e5d49da6 100644
--- a/test/CodeGen/BPF/ex1.ll
+++ b/test/CodeGen/BPF/ex1.ll
@@ -12,7 +12,7 @@ define i32 @bpf_prog1(%struct.bpf_context* nocapture %ctx) #0 section "events/ne
   %devname = alloca [3 x i8], align 1
   %fmt = alloca [15 x i8], align 1
   %1 = getelementptr inbounds [3 x i8], [3 x i8]* %devname, i64 0, i64 0
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %1, i8* getelementptr inbounds ([3 x i8], [3 x i8]* @bpf_prog1.devname, i64 0, i64 0), i64 3, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %1, i8* getelementptr inbounds ([3 x i8], [3 x i8]* @bpf_prog1.devname, i64 0, i64 0), i64 3, i32 1, i1 false)
   %2 = getelementptr inbounds %struct.bpf_context, %struct.bpf_context* %ctx, i64 0, i32 0
   %3 = load i64, i64* %2, align 8
   %4 = inttoptr i64 %3 to %struct.sk_buff*
@@ -25,7 +25,7 @@ define i32 @bpf_prog1(%struct.bpf_context* nocapture %ctx) #0 section "events/ne
 
 ; <label>:10                                      ; preds = %0
   %11 = getelementptr inbounds [15 x i8], [15 x i8]* %fmt, i64 0, i64 0
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %11, i8* getelementptr inbounds ([15 x i8], [15 x i8]* @bpf_prog1.fmt, i64 0, i64 0), i64 15, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %11, i8* getelementptr inbounds ([15 x i8], [15 x i8]* @bpf_prog1.fmt, i64 0, i64 0), i64 15, i32 1, i1 false)
   %12 = call i32 (i8*, i32, ...) inttoptr (i64 11 to i32 (i8*, i32, ...)*)(i8* %11, i32 15, %struct.sk_buff* %4, i8* %7) #1
 ; CHECK-LABEL: bpf_prog1:
 ; CHECK: call 4
@@ -43,4 +43,4 @@ define i32 @bpf_prog1(%struct.bpf_context* nocapture %ctx) #0 section "events/ne
 }
 
 ; Function Attrs: nounwind
-declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i1) #1
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i32, i1) #1
diff --git a/test/CodeGen/BPF/sanity.ll b/test/CodeGen/BPF/sanity.ll
index a2f18dcc989..7f0ef889ff9 100644
--- a/test/CodeGen/BPF/sanity.ll
+++ b/test/CodeGen/BPF/sanity.ll
@@ -103,7 +103,7 @@ declare i32 @manyarg(i32, i32, i32, i32, i32) #2
 define void @foo_printf() #1 {
   %fmt = alloca [9 x i8], align 1
   %1 = getelementptr inbounds [9 x i8], [9 x i8]* %fmt, i64 0, i64 0
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %1, i8* getelementptr inbounds ([9 x i8], [9 x i8]* @foo_printf.fmt, i64 0, i64 0), i64 9, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %1, i8* getelementptr inbounds ([9 x i8], [9 x i8]* @foo_printf.fmt, i64 0, i64 0), i64 9, i32 1, i1 false)
 ; CHECK-LABEL: foo_printf:
 ; CHECK: ld_64 r1, 729618802566522216
   %2 = call i32 (i8*, ...) @printf(i8* %1) #3
@@ -111,7 +111,7 @@ define void @foo_printf() #1 {
 }
 
 ; Function Attrs: nounwind
-declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i1) #3
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i32, i1) #3
 
 ; Function Attrs: nounwind
 declare i32 @printf(i8* nocapture, ...) #4
diff --git a/test/CodeGen/Generic/ForceStackAlign.ll b/test/CodeGen/Generic/ForceStackAlign.ll
index 7eed8321308..57ccb2c41d7 100644
--- a/test/CodeGen/Generic/ForceStackAlign.ll
+++ b/test/CodeGen/Generic/ForceStackAlign.ll
@@ -18,10 +18,10 @@ entry:
 
 if.then:
   %0 = alloca i8, i32 %i
-  call void @llvm.memset.p0i8.i32(i8* %0, i8 0, i32 %i, i1 false)
+  call void @llvm.memset.p0i8.i32(i8* %0, i8 0, i32 %i, i32 1, i1 false)
   %call = call i32 @f(i8* %0)
   %conv = sext i32 %call to i64
   ret i64 %conv
 }
 
-declare void @llvm.memset.p0i8.i32(i8*, i8, i32, i1) nounwind
+declare void @llvm.memset.p0i8.i32(i8*, i8, i32, i32, i1) nounwind
diff --git a/test/CodeGen/Generic/invalid-memcpy.ll b/test/CodeGen/Generic/invalid-memcpy.ll
index 3fcb5d6d966..d4252bc9d98 100644
--- a/test/CodeGen/Generic/invalid-memcpy.ll
+++ b/test/CodeGen/Generic/invalid-memcpy.ll
@@ -10,8 +10,8 @@ define void @Bork() {
 entry:
   %Qux = alloca [33 x i8]
   %Qux1 = bitcast [33 x i8]* %Qux to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %Qux1, i8* getelementptr inbounds ([33 x i8], [33 x i8]* @C.0.1173, i32 0, i32 0), i64 33, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %Qux1, i8* getelementptr inbounds ([33 x i8], [33 x i8]* @C.0.1173, i32 0, i32 0), i64 33, i32 8, i1 false)
   ret void
 }
 
-declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i32, i1) nounwind
diff --git a/test/CodeGen/Hexagon/mem-fi-add.ll b/test/CodeGen/Hexagon/mem-fi-add.ll
index c4529ecb8f2..a46029fdb5e 100644
--- a/test/CodeGen/Hexagon/mem-fi-add.ll
+++ b/test/CodeGen/Hexagon/mem-fi-add.ll
@@ -13,14 +13,14 @@ define void @foo() #0 {
 entry:
   %t = alloca [4 x [2 x i32]], align 8
   %0 = bitcast [4 x [2 x i32]]* %t to i8*
-  call void @llvm.memset.p0i8.i32(i8* %0, i8 0, i32 32, i1 false)
+  call void @llvm.memset.p0i8.i32(i8* %0, i8 0, i32 32, i32 8, i1 false)
   %arraydecay = getelementptr inbounds [4 x [2 x i32]], [4 x [2 x i32]]* %t, i32 0, i32 0
   call void @bar([2 x i32]* %arraydecay) #1
   ret void
 }
 
 ; Function Attrs: nounwind
-declare void @llvm.memset.p0i8.i32(i8* nocapture, i8, i32, i1) #1
+declare void @llvm.memset.p0i8.i32(i8* nocapture, i8, i32, i32, i1) #1
 
 declare void @bar([2 x i32]*) #2
 
diff --git a/test/CodeGen/Hexagon/tail-call-mem-intrinsics.ll b/test/CodeGen/Hexagon/tail-call-mem-intrinsics.ll
index 7f0fb6281ff..90fb75e5be0 100644
--- a/test/CodeGen/Hexagon/tail-call-mem-intrinsics.ll
+++ b/test/CodeGen/Hexagon/tail-call-mem-intrinsics.ll
@@ -4,7 +4,7 @@
 ; CHECK: jump memcpy
 define void @tail_memcpy(i8* nocapture %p, i8* nocapture readonly %q, i32 %n) #0 {
 entry:
-  tail call void @llvm.memcpy.p0i8.p0i8.i32(i8* %p, i8* %q, i32 %n, i1 false)
+  tail call void @llvm.memcpy.p0i8.p0i8.i32(i8* %p, i8* %q, i32 %n, i32 1, i1 false)
   ret void
 }
 
@@ -12,7 +12,7 @@ entry:
 ; CHECK: jump memmove
 define void @tail_memmove(i8* nocapture %p, i8* nocapture readonly %q, i32 %n) #0 {
 entry:
-  tail call void @llvm.memmove.p0i8.p0i8.i32(i8* %p, i8* %q, i32 %n, i1 false)
+  tail call void @llvm.memmove.p0i8.p0i8.i32(i8* %p, i8* %q, i32 %n, i32 1, i1 false)
   ret void
 }
 
@@ -20,12 +20,12 @@ entry:
 ; CHECK: jump memset
 define void @tail_memset(i8* nocapture %p, i8 %c, i32 %n) #0 {
 entry:
-  tail call void @llvm.memset.p0i8.i32(i8* %p, i8 %c, i32 %n, i1 false)
+  tail call void @llvm.memset.p0i8.i32(i8* %p, i8 %c, i32 %n, i32 1, i1 false)
   ret void
 }
 
-declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture readonly, i32, i1) #0
-declare void @llvm.memmove.p0i8.p0i8.i32(i8* nocapture, i8* nocapture readonly, i32, i1) #0
-declare void @llvm.memset.p0i8.i32(i8* nocapture, i8, i32, i1) #0
+declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture readonly, i32, i32, i1) #0
+declare void @llvm.memmove.p0i8.p0i8.i32(i8* nocapture, i8* nocapture readonly, i32, i32, i1) #0
+declare void @llvm.memset.p0i8.i32(i8* nocapture, i8, i32, i32, i1) #0
 
 attributes #0 = { nounwind }
diff --git a/test/CodeGen/MSP430/memset.ll b/test/CodeGen/MSP430/memset.ll
index 68c4f8583dd..76cfb29586d 100644
--- a/test/CodeGen/MSP430/memset.ll
+++ b/test/CodeGen/MSP430/memset.ll
@@ -13,10 +13,10 @@ entry:
 ; CHECK-NEXT: mov.w #5, r14
 ; CHECK-NEXT: mov.w #128, r13
 ; CHECK-NEXT: call #memset
-  call void @llvm.memset.p0i8.i16(i8* %0, i8 5, i16 128, i1 false)
+  call void @llvm.memset.p0i8.i16(i8* %0, i8 5, i16 128, i32 1, i1 false)
   ret void
 }
 
 ; Function Attrs: nounwind
-declare void @llvm.memset.p0i8.i16(i8* nocapture, i8, i16, i1) nounwind
+declare void @llvm.memset.p0i8.i16(i8* nocapture, i8, i16, i32, i1) nounwind
 
diff --git a/test/CodeGen/Mips/2012-12-12-ExpandMemcpy.ll b/test/CodeGen/Mips/2012-12-12-ExpandMemcpy.ll
index 2964b19c1d3..24bcfaee8ba 100644
--- a/test/CodeGen/Mips/2012-12-12-ExpandMemcpy.ll
+++ b/test/CodeGen/Mips/2012-12-12-ExpandMemcpy.ll
@@ -4,8 +4,8 @@
 
 define void @t(i8* %ptr) {
 entry:
-  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %ptr, i8* getelementptr inbounds ([7 x i8], [7 x i8]* @.str, i64 0, i64 0), i64 7, i1 false)
+  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %ptr, i8* getelementptr inbounds ([7 x i8], [7 x i8]* @.str, i64 0, i64 0), i64 7, i32 1, i1 false)
   ret void
 }
 
-declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i32, i1) nounwind
diff --git a/test/CodeGen/Mips/Fast-ISel/memtest1.ll b/test/CodeGen/Mips/Fast-ISel/memtest1.ll
index 5ff9c290ce6..b98200d7456 100644
--- a/test/CodeGen/Mips/Fast-ISel/memtest1.ll
+++ b/test/CodeGen/Mips/Fast-ISel/memtest1.ll
@@ -10,9 +10,9 @@
 @i = global i32 12, align 4
 @dest = common global [50 x i8] zeroinitializer, align 1
 
-declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture readonly, i32, i1)
-declare void @llvm.memmove.p0i8.p0i8.i32(i8* nocapture, i8* nocapture readonly, i32, i1)
-declare void @llvm.memset.p0i8.i32(i8* nocapture, i8, i32, i1)
+declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture readonly, i32, i32, i1)
+declare void @llvm.memmove.p0i8.p0i8.i32(i8* nocapture, i8* nocapture readonly, i32, i32, i1)
+declare void @llvm.memset.p0i8.i32(i8* nocapture, i8, i32, i32, i1)
 
 define void @cpy(i8* %src, i32 %i) {
   ; ALL-LABEL:  cpy:
@@ -29,7 +29,7 @@ define void @cpy(i8* %src, i32 %i) {
   ; ALL-NEXT:       nop
   ; ALL-NOT:        {{.*}}$2{{.*}}
   call void @llvm.memcpy.p0i8.p0i8.i32(i8* getelementptr inbounds ([50 x i8], [50 x i8]* @dest, i32 0, i32 0),
-                                       i8* %src, i32 %i, i1 false)
+                                       i8* %src, i32 %i, i32 1, i1 false)
   ret void
 }
 
@@ -49,7 +49,7 @@ define void @mov(i8* %src, i32 %i) {
   ; ALL-NEXT:       nop
   ; ALL-NOT:        {{.*}}$2{{.*}}
   call void @llvm.memmove.p0i8.p0i8.i32(i8* getelementptr inbounds ([50 x i8], [50 x i8]* @dest, i32 0, i32 0),
-                                        i8* %src, i32 %i, i1 false)
+                                        i8* %src, i32 %i, i32 1, i1 false)
   ret void
 }
 
@@ -69,6 +69,6 @@ define void @clear(i32 %i) {
   ; ALL-NEXT:       nop
   ; ALL-NOT:        {{.*}}$2{{.*}}
   call void @llvm.memset.p0i8.i32(i8* getelementptr inbounds ([50 x i8], [50 x i8]* @dest, i32 0, i32 0),
-                                  i8 42, i32 %i, i1 false)
+                                  i8 42, i32 %i, i32 1, i1 false)
   ret void
 }
diff --git a/test/CodeGen/Mips/biggot.ll b/test/CodeGen/Mips/biggot.ll
index b4fff2138bf..b56ce6ba87b 100644
--- a/test/CodeGen/Mips/biggot.ll
+++ b/test/CodeGen/Mips/biggot.ll
@@ -43,8 +43,8 @@ entry:
 
   %0 = bitcast i32* %d to i8*
   %1 = bitcast i32* %s to i8*
-  tail call void @llvm.memcpy.p0i8.p0i8.i32(i8* %0, i8* %1, i32 %n, i1 false)
+  tail call void @llvm.memcpy.p0i8.p0i8.i32(i8* %0, i8* %1, i32 %n, i32 4, i1 false)
   ret void
 }
 
-declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i32, i1) nounwind
diff --git a/test/CodeGen/Mips/cconv/arguments-small-structures-bigger-than-32bits.ll b/test/CodeGen/Mips/cconv/arguments-small-structures-bigger-than-32bits.ll
index e0f0382b256..087a0515f37 100644
--- a/test/CodeGen/Mips/cconv/arguments-small-structures-bigger-than-32bits.ll
+++ b/test/CodeGen/Mips/cconv/arguments-small-structures-bigger-than-32bits.ll
@@ -43,7 +43,7 @@
 declare void @fS1(i48 inreg) #1
 declare void @fS2(i40 inreg) #1
 
-declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture readonly, i64, i1) #2
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture readonly, i64, i32, i1) #2
 
 define void @f1() #0 {
 entry:
@@ -51,7 +51,7 @@ entry:
   %s1_1.coerce = alloca { i48 }
   %0 = bitcast { i48 }* %s1_1.coerce to i8*
   %1 = bitcast %struct.S1* %s1_1 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %0, i8* %1, i64 6, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %0, i8* %1, i64 6, i32 0, i1 false)
   %2 = getelementptr { i48 }, { i48 }* %s1_1.coerce, i32 0, i32 0
   %3 = load i48, i48* %2, align 1
   call void @fS1(i48 inreg %3)
@@ -68,7 +68,7 @@ entry:
   %s2_1.coerce = alloca { i40 }
   %0 = bitcast { i40 }* %s2_1.coerce to i8*
   %1 = bitcast %struct.S2* %s2_1 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %0, i8* %1, i64 5, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %0, i8* %1, i64 5, i32 0, i1 false)
   %2 = getelementptr { i40 }, { i40 }* %s2_1.coerce, i32 0, i32 0
   %3 = load i40, i40* %2, align 1
   call void @fS2(i40 inreg %3)
diff --git a/test/CodeGen/Mips/cconv/arguments-varargs-small-structs-byte.ll b/test/CodeGen/Mips/cconv/arguments-varargs-small-structs-byte.ll
index f607e119990..ba3aeb598f5 100644
--- a/test/CodeGen/Mips/cconv/arguments-varargs-small-structs-byte.ll
+++ b/test/CodeGen/Mips/cconv/arguments-varargs-small-structs-byte.ll
@@ -172,7 +172,7 @@ entry:
   %0 = load %struct.SmallStruct_3b*, %struct.SmallStruct_3b** %ss.addr, align 8
   %1 = bitcast { i24 }* %.coerce to i8*
   %2 = bitcast %struct.SmallStruct_3b* %0 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %1, i8* %2, i64 3, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %1, i8* %2, i64 3, i32 0, i1 false)
   %3 = getelementptr { i24 }, { i24 }* %.coerce, i32 0, i32 0
   %4 = load i24, i24* %3, align 1
   call void (i8*, ...) @varArgF_SmallStruct(i8* getelementptr inbounds ([3 x i8], [3 x i8]* @.str, i32 0, i32 0), i24 inreg %4)
@@ -181,7 +181,7 @@ entry:
  ; CHECK: dsll $[[R1:[0-9]+]], $[[R2:[0-9]+]], 40
 }
 
-declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture readonly, i64, i1) #1
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture readonly, i64, i32, i1) #1
 
 define void @smallStruct_4b(%struct.SmallStruct_4b* %ss) #0 {
 entry:
@@ -205,7 +205,7 @@ entry:
   %0 = load %struct.SmallStruct_5b*, %struct.SmallStruct_5b** %ss.addr, align 8
   %1 = bitcast { i40 }* %.coerce to i8*
   %2 = bitcast %struct.SmallStruct_5b* %0 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %1, i8* %2, i64 5, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %1, i8* %2, i64 5, i32 0, i1 false)
   %3 = getelementptr { i40 }, { i40 }* %.coerce, i32 0, i32 0
   %4 = load i40, i40* %3, align 1
   call void (i8*, ...) @varArgF_SmallStruct(i8* getelementptr inbounds ([3 x i8], [3 x i8]* @.str, i32 0, i32 0), i40 inreg %4)
@@ -222,7 +222,7 @@ entry:
   %0 = load %struct.SmallStruct_6b*, %struct.SmallStruct_6b** %ss.addr, align 8
   %1 = bitcast { i48 }* %.coerce to i8*
   %2 = bitcast %struct.SmallStruct_6b* %0 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %1, i8* %2, i64 6, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %1, i8* %2, i64 6, i32 0, i1 false)
   %3 = getelementptr { i48 }, { i48 }* %.coerce, i32 0, i32 0
   %4 = load i48, i48* %3, align 1
   call void (i8*, ...) @varArgF_SmallStruct(i8* getelementptr inbounds ([3 x i8], [3 x i8]* @.str, i32 0, i32 0), i48 inreg %4)
@@ -239,7 +239,7 @@ entry:
   %0 = load %struct.SmallStruct_7b*, %struct.SmallStruct_7b** %ss.addr, align 8
   %1 = bitcast { i56 }* %.coerce to i8*
   %2 = bitcast %struct.SmallStruct_7b* %0 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %1, i8* %2, i64 7, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %1, i8* %2, i64 7, i32 0, i1 false)
   %3 = getelementptr { i56 }, { i56 }* %.coerce, i32 0, i32 0
   %4 = load i56, i56* %3, align 1
   call void (i8*, ...) @varArgF_SmallStruct(i8* getelementptr inbounds ([3 x i8], [3 x i8]* @.str, i32 0, i32 0), i56 inreg %4)
@@ -270,7 +270,7 @@ entry:
   %0 = load %struct.SmallStruct_9b*, %struct.SmallStruct_9b** %ss.addr, align 8
   %1 = bitcast { i64, i8 }* %.coerce to i8*
   %2 = bitcast %struct.SmallStruct_9b* %0 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %1, i8* %2, i64 9, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %1, i8* %2, i64 9, i32 0, i1 false)
   %3 = getelementptr { i64, i8 }, { i64, i8 }* %.coerce, i32 0, i32 0
   %4 = load i64, i64* %3, align 1
   %5 = getelementptr { i64, i8 }, { i64, i8 }* %.coerce, i32 0, i32 1
diff --git a/test/CodeGen/Mips/cconv/arguments-varargs-small-structs-combinations.ll b/test/CodeGen/Mips/cconv/arguments-varargs-small-structs-combinations.ll
index f01546457f4..74d3d859ed7 100644
--- a/test/CodeGen/Mips/cconv/arguments-varargs-small-structs-combinations.ll
+++ b/test/CodeGen/Mips/cconv/arguments-varargs-small-structs-combinations.ll
@@ -106,7 +106,7 @@ entry:
   %0 = load %struct.SmallStruct_1b1s1b*, %struct.SmallStruct_1b1s1b** %ss.addr, align 8
   %1 = bitcast { i48 }* %.coerce to i8*
   %2 = bitcast %struct.SmallStruct_1b1s1b* %0 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %1, i8* %2, i64 6, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %1, i8* %2, i64 6, i32 0, i1 false)
   %3 = getelementptr { i48 }, { i48 }* %.coerce, i32 0, i32 0
   %4 = load i48, i48* %3, align 1
   call void (i8*, ...) @varArgF_SmallStruct(i8* getelementptr inbounds ([3 x i8], [3 x i8]* @.str, i32 0, i32 0), i48 inreg %4)
@@ -115,7 +115,7 @@ entry:
  ; CHECK: dsll $[[R1:[0-9]+]], $[[R2:[0-9]+]], 16
 }
 
-declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture readonly, i64, i1) #1
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture readonly, i64, i32, i1) #1
 
 define void @smallStruct_1s1i(%struct.SmallStruct_1s1i* %ss) #0 {
 entry:
@@ -139,7 +139,7 @@ entry:
   %0 = load %struct.SmallStruct_3b1s*, %struct.SmallStruct_3b1s** %ss.addr, align 8
   %1 = bitcast { i48 }* %.coerce to i8*
   %2 = bitcast %struct.SmallStruct_3b1s* %0 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %1, i8* %2, i64 6, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %1, i8* %2, i64 6, i32 0, i1 false)
   %3 = getelementptr { i48 }, { i48 }* %.coerce, i32 0, i32 0
   %4 = load i48, i48* %3, align 1
   call void (i8*, ...) @varArgF_SmallStruct(i8* getelementptr inbounds ([3 x i8], [3 x i8]* @.str, i32 0, i32 0), i48 inreg %4)
diff --git a/test/CodeGen/Mips/cconv/return-struct.ll b/test/CodeGen/Mips/cconv/return-struct.ll
index 1919c7dd000..8decd04f089 100644
--- a/test/CodeGen/Mips/cconv/return-struct.ll
+++ b/test/CodeGen/Mips/cconv/return-struct.ll
@@ -18,7 +18,7 @@
 @struct_6xi32 = global {[6 x i32]} zeroinitializer
 @struct_128xi16 = global {[128 x i16]} zeroinitializer
 
-declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture readonly, i64, i1)
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture readonly, i64, i32, i1)
 
 define inreg {i8} @ret_struct_i8() nounwind {
 entry:
@@ -52,7 +52,7 @@ define inreg {i16} @ret_struct_i16() nounwind {
 entry:
         %retval = alloca {i8,i8}, align 1
         %0 = bitcast {i8,i8}* %retval to i8*
-        call void @llvm.memcpy.p0i8.p0i8.i64(i8* %0, i8* getelementptr inbounds ({i8,i8}, {i8,i8}* @struct_2byte, i32 0, i32 0), i64 2, i1 false)
+        call void @llvm.memcpy.p0i8.p0i8.i64(i8* %0, i8* getelementptr inbounds ({i8,i8}, {i8,i8}* @struct_2byte, i32 0, i32 0), i64 2, i32 1, i1 false)
         %1 = bitcast {i8,i8}* %retval to {i16}*
         %2 = load volatile {i16}, {i16}* %1
         ret {i16} %2
@@ -144,7 +144,7 @@ entry:
 define void @ret_struct_128xi16({[128 x i16]}* sret %returnval) {
 entry:
         %0 = bitcast {[128 x i16]}* %returnval to i8*
-        call void @llvm.memcpy.p0i8.p0i8.i64(i8* %0, i8* bitcast ({[128 x i16]}* @struct_128xi16 to i8*), i64 256, i1 false)
+        call void @llvm.memcpy.p0i8.p0i8.i64(i8* %0, i8* bitcast ({[128 x i16]}* @struct_128xi16 to i8*), i64 256, i32 2, i1 false)
         ret void
 }
 
diff --git a/test/CodeGen/Mips/largeimmprinting.ll b/test/CodeGen/Mips/largeimmprinting.ll
index bafce23951d..a53a953a788 100644
--- a/test/CodeGen/Mips/largeimmprinting.ll
+++ b/test/CodeGen/Mips/largeimmprinting.ll
@@ -28,11 +28,11 @@ entry:
 
   %agg.tmp = alloca %struct.S1, align 1
   %tmp = getelementptr inbounds %struct.S1, %struct.S1* %agg.tmp, i32 0, i32 0, i32 0
-  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %tmp, i8* getelementptr inbounds (%struct.S1, %struct.S1* @s1, i32 0, i32 0, i32 0), i32 65536, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %tmp, i8* getelementptr inbounds (%struct.S1, %struct.S1* @s1, i32 0, i32 0, i32 0), i32 65536, i32 1, i1 false)
   call void @f2(%struct.S1* byval %agg.tmp) nounwind
   ret void
 }
 
 declare void @f2(%struct.S1* byval)
 
-declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i32, i1) nounwind
diff --git a/test/CodeGen/Mips/memcpy.ll b/test/CodeGen/Mips/memcpy.ll
index 21661af067e..5c4ebb27dde 100644
--- a/test/CodeGen/Mips/memcpy.ll
+++ b/test/CodeGen/Mips/memcpy.ll
@@ -9,11 +9,11 @@ entry:
 ; CHECK-NOT: call16(memcpy
 
   %arraydecay = getelementptr inbounds %struct.S1, %struct.S1* %s1, i32 0, i32 1, i32 0
-  tail call void @llvm.memcpy.p0i8.p0i8.i32(i8* %arraydecay, i8* getelementptr inbounds ([31 x i8], [31 x i8]* @.str, i32 0, i32 0), i32 31, i1 false)
+  tail call void @llvm.memcpy.p0i8.p0i8.i32(i8* %arraydecay, i8* getelementptr inbounds ([31 x i8], [31 x i8]* @.str, i32 0, i32 0), i32 31, i32 1, i1 false)
   %arrayidx = getelementptr inbounds %struct.S1, %struct.S1* %s1, i32 0, i32 1, i32 40
   store i8 %n, i8* %arrayidx, align 1
   ret void
 }
 
-declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i32, i1) nounwind
 
diff --git a/test/CodeGen/Mips/tailcall.ll b/test/CodeGen/Mips/tailcall.ll
index f7674a6e220..b0868255053 100644
--- a/test/CodeGen/Mips/tailcall.ll
+++ b/test/CodeGen/Mips/tailcall.ll
@@ -207,7 +207,7 @@ entry:
 
 declare i32 @callee12()
 
-declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i32, i1) nounwind
 
 define i32 @caller12(%struct.S* nocapture byval %a0) nounwind {
 entry:
@@ -221,7 +221,7 @@ entry:
 ; PIC16: jalrc
 
   %0 = bitcast %struct.S* %a0 to i8*
-  tail call void @llvm.memcpy.p0i8.p0i8.i32(i8* bitcast (%struct.S* @gs1 to i8*), i8* %0, i32 8, i1 false)
+  tail call void @llvm.memcpy.p0i8.p0i8.i32(i8* bitcast (%struct.S* @gs1 to i8*), i8* %0, i32 8, i32 4, i1 false)
   %call = tail call i32 @callee12() nounwind
   ret i32 %call
 }
diff --git a/test/CodeGen/NVPTX/lower-aggr-copies.ll b/test/CodeGen/NVPTX/lower-aggr-copies.ll
index 7c22a3e0cd9..ef570982b80 100644
--- a/test/CodeGen/NVPTX/lower-aggr-copies.ll
+++ b/test/CodeGen/NVPTX/lower-aggr-copies.ll
@@ -7,13 +7,13 @@
 target datalayout = "e-m:e-i64:64-f80:128-n8:16:32:64-S128"
 target triple = "nvptx64-unknown-unknown"
 
-declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture readonly, i64, i1) #1
-declare void @llvm.memmove.p0i8.p0i8.i64(i8* nocapture, i8* nocapture readonly, i64, i1) #1
-declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i1) #1
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture readonly, i64, i32, i1) #1
+declare void @llvm.memmove.p0i8.p0i8.i64(i8* nocapture, i8* nocapture readonly, i64, i32, i1) #1
+declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i32, i1) #1
 
 define i8* @memcpy_caller(i8* %dst, i8* %src, i64 %n) #0 {
 entry:
-  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %dst, i8* %src, i64 %n, i1 false)
+  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %dst, i8* %src, i64 %n, i32 1, i1 false)
   ret i8* %dst
 
 ; IR-LABEL:   @memcpy_caller
@@ -34,7 +34,7 @@ entry:
 
 define i8* @memcpy_volatile_caller(i8* %dst, i8* %src, i64 %n) #0 {
 entry:
-  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %dst, i8* %src, i64 %n, i1 true)
+  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %dst, i8* %src, i64 %n, i32 1, i1 true)
   ret i8* %dst
 
 ; IR-LABEL:   @memcpy_volatile_caller
@@ -54,7 +54,7 @@ define i8* @memcpy_casting_caller(i32* %dst, i32* %src, i64 %n) #0 {
 entry:
   %0 = bitcast i32* %dst to i8*
   %1 = bitcast i32* %src to i8*
-  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %0, i8* %1, i64 %n, i1 false)
+  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %0, i8* %1, i64 %n, i32 1, i1 false)
   ret i8* %0
 
 ; Check that casts in calls to memcpy are handled properly
@@ -68,7 +68,7 @@ entry:
 define i8* @memset_caller(i8* %dst, i32 %c, i64 %n) #0 {
 entry:
   %0 = trunc i32 %c to i8
-  tail call void @llvm.memset.p0i8.i64(i8* %dst, i8 %0, i64 %n, i1 false)
+  tail call void @llvm.memset.p0i8.i64(i8* %dst, i8 %0, i64 %n, i32 1, i1 false)
   ret i8* %dst
 
 ; IR-LABEL:   @memset_caller
@@ -88,7 +88,7 @@ entry:
 
 define i8* @memmove_caller(i8* %dst, i8* %src, i64 %n) #0 {
 entry:
-  tail call void @llvm.memmove.p0i8.p0i8.i64(i8* %dst, i8* %src, i64 %n, i1 false)
+  tail call void @llvm.memmove.p0i8.p0i8.i64(i8* %dst, i8* %src, i64 %n, i32 1, i1 false)
   ret i8* %dst
 
 ; IR-LABEL:   @memmove_caller
diff --git a/test/CodeGen/PowerPC/2011-12-05-NoSpillDupCR.ll b/test/CodeGen/PowerPC/2011-12-05-NoSpillDupCR.ll
index 2a6ddd5db10..e5920911ee2 100644
--- a/test/CodeGen/PowerPC/2011-12-05-NoSpillDupCR.ll
+++ b/test/CodeGen/PowerPC/2011-12-05-NoSpillDupCR.ll
@@ -179,7 +179,7 @@ for.end.7:                                        ; preds = %entry, %for.end.7
   br i1 %exitcond.7, label %for.end12, label %for.end.7
 }
 
-declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i32, i1) nounwind
 
 declare i32 @puts(i8* nocapture) nounwind
 
diff --git a/test/CodeGen/PowerPC/2011-12-06-SpillAndRestoreCR.ll b/test/CodeGen/PowerPC/2011-12-06-SpillAndRestoreCR.ll
index 3271466cbcc..93476827949 100644
--- a/test/CodeGen/PowerPC/2011-12-06-SpillAndRestoreCR.ll
+++ b/test/CodeGen/PowerPC/2011-12-06-SpillAndRestoreCR.ll
@@ -213,7 +213,7 @@ for.end23:                                        ; preds = %for.end17
   ret i32 0
 }
 
-declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i32, i1) nounwind
 
 declare i32 @puts(i8* nocapture) nounwind
 
diff --git a/test/CodeGen/PowerPC/ctrloop-reg.ll b/test/CodeGen/PowerPC/ctrloop-reg.ll
index 7f4946fcc12..874e5712bb9 100644
--- a/test/CodeGen/PowerPC/ctrloop-reg.ll
+++ b/test/CodeGen/PowerPC/ctrloop-reg.ll
@@ -74,7 +74,7 @@ declare i32 @interp(%struct.ref_s.1.49.91.115.121.139.145.151.157.163.169.175.18
 
 declare i32 @dict_lookup(%struct.ref_s.1.49.91.115.121.139.145.151.157.163.169.175.181.211*, %struct.ref_s.1.49.91.115.121.139.145.151.157.163.169.175.181.211*, %struct.ref_s.1.49.91.115.121.139.145.151.157.163.169.175.181.211*, %struct.ref_s.1.49.91.115.121.139.145.151.157.163.169.175.181.211**)
 
-declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i32, i1) nounwind
 
 declare i32 @obj_compare(...)
 
diff --git a/test/CodeGen/PowerPC/emptystruct.ll b/test/CodeGen/PowerPC/emptystruct.ll
index 49a9c13c7cf..66cada14f9d 100644
--- a/test/CodeGen/PowerPC/emptystruct.ll
+++ b/test/CodeGen/PowerPC/emptystruct.ll
@@ -21,7 +21,7 @@ entry:
   %0 = load %struct.empty*, %struct.empty** %a2.addr, align 8
   %1 = bitcast %struct.empty* %agg.result to i8*
   %2 = bitcast %struct.empty* %0 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %1, i8* %2, i64 0, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %1, i8* %2, i64 0, i32 1, i1 false)
   ret void
 }
 
@@ -31,7 +31,7 @@ entry:
 ; CHECK-NOT: std 6,
 ; CHECK: blr
 
-declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i32, i1) nounwind
 
 define void @caller(%struct.empty* noalias sret %agg.result) nounwind {
 entry:
diff --git a/test/CodeGen/PowerPC/fsl-e500mc.ll b/test/CodeGen/PowerPC/fsl-e500mc.ll
index 26d9a09b8f7..09b7e41b189 100644
--- a/test/CodeGen/PowerPC/fsl-e500mc.ll
+++ b/test/CodeGen/PowerPC/fsl-e500mc.ll
@@ -15,8 +15,8 @@ entry:
 ; CHECK-NOT: bl memcpy
   %0 = bitcast %struct.teststruct* %agg.result to i8*
   %1 = bitcast %struct.teststruct* %in to i8*
-  tail call void @llvm.memcpy.p0i8.p0i8.i32(i8* %0, i8* %1, i32 52, i1 false)
+  tail call void @llvm.memcpy.p0i8.p0i8.i32(i8* %0, i8* %1, i32 52, i32 4, i1 false)
   ret void
 }
 
-declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i32, i1) nounwind
diff --git a/test/CodeGen/PowerPC/fsl-e5500.ll b/test/CodeGen/PowerPC/fsl-e5500.ll
index 664a449eb1b..d47d8c8ed4f 100644
--- a/test/CodeGen/PowerPC/fsl-e5500.ll
+++ b/test/CodeGen/PowerPC/fsl-e5500.ll
@@ -15,8 +15,8 @@ entry:
 ; CHECK-NOT: bl memcpy
   %0 = bitcast %struct.teststruct* %agg.result to i8*
   %1 = bitcast %struct.teststruct* %in to i8*
-  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %0, i8* %1, i64 100, i1 false)
+  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %0, i8* %1, i64 100, i32 4, i1 false)
   ret void
 }
 
-declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i32, i1) nounwind
diff --git a/test/CodeGen/PowerPC/glob-comp-aa-crash.ll b/test/CodeGen/PowerPC/glob-comp-aa-crash.ll
index 0d5bd846239..88530a3f303 100644
--- a/test/CodeGen/PowerPC/glob-comp-aa-crash.ll
+++ b/test/CodeGen/PowerPC/glob-comp-aa-crash.ll
@@ -35,7 +35,7 @@ entry:
 invoke.cont:                                      ; preds = %entry
   %__exception_ = getelementptr inbounds %"class.std::__1::__assoc_sub_state", %"class.std::__1::__assoc_sub_state"* %this, i64 0, i32 1
   %0 = bitcast { i64, i64 }* %tmp to i8*
-  call void @llvm.memset.p0i8.i64(i8* %0, i8 0, i64 16, i1 false)
+  call void @llvm.memset.p0i8.i64(i8* %0, i8 0, i64 16, i32 8, i1 false)
   call void @_ZNSt15__exception_ptr13exception_ptrC1EMS0_FvvE(%"class.std::__exception_ptr::exception_ptr"* %ref.tmp, { i64, i64 }* byval %tmp) #5
   %call = call zeroext i1 @_ZNSt15__exception_ptrneERKNS_13exception_ptrES2_(%"class.std::__exception_ptr::exception_ptr"* %__exception_, %"class.std::__exception_ptr::exception_ptr"* %ref.tmp) #5
   call void @_ZNSt15__exception_ptr13exception_ptrD1Ev(%"class.std::__exception_ptr::exception_ptr"* %ref.tmp) #5
@@ -120,7 +120,7 @@ declare void @_ZNSt3__15mutex6unlockEv(%"class.std::__1::mutex"*) #1
 declare void @_ZNSt3__15mutex4lockEv(%"class.std::__1::mutex"*) #0
 
 ; Function Attrs: nounwind
-declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i1) #3
+declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i32, i1) #3
 
 attributes #0 = { optsize "less-precise-fpmad"="false" "no-frame-pointer-elim"="true" "no-frame-pointer-elim-non-leaf" "no-infs-fp-math"="false" "no-nans-fp-math"="false" "stack-protector-buffer-size"="8" "unsafe-fp-math"="false" "use-soft-float"="false" }
 attributes #1 = { nounwind optsize "less-precise-fpmad"="false" "no-frame-pointer-elim"="true" "no-frame-pointer-elim-non-leaf" "no-infs-fp-math"="false" "no-nans-fp-math"="false" "stack-protector-buffer-size"="8" "unsafe-fp-math"="false" "use-soft-float"="false" }
diff --git a/test/CodeGen/PowerPC/isel-rc-nox0.ll b/test/CodeGen/PowerPC/isel-rc-nox0.ll
index 93d1b6bc1d8..7475e12c447 100644
--- a/test/CodeGen/PowerPC/isel-rc-nox0.ll
+++ b/test/CodeGen/PowerPC/isel-rc-nox0.ll
@@ -21,7 +21,7 @@ crc32_gentab.exit:                                ; preds = %for.cond1.preheader
   br label %for.cond1.preheader.i2961.i
 
 for.cond1.preheader.i2961.i:                      ; preds = %for.inc44.i2977.i, %crc32_gentab.exit
-  call void @llvm.memset.p0i8.i64(i8* bitcast ([1 x [9 x i32]]* @g_62 to i8*), i8 -1, i64 36, i1 false) #1
+  call void @llvm.memset.p0i8.i64(i8* bitcast ([1 x [9 x i32]]* @g_62 to i8*), i8 -1, i64 36, i32 4, i1 false) #1
   %0 = load i32, i32* %retval.0.i.i.i, align 4
   %tobool.i2967.i = icmp eq i32 %0, 0
   br label %for.body21.i2968.i
@@ -40,7 +40,7 @@ func_80.exit2978.i:                               ; preds = %for.inc44.i2977.i
 }
 
 ; Function Attrs: nounwind
-declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i1) #1
+declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i32, i1) #1
 
 attributes #0 = { nounwind "less-precise-fpmad"="false" "no-frame-pointer-elim"="true" "no-frame-pointer-elim-non-leaf" "no-infs-fp-math"="false" "no-nans-fp-math"="false" "ssp-buffer-size"="8" "unsafe-fp-math"="false" "use-soft-float"="false" }
 attributes #1 = { nounwind }
diff --git a/test/CodeGen/PowerPC/memcpy-vec.ll b/test/CodeGen/PowerPC/memcpy-vec.ll
index 27aed429d1b..29baef55ce1 100644
--- a/test/CodeGen/PowerPC/memcpy-vec.ll
+++ b/test/CodeGen/PowerPC/memcpy-vec.ll
@@ -9,7 +9,7 @@ define void @foo1(double* nocapture %x, double* nocapture readonly %y) #0 {
 entry:
   %0 = bitcast double* %x to i8*
   %1 = bitcast double* %y to i8*
-  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 8 %0, i8* align 8 %1, i64 32, i1 false)
+  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %0, i8* %1, i64 32, i32 8, i1 false)
   ret void
 
 ; PWR7-LABEL: @foo1
@@ -34,14 +34,14 @@ entry:
 }
 
 ; Function Attrs: nounwind
-declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture readonly, i64, i1) #0
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture readonly, i64, i32, i1) #0
 
 ; Function Attrs: nounwind
 define void @foo2(double* nocapture %x, double* nocapture readonly %y) #0 {
 entry:
   %0 = bitcast double* %x to i8*
   %1 = bitcast double* %y to i8*
-  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 8 %0, i8* align 8 %1, i64 128, i1 false)
+  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %0, i8* %1, i64 128, i32 8, i1 false)
   ret void
 
 ; PWR7-LABEL: @foo2
@@ -64,7 +64,7 @@ entry:
 define void @bar1(double* nocapture %x) #0 {
 entry:
   %0 = bitcast double* %x to i8*
-  tail call void @llvm.memset.p0i8.i64(i8* align 8 %0, i8 0, i64 128, i1 false)
+  tail call void @llvm.memset.p0i8.i64(i8* %0, i8 0, i64 128, i32 8, i1 false)
   ret void
 
 ; PWR7-LABEL: @bar1
@@ -87,7 +87,7 @@ entry:
 define void @bar2(double* nocapture %x) #0 {
 entry:
   %0 = bitcast double* %x to i8*
-  tail call void @llvm.memset.p0i8.i64(i8* align 32 %0, i8 0, i64 128, i1 false)
+  tail call void @llvm.memset.p0i8.i64(i8* %0, i8 0, i64 128, i32 32, i1 false)
   ret void
 
 ; PWR7-LABEL: @bar2
@@ -107,7 +107,7 @@ entry:
 }
 
 ; Function Attrs: nounwind
-declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i1) #0
+declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i32, i1) #0
 
 attributes #0 = { nounwind }
 
diff --git a/test/CodeGen/PowerPC/memset-nc-le.ll b/test/CodeGen/PowerPC/memset-nc-le.ll
index fa31af50a07..af8e9c3fb4f 100644
--- a/test/CodeGen/PowerPC/memset-nc-le.ll
+++ b/test/CodeGen/PowerPC/memset-nc-le.ll
@@ -7,7 +7,7 @@ define void @test_vsx() unnamed_addr #0 align 2 {
 entry:
   %0 = load i32, i32* undef, align 4
   %1 = trunc i32 %0 to i8
-  call void @llvm.memset.p0i8.i64(i8* null, i8 %1, i64 32, i1 false)
+  call void @llvm.memset.p0i8.i64(i8* null, i8 %1, i64 32, i32 1, i1 false)
   ret void
 
 ; CHECK-LABEL: @test_vsx
@@ -17,7 +17,7 @@ entry:
 }
 
 ; Function Attrs: nounwind
-declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i1) #1
+declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i32, i1) #1
 
 attributes #0 = { nounwind "target-cpu"="pwr8" }
 attributes #1 = { nounwind }
diff --git a/test/CodeGen/PowerPC/memset-nc.ll b/test/CodeGen/PowerPC/memset-nc.ll
index 041d26f0dbf..414a987a56f 100644
--- a/test/CodeGen/PowerPC/memset-nc.ll
+++ b/test/CodeGen/PowerPC/memset-nc.ll
@@ -8,7 +8,7 @@ define void @test_qpx() unnamed_addr #0 align 2 {
 entry:
   %0 = load i32, i32* undef, align 4
   %1 = trunc i32 %0 to i8
-  call void @llvm.memset.p0i8.i64(i8* align 32 null, i8 %1, i64 64, i1 false)
+  call void @llvm.memset.p0i8.i64(i8* null, i8 %1, i64 64, i32 32, i1 false)
   ret void
 
 ; CHECK-LABEL: @test_qpx
@@ -22,14 +22,14 @@ entry:
 }
 
 ; Function Attrs: nounwind
-declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i1) #1
+declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i32, i1) #1
 
 ; Function Attrs: nounwind
 define void @test_vsx() unnamed_addr #2 align 2 {
 entry:
   %0 = load i32, i32* undef, align 4
   %1 = trunc i32 %0 to i8
-  call void @llvm.memset.p0i8.i64(i8* align 1 null, i8 %1, i64 32, i1 false)
+  call void @llvm.memset.p0i8.i64(i8* null, i8 %1, i64 32, i32 1, i1 false)
   ret void
 
 ; CHECK-LABEL: @test_vsx
diff --git a/test/CodeGen/PowerPC/ppc-empty-fs.ll b/test/CodeGen/PowerPC/ppc-empty-fs.ll
index 9c0af17009b..8c0a2fb1dc0 100644
--- a/test/CodeGen/PowerPC/ppc-empty-fs.ll
+++ b/test/CodeGen/PowerPC/ppc-empty-fs.ll
@@ -15,14 +15,14 @@ entry:
   store i64 %x.coerce, i64* %0, align 1
   %1 = bitcast %struct.fab* %agg.result to i8*
   %2 = bitcast %struct.fab* %x to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %1, i8* %2, i64 8, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %1, i8* %2, i64 8, i32 4, i1 false)
   ret void
 }
 
 ; CHECK: func_fab
 
 ; Function Attrs: nounwind
-declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture readonly, i64, i1) #1
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture readonly, i64, i32, i1) #1
 
 attributes #0 = { nounwind "less-precise-fpmad"="false" "no-frame-pointer-elim"="false" "no-infs-fp-math"="false" "no-nans-fp-math"="false" "no-realign-stack" "stack-protector-buffer-size"="8" "target-features"="" "unsafe-fp-math"="false" "use-soft-float"="false" }
 attributes #1 = { nounwind }
diff --git a/test/CodeGen/PowerPC/resolvefi-basereg.ll b/test/CodeGen/PowerPC/resolvefi-basereg.ll
index c3e63d5342f..a613c3310a5 100644
--- a/test/CodeGen/PowerPC/resolvefi-basereg.ll
+++ b/test/CodeGen/PowerPC/resolvefi-basereg.ll
@@ -32,9 +32,9 @@ entry:
   %agg.tmp117 = alloca %struct.S1998, align 16
   %agg.tmp118 = alloca %struct.S1998, align 16
   %agg.tmp119 = alloca %struct.S1998, align 16
-  call void @llvm.memset.p0i8.i64(i8* bitcast (%struct.S1998* @s1998 to i8*), i8 0, i64 5168, i1 false)
-  call void @llvm.memset.p0i8.i64(i8* bitcast ([5 x %struct.S1998]* @a1998 to i8*), i8 0, i64 25840, i1 false)
-  call void @llvm.memset.p0i8.i64(i8* bitcast (%struct.Info* @info to i8*), i8 0, i64 832, i1 false)
+  call void @llvm.memset.p0i8.i64(i8* bitcast (%struct.S1998* @s1998 to i8*), i8 0, i64 5168, i32 16, i1 false)
+  call void @llvm.memset.p0i8.i64(i8* bitcast ([5 x %struct.S1998]* @a1998 to i8*), i8 0, i64 25840, i32 16, i1 false)
+  call void @llvm.memset.p0i8.i64(i8* bitcast (%struct.Info* @info to i8*), i8 0, i64 832, i32 8, i1 false)
   store i8* bitcast (%struct.S1998* @s1998 to i8*), i8** getelementptr inbounds (%struct.Info, %struct.Info* @info, i32 0, i32 2), align 8
   store i8* bitcast ([5 x %struct.S1998]* @a1998 to i8*), i8** getelementptr inbounds (%struct.Info, %struct.Info* @info, i32 0, i32 3), align 8
   store i8* bitcast (%struct.S1998* getelementptr inbounds ([5 x %struct.S1998], [5 x %struct.S1998]* @a1998, i32 0, i64 3) to i8*), i8** getelementptr inbounds (%struct.Info, %struct.Info* @info, i32 0, i32 4), align 8
@@ -329,32 +329,32 @@ if.end:                                           ; preds = %if.then, %entry
   %61 = load i32, i32* %j, align 4
   store i32 %61, i32* getelementptr inbounds (%struct.Info, %struct.Info* @info, i32 0, i32 1), align 4
   %62 = bitcast %struct.S1998* %agg.tmp111 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %62, i8* bitcast (%struct.S1998* @s1998 to i8*), i64 5168, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %62, i8* bitcast (%struct.S1998* @s1998 to i8*), i64 5168, i32 16, i1 false)
   %63 = bitcast %struct.S1998* %agg.tmp112 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %63, i8* bitcast (%struct.S1998* getelementptr inbounds ([5 x %struct.S1998], [5 x %struct.S1998]* @a1998, i32 0, i64 2) to i8*), i64 5168, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %63, i8* bitcast (%struct.S1998* getelementptr inbounds ([5 x %struct.S1998], [5 x %struct.S1998]* @a1998, i32 0, i64 2) to i8*), i64 5168, i32 16, i1 false)
   call void @check1998(%struct.S1998* sret %agg.tmp, %struct.S1998* byval align 16 %agg.tmp111, %struct.S1998* getelementptr inbounds ([5 x %struct.S1998], [5 x %struct.S1998]* @a1998, i32 0, i64 1), %struct.S1998* byval align 16 %agg.tmp112)
   call void @checkx1998(%struct.S1998* byval align 16 %agg.tmp)
   %64 = bitcast %struct.S1998* %agg.tmp113 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %64, i8* bitcast (%struct.S1998* @s1998 to i8*), i64 5168, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %64, i8* bitcast (%struct.S1998* @s1998 to i8*), i64 5168, i32 16, i1 false)
   %65 = bitcast %struct.S1998* %agg.tmp114 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %65, i8* bitcast (%struct.S1998* getelementptr inbounds ([5 x %struct.S1998], [5 x %struct.S1998]* @a1998, i32 0, i64 2) to i8*), i64 5168, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %65, i8* bitcast (%struct.S1998* getelementptr inbounds ([5 x %struct.S1998], [5 x %struct.S1998]* @a1998, i32 0, i64 2) to i8*), i64 5168, i32 16, i1 false)
   %66 = bitcast %struct.S1998* %agg.tmp115 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %66, i8* bitcast (%struct.S1998* getelementptr inbounds ([5 x %struct.S1998], [5 x %struct.S1998]* @a1998, i32 0, i64 2) to i8*), i64 5168, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %66, i8* bitcast (%struct.S1998* getelementptr inbounds ([5 x %struct.S1998], [5 x %struct.S1998]* @a1998, i32 0, i64 2) to i8*), i64 5168, i32 16, i1 false)
   call void (i32, ...) @check1998va(i32 signext 1, double 1.000000e+00, %struct.S1998* byval align 16 %agg.tmp113, i64 2, %struct.S1998* byval align 16 %agg.tmp114, %struct.S1998* byval align 16 %agg.tmp115)
   %67 = bitcast %struct.S1998* %agg.tmp116 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %67, i8* bitcast (%struct.S1998* @s1998 to i8*), i64 5168, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %67, i8* bitcast (%struct.S1998* @s1998 to i8*), i64 5168, i32 16, i1 false)
   %68 = bitcast %struct.S1998* %agg.tmp117 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %68, i8* bitcast (%struct.S1998* @s1998 to i8*), i64 5168, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %68, i8* bitcast (%struct.S1998* @s1998 to i8*), i64 5168, i32 16, i1 false)
   %69 = bitcast %struct.S1998* %agg.tmp118 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %69, i8* bitcast (%struct.S1998* getelementptr inbounds ([5 x %struct.S1998], [5 x %struct.S1998]* @a1998, i32 0, i64 2) to i8*), i64 5168, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %69, i8* bitcast (%struct.S1998* getelementptr inbounds ([5 x %struct.S1998], [5 x %struct.S1998]* @a1998, i32 0, i64 2) to i8*), i64 5168, i32 16, i1 false)
   %70 = bitcast %struct.S1998* %agg.tmp119 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %70, i8* bitcast (%struct.S1998* @s1998 to i8*), i64 5168, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %70, i8* bitcast (%struct.S1998* @s1998 to i8*), i64 5168, i32 16, i1 false)
   call void (i32, ...) @check1998va(i32 signext 2, %struct.S1998* byval align 16 %agg.tmp116, %struct.S1998* byval align 16 %agg.tmp117, ppc_fp128 0xM40000000000000000000000000000000, %struct.S1998* byval align 16 %agg.tmp118, %struct.S1998* byval align 16 %agg.tmp119)
   ret void
 }
 
-declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i1)
-declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture readonly, i64, i1)
+declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i32, i1)
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture readonly, i64, i32, i1)
 
 declare void @check1998(%struct.S1998* sret, %struct.S1998* byval align 16, %struct.S1998*, %struct.S1998* byval align 16)
 declare void @check1998va(i32 signext, ...)
diff --git a/test/CodeGen/PowerPC/resolvefi-disp.ll b/test/CodeGen/PowerPC/resolvefi-disp.ll
index da7025c121e..a1c2070a6f4 100644
--- a/test/CodeGen/PowerPC/resolvefi-disp.ll
+++ b/test/CodeGen/PowerPC/resolvefi-disp.ll
@@ -30,17 +30,17 @@ entry:
   %b2 = alloca %struct.S2760, align 32
   %2 = bitcast %struct.S2760* %arg0 to i8*
   %3 = bitcast %struct.S2760* %0 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %2, i8* %3, i64 11104, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %2, i8* %3, i64 11104, i32 16, i1 false)
   %4 = bitcast %struct.S2760* %arg2 to i8*
   %5 = bitcast %struct.S2760* %1 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %4, i8* %5, i64 11104, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %4, i8* %5, i64 11104, i32 16, i1 false)
   store %struct.S2760* %arg1, %struct.S2760** %arg1.addr, align 8
   %6 = bitcast %struct.S2760* %ret to i8*
-  call void @llvm.memset.p0i8.i64(i8* %6, i8 0, i64 11104, i1 false)
+  call void @llvm.memset.p0i8.i64(i8* %6, i8 0, i64 11104, i32 32, i1 false)
   %7 = bitcast %struct.S2760* %b1 to i8*
-  call void @llvm.memset.p0i8.i64(i8* %7, i8 0, i64 11104, i1 false)
+  call void @llvm.memset.p0i8.i64(i8* %7, i8 0, i64 11104, i32 32, i1 false)
   %8 = bitcast %struct.S2760* %b2 to i8*
-  call void @llvm.memset.p0i8.i64(i8* %8, i8 0, i64 11104, i1 false)
+  call void @llvm.memset.p0i8.i64(i8* %8, i8 0, i64 11104, i32 32, i1 false)
   %b = getelementptr inbounds %struct.S2760, %struct.S2760* %arg0, i32 0, i32 1
   %g = getelementptr inbounds %struct.anon, %struct.anon* %b, i32 0, i32 1
   %9 = load i64, i64* %g, align 8
@@ -61,11 +61,11 @@ if.end:                                           ; preds = %if.then, %entry
   store i64 %12, i64* %g4, align 8
   %13 = bitcast %struct.S2760* %agg.result to i8*
   %14 = bitcast %struct.S2760* %ret to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %13, i8* %14, i64 11104, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %13, i8* %14, i64 11104, i32 32, i1 false)
   ret void
 }
 
-declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture readonly, i64, i1)
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture readonly, i64, i32, i1)
 
-declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i1)
+declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i32, i1)
 
diff --git a/test/CodeGen/PowerPC/structsinmem.ll b/test/CodeGen/PowerPC/structsinmem.ll
index c240c6e2656..bfada4c6371 100644
--- a/test/CodeGen/PowerPC/structsinmem.ll
+++ b/test/CodeGen/PowerPC/structsinmem.ll
@@ -43,19 +43,19 @@ entry:
   %p6 = alloca %struct.s6, align 4
   %p7 = alloca %struct.s7, align 4
   %0 = bitcast %struct.s1* %p1 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %0, i8* getelementptr inbounds (%struct.s1, %struct.s1* @caller1.p1, i32 0, i32 0), i64 1, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %0, i8* getelementptr inbounds (%struct.s1, %struct.s1* @caller1.p1, i32 0, i32 0), i64 1, i32 1, i1 false)
   %1 = bitcast %struct.s2* %p2 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %1, i8* bitcast (%struct.s2* @caller1.p2 to i8*), i64 2, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %1, i8* bitcast (%struct.s2* @caller1.p2 to i8*), i64 2, i32 2, i1 false)
   %2 = bitcast %struct.s3* %p3 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %2, i8* bitcast ({ i16, i8, i8 }* @caller1.p3 to i8*), i64 4, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %2, i8* bitcast ({ i16, i8, i8 }* @caller1.p3 to i8*), i64 4, i32 2, i1 false)
   %3 = bitcast %struct.s4* %p4 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %3, i8* bitcast (%struct.s4* @caller1.p4 to i8*), i64 4, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %3, i8* bitcast (%struct.s4* @caller1.p4 to i8*), i64 4, i32 4, i1 false)
   %4 = bitcast %struct.s5* %p5 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %4, i8* bitcast ({ i32, i8, [3 x i8] }* @caller1.p5 to i8*), i64 8, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %4, i8* bitcast ({ i32, i8, [3 x i8] }* @caller1.p5 to i8*), i64 8, i32 4, i1 false)
   %5 = bitcast %struct.s6* %p6 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %5, i8* bitcast ({ i32, i16, [2 x i8] }* @caller1.p6 to i8*), i64 8, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %5, i8* bitcast ({ i32, i16, [2 x i8] }* @caller1.p6 to i8*), i64 8, i32 4, i1 false)
   %6 = bitcast %struct.s7* %p7 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %6, i8* bitcast ({ i32, i16, i8, i8 }* @caller1.p7 to i8*), i64 8, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %6, i8* bitcast ({ i32, i16, i8, i8 }* @caller1.p7 to i8*), i64 8, i32 4, i1 false)
   %call = call i32 @callee1(i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, %struct.s1* byval %p1, %struct.s2* byval %p2, %struct.s3* byval %p3, %struct.s4* byval %p4, %struct.s5* byval %p5, %struct.s6* byval %p6, %struct.s7* byval %p7)
   ret i32 %call
 
@@ -68,7 +68,7 @@ entry:
 ; CHECK: std {{[0-9]+}}, 160(1)
 }
 
-declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i32, i1) nounwind
 
 define internal i32 @callee1(i32 %z1, i32 %z2, i32 %z3, i32 %z4, i32 %z5, i32 %z6, i32 %z7, i32 %z8, %struct.s1* byval %v1, %struct.s2* byval %v2, %struct.s3* byval %v3, %struct.s4* byval %v4, %struct.s5* byval %v5, %struct.s6* byval %v6, %struct.s7* byval %v7) nounwind {
 entry:
@@ -132,19 +132,19 @@ entry:
   %p6 = alloca %struct.t6, align 1
   %p7 = alloca %struct.t7, align 1
   %0 = bitcast %struct.t1* %p1 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %0, i8* getelementptr inbounds (%struct.t1, %struct.t1* @caller2.p1, i32 0, i32 0), i64 1, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %0, i8* getelementptr inbounds (%struct.t1, %struct.t1* @caller2.p1, i32 0, i32 0), i64 1, i32 1, i1 false)
   %1 = bitcast %struct.t2* %p2 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %1, i8* bitcast ({ i16 }* @caller2.p2 to i8*), i64 2, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %1, i8* bitcast ({ i16 }* @caller2.p2 to i8*), i64 2, i32 1, i1 false)
   %2 = bitcast %struct.t3* %p3 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %2, i8* bitcast (%struct.t3* @caller2.p3 to i8*), i64 3, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %2, i8* bitcast (%struct.t3* @caller2.p3 to i8*), i64 3, i32 1, i1 false)
   %3 = bitcast %struct.t4* %p4 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %3, i8* bitcast ({ i32 }* @caller2.p4 to i8*), i64 4, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %3, i8* bitcast ({ i32 }* @caller2.p4 to i8*), i64 4, i32 1, i1 false)
   %4 = bitcast %struct.t5* %p5 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %4, i8* bitcast (%struct.t5* @caller2.p5 to i8*), i64 5, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %4, i8* bitcast (%struct.t5* @caller2.p5 to i8*), i64 5, i32 1, i1 false)
   %5 = bitcast %struct.t6* %p6 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %5, i8* bitcast (%struct.t6* @caller2.p6 to i8*), i64 6, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %5, i8* bitcast (%struct.t6* @caller2.p6 to i8*), i64 6, i32 1, i1 false)
   %6 = bitcast %struct.t7* %p7 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %6, i8* bitcast (%struct.t7* @caller2.p7 to i8*), i64 7, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %6, i8* bitcast (%struct.t7* @caller2.p7 to i8*), i64 7, i32 1, i1 false)
   %call = call i32 @callee2(i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, %struct.t1* byval %p1, %struct.t2* byval %p2, %struct.t3* byval %p3, %struct.t4* byval %p4, %struct.t5* byval %p5, %struct.t6* byval %p6, %struct.t7* byval %p7)
   ret i32 %call
 
diff --git a/test/CodeGen/PowerPC/structsinregs.ll b/test/CodeGen/PowerPC/structsinregs.ll
index 006d06cebed..0fb9895a622 100644
--- a/test/CodeGen/PowerPC/structsinregs.ll
+++ b/test/CodeGen/PowerPC/structsinregs.ll
@@ -43,19 +43,19 @@ entry:
   %p6 = alloca %struct.s6, align 4
   %p7 = alloca %struct.s7, align 4
   %0 = bitcast %struct.s1* %p1 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %0, i8* getelementptr inbounds (%struct.s1, %struct.s1* @caller1.p1, i32 0, i32 0), i64 1, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %0, i8* getelementptr inbounds (%struct.s1, %struct.s1* @caller1.p1, i32 0, i32 0), i64 1, i32 1, i1 false)
   %1 = bitcast %struct.s2* %p2 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %1, i8* bitcast (%struct.s2* @caller1.p2 to i8*), i64 2, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %1, i8* bitcast (%struct.s2* @caller1.p2 to i8*), i64 2, i32 2, i1 false)
   %2 = bitcast %struct.s3* %p3 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %2, i8* bitcast ({ i16, i8, i8 }* @caller1.p3 to i8*), i64 4, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %2, i8* bitcast ({ i16, i8, i8 }* @caller1.p3 to i8*), i64 4, i32 2, i1 false)
   %3 = bitcast %struct.s4* %p4 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %3, i8* bitcast (%struct.s4* @caller1.p4 to i8*), i64 4, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %3, i8* bitcast (%struct.s4* @caller1.p4 to i8*), i64 4, i32 4, i1 false)
   %4 = bitcast %struct.s5* %p5 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %4, i8* bitcast ({ i32, i8, [3 x i8] }* @caller1.p5 to i8*), i64 8, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %4, i8* bitcast ({ i32, i8, [3 x i8] }* @caller1.p5 to i8*), i64 8, i32 4, i1 false)
   %5 = bitcast %struct.s6* %p6 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %5, i8* bitcast ({ i32, i16, [2 x i8] }* @caller1.p6 to i8*), i64 8, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %5, i8* bitcast ({ i32, i16, [2 x i8] }* @caller1.p6 to i8*), i64 8, i32 4, i1 false)
   %6 = bitcast %struct.s7* %p7 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %6, i8* bitcast ({ i32, i16, i8, i8 }* @caller1.p7 to i8*), i64 8, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %6, i8* bitcast ({ i32, i16, i8, i8 }* @caller1.p7 to i8*), i64 8, i32 4, i1 false)
   %call = call i32 @callee1(%struct.s1* byval %p1, %struct.s2* byval %p2, %struct.s3* byval %p3, %struct.s4* byval %p4, %struct.s5* byval %p5, %struct.s6* byval %p6, %struct.s7* byval %p7)
   ret i32 %call
 
@@ -68,7 +68,7 @@ entry:
 ; CHECK: lbz 3, 160(31)
 }
 
-declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i32, i1) nounwind
 
 define internal i32 @callee1(%struct.s1* byval %v1, %struct.s2* byval %v2, %struct.s3* byval %v3, %struct.s4* byval %v4, %struct.s5* byval %v5, %struct.s6* byval %v6, %struct.s7* byval %v7) nounwind {
 entry:
@@ -123,19 +123,19 @@ entry:
   %p6 = alloca %struct.t6, align 1
   %p7 = alloca %struct.t7, align 1
   %0 = bitcast %struct.t1* %p1 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %0, i8* getelementptr inbounds (%struct.t1, %struct.t1* @caller2.p1, i32 0, i32 0), i64 1, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %0, i8* getelementptr inbounds (%struct.t1, %struct.t1* @caller2.p1, i32 0, i32 0), i64 1, i32 1, i1 false)
   %1 = bitcast %struct.t2* %p2 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %1, i8* bitcast ({ i16 }* @caller2.p2 to i8*), i64 2, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %1, i8* bitcast ({ i16 }* @caller2.p2 to i8*), i64 2, i32 1, i1 false)
   %2 = bitcast %struct.t3* %p3 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %2, i8* bitcast (%struct.t3* @caller2.p3 to i8*), i64 3, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %2, i8* bitcast (%struct.t3* @caller2.p3 to i8*), i64 3, i32 1, i1 false)
   %3 = bitcast %struct.t4* %p4 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %3, i8* bitcast ({ i32 }* @caller2.p4 to i8*), i64 4, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %3, i8* bitcast ({ i32 }* @caller2.p4 to i8*), i64 4, i32 1, i1 false)
   %4 = bitcast %struct.t5* %p5 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %4, i8* bitcast (%struct.t5* @caller2.p5 to i8*), i64 5, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %4, i8* bitcast (%struct.t5* @caller2.p5 to i8*), i64 5, i32 1, i1 false)
   %5 = bitcast %struct.t6* %p6 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %5, i8* bitcast (%struct.t6* @caller2.p6 to i8*), i64 6, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %5, i8* bitcast (%struct.t6* @caller2.p6 to i8*), i64 6, i32 1, i1 false)
   %6 = bitcast %struct.t7* %p7 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %6, i8* bitcast (%struct.t7* @caller2.p7 to i8*), i64 7, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %6, i8* bitcast (%struct.t7* @caller2.p7 to i8*), i64 7, i32 1, i1 false)
   %call = call i32 @callee2(%struct.t1* byval %p1, %struct.t2* byval %p2, %struct.t3* byval %p3, %struct.t4* byval %p4, %struct.t5* byval %p5, %struct.t6* byval %p6, %struct.t7* byval %p7)
   ret i32 %call
 
diff --git a/test/CodeGen/PowerPC/stwu8.ll b/test/CodeGen/PowerPC/stwu8.ll
index 7b680264993..bb2748432d7 100644
--- a/test/CodeGen/PowerPC/stwu8.ll
+++ b/test/CodeGen/PowerPC/stwu8.ll
@@ -13,7 +13,7 @@ define void @test1(%class.spell_checker.21.103.513.538* %this) unnamed_addr alig
 entry:
   %_M_header.i.i.i.i.i.i = getelementptr inbounds %class.spell_checker.21.103.513.538, %class.spell_checker.21.103.513.538* %this, i64 0, i32 0, i32 0, i32 0, i32 1
   %0 = bitcast %"struct.std::_Rb_tree_node_base.17.99.509.534"* %_M_header.i.i.i.i.i.i to i8*
-  call void @llvm.memset.p0i8.i64(i8* align 4 %0, i8 0, i64 40, i1 false) nounwind
+  call void @llvm.memset.p0i8.i64(i8* %0, i8 0, i64 40, i32 4, i1 false) nounwind
   store %"struct.std::_Rb_tree_node_base.17.99.509.534"* %_M_header.i.i.i.i.i.i, %"struct.std::_Rb_tree_node_base.17.99.509.534"** undef, align 8
   unreachable
 }
@@ -21,4 +21,4 @@ entry:
 ; CHECK: @test1
 ; CHECK: stwu
 
-declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i1) nounwind
+declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i32, i1) nounwind
diff --git a/test/CodeGen/PowerPC/toc-load-sched-bug.ll b/test/CodeGen/PowerPC/toc-load-sched-bug.ll
index 8717183e22c..5ac4e363502 100644
--- a/test/CodeGen/PowerPC/toc-load-sched-bug.ll
+++ b/test/CodeGen/PowerPC/toc-load-sched-bug.ll
@@ -222,7 +222,7 @@ if.then:                                          ; preds = %_ZNK4llvm7ErrorOrIS
   %Filename.i = getelementptr inbounds %"class.llvm::SMDiagnostic", %"class.llvm::SMDiagnostic"* %ref.tmp, i64 0, i32 2
   %10 = getelementptr inbounds %"class.std::allocator", %"class.std::allocator"* %ref.tmp.i.i2.i, i64 0, i32 0
   %11 = bitcast %"class.llvm::SMDiagnostic"* %ref.tmp to i8*
-  call void @llvm.memset.p0i8.i64(i8* %11, i8 0, i64 16, i1 false) #3
+  call void @llvm.memset.p0i8.i64(i8* %11, i8 0, i64 16, i32 8, i1 false) #3
   call void @llvm.lifetime.start(i64 1, i8* %10) #3
   %tobool.i.i4.i = icmp eq i8* %4, null
   br i1 %tobool.i.i4.i, label %if.then.i.i6.i, label %if.end.i.i8.i
@@ -265,7 +265,7 @@ _ZN4llvm12SMDiagnosticC2ENS_9StringRefENS_9SourceMgr8DiagKindES1_.exit: ; preds
   store i8* bitcast (i64* getelementptr inbounds ([0 x i64], [0 x i64]* @_ZNSs4_Rep20_S_empty_rep_storageE, i64 0, i64 3) to i8*), i8** %_M_p.i.i.i.i.i, align 8, !tbaa !13
   %Ranges.i = getelementptr inbounds %"class.llvm::SMDiagnostic", %"class.llvm::SMDiagnostic"* %ref.tmp, i64 0, i32 8
   %13 = bitcast %"class.std::vector.79"* %Ranges.i to i8*
-  call void @llvm.memset.p0i8.i64(i8* %13, i8 0, i64 24, i1 false) #3
+  call void @llvm.memset.p0i8.i64(i8* %13, i8 0, i64 24, i32 8, i1 false) #3
   %14 = getelementptr inbounds %"class.llvm::SMDiagnostic", %"class.llvm::SMDiagnostic"* %ref.tmp, i64 0, i32 9, i32 0, i32 0, i32 0, i32 1, i32 0, i32 0, i64 0
   %BeginX.i.i.i.i.i.i = getelementptr inbounds %"class.llvm::SMDiagnostic", %"class.llvm::SMDiagnostic"* %ref.tmp, i64 0, i32 9, i32 0, i32 0, i32 0, i32 0, i32 0
   store i8* %14, i8** %BeginX.i.i.i.i.i.i, align 8, !tbaa !23
@@ -275,13 +275,13 @@ _ZN4llvm12SMDiagnosticC2ENS_9StringRefENS_9SourceMgr8DiagKindES1_.exit: ; preds
   %add.ptr.i.i.i.i.i.i = getelementptr inbounds %"class.llvm::SMDiagnostic", %"class.llvm::SMDiagnostic"* %ref.tmp, i64 0, i32 9, i32 0, i32 0, i32 0, i32 1, i32 0, i32 0, i64 96
   store i8* %add.ptr.i.i.i.i.i.i, i8** %CapacityX.i.i.i.i.i.i, align 8, !tbaa !26
   %15 = bitcast %"class.llvm::SMDiagnostic"* %Err to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %15, i8* %11, i64 16, i1 false) #3
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %15, i8* %11, i64 16, i32 8, i1 false) #3
   %Filename.i38 = getelementptr inbounds %"class.llvm::SMDiagnostic", %"class.llvm::SMDiagnostic"* %Err, i64 0, i32 2
   call void @_ZNSs4swapERSs(%"class.std::basic_string"* %Filename.i38, %"class.std::basic_string"* dereferenceable(8) %Filename.i) #3
   %LineNo.i39 = getelementptr inbounds %"class.llvm::SMDiagnostic", %"class.llvm::SMDiagnostic"* %Err, i64 0, i32 3
   %16 = bitcast i32* %LineNo.i39 to i8*
   %17 = bitcast i32* %LineNo.i to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %16, i8* %17, i64 12, i1 false) #3
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %16, i8* %17, i64 12, i32 4, i1 false) #3
   %Message.i40 = getelementptr inbounds %"class.llvm::SMDiagnostic", %"class.llvm::SMDiagnostic"* %Err, i64 0, i32 6
   call void @_ZNSs4swapERSs(%"class.std::basic_string"* %Message.i40, %"class.std::basic_string"* dereferenceable(8) %Message.i) #3
   %LineContents.i = getelementptr inbounds %"class.llvm::SMDiagnostic", %"class.llvm::SMDiagnostic"* %Err, i64 0, i32 7
@@ -294,7 +294,7 @@ _ZN4llvm12SMDiagnosticC2ENS_9StringRefENS_9SourceMgr8DiagKindES1_.exit: ; preds
   %_M_end_of_storage.i11.i.i.i = getelementptr inbounds %"class.llvm::SMDiagnostic", %"class.llvm::SMDiagnostic"* %Err, i64 0, i32 8, i32 0, i32 0, i32 2
   %_M_start2.i.i.i.i = getelementptr inbounds %"class.llvm::SMDiagnostic", %"class.llvm::SMDiagnostic"* %ref.tmp, i64 0, i32 8, i32 0, i32 0, i32 0
   %19 = bitcast %"class.std::vector.79"* %Ranges.i41 to i8*
-  call void @llvm.memset.p0i8.i64(i8* %19, i8 0, i64 16, i1 false) #3
+  call void @llvm.memset.p0i8.i64(i8* %19, i8 0, i64 16, i32 8, i1 false) #3
   %20 = load %"struct.std::pair"*, %"struct.std::pair"** %_M_start2.i.i.i.i, align 8, !tbaa !27
   store %"struct.std::pair"* %20, %"struct.std::pair"** %_M_start.i7.i.i.i, align 8, !tbaa !27
   store %"struct.std::pair"* null, %"struct.std::pair"** %_M_start2.i.i.i.i, align 8, !tbaa !27
@@ -449,7 +449,7 @@ declare void @__assert_fail(i8*, i8*, i32 zeroext, i8*) #4
 declare dereferenceable(8) %"class.std::basic_string"* @_ZNSs6insertEmPKcm(%"class.std::basic_string"*, i64, i8*, i64) #1
 
 ; Function Attrs: nounwind
-declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture readonly, i64, i1) #3
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture readonly, i64, i32, i1) #3
 
 ; Function Attrs: nounwind
 declare void @_ZNSs4_Rep10_M_destroyERKSaIcE(%"struct.std::basic_string<char, std::char_traits<char>, std::allocator<char> >::_Rep"*, %"class.std::allocator"* dereferenceable(1)) #0
@@ -471,7 +471,7 @@ declare %"class.llvm::Module"* @_ZN4llvm7ParseIREPNS_12MemoryBufferERNS_12SMDiag
 declare void @_ZNSs4swapERSs(%"class.std::basic_string"*, %"class.std::basic_string"* dereferenceable(8)) #1
 
 ; Function Attrs: nounwind
-declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i1) #3
+declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i32, i1) #3
 
 attributes #0 = { nounwind "less-precise-fpmad"="false" "no-frame-pointer-elim"="false" "no-infs-fp-math"="false" "no-nans-fp-math"="false" "stack-protector-buffer-size"="8" "unsafe-fp-math"="false" "use-soft-float"="false" }
 attributes #1 = { "less-precise-fpmad"="false" "no-frame-pointer-elim"="false" "no-infs-fp-math"="false" "no-nans-fp-math"="false" "stack-protector-buffer-size"="8" "unsafe-fp-math"="false" "use-soft-float"="false" }
diff --git a/test/CodeGen/SystemZ/memcpy-01.ll b/test/CodeGen/SystemZ/memcpy-01.ll
index b13967803b7..1d7b28e940b 100644
--- a/test/CodeGen/SystemZ/memcpy-01.ll
+++ b/test/CodeGen/SystemZ/memcpy-01.ll
@@ -2,8 +2,8 @@
 ;
 ; RUN: llc < %s -mtriple=s390x-linux-gnu | FileCheck %s
 
-declare void @llvm.memcpy.p0i8.p0i8.i32(i8 *nocapture, i8 *nocapture, i32, i1) nounwind
-declare void @llvm.memcpy.p0i8.p0i8.i64(i8 *nocapture, i8 *nocapture, i64, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i32(i8 *nocapture, i8 *nocapture, i32, i32, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8 *nocapture, i8 *nocapture, i64, i32, i1) nounwind
 declare void @foo(i8 *, i8 *)
 
 ; Test a no-op move, i32 version.
@@ -12,7 +12,7 @@ define void @f1(i8 *%dest, i8 *%src) {
 ; CHECK-NOT: %r2
 ; CHECK-NOT: %r3
 ; CHECK: br %r14
-  call void @llvm.memcpy.p0i8.p0i8.i32(i8 *%dest, i8 *%src, i32 0,
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8 *%dest, i8 *%src, i32 0, i32 1,
                                        i1 false)
   ret void
 }
@@ -23,7 +23,7 @@ define void @f2(i8 *%dest, i8 *%src) {
 ; CHECK-NOT: %r2
 ; CHECK-NOT: %r3
 ; CHECK: br %r14
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8 *%dest, i8 *%src, i64 0,
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8 *%dest, i8 *%src, i64 0, i32 1,
                                        i1 false)
   ret void
 }
@@ -33,7 +33,7 @@ define void @f3(i8 *%dest, i8 *%src) {
 ; CHECK-LABEL: f3:
 ; CHECK: mvc 0(1,%r2), 0(%r3)
 ; CHECK: br %r14
-  call void @llvm.memcpy.p0i8.p0i8.i32(i8 *%dest, i8 *%src, i32 1,
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8 *%dest, i8 *%src, i32 1, i32 1,
                                        i1 false)
   ret void
 }
@@ -43,7 +43,7 @@ define void @f4(i8 *%dest, i8 *%src) {
 ; CHECK-LABEL: f4:
 ; CHECK: mvc 0(1,%r2), 0(%r3)
 ; CHECK: br %r14
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8 *%dest, i8 *%src, i64 1,
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8 *%dest, i8 *%src, i64 1, i32 1,
                                        i1 false)
   ret void
 }
@@ -53,7 +53,7 @@ define void @f5(i8 *%dest, i8 *%src) {
 ; CHECK-LABEL: f5:
 ; CHECK: mvc 0(256,%r2), 0(%r3)
 ; CHECK: br %r14
-  call void @llvm.memcpy.p0i8.p0i8.i32(i8 *%dest, i8 *%src, i32 256,
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8 *%dest, i8 *%src, i32 256, i32 1,
                                        i1 false)
   ret void
 }
@@ -63,7 +63,7 @@ define void @f6(i8 *%dest, i8 *%src) {
 ; CHECK-LABEL: f6:
 ; CHECK: mvc 0(256,%r2), 0(%r3)
 ; CHECK: br %r14
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8 *%dest, i8 *%src, i64 256,
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8 *%dest, i8 *%src, i64 256, i32 1,
                                        i1 false)
   ret void
 }
@@ -74,7 +74,7 @@ define void @f7(i8 *%dest, i8 *%src) {
 ; CHECK: mvc 0(256,%r2), 0(%r3)
 ; CHECK: mvc 256(1,%r2), 256(%r3)
 ; CHECK: br %r14
-  call void @llvm.memcpy.p0i8.p0i8.i32(i8 *%dest, i8 *%src, i32 257,
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8 *%dest, i8 *%src, i32 257, i32 1,
                                        i1 false)
   ret void
 }
@@ -85,7 +85,7 @@ define void @f8(i8 *%dest, i8 *%src) {
 ; CHECK: mvc 0(256,%r2), 0(%r3)
 ; CHECK: mvc 256(255,%r2), 256(%r3)
 ; CHECK: br %r14
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8 *%dest, i8 *%src, i64 511,
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8 *%dest, i8 *%src, i64 511, i32 1,
                                        i1 false)
   ret void
 }
@@ -96,7 +96,7 @@ define void @f9(i8 *%dest, i8 *%src) {
 ; CHECK: mvc 0(256,%r2), 0(%r3)
 ; CHECK: mvc 256(256,%r2), 256(%r3)
 ; CHECK: br %r14
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8 *%dest, i8 *%src, i64 512,
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8 *%dest, i8 *%src, i64 512, i32 1,
                                        i1 false)
   ret void
 }
@@ -110,7 +110,7 @@ define void @f10(i8 *%dest, i8 *%src) {
 ; CHECK: mvc 768(256,%r2), 768(%r3)
 ; CHECK: mvc 1024(255,%r2), 1024(%r3)
 ; CHECK: br %r14
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8 *%dest, i8 *%src, i64 1279,
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8 *%dest, i8 *%src, i64 1279, i32 1,
                                        i1 false)
   ret void
 }
@@ -128,7 +128,7 @@ define void @f11(i8 *%srcbase, i8 *%destbase) {
 ; CHECK: br %r14
   %dest = getelementptr i8, i8 *%srcbase, i64 4000
   %src = getelementptr i8, i8* %destbase, i64 3500
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8 *%dest, i8 *%src, i64 1279,
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8 *%dest, i8 *%src, i64 1279, i32 1,
                                        i1 false)
   ret void
 }
@@ -149,7 +149,7 @@ define void @f12() {
   %dest = getelementptr [6000 x i8], [6000 x i8] *%arr, i64 0, i64 3900
   %src = getelementptr [6000 x i8], [6000 x i8] *%arr, i64 0, i64 1924
   call void @foo(i8 *%dest, i8 *%src)
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8 *%dest, i8 *%src, i64 1279,
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8 *%dest, i8 *%src, i64 1279, i32 1,
                                        i1 false)
   call void @foo(i8 *%dest, i8 *%src)
   ret void
@@ -171,7 +171,7 @@ define void @f13() {
   %dest = getelementptr [6000 x i8], [6000 x i8] *%arr, i64 0, i64 24
   %src = getelementptr [6000 x i8], [6000 x i8] *%arr, i64 0, i64 3650
   call void @foo(i8 *%dest, i8 *%src)
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8 *%dest, i8 *%src, i64 1279,
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8 *%dest, i8 *%src, i64 1279, i32 1,
                                        i1 false)
   call void @foo(i8 *%dest, i8 *%src)
   ret void
@@ -187,7 +187,7 @@ define void @f14(i8 *%dest, i8 *%src) {
 ; CHECK: mvc 1024(256,%r2), 1024(%r3)
 ; CHECK: mvc 1280(256,%r2), 1280(%r3)
 ; CHECK: br %r14
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8 *%dest, i8 *%src, i64 1536,
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8 *%dest, i8 *%src, i64 1536, i32 1,
                                        i1 false)
   ret void
 }
@@ -204,7 +204,7 @@ define void @f15(i8 *%dest, i8 *%src) {
 ; CHECK: brctg [[COUNT]], [[LABEL]]
 ; CHECK: mvc 0(1,%r2), 0(%r3)
 ; CHECK: br %r14
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8 *%dest, i8 *%src, i64 1537,
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8 *%dest, i8 *%src, i64 1537, i32 1,
                                        i1 false)
   ret void
 }
@@ -228,7 +228,7 @@ define void @f16() {
   %dest = getelementptr [3200 x i8], [3200 x i8] *%arr, i64 0, i64 1600
   %src = getelementptr [3200 x i8], [3200 x i8] *%arr, i64 0, i64 0
   call void @foo(i8 *%dest, i8 *%src)
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8 *%dest, i8 *%src, i64 1537,
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8 *%dest, i8 *%src, i64 1537, i32 1,
                                        i1 false)
   call void @foo(i8 *%dest, i8 *%src)
   ret void
diff --git a/test/CodeGen/SystemZ/memset-01.ll b/test/CodeGen/SystemZ/memset-01.ll
index d0295ce1fbf..f17901cc73a 100644
--- a/test/CodeGen/SystemZ/memset-01.ll
+++ b/test/CodeGen/SystemZ/memset-01.ll
@@ -2,8 +2,8 @@
 ;
 ; RUN: llc < %s -mtriple=s390x-linux-gnu | FileCheck %s
 
-declare void @llvm.memset.p0i8.i32(i8 *nocapture, i8, i32, i1) nounwind
-declare void @llvm.memset.p0i8.i64(i8 *nocapture, i8, i64, i1) nounwind
+declare void @llvm.memset.p0i8.i32(i8 *nocapture, i8, i32, i32, i1) nounwind
+declare void @llvm.memset.p0i8.i64(i8 *nocapture, i8, i64, i32, i1) nounwind
 
 ; No bytes, i32 version.
 define void @f1(i8 *%dest, i8 %val) {
@@ -11,7 +11,7 @@ define void @f1(i8 *%dest, i8 %val) {
 ; CHECK-NOT: %r2
 ; CHECK-NOT: %r3
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 %val, i32 0, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 %val, i32 0, i32 1, i1 false)
   ret void
 }
 
@@ -21,7 +21,7 @@ define void @f2(i8 *%dest, i8 %val) {
 ; CHECK-NOT: %r2
 ; CHECK-NOT: %r3
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 %val, i64 0, i1 false)
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 %val, i64 0, i32 1, i1 false)
   ret void
 }
 
@@ -30,7 +30,7 @@ define void @f3(i8 *%dest, i8 %val) {
 ; CHECK-LABEL: f3:
 ; CHECK: stc %r3, 0(%r2)
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 %val, i32 1, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 %val, i32 1, i32 1, i1 false)
   ret void
 }
 
@@ -39,7 +39,7 @@ define void @f4(i8 *%dest, i8 %val) {
 ; CHECK-LABEL: f4:
 ; CHECK: stc %r3, 0(%r2)
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 %val, i64 1, i1 false)
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 %val, i64 1, i32 1, i1 false)
   ret void
 }
 
@@ -49,7 +49,7 @@ define void @f5(i8 *%dest, i8 %val) {
 ; CHECK-DAG: stc %r3, 0(%r2)
 ; CHECK-DAG: stc %r3, 1(%r2)
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 %val, i32 2, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 %val, i32 2, i32 1, i1 false)
   ret void
 }
 
@@ -59,7 +59,7 @@ define void @f6(i8 *%dest, i8 %val) {
 ; CHECK-DAG: stc %r3, 0(%r2)
 ; CHECK-DAG: stc %r3, 1(%r2)
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 %val, i64 2, i1 false)
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 %val, i64 2, i32 1, i1 false)
   ret void
 }
 
@@ -69,7 +69,7 @@ define void @f7(i8 *%dest, i8 %val) {
 ; CHECK: stc %r3, 0(%r2)
 ; CHECK: mvc 1(2,%r2), 0(%r2)
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 %val, i32 3, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 %val, i32 3, i32 1, i1 false)
   ret void
 }
 
@@ -79,7 +79,7 @@ define void @f8(i8 *%dest, i8 %val) {
 ; CHECK: stc %r3, 0(%r2)
 ; CHECK: mvc 1(2,%r2), 0(%r2)
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 %val, i64 3, i1 false)
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 %val, i64 3, i32 1, i1 false)
   ret void
 }
 
@@ -89,7 +89,7 @@ define void @f9(i8 *%dest, i8 %val) {
 ; CHECK: stc %r3, 0(%r2)
 ; CHECK: mvc 1(256,%r2), 0(%r2)
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 %val, i32 257, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 %val, i32 257, i32 1, i1 false)
   ret void
 }
 
@@ -99,7 +99,7 @@ define void @f10(i8 *%dest, i8 %val) {
 ; CHECK: stc %r3, 0(%r2)
 ; CHECK: mvc 1(256,%r2), 0(%r2)
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 %val, i64 257, i1 false)
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 %val, i64 257, i32 1, i1 false)
   ret void
 }
 
@@ -110,7 +110,7 @@ define void @f11(i8 *%dest, i8 %val) {
 ; CHECK: mvc 1(256,%r2), 0(%r2)
 ; CHECK: mvc 257(1,%r2), 256(%r2)
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 %val, i32 258, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 %val, i32 258, i32 1, i1 false)
   ret void
 }
 
@@ -121,7 +121,7 @@ define void @f12(i8 *%dest, i8 %val) {
 ; CHECK: mvc 1(256,%r2), 0(%r2)
 ; CHECK: mvc 257(1,%r2), 256(%r2)
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 %val, i64 258, i1 false)
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 %val, i64 258, i32 1, i1 false)
   ret void
 }
 
@@ -136,7 +136,7 @@ define void @f13(i8 *%dest, i8 %val) {
 ; CHECK: mvc 1025(256,%r2), 1024(%r2)
 ; CHECK: mvc 1281(256,%r2), 1280(%r2)
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 %val, i64 1537,
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 %val, i64 1537, i32 1,
                                   i1 false)
   ret void
 }
@@ -154,7 +154,7 @@ define void @f14(i8 *%dest, i8 %val) {
 ; CHECK: brctg [[COUNT]], [[LABEL]]
 ; CHECK: mvc 1(1,%r2), 0(%r2)
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 %val, i64 1538,
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 %val, i64 1538, i32 1,
                                   i1 false)
   ret void
 }
diff --git a/test/CodeGen/SystemZ/memset-02.ll b/test/CodeGen/SystemZ/memset-02.ll
index 7c59b13410d..b4724c0b574 100644
--- a/test/CodeGen/SystemZ/memset-02.ll
+++ b/test/CodeGen/SystemZ/memset-02.ll
@@ -2,15 +2,15 @@
 ;
 ; RUN: llc < %s -mtriple=s390x-linux-gnu | FileCheck %s
 
-declare void @llvm.memset.p0i8.i32(i8 *nocapture, i8, i32, i1) nounwind
-declare void @llvm.memset.p0i8.i64(i8 *nocapture, i8, i64, i1) nounwind
+declare void @llvm.memset.p0i8.i32(i8 *nocapture, i8, i32, i32, i1) nounwind
+declare void @llvm.memset.p0i8.i64(i8 *nocapture, i8, i64, i32, i1) nounwind
 
 ; No bytes, i32 version.
 define void @f1(i8 *%dest) {
 ; CHECK-LABEL: f1:
 ; CHECK-NOT: %r2
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 128, i32 0, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 128, i32 0, i32 1, i1 false)
   ret void
 }
 
@@ -19,7 +19,7 @@ define void @f2(i8 *%dest) {
 ; CHECK-LABEL: f2:
 ; CHECK-NOT: %r2
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 128, i64 0, i1 false)
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 128, i64 0, i32 1, i1 false)
   ret void
 }
 
@@ -28,7 +28,7 @@ define void @f3(i8 *%dest) {
 ; CHECK-LABEL: f3:
 ; CHECK: mvi 0(%r2), 128
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 128, i32 1, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 128, i32 1, i32 1, i1 false)
   ret void
 }
 
@@ -37,7 +37,7 @@ define void @f4(i8 *%dest) {
 ; CHECK-LABEL: f4:
 ; CHECK: mvi 0(%r2), 128
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 128, i64 1, i1 false)
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 128, i64 1, i32 1, i1 false)
   ret void
 }
 
@@ -46,7 +46,7 @@ define void @f5(i8 *%dest) {
 ; CHECK-LABEL: f5:
 ; CHECK: mvhhi 0(%r2), -32640
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 128, i32 2, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 128, i32 2, i32 1, i1 false)
   ret void
 }
 
@@ -55,7 +55,7 @@ define void @f6(i8 *%dest) {
 ; CHECK-LABEL: f6:
 ; CHECK: mvhhi 0(%r2), -32640
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 128, i64 2, i1 false)
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 128, i64 2, i32 1, i1 false)
   ret void
 }
 
@@ -65,7 +65,7 @@ define void @f7(i8 *%dest) {
 ; CHECK-DAG: mvhhi 0(%r2), -32640
 ; CHECK-DAG: mvi 2(%r2), 128
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 128, i32 3, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 128, i32 3, i32 1, i1 false)
   ret void
 }
 
@@ -75,7 +75,7 @@ define void @f8(i8 *%dest) {
 ; CHECK-DAG: mvhhi 0(%r2), -32640
 ; CHECK-DAG: mvi 2(%r2), 128
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 128, i64 3, i1 false)
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 128, i64 3, i32 1, i1 false)
   ret void
 }
 
@@ -85,7 +85,7 @@ define void @f9(i8 *%dest) {
 ; CHECK: iilf [[REG:%r[0-5]]], 2155905152
 ; CHECK: st [[REG]], 0(%r2)
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 128, i32 4, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 128, i32 4, i32 1, i1 false)
   ret void
 }
 
@@ -95,7 +95,7 @@ define void @f10(i8 *%dest) {
 ; CHECK: iilf [[REG:%r[0-5]]], 2155905152
 ; CHECK: st [[REG]], 0(%r2)
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 128, i64 4, i1 false)
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 128, i64 4, i32 1, i1 false)
   ret void
 }
 
@@ -105,7 +105,7 @@ define void @f11(i8 *%dest) {
 ; CHECK: mvi 0(%r2), 128
 ; CHECK: mvc 1(4,%r2), 0(%r2)
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 128, i32 5, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 128, i32 5, i32 1, i1 false)
   ret void
 }
 
@@ -115,7 +115,7 @@ define void @f12(i8 *%dest) {
 ; CHECK: mvi 0(%r2), 128
 ; CHECK: mvc 1(4,%r2), 0(%r2)
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 128, i64 5, i1 false)
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 128, i64 5, i32 1, i1 false)
   ret void
 }
 
@@ -125,7 +125,7 @@ define void @f13(i8 *%dest) {
 ; CHECK: mvi 0(%r2), 128
 ; CHECK: mvc 1(256,%r2), 0(%r2)
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 128, i32 257, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 128, i32 257, i32 1, i1 false)
   ret void
 }
 
@@ -135,7 +135,7 @@ define void @f14(i8 *%dest) {
 ; CHECK: mvi 0(%r2), 128
 ; CHECK: mvc 1(256,%r2), 0(%r2)
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 128, i64 257, i1 false)
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 128, i64 257, i32 1, i1 false)
   ret void
 }
 
@@ -146,7 +146,7 @@ define void @f15(i8 *%dest) {
 ; CHECK: mvc 1(256,%r2), 0(%r2)
 ; CHECK: mvc 257(1,%r2), 256(%r2)
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 128, i32 258, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 128, i32 258, i32 1, i1 false)
   ret void
 }
 
@@ -157,6 +157,6 @@ define void @f16(i8 *%dest) {
 ; CHECK: mvc 1(256,%r2), 0(%r2)
 ; CHECK: mvc 257(1,%r2), 256(%r2)
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 128, i64 258, i1 false)
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 128, i64 258, i32 1, i1 false)
   ret void
 }
diff --git a/test/CodeGen/SystemZ/memset-03.ll b/test/CodeGen/SystemZ/memset-03.ll
index 65d062f7731..a95f89fc7c0 100644
--- a/test/CodeGen/SystemZ/memset-03.ll
+++ b/test/CodeGen/SystemZ/memset-03.ll
@@ -2,15 +2,15 @@
 ;
 ; RUN: llc < %s -mtriple=s390x-linux-gnu | FileCheck %s
 
-declare void @llvm.memset.p0i8.i32(i8 *nocapture, i8, i32, i1) nounwind
-declare void @llvm.memset.p0i8.i64(i8 *nocapture, i8, i64, i1) nounwind
+declare void @llvm.memset.p0i8.i32(i8 *nocapture, i8, i32, i32, i1) nounwind
+declare void @llvm.memset.p0i8.i64(i8 *nocapture, i8, i64, i32, i1) nounwind
 
 ; No bytes, i32 version.
 define void @f1(i8 *%dest) {
 ; CHECK-LABEL: f1:
 ; CHECK-NOT: %r2
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 0, i32 0, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 0, i32 0, i32 1, i1 false)
   ret void
 }
 
@@ -19,7 +19,7 @@ define void @f2(i8 *%dest) {
 ; CHECK-LABEL: f2:
 ; CHECK-NOT: %r2
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 0, i64 0, i1 false)
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 0, i64 0, i32 1, i1 false)
   ret void
 }
 
@@ -28,7 +28,7 @@ define void @f3(i8 *%dest) {
 ; CHECK-LABEL: f3:
 ; CHECK: mvi 0(%r2), 0
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 0, i32 1, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 0, i32 1, i32 1, i1 false)
   ret void
 }
 
@@ -37,7 +37,7 @@ define void @f4(i8 *%dest) {
 ; CHECK-LABEL: f4:
 ; CHECK: mvi 0(%r2), 0
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 0, i64 1, i1 false)
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 0, i64 1, i32 1, i1 false)
   ret void
 }
 
@@ -46,7 +46,7 @@ define void @f5(i8 *%dest) {
 ; CHECK-LABEL: f5:
 ; CHECK: mvhhi 0(%r2), 0
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 0, i32 2, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 0, i32 2, i32 1, i1 false)
   ret void
 }
 
@@ -55,7 +55,7 @@ define void @f6(i8 *%dest) {
 ; CHECK-LABEL: f6:
 ; CHECK: mvhhi 0(%r2), 0
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 0, i64 2, i1 false)
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 0, i64 2, i32 1, i1 false)
   ret void
 }
 
@@ -65,7 +65,7 @@ define void @f7(i8 *%dest) {
 ; CHECK-DAG: mvhhi 0(%r2), 0
 ; CHECK-DAG: mvi 2(%r2), 0
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 0, i32 3, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 0, i32 3, i32 1, i1 false)
   ret void
 }
 
@@ -75,7 +75,7 @@ define void @f8(i8 *%dest) {
 ; CHECK-DAG: mvhhi 0(%r2), 0
 ; CHECK-DAG: mvi 2(%r2), 0
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 0, i64 3, i1 false)
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 0, i64 3, i32 1, i1 false)
   ret void
 }
 
@@ -84,7 +84,7 @@ define void @f9(i8 *%dest) {
 ; CHECK-LABEL: f9:
 ; CHECK: mvhi 0(%r2), 0
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 0, i32 4, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 0, i32 4, i32 1, i1 false)
   ret void
 }
 
@@ -93,7 +93,7 @@ define void @f10(i8 *%dest) {
 ; CHECK-LABEL: f10:
 ; CHECK: mvhi 0(%r2), 0
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 0, i64 4, i1 false)
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 0, i64 4, i32 1, i1 false)
   ret void
 }
 
@@ -103,7 +103,7 @@ define void @f11(i8 *%dest) {
 ; CHECK-DAG: mvhi 0(%r2), 0
 ; CHECK-DAG: mvi 4(%r2), 0
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 0, i32 5, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 0, i32 5, i32 1, i1 false)
   ret void
 }
 
@@ -113,7 +113,7 @@ define void @f12(i8 *%dest) {
 ; CHECK-DAG: mvhi 0(%r2), 0
 ; CHECK-DAG: mvi 4(%r2), 0
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 0, i64 5, i1 false)
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 0, i64 5, i32 1, i1 false)
   ret void
 }
 
@@ -123,7 +123,7 @@ define void @f13(i8 *%dest) {
 ; CHECK-DAG: mvhi 0(%r2), 0
 ; CHECK-DAG: mvhhi 4(%r2), 0
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 0, i32 6, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 0, i32 6, i32 1, i1 false)
   ret void
 }
 
@@ -133,7 +133,7 @@ define void @f14(i8 *%dest) {
 ; CHECK-DAG: mvhi 0(%r2), 0
 ; CHECK-DAG: mvhhi 4(%r2), 0
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 0, i64 6, i1 false)
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 0, i64 6, i32 1, i1 false)
   ret void
 }
 
@@ -142,7 +142,7 @@ define void @f15(i8 *%dest) {
 ; CHECK-LABEL: f15:
 ; CHECK: xc 0(7,%r2), 0(%r2)
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 0, i32 7, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 0, i32 7, i32 1, i1 false)
   ret void
 }
 
@@ -151,7 +151,7 @@ define void @f16(i8 *%dest) {
 ; CHECK-LABEL: f16:
 ; CHECK: xc 0(7,%r2), 0(%r2)
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 0, i64 7, i1 false)
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 0, i64 7, i32 1, i1 false)
   ret void
 }
 
@@ -160,7 +160,7 @@ define void @f17(i8 *%dest) {
 ; CHECK-LABEL: f17:
 ; CHECK: mvghi 0(%r2), 0
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 0, i32 8, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 0, i32 8, i32 1, i1 false)
   ret void
 }
 
@@ -169,7 +169,7 @@ define void @f18(i8 *%dest) {
 ; CHECK-LABEL: f18:
 ; CHECK: mvghi 0(%r2), 0
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 0, i64 8, i1 false)
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 0, i64 8, i32 1, i1 false)
   ret void
 }
 
@@ -179,7 +179,7 @@ define void @f19(i8 *%dest) {
 ; CHECK-DAG: mvghi 0(%r2), 0
 ; CHECK-DAG: mvi 8(%r2), 0
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 0, i32 9, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 0, i32 9, i32 1, i1 false)
   ret void
 }
 
@@ -189,7 +189,7 @@ define void @f20(i8 *%dest) {
 ; CHECK-DAG: mvghi 0(%r2), 0
 ; CHECK-DAG: mvi 8(%r2), 0
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 0, i64 9, i1 false)
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 0, i64 9, i32 1, i1 false)
   ret void
 }
 
@@ -199,7 +199,7 @@ define void @f21(i8 *%dest) {
 ; CHECK-DAG: mvghi 0(%r2), 0
 ; CHECK-DAG: mvhhi 8(%r2), 0
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 0, i32 10, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 0, i32 10, i32 1, i1 false)
   ret void
 }
 
@@ -209,7 +209,7 @@ define void @f22(i8 *%dest) {
 ; CHECK-DAG: mvghi 0(%r2), 0
 ; CHECK-DAG: mvhhi 8(%r2), 0
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 0, i64 10, i1 false)
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 0, i64 10, i32 1, i1 false)
   ret void
 }
 
@@ -218,7 +218,7 @@ define void @f23(i8 *%dest) {
 ; CHECK-LABEL: f23:
 ; CHECK: xc 0(11,%r2), 0(%r2)
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 0, i32 11, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 0, i32 11, i32 1, i1 false)
   ret void
 }
 
@@ -227,7 +227,7 @@ define void @f24(i8 *%dest) {
 ; CHECK-LABEL: f24:
 ; CHECK: xc 0(11,%r2), 0(%r2)
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 0, i64 11, i1 false)
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 0, i64 11, i32 1, i1 false)
   ret void
 }
 
@@ -237,7 +237,7 @@ define void @f25(i8 *%dest) {
 ; CHECK-DAG: mvghi 0(%r2), 0
 ; CHECK-DAG: mvhi 8(%r2), 0
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 0, i32 12, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 0, i32 12, i32 1, i1 false)
   ret void
 }
 
@@ -247,7 +247,7 @@ define void @f26(i8 *%dest) {
 ; CHECK-DAG: mvghi 0(%r2), 0
 ; CHECK-DAG: mvhi 8(%r2), 0
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 0, i32 12, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 0, i32 12, i32 1, i1 false)
   ret void
 }
 
@@ -256,7 +256,7 @@ define void @f27(i8 *%dest) {
 ; CHECK-LABEL: f27:
 ; CHECK: xc 0(13,%r2), 0(%r2)
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 0, i32 13, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 0, i32 13, i32 1, i1 false)
   ret void
 }
 
@@ -265,7 +265,7 @@ define void @f28(i8 *%dest) {
 ; CHECK-LABEL: f28:
 ; CHECK: xc 0(13,%r2), 0(%r2)
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 0, i64 13, i1 false)
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 0, i64 13, i32 1, i1 false)
   ret void
 }
 
@@ -274,7 +274,7 @@ define void @f29(i8 *%dest) {
 ; CHECK-LABEL: f29:
 ; CHECK: xc 0(14,%r2), 0(%r2)
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 0, i32 14, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 0, i32 14, i32 1, i1 false)
   ret void
 }
 
@@ -283,7 +283,7 @@ define void @f30(i8 *%dest) {
 ; CHECK-LABEL: f30:
 ; CHECK: xc 0(14,%r2), 0(%r2)
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 0, i64 14, i1 false)
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 0, i64 14, i32 1, i1 false)
   ret void
 }
 
@@ -292,7 +292,7 @@ define void @f31(i8 *%dest) {
 ; CHECK-LABEL: f31:
 ; CHECK: xc 0(15,%r2), 0(%r2)
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 0, i32 15, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 0, i32 15, i32 1, i1 false)
   ret void
 }
 
@@ -301,7 +301,7 @@ define void @f32(i8 *%dest) {
 ; CHECK-LABEL: f32:
 ; CHECK: xc 0(15,%r2), 0(%r2)
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 0, i64 15, i1 false)
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 0, i64 15, i32 1, i1 false)
   ret void
 }
 
@@ -311,7 +311,7 @@ define void @f33(i8 *%dest) {
 ; CHECK-DAG: mvghi 0(%r2), 0
 ; CHECK-DAG: mvghi 8(%r2), 0
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 0, i32 16, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 0, i32 16, i32 1, i1 false)
   ret void
 }
 
@@ -321,7 +321,7 @@ define void @f34(i8 *%dest) {
 ; CHECK-DAG: mvghi 0(%r2), 0
 ; CHECK-DAG: mvghi 8(%r2), 0
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 0, i64 16, i1 false)
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 0, i64 16, i32 1, i1 false)
   ret void
 }
 
@@ -330,7 +330,7 @@ define void @f35(i8 *%dest) {
 ; CHECK-LABEL: f35:
 ; CHECK: xc 0(17,%r2), 0(%r2)
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 0, i32 17, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 0, i32 17, i32 1, i1 false)
   ret void
 }
 
@@ -339,7 +339,7 @@ define void @f36(i8 *%dest) {
 ; CHECK-LABEL: f36:
 ; CHECK: xc 0(17,%r2), 0(%r2)
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 0, i64 17, i1 false)
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 0, i64 17, i32 1, i1 false)
   ret void
 }
 
@@ -348,7 +348,7 @@ define void @f37(i8 *%dest) {
 ; CHECK-LABEL: f37:
 ; CHECK: xc 0(256,%r2), 0(%r2)
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 0, i32 256, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 0, i32 256, i32 1, i1 false)
   ret void
 }
 
@@ -357,7 +357,7 @@ define void @f38(i8 *%dest) {
 ; CHECK-LABEL: f38:
 ; CHECK: xc 0(256,%r2), 0(%r2)
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 0, i64 256, i1 false)
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 0, i64 256, i32 1, i1 false)
   ret void
 }
 
@@ -367,7 +367,7 @@ define void @f39(i8 *%dest) {
 ; CHECK: xc 0(256,%r2), 0(%r2)
 ; CHECK: xc 256(1,%r2), 256(%r2)
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 0, i32 257, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 0, i32 257, i32 1, i1 false)
   ret void
 }
 
@@ -377,6 +377,6 @@ define void @f40(i8 *%dest) {
 ; CHECK: xc 0(256,%r2), 0(%r2)
 ; CHECK: xc 256(1,%r2), 256(%r2)
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 0, i64 257, i1 false)
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 0, i64 257, i32 1, i1 false)
   ret void
 }
diff --git a/test/CodeGen/SystemZ/memset-04.ll b/test/CodeGen/SystemZ/memset-04.ll
index 1e0ba61877b..7906e8d10a1 100644
--- a/test/CodeGen/SystemZ/memset-04.ll
+++ b/test/CodeGen/SystemZ/memset-04.ll
@@ -2,15 +2,15 @@
 ;
 ; RUN: llc < %s -mtriple=s390x-linux-gnu | FileCheck %s
 
-declare void @llvm.memset.p0i8.i32(i8 *nocapture, i8, i32, i1) nounwind
-declare void @llvm.memset.p0i8.i64(i8 *nocapture, i8, i64, i1) nounwind
+declare void @llvm.memset.p0i8.i32(i8 *nocapture, i8, i32, i32, i1) nounwind
+declare void @llvm.memset.p0i8.i64(i8 *nocapture, i8, i64, i32, i1) nounwind
 
 ; No bytes, i32 version.
 define void @f1(i8 *%dest) {
 ; CHECK-LABEL: f1:
 ; CHECK-NOT: %r2
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 -1, i32 0, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 -1, i32 0, i32 1, i1 false)
   ret void
 }
 
@@ -19,7 +19,7 @@ define void @f2(i8 *%dest) {
 ; CHECK-LABEL: f2:
 ; CHECK-NOT: %r2
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 -1, i64 0, i1 false)
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 -1, i64 0, i32 1, i1 false)
   ret void
 }
 
@@ -28,7 +28,7 @@ define void @f3(i8 *%dest) {
 ; CHECK-LABEL: f3:
 ; CHECK: mvi 0(%r2), 255
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 -1, i32 1, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 -1, i32 1, i32 1, i1 false)
   ret void
 }
 
@@ -37,7 +37,7 @@ define void @f4(i8 *%dest) {
 ; CHECK-LABEL: f4:
 ; CHECK: mvi 0(%r2), 255
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 -1, i64 1, i1 false)
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 -1, i64 1, i32 1, i1 false)
   ret void
 }
 
@@ -46,7 +46,7 @@ define void @f5(i8 *%dest) {
 ; CHECK-LABEL: f5:
 ; CHECK: mvhhi 0(%r2), -1
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 -1, i32 2, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 -1, i32 2, i32 1, i1 false)
   ret void
 }
 
@@ -55,7 +55,7 @@ define void @f6(i8 *%dest) {
 ; CHECK-LABEL: f6:
 ; CHECK: mvhhi 0(%r2), -1
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 -1, i64 2, i1 false)
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 -1, i64 2, i32 1, i1 false)
   ret void
 }
 
@@ -65,7 +65,7 @@ define void @f7(i8 *%dest) {
 ; CHECK-DAG: mvhhi 0(%r2), -1
 ; CHECK-DAG: mvi 2(%r2), 255
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 -1, i32 3, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 -1, i32 3, i32 1, i1 false)
   ret void
 }
 
@@ -75,7 +75,7 @@ define void @f8(i8 *%dest) {
 ; CHECK-DAG: mvhhi 0(%r2), -1
 ; CHECK-DAG: mvi 2(%r2), 255
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 -1, i64 3, i1 false)
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 -1, i64 3, i32 1, i1 false)
   ret void
 }
 
@@ -84,7 +84,7 @@ define void @f9(i8 *%dest) {
 ; CHECK-LABEL: f9:
 ; CHECK: mvhi 0(%r2), -1
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 -1, i32 4, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 -1, i32 4, i32 1, i1 false)
   ret void
 }
 
@@ -93,7 +93,7 @@ define void @f10(i8 *%dest) {
 ; CHECK-LABEL: f10:
 ; CHECK: mvhi 0(%r2), -1
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 -1, i64 4, i1 false)
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 -1, i64 4, i32 1, i1 false)
   ret void
 }
 
@@ -103,7 +103,7 @@ define void @f11(i8 *%dest) {
 ; CHECK-DAG: mvhi 0(%r2), -1
 ; CHECK-DAG: mvi 4(%r2), 255
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 -1, i32 5, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 -1, i32 5, i32 1, i1 false)
   ret void
 }
 
@@ -113,7 +113,7 @@ define void @f12(i8 *%dest) {
 ; CHECK-DAG: mvhi 0(%r2), -1
 ; CHECK-DAG: mvi 4(%r2), 255
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 -1, i64 5, i1 false)
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 -1, i64 5, i32 1, i1 false)
   ret void
 }
 
@@ -123,7 +123,7 @@ define void @f13(i8 *%dest) {
 ; CHECK-DAG: mvhi 0(%r2), -1
 ; CHECK-DAG: mvhhi 4(%r2), -1
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 -1, i32 6, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 -1, i32 6, i32 1, i1 false)
   ret void
 }
 
@@ -133,7 +133,7 @@ define void @f14(i8 *%dest) {
 ; CHECK-DAG: mvhi 0(%r2), -1
 ; CHECK-DAG: mvhhi 4(%r2), -1
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 -1, i64 6, i1 false)
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 -1, i64 6, i32 1, i1 false)
   ret void
 }
 
@@ -143,7 +143,7 @@ define void @f15(i8 *%dest) {
 ; CHECK: mvi 0(%r2), 255
 ; CHECK: mvc 1(6,%r2), 0(%r2)
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 -1, i32 7, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 -1, i32 7, i32 1, i1 false)
   ret void
 }
 
@@ -153,7 +153,7 @@ define void @f16(i8 *%dest) {
 ; CHECK: mvi 0(%r2), 255
 ; CHECK: mvc 1(6,%r2), 0(%r2)
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 -1, i64 7, i1 false)
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 -1, i64 7, i32 1, i1 false)
   ret void
 }
 
@@ -162,7 +162,7 @@ define void @f17(i8 *%dest) {
 ; CHECK-LABEL: f17:
 ; CHECK: mvghi 0(%r2), -1
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 -1, i32 8, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 -1, i32 8, i32 1, i1 false)
   ret void
 }
 
@@ -171,7 +171,7 @@ define void @f18(i8 *%dest) {
 ; CHECK-LABEL: f18:
 ; CHECK: mvghi 0(%r2), -1
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 -1, i64 8, i1 false)
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 -1, i64 8, i32 1, i1 false)
   ret void
 }
 
@@ -181,7 +181,7 @@ define void @f19(i8 *%dest) {
 ; CHECK-DAG: mvghi 0(%r2), -1
 ; CHECK-DAG: mvi 8(%r2), 255
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 -1, i32 9, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 -1, i32 9, i32 1, i1 false)
   ret void
 }
 
@@ -191,7 +191,7 @@ define void @f20(i8 *%dest) {
 ; CHECK-DAG: mvghi 0(%r2), -1
 ; CHECK-DAG: mvi 8(%r2), 255
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 -1, i64 9, i1 false)
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 -1, i64 9, i32 1, i1 false)
   ret void
 }
 
@@ -201,7 +201,7 @@ define void @f21(i8 *%dest) {
 ; CHECK-DAG: mvghi 0(%r2), -1
 ; CHECK-DAG: mvhhi 8(%r2), -1
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 -1, i32 10, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 -1, i32 10, i32 1, i1 false)
   ret void
 }
 
@@ -211,7 +211,7 @@ define void @f22(i8 *%dest) {
 ; CHECK-DAG: mvghi 0(%r2), -1
 ; CHECK-DAG: mvhhi 8(%r2), -1
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 -1, i64 10, i1 false)
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 -1, i64 10, i32 1, i1 false)
   ret void
 }
 
@@ -221,7 +221,7 @@ define void @f23(i8 *%dest) {
 ; CHECK: mvi 0(%r2), 255
 ; CHECK: mvc 1(10,%r2), 0(%r2)
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 -1, i32 11, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 -1, i32 11, i32 1, i1 false)
   ret void
 }
 
@@ -231,7 +231,7 @@ define void @f24(i8 *%dest) {
 ; CHECK: mvi 0(%r2), 255
 ; CHECK: mvc 1(10,%r2), 0(%r2)
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 -1, i64 11, i1 false)
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 -1, i64 11, i32 1, i1 false)
   ret void
 }
 
@@ -241,7 +241,7 @@ define void @f25(i8 *%dest) {
 ; CHECK-DAG: mvghi 0(%r2), -1
 ; CHECK-DAG: mvhi 8(%r2), -1
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 -1, i32 12, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 -1, i32 12, i32 1, i1 false)
   ret void
 }
 
@@ -251,7 +251,7 @@ define void @f26(i8 *%dest) {
 ; CHECK-DAG: mvghi 0(%r2), -1
 ; CHECK-DAG: mvhi 8(%r2), -1
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 -1, i64 12, i1 false)
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 -1, i64 12, i32 1, i1 false)
   ret void
 }
 
@@ -261,7 +261,7 @@ define void @f27(i8 *%dest) {
 ; CHECK: mvi 0(%r2), 255
 ; CHECK: mvc 1(12,%r2), 0(%r2)
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 -1, i32 13, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 -1, i32 13, i32 1, i1 false)
   ret void
 }
 
@@ -271,7 +271,7 @@ define void @f28(i8 *%dest) {
 ; CHECK: mvi 0(%r2), 255
 ; CHECK: mvc 1(12,%r2), 0(%r2)
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 -1, i64 13, i1 false)
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 -1, i64 13, i32 1, i1 false)
   ret void
 }
 
@@ -281,7 +281,7 @@ define void @f29(i8 *%dest) {
 ; CHECK: mvi 0(%r2), 255
 ; CHECK: mvc 1(13,%r2), 0(%r2)
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 -1, i32 14, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 -1, i32 14, i32 1, i1 false)
   ret void
 }
 
@@ -291,7 +291,7 @@ define void @f30(i8 *%dest) {
 ; CHECK: mvi 0(%r2), 255
 ; CHECK: mvc 1(13,%r2), 0(%r2)
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 -1, i64 14, i1 false)
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 -1, i64 14, i32 1, i1 false)
   ret void
 }
 
@@ -301,7 +301,7 @@ define void @f31(i8 *%dest) {
 ; CHECK: mvi 0(%r2), 255
 ; CHECK: mvc 1(14,%r2), 0(%r2)
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 -1, i32 15, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 -1, i32 15, i32 1, i1 false)
   ret void
 }
 
@@ -311,7 +311,7 @@ define void @f32(i8 *%dest) {
 ; CHECK: mvi 0(%r2), 255
 ; CHECK: mvc 1(14,%r2), 0(%r2)
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 -1, i64 15, i1 false)
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 -1, i64 15, i32 1, i1 false)
   ret void
 }
 
@@ -321,7 +321,7 @@ define void @f33(i8 *%dest) {
 ; CHECK-DAG: mvghi 0(%r2), -1
 ; CHECK-DAG: mvghi 8(%r2), -1
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 -1, i32 16, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 -1, i32 16, i32 1, i1 false)
   ret void
 }
 
@@ -331,7 +331,7 @@ define void @f34(i8 *%dest) {
 ; CHECK-DAG: mvghi 0(%r2), -1
 ; CHECK-DAG: mvghi 8(%r2), -1
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 -1, i64 16, i1 false)
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 -1, i64 16, i32 1, i1 false)
   ret void
 }
 
@@ -341,7 +341,7 @@ define void @f35(i8 *%dest) {
 ; CHECK: mvi 0(%r2), 255
 ; CHECK: mvc 1(16,%r2), 0(%r2)
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 -1, i32 17, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 -1, i32 17, i32 1, i1 false)
   ret void
 }
 
@@ -351,7 +351,7 @@ define void @f36(i8 *%dest) {
 ; CHECK: mvi 0(%r2), 255
 ; CHECK: mvc 1(16,%r2), 0(%r2)
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 -1, i64 17, i1 false)
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 -1, i64 17, i32 1, i1 false)
   ret void
 }
 
@@ -361,7 +361,7 @@ define void @f37(i8 *%dest) {
 ; CHECK: mvi 0(%r2), 255
 ; CHECK: mvc 1(256,%r2), 0(%r2)
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 -1, i32 257, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 -1, i32 257, i32 1, i1 false)
   ret void
 }
 
@@ -371,7 +371,7 @@ define void @f38(i8 *%dest) {
 ; CHECK: mvi 0(%r2), 255
 ; CHECK: mvc 1(256,%r2), 0(%r2)
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 -1, i64 257, i1 false)
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 -1, i64 257, i32 1, i1 false)
   ret void
 }
 
@@ -382,7 +382,7 @@ define void @f39(i8 *%dest) {
 ; CHECK: mvc 1(256,%r2), 0(%r2)
 ; CHECK: mvc 257(1,%r2), 256(%r2)
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 -1, i32 258, i1 false)
+  call void @llvm.memset.p0i8.i32(i8 *%dest, i8 -1, i32 258, i32 1, i1 false)
   ret void
 }
 
@@ -393,6 +393,6 @@ define void @f40(i8 *%dest) {
 ; CHECK: mvc 1(256,%r2), 0(%r2)
 ; CHECK: mvc 257(1,%r2), 256(%r2)
 ; CHECK: br %r14
-  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 -1, i64 258, i1 false)
+  call void @llvm.memset.p0i8.i64(i8 *%dest, i8 -1, i64 258, i32 1, i1 false)
   ret void
 }
diff --git a/test/CodeGen/SystemZ/tail-call-mem-intrinsics.ll b/test/CodeGen/SystemZ/tail-call-mem-intrinsics.ll
index 0290d425e57..65cc394f8a9 100644
--- a/test/CodeGen/SystemZ/tail-call-mem-intrinsics.ll
+++ b/test/CodeGen/SystemZ/tail-call-mem-intrinsics.ll
@@ -4,7 +4,7 @@
 ; CHECK: jg memcpy
 define void @tail_memcpy(i8* nocapture %p, i8* nocapture readonly %q, i32 %n) #0 {
 entry:
-  tail call void @llvm.memcpy.p0i8.p0i8.i32(i8* %p, i8* %q, i32 %n, i1 false)
+  tail call void @llvm.memcpy.p0i8.p0i8.i32(i8* %p, i8* %q, i32 %n, i32 1, i1 false)
   ret void
 }
 
@@ -12,7 +12,7 @@ entry:
 ; CHECK: jg memmove
 define void @tail_memmove(i8* nocapture %p, i8* nocapture readonly %q, i32 %n) #0 {
 entry:
-  tail call void @llvm.memmove.p0i8.p0i8.i32(i8* %p, i8* %q, i32 %n, i1 false)
+  tail call void @llvm.memmove.p0i8.p0i8.i32(i8* %p, i8* %q, i32 %n, i32 1, i1 false)
   ret void
 }
 
@@ -20,12 +20,12 @@ entry:
 ; CHECK: jg memset
 define void @tail_memset(i8* nocapture %p, i8 %c, i32 %n) #0 {
 entry:
-  tail call void @llvm.memset.p0i8.i32(i8* %p, i8 %c, i32 %n, i1 false)
+  tail call void @llvm.memset.p0i8.i32(i8* %p, i8 %c, i32 %n, i32 1, i1 false)
   ret void
 }
 
-declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture readonly, i32, i1) #0
-declare void @llvm.memmove.p0i8.p0i8.i32(i8* nocapture, i8* nocapture readonly, i32, i1) #0
-declare void @llvm.memset.p0i8.i32(i8* nocapture, i8, i32, i1) #0
+declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture readonly, i32, i32, i1) #0
+declare void @llvm.memmove.p0i8.p0i8.i32(i8* nocapture, i8* nocapture readonly, i32, i32, i1) #0
+declare void @llvm.memset.p0i8.i32(i8* nocapture, i8, i32, i32, i1) #0
 
 attributes #0 = { nounwind }
diff --git a/test/CodeGen/Thumb/2011-05-11-DAGLegalizer.ll b/test/CodeGen/Thumb/2011-05-11-DAGLegalizer.ll
index 2ddf7854193..d8e165145bd 100644
--- a/test/CodeGen/Thumb/2011-05-11-DAGLegalizer.ll
+++ b/test/CodeGen/Thumb/2011-05-11-DAGLegalizer.ll
@@ -29,25 +29,25 @@ do.body:                                          ; preds = %entry
   %arrayidx = getelementptr inbounds [4 x %struct.RRRRRRRR], [4 x %struct.RRRRRRRR]* %eph, i32 0, i32 0
   %tmp2 = bitcast %struct.RRRRRRRR* %agg.tmp to i8*
   %tmp3 = bitcast %struct.RRRRRRRR* %arrayidx to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %tmp2, i8* %tmp3, i32 312, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %tmp2, i8* %tmp3, i32 312, i32 4, i1 false)
   %tmp5 = load %struct.MMMMMMMMMMMM*, %struct.MMMMMMMMMMMM** %aidData.addr
   %eph6 = getelementptr inbounds %struct.MMMMMMMMMMMM, %struct.MMMMMMMMMMMM* %tmp5, i32 0, i32 0
   %arrayidx7 = getelementptr inbounds [4 x %struct.RRRRRRRR], [4 x %struct.RRRRRRRR]* %eph6, i32 0, i32 1
   %tmp8 = bitcast %struct.RRRRRRRR* %agg.tmp4 to i8*
   %tmp9 = bitcast %struct.RRRRRRRR* %arrayidx7 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %tmp8, i8* %tmp9, i32 312, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %tmp8, i8* %tmp9, i32 312, i32 4, i1 false)
   %tmp11 = load %struct.MMMMMMMMMMMM*, %struct.MMMMMMMMMMMM** %aidData.addr
   %eph12 = getelementptr inbounds %struct.MMMMMMMMMMMM, %struct.MMMMMMMMMMMM* %tmp11, i32 0, i32 0
   %arrayidx13 = getelementptr inbounds [4 x %struct.RRRRRRRR], [4 x %struct.RRRRRRRR]* %eph12, i32 0, i32 2
   %tmp14 = bitcast %struct.RRRRRRRR* %agg.tmp10 to i8*
   %tmp15 = bitcast %struct.RRRRRRRR* %arrayidx13 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %tmp14, i8* %tmp15, i32 312, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %tmp14, i8* %tmp15, i32 312, i32 4, i1 false)
   %tmp17 = load %struct.MMMMMMMMMMMM*, %struct.MMMMMMMMMMMM** %aidData.addr
   %eph18 = getelementptr inbounds %struct.MMMMMMMMMMMM, %struct.MMMMMMMMMMMM* %tmp17, i32 0, i32 0
   %arrayidx19 = getelementptr inbounds [4 x %struct.RRRRRRRR], [4 x %struct.RRRRRRRR]* %eph18, i32 0, i32 3
   %tmp20 = bitcast %struct.RRRRRRRR* %agg.tmp16 to i8*
   %tmp21 = bitcast %struct.RRRRRRRR* %arrayidx19 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %tmp20, i8* %tmp21, i32 312, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %tmp20, i8* %tmp21, i32 312, i32 4, i1 false)
   call void (i8*, i32, i8*, i8*, ...) @CLLoggingLog(i8* %tmp, i32 2, i8* getelementptr inbounds ([62 x i8], [62 x i8]* @__PRETTY_FUNCTION__._ZN12CLGll, i32 0, i32 0), i8* getelementptr inbounds ([75 x i8], [75 x i8]* @.str, i32 0, i32 0), %struct.RRRRRRRR* byval %agg.tmp, %struct.RRRRRRRR* byval %agg.tmp4, %struct.RRRRRRRR* byval %agg.tmp10, %struct.RRRRRRRR* byval %agg.tmp16)
   br label %do.end
 
@@ -57,4 +57,4 @@ do.end:                                           ; preds = %do.body
 
 declare void @CLLoggingLog(i8*, i32, i8*, i8*, ...)
 
-declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i32, i1) nounwind
diff --git a/test/CodeGen/Thumb/dyn-stackalloc.ll b/test/CodeGen/Thumb/dyn-stackalloc.ll
index ba323b2cda7..c94c904e4cd 100644
--- a/test/CodeGen/Thumb/dyn-stackalloc.ll
+++ b/test/CodeGen/Thumb/dyn-stackalloc.ll
@@ -61,7 +61,7 @@ define void @t2(%struct.comment* %vc, i8* %tag, i8* %contents) {
 	%tmp9 = call i8* @strcpy( i8* %tmp6, i8* %tag )
 	%tmp6.len = call i32 @strlen( i8* %tmp6 )
 	%tmp6.indexed = getelementptr i8, i8* %tmp6, i32 %tmp6.len
-	call void @llvm.memcpy.p0i8.p0i8.i32(i8* %tmp6.indexed, i8* getelementptr inbounds ([2 x i8], [2 x i8]* @str215, i32 0, i32 0), i32 2, i1 false)
+	call void @llvm.memcpy.p0i8.p0i8.i32(i8* %tmp6.indexed, i8* getelementptr inbounds ([2 x i8], [2 x i8]* @str215, i32 0, i32 0), i32 2, i32 1, i1 false)
 	%tmp15 = call i8* @strcat( i8* %tmp6, i8* %contents )
 	call fastcc void @comment_add( %struct.comment* %vc, i8* %tmp6 )
 	ret void
@@ -73,6 +73,6 @@ declare i8* @strcat(i8*, i8*)
 
 declare fastcc void @comment_add(%struct.comment*, i8*)
 
-declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i32, i1) nounwind
 
 declare i8* @strcpy(i8*, i8*)
diff --git a/test/CodeGen/Thumb/ldm-stm-base-materialization.ll b/test/CodeGen/Thumb/ldm-stm-base-materialization.ll
index 355fe804ebc..0be796eb8f8 100644
--- a/test/CodeGen/Thumb/ldm-stm-base-materialization.ll
+++ b/test/CodeGen/Thumb/ldm-stm-base-materialization.ll
@@ -23,7 +23,7 @@ entry:
   %2 = load i32*, i32** @b, align 4
   %arrayidx1 = getelementptr inbounds i32, i32* %2, i32 1
   %3 = bitcast i32* %arrayidx1 to i8*
-  tail call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 4 %1, i8* align 4 %3, i32 24, i1 false)
+  tail call void @llvm.memcpy.p0i8.p0i8.i32(i8* %1, i8* %3, i32 24, i32 4, i1 false)
   ret void
 }
 
@@ -44,7 +44,7 @@ entry:
   %2 = load i32*, i32** @b, align 4
   %arrayidx1 = getelementptr inbounds i32, i32* %2, i32 1
   %3 = bitcast i32* %arrayidx1 to i8*
-  tail call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 4 %1, i8* align 4 %3, i32 28, i1 false)
+  tail call void @llvm.memcpy.p0i8.p0i8.i32(i8* %1, i8* %3, i32 28, i32 4, i1 false)
   ret void
 }
 
@@ -65,7 +65,7 @@ entry:
   %2 = load i32*, i32** @b, align 4
   %arrayidx1 = getelementptr inbounds i32, i32* %2, i32 1
   %3 = bitcast i32* %arrayidx1 to i8*
-  tail call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 4 %1, i8* align 4 %3, i32 32, i1 false)
+  tail call void @llvm.memcpy.p0i8.p0i8.i32(i8* %1, i8* %3, i32 32, i32 4, i1 false)
   ret void
 }
 
@@ -88,9 +88,9 @@ entry:
   %2 = load i32*, i32** @b, align 4
   %arrayidx1 = getelementptr inbounds i32, i32* %2, i32 1
   %3 = bitcast i32* %arrayidx1 to i8*
-  tail call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 4 %1, i8* align 4 %3, i32 36, i1 false)
+  tail call void @llvm.memcpy.p0i8.p0i8.i32(i8* %1, i8* %3, i32 36, i32 4, i1 false)
   ret void
 }
 
 ; Function Attrs: nounwind
-declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture readonly, i32, i1) #1
+declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture readonly, i32, i32, i1) #1
diff --git a/test/CodeGen/Thumb/stack-coloring-without-frame-ptr.ll b/test/CodeGen/Thumb/stack-coloring-without-frame-ptr.ll
index 06763f9a53b..97c66d9dc86 100644
--- a/test/CodeGen/Thumb/stack-coloring-without-frame-ptr.ll
+++ b/test/CodeGen/Thumb/stack-coloring-without-frame-ptr.ll
@@ -13,7 +13,7 @@ entry:
   %0 = bitcast %deque* %var3 to i8*
   %1 = bitcast %iterator* %var1 to i8*
   call void @llvm.lifetime.start(i64 16, i8* %1) nounwind
-  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %1, i8* %0, i32 16, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %1, i8* %0, i32 16, i32 4, i1 false)
   call void @llvm.lifetime.end(i64 16, i8* %1) nounwind
 
   %2 = bitcast %insert_iterator* %var2 to i8*
@@ -22,7 +22,7 @@ entry:
   ret i32 0
 }
 
-declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i32, i1) nounwind
 
 declare void @llvm.lifetime.start(i64, i8* nocapture) nounwind
 
diff --git a/test/CodeGen/Thumb2/2009-08-04-SubregLoweringBug.ll b/test/CodeGen/Thumb2/2009-08-04-SubregLoweringBug.ll
index 951cf40e43b..779e100d419 100644
--- a/test/CodeGen/Thumb2/2009-08-04-SubregLoweringBug.ll
+++ b/test/CodeGen/Thumb2/2009-08-04-SubregLoweringBug.ll
@@ -12,7 +12,7 @@ entry:
 	br i1 undef, label %bb, label %bb6.preheader
 
 bb6.preheader:		; preds = %entry
-        call void @llvm.memcpy.p0i8.p0i8.i32(i8* undef, i8* undef, i32 12, i1 false)
+        call void @llvm.memcpy.p0i8.p0i8.i32(i8* undef, i8* undef, i32 12, i32 4, i1 false)
 	br i1 undef, label %bb15, label %bb13
 
 bb:		; preds = %entry
@@ -30,4 +30,4 @@ bb15:		; preds = %bb13, %bb6.preheader
 	ret void
 }
 
-declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i32, i1) nounwind
diff --git a/test/CodeGen/Thumb2/2012-01-13-CBNZBug.ll b/test/CodeGen/Thumb2/2012-01-13-CBNZBug.ll
index ff41777b77f..9121044be4f 100644
--- a/test/CodeGen/Thumb2/2012-01-13-CBNZBug.ll
+++ b/test/CodeGen/Thumb2/2012-01-13-CBNZBug.ll
@@ -9,7 +9,7 @@
 
 @lookup_list = external hidden unnamed_addr global %struct.Dict_node_struct*, align 4
 
-declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i32, i1) nounwind
 
 define hidden fastcc void @rdictionary_lookup(%struct.Dict_node_struct* %dn, i8* nocapture %s) nounwind ssp {
 ; CHECK-LABEL: rdictionary_lookup:
@@ -78,7 +78,7 @@ if.then5:                                         ; preds = %if.end3
   %call6 = tail call fastcc i8* @xalloc(i32 20)
   %5 = bitcast i8* %call6 to %struct.Dict_node_struct*
   %6 = bitcast %struct.Dict_node_struct* %dn.tr to i8*
-  tail call void @llvm.memcpy.p0i8.p0i8.i32(i8* %call6, i8* %6, i32 16, i1 false)
+  tail call void @llvm.memcpy.p0i8.p0i8.i32(i8* %call6, i8* %6, i32 16, i32 4, i1 false)
   %7 = load %struct.Dict_node_struct*, %struct.Dict_node_struct** @lookup_list, align 4
   %right7 = getelementptr inbounds i8, i8* %call6, i32 16
   %8 = bitcast i8* %right7 to %struct.Dict_node_struct**
diff --git a/test/CodeGen/X86/2009-01-25-NoSSE.ll b/test/CodeGen/X86/2009-01-25-NoSSE.ll
index bc108c82af8..c655f2c374a 100644
--- a/test/CodeGen/X86/2009-01-25-NoSSE.ll
+++ b/test/CodeGen/X86/2009-01-25-NoSSE.ll
@@ -12,11 +12,11 @@ define void @foo() nounwind {
 entry:
   %termios = alloca %struct.ktermios, align 8
   %termios1 = bitcast %struct.ktermios* %termios to i8*
-  call void @llvm.memset.p0i8.i64(i8* %termios1, i8 0, i64 44, i1 false)
+  call void @llvm.memset.p0i8.i64(i8* %termios1, i8 0, i64 44, i32 8, i1 false)
   call void @bar(%struct.ktermios* %termios) nounwind
   ret void
 }
 
 declare void @bar(%struct.ktermios*)
 
-declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i1) nounwind
+declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i32, i1) nounwind
diff --git a/test/CodeGen/X86/2009-11-16-UnfoldMemOpBug.ll b/test/CodeGen/X86/2009-11-16-UnfoldMemOpBug.ll
index b58ee5be82e..1cfd108db65 100644
--- a/test/CodeGen/X86/2009-11-16-UnfoldMemOpBug.ll
+++ b/test/CodeGen/X86/2009-11-16-UnfoldMemOpBug.ll
@@ -17,7 +17,7 @@ bb1:
 ; CHECK: movups %xmm0, 12(%rsp)
 ; CHECK: movaps %xmm1, (%rsp)
   %tmp2 = phi i32 [ %tmp3, %bb1 ], [ 0, %entry ]
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %tmp1, i8* getelementptr inbounds ([28 x i8], [28 x i8]* @str, i64 0, i64 0), i64 28, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %tmp1, i8* getelementptr inbounds ([28 x i8], [28 x i8]* @str, i64 0, i64 0), i64 28, i32 1, i1 false)
   %tmp3 = add i32 %tmp2, 1
   %tmp4 = icmp eq i32 %tmp3, %count
   br i1 %tmp4, label %bb2, label %bb1
@@ -26,4 +26,4 @@ bb2:
   ret void
 }
 
-declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i32, i1) nounwind
diff --git a/test/CodeGen/X86/2010-04-08-CoalescerBug.ll b/test/CodeGen/X86/2010-04-08-CoalescerBug.ll
index 52de365464f..5adf99e3e47 100644
--- a/test/CodeGen/X86/2010-04-08-CoalescerBug.ll
+++ b/test/CodeGen/X86/2010-04-08-CoalescerBug.ll
@@ -19,8 +19,8 @@ entry:
   %tmp4 = getelementptr inbounds %struct.FC, %struct.FC* %tmp3, i64 0, i32 1, i64 0
   %tmp5 = bitcast [32 x i32]* %BitValueArray to i8*
   %tmp6 = bitcast i32* %tmp4 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %tmp5, i8* %tmp6, i64 128, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %tmp5, i8* %tmp6, i64 128, i32 4, i1 false)
   unreachable
 }
 
-declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i32, i1) nounwind
diff --git a/test/CodeGen/X86/2010-04-21-CoalescerBug.ll b/test/CodeGen/X86/2010-04-21-CoalescerBug.ll
index f2d7fd10fb3..d5987645cfc 100644
--- a/test/CodeGen/X86/2010-04-21-CoalescerBug.ll
+++ b/test/CodeGen/X86/2010-04-21-CoalescerBug.ll
@@ -8,8 +8,8 @@
 define void @t(%struct.CMTimeMapping* noalias nocapture sret %agg.result) nounwind optsize ssp {
 entry:
   %agg.result1 = bitcast %struct.CMTimeMapping* %agg.result to i8* ; <i8*> [#uses=1]
-  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %agg.result1, i8* null, i64 96, i1 false)
+  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %agg.result1, i8* null, i64 96, i32 4, i1 false)
   ret void
 }
 
-declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i32, i1) nounwind
diff --git a/test/CodeGen/X86/2010-06-25-CoalescerSubRegDefDead.ll b/test/CodeGen/X86/2010-06-25-CoalescerSubRegDefDead.ll
index dd7c3fa571c..ffb51572a30 100644
--- a/test/CodeGen/X86/2010-06-25-CoalescerSubRegDefDead.ll
+++ b/test/CodeGen/X86/2010-06-25-CoalescerSubRegDefDead.ll
@@ -26,7 +26,7 @@ bb:
 ; CHECK: rep;stosl
 
   %tmp5 = bitcast i32* %tmp4 to i8*
-  call void @llvm.memset.p0i8.i64(i8* align 4 %tmp5, i8 0, i64 84, i1 false)
+  call void @llvm.memset.p0i8.i64(i8* %tmp5, i8 0, i64 84, i32 4, i1 false)
   %tmp6 = getelementptr inbounds %struct.type, %struct.type* %s, i32 0, i32 62
   store i32* null, i32** %tmp6, align 8
   br label %bb1
@@ -36,4 +36,4 @@ bb1:
   ret i32 42
 }
 
-declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i1) nounwind
+declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i32, i1) nounwind
diff --git a/test/CodeGen/X86/2010-09-17-SideEffectsInChain.ll b/test/CodeGen/X86/2010-09-17-SideEffectsInChain.ll
index 6ddd4797b32..b7380196bd9 100644
--- a/test/CodeGen/X86/2010-09-17-SideEffectsInChain.ll
+++ b/test/CodeGen/X86/2010-09-17-SideEffectsInChain.ll
@@ -2,7 +2,7 @@
 
 target datalayout = "e-p:64:64:64-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f32:32:32-f64:64:64-v64:64:64-v128:128:128-a0:0:64-s0:64:64-f80:128:128-n8:16:32:64"
 target triple = "x86_64-apple-darwin10.4"
-declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i32, i1) nounwind
 
 define fastcc i32 @cli_magic_scandesc(i8* %in) nounwind ssp {
 entry:
@@ -12,7 +12,7 @@ entry:
   %d = load i8, i8* %b, align 8
   %e = load i8, i8* %c, align 8
   %f = bitcast [64 x i8]* %a to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %f, i8* %in, i64 64, i1 false) nounwind
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %f, i8* %in, i64 64, i32 8, i1 false) nounwind
   store i8 %d, i8* %b, align 8
   store i8 %e, i8* %c, align 8
   ret i32 0
diff --git a/test/CodeGen/X86/2012-01-10-UndefExceptionEdge.ll b/test/CodeGen/X86/2012-01-10-UndefExceptionEdge.ll
index 97a33893fa0..20615afdfa1 100644
--- a/test/CodeGen/X86/2012-01-10-UndefExceptionEdge.ll
+++ b/test/CodeGen/X86/2012-01-10-UndefExceptionEdge.ll
@@ -14,7 +14,7 @@ target triple = "i386-apple-macosx10.7"
 
 @Exception = external unnamed_addr constant { i8*, i8* }
 
-declare void @llvm.memset.p0i8.i32(i8* nocapture, i8, i32, i1) nounwind
+declare void @llvm.memset.p0i8.i32(i8* nocapture, i8, i32, i32, i1) nounwind
 
 define void @f(i32* nocapture %arg, i32* nocapture %arg1, i32* nocapture %arg2, i32* nocapture %arg3, i32 %arg4, i32 %arg5) optsize ssp personality i8* bitcast (i32 (...)* @__gxx_personality_v0 to i8*) {
 bb:
@@ -85,7 +85,7 @@ bb41:                                             ; preds = %bb38
           to label %bb42 unwind label %bb20
 
 bb42:                                             ; preds = %bb41
-  tail call void @llvm.memset.p0i8.i32(i8* %tmp32, i8 0, i32 %tmp9, i1 false) nounwind
+  tail call void @llvm.memset.p0i8.i32(i8* %tmp32, i8 0, i32 %tmp9, i32 1, i1 false) nounwind
   br i1 %tmp35, label %bb43, label %bb45
 
 bb43:                                             ; preds = %bb42
@@ -101,7 +101,7 @@ bb45:                                             ; preds = %bb57, %bb42
   br i1 %tmp47, label %bb48, label %bb59
 
 bb48:                                             ; preds = %bb45
-  tail call void @llvm.memset.p0i8.i32(i8* %tmp32, i8 0, i32 %tmp9, i1 false) nounwind
+  tail call void @llvm.memset.p0i8.i32(i8* %tmp32, i8 0, i32 %tmp9, i32 1, i1 false) nounwind
   br i1 %tmp36, label %bb49, label %bb57
 
 bb49:                                             ; preds = %bb49, %bb48
@@ -120,7 +120,7 @@ bb57:                                             ; preds = %bb49, %bb48
 
 bb59:                                             ; preds = %bb45
   %tmp60 = ashr i32 %tmp46, 31
-  tail call void @llvm.memset.p0i8.i32(i8* null, i8 0, i32 %tmp37, i1 false) nounwind
+  tail call void @llvm.memset.p0i8.i32(i8* null, i8 0, i32 %tmp37, i32 1, i1 false) nounwind
   br i1 %tmp36, label %bb61, label %bb67
 
 bb61:                                             ; preds = %bb61, %bb59
diff --git a/test/CodeGen/X86/alignment-2.ll b/test/CodeGen/X86/alignment-2.ll
index f66960c675b..a38a3626702 100644
--- a/test/CodeGen/X86/alignment-2.ll
+++ b/test/CodeGen/X86/alignment-2.ll
@@ -23,8 +23,8 @@ bb:
 ; CHECK-NOT: movaps {{[0-9]*}}(%{{[a-z]*}}), {{%xmm[0-9]}}
   %myopt = alloca %struct.printQueryOpt, align 4
   %tmp = bitcast %struct.printQueryOpt* %myopt to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %tmp, i8* bitcast (%struct.printQueryOpt* getelementptr inbounds (%struct._psqlSettings, %struct._psqlSettings* @pset, i32 0, i32 4) to i8*), i32 76, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %tmp, i8* bitcast (%struct.printQueryOpt* getelementptr inbounds (%struct._psqlSettings, %struct._psqlSettings* @pset, i32 0, i32 4) to i8*), i32 76, i32 4, i1 false)
   ret i8 0
 }
 
-declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i32, i1) nounwind
diff --git a/test/CodeGen/X86/darwin-bzero.ll b/test/CodeGen/X86/darwin-bzero.ll
index 1f2c5fe5f2a..3099526028a 100644
--- a/test/CodeGen/X86/darwin-bzero.ll
+++ b/test/CodeGen/X86/darwin-bzero.ll
@@ -1,8 +1,8 @@
 ; RUN: llc < %s -mtriple=i386-apple-darwin10 | grep __bzero
 
-declare void @llvm.memset.p0i8.i32(i8* nocapture, i8, i32, i1) nounwind
+declare void @llvm.memset.p0i8.i32(i8* nocapture, i8, i32, i32, i1) nounwind
 
 define void @foo(i8* %p, i32 %len) {
-  call void @llvm.memset.p0i8.i32(i8* %p, i8 0, i32 %len, i1 false)
+  call void @llvm.memset.p0i8.i32(i8* %p, i8 0, i32 %len, i32 1, i1 false)
   ret void
 }
diff --git a/test/CodeGen/X86/fast-isel-call.ll b/test/CodeGen/X86/fast-isel-call.ll
index 439812ffb8c..9fd07b521ab 100644
--- a/test/CodeGen/X86/fast-isel-call.ll
+++ b/test/CodeGen/X86/fast-isel-call.ll
@@ -30,10 +30,10 @@ define void @test2(%struct.s* %d) nounwind {
 ; CHECK: movl {{.*}}, 8(%esp)
 }
 
-declare void @llvm.memset.p0i8.i32(i8* nocapture, i8, i32, i1) nounwind
+declare void @llvm.memset.p0i8.i32(i8* nocapture, i8, i32, i32, i1) nounwind
 
 define void @test3(i8* %a) {
-  call void @llvm.memset.p0i8.i32(i8* %a, i8 0, i32 100, i1 false)
+  call void @llvm.memset.p0i8.i32(i8* %a, i8 0, i32 100, i32 1, i1 false)
   ret void
 ; CHECK-LABEL: test3:
 ; CHECK:   movl	{{.*}}, (%esp)
@@ -42,10 +42,10 @@ define void @test3(i8* %a) {
 ; CHECK:   calll {{.*}}memset
 }
 
-declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i32, i1) nounwind
 
 define void @test4(i8* %a, i8* %b) {
-  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %a, i8* %b, i32 100, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %a, i8* %b, i32 100, i32 1, i1 false)
   ret void
 ; CHECK-LABEL: test4:
 ; CHECK:   movl	{{.*}}, (%esp)
diff --git a/test/CodeGen/X86/fast-isel-x86-64.ll b/test/CodeGen/X86/fast-isel-x86-64.ll
index b6235ac6668..d748cba2f8f 100644
--- a/test/CodeGen/X86/fast-isel-x86-64.ll
+++ b/test/CodeGen/X86/fast-isel-x86-64.ll
@@ -171,11 +171,11 @@ entry:
 ; CHECK: callq
 }
 
-declare void @llvm.memcpy.p0i8.p0i8.i64(i8*, i8*, i64, i1)
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8*, i8*, i64, i32, i1)
 
 ; rdar://9289488 - fast-isel shouldn't bail out on llvm.memcpy
 define void @test15(i8* %a, i8* %b) nounwind {
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %a, i8* %b, i64 4, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %a, i8* %b, i64 4, i32 4, i1 false)
   ret void
 ; CHECK-LABEL: test15:
 ; CHECK-NEXT: movl	(%rsi), %eax
diff --git a/test/CodeGen/X86/force-align-stack-alloca.ll b/test/CodeGen/X86/force-align-stack-alloca.ll
index daaadf32b98..d0cf3417008 100644
--- a/test/CodeGen/X86/force-align-stack-alloca.ll
+++ b/test/CodeGen/X86/force-align-stack-alloca.ll
@@ -61,10 +61,10 @@ entry:
 
 if.then:
   %0 = alloca i8, i32 %i
-  call void @llvm.memset.p0i8.i32(i8* %0, i8 0, i32 %i, i1 false)
+  call void @llvm.memset.p0i8.i32(i8* %0, i8 0, i32 %i, i32 1, i1 false)
   %call = call i32 @f(i8* %0)
   %conv = sext i32 %call to i64
   ret i64 %conv
 }
 
-declare void @llvm.memset.p0i8.i32(i8*, i8, i32, i1) nounwind
+declare void @llvm.memset.p0i8.i32(i8*, i8, i32, i32, i1) nounwind
diff --git a/test/CodeGen/X86/immediate_merging.ll b/test/CodeGen/X86/immediate_merging.ll
index 27b6b5a3971..8aef9c279b3 100644
--- a/test/CodeGen/X86/immediate_merging.ll
+++ b/test/CodeGen/X86/immediate_merging.ll
@@ -64,7 +64,7 @@ entry:
   ret i32 0
 }
 
-declare void @llvm.memset.p0i8.i32(i8* nocapture, i8, i32, i1) #1
+declare void @llvm.memset.p0i8.i32(i8* nocapture, i8, i32, i32, i1) #1
 
 @AA = common global [100 x i8] zeroinitializer, align 1
 
@@ -77,6 +77,6 @@ define void @foomemset() optsize {
 ; CHECK: mov{{l|q}} %{{e|r}}ax, AA
 
 entry:
-  call void @llvm.memset.p0i8.i32(i8* getelementptr inbounds ([100 x i8], [100 x i8]* @AA, i32 0, i32 0), i8 33, i32 24, i1 false)
+  call void @llvm.memset.p0i8.i32(i8* getelementptr inbounds ([100 x i8], [100 x i8]* @AA, i32 0, i32 0), i8 33, i32 24, i32 1, i1 false)
   ret void
 }
diff --git a/test/CodeGen/X86/load-slice.ll b/test/CodeGen/X86/load-slice.ll
index f07d042b28d..2f90f819d47 100644
--- a/test/CodeGen/X86/load-slice.ll
+++ b/test/CodeGen/X86/load-slice.ll
@@ -70,7 +70,7 @@ entry:
 }
 
 ; Function Attrs: nounwind
-declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture readonly, i64, i1) #1
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture readonly, i64, i32, i1) #1
 
 ; Function Attrs: nounwind
 declare void @llvm.lifetime.start(i64, i8* nocapture)
diff --git a/test/CodeGen/X86/lsr-normalization.ll b/test/CodeGen/X86/lsr-normalization.ll
index 295839c1e35..09c892c9fc8 100644
--- a/test/CodeGen/X86/lsr-normalization.ll
+++ b/test/CodeGen/X86/lsr-normalization.ll
@@ -21,7 +21,7 @@ define i32 @main(i32 %arg, i8** nocapture %arg1) nounwind {
 bb:
   %tmp = alloca %0, align 8                       ; <%0*> [#uses=11]
   %tmp2 = bitcast %0* %tmp to i8*                 ; <i8*> [#uses=1]
-  call void @llvm.memset.p0i8.i64(i8* %tmp2, i8 0, i64 16, i1 false) nounwind
+  call void @llvm.memset.p0i8.i64(i8* %tmp2, i8 0, i64 16, i32 8, i1 false) nounwind
   %tmp3 = getelementptr inbounds %0, %0* %tmp, i64 0, i32 0 ; <%0**> [#uses=3]
   store %0* %tmp, %0** %tmp3
   %tmp4 = getelementptr inbounds %0, %0* %tmp, i64 0, i32 1 ; <%0**> [#uses=1]
@@ -98,7 +98,7 @@ declare void @_ZNSt15_List_node_base4hookEPS_(%0*, %0*)
 
 declare noalias i8* @_Znwm(i64)
 
-declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i1) nounwind
+declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i32, i1) nounwind
 
 declare void @_ZdlPv(i8*) nounwind
 
diff --git a/test/CodeGen/X86/mem-intrin-base-reg.ll b/test/CodeGen/X86/mem-intrin-base-reg.ll
index 609cb216c0a..9bace29e185 100644
--- a/test/CodeGen/X86/mem-intrin-base-reg.ll
+++ b/test/CodeGen/X86/mem-intrin-base-reg.ll
@@ -13,11 +13,11 @@ target triple = "i686-pc-windows-msvc"
 ; shouldn't be aligned.
 
 declare void @escape_vla_and_icmp(i8*, i1 zeroext)
-declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture readonly, i32, i1)
-declare void @llvm.memset.p0i8.i32(i8* nocapture, i8, i32, i1)
+declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture readonly, i32, i32, i1)
+declare void @llvm.memset.p0i8.i32(i8* nocapture, i8, i32, i32, i1)
 
 define i32 @memcpy_novla_vector(<4 x i32>* %vp0, i8* %a, i8* %b, i32 %n, i1 zeroext %cond) {
-  call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 4 %a, i8* align 4 %b, i32 128, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %a, i8* %b, i32 128, i32 4, i1 false)
   br i1 %cond, label %spill_vectors, label %no_vectors
 
 no_vectors:
@@ -42,7 +42,7 @@ spill_vectors:
 ; CHECK: rep;movsl
 
 define i32 @memcpy_vla_vector(<4 x i32>* %vp0, i8* %a, i8* %b, i32 %n, i1 zeroext %cond) {
-  call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 4 %a, i8* align 4 %b, i32 128, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %a, i8* %b, i32 128, i32 4, i1 false)
   br i1 %cond, label %spill_vectors, label %no_vectors
 
 no_vectors:
@@ -70,7 +70,7 @@ spill_vectors:
 ; stosd doesn't clobber esi, so we can use it.
 
 define i32 @memset_vla_vector(<4 x i32>* %vp0, i8* %a, i32 %n, i1 zeroext %cond) {
-  call void @llvm.memset.p0i8.i32(i8* align 4 %a, i8 42, i32 128, i1 false)
+  call void @llvm.memset.p0i8.i32(i8* %a, i8 42, i32 128, i32 4, i1 false)
   br i1 %cond, label %spill_vectors, label %no_vectors
 
 no_vectors:
diff --git a/test/CodeGen/X86/memcpy-2.ll b/test/CodeGen/X86/memcpy-2.ll
index 7e8b15e6159..7ef61c9a677 100644
--- a/test/CodeGen/X86/memcpy-2.ll
+++ b/test/CodeGen/X86/memcpy-2.ll
@@ -48,7 +48,7 @@ entry:
 ; X86-64: movq $0
   %tmp1 = alloca [25 x i8]
   %tmp2 = bitcast [25 x i8]* %tmp1 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %tmp2, i8* getelementptr inbounds ([25 x i8], [25 x i8]* @.str, i32 0, i32 0), i32 25, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %tmp2, i8* getelementptr inbounds ([25 x i8], [25 x i8]* @.str, i32 0, i32 0), i32 25, i32 1, i1 false)
   unreachable
 }
 
@@ -86,7 +86,7 @@ entry:
 ; X86-64: movaps %xmm0, (%rdi)
   %tmp2 = bitcast %struct.s0* %a to i8*           ; <i8*> [#uses=1]
   %tmp3 = bitcast %struct.s0* %b to i8*           ; <i8*> [#uses=1]
-  tail call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 16 %tmp2, i8* align 16 %tmp3, i32 16, i1 false)
+  tail call void @llvm.memcpy.p0i8.p0i8.i32(i8* %tmp2, i8* %tmp3, i32 16, i32 16, i1 false)
   ret void
 }
 
@@ -135,7 +135,7 @@ entry:
 ; X86-64: movq %rax, (%rdi)
   %tmp2 = bitcast %struct.s0* %a to i8*           ; <i8*> [#uses=1]
   %tmp3 = bitcast %struct.s0* %b to i8*           ; <i8*> [#uses=1]
-  tail call void @llvm.memcpy.p0i8.p0i8.i32(i8* %tmp2, i8* %tmp3, i32 16, i1 false)
+  tail call void @llvm.memcpy.p0i8.p0i8.i32(i8* %tmp2, i8* %tmp3, i32 16, i32 8, i1 false)
   ret void
 }
 
@@ -202,8 +202,8 @@ entry:
 
   %tmp1 = alloca [30 x i8]
   %tmp2 = bitcast [30 x i8]* %tmp1 to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %tmp2, i8* getelementptr inbounds ([30 x i8], [30 x i8]* @.str2, i32 0, i32 0), i32 30, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %tmp2, i8* getelementptr inbounds ([30 x i8], [30 x i8]* @.str2, i32 0, i32 0), i32 30, i32 1, i1 false)
   unreachable
 }
 
-declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i32, i1) nounwind
diff --git a/test/CodeGen/X86/memcpy.ll b/test/CodeGen/X86/memcpy.ll
index d0e1dda2e3b..00669443d6d 100644
--- a/test/CodeGen/X86/memcpy.ll
+++ b/test/CodeGen/X86/memcpy.ll
@@ -1,13 +1,13 @@
 ; RUN: llc < %s -mtriple=x86_64-unknown-linux-gnu -mcpu=core2 | FileCheck %s -check-prefix=LINUX
 ; RUN: llc < %s -mtriple=x86_64-apple-darwin -mcpu=core2 | FileCheck %s -check-prefix=DARWIN
 
-declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i32, i1) nounwind
 
 
 ; Variable memcpy's should lower to calls.
 define i8* @test1(i8* %a, i8* %b, i64 %n) nounwind {
 entry:
-	tail call void @llvm.memcpy.p0i8.p0i8.i64( i8* %a, i8* %b, i64 %n, i1 0 )
+	tail call void @llvm.memcpy.p0i8.p0i8.i64( i8* %a, i8* %b, i64 %n, i32 1, i1 0 )
 	ret i8* %a
         
 ; LINUX-LABEL: test1:
@@ -19,7 +19,7 @@ define i8* @test2(i64* %a, i64* %b, i64 %n) nounwind {
 entry:
 	%tmp14 = bitcast i64* %a to i8*
 	%tmp25 = bitcast i64* %b to i8*
-	tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %tmp14, i8* %tmp25, i64 %n, i1 0 )
+	tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %tmp14, i8* %tmp25, i64 %n, i32 8, i1 0 )
 	ret i8* %tmp14
         
 ; LINUX-LABEL: test2:
@@ -34,7 +34,7 @@ entry:
 ; rdar://8821501
 define void @test3(i8* nocapture %A, i8* nocapture %B) nounwind optsize noredzone {
 entry:
-  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %A, i8* %B, i64 64, i1 false)
+  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %A, i8* %B, i64 64, i32 1, i1 false)
   ret void
 ; LINUX-LABEL: test3:
 ; LINUX: memcpy
@@ -60,7 +60,7 @@ entry:
 }
 
 define void @test3_minsize(i8* nocapture %A, i8* nocapture %B) nounwind minsize noredzone {
-  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %A, i8* %B, i64 64, i1 false)
+  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %A, i8* %B, i64 64, i32 1, i1 false)
   ret void
 ; LINUX-LABEL: test3_minsize:
 ; LINUX: memcpy
@@ -70,7 +70,7 @@ define void @test3_minsize(i8* nocapture %A, i8* nocapture %B) nounwind minsize
 }
 
 define void @test3_minsize_optsize(i8* nocapture %A, i8* nocapture %B) nounwind optsize minsize noredzone {
-  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %A, i8* %B, i64 64, i1 false)
+  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %A, i8* %B, i64 64, i32 1, i1 false)
   ret void
 ; LINUX-LABEL: test3_minsize_optsize:
 ; LINUX: memcpy
@@ -82,7 +82,7 @@ define void @test3_minsize_optsize(i8* nocapture %A, i8* nocapture %B) nounwind
 ; Large constant memcpy's should be inlined when not optimizing for size.
 define void @test4(i8* nocapture %A, i8* nocapture %B) nounwind noredzone {
 entry:
-  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %A, i8* %B, i64 64, i1 false)
+  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %A, i8* %B, i64 64, i32 1, i1 false)
   ret void
 ; LINUX-LABEL: test4:
 ; LINUX: movq
@@ -104,7 +104,7 @@ entry:
 
 define void @test5(i8* nocapture %C) nounwind uwtable ssp {
 entry:
-  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %C, i8* getelementptr inbounds ([30 x i8], [30 x i8]* @.str, i64 0, i64 0), i64 16, i1 false)
+  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %C, i8* getelementptr inbounds ([30 x i8], [30 x i8]* @.str, i64 0, i64 0), i64 16, i32 1, i1 false)
   ret void
 
 ; DARWIN-LABEL: test5:
@@ -121,7 +121,7 @@ entry:
 ; DARWIN: test6
 ; DARWIN: movw $0, 8
 ; DARWIN: movq $120, 0
-  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* null, i8* getelementptr inbounds ([2 x i8], [2 x i8]* @.str2, i64 0, i64 0), i64 10, i1 false)
+  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* null, i8* getelementptr inbounds ([2 x i8], [2 x i8]* @.str2, i64 0, i64 0), i64 10, i32 1, i1 false)
   ret void
 }
 
@@ -135,6 +135,6 @@ define void @PR15348(i8* %a, i8* %b) {
 ; LINUX: movq
 ; LINUX: movq
 ; LINUX: movq
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %a, i8* %b, i64 17, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %a, i8* %b, i64 17, i32 0, i1 false)
   ret void
 }
diff --git a/test/CodeGen/X86/memset-2.ll b/test/CodeGen/X86/memset-2.ll
index c6a4bd99b5c..a87ef2e15a5 100644
--- a/test/CodeGen/X86/memset-2.ll
+++ b/test/CodeGen/X86/memset-2.ll
@@ -6,7 +6,7 @@ define fastcc void @t1() nounwind {
 entry:
 ; CHECK-LABEL: t1:
 ; CHECK: calll L_memset$stub
-  call void @llvm.memset.p0i8.i32(i8* null, i8 0, i32 188, i1 false)
+  call void @llvm.memset.p0i8.i32(i8* null, i8 0, i32 188, i32 1, i1 false)
   unreachable
 }
 
@@ -14,15 +14,15 @@ define fastcc void @t2(i8 signext %c) nounwind {
 entry:
 ; CHECK-LABEL: t2:
 ; CHECK: calll L_memset$stub
-  call void @llvm.memset.p0i8.i32(i8* undef, i8 %c, i32 76, i1 false)
+  call void @llvm.memset.p0i8.i32(i8* undef, i8 %c, i32 76, i32 1, i1 false)
   unreachable
 }
 
-declare void @llvm.memset.p0i8.i32(i8* nocapture, i8, i32, i1) nounwind
+declare void @llvm.memset.p0i8.i32(i8* nocapture, i8, i32, i32, i1) nounwind
 
 define void @t3(i8* nocapture %s, i8 %a) nounwind {
 entry:
-  tail call void @llvm.memset.p0i8.i32(i8* %s, i8 %a, i32 8, i1 false)
+  tail call void @llvm.memset.p0i8.i32(i8* %s, i8 %a, i32 8, i32 1, i1 false)
   ret void
 ; CHECK-LABEL: t3:
 ; CHECK: imull $16843009
@@ -30,7 +30,7 @@ entry:
 
 define void @t4(i8* nocapture %s, i8 %a) nounwind {
 entry:
-  tail call void @llvm.memset.p0i8.i32(i8* %s, i8 %a, i32 15, i1 false)
+  tail call void @llvm.memset.p0i8.i32(i8* %s, i8 %a, i32 15, i32 1, i1 false)
   ret void
 ; CHECK-LABEL: t4:
 ; CHECK: imull $16843009
diff --git a/test/CodeGen/X86/memset-3.ll b/test/CodeGen/X86/memset-3.ll
index 47c7ab99d29..455e6756013 100644
--- a/test/CodeGen/X86/memset-3.ll
+++ b/test/CodeGen/X86/memset-3.ll
@@ -5,8 +5,8 @@ define void @t() nounwind ssp {
 entry:
   %buf = alloca [512 x i8], align 1
   %ptr = getelementptr inbounds [512 x i8], [512 x i8]* %buf, i32 0, i32 0
-  call void @llvm.memset.p0i8.i32(i8* %ptr, i8 undef, i32 512, i1 false)
+  call void @llvm.memset.p0i8.i32(i8* %ptr, i8 undef, i32 512, i32 1, i1 false)
   unreachable
 }
 
-declare void @llvm.memset.p0i8.i32(i8* nocapture, i8, i32, i1) nounwind
+declare void @llvm.memset.p0i8.i32(i8* nocapture, i8, i32, i32, i1) nounwind
diff --git a/test/CodeGen/X86/memset-sse-stack-realignment.ll b/test/CodeGen/X86/memset-sse-stack-realignment.ll
index 503d375eee8..d77a7ed3816 100644
--- a/test/CodeGen/X86/memset-sse-stack-realignment.ll
+++ b/test/CodeGen/X86/memset-sse-stack-realignment.ll
@@ -9,7 +9,7 @@
 
 define void @test1(i32 %t) nounwind {
   %tmp1210 = alloca i8, i32 32, align 4
-  call void @llvm.memset.p0i8.i64(i8* %tmp1210, i8 0, i64 32, i1 false)
+  call void @llvm.memset.p0i8.i64(i8* %tmp1210, i8 0, i64 32, i32 4, i1 false)
   %x = alloca i8, i32 %t
   call void @dummy(i8* %x)
   ret void
@@ -42,7 +42,7 @@ define void @test1(i32 %t) nounwind {
 
 define void @test2(i32 %t) nounwind {
   %tmp1210 = alloca i8, i32 16, align 4
-  call void @llvm.memset.p0i8.i64(i8* %tmp1210, i8 0, i64 16, i1 false)
+  call void @llvm.memset.p0i8.i64(i8* %tmp1210, i8 0, i64 16, i32 4, i1 false)
   %x = alloca i8, i32 %t
   call void @dummy(i8* %x)
   ret void
@@ -74,4 +74,4 @@ define void @test2(i32 %t) nounwind {
 
 declare void @dummy(i8*)
 
-declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i1) nounwind
+declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i32, i1) nounwind
diff --git a/test/CodeGen/X86/memset.ll b/test/CodeGen/X86/memset.ll
index e96551aed72..96a22e88567 100644
--- a/test/CodeGen/X86/memset.ll
+++ b/test/CodeGen/X86/memset.ll
@@ -10,7 +10,7 @@ entry:
 	%up_mvd116 = getelementptr [8 x %struct.x], [8 x %struct.x]* %up_mvd, i32 0, i32 0		; <%struct.x*> [#uses=1]
 	%tmp110117 = bitcast [8 x %struct.x]* %up_mvd to i8*		; <i8*> [#uses=1]
 
-	call void @llvm.memset.p0i8.i64(i8* %tmp110117, i8 0, i64 32, i1 false)
+	call void @llvm.memset.p0i8.i64(i8* %tmp110117, i8 0, i64 32, i32 8, i1 false)
 ; X86: movl $0,
 ; X86: movl $0,
 ; X86: movl $0,
@@ -39,7 +39,7 @@ entry:
 
 declare void @foo(%struct.x*)
 
-declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i1) nounwind
+declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i32, i1) nounwind
 
 define void @PR15348(i8* %a) {
 ; Ensure that alignment of '0' in an @llvm.memset intrinsic results in
@@ -50,6 +50,6 @@ define void @PR15348(i8* %a) {
 ; XMM: movl $0,
 ; XMM: movl $0,
 ; XMM: movl $0,
-  call void @llvm.memset.p0i8.i64(i8* %a, i8 0, i64 17, i1 false)
+  call void @llvm.memset.p0i8.i64(i8* %a, i8 0, i64 17, i32 0, i1 false)
   ret void
 }
diff --git a/test/CodeGen/X86/memset64-on-x86-32.ll b/test/CodeGen/X86/memset64-on-x86-32.ll
index 2f2a76243fb..8cfa032797f 100644
--- a/test/CodeGen/X86/memset64-on-x86-32.ll
+++ b/test/CodeGen/X86/memset64-on-x86-32.ll
@@ -5,8 +5,8 @@
 
 define void @bork() nounwind {
 entry:
-  call void @llvm.memset.p0i8.i64(i8* null, i8 0, i64 80, i1 false)
+  call void @llvm.memset.p0i8.i64(i8* null, i8 0, i64 80, i32 4, i1 false)
   ret void
 }
 
-declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i1) nounwind
+declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i32, i1) nounwind
diff --git a/test/CodeGen/X86/misaligned-memset.ll b/test/CodeGen/X86/misaligned-memset.ll
index f7a6d577c08..ef8e0e81ad7 100644
--- a/test/CodeGen/X86/misaligned-memset.ll
+++ b/test/CodeGen/X86/misaligned-memset.ll
@@ -7,9 +7,9 @@ define i32 @main() nounwind ssp {
 entry:
   %retval = alloca i32, align 4
   store i32 0, i32* %retval
-  call void @llvm.memset.p0i8.i64(i8* bitcast (i64* getelementptr inbounds ([3 x i64], [3 x i64]* @a, i32 0, i64 1) to i8*), i8 0, i64 16, i1 false)
+  call void @llvm.memset.p0i8.i64(i8* bitcast (i64* getelementptr inbounds ([3 x i64], [3 x i64]* @a, i32 0, i64 1) to i8*), i8 0, i64 16, i32 1, i1 false)
   %0 = load i32, i32* %retval
   ret i32 %0
 }
 
-declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i1) nounwind
+declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i32, i1) nounwind
diff --git a/test/CodeGen/X86/misched-new.ll b/test/CodeGen/X86/misched-new.ll
index 8799b87e707..410a7f32064 100644
--- a/test/CodeGen/X86/misched-new.ll
+++ b/test/CodeGen/X86/misched-new.ll
@@ -11,7 +11,7 @@
 ; FIXME: There should be an assert in the coalescer that we're not rematting
 ; "not-quite-dead" copies, but that breaks a lot of tests <rdar://problem/11148682>.
 
-declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i32, i1) nounwind
 
 ; From oggenc.
 ; After coalescing, we have a dead superreg (RAX) definition.
@@ -24,7 +24,7 @@ entry:
   br i1 undef, label %for.cond.preheader, label %if.end
 
 for.cond.preheader:                               ; preds = %entry
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 4 undef, i8* align 4 null, i64 128, i1 false) nounwind
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* undef, i8* null, i64 128, i32 4, i1 false) nounwind
   unreachable
 
 if.end:                                           ; preds = %entry
diff --git a/test/CodeGen/X86/optimize-max-0.ll b/test/CodeGen/X86/optimize-max-0.ll
index 136fb611964..006592aaade 100644
--- a/test/CodeGen/X86/optimize-max-0.ll
+++ b/test/CodeGen/X86/optimize-max-0.ll
@@ -173,7 +173,7 @@ bb23:                                             ; preds = %bb24, %bb.nph
   %47 = mul i32 %y.21, %w
   %.sum5 = add i32 %47, %.sum3
   %48 = getelementptr i8, i8* %j, i32 %.sum5
-  tail call void @llvm.memcpy.p0i8.p0i8.i32(i8* %48, i8* %46, i32 %w, i1 false)
+  tail call void @llvm.memcpy.p0i8.p0i8.i32(i8* %48, i8* %46, i32 %w, i32 1, i1 false)
   br label %bb24
 
 bb24:                                             ; preds = %bb23
@@ -190,7 +190,7 @@ bb26:                                             ; preds = %bb24.bb26_crit_edge
   %50 = getelementptr i8, i8* %j, i32 %.sum4
   %51 = mul i32 %x, %w
   %52 = sdiv i32 %51, 2
-  tail call void @llvm.memset.p0i8.i32(i8* %50, i8 -128, i32 %52, i1 false)
+  tail call void @llvm.memset.p0i8.i32(i8* %50, i8 -128, i32 %52, i32 1, i1 false)
   ret void
 
 bb29:                                             ; preds = %bb20, %entry
@@ -208,7 +208,7 @@ bb30:                                             ; preds = %bb31, %bb.nph11
   %57 = getelementptr i8, i8* %r, i32 %56
   %58 = mul i32 %y.310, %w
   %59 = getelementptr i8, i8* %j, i32 %58
-  tail call void @llvm.memcpy.p0i8.p0i8.i32(i8* %59, i8* %57, i32 %w, i1 false)
+  tail call void @llvm.memcpy.p0i8.p0i8.i32(i8* %59, i8* %57, i32 %w, i32 1, i1 false)
   br label %bb31
 
 bb31:                                             ; preds = %bb30
@@ -224,7 +224,7 @@ bb33:                                             ; preds = %bb31.bb33_crit_edge
   %61 = getelementptr i8, i8* %j, i32 %60
   %62 = mul i32 %x, %w
   %63 = sdiv i32 %62, 2
-  tail call void @llvm.memset.p0i8.i32(i8* %61, i8 -128, i32 %63, i1 false)
+  tail call void @llvm.memset.p0i8.i32(i8* %61, i8 -128, i32 %63, i32 1, i1 false)
   ret void
 
 return:                                           ; preds = %bb20
@@ -398,7 +398,7 @@ bb23:                                             ; preds = %bb24, %bb.nph
   %47 = mul i32 %y.21, %w
   %.sum5 = add i32 %47, %.sum3
   %48 = getelementptr i8, i8* %j, i32 %.sum5
-  tail call void @llvm.memcpy.p0i8.p0i8.i32(i8* %48, i8* %46, i32 %w, i1 false)
+  tail call void @llvm.memcpy.p0i8.p0i8.i32(i8* %48, i8* %46, i32 %w, i32 1, i1 false)
   br label %bb24
 
 bb24:                                             ; preds = %bb23
@@ -415,7 +415,7 @@ bb26:                                             ; preds = %bb24.bb26_crit_edge
   %50 = getelementptr i8, i8* %j, i32 %.sum4
   %51 = mul i32 %x, %w
   %52 = udiv i32 %51, 2
-  tail call void @llvm.memset.p0i8.i32(i8* %50, i8 -128, i32 %52, i1 false)
+  tail call void @llvm.memset.p0i8.i32(i8* %50, i8 -128, i32 %52, i32 1, i1 false)
   ret void
 
 bb29:                                             ; preds = %bb20, %entry
@@ -433,7 +433,7 @@ bb30:                                             ; preds = %bb31, %bb.nph11
   %57 = getelementptr i8, i8* %r, i32 %56
   %58 = mul i32 %y.310, %w
   %59 = getelementptr i8, i8* %j, i32 %58
-  tail call void @llvm.memcpy.p0i8.p0i8.i32(i8* %59, i8* %57, i32 %w, i1 false)
+  tail call void @llvm.memcpy.p0i8.p0i8.i32(i8* %59, i8* %57, i32 %w, i32 1, i1 false)
   br label %bb31
 
 bb31:                                             ; preds = %bb30
@@ -449,13 +449,13 @@ bb33:                                             ; preds = %bb31.bb33_crit_edge
   %61 = getelementptr i8, i8* %j, i32 %60
   %62 = mul i32 %x, %w
   %63 = udiv i32 %62, 2
-  tail call void @llvm.memset.p0i8.i32(i8* %61, i8 -128, i32 %63, i1 false)
+  tail call void @llvm.memset.p0i8.i32(i8* %61, i8 -128, i32 %63, i32 1, i1 false)
   ret void
 
 return:                                           ; preds = %bb20
   ret void
 }
 
-declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i32, i1) nounwind
 
-declare void @llvm.memset.p0i8.i32(i8* nocapture, i8, i32, i1) nounwind
+declare void @llvm.memset.p0i8.i32(i8* nocapture, i8, i32, i32, i1) nounwind
diff --git a/test/CodeGen/X86/pr11985.ll b/test/CodeGen/X86/pr11985.ll
index 02cb5094076..aae00de112d 100644
--- a/test/CodeGen/X86/pr11985.ll
+++ b/test/CodeGen/X86/pr11985.ll
@@ -24,7 +24,7 @@ define float @foo(i8* nocapture %buf, float %a, float %b) nounwind uwtable {
 ; NEHALEM-NEXT:    movups %xmm2, (%rdi)
 
 entry:
-  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %buf, i8* blockaddress(@foo, %out), i64 22, i1 false)
+  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %buf, i8* blockaddress(@foo, %out), i64 22, i32 1, i1 false)
   br label %out
 
 out:                                              ; preds = %entry
@@ -32,4 +32,4 @@ out:                                              ; preds = %entry
   ret float %add
 }
 
-declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i32, i1) nounwind
diff --git a/test/CodeGen/X86/pr14333.ll b/test/CodeGen/X86/pr14333.ll
index aac20d95159..89779302d7f 100644
--- a/test/CodeGen/X86/pr14333.ll
+++ b/test/CodeGen/X86/pr14333.ll
@@ -6,7 +6,7 @@ define void @bar(%foo* %zed) {
   %tmp2 = getelementptr inbounds %foo, %foo* %zed, i64 0, i32 1
   store i64 0, i64* %tmp2, align 8
   %tmp3 = bitcast %foo* %zed to i8*
-  call void @llvm.memset.p0i8.i64(i8* %tmp3, i8 0, i64 16, i1 false)
+  call void @llvm.memset.p0i8.i64(i8* %tmp3, i8 0, i64 16, i32 8, i1 false)
   ret void
 }
-declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i1) nounwind
+declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i32, i1) nounwind
diff --git a/test/CodeGen/X86/ragreedy-hoist-spill.ll b/test/CodeGen/X86/ragreedy-hoist-spill.ll
index 0a4c2bdcc8e..e7dda534956 100644
--- a/test/CodeGen/X86/ragreedy-hoist-spill.ll
+++ b/test/CodeGen/X86/ragreedy-hoist-spill.ll
@@ -63,7 +63,7 @@ SyTime.exit2720:
   br i1 %cmp293427, label %for.body.lr.ph, label %while.body.preheader
 
 for.body.lr.ph:
-  call void @llvm.memset.p0i8.i64(i8* undef, i8 32, i64 512, i1 false)
+  call void @llvm.memset.p0i8.i64(i8* undef, i8 32, i64 512, i32 16, i1 false)
   br label %while.body.preheader
 
 while.body.preheader:
@@ -373,7 +373,7 @@ cleanup:
 declare i32 @fileno(%struct.TMP.2* nocapture)
 declare i64 @"\01_write"(i32, i8*, i64)
 declare i32 @__maskrune(i32, i64)
-declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i1)
+declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i32, i1)
 
 !llvm.ident = !{!0}
 
diff --git a/test/CodeGen/X86/remat-fold-load.ll b/test/CodeGen/X86/remat-fold-load.ll
index e640974bdd2..3478033bfbf 100644
--- a/test/CodeGen/X86/remat-fold-load.ll
+++ b/test/CodeGen/X86/remat-fold-load.ll
@@ -16,7 +16,7 @@ target triple = "i386-unknown-linux-gnu"
 %type_d = type { i64 }
 %type_e = type { %type_c, i64 }
 
-declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i32, i1) nounwind
 
 define linkonce_odr void @test() nounwind {
 entry:
@@ -41,7 +41,7 @@ if.then.i.i.i.i71:                                ; preds = %while.body12
   %tmp1 = getelementptr inbounds %type_a, %type_a* %tmp, i32 0, i32 1, i32 0, i32 1
   %buf_6.i.i.i.i70 = bitcast %type_d* %tmp1 to i8**
   %tmp2 = load i8*, i8** %buf_6.i.i.i.i70, align 4
-  call void @llvm.memcpy.p0i8.p0i8.i32(i8* undef, i8* %tmp2, i32 undef, i1 false) nounwind
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* undef, i8* %tmp2, i32 undef, i32 1, i1 false) nounwind
   unreachable
 
 if.else.i.i.i.i74:                                ; preds = %while.body12
@@ -69,7 +69,7 @@ if.then.i.i.i.i92:                                ; preds = %if.else.i.i.i.i74
   %tmp12 = getelementptr inbounds %type_e, %type_e* %tmp9, i32 0, i32 0, i32 1
   %buf_6.i.i.i.i91 = bitcast %type_d* %tmp12 to i8**
   %tmp13 = load i8*, i8** %buf_6.i.i.i.i91, align 4
-  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %call4.i.i.i.i89, i8* %tmp13, i32 %tmp10, i1 false) nounwind
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %call4.i.i.i.i89, i8* %tmp13, i32 %tmp10, i32 1, i1 false) nounwind
   br label %A
 
 if.else.i.i.i.i95:                                ; preds = %if.else.i.i.i.i74
diff --git a/test/CodeGen/X86/small-byval-memcpy.ll b/test/CodeGen/X86/small-byval-memcpy.ll
index c5c9a3d8416..3c03750199c 100644
--- a/test/CodeGen/X86/small-byval-memcpy.ll
+++ b/test/CodeGen/X86/small-byval-memcpy.ll
@@ -2,10 +2,10 @@
 ; RUN: llc < %s -mtriple=x86_64-apple-darwin -mcpu=nehalem | FileCheck %s --check-prefix=NEHALEM
 ; RUN: llc < %s -mtriple=x86_64-apple-darwin -mcpu=btver2 | FileCheck %s --check-prefix=BTVER2
 
-declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture readonly, i64, i1)
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture readonly, i64, i32, i1)
 
 define void @copy16bytes(i8* nocapture %a, i8* nocapture readonly %b) {
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %a, i8* %b, i64 16, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %a, i8* %b, i64 16, i32 1, i1 false)
   ret void
 
   ; CHECK-LABEL: copy16bytes
diff --git a/test/CodeGen/X86/stack-protector.ll b/test/CodeGen/X86/stack-protector.ll
index 43c6516a47c..398b8548747 100644
--- a/test/CodeGen/X86/stack-protector.ll
+++ b/test/CodeGen/X86/stack-protector.ll
@@ -3342,7 +3342,7 @@ entry:
   %test.coerce = alloca { i64, i8 }
   %0 = bitcast { i64, i8 }* %test.coerce to i8*
   %1 = bitcast %struct.small_char* %test to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %0, i8* %1, i64 12, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %0, i8* %1, i64 12, i32 0, i1 false)
   %2 = getelementptr { i64, i8 }, { i64, i8 }* %test.coerce, i32 0, i32 0
   %3 = load i64, i64* %2, align 1
   %4 = getelementptr { i64, i8 }, { i64, i8 }* %test.coerce, i32 0, i32 1
@@ -3376,7 +3376,7 @@ entry:
   %test.coerce = alloca { i64, i8 }
   %0 = bitcast { i64, i8 }* %test.coerce to i8*
   %1 = bitcast %struct.small_char* %test to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %0, i8* %1, i64 12, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %0, i8* %1, i64 12, i32 0, i1 false)
   %2 = getelementptr { i64, i8 }, { i64, i8 }* %test.coerce, i32 0, i32 0
   %3 = load i64, i64* %2, align 1
   %4 = getelementptr { i64, i8 }, { i64, i8 }* %test.coerce, i32 0, i32 1
@@ -3453,7 +3453,7 @@ declare void @_Z3exceptPi(i32*)
 declare i32 @__gxx_personality_v0(...)
 declare i32* @getp()
 declare i32 @dummy(...)
-declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture readonly, i64, i1)
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture readonly, i64, i32, i1)
 
 attributes #0 = { ssp }
 attributes #1 = { sspstrong }
diff --git a/test/CodeGen/X86/tailcall-mem-intrinsics.ll b/test/CodeGen/X86/tailcall-mem-intrinsics.ll
index 6a589946752..8e1e4f464ba 100644
--- a/test/CodeGen/X86/tailcall-mem-intrinsics.ll
+++ b/test/CodeGen/X86/tailcall-mem-intrinsics.ll
@@ -4,7 +4,7 @@
 ; CHECK: jmp memcpy
 define void @tail_memcpy(i8* nocapture %p, i8* nocapture readonly %q, i32 %n) #0 {
 entry:
-  tail call void @llvm.memcpy.p0i8.p0i8.i32(i8* %p, i8* %q, i32 %n, i1 false)
+  tail call void @llvm.memcpy.p0i8.p0i8.i32(i8* %p, i8* %q, i32 %n, i32 1, i1 false)
   ret void
 }
 
@@ -12,7 +12,7 @@ entry:
 ; CHECK: jmp memmove
 define void @tail_memmove(i8* nocapture %p, i8* nocapture readonly %q, i32 %n) #0 {
 entry:
-  tail call void @llvm.memmove.p0i8.p0i8.i32(i8* %p, i8* %q, i32 %n, i1 false)
+  tail call void @llvm.memmove.p0i8.p0i8.i32(i8* %p, i8* %q, i32 %n, i32 1, i1 false)
   ret void
 }
 
@@ -20,12 +20,12 @@ entry:
 ; CHECK: jmp memset
 define void @tail_memset(i8* nocapture %p, i8 %c, i32 %n) #0 {
 entry:
-  tail call void @llvm.memset.p0i8.i32(i8* %p, i8 %c, i32 %n, i1 false)
+  tail call void @llvm.memset.p0i8.i32(i8* %p, i8 %c, i32 %n, i32 1, i1 false)
   ret void
 }
 
-declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture readonly, i32, i1) #0
-declare void @llvm.memmove.p0i8.p0i8.i32(i8* nocapture, i8* nocapture readonly, i32, i1) #0
-declare void @llvm.memset.p0i8.i32(i8* nocapture, i8, i32, i1) #0
+declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture readonly, i32, i32, i1) #0
+declare void @llvm.memmove.p0i8.p0i8.i32(i8* nocapture, i8* nocapture readonly, i32, i32, i1) #0
+declare void @llvm.memset.p0i8.i32(i8* nocapture, i8, i32, i32, i1) #0
 
 attributes #0 = { nounwind }
diff --git a/test/CodeGen/X86/tlv-1.ll b/test/CodeGen/X86/tlv-1.ll
index 0dbd00c55eb..5f017d31dbb 100644
--- a/test/CodeGen/X86/tlv-1.ll
+++ b/test/CodeGen/X86/tlv-1.ll
@@ -7,7 +7,7 @@
 define void @main() nounwind ssp {
 ; CHECK-LABEL: main:
 entry:
-  call void @llvm.memset.p0i8.i64(i8* getelementptr inbounds (%struct.A, %struct.A* @c, i32 0, i32 0, i32 0), i8 0, i64 60, i1 false)
+  call void @llvm.memset.p0i8.i64(i8* getelementptr inbounds (%struct.A, %struct.A* @c, i32 0, i32 0, i32 0), i8 0, i64 60, i32 1, i1 false)
   unreachable  
   ; CHECK: movq    _c@TLVP(%rip), %rdi
   ; CHECK-NEXT: callq   *(%rdi)
@@ -31,7 +31,7 @@ entry:
   ret i32 %sub
 }
 
-declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i1) nounwind
+declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i32, i1) nounwind
 
 @a = thread_local global i32 0                    ; <i32*> [#uses=0]
 @b = thread_local global i32 0                    ; <i32*> [#uses=0]
diff --git a/test/CodeGen/X86/unaligned-load.ll b/test/CodeGen/X86/unaligned-load.ll
index c81d6e37d2e..ffbbcff2e5d 100644
--- a/test/CodeGen/X86/unaligned-load.ll
+++ b/test/CodeGen/X86/unaligned-load.ll
@@ -12,14 +12,14 @@ entry:
 
 bb:                                               ; preds = %bb, %entry
   %String2Loc9 = getelementptr inbounds [31 x i8], [31 x i8]* %String2Loc, i64 0, i64 0
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %String2Loc9, i8* getelementptr inbounds ([31 x i8], [31 x i8]* @.str3, i64 0, i64 0), i64 31, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %String2Loc9, i8* getelementptr inbounds ([31 x i8], [31 x i8]* @.str3, i64 0, i64 0), i64 31, i32 1, i1 false)
   br label %bb
 
 return:                                           ; No predecessors!
   ret void
 }
 
-declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i32, i1) nounwind
 
 ; I386: calll {{_?}}memcpy
 
diff --git a/test/CodeGen/X86/unwindraise.ll b/test/CodeGen/X86/unwindraise.ll
index 2f6410fadd4..fb8319b63c2 100644
--- a/test/CodeGen/X86/unwindraise.ll
+++ b/test/CodeGen/X86/unwindraise.ll
@@ -34,7 +34,7 @@ entry:
   call fastcc void @uw_init_context_1(%struct._Unwind_Context* %this_context, i8* %0, i8* %1)
   %2 = bitcast %struct._Unwind_Context* %cur_context to i8*
   %3 = bitcast %struct._Unwind_Context* %this_context to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %2, i8* %3, i64 240, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %2, i8* %3, i64 240, i32 8, i1 false)
   %personality = getelementptr inbounds %struct._Unwind_FrameState, %struct._Unwind_FrameState* %fs, i64 0, i32 6
   %retaddr_column.i = getelementptr inbounds %struct._Unwind_FrameState, %struct._Unwind_FrameState* %fs, i64 0, i32 9
   %flags.i.i.i.i = getelementptr inbounds %struct._Unwind_Context, %struct._Unwind_Context* %cur_context, i64 0, i32 5
@@ -121,7 +121,7 @@ while.end:                                        ; preds = %if.then4
   %16 = ptrtoint i8* %15 to i64
   %private_2 = getelementptr inbounds %struct._Unwind_Exception, %struct._Unwind_Exception* %exc, i64 0, i32 3
   store i64 %16, i64* %private_2, align 8
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %2, i8* %3, i64 240, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %2, i8* %3, i64 240, i32 8, i1 false)
   %17 = bitcast %struct._Unwind_FrameState* %fs.i to i8*
   call void @llvm.lifetime.start(i64 -1, i8* %17)
   %personality.i = getelementptr inbounds %struct._Unwind_FrameState, %struct._Unwind_FrameState* %fs.i, i64 0, i32 6
@@ -234,7 +234,7 @@ declare i8* @llvm.eh.dwarf.cfa(i32) nounwind
 
 declare i8* @llvm.returnaddress(i32) nounwind readnone
 
-declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i32, i1) nounwind
 
 declare fastcc i64 @uw_install_context_1(%struct._Unwind_Context*, %struct._Unwind_Context*) uwtable
 
diff --git a/test/CodeGen/X86/variable-sized-darwin-bzero.ll b/test/CodeGen/X86/variable-sized-darwin-bzero.ll
index 43032fec15c..1e86d75bf09 100644
--- a/test/CodeGen/X86/variable-sized-darwin-bzero.ll
+++ b/test/CodeGen/X86/variable-sized-darwin-bzero.ll
@@ -1,8 +1,8 @@
 ; RUN: llc < %s -march=x86 -mtriple=i686-apple-darwin10 | grep __bzero
 
 define void @foo(i8* %p, i64 %n) {
-  call void @llvm.memset.p0i8.i64(i8* %p, i8 0, i64 %n, i1 false)
+  call void @llvm.memset.p0i8.i64(i8* %p, i8 0, i64 %n, i32 4, i1 false)
   ret void
 }
 
-declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i1) nounwind
+declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i32, i1) nounwind
diff --git a/test/CodeGen/X86/x86-64-static-relo-movl.ll b/test/CodeGen/X86/x86-64-static-relo-movl.ll
index 09a8c73599e..5da3a470503 100644
--- a/test/CodeGen/X86/x86-64-static-relo-movl.ll
+++ b/test/CodeGen/X86/x86-64-static-relo-movl.ll
@@ -10,7 +10,7 @@
 define void @setup() {
   %pending = alloca %struct.MatchInfo, align 8
   %t = bitcast %struct.MatchInfo* %pending to i8*
-  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %t, i8* bitcast (%struct.MatchInfo* @NO_MATCH to i8*), i64 512, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %t, i8* bitcast (%struct.MatchInfo* @NO_MATCH to i8*), i64 512, i32 8, i1 false)
   %u = getelementptr inbounds %struct.MatchInfo, %struct.MatchInfo* %pending, i32 0, i32 2
   %v = load i64, i64* %u, align 8
   br label %done
@@ -21,4 +21,4 @@ done:
 }
 
 ; Function Attrs: nounwind
-declare void @llvm.memcpy.p0i8.p0i8.i64(i8*, i8*, i64, i1)
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8*, i8*, i64, i32, i1)
diff --git a/test/CodeGen/XCore/memcpy.ll b/test/CodeGen/XCore/memcpy.ll
index 934061df173..fe424c50cb2 100644
--- a/test/CodeGen/XCore/memcpy.ll
+++ b/test/CodeGen/XCore/memcpy.ll
@@ -6,7 +6,7 @@ define void @f1(i8* %dst, i8* %src, i32 %n) nounwind {
 ; CHECK: bl __memcpy_4
 entry:
   %0 = shl i32 %n, 2
-  call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 4 %dst, i8* align 4 %src, i32 %0, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dst, i8* %src, i32 %0, i32 4, i1 false)
   ret void
 }
 
@@ -15,7 +15,7 @@ define void @f2(i8* %dst, i8* %src, i32 %n) nounwind {
 ; CHECK-LABEL: f2:
 ; CHECK: bl memcpy
 entry:
-  call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 4 %dst, i8* align 4 %src, i32 %n, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dst, i8* %src, i32 %n, i32 4, i1 false)
   ret void
 }
 
@@ -25,8 +25,8 @@ define void @f3(i8* %dst, i8* %src, i32 %n) nounwind {
 ; CHECK: bl memcpy
 entry:
   %0 = shl i32 %n, 2
-  call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 1 %dst, i8* align 1 %src, i32 %0, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dst, i8* %src, i32 %0, i32 2, i1 false)
   ret void
 }
 
-declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i32, i1) nounwind
author	Pete Cooper <peter_cooper@apple.com>	2015-11-19 05:56:52 +0000
committer	Pete Cooper <peter_cooper@apple.com>	2015-11-19 05:56:52 +0000
commit	6d024c616a2a4959f8dfe5c64d27f89b394cf042 (patch)
tree	3010d09bb33fb9c353cea4ee911a8023854f9a1d /test/CodeGen
parent	0ab4d52b0ad3b1d16238aed4c3f8d50fa8bf8b40 (diff)