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

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

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

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

There are a few places in the code for which the code needs to be
checked by an expert as to whether using only src/dest alignment is
safe.  For those places, they currently take the minimum of src/dest
alignments which matches the current behaviour.

For example, code which used to read:
  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dest, i8* %src, i32 500, i32 8, i1 false)
will now read:
  call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 8 %dest, i8* align 8 %src, i32 500, i1 false)

For out of tree owners, I was able to strip alignment from calls using sed by replacing:
  (call.*llvm\.memset.*)i32\ [0-9]*\,\ i1 false\)
with:
  $1i1 false)

and similarly for memmove and memcpy.

I then added back in alignment to test cases which needed it.

A similar commit will be made to clang which actually has many differences in alignment as now
IRBuilder can generate different source/dest alignments on calls.

In IRBuilder itself, a new argument was added.  Instead of calling:
  CreateMemCpy(Dst, Src, getInt64(Size), DstAlign, /* isVolatile */ false)
you now call
  CreateMemCpy(Dst, Src, getInt64(Size), DstAlign, SrcAlign, /* isVolatile */ false)

There is a temporary class (IntegerAlignment) which takes the source alignment and rejects
implicit conversion from bool.  This is to prevent isVolatile here from passing its default
parameter to the source alignment.

Note, changes in future can now be made to codegen.  I didn't change anything here, but this
change should enable better memcpy code sequences.

Reviewed by Hal Finkel.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@253511 91177308-0d34-0410-b5e6-96231b3b80d8

18 files changed, 87 insertions, 87 deletions

diff --git a/test/CodeGen/PowerPC/2011-12-05-NoSpillDupCR.ll b/test/CodeGen/PowerPC/2011-12-05-NoSpillDupCR.ll
index e5920911ee2..2a6ddd5db10 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, i32, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, 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 93476827949..3271466cbcc 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, i32, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, 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 874e5712bb9..7f4946fcc12 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, i32, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i1) nounwind
 
 declare i32 @obj_compare(...)
 
diff --git a/test/CodeGen/PowerPC/emptystruct.ll b/test/CodeGen/PowerPC/emptystruct.ll
index 66cada14f9d..49a9c13c7cf 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, i32 1, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %1, i8* %2, i64 0, 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, i32, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, 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 09b7e41b189..26d9a09b8f7 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, i32 4, i1 false)
+  tail call void @llvm.memcpy.p0i8.p0i8.i32(i8* %0, i8* %1, i32 52, i1 false)
   ret void
 }
 
-declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i32, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i1) nounwind
diff --git a/test/CodeGen/PowerPC/fsl-e5500.ll b/test/CodeGen/PowerPC/fsl-e5500.ll
index d47d8c8ed4f..664a449eb1b 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, i32 4, i1 false)
+  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* %0, i8* %1, i64 100, i1 false)
   ret void
 }
 
-declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i32, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i1) nounwind
diff --git a/test/CodeGen/PowerPC/glob-comp-aa-crash.ll b/test/CodeGen/PowerPC/glob-comp-aa-crash.ll
index 88530a3f303..0d5bd846239 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, i32 8, i1 false)
+  call void @llvm.memset.p0i8.i64(i8* %0, i8 0, i64 16, 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, i32, i1) #3
+declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, 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 7475e12c447..93d1b6bc1d8 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, i32 4, i1 false) #1
+  call void @llvm.memset.p0i8.i64(i8* bitcast ([1 x [9 x i32]]* @g_62 to i8*), i8 -1, i64 36, 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, i32, i1) #1
+declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, 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 29baef55ce1..27aed429d1b 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* %0, i8* %1, i64 32, i32 8, i1 false)
+  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 8 %0, i8* align 8 %1, i64 32, 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, i32, i1) #0
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture readonly, i64, 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* %0, i8* %1, i64 128, i32 8, i1 false)
+  tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* align 8 %0, i8* align 8 %1, i64 128, 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* %0, i8 0, i64 128, i32 8, i1 false)
+  tail call void @llvm.memset.p0i8.i64(i8* align 8 %0, i8 0, i64 128, 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* %0, i8 0, i64 128, i32 32, i1 false)
+  tail call void @llvm.memset.p0i8.i64(i8* align 32 %0, i8 0, i64 128, 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, i32, i1) #0
+declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i1) #0
 
 attributes #0 = { nounwind }
 
diff --git a/test/CodeGen/PowerPC/memset-nc-le.ll b/test/CodeGen/PowerPC/memset-nc-le.ll
index af8e9c3fb4f..fa31af50a07 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, i32 1, i1 false)
+  call void @llvm.memset.p0i8.i64(i8* null, i8 %1, i64 32, 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, i32, i1) #1
+declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, 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 414a987a56f..041d26f0dbf 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* null, i8 %1, i64 64, i32 32, i1 false)
+  call void @llvm.memset.p0i8.i64(i8* align 32 null, i8 %1, i64 64, 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, i32, i1) #1
+declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, 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* null, i8 %1, i64 32, i32 1, i1 false)
+  call void @llvm.memset.p0i8.i64(i8* align 1 null, i8 %1, i64 32, 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 8c0a2fb1dc0..9c0af17009b 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, i32 4, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %1, i8* %2, i64 8, i1 false)
   ret void
 }
 
 ; CHECK: func_fab
 
 ; Function Attrs: nounwind
-declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture readonly, i64, i32, i1) #1
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture readonly, i64, 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 a613c3310a5..c3e63d5342f 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, 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)
+  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)
   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, i32 16, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %62, i8* bitcast (%struct.S1998* @s1998 to i8*), i64 5168, 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, i32 16, 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, 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, i32 16, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %64, i8* bitcast (%struct.S1998* @s1998 to i8*), i64 5168, 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, i32 16, 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, 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, i32 16, 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, 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, i32 16, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %67, i8* bitcast (%struct.S1998* @s1998 to i8*), i64 5168, 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, i32 16, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %68, i8* bitcast (%struct.S1998* @s1998 to i8*), i64 5168, 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, i32 16, 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, 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, i32 16, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %70, i8* bitcast (%struct.S1998* @s1998 to i8*), i64 5168, 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, i32, 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.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture readonly, i64, 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 a1c2070a6f4..da7025c121e 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, i32 16, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %2, i8* %3, i64 11104, 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, i32 16, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %4, i8* %5, i64 11104, 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, i32 32, i1 false)
+  call void @llvm.memset.p0i8.i64(i8* %6, i8 0, i64 11104, i1 false)
   %7 = bitcast %struct.S2760* %b1 to i8*
-  call void @llvm.memset.p0i8.i64(i8* %7, i8 0, i64 11104, i32 32, i1 false)
+  call void @llvm.memset.p0i8.i64(i8* %7, i8 0, i64 11104, i1 false)
   %8 = bitcast %struct.S2760* %b2 to i8*
-  call void @llvm.memset.p0i8.i64(i8* %8, i8 0, i64 11104, i32 32, i1 false)
+  call void @llvm.memset.p0i8.i64(i8* %8, i8 0, i64 11104, 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, i32 32, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %13, i8* %14, i64 11104, i1 false)
   ret void
 }
 
-declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture readonly, i64, i32, 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.memset.p0i8.i64(i8* nocapture, i8, i64, i1)
 
diff --git a/test/CodeGen/PowerPC/structsinmem.ll b/test/CodeGen/PowerPC/structsinmem.ll
index bfada4c6371..c240c6e2656 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, i32 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, 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, i32 2, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %1, i8* bitcast (%struct.s2* @caller1.p2 to i8*), i64 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, i32 2, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %2, i8* bitcast ({ i16, i8, i8 }* @caller1.p3 to i8*), i64 4, 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, i32 4, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %3, i8* bitcast (%struct.s4* @caller1.p4 to i8*), i64 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, i32 4, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %4, i8* bitcast ({ i32, i8, [3 x i8] }* @caller1.p5 to i8*), i64 8, 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, i32 4, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %5, i8* bitcast ({ i32, i16, [2 x i8] }* @caller1.p6 to i8*), i64 8, 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, i32 4, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %6, i8* bitcast ({ i32, i16, i8, i8 }* @caller1.p7 to i8*), i64 8, 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, i32, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, 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, i32 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, 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, i32 1, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %1, i8* bitcast ({ i16 }* @caller2.p2 to i8*), i64 2, 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, i32 1, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %2, i8* bitcast (%struct.t3* @caller2.p3 to i8*), i64 3, 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, i32 1, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %3, i8* bitcast ({ i32 }* @caller2.p4 to i8*), i64 4, 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, i32 1, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %4, i8* bitcast (%struct.t5* @caller2.p5 to i8*), i64 5, 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, i32 1, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %5, i8* bitcast (%struct.t6* @caller2.p6 to i8*), i64 6, 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, i32 1, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %6, i8* bitcast (%struct.t7* @caller2.p7 to i8*), i64 7, 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 0fb9895a622..006d06cebed 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, i32 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, 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, i32 2, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %1, i8* bitcast (%struct.s2* @caller1.p2 to i8*), i64 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, i32 2, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %2, i8* bitcast ({ i16, i8, i8 }* @caller1.p3 to i8*), i64 4, 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, i32 4, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %3, i8* bitcast (%struct.s4* @caller1.p4 to i8*), i64 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, i32 4, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %4, i8* bitcast ({ i32, i8, [3 x i8] }* @caller1.p5 to i8*), i64 8, 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, i32 4, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %5, i8* bitcast ({ i32, i16, [2 x i8] }* @caller1.p6 to i8*), i64 8, 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, i32 4, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %6, i8* bitcast ({ i32, i16, i8, i8 }* @caller1.p7 to i8*), i64 8, 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, i32, i1) nounwind
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, 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, i32 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, 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, i32 1, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %1, i8* bitcast ({ i16 }* @caller2.p2 to i8*), i64 2, 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, i32 1, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %2, i8* bitcast (%struct.t3* @caller2.p3 to i8*), i64 3, 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, i32 1, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %3, i8* bitcast ({ i32 }* @caller2.p4 to i8*), i64 4, 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, i32 1, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %4, i8* bitcast (%struct.t5* @caller2.p5 to i8*), i64 5, 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, i32 1, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %5, i8* bitcast (%struct.t6* @caller2.p6 to i8*), i64 6, 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, i32 1, i1 false)
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %6, i8* bitcast (%struct.t7* @caller2.p7 to i8*), i64 7, 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 bb2748432d7..7b680264993 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* %0, i8 0, i64 40, i32 4, i1 false) nounwind
+  call void @llvm.memset.p0i8.i64(i8* align 4 %0, i8 0, i64 40, 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, i32, i1) nounwind
+declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, i1) nounwind
diff --git a/test/CodeGen/PowerPC/toc-load-sched-bug.ll b/test/CodeGen/PowerPC/toc-load-sched-bug.ll
index 5ac4e363502..8717183e22c 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, i32 8, i1 false) #3
+  call void @llvm.memset.p0i8.i64(i8* %11, i8 0, i64 16, 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, i32 8, i1 false) #3
+  call void @llvm.memset.p0i8.i64(i8* %13, i8 0, i64 24, 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, i32 8, i1 false) #3
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %15, i8* %11, i64 16, 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, i32 4, i1 false) #3
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %16, i8* %17, i64 12, 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, i32 8, i1 false) #3
+  call void @llvm.memset.p0i8.i64(i8* %19, i8 0, i64 16, 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, i32, i1) #3
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture readonly, i64, 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, i32, i1) #3
+declare void @llvm.memset.p0i8.i64(i8* nocapture, i8, i64, 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" }