Merge master branch

git-svn-id: https://llvm.org/svn/llvm-project/llvm/branches/R600/@166411 91177308-0d34-0410-b5e6-96231b3b80d8
author: tstellar <tstellar@91177308-0d34-0410-b5e6-96231b3b80d8> 2012-10-22 14:55:30 +0000
committer: tstellar <tstellar@91177308-0d34-0410-b5e6-96231b3b80d8> 2012-10-22 14:55:30 +0000
commit: 81d3998812b5977a220d62076cc446822747af21 (patch)
tree: b073cb1be9c65dbdd6cd312c322efd3c04abb0af /test
parent: 63a079bf76d139b91e170557cf16b17bdbecd58d (diff)
51 files changed, 2050 insertions, 76 deletions
diff --git a/test/BugPoint/crash-narrowfunctiontest.ll b/test/BugPoint/crash-narrowfunctiontest.ll
index d080d9dd4b0..9df823ab973 100644
--- a/test/BugPoint/crash-narrowfunctiontest.ll
+++ b/test/BugPoint/crash-narrowfunctiontest.ll
@@ -2,6 +2,7 @@
 ;
 ; RUN: bugpoint -load %llvmshlibdir/BugpointPasses%shlibext %s -output-prefix %t -bugpoint-crashcalls -silence-passes > /dev/null
 ; REQUIRES: loadable_module
+; XFAIL: lto
 
 define i32 @foo() { ret i32 1 }
 
diff --git a/test/BugPoint/metadata.ll b/test/BugPoint/metadata.ll
index 0eda5667ba4..98c79ee03a6 100644
--- a/test/BugPoint/metadata.ll
+++ b/test/BugPoint/metadata.ll
@@ -1,6 +1,7 @@
 ; RUN: bugpoint -load %llvmshlibdir/BugpointPasses%shlibext %s -output-prefix %t -bugpoint-crashcalls -silence-passes > /dev/null
 ; RUN: llvm-dis %t-reduced-simplified.bc -o - | FileCheck %s
 ; REQUIRES: loadable_module
+; XFAIL: lto
 
 ; Bugpoint should keep the call's metadata attached to the call.
 
diff --git a/test/BugPoint/remove_arguments_test.ll b/test/BugPoint/remove_arguments_test.ll
index 29a03b83107..13aa9c5a653 100644
--- a/test/BugPoint/remove_arguments_test.ll
+++ b/test/BugPoint/remove_arguments_test.ll
@@ -1,6 +1,7 @@
 ; RUN: bugpoint -load %llvmshlibdir/BugpointPasses%shlibext %s -output-prefix %t -bugpoint-crashcalls -silence-passes
 ; RUN: llvm-dis %t-reduced-simplified.bc -o - | FileCheck %s
 ; REQUIRES: loadable_module
+; XFAIL: lto
 
 ; Test to make sure that arguments are removed from the function if they are 
 ; unnecessary. And clean up any types that that frees up too.
diff --git a/test/CodeGen/ARM/2012-10-18-PR14099-ByvalFrameAddress.ll b/test/CodeGen/ARM/2012-10-18-PR14099-ByvalFrameAddress.ll
new file mode 100644
index 00000000000..fcc6a7f7e96
--- /dev/null
+++ b/test/CodeGen/ARM/2012-10-18-PR14099-ByvalFrameAddress.ll
@@ -0,0 +1,29 @@
+; RUN: llc < %s -mtriple=armv7-linux-gnueabi | FileCheck %s
+
+%struct.s = type { [4 x i32] }
+@v = constant %struct.s zeroinitializer; 
+
+declare void @f(%struct.s* %p);
+
+; CHECK: t:
+define void @t(i32 %a, %struct.s* byval %s) nounwind {
+entry:
+
+; Here we need to only check proper start address of restored %s argument.
+; CHECK:      sub     sp, sp, #16
+; CHECK:      push    {r11, lr}
+; CHECK:      add     r0, sp, #12
+; CHECK:      stm     r0, {r1, r2, r3}
+; CHECK:      add     r0, sp, #12
+; CHECK-NEXT: bl f
+  call void @f(%struct.s* %s)
+  ret void
+}
+
+; CHECK: caller:
+define void @caller() {
+
+; CHECK:      ldm     r0, {r1, r2, r3}
+  call void @t(i32 0, %struct.s* @v);
+  ret void
+}
diff --git a/test/CodeGen/ARM/trap.ll b/test/CodeGen/ARM/trap.ll
index 38842a9646f..21865f8e4ae 100644
--- a/test/CodeGen/ARM/trap.ll
+++ b/test/CodeGen/ARM/trap.ll
@@ -14,4 +14,16 @@ entry:
   unreachable
 }
 
+define void @t2() nounwind {
+entry:
+; INSTR: t2:
+; INSTR: trap
+
+; FUNC: t2:
+; FUNC: bl __trap
+  call void @llvm.debugtrap()
+  unreachable
+}
+
 declare void @llvm.trap() nounwind
+declare void @llvm.debugtrap() nounwind
diff --git a/test/CodeGen/MSP430/fp.ll b/test/CodeGen/MSP430/fp.ll
new file mode 100644
index 00000000000..c3273eff05c
--- /dev/null
+++ b/test/CodeGen/MSP430/fp.ll
@@ -0,0 +1,17 @@
+; RUN: llc -O0 -disable-fp-elim < %s | FileCheck %s
+
+target datalayout = "e-p:16:16:16-i8:8:8-i16:16:16-i32:16:32-n8:16"
+target triple = "msp430---elf"
+
+define void @fp() nounwind {
+entry:
+; CHECK: fp:
+; CHECK: push.w r4
+; CHECK: mov.w r1, r4
+; CHECK: sub.w #2, r1
+  %i = alloca i16, align 2
+; CHECK: mov.w #0, -2(r4)
+  store i16 0, i16* %i, align 2
+; CHECK: pop.w r4
+  ret void
+}
diff --git a/test/CodeGen/Mips/brconeq.ll b/test/CodeGen/Mips/brconeq.ll
new file mode 100644
index 00000000000..613391557ef
--- /dev/null
+++ b/test/CodeGen/Mips/brconeq.ll
@@ -0,0 +1,38 @@
+; RUN: llc  -march=mipsel -mcpu=mips16 -relocation-model=pic -O3 < %s | FileCheck %s -check-prefix=16
+
+@i = global i32 5, align 4
+@j = global i32 10, align 4
+@result = global i32 0, align 4
+
+define void @test() nounwind {
+entry:
+  %0 = load i32* @i, align 4
+  %1 = load i32* @j, align 4
+  %cmp = icmp eq i32 %0, %1
+; 16:	cmp	${{[0-9]+}}, ${{[0-9]+}}
+; 16:	bteqz	$[[LABEL:[0-9A-Ba-b_]+]]
+; 16: $[[LABEL]]:
+  br i1 %cmp, label %if.end, label %if.then
+
+if.then:                                          ; preds = %entry
+  store i32 1, i32* @result, align 4
+  br label %if.end
+
+if.end:                                           ; preds = %entry, %if.then
+  ret void
+}
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
diff --git a/test/CodeGen/Mips/brconeqk.ll b/test/CodeGen/Mips/brconeqk.ll
new file mode 100644
index 00000000000..2c0e72dabd2
--- /dev/null
+++ b/test/CodeGen/Mips/brconeqk.ll
@@ -0,0 +1,22 @@
+; RUN: llc  -march=mipsel -mcpu=mips16 -relocation-model=pic -O3 < %s | FileCheck %s -check-prefix=16
+
+@i = global i32 5, align 4
+@result = global i32 0, align 4
+
+define void @test() nounwind {
+entry:
+  %0 = load i32* @i, align 4
+  %cmp = icmp eq i32 %0, 10
+  br i1 %cmp, label %if.end, label %if.then
+; 16:	cmpi	${{[0-9]+}}, {{[0-9]+}}
+; 16:	bteqz	$[[LABEL:[0-9A-Ba-b_]+]]
+; 16: $[[LABEL]]:
+if.then:                                          ; preds = %entry
+  store i32 1, i32* @result, align 4
+  br label %if.end
+
+if.end:                                           ; preds = %entry, %if.then
+  ret void
+}
+
+
diff --git a/test/CodeGen/Mips/brconeqz.ll b/test/CodeGen/Mips/brconeqz.ll
new file mode 100644
index 00000000000..5586e7b976d
--- /dev/null
+++ b/test/CodeGen/Mips/brconeqz.ll
@@ -0,0 +1,20 @@
+; RUN: llc  -march=mipsel -mcpu=mips16 -relocation-model=pic -O3 < %s | FileCheck %s -check-prefix=16
+
+@i = global i32 5, align 4
+@result = global i32 0, align 4
+
+define void @test() nounwind {
+entry:
+  %0 = load i32* @i, align 4
+  %cmp = icmp eq i32 %0, 0
+  br i1 %cmp, label %if.end, label %if.then
+; 16:	beqz	${{[0-9]+}}, $[[LABEL:[0-9A-Ba-b_]+]]
+; 16: $[[LABEL]]:
+if.then:                                          ; preds = %entry
+  store i32 1, i32* @result, align 4
+  br label %if.end
+
+if.end:                                           ; preds = %entry, %if.then
+  ret void
+}
+
diff --git a/test/CodeGen/Mips/brconge.ll b/test/CodeGen/Mips/brconge.ll
new file mode 100644
index 00000000000..02f0a633b31
--- /dev/null
+++ b/test/CodeGen/Mips/brconge.ll
@@ -0,0 +1,37 @@
+; RUN: llc  -march=mipsel -mcpu=mips16 -relocation-model=pic -O3 < %s | FileCheck %s -check-prefix=16
+
+@i = global i32 5, align 4
+@j = global i32 10, align 4
+@k = global i32 5, align 4
+@result1 = global i32 0, align 4
+@result2 = global i32 1, align 4
+
+define void @test() nounwind {
+entry:
+  %0 = load i32* @i, align 4
+  %1 = load i32* @j, align 4
+  %cmp = icmp slt i32 %0, %1
+  br i1 %cmp, label %if.then, label %if.end
+
+; 16:	slt	${{[0-9]+}}, ${{[0-9]+}}
+; 16:	bteqz	$[[LABEL:[0-9A-Ba-b_]+]]
+; 16: $[[LABEL]]:
+
+if.then:                                          ; preds = %entry
+  store i32 1, i32* @result1, align 4
+  br label %if.end
+
+if.end:                                           ; preds = %if.then, %entry
+  %2 = load i32* @k, align 4
+  %cmp1 = icmp slt i32 %0, %2
+  br i1 %cmp1, label %if.then2, label %if.end3
+
+if.then2:                                         ; preds = %if.end
+  store i32 1, i32* @result1, align 4
+  br label %if.end3
+
+if.end3:                                          ; preds = %if.then2, %if.end
+  ret void
+}
+
+
diff --git a/test/CodeGen/Mips/brcongt.ll b/test/CodeGen/Mips/brcongt.ll
new file mode 100644
index 00000000000..767b51b21b9
--- /dev/null
+++ b/test/CodeGen/Mips/brcongt.ll
@@ -0,0 +1,25 @@
+; RUN: llc  -march=mipsel -mcpu=mips16 -relocation-model=pic -O3 < %s | FileCheck %s -check-prefix=16
+
+@i = global i32 5, align 4
+@j = global i32 10, align 4
+@k = global i32 5, align 4
+@result = global i32 0, align 4
+
+define void @test() nounwind {
+entry:
+  %0 = load i32* @i, align 4
+  %1 = load i32* @j, align 4
+  %cmp = icmp sgt i32 %0, %1
+  br i1 %cmp, label %if.end, label %if.then
+; 16:	slt	${{[0-9]+}}, ${{[0-9]+}}
+; 16:	btnez	$[[LABEL:[0-9A-Ba-b_]+]]
+; 16: $[[LABEL]]:
+if.then:                                          ; preds = %entry
+  store i32 1, i32* @result, align 4
+  br label %if.end
+
+if.end:                                           ; preds = %entry, %if.then
+  ret void
+}
+
+
diff --git a/test/CodeGen/Mips/brconle.ll b/test/CodeGen/Mips/brconle.ll
new file mode 100644
index 00000000000..854b2481c6e
--- /dev/null
+++ b/test/CodeGen/Mips/brconle.ll
@@ -0,0 +1,37 @@
+; RUN: llc  -march=mipsel -mcpu=mips16 -relocation-model=pic -O3 < %s | FileCheck %s -check-prefix=16
+
+@i = global i32 -5, align 4
+@j = global i32 10, align 4
+@k = global i32 -5, align 4
+@result1 = global i32 0, align 4
+@result2 = global i32 1, align 4
+
+define void @test() nounwind {
+entry:
+  %0 = load i32* @j, align 4
+  %1 = load i32* @i, align 4
+  %cmp = icmp sgt i32 %0, %1
+  br i1 %cmp, label %if.then, label %if.end
+
+; 16:	slt	${{[0-9]+}}, ${{[0-9]+}}
+; 16:	bteqz	$[[LABEL:[0-9A-Ba-b_]+]]
+; 16: $[[LABEL]]:
+
+if.then:                                          ; preds = %entry
+  store i32 1, i32* @result1, align 4
+  br label %if.end
+
+if.end:                                           ; preds = %if.then, %entry
+  %2 = load i32* @k, align 4
+  %cmp1 = icmp sgt i32 %1, %2
+  br i1 %cmp1, label %if.then2, label %if.end3
+
+if.then2:                                         ; preds = %if.end
+  store i32 0, i32* @result1, align 4
+  br label %if.end3
+
+if.end3:                                          ; preds = %if.then2, %if.end
+  ret void
+}
+
+
diff --git a/test/CodeGen/Mips/brconlt.ll b/test/CodeGen/Mips/brconlt.ll
new file mode 100644
index 00000000000..931a3e8c7ba
--- /dev/null
+++ b/test/CodeGen/Mips/brconlt.ll
@@ -0,0 +1,27 @@
+; RUN: llc  -march=mipsel -mcpu=mips16 -relocation-model=pic -O3 < %s | FileCheck %s -check-prefix=16
+
+@i = global i32 5, align 4
+@j = global i32 10, align 4
+@k = global i32 5, align 4
+@result = global i32 0, align 4
+
+define void @test() nounwind {
+entry:
+  %0 = load i32* @j, align 4
+  %1 = load i32* @i, align 4
+  %cmp = icmp slt i32 %0, %1
+  br i1 %cmp, label %if.end, label %if.then
+
+; 16:	slt	${{[0-9]+}}, ${{[0-9]+}}
+; 16:	btnez	$[[LABEL:[0-9A-Ba-b_]+]]
+; 16: $[[LABEL]]:
+
+if.then:                                          ; preds = %entry
+  store i32 1, i32* @result, align 4
+  br label %if.end
+
+if.end:                                           ; preds = %entry, %if.then
+  ret void
+}
+
+
diff --git a/test/CodeGen/Mips/brconne.ll b/test/CodeGen/Mips/brconne.ll
new file mode 100644
index 00000000000..5d5bde3fcf9
--- /dev/null
+++ b/test/CodeGen/Mips/brconne.ll
@@ -0,0 +1,26 @@
+; RUN: llc  -march=mipsel -mcpu=mips16 -relocation-model=pic -O3 < %s | FileCheck %s -check-prefix=16
+
+@i = global i32 5, align 4
+@j = global i32 5, align 4
+@result = global i32 0, align 4
+
+define void @test() nounwind {
+entry:
+  %0 = load i32* @j, align 4
+  %1 = load i32* @i, align 4
+  %cmp = icmp eq i32 %0, %1
+  br i1 %cmp, label %if.then, label %if.end
+; 16:	cmp	${{[0-9]+}}, ${{[0-9]+}}
+; 16:	btnez	$[[LABEL:[0-9A-Ba-b_]+]]
+; 16:   lw ${{[0-9]+}}, %got(result)(${{[0-9]+}})
+; 16: $[[LABEL]]:
+
+if.then:                                          ; preds = %entry
+  store i32 1, i32* @result, align 4
+  br label %if.end
+
+if.end:                                           ; preds = %if.then, %entry
+  ret void
+}
+
+
diff --git a/test/CodeGen/Mips/brconnek.ll b/test/CodeGen/Mips/brconnek.ll
new file mode 100644
index 00000000000..6208d7c5a04
--- /dev/null
+++ b/test/CodeGen/Mips/brconnek.ll
@@ -0,0 +1,25 @@
+; RUN: llc  -march=mipsel -mcpu=mips16 -relocation-model=pic -O3 < %s | FileCheck %s -check-prefix=16
+
+@j = global i32 5, align 4
+@result = global i32 0, align 4
+
+define void @test() nounwind {
+entry:
+  %0 = load i32* @j, align 4
+  %cmp = icmp eq i32 %0, 5
+  br i1 %cmp, label %if.then, label %if.end
+
+; 16:	cmpi	${{[0-9]+}}, {{[0-9]+}}
+; 16:	btnez	$[[LABEL:[0-9A-Ba-b_]+]]
+; 16:   lw ${{[0-9]+}}, %got(result)(${{[0-9]+}})
+; 16: $[[LABEL]]:
+
+if.then:                                          ; preds = %entry
+  store i32 1, i32* @result, align 4
+  br label %if.end
+
+if.end:                                           ; preds = %if.then, %entry
+  ret void
+}
+
+
diff --git a/test/CodeGen/Mips/brconnez.ll b/test/CodeGen/Mips/brconnez.ll
new file mode 100644
index 00000000000..47db7901b51
--- /dev/null
+++ b/test/CodeGen/Mips/brconnez.ll
@@ -0,0 +1,24 @@
+; RUN: llc  -march=mipsel -mcpu=mips16 -relocation-model=pic -O3 < %s | FileCheck %s -check-prefix=16
+
+@j = global i32 0, align 4
+@result = global i32 0, align 4
+
+define void @test() nounwind {
+entry:
+  %0 = load i32* @j, align 4
+  %cmp = icmp eq i32 %0, 0
+  br i1 %cmp, label %if.then, label %if.end
+
+; 16:	bnez	${{[0-9]+}}, $[[LABEL:[0-9A-Ba-b_]+]]
+; 16:   lw ${{[0-9]+}}, %got(result)(${{[0-9]+}})
+; 16: $[[LABEL]]:
+
+if.then:                                          ; preds = %entry
+  store i32 1, i32* @result, align 4
+  br label %if.end
+
+if.end:                                           ; preds = %if.then, %entry
+  ret void
+}
+
+
diff --git a/test/CodeGen/Mips/mips64-sret.ll b/test/CodeGen/Mips/mips64-sret.ll
new file mode 100644
index 00000000000..498c5fe1747
--- /dev/null
+++ b/test/CodeGen/Mips/mips64-sret.ll
@@ -0,0 +1,14 @@
+; RUN: llc -march=mips64el -mcpu=mips64r2 -mattr=n64 -O0 < %s
+
+%struct.S = type { [8 x i32] }
+
+@g = common global %struct.S zeroinitializer, align 4
+
+define void @f(%struct.S* noalias sret %agg.result) nounwind {
+entry:
+  %0 = bitcast %struct.S* %agg.result to i8*
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %0, i8* bitcast (%struct.S* @g to i8*), i64 32, i32 4, i1 false)
+  ret void
+}
+
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i32, i1) nounwind
diff --git a/test/CodeGen/Mips/tailcall.ll b/test/CodeGen/Mips/tailcall.ll
new file mode 100644
index 00000000000..4989636a20d
--- /dev/null
+++ b/test/CodeGen/Mips/tailcall.ll
@@ -0,0 +1,100 @@
+; RUN: llc -march=mipsel -relocation-model=pic -enable-mips-tail-calls < %s | \
+; RUN: FileCheck %s -check-prefix=PIC32
+; RUN: llc -march=mipsel -relocation-model=static \
+; RUN: -enable-mips-tail-calls < %s | FileCheck %s -check-prefix=STATIC32
+; RUN: llc -march=mips64el -mcpu=mips64r2 -mattr=+n64 -enable-mips-tail-calls \
+; RUN: < %s | FileCheck %s -check-prefix=N64
+
+@g0 = common global i32 0, align 4
+@g1 = common global i32 0, align 4
+@g2 = common global i32 0, align 4
+@g3 = common global i32 0, align 4
+@g4 = common global i32 0, align 4
+@g5 = common global i32 0, align 4
+@g6 = common global i32 0, align 4
+@g7 = common global i32 0, align 4
+@g8 = common global i32 0, align 4
+@g9 = common global i32 0, align 4
+
+define i32 @caller1(i32 %a0) nounwind {
+entry:
+; PIC32-NOT: jalr
+; STATIC32-NOT: jal
+; N64-NOT: jalr
+
+  %call = tail call i32 @callee1(i32 1, i32 1, i32 1, i32 %a0) nounwind
+  ret i32 %call
+}
+
+declare i32 @callee1(i32, i32, i32, i32)
+
+define i32 @caller2(i32 %a0, i32 %a1, i32 %a2, i32 %a3) nounwind {
+entry:
+; PIC32: jalr
+; STATIC32: jal
+; N64-NOT: jalr
+
+  %call = tail call i32 @callee2(i32 1, i32 %a0, i32 %a1, i32 %a2, i32 %a3) nounwind
+  ret i32 %call
+}
+
+declare i32 @callee2(i32, i32, i32, i32, i32)
+
+define i32 @caller3(i32 %a0, i32 %a1, i32 %a2, i32 %a3, i32 %a4) nounwind {
+entry:
+; PIC32: jalr
+; STATIC32: jal
+; N64-NOT: jalr
+
+  %call = tail call i32 @callee3(i32 1, i32 1, i32 1, i32 %a0, i32 %a1, i32 %a2, i32 %a3, i32 %a4) nounwind
+  ret i32 %call
+}
+
+declare i32 @callee3(i32, i32, i32, i32, i32, i32, i32, i32)
+
+define i32 @caller4(i32 %a0, i32 %a1, i32 %a2, i32 %a3, i32 %a4, i32 %a5, i32 %a6, i32 %a7) nounwind {
+entry:
+; PIC32: jalr
+; STATIC32: jal
+; N64: jalr
+
+  %call = tail call i32 @callee4(i32 1, i32 %a0, i32 %a1, i32 %a2, i32 %a3, i32 %a4, i32 %a5, i32 %a6, i32 %a7) nounwind
+  ret i32 %call
+}
+
+declare i32 @callee4(i32, i32, i32, i32, i32, i32, i32, i32, i32)
+
+define i32 @caller5() nounwind readonly {
+entry:
+; PIC32-NOT: jalr
+; STATIC32-NOT: jal
+; N64-NOT: jalr
+
+  %0 = load i32* @g0, align 4
+  %1 = load i32* @g1, align 4
+  %2 = load i32* @g2, align 4
+  %3 = load i32* @g3, align 4
+  %4 = load i32* @g4, align 4
+  %5 = load i32* @g5, align 4
+  %6 = load i32* @g6, align 4
+  %7 = load i32* @g7, align 4
+  %8 = load i32* @g8, align 4
+  %9 = load i32* @g9, align 4
+  %call = tail call fastcc i32 @callee5(i32 %0, i32 %1, i32 %2, i32 %3, i32 %4, i32 %5, i32 %6, i32 %7, i32 %8, i32 %9)
+  ret i32 %call
+}
+
+define internal fastcc i32 @callee5(i32 %a0, i32 %a1, i32 %a2, i32 %a3, i32 %a4, i32 %a5, i32 %a6, i32 %a7, i32 %a8, i32 %a9) nounwind readnone noinline {
+entry:
+  %add = add nsw i32 %a1, %a0
+  %add1 = add nsw i32 %add, %a2
+  %add2 = add nsw i32 %add1, %a3
+  %add3 = add nsw i32 %add2, %a4
+  %add4 = add nsw i32 %add3, %a5
+  %add5 = add nsw i32 %add4, %a6
+  %add6 = add nsw i32 %add5, %a7
+  %add7 = add nsw i32 %add6, %a8
+  %add8 = add nsw i32 %add7, %a9
+  ret i32 %add8
+}
+
diff --git a/test/CodeGen/PowerPC/i64_fp_round.ll b/test/CodeGen/PowerPC/i64_fp_round.ll
new file mode 100644
index 00000000000..5a0c072c9c5
--- /dev/null
+++ b/test/CodeGen/PowerPC/i64_fp_round.ll
@@ -0,0 +1,27 @@
+; RUN: llc -mcpu=pwr7 < %s | FileCheck %s
+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-v128:128:128-n32:64"
+target triple = "powerpc64-unknown-linux-gnu"
+
+define float @test(i64 %x) nounwind readnone {
+entry:
+  %conv = sitofp i64 %x to float
+  ret float %conv
+}
+
+; Verify that we get the code sequence needed to avoid double-rounding.
+; Note that only parts of the sequence are checked for here, to allow
+; for minor code generation differences.
+
+; CHECK: sradi [[REGISTER:[0-9]+]], 3, 53
+; CHECK: addi [[REGISTER:[0-9]+]], [[REGISTER]], 1
+; CHECK: cmpldi 0, [[REGISTER]], 1
+; CHECK: isel [[REGISTER:[0-9]+]], {{[0-9]+}}, 3, 1
+; CHECK: std [[REGISTER]], -{{[0-9]+}}(1)
+
+
+; Also check that with -enable-unsafe-fp-math we do not get that extra
+; code sequence.  Simply verify that there is no "isel" present.
+
+; RUN: llc -mcpu=pwr7 -enable-unsafe-fp-math < %s | FileCheck %s -check-prefix=UNSAFE
+; CHECK-UNSAFE-NOT: isel
+
diff --git a/test/CodeGen/X86/2012-10-18-crash-dagco.ll b/test/CodeGen/X86/2012-10-18-crash-dagco.ll
new file mode 100644
index 00000000000..5b98624a37b
--- /dev/null
+++ b/test/CodeGen/X86/2012-10-18-crash-dagco.ll
@@ -0,0 +1,61 @@
+; RUN: llc -march=x86-64 -mcpu=corei7 -disable-cgp-select2branch < %s
+
+; We should not crash on this test.
+
+target datalayout = "e-p:32:32:32-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:32:64-f32:32:32-f64:32:64-v64:64:64-v128:128:128-a0:0:64-f80:128:128-n8:16:32"
+target triple = "i386-apple-darwin9.0.0"
+
+@global = external constant [411 x i8], align 1
+
+define void @snork() nounwind {
+bb:
+  br i1 undef, label %bb26, label %bb27
+
+bb26:                                             ; preds = %bb48, %bb26, %bb
+  switch i32 undef, label %bb26 [
+    i32 142771596, label %bb28
+  ]
+
+bb27:                                             ; preds = %bb48, %bb
+  switch i32 undef, label %bb49 [
+    i32 142771596, label %bb28
+  ]
+
+bb28:                                             ; preds = %bb27, %bb26
+  %tmp = load i32* null
+  %tmp29 = trunc i32 %tmp to i8
+  store i8* undef, i8** undef
+  %tmp30 = load i32* null
+  %tmp31 = icmp eq i32 %tmp30, 0
+  %tmp32 = getelementptr inbounds [411 x i8]* @global, i32 0, i32 undef
+  %tmp33 = load i8* %tmp32, align 1
+  %tmp34 = getelementptr inbounds [411 x i8]* @global, i32 0, i32 0
+  %tmp35 = load i8* %tmp34, align 1
+  %tmp36 = select i1 %tmp31, i8 %tmp35, i8 %tmp33
+  %tmp37 = select i1 undef, i8 %tmp29, i8 %tmp36
+  %tmp38 = zext i8 %tmp37 to i32
+  %tmp39 = select i1 undef, i32 0, i32 %tmp38
+  %tmp40 = getelementptr inbounds i32* null, i32 %tmp39
+  %tmp41 = load i32* %tmp40, align 4
+  %tmp42 = load i32* undef, align 4
+  %tmp43 = load i32* undef
+  %tmp44 = xor i32 %tmp42, %tmp43
+  %tmp45 = lshr i32 %tmp44, 8
+  %tmp46 = lshr i32 %tmp44, 7
+  call void @spam()
+  unreachable
+
+bb47:                                             ; No predecessors!
+  ret void
+
+bb48:                                             ; No predecessors!
+  br i1 undef, label %bb27, label %bb26
+
+bb49:                                             ; preds = %bb49, %bb27
+  br label %bb49
+
+bb50:                                             ; preds = %bb50
+  br label %bb50
+}
+
+declare void @spam() noreturn nounwind
diff --git a/test/CodeGen/X86/buildvec-insertvec.ll b/test/CodeGen/X86/buildvec-insertvec.ll
new file mode 100644
index 00000000000..3fb69a48b3c
--- /dev/null
+++ b/test/CodeGen/X86/buildvec-insertvec.ll
@@ -0,0 +1,15 @@
+; RUN: llc < %s -mcpu=corei7 -mtriple=x86_64-unknown-linux-gnu | FileCheck %s
+
+define void @foo(<3 x float> %in, <4 x i8>* nocapture %out) nounwind {
+  %t0 = fptoui <3 x float> %in to <3 x i8>
+  %t1 = shufflevector <3 x i8> %t0, <3 x i8> undef, <4 x i32> <i32 0, i32 1, i32 2, i32 undef>
+  %t2 = insertelement <4 x i8> %t1, i8 -1, i32 3
+  store <4 x i8> %t2, <4 x i8>* %out, align 4
+  ret void
+; CHECK: foo
+; CHECK: cvttps2dq
+; CHECK-NOT: pextrd
+; CHECK: pinsrd
+; CHECK-NEXT: pshufb
+; CHECK: ret
+}
diff --git a/test/CodeGen/X86/extract-concat.ll b/test/CodeGen/X86/extract-concat.ll
new file mode 100644
index 00000000000..704309eb650
--- /dev/null
+++ b/test/CodeGen/X86/extract-concat.ll
@@ -0,0 +1,17 @@
+; RUN: llc < %s -mcpu=corei7 -mtriple=x86_64-unknown-linux-gnu | FileCheck %s
+
+define void @foo(<4 x float> %in, <4 x i8>* %out) {
+  %t0 = fptosi <4 x float> %in to <4 x i32>
+  %t1 = trunc <4 x i32> %t0 to <4 x i16>
+  %t2 = shufflevector <4 x i16> %t1, <4 x i16> undef, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
+  %t3 = trunc <8 x i16> %t2 to <8 x i8>
+  %t4 = shufflevector <8 x i8> %t3, <8 x i8> undef, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
+  %t5 = insertelement <4 x i8> %t4, i8 -1, i32 3
+  store <4 x i8> %t5, <4 x i8>* %out
+  ret void
+; CHECK: foo
+; CHECK: cvttps2dq
+; CHECK-NOT: pextrd
+; CHECK: pshufb
+; CHECK: ret
+}
diff --git a/test/CodeGen/X86/pr14090.ll b/test/CodeGen/X86/pr14090.ll
new file mode 100644
index 00000000000..d76b912fd8e
--- /dev/null
+++ b/test/CodeGen/X86/pr14090.ll
@@ -0,0 +1,76 @@
+; RUN: llc < %s -march=x86-64 -print-before=stack-coloring -print-after=stack-coloring >%t 2>&1 && FileCheck <%t %s
+
+define void @foo(i64* %retval.i, i32 %call, i32* %.ph.i80, i32 %fourteen, i32* %out.lo, i32* %out.hi) nounwind align 2 {
+entry:
+  %_Tmp.i39 = alloca i64, align 8
+  %retval.i33 = alloca i64, align 8
+  %_Tmp.i = alloca i64, align 8
+  %retval.i.i = alloca i64, align 8
+  %_First.i = alloca i64, align 8
+
+  %0 = load i64* %retval.i, align 8
+
+  %1 = load i64* %retval.i, align 8
+
+  %_Tmp.i39.0.cast73 = bitcast i64* %_Tmp.i39 to i8*
+  call void @llvm.lifetime.start(i64 8, i8* %_Tmp.i39.0.cast73)
+  store i64 %1, i64* %_Tmp.i39, align 8
+  %cmp.i.i.i40 = icmp slt i32 %call, 0
+  %2 = lshr i64 %1, 32
+  %3 = trunc i64 %2 to i32
+  %sub.i.i.i44 = sub i32 0, %call
+  %cmp2.i.i.i45 = icmp ult i32 %3, %sub.i.i.i44
+  %or.cond.i.i.i46 = and i1 %cmp.i.i.i40, %cmp2.i.i.i45
+  %add.i.i.i47 = add i32 %3, %call
+  %sub5.i.i.i48 = lshr i32 %add.i.i.i47, 5
+  %trunc.i50 = trunc i64 %1 to i32
+  %inttoptr.i51 = inttoptr i32 %trunc.i50 to i32*
+  %add61617.i.i.i52 = or i32 %sub5.i.i.i48, -134217728
+  %add61617.i.sub5.i.i.i53 = select i1 %or.cond.i.i.i46, i32 %add61617.i.i.i52, i32 %sub5.i.i.i48
+  %storemerge2.i.i54 = getelementptr inbounds i32* %inttoptr.i51, i32 %add61617.i.sub5.i.i.i53
+  %_Tmp.i39.0.cast74 = bitcast i64* %_Tmp.i39 to i32**
+  store i32* %storemerge2.i.i54, i32** %_Tmp.i39.0.cast74, align 8
+  %storemerge.i.i55 = and i32 %add.i.i.i47, 31
+  %_Tmp.i39.4.raw_idx = getelementptr inbounds i8* %_Tmp.i39.0.cast73, i32 4
+  %_Tmp.i39.4.cast = bitcast i8* %_Tmp.i39.4.raw_idx to i32*
+  store i32 %storemerge.i.i55, i32* %_Tmp.i39.4.cast, align 4
+  %srcval.i56 = load i64* %_Tmp.i39, align 8
+  call void @llvm.lifetime.end(i64 8, i8* %_Tmp.i39.0.cast73)
+
+; CHECK: Before Merge disjoint stack slots
+; CHECK: [[PREFIX15:MOV64mr.*<fi#]]{{[0-9]}}[[SUFFIX15:.*;]] mem:ST8[%fifteen]
+; CHECK: [[PREFIX87:MOV32mr.*;]] mem:ST4[%sunkaddr87]
+
+; CHECK: After Merge disjoint stack slots
+; CHECK: [[PREFIX15]]{{[0-9]}}[[SUFFIX15]] mem:ST8[%_Tmp.i39]
+; CHECK: [[PREFIX87]] mem:ST4[<unknown>]
+
+  %fifteen = bitcast i64* %retval.i.i to i32**
+  %sixteen = bitcast i64* %retval.i.i to i8*
+  call void @llvm.lifetime.start(i64 8, i8* %sixteen)
+  store i32* %.ph.i80, i32** %fifteen, align 8, !tbaa !0
+  %sunkaddr = ptrtoint i64* %retval.i.i to i32
+  %sunkaddr86 = add i32 %sunkaddr, 4
+  %sunkaddr87 = inttoptr i32 %sunkaddr86 to i32*
+  store i32 %fourteen, i32* %sunkaddr87, align 4, !tbaa !3
+  %seventeen = load i64* %retval.i.i, align 8
+  call void @llvm.lifetime.end(i64 8, i8* %sixteen)
+  %eighteen = lshr i64 %seventeen, 32
+  %nineteen = trunc i64 %eighteen to i32
+  %shl.i.i.i = shl i32 1, %nineteen
+
+  store i32 %shl.i.i.i, i32* %out.lo, align 8
+  store i32 %nineteen, i32* %out.hi, align 8
+
+  ret void
+}
+
+declare void @llvm.lifetime.start(i64, i8* nocapture) nounwind
+
+declare void @llvm.lifetime.end(i64, i8* nocapture) nounwind
+
+!0 = metadata !{metadata !"int", metadata !1}
+!1 = metadata !{metadata !"omnipotent char", metadata !2}
+!2 = metadata !{metadata !"Simple C/C++ TBAA"}
+!3 = metadata !{metadata !"any pointer", metadata !1}
+!4 = metadata !{metadata !"vtable pointer", metadata !2}
diff --git a/test/CodeGen/X86/pr14098.ll b/test/CodeGen/X86/pr14098.ll
new file mode 100644
index 00000000000..6ce2449ab6a
--- /dev/null
+++ b/test/CodeGen/X86/pr14098.ll
@@ -0,0 +1,23 @@
+; RUN: llc -mtriple i386-unknown-linux-gnu -relocation-model=pic -verify-machineinstrs < %s
+; We used to crash on this.
+
+declare void @foo()
+declare void @foo3(i1 %x)
+define void @bar(i1 %a1, i16 %a2) nounwind align 2 {
+bb0:
+  %a3 = trunc i16 %a2 to i8
+  %a4 = lshr i16 %a2, 8
+  %a5 = trunc i16 %a4 to i8
+  br i1 %a1, label %bb1, label %bb2
+bb1:
+  br label %bb2
+bb2:
+  %a6 = phi i8 [ 3, %bb0 ], [ %a5, %bb1 ]
+  %a7 = phi i8 [ 9, %bb0 ], [ %a3, %bb1 ]
+  %a8 = icmp eq i8 %a6, 1
+  call void @foo()
+  %a9 = icmp eq i8 %a7, 0
+  call void @foo3(i1 %a9)
+  call void @foo3(i1 %a8)
+  ret void
+}
diff --git a/test/CodeGen/X86/sjlj.ll b/test/CodeGen/X86/sjlj.ll
index d594e982994..681db009438 100644
--- a/test/CodeGen/X86/sjlj.ll
+++ b/test/CodeGen/X86/sjlj.ll
@@ -1,5 +1,7 @@
 ; RUN: llc < %s -mtriple=i386-pc-linux -mcpu=corei7 -relocation-model=static | FileCheck --check-prefix=X86 %s
-; RUN: llc < %s -mtriple=x86_64-pc-linux -mcpu=corei7 | FileCheck --check-prefix=X64 %s
+; RUN: llc < %s -mtriple=i386-pc-linux -mcpu=corei7 -relocation-model=pic | FileCheck --check-prefix=PIC86 %s
+; RUN: llc < %s -mtriple=x86_64-pc-linux -mcpu=corei7 -relocation-model=static | FileCheck --check-prefix=X64 %s
+; RUN: llc < %s -mtriple=x86_64-pc-linux -mcpu=corei7 -relocation-model=pic | FileCheck --check-prefix=PIC64 %s
 
 @buf = internal global [5 x i8*] zeroinitializer
 
@@ -20,14 +22,26 @@ define i32 @sj0() nounwind {
   ret i32 %r
 ; X86: sj0
 ; x86: movl %ebp, buf
-; x86: movl ${{.*LBB.*}}, buf+4
 ; X86: movl %esp, buf+8
+; x86: movl ${{.*LBB.*}}, buf+4
 ; X86: ret
+; PIC86: sj0
+; PIC86: movl %ebp, buf@GOTOFF(%[[GOT:.*]])
+; PIC86: movl %esp, buf@GOTOFF+8(%[[GOT]])
+; PIC86: leal {{.*LBB.*}}@GOTOFF(%[[GOT]]), %[[LREG:.*]]
+; PIC86: movl %[[LREG]], buf@GOTOFF+4
+; PIC86: ret
 ; X64: sj0
 ; x64: movq %rbp, buf(%rip)
 ; x64: movq ${{.*LBB.*}}, buf+8(%rip)
 ; X64: movq %rsp, buf+16(%rip)
 ; X64: ret
+; PIC64: sj0
+; PIC64: movq %rbp, buf(%rip)
+; PIC64: movq %rsp, buf+16(%rip)
+; PIC64: leaq {{.*LBB.*}}(%rip), %[[LREG:.*]]
+; PIC64: movq %[[LREG]], buf+8(%rip)
+; PIC64: ret
 }
 
 define void @lj0() nounwind {
diff --git a/test/Instrumentation/AddressSanitizer/basic.ll b/test/Instrumentation/AddressSanitizer/basic.ll
index d1900018706..655f69c16fd 100644
--- a/test/Instrumentation/AddressSanitizer/basic.ll
+++ b/test/Instrumentation/AddressSanitizer/basic.ll
@@ -69,3 +69,23 @@ entry:
   store i32 42, i32* %a
   ret void
 }
+
+; Check that asan leaves just one alloca.
+
+declare void @alloca_test_use([10 x i8]*)
+define void @alloca_test() address_safety {
+entry:
+  %x = alloca [10 x i8], align 1
+  %y = alloca [10 x i8], align 1
+  %z = alloca [10 x i8], align 1
+  call void @alloca_test_use([10 x i8]* %x)
+  call void @alloca_test_use([10 x i8]* %y)
+  call void @alloca_test_use([10 x i8]* %z)
+  ret void
+}
+
+; CHECK: define void @alloca_test()
+; CHECK: = alloca
+; CHECK-NOT: = alloca
+; CHECK: ret void
+
diff --git a/test/Makefile b/test/Makefile
index 3c6b5b68a4f..810fdded465 100644
--- a/test/Makefile
+++ b/test/Makefile
@@ -117,6 +117,16 @@ else
 ENABLE_ASSERTIONS=1
 endif
 
+# Derive whether or not LTO is enabled by checking the extra options.
+LTO_IS_ENABLED := 0
+ifneq ($(findstring -flto,$(CompileCommonOpts)),)
+LTO_IS_ENABLED := 1
+else
+ifneq ($(findstring -O4,$(CompileCommonOpts)),)
+LTO_IS_ENABLED := 1
+endif
+endif
+
 lit.site.cfg: FORCE
 	@echo "Making LLVM 'lit.site.cfg' file..."
 	@$(ECHOPATH) s=@TARGET_TRIPLE@=$(TARGET_TRIPLE)=g > lit.tmp
@@ -129,6 +139,7 @@ lit.site.cfg: FORCE
 	@$(ECHOPATH) s=@OCAMLOPT@=$(OCAMLOPT) -cc $(subst *,'\\\"',*$(subst =,"\\=",$(CXX_FOR_OCAMLOPT))*) -I $(LibDir)/ocaml=g >> lit.tmp
 	@$(ECHOPATH) s=@ENABLE_SHARED@=$(ENABLE_SHARED)=g >> lit.tmp
 	@$(ECHOPATH) s=@ENABLE_ASSERTIONS@=$(ENABLE_ASSERTIONS)=g >> lit.tmp
+	@$(ECHOPATH) s=@LTO_IS_ENABLED@=$(LTO_IS_ENABLED)=g >> lit.tmp
 	@$(ECHOPATH) s=@TARGETS_TO_BUILD@=$(TARGETS_TO_BUILD)=g >> lit.tmp
 	@$(ECHOPATH) s=@LLVM_BINDINGS@=$(BINDINGS_TO_BUILD)=g >> lit.tmp
 	@$(ECHOPATH) s=@HOST_OS@=$(HOST_OS)=g >> lit.tmp
diff --git a/test/Transforms/BBVectorize/simple-ldstr-ptrs.ll b/test/Transforms/BBVectorize/simple-ldstr-ptrs.ll
index f992d415477..6294543cd81 100644
--- a/test/Transforms/BBVectorize/simple-ldstr-ptrs.ll
+++ b/test/Transforms/BBVectorize/simple-ldstr-ptrs.ll
@@ -79,3 +79,53 @@ entry:
 ; CHECK-AO-NOT: <2 x
 }
 
+; Simple 3-pair chain with loads and stores (using ptrs and gep)
+; using pointer vectors.
+define void @test3(<2 x i64*>* %a, <2 x i64*>* %b, <2 x i64*>* %c) nounwind uwtable readonly {
+entry:
+  %i0 = load <2 x i64*>* %a, align 8
+  %i1 = load <2 x i64*>* %b, align 8
+  %arrayidx3 = getelementptr inbounds <2 x i64*>* %a, i64 1
+  %i3 = load <2 x i64*>* %arrayidx3, align 8
+  %arrayidx4 = getelementptr inbounds <2 x i64*>* %b, i64 1
+  %i4 = load <2 x i64*>* %arrayidx4, align 8
+  %j1 = extractelement <2 x i64*> %i1, i32 0
+  %j4 = extractelement <2 x i64*> %i4, i32 0
+  %o1 = load i64* %j1, align 8
+  %o4 = load i64* %j4, align 8
+  %j0 = extractelement <2 x i64*> %i0, i32 0
+  %j3 = extractelement <2 x i64*> %i3, i32 0
+  %ptr0 = getelementptr inbounds i64* %j0, i64 %o1
+  %ptr3 = getelementptr inbounds i64* %j3, i64 %o4
+  %qtr0 = insertelement <2 x i64*> undef, i64* %ptr0, i32 0
+  %rtr0 = insertelement <2 x i64*> %qtr0, i64* %ptr0, i32 1
+  %qtr3 = insertelement <2 x i64*> undef, i64* %ptr3, i32 0
+  %rtr3 = insertelement <2 x i64*> %qtr3, i64* %ptr3, i32 1
+  store <2 x i64*> %rtr0, <2 x i64*>* %c, align 8
+  %arrayidx5 = getelementptr inbounds <2 x i64*>* %c, i64 1
+  store <2 x i64*> %rtr3, <2 x i64*>* %arrayidx5, align 8
+  ret void
+; CHECK: @test3
+; CHECK: %i0.v.i0 = bitcast <2 x i64*>* %a to <4 x i64*>*
+; CHECK: %i1 = load <2 x i64*>* %b, align 8
+; CHECK: %i0 = load <4 x i64*>* %i0.v.i0, align 8
+; CHECK: %arrayidx4 = getelementptr inbounds <2 x i64*>* %b, i64 1
+; CHECK: %i4 = load <2 x i64*>* %arrayidx4, align 8
+; CHECK: %j1 = extractelement <2 x i64*> %i1, i32 0
+; CHECK: %j4 = extractelement <2 x i64*> %i4, i32 0
+; CHECK: %o1 = load i64* %j1, align 8
+; CHECK: %o4 = load i64* %j4, align 8
+; CHECK: %ptr0.v.i1.1 = insertelement <2 x i64> undef, i64 %o1, i32 0
+; CHECK: %ptr0.v.i1.2 = insertelement <2 x i64> %ptr0.v.i1.1, i64 %o4, i32 1
+; CHECK: %ptr0.v.i0 = shufflevector <4 x i64*> %i0, <4 x i64*> undef, <2 x i32> <i32 0, i32 2>
+; CHECK: %ptr0 = getelementptr inbounds <2 x i64*> %ptr0.v.i0, <2 x i64> %ptr0.v.i1.2
+; CHECK: %rtr0 = shufflevector <2 x i64*> %ptr0, <2 x i64*> undef, <2 x i32> zeroinitializer
+; CHECK: %rtr3 = shufflevector <2 x i64*> %ptr0, <2 x i64*> undef, <2 x i32> <i32 1, i32 1>
+; CHECK: %0 = bitcast <2 x i64*>* %c to <4 x i64*>*
+; CHECK: %1 = shufflevector <2 x i64*> %rtr0, <2 x i64*> %rtr3, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
+; CHECK: store <4 x i64*> %1, <4 x i64*>* %0, align 8
+; CHECK: ret void
+; CHECK-AO: @test3
+; CHECK-AO-NOT: <4 x
+}
+
diff --git a/test/Transforms/IndVarSimplify/2012-10-19-congruent-constant.ll b/test/Transforms/IndVarSimplify/2012-10-19-congruent-constant.ll
new file mode 100644
index 00000000000..5c478669d29
--- /dev/null
+++ b/test/Transforms/IndVarSimplify/2012-10-19-congruent-constant.ll
@@ -0,0 +1,27 @@
+; RUN: opt -S -indvars < %s | FileCheck %s
+
+; PR12627
+define void @test1(i32 %x) nounwind uwtable ssp {
+entry:
+  br label %for.body
+
+for.body:                                         ; preds = %for.body, %entry
+  %phi1 = phi i1 [ false, %entry ], [ %cmpa, %for.body ]
+  %phi2 = phi i1 [ false, %entry ], [ %cmpb, %for.body ]
+  %i.07 = phi i32 [ 0, %entry ], [ %inc, %for.body ]
+  tail call void @aux(i1 %phi1, i1 %phi2) nounwind
+  %cmpa = icmp sgt i32 %i.07, 200
+  %cmpb = icmp sgt i32 %i.07, 100
+  %inc = add nsw i32 %i.07, 1
+  %exitcond = icmp eq i32 %inc, 100
+  br i1 %exitcond, label %for.end, label %for.body
+
+for.end:                                          ; preds = %for.body
+  ret void
+
+; CHECK: @test1
+; CHECK-NOT: phi i1
+; CHECK: call void @aux(i1 false, i1 false)
+}
+
+declare void @aux(i1, i1)
diff --git a/test/Transforms/IndVarSimplify/crash.ll b/test/Transforms/IndVarSimplify/crash.ll
index 62af42b9d68..1b702a3b1a3 100644
--- a/test/Transforms/IndVarSimplify/crash.ll
+++ b/test/Transforms/IndVarSimplify/crash.ll
@@ -113,3 +113,21 @@ bb9:
   ret void
 }
 
+; PR12536
+define void @fn1() noreturn nounwind {
+entry:
+  br label %for.cond
+
+for.cond:                                         ; preds = %for.end, %entry
+  %b.0 = phi i32 [ undef, %entry ], [ %conv, %for.end ]
+  br label %for.cond1
+
+for.cond1:                                        ; preds = %for.cond1, %for.cond
+  %c.0 = phi i32 [ %b.0, %for.cond1 ], [ 0, %for.cond ]
+  br i1 undef, label %for.cond1, label %for.end
+
+for.end:                                          ; preds = %for.cond1
+  %cmp2 = icmp slt i32 %c.0, 1
+  %conv = zext i1 %cmp2 to i32
+  br label %for.cond
+}
diff --git a/test/Transforms/IndVarSimplify/no-iv-rewrite.ll b/test/Transforms/IndVarSimplify/no-iv-rewrite.ll
index bfdd000e38e..507f695e67c 100644
--- a/test/Transforms/IndVarSimplify/no-iv-rewrite.ll
+++ b/test/Transforms/IndVarSimplify/no-iv-rewrite.ll
@@ -199,7 +199,6 @@ entry:
 ; back to the loop iv.
 ;
 ; CHECK: loop:
-; CHECK: phi i32
 ; CHECK-NOT: phi
 ; CHECK: exit:
 loop:
diff --git a/test/Transforms/InstCombine/obfuscated_splat.ll b/test/Transforms/InstCombine/obfuscated_splat.ll
new file mode 100644
index 00000000000..c25dade168a
--- /dev/null
+++ b/test/Transforms/InstCombine/obfuscated_splat.ll
@@ -0,0 +1,11 @@
+; RUN: opt -instcombine -S %s | FileCheck %s
+
+define void @test(<4 x float> *%in_ptr, <4 x float> *%out_ptr) {
+  %A = load <4 x float>* %in_ptr, align 16
+  %B = shufflevector <4 x float> %A, <4 x float> undef, <4 x i32> <i32 0, i32 0, i32 undef, i32 undef>
+  %C = shufflevector <4 x float> %B, <4 x float> %A, <4 x i32> <i32 0, i32 1, i32 4, i32 undef>
+  %D = shufflevector <4 x float> %C, <4 x float> %A, <4 x i32> <i32 0, i32 1, i32 2, i32 4>
+; CHECK:  %D = shufflevector <4 x float> %A, <4 x float> undef, <4 x i32> zeroinitializer
+  store <4 x float> %D, <4 x float> *%out_ptr
+  ret void
+}
diff --git a/test/Transforms/InstCombine/select.ll b/test/Transforms/InstCombine/select.ll
index 4baae2618dd..cc3aacdce3c 100644
--- a/test/Transforms/InstCombine/select.ll
+++ b/test/Transforms/InstCombine/select.ll
@@ -829,3 +829,37 @@ define i1 @test63(i1 %A, i1 %B) {
 ; CHECK: %C = or i1 %B, %not
 ; CHECK: ret i1 %C
 }
+
+; PR14131
+define void @test64(i32 %p, i16 %b) noreturn nounwind {
+entry:
+  %p.addr.0.insert.mask = and i32 %p, -65536
+  %conv2 = and i32 %p, 65535
+  br i1 undef, label %lor.rhs, label %lor.end
+
+lor.rhs:
+  %p.addr.0.extract.trunc = trunc i32 %p.addr.0.insert.mask to i16
+  %phitmp = zext i16 %p.addr.0.extract.trunc to i32
+  br label %lor.end
+
+lor.end:
+  %t.1 = phi i32 [ 0, %entry ], [ %phitmp, %lor.rhs ]
+  %conv6 = zext i16 %b to i32
+  %div = udiv i32 %conv6, %t.1
+  %tobool8 = icmp eq i32 %div, 0
+  %cmp = icmp eq i32 %t.1, 0
+  %cmp12 = icmp ult i32 %conv2, 2
+  %cmp.sink = select i1 %tobool8, i1 %cmp12, i1 %cmp
+  br i1 %cmp.sink, label %cond.end17, label %cond.false16
+
+cond.false16:
+  br label %cond.end17
+
+cond.end17:
+  br label %while.body
+
+while.body:
+  br label %while.body
+; CHECK: @test64
+; CHECK-NOT: select
+}
diff --git a/test/Transforms/InstCombine/strcpy-1.ll b/test/Transforms/InstCombine/strcpy-1.ll
new file mode 100644
index 00000000000..b6cf048b2a8
--- /dev/null
+++ b/test/Transforms/InstCombine/strcpy-1.ll
@@ -0,0 +1,45 @@
+; Test that the strcpy library call simplifier works correctly.
+; rdar://6839935
+; RUN: opt < %s -instcombine -S | FileCheck %s
+;
+; This transformation requires the pointer size, as it assumes that size_t is
+; the size of a pointer.
+target datalayout = "e-p:32:32:32-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:32:64-f32:32:32-f64:32:64-v64:64:64-v128:128:128-a0:0:64-f80:128:128-n8:16:32"
+
+@hello = constant [6 x i8] c"hello\00"
+@a = common global [32 x i8] zeroinitializer, align 1
+@b = common global [32 x i8] zeroinitializer, align 1
+
+declare i8* @strcpy(i8*, i8*)
+
+define void @test_simplify1() {
+; CHECK: @test_simplify1
+
+  %dst = getelementptr [32 x i8]* @a, i32 0, i32 0
+  %src = getelementptr [6 x i8]* @hello, i32 0, i32 0
+
+  call i8* @strcpy(i8* %dst, i8* %src)
+; CHECK: @llvm.memcpy.p0i8.p0i8.i32
+  ret void
+}
+
+define i8* @test_simplify2() {
+; CHECK: @test_simplify2
+
+  %dst = getelementptr [32 x i8]* @a, i32 0, i32 0
+
+  %ret = call i8* @strcpy(i8* %dst, i8* %dst)
+; CHECK: ret i8* getelementptr inbounds ([32 x i8]* @a, i32 0, i32 0)
+  ret i8* %ret
+}
+
+define i8* @test_no_simplify1() {
+; CHECK: @test_no_simplify1
+
+  %dst = getelementptr [32 x i8]* @a, i32 0, i32 0
+  %src = getelementptr [32 x i8]* @b, i32 0, i32 0
+
+  %ret = call i8* @strcpy(i8* %dst, i8* %src)
+; CHECK: call i8* @strcpy
+  ret i8* %ret
+}
diff --git a/test/Transforms/InstCombine/strcpy-2.ll b/test/Transforms/InstCombine/strcpy-2.ll
new file mode 100644
index 00000000000..779e9fdd959
--- /dev/null
+++ b/test/Transforms/InstCombine/strcpy-2.ll
@@ -0,0 +1,22 @@
+; Test that the strcpy library call simplifier works correctly.
+; RUN: opt < %s -instcombine -S | FileCheck %s
+;
+; This transformation requires the pointer size, as it assumes that size_t is
+; the size of a pointer.
+target datalayout = "e-p:32:32:32-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:32:64-f32:32:32-f64:32:64-v64:64:64-v128:128:128-a0:0:64-f80:128:128-n8:16:32"
+
+@hello = constant [6 x i8] c"hello\00"
+@a = common global [32 x i8] zeroinitializer, align 1
+
+declare i16* @strcpy(i8*, i8*)
+
+define void @test_no_simplify1() {
+; CHECK: @test_no_simplify1
+
+  %dst = getelementptr [32 x i8]* @a, i32 0, i32 0
+  %src = getelementptr [6 x i8]* @hello, i32 0, i32 0
+
+  call i16* @strcpy(i8* %dst, i8* %src)
+; CHECK: call i16* @strcpy
+  ret void
+}
diff --git a/test/Transforms/InstCombine/strcpy_chk-1.ll b/test/Transforms/InstCombine/strcpy_chk-1.ll
index c03e8a348b8..3e48f4fd305 100644
--- a/test/Transforms/InstCombine/strcpy_chk-1.ll
+++ b/test/Transforms/InstCombine/strcpy_chk-1.ll
@@ -7,16 +7,16 @@ target datalayout = "e-p:32:32:32-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:32:64-f3
 
 @a = common global [60 x i8] zeroinitializer, align 1
 @b = common global [60 x i8] zeroinitializer, align 1
-@.str = private constant [8 x i8] c"abcdefg\00"
+@.str = private constant [12 x i8] c"abcdefghijk\00"
 
 ; Check cases where slen >= strlen (src).
 
 define void @test_simplify1() {
 ; CHECK: @test_simplify1
   %dst = getelementptr inbounds [60 x i8]* @a, i32 0, i32 0
-  %src = getelementptr inbounds [8 x i8]* @.str, i32 0, i32 0
+  %src = getelementptr inbounds [12 x i8]* @.str, i32 0, i32 0
 
-; CHECK-NEXT: call i8* @strcpy
+; CHECK-NEXT: call void @llvm.memcpy.p0i8.p0i8.i32
   call i8* @__strcpy_chk(i8* %dst, i8* %src, i32 60)
   ret void
 }
@@ -24,19 +24,19 @@ define void @test_simplify1() {
 define void @test_simplify2() {
 ; CHECK: @test_simplify2
   %dst = getelementptr inbounds [60 x i8]* @a, i32 0, i32 0
-  %src = getelementptr inbounds [8 x i8]* @.str, i32 0, i32 0
+  %src = getelementptr inbounds [12 x i8]* @.str, i32 0, i32 0
 
-; CHECK-NEXT: call i8* @strcpy
-  call i8* @__strcpy_chk(i8* %dst, i8* %src, i32 8)
+; CHECK-NEXT: call void @llvm.memcpy.p0i8.p0i8.i32
+  call i8* @__strcpy_chk(i8* %dst, i8* %src, i32 12)
   ret void
 }
 
 define void @test_simplify3() {
 ; CHECK: @test_simplify3
   %dst = getelementptr inbounds [60 x i8]* @a, i32 0, i32 0
-  %src = getelementptr inbounds [8 x i8]* @.str, i32 0, i32 0
+  %src = getelementptr inbounds [12 x i8]* @.str, i32 0, i32 0
 
-; CHECK-NEXT: call i8* @strcpy
+; CHECK-NEXT: call void @llvm.memcpy.p0i8.p0i8.i32
   call i8* @__strcpy_chk(i8* %dst, i8* %src, i32 -1)
   ret void
 }
@@ -53,36 +53,42 @@ define void @test_simplify4() {
   ret void
 }
 
-define void @test_no_simplify1() {
-; CHECK: @test_no_simplify1
+; Check case where the string length is not constant.
+
+define void @test_simplify5() {
+; CHECK: @test_simplify5
   %dst = getelementptr inbounds [60 x i8]* @a, i32 0, i32 0
-  %src = getelementptr inbounds [60 x i8]* @b, i32 0, i32 0
+  %src = getelementptr inbounds [12 x i8]* @.str, i32 0, i32 0
 
-; CHECK-NEXT: call i8* @__strcpy_chk
-  call i8* @__strcpy_chk(i8* %dst, i8* %src, i32 8)
+; CHECK: @__memcpy_chk
+  %len = call i32 @llvm.objectsize.i32(i8* %dst, i1 false)
+  call i8* @__strcpy_chk(i8* %dst, i8* %src, i32 %len)
   ret void
 }
 
-; Check case were slen < strlen (src).
+; Check case where the source and destination are the same.
 
-define void @test_no_simplify2() {
-; CHECK: @test_no_simplify2
+define i8* @test_simplify6() {
+; CHECK: @test_simplify6
   %dst = getelementptr inbounds [60 x i8]* @a, i32 0, i32 0
-  %src = getelementptr inbounds [8 x i8]* @.str, i32 0, i32 0
 
-; CHECK-NEXT: call i8* @__strcpy_chk
-  call i8* @__strcpy_chk(i8* %dst, i8* %src, i32 3)
-  ret void
+; CHECK: getelementptr inbounds ([60 x i8]* @a, i32 0, i32 0)
+  %len = call i32 @llvm.objectsize.i32(i8* %dst, i1 false)
+  %ret = call i8* @__strcpy_chk(i8* %dst, i8* %dst, i32 %len)
+  ret i8* %ret
 }
 
-define void @test_no_simplify3() {
-; CHECK: @test_no_simplify3
+; Check case where slen < strlen (src).
+
+define void @test_no_simplify1() {
+; CHECK: @test_no_simplify1
   %dst = getelementptr inbounds [60 x i8]* @a, i32 0, i32 0
-  %src = getelementptr inbounds [8 x i8]* @.str, i32 0, i32 0
+  %src = getelementptr inbounds [60 x i8]* @b, i32 0, i32 0
 
 ; CHECK-NEXT: call i8* @__strcpy_chk
-  call i8* @__strcpy_chk(i8* %dst, i8* %src, i32 0)
+  call i8* @__strcpy_chk(i8* %dst, i8* %src, i32 8)
   ret void
 }
 
 declare i8* @__strcpy_chk(i8*, i8*, i32) nounwind
+declare i32 @llvm.objectsize.i32(i8*, i1) nounwind readonly
diff --git a/test/Transforms/LoopVectorize/2012-10-20-infloop.ll b/test/Transforms/LoopVectorize/2012-10-20-infloop.ll
new file mode 100644
index 00000000000..5caaffc8dde
--- /dev/null
+++ b/test/Transforms/LoopVectorize/2012-10-20-infloop.ll
@@ -0,0 +1,27 @@
+; RUN: opt < %s  -loop-vectorize -dce
+
+; Check that we don't fall into an infinite loop.
+define void @test() nounwind {
+entry:
+ br label %for.body
+
+for.body:
+ %0 = phi i32 [ 1, %entry ], [ 0, %for.body ]
+ br label %for.body
+}
+
+
+
+define void @test2() nounwind {
+entry:
+ br label %for.body
+
+for.body:                                         ; preds = %for.body, %entry
+ %indvars.iv47 = phi i64 [ 0, %entry ], [ %indvars.iv.next48, %for.body ]
+ %0 = phi i32 [ 1, %entry ], [ 0, %for.body ]
+ %indvars.iv.next48 = add i64 %indvars.iv47, 1
+ br i1 undef, label %for.end, label %for.body
+
+for.end:                                          ; preds = %for.body
+ unreachable
+}
diff --git a/test/Transforms/LoopVectorize/gcc-examples.ll b/test/Transforms/LoopVectorize/gcc-examples.ll
new file mode 100644
index 00000000000..6fb1792b2c8
--- /dev/null
+++ b/test/Transforms/LoopVectorize/gcc-examples.ll
@@ -0,0 +1,648 @@
+; RUN: opt < %s  -loop-vectorize -dce -instcombine -licm -S | FileCheck %s
+
+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-S128"
+target triple = "x86_64-apple-macosx10.8.0"
+
+@b = common global [2048 x i32] zeroinitializer, align 16
+@c = common global [2048 x i32] zeroinitializer, align 16
+@a = common global [2048 x i32] zeroinitializer, align 16
+@G = common global [32 x [1024 x i32]] zeroinitializer, align 16
+@ub = common global [1024 x i32] zeroinitializer, align 16
+@uc = common global [1024 x i32] zeroinitializer, align 16
+@d = common global [2048 x i32] zeroinitializer, align 16
+@fa = common global [1024 x float] zeroinitializer, align 16
+@fb = common global [1024 x float] zeroinitializer, align 16
+@ic = common global [1024 x i32] zeroinitializer, align 16
+@da = common global [1024 x float] zeroinitializer, align 16
+@db = common global [1024 x float] zeroinitializer, align 16
+@dc = common global [1024 x float] zeroinitializer, align 16
+@dd = common global [1024 x float] zeroinitializer, align 16
+@dj = common global [1024 x i32] zeroinitializer, align 16
+
+;CHECK: @example1
+;CHECK: load <4 x i32>
+;CHECK: add <4 x i32>
+;CHECK: store <4 x i32>
+;CHECK: ret void
+define void @example1() nounwind uwtable ssp {
+  br label %1
+
+; <label>:1                                       ; preds = %1, %0
+  %indvars.iv = phi i64 [ 0, %0 ], [ %indvars.iv.next, %1 ]
+  %2 = getelementptr inbounds [2048 x i32]* @b, i64 0, i64 %indvars.iv
+  %3 = load i32* %2, align 4
+  %4 = getelementptr inbounds [2048 x i32]* @c, i64 0, i64 %indvars.iv
+  %5 = load i32* %4, align 4
+  %6 = add nsw i32 %5, %3
+  %7 = getelementptr inbounds [2048 x i32]* @a, i64 0, i64 %indvars.iv
+  store i32 %6, i32* %7, align 4
+  %indvars.iv.next = add i64 %indvars.iv, 1
+  %lftr.wideiv = trunc i64 %indvars.iv.next to i32
+  %exitcond = icmp eq i32 %lftr.wideiv, 256
+  br i1 %exitcond, label %8, label %1
+
+; <label>:8                                       ; preds = %1
+  ret void
+}
+
+;CHECK: @example2
+;CHECK: store <4 x i32>
+;CHECK: ret void
+define void @example2(i32 %n, i32 %x) nounwind uwtable ssp {
+  %1 = icmp sgt i32 %n, 0
+  br i1 %1, label %.lr.ph5, label %.preheader
+
+..preheader_crit_edge:                            ; preds = %.lr.ph5
+  %phitmp = sext i32 %n to i64
+  br label %.preheader
+
+.preheader:                                       ; preds = %..preheader_crit_edge, %0
+  %i.0.lcssa = phi i64 [ %phitmp, %..preheader_crit_edge ], [ 0, %0 ]
+  %2 = icmp eq i32 %n, 0
+  br i1 %2, label %._crit_edge, label %.lr.ph
+
+.lr.ph5:                                          ; preds = %0, %.lr.ph5
+  %indvars.iv6 = phi i64 [ %indvars.iv.next7, %.lr.ph5 ], [ 0, %0 ]
+  %3 = getelementptr inbounds [2048 x i32]* @b, i64 0, i64 %indvars.iv6
+  store i32 %x, i32* %3, align 4
+  %indvars.iv.next7 = add i64 %indvars.iv6, 1
+  %lftr.wideiv = trunc i64 %indvars.iv.next7 to i32
+  %exitcond = icmp eq i32 %lftr.wideiv, %n
+  br i1 %exitcond, label %..preheader_crit_edge, label %.lr.ph5
+
+.lr.ph:                                           ; preds = %.preheader, %.lr.ph
+  %indvars.iv = phi i64 [ %indvars.iv.next, %.lr.ph ], [ %i.0.lcssa, %.preheader ]
+  %.02 = phi i32 [ %4, %.lr.ph ], [ %n, %.preheader ]
+  %4 = add nsw i32 %.02, -1
+  %5 = getelementptr inbounds [2048 x i32]* @b, i64 0, i64 %indvars.iv
+  %6 = load i32* %5, align 4
+  %7 = getelementptr inbounds [2048 x i32]* @c, i64 0, i64 %indvars.iv
+  %8 = load i32* %7, align 4
+  %9 = and i32 %8, %6
+  %10 = getelementptr inbounds [2048 x i32]* @a, i64 0, i64 %indvars.iv
+  store i32 %9, i32* %10, align 4
+  %indvars.iv.next = add i64 %indvars.iv, 1
+  %11 = icmp eq i32 %4, 0
+  br i1 %11, label %._crit_edge, label %.lr.ph
+
+._crit_edge:                                      ; preds = %.lr.ph, %.preheader
+  ret void
+}
+
+; We can't vectorize this loop because it has non constant loop bounds.
+;CHECK: @example3
+;CHECK-NOT: <4 x i32>
+;CHECK: ret void
+define void @example3(i32 %n, i32* noalias nocapture %p, i32* noalias nocapture %q) nounwind uwtable ssp {
+  %1 = icmp eq i32 %n, 0
+  br i1 %1, label %._crit_edge, label %.lr.ph
+
+.lr.ph:                                           ; preds = %0, %.lr.ph
+  %.05 = phi i32 [ %2, %.lr.ph ], [ %n, %0 ]
+  %.014 = phi i32* [ %5, %.lr.ph ], [ %p, %0 ]
+  %.023 = phi i32* [ %3, %.lr.ph ], [ %q, %0 ]
+  %2 = add nsw i32 %.05, -1
+  %3 = getelementptr inbounds i32* %.023, i64 1
+  %4 = load i32* %.023, align 16
+  %5 = getelementptr inbounds i32* %.014, i64 1
+  store i32 %4, i32* %.014, align 16
+  %6 = icmp eq i32 %2, 0
+  br i1 %6, label %._crit_edge, label %.lr.ph
+
+._crit_edge:                                      ; preds = %.lr.ph, %0
+  ret void
+}
+
+;CHECK: @example4
+;CHECK: load <4 x i32>
+;CHECK: ret void
+define void @example4(i32 %n, i32* noalias nocapture %p, i32* noalias nocapture %q) nounwind uwtable ssp {
+  %1 = add nsw i32 %n, -1
+  %2 = icmp eq i32 %n, 0
+  br i1 %2, label %.preheader4, label %.lr.ph10
+
+.preheader4:                                      ; preds = %0
+  %3 = icmp sgt i32 %1, 0
+  br i1 %3, label %.lr.ph6, label %._crit_edge
+
+.lr.ph10:                                         ; preds = %0, %.lr.ph10
+  %4 = phi i32 [ %9, %.lr.ph10 ], [ %1, %0 ]
+  %.018 = phi i32* [ %8, %.lr.ph10 ], [ %p, %0 ]
+  %.027 = phi i32* [ %5, %.lr.ph10 ], [ %q, %0 ]
+  %5 = getelementptr inbounds i32* %.027, i64 1
+  %6 = load i32* %.027, align 16
+  %7 = add nsw i32 %6, 5
+  %8 = getelementptr inbounds i32* %.018, i64 1
+  store i32 %7, i32* %.018, align 16
+  %9 = add nsw i32 %4, -1
+  %10 = icmp eq i32 %4, 0
+  br i1 %10, label %._crit_edge, label %.lr.ph10
+
+.preheader:                                       ; preds = %.lr.ph6
+  br i1 %3, label %.lr.ph, label %._crit_edge
+
+.lr.ph6:                                          ; preds = %.preheader4, %.lr.ph6
+  %indvars.iv11 = phi i64 [ %indvars.iv.next12, %.lr.ph6 ], [ 0, %.preheader4 ]
+  %indvars.iv.next12 = add i64 %indvars.iv11, 1
+  %11 = getelementptr inbounds [2048 x i32]* @b, i64 0, i64 %indvars.iv.next12
+  %12 = load i32* %11, align 4
+  %13 = add nsw i64 %indvars.iv11, 3
+  %14 = getelementptr inbounds [2048 x i32]* @c, i64 0, i64 %13
+  %15 = load i32* %14, align 4
+  %16 = add nsw i32 %15, %12
+  %17 = getelementptr inbounds [2048 x i32]* @a, i64 0, i64 %indvars.iv11
+  store i32 %16, i32* %17, align 4
+  %lftr.wideiv13 = trunc i64 %indvars.iv.next12 to i32
+  %exitcond14 = icmp eq i32 %lftr.wideiv13, %1
+  br i1 %exitcond14, label %.preheader, label %.lr.ph6
+
+.lr.ph:                                           ; preds = %.preheader, %.lr.ph
+  %indvars.iv = phi i64 [ %indvars.iv.next, %.lr.ph ], [ 0, %.preheader ]
+  %18 = getelementptr inbounds [2048 x i32]* @a, i64 0, i64 %indvars.iv
+  %19 = load i32* %18, align 4
+  %20 = icmp sgt i32 %19, 4
+  %21 = select i1 %20, i32 4, i32 0
+  %22 = getelementptr inbounds [2048 x i32]* @b, i64 0, i64 %indvars.iv
+  store i32 %21, i32* %22, align 4
+  %indvars.iv.next = add i64 %indvars.iv, 1
+  %lftr.wideiv = trunc i64 %indvars.iv.next to i32
+  %exitcond = icmp eq i32 %lftr.wideiv, %1
+  br i1 %exitcond, label %._crit_edge, label %.lr.ph
+
+._crit_edge:                                      ; preds = %.lr.ph10, %.preheader4, %.lr.ph, %.preheader
+  ret void
+}
+
+;CHECK: @example8
+;CHECK: store <4 x i32>
+;CHECK: ret void
+define void @example8(i32 %x) nounwind uwtable ssp {
+  br label %.preheader
+
+.preheader:                                       ; preds = %3, %0
+  %indvars.iv3 = phi i64 [ 0, %0 ], [ %indvars.iv.next4, %3 ]
+  br label %1
+
+; <label>:1                                       ; preds = %1, %.preheader
+  %indvars.iv = phi i64 [ 0, %.preheader ], [ %indvars.iv.next, %1 ]
+  %2 = getelementptr inbounds [32 x [1024 x i32]]* @G, i64 0, i64 %indvars.iv3, i64 %indvars.iv
+  store i32 %x, i32* %2, align 4
+  %indvars.iv.next = add i64 %indvars.iv, 1
+  %lftr.wideiv = trunc i64 %indvars.iv.next to i32
+  %exitcond = icmp eq i32 %lftr.wideiv, 1024
+  br i1 %exitcond, label %3, label %1
+
+; <label>:3                                       ; preds = %1
+  %indvars.iv.next4 = add i64 %indvars.iv3, 1
+  %lftr.wideiv5 = trunc i64 %indvars.iv.next4 to i32
+  %exitcond6 = icmp eq i32 %lftr.wideiv5, 32
+  br i1 %exitcond6, label %4, label %.preheader
+
+; <label>:4                                       ; preds = %3
+  ret void
+}
+
+;CHECK: @example9
+;CHECK: phi <4 x i32>
+;CHECK: ret i32
+define i32 @example9() nounwind uwtable readonly ssp {
+  br label %1
+
+; <label>:1                                       ; preds = %1, %0
+  %indvars.iv = phi i64 [ 0, %0 ], [ %indvars.iv.next, %1 ]
+  %diff.01 = phi i32 [ 0, %0 ], [ %7, %1 ]
+  %2 = getelementptr inbounds [1024 x i32]* @ub, i64 0, i64 %indvars.iv
+  %3 = load i32* %2, align 4
+  %4 = getelementptr inbounds [1024 x i32]* @uc, i64 0, i64 %indvars.iv
+  %5 = load i32* %4, align 4
+  %6 = add i32 %3, %diff.01
+  %7 = sub i32 %6, %5
+  %indvars.iv.next = add i64 %indvars.iv, 1
+  %lftr.wideiv = trunc i64 %indvars.iv.next to i32
+  %exitcond = icmp eq i32 %lftr.wideiv, 1024
+  br i1 %exitcond, label %8, label %1
+
+; <label>:8                                       ; preds = %1
+  ret i32 %7
+}
+
+;CHECK: @example10a
+;CHECK: load <4 x i16>
+;CHECK: add <4 x i16>
+;CHECK: store <4 x i16>
+;CHECK: ret void
+define void @example10a(i16* noalias nocapture %sa, i16* noalias nocapture %sb, i16* noalias nocapture %sc, i32* noalias nocapture %ia, i32* noalias nocapture %ib, i32* noalias nocapture %ic) nounwind uwtable ssp {
+  br label %1
+
+; <label>:1                                       ; preds = %1, %0
+  %indvars.iv = phi i64 [ 0, %0 ], [ %indvars.iv.next, %1 ]
+  %2 = getelementptr inbounds i32* %ib, i64 %indvars.iv
+  %3 = load i32* %2, align 4
+  %4 = getelementptr inbounds i32* %ic, i64 %indvars.iv
+  %5 = load i32* %4, align 4
+  %6 = add nsw i32 %5, %3
+  %7 = getelementptr inbounds i32* %ia, i64 %indvars.iv
+  store i32 %6, i32* %7, align 4
+  %8 = getelementptr inbounds i16* %sb, i64 %indvars.iv
+  %9 = load i16* %8, align 2
+  %10 = getelementptr inbounds i16* %sc, i64 %indvars.iv
+  %11 = load i16* %10, align 2
+  %12 = add i16 %11, %9
+  %13 = getelementptr inbounds i16* %sa, i64 %indvars.iv
+  store i16 %12, i16* %13, align 2
+  %indvars.iv.next = add i64 %indvars.iv, 1
+  %lftr.wideiv = trunc i64 %indvars.iv.next to i32
+  %exitcond = icmp eq i32 %lftr.wideiv, 1024
+  br i1 %exitcond, label %14, label %1
+
+; <label>:14                                      ; preds = %1
+  ret void
+}
+
+;CHECK: @example10b
+;CHECK: load <4 x i16>
+;CHECK: sext <4 x i16>
+;CHECK: store <4 x i32>
+;CHECK: ret void
+define void @example10b(i16* noalias nocapture %sa, i16* noalias nocapture %sb, i16* noalias nocapture %sc, i32* noalias nocapture %ia, i32* noalias nocapture %ib, i32* noalias nocapture %ic) nounwind uwtable ssp {
+  br label %1
+
+; <label>:1                                       ; preds = %1, %0
+  %indvars.iv = phi i64 [ 0, %0 ], [ %indvars.iv.next, %1 ]
+  %2 = getelementptr inbounds i16* %sb, i64 %indvars.iv
+  %3 = load i16* %2, align 2
+  %4 = sext i16 %3 to i32
+  %5 = getelementptr inbounds i32* %ia, i64 %indvars.iv
+  store i32 %4, i32* %5, align 4
+  %indvars.iv.next = add i64 %indvars.iv, 1
+  %lftr.wideiv = trunc i64 %indvars.iv.next to i32
+  %exitcond = icmp eq i32 %lftr.wideiv, 1024
+  br i1 %exitcond, label %6, label %1
+
+; <label>:6                                       ; preds = %1
+  ret void
+}
+
+;CHECK: @example11
+;CHECK: load i32
+;CHECK: load i32
+;CHECK: load i32
+;CHECK: load i32
+;CHECK: insertelement
+;CHECK: insertelement
+;CHECK: insertelement
+;CHECK: insertelement
+;CHECK: ret void
+define void @example11() nounwind uwtable ssp {
+  br label %1
+
+; <label>:1                                       ; preds = %1, %0
+  %indvars.iv = phi i64 [ 0, %0 ], [ %indvars.iv.next, %1 ]
+  %2 = shl nsw i64 %indvars.iv, 1
+  %3 = or i64 %2, 1
+  %4 = getelementptr inbounds [2048 x i32]* @b, i64 0, i64 %3
+  %5 = load i32* %4, align 4
+  %6 = getelementptr inbounds [2048 x i32]* @c, i64 0, i64 %3
+  %7 = load i32* %6, align 4
+  %8 = mul nsw i32 %7, %5
+  %9 = getelementptr inbounds [2048 x i32]* @b, i64 0, i64 %2
+  %10 = load i32* %9, align 8
+  %11 = getelementptr inbounds [2048 x i32]* @c, i64 0, i64 %2
+  %12 = load i32* %11, align 8
+  %13 = mul nsw i32 %12, %10
+  %14 = sub nsw i32 %8, %13
+  %15 = getelementptr inbounds [2048 x i32]* @a, i64 0, i64 %indvars.iv
+  store i32 %14, i32* %15, align 4
+  %16 = mul nsw i32 %7, %10
+  %17 = mul nsw i32 %12, %5
+  %18 = add nsw i32 %17, %16
+  %19 = getelementptr inbounds [2048 x i32]* @d, i64 0, i64 %indvars.iv
+  store i32 %18, i32* %19, align 4
+  %indvars.iv.next = add i64 %indvars.iv, 1
+  %lftr.wideiv = trunc i64 %indvars.iv.next to i32
+  %exitcond = icmp eq i32 %lftr.wideiv, 512
+  br i1 %exitcond, label %20, label %1
+
+; <label>:20                                      ; preds = %1
+  ret void
+}
+
+;CHECK: @example12
+;CHECK: trunc <4 x i64>
+;CHECK: store <4 x i32>
+;CHECK: ret void
+define void @example12() nounwind uwtable ssp {
+  br label %1
+
+; <label>:1                                       ; preds = %1, %0
+  %indvars.iv = phi i64 [ 0, %0 ], [ %indvars.iv.next, %1 ]
+  %2 = getelementptr inbounds [2048 x i32]* @a, i64 0, i64 %indvars.iv
+  %3 = trunc i64 %indvars.iv to i32
+  store i32 %3, i32* %2, align 4
+  %indvars.iv.next = add i64 %indvars.iv, 1
+  %lftr.wideiv = trunc i64 %indvars.iv.next to i32
+  %exitcond = icmp eq i32 %lftr.wideiv, 1024
+  br i1 %exitcond, label %4, label %1
+
+; <label>:4                                       ; preds = %1
+  ret void
+}
+
+; Can't vectorize because of reductions.
+;CHECK: @example13
+;CHECK-NOT: <4 x i32>
+;CHECK: ret void
+define void @example13(i32** nocapture %A, i32** nocapture %B, i32* nocapture %out) nounwind uwtable ssp {
+  br label %.preheader
+
+.preheader:                                       ; preds = %14, %0
+  %indvars.iv4 = phi i64 [ 0, %0 ], [ %indvars.iv.next5, %14 ]
+  %1 = getelementptr inbounds i32** %A, i64 %indvars.iv4
+  %2 = load i32** %1, align 8
+  %3 = getelementptr inbounds i32** %B, i64 %indvars.iv4
+  %4 = load i32** %3, align 8
+  br label %5
+
+; <label>:5                                       ; preds = %.preheader, %5
+  %indvars.iv = phi i64 [ 0, %.preheader ], [ %indvars.iv.next, %5 ]
+  %diff.02 = phi i32 [ 0, %.preheader ], [ %11, %5 ]
+  %6 = getelementptr inbounds i32* %2, i64 %indvars.iv
+  %7 = load i32* %6, align 4
+  %8 = getelementptr inbounds i32* %4, i64 %indvars.iv
+  %9 = load i32* %8, align 4
+  %10 = add i32 %7, %diff.02
+  %11 = sub i32 %10, %9
+  %indvars.iv.next = add i64 %indvars.iv, 8
+  %12 = trunc i64 %indvars.iv.next to i32
+  %13 = icmp slt i32 %12, 1024
+  br i1 %13, label %5, label %14
+
+; <label>:14                                      ; preds = %5
+  %15 = getelementptr inbounds i32* %out, i64 %indvars.iv4
+  store i32 %11, i32* %15, align 4
+  %indvars.iv.next5 = add i64 %indvars.iv4, 1
+  %lftr.wideiv = trunc i64 %indvars.iv.next5 to i32
+  %exitcond = icmp eq i32 %lftr.wideiv, 32
+  br i1 %exitcond, label %16, label %.preheader
+
+; <label>:16                                      ; preds = %14
+  ret void
+}
+
+; Can't vectorize because of reductions.
+;CHECK: @example14
+;CHECK-NOT: <4 x i32>
+;CHECK: ret void
+define void @example14(i32** nocapture %in, i32** nocapture %coeff, i32* nocapture %out) nounwind uwtable ssp {
+.preheader3:
+  br label %.preheader
+
+.preheader:                                       ; preds = %11, %.preheader3
+  %indvars.iv7 = phi i64 [ 0, %.preheader3 ], [ %indvars.iv.next8, %11 ]
+  %sum.05 = phi i32 [ 0, %.preheader3 ], [ %10, %11 ]
+  br label %0
+
+; <label>:0                                       ; preds = %0, %.preheader
+  %indvars.iv = phi i64 [ 0, %.preheader ], [ %indvars.iv.next, %0 ]
+  %sum.12 = phi i32 [ %sum.05, %.preheader ], [ %10, %0 ]
+  %1 = getelementptr inbounds i32** %in, i64 %indvars.iv
+  %2 = load i32** %1, align 8
+  %3 = getelementptr inbounds i32* %2, i64 %indvars.iv7
+  %4 = load i32* %3, align 4
+  %5 = getelementptr inbounds i32** %coeff, i64 %indvars.iv
+  %6 = load i32** %5, align 8
+  %7 = getelementptr inbounds i32* %6, i64 %indvars.iv7
+  %8 = load i32* %7, align 4
+  %9 = mul nsw i32 %8, %4
+  %10 = add nsw i32 %9, %sum.12
+  %indvars.iv.next = add i64 %indvars.iv, 1
+  %lftr.wideiv = trunc i64 %indvars.iv.next to i32
+  %exitcond = icmp eq i32 %lftr.wideiv, 1024
+  br i1 %exitcond, label %11, label %0
+
+; <label>:11                                      ; preds = %0
+  %indvars.iv.next8 = add i64 %indvars.iv7, 1
+  %lftr.wideiv9 = trunc i64 %indvars.iv.next8 to i32
+  %exitcond10 = icmp eq i32 %lftr.wideiv9, 32
+  br i1 %exitcond10, label %.preheader3.1, label %.preheader
+
+.preheader3.1:                                    ; preds = %11
+  store i32 %10, i32* %out, align 4
+  br label %.preheader.1
+
+.preheader.1:                                     ; preds = %24, %.preheader3.1
+  %indvars.iv7.1 = phi i64 [ 0, %.preheader3.1 ], [ %indvars.iv.next8.1, %24 ]
+  %sum.05.1 = phi i32 [ 0, %.preheader3.1 ], [ %23, %24 ]
+  br label %12
+
+; <label>:12                                      ; preds = %12, %.preheader.1
+  %indvars.iv.1 = phi i64 [ 0, %.preheader.1 ], [ %13, %12 ]
+  %sum.12.1 = phi i32 [ %sum.05.1, %.preheader.1 ], [ %23, %12 ]
+  %13 = add nsw i64 %indvars.iv.1, 1
+  %14 = getelementptr inbounds i32** %in, i64 %13
+  %15 = load i32** %14, align 8
+  %16 = getelementptr inbounds i32* %15, i64 %indvars.iv7.1
+  %17 = load i32* %16, align 4
+  %18 = getelementptr inbounds i32** %coeff, i64 %indvars.iv.1
+  %19 = load i32** %18, align 8
+  %20 = getelementptr inbounds i32* %19, i64 %indvars.iv7.1
+  %21 = load i32* %20, align 4
+  %22 = mul nsw i32 %21, %17
+  %23 = add nsw i32 %22, %sum.12.1
+  %lftr.wideiv.1 = trunc i64 %13 to i32
+  %exitcond.1 = icmp eq i32 %lftr.wideiv.1, 1024
+  br i1 %exitcond.1, label %24, label %12
+
+; <label>:24                                      ; preds = %12
+  %indvars.iv.next8.1 = add i64 %indvars.iv7.1, 1
+  %lftr.wideiv9.1 = trunc i64 %indvars.iv.next8.1 to i32
+  %exitcond10.1 = icmp eq i32 %lftr.wideiv9.1, 32
+  br i1 %exitcond10.1, label %.preheader3.2, label %.preheader.1
+
+.preheader3.2:                                    ; preds = %24
+  %25 = getelementptr inbounds i32* %out, i64 1
+  store i32 %23, i32* %25, align 4
+  br label %.preheader.2
+
+.preheader.2:                                     ; preds = %38, %.preheader3.2
+  %indvars.iv7.2 = phi i64 [ 0, %.preheader3.2 ], [ %indvars.iv.next8.2, %38 ]
+  %sum.05.2 = phi i32 [ 0, %.preheader3.2 ], [ %37, %38 ]
+  br label %26
+
+; <label>:26                                      ; preds = %26, %.preheader.2
+  %indvars.iv.2 = phi i64 [ 0, %.preheader.2 ], [ %indvars.iv.next.2, %26 ]
+  %sum.12.2 = phi i32 [ %sum.05.2, %.preheader.2 ], [ %37, %26 ]
+  %27 = add nsw i64 %indvars.iv.2, 2
+  %28 = getelementptr inbounds i32** %in, i64 %27
+  %29 = load i32** %28, align 8
+  %30 = getelementptr inbounds i32* %29, i64 %indvars.iv7.2
+  %31 = load i32* %30, align 4
+  %32 = getelementptr inbounds i32** %coeff, i64 %indvars.iv.2
+  %33 = load i32** %32, align 8
+  %34 = getelementptr inbounds i32* %33, i64 %indvars.iv7.2
+  %35 = load i32* %34, align 4
+  %36 = mul nsw i32 %35, %31
+  %37 = add nsw i32 %36, %sum.12.2
+  %indvars.iv.next.2 = add i64 %indvars.iv.2, 1
+  %lftr.wideiv.2 = trunc i64 %indvars.iv.next.2 to i32
+  %exitcond.2 = icmp eq i32 %lftr.wideiv.2, 1024
+  br i1 %exitcond.2, label %38, label %26
+
+; <label>:38                                      ; preds = %26
+  %indvars.iv.next8.2 = add i64 %indvars.iv7.2, 1
+  %lftr.wideiv9.2 = trunc i64 %indvars.iv.next8.2 to i32
+  %exitcond10.2 = icmp eq i32 %lftr.wideiv9.2, 32
+  br i1 %exitcond10.2, label %.preheader3.3, label %.preheader.2
+
+.preheader3.3:                                    ; preds = %38
+  %39 = getelementptr inbounds i32* %out, i64 2
+  store i32 %37, i32* %39, align 4
+  br label %.preheader.3
+
+.preheader.3:                                     ; preds = %52, %.preheader3.3
+  %indvars.iv7.3 = phi i64 [ 0, %.preheader3.3 ], [ %indvars.iv.next8.3, %52 ]
+  %sum.05.3 = phi i32 [ 0, %.preheader3.3 ], [ %51, %52 ]
+  br label %40
+
+; <label>:40                                      ; preds = %40, %.preheader.3
+  %indvars.iv.3 = phi i64 [ 0, %.preheader.3 ], [ %indvars.iv.next.3, %40 ]
+  %sum.12.3 = phi i32 [ %sum.05.3, %.preheader.3 ], [ %51, %40 ]
+  %41 = add nsw i64 %indvars.iv.3, 3
+  %42 = getelementptr inbounds i32** %in, i64 %41
+  %43 = load i32** %42, align 8
+  %44 = getelementptr inbounds i32* %43, i64 %indvars.iv7.3
+  %45 = load i32* %44, align 4
+  %46 = getelementptr inbounds i32** %coeff, i64 %indvars.iv.3
+  %47 = load i32** %46, align 8
+  %48 = getelementptr inbounds i32* %47, i64 %indvars.iv7.3
+  %49 = load i32* %48, align 4
+  %50 = mul nsw i32 %49, %45
+  %51 = add nsw i32 %50, %sum.12.3
+  %indvars.iv.next.3 = add i64 %indvars.iv.3, 1
+  %lftr.wideiv.3 = trunc i64 %indvars.iv.next.3 to i32
+  %exitcond.3 = icmp eq i32 %lftr.wideiv.3, 1024
+  br i1 %exitcond.3, label %52, label %40
+
+; <label>:52                                      ; preds = %40
+  %indvars.iv.next8.3 = add i64 %indvars.iv7.3, 1
+  %lftr.wideiv9.3 = trunc i64 %indvars.iv.next8.3 to i32
+  %exitcond10.3 = icmp eq i32 %lftr.wideiv9.3, 32
+  br i1 %exitcond10.3, label %53, label %.preheader.3
+
+; <label>:53                                      ; preds = %52
+  %54 = getelementptr inbounds i32* %out, i64 3
+  store i32 %51, i32* %54, align 4
+  ret void
+}
+
+; Can't vectorize because the src and dst pointers are not disjoint.
+;CHECK: @example21
+;CHECK-NOT: <4 x i32>
+;CHECK: ret i32
+define i32 @example21(i32* nocapture %b, i32 %n) nounwind uwtable readonly ssp {
+  %1 = icmp sgt i32 %n, 0
+  br i1 %1, label %.lr.ph, label %._crit_edge
+
+.lr.ph:                                           ; preds = %0
+  %2 = sext i32 %n to i64
+  br label %3
+
+; <label>:3                                       ; preds = %.lr.ph, %3
+  %indvars.iv = phi i64 [ %2, %.lr.ph ], [ %indvars.iv.next, %3 ]
+  %a.02 = phi i32 [ 0, %.lr.ph ], [ %6, %3 ]
+  %indvars.iv.next = add i64 %indvars.iv, -1
+  %4 = getelementptr inbounds i32* %b, i64 %indvars.iv.next
+  %5 = load i32* %4, align 4
+  %6 = add nsw i32 %5, %a.02
+  %7 = trunc i64 %indvars.iv.next to i32
+  %8 = icmp sgt i32 %7, 0
+  br i1 %8, label %3, label %._crit_edge
+
+._crit_edge:                                      ; preds = %3, %0
+  %a.0.lcssa = phi i32 [ 0, %0 ], [ %6, %3 ]
+  ret i32 %a.0.lcssa
+}
+
+; Can't vectorize because there are multiple PHIs.
+;CHECK: @example23
+;CHECK-NOT: <4 x i32>
+;CHECK: ret void
+define void @example23(i16* nocapture %src, i32* nocapture %dst) nounwind uwtable ssp {
+  br label %1
+
+; <label>:1                                       ; preds = %1, %0
+  %.04 = phi i16* [ %src, %0 ], [ %2, %1 ]
+  %.013 = phi i32* [ %dst, %0 ], [ %6, %1 ]
+  %i.02 = phi i32 [ 0, %0 ], [ %7, %1 ]
+  %2 = getelementptr inbounds i16* %.04, i64 1
+  %3 = load i16* %.04, align 2
+  %4 = zext i16 %3 to i32
+  %5 = shl nuw nsw i32 %4, 7
+  %6 = getelementptr inbounds i32* %.013, i64 1
+  store i32 %5, i32* %.013, align 4
+  %7 = add nsw i32 %i.02, 1
+  %exitcond = icmp eq i32 %7, 256
+  br i1 %exitcond, label %8, label %1
+
+; <label>:8                                       ; preds = %1
+  ret void
+}
+
+;CHECK: @example24
+;CHECK: shufflevector <4 x i16>
+;CHECK: ret void
+define void @example24(i16 signext %x, i16 signext %y) nounwind uwtable ssp {
+  br label %1
+
+; <label>:1                                       ; preds = %1, %0
+  %indvars.iv = phi i64 [ 0, %0 ], [ %indvars.iv.next, %1 ]
+  %2 = getelementptr inbounds [1024 x float]* @fa, i64 0, i64 %indvars.iv
+  %3 = load float* %2, align 4
+  %4 = getelementptr inbounds [1024 x float]* @fb, i64 0, i64 %indvars.iv
+  %5 = load float* %4, align 4
+  %6 = fcmp olt float %3, %5
+  %x.y = select i1 %6, i16 %x, i16 %y
+  %7 = sext i16 %x.y to i32
+  %8 = getelementptr inbounds [1024 x i32]* @ic, i64 0, i64 %indvars.iv
+  store i32 %7, i32* %8, align 4
+  %indvars.iv.next = add i64 %indvars.iv, 1
+  %lftr.wideiv = trunc i64 %indvars.iv.next to i32
+  %exitcond = icmp eq i32 %lftr.wideiv, 1024
+  br i1 %exitcond, label %9, label %1
+
+; <label>:9                                       ; preds = %1
+  ret void
+}
+
+;CHECK: @example25
+;CHECK: and <4 x i1>
+;CHECK: zext <4 x i1>
+;CHECK: ret void
+define void @example25() nounwind uwtable ssp {
+  br label %1
+
+; <label>:1                                       ; preds = %1, %0
+  %indvars.iv = phi i64 [ 0, %0 ], [ %indvars.iv.next, %1 ]
+  %2 = getelementptr inbounds [1024 x float]* @da, i64 0, i64 %indvars.iv
+  %3 = load float* %2, align 4
+  %4 = getelementptr inbounds [1024 x float]* @db, i64 0, i64 %indvars.iv
+  %5 = load float* %4, align 4
+  %6 = fcmp olt float %3, %5
+  %7 = getelementptr inbounds [1024 x float]* @dc, i64 0, i64 %indvars.iv
+  %8 = load float* %7, align 4
+  %9 = getelementptr inbounds [1024 x float]* @dd, i64 0, i64 %indvars.iv
+  %10 = load float* %9, align 4
+  %11 = fcmp olt float %8, %10
+  %12 = and i1 %6, %11
+  %13 = zext i1 %12 to i32
+  %14 = getelementptr inbounds [1024 x i32]* @dj, i64 0, i64 %indvars.iv
+  store i32 %13, i32* %14, align 4
+  %indvars.iv.next = add i64 %indvars.iv, 1
+  %lftr.wideiv = trunc i64 %indvars.iv.next to i32
+  %exitcond = icmp eq i32 %lftr.wideiv, 1024
+  br i1 %exitcond, label %15, label %1
+
+; <label>:15                                      ; preds = %1
+  ret void
+}
+
diff --git a/test/Transforms/LoopVectorize/increment.ll b/test/Transforms/LoopVectorize/increment.ll
new file mode 100644
index 00000000000..e818d685626
--- /dev/null
+++ b/test/Transforms/LoopVectorize/increment.ll
@@ -0,0 +1,66 @@
+; RUN: opt < %s  -loop-vectorize -dce -instcombine -licm -S | FileCheck %s
+
+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-S128"
+target triple = "x86_64-apple-macosx10.8.0"
+
+@a = common global [2048 x i32] zeroinitializer, align 16
+
+; This is the loop.
+;  for (i=0; i<n; i++){
+;    a[i] += i;
+;  }
+;CHECK: @inc
+;CHECK: load <4 x i32>
+;CHECK: add <4 x i32>
+;CHECK: store <4 x i32>
+;CHECK: ret void
+define void @inc(i32 %n) nounwind uwtable noinline ssp {
+  %1 = icmp sgt i32 %n, 0
+  br i1 %1, label %.lr.ph, label %._crit_edge
+
+.lr.ph:                                           ; preds = %0, %.lr.ph
+  %indvars.iv = phi i64 [ %indvars.iv.next, %.lr.ph ], [ 0, %0 ]
+  %2 = getelementptr inbounds [2048 x i32]* @a, i64 0, i64 %indvars.iv
+  %3 = load i32* %2, align 4
+  %4 = trunc i64 %indvars.iv to i32
+  %5 = add nsw i32 %3, %4
+  store i32 %5, i32* %2, align 4
+  %indvars.iv.next = add i64 %indvars.iv, 1
+  %lftr.wideiv = trunc i64 %indvars.iv.next to i32
+  %exitcond = icmp eq i32 %lftr.wideiv, %n
+  br i1 %exitcond, label %._crit_edge, label %.lr.ph
+
+._crit_edge:                                      ; preds = %.lr.ph, %0
+  ret void
+}
+
+; Can't vectorize this loop because the access to A[X] is non linear.
+;
+;  for (i = 0; i < n; ++i) {
+;    A[B[i]]++;
+;
+;CHECK: @histogram
+;CHECK-NOT: <4 x i32>
+;CHECK: ret i32
+define i32 @histogram(i32* nocapture noalias %A, i32* nocapture noalias %B, i32 %n) nounwind uwtable ssp {
+entry:
+  %cmp6 = icmp sgt i32 %n, 0
+  br i1 %cmp6, label %for.body, label %for.end
+
+for.body:                                         ; preds = %entry, %for.body
+  %indvars.iv = phi i64 [ %indvars.iv.next, %for.body ], [ 0, %entry ]
+  %arrayidx = getelementptr inbounds i32* %B, i64 %indvars.iv
+  %0 = load i32* %arrayidx, align 4
+  %idxprom1 = sext i32 %0 to i64
+  %arrayidx2 = getelementptr inbounds i32* %A, i64 %idxprom1
+  %1 = load i32* %arrayidx2, align 4
+  %inc = add nsw i32 %1, 1
+  store i32 %inc, i32* %arrayidx2, align 4
+  %indvars.iv.next = add i64 %indvars.iv, 1
+  %lftr.wideiv = trunc i64 %indvars.iv.next to i32
+  %exitcond = icmp eq i32 %lftr.wideiv, %n
+  br i1 %exitcond, label %for.end, label %for.body
+
+for.end:                                          ; preds = %for.body, %entry
+  ret i32 0
+}
diff --git a/test/Transforms/LoopVectorize/induction_plus.ll b/test/Transforms/LoopVectorize/induction_plus.ll
new file mode 100644
index 00000000000..bd90113e523
--- /dev/null
+++ b/test/Transforms/LoopVectorize/induction_plus.ll
@@ -0,0 +1,30 @@
+; RUN: opt < %s -loop-vectorize -instcombine -S | FileCheck %s
+
+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-S128"
+target triple = "x86_64-apple-macosx10.8.0"
+
+@array = common global [1024 x i32] zeroinitializer, align 16
+
+;CHECK: @array_at_plus_one
+;CHECK: add <4 x i64>
+;CHECK: trunc <4 x i64>
+;CHECK: add i64 %index, 12
+;CHECK: ret i32
+define i32 @array_at_plus_one(i32 %n) nounwind uwtable ssp {
+  %1 = icmp sgt i32 %n, 0
+  br i1 %1, label %.lr.ph, label %._crit_edge
+
+.lr.ph:                                           ; preds = %0, %.lr.ph
+  %indvars.iv = phi i64 [ %indvars.iv.next, %.lr.ph ], [ 0, %0 ]
+  %2 = add nsw i64 %indvars.iv, 12
+  %3 = getelementptr inbounds [1024 x i32]* @array, i64 0, i64 %2
+  %4 = trunc i64 %indvars.iv to i32
+  store i32 %4, i32* %3, align 4
+  %indvars.iv.next = add i64 %indvars.iv, 1
+  %lftr.wideiv = trunc i64 %indvars.iv.next to i32
+  %exitcond = icmp eq i32 %lftr.wideiv, %n
+  br i1 %exitcond, label %._crit_edge, label %.lr.ph
+
+._crit_edge:                                      ; preds = %.lr.ph, %0
+  ret i32 undef
+}
diff --git a/test/Transforms/LoopVectorize/lit.local.cfg b/test/Transforms/LoopVectorize/lit.local.cfg
new file mode 100644
index 00000000000..19eebc0ac7a
--- /dev/null
+++ b/test/Transforms/LoopVectorize/lit.local.cfg
@@ -0,0 +1 @@
+config.suffixes = ['.ll', '.c', '.cpp']
diff --git a/test/Transforms/LoopVectorize/non-const-n.ll b/test/Transforms/LoopVectorize/non-const-n.ll
new file mode 100644
index 00000000000..04c5c84a4f4
--- /dev/null
+++ b/test/Transforms/LoopVectorize/non-const-n.ll
@@ -0,0 +1,38 @@
+; RUN: opt < %s  -loop-vectorize -dce -instcombine -licm -S | FileCheck %s
+
+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-S128"
+target triple = "x86_64-apple-macosx10.8.0"
+
+@b = common global [2048 x i32] zeroinitializer, align 16
+@c = common global [2048 x i32] zeroinitializer, align 16
+@a = common global [2048 x i32] zeroinitializer, align 16
+
+;CHECK: @example1
+;CHECK: shl i32
+;CHECK: zext i32
+;CHECK: load <4 x i32>
+;CHECK: add <4 x i32>
+;CHECK: store <4 x i32>
+;CHECK: ret void
+define void @example1(i32 %n) nounwind uwtable ssp {
+  %n4 = shl i32 %n, 2
+  br label %1
+
+; <label>:1                                       ; preds = %1, %0
+  %indvars.iv = phi i64 [ 0, %0 ], [ %indvars.iv.next, %1 ]
+  %2 = getelementptr inbounds [2048 x i32]* @b, i64 0, i64 %indvars.iv
+  %3 = load i32* %2, align 4
+  %4 = getelementptr inbounds [2048 x i32]* @c, i64 0, i64 %indvars.iv
+  %5 = load i32* %4, align 4
+  %6 = add nsw i32 %5, %3
+  %7 = getelementptr inbounds [2048 x i32]* @a, i64 0, i64 %indvars.iv
+  store i32 %6, i32* %7, align 4
+  %indvars.iv.next = add i64 %indvars.iv, 1
+  %lftr.wideiv = trunc i64 %indvars.iv.next to i32
+  %exitcond = icmp eq i32 %lftr.wideiv, %n4
+  br i1 %exitcond, label %8, label %1
+
+; <label>:8                                       ; preds = %1
+  ret void
+}
+
diff --git a/test/Transforms/LoopVectorize/read-only.ll b/test/Transforms/LoopVectorize/read-only.ll
new file mode 100644
index 00000000000..4095ea68ef7
--- /dev/null
+++ b/test/Transforms/LoopVectorize/read-only.ll
@@ -0,0 +1,32 @@
+; RUN: opt < %s  -loop-vectorize -dce -instcombine -licm -S | FileCheck %s
+
+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-S128"
+target triple = "x86_64-apple-macosx10.8.0"
+
+;CHECK: @read_only_func
+;CHECK: load <4 x i32>
+;CHECK: ret i32
+define i32 @read_only_func(i32* nocapture %A, i32* nocapture %B, i32 %n) nounwind uwtable readonly ssp {
+  %1 = icmp sgt i32 %n, 0
+  br i1 %1, label %.lr.ph, label %._crit_edge
+
+.lr.ph:                                           ; preds = %0, %.lr.ph
+  %indvars.iv = phi i64 [ %indvars.iv.next, %.lr.ph ], [ 0, %0 ]
+  %sum.02 = phi i32 [ %9, %.lr.ph ], [ 0, %0 ]
+  %2 = getelementptr inbounds i32* %A, i64 %indvars.iv
+  %3 = load i32* %2, align 4
+  %4 = add nsw i64 %indvars.iv, 13
+  %5 = getelementptr inbounds i32* %B, i64 %4
+  %6 = load i32* %5, align 4
+  %7 = shl i32 %6, 1
+  %8 = add i32 %3, %sum.02
+  %9 = add i32 %8, %7
+  %indvars.iv.next = add i64 %indvars.iv, 1
+  %lftr.wideiv = trunc i64 %indvars.iv.next to i32
+  %exitcond = icmp eq i32 %lftr.wideiv, %n
+  br i1 %exitcond, label %._crit_edge, label %.lr.ph
+
+._crit_edge:                                      ; preds = %.lr.ph, %0
+  %sum.0.lcssa = phi i32 [ 0, %0 ], [ %9, %.lr.ph ]
+  ret i32 %sum.0.lcssa
+}
diff --git a/test/Transforms/LoopVectorize/reduction.ll b/test/Transforms/LoopVectorize/reduction.ll
new file mode 100644
index 00000000000..3e871b229b5
--- /dev/null
+++ b/test/Transforms/LoopVectorize/reduction.ll
@@ -0,0 +1,152 @@
+; RUN: opt < %s  -loop-vectorize -dce -instcombine -licm -S | FileCheck %s
+
+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-S128"
+target triple = "x86_64-apple-macosx10.8.0"
+
+;CHECK: @reduction_sum
+;CHECK: phi <4 x i32>
+;CHECK: load <4 x i32>
+;CHECK: add <4 x i32>
+;CHECK: ret i32
+define i32 @reduction_sum(i32 %n, i32* noalias nocapture %A, i32* noalias nocapture %B) nounwind uwtable readonly noinline ssp {
+  %1 = icmp sgt i32 %n, 0
+  br i1 %1, label %.lr.ph, label %._crit_edge
+
+.lr.ph:                                           ; preds = %0, %.lr.ph
+  %indvars.iv = phi i64 [ %indvars.iv.next, %.lr.ph ], [ 0, %0 ]
+  %sum.02 = phi i32 [ %9, %.lr.ph ], [ 0, %0 ]
+  %2 = getelementptr inbounds i32* %A, i64 %indvars.iv
+  %3 = load i32* %2, align 4
+  %4 = getelementptr inbounds i32* %B, i64 %indvars.iv
+  %5 = load i32* %4, align 4
+  %6 = trunc i64 %indvars.iv to i32
+  %7 = add i32 %sum.02, %6
+  %8 = add i32 %7, %3
+  %9 = add i32 %8, %5
+  %indvars.iv.next = add i64 %indvars.iv, 1
+  %lftr.wideiv = trunc i64 %indvars.iv.next to i32
+  %exitcond = icmp eq i32 %lftr.wideiv, %n
+  br i1 %exitcond, label %._crit_edge, label %.lr.ph
+
+._crit_edge:                                      ; preds = %.lr.ph, %0
+  %sum.0.lcssa = phi i32 [ 0, %0 ], [ %9, %.lr.ph ]
+  ret i32 %sum.0.lcssa
+}
+
+;CHECK: @reduction_prod
+;CHECK: phi <4 x i32>
+;CHECK: load <4 x i32>
+;CHECK: mul <4 x i32>
+;CHECK: ret i32
+define i32 @reduction_prod(i32 %n, i32* noalias nocapture %A, i32* noalias nocapture %B) nounwind uwtable readonly noinline ssp {
+  %1 = icmp sgt i32 %n, 0
+  br i1 %1, label %.lr.ph, label %._crit_edge
+
+.lr.ph:                                           ; preds = %0, %.lr.ph
+  %indvars.iv = phi i64 [ %indvars.iv.next, %.lr.ph ], [ 0, %0 ]
+  %prod.02 = phi i32 [ %9, %.lr.ph ], [ 1, %0 ]
+  %2 = getelementptr inbounds i32* %A, i64 %indvars.iv
+  %3 = load i32* %2, align 4
+  %4 = getelementptr inbounds i32* %B, i64 %indvars.iv
+  %5 = load i32* %4, align 4
+  %6 = trunc i64 %indvars.iv to i32
+  %7 = mul i32 %prod.02, %6
+  %8 = mul i32 %7, %3
+  %9 = mul i32 %8, %5
+  %indvars.iv.next = add i64 %indvars.iv, 1
+  %lftr.wideiv = trunc i64 %indvars.iv.next to i32
+  %exitcond = icmp eq i32 %lftr.wideiv, %n
+  br i1 %exitcond, label %._crit_edge, label %.lr.ph
+
+._crit_edge:                                      ; preds = %.lr.ph, %0
+  %prod.0.lcssa = phi i32 [ 1, %0 ], [ %9, %.lr.ph ]
+  ret i32 %prod.0.lcssa
+}
+
+;CHECK: @reduction_mix
+;CHECK: phi <4 x i32>
+;CHECK: load <4 x i32>
+;CHECK: mul <4 x i32>
+;CHECK: ret i32
+define i32 @reduction_mix(i32 %n, i32* noalias nocapture %A, i32* noalias nocapture %B) nounwind uwtable readonly noinline ssp {
+  %1 = icmp sgt i32 %n, 0
+  br i1 %1, label %.lr.ph, label %._crit_edge
+
+.lr.ph:                                           ; preds = %0, %.lr.ph
+  %indvars.iv = phi i64 [ %indvars.iv.next, %.lr.ph ], [ 0, %0 ]
+  %sum.02 = phi i32 [ %9, %.lr.ph ], [ 0, %0 ]
+  %2 = getelementptr inbounds i32* %A, i64 %indvars.iv
+  %3 = load i32* %2, align 4
+  %4 = getelementptr inbounds i32* %B, i64 %indvars.iv
+  %5 = load i32* %4, align 4
+  %6 = mul nsw i32 %5, %3
+  %7 = trunc i64 %indvars.iv to i32
+  %8 = add i32 %sum.02, %7
+  %9 = add i32 %8, %6
+  %indvars.iv.next = add i64 %indvars.iv, 1
+  %lftr.wideiv = trunc i64 %indvars.iv.next to i32
+  %exitcond = icmp eq i32 %lftr.wideiv, %n
+  br i1 %exitcond, label %._crit_edge, label %.lr.ph
+
+._crit_edge:                                      ; preds = %.lr.ph, %0
+  %sum.0.lcssa = phi i32 [ 0, %0 ], [ %9, %.lr.ph ]
+  ret i32 %sum.0.lcssa
+}
+
+;CHECK: @reduction_mul
+;CHECK: mul <4 x i32>
+;CHECK: ret i32
+define i32 @reduction_mul(i32 %n, i32* noalias nocapture %A, i32* noalias nocapture %B) nounwind uwtable readonly noinline ssp {
+  %1 = icmp sgt i32 %n, 0
+  br i1 %1, label %.lr.ph, label %._crit_edge
+
+.lr.ph:                                           ; preds = %0, %.lr.ph
+  %indvars.iv = phi i64 [ %indvars.iv.next, %.lr.ph ], [ 0, %0 ]
+  %sum.02 = phi i32 [ %9, %.lr.ph ], [ 19, %0 ]
+  %2 = getelementptr inbounds i32* %A, i64 %indvars.iv
+  %3 = load i32* %2, align 4
+  %4 = getelementptr inbounds i32* %B, i64 %indvars.iv
+  %5 = load i32* %4, align 4
+  %6 = trunc i64 %indvars.iv to i32
+  %7 = add i32 %3, %6
+  %8 = add i32 %7, %5
+  %9 = mul i32 %8, %sum.02
+  %indvars.iv.next = add i64 %indvars.iv, 1
+  %lftr.wideiv = trunc i64 %indvars.iv.next to i32
+  %exitcond = icmp eq i32 %lftr.wideiv, %n
+  br i1 %exitcond, label %._crit_edge, label %.lr.ph
+
+._crit_edge:                                      ; preds = %.lr.ph, %0
+  %sum.0.lcssa = phi i32 [ 0, %0 ], [ %9, %.lr.ph ]
+  ret i32 %sum.0.lcssa
+}
+
+;CHECK: @start_at_non_zero
+;CHECK: phi <4 x i32>
+;CHECK: <i32 120, i32 0, i32 0, i32 0>
+;CHECK: ret i32
+define i32 @start_at_non_zero(i32* nocapture %in, i32* nocapture %coeff, i32* nocapture %out, i32 %n) nounwind uwtable readonly ssp {
+entry:
+  %cmp7 = icmp sgt i32 %n, 0
+  br i1 %cmp7, label %for.body, label %for.end
+
+for.body:                                         ; preds = %entry, %for.body
+  %indvars.iv = phi i64 [ %indvars.iv.next, %for.body ], [ 0, %entry ]
+  %sum.09 = phi i32 [ %add, %for.body ], [ 120, %entry ]
+  %arrayidx = getelementptr inbounds i32* %in, i64 %indvars.iv
+  %0 = load i32* %arrayidx, align 4
+  %arrayidx2 = getelementptr inbounds i32* %coeff, i64 %indvars.iv
+  %1 = load i32* %arrayidx2, align 4
+  %mul = mul nsw i32 %1, %0
+  %add = add nsw i32 %mul, %sum.09
+  %indvars.iv.next = add i64 %indvars.iv, 1
+  %lftr.wideiv = trunc i64 %indvars.iv.next to i32
+  %exitcond = icmp eq i32 %lftr.wideiv, %n
+  br i1 %exitcond, label %for.end, label %for.body
+
+for.end:                                          ; preds = %for.body, %entry
+  %sum.0.lcssa = phi i32 [ 120, %entry ], [ %add, %for.body ]
+  ret i32 %sum.0.lcssa
+}
+
+
diff --git a/test/Transforms/LoopVectorize/scalar-select.ll b/test/Transforms/LoopVectorize/scalar-select.ll
new file mode 100644
index 00000000000..8d5b6fd8aff
--- /dev/null
+++ b/test/Transforms/LoopVectorize/scalar-select.ll
@@ -0,0 +1,37 @@
+; RUN: opt < %s  -loop-vectorize -dce -instcombine -licm -S | FileCheck %s
+
+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-S128"
+target triple = "x86_64-apple-macosx10.8.0"
+
+@a = common global [2048 x i32] zeroinitializer, align 16
+@b = common global [2048 x i32] zeroinitializer, align 16
+@c = common global [2048 x i32] zeroinitializer, align 16
+
+;CHECK: @example1
+;CHECK: load <4 x i32>
+; make sure that we have a scalar condition and a vector operand.
+;CHECK: select i1 %cond, <4 x i32>
+;CHECK: store <4 x i32>
+;CHECK: ret void
+define void @example1(i1 %cond) nounwind uwtable ssp {
+  br label %1
+
+; <label>:1                                       ; preds = %1, %0
+  %indvars.iv = phi i64 [ 0, %0 ], [ %indvars.iv.next, %1 ]
+  %2 = getelementptr inbounds [2048 x i32]* @b, i64 0, i64 %indvars.iv
+  %3 = load i32* %2, align 4
+  %4 = getelementptr inbounds [2048 x i32]* @c, i64 0, i64 %indvars.iv
+  %5 = load i32* %4, align 4
+  %6 = add nsw i32 %5, %3
+  %7 = getelementptr inbounds [2048 x i32]* @a, i64 0, i64 %indvars.iv
+  %sel = select i1 %cond, i32 %6, i32 zeroinitializer
+  store i32 %sel, i32* %7, align 4
+  %indvars.iv.next = add i64 %indvars.iv, 1
+  %lftr.wideiv = trunc i64 %indvars.iv.next to i32
+  %exitcond = icmp eq i32 %lftr.wideiv, 256
+  br i1 %exitcond, label %8, label %1
+
+; <label>:8                                       ; preds = %1
+  ret void
+}
+
diff --git a/test/Transforms/SROA/basictest.ll b/test/Transforms/SROA/basictest.ll
index 644fda167d4..03120f7a327 100644
--- a/test/Transforms/SROA/basictest.ll
+++ b/test/Transforms/SROA/basictest.ll
@@ -1063,3 +1063,23 @@ entry:
   call void @llvm.lifetime.end(i64 -1, i8* %0)
   ret void
 }
+
+define void @PR14105({ [16 x i8] }* %ptr) {
+; Ensure that when rewriting the GEP index '-1' for this alloca we preserve is
+; sign as negative. We use a volatile memcpy to ensure promotion never actually
+; occurs.
+; CHECK: @PR14105
+
+entry:
+  %a = alloca { [16 x i8] }, align 8
+; CHECK: alloca [16 x i8], align 8
+
+  %gep = getelementptr inbounds { [16 x i8] }* %ptr, i64 -1
+; CHECK-NEXT: getelementptr inbounds { [16 x i8] }* %ptr, i64 -1, i32 0, i64 0
+
+  %cast1 = bitcast { [16 x i8 ] }* %gep to i8*
+  %cast2 = bitcast { [16 x i8 ] }* %a to i8*
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %cast1, i8* %cast2, i32 16, i32 8, i1 true)
+  ret void
+; CHECK: ret
+}
diff --git a/test/Transforms/SimplifyLibCalls/FFS.ll b/test/Transforms/SimplifyLibCalls/FFS.ll
index e38d78349d4..6aecbeacd7e 100644
--- a/test/Transforms/SimplifyLibCalls/FFS.ll
+++ b/test/Transforms/SimplifyLibCalls/FFS.ll
@@ -1,6 +1,7 @@
-; Test that the ToAsciiOptimizer works correctly
-; RUN: opt < %s -simplify-libcalls -S | \
-; RUN:   not grep "call.*@ffs"
+; Test that FFSOpt works correctly
+; RUN: opt < %s -simplify-libcalls -S | FileCheck %s
+
+; CHECK-NOT: call{{.*}}@ffs
 
 @non_const = external global i32		; <i32*> [#uses=1]
 
@@ -34,3 +35,11 @@ define i32 @a(i64) nounwind {
         %2 = call i32 @ffsll(i64 %0)            ; <i32> [#uses=1]
         ret i32 %2
 }
+
+; PR13028
+define i32 @b() nounwind {
+  %ffs = call i32 @ffsll(i64 0)
+  ret i32 %ffs
+; CHECK: @b
+; CHECK-NEXT: ret i32 0
+}
diff --git a/test/Transforms/SimplifyLibCalls/StrCpy.ll b/test/Transforms/SimplifyLibCalls/StrCpy.ll
deleted file mode 100644
index 83406ff8f86..00000000000
--- a/test/Transforms/SimplifyLibCalls/StrCpy.ll
+++ /dev/null
@@ -1,37 +0,0 @@
-; Test that the StrCpyOptimizer works correctly
-; RUN: opt < %s -simplify-libcalls -S | FileCheck %s
-
-; This transformation requires the pointer size, as it assumes that size_t is
-; the size of a pointer.
-target datalayout = "e-p:32:32:32-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:32:64-f32:32:32-f64:32:64-v64:64:64-v128:128:128-a0:0:64-f80:128:128-n8:16:32"
-
-@hello = constant [6 x i8] c"hello\00"
-
-declare i8* @strcpy(i8*, i8*)
-
-declare i8* @__strcpy_chk(i8*, i8*, i32) nounwind
-
-declare i32 @llvm.objectsize.i32(i8*, i1) nounwind readonly
-
-; rdar://6839935
-
-define i32 @t1() {
-; CHECK: @t1
-  %target = alloca [1024 x i8]
-  %arg1 = getelementptr [1024 x i8]* %target, i32 0, i32 0
-  %arg2 = getelementptr [6 x i8]* @hello, i32 0, i32 0
-  %rslt1 = call i8* @strcpy( i8* %arg1, i8* %arg2 )
-; CHECK: @llvm.memcpy.p0i8.p0i8.i32
-  ret i32 0
-}
-
-define i32 @t2() {
-; CHECK: @t2
-  %target = alloca [1024 x i8]
-  %arg1 = getelementptr [1024 x i8]* %target, i32 0, i32 0
-  %arg2 = getelementptr [6 x i8]* @hello, i32 0, i32 0
-  %tmp1 = call i32 @llvm.objectsize.i32(i8* %arg1, i1 false)
-  %rslt1 = call i8* @__strcpy_chk(i8* %arg1, i8* %arg2, i32 %tmp1)
-; CHECK: @__memcpy_chk
-  ret i32 0
-}
diff --git a/test/Transforms/TailCallElim/nocapture.ll b/test/Transforms/TailCallElim/nocapture.ll
index 87cb9dd427b..5a1a9a6e7ce 100644
--- a/test/Transforms/TailCallElim/nocapture.ll
+++ b/test/Transforms/TailCallElim/nocapture.ll
@@ -1,9 +1,9 @@
 ; RUN: opt %s -tailcallelim -S | FileCheck %s
-; XFAIL: *
 
 declare void @use(i8* nocapture, i8* nocapture)
+declare void @boring()
 
-define i8* @foo(i8* nocapture %A, i1 %cond) {
+define i8* @test1(i8* nocapture %A, i1 %cond) {
 ; CHECK: tailrecurse:
 ; CHECK: %A.tr = phi i8* [ %A, %0 ], [ %B, %cond_true ]
 ; CHECK: %cond.tr = phi i1 [ %cond, %0 ], [ false, %cond_true ]
@@ -14,12 +14,27 @@ define i8* @foo(i8* nocapture %A, i1 %cond) {
 cond_true:
 ; CHECK: cond_true:
 ; CHECK: br label %tailrecurse
-  call i8* @foo(i8* %B, i1 false)
+  call i8* @test1(i8* %B, i1 false)
   ret i8* null
 cond_false:
 ; CHECK: cond_false
   call void @use(i8* %A, i8* %B)
-; CHECK: tail call void @use(i8* %A.tr, i8* %B)
+; CHECK: call void @use(i8* %A.tr, i8* %B)
+  call void @boring()
+; CHECK: tail call void @boring()
   ret i8* null
 ; CHECK: ret i8* null
 }
+
+; PR14143
+define void @test2(i8* %a, i8* %b) {
+; CHECK: @test2
+; CHECK-NOT: tail call
+; CHECK: ret void
+  %c = alloca [100 x i8], align 16
+  %tmp = bitcast [100 x i8]* %c to i8*
+  call void @llvm.memcpy.p0i8.p0i8.i64(i8* %b, i8* %tmp, i64 100, i32 1, i1 false)
+  ret void
+}
+
+declare void @llvm.memcpy.p0i8.p0i8.i64(i8*, i8*, i64, i32, i1)
diff --git a/test/lit.cfg b/test/lit.cfg
index dc37317ba95..7e6760e95ae 100644
--- a/test/lit.cfg
+++ b/test/lit.cfg
@@ -146,13 +146,12 @@ if re.search(r'cygwin|mingw32|win32', mcjit_triple):
   mcjit_triple += "-elf"
 config.substitutions.append( ('%mcjit_triple', mcjit_triple) )
 
-# When running under valgrind, we mangle '-vg' or '-vg_leak' onto the end of the
-# triple so we can check it with XFAIL and XTARGET.
-config.target_triple += lit.valgrindTriple
-
 # Provide a substition for those tests that need to run the jit to obtain data
 # but simply want use the currently considered most reliable jit for platform
-defaultIsMCJIT='true' if 'arm' in config.target_triple else 'false'
+if 'arm' in config.target_triple:
+    defaultIsMCJIT = 'true'
+else:
+    defaultIsMCJIT = 'false'
 config.substitutions.append( ('%defaultjit', '-use-mcjit='+defaultIsMCJIT) )
 
 # Process jit implementation option
@@ -242,6 +241,10 @@ else:
 if loadable_module:
     config.available_features.add('loadable_module')
 
+# LTO
+if config.lto_is_enabled == "1":
+    config.available_features.add('lto')
+
 # llc knows whether he is compiled with -DNDEBUG.
 import subprocess
 try:
diff --git a/test/lit.site.cfg.in b/test/lit.site.cfg.in
index 178b22f10f3..2bbe63e6348 100644
--- a/test/lit.site.cfg.in
+++ b/test/lit.site.cfg.in
@@ -11,6 +11,7 @@ config.python_executable = "@PYTHON_EXECUTABLE@"
 config.ocamlopt_executable = "@OCAMLOPT@"
 config.enable_shared = @ENABLE_SHARED@
 config.enable_assertions = @ENABLE_ASSERTIONS@
+config.lto_is_enabled = "@LTO_IS_ENABLED@"
 config.targets_to_build = "@TARGETS_TO_BUILD@"
 config.llvm_bindings = "@LLVM_BINDINGS@"
 config.host_os = "@HOST_OS@"
author	tstellar <tstellar@91177308-0d34-0410-b5e6-96231b3b80d8>	2012-10-22 14:55:30 +0000
committer	tstellar <tstellar@91177308-0d34-0410-b5e6-96231b3b80d8>	2012-10-22 14:55:30 +0000
commit	81d3998812b5977a220d62076cc446822747af21 (patch)
tree	b073cb1be9c65dbdd6cd312c322efd3c04abb0af /test
parent	63a079bf76d139b91e170557cf16b17bdbecd58d (diff)