DAGCombiner: Turn divs of vector splats into vectorized multiplications.

[oota-llvm.git] / test / CodeGen / X86 / stride-nine-with-base-reg.ll

diff --git a/test/CodeGen/X86/stride-nine-with-base-reg.ll b/test/CodeGen/X86/stride-nine-with-base-reg.ll
index f443c76015dd5d7fbc2daa277c5267bdaba6998a..ddf059c675df02799dc4bd9f79826b12701306fd 100644 (file)
--- a/test/CodeGen/X86/stride-nine-with-base-reg.ll
+++ b/test/CodeGen/X86/stride-nine-with-base-reg.ll
@@ -1,14 +1,16 @@
-; RUN: llvm-as < %s | llc -march=x86 | grep lea | count 1
-; RUN: llvm-as < %s | llc -march=x86-64 | not grep lea
+; RUN: llc < %s -march=x86 -relocation-model=static | FileCheck %s
+; RUN: llc < %s -mtriple=x86_64-linux               | FileCheck %s
+; CHECK-NOT:     lea
 
-; For x86 there's an lea above the loop. In both cases, there shouldn't
-; be any lea instructions inside the loop.
+; P should be sunk into the loop and folded into the address mode. There
+; shouldn't be any lea instructions inside the loop.
 
 @B = external global [1000 x i8], align 32
 @A = external global [1000 x i8], align 32
 @P = external global [1000 x i8], align 32
+@Q = external global [1000 x i8], align 32
 
-define void @foo(i32 %m, i32 %p) {
+define void @foo(i32 %m, i32 %p) nounwind {
 entry:
        %tmp1 = icmp sgt i32 %m, 0
        br i1 %tmp1, label %bb, label %return
@@ -24,6 +26,8 @@ bb:
         %tmp0 = add i32 %tmp8, %p
        %tmp10 = getelementptr [1000 x i8]* @P, i32 0, i32 %tmp0
        store i8 17, i8* %tmp10, align 4
+       %tmp11 = getelementptr [1000 x i8]* @Q, i32 0, i32 %tmp0
+       store i8 19, i8* %tmp11, align 4
        %indvar.next = add i32 %i.019.0, 1
        %exitcond = icmp eq i32 %indvar.next, %m
        br i1 %exitcond, label %return, label %bb