diff options
| author | Nadav Rotem <nadav.rotem@intel.com> | 2012-04-07 21:19:08 +0000 |
|---|---|---|
| committer | Nadav Rotem <nadav.rotem@intel.com> | 2012-04-07 21:19:08 +0000 |
| commit | 71d07ae5cb12ba21cff64e2ce610bebb9cd65f18 (patch) | |
| tree | b83876bb7571b4ca2b42de0bcd612e63022d0aed /llvm/test/CodeGen/X86/vec_shuffle-37.ll | |
| parent | 5f8397a9344aafd3591f03d1dbb93688d1fee728 (diff) | |
| download | bcm5719-llvm-71d07ae5cb12ba21cff64e2ce610bebb9cd65f18.tar.gz bcm5719-llvm-71d07ae5cb12ba21cff64e2ce610bebb9cd65f18.zip | |
1. Remove the part of r153848 which optimizes shuffle-of-shuffle into a new
shuffle node because it could introduce new shuffle nodes that were not
supported efficiently by the target.
2. Add a more restrictive shuffle-of-shuffle optimization for cases where the
second shuffle reverses the transformation of the first shuffle.
llvm-svn: 154266
Diffstat (limited to 'llvm/test/CodeGen/X86/vec_shuffle-37.ll')
| -rw-r--r-- | llvm/test/CodeGen/X86/vec_shuffle-37.ll | 10 |
1 files changed, 5 insertions, 5 deletions
diff --git a/llvm/test/CodeGen/X86/vec_shuffle-37.ll b/llvm/test/CodeGen/X86/vec_shuffle-37.ll index 619652aff15..430aa046afa 100644 --- a/llvm/test/CodeGen/X86/vec_shuffle-37.ll +++ b/llvm/test/CodeGen/X86/vec_shuffle-37.ll @@ -27,11 +27,11 @@ entry: define void @t02(<8 x i32>* %source, <2 x i32>* %dest) nounwind noinline { entry: ; CHECK: t02 -; CHECK: mov -; CHECK-NEXT: mov -; CHECK-NEXT: mov -; CHECK-NEXT: mov -; CHECK-NEXT: ret +; CHECK: movaps +; CHECK: shufps +; CHECK: pshufd +; CHECK: movq +; CHECK: ret %0 = bitcast <8 x i32>* %source to <4 x i32>* %arrayidx = getelementptr inbounds <4 x i32>* %0, i64 3 %tmp2 = load <4 x i32>* %arrayidx, align 16 |
