Get rid of storelv4si builtin as it can be expressed directly. This is general

goodness because it provides opportunites to cleanup things.  For example,

uint64_t t1(__m128i vA)
{
  uint64_t Alo;
  _mm_storel_epi64((__m128i*)&Alo, vA);
  return Alo;
}

was generating 

	movq	%xmm0, -8(%rbp)
	movq	-8(%rbp), %rax

and now generates

	movd	%xmm0, %rax

rdar://11282581

llvm-svn: 155924

2 files changed, 6 insertions, 1 deletions

diff --git a/clang/test/CodeGen/builtins-x86.c b/clang/test/CodeGen/builtins-x86.c
index acb5554db4d..fcf1512ca16 100644
--- a/clang/test/CodeGen/builtins-x86.c
+++ b/clang/test/CodeGen/builtins-x86.c
@@ -344,7 +344,6 @@ void f0() {
   tmp_V16c = __builtin_ia32_lddqu(tmp_cCp);
   tmp_V2LLi = __builtin_ia32_palignr128(tmp_V2LLi, tmp_V2LLi, imm_i);
   tmp_V1LLi = __builtin_ia32_palignr(tmp_V1LLi, tmp_V1LLi, imm_i);
-  (void) __builtin_ia32_storelv4si(tmp_V2ip, tmp_V2LLi);
 #ifdef USE_SSE4
   tmp_V16c = __builtin_ia32_pblendvb128(tmp_V16c, tmp_V16c, tmp_V16c);
   tmp_V8s = __builtin_ia32_pblendw128(tmp_V8s, tmp_V8s, imm_i_0_256);
diff --git a/clang/test/CodeGen/sse-builtins.c b/clang/test/CodeGen/sse-builtins.c
index 2d574251555..0e48560b086 100644
--- a/clang/test/CodeGen/sse-builtins.c
+++ b/clang/test/CodeGen/sse-builtins.c
@@ -151,3 +151,9 @@ __m128d test_mm_round_sd(__m128d x, __m128d y) {
   // CHECK: @llvm.x86.sse41.round.sd
   return _mm_round_sd(x, y, 2);
 }
+
+void test_storel_epi64(__m128i x, void* y) {
+  // CHECK: define void @test_storel_epi64
+  // CHECK: store {{.*}} i64* {{.*}}, align 1{{$}}
+  _mm_storel_epi64(y, x);
+}