Remove vpermilp* builtins. Same effect can be achieved with builtin_shufflevector.

llvm-svn: 150056

2 files changed, 24 insertions, 4 deletions

diff --git a/clang/test/CodeGen/avx-shuffle-builtins.c b/clang/test/CodeGen/avx-shuffle-builtins.c
index c11780a5e39..6b2b1b1b186 100644
--- a/clang/test/CodeGen/avx-shuffle-builtins.c
+++ b/clang/test/CodeGen/avx-shuffle-builtins.c
@@ -14,3 +14,27 @@ __m256 x(__m256 a, __m256 b) {
   // CHECK: shufflevector{{.*}}<i32 3, i32 2, i32 8, i32 11, i32 7, i32 6, i32 12, i32 15>
   return _mm256_shuffle_ps(a, b, 203);
 }
+
+__m128d test_mm_permute_pd(__m128d a) {
+  // Check if the mask is correct
+  // CHECK: shufflevector{{.*}}<i32 1, i32 0>
+  return _mm_permute_pd(a, 1);
+}
+
+__m256d test_mm256_permute_pd(__m256d a) {
+  // Check if the mask is correct
+  // CHECK: shufflevector{{.*}}<i32 1, i32 0, i32 3, i32 2>
+  return _mm256_permute_pd(a, 5);
+}
+
+__m128 test_mm_permute_ps(__m128 a) {
+  // Check if the mask is correct
+  // CHECK: shufflevector{{.*}}<i32 3, i32 2, i32 1, i32 0>
+  return _mm_permute_ps(a, 0x1b);
+}
+
+__m256 test_mm256_permute_ps(__m256 a) {
+  // Check if the mask is correct
+  // CHECK: shufflevector{{.*}}<i32 3, i32 2, i32 1, i32 0, i32 7, i32 6, i32 5, i32 4>
+  return _mm256_permute_ps(a, 0x1b);
+}
diff --git a/clang/test/CodeGen/builtins-x86.c b/clang/test/CodeGen/builtins-x86.c
index a8121526b69..acb5554db4d 100644
--- a/clang/test/CodeGen/builtins-x86.c
+++ b/clang/test/CodeGen/builtins-x86.c
@@ -417,10 +417,6 @@ void f0() {
   tmp_V4d = __builtin_ia32_vperm2f128_pd256(tmp_V4d, tmp_V4d, 0x7);
   tmp_V8f = __builtin_ia32_vperm2f128_ps256(tmp_V8f, tmp_V8f, 0x7);
   tmp_V8i = __builtin_ia32_vperm2f128_si256(tmp_V8i, tmp_V8i, 0x7);
-  tmp_V2d = __builtin_ia32_vpermilpd(tmp_V2d, 0x7);
-  tmp_V4f = __builtin_ia32_vpermilps(tmp_V4f, 0x7);
-  tmp_V4d = __builtin_ia32_vpermilpd256(tmp_V4d, 0x7);
-  tmp_V8f = __builtin_ia32_vpermilps256(tmp_V8f, 0x7);
   tmp_V4d = __builtin_ia32_vinsertf128_pd256(tmp_V4d, tmp_V2d, 0x7);
   tmp_V8f = __builtin_ia32_vinsertf128_ps256(tmp_V8f, tmp_V4f, 0x7);
   tmp_V8i = __builtin_ia32_vinsertf128_si256(tmp_V8i, tmp_V4i, 0x7);