2 files changed, 48 insertions, 0 deletions
diff --git a/clang/test/CodeGen/X86/avx512vlbw-builtins.c b/clang/test/CodeGen/X86/avx512vlbw-builtins.c
index 116d86f..febef46 100644
--- a/clang/test/CodeGen/X86/avx512vlbw-builtins.c
+++ b/clang/test/CodeGen/X86/avx512vlbw-builtins.c
@@ -645,6 +645,21 @@ __mmask16 test_mm_cmp_epi8_mask(__m128i __a, __m128i __b) {
   return (__mmask16)_mm_cmp_epi8_mask(__a, __b, 0);
 }
 
+TEST_CONSTEXPR(_mm_cmpeq_epi8_mask(
+    ((__m128i)(__v16qi){5, 3, 7, 2, 9, 3, 7, 1, 5, 4, 8, 2, 9, 6, 7, 5}),
+    ((__m128i)(__v16qi){5, 2, 7, 3, 9, 4, 6, 1, 5, 3, 8, 1, 9, 5, 7, 5})
+) == (__mmask16)0xd595);
+
+TEST_CONSTEXPR(_mm_cmplt_epi8_mask(
+    ((__m128i)(__v16qi){1, 5, 3, 7, 2, 8, 4, 6, 9, 5, 3, 11, 2, 6, 15, 8}),
+    ((__m128i)(__v16qi){2, 4, 6, 8, 3, 5, 7, 9, 4, 6, 8, 10, 5, 7, 9, 11})
+) == (__mmask16)0xb6dd);
+
+TEST_CONSTEXPR(_mm_cmple_epi8_mask(
+    ((__m128i)(__v16qi){1, 3, 5, 7, 2, 6, 6, 8, 1, 3, 9, 7, 2, 4, 6, 10}),
+    ((__m128i)(__v16qi){2, 3, 4, 7, 3, 4, 5, 8, 2, 3, 4, 7, 3, 4, 5, 8})
+) == (__mmask16)0x3b9b);
+
 __mmask16 test_mm_mask_cmp_epi8_mask(__mmask16 __u, __m128i __a, __m128i __b) {
   // CHECK-LABEL: test_mm_mask_cmp_epi8_mask
   // CHECK: icmp eq <16 x i8> %{{.*}}, %{{.*}}
@@ -2894,6 +2909,12 @@ __mmask16 test_mm_test_epi8_mask(__m128i __A, __m128i __B) {
   return _mm_test_epi8_mask(__A, __B); 
 }
 
+TEST_CONSTEXPR(_mm_test_epi8_mask(
+    (__m128i)(__v16qi){1, 2, 2, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16},
+    (__m128i)(__v16qi){1, 2, 4, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16}
+)
+== (__mmask16)0xfffb);
+
 __mmask16 test_mm_mask_test_epi8_mask(__mmask16 __U, __m128i __A, __m128i __B) {
   // CHECK-LABEL: test_mm_mask_test_epi8_mask
   // CHECK: and <2 x i64> %{{.*}}, %{{.*}}
@@ -2901,6 +2922,12 @@ __mmask16 test_mm_mask_test_epi8_mask(__mmask16 __U, __m128i __A, __m128i __B) {
   // CHECK: and <16 x i1> %{{.*}}, %{{.*}}
   return _mm_mask_test_epi8_mask(__U, __A, __B); 
 }
+TEST_CONSTEXPR(_mm_mask_test_epi8_mask(
+    0xFFFF,
+    (__m128i)(__v16qi){1, 2, 2, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16},
+    (__m128i)(__v16qi){1, 2, 4, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16}
+)
+== (__mmask16)0xfffb);
 
 __mmask32 test_mm256_test_epi8_mask(__m256i __A, __m256i __B) {
   // CHECK-LABEL: test_mm256_test_epi8_mask
@@ -2908,6 +2935,11 @@ __mmask32 test_mm256_test_epi8_mask(__m256i __A, __m256i __B) {
   // CHECK: icmp ne <32 x i8> %{{.*}}, %{{.*}}
   return _mm256_test_epi8_mask(__A, __B); 
 }
+TEST_CONSTEXPR(_mm256_test_epi8_mask(
+    (__m256i)(__v32qi){1, 2, 2, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 1, 2, 2, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16},
+    (__m256i)(__v32qi){1, 2, 4, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 1, 2, 4, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16}
+)
+== (__mmask32)0xfffbfffb);
 
 __mmask32 test_mm256_mask_test_epi8_mask(__mmask32 __U, __m256i __A, __m256i __B) {
   // CHECK-LABEL: test_mm256_mask_test_epi8_mask
@@ -2954,6 +2986,12 @@ __mmask16 test_mm_testn_epi8_mask(__m128i __A, __m128i __B) {
   return _mm_testn_epi8_mask(__A, __B); 
 }
 
+TEST_CONSTEXPR(_mm_testn_epi8_mask(
+    (__m128i)(__v16qi){1, 2, 77, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 1, 16, 16},
+    (__m128i)(__v16qi){2, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 15}
+)
+== (__mmask16)0xe001);
+
 __mmask16 test_mm_mask_testn_epi8_mask(__mmask16 __U, __m128i __A, __m128i __B) {
   // CHECK-LABEL: test_mm_mask_testn_epi8_mask
   // CHECK: and <2 x i64> %{{.*}}, %{{.*}}
diff --git a/clang/test/CodeGen/X86/sse41-builtins.c b/clang/test/CodeGen/X86/sse41-builtins.c
index 62cd392..35fa65a 100644
--- a/clang/test/CodeGen/X86/sse41-builtins.c
+++ b/clang/test/CodeGen/X86/sse41-builtins.c
@@ -307,6 +307,16 @@ __m128 test_mm_insert_ps(__m128 x, __m128 y) {
   return _mm_insert_ps(x, y, 4);
 }
 
+TEST_CONSTEXPR((match_m128(_mm_insert_ps(((__m128)(__v4sf){1.0f, 2.0f, 3.0f, 4.0f}), ((__m128)(__v4sf){10.0f, 20.0f, 30.0f, 40.0f}), 0x10), 1.0f, 10.0f, 3.0f, 4.0f))); // Insert Y[0] into X[1]
+TEST_CONSTEXPR((match_m128(_mm_insert_ps(((__m128)(__v4sf){1.0f, 2.0f, 3.0f, 4.0f}), ((__m128)(__v4sf){10.0f, 20.0f, 30.0f, 40.0f}), 0x00), 10.0f, 2.0f, 3.0f, 4.0f))); // Insert Y[0] into X[0]
+TEST_CONSTEXPR((match_m128(_mm_insert_ps(((__m128)(__v4sf){1.0f, 2.0f, 3.0f, 4.0f}), ((__m128)(__v4sf){10.0f, 20.0f, 30.0f, 40.0f}), 0x20), 1.0f, 2.0f, 10.0f, 4.0f))); // Insert Y[0] into X[2]
+TEST_CONSTEXPR((match_m128(_mm_insert_ps(((__m128)(__v4sf){1.0f, 2.0f, 3.0f, 4.0f}), ((__m128)(__v4sf){10.0f, 20.0f, 30.0f, 40.0f}), 0x30), 1.0f, 2.0f, 3.0f, 10.0f))); // Insert Y[0] into X[3]
+TEST_CONSTEXPR((match_m128(_mm_insert_ps(((__m128)(__v4sf){1.0f, 2.0f, 3.0f, 4.0f}), ((__m128)(__v4sf){10.0f, 20.0f, 30.0f, 40.0f}), 0x80), 30.0f, 2.0f, 3.0f, 4.0f))); // Insert Y[2] into X[0]
+TEST_CONSTEXPR((match_m128(_mm_insert_ps(((__m128)(__v4sf){1.0f, 2.0f, 3.0f, 4.0f}), ((__m128)(__v4sf){10.0f, 20.0f, 30.0f, 40.0f}), 0x01), 0.0f, 2.0f, 3.0f, 4.0f))); // Insert Y[0] into X[0], zero X[0]
+TEST_CONSTEXPR((match_m128(_mm_insert_ps(((__m128)(__v4sf){1.0f, 2.0f, 3.0f, 4.0f}), ((__m128)(__v4sf){10.0f, 20.0f, 30.0f, 40.0f}), 0x0A), 10.0f, 0.0f, 3.0f, 0.0f))); // Insert Y[0] into X[0], zero X[1] and X[3]
+TEST_CONSTEXPR((match_m128(_mm_insert_ps(((__m128)(__v4sf){1.0f, 2.0f, 3.0f, 4.0f}), ((__m128)(__v4sf){10.0f, 20.0f, 30.0f, 40.0f}), 0x0F), 0.0f, 0.0f, 0.0f, 0.0f))); // Insert Y[0] into X[0], zero all
+TEST_CONSTEXPR((match_m128(_mm_insert_ps(((__m128)(__v4sf){1.0f, 2.0f, 3.0f, 4.0f}), ((__m128)(__v4sf){10.0f, 20.0f, 30.0f, 40.0f}), 0xCF), 0.0f, 0.0f, 0.0f, 0.0f))); // Insert Y[3] into X[0], zero all
+
 __m128i test_mm_max_epi8(__m128i x, __m128i y) {
   // CHECK-LABEL: test_mm_max_epi8
   // CHECK: call <16 x i8> @llvm.smax.v16i8(<16 x i8> %{{.*}}, <16 x i8> %{{.*}})