1 // RUN: %clang_cc1 %s -triple=x86_64-apple-darwin -target-feature +sse4a -emit-llvm -o - -Werror | FileCheck %s 2 3 // Don't include mm_malloc.h, it's system specific. 4 #define __MM_MALLOC_H 5 6 #include <x86intrin.h> 7 8 // NOTE: This should match the tests in llvm/test/CodeGen/X86/sse4a-intrinsics-fast-isel.ll 9 10 __m128i test_mm_extracti_si64(__m128i x) { 11 // CHECK-LABEL: test_mm_extracti_si64 12 // CHECK: call <2 x i64> @llvm.x86.sse4a.extrqi(<2 x i64> %{{[^,]+}}, i8 3, i8 2) 13 return _mm_extracti_si64(x, 3, 2); 14 } 15 16 __m128i test_mm_extract_si64(__m128i x, __m128i y) { 17 // CHECK-LABEL: test_mm_extract_si64 18 // CHECK: call <2 x i64> @llvm.x86.sse4a.extrq(<2 x i64> %{{[^,]+}}, <16 x i8> %{{[^,]+}}) 19 return _mm_extract_si64(x, y); 20 } 21 22 __m128i test_mm_inserti_si64(__m128i x, __m128i y) { 23 // CHECK-LABEL: test_mm_inserti_si64 24 // CHECK: call <2 x i64> @llvm.x86.sse4a.insertqi(<2 x i64> %{{[^,]+}}, <2 x i64> %{{[^,]+}}, i8 5, i8 6) 25 return _mm_inserti_si64(x, y, 5, 6); 26 } 27 28 __m128i test_mm_insert_si64(__m128i x, __m128i y) { 29 // CHECK-LABEL: test_mm_insert_si64 30 // CHECK: call <2 x i64> @llvm.x86.sse4a.insertq(<2 x i64> %{{[^,]+}}, <2 x i64> %{{[^,]+}}) 31 return _mm_insert_si64(x, y); 32 } 33 34 void test_mm_stream_sd(double *p, __m128d a) { 35 // CHECK-LABEL: test_mm_stream_sd 36 // CHECK: extractelement <2 x double> %{{.*}}, i64 0 37 // CHECK: store double %{{.*}}, double* %{{.*}}, align 1, !nontemporal 38 _mm_stream_sd(p, a); 39 } 40 41 void test_mm_stream_ss(float *p, __m128 a) { 42 // CHECK-LABEL: test_mm_stream_ss 43 // CHECK: extractelement <4 x float> %{{.*}}, i64 0 44 // CHECK: store float %{{.*}}, float* %{{.*}}, align 1, !nontemporal 45 _mm_stream_ss(p, a); 46 } 47
