1 // RUN: %clang_cc1 -x c -flax-vector-conversions=none -ffreestanding %s -triple=x86_64-apple-darwin -target-feature +sse3 -emit-llvm -o - -Wall -Werror | FileCheck %s 2 // RUN: %clang_cc1 -x c -flax-vector-conversions=none -ffreestanding %s -triple=i386-apple-darwin -target-feature +sse3 -emit-llvm -o - -Wall -Werror | FileCheck %s 3 // RUN: %clang_cc1 -x c++ -flax-vector-conversions=none -ffreestanding %s -triple=x86_64-apple-darwin -target-feature +sse3 -emit-llvm -o - -Wall -Werror | FileCheck %s 4 // RUN: %clang_cc1 -x c++ -flax-vector-conversions=none -ffreestanding %s -triple=i386-apple-darwin -target-feature +sse3 -emit-llvm -o - -Wall -Werror | FileCheck %s 5 6 7 #include <immintrin.h> 8 #include "builtin_test_helpers.h" 9 10 // NOTE: This should match the tests in llvm/test/CodeGen/X86/sse3-intrinsics-fast-isel.ll 11 12 __m128d test_mm_addsub_pd(__m128d A, __m128d B) { 13 // CHECK-LABEL: test_mm_addsub_pd 14 // CHECK: call {{.*}}<2 x double> @llvm.x86.sse3.addsub.pd(<2 x double> %{{.*}}, <2 x double> %{{.*}}) 15 return _mm_addsub_pd(A, B); 16 } 17 18 __m128 test_mm_addsub_ps(__m128 A, __m128 B) { 19 // CHECK-LABEL: test_mm_addsub_ps 20 // CHECK: call {{.*}}<4 x float> @llvm.x86.sse3.addsub.ps(<4 x float> %{{.*}}, <4 x float> %{{.*}}) 21 return _mm_addsub_ps(A, B); 22 } 23 24 __m128d test_mm_hadd_pd(__m128d A, __m128d B) { 25 // CHECK-LABEL: test_mm_hadd_pd 26 // CHECK: call {{.*}}<2 x double> @llvm.x86.sse3.hadd.pd(<2 x double> %{{.*}}, <2 x double> %{{.*}}) 27 return _mm_hadd_pd(A, B); 28 } 29 30 __m128 test_mm_hadd_ps(__m128 A, __m128 B) { 31 // CHECK-LABEL: test_mm_hadd_ps 32 // CHECK: call {{.*}}<4 x float> @llvm.x86.sse3.hadd.ps(<4 x float> %{{.*}}, <4 x float> %{{.*}}) 33 return _mm_hadd_ps(A, B); 34 } 35 36 __m128d test_mm_hsub_pd(__m128d A, __m128d B) { 37 // CHECK-LABEL: test_mm_hsub_pd 38 // CHECK: call {{.*}}<2 x double> @llvm.x86.sse3.hsub.pd(<2 x double> %{{.*}}, <2 x double> %{{.*}}) 39 return _mm_hsub_pd(A, B); 40 } 41 42 __m128 test_mm_hsub_ps(__m128 A, __m128 B) { 43 // CHECK-LABEL: test_mm_hsub_ps 44 // CHECK: call {{.*}}<4 x float> @llvm.x86.sse3.hsub.ps(<4 x float> %{{.*}}, <4 x float> %{{.*}}) 45 return _mm_hsub_ps(A, B); 46 } 47 48 __m128i test_mm_lddqu_si128(__m128i const* P) { 49 // CHECK-LABEL: test_mm_lddqu_si128 50 // CHECK: call <16 x i8> @llvm.x86.sse3.ldu.dq(ptr %{{.*}}) 51 return _mm_lddqu_si128(P); 52 } 53 54 __m128d test_mm_loaddup_pd(double const* P) { 55 // CHECK-LABEL: test_mm_loaddup_pd 56 // CHECK: load ptr 57 // CHECK: insertelement <2 x double> poison, double %{{.*}}, i32 0 58 // CHECK: insertelement <2 x double> %{{.*}}, double %{{.*}}, i32 1 59 return _mm_loaddup_pd(P); 60 } 61 62 __m128d test_mm_movedup_pd(__m128d A) { 63 // CHECK-LABEL: test_mm_movedup_pd 64 // CHECK: shufflevector <2 x double> %{{.*}}, <2 x double> %{{.*}}, <2 x i32> zeroinitializer 65 return _mm_movedup_pd(A); 66 } 67 TEST_CONSTEXPR(match_m128d(_mm_movedup_pd((__m128d){+7.0, -7.0}), +7.0, +7.0)); 68 69 __m128 test_mm_movehdup_ps(__m128 A) { 70 // CHECK-LABEL: test_mm_movehdup_ps 71 // CHECK: shufflevector <4 x float> %{{.*}}, <4 x float> %{{.*}}, <4 x i32> <i32 1, i32 1, i32 3, i32 3> 72 return _mm_movehdup_ps(A); 73 } 74 TEST_CONSTEXPR(match_m128(_mm_movehdup_ps((__m128){+1.0f,-1.0f,+2.0f,+4.0f}), -1.0f, -1.0f, +4.0f, +4.0f)); 75 76 __m128 test_mm_moveldup_ps(__m128 A) { 77 // CHECK-LABEL: test_mm_moveldup_ps 78 // CHECK: shufflevector <4 x float> %{{.*}}, <4 x float> %{{.*}}, <4 x i32> <i32 0, i32 0, i32 2, i32 2> 79 return _mm_moveldup_ps(A); 80 } 81 TEST_CONSTEXPR(match_m128(_mm_moveldup_ps((__m128){+1.0f,-1.0f,+2.0f,+4.0f}), +1.0f, +1.0f, +2.0f, +2.0f)); 82