|
1 | | -// RUN: %clang_cc1 -ffreestanding %s -O0 -triple=x86_64-apple-darwin -target-cpu skylake-avx512 -emit-llvm -o - -Wall -Werror | FileCheck %s |
| 1 | +// RUN: %clang_cc1 -x c -ffreestanding %s -O0 -triple=x86_64-apple-darwin -target-cpu skylake-avx512 -emit-llvm -o - -Wall -Werror | FileCheck %s |
| 2 | +// RUN: %clang_cc1 -x c++ -ffreestanding %s -O0 -triple=x86_64-apple-darwin -target-cpu skylake-avx512 -emit-llvm -o - -Wall -Werror | FileCheck %s |
2 | 3 |
|
3 | 4 | #include <immintrin.h> |
4 | 5 | #include "builtin_test_helpers.h" |
5 | 6 |
|
6 | 7 | long long test_mm512_reduce_add_epi64(__m512i __W){ |
7 | | -// CHECK-LABEL: @test_mm512_reduce_add_epi64( |
8 | | -// CHECK: call i64 @llvm.vector.reduce.add.v8i64(<8 x i64> %{{.*}}) |
| 8 | +// CHECK-LABEL: test_mm512_reduce_add_epi64 |
| 9 | +// CHECK: call {{.*}}i64 @llvm.vector.reduce.add.v8i64(<8 x i64> %{{.*}}) |
9 | 10 | return _mm512_reduce_add_epi64(__W); |
10 | 11 | } |
11 | 12 | TEST_CONSTEXPR(_mm512_reduce_add_epi64((__m512i)(__v8di){-4, -3, -2, -1, 0, 1, 2, 3}) == -4); |
12 | 13 |
|
13 | 14 | long long test_mm512_reduce_mul_epi64(__m512i __W){ |
14 | | -// CHECK-LABEL: @test_mm512_reduce_mul_epi64( |
15 | | -// CHECK: call i64 @llvm.vector.reduce.mul.v8i64(<8 x i64> %{{.*}}) |
| 15 | +// CHECK-LABEL: test_mm512_reduce_mul_epi64 |
| 16 | +// CHECK: call {{.*}}i64 @llvm.vector.reduce.mul.v8i64(<8 x i64> %{{.*}}) |
16 | 17 | return _mm512_reduce_mul_epi64(__W); |
17 | 18 | } |
18 | 19 | TEST_CONSTEXPR(_mm512_reduce_mul_epi64((__m512i)(__v8di){1, 2, 3, 4, 5, 6, 7, 8}) == 40320); |
19 | 20 |
|
20 | 21 | long long test_mm512_reduce_or_epi64(__m512i __W){ |
21 | | -// CHECK-LABEL: @test_mm512_reduce_or_epi64( |
22 | | -// CHECK: call i64 @llvm.vector.reduce.or.v8i64(<8 x i64> %{{.*}}) |
| 22 | +// CHECK-LABEL: test_mm512_reduce_or_epi64 |
| 23 | +// CHECK: call {{.*}}i64 @llvm.vector.reduce.or.v8i64(<8 x i64> %{{.*}}) |
23 | 24 | return _mm512_reduce_or_epi64(__W); |
24 | 25 | } |
25 | 26 | TEST_CONSTEXPR(_mm512_reduce_or_epi64((__m512i)(__v8di){0x100, 0x200, 0x400, 0x800, 0, 0, 0, 0}) == 0xF00); |
26 | 27 |
|
27 | 28 | long long test_mm512_reduce_and_epi64(__m512i __W){ |
28 | | -// CHECK-LABEL: @test_mm512_reduce_and_epi64( |
29 | | -// CHECK: call i64 @llvm.vector.reduce.and.v8i64(<8 x i64> %{{.*}}) |
| 29 | +// CHECK-LABEL: test_mm512_reduce_and_epi64 |
| 30 | +// CHECK: call {{.*}}i64 @llvm.vector.reduce.and.v8i64(<8 x i64> %{{.*}}) |
30 | 31 | return _mm512_reduce_and_epi64(__W); |
31 | 32 | } |
32 | 33 | TEST_CONSTEXPR(_mm512_reduce_and_epi64((__m512i)(__v8di){0xFFFF, 0xFF00, 0x00FF, 0xFFFF, 0xFFFF, 0xFFFF, 0xFF00, 0x00FF}) == 0x0000); |
33 | 34 |
|
34 | 35 | long long test_mm512_mask_reduce_add_epi64(__mmask8 __M, __m512i __W){ |
35 | | -// CHECK-LABEL: @test_mm512_mask_reduce_add_epi64( |
| 36 | +// CHECK-LABEL: test_mm512_mask_reduce_add_epi64 |
36 | 37 | // CHECK: select <8 x i1> %{{.*}}, <8 x i64> %{{.*}}, <8 x i64> %{{.*}} |
37 | | -// CHECK: call i64 @llvm.vector.reduce.add.v8i64(<8 x i64> %{{.*}}) |
| 38 | +// CHECK: call {{.*}}i64 @llvm.vector.reduce.add.v8i64(<8 x i64> %{{.*}}) |
38 | 39 | return _mm512_mask_reduce_add_epi64(__M, __W); |
39 | 40 | } |
40 | 41 |
|
41 | 42 | long long test_mm512_mask_reduce_mul_epi64(__mmask8 __M, __m512i __W){ |
42 | | -// CHECK-LABEL: @test_mm512_mask_reduce_mul_epi64( |
| 43 | +// CHECK-LABEL: test_mm512_mask_reduce_mul_epi64 |
43 | 44 | // CHECK: select <8 x i1> %{{.*}}, <8 x i64> %{{.*}}, <8 x i64> %{{.*}} |
44 | | -// CHECK: call i64 @llvm.vector.reduce.mul.v8i64(<8 x i64> %{{.*}}) |
| 45 | +// CHECK: call {{.*}}i64 @llvm.vector.reduce.mul.v8i64(<8 x i64> %{{.*}}) |
45 | 46 | return _mm512_mask_reduce_mul_epi64(__M, __W); |
46 | 47 | } |
47 | 48 |
|
48 | 49 | long long test_mm512_mask_reduce_and_epi64(__mmask8 __M, __m512i __W){ |
49 | | -// CHECK-LABEL: @test_mm512_mask_reduce_and_epi64( |
| 50 | +// CHECK-LABEL: test_mm512_mask_reduce_and_epi64 |
50 | 51 | // CHECK: select <8 x i1> %{{.*}}, <8 x i64> %{{.*}}, <8 x i64> %{{.*}} |
51 | | -// CHECK: call i64 @llvm.vector.reduce.and.v8i64(<8 x i64> %{{.*}}) |
| 52 | +// CHECK: call {{.*}}i64 @llvm.vector.reduce.and.v8i64(<8 x i64> %{{.*}}) |
52 | 53 | return _mm512_mask_reduce_and_epi64(__M, __W); |
53 | 54 | } |
54 | 55 |
|
55 | 56 | long long test_mm512_mask_reduce_or_epi64(__mmask8 __M, __m512i __W){ |
56 | | -// CHECK-LABEL: @test_mm512_mask_reduce_or_epi64( |
| 57 | +// CHECK-LABEL: test_mm512_mask_reduce_or_epi64 |
57 | 58 | // CHECK: select <8 x i1> %{{.*}}, <8 x i64> %{{.*}}, <8 x i64> %{{.*}} |
58 | | -// CHECK: call i64 @llvm.vector.reduce.or.v8i64(<8 x i64> %{{.*}}) |
| 59 | +// CHECK: call {{.*}}i64 @llvm.vector.reduce.or.v8i64(<8 x i64> %{{.*}}) |
59 | 60 | return _mm512_mask_reduce_or_epi64(__M, __W); |
60 | 61 | } |
61 | 62 |
|
62 | 63 | int test_mm512_reduce_add_epi32(__m512i __W){ |
63 | | -// CHECK-LABEL: @test_mm512_reduce_add_epi32( |
64 | | -// CHECK: call i32 @llvm.vector.reduce.add.v16i32(<16 x i32> %{{.*}}) |
| 64 | +// CHECK-LABEL: test_mm512_reduce_add_epi32 |
| 65 | +// CHECK: call {{.*}}i32 @llvm.vector.reduce.add.v16i32(<16 x i32> %{{.*}}) |
65 | 66 | return _mm512_reduce_add_epi32(__W); |
66 | 67 | } |
67 | 68 | TEST_CONSTEXPR(_mm512_reduce_add_epi32((__m512i)(__v16si){-8, -7, -6, -5, -4, -3, -2, -1, 0, 1, 2, 3, 4, 5, 6, 7}) == -8); |
68 | 69 |
|
69 | 70 | int test_mm512_reduce_mul_epi32(__m512i __W){ |
70 | | -// CHECK-LABEL: @test_mm512_reduce_mul_epi32( |
71 | | -// CHECK: call i32 @llvm.vector.reduce.mul.v16i32(<16 x i32> %{{.*}}) |
| 71 | +// CHECK-LABEL: test_mm512_reduce_mul_epi32 |
| 72 | +// CHECK: call {{.*}}i32 @llvm.vector.reduce.mul.v16i32(<16 x i32> %{{.*}}) |
72 | 73 | return _mm512_reduce_mul_epi32(__W); |
73 | 74 | } |
74 | 75 | TEST_CONSTEXPR(_mm512_reduce_mul_epi32((__m512i)(__v16si){1, 1, 1, 1, 1, 1, 1, 1, 2, 2, 3, 1, 1, -3, 1, 1}) == -36); |
75 | 76 |
|
76 | 77 | int test_mm512_reduce_or_epi32(__m512i __W){ |
77 | | -// CHECK: call i32 @llvm.vector.reduce.or.v16i32(<16 x i32> %{{.*}}) |
| 78 | +// CHECK: call {{.*}}i32 @llvm.vector.reduce.or.v16i32(<16 x i32> %{{.*}}) |
78 | 79 | return _mm512_reduce_or_epi32(__W); |
79 | 80 | } |
80 | 81 | TEST_CONSTEXPR(_mm512_reduce_or_epi32((__m512i)(__v16si){0x01, 0x02, 0x04, 0x08, 0x10, 0x20, 0x40, 0x80, 0, 0, 0, 0, 0, 0, 0, 0}) == 0xFF); |
81 | 82 |
|
82 | 83 | int test_mm512_reduce_and_epi32(__m512i __W){ |
83 | | -// CHECK-LABEL: @test_mm512_reduce_and_epi32( |
84 | | -// CHECK: call i32 @llvm.vector.reduce.and.v16i32(<16 x i32> %{{.*}}) |
| 84 | +// CHECK-LABEL: test_mm512_reduce_and_epi32 |
| 85 | +// CHECK: call {{.*}}i32 @llvm.vector.reduce.and.v16i32(<16 x i32> %{{.*}}) |
85 | 86 | return _mm512_reduce_and_epi32(__W); |
86 | 87 | } |
87 | 88 | TEST_CONSTEXPR(_mm512_reduce_and_epi32((__m512i)(__v16si){0xFF, 0xF0, 0x0F, 0xFF, 0xFF, 0xFF, 0xF0, 0x0F, 0xFF, 0xFF, 0xFF, 0xFF, 0xF0, 0xF0, 0x0F, 0x0F}) == 0x00); |
88 | 89 |
|
89 | 90 | int test_mm512_mask_reduce_add_epi32(__mmask16 __M, __m512i __W){ |
90 | | -// CHECK-LABEL: @test_mm512_mask_reduce_add_epi32( |
| 91 | +// CHECK-LABEL: test_mm512_mask_reduce_add_epi32 |
91 | 92 | // CHECK: select <16 x i1> %{{.*}}, <16 x i32> %{{.*}}, <16 x i32> %{{.*}} |
92 | | -// CHECK: call i32 @llvm.vector.reduce.add.v16i32(<16 x i32> %{{.*}}) |
| 93 | +// CHECK: call {{.*}}i32 @llvm.vector.reduce.add.v16i32(<16 x i32> %{{.*}}) |
93 | 94 | return _mm512_mask_reduce_add_epi32(__M, __W); |
94 | 95 | } |
95 | 96 |
|
96 | 97 | int test_mm512_mask_reduce_mul_epi32(__mmask16 __M, __m512i __W){ |
97 | | -// CHECK-LABEL: @test_mm512_mask_reduce_mul_epi32( |
| 98 | +// CHECK-LABEL: test_mm512_mask_reduce_mul_epi32 |
98 | 99 | // CHECK: select <16 x i1> %{{.*}}, <16 x i32> %{{.*}}, <16 x i32> %{{.*}} |
99 | | -// CHECK: call i32 @llvm.vector.reduce.mul.v16i32(<16 x i32> %{{.*}}) |
| 100 | +// CHECK: call {{.*}}i32 @llvm.vector.reduce.mul.v16i32(<16 x i32> %{{.*}}) |
100 | 101 | return _mm512_mask_reduce_mul_epi32(__M, __W); |
101 | 102 | } |
102 | 103 |
|
103 | 104 | int test_mm512_mask_reduce_and_epi32(__mmask16 __M, __m512i __W){ |
104 | | -// CHECK-LABEL: @test_mm512_mask_reduce_and_epi32( |
| 105 | +// CHECK-LABEL: test_mm512_mask_reduce_and_epi32 |
105 | 106 | // CHECK: select <16 x i1> %{{.*}}, <16 x i32> %{{.*}}, <16 x i32> %{{.*}} |
106 | | -// CHECK: call i32 @llvm.vector.reduce.and.v16i32(<16 x i32> %{{.*}}) |
| 107 | +// CHECK: call {{.*}}i32 @llvm.vector.reduce.and.v16i32(<16 x i32> %{{.*}}) |
107 | 108 | return _mm512_mask_reduce_and_epi32(__M, __W); |
108 | 109 | } |
109 | 110 |
|
110 | 111 | int test_mm512_mask_reduce_or_epi32(__mmask16 __M, __m512i __W){ |
111 | | -// CHECK-LABEL: @test_mm512_mask_reduce_or_epi32( |
| 112 | +// CHECK-LABEL: test_mm512_mask_reduce_or_epi32 |
112 | 113 | // CHECK: select <16 x i1> %{{.*}}, <16 x i32> %{{.*}}, <16 x i32> %{{.*}} |
113 | | -// CHECK: call i32 @llvm.vector.reduce.or.v16i32(<16 x i32> %{{.*}}) |
| 114 | +// CHECK: call {{.*}}i32 @llvm.vector.reduce.or.v16i32(<16 x i32> %{{.*}}) |
114 | 115 | return _mm512_mask_reduce_or_epi32(__M, __W); |
115 | 116 | } |
116 | 117 |
|
117 | 118 | double test_mm512_reduce_add_pd(__m512d __W, double ExtraAddOp){ |
118 | | -// CHECK-LABEL: @test_mm512_reduce_add_pd( |
| 119 | +// CHECK-LABEL: test_mm512_reduce_add_pd |
119 | 120 | // CHECK-NOT: reassoc |
120 | | -// CHECK: call reassoc double @llvm.vector.reduce.fadd.v8f64(double -0.000000e+00, <8 x double> %{{.*}}) |
| 121 | +// CHECK: call reassoc {{.*}}double @llvm.vector.reduce.fadd.v8f64(double -0.000000e+00, <8 x double> %{{.*}}) |
121 | 122 | // CHECK-NOT: reassoc |
122 | 123 | return _mm512_reduce_add_pd(__W) + ExtraAddOp; |
123 | 124 | } |
124 | 125 |
|
125 | 126 | double test_mm512_reduce_mul_pd(__m512d __W, double ExtraMulOp){ |
126 | | -// CHECK-LABEL: @test_mm512_reduce_mul_pd( |
| 127 | +// CHECK-LABEL: test_mm512_reduce_mul_pd |
127 | 128 | // CHECK-NOT: reassoc |
128 | | -// CHECK: call reassoc double @llvm.vector.reduce.fmul.v8f64(double 1.000000e+00, <8 x double> %{{.*}}) |
| 129 | +// CHECK: call reassoc {{.*}}double @llvm.vector.reduce.fmul.v8f64(double 1.000000e+00, <8 x double> %{{.*}}) |
129 | 130 | // CHECK-NOT: reassoc |
130 | 131 | return _mm512_reduce_mul_pd(__W) * ExtraMulOp; |
131 | 132 | } |
132 | 133 |
|
133 | 134 | float test_mm512_reduce_add_ps(__m512 __W){ |
134 | | -// CHECK-LABEL: @test_mm512_reduce_add_ps( |
135 | | -// CHECK: call reassoc float @llvm.vector.reduce.fadd.v16f32(float -0.000000e+00, <16 x float> %{{.*}}) |
| 135 | +// CHECK-LABEL: test_mm512_reduce_add_ps |
| 136 | +// CHECK: call reassoc {{.*}}float @llvm.vector.reduce.fadd.v16f32(float -0.000000e+00, <16 x float> %{{.*}}) |
136 | 137 | return _mm512_reduce_add_ps(__W); |
137 | 138 | } |
138 | 139 |
|
139 | 140 | float test_mm512_reduce_mul_ps(__m512 __W){ |
140 | | -// CHECK-LABEL: @test_mm512_reduce_mul_ps( |
141 | | -// CHECK: call reassoc float @llvm.vector.reduce.fmul.v16f32(float 1.000000e+00, <16 x float> %{{.*}}) |
| 141 | +// CHECK-LABEL: test_mm512_reduce_mul_ps |
| 142 | +// CHECK: call reassoc {{.*}}float @llvm.vector.reduce.fmul.v16f32(float 1.000000e+00, <16 x float> %{{.*}}) |
142 | 143 | return _mm512_reduce_mul_ps(__W); |
143 | 144 | } |
144 | 145 |
|
145 | 146 | double test_mm512_mask_reduce_add_pd(__mmask8 __M, __m512d __W){ |
146 | | -// CHECK-LABEL: @test_mm512_mask_reduce_add_pd( |
| 147 | +// CHECK-LABEL: test_mm512_mask_reduce_add_pd |
147 | 148 | // CHECK: select <8 x i1> %{{.*}}, <8 x double> %{{.*}}, <8 x double> %{{.*}} |
148 | | -// CHECK: call reassoc double @llvm.vector.reduce.fadd.v8f64(double -0.000000e+00, <8 x double> %{{.*}}) |
| 149 | +// CHECK: call reassoc {{.*}}double @llvm.vector.reduce.fadd.v8f64(double -0.000000e+00, <8 x double> %{{.*}}) |
149 | 150 | return _mm512_mask_reduce_add_pd(__M, __W); |
150 | 151 | } |
151 | 152 |
|
152 | 153 | double test_mm512_mask_reduce_mul_pd(__mmask8 __M, __m512d __W){ |
153 | | -// CHECK-LABEL: @test_mm512_mask_reduce_mul_pd( |
| 154 | +// CHECK-LABEL: test_mm512_mask_reduce_mul_pd |
154 | 155 | // CHECK: select <8 x i1> %{{.*}}, <8 x double> %{{.*}}, <8 x double> %{{.*}} |
155 | | -// CHECK: call reassoc double @llvm.vector.reduce.fmul.v8f64(double 1.000000e+00, <8 x double> %{{.*}}) |
| 156 | +// CHECK: call reassoc {{.*}}double @llvm.vector.reduce.fmul.v8f64(double 1.000000e+00, <8 x double> %{{.*}}) |
156 | 157 | return _mm512_mask_reduce_mul_pd(__M, __W); |
157 | 158 | } |
158 | 159 |
|
159 | 160 | float test_mm512_mask_reduce_add_ps(__mmask16 __M, __m512 __W){ |
160 | | -// CHECK-LABEL: @test_mm512_mask_reduce_add_ps( |
| 161 | +// CHECK-LABEL: test_mm512_mask_reduce_add_ps |
161 | 162 | // CHECK: select <16 x i1> %{{.*}}, <16 x float> {{.*}}, <16 x float> {{.*}} |
162 | | -// CHECK: call reassoc float @llvm.vector.reduce.fadd.v16f32(float -0.000000e+00, <16 x float> %{{.*}}) |
| 163 | +// CHECK: call reassoc {{.*}}float @llvm.vector.reduce.fadd.v16f32(float -0.000000e+00, <16 x float> %{{.*}}) |
163 | 164 | return _mm512_mask_reduce_add_ps(__M, __W); |
164 | 165 | } |
165 | 166 |
|
166 | 167 | float test_mm512_mask_reduce_mul_ps(__mmask16 __M, __m512 __W){ |
167 | | -// CHECK-LABEL: @test_mm512_mask_reduce_mul_ps( |
| 168 | +// CHECK-LABEL: test_mm512_mask_reduce_mul_ps |
168 | 169 | // CHECK: select <16 x i1> %{{.*}}, <16 x float> {{.*}}, <16 x float> %{{.*}} |
169 | | -// CHECK: call reassoc float @llvm.vector.reduce.fmul.v16f32(float 1.000000e+00, <16 x float> %{{.*}}) |
| 170 | +// CHECK: call reassoc {{.*}}float @llvm.vector.reduce.fmul.v16f32(float 1.000000e+00, <16 x float> %{{.*}}) |
170 | 171 | return _mm512_mask_reduce_mul_ps(__M, __W); |
171 | 172 | } |
0 commit comments