1 /* { dg-do compile { target { ! ia32 } } } */
2 /* { dg-options "-O2 -mavx512vl -mavx512dq" } */
3
4 #include <x86intrin.h>
5
6 __m128d
f1(__m128d a,__m128d b)7 f1 (__m128d a, __m128d b)
8 {
9 register __m128d c __asm ("xmm16") = a;
10 asm volatile ("" : "+v" (c));
11 c = _mm_and_pd (c, b);
12 asm volatile ("" : "+v" (c));
13 return c;
14 }
15
16 /* { dg-final { scan-assembler-times "vandpd\[^\n\r\]*xmm\[0-9\]" 1 } } */
17
18 __m128d
f2(__m128d a,__m128d b)19 f2 (__m128d a, __m128d b)
20 {
21 register __m128d c __asm ("xmm16") = a;
22 asm volatile ("" : "+v" (c));
23 c = _mm_or_pd (c, b);
24 asm volatile ("" : "+v" (c));
25 return c;
26 }
27
28 /* { dg-final { scan-assembler-times "vorpd\[^\n\r\]*xmm\[0-9\]" 1 } } */
29
30 __m128d
f3(__m128d a,__m128d b)31 f3 (__m128d a, __m128d b)
32 {
33 register __m128d c __asm ("xmm16") = a;
34 asm volatile ("" : "+v" (c));
35 c = _mm_xor_pd (c, b);
36 asm volatile ("" : "+v" (c));
37 return c;
38 }
39
40 /* { dg-final { scan-assembler-times "vxorpd\[^\n\r\]*xmm\[0-9\]" 1 } } */
41
42 __m128d
f4(__m128d a,__m128d b)43 f4 (__m128d a, __m128d b)
44 {
45 register __m128d c __asm ("xmm16") = a;
46 asm volatile ("" : "+v" (c));
47 c = _mm_andnot_pd (c, b);
48 asm volatile ("" : "+v" (c));
49 return c;
50 }
51
52 /* { dg-final { scan-assembler-times "vandnpd\[^\n\r\]*xmm\[0-9\]" 1 } } */
53
54 __m128
f5(__m128 a,__m128 b)55 f5 (__m128 a, __m128 b)
56 {
57 register __m128 c __asm ("xmm16") = a;
58 asm volatile ("" : "+v" (c));
59 c = _mm_and_ps (c, b);
60 asm volatile ("" : "+v" (c));
61 return c;
62 }
63
64 /* { dg-final { scan-assembler-times "vandps\[^\n\r\]*xmm\[0-9\]" 1 } } */
65
66 __m128
f6(__m128 a,__m128 b)67 f6 (__m128 a, __m128 b)
68 {
69 register __m128 c __asm ("xmm16") = a;
70 asm volatile ("" : "+v" (c));
71 c = _mm_or_ps (c, b);
72 asm volatile ("" : "+v" (c));
73 return c;
74 }
75
76 /* { dg-final { scan-assembler-times "vorps\[^\n\r\]*xmm\[0-9\]" 1 } } */
77
78 __m128
f7(__m128 a,__m128 b)79 f7 (__m128 a, __m128 b)
80 {
81 register __m128 c __asm ("xmm16") = a;
82 asm volatile ("" : "+v" (c));
83 c = _mm_xor_ps (c, b);
84 asm volatile ("" : "+v" (c));
85 return c;
86 }
87
88 /* { dg-final { scan-assembler-times "vxorps\[^\n\r\]*xmm\[0-9\]" 1 } } */
89
90 __m128
f8(__m128 a,__m128 b)91 f8 (__m128 a, __m128 b)
92 {
93 register __m128 c __asm ("xmm16") = a;
94 asm volatile ("" : "+v" (c));
95 c = _mm_andnot_ps (c, b);
96 asm volatile ("" : "+v" (c));
97 return c;
98 }
99
100 /* { dg-final { scan-assembler-times "vandnps\[^\n\r\]*xmm\[0-9\]" 1 } } */
101
102 __m256d
f9(__m256d a,__m256d b)103 f9 (__m256d a, __m256d b)
104 {
105 register __m256d c __asm ("xmm16") = a;
106 asm volatile ("" : "+v" (c));
107 c = _mm256_and_pd (c, b);
108 asm volatile ("" : "+v" (c));
109 return c;
110 }
111
112 /* { dg-final { scan-assembler-times "vandpd\[^\n\r\]*ymm\[0-9\]" 1 } } */
113
114 __m256d
f10(__m256d a,__m256d b)115 f10 (__m256d a, __m256d b)
116 {
117 register __m256d c __asm ("xmm16") = a;
118 asm volatile ("" : "+v" (c));
119 c = _mm256_or_pd (c, b);
120 asm volatile ("" : "+v" (c));
121 return c;
122 }
123
124 /* { dg-final { scan-assembler-times "vorpd\[^\n\r\]*ymm\[0-9\]" 1 } } */
125
126 __m256d
f11(__m256d a,__m256d b)127 f11 (__m256d a, __m256d b)
128 {
129 register __m256d c __asm ("xmm16") = a;
130 asm volatile ("" : "+v" (c));
131 c = _mm256_xor_pd (c, b);
132 asm volatile ("" : "+v" (c));
133 return c;
134 }
135
136 /* { dg-final { scan-assembler-times "vxorpd\[^\n\r\]*ymm\[0-9\]" 1 } } */
137
138 __m256d
f12(__m256d a,__m256d b)139 f12 (__m256d a, __m256d b)
140 {
141 register __m256d c __asm ("xmm16") = a;
142 asm volatile ("" : "+v" (c));
143 c = _mm256_andnot_pd (c, b);
144 asm volatile ("" : "+v" (c));
145 return c;
146 }
147
148 /* { dg-final { scan-assembler-times "vandnpd\[^\n\r\]*ymm\[0-9\]" 1 } } */
149
150 __m256
f13(__m256 a,__m256 b)151 f13 (__m256 a, __m256 b)
152 {
153 register __m256 c __asm ("xmm16") = a;
154 asm volatile ("" : "+v" (c));
155 c = _mm256_and_ps (c, b);
156 asm volatile ("" : "+v" (c));
157 return c;
158 }
159
160 /* { dg-final { scan-assembler-times "vandps\[^\n\r\]*ymm\[0-9\]" 1 } } */
161
162 __m256
f14(__m256 a,__m256 b)163 f14 (__m256 a, __m256 b)
164 {
165 register __m256 c __asm ("xmm16") = a;
166 asm volatile ("" : "+v" (c));
167 c = _mm256_or_ps (c, b);
168 asm volatile ("" : "+v" (c));
169 return c;
170 }
171
172 /* { dg-final { scan-assembler-times "vorps\[^\n\r\]*ymm\[0-9\]" 1 } } */
173
174 __m256
f15(__m256 a,__m256 b)175 f15 (__m256 a, __m256 b)
176 {
177 register __m256 c __asm ("xmm16") = a;
178 asm volatile ("" : "+v" (c));
179 c = _mm256_xor_ps (c, b);
180 asm volatile ("" : "+v" (c));
181 return c;
182 }
183
184 /* { dg-final { scan-assembler-times "vxorps\[^\n\r\]*ymm\[0-9\]" 1 } } */
185
186 __m256
f16(__m256 a,__m256 b)187 f16 (__m256 a, __m256 b)
188 {
189 register __m256 c __asm ("xmm16") = a;
190 asm volatile ("" : "+v" (c));
191 c = _mm256_andnot_ps (c, b);
192 asm volatile ("" : "+v" (c));
193 return c;
194 }
195
196 /* { dg-final { scan-assembler-times "vandnps\[^\n\r\]*ymm\[0-9\]" 1 } } */
197