Debian Codegraph @

gcc/testsuite/gcc.target/aarch64/sme2/acle-asm/mla_za64_u16_vg4x2.c

raw
1 /* { dg-final { check-function-bodies "**" "" "-DCHECK_ASM" } } */
2
3 #pragma GCC target "+sme-i16i64"
4
5 #include "test_sme2_acle.h"
6
7 /*
8 ** mla_0_z0_z0:
9 ** mov (w8|w9|w10|w11), #?0
10 ** umlall za\.d\[\1, 0:3, vgx2\], {z0\.h - z1\.h}, {z0\.h - z1\.h}
11 ** ret
12 */
13 TEST_ZA_XN (mla_0_z0_z0, svuint16x2_t,
14 svmla_za64_u16_vg4x2 (0, z0, z0),
15 svmla_za64_vg4x2 (0, z0, z0))
16
17 /*
18 ** mla_w0_z0_z0:
19 ** mov (w8|w9|w10|w11), w0
20 ** umlall za\.d\[\1, 0:3, vgx2\], {z0\.h - z1\.h}, {z0\.h - z1\.h}
21 ** ret
22 */
23 TEST_ZA_XN (mla_w0_z0_z0, svuint16x2_t,
24 svmla_za64_u16_vg4x2 (w0, z0, z0),
25 svmla_za64_vg4x2 (w0, z0, z0))
26
27 /*
28 ** mla_w8_z0_z4:
29 ** umlall za\.d\[w8, 0:3, vgx2\], {z0\.h - z1\.h}, {z4\.h - z5\.h}
30 ** ret
31 */
32 TEST_ZA_XN (mla_w8_z0_z4, svuint16x2_t,
33 svmla_za64_u16_vg4x2 (w8, z0, z4),
34 svmla_za64_vg4x2 (w8, z0, z4))
35
36 /*
37 ** mla_w8_z4_z18:
38 ** umlall za\.d\[w8, 0:3, vgx2\], {z4\.h - z5\.h}, {z18\.h - z19\.h}
39 ** ret
40 */
41 TEST_ZA_XN (mla_w8_z4_z18, svuint16x2_t,
42 svmla_za64_u16_vg4x2 (w8, z4, z18),
43 svmla_za64_vg4x2 (w8, z4, z18))
44
45 /* Leave the assembler to check for correctness for misaligned registers. */
46
47 /*
48 ** mla_w8_z0_z23:
49 ** ...
50 ** umlall za\.d\[w8, 0:3, vgx2\], {z0\.h - z1\.h}, [^\n]+
51 ** ret
52 */
53 TEST_ZA_XN (mla_w8_z0_z23, svuint16x2_t,
54 svmla_za64_u16_vg4x2 (w8, z0, z23),
55 svmla_za64_vg4x2 (w8, z0, z23))
56
57 /*
58 ** mla_w8_z23_z0:
59 ** ...
60 ** umlall za\.d\[w8, 0:3, vgx2\], [^\n]+, {z0\.h - z1\.h}
61 ** ret
62 */
63 TEST_ZA_XN (mla_w8_z23_z0, svuint16x2_t,
64 svmla_za64_u16_vg4x2 (w8, z23, z0),
65 svmla_za64_vg4x2 (w8, z23, z0))
66
67 /*
68 ** mla_w8_z18_z28:
69 ** umlall za\.d\[w8, 0:3, vgx2\], {z18\.h - z19\.h}, {z28\.h - z29\.h}
70 ** ret
71 */
72 TEST_ZA_XN (mla_w8_z18_z28, svuint16x2_t,
73 svmla_za64_u16_vg4x2 (w8, z18, z28),
74 svmla_za64_vg4x2 (w8, z18, z28))
75
76 /*
77 ** mla_w8_z28_z4:
78 ** umlall za\.d\[w8, 0:3, vgx2\], {z28\.h - z29\.h}, {z4\.h - z5\.h}
79 ** ret
80 */
81 TEST_ZA_XN (mla_w8_z28_z4, svuint16x2_t,
82 svmla_za64_u16_vg4x2 (w8, z28, z4),
83 svmla_za64_vg4x2 (w8, z28, z4))
84
85 /*
86 ** mla_w8p1_z4_z0:
87 ** add (w8|w9|w10|w11), w8, #?1
88 ** umlall za\.d\[\1, 0:3, vgx2\], {z4\.h - z5\.h}, {z0\.h - z1\.h}
89 ** ret
90 */
91 TEST_ZA_XN (mla_w8p1_z4_z0, svuint16x2_t,
92 svmla_za64_u16_vg4x2 (w8 + 1, z4, z0),
93 svmla_za64_vg4x2 (w8 + 1, z4, z0))
94
95 /*
96 ** mla_w8p2_z4_z0:
97 ** add (w8|w9|w10|w11), w8, #?2
98 ** umlall za\.d\[w8, 0:3, vgx2\], {z4\.h - z5\.h}, {z0\.h - z1\.h}
99 ** ret
100 */
101 TEST_ZA_XN (mla_w8p2_z4_z0, svuint16x2_t,
102 svmla_za64_u16_vg4x2 (w8 + 2, z4, z0),
103 svmla_za64_vg4x2 (w8 + 2, z4, z0))
104
105 /*
106 ** mla_w11p4_z4_z0:
107 ** umlall za\.d\[w11, 4:7, vgx2\], {z4\.h - z5\.h}, {z0\.h - z1\.h}
108 ** ret
109 */
110 TEST_ZA_XN (mla_w11p4_z4_z0, svuint16x2_t,
111 svmla_za64_u16_vg4x2 (w11 + 4, z4, z0),
112 svmla_za64_vg4x2 (w11 + 4, z4, z0))
113
114 /*
115 ** mla_w8p7_z4_z0:
116 ** add (w8|w9|w10|w11), w8, #?7
117 ** umlall za\.d\[\1, 0:3, vgx2\], {z4\.h - z5\.h}, {z0\.h - z1\.h}
118 ** ret
119 */
120 TEST_ZA_XN (mla_w8p7_z4_z0, svuint16x2_t,
121 svmla_za64_u16_vg4x2 (w8 + 7, z4, z0),
122 svmla_za64_vg4x2 (w8 + 7, z4, z0))
123
124 /*
125 ** mla_w8p8_z4_z4:
126 ** add (w8|w9|w10|w11), w8, #?8
127 ** umlall za\.d\[\1, 0:3, vgx2\], {z4\.h - z5\.h}, {z4\.h - z5\.h}
128 ** ret
129 */
130 TEST_ZA_XN (mla_w8p8_z4_z4, svuint16x2_t,
131 svmla_za64_u16_vg4x2 (w8 + 8, z4, z4),
132 svmla_za64_vg4x2 (w8 + 8, z4, z4))
133
134 /*
135 ** mla_w8m1_z4_z0:
136 ** sub (w8|w9|w10|w11), w8, #?1
137 ** umlall za\.d\[\1, 0:3, vgx2\], {z4\.h - z5\.h}, {z0\.h - z1\.h}
138 ** ret
139 */
140 TEST_ZA_XN (mla_w8m1_z4_z0, svuint16x2_t,
141 svmla_za64_u16_vg4x2 (w8 - 1, z4, z0),
142 svmla_za64_vg4x2 (w8 - 1, z4, z0))
143
144 /*
145 ** mla_single_0_z1_z0:
146 ** mov (w8|w9|w10|w11), #?0
147 ** umlall za\.d\[\1, 0:3, vgx2\], {z1\.h - z2\.h}, z0\.h
148 ** ret
149 */
150 TEST_ZA_SINGLE (mla_single_0_z1_z0, svuint16x2_t, svuint16_t,
151 svmla_single_za64_u16_vg4x2 (0, z1, z0),
152 svmla_za64_vg4x2 (0, z1, z0))
153
154 /*
155 ** mla_single_w0_z1_z0:
156 ** mov (w8|w9|w10|w11), w0
157 ** umlall za\.d\[\1, 0:3, vgx2\], {z1\.h - z2\.h}, z0\.h
158 ** ret
159 */
160 TEST_ZA_SINGLE (mla_single_w0_z1_z0, svuint16x2_t, svuint16_t,
161 svmla_single_za64_u16_vg4x2 (w0, z1, z0),
162 svmla_za64_vg4x2 (w0, z1, z0))
163
164 /*
165 ** mla_single_w8_z1_z0:
166 ** umlall za\.d\[w8, 0:3, vgx2\], {z1\.h - z2\.h}, z0\.h
167 ** ret
168 */
169 TEST_ZA_SINGLE (mla_single_w8_z1_z0, svuint16x2_t, svuint16_t,
170 svmla_single_za64_u16_vg4x2 (w8, z1, z0),
171 svmla_za64_vg4x2 (w8, z1, z0))
172
173 /*
174 ** mla_single_w8p1_z1_z0:
175 ** add (w8|w9|w10|w11), w8, #?1
176 ** umlall za\.d\[\1, 0:3, vgx2\], {z1\.h - z2\.h}, z0\.h
177 ** ret
178 */
179 TEST_ZA_SINGLE (mla_single_w8p1_z1_z0, svuint16x2_t, svuint16_t,
180 svmla_single_za64_u16_vg4x2 (w8 + 1, z1, z0),
181 svmla_za64_vg4x2 (w8 + 1, z1, z0))
182
183 /*
184 ** mla_single_w8p2_z20_z0:
185 ** add (w8|w9|w10|w11), w8, #?2
186 ** umlall za\.d\[\1, 0:3, vgx2\], {z20\.h - z21\.h}, z0\.h
187 ** ret
188 */
189 TEST_ZA_SINGLE (mla_single_w8p2_z20_z0, svuint16x2_t, svuint16_t,
190 svmla_single_za64_u16_vg4x2 (w8 + 2, z20, z0),
191 svmla_za64_vg4x2 (w8 + 2, z20, z0))
192
193 /*
194 ** mla_single_w11p4_z27_z0:
195 ** umlall za\.d\[w11, 4:7, vgx2\], {z27\.h - z28\.h}, z0\.h
196 ** ret
197 */
198 TEST_ZA_SINGLE (mla_single_w11p4_z27_z0, svuint16x2_t, svuint16_t,
199 svmla_single_za64_u16_vg4x2 (w11 + 4, z27, z0),
200 svmla_za64_vg4x2 (w11 + 4, z27, z0))
201
202 /*
203 ** mla_single_w8p7_z1_z0:
204 ** add (w8|w9|w10|w11), w8, #?7
205 ** umlall za\.d\[\1, 0:3, vgx2\], {z1\.h - z2\.h}, z0\.h
206 ** ret
207 */
208 TEST_ZA_SINGLE (mla_single_w8p7_z1_z0, svuint16x2_t, svuint16_t,
209 svmla_single_za64_u16_vg4x2 (w8 + 7, z1, z0),
210 svmla_za64_vg4x2 (w8 + 7, z1, z0))
211
212 /*
213 ** mla_single_w8p8_z1_z0:
214 ** add (w8|w9|w10|w11), w8, #?8
215 ** umlall za\.d\[\1, 0:3, vgx2\], {z1\.h - z2\.h}, z0\.h
216 ** ret
217 */
218 TEST_ZA_SINGLE (mla_single_w8p8_z1_z0, svuint16x2_t, svuint16_t,
219 svmla_single_za64_u16_vg4x2 (w8 + 8, z1, z0),
220 svmla_za64_vg4x2 (w8 + 8, z1, z0))
221
222 /*
223 ** mla_single_w0m1_z1_z0:
224 ** sub (w8|w9|w10|w11), w0, #?1
225 ** umlall za\.d\[\1, 0:3, vgx2\], {z1\.h - z2\.h}, z0\.h
226 ** ret
227 */
228 TEST_ZA_SINGLE (mla_single_w0m1_z1_z0, svuint16x2_t, svuint16_t,
229 svmla_single_za64_u16_vg4x2 (w0 - 1, z1, z0),
230 svmla_za64_vg4x2 (w0 - 1, z1, z0))
231
232 /*
233 ** mla_single_w8_z0_z15:
234 ** str d15, \[sp, #?-16\]!
235 ** umlall za\.d\[w8, 0:3, vgx2\], {z0\.h - z1\.h}, z15\.h
236 ** ldr d15, \[sp\], #?16
237 ** ret
238 */
239 TEST_ZA_SINGLE_Z15 (mla_single_w8_z0_z15, svuint16x2_t, svuint16_t,
240 svmla_single_za64_u16_vg4x2 (w8, z0, z15),
241 svmla_za64_vg4x2 (w8, z0, z15))
242
243 /*
244 ** mla_single_w8_z20_z16:
245 ** mov (z[0-7]).d, z16.d
246 ** umlall za\.d\[w8, 0:3, vgx2\], {z20\.h - z21\.h}, \1\.h
247 ** ret
248 */
249 TEST_ZA_SINGLE (mla_single_w8_z20_z16, svuint16x2_t, svuint16_t,
250 svmla_single_za64_u16_vg4x2 (w8, z20, z16),
251 svmla_za64_vg4x2 (w8, z20, z16))