1 /* SPDX-License-Identifier: MIT
2 *
3 * Permission is hereby granted, free of charge, to any person
4 * obtaining a copy of this software and associated documentation
5 * files (the "Software"), to deal in the Software without
6 * restriction, including without limitation the rights to use, copy,
7 * modify, merge, publish, distribute, sublicense, and/or sell copies
8 * of the Software, and to permit persons to whom the Software is
9 * furnished to do so, subject to the following conditions:
10 *
11 * The above copyright notice and this permission notice shall be
12 * included in all copies or substantial portions of the Software.
13 *
14 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
15 * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
16 * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
17 * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
18 * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
19 * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
20 * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
21 * SOFTWARE.
22 *
23 * Copyright:
24 * 2020 Evan Nemerson <evan@nemerson.com>
25 * 2020 Sean Maher <seanptmaher@gmail.com> (Copyright owned by Google, LLC)
26 */
27
28 #if !defined(SIMDE_ARM_NEON_ADDW_H)
29 #define SIMDE_ARM_NEON_ADDW_H
30
31 #include "types.h"
32 #include "add.h"
33 #include "movl.h"
34
35 HEDLEY_DIAGNOSTIC_PUSH
36 SIMDE_DISABLE_UNWANTED_DIAGNOSTICS
37 SIMDE_BEGIN_DECLS_
38
39 SIMDE_FUNCTION_ATTRIBUTES
40 simde_int16x8_t
simde_vaddw_s8(simde_int16x8_t a,simde_int8x8_t b)41 simde_vaddw_s8(simde_int16x8_t a, simde_int8x8_t b) {
42 #if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
43 return vaddw_s8(a, b);
44 #elif SIMDE_NATURAL_VECTOR_SIZE_GE(128)
45 return simde_vaddq_s16(a, simde_vmovl_s8(b));
46 #else
47 simde_int16x8_private r_;
48 simde_int16x8_private a_ = simde_int16x8_to_private(a);
49 simde_int8x8_private b_ = simde_int8x8_to_private(b);
50
51 #if (SIMDE_NATURAL_VECTOR_SIZE > 0) && defined(SIMDE_VECTOR_SUBSCRIPT_OPS) && defined(SIMDE_CONVERT_VECTOR_)
52 SIMDE_CONVERT_VECTOR_(r_.values, b_.values);
53 r_.values += a_.values;
54 #else
55 SIMDE_VECTORIZE
56 for (size_t i = 0 ; i < (sizeof(r_.values) / sizeof(r_.values[0])) ; i++) {
57 r_.values[i] = a_.values[i] + b_.values[i];
58 }
59 #endif
60
61 return simde_int16x8_from_private(r_);
62 #endif
63 }
64 #if defined(SIMDE_ARM_NEON_A32V7_ENABLE_NATIVE_ALIASES)
65 #undef vaddw_s8
66 #define vaddw_s8(a, b) simde_vaddw_s8((a), (b))
67 #endif
68
69 SIMDE_FUNCTION_ATTRIBUTES
70 simde_int32x4_t
simde_vaddw_s16(simde_int32x4_t a,simde_int16x4_t b)71 simde_vaddw_s16(simde_int32x4_t a, simde_int16x4_t b) {
72 #if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
73 return vaddw_s16(a, b);
74 #elif SIMDE_NATURAL_VECTOR_SIZE_GE(128)
75 return simde_vaddq_s32(a, simde_vmovl_s16(b));
76 #else
77 simde_int32x4_private r_;
78 simde_int32x4_private a_ = simde_int32x4_to_private(a);
79 simde_int16x4_private b_ = simde_int16x4_to_private(b);
80
81 #if (SIMDE_NATURAL_VECTOR_SIZE > 0) && defined(SIMDE_VECTOR_SUBSCRIPT_OPS) && defined(SIMDE_CONVERT_VECTOR_)
82 SIMDE_CONVERT_VECTOR_(r_.values, b_.values);
83 r_.values += a_.values;
84 #else
85 SIMDE_VECTORIZE
86 for (size_t i = 0 ; i < (sizeof(r_.values) / sizeof(r_.values[0])) ; i++) {
87 r_.values[i] = a_.values[i] + b_.values[i];
88 }
89 #endif
90
91 return simde_int32x4_from_private(r_);
92 #endif
93 }
94 #if defined(SIMDE_ARM_NEON_A32V7_ENABLE_NATIVE_ALIASES)
95 #undef vaddw_s16
96 #define vaddw_s16(a, b) simde_vaddw_s16((a), (b))
97 #endif
98
99 SIMDE_FUNCTION_ATTRIBUTES
100 simde_int64x2_t
simde_vaddw_s32(simde_int64x2_t a,simde_int32x2_t b)101 simde_vaddw_s32(simde_int64x2_t a, simde_int32x2_t b) {
102 #if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
103 return vaddw_s32(a, b);
104 #elif SIMDE_NATURAL_VECTOR_SIZE_GE(128)
105 return simde_vaddq_s64(a, simde_vmovl_s32(b));
106 #else
107 simde_int64x2_private r_;
108 simde_int64x2_private a_ = simde_int64x2_to_private(a);
109 simde_int32x2_private b_ = simde_int32x2_to_private(b);
110
111 #if (SIMDE_NATURAL_VECTOR_SIZE > 0) && defined(SIMDE_VECTOR_SUBSCRIPT_OPS) && defined(SIMDE_CONVERT_VECTOR_)
112 SIMDE_CONVERT_VECTOR_(r_.values, b_.values);
113 r_.values += a_.values;
114 #else
115 SIMDE_VECTORIZE
116 for (size_t i = 0 ; i < (sizeof(r_.values) / sizeof(r_.values[0])) ; i++) {
117 r_.values[i] = a_.values[i] + b_.values[i];
118 }
119 #endif
120
121 return simde_int64x2_from_private(r_);
122 #endif
123 }
124 #if defined(SIMDE_ARM_NEON_A32V7_ENABLE_NATIVE_ALIASES)
125 #undef vaddw_s32
126 #define vaddw_s32(a, b) simde_vaddw_s32((a), (b))
127 #endif
128
129 SIMDE_FUNCTION_ATTRIBUTES
130 simde_uint16x8_t
simde_vaddw_u8(simde_uint16x8_t a,simde_uint8x8_t b)131 simde_vaddw_u8(simde_uint16x8_t a, simde_uint8x8_t b) {
132 #if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
133 return vaddw_u8(a, b);
134 #elif SIMDE_NATURAL_VECTOR_SIZE_GE(128)
135 return simde_vaddq_u16(a, simde_vmovl_u8(b));
136 #else
137 simde_uint16x8_private r_;
138 simde_uint16x8_private a_ = simde_uint16x8_to_private(a);
139 simde_uint8x8_private b_ = simde_uint8x8_to_private(b);
140
141 #if (SIMDE_NATURAL_VECTOR_SIZE > 0) && defined(SIMDE_VECTOR_SUBSCRIPT_OPS) && defined(SIMDE_CONVERT_VECTOR_)
142 SIMDE_CONVERT_VECTOR_(r_.values, b_.values);
143 r_.values += a_.values;
144 #else
145 SIMDE_VECTORIZE
146 for (size_t i = 0 ; i < (sizeof(r_.values) / sizeof(r_.values[0])) ; i++) {
147 r_.values[i] = a_.values[i] + b_.values[i];
148 }
149 #endif
150
151 return simde_uint16x8_from_private(r_);
152 #endif
153 }
154 #if defined(SIMDE_ARM_NEON_A32V7_ENABLE_NATIVE_ALIASES)
155 #undef vaddw_u8
156 #define vaddw_u8(a, b) simde_vaddw_u8((a), (b))
157 #endif
158
159 SIMDE_FUNCTION_ATTRIBUTES
160 simde_uint32x4_t
simde_vaddw_u16(simde_uint32x4_t a,simde_uint16x4_t b)161 simde_vaddw_u16(simde_uint32x4_t a, simde_uint16x4_t b) {
162 #if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
163 return vaddw_u16(a, b);
164 #elif SIMDE_NATURAL_VECTOR_SIZE_GE(128)
165 return simde_vaddq_u32(a, simde_vmovl_u16(b));
166 #else
167 simde_uint32x4_private r_;
168 simde_uint32x4_private a_ = simde_uint32x4_to_private(a);
169 simde_uint16x4_private b_ = simde_uint16x4_to_private(b);
170
171 #if (SIMDE_NATURAL_VECTOR_SIZE > 0) && defined(SIMDE_VECTOR_SUBSCRIPT_OPS) && defined(SIMDE_CONVERT_VECTOR_)
172 SIMDE_CONVERT_VECTOR_(r_.values, b_.values);
173 r_.values += a_.values;
174 #else
175 SIMDE_VECTORIZE
176 for (size_t i = 0 ; i < (sizeof(r_.values) / sizeof(r_.values[0])) ; i++) {
177 r_.values[i] = a_.values[i] + b_.values[i];
178 }
179 #endif
180
181 return simde_uint32x4_from_private(r_);
182 #endif
183 }
184 #if defined(SIMDE_ARM_NEON_A32V7_ENABLE_NATIVE_ALIASES)
185 #undef vaddw_u16
186 #define vaddw_u16(a, b) simde_vaddw_u16((a), (b))
187 #endif
188
189 SIMDE_FUNCTION_ATTRIBUTES
190 simde_uint64x2_t
simde_vaddw_u32(simde_uint64x2_t a,simde_uint32x2_t b)191 simde_vaddw_u32(simde_uint64x2_t a, simde_uint32x2_t b) {
192 #if defined(SIMDE_ARM_NEON_A32V7_NATIVE)
193 return vaddw_u32(a, b);
194 #elif SIMDE_NATURAL_VECTOR_SIZE_GE(128)
195 return simde_vaddq_u64(a, simde_vmovl_u32(b));
196 #else
197 simde_uint64x2_private r_;
198 simde_uint64x2_private a_ = simde_uint64x2_to_private(a);
199 simde_uint32x2_private b_ = simde_uint32x2_to_private(b);
200
201 #if (SIMDE_NATURAL_VECTOR_SIZE > 0) && defined(SIMDE_VECTOR_SUBSCRIPT_OPS) && defined(SIMDE_CONVERT_VECTOR_)
202 SIMDE_CONVERT_VECTOR_(r_.values, b_.values);
203 r_.values += a_.values;
204 #else
205 SIMDE_VECTORIZE
206 for (size_t i = 0 ; i < (sizeof(r_.values) / sizeof(r_.values[0])) ; i++) {
207 r_.values[i] = a_.values[i] + b_.values[i];
208 }
209 #endif
210
211 return simde_uint64x2_from_private(r_);
212 #endif
213 }
214 #if defined(SIMDE_ARM_NEON_A32V7_ENABLE_NATIVE_ALIASES)
215 #undef vaddw_u32
216 #define vaddw_u32(a, b) simde_vaddw_u32((a), (b))
217 #endif
218
219 SIMDE_END_DECLS_
220 HEDLEY_DIAGNOSTIC_POP
221
222 #endif /* !defined(SIMDE_ARM_NEON_ADDW_H) */
223