|
yading@10
|
1 /*
|
|
yading@10
|
2 * Copyright (c) 2008 Mans Rullgard <mans@mansr.com>
|
|
yading@10
|
3 *
|
|
yading@10
|
4 * This file is part of FFmpeg.
|
|
yading@10
|
5 *
|
|
yading@10
|
6 * FFmpeg is free software; you can redistribute it and/or
|
|
yading@10
|
7 * modify it under the terms of the GNU Lesser General Public
|
|
yading@10
|
8 * License as published by the Free Software Foundation; either
|
|
yading@10
|
9 * version 2.1 of the License, or (at your option) any later version.
|
|
yading@10
|
10 *
|
|
yading@10
|
11 * FFmpeg is distributed in the hope that it will be useful,
|
|
yading@10
|
12 * but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
yading@10
|
13 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
|
|
yading@10
|
14 * Lesser General Public License for more details.
|
|
yading@10
|
15 *
|
|
yading@10
|
16 * You should have received a copy of the GNU Lesser General Public
|
|
yading@10
|
17 * License along with FFmpeg; if not, write to the Free Software
|
|
yading@10
|
18 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
|
|
yading@10
|
19 */
|
|
yading@10
|
20
|
|
yading@10
|
21 #include "libavutil/arm/asm.S"
|
|
yading@10
|
22
|
|
yading@10
|
23 function ff_h264_idct_add_neon, export=1
|
|
yading@10
|
24 vld1.64 {d0-d3}, [r1,:128]
|
|
yading@10
|
25 vmov.i16 q15, #0
|
|
yading@10
|
26
|
|
yading@10
|
27 vswp d1, d2
|
|
yading@10
|
28 vst1.16 {q15}, [r1,:128]!
|
|
yading@10
|
29 vadd.i16 d4, d0, d1
|
|
yading@10
|
30 vst1.16 {q15}, [r1,:128]!
|
|
yading@10
|
31 vshr.s16 q8, q1, #1
|
|
yading@10
|
32 vsub.i16 d5, d0, d1
|
|
yading@10
|
33 vadd.i16 d6, d2, d17
|
|
yading@10
|
34 vsub.i16 d7, d16, d3
|
|
yading@10
|
35 vadd.i16 q0, q2, q3
|
|
yading@10
|
36 vsub.i16 q1, q2, q3
|
|
yading@10
|
37
|
|
yading@10
|
38 vtrn.16 d0, d1
|
|
yading@10
|
39 vtrn.16 d3, d2
|
|
yading@10
|
40 vtrn.32 d0, d3
|
|
yading@10
|
41 vtrn.32 d1, d2
|
|
yading@10
|
42
|
|
yading@10
|
43 vadd.i16 d4, d0, d3
|
|
yading@10
|
44 vld1.32 {d18[0]}, [r0,:32], r2
|
|
yading@10
|
45 vswp d1, d3
|
|
yading@10
|
46 vshr.s16 q8, q1, #1
|
|
yading@10
|
47 vld1.32 {d19[1]}, [r0,:32], r2
|
|
yading@10
|
48 vsub.i16 d5, d0, d1
|
|
yading@10
|
49 vld1.32 {d18[1]}, [r0,:32], r2
|
|
yading@10
|
50 vadd.i16 d6, d16, d3
|
|
yading@10
|
51 vld1.32 {d19[0]}, [r0,:32], r2
|
|
yading@10
|
52 vsub.i16 d7, d2, d17
|
|
yading@10
|
53 sub r0, r0, r2, lsl #2
|
|
yading@10
|
54 vadd.i16 q0, q2, q3
|
|
yading@10
|
55 vsub.i16 q1, q2, q3
|
|
yading@10
|
56
|
|
yading@10
|
57 vrshr.s16 q0, q0, #6
|
|
yading@10
|
58 vrshr.s16 q1, q1, #6
|
|
yading@10
|
59
|
|
yading@10
|
60 vaddw.u8 q0, q0, d18
|
|
yading@10
|
61 vaddw.u8 q1, q1, d19
|
|
yading@10
|
62
|
|
yading@10
|
63 vqmovun.s16 d0, q0
|
|
yading@10
|
64 vqmovun.s16 d1, q1
|
|
yading@10
|
65
|
|
yading@10
|
66 vst1.32 {d0[0]}, [r0,:32], r2
|
|
yading@10
|
67 vst1.32 {d1[1]}, [r0,:32], r2
|
|
yading@10
|
68 vst1.32 {d0[1]}, [r0,:32], r2
|
|
yading@10
|
69 vst1.32 {d1[0]}, [r0,:32], r2
|
|
yading@10
|
70
|
|
yading@10
|
71 sub r1, r1, #32
|
|
yading@10
|
72 bx lr
|
|
yading@10
|
73 endfunc
|
|
yading@10
|
74
|
|
yading@10
|
75 function ff_h264_idct_dc_add_neon, export=1
|
|
yading@10
|
76 mov r3, #0
|
|
yading@10
|
77 vld1.16 {d2[],d3[]}, [r1,:16]
|
|
yading@10
|
78 strh r3, [r1]
|
|
yading@10
|
79 vrshr.s16 q1, q1, #6
|
|
yading@10
|
80 vld1.32 {d0[0]}, [r0,:32], r2
|
|
yading@10
|
81 vld1.32 {d0[1]}, [r0,:32], r2
|
|
yading@10
|
82 vaddw.u8 q2, q1, d0
|
|
yading@10
|
83 vld1.32 {d1[0]}, [r0,:32], r2
|
|
yading@10
|
84 vld1.32 {d1[1]}, [r0,:32], r2
|
|
yading@10
|
85 vaddw.u8 q1, q1, d1
|
|
yading@10
|
86 vqmovun.s16 d0, q2
|
|
yading@10
|
87 vqmovun.s16 d1, q1
|
|
yading@10
|
88 sub r0, r0, r2, lsl #2
|
|
yading@10
|
89 vst1.32 {d0[0]}, [r0,:32], r2
|
|
yading@10
|
90 vst1.32 {d0[1]}, [r0,:32], r2
|
|
yading@10
|
91 vst1.32 {d1[0]}, [r0,:32], r2
|
|
yading@10
|
92 vst1.32 {d1[1]}, [r0,:32], r2
|
|
yading@10
|
93 bx lr
|
|
yading@10
|
94 endfunc
|
|
yading@10
|
95
|
|
yading@10
|
96 function ff_h264_idct_add16_neon, export=1
|
|
yading@10
|
97 push {r4-r8,lr}
|
|
yading@10
|
98 mov r4, r0
|
|
yading@10
|
99 mov r5, r1
|
|
yading@10
|
100 mov r1, r2
|
|
yading@10
|
101 mov r2, r3
|
|
yading@10
|
102 ldr r6, [sp, #24]
|
|
yading@10
|
103 movrel r7, scan8
|
|
yading@10
|
104 mov ip, #16
|
|
yading@10
|
105 1: ldrb r8, [r7], #1
|
|
yading@10
|
106 ldr r0, [r5], #4
|
|
yading@10
|
107 ldrb r8, [r6, r8]
|
|
yading@10
|
108 subs r8, r8, #1
|
|
yading@10
|
109 blt 2f
|
|
yading@10
|
110 ldrsh lr, [r1]
|
|
yading@10
|
111 add r0, r0, r4
|
|
yading@10
|
112 it ne
|
|
yading@10
|
113 movne lr, #0
|
|
yading@10
|
114 cmp lr, #0
|
|
yading@10
|
115 ite ne
|
|
yading@10
|
116 adrne lr, ff_h264_idct_dc_add_neon + CONFIG_THUMB
|
|
yading@10
|
117 adreq lr, ff_h264_idct_add_neon + CONFIG_THUMB
|
|
yading@10
|
118 blx lr
|
|
yading@10
|
119 2: subs ip, ip, #1
|
|
yading@10
|
120 add r1, r1, #32
|
|
yading@10
|
121 bne 1b
|
|
yading@10
|
122 pop {r4-r8,pc}
|
|
yading@10
|
123 endfunc
|
|
yading@10
|
124
|
|
yading@10
|
125 function ff_h264_idct_add16intra_neon, export=1
|
|
yading@10
|
126 push {r4-r8,lr}
|
|
yading@10
|
127 mov r4, r0
|
|
yading@10
|
128 mov r5, r1
|
|
yading@10
|
129 mov r1, r2
|
|
yading@10
|
130 mov r2, r3
|
|
yading@10
|
131 ldr r6, [sp, #24]
|
|
yading@10
|
132 movrel r7, scan8
|
|
yading@10
|
133 mov ip, #16
|
|
yading@10
|
134 1: ldrb r8, [r7], #1
|
|
yading@10
|
135 ldr r0, [r5], #4
|
|
yading@10
|
136 ldrb r8, [r6, r8]
|
|
yading@10
|
137 add r0, r0, r4
|
|
yading@10
|
138 cmp r8, #0
|
|
yading@10
|
139 ldrsh r8, [r1]
|
|
yading@10
|
140 iteet ne
|
|
yading@10
|
141 adrne lr, ff_h264_idct_add_neon + CONFIG_THUMB
|
|
yading@10
|
142 adreq lr, ff_h264_idct_dc_add_neon + CONFIG_THUMB
|
|
yading@10
|
143 cmpeq r8, #0
|
|
yading@10
|
144 blxne lr
|
|
yading@10
|
145 subs ip, ip, #1
|
|
yading@10
|
146 add r1, r1, #32
|
|
yading@10
|
147 bne 1b
|
|
yading@10
|
148 pop {r4-r8,pc}
|
|
yading@10
|
149 endfunc
|
|
yading@10
|
150
|
|
yading@10
|
151 function ff_h264_idct_add8_neon, export=1
|
|
yading@10
|
152 push {r4-r10,lr}
|
|
yading@10
|
153 ldm r0, {r4,r9}
|
|
yading@10
|
154 add r5, r1, #16*4
|
|
yading@10
|
155 add r1, r2, #16*32
|
|
yading@10
|
156 mov r2, r3
|
|
yading@10
|
157 mov r10, r1
|
|
yading@10
|
158 ldr r6, [sp, #32]
|
|
yading@10
|
159 movrel r7, scan8+16
|
|
yading@10
|
160 mov r12, #0
|
|
yading@10
|
161 1: ldrb r8, [r7, r12]
|
|
yading@10
|
162 ldr r0, [r5, r12, lsl #2]
|
|
yading@10
|
163 ldrb r8, [r6, r8]
|
|
yading@10
|
164 add r0, r0, r4
|
|
yading@10
|
165 add r1, r10, r12, lsl #5
|
|
yading@10
|
166 cmp r8, #0
|
|
yading@10
|
167 ldrsh r8, [r1]
|
|
yading@10
|
168 iteet ne
|
|
yading@10
|
169 adrne lr, ff_h264_idct_add_neon + CONFIG_THUMB
|
|
yading@10
|
170 adreq lr, ff_h264_idct_dc_add_neon + CONFIG_THUMB
|
|
yading@10
|
171 cmpeq r8, #0
|
|
yading@10
|
172 blxne lr
|
|
yading@10
|
173 add r12, r12, #1
|
|
yading@10
|
174 cmp r12, #4
|
|
yading@10
|
175 itt eq
|
|
yading@10
|
176 moveq r12, #16
|
|
yading@10
|
177 moveq r4, r9
|
|
yading@10
|
178 cmp r12, #20
|
|
yading@10
|
179 blt 1b
|
|
yading@10
|
180 pop {r4-r10,pc}
|
|
yading@10
|
181 endfunc
|
|
yading@10
|
182
|
|
yading@10
|
183 .macro idct8x8_cols pass
|
|
yading@10
|
184 .if \pass == 0
|
|
yading@10
|
185 qa .req q2
|
|
yading@10
|
186 qb .req q14
|
|
yading@10
|
187 vshr.s16 q2, q10, #1
|
|
yading@10
|
188 vadd.i16 q0, q8, q12
|
|
yading@10
|
189 vld1.16 {q14-q15},[r1,:128]
|
|
yading@10
|
190 vst1.16 {q3}, [r1,:128]!
|
|
yading@10
|
191 vst1.16 {q3}, [r1,:128]!
|
|
yading@10
|
192 vsub.i16 q1, q8, q12
|
|
yading@10
|
193 vshr.s16 q3, q14, #1
|
|
yading@10
|
194 vsub.i16 q2, q2, q14
|
|
yading@10
|
195 vadd.i16 q3, q3, q10
|
|
yading@10
|
196 .else
|
|
yading@10
|
197 qa .req q14
|
|
yading@10
|
198 qb .req q2
|
|
yading@10
|
199 vtrn.32 q8, q10
|
|
yading@10
|
200 vtrn.16 q12, q13
|
|
yading@10
|
201 vtrn.32 q9, q11
|
|
yading@10
|
202 vtrn.32 q12, q2
|
|
yading@10
|
203 vtrn.32 q13, q15
|
|
yading@10
|
204 vswp d21, d4
|
|
yading@10
|
205 vshr.s16 q14, q10, #1
|
|
yading@10
|
206 vswp d17, d24
|
|
yading@10
|
207 vshr.s16 q3, q2, #1
|
|
yading@10
|
208 vswp d19, d26
|
|
yading@10
|
209 vadd.i16 q0, q8, q12
|
|
yading@10
|
210 vswp d23, d30
|
|
yading@10
|
211 vsub.i16 q1, q8, q12
|
|
yading@10
|
212 vsub.i16 q14, q14, q2
|
|
yading@10
|
213 vadd.i16 q3, q3, q10
|
|
yading@10
|
214 .endif
|
|
yading@10
|
215 vadd.i16 q10, q1, qa
|
|
yading@10
|
216 vsub.i16 q12, q1, qa
|
|
yading@10
|
217 vadd.i16 q8, q0, q3
|
|
yading@10
|
218 vsub.i16 qb, q0, q3
|
|
yading@10
|
219 vsub.i16 q0, q13, q11
|
|
yading@10
|
220 vadd.i16 q1, q15, q9
|
|
yading@10
|
221 vsub.i16 qa, q15, q9
|
|
yading@10
|
222 vadd.i16 q3, q13, q11
|
|
yading@10
|
223 vsub.i16 q0, q0, q15
|
|
yading@10
|
224 vsub.i16 q1, q1, q11
|
|
yading@10
|
225 vadd.i16 qa, qa, q13
|
|
yading@10
|
226 vadd.i16 q3, q3, q9
|
|
yading@10
|
227 vshr.s16 q9, q9, #1
|
|
yading@10
|
228 vshr.s16 q11, q11, #1
|
|
yading@10
|
229 vshr.s16 q13, q13, #1
|
|
yading@10
|
230 vshr.s16 q15, q15, #1
|
|
yading@10
|
231 vsub.i16 q0, q0, q15
|
|
yading@10
|
232 vsub.i16 q1, q1, q11
|
|
yading@10
|
233 vadd.i16 qa, qa, q13
|
|
yading@10
|
234 vadd.i16 q3, q3, q9
|
|
yading@10
|
235 vshr.s16 q9, q0, #2
|
|
yading@10
|
236 vshr.s16 q11, q1, #2
|
|
yading@10
|
237 vshr.s16 q13, qa, #2
|
|
yading@10
|
238 vshr.s16 q15, q3, #2
|
|
yading@10
|
239 vsub.i16 q3, q3, q9
|
|
yading@10
|
240 vsub.i16 qa, q11, qa
|
|
yading@10
|
241 vadd.i16 q1, q1, q13
|
|
yading@10
|
242 vadd.i16 q0, q0, q15
|
|
yading@10
|
243 .if \pass == 0
|
|
yading@10
|
244 vsub.i16 q15, q8, q3
|
|
yading@10
|
245 vadd.i16 q8, q8, q3
|
|
yading@10
|
246 vadd.i16 q9, q10, q2
|
|
yading@10
|
247 vsub.i16 q2, q10, q2
|
|
yading@10
|
248 vtrn.16 q8, q9
|
|
yading@10
|
249 vadd.i16 q10, q12, q1
|
|
yading@10
|
250 vtrn.16 q2, q15
|
|
yading@10
|
251 vadd.i16 q11, q14, q0
|
|
yading@10
|
252 vsub.i16 q13, q12, q1
|
|
yading@10
|
253 vtrn.16 q10, q11
|
|
yading@10
|
254 vsub.i16 q12, q14, q0
|
|
yading@10
|
255 .else
|
|
yading@10
|
256 vsub.i16 q15, q8, q3
|
|
yading@10
|
257 vadd.i16 q8, q8, q3
|
|
yading@10
|
258 vadd.i16 q9, q10, q14
|
|
yading@10
|
259 vsub.i16 q14, q10, q14
|
|
yading@10
|
260 vadd.i16 q10, q12, q1
|
|
yading@10
|
261 vsub.i16 q13, q12, q1
|
|
yading@10
|
262 vadd.i16 q11, q2, q0
|
|
yading@10
|
263 vsub.i16 q12, q2, q0
|
|
yading@10
|
264 .endif
|
|
yading@10
|
265 .unreq qa
|
|
yading@10
|
266 .unreq qb
|
|
yading@10
|
267 .endm
|
|
yading@10
|
268
|
|
yading@10
|
269 function ff_h264_idct8_add_neon, export=1
|
|
yading@10
|
270 vmov.i16 q3, #0
|
|
yading@10
|
271 vld1.16 {q8-q9}, [r1,:128]
|
|
yading@10
|
272 vst1.16 {q3}, [r1,:128]!
|
|
yading@10
|
273 vst1.16 {q3}, [r1,:128]!
|
|
yading@10
|
274 vld1.16 {q10-q11},[r1,:128]
|
|
yading@10
|
275 vst1.16 {q3}, [r1,:128]!
|
|
yading@10
|
276 vst1.16 {q3}, [r1,:128]!
|
|
yading@10
|
277 vld1.16 {q12-q13},[r1,:128]
|
|
yading@10
|
278 vst1.16 {q3}, [r1,:128]!
|
|
yading@10
|
279 vst1.16 {q3}, [r1,:128]!
|
|
yading@10
|
280
|
|
yading@10
|
281 idct8x8_cols 0
|
|
yading@10
|
282 idct8x8_cols 1
|
|
yading@10
|
283
|
|
yading@10
|
284 mov r3, r0
|
|
yading@10
|
285 vrshr.s16 q8, q8, #6
|
|
yading@10
|
286 vld1.8 {d0}, [r0,:64], r2
|
|
yading@10
|
287 vrshr.s16 q9, q9, #6
|
|
yading@10
|
288 vld1.8 {d1}, [r0,:64], r2
|
|
yading@10
|
289 vrshr.s16 q10, q10, #6
|
|
yading@10
|
290 vld1.8 {d2}, [r0,:64], r2
|
|
yading@10
|
291 vrshr.s16 q11, q11, #6
|
|
yading@10
|
292 vld1.8 {d3}, [r0,:64], r2
|
|
yading@10
|
293 vrshr.s16 q12, q12, #6
|
|
yading@10
|
294 vld1.8 {d4}, [r0,:64], r2
|
|
yading@10
|
295 vrshr.s16 q13, q13, #6
|
|
yading@10
|
296 vld1.8 {d5}, [r0,:64], r2
|
|
yading@10
|
297 vrshr.s16 q14, q14, #6
|
|
yading@10
|
298 vld1.8 {d6}, [r0,:64], r2
|
|
yading@10
|
299 vrshr.s16 q15, q15, #6
|
|
yading@10
|
300 vld1.8 {d7}, [r0,:64], r2
|
|
yading@10
|
301 vaddw.u8 q8, q8, d0
|
|
yading@10
|
302 vaddw.u8 q9, q9, d1
|
|
yading@10
|
303 vaddw.u8 q10, q10, d2
|
|
yading@10
|
304 vqmovun.s16 d0, q8
|
|
yading@10
|
305 vaddw.u8 q11, q11, d3
|
|
yading@10
|
306 vqmovun.s16 d1, q9
|
|
yading@10
|
307 vaddw.u8 q12, q12, d4
|
|
yading@10
|
308 vqmovun.s16 d2, q10
|
|
yading@10
|
309 vst1.8 {d0}, [r3,:64], r2
|
|
yading@10
|
310 vaddw.u8 q13, q13, d5
|
|
yading@10
|
311 vqmovun.s16 d3, q11
|
|
yading@10
|
312 vst1.8 {d1}, [r3,:64], r2
|
|
yading@10
|
313 vaddw.u8 q14, q14, d6
|
|
yading@10
|
314 vqmovun.s16 d4, q12
|
|
yading@10
|
315 vst1.8 {d2}, [r3,:64], r2
|
|
yading@10
|
316 vaddw.u8 q15, q15, d7
|
|
yading@10
|
317 vqmovun.s16 d5, q13
|
|
yading@10
|
318 vst1.8 {d3}, [r3,:64], r2
|
|
yading@10
|
319 vqmovun.s16 d6, q14
|
|
yading@10
|
320 vqmovun.s16 d7, q15
|
|
yading@10
|
321 vst1.8 {d4}, [r3,:64], r2
|
|
yading@10
|
322 vst1.8 {d5}, [r3,:64], r2
|
|
yading@10
|
323 vst1.8 {d6}, [r3,:64], r2
|
|
yading@10
|
324 vst1.8 {d7}, [r3,:64], r2
|
|
yading@10
|
325
|
|
yading@10
|
326 sub r1, r1, #128
|
|
yading@10
|
327 bx lr
|
|
yading@10
|
328 endfunc
|
|
yading@10
|
329
|
|
yading@10
|
330 function ff_h264_idct8_dc_add_neon, export=1
|
|
yading@10
|
331 mov r3, #0
|
|
yading@10
|
332 vld1.16 {d30[],d31[]},[r1,:16]
|
|
yading@10
|
333 strh r3, [r1]
|
|
yading@10
|
334 vld1.32 {d0}, [r0,:64], r2
|
|
yading@10
|
335 vrshr.s16 q15, q15, #6
|
|
yading@10
|
336 vld1.32 {d1}, [r0,:64], r2
|
|
yading@10
|
337 vld1.32 {d2}, [r0,:64], r2
|
|
yading@10
|
338 vaddw.u8 q8, q15, d0
|
|
yading@10
|
339 vld1.32 {d3}, [r0,:64], r2
|
|
yading@10
|
340 vaddw.u8 q9, q15, d1
|
|
yading@10
|
341 vld1.32 {d4}, [r0,:64], r2
|
|
yading@10
|
342 vaddw.u8 q10, q15, d2
|
|
yading@10
|
343 vld1.32 {d5}, [r0,:64], r2
|
|
yading@10
|
344 vaddw.u8 q11, q15, d3
|
|
yading@10
|
345 vld1.32 {d6}, [r0,:64], r2
|
|
yading@10
|
346 vaddw.u8 q12, q15, d4
|
|
yading@10
|
347 vld1.32 {d7}, [r0,:64], r2
|
|
yading@10
|
348 vaddw.u8 q13, q15, d5
|
|
yading@10
|
349 vaddw.u8 q14, q15, d6
|
|
yading@10
|
350 vaddw.u8 q15, q15, d7
|
|
yading@10
|
351 vqmovun.s16 d0, q8
|
|
yading@10
|
352 vqmovun.s16 d1, q9
|
|
yading@10
|
353 vqmovun.s16 d2, q10
|
|
yading@10
|
354 vqmovun.s16 d3, q11
|
|
yading@10
|
355 sub r0, r0, r2, lsl #3
|
|
yading@10
|
356 vst1.32 {d0}, [r0,:64], r2
|
|
yading@10
|
357 vqmovun.s16 d4, q12
|
|
yading@10
|
358 vst1.32 {d1}, [r0,:64], r2
|
|
yading@10
|
359 vqmovun.s16 d5, q13
|
|
yading@10
|
360 vst1.32 {d2}, [r0,:64], r2
|
|
yading@10
|
361 vqmovun.s16 d6, q14
|
|
yading@10
|
362 vst1.32 {d3}, [r0,:64], r2
|
|
yading@10
|
363 vqmovun.s16 d7, q15
|
|
yading@10
|
364 vst1.32 {d4}, [r0,:64], r2
|
|
yading@10
|
365 vst1.32 {d5}, [r0,:64], r2
|
|
yading@10
|
366 vst1.32 {d6}, [r0,:64], r2
|
|
yading@10
|
367 vst1.32 {d7}, [r0,:64], r2
|
|
yading@10
|
368 bx lr
|
|
yading@10
|
369 endfunc
|
|
yading@10
|
370
|
|
yading@10
|
371 function ff_h264_idct8_add4_neon, export=1
|
|
yading@10
|
372 push {r4-r8,lr}
|
|
yading@10
|
373 mov r4, r0
|
|
yading@10
|
374 mov r5, r1
|
|
yading@10
|
375 mov r1, r2
|
|
yading@10
|
376 mov r2, r3
|
|
yading@10
|
377 ldr r6, [sp, #24]
|
|
yading@10
|
378 movrel r7, scan8
|
|
yading@10
|
379 mov r12, #16
|
|
yading@10
|
380 1: ldrb r8, [r7], #4
|
|
yading@10
|
381 ldr r0, [r5], #16
|
|
yading@10
|
382 ldrb r8, [r6, r8]
|
|
yading@10
|
383 subs r8, r8, #1
|
|
yading@10
|
384 blt 2f
|
|
yading@10
|
385 ldrsh lr, [r1]
|
|
yading@10
|
386 add r0, r0, r4
|
|
yading@10
|
387 it ne
|
|
yading@10
|
388 movne lr, #0
|
|
yading@10
|
389 cmp lr, #0
|
|
yading@10
|
390 ite ne
|
|
yading@10
|
391 adrne lr, ff_h264_idct8_dc_add_neon + CONFIG_THUMB
|
|
yading@10
|
392 adreq lr, ff_h264_idct8_add_neon + CONFIG_THUMB
|
|
yading@10
|
393 blx lr
|
|
yading@10
|
394 2: subs r12, r12, #4
|
|
yading@10
|
395 add r1, r1, #128
|
|
yading@10
|
396 bne 1b
|
|
yading@10
|
397 pop {r4-r8,pc}
|
|
yading@10
|
398 endfunc
|
|
yading@10
|
399
|
|
yading@10
|
400 const scan8
|
|
yading@10
|
401 .byte 4+ 1*8, 5+ 1*8, 4+ 2*8, 5+ 2*8
|
|
yading@10
|
402 .byte 6+ 1*8, 7+ 1*8, 6+ 2*8, 7+ 2*8
|
|
yading@10
|
403 .byte 4+ 3*8, 5+ 3*8, 4+ 4*8, 5+ 4*8
|
|
yading@10
|
404 .byte 6+ 3*8, 7+ 3*8, 6+ 4*8, 7+ 4*8
|
|
yading@10
|
405 .byte 4+ 6*8, 5+ 6*8, 4+ 7*8, 5+ 7*8
|
|
yading@10
|
406 .byte 6+ 6*8, 7+ 6*8, 6+ 7*8, 7+ 7*8
|
|
yading@10
|
407 .byte 4+ 8*8, 5+ 8*8, 4+ 9*8, 5+ 9*8
|
|
yading@10
|
408 .byte 6+ 8*8, 7+ 8*8, 6+ 9*8, 7+ 9*8
|
|
yading@10
|
409 .byte 4+11*8, 5+11*8, 4+12*8, 5+12*8
|
|
yading@10
|
410 .byte 6+11*8, 7+11*8, 6+12*8, 7+12*8
|
|
yading@10
|
411 .byte 4+13*8, 5+13*8, 4+14*8, 5+14*8
|
|
yading@10
|
412 .byte 6+13*8, 7+13*8, 6+14*8, 7+14*8
|
|
yading@10
|
413 endconst
|
