4
* Copyright (c) 2001 Michael Niedermayer <michaelni@gmx.at>
5
* Copyright (c) 2006 Mans Rullgard <mans@mansr.com>
7
* This file is part of FFmpeg.
9
* FFmpeg is free software; you can redistribute it and/or
10
* modify it under the terms of the GNU Lesser General Public
11
* License as published by the Free Software Foundation; either
12
* version 2.1 of the License, or (at your option) any later version.
14
* FFmpeg is distributed in the hope that it will be useful,
15
* but WITHOUT ANY WARRANTY; without even the implied warranty of
16
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
17
* Lesser General Public License for more details.
19
* You should have received a copy of the GNU Lesser General Public
20
* License along with FFmpeg; if not, write to the Free Software
21
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
26
#define W1 22725 /* cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 */
27
#define W2 21407 /* cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 */
28
#define W3 19266 /* cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 */
29
#define W4 16383 /* cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 */
30
#define W5 12873 /* cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 */
31
#define W6 8867 /* cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 */
32
#define W7 4520 /* cos(i*M_PI/16)*sqrt(2)*(1<<14) + 0.5 */
36
#define W13 (W1 | (W3 << 16))
37
#define W26 (W2 | (W6 << 16))
38
#define W57 (W5 | (W7 << 16))
46
function idct_row_armv5te
50
ldrd a3, [a1] /* a3 = row[1:0], a4 = row[3:2] */
56
mov v1, #(1<<(ROW_SHIFT-1))
58
sub ip, ip, #1 /* ip = W4 */
59
smlabb v1, ip, a3, v1 /* v1 = W4*row[0]+(1<<(RS-1)) */
60
ldr ip, w26 /* ip = W2 | (W6 << 16) */
68
ldr ip, w13 /* ip = W1 | (W3 << 16) */
69
ldr lr, w57 /* lr = W5 | (W7 << 16) */
80
ldrd a3, [a1, #8] /* a3=row[5:4] a4=row[7:6] */
96
ldr ip, w26 /* ip = W2 | (W6 << 16) */
98
sub a2, a2, #1 /* a2 = W4 */
99
smulbb a2, a2, a3 /* a2 = W4*row[4] */
100
smultb lr, ip, a4 /* lr = W6*row[6] */
101
add v1, v1, a2 /* v1 += W4*row[4] */
102
add v1, v1, lr /* v1 += W6*row[6] */
103
add v4, v4, a2 /* v4 += W4*row[4] */
104
sub v4, v4, lr /* v4 -= W6*row[6] */
105
smulbb lr, ip, a4 /* lr = W2*row[6] */
106
sub v2, v2, a2 /* v2 -= W4*row[4] */
107
sub v2, v2, lr /* v2 -= W2*row[6] */
108
sub v3, v3, a2 /* v3 -= W4*row[4] */
109
add v3, v3, lr /* v3 += W2*row[6] */
113
bic a3, a3, #0x1f0000
116
add a3, a3, a2, lsl #16
119
bic a4, a4, #0x1f0000
122
add a4, a4, a2, lsl #16
127
bic a3, a3, #0x1f0000
130
add a3, a3, a2, lsl #16
133
bic a4, a4, #0x1f0000
136
add a4, a4, a2, lsl #16
142
orr a3, a3, a3, lsl #16
153
ldr a4, [a1] /* a4 = col[1:0] */
155
sub ip, ip, #1 /* ip = W4 */
157
mov v1, #(1<<(COL_SHIFT-1))
158
smlabt v2, ip, a4, v1 /* v2 = W4*col[1] + (1<<(COL_SHIFT-1)) */
159
smlabb v1, ip, a4, v1 /* v1 = W4*col[0] + (1<<(COL_SHIFT-1)) */
160
ldr a4, [a1, #(16*4)]
162
mov v1, #((1<<(COL_SHIFT-1))/W4) /* this matches the C version */
163
add v2, v1, a4, asr #16
164
rsb v2, v2, v2, lsl #14
166
add v1, v1, a4, asr #16
167
ldr a4, [a1, #(16*4)]
168
rsb v1, v1, v1, lsl #14
181
ldr a4, [a1, #(16*2)]
195
ldr a4, [a1, #(16*6)]
211
stmfd sp!, {v1, v2, v3, v4, v5, v6, v7, fp}
214
ldr a4, [a1, #(16*1)]
225
ldr a4, [a1, #(16*3)]
228
smlatb v1, ip, a4, v1
229
smlatb v3, lr, a4, v3
234
smlatt v2, ip, a4, v2
235
smlatt v4, lr, a4, v4
239
ldr a4, [a1, #(16*5)]
242
smlabb v1, lr, a4, v1
243
smlabb v3, ip, a4, v3
244
smlatb v5, lr, a4, v5
245
smlatb v7, ip, a4, v7
246
smlabt v2, lr, a4, v2
247
smlabt v4, ip, a4, v4
248
smlatt v6, lr, a4, v6
249
ldr a3, [a1, #(16*7)]
250
smlatt fp, ip, a4, fp
252
smlatb v1, lr, a3, v1
253
smlabb v3, lr, a3, v3
254
smlatb v5, ip, a3, v5
256
smlatt v2, lr, a3, v2
258
smlabt v4, lr, a3, v4
260
smlatt v6, ip, a3, v6
264
function idct_col_armv5te
272
orrmi a2, a2, #0xf000
275
orr a2, a2, ip, lsl #16
279
orrmi a2, a2, #0xf000
282
orr a2, a2, a4, lsl #16
284
str a2, [a1, #(16*7)]
288
orrmi a2, a2, #0xf000
291
orr a2, a2, ip, lsl #16
292
str a2, [a1, #(16*1)]
295
orrmi a2, a2, #0xf000
298
orr a2, a2, a4, lsl #16
300
str a2, [a1, #(16*6)]
304
orrmi a2, a2, #0xf000
307
orr a2, a2, ip, lsl #16
308
str a2, [a1, #(16*2)]
311
orrmi a2, a2, #0xf000
314
orr a2, a2, a4, lsl #16
316
str a2, [a1, #(16*5)]
320
orrmi a2, a2, #0xf000
323
orr a2, a2, ip, lsl #16
324
str a2, [a1, #(16*3)]
327
orrmi a2, a2, #0xf000
330
orr a2, a2, a4, lsl #16
331
str a2, [a1, #(16*4)]
336
function idct_col_put_armv5te
353
orr a2, a2, ip, lsl #8
368
orr a2, a3, a4, lsl #8
369
rsb v2, lr, lr, lsl #3
383
orr a2, a2, ip, lsl #8
395
orr a2, a2, a4, lsl #8
409
orr a2, a2, ip, lsl #8
421
orr a2, a2, a4, lsl #8
435
orr a2, a2, ip, lsl #8
447
orr a2, a2, a4, lsl #8
453
function idct_col_add_armv5te
472
adds v1, v1, ip, lsr #8
476
orr a2, a2, v1, lsl #8
479
rsb v2, v1, v1, lsl #3
489
adds a4, a4, ip, lsr #8
495
orr a2, a3, a4, lsl #8
510
adds v3, v3, ip, lsr #8
514
orr a2, a2, v3, lsl #8
525
adds a4, a4, ip, lsr #8
529
orr a2, a3, a4, lsl #8
544
adds v3, v3, ip, lsr #8
548
orr a2, a2, v3, lsl #8
559
adds a4, a4, ip, lsr #8
563
orr a2, a3, a4, lsl #8
578
adds v3, v3, ip, lsr #8
582
orr a2, a2, v3, lsl #8
593
adds a4, a4, ip, lsr #8
597
orr a2, a3, a4, lsl #8
603
function ff_simple_idct_armv5te, export=1
604
stmfd sp!, {v1, v2, v3, v4, v5, v6, v7, fp, lr}
632
ldmfd sp!, {v1, v2, v3, v4, v5, v6, v7, fp, pc}
635
function ff_simple_idct_add_armv5te, export=1
636
stmfd sp!, {a1, a2, v1, v2, v3, v4, v5, v6, v7, fp, lr}
658
bl idct_col_add_armv5te
660
bl idct_col_add_armv5te
662
bl idct_col_add_armv5te
664
bl idct_col_add_armv5te
667
ldmfd sp!, {v1, v2, v3, v4, v5, v6, v7, fp, pc}
670
function ff_simple_idct_put_armv5te, export=1
671
stmfd sp!, {a1, a2, v1, v2, v3, v4, v5, v6, v7, fp, lr}
693
bl idct_col_put_armv5te
695
bl idct_col_put_armv5te
697
bl idct_col_put_armv5te
699
bl idct_col_put_armv5te
702
ldmfd sp!, {v1, v2, v3, v4, v5, v6, v7, fp, pc}