1 ;******************************************************************************
2 ;* x86-SIMD-optimized IDCT for prores
3 ;* this is identical to "simple" IDCT written by Michael Niedermayer
4 ;* except for the clip range
6 ;* Copyright (c) 2011 Ronald S. Bultje <rsbultje@gmail.com>
7 ;* Copyright (c) 2015 Christophe Gisquet
9 ;* This file is part of FFmpeg.
11 ;* FFmpeg is free software; you can redistribute it and/or
12 ;* modify it under the terms of the GNU Lesser General Public
13 ;* License as published by the Free Software Foundation; either
14 ;* version 2.1 of the License, or (at your option) any later version.
16 ;* FFmpeg is distributed in the hope that it will be useful,
17 ;* but WITHOUT ANY WARRANTY; without even the implied warranty of
18 ;* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
19 ;* Lesser General Public License for more details.
21 ;* You should have received a copy of the GNU Lesser General Public
22 ;* License along with FFmpeg; if not, write to the Free Software
23 ;* 51, Inc., Foundation Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
24 ;******************************************************************************
26 %include "libavutil/x86/x86util.asm"
37 pd_round_11: times 4 dd 1<<(11-1)
38 pd_round_12: times 4 dd 1<<(12-1)
39 pd_round_15: times 4 dd 1<<(15-1)
40 pd_round_19: times 4 dd 1<<(19-1)
41 pd_round_20: times 4 dd 1<<(20-1)
48 %define W1sh2 22725 ; W1 = 90901 = 22725<<2 + 1
49 %define W2sh2 21407 ; W2 = 85627 = 21407<<2 - 1
50 %define W3sh2 19265 ; W3 = 77062 = 19265<<2 + 2
51 %define W4sh2 16384 ; W4 = 65535 = 16384<<2 - 1
52 %define W3sh2_lo 19266
53 %define W4sh2_lo 16383
54 %define W5sh2 12873 ; W5 = 51491 = 12873<<2 - 1
55 %define W6sh2 8867 ; W6 = 35468 = 8867<<2
56 %define W7sh2 4520 ; W7 = 18081 = 4520<<2 + 1
58 CONST_DEC w4_plus_w2_hi, W4sh2, +W2sh2
59 CONST_DEC w4_min_w2_hi, W4sh2, -W2sh2
60 CONST_DEC w4_plus_w6_hi, W4sh2, +W6sh2
61 CONST_DEC w4_min_w6_hi, W4sh2, -W6sh2
62 CONST_DEC w1_plus_w3_hi, W1sh2, +W3sh2
63 CONST_DEC w3_min_w1_hi, W3sh2, -W1sh2
64 CONST_DEC w7_plus_w3_hi, W7sh2, +W3sh2
65 CONST_DEC w3_min_w7_hi, W3sh2, -W7sh2
66 CONST_DEC w1_plus_w5, W1sh2, +W5sh2
67 CONST_DEC w5_min_w1, W5sh2, -W1sh2
68 CONST_DEC w5_plus_w7, W5sh2, +W7sh2
69 CONST_DEC w7_min_w5, W7sh2, -W5sh2
70 CONST_DEC w4_plus_w2_lo, W4sh2_lo, +W2sh2
71 CONST_DEC w4_min_w2_lo, W4sh2_lo, -W2sh2
72 CONST_DEC w4_plus_w6_lo, W4sh2_lo, +W6sh2
73 CONST_DEC w4_min_w6_lo, W4sh2_lo, -W6sh2
74 CONST_DEC w1_plus_w3_lo, W1sh2, +W3sh2_lo
75 CONST_DEC w3_min_w1_lo, W3sh2_lo, -W1sh2
76 CONST_DEC w7_plus_w3_lo, W7sh2, +W3sh2_lo
77 CONST_DEC w3_min_w7_lo, W3sh2_lo, -W7sh2
79 %include "libavcodec/x86/simple_idct10_template.asm"
116 %define PASS4ROWS(base, stride, stride3) \
117 [base], [base + stride], [base + 2*stride], [base + stride3]
123 cglobal simple_idct8, 1, 1, 16, 32, block
124 IDCT_FN "", 11, pw_32, 20, "store"
127 cglobal simple_idct8_put, 3, 4, 16, 32, pixels, lsize, block
128 IDCT_FN "", 11, pw_32, 20
130 lea r2, [pixelsq + r3]
135 STORE_HI_LO PASS8ROWS(pixelsq, r2, lsizeq, r3), m8, m1, m4, m9
138 cglobal simple_idct8_add, 3, 4, 16, 32, pixels, lsize, block
139 IDCT_FN "", 11, pw_32, 20
142 lea r3, [pixelsq + r2]
143 LOAD_ZXBW_8 m3, m5, m6, m7, m12, m13, m14, m15, PASS8ROWS(pixelsq, r3, lsizeq, r2)
154 LOAD_ZXBW_4 m3, m5, m6, m7, PASS4ROWS(pixelsq, lsizeq, r2), m12
159 lea r3, [pixelsq + 4*lsizeq]
160 LOAD_ZXBW_4 m3, m5, m6, m7, PASS4ROWS(r3, lsizeq, r2), m12
165 lea r3, [pixelsq + r2]
171 STORE_HI_LO PASS8ROWS(pixelsq, r3, lsizeq, r2), m8, m1, m4, m9
176 cglobal simple_idct10, 1, 1, 16, block
177 IDCT_FN "", 12, "", 19, "store"
180 cglobal simple_idct10_put, 3, 3, 16, pixels, lsize, block
181 IDCT_FN "", 12, "", 19, "put", 0, pw_1023
184 cglobal simple_idct12, 1, 1, 16, block
185 ; coeffs are already 15bits, adding the offset would cause
186 ; overflow in the input
187 IDCT_FN "", 15, pw_2, 16, "store"
190 cglobal simple_idct12_put, 3, 3, 16, pixels, lsize, block
191 ; range isn't known, so the C simple_idct range is used
192 ; Also, using a bias on input overflows, so use the bias
193 ; on output of the first butterfly instead
194 IDCT_FN "", 15, pw_2, 16, "put", 0, pw_4095
200 %if HAVE_AVX_EXTERNAL