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
5;*
6;* Copyright (c) 2011 Ronald S. Bultje <rsbultje@gmail.com>
7;*
8;* This file is part of FFmpeg.
9;*
10;* FFmpeg is free software; you can redistribute it and/or
11;* modify it under the terms of the GNU Lesser General Public
12;* License as published by the Free Software Foundation; either
13;* version 2.1 of the License, or (at your option) any later version.
14;*
15;* FFmpeg is distributed in the hope that it will be useful,
16;* but WITHOUT ANY WARRANTY; without even the implied warranty of
17;* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
18;* Lesser General Public License for more details.
19;*
20;* You should have received a copy of the GNU Lesser General Public
21;* License along with FFmpeg; if not, write to the Free Software
22;* 51, Inc., Foundation Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
23;******************************************************************************
24
25%include "libavutil/x86/x86util.asm"
26
27%if ARCH_X86_64
28
29SECTION_RODATA
30
31pw_88:      times 8 dw 0x2008
32cextern pw_1
33cextern pw_4
34cextern pw_1019
35; Below are defined in simple_idct10.asm built from selecting idctdsp
36cextern w4_plus_w2_hi
37cextern w4_min_w2_hi
38cextern w4_plus_w6_hi
39cextern w4_min_w6_hi
40cextern w1_plus_w3_hi
41cextern w3_min_w1_hi
42cextern w7_plus_w3_hi
43cextern w3_min_w7_hi
44cextern w1_plus_w5
45cextern w5_min_w1
46cextern w5_plus_w7
47cextern w7_min_w5
48
49%include "libavcodec/x86/simple_idct10_template.asm"
50
51SECTION .text
52
53define_constants _hi
54
55%macro idct_fn 0
56cglobal prores_idct_put_10, 4, 4, 15, pixels, lsize, block, qmat
57    IDCT_FN    pw_1, 15, pw_88, 18, "put", pw_4, pw_1019, r3
58    RET
59%endmacro
60
61INIT_XMM sse2
62idct_fn
63%if HAVE_AVX_EXTERNAL
64INIT_XMM avx
65idct_fn
66%endif
67
68%endif
69