2 * Copyright (c) 2002 Falk Hueffner <falk@debian.org>
4 * This file is part of FFmpeg.
6 * FFmpeg is free software; you can redistribute it and/or
7 * modify it under the terms of the GNU Lesser General Public
8 * License as published by the Free Software Foundation; either
9 * version 2.1 of the License, or (at your option) any later version.
11 * FFmpeg is distributed in the hope that it will be useful,
12 * but WITHOUT ANY WARRANTY; without even the implied warranty of
13 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
14 * Lesser General Public License for more details.
16 * You should have received a copy of the GNU Lesser General Public
17 * License along with FFmpeg; if not, write to the Free Software
18 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
21 #include "libavutil/attributes.h"
22 #include "libavcodec/idctdsp.h"
23 #include "idctdsp_alpha.h"
26 void put_pixels_clamped_mvi_asm(const int16_t *block
, uint8_t *pixels
,
28 void add_pixels_clamped_mvi_asm(const int16_t *block
, uint8_t *pixels
,
31 void (*put_pixels_clamped_axp_p
)(const int16_t *block
, uint8_t *pixels
,
33 void (*add_pixels_clamped_axp_p
)(const int16_t *block
, uint8_t *pixels
,
37 /* These functions were the base for the optimized assembler routines,
38 and remain here for documentation purposes. */
39 static void put_pixels_clamped_mvi(const int16_t *block
, uint8_t *pixels
,
43 uint64_t clampmask
= zap(-1, 0xaa); /* 0x00ff00ff00ff00ff */
46 uint64_t shorts0
, shorts1
;
49 shorts0
= maxsw4(shorts0
, 0);
50 shorts0
= minsw4(shorts0
, clampmask
);
51 stl(pkwb(shorts0
), pixels
);
53 shorts1
= ldq(block
+ 4);
54 shorts1
= maxsw4(shorts1
, 0);
55 shorts1
= minsw4(shorts1
, clampmask
);
56 stl(pkwb(shorts1
), pixels
+ 4);
63 void add_pixels_clamped_mvi(const int16_t *block
, uint8_t *pixels
,
67 /* Keep this function a leaf function by generating the constants
68 manually (mainly for the hack value ;-). */
69 uint64_t clampmask
= zap(-1, 0xaa); /* 0x00ff00ff00ff00ff */
70 uint64_t signmask
= zap(-1, 0x33);
71 signmask
^= signmask
>> 1; /* 0x8000800080008000 */
74 uint64_t shorts0
, pix0
, signs0
;
75 uint64_t shorts1
, pix1
, signs1
;
78 shorts1
= ldq(block
+ 4);
80 pix0
= unpkbw(ldl(pixels
));
81 /* Signed subword add (MMX paddw). */
82 signs0
= shorts0
& signmask
;
87 shorts0
= maxsw4(shorts0
, 0);
88 shorts0
= minsw4(shorts0
, clampmask
);
91 pix1
= unpkbw(ldl(pixels
+ 4));
92 signs1
= shorts1
& signmask
;
96 shorts1
= maxsw4(shorts1
, 0);
97 shorts1
= minsw4(shorts1
, clampmask
);
99 stl(pkwb(shorts0
), pixels
);
100 stl(pkwb(shorts1
), pixels
+ 4);
108 av_cold
void ff_idctdsp_init_alpha(IDCTDSPContext
*c
, AVCodecContext
*avctx
,
109 unsigned high_bit_depth
)
111 /* amask clears all bits that correspond to present features. */
112 if (amask(AMASK_MVI
) == 0) {
113 c
->put_pixels_clamped
= put_pixels_clamped_mvi_asm
;
114 c
->add_pixels_clamped
= add_pixels_clamped_mvi_asm
;
117 put_pixels_clamped_axp_p
= c
->put_pixels_clamped
;
118 add_pixels_clamped_axp_p
= c
->add_pixels_clamped
;
120 if (!high_bit_depth
&& !avctx
->lowres
&&
121 (avctx
->idct_algo
== FF_IDCT_AUTO
||
122 avctx
->idct_algo
== FF_IDCT_SIMPLEALPHA
)) {
123 c
->idct_put
= ff_simple_idct_put_axp
;
124 c
->idct_add
= ff_simple_idct_add_axp
;
125 c
->idct
= ff_simple_idct_axp
;