/*
Copyright (C) 2001-2002 Michael Niedermayer <michaelni@gmx.at>
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program; if not, write to the Free Software
Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
*/
/*
supported Input formats: YV12 (grayscale soon too)
supported output formats: YV12, BGR15, BGR16, BGR24, BGR32 (grayscale soon too)
*/
#include <inttypes.h>
#include <string.h>
#include <math.h>
#include <stdio.h>
#include "../config.h"
#include "../mangle.h"
#ifdef HAVE_MALLOC_H
#include <malloc.h>
#endif
#include "swscale.h"
#include "../cpudetect.h"
#include "../libvo/img_format.h"
#undef MOVNTQ
#undef PAVGB
//#undef HAVE_MMX2
//#define HAVE_3DNOW
//#undef HAVE_MMX
//#undef ARCH_X86
#define DITHER1XBPP
#define RET 0xC3 //near return opcode
#ifdef MP_DEBUG
#define ASSERT(x) if(!(x)) { printf("ASSERT " #x " failed\n"); *((int*)0)=0; }
#else
#define ASSERT(x) ;
#endif
#ifdef M_PI
#define PI M_PI
#else
#define PI 3.14159265358979323846
#endif
extern int verbose; // defined in mplayer.c
/*
NOTES
known BUGS with known cause (no bugreports please!, but patches are welcome :) )
horizontal fast_bilinear MMX2 scaler reads 1-7 samples too much (might cause a sig11)
Supported output formats BGR15 BGR16 BGR24 BGR32 YV12
BGR15 & BGR16 MMX verions support dithering
Special versions: fast Y 1:1 scaling (no interpolation in y direction)
TODO
more intelligent missalignment avoidance for the horizontal scaler
dither in C
change the distance of the u & v buffer
write special vertical cubic upscale version
Optimize C code (yv12 / minmax)
*/
#define ABS(a) ((a) > 0 ? (a) : (-(a)))
#define MIN(a,b) ((a) > (b) ? (b) : (a))
#define MAX(a,b) ((a) < (b) ? (b) : (a))
#ifdef ARCH_X86
#define CAN_COMPILE_X86_ASM
#endif
#ifdef CAN_COMPILE_X86_ASM
static uint64_t __attribute__((aligned(8))) yCoeff= 0x2568256825682568LL;
static uint64_t __attribute__((aligned(8))) vrCoeff= 0x3343334333433343LL;
static uint64_t __attribute__((aligned(8))) ubCoeff= 0x40cf40cf40cf40cfLL;
static uint64_t __attribute__((aligned(8))) vgCoeff= 0xE5E2E5E2E5E2E5E2LL;
static uint64_t __attribute__((aligned(8))) ugCoeff= 0xF36EF36EF36EF36ELL;
static uint64_t __attribute__((aligned(8))) bF8= 0xF8F8F8F8F8F8F8F8LL;
static uint64_t __attribute__((aligned(8))) bFC= 0xFCFCFCFCFCFCFCFCLL;
static uint64_t __attribute__((aligned(8))) w400= 0x0400040004000400LL;
static uint64_t __attribute__((aligned(8))) w80= 0x0080008000800080LL;
static uint64_t __attribute__((aligned(8))) w10= 0x0010001000100010LL;
static uint64_t __attribute__((aligned(8))) w02= 0x0002000200020002LL;
static uint64_t __attribute__((aligned(8))) bm00001111=0x00000000FFFFFFFFLL;
static uint64_t __attribute__((aligned(8))) bm00000111=0x0000000000FFFFFFLL;
static uint64_t __attribute__((aligned(8))) bm11111000=0xFFFFFFFFFF000000LL;
static volatile uint64_t __attribute__((aligned(8))) b5Dither;
static volatile uint64_t __attribute__((aligned(8))) g5Dither;
static volatile uint64_t __attribute__((aligned(8))) g6Dither;
static volatile uint64_t __attribute__((aligned(8))) r5Dither;
static uint64_t __attribute__((aligned(8))) dither4[2]={
0x0103010301030103LL,
0x0200020002000200LL,};
static uint64_t __attribute__((aligned(8))) dither8[2]={
0x0602060206020602LL,
0x0004000400040004LL,};
static uint64_t __attribute__((aligned(8))) b16Mask= 0x001F001F001F001FLL;
static uint64_t __attribute__((aligned(8))) g16Mask= 0x07E007E007E007E0LL;
static uint64_t __attribute__((aligned(8))) r16Mask= 0xF800F800F800F800LL;
static uint64_t __attribute__((aligned(8))) b15Mask= 0x001F001F001F001FLL;
static uint64_t __attribute__((aligned(8))) g15Mask= 0x03E003E003E003E0LL;
static uint64_t __attribute__((aligned(8))) r15Mask= 0x7C007C007C007C00LL;
static uint64_t __attribute__((aligned(8))) M24A= 0x00FF0000FF0000FFLL;
static uint64_t __attribute__((aligned(8))) M24B= 0xFF0000FF0000FF00LL;
static uint64_t __attribute__((aligned(8))) M24C= 0x0000FF0000FF0000LL;
// FIXME remove
static uint64_t __attribute__((aligned(8))) asm_yalpha1;
static uint64_t __attribute__((aligned(8))) asm_uvalpha1;
#endif
// clipping helper table for C implementations:
static unsigned char clip_table[768];
static unsigned short clip_table16b[768];
static unsigned short clip_table16g[768];
static unsigned short clip_table16r[768];
static unsigned short clip_table15b[768];
static unsigned short clip_table15g[768];
static unsigned short clip_table15r[768];
// yuv->rgb conversion tables:
static int yuvtab_2568[256];
static int yuvtab_3343[256];
static int yuvtab_0c92[256];
static i
|