Skip to content

F
ffmpeg.wasm-core
Commit d1adad3c authored by Michael Niedermayer's avatar Michael Niedermayer
Browse files
Options

Merge swscale bloatup 

This will be cleaned up in the next merge

Authorship / merged commits:
commit f668afd4
Author: Janne Grunau <janne-libav@jannau.net>
Date:   Fri Apr 15 09:12:34 2011 +0200

    swscale: fix "ISO C90 forbids mixed declarations and code" warning

    only hit with --enable-runtime-cpudetect

commit 7f2ae5c7
Author: Janne Grunau <janne-libav@jannau.net>
Date:   Fri Apr 15 02:09:44 2011 +0200

    swscale: fix compilation with --enable-runtime-cpudetect

commit b6cad3df
Author: Janne Grunau <janne-libav@jannau.net>
Date:   Fri Apr 15 00:31:04 2011 +0200

    swscale: correct include path to fix ppc altivec build

commit 6216fc70
Author: Luca Barbato <lu_zero@gentoo.org>
Date:   Thu Apr 14 22:03:45 2011 +0200

    swscale: simplify rgb2rgb templating

    MMX is always built. Drop the ifdefs

commit 33a0421b
Author: Josh Allmann <joshua.allmann@gmail.com>
Date:   Wed Apr 13 20:57:32 2011 +0200

    swscale: simplify initialization code

    Simplify the fallthrough case when no accelerated functions
    can be initialized.

commit 735bf195
Author: Josh Allmann <joshua.allmann@gmail.com>
Date:   Wed Apr 13 20:57:31 2011 +0200

    swscale: further cleanup swscale.c

    Move x86-specific constants out of swscale.c

commit 86330b4c
Author: Luca Barbato <lu_zero@gentoo.org>
Date:   Wed Apr 13 20:57:30 2011 +0200

    swscale: partially move the arch specific code left

    PPC and x86 code is split off from swscale_template.c. Lots of code is
    still duplicated and should be removed later.

    Again uniformize the init system to be more similar to the dsputil one.

    Unset h*scale_fast in the x86 init in order to make the output
    consistent with the previous status. Thanks to Josh for spotting it.

commit c0038328
Author: Luca Barbato <lu_zero@gentoo.org>
Date:   Wed Apr 13 20:57:29 2011 +0200

    swscale: move away x86 specific code from rgb2rgb

    Keep only the plain C code in the main rgb2rgb.c and move the x86
    specific optimizations to x86/rgb2rgb.c
    Change the initialization pattern a little so some of it can be
    factorized to behave more like dsputils.

Conflicts:
	libswscale/rgb2rgb.c
	libswscale/swscale_template.c
parent d9d56036
Expand all Hide whitespace changes
Inline Side-by-side
Showing with 6916 additions and 4608 deletions
libswscale/Makefile
View file @ d1adad3c
......@@ -12,7 +12,8 @@ OBJS-$(ARCH_BFIN) += bfin/internal_bfin.o \
bfin/yuv2rgb_bfin.o
OBJS-$(CONFIG_MLIB) += mlib/yuv2rgb_mlib.o
OBJS-$(HAVE_ALTIVEC) += ppc/yuv2rgb_altivec.o
OBJS-$(HAVE_MMX) += x86/yuv2rgb_mmx.o
OBJS-$(HAVE_MMX) += x86/rgb2rgb.o \
x86/yuv2rgb_mmx.o
OBJS-$(HAVE_VIS) += sparc/yuv2rgb_vis.o
TESTPROGS = colorspace swscale
......
libswscale/ppc/swscale_template.c 0 → 100644
View file @ d1adad3c
This diff is collapsed.
libswscale/rgb2rgb.c
View file @ d1adad3c
......@@ -24,7 +24,6 @@
*/
#include <inttypes.h>
#include "config.h"
#include "libavutil/x86_cpu.h"
#include "libavutil/bswap.h"
#include "rgb2rgb.h"
#include "swscale.h"
......@@ -95,45 +94,6 @@ void (*yuyvtoyuv422)(uint8_t *ydst, uint8_t *udst, uint8_t *vdst, const uint8_t
long width, long height,
long lumStride, long chromStride, long srcStride);
#if ARCH_X86
DECLARE_ASM_CONST(8, uint64_t, mmx_ff) = 0x00000000000000FFULL;
DECLARE_ASM_CONST(8, uint64_t, mmx_null) = 0x0000000000000000ULL;
DECLARE_ASM_CONST(8, uint64_t, mmx_one) = 0xFFFFFFFFFFFFFFFFULL;
DECLARE_ASM_CONST(8, uint64_t, mask32b) = 0x000000FF000000FFULL;
DECLARE_ASM_CONST(8, uint64_t, mask32g) = 0x0000FF000000FF00ULL;
DECLARE_ASM_CONST(8, uint64_t, mask32r) = 0x00FF000000FF0000ULL;
DECLARE_ASM_CONST(8, uint64_t, mask32a) = 0xFF000000FF000000ULL;
DECLARE_ASM_CONST(8, uint64_t, mask32) = 0x00FFFFFF00FFFFFFULL;
DECLARE_ASM_CONST(8, uint64_t, mask3216br) = 0x00F800F800F800F8ULL;
DECLARE_ASM_CONST(8, uint64_t, mask3216g) = 0x0000FC000000FC00ULL;
DECLARE_ASM_CONST(8, uint64_t, mask3215g) = 0x0000F8000000F800ULL;
DECLARE_ASM_CONST(8, uint64_t, mul3216) = 0x2000000420000004ULL;
DECLARE_ASM_CONST(8, uint64_t, mul3215) = 0x2000000820000008ULL;
DECLARE_ASM_CONST(8, uint64_t, mask24b) = 0x00FF0000FF0000FFULL;
DECLARE_ASM_CONST(8, uint64_t, mask24g) = 0xFF0000FF0000FF00ULL;
DECLARE_ASM_CONST(8, uint64_t, mask24r) = 0x0000FF0000FF0000ULL;
DECLARE_ASM_CONST(8, uint64_t, mask24l) = 0x0000000000FFFFFFULL;
DECLARE_ASM_CONST(8, uint64_t, mask24h) = 0x0000FFFFFF000000ULL;
DECLARE_ASM_CONST(8, uint64_t, mask24hh) = 0xffff000000000000ULL;
DECLARE_ASM_CONST(8, uint64_t, mask24hhh) = 0xffffffff00000000ULL;
DECLARE_ASM_CONST(8, uint64_t, mask24hhhh) = 0xffffffffffff0000ULL;
DECLARE_ASM_CONST(8, uint64_t, mask15b) = 0x001F001F001F001FULL; /* 00000000 00011111 xxB */
DECLARE_ASM_CONST(8, uint64_t, mask15rg) = 0x7FE07FE07FE07FE0ULL; /* 01111111 11100000 RGx */
DECLARE_ASM_CONST(8, uint64_t, mask15s) = 0xFFE0FFE0FFE0FFE0ULL;
DECLARE_ASM_CONST(8, uint64_t, mask15g) = 0x03E003E003E003E0ULL;
DECLARE_ASM_CONST(8, uint64_t, mask15r) = 0x7C007C007C007C00ULL;
#define mask16b mask15b
DECLARE_ASM_CONST(8, uint64_t, mask16g) = 0x07E007E007E007E0ULL;
DECLARE_ASM_CONST(8, uint64_t, mask16r) = 0xF800F800F800F800ULL;
DECLARE_ASM_CONST(8, uint64_t, red_16mask) = 0x0000f8000000f800ULL;
DECLARE_ASM_CONST(8, uint64_t, green_16mask) = 0x000007e0000007e0ULL;
DECLARE_ASM_CONST(8, uint64_t, blue_16mask) = 0x0000001f0000001fULL;
DECLARE_ASM_CONST(8, uint64_t, red_15mask) = 0x00007c0000007c00ULL;
DECLARE_ASM_CONST(8, uint64_t, green_15mask) = 0x000003e0000003e0ULL;
DECLARE_ASM_CONST(8, uint64_t, blue_15mask) = 0x0000001f0000001fULL;
#endif /* ARCH_X86 */
#define RGB2YUV_SHIFT 8
#define BY ((int)( 0.098*(1<<RGB2YUV_SHIFT)+0.5))
#define BV ((int)(-0.071*(1<<RGB2YUV_SHIFT)+0.5))
......@@ -145,50 +105,9 @@ DECLARE_ASM_CONST(8, uint64_t, blue_15mask) = 0x0000001f0000001fULL;
#define RV ((int)( 0.439*(1<<RGB2YUV_SHIFT)+0.5))
#define RU ((int)(-0.148*(1<<RGB2YUV_SHIFT)+0.5))
//Note: We have C, MMX, MMX2, 3DNOW versions, there is no 3DNOW + MMX2 one.
//plain C versions
#define COMPILE_TEMPLATE_MMX 0
#define COMPILE_TEMPLATE_MMX2 0
#define COMPILE_TEMPLATE_AMD3DNOW 0
#define COMPILE_TEMPLATE_SSE2 0
#define RENAME(a) a ## _C
#include "rgb2rgb_template.c"
#if ARCH_X86
//MMX versions
#undef RENAME
#undef COMPILE_TEMPLATE_MMX
#define COMPILE_TEMPLATE_MMX 1
#define RENAME(a) a ## _MMX
#include "rgb2rgb_template.c"
//MMX2 versions
#undef RENAME
#undef COMPILE_TEMPLATE_MMX2
#define COMPILE_TEMPLATE_MMX2 1
#define RENAME(a) a ## _MMX2
#include "rgb2rgb_template.c"
//SSE2 versions
#undef RENAME
#undef COMPILE_TEMPLATE_SSE2
#define COMPILE_TEMPLATE_SSE2 1
#define RENAME(a) a ## _SSE2
#include "rgb2rgb_template.c"
//3DNOW versions
#undef RENAME
#undef COMPILE_TEMPLATE_MMX2
#undef COMPILE_TEMPLATE_SSE2
#undef COMPILE_TEMPLATE_AMD3DNOW
#define COMPILE_TEMPLATE_MMX2 0
#define COMPILE_TEMPLATE_SSE2 1
#define COMPILE_TEMPLATE_AMD3DNOW 1
#define RENAME(a) a ## _3DNOW
#include "rgb2rgb_template.c"
#endif //ARCH_X86 || ARCH_X86_64
/*
RGB15->RGB16 original by Strepto/Astral
......@@ -199,18 +118,10 @@ DECLARE_ASM_CONST(8, uint64_t, blue_15mask) = 0x0000001f0000001fULL;
void sws_rgb2rgb_init(int flags)
{
#if HAVE_SSE2 || HAVE_MMX2 || HAVE_AMD3DNOW || HAVE_MMX
if (flags & SWS_CPU_CAPS_SSE2)
rgb2rgb_init_SSE2();
else if (flags & SWS_CPU_CAPS_MMX2)
rgb2rgb_init_MMX2();
else if (flags & SWS_CPU_CAPS_3DNOW)
rgb2rgb_init_3DNOW();
else if (flags & SWS_CPU_CAPS_MMX)
rgb2rgb_init_MMX();
else
rgb2rgb_init_c();
#if HAVE_MMX2 || HAVE_AMD3DNOW || HAVE_MMX
rgb2rgb_init_x86(flags);
#endif /* HAVE_MMX2 || HAVE_AMD3DNOW || HAVE_MMX */
rgb2rgb_init_C();
}
#if LIBSWSCALE_VERSION_MAJOR < 1
......
libswscale/rgb2rgb.h
View file @ d1adad3c
......@@ -168,4 +168,6 @@ extern void (*yuyvtoyuv422)(uint8_t *ydst, uint8_t *udst, uint8_t *vdst, const u
void sws_rgb2rgb_init(int flags);
void rgb2rgb_init_x86(int flags);
#endif /* SWSCALE_RGB2RGB_H */
libswscale/rgb2rgb_template.c
View file @ d1adad3c
This source diff could not be displayed because it is too large. You can view the blob instead.
libswscale/swscale.c
View file @ d1adad3c
......@@ -122,63 +122,6 @@ add BGR4 output support
write special BGR->BGR scaler
*/
#if ARCH_X86
DECLARE_ASM_CONST(8, uint64_t, bF8)= 0xF8F8F8F8F8F8F8F8LL;
DECLARE_ASM_CONST(8, uint64_t, bFC)= 0xFCFCFCFCFCFCFCFCLL;
DECLARE_ASM_CONST(8, uint64_t, w10)= 0x0010001000100010LL;
DECLARE_ASM_CONST(8, uint64_t, w02)= 0x0002000200020002LL;
DECLARE_ASM_CONST(8, uint64_t, bm00001111)=0x00000000FFFFFFFFLL;
DECLARE_ASM_CONST(8, uint64_t, bm00000111)=0x0000000000FFFFFFLL;
DECLARE_ASM_CONST(8, uint64_t, bm11111000)=0xFFFFFFFFFF000000LL;
DECLARE_ASM_CONST(8, uint64_t, bm01010101)=0x00FF00FF00FF00FFLL;
const DECLARE_ALIGNED(8, uint64_t, ff_dither4)[2] = {
0x0103010301030103LL,
0x0200020002000200LL,};
const DECLARE_ALIGNED(8, uint64_t, ff_dither8)[2] = {
0x0602060206020602LL,
0x0004000400040004LL,};
DECLARE_ASM_CONST(8, uint64_t, b16Mask)= 0x001F001F001F001FLL;
DECLARE_ASM_CONST(8, uint64_t, g16Mask)= 0x07E007E007E007E0LL;
DECLARE_ASM_CONST(8, uint64_t, r16Mask)= 0xF800F800F800F800LL;
DECLARE_ASM_CONST(8, uint64_t, b15Mask)= 0x001F001F001F001FLL;
DECLARE_ASM_CONST(8, uint64_t, g15Mask)= 0x03E003E003E003E0LL;
DECLARE_ASM_CONST(8, uint64_t, r15Mask)= 0x7C007C007C007C00LL;
DECLARE_ALIGNED(8, const uint64_t, ff_M24A) = 0x00FF0000FF0000FFLL;
DECLARE_ALIGNED(8, const uint64_t, ff_M24B) = 0xFF0000FF0000FF00LL;
DECLARE_ALIGNED(8, const uint64_t, ff_M24C) = 0x0000FF0000FF0000LL;
#ifdef FAST_BGR2YV12
DECLARE_ALIGNED(8, const uint64_t, ff_bgr2YCoeff) = 0x000000210041000DULL;
DECLARE_ALIGNED(8, const uint64_t, ff_bgr2UCoeff) = 0x0000FFEEFFDC0038ULL;
DECLARE_ALIGNED(8, const uint64_t, ff_bgr2VCoeff) = 0x00000038FFD2FFF8ULL;
#else
DECLARE_ALIGNED(8, const uint64_t, ff_bgr2YCoeff) = 0x000020E540830C8BULL;
DECLARE_ALIGNED(8, const uint64_t, ff_bgr2UCoeff) = 0x0000ED0FDAC23831ULL;
DECLARE_ALIGNED(8, const uint64_t, ff_bgr2VCoeff) = 0x00003831D0E6F6EAULL;
#endif /* FAST_BGR2YV12 */
DECLARE_ALIGNED(8, const uint64_t, ff_bgr2YOffset) = 0x1010101010101010ULL;
DECLARE_ALIGNED(8, const uint64_t, ff_bgr2UVOffset) = 0x8080808080808080ULL;
DECLARE_ALIGNED(8, const uint64_t, ff_w1111) = 0x0001000100010001ULL;
DECLARE_ASM_CONST(8, uint64_t, ff_bgr24toY1Coeff) = 0x0C88000040870C88ULL;
DECLARE_ASM_CONST(8, uint64_t, ff_bgr24toY2Coeff) = 0x20DE4087000020DEULL;
DECLARE_ASM_CONST(8, uint64_t, ff_rgb24toY1Coeff) = 0x20DE0000408720DEULL;
DECLARE_ASM_CONST(8, uint64_t, ff_rgb24toY2Coeff) = 0x0C88408700000C88ULL;
DECLARE_ASM_CONST(8, uint64_t, ff_bgr24toYOffset) = 0x0008400000084000ULL;
DECLARE_ASM_CONST(8, uint64_t, ff_bgr24toUV)[2][4] = {
{0x38380000DAC83838ULL, 0xECFFDAC80000ECFFULL, 0xF6E40000D0E3F6E4ULL, 0x3838D0E300003838ULL},
{0xECFF0000DAC8ECFFULL, 0x3838DAC800003838ULL, 0x38380000D0E33838ULL, 0xF6E4D0E30000F6E4ULL},
};
DECLARE_ASM_CONST(8, uint64_t, ff_bgr24toUVOffset)= 0x0040400000404000ULL;
#endif /* ARCH_X86 */
DECLARE_ALIGNED(8, static const uint8_t, dither_2x2_4)[2][8]={
{ 1, 3, 1, 3, 1, 3, 1, 3, },
{ 2, 0, 2, 0, 2, 0, 2, 0, },
......@@ -1367,17 +1310,14 @@ static inline void monoblack2Y(uint8_t *dst, const uint8_t *src, long width, uin
#define COMPILE_TEMPLATE_AMD3DNOW 0
#define COMPILE_TEMPLATE_ALTIVEC 0
#if COMPILE_C
#define RENAME(a) a ## _C
#include "swscale_template.c"
#endif
#if COMPILE_ALTIVEC
#undef RENAME
#undef COMPILE_TEMPLATE_ALTIVEC
#define COMPILE_TEMPLATE_ALTIVEC 1
#define RENAME(a) a ## _altivec
#include "swscale_template.c"
#include "ppc/swscale_template.c"
#endif
#if ARCH_X86
......@@ -1392,7 +1332,7 @@ static inline void monoblack2Y(uint8_t *dst, const uint8_t *src, long width, uin
#define COMPILE_TEMPLATE_MMX2 0
#define COMPILE_TEMPLATE_AMD3DNOW 0
#define RENAME(a) a ## _MMX
#include "swscale_template.c"
#include "x86/swscale_template.c"
#endif
//MMX2 versions
......@@ -1405,7 +1345,7 @@ static inline void monoblack2Y(uint8_t *dst, const uint8_t *src, long width, uin
#define COMPILE_TEMPLATE_MMX2 1
#define COMPILE_TEMPLATE_AMD3DNOW 0
#define RENAME(a) a ## _MMX2
#include "swscale_template.c"
#include "x86/swscale_template.c"
#endif
//3DNOW versions
......@@ -1418,44 +1358,36 @@ static inline void monoblack2Y(uint8_t *dst, const uint8_t *src, long width, uin
#define COMPILE_TEMPLATE_MMX2 0
#define COMPILE_TEMPLATE_AMD3DNOW 1
#define RENAME(a) a ## _3DNow
#include "swscale_template.c"
#include "x86/swscale_template.c"
#endif
#endif //ARCH_X86
SwsFunc ff_getSwsFunc(SwsContext *c)
{
#if CONFIG_RUNTIME_CPUDETECT
int flags = c->flags;
sws_init_swScale_c(c);
#if CONFIG_RUNTIME_CPUDETECT
#if ARCH_X86
// ordered per speed fastest first
if (flags & SWS_CPU_CAPS_MMX2) {
if (c->flags & SWS_CPU_CAPS_MMX2) {
sws_init_swScale_MMX2(c);
return swScale_MMX2;
} else if (flags & SWS_CPU_CAPS_3DNOW) {
} else if (c->flags & SWS_CPU_CAPS_3DNOW) {
sws_init_swScale_3DNow(c);
return swScale_3DNow;
} else if (flags & SWS_CPU_CAPS_MMX) {
} else if (c->flags & SWS_CPU_CAPS_MMX) {
sws_init_swScale_MMX(c);
return swScale_MMX;
} else {
sws_init_swScale_C(c);
return swScale_C;
}
#else
#if COMPILE_ALTIVEC
if (flags & SWS_CPU_CAPS_ALTIVEC) {
if (c->flags & SWS_CPU_CAPS_ALTIVEC) {
sws_init_swScale_altivec(c);
return swScale_altivec;
} else {
sws_init_swScale_C(c);
return swScale_C;
}
#endif
sws_init_swScale_C(c);
return swScale_C;
#endif /* ARCH_X86 */
#else //CONFIG_RUNTIME_CPUDETECT
#if COMPILE_TEMPLATE_MMX2
......@@ -1470,11 +1402,10 @@ SwsFunc ff_getSwsFunc(SwsContext *c)
#elif COMPILE_TEMPLATE_ALTIVEC
sws_init_swScale_altivec(c);
return swScale_altivec;
#else
sws_init_swScale_C(c);
return swScale_C;
#endif
#endif //!CONFIG_RUNTIME_CPUDETECT
return swScale_c;
}
static void copyPlane(const uint8_t *src, int srcStride,
......
libswscale/swscale_template.c
View file @ d1adad3c
This source diff could not be displayed because it is too large. You can view the blob instead.
libswscale/x86/rgb2rgb.c 0 → 100644
View file @ d1adad3c
/*
* software RGB to RGB converter
* pluralize by software PAL8 to RGB converter
* software YUV to YUV converter
* software YUV to RGB converter
* Written by Nick Kurshev.
* palette & YUV & runtime CPU stuff by Michael (michaelni@gmx.at)
*
* This file is part of FFmpeg.
*
* FFmpeg is free software; you can redistribute it and/or
* modify it under the terms of the GNU Lesser General Public
* License as published by the Free Software Foundation; either
* version 2.1 of the License, or (at your option) any later version.
*
* FFmpeg 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
* Lesser General Public License for more details.
*
* You should have received a copy of the GNU Lesser General Public
* License along with FFmpeg; if not, write to the Free Software
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
*/
#include <stdint.h>
#include "config.h"
#include "libavutil/x86_cpu.h"
#include "libavutil/bswap.h"
#include "libswscale/rgb2rgb.h"
#include "libswscale/swscale.h"
#include "libswscale/swscale_internal.h"
DECLARE_ASM_CONST(8, uint64_t, mmx_ff) = 0x00000000000000FFULL;
DECLARE_ASM_CONST(8, uint64_t, mmx_null) = 0x0000000000000000ULL;
DECLARE_ASM_CONST(8, uint64_t, mmx_one) = 0xFFFFFFFFFFFFFFFFULL;
DECLARE_ASM_CONST(8, uint64_t, mask32b) = 0x000000FF000000FFULL;
DECLARE_ASM_CONST(8, uint64_t, mask32g) = 0x0000FF000000FF00ULL;
DECLARE_ASM_CONST(8, uint64_t, mask32r) = 0x00FF000000FF0000ULL;
DECLARE_ASM_CONST(8, uint64_t, mask32a) = 0xFF000000FF000000ULL;
DECLARE_ASM_CONST(8, uint64_t, mask32) = 0x00FFFFFF00FFFFFFULL;
DECLARE_ASM_CONST(8, uint64_t, mask3216br) = 0x00F800F800F800F8ULL;
DECLARE_ASM_CONST(8, uint64_t, mask3216g) = 0x0000FC000000FC00ULL;
DECLARE_ASM_CONST(8, uint64_t, mask3215g) = 0x0000F8000000F800ULL;
DECLARE_ASM_CONST(8, uint64_t, mul3216) = 0x2000000420000004ULL;
DECLARE_ASM_CONST(8, uint64_t, mul3215) = 0x2000000820000008ULL;
DECLARE_ASM_CONST(8, uint64_t, mask24b) = 0x00FF0000FF0000FFULL;
DECLARE_ASM_CONST(8, uint64_t, mask24g) = 0xFF0000FF0000FF00ULL;
DECLARE_ASM_CONST(8, uint64_t, mask24r) = 0x0000FF0000FF0000ULL;
DECLARE_ASM_CONST(8, uint64_t, mask24l) = 0x0000000000FFFFFFULL;
DECLARE_ASM_CONST(8, uint64_t, mask24h) = 0x0000FFFFFF000000ULL;
DECLARE_ASM_CONST(8, uint64_t, mask24hh) = 0xffff000000000000ULL;
DECLARE_ASM_CONST(8, uint64_t, mask24hhh) = 0xffffffff00000000ULL;
DECLARE_ASM_CONST(8, uint64_t, mask24hhhh) = 0xffffffffffff0000ULL;
DECLARE_ASM_CONST(8, uint64_t, mask15b) = 0x001F001F001F001FULL; /* 00000000 00011111 xxB */
DECLARE_ASM_CONST(8, uint64_t, mask15rg) = 0x7FE07FE07FE07FE0ULL; /* 01111111 11100000 RGx */
DECLARE_ASM_CONST(8, uint64_t, mask15s) = 0xFFE0FFE0FFE0FFE0ULL;
DECLARE_ASM_CONST(8, uint64_t, mask15g) = 0x03E003E003E003E0ULL;
DECLARE_ASM_CONST(8, uint64_t, mask15r) = 0x7C007C007C007C00ULL;
#define mask16b mask15b
DECLARE_ASM_CONST(8, uint64_t, mask16g) = 0x07E007E007E007E0ULL;
DECLARE_ASM_CONST(8, uint64_t, mask16r) = 0xF800F800F800F800ULL;
DECLARE_ASM_CONST(8, uint64_t, red_16mask) = 0x0000f8000000f800ULL;
DECLARE_ASM_CONST(8, uint64_t, green_16mask) = 0x000007e0000007e0ULL;
DECLARE_ASM_CONST(8, uint64_t, blue_16mask) = 0x0000001f0000001fULL;
DECLARE_ASM_CONST(8, uint64_t, red_15mask) = 0x00007c0000007c00ULL;
DECLARE_ASM_CONST(8, uint64_t, green_15mask) = 0x000003e0000003e0ULL;
DECLARE_ASM_CONST(8, uint64_t, blue_15mask) = 0x0000001f0000001fULL;
#define RGB2YUV_SHIFT 8
#define BY ((int)( 0.098*(1<<RGB2YUV_SHIFT)+0.5))
#define BV ((int)(-0.071*(1<<RGB2YUV_SHIFT)+0.5))
#define BU ((int)( 0.439*(1<<RGB2YUV_SHIFT)+0.5))
#define GY ((int)( 0.504*(1<<RGB2YUV_SHIFT)+0.5))
#define GV ((int)(-0.368*(1<<RGB2YUV_SHIFT)+0.5))
#define GU ((int)(-0.291*(1<<RGB2YUV_SHIFT)+0.5))
#define RY ((int)( 0.257*(1<<RGB2YUV_SHIFT)+0.5))
#define RV ((int)( 0.439*(1<<RGB2YUV_SHIFT)+0.5))
#define RU ((int)(-0.148*(1<<RGB2YUV_SHIFT)+0.5))
//Note: We have C, MMX, MMX2, 3DNOW versions, there is no 3DNOW + MMX2 one.
#define COMPILE_TEMPLATE_MMX2 0
#define COMPILE_TEMPLATE_AMD3DNOW 0
#define COMPILE_TEMPLATE_SSE2 0
//MMX versions
#undef RENAME
#define RENAME(a) a ## _MMX
#include "rgb2rgb_template.c"
//MMX2 versions
#undef RENAME
#undef COMPILE_TEMPLATE_MMX2
#define COMPILE_TEMPLATE_MMX2 1
#define RENAME(a) a ## _MMX2
#include "rgb2rgb_template.c"
//SSE2 versions
#undef RENAME
#undef COMPILE_TEMPLATE_SSE2
#define COMPILE_TEMPLATE_SSE2 1
#define RENAME(a) a ## _SSE2
#include "rgb2rgb_template.c"
//3DNOW versions
#undef RENAME
#undef COMPILE_TEMPLATE_MMX2
#undef COMPILE_TEMPLATE_SSE2
#undef COMPILE_TEMPLATE_AMD3DNOW
#define COMPILE_TEMPLATE_MMX2 0
#define COMPILE_TEMPLATE_SSE2 1
#define COMPILE_TEMPLATE_AMD3DNOW 1
#define RENAME(a) a ## _3DNOW
#include "rgb2rgb_template.c"
/*
RGB15->RGB16 original by Strepto/Astral
ported to gcc & bugfixed : A'rpi
MMX2, 3DNOW optimization by Nick Kurshev
32-bit C version, and and&add trick by Michael Niedermayer
*/
void rgb2rgb_init_x86(int flags)
{
#if HAVE_MMX2 || HAVE_AMD3DNOW || HAVE_MMX
if (flags & SWS_CPU_CAPS_SSE2)
rgb2rgb_init_SSE2();
else if (flags & SWS_CPU_CAPS_MMX2)
rgb2rgb_init_MMX2();
else if (flags & SWS_CPU_CAPS_3DNOW)
rgb2rgb_init_3DNOW();
else if (flags & SWS_CPU_CAPS_MMX)
rgb2rgb_init_MMX();
#endif /* HAVE_MMX2 || HAVE_AMD3DNOW || HAVE_MMX */
}
libswscale/x86/rgb2rgb_template.c 0 → 100644
View file @ d1adad3c
This source diff could not be displayed because it is too large. You can view the blob instead.
libswscale/x86/swscale_template.c 0 → 100644
View file @ d1adad3c
This source diff could not be displayed because it is too large. You can view the blob instead.
libswscale/x86/swscale_template.h 0 → 100644
View file @ d1adad3c
/*
* Copyright (C) 2001-2003 Michael Niedermayer <michaelni@gmx.at>
*
* This file is part of FFmpeg.
*
* FFmpeg is free software; you can redistribute it and/or
* modify it under the terms of the GNU Lesser General Public
* License as published by the Free Software Foundation; either
* version 2.1 of the License, or (at your option) any later version.
*
* FFmpeg 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
* Lesser General Public License for more details.
*
* You should have received a copy of the GNU Lesser General Public
* License along with FFmpeg; if not, write to the Free Software
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
*/
#ifndef SWSCALE_X86_SWSCALE_TEMPLATE_H
#define SWSCALE_X86_SWSCALE_TEMPLATE_H
DECLARE_ASM_CONST(8, uint64_t, bF8)= 0xF8F8F8F8F8F8F8F8LL;
DECLARE_ASM_CONST(8, uint64_t, bFC)= 0xFCFCFCFCFCFCFCFCLL;
DECLARE_ASM_CONST(8, uint64_t, w10)= 0x0010001000100010LL;
DECLARE_ASM_CONST(8, uint64_t, w02)= 0x0002000200020002LL;
DECLARE_ASM_CONST(8, uint64_t, bm00001111)=0x00000000FFFFFFFFLL;
DECLARE_ASM_CONST(8, uint64_t, bm00000111)=0x0000000000FFFFFFLL;
DECLARE_ASM_CONST(8, uint64_t, bm11111000)=0xFFFFFFFFFF000000LL;
DECLARE_ASM_CONST(8, uint64_t, bm01010101)=0x00FF00FF00FF00FFLL;
const DECLARE_ALIGNED(8, uint64_t, ff_dither4)[2] = {
0x0103010301030103LL,
0x0200020002000200LL,};
const DECLARE_ALIGNED(8, uint64_t, ff_dither8)[2] = {
0x0602060206020602LL,
0x0004000400040004LL,};
DECLARE_ASM_CONST(8, uint64_t, b16Mask)= 0x001F001F001F001FLL;
DECLARE_ASM_CONST(8, uint64_t, g16Mask)= 0x07E007E007E007E0LL;
DECLARE_ASM_CONST(8, uint64_t, r16Mask)= 0xF800F800F800F800LL;
DECLARE_ASM_CONST(8, uint64_t, b15Mask)= 0x001F001F001F001FLL;
DECLARE_ASM_CONST(8, uint64_t, g15Mask)= 0x03E003E003E003E0LL;
DECLARE_ASM_CONST(8, uint64_t, r15Mask)= 0x7C007C007C007C00LL;
DECLARE_ALIGNED(8, const uint64_t, ff_M24A) = 0x00FF0000FF0000FFLL;
DECLARE_ALIGNED(8, const uint64_t, ff_M24B) = 0xFF0000FF0000FF00LL;
DECLARE_ALIGNED(8, const uint64_t, ff_M24C) = 0x0000FF0000FF0000LL;
#ifdef FAST_BGR2YV12
DECLARE_ALIGNED(8, const uint64_t, ff_bgr2YCoeff) = 0x000000210041000DULL;
DECLARE_ALIGNED(8, const uint64_t, ff_bgr2UCoeff) = 0x0000FFEEFFDC0038ULL;
DECLARE_ALIGNED(8, const uint64_t, ff_bgr2VCoeff) = 0x00000038FFD2FFF8ULL;
#else
DECLARE_ALIGNED(8, const uint64_t, ff_bgr2YCoeff) = 0x000020E540830C8BULL;
DECLARE_ALIGNED(8, const uint64_t, ff_bgr2UCoeff) = 0x0000ED0FDAC23831ULL;
DECLARE_ALIGNED(8, const uint64_t, ff_bgr2VCoeff) = 0x00003831D0E6F6EAULL;
#endif /* FAST_BGR2YV12 */
DECLARE_ALIGNED(8, const uint64_t, ff_bgr2YOffset) = 0x1010101010101010ULL;
DECLARE_ALIGNED(8, const uint64_t, ff_bgr2UVOffset) = 0x8080808080808080ULL;
DECLARE_ALIGNED(8, const uint64_t, ff_w1111) = 0x0001000100010001ULL;
DECLARE_ASM_CONST(8, uint64_t, ff_bgr24toY1Coeff) = 0x0C88000040870C88ULL;
DECLARE_ASM_CONST(8, uint64_t, ff_bgr24toY2Coeff) = 0x20DE4087000020DEULL;
DECLARE_ASM_CONST(8, uint64_t, ff_rgb24toY1Coeff) = 0x20DE0000408720DEULL;
DECLARE_ASM_CONST(8, uint64_t, ff_rgb24toY2Coeff) = 0x0C88408700000C88ULL;
DECLARE_ASM_CONST(8, uint64_t, ff_bgr24toYOffset) = 0x0008400000084000ULL;
DECLARE_ASM_CONST(8, uint64_t, ff_bgr24toUV)[2][4] = {
{0x38380000DAC83838ULL, 0xECFFDAC80000ECFFULL, 0xF6E40000D0E3F6E4ULL, 0x3838D0E300003838ULL},
{0xECFF0000DAC8ECFFULL, 0x3838DAC800003838ULL, 0x38380000D0E33838ULL, 0xF6E4D0E30000F6E4ULL},
};
DECLARE_ASM_CONST(8, uint64_t, ff_bgr24toUVOffset)= 0x0040400000404000ULL;
#endif /* SWSCALE_X86_SWSCALE_TEMPLATE_H */
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment