SDL_yuv_mmx.c
资源名称:NETVIDEO.rar [点击查看]
上传用户:sun1608
上传日期:2007-02-02
资源大小:6116k
文件大小:20k
源码类别:
流媒体/Mpeg4/MP4
开发平台:
Visual C++
- /*
- SDL - Simple DirectMedia Layer
- Copyright (C) 1997, 1998, 1999, 2000, 2001, 2002 Sam Lantinga
- This library is free software; you can redistribute it and/or
- modify it under the terms of the GNU Library General Public
- License as published by the Free Software Foundation; either
- version 2 of the License, or (at your option) any later version.
- This library 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
- Library General Public License for more details.
- You should have received a copy of the GNU Library General Public
- License along with this library; if not, write to the Free
- Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
- Sam Lantinga
- slouken@libsdl.org
- */
- #ifdef SAVE_RCSID
- static char rcsid =
- "@(#) $Id: SDL_yuv_mmx.c,v 1.4 2002/04/22 21:38:03 wmay Exp $";
- #endif
- #if defined(i386) && defined(__GNUC__) && defined(USE_ASMBLIT)
- #include "SDL_types.h"
- #ifdef __ELF__
- #define ASM_VAR(X) _##X
- #else
- #define ASM_VAR(X) X
- #endif
- static unsigned int ASM_VAR(MMX_0080w)[] = {0x00800080, 0x00800080};
- static unsigned int ASM_VAR(MMX_00FFw)[] = {0x00ff00ff, 0x00ff00ff};
- static unsigned int ASM_VAR(MMX_FF00w)[] = {0xff00ff00, 0xff00ff00};
- static unsigned short ASM_VAR(MMX_Ycoeff)[] = {0x004a, 0x004a, 0x004a, 0x004a};
- static unsigned short ASM_VAR(MMX_UbluRGB)[] = {0x0072, 0x0072, 0x0072, 0x0072};
- static unsigned short ASM_VAR(MMX_VredRGB)[] = {0x0059, 0x0059, 0x0059, 0x0059};
- static unsigned short ASM_VAR(MMX_UgrnRGB)[] = {0xffea, 0xffea, 0xffea, 0xffea};
- static unsigned short ASM_VAR(MMX_VgrnRGB)[] = {0xffd2, 0xffd2, 0xffd2, 0xffd2};
- static unsigned short ASM_VAR(MMX_Ublu5x5)[] = {0x0081, 0x0081, 0x0081, 0x0081};
- static unsigned short ASM_VAR(MMX_Vred5x5)[] = {0x0066, 0x0066, 0x0066, 0x0066};
- static unsigned short ASM_VAR(MMX_Ugrn555)[] = {0xffe7, 0xffe7, 0xffe7, 0xffe7};
- static unsigned short ASM_VAR(MMX_Vgrn555)[] = {0xffcc, 0xffcc, 0xffcc, 0xffcc};
- static unsigned short ASM_VAR(MMX_Ugrn565)[] = {0xffe8, 0xffe8, 0xffe8, 0xffe8};
- static unsigned short ASM_VAR(MMX_Vgrn565)[] = {0xffcd, 0xffcd, 0xffcd, 0xffcd};
- static unsigned short ASM_VAR(MMX_red555)[] = {0x7c00, 0x7c00, 0x7c00, 0x7c00};
- static unsigned short ASM_VAR(MMX_red565)[] = {0xf800, 0xf800, 0xf800, 0xf800};
- static unsigned short ASM_VAR(MMX_grn555)[] = {0x03e0, 0x03e0, 0x03e0, 0x03e0};
- static unsigned short ASM_VAR(MMX_grn565)[] = {0x07e0, 0x07e0, 0x07e0, 0x07e0};
- static unsigned short ASM_VAR(MMX_blu5x5)[] = {0x001f, 0x001f, 0x001f, 0x001f};
- /**
- This MMX assembler is my first assembler/MMX program ever.
- Thus it maybe buggy.
- Send patches to:
- mvogt@rhrk.uni-kl.de
- After it worked fine I have "obfuscated" the code a bit to have
- more parallism in the MMX units. This means I moved
- initilisation around and delayed other instruction.
- Performance measurement did not show that this brought any advantage
- but in theory it _should_ be faster this way.
- The overall performanve gain to the C based dither was 30%-40%.
- The MMX routine calculates 256bit=8RGB values in each cycle
- (4 for row1 & 4 for row2)
- The red/green/blue.. coefficents are taken from the mpeg_play
- player. They look nice, but I dont know if you can have
- better values, to avoid integer rounding errors.
- IMPORTANT:
- ==========
- It is a requirement that the cr/cb/lum are 8 byte aligned and
- the out are 16byte aligned or you will/may get segfaults
- */
- void ColorRGBDitherYV12MMX1X( int *colortab, Uint32 *rgb_2_pix,
- unsigned char *lum, unsigned char *cr,
- unsigned char *cb, unsigned char *out,
- int rows, int cols, int mod )
- {
- Uint32 *row1;
- Uint32 *row2;
- unsigned char* y = lum +cols*rows; // Pointer to the end
- int x=0;
- row1 = (Uint32 *)out; // 32 bit target
- row2 = (Uint32 *)out+cols+mod; // start of second row
- mod = (mod+cols+mod)*4; // increment for row1 in byte
- __asm__ __volatile__ (
- /* We don't really care about PIC - the code should be rewritten to use
- relative addressing for the static tables, so right now we take the
- COW hit on the pages this code resides. Big deal.
- This spill is just to reduce register pressure in the PIC case. */
- "pushl %%ebxn"
- "movl %0, %%ebxn"
- ".align 8n"
- "1:n"
- // create Cr (result in mm1)
- "movd (%%ebx), %%mm1n" // 0 0 0 0 v3 v2 v1 v0
- "pxor %%mm7,%%mm7n" // 00 00 00 00 00 00 00 00
- "movd (%2), %%mm2n" // 0 0 0 0 l3 l2 l1 l0
- "punpcklbw %%mm7,%%mm1n" // 0 v3 0 v2 00 v1 00 v0
- "punpckldq %%mm1,%%mm1n" // 00 v1 00 v0 00 v1 00 v0
- "psubw _MMX_0080w,%%mm1n" // mm1-128:r1 r1 r0 r0 r1 r1 r0 r0
- // create Cr_g (result in mm0)
- "movq %%mm1,%%mm0n" // r1 r1 r0 r0 r1 r1 r0 r0
- "pmullw _MMX_VgrnRGB,%%mm0n"// red*-46dec=0.7136*64
- "pmullw _MMX_VredRGB,%%mm1n"// red*89dec=1.4013*64
- "psraw $6, %%mm0n" // red=red/64
- "psraw $6, %%mm1n" // red=red/64
- // create L1 L2 (result in mm2,mm4)
- // L2=lum+cols
- "movq (%2,%4),%%mm3n" // 0 0 0 0 L3 L2 L1 L0
- "punpckldq %%mm3,%%mm2n" // L3 L2 L1 L0 l3 l2 l1 l0
- "movq %%mm2,%%mm4n" // L3 L2 L1 L0 l3 l2 l1 l0
- "pand _MMX_FF00w,%%mm2n" // L3 0 L1 0 l3 0 l1 0
- "pand _MMX_00FFw,%%mm4n" // 0 L2 0 L0 0 l2 0 l0
- "psrlw $8,%%mm2n" // 0 L3 0 L1 0 l3 0 l1
- // create R (result in mm6)
- "movq %%mm2,%%mm5n" // 0 L3 0 L1 0 l3 0 l1
- "movq %%mm4,%%mm6n" // 0 L2 0 L0 0 l2 0 l0
- "paddsw %%mm1, %%mm5n" // lum1+red:x R3 x R1 x r3 x r1
- "paddsw %%mm1, %%mm6n" // lum1+red:x R2 x R0 x r2 x r0
- "packuswb %%mm5,%%mm5n" // R3 R1 r3 r1 R3 R1 r3 r1
- "packuswb %%mm6,%%mm6n" // R2 R0 r2 r0 R2 R0 r2 r0
- "pxor %%mm7,%%mm7n" // 00 00 00 00 00 00 00 00
- "punpcklbw %%mm5,%%mm6n" // R3 R2 R1 R0 r3 r2 r1 r0
- // create Cb (result in mm1)
- "movd (%1), %%mm1n" // 0 0 0 0 u3 u2 u1 u0
- "punpcklbw %%mm7,%%mm1n" // 0 u3 0 u2 00 u1 00 u0
- "punpckldq %%mm1,%%mm1n" // 00 u1 00 u0 00 u1 00 u0
- "psubw _MMX_0080w,%%mm1n" // mm1-128:u1 u1 u0 u0 u1 u1 u0 u0
- // create Cb_g (result in mm5)
- "movq %%mm1,%%mm5n" // u1 u1 u0 u0 u1 u1 u0 u0
- "pmullw _MMX_UgrnRGB,%%mm5n" // blue*-109dec=1.7129*64
- "pmullw _MMX_UbluRGB,%%mm1n" // blue*114dec=1.78125*64
- "psraw $6, %%mm5n" // blue=red/64
- "psraw $6, %%mm1n" // blue=blue/64
- // create G (result in mm7)
- "movq %%mm2,%%mm3n" // 0 L3 0 L1 0 l3 0 l1
- "movq %%mm4,%%mm7n" // 0 L2 0 L0 0 l2 0 l1
- "paddsw %%mm5, %%mm3n" // lum1+Cb_g:x G3t x G1t x g3t x g1t
- "paddsw %%mm5, %%mm7n" // lum1+Cb_g:x G2t x G0t x g2t x g0t
- "paddsw %%mm0, %%mm3n" // lum1+Cr_g:x G3 x G1 x g3 x g1
- "paddsw %%mm0, %%mm7n" // lum1+blue:x G2 x G0 x g2 x g0
- "packuswb %%mm3,%%mm3n" // G3 G1 g3 g1 G3 G1 g3 g1
- "packuswb %%mm7,%%mm7n" // G2 G0 g2 g0 G2 G0 g2 g0
- "punpcklbw %%mm3,%%mm7n" // G3 G2 G1 G0 g3 g2 g1 g0
- // create B (result in mm5)
- "movq %%mm2,%%mm3n" // 0 L3 0 L1 0 l3 0 l1
- "movq %%mm4,%%mm5n" // 0 L2 0 L0 0 l2 0 l1
- "paddsw %%mm1, %%mm3n" // lum1+blue:x B3 x B1 x b3 x b1
- "paddsw %%mm1, %%mm5n" // lum1+blue:x B2 x B0 x b2 x b0
- "packuswb %%mm3,%%mm3n" // B3 B1 b3 b1 B3 B1 b3 b1
- "packuswb %%mm5,%%mm5n" // B2 B0 b2 b0 B2 B0 b2 b0
- "punpcklbw %%mm3,%%mm5n" // B3 B2 B1 B0 b3 b2 b1 b0
- // fill destination row1 (needed are mm6=Rr,mm7=Gg,mm5=Bb)
- "pxor %%mm2,%%mm2n" // 0 0 0 0 0 0 0 0
- "pxor %%mm4,%%mm4n" // 0 0 0 0 0 0 0 0
- "movq %%mm6,%%mm1n" // R3 R2 R1 R0 r3 r2 r1 r0
- "movq %%mm5,%%mm3n" // B3 B2 B1 B0 b3 b2 b1 b0
- // process lower lum
- "punpcklbw %%mm4,%%mm1n" // 0 r3 0 r2 0 r1 0 r0
- "punpcklbw %%mm4,%%mm3n" // 0 b3 0 b2 0 b1 0 b0
- "movq %%mm1,%%mm2n" // 0 r3 0 r2 0 r1 0 r0
- "movq %%mm3,%%mm0n" // 0 b3 0 b2 0 b1 0 b0
- "punpcklwd %%mm1,%%mm3n" // 0 r1 0 b1 0 r0 0 b0
- "punpckhwd %%mm2,%%mm0n" // 0 r3 0 b3 0 r2 0 b2
- "pxor %%mm2,%%mm2n" // 0 0 0 0 0 0 0 0
- "movq %%mm7,%%mm1n" // G3 G2 G1 G0 g3 g2 g1 g0
- "punpcklbw %%mm1,%%mm2n" // g3 0 g2 0 g1 0 g0 0
- "punpcklwd %%mm4,%%mm2n" // 0 0 g1 0 0 0 g0 0
- "por %%mm3, %%mm2n" // 0 r1 g1 b1 0 r0 g0 b0
- "movq %%mm2,(%3)n" // wrote out ! row1
- "pxor %%mm2,%%mm2n" // 0 0 0 0 0 0 0 0
- "punpcklbw %%mm1,%%mm4n" // g3 0 g2 0 g1 0 g0 0
- "punpckhwd %%mm2,%%mm4n" // 0 0 g3 0 0 0 g2 0
- "por %%mm0, %%mm4n" // 0 r3 g3 b3 0 r2 g2 b2
- "movq %%mm4,8(%3)n" // wrote out ! row1
- // fill destination row2 (needed are mm6=Rr,mm7=Gg,mm5=Bb)
- // this can be done "destructive"
- "pxor %%mm2,%%mm2n" // 0 0 0 0 0 0 0 0
- "punpckhbw %%mm2,%%mm6n" // 0 R3 0 R2 0 R1 0 R0
- "punpckhbw %%mm1,%%mm5n" // G3 B3 G2 B2 G1 B1 G0 B0
- "movq %%mm5,%%mm1n" // G3 B3 G2 B2 G1 B1 G0 B0
- "punpcklwd %%mm6,%%mm1n" // 0 R1 G1 B1 0 R0 G0 B0
- "movq %%mm1,(%5)n" // wrote out ! row2
- "punpckhwd %%mm6,%%mm5n" // 0 R3 G3 B3 0 R2 G2 B2
- "movq %%mm5,8(%5)n" // wrote out ! row2
- "addl $4,%2n" // lum+4
- "leal 16(%3),%3n" // row1+16
- "leal 16(%5),%5n" // row2+16
- "addl $2, %%ebxn" // cr+2
- "addl $2, %1n" // cb+2
- "addl $4,%6n" // x+4
- "cmpl %4,%6n"
- "jl 1bn"
- "addl %4, %2n" // lum += cols
- "addl %8, %3n" // row1+= mod
- "addl %8, %5n" // row2+= mod
- "movl $0, %6n" // x=0
- "cmpl %7, %2n"
- "jl 1bn"
- "emmsn"
- "popl %%ebxn"
- :
- : "m" (cr), "r"(cb),"r"(lum),
- "r"(row1),"r"(cols),"r"(row2),"m"(x),"m"(y),"m"(mod)
- : "%ebx"
- );
- }
- void Color565DitherYV12MMX1X( int *colortab, Uint32 *rgb_2_pix,
- unsigned char *lum, unsigned char *cr,
- unsigned char *cb, unsigned char *out,
- int rows, int cols, int mod )
- {
- Uint16 *row1;
- Uint16 *row2;
- unsigned char* y = lum +cols*rows; /* Pointer to the end */
- int x=0;
- row1 = (Uint16 *)out; /* 16 bit target */
- row2 = (Uint16 *)out+cols+mod; /* start of second row */
- mod = (mod+cols+mod)*2; /* increment for row1 in byte */
- __asm__ __volatile__(
- "pushl %%ebxn"
- "movl %0, %%ebxn"
- ".align 8n"
- "1:n"
- "movd (%1), %%mm0n" // 4 Cb 0 0 0 0 u3 u2 u1 u0
- "pxor %%mm7, %%mm7n"
- "movd (%%ebx), %%mm1n" // 4 Cr 0 0 0 0 v3 v2 v1 v0
- "punpcklbw %%mm7, %%mm0n" // 4 W cb 0 u3 0 u2 0 u1 0 u0
- "punpcklbw %%mm7, %%mm1n" // 4 W cr 0 v3 0 v2 0 v1 0 v0
- "psubw _MMX_0080w, %%mm0n"
- "psubw _MMX_0080w, %%mm1n"
- "movq %%mm0, %%mm2n" // Cb 0 u3 0 u2 0 u1 0 u0
- "movq %%mm1, %%mm3n" // Cr
- "pmullw _MMX_Ugrn565, %%mm2n" // Cb2green 0 R3 0 R2 0 R1 0 R0
- "movq (%2), %%mm6n" // L1 l7 L6 L5 L4 L3 L2 L1 L0
- "pmullw _MMX_Ublu5x5, %%mm0n" // Cb2blue
- "pand _MMX_00FFw, %%mm6n" // L1 00 L6 00 L4 00 L2 00 L0
- "pmullw _MMX_Vgrn565, %%mm3n" // Cr2green
- "movq (%2), %%mm7n" // L2
- "pmullw _MMX_Vred5x5, %%mm1n" // Cr2red
- "psrlw $8, %%mm7n" // L2 00 L7 00 L5 00 L3 00 L1
- "pmullw _MMX_Ycoeff, %%mm6n" // lum1
- "paddw %%mm3, %%mm2n" // Cb2green + Cr2green == green
- "pmullw _MMX_Ycoeff, %%mm7n" // lum2
- "movq %%mm6, %%mm4n" // lum1
- "paddw %%mm0, %%mm6n" // lum1 +blue 00 B6 00 B4 00 B2 00 B0
- "movq %%mm4, %%mm5n" // lum1
- "paddw %%mm1, %%mm4n" // lum1 +red 00 R6 00 R4 00 R2 00 R0
- "paddw %%mm2, %%mm5n" // lum1 +green 00 G6 00 G4 00 G2 00 G0
- "psraw $6, %%mm4n" // R1 0 .. 64
- "movq %%mm7, %%mm3n" // lum2 00 L7 00 L5 00 L3 00 L1
- "psraw $6, %%mm5n" // G1 - .. +
- "paddw %%mm0, %%mm7n" // Lum2 +blue 00 B7 00 B5 00 B3 00 B1
- "psraw $6, %%mm6n" // B1 0 .. 64
- "packuswb %%mm4, %%mm4n" // R1 R1
- "packuswb %%mm5, %%mm5n" // G1 G1
- "packuswb %%mm6, %%mm6n" // B1 B1
- "punpcklbw %%mm4, %%mm4n"
- "punpcklbw %%mm5, %%mm5n"
- "pand _MMX_red565, %%mm4n"
- "psllw $3, %%mm5n" // GREEN 1
- "punpcklbw %%mm6, %%mm6n"
- "pand _MMX_grn565, %%mm5n"
- "pand _MMX_red565, %%mm6n"
- "por %%mm5, %%mm4n" //
- "psrlw $11, %%mm6n" // BLUE 1
- "movq %%mm3, %%mm5n" // lum2
- "paddw %%mm1, %%mm3n" // lum2 +red 00 R7 00 R5 00 R3 00 R1
- "paddw %%mm2, %%mm5n" // lum2 +green 00 G7 00 G5 00 G3 00 G1
- "psraw $6, %%mm3n" // R2
- "por %%mm6, %%mm4n" // MM4
- "psraw $6, %%mm5n" // G2
- "movq (%2, %4), %%mm6n" // L3 load lum2
- "psraw $6, %%mm7n"
- "packuswb %%mm3, %%mm3n"
- "packuswb %%mm5, %%mm5n"
- "packuswb %%mm7, %%mm7n"
- "pand _MMX_00FFw, %%mm6n" // L3
- "punpcklbw %%mm3, %%mm3n"
- "punpcklbw %%mm5, %%mm5n"
- "pmullw _MMX_Ycoeff, %%mm6n" // lum3
- "punpcklbw %%mm7, %%mm7n"
- "psllw $3, %%mm5n" // GREEN 2
- "pand _MMX_red565, %%mm7n"
- "pand _MMX_red565, %%mm3n"
- "psrlw $11, %%mm7n" // BLUE 2
- "pand _MMX_grn565, %%mm5n"
- "por %%mm7, %%mm3n"
- "movq (%2,%4), %%mm7n" // L4 load lum2
- "por %%mm5, %%mm3n" //
- "psrlw $8, %%mm7n" // L4
- "movq %%mm4, %%mm5n"
- "punpcklwd %%mm3, %%mm4n"
- "pmullw _MMX_Ycoeff, %%mm7n" // lum4
- "punpckhwd %%mm3, %%mm5n"
- "movq %%mm4, (%3)n" // write row1
- "movq %%mm5, 8(%3)n" // write row1
- "movq %%mm6, %%mm4n" // Lum3
- "paddw %%mm0, %%mm6n" // Lum3 +blue
- "movq %%mm4, %%mm5n" // Lum3
- "paddw %%mm1, %%mm4n" // Lum3 +red
- "paddw %%mm2, %%mm5n" // Lum3 +green
- "psraw $6, %%mm4n"
- "movq %%mm7, %%mm3n" // Lum4
- "psraw $6, %%mm5n"
- "paddw %%mm0, %%mm7n" // Lum4 +blue
- "psraw $6, %%mm6n" // Lum3 +blue
- "movq %%mm3, %%mm0n" // Lum4
- "packuswb %%mm4, %%mm4n"
- "paddw %%mm1, %%mm3n" // Lum4 +red
- "packuswb %%mm5, %%mm5n"
- "paddw %%mm2, %%mm0n" // Lum4 +green
- "packuswb %%mm6, %%mm6n"
- "punpcklbw %%mm4, %%mm4n"
- "punpcklbw %%mm5, %%mm5n"
- "punpcklbw %%mm6, %%mm6n"
- "psllw $3, %%mm5n" // GREEN 3
- "pand _MMX_red565, %%mm4n"
- "psraw $6, %%mm3n" // psr 6
- "psraw $6, %%mm0n"
- "pand _MMX_red565, %%mm6n" // BLUE
- "pand _MMX_grn565, %%mm5n"
- "psrlw $11, %%mm6n" // BLUE 3
- "por %%mm5, %%mm4n"
- "psraw $6, %%mm7n"
- "por %%mm6, %%mm4n"
- "packuswb %%mm3, %%mm3n"
- "packuswb %%mm0, %%mm0n"
- "packuswb %%mm7, %%mm7n"
- "punpcklbw %%mm3, %%mm3n"
- "punpcklbw %%mm0, %%mm0n"
- "punpcklbw %%mm7, %%mm7n"
- "pand _MMX_red565, %%mm3n"
- "pand _MMX_red565, %%mm7n" // BLUE
- "psllw $3, %%mm0n" // GREEN 4
- "psrlw $11, %%mm7n"
- "pand _MMX_grn565, %%mm0n"
- "por %%mm7, %%mm3n"
- "por %%mm0, %%mm3n"
- "movq %%mm4, %%mm5n"
- "punpcklwd %%mm3, %%mm4n"
- "punpckhwd %%mm3, %%mm5n"
- "movq %%mm4, (%5)n"
- "movq %%mm5, 8(%5)n"
- "addl $8, %6n"
- "addl $8, %2n"
- "addl $4, %%ebxn"
- "addl $4, %1n"
- "cmpl %4, %6n"
- "leal 16(%3), %3n"
- "leal 16(%5),%5n" // row2+16
- "jl 1bn"
- "addl %4, %2n" // lum += cols
- "addl %8, %3n" // row1+= mod
- "addl %8, %5n" // row2+= mod
- "movl $0, %6n" // x=0
- "cmpl %7, %2n"
- "jl 1bn"
- "emmsn"
- "popl %%ebxn"
- :
- :"m" (cr), "r"(cb),"r"(lum),
- "r"(row1),"r"(cols),"r"(row2),"m"(x),"m"(y),"m"(mod)
- : "%ebx"
- );
- }
- #endif /* GCC i386 inline assembly */