Skip to content

L
libswscale
Commit 61f36e03 authored by michael's avatar michael
Browse files
Options

increased precission of s_xinc s_xinc2 (needed for the mmx2 bugfix) 

moved mmx variables to top to avoid alignment issues
mmx2 code should work fine now if and only if the input width is %16=0 and the output width is %32=0
reordered some code (5% faster with a simply -benchmark)
first line bug fixed (i hope i didnt introduce any new bugs with that ...)
changed a lot of the vertical scale setup code, i hope i fixed something and didnt mess it up :)
a few known bugs left (rightmost line is wrong)
MMX2 code will only be used for upscaling & acceptable width´s
16bit dithering can be disabled


git-svn-id: file:///var/local/repositories/mplayer/trunk/postproc@2265 b3059339-0415-0410-9bf9-f77b7e298cf2
parent bd0d5ef6
Show whitespace changes
Inline Side-by-side
Showing with 542 additions and 412 deletions
swscale.c
View file @ 61f36e03
// Software scaling and colorspace conversion routines for MPlayer
// Orginal C implementation by ?
// current version mostly by Michael Niedermayer (michaelni@gmx.at)
#include <inttypes.h>
#include "../config.h"
#undef HAVE_MMX2 //code is buggy
//#undef HAVE_MMX2
//#undef HAVE_MMX
//#undef ARCH_X86
#define DITHER16BPP
#define ALT_ERROR
#define RET 0xC3 //near return opcode
/*
NOTES
// temporary storage for 4 yuv lines:
// 16bit for now (mmx likes it more compact)
static uint16_t pix_buf_y[4][2048];
static uint16_t pix_buf_uv[2][2048*2];
known BUGS with known cause (no bugreports please!)
line at the right (c,asm and mmx2)
code reads 1 sample too much (might cause a sig11)
// clipping helper table for C implementations:
static unsigned char clip_table[768];
// yuv->rgb conversion tables:
static int yuvtab_2568[256];
static int yuvtab_3343[256];
static int yuvtab_0c92[256];
static int yuvtab_1a1e[256];
static int yuvtab_40cf[256];
TODO
check alignment off everything
*/
static uint64_t yCoeff= 0x2568256825682568LL;
static uint64_t ubCoeff= 0x3343334333433343LL;
......@@ -46,11 +47,27 @@ static uint64_t g16Mask= 0x07E007E007E007E0LL;
static uint64_t r16Mask= 0xF800F800F800F800LL;
static uint64_t temp0;
// temporary storage for 4 yuv lines:
// 16bit for now (mmx likes it more compact)
static uint16_t pix_buf_y[4][2048];
static uint16_t pix_buf_uv[2][2048*2];
// clipping helper table for C implementations:
static unsigned char clip_table[768];
// yuv->rgb conversion tables:
static int yuvtab_2568[256];
static int yuvtab_3343[256];
static int yuvtab_0c92[256];
static int yuvtab_1a1e[256];
static int yuvtab_40cf[256];
static uint8_t funnyYCode[10000];
static uint8_t funnyUVCode[10000];
// *** bilinear scaling and yuv->rgb conversion of yv12 slices:
// *** Note: it's called multiple times while decoding a frame, first time y==0
// *** Designed to upscale, but may work for downscale too.
......@@ -64,27 +81,43 @@ void SwScale_YV12slice_brg24(unsigned char* srcptr[],int stride[], int y, int h,
//static int s_yinc=(vo_dga_src_height<<16)/vo_dga_vp_height;
//static int s_xinc=(vo_dga_src_width<<8)/vo_dga_vp_width;
unsigned int s_xinc2=s_xinc>>1;
unsigned int s_xinc2;
static int s_srcypos;
static int s_srcypos; // points to the dst Pixels center in the source (0 is the center of pixel 0,0 in src)
static int s_ypos;
// last horzontally interpolated lines, used to avoid unnecessary calculations
static int s_last_ypos;
static int s_last_y1pos;
static int static_dstw;
#ifdef HAVE_MMX2
// used to detect a horizontal size change
static int old_dstw= -1;
static int old_s_xinc= -1;
// difference between the requested xinc and the required one for the mmx2 routine
static int s_xinc_diff=0;
static int s_xinc2_diff=0;
#endif
int canMMX2BeUsed;
s_xinc&= -2; //clear last bit or uv and y might be shifted relative to each other
// we need that precission at least for the mmx2 code
s_xinc*= 256;
s_xinc2=s_xinc>>1;
canMMX2BeUsed= (s_xinc <= 0x10000 && (dstw&31)==0) ? 1 : 0;
if(y==0){
s_srcypos=-2*s_yinc;
s_ypos=-2;
s_last_ypos=-2;
s_srcypos= s_yinc/2 - 0x8000;
s_ypos=0;
// force calculation of the horizontal interpolation of the first line
s_last_ypos=-99;
s_last_y1pos=-99;
#ifdef HAVE_MMX2
// cant downscale !!!
if(old_s_xinc != s_xinc || old_dstw!=dstw)
if((old_s_xinc != s_xinc || old_dstw!=dstw) && canMMX2BeUsed)
{
uint8_t *fragment;
int imm8OfPShufW1;
......@@ -102,32 +135,30 @@ s_xinc&= -2; //clear last bit or uv and y might be shifted relative to each othe
//code fragment
// fragmentLength=0;
// printf("%d, %d\n", fragmentLength,imm8OfPShufW1);
asm volatile(
"jmp 9f \n\t"
// Begin
"0: \n\t"
"movq (%%esi, %%ebx), %%mm0 \n\t" //FIXME Alignment
"movq (%%esi), %%mm0 \n\t" //FIXME Alignment
"movq %%mm0, %%mm1 \n\t"
"psrlq $8, %%mm0 \n\t"
"punpcklbw %%mm7, %%mm1 \n\t"
"movq %%mm2, %%mm3 \n\t"
"punpcklbw %%mm7, %%mm0 \n\t"
"addw %%bx, %%cx \n\t" //2*xalpha += (4*s_xinc)&0xFFFF
"pshufw $0xFF, %%mm1, %%mm1 \n\t"
"1: \n\t"
"adcl %%edx, %%esi \n\t" //xx+= (4*s_xinc)>>16 + carry
"pshufw $0xFF, %%mm0, %%mm0 \n\t"
"2: \n\t"
"psrlw $9, %%mm3 \n\t"
"psubw %%mm1, %%mm0 \n\t"
"psraw $1, %%mm0 \n\t"
"pmullw %%mm2, %%mm0 \n\t"
"pmullw %%mm3, %%mm0 \n\t"
"paddw %%mm6, %%mm2 \n\t" // 2*alpha += xpos&0xFFFF
"psllw $7, %%mm1 \n\t"
"paddw %%mm1, %%mm0 \n\t"
"movq %%mm0, (%%edi, %%eax) \n\t"
"paddb %%mm6, %%mm2 \n\t" // 2*alpha += xpos&0xFF
"addb %%ch, %%cl \n\t" //2*xalpha += (4*s_xinc)&0xFF
"adcl %%edx, %%ebx \n\t" //xx+= (4*s_xinc)>>8 + carry
"movq %%mm0, (%%edi, %%eax) \n\t"
"addl $8, %%eax \n\t"
// End
......@@ -147,17 +178,28 @@ s_xinc&= -2; //clear last bit or uv and y might be shifted relative to each othe
);
xpos= xx=xalpha= 0;
//FIXME choose size and or xinc so that they fit exactly
/* choose xinc so that all 8 parts fit exactly
Note: we cannot use just 1 part because it would not fit in the code cache */
s_xinc2_diff= -((((s_xinc2*(dstw/8))&0xFFFF))/(dstw/8))+10;
// s_xinc_diff= -((((s_xinc*(dstw/8))&0xFFFF))/(dstw/8));
#ifdef ALT_ERROR
s_xinc2_diff+= ((0x10000/(dstw/8)));
#endif
s_xinc_diff= s_xinc2_diff*2;
s_xinc2+= s_xinc2_diff;
s_xinc+= s_xinc_diff;
for(i=0; i<dstw/8; i++)
{
int xx=xpos>>8;
int xx=xpos>>16;
if((i&3) == 0)
{
int a=0;
int b=((xpos+s_xinc)>>8) - xx;
int c=((xpos+s_xinc*2)>>8) - xx;
int d=((xpos+s_xinc*3)>>8) - xx;
int b=((xpos+s_xinc)>>16) - xx;
int c=((xpos+s_xinc*2)>>16) - xx;
int d=((xpos+s_xinc*3)>>16) - xx;
memcpy(funnyYCode + fragmentLength*i/4, fragment, fragmentLength);
......@@ -174,14 +216,14 @@ s_xinc&= -2; //clear last bit or uv and y might be shifted relative to each othe
//FIXME choose size and or xinc so that they fit exactly
for(i=0; i<dstw/8; i++)
{
int xx=xpos>>8;
int xx=xpos>>16;
if((i&3) == 0)
{
int a=0;
int b=((xpos+s_xinc2)>>8) - xx;
int c=((xpos+s_xinc2*2)>>8) - xx;
int d=((xpos+s_xinc2*3)>>8) - xx;
int b=((xpos+s_xinc2)>>16) - xx;
int c=((xpos+s_xinc2*2)>>16) - xx;
int d=((xpos+s_xinc2*3)>>16) - xx;
memcpy(funnyUVCode + fragmentLength*i/4, fragment, fragmentLength);
......@@ -197,86 +239,117 @@ s_xinc&= -2; //clear last bit or uv and y might be shifted relative to each othe
}
#endif
if(canMMX2BeUsed)
{
s_xinc+= s_xinc_diff;
s_xinc2+= s_xinc2_diff;
}
#endif // HAVE_MMX2
} // reset counters
while(1){
unsigned char *dest=dstptr+dststride*s_ypos;
int y0=2+(s_srcypos>>16);
int y1=1+(s_srcypos>>17);
int y0=(s_srcypos + 0xFFFF)>>16; // first luminance source line number below the dst line
// points to the dst Pixels center in the source (0 is the center of pixel 0,0 in src)
int srcuvpos= s_srcypos + s_yinc/2 - 0x8000;
int y1=(srcuvpos + 0x1FFFF)>>17; // first chrominance source line number below the dst line
int yalpha=(s_srcypos&0xFFFF)>>7;
int yalpha1=yalpha^511;
int uvalpha=((s_srcypos>>1)&0xFFFF)>>7;
int uvalpha=(srcuvpos&0x1FFFF)>>8;
int uvalpha1=uvalpha^511;
uint16_t *buf0=pix_buf_y[y0&3];
uint16_t *buf1=pix_buf_y[((y0+1)&3)];
uint16_t *uvbuf0=pix_buf_uv[y1&1];
uint16_t *uvbuf1=pix_buf_uv[(y1&1)^1];
uint16_t *buf0=pix_buf_y[y0&1]; // top line of the interpolated slice
uint16_t *buf1=pix_buf_y[((y0+1)&1)]; // bottom line of the interpolated slice
uint16_t *uvbuf0=pix_buf_uv[y1&1]; // top line of the interpolated slice
uint16_t *uvbuf1=pix_buf_uv[(y1+1)&1]; // bottom line of the interpolated slice
int i;
if(y0>=y+h) break;
// if this is before the first line than use only the first src line
if(y0==0) buf0= buf1;
if(y1==0) uvbuf0= uvbuf1; // yes we do have to check this, its not the same as y0==0
if(y0>=y+h) break; // FIXME wrong, skips last lines, but they are dupliactes anyway
// if this is after the last line than use only the last src line
if(y0>=y+h)
{
buf1= buf0;
s_last_ypos=y0;
}
if(y1>=(y+h)/2)
{
uvbuf1= uvbuf0;
s_last_y1pos=y1;
}
s_ypos++; s_srcypos+=s_yinc;
//only interpolate the src line horizontally if we didnt do it allready
if(s_last_ypos!=y0){
unsigned char *src=srcptr[0]+(y0-y)*stride[0];
unsigned int xpos=0;
s_last_ypos=y0;
// *** horizontal scale Y line to temp buffer
// this loop should be rewritten in MMX assembly!!!!
#ifdef ARCH_X86
#ifdef HAVE_MMX2
if(canMMX2BeUsed)
{
asm volatile(
"pxor %%mm7, %%mm7 \n\t"
"pxor %%mm2, %%mm2 \n\t" // 2*xalpha
"movd %5, %%mm6 \n\t" // s_xinc&0xFF
"movd %5, %%mm6 \n\t" // s_xinc&0xFFFF
"punpcklwd %%mm6, %%mm6 \n\t"
"punpcklwd %%mm6, %%mm6 \n\t"
"movq %%mm6, %%mm2 \n\t"
"psllq $16, %%mm2 \n\t"
"paddb %%mm6, %%mm2 \n\t"
"paddw %%mm6, %%mm2 \n\t"
"psllq $16, %%mm2 \n\t"
"paddb %%mm6, %%mm2 \n\t"
"paddw %%mm6, %%mm2 \n\t"
"psllq $16, %%mm2 \n\t" //0,t,2t,3t t=s_xinc&0xFF
"movq %%mm2, temp0 \n\t"
"movd %4, %%mm6 \n\t" //(s_xinc*4)&0xFF
"movd %4, %%mm6 \n\t" //(s_xinc*4)&0xFFFF
"punpcklwd %%mm6, %%mm6 \n\t"
"punpcklwd %%mm6, %%mm6 \n\t"
"xorl %%eax, %%eax \n\t" // i
"xorl %%ebx, %%ebx \n\t" // xx
"movl %0, %%esi \n\t" // src
"movl %1, %%edi \n\t" // buf1
"movl %3, %%edx \n\t" // (s_xinc*4)>>8
"movl %3, %%edx \n\t" // (s_xinc*4)>>16
"xorl %%ecx, %%ecx \n\t"
"movb %4, %%ch \n\t" // (s_xinc*4)&0xFF
// "int $3\n\t"
"xorl %%ebx, %%ebx \n\t"
"movw %4, %%bx \n\t" // (s_xinc*4)&0xFFFF
// "int $3\n\t"
"call funnyYCode \n\t"
"movq temp0, %%mm2 \n\t"
"xorb %%cl, %%cl \n\t"
"xorl %%ecx, %%ecx \n\t"
"call funnyYCode \n\t"
"movq temp0, %%mm2 \n\t"
"xorb %%cl, %%cl \n\t"
"xorl %%ecx, %%ecx \n\t"
"call funnyYCode \n\t"
"movq temp0, %%mm2 \n\t"
"xorb %%cl, %%cl \n\t"
"xorl %%ecx, %%ecx \n\t"
"call funnyYCode \n\t"
"movq temp0, %%mm2 \n\t"
"xorb %%cl, %%cl \n\t"
"xorl %%ecx, %%ecx \n\t"
"call funnyYCode \n\t"
"movq temp0, %%mm2 \n\t"
"xorb %%cl, %%cl \n\t"
"xorl %%ecx, %%ecx \n\t"
"call funnyYCode \n\t"
"movq temp0, %%mm2 \n\t"
"xorb %%cl, %%cl \n\t"
"xorl %%ecx, %%ecx \n\t"
"call funnyYCode \n\t"
"movq temp0, %%mm2 \n\t"
"xorb %%cl, %%cl \n\t"
"xorl %%ecx, %%ecx \n\t"
"call funnyYCode \n\t"
:: "m" (src), "m" (buf1), "m" (dstw), "m" ((s_xinc*4)>>8),
"m" ((s_xinc*4)&0xFF), "m" (s_xinc&0xFF)
:: "m" (src), "m" (buf1), "m" (dstw), "m" ((s_xinc*4)>>16),
"m" ((s_xinc*4)&0xFFFF), "m" (s_xinc&0xFFFF)
: "%eax", "%ebx", "%ecx", "%edx", "%esi", "%edi"
);
#elif defined (ARCH_X86)
}
else
{
#endif
//NO MMX just normal asm ... FIXME try/write funny MMX2 variant
//FIXME add prefetch
asm volatile(
......@@ -288,24 +361,24 @@ s_xinc&= -2; //clear last bit or uv and y might be shifted relative to each othe
"movzbl 1(%0, %%ebx), %%esi \n\t" //src[xx+1]
"subl %%edi, %%esi \n\t" //src[xx+1] - src[xx]
"imull %%ecx, %%esi \n\t" //(src[xx+1] - src[xx])*2*xalpha
"shll $8, %%edi \n\t"
"shll $16, %%edi \n\t"
"addl %%edi, %%esi \n\t" //src[xx+1]*2*xalpha + src[xx]*(1-2*xalpha)
"movl %1, %%edi \n\t"
"shrl $1, %%esi \n\t"
"shrl $9, %%esi \n\t"
"movw %%si, (%%edi, %%eax, 2) \n\t"
"addb %4, %%cl \n\t" //2*xalpha += s_xinc&0xFF
"addw %4, %%cx \n\t" //2*xalpha += s_xinc&0xFF
"adcl %3, %%ebx \n\t" //xx+= s_xinc>>8 + carry
"movzbl (%0, %%ebx), %%edi \n\t" //src[xx]
"movzbl 1(%0, %%ebx), %%esi \n\t" //src[xx+1]
"subl %%edi, %%esi \n\t" //src[xx+1] - src[xx]
"imull %%ecx, %%esi \n\t" //(src[xx+1] - src[xx])*2*xalpha
"shll $8, %%edi \n\t"
"shll $16, %%edi \n\t"
"addl %%edi, %%esi \n\t" //src[xx+1]*2*xalpha + src[xx]*(1-2*xalpha)
"movl %1, %%edi \n\t"
"shrl $1, %%esi \n\t"
"shrl $9, %%esi \n\t"
"movw %%si, 2(%%edi, %%eax, 2) \n\t"
"addb %4, %%cl \n\t" //2*xalpha += s_xinc&0xFF
"addw %4, %%cx \n\t" //2*xalpha += s_xinc&0xFF
"adcl %3, %%ebx \n\t" //xx+= s_xinc>>8 + carry
......@@ -314,106 +387,96 @@ s_xinc&= -2; //clear last bit or uv and y might be shifted relative to each othe
" jb 1b \n\t"
:: "r" (src), "m" (buf1), "m" (dstw), "m" (s_xinc>>8), "m" (s_xinc&0xFF)
:: "r" (src), "m" (buf1), "m" (dstw), "m" (s_xinc>>16), "m" (s_xinc&0xFFFF)
: "%eax", "%ebx", "%ecx", "%edi", "%esi"
);
#ifdef HAVE_MMX2
} //if MMX2 cant be used
#endif
#else
for(i=0;i<dstw;i++){
register unsigned int xx=xpos>>8;
register unsigned int xalpha=(xpos&0xFF)>>1;
register unsigned int xx=xpos>>16;
register unsigned int xalpha=(xpos&0xFFFF)>>9;
buf1[i]=(src[xx]*(xalpha^127)+src[xx+1]*xalpha);
xpos+=s_xinc;
}
#endif
}
// *** horizontal scale U and V lines to temp buffer
if(!(y0&1)){
if(s_last_y1pos!=y1){
unsigned char *src1=srcptr[1]+(y1-y/2)*stride[1];
unsigned char *src2=srcptr[2]+(y1-y/2)*stride[2];
xpos=0;
// this loop should be rewritten in MMX assembly!!!!
int xpos=0;
s_last_y1pos= y1;
#ifdef ARCH_X86
#ifdef HAVE_MMX2
if(canMMX2BeUsed)
{
asm volatile(
"pxor %%mm7, %%mm7 \n\t"
"pxor %%mm2, %%mm2 \n\t" // 2*xalpha
"movd %5, %%mm6 \n\t" // s_xinc&0xFF
"movd %5, %%mm6 \n\t" // s_xinc&0xFFFF
"punpcklwd %%mm6, %%mm6 \n\t"
"punpcklwd %%mm6, %%mm6 \n\t"
"movq %%mm6, %%mm2 \n\t"
"psllq $16, %%mm2 \n\t"
"paddb %%mm6, %%mm2 \n\t"
"paddw %%mm6, %%mm2 \n\t"
"psllq $16, %%mm2 \n\t"
"paddb %%mm6, %%mm2 \n\t"
"psllq $16, %%mm2 \n\t" //0,t,2t,3t t=s_xinc&0xFF
"paddw %%mm6, %%mm2 \n\t"
"psllq $16, %%mm2 \n\t" //0,t,2t,3t t=s_xinc&0xFFFF
"movq %%mm2, temp0 \n\t"
"movd %4, %%mm6 \n\t" //(s_xinc*4)&0xFF
"movd %4, %%mm6 \n\t" //(s_xinc*4)&0xFFFF
"punpcklwd %%mm6, %%mm6 \n\t"
"punpcklwd %%mm6, %%mm6 \n\t"
"xorl %%eax, %%eax \n\t" // i
"xorl %%ebx, %%ebx \n\t" // xx
"movl %0, %%esi \n\t" // src
"movl %1, %%edi \n\t" // buf1
"movl %3, %%edx \n\t" // (s_xinc*4)>>8
"movl %3, %%edx \n\t" // (s_xinc*4)>>16
"xorl %%ecx, %%ecx \n\t"
"movb %4, %%ch \n\t" // (s_xinc*4)&0xFF
"xorl %%ebx, %%ebx \n\t"
"movw %4, %%bx \n\t" // (s_xinc*4)&0xFFFF
// "int $3\n\t"
"call funnyUVCode \n\t"
"movq temp0, %%mm2 \n\t"
"xorb %%cl, %%cl \n\t"
"call funnyUVCode \n\t"
"movq temp0, %%mm2 \n\t"
"xorb %%cl, %%cl \n\t"
"call funnyUVCode \n\t"
"movq temp0, %%mm2 \n\t"
"xorb %%cl, %%cl \n\t"
"call funnyUVCode \n\t"
"movq temp0, %%mm2 \n\t"
"xorb %%cl, %%cl \n\t"
"call funnyUVCode \n\t"
"movq temp0, %%mm2 \n\t"
"xorb %%cl, %%cl \n\t"
"call funnyUVCode \n\t"
"movq temp0, %%mm2 \n\t"
"xorb %%cl, %%cl \n\t"
"call funnyUVCode \n\t"
"movq temp0, %%mm2 \n\t"
"xorb %%cl, %%cl \n\t"
"call funnyUVCode \n\t"
#define FUNNYUVCODE \
"call funnyUVCode \n\t"\
"movq temp0, %%mm2 \n\t"\
"xorl %%ecx, %%ecx \n\t"
FUNNYUVCODE
FUNNYUVCODE
FUNNYUVCODE
FUNNYUVCODE
FUNNYUVCODE
FUNNYUVCODE
FUNNYUVCODE
FUNNYUVCODE
"xorl %%eax, %%eax \n\t" // i
"xorl %%ebx, %%ebx \n\t" // xx
"movl %6, %%esi \n\t" // src
"movl %1, %%edi \n\t" // buf1
"addl $4096, %%edi \n\t"
"call funnyUVCode \n\t"
"movq temp0, %%mm2 \n\t"
"xorb %%cl, %%cl \n\t"
"call funnyUVCode \n\t"
"movq temp0, %%mm2 \n\t"
"xorb %%cl, %%cl \n\t"
"call funnyUVCode \n\t"
"movq temp0, %%mm2 \n\t"
"xorb %%cl, %%cl \n\t"
"call funnyUVCode \n\t"
"movq temp0, %%mm2 \n\t"
"xorb %%cl, %%cl \n\t"
"call funnyUVCode \n\t"
"movq temp0, %%mm2 \n\t"
"xorb %%cl, %%cl \n\t"
"call funnyUVCode \n\t"
"movq temp0, %%mm2 \n\t"
"xorb %%cl, %%cl \n\t"
"call funnyUVCode \n\t"
"movq temp0, %%mm2 \n\t"
"xorb %%cl, %%cl \n\t"
"call funnyUVCode \n\t"
FUNNYUVCODE
FUNNYUVCODE
FUNNYUVCODE
FUNNYUVCODE
FUNNYUVCODE
FUNNYUVCODE
FUNNYUVCODE
FUNNYUVCODE
:: "m" (src1), "m" (uvbuf1), "m" (dstw), "m" ((s_xinc2*4)>>8),
"m" ((s_xinc2*4)&0xFF), "m" (s_xinc2&0xFF), "m" (src2)
:: "m" (src1), "m" (uvbuf1), "m" (dstw), "m" ((s_xinc2*4)>>16),
"m" ((s_xinc2*4)&0xFFFF), "m" (s_xinc2&0xFFFF), "m" (src2)
: "%eax", "%ebx", "%ecx", "%edx", "%esi", "%edi"
);
#elif defined (ARCH_X86)
}
else
{
#endif
asm volatile(
"xorl %%eax, %%eax \n\t" // i
"xorl %%ebx, %%ebx \n\t" // xx
......@@ -424,45 +487,47 @@ s_xinc&= -2; //clear last bit or uv and y might be shifted relative to each othe
"movzbl 1(%%esi, %%ebx), %%esi \n\t" //src[xx+1]
"subl %%edi, %%esi \n\t" //src[xx+1] - src[xx]
"imull %%ecx, %%esi \n\t" //(src[xx+1] - src[xx])*2*xalpha
"shll $8, %%edi \n\t"
"shll $16, %%edi \n\t"
"addl %%edi, %%esi \n\t" //src[xx+1]*2*xalpha + src[xx]*(1-2*xalpha)
"movl %1, %%edi \n\t"
"shrl $1, %%esi \n\t"
"shrl $9, %%esi \n\t"
"movw %%si, (%%edi, %%eax, 2) \n\t"
"movzbl (%5, %%ebx), %%edi \n\t" //src[xx]
"movzbl 1(%5, %%ebx), %%esi \n\t" //src[xx+1]
"subl %%edi, %%esi \n\t" //src[xx+1] - src[xx]
"imull %%ecx, %%esi \n\t" //(src[xx+1] - src[xx])*2*xalpha
"shll $8, %%edi \n\t"
"shll $16, %%edi \n\t"
"addl %%edi, %%esi \n\t" //src[xx+1]*2*xalpha + src[xx]*(1-2*xalpha)
"movl %1, %%edi \n\t"
"shrl $1, %%esi \n\t"
"shrl $9, %%esi \n\t"
"movw %%si, 4096(%%edi, %%eax, 2)\n\t"
"addb %4, %%cl \n\t" //2*xalpha += s_xinc&0xFF
"addw %4, %%cx \n\t" //2*xalpha += s_xinc&0xFF
"adcl %3, %%ebx \n\t" //xx+= s_xinc>>8 + carry
"addl $1, %%eax \n\t"
"cmpl %2, %%eax \n\t"
" jb 1b \n\t"
:: "m" (src1), "m" (uvbuf1), "m" (dstw), "m" (s_xinc2>>8), "m" (s_xinc2&0xFF),
:: "m" (src1), "m" (uvbuf1), "m" (dstw), "m" (s_xinc2>>16), "m" (s_xinc2&0xFFFF),
"r" (src2)
: "%eax", "%ebx", "%ecx", "%edi", "%esi"
);
#ifdef HAVE_MMX2
} //if MMX2 cant be used
#endif
#else
for(i=0;i<dstw;i++){
register unsigned int xx=xpos>>8;
register unsigned int xalpha=(xpos&0xFF)>>1;
register unsigned int xx=xpos>>16;
register unsigned int xalpha=(xpos&0xFFFF)>>9;
uvbuf1[i]=(src1[xx]*(xalpha^127)+src1[xx+1]*xalpha);
uvbuf1[i+2048]=(src2[xx]*(xalpha^127)+src2[xx+1]*xalpha);
xpos+=s_xinc2;
}
#endif
}
if(!y0) continue;
}
// Note1: this code can be resticted to n*8 (or n*16) width lines to simplify optimization...
// Re: Note1: ok n*4 for now
......@@ -489,47 +554,47 @@ s_xinc&= -2; //clear last bit or uv and y might be shifted relative to each othe
"1: \n\t"\
"movq (%0, %%eax, 2), %%mm0 \n\t" /*buf0[eax]*/\
"movq (%1, %%eax, 2), %%mm1 \n\t" /*buf1[eax]*/\
"psubw %%mm1, %%mm0 \n\t" /* buf0[eax] - buf1[eax]*/\
"pmulhw %%mm6, %%mm0 \n\t" /* (buf0[eax] - buf1[eax])yalpha1>>16*/\
"psraw $7, %%mm1 \n\t" /* buf0[eax] - buf1[eax] >>7*/\
"paddw %%mm0, %%mm1 \n\t" /* buf0[eax]yalpha1 + buf1[eax](1-yalpha1) >>16*/\
"psubw w10, %%mm1 \n\t" /* Y-16*/\
"psllw $3, %%mm1 \n\t" /* (y-16)*8*/\
"pmulhw yCoeff, %%mm1 \n\t"\
\
"movq (%2, %%eax,2), %%mm2 \n\t" /* uvbuf0[eax]*/\
"movq (%3, %%eax,2), %%mm3 \n\t" /* uvbuf1[eax]*/\
"psubw %%mm1, %%mm0 \n\t" /* buf0[eax] - buf1[eax]*/\
"psubw %%mm3, %%mm2 \n\t" /* uvbuf0[eax] - uvbuf1[eax]*/\
"pmulhw %%mm6, %%mm0 \n\t" /* (buf0[eax] - buf1[eax])yalpha1>>16*/\
"pmulhw %%mm5, %%mm2 \n\t" /* (uvbuf0[eax] - uvbuf1[eax])uvalpha1>>16*/\
"psraw $7, %%mm1 \n\t" /* buf0[eax] - buf1[eax] >>7*/\
"movq 4096(%2, %%eax,2), %%mm4 \n\t" /* uvbuf0[eax+2048]*/\
"psraw $7, %%mm3 \n\t" /* uvbuf0[eax] - uvbuf1[eax] >>7*/\
"paddw %%mm0, %%mm1 \n\t" /* buf0[eax]yalpha1 + buf1[eax](1-yalpha1) >>16*/\
"movq 4096(%3, %%eax,2), %%mm0 \n\t" /* uvbuf1[eax+2048]*/\
"paddw %%mm2, %%mm3 \n\t" /* uvbuf0[eax]uvalpha1 - uvbuf1[eax](1-uvalpha1)*/\
"psubw %%mm0, %%mm4 \n\t" /* uvbuf0[eax+2048] - uvbuf1[eax+2048]*/\
"psubw w10, %%mm1 \n\t" /* Y-16*/\
"psubw w80, %%mm3 \n\t" /* (U-128)*/\
"psllw $3, %%mm1 \n\t" /* (y-16)*8*/\
"psllw $3, %%mm3 \n\t" /*(U-128)8*/\
"pmulhw yCoeff, %%mm1 \n\t"\
\
\
"movq 4096(%2, %%eax,2), %%mm4 \n\t" /* uvbuf0[eax+2048]*/\
"movq 4096(%3, %%eax,2), %%mm0 \n\t" /* uvbuf1[eax+2048]*/\
"psubw %%mm0, %%mm4 \n\t" /* uvbuf0[eax+2048] - uvbuf1[eax+2048]*/\
"pmulhw %%mm5, %%mm4 \n\t" /* (uvbuf0[eax+2048] - uvbuf1[eax+2048])uvalpha1>>16*/\
"movq %%mm3, %%mm2 \n\t" /* (U-128)8*/\
"pmulhw ubCoeff, %%mm3 \n\t"\
"psraw $7, %%mm0 \n\t" /* uvbuf0[eax+2048] - uvbuf1[eax+2048] >>7*/\
"pmulhw ugCoeff, %%mm2 \n\t"\
"paddw %%mm4, %%mm0 \n\t" /* uvbuf0[eax+2048]uvalpha1 - uvbuf1[eax+2048](1-uvalpha1)*/\
"psubw w80, %%mm0 \n\t" /* (V-128)*/\
"psllw $3, %%mm0 \n\t" /* (V-128)8*/\
\
"movq %%mm3, %%mm2 \n\t" /* (U-128)8*/\
"pmulhw ubCoeff, %%mm3 \n\t"\
"paddw %%mm1, %%mm3 \n\t" /* B*/\
\
"movq %%mm0, %%mm4 \n\t" /* (V-128)8*/\
"pmulhw vrCoeff, %%mm0 \n\t"\
"pmulhw vgCoeff, %%mm4 \n\t"\
"paddw %%mm1, %%mm3 \n\t" /* B*/\
"paddw %%mm1, %%mm0 \n\t" /* R*/\
"packuswb %%mm3, %%mm3 \n\t"\
\
"pmulhw ugCoeff, %%mm2 \n\t"\
"pmulhw vgCoeff, %%mm4 \n\t"\
"packuswb %%mm0, %%mm0 \n\t"\
"paddw %%mm4, %%mm2 \n\t"\
"paddw %%mm2, %%mm1 \n\t" /* G*/\
\
"packuswb %%mm3, %%mm3 \n\t"\
"packuswb %%mm0, %%mm0 \n\t"\
"packuswb %%mm1, %%mm1 \n\t"
YSCALEYUV2RGB
......@@ -610,9 +675,11 @@ YSCALEYUV2RGB
asm volatile(
YSCALEYUV2RGB
#ifdef DITHER16BPP
"paddusb g16Dither, %%mm1 \n\t"
"paddusb b16Dither, %%mm0 \n\t"
"paddusb b16Dither, %%mm3 \n\t"
#endif
"punpcklbw %%mm7, %%mm1 \n\t" // 0G0G0G0G
"punpcklbw %%mm7, %%mm3 \n\t" // 0B0B0B0B
"punpcklbw %%mm7, %%mm0 \n\t" // 0R0R0R0R
......@@ -699,8 +766,6 @@ YSCALEYUV2RGB
#elif defined (HAVE_MMX)
asm volatile("emms");
#endif
}
......
swscale_template.c
View file @ 61f36e03
// Software scaling and colorspace conversion routines for MPlayer
// Orginal C implementation by ?
// current version mostly by Michael Niedermayer (michaelni@gmx.at)
#include <inttypes.h>
#include "../config.h"
#undef HAVE_MMX2 //code is buggy
//#undef HAVE_MMX2
//#undef HAVE_MMX
//#undef ARCH_X86
#define DITHER16BPP
#define ALT_ERROR
#define RET 0xC3 //near return opcode
/*
NOTES
// temporary storage for 4 yuv lines:
// 16bit for now (mmx likes it more compact)
static uint16_t pix_buf_y[4][2048];
static uint16_t pix_buf_uv[2][2048*2];
known BUGS with known cause (no bugreports please!)
line at the right (c,asm and mmx2)
code reads 1 sample too much (might cause a sig11)
// clipping helper table for C implementations:
static unsigned char clip_table[768];
// yuv->rgb conversion tables:
static int yuvtab_2568[256];
static int yuvtab_3343[256];
static int yuvtab_0c92[256];
static int yuvtab_1a1e[256];
static int yuvtab_40cf[256];
TODO
check alignment off everything
*/
static uint64_t yCoeff= 0x2568256825682568LL;
static uint64_t ubCoeff= 0x3343334333433343LL;
......@@ -46,11 +47,27 @@ static uint64_t g16Mask= 0x07E007E007E007E0LL;
static uint64_t r16Mask= 0xF800F800F800F800LL;
static uint64_t temp0;
// temporary storage for 4 yuv lines:
// 16bit for now (mmx likes it more compact)
static uint16_t pix_buf_y[4][2048];
static uint16_t pix_buf_uv[2][2048*2];
// clipping helper table for C implementations:
static unsigned char clip_table[768];
// yuv->rgb conversion tables:
static int yuvtab_2568[256];
static int yuvtab_3343[256];
static int yuvtab_0c92[256];
static int yuvtab_1a1e[256];
static int yuvtab_40cf[256];
static uint8_t funnyYCode[10000];
static uint8_t funnyUVCode[10000];
// *** bilinear scaling and yuv->rgb conversion of yv12 slices:
// *** Note: it's called multiple times while decoding a frame, first time y==0
// *** Designed to upscale, but may work for downscale too.
......@@ -64,27 +81,43 @@ void SwScale_YV12slice_brg24(unsigned char* srcptr[],int stride[], int y, int h,
//static int s_yinc=(vo_dga_src_height<<16)/vo_dga_vp_height;
//static int s_xinc=(vo_dga_src_width<<8)/vo_dga_vp_width;
unsigned int s_xinc2=s_xinc>>1;
unsigned int s_xinc2;
static int s_srcypos;
static int s_srcypos; // points to the dst Pixels center in the source (0 is the center of pixel 0,0 in src)
static int s_ypos;
// last horzontally interpolated lines, used to avoid unnecessary calculations
static int s_last_ypos;
static int s_last_y1pos;
static int static_dstw;
#ifdef HAVE_MMX2
// used to detect a horizontal size change
static int old_dstw= -1;
static int old_s_xinc= -1;
// difference between the requested xinc and the required one for the mmx2 routine
static int s_xinc_diff=0;
static int s_xinc2_diff=0;
#endif
int canMMX2BeUsed;
s_xinc&= -2; //clear last bit or uv and y might be shifted relative to each other
// we need that precission at least for the mmx2 code
s_xinc*= 256;
s_xinc2=s_xinc>>1;
canMMX2BeUsed= (s_xinc <= 0x10000 && (dstw&31)==0) ? 1 : 0;
if(y==0){
s_srcypos=-2*s_yinc;
s_ypos=-2;
s_last_ypos=-2;
s_srcypos= s_yinc/2 - 0x8000;
s_ypos=0;
// force calculation of the horizontal interpolation of the first line
s_last_ypos=-99;
s_last_y1pos=-99;
#ifdef HAVE_MMX2
// cant downscale !!!
if(old_s_xinc != s_xinc || old_dstw!=dstw)
if((old_s_xinc != s_xinc || old_dstw!=dstw) && canMMX2BeUsed)
{
uint8_t *fragment;
int imm8OfPShufW1;
......@@ -102,32 +135,30 @@ s_xinc&= -2; //clear last bit or uv and y might be shifted relative to each othe
//code fragment
// fragmentLength=0;
// printf("%d, %d\n", fragmentLength,imm8OfPShufW1);
asm volatile(
"jmp 9f \n\t"
// Begin
"0: \n\t"
"movq (%%esi, %%ebx), %%mm0 \n\t" //FIXME Alignment
"movq (%%esi), %%mm0 \n\t" //FIXME Alignment
"movq %%mm0, %%mm1 \n\t"
"psrlq $8, %%mm0 \n\t"
"punpcklbw %%mm7, %%mm1 \n\t"
"movq %%mm2, %%mm3 \n\t"
"punpcklbw %%mm7, %%mm0 \n\t"
"addw %%bx, %%cx \n\t" //2*xalpha += (4*s_xinc)&0xFFFF
"pshufw $0xFF, %%mm1, %%mm1 \n\t"
"1: \n\t"
"adcl %%edx, %%esi \n\t" //xx+= (4*s_xinc)>>16 + carry
"pshufw $0xFF, %%mm0, %%mm0 \n\t"
"2: \n\t"
"psrlw $9, %%mm3 \n\t"
"psubw %%mm1, %%mm0 \n\t"
"psraw $1, %%mm0 \n\t"
"pmullw %%mm2, %%mm0 \n\t"
"pmullw %%mm3, %%mm0 \n\t"
"paddw %%mm6, %%mm2 \n\t" // 2*alpha += xpos&0xFFFF
"psllw $7, %%mm1 \n\t"
"paddw %%mm1, %%mm0 \n\t"
"movq %%mm0, (%%edi, %%eax) \n\t"
"paddb %%mm6, %%mm2 \n\t" // 2*alpha += xpos&0xFF
"addb %%ch, %%cl \n\t" //2*xalpha += (4*s_xinc)&0xFF
"adcl %%edx, %%ebx \n\t" //xx+= (4*s_xinc)>>8 + carry
"movq %%mm0, (%%edi, %%eax) \n\t"
"addl $8, %%eax \n\t"
// End
......@@ -147,17 +178,28 @@ s_xinc&= -2; //clear last bit or uv and y might be shifted relative to each othe
);
xpos= xx=xalpha= 0;
//FIXME choose size and or xinc so that they fit exactly
/* choose xinc so that all 8 parts fit exactly
Note: we cannot use just 1 part because it would not fit in the code cache */
s_xinc2_diff= -((((s_xinc2*(dstw/8))&0xFFFF))/(dstw/8))+10;
// s_xinc_diff= -((((s_xinc*(dstw/8))&0xFFFF))/(dstw/8));
#ifdef ALT_ERROR
s_xinc2_diff+= ((0x10000/(dstw/8)));
#endif
s_xinc_diff= s_xinc2_diff*2;
s_xinc2+= s_xinc2_diff;
s_xinc+= s_xinc_diff;
for(i=0; i<dstw/8; i++)
{
int xx=xpos>>8;
int xx=xpos>>16;
if((i&3) == 0)
{
int a=0;
int b=((xpos+s_xinc)>>8) - xx;
int c=((xpos+s_xinc*2)>>8) - xx;
int d=((xpos+s_xinc*3)>>8) - xx;
int b=((xpos+s_xinc)>>16) - xx;
int c=((xpos+s_xinc*2)>>16) - xx;
int d=((xpos+s_xinc*3)>>16) - xx;
memcpy(funnyYCode + fragmentLength*i/4, fragment, fragmentLength);
......@@ -174,14 +216,14 @@ s_xinc&= -2; //clear last bit or uv and y might be shifted relative to each othe
//FIXME choose size and or xinc so that they fit exactly
for(i=0; i<dstw/8; i++)
{
int xx=xpos>>8;
int xx=xpos>>16;
if((i&3) == 0)
{
int a=0;
int b=((xpos+s_xinc2)>>8) - xx;
int c=((xpos+s_xinc2*2)>>8) - xx;
int d=((xpos+s_xinc2*3)>>8) - xx;
int b=((xpos+s_xinc2)>>16) - xx;
int c=((xpos+s_xinc2*2)>>16) - xx;
int d=((xpos+s_xinc2*3)>>16) - xx;
memcpy(funnyUVCode + fragmentLength*i/4, fragment, fragmentLength);
......@@ -197,86 +239,117 @@ s_xinc&= -2; //clear last bit or uv and y might be shifted relative to each othe
}
#endif
if(canMMX2BeUsed)
{
s_xinc+= s_xinc_diff;
s_xinc2+= s_xinc2_diff;
}
#endif // HAVE_MMX2
} // reset counters
while(1){
unsigned char *dest=dstptr+dststride*s_ypos;
int y0=2+(s_srcypos>>16);
int y1=1+(s_srcypos>>17);
int y0=(s_srcypos + 0xFFFF)>>16; // first luminance source line number below the dst line
// points to the dst Pixels center in the source (0 is the center of pixel 0,0 in src)
int srcuvpos= s_srcypos + s_yinc/2 - 0x8000;
int y1=(srcuvpos + 0x1FFFF)>>17; // first chrominance source line number below the dst line
int yalpha=(s_srcypos&0xFFFF)>>7;
int yalpha1=yalpha^511;
int uvalpha=((s_srcypos>>1)&0xFFFF)>>7;
int uvalpha=(srcuvpos&0x1FFFF)>>8;
int uvalpha1=uvalpha^511;
uint16_t *buf0=pix_buf_y[y0&3];
uint16_t *buf1=pix_buf_y[((y0+1)&3)];
uint16_t *uvbuf0=pix_buf_uv[y1&1];
uint16_t *uvbuf1=pix_buf_uv[(y1&1)^1];
uint16_t *buf0=pix_buf_y[y0&1]; // top line of the interpolated slice
uint16_t *buf1=pix_buf_y[((y0+1)&1)]; // bottom line of the interpolated slice
uint16_t *uvbuf0=pix_buf_uv[y1&1]; // top line of the interpolated slice
uint16_t *uvbuf1=pix_buf_uv[(y1+1)&1]; // bottom line of the interpolated slice
int i;
if(y0>=y+h) break;
// if this is before the first line than use only the first src line
if(y0==0) buf0= buf1;
if(y1==0) uvbuf0= uvbuf1; // yes we do have to check this, its not the same as y0==0
if(y0>=y+h) break; // FIXME wrong, skips last lines, but they are dupliactes anyway
// if this is after the last line than use only the last src line
if(y0>=y+h)
{
buf1= buf0;
s_last_ypos=y0;
}
if(y1>=(y+h)/2)
{
uvbuf1= uvbuf0;
s_last_y1pos=y1;
}
s_ypos++; s_srcypos+=s_yinc;
//only interpolate the src line horizontally if we didnt do it allready
if(s_last_ypos!=y0){
unsigned char *src=srcptr[0]+(y0-y)*stride[0];
unsigned int xpos=0;
s_last_ypos=y0;
// *** horizontal scale Y line to temp buffer
// this loop should be rewritten in MMX assembly!!!!
#ifdef ARCH_X86
#ifdef HAVE_MMX2
if(canMMX2BeUsed)
{
asm volatile(
"pxor %%mm7, %%mm7 \n\t"
"pxor %%mm2, %%mm2 \n\t" // 2*xalpha
"movd %5, %%mm6 \n\t" // s_xinc&0xFF
"movd %5, %%mm6 \n\t" // s_xinc&0xFFFF
"punpcklwd %%mm6, %%mm6 \n\t"
"punpcklwd %%mm6, %%mm6 \n\t"
"movq %%mm6, %%mm2 \n\t"
"psllq $16, %%mm2 \n\t"
"paddb %%mm6, %%mm2 \n\t"
"paddw %%mm6, %%mm2 \n\t"
"psllq $16, %%mm2 \n\t"
"paddb %%mm6, %%mm2 \n\t"
"paddw %%mm6, %%mm2 \n\t"
"psllq $16, %%mm2 \n\t" //0,t,2t,3t t=s_xinc&0xFF
"movq %%mm2, temp0 \n\t"
"movd %4, %%mm6 \n\t" //(s_xinc*4)&0xFF
"movd %4, %%mm6 \n\t" //(s_xinc*4)&0xFFFF
"punpcklwd %%mm6, %%mm6 \n\t"
"punpcklwd %%mm6, %%mm6 \n\t"
"xorl %%eax, %%eax \n\t" // i
"xorl %%ebx, %%ebx \n\t" // xx
"movl %0, %%esi \n\t" // src
"movl %1, %%edi \n\t" // buf1
"movl %3, %%edx \n\t" // (s_xinc*4)>>8
"movl %3, %%edx \n\t" // (s_xinc*4)>>16
"xorl %%ecx, %%ecx \n\t"
"movb %4, %%ch \n\t" // (s_xinc*4)&0xFF
// "int $3\n\t"
"xorl %%ebx, %%ebx \n\t"
"movw %4, %%bx \n\t" // (s_xinc*4)&0xFFFF
// "int $3\n\t"
"call funnyYCode \n\t"
"movq temp0, %%mm2 \n\t"
"xorb %%cl, %%cl \n\t"
"xorl %%ecx, %%ecx \n\t"
"call funnyYCode \n\t"
"movq temp0, %%mm2 \n\t"
"xorb %%cl, %%cl \n\t"
"xorl %%ecx, %%ecx \n\t"
"call funnyYCode \n\t"
"movq temp0, %%mm2 \n\t"
"xorb %%cl, %%cl \n\t"
"xorl %%ecx, %%ecx \n\t"
"call funnyYCode \n\t"
"movq temp0, %%mm2 \n\t"
"xorb %%cl, %%cl \n\t"
"xorl %%ecx, %%ecx \n\t"
"call funnyYCode \n\t"
"movq temp0, %%mm2 \n\t"
"xorb %%cl, %%cl \n\t"
"xorl %%ecx, %%ecx \n\t"
"call funnyYCode \n\t"
"movq temp0, %%mm2 \n\t"
"xorb %%cl, %%cl \n\t"
"xorl %%ecx, %%ecx \n\t"
"call funnyYCode \n\t"
"movq temp0, %%mm2 \n\t"
"xorb %%cl, %%cl \n\t"
"xorl %%ecx, %%ecx \n\t"
"call funnyYCode \n\t"
:: "m" (src), "m" (buf1), "m" (dstw), "m" ((s_xinc*4)>>8),
"m" ((s_xinc*4)&0xFF), "m" (s_xinc&0xFF)
:: "m" (src), "m" (buf1), "m" (dstw), "m" ((s_xinc*4)>>16),
"m" ((s_xinc*4)&0xFFFF), "m" (s_xinc&0xFFFF)
: "%eax", "%ebx", "%ecx", "%edx", "%esi", "%edi"
);
#elif defined (ARCH_X86)
}
else
{
#endif
//NO MMX just normal asm ... FIXME try/write funny MMX2 variant
//FIXME add prefetch
asm volatile(
......@@ -288,24 +361,24 @@ s_xinc&= -2; //clear last bit or uv and y might be shifted relative to each othe
"movzbl 1(%0, %%ebx), %%esi \n\t" //src[xx+1]
"subl %%edi, %%esi \n\t" //src[xx+1] - src[xx]
"imull %%ecx, %%esi \n\t" //(src[xx+1] - src[xx])*2*xalpha
"shll $8, %%edi \n\t"
"shll $16, %%edi \n\t"
"addl %%edi, %%esi \n\t" //src[xx+1]*2*xalpha + src[xx]*(1-2*xalpha)
"movl %1, %%edi \n\t"
"shrl $1, %%esi \n\t"
"shrl $9, %%esi \n\t"
"movw %%si, (%%edi, %%eax, 2) \n\t"
"addb %4, %%cl \n\t" //2*xalpha += s_xinc&0xFF
"addw %4, %%cx \n\t" //2*xalpha += s_xinc&0xFF
"adcl %3, %%ebx \n\t" //xx+= s_xinc>>8 + carry
"movzbl (%0, %%ebx), %%edi \n\t" //src[xx]
"movzbl 1(%0, %%ebx), %%esi \n\t" //src[xx+1]
"subl %%edi, %%esi \n\t" //src[xx+1] - src[xx]
"imull %%ecx, %%esi \n\t" //(src[xx+1] - src[xx])*2*xalpha
"shll $8, %%edi \n\t"
"shll $16, %%edi \n\t"
"addl %%edi, %%esi \n\t" //src[xx+1]*2*xalpha + src[xx]*(1-2*xalpha)
"movl %1, %%edi \n\t"
"shrl $1, %%esi \n\t"
"shrl $9, %%esi \n\t"
"movw %%si, 2(%%edi, %%eax, 2) \n\t"
"addb %4, %%cl \n\t" //2*xalpha += s_xinc&0xFF
"addw %4, %%cx \n\t" //2*xalpha += s_xinc&0xFF
"adcl %3, %%ebx \n\t" //xx+= s_xinc>>8 + carry
......@@ -314,106 +387,96 @@ s_xinc&= -2; //clear last bit or uv and y might be shifted relative to each othe
" jb 1b \n\t"
:: "r" (src), "m" (buf1), "m" (dstw), "m" (s_xinc>>8), "m" (s_xinc&0xFF)
:: "r" (src), "m" (buf1), "m" (dstw), "m" (s_xinc>>16), "m" (s_xinc&0xFFFF)
: "%eax", "%ebx", "%ecx", "%edi", "%esi"
);
#ifdef HAVE_MMX2
} //if MMX2 cant be used
#endif
#else
for(i=0;i<dstw;i++){
register unsigned int xx=xpos>>8;
register unsigned int xalpha=(xpos&0xFF)>>1;
register unsigned int xx=xpos>>16;
register unsigned int xalpha=(xpos&0xFFFF)>>9;
buf1[i]=(src[xx]*(xalpha^127)+src[xx+1]*xalpha);
xpos+=s_xinc;
}
#endif
}
// *** horizontal scale U and V lines to temp buffer
if(!(y0&1)){
if(s_last_y1pos!=y1){
unsigned char *src1=srcptr[1]+(y1-y/2)*stride[1];
unsigned char *src2=srcptr[2]+(y1-y/2)*stride[2];
xpos=0;
// this loop should be rewritten in MMX assembly!!!!
int xpos=0;
s_last_y1pos= y1;
#ifdef ARCH_X86
#ifdef HAVE_MMX2
if(canMMX2BeUsed)
{
asm volatile(
"pxor %%mm7, %%mm7 \n\t"
"pxor %%mm2, %%mm2 \n\t" // 2*xalpha
"movd %5, %%mm6 \n\t" // s_xinc&0xFF
"movd %5, %%mm6 \n\t" // s_xinc&0xFFFF
"punpcklwd %%mm6, %%mm6 \n\t"
"punpcklwd %%mm6, %%mm6 \n\t"
"movq %%mm6, %%mm2 \n\t"
"psllq $16, %%mm2 \n\t"
"paddb %%mm6, %%mm2 \n\t"
"paddw %%mm6, %%mm2 \n\t"
"psllq $16, %%mm2 \n\t"
"paddb %%mm6, %%mm2 \n\t"
"psllq $16, %%mm2 \n\t" //0,t,2t,3t t=s_xinc&0xFF
"paddw %%mm6, %%mm2 \n\t"
"psllq $16, %%mm2 \n\t" //0,t,2t,3t t=s_xinc&0xFFFF
"movq %%mm2, temp0 \n\t"
"movd %4, %%mm6 \n\t" //(s_xinc*4)&0xFF
"movd %4, %%mm6 \n\t" //(s_xinc*4)&0xFFFF
"punpcklwd %%mm6, %%mm6 \n\t"
"punpcklwd %%mm6, %%mm6 \n\t"
"xorl %%eax, %%eax \n\t" // i
"xorl %%ebx, %%ebx \n\t" // xx
"movl %0, %%esi \n\t" // src
"movl %1, %%edi \n\t" // buf1
"movl %3, %%edx \n\t" // (s_xinc*4)>>8
"movl %3, %%edx \n\t" // (s_xinc*4)>>16
"xorl %%ecx, %%ecx \n\t"
"movb %4, %%ch \n\t" // (s_xinc*4)&0xFF
"xorl %%ebx, %%ebx \n\t"
"movw %4, %%bx \n\t" // (s_xinc*4)&0xFFFF
// "int $3\n\t"
"call funnyUVCode \n\t"
"movq temp0, %%mm2 \n\t"
"xorb %%cl, %%cl \n\t"
"call funnyUVCode \n\t"
"movq temp0, %%mm2 \n\t"
"xorb %%cl, %%cl \n\t"
"call funnyUVCode \n\t"
"movq temp0, %%mm2 \n\t"
"xorb %%cl, %%cl \n\t"
"call funnyUVCode \n\t"
"movq temp0, %%mm2 \n\t"
"xorb %%cl, %%cl \n\t"
"call funnyUVCode \n\t"
"movq temp0, %%mm2 \n\t"
"xorb %%cl, %%cl \n\t"
"call funnyUVCode \n\t"
"movq temp0, %%mm2 \n\t"
"xorb %%cl, %%cl \n\t"
"call funnyUVCode \n\t"
"movq temp0, %%mm2 \n\t"
"xorb %%cl, %%cl \n\t"
"call funnyUVCode \n\t"
#define FUNNYUVCODE \
"call funnyUVCode \n\t"\
"movq temp0, %%mm2 \n\t"\
"xorl %%ecx, %%ecx \n\t"
FUNNYUVCODE
FUNNYUVCODE
FUNNYUVCODE
FUNNYUVCODE
FUNNYUVCODE
FUNNYUVCODE
FUNNYUVCODE
FUNNYUVCODE
"xorl %%eax, %%eax \n\t" // i
"xorl %%ebx, %%ebx \n\t" // xx
"movl %6, %%esi \n\t" // src
"movl %1, %%edi \n\t" // buf1
"addl $4096, %%edi \n\t"
"call funnyUVCode \n\t"
"movq temp0, %%mm2 \n\t"
"xorb %%cl, %%cl \n\t"
"call funnyUVCode \n\t"
"movq temp0, %%mm2 \n\t"
"xorb %%cl, %%cl \n\t"
"call funnyUVCode \n\t"
"movq temp0, %%mm2 \n\t"
"xorb %%cl, %%cl \n\t"
"call funnyUVCode \n\t"
"movq temp0, %%mm2 \n\t"
"xorb %%cl, %%cl \n\t"
"call funnyUVCode \n\t"
"movq temp0, %%mm2 \n\t"
"xorb %%cl, %%cl \n\t"
"call funnyUVCode \n\t"
"movq temp0, %%mm2 \n\t"
"xorb %%cl, %%cl \n\t"
"call funnyUVCode \n\t"
"movq temp0, %%mm2 \n\t"
"xorb %%cl, %%cl \n\t"
"call funnyUVCode \n\t"
FUNNYUVCODE
FUNNYUVCODE
FUNNYUVCODE
FUNNYUVCODE
FUNNYUVCODE
FUNNYUVCODE
FUNNYUVCODE
FUNNYUVCODE
:: "m" (src1), "m" (uvbuf1), "m" (dstw), "m" ((s_xinc2*4)>>8),
"m" ((s_xinc2*4)&0xFF), "m" (s_xinc2&0xFF), "m" (src2)
:: "m" (src1), "m" (uvbuf1), "m" (dstw), "m" ((s_xinc2*4)>>16),
"m" ((s_xinc2*4)&0xFFFF), "m" (s_xinc2&0xFFFF), "m" (src2)
: "%eax", "%ebx", "%ecx", "%edx", "%esi", "%edi"
);
#elif defined (ARCH_X86)
}
else
{
#endif
asm volatile(
"xorl %%eax, %%eax \n\t" // i
"xorl %%ebx, %%ebx \n\t" // xx
......@@ -424,45 +487,47 @@ s_xinc&= -2; //clear last bit or uv and y might be shifted relative to each othe
"movzbl 1(%%esi, %%ebx), %%esi \n\t" //src[xx+1]
"subl %%edi, %%esi \n\t" //src[xx+1] - src[xx]
"imull %%ecx, %%esi \n\t" //(src[xx+1] - src[xx])*2*xalpha
"shll $8, %%edi \n\t"
"shll $16, %%edi \n\t"
"addl %%edi, %%esi \n\t" //src[xx+1]*2*xalpha + src[xx]*(1-2*xalpha)
"movl %1, %%edi \n\t"
"shrl $1, %%esi \n\t"
"shrl $9, %%esi \n\t"
"movw %%si, (%%edi, %%eax, 2) \n\t"
"movzbl (%5, %%ebx), %%edi \n\t" //src[xx]
"movzbl 1(%5, %%ebx), %%esi \n\t" //src[xx+1]
"subl %%edi, %%esi \n\t" //src[xx+1] - src[xx]
"imull %%ecx, %%esi \n\t" //(src[xx+1] - src[xx])*2*xalpha
"shll $8, %%edi \n\t"
"shll $16, %%edi \n\t"
"addl %%edi, %%esi \n\t" //src[xx+1]*2*xalpha + src[xx]*(1-2*xalpha)
"movl %1, %%edi \n\t"
"shrl $1, %%esi \n\t"
"shrl $9, %%esi \n\t"
"movw %%si, 4096(%%edi, %%eax, 2)\n\t"
"addb %4, %%cl \n\t" //2*xalpha += s_xinc&0xFF
"addw %4, %%cx \n\t" //2*xalpha += s_xinc&0xFF
"adcl %3, %%ebx \n\t" //xx+= s_xinc>>8 + carry
"addl $1, %%eax \n\t"
"cmpl %2, %%eax \n\t"
" jb 1b \n\t"
:: "m" (src1), "m" (uvbuf1), "m" (dstw), "m" (s_xinc2>>8), "m" (s_xinc2&0xFF),
:: "m" (src1), "m" (uvbuf1), "m" (dstw), "m" (s_xinc2>>16), "m" (s_xinc2&0xFFFF),
"r" (src2)
: "%eax", "%ebx", "%ecx", "%edi", "%esi"
);
#ifdef HAVE_MMX2
} //if MMX2 cant be used
#endif
#else
for(i=0;i<dstw;i++){
register unsigned int xx=xpos>>8;
register unsigned int xalpha=(xpos&0xFF)>>1;
register unsigned int xx=xpos>>16;
register unsigned int xalpha=(xpos&0xFFFF)>>9;
uvbuf1[i]=(src1[xx]*(xalpha^127)+src1[xx+1]*xalpha);
uvbuf1[i+2048]=(src2[xx]*(xalpha^127)+src2[xx+1]*xalpha);
xpos+=s_xinc2;
}
#endif
}
if(!y0) continue;
}
// Note1: this code can be resticted to n*8 (or n*16) width lines to simplify optimization...
// Re: Note1: ok n*4 for now
......@@ -489,47 +554,47 @@ s_xinc&= -2; //clear last bit or uv and y might be shifted relative to each othe
"1: \n\t"\
"movq (%0, %%eax, 2), %%mm0 \n\t" /*buf0[eax]*/\
"movq (%1, %%eax, 2), %%mm1 \n\t" /*buf1[eax]*/\
"psubw %%mm1, %%mm0 \n\t" /* buf0[eax] - buf1[eax]*/\
"pmulhw %%mm6, %%mm0 \n\t" /* (buf0[eax] - buf1[eax])yalpha1>>16*/\
"psraw $7, %%mm1 \n\t" /* buf0[eax] - buf1[eax] >>7*/\
"paddw %%mm0, %%mm1 \n\t" /* buf0[eax]yalpha1 + buf1[eax](1-yalpha1) >>16*/\
"psubw w10, %%mm1 \n\t" /* Y-16*/\
"psllw $3, %%mm1 \n\t" /* (y-16)*8*/\
"pmulhw yCoeff, %%mm1 \n\t"\
\
"movq (%2, %%eax,2), %%mm2 \n\t" /* uvbuf0[eax]*/\
"movq (%3, %%eax,2), %%mm3 \n\t" /* uvbuf1[eax]*/\
"psubw %%mm1, %%mm0 \n\t" /* buf0[eax] - buf1[eax]*/\
"psubw %%mm3, %%mm2 \n\t" /* uvbuf0[eax] - uvbuf1[eax]*/\
"pmulhw %%mm6, %%mm0 \n\t" /* (buf0[eax] - buf1[eax])yalpha1>>16*/\
"pmulhw %%mm5, %%mm2 \n\t" /* (uvbuf0[eax] - uvbuf1[eax])uvalpha1>>16*/\
"psraw $7, %%mm1 \n\t" /* buf0[eax] - buf1[eax] >>7*/\
"movq 4096(%2, %%eax,2), %%mm4 \n\t" /* uvbuf0[eax+2048]*/\
"psraw $7, %%mm3 \n\t" /* uvbuf0[eax] - uvbuf1[eax] >>7*/\
"paddw %%mm0, %%mm1 \n\t" /* buf0[eax]yalpha1 + buf1[eax](1-yalpha1) >>16*/\
"movq 4096(%3, %%eax,2), %%mm0 \n\t" /* uvbuf1[eax+2048]*/\
"paddw %%mm2, %%mm3 \n\t" /* uvbuf0[eax]uvalpha1 - uvbuf1[eax](1-uvalpha1)*/\
"psubw %%mm0, %%mm4 \n\t" /* uvbuf0[eax+2048] - uvbuf1[eax+2048]*/\
"psubw w10, %%mm1 \n\t" /* Y-16*/\
"psubw w80, %%mm3 \n\t" /* (U-128)*/\
"psllw $3, %%mm1 \n\t" /* (y-16)*8*/\
"psllw $3, %%mm3 \n\t" /*(U-128)8*/\
"pmulhw yCoeff, %%mm1 \n\t"\
\
\
"movq 4096(%2, %%eax,2), %%mm4 \n\t" /* uvbuf0[eax+2048]*/\
"movq 4096(%3, %%eax,2), %%mm0 \n\t" /* uvbuf1[eax+2048]*/\
"psubw %%mm0, %%mm4 \n\t" /* uvbuf0[eax+2048] - uvbuf1[eax+2048]*/\
"pmulhw %%mm5, %%mm4 \n\t" /* (uvbuf0[eax+2048] - uvbuf1[eax+2048])uvalpha1>>16*/\
"movq %%mm3, %%mm2 \n\t" /* (U-128)8*/\
"pmulhw ubCoeff, %%mm3 \n\t"\
"psraw $7, %%mm0 \n\t" /* uvbuf0[eax+2048] - uvbuf1[eax+2048] >>7*/\
"pmulhw ugCoeff, %%mm2 \n\t"\
"paddw %%mm4, %%mm0 \n\t" /* uvbuf0[eax+2048]uvalpha1 - uvbuf1[eax+2048](1-uvalpha1)*/\
"psubw w80, %%mm0 \n\t" /* (V-128)*/\
"psllw $3, %%mm0 \n\t" /* (V-128)8*/\
\
"movq %%mm3, %%mm2 \n\t" /* (U-128)8*/\
"pmulhw ubCoeff, %%mm3 \n\t"\
"paddw %%mm1, %%mm3 \n\t" /* B*/\
\
"movq %%mm0, %%mm4 \n\t" /* (V-128)8*/\
"pmulhw vrCoeff, %%mm0 \n\t"\
"pmulhw vgCoeff, %%mm4 \n\t"\
"paddw %%mm1, %%mm3 \n\t" /* B*/\
"paddw %%mm1, %%mm0 \n\t" /* R*/\
"packuswb %%mm3, %%mm3 \n\t"\
\
"pmulhw ugCoeff, %%mm2 \n\t"\
"pmulhw vgCoeff, %%mm4 \n\t"\
"packuswb %%mm0, %%mm0 \n\t"\
"paddw %%mm4, %%mm2 \n\t"\
"paddw %%mm2, %%mm1 \n\t" /* G*/\
\
"packuswb %%mm3, %%mm3 \n\t"\
"packuswb %%mm0, %%mm0 \n\t"\
"packuswb %%mm1, %%mm1 \n\t"
YSCALEYUV2RGB
......@@ -610,9 +675,11 @@ YSCALEYUV2RGB
asm volatile(
YSCALEYUV2RGB
#ifdef DITHER16BPP
"paddusb g16Dither, %%mm1 \n\t"
"paddusb b16Dither, %%mm0 \n\t"
"paddusb b16Dither, %%mm3 \n\t"
#endif
"punpcklbw %%mm7, %%mm1 \n\t" // 0G0G0G0G
"punpcklbw %%mm7, %%mm3 \n\t" // 0B0B0B0B
"punpcklbw %%mm7, %%mm0 \n\t" // 0R0R0R0R
......@@ -699,8 +766,6 @@ YSCALEYUV2RGB
#elif defined (HAVE_MMX)
asm volatile("emms");
#endif
}
......
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