diff options
Diffstat (limited to 'src/post/deinterlace/speedy.c')
| -rw-r--r-- | src/post/deinterlace/speedy.c | 220 |
1 files changed, 218 insertions, 2 deletions
diff --git a/src/post/deinterlace/speedy.c b/src/post/deinterlace/speedy.c index 32c8b03e3..4c9d5c0d1 100644 --- a/src/post/deinterlace/speedy.c +++ b/src/post/deinterlace/speedy.c @@ -62,8 +62,9 @@ #include <stdint.h> #endif -#include "attributes.h" -#include "xineutils.h" +#include <xine/attributes.h> +#include <xine/xineutils.h> +#include "xine_mmx.h" #include "speedtools.h" #include "speedy.h" @@ -343,6 +344,89 @@ static unsigned int diff_factor_packed422_scanline_mmx( uint8_t *cur, uint8_t *o } #endif +#if defined(ARCH_X86) || defined(ARCH_X86_64) + +static const sse_t dqwYMask = { uq: { 0x00ff00ff00ff00ffULL, 0x00ff00ff00ff00ffULL }}; +static const sse_t dqwCMask = { uq: { 0xff00ff00ff00ff00ULL, 0xff00ff00ff00ff00ULL }}; + +static unsigned int diff_factor_packed422_scanline_sse2_aligned( uint8_t *cur, uint8_t *old, int width ) +{ + register unsigned int temp; + + width /= 8; + + movdqa_m2r( dqwYMask, xmm1 ); + movd_m2r( BitShift, xmm7 ); + pxor_r2r( xmm0, xmm0 ); + + while( width-- ) { + movdqa_m2r( *cur, xmm4 ); + movdqa_m2r( *old, xmm5 ); + + pand_r2r( xmm1, xmm4 ); + pand_r2r( xmm1, xmm5 ); + + psubw_r2r( xmm5, xmm4 ); /* mm4 = Y1 - Y2 */ + pmaddwd_r2r( xmm4, xmm4 ); /* mm4 = (Y1 - Y2)^2 */ + psrld_r2r( xmm7, xmm4 ); /* divide mm4 by 2^BitShift */ + paddd_r2r( xmm4, xmm0 ); /* keep total in mm0 */ + + cur += 16; + old += 16; + } + + pshufd_r2r(xmm0, xmm1, 0x0e); + paddd_r2r(xmm1, xmm0); + pshufd_r2r(xmm0, xmm1, 0x01); + paddd_r2r(xmm1, xmm0); + + movd_r2a(xmm0, temp); + return temp; +} +#endif + +#if defined(ARCH_X86) || defined(ARCH_X86_64) +static unsigned int diff_factor_packed422_scanline_sse2( uint8_t *cur, uint8_t *old, int width ) +{ + if (0 == (((unsigned int)cur|(unsigned int)old) & 15)) { + return diff_factor_packed422_scanline_sse2_aligned(cur, old, width); + } + + register unsigned int temp; + + width /= 8; + + movdqa_m2r( dqwYMask, xmm1 ); + movd_m2r( BitShift, xmm7 ); + pxor_r2r( xmm0, xmm0 ); + + while( width-- ) { + movdqu_m2r( *cur, xmm4 ); + movdqu_m2r( *old, xmm5 ); + + pand_r2r( xmm1, xmm4 ); + pand_r2r( xmm1, xmm5 ); + + psubw_r2r( xmm5, xmm4 ); /* mm4 = Y1 - Y2 */ + pmaddwd_r2r( xmm4, xmm4 ); /* mm4 = (Y1 - Y2)^2 */ + psrld_r2r( xmm7, xmm4 ); /* divide mm4 by 2^BitShift */ + paddd_r2r( xmm4, xmm0 ); /* keep total in mm0 */ + + cur += 16; + old += 16; + } + + pshufd_r2r(xmm0, xmm1, 0x0e); + paddd_r2r(xmm1, xmm0); + pshufd_r2r(xmm0, xmm1, 0x01); + paddd_r2r(xmm1, xmm0); + + movd_r2a(xmm0, temp); + + return temp; +} +#endif + #define ABS(a) (((a) < 0)?-(a):(a)) #if defined(ARCH_X86) || defined(ARCH_X86_64) @@ -716,6 +800,130 @@ static void vfilter_chroma_332_packed422_scanline_mmx( uint8_t *output, int widt } #endif +#if defined(ARCH_X86) || defined(ARCH_X86_64) +static void vfilter_chroma_332_packed422_scanline_sse2_aligned( uint8_t *output, int width, + uint8_t *m, uint8_t *t, uint8_t *b ) +{ + int i; + + // Get width in bytes. + width *= 2; + i = width / 16; + width -= i * 16; + + movdqa_m2r( dqwYMask, xmm7 ); + movdqa_m2r( dqwCMask, xmm6 ); + + while( i-- ) { + movdqa_m2r ( *t, xmm0 ); + movdqa_m2r ( *b, xmm1 ); + movdqa_m2r ( *m, xmm2 ); + + movdqa_r2r ( xmm2, xmm3 ); + pand_r2r ( xmm7, xmm3 ); + + pand_r2r ( xmm6, xmm0 ); + pand_r2r ( xmm6, xmm1 ); + pand_r2r ( xmm6, xmm2 ); + + psrlq_i2r ( 8, xmm0 ); + psrlq_i2r ( 7, xmm1 ); + psrlq_i2r ( 8, xmm2 ); + + movdqa_r2r ( xmm0, xmm4 ); + movdqa_r2r ( xmm2, xmm5 ); + psllw_i2r ( 1, xmm4 ); + psllw_i2r ( 1, xmm5 ); + paddw_r2r ( xmm4, xmm0 ); + paddw_r2r ( xmm5, xmm2 ); + + paddw_r2r ( xmm0, xmm2 ); + paddw_r2r ( xmm1, xmm2 ); + + psllw_i2r ( 5, xmm2 ); + pand_r2r ( xmm6, xmm2 ); + + por_r2r ( xmm3, xmm2 ); + + movdqa_r2m( xmm2, *output ); + output += 16; + t += 16; + b += 16; + m += 16; + } + output++; t++; b++; m++; + while( width-- ) { + *output = (3 * *t + 3 * *m + 2 * *b) >> 3; + output +=2; t+=2; b+=2; m+=2; + } +} +#endif + +#if defined(ARCH_X86) || defined(ARCH_X86_64) +static void vfilter_chroma_332_packed422_scanline_sse2( uint8_t *output, int width, + uint8_t *m, uint8_t *t, uint8_t *b ) +{ + int i; + + if (0 == (((unsigned int)output|(unsigned int)m|(unsigned int)t|(unsigned int)b) & 15)) { + vfilter_chroma_332_packed422_scanline_sse2_aligned(output, width, m, t, b); + return; + } + + // Get width in bytes. + width *= 2; + i = width / 16; + width -= i * 16; + + movdqa_m2r( dqwYMask, xmm7 ); + movdqa_m2r( dqwCMask, xmm6 ); + + while( i-- ) { + movdqu_m2r ( *t, xmm0 ); + movdqu_m2r ( *b, xmm1 ); + movdqu_m2r ( *m, xmm2 ); + + movdqa_r2r ( xmm2, xmm3 ); + pand_r2r ( xmm7, xmm3 ); + + pand_r2r ( xmm6, xmm0 ); + pand_r2r ( xmm6, xmm1 ); + pand_r2r ( xmm6, xmm2 ); + + psrlq_i2r ( 8, xmm0 ); + psrlq_i2r ( 7, xmm1 ); + psrlq_i2r ( 8, xmm2 ); + + movdqa_r2r ( xmm0, xmm4 ); + movdqa_r2r ( xmm2, xmm5 ); + psllw_i2r ( 1, xmm4 ); + psllw_i2r ( 1, xmm5 ); + paddw_r2r ( xmm4, xmm0 ); + paddw_r2r ( xmm5, xmm2 ); + + paddw_r2r ( xmm0, xmm2 ); + paddw_r2r ( xmm1, xmm2 ); + + psllw_i2r ( 5, xmm2 ); + pand_r2r ( xmm6, xmm2 ); + + por_r2r ( xmm3, xmm2 ); + + movdqu_r2m( xmm2, *output ); + output += 16; + t += 16; + b += 16; + m += 16; + } + output++; t++; b++; m++; + while( width-- ) { + *output = (3 * *t + 3 * *m + 2 * *b) >> 3; + output +=2; t+=2; b+=2; m+=2; + } +} +#endif + + static void vfilter_chroma_332_packed422_scanline_c( uint8_t *output, int width, uint8_t *m, uint8_t *t, uint8_t *b ) { @@ -2458,6 +2666,14 @@ void setup_speedy_calls( uint32_t accel, int verbose ) printf( "speedycode: No MMX or MMXEXT support detected, using C fallbacks.\n" ); } } + + if( speedy_accel & MM_ACCEL_X86_SSE2 ) { + if( verbose ) { + printf( "speedycode: Using SSE2 optimized functions.\n" ); + } + diff_factor_packed422_scanline = diff_factor_packed422_scanline_sse2; + vfilter_chroma_332_packed422_scanline = vfilter_chroma_332_packed422_scanline_sse2; + } #endif } |