do not use sse for yuv420p to rgb conversion when width is not multiple of 16

This commit is contained in:
jfmcarreira committed 2015-02-28 03:51:49 +00:00
1 parent 01c0bb5769
commit 9bbcfab99b
1 file changed
+119 -113
+119 -113
View File
@@ -119,145 +119,151 @@ Void bufferFromFrameYUV420p( Pel ***in, Pel* out, UInt width, UInt height )
}
Void fillRGBbufferYUV420p( Pel*** in, UChar* out, UInt width, UInt height )
{
#ifndef USE_SSE
fillRGBbufferYUV( in, out, width, height, 1, 1 );
#else
uint8_t *yp = in[LUMA][0];
uint8_t *up = in[CHROMA_U][0];
uint8_t *vp = in[CHROMA_V][0];
uint32_t suv = width >> 1;
uint32_t *rgb = ( uint32_t* )out;
__m128i y0r0, y0r1, u0, v0;
__m128i y00r0, y01r0, y00r1, y01r1;
__m128i u00, u01, v00, v01;
__m128i rv00, rv01, gu00, gu01, gv00, gv01, bu00, bu01;
__m128i r00, r01, g00, g01, b00, b01;
__m128i rgb0123, rgb4567, rgb89ab, rgbcdef;
__m128i gbgb;
__m128i ysub, uvsub;
__m128i zero, facy, facrv, facgu, facgv, facbu;
__m128i *srcy128r0, *srcy128r1;
__m128i *dstrgb128r0, *dstrgb128r1;
__m64 *srcu64, *srcv64;
UInt x, y;
ysub = _mm_set1_epi32( 0x00100010 );
uvsub = _mm_set1_epi32( 0x00800080 );
facy = _mm_set1_epi32( 0x004a004a );
facrv = _mm_set1_epi32( 0x00660066 );
facgu = _mm_set1_epi32( 0x00190019 );
facgv = _mm_set1_epi32( 0x00340034 );
facbu = _mm_set1_epi32( 0x00810081 );
zero = _mm_set1_epi32( 0x00000000 );
for( y = 0; y < height; y += 2 )
#ifdef USE_SSE
if( ( width % 16 ) == 0 )
{
uint8_t *yp = in[LUMA][0];
uint8_t *up = in[CHROMA_U][0];
uint8_t *vp = in[CHROMA_V][0];
uint32_t suv = width >> 1;
uint32_t *rgb = ( uint32_t* )out;
srcy128r0 = ( __m128i * )( yp + width * y );
srcy128r1 = ( __m128i * )( yp + width * y + width );
srcu64 = ( __m64 * )( up + suv * ( y / 2 ) );
srcv64 = ( __m64 * )( vp + suv * ( y / 2 ) );
__m128i y0r0, y0r1, u0, v0;
__m128i y00r0, y01r0, y00r1, y01r1;
__m128i u00, u01, v00, v01;
__m128i rv00, rv01, gu00, gu01, gv00, gv01, bu00, bu01;
__m128i r00, r01, g00, g01, b00, b01;
__m128i rgb0123, rgb4567, rgb89ab, rgbcdef;
__m128i gbgb;
__m128i ysub, uvsub;
__m128i zero, facy, facrv, facgu, facgv, facbu;
__m128i *srcy128r0, *srcy128r1;
__m128i *dstrgb128r0, *dstrgb128r1;
__m64 *srcu64, *srcv64;
UInt x, y;
dstrgb128r0 = ( __m128i * )( rgb + width * y );
dstrgb128r1 = ( __m128i * )( rgb + width * y + width );
ysub = _mm_set1_epi32( 0x00100010 );
uvsub = _mm_set1_epi32( 0x00800080 );
for( x = 0; x < width; x += 16 )
facy = _mm_set1_epi32( 0x004a004a );
facrv = _mm_set1_epi32( 0x00660066 );
facgu = _mm_set1_epi32( 0x00190019 );
facgv = _mm_set1_epi32( 0x00340034 );
facbu = _mm_set1_epi32( 0x00810081 );
zero = _mm_set1_epi32( 0x00000000 );
for( y = 0; y < height; y += 2 )
{
u0 = _mm_loadl_epi64( ( __m128i * )srcu64 );
srcu64++;
v0 = _mm_loadl_epi64( ( __m128i * )srcv64 );
srcv64++;
srcy128r0 = ( __m128i * )( yp + width * y );
srcy128r1 = ( __m128i * )( yp + width * y + width );
srcu64 = ( __m64 * )( up + suv * ( y / 2 ) );
srcv64 = ( __m64 * )( vp + suv * ( y / 2 ) );
y0r0 = _mm_load_si128( srcy128r0++ );
y0r1 = _mm_load_si128( srcy128r1++ );
dstrgb128r0 = ( __m128i * )( rgb + width * y );
dstrgb128r1 = ( __m128i * )( rgb + width * y + width );
// constant y factors
y00r0 = _mm_mullo_epi16( _mm_sub_epi16( _mm_unpacklo_epi8( y0r0, zero ), ysub ), facy );
y01r0 = _mm_mullo_epi16( _mm_sub_epi16( _mm_unpackhi_epi8( y0r0, zero ), ysub ), facy );
y00r1 = _mm_mullo_epi16( _mm_sub_epi16( _mm_unpacklo_epi8( y0r1, zero ), ysub ), facy );
y01r1 = _mm_mullo_epi16( _mm_sub_epi16( _mm_unpackhi_epi8( y0r1, zero ), ysub ), facy );
for( x = 0; x < width; x += 16 )
{
// expand u and v so they're aligned with y values
u0 = _mm_unpacklo_epi8( u0, zero );
u00 = _mm_sub_epi16( _mm_unpacklo_epi16( u0, u0 ), uvsub );
u01 = _mm_sub_epi16( _mm_unpackhi_epi16( u0, u0 ), uvsub );
u0 = _mm_loadl_epi64( ( __m128i * )srcu64 );
srcu64++;
v0 = _mm_loadl_epi64( ( __m128i * )srcv64 );
srcv64++;
v0 = _mm_unpacklo_epi8( v0, zero );
v00 = _mm_sub_epi16( _mm_unpacklo_epi16( v0, v0 ), uvsub );
v01 = _mm_sub_epi16( _mm_unpackhi_epi16( v0, v0 ), uvsub );
y0r0 = _mm_load_si128( srcy128r0++ );
y0r1 = _mm_load_si128( srcy128r1++ );
// common factors on both rows.
rv00 = _mm_mullo_epi16( facrv, v00 );
rv01 = _mm_mullo_epi16( facrv, v01 );
gu00 = _mm_mullo_epi16( facgu, u00 );
gu01 = _mm_mullo_epi16( facgu, u01 );
gv00 = _mm_mullo_epi16( facgv, v00 );
gv01 = _mm_mullo_epi16( facgv, v01 );
bu00 = _mm_mullo_epi16( facbu, u00 );
bu01 = _mm_mullo_epi16( facbu, u01 );
// constant y factors
y00r0 = _mm_mullo_epi16( _mm_sub_epi16( _mm_unpacklo_epi8( y0r0, zero ), ysub ), facy );
y01r0 = _mm_mullo_epi16( _mm_sub_epi16( _mm_unpackhi_epi8( y0r0, zero ), ysub ), facy );
y00r1 = _mm_mullo_epi16( _mm_sub_epi16( _mm_unpacklo_epi8( y0r1, zero ), ysub ), facy );
y01r1 = _mm_mullo_epi16( _mm_sub_epi16( _mm_unpackhi_epi8( y0r1, zero ), ysub ), facy );
// row 0
r00 = _mm_srai_epi16( _mm_add_epi16( y00r0, rv00 ), 6 );
r01 = _mm_srai_epi16( _mm_add_epi16( y01r0, rv01 ), 6 );
g00 = _mm_srai_epi16( _mm_sub_epi16( _mm_sub_epi16( y00r0, gu00 ), gv00 ), 6 );
g01 = _mm_srai_epi16( _mm_sub_epi16( _mm_sub_epi16( y01r0, gu01 ), gv01 ), 6 );
b00 = _mm_srai_epi16( _mm_add_epi16( y00r0, bu00 ), 6 );
b01 = _mm_srai_epi16( _mm_add_epi16( y01r0, bu01 ), 6 );
// expand u and v so they're aligned with y values
u0 = _mm_unpacklo_epi8( u0, zero );
u00 = _mm_sub_epi16( _mm_unpacklo_epi16( u0, u0 ), uvsub );
u01 = _mm_sub_epi16( _mm_unpackhi_epi16( u0, u0 ), uvsub );
r00 = _mm_packus_epi16( r00, r01 );// rrrr.. saturated
g00 = _mm_packus_epi16( g00, g01 );// gggg.. saturated
b00 = _mm_packus_epi16( b00, b01 );// bbbb.. saturated
v0 = _mm_unpacklo_epi8( v0, zero );
v00 = _mm_sub_epi16( _mm_unpacklo_epi16( v0, v0 ), uvsub );
v01 = _mm_sub_epi16( _mm_unpackhi_epi16( v0, v0 ), uvsub );
r01 = _mm_unpacklo_epi8( r00, zero );// 0r0r..
gbgb = _mm_unpacklo_epi8( b00, g00 );// gbgb..
rgb0123 = _mm_unpacklo_epi16( gbgb, r01 );// 0rgb0rgb..
rgb4567 = _mm_unpackhi_epi16( gbgb, r01 );// 0rgb0rgb..
// common factors on both rows.
rv00 = _mm_mullo_epi16( facrv, v00 );
rv01 = _mm_mullo_epi16( facrv, v01 );
gu00 = _mm_mullo_epi16( facgu, u00 );
gu01 = _mm_mullo_epi16( facgu, u01 );
gv00 = _mm_mullo_epi16( facgv, v00 );
gv01 = _mm_mullo_epi16( facgv, v01 );
bu00 = _mm_mullo_epi16( facbu, u00 );
bu01 = _mm_mullo_epi16( facbu, u01 );
r01 = _mm_unpackhi_epi8( r00, zero );
gbgb = _mm_unpackhi_epi8( b00, g00 );
rgb89ab = _mm_unpacklo_epi16( gbgb, r01 );
rgbcdef = _mm_unpackhi_epi16( gbgb, r01 );
// row 0
r00 = _mm_srai_epi16( _mm_add_epi16( y00r0, rv00 ), 6 );
r01 = _mm_srai_epi16( _mm_add_epi16( y01r0, rv01 ), 6 );
g00 = _mm_srai_epi16( _mm_sub_epi16( _mm_sub_epi16( y00r0, gu00 ), gv00 ), 6 );
g01 = _mm_srai_epi16( _mm_sub_epi16( _mm_sub_epi16( y01r0, gu01 ), gv01 ), 6 );
b00 = _mm_srai_epi16( _mm_add_epi16( y00r0, bu00 ), 6 );
b01 = _mm_srai_epi16( _mm_add_epi16( y01r0, bu01 ), 6 );
_mm_store_si128( dstrgb128r0++, rgb0123 );
_mm_store_si128( dstrgb128r0++, rgb4567 );
_mm_store_si128( dstrgb128r0++, rgb89ab );
_mm_store_si128( dstrgb128r0++, rgbcdef );
r00 = _mm_packus_epi16( r00, r01 ); // rrrr.. saturated
g00 = _mm_packus_epi16( g00, g01 ); // gggg.. saturated
b00 = _mm_packus_epi16( b00, b01 ); // bbbb.. saturated
// row 1
r00 = _mm_srai_epi16( _mm_add_epi16( y00r1, rv00 ), 6 );
r01 = _mm_srai_epi16( _mm_add_epi16( y01r1, rv01 ), 6 );
g00 = _mm_srai_epi16( _mm_sub_epi16( _mm_sub_epi16( y00r1, gu00 ), gv00 ), 6 );
g01 = _mm_srai_epi16( _mm_sub_epi16( _mm_sub_epi16( y01r1, gu01 ), gv01 ), 6 );
b00 = _mm_srai_epi16( _mm_add_epi16( y00r1, bu00 ), 6 );
b01 = _mm_srai_epi16( _mm_add_epi16( y01r1, bu01 ), 6 );
r01 = _mm_unpacklo_epi8( r00, zero ); // 0r0r..
gbgb = _mm_unpacklo_epi8( b00, g00 ); // gbgb..
rgb0123 = _mm_unpacklo_epi16( gbgb, r01 ); // 0rgb0rgb..
rgb4567 = _mm_unpackhi_epi16( gbgb, r01 ); // 0rgb0rgb..
r00 = _mm_packus_epi16( r00, r01 );// rrrr.. saturated
g00 = _mm_packus_epi16( g00, g01 );// gggg.. saturated
b00 = _mm_packus_epi16( b00, b01 );// bbbb.. saturated
r01 = _mm_unpackhi_epi8( r00, zero );
gbgb = _mm_unpackhi_epi8( b00, g00 );
rgb89ab = _mm_unpacklo_epi16( gbgb, r01 );
rgbcdef = _mm_unpackhi_epi16( gbgb, r01 );
r01 = _mm_unpacklo_epi8( r00, zero );// 0r0r..
gbgb = _mm_unpacklo_epi8( b00, g00 );// gbgb..
rgb0123 = _mm_unpacklo_epi16( gbgb, r01 );// 0rgb0rgb..
rgb4567 = _mm_unpackhi_epi16( gbgb, r01 );// 0rgb0rgb..
_mm_store_si128( dstrgb128r0++, rgb0123 );
_mm_store_si128( dstrgb128r0++, rgb4567 );
_mm_store_si128( dstrgb128r0++, rgb89ab );
_mm_store_si128( dstrgb128r0++, rgbcdef );
r01 = _mm_unpackhi_epi8( r00, zero );
gbgb = _mm_unpackhi_epi8( b00, g00 );
rgb89ab = _mm_unpacklo_epi16( gbgb, r01 );
rgbcdef = _mm_unpackhi_epi16( gbgb, r01 );
// row 1
r00 = _mm_srai_epi16( _mm_add_epi16( y00r1, rv00 ), 6 );
r01 = _mm_srai_epi16( _mm_add_epi16( y01r1, rv01 ), 6 );
g00 = _mm_srai_epi16( _mm_sub_epi16( _mm_sub_epi16( y00r1, gu00 ), gv00 ), 6 );
g01 = _mm_srai_epi16( _mm_sub_epi16( _mm_sub_epi16( y01r1, gu01 ), gv01 ), 6 );
b00 = _mm_srai_epi16( _mm_add_epi16( y00r1, bu00 ), 6 );
b01 = _mm_srai_epi16( _mm_add_epi16( y01r1, bu01 ), 6 );
_mm_store_si128( dstrgb128r1++, rgb0123 );
_mm_store_si128( dstrgb128r1++, rgb4567 );
_mm_store_si128( dstrgb128r1++, rgb89ab );
_mm_store_si128( dstrgb128r1++, rgbcdef );
r00 = _mm_packus_epi16( r00, r01 ); // rrrr.. saturated
g00 = _mm_packus_epi16( g00, g01 ); // gggg.. saturated
b00 = _mm_packus_epi16( b00, b01 ); // bbbb.. saturated
r01 = _mm_unpacklo_epi8( r00, zero ); // 0r0r..
gbgb = _mm_unpacklo_epi8( b00, g00 ); // gbgb..
rgb0123 = _mm_unpacklo_epi16( gbgb, r01 ); // 0rgb0rgb..
rgb4567 = _mm_unpackhi_epi16( gbgb, r01 ); // 0rgb0rgb..
r01 = _mm_unpackhi_epi8( r00, zero );
gbgb = _mm_unpackhi_epi8( b00, g00 );
rgb89ab = _mm_unpacklo_epi16( gbgb, r01 );
rgbcdef = _mm_unpackhi_epi16( gbgb, r01 );
_mm_store_si128( dstrgb128r1++, rgb0123 );
_mm_store_si128( dstrgb128r1++, rgb4567 );
_mm_store_si128( dstrgb128r1++, rgb89ab );
_mm_store_si128( dstrgb128r1++, rgbcdef );
}
}
}
else
#endif
{
fillRGBbufferYUV( in, out, width, height, 1, 1 );
}
}
PlaYUVerFramePelFormat yuv420p =
{
"YUV420p",