/*****************************************************************************
* deblock.c: deblocking
*****************************************************************************
- * Copyright (C) 2003-2011 x264 project
+ * Copyright (C) 2003-2015 x264 project
*
* Authors: Laurent Aimar <fenrir@via.ecp.fr>
* Loren Merritt <lorenm@u.washington.edu>
* Fiona Glaser <fiona@x264.com>
- * Henrik Gramner <hengar-6@student.ltu.se>
+ * Henrik Gramner <henrik@gramner.com>
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
#define tc0_table(x) i_tc0_table[(x)+24]
/* From ffmpeg */
-static ALWAYS_INLINE void deblock_edge_luma_c( pixel *pix, int xstride, int alpha, int beta, int8_t tc0 )
+static ALWAYS_INLINE void deblock_edge_luma_c( pixel *pix, intptr_t xstride, int alpha, int beta, int8_t tc0 )
{
int p2 = pix[-3*xstride];
int p1 = pix[-2*xstride];
pix[ 0*xstride] = x264_clip_pixel( q0 - delta ); /* q0' */
}
}
-static inline void deblock_luma_c( pixel *pix, int xstride, int ystride, int alpha, int beta, int8_t *tc0 )
+static inline void deblock_luma_c( pixel *pix, intptr_t xstride, intptr_t ystride, int alpha, int beta, int8_t *tc0 )
{
for( int i = 0; i < 4; i++ )
{
deblock_edge_luma_c( pix, xstride, alpha, beta, tc0[i] );
}
}
-static void deblock_h_luma_mbaff_c( pixel *pix, int stride, int alpha, int beta, int8_t *tc0 )
+static void deblock_h_luma_mbaff_c( pixel *pix, intptr_t stride, int alpha, int beta, int8_t *tc0 )
{
for( int d = 0; d < 8; d++, pix += stride )
deblock_edge_luma_c( pix, 1, alpha, beta, tc0[d>>1] );
}
-static void deblock_v_luma_c( pixel *pix, int stride, int alpha, int beta, int8_t *tc0 )
+static void deblock_v_luma_c( pixel *pix, intptr_t stride, int alpha, int beta, int8_t *tc0 )
{
deblock_luma_c( pix, stride, 1, alpha, beta, tc0 );
}
-static void deblock_h_luma_c( pixel *pix, int stride, int alpha, int beta, int8_t *tc0 )
+static void deblock_h_luma_c( pixel *pix, intptr_t stride, int alpha, int beta, int8_t *tc0 )
{
deblock_luma_c( pix, 1, stride, alpha, beta, tc0 );
}
-static ALWAYS_INLINE void deblock_edge_chroma_c( pixel *pix, int xstride, int alpha, int beta, int8_t tc )
+static ALWAYS_INLINE void deblock_edge_chroma_c( pixel *pix, intptr_t xstride, int alpha, int beta, int8_t tc )
{
int p1 = pix[-2*xstride];
int p0 = pix[-1*xstride];
pix[ 0*xstride] = x264_clip_pixel( q0 - delta ); /* q0' */
}
}
-static ALWAYS_INLINE void deblock_chroma_c( pixel *pix, int height, int xstride, int ystride, int alpha, int beta, int8_t *tc0 )
+static ALWAYS_INLINE void deblock_chroma_c( pixel *pix, int height, intptr_t xstride, intptr_t ystride, int alpha, int beta, int8_t *tc0 )
{
for( int i = 0; i < 4; i++ )
{
deblock_edge_chroma_c( pix, xstride, alpha, beta, tc0[i] );
}
}
-static void deblock_h_chroma_mbaff_c( pixel *pix, int stride, int alpha, int beta, int8_t *tc0 )
+static void deblock_h_chroma_mbaff_c( pixel *pix, intptr_t stride, int alpha, int beta, int8_t *tc0 )
{
- for( int i = 0; i < 4; i++, pix += stride )
- deblock_edge_chroma_c( pix, 2, alpha, beta, tc0[i] );
+ deblock_chroma_c( pix, 1, 2, stride, alpha, beta, tc0 );
}
-static void deblock_h_chroma_422_mbaff_c( pixel *pix, int stride, int alpha, int beta, int8_t *tc0 )
-{
- for( int i = 0; i < 8; i++, pix += stride )
- deblock_edge_chroma_c( pix, 2, alpha, beta, tc0[i>>1] );
-}
-static void deblock_v_chroma_c( pixel *pix, int stride, int alpha, int beta, int8_t *tc0 )
+static void deblock_v_chroma_c( pixel *pix, intptr_t stride, int alpha, int beta, int8_t *tc0 )
{
deblock_chroma_c( pix, 2, stride, 2, alpha, beta, tc0 );
}
-static void deblock_h_chroma_c( pixel *pix, int stride, int alpha, int beta, int8_t *tc0 )
+static void deblock_h_chroma_c( pixel *pix, intptr_t stride, int alpha, int beta, int8_t *tc0 )
{
deblock_chroma_c( pix, 2, 2, stride, alpha, beta, tc0 );
}
-static void deblock_h_chroma_422_c( pixel *pix, int stride, int alpha, int beta, int8_t *tc0 )
+static void deblock_h_chroma_422_c( pixel *pix, intptr_t stride, int alpha, int beta, int8_t *tc0 )
{
deblock_chroma_c( pix, 4, 2, stride, alpha, beta, tc0 );
}
-static ALWAYS_INLINE void deblock_edge_luma_intra_c( pixel *pix, int xstride, int alpha, int beta )
+static ALWAYS_INLINE void deblock_edge_luma_intra_c( pixel *pix, intptr_t xstride, int alpha, int beta )
{
int p2 = pix[-3*xstride];
int p1 = pix[-2*xstride];
}
}
}
-static inline void deblock_luma_intra_c( pixel *pix, int xstride, int ystride, int alpha, int beta )
+static inline void deblock_luma_intra_c( pixel *pix, intptr_t xstride, intptr_t ystride, int alpha, int beta )
{
for( int d = 0; d < 16; d++, pix += ystride )
deblock_edge_luma_intra_c( pix, xstride, alpha, beta );
}
-static void deblock_h_luma_intra_mbaff_c( pixel *pix, int ystride, int alpha, int beta )
+static void deblock_h_luma_intra_mbaff_c( pixel *pix, intptr_t ystride, int alpha, int beta )
{
for( int d = 0; d < 8; d++, pix += ystride )
deblock_edge_luma_intra_c( pix, 1, alpha, beta );
}
-static void deblock_v_luma_intra_c( pixel *pix, int stride, int alpha, int beta )
+static void deblock_v_luma_intra_c( pixel *pix, intptr_t stride, int alpha, int beta )
{
deblock_luma_intra_c( pix, stride, 1, alpha, beta );
}
-static void deblock_h_luma_intra_c( pixel *pix, int stride, int alpha, int beta )
+static void deblock_h_luma_intra_c( pixel *pix, intptr_t stride, int alpha, int beta )
{
deblock_luma_intra_c( pix, 1, stride, alpha, beta );
}
-static ALWAYS_INLINE void deblock_edge_chroma_intra_c( pixel *pix, int xstride, int alpha, int beta )
+static ALWAYS_INLINE void deblock_edge_chroma_intra_c( pixel *pix, intptr_t xstride, int alpha, int beta )
{
int p1 = pix[-2*xstride];
int p0 = pix[-1*xstride];
pix[ 0*xstride] = (2*q1 + q0 + p1 + 2) >> 2; /* q0' */
}
}
-static ALWAYS_INLINE void deblock_chroma_intra_c( pixel *pix, int width, int height, int xstride, int ystride, int alpha, int beta )
+static ALWAYS_INLINE void deblock_chroma_intra_c( pixel *pix, int width, int height, intptr_t xstride, intptr_t ystride, int alpha, int beta )
{
for( int d = 0; d < height; d++, pix += ystride-2 )
for( int e = 0; e < width; e++, pix++ )
deblock_edge_chroma_intra_c( pix, xstride, alpha, beta );
}
-static void deblock_h_chroma_intra_mbaff_c( pixel *pix, int stride, int alpha, int beta )
-{
- for( int i = 0; i < 4; i++, pix += stride )
- deblock_edge_chroma_intra_c( pix, 2, alpha, beta );
-}
-static void deblock_h_chroma_422_intra_mbaff_c( pixel *pix, int stride, int alpha, int beta )
+static void deblock_h_chroma_intra_mbaff_c( pixel *pix, intptr_t stride, int alpha, int beta )
{
- for( int i = 0; i < 8; i++, pix += stride )
- deblock_edge_chroma_intra_c( pix, 2, alpha, beta );
+ deblock_chroma_intra_c( pix, 2, 4, 2, stride, alpha, beta );
}
-static void deblock_v_chroma_intra_c( pixel *pix, int stride, int alpha, int beta )
+static void deblock_v_chroma_intra_c( pixel *pix, intptr_t stride, int alpha, int beta )
{
deblock_chroma_intra_c( pix, 1, 16, stride, 2, alpha, beta );
}
-static void deblock_h_chroma_intra_c( pixel *pix, int stride, int alpha, int beta )
+static void deblock_h_chroma_intra_c( pixel *pix, intptr_t stride, int alpha, int beta )
{
deblock_chroma_intra_c( pix, 2, 8, 2, stride, alpha, beta );
}
-static void deblock_h_chroma_422_intra_c( pixel *pix, int stride, int alpha, int beta )
+static void deblock_h_chroma_422_intra_c( pixel *pix, intptr_t stride, int alpha, int beta )
{
deblock_chroma_intra_c( pix, 2, 16, 2, stride, alpha, beta );
}
}
}
-static ALWAYS_INLINE void deblock_edge( x264_t *h, pixel *pix, int i_stride, uint8_t bS[4], int i_qp, int a, int b, int b_chroma, x264_deblock_inter_t pf_inter )
+static ALWAYS_INLINE void deblock_edge( x264_t *h, pixel *pix, intptr_t i_stride, uint8_t bS[4], int i_qp,
+ int a, int b, int b_chroma, x264_deblock_inter_t pf_inter )
{
int index_a = i_qp + a;
int index_b = i_qp + b;
pf_inter( pix, i_stride, alpha, beta, tc );
}
-static ALWAYS_INLINE void deblock_edge_intra( x264_t *h, pixel *pix, int i_stride, uint8_t bS[4], int i_qp, int a, int b, int b_chroma, x264_deblock_intra_t pf_intra )
+static ALWAYS_INLINE void deblock_edge_intra( x264_t *h, pixel *pix, intptr_t i_stride, uint8_t bS[4], int i_qp,
+ int a, int b, int b_chroma, x264_deblock_intra_t pf_intra )
{
int index_a = i_qp + a;
int index_b = i_qp + b;
x264_macroblock_cache_load_neighbours_deblock( h, mb_x, mb_y );
int mb_xy = h->mb.i_mb_xy;
- int transform_8x8 = h->mb.mb_transform_size[h->mb.i_mb_xy];
+ int transform_8x8 = h->mb.mb_transform_size[mb_xy];
int intra_cur = IS_INTRA( h->mb.type[mb_xy] );
- uint8_t (*bs)[8][4] = h->deblock_strength[mb_y&1][mb_x];
+ uint8_t (*bs)[8][4] = h->deblock_strength[mb_y&1][h->param.b_sliced_threads?mb_xy:mb_x];
pixel *pixy = h->fdec->plane[0] + 16*mb_y*stridey + 16*mb_x;
pixel *pixuv = h->fdec->plane[1] + chroma_height*mb_y*strideuv + 16*mb_x;
{
deblock_edge_intra( h, pixy, 2*stridey, bs[0][0], luma_qp[0], a, b, 0, luma_intra_deblock );
deblock_edge_intra( h, pixuv, 2*strideuv, bs[0][0], chroma_qp[0], a, b, c, chroma_intra_deblock );
- deblock_edge_intra( h, pixuv + uvdiff, 2*strideuv, bs[0][0], chroma_qp[0], a, b, c, chroma_intra_deblock );
+ if( chroma444 )
+ deblock_edge_intra( h, pixuv + uvdiff, 2*strideuv, bs[0][0], chroma_qp[0], a, b, c, chroma_intra_deblock );
}
else
{
deblock_edge( h, pixy, 2*stridey, bs[0][0], luma_qp[0], a, b, 0, luma_deblock );
deblock_edge( h, pixuv, 2*strideuv, bs[0][0], chroma_qp[0], a, b, c, chroma_deblock );
- deblock_edge( h, pixuv + uvdiff, 2*strideuv, bs[0][0], chroma_qp[0], a, b, c, chroma_deblock );
+ if( chroma444 )
+ deblock_edge( h, pixuv + uvdiff, 2*strideuv, bs[0][0], chroma_qp[0], a, b, c, chroma_deblock );
}
int offy = MB_INTERLACED ? 4 : 0;
{
deblock_edge_intra( h, pixy + (stridey<<offy), 2*stridey, bs[0][4], luma_qp[1], a, b, 0, luma_intra_deblock );
deblock_edge_intra( h, pixuv + (strideuv<<offuv), 2*strideuv, bs[0][4], chroma_qp[1], a, b, c, chroma_intra_deblock );
- deblock_edge_intra( h, pixuv + uvdiff + (strideuv<<offuv), 2*strideuv, bs[0][4], chroma_qp[1], a, b, c, chroma_intra_deblock );
+ if( chroma444 )
+ deblock_edge_intra( h, pixuv + uvdiff + (strideuv<<offuv), 2*strideuv, bs[0][4], chroma_qp[1], a, b, c, chroma_intra_deblock );
}
else
{
deblock_edge( h, pixy + (stridey<<offy), 2*stridey, bs[0][4], luma_qp[1], a, b, 0, luma_deblock );
deblock_edge( h, pixuv + (strideuv<<offuv), 2*strideuv, bs[0][4], chroma_qp[1], a, b, c, chroma_deblock );
- deblock_edge( h, pixuv + uvdiff + (strideuv<<offuv), 2*strideuv, bs[0][4], chroma_qp[1], a, b, c, chroma_deblock );
+ if( chroma444 )
+ deblock_edge( h, pixuv + uvdiff + (strideuv<<offuv), 2*strideuv, bs[0][4], chroma_qp[1], a, b, c, chroma_deblock );
}
}
else
int qp_left = (qp + qpl + 1) >> 1;
int qpc_left = (qpc + h->chroma_qp_table[qpl] + 1) >> 1;
int intra_left = IS_INTRA( h->mb.type[h->mb.i_mb_xy-1] );
+ int intra_deblock = intra_cur || intra_left;
+
+ /* Any MB that was coded, or that analysis decided to skip, has quality commensurate with its QP.
+ * But if deblocking affects neighboring MBs that were force-skipped, blur might accumulate there.
+ * So reset their effective QP to max, to indicate that lack of guarantee. */
+ if( h->fdec->mb_info && M32( bs[0][0] ) )
+ {
+#define RESET_EFFECTIVE_QP(xy) h->fdec->effective_qp[xy] |= 0xff * !!(h->fdec->mb_info[xy] & X264_MBINFO_CONSTANT);
+ RESET_EFFECTIVE_QP(mb_xy);
+ RESET_EFFECTIVE_QP(h->mb.i_mb_left_xy[0]);
+ }
- if( intra_cur || intra_left )
+ if( intra_deblock )
FILTER( _intra, 0, 0, qp_left, qpc_left );
else
FILTER( , 0, 0, qp_left, qpc_left );
int qp_top = (qp + qpt + 1) >> 1;
int qpc_top = (qpc + h->chroma_qp_table[qpt] + 1) >> 1;
int intra_top = IS_INTRA( h->mb.type[h->mb.i_mb_top_xy] );
+ int intra_deblock = intra_cur || intra_top;
- if( (!b_interlaced || (!MB_INTERLACED && !h->mb.field[h->mb.i_mb_top_xy]))
- && (intra_cur || intra_top) )
+ /* This edge has been modified, reset effective qp to max. */
+ if( h->fdec->mb_info && M32( bs[1][0] ) )
+ {
+ RESET_EFFECTIVE_QP(mb_xy);
+ RESET_EFFECTIVE_QP(h->mb.i_mb_top_xy);
+ }
+
+ if( (!b_interlaced || (!MB_INTERLACED && !h->mb.field[h->mb.i_mb_top_xy])) && intra_deblock )
{
FILTER( _intra, 1, 0, qp_top, qpc_top );
}
else
{
- if( intra_cur || intra_top )
+ if( intra_deblock )
M32( bs[1][0] ) = 0x03030303;
FILTER( , 1, 0, qp_top, qpc_top );
}
* support analysis partitions smaller than 16x16
* deblock chroma for 4:2:0/4:2:2
* handle duplicate refs correctly
- * handle cavlc+8x8dct correctly
*/
void x264_macroblock_deblock( x264_t *h )
{
if( (h->mb.i_partition == D_16x16 && !h->mb.i_cbp_luma && !intra_cur) || qp <= qp_thresh )
return;
- uint8_t (*bs)[8][4] = h->deblock_strength[h->mb.i_mb_y&1][h->mb.i_mb_x];
+ uint8_t (*bs)[8][4] = h->mb.cache.deblock_strength;
if( intra_cur )
{
- memset( &bs[0][1], 3, 3*4*sizeof(uint8_t) );
- memset( &bs[1][1], 3, 3*4*sizeof(uint8_t) );
+ M32( bs[0][1] ) = 0x03030303;
+ M64( bs[0][2] ) = 0x0303030303030303ULL;
+ M32( bs[1][1] ) = 0x03030303;
+ M64( bs[1][2] ) = 0x0303030303030303ULL;
}
else
h->loopf.deblock_strength( h->mb.cache.non_zero_count, h->mb.cache.ref, h->mb.cache.mv,
}
#if HAVE_MMX
-void x264_deblock_v_luma_sse2( pixel *pix, int stride, int alpha, int beta, int8_t *tc0 );
-void x264_deblock_v_luma_avx ( pixel *pix, int stride, int alpha, int beta, int8_t *tc0 );
-void x264_deblock_h_luma_sse2( pixel *pix, int stride, int alpha, int beta, int8_t *tc0 );
-void x264_deblock_h_luma_avx ( pixel *pix, int stride, int alpha, int beta, int8_t *tc0 );
-void x264_deblock_v_chroma_sse2( pixel *pix, int stride, int alpha, int beta, int8_t *tc0 );
-void x264_deblock_v_chroma_avx ( pixel *pix, int stride, int alpha, int beta, int8_t *tc0 );
-void x264_deblock_h_chroma_sse2( pixel *pix, int stride, int alpha, int beta, int8_t *tc0 );
-void x264_deblock_h_chroma_avx ( pixel *pix, int stride, int alpha, int beta, int8_t *tc0 );
-void x264_deblock_v_luma_intra_sse2( pixel *pix, int stride, int alpha, int beta );
-void x264_deblock_v_luma_intra_avx ( pixel *pix, int stride, int alpha, int beta );
-void x264_deblock_h_luma_intra_sse2( pixel *pix, int stride, int alpha, int beta );
-void x264_deblock_h_luma_intra_avx ( pixel *pix, int stride, int alpha, int beta );
-void x264_deblock_v_chroma_intra_sse2( pixel *pix, int stride, int alpha, int beta );
-void x264_deblock_v_chroma_intra_avx ( pixel *pix, int stride, int alpha, int beta );
-void x264_deblock_h_chroma_intra_sse2( pixel *pix, int stride, int alpha, int beta );
-void x264_deblock_h_chroma_intra_avx ( pixel *pix, int stride, int alpha, int beta );
+void x264_deblock_v_luma_sse2( pixel *pix, intptr_t stride, int alpha, int beta, int8_t *tc0 );
+void x264_deblock_v_luma_avx ( pixel *pix, intptr_t stride, int alpha, int beta, int8_t *tc0 );
+void x264_deblock_h_luma_sse2( pixel *pix, intptr_t stride, int alpha, int beta, int8_t *tc0 );
+void x264_deblock_h_luma_avx ( pixel *pix, intptr_t stride, int alpha, int beta, int8_t *tc0 );
+void x264_deblock_v_chroma_sse2( pixel *pix, intptr_t stride, int alpha, int beta, int8_t *tc0 );
+void x264_deblock_v_chroma_avx ( pixel *pix, intptr_t stride, int alpha, int beta, int8_t *tc0 );
+void x264_deblock_h_chroma_sse2( pixel *pix, intptr_t stride, int alpha, int beta, int8_t *tc0 );
+void x264_deblock_h_chroma_avx ( pixel *pix, intptr_t stride, int alpha, int beta, int8_t *tc0 );
+void x264_deblock_h_chroma_mbaff_sse2( pixel *pix, intptr_t stride, int alpha, int beta, int8_t *tc0 );
+void x264_deblock_h_chroma_mbaff_avx ( pixel *pix, intptr_t stride, int alpha, int beta, int8_t *tc0 );
+void x264_deblock_h_chroma_422_mmx2( pixel *pix, intptr_t stride, int alpha, int beta, int8_t *tc0 );
+void x264_deblock_h_chroma_422_sse2( pixel *pix, intptr_t stride, int alpha, int beta, int8_t *tc0 );
+void x264_deblock_h_chroma_422_avx ( pixel *pix, intptr_t stride, int alpha, int beta, int8_t *tc0 );
+void x264_deblock_v_luma_intra_sse2( pixel *pix, intptr_t stride, int alpha, int beta );
+void x264_deblock_v_luma_intra_avx ( pixel *pix, intptr_t stride, int alpha, int beta );
+void x264_deblock_h_luma_intra_sse2( pixel *pix, intptr_t stride, int alpha, int beta );
+void x264_deblock_h_luma_intra_avx ( pixel *pix, intptr_t stride, int alpha, int beta );
+void x264_deblock_v_chroma_intra_sse2( pixel *pix, intptr_t stride, int alpha, int beta );
+void x264_deblock_v_chroma_intra_avx ( pixel *pix, intptr_t stride, int alpha, int beta );
+void x264_deblock_h_chroma_intra_sse2( pixel *pix, intptr_t stride, int alpha, int beta );
+void x264_deblock_h_chroma_intra_avx ( pixel *pix, intptr_t stride, int alpha, int beta );
+void x264_deblock_h_chroma_422_intra_mmx2( pixel *pix, intptr_t stride, int alpha, int beta );
+void x264_deblock_h_chroma_422_intra_sse2( pixel *pix, intptr_t stride, int alpha, int beta );
+void x264_deblock_h_chroma_422_intra_avx ( pixel *pix, intptr_t stride, int alpha, int beta );
void x264_deblock_strength_mmx2 ( uint8_t nnz[X264_SCAN8_SIZE], int8_t ref[2][X264_SCAN8_LUMA_SIZE],
int16_t mv[2][X264_SCAN8_LUMA_SIZE][2], uint8_t bs[2][8][4],
int mvy_limit, int bframe );
void x264_deblock_strength_avx ( uint8_t nnz[X264_SCAN8_SIZE], int8_t ref[2][X264_SCAN8_LUMA_SIZE],
int16_t mv[2][X264_SCAN8_LUMA_SIZE][2], uint8_t bs[2][8][4],
int mvy_limit, int bframe );
+void x264_deblock_strength_avx2 ( uint8_t nnz[X264_SCAN8_SIZE], int8_t ref[2][X264_SCAN8_LUMA_SIZE],
+ int16_t mv[2][X264_SCAN8_LUMA_SIZE][2], uint8_t bs[2][8][4],
+ int mvy_limit, int bframe );
+
+void x264_deblock_h_chroma_intra_mbaff_mmx2( pixel *pix, intptr_t stride, int alpha, int beta );
+void x264_deblock_h_chroma_intra_mbaff_sse2( pixel *pix, intptr_t stride, int alpha, int beta );
+void x264_deblock_h_chroma_intra_mbaff_avx ( pixel *pix, intptr_t stride, int alpha, int beta );
#if ARCH_X86
-void x264_deblock_h_luma_mmx2( pixel *pix, int stride, int alpha, int beta, int8_t *tc0 );
-void x264_deblock_v8_luma_mmx2( uint8_t *pix, int stride, int alpha, int beta, int8_t *tc0 );
-void x264_deblock_v_chroma_mmx2( pixel *pix, int stride, int alpha, int beta, int8_t *tc0 );
-void x264_deblock_h_chroma_mmx2( pixel *pix, int stride, int alpha, int beta, int8_t *tc0 );
-void x264_deblock_h_luma_intra_mmx2( pixel *pix, int stride, int alpha, int beta );
-void x264_deblock_v8_luma_intra_mmx2( uint8_t *pix, int stride, int alpha, int beta );
-void x264_deblock_v_chroma_intra_mmx2( pixel *pix, int stride, int alpha, int beta );
-void x264_deblock_h_chroma_intra_mmx2( pixel *pix, int stride, int alpha, int beta );
+void x264_deblock_h_luma_mmx2( pixel *pix, intptr_t stride, int alpha, int beta, int8_t *tc0 );
+void x264_deblock_v8_luma_mmx2( uint8_t *pix, intptr_t stride, int alpha, int beta, int8_t *tc0 );
+void x264_deblock_v_chroma_mmx2( pixel *pix, intptr_t stride, int alpha, int beta, int8_t *tc0 );
+void x264_deblock_h_chroma_mmx2( pixel *pix, intptr_t stride, int alpha, int beta, int8_t *tc0 );
+void x264_deblock_h_chroma_mbaff_mmx2( pixel *pix, intptr_t stride, int alpha, int beta, int8_t *tc0 );
+void x264_deblock_h_luma_intra_mmx2( pixel *pix, intptr_t stride, int alpha, int beta );
+void x264_deblock_v8_luma_intra_mmx2( uint8_t *pix, intptr_t stride, int alpha, int beta );
+void x264_deblock_v_chroma_intra_mmx2( pixel *pix, intptr_t stride, int alpha, int beta );
+void x264_deblock_h_chroma_intra_mmx2( pixel *pix, intptr_t stride, int alpha, int beta );
+void x264_deblock_h_chroma_intra_mbaff_mmx2( pixel *pix, intptr_t stride, int alpha, int beta );
#if HIGH_BIT_DEPTH
-void x264_deblock_v_luma_mmx2( pixel *pix, int stride, int alpha, int beta, int8_t *tc0 );
-void x264_deblock_v_luma_intra_mmx2( pixel *pix, int stride, int alpha, int beta );
+void x264_deblock_v_luma_mmx2( pixel *pix, intptr_t stride, int alpha, int beta, int8_t *tc0 );
+void x264_deblock_v_luma_intra_mmx2( pixel *pix, intptr_t stride, int alpha, int beta );
#else
// FIXME this wrapper has a significant cpu cost
-static void x264_deblock_v_luma_mmx2( uint8_t *pix, int stride, int alpha, int beta, int8_t *tc0 )
+static void x264_deblock_v_luma_mmx2( uint8_t *pix, intptr_t stride, int alpha, int beta, int8_t *tc0 )
{
x264_deblock_v8_luma_mmx2( pix, stride, alpha, beta, tc0 );
x264_deblock_v8_luma_mmx2( pix+8, stride, alpha, beta, tc0+2 );
}
-static void x264_deblock_v_luma_intra_mmx2( uint8_t *pix, int stride, int alpha, int beta )
+static void x264_deblock_v_luma_intra_mmx2( uint8_t *pix, intptr_t stride, int alpha, int beta )
{
x264_deblock_v8_luma_intra_mmx2( pix, stride, alpha, beta );
x264_deblock_v8_luma_intra_mmx2( pix+8, stride, alpha, beta );
#endif
#if ARCH_PPC
-void x264_deblock_v_luma_altivec( uint8_t *pix, int stride, int alpha, int beta, int8_t *tc0 );
-void x264_deblock_h_luma_altivec( uint8_t *pix, int stride, int alpha, int beta, int8_t *tc0 );
+void x264_deblock_v_luma_altivec( uint8_t *pix, intptr_t stride, int alpha, int beta, int8_t *tc0 );
+void x264_deblock_h_luma_altivec( uint8_t *pix, intptr_t stride, int alpha, int beta, int8_t *tc0 );
#endif // ARCH_PPC
-#if HAVE_ARMV6
-void x264_deblock_v_luma_neon( uint8_t *, int, int, int, int8_t * );
-void x264_deblock_h_luma_neon( uint8_t *, int, int, int, int8_t * );
-void x264_deblock_v_chroma_neon( uint8_t *, int, int, int, int8_t * );
-void x264_deblock_h_chroma_neon( uint8_t *, int, int, int, int8_t * );
+#if HAVE_ARMV6 || ARCH_AARCH64
+void x264_deblock_v_luma_neon ( uint8_t *pix, intptr_t stride, int alpha, int beta, int8_t *tc0 );
+void x264_deblock_h_luma_neon ( uint8_t *pix, intptr_t stride, int alpha, int beta, int8_t *tc0 );
+void x264_deblock_v_chroma_neon( uint8_t *pix, intptr_t stride, int alpha, int beta, int8_t *tc0 );
+void x264_deblock_h_chroma_neon( uint8_t *pix, intptr_t stride, int alpha, int beta, int8_t *tc0 );
+void x264_deblock_strength_neon( uint8_t nnz[X264_SCAN8_SIZE], int8_t ref[2][X264_SCAN8_LUMA_SIZE],
+ int16_t mv[2][X264_SCAN8_LUMA_SIZE][2], uint8_t bs[2][8][4],
+ int mvy_limit, int bframe );
+void x264_deblock_h_chroma_422_neon( uint8_t *pix, intptr_t stride, int alpha, int beta, int8_t *tc0 );
+void x264_deblock_h_chroma_mbaff_neon( uint8_t *pix, intptr_t stride, int alpha, int beta, int8_t *tc0 );
+void x264_deblock_h_chroma_intra_mbaff_neon( uint8_t *pix, intptr_t stride, int alpha, int beta );
+void x264_deblock_h_chroma_intra_neon( uint8_t *pix, intptr_t stride, int alpha, int beta );
+void x264_deblock_h_chroma_422_intra_neon( uint8_t *pix, intptr_t stride, int alpha, int beta );
+void x264_deblock_v_chroma_intra_neon( uint8_t *pix, intptr_t stride, int alpha, int beta );
+#if ARCH_AARCH64
+void x264_deblock_h_luma_intra_neon( uint8_t *pix, intptr_t stride, int alpha, int beta );
+void x264_deblock_v_luma_intra_neon( uint8_t *pix, intptr_t stride, int alpha, int beta );
+#endif
+#endif
+
+#if !HIGH_BIT_DEPTH
+#if HAVE_MSA
+void x264_deblock_v_luma_msa( uint8_t *pix, intptr_t stride, int alpha, int beta, int8_t *tc0 );
+void x264_deblock_h_luma_msa( uint8_t *pix, intptr_t stride, int alpha, int beta, int8_t *tc0 );
+void x264_deblock_v_chroma_msa( uint8_t *pix, intptr_t stride, int alpha, int beta, int8_t *tc0 );
+void x264_deblock_h_chroma_msa( uint8_t *pix, intptr_t stride, int alpha, int beta, int8_t *tc0 );
+void x264_deblock_v_luma_intra_msa( uint8_t *pix, intptr_t stride, int alpha, int beta );
+void x264_deblock_h_luma_intra_msa( uint8_t *pix, intptr_t stride, int alpha, int beta );
+void x264_deblock_v_chroma_intra_msa( uint8_t *pix, intptr_t stride, int alpha, int beta );
+void x264_deblock_h_chroma_intra_msa( uint8_t *pix, intptr_t stride, int alpha, int beta );
+void x264_deblock_strength_msa( uint8_t nnz[X264_SCAN8_SIZE], int8_t ref[2][X264_SCAN8_LUMA_SIZE],
+ int16_t mv[2][X264_SCAN8_LUMA_SIZE][2], uint8_t bs[2][8][4], int mvy_limit,
+ int bframe );
+#endif
#endif
void x264_deblock_init( int cpu, x264_deblock_function_t *pf, int b_mbaff )
pf->deblock_h_chroma_422_intra = deblock_h_chroma_422_intra_c;
pf->deblock_luma_mbaff = deblock_h_luma_mbaff_c;
pf->deblock_chroma_420_mbaff = deblock_h_chroma_mbaff_c;
- pf->deblock_chroma_422_mbaff = deblock_h_chroma_422_mbaff_c;
pf->deblock_luma_intra_mbaff = deblock_h_luma_intra_mbaff_c;
pf->deblock_chroma_420_intra_mbaff = deblock_h_chroma_intra_mbaff_c;
- pf->deblock_chroma_422_intra_mbaff = deblock_h_chroma_422_intra_mbaff_c;
pf->deblock_strength = deblock_strength_c;
#if HAVE_MMX
pf->deblock_luma[0] = x264_deblock_h_luma_mmx2;
pf->deblock_chroma[1] = x264_deblock_v_chroma_mmx2;
pf->deblock_h_chroma_420 = x264_deblock_h_chroma_mmx2;
+ pf->deblock_chroma_420_mbaff = x264_deblock_h_chroma_mbaff_mmx2;
+ pf->deblock_h_chroma_422 = x264_deblock_h_chroma_422_mmx2;
+ pf->deblock_h_chroma_422_intra = x264_deblock_h_chroma_422_intra_mmx2;
pf->deblock_luma_intra[1] = x264_deblock_v_luma_intra_mmx2;
pf->deblock_luma_intra[0] = x264_deblock_h_luma_intra_mmx2;
pf->deblock_chroma_intra[1] = x264_deblock_v_chroma_intra_mmx2;
pf->deblock_h_chroma_420_intra = x264_deblock_h_chroma_intra_mmx2;
+ pf->deblock_chroma_420_intra_mbaff = x264_deblock_h_chroma_intra_mbaff_mmx2;
+#endif
+#if !HIGH_BIT_DEPTH
+ pf->deblock_chroma_420_intra_mbaff = x264_deblock_h_chroma_intra_mbaff_mmx2;
#endif
pf->deblock_strength = x264_deblock_strength_mmx2;
if( cpu&X264_CPU_SSE2 )
{
pf->deblock_strength = x264_deblock_strength_sse2;
+ pf->deblock_h_chroma_420 = x264_deblock_h_chroma_sse2;
+ pf->deblock_h_chroma_422 = x264_deblock_h_chroma_422_sse2;
+ pf->deblock_h_chroma_422_intra = x264_deblock_h_chroma_422_intra_sse2;
+ pf->deblock_chroma_420_mbaff = x264_deblock_h_chroma_mbaff_sse2;
+ pf->deblock_luma[1] = x264_deblock_v_luma_sse2;
+ pf->deblock_luma[0] = x264_deblock_h_luma_sse2;
+ pf->deblock_luma_intra[1] = x264_deblock_v_luma_intra_sse2;
+ pf->deblock_luma_intra[0] = x264_deblock_h_luma_intra_sse2;
if( !(cpu&X264_CPU_STACK_MOD4) )
{
- pf->deblock_luma[1] = x264_deblock_v_luma_sse2;
- pf->deblock_luma[0] = x264_deblock_h_luma_sse2;
pf->deblock_chroma[1] = x264_deblock_v_chroma_sse2;
- pf->deblock_h_chroma_420 = x264_deblock_h_chroma_sse2;
- pf->deblock_luma_intra[1] = x264_deblock_v_luma_intra_sse2;
- pf->deblock_luma_intra[0] = x264_deblock_h_luma_intra_sse2;
pf->deblock_chroma_intra[1] = x264_deblock_v_chroma_intra_sse2;
pf->deblock_h_chroma_420_intra = x264_deblock_h_chroma_intra_sse2;
+#if HIGH_BIT_DEPTH
+ pf->deblock_chroma_420_intra_mbaff= x264_deblock_h_chroma_intra_mbaff_sse2;
+#endif
}
}
if( cpu&X264_CPU_SSSE3 )
if( cpu&X264_CPU_AVX )
{
pf->deblock_strength = x264_deblock_strength_avx;
+ pf->deblock_h_chroma_420 = x264_deblock_h_chroma_avx;
+ pf->deblock_h_chroma_422 = x264_deblock_h_chroma_422_avx;
+ pf->deblock_h_chroma_422_intra = x264_deblock_h_chroma_422_intra_avx;
+ pf->deblock_luma[1] = x264_deblock_v_luma_avx;
+ pf->deblock_luma[0] = x264_deblock_h_luma_avx;
+ pf->deblock_luma_intra[1] = x264_deblock_v_luma_intra_avx;
+ pf->deblock_luma_intra[0] = x264_deblock_h_luma_intra_avx;
if( !(cpu&X264_CPU_STACK_MOD4) )
{
- pf->deblock_luma[1] = x264_deblock_v_luma_avx;
- pf->deblock_luma[0] = x264_deblock_h_luma_avx;
pf->deblock_chroma[1] = x264_deblock_v_chroma_avx;
- pf->deblock_h_chroma_420 = x264_deblock_h_chroma_avx;
- pf->deblock_luma_intra[1] = x264_deblock_v_luma_intra_avx;
- pf->deblock_luma_intra[0] = x264_deblock_h_luma_intra_avx;
pf->deblock_chroma_intra[1] = x264_deblock_v_chroma_intra_avx;
pf->deblock_h_chroma_420_intra = x264_deblock_h_chroma_intra_avx;
+#if HIGH_BIT_DEPTH
+ pf->deblock_chroma_420_mbaff = x264_deblock_h_chroma_mbaff_avx;
+ pf->deblock_chroma_420_intra_mbaff = x264_deblock_h_chroma_intra_mbaff_avx;
+#endif
}
}
+ if( cpu&X264_CPU_AVX2 )
+ {
+ pf->deblock_strength = x264_deblock_strength_avx2;
+ }
}
#endif
{
pf->deblock_luma[1] = x264_deblock_v_luma_altivec;
pf->deblock_luma[0] = x264_deblock_h_luma_altivec;
- }
+ }
#endif // HAVE_ALTIVEC
-#if HAVE_ARMV6
- if( cpu&X264_CPU_NEON )
- {
+#if HAVE_ARMV6 || ARCH_AARCH64
+ if( cpu&X264_CPU_NEON )
+ {
pf->deblock_luma[1] = x264_deblock_v_luma_neon;
pf->deblock_luma[0] = x264_deblock_h_luma_neon;
-// pf->deblock_chroma[1] = x264_deblock_v_chroma_neon;
-// pf->deblock_h_chroma_420 = x264_deblock_h_chroma_neon;
- }
+ pf->deblock_chroma[1] = x264_deblock_v_chroma_neon;
+ pf->deblock_h_chroma_420 = x264_deblock_h_chroma_neon;
+ pf->deblock_h_chroma_422 = x264_deblock_h_chroma_422_neon;
+ pf->deblock_chroma_420_mbaff = x264_deblock_h_chroma_mbaff_neon;
+ pf->deblock_chroma_420_intra_mbaff = x264_deblock_h_chroma_intra_mbaff_neon;
+ pf->deblock_h_chroma_420_intra = x264_deblock_h_chroma_intra_neon;
+ pf->deblock_h_chroma_422_intra = x264_deblock_h_chroma_422_intra_neon;
+ pf->deblock_chroma_intra[1] = x264_deblock_v_chroma_intra_neon;
+#if ARCH_AARCH64
+ pf->deblock_luma_intra[0] = x264_deblock_h_luma_intra_neon;
+ pf->deblock_luma_intra[1] = x264_deblock_v_luma_intra_neon;
+#endif
+ pf->deblock_strength = x264_deblock_strength_neon;
+ }
+#endif
+
+#if HAVE_MSA
+ if( cpu&X264_CPU_MSA )
+ {
+ pf->deblock_luma[1] = x264_deblock_v_luma_msa;
+ pf->deblock_luma[0] = x264_deblock_h_luma_msa;
+ pf->deblock_chroma[1] = x264_deblock_v_chroma_msa;
+ pf->deblock_h_chroma_420 = x264_deblock_h_chroma_msa;
+ pf->deblock_luma_intra[1] = x264_deblock_v_luma_intra_msa;
+ pf->deblock_luma_intra[0] = x264_deblock_h_luma_intra_msa;
+ pf->deblock_chroma_intra[1] = x264_deblock_v_chroma_intra_msa;
+ pf->deblock_h_chroma_420_intra = x264_deblock_h_chroma_intra_msa;
+ pf->deblock_strength = x264_deblock_strength_msa;
+ }
#endif
#endif // !HIGH_BIT_DEPTH
+
+ /* These functions are equivalent, so don't duplicate them. */
+ pf->deblock_chroma_422_mbaff = pf->deblock_h_chroma_420;
+ pf->deblock_chroma_422_intra_mbaff = pf->deblock_h_chroma_420_intra;
}