predict.c
资源名称:X264CODEC.rar [点击查看]
上传用户:lctgjx
上传日期:2022-06-04
资源大小:8887k
文件大小:7k
源码类别:
流媒体/Mpeg4/MP4
开发平台:
Visual C++
- /*****************************************************************************
- * predict.c: h264 encoder
- *****************************************************************************
- * Copyright (C) 2007-2009 Guillaume Poirier <gpoirier@mplayerhq.hu>
- *
- * 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
- * the Free Software Foundation; either version 2 of the License, or
- * (at your option) any later version.
- *
- * This program 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 General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02111, USA.
- *****************************************************************************/
- #include "common/common.h"
- #include "predict.h"
- #include "pixel.h"
- #include "ppccommon.h"
- static void predict_8x8c_p_altivec( uint8_t *src )
- {
- int i;
- int a, b, c;
- int H = 0;
- int V = 0;
- int i00;
- for( i = 0; i < 4; i++ )
- {
- H += ( i + 1 ) * ( src[4+i - FDEC_STRIDE] - src[2 - i -FDEC_STRIDE] );
- V += ( i + 1 ) * ( src[-1 +(i+4)*FDEC_STRIDE] - src[-1+(2-i)*FDEC_STRIDE] );
- }
- a = 16 * ( src[-1+7*FDEC_STRIDE] + src[7 - FDEC_STRIDE] );
- b = ( 17 * H + 16 ) >> 5;
- c = ( 17 * V + 16 ) >> 5;
- i00 = a -3*b -3*c + 16;
- vec_s16_u i00_u, b_u, c_u;
- i00_u.s[0] = i00;
- b_u.s[0] = b;
- c_u.s[0] = c;
- vec_u16_t val5_v = vec_splat_u16(5);
- vec_s16_t i00_v, b_v, c_v;
- i00_v = vec_splat(i00_u.v, 0);
- b_v = vec_splat(b_u.v, 0);
- c_v = vec_splat(c_u.v, 0);
- vec_s16_t induc_v = (vec_s16_t) CV(0, 1, 2, 3, 4, 5, 6, 7);
- vec_s16_t add_i0_b_0v = vec_mladd(induc_v, b_v, i00_v);
- PREP_STORE8;
- for( i = 0; i < 8; ++i )
- {
- vec_s16_t shift_0_v = vec_sra(add_i0_b_0v, val5_v);
- vec_u8_t com_sat_v = vec_packsu(shift_0_v, shift_0_v);
- VEC_STORE8(com_sat_v, &src[0]);
- src += FDEC_STRIDE;
- add_i0_b_0v = vec_adds(add_i0_b_0v, c_v);
- }
- }
- /****************************************************************************
- * 16x16 prediction for intra luma block
- ****************************************************************************/
- static void predict_16x16_p_altivec( uint8_t *src )
- {
- int16_t a, b, c, i;
- int H = 0;
- int V = 0;
- int16_t i00;
- for( i = 1; i <= 8; i++ )
- {
- H += i * ( src[7+i - FDEC_STRIDE ] - src[7-i - FDEC_STRIDE ] );
- V += i * ( src[(7+i)*FDEC_STRIDE -1] - src[(7-i)*FDEC_STRIDE -1] );
- }
- a = 16 * ( src[15*FDEC_STRIDE -1] + src[15 - FDEC_STRIDE] );
- b = ( 5 * H + 32 ) >> 6;
- c = ( 5 * V + 32 ) >> 6;
- i00 = a - b * 7 - c * 7 + 16;
- vec_s16_u i00_u, b_u, c_u;
- i00_u.s[0] = i00;
- b_u.s[0] = b;
- c_u.s[0] = c;
- vec_u16_t val5_v = vec_splat_u16(5);
- vec_s16_t i00_v, b_v, c_v;
- i00_v = vec_splat(i00_u.v, 0);
- b_v = vec_splat(b_u.v, 0);
- c_v = vec_splat(c_u.v, 0);
- vec_s16_t induc_v = (vec_s16_t) CV(0, 1, 2, 3, 4, 5, 6, 7);
- vec_s16_t b8_v = vec_sl(b_v, vec_splat_u16(3));
- vec_s16_t add_i0_b_0v = vec_mladd(induc_v, b_v, i00_v);
- vec_s16_t add_i0_b_8v = vec_adds(b8_v, add_i0_b_0v);
- int y;
- for( y = 0; y < 16; y++ )
- {
- vec_s16_t shift_0_v = vec_sra(add_i0_b_0v, val5_v);
- vec_s16_t shift_8_v = vec_sra(add_i0_b_8v, val5_v);
- vec_u8_t com_sat_v = vec_packsu(shift_0_v, shift_8_v);
- vec_st( com_sat_v, 0, &src[0]);
- src += FDEC_STRIDE;
- add_i0_b_0v = vec_adds(add_i0_b_0v, c_v);
- add_i0_b_8v = vec_adds(add_i0_b_8v, c_v);
- }
- }
- #define PREDICT_16x16_DC_ALTIVEC(v)
- for (i=0; i<16; i+=2)
- {
- vec_st(v, 0, src);
- vec_st(v, FDEC_STRIDE, src);
- src += FDEC_STRIDE*2;
- }
- static void predict_16x16_dc_altivec( uint8_t *src )
- {
- uint32_t dc = 0;
- int i;
- for( i = 0; i < 16; i++ )
- {
- dc += src[-1 + i * FDEC_STRIDE];
- dc += src[i - FDEC_STRIDE];
- }
- vec_u8_u v ; v.s[0] = (( dc + 16 ) >> 5);
- vec_u8_t bc_v = vec_splat(v.v, 0);
- PREDICT_16x16_DC_ALTIVEC(bc_v);
- }
- static void predict_16x16_dc_left_altivec( uint8_t *src )
- {
- uint32_t dc = 0;
- int i;
- for( i = 0; i < 16; i++ )
- {
- dc += src[-1 + i * FDEC_STRIDE];
- }
- vec_u8_u v ; v.s[0] = (( dc + 8 ) >> 4);
- vec_u8_t bc_v = vec_splat(v.v, 0);
- PREDICT_16x16_DC_ALTIVEC(bc_v);
- }
- static void predict_16x16_dc_top_altivec( uint8_t *src )
- {
- uint32_t dc = 0;
- int i;
- for( i = 0; i < 16; i++ )
- {
- dc += src[i - FDEC_STRIDE];
- }
- vec_u8_u v ; v.s[0] = (( dc + 8 ) >> 4);
- vec_u8_t bc_v = vec_splat(v.v, 0);
- PREDICT_16x16_DC_ALTIVEC(bc_v);
- }
- static void predict_16x16_dc_128_altivec( uint8_t *src )
- {
- int i;
- /* test if generating the constant is faster than loading it.
- vector unsigned int bc_v = (vector unsigned int)CV(0x80808080, 0x80808080, 0x80808080, 0x80808080);
- */
- vec_u8_t bc_v = vec_vslb((vec_u8_t)vec_splat_u8(1),(vec_u8_t)vec_splat_u8(7));
- PREDICT_16x16_DC_ALTIVEC(bc_v);
- }
- static void predict_16x16_h_altivec( uint8_t *src )
- {
- int i;
- for( i = 0; i < 16; i++ )
- {
- vec_u8_t v = vec_ld(-1, src);
- vec_u8_t v_v = vec_splat(v, 15);
- vec_st(v_v, 0, src);
- src += FDEC_STRIDE;
- }
- }
- static void predict_16x16_v_altivec( uint8_t *src )
- {
- vec_u32_u v;
- v.s[0] = *(uint32_t*)&src[ 0-FDEC_STRIDE];
- v.s[1] = *(uint32_t*)&src[ 4-FDEC_STRIDE];
- v.s[2] = *(uint32_t*)&src[ 8-FDEC_STRIDE];
- v.s[3] = *(uint32_t*)&src[12-FDEC_STRIDE];
- int i;
- for( i = 0; i < 16; i++ )
- {
- vec_st(v.v, 0, (uint32_t*)src);
- src += FDEC_STRIDE;
- }
- }
- /****************************************************************************
- * Exported functions:
- ****************************************************************************/
- void x264_predict_16x16_init_altivec( x264_predict_t pf[7] )
- {
- pf[I_PRED_16x16_V ] = predict_16x16_v_altivec;
- pf[I_PRED_16x16_H ] = predict_16x16_h_altivec;
- pf[I_PRED_16x16_DC] = predict_16x16_dc_altivec;
- pf[I_PRED_16x16_P ] = predict_16x16_p_altivec;
- pf[I_PRED_16x16_DC_LEFT] = predict_16x16_dc_left_altivec;
- pf[I_PRED_16x16_DC_TOP ] = predict_16x16_dc_top_altivec;
- pf[I_PRED_16x16_DC_128 ] = predict_16x16_dc_128_altivec;
- }
- void x264_predict_8x8c_init_altivec( x264_predict_t pf[7] )
- {
- pf[I_PRED_CHROMA_P] = predict_8x8c_p_altivec;
- }