/***************************************************************************** * * XVID MPEG-4 VIDEO CODEC * - Native API implementation - * * Copyright(C) 2001-2003 Peter Ross * * 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., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA * * $Id: xvid.c,v 1.48.2.5 2004/08/29 10:56:57 edgomez Exp $ * ****************************************************************************/ #include #include #include #include #include "xvid.h" #include "decoder.h" #include "encoder.h" #include "bitstream/cbp.h" #include "dct/idct.h" #include "dct/fdct.h" #include "image/colorspace.h" #include "image/interpolate8x8.h" #include "image/reduced.h" #include "utils/mem_transfer.h" #include "utils/mbfunctions.h" #include "quant/quant.h" #include "motion/motion.h" #include "motion/sad.h" #include "utils/emms.h" #include "utils/timer.h" #include "bitstream/mbcoding.h" #include "image/qpel.h" #include "image/postprocessing.h" #if defined(_DEBUG) unsigned int xvid_debug = 0; /* xvid debug mask */ #endif #if defined(ARCH_IS_IA32) #if defined(_MSC_VER) # include #else # include # include static jmp_buf mark; static void sigill_handler(int signal) { longjmp(mark, 1); } #endif /* * Calls the funcptr, and returns whether SIGILL (illegal instruction) was * signalled * * Return values: * -1 : could not determine * 0 : SIGILL was *not* signalled * 1 : SIGILL was signalled */ int sigill_check(void (*func)()) { #if defined(_MSC_VER) _try { func(); } _except(EXCEPTION_EXECUTE_HANDLER) { if (_exception_code() == STATUS_ILLEGAL_INSTRUCTION) return 1; } return 0; #else void * old_handler; int jmpret; old_handler = signal(SIGILL, sigill_handler); if (old_handler == SIG_ERR) { return -1; } jmpret = setjmp(mark); if (jmpret == 0) { func(); } signal(SIGILL, old_handler); return jmpret; #endif } #endif /* detect cpu flags */ static unsigned int detect_cpu_flags() { /* enable native assembly optimizations by default */ unsigned int cpu_flags = XVID_CPU_ASM; #if defined(ARCH_IS_IA32) cpu_flags |= check_cpu_features(); if ((cpu_flags & XVID_CPU_SSE) && sigill_check(sse_os_trigger)) cpu_flags &= ~XVID_CPU_SSE; if ((cpu_flags & XVID_CPU_SSE2) && sigill_check(sse2_os_trigger)) cpu_flags &= ~XVID_CPU_SSE2; #endif #if defined(ARCH_IS_PPC) #if defined(ARCH_IS_PPC_ALTIVEC) cpu_flags |= XVID_CPU_ALTIVEC; #endif #endif return cpu_flags; } /***************************************************************************** * XviD Init Entry point * * Well this function initialize all internal function pointers according * to the CPU features forced by the library client or autodetected (depending * on the XVID_CPU_FORCE flag). It also initializes vlc coding tables and all * image colorspace transformation tables. * * Returned value : XVID_ERR_OK * + API_VERSION in the input XVID_INIT_PARAM structure * + core build " " " " " * ****************************************************************************/ static int xvid_gbl_init(xvid_gbl_init_t * init) { unsigned int cpu_flags; if (XVID_VERSION_MAJOR(init->version) != 1) /* v1.x.x */ return XVID_ERR_VERSION; cpu_flags = (init->cpu_flags & XVID_CPU_FORCE) ? init->cpu_flags : detect_cpu_flags(); /* Initialize the function pointers */ idct_int32_init(); init_vlc_tables(); /* Fixed Point Forward/Inverse DCT transformations */ fdct = fdct_int32; idct = idct_int32; /* Only needed on PPC Altivec archs */ sadInit = 0; /* Restore FPU context : emms_c is a nop functions */ emms = emms_c; /* Qpel stuff */ xvid_QP_Funcs = &xvid_QP_Funcs_C; xvid_QP_Add_Funcs = &xvid_QP_Add_Funcs_C; xvid_Init_QP(); /* Quantization functions */ quant_h263_intra = quant_h263_intra_c; quant_h263_inter = quant_h263_inter_c; dequant_h263_intra = dequant_h263_intra_c; dequant_h263_inter = dequant_h263_inter_c; quant_mpeg_intra = quant_mpeg_intra_c; quant_mpeg_inter = quant_mpeg_inter_c; dequant_mpeg_intra = dequant_mpeg_intra_c; dequant_mpeg_inter = dequant_mpeg_inter_c; /* Block transfer related functions */ transfer_8to16copy = transfer_8to16copy_c; transfer_16to8copy = transfer_16to8copy_c; transfer_8to16sub = transfer_8to16sub_c; transfer_8to16subro = transfer_8to16subro_c; transfer_8to16sub2 = transfer_8to16sub2_c; transfer_16to8add = transfer_16to8add_c; transfer8x8_copy = transfer8x8_copy_c; /* Interlacing functions */ MBFieldTest = MBFieldTest_c; /* Image interpolation related functions */ interpolate8x8_halfpel_h = interpolate8x8_halfpel_h_c; interpolate8x8_halfpel_v = interpolate8x8_halfpel_v_c; interpolate8x8_halfpel_hv = interpolate8x8_halfpel_hv_c; interpolate16x16_lowpass_h = interpolate16x16_lowpass_h_c; interpolate16x16_lowpass_v = interpolate16x16_lowpass_v_c; interpolate16x16_lowpass_hv = interpolate16x16_lowpass_hv_c; interpolate8x8_lowpass_h = interpolate8x8_lowpass_h_c; interpolate8x8_lowpass_v = interpolate8x8_lowpass_v_c; interpolate8x8_lowpass_hv = interpolate8x8_lowpass_hv_c; interpolate8x8_6tap_lowpass_h = interpolate8x8_6tap_lowpass_h_c; interpolate8x8_6tap_lowpass_v = interpolate8x8_6tap_lowpass_v_c; interpolate8x8_avg2 = interpolate8x8_avg2_c; interpolate8x8_avg4 = interpolate8x8_avg4_c; /* reduced resolution */ copy_upsampled_8x8_16to8 = xvid_Copy_Upsampled_8x8_16To8_C; add_upsampled_8x8_16to8 = xvid_Add_Upsampled_8x8_16To8_C; vfilter_31 = xvid_VFilter_31_C; hfilter_31 = xvid_HFilter_31_C; filter_18x18_to_8x8 = xvid_Filter_18x18_To_8x8_C; filter_diff_18x18_to_8x8 = xvid_Filter_Diff_18x18_To_8x8_C; /* Initialize internal colorspace transformation tables */ colorspace_init(); /* All colorspace transformation functions User Format->YV12 */ yv12_to_yv12 = yv12_to_yv12_c; rgb555_to_yv12 = rgb555_to_yv12_c; rgb565_to_yv12 = rgb565_to_yv12_c; bgr_to_yv12 = bgr_to_yv12_c; bgra_to_yv12 = bgra_to_yv12_c; abgr_to_yv12 = abgr_to_yv12_c; rgba_to_yv12 = rgba_to_yv12_c; argb_to_yv12 = argb_to_yv12_c; yuyv_to_yv12 = yuyv_to_yv12_c; uyvy_to_yv12 = uyvy_to_yv12_c; rgb555i_to_yv12 = rgb555i_to_yv12_c; rgb565i_to_yv12 = rgb565i_to_yv12_c; bgri_to_yv12 = bgri_to_yv12_c; bgrai_to_yv12 = bgrai_to_yv12_c; abgri_to_yv12 = abgri_to_yv12_c; rgbai_to_yv12 = rgbai_to_yv12_c; argbi_to_yv12 = argbi_to_yv12_c; yuyvi_to_yv12 = yuyvi_to_yv12_c; uyvyi_to_yv12 = uyvyi_to_yv12_c; /* All colorspace transformation functions YV12->User format */ yv12_to_rgb555 = yv12_to_rgb555_c; yv12_to_rgb565 = yv12_to_rgb565_c; yv12_to_bgr = yv12_to_bgr_c; yv12_to_bgra = yv12_to_bgra_c; yv12_to_abgr = yv12_to_abgr_c; yv12_to_rgba = yv12_to_rgba_c; yv12_to_argb = yv12_to_argb_c; yv12_to_yuyv = yv12_to_yuyv_c; yv12_to_uyvy = yv12_to_uyvy_c; yv12_to_rgb555i = yv12_to_rgb555i_c; yv12_to_rgb565i = yv12_to_rgb565i_c; yv12_to_bgri = yv12_to_bgri_c; yv12_to_bgrai = yv12_to_bgrai_c; yv12_to_abgri = yv12_to_abgri_c; yv12_to_rgbai = yv12_to_rgbai_c; yv12_to_argbi = yv12_to_argbi_c; yv12_to_yuyvi = yv12_to_yuyvi_c; yv12_to_uyvyi = yv12_to_uyvyi_c; /* Functions used in motion estimation algorithms */ calc_cbp = calc_cbp_c; sad16 = sad16_c; sad8 = sad8_c; sad16bi = sad16bi_c; sad8bi = sad8bi_c; dev16 = dev16_c; sad16v = sad16v_c; sse8_16bit = sse8_16bit_c; #if defined(ARCH_IS_IA32) if ((cpu_flags & XVID_CPU_ASM)) { vfilter_31 = xvid_VFilter_31_x86; hfilter_31 = xvid_HFilter_31_x86; } if ((cpu_flags & XVID_CPU_MMX) || (cpu_flags & XVID_CPU_MMXEXT) || (cpu_flags & XVID_CPU_3DNOW) || (cpu_flags & XVID_CPU_3DNOWEXT) || (cpu_flags & XVID_CPU_SSE) || (cpu_flags & XVID_CPU_SSE2)) { /* Restore FPU context : emms_c is a nop functions */ emms = emms_mmx; } if ((cpu_flags & XVID_CPU_MMX)) { /* Forward and Inverse Discrete Cosine Transformation functions */ fdct = fdct_mmx_skal; idct = idct_mmx; /* Qpel stuff */ xvid_QP_Funcs = &xvid_QP_Funcs_mmx; xvid_QP_Add_Funcs = &xvid_QP_Add_Funcs_mmx; /* Quantization related functions */ quant_h263_intra = quant_h263_intra_mmx; quant_h263_inter = quant_h263_inter_mmx; dequant_h263_intra = dequant_h263_intra_mmx; dequant_h263_inter = dequant_h263_inter_mmx; quant_mpeg_intra = quant_mpeg_intra_mmx; quant_mpeg_inter = quant_mpeg_inter_mmx; dequant_mpeg_intra = dequant_mpeg_intra_mmx; dequant_mpeg_inter = dequant_mpeg_inter_mmx; /* Block related functions */ transfer_8to16copy = transfer_8to16copy_mmx; transfer_16to8copy = transfer_16to8copy_mmx; transfer_8to16sub = transfer_8to16sub_mmx; transfer_8to16subro = transfer_8to16subro_mmx; transfer_8to16sub2 = transfer_8to16sub2_mmx; transfer_16to8add = transfer_16to8add_mmx; transfer8x8_copy = transfer8x8_copy_mmx; /* Interlacing Functions */ MBFieldTest = MBFieldTest_mmx; /* Image Interpolation related functions */ interpolate8x8_halfpel_h = interpolate8x8_halfpel_h_mmx; interpolate8x8_halfpel_v = interpolate8x8_halfpel_v_mmx; interpolate8x8_halfpel_hv = interpolate8x8_halfpel_hv_mmx; interpolate8x8_6tap_lowpass_h = interpolate8x8_6tap_lowpass_h_mmx; interpolate8x8_6tap_lowpass_v = interpolate8x8_6tap_lowpass_v_mmx; interpolate8x8_avg2 = interpolate8x8_avg2_mmx; interpolate8x8_avg4 = interpolate8x8_avg4_mmx; /* reduced resolution */ copy_upsampled_8x8_16to8 = xvid_Copy_Upsampled_8x8_16To8_mmx; add_upsampled_8x8_16to8 = xvid_Add_Upsampled_8x8_16To8_mmx; hfilter_31 = xvid_HFilter_31_mmx; filter_18x18_to_8x8 = xvid_Filter_18x18_To_8x8_mmx; filter_diff_18x18_to_8x8 = xvid_Filter_Diff_18x18_To_8x8_mmx; /* image input xxx_to_yv12 related functions */ yv12_to_yv12 = yv12_to_yv12_mmx; bgr_to_yv12 = bgr_to_yv12_mmx; bgra_to_yv12 = bgra_to_yv12_mmx; yuyv_to_yv12 = yuyv_to_yv12_mmx; uyvy_to_yv12 = uyvy_to_yv12_mmx; /* image output yv12_to_xxx related functions */ yv12_to_bgr = yv12_to_bgr_mmx; yv12_to_bgra = yv12_to_bgra_mmx; yv12_to_yuyv = yv12_to_yuyv_mmx; yv12_to_uyvy = yv12_to_uyvy_mmx; yv12_to_yuyvi = yv12_to_yuyvi_mmx; yv12_to_uyvyi = yv12_to_uyvyi_mmx; /* Motion estimation related functions */ calc_cbp = calc_cbp_mmx; sad16 = sad16_mmx; sad8 = sad8_mmx; sad16bi = sad16bi_mmx; sad8bi = sad8bi_mmx; dev16 = dev16_mmx; sad16v = sad16v_mmx; sse8_16bit = sse8_16bit_mmx; } /* these 3dnow functions are faster than mmx, but slower than xmm. */ if ((cpu_flags & XVID_CPU_3DNOW)) { emms = emms_3dn; /* ME functions */ sad16bi = sad16bi_3dn; sad8bi = sad8bi_3dn; yuyv_to_yv12 = yuyv_to_yv12_3dn; uyvy_to_yv12 = uyvy_to_yv12_3dn; } if ((cpu_flags & XVID_CPU_MMXEXT)) { /* DCT */ fdct = fdct_xmm_skal; idct = idct_xmm; /* Interpolation */ interpolate8x8_halfpel_h = interpolate8x8_halfpel_h_xmm; interpolate8x8_halfpel_v = interpolate8x8_halfpel_v_xmm; interpolate8x8_halfpel_hv = interpolate8x8_halfpel_hv_xmm; /* reduced resolution */ copy_upsampled_8x8_16to8 = xvid_Copy_Upsampled_8x8_16To8_xmm; add_upsampled_8x8_16to8 = xvid_Add_Upsampled_8x8_16To8_xmm; /* Quantization */ quant_mpeg_intra = quant_mpeg_intra_xmm; quant_mpeg_inter = quant_mpeg_inter_xmm; dequant_h263_intra = dequant_h263_intra_xmm; dequant_h263_inter = dequant_h263_inter_xmm; /* Buffer transfer */ transfer_8to16sub2 = transfer_8to16sub2_xmm; /* Colorspace transformation */ yv12_to_yv12 = yv12_to_yv12_xmm; yuyv_to_yv12 = yuyv_to_yv12_xmm; uyvy_to_yv12 = uyvy_to_yv12_xmm; /* ME functions */ sad16 = sad16_xmm; sad8 = sad8_xmm; sad16bi = sad16bi_xmm; sad8bi = sad8bi_xmm; dev16 = dev16_xmm; sad16v = sad16v_xmm; } if ((cpu_flags & XVID_CPU_3DNOW)) { /* Interpolation */ interpolate8x8_halfpel_h = interpolate8x8_halfpel_h_3dn; interpolate8x8_halfpel_v = interpolate8x8_halfpel_v_3dn; interpolate8x8_halfpel_hv = interpolate8x8_halfpel_hv_3dn; } if ((cpu_flags & XVID_CPU_3DNOWEXT)) { /* Buffer transfer */ transfer_8to16copy = transfer_8to16copy_3dne; transfer_16to8copy = transfer_16to8copy_3dne; transfer_8to16sub = transfer_8to16sub_3dne; transfer_8to16subro = transfer_8to16subro_3dne; transfer_16to8add = transfer_16to8add_3dne; transfer8x8_copy = transfer8x8_copy_3dne; if ((cpu_flags & XVID_CPU_MMXEXT)) { /* Inverse DCT */ idct = idct_3dne; /* Buffer transfer */ transfer_8to16sub2 = transfer_8to16sub2_3dne; /* Interpolation */ interpolate8x8_halfpel_h = interpolate8x8_halfpel_h_3dne; interpolate8x8_halfpel_v = interpolate8x8_halfpel_v_3dne; interpolate8x8_halfpel_hv = interpolate8x8_halfpel_hv_3dne; /* Quantization */ quant_h263_intra = quant_h263_intra_3dne; /* cmov only */ quant_h263_inter = quant_h263_inter_3dne; dequant_mpeg_intra = dequant_mpeg_intra_3dne; /* cmov only */ dequant_mpeg_inter = dequant_mpeg_inter_3dne; dequant_h263_intra = dequant_h263_intra_3dne; dequant_h263_inter = dequant_h263_inter_3dne; /* ME functions */ calc_cbp = calc_cbp_3dne; sad16 = sad16_3dne; sad8 = sad8_3dne; sad16bi = sad16bi_3dne; sad8bi = sad8bi_3dne; dev16 = dev16_3dne; } } if ((cpu_flags & XVID_CPU_SSE2)) { calc_cbp = calc_cbp_sse2; /* Quantization */ quant_h263_intra = quant_h263_intra_sse2; quant_h263_inter = quant_h263_inter_sse2; dequant_h263_intra = dequant_h263_intra_sse2; dequant_h263_inter = dequant_h263_inter_sse2; /* SAD operators */ sad16 = sad16_sse2; dev16 = dev16_sse2; /* DCT operators * no iDCT because it's not "Walken matching" */ fdct = fdct_sse2_skal; } #endif /* ARCH_IS_IA32 */ #if defined(ARCH_IS_IA64) if ((cpu_flags & XVID_CPU_ASM)) { /* use assembler routines? */ idct_ia64_init(); fdct = fdct_ia64; idct = idct_ia64; /*not yet working, crashes */ interpolate8x8_halfpel_h = interpolate8x8_halfpel_h_ia64; interpolate8x8_halfpel_v = interpolate8x8_halfpel_v_ia64; interpolate8x8_halfpel_hv = interpolate8x8_halfpel_hv_ia64; sad16 = sad16_ia64; sad16bi = sad16bi_ia64; sad8 = sad8_ia64; dev16 = dev16_ia64; /* Halfpel8_Refine = Halfpel8_Refine_ia64; */ quant_h263_intra = quant_h263_intra_ia64; quant_h263_inter = quant_h263_inter_ia64; dequant_h263_intra = dequant_h263_intra_ia64; dequant_h263_inter = dequant_h263_inter_ia64; transfer_8to16copy = transfer_8to16copy_ia64; transfer_16to8copy = transfer_16to8copy_ia64; transfer_8to16sub = transfer_8to16sub_ia64; transfer_8to16sub2 = transfer_8to16sub2_ia64; transfer_16to8add = transfer_16to8add_ia64; transfer8x8_copy = transfer8x8_copy_ia64; } #endif #if defined(ARCH_IS_PPC) if ((cpu_flags & XVID_CPU_ASM)) { calc_cbp = calc_cbp_ppc; } if ((cpu_flags & XVID_CPU_ALTIVEC)) { calc_cbp = calc_cbp_altivec; fdct = fdct_altivec; idct = idct_altivec; sadInit = sadInit_altivec; sad16 = sad16_altivec; sad8 = sad8_altivec; dev16 = dev16_altivec; } #endif #if defined(_DEBUG) xvid_debug = init->debug; #endif return 0; } static int xvid_gbl_info(xvid_gbl_info_t * info) { if (XVID_VERSION_MAJOR(info->version) != 1) /* v1.x.x */ return XVID_ERR_VERSION; info->actual_version = XVID_VERSION; info->build = "xvid-1.0.2"; info->cpu_flags = detect_cpu_flags(); #if defined(_SMP) && defined(WIN32) info->num_threads = pthread_num_processors_np();; #else info->num_threads = 0; #endif return 0; } static int xvid_gbl_convert(xvid_gbl_convert_t* convert) { int width; int height; int width2; int height2; IMAGE img; if (XVID_VERSION_MAJOR(convert->version) != 1) /* v1.x.x */ return XVID_ERR_VERSION; #if 0 const int flip1 = (convert->input.colorspace & XVID_CSP_VFLIP) ^ (convert->output.colorspace & XVID_CSP_VFLIP); #endif width = convert->width; height = convert->height; width2 = convert->width/2; height2 = convert->height/2; switch (convert->input.csp & ~XVID_CSP_VFLIP) { case XVID_CSP_YV12 : img.y = convert->input.plane[0]; img.v = (uint8_t*)convert->input.plane[0] + convert->input.stride[0]*height; img.u = (uint8_t*)convert->input.plane[0] + convert->input.stride[0]*height + (convert->input.stride[0]/2)*height2; image_output(&img, width, height, width, (uint8_t**)convert->output.plane, convert->output.stride, convert->output.csp, convert->interlacing); break; default : return XVID_ERR_FORMAT; } emms(); return 0; } /***************************************************************************** * XviD Global Entry point * * Well this function initialize all internal function pointers according * to the CPU features forced by the library client or autodetected (depending * on the XVID_CPU_FORCE flag). It also initializes vlc coding tables and all * image colorspace transformation tables. * ****************************************************************************/ int xvid_global(void *handle, int opt, void *param1, void *param2) { switch(opt) { case XVID_GBL_INIT : return xvid_gbl_init((xvid_gbl_init_t*)param1); case XVID_GBL_INFO : return xvid_gbl_info((xvid_gbl_info_t*)param1); case XVID_GBL_CONVERT : return xvid_gbl_convert((xvid_gbl_convert_t*)param1); default : return XVID_ERR_FAIL; } } /***************************************************************************** * XviD Native decoder entry point * * This function is just a wrapper to all the option cases. * * Returned values : XVID_ERR_FAIL when opt is invalid * else returns the wrapped function result * ****************************************************************************/ int xvid_decore(void *handle, int opt, void *param1, void *param2) { switch (opt) { case XVID_DEC_CREATE: return decoder_create((xvid_dec_create_t *) param1); case XVID_DEC_DESTROY: return decoder_destroy((DECODER *) handle); case XVID_DEC_DECODE: return decoder_decode((DECODER *) handle, (xvid_dec_frame_t *) param1, (xvid_dec_stats_t*) param2); default: return XVID_ERR_FAIL; } } /***************************************************************************** * XviD Native encoder entry point * * This function is just a wrapper to all the option cases. * * Returned values : XVID_ERR_FAIL when opt is invalid * else returns the wrapped function result * ****************************************************************************/ int xvid_encore(void *handle, int opt, void *param1, void *param2) { switch (opt) { case XVID_ENC_ENCODE: return enc_encode((Encoder *) handle, (xvid_enc_frame_t *) param1, (xvid_enc_stats_t *) param2); case XVID_ENC_CREATE: return enc_create((xvid_enc_create_t *) param1); case XVID_ENC_DESTROY: return enc_destroy((Encoder *) handle); default: return XVID_ERR_FAIL; } }