--- xvid.c 2002/07/09 01:48:49 1.29 +++ xvid.c 2003/02/09 19:32:52 1.40 @@ -3,41 +3,54 @@ * XVID MPEG-4 VIDEO CODEC * - Native API implementation - * - * This program is an implementation of a part of one or more MPEG-4 - * Video tools as specified in ISO/IEC 14496-2 standard. Those intending - * to use this software module in hardware or software products are - * advised that its use may infringe existing patents or copyrights, and - * any such use would be at such party's own risk. The original - * developer of this software module and his/her company, and subsequent - * editors and their companies, will have no liability for use of this - * software or modifications or derivatives thereof. - * - * 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 + * Copyright(C) 2001-2002 Peter Ross + * + * This file is part of XviD, a free MPEG-4 video encoder/decoder + * + * XviD 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 + * 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 + * along with this program; if not, write to the Free Software * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA * - ****************************************************************************/ - -/***************************************************************************** - * - * History + * Under section 8 of the GNU General Public License, the copyright + * holders of XVID explicitly forbid distribution in the following + * countries: + * + * - Japan + * - United States of America + * + * Linking XviD statically or dynamically with other modules is making a + * combined work based on XviD. Thus, the terms and conditions of the + * GNU General Public License cover the whole combination. + * + * As a special exception, the copyright holders of XviD give you + * permission to link XviD with independent modules that communicate with + * XviD solely through the VFW1.1 and DShow interfaces, regardless of the + * license terms of these independent modules, and to copy and distribute + * the resulting combined work under terms of your choice, provided that + * every copy of the combined work is accompanied by a complete copy of + * the source code of XviD (the version of XviD used to produce the + * combined work), being distributed under the terms of the GNU General + * Public License plus this exception. An independent module is a module + * which is not derived from or based on XviD. + * + * Note that people who make modified versions of XviD are not obligated + * to grant this special exception for their modified versions; it is + * their choice whether to do so. The GNU General Public License gives + * permission to release a modified version without this exception; this + * exception also makes it possible to release a modified version which + * carries forward this exception. * - * - 23.06.2002 added XVID_CPU_CHKONLY - * - 17.03.2002 Added interpolate8x8_halfpel_hv_xmm - * - 22.12.2001 API change: added xvid_init() - Isibaar - * - 16.12.2001 inital version; (c)2001 peter ross - * - * $Id: xvid.c,v 1.29 2002/07/09 01:48:49 chenm001 Exp $ + * $Id: xvid.c,v 1.40 2003/02/09 19:32:52 edgomez Exp $ * ****************************************************************************/ @@ -52,11 +65,79 @@ #include "utils/mem_transfer.h" #include "quant/quant_h263.h" #include "quant/quant_mpeg4.h" +#include "motion/motion.h" #include "motion/sad.h" #include "utils/emms.h" #include "utils/timer.h" #include "bitstream/mbcoding.h" +#if defined(ARCH_IS_IA32) && defined(EXPERIMENTAL_SSE2_CODE) + +#ifdef _MSC_VER +#include +#else +#include +#include +#endif + + +#ifndef _MSC_VER + +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)()) +{ +#ifdef _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 + /***************************************************************************** * XviD Init Entry point * @@ -88,21 +169,33 @@ /* Inform the client the core build - unused because we're still alpha */ init_param->core_build = 1000; - if ((init_param->cpu_flags & XVID_CPU_CHKONLY)) - { - init_param->cpu_flags = check_cpu_features(); - return XVID_ERR_OK; - } - /* Do we have to force CPU features ? */ - if ((init_param->cpu_flags & XVID_CPU_FORCE) > 0) { + if ((init_param->cpu_flags & XVID_CPU_FORCE)) { + cpu_flags = init_param->cpu_flags; + } else { cpu_flags = check_cpu_features(); + +#if defined(ARCH_IS_IA32) && defined(EXPERIMENTAL_SSE2_CODE) + 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 ((init_param->cpu_flags & XVID_CPU_CHKONLY)) + { init_param->cpu_flags = cpu_flags; + return XVID_ERR_OK; } + init_param->cpu_flags = cpu_flags; + + /* Initialize the function pointers */ idct_int32_init(); init_vlc_tables(); @@ -165,11 +258,14 @@ /* Functions used in motion estimation algorithms */ calc_cbp = calc_cbp_c; sad16 = sad16_c; - sad16bi = sad16bi_c; sad8 = sad8_c; + sad16bi = sad16bi_c; + sad8bi = sad8bi_c; dev16 = dev16_c; + + Halfpel8_Refine = Halfpel8_Refine_c; -#ifdef ARCH_X86 +#ifdef ARCH_IS_IA32 if ((cpu_flags & XVID_CPU_MMX) > 0) { /* Forward and Inverse Discrete Cosine Transformation functions */ @@ -221,10 +317,21 @@ calc_cbp = calc_cbp_mmx; sad16 = sad16_mmx; sad8 = sad8_mmx; + sad16bi = sad16bi_mmx; + sad8bi = sad8bi_mmx; dev16 = dev16_mmx; } + /* these 3dnow functions are faster than mmx, but slower than xmm. */ + if ((cpu_flags & XVID_CPU_3DNOW) > 0) { + + /* ME functions */ + sad16bi = sad16bi_3dn; + sad8bi = sad8bi_3dn; + } + + if ((cpu_flags & XVID_CPU_MMXEXT) > 0) { /* Inverse DCT */ @@ -247,8 +354,9 @@ /* ME functions */ sad16 = sad16_xmm; - sad16bi = sad16bi_xmm; sad8 = sad8_xmm; + sad16bi = sad16bi_xmm; + sad8bi = sad8bi_xmm; dev16 = dev16_xmm; } @@ -284,11 +392,11 @@ #endif -#ifdef ARCH_IA64 +#ifdef ARCH_IS_IA64 if ((cpu_flags & XVID_CPU_IA64) > 0) { //use assembler routines? idct_ia64_init(); fdct = fdct_ia64; - idct = idct_ia64; //not yet working, crashes + idct = idct_ia64; interpolate8x8_halfpel_h = interpolate8x8_halfpel_h_ia64; interpolate8x8_halfpel_v = interpolate8x8_halfpel_v_ia64; interpolate8x8_halfpel_hv = interpolate8x8_halfpel_hv_ia64; @@ -296,6 +404,7 @@ sad16bi = sad16bi_ia64; sad8 = sad8_ia64; dev16 = dev16_ia64; + Halfpel8_Refine = Halfpel8_Refine_ia64; quant_intra = quant_intra_ia64; dequant_intra = dequant_intra_ia64; quant_inter = quant_inter_ia64; @@ -306,12 +415,12 @@ transfer_8to16sub2 = transfer_8to16sub2_ia64; transfer_16to8add = transfer_16to8add_ia64; transfer8x8_copy = transfer8x8_copy_ia64; - DEBUG("Using IA-64 assembler routines.\n"); +// DEBUG("Using IA-64 assembler routines.\n"); } #endif -#ifdef ARCH_PPC -#ifdef ARCH_PPC_ALTIVEC +#ifdef ARCH_IS_PPC +#ifdef ARCH_IS_PPC_ALTIVEC calc_cbp = calc_cbp_altivec; fdct = fdct_altivec; idct = idct_altivec; @@ -377,12 +486,6 @@ { switch (opt) { case XVID_ENC_ENCODE: -#ifdef BFRAMES - if (((Encoder *) handle)->mbParam.max_bframes >= 0) - return encoder_encode_bframes((Encoder *) handle, (XVID_ENC_FRAME *) param1, - (XVID_ENC_STATS *) param2); - else -#endif return encoder_encode((Encoder *) handle, (XVID_ENC_FRAME *) param1, (XVID_ENC_STATS *) param2);