3 |
* XVID MPEG-4 VIDEO CODEC |
* XVID MPEG-4 VIDEO CODEC |
4 |
* - Native API implementation - |
* - Native API implementation - |
5 |
* |
* |
6 |
* This program is an implementation of a part of one or more MPEG-4 |
* Copyright(C) 2001-2002 Peter Ross <pross@xvid.org> |
|
* 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. |
|
7 |
* |
* |
8 |
* This program is free software ; you can redistribute it and/or modify |
* This file is part of XviD, a free MPEG-4 video encoder/decoder |
9 |
* it under the terms of the GNU General Public License as published by |
* |
10 |
|
* XviD is free software; you can redistribute it and/or modify it |
11 |
|
* under the terms of the GNU General Public License as published by |
12 |
* the Free Software Foundation ; either version 2 of the License, or |
* the Free Software Foundation ; either version 2 of the License, or |
13 |
* (at your option) any later version. |
* (at your option) any later version. |
14 |
* |
* |
21 |
* along with this program ; if not, write to the Free Software |
* along with this program ; if not, write to the Free Software |
22 |
* Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA |
* Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA |
23 |
* |
* |
24 |
****************************************************************************/ |
* Under section 8 of the GNU General Public License, the copyright |
25 |
|
* holders of XVID explicitly forbid distribution in the following |
26 |
/***************************************************************************** |
* countries: |
27 |
* |
* |
28 |
* History |
* - Japan |
29 |
* |
* - United States of America |
30 |
* - 23.06.2002 added XVID_CPU_CHKONLY |
* |
31 |
* - 17.03.2002 Added interpolate8x8_halfpel_hv_xmm |
* Linking XviD statically or dynamically with other modules is making a |
32 |
* - 22.12.2001 API change: added xvid_init() - Isibaar |
* combined work based on XviD. Thus, the terms and conditions of the |
33 |
* - 16.12.2001 inital version; (c)2001 peter ross <pross@cs.rmit.edu.au> |
* GNU General Public License cover the whole combination. |
34 |
|
* |
35 |
|
* As a special exception, the copyright holders of XviD give you |
36 |
|
* permission to link XviD with independent modules that communicate with |
37 |
|
* XviD solely through the VFW1.1 and DShow interfaces, regardless of the |
38 |
|
* license terms of these independent modules, and to copy and distribute |
39 |
|
* the resulting combined work under terms of your choice, provided that |
40 |
|
* every copy of the combined work is accompanied by a complete copy of |
41 |
|
* the source code of XviD (the version of XviD used to produce the |
42 |
|
* combined work), being distributed under the terms of the GNU General |
43 |
|
* Public License plus this exception. An independent module is a module |
44 |
|
* which is not derived from or based on XviD. |
45 |
|
* |
46 |
|
* Note that people who make modified versions of XviD are not obligated |
47 |
|
* to grant this special exception for their modified versions; it is |
48 |
|
* their choice whether to do so. The GNU General Public License gives |
49 |
|
* permission to release a modified version without this exception; this |
50 |
|
* exception also makes it possible to release a modified version which |
51 |
|
* carries forward this exception. |
52 |
* |
* |
53 |
* $Id: xvid.c,v 1.32 2002-07-21 23:34:08 chl Exp $ |
* $Id: xvid.c,v 1.40 2003-02-09 19:32:52 edgomez Exp $ |
54 |
* |
* |
55 |
****************************************************************************/ |
****************************************************************************/ |
56 |
|
|
71 |
#include "utils/timer.h" |
#include "utils/timer.h" |
72 |
#include "bitstream/mbcoding.h" |
#include "bitstream/mbcoding.h" |
73 |
|
|
74 |
#if defined(ARCH_X86) && defined(EXPERIMENTAL_SSE2_CODE) |
#if defined(ARCH_IS_IA32) && defined(EXPERIMENTAL_SSE2_CODE) |
75 |
|
|
76 |
#ifdef WIN32 |
#ifdef _MSC_VER |
77 |
#include <windows.h> |
#include <windows.h> |
78 |
#else |
#else |
79 |
#include <signal.h> |
#include <signal.h> |
81 |
#endif |
#endif |
82 |
|
|
83 |
|
|
84 |
#ifndef WIN32 |
#ifndef _MSC_VER |
85 |
|
|
86 |
static jmp_buf mark; |
static jmp_buf mark; |
87 |
|
|
94 |
|
|
95 |
|
|
96 |
/* |
/* |
97 |
calls the funcptr, and returns whether SIGILL (illegal instruction) was signalled |
* Calls the funcptr, and returns whether SIGILL (illegal instruction) was signalled |
98 |
return values: |
* Return values: |
99 |
-1 : could not determine |
* -1 : could not determine |
100 |
0 : SIGILL was *not* signalled |
* 0 : SIGILL was *not* signalled |
101 |
1 : SIGILL was signalled |
* 1 : SIGILL was signalled |
102 |
*/ |
*/ |
103 |
|
|
104 |
int |
int |
105 |
sigill_check(void (*func)()) |
sigill_check(void (*func)()) |
106 |
{ |
{ |
107 |
#ifdef WIN32 |
#ifdef _MSC_VER |
108 |
_try { |
_try { |
109 |
func(); |
func(); |
110 |
} |
} |
178 |
|
|
179 |
cpu_flags = check_cpu_features(); |
cpu_flags = check_cpu_features(); |
180 |
|
|
181 |
#if defined(ARCH_X86) && defined(EXPERIMENTAL_SSE2_CODE) |
#if defined(ARCH_IS_IA32) && defined(EXPERIMENTAL_SSE2_CODE) |
182 |
if ((cpu_flags & XVID_CPU_SSE) && sigill_check(sse_os_trigger)) |
if ((cpu_flags & XVID_CPU_SSE) && sigill_check(sse_os_trigger)) |
183 |
cpu_flags &= ~XVID_CPU_SSE; |
cpu_flags &= ~XVID_CPU_SSE; |
184 |
|
|
258 |
/* Functions used in motion estimation algorithms */ |
/* Functions used in motion estimation algorithms */ |
259 |
calc_cbp = calc_cbp_c; |
calc_cbp = calc_cbp_c; |
260 |
sad16 = sad16_c; |
sad16 = sad16_c; |
|
sad16bi = sad16bi_c; |
|
261 |
sad8 = sad8_c; |
sad8 = sad8_c; |
262 |
dev16 = dev16_c; |
sad16bi = sad16bi_c; |
263 |
sad8bi = sad8bi_c; |
sad8bi = sad8bi_c; |
264 |
|
dev16 = dev16_c; |
265 |
|
|
266 |
Halfpel8_Refine = Halfpel8_Refine_c; |
Halfpel8_Refine = Halfpel8_Refine_c; |
267 |
|
|
268 |
#ifdef ARCH_X86 |
#ifdef ARCH_IS_IA32 |
269 |
if ((cpu_flags & XVID_CPU_MMX) > 0) { |
if ((cpu_flags & XVID_CPU_MMX) > 0) { |
270 |
|
|
271 |
/* Forward and Inverse Discrete Cosine Transformation functions */ |
/* Forward and Inverse Discrete Cosine Transformation functions */ |
317 |
calc_cbp = calc_cbp_mmx; |
calc_cbp = calc_cbp_mmx; |
318 |
sad16 = sad16_mmx; |
sad16 = sad16_mmx; |
319 |
sad8 = sad8_mmx; |
sad8 = sad8_mmx; |
320 |
|
sad16bi = sad16bi_mmx; |
321 |
|
sad8bi = sad8bi_mmx; |
322 |
dev16 = dev16_mmx; |
dev16 = dev16_mmx; |
323 |
|
|
324 |
} |
} |
325 |
|
|
326 |
|
/* these 3dnow functions are faster than mmx, but slower than xmm. */ |
327 |
|
if ((cpu_flags & XVID_CPU_3DNOW) > 0) { |
328 |
|
|
329 |
|
/* ME functions */ |
330 |
|
sad16bi = sad16bi_3dn; |
331 |
|
sad8bi = sad8bi_3dn; |
332 |
|
} |
333 |
|
|
334 |
|
|
335 |
if ((cpu_flags & XVID_CPU_MMXEXT) > 0) { |
if ((cpu_flags & XVID_CPU_MMXEXT) > 0) { |
336 |
|
|
337 |
/* Inverse DCT */ |
/* Inverse DCT */ |
354 |
|
|
355 |
/* ME functions */ |
/* ME functions */ |
356 |
sad16 = sad16_xmm; |
sad16 = sad16_xmm; |
|
sad16bi = sad16bi_xmm; |
|
357 |
sad8 = sad8_xmm; |
sad8 = sad8_xmm; |
358 |
|
sad16bi = sad16bi_xmm; |
359 |
|
sad8bi = sad8bi_xmm; |
360 |
dev16 = dev16_xmm; |
dev16 = dev16_xmm; |
361 |
|
|
362 |
} |
} |
392 |
|
|
393 |
#endif |
#endif |
394 |
|
|
395 |
#ifdef ARCH_IA64 |
#ifdef ARCH_IS_IA64 |
396 |
if ((cpu_flags & XVID_CPU_IA64) > 0) { //use assembler routines? |
if ((cpu_flags & XVID_CPU_IA64) > 0) { //use assembler routines? |
397 |
idct_ia64_init(); |
idct_ia64_init(); |
398 |
fdct = fdct_ia64; |
fdct = fdct_ia64; |
399 |
idct = idct_ia64; //not yet working, crashes |
idct = idct_ia64; |
400 |
interpolate8x8_halfpel_h = interpolate8x8_halfpel_h_ia64; |
interpolate8x8_halfpel_h = interpolate8x8_halfpel_h_ia64; |
401 |
interpolate8x8_halfpel_v = interpolate8x8_halfpel_v_ia64; |
interpolate8x8_halfpel_v = interpolate8x8_halfpel_v_ia64; |
402 |
interpolate8x8_halfpel_hv = interpolate8x8_halfpel_hv_ia64; |
interpolate8x8_halfpel_hv = interpolate8x8_halfpel_hv_ia64; |
415 |
transfer_8to16sub2 = transfer_8to16sub2_ia64; |
transfer_8to16sub2 = transfer_8to16sub2_ia64; |
416 |
transfer_16to8add = transfer_16to8add_ia64; |
transfer_16to8add = transfer_16to8add_ia64; |
417 |
transfer8x8_copy = transfer8x8_copy_ia64; |
transfer8x8_copy = transfer8x8_copy_ia64; |
418 |
DEBUG("Using IA-64 assembler routines.\n"); |
// DEBUG("Using IA-64 assembler routines.\n"); |
419 |
} |
} |
420 |
#endif |
#endif |
421 |
|
|
422 |
#ifdef ARCH_PPC |
#ifdef ARCH_IS_PPC |
423 |
#ifdef ARCH_PPC_ALTIVEC |
#ifdef ARCH_IS_PPC_ALTIVEC |
424 |
calc_cbp = calc_cbp_altivec; |
calc_cbp = calc_cbp_altivec; |
425 |
fdct = fdct_altivec; |
fdct = fdct_altivec; |
426 |
idct = idct_altivec; |
idct = idct_altivec; |
486 |
{ |
{ |
487 |
switch (opt) { |
switch (opt) { |
488 |
case XVID_ENC_ENCODE: |
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 |
|
489 |
return encoder_encode((Encoder *) handle, (XVID_ENC_FRAME *) param1, |
return encoder_encode((Encoder *) handle, (XVID_ENC_FRAME *) param1, |
490 |
(XVID_ENC_STATS *) param2); |
(XVID_ENC_STATS *) param2); |
491 |
|
|