1 /*****************************************************************************
2 * i422_yuy2.c : YUV to YUV conversion module for vlc
3 *****************************************************************************
4 * Copyright (C) 2000, 2001 the VideoLAN team
7 * Authors: Samuel Hocevar <sam@zoy.org>
8 * Damien Fouilleul <damienf@videolan.org>
10 * This program is free software; you can redistribute it and/or modify
11 * it under the terms of the GNU General Public License as published by
12 * the Free Software Foundation; either version 2 of the License, or
13 * (at your option) any later version.
15 * This program is distributed in the hope that it will be useful,
16 * but WITHOUT ANY WARRANTY; without even the implied warranty of
17 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
18 * GNU General Public License for more details.
20 * You should have received a copy of the GNU General Public License
21 * along with this program; if not, write to the Free Software
22 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston MA 02110-1301, USA.
23 *****************************************************************************/
25 /*****************************************************************************
27 *****************************************************************************/
28 #include <string.h> /* strerror() */
29 #include <stdlib.h> /* malloc(), free() */
34 #include "i422_yuy2.h"
36 #define SRC_FOURCC "I422"
37 #if defined (MODULE_NAME_IS_i422_yuy2)
38 # define DEST_FOURCC "YUY2,YUNV,YVYU,UYVY,UYNV,Y422,IUYV,cyuv,Y211"
40 # define DEST_FOURCC "YUY2,YUNV,YVYU,UYVY,UYNV,Y422,IUYV,cyuv"
43 /*****************************************************************************
44 * Local and extern prototypes.
45 *****************************************************************************/
46 static int Activate ( vlc_object_t * );
48 static void I422_YUY2 ( vout_thread_t *, picture_t *, picture_t * );
49 static void I422_YVYU ( vout_thread_t *, picture_t *, picture_t * );
50 static void I422_UYVY ( vout_thread_t *, picture_t *, picture_t * );
51 static void I422_IUYV ( vout_thread_t *, picture_t *, picture_t * );
52 static void I422_cyuv ( vout_thread_t *, picture_t *, picture_t * );
53 #if defined (MODULE_NAME_IS_i422_yuy2)
54 static void I422_Y211 ( vout_thread_t *, picture_t *, picture_t * );
55 static void I422_Y211 ( vout_thread_t *, picture_t *, picture_t * );
56 static void I422_YV12 ( vout_thread_t *, picture_t *, picture_t * );
59 /*****************************************************************************
61 *****************************************************************************/
63 #if defined (MODULE_NAME_IS_i422_yuy2)
64 set_description( _("Conversions from " SRC_FOURCC " to " DEST_FOURCC) );
65 set_capability( "chroma", 80 );
66 #elif defined (MODULE_NAME_IS_i422_yuy2_mmx)
67 set_description( _("MMX conversions from " SRC_FOURCC " to " DEST_FOURCC) );
68 set_capability( "chroma", 100 );
69 add_requirement( MMX );
70 #elif defined (MODULE_NAME_IS_i422_yuy2_sse2)
71 set_description( _("SSE2 conversions from " SRC_FOURCC " to " DEST_FOURCC) );
72 set_capability( "chroma", 120 );
73 add_requirement( MMX );
75 set_callbacks( Activate, NULL );
78 /*****************************************************************************
79 * Activate: allocate a chroma function
80 *****************************************************************************
81 * This function allocates and initializes a chroma function
82 *****************************************************************************/
83 static int Activate( vlc_object_t *p_this )
85 vout_thread_t *p_vout = (vout_thread_t *)p_this;
87 if( p_vout->render.i_width & 1 || p_vout->render.i_height & 1 )
92 switch( p_vout->render.i_chroma )
94 case VLC_FOURCC('I','4','2','2'):
95 switch( p_vout->output.i_chroma )
97 case VLC_FOURCC('Y','U','Y','2'):
98 case VLC_FOURCC('Y','U','N','V'):
99 p_vout->chroma.pf_convert = I422_YUY2;
102 case VLC_FOURCC('Y','V','Y','U'):
103 p_vout->chroma.pf_convert = I422_YVYU;
106 case VLC_FOURCC('U','Y','V','Y'):
107 case VLC_FOURCC('U','Y','N','V'):
108 case VLC_FOURCC('Y','4','2','2'):
109 p_vout->chroma.pf_convert = I422_UYVY;
112 case VLC_FOURCC('I','U','Y','V'):
113 p_vout->chroma.pf_convert = I422_IUYV;
116 case VLC_FOURCC('c','y','u','v'):
117 p_vout->chroma.pf_convert = I422_cyuv;
120 #if defined (MODULE_NAME_IS_i422_yuy2)
121 case VLC_FOURCC('Y','2','1','1'):
122 p_vout->chroma.pf_convert = I422_Y211;
125 case VLC_FOURCC('Y','V','1','2'):
126 p_vout->chroma.pf_convert = I422_YV12;
142 /* Following functions are local */
144 /*****************************************************************************
145 * I422_YUY2: planar YUV 4:2:2 to packed YUY2 4:2:2
146 *****************************************************************************/
147 static void I422_YUY2( vout_thread_t *p_vout, picture_t *p_source,
150 uint8_t *p_line = p_dest->p->p_pixels;
151 uint8_t *p_y = p_source->Y_PIXELS;
152 uint8_t *p_u = p_source->U_PIXELS;
153 uint8_t *p_v = p_source->V_PIXELS;
157 const int i_source_margin = p_source->p[0].i_pitch
158 - p_source->p[0].i_visible_pitch;
159 const int i_source_margin_c = p_source->p[1].i_pitch
160 - p_source->p[1].i_visible_pitch;
161 const int i_dest_margin = p_dest->p->i_pitch
162 - p_dest->p->i_visible_pitch;
164 #if defined (MODULE_NAME_IS_i422_yuy2_sse2)
166 if( 0 == (15 & (p_source->p[Y_PLANE].i_pitch|p_dest->p->i_pitch|
167 ((int)p_line|(int)p_y))) )
169 /* use faster SSE2 aligned fetch and store */
170 for( i_y = p_vout->render.i_height ; i_y-- ; )
172 for( i_x = p_vout->render.i_width / 16 ; i_x-- ; )
174 SSE2_CALL( SSE2_YUV422_YUYV_ALIGNED );
176 for( i_x = ( p_vout->render.i_width % 16 ) / 2; i_x-- ; )
178 C_YUV422_YUYV( p_line, p_y, p_u, p_v );
180 p_y += i_source_margin;
181 p_u += i_source_margin_c;
182 p_v += i_source_margin_c;
183 p_line += i_dest_margin;
187 /* use slower SSE2 unaligned fetch and store */
188 for( i_y = p_vout->render.i_height ; i_y-- ; )
190 for( i_x = p_vout->render.i_width / 16 ; i_x-- ; )
192 SSE2_CALL( SSE2_YUV422_YUYV_UNALIGNED );
194 for( i_x = ( p_vout->render.i_width % 16 ) / 2; i_x-- ; )
196 C_YUV422_YUYV( p_line, p_y, p_u, p_v );
198 p_y += i_source_margin;
199 p_u += i_source_margin_c;
200 p_v += i_source_margin_c;
201 p_line += i_dest_margin;
208 for( i_y = p_vout->render.i_height ; i_y-- ; )
210 for( i_x = p_vout->render.i_width / 8 ; i_x-- ; )
212 #if defined (MODULE_NAME_IS_i422_yuy2)
213 C_YUV422_YUYV( p_line, p_y, p_u, p_v );
214 C_YUV422_YUYV( p_line, p_y, p_u, p_v );
215 C_YUV422_YUYV( p_line, p_y, p_u, p_v );
216 C_YUV422_YUYV( p_line, p_y, p_u, p_v );
217 #elif defined (MODULE_NAME_IS_i422_yuy2_mmx)
218 MMX_CALL( MMX_YUV422_YUYV );
221 for( i_x = ( p_vout->render.i_width % 8 ) / 2; i_x-- ; )
223 C_YUV422_YUYV( p_line, p_y, p_u, p_v );
225 p_y += i_source_margin;
226 p_u += i_source_margin_c;
227 p_v += i_source_margin_c;
228 p_line += i_dest_margin;
230 #if defined (MODULE_NAME_IS_i422_yuy2_mmx)
237 /*****************************************************************************
238 * I422_YVYU: planar YUV 4:2:2 to packed YVYU 4:2:2
239 *****************************************************************************/
240 static void I422_YVYU( vout_thread_t *p_vout, picture_t *p_source,
243 uint8_t *p_line = p_dest->p->p_pixels;
244 uint8_t *p_y = p_source->Y_PIXELS;
245 uint8_t *p_u = p_source->U_PIXELS;
246 uint8_t *p_v = p_source->V_PIXELS;
250 const int i_source_margin = p_source->p[0].i_pitch
251 - p_source->p[0].i_visible_pitch;
252 const int i_source_margin_c = p_source->p[1].i_pitch
253 - p_source->p[1].i_visible_pitch;
254 const int i_dest_margin = p_dest->p->i_pitch
255 - p_dest->p->i_visible_pitch;
257 #if defined (MODULE_NAME_IS_i422_yuy2_sse2)
259 if( 0 == (15 & (p_source->p[Y_PLANE].i_pitch|p_dest->p->i_pitch|
260 ((int)p_line|(int)p_y))) )
262 /* use faster SSE2 aligned fetch and store */
263 for( i_y = p_vout->render.i_height ; i_y-- ; )
265 for( i_x = p_vout->render.i_width / 16 ; i_x-- ; )
267 SSE2_CALL( SSE2_YUV422_YVYU_ALIGNED );
269 for( i_x = ( p_vout->render.i_width % 16 ) / 2; i_x-- ; )
271 C_YUV422_YVYU( p_line, p_y, p_u, p_v );
273 p_y += i_source_margin;
274 p_u += i_source_margin_c;
275 p_v += i_source_margin_c;
276 p_line += i_dest_margin;
280 /* use slower SSE2 unaligned fetch and store */
281 for( i_y = p_vout->render.i_height ; i_y-- ; )
283 for( i_x = p_vout->render.i_width / 16 ; i_x-- ; )
285 SSE2_CALL( SSE2_YUV422_YVYU_UNALIGNED );
287 for( i_x = ( p_vout->render.i_width % 16 ) / 2; i_x-- ; )
289 C_YUV422_YVYU( p_line, p_y, p_u, p_v );
291 p_y += i_source_margin;
292 p_u += i_source_margin_c;
293 p_v += i_source_margin_c;
294 p_line += i_dest_margin;
301 for( i_y = p_vout->render.i_height ; i_y-- ; )
303 for( i_x = p_vout->render.i_width / 8 ; i_x-- ; )
305 #if defined (MODULE_NAME_IS_i422_yuy2)
306 C_YUV422_YVYU( p_line, p_y, p_u, p_v );
307 C_YUV422_YVYU( p_line, p_y, p_u, p_v );
308 C_YUV422_YVYU( p_line, p_y, p_u, p_v );
309 C_YUV422_YVYU( p_line, p_y, p_u, p_v );
310 #elif defined (MODULE_NAME_IS_i422_yuy2_mmx)
311 MMX_CALL( MMX_YUV422_YVYU );
314 for( i_x = ( p_vout->render.i_width % 8 ) / 2; i_x-- ; )
316 C_YUV422_YVYU( p_line, p_y, p_u, p_v );
318 p_y += i_source_margin;
319 p_u += i_source_margin_c;
320 p_v += i_source_margin_c;
321 p_line += i_dest_margin;
323 #if defined (MODULE_NAME_IS_i422_yuy2_mmx)
330 /*****************************************************************************
331 * I422_UYVY: planar YUV 4:2:2 to packed UYVY 4:2:2
332 *****************************************************************************/
333 static void I422_UYVY( vout_thread_t *p_vout, picture_t *p_source,
336 uint8_t *p_line = p_dest->p->p_pixels;
337 uint8_t *p_y = p_source->Y_PIXELS;
338 uint8_t *p_u = p_source->U_PIXELS;
339 uint8_t *p_v = p_source->V_PIXELS;
343 const int i_source_margin = p_source->p[0].i_pitch
344 - p_source->p[0].i_visible_pitch;
345 const int i_source_margin_c = p_source->p[1].i_pitch
346 - p_source->p[1].i_visible_pitch;
347 const int i_dest_margin = p_dest->p->i_pitch
348 - p_dest->p->i_visible_pitch;
350 #if defined (MODULE_NAME_IS_i422_yuy2_sse2)
352 if( 0 == (15 & (p_source->p[Y_PLANE].i_pitch|p_dest->p->i_pitch|
353 ((int)p_line|(int)p_y))) )
355 /* use faster SSE2 aligned fetch and store */
356 for( i_y = p_vout->render.i_height ; i_y-- ; )
358 for( i_x = p_vout->render.i_width / 16 ; i_x-- ; )
360 SSE2_CALL( SSE2_YUV422_UYVY_ALIGNED );
362 for( i_x = ( p_vout->render.i_width % 16 ) / 2; i_x-- ; )
364 C_YUV422_UYVY( p_line, p_y, p_u, p_v );
366 p_y += i_source_margin;
367 p_u += i_source_margin_c;
368 p_v += i_source_margin_c;
369 p_line += i_dest_margin;
373 /* use slower SSE2 unaligned fetch and store */
374 for( i_y = p_vout->render.i_height ; i_y-- ; )
376 for( i_x = p_vout->render.i_width / 16 ; i_x-- ; )
378 SSE2_CALL( SSE2_YUV422_UYVY_UNALIGNED );
380 for( i_x = ( p_vout->render.i_width % 16 ) / 2; i_x-- ; )
382 C_YUV422_UYVY( p_line, p_y, p_u, p_v );
384 p_y += i_source_margin;
385 p_u += i_source_margin_c;
386 p_v += i_source_margin_c;
387 p_line += i_dest_margin;
394 for( i_y = p_vout->render.i_height ; i_y-- ; )
396 for( i_x = p_vout->render.i_width / 8 ; i_x-- ; )
398 #if defined (MODULE_NAME_IS_i422_yuy2)
399 C_YUV422_UYVY( p_line, p_y, p_u, p_v );
400 C_YUV422_UYVY( p_line, p_y, p_u, p_v );
401 C_YUV422_UYVY( p_line, p_y, p_u, p_v );
402 C_YUV422_UYVY( p_line, p_y, p_u, p_v );
403 #elif defined (MODULE_NAME_IS_i422_yuy2_mmx)
404 MMX_CALL( MMX_YUV422_UYVY );
407 for( i_x = ( p_vout->render.i_width % 8 ) / 2; i_x-- ; )
409 C_YUV422_UYVY( p_line, p_y, p_u, p_v );
411 p_y += i_source_margin;
412 p_u += i_source_margin_c;
413 p_v += i_source_margin_c;
414 p_line += i_dest_margin;
416 #if defined (MODULE_NAME_IS_i422_yuy2_mmx)
423 /*****************************************************************************
424 * I422_IUYV: planar YUV 4:2:2 to interleaved packed IUYV 4:2:2
425 *****************************************************************************/
426 static void I422_IUYV( vout_thread_t *p_vout, picture_t *p_source,
430 msg_Err( p_vout, "I422_IUYV unimplemented, please harass <sam@zoy.org>" );
433 /*****************************************************************************
434 * I422_cyuv: planar YUV 4:2:2 to upside-down packed UYVY 4:2:2
435 *****************************************************************************/
436 static void I422_cyuv( vout_thread_t *p_vout, picture_t *p_source,
439 uint8_t *p_line = p_dest->p->p_pixels + p_dest->p->i_visible_lines * p_dest->p->i_pitch;
440 uint8_t *p_y = p_source->Y_PIXELS;
441 uint8_t *p_u = p_source->U_PIXELS;
442 uint8_t *p_v = p_source->V_PIXELS;
446 for( i_y = p_vout->render.i_height ; i_y-- ; )
448 for( i_x = p_vout->render.i_width / 8 ; i_x-- ; )
450 p_line -= 2 * p_dest->p->i_pitch;
452 #if defined (MODULE_NAME_IS_i422_yuy2)
453 C_YUV422_UYVY( p_line, p_y, p_u, p_v );
454 C_YUV422_UYVY( p_line, p_y, p_u, p_v );
455 C_YUV422_UYVY( p_line, p_y, p_u, p_v );
456 C_YUV422_UYVY( p_line, p_y, p_u, p_v );
457 #elif defined (MODULE_NAME_IS_i422_yuy2_mmx)
458 MMX_CALL( MMX_YUV422_UYVY );
462 #if defined (MODULE_NAME_IS_i422_yuy2_mmx)
464 #elif defined (MODULE_NAME_IS_i422_yuy2_sse2)
469 /*****************************************************************************
470 * I422_Y211: planar YUV 4:2:2 to packed YUYV 2:1:1
471 *****************************************************************************/
472 #if defined (MODULE_NAME_IS_i422_yuy2)
473 static void I422_Y211( vout_thread_t *p_vout, picture_t *p_source,
476 uint8_t *p_line = p_dest->p->p_pixels + p_dest->p->i_visible_lines * p_dest->p->i_pitch;
477 uint8_t *p_y = p_source->Y_PIXELS;
478 uint8_t *p_u = p_source->U_PIXELS;
479 uint8_t *p_v = p_source->V_PIXELS;
483 for( i_y = p_vout->render.i_height ; i_y-- ; )
485 for( i_x = p_vout->render.i_width / 8 ; i_x-- ; )
487 C_YUV422_Y211( p_line, p_y, p_u, p_v );
488 C_YUV422_Y211( p_line, p_y, p_u, p_v );
495 /*****************************************************************************
496 * I422_YV12: planar YUV 4:2:2 to planar YV12
497 *****************************************************************************/
498 #if defined (MODULE_NAME_IS_i422_yuy2)
499 static void I422_YV12( vout_thread_t *p_vout, picture_t *p_source,
502 uint16_t i_dpy = p_dest->p[Y_PLANE].i_pitch;
503 uint16_t i_spy = p_source->p[Y_PLANE].i_pitch;
504 uint16_t i_dpuv = p_dest->p[U_PLANE].i_pitch;
505 uint16_t i_spuv = p_source->p[U_PLANE].i_pitch;
506 uint16_t i_width = p_vout->render.i_width;
507 uint16_t i_y = p_vout->render.i_height;
508 uint8_t *p_dy = p_dest->Y_PIXELS + (i_y-1)*i_dpy;
509 uint8_t *p_y = p_source->Y_PIXELS + (i_y-1)*i_spy;
510 uint8_t *p_du = p_dest->U_PIXELS + (i_y/2-1)*i_dpuv;
511 uint8_t *p_u = p_source->U_PIXELS + (i_y-1)*i_spuv;
512 uint8_t *p_dv = p_dest->V_PIXELS + (i_y/2-1)*i_dpuv;
513 uint8_t *p_v = p_source->V_PIXELS + (i_y-1)*i_spuv;
518 memcpy(p_dy, p_y, i_width); p_dy -= i_dpy; p_y -= i_spy;
519 memcpy(p_dy, p_y, i_width); p_dy -= i_dpy; p_y -= i_spy;
520 memcpy(p_du, p_u, i_width/2); p_du -= i_dpuv; p_u -= 2*i_spuv;
521 memcpy(p_dv, p_v, i_width/2); p_dv -= i_dpuv; p_v -= 2*i_spuv;