April 2006 - MPlayer-dev-eng - lists.mplayerhq.hu

[PATCH] Prevent extra flipping of frame buffer at end of movie
by Timothy Lee 05 May '06

05 May '06

Dear all, This one line patch clears blit_frame after each page flip to prevent an undesired flip at the end of a movie when double-buffering is enabled. Regards, Timothy

4 4

pre8 update
by Roberto Togni 04 May '06

04 May '06

Hi all Update on pre8 status. * old release critical bugs: - realaudio in mkv: fixed - realvideo in mkv on ppc: fixed - faad2 update: done * setting up the mirror system no idea, but attila told that the server can handle the load alone * new release critical bugs: - 3gp with amr audio broken: anybody? I can make a trivial hack on 4cc or revert the twos fix if no other patch is available before the release * Name: there are already some binary packages called pre8, so I propose to skip it and go to pre9 to avoid confusion. * Target: this weekend (else Monday or Tuesday) * Open tasks (help needed): - check and complete changelog, including ffmpeg part Do you know other release critical bugs? Please commit all safe patches you have pending and delay big/risky patches unless you have a good reason. Scream stop asap if you have a reason to halt the release process. Ciao, Roberto

26 53

[PATCH] -endpos support in mplayer::4th try
by Paul TT 03 May '06

03 May '06

adds option -endpos to mplayer patch reworked to apply to latest cvs, corrected a bug in time/size parsing function. includes modifications to mplayer.1 man and externalization of parse_end_at function and field, this last passage thanx to alban bedel :-) discuss and apply or tell me to apply it ;-P PaulTT PS: reminding that i'll be working on -ss/-sb code and better -endpos <time> for dvd after apllication of this patch :-)

3 3

MPlayer GUI
by Piotr Stachura 02 May '06

02 May '06

Hi, some time ago, there was a little discussion about GUI for MPlayer. At the end nothing was born. I'd like to make new one, but based on QT. I know that you guys dislike C++ code but is there any chance that C++ style GUI will be included into MPlayer, or it's forbidden? -- Piotr Stachura

18 46

[PATCH] Solaris/x86 sysi86() type conflict
by Derek E. Lewis 30 Apr '06

30 Apr '06

sysi86() is used in loader/ldt_keeper.c. sysi86() was introduced in Solaris 2.0 and was compatible until Solaris 9. There was a tweak in Solaris 10 with the args -- the function went from receiving them in a struct to them being unpacked. The current ldt_keeper.c assumes that sysi86()'s args aren't unpacked, and thus provides a "correct" declaration. I've added a new test to the 'configure' script to test for the existence of the new sysi86() declaration. This patch avoids a build failure that would normally occur because of the type conflict in releases >= 10 of Solaris. This bug should be addressed before pre8 is released, as it breaks all builds on Solaris 10 or greater on x86. Thanks, -- Derek E. Lewis http://riemann.solnetworks.net/~dlewis/weblog dlewis(a)solnetworks.net

2 3

HEADS UP: Recent floats to doubles commit breaks v4l
by Ismail Donmez 28 Apr '06

28 Apr '06

Hi, Latest commits to libmpdemux/demuxer.c and libmpdemux/demuxer.h by uau breaks v4l : Before the change I get : cartman@southpark mplayer $ ./mplayer tv:// -tv driver=v4l:device=/dev/video0 MPlayer dev-CVS-060428-14:53-3.4.6 (C) 2000-2006 MPlayer Team CPU: Intel Pentium M Dothan (Family: 6, Stepping: 8) CPUflags: MMX: 1 MMX2: 1 3DNow: 0 3DNow2: 0 SSE: 1 SSE2: 1 Compiled for x86 CPU with extensions: MMX MMX2 SSE SSE2 Playing tv://. Selected driver: v4l name: Video 4 Linux input author: Alex Beregszaszi comment: under development ================================================================= WARNING: YOU ARE USING V4L DEMUXER WITH V4L2 DRIVERS!!! As the V4L1 compatibility layer is broken, this may not work. If you encounter any problems, use driver=v4l2 instead. Bugreports on driver=v4l with v4l2 drivers will be ignored. ================================================================= Selected device: Philips SPC 900NC webcam Capabilites: capture Device type: 1 Supported sizes: 160x120 => 640x480 Inputs: 1 0: Webcam: (tuner:0, norm:pal) Using input 'Webcam' Selected input hasn't got a tuner! Opening video filter: [pp] ========================================================================== Opening video decoder: [raw] RAW Uncompressed Video VDec: vo config request - 160 x 120 (preferred colorspace: Planar YV12) [PP] Using external postprocessing filter, max q = 6. VDec: using Planar YV12 as output csp (no 0) Movie-Aspect is undefined - no prescaling applied. VO: [xv] 160x120 => 160x120 Planar YV12 Selected video codec: [rawyv12] vfm: raw (RAW YV12) ========================================================================== Audio: no sound Starting playback... MJP: returning! ?% ??% ??,?% 0 0 Exiting... (Quit) After the change: cartman@southpark mplayer $ ./mplayer tv:// -tv driver=v4l:device=/dev/video0 MPlayer dev-CVS-060428-14:53-3.4.6 (C) 2000-2006 MPlayer Team CPU: Intel Pentium M Dothan (Family: 6, Stepping: 8) CPUflags: MMX: 1 MMX2: 1 3DNow: 0 3DNow2: 0 SSE: 1 SSE2: 1 Compiled for x86 CPU with extensions: MMX MMX2 SSE SSE2 Playing tv://. Selected driver: v4l name: Video 4 Linux input author: Alex Beregszaszi comment: under development ================================================================= WARNING: YOU ARE USING V4L DEMUXER WITH V4L2 DRIVERS!!! As the V4L1 compatibility layer is broken, this may not work. If you encounter any problems, use driver=v4l2 instead. Bugreports on driver=v4l with v4l2 drivers will be ignored. ================================================================= Selected device: Philips SPC 900NC webcam Capabilites: capture Device type: 1 Supported sizes: 160x120 => 640x480 Inputs: 1 0: Webcam: (tuner:0, norm:pal) Using input 'Webcam' Selected input hasn't got a tuner! MPlayer interrupted by signal 11 in module: video_read_properties - MPlayer crashed by bad usage of CPU/FPU/RAM. Recompile MPlayer with --enable-debug and make a 'gdb' backtrace and disassembly. Details in DOCS/HTML/en/bugreports_what.html#bugreports_crash. - MPlayer crashed. This shouldn't happen. It can be a bug in the MPlayer code _or_ in your drivers _or_ in your gcc version. If you think it's MPlayer's fault, please read DOCS/HTML/en/bugreports.html and follow the instructions there. We can't and won't help unless you provide this information when reporting a possible bug. MPlayer interrupted by signal 11 in module: free_demuxer --

1 1

[PATCH] dis-/enable playlist advancement
by Mark Pustjens 28 Apr '06

28 Apr '06

Hello, This patch adds commandline options to dis- or enable playlist advancement. It also add a new command which can be bound to a key. You might want to be able to disable playlist advancement if you are particulairy fond of a song. It is also usefull if mplayer is used in a kiosk type application. I have modified the manpage to include the commandline options. Please let me know if there is any other documentation i should update. Greetings, Mark Pustjens -- The class was learning about some revolt in which some peasants had wanted to stop being peasants and, since the nobles had won, had stopped being peasants *really quickly*. (Soul Music)

2 1

vf_eq{,2} buffer alignment
by Alan Curry 27 Apr '06

27 Apr '06

vf_eq and vf_eq2 both pass on a malloc()ed buffer to the next filter in the chain as part of an mp_image_t. When the next filter is vf_scale using altivec operations, it expects the image plans to begin at 16-byte boundaries so it ends up putting the pixels in the wrong place. If you're altivec-capable, the effect is observable with mplayer -vf eq2=1.0:1.0:0.1,scale -vo x11 yourmovie The misalignment depends on the whims of malloc, but in my case I usually get everything shifted 8 pixels to the right, with what should be the rightmost 8 pixels appearing instead on the left. The following patch changes eq and eq2 to align their buffers at 16-byte boundaries. Is this OK, or should I instead try to make scale more tolerant of unaligned input? Index: libmpcodecs/vf_eq.c =================================================================== RCS file: /cvsroot/mplayer/main/libmpcodecs/vf_eq.c,v retrieving revision 1.12 diff -u -r1.12 vf_eq.c --- libmpcodecs/vf_eq.c 15 Apr 2006 20:46:54 -0000 1.12 +++ libmpcodecs/vf_eq.c 22 Apr 2006 21:45:03 -0000 @@ -4,6 +4,9 @@ #include <inttypes.h> #include "config.h" +#ifdef HAVE_MALLOC_H +#include <malloc.h> +#endif #include "mp_msg.h" #include "cpudetect.h" #include "asmalign.h" @@ -131,7 +134,7 @@ dmpi->stride[1] = mpi->stride[1]; dmpi->stride[2] = mpi->stride[2]; - if (!vf->priv->buf) vf->priv->buf = malloc(mpi->stride[0]*mpi->h); + if (!vf->priv->buf) vf->priv->buf = memalign(16, mpi->stride[0]*mpi->h); if ((vf->priv->brightness == 0) && (vf->priv->contrast == 0)) dmpi->planes[0] = mpi->planes[0]; Index: libmpcodecs/vf_eq2.c =================================================================== RCS file: /cvsroot/mplayer/main/libmpcodecs/vf_eq2.c,v retrieving revision 1.12 diff -u -r1.12 vf_eq2.c --- libmpcodecs/vf_eq2.c 15 Apr 2006 20:46:54 -0000 1.12 +++ libmpcodecs/vf_eq2.c 22 Apr 2006 21:45:03 -0000 @@ -16,6 +16,9 @@ #include <inttypes.h> #include "config.h" +#ifdef HAVE_MALLOC_H +#include <malloc.h> +#endif #include "mp_msg.h" #include "cpudetect.h" #include "asmalign.h" @@ -240,13 +243,14 @@ eq2->buf_w[1] = eq2->buf_w[2] = src->w >> src->chroma_x_shift; eq2->buf_h[1] = eq2->buf_h[2] = src->h >> src->chroma_y_shift; img_n = eq2->buf_w[0]*eq2->buf_h[0]; + free(eq2->buf[0]); if(src->num_planes>1){ img_c = eq2->buf_w[1]*eq2->buf_h[1]; - eq2->buf[0] = (unsigned char *) realloc (eq2->buf[0], img_n + 2*img_c); + eq2->buf[0] = memalign (16, img_n + 2*img_c); eq2->buf[1] = eq2->buf[0] + img_n; eq2->buf[2] = eq2->buf[1] + img_c; } else - eq2->buf[0] = (unsigned char *) realloc (eq2->buf[0], img_n); + eq2->buf[0] = memalign (16, img_n); } dst = vf_get_image (vf->next, src->imgfmt, MP_IMGTYPE_EXPORT, 0, src->w, src->h);

4 9

[PATCH] Chroma plane copy in vf_expand
by John Koleszar 27 Apr '06

27 Apr '06

The attached patch fixes a bug in vf_expand when copying the image into a destination buffer smaller than the source. This occurs when the destination buffer is packed but the source buffer uses aligned h&w. As I recall, this causes chroma problems on the top few rows in the best case and worst case segfaults. Since mpi.chroma_height is based off of mpi.height and not mpi.h, the padding data is copied into a buffer that might not be big enough. Theoretically this should provide a (small) speedup in some cases, since the padding data isn't being copied around. I can't reproduce this with mplayer CVS, since AFAIK my other pending patches to ve_raw are the only case where packed images are used, so this will have to be debated on its technical merits alone. John

4 3

RE: compiling vo_v4lw module, please help !!!!
by Scott Franzyshen 27 Apr '06

27 Apr '06

>>Author: Raul Lucientes >>Date: 2004-09-21 02:41 -600 >>To: mplayer-dev-eng >>Subject: [MPlayer-dev-eng] compiling vo_v4lw module, please help !!!! >>Hi, >> >>Im triyng to compile the mplayer (MPlayer-1.0pre5) with support for >>"vloopback" (patch vo_v4l-patch-04-09-2002.p1), I have followed the >>TilmannŽs procedure: >> >>http://www1.mplayerhq.hu/pipermail/mplayer-dev-eng/2001-November/002591.html >> >>But, when the compiling process arrive to the "vo_v4lw.c" module this >>error >>happens :-( >> OK! I know this hasn't been in the news lately, but I was think how useful this would be to have around. So I hacked it up to work for now, but I would like to work this back into main stream development. Any help would be greatly appreciated. I do not know most of the mplayer internal stuff. Anyway, here is the patch (below) against MPlayer-1.0pre7try2 (for now ... I am working on bringing it up to cvs) Just a reminder to you all about this back in 2001 Tilmann Bitterberg created this patch; >>Author: Tilmann Bitterberg >>Date: 2001-11-27 09:49 -700 >>To: mplayer-dev-eng >>Subject: [MPlayer-dev-eng] New vo_v4lw video output module >>Hi, >> >>I hacked together a new libvo module called vo_v4lw.c. It >>provides write support for the video4linux api. This means that >>mplayer can act as a video4linux device providing (uncompressed) >>YUV/RGB data to other v4l applications like xawtv, mp1e etc. So why do I think this would be handy? Well I was thinking about directfb and how nice it would be to use the output of mplayer as a videoprovider within a directfb application. However, the "-vo directfb" is limited to only outputting fullscreen to the directfb screen. This doesn't allow any directfb api interaction. (No overlay, alpha channel, etc.) So, If I was able to use directfb's videodev(v4l) videoprovider, I could use an mplayer's ( or mplayers' ) output(s) within a direcfb application rendered to any directfb surface.(not just the whole screen) I got this to work using the following setup; MPlayer-1.0pre7try2 - Might work against CVS (haven't tried yet) This Patch vloopback-0.91 linux-2.4.27 - debian sarge source - I patched MPlayer with this patch - I built vloopback kernel module and insmod it (if you don't have a real v4l device don't forget to modprobe videodev first) - I then built MPlayer with --enable-v4lw - Then I start player outputting to v4l; mplayer -vo v4lw -vf scale=640:480 The.Simpsons.avi - The I start another instance of MPlayer playing the v4l stream; mplayer -tv device=/dev/video1:driver=v4l:width=640:height=480 -vo directfb tv:// - This works for me, however this patch needs to be cleaned up and brought up to 2006 development standards. also there seems to be a problem with the current directfb cvs version of the v4l videoprovider. It isn't working. I should be able to do something like dfbsee /dev/video1 to watch the v4l stream ... but it don't work :( Besides directfb. There are tons of applications out there that use the v4l standard ... This would also add a level of virtualization to real hardware devices. Thanks sfranzyshen ==================== PATCH ==================================== diff -uraN MPlayer-1.0pre7try2/configure MPlayer-1.0pre7try2-v4lw/configure --- MPlayer-1.0pre7try2/configure 2005-04-13 05:46:35.000000000 -0600 +++ MPlayer-1.0pre7try2-v4lw/configure 2006-04-26 13:54:47.000000000 -0600 @@ -249,6 +249,7 @@ (check for /dev/mga_vid) [autodetect] --enable-xmga build with mga_vid X Window support (check for X & /dev/mga_vid) [autodetect] + --enable-v4lw build with video4linux write support [autodetect] --enable-xv build with Xv render support for X 4.x [autodetect] --enable-xvmc build with XvMC acceleration for X 4.x [disable] --enable-vm build with XF86VidMode support for X11 [autodetect] @@ -1343,6 +1344,7 @@ _real=auto _live=auto _xinerama=auto +_v4lw=auto _mga=auto _xmga=auto _vm=auto @@ -1550,6 +1552,8 @@ --disable-live) _live=no ;; --enable-xinerama) _xinerama=yes ;; --disable-xinerama) _xinerama=no ;; + --enable-v4lw) _v4lw=yes ;; + --disable-v4lw) _v4lw=no ;; --enable-mga) _mga=yes ;; --disable-mga) _mga=no ;; --enable-xmga) _xmga=yes ;; @@ -3704,6 +3708,19 @@ fi echores "$_gl" +echocheck "/proc/video/vloopback" +if test "$_v4lw" = auto ; then + _v4lw=no + test -d /proc/video/vloopback && _v4lw=yes +fi +if test "$_v4lw" = yes ; then + _def_v4lw='#define HAVE_V4LW 1' + _vosrc="$_vosrc vo_v4lw.c" + _vomodules="v4lw $_vomodules" +else + _def_v4lw='#undef HAVE_V4LW' +fi +echores "$_v4lw" echocheck "/dev/mga_vid" if test "$_mga" = auto ; then @@ -7466,6 +7483,7 @@ $_def_directx $_def_ggi $_def_3dfx +$_def_v4lw $_def_tdfxfb $_def_tdfxvid $_def_directfb diff -uraN MPlayer-1.0pre7try2/libvo/video_out.c MPlayer-1.0pre7try2-v4lw/libvo/video_out.c --- MPlayer-1.0pre7try2/libvo/video_out.c 2004-12-21 13:33:51.000000000 -0700 +++ MPlayer-1.0pre7try2-v4lw/libvo/video_out.c 2006-04-26 13:58:19.000000000 -0600 @@ -80,6 +80,7 @@ extern vo_functions_t video_out_syncfb; extern vo_functions_t video_out_fbdev; extern vo_functions_t video_out_fbdev2; +extern vo_functions_t video_out_v4lw; extern vo_functions_t video_out_svga; extern vo_functions_t video_out_png; extern vo_functions_t video_out_ggi; @@ -192,6 +193,9 @@ #ifdef HAVE_AA &video_out_aa, #endif +#ifdef HAVE_V4LW + &video_out_v4lw, +#endif #ifdef HAVE_CACA &video_out_caca, #endif diff -uraN MPlayer-1.0pre7try2/libvo/vo_v4lw.c MPlayer-1.0pre7try2-v4lw/libvo/vo_v4lw.c --- MPlayer-1.0pre7try2/libvo/vo_v4lw.c 1969-12-31 17:00:00.000000000 -0700 +++ MPlayer-1.0pre7try2-v4lw/libvo/vo_v4lw.c 2006-04-26 16:02:56.000000000 -0600 @@ -0,0 +1,423 @@ +/* + * vo_v4lw.c + * + * Copyright (C) Tilmann Bitterberg - Nov 2001 + * + * This 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, or (at your option) + * any later version. + * + * This 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 GNU Make; see the file COPYING. If not, write to + * the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA. + * + * Updated 4/27/2006 - sfranzyshen <sfranzyshen(a)hotmail.com> + * compiled against MPlayer-1.0pre7try2 + */ + +/* + Quick v4lw loopback howto: + + - get the vloopback kernel module from + <http://motion.technolust.cx/vloopback/index.html> + or (alternative) my hacked version from + <http://tibit.org/video/vloopback-0.90-tibit.tar.gz> + + - if using my version, you don't need to specify dev_offset. + + - untar and compile it + tar xzf vloopback-VERSION.tar.gz && cd vloopback-VERSION + make + + - load the vloopback module + insmod vloopback.o dev_offset=2 pipes=1 + + - Make the devices accessible + chmod og+rw /dev/v4l/video2 + chmod og+rw /dev/v4l/video3 + + - eventually you have to recompile mplayer to let + it see that it can talk v4l. + ./configure --enable-v4lw + to be sure + + - start mplayer + mplayer -vo v4lw movie.avi + + - start xawtv + xawtv -c /dev/v4l/video3 -geometry 800x600 + + - done + easy, isn't it. + + NOTES: + - avicap doesn't work yet. + I got mp1e to work with a my hacked vloopback.o, + you need to specify the exact size. + - no scaling of video output is possible right now. + - be sure to insmod vloopback with a dev_offset appropriate + for your system. If you have a real grabber card (eg. a bttv + which is at /dev/v4l/video0) you _HAVE_ to use at least a + dev_offset=1 or you'll crash your kernel. + - I am on devfs, if you're not you have to use /dev/video2 + instead of /dev/v4l/video2 and may have to create the + devices manually. + Eg. for video2: mknod c 81 2 /dev/video2 + - mplayer will tell you the v4lw output device from which you + have to read. + + Performance: + System: 2 * 1GHz Intel PIII (Coppermine) #SMP + MPEG1 at 768x576 + mplayer CPU Usage ca. 35%, xawtv 11% + Divx at 512x384 + mplayer CPU Usage ca. 20%, xawtv 10% + */ + +#include <unistd.h> +#include <stdlib.h> +#include <stdio.h> +#include <fcntl.h> +#include <string.h> +#include <sys/ioctl.h> +#include <sys/mman.h> +#include <signal.h> +#include <sys/wait.h> +#include <linux/videodev.h> + +#include "config.h" +#include "video_out.h" +#include "video_out_internal.h" + +#include "fastmemcpy.h" +#include "sub.h" +#include "aspect.h" +#include "mp_msg.h" + +static uint32_t image_width, image_height; +static int devout; +static uint8_t *image; +static unsigned int image_format=0; +static int bpp = 24; +static int fmt; + +static vo_info_t info = +{ + "video4linux write output", + "v4lw", + "Tilmann Bitterberg <god(a)tibit.org>", + "update 4/27/06 sfanzyshen <sfranzyshen(a)hotmail.com>" +}; + +LIBVO_EXTERN(v4lw) + +/* stolen from: + * vloopback/example/dummy.c + * returns 0 on successfull opening, !0 otherwise + */ +int check_vidpipe(void) +{ + FILE *vloopbacks; + char pipepath[255]; + char buffer[255]; + char *loop; /* internal loop nr */ + char *input; /* our output dev */ + char *istatus; + char *output; /* user should read from */ + char *ostatus; + + vloopbacks=fopen("/proc/video/vloopback/vloopbacks", "r"); + if (!vloopbacks) { + perror ("V4lW: Failed to open '/proc/video/vloopback/vloopbacks'"); + printf ("V4lW: You need to load the vlooback kernel module\n"); + return -1; + } + /* Read vloopback version */ + fgets(buffer, 255, vloopbacks); + /* Read explaination line */ + fgets(buffer, 255, vloopbacks); + while (fgets(buffer, 255, vloopbacks)) { + if (strlen(buffer)>1) { + buffer[strlen(buffer)-1]=0; + loop = strtok(buffer, "\t"); + input = strtok(NULL, "\t"); + istatus = strtok(NULL, "\t"); + output = strtok(NULL, "\t"); + ostatus = strtok(NULL, "\t"); + if (istatus[0] == '-') { + sprintf(pipepath, "/dev/%s", input); + devout = open (pipepath, O_RDWR); + if (devout >= 0) { + printf("V4lW: I am writing to: /dev/%s\n", input); + printf("V4lW: You should read from: /dev/%s\n", output); + /* Ok, found one */ + return 0; + } else { + printf ("V4lW: Error opening /dev/%s\n", input); + perror ("V4lW: Failed to open output video device"); + /* scan ahead */ + } + } + } + } + printf("V4lW: Unable to find a available vloopback device\n"); + printf("V4lW: Maybe insmod with a greater pipes= value helps\n"); + return -1; +} +/* copied from + * vloopback-0.90/example/invert.c + */ + +int start_pipe (int dev, int width, int height) +{ + struct video_capability vid_caps; + struct video_window vid_win; + struct video_picture vid_pic; + + if (ioctl (dev, VIDIOCGCAP, &vid_caps) == -1) { + perror ("ioctl (VIDIOCGCAP)"); + return (1); + } + if (ioctl (dev, VIDIOCGPICT, &vid_pic)== -1) { + perror ("ioctl VIDIOCGPICT"); + return (1); + } + /* this is probably brain dead */ + /* vid_pic.palette=VIDEO_PALETTE_RGB24; */ + /* vid_pic.palette=VIDEO_PALETTE_YUV420P; */ + vid_pic.palette = fmt; + + if (ioctl (dev, VIDIOCSPICT, &vid_pic)== -1) { + perror ("ioctl VIDIOCSPICT"); + return (1); + } + if (ioctl (dev, VIDIOCGWIN, &vid_win)== -1) { + perror ("ioctl VIDIOCGWIN"); + return (1); + } + vid_win.width = width; + vid_win.height = height; + if (ioctl (dev, VIDIOCSWIN, &vid_win)== -1) { + perror ("ioctl VIDIOCSWIN"); + return (1); + } + return 0; +} + +/* stolen from + vo_odivx.c + */ +/* orig size -- working */ +static uint32_t draw_slice(uint8_t *src[], int stride[], int w,int h,int x,int y) +{ + uint8_t *s; + uint8_t *d; + int i; + int dstride=image_width; + int size = image_width*image_height; + + // copy Y + d=image+dstride*y+x; + s=src[0]; + for(i=0;i<h;i++){ + memcpy(d,s,w); + s+=stride[0]; + d+=dstride; + } + + w/=2;h/=2;x/=2;y/=2; dstride/=2; + + // copy U + d=image + size + dstride*y+x; + s=src[1]; + for(i=0;i<h;i++){ + memcpy(d,s,w); + s+=stride[1]; + d+=dstride; + } + + // copy V + d=image + size + size/4 + dstride*y+x; + s=src[2]; + for(i=0;i<h;i++){ + memcpy(d,s,w); + s+=stride[2]; + d+=dstride; + } + + return 0; +} +/* + printf("%d, %d, %d, %d, %d, %d, %d\n", stride[0], stride[1], stride[2], w, h, x, y); + 720, 360, 360, 720, 16, 0, 0 + 720, 360, 360, 720, 16, 0, 16 + ... + 720, 360, 360, 720, 16, 0, 544 + 720, 360, 360, 720, 16, 0, 560 +*/ +#if 0 +/* Half size */ +static uint32_t draw_slice(uint8_t *src[], int stride[], int w,int h,int x,int y) +{ + uint8_t *s; + uint8_t *d; + int i; + int j; + int dstride = image_width; + int size = image_width*image_height; + + stride[0]<<=1; + stride[1]<<=1; + stride[2]<<=1; + + // copy Y + d=image+dstride/2*y+x; + s=src[0]; + for(i=0;i<h;i+=2){ + for (j=0; j<w; j+=2) + d[j>>1]=s[j]; + s+=stride[0]; + d+=dstride; + } + + w/=2;h/=2;x/=2;y/=2; dstride/=2; + + // copy U + d=image + size + dstride/2*y+x; + s=src[1]; + for(i=0;i<h;i+=2){ + for (j=0; j<w; j+=2) + d[j>>1]=s[j]; + s+=stride[1]; + d+=dstride; + } + + // copy V + d=image + size + size/4 + dstride/2*y+x; + s=src[2]; + for(i=0;i<h;i+=2){ + for (j=0; j<w; j+=2) + d[j>>1]=s[j]; + s+=stride[2]; + d+=dstride; + } + + return 0; +} +#endif +static void draw_osd(void) +{ +} + +static void put_image(void) +{ + int size = image_width*image_height*3; + if (write(devout, image, size) != size) { + perror("V4lW: Error writing image to pipe!"); + } +} + +static void flip_page(void) +{ + put_image(); +} + +static uint32_t draw_frame(uint8_t *src[]) +{ + /* untested */ + memcpy (image, src, image_width * image_height * 3); + return 0; +} + +static uint32_t query_format(uint32_t format) +{ + switch(format){ + case IMGFMT_YV12: + case IMGFMT_I420: + case IMGFMT_YUY2: + case IMGFMT_RGB24: + case IMGFMT_BGR24: + return 1; + } + return 0; +} + +static uint32_t config (uint32_t width, uint32_t height, uint32_t d_width, + uint32_t d_height, uint32_t fullscreen, char *title, + uint32_t format) +{ + + image_width = width; + image_height = height; + image_format = format; + + switch (image_format) { + case IMGFMT_YV12: + case IMGFMT_I420: + fmt = VIDEO_PALETTE_YUV420P; + break; + case IMGFMT_RGB24: + case IMGFMT_BGR24: + fmt = VIDEO_PALETTE_RGB24; + break; + } + + if (check_vidpipe() != 0) { + return 1; + } + + + if (start_pipe(devout, width, height) != 0) { + return 1; + } + + /* printf("V4lW: width: %d; height: %d; format: %d\n", width, height, format); */ + + image = (uint8_t *) malloc(width*height*3); + + if (!image) { + close (devout); + return 1; + } + /* clear the buffer, grey */ + memset(image,0x80,width*height*3); + + return 0; +} + +static const vo_info_t* +get_info(void) +{ + return &info; +} + +static void +uninit(void) +{ + if (image) + free(image); + close (devout); +} + + +static void check_events(void) +{ +} +static uint32_t preinit(const char *arg) +{ + return 0; +} + +static uint32_t control(uint32_t request, void *data, ...) +{ + switch (request) { + case VOCTRL_QUERY_FORMAT: + return query_format(*((uint32_t*)data)); + } + return VO_NOTIMPL; +} + ==================== PATCH ====================================

1 0