Francesco Romani wrote: > On Wed, 2008-10-22 at 11:37 -0500, Carl Karsten wrote: >> Getting back to my vga2usb project that I left off 6 months ago. I am >> starting >> with what is simple, and expecting to be told "build from current source" and >> "those options don't make sense" > > It may be the case (and it can be the case) but one way or another a > segfault or anything else as serious as that is not acceptable, > especially in beta or later releases. > > As always, please be patient if my feedback comes late and don't be > afraid to remind to the list/bugtracker (berlios!)/me the still unsolved > issues.
http://developer.berlios.de/bugs/?group_id=10094 > >> Does anyone (ping Christian) know how up to date that is? > > It is two releases old. No critical fixes committed in the meantime, > anyway. So the segfault you're reporting is even less acceptable. Well good that I am not too far off :) > >> Where do I get trunk from these days, and is that the most cutting edge? > > The official repo is still (and will remain) CVS on cvs.exit1.org > on berlios you can find: > - mirror of official tarballs > - special-purposes branches (I feel SVN MUCH more handy for that, even > using a repository located on a different server). > stop (= well, there is also the now-official bugtracker and the > development blog, but the coding related stuff ends here :) > >> I don't need this to work until march, so I have some time to help debug >> whatever >> is active development. > > I hope the debug will take much less time :) well, that depends on how much time I can spare, and how quick the usb2vga driver dev reacts. I just hit him with What: Video4Linux API 1 ioctls and video_decoder.h from Video devices. When: December 2008 http://www.kernel.org/doc/Documentation/feature-removal-schedule.txt > >> [26944.623816] Linux video capture interface: v2.00 >> [26944.732571] Epiphan Systems VGA frame grabber driver version 3.19.2.0000 >> loaded >> [26944.733692] usbcore: registered new interface driver vga2usb >> [27202.812026] usb 5-4: new high speed USB device using ehci_hcd and address >> 3 >> [27202.944605] usb 5-4: configuration #1 chosen from 1 choice >> [27202.979123] VGA2USB is now attached to vga2usb-192 >> >> Is there some way to figure out that vga2usb-192 is /dev/video0? (I need to >> submit a report to vga2usb dev and tell him to log what /dev/videoX gets >> used.) > > You can just use the driver name reported by transcode's import module. I don't understand. > >> [EMAIL PROTECTED]:~/vga2usb/v2u$ transcode -x v4l,null -V rgb24 -i >> /dev/video0 -f 3 >> -g 1600x1200 -y raw,null -z -r 2 -f 10 -o test.avi >> transcode v1.1.0-beta1 (C) 2001-2003 Thomas Oestreich, 2003-2008 Transcode >> Team >> tcprobe: error while loading shared libraries: libmjpegutils-1.8.so.0: cannot >> open shared object file: No such file or directory > > That is not good, please install mjpegutils oh, that's the file it couldn't find. I thought it had to do with /dev/video0 > >> [transcode] V: auto-probing | /dev/video0 (FAILED) >> [transcode] V: import format | (null) in (module=v4l) >> [transcode] A: auto-probing | /dev/video0 (FAILED) >> [transcode] A: import format | AC3 in (module=null) >> [transcode] V: AV demux/sync | (1) sync AV at initial MPEG sequence >> [transcode] V: import frame | 1600x1200 1.33:1 encoded @ UNKNOWN >> [transcode] V: rescale frame | 800x600 1.33:1 (-r) >> [transcode] V: flip frame | yes >> [transcode] V: bits/pixel | 0.375 >> [transcode] V: decoding fps,frc | 10.000,0 >> [transcode] V: video format | RGB24 >> [transcode] A: import format | 0x2000 AC3 [48000,16,2] >> [transcode] A: export | disabled >> [transcode] V: encoding fps,frc | 10.000,11 >> [transcode] A: bytes per frame | 19200 (19200.000000) >> [transcode] A: adjustment | [EMAIL PROTECTED] >> [transcode] V: IA32/AMD64 accel | sse2 sse mmx cmove asm >> [transcode] V: video buffer | 10 @ 1600x1200 [0x1] >> [transcode] A: audio buffer | 10 @ 48000x2x16 >> [import_null.so] v0.2.0 (2002-01-19) (video) null | (audio) null >> [import_v4l.so] v0.0.6 (2006-05-08) (video) v4l | (audio) PCM >> [export_null.so] v0.1.2 (2001-08-17) (video) null | (audio) null >> [export_raw.so] v0.3.12 (2003-08-04) (video) * | (audio) MPEG/AC3/PCM >> [import_v4l.so] warning: this module is deprecated: please use import_v4l2 >> instead >> [import_v4l.so] video4linux video grabbing >> [import_v4l.so] device has no audio channel >> [import_v4l.so] (audio-): >> [import_v4l.so] picture: brightness= 0% hue= 0% colour= 0% contrast= 0% >> [import_v4l.so] 1 frame buffer(s) available >> [import_v4l.so] dropping -2147483623 video frames for AV sync >> [export_raw.so] codec=RGB, fps=10.000, width=800, height=600 >> Segmentation fault > > could you please report a backtrace? > you'll need: > - transcode compiled with debug symbols: it is the default when > compiling from sources, binary packages usually don't have them > - gdb > - enabling core dump on your box (ulimit -c unlimited once before to run > transcode and a writable working directory) > - after the segfault, you'll find the core on your working directory > (beware: it can be LARGE). > - now you can follow the instructions here: > http://www.transcoding.org/cgi-bin/transcode?Reporting_Crashes > - after obtaining the backtrace, you can safely delete the core. > >> [import_v4l.so] picture: brightness=50% hue=50% colour=50% contrast= 6% >> [import_v4l.so] critical: Cannot not set RGB picture attributes. ioctl >> VIDIOCSPICT: Invalid argument >> [import_v4l.so] critical: error grab init >> [transcode] critical: video import module error: OPEN failed >> [transcode] critical: failed to open input source > > I'll try to replicate (and solve it ASAP). Thanks. working on the bt now. Carl K
