Viggo Fredriksen wrote: > Hi, > > upgraded to cvs today, and got the errors appended in the log. > The error is 100% reproducable when playing video files with > mplayer (using pygame/dfbmga on crtc2). The video starts, but > will not end properly when sending STOP (the screen stays black). > If stop is pushed once again, the freevo display is showed. > > I don't have the means to check what really causes this atm., but > just wanted to give a heads up incase it's a simple bug :)
It's a very strange bug we (crunchy and I) don't understand. It only happens when using dfb. | (!!!) *** ONCE [accessing video memory during suspend] *** [surfaces.c:772 in dfb_surface_software_lock()] That doesn't sound good. Would be nice to know when this happens. | __sockets[ IO_EXCEPT ].keys(), timeout / 1000.0 ) | error: (4, 'Interrupted system call') That is a very strange bug. Why Interrupted system call? The system call running is the select, but who interrupted it? Is it dfb? Why only at this point? Maybe someone is sending a signal? Can you add a signal handler for all signals to find out? | (*) Direct/Thread: Running 'Keyboard Input' (INPUT, 24588)... Thread doesn't sound good. Freevo doesn't use threads, but maybe dfb? Can you check this? How many freevo processes to you see (it should be two). Dischi -- Time passed, which, basically, is its job. -- Terry Pratchett (in: Equal Rites)
pgpO7YppZDSjj.pgp
Description: PGP signature
