Jason Tackaberry wrote: > On Tue, 2009-01-27 at 18:36 +0100, Duncan Webb wrote: >> The CPU is running at 80% which is way to much. > > Wow. When the UI is otherwise idle, anything over 2% is too much. > > >> The poll time is set to 200 ms, which is much higher than I would >> normally use (50 ms). > > If nothing else is going on (e.g. no timers or activity on any of the > monitored file descriptors), the select() timeout might as well be > infinity. The notifier will use 30 seconds as its "infinity" value. > > Waking up every 50ms or even 200ms for that matter when there is no need > to is just wasting power, really. > > >> I reckon this problem is going to be quite tricky to find. > > It doesn't sound fun. Knowing nothing about the 1.x codebase, I'm not > sure what else to say except to wish you luck. :)
I've found the source of the problem, not freevo or Python related... It seems to be either SDL or DirectFB that is using a lot of CPU. I wrote a little test program for Andrea Mennucc to test screen bliting with numpy. Adding a 1 second sleep between blits the CPU usage is about 50% using pygame, SDL and DirectFB. When running this on X the CPU usage was minimal 0-0.5%. Also using the pygoom test program on DirectFB or X gave some surprising results. The framerate on DirectFB was about 5 and on X about 27. In conclusion the problem lies with either SDL or DirectFB or the interface between them. The next problem is to find out what is the cause. Back to recompiling old version of DirectFB... Duncan ------------------------------------------------------------------------------ This SF.net email is sponsored by: SourcForge Community SourceForge wants to tell your story. http://p.sf.net/sfu/sf-spreadtheword _______________________________________________ Freevo-devel mailing list Freevo-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/freevo-devel
