Heh, there's like 6 threads. I don't have time to play with it now, but this is what I recall. Most of the threads are busy doing there normal poll() and select() loops. One of them gets stuck in sig_procmask(SUSPEND,...). If I sent it SIGCONT, strace shows that it gets the CONT and then suspends itself again. If I have time tonight I'll play with it in more detail.
It's easier to recompile and test new versions while at work because most the work is idle. But watching straces requires more attention =P On Wed, Mar 06, 2002 at 12:56:18PM +0100, [EMAIL PROTECTED] alleged: > Garrick, maybe you can run a trace on xmms to get a clue? > > I cannot reproduce it either. > > danny > > On 6 Mar 2002, Guillaume Cottenceau wrote: > > > garrick <[EMAIL PROTECTED]> writes: > > > > > No love. I compiled xmms-1.2.7-pre1 with -O0 using gcc-2.96-0.76mdk, > > > gcc3.0-3.0.4-2mdk, and a gcc-2.95.3 that I compiled myself. The longest > > > I got was 38 minutes with the 3.0 version. > > > > maybe you could try with latest 1.2.7 stable version? but there > > is little chance it changes anything, there is nothing relevant > > in changelog... > > > > > The display is not getting locked. I can still minimize/maximize, > > > repeatedly pause/unpause, and a few other buttons. But as soon as I > > > press the stop or play buttons, *then* the display is frozen. > > > > we have two p4 here (one i845 and one i850) and on both we could > > not reproduce the problem :-(. > > > > > Oh well... freeamp is working ok. > > > > fine ;p. > > > > > > >
