On 06 Jul 2002 14:23:01 +0200, Dominik Vogt wrote: > > Mikhael, I found this code snippet in FvwmEvent.c: > > /* migo (19-Aug-2000): synchronize on M_DESTROY_WINDOW */ > SetSyncMask(fd, M_DESTROY_WINDOW); > > I believe this causes fvwm and FvwmEvent to freeze sometimes when > the pointer is grabbed and windows are spawned and killed really > fast. The best way to reproduce this is to enable the "autoraise" > config I posted in another thread, then open any gtk application > (e.g. gedit) and tear out its menu or tool bar. Drag it around > (using the button on the menu bar, not the window title) so that > it's repeatedly swallowed by the gedit window and spit out again. > Sometimes it takes only two seconds until FvwmEvent and fvwm have > a deadlock, sometimes I can drag it around for five minutes > without any problems. > > I still don't understand the circumstances that lead to the > deadlock. Any additional debug code (frpintfs, efence, etc.) > dramatically reduces the probability of this happening, so its > certainly a race condition somewhere. If you can help with this > problem, please do. > > By the way, what was that patch good for in the first place? > > P.S.: It may be a problem with packets that are not sent > completely. I notice this effect in the past with FvwmConsole.
There was one bug report (saying that the destroy_window behaviour changed since 2.2.4) that I fixed (in both FvwmEvent and PositiveWrite): http://www.hpc.uh.edu/fvwm/archive/0008/msg00011.html http://www.hpc.uh.edu/fvwm/archive/0008/msg00133.html If you know a better fix, do it. I don't think I will have a time for FvwmEvent debugging this/next week. Regards, Mikhael. -- Visit the official FVWM web page at <URL:http://www.fvwm.org/>. To unsubscribe from the list, send "unsubscribe fvwm-workers" in the body of a message to [EMAIL PROTECTED] To report problems, send mail to [EMAIL PROTECTED]
