[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Iris-devel] more on fs problem & segfaults
From: |
Cédric Delfosse |
Subject: |
Re: [Iris-devel] more on fs problem & segfaults |
Date: |
15 May 2002 09:44:51 +0200 |
Le mar 14/05/2002 à 14:51, foser a écrit :
> Lo,
>
> > For me, it happens only when the desktop resolution and the fs
> > resolution is the same. It bugs at home and at work, same gnome, same
> > sawfish.
> >
> > The problem appears since "pressing 'f' to switch fullscreen".
>
> Even if i set desktop size == iris fs size, i still only get this
> sometimes when i keep f pressed for several seconds (but it's probably
> better for my monitors health not having to switch resolutions 10 times
> a second). You using something like dnetc , whats your proc usage when
> it happens (it only seems to happen when the usage is high here). Now I
> just made some changes and i can't reproduce it anymore, but that might
> just be luck or something. I can hardly ever reproduce it.
>
> I mean the window setup is the same as when it starts up with fullscreen
> config option toggled, it's hard to imagine that is the problem (since
> that worked perfectly for you it seems ?). The only things i can think
> off are that the GLWin window stuff doesn't get properly cleaned (in
> time?) or the X server just wasnt meant for this sort of mad updating
> (in my case, it takes minutes of continuously pressing 'f' to get this
> situation). Maybe it has todo with some sort of async situation with the
> draw thread, but since this all happens in the thread i don't think so.
> I'm afraid since i can't reproduce it i can't be of much help.
>
> I noticed btw that you reverted the knot change, now it segfaults again.
> It only does so when disabling and then re-enabling the plugin. I'm
> surprised it doesn't do so for you, but i suggest you leave it in as
> long as it helps me :)
That's very weird ! Maybe we have a memory corruption problem somewhere,
but it's not easy to debug a shared lib.
>
> Oh and hm, the disabling of the window size stuff.. i like to have a
> small window when i start up :) I know it's not very clean and i suppose
> it should be done by using some window resize event, but could we leave
> it around as long as there isn't anything better?
Ok, but please don't call iris_config_write (&config);
Do this directly:
xmms_cfg_write_int (f, section_name, "window_width",
config->window_width);
xmms_cfg_write_int (f, section_name, "window_height",
So that all the config stuff is not saved.
Thanks for your time,
-- Cédric