On Mon, 2011-04-18 at 23:19 +0200, GÃran Uddeborg wrote:
> I've been experimenting a bit by installing bits and pieces from F15,
> without doing a complete install. Now I realised my beep had
> disappeared. I'm talking about the nice classical square wave PC
> speaker beep. The thing that sounds when I, for example, try to do
> tab-expansion in an xterm, and there are several alternative ways to
> continue.
What a wretched, wretched noise.
> I note that "xkbbell" is silent, but "xkbbell -force" still generates
> the beep.
The xkbbell source has:
if (force) ok= XkbForceBell(dpy,volume);
else if (nobeep) ok= XkbBellEvent(dpy,win,volume,nameAtom);
else ok= XkbBell(dpy,win,volume,nameAtom);
Drilling down into the implementation, the only real difference is that
XkbForceBell sets a "force" flag in the X request, so now it's off to
the server to see what that implements. That ends up doing (apologies
again for the style, the XKB code really is repulsive):
if ((force||(xkbi->desc->ctrls->enabled_ctrls&XkbAudibleBellMask))&&
(!eventOnly)) {
if (kbd->kbdfeed->BellProc)
(*kbd->kbdfeed->BellProc)(percent,kbd,(pointer)pCtrl,class);
}
Which implies that, somehow, the device is getting its audible bell
control disabled. However, the keyboard code looks like it creates
devices with that bit set, so I would assume it's something in your
session doing it.
If I were to go about debugging this any further, I'd ssh into the
machine under test, launch my X server under gdb with a breakpoint set
on ProcXkbSetControls (which is how those flags would get modified), and
then launch my session and see who's calling it.
- ajax
Attachment:
signature.asc
Description: This is a digitally signed message part
-- test mailing list test@xxxxxxxxxxxxxxxxxxxxxxx To unsubscribe: https://admin.fedoraproject.org/mailman/listinfo/test
