Re: Possible bug in keyboard.c (2.6.10)

[Dmitry Torokhov <dtor_core@ameritech.net>]

On Sunday 30 January 2005 18:21, Dmitry Torokhov wrote:
> On Sunday 30 January 2005 03:41, Al Viro wrote:
> > On Sat, Jan 29, 2005 at 12:25:10PM +0100, Vojtech Pavlik wrote:
> > > I know. As I said, this is a problem I know about, and will be fixed. I
> > > was mainly interested whether anyone sees further problems in scenarios
> > > which don't include device addition/removal.
> > > 
> > > We already fixed this in serio, and input and gameport are next in the
> > > list.
> > 
> > OK, I'll bite.  What's to guarantee that no events will happen in
> > the middle of serio_unregister_port(), right after we'd done
> > serio_remove_pending_events()?
> 
> At this point serio is disconnected from driver and serio_interrupt
> will only queue rescans only if serio->registered. I guess I will need
> to protect change to serio->registered and take serio->lock to be
> completely in clear.
> 

Thinking about it some more - serio driver should take care of shutting
off interrupt delivery either in ->close() or ->stop() methods so we
don't really need to worry about having new events delivered here.

-- 
Dmitry