Re: [patch] 2.4.2 Advantech WDT driver

[Jeff Garzik <jgarzik@mandrakesoft.com>]

Marek Michalkiewicz wrote:
> 
> > Why is lock_kernel necessary?
> 
> Well, it is there in 2.4.2 acquirewdt.c (which this driver is based on -
> really only minimal changes, the hardware is only slightly different).
> I can remove it if you tell me it's really not necessary.

Sounds like it got caught in an audit and is necessary, I didn't look
closely...


> > > +       spin_lock_init(&advwdt_lock);
> > > +       misc_register(&advwdt_miscdev);
> >
> > check return code for error
> 
> Again, acquirewdt.c doesn't do it with the misc_register, request_region
> and register_reboot_notifier calls.  (Should it?  Cc: Alan Cox)

Yes, the return codes should be checked.  :)

-- 
Jeff Garzik       | "You see, in this world there's two kinds of
Building 1024     |  people, my friend: Those with loaded guns
MandrakeSoft      |  and those who dig. You dig."  --Blondie