Re: [RFC][PATCH 2/6] x86, nmi: create new NMI handler routines

[Don Zickus <dzickus@redhat.com>]

On Mon, Aug 22, 2011 at 04:13:49PM +0200, Peter Zijlstra wrote:
> On Fri, 2011-08-19 at 16:37 -0400, Don Zickus wrote:
> > +int register_nmi_handler(unsigned int type, nmi_handler_t handler,
> > +                       unsigned long nmiflags, const char *devname)
> > +{
> > +       struct nmiaction *action;
> > +       int retval;
> > +
> > +       if (!handler)
> > +               return -EINVAL;
> > +
> > +       action = kzalloc(sizeof(struct nmiaction), GFP_KERNEL);
> > +       if (!action)
> > +               return -ENOMEM;
> > +
> > +       action->handler = handler;
> > +       action->flags = nmiflags;
> > +       action->name = devname; 
> 
> So you assume the string passed as devname remains stable..

yeah, I just kinda copied the irq stuff, thinking if it works there, it
should work here.  It does seem odd, but I can change that to strncpy to
be safer.

> 
> 
> > +
> > +       retval = __setup_nmi(type, action);
> > +
> > +       if (retval)
> > +               kfree(action);
> > +       
> > +       return retval;
> > +}
> > +EXPORT_SYMBOL(register_nmi_handler);
> 
> Also, I'd use _GPL, the current register_die_notifier() is also.
> Preferably I'd not export NMI stuff at all.

Sure.  Well there are a few drivers that want to register.  Unless you
want to make a rule that all nmi handlers have to be built-in?

Cheers,
Don