Re: [PATCH] make kernel ignore bogus partitions

["Mike Miller (OS Dev)" <mikem@beardog.cca.cpqcorp.net>]

On Thu, May 11, 2006 at 04:00:14AM -0700, Andrew Morton wrote:
> Andries Brouwer <Andries.Brouwer@cwi.nl> wrote:
> >
> > On Tue, May 09, 2006 at 12:41:38PM -0700, Andrew Morton wrote:
> > 
> > > > +		if (from+size-1 > get_capacity(disk)) {
> > > > +			printk(" %s: p%d exceeds device capacity, ignoring.\n", 
> > > > +				disk->disk_name, p);
> > > > +			continue;
> > > > +		}
> > > >  		add_partition(disk, p, from, size);
> > 
> > > Shouldn't that be
> > > 
> > > 	if (from+size > get_capacity(disk)) {
> > > 
> > > ?
> > 
> > Ha, you are awake.
> 
> Opinions differ.
> 
> > Yes, it should.
> > And no "ignoring". And no "continue". E.g.:
> > 
> > 	printk(" %s: warning: p%d exceeds device capacity.\n", ...);
> > 
> 
> So you're saying that after detecting this inconsistency we should proceed
> to use the partition anyway?
> 
> For what reason?

Using the partition will result in io errors when accessing beyond the end
of the device. Most users don't appreciate that. It makes them nervous
about the hw.

mikem