Re: [PATCH] VFS: Sanity check mount flags passed to change_mnt_propagation()

[Valerie Aurora <vaurora@redhat.com>]

On Thu, Aug 26, 2010 at 07:14:36PM -0600, Matthew Wilcox wrote:
> On Thu, Aug 26, 2010 at 04:03:18PM -0400, Valerie Aurora wrote:
> > +	/* Fail if any non-propagation flags are set */
> > +	if (type & ~(MS_SHARED | MS_PRIVATE | MS_SLAVE | MS_UNBINDABLE))
> > +		return 0;
> > +	/* Only one propagation flag should be set */
> > +	if (((type & (MS_SHARED)) && (type & ~MS_SHARED)) ||
> > +	    ((type & (MS_PRIVATE)) && (type & ~MS_PRIVATE)) ||
> > +	    ((type & (MS_SLAVE)) && (type & ~MS_SLAVE)) ||
> > +	    ((type & (MS_UNBINDABLE)) && (type & ~MS_UNBINDABLE)))
> > +		return 0;
> 
> Hrm.  I think we can do this a bit more pithily.
> 
> 	/* Only one propagation flag should be set, and no others */
> 	if (hweight32(type) != 1 &&
> 	    (type & ~(MS_SHARED | MS_PRIVATE | MS_SLAVE | MS_UNBINDABLE))
> 		return 0;
> 
> Too clever?

I was hoping someone would go find the best bitop for me, thanks. :)
hweight32() is an awkward name but the comment makes it clear.  I'm
happy with either.

Thanks for the help,

-VAL