Re: [patch v2] epoll use a single inode ...

[Davide Libenzi]

On Mon, 5 Mar 2007, Davide Libenzi wrote:

> On Mon, 5 Mar 2007, Linus Torvalds wrote:
> 
> > On Mon, 5 Mar 2007, Davide Libenzi wrote:
> > > @@ -763,15 +767,17 @@
> > >  	 * using the inode number.
> > >  	 */
> > >  	error = -ENOMEM;
> > > -	sprintf(name, "[%lu]", inode->i_ino);
> > >  	this.name = name;
> > > -	this.len = strlen(name);
> > > -	this.hash = inode->i_ino;
> > > +	this.len = sprintf(name, "[%p]", ep);
> > > +	this.hash = 0;
> > 
> > Please don't expose kernel pointers to user space.
> > 
> > It's much better to do something like
> > 
> > 	static unsigned int epoll_inode;
> > 
> > 	this.len = sprintf(name, "[%u]", ++epoll_inode);
> > 
> > if you just need some pseudo-unique name to distinguish two epoll things 
> > from each other (vs from a dup'ed fd).
> 
> Heh, this is what Al was saying ;)
> I'm fine with that, but how about counter cycles (going back to zero)? 
> Should we care to handle them correctly? 

Maybe a [TID.FD] format?



- Davide


-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/