Re: [PATCH 1/4] idr: proper invalid argument handling

[Andrew Morton <akpm@linux-foundation.org>]

On Tue, 22 Apr 2014 12:16:56 -0700 Andrew Morton <akpm@linux-foundation.org> wrote:

> On Tue, 22 Apr 2014 18:16:18 +0800 Lai Jiangshan <laijs@cn.fujitsu.com> wrote:
> 
> > When the arguments passed by the caller are invalid, WARN_ON_ONCE()
> > is proper than BUG_ON() which may crash the kernel.
> > 
> > ida_remove()/idr_remove() add checks for "id < 0".
> > BUG_ON() in ida_simple_remove() is simply removed, due to
> > ida_remove() already checks for "id < 0".
> > 
> > In idr_alloc(), it still returns -ENOSPC when "start == end",
> > but it returns -EINVAL when "max < start" while old code returns
> > -ENOSPC. -EINVAL is proper here, the caller must passed wrong
> > arguments.
> > 
> > ida_simple_get()'s argument-checks are changed as the same as
> > idr_alloc().
> 
> This patch doesn't apply.
> 
> > @@ -551,10 +553,7 @@ void idr_remove(struct idr *idp, int id)
> >  	struct idr_layer *p;
> >  	struct idr_layer *to_free;
> >  
> > -	if (id < 0)
> > -		return;
> > -
> > -	if (id > idr_max(idp->layers)) {
> > +	if (id < 0 || id > idr_max(idp->layers)) {
> >  		idr_remove_warning(id);
> >  		return;
> >  	}
> 
> 3.15-rc2's idr_remove() has a call to sub_remove() in there, but
> whatever-kernel-you're-using does not.

Ah, it's based on your other idr patchset.  That's what I get for
working in reverse time order.