On Tue, 9 Sep 2014, Andrew Morton wrote:
> > kfree() is happy to accept NULL pointer and does nothing in such case.
> > It's reasonable to expect it to behave the same if ERR_PTR is passed to
> > it.
> >
> > Inspired by a9cfcd63e8d ("ext4: avoid trying to kfree an ERR_PTR
> > pointer").
> >
> > ...
> >
> > --- a/mm/slab.c
> > +++ b/mm/slab.c
> > @@ -3612,7 +3612,7 @@ void kfree(const void *objp)
> >
> > trace_kfree(_RET_IP_, objp);
> >
> > - if (unlikely(ZERO_OR_NULL_PTR(objp)))
> > + if (unlikely(ZERO_OR_NULL_PTR(objp) || IS_ERR(objp)))
> > return;
>
> kfree() is quite a hot path to which this will add overhead. And we
> have (as far as we know) no code which will actually use this at
> present.
We obviously don't, as such code will be causing explosions. This is meant
as a prevention of problems such as the one that has just been fixed in
ext4.
> How about a new
>
> kfree_safe(...)
> {
> if (IS_ERR(...))
> return;
> if (other-stuff-when-we-think-of-it)
> return;
> kfree(...);
> }
I think this unfortunately undermines the whole point of the patch ... if
the caller knows that he might potentially be feeding ERR_PTR() to
kfree(), he can as well check the pointer himself.
--
Jiri Kosina
SUSE Labs
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@xxxxxxxxx. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>