Re: [PATCH v4 17/17] mm/sl[au]b: check if large object is valid in __ksize()

[Hyeonggon Yoo <42.hyeyoo@gmail.com>]

On Tue, Aug 23, 2022 at 05:12:01PM +0200, Vlastimil Babka wrote:
> On 8/17/22 12:18, Hyeonggon Yoo wrote:
> > If address of large object is not beginning of folio or size of
> > the folio is too small, it must be invalid. BUG() in such cases.
> > 
> > Cc: Marco Elver <elver@google.com>
> > Suggested-by: Vlastimil Babka <vbabka@suse.cz>
> > Signed-off-by: Hyeonggon Yoo <42.hyeyoo@gmail.com>
> > Reviewed-by: Vlastimil Babka <vbabka@suse.cz>
> > ---
> >  mm/slab_common.c | 5 ++++-
> >  1 file changed, 4 insertions(+), 1 deletion(-)
> > 
> > diff --git a/mm/slab_common.c b/mm/slab_common.c
> > index 9c273a5fb0d7..98d029212682 100644
> > --- a/mm/slab_common.c
> > +++ b/mm/slab_common.c
> > @@ -984,8 +984,11 @@ size_t __ksize(const void *object)
> >  
> >  	folio = virt_to_folio(object);
> >  
> > -	if (unlikely(!folio_test_slab(folio)))
> > +	if (unlikely(!folio_test_slab(folio))) {
> > +		BUG_ON(folio_size(folio) <= KMALLOC_MAX_CACHE_SIZE);
> > +		BUG_ON(object != folio_address(folio));
> >  		return folio_size(folio);
> > +	}
> >  
> >  	return slab_ksize(folio_slab(folio)->slab_cache);
> >  }
> 
> In light of latest Linus' rant on BUG_ON() [1] I'm changing it to WARN_ON
> and return 0, as it was in v3.
> 
> [1] https://lore.kernel.org/all/CAHk-=wiEAH+ojSpAgx_Ep=NKPWHU8AdO3V56BXcCsU97oYJ1EA@mail.gmail.com/

Okay. I'm fine with that.

> 
> 
> diff --git a/mm/slab_common.c b/mm/slab_common.c
> index 98d029212682..a80c3a5e194d 100644
> --- a/mm/slab_common.c
> +++ b/mm/slab_common.c
> @@ -985,8 +985,10 @@ size_t __ksize(const void *object)
>  	folio = virt_to_folio(object);
>  
>  	if (unlikely(!folio_test_slab(folio))) {
> -		BUG_ON(folio_size(folio) <= KMALLOC_MAX_CACHE_SIZE);
> -		BUG_ON(object != folio_address(folio));
> +		if (WARN_ON(folio_size(folio) <= KMALLOC_MAX_CACHE_SIZE))
> +			return 0;
> +		if (WARN_ON(object != folio_address(folio)))
> +			return 0;
>  		return folio_size(folio);
>  	}
>  
> 

-- 
Thanks,
Hyeonggon