Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Date:   Tue, 12 Jul 2022 17:13:44 +0200 (CEST)
From:   Christoph Lameter <cl@gentwo.de>
To:     Hyeonggon Yoo <42.hyeyoo@gmail.com>
cc:     Pekka Enberg <penberg@kernel.org>,
        David Rientjes <rientjes@google.com>,
        Joonsoo Kim <iamjoonsoo.kim@lge.com>,
        Andrew Morton <akpm@linux-foundation.org>,
        Vlastimil Babka <vbabka@suse.cz>,
        Roman Gushchin <roman.gushchin@linux.dev>,
        Joe Perches <joe@perches.com>,
        Vasily Averin <vasily.averin@linux.dev>,
        Matthew WilCox <willy@infradead.org>,
        linux-kernel@vger.kernel.org, linux-mm@kvack.org
Subject: Re: [PATCH 16/16] mm/sl[au]b: check if large object is valid in
 __ksize()
In-Reply-To: <20220712133946.307181-17-42.hyeyoo@gmail.com>
Message-ID: <alpine.DEB.2.22.394.2207121701070.57893@gentwo.de>
References: <20220712133946.307181-1-42.hyeyoo@gmail.com> <20220712133946.307181-17-42.hyeyoo@gmail.com>
User-Agent: Alpine 2.22 (DEB 394 2020-01-19)
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Precedence: bulk

On Tue, 12 Jul 2022, Hyeonggon Yoo wrote:

> __ksize() returns size of objects allocated from slab allocator.
> When invalid object is passed to __ksize(), returning zero
> prevents further memory corruption and makes caller be able to
> check if there is an error.
>
> If address of large object is not beginning of folio or size of
> the folio is too small, it must be invalid. Return zero in such cases.

Why return 0 if there is an error and why bother the callers with these
checks. BUG()?

> diff --git a/mm/slab_common.c b/mm/slab_common.c
> index 1f8db7959366..0d6cbe9d7ad0 100644
> --- a/mm/slab_common.c
> +++ b/mm/slab_common.c
> @@ -1013,8 +1013,12 @@ size_t __ksize(const void *object)
>
>  	folio = virt_to_folio(object);
>
> -	if (unlikely(!folio_test_slab(folio)))
> +	if (unlikely(!folio_test_slab(folio))) {
> +		if (WARN_ON(object != folio_address(folio) ||
> +				folio_size(folio) <= KMALLOC_MAX_CACHE_SIZE))

Hmmm... This may change things a bit. Before this patch it was possible to
determine the storage size of order-0 pages using ksize(). Now this
returns 0?

I guess this is an error since the order-0 page cannot come from slab
allocations.