From mboxrd@z Thu Jan 1 00:00:00 1970 From: David Miller Subject: Re: [Bugme-new] [Bug 27212] New: Warning kmemcheck: Caught 64-bit read from uninitialized memory in netlink_broadcast_filtered Date: Mon, 14 Feb 2011 11:43:18 -0800 (PST) Message-ID: <20110214.114318.112601812.davem@davemloft.net> References: <1295556085.2613.22.camel@edumazet-laptop> <4D393A99.9060104@kernel.org> <1297704922.2996.60.camel@edumazet-laptop> Mime-Version: 1.0 Content-Type: Text/Plain; charset=iso-8859-1 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: penberg@kernel.org, akpm@linux-foundation.org, netdev@vger.kernel.org, bugzilla-daemon@bugzilla.kernel.org, bugme-daemon@bugzilla.kernel.org, casteyde.christian@free.fr, xiaosuo@gmail.com, vegardno@ifi.uio.no, linux-kernel@vger.kernel.org To: eric.dumazet@gmail.com Return-path: In-Reply-To: <1297704922.2996.60.camel@edumazet-laptop> Sender: linux-kernel-owner@vger.kernel.org List-Id: netdev.vger.kernel.org =46rom: Eric Dumazet Date: Mon, 14 Feb 2011 18:35:22 +0100 > Le vendredi 21 janvier 2011 =E0 09:49 +0200, Pekka Enberg a =E9crit : >=20 >> It actually looks like a bug in SLUB+kmemcheck. The=20 >> kmemcheck_slab_alloc() call in slab_post_alloc_hook() should use ksi= ze()=20 >> instead of s->objsize. SLAB seems to do the right thing already. Any= one=20 >> care to send a patch my way? >>=20 >=20 > Hmm, what do you think of following patch ? >=20 > Thanks, and sorry for the delay. >=20 > [PATCH] slub: fix kmemcheck calls to match ksize() hints >=20 > Recent use of ksize() in network stack (commit ca44ac38 : net: don't > reallocate skb->head unless the current one hasn't the needed extra s= ize > or is shared) triggers kmemcheck warnings, because ksize() can return > more space than kmemcheck is aware of. >=20 > Pekka Enberg noticed SLAB+kmemcheck is doing the right thing, while S= LUB > +kmemcheck doesnt. >=20 > Bugzilla reference #27212 >=20 > Reported-by: Christian Casteyde > Suggested-by: Pekka Enberg > Signed-off-by: Eric Dumazet Acked-by: David S. Miller