All of lore.kernel.org
 help / color / mirror / Atom feed
From: Sven Eckelmann <sven@narfation.org>
To: Marek Lindner <mareklindner@neomailbox.ch>,
	Simon Wunderlich <sw@simonwunderlich.de>,
	Antonio Quartulli <a@unstable.cc>,
	"David S. Miller" <davem@davemloft.net>,
	Jakub Kicinski <kuba@kernel.org>,
	b.a.t.m.a.n@lists.open-mesh.org, netdev@vger.kernel.org,
	linux-kernel@vger.kernel.org, Xiyu Yang <xiyuyang19@fudan.edu.cn>
Cc: yuanxzhang@fudan.edu.cn, kjlu@umn.edu,
	Xiyu Yang <xiyuyang19@fudan.edu.cn>,
	Xin Tan <tanxin.ctf@gmail.com>
Subject: Re: [PATCH] batman-adv: Fix orig node refcnt leak when creating neigh node
Date: Fri, 18 Sep 2020 08:22:43 +0200	[thread overview]
Message-ID: <3173635.NQHa8YD4nL@ripper> (raw)
In-Reply-To: <1600398200-8198-1-git-send-email-xiyuyang19@fudan.edu.cn>

[-- Attachment #1: Type: text/plain, Size: 1416 bytes --]

On Friday, 18 September 2020 05:03:19 CEST Xiyu Yang wrote:
> batadv_neigh_node_create() is used to create a neigh node object, whose
> fields will be initialized with the specific object. When a new
> reference of the specific object is created during the initialization,
> its refcount should be increased.
> 
> However, when "neigh_node" object initializes its orig_node field with
> the "orig_node" object, the function forgets to hold the refcount of the
> "orig_node", causing a potential refcount leak and use-after-free issue
> for the reason that the object can be freed in other places.
> 
> Fix this issue by increasing the refcount of orig_node object during the
> initialization and adding corresponding batadv_orig_node_put() in
> batadv_neigh_node_release().


I will most likely not add this patch because I have concerns that this would 
need an active garbage collector to fix the reference counter loop.

Please check batadv_neigh_node::orig_node (whose reference counter you've just 
incremented) and batadv_orig_node::neigh_list (with batadv_neigh_node). And at 
the same time the batadv_neigh_node_release and batadv_orig_node_release. So 
the originator will only free the reference (and thus potentially call 
batadv_neigh_node_release) when its own reference counter is zero. But it 
cannot become zero because the neigh_node is holding a reference to this 
originator.

Kind regards,
	Sven

[-- Attachment #2: This is a digitally signed message part. --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

      reply	other threads:[~2020-09-18  6:22 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-09-18  3:03 [PATCH] batman-adv: Fix orig node refcnt leak when creating neigh node Xiyu Yang
2020-09-18  6:22 ` Sven Eckelmann [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=3173635.NQHa8YD4nL@ripper \
    --to=sven@narfation.org \
    --cc=a@unstable.cc \
    --cc=b.a.t.m.a.n@lists.open-mesh.org \
    --cc=davem@davemloft.net \
    --cc=kjlu@umn.edu \
    --cc=kuba@kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mareklindner@neomailbox.ch \
    --cc=netdev@vger.kernel.org \
    --cc=sw@simonwunderlich.de \
    --cc=tanxin.ctf@gmail.com \
    --cc=xiyuyang19@fudan.edu.cn \
    --cc=yuanxzhang@fudan.edu.cn \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.