From: duoming@zju.edu.cn
To: Jakub Kicinski <kuba@kernel.org>
Cc: Thomas Osterried <thomas@osterried.de>,
"David S. Miller" <davem@davemloft.net>,
Paolo Abeni <pabeni@redhat.com>,
linux-hams@vger.kernel.org
Subject: Re: [PATCH net] ax25: Fix ax25 session cleanup problem in ax25_release
Date: Sat, 28 May 2022 21:29:20 +0800 (GMT+08:00) [thread overview]
Message-ID: <20a99b3a.3a4c0.1810adb6a1a.Coremail.duoming@zju.edu.cn> (raw)
In-Reply-To: <20220527173312.71122dbe@kernel.org>
Hello,
On Fri, 27 May 2022 13:29:30 +0200 Thomas Osterried wrote:
> > > I Tested several cases: this patch works as expected.
> > If you agree, that no concurrent process is able to re-use this ax25_cb,
> > and because all timers are stoppeed, the cleanup with ax25_cb_del(s);
> > should be safe.
> >
> >
> > My successfull test was this:
> >
> > diff --git a/net/ax25/af_ax25.c b/net/ax25/af_ax25.c
> > index 363d47f94532..de417b974c07 100644
> > --- a/net/ax25/af_ax25.c
> > +++ b/net/ax25/af_ax25.c
> > @@ -91,6 +92,7 @@ static void ax25_kill_by_device(struct net_device *dev)
> > spin_unlock_bh(&ax25_list_lock);
> > ax25_disconnect(s, ENETUNREACH);
> > s->ax25_dev = NULL;
> > + ax25_cb_del(s);
> > spin_lock_bh(&ax25_list_lock);
> > goto again;
> > }
The above code also exists a race condition:
(Thread 1) | (Thread 2)
ax25_kill_by_device() | ax25_rcv()
... | ...
if (!sk) { |
ax25_cb_del() | ax25_destroy_socket()
if (!hlist_unhashed(..)) | ax25_cb_del()
... | if (!hlist_unhashed(..))
hlist_del_init() |
ax25_cb_put(ax25) | ...
| ax25_cb_put(ax25)
| ...
If we use ax25_cb_del() to delete ax25 node from the hlist, the ax25_cb_put() should only be
called once. But the above picture shows that ax25_cb_put() is called twice and the refcount
of the same ax25_cb is decreased twice. As a result, it will cause "refcount_t: underflow".
The following is a solution that has been tested.
@@ -62,12 +62,12 @@ static void ax25_free_sock(struct sock *sk)
*/
static void ax25_cb_del(ax25_cb *ax25)
{
+ spin_lock_bh(&ax25_list_lock);
if (!hlist_unhashed(&ax25->ax25_node)) {
- spin_lock_bh(&ax25_list_lock);
hlist_del_init(&ax25->ax25_node);
- spin_unlock_bh(&ax25_list_lock);
ax25_cb_put(ax25);
}
+ spin_unlock_bh(&ax25_list_lock);
}
Best regards,
Duoming Zhou
prev parent reply other threads:[~2022-05-28 13:29 UTC|newest]
Thread overview: 6+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-05-25 11:28 [PATCH net] ax25: Fix ax25 session cleanup problem in ax25_release Duoming Zhou
2022-05-27 8:11 ` Thomas Osterried
2022-05-27 11:29 ` Thomas Osterried
2022-05-27 14:23 ` duoming
2022-05-28 0:33 ` Jakub Kicinski
2022-05-28 13:29 ` duoming [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=20a99b3a.3a4c0.1810adb6a1a.Coremail.duoming@zju.edu.cn \
--to=duoming@zju.edu.cn \
--cc=davem@davemloft.net \
--cc=kuba@kernel.org \
--cc=linux-hams@vger.kernel.org \
--cc=pabeni@redhat.com \
--cc=thomas@osterried.de \
/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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).