public inbox for netdev@vger.kernel.org 
 help / color / mirror / Atom feed
From: Ding Tianhong <dingtianhong@huawei•com>
To: David Miller <davem@davemloft•net>
Cc: <gaofeng@cn•fujitsu.com>, <yoshfuji@linux-ipv6•org>,
	<joe@perches•com>, <vfalico@redhat•com>, <netdev@vger•kernel.org>
Subject: Re: [PATCH net] net: neighbour: add neighbour dead check for neigh_timer_handler()
Date: Wed, 4 Dec 2013 14:19:06 +0800	[thread overview]
Message-ID: <529EC95A.5080908@huawei.com> (raw)
In-Reply-To: <20131203.232122.852236751455974887.davem@davemloft.net>

On 2013/12/4 12:21, David Miller wrote:
> From: Ding Tianhong <dingtianhong@huawei•com>
> Date: Wed, 4 Dec 2013 12:04:31 +0800
> 
>> The destroying neigh could be trigger by userspace, just like set the ip address which
>> in arp table to the local device ip, some I could not control it, it maybe anytime,
>> but the timer handler is execute by logic, this is normal, so I think the logic
>> is no problem, and the process of destroying neigh may conflict with the timer handler,
>> it is a synchronous problem to make sure the timer should be finished before the
>> reference neigh is freed.
> 
> The more I think about this, the more none of the explanations for this bug
> make any sense.
> 
> neigh_destroy() _ONLY_ runs when:
> 
> 	if (atomic_dec_and_test(&neigh->refcnt))
> 
> triggers in neigh_release().
> 
> This means it triggers if, and only if, neigh_refcnt goes to zero.
> 
> If the refcnt goes to zero, NO TIMER can be running.  If the timer is
> running, then there refcnt must be at least '1'.

Hi David:

Yes, you are right, but when the timer is running and prior to get the neigh->lock, the refcnt
could be dec to 0, you could not stop it by existing mechanism.

the refcnt of neighbour could only be inc by these actions:

1.create neighbour, the refcnt will be set to 1.
2.add timer, the refcnt++.
3.neigh_lookup, if found the neigh, refcnt++.

I can show the whole process of my analysis:

		CPU 0				CPU 1
		-----				-----
	create_neigh() => refcnt = 1;		
	add timer =>	refcnt++;
						<SOFTIRQ>
						base->running_timer = neigh->timer;
						neigh_timer_handler() => at this time, refcnt is 2;

user->	neigh_changeaddr()
	neigh_flush_dev();
	neigh_del_imer, refcnt dec to 1;
	release_neigh(), refcnt is 0,
	destroy_neigh()
	kfree(neighbour);
						write(neigh->lock)

So in my opinion, the point of the problem is that I should not kfree the neighbour until
the timer is not running on CPUs and not pending.

If I miss someghing, pls point out.

Regards
Ding

> 
> The only plausible theory would be that something is releasing a neigh
> too early, when references to the neigh still actually exist.
> 
> And that's a bug that should be fixed.
> 
> 

  reply	other threads:[~2013-12-04  6:19 UTC|newest]

Thread overview: 30+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-12-03 13:48 [PATCH net] net: neighbour: add neighbour dead check for neigh_timer_handler() Ding Tianhong
2013-12-03 15:03 ` Hannes Frederic Sowa
2013-12-04  1:36   ` Ding Tianhong
2013-12-03 16:28 ` Eric Dumazet
2013-12-04  1:59   ` Ding Tianhong
2013-12-03 16:37 ` David Miller
2013-12-04  2:37 ` Gao feng
2013-12-04  4:04   ` Ding Tianhong
2013-12-04  4:21     ` David Miller
2013-12-04  6:19       ` Ding Tianhong [this message]
2013-12-04  6:27         ` Eric Dumazet
2013-12-04  9:16           ` Ding Tianhong
2013-12-04 10:10             ` Gao feng
2013-12-04 15:24             ` Eric Dumazet
2013-12-05  0:32               ` Gao feng
2013-12-05  3:17                 ` Ding Tianhong
2013-12-18  6:37                   ` Ding Tianhong
2013-12-18  7:51                     ` Hannes Frederic Sowa
2013-12-18  8:19                       ` Ding Tianhong
2013-12-18  8:41                         ` Hannes Frederic Sowa
2013-12-18  8:57                           ` Ding Tianhong
2013-12-18  9:28                             ` Hannes Frederic Sowa
2013-12-18 10:02                               ` Ding Tianhong
2013-12-18 10:21                                 ` Hannes Frederic Sowa
2013-12-18 11:57                                   ` Ding Tianhong
2013-12-18 14:27                                     ` Hannes Frederic Sowa
2013-12-18 15:12                                       ` Ding Tianhong
2013-12-18 15:46                                         ` Hannes Frederic Sowa
2013-12-19  3:32                                           ` Ding Tianhong
2013-12-04  6:36         ` David Miller

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=529EC95A.5080908@huawei.com \
    --to=dingtianhong@huawei$(echo .)com \
    --cc=davem@davemloft$(echo .)net \
    --cc=gaofeng@cn$(echo .)fujitsu.com \
    --cc=joe@perches$(echo .)com \
    --cc=netdev@vger$(echo .)kernel.org \
    --cc=vfalico@redhat$(echo .)com \
    --cc=yoshfuji@linux-ipv6$(echo .)org \
    /path/to/YOUR_REPLY

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