public inbox for netdev@vger.kernel.org 
 help / color / mirror / Atom feed
From: Jiri Bohac <jbohac@suse•cz>
To: netdev@vger•kernel.org
Cc: Hideaki YOSHIFUJI <yoshfuji@linux-ipv6•org>,
	David Miller <davem@davemloft•net>
Subject: IPv6: race condition in __ipv6_ifa_notify() and dst_free() ?
Date: Tue, 20 Apr 2010 19:44:01 +0200	[thread overview]
Message-ID: <20100420174401.GB1334@midget.suse.cz> (raw)

Hi,

I found what I believe is a race condition in __ipv6_ifa_notify(), in the call
to dst_free().

__ipv6_ifa_notify() contains:

        case RTM_DELADDR:
                if (ifp->idev->cnf.forwarding)
                        addrconf_leave_anycast(ifp);
                addrconf_leave_solict(ifp->idev, &ifp->addr);
                dst_hold(&ifp->rt->u.dst);
                if (ip6_del_rt(ifp->rt))
                        dst_free(&ifp->rt->u.dst);
                break;

AFAICT, ip6_del_rt() will call dst_free() itself if it finds and actually
deletes the route: 
	ip6_del_rt() -> __ip6_del_rt() -> fib6_del() -> fib6_del_route() ->
	-> rt6_release() -> dst_free()

If it fails (like when it races with another invocation of ip6_del_rt()), it
will return nonzero and this will cause the above code to call dst_free() on its own.

dst_free() has no protection against concurrent invocation and if
two invocations make it through the "if (dst->obsolete > 1)"
check before one of them calls __dst_free(), the same dst_entry
may end up either:
	1) dst_destroy()ed and put on the dst_garbage.list, or
	2) put on the dst_garbage.list twice
both resulting in trouble once the GC is run.

One possible code path leading to two invocations of __ipv6_ifa_notify() seems
to be when two bonding slaves receive a NS/NA with the bonds IPv6 address when
the bonding master is in the DAD phase with a tentative address:

netif_receive_skb() gets invoked on two CPUs and sets skb->dev to the bonding master ...
... ip6_mc_input() -> ip6_input_finish() -> icmpv6_rcv() -> ndisc_rcv() ->
 -> ndisc_recv_ns() -> addrconf_dad_failure() -> ipv6_del_addr() -> ipv6_ifa_notify() ->
 -> __ipv6_ifa_notify


What is the reason __ipv6_ifa_notify() calls dst_free() when
ip6_del_rt() fails? I don't see a way ip6_del_rt() could fail
with the dst still needing to be freed.

I am just testing whether the following will help:

--- a/net/ipv6/addrconf.c	2010-04-17 00:12:32.000000000 +0200
+++ b/net/ipv6/addrconf.c	2010-04-20 19:07:35.000000000 +0200
@@ -3974,8 +3974,7 @@ static void __ipv6_ifa_notify(int event,
 			addrconf_leave_anycast(ifp);
 		addrconf_leave_solict(ifp->idev, &ifp->addr);
 		dst_hold(&ifp->rt->u.dst);
-		if (ip6_del_rt(ifp->rt))
-			dst_free(&ifp->rt->u.dst);
+		ip6_del_rt(ifp->rt);
 		break;
 	}
 }

Thanks,

-- 
Jiri Bohac <jbohac@suse•cz>
SUSE Labs, SUSE CZ


             reply	other threads:[~2010-04-20 17:44 UTC|newest]

Thread overview: 28+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2010-04-20 17:44 Jiri Bohac [this message]
2010-04-20 17:57 ` IPv6: race condition in __ipv6_ifa_notify() and dst_free() ? Eric Dumazet
2010-04-20 20:49   ` Jiri Bohac
2010-04-20 20:57     ` Eric Dumazet
2010-04-20 21:16       ` Stephen Hemminger
2010-04-20 21:35         ` Jiri Bohac
2010-04-21 21:34 ` Jiri Bohac
2010-04-22  2:32   ` Herbert Xu
2010-04-22  7:43     ` David Miller
2010-04-22 14:25       ` Herbert Xu
2010-04-22 15:49         ` Jiri Bohac
2010-04-22 16:17           ` Stephen Hemminger
2010-04-23  1:54           ` David Miller
2010-04-23  2:10             ` Herbert Xu
2010-04-23 15:05               ` Herbert Xu
2010-05-18 11:02                 ` [0/4] Fix addrconf race conditions Herbert Xu
2010-05-18 11:04                   ` [PATCH 1/4] ipv6: Replace inet6_ifaddr->dead with state Herbert Xu
2010-05-18 17:23                     ` Stephen Hemminger
2010-05-18 22:27                       ` David Miller
2010-05-18 11:04                   ` [PATCH 2/4] ipv6: Use state_lock to protect ifa state Herbert Xu
2010-05-18 11:04                   ` [PATCH 3/4] ipv6: Use POSTDAD state Herbert Xu
2010-05-18 11:04                   ` [PATCH 4/4] ipv6: Never schedule DAD timer on dead address Herbert Xu
2010-05-18 17:25                   ` [0/4] Fix addrconf race conditions Stephen Hemminger
2010-05-18 22:27                     ` David Miller
2010-05-18 22:35                       ` Herbert Xu
2010-04-27 15:50             ` IPv6: race condition in __ipv6_ifa_notify() and dst_free() ? Jiri Bohac
2010-04-27 15:55               ` Herbert Xu
2010-05-09  6:48                 ` Herbert Xu

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=20100420174401.GB1334@midget.suse.cz \
    --to=jbohac@suse$(echo .)cz \
    --cc=davem@davemloft$(echo .)net \
    --cc=netdev@vger$(echo .)kernel.org \
    --cc=yoshfuji@linux-ipv6$(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