From: Patrick McHardy <kaber@trash.net>
To: Eric Dumazet <eric.dumazet@gmail.com>
Cc: Ingo Molnar <mingo@elte.hu>, David Miller <davem@davemloft.net>,
Thomas Gleixner <tglx@linutronix.de>,
torvalds@linux-foundation.org, akpm@linux-foundation.org,
netdev@vger.kernel.org, linux-kernel@vger.kernel.org
Subject: Re: [bug] __nf_ct_refresh_acct(): WARNING: at lib/list_debug.c:30 __list_add+0x7d/0xad()
Date: Wed, 17 Jun 2009 13:51:45 +0200 [thread overview]
Message-ID: <4A38D8D1.6060004@trash.net> (raw)
In-Reply-To: <4A38D5BD.2040502@trash.net>
[-- Attachment #1: Type: text/plain, Size: 846 bytes --]
Patrick McHardy wrote:
> Eric Dumazet wrote:
>> IPS_CONFIRMED_BIT is set under nf_conntrack_lock (in
>> __nf_conntrack_confirm()),
>> we probably want to add a synchronisation under ct->lock as well,
>> or __nf_ct_refresh_acct() could set ct->timeout.expires to extra_jiffies,
>> while a different cpu could confirm the conntrack.
>
> Before the conntrack is confirmed, it is exclusively handled by a
> single CPU. I agree that we need to make sure the IPS_CONFIRMED_BIT
> is visible before we add the conntrack to the hash table since the
> lookup is lockless, but simply moving the set_bit before the hash
> insertion should be fine I think.
>
A slightly changed version which moves hash insertion to the end
and adds a comment about ordering. This make sure the timer is
actually running before the conntrack can be found be other CPUs.
[-- Attachment #2: x --]
[-- Type: text/plain, Size: 1174 bytes --]
diff --git a/net/netfilter/nf_conntrack_core.c b/net/netfilter/nf_conntrack_core.c
index 5f72b94..e2cc707 100644
--- a/net/netfilter/nf_conntrack_core.c
+++ b/net/netfilter/nf_conntrack_core.c
@@ -425,7 +425,6 @@ __nf_conntrack_confirm(struct sk_buff *skb)
/* Remove from unconfirmed list */
hlist_nulls_del_rcu(&ct->tuplehash[IP_CT_DIR_ORIGINAL].hnnode);
- __nf_conntrack_hash_insert(ct, hash, repl_hash);
/* Timer relative to confirmation time, not original
setting time, otherwise we'd get timer wrap in
weird delay cases. */
@@ -433,8 +432,15 @@ __nf_conntrack_confirm(struct sk_buff *skb)
add_timer(&ct->timeout);
atomic_inc(&ct->ct_general.use);
set_bit(IPS_CONFIRMED_BIT, &ct->status);
+
+ /* Since the lookup is lockless, hash insertion must be after starting the
+ * timer and setting the CONFIRMED bit. The RCU barriers guarantee that no
+ * other CPU can find the conntrack before the above stores are visible.
+ */
+ __nf_conntrack_hash_insert(ct, hash, repl_hash);
NF_CT_STAT_INC(net, insert);
spin_unlock_bh(&nf_conntrack_lock);
+
help = nfct_help(ct);
if (help && help->helper)
nf_conntrack_event_cache(IPCT_HELPER, ct);
next prev parent reply other threads:[~2009-06-17 11:51 UTC|newest]
Thread overview: 55+ messages / expand[flat|nested] mbox.gz Atom feed top
2009-06-15 12:04 [GIT]: Networking David Miller
2009-06-16 9:15 ` Ingo Molnar
2009-06-16 9:19 ` David Miller
2009-06-16 9:33 ` Ingo Molnar
2009-06-16 9:44 ` Ingo Molnar
2009-06-16 9:44 ` Alan Cox
2009-06-16 9:48 ` Ingo Molnar
2009-06-16 9:56 ` David Miller
2009-06-16 10:11 ` Ingo Molnar
2009-06-16 10:35 ` David Miller
2009-06-16 13:38 ` Eric Dumazet
2009-06-16 14:19 ` Eric Dumazet
2009-06-16 18:59 ` Linus Torvalds
2009-06-16 19:08 ` David Miller
2009-06-16 19:37 ` Eric Dumazet
2009-06-16 20:12 ` Eric Dumazet
2009-06-17 6:41 ` [PATCH] net: correct off-by-one write allocations reports Eric Dumazet
2009-06-17 11:31 ` [PATCH] atm: sk_wmem_alloc initial value is one Eric Dumazet
2009-06-18 2:06 ` David Miller
2009-06-18 2:05 ` [PATCH] net: correct off-by-one write allocations reports David Miller
2009-06-17 11:32 ` [GIT]: Networking David Miller
2009-06-16 9:21 ` David Miller
2009-06-16 9:26 ` Ingo Molnar
2009-06-16 10:47 ` David Miller
2009-06-16 10:53 ` Ingo Molnar
2009-06-16 12:24 ` Ingo Molnar
2009-06-16 12:39 ` David Miller
2009-06-17 9:21 ` [bug] __nf_ct_refresh_acct(): WARNING: at lib/list_debug.c:30 __list_add+0x7d/0xad() Ingo Molnar
2009-06-17 10:18 ` Eric Dumazet
2009-06-17 11:08 ` Ingo Molnar
2009-06-17 11:35 ` David Miller
2009-06-18 5:23 ` Ingo Molnar
2009-06-18 5:58 ` Eric Dumazet
2009-06-18 9:47 ` Patrick McHardy
2009-06-18 14:56 ` Patrick McHardy
2009-06-18 15:46 ` Eric Dumazet
2009-06-18 16:09 ` Patrick McHardy
2009-06-18 16:13 ` Pablo Neira Ayuso
2009-06-18 22:46 ` [PATCH] netfilter: conntrack: death_by_timeout() fix Eric Dumazet
2009-06-19 11:15 ` Patrick McHardy
2009-06-20 15:47 ` [bug] __nf_ct_refresh_acct(): WARNING: at lib/list_debug.c:30 __list_add+0x7d/0xad() Ingo Molnar
2009-06-20 15:56 ` Patrick McHardy
2009-06-17 11:38 ` Patrick McHardy
2009-06-17 11:51 ` Patrick McHardy [this message]
2009-06-17 11:55 ` Eric Dumazet
2009-06-17 12:00 ` Patrick McHardy
2009-06-17 12:33 ` Eric Dumazet
2009-06-17 12:36 ` Patrick McHardy
2009-06-17 13:27 ` Eric Dumazet
2009-06-17 13:29 ` Patrick McHardy
2009-06-17 14:23 ` Patrick McHardy
2009-06-17 15:29 ` Eric Dumazet
2009-06-17 15:34 ` Patrick McHardy
2009-06-18 23:18 ` [GIT]: Networking Tilman Schmidt
2009-06-19 3:36 ` 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=4A38D8D1.6060004@trash.net \
--to=kaber@trash.net \
--cc=akpm@linux-foundation.org \
--cc=davem@davemloft.net \
--cc=eric.dumazet@gmail.com \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@elte.hu \
--cc=netdev@vger.kernel.org \
--cc=tglx@linutronix.de \
--cc=torvalds@linux-foundation.org \
/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 NNTP newsgroup(s).