From: Patrick McHardy <kaber@trash.net>
To: "David S. Miller" <davem@davemloft.net>
Cc: David Stevens <dlstevens@us.ibm.com>,
davem@redhat.com, laforge@netfilter.org, netdev@oss.sgi.com,
netdev-bounce@oss.sgi.com, netfilter-devel@lists.netfilter.org,
okir@suse.de
Subject: Re: [PATCH] Prevent crash on ip_conntrack removal
Date: Sun, 29 Aug 2004 21:36:28 +0200 [thread overview]
Message-ID: <4132303C.2060807@trash.net> (raw)
In-Reply-To: <20040828231529.051a73cc.davem@davemloft.net>
[-- Attachment #1: Type: text/plain, Size: 424 bytes --]
David S. Miller wrote:
>I think I agree with David now that I've read through this a few
>times. Can someone send me a patch which does this?
>
>
Attached. The first patch still crashed, we need to prevent new
fragments from getting queued after the queue is flushed until the
hook in unregistered.
>Does 2.4.x have this problem too? I thought it didn't.
>
>
I'll have a look, but I think it does.
Regards
Patrick
[-- Attachment #2: x --]
[-- Type: text/plain, Size: 4121 bytes --]
# This is a BitKeeper generated diff -Nru style patch.
#
# ChangeSet
# 2004/08/29 20:24:07+02:00 kaber@coreworks.de
# [NETFILTER]: Flush fragment queue on conntrack unload
#
# Signed-off-by: Patrick McHardy <kaber@trash.net>
#
# net/ipv4/netfilter/ip_conntrack_standalone.c
# 2004/08/29 20:22:57+02:00 kaber@coreworks.de +6 -0
# [NETFILTER]: Flush fragment queue on conntrack unload
#
# net/ipv4/netfilter/ip_conntrack_core.c
# 2004/08/29 20:22:57+02:00 kaber@coreworks.de +8 -0
# [NETFILTER]: Flush fragment queue on conntrack unload
#
# net/ipv4/ip_fragment.c
# 2004/08/29 20:22:57+02:00 kaber@coreworks.de +14 -3
# [NETFILTER]: Flush fragment queue on conntrack unload
#
# include/net/ip.h
# 2004/08/29 20:22:57+02:00 kaber@coreworks.de +1 -0
# [NETFILTER]: Flush fragment queue on conntrack unload
#
# include/linux/netfilter_ipv4/ip_conntrack.h
# 2004/08/29 20:22:57+02:00 kaber@coreworks.de +1 -0
# [NETFILTER]: Flush fragment queue on conntrack unload
#
diff -Nru a/include/linux/netfilter_ipv4/ip_conntrack.h b/include/linux/netfilter_ipv4/ip_conntrack.h
--- a/include/linux/netfilter_ipv4/ip_conntrack.h 2004-08-29 20:55:13 +02:00
+++ b/include/linux/netfilter_ipv4/ip_conntrack.h 2004-08-29 20:55:13 +02:00
@@ -275,6 +275,7 @@
/* Fake conntrack entry for untracked connections */
extern struct ip_conntrack ip_conntrack_untracked;
+extern int ip_ct_no_defrag;
/* Returns new sk_buff, or NULL */
struct sk_buff *
ip_ct_gather_frags(struct sk_buff *skb);
diff -Nru a/include/net/ip.h b/include/net/ip.h
--- a/include/net/ip.h 2004-08-29 20:55:13 +02:00
+++ b/include/net/ip.h 2004-08-29 20:55:13 +02:00
@@ -255,6 +255,7 @@
*/
struct sk_buff *ip_defrag(struct sk_buff *skb);
+extern void ipfrag_flush(void);
extern int ip_frag_nqueues;
extern atomic_t ip_frag_mem;
diff -Nru a/net/ipv4/ip_fragment.c b/net/ipv4/ip_fragment.c
--- a/net/ipv4/ip_fragment.c 2004-08-29 20:55:13 +02:00
+++ b/net/ipv4/ip_fragment.c 2004-08-29 20:55:13 +02:00
@@ -241,15 +241,15 @@
}
/* Memory limiting on fragments. Evictor trashes the oldest
- * fragment queue until we are back under the low threshold.
+ * fragment queue until we are back under the threshold.
*/
-static void ip_evictor(void)
+static void __ip_evictor(int threshold)
{
struct ipq *qp;
struct list_head *tmp;
int work;
- work = atomic_read(&ip_frag_mem) - sysctl_ipfrag_low_thresh;
+ work = atomic_read(&ip_frag_mem) - threshold;
if (work <= 0)
return;
@@ -274,6 +274,11 @@
}
}
+static inline void ip_evictor(void)
+{
+ __ip_evictor(sysctl_ipfrag_low_thresh);
+}
+
/*
* Oops, a fragment queue timed out. Kill it and send an ICMP reply.
*/
@@ -684,4 +689,10 @@
add_timer(&ipfrag_secret_timer);
}
+void ipfrag_flush(void)
+{
+ __ip_evictor(0);
+}
+
EXPORT_SYMBOL(ip_defrag);
+EXPORT_SYMBOL(ipfrag_flush);
diff -Nru a/net/ipv4/netfilter/ip_conntrack_core.c b/net/ipv4/netfilter/ip_conntrack_core.c
--- a/net/ipv4/netfilter/ip_conntrack_core.c 2004-08-29 20:55:13 +02:00
+++ b/net/ipv4/netfilter/ip_conntrack_core.c 2004-08-29 20:55:13 +02:00
@@ -1173,6 +1173,8 @@
}
}
+int ip_ct_no_defrag;
+
/* Returns new sk_buff, or NULL */
struct sk_buff *
ip_ct_gather_frags(struct sk_buff *skb)
@@ -1181,6 +1183,12 @@
#ifdef CONFIG_NETFILTER_DEBUG
unsigned int olddebug = skb->nf_debug;
#endif
+
+ if (unlikely(ip_ct_no_defrag)) {
+ kfree_skb(skb);
+ return NULL;
+ }
+
if (sk) {
sock_hold(sk);
skb_orphan(skb);
diff -Nru a/net/ipv4/netfilter/ip_conntrack_standalone.c b/net/ipv4/netfilter/ip_conntrack_standalone.c
--- a/net/ipv4/netfilter/ip_conntrack_standalone.c 2004-08-29 20:55:13 +02:00
+++ b/net/ipv4/netfilter/ip_conntrack_standalone.c 2004-08-29 20:55:13 +02:00
@@ -805,6 +805,12 @@
cleanup_defraglocalops:
nf_unregister_hook(&ip_conntrack_defrag_local_out_ops);
cleanup_defragops:
+ /* Frag queues may hold fragments with skb->dst == NULL */
+ ip_ct_no_defrag = 1;
+ smp_wmb();
+ local_bh_disable();
+ ipfrag_flush();
+ local_bh_enable();
nf_unregister_hook(&ip_conntrack_defrag_ops);
cleanup_proc_stat:
proc_net_remove("ip_conntrack_stat");
next prev parent reply other threads:[~2004-08-29 19:36 UTC|newest]
Thread overview: 23+ messages / expand[flat|nested] mbox.gz Atom feed top
2004-08-18 9:13 [PATCH] Prevent crash on ip_conntrack removal Olaf Kirch
2004-08-19 10:11 ` Harald Welte
2004-08-19 14:18 ` David S. Miller
2004-08-19 14:55 ` Patrick McHardy
2004-08-19 15:14 ` David S. Miller
2004-08-21 15:10 ` Patrick McHardy
2004-08-22 5:13 ` David S. Miller
2004-08-22 12:58 ` Patrick McHardy
2004-08-23 5:03 ` David S. Miller
2004-08-23 21:18 ` David Stevens
2004-08-24 0:45 ` Nivedita Singhvi
2004-08-24 0:45 ` Patrick McHardy
2004-08-24 21:28 ` David Stevens
2004-08-29 6:15 ` David S. Miller
2004-08-29 19:36 ` Patrick McHardy [this message]
2004-08-29 19:57 ` David S. Miller
2004-08-29 20:06 ` Patrick McHardy
2004-08-29 21:58 ` Patrick McHardy
2004-08-29 23:38 ` David S. Miller
2004-08-30 0:50 ` Patrick McHardy
2004-08-30 4:28 ` David S. Miller
2004-08-29 21:48 ` Patrick McHardy
2004-08-30 7:57 ` Olaf Kirch
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=4132303C.2060807@trash.net \
--to=kaber@trash.net \
--cc=davem@davemloft.net \
--cc=davem@redhat.com \
--cc=dlstevens@us.ibm.com \
--cc=laforge@netfilter.org \
--cc=netdev-bounce@oss.sgi.com \
--cc=netdev@oss.sgi.com \
--cc=netfilter-devel@lists.netfilter.org \
--cc=okir@suse.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 NNTP newsgroup(s).