2006-11-29 01:35:01 +00:00
|
|
|
/* Event cache for netfilter. */
|
|
|
|
|
|
|
|
/* (C) 1999-2001 Paul `Rusty' Russell
|
|
|
|
* (C) 2002-2006 Netfilter Core Team <coreteam@netfilter.org>
|
|
|
|
* (C) 2003,2004 USAGI/WIDE Project <http://www.linux-ipv6.org>
|
|
|
|
*
|
|
|
|
* This program is free software; you can redistribute it and/or modify
|
|
|
|
* it under the terms of the GNU General Public License version 2 as
|
|
|
|
* published by the Free Software Foundation.
|
|
|
|
*/
|
|
|
|
|
|
|
|
#include <linux/types.h>
|
|
|
|
#include <linux/netfilter.h>
|
|
|
|
#include <linux/skbuff.h>
|
|
|
|
#include <linux/vmalloc.h>
|
|
|
|
#include <linux/stddef.h>
|
|
|
|
#include <linux/err.h>
|
|
|
|
#include <linux/percpu.h>
|
|
|
|
#include <linux/kernel.h>
|
|
|
|
#include <linux/netdevice.h>
|
include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h
percpu.h is included by sched.h and module.h and thus ends up being
included when building most .c files. percpu.h includes slab.h which
in turn includes gfp.h making everything defined by the two files
universally available and complicating inclusion dependencies.
percpu.h -> slab.h dependency is about to be removed. Prepare for
this change by updating users of gfp and slab facilities include those
headers directly instead of assuming availability. As this conversion
needs to touch large number of source files, the following script is
used as the basis of conversion.
http://userweb.kernel.org/~tj/misc/slabh-sweep.py
The script does the followings.
* Scan files for gfp and slab usages and update includes such that
only the necessary includes are there. ie. if only gfp is used,
gfp.h, if slab is used, slab.h.
* When the script inserts a new include, it looks at the include
blocks and try to put the new include such that its order conforms
to its surrounding. It's put in the include block which contains
core kernel includes, in the same order that the rest are ordered -
alphabetical, Christmas tree, rev-Xmas-tree or at the end if there
doesn't seem to be any matching order.
* If the script can't find a place to put a new include (mostly
because the file doesn't have fitting include block), it prints out
an error message indicating which .h file needs to be added to the
file.
The conversion was done in the following steps.
1. The initial automatic conversion of all .c files updated slightly
over 4000 files, deleting around 700 includes and adding ~480 gfp.h
and ~3000 slab.h inclusions. The script emitted errors for ~400
files.
2. Each error was manually checked. Some didn't need the inclusion,
some needed manual addition while adding it to implementation .h or
embedding .c file was more appropriate for others. This step added
inclusions to around 150 files.
3. The script was run again and the output was compared to the edits
from #2 to make sure no file was left behind.
4. Several build tests were done and a couple of problems were fixed.
e.g. lib/decompress_*.c used malloc/free() wrappers around slab
APIs requiring slab.h to be added manually.
5. The script was run on all .h files but without automatically
editing them as sprinkling gfp.h and slab.h inclusions around .h
files could easily lead to inclusion dependency hell. Most gfp.h
inclusion directives were ignored as stuff from gfp.h was usually
wildly available and often used in preprocessor macros. Each
slab.h inclusion directive was examined and added manually as
necessary.
6. percpu.h was updated not to include slab.h.
7. Build test were done on the following configurations and failures
were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my
distributed build env didn't work with gcov compiles) and a few
more options had to be turned off depending on archs to make things
build (like ipr on powerpc/64 which failed due to missing writeq).
* x86 and x86_64 UP and SMP allmodconfig and a custom test config.
* powerpc and powerpc64 SMP allmodconfig
* sparc and sparc64 SMP allmodconfig
* ia64 SMP allmodconfig
* s390 SMP allmodconfig
* alpha SMP allmodconfig
* um on x86_64 SMP allmodconfig
8. percpu.h modifications were reverted so that it could be applied as
a separate patch and serve as bisection point.
Given the fact that I had only a couple of failures from tests on step
6, I'm fairly confident about the coverage of this conversion patch.
If there is a breakage, it's likely to be something in one of the arch
headers which should be easily discoverable easily on most builds of
the specific arch.
Signed-off-by: Tejun Heo <tj@kernel.org>
Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
2010-03-24 08:04:11 +00:00
|
|
|
#include <linux/slab.h>
|
2011-07-15 15:47:34 +00:00
|
|
|
#include <linux/export.h>
|
2006-11-29 01:35:01 +00:00
|
|
|
|
|
|
|
#include <net/netfilter/nf_conntrack.h>
|
|
|
|
#include <net/netfilter/nf_conntrack_core.h>
|
2009-06-13 10:26:29 +00:00
|
|
|
#include <net/netfilter/nf_conntrack_extend.h>
|
2006-11-29 01:35:01 +00:00
|
|
|
|
2009-06-03 08:32:06 +00:00
|
|
|
static DEFINE_MUTEX(nf_ct_ecache_mutex);
|
2006-12-03 06:11:25 +00:00
|
|
|
|
2006-11-29 01:35:01 +00:00
|
|
|
/* deliver cached events and clear cache entry - must be called with locally
|
|
|
|
* disabled softirqs */
|
2009-06-13 10:26:29 +00:00
|
|
|
void nf_ct_deliver_cached_events(struct nf_conn *ct)
|
2006-11-29 01:35:01 +00:00
|
|
|
{
|
2011-11-21 23:16:51 +00:00
|
|
|
struct net *net = nf_ct_net(ct);
|
2012-02-22 06:48:01 +00:00
|
|
|
unsigned long events, missed;
|
2009-06-03 08:32:06 +00:00
|
|
|
struct nf_ct_event_notifier *notify;
|
2009-06-13 10:26:29 +00:00
|
|
|
struct nf_conntrack_ecache *e;
|
2012-02-22 06:48:01 +00:00
|
|
|
struct nf_ct_event item;
|
|
|
|
int ret;
|
2009-06-03 08:32:06 +00:00
|
|
|
|
|
|
|
rcu_read_lock();
|
2011-11-21 23:16:51 +00:00
|
|
|
notify = rcu_dereference(net->ct.nf_conntrack_event_cb);
|
2009-06-03 08:32:06 +00:00
|
|
|
if (notify == NULL)
|
|
|
|
goto out_unlock;
|
|
|
|
|
2009-06-13 10:26:29 +00:00
|
|
|
e = nf_ct_ecache_find(ct);
|
|
|
|
if (e == NULL)
|
|
|
|
goto out_unlock;
|
|
|
|
|
|
|
|
events = xchg(&e->cache, 0);
|
|
|
|
|
2012-02-22 06:48:01 +00:00
|
|
|
if (!nf_ct_is_confirmed(ct) || nf_ct_is_dying(ct) || !events)
|
|
|
|
goto out_unlock;
|
|
|
|
|
|
|
|
/* We make a copy of the missed event cache without taking
|
|
|
|
* the lock, thus we may send missed events twice. However,
|
|
|
|
* this does not harm and it happens very rarely. */
|
|
|
|
missed = e->missed;
|
|
|
|
|
|
|
|
if (!((events | missed) & e->ctmask))
|
|
|
|
goto out_unlock;
|
|
|
|
|
|
|
|
item.ct = ct;
|
|
|
|
item.pid = 0;
|
|
|
|
item.report = 0;
|
|
|
|
|
|
|
|
ret = notify->fcn(events | missed, &item);
|
|
|
|
|
|
|
|
if (likely(ret >= 0 && !missed))
|
|
|
|
goto out_unlock;
|
|
|
|
|
|
|
|
spin_lock_bh(&ct->lock);
|
|
|
|
if (ret < 0)
|
|
|
|
e->missed |= events;
|
|
|
|
else
|
|
|
|
e->missed &= ~missed;
|
|
|
|
spin_unlock_bh(&ct->lock);
|
2006-11-29 01:35:01 +00:00
|
|
|
|
2009-06-03 08:32:06 +00:00
|
|
|
out_unlock:
|
|
|
|
rcu_read_unlock();
|
2006-11-29 01:35:01 +00:00
|
|
|
}
|
2006-12-03 06:11:25 +00:00
|
|
|
EXPORT_SYMBOL_GPL(nf_ct_deliver_cached_events);
|
2006-11-29 01:35:01 +00:00
|
|
|
|
2011-11-21 23:16:51 +00:00
|
|
|
int nf_conntrack_register_notifier(struct net *net,
|
|
|
|
struct nf_ct_event_notifier *new)
|
2007-03-14 23:40:10 +00:00
|
|
|
{
|
2009-06-03 08:32:06 +00:00
|
|
|
int ret = 0;
|
2010-05-10 16:47:57 +00:00
|
|
|
struct nf_ct_event_notifier *notify;
|
2009-06-03 08:32:06 +00:00
|
|
|
|
|
|
|
mutex_lock(&nf_ct_ecache_mutex);
|
2011-11-21 23:16:51 +00:00
|
|
|
notify = rcu_dereference_protected(net->ct.nf_conntrack_event_cb,
|
2010-05-10 16:47:57 +00:00
|
|
|
lockdep_is_held(&nf_ct_ecache_mutex));
|
|
|
|
if (notify != NULL) {
|
2009-06-03 08:32:06 +00:00
|
|
|
ret = -EBUSY;
|
|
|
|
goto out_unlock;
|
|
|
|
}
|
2012-01-12 04:41:32 +00:00
|
|
|
rcu_assign_pointer(net->ct.nf_conntrack_event_cb, new);
|
2009-06-03 08:32:06 +00:00
|
|
|
mutex_unlock(&nf_ct_ecache_mutex);
|
|
|
|
return ret;
|
|
|
|
|
|
|
|
out_unlock:
|
|
|
|
mutex_unlock(&nf_ct_ecache_mutex);
|
|
|
|
return ret;
|
2007-03-14 23:40:10 +00:00
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(nf_conntrack_register_notifier);
|
|
|
|
|
2011-11-21 23:16:51 +00:00
|
|
|
void nf_conntrack_unregister_notifier(struct net *net,
|
|
|
|
struct nf_ct_event_notifier *new)
|
2007-03-14 23:40:10 +00:00
|
|
|
{
|
2010-05-10 16:47:57 +00:00
|
|
|
struct nf_ct_event_notifier *notify;
|
|
|
|
|
2009-06-03 08:32:06 +00:00
|
|
|
mutex_lock(&nf_ct_ecache_mutex);
|
2011-11-21 23:16:51 +00:00
|
|
|
notify = rcu_dereference_protected(net->ct.nf_conntrack_event_cb,
|
2010-05-10 16:47:57 +00:00
|
|
|
lockdep_is_held(&nf_ct_ecache_mutex));
|
|
|
|
BUG_ON(notify != new);
|
2011-11-21 23:16:51 +00:00
|
|
|
RCU_INIT_POINTER(net->ct.nf_conntrack_event_cb, NULL);
|
2009-06-03 08:32:06 +00:00
|
|
|
mutex_unlock(&nf_ct_ecache_mutex);
|
2007-03-14 23:40:10 +00:00
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(nf_conntrack_unregister_notifier);
|
|
|
|
|
2011-11-21 23:16:51 +00:00
|
|
|
int nf_ct_expect_register_notifier(struct net *net,
|
|
|
|
struct nf_exp_event_notifier *new)
|
2007-03-14 23:40:10 +00:00
|
|
|
{
|
2009-06-03 08:32:06 +00:00
|
|
|
int ret = 0;
|
2010-05-10 16:47:57 +00:00
|
|
|
struct nf_exp_event_notifier *notify;
|
2009-06-03 08:32:06 +00:00
|
|
|
|
|
|
|
mutex_lock(&nf_ct_ecache_mutex);
|
2011-11-21 23:16:51 +00:00
|
|
|
notify = rcu_dereference_protected(net->ct.nf_expect_event_cb,
|
2010-05-10 16:47:57 +00:00
|
|
|
lockdep_is_held(&nf_ct_ecache_mutex));
|
|
|
|
if (notify != NULL) {
|
2009-06-03 08:32:06 +00:00
|
|
|
ret = -EBUSY;
|
|
|
|
goto out_unlock;
|
|
|
|
}
|
2012-01-12 04:41:32 +00:00
|
|
|
rcu_assign_pointer(net->ct.nf_expect_event_cb, new);
|
2009-06-03 08:32:06 +00:00
|
|
|
mutex_unlock(&nf_ct_ecache_mutex);
|
|
|
|
return ret;
|
|
|
|
|
|
|
|
out_unlock:
|
|
|
|
mutex_unlock(&nf_ct_ecache_mutex);
|
|
|
|
return ret;
|
2007-03-14 23:40:10 +00:00
|
|
|
}
|
2007-07-08 05:30:49 +00:00
|
|
|
EXPORT_SYMBOL_GPL(nf_ct_expect_register_notifier);
|
2007-03-14 23:40:10 +00:00
|
|
|
|
2011-11-21 23:16:51 +00:00
|
|
|
void nf_ct_expect_unregister_notifier(struct net *net,
|
|
|
|
struct nf_exp_event_notifier *new)
|
2007-03-14 23:40:10 +00:00
|
|
|
{
|
2010-05-10 16:47:57 +00:00
|
|
|
struct nf_exp_event_notifier *notify;
|
|
|
|
|
2009-06-03 08:32:06 +00:00
|
|
|
mutex_lock(&nf_ct_ecache_mutex);
|
2011-11-21 23:16:51 +00:00
|
|
|
notify = rcu_dereference_protected(net->ct.nf_expect_event_cb,
|
2010-05-10 16:47:57 +00:00
|
|
|
lockdep_is_held(&nf_ct_ecache_mutex));
|
|
|
|
BUG_ON(notify != new);
|
2011-11-21 23:16:51 +00:00
|
|
|
RCU_INIT_POINTER(net->ct.nf_expect_event_cb, NULL);
|
2009-06-03 08:32:06 +00:00
|
|
|
mutex_unlock(&nf_ct_ecache_mutex);
|
2007-03-14 23:40:10 +00:00
|
|
|
}
|
2007-07-08 05:30:49 +00:00
|
|
|
EXPORT_SYMBOL_GPL(nf_ct_expect_unregister_notifier);
|
2009-06-13 10:26:29 +00:00
|
|
|
|
|
|
|
#define NF_CT_EVENTS_DEFAULT 1
|
|
|
|
static int nf_ct_events __read_mostly = NF_CT_EVENTS_DEFAULT;
|
netfilter: conntrack: optional reliable conntrack event delivery
This patch improves ctnetlink event reliability if one broadcast
listener has set the NETLINK_BROADCAST_ERROR socket option.
The logic is the following: if an event delivery fails, we keep
the undelivered events in the missed event cache. Once the next
packet arrives, we add the new events (if any) to the missed
events in the cache and we try a new delivery, and so on. Thus,
if ctnetlink fails to deliver an event, we try to deliver them
once we see a new packet. Therefore, we may lose state
transitions but the userspace process gets in sync at some point.
At worst case, if no events were delivered to userspace, we make
sure that destroy events are successfully delivered. Basically,
if ctnetlink fails to deliver the destroy event, we remove the
conntrack entry from the hashes and we insert them in the dying
list, which contains inactive entries. Then, the conntrack timer
is added with an extra grace timeout of random32() % 15 seconds
to trigger the event again (this grace timeout is tunable via
/proc). The use of a limited random timeout value allows
distributing the "destroy" resends, thus, avoiding accumulating
lots "destroy" events at the same time. Event delivery may
re-order but we can identify them by means of the tuple plus
the conntrack ID.
The maximum number of conntrack entries (active or inactive) is
still handled by nf_conntrack_max. Thus, we may start dropping
packets at some point if we accumulate a lot of inactive conntrack
entries that did not successfully report the destroy event to
userspace.
During my stress tests consisting of setting a very small buffer
of 2048 bytes for conntrackd and the NETLINK_BROADCAST_ERROR socket
flag, and generating lots of very small connections, I noticed
very few destroy entries on the fly waiting to be resend.
A simple way to test this patch consist of creating a lot of
entries, set a very small Netlink buffer in conntrackd (+ a patch
which is not in the git tree to set the BROADCAST_ERROR flag)
and invoke `conntrack -F'.
For expectations, no changes are introduced in this patch.
Currently, event delivery is only done for new expectations (no
events from expectation expiration, removal and confirmation).
In that case, they need a per-expectation event cache to implement
the same idea that is exposed in this patch.
This patch can be useful to provide reliable flow-accouting. We
still have to add a new conntrack extension to store the creation
and destroy time.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: Patrick McHardy <kaber@trash.net>
2009-06-13 10:30:52 +00:00
|
|
|
static int nf_ct_events_retry_timeout __read_mostly = 15*HZ;
|
2009-06-13 10:26:29 +00:00
|
|
|
|
|
|
|
#ifdef CONFIG_SYSCTL
|
|
|
|
static struct ctl_table event_sysctl_table[] = {
|
|
|
|
{
|
|
|
|
.procname = "nf_conntrack_events",
|
|
|
|
.data = &init_net.ct.sysctl_events,
|
|
|
|
.maxlen = sizeof(unsigned int),
|
|
|
|
.mode = 0644,
|
|
|
|
.proc_handler = proc_dointvec,
|
|
|
|
},
|
netfilter: conntrack: optional reliable conntrack event delivery
This patch improves ctnetlink event reliability if one broadcast
listener has set the NETLINK_BROADCAST_ERROR socket option.
The logic is the following: if an event delivery fails, we keep
the undelivered events in the missed event cache. Once the next
packet arrives, we add the new events (if any) to the missed
events in the cache and we try a new delivery, and so on. Thus,
if ctnetlink fails to deliver an event, we try to deliver them
once we see a new packet. Therefore, we may lose state
transitions but the userspace process gets in sync at some point.
At worst case, if no events were delivered to userspace, we make
sure that destroy events are successfully delivered. Basically,
if ctnetlink fails to deliver the destroy event, we remove the
conntrack entry from the hashes and we insert them in the dying
list, which contains inactive entries. Then, the conntrack timer
is added with an extra grace timeout of random32() % 15 seconds
to trigger the event again (this grace timeout is tunable via
/proc). The use of a limited random timeout value allows
distributing the "destroy" resends, thus, avoiding accumulating
lots "destroy" events at the same time. Event delivery may
re-order but we can identify them by means of the tuple plus
the conntrack ID.
The maximum number of conntrack entries (active or inactive) is
still handled by nf_conntrack_max. Thus, we may start dropping
packets at some point if we accumulate a lot of inactive conntrack
entries that did not successfully report the destroy event to
userspace.
During my stress tests consisting of setting a very small buffer
of 2048 bytes for conntrackd and the NETLINK_BROADCAST_ERROR socket
flag, and generating lots of very small connections, I noticed
very few destroy entries on the fly waiting to be resend.
A simple way to test this patch consist of creating a lot of
entries, set a very small Netlink buffer in conntrackd (+ a patch
which is not in the git tree to set the BROADCAST_ERROR flag)
and invoke `conntrack -F'.
For expectations, no changes are introduced in this patch.
Currently, event delivery is only done for new expectations (no
events from expectation expiration, removal and confirmation).
In that case, they need a per-expectation event cache to implement
the same idea that is exposed in this patch.
This patch can be useful to provide reliable flow-accouting. We
still have to add a new conntrack extension to store the creation
and destroy time.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: Patrick McHardy <kaber@trash.net>
2009-06-13 10:30:52 +00:00
|
|
|
{
|
|
|
|
.procname = "nf_conntrack_events_retry_timeout",
|
|
|
|
.data = &init_net.ct.sysctl_events_retry_timeout,
|
|
|
|
.maxlen = sizeof(unsigned int),
|
|
|
|
.mode = 0644,
|
|
|
|
.proc_handler = proc_dointvec_jiffies,
|
|
|
|
},
|
2009-06-13 10:26:29 +00:00
|
|
|
{}
|
|
|
|
};
|
|
|
|
#endif /* CONFIG_SYSCTL */
|
|
|
|
|
|
|
|
static struct nf_ct_ext_type event_extend __read_mostly = {
|
|
|
|
.len = sizeof(struct nf_conntrack_ecache),
|
|
|
|
.align = __alignof__(struct nf_conntrack_ecache),
|
|
|
|
.id = NF_CT_EXT_ECACHE,
|
|
|
|
};
|
|
|
|
|
|
|
|
#ifdef CONFIG_SYSCTL
|
|
|
|
static int nf_conntrack_event_init_sysctl(struct net *net)
|
|
|
|
{
|
|
|
|
struct ctl_table *table;
|
|
|
|
|
|
|
|
table = kmemdup(event_sysctl_table, sizeof(event_sysctl_table),
|
|
|
|
GFP_KERNEL);
|
|
|
|
if (!table)
|
|
|
|
goto out;
|
|
|
|
|
|
|
|
table[0].data = &net->ct.sysctl_events;
|
netfilter: conntrack: optional reliable conntrack event delivery
This patch improves ctnetlink event reliability if one broadcast
listener has set the NETLINK_BROADCAST_ERROR socket option.
The logic is the following: if an event delivery fails, we keep
the undelivered events in the missed event cache. Once the next
packet arrives, we add the new events (if any) to the missed
events in the cache and we try a new delivery, and so on. Thus,
if ctnetlink fails to deliver an event, we try to deliver them
once we see a new packet. Therefore, we may lose state
transitions but the userspace process gets in sync at some point.
At worst case, if no events were delivered to userspace, we make
sure that destroy events are successfully delivered. Basically,
if ctnetlink fails to deliver the destroy event, we remove the
conntrack entry from the hashes and we insert them in the dying
list, which contains inactive entries. Then, the conntrack timer
is added with an extra grace timeout of random32() % 15 seconds
to trigger the event again (this grace timeout is tunable via
/proc). The use of a limited random timeout value allows
distributing the "destroy" resends, thus, avoiding accumulating
lots "destroy" events at the same time. Event delivery may
re-order but we can identify them by means of the tuple plus
the conntrack ID.
The maximum number of conntrack entries (active or inactive) is
still handled by nf_conntrack_max. Thus, we may start dropping
packets at some point if we accumulate a lot of inactive conntrack
entries that did not successfully report the destroy event to
userspace.
During my stress tests consisting of setting a very small buffer
of 2048 bytes for conntrackd and the NETLINK_BROADCAST_ERROR socket
flag, and generating lots of very small connections, I noticed
very few destroy entries on the fly waiting to be resend.
A simple way to test this patch consist of creating a lot of
entries, set a very small Netlink buffer in conntrackd (+ a patch
which is not in the git tree to set the BROADCAST_ERROR flag)
and invoke `conntrack -F'.
For expectations, no changes are introduced in this patch.
Currently, event delivery is only done for new expectations (no
events from expectation expiration, removal and confirmation).
In that case, they need a per-expectation event cache to implement
the same idea that is exposed in this patch.
This patch can be useful to provide reliable flow-accouting. We
still have to add a new conntrack extension to store the creation
and destroy time.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: Patrick McHardy <kaber@trash.net>
2009-06-13 10:30:52 +00:00
|
|
|
table[1].data = &net->ct.sysctl_events_retry_timeout;
|
2009-06-13 10:26:29 +00:00
|
|
|
|
|
|
|
net->ct.event_sysctl_header =
|
|
|
|
register_net_sysctl_table(net,
|
|
|
|
nf_net_netfilter_sysctl_path, table);
|
|
|
|
if (!net->ct.event_sysctl_header) {
|
|
|
|
printk(KERN_ERR "nf_ct_event: can't register to sysctl.\n");
|
|
|
|
goto out_register;
|
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
out_register:
|
|
|
|
kfree(table);
|
|
|
|
out:
|
|
|
|
return -ENOMEM;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void nf_conntrack_event_fini_sysctl(struct net *net)
|
|
|
|
{
|
|
|
|
struct ctl_table *table;
|
|
|
|
|
|
|
|
table = net->ct.event_sysctl_header->ctl_table_arg;
|
|
|
|
unregister_net_sysctl_table(net->ct.event_sysctl_header);
|
|
|
|
kfree(table);
|
|
|
|
}
|
|
|
|
#else
|
|
|
|
static int nf_conntrack_event_init_sysctl(struct net *net)
|
|
|
|
{
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void nf_conntrack_event_fini_sysctl(struct net *net)
|
|
|
|
{
|
|
|
|
}
|
|
|
|
#endif /* CONFIG_SYSCTL */
|
|
|
|
|
|
|
|
int nf_conntrack_ecache_init(struct net *net)
|
|
|
|
{
|
|
|
|
int ret;
|
|
|
|
|
|
|
|
net->ct.sysctl_events = nf_ct_events;
|
netfilter: conntrack: optional reliable conntrack event delivery
This patch improves ctnetlink event reliability if one broadcast
listener has set the NETLINK_BROADCAST_ERROR socket option.
The logic is the following: if an event delivery fails, we keep
the undelivered events in the missed event cache. Once the next
packet arrives, we add the new events (if any) to the missed
events in the cache and we try a new delivery, and so on. Thus,
if ctnetlink fails to deliver an event, we try to deliver them
once we see a new packet. Therefore, we may lose state
transitions but the userspace process gets in sync at some point.
At worst case, if no events were delivered to userspace, we make
sure that destroy events are successfully delivered. Basically,
if ctnetlink fails to deliver the destroy event, we remove the
conntrack entry from the hashes and we insert them in the dying
list, which contains inactive entries. Then, the conntrack timer
is added with an extra grace timeout of random32() % 15 seconds
to trigger the event again (this grace timeout is tunable via
/proc). The use of a limited random timeout value allows
distributing the "destroy" resends, thus, avoiding accumulating
lots "destroy" events at the same time. Event delivery may
re-order but we can identify them by means of the tuple plus
the conntrack ID.
The maximum number of conntrack entries (active or inactive) is
still handled by nf_conntrack_max. Thus, we may start dropping
packets at some point if we accumulate a lot of inactive conntrack
entries that did not successfully report the destroy event to
userspace.
During my stress tests consisting of setting a very small buffer
of 2048 bytes for conntrackd and the NETLINK_BROADCAST_ERROR socket
flag, and generating lots of very small connections, I noticed
very few destroy entries on the fly waiting to be resend.
A simple way to test this patch consist of creating a lot of
entries, set a very small Netlink buffer in conntrackd (+ a patch
which is not in the git tree to set the BROADCAST_ERROR flag)
and invoke `conntrack -F'.
For expectations, no changes are introduced in this patch.
Currently, event delivery is only done for new expectations (no
events from expectation expiration, removal and confirmation).
In that case, they need a per-expectation event cache to implement
the same idea that is exposed in this patch.
This patch can be useful to provide reliable flow-accouting. We
still have to add a new conntrack extension to store the creation
and destroy time.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: Patrick McHardy <kaber@trash.net>
2009-06-13 10:30:52 +00:00
|
|
|
net->ct.sysctl_events_retry_timeout = nf_ct_events_retry_timeout;
|
2009-06-13 10:26:29 +00:00
|
|
|
|
|
|
|
if (net_eq(net, &init_net)) {
|
|
|
|
ret = nf_ct_extend_register(&event_extend);
|
|
|
|
if (ret < 0) {
|
|
|
|
printk(KERN_ERR "nf_ct_event: Unable to register "
|
|
|
|
"event extension.\n");
|
|
|
|
goto out_extend_register;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
ret = nf_conntrack_event_init_sysctl(net);
|
|
|
|
if (ret < 0)
|
|
|
|
goto out_sysctl;
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
out_sysctl:
|
|
|
|
if (net_eq(net, &init_net))
|
|
|
|
nf_ct_extend_unregister(&event_extend);
|
|
|
|
out_extend_register:
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
|
|
|
void nf_conntrack_ecache_fini(struct net *net)
|
|
|
|
{
|
|
|
|
nf_conntrack_event_fini_sysctl(net);
|
|
|
|
if (net_eq(net, &init_net))
|
|
|
|
nf_ct_extend_unregister(&event_extend);
|
|
|
|
}
|