netfilter: remove extra timer from ecache extension
Subsystems (2)
Florian Westphal
maintainerPATCH 1/1 netfilter: remove extra timer from ecache extension
This brings the (per-conntrack) ecache extension back to 24 bytes in size (was 112 byte on x86_64 with lockdep on).
Instead we use a per-ns tasklet to re-trigger event delivery. When we enqueue a ct entry into the dying list, the tasklet is scheduled.
The tasklet will then deliver up to 20 entries. It will re-sched itself if not all the pending events could be delivered.
While at it, dying list handling is moved into ecache.c, since its only revlevant if ct events are enabled.
---
Note: Conflicts with "improve conntrack object traceability".
The patch assumes the dying list only contains entries where the delete event has not been delivered yet.
With that patch, all conntracks are put on the dying list, including those who are about to be free'd.
I THINK that this is fixable by skipping dying-list entries with IPS_DYING_BIT set. However, this will increase the tasket workload.
Any other ideas?
include/net/netfilter/nf_conntrack.h | 1 - include/net/netfilter/nf_conntrack_ecache.h | 9 +++- include/net/netns/conntrack.h | 5 +- net/netfilter/nf_conntrack_core.c | 58 ------------------- net/netfilter/nf_conntrack_ecache.c | 80 +++++++++++++++++++++++--- 5 files changed, 82 insertions(+), 71 deletions(-)
include/net/netfilter/nf_conntrack.h +0 -1
@@ -182,7 +182,6 @@ __nf_conntrack_find(struct net *net, u16 zone,extern int nf_conntrack_hash_check_insert (struct nf_conn *ct);extern void nf_ct_delete_from_lists (struct nf_conn *ct); -extern void nf_ct_insert_dying_list (struct nf_conn *ct);extern void nf_conntrack_flush_report (struct net *net,u32 pid,int report);
include/net/netfilter/nf_conntrack_ecache.h +8 -1
@@ -18,7 +18,6 @@ struct nf_conntrack_ecache {u16 ctmask;/* bitmask of ct events to be delivered */ u16 expmask;/* bitmask of expect events to be delivered */ u32 portid;/* netlink portid of destroyer */ -struct timer_list timeout; }; static inline struct nf_conntrack_ecache *
@@ -207,6 +206,8 @@ nf_ct_expect_event(enum ip_conntrack_expect_events event,nf_ct_expect_event_report (event, exp,0 ,0 ); } +extern void nf_ct_insert_dying_list (struct nf_conn *ct); +extern void nf_ct_release_dying_list (struct net *net);extern int nf_conntrack_ecache_init (struct net *net);extern void nf_conntrack_ecache_fini (struct net *net);
@@ -232,6 +233,12 @@ static inline void nf_ct_expect_event_report(enum ip_conntrack_expect_events e, u32 portid, int report) {} +static inline void nf_ct_insert_dying_list (struct nf_conn *ct) +{ +WARN_ON_ONCE (1 );/* never called when CONFIG_NF_CONNTRACK_EVENTS=n */ +} +static inline void nf_ct_release_dying_list (struct net *net) {} +static inline int nf_conntrack_ecache_init (struct net *net) {return 0 ;
include/net/netns/conntrack.h +4 -1
@@ -4,6 +4,7 @@#include <linux/list.h> #include <linux/list_nulls.h> #include <linux/atomic.h> +#include <linux/interrupt.h> #include <linux/netfilter/nf_conntrack_tcp.h> struct ctl_table_header ;
@@ -71,11 +72,13 @@ struct netns_ct {struct hlist_head *expect_hash;struct hlist_nulls_head unconfirmed;struct hlist_nulls_head dying; +#ifdef CONFIG_NF_CONNTRACK_EVENTS +struct tasklet_struct dying_tasklet; +#endif struct ip_conntrack_stat __percpu *stat;struct nf_ct_event_notifier __rcu *nf_conntrack_event_cb;struct nf_exp_event_notifier __rcu *nf_expect_event_cb;int sysctl_events; -unsigned int sysctl_events_retry_timeout;int sysctl_acct;int sysctl_tstamp;int sysctl_checksum;
net/netfilter/nf_conntrack_core.c +0 -58
70 lines not on this page · raw message
net/netfilter/nf_conntrack_ecache.c +70 -10
122 lines not on this page · raw message
Signature · 1 line
1.7.8.6
Pablo Neira Ayuso
maintainerQuoted 2 lines · Florian Westphal
This brings the (per-conntrack) ecache extension back to 24 bytes in size (was 112 byte on
x86_64with lockdep on).
I like this.
Quoted 6 lines
Instead we use a per-ns tasklet to re-trigger event delivery. When we enqueue a ct entry into the dying list, the tasklet is scheduled.
The tasklet will then deliver up to 20 entries. It will re-sched itself if not all the pending events could be delivered.
I would like to give a test to this patch in my testbed.
And I wonder if we can make it better with some timer-based garbage collector that randomly / adaptively runs to give tries to deliver events.
I remember that insisting too often in the delivery of missed events does not make any good.
Quoted 2 lines
While at it, dying list handling is moved into
ecache.c, since its only revlevant if ct events are enabled.
I like this too. That code really belongs where you moved it.
Quoted 4 lines
Signed-off-by Florian Westphal <[email protected]>---
Note: Conflicts with "improve conntrack object traceability".
Quoted 8 lines
The patch assumes the dying list only contains entries where the delete event has not been delivered yet.
With that patch, all conntracks are put on the dying list, including those who are about to be free'd.
I THINK that this is fixable by skipping dying-list entries with
IPS_DYING_BITset. However, this will increase the tasket workload.
I think you will mostly find entries that are waiting for its event to be delivered. So playing with IPS_DYING_BIT seems the right way to go to me.
Quoted 299 lines
Any other ideas?
include/net/netfilter/nf_conntrack.h | 1 - include/net/netfilter/nf_conntrack_ecache.h | 9 +++- include/net/netns/conntrack.h | 5 +- net/netfilter/nf_conntrack_core.c | 58 ------------------- net/netfilter/nf_conntrack_ecache.c | 80 +++++++++++++++++++++++--- 5 files changed, 82 insertions(+), 71 deletions(-)include/net/netfilter/nf_conntrack.h +0
-1@@ -182,7 +182,6 @@ __nf_conntrack_find(struct net *net, u16 zone,extern int nf_conntrack_hash_check_insert (struct nf_conn *ct);extern void nf_ct_delete_from_lists (struct nf_conn *ct); -extern void nf_ct_insert_dying_list (struct nf_conn *ct);extern void nf_conntrack_flush_report (struct net *net,u32 pid,int report);include/net/netfilter/nf_conntrack_ecache.h +8
-1@@ -18,7 +18,6 @@ struct nf_conntrack_ecache {u16 ctmask;/* bitmask of ct events to be delivered */ u16 expmask;/* bitmask of expect events to be delivered */ u32 portid;/* netlink portid of destroyer */ -struct timer_list timeout; }; static inline struct nf_conntrack_ecache *@@ -207,6 +206,8 @@ nf_ct_expect_event(enum ip_conntrack_expect_events event,nf_ct_expect_event_report (event, exp,0 ,0 ); } +extern void nf_ct_insert_dying_list (struct nf_conn *ct); +extern void nf_ct_release_dying_list (struct net *net);extern int nf_conntrack_ecache_init (struct net *net);extern void nf_conntrack_ecache_fini (struct net *net);@@ -232,6 +233,12 @@ static inline void nf_ct_expect_event_report(enum ip_conntrack_expect_events e, u32 portid, int report) {} +static inline void nf_ct_insert_dying_list (struct nf_conn *ct) +{ +WARN_ON_ONCE (1 );/* never called when CONFIG_NF_CONNTRACK_EVENTS=n */ +} +static inline void nf_ct_release_dying_list (struct net *net) {} +static inline int nf_conntrack_ecache_init (struct net *net) {return 0 ;include/net/netns/conntrack.h +4
-1@@ -4,6 +4,7 @@#include <linux/list.h> #include <linux/list_nulls.h> #include <linux/atomic.h> +#include <linux/interrupt.h> #include <linux/netfilter/nf_conntrack_tcp.h> struct ctl_table_header ;@@ -71,11 +72,13 @@ struct netns_ct {struct hlist_head *expect_hash;struct hlist_nulls_head unconfirmed;struct hlist_nulls_head dying; +#ifdef CONFIG_NF_CONNTRACK_EVENTS +struct tasklet_struct dying_tasklet; +#endif struct ip_conntrack_stat __percpu *stat;struct nf_ct_event_notifier __rcu *nf_conntrack_event_cb;struct nf_exp_event_notifier __rcu *nf_expect_event_cb;int sysctl_events; -unsigned int sysctl_events_retry_timeout;int sysctl_acct;int sysctl_tstamp;int sysctl_checksum;net/netfilter/nf_conntrack_core.c +0
-58@@ -251,49 +251,6 @@ void nf_ct_delete_from_lists(struct nf_conn *ct) }EXPORT_SYMBOL_GPL (nf_ct_delete_from_lists); -static void death_by_event (unsigned long ul_conntrack) -{ -struct nf_conn *ct = (void *)ul_conntrack; -struct net *net =nf_ct_net (ct); -struct nf_conntrack_ecache *ecache =nf_ct_ecache_find (ct); - -BUG_ON (ecache ==NULL ); - -if (nf_conntrack_event (IPCT_DESTROY , ct) <0 ) { -/* bad luck, let's retry again */ - ecache->timeout.expires = jiffies + - (random32 () % net->ct.sysctl_events_retry_timeout); -add_timer (&ecache->timeout); -return ; - } -/* we've got the event delivered, now it's dying */ -set_bit (IPS_DYING_BIT , &ct->status); -spin_lock (&nf_conntrack_lock); -hlist_nulls_del (&ct->tuplehash[IP_CT_DIR_ORIGINAL ].hnnode); -spin_unlock (&nf_conntrack_lock); -nf_ct_put (ct); -} - -void nf_ct_insert_dying_list (struct nf_conn *ct) -{ -struct net *net =nf_ct_net (ct); -struct nf_conntrack_ecache *ecache =nf_ct_ecache_find (ct); - -BUG_ON (ecache ==NULL ); - -/* add this conntrack to the dying list */ -spin_lock_bh (&nf_conntrack_lock); -hlist_nulls_add_head (&ct->tuplehash[IP_CT_DIR_ORIGINAL ].hnnode, - &net->ct.dying); -spin_unlock_bh (&nf_conntrack_lock); -/* set a new timer to retry event delivery */ -setup_timer (&ecache->timeout, death_by_event, (unsigned long )ct); - ecache->timeout.expires = jiffies + - (random32 () % net->ct.sysctl_events_retry_timeout); -add_timer (&ecache->timeout); -} -EXPORT_SYMBOL_GPL (nf_ct_insert_dying_list); -static void death_by_timeout (unsigned long ul_conntrack) {struct nf_conn *ct = (void *)ul_conntrack;@@ -1311,21 +1268,6 @@ void nf_conntrack_flush_report(struct net *net, u32 pid, int report) }EXPORT_SYMBOL_GPL (nf_conntrack_flush_report); -static void nf_ct_release_dying_list (struct net *net) -{ -struct nf_conntrack_tuple_hash *h; -struct nf_conn *ct; -struct hlist_nulls_node *n; - -spin_lock_bh (&nf_conntrack_lock); -hlist_nulls_for_each_entry (h, n, &net->ct.dying, hnnode) { - ct =nf_ct_tuplehash_to_ctrack (h); -/* never fails to remove them, no listeners at this point */ -nf_ct_kill (ct); - } -spin_unlock_bh (&nf_conntrack_lock); -} -static int untrack_refs (void ) {int cnt =0 , cpu;net/netfilter/nf_conntrack_ecache.c +70
-10@@ -27,6 +27,38 @@static DEFINE_MUTEX (nf_ct_ecache_mutex); +static void dying_tasklet_retry_events (unsigned long ctnetns) +{ +struct netns_ct *ctnet = (void *) ctnetns; +struct nf_conntrack_tuple_hash *h; +struct hlist_nulls_node *n; +struct nf_conn *ct; +int err =0 ; +int budget =20 ; + +hlist_nulls_for_each_entry (h, n, &ctnet->dying, hnnode) { + ct =nf_ct_tuplehash_to_ctrack (h); + + err =nf_conntrack_event (IPCT_DESTROY , ct); +if (err) +break ; +spin_lock_bh (&nf_conntrack_lock); +hlist_nulls_del (&ct->tuplehash[IP_CT_DIR_ORIGINAL ].hnnode); +spin_unlock_bh (&nf_conntrack_lock); + +/* we've got the event delivered, now it's dying */ +set_bit (IPS_DYING_BIT , &ct->status); + +nf_ct_put (ct); +if (--budget ==0 ) +break ; + } + +/* err or budget exhausted? -> dying list not empty -- resched. */ +if (err || budget ==0 ) +tasklet_schedule (&ctnet->dying_tasklet); +} +/* deliver cached events and clear cache entry - must be called with locally * disabled softirqs */ void nf_ct_deliver_cached_events (struct nf_conn *ct)@@ -81,6 +113,39 @@ out_unlock: }EXPORT_SYMBOL_GPL (nf_ct_deliver_cached_events); +void nf_ct_insert_dying_list (struct nf_conn *ct) +{ +struct net *net =nf_ct_net (ct); + +/* add this conntrack to the dying list */ +spin_lock_bh (&nf_conntrack_lock); +hlist_nulls_add_head (&ct->tuplehash[IP_CT_DIR_ORIGINAL ].hnnode, + &net->ct.dying); +spin_unlock_bh (&nf_conntrack_lock); + +/* retry event delivery */ +tasklet_schedule (&net->ct.dying_tasklet); +} +EXPORT_SYMBOL_GPL (nf_ct_insert_dying_list); + +void nf_ct_release_dying_list (struct net *net) +{ +struct nf_conntrack_tuple_hash *h; +struct nf_conn *ct; +struct hlist_nulls_node *n; + +/* no listeners at this point */ +hlist_nulls_for_each_entry (h, n, &net->ct.dying, hnnode) { + ct =nf_ct_tuplehash_to_ctrack (h); + +spin_lock_bh (&nf_conntrack_lock); +hlist_nulls_del (&ct->tuplehash[IP_CT_DIR_ORIGINAL ].hnnode); +spin_unlock_bh (&nf_conntrack_lock); + +nf_ct_put (ct); + } +} +int nf_conntrack_register_notifier (struct net *net,struct nf_ct_event_notifier *new) {@@ -155,7 +220,6 @@ EXPORT_SYMBOL_GPL(nf_ct_expect_unregister_notifier);#define NF_CT_EVENTS_DEFAULT 1static int nf_ct_events__read_mostly =NF_CT_EVENTS_DEFAULT ; -static int nf_ct_events_retry_timeout__read_mostly =15 *HZ ;#ifdef CONFIG_SYSCTL static struct ctl_table event_sysctl_table[] = {@@ -166,13 +230,6 @@ static struct ctl_table event_sysctl_table[] = { .mode = 0644, .proc_handler = proc_dointvec, }, - { - .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, - }, {} };#endif /* CONFIG_SYSCTL */ @@ -194,7 +251,6 @@ static int nf_conntrack_event_init_sysctl(struct net *net) goto out; table[0 ].data = &net->ct.sysctl_events; - table[1 ].data = &net->ct.sysctl_events_retry_timeout; net->ct.event_sysctl_header =register_net_sysctl (net,"net/netfilter" , table);@@ -234,7 +290,9 @@ int nf_conntrack_ecache_init(struct net *net)int ret; net->ct.sysctl_events = nf_ct_events; - net->ct.sysctl_events_retry_timeout = nf_ct_events_retry_timeout; + +tasklet_init (&net->ct.dying_tasklet, dying_tasklet_retry_events, + (unsigned long ) &net->ct);if (net_eq (net, &init_net)) { ret =nf_ct_extend_register (&event_extend);@@ -260,6 +318,8 @@ out_extend_register:void nf_conntrack_ecache_fini (struct net *net) { +tasklet_kill (&net->ct.dying_tasklet); +nf_conntrack_event_fini_sysctl (net);if (net_eq (net, &init_net))nf_ct_extend_unregister (&event_extend);Signature · 6 lines
1.7.8.6 -- To unsubscribe from this list: send the line "unsubscribe netfilter-devel" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html
Florian Westphal
maintainerQuoted 8 lines · Pablo Neira Ayuso
Quoted 6 lines
Instead we use a per-ns tasklet to re-trigger event delivery. When we enqueue a ct entry into the dying list, the tasklet is scheduled.
The tasklet will then deliver up to 20 entries. It will re-sched itself if not all the pending events could be delivered.
I would like to give a test to this patch in my testbed.
That would be great.
I can re-spin the patch on top of your dying-list changes, if you prefer to test all patches at the same time.
Quoted 6 lines
And I wonder if we can make it better with some timer-based garbage collector that randomly / adaptively runs to give tries to deliver events.
I remember that insisting too often in the delivery of missed events does not make any good.
Yes, from my tests only very few attempts are successful.
However, I've failed to come up with a scheme where events are re-tried in a timely manner without risk of acummulating a large event/entry backlog.
Quoted 8 lines
Quoted 8 lines
Note: Conflicts with "improve conntrack object traceability".
The patch assumes the dying list only contains entries where the delete event has not been delivered yet.
With that patch, all conntracks are put on the dying list, including those who are about to be free'd.
Quoted 6 lines
Quoted 2 lines
I THINK that this is fixable by skipping dying-list entries with
IPS_DYING_BITset. However, this will increase the tasket workload.I think you will mostly find entries that are waiting for its event to be delivered. So playing with
IPS_DYING_BITseems the right way to go to me.
Perfect. Thats what I'll do, then.
Pablo Neira Ayuso
maintainerOn Mon, Dec 03, 2012 at 04:17:06PM +0100, Florian Westphal wrote:
[...]
Quoted 2 lines
I can re-spin the patch on top of your dying-list changes, if you prefer to test all patches at the same time.
Please, do so. You can use nf-next tree, I have applied those to it.
As promised, I'll send a follow-up patch to fix the issue that you spotted regarding the dying list and the output path of nf_conntrack.
Thanks.