From mboxrd@z Thu Jan 1 00:00:00 1970 From: Eric Dumazet Subject: [PATCH net-next v2] net: remove delay at device dismantle Date: Fri, 10 Aug 2012 17:42:49 +0200 Message-ID: <1344613369.31104.2734.camel@edumazet-glaptop> References: <1344590824.31104.1953.camel@edumazet-glaptop> <20120810.034211.994338127277150687.davem@davemloft.net> <1344596809.31104.2358.camel@edumazet-glaptop> <1344609966.31104.2705.camel@edumazet-glaptop> Mime-Version: 1.0 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: 7bit Cc: netdev , Tom Herbert , Mahesh Bandewar , "Eric W. Biederman" To: David Miller Return-path: Received: from mail-bk0-f46.google.com ([209.85.214.46]:63099 "EHLO mail-bk0-f46.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754825Ab2HJPmz (ORCPT ); Fri, 10 Aug 2012 11:42:55 -0400 Received: by bkwj10 with SMTP id j10so649620bkw.19 for ; Fri, 10 Aug 2012 08:42:53 -0700 (PDT) In-Reply-To: <1344609966.31104.2705.camel@edumazet-glaptop> Sender: netdev-owner@vger.kernel.org List-ID: From: Eric Dumazet I noticed extra one second delay in device dismantle, tracked down to a call to dst_dev_event() while some call_rcu() are still in RCU queues. These call_rcu() were posted by rt_free(struct rtable *rt) calls. We then wait a little (but one second) in netdev_wait_allrefs() before kicking again NETDEV_UNREGISTER. As the call_rcu() are now completed, dst_dev_event() can do the needed device swap on busy dst. To solve this problem, add a new NETDEV_UNREGISTER_FINAL, called after a rcu_barrier(), but outside of RTNL lock. Use NETDEV_UNREGISTER_FINAL with care ! Change dst_dev_event() handler to react to NETDEV_UNREGISTER_FINAL Also remove NETDEV_UNREGISTER_BATCH, as its not used anymore after IP cache removal. Signed-off-by: Eric Dumazet Cc: Tom Herbert Cc: Mahesh Bandewar Cc: "Eric W. Biederman" --- v2: NETDEV_UNREGISTER_FINAL called outside of rtnl lock as its more risky, base this patch on net-next include/linux/netdevice.h | 2 +- net/core/dev.c | 22 ++++++++-------------- net/core/dst.c | 2 +- net/core/fib_rules.c | 3 ++- net/core/rtnetlink.c | 2 +- net/ipv4/devinet.c | 6 +++++- net/ipv4/fib_frontend.c | 2 -- 7 files changed, 18 insertions(+), 21 deletions(-) diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index a9db4f3..e3a43fd 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -1550,7 +1550,7 @@ struct packet_type { #define NETDEV_PRE_TYPE_CHANGE 0x000E #define NETDEV_POST_TYPE_CHANGE 0x000F #define NETDEV_POST_INIT 0x0010 -#define NETDEV_UNREGISTER_BATCH 0x0011 +#define NETDEV_UNREGISTER_FINAL 0x0011 #define NETDEV_RELEASE 0x0012 #define NETDEV_NOTIFY_PEERS 0x0013 #define NETDEV_JOIN 0x0014 diff --git a/net/core/dev.c b/net/core/dev.c index 1f06df8..4c0f837 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -1391,7 +1391,6 @@ rollback: nb->notifier_call(nb, NETDEV_DOWN, dev); } nb->notifier_call(nb, NETDEV_UNREGISTER, dev); - nb->notifier_call(nb, NETDEV_UNREGISTER_BATCH, dev); } } @@ -1433,7 +1432,6 @@ int unregister_netdevice_notifier(struct notifier_block *nb) nb->notifier_call(nb, NETDEV_DOWN, dev); } nb->notifier_call(nb, NETDEV_UNREGISTER, dev); - nb->notifier_call(nb, NETDEV_UNREGISTER_BATCH, dev); } } unlock: @@ -1453,7 +1451,8 @@ EXPORT_SYMBOL(unregister_netdevice_notifier); int call_netdevice_notifiers(unsigned long val, struct net_device *dev) { - ASSERT_RTNL(); + if (val != NETDEV_UNREGISTER_FINAL) + ASSERT_RTNL(); return raw_notifier_call_chain(&netdev_chain, val, dev); } EXPORT_SYMBOL(call_netdevice_notifiers); @@ -5304,10 +5303,6 @@ static void rollback_registered_many(struct list_head *head) netdev_unregister_kobject(dev); } - /* Process any work delayed until the end of the batch */ - dev = list_first_entry(head, struct net_device, unreg_list); - call_netdevice_notifiers(NETDEV_UNREGISTER_BATCH, dev); - synchronize_net(); list_for_each_entry(dev, head, unreg_list) @@ -5759,9 +5754,8 @@ static void netdev_wait_allrefs(struct net_device *dev) /* Rebroadcast unregister notification */ call_netdevice_notifiers(NETDEV_UNREGISTER, dev); - /* don't resend NETDEV_UNREGISTER_BATCH, _BATCH users - * should have already handle it the first time */ - + rcu_barrier(); + call_netdevice_notifiers(NETDEV_UNREGISTER_FINAL, dev); if (test_bit(__LINK_STATE_LINKWATCH_PENDING, &dev->state)) { /* We must not have linkwatch events @@ -5823,9 +5817,8 @@ void netdev_run_todo(void) __rtnl_unlock(); - /* Wait for rcu callbacks to finish before attempting to drain - * the device list. This usually avoids a 250ms wait. - */ + + /* Wait for rcu callbacks to finish before next phase */ if (!list_empty(&list)) rcu_barrier(); @@ -5834,6 +5827,8 @@ void netdev_run_todo(void) = list_first_entry(&list, struct net_device, todo_list); list_del(&dev->todo_list); + call_netdevice_notifiers(NETDEV_UNREGISTER_FINAL, dev); + if (unlikely(dev->reg_state != NETREG_UNREGISTERING)) { pr_err("network todo '%s' but state %d\n", dev->name, dev->reg_state); @@ -6228,7 +6223,6 @@ int dev_change_net_namespace(struct net_device *dev, struct net *net, const char the device is just moving and can keep their slaves up. */ call_netdevice_notifiers(NETDEV_UNREGISTER, dev); - call_netdevice_notifiers(NETDEV_UNREGISTER_BATCH, dev); rtmsg_ifinfo(RTM_DELLINK, dev, ~0U); /* diff --git a/net/core/dst.c b/net/core/dst.c index 069d51d..a2d3624 100644 --- a/net/core/dst.c +++ b/net/core/dst.c @@ -366,7 +366,7 @@ static int dst_dev_event(struct notifier_block *this, unsigned long event, struct dst_entry *dst, *last = NULL; switch (event) { - case NETDEV_UNREGISTER: + case NETDEV_UNREGISTER_FINAL: case NETDEV_DOWN: mutex_lock(&dst_gc_mutex); for (dst = dst_busy_list; dst; dst = dst->next) { diff --git a/net/core/fib_rules.c b/net/core/fib_rules.c index ab7db83..5850937 100644 --- a/net/core/fib_rules.c +++ b/net/core/fib_rules.c @@ -711,15 +711,16 @@ static int fib_rules_event(struct notifier_block *this, unsigned long event, struct net *net = dev_net(dev); struct fib_rules_ops *ops; - ASSERT_RTNL(); switch (event) { case NETDEV_REGISTER: + ASSERT_RTNL(); list_for_each_entry(ops, &net->rules_ops, list) attach_rules(&ops->rules_list, dev); break; case NETDEV_UNREGISTER: + ASSERT_RTNL(); list_for_each_entry(ops, &net->rules_ops, list) detach_rules(&ops->rules_list, dev); break; diff --git a/net/core/rtnetlink.c b/net/core/rtnetlink.c index 34d975b..c64efcf 100644 --- a/net/core/rtnetlink.c +++ b/net/core/rtnetlink.c @@ -2358,7 +2358,7 @@ static int rtnetlink_event(struct notifier_block *this, unsigned long event, voi case NETDEV_PRE_TYPE_CHANGE: case NETDEV_GOING_DOWN: case NETDEV_UNREGISTER: - case NETDEV_UNREGISTER_BATCH: + case NETDEV_UNREGISTER_FINAL: case NETDEV_RELEASE: case NETDEV_JOIN: break; diff --git a/net/ipv4/devinet.c b/net/ipv4/devinet.c index adf273f..6a5e6e4 100644 --- a/net/ipv4/devinet.c +++ b/net/ipv4/devinet.c @@ -1147,8 +1147,12 @@ static int inetdev_event(struct notifier_block *this, unsigned long event, void *ptr) { struct net_device *dev = ptr; - struct in_device *in_dev = __in_dev_get_rtnl(dev); + struct in_device *in_dev; + if (event == NETDEV_UNREGISTER_FINAL) + goto out; + + in_dev = __in_dev_get_rtnl(dev); ASSERT_RTNL(); if (!in_dev) { diff --git a/net/ipv4/fib_frontend.c b/net/ipv4/fib_frontend.c index 7f073a3..43d6fea 100644 --- a/net/ipv4/fib_frontend.c +++ b/net/ipv4/fib_frontend.c @@ -1071,8 +1071,6 @@ static int fib_netdev_event(struct notifier_block *this, unsigned long event, vo case NETDEV_CHANGE: rt_cache_flush(dev_net(dev), 0); break; - case NETDEV_UNREGISTER_BATCH: - break; } return NOTIFY_DONE; }