From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jiri Pirko Subject: Re: [PATCH net-next 1/2] net: add NETDEV_PRECHANGEMTU to notify before mtu change happens Date: Thu, 16 Jan 2014 08:18:03 +0100 Message-ID: <20140116071803.GB2815@minipsycho.orion> References: <1389826939-20691-1-git-send-email-vfalico@redhat.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: netdev@vger.kernel.org, "David S. Miller" , Eric Dumazet , Nicolas Dichtel , Cong Wang To: Veaceslav Falico Return-path: Received: from mail-ea0-f169.google.com ([209.85.215.169]:61850 "EHLO mail-ea0-f169.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750957AbaAPHSH (ORCPT ); Thu, 16 Jan 2014 02:18:07 -0500 Received: by mail-ea0-f169.google.com with SMTP id l9so798067eaj.14 for ; Wed, 15 Jan 2014 23:18:05 -0800 (PST) Content-Disposition: inline In-Reply-To: <1389826939-20691-1-git-send-email-vfalico@redhat.com> Sender: netdev-owner@vger.kernel.org List-ID: Thu, Jan 16, 2014 at 12:02:18AM CET, vfalico@redhat.com wrote: >Currently, if a device changes its mtu, first the change happens (invloving >all the side effects), and after that the NETDEV_CHANGEMTU is sent so that >other devices can catch up with the new mtu. However, if they return >NOTIFY_BAD, then the change is reverted and error returned. > >This is a really long and costy operation (sometimes). To fix this, add >NETDEV_PRECHANGEMTU notification which is called prior to any change >actually happening, and if any callee returns NOTIFY_BAD - the change is >aborted. This way we're skipping all the playing with apply/revert the mtu. > >CC: "David S. Miller" >CC: Jiri Pirko >CC: Eric Dumazet >CC: Nicolas Dichtel >CC: Cong Wang >Signed-off-by: Veaceslav Falico >--- > include/linux/netdevice.h | 3 ++- > net/core/dev.c | 5 +++++ > 2 files changed, 7 insertions(+), 1 deletion(-) > >diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h >index 5c88ab1..6fa2c91 100644 >--- a/include/linux/netdevice.h >+++ b/include/linux/netdevice.h >@@ -1718,7 +1718,7 @@ struct pcpu_sw_netstats { > #define NETDEV_CHANGE 0x0004 /* Notify device state change */ > #define NETDEV_REGISTER 0x0005 > #define NETDEV_UNREGISTER 0x0006 >-#define NETDEV_CHANGEMTU 0x0007 >+#define NETDEV_CHANGEMTU 0x0007 /* notify after mtu change happened */ > #define NETDEV_CHANGEADDR 0x0008 > #define NETDEV_GOING_DOWN 0x0009 > #define NETDEV_CHANGENAME 0x000A >@@ -1734,6 +1734,7 @@ struct pcpu_sw_netstats { > #define NETDEV_JOIN 0x0014 > #define NETDEV_CHANGEUPPER 0x0015 > #define NETDEV_RESEND_IGMP 0x0016 >+#define NETDEV_PRECHANGEMTU 0x0017 /* notify before mtu change happened */ > > int register_netdevice_notifier(struct notifier_block *nb); > int unregister_netdevice_notifier(struct notifier_block *nb); >diff --git a/net/core/dev.c b/net/core/dev.c >index 20c834e..cb78923 100644 >--- a/net/core/dev.c >+++ b/net/core/dev.c >@@ -5360,6 +5360,11 @@ int dev_set_mtu(struct net_device *dev, int new_mtu) > if (!netif_device_present(dev)) > return -ENODEV; > >+ err = call_netdevice_notifiers(NETDEV_PRECHANGEMTU, dev); >+ err = notifier_to_errno(err); >+ if (err) >+ return err; >+ > orig_mtu = dev->mtu; > err = __dev_set_mtu(dev, new_mtu); > >-- >1.8.4 > I like this Acked-by: Jiri Pirko