From: Stephen Hemminger <stephen@networkplumber.org>
To: Pravin B Shelar <pshelar@nicira.com>
Cc: netdev@vger.kernel.org, dev@openvswitch.org
Subject: Re: [PATCH vxlan v2 1/8] vxlan: Fix module cleanup.
Date: Wed, 26 Jun 2013 14:48:40 -0700 [thread overview]
Message-ID: <20130626144840.05ff8c80@nehalam.linuxnetplumber.net> (raw)
In-Reply-To: <1372277609-17467-1-git-send-email-pshelar@nicira.com>
On Wed, 26 Jun 2013 13:13:29 -0700
Pravin B Shelar <pshelar@nicira.com> wrote:
> vxlan private per net object can be accessed in device unregister.
> therefore unregister pernet device at the end.
>
> Signed-off-by: Pravin B Shelar <pshelar@nicira.com>
> ---
> drivers/net/vxlan.c | 2 +-
> 1 files changed, 1 insertions(+), 1 deletions(-)
>
> diff --git a/drivers/net/vxlan.c b/drivers/net/vxlan.c
> index 227b54a..0ba1e7e 100644
> --- a/drivers/net/vxlan.c
> +++ b/drivers/net/vxlan.c
> @@ -1916,9 +1916,9 @@ late_initcall(vxlan_init_module);
>
> static void __exit vxlan_cleanup_module(void)
> {
> - unregister_pernet_device(&vxlan_net_ops);
> rtnl_link_unregister(&vxlan_link_ops);
> destroy_workqueue(vxlan_wq);
> + unregister_pernet_device(&vxlan_net_ops);
> rcu_barrier();
> }
> module_exit(vxlan_cleanup_module);
No. This won't work.
The problem is that:
rtnl_link_unregister calls __rtnl_kill_links which calls
vxlan_dellink
vxlan_dellink assumes link is last thing and that the vxlan
device is already down. vxlan_dellink will unregister the device
and schedule work queue for closing socket.
When you call unregister_pernet_device last, nothing happens.
Since all vxlan device's are unregistered, the per-net vxlan
device list would be empty. And since vxlan_stop was never called.
The age_timer would never be stopped, and the multicast group
would never left, and the forwarding table will leak memory.
commit b7153984074e51a50dad905871b705e0d67aa147
Author: Stephen Hemminger <stephen@networkplumber.org>
Date: Mon Jun 17 14:16:09 2013 -0700
vxlan: fix out of order operation on module removal
If vxlan is removed with active vxlan's it would crash because
rtnl_link_unregister (which calls vxlan_dellink), was invoked
before unregister_pernet_device (which calls vxlan_stop).
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
diff --git a/drivers/net/vxlan.c b/drivers/net/vxlan.c
index 284c6c0..d3005d3 100644
--- a/drivers/net/vxlan.c
+++ b/drivers/net/vxlan.c
@@ -1771,8 +1771,8 @@ late_initcall(vxlan_init_module);
static void __exit vxlan_cleanup_module(void)
{
- rtnl_link_unregister(&vxlan_link_ops);
unregister_pernet_device(&vxlan_net_ops);
+ rtnl_link_unregister(&vxlan_link_ops);
rcu_barrier();
}
module_exit(vxlan_cleanup_module);
next prev parent reply other threads:[~2013-06-26 21:48 UTC|newest]
Thread overview: 3+ messages / expand[flat|nested] mbox.gz Atom feed top
2013-06-26 20:13 [PATCH vxlan v2 1/8] vxlan: Fix module cleanup Pravin B Shelar
2013-06-26 21:48 ` Stephen Hemminger [this message]
2013-06-27 5:15 ` Pravin Shelar
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=20130626144840.05ff8c80@nehalam.linuxnetplumber.net \
--to=stephen@networkplumber.org \
--cc=dev@openvswitch.org \
--cc=netdev@vger.kernel.org \
--cc=pshelar@nicira.com \
/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).