[Devel] Re: [PATCH net-2.6.26 3/3][TUN][NETNS]: Allow to register tun devices in namespace.

Pavel Emelyanov xemul at openvz.org
Thu Apr 3 05:34:44 PDT 2008


Max Krasnyanskiy wrote:
> Hi Pavel,
> 
> All three patches look perfectly fine to me.
> Feel free to add Acked-by: Max Krasnyansky <maxk at qualcomm.com>

:)

> Pavel Emelyanov wrote:
>> This is basically means that a net is set for a new device, but
>> actually this involves two more steps:
>>
>> 1. mark the tun device as "local", i.e. do not allow for it to
>>    move across namespaces.
>>
>> This is done so, since tun device is most often associated to some
>> file (and thus to some process) and moving the device alone is not
>> valid while keeping the file and the process outside.
>>
>> 2. get the tun device's net when tun becomes attached and put one
>>    when it becomes detached.
>>
>> This is needed to handle the case when a task owning the tun dies,
>> but a files lives for some more time - in this case we must not
>> allow for net to be freed, since its exit hook will spoil that file's
>> private data by unregistering the tun from under tun_chr_close.
> I'm not sure what you mean "by file lives on" here. I believe you're talking

I meant the following. When a task dies it releases it's net namespace
and thus this net cleanup job may start. But the file, that holds the
(not yet detached) tun device is not released immediately - so before this
file is being released the tun_exit_net callback may be called and the 
file's tun device will be unregistered in it. I catch this race by getting
the net by attached tun device, so that net is cleaned up only when there
are only detached tuns in it. In this case no tasks will pick up these tuns
and everything will be OK.
 
> about persistent tun devices. ie Those that exist in detached state and are 
> not attached to any file descriptors.
> "net" refcounting logic there looks fine to me.

Great!

> Max

Thanks,
Pavel




More information about the Devel mailing list