netdev - Re: namespace support requires network modules to say "GPL"

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <4755A025.8090903@candelatech.com>
Date:	Tue, 04 Dec 2007 10:44:53 -0800
From:	Ben Greear <greearb@...delatech.com>
To:	"Eric W. Biederman" <ebiederm@...ssion.com>
CC:	Daniel Lezcano <daniel.lezcano@...e.fr>, netdev@...r.kernel.org,
	linux-kernel@...r.kernel.org, containers@...ts.osdl.org,
	Mark Lord <lkml@....ca>,
	Stephen Hemminger <shemminger@...ux-foundation.org>,
	David Miller <davem@...emloft.net>
Subject: Re: namespace support requires network modules to say "GPL"

Eric W. Biederman wrote:
> However there also seem to be simpler cases like Ben's bridge module,
> that don't appear to have any global state.
>   
Well, my module has some global state, but I don't think it needs to 
care about
namespaces.  My first impression is that my module should be able to bridge
namespaces...not be contained within one.   I can have user-space make 
sure that I don't bridge between
devices in different name-spaces, or perhaps bridging between namespaces
wouldn't be a problem anyway.  If I *do* need to add some sort of namespace
awareness to just achieve today's functionality, I don't mind making the 
changes,
so long as I don't need to change to GPL licensing.  Perhaps at the 
least you
can export enough symbols w/out GPL tag to achieve backwards compat with .23
and previous kernels, or rework dev_get_by_* etc to not need GPL'd namespace
symbols and just return the device in the default namespace?
> Ben I don't have a clue how your user space interface works.  My gut
> feel is that you can likely use sk->sk_net (if your configuration is
> through a socket), or failing that current->nsproxy->net_ns.  To get
> the network namespace to look up "eth0" and "eth1".
>   
Currently I use procfs and ioctls bound to a procfs file descriptor.

For namespaces in general, will there be a way to just do a dev_get_by_* 
and find the
device in *any* namespace and query the device to see what namespace it 
is in?
Then my module or some other more clever piece of code can determine the 
namespaces
(by comparing pointers if nothing else) and make proper decision.  For 
instance, maybe
we want to bridge two namespaces, or maybe we want to forbid that ever 
happening...
> This however still begs the question how do we want to handle this
> so there is a minimum of pain.
>
> Since using register_pernet_subsys implies you need your own member
> in struct net.  I am inclined to leave that with the GPL hint on
> the EXPORT as you need to be really tight with the system to use that.
>   
I certainly don't want to have to muck with struct net unless you have 
some way to
register anonymous (and non GPL) subsystems.  I'm guessing you do not 
want to
support that....

Thanks,
Ben

-- 
Ben Greear <greearb@...delatech.com> 
Candela Technologies Inc  http://www.candelatech.com


--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html