netdev - Re: [PATCH net-next 00/15] Simplify netfilter and network namespaces

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: <87wpz25a2q.fsf@x220.int.ebiederm.org>
Date:	Wed, 17 Jun 2015 15:55:25 -0500
From:	ebiederm@...ssion.com (Eric W. Biederman)
To:	Julian Anastasov <ja@....bg>
Cc:	David Miller <davem@...emloft.net>, netdev@...r.kernel.org,
	netfilter-devel@...r.kernel.org,
	Stephen Hemminger <stephen@...workplumber.org>,
	Juanjo Ciarlante <jjciarla@...z.uncu.edu.ar>,
	Wensong Zhang <wensong@...ux-vs.org>,
	Simon Horman <horms@...ge.net.au>,
	Pablo Neira Ayuso <pablo@...filter.org>,
	Patrick McHardy <kaber@...sh.net>,
	Jozsef Kadlecsik <kadlec@...ckhole.kfki.hu>,
	Jamal Hadi Salim <jhs@...atatu.com>,
	Steffen Klassert <steffen.klassert@...unet.com>,
	Herbert Xu <herbert@...dor.apana.org.au>
Subject: Re: [PATCH net-next 00/15] Simplify netfilter and network namespaces

Julian Anastasov <ja@....bg> writes:

> 	Hello,
>
> On Sun, 14 Jun 2015, Eric W. Biederman wrote:
>
>> This patshset roots out all of the very weird network namespace
>> computation logic (except for the code in ipvs) and fixes it.  I really
>> don't like how the code has been essentially guessing  which network
>> namespace to use.
>> 
>> Probably the worst guessing is in ipvs in the function skb_net. I have
>> some preliminary changes to fix ipvs but they are not quite ready yet.
>> Cleaning up ipvs enough that I can kill skb_net is on my short list.
>
> 	For IPVS skb_net is too complicated. One of
> the first things we do in hook handler is to check
> skb_dst, so even now dev_net(skb_dst(skb)->dev) should
> work.

A couple of things:
- You don't check skb_dst() first thing in all hooks.

- All hooks have state->net after my changes so it gets
  a lot easier.

- I am trying to avoid dev_net(skb_dst(skb)->dev) if I can
  because of my ulterior motive of allowing a destination
  network device in another network namespace.

>  sock_net is used for administration via netlink,
> skb_net is not used there (ip_vs_ctl.c).

Agreed.  sock_net is the right thing for the netlink sockets,
and while obscured by skb_sknet that works fine.

> 	As for removing/replacing skb_net, ip_vs_conn_net()
> is ok when cp is present,

Yes I have seen that.

> we have also svc->net,

And I have seen this.

>  otherwise
> we can store net into struct ip_vs_iphdr, it is filled by
> ip_vs_fill_iph_skb from small number of places when hook
> handler is entered.

Interesting I had not thought of changing ip_vs_iphdr.

I actually have a patch that passes state->net down to where it is
needed or uses ip_vs_conn_net() and it it looks ok, but the patch
is so busy with small tweaks that I need to break it up.

Further I am staring at things and trying to see if there is a way to
make the mass of parameters more readable/obvious.   Something like
the passing of struct nf_hook_state into the netfilter hooks.

Or in other words I want to make the change to ip_vs deadly boring,
obviously correct, and trivial to maintain, and I am not quite there
yet.

Eric

p.s.  I do have my patch that I can toss in your direction if you are
interested.
--
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