Re: NFNL_NFA_NEST

[Pablo Neira Ayuso <pablo@netfilter.org>]

Patrick McHardy wrote:
> Not really. The NEST bit allows to walk nested structures without
> being aware of the structure. So all you need to do is make it aware
> of which attributes are nested - something you need anyway for
> parsing I suppose.

OK, then I have two choices here, define a proxy function that is
structure aware to convert TL to network byte order that will be almost
a copy and paste of the parsing part, or alternatively integrate the
byte order conversion as an option for the build/parse functions, as for
now the latter seems more natural to me.

> The byteorder is already clear, all netlink header values are in
> host byte order, all attributes in network byte order.
> 
>> I'm fine with using the new netlink infrastructure. We could remove the
>> nested bit, release new libraries that don't send it to kernel anymore
>> and keep the old nfnetlink code that understand the nest bit thing for
>> quite some time. Thus, doing the port to the new infrastructure later.
> 
> Yes, we should probably wait for at least one year.

Fine with it. I'm about to release a new version of libnfnetlink, should
we stop sending the nested bit thing to kernel since now?

-- 
The dawn of the fourth age of Linux firewalling is coming; a time of
great struggle and heroic deeds -- J.Kadlecsik got inspired by J.Morris