Re: [ECOS] Temporary IP address assignment trick

[Grant Edwards <grante at visi dot com>]

On Thu, Mar 20, 2003 at 04:20:55PM +0000, Jonathan Larmour wrote:

>> So there's no way to open a raw socket if the stack doesn't
>> have an IP address?
> 
> I can't see a way for a _raw_ socket to do that no... I don't
> even see raw sockets at all in this stack. I hope I'm wrong.

I don't actually know. IIRC, the old stack didn't support raw
sockets when I first started using 1.3.1, so I added my own raw
packet API to my Ethernet driver (which is rather feature-poor,
and in my app can't be used for this purpose since it's in use
for something else).

> The thing that I mean about the "not working transparently"
> thing is that once you eat that packet, it's gone, so it's not
> transparent - that's possible but would probably be difficult
> and invasive.

I'll have to do a network trace to see if the competition eats
that first packet or if it's handled transparently.  I had
always assumed the former.

> If you don't mind the packet being lost (e.g. wait for TCP SYN
> retry) then,

That would certainly be good enough for my indended use, which
is as a one-time manual setup operation.

> with some changes in net/common you should be able to do
> something very similar to what BOOTP does now in do_bootp(),
> but just without sending a request out first.

Thanks -- I'll take a look at it.

-- 
Grant Edwards
grante at visi dot com

-- 
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss