lwip-users
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[lwip-users] Re: [lwip] Gigabit Ethernet and LWIP


From: Adam Dunkels
Subject: [lwip-users] Re: [lwip] Gigabit Ethernet and LWIP
Date: Thu, 09 Jan 2003 01:33:36 -0000

On Wednesday 08 May 2002 12.56, you wrote:
> On Tue, 7 May 2002, Adam Dunkels wrote:
> > I see - but the stack and the network device driver doesn't have to run
> > in different threads. In fact, in small systems without support from a
> > multitasking operating system, the only way to do it is to have the stack
> > and the device driver in the same thread. Polling or interrupt driven
> > operation works equally well since there aren't any other threads using
> > the CPU so busy waiting can be used for the polling loop.
>
> Would be nice if it were that simple.  Unfortunately I'm using it in an
> operating system environment, so it has to cooperate with other processes,
> which rules out polling (pretty much).  Interrupts are the straightforward
> way of doing it, but involve just too much overhead.  The approach I'm
> taking is to run the stack at user level, access the NIC directly for data
> transfer, and so avoid interrupts altogether.  There's a lot more too it
> than that obviously, but you get the idea.

Yes, in such an environment there are additional problems that aren't present 
in the smaller environments. How do you solve the problem of accessing the 
NIC from userspace by the way? Is it possible to allow direct I/O access to 
user processes or have you been forced to change the kernel?

> > A socket interface for an application running in the same thread as the
> > stack - that's an interesting problem. I would presume that only works
> > well for communication-centric applications, which spends most of their
> > time waiting in read() or accept()
>
> Not necessarily - I have way's round this, but haven't tried them out yet
>
> :)
> :
> > I'm definately looking forward to hearing about your
> > work with this!
>
> Well, yes - me too!  It's forming my PhD research.

Ok :-)

> > (By the way, is the direct application<->NIC communication in any way
> > related to Van Jacobson's proposed 4.4BSD TCP/IP stack enhancements? His
> > slides are avaliable at ftp://ftp.ee.lbl.gov/talks/vj-nkarch.ps.Z - they
> > are a very interesting read!)
>
> I'm not sure - I'll take a look.  You can read about our network at
> http://www.uk.research.att.com/clan/ - there are publications at the
> bottom.  Unfortunately AT&T Labs Cambridge is now defunct, so the link
> above might not work for too much longer.  The TCP/IP described in the
> papers is an older kernel based traditional stack, although the most
> recent "Distributed Computing with the CLAN Network" does mention lwIP,
> and I think even cites one of you publications.

I printed your CLAN paper out a few days ago but haven't had the time to 
thuroughly read it yet, but it definately looks interesting! (I was extremely 
proud to see lwIP being mentioned, by the way ;-)

/adam
-- 
Adam Dunkels <address@hidden>
http://www.sics.se/~adam
[This message was sent through the lwip discussion list.]




reply via email to

[Prev in Thread] Current Thread [Next in Thread]