Corrupt packets in Jnet (Was: Re: rtentry and rtrequest)

[Yar Tikhiy]

On Wed, Apr 25, 2007 at 07:37:06AM +1000, Peter Jeremy wrote:
> On 2007-Apr-23 18:54:30 +0400, Yar Tikhiy <yar at comp.chem.msu.su> wrote:
> >Perhaps the bug is triggered when the outgoing packet consists of
> >multiple mbufs.
> 
> Given that we are effectivly dealing with a shared memory block, how
> does the SP now when the server has finished writing and vice versa?

This is a pretty good question!  I don't remember seeing any
synchronization with the SP in jnet_start() so it cannot be
sure the SP has finished reading the previous packet.

> Is jnet's handling of multiple mbufs making the SP think there are
> multiple packets?

It shouldn't.  The data from all mbufs in the chain are just written
to the buffer to form the whole packet in it.

> >I'd also test if the ssh from SP can work OK with a FreeBSD host
> >(the same FreeBSD version as on the platform side would be the best)
> >via the external Ethernet.
> 
> It definitely can.  I haven't used anything else to talk to my SP.

I meant using ssh(1) on the SP and sshd(8) on an external FreeBSD
host just to be 100% sure that combination works, too.

-- 
Yar