Re: 2.0.6 Networking strangeness
Carsten Paeth (calle@calle.in-berlin.de)
Wed, 17 Jul 1996 01:53:34 +0200 (MET DST)
>
>
>
> On Mon, 15 Jul 1996 tytso@MIT.EDU wrote:
> >
> > Date: Sun, 14 Jul 1996 13:37:22 +0300 (EET DST)
> > From: Linus Torvalds <torvalds@cs.helsinki.fi>
> >
> > However, this behaviour makes me suspect more and more that there is
> > something broken in the PPP (or serial code) timings,
> >
> > Well, there is a known problem in the tty layer, which is that upper
> > layer processing doesn't happen except on the next clock tick. Hence,
> > there is on average a .005 second delay between when the PPP framing
> > character comes in and when it actually gets processed.
>
> I know about this, and I consider it reasonable - it does mean that
> latency for serial lines are about 10ms regardless of the speed of the
> line, but I don't _really_ consider this a problem (the only place it
> shows up is "ping", because TCP should be able to handle it with the
> windows anyway for any reasonable speed for the serial line).
I'm not sure, what will happen with faster lines.
I'm will get syncron-modems with a converter to async next month,
to have a 128Kbit connection over an analog line.
(Sure not using normal serial card ...)
I will report.
>
> We might want to fix it some day, but that one shouldn't have been the
> cause of this (when the delays were on the order of hundreds of ms,
> rather than 10ms). And as it turns out, it seems the problem was a false
> alarm anyway.
>
> Linus
>
calle
--
calle@calle.in-berlin.de