According to Joerg Reuter: While burning my CPU.
>
> Hi Richard,
Hello Joerg.
>
> > It would seem buffers get overloaded, or something to that effect, ie, =
> the
> > 9k6 interface works twice as fast as the data can be retransimited on t=
> he
> > 4k8 interface causing buffer/memory problems. =
>
>
> Assuming that you are using either z8530drv-2.4c in its network
> driver mode or z8530drv-3.0 it is most likely not a buffer problem.
> What seems to happen is that the driver looses not only one TX
> interrupt but also the "exception" interrupt that notifies the
> driver of the TX underrun. Even worse, it seems that in some cases
> (probably depending on the mask version) the Z8530 needs a complete
> re-initialization of that channel to be able to transmit again.
I use 2.4c on all of my machines, i dont have 8530's i use 85230's.
I dont think they are Zilog either, if need be i could extracht the baord
and look, the computer is not easy to get at in my shack, my shack is a
typical shack filled with what some will call "junk", i never throw
anything away ;-)
> I'd recommend a DMA capable SCC card instead.
Yes times are changing, but i am stuck with the old one, my motto is, if
it works why break it.
>
> Seems that I'll have to add yet another timer to monitor transmit
> interrupts, sigh.
That would be "highly appricated Joreg".
>
> Yes, but messages with little more information than "it does not work und=
> er
> high load" don't help much when trying to find the exact condition
> causing it.
That i agree with, but when there is absolutly "no sign" of why it happened
then its dam hard to say anything more.
Considering what you say about interrupts, i could write a small script to
monitor TxUnderruns, then when the interface stops working i would then see
if the count increased or not, is that a good idea >?? while i'm at it, i
will include NoSpace errors as well.
>
> > Joreg DK1BKE has never responded to any of the comments, hints or kinks=
> ,
> > made about this problem, =
>
>
> This is not true. I've responded to several of your mails, asked for
> more information and never got any answers or replies afterwards. Seems
> that the path to your mailbox is not very reliable.
Well maybe my expression was a little overzellous, we did have good
communications during the, using "soft_dcd" disscussion, which i think was
the last time you replyed to mails sent by me and others with suggestions
about the interface stopping transmitting.
I did say that the soft_dcd problem helpt one machine i have here, but not
another because soft_dcd is not used there.
I received no more mails from you after that, that is what i meant with my
comment.
> Has it ever occured to you that the fault may not be a more or less
> "buggy" SCC driver but rather broken SCC board design equiped with
> broken second-source versions of Zilog's Z8530 on a broken PC
> architecture?
Yes, but an argument to that one is, the problem does not show up on dos
machines running with "net" written by Rob Jansen pe1chl.
> =2E..and perhaps pass me copy when this happens, with the version of
> the z8530drv you are using and the output of cat /proc/net/z8530drv.
That has been noted here.
Thanks for your time.
>
> 73,
>
> Joerg Reuter http://poboxes.com/jreuter/
> And I make my way to where the warm scent of soil fills the evening air. =
>
> Everything is waiting quietly out there.... (Anne Clark)
--
Regards Richard.
[EMAIL PROTECTED]
