On Fri, 19 Apr 2002, Allen Briggs wrote:

> On Fri, Apr 19, 2002 at 01:30:52PM -0500, Donald Lee wrote:
> > The flow control on ttya is unknown.
>
> Do you have 'local' in your line for that port in /etc/ttys?
> You should be able to determine what NetBSD thinks of the flow control
> with 'stty -a' after logging in.
>
> There are definitely some bugs in the serial code.  I've seen my
> serial console lock up when there's "too much" output (whatever
> that means).  So I've learned to not use the box from the console
> unless I need to.  It's been on my list for a LONG time to see if
> I can figure out what's wrong and fix it.  I don't think I've seen
> the mac68k serial console do this, and it's using the same serial
> chip and pretty much the same driver...

It's using the same driver, but not the same chip. The powerpc macs have
an "updated" version of the chip, whose features we don't take advantage
of.

My suspicion is that it's actually interupt lossage. I don't think our
interupt handling code deals with interupts comming in quick succession.
As Donald pointed out with the cyclades-Y card, there are windows where we
will drop interupts. I am not sure if this has been fixed yet, and the bug
certianly used to be there. The bug was that if an interupt came in while
we were servicing one on the same pin, we might drop it.

Take care,

Bill