From: Marc G Fournier (marcf@nexus.yorku.ca)
Date: 01/02/93


From: marcf@nexus.yorku.ca (Marc G Fournier)
Subject: Re: Appendum to problems with AST board
Date: Sat, 2 Jan 1993 17:27:17 GMT

tytso@ATHENA.MIT.EDU (Theodore Ts'o) writes:

> From: marcf@nexus.yorku.ca (Marc G Fournier)
> Date: Sat, 2 Jan 1993 03:50:08 GMT

> When I boot up the system, the kernel properly finds the
> AST board, except it 'recognizes' them as ttys[8-11], which
> do not exist. I figured that I had to mknod them, so I tried it
> out, but it didn't seem to work. After thinking about it, over
> dinner, it struck me like lightening...it is 'half' working...I'm
> getting communications in one direction, but not the other. Is
> that possible? It is sending out /etc/issue and login prompts,
> but isn't allowing the other end to enter anything.

>Sounds like interrupts to the FourPort board aren't working right. If
>you don't have AUTO_IRQ (the automatic IRQ detection code) enabled, then
>you either need to jumper your FourPort card for IRQ 5, or you need to
>modify serial.c to use the proper IRQ for your card, or you need to use
>the setserial program to let the kernel know what IRQ the card is using.

        This is basically what I found out. I did have AUTO_IRQ enabled
in the kernel, but for some reason, it wasn't picking up the IRQ correctly.
I had the IRQ set at 2, with the 'addresses' set at 2a0->2b8(?), but
the IRQ that the kernel was finding was 5 :( What I ended up doing was
just changing the IRQ to 5, and I seem to be having at least hard-wired
communications.

        Now, to get the modems to work :)

        Thanks loads...

Marc