On Mon, 24 Jul 2000, Andreas Wrede wrote:

> On Mon, 24 Jul 2000, Bill Studenmund wrote:
> 
> > I know it might not help much, but the cz serial cards work just fine. I
> > debugged a Cyclades-Z 16e on my G3. :-)
> 
> :-) I suspect the driver will work just fine once the interrupts get
> delivered from the card - I have never seem a single one. The card
> currently is an B&W G3, due to the problem with the PCI0646 ide
> controller chip initialization after the G4 firmware has had its hands
> on it.  The Cyclom8Y has the same problem in both the G4 and G3 - no
> interrupts. Modem signals are delivered properly.

I think the B&W G3 has the same ide problem. I'll assume that you just
also have a SCSI card too. :-)

> > > Is someone using a PCI Cyclades-Y board on G3 or G4 Powermac with
> > > NetBSD 1.4/5? If not, where should I start to look in the code to
> > > trace the missing interrupts?
> > 
> > syssrc/sys/dec/pci/cy_pci.c is the pci-specific code for the driver, and
> > syssrc/sys/dev/ic/cy.c is the bus-independent part of the driver.
> 
> I found these two. Where does PCI bus interrupt routing happen? Or is
> that to naive a question to ask?

I'm not sure. Though I'd not suspect interrupt routing problems at
first. To test, get another PCI card which will interrupt, and put it in
the slot. make sure it says it's using the same interrupt # as the cy card
(PCI cards have 4 interrupts each). If it works, you have verified the
interrupt and routing is fine.

> > > $  ls -l /dev/*tyCY00
> > > crw-------  1 uucp  wheel  47, 128 Jul 13 11:50 /dev/dtyCY00
> > > crw-------  1 uucp  wheel  47,   0 Jul 13 10:26 /dev/ttyCY00
> > 
> > You might also talk with tsubai, as he was the one who enabled cy
> > support. :-)
> 
> I Cc:'d tsubai, maybe he is not on port-macppc anymore?

He's the portmaster, so I'd surprised if he weren't around. :-)

> I compiled the cy driver with CY_DEBUG on. Can anyone find a clue as
> to why interrupts are not working at all??

No, but then I don't know much about this card. I suspect that the real
problem is that interrupts on the card aren't getting enabled right. I
can't figure out why, though.

Ask on teck-kern. I'm not too familiar with this driver, and I think the
folks who've worked on it will be there. :-)

Take care,

Bill