[vortex] Mini PCI - no link

Dave Dribin dave-ml@dribin.org
Sat Jun 22 12:52:01 2002


On Sat, Jun 22, 2002 at 12:09:47PM -0400, Donald Becker wrote:
> > Can we use the fact that the card's transceiver comes up in Windows
> > (and stays up in Linux afterward) for anything?  Something like: dump
> > the registers with vortex-diag, reboot to Windows (to let the
> > transceiver come up), reboot back to Linux and dump the registers
> > again.  Then compare the before-Windows registers with the
> > after-Windows registers to see if we can find any differences.
> 
> Yes.  That's an excellent approach!
> It's possible that the changed register is in the MII management
> section, so dump those registers as well (the -m option to the diag
> program).

Ok, I'll try this out on Monday.  I don't have Windows on my laptop
(and I don't have any free disk space either. :)

> > I've also tried a co-workers Mini-PCI card in my machine just to make
> > sure my machine works.  It, too, gets identified as a 3c1556B, but it
> > is obviously different some how 'cuz it works.  :)
> 
> Very useful info!  We do have reports of the card working.  This
> confirms that there are multiple version of the card

Now that I think of it, when I tested the other card, I used the
kernel's 3c59x driver.  I know my co-worker uses the kernel driver,
too.  I'll re-test with your driver just to make sure we're comparing
apples to apples, but that will have to wait until Monday, too.

I'm guessing that the other card will only work with 0.99Xa due to the
EEPROM offset issue you resolved earlier (which is already resolved in
the kernel's driver).  I'll try 0.99X and 0.99Xa.

> The driver uses the table entry:
> 	{"3c1556B mini-PCI",{ 0x605610B7, 0xf0ffffff },
> which matches IDs 0x6*56 (0x6056 0x6156 .. 0x6f56).
> What are the IDs of the two cards?

Which IDs is it comparing?  Device and vendor or subsystem?  Here's
both for my card:

00:03.0 Class 0200: 10b7:6056 (rev 20)
        Subsystem: 10b7:6556
        Flags: bus master, medium devsel, latency 80, IRQ 9
        I/O ports at 1800
        Memory at e8101400 (32-bit, non-prefetchable)
        Memory at e8101000 (32-bit, non-prefetchable)
        Capabilities: [50] Power Management version 2

On Monday, I'll grab the IDs for the other card.  I think we did this
once (of course I didn't write down the results) and the subsystem IDs
were different.

Sweet, looks like we'll be in touch Monday, unless you get an "Ah-Ha!"
moment. :)  Have a good weekend.

-Dave