[vortex] 3c556 problems (3Com Mini 10/100 PCI)

[Donald Becker]

On Sun, 21 Oct 2001, Tim Miller wrote:

> On Sun, Oct 21, 2001 at 07:23:19AM -0400, Donald Becker wrote:
> 
> > Is the transceiver always detected at address #0?  Or only when the
> > driver fails?  (#0 might be correct for your card, but that address
...
> Apparantly not.  If I'm reading this right, it thinks it's at #24 at
> boot, and can't find it.  (Ref. attached 2.2.19-failed).

The labels for the attachments didn't come through.

The important information is the driver detection message when the
driver fails, and the detection message when it works.

> > Does
> >   mii-diag --Reset
> > restore operation?
> No.  (Ref. attached 2.2.19-reset-attempt.)

OK, then it's likely a card activation problem, rather than a
transceiver reset issue.

> > [Re; 2.4.x operation]
> > Interrupt problem.  No interrupts are getting through.  This is a kernel
> 
> I'll buy that.  However, I've crawled through the system, and there's nothing
> that could be preempting that interrupt as far as I can tell.  And it begs
> the question of why does the interrupt work under 2.2.19?

2.2.19 uses the external, separately tested PCMCIA package from David
Hinds.
    
2.4.* uses PCMCIA/CardBus control built into the kernel, and tied to
that specific kernel version.  The "stable" kernels are used for
experimenting with major structural changes, API changes, as well as
minor bug fixes.  The likelyhood of all device drivers and the core
services both working at once is much reduced.

Donald Becker				becker@scyld.com
Scyld Computing Corporation		http://www.scyld.com
410 Severn Ave. Suite 210		Second Generation Beowulf Clusters
Annapolis MD 21403			410-990-9993