[tulip] I/O region too small problem

[Donald Becker]

On Thu, 29 Nov 2001, David Annett wrote:
> I have recently installed Red Hat 7.2 on an IBM i1400 ThinkPad.  I
> have be unable to get the network card to go and I am unsure if the
> problem is with the PCMCIA side of things or the network card. 

PCMCIA control bug, not the driver.

> The card is identified by Windows as:
> 
> Linksys EtherFast 10/100 CardBus PC Card (PCMPC200)
> I/O FE80 - FEFF
> Mem FFEFFC00 - FFEFFFFF
> Irq 9

The I/O, memory map and (to a limited set) IRQ are decided by the OS, so
these settings are not directly useful.

> Nov 20 22:53:04 localhost kernel: Yenta IRQ list 08b8, PCI irq9

Yenta...

> Nov 20 22:53:05 localhost kernel: cs: cb_alloc(bus 2): vendor 0x1011, device 0x0019

It's a 21143, the tulip driver is correct.

> Nov 20 22:53:05 localhost kernel: PCI: Failed to allocate resource 0(1000-fff) for 02:00.0

The PCMCIA control code read the range incorrectly.
This was a known bug with the beta releases of RH 7.2.  Are you certain
that you are using a final release?  (I don't for certain that the bug
was fixed, but it was known well before release.)

> Nov 20 22:53:05 localhost kernel: tulip: 02:00.0: I/O region (0x0@0x1000) too small, aborting

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