[realtek] Problems with GF100TXII & V1.11

Martin Hillier martin.hillier@sykes.demon.co.uk
Wed, 9 Aug 2000 09:43:01 +0100

I have found out the problem, the pci slot is duffy on the motherboard!.

Changed the nic to another slot and it worked, for a bit. Now I have a
different problem.

Kernel message (fault/error etc) in swapper task (something like this)

and the machine hangs, after transferring a few meg. I am not sure what this
is related to, network or scsi drive? ( the new card is a 100meg and I know
there has been problems with the scsi on this machine before)



----- Original Message -----
From: Donald Becker <becker@scyld.com>
To: Martin Hillier <martin.hillier@sykes.demon.co.uk>
Cc: <realtek@scyld.com>
Sent: Tuesday, August 08, 2000 3:29 PM
Subject: Re: [realtek] Problems with GF100TXII & V1.11

> On Tue, 8 Aug 2000, Martin Hillier wrote:
> > I have set up a second nic on a machine with kernel 2.0.33, but as soon
as I
> > plug the card into the network I get lots of
> >
> > Abnormal interrupt, status 0xffff
> >
> > with the debug level is set to 4, and the machine locks up (keeps
> > the kernel message as above)
> The chip doesn't exist in I/O space.
> Run 'rtl8139-diag' to verify this.
>   http://www.scyld.com/diag/index.html
> The driver does handle this problem gracefully -- the CardBus version
> disappear before the driver is notified of the eject.  But in your case
> other device (or a system error) is repeatedly raising an interrupt that
> is not cleared.
> > as far as I can tell the ifconfig and route is correct now and cabling
is ok
> This is unrelated to IP routing or your cables.
> It might be as simple as the card not being fully seated in the PCI slot.
> Donald Becker becker@scyld.com
> Scyld Computing Corporation http://www.scyld.com
> 410 Severn Ave. Suite 210 Beowulf-II Cluster Distribution
> Annapolis MD 21403