[tulip] PCI bus fatal errors with Linksys LNE100tx

Donald Becker becker@scyld.com
Thu, 21 Sep 2000 13:37:40 -0400 (EDT)


On Thu, 21 Sep 2000, Mario Simoni wrote:

> This did fix my problem and I now have my network card working. However,
> when I look at the /var/log/messages file I hundreds of lines of PCI Fatal
> Bus errors of different types as shown below.  This is the point in the
> file where it begins to give the error messages after booting up. After
> this there are hundreds more messages of the same types listed below.


> Sep 20 22:36:22 rivendel kernel: tulip.c:v0.92i 7/31/2000  Written by Donald Becker
> <becker@scyld.com>
> Sep 20 22:36:22 rivendel kernel:   http://www.scyld.com/network/tulip.html
> Sep 20 22:36:22 rivendel kernel: eth0: ADMtek Comet rev 17 at 0xc2852c00,
> 00:20:78:1E:53:F4, IRQ 11.
> Sep 20 22:36:22 rivendel kernel: eth0:  MII transceiver #1 config 1000 status 786d
> advertising 01e1.
> Sep 20 22:36:22 rivendel kernel: eth0: PCI Fatal Bus Error, fc66e010.
> Sep 20 22:36:22 rivendel kernel: eth0: PCI Fatal Bus Error, fc27e014.

These are Very Bad.
They indicate a hardware problem.

Try moving the card to a different PCI slot.

Check that your power supply is putting out the correct voltage.
Verify that your system is not overclocked.

> Occasionally I see the following messages mixed in with the PCI Fatal bus errors
> 
> Sep 20 22:50:01 rivendel kernel: eth0: Too much work during an interrupt,
> csr5=0xfc174014.
> Sep 20 22:50:01 rivendel kernel: eth0: Re-enabling interrupts, fc66c810

What else is using the PCI bus?

Donald Becker				becker@scyld.com
Scyld Computing Corporation		http://www.scyld.com
410 Severn Ave. Suite 210		Beowulf-II Cluster Distribution
Annapolis MD 21403