Re: Command 80 was not immediately accepted, 10001 ticks!

On Tue, 2 Dec 2003, Mark Adams wrote:

> I'm having a problem with the ethernet interface in a Dell PowerEdge 1550.
> 
> It has two 10/100 ethernet ports, using the eepro100 driver (latest
> from scyld.com) on RedHat 7.1, with kernel 2.4.20-20.7.
> 
> I upgraded to the latest driver after having problems that gave me lots
> of "wait_for_cmd_done timeout!" errors after a day or so of uptime.
> 
> I don't see that error any more, but I now have, after a few days:
> 
> Nov 29 13:10:37 kernel: eth0: Command 80 was not immediately accepted, 10001 ticks!
> Nov 29 13:10:44 kernel: eth0: Transmit timed out: status 0050  0080 at 39064/39066 commands 000c0000 400c0000 000ca000.

This is the driver trying to restart the chip, but the chip firmware is
now confused.  There might be some useful information from the very
first problem message.

The usual cause of this problem is an incorrect motherboard chipset
configuration.  The solution is a BIOS update.

> pci-scan.c:v1.12 7/30/2003  Donald Becker <[email protected]> http://www.scyld.com/linux/drivers.html
> eepro100.c:v1.28 7/22/2003 Donald Becker <[email protected]>
>   http://www.scyld.com/network/eepro100.html
> divert: allocating divert_blk for eth0
> eth0: OEM Intel i82559 rev 8 at 0xd089f000, 00:06:5B:3D:28:B8, IRQ 11.
>   Board assembly 02d484-000, Physical connectors present: RJ45
>   Primary interface chip i82555 PHY #1.
>   General self-test: passed.
>   Serial sub-system self-test: passed.
>   Internal registers self-test: passed.
>   ROM checksum self-test: passed (0x04f4518b).

Thanks -- This confirms that sleep mode is not configured, which causes
a similar problem (although the bus transaction screw-up is then the
fault of the eepro100 chip rather than triggered by the motherboard chipset).

-- 
Donald Becker				[email protected]
Scyld Computing Corporation		http://www.scyld.com
914 Bay Ridge Road, Suite 220		Scyld Beowulf cluster system
Annapolis MD 21403			410-990-9993