Re: PRO/100 VE on thinkpad R40 is missing PHY - no MII?

On Mon, 14 Apr 2003, Koukopoulos Konstantinos wrote:

> I just bought a thinkpad R40, and after updating the default
> eepro100 driver in debian woody to a newer version that supports the
> Intel PRO/100 VE card it seems that it's working OK.  Except, it seems
> that either there is no MII transceiver on this card or the driver can't
> find it properly.  This is a problem because I wish to use a program
> (laptop-net) which must monitor the link status through MII.

What version of the driver are you using?

>    eth0: Intel Corp. 82801BD PRO/100 VE (MOB) Ethernet Controller,
>    00:06:1B:C0:8A:95, IRQ 11.
>      Board assembly a43917-002, Physical connectors present:
>      Primary interface chip None PHY #0.

This is flawed: there should be both a transceiver and connector type
programmed into the configuration EEPROM.

>    root@panic:~# eepro100-diag 
>    eepro100-diag.c:v2.09 7/15/2002 Donald Becker ([email protected])
>     http://www.scyld.com/diag/index.html
>     Unable to find a recognized card in /proc/pci.

What is the output of
  lspci -n

>    root@panic:~# eepro100-diag -p 0x8000 -f -aaeemm 
>    eepro100-diag.c:v2.09 7/15/2002 Donald Becker ([email protected])
>     http://www.scyld.com/diag/index.html
>    Assuming a Intel i82557/8/9 EtherExpressPro100 adapter at 0x8000.
>    i82557 chip registers at 0x8000:
>      0c000050 0f516000 00000000 00080002 1801ffff 00000600
...
>    EEPROM contents, size 64x16:
>        00: 0600 c01b 958a 0000 0000 0000 0000 0000

Hmmm, it appears that this EEPROM was not fully programmed.  The
transceiver type and connector type should be at offset 5, 6 and 7.  But
these are all left as zero.  It appears the laptop manufacturer used the
Intel example programming and ignored the "insert correct value here"
part.

>       Sleep mode is enabled.  This is not recommended.
>       Under high load the card may not respond to
>       PCI requests, and thus cause a master abort.
>       To clear sleep mode use the '-G 0 -w -w -f' options.

There is a chance that this chip might recent enough that it has fixed
the sleep mode bug.  But I suspect not.

It also appears that you are not using a recent version of my driver
release, or the warning would have appeared when you loaded the driver.
I've just added a new warning message to 'eepro100.c:v1.27b 4/14/2003'
to catch this flaw.

-- 
Donald Becker				[email protected]
Scyld Computing Corporation		http://www.scyld.com
410 Severn Ave. Suite 210		Scyld Beowulf cluster system
Annapolis MD 21403			410-990-9993