[vortex] The mysteries of on-chip MII.

Bogdan Costescu bogdan.costescu@iwr.uni-heidelberg.de
Fri Mar 1 13:32:11 2002


>From the loud silence, you might understand that there is little hope of 
getting past the current state. Some comments inlined.

On Mon, 25 Feb 2002, Bill Cattey wrote:

> 1. Live with the fact that we have chips that will come from the factory
> with EEPROM defaults that cause our drivers to be unable to properly
> initialize the MII.

I would say that this is actually the best solution. It's known to work 
now and Autonegotiate/NWAY is the recommended (by Donald and me, at least) 
media setting.

> 2. Contact whatever 3Com liaison negotiated their generous disclosures
> of the past...

I have nothing like this. The one and only documentation is the PDF that 
3Com made available on their web site (after free registration as a 

> 3. Attempt to guess at an interpretation of the documentation that
> resolves this case.

I'm gone for a week and I'll take the documentation with me. I'll report 
back to the list when I'm back if I find something of interest; however, I 
already passed through it some time ago and found the initialization 
sequence from the driver to not do anything wrong.

> With regards to #3 I ask what may be a naive question:  What kind of
> reset sequence might one need if one first needed to verify that there
> were no EXTERNAL chips active for MII? 

Reset is one thing, finding MII transceivers is another. The driver does 
scan all the possible MII addresses, looking for valid data. The first 
found is then used, as the driver can only handle one MII transceiver. For 
Cyclone and later chips, 3Com recommends to start scanning at address 24, 
where the on-chip transceiver should be found; we found out that this is 
imperatively necessary (as opposed to recommended) for the CX cards. Given 
these, when using Cyclone and later chips, the internal transceiver is the 
only one which can be used by the driver - if it's active and recognized 
as such during the scan - no external transceivers could be used.

However, this has nothing to do with your problem. The transceiver that is 
not activated in your case is the _internal_ one, the one which should 
always be found at address 24. Even if the scan would be eliminated and 
the transceiver at address 24 would always be used, it would not help, as 
it's not active and the driver (and diag tools) cannot communicate with 

Bogdan Costescu

IWR - Interdisziplinaeres Zentrum fuer Wissenschaftliches Rechnen
Universitaet Heidelberg, INF 368, D-69120 Heidelberg, GERMANY
Telephone: +49 6221 54 8869, Telefax: +49 6221 54 8868
E-mail: Bogdan.Costescu@IWR.Uni-Heidelberg.De