[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: Ethernet transceiver - PHY address



>The linux driver should be no problem, but I'm a bit worried about 
>network-booting.
>Where does the phy address come from in that case?

The phy address is only used for the mdio interface to manage
the transceiver. No such management is performaned by the on-chip
ROM and hence the phy address doesn't matter. (The MII interface
used by Ethernet data transfers is not controlled by the
phy address at all).

>I haven't connected it yet, but I read in the datasheet that it sets its 
>outputs to tristate when address 0 is used.

Which outputs? All? Seams stupid but probably they have some reason 
for it.

-----Original Message-----
From: Pieter Grimmerink [mailto:mailinglists@xxxxxxx.nl]
Sent: Monday, April 07, 2003 6:37 PM
To: Mikael Starvik; dev-etrax
Subject: Re: Ethernet transceiver - PHY address


On Monday 07 April 2003 15:56, Mikael Starvik wrote:
> The current software has phy address hardcoded to 0 but it is easy
> to modify the software to use any other value (or probe it).
> In arch/cris/drivers/ethernet.c you'll find the constant
> MDIO_HPYS_ADDR (or something like that).

The linux driver should be no problem, but I'm a bit worried about 
network-booting.
Where does the phy address come from in that case?

> What is the problem with ICS1893af? Do you get 0xffff when you
> read registers? Check tat the 20 MHz clock is running and that
> the reset signal to the transceiver is correct.

I haven't connected it yet, but I read in the datasheet that it sets its 
outputs to tristate when address 0 is used.

Best regards,

Pieter