[vortex] Two quick questions...

[Bogdan Costescu]

On Sun, 14 Aug 2005, Peter Szmrecsanyi wrote:

> 1. Does the vortex driver (the one that is included in the 2.6.11 source)
> accept kernel boot parameters?

I can't answer this question as I didn't pay attention for some time
to how the things are initialized when the kernel start. I don't use 
statically compiled network drivers for a few years...

> If so what would the equivalent be of "options=0x204",
> "ether=0x204,eth2"?

To specify options for more than one card (up to 8), it used to be 
something like:

ether=0x204,0x209,...

so you just need to put the options one after the other separated by 
commas and without any space.

> 2. Does the vortex driver support running the 3C597 (a.k.a. TCM5970) at
> 100Mbps? Or maybe at 10Mbps at FD? Is there a reason as to why "mii-tool"
> wont change the speed of an EISA device that uses a vortex driver built into
> the kernel?

I don't have documentation for these old cards, but from the driver I
understand that they might not use MII compatible transceivers, so
they need to be told to "force" the transceiver to a certain media
type (including duplex, if supported), by either using the options
(like you try) or by setting a default media in EEPROM (which you can
change using the vortex-diag tool or the DOS-based config utility from
3Com). "mii-tool" cannot work in this case, as it requires access to
the MII registers.

> My problem is with the 3C597 ... so it should auto detect

Not quite. Auto detection works well only for NWAY (autonegotiation) 
capable MII transceivers. It's also possible to do "autoselect" (which 
might work with the driver, but I don't remember to ever test it) 
which tries to "force" several media types until it finds one that 
works, but this is not so reliable.

> but ether=0x204,eth2 seems to have no affect!

I don't think that this is the right syntax, you should not specify an 
interface name (eth2), the options are passed based on their position 
(first options are passed to eth0, second to eth1, and so on). At 
least this is how it used to work, don't shoot me if it changed in the 
mean time...

-- 
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 at IWR.Uni-Heidelberg.De