Linksys EG1032 rev. 3 patch

[Fredrik Lindberg]

Benjamin Lutz wrote:
>>>Sounds like the if_sk driver needs to be made more intelligent so that
>>>it doesn't try to claim this card, yes?
>>>
>>
>>Something like this...
> 
> 
> Shouldn't this be applied to the re and maybe the nge driver as well
> then? To prevent them from trying to attach to the other revisions?

Yes, you are probably right. I realized this too, a few moments after
I sent my previous mail.

> The revision of the EG1032 that uses the Realtek chip is 0x10, according
> to pciconf. (Full output for completeness' sake:
> 
> re0 at pci2:10:0:  class=0x020000 card=0x00241737 chip=0x10321737 rev=0x10
> hdr=0x00
>     vendor   = 'Linksys'
>     device   = 'EG1032 Gigabit Ethernet'
>     class    = network
>     subclass = ethernet
> )
> 

I don't think we need to take nge into account since it only
probes for NGE_VENDORID (0x100B) and NGE_DEVICEID (0x0022).

rev 3 0x10, should be claimed by re
rev 2(?) 0x12, should be claimed by sk
rev 1(?), claimed by nge(4)

	Fredrik