Steven Stallion wrote:
> Garrett/All,
> 
> There is an issue with dnet(7D) in which it does not select the MII PHY
> on a DEC 21143 (the faulty behavior seems to be isolated in
> find_active_media). This is making the remainder of the conversion very
> difficult (i.e. not being able to rely on mii soft state for link
> notification).
> 
> I'd like to put together a separate patch for this problem since it is
> fairly limited in scope and really has nothing to do with the conversion
> itself. As it is, the conversion patch affects nearly 1,000 lines
> (roughly 20% of the overall source).
> 
> (FWIW, I've already submitted a bug report to bugs under driver/dnet)
> 
> Thoughts?
> 
> Steve
> 

After digging into this in more detail, it seems that the SROM for the
21143 only returns 2 different block types: SIA and SYM. However, there
 there are 4 blocks provided (types 2 and 4 are repeated; the sequence
is: 2 -> 2 -> 4 -> 4, which indicates the SYM block as the highest
priority).

This seems odd... the DEC SROM Specification seems to imply that a 21143
would support MII blocks as well. Should I perhaps look into updating
the dnet code to better support SYM PHY's?

I am still double checking the SROM parsing code, but so far it appears
to be correct.

Any ideas ?

Steve

-- 
Yet magic and hierarchy
arise from the same source,
and this source has a null pointer.

Reference the NULL within NULL,
it is the gateway to all wizardry.
_______________________________________________
driver-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/driver-discuss

Reply via email to