Garrett D'Amore wrote:
>> I'm thinking that m_setprop may end up being fairly lazy - from what I
>> have seen so far dnet will probably not respond very well to dynamic
>> updates without a full reset/init. As it is, I do not think changing the
>> 'default' properties warrant completely interrupting chip state 
>> without at
>> least downing the interface first. (FWIW, I believe the Jumbo Frames 
>> study
>> done in the Brussels spec did something similiar, i.e. lazy
>> changes/waiting for the next reset/sync)
>>   
> 
> Uh, no.  You should make MII changes take effect immediately.
> 

I should have been way more specific :)

MII changes definitely need to take affect immediately.

I was referring to the handful of private properties dnet currently 
supports (i.e. sromdump). I didnt really see the advantage to 
interrupting state for a trivial set of properties.

-- 
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