QFE/HME merge [PSARC/2007/404 FastTrack timeout 07/18/2007]

[Garrett D'Amore]

Randy Fishel wrote:
>>>> Adding such support is not too hard, but it isn't trivial
>>>> either.  Right now only you and Darren Reed have expressed interest.
>>>> (Assuming you are expressing interest?)
>>>>         
>>> I am expressing interest.
>>>       
>> Understood.
>>
>>     -- Garrett
>>     
>>> Irek
>>>       
>
>
>   The common case is that unless there is an significant reason for not
> supporting a driver on all platforms (i.e. nge is a network driver for
> chipsets _only_ found on x86 motherboards), then a driver must support
> all platforms.
>
>   So if this hardware is available on both Sparc and x86, then
> architectual correctness indicates that it must be supported on those
> platforms.  Phased delivery is between the implementor and the c-team
> the driver delivers into.
>   

I would tend to agree, but there are considerations as well.

1) the hardware was only ever sold into SPARC markets, never to x86 markets
2) the hardware has long since stopped shipping
3) because of the inability to evaluate FCode, we cannot distinguish 
between qfe and hme boards on x86
4) which means that there may be different driver names used on x86 and 
SPARC.

Anyway, we've already agreed (at PSARC) that the issue of x86 delivery 
is out of scope for this case, and may be handled by follow on work.

    -- Garrett
>
> 	---- Randy
>