If you've got lots of ge cards, and are willing to part with two, I
might be willing to work on it later. (Two, because I don't have any
fiber gear at home. I guess I'd also need a cable...)
It has the same MAC controller as eri, and its likely that the same kind
of work that made qfe possible could be done for ge as well.
Cassini is a non-starter though.
-- Garrett
Volker A. Brandt wrote:
> Hi Garrett, hi James!
>
>
> Garrett D'Amore writes:
>
>> As soon as my current workload settles, I've agreed to convert hme/qfe,
>> as well as several other drivers (eri, afe, mxfe) to Brussels. Its
>> likely to be about a month before I dedicate the time this project needs.
>>
>
> You da man! [Apart from top-posting, that is :-)]
>
>
>
>> James Carlson wrote:
>>
>>> In the current Nevada code, hme and qfe are supported by the same
>>> GLDv3 driver. Stepping up to Brussels ought to be fairly easy as long
>>> as someone interested in doing the work appears.
>>>
>
> This sounds great! (See above :-)
>
>
>>> The ce and ge drivers, though, are in the old NSPG/NSN/CS
>>> consolidation, which isn't open source, and which is unlikely to get
>>> attention. There's a project running to produce a new open driver for
>>> Cassini hardware (likely leaving only ge and gem out in the cold), but
>>> I don't know the current status of that. (Given the special
>>> interactions between 'ce' and Sun Cluster and other odd features, I
>>> wouldn't be surprised to hear that an open driver would not be usable
>>> on a regular Sun distribution. But that's just my guess.)
>>>
>
> Hmmm.... this means there is some likelihood for a schism like
> either "S10 + old ce + Solaris Cluster + no Brussels"
> or "OpenSolaris + new ce + OpenHA + Brussels"?
>
> Well, better that than nothing I suppose. The part about ge in the
> cold made all my nice ge cards on the shelf cringe and squeak though.
>
>
>>>> Also, you define build 88 to be the next version of OpenSolaris.
>>>> I would just say "snv_88". In a little while, the "next version"
>>>> will confuse people.
>>>>
>>> I think that refers to the OpenSolaris Developer Preview release,
>>> rather than a particular Nevada build.
>>>
>
> I thought the authors just wanted to point out that snv_88 has not
> been released at the time of writing.
>
>
>>> I agree that having the code base and one particular distribution
>>> having the same name ("OpenSolaris") is really confusing, but there's
>>> not much I can do about that. ;-}
>>>
>
> Explaining confusion? No problem. That's what we top-notch high-tech
> consultants are for. :-)
>
>
> Regards -- Volker
>
