sowmini.varadhan at sun.com wrote:
> On (02/08/08 11:45), Garrett D'Amore wrote:
>
>> *HOWEVER*, I agree that we want to discourage use of ndd for device driver
>> tuning. THEREFORE, I recommend the following strategies:
>>
>
> another option that I thought of is that we could add register a stub entry
> for legacy drivers (if needed, I think every driver that uses ndd has
> at least 1 outlier, so coming up with a stub registration will
> not be a problem), and only do the mapping if the driver has a stub
> entry registered.
>
I know of several drivers that have no outliers. I wrote two of them. :-)
What I really think, is that these should just be converted to Brussels
private properties. I hate continuing to carry around any more legacy
baggage than absolutely necessary for NDD.
As a compromise, perhaps the Brussels code can just pass the IOCTL down
to the driver if it can't find either a public or a private property
matching the name?
-- Garrett
