domain validation

[Scott Long]

Matt Jacob wrote:
>>From the point of view of how the probe code is currently structured, it
> seems that this is the natural place to insert DOMAIN VALIDATION.
> 
> That is, don't make it a function of each HBA, but rather that of CAM
> when it is doing device attaching.
> 
> Thoughts?

Correct, it needs to be in CAM.  The optimal place in the the probe 
device.  Gibbs and I wrote a DV state machine for the linux Adaptec
drivers that I've wanted to port back but have never had time to do.

Scott