I don't believe the fix is appropriate; it's unsafe to recognize
controllers that are not using the right self-id information. That's why
those lines were added (because we know of such devices).
Thomas Tornblom wrote:
> CR 5031620 has a trivial fix, removing two source lines, which if
> implemented allows easy addition of new and unknown ATA controllers.
> One might even argue that is_pciide() has a bug that is unnecessarily
> stopping some adapters from working.
> Thomas Tornblom (Sun employee)
> request-sponsor mailing list
> request-sponsor at opensolaris.org