On Nov 13, Ken Hornstein wrote
> Well, I could make it conditional to a PCMCIA bus (would that violate
> some abstraction, though), or perhaps just put it under the
> NO_EXTRA_RESETS quirk, since this is the only device (so far) that
> we've seen that does this.

No, I'd prefer a separate flag (something like ISA_PROBE, or NON_ISA_PROBE)
as other busses may have the same problem.
The question is: is it better to have the MI part handle this flag, or
move this specific part of the specific part of the probe in the ISA
front-ends.
For now I think this test should be here for ISA, ISAPnP and
PCI front-end, at last, as well as some MD front-end (that is, all
controllers which uses standart IDE devices and have pull-up resitors).

--
Manuel Bouyer, LIP6, Universite Paris VI.           Manuel.Bouyer@lip6.fr
--