Re: Aic7xxx v6.2.22 and Aic79xx v1.3.0Alpha2 Released

["Justin T. Gibbs" <gibbs@scsiguy.com>]

> On Fri, Dec 13, 2002 at 09:02:15PM +0000, Christoph Hellwig wrote:
>> On Thu, Dec 12, 2002 at 03:20:52PM -0500, Doug Ledford wrote:
>> > On Wed, Dec 11, 2002 at 06:17:46PM +0000, Christoph Hellwig wrote:
>> > > 
>> > > Yupp, it currently crashes when I have both compiled in.  Dough, any
>> > > chance you could fix that?  A PCI driver is not supposed to stop
>> > > over already claimed device.
>> > 
>> > If you've got a test machine, the attached patch would I think fix the 
>> > problem.  Let me know if it does in fact work for you and I'll commit
>> > it  to the tree.
>> 
>> Still doesn't work.  I think you really need to set and check the driver
>> field like justin suggested.
> 
> Hmmm...what's the failure mode here?

I haven't tested it locally yet, so I'm not sure.  My concern was that
the PCI subsystem would simply not know that your driver had attached
which could lead to a second attach playing with BARs or other configuration
space stuff before the aic7xxx driver even tries to map registers.
This doesn't occur in the non-PCI case, so the request region stuff is
probably sufficient there.

--
Justin