Manuel Bouyer wrote:
> Ok, pciide releases the I/O address space in this case, but keep the interrupt
> mapped. Previously this worked by acciddent, because we still attached
> a atabus to a disabled channel, and the drive was probed on this atabus,
> and was getting interrupts though the (disabled !) pciide0 device.
> Now the drive gets probed on the atabus attached to wdc at isa, but wdc at isa
> doesn't gets the interrupts.
 
> The attached patch should fix this, can you give it a try ?

Yes, your patch fixes the problem. Thanks.

-- 
Christian