Re: My understanding of the IDE mess, and why it does not make sense to apply the proposed patch

[Anthony DeRobertis <aderobertis@metrics.net>]

Jurij Smakov wrote:

> Looking at the code I cannot see how the
> native drivers can depend in any way on the ide-generic being loaded
> before them.

While I have not thoroughly tested 2.6.15 in this respect, in 2.6.8 and
2.6.12 some IDE drivers, on some hardware, absolutely does[0]. Maybe
this is fixed in 2.6.15/16.

On those older kernels, the correct sequence of module loadings is:

1. hardware-specific-driver
2. ide-generic

If you stop after step 1, you don't have any IDE devices. If you reverse
steps 1 and 2, you don't have DMA.


[0] Honestly, I had previously thought *all* hardware did...