Bugfix for kernel page fault on apci_wait_irq (correct branch).

[sunipkm]

Fixes a bug where a page fault occurs when a device incapable of generating IRQs gets an wait_irq ioctl.

[jdolanIV]

The target branch is now the feature branch. I don't know how long it's going to be before the feature branch gets merged back into master. There are a couple of ways to do this, but what I would do is commit this fix to master and then rebase the feature branch off of master. That way the bugfix is already in and not waiting for the feature branch to be finished.
Let me know if you want to edit this PR, create a new one, or have me try to cherry-pick the fix to master.
Once it's in master either I or @JHentges will rebase the feature branch. Which means a forced update, but it's a feature branch.

[sunipkm]

You can cherry-pick the fix to master, whatever is easier for you guys really!

[jdolanIV]

This adventure continues. The patch refers to an irq_disabled field of ddata which doesn't exist in master which breaks compilation. I'm guessing it's something added in the alpha branch. I'm just going to leave it for @JHentges to deal with :P