Re: svwsata: lost/missing interrupt

EF> Initially, the fix was #ifdef __powerpc__.
M> ... which is technically wrong since this isn't really powerpc-specific.
Yes, sure. So they changed the condition later on.

> This seems to work around a hardware bug in K2, so we should do the
> same and use the 32bit read only on those.
Yes, probably.

Since svwisata_intr() mostly duplicates pciide_pci_intr(), I also checked 
the latter for changes (after -8) not incorporated into the former, but 
couldn't find any.

That code duplication looks ugly, but I can't think of a better solution since 
the changes are utterly hardware-dependent.
Perhaps pciide_pci_intr() could get a comment that the code is duplicated 
elsewhere.