Re: pci_attach_hook() tweaks

[Frank Wille <frank%phoenix.owl.de@localhost>]

Tim Rightnour wrote:

>> Ok, I have cleared the native-mode bits in the PCI-interface register,
>> but the result is that the controller is dead, and no more drives will be
>> found.
> 
> I know someone told us we have to futz with it, but.. umm.. why?

Because the secondary channel didn't work. I guessed it was firing irq 15,
which is not recognized.


>  Are we
> actually seeing a problem here, or are we just doing it because linux did?

The primary channel works fine, and my hard drive is recognized, but the
secondary channel runs into many timeouts and "lost interrupt", before
showing the name of my CD-ROM.


> I'm kinda inclined to punt on it until we have the machine coming up in
> single and we can test the actual drives out, specifically because you are
> having problems with it once you fiddle the bits.

Ok, you're right that this is not a priority item at the moment. It just
takes a lot of time until all timeouts have passed. Or I have to disable the
secondary channel for now.


> Speaking of which, there is definately a problem with userland, as I can't
> seem to get fork to work properly, and init just kinda freaks out.

That's true. I tried to debug it some weeks ago, and it seems to me that
copyin()/copyout() don't work correctly?


> haven't looked into this at all yet, as this was the first time I decided
> to actually give it a go. Maybe one of the header files is missing
> something, or there is a clock issue. (Clock issue strikes me as a
> definate possibility, I don't think I ever fiddled with the clock other
> than the rtas initialization of it)

Let's do it this way: you will look into the /sbin/init-fork issue, and I
will look into the viaide issue (I think I should ask somebody in the MorphOS
team again).


-- 
    _  Frank Wille (frank%phoenix.owl.de@localhost)
 _ //  http://sun.hasenbraten.de/~frank/
 \X/   Phx @ #AmigaGer