Re: PCI expansion card working?

[Tim Rightnour <root%garbled.net@localhost>]

On 16-Jan-2008 KIYOHARA Takashi wrote:
> That log attach this mail.  I am inserting fwohci+ehci and bktr (my best
> combination ;-).
> The combination cannot be booted. Therefore, the following correction
> was done.

Ok..this is a bit more complex than I originally hoped.  But if you are willing
to do a little debugging, we might be able to figure it out.   It will be
easier for you to debug this on -current, if thats OK. (me too, since I'm
beginning to forget how 4.0 interrupts worked)

First, since siop is usually the one that hangs the system, can you disable
siop in the config file and see if the machine still hangs?  (perhaps mount
root on nfs if you needed siop to boot)

Second, in powerpc/pic/intr.c:pic_handle_intr(), there is a section where we
loop through all the handlers for an interrupt.

find this:
(*ih->ih_fun)(ih->ih_arg);

and change it to something like this:

printf("calling handler\n");
foo = (*ih->ih_fun)(ih->ih_arg);
printf("returned %d\n", foo);

It will also be easier to debug, if you go into each of the actual interrupt
handlers for the devices and printf the top of them, so we know which one got
called.  (I'm not aware of a quick way to printf *ih->ih_fun from
pic_handle_intr(), if you are, do that instead, and tell me what it is)

My first theory, would be that someone else is handling siop's interrupts and
somehow clearing them, or it's not descending down enough to get to siop.

My second theory would be that something disabled the interrupt somehow.

---
Tim Rightnour <root%garbled.net@localhost>
NetBSD: Free multi-architecture OS http://www.netbsd.org/
Genecys: Open Source 3D MMORPG: http://www.genecys.org/