Re: [current tinderbox] failure on i386/pc98

From: John Baldwin <jhb_at_FreeBSD.org>
Date: Fri, 20 Feb 2004 17:15:25 -0500
On Friday 20 February 2004 02:09 pm, M. Warner Losh wrote:
> In message: <200402201352.57589.jhb_at_FreeBSD.org>
>
>             John Baldwin <jhb_at_FreeBSD.org> writes:
> : On Thursday 19 February 2004 08:39 pm, M. Warner Losh wrote:
> : > : /other/tinderbox/CURRENT/i386/pc98/src/sys/i386/pci/pci_cfgreg.c:116:
> : > : undefined reference to `pci_pir_open'
> : >
> : > pc98 machines don't have a $PIR...
> :
> : They still use the common legacy_pcib host-pci bridge driver.  That
> : is, until someone implements a pc98 set of pci bridge drivers that
> : route interrupts (if that is even possible).  Are you sure none of
> : them have a PIR btw?  I thought there were some PII-based PC98
> : machines, and those should have a PIR.
>
> There are many PII based PC98 machines.  I have one in my basement
> even.  There's no PCIBIOS or $PIR that I can find.  I've been told by
> the pc98 guys that there is no PCIBIOS, but I've not been told there's
> no PIR.  It is possible to do interrupt routing on these machines, but
> no one has written an interrupt router for it.  The interrupt routing
> device that lives on the PCI bus (since it bridges to the CBUS bus,
> and the interrupts from that are connected to the PIC).  The good news
> is that one can easily write something that will do this, but the pc98
> bridge layer is going to be the one that copes.  One could shoe-horn
> this into the current scheme since there's really only ever one of
> these things as far as I know, so globals works.
>
> At the very least, the build needs to be fixed, maybe even with a 'you
> lose' stub.

I fixed the build yesterday.

-- 
John Baldwin <jhb_at_FreeBSD.org>  <><  http://www.FreeBSD.org/~jhb/
"Power Users Use the Power to Serve"  =  http://www.FreeBSD.org
Received on Fri Feb 20 2004 - 13:14:10 UTC

This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:37:44 UTC