Re: [Xen-devel] [rfc 00/18] ioemu: use devfn instead of slots as the unit for passthrough

[Simon Horman <horms@xxxxxxxxxxxx>]

On Wed, Mar 04, 2009 at 10:32:11PM +0000, Keir Fraser wrote:
> On 04/03/2009 22:26, "Simon Horman" <horms@xxxxxxxxxxxx> wrote:
> 
> >> Please note that _PRT method in ACPI AML should reflect GSIs. If you expand
> >> GSIs, it will be necessary to change the _PRT method. Please see
> >> tools/firmware/hvmloader/acpi/dsdt.asl.
> > 
> > Could someone explain why the current code uses 32 GSIs rather than using 
> > 128?
> 
> 32 should be plenty, and I'd want to emulate multiple IO-APICs beyond that
> just out of fear that some OS will choke on an IO-APIC with 128 pins (since
> no real single IO-APIC has so many pins). It just seemed a bit unnecessary
> when this support was originally implemented. At this point I haven't heard
> any concrete evidence so far that 32 GSIs is insufficient in any real
> scenario. So that's what we'll be sticking with for the time being, I think.

Thanks, I will see if I can dig out any real-world reasons.

-- 
Simon Horman
  VA Linux Systems Japan K.K., Sydney, Australia Satellite Office
  H: www.vergenet.net/~horms/             W: www.valinux.co.jp/en


_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-devel