Re: HPET regression in 2.6.26 versus 2.6.25 -- found another user with the same regression

[Jesse Barnes <jbarnes@virtuousgeek.org>]

On Monday, August 25, 2008 1:00 am Yinghai Lu wrote:
> On Mon, Aug 25, 2008 at 12:31 AM, Ingo Molnar <mingo@elte.hu> wrote:
> > * Yinghai Lu <yhlu.kernel@gmail.com> wrote:
> >> this one should work. please apply this one only.
> >>
> >> YH
> >>
> >> [PATCH] x86: check hpet with BAR v2
> >
> > great. I've cleaned it up a bit (see the final commit below) and queued
> > it up in tip/x86/urgent for some testing. But there are a few open
> > questions, and an Ack/feedback from Jesse/Linus would be nice as well:
> >
> > - the forced insertion and the embedded knowledge about iomem_resource
> >  and ioport_resource looks ugly to me.
> >
> > - we should also extend this to other platform resource types that we
> >  know about: ioapic address(es) might be a prime candidate. (local
> >  APICs are CPU entities and should never show up as PCI devices) The
> >  mmconfig range is already properly accounted for by the PCI code
> >  itself, right?
>
> please check v3
>
> > - plus a more highlevel approach would be nice as well i think - making
> >  sure that the hpet driver runs before any of the PCI code, and
> >  inserting a special "sticky" resource there which would keep any
> >  potential followup generic PCI resource that overlaps this resource
> >  untouched. (with a proper kernel warning emitted as well - such
> >  situations are likely BIOS bugs.)
>
> yes, can remove that late_initcall tricky code to insert resource for
> lapic and ioapic, and mmconfig etc.

I agree, a higher level way of dealing with resource reservation might be 
nice.  I'm hoping to polish up TJ's PCI allocation code 
(http://tjworld.net/wiki/Linux/PCIDynamicResourceAllocationManagement) for 
2.6.28; it may have some stuff that can help.

Thanks,
-- 
Jesse Barnes, Intel Open Source Technology Center