Messages in this thread | | | Date | Wed, 4 Jun 2008 19:51:15 +0100 (BST) | From | "Maciej W. Rozycki" <> | Subject | Re: [PATCH 0/7] Boot IRQ quirks and rerouting |
| |
On Wed, 4 Jun 2008, Thomas Gleixner wrote:
> > Anyway, my other questions remain open. How is the INTx message actually > > delivered to the ICH? > > I have roughly as much clue as you about that. The datashi^Heet is not > really helpful.
I haven't been following Intel's datasheets recently. They usually do publish all the necessary bits, but they are scattered around, so you really have to read them all. Chances are it's mentioned somewhere. Besides, I have no access to the PCI Express spec and there may be further interesting bits there. I suppose Intel can be queried directly too, especially by the interested corporate parties.
> > And why is the command line option needed? > > We don't want to impose that in the first place, but OTOH to get test > coverage it's probably better to do it unconditionally.
I gather with the workarounds disabled affected systems are useless anyway, so what's the point?
Maciej
| |