Re: [Qemu-devel] [PATCH] hw/pci-host: Fix x86 Host Bridges 64bit PCI hole

[Marcel Apfelbaum]

Hi Gerd,

On 20/10/2017 9:55, Gerd Hoffmann wrote:
>    Hi,
>
> > > commit message says:
> > >
> > > <quote>
> > >       It turns out that some 32 bit windows guests crash
> > >       if 64 bit PCI hole size is >2G.
> > > </quote>
> > >
> > > Why this suddenly isn't a problem any more?
> >
> > I suppose it is, so we need a way to turn it "off".
>
> Or have machine types behave differently, i.e. give q35 a large 64bit
> hole and leave pc as-is.
>
> Devices with that large bars are most likely pci express anyway ...

Agreed

> > This is how I started, however Eduardo and (and maybe Michael ?)
> > where against letting the upper management software to deal with
> > such a low low level detail. They simply can't take such a decision.
> > This is why the property you mentioned is not ever linked
> > in code anywhere! It is simply not implemented and not used.
>
> Makes sense.
>
> BTW:  Is it safe to just assume 40 bits physical is going to work?  My
> workstation:
>
> model name      : Intel(R) Core(TM) i7-7700K CPU @ 4.20GHz
> address sizes   : 39 bits physical, 48 bits virtual
>
> Does this imply ept is limited 39 bits physical too?

The guest will still run with 40 bits physical!
(judging the code anyway, I hope I am wrong)

Then, the stakes are not so big, the Guest kernel will disregard
the 64bit hole since is not CPU addressable and go on.
At least Linux does it, but I think Win10 also can leave with that.

Thanks,
Marcel

> cheers,
>    Gerd