Re: [Xen-devel] [question] Does HVM domain support xen-pcifront/xen-pciback?

[Konrad Rzeszutek Wilk <konrad.wilk@xxxxxxxxxx>]

On Tue, Feb 07, 2012 at 09:36:31PM +0800, cody wrote:
> On 02/07/2012 01:58 AM, Konrad Rzeszutek Wilk wrote:
> >On Mon, Feb 06, 2012 at 04:32:05PM +0800, Kai Huang wrote:
> >>Hi,
> >>
> >>I see in pcifront_init, if domain is not PV domain, pcifront_init just
> >>returns error. So seems HVM domain does not support
> >>xen-pcifront/xen-pciback mechanism? If it is true, why? I think
> >Yup. B/c the only thing that the PV PCI protocol does is enable
> >PCI configuration emulation. And if you boot an HVM guest - QEMU does
> >that already.
> >
> I heard qemu does not support PCIE simulation, and Xen does not
> provides MMIO mechanism but only legacy IO port mechanism to guest
> for configuration space access. Is this true?

The upstream version has a X58 north bridge implementation to support this.
(ioh3420.c).

In regards to MMIO mechanism are you talking about MSI-X and such? Then
the answer is it does. QEMU traps when a guest tries to write MSI vectors
in the BAR space and translates those to appropiate xen calls to setup
vectors for the guest.

> 
> If using IO port mechanism, we can only access first 256B of
> configuration space, but if using PV PCI protocol, we will not have
> such limitation. I think this is an advantage of PCI PV protocol. Of
> course if Xen provides MMIO mechanism to guest for configuration
> space, it will not have this limitation too.

What do you mean by MMIO mechanism? Setup of MSI-x vectors?
> 
> >>technically there's nothing that can block to support pcifront/pciback
> >>in HVM, and for performance reason, there will be benefits if HVM
> >>supports PV PCI operation.
> >Nope. The PCI operations are just for writting configuration deta in the
> >PCI space. Whcih is done mostly when a driver starts/stops and no more.
> >
> >The performance is with interrupts and how to inject them in a guest - and
> >in there the PV is much faster than HVM due to the emulation layer 
> >complexity.
> >However, work by Stefano on making that path shorter has made the interrupt
> >injection much much faster.
> I think PV PCI protocol can be used for other purpose in the future,
> such as PF/VF communication. In this case it will be better if HVM
> domain can support PV PCI protocol.

What is 'PF/VF' communication? Is it just the negoting in the guest of
parameters for the PF to inject in the VF? That seems a huge security risk.

Or is it some other esoteric PCI configuration options that are not 
available in the VF's BARs?

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

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