Re: [PULL] virtio

On Wed, 2011-11-02 at 11:39 +1030, Rusty Russell wrote:

On Tue, 1 Nov 2011 14:42:23 +0200, "Michael S. Tsirkin" [off-list ref] wrote:

quoted

On Tue, Nov 01, 2011 at 02:33:33PM +0200, Sasha Levin wrote:

quoted

Unless Michael pointed this patch out, it would have broken (at least)
the kvm tool in a non obvious way which would require a rather long
session of 'git bisect' to figure out whats wrong.

I assumed not, until QEMU provided the second BAR it would simply fall
back.  Or am I completely misunderstanding?

QEMU uses two bars within each virtio-pci device. First one (BAR 0) is
the PIO config BAR, and the second one (BAR 1) is the MSI-X BAR used for
the MSI-X table and the PBA.

Since the PCI spec allows you to have the MSI-X table and the PBA in
completely different regions pointed by two different BARs, we did so in
kvm tools and have BAR 1 pointing to the MSI-X table, and BAR 2 pointing
to the PBA ( - we *already* provided the second BAR and used it for
something else).

This means that with this patch, we'd load BAR 2 and use it as the
config space, instead of the PBA table which it really represents.

-- 

Sasha.

On Wed, 02 Nov 2011 06:52:44 +0200, Sasha Levin [off-list ref] wrote:

Since the PCI spec allows you to have the MSI-X table and the PBA in
completely different regions pointed by two different BARs, we did so in
kvm tools and have BAR 1 pointing to the MSI-X table, and BAR 2 pointing
to the PBA ( - we *already* provided the second BAR and used it for
something else).

Got it!  Thanks for the explanation.

Cheers,
Rusty.