On Thu, Apr 01, 2021 at 12:07:30PM +0200, Vincent Bernat wrote:
> ❦ 1 avril 2021 10:38 +01, Daniel P. Berrangé:
>
> >> hw/smbios/smbios.c | 47 +++++++++++++++++++++-------------------------
> >> qemu-options.hx | 2 +-
> >> 2 files changed, 22 insertions(+), 27 deletions(-)
> >
> > It doesn't really make sense to have this as a separate patch
> > when it is deleting half the code you added in the previous
> > patch. Just merge them together as one.
>
> I'll do that.
>
> >> + /*
> >> + * TODO: Extract the appropriate value. Most of the
> >> + * time, this will be 0.
> >> + */
> >> + t->segment_group_number = cpu_to_le16(0);
> >
> > Hmm, tricky, as it requires interpreting the PCI topology. Wonder if
> > there's any helper that can do the hard work for you
>
> There is pci_root_bus_path(), but it returns a string which could just
> contain a segment or several segments. It seems the SMBIOS standard
> didn't account for complex topologies. I could parse the string. and
> keep only the right-most segment.
>
> >> + t->bus_number = pci_dev_bus_num(pdev);
> >> + t->device_number = pdev->devfn;
> >> + } else {
> >> + fprintf(stderr, "%s: cannot find PCI device %s\n",
> >> + __func__, t41->pcidev);
> >
> > This isn't terminating execution which looks like a bug.
>
> It was my intention. The PCI address will then be 00:00:00.0. If you
> think it's better to terminate, I can do what you suggest.
If we can't find the PCI device, that's user configuration error, and
we prefer to report those & exit, rather than continuing with likely
bogus data.
Regards,
Daniel
--
|: https://berrange.com -o- https://www.flickr.com/photos/dberrange :|
|: https://libvirt.org -o- https://fstop138.berrange.com :|
|: https://entangle-photo.org -o- https://www.instagram.com/dberrange :|
