On Fri, 2016-04-15 at 06:40 +0200, Martin Kletzander wrote:
On Thu, Apr 14, 2016 at 06:24:26PM +0200, Andrea Bolognani wrote:
> On Wed, 2016-04-13 at 16:37 +0200, Martin Kletzander wrote:
> >
> > v2:
> > - Just a rebase
> > - I did *not* use virPCIDeviceAddress wording instead as discussed in
> > the v1 thread. That's because we have lot of functions working
> > with virDevicePCIAddress named exactly after that and renaming
> > those would be ugly IMHO.
>
> Sorry, but I feel pretty strongly the other way around: if
> it's defined in virpci.h, it should be called virPCI*.
>
> virDevicePCIAddress is used a lot but AFAICT the number of
> functions whose name is derived from it is just six.
>
> Moreover, we don't have other virDevice*Address types (or
> even just virDevice*) to set a precedent, but we have a
> bunch of virPCI* stuff including virPCIDevice, which happens
> to have a virPCIDeviceAddress among its members.
>
> Bikeshedding, I know, but there you have it :)
No problem, I'm not against discussing this. Feel free to try it out if
you want. I actually should've posted something like 2a/4 and 2b/4,
i.e. two versions of patches so that it's visible what's nice and what's
not :) If we don't reach a conclusion, I might do that later.
ACK series with the stuff pointed out in 4/4 taken care of
on the condition that you *do not* rename stuff from
virPCIDeviceAddress to virDevicePCIAddress.
If you're not okay with that condition we can discuss the
matter further next week :)
Cheers.
--
Andrea Bolognani
Software Engineer - Virtualization Team