[PATCH 11/33] conf: add 'pnv-phb3-root-port' domain definition

[Peter Krempa]

On Tue, Jan 25, 2022 at 09:40:48 -0300, Daniel Henrique Barboza wrote:
> 
> 
> On 1/21/22 11:17, Peter Krempa wrote:
> > On Thu, Jan 20, 2022 at 10:52:14 -0300, Daniel Henrique Barboza wrote:
> > > Apart from being usable only with pnv-phb3 PCIE host bridges (to be
> > > added soon), this device acts as a regular pcie-root-port but with a
> > > specific model name.
> > > 
> > > Signed-off-by: Daniel Henrique Barboza <danielhb413 at gmail.com>
> > > ---
> > >   docs/schemas/domaincommon.rng | 1 +
> > >   src/conf/domain_conf.c        | 1 +
> > >   src/conf/domain_conf.h        | 1 +
> > >   src/qemu/qemu_validate.c      | 2 ++
> > >   4 files changed, 5 insertions(+)
> > 
> > [...]
> > 
> > > diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c
> > > index 393f9d9478..c540b740df 100644
> > > --- a/src/conf/domain_conf.c
> > > +++ b/src/conf/domain_conf.c
> > > @@ -437,6 +437,7 @@ VIR_ENUM_IMPL(virDomainControllerPCIModelName,
> > >                 "pcie-root-port",
> > >                 "spapr-pci-host-bridge",
> > >                 "pcie-pci-bridge",
> > > +              "pnv-phb3-root-port",
> > >   );
> > 
> > Missing corresponding 'docs/formatdomain.rst' change.
> 
> The reason I didn't add a note in that doc is because of this specific paragraph:
> 
> "PCI controllers also have an optional subelement <model> with an attribute name. The name
> attribute holds the name of the specific device that qemu is emulating (e.g. "i82801b11-bridge")
> rather than simply the class of device ("pcie-to-pci-bridge", "pci-bridge"), which is set in
> the controller element's model attribute. In almost all cases, you should not manually add a
> <model> subelement to a controller, nor should you modify one that is automatically generated
> by libvirt. Since 1.2.19 (QEMU only)."
> 
> This, summed up with the fact that not all PCI model names are documented in this doc (e.g.
> the ioh3420 root port model name), gave me the impression that we don't want/bother to
> specify these details to the user.
> 
> 
> All this said, I can add a patch that documents all the model names currently supported,
> then I can add the new stuff on top of it.

No with the documentation stating that it's not really for users and
having an already existing status-quo of not mentioning what we already
have it's okay.

Just mention in the commit message the reason for not adding docs.