Re: [PATCH] hw/pci/pci: Fix slot check for plugged devices

[Michael S. Tsirkin]

On Wed, Sep 23, 2020 at 05:57:19PM +0200, Julia Suvorova wrote:
> On Wed, Sep 23, 2020 at 5:03 PM Michael S. Tsirkin <mst@redhat.com> wrote:
> >
> > On Wed, Sep 23, 2020 at 11:26:36AM +0200, Julia Suvorova wrote:
> > > If devfn is assigned automatically, 'else' clauses will never be
> > > executed. And if it does not matter for the reserved and available
> > > devfn, because we have already checked it, the check for function0
> > > needs to be done again.
> > >
> > > Signed-off-by: Julia Suvorova <jusual@redhat.com>
> >
> > This is just cosmetics right? I wouldn't describe this as
> > a "fix" then - "simplify" would be clearer.
> 
> No, this is a bug fix. For example, if you had a root port with a
> device on it already, and you do
> 'device_add new_device,bus=the_same_root_port', then it will miss the
> last 'if' and will be added to slot 1.

OK and I think that is the only example - if devfn is not supplied
and is auto-assigned. Can you add this to commit log pls?


> > > ---
> > >  hw/pci/pci.c | 5 +++--
> > >  1 file changed, 3 insertions(+), 2 deletions(-)
> > >
> > > diff --git a/hw/pci/pci.c b/hw/pci/pci.c
> > > index de0fae10ab..ae132b0b52 100644
> > > --- a/hw/pci/pci.c
> > > +++ b/hw/pci/pci.c
> > > @@ -1034,8 +1034,9 @@ static PCIDevice *do_pci_register_device(PCIDevice 
> > > *pci_dev,
> > >                     PCI_SLOT(devfn), PCI_FUNC(devfn), name,
> > >                     bus->devices[devfn]->name);
> > >          return NULL;
> > > -    } else if (dev->hotplugged &&
> > > -               pci_get_function_0(pci_dev)) {
> > > +    };
> > > +
> > > +    if (dev->hotplugged && pci_get_function_0(pci_dev)) {
> > >          error_setg(errp, "PCI: slot %d function 0 already ocuppied by 
> > > %s,"
> > >                     " new func %s cannot be exposed to guest.",
> > >                     PCI_SLOT(pci_get_function_0(pci_dev)->devfn),
> > > --
> > > 2.25.4
> >