On Fri, Jul 18, 2025 at 05:41:47PM -0400, Donald Dutile wrote: > > Not that, at the start of the function there is a pci_physfn(), the > > entire function never works on a VF, so bus is never a VF's bus. > > > Well, i guess it depends on what you call 'a VF's bus' -- it returns the VF's->PF(pdev)->bus if virt-fn, > which I would call the VF's bus. There is "VF bus" that the PCI core creates if the VF's RID reaches outside the PF's bus. This bus - the "SRIOV virtual bus" has a NULL self. Jason
