Re: [libvirt] [PATCH v6 10/17] qemu: auto-add pcie-root-port/dmi-to-pci-bridge controllers as needed

> 2) pci-bridge was requested with user-supplied address that is on a  > nonexistent bus, and that bus number is >1 over the current last bus  > (i.e. there is a gap in the controller indexes). This would happen if,  > for example, someone put this in their xml: >  >         <controller type='pci' index='0' model='pci-root'/> >         <controller type='pci' index='2' model='pci-bridge'/> >  > (if libvirt was setting the index, it would have set the index for the  > pci-bridge to '1'). >  >     a) pci-root - okay, here's where what you're saying would apply -  > we need to add extra pci-bridges >         to fill in the "empty space" in the list of controllers. >  >     b) pcie-root - actually we have the same issue in this case. >  > Personally I think that it's nonsensical for a user to do that, and  > trying to support "empty regions" in the PCI address space by  > auto-adding pci-bridges is pointless busywork, but we have supported  > it in the past, so I guess we need to now. Sigh.   But wait! Look at the example above - that's *exactly* the odd case that  patch 11/17 (the RFC patch that we agreed wasn't worthwhile) remedies.   Even without that RFC patch, if you look at the code that creates a  PCIAddressSet (qemuDomainPCIAddressSetCreate()), you'll see that during  the initial creation, the unused bus numbers are filled in with  pcie-root-port or pci-bridge controllers, i.e. the PCIAddressSet is  never "sparse", and thus virDomainPCIAddressSetGrow() is guaranteed  never to encounter the case above. So you're asking for a solution to a  nonexistent problem.