On Tue, Dec 29, 2015 at 4:36 PM, Burakov, Anatoly <anatoly.burakov at intel.com> wrote: > Hi Santosh, > >> Look at kernel/resource.c, it exports two symbol ioport_resource and >> iomem_resource and sets appropriate flag type i.e.. IORESOURCE_IO and >> IORESOURCE_MEM. In virtio-net case; it creates both pci region i.e.. >> _io bar and _mem bar. dpdk virtio pmd driver (<= 0.95 virtio spec) uses pci >> _io bar region for device initialization as virtio headers are locate at pci >> _io bar >> region. Since it uses pci _iobar region so likely it update >> pci_resource.[index].flag = IORESOURCE_IO. and vfio mmap function wont >> handle ioresource_io (i guess). And that is why I asked same to lkml thread. > > Yes, I can see that this is what I was misunderstanding (that IORESOURCE_IO > isn't a BAR but rather a flag). Technically, if you also set IORESOURCE_MEM > along with IORESOURCE_IO, then the BAR would be mapped by VFIO, but this > sounds like a dangerous hack :) other than that, yes, I'm afraid it's up to > kernel guys to add support for it. Once done, provided IO BARs are meant to > be worked with the same way as MEM BARs, there's a good chance it would work > out of the box with DPDK. >
I guess so, I'll give it a try; although before that I need to port / use your vfio/noiommu patch for virtio pmd driver and perhaps do dependant changes, I will post my test feedback. Thanks > Thanks, > Anatoly
