0

I have few queries regarding DPDK's virtio driver when used on ARM machine. We do unbind virti-pci driver(of guest Kernel) from virto-net device emualted by QEMU, and then bind it to VFIO-PCI driver(of guest Kernel) but I didn't understand why we bind virtio-net device to VFIO-PCI driver?

In normal world, VFIO is used to assign device to guest but how it work in case of virtio-pmd driver for DPDK?

Also, is it necessary to have vhost-net in case of virtio-pmd and how it fit in there ?

Milan
  • 1,447
  • 5
  • 19
  • 27
  • By default for any virtio-net device is bind with virtio-pci. To make it available to UIO or any userspace one has to bind with virtio-pci. So is question, why we bind with `igb_uio` or `vfio-pci` to access virtio-net device? – Vipin Varghese Jan 10 '21 at 06:44
  • Yes, question is why we bind virtio-net device to vfio-pci, is it like vfio-pci driver does device configuration, for instance access to PCI device config space and actual packet processing is done virto pmd of DPDK? – Milan Jan 10 '21 at 08:47
  • I assume the hint will help, `vrtio-pci` kernel driver allows the device configuration and conversion of virtio packets to SKB. while `igb_uio or vfio-pci` allows configuration and buffer copy or zero copy to user space as it is. – Vipin Varghese Jan 12 '21 at 05:32

0 Answers0