John,
On 2/19/25 15:48, John Levon wrote:
This is the 8th revision of the vfio-user client implementation. The vfio-user
protocol allows for implementing (PCI) devices in another userspace process;
SPDK is one example, which includes a virtual NVMe implementation.
The vfio-user framework consists of 3 parts:
1) The VFIO user protocol specification.
2) A client - the VFIO device in QEMU that encapsulates VFIO messages
and sends them to the server.
3) A server - a remote process that emulates a device.
This patchset implements parts 1 and 2.
It has been tested against libvfio-user test servers as well as SPDK.
Thanks for previous reviews & comments.
Changes since v7:
- split up pci patches for easier reviewing
- fixed lots of device ops error handling
- vfio-user code now in hw/vfio-user
- improved commit messages
- various other small cleanups
Could please resend next version on top
https://github.com/legoater/qemu/commits/vfio-10.1
This is the branch in which I put the experimental stuff for the
current cycle. It contains today the "spring cleanup" v2 series
I plan to send when QEMU 10.1 opens. Well, not exactly it because
I am waiting for some emails from Avihai to reach the list before
doing an update. It should be very close.
Also, could you please resend the vfio prerequisites only ? The
first 10 patches or so. It seems these could be merged soon. At
least checked for functional and performance regressions.
There are still relatively big changes in VFIO to discuss in these
patches:
vfio-user: implement VFIO_USER_DEVICE_GET_REGION_INFO
vfio-user: implement VFIO_USER_REGION_READ/WRITE
vfio-user: set up PCI in vfio_user_pci_realize()
vfio-user: forward MSI-X PBA BAR accesses to server
Something for later.
Thanks,
C.
