Dropped virtio-dev list.

On 20-05-24, 12:20, Stefano Garzarella wrote:

I haven't followed v1 and v2, so apologies in advance if my comments have already been discussed.

No issues.

On Mon, May 20, 2024 at 02:59:40PM GMT, Viresh Kumar wrote:

...

+\transportnormative{\subsection}{Virtio Transport Requirements}{Virtio Transport Options / Virtio Transport Requirements} +The transport MUST provide a mechanism for the driver to discover the +device.

This seems a little too stringent. For example, does MMIO offer a discovery mechanism? IIRC we have to pass the address via device tree or kernel command line.

Right. Is converting this to SHOULD okay or do you suggest some other change ?

...

+The transport MUST provide a mechanism to implement config space

"configuration space"

Sure.

...

between +the device and the driver.

+\devicenormative{\subsection}{Virtio Transport Requirements}{Virtio Transport Options / Virtio Transport Requirements}

+The device MUST keep any data associated with a device-initiated +transaction accessible to the driver until the driver acknowledges the +transaction to be complete.

+The device MUST NOT access the contents of a virtqueue before the +driver notifies, in a transport defined way, the device that the +virtqueue is ready to be accessed.

+The device MUST NOT access or modify buffers on a virtqueue after it has +notified the driver about their availability.

+The device MUST reset the virtqueues if requested by the driver, in a

Should we also talk about the possibility of resetting the entire device?

Do we support that today ? Maybe we can if the maintainers agree.