A doc laying out the rationale for the choice of primitives in virtio-crypto [1] would be very nice. The choice is weird. Also, I couldn't find links to actual specs of the named supported primitives, seems we should understand what they mean by name alone.

1 - https://docs.oasis-open.org/virtio/virtio/v1.1/cs01/virtio-v...

I see pretty high overlap with dpdk cryptodev primitives[1]. If I had to guess, they are related and probably based on what actual hardware provides and that is known to be used in some key applications (e.g. 4G/LTE infrastructure).

https://doc.dpdk.org/guides-18.11/cryptodevs/overview.html

I wonder what has happened with the VirtFS filesystem over virtio device. This presented a 9p interface directly in virtio to share filesystems between machines. The only reference I see in the spec is the 0x1009 device id being allocated to the 9P transport.

I believe some of the thinking has shifted toward virtio-fs[1] instead of the 9p/network oriented models. Though this requires hypervisor locality for sharing among VMs.

[1] https://virtio-fs.gitlab.io/

At some level yes. But virtio 9pfs is still widely used. It's just that no one yet bothered writing up the spec.

What's changed?

Stefan Hajnoczi has a nice summary:

http://blog.vmsplice.net/2019/04/whats-new-in-virtio-11.html

If you go here: https://docs.oasis-open.org/virtio/virtio/v1.1/cs01/ you will find a version with changes from v1.0 cs04 highlighted. In case you participated in a public review, changes since the public review draft are also listed in the specification itself.