On Wed, Sep 01, 2021 at 01:53:34PM +0100, Alex Bennée wrote:
Stefan Hajnoczi stefanha@redhat.com writes:
[[PGP Signed Part:Undecided]] On Wed, Aug 04, 2021 at 12:20:01PM -0700, Stefano Stabellini wrote:
Could we consider the kernel internally converting IOREQ messages from the Xen hypervisor to eventfd events? Would this scale with other kernel hypercall interfaces?
So any thoughts on what directions are worth experimenting with?
One option we should consider is for each backend to connect to Xen via the IOREQ interface. We could generalize the IOREQ interface and make it hypervisor agnostic. The interface is really trivial and easy to add. The only Xen-specific part is the notification mechanism, which is an event channel. If we replaced the event channel with something else the interface would be generic. See: https://gitlab.com/xen-project/xen/-/blob/staging/xen/include/public/hvm/ior...
There have been experiments with something kind of similar in KVM recently (see struct ioregionfd_cmd): https://lore.kernel.org/kvm/dad3d025bcf15ece11d9df0ff685e8ab0a4f2edd.1613828...
Reading the cover letter was very useful in showing how this provides a separate channel for signalling IO events to userspace instead of using the normal type-2 vmexit type event. I wonder how deeply tied the userspace facing side of this is to KVM? Could it provide a common FD type interface to IOREQ?
I wondered this too after reading Stefano's link to Xen's ioreq. They seem to be quite similar. ioregionfd is closer to have PIO/MMIO vmexits are handled in KVM while I guess ioreq is closer to how Xen handles them, but those are small details.
It may be possible to use the ioreq struct instead of ioregionfd in KVM, but I haven't checked each field.
As I understand IOREQ this is currently a direct communication between userspace and the hypervisor using the existing Xen message bus. My worry would be that by adding knowledge of what the underlying hypervisor is we'd end up with excess complexity in the kernel. For one thing we certainly wouldn't want an API version dependency on the kernel to understand which version of the Xen hypervisor it was running on.
There is also another problem. IOREQ is probably not be the only interface needed. Have a look at https://marc.info/?l=xen-devel&m=162373754705233&w=2. Don't we also need an interface for the backend to inject interrupts into the frontend? And if the backend requires dynamic memory mappings of frontend pages, then we would also need an interface to map/unmap domU pages.
These interfaces are a lot more problematic than IOREQ: IOREQ is tiny and self-contained. It is easy to add anywhere. A new interface to inject interrupts or map pages is more difficult to manage because it would require changes scattered across the various emulators.
Something like ioreq is indeed necessary to implement arbitrary devices, but if you are willing to restrict yourself to VIRTIO then other interfaces are possible too because the VIRTIO device model is different from the general purpose x86 PIO/MMIO that Xen's ioreq seems to support.
It's true our focus is just VirtIO which does support alternative transport options however most implementations seem to be targeting virtio-mmio for it's relative simplicity and understood semantics (modulo a desire for MSI to reduce round trip latency handling signalling).
Okay.
Stefan