[RFC PATCH 1/5] nvme-pci: add function nvme_submit_vf_cmd to issue admin commands for VF driver.

Christoph Hellwig hch at lst.de
Wed Dec 7 05:52:03 PST 2022


On Wed, Dec 07, 2022 at 09:34:14AM -0400, Jason Gunthorpe wrote:
> The VFIO design assumes that the "vfio migration driver" will talk to
> both functions under the hood, and I don't see a fundamental problem
> with this beyond it being awkward with the driver core.

And while that is a fine concept per see, the current incarnation of
that is fundamentally broken is it centered around the controlled
VM.  Which really can't work.

> Even the basic assumption that there would be a controlling/controlled
> relationship is not universally true. The mdev type drivers, and 
> SIOV-like devices are unlikely to have that. Once you can use PASID
> the reasons to split things at the HW level go away, and a VF could
> certainly self-migrate.

Even then you need a controlling and a controlled entity.  The
controlling entity even in SIOV remains a PCIe function.  The
controlled entity might just be a bunch of hardware resoures and
a PASID.  Making it important again that all migration is driven
by the controlling entity.

Also the whole concept that only VFIO can do live migration is
a little bogus.  With checkpoint and restart it absolutely
does make sense to live migrate a container, and with that
the hardware interface (e.g. nvme controller) assigned to it.

> So, when you see both Intel and Pensando proposing this kind of
> layered model for NVMe where migration is subsystem-local to VFIO, I
> think this is where the inspiration is coming from. Their native DPU
> drivers already work this way.

Maybe they should have talked to someone not high on their own
supply before designing this.



More information about the Linux-nvme mailing list