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

[Max Gurtovoy]

On 12/12/2022 9:55 AM, Christoph Hellwig wrote:
> On Sun, Dec 11, 2022 at 01:39:37PM +0200, Max Gurtovoy wrote:
> > 1. Need to define a concept of a "virtual subsystem". A primary controller
> > will be able to create a virtual subsystem. Inside this subsystem the
> > primary controller will be the master ("the controlling") of the migration
> > process. It will also be able to add secondary controllers to this virtual
> > subsystem and assign "virtual controller ID" to it.
> > something like:
> > - nvme virtual_subsys_create --dev=/dev/nvme1 --virtual_nqn="my_v_nqn_1"
> > --dev_vcid = 1
> > - nvme virtual_subsys_add --dev=/dev/nvme1 --virtual_nqn="my_v_nqn_1"
> > --secondary_dev=/dev/nvme2 --secondary_dev_vcid=20
> Yes. Note that there is a bit more state than just the NQN.  You also
> need at least a serial number, and also probably a different vendor
> ID (the PCI vendor ID that is also mirror in Identify Controller and
> the IEEE OUI), and new unique namespace identifier.

Yes for sure there is more bits we should consider.

I wanted to describe the high level.

I think that we can maybe say the when a secondary function is moved to 
a virtual subsystem its feature set of the virtual ctrl is narrowed to 
mandatory NVMe features set.
And we'll provide an API to set/extend it's feature set to a maximum of 
the feature set that the original ctrl of the secondary function had.
Then the sys admin will configure the virtual ctrl in both src/dst hosts.

The high level idea is to have a programmable way to set the features of 
a virtual controller inside a virtual subsystem that is also programmable.