[PATCH 0/5] pci: nvmet: support completion queue sharing by multiple submission queues

[Christoph Hellwig]

On Thu, Apr 24, 2025 at 03:16:54PM +0200, Niklas Cassel wrote:
> Perhaps it is a bit too harsh to say that we are non-spec compliant.

It is not too harsh but simply a true statement.  If the spec
compliance matters or not might be a different question.

> We don't implement every feature in the NVMe spec in the Linux drivers
> on purpose, because there is a lot of feature creep in NVMe.

That is a valid point for optional features.  Sharing a CQ for multiple
SQs is a mandatory feature in PCIe.

This should also get us to the point of running verification test
suite on the PCIe nvmet code, which I'm really looking forward to.