[PATCH v2 1/8] nvme-fabrics: allow to queue requests for live queues

[Christoph Hellwig]

On Thu, Aug 06, 2020 at 12:11:20PM -0700, Sagi Grimberg wrote:
> Right now we are failing requests based on the controller
> state (which is checked inline in nvmf_check_ready) however
> we should definitely accept requests if the queue is live.
> 
> When entering controller reset, we transition the controller
> into NVME_CTRL_RESETTING, and then return BLK_STS_RESOURCE for
> non-mpath requests (have blk_noretry_request set).
> 
> This is also the case for NVME_REQ_USER for the wrong reason.
> There shouldn't be any reason for us to reject this I/O in a
> controller reset. We do want to prevent passthru commands on
> the admin queue because we need the controller to fully initialize
> first before we let user passthru admin commands to be issued.
> 
> In a non-mpath setup, this means that the requests will simply
> be requeued over and over forever not allowing the q_usage_counter
> to drop its final reference, causing controller reset to hang
> if running concurrently with heavy I/O.

Which will still happen with the admin queue user passthrough
commands with this patch, so I don't think it actually solves anything,
it just reduces the exposure a bit.

> 
> While we are at it, remove the redundant NVME_CTRL_NEW case, which
> should never see any I/O as it must first transition to
> NVME_CTRL_CONNECTING.

That probablyly should be a separate patch.

> -		if (nvme_is_fabrics(req->cmd) &&
> +		if (blk_rq_is_passthrough(rq) && nvme_is_fabrics(req->cmd) &&

And this (make sure we don't access garbage in ->cmd for non-passthrough)
should probably be a separate fix as well.