[PATCH v5 06/14] nvmet-fcloop: add missing fcloop_callback_host_done

[Daniel Wagner]

On Thu, Apr 24, 2025 at 12:13:10PM +0200, Hannes Reinecke wrote:
> > +	if (!tfcp_req) {
> >   		/* abort has already been called */
> > -		return;
> > +		goto out_host_done;
> > +	}
> Sure this is correct?
> If the abort has been called I would have expected that all resources
> are cleaned up by the abort, so we wouldn't need to do that here...

Yes, it is still necessary to call fcpreq->done for this particular request.

> > @@ -983,7 +984,7 @@ fcloop_fcp_abort(struct nvme_fc_local_port *localport,
> >   	default:
> >   		spin_unlock_irqrestore(&tfcp_req->reqlock, flags);
> >   		WARN_ON(1);
> > -		return;
> > +		goto out_host_done;
> 
> Do we still need the WARN_ON()? We can now gracefully recover, so a
> simple log message would be sufficient here, no?

The WARN_ON is there because it catches when tfcp_req is in an wrong
state. It should be either in INI_IO_START, INI_IO_ACTIVE or
INI_IO_COMPLETED but never INI_IO_ABORT.