On 9/14/21 9:06 AM, Sagi Grimberg wrote:

quoted

quoted

quoted

@@ -361,11 +366,13 @@ static inline void nvme_end_req(struct request

*req)
     void nvme_complete_rq(struct request *req)
   {
+    struct nvme_ctrl *ctrl = nvme_req(req)->ctrl;
+
       trace_nvme_complete_rq(req);
       nvme_cleanup_cmd(req);
   -    if (nvme_req(req)->ctrl->kas)
-        nvme_req(req)->ctrl->comp_seen = true;
+    if (ctrl->kas)
+        ctrl->comp_seen = true;
         switch (nvme_decide_disposition(req)) {
       case COMPLETE:

@@ -377,6 +384,15 @@ void nvme_complete_rq(struct request *req)

       case FAILOVER:
           nvme_failover_req(req);
           return;
+    case AUTHENTICATE:
+#ifdef CONFIG_NVME_AUTH
+        if (nvme_change_ctrl_state(ctrl, NVME_CTRL_RESETTING))
+            queue_work(nvme_wq, &ctrl->dhchap_auth_work);

Why is the state change here and not in nvme_dhchap_auth_work?

Because switching to 'resetting' is an easy way to synchronize with the
admin queue.

Maybe fold this into nvme_authenticate_ctrl? in case someone adds/moves
this in the future and forgets the ctrl state serialization?

Yeah; not a bad idea. Will be looking into it.

Cheers,

Hannes
-- 
Dr. Hannes Reinecke		           Kernel Storage Architect
hare@suse.de			                  +49 911 74053 688
SUSE Software Solutions Germany GmbH, Maxfeldstr. 5, 90409 Nürnberg
HRB 36809 (AG Nürnberg), GF: Felix Imendörffer

_______________________________________________
Linux-nvme mailing list
Linux-nvme@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-nvme