scsi: lpfc: Fix driver not recovering NVME rports during target link …

…faults During target-side port faults, the driver would not recover all target port logins. This resulted in a loss of nvme device discovery. The driver is coded to wait for all GID_FT requests to complete before restarting discovery. A fault is seen where the outstanding GIT_FT counts are not properly decremented, thus discovery would never start. Another fault was found in the clearing of the gidft_inp counter that would be skipped in this condition. And a third fault found with lpfc_nvme_register_port that would remove a reverence on the ndlp which then allows a node swap on a port address change to prematurely remove the reference and release the ndlp. The following changes are made: - Correct the decrementing of the outstanding GID_FT counters. - In RSCN handling, no longer zero the counter before calling to issue another GID_FT. - No longer remove the reference on the dlp when the ndlp->nrport value is not yet null. Signed-off-by: Dick Kennedy <[email protected]> Signed-off-by: James Smart <[email protected]> Reviewed-by: Hannes Reinecke <[email protected]> Signed-off-by: Martin K. Petersen <[email protected]>
zandrey · Apr 18, 2018 · b04744c · b04744c
1 parent bf316c7
commit b04744c
Show file tree

Hide file tree

Showing 3 changed files with 15 additions and 3 deletions.
diff --git a/drivers/scsi/lpfc/lpfc_ct.c b/drivers/scsi/lpfc/lpfc_ct.c
@@ -691,6 +691,11 @@ lpfc_cmpl_ct_cmd_gid_ft(struct lpfc_hba *phba, struct lpfc_iocbq *cmdiocb,
 		vport->fc_flag &= ~FC_RSCN_DEFERRED;
 		spin_unlock_irq(shost->host_lock);
 
+		/* This is a GID_FT completing so the gidft_inp counter was
+		 * incremented before the GID_FT was issued to the wire.
+		 */
+		vport->gidft_inp--;
+
 		/*
 		 * Skip processing the NS response
 		 * Re-issue the NS cmd

diff --git a/drivers/scsi/lpfc/lpfc_els.c b/drivers/scsi/lpfc/lpfc_els.c
@@ -6268,7 +6268,6 @@ lpfc_els_handle_rscn(struct lpfc_vport *vport)
 		 * flush the RSCN.  Otherwise, the outstanding requests
 		 * need to complete.
 		 */
-		vport->gidft_inp = 0;
 		if (lpfc_issue_gidft(vport) > 0)
 			return 1;
 	} else {

diff --git a/drivers/scsi/lpfc/lpfc_nvme.c b/drivers/scsi/lpfc/lpfc_nvme.c
@@ -2721,8 +2721,16 @@ lpfc_nvme_register_port(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp)
 			spin_unlock_irq(&vport->phba->hbalock);
 			rport->ndlp = NULL;
 			rport->remoteport = NULL;
-			if (prev_ndlp)
-				lpfc_nlp_put(ndlp);
+
+			/* Reference only removed if previous NDLP is no longer
+			 * active. It might be just a swap and removing the
+			 * reference would cause a premature cleanup.
+			 */
+			if (prev_ndlp && prev_ndlp != ndlp) {
+				if ((!NLP_CHK_NODE_ACT(prev_ndlp)) ||
+				    (!prev_ndlp->nrport))
+					lpfc_nlp_put(prev_ndlp);
+			}
 		}
 
 		/* Clean bind the rport to the ndlp. */