Donate to e Foundation | Murena handsets with /e/OS | Own a part of Murena! Learn more

Commit b04744ce authored Apr 09, 2018 by James Smart Committed by Martin K. Petersen Apr 18, 2018

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 <dick.kennedy@broadcom.com>
Signed-off-by: James Smart <james.smart@broadcom.com>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

parent bf316c78

drivers/scsi/lpfc/lpfc_ct.c

+5 −0

Original line number	Original line	Diff line number	Diff line
	@@ -691,6 +691,11 @@ lpfc_cmpl_ct_cmd_gid_ft(struct lpfc_hba phba, struct lpfc_iocbq cmdiocb,
	vport->fc_flag &= ~FC_RSCN_DEFERRED;		vport->fc_flag &= ~FC_RSCN_DEFERRED;
	spin_unlock_irq(shost->host_lock);		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		* Skip processing the NS response
	* Re-issue the NS cmd		* Re-issue the NS cmd

drivers/scsi/lpfc/lpfc_els.c

+0 −1

Original line number	Original line	Diff line number	Diff line
	@@ -6268,7 +6268,6 @@ lpfc_els_handle_rscn(struct lpfc_vport *vport)
	* flush the RSCN. Otherwise, the outstanding requests		* flush the RSCN. Otherwise, the outstanding requests
	* need to complete.		* need to complete.
	*/		*/
	vport->gidft_inp = 0;
	if (lpfc_issue_gidft(vport) > 0)		if (lpfc_issue_gidft(vport) > 0)
	return 1;		return 1;
	} else {		} else {

drivers/scsi/lpfc/lpfc_nvme.c

+10 −2

Original line number	Original line	Diff line number	Diff line
	@@ -2721,8 +2721,16 @@ lpfc_nvme_register_port(struct lpfc_vport vport, struct lpfc_nodelist ndlp)
	spin_unlock_irq(&vport->phba->hbalock);		spin_unlock_irq(&vport->phba->hbalock);
	rport->ndlp = NULL;		rport->ndlp = NULL;
	rport->remoteport = 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. */		/* Clean bind the rport to the ndlp. */