scsi: lpfc: Make fabric zone discovery more robust when handling unsolicited LOGO

This patch provides better target rport recovery when a target rport is
running in initiator mode to discover the fabric.  Such a target will issue
a LOGO before switching back to strict target mode and changes are made to
recover the login.  Log messages are also updated accordingly.
Signed-off-by: Justin Tee <justin.tee@broadcom.com>
Link: https://lore.kernel.org/r/20230712180522.112722-8-justintee8345@gmail.comSigned-off-by: Martin K. Petersen <martin.petersen@oracle.com>

drivers/scsi/lpfc/lpfc_ct.c

@@ -1557,7 +1557,8 @@ lpfc_cmpl_ct_cmd_gft_id(struct lpfc_hba *phba, struct lpfc_iocbq *cmdiocb,
 				ndlp->nlp_fc4_type |= NLP_FC4_FCP;
 			if (fc4_data_1 &  LPFC_FC4_TYPE_BITMASK)
 				ndlp->nlp_fc4_type |= NLP_FC4_NVME;
-			lpfc_printf_vlog(vport, KERN_INFO, LOG_DISCOVERY,
+			lpfc_printf_vlog(vport, KERN_INFO,
+					 LOG_DISCOVERY | LOG_NODE,
 					 "3064 Setting ndlp x%px, DID x%06x "
 					 "with FC4 x%08x, Data: x%08x x%08x "
 					 "%d\n",
@@ -1568,14 +1569,21 @@ lpfc_cmpl_ct_cmd_gft_id(struct lpfc_hba *phba, struct lpfc_iocbq *cmdiocb,
 			if (ndlp->nlp_state == NLP_STE_REG_LOGIN_ISSUE &&
 			    ndlp->nlp_fc4_type) {
 				ndlp->nlp_prev_state = NLP_STE_REG_LOGIN_ISSUE;
-
+				/* This is a fabric topology so if discovery
+				 * started with an unsolicited PLOGI, don't
+				 * send a PRLI.  Targets don't issue PLOGI or
+				 * PRLI when acting as a target. Likely this is
+				 * an initiator function.
+				 */
+				if (!(ndlp->nlp_flag & NLP_RCV_PLOGI)) {
 					lpfc_nlp_set_state(vport, ndlp,
 							   NLP_STE_PRLI_ISSUE);
 					lpfc_issue_els_prli(vport, ndlp, 0);
+				}
 			} else if (!ndlp->nlp_fc4_type) {
 				/* If fc4 type is still unknown, then LOGO */
 				lpfc_printf_vlog(vport, KERN_INFO,
-						 LOG_DISCOVERY,
+						 LOG_DISCOVERY | LOG_NODE,
 						 "6443 Sending LOGO ndlp x%px,"
 						 "DID x%06x with fc4_type: "
 						 "x%08x, state: %d\n",

drivers/scsi/lpfc/lpfc_els.c

@@ -2376,10 +2376,10 @@ lpfc_cmpl_els_prli(struct lpfc_hba *phba, struct lpfc_iocbq *cmdiocb,
 		/* PRLI failed */
 		lpfc_printf_vlog(vport, mode, loglevel,
 				 "2754 PRLI failure DID:%06X Status:x%x/x%x, "
-				 "data: x%x x%x\n",
+				 "data: x%x x%x x%x\n",
 				 ndlp->nlp_DID, ulp_status,
 				 ulp_word4, ndlp->nlp_state,
-				 ndlp->fc4_prli_sent);
+				 ndlp->fc4_prli_sent, ndlp->nlp_flag);
 
 		/* Do not call DSM for lpfc_els_abort'ed ELS cmds */
 		if (!lpfc_error_lost_link(vport, ulp_status, ulp_word4))
@@ -2390,10 +2390,12 @@ lpfc_cmpl_els_prli(struct lpfc_hba *phba, struct lpfc_iocbq *cmdiocb,
 		 * mismatch typically caused by an RSCN. Skip any
 		 * processing to allow recovery.
 		 */
-		if (ndlp->nlp_state >= NLP_STE_PLOGI_ISSUE &&
-		    ndlp->nlp_state <= NLP_STE_REG_LOGIN_ISSUE) {
+		if ((ndlp->nlp_state >= NLP_STE_PLOGI_ISSUE &&
+		     ndlp->nlp_state <= NLP_STE_REG_LOGIN_ISSUE) ||
+		    (ndlp->nlp_state == NLP_STE_NPR_NODE &&
+		     ndlp->nlp_flag & NLP_DELAY_TMO)) {
 			lpfc_printf_vlog(vport, KERN_WARNING, LOG_NODE,
-					 "2784 PRLI cmpl: state mismatch "
+					 "2784 PRLI cmpl: Allow Node recovery "
 					 "DID x%06x nstate x%x nflag x%x\n",
 					 ndlp->nlp_DID, ndlp->nlp_state,
 					 ndlp->nlp_flag);

drivers/scsi/lpfc/lpfc_nportdisc.c

@@ -879,23 +879,34 @@ lpfc_rcv_logo(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp,
 			spin_unlock_irq(shost->host_lock);
 			lpfc_retry_pport_discovery(phba);
 		}
-	} else if ((!(ndlp->nlp_type & NLP_FABRIC) &&
-		((ndlp->nlp_type & NLP_FCP_TARGET) ||
-		(ndlp->nlp_type & NLP_NVME_TARGET) ||
-		(vport->fc_flag & FC_PT2PT))) ||
-		(ndlp->nlp_state == NLP_STE_ADISC_ISSUE)) {
-		/* Only try to re-login if this is NOT a Fabric Node
-		 * AND the remote NPORT is a FCP/NVME Target or we
-		 * are in pt2pt mode. NLP_STE_ADISC_ISSUE is a special
-		 * case for LOGO as a response to ADISC behavior.
-		 */
+	} else {
+		lpfc_printf_vlog(vport, KERN_INFO,
+				 LOG_NODE | LOG_ELS | LOG_DISCOVERY,
+				 "3203 LOGO recover nport x%06x state x%x "
+				 "ntype x%x fc_flag x%x\n",
+				 ndlp->nlp_DID, ndlp->nlp_state,
+				 ndlp->nlp_type, vport->fc_flag);
+
+		/* Special cases for rports that recover post LOGO. */
+		if ((!(ndlp->nlp_type == NLP_FABRIC) &&
+		     (ndlp->nlp_type & (NLP_FCP_TARGET | NLP_NVME_TARGET) ||
+		      vport->fc_flag & FC_PT2PT)) ||
+		    (ndlp->nlp_state >= NLP_STE_ADISC_ISSUE ||
+		     ndlp->nlp_state <= NLP_STE_PRLI_ISSUE)) {
 			mod_timer(&ndlp->nlp_delayfunc,
 				  jiffies + msecs_to_jiffies(1000 * 1));
 			spin_lock_irq(&ndlp->lock);
 			ndlp->nlp_flag |= NLP_DELAY_TMO;
 			spin_unlock_irq(&ndlp->lock);
-
 			ndlp->nlp_last_elscmd = ELS_CMD_PLOGI;
+			lpfc_printf_vlog(vport, KERN_INFO,
+					 LOG_NODE | LOG_ELS | LOG_DISCOVERY,
+					 "3204 Start nlpdelay on DID x%06x "
+					 "nflag x%x lastels x%x ref cnt %u",
+					 ndlp->nlp_DID, ndlp->nlp_flag,
+					 ndlp->nlp_last_elscmd,
+					 kref_read(&ndlp->kref));
+		}
 	}
 out:
 	/* Unregister from backend, could have been skipped due to ADISC */
@@ -1854,7 +1865,6 @@ lpfc_rcv_logo_reglogin_issue(struct lpfc_vport *vport,
 	struct lpfc_iocbq *cmdiocb = (struct lpfc_iocbq *) arg;
 	LPFC_MBOXQ_t	  *mb;
 	LPFC_MBOXQ_t	  *nextmb;
-	struct lpfc_nodelist *ns_ndlp;
 
 	cmdiocb = (struct lpfc_iocbq *) arg;
 
@@ -1882,13 +1892,6 @@ lpfc_rcv_logo_reglogin_issue(struct lpfc_vport *vport,
 	}
 	spin_unlock_irq(&phba->hbalock);
 
-	/* software abort if any GID_FT is outstanding */
-	if (vport->cfg_enable_fc4_type != LPFC_ENABLE_FCP) {
-		ns_ndlp = lpfc_findnode_did(vport, NameServer_DID);
-		if (ns_ndlp)
-			lpfc_els_abort(phba, ns_ndlp);
-	}
-
 	lpfc_rcv_logo(vport, ndlp, cmdiocb, ELS_CMD_LOGO);
 	return ndlp->nlp_state;
 }