[SCSI] lpfc 8.3.34: Add LOGO support after ABTS compliance

Make compliant with FC specs by sending LOGO after ABTS timeouts

Signed-off-by: James Smart <james.smart@emulex.com>
Signed-off-by: James Bottomley <JBottomley@Parallels.com>

1 files changed, 115 insertions, 22 deletions

diff --git a/drivers/scsi/lpfc/lpfc_els.c b/drivers/scsi/lpfc/lpfc_els.c
index 3c624a5494a..e0e0dd0a5df 100644
--- a/drivers/scsi/lpfc/lpfc_els.c
+++ b/drivers/scsi/lpfc/lpfc_els.c
@@ -2385,6 +2385,8 @@ lpfc_cmpl_els_logo(struct lpfc_hba *phba, struct lpfc_iocbq *cmdiocb,
 	IOCB_t *irsp;
 	struct lpfc_sli *psli;
 	struct lpfcMboxq *mbox;
+	unsigned long flags;
+	uint32_t skip_recovery = 0;
 
 	psli = &phba->sli;
 	/* we pass cmdiocb to state machine which needs rspiocb as well */
@@ -2399,47 +2401,52 @@ lpfc_cmpl_els_logo(struct lpfc_hba *phba, struct lpfc_iocbq *cmdiocb,
 		"LOGO cmpl:       status:x%x/x%x did:x%x",
 		irsp->ulpStatus, irsp->un.ulpWord[4],
 		ndlp->nlp_DID);
+
 	/* LOGO completes to NPort <nlp_DID> */
 	lpfc_printf_vlog(vport, KERN_INFO, LOG_ELS,
 			 "0105 LOGO completes to NPort x%x "
 			 "Data: x%x x%x x%x x%x\n",
 			 ndlp->nlp_DID, irsp->ulpStatus, irsp->un.ulpWord[4],
 			 irsp->ulpTimeout, vport->num_disc_nodes);
-	/* Check to see if link went down during discovery */
-	if (lpfc_els_chk_latt(vport))
+
+	if (lpfc_els_chk_latt(vport)) {
+		skip_recovery = 1;
 		goto out;
+	}
 
+	/* Check to see if link went down during discovery */
 	if (ndlp->nlp_flag & NLP_TARGET_REMOVE) {
 	        /* NLP_EVT_DEVICE_RM should unregister the RPI
 		 * which should abort all outstanding IOs.
 		 */
 		lpfc_disc_state_machine(vport, ndlp, cmdiocb,
 					NLP_EVT_DEVICE_RM);
+		skip_recovery = 1;
 		goto out;
 	}
 
 	if (irsp->ulpStatus) {
 		/* Check for retry */
-		if (lpfc_els_retry(phba, cmdiocb, rspiocb))
+		if (lpfc_els_retry(phba, cmdiocb, rspiocb)) {
 			/* ELS command is being retried */
+			skip_recovery = 1;
 			goto out;
+		}
 		/* LOGO failed */
 		lpfc_printf_vlog(vport, KERN_ERR, LOG_ELS,
 				 "2756 LOGO failure DID:%06X Status:x%x/x%x\n",
 				 ndlp->nlp_DID, irsp->ulpStatus,
 				 irsp->un.ulpWord[4]);
 		/* Do not call DSM for lpfc_els_abort'ed ELS cmds */
-		if (lpfc_error_lost_link(irsp))
+		if (lpfc_error_lost_link(irsp)) {
+			skip_recovery = 1;
 			goto out;
-		else
-			lpfc_disc_state_machine(vport, ndlp, cmdiocb,
-						NLP_EVT_CMPL_LOGO);
-	} else
-		/* Good status, call state machine.
-		 * This will unregister the rpi if needed.
-		 */
-		lpfc_disc_state_machine(vport, ndlp, cmdiocb,
-					NLP_EVT_CMPL_LOGO);
+		}
+	}
+
+	/* Call state machine. This will unregister the rpi if needed. */
+	lpfc_disc_state_machine(vport, ndlp, cmdiocb, NLP_EVT_CMPL_LOGO);
+
 out:
 	lpfc_els_free_iocb(phba, cmdiocb);
 	/* If we are in pt2pt mode, we could rcv new S_ID on PLOGI */
@@ -2454,9 +2461,30 @@ out:
 			if (lpfc_sli_issue_mbox(phba, mbox, MBX_NOWAIT) ==
 				MBX_NOT_FINISHED) {
 				mempool_free(mbox, phba->mbox_mem_pool);
+				skip_recovery = 1;
 			}
 		}
 	}
+
+	/*
+	 * If the node is a target, the handling attempts to recover the port.
+	 * For any other port type, the rpi is unregistered as an implicit
+	 * LOGO.
+	 */
+	if ((ndlp->nlp_type & NLP_FCP_TARGET) && (skip_recovery == 0)) {
+		lpfc_cancel_retry_delay_tmo(vport, ndlp);
+		spin_lock_irqsave(shost->host_lock, flags);
+		ndlp->nlp_flag |= NLP_NPR_2B_DISC;
+		spin_unlock_irqrestore(shost->host_lock, flags);
+
+		lpfc_printf_vlog(vport, KERN_INFO, LOG_ELS,
+				 "3187 LOGO completes to NPort x%x: Start "
+				 "Recovery Data: x%x x%x x%x x%x\n",
+				 ndlp->nlp_DID, irsp->ulpStatus,
+				 irsp->un.ulpWord[4], irsp->ulpTimeout,
+				 vport->num_disc_nodes);
+		lpfc_disc_start(vport);
+	}
 	return;
 }
 
@@ -2519,10 +2547,27 @@ lpfc_issue_els_logo(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp,
 		"Issue LOGO:      did:x%x",
 		ndlp->nlp_DID, 0, 0);
 
+	/*
+	 * If we are issuing a LOGO, we may try to recover the remote NPort
+	 * by issuing a PLOGI later. Even though we issue ELS cmds by the
+	 * VPI, if we have a valid RPI, and that RPI gets unreg'ed while
+	 * that ELS command is in-flight, the HBA returns a IOERR_INVALID_RPI
+	 * for that ELS cmd. To avoid this situation, lets get rid of the
+	 * RPI right now, before any ELS cmds are sent.
+	 */
+	spin_lock_irq(shost->host_lock);
+	ndlp->nlp_flag |= NLP_ISSUE_LOGO;
+	spin_unlock_irq(shost->host_lock);
+	if (lpfc_unreg_rpi(vport, ndlp)) {
+		lpfc_els_free_iocb(phba, elsiocb);
+		return 0;
+	}
+
 	phba->fc_stat.elsXmitLOGO++;
 	elsiocb->iocb_cmpl = lpfc_cmpl_els_logo;
 	spin_lock_irq(shost->host_lock);
 	ndlp->nlp_flag |= NLP_LOGO_SND;
+	ndlp->nlp_flag &= ~NLP_ISSUE_LOGO;
 	spin_unlock_irq(shost->host_lock);
 	rc = lpfc_sli_issue_iocb(phba, LPFC_ELS_RING, elsiocb, 0);
 
@@ -2938,7 +2983,7 @@ lpfc_els_retry_delay_handler(struct lpfc_nodelist *ndlp)
 	case ELS_CMD_LOGO:
 		if (!lpfc_issue_els_logo(vport, ndlp, retry)) {
 			ndlp->nlp_prev_state = ndlp->nlp_state;
-			lpfc_nlp_set_state(vport, ndlp, NLP_STE_NPR_NODE);
+			lpfc_nlp_set_state(vport, ndlp, NLP_STE_LOGO_ISSUE);
 		}
 		break;
 	case ELS_CMD_FDISC:
@@ -3291,7 +3336,7 @@ lpfc_els_retry(struct lpfc_hba *phba, struct lpfc_iocbq *cmdiocb,
 			return 1;
 		case ELS_CMD_LOGO:
 			ndlp->nlp_prev_state = ndlp->nlp_state;
-			lpfc_nlp_set_state(vport, ndlp, NLP_STE_NPR_NODE);
+			lpfc_nlp_set_state(vport, ndlp, NLP_STE_LOGO_ISSUE);
 			lpfc_issue_els_logo(vport, ndlp, cmdiocb->retry);
 			return 1;
 		}
@@ -3551,13 +3596,17 @@ lpfc_mbx_cmpl_dflt_rpi(struct lpfc_hba *phba, LPFC_MBOXQ_t *pmb)
 	lpfc_mbuf_free(phba, mp->virt, mp->phys);
 	kfree(mp);
 	mempool_free(pmb, phba->mbox_mem_pool);
-	if (ndlp && NLP_CHK_NODE_ACT(ndlp)) {
-		lpfc_nlp_put(ndlp);
-		/* This is the end of the default RPI cleanup logic for this
-		 * ndlp. If no other discovery threads are using this ndlp.
-		 * we should free all resources associated with it.
-		 */
-		lpfc_nlp_not_used(ndlp);
+	if (ndlp) {
+		if (NLP_CHK_NODE_ACT(ndlp)) {
+			lpfc_nlp_put(ndlp);
+			/* This is the end of the default RPI cleanup logic for
+			 * this ndlp. If no other discovery threads are using
+			 * this ndlp, free all resources associated with it.
+			 */
+			lpfc_nlp_not_used(ndlp);
+		} else {
+			lpfc_drop_node(ndlp->vport, ndlp);
+		}
 	}
 
 	return;
@@ -8003,3 +8052,47 @@ lpfc_sli4_els_xri_aborted(struct lpfc_hba *phba,
 	spin_unlock_irqrestore(&phba->hbalock, iflag);
 	return;
 }
+
+/* lpfc_sli_abts_recover_port - Recover a port that failed a BLS_ABORT req.
+ * @vport: pointer to virtual port object.
+ * @ndlp: nodelist pointer for the impacted node.
+ *
+ * The driver calls this routine in response to an SLI4 XRI ABORT CQE
+ * or an SLI3 ASYNC_STATUS_CN event from the port.  For either event,
+ * the driver is required to send a LOGO to the remote node before it
+ * attempts to recover its login to the remote node.
+ */
+void
+lpfc_sli_abts_recover_port(struct lpfc_vport *vport,
+			   struct lpfc_nodelist *ndlp)
+{
+	struct Scsi_Host *shost;
+	struct lpfc_hba *phba;
+	unsigned long flags = 0;
+
+	shost = lpfc_shost_from_vport(vport);
+	phba = vport->phba;
+	if (ndlp->nlp_state != NLP_STE_MAPPED_NODE) {
+		lpfc_printf_log(phba, KERN_INFO,
+				LOG_SLI, "3093 No rport recovery needed. "
+				"rport in state 0x%x\n", ndlp->nlp_state);
+		return;
+	}
+	lpfc_printf_log(phba, KERN_INFO, LOG_SLI,
+			"3094 Start rport recovery on shost id 0x%x "
+			"fc_id 0x%06x vpi 0x%x rpi 0x%x state 0x%x "
+			"flags 0x%x\n",
+			shost->host_no, ndlp->nlp_DID,
+			vport->vpi, ndlp->nlp_rpi, ndlp->nlp_state,
+			ndlp->nlp_flag);
+	/*
+	 * The rport is not responding.  Remove the FCP-2 flag to prevent
+	 * an ADISC in the follow-up recovery code.
+	 */
+	spin_lock_irqsave(shost->host_lock, flags);
+	ndlp->nlp_fcp_info &= ~NLP_FCP_2_DEVICE;
+	spin_unlock_irqrestore(shost->host_lock, flags);
+	lpfc_issue_els_logo(vport, ndlp, 0);
+	lpfc_nlp_set_state(vport, ndlp, NLP_STE_LOGO_ISSUE);
+}
+