[PATCH 0/9] lpfc: Update lpfc to revision 14.2.0.4

[PATCH 0/9] lpfc: Update lpfc to revision 14.2.0.4

[James Smart]

Update lpfc to revision 14.2.0.4

This patch set contains mainly fixes. Also adds some additional logging
for NVMe command completions when errors, and enabling of nvme-fc flush
handling on async event cmds.

The patches were cut against Martin's 5.19/scsi-queue tree

James Smart (9):
  lpfc: Correct BDE type for XMIT_SEQ64_WQE in ct_reject_event
  lpfc: Resolve some cleanup issues following abort path refactoring
  lpfc: Resolve some cleanup issues following SLI path refactoring
  lpfc: Address null pointer dereference after starget_to_rport()
  lpfc: Resolve null ptr dereference after an ELS LOGO is aborted
  lpfc: Fix port stuck in bypassed state after lip in PT2PT topology
  lpfc: Add more logging of cmd and cqe information for aborted NVME
    cmds
  lpfc: Allow reduced polling rate for nvme_admin_async_event cmd
    completion
  lpfc: Update lpfc version to 14.2.0.4

 drivers/scsi/lpfc/lpfc_crtn.h      |  4 +--
 drivers/scsi/lpfc/lpfc_ct.c        |  2 +-
 drivers/scsi/lpfc/lpfc_els.c       | 21 ++++++------
 drivers/scsi/lpfc/lpfc_hw4.h       |  3 ++
 drivers/scsi/lpfc/lpfc_init.c      |  2 +-
 drivers/scsi/lpfc/lpfc_nportdisc.c |  3 +-
 drivers/scsi/lpfc/lpfc_nvme.c      | 52 +++++++++++++++++++++++-------
 drivers/scsi/lpfc/lpfc_scsi.c      |  6 ++++
 drivers/scsi/lpfc/lpfc_sli.c       | 25 +++++++-------
 drivers/scsi/lpfc/lpfc_version.h   |  2 +-
 10 files changed, 76 insertions(+), 44 deletions(-)

-- 
2.26.2

[PATCH 1/9] lpfc: Correct BDE type for XMIT_SEQ64_WQE in ct_reject_event

[James Smart]

A previous commit assumed all XMIT_SEQ64_WQEs are prepped with the correct
BDE type in word 0-2.  However, ct_reject_event routine was missed and is
still filling out the incorrect BDE type.

Fix in ct_reject_event routine so that type BUFF_TYPE_BDE_64 is set instead
of BUFF_TYPE_BLP_64.

Fixes: 596fc8adb171 ("scsi: lpfc: Fix dmabuf ptr assignment in lpfc_ct_reject_event()")
Co-developed-by: Justin Tee <justin.tee@broadcom.com>
Signed-off-by: Justin Tee <justin.tee@broadcom.com>
Signed-off-by: James Smart <jsmart2021@gmail.com>
---
 drivers/scsi/lpfc/lpfc_ct.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/scsi/lpfc/lpfc_ct.c b/drivers/scsi/lpfc/lpfc_ct.c
index 9d36b20fb878..13dfe285493d 100644
--- a/drivers/scsi/lpfc/lpfc_ct.c
+++ b/drivers/scsi/lpfc/lpfc_ct.c
@@ -197,7 +197,7 @@ lpfc_ct_reject_event(struct lpfc_nodelist *ndlp,
 	memset(bpl, 0, sizeof(struct ulp_bde64));
 	bpl->addrHigh = le32_to_cpu(putPaddrHigh(mp->phys));
 	bpl->addrLow = le32_to_cpu(putPaddrLow(mp->phys));
-	bpl->tus.f.bdeFlags = BUFF_TYPE_BLP_64;
+	bpl->tus.f.bdeFlags = BUFF_TYPE_BDE_64;
 	bpl->tus.f.bdeSize = (LPFC_CT_PREAMBLE - 4);
 	bpl->tus.w = le32_to_cpu(bpl->tus.w);
 
-- 
2.26.2

[PATCH 2/9] lpfc: Resolve some cleanup issues following abort path refactoring

[James Smart]

Following refactoring and consolidation of abort paths,

- lpfc_sli4_abort_fcp_cmpl and lpfc_sli_abort_fcp_cmpl are combined into
  a single generic lpfc_sli_abort_fcp_cmpl routine.  Thus, remove
  extraneous lpfc_sli4_abort_fcp_cmpl prototype declaration.

- lpfc_nvme_abort_fcreq_cmpl abort completion routine is called with a
  mismatched argument type.  This may result in misleading log message
  content.  Update to the correct argument type of lpfc_iocbq instead
  of lpfc_wcqe_complete.  The lpfc_wcqe_complete should be derived from
  the lpfc_iocbq structure.

Fixes: 31a59f75702f ("scsi: lpfc: SLI path split: Refactor Abort paths")
Cc: <stable@vger.kernel.org> # v5.18
Co-developed-by: Justin Tee <justin.tee@broadcom.com>
Signed-off-by: Justin Tee <justin.tee@broadcom.com>
Signed-off-by: James Smart <jsmart2021@gmail.com>
---
 drivers/scsi/lpfc/lpfc_crtn.h | 4 +---
 drivers/scsi/lpfc/lpfc_nvme.c | 6 ++++--
 2 files changed, 5 insertions(+), 5 deletions(-)

diff --git a/drivers/scsi/lpfc/lpfc_crtn.h b/drivers/scsi/lpfc/lpfc_crtn.h
index b1be0dd0337a..f5d74958b664 100644
--- a/drivers/scsi/lpfc/lpfc_crtn.h
+++ b/drivers/scsi/lpfc/lpfc_crtn.h
@@ -420,8 +420,6 @@ int lpfc_sli_issue_iocb_wait(struct lpfc_hba *, uint32_t,
 			     uint32_t);
 void lpfc_sli_abort_fcp_cmpl(struct lpfc_hba *, struct lpfc_iocbq *,
 			     struct lpfc_iocbq *);
-void lpfc_sli4_abort_fcp_cmpl(struct lpfc_hba *h, struct lpfc_iocbq *i,
-			      struct lpfc_wcqe_complete *w);
 
 void lpfc_sli_free_hbq(struct lpfc_hba *, struct hbq_dmabuf *);
 
@@ -630,7 +628,7 @@ void lpfc_nvmet_invalidate_host(struct lpfc_hba *phba,
 			struct lpfc_nodelist *ndlp);
 void lpfc_nvme_abort_fcreq_cmpl(struct lpfc_hba *phba,
 				struct lpfc_iocbq *cmdiocb,
-				struct lpfc_wcqe_complete *abts_cmpl);
+				struct lpfc_iocbq *rspiocb);
 void lpfc_create_multixri_pools(struct lpfc_hba *phba);
 void lpfc_create_destroy_pools(struct lpfc_hba *phba);
 void lpfc_move_xri_pvt_to_pbl(struct lpfc_hba *phba, u32 hwqid);
diff --git a/drivers/scsi/lpfc/lpfc_nvme.c b/drivers/scsi/lpfc/lpfc_nvme.c
index 335e90633933..88fa630ab93a 100644
--- a/drivers/scsi/lpfc/lpfc_nvme.c
+++ b/drivers/scsi/lpfc/lpfc_nvme.c
@@ -1787,7 +1787,7 @@ lpfc_nvme_fcp_io_submit(struct nvme_fc_local_port *pnvme_lport,
  * lpfc_nvme_abort_fcreq_cmpl - Complete an NVME FCP abort request.
  * @phba: Pointer to HBA context object
  * @cmdiocb: Pointer to command iocb object.
- * @abts_cmpl: Pointer to wcqe complete object.
+ * @rspiocb: Pointer to response iocb object.
  *
  * This is the callback function for any NVME FCP IO that was aborted.
  *
@@ -1796,8 +1796,10 @@ lpfc_nvme_fcp_io_submit(struct nvme_fc_local_port *pnvme_lport,
  **/
 void
 lpfc_nvme_abort_fcreq_cmpl(struct lpfc_hba *phba, struct lpfc_iocbq *cmdiocb,
-			   struct lpfc_wcqe_complete *abts_cmpl)
+			   struct lpfc_iocbq *rspiocb)
 {
+	struct lpfc_wcqe_complete *abts_cmpl = &rspiocb->wcqe_cmpl;
+
 	lpfc_printf_log(phba, KERN_INFO, LOG_NVME,
 			"6145 ABORT_XRI_CN completing on rpi x%x "
 			"original iotag x%x, abort cmd iotag x%x "
-- 
2.26.2

[PATCH 3/9] lpfc: Resolve some cleanup issues following SLI path refactoring

[James Smart]

Following refactoring and consolidation in SLI processing, fixup some
minor issues related to SLI path.

- Correct the setting of LPFC_EXCHANGE_BUSY flag in response iocb.
- Fix some typographical errors.
- Fix duplicate log messages.

Fixes: 1b64aa9eae28 ("scsi: lpfc: SLI path split: Refactor fast and slow paths to native SLI4")
Cc: <stable@vger.kernel.org> # v5.18
Co-developed-by: Justin Tee <justin.tee@broadcom.com>
Signed-off-by: Justin Tee <justin.tee@broadcom.com>
Signed-off-by: James Smart <jsmart2021@gmail.com>
---
 drivers/scsi/lpfc/lpfc_init.c |  2 +-
 drivers/scsi/lpfc/lpfc_sli.c  | 25 ++++++++++++-------------
 2 files changed, 13 insertions(+), 14 deletions(-)

diff --git a/drivers/scsi/lpfc/lpfc_init.c b/drivers/scsi/lpfc/lpfc_init.c
index 93b94c64518d..750dd1e9f2cc 100644
--- a/drivers/scsi/lpfc/lpfc_init.c
+++ b/drivers/scsi/lpfc/lpfc_init.c
@@ -12188,7 +12188,7 @@ lpfc_sli_enable_msi(struct lpfc_hba *phba)
 	rc = pci_enable_msi(phba->pcidev);
 	if (!rc)
 		lpfc_printf_log(phba, KERN_INFO, LOG_INIT,
-				"0462 PCI enable MSI mode success.\n");
+				"0012 PCI enable MSI mode success.\n");
 	else {
 		lpfc_printf_log(phba, KERN_INFO, LOG_INIT,
 				"0471 PCI enable MSI mode failed (%d)\n", rc);
diff --git a/drivers/scsi/lpfc/lpfc_sli.c b/drivers/scsi/lpfc/lpfc_sli.c
index 79d2ef5f0f05..dd96fc09105d 100644
--- a/drivers/scsi/lpfc/lpfc_sli.c
+++ b/drivers/scsi/lpfc/lpfc_sli.c
@@ -1930,7 +1930,7 @@ lpfc_issue_cmf_sync_wqe(struct lpfc_hba *phba, u32 ms, u64 total)
 	sync_buf = __lpfc_sli_get_iocbq(phba);
 	if (!sync_buf) {
 		lpfc_printf_log(phba, KERN_ERR, LOG_CGN_MGMT,
-				"6213 No available WQEs for CMF_SYNC_WQE\n");
+				"6244 No available WQEs for CMF_SYNC_WQE\n");
 		ret_val = ENOMEM;
 		goto out_unlock;
 	}
@@ -3805,7 +3805,7 @@ lpfc_sli_process_sol_iocb(struct lpfc_hba *phba, struct lpfc_sli_ring *pring,
 						set_job_ulpword4(cmdiocbp,
 								 IOERR_ABORT_REQUESTED);
 						/*
-						 * For SLI4, irsiocb contains
+						 * For SLI4, irspiocb contains
 						 * NO_XRI in sli_xritag, it
 						 * shall not affect releasing
 						 * sgl (xri) process.
@@ -3823,7 +3823,7 @@ lpfc_sli_process_sol_iocb(struct lpfc_hba *phba, struct lpfc_sli_ring *pring,
 					}
 				}
 			}
-			(cmdiocbp->cmd_cmpl) (phba, cmdiocbp, saveq);
+			cmdiocbp->cmd_cmpl(phba, cmdiocbp, saveq);
 		} else
 			lpfc_sli_release_iocbq(phba, cmdiocbp);
 	} else {
@@ -4063,8 +4063,7 @@ lpfc_sli_handle_fast_ring_event(struct lpfc_hba *phba,
 				cmdiocbq->cmd_flag &= ~LPFC_DRIVER_ABORTED;
 			if (cmdiocbq->cmd_cmpl) {
 				spin_unlock_irqrestore(&phba->hbalock, iflag);
-				(cmdiocbq->cmd_cmpl)(phba, cmdiocbq,
-						      &rspiocbq);
+				cmdiocbq->cmd_cmpl(phba, cmdiocbq, &rspiocbq);
 				spin_lock_irqsave(&phba->hbalock, iflag);
 			}
 			break;
@@ -10288,7 +10287,7 @@ __lpfc_sli_issue_iocb_s3(struct lpfc_hba *phba, uint32_t ring_number,
  * @flag: Flag indicating if this command can be put into txq.
  *
  * __lpfc_sli_issue_fcp_io_s3 is wrapper function to invoke lockless func to
- * send  an iocb command to an HBA with SLI-4 interface spec.
+ * send  an iocb command to an HBA with SLI-3 interface spec.
  *
  * This function takes the hbalock before invoking the lockless version.
  * The function will return success after it successfully submit the wqe to
@@ -12740,7 +12739,7 @@ lpfc_sli_wake_iocb_wait(struct lpfc_hba *phba,
 		cmdiocbq->cmd_cmpl = cmdiocbq->wait_cmd_cmpl;
 		cmdiocbq->wait_cmd_cmpl = NULL;
 		if (cmdiocbq->cmd_cmpl)
-			(cmdiocbq->cmd_cmpl)(phba, cmdiocbq, NULL);
+			cmdiocbq->cmd_cmpl(phba, cmdiocbq, NULL);
 		else
 			lpfc_sli_release_iocbq(phba, cmdiocbq);
 		return;
@@ -12754,9 +12753,9 @@ lpfc_sli_wake_iocb_wait(struct lpfc_hba *phba,
 
 	/* Set the exchange busy flag for task management commands */
 	if ((cmdiocbq->cmd_flag & LPFC_IO_FCP) &&
-		!(cmdiocbq->cmd_flag & LPFC_IO_LIBDFC)) {
+	    !(cmdiocbq->cmd_flag & LPFC_IO_LIBDFC)) {
 		lpfc_cmd = container_of(cmdiocbq, struct lpfc_io_buf,
-			cur_iocbq);
+					cur_iocbq);
 		if (rspiocbq && (rspiocbq->cmd_flag & LPFC_EXCHANGE_BUSY))
 			lpfc_cmd->flags |= LPFC_SBUF_XBUSY;
 		else
@@ -13896,7 +13895,7 @@ void lpfc_sli4_els_xri_abort_event_proc(struct lpfc_hba *phba)
  * @irspiocbq: Pointer to work-queue completion queue entry.
  *
  * This routine handles an ELS work-queue completion event and construct
- * a pseudo response ELS IODBQ from the SLI4 ELS WCQE for the common
+ * a pseudo response ELS IOCBQ from the SLI4 ELS WCQE for the common
  * discovery engine to handle.
  *
  * Return: Pointer to the receive IOCBQ, NULL otherwise.
@@ -13940,7 +13939,7 @@ lpfc_sli4_els_preprocess_rspiocbq(struct lpfc_hba *phba,
 
 	if (bf_get(lpfc_wcqe_c_xb, wcqe)) {
 		spin_lock_irqsave(&phba->hbalock, iflags);
-		cmdiocbq->cmd_flag |= LPFC_EXCHANGE_BUSY;
+		irspiocbq->cmd_flag |= LPFC_EXCHANGE_BUSY;
 		spin_unlock_irqrestore(&phba->hbalock, iflags);
 	}
 
@@ -14799,7 +14798,7 @@ lpfc_sli4_fp_handle_fcp_wcqe(struct lpfc_hba *phba, struct lpfc_queue *cq,
 		/* Pass the cmd_iocb and the wcqe to the upper layer */
 		memcpy(&cmdiocbq->wcqe_cmpl, wcqe,
 		       sizeof(struct lpfc_wcqe_complete));
-		(cmdiocbq->cmd_cmpl)(phba, cmdiocbq, cmdiocbq);
+		cmdiocbq->cmd_cmpl(phba, cmdiocbq, cmdiocbq);
 	} else {
 		lpfc_printf_log(phba, KERN_WARNING, LOG_SLI,
 				"0375 FCP cmdiocb not callback function "
@@ -18956,7 +18955,7 @@ lpfc_sli4_send_seq_to_ulp(struct lpfc_vport *vport,
 
 	/* Free iocb created in lpfc_prep_seq */
 	list_for_each_entry_safe(curr_iocb, next_iocb,
-		&iocbq->list, list) {
+				 &iocbq->list, list) {
 		list_del_init(&curr_iocb->list);
 		lpfc_sli_release_iocbq(phba, curr_iocb);
 	}
-- 
2.26.2

[PATCH 4/9] lpfc: Address null pointer dereference after starget_to_rport()

[James Smart]

Calls to starget_to_rport() may return null.  Add check for null rport
before dereference.

Fixes: bb21fc9911ee ("scsi: lpfc: Use fc_block_rport()")
Cc: <stable@vger.kernel.org> # v5.18
Co-developed-by: Justin Tee <justin.tee@broadcom.com>
Signed-off-by: Justin Tee <justin.tee@broadcom.com>
Signed-off-by: James Smart <jsmart2021@gmail.com>
---
 drivers/scsi/lpfc/lpfc_scsi.c | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/drivers/scsi/lpfc/lpfc_scsi.c b/drivers/scsi/lpfc/lpfc_scsi.c
index f5f4409e24cd..cb5e4e63ac44 100644
--- a/drivers/scsi/lpfc/lpfc_scsi.c
+++ b/drivers/scsi/lpfc/lpfc_scsi.c
@@ -6060,6 +6060,9 @@ lpfc_device_reset_handler(struct scsi_cmnd *cmnd)
 	int status;
 	u32 logit = LOG_FCP;
 
+	if (!rport)
+		return FAILED;
+
 	rdata = rport->dd_data;
 	if (!rdata || !rdata->pnode) {
 		lpfc_printf_vlog(vport, KERN_ERR, LOG_TRACE_EVENT,
@@ -6138,6 +6141,9 @@ lpfc_target_reset_handler(struct scsi_cmnd *cmnd)
 	unsigned long flags;
 	DECLARE_WAIT_QUEUE_HEAD_ONSTACK(waitq);
 
+	if (!rport)
+		return FAILED;
+
 	rdata = rport->dd_data;
 	if (!rdata || !rdata->pnode) {
 		lpfc_printf_vlog(vport, KERN_ERR, LOG_TRACE_EVENT,
-- 
2.26.2

[PATCH 5/9] lpfc: Resolve null ptr dereference after an ELS LOGO is aborted

[James Smart]

A use-after-free crash can occur after an ELS LOGO is aborted.

Specifically, a nodelist structure is freed and then
ndlp->vport->cfg_log_verbose is dereferenced in lpfc_nlp_get when the
discovery state machine is mistakenly called a second time with
NLP_EVT_DEVICE_RM argument.

Rework lpfc_cmpl_els_logo to prevent the duplicate calls to release a
nodelist structure.

Co-developed-by: Justin Tee <justin.tee@broadcom.com>
Signed-off-by: Justin Tee <justin.tee@broadcom.com>
Signed-off-by: James Smart <jsmart2021@gmail.com>
---
 drivers/scsi/lpfc/lpfc_els.c | 21 +++++++++------------
 1 file changed, 9 insertions(+), 12 deletions(-)

diff --git a/drivers/scsi/lpfc/lpfc_els.c b/drivers/scsi/lpfc/lpfc_els.c
index 51c505d15410..57e189f62e42 100644
--- a/drivers/scsi/lpfc/lpfc_els.c
+++ b/drivers/scsi/lpfc/lpfc_els.c
@@ -2998,10 +2998,7 @@ lpfc_cmpl_els_logo(struct lpfc_hba *phba, struct lpfc_iocbq *cmdiocb,
 				 ndlp->nlp_DID, ulp_status,
 				 ulp_word4);
 
-		/* Call NLP_EVT_DEVICE_RM if link is down or LOGO is aborted */
 		if (lpfc_error_lost_link(ulp_status, ulp_word4)) {
-			lpfc_disc_state_machine(vport, ndlp, cmdiocb,
-						NLP_EVT_DEVICE_RM);
 			skip_recovery = 1;
 			goto out;
 		}
@@ -3021,18 +3018,10 @@ lpfc_cmpl_els_logo(struct lpfc_hba *phba, struct lpfc_iocbq *cmdiocb,
 		spin_unlock_irq(&ndlp->lock);
 		lpfc_disc_state_machine(vport, ndlp, cmdiocb,
 					NLP_EVT_DEVICE_RM);
-		lpfc_els_free_iocb(phba, cmdiocb);
-		lpfc_nlp_put(ndlp);
-
-		/* Presume the node was released. */
-		return;
+		goto out_rsrc_free;
 	}
 
 out:
-	/* Driver is done with the IO.  */
-	lpfc_els_free_iocb(phba, cmdiocb);
-	lpfc_nlp_put(ndlp);
-
 	/* At this point, the LOGO processing is complete. NOTE: For a
 	 * pt2pt topology, we are assuming the NPortID will only change
 	 * on link up processing. For a LOGO / PLOGI initiated by the
@@ -3059,6 +3048,10 @@ lpfc_cmpl_els_logo(struct lpfc_hba *phba, struct lpfc_iocbq *cmdiocb,
 				 ndlp->nlp_DID, ulp_status,
 				 ulp_word4, tmo,
 				 vport->num_disc_nodes);
+
+		lpfc_els_free_iocb(phba, cmdiocb);
+		lpfc_nlp_put(ndlp);
+
 		lpfc_disc_start(vport);
 		return;
 	}
@@ -3075,6 +3068,10 @@ lpfc_cmpl_els_logo(struct lpfc_hba *phba, struct lpfc_iocbq *cmdiocb,
 		lpfc_disc_state_machine(vport, ndlp, cmdiocb,
 					NLP_EVT_DEVICE_RM);
 	}
+out_rsrc_free:
+	/* Driver is done with the I/O. */
+	lpfc_els_free_iocb(phba, cmdiocb);
+	lpfc_nlp_put(ndlp);
 }
 
 /**
-- 
2.26.2

[PATCH 6/9] lpfc: Fix port stuck in bypassed state after lip in PT2PT topology

[James Smart]

After issuing a lip, a specific target vendor does not ACC the FLOGI that
lpfc sends.  However, it does send its own FLOGI that lpfc ACCs.  The
target then establishes the port IDs by sending a PLOGI.  Lpfc PLOGI_ACCs
and starts the RPI registration for DID 0x000001.  The target then sends
a LOGO to the fabric DID.  Lpfc is currently treating the LOGO from the
fabric DID as a link down and cleans up all the ndlps.  The ndlp for DID
0x000001 is put back into NPR and discovery stops, leaving the port in
stuck in bypassed mode.

Change lpfc behavior such that if a LOGO is received for the fabric DID in
PT2PT topology skip the lpfc_linkdown_port routine and just move the fabric
DID back to NPR.

Co-developed-by: Justin Tee <justin.tee@broadcom.com>
Signed-off-by: Justin Tee <justin.tee@broadcom.com>
Signed-off-by: James Smart <jsmart2021@gmail.com>
---
 drivers/scsi/lpfc/lpfc_nportdisc.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/scsi/lpfc/lpfc_nportdisc.c b/drivers/scsi/lpfc/lpfc_nportdisc.c
index 639f86635127..b86ff9fcdf0c 100644
--- a/drivers/scsi/lpfc/lpfc_nportdisc.c
+++ b/drivers/scsi/lpfc/lpfc_nportdisc.c
@@ -834,7 +834,8 @@ lpfc_rcv_logo(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp,
 		lpfc_nvmet_invalidate_host(phba, ndlp);
 
 	if (ndlp->nlp_DID == Fabric_DID) {
-		if (vport->port_state <= LPFC_FDISC)
+		if (vport->port_state <= LPFC_FDISC ||
+		    vport->fc_flag & FC_PT2PT)
 			goto out;
 		lpfc_linkdown_port(vport);
 		spin_lock_irq(shost->host_lock);
-- 
2.26.2

[PATCH 7/9] lpfc: Add more logging of cmd and cqe information for aborted NVME cmds

[James Smart]

When an NVME command is aborted or completes with an ERSP, log the opcode
and command id fields to help provide more detail on the failed command.

Co-developed-by: Justin Tee <justin.tee@broadcom.com>
Signed-off-by: Justin Tee <justin.tee@broadcom.com>
Signed-off-by: James Smart <jsmart2021@gmail.com>
---
 drivers/scsi/lpfc/lpfc_nvme.c | 35 +++++++++++++++++++++++++++--------
 1 file changed, 27 insertions(+), 8 deletions(-)

diff --git a/drivers/scsi/lpfc/lpfc_nvme.c b/drivers/scsi/lpfc/lpfc_nvme.c
index 88fa630ab93a..9cc22cefcb37 100644
--- a/drivers/scsi/lpfc/lpfc_nvme.c
+++ b/drivers/scsi/lpfc/lpfc_nvme.c
@@ -1065,25 +1065,37 @@ lpfc_nvme_io_cmd_cmpl(struct lpfc_hba *phba, struct lpfc_iocbq *pwqeIn,
 			nCmd->rcv_rsplen = wcqe->parameter;
 			nCmd->status = 0;
 
+			/* Get the NVME cmd details for this unique error. */
+			cp = (struct nvme_fc_cmd_iu *)nCmd->cmdaddr;
+			ep = (struct nvme_fc_ersp_iu *)nCmd->rspaddr;
+
 			/* Check if this is really an ERSP */
 			if (nCmd->rcv_rsplen == LPFC_NVME_ERSP_LEN) {
 				lpfc_ncmd->status = IOSTAT_SUCCESS;
 				lpfc_ncmd->result = 0;
 
 				lpfc_printf_vlog(vport, KERN_INFO, LOG_NVME,
-					 "6084 NVME Completion ERSP: "
-					 "xri %x placed x%x\n",
-					 lpfc_ncmd->cur_iocbq.sli4_xritag,
-					 wcqe->total_data_placed);
+					"6084 NVME FCP_ERR ERSP: "
+					"xri %x placed x%x opcode x%x cmd_id "
+					"x%x cqe_status x%x\n",
+					lpfc_ncmd->cur_iocbq.sli4_xritag,
+					wcqe->total_data_placed,
+					cp->sqe.common.opcode,
+					cp->sqe.common.command_id,
+					ep->cqe.status);
 				break;
 			}
 			lpfc_printf_vlog(vport, KERN_ERR, LOG_TRACE_EVENT,
 					 "6081 NVME Completion Protocol Error: "
 					 "xri %x status x%x result x%x "
-					 "placed x%x\n",
+					 "placed x%x opcode x%x cmd_id x%x, "
+					 "cqe_status x%x\n",
 					 lpfc_ncmd->cur_iocbq.sli4_xritag,
 					 lpfc_ncmd->status, lpfc_ncmd->result,
-					 wcqe->total_data_placed);
+					 wcqe->total_data_placed,
+					 cp->sqe.common.opcode,
+					 cp->sqe.common.command_id,
+					 ep->cqe.status);
 			break;
 		case IOSTAT_LOCAL_REJECT:
 			/* Let fall through to set command final state. */
@@ -1842,6 +1854,7 @@ lpfc_nvme_fcp_abort(struct nvme_fc_local_port *pnvme_lport,
 	struct lpfc_nvme_fcpreq_priv *freqpriv;
 	unsigned long flags;
 	int ret_val;
+	struct nvme_fc_cmd_iu *cp;
 
 	/* Validate pointers. LLDD fault handling with transport does
 	 * have timing races.
@@ -1965,10 +1978,16 @@ lpfc_nvme_fcp_abort(struct nvme_fc_local_port *pnvme_lport,
 		return;
 	}
 
+	/*
+	 * Get Command Id from cmd to plug into response. This
+	 * code is not needed in the next NVME Transport drop.
+	 */
+	cp = (struct nvme_fc_cmd_iu *)lpfc_nbuf->nvmeCmd->cmdaddr;
 	lpfc_printf_vlog(vport, KERN_INFO, LOG_NVME_ABTS,
 			 "6138 Transport Abort NVME Request Issued for "
-			 "ox_id x%x\n",
-			 nvmereq_wqe->sli4_xritag);
+			 "ox_id x%x nvme opcode x%x nvme cmd_id x%x\n",
+			 nvmereq_wqe->sli4_xritag, cp->sqe.common.opcode,
+			 cp->sqe.common.command_id);
 	return;
 
 out_unlock:
-- 
2.26.2

[PATCH 8/9] lpfc: Allow reduced polling rate for nvme_admin_async_event cmd completion

[James Smart]

NVME Asynchronous Event Request commands have no command timeout value per
specifications.

Set WQE option to allow a reduced FLUSH polling rate for I/O error
detection specifically for nvme_admin_async_event commands.

Co-developed-by: Justin Tee <justin.tee@broadcom.com>
Signed-off-by: Justin Tee <justin.tee@broadcom.com>
Signed-off-by: James Smart <jsmart2021@gmail.com>
---
 drivers/scsi/lpfc/lpfc_hw4.h  |  3 +++
 drivers/scsi/lpfc/lpfc_nvme.c | 11 +++++++++--
 2 files changed, 12 insertions(+), 2 deletions(-)

diff --git a/drivers/scsi/lpfc/lpfc_hw4.h b/drivers/scsi/lpfc/lpfc_hw4.h
index 8511369d2cf8..f024415731ac 100644
--- a/drivers/scsi/lpfc/lpfc_hw4.h
+++ b/drivers/scsi/lpfc/lpfc_hw4.h
@@ -4487,6 +4487,9 @@ struct wqe_common {
 #define wqe_sup_SHIFT         6
 #define wqe_sup_MASK          0x00000001
 #define wqe_sup_WORD          word11
+#define wqe_ffrq_SHIFT         6
+#define wqe_ffrq_MASK          0x00000001
+#define wqe_ffrq_WORD          word11
 #define wqe_wqec_SHIFT        7
 #define wqe_wqec_MASK         0x00000001
 #define wqe_wqec_WORD         word11
diff --git a/drivers/scsi/lpfc/lpfc_nvme.c b/drivers/scsi/lpfc/lpfc_nvme.c
index 9cc22cefcb37..cd10ee6482fc 100644
--- a/drivers/scsi/lpfc/lpfc_nvme.c
+++ b/drivers/scsi/lpfc/lpfc_nvme.c
@@ -1207,7 +1207,8 @@ lpfc_nvme_prep_io_cmd(struct lpfc_vport *vport,
 {
 	struct lpfc_hba *phba = vport->phba;
 	struct nvmefc_fcp_req *nCmd = lpfc_ncmd->nvmeCmd;
-	struct lpfc_iocbq *pwqeq = &(lpfc_ncmd->cur_iocbq);
+	struct nvme_common_command *sqe;
+	struct lpfc_iocbq *pwqeq = &lpfc_ncmd->cur_iocbq;
 	union lpfc_wqe128 *wqe = &pwqeq->wqe;
 	uint32_t req_len;
 
@@ -1264,8 +1265,14 @@ lpfc_nvme_prep_io_cmd(struct lpfc_vport *vport,
 		cstat->control_requests++;
 	}
 
-	if (pnode->nlp_nvme_info & NLP_NVME_NSLER)
+	if (pnode->nlp_nvme_info & NLP_NVME_NSLER) {
 		bf_set(wqe_erp, &wqe->generic.wqe_com, 1);
+		sqe = &((struct nvme_fc_cmd_iu *)
+			nCmd->cmdaddr)->sqe.common;
+		if (sqe->opcode == nvme_admin_async_event)
+			bf_set(wqe_ffrq, &wqe->generic.wqe_com, 1);
+	}
+
 	/*
 	 * Finish initializing those WQE fields that are independent
 	 * of the nvme_cmnd request_buffer
-- 
2.26.2

[PATCH 9/9] lpfc: Update lpfc version to 14.2.0.4

[James Smart]

Update lpfc version to 14.2.0.4

Co-developed-by: Justin Tee <justin.tee@broadcom.com>
Signed-off-by: Justin Tee <justin.tee@broadcom.com>
Signed-off-by: James Smart <jsmart2021@gmail.com>
---
 drivers/scsi/lpfc/lpfc_version.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/scsi/lpfc/lpfc_version.h b/drivers/scsi/lpfc/lpfc_version.h
index 4fab79ed58ed..2ab6f7db64d8 100644
--- a/drivers/scsi/lpfc/lpfc_version.h
+++ b/drivers/scsi/lpfc/lpfc_version.h
@@ -20,7 +20,7 @@
  * included with this package.                                     *
  *******************************************************************/
 
-#define LPFC_DRIVER_VERSION "14.2.0.3"
+#define LPFC_DRIVER_VERSION "14.2.0.4"
 #define LPFC_DRIVER_NAME		"lpfc"
 
 /* Used for SLI 2/3 */
-- 
2.26.2

Re: [PATCH 0/9] lpfc: Update lpfc to revision 14.2.0.4

[Martin K. Petersen]

On Fri, 3 Jun 2022 10:43:20 -0700, James Smart wrote:

> Update lpfc to revision 14.2.0.4
> 
> This patch set contains mainly fixes. Also adds some additional logging
> for NVMe command completions when errors, and enabling of nvme-fc flush
> handling on async event cmds.
> 
> The patches were cut against Martin's 5.19/scsi-queue tree
> 
> [...]

Applied to 5.19/scsi-fixes, thanks!

[1/9] lpfc: Correct BDE type for XMIT_SEQ64_WQE in ct_reject_event
      https://git.kernel.org/mkp/scsi/c/44ba9786b673
[2/9] lpfc: Resolve some cleanup issues following abort path refactoring
      https://git.kernel.org/mkp/scsi/c/24e1f056677e
[3/9] lpfc: Resolve some cleanup issues following SLI path refactoring
      https://git.kernel.org/mkp/scsi/c/e27f05147bff
[4/9] lpfc: Address null pointer dereference after starget_to_rport()
      https://git.kernel.org/mkp/scsi/c/6f808bd78e82
[5/9] lpfc: Resolve null ptr dereference after an ELS LOGO is aborted
      https://git.kernel.org/mkp/scsi/c/b1b3440f437b
[6/9] lpfc: Fix port stuck in bypassed state after lip in PT2PT topology
      https://git.kernel.org/mkp/scsi/c/336d63615466
[7/9] lpfc: Add more logging of cmd and cqe information for aborted NVME cmds
      https://git.kernel.org/mkp/scsi/c/ea7bd1f39331
[8/9] lpfc: Allow reduced polling rate for nvme_admin_async_event cmd completion
      https://git.kernel.org/mkp/scsi/c/2e7e9c0c1ec0
[9/9] lpfc: Update lpfc version to 14.2.0.4
      https://git.kernel.org/mkp/scsi/c/1af48fffd7ff

-- 
Martin K. Petersen	Oracle Linux Engineering