From: Justin Tee <justintee8345@gmail.com>
To: linux-scsi@vger.kernel.org
Cc: jsmart833426@gmail.com, justin.tee@broadcom.com,
Justin Tee <justintee8345@gmail.com>
Subject: [PATCH 08/14] lpfc: Improve PLOGI retry handling for large SAN configurations
Date: Thu, 4 Jun 2026 12:29:31 -0700 [thread overview]
Message-ID: <20260604192937.65605-9-justintee8345@gmail.com> (raw)
In-Reply-To: <20260604192937.65605-1-justintee8345@gmail.com>
In large SAN configurations with link perturbations, rediscovery of target
ports is problematic due to PLOGI retry race conditions.
This patch improves target rediscovery by ensuring PLOGI retries are
serialized in unregistration and retry handler paths.
Signed-off-by: Justin Tee <justintee8345@gmail.com>
---
drivers/scsi/lpfc/lpfc_els.c | 62 +++++++++++++++++++--
drivers/scsi/lpfc/lpfc_nportdisc.c | 58 ++++++++++++++++++++
drivers/scsi/lpfc/lpfc_sli.c | 87 +++++++++++++++++-------------
3 files changed, 167 insertions(+), 40 deletions(-)
diff --git a/drivers/scsi/lpfc/lpfc_els.c b/drivers/scsi/lpfc/lpfc_els.c
index c67f8581f584..056d63a3d166 100644
--- a/drivers/scsi/lpfc/lpfc_els.c
+++ b/drivers/scsi/lpfc/lpfc_els.c
@@ -2159,6 +2159,8 @@ lpfc_cmpl_els_plogi(struct lpfc_hba *phba, struct lpfc_iocbq *cmdiocb,
goto out_freeiocb;
}
+ clear_bit(NLP_PLOGI_SND, &ndlp->nlp_flag);
+
/* Since ndlp can be freed in the disc state machine, note if this node
* is being used during discovery.
*/
@@ -2329,17 +2331,43 @@ lpfc_issue_els_plogi(struct lpfc_vport *vport, uint32_t did, uint8_t retry)
test_bit(NLP_UNREG_INP, &ndlp->nlp_flag)) &&
((ndlp->nlp_DID & Fabric_DID_MASK) != Fabric_DID_MASK) &&
!test_bit(FC_OFFLINE_MODE, &vport->fc_flag)) {
- lpfc_printf_vlog(vport, KERN_INFO, LOG_DISCOVERY,
+ lpfc_printf_vlog(vport, KERN_INFO,
+ LOG_ELS | LOG_NODE | LOG_DISCOVERY,
"4110 Issue PLOGI x%x deferred "
"on NPort x%x rpi x%x flg x%lx Data:"
" x%px\n",
ndlp->nlp_defer_did, ndlp->nlp_DID,
ndlp->nlp_rpi, ndlp->nlp_flag, ndlp);
- /* We can only defer 1st PLOGI */
- if (ndlp->nlp_defer_did == NLP_EVT_NOTHING_PENDING)
+ /* Don't defer a PLOGI that is already in that condition.
+ * Also set the nlp_last_elscmd to PLOGI to get the retry.
+ */
+ if (ndlp->nlp_defer_did == NLP_EVT_NOTHING_PENDING) {
ndlp->nlp_defer_did = did;
- return 0;
+ ndlp->nlp_last_elscmd = ELS_CMD_PLOGI;
+ }
+ return 1;
+ }
+
+ if (test_bit(NLP_PLOGI_SND, &ndlp->nlp_flag)) {
+ lpfc_printf_vlog(vport, KERN_INFO,
+ LOG_ELS | LOG_NODE | LOG_DISCOVERY,
+ "4113 Reject PLOGI issue, PLOGI in-flight "
+ "x%px, DID x%x nflag x%lx\n",
+ ndlp, ndlp->nlp_DID, ndlp->nlp_flag);
+ return 1;
+ }
+
+ if (ndlp->nlp_state > NLP_STE_PLOGI_ISSUE &&
+ ndlp->nlp_state <= NLP_STE_MAPPED_NODE) {
+ lpfc_printf_vlog(vport, KERN_INFO,
+ LOG_ELS | LOG_NODE | LOG_DISCOVERY,
+ "4114 Reject PLOGI issue, Node in "
+ "unexpected state x%px, DID x%x nflag x%lx "
+ "in State x%x\n",
+ ndlp, ndlp->nlp_DID,
+ ndlp->nlp_flag, ndlp->nlp_state);
+ return 1;
}
cmdsize = (sizeof(uint32_t) + sizeof(struct serv_parm));
@@ -2415,6 +2443,7 @@ lpfc_issue_els_plogi(struct lpfc_vport *vport, uint32_t did, uint8_t retry)
return 1;
}
+ set_bit(NLP_PLOGI_SND, &ndlp->nlp_flag);
return 0;
}
@@ -4614,6 +4643,31 @@ lpfc_els_retry_delay_handler(struct lpfc_nodelist *ndlp)
lpfc_issue_els_flogi(vport, ndlp, retry);
break;
case ELS_CMD_PLOGI:
+ /* The driver delayed a PLOGI via the nlp_delayfunc, but
+ * it's possible the PLOGI is already on a deferred retry.
+ * Catch this case and skip this delayed PLOGI. This prevents
+ * multiple PLOGIs in flight. The defer code flow cleans
+ * up.
+ */
+ if ((test_bit(NLP_IGNR_REG_CMPL, &ndlp->nlp_flag) ||
+ test_bit(NLP_UNREG_INP, &ndlp->nlp_flag)) &&
+ ndlp->nlp_defer_did != NLP_EVT_NOTHING_PENDING &&
+ ((ndlp->nlp_DID & Fabric_DID_MASK) != Fabric_DID_MASK) &&
+ !test_bit(FC_OFFLINE_MODE, &vport->fc_flag)) {
+ /* When UNREG_RPI completes we need to have the
+ * nlp_last_elscmd set.
+ */
+ ndlp->nlp_last_elscmd = ELS_CMD_PLOGI;
+ lpfc_printf_vlog(vport, KERN_INFO,
+ LOG_ELS | LOG_NODE | LOG_DISCOVERY,
+ "4112 Skip delayed PLOGI x%x deferred "
+ "on NPort x%x rpi x%x flg x%lx Data:"
+ " x%px\n",
+ ndlp->nlp_defer_did, ndlp->nlp_DID,
+ ndlp->nlp_rpi, ndlp->nlp_flag, ndlp);
+ break;
+ }
+
if (!lpfc_issue_els_plogi(vport, ndlp->nlp_DID, retry)) {
ndlp->nlp_prev_state = ndlp->nlp_state;
lpfc_nlp_set_state(vport, ndlp, NLP_STE_PLOGI_ISSUE);
diff --git a/drivers/scsi/lpfc/lpfc_nportdisc.c b/drivers/scsi/lpfc/lpfc_nportdisc.c
index f917a5bcfd02..0270ab7e602f 100644
--- a/drivers/scsi/lpfc/lpfc_nportdisc.c
+++ b/drivers/scsi/lpfc/lpfc_nportdisc.c
@@ -920,6 +920,64 @@ lpfc_rcv_logo(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp,
ndlp->nlp_DID, ndlp->nlp_state,
ndlp->nlp_type, vport->fc_flag);
+ /* The driver wants to schedule a delayed PLOGI to recover
+ * the remote Nport. However, there are two cases that
+ * stop this so that multiple PLOGI are not inflight to
+ * the same NPortID
+ *
+ * Do not schedule a delayed PLOGI if the deferred PLOGI
+ * code already set up a PLOGI retry after an UNREG_RPI
+ * mailbox completes.
+ */
+ if (test_bit(NLP_UNREG_INP, &ndlp->nlp_flag) &&
+ ndlp->nlp_defer_did == ndlp->nlp_DID &&
+ ndlp->nlp_last_elscmd == ELS_CMD_PLOGI) {
+ lpfc_printf_vlog(vport, KERN_INFO,
+ LOG_NODE | LOG_ELS | LOG_DISCOVERY,
+ "3206 No PLOGI delay, defer PLOGI "
+ "waiting on DID x%06x UNREG_RPI "
+ "nflag x%lx state x%x lastels x%x "
+ "defer_did x%x\n",
+ ndlp->nlp_DID, ndlp->nlp_flag,
+ ndlp->nlp_state, ndlp->nlp_last_elscmd,
+ ndlp->nlp_defer_did);
+ goto out;
+ }
+
+ /* A delayed PLOGI retry is not required if the ndlp's delay
+ * timer is running and the last command was PLOGI.
+ */
+ if (test_bit(NLP_DELAY_TMO, &ndlp->nlp_flag) &&
+ ndlp->nlp_last_elscmd == ELS_CMD_PLOGI) {
+ lpfc_printf_vlog(vport, KERN_INFO,
+ LOG_NODE | LOG_ELS | LOG_DISCOVERY,
+ "3207 No PLOGI delay, PLOGI_DELAY_TMO "
+ "active on DID x%06x "
+ "nflag x%lx state x%x lastels x%x "
+ "defer_did x%x\n",
+ ndlp->nlp_DID, ndlp->nlp_flag,
+ ndlp->nlp_state, ndlp->nlp_last_elscmd,
+ ndlp->nlp_defer_did);
+ goto out;
+ }
+
+ /* Do not schedule a PLOGI retry if the ndlp state is NPR
+ * and vport has received an RSCN
+ */
+ if (ndlp->nlp_state == NLP_STE_NPR_NODE &&
+ test_bit(FC_RSCN_MODE, &vport->fc_flag)) {
+ lpfc_printf_vlog(vport, KERN_INFO,
+ LOG_NODE | LOG_ELS | LOG_DISCOVERY,
+ "3939 No PLOGI delay, RSCN in "
+ "progress for NPR DID x%06x "
+ "nflag x%lx state x%x last_els x%x "
+ "defer_did x%06x\n",
+ ndlp->nlp_DID, ndlp->nlp_flag,
+ ndlp->nlp_state, ndlp->nlp_last_elscmd,
+ ndlp->nlp_defer_did);
+ goto out;
+ }
+
/* Special cases for rports that recover post LOGO. */
if ((!(ndlp->nlp_type == NLP_FABRIC) &&
(ndlp->nlp_type & (NLP_FCP_TARGET | NLP_NVME_TARGET) ||
diff --git a/drivers/scsi/lpfc/lpfc_sli.c b/drivers/scsi/lpfc/lpfc_sli.c
index 20ee8171e31f..efde944a2693 100644
--- a/drivers/scsi/lpfc/lpfc_sli.c
+++ b/drivers/scsi/lpfc/lpfc_sli.c
@@ -2914,7 +2914,19 @@ lpfc_sli_def_mbox_cmpl(struct lpfc_hba *phba, LPFC_MBOXQ_t *pmb)
ndlp->nlp_defer_did != NLP_EVT_NOTHING_PENDING) {
clear_bit(NLP_UNREG_INP, &ndlp->nlp_flag);
ndlp->nlp_defer_did = NLP_EVT_NOTHING_PENDING;
- lpfc_issue_els_plogi(vport, ndlp->nlp_DID, 0);
+
+ if (!test_bit(NLP_DELAY_TMO, &ndlp->nlp_flag) &&
+ ndlp->nlp_last_elscmd == ELS_CMD_PLOGI) {
+ rc = lpfc_issue_els_plogi(vport,
+ ndlp->nlp_DID,
+ 0);
+ if (!rc) {
+ ndlp->nlp_prev_state =
+ ndlp->nlp_state;
+ lpfc_nlp_set_state(vport, ndlp,
+ NLP_STE_PLOGI_ISSUE);
+ }
+ }
} else {
clear_bit(NLP_UNREG_INP, &ndlp->nlp_flag);
}
@@ -2967,52 +2979,55 @@ lpfc_sli4_unreg_rpi_cmpl_clr(struct lpfc_hba *phba, LPFC_MBOXQ_t *pmb)
bool unreg_inp;
ndlp = pmb->ctx_ndlp;
- if (pmb->u.mb.mbxCommand == MBX_UNREG_LOGIN) {
+ if (pmb->u.mb.mbxCommand == MBX_UNREG_LOGIN && ndlp) {
if (phba->sli_rev == LPFC_SLI_REV4 &&
(bf_get(lpfc_sli_intf_if_type,
&phba->sli4_hba.sli_intf) >=
LPFC_SLI_INTF_IF_TYPE_2)) {
- if (ndlp) {
- lpfc_printf_vlog(
- vport, KERN_INFO,
+ lpfc_printf_vlog(vport, KERN_INFO,
LOG_MBOX | LOG_SLI | LOG_NODE,
- "0010 UNREG_LOGIN vpi:x%x "
- "rpi:%x DID:%x defer x%x flg x%lx "
- "x%px\n",
+ "0010 UNREG_LOGIN vpi:x%x rpi:%x "
+ "DID:%x defer x%x flg x%lx x%px\n",
vport->vpi, ndlp->nlp_rpi,
ndlp->nlp_DID, ndlp->nlp_defer_did,
- ndlp->nlp_flag,
- ndlp);
+ ndlp->nlp_flag, ndlp);
- /* Cleanup the nlp_flag now that the UNREG RPI
- * has completed.
- */
- unreg_inp = test_and_clear_bit(NLP_UNREG_INP,
- &ndlp->nlp_flag);
- clear_bit(NLP_LOGO_ACC, &ndlp->nlp_flag);
+ /* Cleanup the nlp_flag now that the UNREG RPI
+ * has completed.
+ */
+ unreg_inp = test_and_clear_bit(NLP_UNREG_INP,
+ &ndlp->nlp_flag);
+ clear_bit(NLP_LOGO_ACC, &ndlp->nlp_flag);
- /* Check to see if there are any deferred
- * events to process
- */
- if (unreg_inp &&
- ndlp->nlp_defer_did !=
- NLP_EVT_NOTHING_PENDING) {
- lpfc_printf_vlog(
- vport, KERN_INFO,
- LOG_MBOX | LOG_SLI | LOG_NODE,
- "4111 UNREG cmpl deferred "
- "clr x%x on "
- "NPort x%x Data: x%x x%px\n",
- ndlp->nlp_rpi, ndlp->nlp_DID,
- ndlp->nlp_defer_did, ndlp);
- ndlp->nlp_defer_did =
- NLP_EVT_NOTHING_PENDING;
- lpfc_issue_els_plogi(
- vport, ndlp->nlp_DID, 0);
- }
+ /* Check to see if there are any deferred
+ * events to process
+ */
+ if (unreg_inp &&
+ ndlp->nlp_defer_did != NLP_EVT_NOTHING_PENDING) {
+ lpfc_printf_vlog(vport, KERN_INFO,
+ LOG_MBOX | LOG_SLI | LOG_NODE,
+ "4111 UNREG cmpl deferred "
+ "clr x%x on NPort x%x "
+ "Data: x%x x%x x%px\n",
+ ndlp->nlp_rpi, ndlp->nlp_DID,
+ ndlp->nlp_defer_did,
+ ndlp->nlp_last_elscmd, ndlp);
+ ndlp->nlp_defer_did = NLP_EVT_NOTHING_PENDING;
- lpfc_nlp_put(ndlp);
+ if (!test_bit(NLP_DELAY_TMO, &ndlp->nlp_flag) &&
+ ndlp->nlp_last_elscmd == ELS_CMD_PLOGI) {
+ if (lpfc_issue_els_plogi(vport,
+ ndlp->nlp_DID,
+ 0))
+ goto out;
+
+ ndlp->nlp_prev_state = ndlp->nlp_state;
+ lpfc_nlp_set_state(vport, ndlp,
+ NLP_STE_PLOGI_ISSUE);
+ }
}
+out:
+ lpfc_nlp_put(ndlp);
}
}
--
2.38.0
next prev parent reply other threads:[~2026-06-04 18:50 UTC|newest]
Thread overview: 15+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-06-04 19:29 [PATCH 00/14] Update lpfc to revision 15.0.0.1 Justin Tee
2026-06-04 19:29 ` [PATCH 01/14] lpfc: Fix use-after-free in lpfc_cmpl_ct_cmd_vmid Justin Tee
2026-06-04 19:29 ` [PATCH 02/14] lpfc: Early return out of lpfc_els_abort when HBA_SETUP flag is not set Justin Tee
2026-06-04 19:29 ` [PATCH 03/14] lpfc: Fix kernel oops when unmapping scsi dma buffers for an aborted cmd Justin Tee
2026-06-04 19:29 ` [PATCH 04/14] lpfc: Check fc4_xpt_flags before decrementing ndlp kref on FDISC error Justin Tee
2026-06-04 19:29 ` [PATCH 05/14] lpfc: Add handling for when PLOGI or PRLI is dropped during link failure Justin Tee
2026-06-04 19:29 ` [PATCH 06/14] lpfc: Fix ndlp use-after-free during repeated RSCN and rediscovery sequence Justin Tee
2026-06-04 19:29 ` [PATCH 07/14] lpfc: Rework I/O flush ordering when unloading driver Justin Tee
2026-06-04 19:29 ` Justin Tee [this message]
2026-06-04 19:29 ` [PATCH 09/14] lpfc: Send inhibited ABORT_WQE when PLOGI CQE SEQUENCE_TMO is received Justin Tee
2026-06-04 19:29 ` [PATCH 10/14] lpfc: Remove slowpath cqe process limiter in slow ring event handler Justin Tee
2026-06-04 19:29 ` [PATCH 11/14] lpfc: Put iocbq on phba->txq when ELS WQ is full or ELS SGL unavailable Justin Tee
2026-06-04 19:29 ` [PATCH 12/14] lpfc: Update ELS ACC logging for diagnostic troubleshooting Justin Tee
2026-06-04 19:29 ` [PATCH 13/14] lpfc: Refactor calls on fc_disctmo to lpfc_set_disctmo in RSCN handler Justin Tee
2026-06-04 19:29 ` [PATCH 14/14] lpfc: Update lpfc version to 15.0.0.1 Justin Tee
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20260604192937.65605-9-justintee8345@gmail.com \
--to=justintee8345@gmail.com \
--cc=jsmart833426@gmail.com \
--cc=justin.tee@broadcom.com \
--cc=linux-scsi@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox