[PATCH AUTOSEL 6.12 096/486] scsi: st: Tighten the page format heuristics with MODE SELECT

[PATCH AUTOSEL 6.12 096/486] scsi: st: Tighten the page format heuristics with MODE SELECT

[Sasha Levin]

From: Kai Mäkisara <Kai.Makisara@kolumbus.fi>

[ Upstream commit 8db816c6f176321e42254badd5c1a8df8bfcfdb4 ]

In the days when SCSI-2 was emerging, some drives did claim SCSI-2 but did
not correctly implement it. The st driver first tries MODE SELECT with the
page format bit set to set the block descriptor.  If not successful, the
non-page format is tried.

The test only tests the sense code and this triggers also from illegal
parameter in the parameter list. The test is limited to "old" devices and
made more strict to remove false alarms.

Signed-off-by: Kai Mäkisara <Kai.Makisara@kolumbus.fi>
Link: https://lore.kernel.org/r/20250311112516.5548-4-Kai.Makisara@kolumbus.fi
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 drivers/scsi/st.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/drivers/scsi/st.c b/drivers/scsi/st.c
index a17441635ff3a..1c0951168f646 100644
--- a/drivers/scsi/st.c
+++ b/drivers/scsi/st.c
@@ -3081,7 +3081,9 @@ static int st_int_ioctl(struct scsi_tape *STp, unsigned int cmd_in, unsigned lon
 			   cmd_in == MTSETDRVBUFFER ||
 			   cmd_in == SET_DENS_AND_BLK) {
 			if (cmdstatp->sense_hdr.sense_key == ILLEGAL_REQUEST &&
-			    !(STp->use_pf & PF_TESTED)) {
+				cmdstatp->sense_hdr.asc == 0x24 &&
+				(STp->device)->scsi_level <= SCSI_2 &&
+				!(STp->use_pf & PF_TESTED)) {
 				/* Try the other possible state of Page Format if not
 				   already tried */
 				STp->use_pf = (STp->use_pf ^ USE_PF) | PF_TESTED;
-- 
2.39.5

[PATCH AUTOSEL 6.12 097/486] scsi: st: ERASE does not change tape location

[Sasha Levin]

From: Kai Mäkisara <Kai.Makisara@kolumbus.fi>

[ Upstream commit ad77cebf97bd42c93ab4e3bffd09f2b905c1959a ]

The SCSI ERASE command erases from the current position onwards.  Don't
clear the position variables.

Signed-off-by: Kai Mäkisara <Kai.Makisara@kolumbus.fi>
Link: https://lore.kernel.org/r/20250311112516.5548-3-Kai.Makisara@kolumbus.fi
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 drivers/scsi/st.c | 1 -
 1 file changed, 1 deletion(-)

diff --git a/drivers/scsi/st.c b/drivers/scsi/st.c
index 1c0951168f646..1cfd7e71dcdde 100644
--- a/drivers/scsi/st.c
+++ b/drivers/scsi/st.c
@@ -2894,7 +2894,6 @@ static int st_int_ioctl(struct scsi_tape *STp, unsigned int cmd_in, unsigned lon
 			timeout = STp->long_timeout * 8;
 
 		DEBC_printk(STp, "Erasing tape.\n");
-		fileno = blkno = at_sm = 0;
 		break;
 	case MTSETBLK:		/* Set block length */
 	case MTSETDENSITY:	/* Set tape density */
-- 
2.39.5

[PATCH AUTOSEL 6.12 215/486] scsi: logging: Fix scsi_logging_level bounds

[Sasha Levin]

From: Nicolas Bouchinet <nicolas.bouchinet@ssi.gouv.fr>

[ Upstream commit 2cef5b4472c602e6c5a119aca869d9d4050586f3 ]

Bound scsi_logging_level sysctl writings between SYSCTL_ZERO and
SYSCTL_INT_MAX.

The proc_handler has thus been updated to proc_dointvec_minmax.

Signed-off-by: Nicolas Bouchinet <nicolas.bouchinet@ssi.gouv.fr>
Link: https://lore.kernel.org/r/20250224095826.16458-5-nicolas.bouchinet@clip-os.org
Reviewed-by: Joel Granados <joel.granados@kernel.org>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 drivers/scsi/scsi_sysctl.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/drivers/scsi/scsi_sysctl.c b/drivers/scsi/scsi_sysctl.c
index 093774d775346..daa160459c9b3 100644
--- a/drivers/scsi/scsi_sysctl.c
+++ b/drivers/scsi/scsi_sysctl.c
@@ -17,7 +17,9 @@ static struct ctl_table scsi_table[] = {
 	  .data		= &scsi_logging_level,
 	  .maxlen	= sizeof(scsi_logging_level),
 	  .mode		= 0644,
-	  .proc_handler	= proc_dointvec },
+	  .proc_handler	= proc_dointvec_minmax,
+	  .extra1	= SYSCTL_ZERO,
+	  .extra2	= SYSCTL_INT_MAX },
 };
 
 static struct ctl_table_header *scsi_table_header;
-- 
2.39.5

[PATCH AUTOSEL 6.12 264/486] scsi: mpi3mr: Update timestamp only for supervisor IOCs

[Sasha Levin]

From: Ranjan Kumar <ranjan.kumar@broadcom.com>

[ Upstream commit 83a9d30d29f275571f6e8f879f04b2379be7eb6c ]

The driver issues the time stamp update command periodically. Even if the
command fails with supervisor only IOC Status.

Instead check the Non-Supervisor capability bit reported by IOC as part of
IOC Facts.

Co-developed-by: Sumit Saxena <sumit.saxena@broadcom.com>
Signed-off-by: Sumit Saxena <sumit.saxena@broadcom.com>
Signed-off-by: Ranjan Kumar <ranjan.kumar@broadcom.com>
Link: https://lore.kernel.org/r/20250220142528.20837-3-ranjan.kumar@broadcom.com
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 drivers/scsi/mpi3mr/mpi3mr_fw.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/drivers/scsi/mpi3mr/mpi3mr_fw.c b/drivers/scsi/mpi3mr/mpi3mr_fw.c
index c0a372868e1d7..dee3ea8d4837e 100644
--- a/drivers/scsi/mpi3mr/mpi3mr_fw.c
+++ b/drivers/scsi/mpi3mr/mpi3mr_fw.c
@@ -2744,7 +2744,10 @@ static void mpi3mr_watchdog_work(struct work_struct *work)
 		return;
 	}
 
-	if (mrioc->ts_update_counter++ >= mrioc->ts_update_interval) {
+	if (!(mrioc->facts.ioc_capabilities &
+		MPI3_IOCFACTS_CAPABILITY_NON_SUPERVISOR_IOC) &&
+		(mrioc->ts_update_counter++ >= mrioc->ts_update_interval)) {
+
 		mrioc->ts_update_counter = 0;
 		mpi3mr_sync_timestamp(mrioc);
 	}
-- 
2.39.5

[PATCH AUTOSEL 6.12 269/486] scsi: scsi_debug: First fixes for tapes

[Sasha Levin]

From: Kai Mäkisara <Kai.Makisara@kolumbus.fi>

[ Upstream commit f69da85d5d5cc5b7dfb963a6c6c1ac0dd9002341 ]

Patch includes the following:

 - Enable MODE SENSE/SELECT without actual page (to read/write only the
   Block Descriptor)

 - Store the density code and block size in the Block Descriptor (only
   short version for tapes)

 - Fix REWIND not to use the wrong page filling function

Signed-off-by: Kai Mäkisara <Kai.Makisara@kolumbus.fi>
Link: https://lore.kernel.org/r/20250213092636.2510-2-Kai.Makisara@kolumbus.fi
Reviewed-by: John Meneghini <jmeneghi@redhat.com>
Tested-by: John Meneghini <jmeneghi@redhat.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 drivers/scsi/scsi_debug.c | 55 ++++++++++++++++++++++++++++++++++-----
 1 file changed, 49 insertions(+), 6 deletions(-)

diff --git a/drivers/scsi/scsi_debug.c b/drivers/scsi/scsi_debug.c
index 680ba180a6725..89a2aaccdcfce 100644
--- a/drivers/scsi/scsi_debug.c
+++ b/drivers/scsi/scsi_debug.c
@@ -173,6 +173,10 @@ static const char *sdebug_version_date = "20210520";
 #define DEF_ZBC_MAX_OPEN_ZONES	8
 #define DEF_ZBC_NR_CONV_ZONES	1
 
+/* Default parameters for tape drives */
+#define TAPE_DEF_DENSITY  0x0
+#define TAPE_DEF_BLKSIZE  0
+
 #define SDEBUG_LUN_0_VAL 0
 
 /* bit mask values for sdebug_opts */
@@ -363,6 +367,10 @@ struct sdebug_dev_info {
 	ktime_t create_ts;	/* time since bootup that this device was created */
 	struct sdeb_zone_state *zstate;
 
+	/* For tapes */
+	unsigned int tape_blksize;
+	unsigned int tape_density;
+
 	struct dentry *debugfs_entry;
 	struct spinlock list_lock;
 	struct list_head inject_err_list;
@@ -773,7 +781,7 @@ static const struct opcode_info_t opcode_info_arr[SDEB_I_LAST_ELEM_P1 + 1] = {
 /* 20 */
 	{0, 0x1e, 0, 0, NULL, NULL, /* ALLOW REMOVAL */
 	    {6,  0, 0, 0, 0x3, 0xc7, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0} },
-	{0, 0x1, 0, 0, resp_start_stop, NULL, /* REWIND ?? */
+	{0, 0x1, 0, 0, NULL, NULL, /* REWIND ?? */
 	    {6,  0x1, 0, 0, 0, 0xc7, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0} },
 	{0, 0, 0, F_INV_OP | FF_RESPOND, NULL, NULL, /* ATA_PT */
 	    {0,  0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0} },
@@ -2742,7 +2750,7 @@ static int resp_mode_sense(struct scsi_cmnd *scp,
 	unsigned char *ap;
 	unsigned char *arr __free(kfree);
 	unsigned char *cmd = scp->cmnd;
-	bool dbd, llbaa, msense_6, is_disk, is_zbc;
+	bool dbd, llbaa, msense_6, is_disk, is_zbc, is_tape;
 
 	arr = kzalloc(SDEBUG_MAX_MSENSE_SZ, GFP_ATOMIC);
 	if (!arr)
@@ -2755,7 +2763,8 @@ static int resp_mode_sense(struct scsi_cmnd *scp,
 	llbaa = msense_6 ? false : !!(cmd[1] & 0x10);
 	is_disk = (sdebug_ptype == TYPE_DISK);
 	is_zbc = devip->zoned;
-	if ((is_disk || is_zbc) && !dbd)
+	is_tape = (sdebug_ptype == TYPE_TAPE);
+	if ((is_disk || is_zbc || is_tape) && !dbd)
 		bd_len = llbaa ? 16 : 8;
 	else
 		bd_len = 0;
@@ -2793,15 +2802,25 @@ static int resp_mode_sense(struct scsi_cmnd *scp,
 			put_unaligned_be32(0xffffffff, ap + 0);
 		else
 			put_unaligned_be32(sdebug_capacity, ap + 0);
-		put_unaligned_be16(sdebug_sector_size, ap + 6);
+		if (is_tape) {
+			ap[0] = devip->tape_density;
+			put_unaligned_be16(devip->tape_blksize, ap + 6);
+		} else
+			put_unaligned_be16(sdebug_sector_size, ap + 6);
 		offset += bd_len;
 		ap = arr + offset;
 	} else if (16 == bd_len) {
+		if (is_tape) {
+			mk_sense_invalid_fld(scp, SDEB_IN_DATA, 1, 4);
+			return check_condition_result;
+		}
 		put_unaligned_be64((u64)sdebug_capacity, ap + 0);
 		put_unaligned_be32(sdebug_sector_size, ap + 12);
 		offset += bd_len;
 		ap = arr + offset;
 	}
+	if (cmd[2] == 0)
+		goto only_bd; /* Only block descriptor requested */
 
 	/*
 	 * N.B. If len>0 before resp_*_pg() call, then form of that call should be:
@@ -2902,6 +2921,7 @@ static int resp_mode_sense(struct scsi_cmnd *scp,
 	default:
 		goto bad_pcode;
 	}
+only_bd:
 	if (msense_6)
 		arr[0] = offset - 1;
 	else
@@ -2945,8 +2965,27 @@ static int resp_mode_select(struct scsi_cmnd *scp,
 			    __func__, param_len, res);
 	md_len = mselect6 ? (arr[0] + 1) : (get_unaligned_be16(arr + 0) + 2);
 	bd_len = mselect6 ? arr[3] : get_unaligned_be16(arr + 6);
-	off = bd_len + (mselect6 ? 4 : 8);
-	if (md_len > 2 || off >= res) {
+	off = (mselect6 ? 4 : 8);
+	if (sdebug_ptype == TYPE_TAPE) {
+		int blksize;
+
+		if (bd_len != 8) {
+			mk_sense_invalid_fld(scp, SDEB_IN_DATA,
+					mselect6 ? 3 : 6, -1);
+			return check_condition_result;
+		}
+		blksize = get_unaligned_be16(arr + off + 6);
+		if ((blksize % 4) != 0) {
+			mk_sense_invalid_fld(scp, SDEB_IN_DATA, off + 6, -1);
+			return check_condition_result;
+		}
+		devip->tape_density = arr[off];
+		devip->tape_blksize = blksize;
+	}
+	off += bd_len;
+	if (off >= res)
+		return 0; /* No page written, just descriptors */
+	if (md_len > 2) {
 		mk_sense_invalid_fld(scp, SDEB_IN_DATA, 0, -1);
 		return check_condition_result;
 	}
@@ -5835,6 +5874,10 @@ static struct sdebug_dev_info *sdebug_device_create(
 		} else {
 			devip->zoned = false;
 		}
+		if (sdebug_ptype == TYPE_TAPE) {
+			devip->tape_density = TAPE_DEF_DENSITY;
+			devip->tape_blksize = TAPE_DEF_BLKSIZE;
+		}
 		devip->create_ts = ktime_get_boottime();
 		atomic_set(&devip->stopped, (sdeb_tur_ms_to_ready > 0 ? 2 : 0));
 		spin_lock_init(&devip->list_lock);
-- 
2.39.5

[PATCH AUTOSEL 6.12 283/486] scsi: mpt3sas: Send a diag reset if target reset fails

[Sasha Levin]

From: Shivasharan S <shivasharan.srikanteshwara@broadcom.com>

[ Upstream commit 5612d6d51ed2634a033c95de2edec7449409cbb9 ]

When an IOCTL times out and driver issues a target reset, if firmware
fails the task management elevate the recovery by issuing a diag reset to
controller.

Signed-off-by: Shivasharan S <shivasharan.srikanteshwara@broadcom.com>
Link: https://lore.kernel.org/r/1739410016-27503-5-git-send-email-shivasharan.srikanteshwara@broadcom.com
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 drivers/scsi/mpt3sas/mpt3sas_ctl.c | 12 ++++++++++--
 1 file changed, 10 insertions(+), 2 deletions(-)

diff --git a/drivers/scsi/mpt3sas/mpt3sas_ctl.c b/drivers/scsi/mpt3sas/mpt3sas_ctl.c
index 87784c96249a7..47faa27bc3559 100644
--- a/drivers/scsi/mpt3sas/mpt3sas_ctl.c
+++ b/drivers/scsi/mpt3sas/mpt3sas_ctl.c
@@ -679,6 +679,7 @@ _ctl_do_mpt_command(struct MPT3SAS_ADAPTER *ioc, struct mpt3_ioctl_command karg,
 	size_t data_in_sz = 0;
 	long ret;
 	u16 device_handle = MPT3SAS_INVALID_DEVICE_HANDLE;
+	int tm_ret;
 
 	issue_reset = 0;
 
@@ -1120,18 +1121,25 @@ _ctl_do_mpt_command(struct MPT3SAS_ADAPTER *ioc, struct mpt3_ioctl_command karg,
 			if (pcie_device && (!ioc->tm_custom_handling) &&
 			    (!(mpt3sas_scsih_is_pcie_scsi_device(
 			    pcie_device->device_info))))
-				mpt3sas_scsih_issue_locked_tm(ioc,
+				tm_ret = mpt3sas_scsih_issue_locked_tm(ioc,
 				  le16_to_cpu(mpi_request->FunctionDependent1),
 				  0, 0, 0,
 				  MPI2_SCSITASKMGMT_TASKTYPE_TARGET_RESET, 0,
 				  0, pcie_device->reset_timeout,
 			MPI26_SCSITASKMGMT_MSGFLAGS_PROTOCOL_LVL_RST_PCIE);
 			else
-				mpt3sas_scsih_issue_locked_tm(ioc,
+				tm_ret = mpt3sas_scsih_issue_locked_tm(ioc,
 				  le16_to_cpu(mpi_request->FunctionDependent1),
 				  0, 0, 0,
 				  MPI2_SCSITASKMGMT_TASKTYPE_TARGET_RESET, 0,
 				  0, 30, MPI2_SCSITASKMGMT_MSGFLAGS_LINK_RESET);
+
+			if (tm_ret != SUCCESS) {
+				ioc_info(ioc,
+					 "target reset failed, issue hard reset: handle (0x%04x)\n",
+					 le16_to_cpu(mpi_request->FunctionDependent1));
+				mpt3sas_base_hard_reset_handler(ioc, FORCE_BIG_HAMMER);
+			}
 		} else
 			mpt3sas_base_hard_reset_handler(ioc, FORCE_BIG_HAMMER);
 	}
-- 
2.39.5

[PATCH AUTOSEL 6.12 378/486] scsi: target: spc: Fix loop traversal in spc_rsoc_get_descr()

[Sasha Levin]

From: Chaohai Chen <wdhh66@163.com>

[ Upstream commit 04ad06e41d1c74cc323b20a7bd023c47bd0e0c38 ]

Stop traversing after finding the appropriate descriptor.

Signed-off-by: Chaohai Chen <wdhh66@163.com>
Link: https://lore.kernel.org/r/20250124085542.109088-1-wdhh66@163.com
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 drivers/target/target_core_spc.c | 14 ++++++++++----
 1 file changed, 10 insertions(+), 4 deletions(-)

diff --git a/drivers/target/target_core_spc.c b/drivers/target/target_core_spc.c
index 61c065702350e..701dcbd7b63cf 100644
--- a/drivers/target/target_core_spc.c
+++ b/drivers/target/target_core_spc.c
@@ -2151,8 +2151,10 @@ spc_rsoc_get_descr(struct se_cmd *cmd, struct target_opcode_descriptor **opcode)
 			if (descr->serv_action_valid)
 				return TCM_INVALID_CDB_FIELD;
 
-			if (!descr->enabled || descr->enabled(descr, cmd))
+			if (!descr->enabled || descr->enabled(descr, cmd)) {
 				*opcode = descr;
+				return TCM_NO_SENSE;
+			}
 			break;
 		case 0x2:
 			/*
@@ -2166,8 +2168,10 @@ spc_rsoc_get_descr(struct se_cmd *cmd, struct target_opcode_descriptor **opcode)
 			if (descr->serv_action_valid &&
 			    descr->service_action == requested_sa) {
 				if (!descr->enabled || descr->enabled(descr,
-								      cmd))
+								      cmd)) {
 					*opcode = descr;
+					return TCM_NO_SENSE;
+				}
 			} else if (!descr->serv_action_valid)
 				return TCM_INVALID_CDB_FIELD;
 			break;
@@ -2180,13 +2184,15 @@ spc_rsoc_get_descr(struct se_cmd *cmd, struct target_opcode_descriptor **opcode)
 			 */
 			if (descr->service_action == requested_sa)
 				if (!descr->enabled || descr->enabled(descr,
-								      cmd))
+								      cmd)) {
 					*opcode = descr;
+					return TCM_NO_SENSE;
+				}
 			break;
 		}
 	}
 
-	return 0;
+	return TCM_NO_SENSE;
 }
 
 static sense_reason_t
-- 
2.39.5

[PATCH AUTOSEL 6.12 441/486] scsi: lpfc: Handle duplicate D_IDs in ndlp search-by D_ID routine

[Sasha Levin]

From: Justin Tee <justin.tee@broadcom.com>

[ Upstream commit 56c3d809b7b450379162d0b8a70bbe71ab8db706 ]

After a port swap between separate fabrics, there may be multiple nodes in
the vport's fc_nodes list with the same fabric well known address.
Duplication is temporary and eventually resolves itself after dev_loss_tmo
expires, but nameserver queries may still occur before dev_loss_tmo.  This
possibly results in returning stale fabric ndlp objects.  Fix by adding an
nlp_state check to ensure the ndlp search routine returns the correct newer
allocated ndlp fabric object.

Signed-off-by: Justin Tee <justin.tee@broadcom.com>
Link: https://lore.kernel.org/r/20250131000524.163662-5-justintee8345@gmail.com
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 drivers/scsi/lpfc/lpfc_hbadisc.c | 17 ++++++++++++-----
 1 file changed, 12 insertions(+), 5 deletions(-)

diff --git a/drivers/scsi/lpfc/lpfc_hbadisc.c b/drivers/scsi/lpfc/lpfc_hbadisc.c
index 6e8d8a96c54fb..31dcabebc9b6d 100644
--- a/drivers/scsi/lpfc/lpfc_hbadisc.c
+++ b/drivers/scsi/lpfc/lpfc_hbadisc.c
@@ -5622,6 +5622,7 @@ static struct lpfc_nodelist *
 __lpfc_findnode_did(struct lpfc_vport *vport, uint32_t did)
 {
 	struct lpfc_nodelist *ndlp;
+	struct lpfc_nodelist *np = NULL;
 	uint32_t data1;
 
 	list_for_each_entry(ndlp, &vport->fc_nodes, nlp_listp) {
@@ -5636,14 +5637,20 @@ __lpfc_findnode_did(struct lpfc_vport *vport, uint32_t did)
 					 ndlp, ndlp->nlp_DID,
 					 ndlp->nlp_flag, data1, ndlp->nlp_rpi,
 					 ndlp->active_rrqs_xri_bitmap);
-			return ndlp;
+
+			/* Check for new or potentially stale node */
+			if (ndlp->nlp_state != NLP_STE_UNUSED_NODE)
+				return ndlp;
+			np = ndlp;
 		}
 	}
 
-	/* FIND node did <did> NOT FOUND */
-	lpfc_printf_vlog(vport, KERN_INFO, LOG_NODE,
-			 "0932 FIND node did x%x NOT FOUND.\n", did);
-	return NULL;
+	if (!np)
+		/* FIND node did <did> NOT FOUND */
+		lpfc_printf_vlog(vport, KERN_INFO, LOG_NODE,
+				 "0932 FIND node did x%x NOT FOUND.\n", did);
+
+	return np;
 }
 
 struct lpfc_nodelist *
-- 
2.39.5

[PATCH AUTOSEL 6.12 442/486] scsi: lpfc: Ignore ndlp rport mismatch in dev_loss_tmo callbk

[Sasha Levin]

From: Justin Tee <justin.tee@broadcom.com>

[ Upstream commit 23ed62897746f49f195d819ce6edeb1db27d1b72 ]

With repeated port swaps between separate fabrics, there can be multiple
registrations for fabric well known address 0xfffffe.  This can cause ndlp
reference confusion due to the usage of a single ndlp ptr that stores the
rport object in fc_rport struct private storage during transport
registration.  Subsequent registrations update the ndlp->rport field with
the newer rport, so when transport layer triggers dev_loss_tmo for the
earlier registered rport the ndlp->rport private storage is referencing the
newer rport instead of the older rport in dev_loss_tmo callbk.

Because the older ndlp->rport object is already cleaned up elsewhere in
driver code during the time of fabric swap, check that the rport provided
in dev_loss_tmo callbk actually matches the rport stored in the LLDD's
ndlp->rport field.  Otherwise, skip dev_loss_tmo work on a stale rport.

Signed-off-by: Justin Tee <justin.tee@broadcom.com>
Link: https://lore.kernel.org/r/20250131000524.163662-4-justintee8345@gmail.com
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 drivers/scsi/lpfc/lpfc_hbadisc.c | 12 +++++++++---
 1 file changed, 9 insertions(+), 3 deletions(-)

diff --git a/drivers/scsi/lpfc/lpfc_hbadisc.c b/drivers/scsi/lpfc/lpfc_hbadisc.c
index 31dcabebc9b6d..f2e4237ff3d99 100644
--- a/drivers/scsi/lpfc/lpfc_hbadisc.c
+++ b/drivers/scsi/lpfc/lpfc_hbadisc.c
@@ -227,10 +227,16 @@ lpfc_dev_loss_tmo_callbk(struct fc_rport *rport)
 	if (ndlp->nlp_state == NLP_STE_MAPPED_NODE)
 		return;
 
-	/* check for recovered fabric node */
-	if (ndlp->nlp_state == NLP_STE_UNMAPPED_NODE &&
-	    ndlp->nlp_DID == Fabric_DID)
+	/* Ignore callback for a mismatched (stale) rport */
+	if (ndlp->rport != rport) {
+		lpfc_vlog_msg(vport, KERN_WARNING, LOG_NODE,
+			      "6788 fc rport mismatch: d_id x%06x ndlp x%px "
+			      "fc rport x%px node rport x%px state x%x "
+			      "refcnt %u\n",
+			      ndlp->nlp_DID, ndlp, rport, ndlp->rport,
+			      ndlp->nlp_state, kref_read(&ndlp->kref));
 		return;
+	}
 
 	if (rport->port_name != wwn_to_u64(ndlp->nlp_portname.u.wwn))
 		lpfc_printf_vlog(vport, KERN_ERR, LOG_TRACE_EVENT,
-- 
2.39.5

[PATCH AUTOSEL 6.12 443/486] scsi: lpfc: Free phba irq in lpfc_sli4_enable_msi() when pci_irq_vector() fails

[Sasha Levin]

From: Justin Tee <justin.tee@broadcom.com>

[ Upstream commit f0842902b383982d1f72c490996aa8fc29a7aa0d ]

Fix smatch warning regarding missed calls to free_irq().  Free the phba IRQ
in the failed pci_irq_vector cases.

lpfc_init.c: lpfc_sli4_enable_msi() warn: 'phba->pcidev->irq' from
             request_irq() not released.

Signed-off-by: Justin Tee <justin.tee@broadcom.com>
Link: https://lore.kernel.org/r/20250131000524.163662-3-justintee8345@gmail.com
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 drivers/scsi/lpfc/lpfc_init.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/drivers/scsi/lpfc/lpfc_init.c b/drivers/scsi/lpfc/lpfc_init.c
index a3658ef1141b2..50c761991191f 100644
--- a/drivers/scsi/lpfc/lpfc_init.c
+++ b/drivers/scsi/lpfc/lpfc_init.c
@@ -13190,6 +13190,7 @@ lpfc_sli4_enable_msi(struct lpfc_hba *phba)
 	eqhdl = lpfc_get_eq_hdl(0);
 	rc = pci_irq_vector(phba->pcidev, 0);
 	if (rc < 0) {
+		free_irq(phba->pcidev->irq, phba);
 		pci_free_irq_vectors(phba->pcidev);
 		lpfc_printf_log(phba, KERN_WARNING, LOG_INIT,
 				"0496 MSI pci_irq_vec failed (%d)\n", rc);
@@ -13270,6 +13271,7 @@ lpfc_sli4_enable_intr(struct lpfc_hba *phba, uint32_t cfg_mode)
 			eqhdl = lpfc_get_eq_hdl(0);
 			retval = pci_irq_vector(phba->pcidev, 0);
 			if (retval < 0) {
+				free_irq(phba->pcidev->irq, phba);
 				lpfc_printf_log(phba, KERN_WARNING, LOG_INIT,
 					"0502 INTR pci_irq_vec failed (%d)\n",
 					 retval);
-- 
2.39.5

[PATCH AUTOSEL 6.12 444/486] scsi: st: Restore some drive settings after reset

[Sasha Levin]

From: Kai Mäkisara <Kai.Makisara@kolumbus.fi>

[ Upstream commit 7081dc75df79696d8322d01821c28e53416c932c ]

Some of the allowed operations put the tape into a known position to
continue operation assuming only the tape position has changed.  But reset
sets partition, density and block size to drive default values. These
should be restored to the values before reset.

Normally the current block size and density are stored by the drive.  If
the settings have been changed, the changed values have to be saved by the
driver across reset.

Signed-off-by: Kai Mäkisara <Kai.Makisara@kolumbus.fi>
Link: https://lore.kernel.org/r/20250120194925.44432-2-Kai.Makisara@kolumbus.fi
Reviewed-by: John Meneghini <jmeneghi@redhat.com>
Tested-by: John Meneghini <jmeneghi@redhat.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 drivers/scsi/st.c | 24 +++++++++++++++++++++---
 drivers/scsi/st.h |  2 ++
 2 files changed, 23 insertions(+), 3 deletions(-)

diff --git a/drivers/scsi/st.c b/drivers/scsi/st.c
index 1cfd7e71dcdde..3e982c166baf9 100644
--- a/drivers/scsi/st.c
+++ b/drivers/scsi/st.c
@@ -952,7 +952,6 @@ static void reset_state(struct scsi_tape *STp)
 		STp->partition = find_partition(STp);
 		if (STp->partition < 0)
 			STp->partition = 0;
-		STp->new_partition = STp->partition;
 	}
 }
 \f
@@ -2926,14 +2925,17 @@ static int st_int_ioctl(struct scsi_tape *STp, unsigned int cmd_in, unsigned lon
 		if (cmd_in == MTSETDENSITY) {
 			(STp->buffer)->b_data[4] = arg;
 			STp->density_changed = 1;	/* At least we tried ;-) */
+			STp->changed_density = arg;
 		} else if (cmd_in == SET_DENS_AND_BLK)
 			(STp->buffer)->b_data[4] = arg >> 24;
 		else
 			(STp->buffer)->b_data[4] = STp->density;
 		if (cmd_in == MTSETBLK || cmd_in == SET_DENS_AND_BLK) {
 			ltmp = arg & MT_ST_BLKSIZE_MASK;
-			if (cmd_in == MTSETBLK)
+			if (cmd_in == MTSETBLK) {
 				STp->blksize_changed = 1; /* At least we tried ;-) */
+				STp->changed_blksize = arg;
+			}
 		} else
 			ltmp = STp->block_size;
 		(STp->buffer)->b_data[9] = (ltmp >> 16);
@@ -3634,9 +3636,25 @@ static long st_ioctl(struct file *file, unsigned int cmd_in, unsigned long arg)
 				retval = (-EIO);
 				goto out;
 			}
-			reset_state(STp);
+			reset_state(STp); /* Clears pos_unknown */
 			/* remove this when the midlevel properly clears was_reset */
 			STp->device->was_reset = 0;
+
+			/* Fix the device settings after reset, ignore errors */
+			if (mtc.mt_op == MTREW || mtc.mt_op == MTSEEK ||
+				mtc.mt_op == MTEOM) {
+				if (STp->can_partitions) {
+					/* STp->new_partition contains the
+					 *  latest partition set
+					 */
+					STp->partition = 0;
+					switch_partition(STp);
+				}
+				if (STp->density_changed)
+					st_int_ioctl(STp, MTSETDENSITY, STp->changed_density);
+				if (STp->blksize_changed)
+					st_int_ioctl(STp, MTSETBLK, STp->changed_blksize);
+			}
 		}
 
 		if (mtc.mt_op != MTNOP && mtc.mt_op != MTSETBLK &&
diff --git a/drivers/scsi/st.h b/drivers/scsi/st.h
index 1aaaf5369a40f..6d31b894ee84c 100644
--- a/drivers/scsi/st.h
+++ b/drivers/scsi/st.h
@@ -165,6 +165,7 @@ struct scsi_tape {
 	unsigned char compression_changed;
 	unsigned char drv_buffer;
 	unsigned char density;
+	unsigned char changed_density;
 	unsigned char door_locked;
 	unsigned char autorew_dev;   /* auto-rewind device */
 	unsigned char rew_at_close;  /* rewind necessary at close */
@@ -172,6 +173,7 @@ struct scsi_tape {
 	unsigned char cleaning_req;  /* cleaning requested? */
 	unsigned char first_tur;     /* first TEST UNIT READY */
 	int block_size;
+	int changed_blksize;
 	int min_block;
 	int max_block;
 	int recover_count;     /* From tape opening */
-- 
2.39.5