[PATCH AUTOSEL 4.19 05/39] scsi: hisi_sas: Propagate errors in interrupt_init_v1_hw()

[PATCH AUTOSEL 4.19 05/39] scsi: hisi_sas: Propagate errors in interrupt_init_v1_hw()

[Sasha Levin]

From: Sergey Shtylyov <s.shtylyov@omp.ru>

[ Upstream commit ab17122e758ef68fb21033e25c041144067975f5 ]

After commit 6c11dc060427 ("scsi: hisi_sas: Fix IRQ checks") we have the
error codes returned by platform_get_irq() ready for the propagation
upsream in interrupt_init_v1_hw() -- that will fix still broken deferred
probing. Let's propagate the error codes from devm_request_irq() as well
since I don't see the reason to override them with -ENOENT...

Link: https://lore.kernel.org/r/49ba93a3-d427-7542-d85a-b74fe1a33a73@omp.ru
Acked-by: John Garry <john.garry@huawei.com>
Signed-off-by: Sergey Shtylyov <s.shtylyov@omp.ru>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 drivers/scsi/hisi_sas/hisi_sas_v1_hw.c | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)

diff --git a/drivers/scsi/hisi_sas/hisi_sas_v1_hw.c b/drivers/scsi/hisi_sas/hisi_sas_v1_hw.c
index 8aa3222fe486..5a777e48963b 100644
--- a/drivers/scsi/hisi_sas/hisi_sas_v1_hw.c
+++ b/drivers/scsi/hisi_sas/hisi_sas_v1_hw.c
@@ -1697,7 +1697,7 @@ static int interrupt_init_v1_hw(struct hisi_hba *hisi_hba)
 				dev_err(dev,
 					"irq init: fail map phy interrupt %d\n",
 					idx);
-				return -ENOENT;
+				return irq;
 			}
 
 			rc = devm_request_irq(dev, irq, phy_interrupts[j], 0,
@@ -1706,7 +1706,7 @@ static int interrupt_init_v1_hw(struct hisi_hba *hisi_hba)
 				dev_err(dev, "irq init: could not request "
 					"phy interrupt %d, rc=%d\n",
 					irq, rc);
-				return -ENOENT;
+				return rc;
 			}
 		}
 	}
@@ -1717,7 +1717,7 @@ static int interrupt_init_v1_hw(struct hisi_hba *hisi_hba)
 		if (!irq) {
 			dev_err(dev, "irq init: could not map cq interrupt %d\n",
 				idx);
-			return -ENOENT;
+			return irq;
 		}
 
 		rc = devm_request_irq(dev, irq, cq_interrupt_v1_hw, 0,
@@ -1725,7 +1725,7 @@ static int interrupt_init_v1_hw(struct hisi_hba *hisi_hba)
 		if (rc) {
 			dev_err(dev, "irq init: could not request cq interrupt %d, rc=%d\n",
 				irq, rc);
-			return -ENOENT;
+			return rc;
 		}
 	}
 
@@ -1735,7 +1735,7 @@ static int interrupt_init_v1_hw(struct hisi_hba *hisi_hba)
 		if (!irq) {
 			dev_err(dev, "irq init: could not map fatal interrupt %d\n",
 				idx);
-			return -ENOENT;
+			return irq;
 		}
 
 		rc = devm_request_irq(dev, irq, fatal_interrupts[i], 0,
@@ -1744,7 +1744,7 @@ static int interrupt_init_v1_hw(struct hisi_hba *hisi_hba)
 			dev_err(dev,
 				"irq init: could not request fatal interrupt %d, rc=%d\n",
 				irq, rc);
-			return -ENOENT;
+			return rc;
 		}
 	}
 
-- 
2.30.2

Re: [PATCH AUTOSEL 4.19 05/39] scsi: hisi_sas: Propagate errors in interrupt_init_v1_hw()

[Sergey Shtylyov]

On 10.07.2021 5:31, Sasha Levin wrote:

> From: Sergey Shtylyov <s.shtylyov@omp.ru>
> 
> [ Upstream commit ab17122e758ef68fb21033e25c041144067975f5 ]
> 
> After commit 6c11dc060427 ("scsi: hisi_sas: Fix IRQ checks") we have the
> error codes returned by platform_get_irq() ready for the propagation
> upsream in interrupt_init_v1_hw() -- that will fix still broken deferred
> probing. Let's propagate the error codes from devm_request_irq() as well
> since I don't see the reason to override them with -ENOENT...
> 
> Link: https://lore.kernel.org/r/49ba93a3-d427-7542-d85a-b74fe1a33a73@omp.ru
> Acked-by: John Garry <john.garry@huawei.com>
> Signed-off-by: Sergey Shtylyov <s.shtylyov@omp.ru>
> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
> Signed-off-by: Sasha Levin <sashal@kernel.org>
> ---
>   drivers/scsi/hisi_sas/hisi_sas_v1_hw.c | 12 ++++++------
>   1 file changed, 6 insertions(+), 6 deletions(-)
> 
> diff --git a/drivers/scsi/hisi_sas/hisi_sas_v1_hw.c b/drivers/scsi/hisi_sas/hisi_sas_v1_hw.c
> index 8aa3222fe486..5a777e48963b 100644
> --- a/drivers/scsi/hisi_sas/hisi_sas_v1_hw.c
> +++ b/drivers/scsi/hisi_sas/hisi_sas_v1_hw.c
[...]
> @@ -1717,7 +1717,7 @@ static int interrupt_init_v1_hw(struct hisi_hba *hisi_hba)
>   		if (!irq) {
>   			dev_err(dev, "irq init: could not map cq interrupt %d\n",
>   				idx);
> -			return -ENOENT;
> +			return irq;

    This patch is borked too, we don't want to return 0 here...

[...]

MBR, Sergei

Re: [PATCH AUTOSEL 4.19 05/39] scsi: hisi_sas: Propagate errors in interrupt_init_v1_hw()

[Sasha Levin]

On Sat, Jul 10, 2021 at 12:20:43PM +0300, Sergey Shtylyov wrote:
>On 10.07.2021 5:31, Sasha Levin wrote:
>
>>From: Sergey Shtylyov <s.shtylyov@omp.ru>
>>
>>[ Upstream commit ab17122e758ef68fb21033e25c041144067975f5 ]
>>
>>After commit 6c11dc060427 ("scsi: hisi_sas: Fix IRQ checks") we have the
>>error codes returned by platform_get_irq() ready for the propagation
>>upsream in interrupt_init_v1_hw() -- that will fix still broken deferred
>>probing. Let's propagate the error codes from devm_request_irq() as well
>>since I don't see the reason to override them with -ENOENT...
>>
>>Link: https://lore.kernel.org/r/49ba93a3-d427-7542-d85a-b74fe1a33a73@omp.ru
>>Acked-by: John Garry <john.garry@huawei.com>
>>Signed-off-by: Sergey Shtylyov <s.shtylyov@omp.ru>
>>Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
>>Signed-off-by: Sasha Levin <sashal@kernel.org>
>>---
>>  drivers/scsi/hisi_sas/hisi_sas_v1_hw.c | 12 ++++++------
>>  1 file changed, 6 insertions(+), 6 deletions(-)
>>
>>diff --git a/drivers/scsi/hisi_sas/hisi_sas_v1_hw.c b/drivers/scsi/hisi_sas/hisi_sas_v1_hw.c
>>index 8aa3222fe486..5a777e48963b 100644
>>--- a/drivers/scsi/hisi_sas/hisi_sas_v1_hw.c
>>+++ b/drivers/scsi/hisi_sas/hisi_sas_v1_hw.c
>[...]
>>@@ -1717,7 +1717,7 @@ static int interrupt_init_v1_hw(struct hisi_hba *hisi_hba)
>>  		if (!irq) {
>>  			dev_err(dev, "irq init: could not map cq interrupt %d\n",
>>  				idx);
>>-			return -ENOENT;
>>+			return irq;
>
>   This patch is borked too, we don't want to return 0 here...

Looks like it's broken on <=4.19, I'll drop it. Thanks!

-- 
Thanks,
Sasha

Re: [PATCH AUTOSEL 4.19 05/39] scsi: hisi_sas: Propagate errors in interrupt_init_v1_hw()

[Sergey Shtylyov]

Hello!

On 18.07.2021 4:20, Sasha Levin wrote:

[...]
>>> [ Upstream commit ab17122e758ef68fb21033e25c041144067975f5 ]
>>>
>>> After commit 6c11dc060427 ("scsi: hisi_sas: Fix IRQ checks") we have the
>>> error codes returned by platform_get_irq() ready for the propagation
>>> upsream in interrupt_init_v1_hw() -- that will fix still broken deferred
>>> probing. Let's propagate the error codes from devm_request_irq() as well
>>> since I don't see the reason to override them with -ENOENT...
>>>
>>> Link: https://lore.kernel.org/r/49ba93a3-d427-7542-d85a-b74fe1a33a73@omp.ru
>>> Acked-by: John Garry <john.garry@huawei.com>
>>> Signed-off-by: Sergey Shtylyov <s.shtylyov@omp.ru>
>>> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
>>> Signed-off-by: Sasha Levin <sashal@kernel.org>
>>> ---
>>>  drivers/scsi/hisi_sas/hisi_sas_v1_hw.c | 12 ++++++------
>>>  1 file changed, 6 insertions(+), 6 deletions(-)
>>>
>>> diff --git a/drivers/scsi/hisi_sas/hisi_sas_v1_hw.c 
>>> b/drivers/scsi/hisi_sas/hisi_sas_v1_hw.c
>>> index 8aa3222fe486..5a777e48963b 100644
>>> --- a/drivers/scsi/hisi_sas/hisi_sas_v1_hw.c
>>> +++ b/drivers/scsi/hisi_sas/hisi_sas_v1_hw.c
>> [...]
>>> @@ -1717,7 +1717,7 @@ static int interrupt_init_v1_hw(struct hisi_hba 
>>> *hisi_hba)
>>>          if (!irq) {
>>>              dev_err(dev, "irq init: could not map cq interrupt %d\n",
>>>                  idx);
>>> -            return -ENOENT;
>>> +            return irq;
>>
>>   This patch is borked too, we don't want to return 0 here...
> 
> Looks like it's broken on <=4.19, I'll drop it. Thanks!

    You might to want backport the below patch (before this one),
the same way yoo did for 5.x-stable):

https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=6c11dc060427e07ca144eacaccd696106b361b06

MBR, Sergei

[PATCH AUTOSEL 4.19 06/39] scsi: lpfc: Fix "Unexpected timeout" error in direct attach topology

[Sasha Levin]

From: James Smart <jsmart2021@gmail.com>

[ Upstream commit e30d55137edef47434c40d7570276a0846fe922c ]

An 'unexpected timeout' message may be seen in a point-2-point topology.
The message occurs when a PLOGI is received before the driver is notified
of FLOGI completion. The FLOGI completion failure causes discovery to be
triggered for a second time. The discovery timer is restarted but no new
discovery activity is initiated, thus the timeout message eventually
appears.

In point-2-point, when discovery has progressed before the FLOGI completion
is processed, it is not a failure. Add code to FLOGI completion to detect
that discovery has progressed and exit the FLOGI handling (noop'ing it).

Link: https://lore.kernel.org/r/20210514195559.119853-4-jsmart2021@gmail.com
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>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 drivers/scsi/lpfc/lpfc_els.c | 9 +++++++++
 1 file changed, 9 insertions(+)

diff --git a/drivers/scsi/lpfc/lpfc_els.c b/drivers/scsi/lpfc/lpfc_els.c
index 6a4b496081e4..c4428249d3d7 100644
--- a/drivers/scsi/lpfc/lpfc_els.c
+++ b/drivers/scsi/lpfc/lpfc_els.c
@@ -1176,6 +1176,15 @@ lpfc_cmpl_els_flogi(struct lpfc_hba *phba, struct lpfc_iocbq *cmdiocb,
 			phba->fcf.fcf_redisc_attempted = 0; /* reset */
 			goto out;
 		}
+	} else if (vport->port_state > LPFC_FLOGI &&
+		   vport->fc_flag & FC_PT2PT) {
+		/*
+		 * In a p2p topology, it is possible that discovery has
+		 * already progressed, and this completion can be ignored.
+		 * Recheck the indicated topology.
+		 */
+		if (!sp->cmn.fPort)
+			goto out;
 	}
 
 flogifail:
-- 
2.30.2

[PATCH AUTOSEL 4.19 07/39] scsi: lpfc: Fix crash when lpfc_sli4_hba_setup() fails to initialize the SGLs

[Sasha Levin]

From: James Smart <jsmart2021@gmail.com>

[ Upstream commit 5aa615d195f1e142c662cb2253f057c9baec7531 ]

The driver is encountering a crash in lpfc_free_iocb_list() while
performing initial attachment.

Code review found this to be an errant failure path that was taken, jumping
to a tag that then referenced structures that were uninitialized.

Fix the failure path.

Link: https://lore.kernel.org/r/20210514195559.119853-9-jsmart2021@gmail.com
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>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 drivers/scsi/lpfc/lpfc_sli.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/drivers/scsi/lpfc/lpfc_sli.c b/drivers/scsi/lpfc/lpfc_sli.c
index f4633c9f8183..40d6537e64dd 100644
--- a/drivers/scsi/lpfc/lpfc_sli.c
+++ b/drivers/scsi/lpfc/lpfc_sli.c
@@ -7281,7 +7281,7 @@ lpfc_sli4_hba_setup(struct lpfc_hba *phba)
 				"0393 Error %d during rpi post operation\n",
 				rc);
 		rc = -ENODEV;
-		goto out_destroy_queue;
+		goto out_free_iocblist;
 	}
 	lpfc_sli4_node_prep(phba);
 
@@ -7406,8 +7406,9 @@ lpfc_sli4_hba_setup(struct lpfc_hba *phba)
 out_unset_queue:
 	/* Unset all the queues set up in this routine when error out */
 	lpfc_sli4_queue_unset(phba);
-out_destroy_queue:
+out_free_iocblist:
 	lpfc_free_iocb_list(phba);
+out_destroy_queue:
 	lpfc_sli4_queue_destroy(phba);
 out_stop_timers:
 	lpfc_stop_hba_timers(phba);
-- 
2.30.2

[PATCH AUTOSEL 4.19 08/39] scsi: core: Cap scsi_host cmd_per_lun at can_queue

[Sasha Levin]

From: John Garry <john.garry@huawei.com>

[ Upstream commit ea2f0f77538c50739b9fb4de4700cee5535e1f77 ]

The sysfs handling function sdev_store_queue_depth() enforces that the sdev
queue depth cannot exceed shost can_queue. The initial sdev queue depth
comes from shost cmd_per_lun. However, the LLDD may manually set
cmd_per_lun to be larger than can_queue, which leads to an initial sdev
queue depth greater than can_queue.

Such an issue was reported in [0], which caused a hang. That has since been
fixed in commit fc09acb7de31 ("scsi: scsi_debug: Fix cmd_per_lun, set to
max_queue").

Stop this possibly happening for other drivers by capping shost cmd_per_lun
at shost can_queue.

[0] https://lore.kernel.org/linux-scsi/YHaez6iN2HHYxYOh@T590/

Link: https://lore.kernel.org/r/1621434662-173079-1-git-send-email-john.garry@huawei.com
Reviewed-by: Ming Lei <ming.lei@redhat.com>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Signed-off-by: John Garry <john.garry@huawei.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 drivers/scsi/hosts.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/drivers/scsi/hosts.c b/drivers/scsi/hosts.c
index fa03be813f2c..9462564b2147 100644
--- a/drivers/scsi/hosts.c
+++ b/drivers/scsi/hosts.c
@@ -218,6 +218,9 @@ int scsi_add_host_with_dma(struct Scsi_Host *shost, struct device *dev,
 		goto fail;
 	}
 
+	shost->cmd_per_lun = min_t(short, shost->cmd_per_lun,
+				   shost->can_queue);
+
 	error = scsi_init_sense_cache(shost);
 	if (error)
 		goto fail;
-- 
2.30.2

[PATCH AUTOSEL 4.19 11/39] scsi: scsi_dh_alua: Check for negative result value

[Sasha Levin]

From: Hannes Reinecke <hare@suse.de>

[ Upstream commit 7e26e3ea028740f934477ec01ba586ab033c35aa ]

scsi_execute() will now return a negative error if there was an error prior
to command submission; evaluate that instead if checking for DRIVER_ERROR.

[mkp: build fix]

Link: https://lore.kernel.org/r/20210427083046.31620-6-hare@suse.de
Signed-off-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 drivers/scsi/device_handler/scsi_dh_alua.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/drivers/scsi/device_handler/scsi_dh_alua.c b/drivers/scsi/device_handler/scsi_dh_alua.c
index efd2b4312528..41e8c9e68878 100644
--- a/drivers/scsi/device_handler/scsi_dh_alua.c
+++ b/drivers/scsi/device_handler/scsi_dh_alua.c
@@ -562,12 +562,12 @@ static int alua_rtpg(struct scsi_device *sdev, struct alua_port_group *pg)
 			kfree(buff);
 			return SCSI_DH_OK;
 		}
-		if (!scsi_sense_valid(&sense_hdr)) {
+		if (retval < 0 || !scsi_sense_valid(&sense_hdr)) {
 			sdev_printk(KERN_INFO, sdev,
 				    "%s: rtpg failed, result %d\n",
 				    ALUA_DH_NAME, retval);
 			kfree(buff);
-			if (driver_byte(retval) == DRIVER_ERROR)
+			if (retval < 0)
 				return SCSI_DH_DEV_TEMP_BUSY;
 			return SCSI_DH_IO;
 		}
@@ -789,11 +789,11 @@ static unsigned alua_stpg(struct scsi_device *sdev, struct alua_port_group *pg)
 	retval = submit_stpg(sdev, pg->group_id, &sense_hdr);
 
 	if (retval) {
-		if (!scsi_sense_valid(&sense_hdr)) {
+		if (retval < 0 || !scsi_sense_valid(&sense_hdr)) {
 			sdev_printk(KERN_INFO, sdev,
 				    "%s: stpg failed, result %d",
 				    ALUA_DH_NAME, retval);
-			if (driver_byte(retval) == DRIVER_ERROR)
+			if (retval < 0)
 				return SCSI_DH_DEV_TEMP_BUSY;
 		} else {
 			sdev_printk(KERN_INFO, sdev, "%s: stpg failed\n",
-- 
2.30.2

[PATCH AUTOSEL 4.19 13/39] scsi: iscsi: Add iscsi_cls_conn refcount helpers

[Sasha Levin]

From: Mike Christie <michael.christie@oracle.com>

[ Upstream commit b1d19e8c92cfb0ded180ef3376c20e130414e067 ]

There are a couple places where we could free the iscsi_cls_conn while it's
still in use. This adds some helpers to get/put a refcount on the struct
and converts an exiting user. Subsequent commits will then use the helpers
to fix 2 bugs in the eh code.

Link: https://lore.kernel.org/r/20210525181821.7617-11-michael.christie@oracle.com
Reviewed-by: Lee Duncan <lduncan@suse.com>
Signed-off-by: Mike Christie <michael.christie@oracle.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 drivers/scsi/libiscsi.c             |  7 ++-----
 drivers/scsi/scsi_transport_iscsi.c | 12 ++++++++++++
 include/scsi/scsi_transport_iscsi.h |  2 ++
 3 files changed, 16 insertions(+), 5 deletions(-)

diff --git a/drivers/scsi/libiscsi.c b/drivers/scsi/libiscsi.c
index 81471c304991..52521b68f0a7 100644
--- a/drivers/scsi/libiscsi.c
+++ b/drivers/scsi/libiscsi.c
@@ -1385,7 +1385,6 @@ void iscsi_session_failure(struct iscsi_session *session,
 			   enum iscsi_err err)
 {
 	struct iscsi_conn *conn;
-	struct device *dev;
 
 	spin_lock_bh(&session->frwd_lock);
 	conn = session->leadconn;
@@ -1394,10 +1393,8 @@ void iscsi_session_failure(struct iscsi_session *session,
 		return;
 	}
 
-	dev = get_device(&conn->cls_conn->dev);
+	iscsi_get_conn(conn->cls_conn);
 	spin_unlock_bh(&session->frwd_lock);
-	if (!dev)
-	        return;
 	/*
 	 * if the host is being removed bypass the connection
 	 * recovery initialization because we are going to kill
@@ -1407,7 +1404,7 @@ void iscsi_session_failure(struct iscsi_session *session,
 		iscsi_conn_error_event(conn->cls_conn, err);
 	else
 		iscsi_conn_failure(conn, err);
-	put_device(dev);
+	iscsi_put_conn(conn->cls_conn);
 }
 EXPORT_SYMBOL_GPL(iscsi_session_failure);
 
diff --git a/drivers/scsi/scsi_transport_iscsi.c b/drivers/scsi/scsi_transport_iscsi.c
index e340b05278b6..2aaa5a2bd613 100644
--- a/drivers/scsi/scsi_transport_iscsi.c
+++ b/drivers/scsi/scsi_transport_iscsi.c
@@ -2306,6 +2306,18 @@ int iscsi_destroy_conn(struct iscsi_cls_conn *conn)
 }
 EXPORT_SYMBOL_GPL(iscsi_destroy_conn);
 
+void iscsi_put_conn(struct iscsi_cls_conn *conn)
+{
+	put_device(&conn->dev);
+}
+EXPORT_SYMBOL_GPL(iscsi_put_conn);
+
+void iscsi_get_conn(struct iscsi_cls_conn *conn)
+{
+	get_device(&conn->dev);
+}
+EXPORT_SYMBOL_GPL(iscsi_get_conn);
+
 /*
  * iscsi interface functions
  */
diff --git a/include/scsi/scsi_transport_iscsi.h b/include/scsi/scsi_transport_iscsi.h
index b266d2a3bcb1..484e9787d817 100644
--- a/include/scsi/scsi_transport_iscsi.h
+++ b/include/scsi/scsi_transport_iscsi.h
@@ -436,6 +436,8 @@ extern void iscsi_remove_session(struct iscsi_cls_session *session);
 extern void iscsi_free_session(struct iscsi_cls_session *session);
 extern struct iscsi_cls_conn *iscsi_create_conn(struct iscsi_cls_session *sess,
 						int dd_size, uint32_t cid);
+extern void iscsi_put_conn(struct iscsi_cls_conn *conn);
+extern void iscsi_get_conn(struct iscsi_cls_conn *conn);
 extern int iscsi_destroy_conn(struct iscsi_cls_conn *conn);
 extern void iscsi_unblock_session(struct iscsi_cls_session *session);
 extern void iscsi_block_session(struct iscsi_cls_session *session);
-- 
2.30.2

[PATCH AUTOSEL 4.19 14/39] scsi: iscsi: Fix conn use after free during resets

[Sasha Levin]

From: Mike Christie <michael.christie@oracle.com>

[ Upstream commit ec29d0ac29be366450a7faffbcf8cba3a6a3b506 ]

If we haven't done a unbind target call we can race where
iscsi_conn_teardown wakes up the EH thread and then frees the conn while
those threads are still accessing the conn ehwait.

We can only do one TMF per session so this just moves the TMF fields from
the conn to the session. We can then rely on the
iscsi_session_teardown->iscsi_remove_session->__iscsi_unbind_session call
to remove the target and it's devices, and know after that point there is
no device or scsi-ml callout trying to access the session.

Link: https://lore.kernel.org/r/20210525181821.7617-14-michael.christie@oracle.com
Reviewed-by: Lee Duncan <lduncan@suse.com>
Signed-off-by: Mike Christie <michael.christie@oracle.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 drivers/scsi/libiscsi.c | 115 +++++++++++++++++++---------------------
 include/scsi/libiscsi.h |  11 ++--
 2 files changed, 60 insertions(+), 66 deletions(-)

diff --git a/drivers/scsi/libiscsi.c b/drivers/scsi/libiscsi.c
index 52521b68f0a7..5607fe8541c3 100644
--- a/drivers/scsi/libiscsi.c
+++ b/drivers/scsi/libiscsi.c
@@ -259,11 +259,11 @@ static int iscsi_prep_bidi_ahs(struct iscsi_task *task)
  */
 static int iscsi_check_tmf_restrictions(struct iscsi_task *task, int opcode)
 {
-	struct iscsi_conn *conn = task->conn;
-	struct iscsi_tm *tmf = &conn->tmhdr;
+	struct iscsi_session *session = task->conn->session;
+	struct iscsi_tm *tmf = &session->tmhdr;
 	u64 hdr_lun;
 
-	if (conn->tmf_state == TMF_INITIAL)
+	if (session->tmf_state == TMF_INITIAL)
 		return 0;
 
 	if ((tmf->opcode & ISCSI_OPCODE_MASK) != ISCSI_OP_SCSI_TMFUNC)
@@ -283,24 +283,19 @@ static int iscsi_check_tmf_restrictions(struct iscsi_task *task, int opcode)
 		 * Fail all SCSI cmd PDUs
 		 */
 		if (opcode != ISCSI_OP_SCSI_DATA_OUT) {
-			iscsi_conn_printk(KERN_INFO, conn,
-					  "task [op %x itt "
-					  "0x%x/0x%x] "
-					  "rejected.\n",
-					  opcode, task->itt,
-					  task->hdr_itt);
+			iscsi_session_printk(KERN_INFO, session,
+					     "task [op %x itt 0x%x/0x%x] rejected.\n",
+					     opcode, task->itt, task->hdr_itt);
 			return -EACCES;
 		}
 		/*
 		 * And also all data-out PDUs in response to R2T
 		 * if fast_abort is set.
 		 */
-		if (conn->session->fast_abort) {
-			iscsi_conn_printk(KERN_INFO, conn,
-					  "task [op %x itt "
-					  "0x%x/0x%x] fast abort.\n",
-					  opcode, task->itt,
-					  task->hdr_itt);
+		if (session->fast_abort) {
+			iscsi_session_printk(KERN_INFO, session,
+					     "task [op %x itt 0x%x/0x%x] fast abort.\n",
+					     opcode, task->itt, task->hdr_itt);
 			return -EACCES;
 		}
 		break;
@@ -313,7 +308,7 @@ static int iscsi_check_tmf_restrictions(struct iscsi_task *task, int opcode)
 		 */
 		if (opcode == ISCSI_OP_SCSI_DATA_OUT &&
 		    task->hdr_itt == tmf->rtt) {
-			ISCSI_DBG_SESSION(conn->session,
+			ISCSI_DBG_SESSION(session,
 					  "Preventing task %x/%x from sending "
 					  "data-out due to abort task in "
 					  "progress\n", task->itt,
@@ -970,20 +965,21 @@ iscsi_data_in_rsp(struct iscsi_conn *conn, struct iscsi_hdr *hdr,
 static void iscsi_tmf_rsp(struct iscsi_conn *conn, struct iscsi_hdr *hdr)
 {
 	struct iscsi_tm_rsp *tmf = (struct iscsi_tm_rsp *)hdr;
+	struct iscsi_session *session = conn->session;
 
 	conn->exp_statsn = be32_to_cpu(hdr->statsn) + 1;
 	conn->tmfrsp_pdus_cnt++;
 
-	if (conn->tmf_state != TMF_QUEUED)
+	if (session->tmf_state != TMF_QUEUED)
 		return;
 
 	if (tmf->response == ISCSI_TMF_RSP_COMPLETE)
-		conn->tmf_state = TMF_SUCCESS;
+		session->tmf_state = TMF_SUCCESS;
 	else if (tmf->response == ISCSI_TMF_RSP_NO_TASK)
-		conn->tmf_state = TMF_NOT_FOUND;
+		session->tmf_state = TMF_NOT_FOUND;
 	else
-		conn->tmf_state = TMF_FAILED;
-	wake_up(&conn->ehwait);
+		session->tmf_state = TMF_FAILED;
+	wake_up(&session->ehwait);
 }
 
 static int iscsi_send_nopout(struct iscsi_conn *conn, struct iscsi_nopin *rhdr)
@@ -1822,15 +1818,14 @@ EXPORT_SYMBOL_GPL(iscsi_target_alloc);
 
 static void iscsi_tmf_timedout(struct timer_list *t)
 {
-	struct iscsi_conn *conn = from_timer(conn, t, tmf_timer);
-	struct iscsi_session *session = conn->session;
+	struct iscsi_session *session = from_timer(session, t, tmf_timer);
 
 	spin_lock(&session->frwd_lock);
-	if (conn->tmf_state == TMF_QUEUED) {
-		conn->tmf_state = TMF_TIMEDOUT;
+	if (session->tmf_state == TMF_QUEUED) {
+		session->tmf_state = TMF_TIMEDOUT;
 		ISCSI_DBG_EH(session, "tmf timedout\n");
 		/* unblock eh_abort() */
-		wake_up(&conn->ehwait);
+		wake_up(&session->ehwait);
 	}
 	spin_unlock(&session->frwd_lock);
 }
@@ -1853,8 +1848,8 @@ static int iscsi_exec_task_mgmt_fn(struct iscsi_conn *conn,
 		return -EPERM;
 	}
 	conn->tmfcmd_pdus_cnt++;
-	conn->tmf_timer.expires = timeout * HZ + jiffies;
-	add_timer(&conn->tmf_timer);
+	session->tmf_timer.expires = timeout * HZ + jiffies;
+	add_timer(&session->tmf_timer);
 	ISCSI_DBG_EH(session, "tmf set timeout\n");
 
 	spin_unlock_bh(&session->frwd_lock);
@@ -1868,12 +1863,12 @@ static int iscsi_exec_task_mgmt_fn(struct iscsi_conn *conn,
 	 * 3) session is terminated or restarted or userspace has
 	 * given up on recovery
 	 */
-	wait_event_interruptible(conn->ehwait, age != session->age ||
+	wait_event_interruptible(session->ehwait, age != session->age ||
 				 session->state != ISCSI_STATE_LOGGED_IN ||
-				 conn->tmf_state != TMF_QUEUED);
+				 session->tmf_state != TMF_QUEUED);
 	if (signal_pending(current))
 		flush_signals(current);
-	del_timer_sync(&conn->tmf_timer);
+	del_timer_sync(&session->tmf_timer);
 
 	mutex_lock(&session->eh_mutex);
 	spin_lock_bh(&session->frwd_lock);
@@ -2233,17 +2228,17 @@ int iscsi_eh_abort(struct scsi_cmnd *sc)
 	}
 
 	/* only have one tmf outstanding at a time */
-	if (conn->tmf_state != TMF_INITIAL)
+	if (session->tmf_state != TMF_INITIAL)
 		goto failed;
-	conn->tmf_state = TMF_QUEUED;
+	session->tmf_state = TMF_QUEUED;
 
-	hdr = &conn->tmhdr;
+	hdr = &session->tmhdr;
 	iscsi_prep_abort_task_pdu(task, hdr);
 
 	if (iscsi_exec_task_mgmt_fn(conn, hdr, age, session->abort_timeout))
 		goto failed;
 
-	switch (conn->tmf_state) {
+	switch (session->tmf_state) {
 	case TMF_SUCCESS:
 		spin_unlock_bh(&session->frwd_lock);
 		/*
@@ -2258,7 +2253,7 @@ int iscsi_eh_abort(struct scsi_cmnd *sc)
 		 */
 		spin_lock_bh(&session->frwd_lock);
 		fail_scsi_task(task, DID_ABORT);
-		conn->tmf_state = TMF_INITIAL;
+		session->tmf_state = TMF_INITIAL;
 		memset(hdr, 0, sizeof(*hdr));
 		spin_unlock_bh(&session->frwd_lock);
 		iscsi_start_tx(conn);
@@ -2269,7 +2264,7 @@ int iscsi_eh_abort(struct scsi_cmnd *sc)
 		goto failed_unlocked;
 	case TMF_NOT_FOUND:
 		if (!sc->SCp.ptr) {
-			conn->tmf_state = TMF_INITIAL;
+			session->tmf_state = TMF_INITIAL;
 			memset(hdr, 0, sizeof(*hdr));
 			/* task completed before tmf abort response */
 			ISCSI_DBG_EH(session, "sc completed while abort	in "
@@ -2278,7 +2273,7 @@ int iscsi_eh_abort(struct scsi_cmnd *sc)
 		}
 		/* fall through */
 	default:
-		conn->tmf_state = TMF_INITIAL;
+		session->tmf_state = TMF_INITIAL;
 		goto failed;
 	}
 
@@ -2335,11 +2330,11 @@ int iscsi_eh_device_reset(struct scsi_cmnd *sc)
 	conn = session->leadconn;
 
 	/* only have one tmf outstanding at a time */
-	if (conn->tmf_state != TMF_INITIAL)
+	if (session->tmf_state != TMF_INITIAL)
 		goto unlock;
-	conn->tmf_state = TMF_QUEUED;
+	session->tmf_state = TMF_QUEUED;
 
-	hdr = &conn->tmhdr;
+	hdr = &session->tmhdr;
 	iscsi_prep_lun_reset_pdu(sc, hdr);
 
 	if (iscsi_exec_task_mgmt_fn(conn, hdr, session->age,
@@ -2348,7 +2343,7 @@ int iscsi_eh_device_reset(struct scsi_cmnd *sc)
 		goto unlock;
 	}
 
-	switch (conn->tmf_state) {
+	switch (session->tmf_state) {
 	case TMF_SUCCESS:
 		break;
 	case TMF_TIMEDOUT:
@@ -2356,7 +2351,7 @@ int iscsi_eh_device_reset(struct scsi_cmnd *sc)
 		iscsi_conn_failure(conn, ISCSI_ERR_SCSI_EH_SESSION_RST);
 		goto done;
 	default:
-		conn->tmf_state = TMF_INITIAL;
+		session->tmf_state = TMF_INITIAL;
 		goto unlock;
 	}
 
@@ -2368,7 +2363,7 @@ int iscsi_eh_device_reset(struct scsi_cmnd *sc)
 	spin_lock_bh(&session->frwd_lock);
 	memset(hdr, 0, sizeof(*hdr));
 	fail_scsi_tasks(conn, sc->device->lun, DID_ERROR);
-	conn->tmf_state = TMF_INITIAL;
+	session->tmf_state = TMF_INITIAL;
 	spin_unlock_bh(&session->frwd_lock);
 
 	iscsi_start_tx(conn);
@@ -2391,8 +2386,7 @@ void iscsi_session_recovery_timedout(struct iscsi_cls_session *cls_session)
 	spin_lock_bh(&session->frwd_lock);
 	if (session->state != ISCSI_STATE_LOGGED_IN) {
 		session->state = ISCSI_STATE_RECOVERY_FAILED;
-		if (session->leadconn)
-			wake_up(&session->leadconn->ehwait);
+		wake_up(&session->ehwait);
 	}
 	spin_unlock_bh(&session->frwd_lock);
 }
@@ -2437,7 +2431,7 @@ int iscsi_eh_session_reset(struct scsi_cmnd *sc)
 	iscsi_conn_failure(conn, ISCSI_ERR_SCSI_EH_SESSION_RST);
 
 	ISCSI_DBG_EH(session, "wait for relogin\n");
-	wait_event_interruptible(conn->ehwait,
+	wait_event_interruptible(session->ehwait,
 				 session->state == ISCSI_STATE_TERMINATE ||
 				 session->state == ISCSI_STATE_LOGGED_IN ||
 				 session->state == ISCSI_STATE_RECOVERY_FAILED);
@@ -2498,11 +2492,11 @@ static int iscsi_eh_target_reset(struct scsi_cmnd *sc)
 	conn = session->leadconn;
 
 	/* only have one tmf outstanding at a time */
-	if (conn->tmf_state != TMF_INITIAL)
+	if (session->tmf_state != TMF_INITIAL)
 		goto unlock;
-	conn->tmf_state = TMF_QUEUED;
+	session->tmf_state = TMF_QUEUED;
 
-	hdr = &conn->tmhdr;
+	hdr = &session->tmhdr;
 	iscsi_prep_tgt_reset_pdu(sc, hdr);
 
 	if (iscsi_exec_task_mgmt_fn(conn, hdr, session->age,
@@ -2511,7 +2505,7 @@ static int iscsi_eh_target_reset(struct scsi_cmnd *sc)
 		goto unlock;
 	}
 
-	switch (conn->tmf_state) {
+	switch (session->tmf_state) {
 	case TMF_SUCCESS:
 		break;
 	case TMF_TIMEDOUT:
@@ -2519,7 +2513,7 @@ static int iscsi_eh_target_reset(struct scsi_cmnd *sc)
 		iscsi_conn_failure(conn, ISCSI_ERR_SCSI_EH_SESSION_RST);
 		goto done;
 	default:
-		conn->tmf_state = TMF_INITIAL;
+		session->tmf_state = TMF_INITIAL;
 		goto unlock;
 	}
 
@@ -2531,7 +2525,7 @@ static int iscsi_eh_target_reset(struct scsi_cmnd *sc)
 	spin_lock_bh(&session->frwd_lock);
 	memset(hdr, 0, sizeof(*hdr));
 	fail_scsi_tasks(conn, -1, DID_ERROR);
-	conn->tmf_state = TMF_INITIAL;
+	session->tmf_state = TMF_INITIAL;
 	spin_unlock_bh(&session->frwd_lock);
 
 	iscsi_start_tx(conn);
@@ -2836,7 +2830,10 @@ iscsi_session_setup(struct iscsi_transport *iscsit, struct Scsi_Host *shost,
 	session->tt = iscsit;
 	session->dd_data = cls_session->dd_data + sizeof(*session);
 
+	session->tmf_state = TMF_INITIAL;
+	timer_setup(&session->tmf_timer, iscsi_tmf_timedout, 0);
 	mutex_init(&session->eh_mutex);
+
 	spin_lock_init(&session->frwd_lock);
 	spin_lock_init(&session->back_lock);
 
@@ -2940,7 +2937,6 @@ iscsi_conn_setup(struct iscsi_cls_session *cls_session, int dd_size,
 	conn->c_stage = ISCSI_CONN_INITIAL_STAGE;
 	conn->id = conn_idx;
 	conn->exp_statsn = 0;
-	conn->tmf_state = TMF_INITIAL;
 
 	timer_setup(&conn->transport_timer, iscsi_check_transport_timeouts, 0);
 
@@ -2966,8 +2962,7 @@ iscsi_conn_setup(struct iscsi_cls_session *cls_session, int dd_size,
 		goto login_task_data_alloc_fail;
 	conn->login_task->data = conn->data = data;
 
-	timer_setup(&conn->tmf_timer, iscsi_tmf_timedout, 0);
-	init_waitqueue_head(&conn->ehwait);
+	init_waitqueue_head(&session->ehwait);
 
 	return cls_conn;
 
@@ -3002,7 +2997,7 @@ void iscsi_conn_teardown(struct iscsi_cls_conn *cls_conn)
 		 * leading connection? then give up on recovery.
 		 */
 		session->state = ISCSI_STATE_TERMINATE;
-		wake_up(&conn->ehwait);
+		wake_up(&session->ehwait);
 	}
 	spin_unlock_bh(&session->frwd_lock);
 
@@ -3077,7 +3072,7 @@ int iscsi_conn_start(struct iscsi_cls_conn *cls_conn)
 		 * commands after successful recovery
 		 */
 		conn->stop_stage = 0;
-		conn->tmf_state = TMF_INITIAL;
+		session->tmf_state = TMF_INITIAL;
 		session->age++;
 		if (session->age == 16)
 			session->age = 0;
@@ -3091,7 +3086,7 @@ int iscsi_conn_start(struct iscsi_cls_conn *cls_conn)
 	spin_unlock_bh(&session->frwd_lock);
 
 	iscsi_unblock_session(session->cls_session);
-	wake_up(&conn->ehwait);
+	wake_up(&session->ehwait);
 	return 0;
 }
 EXPORT_SYMBOL_GPL(iscsi_conn_start);
@@ -3177,7 +3172,7 @@ static void iscsi_start_session_recovery(struct iscsi_session *session,
 	spin_lock_bh(&session->frwd_lock);
 	fail_scsi_tasks(conn, -1, DID_TRANSPORT_DISRUPTED);
 	fail_mgmt_tasks(session, conn);
-	memset(&conn->tmhdr, 0, sizeof(conn->tmhdr));
+	memset(&session->tmhdr, 0, sizeof(session->tmhdr));
 	spin_unlock_bh(&session->frwd_lock);
 	mutex_unlock(&session->eh_mutex);
 }
diff --git a/include/scsi/libiscsi.h b/include/scsi/libiscsi.h
index 1ee0f30ae190..647f1e0e726c 100644
--- a/include/scsi/libiscsi.h
+++ b/include/scsi/libiscsi.h
@@ -208,12 +208,6 @@ struct iscsi_conn {
 	unsigned long		suspend_tx;	/* suspend Tx */
 	unsigned long		suspend_rx;	/* suspend Rx */
 
-	/* abort */
-	wait_queue_head_t	ehwait;		/* used in eh_abort() */
-	struct iscsi_tm		tmhdr;
-	struct timer_list	tmf_timer;
-	int			tmf_state;	/* see TMF_INITIAL, etc.*/
-
 	/* negotiated params */
 	unsigned		max_recv_dlength; /* initiator_max_recv_dsl*/
 	unsigned		max_xmit_dlength; /* target_max_recv_dsl */
@@ -283,6 +277,11 @@ struct iscsi_session {
 	 * and recv lock.
 	 */
 	struct mutex		eh_mutex;
+	/* abort */
+	wait_queue_head_t	ehwait;		/* used in eh_abort() */
+	struct iscsi_tm		tmhdr;
+	struct timer_list	tmf_timer;
+	int			tmf_state;	/* see TMF_INITIAL, etc.*/
 
 	/* iSCSI session-wide sequencing */
 	uint32_t		cmdsn;
-- 
2.30.2

[PATCH AUTOSEL 4.19 15/39] scsi: iscsi: Fix shost->max_id use

[Sasha Levin]

From: Mike Christie <michael.christie@oracle.com>

[ Upstream commit bdd4aad7ff92ae39c2e93c415bb6761cb8b584da ]

The iscsi offload drivers are setting the shost->max_id to the max number
of sessions they support. The problem is that max_id is not the max number
of targets but the highest identifier the targets can have. To use it to
limit the number of targets we need to set it to max sessions - 1, or we
can end up with a session we might not have preallocated resources for.

Link: https://lore.kernel.org/r/20210525181821.7617-15-michael.christie@oracle.com
Reviewed-by: Lee Duncan <lduncan@suse.com>
Signed-off-by: Mike Christie <michael.christie@oracle.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 drivers/scsi/be2iscsi/be_main.c  | 4 ++--
 drivers/scsi/bnx2i/bnx2i_iscsi.c | 2 +-
 drivers/scsi/cxgbi/libcxgbi.c    | 4 ++--
 drivers/scsi/qedi/qedi_main.c    | 2 +-
 4 files changed, 6 insertions(+), 6 deletions(-)

diff --git a/drivers/scsi/be2iscsi/be_main.c b/drivers/scsi/be2iscsi/be_main.c
index 3660059784f7..a5b807c676fc 100644
--- a/drivers/scsi/be2iscsi/be_main.c
+++ b/drivers/scsi/be2iscsi/be_main.c
@@ -423,7 +423,7 @@ static struct beiscsi_hba *beiscsi_hba_alloc(struct pci_dev *pcidev)
 			"beiscsi_hba_alloc - iscsi_host_alloc failed\n");
 		return NULL;
 	}
-	shost->max_id = BE2_MAX_SESSIONS;
+	shost->max_id = BE2_MAX_SESSIONS - 1;
 	shost->max_channel = 0;
 	shost->max_cmd_len = BEISCSI_MAX_CMD_LEN;
 	shost->max_lun = BEISCSI_NUM_MAX_LUN;
@@ -5336,7 +5336,7 @@ static int beiscsi_enable_port(struct beiscsi_hba *phba)
 	/* Re-enable UER. If different TPE occurs then it is recoverable. */
 	beiscsi_set_uer_feature(phba);
 
-	phba->shost->max_id = phba->params.cxns_per_ctrl;
+	phba->shost->max_id = phba->params.cxns_per_ctrl - 1;
 	phba->shost->can_queue = phba->params.ios_per_ctrl;
 	ret = beiscsi_init_port(phba);
 	if (ret < 0) {
diff --git a/drivers/scsi/bnx2i/bnx2i_iscsi.c b/drivers/scsi/bnx2i/bnx2i_iscsi.c
index 718a0335cfbb..9a5795d5a347 100644
--- a/drivers/scsi/bnx2i/bnx2i_iscsi.c
+++ b/drivers/scsi/bnx2i/bnx2i_iscsi.c
@@ -793,7 +793,7 @@ struct bnx2i_hba *bnx2i_alloc_hba(struct cnic_dev *cnic)
 		return NULL;
 	shost->dma_boundary = cnic->pcidev->dma_mask;
 	shost->transportt = bnx2i_scsi_xport_template;
-	shost->max_id = ISCSI_MAX_CONNS_PER_HBA;
+	shost->max_id = ISCSI_MAX_CONNS_PER_HBA - 1;
 	shost->max_channel = 0;
 	shost->max_lun = 512;
 	shost->max_cmd_len = 16;
diff --git a/drivers/scsi/cxgbi/libcxgbi.c b/drivers/scsi/cxgbi/libcxgbi.c
index 559d8bda4cca..01f5ebee3229 100644
--- a/drivers/scsi/cxgbi/libcxgbi.c
+++ b/drivers/scsi/cxgbi/libcxgbi.c
@@ -339,7 +339,7 @@ void cxgbi_hbas_remove(struct cxgbi_device *cdev)
 EXPORT_SYMBOL_GPL(cxgbi_hbas_remove);
 
 int cxgbi_hbas_add(struct cxgbi_device *cdev, u64 max_lun,
-		unsigned int max_id, struct scsi_host_template *sht,
+		unsigned int max_conns, struct scsi_host_template *sht,
 		struct scsi_transport_template *stt)
 {
 	struct cxgbi_hba *chba;
@@ -359,7 +359,7 @@ int cxgbi_hbas_add(struct cxgbi_device *cdev, u64 max_lun,
 
 		shost->transportt = stt;
 		shost->max_lun = max_lun;
-		shost->max_id = max_id;
+		shost->max_id = max_conns - 1;
 		shost->max_channel = 0;
 		shost->max_cmd_len = 16;
 
diff --git a/drivers/scsi/qedi/qedi_main.c b/drivers/scsi/qedi/qedi_main.c
index fe26144d390a..7665fd641886 100644
--- a/drivers/scsi/qedi/qedi_main.c
+++ b/drivers/scsi/qedi/qedi_main.c
@@ -629,7 +629,7 @@ static struct qedi_ctx *qedi_host_alloc(struct pci_dev *pdev)
 		goto exit_setup_shost;
 	}
 
-	shost->max_id = QEDI_MAX_ISCSI_CONNS_PER_HBA;
+	shost->max_id = QEDI_MAX_ISCSI_CONNS_PER_HBA - 1;
 	shost->max_channel = 0;
 	shost->max_lun = ~0;
 	shost->max_cmd_len = 16;
-- 
2.30.2

[PATCH AUTOSEL 4.19 16/39] scsi: qedi: Fix null ref during abort handling

[Sasha Levin]

From: Mike Christie <michael.christie@oracle.com>

[ Upstream commit 5777b7f0f03ce49372203b6521631f62f2810c8f ]

If qedi_process_cmd_cleanup_resp finds the cmd it frees the work and sets
list_tmf_work to NULL, so qedi_tmf_work should check if list_tmf_work is
non-NULL when it wants to force cleanup.

Link: https://lore.kernel.org/r/20210525181821.7617-20-michael.christie@oracle.com
Reviewed-by: Manish Rangankar <mrangankar@marvell.com>
Signed-off-by: Mike Christie <michael.christie@oracle.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 drivers/scsi/qedi/qedi_fw.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/scsi/qedi/qedi_fw.c b/drivers/scsi/qedi/qedi_fw.c
index 357a0acc5ed2..b60b48f3b984 100644
--- a/drivers/scsi/qedi/qedi_fw.c
+++ b/drivers/scsi/qedi/qedi_fw.c
@@ -1466,7 +1466,7 @@ static void qedi_tmf_work(struct work_struct *work)
 
 ldel_exit:
 	spin_lock_bh(&qedi_conn->tmf_work_lock);
-	if (!qedi_cmd->list_tmf_work) {
+	if (qedi_cmd->list_tmf_work) {
 		list_del_init(&list_work->list);
 		qedi_cmd->list_tmf_work = NULL;
 		kfree(list_work);
-- 
2.30.2