All of lore.kernel.org
 help / color / mirror / Atom feed
* [PATCH] [scsi] : pm80xx : fix problem of pm8001_work_fn reseting uncorrect phy device
@ 2014-02-14  8:01 XinHong Zhu
  2014-02-14  9:00 ` lindar_liu
  0 siblings, 1 reply; 3+ messages in thread
From: XinHong Zhu @ 2014-02-14  8:01 UTC (permalink / raw)
  To: linux-scsi; +Cc: lindar_liu, XinHong Zhu

If a phy device is removed ,the device can get error of I/O and HBA maybe receieve
IO_OPEN_CNX_ERROR_IT_NEXUS_LOSS of event which causes pm8001_work_fn to reset the phy
device ,but in pm8001_task_exec don't assign a value for field device of ccb  and in
other case a ccb used have device field set , when  ccb is freeed the field device of
the ccb don't be set NULL.So there is possibility of getting another devcie reset in fun
of mpi_ssp_completion .Also there are another way to solve problem by adding following
code in mpi_SSP_completion:
  pm8001_dev = t->dev->lldd_dev;
Signed-off-by: zhuxh <zxh3737@163.com>
---
 drivers/scsi/pm8001/pm8001_sas.c |    1 +
 1 files changed, 1 insertions(+), 0 deletions(-)

diff --git a/drivers/scsi/pm8001/pm8001_sas.c b/drivers/scsi/pm8001/pm8001_sas.c
index f50ac44..f0ea5db 100644
--- a/drivers/scsi/pm8001/pm8001_sas.c
+++ b/drivers/scsi/pm8001/pm8001_sas.c
@@ -434,6 +434,7 @@ static int pm8001_task_exec(struct sas_task *task, const int num,
 		ccb->n_elem = n_elem;
 		ccb->ccb_tag = tag;
 		ccb->task = t;
+		ccb->device = pm8001_dev;
 		switch (t->task_proto) {
 		case SAS_PROTOCOL_SMP:
 			rc = pm8001_task_prep_smp(pm8001_ha, ccb);
-- 
1.7.9



^ permalink raw reply related	[flat|nested] 3+ messages in thread

* RE: [PATCH] [scsi] : pm80xx : fix problem of pm8001_work_fn reseting uncorrect phy device
  2014-02-14  8:01 [PATCH] [scsi] : pm80xx : fix problem of pm8001_work_fn reseting uncorrect phy device XinHong Zhu
@ 2014-02-14  9:00 ` lindar_liu
  0 siblings, 0 replies; 3+ messages in thread
From: lindar_liu @ 2014-02-14  9:00 UTC (permalink / raw)
  To: 'XinHong Zhu', linux-scsi

On Friday, February 14, 2014 4:01 PM, XinHong Zhu wrote: 
> If a phy device is removed ,the device can get error of I/O and HBA maybe
> receieve IO_OPEN_CNX_ERROR_IT_NEXUS_LOSS of event which causes
pm8001_work_fn
> to reset the phy device ,but in pm8001_task_exec don't assign a value for
> field device of ccb  and in other case a ccb used have device field set ,
when
> ccb is freeed the field device of the ccb don't be set NULL.So there is
> possibility of getting another devcie reset in fun of mpi_ssp_completion
.Also
> there are another way to solve problem by adding following code in
> mpi_SSP_completion:
>   pm8001_dev = t->dev->lldd_dev;
> Signed-off-by: zhuxh <zxh3737@163.com>
> ---
>  drivers/scsi/pm8001/pm8001_sas.c |    1 +
>  1 files changed, 1 insertions(+), 0 deletions(-)
> 
> diff --git a/drivers/scsi/pm8001/pm8001_sas.c
> b/drivers/scsi/pm8001/pm8001_sas.c
> index f50ac44..f0ea5db 100644
> --- a/drivers/scsi/pm8001/pm8001_sas.c
> +++ b/drivers/scsi/pm8001/pm8001_sas.c
> @@ -434,6 +434,7 @@ static int pm8001_task_exec(struct sas_task *task,
const
> int num,
>  		ccb->n_elem = n_elem;
>  		ccb->ccb_tag = tag;
>  		ccb->task = t;
> +		ccb->device = pm8001_dev;
>  		switch (t->task_proto) {
>  		case SAS_PROTOCOL_SMP:
>  			rc = pm8001_task_prep_smp(pm8001_ha, ccb);

Acked-by: Lindar Liu <lindar_liu@usish.com>

This is the best way to fix such issue. Thanks.

> --
> 1.7.9



^ permalink raw reply	[flat|nested] 3+ messages in thread

* [PATCH] [scsi] : pm80xx : fix problem of pm8001_work_fn reseting uncorrect phy device
@ 2014-01-14  9:41 XinHong Zhu
  0 siblings, 0 replies; 3+ messages in thread
From: XinHong Zhu @ 2014-01-14  9:41 UTC (permalink / raw)
  To: linux-scsi; +Cc: XinHong Zhu

If a phy device is removed ,the device can get error of I/O and HBA maybe receieve
IO_OPEN_CNX_ERROR_IT_NEXUS_LOSS of event which causes pm8001_work_fn to reset the phy
device ,but in pm8001_task_exec don't assign a value for field device of ccb  and in
other case a ccb used have device field set , when  ccb is freeed the field device of
the ccb don't be set NULL.So there is possibility of getting another devcie reset in fun
of mpi_ssp_completion .Also there are another way to solve problem by adding following
code in mpi_SSP_completion:
  pm8001_dev = t->dev->lldd_dev;
Signed-off-by: zhuxh <zxh3737@163.com>
---
 drivers/scsi/pm8001/pm8001_sas.c |    1 +
 1 files changed, 1 insertions(+), 0 deletions(-)

diff --git a/drivers/scsi/pm8001/pm8001_sas.c b/drivers/scsi/pm8001/pm8001_sas.c
index f50ac44..f0ea5db 100644
--- a/drivers/scsi/pm8001/pm8001_sas.c
+++ b/drivers/scsi/pm8001/pm8001_sas.c
@@ -434,6 +434,7 @@ static int pm8001_task_exec(struct sas_task *task, const int num,
 		ccb->n_elem = n_elem;
 		ccb->ccb_tag = tag;
 		ccb->task = t;
+		ccb->device = pm8001_dev;
 		switch (t->task_proto) {
 		case SAS_PROTOCOL_SMP:
 			rc = pm8001_task_prep_smp(pm8001_ha, ccb);
-- 
1.7.9



^ permalink raw reply related	[flat|nested] 3+ messages in thread

end of thread, other threads:[~2014-02-14  9:00 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2014-02-14  8:01 [PATCH] [scsi] : pm80xx : fix problem of pm8001_work_fn reseting uncorrect phy device XinHong Zhu
2014-02-14  9:00 ` lindar_liu
  -- strict thread matches above, loose matches on Subject: below --
2014-01-14  9:41 XinHong Zhu

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.