scsi: Synchronize request queue PM status only on successful resume
Subsystems (3)
PATCH v2 1/1 scsi: Synchronize request queue PM status only on successful resume
The commit 356fd2663cff ("scsi: Set request queue runtime PM status back to active on resume") fixed up the inconsistent RPM status between request queue and device. However changing request queue RPM status shall be done only on successful resume, otherwise status may be still inconsistent as below,
Request queue: RPM_ACTIVE
Device: RPM_SUSPENDED
This ends up soft lockup because requests can be submitted to underlying devices but those devices and their required resource are not resumed.
A message that is not in the archive (sent off-list or not yet received): 1546498395-4184-1-git-send-email-stanley.chu-NuS5LvNUpcJWk0Htik3J/[email protected]
v2 cover scsi: Synchronize request queue PM status only on successful resume
From: Stanley Chu <stanley.chu-NuS5LvNUpcJWk0Htik3J/[email protected]>
Stanley Chu (1):
scsi: Synchronize request queue PM status only on successful resume
drivers/scsi/scsi_pm.c | 26 +++++++++++++++----------- 1 file changed, 15 insertions(+), 11 deletions(-)
Signature · 1 line
2.18.0
PATCH v2 1/1 scsi: Synchronize request queue PM status only on successful resume
In reply to a message not in the archiveFrom: Stanley Chu <stanley.chu-NuS5LvNUpcJWk0Htik3J/[email protected]>
The commit 356fd2663cff ("scsi: Set request queue runtime PM status back to active on resume") fixed up the inconsistent RPM status between request queue and device. However changing request queue RPM status shall be done only on successful resume, otherwise status may be still inconsistent as below,
Request queue: RPM_ACTIVE
Device: RPM_SUSPENDED
This ends up soft lockup because requests can be submitted to underlying devices but those devices and their required resource are not resumed.
back to active on resume")
--- drivers/scsi/scsi_pm.c | 26 +++++++++++++++----------- 1 file changed, 15 insertions(+), 11 deletions(-)
drivers/scsi/scsi_pm.c +15 -11
@@ -80,8 +80,22 @@ static int scsi_dev_type_resume(struct device *dev,if (err ==0 ) {pm_runtime_disable (dev); -pm_runtime_set_active (dev); + err =pm_runtime_set_active (dev);pm_runtime_enable (dev); + +/* +* Forcibly set runtime PM status of request queue to "active" +* to make sure we can again get requests from the queue +* (see also blk_pm_peek_request()). +* +* The resume hook will correct runtime PM status of the disk. +*/ +if (!err &&scsi_is_sdev_device (dev)) { +struct scsi_device *sdev =to_scsi_device (dev); + +if (sdev->request_queue->dev) +blk_set_runtime_active (sdev->request_queue); + } }return err;
@@ -140,16 +154,6 @@ static int scsi_bus_resume_common(struct device *dev,else fn =NULL ; -/* -* Forcibly set runtime PM status of request queue to "active" to -* make sure we can again get requests from the queue (see also -* blk_pm_peek_request()). -* -* The resume hook will correct runtime PM status of the disk. -*/ -if (scsi_is_sdev_device (dev) &&pm_runtime_suspended (dev)) -blk_set_runtime_active (to_scsi_device (dev)->request_queue); -if (fn) {async_schedule_domain (fn, dev, &scsi_sd_pm_domain);
Signature · 1 line
2.18.0
Hi,
Quoted 8 lines · On Thu, Jan 03, 2019 at 02:53:15PM +0800, [email protected] wrote:
From: Stanley Chu <[email protected]>
The commit
356fd2663cff("scsi: Set request queue runtime PM status back to active on resume") fixed up the inconsistent RPM status between request queue and device. However changing request queue RPM status shall be done only on successful resume, otherwise status may be still inconsistent as below,
Quoted 6 lines
Request queue:
RPM_ACTIVE
Device:RPM_SUSPENDEDThis ends up soft lockup because requests can be submitted to underlying devices but those devices and their required resource are not resumed.
It would be good to add some example of the soft lockup you are seeing here.
Quoted 2 lines
Fixes 356fd2663cff ("scsi: Set request queue runtime PM statusback to active on resume")
You don't need to wrap this.
The change itself looks fine.
Hi Mika,
Quoted 14 lines · Mika Westerberg
Hi,
Quoted 11 lines · On Thu, Jan 03, 2019 at 02:53:15PM +0800, [email protected] wrote:
From: Stanley Chu <[email protected]>
The commit
356fd2663cff("scsi: Set request queue runtime PM status back to active on resume") fixed up the inconsistent RPM status between request queue and device. However changing request queue RPM status shall be done only on successful resume, otherwise status may be still inconsistent as below,Request queue:
RPM_ACTIVE
Device:RPM_SUSPENDED
Quoted 6 lines
Quoted 3 lines
This ends up soft lockup because requests can be submitted to underlying devices but those devices and their required resource are not resumed.
It would be good to add some example of the soft lockup you are seeing here.
Thanks for remind, I will add below example in commit message in v3.
For example,
After above inconsistent status happens, IO request can be submitted to UFS device driver but required resource (like clock) is not resumed yet thus lead to warning as below call stack,
WARN_ON(hba->clk_gating.state != CLKS_ON);ufshcd_queuecommandscsi_dispatch_cmdscsi_request_fn__blk_run_queuecfq_insert_request__elv_add_requestblk_flush_plug_listblk_finish_plugjbd2_journal_commit_transaction
kjournald2
We may see all behind IO requests hang because of no response from storage host or device and then soft lockup happens in system. In the end, system may crash in many ways.
Quoted 5 lines
Quoted 3 lines
Fixes 356fd2663cff ("scsi: Set request queue runtime PM statusback to active on resume")
You don't need to wrap this.
OK! Will fix it.
Quoted 2 lines
The change itself looks fine.
Thanks.
Stanley
Quoted 5 lines
_______________________________________________
Linux-mediatek mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-mediatek