scsi: ufs: Let devices remain runtime suspended during system suspend
v6 cover scsi: ufs: Let devices remain runtime suspended during system suspend
Hi
UFS devices can remain runtime suspended at system suspend time, if the conditions are right. Add support for that, first fixing the impediments.
Changes in V6:
scsi: ufs: Fix error handler clear ua deadlock
Ensure data byte count bits 1:0 are 11b Use ufshcd_compose_dev_cmd() to set up command
Changes in V5:
scsi: ufs: Fix error handler clear ua deadlock
Update commit message
Try to abort REQUEST SENSE if it times out
Changes in V4:
scsi: ufs: Fix error handler clear ua deadlock
Do request-sense directly
Changes in V3:
scsi: ufs: Fix error handler clear ua deadlock
Correct commit message.
Amend stable tags to add dependent cherry picks
Changes in V2:
scsi: ufs: Let devices remain runtime suspended during system suspend
The ufs-hisi driver uses different RPM and SPM, but it is made explicit by a new parameter to suspend prepare.
Adrian Hunter (3):
scsi: ufs: Fix error handler clear ua deadlock
scsi: ufs: Fix runtime PM dependencies getting broken
scsi: ufs: Let devices remain runtime suspended during system suspend
drivers/scsi/scsi_pm.c | 16 ++- drivers/scsi/ufs/ufs-hisi.c | 8 +- drivers/scsi/ufs/ufs.h | 3 +- drivers/scsi/ufs/ufshcd.c | 296 ++++++++++++++++++++++++++++++-------------- drivers/scsi/ufs/ufshcd.h | 24 +++- include/scsi/scsi_device.h | 1 + include/trace/events/ufs.h | 5 +- 7 files changed, 251 insertions(+), 102 deletions(-)
Regards
Adrian
PATCH v6 1/3 scsi: ufs: Fix error handler clear ua deadlock
There is no guarantee to be able to enter the queue if requests are blocked. That is because freezing the queue will block entry to the queue, but freezing also waits for outstanding requests which can make no progress while the queue is blocked.
That situation can happen when the error handler issues requests to clear unit attention condition. Requests are blocked if the ufshcd_state is UFSHCD_STATE_EH_SCHEDULED_FATAL, which can happen as a result of prior error handler activity. Requests cannot make progress when ufshcd_state is UFSHCD_STATE_EH_SCHEDULED_FATAL, and only the error handler can change that, so if the error handler is waiting to enter the queue and blk_mq_freeze_queue() is waiting for outstanding requests, they will deadlock.
Fix by issuing REQUEST_SENSE directly avoiding the SCSI queues, in a similar fashion as other device commands.
--- drivers/scsi/ufs/ufs.h | 3 +- drivers/scsi/ufs/ufshcd.c | 242 +++++++++++++++++++++++++------------ drivers/scsi/ufs/ufshcd.h | 12 ++ include/trace/events/ufs.h | 5 +- 4 files changed, 180 insertions(+), 82 deletions(-)
drivers/scsi/ufs/ufs.h +2 -1
@@ -612,7 +612,8 @@ struct ufs_dev_info {enum ufs_trace_str_t {UFS_CMD_SEND ,UFS_CMD_COMP ,UFS_DEV_COMP ,UFS_QUERY_SEND ,UFS_QUERY_COMP ,UFS_QUERY_ERR , -UFS_TM_SEND ,UFS_TM_COMP ,UFS_TM_ERR +UFS_TM_SEND ,UFS_TM_COMP ,UFS_TM_ERR , +UFS_SENSE_SEND ,UFS_SENSE_COMP ,UFS_SENSE_ERR , };/*
drivers/scsi/ufs/ufshcd.c +162 -80
386 lines not on this page · raw message
drivers/scsi/ufs/ufshcd.h +12 -0
@@ -56,6 +56,8 @@ struct ufs_hba;enum dev_cmd_type {DEV_CMD_TYPE_NOP =0x0 ,DEV_CMD_TYPE_QUERY =0x1 , +DEV_CMD_TYPE_SENSE =0x2 , +DEV_CMD_TYPE_INVALID =0xff , };enum ufs_event_type {
@@ -236,17 +238,27 @@ struct ufs_query {struct ufs_query_res response; }; +/* +* UFS_DEV_CMD_BUF_SZ must be a power of 2 and bigger than UFS_SENSE_SIZE +* rounded up to a multiple of 4. +*/ +#define UFS_DEV_CMD_BUF_SZ 32 +/** * struct ufs_dev_cmd - all assosiated fields with device management commands * @type: device management command type - Query, NOP OUT * @lock: lock to allow one command at a time * @complete: internal commands completion +* @buf: buffer for data +* @buf_dma_addr: DMA address of @buf */ struct ufs_dev_cmd {enum dev_cmd_type type;struct mutex lock;struct completion *complete;struct ufs_query query; +void *buf; +dma_addr_t buf_dma_addr; };/**
include/trace/events/ufs.h +4 -1
@@ -46,7 +46,10 @@ EM(UFS_QUERY_ERR, "query_complete_err") \ EM(UFS_TM_SEND, "tm_send") \ EM(UFS_TM_COMP, "tm_complete") \ - EMe(UFS_TM_ERR, "tm_complete_err") + EM(UFS_TM_ERR, "tm_complete_err") \ + EM(UFS_SENSE_SEND, "sense_send") \ + EM(UFS_SENSE_COMP, "sense_complete") \ + EMe(UFS_SENSE_ERR, "sense_complete_err")#define UFS_CMD_TRACE_TSF_TYPES \ EM(UFS_TSF_CDB, "CDB") \
Signature · 1 line
2.25.1
Re: [PATCH V6 1/3] scsi: ufs: Fix error handler clear ua deadlock
Quoted 14 lines · Adrian Hunter
There is no guarantee to be able to enter the queue if requests are blocked. That is because freezing the queue will block entry to the queue, but freezing also waits for outstanding requests which can make no progress while the queue is blocked.
That situation can happen when the error handler issues requests to clear unit attention condition. Requests are blocked if the
ufshcd_stateisUFSHCD_STATE_EH_SCHEDULED_FATAL, which can happen as a result of prior error handler activity. Requests cannot make progress whenufshcd_stateisUFSHCD_STATE_EH_SCHEDULED_FATAL, and only the error handler can change that, so if the error handler is waiting to enter the queue andblk_mq_freeze_queue()is waiting for outstanding requests, they will deadlock.
Quoted 2 lines
Fix by issuing
REQUEST_SENSEdirectly avoiding the SCSI queues, in a similar fashion as other device commands.
Hi Adrian,
Although I appreciate all the work that you have done on this patch, there is at least one application (the Android Trusty software) that needs the RPMB unit attention information to work correctly. Hence my request to drop this patch and to integrate the following patch series in the upstream kernel:
https://lore.kernel.org/linux-scsi/[email protected]/T/#t
Thanks,
Bart.
Re: [PATCH V6 1/3] scsi: ufs: Fix error handler clear ua deadlock
Quoted 18 lines · Bart Van Assche
Quoted 14 lines · Adrian Hunter
There is no guarantee to be able to enter the queue if requests are blocked. That is because freezing the queue will block entry to the queue, but freezing also waits for outstanding requests which can make no progress while the queue is blocked.
That situation can happen when the error handler issues requests to clear unit attention condition. Requests are blocked if the
ufshcd_stateisUFSHCD_STATE_EH_SCHEDULED_FATAL, which can happen as a result of prior error handler activity. Requests cannot make progress whenufshcd_stateisUFSHCD_STATE_EH_SCHEDULED_FATAL, and only the error handler can change that, so if the error handler is waiting to enter the queue andblk_mq_freeze_queue()is waiting for outstanding requests, they will deadlock.Quoted 2 lines
Fix by issuing
REQUEST_SENSEdirectly avoiding the SCSI queues, in a similar fashion as other device commands.
Quoted 7 lines
Hi Adrian,
Although I appreciate all the work that you have done on this patch, there is at least one application (the Android Trusty software) that needs the RPMB unit attention information to work correctly. Hence my request to drop this patch and to integrate the following patch series in the upstream kernel:
https://lore.kernel.org/linux-scsi/[email protected]/T/#t
That's fine. It is always good to have options.
I made a couple of minor comments on the patch series.
PATCH v6 2/3 scsi: ufs: Fix runtime PM dependencies getting broken
In reply to Adrian HunterUFS SCSI devices make use of device links to establish PM dependencies. However, SCSI PM will force devices' runtime PM state to be active during system resume. That can break runtime PM dependencies for UFS devices. Fix by adding a flag 'preserve_rpm' to let UFS SCSI devices opt-out of the unwanted behaviour.
--- drivers/scsi/scsi_pm.c | 16 +++++++++++----- drivers/scsi/ufs/ufshcd.c | 1 + include/scsi/scsi_device.h | 1 + 3 files changed, 13 insertions(+), 5 deletions(-)
drivers/scsi/scsi_pm.c +11 -5
@@ -73,13 +73,22 @@ static int scsi_dev_type_resume(struct device *dev, int (*cb)(struct device *, const struct dev_pm_ops *)) {const struct dev_pm_ops *pm = dev->driver ? dev->driver->pm :NULL ; +struct scsi_device *sdev =NULL ; +bool preserve_rpm = false;int err =0 ; +if (scsi_is_sdev_device (dev)) { + sdev =to_scsi_device (dev); + preserve_rpm = sdev->preserve_rpm; +if (preserve_rpm &&pm_runtime_suspended (dev)) +return 0 ; + } + err =cb (dev, pm);scsi_device_resume (to_scsi_device (dev));dev_dbg (dev,"scsi resume: %d\n" , err); -if (err ==0 ) { +if (err ==0 && !preserve_rpm) {pm_runtime_disable (dev); err =pm_runtime_set_active (dev);pm_runtime_enable (dev);
@@ -91,11 +100,8 @@ static int scsi_dev_type_resume(struct device *dev,* * 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 (!err && sdev)blk_set_runtime_active (sdev->request_queue); - } }return err;
drivers/scsi/ufs/ufshcd.c +1 -0
@@ -5069,6 +5069,7 @@ static int ufshcd_slave_configure(struct scsi_device *sdev)pm_runtime_get_noresume (&sdev->sdev_gendev); else if (ufshcd_is_rpm_autosuspend_allowed(hba)) sdev->rpm_autosuspend = 1; + sdev->preserve_rpm =1 ;ufshcd_crypto_setup_rq_keyslot_manager (hba, q);
include/scsi/scsi_device.h +1 -0
@@ -196,6 +196,7 @@ struct scsi_device { unsigned no_read_disc_info:1; /* Avoid READ_DISC_INFO cmds */ unsigned no_read_capacity_16:1; /* Avoid READ_CAPACITY_16 cmds */ unsigned try_rc_10_first:1; /* Try READ_CAPACACITY_10 first */ + unsigned preserve_rpm:1; /* Preserve runtime PM */ unsigned security_supported:1; /* Supports Security Protocols */ unsigned is_visible:1; /* is the device visible in sysfs */ unsigned wce_default_on:1; /* Cache is ON by default */
Signature · 1 line
2.25.1
PATCH v6 3/3 scsi: ufs: Let devices remain runtime suspended during system suspend
In reply to Adrian HunterIf the UFS Device WLUN is runtime suspended and is in the same power mode, link state and b_rpm_dev_flush_capable (BKOP or WB buffer flush etc) state, then it can remain runtime suspended instead of being runtime resumed and then system suspended.
The following patches have cleared the way for that to happen:
scsi: ufs: Fix runtime PM dependencies getting broken
scsi: ufs: Fix error handler clear ua deadlock
So amend the logic accordingly.
Note, the ufs-hisi driver uses different RPM and SPM, but it is made explicit by a new parameter to suspend prepare.
--- drivers/scsi/ufs/ufs-hisi.c | 8 +++++- drivers/scsi/ufs/ufshcd.c | 53 ++++++++++++++++++++++++++++--------- drivers/scsi/ufs/ufshcd.h | 12 ++++++++- 3 files changed, 58 insertions(+), 15 deletions(-)
drivers/scsi/ufs/ufs-hisi.c +7 -1
@@ -396,6 +396,12 @@ static int ufs_hisi_pwr_change_notify(struct ufs_hba *hba,return ret; } +static int ufs_hisi_suspend_prepare (struct device *dev) +{ +/* RPM and SPM are different. Refer ufs_hisi_suspend() */ +return __ufshcd_suspend_prepare (dev, false); +} +static int ufs_hisi_suspend (struct ufs_hba *hba,enum ufs_pm_op pm_op) {struct ufs_hisi_host *host =ufshcd_get_variant (hba);
@@ -574,7 +580,7 @@ static int ufs_hisi_remove(struct platform_device *pdev)static const struct dev_pm_ops ufs_hisi_pm_ops = { SET_SYSTEM_SLEEP_PM_OPS(ufshcd_system_suspend, ufshcd_system_resume)SET_RUNTIME_PM_OPS (ufshcd_runtime_suspend, ufshcd_runtime_resume,NULL ) - .prepare = ufshcd_suspend_prepare, + .prepare = ufs_hisi_suspend_prepare, .complete = ufshcd_resume_complete, };
drivers/scsi/ufs/ufshcd.c +40 -13
73 lines not on this page · raw message
drivers/scsi/ufs/ufshcd.h +11 -1
@@ -931,7 +931,6 @@ struct ufs_hba {#endif u32 luns_avail;bool complete_put; -bool rpmb_complete_put; };/* Returns true if clocks can be gated. Otherwise false */
@@ -1191,6 +1190,7 @@ int ufshcd_exec_raw_upiu_cmd(struct ufs_hba *hba,int ufshcd_wb_toggle (struct ufs_hba *hba,bool enable);int ufshcd_suspend_prepare (struct device *dev); +int __ufshcd_suspend_prepare (struct device *dev,bool rpm_ok_for_spm);void ufshcd_resume_complete (struct device *dev);/* Wrapper functions for safely calling variant operations */
@@ -1399,6 +1399,16 @@ static inline int ufshcd_rpm_put_sync(struct ufs_hba *hba)return pm_runtime_put_sync (&hba->sdev_ufs_device->sdev_gendev); } +static inline void ufshcd_rpm_get_noresume (struct ufs_hba *hba) +{ +pm_runtime_get_noresume (&hba->sdev_ufs_device->sdev_gendev); +} + +static inline int ufshcd_rpm_resume (struct ufs_hba *hba) +{ +return pm_runtime_resume (&hba->sdev_ufs_device->sdev_gendev); +} +static inline int ufshcd_rpm_put (struct ufs_hba *hba) {return pm_runtime_put (&hba->sdev_ufs_device->sdev_gendev);
Signature · 1 line
2.25.1