From: Bean Huo <[email protected]>
According to the JEDEC UFS 3.1 Spec, If fWriteBoosterBufferFlushDuringHibernate
is set to one, the device flushes the WriteBooster Buffer data automatically
whenever the link enters the hibernate (HIBERN8) state. While the flushing
operation is in progress, the device should be kept in Active power mode.
Currently, we set this flag during the UFSHCD probe stage, but we didn't deal
with its programming failure. Even this failure is less likely to occur, but
still it is possible.
This patch is to add checkup of fWriteBoosterBufferFlushDuringHibernate setting,
keep the device as "active power mode" only when this flag be successfully set
to 1.
Fixes: 51dd905bd2f6 ("scsi: ufs: Fix WriteBooster flush during runtime suspend")
Signed-off-by: Bean Huo <[email protected]>
---
drivers/scsi/ufs/ufs.h | 2 ++
drivers/scsi/ufs/ufshcd.c | 20 +++++++++++++++-----
2 files changed, 17 insertions(+), 5 deletions(-)
diff --git a/drivers/scsi/ufs/ufs.h b/drivers/scsi/ufs/ufs.h
index d593edb48767..311d5f7a024d 100644
--- a/drivers/scsi/ufs/ufs.h
+++ b/drivers/scsi/ufs/ufs.h
@@ -530,6 +530,8 @@ struct ufs_dev_info {
bool f_power_on_wp_en;
/* Keeps information if any of the LU is power on write protected */
bool is_lu_power_on_wp;
+ /* Indicates if flush WB buffer during hibern8 successfully enabled */
+ bool is_hibern8_wb_flush;
/* Maximum number of general LU supported by the UFS device */
u8 max_lu_supported;
u8 wb_dedicated_lu;
diff --git a/drivers/scsi/ufs/ufshcd.c b/drivers/scsi/ufs/ufshcd.c
index 30332592e624..da38d760944b 100644
--- a/drivers/scsi/ufs/ufshcd.c
+++ b/drivers/scsi/ufs/ufshcd.c
@@ -285,10 +285,16 @@ static inline void ufshcd_wb_config(struct ufs_hba *hba)
dev_err(hba->dev, "%s: Enable WB failed: %d\n", __func__, ret);
else
dev_info(hba->dev, "%s: Write Booster Configured\n", __func__);
+
ret = ufshcd_wb_toggle_flush_during_h8(hba, true);
- if (ret)
+ if (ret) {
dev_err(hba->dev, "%s: En WB flush during H8: failed: %d\n",
__func__, ret);
+ hba->dev_info.is_hibern8_wb_flush = false;
+ } else {
+ hba->dev_info.is_hibern8_wb_flush = true;
+ }
+
ufshcd_wb_toggle_flush(hba, true);
}
@@ -5448,6 +5454,7 @@ static bool ufshcd_wb_need_flush(struct ufs_hba *hba)
if (!ufshcd_is_wb_allowed(hba))
return false;
+
/*
* The ufs device needs the vcc to be ON to flush.
* With user-space reduction enabled, it's enough to enable flush
@@ -8540,6 +8547,7 @@ static int ufshcd_suspend(struct ufs_hba *hba, enum ufs_pm_op pm_op)
enum ufs_pm_level pm_lvl;
enum ufs_dev_pwr_mode req_dev_pwr_mode;
enum uic_link_state req_link_state;
+ bool hibern8;
hba->pm_op_in_progress = 1;
if (!ufshcd_is_shutdown_pm(pm_op)) {
@@ -8599,11 +8607,13 @@ static int ufshcd_suspend(struct ufs_hba *hba, enum ufs_pm_op pm_op)
* Hibern8, keep device power mode as "active power mode"
* and VCC supply.
*/
+ hibern8 = req_link_state == UIC_LINK_HIBERN8_STATE ||
+ (req_link_state == UIC_LINK_ACTIVE_STATE &&
+ ufshcd_is_auto_hibern8_enabled(hba));
+
hba->dev_info.b_rpm_dev_flush_capable =
- hba->auto_bkops_enabled ||
- (((req_link_state == UIC_LINK_HIBERN8_STATE) ||
- ((req_link_state == UIC_LINK_ACTIVE_STATE) &&
- ufshcd_is_auto_hibern8_enabled(hba))) &&
+ hba->auto_bkops_enabled || (hibern8 &&
+ hba->dev_info.is_hibern8_wb_flush &&
ufshcd_wb_need_flush(hba));
}
--
2.17.1
On Sun, 2020-12-06 at 11:13 +0100, Bean Huo wrote:
> From: Bean Huo <[email protected]>
>
> According to the JEDEC UFS 3.1 Spec, If fWriteBoosterBufferFlushDuringHibernate
> is set to one, the device flushes the WriteBooster Buffer data automatically
> whenever the link enters the hibernate (HIBERN8) state. While the flushing
> operation is in progress, the device should be kept in Active power mode.
> Currently, we set this flag during the UFSHCD probe stage, but we didn't deal
> with its programming failure. Even this failure is less likely to occur, but
> still it is possible.
> This patch is to add checkup of fWriteBoosterBufferFlushDuringHibernate setting,
> keep the device as "active power mode" only when this flag be successfully set
> to 1.
>
> Fixes: 51dd905bd2f6 ("scsi: ufs: Fix WriteBooster flush during runtime suspend")
> Signed-off-by: Bean Huo <[email protected]>
> ---
> drivers/scsi/ufs/ufs.h | 2 ++
> drivers/scsi/ufs/ufshcd.c | 20 +++++++++++++++-----
> 2 files changed, 17 insertions(+), 5 deletions(-)
>
> diff --git a/drivers/scsi/ufs/ufs.h b/drivers/scsi/ufs/ufs.h
> index d593edb48767..311d5f7a024d 100644
> --- a/drivers/scsi/ufs/ufs.h
> +++ b/drivers/scsi/ufs/ufs.h
> @@ -530,6 +530,8 @@ struct ufs_dev_info {
> bool f_power_on_wp_en;
> /* Keeps information if any of the LU is power on write protected */
> bool is_lu_power_on_wp;
> + /* Indicates if flush WB buffer during hibern8 successfully enabled */
> + bool is_hibern8_wb_flush;
Perhaps a more comprehensive name?
For example, wb_flush_during_hibern8?
> /* Maximum number of general LU supported by the UFS device */
> u8 max_lu_supported;
> u8 wb_dedicated_lu;
> diff --git a/drivers/scsi/ufs/ufshcd.c b/drivers/scsi/ufs/ufshcd.c
> index 30332592e624..da38d760944b 100644
> --- a/drivers/scsi/ufs/ufshcd.c
> +++ b/drivers/scsi/ufs/ufshcd.c
> @@ -285,10 +285,16 @@ static inline void ufshcd_wb_config(struct ufs_hba *hba)
> dev_err(hba->dev, "%s: Enable WB failed: %d\n", __func__, ret);
> else
> dev_info(hba->dev, "%s: Write Booster Configured\n", __func__);
> +
> ret = ufshcd_wb_toggle_flush_during_h8(hba, true);
> - if (ret)
> + if (ret) {
> dev_err(hba->dev, "%s: En WB flush during H8: failed: %d\n",
> __func__, ret);
> + hba->dev_info.is_hibern8_wb_flush = false;
Perhaps this statement could be dummy because
hba->dev_info.is_hibern8_wb_flush is zero-initialized and
ufshcd_wb_config() is invoked only once during ufs initialization.
Thanks,
Stanley Chu
> + } else {
> + hba->dev_info.is_hibern8_wb_flush = true;
> + }
> +
> ufshcd_wb_toggle_flush(hba, true);
> }
>
> @@ -5448,6 +5454,7 @@ static bool ufshcd_wb_need_flush(struct ufs_hba *hba)
>
> if (!ufshcd_is_wb_allowed(hba))
> return false;
> +
> /*
> * The ufs device needs the vcc to be ON to flush.
> * With user-space reduction enabled, it's enough to enable flush
> @@ -8540,6 +8547,7 @@ static int ufshcd_suspend(struct ufs_hba *hba, enum ufs_pm_op pm_op)
> enum ufs_pm_level pm_lvl;
> enum ufs_dev_pwr_mode req_dev_pwr_mode;
> enum uic_link_state req_link_state;
> + bool hibern8;
>
> hba->pm_op_in_progress = 1;
> if (!ufshcd_is_shutdown_pm(pm_op)) {
> @@ -8599,11 +8607,13 @@ static int ufshcd_suspend(struct ufs_hba *hba, enum ufs_pm_op pm_op)
> * Hibern8, keep device power mode as "active power mode"
> * and VCC supply.
> */
> + hibern8 = req_link_state == UIC_LINK_HIBERN8_STATE ||
> + (req_link_state == UIC_LINK_ACTIVE_STATE &&
> + ufshcd_is_auto_hibern8_enabled(hba));
> +
> hba->dev_info.b_rpm_dev_flush_capable =
> - hba->auto_bkops_enabled ||
> - (((req_link_state == UIC_LINK_HIBERN8_STATE) ||
> - ((req_link_state == UIC_LINK_ACTIVE_STATE) &&
> - ufshcd_is_auto_hibern8_enabled(hba))) &&
> + hba->auto_bkops_enabled || (hibern8 &&
> + hba->dev_info.is_hibern8_wb_flush &&
> ufshcd_wb_need_flush(hba));
> }
>
>
> According to the JEDEC UFS 3.1 Spec, If
> fWriteBoosterBufferFlushDuringHibernate
> is set to one, the device flushes the WriteBooster Buffer data automatically
> whenever the link enters the hibernate (HIBERN8) state. While the flushing
> operation is in progress, the device should be kept in Active power mode.
> Currently, we set this flag during the UFSHCD probe stage, but we didn't deal
> with its programming failure. Even this failure is less likely to occur, but
> still it is possible.
> This patch is to add checkup of fWriteBoosterBufferFlushDuringHibernate
> setting,
> keep the device as "active power mode" only when this flag be successfully
> set
> to 1.
>
> Fixes: 51dd905bd2f6 ("scsi: ufs: Fix WriteBooster flush during runtime
> suspend")
> Signed-off-by: Bean Huo <[email protected]>
You've added the fixes tag, but my previous comment is still unanswered:
you are adding protection to a single device management command.
Why this command in particular?
What makes it so special that it needs this extra care?
Thanks,
Avri
On Mon, 2020-12-07 at 08:02 +0000, Avri Altman wrote:
> > According to the JEDEC UFS 3.1 Spec, If
> > fWriteBoosterBufferFlushDuringHibernate
> > is set to one, the device flushes the WriteBooster Buffer data
> > automatically
> > whenever the link enters the hibernate (HIBERN8) state. While the
> > flushing
> > operation is in progress, the device should be kept in Active power
> > mode.
> > Currently, we set this flag during the UFSHCD probe stage, but we
> > didn't deal
> > with its programming failure. Even this failure is less likely to
> > occur, but
> > still it is possible.
> > This patch is to add checkup of
> > fWriteBoosterBufferFlushDuringHibernate
> > setting,
> > keep the device as "active power mode" only when this flag be
> > successfully
> > set
> > to 1.
> >
> > Fixes: 51dd905bd2f6 ("scsi: ufs: Fix WriteBooster flush during
> > runtime
> > suspend")
> > Signed-off-by: Bean Huo <[email protected]>
>
> You've added the fixes tag,
yes,it is a bug.
> but my previous comment is still unanswered:
> you are adding protection to a single device management command.
> Why this command in particular?
> What makes it so special that it needs this extra care?
>
see the Spec:
"
If fWriteBoosterBufferFlushDuringHibernate is set to one, the device
flushes the WriteBooster Buffer data automatically whenever the link
enters the hibernate (HIBERN8) state.
The device shall stop the flushing operation if
fWriteBoosterBufferFlushDuringHibernate are set to zero.
....
"
If fWriteBoosterBufferFlushDuringHibernate ==0, device will not flush
WB, even if you keep device as "active mode" and LINK in hibernate
state.
Bean
Thanks,
On Mon, 2020-12-07 at 13:36 +0800, Stanley Chu wrote:
> On Sun, 2020-12-06 at 11:13 +0100, Bean Huo wrote:
> > From: Bean Huo <[email protected]>
> >
> > According to the JEDEC UFS 3.1 Spec, If fWriteBoosterBufferFlushDuringHibernate
> > is set to one, the device flushes the WriteBooster Buffer data automatically
> > whenever the link enters the hibernate (HIBERN8) state. While the flushing
> > operation is in progress, the device should be kept in Active power mode.
> > Currently, we set this flag during the UFSHCD probe stage, but we didn't deal
> > with its programming failure. Even this failure is less likely to occur, but
> > still it is possible.
> > This patch is to add checkup of fWriteBoosterBufferFlushDuringHibernate setting,
> > keep the device as "active power mode" only when this flag be successfully set
> > to 1.
> >
> > Fixes: 51dd905bd2f6 ("scsi: ufs: Fix WriteBooster flush during runtime suspend")
> > Signed-off-by: Bean Huo <[email protected]>
> > ---
> > drivers/scsi/ufs/ufs.h | 2 ++
> > drivers/scsi/ufs/ufshcd.c | 20 +++++++++++++++-----
> > 2 files changed, 17 insertions(+), 5 deletions(-)
> >
> > diff --git a/drivers/scsi/ufs/ufs.h b/drivers/scsi/ufs/ufs.h
> > index d593edb48767..311d5f7a024d 100644
> > --- a/drivers/scsi/ufs/ufs.h
> > +++ b/drivers/scsi/ufs/ufs.h
> > @@ -530,6 +530,8 @@ struct ufs_dev_info {
> > bool f_power_on_wp_en;
> > /* Keeps information if any of the LU is power on write protected */
> > bool is_lu_power_on_wp;
> > + /* Indicates if flush WB buffer during hibern8 successfully enabled */
> > + bool is_hibern8_wb_flush;
>
> Perhaps a more comprehensive name?
> For example, wb_flush_during_hibern8?
> > /* Maximum number of general LU supported by the UFS device */
> > u8 max_lu_supported;
> > u8 wb_dedicated_lu;
> > diff --git a/drivers/scsi/ufs/ufshcd.c b/drivers/scsi/ufs/ufshcd.c
> > index 30332592e624..da38d760944b 100644
> > --- a/drivers/scsi/ufs/ufshcd.c
> > +++ b/drivers/scsi/ufs/ufshcd.c
> > @@ -285,10 +285,16 @@ static inline void ufshcd_wb_config(struct ufs_hba *hba)
> > dev_err(hba->dev, "%s: Enable WB failed: %d\n", __func__, ret);
> > else
> > dev_info(hba->dev, "%s: Write Booster Configured\n", __func__);
> > +
> > ret = ufshcd_wb_toggle_flush_during_h8(hba, true);
> > - if (ret)
> > + if (ret) {
> > dev_err(hba->dev, "%s: En WB flush during H8: failed: %d\n",
> > __func__, ret);
> > + hba->dev_info.is_hibern8_wb_flush = false;
>
> Perhaps this statement could be dummy because
> hba->dev_info.is_hibern8_wb_flush is zero-initialized and
> ufshcd_wb_config() is invoked only once during ufs initialization.
Hi Bean,
Sorry to mislead you. ufshcd_wb_config() may be called multiple times by
both initialization and error recovery. Please ignore above suggestion.
Thanks,
Stanley Chu
>
> Thanks,
> Stanley Chu
>
> > + } else {
> > + hba->dev_info.is_hibern8_wb_flush = true;
> > + }
> > +
> > ufshcd_wb_toggle_flush(hba, true);
> > }
> >
> > @@ -5448,6 +5454,7 @@ static bool ufshcd_wb_need_flush(struct ufs_hba *hba)
> >
> > if (!ufshcd_is_wb_allowed(hba))
> > return false;
> > +
> > /*
> > * The ufs device needs the vcc to be ON to flush.
> > * With user-space reduction enabled, it's enough to enable flush
> > @@ -8540,6 +8547,7 @@ static int ufshcd_suspend(struct ufs_hba *hba, enum ufs_pm_op pm_op)
> > enum ufs_pm_level pm_lvl;
> > enum ufs_dev_pwr_mode req_dev_pwr_mode;
> > enum uic_link_state req_link_state;
> > + bool hibern8;
> >
> > hba->pm_op_in_progress = 1;
> > if (!ufshcd_is_shutdown_pm(pm_op)) {
> > @@ -8599,11 +8607,13 @@ static int ufshcd_suspend(struct ufs_hba *hba, enum ufs_pm_op pm_op)
> > * Hibern8, keep device power mode as "active power mode"
> > * and VCC supply.
> > */
> > + hibern8 = req_link_state == UIC_LINK_HIBERN8_STATE ||
> > + (req_link_state == UIC_LINK_ACTIVE_STATE &&
> > + ufshcd_is_auto_hibern8_enabled(hba));
> > +
> > hba->dev_info.b_rpm_dev_flush_capable =
> > - hba->auto_bkops_enabled ||
> > - (((req_link_state == UIC_LINK_HIBERN8_STATE) ||
> > - ((req_link_state == UIC_LINK_ACTIVE_STATE) &&
> > - ufshcd_is_auto_hibern8_enabled(hba))) &&
> > + hba->auto_bkops_enabled || (hibern8 &&
> > + hba->dev_info.is_hibern8_wb_flush &&
> > ufshcd_wb_need_flush(hba));
> > }
> >
>
> On Mon, 2020-12-07 at 08:02 +0000, Avri Altman wrote:
> > > According to the JEDEC UFS 3.1 Spec, If
> > > fWriteBoosterBufferFlushDuringHibernate
> > > is set to one, the device flushes the WriteBooster Buffer data
> > > automatically
> > > whenever the link enters the hibernate (HIBERN8) state. While the
> > > flushing
> > > operation is in progress, the device should be kept in Active power
> > > mode.
> > > Currently, we set this flag during the UFSHCD probe stage, but we
> > > didn't deal
> > > with its programming failure. Even this failure is less likely to
> > > occur, but
> > > still it is possible.
> > > This patch is to add checkup of
> > > fWriteBoosterBufferFlushDuringHibernate
> > > setting,
> > > keep the device as "active power mode" only when this flag be
> > > successfully
> > > set
> > > to 1.
> > >
> > > Fixes: 51dd905bd2f6 ("scsi: ufs: Fix WriteBooster flush during
> > > runtime
> > > suspend")
> > > Signed-off-by: Bean Huo <[email protected]>
> >
> > You've added the fixes tag,
>
> yes,it is a bug.
>
> > but my previous comment is still unanswered:
> > you are adding protection to a single device management command.
> > Why this command in particular?
> > What makes it so special that it needs this extra care?
> >
>
> see the Spec:
> "
> If fWriteBoosterBufferFlushDuringHibernate is set to one, the device
> flushes the WriteBooster Buffer data automatically whenever the link
> enters the hibernate (HIBERN8) state.
>
> The device shall stop the flushing operation if
> fWriteBoosterBufferFlushDuringHibernate are set to zero.
> ....
>
> "
> If fWriteBoosterBufferFlushDuringHibernate ==0, device will not flush
> WB, even if you keep device as "active mode" and LINK in hibernate
> state.
OK, so what you are actually saying, is that since we are only toggling this flag once per
link startup / recovery, in case of a failure, however rare - the host may be still keep vcc on
for nothing, as the device will do nothing in that extra wake time. Right?
But every time ufshcd_wb_need_flush() is called we are reading some other flags/attributes?
What about them? Why not protecting them as well?
Sorry - the whole idea doesn't make any sense to me.
Thanks,
Avri
>
> Bean
> Thanks,
>
On Tue, 2020-12-08 at 09:05 +0000, Avri Altman wrote:
> > "
> > If fWriteBoosterBufferFlushDuringHibernate ==0, device will not
> > flush
> > WB, even if you keep device as "active mode" and LINK in hibernate
> > state.
>
> OK, so what you are actually saying, is that since we are only
> toggling this flag once per
> link startup / recovery, in case of a failure, however rare - the
> host may be still keep vcc on
> for nothing, as the device will do nothing in that extra wake time.
> Right?
>
again, this is a BUG, BUG...
Bug is a Bug, doesn't matter it is rare or not rare.
Tell me why we retry three times in ufshcd_query_flag_retry() in case
of failure?
> But every time ufshcd_wb_need_flush() is called we are reading some
> other flags/attributes?
> What about them? Why not protecting them as well?
>
did you read ufshcd_wb_need_flush() fully? they have been properly
protected.
> Sorry - the whole idea doesn't make any sense to me.
>
Thanks very much for sharing your opinion, I am very happy to hear your
opinion. Because you don't believe this is a bug, I will ask other UFS
guys to review, let Martin make the last decision.
Thanks agaion for your review.
Bean