drivers/i2c/busses/i2c-designware-common.c | 14 +++++++++ drivers/i2c/busses/i2c-designware-core.h | 1 + drivers/i2c/busses/i2c-designware-master.c | 35 ++++++++++++++++++++++ 3 files changed, 50 insertions(+)
It was observed that issuing ABORT bit (IC_ENABLE[1]) will not
work when IC_ENABLE is already disabled.
Check if ENABLE bit (IC_ENABLE[0]) is disabled when the controller
is holding SCL low. If ENABLE bit is disabled, the software need
to enable it before trying to issue ABORT bit. otherwise,
the controller ignores any write to ABORT bit.
These kernel logs show up whenever an I2C transaction is
attempted after this failure.
i2c_designware e95e0000.i2c: timeout waiting for bus ready
i2c_designware e95e0000.i2c: timeout in disabling adapter
The patch can be fix the controller cannot be disabled while
SCL is held low in ENABLE bit is already disabled.
Fixes: 2409205acd3c ("i2c: designware: fix __i2c_dw_disable() in case master is holding SCL low")
Signed-off-by: Kimriver Liu <kimriver.liu@siengine.com>
Reviewed-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Acked-by: Jarkko Nikula <jarkko.nikula@linux.intel.com>
---
V9->V10:
1. return a boolean value for i2c_dw_is_controller_active()
2. update comments and indentation alignment
V8->V9:
1. update commit messages and comments
2. add Acked-by: Jarkko Nikula <jarkko.nikula@linux.intel.com>
3. add fixes tag
4. change function i2c_dw_is_master_idling to i2c_dw_is_controller_active
5. update subject as: fix controller is holding SCL low while ENABLE bit is disabled
V7->V8:
1. calculate this delay based on the actual speed in use
fsleep(DIV_ROUND_CLOSEST_ULL(10 * MICRO, t->bus_freq_hz))
2. add Reviewed-by: Mika Westerberg<mika.westerberg@linux.intel.com>
V6->V7:
1. add Subject versioning [PATCH v7]
2. change fsleep(25) to usleep_range(25, 250)
3. Add macro definition DW_iC_ENABLE_ENABLE to fix compile errors
| Reported-by: kernel test robot <lkp@intel.com>
| Closes:https://lore.kernel.org/oe-kbuild-all/202409082011.9JF6aYsk-lkp@intel.com/
4. base: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?h=master
V5->V6: restore i2c_dw_is_master_idling() function checking
V4->V5: delete master idling checking
V3->V4:
1. update commit messages and add patch version and changelog
2. move print the error message in i2c_dw_xfer
V2->V3: change (!enable) to (!(enable & DW_IC_ENABLE_ENABLE))
V1->V2: used standard words in function names and addressed review comments
link to V1:
https://lore.kernel.org/lkml/20240904064224.2394-1-kimriver.liu@siengine.com/
---
drivers/i2c/busses/i2c-designware-common.c | 14 +++++++++
drivers/i2c/busses/i2c-designware-core.h | 1 +
drivers/i2c/busses/i2c-designware-master.c | 35 ++++++++++++++++++++++
3 files changed, 50 insertions(+)
diff --git a/drivers/i2c/busses/i2c-designware-common.c b/drivers/i2c/busses/i2c-designware-common.c
index e8a688d04aee..48b8d1c1080c 100644
--- a/drivers/i2c/busses/i2c-designware-common.c
+++ b/drivers/i2c/busses/i2c-designware-common.c
@@ -441,6 +441,7 @@ int i2c_dw_set_sda_hold(struct dw_i2c_dev *dev)
void __i2c_dw_disable(struct dw_i2c_dev *dev)
{
+ struct i2c_timings *t = &dev->timings;
unsigned int raw_intr_stats;
unsigned int enable;
int timeout = 100;
@@ -453,6 +454,19 @@ void __i2c_dw_disable(struct dw_i2c_dev *dev)
abort_needed = raw_intr_stats & DW_IC_INTR_MST_ON_HOLD;
if (abort_needed) {
+ if (!(enable & DW_IC_ENABLE_ENABLE)) {
+ regmap_write(dev->map, DW_IC_ENABLE, DW_IC_ENABLE_ENABLE);
+ /*
+ * Wait 10 times the signaling period of the highest I2C
+ * transfer supported by the driver (for 400KHz this is
+ * 25us) to ensure the I2C ENABLE bit is already set
+ * as described in the DesignWare I2C databook.
+ */
+ fsleep(DIV_ROUND_CLOSEST_ULL(10 * MICRO, t->bus_freq_hz));
+ /* Keep ENABLE bit is already set before Setting ABORT.*/
+ enable |= DW_IC_ENABLE_ENABLE;
+ }
+
regmap_write(dev->map, DW_IC_ENABLE, enable | DW_IC_ENABLE_ABORT);
ret = regmap_read_poll_timeout(dev->map, DW_IC_ENABLE, enable,
!(enable & DW_IC_ENABLE_ABORT), 10,
diff --git a/drivers/i2c/busses/i2c-designware-core.h b/drivers/i2c/busses/i2c-designware-core.h
index e9606c00b8d1..e45daedad967 100644
--- a/drivers/i2c/busses/i2c-designware-core.h
+++ b/drivers/i2c/busses/i2c-designware-core.h
@@ -109,6 +109,7 @@
DW_IC_INTR_RX_UNDER | \
DW_IC_INTR_RD_REQ)
+#define DW_IC_ENABLE_ENABLE BIT(0)
#define DW_IC_ENABLE_ABORT BIT(1)
#define DW_IC_STATUS_ACTIVITY BIT(0)
diff --git a/drivers/i2c/busses/i2c-designware-master.c b/drivers/i2c/busses/i2c-designware-master.c
index c7e56002809a..0cc5833214d5 100644
--- a/drivers/i2c/busses/i2c-designware-master.c
+++ b/drivers/i2c/busses/i2c-designware-master.c
@@ -253,6 +253,31 @@ static void i2c_dw_xfer_init(struct dw_i2c_dev *dev)
__i2c_dw_write_intr_mask(dev, DW_IC_INTR_MASTER_MASK);
}
+/*
+ * This functions waits controller idling before disabling I2C
+ * When the controller is not in the IDLE state,
+ * MST_ACTIVITY bit (IC_STATUS[5]) is set.
+ * Values:
+ * 0x1 (ACTIVE): Controller not idle
+ * 0x0 (IDLE): Controller is idle
+ * The function is called after returning the end of the current transfer
+ * Returns:
+ * False when controller is in IDLE state.
+ * True when controller is in ACTIVE state.
+ */
+static bool i2c_dw_is_controller_active(struct dw_i2c_dev *dev)
+{
+ u32 status;
+
+ regmap_read(dev->map, DW_IC_STATUS, &status);
+ if (!(status & DW_IC_STATUS_MASTER_ACTIVITY))
+ return false;
+
+ return regmap_read_poll_timeout(dev->map, DW_IC_STATUS, status,
+ !(status & DW_IC_STATUS_MASTER_ACTIVITY),
+ 1100, 20000) != 0;
+}
+
static int i2c_dw_check_stopbit(struct dw_i2c_dev *dev)
{
u32 val;
@@ -788,6 +813,16 @@ i2c_dw_xfer(struct i2c_adapter *adap, struct i2c_msg msgs[], int num)
goto done;
}
+ /*
+ * This happens rarely (~1:500) and is hard to reproduce. Debug trace
+ * showed that IC_STATUS had value of 0x23 when STOP_DET occurred,
+ * if disable IC_ENABLE.ENABLE immediately that can result in
+ * IC_RAW_INTR_STAT.MASTER_ON_HOLD holding SCL low. Check if
+ * controller is still ACTIVE before disabling I2C.
+ */
+ if (i2c_dw_is_controller_active(dev))
+ dev_err(dev->dev, "controller active\n");
+
/*
* We must disable the adapter before returning and signaling the end
* of the current transfer. Otherwise the hardware might continue
--
2.17.1Thu, Sep 12, 2024 at 02:11:12PM +0800, Kimriver Liu kirjoitti:
> It was observed that issuing ABORT bit (IC_ENABLE[1]) will not
> work when IC_ENABLE is already disabled.
>
> Check if ENABLE bit (IC_ENABLE[0]) is disabled when the controller
> is holding SCL low. If ENABLE bit is disabled, the software need
> to enable it before trying to issue ABORT bit. otherwise,
> the controller ignores any write to ABORT bit.
>
> These kernel logs show up whenever an I2C transaction is
> attempted after this failure.
> i2c_designware e95e0000.i2c: timeout waiting for bus ready
> i2c_designware e95e0000.i2c: timeout in disabling adapter
> The patch can be fix the controller cannot be disabled while
> SCL is held low in ENABLE bit is already disabled.
There are English words but a complete nonsense together.
Fix the condition when controller cannot be disabled while SCL
is held low and ENABLE bit is already disabled.
Reviewed-by: Andy Shevchenko <andy@kernel.org>
...
> + if (!(enable & DW_IC_ENABLE_ENABLE)) {
> + regmap_write(dev->map, DW_IC_ENABLE, DW_IC_ENABLE_ENABLE);
> + /*
> + * Wait 10 times the signaling period of the highest I2C
> + * transfer supported by the driver (for 400KHz this is
> + * 25us) to ensure the I2C ENABLE bit is already set
> + * as described in the DesignWare I2C databook.
> + */
> + fsleep(DIV_ROUND_CLOSEST_ULL(10 * MICRO, t->bus_freq_hz));
> + /* Keep ENABLE bit is already set before Setting ABORT.*/
/* Set ENABLE bit before setting ABORT */
> + enable |= DW_IC_ENABLE_ENABLE;
> + }
...
> +/*
> + * This functions waits controller idling before disabling I2C
s/functions/function/
> + * When the controller is not in the IDLE state,
> + * MST_ACTIVITY bit (IC_STATUS[5]) is set.
> + * Values:
> + * 0x1 (ACTIVE): Controller not idle
> + * 0x0 (IDLE): Controller is idle
> + * The function is called after returning the end of the current transfer
> + * Returns:
> + * False when controller is in IDLE state.
> + * True when controller is in ACTIVE state.
> + */
> +static bool i2c_dw_is_controller_active(struct dw_i2c_dev *dev)
> +{
> + u32 status;
> +
> + regmap_read(dev->map, DW_IC_STATUS, &status);
> + if (!(status & DW_IC_STATUS_MASTER_ACTIVITY))
> + return false;
> +
> + return regmap_read_poll_timeout(dev->map, DW_IC_STATUS, status,
> + !(status & DW_IC_STATUS_MASTER_ACTIVITY),
> + 1100, 20000) != 0;
return regmap_read_poll_timeout(dev->map, DW_IC_STATUS, status,
!(status & DW_IC_STATUS_MASTER_ACTIVITY),
1100, 20000) != 0;
(in the second line replaced 7 spaces by a single TAB to fix the indentation)
> +}
...
> + /*
> + * This happens rarely (~1:500) and is hard to reproduce. Debug trace
> + * showed that IC_STATUS had value of 0x23 when STOP_DET occurred,
> + * if disable IC_ENABLE.ENABLE immediately that can result in
> + * IC_RAW_INTR_STAT.MASTER_ON_HOLD holding SCL low. Check if
> + * controller is still ACTIVE before disabling I2C.
> + */
> + if (i2c_dw_is_controller_active(dev))
> + dev_err(dev->dev, "controller active\n");
--
With Best Regards,
Andy Shevchenko
On Thu, Sep 12, 2024 at 06:32:48PM GMT, Andy Shevchenko wrote:
> Thu, Sep 12, 2024 at 02:11:12PM +0800, Kimriver Liu kirjoitti:
> > It was observed that issuing ABORT bit (IC_ENABLE[1]) will not
> > work when IC_ENABLE is already disabled.
> >
> > Check if ENABLE bit (IC_ENABLE[0]) is disabled when the controller
> > is holding SCL low. If ENABLE bit is disabled, the software need
/need/needs/
> > to enable it before trying to issue ABORT bit. otherwise,
/ABORT/the ABORT/
> > the controller ignores any write to ABORT bit.
> >
> > These kernel logs show up whenever an I2C transaction is
> > attempted after this failure.
> > i2c_designware e95e0000.i2c: timeout waiting for bus ready
> > i2c_designware e95e0000.i2c: timeout in disabling adapter
>
> > The patch can be fix the controller cannot be disabled while
> > SCL is held low in ENABLE bit is already disabled.
>
> There are English words but a complete nonsense together.
We could reword this to:
The patch fixes the issue where the controller cannot be disabled
while SCL is held low if the ENABLE bit is already disabled.
> Fix the condition when controller cannot be disabled while SCL
/when/where/
> is held low and ENABLE bit is already disabled.
/ENABLE/the ENABLE/
If you agree with the changes, I can apply them before merging.
>
> Reviewed-by: Andy Shevchenko <andy@kernel.org>
Thanks a lot, Andy! I really appreciate your reviews!
> ...
>
> > + if (!(enable & DW_IC_ENABLE_ENABLE)) {
> > + regmap_write(dev->map, DW_IC_ENABLE, DW_IC_ENABLE_ENABLE);
> > + /*
> > + * Wait 10 times the signaling period of the highest I2C
> > + * transfer supported by the driver (for 400KHz this is
> > + * 25us) to ensure the I2C ENABLE bit is already set
> > + * as described in the DesignWare I2C databook.
> > + */
> > + fsleep(DIV_ROUND_CLOSEST_ULL(10 * MICRO, t->bus_freq_hz));
>
> > + /* Keep ENABLE bit is already set before Setting ABORT.*/
>
> /* Set ENABLE bit before setting ABORT */
>
>
> > + enable |= DW_IC_ENABLE_ENABLE;
> > + }
>
> ...
>
> > +/*
> > + * This functions waits controller idling before disabling I2C
>
> s/functions/function/
I can fix it before merging.
> > + * When the controller is not in the IDLE state,
> > + * MST_ACTIVITY bit (IC_STATUS[5]) is set.
> > + * Values:
> > + * 0x1 (ACTIVE): Controller not idle
> > + * 0x0 (IDLE): Controller is idle
> > + * The function is called after returning the end of the current transfer
> > + * Returns:
> > + * False when controller is in IDLE state.
> > + * True when controller is in ACTIVE state.
> > + */
> > +static bool i2c_dw_is_controller_active(struct dw_i2c_dev *dev)
> > +{
> > + u32 status;
> > +
> > + regmap_read(dev->map, DW_IC_STATUS, &status);
> > + if (!(status & DW_IC_STATUS_MASTER_ACTIVITY))
> > + return false;
> > +
> > + return regmap_read_poll_timeout(dev->map, DW_IC_STATUS, status,
> > + !(status & DW_IC_STATUS_MASTER_ACTIVITY),
> > + 1100, 20000) != 0;
>
> return regmap_read_poll_timeout(dev->map, DW_IC_STATUS, status,
> !(status & DW_IC_STATUS_MASTER_ACTIVITY),
> 1100, 20000) != 0;
>
> (in the second line replaced 7 spaces by a single TAB to fix the indentation)
I can fix it before merging.
Kimriver, if you agree, i can merge this and queue it for the
next week's fixes merge request.
Andi
> > +}
>
> ...
>
> > + /*
> > + * This happens rarely (~1:500) and is hard to reproduce. Debug trace
> > + * showed that IC_STATUS had value of 0x23 when STOP_DET occurred,
> > + * if disable IC_ENABLE.ENABLE immediately that can result in
> > + * IC_RAW_INTR_STAT.MASTER_ON_HOLD holding SCL low. Check if
> > + * controller is still ACTIVE before disabling I2C.
> > + */
> > + if (i2c_dw_is_controller_active(dev))
> > + dev_err(dev->dev, "controller active\n");
>
> --
> With Best Regards,
> Andy Shevchenko
>
>
> -----Original Message-----
> From: Andi Shyti <andi.shyti@kernel.org>
> Sent: 2024年9月13日 0:36
> To: Andy Shevchenko <andy.shevchenko@gmail.com>
> Cc: Liu Kimriver/刘金河 <kimriver.liu@siengine.com>; jarkko.nikula@linux.intel.com; andriy.shevchenko@linux.intel.com; mika.westerberg@linux.intel.com; jsd@semihalf.com; linux-i2c@vger.kernel.org; linux-> kernel@vger.kernel.org
> Subject: Re: [PATCH v10] i2c: designware: fix controller is holding SCL low while ENABLE bit is disabled
> On Thu, Sep 12, 2024 at 06:32:48PM GMT, Andy Shevchenko wrote:
> > Thu, Sep 12, 2024 at 02:11:12PM +0800, Kimriver Liu kirjoitti:
> > > It was observed that issuing ABORT bit (IC_ENABLE[1]) will not work
> > > when IC_ENABLE is already disabled.
> > >
> > > Check if ENABLE bit (IC_ENABLE[0]) is disabled when the controller
> > > is holding SCL low. If ENABLE bit is disabled, the software need
> /need/needs/
> > > to enable it before trying to issue ABORT bit. otherwise,
> /ABORT/the ABORT/
> > > the controller ignores any write to ABORT bit.
>> >
>> > These kernel logs show up whenever an I2C transaction is attempted
>> > after this failure.
>> > i2c_designware e95e0000.i2c: timeout waiting for bus ready
>> > i2c_designware e95e0000.i2c: timeout in disabling adapter
> >
>> > The patch can be fix the controller cannot be disabled while SCL is
>> > held low in ENABLE bit is already disabled.
>>
> > There are English words but a complete nonsense together.
> We could reword this to:
> The patch fixes the issue where the controller cannot be disabled while SCL is held low if the ENABLE bit is already disabled.
> > Fix the condition when controller cannot be disabled while SCL
> /when/where/
> > is held low and ENABLE bit is already disabled.
> /ENABLE/the ENABLE/
> If you agree with the changes, I can apply them before merging.
> >
> > Reviewed-by: Andy Shevchenko <andy@kernel.org>
> Thanks a lot, Andy! I really appreciate your reviews!
> > ...
> >
> > > + if (!(enable & DW_IC_ENABLE_ENABLE)) {
> > > + regmap_write(dev->map, DW_IC_ENABLE, DW_IC_ENABLE_ENABLE);
> > > + /*
>> > + * Wait 10 times the signaling period of the highest I2C
>> > + * transfer supported by the driver (for 400KHz this is
> > > + * 25us) to ensure the I2C ENABLE bit is already set
> > > + * as described in the DesignWare I2C databook.
> > > + */
> > > + fsleep(DIV_ROUND_CLOSEST_ULL(10 * MICRO, t->bus_freq_hz));
> >
> > > + /* Keep ENABLE bit is already set before Setting ABORT.*/
>>
>> /* Set ENABLE bit before setting ABORT */
>>
>>
>> > + enable |= DW_IC_ENABLE_ENABLE;
>> > + }
> >
>>
> >
> > > +/*
> > > + * This functions waits controller idling before disabling I2C
> >
> > s/functions/function/
> I can fix it before merging.
OK，I agree with your fixing it before merging. Thanks for your review!
> > > + * When the controller is not in the IDLE state,
> > > + * MST_ACTIVITY bit (IC_STATUS[5]) is set.
> > > + * Values:
> > > + * 0x1 (ACTIVE): Controller not idle
>> > + * 0x0 (IDLE): Controller is idle
>> > + * The function is called after returning the end of the current
>> > +transfer
>> > + * Returns:
> > > + * False when controller is in IDLE state.
>> > + * True when controller is in ACTIVE state.
> > > + */
> > > +static bool i2c_dw_is_controller_active(struct dw_i2c_dev *dev) {
> > > + u32 status;
> > > +
>> > + regmap_read(dev->map, DW_IC_STATUS, &status);
> > > + if (!(status & DW_IC_STATUS_MASTER_ACTIVITY))
> > > + return false;
>> > +
>> > + return regmap_read_poll_timeout(dev->map, DW_IC_STATUS, status,
>> > + !(status & DW_IC_STATUS_MASTER_ACTIVITY),
>> > + 1100, 20000) != 0;
> >
> > return regmap_read_poll_timeout(dev->map, DW_IC_STATUS, status,
>> !(status & DW_IC_STATUS_MASTER_ACTIVITY),
> > 1100, 20000) != 0;
>
> > (in the second line replaced 7 spaces by a single TAB to fix the
>> indentation)
> I can fix it before merging.
> Kimriver, if you agree, i can merge this and queue it for the next week's fixes merge request.
>Andi
OK，I agree.
Do I still need to update to patch V11? Or you fix it before merging
Thanks for taking the time to review the patch, I have learned a lot of knowledge .
> > +}
>
> ...
>
> > + /*
> > + * This happens rarely (~1:500) and is hard to reproduce. Debug trace
> > + * showed that IC_STATUS had value of 0x23 when STOP_DET occurred,
> > + * if disable IC_ENABLE.ENABLE immediately that can result in
> > + * IC_RAW_INTR_STAT.MASTER_ON_HOLD holding SCL low. Check if
> > + * controller is still ACTIVE before disabling I2C.
> > + */
> > + if (i2c_dw_is_controller_active(dev))
> > + dev_err(dev->dev, "controller active\n");
>
> --
> With Best Regards,
> Andy Shevchenko
>
>
------------------------------------------
Best Regards
Kimriver Liu
© 2016 - 2024 Red Hat, Inc.