[PATCH 0/2][RESEND] omap i2c interrupt handler fixes

[PATCH 0/2][RESEND] omap i2c interrupt handler fixes

[Alexander Shishkin]

Hi,

This is the second version of the patch that I've sent to linux-omap to
address this issue. This time I've moved the whole errata workaround bit
to a separate function to get rid of too long lines and a couple of extra
levels of indentation.

The actual fix is the same as the first time, it adds a timeout to a busy
loop which happens to take place in an interrupt handler and is capable
of hanging the kernel.

Regards,
--
Alex

[PATCH 1/2] omap i2c: make errata 1.153 workaround a separate function

[Alexander Shishkin]

This is to avoid insanely long lines and levels of indentation.

Signed-off-by: Alexander Shishkin <virtuoso-0lOfPCoBze7YtjvyW6yDsg@public.gmane.org>
CC: linux-i2c-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
CC: linux-omap-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
---
 drivers/i2c/busses/i2c-omap.c |   43 ++++++++++++++++++++++------------------
 1 files changed, 24 insertions(+), 19 deletions(-)

diff --git a/drivers/i2c/busses/i2c-omap.c b/drivers/i2c/busses/i2c-omap.c
index 75bf3ad..ad8242a 100644
--- a/drivers/i2c/busses/i2c-omap.c
+++ b/drivers/i2c/busses/i2c-omap.c
@@ -671,6 +671,27 @@ omap_i2c_rev1_isr(int this_irq, void *dev_id)
 #define omap_i2c_rev1_isr		NULL
 #endif
 
+/*
+ * OMAP3430 Errata 1.153: When an XRDY/XDR is hit, wait for XUDF before writing
+ * data to DATA_REG. Otherwise some data bytes can be lost while transferring
+ * them from the memory to the I2C interface.
+ */
+static int omap3430_workaround(struct omap_i2c_dev *dev, u16 *stat, int *err)
+{
+	while (!(*stat & OMAP_I2C_STAT_XUDF)) {
+		if (*stat & (OMAP_I2C_STAT_NACK | OMAP_I2C_STAT_AL)) {
+			omap_i2c_ack_stat(dev, *stat & (OMAP_I2C_STAT_XRDY |
+							OMAP_I2C_STAT_XDR));
+			*err |= OMAP_I2C_STAT_XUDF;
+			return -1;
+		}
+		cpu_relax();
+		*stat = omap_i2c_read_reg(dev, OMAP_I2C_STAT_REG);
+	}
+
+	return 0;
+}
+
 static irqreturn_t
 omap_i2c_isr(int this_irq, void *dev_id)
 {
@@ -794,25 +815,9 @@ complete:
 					break;
 				}
 
-				/*
-				 * OMAP3430 Errata 1.153: When an XRDY/XDR
-				 * is hit, wait for XUDF before writing data
-				 * to DATA_REG. Otherwise some data bytes can
-				 * be lost while transferring them from the
-				 * memory to the I2C interface.
-				 */
-
-				if (dev->rev <= OMAP_I2C_REV_ON_3430) {
-						while (!(stat & OMAP_I2C_STAT_XUDF)) {
-							if (stat & (OMAP_I2C_STAT_NACK | OMAP_I2C_STAT_AL)) {
-								omap_i2c_ack_stat(dev, stat & (OMAP_I2C_STAT_XRDY | OMAP_I2C_STAT_XDR));
-								err |= OMAP_I2C_STAT_XUDF;
-								goto complete;
-							}
-							cpu_relax();
-							stat = omap_i2c_read_reg(dev, OMAP_I2C_STAT_REG);
-						}
-				}
+				if (dev->rev <= OMAP_I2C_REV_ON_3430 &&
+				    omap3430_workaround(dev, &stat, &err))
+					goto complete;
 
 				omap_i2c_write_reg(dev, OMAP_I2C_DATA_REG, w);
 			}
-- 
1.6.3.3

[PATCH 2/2] omap i2c: add a timeout to the busy waiting

[Alexander Shishkin]

The errata 1.153 workaround is busy waiting on XUDF bit in interrupt
context, which may lead to kernel hangs. The problem can be reproduced
by running the bus with wrong (too high) speed.

Signed-off-by: Alexander Shishkin <virtuoso@slind.org>
CC: linux-i2c@vger.kernel.org
CC: linux-omap@vger.kernel.org
---
 drivers/i2c/busses/i2c-omap.c |    7 ++++++-
 1 files changed, 6 insertions(+), 1 deletions(-)

diff --git a/drivers/i2c/busses/i2c-omap.c b/drivers/i2c/busses/i2c-omap.c
index ad8242a..b474c20 100644
--- a/drivers/i2c/busses/i2c-omap.c
+++ b/drivers/i2c/busses/i2c-omap.c
@@ -678,7 +678,9 @@ omap_i2c_rev1_isr(int this_irq, void *dev_id)
  */
 static int omap3430_workaround(struct omap_i2c_dev *dev, u16 *stat, int *err)
 {
-	while (!(*stat & OMAP_I2C_STAT_XUDF)) {
+	unsigned long timeout = 10000;
+
+	while (!(*stat & OMAP_I2C_STAT_XUDF && --timeout)) {
 		if (*stat & (OMAP_I2C_STAT_NACK | OMAP_I2C_STAT_AL)) {
 			omap_i2c_ack_stat(dev, *stat & (OMAP_I2C_STAT_XRDY |
 							OMAP_I2C_STAT_XDR));
@@ -689,6 +691,9 @@ static int omap3430_workaround(struct omap_i2c_dev *dev, u16 *stat, int *err)
 		*stat = omap_i2c_read_reg(dev, OMAP_I2C_STAT_REG);
 	}
 
+	if (!timeout)
+		dev_err(dev->dev, "timeout waiting on XUDF bit\n");
+
 	return 0;
 }
 
-- 
1.6.3.3

Re: [PATCH 2/2] omap i2c: add a timeout to the busy waiting

[Menon, Nishanth]

Alexander Shishkin said the following on 12/16/2009 07:32 PM:
> The errata 1.153 workaround is busy waiting on XUDF bit in interrupt
> context, which may lead to kernel hangs. The problem can be reproduced
> by running the bus with wrong (too high) speed.
>
> Signed-off-by: Alexander Shishkin <virtuoso@slind.org>
> CC: linux-i2c@vger.kernel.org
> CC: linux-omap@vger.kernel.org
> ---
>  drivers/i2c/busses/i2c-omap.c |    7 ++++++-
>  1 files changed, 6 insertions(+), 1 deletions(-)
>
> diff --git a/drivers/i2c/busses/i2c-omap.c b/drivers/i2c/busses/i2c-omap.c
> index ad8242a..b474c20 100644
> --- a/drivers/i2c/busses/i2c-omap.c
> +++ b/drivers/i2c/busses/i2c-omap.c
> @@ -678,7 +678,9 @@ omap_i2c_rev1_isr(int this_irq, void *dev_id)
>   */
>  static int omap3430_workaround(struct omap_i2c_dev *dev, u16 *stat, int *err)
>  {
> -	while (!(*stat & OMAP_I2C_STAT_XUDF)) {
> +	unsigned long timeout = 10000;
> +
> +	while (!(*stat & OMAP_I2C_STAT_XUDF && --timeout)) {
>   
a) timeout without using an actual delay is not a good idea - consider 
each OPP - we can go upto 1ghz on 3630,
the actual time for 10000 iterations will depend on the MPU speed.
b) how did you arrive at the 10k iteration limit?

>  		if (*stat & (OMAP_I2C_STAT_NACK | OMAP_I2C_STAT_AL)) {
>  			omap_i2c_ack_stat(dev, *stat & (OMAP_I2C_STAT_XRDY |
>  							OMAP_I2C_STAT_XDR));
> @@ -689,6 +691,9 @@ static int omap3430_workaround(struct omap_i2c_dev *dev, u16 *stat, int *err)
>  		*stat = omap_i2c_read_reg(dev, OMAP_I2C_STAT_REG);
>  	}
>  
> +	if (!timeout)
> +		dev_err(dev->dev, "timeout waiting on XUDF bit\n");
> +
>  	return 0;
>  }
>  
>   
Regards,
Nishanth Menon

Re: [PATCH 2/2] omap i2c: add a timeout to the busy waiting

[Alexander Shishkin]

On Thu, Dec 17, 2009 at 08:38:39 +0530, Menon, Nishanth wrote:
> Alexander Shishkin said the following on 12/16/2009 07:32 PM:
> >The errata 1.153 workaround is busy waiting on XUDF bit in interrupt
> >context, which may lead to kernel hangs. The problem can be reproduced
> >by running the bus with wrong (too high) speed.
> >
> >Signed-off-by: Alexander Shishkin <virtuoso-0lOfPCoBze7YtjvyW6yDsg@public.gmane.org>
> >CC: linux-i2c-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
> >CC: linux-omap-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
> >---
> > drivers/i2c/busses/i2c-omap.c |    7 ++++++-
> > 1 files changed, 6 insertions(+), 1 deletions(-)
> >
> >diff --git a/drivers/i2c/busses/i2c-omap.c b/drivers/i2c/busses/i2c-omap.c
> >index ad8242a..b474c20 100644
> >--- a/drivers/i2c/busses/i2c-omap.c
> >+++ b/drivers/i2c/busses/i2c-omap.c
> >@@ -678,7 +678,9 @@ omap_i2c_rev1_isr(int this_irq, void *dev_id)
> >  */
> > static int omap3430_workaround(struct omap_i2c_dev *dev, u16 *stat, int *err)
> > {
> >-	while (!(*stat & OMAP_I2C_STAT_XUDF)) {
> >+	unsigned long timeout = 10000;
> >+
> >+	while (!(*stat & OMAP_I2C_STAT_XUDF && --timeout)) {
> a) timeout without using an actual delay is not a good idea -
> consider each OPP - we can go upto 1ghz on 3630,
> the actual time for 10000 iterations will depend on the MPU speed.

Well, I could calculate the timeout value based on current operating speed,
I guess. Or a delay. Perhaps OMAP_I2C_TIMEOUT can be used here?

> b) how did you arrive at the 10k iteration limit?

It was random, but then it seemed ok considering the l4 latency.

> > 		if (*stat & (OMAP_I2C_STAT_NACK | OMAP_I2C_STAT_AL)) {
> > 			omap_i2c_ack_stat(dev, *stat & (OMAP_I2C_STAT_XRDY |
> > 							OMAP_I2C_STAT_XDR));
> >@@ -689,6 +691,9 @@ static int omap3430_workaround(struct omap_i2c_dev *dev, u16 *stat, int *err)
> > 		*stat = omap_i2c_read_reg(dev, OMAP_I2C_STAT_REG);
> > 	}
> >+	if (!timeout)
> >+		dev_err(dev->dev, "timeout waiting on XUDF bit\n");
> >+
> > 	return 0;
> > }
> Regards,
> Nishanth Menon
> --
> To unsubscribe from this list: send the line "unsubscribe linux-omap" in
> the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH 2/2] omap i2c: add a timeout to the busy waiting

[Menon, Nishanth]

Alexander Shishkin said the following on 12/17/2009 07:01 PM:
> On Thu, Dec 17, 2009 at 08:38:39 +0530, Menon, Nishanth wrote:
>   
>> Alexander Shishkin said the following on 12/16/2009 07:32 PM:
>>     
>>> The errata 1.153 workaround is busy waiting on XUDF bit in interrupt
>>> context, which may lead to kernel hangs. The problem can be reproduced
>>> by running the bus with wrong (too high) speed.
>>>
>>> Signed-off-by: Alexander Shishkin <virtuoso-0lOfPCoBze7YtjvyW6yDsg@public.gmane.org>
>>> CC: linux-i2c-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
>>> CC: linux-omap-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
>>> ---
>>> drivers/i2c/busses/i2c-omap.c |    7 ++++++-
>>> 1 files changed, 6 insertions(+), 1 deletions(-)
>>>
>>> diff --git a/drivers/i2c/busses/i2c-omap.c b/drivers/i2c/busses/i2c-omap.c
>>> index ad8242a..b474c20 100644
>>> --- a/drivers/i2c/busses/i2c-omap.c
>>> +++ b/drivers/i2c/busses/i2c-omap.c
>>> @@ -678,7 +678,9 @@ omap_i2c_rev1_isr(int this_irq, void *dev_id)
>>>  */
>>> static int omap3430_workaround(struct omap_i2c_dev *dev, u16 *stat, int *err)
>>> {
>>> -	while (!(*stat & OMAP_I2C_STAT_XUDF)) {
>>> +	unsigned long timeout = 10000;
>>> +
>>> +	while (!(*stat & OMAP_I2C_STAT_XUDF && --timeout)) {
>>>       
>> a) timeout without using an actual delay is not a good idea -
>> consider each OPP - we can go upto 1ghz on 3630,
>> the actual time for 10000 iterations will depend on the MPU speed.
>>     
>
> Well, I could calculate the timeout value based on current operating speed,
> I guess. Or a delay. Perhaps OMAP_I2C_TIMEOUT can be used here?
>   
it might be an overkill trying to generate counter based on opp speeds. 
wondering if this delay is required in the first place..

Moiz Sonasath / Vikram P, (commit cd086d3a) could probably comment?
Ref: http://www.mail-archive.com/linux-omap-u79uwXL29TY76Z2rM5mHXA@public.gmane.org/msg15317.html

>   
>> b) how did you arrive at the 10k iteration limit?
>>     
>
> It was random, but then it seemed ok considering the l4 latency.
>   
I had guessed this is an emperical value. It will be good to know the 
exact rationale on why a timeout will happen, else we will have all kind 
of questions pending trying to figure out if a timeout happened.

>   
>>> 		if (*stat & (OMAP_I2C_STAT_NACK | OMAP_I2C_STAT_AL)) {
>>> 			omap_i2c_ack_stat(dev, *stat & (OMAP_I2C_STAT_XRDY |
>>> 							OMAP_I2C_STAT_XDR));
>>> @@ -689,6 +691,9 @@ static int omap3430_workaround(struct omap_i2c_dev *dev, u16 *stat, int *err)
>>> 		*stat = omap_i2c_read_reg(dev, OMAP_I2C_STAT_REG);
>>> 	}
>>> +	if (!timeout)
>>> +		dev_err(dev->dev, "timeout waiting on XUDF bit\n");
>>> +
>>> 	return 0;
>>> }
Regards,
Nishanth Menon

Re: [PATCH 2/2] omap i2c: add a timeout to the busy waiting

[Menon, Nishanth]

Alexander Shishkin said the following on 12/17/2009 07:01 PM:
> On Thu, Dec 17, 2009 at 08:38:39 +0530, Menon, Nishanth wrote:
>   
>> Alexander Shishkin said the following on 12/16/2009 07:32 PM:
>>     
>>> The errata 1.153 workaround is busy waiting on XUDF bit in interrupt
>>> context, which may lead to kernel hangs. The problem can be reproduced
>>> by running the bus with wrong (too high) speed.
>>>
>>> Signed-off-by: Alexander Shishkin <virtuoso@slind.org>
>>> CC: linux-i2c@vger.kernel.org
>>> CC: linux-omap@vger.kernel.org
>>> ---
>>> drivers/i2c/busses/i2c-omap.c |    7 ++++++-
>>> 1 files changed, 6 insertions(+), 1 deletions(-)
>>>
>>> diff --git a/drivers/i2c/busses/i2c-omap.c b/drivers/i2c/busses/i2c-omap.c
>>> index ad8242a..b474c20 100644
>>> --- a/drivers/i2c/busses/i2c-omap.c
>>> +++ b/drivers/i2c/busses/i2c-omap.c
>>> @@ -678,7 +678,9 @@ omap_i2c_rev1_isr(int this_irq, void *dev_id)
>>>  */
>>> static int omap3430_workaround(struct omap_i2c_dev *dev, u16 *stat, int *err)
>>> {
>>> -	while (!(*stat & OMAP_I2C_STAT_XUDF)) {
>>> +	unsigned long timeout = 10000;
>>> +
>>> +	while (!(*stat & OMAP_I2C_STAT_XUDF && --timeout)) {
>>>       
>> a) timeout without using an actual delay is not a good idea -
>> consider each OPP - we can go upto 1ghz on 3630,
>> the actual time for 10000 iterations will depend on the MPU speed.
>>     
>
> Well, I could calculate the timeout value based on current operating speed,
> I guess. Or a delay. Perhaps OMAP_I2C_TIMEOUT can be used here?
>   
it might be an overkill trying to generate counter based on opp speeds. 
wondering if this delay is required in the first place..

Moiz Sonasath / Vikram P, (commit cd086d3a) could probably comment?
Ref: http://www.mail-archive.com/linux-omap@vger.kernel.org/msg15317.html

>   
>> b) how did you arrive at the 10k iteration limit?
>>     
>
> It was random, but then it seemed ok considering the l4 latency.
>   
I had guessed this is an emperical value. It will be good to know the 
exact rationale on why a timeout will happen, else we will have all kind 
of questions pending trying to figure out if a timeout happened.

>   
>>> 		if (*stat & (OMAP_I2C_STAT_NACK | OMAP_I2C_STAT_AL)) {
>>> 			omap_i2c_ack_stat(dev, *stat & (OMAP_I2C_STAT_XRDY |
>>> 							OMAP_I2C_STAT_XDR));
>>> @@ -689,6 +691,9 @@ static int omap3430_workaround(struct omap_i2c_dev *dev, u16 *stat, int *err)
>>> 		*stat = omap_i2c_read_reg(dev, OMAP_I2C_STAT_REG);
>>> 	}
>>> +	if (!timeout)
>>> +		dev_err(dev->dev, "timeout waiting on XUDF bit\n");
>>> +
>>> 	return 0;
>>> }
Regards,
Nishanth Menon

RE: [PATCH 2/2] omap i2c: add a timeout to the busy waiting

[Sonasath, Moiz]

Alex,

> -----Original Message-----
> From: Menon, Nishanth
> Sent: Thursday, December 17, 2009 7:59 AM
> To: Menon, Nishanth; ben-linux-elnMNo+KYs3YtjvyW6yDsg@public.gmane.org; linux-omap-u79uwXL29TY76Z2rM5mHXA@public.gmane.org;
> linux-i2c-u79uwXL29TY76Z2rM5mHXA@public.gmane.org; Sonasath, Moiz; Pandita, Vikram
> Subject: Re: [PATCH 2/2] omap i2c: add a timeout to the busy waiting
> 
> Alexander Shishkin said the following on 12/17/2009 07:01 PM:
> > On Thu, Dec 17, 2009 at 08:38:39 +0530, Menon, Nishanth wrote:
> >
> >> Alexander Shishkin said the following on 12/16/2009 07:32 PM:
> >>
> >>> The errata 1.153 workaround is busy waiting on XUDF bit in interrupt
> >>> context, which may lead to kernel hangs. The problem can be reproduced
> >>> by running the bus with wrong (too high) speed.
> >>>
> >>> Signed-off-by: Alexander Shishkin <virtuoso-0lOfPCoBze7YtjvyW6yDsg@public.gmane.org>
> >>> CC: linux-i2c-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
> >>> CC: linux-omap-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
> >>> ---
> >>> drivers/i2c/busses/i2c-omap.c |    7 ++++++-
> >>> 1 files changed, 6 insertions(+), 1 deletions(-)
> >>>
> >>> diff --git a/drivers/i2c/busses/i2c-omap.c b/drivers/i2c/busses/i2c-
> omap.c
> >>> index ad8242a..b474c20 100644
> >>> --- a/drivers/i2c/busses/i2c-omap.c
> >>> +++ b/drivers/i2c/busses/i2c-omap.c
> >>> @@ -678,7 +678,9 @@ omap_i2c_rev1_isr(int this_irq, void *dev_id)
> >>>  */
> >>> static int omap3430_workaround(struct omap_i2c_dev *dev, u16 *stat,
> int *err)
> >>> {
> >>> -	while (!(*stat & OMAP_I2C_STAT_XUDF)) {
> >>> +	unsigned long timeout = 10000;
> >>> +
> >>> +	while (!(*stat & OMAP_I2C_STAT_XUDF && --timeout)) {
> >>>
> >> a) timeout without using an actual delay is not a good idea -
> >> consider each OPP - we can go upto 1ghz on 3630,
> >> the actual time for 10000 iterations will depend on the MPU speed.
> >>
> >
> > Well, I could calculate the timeout value based on current operating
> speed,
> > I guess. Or a delay. Perhaps OMAP_I2C_TIMEOUT can be used here?
> >
> it might be an overkill trying to generate counter based on opp speeds.
> wondering if this delay is required in the first place..

The reason why this timeout was not put in place was that if the I2C controller is functional, the XUDF has to set after XRDY/XDR are set, the only case where it might not be set is when there is a sudden NACK | AL condition produced on the bus and the transfer stops (thereby the FIFO is not getting empty). The code takes care of this error situation so IMHO we can do without a timeout here.

--- Moiz
> 
> Moiz Sonasath / Vikram P, (commit cd086d3a) could probably comment?
> Ref: http://www.mail-archive.com/linux-omap-u79uwXL29TY76Z2rM5mHXA@public.gmane.org/msg15317.html
> 
> >
> >> b) how did you arrive at the 10k iteration limit?
> >>
> >
> > It was random, but then it seemed ok considering the l4 latency.
> >
> I had guessed this is an emperical value. It will be good to know the
> exact rationale on why a timeout will happen, else we will have all kind
> of questions pending trying to figure out if a timeout happened.
> 
> >
> >>> 		if (*stat & (OMAP_I2C_STAT_NACK | OMAP_I2C_STAT_AL)) {
> >>> 			omap_i2c_ack_stat(dev, *stat & (OMAP_I2C_STAT_XRDY |
> >>> 							OMAP_I2C_STAT_XDR));
> >>> @@ -689,6 +691,9 @@ static int omap3430_workaround(struct omap_i2c_dev
> *dev, u16 *stat, int *err)
> >>> 		*stat = omap_i2c_read_reg(dev, OMAP_I2C_STAT_REG);
> >>> 	}
> >>> +	if (!timeout)
> >>> +		dev_err(dev->dev, "timeout waiting on XUDF bit\n");
> >>> +
> >>> 	return 0;
> >>> }
> Regards,
> Nishanth Menon

Re: [PATCH 2/2] omap i2c: add a timeout to the busy waiting

[Aaro Koskinen]

Hi,

Sonasath, Moiz wrote:
>> From: Menon, Nishanth
>> Alexander Shishkin said the following on 12/17/2009 07:01 PM:
>>> Well, I could calculate the timeout value based on current operating
>> speed,
>>> I guess. Or a delay. Perhaps OMAP_I2C_TIMEOUT can be used here?
>>>
>> it might be an overkill trying to generate counter based on opp speeds.
>> wondering if this delay is required in the first place..
> 
> The reason why this timeout was not put in place was that if the I2C
> controller is functional, the XUDF has to set after XRDY/XDR are set,
> the only case where it might not be set is when there is a sudden
> NACK | AL condition produced on the bus and the transfer stops
> (thereby the FIFO is not getting empty). The code takes care of this
> error situation so IMHO we can do without a timeout here.

We are able to reproduce an error situation, where none of XUDF|NACK|AL
gets set. In that case the kernel hangs. If we add the timeout, we are
able to recover. If you want to have a robust driver, you cannot have a
loop which is not guaranteed to terminate.

A.

Re: [PATCH 1/2] omap i2c: make errata 1.153 workaround a separate function

[Menon, Nishanth]

Alexander Shishkin said the following on 12/16/2009 07:32 PM:
> This is to avoid insanely long lines and levels of indentation.
>
> Signed-off-by: Alexander Shishkin <virtuoso-0lOfPCoBze7YtjvyW6yDsg@public.gmane.org>
> CC: linux-i2c-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
> CC: linux-omap-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
> ---
>  drivers/i2c/busses/i2c-omap.c |   43 ++++++++++++++++++++++------------------
>  1 files changed, 24 insertions(+), 19 deletions(-)
>
> diff --git a/drivers/i2c/busses/i2c-omap.c b/drivers/i2c/busses/i2c-omap.c
> index 75bf3ad..ad8242a 100644
> --- a/drivers/i2c/busses/i2c-omap.c
> +++ b/drivers/i2c/busses/i2c-omap.c
> @@ -671,6 +671,27 @@ omap_i2c_rev1_isr(int this_irq, void *dev_id)
>  #define omap_i2c_rev1_isr		NULL
>  #endif
>  
> +/*
> + * OMAP3430 Errata 1.153: When an XRDY/XDR is hit, wait for XUDF before writing
> + * data to DATA_REG. Otherwise some data bytes can be lost while transferring
> + * them from the memory to the I2C interface.
> + */
> +static int omap3430_workaround(struct omap_i2c_dev *dev, u16 *stat, int *err)
>   

note, though this is identified as being part of 3430, it is not really 
restricted to 3430 alone
we might want to rename this as errata_omap3_1p153() perhaps?

> +{
> +	while (!(*stat & OMAP_I2C_STAT_XUDF)) {
> +		if (*stat & (OMAP_I2C_STAT_NACK | OMAP_I2C_STAT_AL)) {
> +			omap_i2c_ack_stat(dev, *stat & (OMAP_I2C_STAT_XRDY |
> +							OMAP_I2C_STAT_XDR));
> +			*err |= OMAP_I2C_STAT_XUDF;
> +			return -1;
> +		}
> +		cpu_relax();
> +		*stat = omap_i2c_read_reg(dev, OMAP_I2C_STAT_REG);
> +	}
> +
> +	return 0;
> +}
>   
wonder if using an inline might help throw away the function call 
overhead (considering it is used only once)?
> +
>  static irqreturn_t
>  omap_i2c_isr(int this_irq, void *dev_id)
>  {
> @@ -794,25 +815,9 @@ complete:
>  					break;
>  				}
>  
> -				/*
> -				 * OMAP3430 Errata 1.153: When an XRDY/XDR
> -				 * is hit, wait for XUDF before writing data
> -				 * to DATA_REG. Otherwise some data bytes can
> -				 * be lost while transferring them from the
> -				 * memory to the I2C interface.
> -				 */
> -
> -				if (dev->rev <= OMAP_I2C_REV_ON_3430) {
> -						while (!(stat & OMAP_I2C_STAT_XUDF)) {
> -							if (stat & (OMAP_I2C_STAT_NACK | OMAP_I2C_STAT_AL)) {
> -								omap_i2c_ack_stat(dev, stat & (OMAP_I2C_STAT_XRDY | OMAP_I2C_STAT_XDR));
> -								err |= OMAP_I2C_STAT_XUDF;
> -								goto complete;
> -							}
> -							cpu_relax();
> -							stat = omap_i2c_read_reg(dev, OMAP_I2C_STAT_REG);
> -						}
> -				}
> +				if (dev->rev <= OMAP_I2C_REV_ON_3430 &&
> +				    omap3430_workaround(dev, &stat, &err))
> +					goto complete;
>  
>  				omap_i2c_write_reg(dev, OMAP_I2C_DATA_REG, w);
>  			}
>   

Regards,
Nishanth Menon

Re: [PATCH 1/2] omap i2c: make errata 1.153 workaround a separate function

[Alexander Shishkin]

On Thu, Dec 17, 2009 at 08:36:30 +0530, Menon, Nishanth wrote:
> Alexander Shishkin said the following on 12/16/2009 07:32 PM:
> >This is to avoid insanely long lines and levels of indentation.
> >
> >Signed-off-by: Alexander Shishkin <virtuoso-0lOfPCoBze7YtjvyW6yDsg@public.gmane.org>
> >CC: linux-i2c-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
> >CC: linux-omap-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
> >---
> > drivers/i2c/busses/i2c-omap.c |   43 ++++++++++++++++++++++------------------
> > 1 files changed, 24 insertions(+), 19 deletions(-)
> >
> >diff --git a/drivers/i2c/busses/i2c-omap.c b/drivers/i2c/busses/i2c-omap.c
> >index 75bf3ad..ad8242a 100644
> >--- a/drivers/i2c/busses/i2c-omap.c
> >+++ b/drivers/i2c/busses/i2c-omap.c
> >@@ -671,6 +671,27 @@ omap_i2c_rev1_isr(int this_irq, void *dev_id)
> > #define omap_i2c_rev1_isr		NULL
> > #endif
> >+/*
> >+ * OMAP3430 Errata 1.153: When an XRDY/XDR is hit, wait for XUDF before writing
> >+ * data to DATA_REG. Otherwise some data bytes can be lost while transferring
> >+ * them from the memory to the I2C interface.
> >+ */
> >+static int omap3430_workaround(struct omap_i2c_dev *dev, u16 *stat, int *err)
> 
> note, though this is identified as being part of 3430, it is not
> really restricted to 3430 alone
> we might want to rename this as errata_omap3_1p153() perhaps?

Ok, I don't see why not.

> >+{
> >+	while (!(*stat & OMAP_I2C_STAT_XUDF)) {
> >+		if (*stat & (OMAP_I2C_STAT_NACK | OMAP_I2C_STAT_AL)) {
> >+			omap_i2c_ack_stat(dev, *stat & (OMAP_I2C_STAT_XRDY |
> >+							OMAP_I2C_STAT_XDR));
> >+			*err |= OMAP_I2C_STAT_XUDF;
> >+			return -1;
> >+		}
> >+		cpu_relax();
> >+		*stat = omap_i2c_read_reg(dev, OMAP_I2C_STAT_REG);
> >+	}
> >+
> >+	return 0;
> >+}
> wonder if using an inline might help throw away the function call
> overhead (considering it is used only once)?

objdump -S says it's implicitly inlined already. I actually had in mind
the conversation about generalizing the features/erratas for chips/IPs
and that somehow stopped me from explicitly inlining this. Do you think
it makes sense (for the next version of this patchset) to explicitly
inline this?

> >+
> > static irqreturn_t
> > omap_i2c_isr(int this_irq, void *dev_id)
> > {
> >@@ -794,25 +815,9 @@ complete:
> > 					break;
> > 				}
> >-				/*
> >-				 * OMAP3430 Errata 1.153: When an XRDY/XDR
> >-				 * is hit, wait for XUDF before writing data
> >-				 * to DATA_REG. Otherwise some data bytes can
> >-				 * be lost while transferring them from the
> >-				 * memory to the I2C interface.
> >-				 */
> >-
> >-				if (dev->rev <= OMAP_I2C_REV_ON_3430) {
> >-						while (!(stat & OMAP_I2C_STAT_XUDF)) {
> >-							if (stat & (OMAP_I2C_STAT_NACK | OMAP_I2C_STAT_AL)) {
> >-								omap_i2c_ack_stat(dev, stat & (OMAP_I2C_STAT_XRDY | OMAP_I2C_STAT_XDR));
> >-								err |= OMAP_I2C_STAT_XUDF;
> >-								goto complete;
> >-							}
> >-							cpu_relax();
> >-							stat = omap_i2c_read_reg(dev, OMAP_I2C_STAT_REG);
> >-						}
> >-				}
> >+				if (dev->rev <= OMAP_I2C_REV_ON_3430 &&
> >+				    omap3430_workaround(dev, &stat, &err))
> >+					goto complete;
> > 				omap_i2c_write_reg(dev, OMAP_I2C_DATA_REG, w);
> > 			}
> 
> Regards,
> Nishanth Menon

Re: [PATCH 1/2] omap i2c: make errata 1.153 workaround a separate function

[Menon, Nishanth]

Alexander Shishkin said the following on 12/17/2009 06:18 PM:
> On Thu, Dec 17, 2009 at 08:36:30 +0530, Menon, Nishanth wrote:
>   
>> Alexander Shishkin said the following on 12/16/2009 07:32 PM:
>>     
>>> This is to avoid insanely long lines and levels of indentation.
>>>
>>> Signed-off-by: Alexander Shishkin <virtuoso@slind.org>
>>> CC: linux-i2c@vger.kernel.org
>>> CC: linux-omap@vger.kernel.org
>>> ---
>>> drivers/i2c/busses/i2c-omap.c |   43 ++++++++++++++++++++++------------------
>>> 1 files changed, 24 insertions(+), 19 deletions(-)
>>>
>>> diff --git a/drivers/i2c/busses/i2c-omap.c b/drivers/i2c/busses/i2c-omap.c
>>> index 75bf3ad..ad8242a 100644
>>> --- a/drivers/i2c/busses/i2c-omap.c
>>> +++ b/drivers/i2c/busses/i2c-omap.c
>>> @@ -671,6 +671,27 @@ omap_i2c_rev1_isr(int this_irq, void *dev_id)
>>> #define omap_i2c_rev1_isr		NULL
>>> #endif
>>> +/*
>>> + * OMAP3430 Errata 1.153: When an XRDY/XDR is hit, wait for XUDF before writing
>>> + * data to DATA_REG. Otherwise some data bytes can be lost while transferring
>>> + * them from the memory to the I2C interface.
>>> + */
>>> +static int omap3430_workaround(struct omap_i2c_dev *dev, u16 *stat, int *err)
>>>       
>> note, though this is identified as being part of 3430, it is not
>> really restricted to 3430 alone
>> we might want to rename this as errata_omap3_1p153() perhaps?
>>     
>
> Ok, I don't see why not.
>   
Thanks..
>   
>>> +{
>>> +	while (!(*stat & OMAP_I2C_STAT_XUDF)) {
>>> +		if (*stat & (OMAP_I2C_STAT_NACK | OMAP_I2C_STAT_AL)) {
>>> +			omap_i2c_ack_stat(dev, *stat & (OMAP_I2C_STAT_XRDY |
>>> +							OMAP_I2C_STAT_XDR));
>>> +			*err |= OMAP_I2C_STAT_XUDF;
>>> +			return -1;
>>> +		}
>>> +		cpu_relax();
>>> +		*stat = omap_i2c_read_reg(dev, OMAP_I2C_STAT_REG);
>>> +	}
>>> +
>>> +	return 0;
>>> +}
>>>       
>> wonder if using an inline might help throw away the function call
>> overhead (considering it is used only once)?
>>     
>
> objdump -S says it's implicitly inlined already. I actually had in mind
> the conversation about generalizing the features/erratas for chips/IPs
> and that somehow stopped me from explicitly inlining this. Do you think
> it makes sense (for the next version of this patchset) to explicitly
> inline this?
>   
I guess that might allow folks to realize that without objdump -S ;)
[...]

regards,
Nishanth Menon

Re: [PATCH 1/2] omap i2c: make errata 1.153 workaround a separate function

[Alexander Shishkin]

On Wed, Dec 16, 2009 at 04:02:23 +0200, Alexander Shishkin wrote:
> This is to avoid insanely long lines and levels of indentation.

These seem to be forgotten. Is there any problem with these I should address?

Regards,
--
Alex

Re: [PATCH 1/2] omap i2c: make errata 1.153 workaround a separate function

[Tony Lindgren]

* Alexander Shishkin <virtuoso-0lOfPCoBze7YtjvyW6yDsg@public.gmane.org> [100316 04:28]:
> On Wed, Dec 16, 2009 at 04:02:23 +0200, Alexander Shishkin wrote:
> > This is to avoid insanely long lines and levels of indentation.
> 
> These seem to be forgotten. Is there any problem with these I should address?

Can you please repost and I'll add those too into omap-testing
first? Or maybe point to the right patchwork.kernel.org link.

Regards,

Tony

Re: [PATCH 1/2] omap i2c: make errata 1.153 workaround a separate function

[Alexander Shishkin]

> Can you please repost and I'll add those too into omap-testing
> first? Or maybe point to the right patchwork.kernel.org link.

Resending.

P.S. I think linux-i2c@ mailing list adds some vicious email headers that
the email clients obey and don't include me in the to/cc in replies. I've
only found Tony's reply in the archives because I suspected something wrong.

Regards,
--
Alex

[PATCH v2 2/2] omap i2c: add a timeout to the busy waiting

[Alexander Shishkin]

The errata 1.153 workaround is busy waiting on XUDF bit in interrupt
context, which may lead to kernel hangs. The problem can be reproduced
by running the bus with wrong (too high) speed.

Signed-off-by: Alexander Shishkin <virtuoso-0lOfPCoBze7YtjvyW6yDsg@public.gmane.org>
Acked-by: Tony Lindgren <tony-4v6yS6AI5VpBDgjK7y7TUQ@public.gmane.org>
CC: linux-i2c-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
CC: linux-omap-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
CC: nm-l0cyMroinI0@public.gmane.org
---
 drivers/i2c/busses/i2c-omap.c |    8 ++++++++
 1 files changed, 8 insertions(+), 0 deletions(-)

diff --git a/drivers/i2c/busses/i2c-omap.c b/drivers/i2c/busses/i2c-omap.c
index 2d146ac..7d56a25 100644
--- a/drivers/i2c/busses/i2c-omap.c
+++ b/drivers/i2c/busses/i2c-omap.c
@@ -678,6 +678,8 @@ omap_i2c_rev1_isr(int this_irq, void *dev_id)
  */
 static int errata_omap3_1p153(struct omap_i2c_dev *dev, u16 *stat, int *err)
 {
+	unsigned long timeout = jiffies + msecs_to_jiffies(1);
+
 	while (!(*stat & OMAP_I2C_STAT_XUDF)) {
 		if (*stat & (OMAP_I2C_STAT_NACK | OMAP_I2C_STAT_AL)) {
 			omap_i2c_ack_stat(dev, *stat & (OMAP_I2C_STAT_XRDY |
@@ -685,6 +687,12 @@ static int errata_omap3_1p153(struct omap_i2c_dev *dev, u16 *stat, int *err)
 			*err |= OMAP_I2C_STAT_XUDF;
 			return -ETIMEDOUT;
 		}
+
+		if (time_after(jiffies, timeout)) {
+			dev_err(dev->dev, "timeout waiting on XUDF bit\n");
+			return 0;
+		}
+
 		cpu_relax();
 		*stat = omap_i2c_read_reg(dev, OMAP_I2C_STAT_REG);
 	}
-- 
1.6.3.3

Re: [PATCH v2 2/2] omap i2c: add a timeout to the busy waiting

[Aaro Koskinen]

Hi,

Alexander Shishkin wrote:
> The errata 1.153 workaround is busy waiting on XUDF bit in interrupt
> context, which may lead to kernel hangs. The problem can be reproduced
> by running the bus with wrong (too high) speed.
> 
> Signed-off-by: Alexander Shishkin <virtuoso-0lOfPCoBze7YtjvyW6yDsg@public.gmane.org>
> Acked-by: Tony Lindgren <tony-4v6yS6AI5VpBDgjK7y7TUQ@public.gmane.org>
> CC: linux-i2c-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
> CC: linux-omap-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
> CC: nm-l0cyMroinI0@public.gmane.org
> ---
>  drivers/i2c/busses/i2c-omap.c |    8 ++++++++
>  1 files changed, 8 insertions(+), 0 deletions(-)
> 
> diff --git a/drivers/i2c/busses/i2c-omap.c b/drivers/i2c/busses/i2c-omap.c
> index 2d146ac..7d56a25 100644
> --- a/drivers/i2c/busses/i2c-omap.c
> +++ b/drivers/i2c/busses/i2c-omap.c
> @@ -678,6 +678,8 @@ omap_i2c_rev1_isr(int this_irq, void *dev_id)
>   */
>  static int errata_omap3_1p153(struct omap_i2c_dev *dev, u16 *stat, int *err)
>  {
> +	unsigned long timeout = jiffies + msecs_to_jiffies(1);
> +
>  	while (!(*stat & OMAP_I2C_STAT_XUDF)) {
>  		if (*stat & (OMAP_I2C_STAT_NACK | OMAP_I2C_STAT_AL)) {
>  			omap_i2c_ack_stat(dev, *stat & (OMAP_I2C_STAT_XRDY |
> @@ -685,6 +687,12 @@ static int errata_omap3_1p153(struct omap_i2c_dev *dev, u16 *stat, int *err)
>  			*err |= OMAP_I2C_STAT_XUDF;
>  			return -ETIMEDOUT;
>  		}
> +
> +		if (time_after(jiffies, timeout)) {

This is called from hard interrupt context, so we cannot use jiffies.

> +			dev_err(dev->dev, "timeout waiting on XUDF bit\n");
> +			return 0;
> +		}
> +
>  		cpu_relax();
>  		*stat = omap_i2c_read_reg(dev, OMAP_I2C_STAT_REG);
>  	}

A.

Re: [PATCH v2 2/2] omap i2c: add a timeout to the busy waiting

[Alexander Shishkin]

On Thu, Mar 25, 2010 at 04:38:09 +0200, Aaro Koskinen wrote:
> Hi,
> 
> Alexander Shishkin wrote:
> > The errata 1.153 workaround is busy waiting on XUDF bit in interrupt
> > context, which may lead to kernel hangs. The problem can be reproduced
> > by running the bus with wrong (too high) speed.
> > 
> > Signed-off-by: Alexander Shishkin <virtuoso-0lOfPCoBze7YtjvyW6yDsg@public.gmane.org>
> > Acked-by: Tony Lindgren <tony-4v6yS6AI5VpBDgjK7y7TUQ@public.gmane.org>
> > CC: linux-i2c-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
> > CC: linux-omap-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
> > CC: nm-l0cyMroinI0@public.gmane.org
> > ---
> >  drivers/i2c/busses/i2c-omap.c |    8 ++++++++
> >  1 files changed, 8 insertions(+), 0 deletions(-)
> > 
> > diff --git a/drivers/i2c/busses/i2c-omap.c b/drivers/i2c/busses/i2c-omap.c
> > index 2d146ac..7d56a25 100644
> > --- a/drivers/i2c/busses/i2c-omap.c
> > +++ b/drivers/i2c/busses/i2c-omap.c
> > @@ -678,6 +678,8 @@ omap_i2c_rev1_isr(int this_irq, void *dev_id)
> >   */
> >  static int errata_omap3_1p153(struct omap_i2c_dev *dev, u16 *stat, int *err)
> >  {
> > +	unsigned long timeout = jiffies + msecs_to_jiffies(1);
> > +
> >  	while (!(*stat & OMAP_I2C_STAT_XUDF)) {
> >  		if (*stat & (OMAP_I2C_STAT_NACK | OMAP_I2C_STAT_AL)) {
> >  			omap_i2c_ack_stat(dev, *stat & (OMAP_I2C_STAT_XRDY |
> > @@ -685,6 +687,12 @@ static int errata_omap3_1p153(struct omap_i2c_dev *dev, u16 *stat, int *err)
> >  			*err |= OMAP_I2C_STAT_XUDF;
> >  			return -ETIMEDOUT;
> >  		}
> > +
> > +		if (time_after(jiffies, timeout)) {
> 
> This is called from hard interrupt context, so we cannot use jiffies.

Ahh, stupid me. Then we are left with the iteration counting, I guess.

Regards,
--
Alex

[PATCH v3] omap i2c: add a timeout to the busy waiting

[Alexander Shishkin]

The errata 1.153 workaround is busy waiting on XUDF bit in interrupt
context, which may lead to kernel hangs. The problem can be reproduced
by running the bus with wrong (too high) speed.

Signed-off-by: Alexander Shishkin <virtuoso@slind.org>
---
 drivers/i2c/busses/i2c-omap.c |   10 +++++++++-
 1 files changed, 9 insertions(+), 1 deletions(-)

diff --git a/drivers/i2c/busses/i2c-omap.c b/drivers/i2c/busses/i2c-omap.c
index ef73483..00fd02e 100644
--- a/drivers/i2c/busses/i2c-omap.c
+++ b/drivers/i2c/busses/i2c-omap.c
@@ -763,17 +763,25 @@ omap_i2c_rev1_isr(int this_irq, void *dev_id)
  */
 static int errata_omap3_1p153(struct omap_i2c_dev *dev, u16 *stat, int *err)
 {
-	while (!(*stat & OMAP_I2C_STAT_XUDF)) {
+	unsigned long timeout = 10000;
+
+	while (--timeout && !(*stat & OMAP_I2C_STAT_XUDF)) {
 		if (*stat & (OMAP_I2C_STAT_NACK | OMAP_I2C_STAT_AL)) {
 			omap_i2c_ack_stat(dev, *stat & (OMAP_I2C_STAT_XRDY |
 							OMAP_I2C_STAT_XDR));
 			*err |= OMAP_I2C_STAT_XUDF;
 			return -ETIMEDOUT;
 		}
+
 		cpu_relax();
 		*stat = omap_i2c_read_reg(dev, OMAP_I2C_STAT_REG);
 	}
 
+	if (!timeout) {
+		dev_err(dev->dev, "timeout waiting on XUDF bit\n");
+		return 0;
+	}
+
 	return 0;
 }
 
-- 
1.7.1.1.g15764

[APPLIED] [PATCH v3] omap i2c: add a timeout to the busy waiting

[Tony Lindgren]

This patch has been applied to the linux-omap
by youw fwiendly patch wobot.

Branch in linux-omap: i2c-omap-for-ben

Initial commit ID (Likely to change): 4c156dc0e7cc486feb106348fb7159d4be0c9889

PatchWorks
http://patchwork.kernel.org/patch/98127/

Git (Likely to change, and takes a while to get mirrored)
http://git.kernel.org/?p=linux/kernel/git/tmlind/linux-omap-2.6.git;a=commit;h=4c156dc0e7cc486feb106348fb7159d4be0c9889

[PATCH v2 1/2] omap i2c: make errata 1.153 workaround a separate function

[Alexander Shishkin]

This is to avoid insanely long lines and levels of indentation.

Signed-off-by: Alexander Shishkin <virtuoso@slind.org>
Acked-by: Tony Lindgren <tony@atomide.com>
CC: linux-i2c@vger.kernel.org
CC: linux-omap@vger.kernel.org
CC: nm@ti.com
---
 drivers/i2c/busses/i2c-omap.c |   43 ++++++++++++++++++++++------------------
 1 files changed, 24 insertions(+), 19 deletions(-)

diff --git a/drivers/i2c/busses/i2c-omap.c b/drivers/i2c/busses/i2c-omap.c
index 75bf3ad..2d146ac 100644
--- a/drivers/i2c/busses/i2c-omap.c
+++ b/drivers/i2c/busses/i2c-omap.c
@@ -671,6 +671,27 @@ omap_i2c_rev1_isr(int this_irq, void *dev_id)
 #define omap_i2c_rev1_isr		NULL
 #endif
 
+/*
+ * OMAP3430 Errata 1.153: When an XRDY/XDR is hit, wait for XUDF before writing
+ * data to DATA_REG. Otherwise some data bytes can be lost while transferring
+ * them from the memory to the I2C interface.
+ */
+static int errata_omap3_1p153(struct omap_i2c_dev *dev, u16 *stat, int *err)
+{
+	while (!(*stat & OMAP_I2C_STAT_XUDF)) {
+		if (*stat & (OMAP_I2C_STAT_NACK | OMAP_I2C_STAT_AL)) {
+			omap_i2c_ack_stat(dev, *stat & (OMAP_I2C_STAT_XRDY |
+							OMAP_I2C_STAT_XDR));
+			*err |= OMAP_I2C_STAT_XUDF;
+			return -ETIMEDOUT;
+		}
+		cpu_relax();
+		*stat = omap_i2c_read_reg(dev, OMAP_I2C_STAT_REG);
+	}
+
+	return 0;
+}
+
 static irqreturn_t
 omap_i2c_isr(int this_irq, void *dev_id)
 {
@@ -794,25 +815,9 @@ complete:
 					break;
 				}
 
-				/*
-				 * OMAP3430 Errata 1.153: When an XRDY/XDR
-				 * is hit, wait for XUDF before writing data
-				 * to DATA_REG. Otherwise some data bytes can
-				 * be lost while transferring them from the
-				 * memory to the I2C interface.
-				 */
-
-				if (dev->rev <= OMAP_I2C_REV_ON_3430) {
-						while (!(stat & OMAP_I2C_STAT_XUDF)) {
-							if (stat & (OMAP_I2C_STAT_NACK | OMAP_I2C_STAT_AL)) {
-								omap_i2c_ack_stat(dev, stat & (OMAP_I2C_STAT_XRDY | OMAP_I2C_STAT_XDR));
-								err |= OMAP_I2C_STAT_XUDF;
-								goto complete;
-							}
-							cpu_relax();
-							stat = omap_i2c_read_reg(dev, OMAP_I2C_STAT_REG);
-						}
-				}
+				if ((dev->rev <= OMAP_I2C_REV_ON_3430) &&
+				    errata_omap3_1p153(dev, &stat, &err))
+					goto complete;
 
 				omap_i2c_write_reg(dev, OMAP_I2C_DATA_REG, w);
 			}
-- 
1.6.3.3

[APPLIED] [PATCH v2 1/2] omap i2c: make errata 1.153 workaround a separate

[Tony Lindgren]

This patch has been applied to the linux-omap
by youw fwiendly patch wobot.

Branch in linux-omap: i2c-omap-for-ben

Initial commit ID (Likely to change): 36d6678b28d8a46cebf3d3adeb9d297e86a18d95

PatchWorks
http://patchwork.kernel.org/patch/88171/

Git (Likely to change, and takes a while to get mirrored)
http://git.kernel.org/?p=linux/kernel/git/tmlind/linux-omap-2.6.git;a=commit;h=36d6678b28d8a46cebf3d3adeb9d297e86a18d95

OMAP3 I2C driver timing problem with multiple messages transfer

[Weng, Wending]

Hi All,

        Some I2C related protocols as SMBus, each command is composed of multiple I2C messages with timing constraint, the delay between two messages must be less than 10 ms, otherwise, the SMBus device will be timeout, and the command fails. As the function wait_for_completion_timeout is used to wait the end of a message, it doesn't guarante the next message will be sent out in 10 ms after the previous message is completed, so the SMBus command fails very often. I developped a patch for fixing this timing problem(2.6.32.RC5), is there anybody willing to review and to incorporate the patch into the furture release?

Wending
Software Engineer
at Rheinmetall Canada

RE: OMAP3 I2C driver timing problem with multiple messages transfer

[Sonasath, Moiz]

Wending,

> -----Original Message-----
> From: linux-omap-owner@vger.kernel.org [mailto:linux-omap-
> owner@vger.kernel.org] On Behalf Of Weng, Wending
> Sent: Wednesday, December 16, 2009 8:54 AM
> To: linux-omap@vger.kernel.org; linux-i2c@vger.kernel.org
> Subject: OMAP3 I2C driver timing problem with multiple messages transfer
> 
> Hi All,
> 
>         Some I2C related protocols as SMBus, each command is composed of
> multiple I2C messages with timing constraint, the delay between two
> messages must be less than 10 ms, otherwise, the SMBus device will be
> timeout, and the command fails. As the function
> wait_for_completion_timeout is used to wait the end of a message, it
> doesn't guarante the next message will be sent out in 10 ms after the
> previous message is completed, so the SMBus command fails very often. I
> developped a patch for fixing this timing problem(2.6.32.RC5), is there
> anybody willing to review and to incorporate the patch into the furture
> release?
> 

Yes, do send out the patch as I am sure people would be interested to review your patch as well as incorporate it if it's needed. Please send out the patch to these mailing lists.

> Wending
> Software Engineer
> at Rheinmetall Canada
> --
> To unsubscribe from this list: send the line "unsubscribe linux-omap" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

RE: OMAP3 I2C driver timing problem with multiple messages transfer

[Weng, Wending]

[-- Attachment #1: Type: text/plain, Size: 2039 bytes --]

Hi All,

        The atached file is the patch for the problem.

Wending

>-----Original Message-----
>From: Sonasath, Moiz [mailto:m-sonasath-l0cyMroinI0@public.gmane.org]
>Sent: December 16, 2009 10:58 AM
>To: Weng, Wending; linux-omap-u79uwXL29TY76Z2rM5mHXA@public.gmane.org; linux-i2c-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
>Subject: RE: OMAP3 I2C driver timing problem with multiple messages
>transfer
>
>
>Wending,
>
>> -----Original Message-----
>> From: linux-omap-owner-u79uwXL29TY76Z2rM5mHXA@public.gmane.org [mailto:linux-omap-
>> owner-u79uwXL29TY76Z2rM5mHXA@public.gmane.org] On Behalf Of Weng, Wending
>> Sent: Wednesday, December 16, 2009 8:54 AM
>> To: linux-omap-u79uwXL29TY76Z2rM5mHXA@public.gmane.org; linux-i2c-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
>> Subject: OMAP3 I2C driver timing problem with multiple messages transfer
>>
>> Hi All,
>>
>>         Some I2C related protocols as SMBus, each command is composed of
>> multiple I2C messages with timing constraint, the delay between two
>> messages must be less than 10 ms, otherwise, the SMBus device will be
>> timeout, and the command fails. As the function
>> wait_for_completion_timeout is used to wait the end of a message, it
>> doesn't guarante the next message will be sent out in 10 ms after the
>> previous message is completed, so the SMBus command fails very often. I
>> developped a patch for fixing this timing problem(2.6.32.RC5), is there
>> anybody willing to review and to incorporate the patch into the furture
>> release?
>>

>Yes, do send out the patch as I am sure people would be interested to review your patch as well as incorporate it if it's
>needed. Please send out the patch to these mailing lists.

>> Wending
>> Software Engineer
>> at Rheinmetall Canada
>> --
>> To unsubscribe from this list: send the line "unsubscribe linux-omap" in
>> the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
>> More majordomo info at  http://vger.kernel.org/majordomo-info.html

[-- Attachment #2: i2c.patch --]
[-- Type: application/octet-stream, Size: 17101 bytes --]

--- 2.6.32.rc5_ori/drivers/i2c/busses/i2c-omap.c	2009-10-30 10:20:51.000000000 -0400
+++ 2.6.32.rc5_new/drivers/i2c/busses/i2c-omap.c	2009-12-14 10:46:01.000000000 -0500
@@ -38,6 +38,15 @@
 #include <linux/clk.h>
 #include <linux/io.h>
 
+/* Hack to enable zero length transfers and smbus quick until clean fix
+   is available */
+#define OMAP_HACK
+
+/* timeout waiting for the controller to respond */
+#define OMAP_I2C_TIMEOUT (msecs_to_jiffies(1000))
+
+
+
 /* I2C controller revisions */
 #define OMAP_I2C_REV_2			0x20
 
@@ -167,6 +176,11 @@
 	struct resource		*ioarea;
 	u32			speed;		/* Speed of bus in Khz */
 	u16			cmd_err;
+
+	u16                     cur_msg;
+	u16                     nbr_msg;
+	struct i2c_msg          *msgs;
+
 	u8			*buf;
 	size_t			buf_len;
 	struct i2c_adapter	adapter;
@@ -436,134 +450,142 @@
 /*
  * Low level master read/write transaction.
  */
-static int omap_i2c_xfer_msg(struct i2c_adapter *adap,
+static void omap_i2c_xfer_msg(struct i2c_adapter *adap,
 			     struct i2c_msg *msg, int stop)
 {
 	struct omap_i2c_dev *dev = i2c_get_adapdata(adap);
-	int r;
+
+#ifdef OMAP_HACK
+        u8 zero_byte = 0;
+#endif
+
 	u16 w;
 
 	dev_dbg(dev->dev, "addr: 0x%04x, len: %d, flags: 0x%x, stop: %d\n",
 		msg->addr, msg->len, msg->flags, stop);
 
-	if (msg->len == 0)
-		return -EINVAL;
+#ifndef OMAP_HACK
+        if (msg->len == 0)
+                return -EINVAL;
+
+        omap_i2c_write_reg(dev, OMAP_I2C_SA_REG, msg->addr);
+
+        /* REVISIT: Could the STB bit of I2C_CON be used with probing? */
+        dev->buf = msg->buf;
+        dev->buf_len = msg->len;
+#else
 
-	omap_i2c_write_reg(dev, OMAP_I2C_SA_REG, msg->addr);
+        omap_i2c_write_reg(dev, OMAP_I2C_SA_REG, msg->addr);
+        /* REVISIT: Remove this hack when we can get I2C chips from board-*.c
+         *          files
+         * Sigh, seems we can't do zero length transactions. Thus, we
+         * can't probe for devices w/o actually sending/receiving at least
+         * a single byte. So we'll set count to 1 for the zero length
+         * transaction case and hope we don't cause grief for some
+         * arbitrary device due to random byte write/read during
+         * probes.
+         */
+        if (msg->len == 0) {
+                dev->buf = &zero_byte;
+                dev->buf_len = 1;
+        } else {
+                dev->buf = msg->buf;
+                dev->buf_len = msg->len;
+        }
+#endif
 
-	/* REVISIT: Could the STB bit of I2C_CON be used with probing? */
-	dev->buf = msg->buf;
-	dev->buf_len = msg->len;
 
 	omap_i2c_write_reg(dev, OMAP_I2C_CNT_REG, dev->buf_len);
 
-	/* Clear the FIFO Buffers */
-	w = omap_i2c_read_reg(dev, OMAP_I2C_BUF_REG);
-	w |= OMAP_I2C_BUF_RXFIF_CLR | OMAP_I2C_BUF_TXFIF_CLR;
-	omap_i2c_write_reg(dev, OMAP_I2C_BUF_REG, w);
+        /* Clear the FIFO Buffers */
+        w = omap_i2c_read_reg(dev, OMAP_I2C_BUF_REG);
+        w |= OMAP_I2C_BUF_RXFIF_CLR | OMAP_I2C_BUF_TXFIF_CLR;
+        omap_i2c_write_reg(dev, OMAP_I2C_BUF_REG, w);
+
+        dev->cmd_err = 0;
+
+        w = OMAP_I2C_CON_EN | OMAP_I2C_CON_MST | OMAP_I2C_CON_STT;
+
+        /* High speed configuration */
+        if (dev->speed > 400)
+                w |= OMAP_I2C_CON_OPMODE_HS;
+
+        if (msg->flags & I2C_M_TEN)
+                w |= OMAP_I2C_CON_XA;
+        if (!(msg->flags & I2C_M_RD))
+                w |= OMAP_I2C_CON_TRX;
+
+        if (!dev->b_hw && stop)
+                w |= OMAP_I2C_CON_STP;
+
+        omap_i2c_write_reg(dev, OMAP_I2C_CON_REG, w);
+
+        if (dev->b_hw && stop) {
+                /* H/w behavior: dont write stt and stp together.. */
+                while (omap_i2c_read_reg(dev, OMAP_I2C_CON_REG) & OMAP_I2C_CON_STT) {
+                        /* Dont do anything - this will come in a couple of loops at max*/
+                }
+                w |= OMAP_I2C_CON_STP;
+                w &= ~OMAP_I2C_CON_STT;
+                omap_i2c_write_reg(dev, OMAP_I2C_CON_REG, w);
+        }
+}
 
-	init_completion(&dev->cmd_complete);
-	dev->cmd_err = 0;
+/*
+ * Prepare controller for a transaction and call omap_i2c_xfer_msg
+ * to do the work during IRQ processing.
+ */
+static int
+omap_i2c_xfer(struct i2c_adapter *adap, struct i2c_msg msgs[], int num)
+{
+	struct omap_i2c_dev *dev = i2c_get_adapdata(adap);
+	int r;
+	u16 w;
 
-	w = OMAP_I2C_CON_EN | OMAP_I2C_CON_MST | OMAP_I2C_CON_STT;
+	omap_i2c_unidle(dev);
 
-	/* High speed configuration */
-	if (dev->speed > 400)
-		w |= OMAP_I2C_CON_OPMODE_HS;
-
-	if (msg->flags & I2C_M_TEN)
-		w |= OMAP_I2C_CON_XA;
-	if (!(msg->flags & I2C_M_RD))
-		w |= OMAP_I2C_CON_TRX;
-
-	if (!dev->b_hw && stop)
-		w |= OMAP_I2C_CON_STP;
-
-	omap_i2c_write_reg(dev, OMAP_I2C_CON_REG, w);
-
-	/*
-	 * Don't write stt and stp together on some hardware.
-	 */
-	if (dev->b_hw && stop) {
-		unsigned long delay = jiffies + OMAP_I2C_TIMEOUT;
-		u16 con = omap_i2c_read_reg(dev, OMAP_I2C_CON_REG);
-		while (con & OMAP_I2C_CON_STT) {
-			con = omap_i2c_read_reg(dev, OMAP_I2C_CON_REG);
-
-			/* Let the user know if i2c is in a bad state */
-			if (time_after(jiffies, delay)) {
-				dev_err(dev->dev, "controller timed out "
-				"waiting for start condition to finish\n");
-				return -ETIMEDOUT;
-			}
-			cpu_relax();
-		}
 
-		w |= OMAP_I2C_CON_STP;
-		w &= ~OMAP_I2C_CON_STT;
-		omap_i2c_write_reg(dev, OMAP_I2C_CON_REG, w);
-	}
+	r = omap_i2c_wait_for_bb(dev);
+	if (r < 0)
+		goto out;
+
+	dev->cur_msg = 0;
+	dev->nbr_msg = num;
+	dev->msgs = msgs;
+
+	init_completion(&dev->cmd_complete);
+
+	omap_i2c_xfer_msg(adap, &msgs[0], num == 1);
+
 
-	/*
-	 * REVISIT: We should abort the transfer on signals, but the bus goes
-	 * into arbitration and we're currently unable to recover from it.
-	 */
 	r = wait_for_completion_timeout(&dev->cmd_complete,
 					OMAP_I2C_TIMEOUT);
-	dev->buf_len = 0;
-	if (r < 0)
-		return r;
 	if (r == 0) {
 		dev_err(dev->dev, "controller timed out\n");
 		omap_i2c_init(dev);
-		return -ETIMEDOUT;
+		r = -ETIMEDOUT;
 	}
-
-	if (likely(!dev->cmd_err))
-		return 0;
+	else
+		if (likely(!dev->cmd_err))
+			r = 0;
 
 	/* We have an error */
 	if (dev->cmd_err & (OMAP_I2C_STAT_AL | OMAP_I2C_STAT_ROVR |
 			    OMAP_I2C_STAT_XUDF)) {
 		omap_i2c_init(dev);
-		return -EIO;
+		r = -EIO;
 	}
 
 	if (dev->cmd_err & OMAP_I2C_STAT_NACK) {
-		if (msg->flags & I2C_M_IGNORE_NAK)
-			return 0;
-		if (stop) {
+               if (msgs[0].flags & I2C_M_IGNORE_NAK)
+                       r = 0;
+               else {
+
 			w = omap_i2c_read_reg(dev, OMAP_I2C_CON_REG);
 			w |= OMAP_I2C_CON_STP;
 			omap_i2c_write_reg(dev, OMAP_I2C_CON_REG, w);
+			r = -EREMOTEIO;
 		}
-		return -EREMOTEIO;
-	}
-	return -EIO;
-}
-
-
-/*
- * Prepare controller for a transaction and call omap_i2c_xfer_msg
- * to do the work during IRQ processing.
- */
-static int
-omap_i2c_xfer(struct i2c_adapter *adap, struct i2c_msg msgs[], int num)
-{
-	struct omap_i2c_dev *dev = i2c_get_adapdata(adap);
-	int i;
-	int r;
-
-	omap_i2c_unidle(dev);
-
-	r = omap_i2c_wait_for_bb(dev);
-	if (r < 0)
-		goto out;
-
-	for (i = 0; i < num; i++) {
-		r = omap_i2c_xfer_msg(adap, &msgs[i], (i == (num - 1)));
-		if (r != 0)
-			break;
 	}
 
 	if (r == 0)
@@ -576,7 +598,11 @@
 static u32
 omap_i2c_func(struct i2c_adapter *adap)
 {
+#ifndef OMAP_HACK
 	return I2C_FUNC_I2C | (I2C_FUNC_SMBUS_EMUL & ~I2C_FUNC_SMBUS_QUICK);
+#else
+	return I2C_FUNC_I2C | I2C_FUNC_SMBUS_EMUL;
+#endif
 }
 
 static inline void
@@ -659,7 +685,7 @@
 	struct omap_i2c_dev *dev = dev_id;
 	u16 bits;
 	u16 stat, w;
-	int err, count = 0;
+	int count = 0;
 
 	if (dev->idle)
 		return IRQ_NONE;
@@ -672,7 +698,6 @@
 			break;
 		}
 
-		err = 0;
 complete:
 		/*
 		 * Ack the stat in one go, but [R/X]DR and [R/X]RDY should be
@@ -683,134 +708,110 @@
 				~(OMAP_I2C_STAT_RRDY | OMAP_I2C_STAT_RDR |
 				OMAP_I2C_STAT_XRDY | OMAP_I2C_STAT_XDR));
 
-		if (stat & OMAP_I2C_STAT_NACK) {
-			err |= OMAP_I2C_STAT_NACK;
-			omap_i2c_write_reg(dev, OMAP_I2C_CON_REG,
-					   OMAP_I2C_CON_STP);
-		}
-		if (stat & OMAP_I2C_STAT_AL) {
-			dev_err(dev->dev, "Arbitration lost\n");
-			err |= OMAP_I2C_STAT_AL;
-		}
-		if (stat & (OMAP_I2C_STAT_ARDY | OMAP_I2C_STAT_NACK |
-					OMAP_I2C_STAT_AL)) {
-			omap_i2c_ack_stat(dev, stat &
-				(OMAP_I2C_STAT_RRDY | OMAP_I2C_STAT_RDR |
-				OMAP_I2C_STAT_XRDY | OMAP_I2C_STAT_XDR));
-			omap_i2c_complete_cmd(dev, err);
-			return IRQ_HANDLED;
-		}
-		if (stat & (OMAP_I2C_STAT_RRDY | OMAP_I2C_STAT_RDR)) {
-			u8 num_bytes = 1;
-			if (dev->fifo_size) {
-				if (stat & OMAP_I2C_STAT_RRDY)
-					num_bytes = dev->fifo_size;
-				else    /* read RXSTAT on RDR interrupt */
-					num_bytes = (omap_i2c_read_reg(dev,
-							OMAP_I2C_BUFSTAT_REG)
-							>> 8) & 0x3F;
-			}
-			while (num_bytes) {
-				num_bytes--;
-				w = omap_i2c_read_reg(dev, OMAP_I2C_DATA_REG);
-				if (dev->buf_len) {
-					*dev->buf++ = w;
-					dev->buf_len--;
-					/* Data reg from 2430 is 8 bit wide */
-					if (!cpu_is_omap2430() &&
-							!cpu_is_omap34xx()) {
-						if (dev->buf_len) {
-							*dev->buf++ = w >> 8;
-							dev->buf_len--;
-						}
-					}
-				} else {
-					if (stat & OMAP_I2C_STAT_RRDY)
-						dev_err(dev->dev,
-							"RRDY IRQ while no data"
-								" requested\n");
-					if (stat & OMAP_I2C_STAT_RDR)
-						dev_err(dev->dev,
-							"RDR IRQ while no data"
-								" requested\n");
-					break;
-				}
-			}
-			omap_i2c_ack_stat(dev,
-				stat & (OMAP_I2C_STAT_RRDY | OMAP_I2C_STAT_RDR));
-			continue;
-		}
-		if (stat & (OMAP_I2C_STAT_XRDY | OMAP_I2C_STAT_XDR)) {
-			u8 num_bytes = 1;
-			if (dev->fifo_size) {
-				if (stat & OMAP_I2C_STAT_XRDY)
-					num_bytes = dev->fifo_size;
-				else    /* read TXSTAT on XDR interrupt */
-					num_bytes = omap_i2c_read_reg(dev,
-							OMAP_I2C_BUFSTAT_REG)
-							& 0x3F;
-			}
-			while (num_bytes) {
-				num_bytes--;
-				w = 0;
-				if (dev->buf_len) {
-					w = *dev->buf++;
-					dev->buf_len--;
-					/* Data reg from  2430 is 8 bit wide */
-					if (!cpu_is_omap2430() &&
-							!cpu_is_omap34xx()) {
-						if (dev->buf_len) {
-							w |= *dev->buf++ << 8;
-							dev->buf_len--;
-						}
-					}
-				} else {
-					if (stat & OMAP_I2C_STAT_XRDY)
-						dev_err(dev->dev,
-							"XRDY IRQ while no "
-							"data to send\n");
-					if (stat & OMAP_I2C_STAT_XDR)
-						dev_err(dev->dev,
-							"XDR IRQ while no "
-							"data to send\n");
-					break;
-				}
-
-				/*
-				 * OMAP3430 Errata 1.153: When an XRDY/XDR
-				 * is hit, wait for XUDF before writing data
-				 * to DATA_REG. Otherwise some data bytes can
-				 * be lost while transferring them from the
-				 * memory to the I2C interface.
-				 */
-
-				if (dev->rev <= OMAP_I2C_REV_ON_3430) {
-						while (!(stat & OMAP_I2C_STAT_XUDF)) {
-							if (stat & (OMAP_I2C_STAT_NACK | OMAP_I2C_STAT_AL)) {
-								omap_i2c_ack_stat(dev, stat & (OMAP_I2C_STAT_XRDY | OMAP_I2C_STAT_XDR));
-								err |= OMAP_I2C_STAT_XUDF;
-								goto complete;
-							}
-							cpu_relax();
-							stat = omap_i2c_read_reg(dev, OMAP_I2C_STAT_REG);
-						}
-				}
+		if (stat & OMAP_I2C_STAT_ARDY) {
+			// the bus is avaiable for next msg.
+			dev->cur_msg ++;
+
+			// Stop if the last msg is sent or errors
+			if ((dev->cur_msg >= dev->nbr_msg) || dev->cmd_err)
+				omap_i2c_complete_cmd(dev, 0);
+			else
+				omap_i2c_xfer_msg(&dev->adapter,
+				&dev->msgs[dev->cur_msg], dev->cur_msg == (dev->nbr_msg - 1));
+			//omap_i2c_complete_cmd(dev, 0);
 
-				omap_i2c_write_reg(dev, OMAP_I2C_DATA_REG, w);
-			}
-			omap_i2c_ack_stat(dev,
-				stat & (OMAP_I2C_STAT_XRDY | OMAP_I2C_STAT_XDR));
 			continue;
 		}
-		if (stat & OMAP_I2C_STAT_ROVR) {
-			dev_err(dev->dev, "Receive overrun\n");
-			dev->cmd_err |= OMAP_I2C_STAT_ROVR;
-		}
-		if (stat & OMAP_I2C_STAT_XUDF) {
-			dev_err(dev->dev, "Transmit underflow\n");
-			dev->cmd_err |= OMAP_I2C_STAT_XUDF;
-		}
-	}
+
+                if (stat & (OMAP_I2C_STAT_RRDY | OMAP_I2C_STAT_RDR)) {
+                         u8 num_bytes = 1;
+                         if (dev->fifo_size) {
+                                 num_bytes = (stat & OMAP_I2C_STAT_RRDY) ? dev->fifo_size :
+                                                 omap_i2c_read_reg(dev, OMAP_I2C_BUFSTAT_REG);
+                         }
+                         while (num_bytes) {
+                                 num_bytes--;
+                                 w = omap_i2c_read_reg(dev, OMAP_I2C_DATA_REG);
+                                 if (dev->buf_len) {
+                                         *dev->buf++ = w;
+                                         dev->buf_len--;
+                                         /* data reg from 2430 is 8 bit wide */
+                                         if (!cpu_is_omap2430() &&
+                                                         !cpu_is_omap34xx()) {
+                                                 if (dev->buf_len) {
+                                                         *dev->buf++ = w >> 8;
+                                                         dev->buf_len--;
+                                                 }
+                                         }
+                                 } else {
+                                         if (stat & OMAP_I2C_STAT_RRDY)
+                                                 dev_err(dev->dev, "RRDY IRQ while no data"
+                                                                 "requested\n");
+                                         if (stat & OMAP_I2C_STAT_RDR)
+                                                 dev_err(dev->dev, "RDR IRQ while no data"
+                                                                 "requested\n");
+                                         break;
+                                 }
+                         }
+                         omap_i2c_ack_stat(dev, stat & (OMAP_I2C_STAT_RRDY | OMAP_I2C_STAT_RDR));
+                         continue;
+                }
+
+                if (stat & (OMAP_I2C_STAT_XRDY | OMAP_I2C_STAT_XDR)) {
+                        u8 num_bytes = 1;
+                        if (dev->fifo_size) {
+                                num_bytes = (stat & OMAP_I2C_STAT_XRDY) ? dev->fifo_size :
+                                                omap_i2c_read_reg(dev, OMAP_I2C_BUFSTAT_REG);
+                        }
+                        while (num_bytes) {
+                                num_bytes--;
+                                w = 0;
+                                if (dev->buf_len) {
+                                        w = *dev->buf++;
+                                        dev->buf_len--;
+                                        /* data reg from 2430 is 8 bit wide */
+                                        if (!cpu_is_omap2430() &&
+                                                        !cpu_is_omap34xx()) {
+                                                if (dev->buf_len) {
+                                                        w |= *dev->buf++ << 8;
+                                                        dev->buf_len--;
+                                                }
+                                        }
+                                } else {
+                                        if (stat & OMAP_I2C_STAT_XRDY)
+                                                dev_err(dev->dev, "XRDY IRQ while no"
+                                                                "data to send\n");
+                                        if (stat & OMAP_I2C_STAT_XDR)
+                                                dev_err(dev->dev, "XDR IRQ while no"
+                                                                "data to send\n");
+                                        break;
+                                }
+                                omap_i2c_write_reg(dev, OMAP_I2C_DATA_REG, w);
+                        }
+                        omap_i2c_ack_stat(dev, stat &
+                                        (OMAP_I2C_STAT_XRDY |
+                                                 OMAP_I2C_STAT_XDR));
+                        continue;
+                }
+                if (stat & OMAP_I2C_STAT_ROVR) {
+                        dev_err(dev->dev, "Receive overrun\n");
+                        dev->cmd_err |= OMAP_I2C_STAT_ROVR;
+                }
+                if (stat & OMAP_I2C_STAT_XUDF) {
+                        dev_err(dev->dev, "Transmit overflow\n");
+                        dev->cmd_err |= OMAP_I2C_STAT_XUDF;
+                }
+                if (stat & OMAP_I2C_STAT_NACK) {
+                        dev_err(dev->dev, "No Acknowledgement\n");
+                        omap_i2c_complete_cmd(dev, OMAP_I2C_STAT_NACK);
+                        omap_i2c_write_reg(dev, OMAP_I2C_CON_REG,
+                                           OMAP_I2C_CON_STP);
+                }
+                if (stat & OMAP_I2C_STAT_AL) {
+                        dev_err(dev->dev, "Arbitration lost\n");
+                        omap_i2c_complete_cmd(dev, OMAP_I2C_STAT_AL);
+                }
+        }
 
 	return count ? IRQ_HANDLED : IRQ_NONE;
 }

[PATCH 1/2] omap i2c: make errata 1.153 workaround a separate function

[Alexander Shishkin]

From: Alexander Shishkin <ext-alexander.shishkin@nokia.com>

This is to avoid insanely long lines and levels of indentation.

Signed-off-by: Alexander Shishkin <ext-alexander.shishkin@nokia.com>
---
 drivers/i2c/busses/i2c-omap.c |   43 ++++++++++++++++++++++------------------
 1 files changed, 24 insertions(+), 19 deletions(-)

diff --git a/drivers/i2c/busses/i2c-omap.c b/drivers/i2c/busses/i2c-omap.c
index 75bf3ad..ad8242a 100644
--- a/drivers/i2c/busses/i2c-omap.c
+++ b/drivers/i2c/busses/i2c-omap.c
@@ -671,6 +671,27 @@ omap_i2c_rev1_isr(int this_irq, void *dev_id)
 #define omap_i2c_rev1_isr		NULL
 #endif
 
+/*
+ * OMAP3430 Errata 1.153: When an XRDY/XDR is hit, wait for XUDF before writing
+ * data to DATA_REG. Otherwise some data bytes can be lost while transferring
+ * them from the memory to the I2C interface.
+ */
+static int omap3430_workaround(struct omap_i2c_dev *dev, u16 *stat, int *err)
+{
+	while (!(*stat & OMAP_I2C_STAT_XUDF)) {
+		if (*stat & (OMAP_I2C_STAT_NACK | OMAP_I2C_STAT_AL)) {
+			omap_i2c_ack_stat(dev, *stat & (OMAP_I2C_STAT_XRDY |
+							OMAP_I2C_STAT_XDR));
+			*err |= OMAP_I2C_STAT_XUDF;
+			return -1;
+		}
+		cpu_relax();
+		*stat = omap_i2c_read_reg(dev, OMAP_I2C_STAT_REG);
+	}
+
+	return 0;
+}
+
 static irqreturn_t
 omap_i2c_isr(int this_irq, void *dev_id)
 {
@@ -794,25 +815,9 @@ complete:
 					break;
 				}
 
-				/*
-				 * OMAP3430 Errata 1.153: When an XRDY/XDR
-				 * is hit, wait for XUDF before writing data
-				 * to DATA_REG. Otherwise some data bytes can
-				 * be lost while transferring them from the
-				 * memory to the I2C interface.
-				 */
-
-				if (dev->rev <= OMAP_I2C_REV_ON_3430) {
-						while (!(stat & OMAP_I2C_STAT_XUDF)) {
-							if (stat & (OMAP_I2C_STAT_NACK | OMAP_I2C_STAT_AL)) {
-								omap_i2c_ack_stat(dev, stat & (OMAP_I2C_STAT_XRDY | OMAP_I2C_STAT_XDR));
-								err |= OMAP_I2C_STAT_XUDF;
-								goto complete;
-							}
-							cpu_relax();
-							stat = omap_i2c_read_reg(dev, OMAP_I2C_STAT_REG);
-						}
-				}
+				if (dev->rev <= OMAP_I2C_REV_ON_3430 &&
+				    omap3430_workaround(dev, &stat, &err))
+					goto complete;
 
 				omap_i2c_write_reg(dev, OMAP_I2C_DATA_REG, w);
 			}
-- 
1.6.3.3

[PATCH 2/2] omap i2c: add a timeout to the busy waiting

[Alexander Shishkin]

The errata 1.153 workaround is busy waiting on XUDF bit in interrupt
context, which may lead to kernel hangs. The problem can be reproduced
by running the bus with wrong (too high) speed.

Signed-off-by: Alexander Shishkin <virtuoso-0lOfPCoBze7YtjvyW6yDsg@public.gmane.org>
CC: linux-i2c-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
CC: linux-omap-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
---
 drivers/i2c/busses/i2c-omap.c |    7 ++++++-
 1 files changed, 6 insertions(+), 1 deletions(-)

diff --git a/drivers/i2c/busses/i2c-omap.c b/drivers/i2c/busses/i2c-omap.c
index ad8242a..b474c20 100644
--- a/drivers/i2c/busses/i2c-omap.c
+++ b/drivers/i2c/busses/i2c-omap.c
@@ -678,7 +678,9 @@ omap_i2c_rev1_isr(int this_irq, void *dev_id)
  */
 static int omap3430_workaround(struct omap_i2c_dev *dev, u16 *stat, int *err)
 {
-	while (!(*stat & OMAP_I2C_STAT_XUDF)) {
+	unsigned long timeout = 10000;
+
+	while (!(*stat & OMAP_I2C_STAT_XUDF && --timeout)) {
 		if (*stat & (OMAP_I2C_STAT_NACK | OMAP_I2C_STAT_AL)) {
 			omap_i2c_ack_stat(dev, *stat & (OMAP_I2C_STAT_XRDY |
 							OMAP_I2C_STAT_XDR));
@@ -689,6 +691,9 @@ static int omap3430_workaround(struct omap_i2c_dev *dev, u16 *stat, int *err)
 		*stat = omap_i2c_read_reg(dev, OMAP_I2C_STAT_REG);
 	}
 
+	if (!timeout)
+		dev_err(dev->dev, "timeout waiting on XUDF bit\n");
+
 	return 0;
 }
 
-- 
1.6.3.3