diff mbox

thermal: rockchip: fix handling of invalid readings

Message ID 20150807205923.GA33949@dtor-ws (mailing list archive)
State New, archived
Headers show

Commit Message

Dmitry Torokhov Aug. 7, 2015, 8:59 p.m. UTC
We attempted to signal invalid code by returning -EAGAIN from
rk_tsadcv2_code_to_temp(), unfortunately the return value was stuffed
directly into the temperature pointer, potentially confusing upper
layers with temperature of -EINVAL.

Let's split temperature from error/success indicator to avoid such
confusion.

Also change the way we scan the temperature table to start with the 2nd
element so that we do not need to worry that we may reference out of
bounds element while doing binary search and keep checking that we end
up with 'mid' equal to 0 (since we are looking for the temperature that
would fall into interval between the 'mid' and 'mid - 1') .

Signed-off-by: Dmitry Torokhov <dtor@chromium.org>
---
 drivers/thermal/rockchip_thermal.c | 28 +++++++++++++---------------
 1 file changed, 13 insertions(+), 15 deletions(-)

Comments

Dmitry Torokhov Aug. 10, 2015, 7:59 a.m. UTC | #1
Hi Caesar,

On Sun, Aug 9, 2015 at 11:30 PM, Caesar Wang <wxt@rock-chips.com> wrote:
> Dear Dmitry,
>
> Thanks your patch.
>
> The code looks like fine,but I don't think the TS-ADC will work on rk3288
> SoC.

What do you mean? It seems to work when I ran it...

>
>
> ? 2015?08?08? 04:59, Dmitry Torokhov ??:
>>
>> We attempted to signal invalid code by returning -EAGAIN from
>> rk_tsadcv2_code_to_temp(), unfortunately the return value was stuffed
>> directly into the temperature pointer, potentially confusing upper
>> layers with temperature of -EINVAL.
>>
>> Let's split temperature from error/success indicator to avoid such
>> confusion.
>>
>> Also change the way we scan the temperature table to start with the 2nd
>> element so that we do not need to worry that we may reference out of
>> bounds element while doing binary search and keep checking that we end
>> up with 'mid' equal to 0 (since we are looking for the temperature that
>> would fall into interval between the 'mid' and 'mid - 1') .
>>
>> Signed-off-by: Dmitry Torokhov <dtor@chromium.org>
>> ---
>>   drivers/thermal/rockchip_thermal.c | 28 +++++++++++++---------------
>>   1 file changed, 13 insertions(+), 15 deletions(-)
>>
>> diff --git a/drivers/thermal/rockchip_thermal.c
>> b/drivers/thermal/rockchip_thermal.c
>> index c89ffb2..93ee307 100644
>> --- a/drivers/thermal/rockchip_thermal.c
>> +++ b/drivers/thermal/rockchip_thermal.c
>> @@ -124,7 +124,7 @@ struct rockchip_thermal_data {
>>   #define TSADCV2_AUTO_PERIOD_HT_TIME           50  /* msec */
>>     struct tsadc_table {
>> -       unsigned long code;
>> +       u32 code;
>>         long temp;
>>   };
>>   @@ -164,7 +164,6 @@ static const struct tsadc_table v2_code_table[] = {
>>         {3452, 115000},
>>         {3437, 120000},
>>         {3421, 125000},
>> -       {0, 125000},
>>   };
>>     static u32 rk_tsadcv2_temp_to_code(long temp)
>> @@ -191,19 +190,21 @@ static u32 rk_tsadcv2_temp_to_code(long temp)
>>         return 0;
>>   }
>>   -static int rk_tsadcv2_code_to_temp(u32 code)
>> +static int rk_tsadcv2_code_to_temp(u32 code, int *temp)
>>   {
>> -       unsigned int low = 0;
>> +       unsigned int low = 1;
>>         unsigned int high = ARRAY_SIZE(v2_code_table) - 1;
>>         unsigned int mid = (low + high) / 2;
>>         unsigned int num;
>>         unsigned long denom;
>>   -     /* Invalid code, return -EAGAIN */
>> -       if (code > TSADCV2_DATA_MASK)
>> -               return -EAGAIN;
>> +       BUILD_BUG_ON(ARRAY_SIZE(v2_code_table) < 2);
>>   -     while (low <= high && mid) {
>> +       code &= TSADCV2_DATA_MASK;
>> +       if (code < v2_code_table[high].code)
>> +               return -EAGAIN;         /* Incorrect reading */
>> +
>> +       while (low <= high) {
>>                 if (code >= v2_code_table[mid].code &&
>>                     code < v2_code_table[mid - 1].code)
>>                         break;
>> @@ -223,7 +224,9 @@ static int rk_tsadcv2_code_to_temp(u32 code)
>>         num = v2_code_table[mid].temp - v2_code_table[mid - 1].temp;
>>         num *= v2_code_table[mid - 1].code - code;
>>         denom = v2_code_table[mid - 1].code - v2_code_table[mid].code;
>> -       return v2_code_table[mid - 1].temp + (num / denom);
>> +       *temp = v2_code_table[mid - 1].temp + (num / denom);
>> +
>> +       return 0;
>>   }
>>     /**
>> @@ -281,14 +284,9 @@ static int rk_tsadcv2_get_temp(int chn, void __iomem
>> *regs, int *temp)
>>   {
>>         u32 val;
>>   -     /* the A/D value of the channel last conversion need some time */
>>         val = readl_relaxed(regs + TSADCV2_DATA(chn));
>> -       if (val == 0)
>> -               return -EAGAIN;
>> -
>
>
> if we reserve the above code, that will get the ADC value.

But if we pass  0 into rk_tsadcv2_code_to_temp() it will trigger:

>> +       if (code < v2_code_table[high].code)
>> +               return -EAGAIN;         /* Incorrect reading */

and still return -EAGAIN, as it was. There is no behavior change as
far as I can see.

>
>
>> -       *temp = rk_tsadcv2_code_to_temp(val);
>>   -     return 0;
>> +       return rk_tsadcv2_code_to_temp(val, temp);
>>   }
>>     static void rk_tsadcv2_tshut_temp(int chn, void __iomem *regs, long
>> temp)
>
>

Thanks,
Dmitry
Caesar Wang Aug. 10, 2015, 10:22 a.m. UTC | #2
? 2015?08?10? 15:59, Dmitry Torokhov ??:
> Hi Caesar,
>
> On Sun, Aug 9, 2015 at 11:30 PM, Caesar Wang <wxt@rock-chips.com> wrote:
>> Dear Dmitry,
>>
>> Thanks your patch.
>>
>> The code looks like fine,but I don't think the TS-ADC will work on rk3288
>> SoC.
> What do you mean? It seems to work when I ran it...

Making a mistake.:-(

>>
>> ? 2015?08?08? 04:59, Dmitry Torokhov ??:
>>> We attempted to signal invalid code by returning -EAGAIN from
>>> rk_tsadcv2_code_to_temp(), unfortunately the return value was stuffed
>>> directly into the temperature pointer, potentially confusing upper
>>> layers with temperature of -EINVAL.
>>>
>>> Let's split temperature from error/success indicator to avoid such
>>> confusion.
>>>
>>> Also change the way we scan the temperature table to start with the 2nd
>>> element so that we do not need to worry that we may reference out of
>>> bounds element while doing binary search and keep checking that we end
>>> up with 'mid' equal to 0 (since we are looking for the temperature that
>>> would fall into interval between the 'mid' and 'mid - 1') .
>>>
>>> Signed-off-by: Dmitry Torokhov <dtor@chromium.org>
>>> ---
>>>    drivers/thermal/rockchip_thermal.c | 28 +++++++++++++---------------
>>>    1 file changed, 13 insertions(+), 15 deletions(-)
>>>
>>> diff --git a/drivers/thermal/rockchip_thermal.c
>>> b/drivers/thermal/rockchip_thermal.c
>>> index c89ffb2..93ee307 100644
>>> --- a/drivers/thermal/rockchip_thermal.c
>>> +++ b/drivers/thermal/rockchip_thermal.c
>>> @@ -124,7 +124,7 @@ struct rockchip_thermal_data {
>>>    #define TSADCV2_AUTO_PERIOD_HT_TIME           50  /* msec */
>>>      struct tsadc_table {
>>> -       unsigned long code;
>>> +       u32 code;
>>>          long temp;
>>>    };
>>>    @@ -164,7 +164,6 @@ static const struct tsadc_table v2_code_table[] = {
>>>          {3452, 115000},
>>>          {3437, 120000},
>>>          {3421, 125000},
>>> -       {0, 125000},
>>>    };
>>>      static u32 rk_tsadcv2_temp_to_code(long temp)
>>> @@ -191,19 +190,21 @@ static u32 rk_tsadcv2_temp_to_code(long temp)
>>>          return 0;
>>>    }
>>>    -static int rk_tsadcv2_code_to_temp(u32 code)
>>> +static int rk_tsadcv2_code_to_temp(u32 code, int *temp)

It's wired that can't git-am this patch.

I search the Eduardo'branch[0] for rockchip-thermal.c driver. The 
function is "static long rk_tsadcv2_code_to_temp(u32 code)"

Maybe i'm missing something, I verified this patch on my board but this 
fixed.

So you can free add it:
Tested-by: Caesar Wang <wxt@rock-chips.com>

[0]:
url = 
git://git.kernel.org/pub/scm/linux/kernel/git/evalenti/linux-soc-thermal.git



---
Thanks,
Caesar



>>>    {
>>> -       unsigned int low = 0;
>>> +       unsigned int low = 1;
>>>          unsigned int high = ARRAY_SIZE(v2_code_table) - 1;
>>>          unsigned int mid = (low + high) / 2;
>>>          unsigned int num;
>>>          unsigned long denom;
>>>    -     /* Invalid code, return -EAGAIN */
>>> -       if (code > TSADCV2_DATA_MASK)
>>> -               return -EAGAIN;
>>> +       BUILD_BUG_ON(ARRAY_SIZE(v2_code_table) < 2);
>>>    -     while (low <= high && mid) {
>>> +       code &= TSADCV2_DATA_MASK;
>>> +       if (code < v2_code_table[high].code)
>>> +               return -EAGAIN;         /* Incorrect reading */
>>> +
>>> +       while (low <= high) {
>>>                  if (code >= v2_code_table[mid].code &&
>>>                      code < v2_code_table[mid - 1].code)
>>>                          break;
>>> @@ -223,7 +224,9 @@ static int rk_tsadcv2_code_to_temp(u32 code)
>>>          num = v2_code_table[mid].temp - v2_code_table[mid - 1].temp;
>>>          num *= v2_code_table[mid - 1].code - code;
>>>          denom = v2_code_table[mid - 1].code - v2_code_table[mid].code;
>>> -       return v2_code_table[mid - 1].temp + (num / denom);
>>> +       *temp = v2_code_table[mid - 1].temp + (num / denom);
>>> +
>>> +       return 0;
>>>    }
>>>      /**
>>> @@ -281,14 +284,9 @@ static int rk_tsadcv2_get_temp(int chn, void __iomem
>>> *regs, int *temp)
>>>    {
>>>          u32 val;
>>>    -     /* the A/D value of the channel last conversion need some time */
>>>          val = readl_relaxed(regs + TSADCV2_DATA(chn));
>>> -       if (val == 0)
>>> -               return -EAGAIN;
>>> -
>>
>> if we reserve the above code, that will get the ADC value.
> But if we pass  0 into rk_tsadcv2_code_to_temp() it will trigger:
>
>>> +       if (code < v2_code_table[high].code)
>>> +               return -EAGAIN;         /* Incorrect reading */
> and still return -EAGAIN, as it was. There is no behavior change as
> far as I can see.
Yup, that's ok for me.

>>
>>> -       *temp = rk_tsadcv2_code_to_temp(val);
>>>    -     return 0;
>>> +       return rk_tsadcv2_code_to_temp(val, temp);
>>>    }
>>>      static void rk_tsadcv2_tshut_temp(int chn, void __iomem *regs, long
>>> temp)
>>
> Thanks,
> Dmitry
>
>
>
Caesar Wang Sept. 14, 2015, 11:46 a.m. UTC | #3
Hello Eduardo,

I guess you forgot to review/pick up this series patchs.

Another patch as follows:

[PATCH] thermal: rockhip: fix setting thermal shutdown polarity
https://patchwork.kernel.org/patch/6973131/

----
Thanks,
Caesar

? 2015?08?10? 18:22, Caesar Wang ??:
>
>
> ? 2015?08?10? 15:59, Dmitry Torokhov ??:
>> Hi Caesar,
>>
>> On Sun, Aug 9, 2015 at 11:30 PM, Caesar Wang <wxt@rock-chips.com> wrote:
>>> Dear Dmitry,
>>>
>>> Thanks your patch.
>>>
>>> The code looks like fine,but I don't think the TS-ADC will work on 
>>> rk3288
>>> SoC.
>> What do you mean? It seems to work when I ran it...
>
> Making a mistake.:-(
>
>>>
>>> ? 2015?08?08? 04:59, Dmitry Torokhov ??:
>>>> We attempted to signal invalid code by returning -EAGAIN from
>>>> rk_tsadcv2_code_to_temp(), unfortunately the return value was stuffed
>>>> directly into the temperature pointer, potentially confusing upper
>>>> layers with temperature of -EINVAL.
>>>>
>>>> Let's split temperature from error/success indicator to avoid such
>>>> confusion.
>>>>
>>>> Also change the way we scan the temperature table to start with the 
>>>> 2nd
>>>> element so that we do not need to worry that we may reference out of
>>>> bounds element while doing binary search and keep checking that we end
>>>> up with 'mid' equal to 0 (since we are looking for the temperature 
>>>> that
>>>> would fall into interval between the 'mid' and 'mid - 1') .
>>>>
>>>> Signed-off-by: Dmitry Torokhov <dtor@chromium.org>
>>>> ---
>>>>    drivers/thermal/rockchip_thermal.c | 28 
>>>> +++++++++++++---------------
>>>>    1 file changed, 13 insertions(+), 15 deletions(-)
>>>>
>>>> diff --git a/drivers/thermal/rockchip_thermal.c
>>>> b/drivers/thermal/rockchip_thermal.c
>>>> index c89ffb2..93ee307 100644
>>>> --- a/drivers/thermal/rockchip_thermal.c
>>>> +++ b/drivers/thermal/rockchip_thermal.c
>>>> @@ -124,7 +124,7 @@ struct rockchip_thermal_data {
>>>>    #define TSADCV2_AUTO_PERIOD_HT_TIME           50  /* msec */
>>>>      struct tsadc_table {
>>>> -       unsigned long code;
>>>> +       u32 code;
>>>>          long temp;
>>>>    };
>>>>    @@ -164,7 +164,6 @@ static const struct tsadc_table 
>>>> v2_code_table[] = {
>>>>          {3452, 115000},
>>>>          {3437, 120000},
>>>>          {3421, 125000},
>>>> -       {0, 125000},
>>>>    };
>>>>      static u32 rk_tsadcv2_temp_to_code(long temp)
>>>> @@ -191,19 +190,21 @@ static u32 rk_tsadcv2_temp_to_code(long temp)
>>>>          return 0;
>>>>    }
>>>>    -static int rk_tsadcv2_code_to_temp(u32 code)
>>>> +static int rk_tsadcv2_code_to_temp(u32 code, int *temp)
>
> It's wired that can't git-am this patch.
>
> I search the Eduardo'branch[0] for rockchip-thermal.c driver. The 
> function is "static long rk_tsadcv2_code_to_temp(u32 code)"
>
> Maybe i'm missing something, I verified this patch on my board but 
> this fixed.
>
> So you can free add it:
> Tested-by: Caesar Wang <wxt@rock-chips.com>
>
> [0]:
> url = 
> git://git.kernel.org/pub/scm/linux/kernel/git/evalenti/linux-soc-thermal.git
>
>
>
> ---
> Thanks,
> Caesar
>
>
>
>>>>    {
>>>> -       unsigned int low = 0;
>>>> +       unsigned int low = 1;
>>>>          unsigned int high = ARRAY_SIZE(v2_code_table) - 1;
>>>>          unsigned int mid = (low + high) / 2;
>>>>          unsigned int num;
>>>>          unsigned long denom;
>>>>    -     /* Invalid code, return -EAGAIN */
>>>> -       if (code > TSADCV2_DATA_MASK)
>>>> -               return -EAGAIN;
>>>> +       BUILD_BUG_ON(ARRAY_SIZE(v2_code_table) < 2);
>>>>    -     while (low <= high && mid) {
>>>> +       code &= TSADCV2_DATA_MASK;
>>>> +       if (code < v2_code_table[high].code)
>>>> +               return -EAGAIN;         /* Incorrect reading */
>>>> +
>>>> +       while (low <= high) {
>>>>                  if (code >= v2_code_table[mid].code &&
>>>>                      code < v2_code_table[mid - 1].code)
>>>>                          break;
>>>> @@ -223,7 +224,9 @@ static int rk_tsadcv2_code_to_temp(u32 code)
>>>>          num = v2_code_table[mid].temp - v2_code_table[mid - 1].temp;
>>>>          num *= v2_code_table[mid - 1].code - code;
>>>>          denom = v2_code_table[mid - 1].code - 
>>>> v2_code_table[mid].code;
>>>> -       return v2_code_table[mid - 1].temp + (num / denom);
>>>> +       *temp = v2_code_table[mid - 1].temp + (num / denom);
>>>> +
>>>> +       return 0;
>>>>    }
>>>>      /**
>>>> @@ -281,14 +284,9 @@ static int rk_tsadcv2_get_temp(int chn, void 
>>>> __iomem
>>>> *regs, int *temp)
>>>>    {
>>>>          u32 val;
>>>>    -     /* the A/D value of the channel last conversion need some 
>>>> time */
>>>>          val = readl_relaxed(regs + TSADCV2_DATA(chn));
>>>> -       if (val == 0)
>>>> -               return -EAGAIN;
>>>> -
>>>
>>> if we reserve the above code, that will get the ADC value.
>> But if we pass  0 into rk_tsadcv2_code_to_temp() it will trigger:
>>
>>>> +       if (code < v2_code_table[high].code)
>>>> +               return -EAGAIN;         /* Incorrect reading */
>> and still return -EAGAIN, as it was. There is no behavior change as
>> far as I can see.
> Yup, that's ok for me.
>
>>>
>>>> -       *temp = rk_tsadcv2_code_to_temp(val);
>>>>    -     return 0;
>>>> +       return rk_tsadcv2_code_to_temp(val, temp);
>>>>    }
>>>>      static void rk_tsadcv2_tshut_temp(int chn, void __iomem *regs, 
>>>> long
>>>> temp)
>>>
>> Thanks,
>> Dmitry
>>
>>
>>
>
diff mbox

Patch

diff --git a/drivers/thermal/rockchip_thermal.c b/drivers/thermal/rockchip_thermal.c
index c89ffb2..93ee307 100644
--- a/drivers/thermal/rockchip_thermal.c
+++ b/drivers/thermal/rockchip_thermal.c
@@ -124,7 +124,7 @@  struct rockchip_thermal_data {
 #define TSADCV2_AUTO_PERIOD_HT_TIME		50  /* msec */
 
 struct tsadc_table {
-	unsigned long code;
+	u32 code;
 	long temp;
 };
 
@@ -164,7 +164,6 @@  static const struct tsadc_table v2_code_table[] = {
 	{3452, 115000},
 	{3437, 120000},
 	{3421, 125000},
-	{0, 125000},
 };
 
 static u32 rk_tsadcv2_temp_to_code(long temp)
@@ -191,19 +190,21 @@  static u32 rk_tsadcv2_temp_to_code(long temp)
 	return 0;
 }
 
-static int rk_tsadcv2_code_to_temp(u32 code)
+static int rk_tsadcv2_code_to_temp(u32 code, int *temp)
 {
-	unsigned int low = 0;
+	unsigned int low = 1;
 	unsigned int high = ARRAY_SIZE(v2_code_table) - 1;
 	unsigned int mid = (low + high) / 2;
 	unsigned int num;
 	unsigned long denom;
 
-	/* Invalid code, return -EAGAIN */
-	if (code > TSADCV2_DATA_MASK)
-		return -EAGAIN;
+	BUILD_BUG_ON(ARRAY_SIZE(v2_code_table) < 2);
 
-	while (low <= high && mid) {
+	code &= TSADCV2_DATA_MASK;
+	if (code < v2_code_table[high].code)
+		return -EAGAIN;		/* Incorrect reading */
+
+	while (low <= high) {
 		if (code >= v2_code_table[mid].code &&
 		    code < v2_code_table[mid - 1].code)
 			break;
@@ -223,7 +224,9 @@  static int rk_tsadcv2_code_to_temp(u32 code)
 	num = v2_code_table[mid].temp - v2_code_table[mid - 1].temp;
 	num *= v2_code_table[mid - 1].code - code;
 	denom = v2_code_table[mid - 1].code - v2_code_table[mid].code;
-	return v2_code_table[mid - 1].temp + (num / denom);
+	*temp = v2_code_table[mid - 1].temp + (num / denom);
+
+	return 0;
 }
 
 /**
@@ -281,14 +284,9 @@  static int rk_tsadcv2_get_temp(int chn, void __iomem *regs, int *temp)
 {
 	u32 val;
 
-	/* the A/D value of the channel last conversion need some time */
 	val = readl_relaxed(regs + TSADCV2_DATA(chn));
-	if (val == 0)
-		return -EAGAIN;
-
-	*temp = rk_tsadcv2_code_to_temp(val);
 
-	return 0;
+	return rk_tsadcv2_code_to_temp(val, temp);
 }
 
 static void rk_tsadcv2_tshut_temp(int chn, void __iomem *regs, long temp)