[PATCH] Blackfin: bf537: rename "CONFIG_ADT7310" and friends

[PATCH] Blackfin: bf537: rename "CONFIG_ADT7310" and friends

[Paul Bolle]

The Kconfig symbol ADT7310 got removed in commit
2b0c856ad9571013db8fc369194b7108dff3c18e ("staging:iio: Consolidate
adt7310 and adt7410 driver"). That commit did not touch the references
to CONFIG_ADT7310 and CONFIG_ADT7310_MODULE in the BF537-STAMP code.
Convert these now, to their ADT7410 equivalents.

Signed-off-by: Paul Bolle <pebolle@tiscali.nl>
---
0) Untested. This took a bit of guesswork. I guessed that multiple AD
drivers can be enabled at the same time in this code. And I also guessed
that the "adt7310" modalias is still correct.

1) If this passes testing it could go in stable (3.8.y).

 arch/blackfin/mach-bf537/boards/stamp.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/arch/blackfin/mach-bf537/boards/stamp.c b/arch/blackfin/mach-bf537/boards/stamp.c
index 95114ed..23bb55d 100644
--- a/arch/blackfin/mach-bf537/boards/stamp.c
+++ b/arch/blackfin/mach-bf537/boards/stamp.c
@@ -700,7 +700,7 @@ static struct bfin5xx_spi_chip ad7816_spi_chip_info = {
 };
 #endif
 
-#if defined(CONFIG_ADT7310) || defined(CONFIG_ADT7310_MODULE)
+#if defined(CONFIG_ADT7410) || defined(CONFIG_ADT7410_MODULE)
 static unsigned long adt7310_platform_data[3] = {
 /* INT bound temperature alarm event. line 1 */
 	IRQ_PG4, IRQF_TRIGGER_LOW,
@@ -1062,7 +1062,7 @@ static struct spi_board_info bfin_spi_board_info[] __initdata = {
 	},
 #endif
 
-#if defined(CONFIG_ADT7310) || defined(CONFIG_ADT7310_MODULE)
+#if defined(CONFIG_ADT7410) || defined(CONFIG_ADT7410_MODULE)
 	{
 		.modalias = "adt7310",
 		.max_speed_hz = 1000000,
-- 
1.7.11.7

Re: [uclinux-dist-devel] [PATCH] Blackfin: bf537: rename "CONFIG_ADT7310" and friends

[Lars-Peter Clausen]

On 03/25/2013 12:43 PM, Paul Bolle wrote:
> The Kconfig symbol ADT7310 got removed in commit
> 2b0c856ad9571013db8fc369194b7108dff3c18e ("staging:iio: Consolidate
> adt7310 and adt7410 driver"). That commit did not touch the references
> to CONFIG_ADT7310 and CONFIG_ADT7310_MODULE in the BF537-STAMP code.
> Convert these now, to their ADT7410 equivalents.
> 
> Signed-off-by: Paul Bolle <pebolle@tiscali.nl>
> ---
> 0) Untested. This took a bit of guesswork. I guessed that multiple AD
> drivers can be enabled at the same time in this code. And I also guessed
> that the "adt7310" modalias is still correct.
> 
> 1) If this passes testing it could go in stable (3.8.y).

Hi,

Thanks for the patch. Unfortunately this has already changed again for 3.10
since we replaced the IIO adt7410/adt7310 driver with a hwmon driver. So the
new config name is CONFIG_SENSORS_ADT7310.

Thanks,
- Lars

> 
>  arch/blackfin/mach-bf537/boards/stamp.c | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)
> 
> diff --git a/arch/blackfin/mach-bf537/boards/stamp.c b/arch/blackfin/mach-bf537/boards/stamp.c
> index 95114ed..23bb55d 100644
> --- a/arch/blackfin/mach-bf537/boards/stamp.c
> +++ b/arch/blackfin/mach-bf537/boards/stamp.c
> @@ -700,7 +700,7 @@ static struct bfin5xx_spi_chip ad7816_spi_chip_info = {
>  };
>  #endif
>  
> -#if defined(CONFIG_ADT7310) || defined(CONFIG_ADT7310_MODULE)
> +#if defined(CONFIG_ADT7410) || defined(CONFIG_ADT7410_MODULE)
>  static unsigned long adt7310_platform_data[3] = {
>  /* INT bound temperature alarm event. line 1 */
>  	IRQ_PG4, IRQF_TRIGGER_LOW,
> @@ -1062,7 +1062,7 @@ static struct spi_board_info bfin_spi_board_info[] __initdata = {
>  	},
>  #endif
>  
> -#if defined(CONFIG_ADT7310) || defined(CONFIG_ADT7310_MODULE)
> +#if defined(CONFIG_ADT7410) || defined(CONFIG_ADT7410_MODULE)
>  	{
>  		.modalias = "adt7310",
>  		.max_speed_hz = 1000000,

Re: [uclinux-dist-devel] [PATCH] Blackfin: bf537: rename "CONFIG_ADT7310" and friends

[Paul Bolle]

On Mon, 2013-03-25 at 12:55 +0100, Lars-Peter Clausen wrote:
> Thanks for the patch. Unfortunately this has already changed again for 3.10
> since we replaced the IIO adt7410/adt7310 driver with a hwmon driver. So the
> new config name is CONFIG_SENSORS_ADT7310.

That leaves v3.8 and v3.9 with a (minor?) problem, doesn't it? If that's
correct, can't something along the lines of my patch be added to v3.9
(and then trivially backported to v3.8)?


Paul Bolle

Re: [PATCH] Blackfin: bf537: rename "CONFIG_ADT7310" and friends

[Guenter Roeck]

On Mon, Mar 25, 2013 at 12:43:51PM +0100, Paul Bolle wrote:
> The Kconfig symbol ADT7310 got removed in commit
> 2b0c856ad9571013db8fc369194b7108dff3c18e ("staging:iio: Consolidate
> adt7310 and adt7410 driver"). That commit did not touch the references
> to CONFIG_ADT7310 and CONFIG_ADT7310_MODULE in the BF537-STAMP code.
> Convert these now, to their ADT7410 equivalents.
> 
> Signed-off-by: Paul Bolle <pebolle@tiscali.nl>
> ---
> 0) Untested. This took a bit of guesswork. I guessed that multiple AD
> drivers can be enabled at the same time in this code. And I also guessed
> that the "adt7310" modalias is still correct.
> 
> 1) If this passes testing it could go in stable (3.8.y).
> 
>  arch/blackfin/mach-bf537/boards/stamp.c | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)
> 
> diff --git a/arch/blackfin/mach-bf537/boards/stamp.c b/arch/blackfin/mach-bf537/boards/stamp.c
> index 95114ed..23bb55d 100644
> --- a/arch/blackfin/mach-bf537/boards/stamp.c
> +++ b/arch/blackfin/mach-bf537/boards/stamp.c
> @@ -700,7 +700,7 @@ static struct bfin5xx_spi_chip ad7816_spi_chip_info = {
>  };
>  #endif
>  
> -#if defined(CONFIG_ADT7310) || defined(CONFIG_ADT7310_MODULE)
> +#if defined(CONFIG_ADT7410) || defined(CONFIG_ADT7410_MODULE)

It might be easier and more future-proof to just drop the conditional.
If the driver isn't there, it won't be loaded anyway. It would also make
back-porting much easier.

Guenter

>  static unsigned long adt7310_platform_data[3] = {
>  /* INT bound temperature alarm event. line 1 */
>  	IRQ_PG4, IRQF_TRIGGER_LOW,
> @@ -1062,7 +1062,7 @@ static struct spi_board_info bfin_spi_board_info[] __initdata = {
>  	},
>  #endif
>  
> -#if defined(CONFIG_ADT7310) || defined(CONFIG_ADT7310_MODULE)
> +#if defined(CONFIG_ADT7410) || defined(CONFIG_ADT7410_MODULE)
>  	{
>  		.modalias = "adt7310",
>  		.max_speed_hz = 1000000,
> -- 
> 1.7.11.7
> 
> 

Re: [PATCH] Blackfin: bf537: rename "CONFIG_ADT7310" and friends

[Lars-Peter Clausen]

On 03/25/2013 04:22 PM, Guenter Roeck wrote:
> On Mon, Mar 25, 2013 at 12:43:51PM +0100, Paul Bolle wrote:
>> The Kconfig symbol ADT7310 got removed in commit
>> 2b0c856ad9571013db8fc369194b7108dff3c18e ("staging:iio: Consolidate
>> adt7310 and adt7410 driver"). That commit did not touch the references
>> to CONFIG_ADT7310 and CONFIG_ADT7310_MODULE in the BF537-STAMP code.
>> Convert these now, to their ADT7410 equivalents.
>>
>> Signed-off-by: Paul Bolle <pebolle@tiscali.nl>
>> ---
>> 0) Untested. This took a bit of guesswork. I guessed that multiple AD
>> drivers can be enabled at the same time in this code. And I also guessed
>> that the "adt7310" modalias is still correct.
>>
>> 1) If this passes testing it could go in stable (3.8.y).
>>
>>  arch/blackfin/mach-bf537/boards/stamp.c | 4 ++--
>>  1 file changed, 2 insertions(+), 2 deletions(-)
>>
>> diff --git a/arch/blackfin/mach-bf537/boards/stamp.c b/arch/blackfin/mach-bf537/boards/stamp.c
>> index 95114ed..23bb55d 100644
>> --- a/arch/blackfin/mach-bf537/boards/stamp.c
>> +++ b/arch/blackfin/mach-bf537/boards/stamp.c
>> @@ -700,7 +700,7 @@ static struct bfin5xx_spi_chip ad7816_spi_chip_info = {
>>  };
>>  #endif
>>  
>> -#if defined(CONFIG_ADT7310) || defined(CONFIG_ADT7310_MODULE)
>> +#if defined(CONFIG_ADT7410) || defined(CONFIG_ADT7410_MODULE)
> 
> It might be easier and more future-proof to just drop the conditional.
> If the driver isn't there, it won't be loaded anyway. It would also make
> back-porting much easier.
> 
> Guenter


The situation is unfortunately really messy. The issue is that these are
add-on board and the same SPI chip select pin is used for multiple devices.
If we add the spi_board_info for all devices unconditionally the SPI core
will scream at us because we tried to register multiple devices with the
same chip select pin and none of the devices gets actually registered.

- Lars

Re: [PATCH] Blackfin: bf537: rename "CONFIG_ADT7310" and friends

[Paul Bolle]

[Replaced previous maintainer and list with current maintainer and
list.]

On Mon, 2013-03-25 at 16:45 +0100, Lars-Peter Clausen wrote:
> On 03/25/2013 04:22 PM, Guenter Roeck wrote:
> > On Mon, Mar 25, 2013 at 12:43:51PM +0100, Paul Bolle wrote:
> >> The Kconfig symbol ADT7310 got removed in commit
> >> 2b0c856ad9571013db8fc369194b7108dff3c18e ("staging:iio: Consolidate
> >> adt7310 and adt7410 driver"). That commit did not touch the references
> >> to CONFIG_ADT7310 and CONFIG_ADT7310_MODULE in the BF537-STAMP code.
> >> Convert these now, to their ADT7410 equivalents.
> >>
> >> Signed-off-by: Paul Bolle <pebolle@tiscali.nl>
> >> ---
> >> 0) Untested. This took a bit of guesswork. I guessed that multiple AD
> >> drivers can be enabled at the same time in this code. And I also guessed
> >> that the "adt7310" modalias is still correct.
> >>
> >> 1) If this passes testing it could go in stable (3.8.y).
> >>
> >>  arch/blackfin/mach-bf537/boards/stamp.c | 4 ++--
> >>  1 file changed, 2 insertions(+), 2 deletions(-)
> >>
> >> diff --git a/arch/blackfin/mach-bf537/boards/stamp.c b/arch/blackfin/mach-bf537/boards/stamp.c
> >> index 95114ed..23bb55d 100644
> >> --- a/arch/blackfin/mach-bf537/boards/stamp.c
> >> +++ b/arch/blackfin/mach-bf537/boards/stamp.c
> >> @@ -700,7 +700,7 @@ static struct bfin5xx_spi_chip ad7816_spi_chip_info = {
> >>  };
> >>  #endif
> >>  
> >> -#if defined(CONFIG_ADT7310) || defined(CONFIG_ADT7310_MODULE)
> >> +#if defined(CONFIG_ADT7410) || defined(CONFIG_ADT7410_MODULE)
> > 
> > It might be easier and more future-proof to just drop the conditional.
> > If the driver isn't there, it won't be loaded anyway. It would also make
> > back-porting much easier.
> > 
> > Guenter
> 
> 
> The situation is unfortunately really messy. The issue is that these are
> add-on board and the same SPI chip select pin is used for multiple devices.
> If we add the spi_board_info for all devices unconditionally the SPI core
> will scream at us because we tried to register multiple devices with the
> same chip select pin and none of the devices gets actually registered.

This issue is still present in v3.14-rc2. Guenter's suggestion is the
easiest way out. Should I submit a trivial patch that just removes the
dead code depending on never defined CONFIG_ADT7310 and
CONFIG_ADT7310_MODULE?


Paul Bolle

Re: [PATCH] Blackfin: bf537: rename "CONFIG_ADT7310" and friends

[Paul Bolle]

On Thu, 2014-02-13 at 10:57 +0100, Paul Bolle wrote:
> This issue is still present in v3.14-rc2. Guenter's suggestion is the
> easiest way out. Should I submit a trivial patch that just removes the
> dead code depending on never defined CONFIG_ADT7310 and
> CONFIG_ADT7310_MODULE?

And that is probably also true for for CONFIG_ADT7410 and
CONFIG_ADT7410_MODULE. See https://lkml.org/lkml/2013/5/13/497 (and my
follow up to that message).


Paul Bolle

Re: [PATCH] Blackfin: bf537: rename "CONFIG_ADT7310" and friends

[Guenter Roeck]

On 02/13/2014 02:31 AM, Paul Bolle wrote:
> On Thu, 2014-02-13 at 10:57 +0100, Paul Bolle wrote:
>> This issue is still present in v3.14-rc2. Guenter's suggestion is the
>> easiest way out. Should I submit a trivial patch that just removes the
>> dead code depending on never defined CONFIG_ADT7310 and
>> CONFIG_ADT7310_MODULE?
>

My suggestion was to remove the ifdefs, not the code, and if I
understand the reply correctly that won't work because in that
case multiple devices would be registered with the SPI core
on the same SPI interface.

> And that is probably also true for for CONFIG_ADT7410 and
> CONFIG_ADT7410_MODULE. See https://lkml.org/lkml/2013/5/13/497 (and my
> follow up to that message).
>

Both Jean and myself sent Reviewed-by/Acked-by feedback.
Not being the maintainers, that is pretty much all we can do.
 From there it is really up to the maintainer to accept
or reject the patches.

Sure, we both suggested that the code is messy, but cleaning
that up would or should be a separate task anyway for someone
who has both the hardware and the time/interest to do it.

Guenter

Re: [PATCH] Blackfin: bf537: rename "CONFIG_ADT7310" and friends

[Paul Bolle]

On Thu, 2014-02-13 at 08:29 -0800, Guenter Roeck wrote:
> On 02/13/2014 02:31 AM, Paul Bolle wrote:
> > On Thu, 2014-02-13 at 10:57 +0100, Paul Bolle wrote:
> >> This issue is still present in v3.14-rc2. Guenter's suggestion is the
> >> easiest way out. Should I submit a trivial patch that just removes the
> >> dead code depending on never defined CONFIG_ADT7310 and
> >> CONFIG_ADT7310_MODULE?
> >
> 
> My suggestion was to remove the ifdefs, not the code, and if I
> understand the reply correctly that won't work because in that
> case multiple devices would be registered with the SPI core
> on the same SPI interface.

Then I misinterpreted your suggestion. But removing the dead code
remains the easiest way out, of course. (Note that the code has only
been dead since v3.8, ie only a year now.)

> > And that is probably also true for for CONFIG_ADT7410 and
> > CONFIG_ADT7410_MODULE. See https://lkml.org/lkml/2013/5/13/497 (and my
> > follow up to that message).
> 
> Both Jean and myself sent Reviewed-by/Acked-by feedback.
> Not being the maintainers, that is pretty much all we can do.
>  From there it is really up to the maintainer to accept
> or reject the patches.

Jean and you are included in (some of) the threads regarding a few
blackfin patches I sent last year. I forgot why you two are. Anyhow,
those patches were (mostly) ignored by the people working on blackfin.
The maintainer appears to be more responsive now. So I figured it might
be worthwhile to send reminders. Maybe I have more luck this time.


Paul Bolle