[U-Boot] Samsung I2C driver

[U-Boot] Samsung I2C driver

[Simon Glass]

Hi Przemyslaw,

At present this driver has a few implementations and supports
non-device-tree and non-driver model. From I can see the odroid and
perhaps one other platform need conversion and then all of this old
stuff can be removed. Do you have any plans to look at this?

Regards,
Simon

[U-Boot] Samsung I2C driver

[Przemyslaw Marczak]

Hello Simon,

On 12/07/2015 02:07 PM, Simon Glass wrote:
> Hi Przemyslaw,
>
> At present this driver has a few implementations and supports
> non-device-tree and non-driver model. From I can see the odroid and
> perhaps one other platform need conversion and then all of this old
> stuff can be removed. Do you have any plans to look at this?
>
> Regards,
> Simon
>
>

Yes, this driver is very dirty and needs a big clean-up, however at 
present I don't have time for doing this.

I can see that there are few boards that are using the old version of 
this driver. I suppose that the "COMPAT" layer can be enabled for those 
boards and everything should work as previous, right?

I can help you with the testing for Trats/Trats2/OdroidU3 if you want 
move it on your own. I could take care of this, but no sooner than in 
the new year.

Best regards,
-- 
Przemyslaw Marczak
Samsung R&D Institute Poland
Samsung Electronics
p.marczak at samsung.com

[U-Boot] Samsung I2C driver

[Simon Glass]

+David, who is the vcma9 maintainer.

Hi Przemyslaw,

On 7 December 2015 at 07:00, Przemyslaw Marczak <p.marczak@samsung.com> wrote:
> Hello Simon,
>
>
> On 12/07/2015 02:07 PM, Simon Glass wrote:
>>
>> Hi Przemyslaw,
>>
>> At present this driver has a few implementations and supports
>> non-device-tree and non-driver model. From I can see the odroid and
>> perhaps one other platform need conversion and then all of this old
>> stuff can be removed. Do you have any plans to look at this?
>>
>> Regards,
>> Simon
>>
>>
>
> Yes, this driver is very dirty and needs a big clean-up, however at present
> I don't have time for doing this.
>
> I can see that there are few boards that are using the old version of this
> driver. I suppose that the "COMPAT" layer can be enabled for those boards
> and everything should work as previous, right?
>
> I can help you with the testing for Trats/Trats2/OdroidU3 if you want move
> it on your own. I could take care of this, but no sooner than in the new
> year.

I can leave it to you for when you have time. David are you able to
move your board to driver model for i2c?

Regards,
Simon

[U-Boot] Samsung I2C driver

[David Müller (ELSOFT AG)]

Simon Glass wrote:

> David are you able to move your board to driver model for i2c?

To be frank, I don't know it (yet). Being a total DM novice, it is
difficult for me to estimate the amount of effort needed to do the
conversion.

Is it really possible to convert just the I2C subsystem to DM and keep
the rest as is?


Dave

[U-Boot] Samsung I2C driver

[Simon Glass]

Hi David,

On 11 December 2015 at 05:07, David M?ller (ELSOFT AG)
<d.mueller@elsoft.ch> wrote:
> Simon Glass wrote:
>
>> David are you able to move your board to driver model for i2c?
>
> To be frank, I don't know it (yet). Being a total DM novice, it is
> difficult for me to estimate the amount of effort needed to do the
> conversion.
>
> Is it really possible to convert just the I2C subsystem to DM and keep
> the rest as is?

Yes, you can define CONFIG_DM_I2C to do this. But you will always need
CONFIG_DM and on exynos you will need a device tree - hopefully you
have one from Linux for your board.

Regards,
Simon

[U-Boot] Samsung I2C driver

[Przemyslaw Marczak]

Hi,

On 12/11/2015 04:04 AM, Simon Glass wrote:
> +David, who is the vcma9 maintainer.
>
> Hi Przemyslaw,
>
> On 7 December 2015 at 07:00, Przemyslaw Marczak <p.marczak@samsung.com> wrote:
>> Hello Simon,
>>
>>
>> On 12/07/2015 02:07 PM, Simon Glass wrote:
>>>
>>> Hi Przemyslaw,
>>>
>>> At present this driver has a few implementations and supports
>>> non-device-tree and non-driver model. From I can see the odroid and
>>> perhaps one other platform need conversion and then all of this old
>>> stuff can be removed. Do you have any plans to look at this?
>>>
>>> Regards,
>>> Simon
>>>
>>>
>>
>> Yes, this driver is very dirty and needs a big clean-up, however at present
>> I don't have time for doing this.
>>
>> I can see that there are few boards that are using the old version of this
>> driver. I suppose that the "COMPAT" layer can be enabled for those boards
>> and everything should work as previous, right?
>>
>> I can help you with the testing for Trats/Trats2/OdroidU3 if you want move
>> it on your own. I could take care of this, but no sooner than in the new
>> year.
>
> I can leave it to you for when you have time. David are you able to
> move your board to driver model for i2c?
>
> Regards,
> Simon
>
>

That's ok for me. I have plan to do few things around this task and I 
think, that David will have enough time to move his board to dm before I 
finish all my tasks.

Best regards,
-- 
Przemyslaw Marczak
Samsung R&D Institute Poland
Samsung Electronics
p.marczak at samsung.com

[U-Boot] Samsung I2C driver

[David Müller (ELSOFT AG)]

Hello

I'm currently implementing DM support in the s3c24xx serial driver and
activating DM support for the VCMA9 board using

CONFIG_DM=y
CONFIG_DM_SERIAL=y
CONFIG_OF_CONTROL=y
CONFIG_OF_EMBED=y
CONFIG_DEFAULT_DEVICE_TREE="s3c24xx-vcma9"


This results in the following problem:

  CC      drivers/i2c/s3c24x0_i2c.o
drivers/i2c/s3c24x0_i2c.c: In function 'process_nodes':
drivers/i2c/s3c24x0_i2c.c:1024:12: error: 'PINMUX_FLAG_HS_MODE'
undeclared (first use in this function)
drivers/i2c/s3c24x0_i2c.c:1024:12: note: each undeclared identifier is
reported only once for each function it appears in
drivers/i2c/s3c24x0_i2c.c:1035:3: warning: implicit declaration of
function 'pinmux_decode_periph_id' [-Wimplicit-function-declaration]
drivers/i2c/s3c24x0_i2c.c:1041:3: warning: implicit declaration of
function 'exynos_pinmux_config' [-Wimplicit-function-declaration]
make[1]: *** [drivers/i2c/s3c24x0_i2c.o] Error 1


Do you have an updated version of the s3c24xx I2C driver available yet?


Thanks

[U-Boot] Samsung I2C driver

[Simon Glass]

Hi David,

On 20 February 2016 at 00:04, David M?ller (ELSOFT AG)
<d.mueller@elsoft.ch> wrote:
> Hello
>
> I'm currently implementing DM support in the s3c24xx serial driver and
> activating DM support for the VCMA9 board using
>
> CONFIG_DM=y
> CONFIG_DM_SERIAL=y
> CONFIG_OF_CONTROL=y
> CONFIG_OF_EMBED=y
> CONFIG_DEFAULT_DEVICE_TREE="s3c24xx-vcma9"
>
>
> This results in the following problem:
>
>   CC      drivers/i2c/s3c24x0_i2c.o
> drivers/i2c/s3c24x0_i2c.c: In function 'process_nodes':
> drivers/i2c/s3c24x0_i2c.c:1024:12: error: 'PINMUX_FLAG_HS_MODE'
> undeclared (first use in this function)
> drivers/i2c/s3c24x0_i2c.c:1024:12: note: each undeclared identifier is
> reported only once for each function it appears in
> drivers/i2c/s3c24x0_i2c.c:1035:3: warning: implicit declaration of
> function 'pinmux_decode_periph_id' [-Wimplicit-function-declaration]
> drivers/i2c/s3c24x0_i2c.c:1041:3: warning: implicit declaration of
> function 'exynos_pinmux_config' [-Wimplicit-function-declaration]
> make[1]: *** [drivers/i2c/s3c24x0_i2c.o] Error 1
>
>
> Do you have an updated version of the s3c24xx I2C driver available yet?

This seems to be because:

#if (defined CONFIG_EXYNOS4 || defined CONFIG_EXYNOS5)
#include <asm/arch/clk.h>
#include <asm/arch/cpu.h>
#include <asm/arch/pinmux.h>
#else
#include <asm/arch/s3c24x0_cpu.h>
#endif

so perhaps you need to check if the pinmux needs to be defined for your chip?

Regards,
Simon