Re: [PATCH v3 1/2] iio: adc: driver for ti adc081s/adc101s/adc121s

[Milan Stevanovic <milan.o.stevanovic@gmail.com>]

On 01/26/2018 07:43 PM, Andy Shevchenko wrote:
> On Fri, Jan 26, 2018 at 8:25 PM, Lars-Peter Clausen <lars@metafoo.de> wrote:
>> On 01/26/2018 07:19 PM, Andy Shevchenko wrote:
>>> On Sun, Jan 14, 2018 at 10:32 PM, Milan Stevanovic
>>> <milan.o.stevanovic@gmail.com> wrote:
>>>>      Add Linux device driver for TI single-channel CMOS
>>>>      8/10/12-bit analog-to-digital converter with a
>>>>      high-speed serial interface.
>>>>
>>>> Signed-off-by: Milan Stevanovic <milan.o.stevanovic@gmail.com>
>>>> + * Analog Devices AD7466/7/8 AD7476/5/7/8 (A) SPI ADC driver
>>>> + * TI ADC081S/ADC101S/ADC121S 8/10/12-bit SPI ADC driver
>>> Guys, I'm not sure I understood this mix.
>> You often have the case where two or even more vendors produce parts that
>> are (mostly) pin and register map compatible. This is typically to fulfill
>> the second source requirement that some customers have.
>>
>> It is not uncommon to see drivers that support parts from different vendors.
> Yep, though in this case we have, it seems, a counterpart (i2c
> variant) in drivers/iio/adc/ti-adc081c.c
>
>>> We have like few TI drivers here:
>>>
>>> drivers/iio/adc/ti-adc0832.c:352:module_spi_driver(adc0832_driver);
>>> drivers/iio/adc/ti-adc084s021.c:269:module_spi_driver(adc084s021_driver);
>>> drivers/iio/adc/ti-adc108s102.c:343:module_spi_driver(adc108s102_driver);
>>> drivers/iio/adc/ti-adc12138.c:547:module_spi_driver(adc12138_driver);
>>> drivers/iio/adc/ti-adc128s052.c:211:module_spi_driver(adc128_driver);
>>> drivers/iio/adc/ti-adc161s626.c:276:module_spi_driver(ti_adc_driver);
>>> drivers/iio/adc/ti-ads7950.c:519:module_spi_driver(ti_ads7950_driver);
>>> drivers/iio/adc/ti-ads8688.c:481:module_spi_driver(ads8688_driver);
>>> drivers/iio/adc/ti-tlc4541.c:266:module_spi_driver(tlc4541_driver);
>>>
>>> What's wrong with them?
>> They are probably not register map compatible with those other drivers. (Or
>> nobody cared to check if they are register map compatible).
> I would believe in the latter than former.
>
>>> Is it here code duplication between two vendors?
> ...and instead of doing such mix I would really rather have a separate
> glue driver to the same code.
>
I spoke about this with Jonathan. Generally we can share a few lines
here and there but not enough to overcome the fact that the
drivers just became a lot less readable.
There are comments for that in patch/10132693