[U-Boot] [PATCH v4 0/3] i2c: sunxi: Support every i2c controller on each supported platform

[Hans de Goede]

Hi,

On 08-04-15 13:05, Paul Kocialkowski wrote:
> Actually, I'd prefer that you do not merge this since some ifdef logic
> around controller 0 is still missing and mvtwsi code still assumes that
> controller 0 will be defined. It may be the case that users wish to only
> enable controller [1-4]. I think the Kconfig logic is also not necessary
> and not a great addition to have since sunxi will be the sole user of it
> (while other boards will keep using the option defined in the config
> header). I'll work those issues out soon and provide a v5.
>
> What do you think about all this?

I've not yet had a chance to take a look at your V4, but reducing
the amount of necessary conditionals sounds good, so go for it.

Regards,

Hans

>
> Le mardi 07 avril 2015 à 22:19 +0200, Paul Kocialkowski a écrit :
>> Changes since v3:
>> * Kconfig support for MVTWSI
>> * Only enable twsi0 by default for platforms that always use it for the AXP
>> * Remove enabling other I2C busses by default on boards that expose them on pin
>>    headers since those might be used for some other functionalities
>>
>> Changes since v2:
>> * I2C/TWI busses enable for Cubietruck as well
>>
>> Changes since v1:
>> * Kconfig option to enable I2C/TWI controllers 1-4 (when applicable)
>> * Following patch to enable exposed busses on a few community-supported
>>    single-board-computers
>>
>> This series adds support for every i2c controller found on
>> sun4i/sun5i/sun6i/sun7i/sun8i platforms and shouldn't break support for Marvell
>> platforms (orion5x, kirkwood, armada xp) the driver was originally written for.
>>
>> Regarding sunxi, I double-checked that this doesn't conflict with
>> VIDEO_LCD_PANEL_I2C.
>>
>> I would be interested in having this tested on sun8i (A23), since I changed TWI0
>> muxing (to PH2-PH3 instead of PB0-PB1), according to the user manual and what
>> is being done on the upstream Linux kernel. I2C was either not working before,
>> or it was being muxed correctly by the bootrom, probably to communicate with the
>> AXP, which luckily made it work in U-Boot too, since the I/O base address was
>> already correct.
>>
>> My use case here is that I'm writing a slave-side bitbang i2c implementation
>> (with an Arduino) for a school project, using a Cubieboard2 as master and
>> U-Boot as POC. However, only TWI1 was available through the expansion pins,
>> hence the need for this series.
>>
>> _______________________________________________
>> U-Boot mailing list
>> U-Boot at lists.denx.de
>> http://lists.denx.de/mailman/listinfo/u-boot
>