Re: [PATCH] mmc: omap_hsmmc: Update driver to support without regulators

[Kishon Vijay Abraham I <kishon@ti.com>]

Hi,

On Wednesday 22 July 2015 05:37 PM, Ulf Hansson wrote:
> [...]
> 
>> Hi Ulf,
>>
>> Just to clarify do you have an issue with this patch or in general
>> you rather nothave !CONFIG_REGULATOR supported? If its the latter
>> then wouldit be better to remove the !CONFIG_REGULATOR portion from
>> this driver all together.
> 
> I don't like this patch as is, as it seems like you are replacing the
> VMMC regulator that actually exist in HW by parsing a DT property.
> 
> I think mmc_of_parse_voltage() shall *only* be used when the OCR mask
> (->ocr_avail) can't be fetched by using a regulator. Typically that's
> when the VMMC voltage levels are managed within the mmc controller
> itself.
> 
> I don't have a strong opinion on how you solve that, but I guess you
> have two options.
> 
> 1.
> Make sure the driver can be build and used both for CONFIG_REGULATOR
> and !CONFIG_REGULATOR. What you need to take care of then, is to
> assign the ->ocr_avail mask to a default value in the case when it
> can't be fetched from a regulator. Most mmc drivers has some kind of
> fall-back mechanism like this.

I'm working on cleaning up omap_hsmmc driver where I have taken care to avoid
dependency with CONFIG_REGULATOR. I'll post the series asap.

Thanks
Kishon