[PATCH v2 1/2] mmc: dw_mmc: Handle late vmmc regulators with EPROBE_DEFER

[Doug Anderson]

Tomasz,

On Fri, Jun 7, 2013 at 10:42 AM, Tomasz Figa <tomasz.figa at gmail.com> wrote:
> A question just out of curiousity: Is it really correct to have one vmmc
> regulator for the whole device, instead of one regulator per slot?
>
> This might be something obvious, but I don't know any details about
> dw_mmc, so sorry if it's the case.

I don't actually know.  I've never seen a multi-slot implementation so
I end up doing lots of speculation when I make changes.  Certainly the
code I'm moving wouldn't have worked correctly in the
more-than-one-slot case.  There's only space to pass one regulator
(it's not per-slot) and that one regulator would have just been
enabled multiple times.  ...and it would have only been disabled once
in places like suspend/resume.

If there is every a multi-slot implementation that needs vmmc on a
per-slot it seems like we could address at that time?

Thanks!

-Doug