[PATCH v3] Samsung: sdhci-s3c: add support for new card detection methods

[Marek Szyprowski]

Hello,

On Monday, July 26, 2010 10:57 PM Andrew Morton wrote:

> On Thu, 22 Jul 2010 10:25:44 +0200
> Marek Szyprowski <m.szyprowski at samsung.com> wrote:
> > On Thursday, July 22, 2010 1:12 AM
> > > On Fri, 16 Jul 2010 08:24:26 +0200
> > > Marek Szyprowski <m.szyprowski at samsung.com> wrote:
> > >
> > > > > >  arch/arm/mach-s3c64xx/setup-sdhci-gpio.c   |   14 +++++++++---
> > > > > >  arch/arm/mach-s5pc100/setup-sdhci-gpio.c   |   21
> ++++++++++++++----
> > > -
> > > > > >  arch/arm/mach-s5pv210/setup-sdhci-gpio.c   |   22
> +++++++++++++++---
> > > --
> > > > > >  arch/arm/plat-samsung/dev-hsmmc.c          |    5 ++++
> > > > > >  arch/arm/plat-samsung/dev-hsmmc1.c         |    5 ++++
> > > > > >  arch/arm/plat-samsung/dev-hsmmc2.c         |    5 ++++
> > > > > >  arch/arm/plat-samsung/dev-hsmmc3.c         |    5 ++++
> > > > > >  arch/arm/plat-samsung/include/plat/sdhci.h |   29
> > > > > ++++++++++++++++++++++++++++
> > > > > >  8 files changed, 90 insertions(+), 16 deletions(-)
> > > > >
> > > > > This is quite confusing.  You've already sent a patch called
> > > > > "sdhci-s3c: add support for new card detection methods".  It had
> the
> > > > > same changelog as this patch and the same title, but the two
> patches
> > > > > are utterly different!
> > > >
> > > > The real patch has been split into two for easier merging: 1. the
> driver
> > > > part and 2. samsung platform related part. The part which patch
> belongs
> > > > to is indicated in the last line of the change log.
> > > >
> > > > I'm really confused how to submit properly a patch that requires
> changes
> > > > to both the driver (which is merged by the proper driver maintainer's
> > > > sub-tree) and the platform (which should go through platform
> maintainer's
> > > > tree).
> > >
> > > Make the relationship very very clear in the changelog.  Send both
> > > patches to both maintainers.  Ask that one of them merge both patches
> > > and that the other ack both patches.
> >
> > Ok. Thanks for the hint.
> >
> > > Anyway, I've forgotten what's happening here.  I appear to be sitting
> on
> > >
> > > sdhci-s3c-add-support-for-the-non-standard-minimal-clock-value.patch
> >
> > This one is independent from the card-detection patches and can be
> applied
> > directly onto linux-next kernel tree.
> >
> > > and
> > > sdhci-s3c-add-support-for-new-card-detection-methods.patch
> > >
> > > the latter of which is below.
> > >
> > > Am I missing something?  Is
> > > sdhci-s3c-add-support-for-new-card-detection-methods.patch up to date?
> > >
> >
> > Ok, I will repost these 2 patches under new names to avoid further
> > confusion. Currently there exists two
> > "sdhci-s3c-add-support-for-new-card-detection-methods.patch", both of
> them
> > are required.
> 
> So the second(?) patch "[PATCH v4] Samsung: add new card detection
> methods in s3c-sdhci driver (platform part)" has already been applied
> to linux-next by Kukjin Kim, along with modifications which we have not
> been shown.
> 
> But the first patch
> (sdhci-s3c-add-support-for-the-non-standard-minimal-clock-value.patch)
> has not been applied to linux-next by anoyne and the third patch
> "[PATCH v4] sdhci-s3c: add support for new card detection methods
> (driver part)" has not been applied either.
> 
> This is a complete mess.  What happens if I apply and merge
> sdhci-s3c-add-support-for-the-non-standard-minimal-clock-value.patch
> and
> sdhci-s3c-add-support-for-new-card-detection-methods-driver-part.patch?
> I have no idea.

"[PATCH v4] Samsung: add new card detection methods in s3c-sdhci driver
(platform part)" perform some important changes in platform setup code.
It does not influence the driver itself (the driver can be still compiled
as before and works without changes).

The only changes added by Kukjin Kim were compilation fixes (looks I've
forgot one include on the platform that has not been tested). 

This patch is also a prerequisite for "[PATCH v4] sdhci-s3c: add support
for new card detection methods (driver part)".

> I think I'll just drop all of it.  Please sort all of this out then let
> me know what you want me to do.

I would to ask not for dropping. As all required platform changes has been
already merged by Kukjin, s
dhci-s3c-add-support-for-the-non-standard-minimal-clock-value.patch and 
sdhci-s3c-add-support-for-new-card-detection-methods-driver-part.patch can
be merged without causing compilation break in the 'next' kernel tree.

Best regards
--
Marek Szyprowski
Samsung Poland R&D Center