[GIT PULL] Urgent fixes (for v3.4 if possible) for Renesas ARM-based platforms

[Rafael J. Wysocki]

On Saturday, May 12, 2012, Olof Johansson wrote:
> On Fri, May 11, 2012 at 11:37 PM, Guennadi Liakhovetski
> <g.liakhovetski at gmx.de> wrote:
> > On Fri, 11 May 2012, Olof Johansson wrote:
> >
> >> Hej Magnus,
> >>
> >> On Fri, May 11, 2012 at 9:28 PM, Magnus Damm <magnus.damm at gmail.com> wrote:
> >>
> >> >>>  git://git.kernel.org/pub/scm/linux/kernel/git/rafael/renesas.git fixes
> >> >>
> >> >> No such branch (nor tag).
> >> >
> >> > I believe Rafael will get back to you about that when he gets to his computer.
> >>
> >> Yep, good.
> >>
> >> >>> Guennadi Liakhovetski (2):
> >> >>>      ARM: mach-shmobile: convert ag5evm to use the generic MMC GPIO hotplug helper
> >> >>>      ARM: mach-shmobile: convert mackerel to use the generic MMC GPIO hotplug helper
> >> >>
> >> >> These seem to fix the long-standing build errors on the two platforms,
> >> >> but that's impossible to tell from reading the commit messages (they
> >> >> are empty). Care to add one describing why they're important to go in?
> >> >
> >> > I only know that a) some boards don't build without these fixes and b)
> >> > they used to build just fine.
> >> >
> >> > Perhaps Guennadi [CC:ed] could be so kind to fill us in with the blanks?
> >>
> >> It was pretty obvious once I tried applying them and building that
> >> they did indeed resolve the build errors. However, the reason I'm
> >> asking for a better description is that when looking at just the pull
> >> request, or even the patch full commit description, it was impossible
> >> to tell why the patch was urgent as a fix. So, while I am not
> >> questioning that the patch should be included, I'm just requesting to
> >> make sure it's properly described. That's also good for historical
> >> purposes when someone is reading the git logs a year from now, etc.
> >
> > Ok, sorry, I wasn't sure from the previous mail - would you prefer to
> > actually extend their commit messages? As described here
> >
> > http://thread.gmane.org/gmane.linux.kernel.mmc/12744/focus=12748
> >
> > These patches are a part of a patch-series, so, when they were submitted
> > in that context, it was pretty obvious how they relate to the rest.
> > However, unfortunately, they are now applied later than the main patches
> > in that series, so, their purpose has become much less obvious, and their
> > main feature now seems to be not the conversion of the affected platforms,
> > but fixing the build.
> 
> Yeah, that seems to be what happened here, so indeed by now the
> commits are a bit out of context.
> 
> > So, I think, we could use something like this in the
> > patch descriptions:
> >
> > This also fixes modular mmc build on this platform by eliminating the use
> > of an inline function, which calls into the mmc core.
> 
> Sounds good, with a cut-and-paste of the build error that happens
> without the patch for extra credit.
> 
> Rafael, can you add the above or something like it to the commit
> messages when you push out the fixes branch?

Well, I've pushed it already, but I can still update it.  Please don't
pull from it yet.

Thanks,
Rafael