[PATCH 1/4] of_gpio: Return GPIO flags from of_get_gpio()

Anton Vorontsov avorontsov at ru.mvista.com
Wed Oct 29 02:16:59 EST 2008


On Tue, Oct 28, 2008 at 08:53:11AM -0600, Grant Likely wrote:
> On Tue, Oct 28, 2008 at 8:39 AM, Anton Vorontsov
> <avorontsov at ru.mvista.com> wrote:
> > Can you repost a fixed version with my Ack and Cc: Andrew Morton,
> > Benjamin Herrenschmidt?
> >
> > I think this change should go into the 2.6.28, so that we can
> > write new code on top of new API. Otherwise this change will cause
> > issues in the next merge window.
> 
> While I like this series, it should not be pushed for .28.  The window
> is closed and we are in bug-fix-only mode.

IIRC, we approve API changes in the early -rcs, exactly to avoid build
break issues for the new code (though the API changes themselves are
discouraged).

> Focus on getting it into Ben's -next tree instead.

Would Benjamin apply the USB FHCI driver then? I doubt it.

So, there are options:

1. Either this patch go in into the .28;

2. Or this patch go in _after_ USB FHCI driver, and Trent takes care
   to convert it to the new API. (I'm trying to push the driver since
   2007, I think. Firstly it has GPIO framework dependency, QE GPIO,
   QE Pinmux, FSL GTM dependency, bindings, e.t.c. And now that.
   No, no deal. I quite tired of the dependencies for this driver.)

3. Or we apply my original patch that won't break the API.
   http://lkml.org/lkml/2008/7/25/236
   I still like it better than Trent's patch, exactly because it
   can't break anything.

> That will deal with any merge window
> conflicts.

This won't deal with build breakage because of API changes, which
means I can't submit FHCI driver to Greg, I have to wait for the
API changes applied.

Thanks,

-- 
Anton Vorontsov
email: cbouatmailru at gmail.com
irc://irc.freenode.net/bd2



More information about the Linuxppc-dev mailing list