[PATCH v2 0/4] Move plat-mxc gpio driver into drivers/gpio

Hi Grant,

On Fri, Jun 03, 2011 at 08:55:58AM -0600, Grant Likely wrote:

On Fri, Jun 3, 2011 at 8:34 AM, Shawn Guo [off-list ref] wrote:

quoted

Hi Russell,

On Fri, Jun 03, 2011 at 08:52:01AM +0100, Russell King - ARM Linux wrote:

quoted

On Thu, Jun 02, 2011 at 11:33:48AM +0800, Shawn Guo wrote:

quoted

?arch/arm/plat-mxc/gpio.c ? ? ? ? ? ? ? ? ? ? ? ?| ?361 -------------------
?drivers/gpio/gpio-mxc.c ? ? ? ? ? ? ? ? ? ? ? ? | ?433 +++++++++++++++++++++++

My bad here. ?I should have used 'git diff --stat -M' to show the
the following.

.../arm/plat-mxc/gpio.c => drivers/gpio/gpio-mxc.c | ?216 +++++++++++++-------

quoted

I'm wondering why just moving this driver into drivers/gpio has
resulted in it growing by 72 lines - and it's not clear from the
diffs why that is because of the way they're broken up.

Yes, I agree. ?But when I did something like that to ease the review,
people think it's not necessary :)

http://permalink.gmane.org/gmane.linux.kernel/1143257

The issue was bisectability: it looked like the build would break
after applying the first patch.  The first patch should move the

Yes, the build would break only if you change Kconfig/Makefile to
actually build it.  The patch does not enable the build of the driver
in the patch.

driver without breaking the build, and then you can follow up with
driver fixes.  I don't want to see functional changes mixed in with
the file move change.

Understood.  Do you want me to resend the gpio-mxs and gpio-mxc patch
sets for that?  Or can I follow the practice you and Russell
suggested in the future posts?  I have learnt the lesson.

quoted

+
+static struct platform_driver mxc_gpio_driver = {
+ ? ? ? .driver ? ? ? ? = {
+ ? ? ? ? ? ? ? .name ? = "gpio-mxc",

.owner = THIS_MODULE,

Do I need to re-spin the patch set to fix it, or maintainer (you or
Sascha) can help to fix it up?

BTW, do you and Sascha get the agreement on which tree the gpio-mxs
and gpio-mxc should go through?

-- 
Regards,
Shawn