[PATCH V1] regulator: Enable supply regulator if child rail is enabled.

[PATCH V1] regulator: Enable supply regulator if child rail is enabled.

[Laxman Dewangan]

During regulator_register, the rail is set on the provided
machine constraints and if it is enabled then it is also
require to enable the supply regulator. This will make sure
that:
   1. Proper reference count for supply regulator to be maintain.
   2. Supply regulator should be enable when given rail is enabled.

Signed-off-by: Laxman Dewangan <ldewangan-DDmLM1+adcrQT0dZR+AlfA@public.gmane.org>
---
When given rail is registered and if it has machine constratins to
enable the rail then it is also require to enable the supply rail for
a given rail. This will make sure that given rail is enabled and
proper reference count is maintained for the supply rail to avoid
disable of supply rail even child rail is enabled.

 drivers/regulator/core.c |    8 ++++++++
 1 files changed, 8 insertions(+), 0 deletions(-)

diff --git a/drivers/regulator/core.c b/drivers/regulator/core.c
index dbdebed..6e7a4f1 100644
--- a/drivers/regulator/core.c
+++ b/drivers/regulator/core.c
@@ -2801,6 +2801,14 @@ struct regulator_dev *regulator_register(struct regulator_desc *regulator_desc,
 		ret = set_supply(rdev, r);
 		if (ret < 0)
 			goto scrub;
+
+		/* Enable supply if rail is enabled */
+		if (rdev->desc->ops->is_enabled &&
+				rdev->desc->ops->is_enabled(rdev)) {
+			ret = regulator_enable(rdev->supply);
+			if (ret < 0)
+				goto scrub;
+		}
 	}
 
 	/* add consumers devices */
-- 
1.7.1.1

Re: [PATCH V1] regulator: Enable supply regulator if child rail is enabled.

[Mark Brown]

On Mon, Jan 02, 2012 at 01:08:45PM +0530, Laxman Dewangan wrote:
> During regulator_register, the rail is set on the provided
> machine constraints and if it is enabled then it is also
> require to enable the supply regulator. This will make sure
> that:
>    1. Proper reference count for supply regulator to be maintain.
>    2. Supply regulator should be enable when given rail is enabled.
> 
> Signed-off-by: Laxman Dewangan <ldewangan-DDmLM1+adcrQT0dZR+AlfA@public.gmane.org>

Applied, thanks.

> When given rail is registered and if it has machine constratins to
> enable the rail then it is also require to enable the supply rail for
> a given rail. This will make sure that given rail is enabled and
> proper reference count is maintained for the supply rail to avoid
> disable of supply rail even child rail is enabled.

Why is this not in the changelog?

RE: [PATCH V1] regulator: Enable supply regulator if child rail is enabled.

[Laxman Dewangan]

> From: Mark Brown [mailto:broonie@opensource.wolfsonmicro.com]
> Sent: Monday, January 02, 2012 5:06 PM
> > During regulator_register, the rail is set on the provided
> > machine constraints and if it is enabled then it is also
> > require to enable the supply regulator. This will make sure
> > that:
> >    1. Proper reference count for supply regulator to be maintain.
> >    2. Supply regulator should be enable when given rail is enabled.
> >
> > Signed-off-by: Laxman Dewangan <ldewangan@nvidia.com>
> 
> Applied, thanks.
> 
> > When given rail is registered and if it has machine constratins to
> > enable the rail then it is also require to enable the supply rail for
> > a given rail. This will make sure that given rail is enabled and
> > proper reference count is maintained for the supply rail to avoid
> > disable of supply rail even child rail is enabled.
> 
> Why is this not in the changelog?
I think this is in just different wording for explaining the change.

Re: [PATCH V1] regulator: Enable supply regulator if child rail is enabled.

[Mark Brown]

On Mon, Jan 02, 2012 at 05:37:04PM +0530, Laxman Dewangan wrote:

> > Why is this not in the changelog?

> I think this is in just different wording for explaining the change.

If something like this is worth saying it should be in the changelog so
that anyone finding the patch in git can see it too - in general stuff
after the --- should be things like differences from previous versions
of the patch or information about dependencies that need to be taken
care of when applying that don't matter once the patch is applied.