[PATCH] regulator: Fix machine constraints apply for min_uV != max_uV

[PATCH] regulator: Fix machine constraints apply for min_uV != max_uV

[Jarkko Nikula]

Constraints are not applied until both min_uV and max_uV are the same. This
looks like a typo in the test.

Signed-off-by: Jarkko Nikula <jhnikula@gmail.com>
---
 drivers/regulator/core.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/drivers/regulator/core.c b/drivers/regulator/core.c
index 422a709..b092070 100644
--- a/drivers/regulator/core.c
+++ b/drivers/regulator/core.c
@@ -724,7 +724,7 @@ static int machine_constraints_voltage(struct regulator_dev *rdev,
 
 	/* do we need to apply the constraint voltage */
 	if (rdev->constraints->apply_uV &&
-		rdev->constraints->min_uV == rdev->constraints->max_uV &&
+		rdev->constraints->min_uV && rdev->constraints->max_uV &&
 		ops->set_voltage) {
 		ret = ops->set_voltage(rdev,
 			rdev->constraints->min_uV, rdev->constraints->max_uV);
-- 
1.7.1

Re: [PATCH] regulator: Fix machine constraints apply for min_uV != max_uV

[Mark Brown]

On Fri, Sep 03, 2010 at 04:43:49PM +0300, Jarkko Nikula wrote:
> Constraints are not applied until both min_uV and max_uV are the same. This
> looks like a typo in the test.
> 
> Signed-off-by: Jarkko Nikula <jhnikula@gmail.com>

This is deliberate behaviour, apply_uV can only be used with a fixed
voltage at startup in order to avoid issues with handover from the
bootloader if we suddenly drop the supply voltage on a device which has
not yet been claimed by a driver.  The expectation is that if the
voltage can vary then something that understands the hardware state
needs to be involved in maintaining the voltage.

Re: [PATCH] regulator: Fix machine constraints apply for min_uV != max_uV

[Jarkko Nikula]

On Fri, 3 Sep 2010 15:54:20 +0100
Mark Brown <broonie@opensource.wolfsonmicro.com> wrote:

> This is deliberate behaviour, apply_uV can only be used with a fixed
> voltage at startup in order to avoid issues with handover from the
> bootloader if we suddenly drop the supply voltage on a device which has
> not yet been claimed by a driver.  The expectation is that if the
> voltage can vary then something that understands the hardware state
> needs to be involved in maintaining the voltage.

Ah, true. It could be possible that machine constraints are specified
broader, like they include standby state voltage which is not enough to
probe & initialize the device etc.


-- 
Jarkko