[PATCH] PM: lenient generic runtime pm callbacks

[PATCH] PM: lenient generic runtime pm callbacks

[Ohad Ben-Cohen]

Allow drivers, that belong to subsystems which use the generic
runtime pm callbacks, not to define runtime pm suspend/resume handlers,
by implicitly assuming success in such cases.

This is needed to eliminate nop handlers that would otherwise be
necessary by drivers which enable runtime pm, but don't need
to do anything when their devices are runtime-suspended/resumed.

Signed-off-by: Ohad Ben-Cohen <ohad@wizery.com>
---
 drivers/base/power/generic_ops.c |    4 ++--
 1 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/base/power/generic_ops.c b/drivers/base/power/generic_ops.c
index 4b29d49..81f2c84 100644
--- a/drivers/base/power/generic_ops.c
+++ b/drivers/base/power/generic_ops.c
@@ -46,7 +46,7 @@ int pm_generic_runtime_suspend(struct device *dev)
 	const struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL;
 	int ret;
 
-	ret = pm && pm->runtime_suspend ? pm->runtime_suspend(dev) : -EINVAL;
+	ret = pm && pm->runtime_suspend ? pm->runtime_suspend(dev) : 0;
 
 	return ret;
 }
@@ -65,7 +65,7 @@ int pm_generic_runtime_resume(struct device *dev)
 	const struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL;
 	int ret;
 
-	ret = pm && pm->runtime_resume ? pm->runtime_resume(dev) : -EINVAL;
+	ret = pm && pm->runtime_resume ? pm->runtime_resume(dev) : 0;
 
 	return ret;
 }
-- 
1.7.0.4

Re: [PATCH] PM: lenient generic runtime pm callbacks

[Kevin Hilman]

Ohad Ben-Cohen <ohad@wizery.com> writes:

> Allow drivers, that belong to subsystems which use the generic
> runtime pm callbacks, not to define runtime pm suspend/resume handlers,
> by implicitly assuming success in such cases.
>
> This is needed to eliminate nop handlers that would otherwise be
> necessary by drivers which enable runtime pm, but don't need
> to do anything when their devices are runtime-suspended/resumed.
>
> Signed-off-by: Ohad Ben-Cohen <ohad@wizery.com>

Acked-by: Kevin Hilman <khilman@deeprootsystems.com>

for runtime PM of on-chip devices on OMAP, this will cleanup several nop
handlers we have had to put in place.

FWIW, the 'no_callbacks' approach won't work for runtime PM of on-chip
devices for us since most of the devices will have callbacks, only some
will not need them.

Kevin

> ---
>  drivers/base/power/generic_ops.c |    4 ++--
>  1 files changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/base/power/generic_ops.c b/drivers/base/power/generic_ops.c
> index 4b29d49..81f2c84 100644
> --- a/drivers/base/power/generic_ops.c
> +++ b/drivers/base/power/generic_ops.c
> @@ -46,7 +46,7 @@ int pm_generic_runtime_suspend(struct device *dev)
>  	const struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL;
>  	int ret;
>  
> -	ret = pm && pm->runtime_suspend ? pm->runtime_suspend(dev) : -EINVAL;
> +	ret = pm && pm->runtime_suspend ? pm->runtime_suspend(dev) : 0;
>  
>  	return ret;
>  }
> @@ -65,7 +65,7 @@ int pm_generic_runtime_resume(struct device *dev)
>  	const struct dev_pm_ops *pm = dev->driver ? dev->driver->pm : NULL;
>  	int ret;
>  
> -	ret = pm && pm->runtime_resume ? pm->runtime_resume(dev) : -EINVAL;
> +	ret = pm && pm->runtime_resume ? pm->runtime_resume(dev) : 0;
>  
>  	return ret;
>  }

Re: [PATCH] PM: lenient generic runtime pm callbacks

[Mark Brown]

On Tue, Sep 07, 2010 at 01:20:42AM +0300, Ohad Ben-Cohen wrote:
> Allow drivers, that belong to subsystems which use the generic
> runtime pm callbacks, not to define runtime pm suspend/resume handlers,
> by implicitly assuming success in such cases.
> 
> This is needed to eliminate nop handlers that would otherwise be
> necessary by drivers which enable runtime pm, but don't need
> to do anything when their devices are runtime-suspended/resumed.
> 
> Signed-off-by: Ohad Ben-Cohen <ohad@wizery.com>

Acked-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

As I said previously this would also remove some dummy suspend functions
I need for some MFDs.

Re: [PATCH] PM: lenient generic runtime pm callbacks

[Rafael J. Wysocki]

On Wednesday, September 08, 2010, Mark Brown wrote:
> On Tue, Sep 07, 2010 at 01:20:42AM +0300, Ohad Ben-Cohen wrote:
> > Allow drivers, that belong to subsystems which use the generic
> > runtime pm callbacks, not to define runtime pm suspend/resume handlers,
> > by implicitly assuming success in such cases.
> > 
> > This is needed to eliminate nop handlers that would otherwise be
> > necessary by drivers which enable runtime pm, but don't need
> > to do anything when their devices are runtime-suspended/resumed.
> > 
> > Signed-off-by: Ohad Ben-Cohen <ohad@wizery.com>
> 
> Acked-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
> 
> As I said previously this would also remove some dummy suspend functions
> I need for some MFDs.

Applied to suspend-2.6/linux-next.

Thanks,
Rafael