[PATCH] i2c: Hook up runtime PM support

[PATCH] i2c: Hook up runtime PM support

[Mark Brown]

Allow I2C drivers to make use of the runtime PM framework by adding
bus implementations of the runtime PM operations. These simply
immediately suspend when the device is idle. The runtime PM framework
provides drivers with off the shelf refcounts for enables and sysfs
control for managing runtime suspend from userspace so is useful even
without meaningful input from the bus.

Signed-off-by: Mark Brown <broonie-yzvPICuk2AATkU/dhu1WVueM+bqZidxxQQ4Iyu8u01E@public.gmane.org>
---

Changed to allow drivers to use the idle callback to veto suspend.

 drivers/i2c/i2c-core.c |   50 ++++++++++++++++++++++++++++++++++++++++++++++++
 1 files changed, 50 insertions(+), 0 deletions(-)

diff --git a/drivers/i2c/i2c-core.c b/drivers/i2c/i2c-core.c
index 10be7b5..4131698 100644
--- a/drivers/i2c/i2c-core.c
+++ b/drivers/i2c/i2c-core.c
@@ -34,6 +34,7 @@
 #include <linux/hardirq.h>
 #include <linux/irqflags.h>
 #include <linux/rwsem.h>
+#include <linux/pm_runtime.h>
 #include <asm/uaccess.h>
 
 #include "i2c-core.h"
@@ -184,6 +185,52 @@ static int i2c_device_pm_resume(struct device *dev)
 #define i2c_device_pm_resume	NULL
 #endif
 
+#ifdef CONFIG_PM_RUNTIME
+static int i2c_device_runtime_suspend(struct device *dev)
+{
+	const struct dev_pm_ops *pm;
+
+	if (!dev->driver)
+		return 0;
+	pm = dev->driver->pm;
+	if (!pm || !pm->runtime_suspend)
+		return 0;
+	return pm->runtime_suspend(dev);
+}
+
+static int i2c_device_runtime_resume(struct device *dev)
+{
+	const struct dev_pm_ops *pm;
+
+	if (!dev->driver)
+		return 0;
+	pm = dev->driver->pm;
+	if (!pm || !pm->runtime_resume)
+		return 0;
+	return pm->runtime_resume(dev);
+}
+
+static int i2c_device_runtime_idle(struct device *dev)
+{
+	const struct dev_pm_ops *pm = NULL;
+	int ret;
+
+	if (dev->driver)
+		pm = dev->driver->pm;
+	if (pm && pm->runtime_idle) {
+		ret = pm->runtime_idle(dev);
+		if (ret)
+			return ret;
+	}
+
+	return pm_runtime_suspend(dev);
+}
+#else
+#define i2c_device_runtime_suspend	NULL
+#define i2c_device_runtime_resume	NULL
+#define i2c_device_runtime_idle		NULL
+#endif
+
 static int i2c_device_suspend(struct device *dev, pm_message_t mesg)
 {
 	struct i2c_client *client = i2c_verify_client(dev);
@@ -251,6 +298,9 @@ static const struct attribute_group *i2c_dev_attr_groups[] = {
 static const struct dev_pm_ops i2c_device_pm_ops = {
 	.suspend = i2c_device_pm_suspend,
 	.resume = i2c_device_pm_resume,
+	.runtime_suspend = i2c_device_runtime_suspend,
+	.runtime_resume = i2c_device_runtime_resume,
+	.runtime_idle = i2c_device_runtime_idle,
 };
 
 struct bus_type i2c_bus_type = {
-- 
1.6.6

Re: [PATCH] i2c: Hook up runtime PM support

[Rafael J. Wysocki]

On Friday 05 February 2010, Mark Brown wrote:
> Allow I2C drivers to make use of the runtime PM framework by adding
> bus implementations of the runtime PM operations. These simply
> immediately suspend when the device is idle. The runtime PM framework
> provides drivers with off the shelf refcounts for enables and sysfs
> control for managing runtime suspend from userspace so is useful even
> without meaningful input from the bus.
> 
> Signed-off-by: Mark Brown <broonie-yzvPICuk2AATkU/dhu1WVueM+bqZidxxQQ4Iyu8u01E@public.gmane.org>

Acked-by: Rafael J. Wysocki <rjw-KKrjLPT3xs0@public.gmane.org>

> ---
> 
> Changed to allow drivers to use the idle callback to veto suspend.
> 
>  drivers/i2c/i2c-core.c |   50 ++++++++++++++++++++++++++++++++++++++++++++++++
>  1 files changed, 50 insertions(+), 0 deletions(-)
> 
> diff --git a/drivers/i2c/i2c-core.c b/drivers/i2c/i2c-core.c
> index 10be7b5..4131698 100644
> --- a/drivers/i2c/i2c-core.c
> +++ b/drivers/i2c/i2c-core.c
> @@ -34,6 +34,7 @@
>  #include <linux/hardirq.h>
>  #include <linux/irqflags.h>
>  #include <linux/rwsem.h>
> +#include <linux/pm_runtime.h>
>  #include <asm/uaccess.h>
>  
>  #include "i2c-core.h"
> @@ -184,6 +185,52 @@ static int i2c_device_pm_resume(struct device *dev)
>  #define i2c_device_pm_resume	NULL
>  #endif
>  
> +#ifdef CONFIG_PM_RUNTIME
> +static int i2c_device_runtime_suspend(struct device *dev)
> +{
> +	const struct dev_pm_ops *pm;
> +
> +	if (!dev->driver)
> +		return 0;
> +	pm = dev->driver->pm;
> +	if (!pm || !pm->runtime_suspend)
> +		return 0;
> +	return pm->runtime_suspend(dev);
> +}
> +
> +static int i2c_device_runtime_resume(struct device *dev)
> +{
> +	const struct dev_pm_ops *pm;
> +
> +	if (!dev->driver)
> +		return 0;
> +	pm = dev->driver->pm;
> +	if (!pm || !pm->runtime_resume)
> +		return 0;
> +	return pm->runtime_resume(dev);
> +}
> +
> +static int i2c_device_runtime_idle(struct device *dev)
> +{
> +	const struct dev_pm_ops *pm = NULL;
> +	int ret;
> +
> +	if (dev->driver)
> +		pm = dev->driver->pm;
> +	if (pm && pm->runtime_idle) {
> +		ret = pm->runtime_idle(dev);
> +		if (ret)
> +			return ret;
> +	}
> +
> +	return pm_runtime_suspend(dev);
> +}
> +#else
> +#define i2c_device_runtime_suspend	NULL
> +#define i2c_device_runtime_resume	NULL
> +#define i2c_device_runtime_idle		NULL
> +#endif
> +
>  static int i2c_device_suspend(struct device *dev, pm_message_t mesg)
>  {
>  	struct i2c_client *client = i2c_verify_client(dev);
> @@ -251,6 +298,9 @@ static const struct attribute_group *i2c_dev_attr_groups[] = {
>  static const struct dev_pm_ops i2c_device_pm_ops = {
>  	.suspend = i2c_device_pm_suspend,
>  	.resume = i2c_device_pm_resume,
> +	.runtime_suspend = i2c_device_runtime_suspend,
> +	.runtime_resume = i2c_device_runtime_resume,
> +	.runtime_idle = i2c_device_runtime_idle,
>  };
>  
>  struct bus_type i2c_bus_type = {
> 

Re: [PATCH] i2c: Hook up runtime PM support

[Jean Delvare]

Hi Mark,

On Fri,  5 Feb 2010 12:30:11 +0000, Mark Brown wrote:
> Allow I2C drivers to make use of the runtime PM framework by adding
> bus implementations of the runtime PM operations. These simply
> immediately suspend when the device is idle. The runtime PM framework
> provides drivers with off the shelf refcounts for enables and sysfs
> control for managing runtime suspend from userspace so is useful even
> without meaningful input from the bus.
> 
> Signed-off-by: Mark Brown <broonie-yzvPICuk2AATkU/dhu1WVueM+bqZidxxQQ4Iyu8u01E@public.gmane.org>
> ---
> 
> Changed to allow drivers to use the idle callback to veto suspend.
> 
>  drivers/i2c/i2c-core.c |   50 ++++++++++++++++++++++++++++++++++++++++++++++++
>  1 files changed, 50 insertions(+), 0 deletions(-)
> 
> diff --git a/drivers/i2c/i2c-core.c b/drivers/i2c/i2c-core.c
> index 10be7b5..4131698 100644
> --- a/drivers/i2c/i2c-core.c
> +++ b/drivers/i2c/i2c-core.c
> @@ -34,6 +34,7 @@
>  #include <linux/hardirq.h>
>  #include <linux/irqflags.h>
>  #include <linux/rwsem.h>
> +#include <linux/pm_runtime.h>
>  #include <asm/uaccess.h>
>  
>  #include "i2c-core.h"
> @@ -184,6 +185,52 @@ static int i2c_device_pm_resume(struct device *dev)
>  #define i2c_device_pm_resume	NULL
>  #endif
>  
> +#ifdef CONFIG_PM_RUNTIME
> +static int i2c_device_runtime_suspend(struct device *dev)
> +{
> +	const struct dev_pm_ops *pm;
> +
> +	if (!dev->driver)
> +		return 0;
> +	pm = dev->driver->pm;
> +	if (!pm || !pm->runtime_suspend)
> +		return 0;
> +	return pm->runtime_suspend(dev);
> +}
> +
> +static int i2c_device_runtime_resume(struct device *dev)
> +{
> +	const struct dev_pm_ops *pm;
> +
> +	if (!dev->driver)
> +		return 0;
> +	pm = dev->driver->pm;
> +	if (!pm || !pm->runtime_resume)
> +		return 0;
> +	return pm->runtime_resume(dev);
> +}
> +
> +static int i2c_device_runtime_idle(struct device *dev)
> +{
> +	const struct dev_pm_ops *pm = NULL;
> +	int ret;
> +
> +	if (dev->driver)
> +		pm = dev->driver->pm;
> +	if (pm && pm->runtime_idle) {
> +		ret = pm->runtime_idle(dev);
> +		if (ret)
> +			return ret;
> +	}
> +
> +	return pm_runtime_suspend(dev);
> +}
> +#else
> +#define i2c_device_runtime_suspend	NULL
> +#define i2c_device_runtime_resume	NULL
> +#define i2c_device_runtime_idle		NULL
> +#endif
> +
>  static int i2c_device_suspend(struct device *dev, pm_message_t mesg)
>  {
>  	struct i2c_client *client = i2c_verify_client(dev);
> @@ -251,6 +298,9 @@ static const struct attribute_group *i2c_dev_attr_groups[] = {
>  static const struct dev_pm_ops i2c_device_pm_ops = {
>  	.suspend = i2c_device_pm_suspend,
>  	.resume = i2c_device_pm_resume,
> +	.runtime_suspend = i2c_device_runtime_suspend,
> +	.runtime_resume = i2c_device_runtime_resume,
> +	.runtime_idle = i2c_device_runtime_idle,
>  };
>  
>  struct bus_type i2c_bus_type = {

Applied, thanks. I am a little surprised to see that these functions
have nothing i2c-specific, so I am wondering why we have to duplicate
them in every bus type... Shouldn't the functions above be part of
drivers/base/power/runtime.c and exported so that all bus types that
want them can reuse them?

-- 
Jean Delvare

Re: [PATCH] i2c: Hook up runtime PM support

[Mark Brown]

On Mon, Feb 15, 2010 at 07:14:09PM +0100, Jean Delvare wrote:

> Applied, thanks. I am a little surprised to see that these functions
> have nothing i2c-specific, so I am wondering why we have to duplicate
> them in every bus type... Shouldn't the functions above be part of
> drivers/base/power/runtime.c and exported so that all bus types that
> want them can reuse them?

I tend to agree - in fact I'd been a little surprised the default for
buses that don't provide anything was to refuse to do runtime PM (though
I can see the transition issues when a bus does want to go and do its
own thing).

My actual plan here was to implement this for a couple of buses and then
present a patch factoring out the common code.

Re: [PATCH] i2c: Hook up runtime PM support

[Rafael J. Wysocki]

On Monday 15 February 2010, Mark Brown wrote:
> On Mon, Feb 15, 2010 at 07:14:09PM +0100, Jean Delvare wrote:
> 
> > Applied, thanks. I am a little surprised to see that these functions
> > have nothing i2c-specific, so I am wondering why we have to duplicate
> > them in every bus type... Shouldn't the functions above be part of
> > drivers/base/power/runtime.c and exported so that all bus types that
> > want them can reuse them?

In fact this is the first bus type that doesn't need anything specific in these
routines so fat.

> I tend to agree - in fact I'd been a little surprised the default for
> buses that don't provide anything was to refuse to do runtime PM (though
> I can see the transition issues when a bus does want to go and do its
> own thing).
> 
> My actual plan here was to implement this for a couple of buses and then
> present a patch factoring out the common code.

That's a good approach IMO.

Rafael

Re: [PATCH] i2c: Hook up runtime PM support

[Jean Delvare]

On Mon, 15 Feb 2010 20:08:24 +0100, Rafael J. Wysocki wrote:
> On Monday 15 February 2010, Mark Brown wrote:
> > On Mon, Feb 15, 2010 at 07:14:09PM +0100, Jean Delvare wrote:
> > 
> > > Applied, thanks. I am a little surprised to see that these functions
> > > have nothing i2c-specific, so I am wondering why we have to duplicate
> > > them in every bus type... Shouldn't the functions above be part of
> > > drivers/base/power/runtime.c and exported so that all bus types that
> > > want them can reuse them?
> 
> In fact this is the first bus type that doesn't need anything specific in these
> routines so fat.
> 
> > I tend to agree - in fact I'd been a little surprised the default for
> > buses that don't provide anything was to refuse to do runtime PM (though
> > I can see the transition issues when a bus does want to go and do its
> > own thing).
> > 
> > My actual plan here was to implement this for a couple of buses and then
> > present a patch factoring out the common code.
> 
> That's a good approach IMO.

OK, fine with me.

-- 
Jean Delvare

Re: [PATCH] i2c: Hook up runtime PM support

[Mark Brown]

On Mon, Feb 15, 2010 at 08:08:24PM +0100, Rafael J. Wysocki wrote:
> On Monday 15 February 2010, Mark Brown wrote:
> > On Mon, Feb 15, 2010 at 07:14:09PM +0100, Jean Delvare wrote:

> > > Applied, thanks. I am a little surprised to see that these functions
> > > have nothing i2c-specific, so I am wondering why we have to duplicate
> > > them in every bus type... Shouldn't the functions above be part of
> > > drivers/base/power/runtime.c and exported so that all bus types that
> > > want them can reuse them?

> In fact this is the first bus type that doesn't need anything specific in these
> routines so fat.

Not really - the platform bus is the same, it's just that some platforms
are also able to do some neat stuff with the information they glean via
runtime PM.  The platform bus is actually a bit of a problem here since
it gets used both for on-SoC devices where that sort of stuff is
possible and also for things like MFDs and random memory mapped things
on the system, meaning that drivers can't take advantage of runtime PM
unless someone goes through and does something per-SoC which is a bit
inconvnient.

There's other buses like SPI that are in the same boat as I2C - they
don't really have a cohesive bus level power management structure, or
the power management is fully handled by with noop methods on the bus
and holding the parent open until all the children are idle.

[PATCH] i2c: Hook up runtime PM support

[Mark Brown]

Allow I2C drivers to make use of the runtime PM framework by adding
bus implementations of the runtime PM operations. These simply
immediately suspend when the device is idle. The runtime PM framework
provides drivers with off the shelf refcounts for enables and sysfs
control for managing runtime suspend from userspace so is useful even
without meaningful input from the bus.

Signed-off-by: Mark Brown <broonie-yzvPICuk2AATkU/dhu1WVueM+bqZidxxQQ4Iyu8u01E@public.gmane.org>
---
 drivers/i2c/i2c-core.c |   41 +++++++++++++++++++++++++++++++++++++++++
 1 files changed, 41 insertions(+), 0 deletions(-)

diff --git a/drivers/i2c/i2c-core.c b/drivers/i2c/i2c-core.c
index 10be7b5..985eaac 100644
--- a/drivers/i2c/i2c-core.c
+++ b/drivers/i2c/i2c-core.c
@@ -34,6 +34,7 @@
 #include <linux/hardirq.h>
 #include <linux/irqflags.h>
 #include <linux/rwsem.h>
+#include <linux/pm_runtime.h>
 #include <asm/uaccess.h>
 
 #include "i2c-core.h"
@@ -184,6 +185,43 @@ static int i2c_device_pm_resume(struct device *dev)
 #define i2c_device_pm_resume	NULL
 #endif
 
+#ifdef CONFIG_PM_RUNTIME
+static int i2c_device_runtime_suspend(struct device *dev)
+{
+	const struct dev_pm_ops *pm;
+
+	if (!dev->driver)
+		return 0;
+	pm = dev->driver->pm;
+	if (!pm || !pm->runtime_suspend)
+		return 0;
+	return pm->runtime_suspend(dev);
+}
+
+static int i2c_device_runtime_resume(struct device *dev)
+{
+	const struct dev_pm_ops *pm;
+
+	if (!dev->driver)
+		return 0;
+	pm = dev->driver->pm;
+	if (!pm || !pm->runtime_resume)
+		return 0;
+	return pm->runtime_resume(dev);
+}
+
+static int i2c_device_runtime_idle(struct device *dev)
+{
+	/* I2C devices are very independent of each other so we
+	 * suspend them immediately. */
+	return pm_runtime_suspend(dev);
+}
+#else
+#define i2c_device_runtime_suspend	NULL
+#define i2c_device_runtime_resume	NULL
+#define i2c_device_runtime_idle		NULL
+#endif
+
 static int i2c_device_suspend(struct device *dev, pm_message_t mesg)
 {
 	struct i2c_client *client = i2c_verify_client(dev);
@@ -251,6 +289,9 @@ static const struct attribute_group *i2c_dev_attr_groups[] = {
 static const struct dev_pm_ops i2c_device_pm_ops = {
 	.suspend = i2c_device_pm_suspend,
 	.resume = i2c_device_pm_resume,
+	.runtime_suspend = i2c_device_runtime_suspend,
+	.runtime_resume = i2c_device_runtime_resume,
+	.runtime_idle = i2c_device_runtime_idle,
 };
 
 struct bus_type i2c_bus_type = {
-- 
1.6.6

Re: [PATCH] i2c: Hook up runtime PM support

[Rafael J. Wysocki]

On Wednesday 03 February 2010, Mark Brown wrote:
> Allow I2C drivers to make use of the runtime PM framework by adding
> bus implementations of the runtime PM operations. These simply
> immediately suspend when the device is idle.

Perhaps it would be a good idea to give the driver a chance to veto
the suspend by calling its _idle callback?  We do that for PCI and turns out to
be quite useful.

> The runtime PM framework provides drivers with off the shelf refcounts for
> enables and sysfs control for managing runtime suspend from userspace so is
> useful even without meaningful input from the bus.
> 
> Signed-off-by: Mark Brown <broonie-yzvPICuk2AATkU/dhu1WVueM+bqZidxxQQ4Iyu8u01E@public.gmane.org>
> ---
>  drivers/i2c/i2c-core.c |   41 +++++++++++++++++++++++++++++++++++++++++
>  1 files changed, 41 insertions(+), 0 deletions(-)
> 
> diff --git a/drivers/i2c/i2c-core.c b/drivers/i2c/i2c-core.c
> index 10be7b5..985eaac 100644
> --- a/drivers/i2c/i2c-core.c
> +++ b/drivers/i2c/i2c-core.c
> @@ -34,6 +34,7 @@
>  #include <linux/hardirq.h>
>  #include <linux/irqflags.h>
>  #include <linux/rwsem.h>
> +#include <linux/pm_runtime.h>
>  #include <asm/uaccess.h>
>  
>  #include "i2c-core.h"
> @@ -184,6 +185,43 @@ static int i2c_device_pm_resume(struct device *dev)
>  #define i2c_device_pm_resume	NULL
>  #endif
>  
> +#ifdef CONFIG_PM_RUNTIME
> +static int i2c_device_runtime_suspend(struct device *dev)
> +{
> +	const struct dev_pm_ops *pm;
> +
> +	if (!dev->driver)
> +		return 0;
> +	pm = dev->driver->pm;
> +	if (!pm || !pm->runtime_suspend)
> +		return 0;
> +	return pm->runtime_suspend(dev);
> +}
> +
> +static int i2c_device_runtime_resume(struct device *dev)
> +{
> +	const struct dev_pm_ops *pm;
> +
> +	if (!dev->driver)
> +		return 0;
> +	pm = dev->driver->pm;
> +	if (!pm || !pm->runtime_resume)
> +		return 0;
> +	return pm->runtime_resume(dev);
> +}
> +
> +static int i2c_device_runtime_idle(struct device *dev)
> +{
> +	/* I2C devices are very independent of each other so we
> +	 * suspend them immediately. */
> +	return pm_runtime_suspend(dev);
> +}
> +#else
> +#define i2c_device_runtime_suspend	NULL
> +#define i2c_device_runtime_resume	NULL
> +#define i2c_device_runtime_idle		NULL
> +#endif
> +
>  static int i2c_device_suspend(struct device *dev, pm_message_t mesg)
>  {
>  	struct i2c_client *client = i2c_verify_client(dev);
> @@ -251,6 +289,9 @@ static const struct attribute_group *i2c_dev_attr_groups[] = {
>  static const struct dev_pm_ops i2c_device_pm_ops = {
>  	.suspend = i2c_device_pm_suspend,
>  	.resume = i2c_device_pm_resume,
> +	.runtime_suspend = i2c_device_runtime_suspend,
> +	.runtime_resume = i2c_device_runtime_resume,
> +	.runtime_idle = i2c_device_runtime_idle,
>  };
>  
>  struct bus_type i2c_bus_type = {
> 

Rafael