32-rc2: runtime_pm is noisy

32-rc2: runtime_pm is noisy

[Pavel Machek]

Hi!

Maybe it is my config, but pm_runtime_idle is a bit too noisy on
zaurus... ok, I do have pm_debug & pm_verbose set... but still, is it
useful to print message about devices that do not support it? 
								Pavel

Serial: 8250/16550 driver, 4 ports, IRQ sharing disabled
PM: Adding info for platform:serial8250
platform serial8250: __pm_runtime_idle()!
platform serial8250: __pm_runtime_idle() returns -11!
...
serial8250 serial8250: __pm_runtime_idle()!
serial8250 serial8250: __pm_runtime_idle() returns -11!
...
sharpsl-nand sharpsl-nand: __pm_runtime_idle()!
sharpsl-nand sharpsl-nand: __pm_runtime_idle() returns -11!
...
gpio-keys gpio-keys: __pm_runtime_idle()!
gpio-keys gpio-keys: __pm_runtime_idle() returns -11!
PM: Adding info for No Bus:input1
input: matrix-keypad as /class/input/input1
PM: Adding info for No Bus:event1
matrix-keypad matrix-keypad: __pm_runtime_idle()!
matrix-keypad matrix-keypad: __pm_runtime_idle() returns -11!
...
ads7846 spi2.0: __pm_runtime_idle()!
ads7846 spi2.0: __pm_runtime_idle() returns -11!
APM Battery Driver
PM: Adding info for No Bus:hwmon1
max1111 spi2.2: __pm_runtime_idle()!
max1111 spi2.2: __pm_runtime_idle() returns -11!
...
pxa2xx-mci pxa2xx-mci.0: __pm_runtime_idle()!
pxa2xx-mci pxa2xx-mci.0: __pm_runtime_idle() returns -11!
PM: Adding info for No Bus:spitz:amber:charge
Registered led device: spitz:amber:charge
PM: Adding info for No Bus:spitz:green:hddactivity
Registered led device: spitz:green:hddactivity
leds-gpio leds-gpio: __pm_runtime_idle()!
leds-gpio leds-gpio: __pm_runtime_idle() returns -11!
...
sharpsl-pm sharpsl-pm: __pm_runtime_idle()!
sharpsl-pm sharpsl-pm: __pm_runtime_idle() returns -11!
...
ide-gd 0.0: __pm_runtime_idle()!
ide-gd 0.0: __pm_runtime_idle() returns -11!
ide-cs: hda: Vpp = 0.0
ide-cs 1.0: __pm_runtime_idle()!
ide-cs 1.0: __pm_runtime_idle() returns -11!

-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html

Re: 32-rc2: runtime_pm is noisy

[Rafael J. Wysocki]

On Tuesday 06 October 2009, Pavel Machek wrote:
> Hi!
> 
> Maybe it is my config,

Yes it is.

> but pm_runtime_idle is a bit too noisy on
> zaurus... ok, I do have pm_debug & pm_verbose

Exactly.

> set... but still, is it useful to print message about devices that do not
> support it?

It may be useful, because, for example, PCI can put devices without drivers
into low power states using this mechanism.

Still, perhaps we can drop the debug messages from __pm_runtime_idle().

Care to send a patch?

Rafael

Re: 32-rc2: runtime_pm is noisy

[Pavel Machek]

On Tue 2009-10-06 23:42:20, Rafael J. Wysocki wrote:
> On Tuesday 06 October 2009, Pavel Machek wrote:
> > Hi!
> > 
> > Maybe it is my config,
> 
> Yes it is.
> 
> > but pm_runtime_idle is a bit too noisy on
> > zaurus... ok, I do have pm_debug & pm_verbose
> 
> Exactly.
> 
> > set... but still, is it useful to print message about devices that do not
> > support it?
> 
> It may be useful, because, for example, PCI can put devices without drivers
> into low power states using this mechanism.
> 
> Still, perhaps we can drop the debug messages from __pm_runtime_idle().
> 
> Care to send a patch?

Yep, something like this?

---

pm_runtime_idle is somewhat noisy. Remove debug prints.

Signed-off-by: Pavel Machek <pavel@ucw.cz>

--- a/drivers/base/power/runtime.c
+++ b/drivers/base/power/runtime.c
@@ -51,8 +51,6 @@ static int __pm_runtime_idle(struct device *dev)
 {
 	int retval = 0;
 
-	dev_dbg(dev, "__pm_runtime_idle()!\n");
-
 	if (dev->power.runtime_error)
 		retval = -EINVAL;
 	else if (dev->power.idle_notification)
@@ -93,8 +91,6 @@ static int __pm_runtime_idle(struct device *dev)
 	wake_up_all(&dev->power.wait_queue);
 
  out:
-	dev_dbg(dev, "__pm_runtime_idle() returns %d!\n", retval);
-
 	return retval;
 }
 

-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html

Re: 32-rc2: runtime_pm is noisy

[Rafael J. Wysocki]

On Tuesday 13 October 2009, Pavel Machek wrote:
> On Tue 2009-10-06 23:42:20, Rafael J. Wysocki wrote:
> > On Tuesday 06 October 2009, Pavel Machek wrote:
> > > Hi!
> > > 
> > > Maybe it is my config,
> > 
> > Yes it is.
> > 
> > > but pm_runtime_idle is a bit too noisy on
> > > zaurus... ok, I do have pm_debug & pm_verbose
> > 
> > Exactly.
> > 
> > > set... but still, is it useful to print message about devices that do not
> > > support it?
> > 
> > It may be useful, because, for example, PCI can put devices without drivers
> > into low power states using this mechanism.
> > 
> > Still, perhaps we can drop the debug messages from __pm_runtime_idle().
> > 
> > Care to send a patch?
> 
> Yep, something like this?

Applied to suspend-2.6/linux-next.

Thanks,
Rafael


> ---
> 
> pm_runtime_idle is somewhat noisy. Remove debug prints.
> 
> Signed-off-by: Pavel Machek <pavel@ucw.cz>
> 
> --- a/drivers/base/power/runtime.c
> +++ b/drivers/base/power/runtime.c
> @@ -51,8 +51,6 @@ static int __pm_runtime_idle(struct device *dev)
>  {
>  	int retval = 0;
>  
> -	dev_dbg(dev, "__pm_runtime_idle()!\n");
> -
>  	if (dev->power.runtime_error)
>  		retval = -EINVAL;
>  	else if (dev->power.idle_notification)
> @@ -93,8 +91,6 @@ static int __pm_runtime_idle(struct device *dev)
>  	wake_up_all(&dev->power.wait_queue);
>  
>   out:
> -	dev_dbg(dev, "__pm_runtime_idle() returns %d!\n", retval);
> -
>  	return retval;
>  }
>  
> 
>