[PATCH] PM: make freezer optional for suspend

[PATCH] PM: make freezer optional for suspend

[Johannes Berg]

This patch makes the freezer optional for suspend to allow the
system to work (or not work) like the original PMU suspend.

Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
---
In a quick test, playing sound across suspend failed to work,
it usually works when the freezer is enabled (unless you use
gstreamer.)

 kernel/power/Kconfig |   10 ++++++++++
 kernel/power/main.c  |    6 ++++++
 2 files changed, 16 insertions(+)

--- everything.orig/kernel/power/main.c	2007-11-13 20:00:16.406263508 +0100
+++ everything/kernel/power/main.c	2007-11-13 20:00:17.626268174 +0100
@@ -79,10 +79,12 @@ static int suspend_prepare(void)
 
 	pm_prepare_console();
 
+#ifdef CONFIG_SUSPEND_FREEZER
 	if (freeze_processes()) {
 		error = -EAGAIN;
 		goto Thaw;
 	}
+#endif
 
 	free_pages = global_page_state(NR_FREE_PAGES);
 	if (free_pages < FREE_PAGE_NUMBER) {
@@ -96,8 +98,10 @@ static int suspend_prepare(void)
 	if (!error)
 		return 0;
 
+#ifdef CONFIG_SUSPEND_FREEZER
  Thaw:
 	thaw_processes();
+#endif
 	pm_restore_console();
  Finish:
 	pm_notifier_call_chain(PM_POST_SUSPEND);
@@ -191,7 +195,9 @@ int suspend_devices_and_enter(suspend_st
  */
 static void suspend_finish(void)
 {
+#ifdef CONFIG_SUSPEND_FREEZER
 	thaw_processes();
+#endif
 	pm_restore_console();
 	pm_notifier_call_chain(PM_POST_SUSPEND);
 }
--- everything.orig/kernel/power/Kconfig	2007-11-13 20:00:16.476274685 +0100
+++ everything/kernel/power/Kconfig	2007-11-13 20:01:10.186257269 +0100
@@ -97,6 +97,16 @@ config SUSPEND
 	  powered and thus its contents are preserved, such as the
 	  suspend-to-RAM state (i.e. the ACPI S3 state).
 
+config SUSPEND_FREEZER
+	bool "Enable freezer for suspend to RAM/standby" if ADB_PMU || BROKEN
+	depends on SUSPEND
+	default y
+	help
+	  This allows you to turn off the freezer for suspend. If this is
+	  done, no tasks are frozen for suspend to RAM/standby.
+
+	  Turning OFF this setting is NOT recommended! If in doubt, say Y.
+
 config HIBERNATION_UP_POSSIBLE
 	bool
 	depends on X86 || PPC64_SWSUSP || PPC32

Re: [PATCH] PM: make freezer optional for suspend

[Rafael J. Wysocki]

On Tuesday, 13 of November 2007, Johannes Berg wrote:
> This patch makes the freezer optional for suspend to allow the
> system to work (or not work) like the original PMU suspend.
> 
> Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
> ---
> In a quick test, playing sound across suspend failed to work,
> it usually works when the freezer is enabled (unless you use
> gstreamer.)
> 
>  kernel/power/Kconfig |   10 ++++++++++
>  kernel/power/main.c  |    6 ++++++
>  2 files changed, 16 insertions(+)
> 
> --- everything.orig/kernel/power/main.c	2007-11-13 20:00:16.406263508 +0100
> +++ everything/kernel/power/main.c	2007-11-13 20:00:17.626268174 +0100
> @@ -79,10 +79,12 @@ static int suspend_prepare(void)
>  
>  	pm_prepare_console();
>  
> +#ifdef CONFIG_SUSPEND_FREEZER
>  	if (freeze_processes()) {
>  		error = -EAGAIN;
>  		goto Thaw;
>  	}
> +#endif
>  
>  	free_pages = global_page_state(NR_FREE_PAGES);
>  	if (free_pages < FREE_PAGE_NUMBER) {
> @@ -96,8 +98,10 @@ static int suspend_prepare(void)
>  	if (!error)
>  		return 0;
>  
> +#ifdef CONFIG_SUSPEND_FREEZER
>   Thaw:
>  	thaw_processes();
> +#endif
>  	pm_restore_console();
>   Finish:
>  	pm_notifier_call_chain(PM_POST_SUSPEND);
> @@ -191,7 +195,9 @@ int suspend_devices_and_enter(suspend_st
>   */
>  static void suspend_finish(void)
>  {
> +#ifdef CONFIG_SUSPEND_FREEZER
>  	thaw_processes();
> +#endif
>  	pm_restore_console();
>  	pm_notifier_call_chain(PM_POST_SUSPEND);
>  }
> --- everything.orig/kernel/power/Kconfig	2007-11-13 20:00:16.476274685 +0100
> +++ everything/kernel/power/Kconfig	2007-11-13 20:01:10.186257269 +0100
> @@ -97,6 +97,16 @@ config SUSPEND
>  	  powered and thus its contents are preserved, such as the
>  	  suspend-to-RAM state (i.e. the ACPI S3 state).
>  
> +config SUSPEND_FREEZER
> +	bool "Enable freezer for suspend to RAM/standby" if ADB_PMU || BROKEN

Hm, is ADB_PMU specific to powerpc?

> +	depends on SUSPEND
> +	default y
> +	help
> +	  This allows you to turn off the freezer for suspend. If this is
> +	  done, no tasks are frozen for suspend to RAM/standby.
> +
> +	  Turning OFF this setting is NOT recommended! If in doubt, say Y.
> +
>  config HIBERNATION_UP_POSSIBLE
>  	bool
>  	depends on X86 || PPC64_SWSUSP || PPC32

Greetings,
Rafael

Re: [PATCH] PM: make freezer optional for suspend

[Johannes Berg]

[-- Attachment #1.1: Type: text/plain, Size: 274 bytes --]


> > +config SUSPEND_FREEZER
> > +	bool "Enable freezer for suspend to RAM/standby" if ADB_PMU || BROKEN
> 
> Hm, is ADB_PMU specific to powerpc?

Yes, and it's the only code that allows you to suspend w/o the freezer
right now (via PMU specific ioctls)

johannes

[-- Attachment #1.2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 828 bytes --]

[-- Attachment #2: Type: text/plain, Size: 0 bytes --]

Re: [PATCH] PM: make freezer optional for suspend

[Rafael J. Wysocki]

On Wednesday, 14 of November 2007, Johannes Berg wrote:
> 
> > > +config SUSPEND_FREEZER
> > > +	bool "Enable freezer for suspend to RAM/standby" if ADB_PMU || BROKEN
> > 
> > Hm, is ADB_PMU specific to powerpc?
> 
> Yes, and it's the only code that allows you to suspend w/o the freezer
> right now (via PMU specific ioctls)

It's a bit confusing for someone who doesn't know what ADB_PMU is.

Perhaps it might depend on something like ARCH_CAN_DO_WITHOUT_FREEZER instead?

Rafael

Re: [PATCH] PM: make freezer optional for suspend

[Johannes Berg]

[-- Attachment #1.1: Type: text/plain, Size: 729 bytes --]


> > > > +config SUSPEND_FREEZER
> > > > +	bool "Enable freezer for suspend to RAM/standby" if ADB_PMU || BROKEN
> > > 
> > > Hm, is ADB_PMU specific to powerpc?
> > 
> > Yes, and it's the only code that allows you to suspend w/o the freezer
> > right now (via PMU specific ioctls)
> 
> It's a bit confusing for someone who doesn't know what ADB_PMU is.
> 
> Perhaps it might depend on something like ARCH_CAN_DO_WITHOUT_FREEZER instead?

It can't really. Things break. Ben and Paul just don't want to hear
about it. I made it depend on BROKEN so people can take it out and see
what breaks (to possibly fix it), and depend on ADB_PMU so they can have
their favourite setting without enabling BROKEN.

johannes

[-- Attachment #1.2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 828 bytes --]

[-- Attachment #2: Type: text/plain, Size: 0 bytes --]

Re: [PATCH] PM: make freezer optional for suspend

[Rafael J. Wysocki]

On Thursday, 15 of November 2007, Johannes Berg wrote:
> 
> > > > > +config SUSPEND_FREEZER
> > > > > +	bool "Enable freezer for suspend to RAM/standby" if ADB_PMU || BROKEN
> > > > 
> > > > Hm, is ADB_PMU specific to powerpc?
> > > 
> > > Yes, and it's the only code that allows you to suspend w/o the freezer
> > > right now (via PMU specific ioctls)
> > 
> > It's a bit confusing for someone who doesn't know what ADB_PMU is.
> > 
> > Perhaps it might depend on something like ARCH_CAN_DO_WITHOUT_FREEZER instead?
> 
> It can't really. Things break. Ben and Paul just don't want to hear
> about it. I made it depend on BROKEN so people can take it out and see
> what breaks (to possibly fix it), and depend on ADB_PMU so they can have
> their favourite setting without enabling BROKEN.

I understand that, but it's a bit confusing to put ADB_PMU, which is specific
to powerpc, into the generic configuration file.

Greetings,
Rafael

Re: [PATCH] PM: make freezer optional for suspend

[Johannes Berg]

[-- Attachment #1.1: Type: text/plain, Size: 695 bytes --]

> > > It's a bit confusing for someone who doesn't know what ADB_PMU is.
> > > 
> > > Perhaps it might depend on something like ARCH_CAN_DO_WITHOUT_FREEZER instead?
> > 
> > It can't really. Things break. Ben and Paul just don't want to hear
> > about it. I made it depend on BROKEN so people can take it out and see
> > what breaks (to possibly fix it), and depend on ADB_PMU so they can have
> > their favourite setting without enabling BROKEN.
> 
> I understand that, but it's a bit confusing to put ADB_PMU, which is specific
> to powerpc, into the generic configuration file.

Alright, I'll give it some thought and repost later. Maybe
ARCH_DOESNT_WANT_FREEZER ;)

johannes

[-- Attachment #1.2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 828 bytes --]

[-- Attachment #2: Type: text/plain, Size: 0 bytes --]

Re: Re: [PATCH] PM: make freezer optional for suspend

[Pavel Machek]

Hi!

> > > It can't really. Things break. Ben and Paul just don't want to hear
> > > about it. I made it depend on BROKEN so people can take it out and see
> > > what breaks (to possibly fix it), and depend on ADB_PMU so they can have
> > > their favourite setting without enabling BROKEN.
> > 
> > I understand that, but it's a bit confusing to put ADB_PMU, which is specific
> > to powerpc, into the generic configuration file.
> 
> Alright, I'll give it some thought and repost later. Maybe
> ARCH_DOESNT_WANT_FREEZER ;)

What about just 

a) reverse the logic, so that it is C_SUSPEND_NO_FREEZER

b) hardcode it in ppc-specific config?

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

Re: Re: [PATCH] PM: make freezer optional for suspend

[Johannes Berg]

[-- Attachment #1.1: Type: text/plain, Size: 847 bytes --]


On Sun, 2007-11-18 at 18:21 +0000, Pavel Machek wrote:
> Hi!
> 
> > > > It can't really. Things break. Ben and Paul just don't want to hear
> > > > about it. I made it depend on BROKEN so people can take it out and see
> > > > what breaks (to possibly fix it), and depend on ADB_PMU so they can have
> > > > their favourite setting without enabling BROKEN.
> > > 
> > > I understand that, but it's a bit confusing to put ADB_PMU, which is specific
> > > to powerpc, into the generic configuration file.
> > 
> > Alright, I'll give it some thought and repost later. Maybe
> > ARCH_DOESNT_WANT_FREEZER ;)
> 
> What about just 
> 
> a) reverse the logic, so that it is C_SUSPEND_NO_FREEZER
> 
> b) hardcode it in ppc-specific config?

I also wanted people interested in debugging things w/o the freezer to
use it.

johannes

[-- Attachment #1.2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 828 bytes --]

[-- Attachment #2: Type: text/plain, Size: 0 bytes --]

Re: Re: [PATCH] PM: make freezer optional for suspend

[Pavel Machek]

Hi!

> > > > > It can't really. Things break. Ben and Paul just don't want to hear
> > > > > about it. I made it depend on BROKEN so people can take it out and see
> > > > > what breaks (to possibly fix it), and depend on ADB_PMU so they can have
> > > > > their favourite setting without enabling BROKEN.
> > > > 
> > > > I understand that, but it's a bit confusing to put ADB_PMU, which is specific
> > > > to powerpc, into the generic configuration file.
> > > 
> > > Alright, I'll give it some thought and repost later. Maybe
> > > ARCH_DOESNT_WANT_FREEZER ;)
> > 
> > What about just 
> > 
> > a) reverse the logic, so that it is C_SUSPEND_NO_FREEZER
> > 
> > b) hardcode it in ppc-specific config?
> 
> I also wanted people interested in debugging things w/o the freezer to
> use it.

I'd say that people interested in debugging can just edit the config
manually :-).
								Pavel
-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html