[PATCH] PM / Hibernate: Thaw processes in SNAPSHOT_CREATE_IMAGE ioctl error/test paths

[PATCH] PM / Hibernate: Thaw processes in SNAPSHOT_CREATE_IMAGE ioctl error/test paths

[Srivatsa S. Bhat]

Commit 2aede851ddf08666f68ffc17be446420e9d2a056 (PM / Hibernate: Freeze
kernel threads after preallocating memory) moved the freezing of kernel
threads to hibernation_snapshot() function.

So now, if the call to hibernation_snapshot() fails or returns early due
to a successful hibernation test, the caller has to thaw processes to
ensure that the system gets back to its original state.

But in SNAPSHOT_CREATE_IMAGE hibernation ioctl, the caller does not thaw
processes. Fix this issue. But note that if hibernation_snapshot() returned
due to a successful freezer test, we still send the value of 'in_suspend'
(which is now 0) to userspace, because we are not in an error path per-se,
and moreover, the value of in_suspend correctly depicts the situation here.

Signed-off-by: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com>
---

 kernel/power/hibernate.c |    2 +-
 kernel/power/power.h     |    2 ++
 kernel/power/user.c      |   13 ++++++++++---
 3 files changed, 13 insertions(+), 4 deletions(-)

diff --git a/kernel/power/hibernate.c b/kernel/power/hibernate.c
index a6b0503..624538a 100644
--- a/kernel/power/hibernate.c
+++ b/kernel/power/hibernate.c
@@ -55,7 +55,7 @@ enum {
 
 static int hibernation_mode = HIBERNATION_SHUTDOWN;
 
-static bool freezer_test_done;
+bool freezer_test_done;
 
 static const struct platform_hibernation_ops *hibernation_ops;
 
diff --git a/kernel/power/power.h b/kernel/power/power.h
index 23a2db1..0c4defe 100644
--- a/kernel/power/power.h
+++ b/kernel/power/power.h
@@ -50,6 +50,8 @@ static inline char *check_image_kernel(struct swsusp_info *info)
 #define SPARE_PAGES	((1024 * 1024) >> PAGE_SHIFT)
 
 /* kernel/power/hibernate.c */
+extern bool freezer_test_done;
+
 extern int hibernation_snapshot(int platform_mode);
 extern int hibernation_restore(int platform_mode);
 extern int hibernation_platform_enter(void);
diff --git a/kernel/power/user.c b/kernel/power/user.c
index 6d8f535..b98b7e1 100644
--- a/kernel/power/user.c
+++ b/kernel/power/user.c
@@ -283,10 +283,17 @@ static long snapshot_ioctl(struct file *filp, unsigned int cmd,
 		}
 		pm_restore_gfp_mask();
 		error = hibernation_snapshot(data->platform_support);
-		if (!error)
+		if (error) {
+			thaw_processes();
+		} else {
 			error = put_user(in_suspend, (int __user *)arg);
-		if (!error)
-			data->ready = 1;
+			if (!error && !freezer_test_done)
+				data->ready = 1;
+			if (freezer_test_done) {
+				freezer_test_done = false;
+				thaw_processes();
+			}
+		}
 		break;
 
 	case SNAPSHOT_ATOMIC_RESTORE:

Re: [PATCH] PM / Hibernate: Thaw processes in SNAPSHOT_CREATE_IMAGE ioctl error/test paths

[Tejun Heo]

On Thu, Dec 01, 2011 at 06:45:37PM +0530, Srivatsa S. Bhat wrote:
> Commit 2aede851ddf08666f68ffc17be446420e9d2a056 (PM / Hibernate: Freeze
> kernel threads after preallocating memory) moved the freezing of kernel
> threads to hibernation_snapshot() function.
> 
> So now, if the call to hibernation_snapshot() fails or returns early due
> to a successful hibernation test, the caller has to thaw processes to
> ensure that the system gets back to its original state.
> 
> But in SNAPSHOT_CREATE_IMAGE hibernation ioctl, the caller does not thaw
> processes. Fix this issue. But note that if hibernation_snapshot() returned
> due to a successful freezer test, we still send the value of 'in_suspend'
> (which is now 0) to userspace, because we are not in an error path per-se,
> and moreover, the value of in_suspend correctly depicts the situation here.
> 
> Signed-off-by: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com>

Is this still applicable on top of pm-freezer branch?

Thanks.

-- 
tejun

Re: [PATCH] PM / Hibernate: Thaw processes in SNAPSHOT_CREATE_IMAGE ioctl error/test paths

[Srivatsa S. Bhat]

On 12/01/2011 09:17 PM, Tejun Heo wrote:

> On Thu, Dec 01, 2011 at 06:45:37PM +0530, Srivatsa S. Bhat wrote:
>> Commit 2aede851ddf08666f68ffc17be446420e9d2a056 (PM / Hibernate: Freeze
>> kernel threads after preallocating memory) moved the freezing of kernel
>> threads to hibernation_snapshot() function.
>>
>> So now, if the call to hibernation_snapshot() fails or returns early due
>> to a successful hibernation test, the caller has to thaw processes to
>> ensure that the system gets back to its original state.
>>
>> But in SNAPSHOT_CREATE_IMAGE hibernation ioctl, the caller does not thaw
>> processes. Fix this issue. But note that if hibernation_snapshot() returned
>> due to a successful freezer test, we still send the value of 'in_suspend'
>> (which is now 0) to userspace, because we are not in an error path per-se,
>> and moreover, the value of in_suspend correctly depicts the situation here.
>>
>> Signed-off-by: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com>
> 
> Is this still applicable on top of pm-freezer branch?
> 


Well, yes, but not in this form. Thanks for bringing that up!

If the call to hibernation_snapshot() returns an error, we need not do
anything since your freezer changes handle that case and call thaw_processes()
automatically.
But in case of return due to a successful freezer test, we would need to
explicitly thaw processes since the freezer successfully froze them in the
previous step.

So the following patch would do that, for the pm-freezer branch:

---
From: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com>
Subject: [PATCH] PM / Hibernate: Thaw processes in SNAPSHOT_CREATE_IMAGE ioctl test path

Commit 2aede851ddf08666f68ffc17be446420e9d2a056 (PM / Hibernate: Freeze
kernel threads after preallocating memory) moved the freezing of kernel
threads to hibernation_snapshot() function.

So now, if the call to hibernation_snapshot() returns early due to a
successful hibernation test, the caller has to thaw processes to ensure
that the system gets back to its original state.

But in SNAPSHOT_CREATE_IMAGE hibernation ioctl, the caller does not thaw
processes in case hibernation_snapshot() returned due to a successful
freezer test. Fix this issue. But note we still send the value of 'in_suspend'
(which is now 0) to userspace, because we are not in an error path per-se,
and moreover, the value of in_suspend correctly depicts the situation here.

Signed-off-by: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com>
---

 kernel/power/hibernate.c |    2 +-
 kernel/power/power.h     |    2 ++
 kernel/power/user.c      |   11 ++++++++---
 3 files changed, 11 insertions(+), 4 deletions(-)

diff --git a/kernel/power/hibernate.c b/kernel/power/hibernate.c
index 80c2d90..57bab63 100644
--- a/kernel/power/hibernate.c
+++ b/kernel/power/hibernate.c
@@ -55,7 +55,7 @@ enum {
 
 static int hibernation_mode = HIBERNATION_SHUTDOWN;
 
-static bool freezer_test_done;
+bool freezer_test_done;
 
 static const struct platform_hibernation_ops *hibernation_ops;
 
diff --git a/kernel/power/power.h b/kernel/power/power.h
index 23a2db1..0c4defe 100644
--- a/kernel/power/power.h
+++ b/kernel/power/power.h
@@ -50,6 +50,8 @@ static inline char *check_image_kernel(struct swsusp_info *info)
 #define SPARE_PAGES	((1024 * 1024) >> PAGE_SHIFT)
 
 /* kernel/power/hibernate.c */
+extern bool freezer_test_done;
+
 extern int hibernation_snapshot(int platform_mode);
 extern int hibernation_restore(int platform_mode);
 extern int hibernation_platform_enter(void);
diff --git a/kernel/power/user.c b/kernel/power/user.c
index 7cc3f5b..c202e2e 100644
--- a/kernel/power/user.c
+++ b/kernel/power/user.c
@@ -281,10 +281,15 @@ static long snapshot_ioctl(struct file *filp, unsigned int cmd,
 		}
 		pm_restore_gfp_mask();
 		error = hibernation_snapshot(data->platform_support);
-		if (!error)
+		if (!error) {
 			error = put_user(in_suspend, (int __user *)arg);
-		if (!error)
-			data->ready = 1;
+			if (!error && !freezer_test_done)
+				data->ready = 1;
+			if (freezer_test_done) {
+				freezer_test_done = false;
+				thaw_processes();
+			}
+		}
 		break;
 
 	case SNAPSHOT_ATOMIC_RESTORE:

Re: [PATCH] PM / Hibernate: Thaw processes in SNAPSHOT_CREATE_IMAGE ioctl error/test paths

[Rafael J. Wysocki]

On Thursday, December 01, 2011, Srivatsa S. Bhat wrote:
> On 12/01/2011 09:17 PM, Tejun Heo wrote:
> 
> > On Thu, Dec 01, 2011 at 06:45:37PM +0530, Srivatsa S. Bhat wrote:
> >> Commit 2aede851ddf08666f68ffc17be446420e9d2a056 (PM / Hibernate: Freeze
> >> kernel threads after preallocating memory) moved the freezing of kernel
> >> threads to hibernation_snapshot() function.
> >>
> >> So now, if the call to hibernation_snapshot() fails or returns early due
> >> to a successful hibernation test, the caller has to thaw processes to
> >> ensure that the system gets back to its original state.
> >>
> >> But in SNAPSHOT_CREATE_IMAGE hibernation ioctl, the caller does not thaw
> >> processes. Fix this issue. But note that if hibernation_snapshot() returned
> >> due to a successful freezer test, we still send the value of 'in_suspend'
> >> (which is now 0) to userspace, because we are not in an error path per-se,
> >> and moreover, the value of in_suspend correctly depicts the situation here.
> >>
> >> Signed-off-by: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com>
> > 
> > Is this still applicable on top of pm-freezer branch?
> > 
> 
> 
> Well, yes, but not in this form. Thanks for bringing that up!
> 
> If the call to hibernation_snapshot() returns an error, we need not do
> anything since your freezer changes handle that case and call thaw_processes()
> automatically.
> But in case of return due to a successful freezer test, we would need to
> explicitly thaw processes since the freezer successfully froze them in the
> previous step.
> 
> So the following patch would do that, for the pm-freezer branch:
> 
> ---
> From: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com>
> Subject: [PATCH] PM / Hibernate: Thaw processes in SNAPSHOT_CREATE_IMAGE ioctl test path
> 
> Commit 2aede851ddf08666f68ffc17be446420e9d2a056 (PM / Hibernate: Freeze
> kernel threads after preallocating memory) moved the freezing of kernel
> threads to hibernation_snapshot() function.
> 
> So now, if the call to hibernation_snapshot() returns early due to a
> successful hibernation test, the caller has to thaw processes to ensure
> that the system gets back to its original state.
> 
> But in SNAPSHOT_CREATE_IMAGE hibernation ioctl, the caller does not thaw
> processes in case hibernation_snapshot() returned due to a successful
> freezer test. Fix this issue. But note we still send the value of 'in_suspend'
> (which is now 0) to userspace, because we are not in an error path per-se,
> and moreover, the value of in_suspend correctly depicts the situation here.
> 
> Signed-off-by: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com>

Applied to linux-pm/linux-next.

Thanks,
Rafael


> ---
> 
>  kernel/power/hibernate.c |    2 +-
>  kernel/power/power.h     |    2 ++
>  kernel/power/user.c      |   11 ++++++++---
>  3 files changed, 11 insertions(+), 4 deletions(-)
> 
> diff --git a/kernel/power/hibernate.c b/kernel/power/hibernate.c
> index 80c2d90..57bab63 100644
> --- a/kernel/power/hibernate.c
> +++ b/kernel/power/hibernate.c
> @@ -55,7 +55,7 @@ enum {
>  
>  static int hibernation_mode = HIBERNATION_SHUTDOWN;
>  
> -static bool freezer_test_done;
> +bool freezer_test_done;
>  
>  static const struct platform_hibernation_ops *hibernation_ops;
>  
> diff --git a/kernel/power/power.h b/kernel/power/power.h
> index 23a2db1..0c4defe 100644
> --- a/kernel/power/power.h
> +++ b/kernel/power/power.h
> @@ -50,6 +50,8 @@ static inline char *check_image_kernel(struct swsusp_info *info)
>  #define SPARE_PAGES	((1024 * 1024) >> PAGE_SHIFT)
>  
>  /* kernel/power/hibernate.c */
> +extern bool freezer_test_done;
> +
>  extern int hibernation_snapshot(int platform_mode);
>  extern int hibernation_restore(int platform_mode);
>  extern int hibernation_platform_enter(void);
> diff --git a/kernel/power/user.c b/kernel/power/user.c
> index 7cc3f5b..c202e2e 100644
> --- a/kernel/power/user.c
> +++ b/kernel/power/user.c
> @@ -281,10 +281,15 @@ static long snapshot_ioctl(struct file *filp, unsigned int cmd,
>  		}
>  		pm_restore_gfp_mask();
>  		error = hibernation_snapshot(data->platform_support);
> -		if (!error)
> +		if (!error) {
>  			error = put_user(in_suspend, (int __user *)arg);
> -		if (!error)
> -			data->ready = 1;
> +			if (!error && !freezer_test_done)
> +				data->ready = 1;
> +			if (freezer_test_done) {
> +				freezer_test_done = false;
> +				thaw_processes();
> +			}
> +		}
>  		break;
>  
>  	case SNAPSHOT_ATOMIC_RESTORE:
> 
> 
> --
> To unsubscribe from this list: send the line "unsubscribe linux-pm" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> 
> 

Re: [PATCH] PM / Hibernate: Thaw processes in SNAPSHOT_CREATE_IMAGE ioctl error/test paths

[Pavel Machek]

> On 12/01/2011 09:17 PM, Tejun Heo wrote:
> 
> > On Thu, Dec 01, 2011 at 06:45:37PM +0530, Srivatsa S. Bhat wrote:
> >> Commit 2aede851ddf08666f68ffc17be446420e9d2a056 (PM / Hibernate: Freeze
> >> kernel threads after preallocating memory) moved the freezing of kernel
> >> threads to hibernation_snapshot() function.
> >>
> >> So now, if the call to hibernation_snapshot() fails or returns early due
> >> to a successful hibernation test, the caller has to thaw processes to
> >> ensure that the system gets back to its original state.
> >>
> >> But in SNAPSHOT_CREATE_IMAGE hibernation ioctl, the caller does not thaw
> >> processes. Fix this issue. But note that if hibernation_snapshot() returned
> >> due to a successful freezer test, we still send the value of 'in_suspend'
> >> (which is now 0) to userspace, because we are not in an error path per-se,
> >> and moreover, the value of in_suspend correctly depicts the situation here.
> >>
> >> Signed-off-by: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com>
> > 
> > Is this still applicable on top of pm-freezer branch?
> > 
> 
> 
> Well, yes, but not in this form. Thanks for bringing that up!
> 
> If the call to hibernation_snapshot() returns an error, we need not do
> anything since your freezer changes handle that case and call thaw_processes()
> automatically.
> But in case of return due to a successful freezer test, we would need to
> explicitly thaw processes since the freezer successfully froze them in the
> previous step.
> 
> So the following patch would do that, for the pm-freezer branch:
> 
> ---
> From: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com>
> Subject: [PATCH] PM / Hibernate: Thaw processes in SNAPSHOT_CREATE_IMAGE ioctl test path
> 
> Commit 2aede851ddf08666f68ffc17be446420e9d2a056 (PM / Hibernate: Freeze
> kernel threads after preallocating memory) moved the freezing of kernel
> threads to hibernation_snapshot() function.
> 
> So now, if the call to hibernation_snapshot() returns early due to a
> successful hibernation test, the caller has to thaw processes to ensure
> that the system gets back to its original state.
> 
> But in SNAPSHOT_CREATE_IMAGE hibernation ioctl, the caller does not thaw
> processes in case hibernation_snapshot() returned due to a successful
> freezer test. Fix this issue. But note we still send the value of 'in_suspend'
> (which is now 0) to userspace, because we are not in an error path per-se,
> and moreover, the value of in_suspend correctly depicts the situation here.
> 
> Signed-off-by: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com>

Ack.

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