PM/Freezer: Thaw only kernel threads if freezing of kernel threads fails

If freezing of kernel threads fails, we are expected to automatically thaw tasks in the error recovery path. However, at times, we encounter situations in which we would like the automatic error recovery path to thaw only the kernel threads, because we want to be able to do some more cleanup before we thaw userspace. Something like:

error = freeze_kernel_threads();

if (error) {
	/* Do some cleanup */

	/* Only then thaw userspace tasks*/
	thaw_processes();
}

An example of such a situation is where we freeze/thaw filesystems during suspend/hibernation. There, if freezing of kernel threads fails, we would like to thaw the frozen filesystems before thawing the userspace tasks.

So, modify freeze_kernel_threads() to thaw only kernel threads in case of freezing failure.

Signed-off-by Srivatsa S. Bhat <[email protected]>

---

 kernel/power/power.h   |   26 ++++++++++++++++++++++++--
 kernel/power/process.c |    7 +++++--
 2 files changed, 29 insertions(+), 4 deletions(-)

kernel/power/power.h +14 -2

index 0c4defe..cf41c03 100644
@@ -231,8 +231,30 @@ extern int pm_test_level;
 #ifdef CONFIG_SUSPEND_FREEZER
 static inline int suspend_freeze_processes(void)
 {
-	int error = freeze_processes();
-	return error ? : freeze_kernel_threads();
+	int error;
+
+	error = freeze_processes();
+
+	/*
+	 * freeze_processes() automatically thaws
+	 * every task if freezing fails. So we need
+	 * not do anything extra upon error.
+	 */
+	if (error)
+		goto Finish;
+
+	error = freeze_kernel_threads();
+

Quoted 14 lines · Srivatsa S. Bhat

If freezing of kernel threads fails, we are expected to automatically thaw tasks in the error recovery path. However, at times, we encounter situations in which we would like the automatic error recovery path to thaw only the kernel threads, because we want to be able to do some more cleanup before we thaw userspace. Something like:

error = freeze_kernel_threads();

if (error) {
	/* Do some cleanup */

	/* Only then thaw userspace tasks*/
	thaw_processes();
}

Quoted 8 lines

An example of such a situation is where we freeze/thaw filesystems during suspend/hibernation. There, if freezing of kernel threads fails, we would like to thaw the frozen filesystems before thawing the userspace tasks.

So, modify freeze_kernel_threads() to thaw only kernel threads in case of freezing failure.

Signed-off-by Srivatsa S. Bhat <[email protected]>

Acked-by Tejun Heo <[email protected]>

with nits below,

Quoted 34 lines

---

 kernel/power/power.h   |   26 ++++++++++++++++++++++++--
 kernel/power/process.c |    7 +++++--
 2 files changed, 29 insertions(+), 4 deletions(-)

kernel/power/power.h +18 -2

index 0c4defe..cf41c03 100644
@@ -231,8 +231,30 @@ extern int pm_test_level;
 #ifdef CONFIG_SUSPEND_FREEZER
 static inline int suspend_freeze_processes(void)
 {
-	int error = freeze_processes();
-	return error ? : freeze_kernel_threads();
+	int error;
+
+	error = freeze_processes();
+
+	/*
+	 * freeze_processes() automatically thaws
+	 * every task if freezing fails. So we need
+	 * not do anything extra upon error.
+	 */
+	if (error)
+		goto Finish;
+
+	error = freeze_kernel_threads();
+
+	/*
+	 * freeze_kernel_threads() thaws only kernel
+	 * threads upon freezing failure. So we have
+	 * to thaw the userspace tasks ourselves.

kernel/power/power.h extern int pm_test_level; · line 257

Quoted 6 lines

kernel/power/power.h +6 -0

+	 */
+	if (error)
+		thaw_processes();
+
+ Finish:
+	return error;

Heh, the original ?: code was scary.

That said, IMHO, the above seems
a bit too verbose. Functions usually shouldn't have side effects on failure and the above code is usual in that sense and it would be nice to mention the reformatting (from cryptic to sane) in the commit log. Also, how many columns are you flowing the text to? It seems quite narrower than they usually should be.

Quoted 4 lines · Srivatsa S. Bhat

So, modify freeze_kernel_threads() to thaw only kernel threads in case of freezing failure.

Signed-off-by Srivatsa S. Bhat <[email protected]>

Acked-by Tejun Heo <[email protected]>

with nits below,

Heh, the original ?: code was scary. That said, IMHO, the above seems a bit too verbose. Functions usually shouldn't have side effects on failure and the above code is usual in that sense and it would be nice to mention the reformatting (from cryptic to sane) in the commit log.

Also, how many columns are you flowing the text to? It seems quite narrower than they usually should be.

Quoted 7 lines · Tejun Heo

Quoted 4 lines · Srivatsa S. Bhat

So, modify freeze_kernel_threads() to thaw only kernel threads in case of freezing failure.

Signed-off-by Srivatsa S. Bhat <[email protected]>

Acked-by Tejun Heo <[email protected]>

Quoted 8 lines · Tejun Heo

Quoted 4 lines · Srivatsa S. Bhat

So, modify freeze_kernel_threads() to thaw only kernel threads in case of freezing failure.

Signed-off-by Srivatsa S. Bhat <[email protected]>

Acked-by Tejun Heo <[email protected]>

Thanks a lot!

Quoted 7 lines

with nits below,

Heh, the original ?: code was scary. That said, IMHO, the above seems a bit too verbose. Functions usually shouldn't have side effects on failure and the above code is usual in that sense and it would be nice to mention the reformatting (from cryptic to sane) in the commit log.

Ok :-)

Quoted 2 lines

Also, how many columns are you flowing the text to? It seems quite narrower than they usually should be.

Oh right, I didn't give attention to that. I'll try to fit it in lesser number of lines. But, should I really cut short on the explanation itself? I remember Nigel had requested some comments on a different patch on roughly similar code path https://lkml.org/lkml/2012/1/28/111.
So I thought it would be a good idea to explicitly document the differences, to be clear...

Here is the updated patch:

From: Srivatsa S. Bhat <[email protected]>
Subject: [PATCH] PM/Freezer: Thaw only kernel threads if freezing of kernel threads fails

If freezing of kernel threads fails, we are expected to automatically thaw tasks in the error recovery path. However, at times, we encounter situations in which we would like the automatic error recovery path to thaw only the kernel threads, because we want to be able to do some more cleanup before we thaw userspace. Something like:

error = freeze_kernel_threads();

if (error) {
	/* Do some cleanup */

	/* Only then thaw userspace tasks*/
	thaw_processes();
}

An example of such a situation is where we freeze/thaw filesystems during suspend/hibernation. There, if freezing of kernel threads fails, we would like to thaw the frozen filesystems before thawing the userspace tasks.

So, modify freeze_kernel_threads() to thaw only kernel threads in case of freezing failure. And change suspend_freeze_processes() accordingly. (At the same time, let us also get rid of the rather cryptic usage of the conditional operator (:?) in that function.)

Signed-off-by Srivatsa S. Bhat <[email protected]>

Acked-by Tejun Heo <[email protected]>

---

 kernel/power/power.h   |   24 ++++++++++++++++++++++--
 kernel/power/process.c |    7 +++++--
 2 files changed, 27 insertions(+), 4 deletions(-)

kernel/power/power.h +22 -2

index 0c4defe..21724ee 100644
@@ -231,8 +231,28 @@ extern int pm_test_level;
 #ifdef CONFIG_SUSPEND_FREEZER
 static inline int suspend_freeze_processes(void)
 {
-	int error = freeze_processes();
-	return error ? : freeze_kernel_threads();
+	int error;
+
+	error = freeze_processes();
+
+	/*
+	 * freeze_processes() automatically thaws every task if freezing
+	 * fails. So we need not do anything extra upon error.
+	 */
+	if (error)
+		goto Finish;
+
+	error = freeze_kernel_threads();
+
+	/*
+	 * freeze_kernel_threads() thaws only kernel threads upon freezing
+	 * failure. So we have to thaw the userspace tasks ourselves.
+	 */
+	if (error)
+		thaw_processes();
+
+ Finish:
+	return error;
 }
 
 static inline void suspend_thaw_processes(void)

kernel/power/process.c +5 -2

index eeca003..7e42645 100644
@@ -143,7 +143,10 @@ int freeze_processes(void)
 /**
  * freeze_kernel_threads - Make freezable kernel threads go to the refrigerator.
  *
- * On success, returns 0.  On failure, -errno and system is fully thawed.
+ * On success, returns 0.  On failure, -errno and only the kernel threads are
+ * thawed, so as to give a chance to the caller to do additional cleanups
+ * (if any) before thawing the userspace tasks. So, it is the responsibility
+ * of the caller to thaw the userspace tasks, when the time is right.
  */
 int freeze_kernel_threads(void)
 {

@@ -159,7 +162,7 @@ int freeze_kernel_threads(void)
 	BUG_ON(in_atomic());
 
 	if (error)
-		thaw_processes();
+		thaw_kernel_threads();
 	return error;
 }
 

From: Srivatsa S. Bhat <[email protected]>
Subject: [PATCH] PM/Freezer: Thaw only kernel threads if freezing of kernel threads fails

If freezing of kernel threads fails, we are expected to automatically thaw tasks in the error recovery path. However, at times, we encounter situations in which we would like the automatic error recovery path to thaw only the kernel threads, because we want to be able to do some more cleanup before we thaw userspace. Something like:

error = freeze_kernel_threads();

if (error) {
	/* Do some cleanup */

	/* Only then thaw userspace tasks*/
	thaw_processes();
}

An example of such a situation is where we freeze/thaw filesystems during suspend/hibernation. There, if freezing of kernel threads fails, we would like to thaw the frozen filesystems before thawing the userspace tasks.

So, modify freeze_kernel_threads() to thaw only kernel threads in case of freezing failure. And change suspend_freeze_processes() accordingly. (At the same time, let us also get rid of the rather cryptic usage of the conditional operator (:?) in that function.)

Signed-off-by Srivatsa S. Bhat <[email protected]>

Acked-by Tejun Heo <[email protected]>

---

Quoted 22 lines · Srivatsa S. Bhat

From: Srivatsa S. Bhat <[email protected]>
Subject: [PATCH] PM/Freezer: Thaw only kernel threads if freezing of kernel threads fails

If freezing of kernel threads fails, we are expected to automatically thaw tasks in the error recovery path. However, at times, we encounter situations in which we would like the automatic error recovery path to thaw only the kernel threads, because we want to be able to do some more cleanup before we thaw userspace. Something like:

error = freeze_kernel_threads();

if (error) {
	/* Do some cleanup */

	/* Only then thaw userspace tasks*/
	thaw_processes();
}

An example of such a situation is where we freeze/thaw filesystems during suspend/hibernation. There, if freezing of kernel threads fails, we would like to thaw the frozen filesystems before thawing the userspace tasks.

Quoted 8 lines

So, modify freeze_kernel_threads() to thaw only kernel threads in case of freezing failure. And change suspend_freeze_processes() accordingly. (At the same time, let us also get rid of the rather cryptic usage of the conditional operator (:?) in that function.)

Signed-off-by Srivatsa S. Bhat <[email protected]>

Acked-by Tejun Heo <[email protected]>

---

Hi Rafael,

I have reworked the changelog of this patch to indicate that this fixes a regression. Here is the updated patch:

Quoted 5 lines · Srivatsa S. Bhat

On 02/02/2012 04:29 AM, Srivatsa S. Bhat wrote:
Hi Rafael,

I have reworked the changelog of this patch to indicate that this fixes a regression. Here is the updated patch:

Well, I've already fixed that up in my tree, but thanks anyway. :-)

Quoted 4 lines · Rafael J. Wysocki

Quoted 3 lines · Srivatsa S. Bhat

On 02/02/2012 04:29 AM, Srivatsa S. Bhat wrote:
Hi Rafael,

Quoted 5 lines

Quoted 2 lines

I have reworked the changelog of this patch to indicate that this fixes a regression. Here is the updated patch:

Well, I've already fixed that up in my tree, but thanks anyway. :-)

Nicer if gurus open wider windows for patches received, no?

Quoted 4 lines · Rafael J. Wysocki

Quoted 3 lines · Srivatsa S. Bhat

On 02/02/2012 04:29 AM, Srivatsa S. Bhat wrote:
Hi Rafael,

Quoted 5 lines

Quoted 2 lines

I have reworked the changelog of this patch to indicate that this fixes a regression. Here is the updated patch:

Well, I've already fixed that up in my tree, but thanks anyway. :-)

Nicer if gurus open wider windows for patches received, no?

Quoted 5 lines · Hillf Danton

Quoted 4 lines · Rafael J. Wysocki

Quoted 3 lines · Srivatsa S. Bhat

On 02/02/2012 04:29 AM, Srivatsa S. Bhat wrote:
Hi Rafael,

Quoted 8 lines

Quoted 5 lines

Quoted 2 lines

I have reworked the changelog of this patch to indicate that this fixes a regression. Here is the updated patch:

Well, I've already fixed that up in my tree, but thanks anyway. :-)

Nicer if gurus open wider windows for patches received, no?

I am not quite sure what you meant by that, but just in case you got the wrong idea by any chance, let me clarify:

Working with Rafael has *always* been an immensely pleasurable experience! There is absolutely no doubt in that :-)

Quoted 5 lines · Hillf Danton

Quoted 4 lines · Rafael J. Wysocki

Quoted 3 lines · Srivatsa S. Bhat

On 02/02/2012 04:29 AM, Srivatsa S. Bhat wrote:
Hi Rafael,

Quoted 8 lines

Quoted 5 lines

Quoted 2 lines

I have reworked the changelog of this patch to indicate that this fixes a regression. Here is the updated patch:

Well, I've already fixed that up in my tree, but thanks anyway. :-)

Nicer if gurus open wider windows for patches received, no?

I am not quite sure what you meant by that, but just in case you got the wrong idea by any chance, let me clarify:

Working with Rafael has *always* been an immensely pleasurable experience! There is absolutely no doubt in that :-)

So there really is no reason to feel bad about the above incident :-)