Hibernation and suspend patchset

Hibernation and suspend patchset

[Rafael J. Wysocki]

Hi,

I've put together a quilt patchset that contains some hibernation and suspend
patches from Johannes, Pekka, me and some other people that either have
already been merged with the mainline, but are not present in 2.6.21, or are
included in the -mm tree, or have been sent for a review to this list.

The patches are available at:

http://www.sisk.pl/kernel/hibernation_and_suspend/

where the '2.6.21' directory contains the patchset against the 2.6.21 kernel,
and will contain daily snapshots of the patchset.  Later on I'll add analogous
directories for subsequent -rc and 'vanilla' kernels.

The most recent patchset agains the latest 'vanilla' or -rc kernel (whichever
is more recent) will always be available at

http://www.sisk.pl/kernel/hibernation_and_suspend/current/patches/

or, as a tarball, at:

http://www.sisk.pl/kernel/hibernation_and_suspend/current_snapshot.tgz

The present patchset against 2.6.21, tested a bit on x86_64, contains some
'glue' patches that I have added to be able to apply the other patches cleany
on top of 2.6.21.  When 2.6.21-rc1 is out, I'll rediff the entire patchset
against it and then, hopefully, no such 'glue' patches will be necessary.

I'm going to track and include in the patchset all patches that affect:
- any files under kernel/power
- any files under Documentation/power
- include/linux/freezer.h
- include/linux/pm.h
- include/linux/suspend.h
- kernel/freezer.c
and some architecture-dependent files related to hibernation/suspend (patches
that don't affect these files may be included too, if they are related to
hibernation/suspend) and that have been posted for a review either to this list
or to the LKML and have not received negative comments.

If you know of a patch that in your opinion should be included but currently is
not, please let me know.

Greetings,
Rafael


-- 
If you don't have the time to read,
you don't have the time or the tools to write.
		- Stephen King

Re: Hibernation and suspend patchset

[Alan Stern]

On Tue, 1 May 2007, Rafael J. Wysocki wrote:

> I'm going to track and include in the patchset all patches that affect:
> - any files under kernel/power
> - any files under Documentation/power
> - include/linux/freezer.h
> - include/linux/pm.h
> - include/linux/suspend.h
> - kernel/freezer.c
> and some architecture-dependent files related to hibernation/suspend (patches
> that don't affect these files may be included too, if they are related to
> hibernation/suspend) and that have been posted for a review either to this list
> or to the LKML and have not received negative comments.
> 
> If you know of a patch that in your opinion should be included but currently is
> not, please let me know.

Shouldn't you also include all files under drivers/base/power?

Alan Stern

Re: Hibernation and suspend patchset

[Rafael J. Wysocki]

On Tuesday, 1 May 2007 16:04, Alan Stern wrote:
> On Tue, 1 May 2007, Rafael J. Wysocki wrote:
> 
> > I'm going to track and include in the patchset all patches that affect:
> > - any files under kernel/power
> > - any files under Documentation/power
> > - include/linux/freezer.h
> > - include/linux/pm.h
> > - include/linux/suspend.h
> > - kernel/freezer.c
> > and some architecture-dependent files related to hibernation/suspend (patches
> > that don't affect these files may be included too, if they are related to
> > hibernation/suspend) and that have been posted for a review either to this list
> > or to the LKML and have not received negative comments.
> > 
> > If you know of a patch that in your opinion should be included but currently is
> > not, please let me know.
> 
> Shouldn't you also include all files under drivers/base/power?

Yes, absolutely.

I forgot to put them on the list. :-)

Greetings,
Rafael

Re: Hibernation and suspend patchset

[Johannes Berg]

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

Hi again,

I just took your whole patchset, removed two patches that wouldn't apply
against Linus's current git tree, put mine for suspend to disk on G5 on
top of it, then applied the freezer fix I just sent and finally tested
the whole thing. It works :)

johannes

[-- Attachment #1.2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 190 bytes --]

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

Re: Hibernation and suspend patchset

[Rafael J. Wysocki]

Hi,

On Thursday, 3 May 2007 15:38, Johannes Berg wrote:
> Hi again,
> 
> I just took your whole patchset, removed two patches that wouldn't apply
> against Linus's current git tree, put mine for suspend to disk on G5 on
> top of it, then applied the freezer fix I just sent and finally tested
> the whole thing. It works :)

Nice to hear that. :-)

What's the current status of your patches for hibernation on G5?

Greetings,
Rafael

Re: Hibernation and suspend patchset

[Johannes Berg]

[-- Attachment #1.1: Type: text/plain, Size: 340 bytes --]

On Fri, 2007-05-04 at 13:36 +0200, Rafael J. Wysocki wrote:

> What's the current status of your patches for hibernation on G5?

I sent most of them to Paul for inclusion, two of them depend on the
stuff you did with pageflags so I still have four patches outstanding
that I'll send once that code trickles into powerpc.

johannes

[-- Attachment #1.2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 190 bytes --]

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

Re: Hibernation and suspend patchset

[Johannes Berg]

Hi Rafael,

Some bugs with freezer_exempt:

kernel/softirq.c: In function ‘ksoftirqd’:
kernel/softirq.c:515: warning: implicit declaration of function ‘freezer_exempt’
[...]

Maybe some header file they all use pulls in freezer.h on x86 but not on
powerpc? This fixes it.

--- linux-2.6-git.orig/drivers/block/loop.c	2007-05-03 15:30:23.274123904 +0200
+++ linux-2.6-git/drivers/block/loop.c	2007-05-03 15:30:43.015123904 +0200
@@ -74,6 +74,7 @@
 #include <linux/highmem.h>
 #include <linux/gfp.h>
 #include <linux/kthread.h>
+#include <linux/freezer.h>
 
 #include <asm/uaccess.h>
 
--- linux-2.6-git.orig/drivers/ieee1394/ieee1394_core.c	2007-05-03 15:30:23.423123904 +0200
+++ linux-2.6-git/drivers/ieee1394/ieee1394_core.c	2007-05-03 15:30:50.023123904 +0200
@@ -34,6 +34,7 @@
 #include <linux/kthread.h>
 #include <linux/preempt.h>
 #include <linux/time.h>
+#include <linux/freezer.h>
 
 #include <asm/system.h>
 #include <asm/byteorder.h>
--- linux-2.6-git.orig/drivers/scsi/scsi_error.c	2007-05-03 15:30:23.498123904 +0200
+++ linux-2.6-git/drivers/scsi/scsi_error.c	2007-05-03 15:30:54.423123904 +0200
@@ -24,6 +24,7 @@
 #include <linux/interrupt.h>
 #include <linux/blkdev.h>
 #include <linux/delay.h>
+#include <linux/freezer.h>
 
 #include <scsi/scsi.h>
 #include <scsi/scsi_cmnd.h>
--- linux-2.6-git.orig/kernel/softirq.c	2007-05-03 15:30:23.187123904 +0200
+++ linux-2.6-git/kernel/softirq.c	2007-05-03 15:30:37.470123904 +0200
@@ -18,6 +18,7 @@
 #include <linux/rcupdate.h>
 #include <linux/smp.h>
 #include <linux/tick.h>
+#include <linux/freezer.h>
 
 #include <asm/irq.h>
 /*
--- linux-2.6-git.orig/kernel/softlockup.c	2007-05-03 15:30:23.348123904 +0200
+++ linux-2.6-git/kernel/softlockup.c	2007-05-03 15:30:46.214123904 +0200
@@ -13,6 +13,7 @@
 #include <linux/kthread.h>
 #include <linux/notifier.h>
 #include <linux/module.h>
+#include <linux/freezer.h>
 
 static DEFINE_SPINLOCK(print_lock);
 
--- wireless-dev.orig/drivers/char/apm-emulation.c      2007-05-03 20:01:08.309154963 +0200
+++ wireless-dev/drivers/char/apm-emulation.c   2007-05-03 20:01:14.119154963 +0200
@@ -27,6 +27,7 @@
 #include <linux/completion.h>
 #include <linux/kthread.h>
 #include <linux/delay.h>
+#include <linux/freezer.h>
 
 #include <asm/system.h>
 

Re: Hibernation and suspend patchset

[Rafael J. Wysocki]

On Thursday, 3 May 2007 21:08, Johannes Berg wrote:
> Hi Rafael,
> 
> Some bugs with freezer_exempt:
> 
> kernel/softirq.c: In function ‘ksoftirqd’:
> kernel/softirq.c:515: warning: implicit declaration of function ‘freezer_exempt’
> [...]
> 
> Maybe some header file they all use pulls in freezer.h on x86 but not on
> powerpc? This fixes it.

Yes, that's possible.

Thanks for the patch, I'll add it to the patchset and send it for inclusion in -mm.

Greetings,
Rafael

Re: Hibernation and suspend patchset

[Rafael J. Wysocki]

Hi,

On Thursday, 3 May 2007 21:08, Johannes Berg wrote:
> Hi Rafael,
> 
> Some bugs with freezer_exempt:
> 
> kernel/softirq.c: In function ‘ksoftirqd’:
> kernel/softirq.c:515: warning: implicit declaration of function ‘freezer_exempt’
> [...]
> 
> Maybe some header file they all use pulls in freezer.h on x86 but not on
> powerpc? This fixes it.

These #includes are added by

008-add-try_to_freeze-everywhere.patch

Perhaps it's one of the patches that don't apply to the current -git?

Rafael

Re: Hibernation and suspend patchset

[Johannes Berg]

[-- Attachment #1.1: Type: text/plain, Size: 270 bytes --]

Hi,

> These #includes are added by
> 
> 008-add-try_to_freeze-everywhere.patch
> 
> Perhaps it's one of the patches that don't apply to the current -git?

Hmm, could be, but shouldn't the patch that adds the freezer_exempt
calls add them as well?

johannes

[-- Attachment #1.2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 190 bytes --]

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

Re: Hibernation and suspend patchset

[Rafael J. Wysocki]

On Friday, 4 May 2007 14:05, Johannes Berg wrote:
> Hi,
> 
> > These #includes are added by
> > 
> > 008-add-try_to_freeze-everywhere.patch
> > 
> > Perhaps it's one of the patches that don't apply to the current -git?
> 
> Hmm, could be, but shouldn't the patch that adds the freezer_exempt
> calls add them as well?

Well, I'm not sure what you mean.  It's on top of
008-add-try_to_freeze-everywhere.patch and the #includes are already there, so
it can't just add them once again.

IOW, the ordering of patches in the patchset is important and if you drop one,
you may need to replace it with another patch that does similar changes.

Greetings,
Rafael

Re: Hibernation and suspend patchset

[Johannes Berg]

[-- Attachment #1.1: Type: text/plain, Size: 379 bytes --]

On Fri, 2007-05-04 at 14:15 +0200, Rafael J. Wysocki wrote:

> Well, I'm not sure what you mean.  It's on top of
> 008-add-try_to_freeze-everywhere.patch and the #includes are already there, so
> it can't just add them once again.

Ah, gotcha. Ok, sorry about the noise. I'll check it out again once you
port your patchset to -rc1 or whatever you plan to do.

johannes

[-- Attachment #1.2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 190 bytes --]

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