Re: [RFC][PATCH] PM / Sleep: Freeze filesystems during system suspend/hibernation

["Rafael J. Wysocki" <rjw@sisk.pl>]

On Friday, February 17, 2012, Josh Boyer wrote:
> On Fri, Feb 17, 2012 at 10:41 AM, Josh Boyer <jwboyer@gmail.com> wrote:
> > On Sat, Jan 28, 2012 at 8:45 AM, Rafael J. Wysocki <rjw@sisk.pl> wrote:
> >> From: Rafael J. Wysocki <rjw@sisk.pl>
> >>
> >> Freeze all filesystems during system suspend and (kernel-driven)
> >> hibernation by calling freeze_supers() for all superblocks and thaw
> >> them during the subsequent resume with the help of thaw_supers().
> >>
> >> This makes filesystems stay in a consistent state in case something
> >> goes wrong between system suspend (or hibernation) and the subsequent
> >> resume (e.g. journal replays won't be necessary in those cases).  In
> >> particular, this should help to solve a long-standing issue that, in
> >> some cases, during resume from hibernation the boot loader causes the
> >> journal to be replied for the filesystem containing the kernel image
> >> and/or initrd causing it to become inconsistent with the information
> >> stored in the hibernation image.
> >>
> >> The user-space-driven hibernation (s2disk) is not covered by this
> >> change, because the freezing of filesystems prevents s2disk from
> >> accessing device special files it needs to do its job.
> >>
> >> This change is based on earlier work by Nigel Cunningham.
> >>
> >> Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
> >
> > Do you happen to have a version of this patch that has been rebased to the
> > latest 3.3 git tree?  I looked in the various archives and haven't found one
> > yet.  I'd like to get this into Fedora rawhide to see if some of the hibernate
> > issues we've been seeing clear up.
> 
> I got impatient and decided to tackle this myself.  If others would like to
> look at the forward port below, I'd appreciate it.

It looks good at first sight, thanks for rebasing!

Rafael


> commit b94887bbc0621e1e8402e7f0ec4bc3adf46c9a6e
> Author: Rafael J. Wysocki <rjw@sisk.pl>
> Date:   Fri Feb 17 12:42:08 2012 -0500
> 
>     Freeze all filesystems during system suspend and (kernel-driven)
>     hibernation by calling freeze_supers() for all superblocks and thaw
>     them during the subsequent resume with the help of thaw_supers().
> 
>     This makes filesystems stay in a consistent state in case something
>     goes wrong between system suspend (or hibernation) and the subsequent
>     resume (e.g. journal replays won't be necessary in those cases).  In
>     particular, this should help to solve a long-standing issue that, in
>     some cases, during resume from hibernation the boot loader causes the
>     journal to be replied for the filesystem containing the kernel image
>     and/or initrd causing it to become inconsistent with the information
>     stored in the hibernation image.
> 
>     The user-space-driven hibernation (s2disk) is not covered by this
>     change, because the freezing of filesystems prevents s2disk from
>     accessing device special files it needs to do its job.
> 
>     This change is based on earlier work by Nigel Cunningham.
> 
>     Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
> 
>     Rebased to 3.3-rc3 by Josh Boyer <jwboyer@redhat.com>
> 
> diff --git a/fs/super.c b/fs/super.c
> index 6015c02..c8057fa 100644
> --- a/fs/super.c
> +++ b/fs/super.c
> @@ -594,6 +594,79 @@ void iterate_supers_type(struct file_system_type *type,
>  EXPORT_SYMBOL(iterate_supers_type);
> 
>  /**
> + *	thaw_supers - call thaw_super() for all superblocks
> + */
> +void thaw_supers(void)
> +{
> +	struct super_block *sb, *p = NULL;
> +
> +	spin_lock(&sb_lock);
> +	list_for_each_entry(sb, &super_blocks, s_list) {
> +		if (hlist_unhashed(&sb->s_instances))
> +			continue;
> +		sb->s_count++;
> +		spin_unlock(&sb_lock);
> +
> +		if (sb->s_flags & MS_FROZEN) {
> +			thaw_super(sb);
> +			sb->s_flags &= ~MS_FROZEN;
> +		}
> +
> +		spin_lock(&sb_lock);
> +		if (p)
> +			__put_super(p);
> +		p = sb;
> +	}
> +	if (p)
> +		__put_super(p);
> +	spin_unlock(&sb_lock);
> +}
> +
> +/**
> + *	freeze_supers - call freeze_super() for all superblocks
> + */
> +int freeze_supers(void)
> +{
> +	struct super_block *sb, *p = NULL;
> +	int error = 0;
> +
> +	spin_lock(&sb_lock);
> +	/*
> +	 * Freeze in reverse order so filesystems depending on others are
> +	 * frozen in the right order (eg. loopback on ext3).
> +	 */
> +	list_for_each_entry_reverse(sb, &super_blocks, s_list) {
> +		if (hlist_unhashed(&sb->s_instances))
> +			continue;
> +		sb->s_count++;
> +		spin_unlock(&sb_lock);
> +
> +		if (sb->s_root && sb->s_frozen != SB_FREEZE_TRANS
> +		    && !(sb->s_flags & MS_RDONLY)) {
> +			error = freeze_super(sb);
> +			if (!error)
> +				sb->s_flags |= MS_FROZEN;
> +		}
> +
> +		spin_lock(&sb_lock);
> +		if (error)
> +			break;
> +		if (p)
> +			__put_super(p);
> +		p = sb;
> +	}
> +	if (p)
> +		__put_super(p);
> +	spin_unlock(&sb_lock);
> +
> +	if (error)
> +		thaw_supers();
> +
> +	return error;
> +}
> +
> +
> +/**
>   *	get_super - get the superblock of a device
>   *	@bdev: device to get the superblock for
>   *	
> diff --git a/include/linux/fs.h b/include/linux/fs.h
> index 386da09..a164f4a 100644
> --- a/include/linux/fs.h
> +++ b/include/linux/fs.h
> @@ -210,6 +210,7 @@ struct inodes_stat_t {
>  #define MS_KERNMOUNT	(1<<22) /* this is a kern_mount call */
>  #define MS_I_VERSION	(1<<23) /* Update inode I_version field */
>  #define MS_STRICTATIME	(1<<24) /* Always perform atime updates */
> +#define MS_FROZEN	(1<<25) /* Frozen filesystem */
>  #define MS_NOSEC	(1<<28)
>  #define MS_BORN		(1<<29)
>  #define MS_ACTIVE	(1<<30)
> @@ -2501,6 +2502,8 @@ extern void drop_super(struct super_block *sb);
>  extern void iterate_supers(void (*)(struct super_block *, void *), void *);
>  extern void iterate_supers_type(struct file_system_type *,
>  			        void (*)(struct super_block *, void *), void *);
> +extern int freeze_supers(void);
> +extern void thaw_supers(void);
> 
>  extern int dcache_dir_open(struct inode *, struct file *);
>  extern int dcache_dir_close(struct inode *, struct file *);
> diff --git a/kernel/power/hibernate.c b/kernel/power/hibernate.c
> index 6d6d288..492fc62 100644
> --- a/kernel/power/hibernate.c
> +++ b/kernel/power/hibernate.c
> @@ -626,12 +626,17 @@ int hibernate(void)
>  	if (error)
>  		goto Finish;
> 
> -	error = hibernation_snapshot(hibernation_mode == HIBERNATION_PLATFORM);
> +	error = freeze_supers();
>  	if (error)
>  		goto Thaw;
> +
> +	error = hibernation_snapshot(hibernation_mode == HIBERNATION_PLATFORM);
> +	if (error)
> +		goto Thaw_fs;
> +
>  	if (freezer_test_done) {
>  		freezer_test_done = false;
> -		goto Thaw;
> +		goto Thaw_fs;
>  	}
> 
>  	if (in_suspend) {
> @@ -655,6 +660,8 @@ int hibernate(void)
>  		pr_debug("PM: Image restored successfully.\n");
>  	}
> 
> + Thaw_fs:
> +	thaw_supers();
>   Thaw:
>  	thaw_processes();
>   Finish:
> diff --git a/kernel/power/power.h b/kernel/power/power.h
> index 21724ee..40d6f64 100644
> --- a/kernel/power/power.h
> +++ b/kernel/power/power.h
> @@ -1,3 +1,4 @@
> +#include <linux/fs.h>
>  #include <linux/suspend.h>
>  #include <linux/suspend_ioctls.h>
>  #include <linux/utsname.h>
> @@ -227,45 +228,3 @@ enum {
>  #define TEST_MAX	(__TEST_AFTER_LAST - 1)
> 
>  extern int pm_test_level;
> -
> -#ifdef CONFIG_SUSPEND_FREEZER
> -static inline int suspend_freeze_processes(void)
> -{
> -	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)
> -{
> -	thaw_processes();
> -}
> -#else
> -static inline int suspend_freeze_processes(void)
> -{
> -	return 0;
> -}
> -
> -static inline void suspend_thaw_processes(void)
> -{
> -}
> -#endif
> diff --git a/kernel/power/suspend.c b/kernel/power/suspend.c
> index 4fd51be..5f51fc7 100644
> --- a/kernel/power/suspend.c
> +++ b/kernel/power/suspend.c
> @@ -29,6 +29,62 @@
> 
>  #include "power.h"
> 
> +#ifdef CONFIG_SUSPEND_FREEZER
> +
> +static inline int suspend_freeze_processes(void)
> +{
> +	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_supers();
> +	if (error) {
> +		thaw_processes();
> +		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_supers();
> +		thaw_processes();
> +	}
> +
> +Finish:
> +	return error;
> +}
> +
> +static inline void suspend_thaw_processes(void)
> +{
> +	thaw_supers();
> +	thaw_processes();
> +}
> +
> +#else /* !CONFIG_SUSPEND_FREEZER */
> +
> +static inline int suspend_freeze_processes(void)
> +{
> +	return 0;
> +}
> +
> +static inline void suspend_thaw_processes(void)
> +{
> +}
> +
> +#endif /* !CONFIG_SUSPEND_FREEZER */
> +
>  const char *const pm_states[PM_SUSPEND_MAX] = {
>  	[PM_SUSPEND_STANDBY]	= "standby",
>  	[PM_SUSPEND_MEM]	= "mem",
> 
>