fs: minor fs thaw fixes and adjustments

PATCH In review Luis R. Rodriguez 4 participants 11 messages Apr 20, 2018 – May 3, 2018

Subsystems (2)

VFS Block layer

lore

Share & tools

lore

Luis R. Rodriguez

author

2018-04-20 23:59 UTC #0 lore raw

cover fs: minor fs thaw fixes and adjustments

Here's a few minor fs thaw fixes and adjustments I ran accross as I started to refresh my development for to use freeze_fs on suspend/hibernation [0]. These are just prep commits for the real work, I'm still reviewing feedback and adjusting the code for that.

I've tested the patches with generic/085 on xfs and found no regressions.

Luis R. Rodriguez (3):
fs: move documentation for thaw_super() where appropriate
fs: make thaw_super_locked() really just a helper
fs: fix corner case race on freeze_bdev() when sb disappears

 fs/block_dev.c |  4 +++-
 fs/super.c     | 39 ++++++++++++++++++++++++++-------------
 2 files changed, 29 insertions(+), 14 deletions(-)

Signature · 1 line

2.16.3

Luis R. Rodriguez

author

2018-04-20 23:59 UTC #1 lore raw

PATCH 1/3 fs: move documentation for thaw_super() where appropriate

On commit 08fdc8a0138a ("buffer.c: call thaw_super during emergency thaw") Mateusz added thaw_super_locked() and made thaw_super() use it, but forgot to move the documentation.

Signed-off-by Luis R. Rodriguez <[email protected]>

---
 fs/super.c | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)

fs/super.c +6 -6

index 5fa9a8d8d865..9d0eb5e20a1f 100644
@@ -1532,12 +1532,6 @@ int freeze_super(struct super_block *sb)
 }
 EXPORT_SYMBOL(freeze_super);
 
-/**
- * thaw_super -- unlock filesystem
- * @sb: the super to thaw
- *
- * Unlocks the filesystem and marks it writeable again after freeze_super().
- */
 static int thaw_super_locked(struct super_block *sb)
 {
 	int error;

@@ -1573,6 +1567,12 @@ static int thaw_super_locked(struct super_block *sb)
 	return 0;
 }
 
+/**
+ * thaw_super -- unlock filesystem
+ * @sb: the super to thaw
+ *
+ * Unlocks the filesystem and marks it writeable again after freeze_super().
+ */
 int thaw_super(struct super_block *sb)
 {
 	down_write(&sb->s_umount);

Signature · 1 line

2.16.3

Bart Van Assche

2018-04-21 00:01 UTC #2 lore raw

Re: [PATCH 1/3] fs: move documentation for thaw_super() where appropriate

Quoted 6 lines · Luis R. Rodriguez

+/**
+ * thaw_super -- unlock filesystem
+ * @sb: the super to thaw
+ *
+ * Unlocks the filesystem and marks it writeable again after freeze_super().
+ */

Have you verified the output generated by scripts/kernel-doc? Last time I checked the output for headers containing a double dash looked weird.

Bart.

Luis R. Rodriguez

author

2018-04-21 00:07 UTC #3 lore raw

Re: [PATCH 1/3] fs: move documentation for thaw_super() where appropriate

Quoted 11 lines · Bart Van Assche

Quoted 6 lines · Luis R. Rodriguez
+/**
+ * thaw_super -- unlock filesystem
+ * @sb: the super to thaw
+ *
+ * Unlocks the filesystem and marks it writeable again after freeze_super().
+ */
Have you verified the output generated by scripts/kernel-doc? Last time I checked the output for headers containing a double dash looked weird.

No, I just moved the block of kdoc crap. Randy would have this memorized I'm sure though so I should just fix the kdoc if its bad while at it.

Luis

Randy Dunlap

2018-04-21 00:09 UTC #4 lore raw

Re: [PATCH 1/3] fs: move documentation for thaw_super() where appropriate

Quoted 9 lines · Luis R. Rodriguez

Quoted 8 lines · Bart Van Assche
Quoted 6 lines · Luis R. Rodriguez
+/**
+ * thaw_super -- unlock filesystem
+ * @sb: the super to thaw
+ *
+ * Unlocks the filesystem and marks it writeable again after freeze_super().
+ */

Quoted 9 lines

Quoted 3 lines
Have you verified the output generated by scripts/kernel-doc? Last time I checked the output for headers containing a double dash looked weird.
No, I just moved the block of kdoc crap. Randy would have this memorized I'm sure though so I should just fix the kdoc if its bad while at it.
Luis

"--" does sound odd. I've never seen it used on purpose. FWIW, I just go by what is in Documentation/doc-guide/kernel-doc.rst:

Function documentation
----------------------

The general format of a function and function-like macro kernel-doc comment is::

/**
* function_name() - Brief description of function.
* @arg1: Describe the first argument.
* @arg2: Describe the second argument.
* One can provide multiple line descriptions
* for arguments.

Signature · 1 line

~Randy

Jan Kara

reviewer Reviewed-by on v1 1/3

2018-05-03 14:53 UTC #5 lore raw

Re: [PATCH 1/3] fs: move documentation for thaw_super() where appropriate

In reply to Luis R. Rodriguez

Quoted 5 lines · Luis R. Rodriguez

On commit 08fdc8a0138a ("buffer.c: call thaw_super during emergency thaw") Mateusz added thaw_super_locked() and made thaw_super() use it, but forgot to move the documentation.
Signed-off-by Luis R. Rodriguez <[email protected]>

Looks good (modulo the -- which is probably worth fixing when touching the comment anyway). You can add:

Reviewed-by Jan Kara <[email protected]>

Honza

Quoted 37 lines

---
 fs/super.c | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)

fs/super.c +6 -6

index 5fa9a8d8d865..9d0eb5e20a1f 100644
@@ -1532,12 +1532,6 @@ int freeze_super(struct super_block *sb)
 }
 EXPORT_SYMBOL(freeze_super);
 
-/**
- * thaw_super -- unlock filesystem
- * @sb: the super to thaw
- *
- * Unlocks the filesystem and marks it writeable again after freeze_super().
- */
 static int thaw_super_locked(struct super_block *sb)
 {
 	int error;

@@ -1573,6 +1567,12 @@ static int thaw_super_locked(struct super_block *sb)
 	return 0;
 }
 
+/**
+ * thaw_super -- unlock filesystem
+ * @sb: the super to thaw
+ *
+ * Unlocks the filesystem and marks it writeable again after freeze_super().
+ */
 int thaw_super(struct super_block *sb)
 {
 	down_write(&sb->s_umount);

Signature · 1 line

2.16.3

Signature · 2 lines

Jan Kara <[email protected]>
SUSE Labs, CR

Luis R. Rodriguez

author

2018-04-20 23:59 UTC #6 lore raw

PATCH 2/3 fs: make thaw_super_locked() really just a helper

In reply to Luis R. Rodriguez

thaw_super_locked() was added via commit 08fdc8a0138a ("buffer.c: call thaw_super during emergency thaw") merged on v4.17 to help with the ability so that the caller can take charge of handling the s_umount lock, however, it has left all* of the failure handling including unlocking lock of s_umount inside thaw_super_locked().

This does not make thaw_super_locked() flexible. For instance we may later want to use it with the abilty to handle unfolding of the locks ourselves.

Change thaw_super_locked() to really just be a helper, and let the callers deal with all the error handling.

This commit introeuces no functional changes.

Signed-off-by Luis R. Rodriguez <[email protected]>

---
 fs/super.c | 27 ++++++++++++++++++++-------
 1 file changed, 20 insertions(+), 7 deletions(-)

fs/super.c +20 -7

index 9d0eb5e20a1f..82bc74a16f06 100644
@@ -937,10 +937,15 @@ void emergency_remount(void)
 
 static void do_thaw_all_callback(struct super_block *sb)
 {
+	int error;
+
 	down_write(&sb->s_umount);
 	if (sb->s_root && sb->s_flags & MS_BORN) {
 		emergency_thaw_bdev(sb);
-		thaw_super_locked(sb);
+		error = thaw_super_locked(sb);
+		if (error)
+			up_write(&sb->s_umount);
+		deactivate_locked_super(sb);
 	} else {
 		up_write(&sb->s_umount);
 	}

@@ -1532,14 +1537,13 @@ int freeze_super(struct super_block *sb)
 }
 EXPORT_SYMBOL(freeze_super);
 
+/* Caller takes the sb->s_umount rw_semaphore lock and handles active count */
 static int thaw_super_locked(struct super_block *sb)
 {
 	int error;
 
-	if (sb->s_writers.frozen != SB_FREEZE_COMPLETE) {
-		up_write(&sb->s_umount);
+	if (sb->s_writers.frozen != SB_FREEZE_COMPLETE)
 		return -EINVAL;
-	}
 
 	if (sb_rdonly(sb)) {
 		sb->s_writers.frozen = SB_UNFROZEN;

@@ -1554,7 +1558,6 @@ static int thaw_super_locked(struct super_block *sb)
 			printk(KERN_ERR
 				"VFS:Filesystem thaw failed\n");
 			lockdep_sb_freeze_release(sb);
-			up_write(&sb->s_umount);
 			return error;
 		}
 	}

@@ -1563,7 +1566,6 @@ static int thaw_super_locked(struct super_block *sb)
 	sb_freeze_unlock(sb);
 out:
 	wake_up(&sb->s_writers.wait_unfrozen);
-	deactivate_locked_super(sb);
 	return 0;
 }

@@ -1575,7 +1577,18 @@ static int thaw_super_locked(struct super_block *sb)
  */
 int thaw_super(struct super_block *sb)
 {
+	int error;
+
 	down_write(&sb->s_umount);
-	return thaw_super_locked(sb);
+	error = thaw_super_locked(sb);
+	if (error) {
+		up_write(&sb->s_umount);
+		goto out;
+	}
+
+	deactivate_locked_super(sb);
+
+out:
+	return error;
 }
 EXPORT_SYMBOL(thaw_super);

Signature · 1 line

2.16.3

Jan Kara

reviewer

2018-05-03 15:07 UTC #7 lore raw

Re: [PATCH 2/3] fs: make thaw_super_locked() really just a helper

Quoted 6 lines · Luis R. Rodriguez

thaw_super_locked() was added via commit 08fdc8a0138a ("buffer.c: call thaw_super during emergency thaw") merged on v4.17 to help with the ability so that the caller can take charge of handling the s_umount lock, however, it has left all* of the failure handling including unlocking lock of s_umount inside thaw_super_locked().

Quoted 6 lines

This does not make thaw_super_locked() flexible. For instance we may later want to use it with the abilty to handle unfolding of the locks ourselves.
Change thaw_super_locked() to really just be a helper, and let the callers deal with all the error handling.

And do you have use for the new thaw_super_locked()? Because the new semantics with having to deal with deactivate_locked_super() does not seem ideal either so as a standalone cleanup patch it does not look too useful.

Quoted 1 line

This commit introeuces no functional changes.

^^^ introduces

Honza

Quoted 81 lines

---
 fs/super.c | 27 ++++++++++++++++++++-------
 1 file changed, 20 insertions(+), 7 deletions(-)

fs/super.c +20 -7

index 9d0eb5e20a1f..82bc74a16f06 100644
@@ -937,10 +937,15 @@ void emergency_remount(void)
 
 static void do_thaw_all_callback(struct super_block *sb)
 {
+	int error;
+
 	down_write(&sb->s_umount);
 	if (sb->s_root && sb->s_flags & MS_BORN) {
 		emergency_thaw_bdev(sb);
-		thaw_super_locked(sb);
+		error = thaw_super_locked(sb);
+		if (error)
+			up_write(&sb->s_umount);
+		deactivate_locked_super(sb);
 	} else {
 		up_write(&sb->s_umount);
 	}

@@ -1532,14 +1537,13 @@ int freeze_super(struct super_block *sb)
 }
 EXPORT_SYMBOL(freeze_super);
 
+/* Caller takes the sb->s_umount rw_semaphore lock and handles active count */
 static int thaw_super_locked(struct super_block *sb)
 {
 	int error;
 
-	if (sb->s_writers.frozen != SB_FREEZE_COMPLETE) {
-		up_write(&sb->s_umount);
+	if (sb->s_writers.frozen != SB_FREEZE_COMPLETE)
 		return -EINVAL;
-	}
 
 	if (sb_rdonly(sb)) {
 		sb->s_writers.frozen = SB_UNFROZEN;

@@ -1554,7 +1558,6 @@ static int thaw_super_locked(struct super_block *sb)
 			printk(KERN_ERR
 				"VFS:Filesystem thaw failed\n");
 			lockdep_sb_freeze_release(sb);
-			up_write(&sb->s_umount);
 			return error;
 		}
 	}

@@ -1563,7 +1566,6 @@ static int thaw_super_locked(struct super_block *sb)
 	sb_freeze_unlock(sb);
 out:
 	wake_up(&sb->s_writers.wait_unfrozen);
-	deactivate_locked_super(sb);
 	return 0;
 }

@@ -1575,7 +1577,18 @@ static int thaw_super_locked(struct super_block *sb)
  */
 int thaw_super(struct super_block *sb)
 {
+	int error;
+
 	down_write(&sb->s_umount);
-	return thaw_super_locked(sb);
+	error = thaw_super_locked(sb);
+	if (error) {
+		up_write(&sb->s_umount);
+		goto out;
+	}
+
+	deactivate_locked_super(sb);
+
+out:
+	return error;
 }
 EXPORT_SYMBOL(thaw_super);

Signature · 1 line

2.16.3

Signature · 2 lines

Jan Kara <[email protected]>
SUSE Labs, CR

Luis R. Rodriguez

author

2018-04-20 23:59 UTC #8 lore raw

PATCH 3/3 fs: fix corner case race on freeze_bdev() when sb disappears

In reply to Luis R. Rodriguez

freeze_bdev() will bail but leave the bd_fsfreeze_count incremented if the get_active_super() does not find the superblock on our super_blocks list to match.

This issue has been present since v2.6.29 during the introduction of the ioctl_fsfreeze() and ioctl_fsthaw() via commit fcccf502540e3 ("filesystem
freeze: implement generic freeze feature").

I am not aware of any existing races which have triggered this situation, however, if it does trigger it could mean leaving a superblock with bd_fsfreeze_count always positive.

Fixes fcccf502540e3 ("filesystem freeze: implement generic freeze feature")

Signed-off-by Luis R. Rodriguez <[email protected]>

---
 fs/block_dev.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

fs/block_dev.c +3 -1

index b54966679833..7a532aa58c07 100644
@@ -507,8 +507,10 @@ struct super_block *freeze_bdev(struct block_device *bdev)
 	}
 
 	sb = get_active_super(bdev);
-	if (!sb)
+	if (!sb) {
+		bdev->bd_fsfreeze_count--;
 		goto out;
+	}
 	if (sb->s_op->freeze_super)
 		error = sb->s_op->freeze_super(sb);
 	else

Signature · 1 line

2.16.3

Jan Kara

reviewer Reviewed-by on v1 3/3

2018-05-03 15:02 UTC #9 lore raw

Re: [PATCH 3/3] fs: fix corner case race on freeze_bdev() when sb disappears

Quoted 8 lines · Luis R. Rodriguez

freeze_bdev() will bail but leave the bd_fsfreeze_count incremented if the get_active_super() does not find the superblock on our super_blocks list to match.
This issue has been present since v2.6.29 during the introduction of the ioctl_fsfreeze() and ioctl_fsthaw() via commit fcccf502540e3 ("filesystem
freeze: implement generic freeze feature").

Quoted 6 lines

I am not aware of any existing races which have triggered this situation, however, if it does trigger it could mean leaving a superblock with bd_fsfreeze_count always positive.
Fixes fcccf502540e3 ("filesystem freeze: implement generic freeze feature")
Signed-off-by Luis R. Rodriguez <[email protected]>

Looks good to me. You can add:

Reviewed-by Jan Kara <[email protected]>

Honza

Quoted 23 lines

---
 fs/block_dev.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

fs/block_dev.c +3 -1

index b54966679833..7a532aa58c07 100644
@@ -507,8 +507,10 @@ struct super_block *freeze_bdev(struct block_device *bdev)
 	}
 
 	sb = get_active_super(bdev);
-	if (!sb)
+	if (!sb) {
+		bdev->bd_fsfreeze_count--;
 		goto out;
+	}
 	if (sb->s_op->freeze_super)
 		error = sb->s_op->freeze_super(sb);
 	else

Signature · 1 line

2.16.3

Signature · 2 lines

Jan Kara <[email protected]>
SUSE Labs, CR

Luis R. Rodriguez

author

2018-04-21 00:02 UTC #10 lore raw

In reply to Luis R. Rodriguez

Quoted 5 lines · Luis R. Rodriguez

Here's a few minor fs thaw fixes and adjustments I ran accross as I started to refresh my development for to use freeze_fs on suspend/hibernation [0]. These are just prep commits for the real work, I'm still reviewing feedback and adjusting the code for that.

And I forgot to provide the URL, for those that missed the old series I was working on:

[0] https://lkml.kernel.org/r/[email protected]

Luis

Quoted 16 lines

I've tested the patches with generic/085 on xfs and found no regressions.
Luis R. Rodriguez (3):
fs: move documentation for thaw_super() where appropriate
fs: make thaw_super_locked() really just a helper
fs: fix corner case race on freeze_bdev() when sb disappears
 fs/block_dev.c |  4 +++-
 fs/super.c     | 39 ++++++++++++++++++++++++++-------------
 2 files changed, 29 insertions(+), 14 deletions(-)
Signature · 1 line
2.16.3

Signature · 1 line

Do not panic