Skip to contentSkip to conversation

ext3: Don't update superblock write time when filesystem is read-only

PATCH In review Theodore Ts'o 2 participants 4 messages Oct 4, 2009 – Oct 5, 2009

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 ext3: Don't update superblock write time when filesystem is read-only

This avoids updating the superblock write time when we are mounting the root file system read/only but we need to replay the journal; at that point, for people who are east of GMT and who make their clock tick in localtime for Windows bug-for-bug compatibility, and this will cause e2fsck to complain and force a full file system check.

---
 fs/ext3/super.c |   13 ++++++++++++-
 1 files changed, 12 insertions(+), 1 deletions(-)
fs/ext3/super.c +12 -1
index 72743d3..7a520a8 100644
@@ -2321,7 +2321,18 @@ static int ext3_commit_super(struct super_block *sb,
 
 	if (!sbh)
 		return error;
-	es->s_wtime = cpu_to_le32(get_seconds());
+	/*
+	 * If the file system is mounted read-only, don't update the
+	 * superblock write time.  This avoids updating the superblock
+	 * write time when we are mounting the root file system
+	 * read/only but we need to replay the journal; at that point,
+	 * for people who are east of GMT and who make their clock
+	 * tick in localtime for Windows bug-for-bug compatibility,
+	 * the clock is set in the future, and this will cause e2fsck
+	 * to complain and force a full file system check.
+	 */
+	if (!(sb->s_flags & MS_RDONLY))
+		es->s_wtime = cpu_to_le32(get_seconds());
 	es->s_free_blocks_count = cpu_to_le32(ext3_count_free_blocks(sb));
 	es->s_free_inodes_count = cpu_to_le32(ext3_count_free_inodes(sb));
 	BUFFER_TRACE(sbh, "marking dirty");
Signature · 1 line
1.6.4.4.2.g31f8e.dirty
#1 lore raw
Quoted 5 lines · Theodore Ts'o

This avoids updating the superblock write time when we are mounting the root file system read/only but we need to replay the journal; at that point, for people who are east of GMT and who make their clock tick in localtime for Windows bug-for-bug compatibility, and this will cause e2fsck to complain and force a full file system check.

Thanks, the patch is queued in my queue already. I just held it for a
while because Eric also had a valid point - maybe fsck is too strict when it forces a full filesystem check just because the superblock write time is in future? On a pc-class hw, clocks go often wrong. Even I remember setting system clock on my desktop a few hours back because it just gets out of sync and I don't run NTP...
So I'm not against your patch because it can help in some cases and
it's not a big deal but I think a proper fix would be in fsck...

Honza

Quoted 33 lines
---
 fs/ext3/super.c |   13 ++++++++++++-
 1 files changed, 12 insertions(+), 1 deletions(-)
fs/ext3/super.c +12 -1
index 72743d3..7a520a8 100644
@@ -2321,7 +2321,18 @@ static int ext3_commit_super(struct super_block *sb,
 
 	if (!sbh)
 		return error;
-	es->s_wtime = cpu_to_le32(get_seconds());
+	/*
+	 * If the file system is mounted read-only, don't update the
+	 * superblock write time.  This avoids updating the superblock
+	 * write time when we are mounting the root file system
+	 * read/only but we need to replay the journal; at that point,
+	 * for people who are east of GMT and who make their clock
+	 * tick in localtime for Windows bug-for-bug compatibility,
+	 * the clock is set in the future, and this will cause e2fsck
+	 * to complain and force a full file system check.
+	 */
+	if (!(sb->s_flags & MS_RDONLY))
+		es->s_wtime = cpu_to_le32(get_seconds());
 	es->s_free_blocks_count = cpu_to_le32(ext3_count_free_blocks(sb));
 	es->s_free_inodes_count = cpu_to_le32(ext3_count_free_inodes(sb));
 	BUFFER_TRACE(sbh, "marking dirty");
Signature · 1 line
1.6.4.4.2.g31f8e.dirty
Signature · 2 lines
Jan Kara <[email protected]>
SUSE Labs, CR
#2 lore raw
Quoted 6 lines · Jan Kara
Quoted 5 lines · Theodore Ts'o

This avoids updating the superblock write time when we are mounting the root file system read/only but we need to replay the journal; at that point, for people who are east of GMT and who make their clock tick in localtime for Windows bug-for-bug compatibility, and this will cause e2fsck to complain and force a full file system check.

Quoted 8 lines

Thanks, the patch is queued in my queue already. I just held it for a
while because Eric also had a valid point - maybe fsck is too strict when it forces a full filesystem check just because the superblock write time is in future? On a pc-class hw, clocks go often wrong. Even I remember setting system clock on my desktop a few hours back because it just gets out of sync and I don't run NTP...
So I'm not against your patch because it can help in some cases and
it's not a big deal but I think a proper fix would be in fsck...

Regardless of whether fsck relaxes its check, if the filesystem is mounted read-only, I don't think the last write time should change...

- Ted

#3 lore raw
Quoted 16 lines · Theodore Tso
Quoted 6 lines · Jan Kara
Quoted 5 lines · Theodore Ts'o

This avoids updating the superblock write time when we are mounting the root file system read/only but we need to replay the journal; at that point, for people who are east of GMT and who make their clock tick in localtime for Windows bug-for-bug compatibility, and this will cause e2fsck to complain and force a full file system check.

Quoted 8 lines

Thanks, the patch is queued in my queue already. I just held it for a
while because Eric also had a valid point - maybe fsck is too strict when it forces a full filesystem check just because the superblock write time is in future? On a pc-class hw, clocks go often wrong. Even I remember setting system clock on my desktop a few hours back because it just gets out of sync and I don't run NTP...
So I'm not against your patch because it can help in some cases and
it's not a big deal but I think a proper fix would be in fsck...

Quoted 2 lines

Regardless of whether fsck relaxes its check, if the filesystem is mounted read-only, I don't think the last write time should change...

Yeah, that's also a reasonable argument, although in fact the filesystem
has changed (we replayed the journal) so I can also see a reason to keep updating it.
Anyway, as I said above, I'm not against the patch. I'll push it to Linus
this week.

Honza

Signature · 2 lines
Jan Kara <[email protected]>
SUSE Labs, CR