[PATCH] ext2: Avoid WARN() messages when failing to write to the superblock

[PATCH] ext2: Avoid WARN() messages when failing to write to the superblock

[Theodore Ts'o]

This fixes a common warning reported by kerneloops.org

[Kernel summit hacking hour]
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
---
 fs/ext2/super.c |   22 ++++++++++++++++++++++
 1 files changed, 22 insertions(+), 0 deletions(-)

diff --git a/fs/ext2/super.c b/fs/ext2/super.c
index 1a9ffee..7b3b6cd 100644
--- a/fs/ext2/super.c
+++ b/fs/ext2/super.c
@@ -1089,9 +1089,30 @@ failed_sbi:
 	return ret;
 }
 
+static void ext2_clear_super_error(struct super_block * sb)
+{
+	struct buffer_head *sbh = EXT2_SB(sb)->s_sbh;
+
+	if (buffer_write_io_error(sbh)) {
+		/*
+		 * Oh, dear.  A previous attempt to write the
+		 * superblock failed.  This could happen because the
+		 * USB device was yanked out.  Or it could happen to
+		 * be a transient write error and maybe the block will
+		 * be remapped.  Nothing we can do but to retry the
+		 * write and hope for the best.
+		 */
+		printk(KERN_ERR "EXT2-fs: %s previous I/O error to "
+		       "superblock detected", sb->s_id);
+		clear_buffer_write_io_error(sbh);
+		set_buffer_uptodate(sbh);
+	}
+}
+
 static void ext2_commit_super (struct super_block * sb,
 			       struct ext2_super_block * es)
 {
+	ext2_clear_super_error(sb);
 	es->s_wtime = cpu_to_le32(get_seconds());
 	mark_buffer_dirty(EXT2_SB(sb)->s_sbh);
 	sb->s_dirt = 0;
@@ -1099,6 +1120,7 @@ static void ext2_commit_super (struct super_block * sb,
 
 static void ext2_sync_super(struct super_block *sb, struct ext2_super_block *es)
 {
+	ext2_clear_super_error(sb);
 	es->s_free_blocks_count = cpu_to_le32(ext2_count_free_blocks(sb));
 	es->s_free_inodes_count = cpu_to_le32(ext2_count_free_inodes(sb));
 	es->s_wtime = cpu_to_le32(get_seconds());
-- 
1.6.5.104.g2567b.dirty

Re: [PATCH] ext2: Avoid WARN() messages when failing to write to the superblock

[Pavel Machek]

On Tue 2009-10-20 01:22:26, Theodore Ts'o wrote:
> This fixes a common warning reported by kerneloops.org
> 

> +static void ext2_clear_super_error(struct super_block * sb)
> +{
> +	struct buffer_head *sbh = EXT2_SB(sb)->s_sbh;
> +
> +	if (buffer_write_io_error(sbh)) {
> +		/*
> +		 * Oh, dear.  A previous attempt to write the
> +		 * superblock failed.  This could happen because the
> +		 * USB device was yanked out.  Or it could happen to
> +		 * be a transient write error and maybe the block will
> +		 * be remapped.  Nothing we can do but to retry the
> +		 * write and hope for the best.
> +		 */
> +		printk(KERN_ERR "EXT2-fs: %s previous I/O error to "
> +		       "superblock detected", sb->s_id);

KERN_ALERT/KERN_CRITICAL?
									Pavel
-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html

Re: [PATCH] ext2: Avoid WARN() messages when failing to write to the superblock

[Theodore Tso]

On Sun, Oct 25, 2009 at 07:21:28AM +0100, Pavel Machek wrote:
> > +	if (buffer_write_io_error(sbh)) {
> > +		/*
> > +		 * Oh, dear.  A previous attempt to write the
> > +		 * superblock failed.  This could happen because the
> > +		 * USB device was yanked out.  Or it could happen to
> > +		 * be a transient write error and maybe the block will
> > +		 * be remapped.  Nothing we can do but to retry the
> > +		 * write and hope for the best.
> > +		 */
> > +		printk(KERN_ERR "EXT2-fs: %s previous I/O error to "
> > +		       "superblock detected", sb->s_id);
> 
> KERN_ALERT/KERN_CRITICAL?

Based on kerneloops.org logs, it's apparently very frequent that
clueless lusers rip out USB drives without bothering to unmount them
first.  (Or people by crappy laptops where the SD cards sticks out
slightly and is easily jostled.)  It's not clear we really want to be
issueing KERN_ALERT or KERN_CRITICAL messages when this very common
event happens.

						- TRed

Re: [PATCH] ext2: Avoid WARN() messages when failing to write to the superblock

[Pavel Machek]

On Mon 2009-10-26 00:08:10, Theodore Tso wrote:
> On Sun, Oct 25, 2009 at 07:21:28AM +0100, Pavel Machek wrote:
> > > +	if (buffer_write_io_error(sbh)) {
> > > +		/*
> > > +		 * Oh, dear.  A previous attempt to write the
> > > +		 * superblock failed.  This could happen because the
> > > +		 * USB device was yanked out.  Or it could happen to
> > > +		 * be a transient write error and maybe the block will
> > > +		 * be remapped.  Nothing we can do but to retry the
> > > +		 * write and hope for the best.
> > > +		 */
> > > +		printk(KERN_ERR "EXT2-fs: %s previous I/O error to "
> > > +		       "superblock detected", sb->s_id);
> > 
> > KERN_ALERT/KERN_CRITICAL?
> 
> Based on kerneloops.org logs, it's apparently very frequent that
> clueless lusers rip out USB drives without bothering to unmount them
> first.  (Or people by crappy laptops where the SD cards sticks out
> slightly and is easily jostled.)  It's not clear we really want to be
> issueing KERN_ALERT or KERN_CRITICAL messages when this very common
> event happens.

I'd say that frequency of the message should not determine the
priority. (If this happens on your root filesystem you are in deep
trouble, and it _should_ be broadcast on all consoles).

OTOH filesystem should be handle device unplug just fine, so not
issuing this in that case would be nice...
								Pavel
-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html