[PATCH] JBD2: round commit timer up to avoid uncommitted transaction

[PATCH] JBD2: round commit timer up to avoid uncommitted transaction

[Andreas Dilger]

jbd and jbd2 fixes: fix jiffie rounding in jbd commit timer setup code.
Rounding down could cause the timer to be fired before the corresponding
transaction has expired.  That transaction can stay not committed forever if
no new transaction is created or expicit sync/umount happens.

Signed-off-by: Alex Zhuravlev (Tomas) <alex.zhuravlev@sun.com>
Signed-off-by: Andreas Dilger <adilger@sun.com>

Index: linux-2.6.27.21-0.1/fs/jbd2/transaction.c
===================================================================
--- linux-2.6.27.21-0.1.orig/fs/jbd2/transaction.c	2009-06-10 11:11:41.000000000 -0600
+++ linux-2.6.27.21-0.1/fs/jbd2/transaction.c	2009-06-10 11:12:32.000000000 -0600
@@ -54,7 +54,7 @@
 	INIT_LIST_HEAD(&transaction->t_inode_list);
 
 	/* Set up the commit timer for the new transaction. */
-	journal->j_commit_timer.expires = round_jiffies(transaction->t_expires);
+	journal->j_commit_timer.expires = round_jiffies_up(transaction->t_expires);
 	add_timer(&journal->j_commit_timer);
 
 	J_ASSERT(journal->j_running_transaction == NULL);

Cheers, Andreas
--
Andreas Dilger
Sr. Staff Engineer, Lustre Group
Sun Microsystems of Canada, Inc.

Re: [PATCH] JBD2: round commit timer up to avoid uncommitted transaction

[Jan Kara]

> jbd and jbd2 fixes: fix jiffie rounding in jbd commit timer setup code.
> Rounding down could cause the timer to be fired before the corresponding
> transaction has expired.  That transaction can stay not committed forever if
> no new transaction is created or expicit sync/umount happens.
> 
> Signed-off-by: Alex Zhuravlev (Tomas) <alex.zhuravlev@sun.com>
> Signed-off-by: Andreas Dilger <adilger@sun.com>
  Looks fine. You can add:
Acked-by: Jan Kara <jack@suse.cz>
  How about a similar JBD fix?

									Honza
> Index: linux-2.6.27.21-0.1/fs/jbd2/transaction.c
> ===================================================================
> --- linux-2.6.27.21-0.1.orig/fs/jbd2/transaction.c	2009-06-10 11:11:41.000000000 -0600
> +++ linux-2.6.27.21-0.1/fs/jbd2/transaction.c	2009-06-10 11:12:32.000000000 -0600
> @@ -54,7 +54,7 @@
>  	INIT_LIST_HEAD(&transaction->t_inode_list);
>  
>  	/* Set up the commit timer for the new transaction. */
> -	journal->j_commit_timer.expires = round_jiffies(transaction->t_expires);
> +	journal->j_commit_timer.expires = round_jiffies_up(transaction->t_expires);
>  	add_timer(&journal->j_commit_timer);
>  
>  	J_ASSERT(journal->j_running_transaction == NULL);
> 
> Cheers, Andreas
> --
> Andreas Dilger
> Sr. Staff Engineer, Lustre Group
> Sun Microsystems of Canada, Inc.
> 
> --
> To unsubscribe from this list: send the line "unsubscribe linux-ext4" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
-- 
Jan Kara <jack@suse.cz>
SuSE CR Labs

Re: [PATCH] JBD2: round commit timer up to avoid uncommitted transaction

[Andreas Dilger]

On Jul 30, 2009  19:47 +0200, Jan Kara wrote:
> > jbd and jbd2 fixes: fix jiffie rounding in jbd commit timer setup code.
> > Rounding down could cause the timer to be fired before the corresponding
> > transaction has expired.  That transaction can stay not committed forever if
> > no new transaction is created or expicit sync/umount happens.
> > 
> > Signed-off-by: Alex Zhuravlev (Tomas) <alex.zhuravlev@sun.com>
> > Signed-off-by: Andreas Dilger <adilger@sun.com>
>
>   Looks fine. You can add:
> Acked-by: Jan Kara <jack@suse.cz>
>   How about a similar JBD fix?

Index: linux-2.6.27.21-0.1/fs/jbd/transaction.c
===================================================================
--- linux-2.6.27.21-0.1.orig/fs/jbd/transaction.c	2009-06-10 11:11:41.000000000 -0600
+++ linux-2.6.27.21-0.1/fs/jbd/transaction.c	2009-06-10 11:12:32.000000000 -0600
@@ -54,7 +54,7 @@
 	spin_lock_init(&transaction->t_inode_list);
 
 	/* Set up the commit timer for the new transaction. */
-	journal->j_commit_timer.expires = round_jiffies(transaction->t_expires);
+	journal->j_commit_timer.expires = round_jiffies_up(transaction->t_expires);
 	add_timer(&journal->j_commit_timer);
 
 	J_ASSERT(journal->j_running_transaction == NULL);

Cheers, Andreas
--
Andreas Dilger
Sr. Staff Engineer, Lustre Group
Sun Microsystems of Canada, Inc.

Re: [PATCH] JBD2: round commit timer up to avoid uncommitted transaction

[Jan Kara]

> On Jul 30, 2009  19:47 +0200, Jan Kara wrote:
> > > jbd and jbd2 fixes: fix jiffie rounding in jbd commit timer setup code.
> > > Rounding down could cause the timer to be fired before the corresponding
> > > transaction has expired.  That transaction can stay not committed forever if
> > > no new transaction is created or expicit sync/umount happens.
> > > 
> > > Signed-off-by: Alex Zhuravlev (Tomas) <alex.zhuravlev@sun.com>
> > > Signed-off-by: Andreas Dilger <adilger@sun.com>
> >
> >   Looks fine. You can add:
> > Acked-by: Jan Kara <jack@suse.cz>
> >   How about a similar JBD fix?
> 
> Index: linux-2.6.27.21-0.1/fs/jbd/transaction.c
> ===================================================================
> --- linux-2.6.27.21-0.1.orig/fs/jbd/transaction.c	2009-06-10 11:11:41.000000000 -0600
> +++ linux-2.6.27.21-0.1/fs/jbd/transaction.c	2009-06-10 11:12:32.000000000 -0600
> @@ -54,7 +54,7 @@
>  	spin_lock_init(&transaction->t_inode_list);
>  
>  	/* Set up the commit timer for the new transaction. */
> -	journal->j_commit_timer.expires = round_jiffies(transaction->t_expires);
> +	journal->j_commit_timer.expires = round_jiffies_up(transaction->t_expires);
>  	add_timer(&journal->j_commit_timer);
>  
>  	J_ASSERT(journal->j_running_transaction == NULL);
  Thanks. Copied over the changelog, added your Signed-off-by and merged
into my tree...

									Honza
-- 
Jan Kara <jack@suse.cz>
SuSE CR Labs

Re: [PATCH] JBD2: round commit timer up to avoid uncommitted transaction

[Theodore Tso]

On Mon, Jul 27, 2009 at 02:22:09PM -0600, Andreas Dilger wrote:
> jbd and jbd2 fixes: fix jiffie rounding in jbd commit timer setup code.
> Rounding down could cause the timer to be fired before the corresponding
> transaction has expired.  That transaction can stay not committed forever if
> no new transaction is created or expicit sync/umount happens.
> 
> Signed-off-by: Alex Zhuravlev (Tomas) <alex.zhuravlev@sun.com>
> Signed-off-by: Andreas Dilger <adilger@sun.com>

Added to the ext4 patch queue; apologies for the delay.

      	     	  	       		     - Ted