From: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
Date: Sun, 28 Dec 2008 07:11:02 +0000 (+0200)
Subject: UBIFS: always commit on unmount
X-Git-Tag: v2.6.29-rc1~533^2~10
X-Git-Url: http://www.pilppa.org/gitweb/gitweb.cgi?a=commitdiff_plain;h=26d05777b0a23062a39e83c369c0a3583918f164;p=linux-2.6-omap-h63xx.git

UBIFS: always commit on unmount

UBIFS commits on unmount to make the next mount faster. Currently,
it commits only if there is more than LEB size bytes in the
journal. This is not very good, because journal size may be
large (512KiB). And there may be few deletions in the journal
which do not take much journal space, but which do introduce
a lot of TNC changes and make mount slow.

Thus, jurt remove this condition and always commit.

Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
---

diff --git a/fs/ubifs/super.c b/fs/ubifs/super.c
index a14703e0a9a..1c1bbe4135c 100644
--- a/fs/ubifs/super.c
+++ b/fs/ubifs/super.c
@@ -1570,20 +1570,24 @@ out:
  * @c: UBIFS file-system description object
  *
  * This function is called during un-mounting and re-mounting, and it commits
- * the journal unless the "fast unmount" mode is enabled. It also avoids
- * committing the journal if it contains too few data.
+ * the journal unless the "fast unmount" mode is enabled.
  */
 static void commit_on_unmount(struct ubifs_info *c)
 {
-	if (!c->fast_unmount) {
-		long long bud_bytes;
+	struct super_block *sb = c->vfs_sb;
+	long long bud_bytes;
 
-		spin_lock(&c->buds_lock);
-		bud_bytes = c->bud_bytes;
-		spin_unlock(&c->buds_lock);
-		if (bud_bytes > c->leb_size)
-			ubifs_run_commit(c);
-	}
+	/*
+	 * This function is called before the background thread is stopped, so
+	 * we may race with ongoing commit, which means we have to take
+	 * @c->bud_lock to access @c->bud_bytes.
+	 */
+	spin_lock(&c->buds_lock);
+	bud_bytes = c->bud_bytes;
+	spin_unlock(&c->buds_lock);
+
+	if (!c->fast_unmount && !(sb->s_flags & MS_RDONLY) && bud_bytes)
+		ubifs_run_commit(c);
 }
 
 /**
@@ -2009,7 +2013,7 @@ static void ubifs_kill_sb(struct super_block *sb)
 	 * We do 'commit_on_unmount()' here instead of 'ubifs_put_super()'
 	 * in order to be outside BKL.
 	 */
-	if (sb->s_root && !(sb->s_flags & MS_RDONLY))
+	if (sb->s_root)
 		commit_on_unmount(c);
 	/* The un-mount routine is actually done in put_super() */
 	generic_shutdown_super(sb);