1 Under rare conditions (filesystem corruption, really) it is possible
2 for ext3_dirty_inode() to require _two_ blocks for the transaction: one
3 for the inode and one to update the superblock - to set
4 EXT3_FEATURE_RO_COMPAT_LARGE_FILE. This causes the filesystem to go
7 So reserve an additional block for that eventuality.
10 fs/ext3/inode.c | 2 +-
11 1 files changed, 1 insertion(+), 1 deletion(-)
13 --- 25/fs/ext3/inode.c~ext3-transaction-reserved-blocks Sat Dec 14 18:28:21 2002
14 +++ 25-akpm/fs/ext3/inode.c Sat Dec 14 18:28:21 2002
15 @@ -2698,7 +2698,7 @@ void ext3_dirty_inode(struct inode *inod
19 - handle = ext3_journal_start(inode, 1);
20 + handle = ext3_journal_start(inode, 2);