diff options
| author | Dave Kleikamp <dave.kleikamp@oracle.com> | 2011-06-20 15:53:46 (GMT) |
|---|---|---|
| committer | Dave Kleikamp <dave.kleikamp@oracle.com> | 2011-06-20 15:53:46 (GMT) |
| commit | d31b53e3cd069e02290ed8a648aa8c7618d6fe77 (patch) | |
| tree | 529f1fab64d7b525e1a77510f3c9b02d24597c65 /fs/jfs/file.c | |
| parent | 28e0fa894cd5996d3007ce82f07226f79beb7286 (diff) | |
| download | linux-d31b53e3cd069e02290ed8a648aa8c7618d6fe77.tar.xz | |
JFS: Don't save agno in the inode
Resizing the file system can result in an in-memory inode being remapped
to a different aggregate group (AG). A cached AG number can cause
problems when trying to free or allocate inodes. Instead, save the IAG's
agstart address and calculate the agno when we need it.
Signed-off-by: Dave Kleikamp <dave.kleikamp@oracle.com>
Diffstat (limited to 'fs/jfs/file.c')
| -rw-r--r-- | fs/jfs/file.c | 6 |
1 files changed, 3 insertions, 3 deletions
diff --git a/fs/jfs/file.c b/fs/jfs/file.c index c5ce6c1..2f3f531 100644 --- a/fs/jfs/file.c +++ b/fs/jfs/file.c @@ -66,9 +66,9 @@ static int jfs_open(struct inode *inode, struct file *file) struct jfs_inode_info *ji = JFS_IP(inode); spin_lock_irq(&ji->ag_lock); if (ji->active_ag == -1) { - ji->active_ag = ji->agno; - atomic_inc( - &JFS_SBI(inode->i_sb)->bmap->db_active[ji->agno]); + struct jfs_sb_info *jfs_sb = JFS_SBI(inode->i_sb); + ji->active_ag = BLKTOAG(addressPXD(&ji->ixpxd), jfs_sb); + atomic_inc( &jfs_sb->bmap->db_active[ji->active_ag]); } spin_unlock_irq(&ji->ag_lock); } |