From: Sergei Trofimovich <slyfox@xxxxxxxxxx>
iput() shouldn't be called for inodes in I_NEW state.
We need to mark inode as constructed first.
WARNING: at fs/inode.c:1309 iput+0x20b/0x210()
Call Trace:
[<ffffffff8103e7ba>] warn_slowpath_common+0x7a/0xb0
[<ffffffff8103e805>] warn_slowpath_null+0x15/0x20
[<ffffffff810eaf0b>] iput+0x20b/0x210
[<ffffffff811b96fb>] btrfs_iget+0x1eb/0x4a0
[<ffffffff811c3ad6>] btrfs_run_defrag_inodes+0x136/0x210
[<ffffffff811ad55f>] cleaner_kthread+0x17f/0x1a0
[<ffffffff81035b7d>] ? sub_preempt_count+0x9d/0xd0
[<ffffffff811ad3e0>] ? transaction_kthread+0x280/0x280
[<ffffffff8105af86>] kthread+0x96/0xa0
[<ffffffff814336d4>] kernel_thread_helper+0x4/0x10
[<ffffffff8105aef0>] ? kthread_worker_fn+0x190/0x190
[<ffffffff814336d0>] ? gs_change+0xb/0xb
Signed-off-by: Sergei Trofimovich <slyfox@xxxxxxxxxx>
CC: Konstantin Khlebnikov <khlebnikov@xxxxxxxxxx>
Tested-by: David Sterba <dsterba@xxxxxxx>
CC: Josef Bacik <josef@xxxxxxxxxx>
CC: Chris Mason <chris.mason@xxxxxxxxxx>
---
Change since v1:
- Added David's positive feedback
Change since Konstantin's patch:
- avoid using explicit destructor for inode, as it can(?)
be shared among other threads. Use unlock_new_inode()/iput()
instead. It fixes 100% CPU load after 'sync()' popped-up in
Konstantin's patch
fs/btrfs/inode.c | 10 +++-------
1 files changed, 3 insertions(+), 7 deletions(-)
diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index 0ccc743..ec4b550 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -3952,7 +3952,6 @@ struct inode *btrfs_iget(struct super_block *s, struct btrfs_key *location,
struct btrfs_root *root, int *new)
{
struct inode *inode;
- int bad_inode = 0;
inode = btrfs_iget_locked(s, location->objectid, root);
if (!inode)
@@ -3968,15 +3967,12 @@ struct inode *btrfs_iget(struct super_block *s, struct btrfs_key *location,
if (new)
*new = 1;
} else {
- bad_inode = 1;
+ unlock_new_inode(inode);
+ iput(inode);
+ inode = ERR_PTR(-ESTALE);
}
}
- if (bad_inode) {
- iput(inode);
- inode = ERR_PTR(-ESTALE);
- }
-
return inode;
}
--
1.7.3.4
--
To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
