[PATCH v2] Btrfs: fix deadlock in btrfs_commit_transaction

[PATCH v2] Btrfs: fix deadlock in btrfs_commit_transaction

[Sage Weil]

We calculate timeout (either 1 or MAX_SCHEDULE_TIMEOUT) based on whether
num_writers > 1 or should_grow at the top of the loop.  Then, much much
later, we wait for that timeout if either num_writers or should_grow is
true.  However, it's possible for a racing process (calling
btrfs_end_transaction()) to decrement num_writers such that we wait
forever instead of for 1.

Fix this by deciding how long to wait when we wait.  Include a smp_mb()
before checking if the waitqueue is active to ensure the num_writers
is visible.

Signed-off-by: Sage Weil <sage@newdream.net>
---
v2:
 - add smp_mb() before waitqueue_active() check to clone another possible race

 fs/btrfs/transaction.c |   13 +++++--------
 1 files changed, 5 insertions(+), 8 deletions(-)

diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c
index 66e4c66..b461fe3 100644
--- a/fs/btrfs/transaction.c
+++ b/fs/btrfs/transaction.c
@@ -392,6 +392,7 @@ static int __btrfs_end_transaction(struct btrfs_trans_handle *trans,
 	WARN_ON(cur_trans->num_writers < 1);
 	cur_trans->num_writers--;
 
+	smp_mb();
 	if (waitqueue_active(&cur_trans->writer_wait))
 		wake_up(&cur_trans->writer_wait);
 	put_transaction(cur_trans);
@@ -992,7 +993,6 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans,
 			     struct btrfs_root *root)
 {
 	unsigned long joined = 0;
-	unsigned long timeout = 1;
 	struct btrfs_transaction *cur_trans;
 	struct btrfs_transaction *prev_trans = NULL;
 	DEFINE_WAIT(wait);
@@ -1063,11 +1063,6 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans,
 			snap_pending = 1;
 
 		WARN_ON(cur_trans != trans->transaction);
-		if (cur_trans->num_writers > 1)
-			timeout = MAX_SCHEDULE_TIMEOUT;
-		else if (should_grow)
-			timeout = 1;
-
 		mutex_unlock(&root->fs_info->trans_mutex);
 
 		if (flush_on_commit || snap_pending) {
@@ -1089,8 +1084,10 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans,
 				TASK_UNINTERRUPTIBLE);
 
 		smp_mb();
-		if (cur_trans->num_writers > 1 || should_grow)
-			schedule_timeout(timeout);
+		if (cur_trans->num_writers > 1)
+			schedule_timeout(MAX_SCHEDULE_TIMEOUT);
+		else if (should_grow)
+			schedule_timeout(1);
 
 		mutex_lock(&root->fs_info->trans_mutex);
 		finish_wait(&cur_trans->writer_wait, &wait);
-- 
1.6.6.1