btrfs: add barriers to btrfs_sync_log before log_commit_wait wakeups

Currently the code assumes that there's an implied barrier by the
sequence of code preceding the wakeup, namely the mutex unlock.

As Nikolay pointed out:

I think this is wrong (not your code) but the original assumption that
the RELEASE semantics provided by mutex_unlock is sufficient.
According to memory-barriers.txt:

Section 'LOCK ACQUISITION FUNCTIONS' states:

(2) RELEASE operation implication:

Memory operations issued before the RELEASE will be completed before the
RELEASE operation has completed.

Memory operations issued after the RELEASE *may* be completed before the
RELEASE operation has completed.

(I've bolded the may portion)

The example given there:

As an example, consider the following:

*A = a;
*B = b;
ACQUIRE
*C = c;
*D = d;
RELEASE
*E = e;
*F = f;

The following sequence of events is acceptable:

ACQUIRE, {*F,*A}, *E, {*C,*D}, *B, RELEASE

So if we assume that *C is modifying the flag which the waitqueue is checking,
and *E is the actual wakeup, then those accesses can be re-ordered...

IMHO this code should be considered broken...
---

To be on the safe side, add the barriers. The synchronization logic
around log using the mutexes and several other threads does not make it
easy to reason for/against the barrier.

CC: Nikolay Borisov <[e-mail]>
Link: https://lkml.kernel.org/r/[e-mail]
Reviewed-by: Nikolay Borisov <[e-mail]>
Signed-off-by: David Sterba <[e-mail]>

fs/btrfs/tree-log.c

···
 	mutex_unlock(&log_root_tree->log_mutex);
 
 	/*
-	 * The barrier before waitqueue_active is implied by mutex_unlock
+	 * The barrier before waitqueue_active is needed so all the updates
+	 * above are seen by the woken threads. It might not be necessary, but
+	 * proving that seems to be hard.
 	 */
+	smp_mb();
 	if (waitqueue_active(&log_root_tree->log_commit_wait[index2]))
 		wake_up(&log_root_tree->log_commit_wait[index2]);
 out:
···
 	mutex_unlock(&root->log_mutex);
 
 	/*
-	 * The barrier before waitqueue_active is implied by mutex_unlock
+	 * The barrier before waitqueue_active is needed so all the updates
+	 * above are seen by the woken threads. It might not be necessary, but
+	 * proving that seems to be hard.
 	 */
+	smp_mb();
 	if (waitqueue_active(&root->log_commit_wait[index1]))
 		wake_up(&root->log_commit_wait[index1]);
 	return ret;