diff mbox series

btrfs: avoid pointless wake ups of drew lock readers

Message ID ed6897d89e53696370450d72a366f96ad954edbd.1710760179.git.fdmanana@suse.com (mailing list archive)
State New, archived
Headers show
Series btrfs: avoid pointless wake ups of drew lock readers | expand

Commit Message

Filipe Manana March 18, 2024, 11:11 a.m. UTC
From: Filipe Manana <fdmanana@suse.com>

When unlocking a write lock on a drew lock, at btrfs_drew_write_unlock(),
it's pointless to wake up tasks waiting to acquire a read lock if we
didn't decrement the 'writers' counter down to 0, since a read lock can
only be acquired when the counter reaches a value of 0. Doing so is
harmless from a functional point of view, but it's not efficient due to
unnecessarily waking up tasks just for them to sleep again on the
waitqueue.

So change this to wake up readers only if we decremented the 'writers'
counter to 0.

Signed-off-by: Filipe Manana <fdmanana@suse.com>
---
 fs/btrfs/locking.c | 8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)

Comments

David Sterba March 18, 2024, 10:14 p.m. UTC | #1
On Mon, Mar 18, 2024 at 11:11:56AM +0000, fdmanana@kernel.org wrote:
> From: Filipe Manana <fdmanana@suse.com>
> 
> When unlocking a write lock on a drew lock, at btrfs_drew_write_unlock(),
> it's pointless to wake up tasks waiting to acquire a read lock if we
> didn't decrement the 'writers' counter down to 0, since a read lock can
> only be acquired when the counter reaches a value of 0. Doing so is
> harmless from a functional point of view, but it's not efficient due to
> unnecessarily waking up tasks just for them to sleep again on the
> waitqueue.
> 
> So change this to wake up readers only if we decremented the 'writers'
> counter to 0.
> 
> Signed-off-by: Filipe Manana <fdmanana@suse.com>

Reviewed-by: David Sterba <dsterba@suse.com>
diff mbox series

Patch

diff --git a/fs/btrfs/locking.c b/fs/btrfs/locking.c
index 508a3fdfcd58..72992e74c479 100644
--- a/fs/btrfs/locking.c
+++ b/fs/btrfs/locking.c
@@ -364,8 +364,12 @@  void btrfs_drew_write_lock(struct btrfs_drew_lock *lock)
 
 void btrfs_drew_write_unlock(struct btrfs_drew_lock *lock)
 {
-	atomic_dec(&lock->writers);
-	cond_wake_up(&lock->pending_readers);
+	/*
+	 * atomic_dec_and_test() implies a full barrier, so woken up readers are
+	 * guaranteed to see the decrement.
+	 */
+	if (atomic_dec_and_test(&lock->writers))
+		wake_up(&lock->pending_readers);
 }
 
 void btrfs_drew_read_lock(struct btrfs_drew_lock *lock)