Re: [locks] 6d390e4b5d: will-it-scale.per_process_ops -96.6% regression

[Jeff Layton]

On Mon, 2020-03-09 at 13:22 -0400, Jeff Layton wrote:
> On Mon, 2020-03-09 at 08:52 -0700, Linus Torvalds wrote:
> > On Mon, Mar 9, 2020 at 7:36 AM Jeff Layton <jlayton@xxxxxxxxxx> wrote:
> > > On Sun, 2020-03-08 at 22:03 +0800, kernel test robot wrote:
> > > > FYI, we noticed a -96.6% regression of will-it-scale.per_process_ops due to commit:
> > > 
> > > This is not completely unexpected as we're banging on the global
> > > blocked_lock_lock now for every unlock. This test just thrashes file
> > > locks and unlocks without doing anything in between, so the workload
> > > looks pretty artificial [1].
> > > 
> > > It would be nice to avoid the global lock in this codepath, but it
> > > doesn't look simple to do. I'll keep thinking about it, but for now I'm
> > > inclined to ignore this result unless we see a problem in more realistic
> > > workloads.
> > 
> > That is a _huge_ regression, though.
> > 
> > What about something like the attached? Wouldn't that work? And make
> > the code actually match the old comment about wow "fl_blocker" being
> > NULL being special.
> > 
> > The old code seemed to not know about things like memory ordering either.
> > 
> > Patch is entirely untested, but aims to have that "smp_store_release()
> > means I'm done and not going to touch it any more", making that
> > smp_load_acquire() test hopefully be valid as per the comment..
> 
> Yeah, something along those lines maybe. I don't think we can use
> fl_blocker that way though, as the wait_event_interruptible is waiting
> on it to go to NULL, and the wake_up happens before fl_blocker is
> cleared.
> 
> Maybe we need to mix in some sort of FL_BLOCK_ACTIVE flag and use that
> instead of testing for !fl_blocker to see whether we can avoid the
> blocked_lock_lock?
>   

How about something like this instead? (untested other than for
compilation)

Basically, this just switches the waiters over to wait for
fl_blocked_member to go empty. That still happens before the wakeup, so
it should be ok to wait on that.

I think we can also eliminate the lockless list_empty check in
locks_delete_block, as the fl_blocker check should be sufficient now.
-- 
Jeff Layton <jlayton@xxxxxxxxxx>
From c179d779c9b72838ed9996a65d686d86679d1639 Mon Sep 17 00:00:00 2001
From: Linus Torvalds <torvalds@xxxxxxxxxxxxxxxxxxxx>
Date: Mon, 9 Mar 2020 14:35:43 -0400
Subject: [PATCH] locks: reinstate locks_delete_lock optimization

...by using smp_load_acquire and smp_store_release to close the race
window.

[ jlayton: wait on the fl_blocked_requests list to go empty instead of
	   the fl_blocker pointer to clear. Remove the list_empty check
	   from locks_delete_lock shortcut. ]

Signed-off-by: Jeff Layton <jlayton@xxxxxxxxxx>
---
 fs/cifs/file.c |  3 ++-
 fs/locks.c     | 42 ++++++++++++++++++++++++++++++++++++------
 2 files changed, 38 insertions(+), 7 deletions(-)

diff --git a/fs/cifs/file.c b/fs/cifs/file.c
index 3b942ecdd4be..8f9d849a0012 100644
--- a/fs/cifs/file.c
+++ b/fs/cifs/file.c
@@ -1169,7 +1169,8 @@ cifs_posix_lock_set(struct file *file, struct file_lock *flock)
 	rc = posix_lock_file(file, flock, NULL);
 	up_write(&cinode->lock_sem);
 	if (rc == FILE_LOCK_DEFERRED) {
-		rc = wait_event_interruptible(flock->fl_wait, !flock->fl_blocker);
+		rc = wait_event_interruptible(flock->fl_wait,
+					list_empty(&flock->fl_blocked_member));
 		if (!rc)
 			goto try_again;
 		locks_delete_block(flock);
diff --git a/fs/locks.c b/fs/locks.c
index 426b55d333d5..30923db708c2 100644
--- a/fs/locks.c
+++ b/fs/locks.c
@@ -725,7 +725,6 @@ static void __locks_delete_block(struct file_lock *waiter)
 {
 	locks_delete_global_blocked(waiter);
 	list_del_init(&waiter->fl_blocked_member);
-	waiter->fl_blocker = NULL;
 }
 
 static void __locks_wake_up_blocks(struct file_lock *blocker)
@@ -740,6 +739,12 @@ static void __locks_wake_up_blocks(struct file_lock *blocker)
 			waiter->fl_lmops->lm_notify(waiter);
 		else
 			wake_up(&waiter->fl_wait);
+
+		/*
+		 * Tell the world we're done with it - see comment at
+		 * top of locks_delete_block().
+		 */
+		smp_store_release(&waiter->fl_blocker, NULL);
 	}
 }
 
@@ -753,11 +758,31 @@ int locks_delete_block(struct file_lock *waiter)
 {
 	int status = -ENOENT;
 
+	/*
+	 * If fl_blocker is NULL, it won't be set again as this thread
+	 * "owns" the lock and is the only one that might try to claim
+	 * the lock.  So it is safe to test fl_blocker locklessly.
+	 * Also if fl_blocker is NULL, this waiter is not listed on
+	 * fl_blocked_requests for some lock, so no other request can
+	 * be added to the list of fl_blocked_requests for this
+	 * request.  So if fl_blocker is NULL, it is safe to
+	 * locklessly check if fl_blocked_requests is empty.  If both
+	 * of these checks succeed, there is no need to take the lock.
+	 */
+	if (!smp_load_acquire(&waiter->fl_blocker))
+		return status;
+
 	spin_lock(&blocked_lock_lock);
 	if (waiter->fl_blocker)
 		status = 0;
 	__locks_wake_up_blocks(waiter);
 	__locks_delete_block(waiter);
+
+	/*
+	 * Tell the world we're done with it - see comment at top
+	 * of this function
+	 */
+	smp_store_release(&waiter->fl_blocker, NULL);
 	spin_unlock(&blocked_lock_lock);
 	return status;
 }
@@ -1350,7 +1375,8 @@ static int posix_lock_inode_wait(struct inode *inode, struct file_lock *fl)
 		error = posix_lock_inode(inode, fl, NULL);
 		if (error != FILE_LOCK_DEFERRED)
 			break;
-		error = wait_event_interruptible(fl->fl_wait, !fl->fl_blocker);
+		error = wait_event_interruptible(fl->fl_wait,
+					list_empty(&fl->fl_blocked_member));
 		if (error)
 			break;
 	}
@@ -1435,7 +1461,8 @@ int locks_mandatory_area(struct inode *inode, struct file *filp, loff_t start,
 		error = posix_lock_inode(inode, &fl, NULL);
 		if (error != FILE_LOCK_DEFERRED)
 			break;
-		error = wait_event_interruptible(fl.fl_wait, !fl.fl_blocker);
+		error = wait_event_interruptible(fl.fl_wait,
+					list_empty(&fl.fl_blocked_member));
 		if (!error) {
 			/*
 			 * If we've been sleeping someone might have
@@ -1638,7 +1665,8 @@ int __break_lease(struct inode *inode, unsigned int mode, unsigned int type)
 
 	locks_dispose_list(&dispose);
 	error = wait_event_interruptible_timeout(new_fl->fl_wait,
-						!new_fl->fl_blocker, break_time);
+					list_empty(&new_fl->fl_blocked_member),
+					break_time);
 
 	percpu_down_read(&file_rwsem);
 	spin_lock(&ctx->flc_lock);
@@ -2122,7 +2150,8 @@ static int flock_lock_inode_wait(struct inode *inode, struct file_lock *fl)
 		error = flock_lock_inode(inode, fl);
 		if (error != FILE_LOCK_DEFERRED)
 			break;
-		error = wait_event_interruptible(fl->fl_wait, !fl->fl_blocker);
+		error = wait_event_interruptible(fl->fl_wait,
+				list_empty(&fl->fl_blocked_member));
 		if (error)
 			break;
 	}
@@ -2399,7 +2428,8 @@ static int do_lock_file_wait(struct file *filp, unsigned int cmd,
 		error = vfs_lock_file(filp, cmd, fl, NULL);
 		if (error != FILE_LOCK_DEFERRED)
 			break;
-		error = wait_event_interruptible(fl->fl_wait, !fl->fl_blocker);
+		error = wait_event_interruptible(fl->fl_wait,
+					list_empty(&fl->fl_blocked_member));
 		if (error)
 			break;
 	}
-- 
2.24.1