From: Al Viro <viro@zeniv.linux.org.uk>
Date: Tue, 31 Oct 2023 04:23:35 +0000 (-0400)
Subject: dentry_kill(): don't bother with retain_dentry() on slow path
X-Git-Url: http://git.maquefel.me/?a=commitdiff_plain;h=b06c684d3984ef64e792ec3e89889c96cab97b5e;p=linux.git

dentry_kill(): don't bother with retain_dentry() on slow path

We have already checked it and dentry used to look not worthy
of keeping.  The only hard obstacle to evicting dentry is
non-zero refcount; everything else is advisory - e.g. memory
pressure could evict any dentry found with refcount zero.
On the slow path in dentry_kill() we had dropped and regained
->d_lock; we must recheck the refcount, but everything else
is not worth bothering with.

Note that filesystem can not count upon ->d_delete() being
called for dentry - not even once.  Again, memory pressure
(as well as d_prune_aliases(), or attempted rmdir() of ancestor,
or...) will not call ->d_delete() at all.

So from the correctness point of view we are fine doing the
check only once.  And it makes things simpler down the road.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---

diff --git a/fs/dcache.c b/fs/dcache.c
index b527db8e5901e..80992e49561c0 100644
--- a/fs/dcache.c
+++ b/fs/dcache.c
@@ -739,14 +739,10 @@ slow_positive:
 	spin_lock(&dentry->d_lock);
 	parent = lock_parent(dentry);
 got_locks:
-	if (unlikely(dentry->d_lockref.count != 1)) {
-		dentry->d_lockref.count--;
-	} else if (likely(!retain_dentry(dentry))) {
-		dentry->d_lockref.count--;
+	dentry->d_lockref.count--;
+	if (likely(dentry->d_lockref.count == 0)) {
 		__dentry_kill(dentry);
 		return parent;
-	} else {
-		dentry->d_lockref.count--;
 	}
 	/* we are keeping it, after all */
 	if (inode)