fs/notify: optimize inotify/fsnotify code for unwatched files

I have a _tiny_ microbenchmark that sits in a loop and writes single
bytes to a file.  Writing one byte to a tmpfs file is around 2x slower
than reading one byte from a file, which is a _bit_ more than I expecte.
This is a dumb benchmark, but I think it's hard to deny that write() is
a hot path and we should avoid unnecessary overhead there.

I did a 'perf record' of 30-second samples of read and write.  The top
item in a diffprofile is srcu_read_lock() from fsnotify().  There are
active inotify fd's from systemd, but nothing is actually listening to
the file or its part of the filesystem.

I *think* we can avoid taking the srcu_read_lock() for the common case
where there are no actual marks on the file.  This means that there will
both be nothing to notify for *and* implies that there is no need for
clearing the ignore mask.

This patch gave a 13.1% speedup in writes/second on my test, which is an
improvement from the 10.8% that I saw with the last version.
Signed-off-by: NDave Hansen <dave.hansen@linux.intel.com>
Reviewed-by: NJan Kara <jack@suse.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Eric Paris <eparis@redhat.com>
Cc: John McCutchan <john@johnmccutchan.com>
Cc: Robert Love <rlove@rlove.org>
Cc: Andi Kleen <ak@linux.intel.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

fs/notify/fsnotify.c

@@ -204,6 +204,16 @@ int fsnotify(struct inode *to_tell, __u32 mask, void *data, int data_is,
 	else
 		mnt = NULL;
 
+	/*
+	 * Optimization: srcu_read_lock() has a memory barrier which can
+	 * be expensive.  It protects walking the *_fsnotify_marks lists.
+	 * However, if we do not walk the lists, we do not have to do
+	 * SRCU because we have no references to any objects and do not
+	 * need SRCU to keep them "alive".
+	 */
+	if (hlist_empty(&to_tell->i_fsnotify_marks) &&
+	    (!mnt || hlist_empty(&mnt->mnt_fsnotify_marks)))
+		return 0;
 	/*
 	 * if this is a modify event we may need to clear the ignored masks
 	 * otherwise return if neither the inode nor the vfsmount care about