file: reinstate f_pos locking optimization for regular files

In commit 20ea1e7d13c1 ("file: always lock position for
FMODE_ATOMIC_POS") we ended up always taking the file pos lock, because
pidfd_getfd() could get a reference to the file even when it didn't have
an elevated file count due to threading of other sharing cases.

But Mateusz Guzik reports that the extra locking is actually measurable,
so let's re-introduce the optimization, and only force the locking for
directory traversal.

Directories need the lock for correctness reasons, while regular files
only need it for "POSIX semantics".  Since pidfd_getfd() is about
debuggers etc special things that are _way_ outside of POSIX, we can
relax the rules for that case.

Reported-by: Mateusz Guzik <mjguzik@gmail.com>
Cc: Christian Brauner <brauner@kernel.org>
Link: https://lore.kernel.org/linux-fsdevel/20230803095311.ijpvhx3fyrbkasul@f/
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

1 files changed, 17 insertions, 1 deletions

diff --git a/fs/file.c b/fs/file.c
index 35c62b54c9d6..dbca26ef7a01 100644
--- a/fs/file.c
+++ b/fs/file.c
@@ -1036,12 +1036,28 @@ unsigned long __fdget_raw(unsigned int fd)
 	return __fget_light(fd, 0);
 }
 
+/*
+ * Try to avoid f_pos locking. We only need it if the
+ * file is marked for FMODE_ATOMIC_POS, and it can be
+ * accessed multiple ways.
+ *
+ * Always do it for directories, because pidfd_getfd()
+ * can make a file accessible even if it otherwise would
+ * not be, and for directories this is a correctness
+ * issue, not a "POSIX requirement".
+ */
+static inline bool file_needs_f_pos_lock(struct file *file)
+{
+	return (file->f_mode & FMODE_ATOMIC_POS) &&
+		(file_count(file) > 1 || S_ISDIR(file_inode(file)->i_mode));
+}
+
 unsigned long __fdget_pos(unsigned int fd)
 {
 	unsigned long v = __fdget(fd);
 	struct file *file = (struct file *)(v & ~3);
 
-	if (file && (file->f_mode & FMODE_ATOMIC_POS)) {
+	if (file && file_needs_f_pos_lock(file)) {
 		v |= FDPUT_POS_UNLOCK;
 		mutex_lock(&file->f_pos_lock);
 	}