[PATCH AUTOSEL 5.15 07/24] fs: use acquire ordering in __fget_light()

[Sasha Levin]

From: Jann Horn <jannh@xxxxxxxxxx>

[ Upstream commit 7ee47dcfff1835ff75a794d1075b6b5f5462cfed ]

We must prevent the CPU from reordering the files->count read with the
FD table access like this, on architectures where read-read reordering is
possible:

    files_lookup_fd_raw()
                                  close_fd()
                                  put_files_struct()
    atomic_read(&files->count)

I would like to mark this for stable, but the stable rules explicitly say
"no theoretical races", and given that the FD table pointer and
files->count are explicitly stored in the same cacheline, this sort of
reordering seems quite unlikely in practice...

Signed-off-by: Jann Horn <jannh@xxxxxxxxxx>
Signed-off-by: Al Viro <viro@xxxxxxxxxxxxxxxxxx>
Signed-off-by: Sasha Levin <sashal@xxxxxxxxxx>
---
 fs/file.c | 11 ++++++++++-
 1 file changed, 10 insertions(+), 1 deletion(-)

diff --git a/fs/file.c b/fs/file.c
index ee9317346702..214364e19d76 100644
--- a/fs/file.c
+++ b/fs/file.c
@@ -1029,7 +1029,16 @@ static unsigned long __fget_light(unsigned int fd, fmode_t mask)
 	struct files_struct *files = current->files;
 	struct file *file;
 
-	if (atomic_read(&files->count) == 1) {
+	/*
+	 * If another thread is concurrently calling close_fd() followed
+	 * by put_files_struct(), we must not observe the old table
+	 * entry combined with the new refcount - otherwise we could
+	 * return a file that is concurrently being freed.
+	 *
+	 * atomic_read_acquire() pairs with atomic_dec_and_test() in
+	 * put_files_struct().
+	 */
+	if (atomic_read_acquire(&files->count) == 1) {
 		file = files_lookup_fd_raw(files, fd);
 		if (!file || unlikely(file->f_mode & mask))
 			return 0;
-- 
2.35.1