Merge patch series "mount: handle mount propagation for detached mount trees"

Christian Brauner <brauner@kernel.org> says:

In commit ee2e3f50629f ("mount: fix mounting of detached mounts onto
targets that reside on shared mounts") I fixed a bug where propagating
the source mount tree of an anonymous mount namespace into a target
mount tree of a non-anonymous mount namespace could be used to trigger
an integer overflow in the non-anonymous mount namespace causing any new
mounts to fail.

The cause of this was that the propagation algorithm was unable to
recognize mounts from the source mount tree that were already propagated
into the target mount tree and then reappeared as propagation targets
when walking the destination propagation mount tree.

When fixing this I disabled mount propagation into anonymous mount
namespaces. Make it possible for anonymous mount namespace to receive
mount propagation events correctly. This is no also a correctness issue
now that we allow mounting detached mount trees onto detached mount
trees.

Mark the source anonymous mount namespace with MNTNS_PROPAGATING
indicating that all mounts belonging to this mount namespace are
currently in the process of being propagated and make the propagation
algorithm discard those if they appear as propagation targets.

* patches from https://lore.kernel.org/r/20250225-work-mount-propagation-v1-0-e6e3724500eb@kernel.org:
  selftests: test subdirectory mounting
  selftests: add test for detached mount tree propagation
  mount: handle mount propagation for detached mount trees

Link: https://lore.kernel.org/r/20250225-work-mount-propagation-v1-0-e6e3724500eb@kernel.org
Signed-off-by: Christian Brauner <brauner@kernel.org>

5 files changed, 146 insertions, 66 deletions

diff --git a/fs/mount.h b/fs/mount.h
index e2501a724688..96862eba2246 100644
--- a/fs/mount.h
+++ b/fs/mount.h
@@ -5,6 +5,12 @@
 #include <linux/ns_common.h>
 #include <linux/fs_pin.h>
 
+extern struct list_head notify_list;
+
+typedef __u32 __bitwise mntns_flags_t;
+
+#define MNTNS_PROPAGATING	((__force mntns_flags_t)(1 << 0))
+
 struct mnt_namespace {
 	struct ns_common	ns;
 	struct mount *	root;
@@ -27,6 +33,7 @@ struct mnt_namespace {
 	struct rb_node		mnt_ns_tree_node; /* node in the mnt_ns_tree */
 	struct list_head	mnt_ns_list; /* entry in the sequential list of mounts namespace */
 	refcount_t		passive; /* number references not pinning @mounts */
+	mntns_flags_t		mntns_flags;
 } __randomize_layout;
 
 struct mnt_pcp {
diff --git a/fs/namespace.c b/fs/namespace.c
index a7da41e81673..59894e3f3e68 100644
--- a/fs/namespace.c
+++ b/fs/namespace.c
@@ -3536,13 +3536,6 @@ static int do_move_mount(struct path *old_path,
 	if (!may_use_mount(p))
 		goto out;
 
-	/*
-	 * Don't allow moving an attached mount tree to an anonymous
-	 * mount tree.
-	 */
-	if (!is_anon_ns(ns) && is_anon_ns(p->mnt_ns))
-		goto out;
-
 	/* The thing moved must be mounted... */
 	if (!is_mounted(&old->mnt))
 		goto out;
@@ -3551,15 +3544,31 @@ static int do_move_mount(struct path *old_path,
 	if (!(attached ? check_mnt(old) : is_anon_ns(ns)))
 		goto out;
 
-	/*
-	 * Ending up with two files referring to the root of the same
-	 * anonymous mount namespace would cause an error as this would
-	 * mean trying to move the same mount twice into the mount tree
-	 * which would be rejected later. But be explicit about it right
-	 * here.
-	 */
-	if (is_anon_ns(ns) && is_anon_ns(p->mnt_ns) && ns == p->mnt_ns)
+	if (is_anon_ns(ns)) {
+		/*
+		 * Ending up with two files referring to the root of the
+		 * same anonymous mount namespace would cause an error
+		 * as this would mean trying to move the same mount
+		 * twice into the mount tree which would be rejected
+		 * later. But be explicit about it right here.
+		 */
+		if ((is_anon_ns(p->mnt_ns) && ns == p->mnt_ns))
+			goto out;
+
+		/*
+		 * If this is an anonymous mount tree ensure that mount
+		 * propagation can detect mounts that were just
+		 * propagated to the target mount tree so we don't
+		 * propagate onto them.
+		 */
+		ns->mntns_flags |= MNTNS_PROPAGATING;
+	} else if (is_anon_ns(p->mnt_ns)) {
+		/*
+		 * Don't allow moving an attached mount tree to an
+		 * anonymous mount tree.
+		 */
 		goto out;
+	}
 
 	if (old->mnt.mnt_flags & MNT_LOCKED)
 		goto out;
@@ -3603,6 +3612,9 @@ static int do_move_mount(struct path *old_path,
 	if (err)
 		goto out;
 
+	if (is_anon_ns(ns))
+		ns->mntns_flags &= ~MNTNS_PROPAGATING;
+
 	/* if the mount is moved, it should no longer be expire
 	 * automatically */
 	list_del_init(&old->mnt_expire);
diff --git a/fs/pnode.c b/fs/pnode.c
index ef048f008bdd..6dd26ca96cdd 100644
--- a/fs/pnode.c
+++ b/fs/pnode.c
@@ -150,7 +150,7 @@ static struct mount *propagation_next(struct mount *m,
 					 struct mount *origin)
 {
 	/* are there any slaves of this mount? */
-	if (!IS_MNT_NEW(m) && !list_empty(&m->mnt_slave_list))
+	if (!IS_MNT_PROPAGATED(m) && !list_empty(&m->mnt_slave_list))
 		return first_slave(m);
 
 	while (1) {
@@ -174,7 +174,7 @@ static struct mount *skip_propagation_subtree(struct mount *m,
 	 * Advance m such that propagation_next will not return
 	 * the slaves of m.
 	 */
-	if (!IS_MNT_NEW(m) && !list_empty(&m->mnt_slave_list))
+	if (!IS_MNT_PROPAGATED(m) && !list_empty(&m->mnt_slave_list))
 		m = last_slave(m);
 
 	return m;
@@ -185,7 +185,7 @@ static struct mount *next_group(struct mount *m, struct mount *origin)
 	while (1) {
 		while (1) {
 			struct mount *next;
-			if (!IS_MNT_NEW(m) && !list_empty(&m->mnt_slave_list))
+			if (!IS_MNT_PROPAGATED(m) && !list_empty(&m->mnt_slave_list))
 				return first_slave(m);
 			next = next_peer(m);
 			if (m->mnt_group_id == origin->mnt_group_id) {
@@ -226,7 +226,7 @@ static int propagate_one(struct mount *m, struct mountpoint *dest_mp)
 	struct mount *child;
 	int type;
 	/* skip ones added by this propagate_mnt() */
-	if (IS_MNT_NEW(m))
+	if (IS_MNT_PROPAGATED(m))
 		return 0;
 	/* skip if mountpoint isn't covered by it */
 	if (!is_subdir(dest_mp->m_dentry, m->mnt.mnt_root))
@@ -380,7 +380,7 @@ bool propagation_would_overmount(const struct mount *from,
 	if (!IS_MNT_SHARED(from))
 		return false;
 
-	if (IS_MNT_NEW(to))
+	if (IS_MNT_PROPAGATED(to))
 		return false;
 
 	if (to->mnt.mnt_root != mp->m_dentry)
diff --git a/fs/pnode.h b/fs/pnode.h
index 0b02a6393891..ddafe0d087ca 100644
--- a/fs/pnode.h
+++ b/fs/pnode.h
@@ -12,7 +12,7 @@
 
 #define IS_MNT_SHARED(m) ((m)->mnt.mnt_flags & MNT_SHARED)
 #define IS_MNT_SLAVE(m) ((m)->mnt_master)
-#define IS_MNT_NEW(m)  (!(m)->mnt_ns || is_anon_ns((m)->mnt_ns))
+#define IS_MNT_PROPAGATED(m) (!(m)->mnt_ns || ((m)->mnt_ns->mntns_flags & MNTNS_PROPAGATING))
 #define CLEAR_MNT_SHARED(m) ((m)->mnt.mnt_flags &= ~MNT_SHARED)
 #define IS_MNT_UNBINDABLE(m) ((m)->mnt.mnt_flags & MNT_UNBINDABLE)
 #define IS_MNT_MARKED(m) ((m)->mnt.mnt_flags & MNT_MARKED)
diff --git a/tools/testing/selftests/mount_setattr/mount_setattr_test.c b/tools/testing/selftests/mount_setattr/mount_setattr_test.c
index 1e0508cb5c2d..48a000cabc97 100644
--- a/tools/testing/selftests/mount_setattr/mount_setattr_test.c
+++ b/tools/testing/selftests/mount_setattr/mount_setattr_test.c
@@ -20,6 +20,7 @@
 #include <stdarg.h>
 #include <linux/mount.h>
 
+#include "../filesystems/overlayfs/wrappers.h"
 #include "../kselftest_harness.h"
 
 #ifndef CLONE_NEWNS
@@ -177,51 +178,6 @@ static inline int sys_open_tree(int dfd, const char *filename, unsigned int flag
 	return syscall(__NR_open_tree, dfd, filename, flags);
 }
 
-/* move_mount() flags */
-#ifndef MOVE_MOUNT_F_SYMLINKS
-#define MOVE_MOUNT_F_SYMLINKS 0x00000001 /* Follow symlinks on from path */
-#endif
-
-#ifndef MOVE_MOUNT_F_AUTOMOUNTS
-#define MOVE_MOUNT_F_AUTOMOUNTS 0x00000002 /* Follow automounts on from path */
-#endif
-
-#ifndef MOVE_MOUNT_F_EMPTY_PATH
-#define MOVE_MOUNT_F_EMPTY_PATH 0x00000004 /* Empty from path permitted */
-#endif
-
-#ifndef MOVE_MOUNT_T_SYMLINKS
-#define MOVE_MOUNT_T_SYMLINKS 0x00000010 /* Follow symlinks on to path */
-#endif
-
-#ifndef MOVE_MOUNT_T_AUTOMOUNTS
-#define MOVE_MOUNT_T_AUTOMOUNTS 0x00000020 /* Follow automounts on to path */
-#endif
-
-#ifndef MOVE_MOUNT_T_EMPTY_PATH
-#define MOVE_MOUNT_T_EMPTY_PATH 0x00000040 /* Empty to path permitted */
-#endif
-
-#ifndef MOVE_MOUNT_SET_GROUP
-#define MOVE_MOUNT_SET_GROUP 0x00000100 /* Set sharing group instead */
-#endif
-
-#ifndef MOVE_MOUNT_BENEATH
-#define MOVE_MOUNT_BENEATH 0x00000200 /* Mount beneath top mount */
-#endif
-
-#ifndef MOVE_MOUNT__MASK
-#define MOVE_MOUNT__MASK 0x00000377
-#endif
-
-static inline int sys_move_mount(int from_dfd, const char *from_pathname,
-				 int to_dfd, const char *to_pathname,
-				 unsigned int flags)
-{
-	return syscall(__NR_move_mount, from_dfd, from_pathname, to_dfd,
-		       to_pathname, flags);
-}
-
 static ssize_t write_nointr(int fd, const void *buf, size_t count)
 {
 	ssize_t ret;
@@ -1789,6 +1745,41 @@ TEST_F(mount_setattr, open_tree_detached_fail3)
 	ASSERT_EQ(errno, EINVAL);
 }
 
+TEST_F(mount_setattr, open_tree_subfolder)
+{
+	int fd_context, fd_tmpfs, fd_tree;
+
+	fd_context = sys_fsopen("tmpfs", 0);
+	ASSERT_GE(fd_context, 0);
+
+	ASSERT_EQ(sys_fsconfig(fd_context, FSCONFIG_CMD_CREATE, NULL, NULL, 0), 0);
+
+	fd_tmpfs = sys_fsmount(fd_context, 0, 0);
+	ASSERT_GE(fd_tmpfs, 0);
+
+	EXPECT_EQ(close(fd_context), 0);
+
+	ASSERT_EQ(mkdirat(fd_tmpfs, "subdir", 0755), 0);
+
+	fd_tree = sys_open_tree(fd_tmpfs, "subdir",
+				AT_NO_AUTOMOUNT | AT_SYMLINK_NOFOLLOW |
+				AT_RECURSIVE | OPEN_TREE_CLOEXEC |
+				OPEN_TREE_CLONE);
+	ASSERT_GE(fd_tree, 0);
+
+	EXPECT_EQ(close(fd_tmpfs), 0);
+
+	ASSERT_EQ(mkdirat(-EBADF, "/mnt/open_tree_subfolder", 0755), 0);
+
+	ASSERT_EQ(sys_move_mount(fd_tree, "", -EBADF, "/mnt/open_tree_subfolder", MOVE_MOUNT_F_EMPTY_PATH), 0);
+
+	EXPECT_EQ(close(fd_tree), 0);
+
+	ASSERT_EQ(umount2("/mnt/open_tree_subfolder", 0), 0);
+
+	EXPECT_EQ(rmdir("/mnt/open_tree_subfolder"), 0);
+}
+
 TEST_F(mount_setattr, mount_detached_mount_on_detached_mount_then_close)
 {
 	int fd_tree_base = -EBADF, fd_tree_subdir = -EBADF;
@@ -2097,4 +2088,74 @@ TEST_F(mount_setattr, two_detached_subtrees_of_same_anonymous_mount_namespace)
 	ASSERT_EQ(move_mount(fd_tree1, "", -EBADF, "/tmp/target1", MOVE_MOUNT_F_EMPTY_PATH), 0);
 }
 
+TEST_F(mount_setattr, detached_tree_propagation)
+{
+	int fd_tree = -EBADF;
+	struct statx stx1, stx2, stx3, stx4;
+
+	ASSERT_EQ(unshare(CLONE_NEWNS), 0);
+	ASSERT_EQ(mount(NULL, "/mnt", NULL, MS_REC | MS_SHARED, NULL), 0);
+
+	/*
+	 * Copy the following mount tree:
+	 *
+         * /mnt                   testing tmpfs
+         * |-/mnt/A               testing tmpfs
+         * | `-/mnt/A/AA          testing tmpfs
+         * |   `-/mnt/A/AA/B      testing tmpfs
+         * |     `-/mnt/A/AA/B/BB testing tmpfs
+         * `-/mnt/B               testing ramfs
+	 */
+	fd_tree = sys_open_tree(-EBADF, "/mnt",
+				 AT_NO_AUTOMOUNT | AT_SYMLINK_NOFOLLOW |
+				 AT_RECURSIVE | OPEN_TREE_CLOEXEC |
+				 OPEN_TREE_CLONE);
+	ASSERT_GE(fd_tree, 0);
+
+	ASSERT_EQ(statx(-EBADF, "/mnt/A", 0, 0, &stx1), 0);
+	ASSERT_EQ(statx(fd_tree, "A", 0, 0, &stx2), 0);
+
+	/*
+	 * Copying the mount namespace like done above doesn't alter the
+	 * mounts in any way so the filesystem mounted on /mnt must be
+	 * identical even though the mounts will differ. Use the device
+	 * information to verify that. Note that tmpfs will have a 0
+	 * major number so comparing the major number is misleading.
+	 */
+	ASSERT_EQ(stx1.stx_dev_minor, stx2.stx_dev_minor);
+
+	/* Mount a tmpfs filesystem over /mnt/A. */
+	ASSERT_EQ(mount(NULL, "/mnt/A", "tmpfs", 0, NULL), 0);
+
+
+	ASSERT_EQ(statx(-EBADF, "/mnt/A", 0, 0, &stx3), 0);
+	ASSERT_EQ(statx(fd_tree, "A", 0, 0, &stx4), 0);
+
+	/*
+	 * A new filesystem has been mounted on top of /mnt/A which
+	 * means that the device information will be different for any
+	 * statx() that was taken from /mnt/A before the mount compared
+	 * to one after the mount.
+	 *
+	 * Since we already now that the device information between the
+	 * stx1 and stx2 samples are identical we also now that stx2 and
+	 * stx3 device information will necessarily differ.
+	 */
+	ASSERT_NE(stx1.stx_dev_minor, stx3.stx_dev_minor);
+
+	/*
+	 * If mount propagation worked correctly then the tmpfs mount
+	 * that was created after the mount namespace was unshared will
+	 * have propagated onto /mnt/A in the detached mount tree.
+	 *
+	 * Verify that the device information for stx3 and stx4 are
+	 * identical. It is already established that stx3 is different
+	 * from both stx1 and stx2 sampled before the tmpfs mount was
+	 * done so if stx3 and stx4 are identical the proof is done.
+	 */
+	ASSERT_EQ(stx3.stx_dev_minor, stx4.stx_dev_minor);
+
+	EXPECT_EQ(close(fd_tree), 0);
+}
+
 TEST_HARNESS_MAIN