Merge tag 'vfs-6.16-rc1.mount' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs

Pull vfs mount updates from Christian Brauner:
 "This contains minor mount updates for this cycle:

   - mnt->mnt_devname can never be NULL so simplify the code handling
     that case

   - Add a comment about concurrent changes during statmount() and
     listmount()

   - Update the STATMOUNT_SUPPORTED macro

   - Convert mount flags to an enum"

* tag 'vfs-6.16-rc1.mount' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs:
  statmount: update STATMOUNT_SUPPORTED macro
  fs: convert mount flags to enum
  ->mnt_devname is never NULL
  mount: add a comment about concurrent changes with statmount()/listmount()

Author: torvalds

fs/namespace.c

@@ -355,12 +355,13 @@ static struct mount *alloc_vfsmnt(const char *name)
 		if (err)
 			goto out_free_cache;
 
-		if (name) {
+		if (name)
 			mnt->mnt_devname = kstrdup_const(name,
 							 GFP_KERNEL_ACCOUNT);
-			if (!mnt->mnt_devname)
-				goto out_free_id;
-		}
+		else
+			mnt->mnt_devname = "none";
+		if (!mnt->mnt_devname)
+			goto out_free_id;
 
 #ifdef CONFIG_SMP
 		mnt->mnt_pcp = alloc_percpu(struct mnt_pcp);
@@ -1264,7 +1265,7 @@ struct vfsmount *vfs_create_mount(struct fs_context *fc)
 	if (!fc->root)
 		return ERR_PTR(-EINVAL);
 
-	mnt = alloc_vfsmnt(fc->source ?: "none");
+	mnt = alloc_vfsmnt(fc->source);
 	if (!mnt)
 		return ERR_PTR(-ENOMEM);
 
@@ -5491,7 +5492,7 @@ static int statmount_sb_source(struct kstatmount *s, struct seq_file *seq)
 		seq->buf[seq->count] = '\0';
 		seq->count = start;
 		seq_commit(seq, string_unescape_inplace(seq->buf + start, UNESCAPE_OCTAL));
-	} else if (r->mnt_devname) {
+	} else {
 		seq_puts(seq, r->mnt_devname);
 	}
 	return 0;
@@ -5804,7 +5805,9 @@ static int grab_requested_root(struct mnt_namespace *ns, struct path *root)
 			     STATMOUNT_SB_SOURCE | \
 			     STATMOUNT_OPT_ARRAY | \
 			     STATMOUNT_OPT_SEC_ARRAY | \
-			     STATMOUNT_SUPPORTED_MASK)
+			     STATMOUNT_SUPPORTED_MASK | \
+			     STATMOUNT_MNT_UIDMAP | \
+			     STATMOUNT_MNT_GIDMAP)
 
 static int do_statmount(struct kstatmount *s, u64 mnt_id, u64 mnt_ns_id,
 			struct mnt_namespace *ns)
@@ -5839,13 +5842,29 @@ static int do_statmount(struct kstatmount *s, u64 mnt_id, u64 mnt_ns_id,
 		return err;
 
 	s->root = root;
-	s->idmap = mnt_idmap(s->mnt);
-	if (s->mask & STATMOUNT_SB_BASIC)
-		statmount_sb_basic(s);
 
+	/*
+	 * Note that mount properties in mnt->mnt_flags, mnt->mnt_idmap
+	 * can change concurrently as we only hold the read-side of the
+	 * namespace semaphore and mount properties may change with only
+	 * the mount lock held.
+	 *
+	 * We could sample the mount lock sequence counter to detect
+	 * those changes and retry. But it's not worth it. Worst that
+	 * happens is that the mnt->mnt_idmap pointer is already changed
+	 * while mnt->mnt_flags isn't or vica versa. So what.
+	 *
+	 * Both mnt->mnt_flags and mnt->mnt_idmap are set and retrieved
+	 * via READ_ONCE()/WRITE_ONCE() and guard against theoretical
+	 * torn read/write. That's all we care about right now.
+	 */
+	s->idmap = mnt_idmap(s->mnt);
 	if (s->mask & STATMOUNT_MNT_BASIC)
 		statmount_mnt_basic(s);
 
+	if (s->mask & STATMOUNT_SB_BASIC)
+		statmount_sb_basic(s);
+
 	if (s->mask & STATMOUNT_PROPAGATE_FROM)
 		statmount_propagate_from(s);
 
@@ -6157,6 +6176,10 @@ SYSCALL_DEFINE4(listmount, const struct mnt_id_req __user *, req,
 	    !ns_capable_noaudit(ns->user_ns, CAP_SYS_ADMIN))
 		return -ENOENT;
 
+	/*
+	 * We only need to guard against mount topology changes as
+	 * listmount() doesn't care about any mount properties.
+	 */
 	scoped_guard(rwsem_read, &namespace_sem)
 		ret = do_listmount(ns, kreq.mnt_id, last_mnt_id, kmnt_ids,
 				   nr_mnt_ids, (flags & LISTMOUNT_REVERSE));

fs/proc_namespace.c

@@ -111,7 +111,7 @@ static int show_vfsmnt(struct seq_file *m, struct vfsmount *mnt)
 		if (err)
 			goto out;
 	} else {
-		mangle(m, r->mnt_devname ? r->mnt_devname : "none");
+		mangle(m, r->mnt_devname);
 	}
 	seq_putc(m, ' ');
 	/* mountpoints outside of chroot jail will give SEQ_SKIP on this */
@@ -177,7 +177,7 @@ static int show_mountinfo(struct seq_file *m, struct vfsmount *mnt)
 		if (err)
 			goto out;
 	} else {
-		mangle(m, r->mnt_devname ? r->mnt_devname : "none");
+		mangle(m, r->mnt_devname);
 	}
 	seq_puts(m, sb_rdonly(sb) ? " ro" : " rw");
 	err = show_sb_opts(m, sb);
@@ -199,17 +199,13 @@ static int show_vfsstat(struct seq_file *m, struct vfsmount *mnt)
 	int err;
 
 	/* device */
+	seq_puts(m, "device ");
 	if (sb->s_op->show_devname) {
-		seq_puts(m, "device ");
 		err = sb->s_op->show_devname(m, mnt_path.dentry);
 		if (err)
 			goto out;
 	} else {
-		if (r->mnt_devname) {
-			seq_puts(m, "device ");
-			mangle(m, r->mnt_devname);
-		} else
-			seq_puts(m, "no device");
+		mangle(m, r->mnt_devname);
 	}
 
 	/* mount point */

include/linux/mount.h

@@ -22,48 +22,51 @@ struct fs_context;
 struct file;
 struct path;
 
-#define MNT_NOSUID	0x01
-#define MNT_NODEV	0x02
-#define MNT_NOEXEC	0x04
-#define MNT_NOATIME	0x08
-#define MNT_NODIRATIME	0x10
-#define MNT_RELATIME	0x20
-#define MNT_READONLY	0x40	/* does the user want this to be r/o? */
-#define MNT_NOSYMFOLLOW	0x80
-
-#define MNT_SHRINKABLE	0x100
-#define MNT_WRITE_HOLD	0x200
-
-#define MNT_SHARED	0x1000	/* if the vfsmount is a shared mount */
-#define MNT_UNBINDABLE	0x2000	/* if the vfsmount is a unbindable mount */
-/*
- * MNT_SHARED_MASK is the set of flags that should be cleared when a
- * mount becomes shared.  Currently, this is only the flag that says a
- * mount cannot be bind mounted, since this is how we create a mount
- * that shares events with another mount.  If you add a new MNT_*
- * flag, consider how it interacts with shared mounts.
- */
-#define MNT_SHARED_MASK	(MNT_UNBINDABLE)
-#define MNT_USER_SETTABLE_MASK  (MNT_NOSUID | MNT_NODEV | MNT_NOEXEC \
-				 | MNT_NOATIME | MNT_NODIRATIME | MNT_RELATIME \
-				 | MNT_READONLY | MNT_NOSYMFOLLOW)
-#define MNT_ATIME_MASK (MNT_NOATIME | MNT_NODIRATIME | MNT_RELATIME )
-
-#define MNT_INTERNAL_FLAGS (MNT_SHARED | MNT_WRITE_HOLD | MNT_INTERNAL | \
-			    MNT_DOOMED | MNT_SYNC_UMOUNT | MNT_MARKED)
-
-#define MNT_INTERNAL	0x4000
-
-#define MNT_LOCK_ATIME		0x040000
-#define MNT_LOCK_NOEXEC		0x080000
-#define MNT_LOCK_NOSUID		0x100000
-#define MNT_LOCK_NODEV		0x200000
-#define MNT_LOCK_READONLY	0x400000
-#define MNT_LOCKED		0x800000
-#define MNT_DOOMED		0x1000000
-#define MNT_SYNC_UMOUNT		0x2000000
-#define MNT_MARKED		0x4000000
-#define MNT_UMOUNT		0x8000000
+enum mount_flags {
+	MNT_NOSUID	= 0x01,
+	MNT_NODEV	= 0x02,
+	MNT_NOEXEC	= 0x04,
+	MNT_NOATIME	= 0x08,
+	MNT_NODIRATIME	= 0x10,
+	MNT_RELATIME	= 0x20,
+	MNT_READONLY	= 0x40, /* does the user want this to be r/o? */
+	MNT_NOSYMFOLLOW	= 0x80,
+
+	MNT_SHRINKABLE	= 0x100,
+	MNT_WRITE_HOLD	= 0x200,
+
+	MNT_SHARED	= 0x1000, /* if the vfsmount is a shared mount */
+	MNT_UNBINDABLE	= 0x2000, /* if the vfsmount is a unbindable mount */
+
+	MNT_INTERNAL	= 0x4000,
+
+	MNT_LOCK_ATIME		= 0x040000,
+	MNT_LOCK_NOEXEC		= 0x080000,
+	MNT_LOCK_NOSUID		= 0x100000,
+	MNT_LOCK_NODEV		= 0x200000,
+	MNT_LOCK_READONLY	= 0x400000,
+	MNT_LOCKED		= 0x800000,
+	MNT_DOOMED		= 0x1000000,
+	MNT_SYNC_UMOUNT		= 0x2000000,
+	MNT_MARKED		= 0x4000000,
+	MNT_UMOUNT		= 0x8000000,
+
+	/*
+	 * MNT_SHARED_MASK is the set of flags that should be cleared when a
+	 * mount becomes shared.  Currently, this is only the flag that says a
+	 * mount cannot be bind mounted, since this is how we create a mount
+	 * that shares events with another mount.  If you add a new MNT_*
+	 * flag, consider how it interacts with shared mounts.
+	 */
+	MNT_SHARED_MASK	= MNT_UNBINDABLE,
+	MNT_USER_SETTABLE_MASK  = MNT_NOSUID | MNT_NODEV | MNT_NOEXEC
+				  | MNT_NOATIME | MNT_NODIRATIME | MNT_RELATIME
+				  | MNT_READONLY | MNT_NOSYMFOLLOW,
+	MNT_ATIME_MASK = MNT_NOATIME | MNT_NODIRATIME | MNT_RELATIME,
+
+	MNT_INTERNAL_FLAGS = MNT_SHARED | MNT_WRITE_HOLD | MNT_INTERNAL |
+			     MNT_DOOMED | MNT_SYNC_UMOUNT | MNT_MARKED,
+};
 
 struct vfsmount {
 	struct dentry *mnt_root;	/* root of the mounted tree */