From: Miklos Szeredi mszeredi@redhat.com

mainline inclusion from mainline-v5.8-rc1 commit 9f6c61f96f2d97cbb5f7fa85607bc398f843ff0f category: bugfix bugzilla: 35619 CVE: NA

--------------------------------

If mounts are deleted after a read(2) call on /proc/self/mounts (or its kin), the subsequent read(2) could miss a mount that comes after the deleted one in the list. This is because the file position is interpreted as the number mount entries from the start of the list.

E.g. first read gets entries #0 to #9; the seq file index will be 10. Then entry #5 is deleted, resulting in #10 becoming #9 and #11 becoming #10, etc... The next read will continue from entry #10, and #9 is missed.

Solve this by adding a cursor entry for each open instance. Taking the global namespace_sem for write seems excessive, since we are only dealing with a per-namespace list. Instead add a per-namespace spinlock and use that together with namespace_sem taken for read to protect against concurrent modification of the mount list. This may reduce parallelism of is_local_mountpoint(), but it's hardly a big contention point. We could also use RCU freeing of cursors to make traversal not need additional locks, if that turns out to be neceesary.

Only move the cursor once for each read (cursor is not added on open) to minimize cacheline invalidation. When EOF is reached, the cursor is taken off the list, in order to prevent an excessive number of cursors due to inactive open file descriptors.

Reported-by: Karel Zak kzak@redhat.com Signed-off-by: Miklos Szeredi mszeredi@redhat.com

Conflicts: fs/mount.h fs/namespace.c

Signed-off-by: Zhang Xiaoxu zhangxiaoxu5@huawei.com Reviewed-by: zhangyi (F) yi.zhang@huawei.com Signed-off-by: Yang Yingliang yangyingliang@huawei.com --- fs/mount.h | 12 ++++-- fs/namespace.c | 91 +++++++++++++++++++++++++++++++++++-------- fs/proc_namespace.c | 4 +- include/linux/mount.h | 4 +- 4 files changed, 90 insertions(+), 21 deletions(-)

diff --git a/fs/mount.h b/fs/mount.h index f39bc9da4d73..b8318db51ea1 100644 --- a/fs/mount.h +++ b/fs/mount.h @@ -9,7 +9,13 @@ struct mnt_namespace { atomic_t count; struct ns_common ns; struct mount * root; + /* + * Traversal and modification of .list is protected by either + * - taking namespace_sem for write, OR + * - taking namespace_sem for read AND taking .ns_lock. + */ struct list_head list; + spinlock_t ns_lock; struct user_namespace *user_ns; struct ucounts *ucounts; u64 seq; /* Sequence number to prevent loops */ @@ -131,9 +137,7 @@ struct proc_mounts { struct mnt_namespace *ns; struct path root; int (*show)(struct seq_file *, struct vfsmount *); - void *cached_mount; - u64 cached_event; - loff_t cached_index; + struct mount cursor; };

extern const struct seq_operations mounts_op; @@ -146,3 +150,5 @@ static inline bool is_local_mountpoint(struct dentry *dentry)

return __is_local_mountpoint(dentry); } + +extern void mnt_cursor_del(struct mnt_namespace *ns, struct mount *cursor); diff --git a/fs/namespace.c b/fs/namespace.c index da90b9c878c1..c582abcdab59 100644 --- a/fs/namespace.c +++ b/fs/namespace.c @@ -653,6 +653,21 @@ struct vfsmount *lookup_mnt(const struct path *path) return m; }

+static inline void lock_ns_list(struct mnt_namespace *ns) +{ + spin_lock(&ns->ns_lock); +} + +static inline void unlock_ns_list(struct mnt_namespace *ns) +{ + spin_unlock(&ns->ns_lock); +} + +static inline bool mnt_is_cursor(struct mount *mnt) +{ + return mnt->mnt.mnt_flags & MNT_CURSOR; +} + /* * __is_local_mountpoint - Test to see if dentry is a mountpoint in the * current mount namespace. @@ -678,11 +693,15 @@ bool __is_local_mountpoint(struct dentry *dentry) goto out;

down_read(&namespace_sem); + lock_ns_list(ns); list_for_each_entry(mnt, &ns->list, mnt_list) { + if (mnt_is_cursor(mnt)) + continue; is_covered = (mnt->mnt_mountpoint == dentry); if (is_covered) break; } + unlock_ns_list(ns); up_read(&namespace_sem); out: return is_covered; @@ -1237,46 +1256,71 @@ struct vfsmount *mnt_clone_internal(const struct path *path) }

#ifdef CONFIG_PROC_FS +static struct mount *mnt_list_next(struct mnt_namespace *ns, + struct list_head *p) +{ + struct mount *mnt, *ret = NULL; + + lock_ns_list(ns); + list_for_each_continue(p, &ns->list) { + mnt = list_entry(p, typeof(*mnt), mnt_list); + if (!mnt_is_cursor(mnt)) { + ret = mnt; + break; + } + } + unlock_ns_list(ns); + + return ret; +} + /* iterator; we want it to have access to namespace_sem, thus here... */ static void *m_start(struct seq_file *m, loff_t *pos) { struct proc_mounts *p = m->private; + struct list_head *prev;

down_read(&namespace_sem); - if (p->cached_event == p->ns->event) { - void *v = p->cached_mount; - if (*pos == p->cached_index) - return v; - if (*pos == p->cached_index + 1) { - v = seq_list_next(v, &p->ns->list, &p->cached_index); - return p->cached_mount = v; - } + if (!*pos) { + prev = &p->ns->list; + } else { + prev = &p->cursor.mnt_list; + + /* Read after we'd reached the end? */ + if (list_empty(prev)) + return NULL; }

- p->cached_event = p->ns->event; - p->cached_mount = seq_list_start(&p->ns->list, *pos); - p->cached_index = *pos; - return p->cached_mount; + return mnt_list_next(p->ns, prev); }

static void *m_next(struct seq_file *m, void *v, loff_t *pos) { struct proc_mounts *p = m->private; + struct mount *mnt = v;

- p->cached_mount = seq_list_next(v, &p->ns->list, pos); - p->cached_index = *pos; - return p->cached_mount; + ++*pos; + return mnt_list_next(p->ns, &mnt->mnt_list); }

static void m_stop(struct seq_file *m, void *v) { + struct proc_mounts *p = m->private; + struct mount *mnt = v; + + lock_ns_list(p->ns); + if (mnt) + list_move_tail(&p->cursor.mnt_list, &mnt->mnt_list); + else + list_del_init(&p->cursor.mnt_list); + unlock_ns_list(p->ns); up_read(&namespace_sem); }

static int m_show(struct seq_file *m, void *v) { struct proc_mounts *p = m->private; - struct mount *r = list_entry(v, struct mount, mnt_list); + struct mount *r = v; return p->show(m, &r->mnt); }

@@ -1286,6 +1330,15 @@ const struct seq_operations mounts_op = { .stop = m_stop, .show = m_show, }; + +void mnt_cursor_del(struct mnt_namespace *ns, struct mount *cursor) +{ + down_read(&namespace_sem); + lock_ns_list(ns); + list_del(&cursor->mnt_list); + unlock_ns_list(ns); + up_read(&namespace_sem); +} #endif /* CONFIG_PROC_FS */

/** @@ -2858,6 +2911,7 @@ static struct mnt_namespace *alloc_mnt_ns(struct user_namespace *user_ns) INIT_LIST_HEAD(&new_ns->list); init_waitqueue_head(&new_ns->poll); new_ns->event = 0; + spin_lock_init(&new_ns->ns_lock); new_ns->user_ns = get_user_ns(user_ns); new_ns->ucounts = ucounts; new_ns->mounts = 0; @@ -3312,10 +3366,14 @@ static bool mnt_already_visible(struct mnt_namespace *ns, struct vfsmount *new, bool visible = false;

down_read(&namespace_sem); + lock_ns_list(ns); list_for_each_entry(mnt, &ns->list, mnt_list) { struct mount *child; int mnt_flags;

+ if (mnt_is_cursor(mnt)) + continue; + if (mnt->mnt.mnt_sb->s_type != new->mnt_sb->s_type) continue;

@@ -3363,6 +3421,7 @@ static bool mnt_already_visible(struct mnt_namespace *ns, struct vfsmount *new, next: ; } found: + unlock_ns_list(ns); up_read(&namespace_sem); return visible; } diff --git a/fs/proc_namespace.c b/fs/proc_namespace.c index e16fb8f2049e..969f9c8fbdc0 100644 --- a/fs/proc_namespace.c +++ b/fs/proc_namespace.c @@ -279,7 +279,8 @@ static int mounts_open_common(struct inode *inode, struct file *file, p->ns = ns; p->root = root; p->show = show; - p->cached_event = ~0ULL; + INIT_LIST_HEAD(&p->cursor.mnt_list); + p->cursor.mnt.mnt_flags = MNT_CURSOR;

return 0;

@@ -296,6 +297,7 @@ static int mounts_release(struct inode *inode, struct file *file) struct seq_file *m = file->private_data; struct proc_mounts *p = m->private; path_put(&p->root); + mnt_cursor_del(p->ns, &p->cursor); put_mnt_ns(p->ns); return seq_release_private(inode, file); } diff --git a/include/linux/mount.h b/include/linux/mount.h index 4b0db4418954..46a77d791870 100644 --- a/include/linux/mount.h +++ b/include/linux/mount.h @@ -49,7 +49,8 @@ struct mnt_namespace; #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) + MNT_DOOMED | MNT_SYNC_UMOUNT | MNT_MARKED | \ + MNT_CURSOR)

#define MNT_INTERNAL 0x4000

@@ -63,6 +64,7 @@ struct mnt_namespace; #define MNT_SYNC_UMOUNT 0x2000000 #define MNT_MARKED 0x4000000 #define MNT_UMOUNT 0x8000000 +#define MNT_CURSOR 0x10000000

struct vfsmount { struct dentry *mnt_root; /* root of the mounted tree */