From e1ff7aa34dec7e650159fd7ca8ec6af7cc428d9f Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Tue, 18 Mar 2025 12:29:21 -0400 Subject: [PATCH] umount: Allow superblock owners to force umount Loosen the permission check on forced umount to allow users holding CAP_SYS_ADMIN privileges in namespaces that are privileged with respect to the userns that originally mounted the filesystem. Signed-off-by: Trond Myklebust Link: https://lore.kernel.org/r/12f212d4ef983714d065a6bb372fbb378753bf4c.1742315194.git.trond.myklebust@hammerspace.com Acked-by: "Eric W. Biederman" Signed-off-by: Christian Brauner --- fs/namespace.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/fs/namespace.c b/fs/namespace.c index 01fb1074c4c9..57c14af6092a 100644 --- a/fs/namespace.c +++ b/fs/namespace.c @@ -2105,6 +2105,7 @@ static void warn_mandlock(void) static int can_umount(const struct path *path, int flags) { struct mount *mnt = real_mount(path->mnt); + struct super_block *sb = path->dentry->d_sb; if (!may_mount()) return -EPERM; @@ -2114,7 +2115,7 @@ static int can_umount(const struct path *path, int flags) return -EINVAL; if (mnt->mnt.mnt_flags & MNT_LOCKED) /* Check optimistically */ return -EINVAL; - if (flags & MNT_FORCE && !capable(CAP_SYS_ADMIN)) + if (flags & MNT_FORCE && !ns_capable(sb->s_user_ns, CAP_SYS_ADMIN)) return -EPERM; return 0; }