Validate extensible ioctls stricter than we do now.

Signed-off-by: Christian Brauner brauner@kernel.org --- fs/nsfs.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/fs/nsfs.c b/fs/nsfs.c index 59aa801347a7..34f0b35d3ead 100644 --- a/fs/nsfs.c +++ b/fs/nsfs.c @@ -169,9 +169,11 @@ static bool nsfs_ioctl_valid(unsigned int cmd) /* Extensible ioctls require some extra handling. */ switch (_IOC_NR(cmd)) { case _IOC_NR(NS_MNT_GET_INFO): + return extensible_ioctl_valid(cmd, NS_MNT_GET_INFO, MNT_NS_INFO_SIZE_VER0); case _IOC_NR(NS_MNT_GET_NEXT): + return extensible_ioctl_valid(cmd, NS_MNT_GET_NEXT, MNT_NS_INFO_SIZE_VER0); case _IOC_NR(NS_MNT_GET_PREV): - return (_IOC_TYPE(cmd) == _IOC_TYPE(cmd)); + return extensible_ioctl_valid(cmd, NS_MNT_GET_PREV, MNT_NS_INFO_SIZE_VER0); }

return false;

Use the new extensible_ioctl_valid() helper which is equivalent to what is done here.

Signed-off-by: Christian Brauner brauner@kernel.org --- block/blk-integrity.c | 8 +++----- 1 file changed, 3 insertions(+), 5 deletions(-)

diff --git a/block/blk-integrity.c b/block/blk-integrity.c index 056b8948369d..609d75d6a39b 100644 --- a/block/blk-integrity.c +++ b/block/blk-integrity.c @@ -58,16 +58,14 @@ int blk_rq_count_integrity_sg(struct request_queue *q, struct bio *bio) int blk_get_meta_cap(struct block_device *bdev, unsigned int cmd, struct logical_block_metadata_cap __user *argp) { - struct blk_integrity *bi = blk_get_integrity(bdev->bd_disk); + struct blk_integrity *bi; struct logical_block_metadata_cap meta_cap = {}; size_t usize = _IOC_SIZE(cmd);

- if (_IOC_DIR(cmd) != _IOC_DIR(FS_IOC_GETLBMD_CAP) || - _IOC_TYPE(cmd) != _IOC_TYPE(FS_IOC_GETLBMD_CAP) || - _IOC_NR(cmd) != _IOC_NR(FS_IOC_GETLBMD_CAP) || - _IOC_SIZE(cmd) < LBMD_SIZE_VER0) + if (extensible_ioctl_valid(cmd, FS_IOC_GETLBMD_CAP, LBMD_SIZE_VER0)) return -ENOIOCTLCMD;

+ bi = blk_get_integrity(bdev->bd_disk); if (!bi) goto out;

Don't cargo-cult the same thing over and over.

Signed-off-by: Christian Brauner brauner@kernel.org --- kernel/utsname.c | 16 ++-------------- 1 file changed, 2 insertions(+), 14 deletions(-)

diff --git a/kernel/utsname.c b/kernel/utsname.c index b1ac3ca870f2..02037010b378 100644 --- a/kernel/utsname.c +++ b/kernel/utsname.c @@ -27,16 +27,6 @@ static void dec_uts_namespaces(struct ucounts *ucounts) dec_ucount(ucounts, UCOUNT_UTS_NAMESPACES); }

-static struct uts_namespace *create_uts_ns(void) -{ - struct uts_namespace *uts_ns; - - uts_ns = kmem_cache_alloc(uts_ns_cache, GFP_KERNEL); - if (uts_ns) - refcount_set(&uts_ns->ns.count, 1); - return uts_ns; -} - /* * Clone a new ns copying an original utsname, setting refcount to 1 * @old_ns: namespace to clone @@ -55,17 +45,15 @@ static struct uts_namespace *clone_uts_ns(struct user_namespace *user_ns, goto fail;

err = -ENOMEM; - ns = create_uts_ns(); + ns = kmem_cache_zalloc(uts_ns_cache, GFP_KERNEL); if (!ns) goto fail_dec;

- err = ns_alloc_inum(&ns->ns); + err = ns_common_init(&ns->ns, &utsns_operations, true); if (err) goto fail_free;

ns->ucounts = ucounts; - ns->ns.ops = &utsns_operations; - down_read(&uts_sem); memcpy(&ns->name, &old_ns->name, sizeof(ns->name)); ns->user_ns = get_user_ns(user_ns);

Don't cargo-cult the same thing over and over.

Signed-off-by: Christian Brauner brauner@kernel.org --- kernel/user_namespace.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-)

diff --git a/kernel/user_namespace.c b/kernel/user_namespace.c index 682f40d5632d..98f4fe84d039 100644 --- a/kernel/user_namespace.c +++ b/kernel/user_namespace.c @@ -124,12 +124,11 @@ int create_user_ns(struct cred *new) goto fail_dec;

ns->parent_could_setfcap = cap_raised(new->cap_effective, CAP_SETFCAP); - ret = ns_alloc_inum(&ns->ns); + + ret = ns_common_init(&ns->ns, &userns_operations, true); if (ret) goto fail_free; - ns->ns.ops = &userns_operations;

- refcount_set(&ns->ns.count, 1); /* Leave the new->user_ns reference with the new user namespace. */ ns->parent = parent_ns; ns->level = parent_ns->level + 1;

Don't cargo-cult the same thing over and over.

Signed-off-by: Christian Brauner brauner@kernel.org --- net/core/net_namespace.c | 46 ++++++++++++++++++++++++++++++++-------------- 1 file changed, 32 insertions(+), 14 deletions(-)

diff --git a/net/core/net_namespace.c b/net/core/net_namespace.c index 1b6f3826dd0e..dafb3d947043 100644 --- a/net/core/net_namespace.c +++ b/net/core/net_namespace.c @@ -397,10 +397,22 @@ static __net_init void preinit_net_sysctl(struct net *net) }

/* init code that must occur even if setup_net() is not called. */ -static __net_init void preinit_net(struct net *net, struct user_namespace *user_ns) +static __net_init int preinit_net(struct net *net, struct user_namespace *user_ns) { + const struct proc_ns_operations *ns_ops; + int ret; + +#ifdef CONFIG_NET_NS + ns_ops = &netns_operations; +#else + ns_ops = NULL; +#endif + + ret = ns_common_init(&net->ns, ns_ops, false); + if (ret) + return ret; + refcount_set(&net->passive, 1); - refcount_set(&net->ns.count, 1); ref_tracker_dir_init(&net->refcnt_tracker, 128, "net_refcnt"); ref_tracker_dir_init(&net->notrefcnt_tracker, 128, "net_notrefcnt");

@@ -420,6 +432,7 @@ static __net_init void preinit_net(struct net *net, struct user_namespace *user_ INIT_LIST_HEAD(&net->ptype_all); INIT_LIST_HEAD(&net->ptype_specific); preinit_net_sysctl(net); + return 0; }

/* @@ -559,7 +572,9 @@ struct net *copy_net_ns(unsigned long flags, goto dec_ucounts; }

- preinit_net(net, user_ns); + rv = preinit_net(net, user_ns); + if (rv < 0) + goto dec_ucounts; net->ucounts = ucounts; get_user_ns(user_ns);

@@ -573,6 +588,7 @@ struct net *copy_net_ns(unsigned long flags,

if (rv < 0) { put_userns: + ns_free_inum(&net->ns); #ifdef CONFIG_KEYS key_remove_domain(net->key_domain); #endif @@ -812,17 +828,14 @@ static void net_ns_net_debugfs(struct net *net)

static __net_init int net_ns_net_init(struct net *net) { -#ifdef CONFIG_NET_NS - net->ns.ops = &netns_operations; -#endif - net->ns.inum = PROC_NET_INIT_INO; - if (net != &init_net) { - int ret = ns_alloc_inum(&net->ns); - if (ret) - return ret; - } + int ret = 0; + + if (net == &init_net) + net->ns.inum = PROC_NET_INIT_INO; + else + ret = proc_alloc_inum(&to_ns_common(net)->inum); net_ns_net_debugfs(net); - return 0; + return ret; }

static __net_exit void net_ns_net_exit(struct net *net) @@ -1282,7 +1295,12 @@ void __init net_ns_init(void) #ifdef CONFIG_KEYS init_net.key_domain = &init_net_key_domain; #endif - preinit_net(&init_net, &init_user_ns); + /* + * This currently cannot fail as the initial network namespace + * has a static inode number. + */ + if (preinit_net(&init_net, &init_user_ns)) + panic("Could not preinitialize the initial network namespace");

down_write(&pernet_ops_rwsem); if (setup_net(&init_net))

Move the mount namespace to the generic iterator. This allows us to drop a bunch of members from struct mnt_namespace. t Signed-off-by: Christian Brauner brauner@kernel.org --- fs/mount.h | 10 +--- fs/namespace.c | 141 +++++++++++++-------------------------------------------- fs/nsfs.c | 4 +- 3 files changed, 35 insertions(+), 120 deletions(-)

diff --git a/fs/mount.h b/fs/mount.h index 97737051a8b9..76bf863c9ae2 100644 --- a/fs/mount.h +++ b/fs/mount.h @@ -17,11 +17,7 @@ struct mnt_namespace { }; struct user_namespace *user_ns; struct ucounts *ucounts; - u64 seq; /* Sequence number to prevent loops */ - union { - wait_queue_head_t poll; - struct rcu_head mnt_ns_rcu; - }; + wait_queue_head_t poll; u64 seq_origin; /* Sequence number of origin mount namespace */ u64 event; #ifdef CONFIG_FSNOTIFY @@ -30,8 +26,6 @@ struct mnt_namespace { #endif unsigned int nr_mounts; /* # of mounts in the namespace */ unsigned int pending_mounts; - 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 */ } __randomize_layout;

@@ -173,7 +167,7 @@ static inline bool is_local_mountpoint(const struct dentry *dentry)

static inline bool is_anon_ns(struct mnt_namespace *ns) { - return ns->seq == 0; + return ns->ns.ns_id == 0; }

static inline bool anon_ns_root(const struct mount *m) diff --git a/fs/namespace.c b/fs/namespace.c index 14c5cdbdd6e1..40a8d75f6b16 100644 --- a/fs/namespace.c +++ b/fs/namespace.c @@ -33,6 +33,7 @@ #include <linux/shmem_fs.h> #include <linux/mnt_idmapping.h> #include <linux/pidfs.h> +#include <linux/nstree.h>

#include "pnode.h" #include "internal.h" @@ -80,13 +81,10 @@ static DECLARE_RWSEM(namespace_sem); static HLIST_HEAD(unmounted); /* protected by namespace_sem */ static LIST_HEAD(ex_mountpoints); /* protected by namespace_sem */ static struct mnt_namespace *emptied_ns; /* protected by namespace_sem */ -static DEFINE_SEQLOCK(mnt_ns_tree_lock);

#ifdef CONFIG_FSNOTIFY LIST_HEAD(notify_list); /* protected by namespace_sem */ #endif -static struct rb_root mnt_ns_tree = RB_ROOT; /* protected by mnt_ns_tree_lock */ -static LIST_HEAD(mnt_ns_list); /* protected by mnt_ns_tree_lock */

enum mount_kattr_flags_t { MOUNT_KATTR_RECURSE = (1 << 0), @@ -119,53 +117,12 @@ __cacheline_aligned_in_smp DEFINE_SEQLOCK(mount_lock);

static inline struct mnt_namespace *node_to_mnt_ns(const struct rb_node *node) { + struct ns_common *ns; + if (!node) return NULL; - return rb_entry(node, struct mnt_namespace, mnt_ns_tree_node); -} - -static int mnt_ns_cmp(struct rb_node *a, const struct rb_node *b) -{ - struct mnt_namespace *ns_a = node_to_mnt_ns(a); - struct mnt_namespace *ns_b = node_to_mnt_ns(b); - u64 seq_a = ns_a->seq; - u64 seq_b = ns_b->seq; - - if (seq_a < seq_b) - return -1; - if (seq_a > seq_b) - return 1; - return 0; -} - -static inline void mnt_ns_tree_write_lock(void) -{ - write_seqlock(&mnt_ns_tree_lock); -} - -static inline void mnt_ns_tree_write_unlock(void) -{ - write_sequnlock(&mnt_ns_tree_lock); -} - -static void mnt_ns_tree_add(struct mnt_namespace *ns) -{ - struct rb_node *node, *prev; - - mnt_ns_tree_write_lock(); - node = rb_find_add_rcu(&ns->mnt_ns_tree_node, &mnt_ns_tree, mnt_ns_cmp); - /* - * If there's no previous entry simply add it after the - * head and if there is add it after the previous entry. - */ - prev = rb_prev(&ns->mnt_ns_tree_node); - if (!prev) - list_add_rcu(&ns->mnt_ns_list, &mnt_ns_list); - else - list_add_rcu(&ns->mnt_ns_list, &node_to_mnt_ns(prev)->mnt_ns_list); - mnt_ns_tree_write_unlock(); - - WARN_ON_ONCE(node); + ns = rb_entry(node, struct ns_common, ns_tree_node); + return container_of(ns, struct mnt_namespace, ns); }

static void mnt_ns_release(struct mnt_namespace *ns) @@ -181,32 +138,16 @@ DEFINE_FREE(mnt_ns_release, struct mnt_namespace *, if (_T) mnt_ns_release(_T))

static void mnt_ns_release_rcu(struct rcu_head *rcu) { - mnt_ns_release(container_of(rcu, struct mnt_namespace, mnt_ns_rcu)); + mnt_ns_release(container_of(rcu, struct mnt_namespace, ns.ns_rcu)); }

static void mnt_ns_tree_remove(struct mnt_namespace *ns) { /* remove from global mount namespace list */ - if (!is_anon_ns(ns)) { - mnt_ns_tree_write_lock(); - rb_erase(&ns->mnt_ns_tree_node, &mnt_ns_tree); - list_bidir_del_rcu(&ns->mnt_ns_list); - mnt_ns_tree_write_unlock(); - } - - call_rcu(&ns->mnt_ns_rcu, mnt_ns_release_rcu); -} - -static int mnt_ns_find(const void *key, const struct rb_node *node) -{ - const u64 mnt_ns_id = *(u64 *)key; - const struct mnt_namespace *ns = node_to_mnt_ns(node); + if (!is_anon_ns(ns)) + ns_tree_remove(ns);

- if (mnt_ns_id < ns->seq) - return -1; - if (mnt_ns_id > ns->seq) - return 1; - return 0; + call_rcu(&ns->ns.ns_rcu, mnt_ns_release_rcu); }

/* @@ -225,28 +166,21 @@ static int mnt_ns_find(const void *key, const struct rb_node *node) */ static struct mnt_namespace *lookup_mnt_ns(u64 mnt_ns_id) { - struct mnt_namespace *ns; - struct rb_node *node; - unsigned int seq; + struct mnt_namespace *mnt_ns; + struct ns_common *ns;

guard(rcu)(); - do { - seq = read_seqbegin(&mnt_ns_tree_lock); - node = rb_find_rcu(&mnt_ns_id, &mnt_ns_tree, mnt_ns_find); - if (node) - break; - } while (read_seqretry(&mnt_ns_tree_lock, seq)); - - if (!node) + ns = ns_tree_lookup_rcu(mnt_ns_id, CLONE_NEWNS); + if (!ns) return NULL;

/* * The last reference count is put with RCU delay so we can * unconditonally acquire a reference here. */ - ns = node_to_mnt_ns(node); - refcount_inc(&ns->passive); - return ns; + mnt_ns = container_of(ns, struct mnt_namespace, ns); + refcount_inc(&mnt_ns->passive); + return mnt_ns; }

static inline void lock_mount_hash(void) @@ -1017,7 +951,7 @@ static inline bool check_anonymous_mnt(struct mount *mnt) return false;

seq = mnt->mnt_ns->seq_origin; - return !seq || (seq == current->nsproxy->mnt_ns->seq); + return !seq || (seq == current->nsproxy->mnt_ns->ns.ns_id); }

/* @@ -2155,19 +2089,16 @@ struct ns_common *from_mnt_ns(struct mnt_namespace *mnt)

struct mnt_namespace *get_sequential_mnt_ns(struct mnt_namespace *mntns, bool previous) { + struct ns_common *ns; + guard(rcu)();

for (;;) { - struct list_head *list; - - if (previous) - list = rcu_dereference(list_bidir_prev_rcu(&mntns->mnt_ns_list)); - else - list = rcu_dereference(list_next_rcu(&mntns->mnt_ns_list)); - if (list_is_head(list, &mnt_ns_list)) - return ERR_PTR(-ENOENT); + ns = ns_tree_adjoined_rcu(mntns, previous); + if (IS_ERR(ns)) + return ERR_CAST(ns);

- mntns = list_entry_rcu(list, struct mnt_namespace, mnt_ns_list); + mntns = to_mnt_ns(ns);

/* * The last passive reference count is put with RCU @@ -2207,7 +2138,7 @@ static bool mnt_ns_loop(struct dentry *dentry) if (!mnt_ns) return false;

- return current->nsproxy->mnt_ns->seq >= mnt_ns->seq; + return current->nsproxy->mnt_ns->ns.ns_id >= mnt_ns->ns.ns_id; }

struct mount *copy_tree(struct mount *src_root, struct dentry *dentry, @@ -3070,7 +3001,7 @@ static struct file *open_detached_copy(struct path *path, bool recursive) if (is_anon_ns(src_mnt_ns)) ns->seq_origin = src_mnt_ns->seq_origin; else - ns->seq_origin = src_mnt_ns->seq; + ns->seq_origin = src_mnt_ns->ns.ns_id; }

mnt = __do_loopback(path, recursive); @@ -4153,15 +4084,6 @@ static void free_mnt_ns(struct mnt_namespace *ns) mnt_ns_tree_remove(ns); }

-/* - * Assign a sequence number so we can detect when we attempt to bind - * mount a reference to an older mount namespace into the current - * mount namespace, preventing reference counting loops. A 64bit - * number incrementing at 10Ghz will take 12,427 years to wrap which - * is effectively never, so we can ignore the possibility. - */ -static atomic64_t mnt_ns_seq = ATOMIC64_INIT(1); - static struct mnt_namespace *alloc_mnt_ns(struct user_namespace *user_ns, bool anon) { struct mnt_namespace *new_ns; @@ -4185,11 +4107,11 @@ static struct mnt_namespace *alloc_mnt_ns(struct user_namespace *user_ns, bool a return ERR_PTR(ret); } if (!anon) - new_ns->seq = atomic64_inc_return(&mnt_ns_seq); + ns_tree_gen_id(&new_ns->ns); + RB_CLEAR_NODE(&new_ns->ns.ns_tree_node); + INIT_LIST_HEAD(&new_ns->ns.ns_list_node); refcount_set(&new_ns->passive, 1); new_ns->mounts = RB_ROOT; - INIT_LIST_HEAD(&new_ns->mnt_ns_list); - RB_CLEAR_NODE(&new_ns->mnt_ns_tree_node); init_waitqueue_head(&new_ns->poll); new_ns->user_ns = get_user_ns(user_ns); new_ns->ucounts = ucounts; @@ -4275,7 +4197,7 @@ struct mnt_namespace *copy_mnt_ns(unsigned long flags, struct mnt_namespace *ns, if (pwdmnt) mntput(pwdmnt);

- mnt_ns_tree_add(new_ns); + ns_tree_add_raw(new_ns); return new_ns; }

@@ -5385,7 +5307,7 @@ static int statmount_sb_source(struct kstatmount *s, struct seq_file *seq) static void statmount_mnt_ns_id(struct kstatmount *s, struct mnt_namespace *ns) { s->sm.mask |= STATMOUNT_MNT_NS_ID; - s->sm.mnt_ns_id = ns->seq; + s->sm.mnt_ns_id = ns->ns.ns_id; }

static int statmount_mnt_opts(struct kstatmount *s, struct seq_file *seq) @@ -6090,7 +6012,6 @@ static void __init init_mount_tree(void) ns = alloc_mnt_ns(&init_user_ns, true); if (IS_ERR(ns)) panic("Can't allocate initial namespace"); - ns->seq = atomic64_inc_return(&mnt_ns_seq); ns->ns.inum = PROC_MNT_INIT_INO; m = real_mount(mnt); ns->root = m; @@ -6105,7 +6026,7 @@ static void __init init_mount_tree(void) set_fs_pwd(current->fs, &root); set_fs_root(current->fs, &root);

- mnt_ns_tree_add(ns); + ns_tree_add(ns); }

void __init mnt_init(void) diff --git a/fs/nsfs.c b/fs/nsfs.c index 34f0b35d3ead..6f8008177133 100644 --- a/fs/nsfs.c +++ b/fs/nsfs.c @@ -139,7 +139,7 @@ static int copy_ns_info_to_user(const struct mnt_namespace *mnt_ns, * the size value will be set to the size the kernel knows about. */ kinfo->size = min(usize, sizeof(*kinfo)); - kinfo->mnt_ns_id = mnt_ns->seq; + kinfo->mnt_ns_id = mnt_ns->ns.ns_id; kinfo->nr_mounts = READ_ONCE(mnt_ns->nr_mounts); /* Subtract the root mount of the mount namespace. */ if (kinfo->nr_mounts) @@ -221,7 +221,7 @@ static long ns_ioctl(struct file *filp, unsigned int ioctl,

mnt_ns = container_of(ns, struct mnt_namespace, ns); idp = (__u64 __user *)arg; - id = mnt_ns->seq; + id = mnt_ns->ns.ns_id; return put_user(id, idp); } case NS_GET_PID_FROM_PIDNS:

Support the generic namespace iterator and lookup infrastructure to support file handles for namespaces.

The network namespace has a separate list with different lifetime rules which we can just leave in tact. We have a similar concept for mount namespaces as well where it is on two differenet lists for different purposes.

Signed-off-by: Christian Brauner brauner@kernel.org --- include/net/net_namespace.h | 1 + net/core/net_namespace.c | 8 ++++++-- 2 files changed, 7 insertions(+), 2 deletions(-)

diff --git a/include/net/net_namespace.h b/include/net/net_namespace.h index 025a7574b275..42075748dff1 100644 --- a/include/net/net_namespace.h +++ b/include/net/net_namespace.h @@ -11,6 +11,7 @@ #include <linux/list.h> #include <linux/sysctl.h> #include <linux/uidgid.h> +#include <linux/nstree.h>

#include <net/flow.h> #include <net/netns/core.h> diff --git a/net/core/net_namespace.c b/net/core/net_namespace.c index dafb3d947043..b85e303400be 100644 --- a/net/core/net_namespace.c +++ b/net/core/net_namespace.c @@ -20,6 +20,7 @@ #include <linux/sched/task.h> #include <linux/uidgid.h> #include <linux/proc_fs.h> +#include <linux/nstree.h>

#include <net/aligned_data.h> #include <net/sock.h> @@ -445,7 +446,7 @@ static __net_init int setup_net(struct net *net) LIST_HEAD(net_exit_list); int error = 0;

- net->net_cookie = atomic64_inc_return(&net_aligned_data.net_cookie); + net->net_cookie = ns_tree_gen_id(&net->ns);

list_for_each_entry(ops, &pernet_list, list) { error = ops_init(ops, net); @@ -455,6 +456,7 @@ static __net_init int setup_net(struct net *net) down_write(&net_rwsem); list_add_tail_rcu(&net->list, &net_namespace_list); up_write(&net_rwsem); + ns_tree_add_raw(net); out: return error;

@@ -675,8 +677,10 @@ static void cleanup_net(struct work_struct *work)

/* Don't let anyone else find us. */ down_write(&net_rwsem); - llist_for_each_entry(net, net_kill_list, cleanup_list) + llist_for_each_entry(net, net_kill_list, cleanup_list) { + ns_tree_remove(net); list_del_rcu(&net->list); + } /* Cache last net. After we unlock rtnl, no one new net * added to net_namespace_list can assign nsid pointer * to a net from net_kill_list (see peernet2id_alloc()).

Support the generic namespace iterator and lookup infrastructure to support file handles for namespaces.

Signed-off-by: Christian Brauner brauner@kernel.org --- include/linux/time_namespace.h | 5 +++++ init/main.c | 2 ++ kernel/time/namespace.c | 13 +++++++++++-- 3 files changed, 18 insertions(+), 2 deletions(-)

diff --git a/include/linux/time_namespace.h b/include/linux/time_namespace.h index bb2c52f4fc94..7f6af7a9771e 100644 --- a/include/linux/time_namespace.h +++ b/include/linux/time_namespace.h @@ -33,6 +33,7 @@ struct time_namespace { extern struct time_namespace init_time_ns;

#ifdef CONFIG_TIME_NS +void __init time_ns_init(void); extern int vdso_join_timens(struct task_struct *task, struct time_namespace *ns); extern void timens_commit(struct task_struct *tsk, struct time_namespace *ns); @@ -108,6 +109,10 @@ static inline ktime_t timens_ktime_to_host(clockid_t clockid, ktime_t tim) }

#else +static inline void __init time_ns_init(void) +{ +} + static inline int vdso_join_timens(struct task_struct *task, struct time_namespace *ns) { diff --git a/init/main.c b/init/main.c index 0ee0ee7b7c2c..e7d2c57c65a7 100644 --- a/init/main.c +++ b/init/main.c @@ -103,6 +103,7 @@ #include <linux/randomize_kstack.h> #include <linux/pidfs.h> #include <linux/ptdump.h> +#include <linux/time_namespace.h> #include <net/net_namespace.h>

#include <asm/io.h> @@ -1072,6 +1073,7 @@ void start_kernel(void) fork_init(); proc_caches_init(); uts_ns_init(); + time_ns_init(); key_init(); security_init(); dbg_late_init(); diff --git a/kernel/time/namespace.c b/kernel/time/namespace.c index 80b3d2ce2fb6..408f60d0a3b6 100644 --- a/kernel/time/namespace.c +++ b/kernel/time/namespace.c @@ -12,6 +12,7 @@ #include <linux/seq_file.h> #include <linux/proc_ns.h> #include <linux/export.h> +#include <linux/nstree.h> #include <linux/time.h> #include <linux/slab.h> #include <linux/cred.h> @@ -88,7 +89,7 @@ static struct time_namespace *clone_time_ns(struct user_namespace *user_ns, goto fail;

err = -ENOMEM; - ns = kmalloc(sizeof(*ns), GFP_KERNEL_ACCOUNT); + ns = kzalloc(sizeof(*ns), GFP_KERNEL_ACCOUNT); if (!ns) goto fail_dec;

@@ -104,6 +105,7 @@ static struct time_namespace *clone_time_ns(struct user_namespace *user_ns, ns->user_ns = get_user_ns(user_ns); ns->offsets = old_ns->offsets; ns->frozen_offsets = false; + ns_tree_add(ns); return ns;

fail_free_page: @@ -250,11 +252,13 @@ static void timens_set_vvar_page(struct task_struct *task,

void free_time_ns(struct time_namespace *ns) { + ns_tree_remove(ns); dec_time_namespaces(ns->ucounts); put_user_ns(ns->user_ns); ns_free_inum(&ns->ns); __free_page(ns->vvar_page); - kfree(ns); + /* Concurrent nstree traversal depends on a grace period. */ + kfree_rcu(ns, ns.ns_rcu); }

static struct time_namespace *to_time_ns(struct ns_common *ns) @@ -487,3 +491,8 @@ struct time_namespace init_time_ns = { .ns.ops = &timens_operations, .frozen_offsets = true, }; + +void __init time_ns_init(void) +{ + ns_tree_add(&init_time_ns); +}

Support the generic namespace iterator and lookup infrastructure to support file handles for namespaces.

Signed-off-by: Christian Brauner brauner@kernel.org --- kernel/user_namespace.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-)

diff --git a/kernel/user_namespace.c b/kernel/user_namespace.c index 98f4fe84d039..ade5b6806c5c 100644 --- a/kernel/user_namespace.c +++ b/kernel/user_namespace.c @@ -21,6 +21,7 @@ #include <linux/fs_struct.h> #include <linux/bsearch.h> #include <linux/sort.h> +#include <linux/nstree.h>

static struct kmem_cache *user_ns_cachep __ro_after_init; static DEFINE_MUTEX(userns_state_mutex); @@ -158,6 +159,7 @@ int create_user_ns(struct cred *new) goto fail_keyring;

set_cred_user_ns(new, ns); + ns_tree_add(ns); return 0; fail_keyring: #ifdef CONFIG_PERSISTENT_KEYRINGS @@ -200,6 +202,7 @@ static void free_user_ns(struct work_struct *work) do { struct ucounts *ucounts = ns->ucounts; parent = ns->parent; + ns_tree_remove(ns); if (ns->gid_map.nr_extents > UID_GID_MAP_MAX_BASE_EXTENTS) { kfree(ns->gid_map.forward); kfree(ns->gid_map.reverse); @@ -218,7 +221,8 @@ static void free_user_ns(struct work_struct *work) retire_userns_sysctls(ns); key_free_user_ns(ns); ns_free_inum(&ns->ns); - kmem_cache_free(user_ns_cachep, ns); + /* Concurrent nstree traversal depends on a grace period. */ + kfree_rcu(ns, ns.ns_rcu); dec_user_namespaces(ucounts); ns = parent; } while (refcount_dec_and_test(&parent->ns.count)); @@ -1412,6 +1416,7 @@ const struct proc_ns_operations userns_operations = { static __init int user_namespaces_init(void) { user_ns_cachep = KMEM_CACHE(user_namespace, SLAB_PANIC | SLAB_ACCOUNT); + ns_tree_add(&init_user_ns); return 0; } subsys_initcall(user_namespaces_init);

Every namespace type has a container_of(ns, <ns_type>, ns) static inline function that is currently not exposed in the header. So we have a bunch of places that open-code it via container_of(). Move it to the headers so we can use it directly.

Signed-off-by: Christian Brauner brauner@kernel.org --- include/linux/cgroup.h | 5 +++++ include/linux/ipc_namespace.h | 5 +++++ include/linux/pid_namespace.h | 5 +++++ include/linux/time_namespace.h | 4 ++++ include/linux/user_namespace.h | 5 +++++ include/linux/utsname.h | 5 +++++ include/net/net_namespace.h | 5 +++++ ipc/namespace.c | 5 ----- kernel/cgroup/namespace.c | 5 ----- kernel/pid_namespace.c | 5 ----- kernel/time/namespace.c | 5 ----- kernel/user_namespace.c | 5 ----- kernel/utsname.c | 5 ----- net/core/net_namespace.c | 5 ----- 14 files changed, 34 insertions(+), 35 deletions(-)

diff --git a/include/linux/cgroup.h b/include/linux/cgroup.h index b18fb5fcb38e..9ca25346f7cb 100644 --- a/include/linux/cgroup.h +++ b/include/linux/cgroup.h @@ -794,6 +794,11 @@ extern struct cgroup_namespace init_cgroup_ns;

#ifdef CONFIG_CGROUPS

+static inline struct cgroup_namespace *to_cg_ns(struct ns_common *ns) +{ + return container_of(ns, struct cgroup_namespace, ns); +} + void free_cgroup_ns(struct cgroup_namespace *ns);

struct cgroup_namespace *copy_cgroup_ns(unsigned long flags, diff --git a/include/linux/ipc_namespace.h b/include/linux/ipc_namespace.h index e8240cf2611a..924e4754374f 100644 --- a/include/linux/ipc_namespace.h +++ b/include/linux/ipc_namespace.h @@ -129,6 +129,11 @@ static inline int mq_init_ns(struct ipc_namespace *ns) { return 0; } #endif

#if defined(CONFIG_IPC_NS) +static inline struct ipc_namespace *to_ipc_ns(struct ns_common *ns) +{ + return container_of(ns, struct ipc_namespace, ns); +} + extern struct ipc_namespace *copy_ipcs(unsigned long flags, struct user_namespace *user_ns, struct ipc_namespace *ns);

diff --git a/include/linux/pid_namespace.h b/include/linux/pid_namespace.h index 7c67a5811199..ba0efc8c8596 100644 --- a/include/linux/pid_namespace.h +++ b/include/linux/pid_namespace.h @@ -54,6 +54,11 @@ extern struct pid_namespace init_pid_ns; #define PIDNS_ADDING (1U << 31)

#ifdef CONFIG_PID_NS +static inline struct pid_namespace *to_pid_ns(struct ns_common *ns) +{ + return container_of(ns, struct pid_namespace, ns); +} + static inline struct pid_namespace *get_pid_ns(struct pid_namespace *ns) { if (ns != &init_pid_ns) diff --git a/include/linux/time_namespace.h b/include/linux/time_namespace.h index 7f6af7a9771e..a47a4ce4183e 100644 --- a/include/linux/time_namespace.h +++ b/include/linux/time_namespace.h @@ -33,6 +33,10 @@ struct time_namespace { extern struct time_namespace init_time_ns;

#ifdef CONFIG_TIME_NS +static inline struct time_namespace *to_time_ns(struct ns_common *ns) +{ + return container_of(ns, struct time_namespace, ns); +} void __init time_ns_init(void); extern int vdso_join_timens(struct task_struct *task, struct time_namespace *ns); diff --git a/include/linux/user_namespace.h b/include/linux/user_namespace.h index a0bb6d012137..a09056ad090e 100644 --- a/include/linux/user_namespace.h +++ b/include/linux/user_namespace.h @@ -168,6 +168,11 @@ static inline void set_userns_rlimit_max(struct user_namespace *ns,

#ifdef CONFIG_USER_NS

+static inline struct user_namespace *to_user_ns(struct ns_common *ns) +{ + return container_of(ns, struct user_namespace, ns); +} + static inline struct user_namespace *get_user_ns(struct user_namespace *ns) { if (ns) diff --git a/include/linux/utsname.h b/include/linux/utsname.h index bf7613ba412b..5d34c4f0f945 100644 --- a/include/linux/utsname.h +++ b/include/linux/utsname.h @@ -30,6 +30,11 @@ struct uts_namespace { extern struct uts_namespace init_uts_ns;

#ifdef CONFIG_UTS_NS +static inline struct uts_namespace *to_uts_ns(struct ns_common *ns) +{ + return container_of(ns, struct uts_namespace, ns); +} + static inline void get_uts_ns(struct uts_namespace *ns) { refcount_inc(&ns->ns.count); diff --git a/include/net/net_namespace.h b/include/net/net_namespace.h index 42075748dff1..b9c5f6c7ee1e 100644 --- a/include/net/net_namespace.h +++ b/include/net/net_namespace.h @@ -263,6 +263,11 @@ void ipx_unregister_sysctl(void); #ifdef CONFIG_NET_NS void __put_net(struct net *net);

+static inline struct net *to_net_ns(struct ns_common *ns) +{ + return container_of(ns, struct net, ns); +} + /* Try using get_net_track() instead */ static inline struct net *get_net(struct net *net) { diff --git a/ipc/namespace.c b/ipc/namespace.c index 9f923c1a1eb3..89588819956b 100644 --- a/ipc/namespace.c +++ b/ipc/namespace.c @@ -209,11 +209,6 @@ void put_ipc_ns(struct ipc_namespace *ns) } }

-static inline struct ipc_namespace *to_ipc_ns(struct ns_common *ns) -{ - return container_of(ns, struct ipc_namespace, ns); -} - static struct ns_common *ipcns_get(struct task_struct *task) { struct ipc_namespace *ns = NULL; diff --git a/kernel/cgroup/namespace.c b/kernel/cgroup/namespace.c index fc12c416dfeb..5a327914b565 100644 --- a/kernel/cgroup/namespace.c +++ b/kernel/cgroup/namespace.c @@ -89,11 +89,6 @@ struct cgroup_namespace *copy_cgroup_ns(unsigned long flags, return new_ns; }

-static inline struct cgroup_namespace *to_cg_ns(struct ns_common *ns) -{ - return container_of(ns, struct cgroup_namespace, ns); -} - static int cgroupns_install(struct nsset *nsset, struct ns_common *ns) { struct nsproxy *nsproxy = nsset->nsproxy; diff --git a/kernel/pid_namespace.c b/kernel/pid_namespace.c index 228ae20299f9..9b327420309e 100644 --- a/kernel/pid_namespace.c +++ b/kernel/pid_namespace.c @@ -345,11 +345,6 @@ int reboot_pid_ns(struct pid_namespace *pid_ns, int cmd) return 0; }

-static inline struct pid_namespace *to_pid_ns(struct ns_common *ns) -{ - return container_of(ns, struct pid_namespace, ns); -} - static struct ns_common *pidns_get(struct task_struct *task) { struct pid_namespace *ns; diff --git a/kernel/time/namespace.c b/kernel/time/namespace.c index 408f60d0a3b6..20b65f90549e 100644 --- a/kernel/time/namespace.c +++ b/kernel/time/namespace.c @@ -261,11 +261,6 @@ void free_time_ns(struct time_namespace *ns) kfree_rcu(ns, ns.ns_rcu); }

-static struct time_namespace *to_time_ns(struct ns_common *ns) -{ - return container_of(ns, struct time_namespace, ns); -} - static struct ns_common *timens_get(struct task_struct *task) { struct time_namespace *ns = NULL; diff --git a/kernel/user_namespace.c b/kernel/user_namespace.c index ade5b6806c5c..cfb0e28f2779 100644 --- a/kernel/user_namespace.c +++ b/kernel/user_namespace.c @@ -1325,11 +1325,6 @@ bool current_in_userns(const struct user_namespace *target_ns) } EXPORT_SYMBOL(current_in_userns);

-static inline struct user_namespace *to_user_ns(struct ns_common *ns) -{ - return container_of(ns, struct user_namespace, ns); -} - static struct ns_common *userns_get(struct task_struct *task) { struct user_namespace *user_ns; diff --git a/kernel/utsname.c b/kernel/utsname.c index 64155417ae0c..a682830742d3 100644 --- a/kernel/utsname.c +++ b/kernel/utsname.c @@ -103,11 +103,6 @@ void free_uts_ns(struct uts_namespace *ns) kfree_rcu(ns, ns.ns_rcu); }

-static inline struct uts_namespace *to_uts_ns(struct ns_common *ns) -{ - return container_of(ns, struct uts_namespace, ns); -} - static struct ns_common *utsns_get(struct task_struct *task) { struct uts_namespace *ns = NULL; diff --git a/net/core/net_namespace.c b/net/core/net_namespace.c index b85e303400be..ca9b06f3925f 100644 --- a/net/core/net_namespace.c +++ b/net/core/net_namespace.c @@ -1539,11 +1539,6 @@ static struct ns_common *netns_get(struct task_struct *task) return net ? &net->ns : NULL; }

-static inline struct net *to_net_ns(struct ns_common *ns) -{ - return container_of(ns, struct net, ns); -} - static void netns_put(struct ns_common *ns) { put_net(to_net_ns(ns));

Add a helper to easily check whether a given namespace is the caller's current namespace. This is currently open-coded in a lot of places. Simply switch on the type and compare the results.

Signed-off-by: Christian Brauner brauner@kernel.org --- include/linux/nsfs.h | 16 +++++++++++++++- 1 file changed, 15 insertions(+), 1 deletion(-)

diff --git a/include/linux/nsfs.h b/include/linux/nsfs.h index fb84aa538091..e5a5fa83d36b 100644 --- a/include/linux/nsfs.h +++ b/include/linux/nsfs.h @@ -5,6 +5,8 @@ #define _LINUX_NSFS_H

#include <linux/ns_common.h> +#include <linux/cred.h> +#include <linux/pid_namespace.h>

struct path; struct task_struct; @@ -22,5 +24,17 @@ int ns_get_name(char *buf, size_t size, struct task_struct *task, const struct proc_ns_operations *ns_ops); void nsfs_init(void);

-#endif /* _LINUX_NSFS_H */ +#define __current_namespace_from_type(__ns) \ + _Generic((__ns), \ + struct cgroup_namespace *: current->nsproxy->cgroup_ns, \ + struct ipc_namespace *: current->nsproxy->ipc_ns, \ + struct net *: current->nsproxy->net_ns, \ + struct pid_namespace *: task_active_pid_ns(current), \ + struct mnt_namespace *: current->nsproxy->mnt_ns, \ + struct time_namespace *: current->nsproxy->time_ns, \ + struct user_namespace *: current_user_ns(), \ + struct uts_namespace *: current->nsproxy->uts_ns) + +#define current_in_namespace(__ns) (__current_namespace_from_type(__ns) == __ns)

+#endif /* _LINUX_NSFS_H */

A while ago we added support for file handles to pidfs so pidfds can be encoded and decoded as file handles. Userspace has adopted this quickly and it's proven very useful. Pidfd file handles are exhaustive meaning they don't require a handle on another pidfd to pass to open_by_handle_at() so it can derive the filesystem to decode in.

Implement the exhaustive file handles for namespaces as well.

Signed-off-by: Christian Brauner brauner@kernel.org --- fs/nsfs.c | 176 +++++++++++++++++++++++++++++++++++++++++++++++ include/linux/exportfs.h | 6 ++ 2 files changed, 182 insertions(+)

diff --git a/fs/nsfs.c b/fs/nsfs.c index 6f8008177133..a1585a2f4f03 100644 --- a/fs/nsfs.c +++ b/fs/nsfs.c @@ -13,6 +13,12 @@ #include <linux/nsfs.h> #include <linux/uaccess.h> #include <linux/mnt_namespace.h> +#include <linux/ipc_namespace.h> +#include <linux/time_namespace.h> +#include <linux/utsname.h> +#include <linux/exportfs.h> +#include <linux/nstree.h> +#include <net/net_namespace.h>

#include "mount.h" #include "internal.h" @@ -417,12 +423,182 @@ static const struct stashed_operations nsfs_stashed_ops = { .put_data = nsfs_put_data, };

+struct nsfs_fid { + u64 ns_id; + u32 ns_type; + u32 ns_inum; +} __attribute__ ((packed)); + +#define NSFS_FID_SIZE (sizeof(struct nsfs_fid) / sizeof(u32)) + +static int nsfs_encode_fh(struct inode *inode, u32 *fh, int *max_len, + struct inode *parent) +{ + struct nsfs_fid *fid = (struct nsfs_fid *)fh; + struct ns_common *ns = inode->i_private; + int len = *max_len; + + /* + * TODO: + * For hierarchical namespaces we should start to encode the + * parent namespace. Then userspace can walk a namespace + * hierarchy purely based on file handles. + */ + if (parent) + return FILEID_INVALID; + + if (len < NSFS_FID_SIZE) { + *max_len = NSFS_FID_SIZE; + return FILEID_INVALID; + } + + len = NSFS_FID_SIZE; + + fid->ns_id = ns->ns_id; + fid->ns_type = ns->ops->type; + fid->ns_inum = inode->i_ino; + *max_len = len; + return FILEID_NSFS; +} + +static struct dentry *nsfs_fh_to_dentry(struct super_block *sb, struct fid *fh, + int fh_len, int fh_type) +{ + struct path path __free(path_put) = {}; + struct nsfs_fid *fid = (struct nsfs_fid *)fh; + struct user_namespace *owning_ns = NULL; + struct ns_common *ns; + int ret; + + if (fh_len < NSFS_FID_SIZE) + return NULL; + + switch (fh_type) { + case FILEID_NSFS: + break; + default: + return NULL; + } + + scoped_guard(rcu) { + ns = ns_tree_lookup_rcu(fid->ns_id, fid->ns_type); + if (!ns) + return NULL; + + VFS_WARN_ON_ONCE(ns->ns_id != fid->ns_id); + VFS_WARN_ON_ONCE(ns->ops->type != fid->ns_type); + VFS_WARN_ON_ONCE(ns->inum != fid->ns_inum); + + if (!refcount_inc_not_zero(&ns->count)) + return NULL; + } + + switch (ns->ops->type) { +#ifdef CONFIG_CGROUPS + case CLONE_NEWCGROUP: + if (!current_in_namespace(to_cg_ns(ns))) + owning_ns = to_cg_ns(ns)->user_ns; + break; +#endif +#ifdef CONFIG_IPC_NS + case CLONE_NEWIPC: + if (!current_in_namespace(to_ipc_ns(ns))) + owning_ns = to_ipc_ns(ns)->user_ns; + break; +#endif + case CLONE_NEWNS: + if (!current_in_namespace(to_mnt_ns(ns))) + owning_ns = to_mnt_ns(ns)->user_ns; + break; +#ifdef CONFIG_NET_NS + case CLONE_NEWNET: + if (!current_in_namespace(to_net_ns(ns))) + owning_ns = to_net_ns(ns)->user_ns; + break; +#endif +#ifdef CONFIG_PID_NS + case CLONE_NEWPID: + if (!current_in_namespace(to_pid_ns(ns))) { + owning_ns = to_pid_ns(ns)->user_ns; + } else if (!READ_ONCE(to_pid_ns(ns)->child_reaper)) { + ns->ops->put(ns); + return ERR_PTR(-EPERM); + } + break; +#endif +#ifdef CONFIG_TIME_NS + case CLONE_NEWTIME: + if (!current_in_namespace(to_time_ns(ns))) + owning_ns = to_time_ns(ns)->user_ns; + break; +#endif +#ifdef CONFIG_USER_NS + case CLONE_NEWUSER: + if (!current_in_namespace(to_user_ns(ns))) + owning_ns = to_user_ns(ns); + break; +#endif +#ifdef CONFIG_UTS_NS + case CLONE_NEWUTS: + if (!current_in_namespace(to_uts_ns(ns))) + owning_ns = to_uts_ns(ns)->user_ns; + break; +#endif + default: + return ERR_PTR(-EOPNOTSUPP); + } + + if (owning_ns && !ns_capable(owning_ns, CAP_SYS_ADMIN)) { + ns->ops->put(ns); + return ERR_PTR(-EPERM); + } + + /* path_from_stashed() unconditionally consumes the reference. */ + ret = path_from_stashed(&ns->stashed, nsfs_mnt, ns, &path); + if (ret) + return ERR_PTR(ret); + + return no_free_ptr(path.dentry); +} + +/* + * Make sure that we reject any nonsensical flags that users pass via + * open_by_handle_at(). + */ +#define VALID_FILE_HANDLE_OPEN_FLAGS \ + (O_RDONLY | O_WRONLY | O_RDWR | O_NONBLOCK | O_CLOEXEC | O_EXCL) + +static int nsfs_export_permission(struct handle_to_path_ctx *ctx, + unsigned int oflags) +{ + if (oflags & ~(VALID_FILE_HANDLE_OPEN_FLAGS | O_LARGEFILE)) + return -EINVAL; + + /* nsfs_fh_to_dentry() is performs further permission checks. */ + return 0; +} + +static struct file *nsfs_export_open(struct path *path, unsigned int oflags) +{ + /* Clear O_LARGEFILE as open_by_handle_at() forces it. */ + oflags &= ~O_LARGEFILE; + return file_open_root(path, "", oflags, 0); +} + +static const struct export_operations nsfs_export_operations = { + .encode_fh = nsfs_encode_fh, + .fh_to_dentry = nsfs_fh_to_dentry, + .open = nsfs_export_open, + .permission = nsfs_export_permission, +}; + static int nsfs_init_fs_context(struct fs_context *fc) { struct pseudo_fs_context *ctx = init_pseudo(fc, NSFS_MAGIC); if (!ctx) return -ENOMEM; ctx->ops = &nsfs_ops; + ctx->eops = &nsfs_export_operations; ctx->dops = &ns_dentry_operations; fc->s_fs_info = (void *)&nsfs_stashed_ops; return 0; diff --git a/include/linux/exportfs.h b/include/linux/exportfs.h index cfb0dd1ea49c..3aac58a520c7 100644 --- a/include/linux/exportfs.h +++ b/include/linux/exportfs.h @@ -122,6 +122,12 @@ enum fid_type { FILEID_BCACHEFS_WITHOUT_PARENT = 0xb1, FILEID_BCACHEFS_WITH_PARENT = 0xb2,

+ /* + * + * 64 bit namespace identifier, 32 bit namespace type, 32 bit inode number. + */ + FILEID_NSFS = 0xf1, + /* * 64 bit unique kernfs id */

Pidfd file handles are exhaustive meaning they don't require a handle on another pidfd to pass to open_by_handle_at() so it can derive the filesystem to decode in. Instead it can be derived from the file handle itself. The same is possible for namespace file handles.

Signed-off-by: Christian Brauner brauner@kernel.org --- fs/fhandle.c | 6 ++++++ fs/internal.h | 1 + fs/nsfs.c | 10 ++++++++++ include/uapi/linux/fcntl.h | 1 + 4 files changed, 18 insertions(+)

diff --git a/fs/fhandle.c b/fs/fhandle.c index 7c236f64cdea..f18c855bb0c2 100644 --- a/fs/fhandle.c +++ b/fs/fhandle.c @@ -11,6 +11,7 @@ #include <linux/personality.h> #include <linux/uaccess.h> #include <linux/compat.h> +#include <linux/nsfs.h> #include "internal.h" #include "mount.h"

@@ -189,6 +190,11 @@ static int get_path_anchor(int fd, struct path *root) return 0; }

+ if (fd == FD_NSFS_ROOT) { + nsfs_get_root(root); + return 0; + } + return -EBADF; }

diff --git a/fs/internal.h b/fs/internal.h index 38e8aab27bbd..a33d18ee5b74 100644 --- a/fs/internal.h +++ b/fs/internal.h @@ -355,3 +355,4 @@ int anon_inode_getattr(struct mnt_idmap *idmap, const struct path *path, int anon_inode_setattr(struct mnt_idmap *idmap, struct dentry *dentry, struct iattr *attr); void pidfs_get_root(struct path *path); +void nsfs_get_root(struct path *path); diff --git a/fs/nsfs.c b/fs/nsfs.c index a1585a2f4f03..3c6fcf652633 100644 --- a/fs/nsfs.c +++ b/fs/nsfs.c @@ -25,6 +25,14 @@

static struct vfsmount *nsfs_mnt;

+static struct path nsfs_root_path = {}; + +void nsfs_get_root(struct path *path) +{ + *path = nsfs_root_path; + path_get(path); +} + static long ns_ioctl(struct file *filp, unsigned int ioctl, unsigned long arg); static const struct file_operations ns_file_operations = { @@ -616,4 +624,6 @@ void __init nsfs_init(void) if (IS_ERR(nsfs_mnt)) panic("can't set nsfs up\n"); nsfs_mnt->mnt_sb->s_flags &= ~SB_NOUSER; + nsfs_root_path.mnt = nsfs_mnt; + nsfs_root_path.dentry = nsfs_mnt->mnt_root; } diff --git a/include/uapi/linux/fcntl.h b/include/uapi/linux/fcntl.h index f291ab4f94eb..3741ea1b73d8 100644 --- a/include/uapi/linux/fcntl.h +++ b/include/uapi/linux/fcntl.h @@ -111,6 +111,7 @@ #define PIDFD_SELF_THREAD_GROUP -10001 /* Current thread group leader. */

#define FD_PIDFS_ROOT -10002 /* Root of the pidfs filesystem */ +#define FD_NSFS_ROOT -10003 /* Root of the nsfs filesystem */ #define FD_INVALID -10009 /* Invalid file descriptor: -10000 - EBADF = -10009 */

/* Generic flags for the *at(2) family of syscalls. */

The mount namespace has supported id retrieval for a while already. Add support for the other types as well.

Signed-off-by: Christian Brauner brauner@kernel.org --- fs/nsfs.c | 74 +++++++++++++++++++++++++++++++++++++++-------- include/uapi/linux/nsfs.h | 12 ++++++-- 2 files changed, 72 insertions(+), 14 deletions(-)

diff --git a/fs/nsfs.c b/fs/nsfs.c index 3c6fcf652633..527480e67fd1 100644 --- a/fs/nsfs.c +++ b/fs/nsfs.c @@ -173,6 +173,13 @@ static bool nsfs_ioctl_valid(unsigned int cmd) case NS_GET_NSTYPE: case NS_GET_OWNER_UID: case NS_GET_MNTNS_ID: + case NS_GET_NETNS_ID: + case NS_GET_CGROUPNS_ID: + case NS_GET_IPCNS_ID: + case NS_GET_UTSNS_ID: + case NS_GET_PIDNS_ID: + case NS_GET_TIMENS_ID: + case NS_GET_USERNS_ID: case NS_GET_PID_FROM_PIDNS: case NS_GET_TGID_FROM_PIDNS: case NS_GET_PID_IN_PIDNS: @@ -226,18 +233,6 @@ static long ns_ioctl(struct file *filp, unsigned int ioctl, argp = (uid_t __user *) arg; uid = from_kuid_munged(current_user_ns(), user_ns->owner); return put_user(uid, argp); - case NS_GET_MNTNS_ID: { - __u64 __user *idp; - __u64 id; - - if (ns->ops->type != CLONE_NEWNS) - return -EINVAL; - - mnt_ns = container_of(ns, struct mnt_namespace, ns); - idp = (__u64 __user *)arg; - id = mnt_ns->ns.ns_id; - return put_user(id, idp); - } case NS_GET_PID_FROM_PIDNS: fallthrough; case NS_GET_TGID_FROM_PIDNS: @@ -283,6 +278,61 @@ static long ns_ioctl(struct file *filp, unsigned int ioctl, ret = -ESRCH; return ret; } + case NS_GET_MNTNS_ID: + fallthrough; + case NS_GET_NETNS_ID: + fallthrough; + case NS_GET_CGROUPNS_ID: + fallthrough; + case NS_GET_IPCNS_ID: + fallthrough; + case NS_GET_UTSNS_ID: + fallthrough; + case NS_GET_PIDNS_ID: + fallthrough; + case NS_GET_TIMENS_ID: + fallthrough; + case NS_GET_USERNS_ID: { + __u64 __user *idp; + __u64 id; + int expected_type; + + switch (ioctl) { + case NS_GET_MNTNS_ID: + expected_type = CLONE_NEWNS; + break; + case NS_GET_NETNS_ID: + expected_type = CLONE_NEWNET; + break; + case NS_GET_CGROUPNS_ID: + expected_type = CLONE_NEWCGROUP; + break; + case NS_GET_IPCNS_ID: + expected_type = CLONE_NEWIPC; + break; + case NS_GET_UTSNS_ID: + expected_type = CLONE_NEWUTS; + break; + case NS_GET_PIDNS_ID: + expected_type = CLONE_NEWPID; + break; + case NS_GET_TIMENS_ID: + expected_type = CLONE_NEWTIME; + break; + case NS_GET_USERNS_ID: + expected_type = CLONE_NEWUSER; + break; + default: + return -EINVAL; + } + + if (ns->ops->type != expected_type) + return -EINVAL; + + idp = (__u64 __user *)arg; + id = ns->ns_id; + return put_user(id, idp); + } }

/* extensible ioctls */ diff --git a/include/uapi/linux/nsfs.h b/include/uapi/linux/nsfs.h index 97d8d80d139f..f7c21840cc09 100644 --- a/include/uapi/linux/nsfs.h +++ b/include/uapi/linux/nsfs.h @@ -16,8 +16,6 @@ #define NS_GET_NSTYPE _IO(NSIO, 0x3) /* Get owner UID (in the caller's user namespace) for a user namespace */ #define NS_GET_OWNER_UID _IO(NSIO, 0x4) -/* Get the id for a mount namespace */ -#define NS_GET_MNTNS_ID _IOR(NSIO, 0x5, __u64) /* Translate pid from target pid namespace into the caller's pid namespace. */ #define NS_GET_PID_FROM_PIDNS _IOR(NSIO, 0x6, int) /* Return thread-group leader id of pid in the callers pid namespace. */ @@ -42,6 +40,16 @@ struct mnt_ns_info { /* Get previous namespace. */ #define NS_MNT_GET_PREV _IOR(NSIO, 12, struct mnt_ns_info)

+/* Retrieve namespace identifiers. */ +#define NS_GET_MNTNS_ID _IOR(NSIO, 5, __u64) +#define NS_GET_NETNS_ID _IOR(NSIO, 13, __u64) +#define NS_GET_CGROUPNS_ID _IOR(NSIO, 14, __u64) +#define NS_GET_IPCNS_ID _IOR(NSIO, 15, __u64) +#define NS_GET_UTSNS_ID _IOR(NSIO, 16, __u64) +#define NS_GET_PIDNS_ID _IOR(NSIO, 17, __u64) +#define NS_GET_TIMENS_ID _IOR(NSIO, 18, __u64) +#define NS_GET_USERNS_ID _IOR(NSIO, 19, __u64) + enum init_ns_ino { IPC_NS_INIT_INO = 0xEFFFFFFFU, UTS_NS_INIT_INO = 0xEFFFFFFEU,

Update the nsfs.h tools header to the uapi/nsfs.h header so we can rely on it in the selftests.

Signed-off-by: Christian Brauner brauner@kernel.org --- tools/include/uapi/linux/nsfs.h | 23 +++++++++++++++++++++-- 1 file changed, 21 insertions(+), 2 deletions(-)

diff --git a/tools/include/uapi/linux/nsfs.h b/tools/include/uapi/linux/nsfs.h index 34127653fd00..f7c21840cc09 100644 --- a/tools/include/uapi/linux/nsfs.h +++ b/tools/include/uapi/linux/nsfs.h @@ -16,8 +16,6 @@ #define NS_GET_NSTYPE _IO(NSIO, 0x3) /* Get owner UID (in the caller's user namespace) for a user namespace */ #define NS_GET_OWNER_UID _IO(NSIO, 0x4) -/* Get the id for a mount namespace */ -#define NS_GET_MNTNS_ID _IOR(NSIO, 0x5, __u64) /* Translate pid from target pid namespace into the caller's pid namespace. */ #define NS_GET_PID_FROM_PIDNS _IOR(NSIO, 0x6, int) /* Return thread-group leader id of pid in the callers pid namespace. */ @@ -42,4 +40,25 @@ struct mnt_ns_info { /* Get previous namespace. */ #define NS_MNT_GET_PREV _IOR(NSIO, 12, struct mnt_ns_info)

+/* Retrieve namespace identifiers. */ +#define NS_GET_MNTNS_ID _IOR(NSIO, 5, __u64) +#define NS_GET_NETNS_ID _IOR(NSIO, 13, __u64) +#define NS_GET_CGROUPNS_ID _IOR(NSIO, 14, __u64) +#define NS_GET_IPCNS_ID _IOR(NSIO, 15, __u64) +#define NS_GET_UTSNS_ID _IOR(NSIO, 16, __u64) +#define NS_GET_PIDNS_ID _IOR(NSIO, 17, __u64) +#define NS_GET_TIMENS_ID _IOR(NSIO, 18, __u64) +#define NS_GET_USERNS_ID _IOR(NSIO, 19, __u64) + +enum init_ns_ino { + IPC_NS_INIT_INO = 0xEFFFFFFFU, + UTS_NS_INIT_INO = 0xEFFFFFFEU, + USER_NS_INIT_INO = 0xEFFFFFFDU, + PID_NS_INIT_INO = 0xEFFFFFFCU, + CGROUP_NS_INIT_INO = 0xEFFFFFFBU, + TIME_NS_INIT_INO = 0xEFFFFFFAU, + NET_NS_INIT_INO = 0xEFFFFFF9U, + MNT_NS_INIT_INO = 0xEFFFFFF8U, +}; + #endif /* __LINUX_NSFS_H */

Add a bunch of selftests for namespace file handles.

Signed-off-by: Christian Brauner brauner@kernel.org --- tools/testing/selftests/namespaces/.gitignore | 1 + tools/testing/selftests/namespaces/Makefile | 2 +- .../selftests/namespaces/file_handle_test.c | 1410 ++++++++++++++++++++ 3 files changed, 1412 insertions(+), 1 deletion(-)

diff --git a/tools/testing/selftests/namespaces/.gitignore b/tools/testing/selftests/namespaces/.gitignore index c1e8d634dd21..7639dbf58bbf 100644 --- a/tools/testing/selftests/namespaces/.gitignore +++ b/tools/testing/selftests/namespaces/.gitignore @@ -1 +1,2 @@ nsid_test +file_handle_test diff --git a/tools/testing/selftests/namespaces/Makefile b/tools/testing/selftests/namespaces/Makefile index 9280c703533e..f6c117ce2c2b 100644 --- a/tools/testing/selftests/namespaces/Makefile +++ b/tools/testing/selftests/namespaces/Makefile @@ -1,7 +1,7 @@ # SPDX-License-Identifier: GPL-2.0-only CFLAGS += -Wall -O0 -g $(KHDR_INCLUDES) $(TOOLS_INCLUDES)

-TEST_GEN_PROGS := nsid_test +TEST_GEN_PROGS := nsid_test file_handle_test

include ../lib.mk

diff --git a/tools/testing/selftests/namespaces/file_handle_test.c b/tools/testing/selftests/namespaces/file_handle_test.c new file mode 100644 index 000000000000..87573fa06990 --- /dev/null +++ b/tools/testing/selftests/namespaces/file_handle_test.c @@ -0,0 +1,1410 @@ +// SPDX-License-Identifier: GPL-2.0 +#define _GNU_SOURCE +#include <errno.h> +#include <fcntl.h> +#include <grp.h> +#include <limits.h> +#include <sched.h> +#include <stdio.h> +#include <stdlib.h> +#include <string.h> +#include <sys/mount.h> +#include <sys/stat.h> +#include <sys/types.h> +#include <sys/wait.h> +#include <unistd.h> +#include <linux/unistd.h> +#include "../kselftest_harness.h" + +#ifndef FD_NSFS_ROOT +#define FD_NSFS_ROOT -10003 /* Root of the nsfs filesystem */ +#endif + +TEST(nsfs_net_handle) +{ + struct file_handle *handle; + int mount_id; + int ret; + int fd; + int ns_fd; + char ns_path[256]; + struct stat st1, st2; + + handle = malloc(sizeof(*handle) + MAX_HANDLE_SZ); + ASSERT_NE(handle, NULL); + + /* Open a namespace file descriptor */ + snprintf(ns_path, sizeof(ns_path), "/proc/self/ns/net"); + ns_fd = open(ns_path, O_RDONLY); + ASSERT_GE(ns_fd, 0); + + /* Get handle for the namespace */ + handle->handle_bytes = MAX_HANDLE_SZ; + ret = name_to_handle_at(ns_fd, "", handle, &mount_id, AT_EMPTY_PATH); + if (ret < 0 && errno == EOPNOTSUPP) { + SKIP(free(handle); close(ns_fd); + return, "nsfs doesn't support file handles"); + } + ASSERT_EQ(ret, 0); + ASSERT_GT(handle->handle_bytes, 0); + + /* Try to open using FD_NSFS_ROOT */ + fd = open_by_handle_at(FD_NSFS_ROOT, handle, O_RDONLY); + if (fd < 0 && (errno == EINVAL || errno == EOPNOTSUPP)) { + SKIP(free(handle); close(ns_fd); + return, + "open_by_handle_at with FD_NSFS_ROOT not supported"); + } + ASSERT_GE(fd, 0); + + /* Verify we opened the correct namespace */ + ASSERT_EQ(fstat(ns_fd, &st1), 0); + ASSERT_EQ(fstat(fd, &st2), 0); + ASSERT_EQ(st1.st_ino, st2.st_ino); + ASSERT_EQ(st1.st_dev, st2.st_dev); + + close(fd); + close(ns_fd); + free(handle); +} + +TEST(nsfs_uts_handle) +{ + struct file_handle *handle; + int mount_id; + int ret; + int fd; + int ns_fd; + char ns_path[256]; + struct stat st1, st2; + + handle = malloc(sizeof(*handle) + MAX_HANDLE_SZ); + ASSERT_NE(handle, NULL); + + /* Open UTS namespace file descriptor */ + snprintf(ns_path, sizeof(ns_path), "/proc/self/ns/uts"); + ns_fd = open(ns_path, O_RDONLY); + ASSERT_GE(ns_fd, 0); + + /* Get handle for the namespace */ + handle->handle_bytes = MAX_HANDLE_SZ; + ret = name_to_handle_at(ns_fd, "", handle, &mount_id, AT_EMPTY_PATH); + if (ret < 0 && errno == EOPNOTSUPP) { + SKIP(free(handle); close(ns_fd); + return, "nsfs doesn't support file handles"); + } + ASSERT_EQ(ret, 0); + ASSERT_GT(handle->handle_bytes, 0); + + /* Try to open using FD_NSFS_ROOT */ + fd = open_by_handle_at(FD_NSFS_ROOT, handle, O_RDONLY); + if (fd < 0 && (errno == EINVAL || errno == EOPNOTSUPP)) { + SKIP(free(handle); close(ns_fd); + return, + "open_by_handle_at with FD_NSFS_ROOT not supported"); + } + ASSERT_GE(fd, 0); + + /* Verify we opened the correct namespace */ + ASSERT_EQ(fstat(ns_fd, &st1), 0); + ASSERT_EQ(fstat(fd, &st2), 0); + ASSERT_EQ(st1.st_ino, st2.st_ino); + ASSERT_EQ(st1.st_dev, st2.st_dev); + + close(fd); + close(ns_fd); + free(handle); +} + +TEST(nsfs_ipc_handle) +{ + struct file_handle *handle; + int mount_id; + int ret; + int fd; + int ns_fd; + char ns_path[256]; + struct stat st1, st2; + + handle = malloc(sizeof(*handle) + MAX_HANDLE_SZ); + ASSERT_NE(handle, NULL); + + /* Open IPC namespace file descriptor */ + snprintf(ns_path, sizeof(ns_path), "/proc/self/ns/ipc"); + ns_fd = open(ns_path, O_RDONLY); + ASSERT_GE(ns_fd, 0); + + /* Get handle for the namespace */ + handle->handle_bytes = MAX_HANDLE_SZ; + ret = name_to_handle_at(ns_fd, "", handle, &mount_id, AT_EMPTY_PATH); + if (ret < 0 && errno == EOPNOTSUPP) { + SKIP(free(handle); close(ns_fd); + return, "nsfs doesn't support file handles"); + } + ASSERT_EQ(ret, 0); + ASSERT_GT(handle->handle_bytes, 0); + + /* Try to open using FD_NSFS_ROOT */ + fd = open_by_handle_at(FD_NSFS_ROOT, handle, O_RDONLY); + if (fd < 0 && (errno == EINVAL || errno == EOPNOTSUPP)) { + SKIP(free(handle); close(ns_fd); + return, + "open_by_handle_at with FD_NSFS_ROOT not supported"); + } + ASSERT_GE(fd, 0); + + /* Verify we opened the correct namespace */ + ASSERT_EQ(fstat(ns_fd, &st1), 0); + ASSERT_EQ(fstat(fd, &st2), 0); + ASSERT_EQ(st1.st_ino, st2.st_ino); + ASSERT_EQ(st1.st_dev, st2.st_dev); + + close(fd); + close(ns_fd); + free(handle); +} + +TEST(nsfs_pid_handle) +{ + struct file_handle *handle; + int mount_id; + int ret; + int fd; + int ns_fd; + char ns_path[256]; + struct stat st1, st2; + + handle = malloc(sizeof(*handle) + MAX_HANDLE_SZ); + ASSERT_NE(handle, NULL); + + /* Open PID namespace file descriptor */ + snprintf(ns_path, sizeof(ns_path), "/proc/self/ns/pid"); + ns_fd = open(ns_path, O_RDONLY); + ASSERT_GE(ns_fd, 0); + + /* Get handle for the namespace */ + handle->handle_bytes = MAX_HANDLE_SZ; + ret = name_to_handle_at(ns_fd, "", handle, &mount_id, AT_EMPTY_PATH); + if (ret < 0 && errno == EOPNOTSUPP) { + SKIP(free(handle); close(ns_fd); + return, "nsfs doesn't support file handles"); + } + ASSERT_EQ(ret, 0); + ASSERT_GT(handle->handle_bytes, 0); + + /* Try to open using FD_NSFS_ROOT */ + fd = open_by_handle_at(FD_NSFS_ROOT, handle, O_RDONLY); + if (fd < 0 && (errno == EINVAL || errno == EOPNOTSUPP)) { + SKIP(free(handle); close(ns_fd); + return, + "open_by_handle_at with FD_NSFS_ROOT not supported"); + } + ASSERT_GE(fd, 0); + + /* Verify we opened the correct namespace */ + ASSERT_EQ(fstat(ns_fd, &st1), 0); + ASSERT_EQ(fstat(fd, &st2), 0); + ASSERT_EQ(st1.st_ino, st2.st_ino); + ASSERT_EQ(st1.st_dev, st2.st_dev); + + close(fd); + close(ns_fd); + free(handle); +} + +TEST(nsfs_mnt_handle) +{ + struct file_handle *handle; + int mount_id; + int ret; + int fd; + int ns_fd; + char ns_path[256]; + struct stat st1, st2; + + handle = malloc(sizeof(*handle) + MAX_HANDLE_SZ); + ASSERT_NE(handle, NULL); + + /* Open mount namespace file descriptor */ + snprintf(ns_path, sizeof(ns_path), "/proc/self/ns/mnt"); + ns_fd = open(ns_path, O_RDONLY); + ASSERT_GE(ns_fd, 0); + + /* Get handle for the namespace */ + handle->handle_bytes = MAX_HANDLE_SZ; + ret = name_to_handle_at(ns_fd, "", handle, &mount_id, AT_EMPTY_PATH); + if (ret < 0 && errno == EOPNOTSUPP) { + SKIP(free(handle); close(ns_fd); + return, "nsfs doesn't support file handles"); + } + ASSERT_EQ(ret, 0); + ASSERT_GT(handle->handle_bytes, 0); + + /* Try to open using FD_NSFS_ROOT */ + fd = open_by_handle_at(FD_NSFS_ROOT, handle, O_RDONLY); + if (fd < 0 && (errno == EINVAL || errno == EOPNOTSUPP)) { + SKIP(free(handle); close(ns_fd); + return, + "open_by_handle_at with FD_NSFS_ROOT not supported"); + } + ASSERT_GE(fd, 0); + + /* Verify we opened the correct namespace */ + ASSERT_EQ(fstat(ns_fd, &st1), 0); + ASSERT_EQ(fstat(fd, &st2), 0); + ASSERT_EQ(st1.st_ino, st2.st_ino); + ASSERT_EQ(st1.st_dev, st2.st_dev); + + close(fd); + close(ns_fd); + free(handle); +} + +TEST(nsfs_user_handle) +{ + struct file_handle *handle; + int mount_id; + int ret; + int fd; + int ns_fd; + char ns_path[256]; + struct stat st1, st2; + + handle = malloc(sizeof(*handle) + MAX_HANDLE_SZ); + ASSERT_NE(handle, NULL); + + /* Open user namespace file descriptor */ + snprintf(ns_path, sizeof(ns_path), "/proc/self/ns/user"); + ns_fd = open(ns_path, O_RDONLY); + ASSERT_GE(ns_fd, 0); + + /* Get handle for the namespace */ + handle->handle_bytes = MAX_HANDLE_SZ; + ret = name_to_handle_at(ns_fd, "", handle, &mount_id, AT_EMPTY_PATH); + if (ret < 0 && errno == EOPNOTSUPP) { + SKIP(free(handle); close(ns_fd); + return, "nsfs doesn't support file handles"); + } + ASSERT_EQ(ret, 0); + ASSERT_GT(handle->handle_bytes, 0); + + /* Try to open using FD_NSFS_ROOT */ + fd = open_by_handle_at(FD_NSFS_ROOT, handle, O_RDONLY); + if (fd < 0 && (errno == EINVAL || errno == EOPNOTSUPP)) { + SKIP(free(handle); close(ns_fd); + return, + "open_by_handle_at with FD_NSFS_ROOT not supported"); + } + ASSERT_GE(fd, 0); + + /* Verify we opened the correct namespace */ + ASSERT_EQ(fstat(ns_fd, &st1), 0); + ASSERT_EQ(fstat(fd, &st2), 0); + ASSERT_EQ(st1.st_ino, st2.st_ino); + ASSERT_EQ(st1.st_dev, st2.st_dev); + + close(fd); + close(ns_fd); + free(handle); +} + +TEST(nsfs_cgroup_handle) +{ + struct file_handle *handle; + int mount_id; + int ret; + int fd; + int ns_fd; + char ns_path[256]; + struct stat st1, st2; + + handle = malloc(sizeof(*handle) + MAX_HANDLE_SZ); + ASSERT_NE(handle, NULL); + + /* Open cgroup namespace file descriptor */ + snprintf(ns_path, sizeof(ns_path), "/proc/self/ns/cgroup"); + ns_fd = open(ns_path, O_RDONLY); + if (ns_fd < 0) { + SKIP(free(handle); return, "cgroup namespace not available"); + } + + /* Get handle for the namespace */ + handle->handle_bytes = MAX_HANDLE_SZ; + ret = name_to_handle_at(ns_fd, "", handle, &mount_id, AT_EMPTY_PATH); + if (ret < 0 && errno == EOPNOTSUPP) { + SKIP(free(handle); close(ns_fd); + return, "nsfs doesn't support file handles"); + } + ASSERT_EQ(ret, 0); + ASSERT_GT(handle->handle_bytes, 0); + + /* Try to open using FD_NSFS_ROOT */ + fd = open_by_handle_at(FD_NSFS_ROOT, handle, O_RDONLY); + if (fd < 0 && (errno == EINVAL || errno == EOPNOTSUPP)) { + SKIP(free(handle); close(ns_fd); + return, + "open_by_handle_at with FD_NSFS_ROOT not supported"); + } + ASSERT_GE(fd, 0); + + /* Verify we opened the correct namespace */ + ASSERT_EQ(fstat(ns_fd, &st1), 0); + ASSERT_EQ(fstat(fd, &st2), 0); + ASSERT_EQ(st1.st_ino, st2.st_ino); + ASSERT_EQ(st1.st_dev, st2.st_dev); + + close(fd); + close(ns_fd); + free(handle); +} + +TEST(nsfs_time_handle) +{ + struct file_handle *handle; + int mount_id; + int ret; + int fd; + int ns_fd; + char ns_path[256]; + struct stat st1, st2; + + handle = malloc(sizeof(*handle) + MAX_HANDLE_SZ); + ASSERT_NE(handle, NULL); + + /* Open time namespace file descriptor */ + snprintf(ns_path, sizeof(ns_path), "/proc/self/ns/time"); + ns_fd = open(ns_path, O_RDONLY); + if (ns_fd < 0) { + SKIP(free(handle); return, "time namespace not available"); + } + + /* Get handle for the namespace */ + handle->handle_bytes = MAX_HANDLE_SZ; + ret = name_to_handle_at(ns_fd, "", handle, &mount_id, AT_EMPTY_PATH); + if (ret < 0 && errno == EOPNOTSUPP) { + SKIP(free(handle); close(ns_fd); + return, "nsfs doesn't support file handles"); + } + ASSERT_EQ(ret, 0); + ASSERT_GT(handle->handle_bytes, 0); + + /* Try to open using FD_NSFS_ROOT */ + fd = open_by_handle_at(FD_NSFS_ROOT, handle, O_RDONLY); + if (fd < 0 && (errno == EINVAL || errno == EOPNOTSUPP)) { + SKIP(free(handle); close(ns_fd); + return, + "open_by_handle_at with FD_NSFS_ROOT not supported"); + } + ASSERT_GE(fd, 0); + + /* Verify we opened the correct namespace */ + ASSERT_EQ(fstat(ns_fd, &st1), 0); + ASSERT_EQ(fstat(fd, &st2), 0); + ASSERT_EQ(st1.st_ino, st2.st_ino); + ASSERT_EQ(st1.st_dev, st2.st_dev); + + close(fd); + close(ns_fd); + free(handle); +} + +TEST(nsfs_user_net_namespace_isolation) +{ + struct file_handle *handle; + int mount_id; + int ret; + int fd; + int ns_fd; + pid_t pid; + int status; + int pipefd[2]; + char result; + + handle = malloc(sizeof(*handle) + MAX_HANDLE_SZ); + ASSERT_NE(handle, NULL); + + /* Create pipe for communication */ + ASSERT_EQ(pipe(pipefd), 0); + + /* Get handle for current network namespace */ + ns_fd = open("/proc/self/ns/net", O_RDONLY); + ASSERT_GE(ns_fd, 0); + + handle->handle_bytes = MAX_HANDLE_SZ; + ret = name_to_handle_at(ns_fd, "", handle, &mount_id, AT_EMPTY_PATH); + if (ret < 0 && errno == EOPNOTSUPP) { + SKIP(free(handle); close(ns_fd); close(pipefd[0]); + close(pipefd[1]); + return, "nsfs doesn't support file handles"); + } + ASSERT_EQ(ret, 0); + close(ns_fd); + + pid = fork(); + ASSERT_GE(pid, 0); + + if (pid == 0) { + /* Child process */ + close(pipefd[0]); + + /* First create new user namespace to drop privileges */ + ret = unshare(CLONE_NEWUSER); + if (ret < 0) { + write(pipefd[1], "U", + 1); /* Unable to create user namespace */ + close(pipefd[1]); + exit(0); + } + + /* Write uid/gid mappings to maintain some capabilities */ + int uid_map_fd = open("/proc/self/uid_map", O_WRONLY); + int gid_map_fd = open("/proc/self/gid_map", O_WRONLY); + int setgroups_fd = open("/proc/self/setgroups", O_WRONLY); + + if (uid_map_fd < 0 || gid_map_fd < 0 || setgroups_fd < 0) { + write(pipefd[1], "M", 1); /* Unable to set mappings */ + close(pipefd[1]); + exit(0); + } + + /* Disable setgroups to allow gid mapping */ + write(setgroups_fd, "deny", 4); + close(setgroups_fd); + + /* Map current uid/gid to root in the new namespace */ + char mapping[64]; + snprintf(mapping, sizeof(mapping), "0 %d 1", getuid()); + write(uid_map_fd, mapping, strlen(mapping)); + close(uid_map_fd); + + snprintf(mapping, sizeof(mapping), "0 %d 1", getgid()); + write(gid_map_fd, mapping, strlen(mapping)); + close(gid_map_fd); + + /* Now create new network namespace */ + ret = unshare(CLONE_NEWNET); + if (ret < 0) { + write(pipefd[1], "N", + 1); /* Unable to create network namespace */ + close(pipefd[1]); + exit(0); + } + + /* Try to open parent's network namespace handle from new user+net namespace */ + fd = open_by_handle_at(FD_NSFS_ROOT, handle, O_RDONLY); + + if (fd >= 0) { + /* Should NOT succeed - we're in a different user namespace */ + write(pipefd[1], "S", 1); /* Unexpected success */ + close(fd); + } else if (errno == ESTALE) { + /* Expected: Stale file handle */ + write(pipefd[1], "P", 1); + } else { + /* Other error */ + write(pipefd[1], "F", 1); + } + + close(pipefd[1]); + exit(0); + } + + /* Parent process */ + close(pipefd[1]); + ASSERT_EQ(read(pipefd[0], &result, 1), 1); + + waitpid(pid, &status, 0); + ASSERT_TRUE(WIFEXITED(status)); + ASSERT_EQ(WEXITSTATUS(status), 0); + + if (result == 'U') { + SKIP(free(handle); close(pipefd[0]); + return, "Cannot create new user namespace"); + } + if (result == 'M') { + SKIP(free(handle); close(pipefd[0]); + return, "Cannot set uid/gid mappings"); + } + if (result == 'N') { + SKIP(free(handle); close(pipefd[0]); + return, "Cannot create new network namespace"); + } + + /* Should fail with permission denied since we're in a different user namespace */ + ASSERT_EQ(result, 'P'); + + close(pipefd[0]); + free(handle); +} + +TEST(nsfs_user_uts_namespace_isolation) +{ + struct file_handle *handle; + int mount_id; + int ret; + int fd; + int ns_fd; + pid_t pid; + int status; + int pipefd[2]; + char result; + + handle = malloc(sizeof(*handle) + MAX_HANDLE_SZ); + ASSERT_NE(handle, NULL); + + /* Create pipe for communication */ + ASSERT_EQ(pipe(pipefd), 0); + + /* Get handle for current UTS namespace */ + ns_fd = open("/proc/self/ns/uts", O_RDONLY); + ASSERT_GE(ns_fd, 0); + + handle->handle_bytes = MAX_HANDLE_SZ; + ret = name_to_handle_at(ns_fd, "", handle, &mount_id, AT_EMPTY_PATH); + if (ret < 0 && errno == EOPNOTSUPP) { + SKIP(free(handle); close(ns_fd); close(pipefd[0]); + close(pipefd[1]); + return, "nsfs doesn't support file handles"); + } + ASSERT_EQ(ret, 0); + close(ns_fd); + + pid = fork(); + ASSERT_GE(pid, 0); + + if (pid == 0) { + /* Child process */ + close(pipefd[0]); + + /* First create new user namespace to drop privileges */ + ret = unshare(CLONE_NEWUSER); + if (ret < 0) { + write(pipefd[1], "U", + 1); /* Unable to create user namespace */ + close(pipefd[1]); + exit(0); + } + + /* Write uid/gid mappings to maintain some capabilities */ + int uid_map_fd = open("/proc/self/uid_map", O_WRONLY); + int gid_map_fd = open("/proc/self/gid_map", O_WRONLY); + int setgroups_fd = open("/proc/self/setgroups", O_WRONLY); + + if (uid_map_fd < 0 || gid_map_fd < 0 || setgroups_fd < 0) { + write(pipefd[1], "M", 1); /* Unable to set mappings */ + close(pipefd[1]); + exit(0); + } + + /* Disable setgroups to allow gid mapping */ + write(setgroups_fd, "deny", 4); + close(setgroups_fd); + + /* Map current uid/gid to root in the new namespace */ + char mapping[64]; + snprintf(mapping, sizeof(mapping), "0 %d 1", getuid()); + write(uid_map_fd, mapping, strlen(mapping)); + close(uid_map_fd); + + snprintf(mapping, sizeof(mapping), "0 %d 1", getgid()); + write(gid_map_fd, mapping, strlen(mapping)); + close(gid_map_fd); + + /* Now create new UTS namespace */ + ret = unshare(CLONE_NEWUTS); + if (ret < 0) { + write(pipefd[1], "N", + 1); /* Unable to create UTS namespace */ + close(pipefd[1]); + exit(0); + } + + /* Try to open parent's UTS namespace handle from new user+uts namespace */ + fd = open_by_handle_at(FD_NSFS_ROOT, handle, O_RDONLY); + + if (fd >= 0) { + /* Should NOT succeed - we're in a different user namespace */ + write(pipefd[1], "S", 1); /* Unexpected success */ + close(fd); + } else if (errno == ESTALE) { + /* Expected: Stale file handle */ + write(pipefd[1], "P", 1); + } else { + /* Other error */ + write(pipefd[1], "F", 1); + } + + close(pipefd[1]); + exit(0); + } + + /* Parent process */ + close(pipefd[1]); + ASSERT_EQ(read(pipefd[0], &result, 1), 1); + + waitpid(pid, &status, 0); + ASSERT_TRUE(WIFEXITED(status)); + ASSERT_EQ(WEXITSTATUS(status), 0); + + if (result == 'U') { + SKIP(free(handle); close(pipefd[0]); + return, "Cannot create new user namespace"); + } + if (result == 'M') { + SKIP(free(handle); close(pipefd[0]); + return, "Cannot set uid/gid mappings"); + } + if (result == 'N') { + SKIP(free(handle); close(pipefd[0]); + return, "Cannot create new UTS namespace"); + } + + /* Should fail with ESTALE since we're in a different user namespace */ + ASSERT_EQ(result, 'P'); + + close(pipefd[0]); + free(handle); +} + +TEST(nsfs_user_ipc_namespace_isolation) +{ + struct file_handle *handle; + int mount_id; + int ret; + int fd; + int ns_fd; + pid_t pid; + int status; + int pipefd[2]; + char result; + + handle = malloc(sizeof(*handle) + MAX_HANDLE_SZ); + ASSERT_NE(handle, NULL); + + /* Create pipe for communication */ + ASSERT_EQ(pipe(pipefd), 0); + + /* Get handle for current IPC namespace */ + ns_fd = open("/proc/self/ns/ipc", O_RDONLY); + ASSERT_GE(ns_fd, 0); + + handle->handle_bytes = MAX_HANDLE_SZ; + ret = name_to_handle_at(ns_fd, "", handle, &mount_id, AT_EMPTY_PATH); + if (ret < 0 && errno == EOPNOTSUPP) { + SKIP(free(handle); close(ns_fd); close(pipefd[0]); + close(pipefd[1]); + return, "nsfs doesn't support file handles"); + } + ASSERT_EQ(ret, 0); + close(ns_fd); + + pid = fork(); + ASSERT_GE(pid, 0); + + if (pid == 0) { + /* Child process */ + close(pipefd[0]); + + /* First create new user namespace to drop privileges */ + ret = unshare(CLONE_NEWUSER); + if (ret < 0) { + write(pipefd[1], "U", + 1); /* Unable to create user namespace */ + close(pipefd[1]); + exit(0); + } + + /* Write uid/gid mappings to maintain some capabilities */ + int uid_map_fd = open("/proc/self/uid_map", O_WRONLY); + int gid_map_fd = open("/proc/self/gid_map", O_WRONLY); + int setgroups_fd = open("/proc/self/setgroups", O_WRONLY); + + if (uid_map_fd < 0 || gid_map_fd < 0 || setgroups_fd < 0) { + write(pipefd[1], "M", 1); /* Unable to set mappings */ + close(pipefd[1]); + exit(0); + } + + /* Disable setgroups to allow gid mapping */ + write(setgroups_fd, "deny", 4); + close(setgroups_fd); + + /* Map current uid/gid to root in the new namespace */ + char mapping[64]; + snprintf(mapping, sizeof(mapping), "0 %d 1", getuid()); + write(uid_map_fd, mapping, strlen(mapping)); + close(uid_map_fd); + + snprintf(mapping, sizeof(mapping), "0 %d 1", getgid()); + write(gid_map_fd, mapping, strlen(mapping)); + close(gid_map_fd); + + /* Now create new IPC namespace */ + ret = unshare(CLONE_NEWIPC); + if (ret < 0) { + write(pipefd[1], "N", + 1); /* Unable to create IPC namespace */ + close(pipefd[1]); + exit(0); + } + + /* Try to open parent's IPC namespace handle from new user+ipc namespace */ + fd = open_by_handle_at(FD_NSFS_ROOT, handle, O_RDONLY); + + if (fd >= 0) { + /* Should NOT succeed - we're in a different user namespace */ + write(pipefd[1], "S", 1); /* Unexpected success */ + close(fd); + } else if (errno == ESTALE) { + /* Expected: Stale file handle */ + write(pipefd[1], "P", 1); + } else { + /* Other error */ + write(pipefd[1], "F", 1); + } + + close(pipefd[1]); + exit(0); + } + + /* Parent process */ + close(pipefd[1]); + ASSERT_EQ(read(pipefd[0], &result, 1), 1); + + waitpid(pid, &status, 0); + ASSERT_TRUE(WIFEXITED(status)); + ASSERT_EQ(WEXITSTATUS(status), 0); + + if (result == 'U') { + SKIP(free(handle); close(pipefd[0]); + return, "Cannot create new user namespace"); + } + if (result == 'M') { + SKIP(free(handle); close(pipefd[0]); + return, "Cannot set uid/gid mappings"); + } + if (result == 'N') { + SKIP(free(handle); close(pipefd[0]); + return, "Cannot create new IPC namespace"); + } + + /* Should fail with ESTALE since we're in a different user namespace */ + ASSERT_EQ(result, 'P'); + + close(pipefd[0]); + free(handle); +} + +TEST(nsfs_user_mnt_namespace_isolation) +{ + struct file_handle *handle; + int mount_id; + int ret; + int fd; + int ns_fd; + pid_t pid; + int status; + int pipefd[2]; + char result; + + handle = malloc(sizeof(*handle) + MAX_HANDLE_SZ); + ASSERT_NE(handle, NULL); + + /* Create pipe for communication */ + ASSERT_EQ(pipe(pipefd), 0); + + /* Get handle for current mount namespace */ + ns_fd = open("/proc/self/ns/mnt", O_RDONLY); + ASSERT_GE(ns_fd, 0); + + handle->handle_bytes = MAX_HANDLE_SZ; + ret = name_to_handle_at(ns_fd, "", handle, &mount_id, AT_EMPTY_PATH); + if (ret < 0 && errno == EOPNOTSUPP) { + SKIP(free(handle); close(ns_fd); close(pipefd[0]); + close(pipefd[1]); + return, "nsfs doesn't support file handles"); + } + ASSERT_EQ(ret, 0); + close(ns_fd); + + pid = fork(); + ASSERT_GE(pid, 0); + + if (pid == 0) { + /* Child process */ + close(pipefd[0]); + + /* First create new user namespace to drop privileges */ + ret = unshare(CLONE_NEWUSER); + if (ret < 0) { + write(pipefd[1], "U", + 1); /* Unable to create user namespace */ + close(pipefd[1]); + exit(0); + } + + /* Write uid/gid mappings to maintain some capabilities */ + int uid_map_fd = open("/proc/self/uid_map", O_WRONLY); + int gid_map_fd = open("/proc/self/gid_map", O_WRONLY); + int setgroups_fd = open("/proc/self/setgroups", O_WRONLY); + + if (uid_map_fd < 0 || gid_map_fd < 0 || setgroups_fd < 0) { + write(pipefd[1], "M", 1); /* Unable to set mappings */ + close(pipefd[1]); + exit(0); + } + + /* Disable setgroups to allow gid mapping */ + write(setgroups_fd, "deny", 4); + close(setgroups_fd); + + /* Map current uid/gid to root in the new namespace */ + char mapping[64]; + snprintf(mapping, sizeof(mapping), "0 %d 1", getuid()); + write(uid_map_fd, mapping, strlen(mapping)); + close(uid_map_fd); + + snprintf(mapping, sizeof(mapping), "0 %d 1", getgid()); + write(gid_map_fd, mapping, strlen(mapping)); + close(gid_map_fd); + + /* Now create new mount namespace */ + ret = unshare(CLONE_NEWNS); + if (ret < 0) { + write(pipefd[1], "N", + 1); /* Unable to create mount namespace */ + close(pipefd[1]); + exit(0); + } + + /* Try to open parent's mount namespace handle from new user+mnt namespace */ + fd = open_by_handle_at(FD_NSFS_ROOT, handle, O_RDONLY); + + if (fd >= 0) { + /* Should NOT succeed - we're in a different user namespace */ + write(pipefd[1], "S", 1); /* Unexpected success */ + close(fd); + } else if (errno == ESTALE) { + /* Expected: Stale file handle */ + write(pipefd[1], "P", 1); + } else { + /* Other error */ + write(pipefd[1], "F", 1); + } + + close(pipefd[1]); + exit(0); + } + + /* Parent process */ + close(pipefd[1]); + ASSERT_EQ(read(pipefd[0], &result, 1), 1); + + waitpid(pid, &status, 0); + ASSERT_TRUE(WIFEXITED(status)); + ASSERT_EQ(WEXITSTATUS(status), 0); + + if (result == 'U') { + SKIP(free(handle); close(pipefd[0]); + return, "Cannot create new user namespace"); + } + if (result == 'M') { + SKIP(free(handle); close(pipefd[0]); + return, "Cannot set uid/gid mappings"); + } + if (result == 'N') { + SKIP(free(handle); close(pipefd[0]); + return, "Cannot create new mount namespace"); + } + + /* Should fail with ESTALE since we're in a different user namespace */ + ASSERT_EQ(result, 'P'); + + close(pipefd[0]); + free(handle); +} + +TEST(nsfs_user_cgroup_namespace_isolation) +{ + struct file_handle *handle; + int mount_id; + int ret; + int fd; + int ns_fd; + pid_t pid; + int status; + int pipefd[2]; + char result; + + handle = malloc(sizeof(*handle) + MAX_HANDLE_SZ); + ASSERT_NE(handle, NULL); + + /* Create pipe for communication */ + ASSERT_EQ(pipe(pipefd), 0); + + /* Get handle for current cgroup namespace */ + ns_fd = open("/proc/self/ns/cgroup", O_RDONLY); + if (ns_fd < 0) { + SKIP(free(handle); close(pipefd[0]); close(pipefd[1]); + return, "cgroup namespace not available"); + } + + handle->handle_bytes = MAX_HANDLE_SZ; + ret = name_to_handle_at(ns_fd, "", handle, &mount_id, AT_EMPTY_PATH); + if (ret < 0 && errno == EOPNOTSUPP) { + SKIP(free(handle); close(ns_fd); close(pipefd[0]); + close(pipefd[1]); + return, "nsfs doesn't support file handles"); + } + ASSERT_EQ(ret, 0); + close(ns_fd); + + pid = fork(); + ASSERT_GE(pid, 0); + + if (pid == 0) { + /* Child process */ + close(pipefd[0]); + + /* First create new user namespace to drop privileges */ + ret = unshare(CLONE_NEWUSER); + if (ret < 0) { + write(pipefd[1], "U", + 1); /* Unable to create user namespace */ + close(pipefd[1]); + exit(0); + } + + /* Write uid/gid mappings to maintain some capabilities */ + int uid_map_fd = open("/proc/self/uid_map", O_WRONLY); + int gid_map_fd = open("/proc/self/gid_map", O_WRONLY); + int setgroups_fd = open("/proc/self/setgroups", O_WRONLY); + + if (uid_map_fd < 0 || gid_map_fd < 0 || setgroups_fd < 0) { + write(pipefd[1], "M", 1); /* Unable to set mappings */ + close(pipefd[1]); + exit(0); + } + + /* Disable setgroups to allow gid mapping */ + write(setgroups_fd, "deny", 4); + close(setgroups_fd); + + /* Map current uid/gid to root in the new namespace */ + char mapping[64]; + snprintf(mapping, sizeof(mapping), "0 %d 1", getuid()); + write(uid_map_fd, mapping, strlen(mapping)); + close(uid_map_fd); + + snprintf(mapping, sizeof(mapping), "0 %d 1", getgid()); + write(gid_map_fd, mapping, strlen(mapping)); + close(gid_map_fd); + + /* Now create new cgroup namespace */ + ret = unshare(CLONE_NEWCGROUP); + if (ret < 0) { + write(pipefd[1], "N", + 1); /* Unable to create cgroup namespace */ + close(pipefd[1]); + exit(0); + } + + /* Try to open parent's cgroup namespace handle from new user+cgroup namespace */ + fd = open_by_handle_at(FD_NSFS_ROOT, handle, O_RDONLY); + + if (fd >= 0) { + /* Should NOT succeed - we're in a different user namespace */ + write(pipefd[1], "S", 1); /* Unexpected success */ + close(fd); + } else if (errno == ESTALE) { + /* Expected: Stale file handle */ + write(pipefd[1], "P", 1); + } else { + /* Other error */ + write(pipefd[1], "F", 1); + } + + close(pipefd[1]); + exit(0); + } + + /* Parent process */ + close(pipefd[1]); + ASSERT_EQ(read(pipefd[0], &result, 1), 1); + + waitpid(pid, &status, 0); + ASSERT_TRUE(WIFEXITED(status)); + ASSERT_EQ(WEXITSTATUS(status), 0); + + if (result == 'U') { + SKIP(free(handle); close(pipefd[0]); + return, "Cannot create new user namespace"); + } + if (result == 'M') { + SKIP(free(handle); close(pipefd[0]); + return, "Cannot set uid/gid mappings"); + } + if (result == 'N') { + SKIP(free(handle); close(pipefd[0]); + return, "Cannot create new cgroup namespace"); + } + + /* Should fail with ESTALE since we're in a different user namespace */ + ASSERT_EQ(result, 'P'); + + close(pipefd[0]); + free(handle); +} + +TEST(nsfs_user_pid_namespace_isolation) +{ + struct file_handle *handle; + int mount_id; + int ret; + int fd; + int ns_fd; + pid_t pid; + int status; + int pipefd[2]; + char result; + + handle = malloc(sizeof(*handle) + MAX_HANDLE_SZ); + ASSERT_NE(handle, NULL); + + /* Create pipe for communication */ + ASSERT_EQ(pipe(pipefd), 0); + + /* Get handle for current PID namespace */ + ns_fd = open("/proc/self/ns/pid", O_RDONLY); + ASSERT_GE(ns_fd, 0); + + handle->handle_bytes = MAX_HANDLE_SZ; + ret = name_to_handle_at(ns_fd, "", handle, &mount_id, AT_EMPTY_PATH); + if (ret < 0 && errno == EOPNOTSUPP) { + SKIP(free(handle); close(ns_fd); close(pipefd[0]); + close(pipefd[1]); + return, "nsfs doesn't support file handles"); + } + ASSERT_EQ(ret, 0); + close(ns_fd); + + pid = fork(); + ASSERT_GE(pid, 0); + + if (pid == 0) { + /* Child process */ + close(pipefd[0]); + + /* First create new user namespace to drop privileges */ + ret = unshare(CLONE_NEWUSER); + if (ret < 0) { + write(pipefd[1], "U", + 1); /* Unable to create user namespace */ + close(pipefd[1]); + exit(0); + } + + /* Write uid/gid mappings to maintain some capabilities */ + int uid_map_fd = open("/proc/self/uid_map", O_WRONLY); + int gid_map_fd = open("/proc/self/gid_map", O_WRONLY); + int setgroups_fd = open("/proc/self/setgroups", O_WRONLY); + + if (uid_map_fd < 0 || gid_map_fd < 0 || setgroups_fd < 0) { + write(pipefd[1], "M", 1); /* Unable to set mappings */ + close(pipefd[1]); + exit(0); + } + + /* Disable setgroups to allow gid mapping */ + write(setgroups_fd, "deny", 4); + close(setgroups_fd); + + /* Map current uid/gid to root in the new namespace */ + char mapping[64]; + snprintf(mapping, sizeof(mapping), "0 %d 1", getuid()); + write(uid_map_fd, mapping, strlen(mapping)); + close(uid_map_fd); + + snprintf(mapping, sizeof(mapping), "0 %d 1", getgid()); + write(gid_map_fd, mapping, strlen(mapping)); + close(gid_map_fd); + + /* Now create new PID namespace - requires fork to take effect */ + ret = unshare(CLONE_NEWPID); + if (ret < 0) { + write(pipefd[1], "N", + 1); /* Unable to create PID namespace */ + close(pipefd[1]); + exit(0); + } + + /* Fork again for PID namespace to take effect */ + pid_t child_pid = fork(); + if (child_pid < 0) { + write(pipefd[1], "N", + 1); /* Unable to fork in PID namespace */ + close(pipefd[1]); + exit(0); + } + + if (child_pid == 0) { + /* Grandchild in new PID namespace */ + /* Try to open parent's PID namespace handle from new user+pid namespace */ + fd = open_by_handle_at(FD_NSFS_ROOT, handle, O_RDONLY); + + if (fd >= 0) { + /* Should NOT succeed - we're in a different user namespace */ + write(pipefd[1], "S", + 1); /* Unexpected success */ + close(fd); + } else if (errno == ESTALE) { + /* Expected: Stale file handle */ + write(pipefd[1], "P", 1); + } else { + /* Other error */ + write(pipefd[1], "F", 1); + } + + close(pipefd[1]); + exit(0); + } + + /* Wait for grandchild */ + waitpid(child_pid, NULL, 0); + exit(0); + } + + /* Parent process */ + close(pipefd[1]); + ASSERT_EQ(read(pipefd[0], &result, 1), 1); + + waitpid(pid, &status, 0); + ASSERT_TRUE(WIFEXITED(status)); + ASSERT_EQ(WEXITSTATUS(status), 0); + + if (result == 'U') { + SKIP(free(handle); close(pipefd[0]); + return, "Cannot create new user namespace"); + } + if (result == 'M') { + SKIP(free(handle); close(pipefd[0]); + return, "Cannot set uid/gid mappings"); + } + if (result == 'N') { + SKIP(free(handle); close(pipefd[0]); + return, "Cannot create new PID namespace"); + } + + /* Should fail with ESTALE since we're in a different user namespace */ + ASSERT_EQ(result, 'P'); + + close(pipefd[0]); + free(handle); +} + +TEST(nsfs_user_time_namespace_isolation) +{ + struct file_handle *handle; + int mount_id; + int ret; + int fd; + int ns_fd; + pid_t pid; + int status; + int pipefd[2]; + char result; + + handle = malloc(sizeof(*handle) + MAX_HANDLE_SZ); + ASSERT_NE(handle, NULL); + + /* Create pipe for communication */ + ASSERT_EQ(pipe(pipefd), 0); + + /* Get handle for current time namespace */ + ns_fd = open("/proc/self/ns/time", O_RDONLY); + if (ns_fd < 0) { + SKIP(free(handle); close(pipefd[0]); close(pipefd[1]); + return, "time namespace not available"); + } + + handle->handle_bytes = MAX_HANDLE_SZ; + ret = name_to_handle_at(ns_fd, "", handle, &mount_id, AT_EMPTY_PATH); + if (ret < 0 && errno == EOPNOTSUPP) { + SKIP(free(handle); close(ns_fd); close(pipefd[0]); + close(pipefd[1]); + return, "nsfs doesn't support file handles"); + } + ASSERT_EQ(ret, 0); + close(ns_fd); + + pid = fork(); + ASSERT_GE(pid, 0); + + if (pid == 0) { + /* Child process */ + close(pipefd[0]); + + /* First create new user namespace to drop privileges */ + ret = unshare(CLONE_NEWUSER); + if (ret < 0) { + write(pipefd[1], "U", + 1); /* Unable to create user namespace */ + close(pipefd[1]); + exit(0); + } + + /* Write uid/gid mappings to maintain some capabilities */ + int uid_map_fd = open("/proc/self/uid_map", O_WRONLY); + int gid_map_fd = open("/proc/self/gid_map", O_WRONLY); + int setgroups_fd = open("/proc/self/setgroups", O_WRONLY); + + if (uid_map_fd < 0 || gid_map_fd < 0 || setgroups_fd < 0) { + write(pipefd[1], "M", 1); /* Unable to set mappings */ + close(pipefd[1]); + exit(0); + } + + /* Disable setgroups to allow gid mapping */ + write(setgroups_fd, "deny", 4); + close(setgroups_fd); + + /* Map current uid/gid to root in the new namespace */ + char mapping[64]; + snprintf(mapping, sizeof(mapping), "0 %d 1", getuid()); + write(uid_map_fd, mapping, strlen(mapping)); + close(uid_map_fd); + + snprintf(mapping, sizeof(mapping), "0 %d 1", getgid()); + write(gid_map_fd, mapping, strlen(mapping)); + close(gid_map_fd); + + /* Now create new time namespace - requires fork to take effect */ + ret = unshare(CLONE_NEWTIME); + if (ret < 0) { + write(pipefd[1], "N", + 1); /* Unable to create time namespace */ + close(pipefd[1]); + exit(0); + } + + /* Fork again for time namespace to take effect */ + pid_t child_pid = fork(); + if (child_pid < 0) { + write(pipefd[1], "N", + 1); /* Unable to fork in time namespace */ + close(pipefd[1]); + exit(0); + } + + if (child_pid == 0) { + /* Grandchild in new time namespace */ + /* Try to open parent's time namespace handle from new user+time namespace */ + fd = open_by_handle_at(FD_NSFS_ROOT, handle, O_RDONLY); + + if (fd >= 0) { + /* Should NOT succeed - we're in a different user namespace */ + write(pipefd[1], "S", + 1); /* Unexpected success */ + close(fd); + } else if (errno == ESTALE) { + /* Expected: Stale file handle */ + write(pipefd[1], "P", 1); + } else { + /* Other error */ + write(pipefd[1], "F", 1); + } + + close(pipefd[1]); + exit(0); + } + + /* Wait for grandchild */ + waitpid(child_pid, NULL, 0); + exit(0); + } + + /* Parent process */ + close(pipefd[1]); + ASSERT_EQ(read(pipefd[0], &result, 1), 1); + + waitpid(pid, &status, 0); + ASSERT_TRUE(WIFEXITED(status)); + ASSERT_EQ(WEXITSTATUS(status), 0); + + if (result == 'U') { + SKIP(free(handle); close(pipefd[0]); + return, "Cannot create new user namespace"); + } + if (result == 'M') { + SKIP(free(handle); close(pipefd[0]); + return, "Cannot set uid/gid mappings"); + } + if (result == 'N') { + SKIP(free(handle); close(pipefd[0]); + return, "Cannot create new time namespace"); + } + + /* Should fail with ESTALE since we're in a different user namespace */ + ASSERT_EQ(result, 'P'); + + close(pipefd[0]); + free(handle); +} + +TEST(nsfs_open_flags) +{ + struct file_handle *handle; + int mount_id; + int ret; + int fd; + int ns_fd; + char ns_path[256]; + + handle = malloc(sizeof(*handle) + MAX_HANDLE_SZ); + ASSERT_NE(handle, NULL); + + /* Open a namespace file descriptor */ + snprintf(ns_path, sizeof(ns_path), "/proc/self/ns/net"); + ns_fd = open(ns_path, O_RDONLY); + ASSERT_GE(ns_fd, 0); + + /* Get handle for the namespace */ + handle->handle_bytes = MAX_HANDLE_SZ; + ret = name_to_handle_at(ns_fd, "", handle, &mount_id, AT_EMPTY_PATH); + if (ret < 0 && errno == EOPNOTSUPP) { + SKIP(free(handle); close(ns_fd); + return, "nsfs doesn't support file handles"); + } + ASSERT_EQ(ret, 0); + ASSERT_GT(handle->handle_bytes, 0); + + /* Test invalid flags that should fail */ + fd = open_by_handle_at(FD_NSFS_ROOT, handle, O_WRONLY); + ASSERT_LT(fd, 0); + ASSERT_EQ(errno, EPERM); + + fd = open_by_handle_at(FD_NSFS_ROOT, handle, O_RDWR); + ASSERT_LT(fd, 0); + ASSERT_EQ(errno, EPERM); + + fd = open_by_handle_at(FD_NSFS_ROOT, handle, O_CREAT); + ASSERT_LT(fd, 0); + ASSERT_EQ(errno, EINVAL); + + fd = open_by_handle_at(FD_NSFS_ROOT, handle, O_TRUNC); + ASSERT_LT(fd, 0); + ASSERT_EQ(errno, EINVAL); + + fd = open_by_handle_at(FD_NSFS_ROOT, handle, O_APPEND); + ASSERT_LT(fd, 0); + ASSERT_EQ(errno, EINVAL); + + fd = open_by_handle_at(FD_NSFS_ROOT, handle, O_DIRECT); + ASSERT_LT(fd, 0); + ASSERT_EQ(errno, EINVAL); + + close(ns_fd); + free(handle); +} + +TEST_HARNESS_MAIN