From 13bcc6a2853435bb5dad368bcbaa9d2a5b9c0ac4 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Sat, 16 Jul 2016 15:22:55 -0500 Subject: sysctl: Stop implicitly passing current into sysctl_table_root.lookup Passing nsproxy into sysctl_table_root.lookup was a premature optimization in attempt to avoid depending on current. The directory /proc/self/sys has not appeared and if and when it does this code will need to be reviewed closely and reworked anyway. So remove the premature optimization. Acked-by: Kees Cook Acked-by: Serge Hallyn Signed-off-by: "Eric W. Biederman" --- net/sysctl_net.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'net') diff --git a/net/sysctl_net.c b/net/sysctl_net.c index 46a71c701e7c..ba9b5d1a31df 100644 --- a/net/sysctl_net.c +++ b/net/sysctl_net.c @@ -27,9 +27,9 @@ #endif static struct ctl_table_set * -net_ctl_header_lookup(struct ctl_table_root *root, struct nsproxy *namespaces) +net_ctl_header_lookup(struct ctl_table_root *root) { - return &namespaces->net_ns->sysctls; + return ¤t->nsproxy->net_ns->sysctls; } static int is_seen(struct ctl_table_set *set) -- cgit v1.2.3 From 703286608a220d53584cca5986aad5305eec75ed Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Mon, 8 Aug 2016 14:33:23 -0500 Subject: netns: Add a limit on the number of net namespaces Acked-by: Kees Cook Signed-off-by: "Eric W. Biederman" --- include/linux/user_namespace.h | 1 + include/net/net_namespace.h | 1 + kernel/ucount.c | 1 + net/core/net_namespace.c | 22 +++++++++++++++++++++- 4 files changed, 24 insertions(+), 1 deletion(-) (limited to 'net') diff --git a/include/linux/user_namespace.h b/include/linux/user_namespace.h index d067f0d3038e..c6bc980b06a9 100644 --- a/include/linux/user_namespace.h +++ b/include/linux/user_namespace.h @@ -29,6 +29,7 @@ enum ucount_type { UCOUNT_PID_NAMESPACES, UCOUNT_UTS_NAMESPACES, UCOUNT_IPC_NAMESPACES, + UCOUNT_NET_NAMESPACES, UCOUNT_CGROUP_NAMESPACES, UCOUNT_COUNTS, }; diff --git a/include/net/net_namespace.h b/include/net/net_namespace.h index 0933c7455a30..fc4f757107df 100644 --- a/include/net/net_namespace.h +++ b/include/net/net_namespace.h @@ -60,6 +60,7 @@ struct net { struct list_head exit_list; /* Use only net_mutex */ struct user_namespace *user_ns; /* Owning user namespace */ + struct ucounts *ucounts; spinlock_t nsid_lock; struct idr netns_ids; diff --git a/kernel/ucount.c b/kernel/ucount.c index 335cc5d2cdd7..205f1a07faac 100644 --- a/kernel/ucount.c +++ b/kernel/ucount.c @@ -71,6 +71,7 @@ static struct ctl_table user_table[] = { UCOUNT_ENTRY("max_pid_namespaces"), UCOUNT_ENTRY("max_uts_namespaces"), UCOUNT_ENTRY("max_ipc_namespaces"), + UCOUNT_ENTRY("max_net_namespaces"), UCOUNT_ENTRY("max_cgroup_namespaces"), { } }; diff --git a/net/core/net_namespace.c b/net/core/net_namespace.c index 2c2eb1b629b1..3e2812aeceb7 100644 --- a/net/core/net_namespace.c +++ b/net/core/net_namespace.c @@ -266,6 +266,16 @@ struct net *get_net_ns_by_id(struct net *net, int id) return peer; } +static struct ucounts *inc_net_namespaces(struct user_namespace *ns) +{ + return inc_ucount(ns, current_euid(), UCOUNT_NET_NAMESPACES); +} + +static void dec_net_namespaces(struct ucounts *ucounts) +{ + dec_ucount(ucounts, UCOUNT_NET_NAMESPACES); +} + /* * setup_net runs the initializers for the network namespace object. */ @@ -351,19 +361,27 @@ void net_drop_ns(void *p) struct net *copy_net_ns(unsigned long flags, struct user_namespace *user_ns, struct net *old_net) { + struct ucounts *ucounts; struct net *net; int rv; if (!(flags & CLONE_NEWNET)) return get_net(old_net); + ucounts = inc_net_namespaces(user_ns); + if (!ucounts) + return ERR_PTR(-ENFILE); + net = net_alloc(); - if (!net) + if (!net) { + dec_net_namespaces(ucounts); return ERR_PTR(-ENOMEM); + } get_user_ns(user_ns); mutex_lock(&net_mutex); + net->ucounts = ucounts; rv = setup_net(net, user_ns); if (rv == 0) { rtnl_lock(); @@ -372,6 +390,7 @@ struct net *copy_net_ns(unsigned long flags, } mutex_unlock(&net_mutex); if (rv < 0) { + dec_net_namespaces(ucounts); put_user_ns(user_ns); net_drop_ns(net); return ERR_PTR(rv); @@ -444,6 +463,7 @@ static void cleanup_net(struct work_struct *work) /* Finally it is safe to free my network namespace structure */ list_for_each_entry_safe(net, tmp, &net_exit_list, exit_list) { list_del_init(&net->exit_list); + dec_net_namespaces(net->ucounts); put_user_ns(net->user_ns); net_drop_ns(net); } -- cgit v1.2.3 From df75e7748bae1c7098bfa358485389b897f71305 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Thu, 22 Sep 2016 13:08:36 -0500 Subject: userns: When the per user per user namespace limit is reached return ENOSPC The current error codes returned when a the per user per user namespace limit are hit (EINVAL, EUSERS, and ENFILE) are wrong. I asked for advice on linux-api and it we made clear that those were the wrong error code, but a correct effor code was not suggested. The best general error code I have found for hitting a resource limit is ENOSPC. It is not perfect but as it is unambiguous it will serve until someone comes up with a better error code. Signed-off-by: "Eric W. Biederman" --- fs/namespace.c | 2 +- ipc/namespace.c | 2 +- kernel/cgroup.c | 2 +- kernel/pid_namespace.c | 2 +- kernel/user_namespace.c | 2 +- kernel/utsname.c | 2 +- net/core/net_namespace.c | 2 +- 7 files changed, 7 insertions(+), 7 deletions(-) (limited to 'net') diff --git a/fs/namespace.c b/fs/namespace.c index 491b8f3e4c9a..cf2cc234c8b4 100644 --- a/fs/namespace.c +++ b/fs/namespace.c @@ -2754,7 +2754,7 @@ static struct mnt_namespace *alloc_mnt_ns(struct user_namespace *user_ns) ucounts = inc_mnt_namespaces(user_ns); if (!ucounts) - return ERR_PTR(-ENFILE); + return ERR_PTR(-ENOSPC); new_ns = kmalloc(sizeof(struct mnt_namespace), GFP_KERNEL); if (!new_ns) { diff --git a/ipc/namespace.c b/ipc/namespace.c index 730914214135..fab727d9fe09 100644 --- a/ipc/namespace.c +++ b/ipc/namespace.c @@ -33,7 +33,7 @@ static struct ipc_namespace *create_ipc_ns(struct user_namespace *user_ns, struct ucounts *ucounts; int err; - err = -ENFILE; + err = -ENOSPC; ucounts = inc_ipc_namespaces(user_ns); if (!ucounts) goto fail; diff --git a/kernel/cgroup.c b/kernel/cgroup.c index e9e4427fec46..f1dd4b076210 100644 --- a/kernel/cgroup.c +++ b/kernel/cgroup.c @@ -6354,7 +6354,7 @@ struct cgroup_namespace *copy_cgroup_ns(unsigned long flags, ucounts = inc_cgroup_namespaces(user_ns); if (!ucounts) - return ERR_PTR(-ENFILE); + return ERR_PTR(-ENOSPC); /* It is not safe to take cgroup_mutex here */ spin_lock_irq(&css_set_lock); diff --git a/kernel/pid_namespace.c b/kernel/pid_namespace.c index 30a7f3351932..7542b28cc929 100644 --- a/kernel/pid_namespace.c +++ b/kernel/pid_namespace.c @@ -98,7 +98,7 @@ static struct pid_namespace *create_pid_namespace(struct user_namespace *user_ns int i; int err; - err = -EINVAL; + err = -ENOSPC; if (level > MAX_PID_NS_LEVEL) goto out; ucounts = inc_pid_namespaces(user_ns); diff --git a/kernel/user_namespace.c b/kernel/user_namespace.c index 0edafe305861..f2c5ba5505f1 100644 --- a/kernel/user_namespace.c +++ b/kernel/user_namespace.c @@ -76,7 +76,7 @@ int create_user_ns(struct cred *new) struct ucounts *ucounts; int ret, i; - ret = -EUSERS; + ret = -ENOSPC; if (parent_ns->level > 32) goto fail; diff --git a/kernel/utsname.c b/kernel/utsname.c index f3b0bb4ac3ba..35587b76faa3 100644 --- a/kernel/utsname.c +++ b/kernel/utsname.c @@ -49,7 +49,7 @@ static struct uts_namespace *clone_uts_ns(struct user_namespace *user_ns, struct ucounts *ucounts; int err; - err = -ENFILE; + err = -ENOSPC; ucounts = inc_uts_namespaces(user_ns); if (!ucounts) goto fail; diff --git a/net/core/net_namespace.c b/net/core/net_namespace.c index 3e2812aeceb7..06af5d6a883c 100644 --- a/net/core/net_namespace.c +++ b/net/core/net_namespace.c @@ -370,7 +370,7 @@ struct net *copy_net_ns(unsigned long flags, ucounts = inc_net_namespaces(user_ns); if (!ucounts) - return ERR_PTR(-ENFILE); + return ERR_PTR(-ENOSPC); net = net_alloc(); if (!net) { -- cgit v1.2.3 From bcac25a58bfc6bd79191ac5d7afb49bea96da8c9 Mon Sep 17 00:00:00 2001 From: Andrey Vagin Date: Tue, 6 Sep 2016 00:47:13 -0700 Subject: kernel: add a helper to get an owning user namespace for a namespace Return -EPERM if an owning user namespace is outside of a process current user namespace. v2: In a first version ns_get_owner returned ENOENT for init_user_ns. This special cases was removed from this version. There is nothing outside of init_user_ns, so we can return EPERM. v3: rename ns->get_owner() to ns->owner(). get_* usually means that it grabs a reference. Acked-by: Serge Hallyn Signed-off-by: Andrei Vagin Signed-off-by: Eric W. Biederman --- fs/namespace.c | 6 ++++++ include/linux/proc_ns.h | 1 + include/linux/user_namespace.h | 7 +++++++ ipc/namespace.c | 6 ++++++ kernel/cgroup.c | 6 ++++++ kernel/pid_namespace.c | 6 ++++++ kernel/user_namespace.c | 24 ++++++++++++++++++++++++ kernel/utsname.c | 6 ++++++ net/core/net_namespace.c | 6 ++++++ 9 files changed, 68 insertions(+) (limited to 'net') diff --git a/fs/namespace.c b/fs/namespace.c index 7bb2cda3bfef..fea56f310547 100644 --- a/fs/namespace.c +++ b/fs/namespace.c @@ -3348,10 +3348,16 @@ static int mntns_install(struct nsproxy *nsproxy, struct ns_common *ns) return 0; } +static struct user_namespace *mntns_owner(struct ns_common *ns) +{ + return to_mnt_ns(ns)->user_ns; +} + const struct proc_ns_operations mntns_operations = { .name = "mnt", .type = CLONE_NEWNS, .get = mntns_get, .put = mntns_put, .install = mntns_install, + .owner = mntns_owner, }; diff --git a/include/linux/proc_ns.h b/include/linux/proc_ns.h index de0e7719d4c5..ca85a4348ffc 100644 --- a/include/linux/proc_ns.h +++ b/include/linux/proc_ns.h @@ -18,6 +18,7 @@ struct proc_ns_operations { struct ns_common *(*get)(struct task_struct *task); void (*put)(struct ns_common *ns); int (*install)(struct nsproxy *nsproxy, struct ns_common *ns); + struct user_namespace *(*owner)(struct ns_common *ns); }; extern const struct proc_ns_operations netns_operations; diff --git a/include/linux/user_namespace.h b/include/linux/user_namespace.h index 9217169c64cb..190cf0760815 100644 --- a/include/linux/user_namespace.h +++ b/include/linux/user_namespace.h @@ -73,6 +73,8 @@ extern ssize_t proc_setgroups_write(struct file *, const char __user *, size_t, extern int proc_setgroups_show(struct seq_file *m, void *v); extern bool userns_may_setgroups(const struct user_namespace *ns); extern bool current_in_userns(const struct user_namespace *target_ns); + +struct ns_common *ns_get_owner(struct ns_common *ns); #else static inline struct user_namespace *get_user_ns(struct user_namespace *ns) @@ -106,6 +108,11 @@ static inline bool current_in_userns(const struct user_namespace *target_ns) { return true; } + +static inline struct ns_common *ns_get_owner(struct ns_common *ns) +{ + return ERR_PTR(-EPERM); +} #endif #endif /* _LINUX_USER_H */ diff --git a/ipc/namespace.c b/ipc/namespace.c index d87e6baa1323..578d93be619d 100644 --- a/ipc/namespace.c +++ b/ipc/namespace.c @@ -165,10 +165,16 @@ static int ipcns_install(struct nsproxy *nsproxy, struct ns_common *new) return 0; } +static struct user_namespace *ipcns_owner(struct ns_common *ns) +{ + return to_ipc_ns(ns)->user_ns; +} + const struct proc_ns_operations ipcns_operations = { .name = "ipc", .type = CLONE_NEWIPC, .get = ipcns_get, .put = ipcns_put, .install = ipcns_install, + .owner = ipcns_owner, }; diff --git a/kernel/cgroup.c b/kernel/cgroup.c index d1c51b7f5221..86b0e8b16426 100644 --- a/kernel/cgroup.c +++ b/kernel/cgroup.c @@ -6403,12 +6403,18 @@ static void cgroupns_put(struct ns_common *ns) put_cgroup_ns(to_cg_ns(ns)); } +static struct user_namespace *cgroupns_owner(struct ns_common *ns) +{ + return to_cg_ns(ns)->user_ns; +} + const struct proc_ns_operations cgroupns_operations = { .name = "cgroup", .type = CLONE_NEWCGROUP, .get = cgroupns_get, .put = cgroupns_put, .install = cgroupns_install, + .owner = cgroupns_owner, }; static __init int cgroup_namespaces_init(void) diff --git a/kernel/pid_namespace.c b/kernel/pid_namespace.c index a65ba137fd15..c02d744225e1 100644 --- a/kernel/pid_namespace.c +++ b/kernel/pid_namespace.c @@ -388,12 +388,18 @@ static int pidns_install(struct nsproxy *nsproxy, struct ns_common *ns) return 0; } +static struct user_namespace *pidns_owner(struct ns_common *ns) +{ + return to_pid_ns(ns)->user_ns; +} + const struct proc_ns_operations pidns_operations = { .name = "pid", .type = CLONE_NEWPID, .get = pidns_get, .put = pidns_put, .install = pidns_install, + .owner = pidns_owner, }; static __init int pid_namespaces_init(void) diff --git a/kernel/user_namespace.c b/kernel/user_namespace.c index 68f594212759..0ef683a03c20 100644 --- a/kernel/user_namespace.c +++ b/kernel/user_namespace.c @@ -1004,12 +1004,36 @@ static int userns_install(struct nsproxy *nsproxy, struct ns_common *ns) return commit_creds(cred); } +struct ns_common *ns_get_owner(struct ns_common *ns) +{ + struct user_namespace *my_user_ns = current_user_ns(); + struct user_namespace *owner, *p; + + /* See if the owner is in the current user namespace */ + owner = p = ns->ops->owner(ns); + for (;;) { + if (!p) + return ERR_PTR(-EPERM); + if (p == my_user_ns) + break; + p = p->parent; + } + + return &get_user_ns(owner)->ns; +} + +static struct user_namespace *userns_owner(struct ns_common *ns) +{ + return to_user_ns(ns)->parent; +} + const struct proc_ns_operations userns_operations = { .name = "user", .type = CLONE_NEWUSER, .get = userns_get, .put = userns_put, .install = userns_install, + .owner = userns_owner, }; static __init int user_namespaces_init(void) diff --git a/kernel/utsname.c b/kernel/utsname.c index 831ea7108232..e1211a8a5c18 100644 --- a/kernel/utsname.c +++ b/kernel/utsname.c @@ -130,10 +130,16 @@ static int utsns_install(struct nsproxy *nsproxy, struct ns_common *new) return 0; } +static struct user_namespace *utsns_owner(struct ns_common *ns) +{ + return to_uts_ns(ns)->user_ns; +} + const struct proc_ns_operations utsns_operations = { .name = "uts", .type = CLONE_NEWUTS, .get = utsns_get, .put = utsns_put, .install = utsns_install, + .owner = utsns_owner, }; diff --git a/net/core/net_namespace.c b/net/core/net_namespace.c index 2c2eb1b629b1..861efa34f08c 100644 --- a/net/core/net_namespace.c +++ b/net/core/net_namespace.c @@ -996,11 +996,17 @@ static int netns_install(struct nsproxy *nsproxy, struct ns_common *ns) return 0; } +static struct user_namespace *netns_owner(struct ns_common *ns) +{ + return to_net_ns(ns)->user_ns; +} + const struct proc_ns_operations netns_operations = { .name = "net", .type = CLONE_NEWNET, .get = netns_get, .put = netns_put, .install = netns_install, + .owner = netns_owner, }; #endif -- cgit v1.2.3 From 2ed6afdee798658fe3c33b50c4a79d1bde45f1d8 Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Fri, 23 Sep 2016 18:06:12 +0200 Subject: netns: move {inc,dec}_net_namespaces into #ifdef With the newly enforced limit on the number of namespaces, we get a build warning if CONFIG_NETNS is disabled: net/core/net_namespace.c:273:13: error: 'dec_net_namespaces' defined but not used [-Werror=unused-function] net/core/net_namespace.c:268:24: error: 'inc_net_namespaces' defined but not used [-Werror=unused-function] This moves the two added functions inside the #ifdef that guards their callers. Fixes: 703286608a22 ("netns: Add a limit on the number of net namespaces") Signed-off-by: Arnd Bergmann Signed-off-by: Eric W. Biederman --- net/core/net_namespace.c | 20 ++++++++++---------- 1 file changed, 10 insertions(+), 10 deletions(-) (limited to 'net') diff --git a/net/core/net_namespace.c b/net/core/net_namespace.c index e8be581b47b0..5e0042693c65 100644 --- a/net/core/net_namespace.c +++ b/net/core/net_namespace.c @@ -266,16 +266,6 @@ struct net *get_net_ns_by_id(struct net *net, int id) return peer; } -static struct ucounts *inc_net_namespaces(struct user_namespace *ns) -{ - return inc_ucount(ns, current_euid(), UCOUNT_NET_NAMESPACES); -} - -static void dec_net_namespaces(struct ucounts *ucounts) -{ - dec_ucount(ucounts, UCOUNT_NET_NAMESPACES); -} - /* * setup_net runs the initializers for the network namespace object. */ @@ -320,6 +310,16 @@ out_undo: #ifdef CONFIG_NET_NS +static struct ucounts *inc_net_namespaces(struct user_namespace *ns) +{ + return inc_ucount(ns, current_euid(), UCOUNT_NET_NAMESPACES); +} + +static void dec_net_namespaces(struct ucounts *ucounts) +{ + dec_ucount(ucounts, UCOUNT_NET_NAMESPACES); +} + static struct kmem_cache *net_cachep; static struct workqueue_struct *netns_wq; -- cgit v1.2.3