From 038e7332b8d4c0629a2965e3ede1a92e8e427bd6 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Thu, 14 Jun 2012 02:31:10 -0700 Subject: userns: make each net (net_ns) belong to a user_ns The user namespace which creates a new network namespace owns that namespace and all resources created in it. This way we can target capability checks for privileged operations against network resources to the user_ns which created the network namespace in which the resource lives. Privilege to the user namespace which owns the network namespace, or any parent user namespace thereof, provides the same privilege to the network resource. This patch is reworked from a version originally by Serge E. Hallyn Acked-by: Serge Hallyn Signed-off-by: Eric W. Biederman --- kernel/nsproxy.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/nsproxy.c b/kernel/nsproxy.c index b576f7f14bc6..7e1c3de1ce45 100644 --- a/kernel/nsproxy.c +++ b/kernel/nsproxy.c @@ -90,7 +90,7 @@ static struct nsproxy *create_new_namespaces(unsigned long flags, goto out_pid; } - new_nsp->net_ns = copy_net_ns(flags, tsk->nsproxy->net_ns); + new_nsp->net_ns = copy_net_ns(flags, task_cred_xxx(tsk, user_ns), tsk->nsproxy->net_ns); if (IS_ERR(new_nsp->net_ns)) { err = PTR_ERR(new_nsp->net_ns); goto out_net; -- cgit v1.2.3 From 49f4d8b93ccf9454284b6f524b96c66d8d7fbccc Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Thu, 2 Aug 2012 04:25:10 -0700 Subject: pidns: Capture the user namespace and filter ns_last_pid - Capture the the user namespace that creates the pid namespace - Use that user namespace to test if it is ok to write to /proc/sys/kernel/ns_last_pid. Zhao Hongjiang noticed I was missing a put_user_ns in when destroying a pid_ns. I have foloded his patch into this one so that bisects will work properly. Acked-by: Serge Hallyn Signed-off-by: "Eric W. Biederman" --- kernel/nsproxy.c | 2 +- kernel/pid.c | 1 + kernel/pid_namespace.c | 17 ++++++++++++----- 3 files changed, 14 insertions(+), 6 deletions(-) (limited to 'kernel') diff --git a/kernel/nsproxy.c b/kernel/nsproxy.c index 7e1c3de1ce45..ca27d2c5264d 100644 --- a/kernel/nsproxy.c +++ b/kernel/nsproxy.c @@ -84,7 +84,7 @@ static struct nsproxy *create_new_namespaces(unsigned long flags, goto out_ipc; } - new_nsp->pid_ns = copy_pid_ns(flags, task_active_pid_ns(tsk)); + new_nsp->pid_ns = copy_pid_ns(flags, task_cred_xxx(tsk, user_ns), task_active_pid_ns(tsk)); if (IS_ERR(new_nsp->pid_ns)) { err = PTR_ERR(new_nsp->pid_ns); goto out_pid; diff --git a/kernel/pid.c b/kernel/pid.c index aebd4f5aaf41..2a624f1486e1 100644 --- a/kernel/pid.c +++ b/kernel/pid.c @@ -78,6 +78,7 @@ struct pid_namespace init_pid_ns = { .last_pid = 0, .level = 0, .child_reaper = &init_task, + .user_ns = &init_user_ns, }; EXPORT_SYMBOL_GPL(init_pid_ns); diff --git a/kernel/pid_namespace.c b/kernel/pid_namespace.c index 7b07cc0dfb75..b2604950aa50 100644 --- a/kernel/pid_namespace.c +++ b/kernel/pid_namespace.c @@ -10,6 +10,7 @@ #include #include +#include #include #include #include @@ -74,7 +75,8 @@ err_alloc: /* MAX_PID_NS_LEVEL is needed for limiting size of 'struct pid' */ #define MAX_PID_NS_LEVEL 32 -static struct pid_namespace *create_pid_namespace(struct pid_namespace *parent_pid_ns) +static struct pid_namespace *create_pid_namespace(struct user_namespace *user_ns, + struct pid_namespace *parent_pid_ns) { struct pid_namespace *ns; unsigned int level = parent_pid_ns->level + 1; @@ -102,6 +104,7 @@ static struct pid_namespace *create_pid_namespace(struct pid_namespace *parent_p kref_init(&ns->kref); ns->level = level; ns->parent = get_pid_ns(parent_pid_ns); + ns->user_ns = get_user_ns(user_ns); set_bit(0, ns->pidmap[0].page); atomic_set(&ns->pidmap[0].nr_free, BITS_PER_PAGE - 1); @@ -117,6 +120,7 @@ static struct pid_namespace *create_pid_namespace(struct pid_namespace *parent_p out_put_parent_pid_ns: put_pid_ns(parent_pid_ns); + put_user_ns(user_ns); out_free_map: kfree(ns->pidmap[0].page); out_free: @@ -131,16 +135,18 @@ static void destroy_pid_namespace(struct pid_namespace *ns) for (i = 0; i < PIDMAP_ENTRIES; i++) kfree(ns->pidmap[i].page); + put_user_ns(ns->user_ns); kmem_cache_free(pid_ns_cachep, ns); } -struct pid_namespace *copy_pid_ns(unsigned long flags, struct pid_namespace *old_ns) +struct pid_namespace *copy_pid_ns(unsigned long flags, + struct user_namespace *user_ns, struct pid_namespace *old_ns) { if (!(flags & CLONE_NEWPID)) return get_pid_ns(old_ns); if (flags & (CLONE_THREAD|CLONE_PARENT)) return ERR_PTR(-EINVAL); - return create_pid_namespace(old_ns); + return create_pid_namespace(user_ns, old_ns); } static void free_pid_ns(struct kref *kref) @@ -239,9 +245,10 @@ void zap_pid_ns_processes(struct pid_namespace *pid_ns) static int pid_ns_ctl_handler(struct ctl_table *table, int write, void __user *buffer, size_t *lenp, loff_t *ppos) { + struct pid_namespace *pid_ns = task_active_pid_ns(current); struct ctl_table tmp = *table; - if (write && !capable(CAP_SYS_ADMIN)) + if (write && !ns_capable(pid_ns->user_ns, CAP_SYS_ADMIN)) return -EPERM; /* @@ -250,7 +257,7 @@ static int pid_ns_ctl_handler(struct ctl_table *table, int write, * it should synchronize its usage with external means. */ - tmp.data = ¤t->nsproxy->pid_ns->last_pid; + tmp.data = &pid_ns->last_pid; return proc_dointvec_minmax(&tmp, write, buffer, lenp, ppos); } -- cgit v1.2.3 From 17cf22c33e1f1b5e435469c84e43872579497653 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Tue, 2 Mar 2010 14:51:53 -0800 Subject: pidns: Use task_active_pid_ns where appropriate The expressions tsk->nsproxy->pid_ns and task_active_pid_ns aka ns_of_pid(task_pid(tsk)) should have the same number of cache line misses with the practical difference that ns_of_pid(task_pid(tsk)) is released later in a processes life. Furthermore by using task_active_pid_ns it becomes trivial to write an unshare implementation for the the pid namespace. So I have used task_active_pid_ns everywhere I can. In fork since the pid has not yet been attached to the process I use ns_of_pid, to achieve the same effect. Signed-off-by: Eric W. Biederman --- kernel/cgroup.c | 2 +- kernel/events/core.c | 2 +- kernel/fork.c | 2 +- kernel/nsproxy.c | 2 +- kernel/pid.c | 8 ++++---- kernel/signal.c | 2 +- kernel/sysctl_binary.c | 2 +- 7 files changed, 10 insertions(+), 10 deletions(-) (limited to 'kernel') diff --git a/kernel/cgroup.c b/kernel/cgroup.c index f24f724620dd..0dbfba2efa77 100644 --- a/kernel/cgroup.c +++ b/kernel/cgroup.c @@ -3390,7 +3390,7 @@ static struct cgroup_pidlist *cgroup_pidlist_find(struct cgroup *cgrp, { struct cgroup_pidlist *l; /* don't need task_nsproxy() if we're looking at ourself */ - struct pid_namespace *ns = current->nsproxy->pid_ns; + struct pid_namespace *ns = task_active_pid_ns(current); /* * We can't drop the pidlist_mutex before taking the l->mutex in case diff --git a/kernel/events/core.c b/kernel/events/core.c index dbccf83c134d..738f3564e83b 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -6155,7 +6155,7 @@ perf_event_alloc(struct perf_event_attr *attr, int cpu, event->parent = parent_event; - event->ns = get_pid_ns(current->nsproxy->pid_ns); + event->ns = get_pid_ns(task_active_pid_ns(current)); event->id = atomic64_inc_return(&perf_event_id); event->state = PERF_EVENT_STATE_INACTIVE; diff --git a/kernel/fork.c b/kernel/fork.c index 8b20ab7d3aa2..7798c247f4b9 100644 --- a/kernel/fork.c +++ b/kernel/fork.c @@ -1442,7 +1442,7 @@ static struct task_struct *copy_process(unsigned long clone_flags, if (thread_group_leader(p)) { if (is_child_reaper(pid)) - p->nsproxy->pid_ns->child_reaper = p; + ns_of_pid(pid)->child_reaper = p; p->signal->leader_pid = pid; p->signal->tty = tty_kref_get(current->signal->tty); diff --git a/kernel/nsproxy.c b/kernel/nsproxy.c index ca27d2c5264d..acc92680381a 100644 --- a/kernel/nsproxy.c +++ b/kernel/nsproxy.c @@ -84,7 +84,7 @@ static struct nsproxy *create_new_namespaces(unsigned long flags, goto out_ipc; } - new_nsp->pid_ns = copy_pid_ns(flags, task_cred_xxx(tsk, user_ns), task_active_pid_ns(tsk)); + new_nsp->pid_ns = copy_pid_ns(flags, task_cred_xxx(tsk, user_ns), tsk->nsproxy->pid_ns); if (IS_ERR(new_nsp->pid_ns)) { err = PTR_ERR(new_nsp->pid_ns); goto out_pid; diff --git a/kernel/pid.c b/kernel/pid.c index 2a624f1486e1..3a5f238c1ca0 100644 --- a/kernel/pid.c +++ b/kernel/pid.c @@ -345,7 +345,7 @@ EXPORT_SYMBOL_GPL(find_pid_ns); struct pid *find_vpid(int nr) { - return find_pid_ns(nr, current->nsproxy->pid_ns); + return find_pid_ns(nr, task_active_pid_ns(current)); } EXPORT_SYMBOL_GPL(find_vpid); @@ -429,7 +429,7 @@ struct task_struct *find_task_by_pid_ns(pid_t nr, struct pid_namespace *ns) struct task_struct *find_task_by_vpid(pid_t vnr) { - return find_task_by_pid_ns(vnr, current->nsproxy->pid_ns); + return find_task_by_pid_ns(vnr, task_active_pid_ns(current)); } struct pid *get_task_pid(struct task_struct *task, enum pid_type type) @@ -484,7 +484,7 @@ EXPORT_SYMBOL_GPL(pid_nr_ns); pid_t pid_vnr(struct pid *pid) { - return pid_nr_ns(pid, current->nsproxy->pid_ns); + return pid_nr_ns(pid, task_active_pid_ns(current)); } EXPORT_SYMBOL_GPL(pid_vnr); @@ -495,7 +495,7 @@ pid_t __task_pid_nr_ns(struct task_struct *task, enum pid_type type, rcu_read_lock(); if (!ns) - ns = current->nsproxy->pid_ns; + ns = task_active_pid_ns(current); if (likely(pid_alive(task))) { if (type != PIDTYPE_PID) task = task->group_leader; diff --git a/kernel/signal.c b/kernel/signal.c index 0af8868525d6..b2445d86f226 100644 --- a/kernel/signal.c +++ b/kernel/signal.c @@ -1752,7 +1752,7 @@ static void do_notify_parent_cldstop(struct task_struct *tsk, * see comment in do_notify_parent() about the following 4 lines */ rcu_read_lock(); - info.si_pid = task_pid_nr_ns(tsk, parent->nsproxy->pid_ns); + info.si_pid = task_pid_nr_ns(tsk, task_active_pid_ns(parent)); info.si_uid = from_kuid_munged(task_cred_xxx(parent, user_ns), task_uid(tsk)); rcu_read_unlock(); diff --git a/kernel/sysctl_binary.c b/kernel/sysctl_binary.c index 65bdcf198d4e..5a6384450501 100644 --- a/kernel/sysctl_binary.c +++ b/kernel/sysctl_binary.c @@ -1344,7 +1344,7 @@ static ssize_t binary_sysctl(const int *name, int nlen, goto out_putname; } - mnt = current->nsproxy->pid_ns->proc_mnt; + mnt = task_active_pid_ns(current)->proc_mnt; file = file_open_root(mnt->mnt_root, mnt, pathname, flags); result = PTR_ERR(file); if (IS_ERR(file)) -- cgit v1.2.3 From 0a01f2cc390e10633a54f72c608cc3fe19a50c3d Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Wed, 1 Aug 2012 10:33:47 -0700 Subject: pidns: Make the pidns proc mount/umount logic obvious. Track the number of pids in the proc hash table. When the number of pids goes to 0 schedule work to unmount the kernel mount of proc. Move the mount of proc into alloc_pid when we allocate the pid for init. Remove the surprising calls of pid_ns_release proc in fork and proc_flush_task. Those code paths really shouldn't know about proc namespace implementation details and people have demonstrated several times that finding and understanding those code paths is difficult and non-obvious. Because of the call path detach pid is alwasy called with the rtnl_lock held free_pid is not allowed to sleep, so the work to unmounting proc is moved to a work queue. This has the side benefit of not blocking the entire world waiting for the unnecessary rcu_barrier in deactivate_locked_super. In the process of making the code clear and obvious this fixes a bug reported by Gao feng where we would leak a mount of proc during clone(CLONE_NEWPID|CLONE_NEWNET) if copy_pid_ns succeeded and copy_net_ns failed. Acked-by: "Serge E. Hallyn" Signed-off-by: "Eric W. Biederman" --- kernel/fork.c | 2 -- kernel/pid.c | 21 +++++++++++++++++---- kernel/pid_namespace.c | 14 +++++++------- 3 files changed, 24 insertions(+), 13 deletions(-) (limited to 'kernel') diff --git a/kernel/fork.c b/kernel/fork.c index 7798c247f4b9..666dc8b06606 100644 --- a/kernel/fork.c +++ b/kernel/fork.c @@ -1476,8 +1476,6 @@ bad_fork_cleanup_io: if (p->io_context) exit_io_context(p); bad_fork_cleanup_namespaces: - if (unlikely(clone_flags & CLONE_NEWPID)) - pid_ns_release_proc(p->nsproxy->pid_ns); exit_task_namespaces(p); bad_fork_cleanup_mm: if (p->mm) diff --git a/kernel/pid.c b/kernel/pid.c index 3a5f238c1ca0..e957f8b09136 100644 --- a/kernel/pid.c +++ b/kernel/pid.c @@ -36,6 +36,7 @@ #include #include #include +#include #define pid_hashfn(nr, ns) \ hash_long((unsigned long)nr + (unsigned long)ns, pidhash_shift) @@ -270,8 +271,12 @@ void free_pid(struct pid *pid) unsigned long flags; spin_lock_irqsave(&pidmap_lock, flags); - for (i = 0; i <= pid->level; i++) - hlist_del_rcu(&pid->numbers[i].pid_chain); + for (i = 0; i <= pid->level; i++) { + struct upid *upid = pid->numbers + i; + hlist_del_rcu(&upid->pid_chain); + if (--upid->ns->nr_hashed == 0) + schedule_work(&upid->ns->proc_work); + } spin_unlock_irqrestore(&pidmap_lock, flags); for (i = 0; i <= pid->level; i++) @@ -293,6 +298,7 @@ struct pid *alloc_pid(struct pid_namespace *ns) goto out; tmp = ns; + pid->level = ns->level; for (i = ns->level; i >= 0; i--) { nr = alloc_pidmap(tmp); if (nr < 0) @@ -303,17 +309,23 @@ struct pid *alloc_pid(struct pid_namespace *ns) tmp = tmp->parent; } + if (unlikely(is_child_reaper(pid))) { + if (pid_ns_prepare_proc(ns)) + goto out_free; + } + get_pid_ns(ns); - pid->level = ns->level; atomic_set(&pid->count, 1); for (type = 0; type < PIDTYPE_MAX; ++type) INIT_HLIST_HEAD(&pid->tasks[type]); upid = pid->numbers + ns->level; spin_lock_irq(&pidmap_lock); - for ( ; upid >= pid->numbers; --upid) + for ( ; upid >= pid->numbers; --upid) { hlist_add_head_rcu(&upid->pid_chain, &pid_hash[pid_hashfn(upid->nr, upid->ns)]); + upid->ns->nr_hashed++; + } spin_unlock_irq(&pidmap_lock); out: @@ -570,6 +582,7 @@ void __init pidmap_init(void) /* Reserve PID 0. We never call free_pidmap(0) */ set_bit(0, init_pid_ns.pidmap[0].page); atomic_dec(&init_pid_ns.pidmap[0].nr_free); + init_pid_ns.nr_hashed = 1; init_pid_ns.pid_cachep = KMEM_CACHE(pid, SLAB_HWCACHE_ALIGN | SLAB_PANIC); diff --git a/kernel/pid_namespace.c b/kernel/pid_namespace.c index b2604950aa50..84591cfeefc1 100644 --- a/kernel/pid_namespace.c +++ b/kernel/pid_namespace.c @@ -72,6 +72,12 @@ err_alloc: return NULL; } +static void proc_cleanup_work(struct work_struct *work) +{ + struct pid_namespace *ns = container_of(work, struct pid_namespace, proc_work); + pid_ns_release_proc(ns); +} + /* MAX_PID_NS_LEVEL is needed for limiting size of 'struct pid' */ #define MAX_PID_NS_LEVEL 32 @@ -105,6 +111,7 @@ static struct pid_namespace *create_pid_namespace(struct user_namespace *user_ns ns->level = level; ns->parent = get_pid_ns(parent_pid_ns); ns->user_ns = get_user_ns(user_ns); + INIT_WORK(&ns->proc_work, proc_cleanup_work); set_bit(0, ns->pidmap[0].page); atomic_set(&ns->pidmap[0].nr_free, BITS_PER_PAGE - 1); @@ -112,15 +119,8 @@ static struct pid_namespace *create_pid_namespace(struct user_namespace *user_ns for (i = 1; i < PIDMAP_ENTRIES; i++) atomic_set(&ns->pidmap[i].nr_free, BITS_PER_PAGE); - err = pid_ns_prepare_proc(ns); - if (err) - goto out_put_parent_pid_ns; - return ns; -out_put_parent_pid_ns: - put_pid_ns(parent_pid_ns); - put_user_ns(user_ns); out_free_map: kfree(ns->pidmap[0].page); out_free: -- cgit v1.2.3 From 5e1182deb81ae8c68494017c4a8a71811659c870 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Mon, 12 Jul 2010 18:50:25 -0700 Subject: pidns: Don't allow new processes in a dead pid namespace. Set nr_hashed to -1 just before we schedule the work to cleanup proc. Test nr_hashed just before we hash a new pid and if nr_hashed is < 0 fail. This guaranteees that processes never enter a pid namespaces after we have cleaned up the state to support processes in a pid namespace. Currently sending SIGKILL to all of the process in a pid namespace as init exists gives us this guarantee but we need something a little stronger to support unsharing and joining a pid namespace. Acked-by: "Serge E. Hallyn" Signed-off-by: Eric W. Biederman --- kernel/pid.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/pid.c b/kernel/pid.c index e957f8b09136..9c219117af36 100644 --- a/kernel/pid.c +++ b/kernel/pid.c @@ -274,8 +274,10 @@ void free_pid(struct pid *pid) for (i = 0; i <= pid->level; i++) { struct upid *upid = pid->numbers + i; hlist_del_rcu(&upid->pid_chain); - if (--upid->ns->nr_hashed == 0) + if (--upid->ns->nr_hashed == 0) { + upid->ns->nr_hashed = -1; schedule_work(&upid->ns->proc_work); + } } spin_unlock_irqrestore(&pidmap_lock, flags); @@ -321,6 +323,8 @@ struct pid *alloc_pid(struct pid_namespace *ns) upid = pid->numbers + ns->level; spin_lock_irq(&pidmap_lock); + if (ns->nr_hashed < 0) + goto out_unlock; for ( ; upid >= pid->numbers; --upid) { hlist_add_head_rcu(&upid->pid_chain, &pid_hash[pid_hashfn(upid->nr, upid->ns)]); @@ -331,6 +335,8 @@ struct pid *alloc_pid(struct pid_namespace *ns) out: return pid; +out_unlock: + spin_unlock(&pidmap_lock); out_free: while (++i <= ns->level) free_pidmap(pid->numbers + i); -- cgit v1.2.3 From af4b8a83add95ef40716401395b44a1b579965f4 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Wed, 1 Aug 2012 15:03:42 -0700 Subject: pidns: Wait in zap_pid_ns_processes until pid_ns->nr_hashed == 1 Looking at pid_ns->nr_hashed is a bit simpler and it works for disjoint process trees that an unshare or a join of a pid_namespace may create. Acked-by: "Serge E. Hallyn" Signed-off-by: "Eric W. Biederman" --- kernel/exit.c | 12 ------------ kernel/pid.c | 16 +++++++++++++--- kernel/pid_namespace.c | 15 ++++----------- 3 files changed, 17 insertions(+), 26 deletions(-) (limited to 'kernel') diff --git a/kernel/exit.c b/kernel/exit.c index 346616c0092c..d7fe58db4527 100644 --- a/kernel/exit.c +++ b/kernel/exit.c @@ -72,18 +72,6 @@ static void __unhash_process(struct task_struct *p, bool group_dead) list_del_rcu(&p->tasks); list_del_init(&p->sibling); __this_cpu_dec(process_counts); - /* - * If we are the last child process in a pid namespace to be - * reaped, notify the reaper sleeping zap_pid_ns_processes(). - */ - if (IS_ENABLED(CONFIG_PID_NS)) { - struct task_struct *parent = p->real_parent; - - if ((task_active_pid_ns(parent)->child_reaper == parent) && - list_empty(&parent->children) && - (parent->flags & PF_EXITING)) - wake_up_process(parent); - } } list_del_rcu(&p->thread_group); } diff --git a/kernel/pid.c b/kernel/pid.c index 9c219117af36..6e8da291de49 100644 --- a/kernel/pid.c +++ b/kernel/pid.c @@ -273,10 +273,20 @@ void free_pid(struct pid *pid) spin_lock_irqsave(&pidmap_lock, flags); for (i = 0; i <= pid->level; i++) { struct upid *upid = pid->numbers + i; + struct pid_namespace *ns = upid->ns; hlist_del_rcu(&upid->pid_chain); - if (--upid->ns->nr_hashed == 0) { - upid->ns->nr_hashed = -1; - schedule_work(&upid->ns->proc_work); + switch(--ns->nr_hashed) { + case 1: + /* When all that is left in the pid namespace + * is the reaper wake up the reaper. The reaper + * may be sleeping in zap_pid_ns_processes(). + */ + wake_up_process(ns->child_reaper); + break; + case 0: + ns->nr_hashed = -1; + schedule_work(&ns->proc_work); + break; } } spin_unlock_irqrestore(&pidmap_lock, flags); diff --git a/kernel/pid_namespace.c b/kernel/pid_namespace.c index 84591cfeefc1..3cc29b830e9e 100644 --- a/kernel/pid_namespace.c +++ b/kernel/pid_namespace.c @@ -217,22 +217,15 @@ void zap_pid_ns_processes(struct pid_namespace *pid_ns) /* * sys_wait4() above can't reap the TASK_DEAD children. - * Make sure they all go away, see __unhash_process(). + * Make sure they all go away, see free_pid(). */ for (;;) { - bool need_wait = false; - - read_lock(&tasklist_lock); - if (!list_empty(¤t->children)) { - __set_current_state(TASK_UNINTERRUPTIBLE); - need_wait = true; - } - read_unlock(&tasklist_lock); - - if (!need_wait) + set_current_state(TASK_UNINTERRUPTIBLE); + if (pid_ns->nr_hashed == 1) break; schedule(); } + __set_current_state(TASK_RUNNING); if (pid_ns->reboot) current->signal->group_exit_code = pid_ns->reboot; -- cgit v1.2.3 From 225778d68d98e7cfe2579f8d8b2d7b76f8541b8b Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Thu, 2 Aug 2012 08:35:35 -0700 Subject: pidns: Deny strange cases when creating pid namespaces. task_active_pid_ns(current) != current->ns_proxy->pid_ns will soon be allowed to support unshare and setns. The definition of creating a child pid namespace when task_active_pid_ns(current) != current->ns_proxy->pid_ns could be that we create a child pid namespace of current->ns_proxy->pid_ns. However that leads to strange cases like trying to have a single process be init in multiple pid namespaces, which is racy and hard to think about. The definition of creating a child pid namespace when task_active_pid_ns(current) != current->ns_proxy->pid_ns could be that we create a child pid namespace of task_active_pid_ns(current). While that seems less racy it does not provide any utility. Therefore define the semantics of creating a child pid namespace when task_active_pid_ns(current) != current->ns_proxy->pid_ns to be that the pid namespace creation fails. That is easy to implement and easy to think about. Signed-off-by: "Eric W. Biederman" --- kernel/pid_namespace.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'kernel') diff --git a/kernel/pid_namespace.c b/kernel/pid_namespace.c index 3cc29b830e9e..0dbbc66b6ec6 100644 --- a/kernel/pid_namespace.c +++ b/kernel/pid_namespace.c @@ -146,6 +146,8 @@ struct pid_namespace *copy_pid_ns(unsigned long flags, return get_pid_ns(old_ns); if (flags & (CLONE_THREAD|CLONE_PARENT)) return ERR_PTR(-EINVAL); + if (task_active_pid_ns(current) != old_ns) + return ERR_PTR(-EINVAL); return create_pid_namespace(user_ns, old_ns); } -- cgit v1.2.3 From 57e8391d327609cbf12d843259c968b9e5c1838f Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Sun, 7 Mar 2010 18:17:03 -0800 Subject: pidns: Add setns support - Pid namespaces are designed to be inescapable so verify that the passed in pid namespace is a child of the currently active pid namespace or the currently active pid namespace itself. Allowing the currently active pid namespace is important so the effects of an earlier setns can be cancelled. Signed-off-by: Eric W. Biederman --- kernel/pid_namespace.c | 54 ++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 54 insertions(+) (limited to 'kernel') diff --git a/kernel/pid_namespace.c b/kernel/pid_namespace.c index 0dbbc66b6ec6..f78fc48c86bc 100644 --- a/kernel/pid_namespace.c +++ b/kernel/pid_namespace.c @@ -301,6 +301,60 @@ int reboot_pid_ns(struct pid_namespace *pid_ns, int cmd) return 0; } +static void *pidns_get(struct task_struct *task) +{ + struct pid_namespace *ns; + + rcu_read_lock(); + ns = get_pid_ns(task_active_pid_ns(task)); + rcu_read_unlock(); + + return ns; +} + +static void pidns_put(void *ns) +{ + put_pid_ns(ns); +} + +static int pidns_install(struct nsproxy *nsproxy, void *ns) +{ + struct pid_namespace *active = task_active_pid_ns(current); + struct pid_namespace *ancestor, *new = ns; + + if (!ns_capable(new->user_ns, CAP_SYS_ADMIN)) + return -EPERM; + + /* + * Only allow entering the current active pid namespace + * or a child of the current active pid namespace. + * + * This is required for fork to return a usable pid value and + * this maintains the property that processes and their + * children can not escape their current pid namespace. + */ + if (new->level < active->level) + return -EINVAL; + + ancestor = new; + while (ancestor->level > active->level) + ancestor = ancestor->parent; + if (ancestor != active) + return -EINVAL; + + put_pid_ns(nsproxy->pid_ns); + nsproxy->pid_ns = get_pid_ns(new); + return 0; +} + +const struct proc_ns_operations pidns_operations = { + .name = "pid", + .type = CLONE_NEWPID, + .get = pidns_get, + .put = pidns_put, + .install = pidns_install, +}; + static __init int pid_namespaces_init(void) { pid_ns_cachep = KMEM_CACHE(pid_namespace, SLAB_PANIC); -- cgit v1.2.3 From 1c4042c29bd2e85aac4110552ca8ade763762e84 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Mon, 12 Jul 2010 17:10:36 -0700 Subject: pidns: Consolidate initialzation of special init task state Instead of setting child_reaper and SIGNAL_UNKILLABLE one way for the system init process, and another way for pid namespace init processes test pid->nr == 1 and use the same code for both. For the global init this results in SIGNAL_UNKILLABLE being set much earlier in the initialization process. This is a small cleanup and it paves the way for allowing unshare and enter of the pid namespace as that path like our global init also will not set CLONE_NEWPID. Signed-off-by: Eric W. Biederman --- kernel/fork.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'kernel') diff --git a/kernel/fork.c b/kernel/fork.c index 666dc8b06606..0f2bbce311fc 100644 --- a/kernel/fork.c +++ b/kernel/fork.c @@ -1039,8 +1039,6 @@ static int copy_signal(unsigned long clone_flags, struct task_struct *tsk) atomic_set(&sig->live, 1); atomic_set(&sig->sigcnt, 1); init_waitqueue_head(&sig->wait_chldexit); - if (clone_flags & CLONE_NEWPID) - sig->flags |= SIGNAL_UNKILLABLE; sig->curr_target = tsk; init_sigpending(&sig->shared_pending); INIT_LIST_HEAD(&sig->posix_timers); @@ -1441,8 +1439,10 @@ static struct task_struct *copy_process(unsigned long clone_flags, ptrace_init_task(p, (clone_flags & CLONE_PTRACE) || trace); if (thread_group_leader(p)) { - if (is_child_reaper(pid)) + if (is_child_reaper(pid)) { ns_of_pid(pid)->child_reaper = p; + p->signal->flags |= SIGNAL_UNKILLABLE; + } p->signal->leader_pid = pid; p->signal->tty = tty_kref_get(current->signal->tty); -- cgit v1.2.3 From 50804fe3737ca6a5942fdc2057a18a8141d00141 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Tue, 2 Mar 2010 15:41:50 -0800 Subject: pidns: Support unsharing the pid namespace. Unsharing of the pid namespace unlike unsharing of other namespaces does not take affect immediately. Instead it affects the children created with fork and clone. The first of these children becomes the init process of the new pid namespace, the rest become oddball children of pid 0. From the point of view of the new pid namespace the process that created it is pid 0, as it's pid does not map. A couple of different semantics were considered but this one was settled on because it is easy to implement and it is usable from pam modules. The core reasons for the existence of unshare. I took a survey of the callers of pam modules and the following appears to be a representative sample of their logic. { setup stuff include pam child = fork(); if (!child) { setuid() exec /bin/bash } waitpid(child); pam and other cleanup } As you can see there is a fork to create the unprivileged user space process. Which means that the unprivileged user space process will appear as pid 1 in the new pid namespace. Further most login processes do not cope with extraneous children which means shifting the duty of reaping extraneous child process to the creator of those extraneous children makes the system more comprehensible. The practical reason for this set of pid namespace semantics is that it is simple to implement and verify they work correctly. Whereas an implementation that requres changing the struct pid on a process comes with a lot more races and pain. Not the least of which is that glibc caches getpid(). These semantics are implemented by having two notions of the pid namespace of a proces. There is task_active_pid_ns which is the pid namspace the process was created with and the pid namespace that all pids are presented to that process in. The task_active_pid_ns is stored in the struct pid of the task. Then there is the pid namespace that will be used for children that pid namespace is stored in task->nsproxy->pid_ns. Signed-off-by: Eric W. Biederman --- kernel/fork.c | 32 +++++++++++++++++++++++++------- kernel/nsproxy.c | 2 +- kernel/pid_namespace.c | 2 -- 3 files changed, 26 insertions(+), 10 deletions(-) (limited to 'kernel') diff --git a/kernel/fork.c b/kernel/fork.c index 0f2bbce311fc..811ffbad7889 100644 --- a/kernel/fork.c +++ b/kernel/fork.c @@ -1565,9 +1565,11 @@ long do_fork(unsigned long clone_flags, * Do some preliminary argument and permissions checking before we * actually start allocating stuff */ - if (clone_flags & CLONE_NEWUSER) { - if (clone_flags & CLONE_THREAD) + if (clone_flags & (CLONE_NEWUSER | CLONE_NEWPID)) { + if (clone_flags & (CLONE_THREAD|CLONE_PARENT)) return -EINVAL; + } + if (clone_flags & CLONE_NEWUSER) { /* hopefully this check will go away when userns support is * complete */ @@ -1692,7 +1694,8 @@ static int check_unshare_flags(unsigned long unshare_flags) { if (unshare_flags & ~(CLONE_THREAD|CLONE_FS|CLONE_NEWNS|CLONE_SIGHAND| CLONE_VM|CLONE_FILES|CLONE_SYSVSEM| - CLONE_NEWUTS|CLONE_NEWIPC|CLONE_NEWNET)) + CLONE_NEWUTS|CLONE_NEWIPC|CLONE_NEWNET| + CLONE_NEWPID)) return -EINVAL; /* * Not implemented, but pretend it works if there is nothing to @@ -1763,15 +1766,30 @@ SYSCALL_DEFINE1(unshare, unsigned long, unshare_flags) int do_sysvsem = 0; int err; - err = check_unshare_flags(unshare_flags); - if (err) - goto bad_unshare_out; - + /* + * If unsharing a pid namespace must also unshare the thread. + */ + if (unshare_flags & CLONE_NEWPID) + unshare_flags |= CLONE_THREAD; + /* + * If unsharing a thread from a thread group, must also unshare vm. + */ + if (unshare_flags & CLONE_THREAD) + unshare_flags |= CLONE_VM; + /* + * If unsharing vm, must also unshare signal handlers. + */ + if (unshare_flags & CLONE_VM) + unshare_flags |= CLONE_SIGHAND; /* * If unsharing namespace, must also unshare filesystem information. */ if (unshare_flags & CLONE_NEWNS) unshare_flags |= CLONE_FS; + + err = check_unshare_flags(unshare_flags); + if (err) + goto bad_unshare_out; /* * CLONE_NEWIPC must also detach from the undolist: after switching * to a new ipc namespace, the semaphore arrays from the old diff --git a/kernel/nsproxy.c b/kernel/nsproxy.c index acc92680381a..b8d4d8709d70 100644 --- a/kernel/nsproxy.c +++ b/kernel/nsproxy.c @@ -188,7 +188,7 @@ int unshare_nsproxy_namespaces(unsigned long unshare_flags, int err = 0; if (!(unshare_flags & (CLONE_NEWNS | CLONE_NEWUTS | CLONE_NEWIPC | - CLONE_NEWNET))) + CLONE_NEWNET | CLONE_NEWPID))) return 0; if (!capable(CAP_SYS_ADMIN)) diff --git a/kernel/pid_namespace.c b/kernel/pid_namespace.c index f78fc48c86bc..68508d330634 100644 --- a/kernel/pid_namespace.c +++ b/kernel/pid_namespace.c @@ -144,8 +144,6 @@ struct pid_namespace *copy_pid_ns(unsigned long flags, { if (!(flags & CLONE_NEWPID)) return get_pid_ns(old_ns); - if (flags & (CLONE_THREAD|CLONE_PARENT)) - return ERR_PTR(-EINVAL); if (task_active_pid_ns(current) != old_ns) return ERR_PTR(-EINVAL); return create_pid_namespace(user_ns, old_ns); -- cgit v1.2.3 From 771b1371686e0a63e938ada28de020b9a0040f55 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Thu, 26 Jul 2012 21:08:32 -0700 Subject: vfs: Add a user namespace reference from struct mnt_namespace This will allow for support for unprivileged mounts in a new user namespace. Acked-by: "Serge E. Hallyn" Signed-off-by: "Eric W. Biederman" --- kernel/nsproxy.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/nsproxy.c b/kernel/nsproxy.c index b8d4d8709d70..7f8b051fc19f 100644 --- a/kernel/nsproxy.c +++ b/kernel/nsproxy.c @@ -66,7 +66,7 @@ static struct nsproxy *create_new_namespaces(unsigned long flags, if (!new_nsp) return ERR_PTR(-ENOMEM); - new_nsp->mnt_ns = copy_mnt_ns(flags, tsk->nsproxy->mnt_ns, new_fs); + new_nsp->mnt_ns = copy_mnt_ns(flags, tsk->nsproxy->mnt_ns, task_cred_xxx(tsk, user_ns), new_fs); if (IS_ERR(new_nsp->mnt_ns)) { err = PTR_ERR(new_nsp->mnt_ns); goto out_ns; -- cgit v1.2.3 From 5eaf563e53294d6696e651466697eb9d491f3946 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Mon, 21 Nov 2011 17:22:31 -0800 Subject: userns: Allow unprivileged users to create user namespaces. Now that we have been through every permission check in the kernel having uid == 0 and gid == 0 in your local user namespace no longer adds any special privileges. Even having a full set of caps in your local user namespace is safe because capabilies are relative to your local user namespace, and do not confer unexpected privileges. Over the long term this should allow much more of the kernels functionality to be safely used by non-root users. Functionality like unsharing the mount namespace that is only unsafe because it can fool applications whose privileges are raised when they are executed. Since those applications have no privileges in a user namespaces it becomes safe to spoof and confuse those applications all you want. Those capabilities will still need to be enabled carefully because we may still need things like rlimits on the number of unprivileged mounts but that is to avoid DOS attacks not to avoid fooling root owned processes. Acked-by: Serge Hallyn Signed-off-by: Eric W. Biederman --- kernel/fork.c | 8 -------- 1 file changed, 8 deletions(-) (limited to 'kernel') diff --git a/kernel/fork.c b/kernel/fork.c index 811ffbad7889..8c29abb19014 100644 --- a/kernel/fork.c +++ b/kernel/fork.c @@ -1569,14 +1569,6 @@ long do_fork(unsigned long clone_flags, if (clone_flags & (CLONE_THREAD|CLONE_PARENT)) return -EINVAL; } - if (clone_flags & CLONE_NEWUSER) { - /* hopefully this check will go away when userns support is - * complete - */ - if (!capable(CAP_SYS_ADMIN) || !capable(CAP_SETUID) || - !capable(CAP_SETGID)) - return -EPERM; - } /* * Determine whether and which event to report to ptracer. When -- cgit v1.2.3 From 37657da3c5d4a3bbbbb9d3b78f53a8134a0abae0 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Fri, 27 Jul 2012 06:21:27 -0700 Subject: userns: Allow setting a userns mapping to your current uid. Acked-by: Serge Hallyn Signed-off-by: "Eric W. Biederman" --- kernel/user_namespace.c | 15 +++++++++++++++ 1 file changed, 15 insertions(+) (limited to 'kernel') diff --git a/kernel/user_namespace.c b/kernel/user_namespace.c index 456a6b9fba34..49096d559e08 100644 --- a/kernel/user_namespace.c +++ b/kernel/user_namespace.c @@ -709,6 +709,21 @@ ssize_t proc_projid_map_write(struct file *file, const char __user *buf, size_t static bool new_idmap_permitted(struct user_namespace *ns, int cap_setid, struct uid_gid_map *new_map) { + /* Allow mapping to your own filesystem ids */ + if ((new_map->nr_extents == 1) && (new_map->extent[0].count == 1)) { + u32 id = new_map->extent[0].lower_first; + if (cap_setid == CAP_SETUID) { + kuid_t uid = make_kuid(ns->parent, id); + if (uid_eq(uid, current_fsuid())) + return true; + } + else if (cap_setid == CAP_SETGID) { + kgid_t gid = make_kgid(ns->parent, id); + if (gid_eq(gid, current_fsgid())) + return true; + } + } + /* Allow anyone to set a mapping that doesn't require privilege */ if (!cap_valid(cap_setid)) return true; -- cgit v1.2.3 From b33c77ef23dd3ec5692c9c0cc739a3f5f0f2baae Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Thu, 26 Jul 2012 00:50:47 -0700 Subject: userns: Allow unprivileged users to create new namespaces If an unprivileged user has the appropriate capabilities in their current user namespace allow the creation of new namespaces. Acked-by: Serge Hallyn Signed-off-by: "Eric W. Biederman" --- kernel/nsproxy.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/nsproxy.c b/kernel/nsproxy.c index 7f8b051fc19f..a214e0e9035f 100644 --- a/kernel/nsproxy.c +++ b/kernel/nsproxy.c @@ -122,6 +122,7 @@ out_ns: int copy_namespaces(unsigned long flags, struct task_struct *tsk) { struct nsproxy *old_ns = tsk->nsproxy; + struct user_namespace *user_ns = task_cred_xxx(tsk, user_ns); struct nsproxy *new_ns; int err = 0; @@ -134,7 +135,7 @@ int copy_namespaces(unsigned long flags, struct task_struct *tsk) CLONE_NEWPID | CLONE_NEWNET))) return 0; - if (!capable(CAP_SYS_ADMIN)) { + if (!ns_capable(user_ns, CAP_SYS_ADMIN)) { err = -EPERM; goto out; } @@ -191,7 +192,7 @@ int unshare_nsproxy_namespaces(unsigned long unshare_flags, CLONE_NEWNET | CLONE_NEWPID))) return 0; - if (!capable(CAP_SYS_ADMIN)) + if (!nsown_capable(CAP_SYS_ADMIN)) return -EPERM; *new_nsp = create_new_namespaces(unshare_flags, current, -- cgit v1.2.3 From 142e1d1d5f088e7a38659daca6e84a730967774a Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Thu, 26 Jul 2012 01:13:20 -0700 Subject: userns: Allow unprivileged use of setns. - Push the permission check from the core setns syscall into the setns install methods where the user namespace of the target namespace can be determined, and used in a ns_capable call. Acked-by: Serge Hallyn Signed-off-by: "Eric W. Biederman" --- kernel/nsproxy.c | 3 --- kernel/utsname.c | 7 ++++++- 2 files changed, 6 insertions(+), 4 deletions(-) (limited to 'kernel') diff --git a/kernel/nsproxy.c b/kernel/nsproxy.c index a214e0e9035f..4357a0a7d17d 100644 --- a/kernel/nsproxy.c +++ b/kernel/nsproxy.c @@ -242,9 +242,6 @@ SYSCALL_DEFINE2(setns, int, fd, int, nstype) struct file *file; int err; - if (!capable(CAP_SYS_ADMIN)) - return -EPERM; - file = proc_ns_fget(fd); if (IS_ERR(file)) return PTR_ERR(file); diff --git a/kernel/utsname.c b/kernel/utsname.c index 679d97a5d3fd..4a9362f9325d 100644 --- a/kernel/utsname.c +++ b/kernel/utsname.c @@ -102,8 +102,13 @@ static void utsns_put(void *ns) put_uts_ns(ns); } -static int utsns_install(struct nsproxy *nsproxy, void *ns) +static int utsns_install(struct nsproxy *nsproxy, void *new) { + struct uts_namespace *ns = new; + + if (!ns_capable(ns->user_ns, CAP_SYS_ADMIN)) + return -EPERM; + get_uts_ns(ns); put_uts_ns(nsproxy->uts_ns); nsproxy->uts_ns = ns; -- cgit v1.2.3 From bcf58e725ddc45d31addbc6627d4f0edccc824c1 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Thu, 26 Jul 2012 04:02:49 -0700 Subject: userns: Make create_new_namespaces take a user_ns parameter Modify create_new_namespaces to explicitly take a user namespace parameter, instead of implicitly through the task_struct. This allows an implementation of unshare(CLONE_NEWUSER) where the new user namespace is not stored onto the current task_struct until after all of the namespaces are created. Acked-by: Serge Hallyn Signed-off-by: "Eric W. Biederman" --- kernel/nsproxy.c | 22 +++++++++++++--------- kernel/utsname.c | 9 ++++----- 2 files changed, 17 insertions(+), 14 deletions(-) (limited to 'kernel') diff --git a/kernel/nsproxy.c b/kernel/nsproxy.c index 4357a0a7d17d..2ddd81657a2a 100644 --- a/kernel/nsproxy.c +++ b/kernel/nsproxy.c @@ -57,7 +57,8 @@ static inline struct nsproxy *create_nsproxy(void) * leave it to the caller to do proper locking and attach it to task. */ static struct nsproxy *create_new_namespaces(unsigned long flags, - struct task_struct *tsk, struct fs_struct *new_fs) + struct task_struct *tsk, struct user_namespace *user_ns, + struct fs_struct *new_fs) { struct nsproxy *new_nsp; int err; @@ -66,31 +67,31 @@ static struct nsproxy *create_new_namespaces(unsigned long flags, if (!new_nsp) return ERR_PTR(-ENOMEM); - new_nsp->mnt_ns = copy_mnt_ns(flags, tsk->nsproxy->mnt_ns, task_cred_xxx(tsk, user_ns), new_fs); + new_nsp->mnt_ns = copy_mnt_ns(flags, tsk->nsproxy->mnt_ns, user_ns, new_fs); if (IS_ERR(new_nsp->mnt_ns)) { err = PTR_ERR(new_nsp->mnt_ns); goto out_ns; } - new_nsp->uts_ns = copy_utsname(flags, tsk); + new_nsp->uts_ns = copy_utsname(flags, user_ns, tsk->nsproxy->uts_ns); if (IS_ERR(new_nsp->uts_ns)) { err = PTR_ERR(new_nsp->uts_ns); goto out_uts; } - new_nsp->ipc_ns = copy_ipcs(flags, tsk); + new_nsp->ipc_ns = copy_ipcs(flags, user_ns, tsk->nsproxy->ipc_ns); if (IS_ERR(new_nsp->ipc_ns)) { err = PTR_ERR(new_nsp->ipc_ns); goto out_ipc; } - new_nsp->pid_ns = copy_pid_ns(flags, task_cred_xxx(tsk, user_ns), tsk->nsproxy->pid_ns); + new_nsp->pid_ns = copy_pid_ns(flags, user_ns, tsk->nsproxy->pid_ns); if (IS_ERR(new_nsp->pid_ns)) { err = PTR_ERR(new_nsp->pid_ns); goto out_pid; } - new_nsp->net_ns = copy_net_ns(flags, task_cred_xxx(tsk, user_ns), tsk->nsproxy->net_ns); + new_nsp->net_ns = copy_net_ns(flags, user_ns, tsk->nsproxy->net_ns); if (IS_ERR(new_nsp->net_ns)) { err = PTR_ERR(new_nsp->net_ns); goto out_net; @@ -152,7 +153,8 @@ int copy_namespaces(unsigned long flags, struct task_struct *tsk) goto out; } - new_ns = create_new_namespaces(flags, tsk, tsk->fs); + new_ns = create_new_namespaces(flags, tsk, + task_cred_xxx(tsk, user_ns), tsk->fs); if (IS_ERR(new_ns)) { err = PTR_ERR(new_ns); goto out; @@ -186,6 +188,7 @@ void free_nsproxy(struct nsproxy *ns) int unshare_nsproxy_namespaces(unsigned long unshare_flags, struct nsproxy **new_nsp, struct fs_struct *new_fs) { + struct user_namespace *user_ns; int err = 0; if (!(unshare_flags & (CLONE_NEWNS | CLONE_NEWUTS | CLONE_NEWIPC | @@ -195,7 +198,8 @@ int unshare_nsproxy_namespaces(unsigned long unshare_flags, if (!nsown_capable(CAP_SYS_ADMIN)) return -EPERM; - *new_nsp = create_new_namespaces(unshare_flags, current, + user_ns = current_user_ns(); + *new_nsp = create_new_namespaces(unshare_flags, current, user_ns, new_fs ? new_fs : current->fs); if (IS_ERR(*new_nsp)) { err = PTR_ERR(*new_nsp); @@ -252,7 +256,7 @@ SYSCALL_DEFINE2(setns, int, fd, int, nstype) if (nstype && (ops->type != nstype)) goto out; - new_nsproxy = create_new_namespaces(0, tsk, tsk->fs); + new_nsproxy = create_new_namespaces(0, tsk, current_user_ns(), tsk->fs); if (IS_ERR(new_nsproxy)) { err = PTR_ERR(new_nsproxy); goto out; diff --git a/kernel/utsname.c b/kernel/utsname.c index 4a9362f9325d..fdc619eb61ef 100644 --- a/kernel/utsname.c +++ b/kernel/utsname.c @@ -32,7 +32,7 @@ static struct uts_namespace *create_uts_ns(void) * @old_ns: namespace to clone * Return NULL on error (failure to kmalloc), new ns otherwise */ -static struct uts_namespace *clone_uts_ns(struct task_struct *tsk, +static struct uts_namespace *clone_uts_ns(struct user_namespace *user_ns, struct uts_namespace *old_ns) { struct uts_namespace *ns; @@ -43,7 +43,7 @@ static struct uts_namespace *clone_uts_ns(struct task_struct *tsk, down_read(&uts_sem); memcpy(&ns->name, &old_ns->name, sizeof(ns->name)); - ns->user_ns = get_user_ns(task_cred_xxx(tsk, user_ns)); + ns->user_ns = get_user_ns(user_ns); up_read(&uts_sem); return ns; } @@ -55,9 +55,8 @@ static struct uts_namespace *clone_uts_ns(struct task_struct *tsk, * versa. */ struct uts_namespace *copy_utsname(unsigned long flags, - struct task_struct *tsk) + struct user_namespace *user_ns, struct uts_namespace *old_ns) { - struct uts_namespace *old_ns = tsk->nsproxy->uts_ns; struct uts_namespace *new_ns; BUG_ON(!old_ns); @@ -66,7 +65,7 @@ struct uts_namespace *copy_utsname(unsigned long flags, if (!(flags & CLONE_NEWUTS)) return old_ns; - new_ns = clone_uts_ns(tsk, old_ns); + new_ns = clone_uts_ns(user_ns, old_ns); put_uts_ns(old_ns); return new_ns; -- cgit v1.2.3 From 4c44aaafa8108f584831850ab48a975e971db2de Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Thu, 26 Jul 2012 05:05:21 -0700 Subject: userns: Kill task_user_ns The task_user_ns function hides the fact that it is getting the user namespace from struct cred on the task. struct cred may go away as soon as the rcu lock is released. This leads to a race where we can dereference a stale user namespace pointer. To make it obvious a struct cred is involved kill task_user_ns. To kill the race modify the users of task_user_ns to only reference the user namespace while the rcu lock is held. Cc: Kees Cook Cc: James Morris Acked-by: Kees Cook Acked-by: Serge Hallyn Signed-off-by: "Eric W. Biederman" --- kernel/ptrace.c | 10 ++++++++-- kernel/sched/core.c | 10 ++++++++-- 2 files changed, 16 insertions(+), 4 deletions(-) (limited to 'kernel') diff --git a/kernel/ptrace.c b/kernel/ptrace.c index 1f5e55dda955..7b09b88862cc 100644 --- a/kernel/ptrace.c +++ b/kernel/ptrace.c @@ -215,8 +215,12 @@ ok: smp_rmb(); if (task->mm) dumpable = get_dumpable(task->mm); - if (!dumpable && !ptrace_has_cap(task_user_ns(task), mode)) + rcu_read_lock(); + if (!dumpable && !ptrace_has_cap(__task_cred(task)->user_ns, mode)) { + rcu_read_unlock(); return -EPERM; + } + rcu_read_unlock(); return security_ptrace_access_check(task, mode); } @@ -280,8 +284,10 @@ static int ptrace_attach(struct task_struct *task, long request, if (seize) flags |= PT_SEIZED; - if (ns_capable(task_user_ns(task), CAP_SYS_PTRACE)) + rcu_read_lock(); + if (ns_capable(__task_cred(task)->user_ns, CAP_SYS_PTRACE)) flags |= PT_PTRACE_CAP; + rcu_read_unlock(); task->ptrace = flags; __ptrace_link(task, current); diff --git a/kernel/sched/core.c b/kernel/sched/core.c index 2d8927fda712..2f5eb1838b3e 100644 --- a/kernel/sched/core.c +++ b/kernel/sched/core.c @@ -4029,8 +4029,14 @@ long sched_setaffinity(pid_t pid, const struct cpumask *in_mask) goto out_free_cpus_allowed; } retval = -EPERM; - if (!check_same_owner(p) && !ns_capable(task_user_ns(p), CAP_SYS_NICE)) - goto out_unlock; + if (!check_same_owner(p)) { + rcu_read_lock(); + if (!ns_capable(__task_cred(p)->user_ns, CAP_SYS_NICE)) { + rcu_read_unlock(); + goto out_unlock; + } + rcu_read_unlock(); + } retval = security_task_setscheduler(p); if (retval) -- cgit v1.2.3 From cde1975bc242f3e1072bde623ef378e547b73f91 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Thu, 26 Jul 2012 06:24:06 -0700 Subject: userns: Implent proc namespace operations This allows entering a user namespace, and the ability to store a reference to a user namespace with a bind mount. Addition of missing userns_ns_put in userns_install from Gao feng Acked-by: Serge Hallyn Signed-off-by: "Eric W. Biederman" --- kernel/user_namespace.c | 90 +++++++++++++++++++++++++++++++++++++++---------- 1 file changed, 73 insertions(+), 17 deletions(-) (limited to 'kernel') diff --git a/kernel/user_namespace.c b/kernel/user_namespace.c index 49096d559e08..a9460774e77d 100644 --- a/kernel/user_namespace.c +++ b/kernel/user_namespace.c @@ -9,6 +9,7 @@ #include #include #include +#include #include #include #include @@ -26,6 +27,24 @@ static struct kmem_cache *user_ns_cachep __read_mostly; static bool new_idmap_permitted(struct user_namespace *ns, int cap_setid, struct uid_gid_map *map); +static void set_cred_user_ns(struct cred *cred, struct user_namespace *user_ns) +{ + /* Start with the same capabilities as init but useless for doing + * anything as the capabilities are bound to the new user namespace. + */ + cred->securebits = SECUREBITS_DEFAULT; + cred->cap_inheritable = CAP_EMPTY_SET; + cred->cap_permitted = CAP_FULL_SET; + cred->cap_effective = CAP_FULL_SET; + cred->cap_bset = CAP_FULL_SET; +#ifdef CONFIG_KEYS + key_put(cred->request_key_auth); + cred->request_key_auth = NULL; +#endif + /* tgcred will be cleared in our caller bc CLONE_THREAD won't be set */ + cred->user_ns = user_ns; +} + /* * Create a new user namespace, deriving the creator from the user in the * passed credentials, and replacing that user with the new root user for the @@ -53,27 +72,12 @@ int create_user_ns(struct cred *new) return -ENOMEM; kref_init(&ns->kref); + /* Leave the new->user_ns reference with the new user namespace. */ ns->parent = parent_ns; ns->owner = owner; ns->group = group; - /* Start with the same capabilities as init but useless for doing - * anything as the capabilities are bound to the new user namespace. - */ - new->securebits = SECUREBITS_DEFAULT; - new->cap_inheritable = CAP_EMPTY_SET; - new->cap_permitted = CAP_FULL_SET; - new->cap_effective = CAP_FULL_SET; - new->cap_bset = CAP_FULL_SET; -#ifdef CONFIG_KEYS - key_put(new->request_key_auth); - new->request_key_auth = NULL; -#endif - /* tgcred will be cleared in our caller bc CLONE_THREAD won't be set */ - - /* Leave the new->user_ns reference with the new user namespace. */ - /* Leave the reference to our user_ns with the new cred. */ - new->user_ns = ns; + set_cred_user_ns(new, ns); return 0; } @@ -737,6 +741,58 @@ static bool new_idmap_permitted(struct user_namespace *ns, int cap_setid, return false; } +static void *userns_get(struct task_struct *task) +{ + struct user_namespace *user_ns; + + rcu_read_lock(); + user_ns = get_user_ns(__task_cred(task)->user_ns); + rcu_read_unlock(); + + return user_ns; +} + +static void userns_put(void *ns) +{ + put_user_ns(ns); +} + +static int userns_install(struct nsproxy *nsproxy, void *ns) +{ + struct user_namespace *user_ns = ns; + struct cred *cred; + + /* Don't allow gaining capabilities by reentering + * the same user namespace. + */ + if (user_ns == current_user_ns()) + return -EINVAL; + + /* Threaded many not enter a different user namespace */ + if (atomic_read(¤t->mm->mm_users) > 1) + return -EINVAL; + + if (!ns_capable(user_ns, CAP_SYS_ADMIN)) + return -EPERM; + + cred = prepare_creds(); + if (!cred) + return -ENOMEM; + + put_user_ns(cred->user_ns); + set_cred_user_ns(cred, get_user_ns(user_ns)); + + return commit_creds(cred); +} + +const struct proc_ns_operations userns_operations = { + .name = "user", + .type = CLONE_NEWUSER, + .get = userns_get, + .put = userns_put, + .install = userns_install, +}; + static __init int user_namespaces_init(void) { user_ns_cachep = KMEM_CACHE(user_namespace, SLAB_PANIC); -- cgit v1.2.3 From b2e0d98705e60e45bbb3c0032c48824ad7ae0704 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Thu, 26 Jul 2012 05:15:35 -0700 Subject: userns: Implement unshare of the user namespace - Add CLONE_THREAD to the unshare flags if CLONE_NEWUSER is selected As changing user namespaces is only valid if all there is only a single thread. - Restore the code to add CLONE_VM if CLONE_THREAD is selected and the code to addCLONE_SIGHAND if CLONE_VM is selected. Making the constraints in the code clear. Acked-by: Serge Hallyn Signed-off-by: "Eric W. Biederman" --- kernel/fork.c | 25 ++++++++++++++++++++++--- kernel/nsproxy.c | 8 ++++---- kernel/user_namespace.c | 15 +++++++++++++++ 3 files changed, 41 insertions(+), 7 deletions(-) (limited to 'kernel') diff --git a/kernel/fork.c b/kernel/fork.c index 8c29abb19014..38e53b87402c 100644 --- a/kernel/fork.c +++ b/kernel/fork.c @@ -1687,7 +1687,7 @@ static int check_unshare_flags(unsigned long unshare_flags) if (unshare_flags & ~(CLONE_THREAD|CLONE_FS|CLONE_NEWNS|CLONE_SIGHAND| CLONE_VM|CLONE_FILES|CLONE_SYSVSEM| CLONE_NEWUTS|CLONE_NEWIPC|CLONE_NEWNET| - CLONE_NEWPID)) + CLONE_NEWUSER|CLONE_NEWPID)) return -EINVAL; /* * Not implemented, but pretend it works if there is nothing to @@ -1754,10 +1754,16 @@ SYSCALL_DEFINE1(unshare, unsigned long, unshare_flags) { struct fs_struct *fs, *new_fs = NULL; struct files_struct *fd, *new_fd = NULL; + struct cred *new_cred = NULL; struct nsproxy *new_nsproxy = NULL; int do_sysvsem = 0; int err; + /* + * If unsharing a user namespace must also unshare the thread. + */ + if (unshare_flags & CLONE_NEWUSER) + unshare_flags |= CLONE_THREAD; /* * If unsharing a pid namespace must also unshare the thread. */ @@ -1795,11 +1801,15 @@ SYSCALL_DEFINE1(unshare, unsigned long, unshare_flags) err = unshare_fd(unshare_flags, &new_fd); if (err) goto bad_unshare_cleanup_fs; - err = unshare_nsproxy_namespaces(unshare_flags, &new_nsproxy, new_fs); + err = unshare_userns(unshare_flags, &new_cred); if (err) goto bad_unshare_cleanup_fd; + err = unshare_nsproxy_namespaces(unshare_flags, &new_nsproxy, + new_cred, new_fs); + if (err) + goto bad_unshare_cleanup_cred; - if (new_fs || new_fd || do_sysvsem || new_nsproxy) { + if (new_fs || new_fd || do_sysvsem || new_cred || new_nsproxy) { if (do_sysvsem) { /* * CLONE_SYSVSEM is equivalent to sys_exit(). @@ -1832,11 +1842,20 @@ SYSCALL_DEFINE1(unshare, unsigned long, unshare_flags) } task_unlock(current); + + if (new_cred) { + /* Install the new user namespace */ + commit_creds(new_cred); + new_cred = NULL; + } } if (new_nsproxy) put_nsproxy(new_nsproxy); +bad_unshare_cleanup_cred: + if (new_cred) + put_cred(new_cred); bad_unshare_cleanup_fd: if (new_fd) put_files_struct(new_fd); diff --git a/kernel/nsproxy.c b/kernel/nsproxy.c index 2ddd81657a2a..78e2ecb20165 100644 --- a/kernel/nsproxy.c +++ b/kernel/nsproxy.c @@ -186,7 +186,7 @@ void free_nsproxy(struct nsproxy *ns) * On success, returns the new nsproxy. */ int unshare_nsproxy_namespaces(unsigned long unshare_flags, - struct nsproxy **new_nsp, struct fs_struct *new_fs) + struct nsproxy **new_nsp, struct cred *new_cred, struct fs_struct *new_fs) { struct user_namespace *user_ns; int err = 0; @@ -195,12 +195,12 @@ int unshare_nsproxy_namespaces(unsigned long unshare_flags, CLONE_NEWNET | CLONE_NEWPID))) return 0; - if (!nsown_capable(CAP_SYS_ADMIN)) + user_ns = new_cred ? new_cred->user_ns : current_user_ns(); + if (!ns_capable(user_ns, CAP_SYS_ADMIN)) return -EPERM; - user_ns = current_user_ns(); *new_nsp = create_new_namespaces(unshare_flags, current, user_ns, - new_fs ? new_fs : current->fs); + new_fs ? new_fs : current->fs); if (IS_ERR(*new_nsp)) { err = PTR_ERR(*new_nsp); goto out; diff --git a/kernel/user_namespace.c b/kernel/user_namespace.c index a9460774e77d..ce92f7e6290a 100644 --- a/kernel/user_namespace.c +++ b/kernel/user_namespace.c @@ -82,6 +82,21 @@ int create_user_ns(struct cred *new) return 0; } +int unshare_userns(unsigned long unshare_flags, struct cred **new_cred) +{ + struct cred *cred; + + if (!(unshare_flags & CLONE_NEWUSER)) + return 0; + + cred = prepare_creds(); + if (!cred) + return -ENOMEM; + + *new_cred = cred; + return create_user_ns(cred); +} + void free_user_ns(struct kref *kref) { struct user_namespace *parent, *ns = -- cgit v1.2.3 From c450f371d48557e3e0fa510a4af27b92f0d8c4cc Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Tue, 14 Aug 2012 21:25:13 -0700 Subject: userns: For /proc/self/{uid,gid}_map derive the lower userns from the struct file To keep things sane in the context of file descriptor passing derive the user namespace that uids are mapped into from the opener of the file instead of from current. When writing to the maps file the lower user namespace must always be the parent user namespace, or setting the mapping simply does not make sense. Enforce that the opener of the file was in the parent user namespace or the user namespace whose mapping is being set. Acked-by: Serge E. Hallyn Signed-off-by: "Eric W. Biederman" --- kernel/user_namespace.c | 12 ++++++++++-- 1 file changed, 10 insertions(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/user_namespace.c b/kernel/user_namespace.c index ce92f7e6290a..89f6eaed067a 100644 --- a/kernel/user_namespace.c +++ b/kernel/user_namespace.c @@ -391,7 +391,7 @@ static int uid_m_show(struct seq_file *seq, void *v) struct user_namespace *lower_ns; uid_t lower; - lower_ns = current_user_ns(); + lower_ns = seq_user_ns(seq); if ((lower_ns == ns) && lower_ns->parent) lower_ns = lower_ns->parent; @@ -412,7 +412,7 @@ static int gid_m_show(struct seq_file *seq, void *v) struct user_namespace *lower_ns; gid_t lower; - lower_ns = current_user_ns(); + lower_ns = seq_user_ns(seq); if ((lower_ns == ns) && lower_ns->parent) lower_ns = lower_ns->parent; @@ -688,10 +688,14 @@ ssize_t proc_uid_map_write(struct file *file, const char __user *buf, size_t siz { struct seq_file *seq = file->private_data; struct user_namespace *ns = seq->private; + struct user_namespace *seq_ns = seq_user_ns(seq); if (!ns->parent) return -EPERM; + if ((seq_ns != ns) && (seq_ns != ns->parent)) + return -EPERM; + return map_write(file, buf, size, ppos, CAP_SETUID, &ns->uid_map, &ns->parent->uid_map); } @@ -700,10 +704,14 @@ ssize_t proc_gid_map_write(struct file *file, const char __user *buf, size_t siz { struct seq_file *seq = file->private_data; struct user_namespace *ns = seq->private; + struct user_namespace *seq_ns = seq_user_ns(seq); if (!ns->parent) return -EPERM; + if ((seq_ns != ns) && (seq_ns != ns->parent)) + return -EPERM; + return map_write(file, buf, size, ppos, CAP_SETGID, &ns->gid_map, &ns->parent->gid_map); } -- cgit v1.2.3 From 98f842e675f96ffac96e6c50315790912b2812be Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Wed, 15 Jun 2011 10:21:48 -0700 Subject: proc: Usable inode numbers for the namespace file descriptors. Assign a unique proc inode to each namespace, and use that inode number to ensure we only allocate at most one proc inode for every namespace in proc. A single proc inode per namespace allows userspace to test to see if two processes are in the same namespace. This has been a long requested feature and only blocked because a naive implementation would put the id in a global space and would ultimately require having a namespace for the names of namespaces, making migration and certain virtualization tricks impossible. We still don't have per superblock inode numbers for proc, which appears necessary for application unaware checkpoint/restart and migrations (if the application is using namespace file descriptors) but that is now allowd by the design if it becomes important. I have preallocated the ipc and uts initial proc inode numbers so their structures can be statically initialized. Signed-off-by: Eric W. Biederman --- kernel/pid.c | 1 + kernel/pid_namespace.c | 12 ++++++++++++ kernel/user.c | 2 ++ kernel/user_namespace.c | 15 +++++++++++++++ kernel/utsname.c | 17 ++++++++++++++++- 5 files changed, 46 insertions(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/pid.c b/kernel/pid.c index 6e8da291de49..3026ddae0a34 100644 --- a/kernel/pid.c +++ b/kernel/pid.c @@ -80,6 +80,7 @@ struct pid_namespace init_pid_ns = { .level = 0, .child_reaper = &init_task, .user_ns = &init_user_ns, + .proc_inum = PROC_PID_INIT_INO, }; EXPORT_SYMBOL_GPL(init_pid_ns); diff --git a/kernel/pid_namespace.c b/kernel/pid_namespace.c index 68508d330634..560da0dab230 100644 --- a/kernel/pid_namespace.c +++ b/kernel/pid_namespace.c @@ -107,6 +107,10 @@ static struct pid_namespace *create_pid_namespace(struct user_namespace *user_ns if (ns->pid_cachep == NULL) goto out_free_map; + err = proc_alloc_inum(&ns->proc_inum); + if (err) + goto out_free_map; + kref_init(&ns->kref); ns->level = level; ns->parent = get_pid_ns(parent_pid_ns); @@ -133,6 +137,7 @@ static void destroy_pid_namespace(struct pid_namespace *ns) { int i; + proc_free_inum(ns->proc_inum); for (i = 0; i < PIDMAP_ENTRIES; i++) kfree(ns->pidmap[i].page); put_user_ns(ns->user_ns); @@ -345,12 +350,19 @@ static int pidns_install(struct nsproxy *nsproxy, void *ns) return 0; } +static unsigned int pidns_inum(void *ns) +{ + struct pid_namespace *pid_ns = ns; + return pid_ns->proc_inum; +} + const struct proc_ns_operations pidns_operations = { .name = "pid", .type = CLONE_NEWPID, .get = pidns_get, .put = pidns_put, .install = pidns_install, + .inum = pidns_inum, }; static __init int pid_namespaces_init(void) diff --git a/kernel/user.c b/kernel/user.c index 750acffbe9ec..33acb5e53a5f 100644 --- a/kernel/user.c +++ b/kernel/user.c @@ -16,6 +16,7 @@ #include #include #include +#include /* * userns count is 1 for root user, 1 for init_uts_ns, @@ -51,6 +52,7 @@ struct user_namespace init_user_ns = { }, .owner = GLOBAL_ROOT_UID, .group = GLOBAL_ROOT_GID, + .proc_inum = PROC_USER_INIT_INO, }; EXPORT_SYMBOL_GPL(init_user_ns); diff --git a/kernel/user_namespace.c b/kernel/user_namespace.c index 89f6eaed067a..f5975ccf9348 100644 --- a/kernel/user_namespace.c +++ b/kernel/user_namespace.c @@ -58,6 +58,7 @@ int create_user_ns(struct cred *new) struct user_namespace *ns, *parent_ns = new->user_ns; kuid_t owner = new->euid; kgid_t group = new->egid; + int ret; /* The creator needs a mapping in the parent user namespace * or else we won't be able to reasonably tell userspace who @@ -71,6 +72,12 @@ int create_user_ns(struct cred *new) if (!ns) return -ENOMEM; + ret = proc_alloc_inum(&ns->proc_inum); + if (ret) { + kmem_cache_free(user_ns_cachep, ns); + return ret; + } + kref_init(&ns->kref); /* Leave the new->user_ns reference with the new user namespace. */ ns->parent = parent_ns; @@ -103,6 +110,7 @@ void free_user_ns(struct kref *kref) container_of(kref, struct user_namespace, kref); parent = ns->parent; + proc_free_inum(ns->proc_inum); kmem_cache_free(user_ns_cachep, ns); put_user_ns(parent); } @@ -808,12 +816,19 @@ static int userns_install(struct nsproxy *nsproxy, void *ns) return commit_creds(cred); } +static unsigned int userns_inum(void *ns) +{ + struct user_namespace *user_ns = ns; + return user_ns->proc_inum; +} + const struct proc_ns_operations userns_operations = { .name = "user", .type = CLONE_NEWUSER, .get = userns_get, .put = userns_put, .install = userns_install, + .inum = userns_inum, }; static __init int user_namespaces_init(void) diff --git a/kernel/utsname.c b/kernel/utsname.c index fdc619eb61ef..f6336d51d64c 100644 --- a/kernel/utsname.c +++ b/kernel/utsname.c @@ -36,11 +36,18 @@ static struct uts_namespace *clone_uts_ns(struct user_namespace *user_ns, struct uts_namespace *old_ns) { struct uts_namespace *ns; + int err; ns = create_uts_ns(); if (!ns) return ERR_PTR(-ENOMEM); + err = proc_alloc_inum(&ns->proc_inum); + if (err) { + kfree(ns); + return ERR_PTR(err); + } + down_read(&uts_sem); memcpy(&ns->name, &old_ns->name, sizeof(ns->name)); ns->user_ns = get_user_ns(user_ns); @@ -77,6 +84,7 @@ void free_uts_ns(struct kref *kref) ns = container_of(kref, struct uts_namespace, kref); put_user_ns(ns->user_ns); + proc_free_inum(ns->proc_inum); kfree(ns); } @@ -114,11 +122,18 @@ static int utsns_install(struct nsproxy *nsproxy, void *new) return 0; } +static unsigned int utsns_inum(void *vp) +{ + struct uts_namespace *ns = vp; + + return ns->proc_inum; +} + const struct proc_ns_operations utsns_operations = { .name = "uts", .type = CLONE_NEWUTS, .get = utsns_get, .put = utsns_put, .install = utsns_install, + .inum = utsns_inum, }; - -- cgit v1.2.3