Blame view
kernel/pid_namespace.c
11.4 KB
74bd59bb3
|
1 2 3 4 5 6 7 8 9 10 11 12 |
/* * Pid namespaces * * Authors: * (C) 2007 Pavel Emelyanov <xemul@openvz.org>, OpenVZ, SWsoft Inc. * (C) 2007 Sukadev Bhattiprolu <sukadev@us.ibm.com>, IBM * Many thanks to Oleg Nesterov for comments and help * */ #include <linux/pid.h> #include <linux/pid_namespace.h> |
49f4d8b93
|
13 |
#include <linux/user_namespace.h> |
74bd59bb3
|
14 |
#include <linux/syscalls.h> |
5b825c3af
|
15 |
#include <linux/cred.h> |
74bd59bb3
|
16 |
#include <linux/err.h> |
0b6b030fc
|
17 |
#include <linux/acct.h> |
5a0e3ad6a
|
18 |
#include <linux/slab.h> |
0bb80f240
|
19 |
#include <linux/proc_ns.h> |
cf3f89214
|
20 |
#include <linux/reboot.h> |
523a6a945
|
21 |
#include <linux/export.h> |
299300258
|
22 |
#include <linux/sched/task.h> |
f361bf4a6
|
23 |
#include <linux/sched/signal.h> |
95846ecf9
|
24 |
#include <linux/idr.h> |
74bd59bb3
|
25 |
|
74bd59bb3
|
26 27 |
static DEFINE_MUTEX(pid_caches_mutex); static struct kmem_cache *pid_ns_cachep; |
dd206bec9
|
28 29 30 31 |
/* MAX_PID_NS_LEVEL is needed for limiting size of 'struct pid' */ #define MAX_PID_NS_LEVEL 32 /* Write once array, filled from the beginning. */ static struct kmem_cache *pid_cache[MAX_PID_NS_LEVEL]; |
74bd59bb3
|
32 33 34 |
/* * creates the kmem cache to allocate pids from. |
dd206bec9
|
35 |
* @level: pid namespace level |
74bd59bb3
|
36 |
*/ |
dd206bec9
|
37 |
static struct kmem_cache *create_pid_cachep(unsigned int level) |
74bd59bb3
|
38 |
{ |
dd206bec9
|
39 40 41 42 43 44 45 46 47 48 49 50 |
/* Level 0 is init_pid_ns.pid_cachep */ struct kmem_cache **pkc = &pid_cache[level - 1]; struct kmem_cache *kc; char name[4 + 10 + 1]; unsigned int len; kc = READ_ONCE(*pkc); if (kc) return kc; snprintf(name, sizeof(name), "pid_%u", level + 1); len = sizeof(struct pid) + level * sizeof(struct upid); |
74bd59bb3
|
51 |
mutex_lock(&pid_caches_mutex); |
dd206bec9
|
52 53 54 |
/* Name collision forces to do allocation under mutex. */ if (!*pkc) *pkc = kmem_cache_create(name, len, 0, SLAB_HWCACHE_ALIGN, 0); |
74bd59bb3
|
55 |
mutex_unlock(&pid_caches_mutex); |
dd206bec9
|
56 57 |
/* current can fail, but someone else can succeed. */ return READ_ONCE(*pkc); |
74bd59bb3
|
58 |
} |
0a01f2cc3
|
59 60 61 62 63 |
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); } |
f333c700c
|
64 65 66 67 68 69 70 71 72 |
static struct ucounts *inc_pid_namespaces(struct user_namespace *ns) { return inc_ucount(ns, current_euid(), UCOUNT_PID_NAMESPACES); } static void dec_pid_namespaces(struct ucounts *ucounts) { dec_ucount(ucounts, UCOUNT_PID_NAMESPACES); } |
49f4d8b93
|
73 74 |
static struct pid_namespace *create_pid_namespace(struct user_namespace *user_ns, struct pid_namespace *parent_pid_ns) |
74bd59bb3
|
75 76 |
{ struct pid_namespace *ns; |
ed469a63c
|
77 |
unsigned int level = parent_pid_ns->level + 1; |
f333c700c
|
78 |
struct ucounts *ucounts; |
f23025057
|
79 |
int err; |
a2b426267
|
80 81 82 |
err = -EINVAL; if (!in_userns(parent_pid_ns->user_ns, user_ns)) goto out; |
df75e7748
|
83 |
err = -ENOSPC; |
f333c700c
|
84 85 86 87 |
if (level > MAX_PID_NS_LEVEL) goto out; ucounts = inc_pid_namespaces(user_ns); if (!ucounts) |
f23025057
|
88 |
goto out; |
74bd59bb3
|
89 |
|
f23025057
|
90 |
err = -ENOMEM; |
84406c153
|
91 |
ns = kmem_cache_zalloc(pid_ns_cachep, GFP_KERNEL); |
74bd59bb3
|
92 |
if (ns == NULL) |
f333c700c
|
93 |
goto out_dec; |
74bd59bb3
|
94 |
|
95846ecf9
|
95 |
idr_init(&ns->idr); |
74bd59bb3
|
96 |
|
dd206bec9
|
97 |
ns->pid_cachep = create_pid_cachep(level); |
74bd59bb3
|
98 |
if (ns->pid_cachep == NULL) |
95846ecf9
|
99 |
goto out_free_idr; |
74bd59bb3
|
100 |
|
6344c433a
|
101 |
err = ns_alloc_inum(&ns->ns); |
98f842e67
|
102 |
if (err) |
95846ecf9
|
103 |
goto out_free_idr; |
33c429405
|
104 |
ns->ns.ops = &pidns_operations; |
98f842e67
|
105 |
|
74bd59bb3
|
106 |
kref_init(&ns->kref); |
74bd59bb3
|
107 |
ns->level = level; |
ed469a63c
|
108 |
ns->parent = get_pid_ns(parent_pid_ns); |
49f4d8b93
|
109 |
ns->user_ns = get_user_ns(user_ns); |
f333c700c
|
110 |
ns->ucounts = ucounts; |
e8cfbc245
|
111 |
ns->pid_allocated = PIDNS_ADDING; |
0a01f2cc3
|
112 |
INIT_WORK(&ns->proc_work, proc_cleanup_work); |
74bd59bb3
|
113 |
|
74bd59bb3
|
114 |
return ns; |
95846ecf9
|
115 116 |
out_free_idr: idr_destroy(&ns->idr); |
74bd59bb3
|
117 |
kmem_cache_free(pid_ns_cachep, ns); |
f333c700c
|
118 119 |
out_dec: dec_pid_namespaces(ucounts); |
74bd59bb3
|
120 |
out: |
4308eebbe
|
121 |
return ERR_PTR(err); |
74bd59bb3
|
122 |
} |
1adfcb03e
|
123 124 |
static void delayed_free_pidns(struct rcu_head *p) { |
add7c65ca
|
125 126 127 128 129 130 |
struct pid_namespace *ns = container_of(p, struct pid_namespace, rcu); dec_pid_namespaces(ns->ucounts); put_user_ns(ns->user_ns); kmem_cache_free(pid_ns_cachep, ns); |
1adfcb03e
|
131 |
} |
74bd59bb3
|
132 133 |
static void destroy_pid_namespace(struct pid_namespace *ns) { |
6344c433a
|
134 |
ns_free_inum(&ns->ns); |
95846ecf9
|
135 136 |
idr_destroy(&ns->idr); |
1adfcb03e
|
137 |
call_rcu(&ns->rcu, delayed_free_pidns); |
74bd59bb3
|
138 |
} |
49f4d8b93
|
139 140 |
struct pid_namespace *copy_pid_ns(unsigned long flags, struct user_namespace *user_ns, struct pid_namespace *old_ns) |
74bd59bb3
|
141 |
{ |
74bd59bb3
|
142 |
if (!(flags & CLONE_NEWPID)) |
dca4a9796
|
143 |
return get_pid_ns(old_ns); |
225778d68
|
144 145 |
if (task_active_pid_ns(current) != old_ns) return ERR_PTR(-EINVAL); |
49f4d8b93
|
146 |
return create_pid_namespace(user_ns, old_ns); |
74bd59bb3
|
147 |
} |
bbc2e3ef8
|
148 |
static void free_pid_ns(struct kref *kref) |
74bd59bb3
|
149 |
{ |
bbc2e3ef8
|
150 |
struct pid_namespace *ns; |
74bd59bb3
|
151 152 |
ns = container_of(kref, struct pid_namespace, kref); |
74bd59bb3
|
153 |
destroy_pid_namespace(ns); |
bbc2e3ef8
|
154 |
} |
74bd59bb3
|
155 |
|
bbc2e3ef8
|
156 157 158 159 160 161 162 163 164 165 |
void put_pid_ns(struct pid_namespace *ns) { struct pid_namespace *parent; while (ns != &init_pid_ns) { parent = ns->parent; if (!kref_put(&ns->kref, free_pid_ns)) break; ns = parent; } |
74bd59bb3
|
166 |
} |
bbc2e3ef8
|
167 |
EXPORT_SYMBOL_GPL(put_pid_ns); |
74bd59bb3
|
168 169 170 171 172 |
void zap_pid_ns_processes(struct pid_namespace *pid_ns) { int nr; int rc; |
00c10bc13
|
173 |
struct task_struct *task, *me = current; |
751c644b9
|
174 |
int init_pids = thread_group_leader(me) ? 1 : 2; |
95846ecf9
|
175 |
struct pid *pid; |
00c10bc13
|
176 |
|
c876ad768
|
177 178 |
/* Don't allow any more processes into the pid namespace */ disable_pid_allocation(pid_ns); |
a53b83154
|
179 180 181 182 183 |
/* * Ignore SIGCHLD causing any terminated children to autoreap. * This speeds up the namespace shutdown, plus see the comment * below. */ |
00c10bc13
|
184 185 186 |
spin_lock_irq(&me->sighand->siglock); me->sighand->action[SIGCHLD - 1].sa.sa_handler = SIG_IGN; spin_unlock_irq(&me->sighand->siglock); |
74bd59bb3
|
187 188 189 190 191 192 193 194 195 196 197 198 199 200 |
/* * The last thread in the cgroup-init thread group is terminating. * Find remaining pid_ts in the namespace, signal and wait for them * to exit. * * Note: This signals each threads in the namespace - even those that * belong to the same thread group, To avoid this, we would have * to walk the entire tasklist looking a processes in this * namespace, but that could be unnecessarily expensive if the * pid namespace has just a few processes. Or we need to * maintain a tasklist for each pid namespace. * */ |
95846ecf9
|
201 |
rcu_read_lock(); |
74bd59bb3
|
202 |
read_lock(&tasklist_lock); |
95846ecf9
|
203 204 205 |
nr = 2; idr_for_each_entry_continue(&pid_ns->idr, pid, nr) { task = pid_task(pid, PIDTYPE_PID); |
a02d6fd64
|
206 207 |
if (task && !__fatal_signal_pending(task)) send_sig_info(SIGKILL, SEND_SIG_FORCED, task); |
74bd59bb3
|
208 209 |
} read_unlock(&tasklist_lock); |
95846ecf9
|
210 |
rcu_read_unlock(); |
74bd59bb3
|
211 |
|
a53b83154
|
212 213 |
/* * Reap the EXIT_ZOMBIE children we had before we ignored SIGCHLD. |
d300b6108
|
214 |
* kernel_wait4() will also block until our children traced from the |
a53b83154
|
215 216 |
* parent namespace are detached and become EXIT_DEAD. */ |
74bd59bb3
|
217 218 |
do { clear_thread_flag(TIF_SIGPENDING); |
d300b6108
|
219 |
rc = kernel_wait4(-1, NULL, __WALL, NULL); |
74bd59bb3
|
220 |
} while (rc != -ECHILD); |
6347e9009
|
221 |
/* |
d300b6108
|
222 |
* kernel_wait4() above can't reap the EXIT_DEAD children but we do not |
a53b83154
|
223 224 |
* really care, we could reparent them to the global init. We could * exit and reap ->child_reaper even if it is not the last thread in |
e8cfbc245
|
225 |
* this pid_ns, free_pid(pid_allocated == 0) calls proc_cleanup_work(), |
a53b83154
|
226 227 228 229 230 231 232 233 234 235 |
* pid_ns can not go away until proc_kill_sb() drops the reference. * * But this ns can also have other tasks injected by setns()+fork(). * Again, ignoring the user visible semantics we do not really need * to wait until they are all reaped, but they can be reparented to * us and thus we need to ensure that pid->child_reaper stays valid * until they all go away. See free_pid()->wake_up_process(). * * We rely on ignored SIGCHLD, an injected zombie must be autoreaped * if reparented. |
6347e9009
|
236 237 |
*/ for (;;) { |
b9a985db9
|
238 |
set_current_state(TASK_INTERRUPTIBLE); |
e8cfbc245
|
239 |
if (pid_ns->pid_allocated == init_pids) |
6347e9009
|
240 241 242 |
break; schedule(); } |
af4b8a83a
|
243 |
__set_current_state(TASK_RUNNING); |
6347e9009
|
244 |
|
cf3f89214
|
245 246 |
if (pid_ns->reboot) current->signal->group_exit_code = pid_ns->reboot; |
0b6b030fc
|
247 |
acct_exit_ns(pid_ns); |
74bd59bb3
|
248 249 |
return; } |
98ed57eef
|
250 |
#ifdef CONFIG_CHECKPOINT_RESTORE |
b8f566b04
|
251 252 253 |
static int pid_ns_ctl_handler(struct ctl_table *table, int write, void __user *buffer, size_t *lenp, loff_t *ppos) { |
49f4d8b93
|
254 |
struct pid_namespace *pid_ns = task_active_pid_ns(current); |
b8f566b04
|
255 |
struct ctl_table tmp = *table; |
95846ecf9
|
256 |
int ret, next; |
b8f566b04
|
257 |
|
49f4d8b93
|
258 |
if (write && !ns_capable(pid_ns->user_ns, CAP_SYS_ADMIN)) |
b8f566b04
|
259 260 261 262 263 264 265 |
return -EPERM; /* * Writing directly to ns' last_pid field is OK, since this field * is volatile in a living namespace anyway and a code writing to * it should synchronize its usage with external means. */ |
95846ecf9
|
266 267 268 269 270 271 272 273 |
next = idr_get_cursor(&pid_ns->idr) - 1; tmp.data = &next; ret = proc_dointvec_minmax(&tmp, write, buffer, lenp, ppos); if (!ret && write) idr_set_cursor(&pid_ns->idr, next + 1); return ret; |
b8f566b04
|
274 |
} |
579035dc5
|
275 276 |
extern int pid_max; static int zero = 0; |
b8f566b04
|
277 278 279 280 281 282 |
static struct ctl_table pid_ns_ctl_table[] = { { .procname = "ns_last_pid", .maxlen = sizeof(int), .mode = 0666, /* permissions are checked in the handler */ .proc_handler = pid_ns_ctl_handler, |
579035dc5
|
283 284 |
.extra1 = &zero, .extra2 = &pid_max, |
b8f566b04
|
285 286 287 |
}, { } }; |
b8f566b04
|
288 |
static struct ctl_path kern_path[] = { { .procname = "kernel", }, { } }; |
98ed57eef
|
289 |
#endif /* CONFIG_CHECKPOINT_RESTORE */ |
b8f566b04
|
290 |
|
cf3f89214
|
291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 |
int reboot_pid_ns(struct pid_namespace *pid_ns, int cmd) { if (pid_ns == &init_pid_ns) return 0; switch (cmd) { case LINUX_REBOOT_CMD_RESTART2: case LINUX_REBOOT_CMD_RESTART: pid_ns->reboot = SIGHUP; break; case LINUX_REBOOT_CMD_POWER_OFF: case LINUX_REBOOT_CMD_HALT: pid_ns->reboot = SIGINT; break; default: return -EINVAL; } read_lock(&tasklist_lock); force_sig(SIGKILL, pid_ns->child_reaper); read_unlock(&tasklist_lock); do_exit(0); /* Not reached */ return 0; } |
3c0411846
|
319 320 321 322 |
static inline struct pid_namespace *to_pid_ns(struct ns_common *ns) { return container_of(ns, struct pid_namespace, ns); } |
64964528b
|
323 |
static struct ns_common *pidns_get(struct task_struct *task) |
57e8391d3
|
324 325 326 327 |
{ struct pid_namespace *ns; rcu_read_lock(); |
d23082257
|
328 329 330 |
ns = task_active_pid_ns(task); if (ns) get_pid_ns(ns); |
57e8391d3
|
331 |
rcu_read_unlock(); |
3c0411846
|
332 |
return ns ? &ns->ns : NULL; |
57e8391d3
|
333 |
} |
eaa0d190b
|
334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 |
static struct ns_common *pidns_for_children_get(struct task_struct *task) { struct pid_namespace *ns = NULL; task_lock(task); if (task->nsproxy) { ns = task->nsproxy->pid_ns_for_children; get_pid_ns(ns); } task_unlock(task); if (ns) { read_lock(&tasklist_lock); if (!ns->child_reaper) { put_pid_ns(ns); ns = NULL; } read_unlock(&tasklist_lock); } return ns ? &ns->ns : NULL; } |
64964528b
|
356 |
static void pidns_put(struct ns_common *ns) |
57e8391d3
|
357 |
{ |
3c0411846
|
358 |
put_pid_ns(to_pid_ns(ns)); |
57e8391d3
|
359 |
} |
64964528b
|
360 |
static int pidns_install(struct nsproxy *nsproxy, struct ns_common *ns) |
57e8391d3
|
361 362 |
{ struct pid_namespace *active = task_active_pid_ns(current); |
3c0411846
|
363 |
struct pid_namespace *ancestor, *new = to_pid_ns(ns); |
57e8391d3
|
364 |
|
5e4a08476
|
365 |
if (!ns_capable(new->user_ns, CAP_SYS_ADMIN) || |
c7b96acf1
|
366 |
!ns_capable(current_user_ns(), CAP_SYS_ADMIN)) |
57e8391d3
|
367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 |
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; |
c2b1df2eb
|
385 386 |
put_pid_ns(nsproxy->pid_ns_for_children); nsproxy->pid_ns_for_children = get_pid_ns(new); |
57e8391d3
|
387 388 |
return 0; } |
a7306ed8d
|
389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 |
static struct ns_common *pidns_get_parent(struct ns_common *ns) { struct pid_namespace *active = task_active_pid_ns(current); struct pid_namespace *pid_ns, *p; /* See if the parent is in the current namespace */ pid_ns = p = to_pid_ns(ns)->parent; for (;;) { if (!p) return ERR_PTR(-EPERM); if (p == active) break; p = p->parent; } return &get_pid_ns(pid_ns)->ns; } |
bcac25a58
|
406 407 408 409 |
static struct user_namespace *pidns_owner(struct ns_common *ns) { return to_pid_ns(ns)->user_ns; } |
57e8391d3
|
410 411 412 413 414 415 |
const struct proc_ns_operations pidns_operations = { .name = "pid", .type = CLONE_NEWPID, .get = pidns_get, .put = pidns_put, .install = pidns_install, |
bcac25a58
|
416 |
.owner = pidns_owner, |
a7306ed8d
|
417 |
.get_parent = pidns_get_parent, |
57e8391d3
|
418 |
}; |
eaa0d190b
|
419 420 421 422 423 424 425 426 427 428 |
const struct proc_ns_operations pidns_for_children_operations = { .name = "pid_for_children", .real_ns_name = "pid", .type = CLONE_NEWPID, .get = pidns_for_children_get, .put = pidns_put, .install = pidns_install, .owner = pidns_owner, .get_parent = pidns_get_parent, }; |
74bd59bb3
|
429 430 431 |
static __init int pid_namespaces_init(void) { pid_ns_cachep = KMEM_CACHE(pid_namespace, SLAB_PANIC); |
98ed57eef
|
432 433 |
#ifdef CONFIG_CHECKPOINT_RESTORE |
b8f566b04
|
434 |
register_sysctl_paths(kern_path, pid_ns_ctl_table); |
98ed57eef
|
435 |
#endif |
74bd59bb3
|
436 437 438 439 |
return 0; } __initcall(pid_namespaces_init); |