]>
git.proxmox.com Git - mirror_lxc.git/blob - src/lxc/namespace.c
2 * lxc: linux Container library
4 * (C) Copyright IBM Corp. 2007, 2009
7 * Daniel Lezcano <daniel.lezcano at free.fr>
9 * This library is free software; you can redistribute it and/or
10 * modify it under the terms of the GNU Lesser General Public
11 * License as published by the Free Software Foundation; either
12 * version 2.1 of the License, or (at your option) any later version.
14 * This library is distributed in the hope that it will be useful,
15 * but WITHOUT ANY WARRANTY; without even the implied warranty of
16 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
17 * Lesser General Public License for more details.
19 * You should have received a copy of the GNU Lesser General Public
20 * License along with this library; if not, write to the Free Software
21 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
30 #include <sys/param.h>
32 #include <sys/syscall.h>
33 #include <sys/types.h>
36 #include "namespace.h"
39 lxc_log_define(lxc_namespace
, lxc
);
46 static int do_clone(void *arg
)
48 struct clone_arg
*clone_arg
= arg
;
49 return clone_arg
->fn(clone_arg
->arg
);
52 pid_t
lxc_clone(int (*fn
)(void *), void *arg
, int flags
)
54 struct clone_arg clone_arg
= {
59 size_t stack_size
= lxc_getpagesize();
60 void *stack
= alloca(stack_size
);
64 ret
= __clone2(do_clone
, stack
, stack_size
, flags
| SIGCHLD
, &clone_arg
);
66 ret
= clone(do_clone
, stack
+ stack_size
, flags
| SIGCHLD
, &clone_arg
);
69 ERROR("Failed to clone (%#x): %s.", flags
, strerror(errno
));
75 * This is based on raw_clone in systemd but adapted to our needs. This uses
76 * copy on write semantics and doesn't pass a stack. CLONE_VM is tricky and
77 * doesn't really matter to us so disallow it.
79 * The nice thing about this is that we get fork() behavior. That is
80 * lxc_raw_clone() returns 0 in the child and the child pid in the parent.
82 pid_t
lxc_raw_clone(unsigned long flags
)
85 /* These flags don't interest at all so we don't jump through any hoopes
86 * of retrieving them and passing them to the kernel.
89 if ((flags
& (CLONE_VM
| CLONE_PARENT_SETTID
| CLONE_CHILD_SETTID
|
90 CLONE_CHILD_CLEARTID
| CLONE_SETTLS
)))
93 #if defined(__s390x__) || defined(__s390__) || defined(__CRIS__)
94 /* On s390/s390x and cris the order of the first and second arguments
95 * of the system call is reversed.
97 return (int)syscall(__NR_clone
, NULL
, flags
| SIGCHLD
);
98 #elif defined(__sparc__) && defined(__arch64__)
101 * sparc64 always returns the other process id in %o0, and
102 * a boolean flag whether this is the child or the parent in
103 * %o1. Inline assembly is needed to get the flag returned
108 asm volatile("mov %2, %%g1\n\t"
114 : "=r"(in_child
), "=r"(child_pid
)
115 : "i"(__NR_clone
), "r"(flags
| SIGCHLD
)
116 : "%o1", "%o0", "%g1");
122 #elif defined(__ia64__)
123 /* On ia64 the stack and stack size are passed as separate arguments. */
124 return (int)syscall(__NR_clone
, flags
| SIGCHLD
, NULL
, 0);
126 return (int)syscall(__NR_clone
, flags
| SIGCHLD
, NULL
);
130 /* Leave the user namespace at the first position in the array of structs so
131 * that we always attach to it first when iterating over the struct and using
132 * setns() to switch namespaces. This especially affects lxc_attach(): Suppose
133 * you cloned a new user namespace and mount namespace as an unprivileged user
134 * on the host and want to setns() to the mount namespace. This requires you to
135 * attach to the user namespace first otherwise the kernel will fail this check:
137 * if (!ns_capable(mnt_ns->user_ns, CAP_SYS_ADMIN) ||
138 * !ns_capable(current_user_ns(), CAP_SYS_CHROOT) ||
139 * !ns_capable(current_user_ns(), CAP_SYS_ADMIN))
144 * linux/fs/namespace.c:mntns_install().
146 const struct ns_info ns_info
[LXC_NS_MAX
] = {
147 [LXC_NS_USER
] = { "user", CLONE_NEWUSER
, "CLONE_NEWUSER", "LXC_USER_NS" },
148 [LXC_NS_MNT
] = { "mnt", CLONE_NEWNS
, "CLONE_NEWNS", "LXC_MNT_NS" },
149 [LXC_NS_PID
] = { "pid", CLONE_NEWPID
, "CLONE_NEWPID", "LXC_PID_NS" },
150 [LXC_NS_UTS
] = { "uts", CLONE_NEWUTS
, "CLONE_NEWUTS", "LXC_UTS_NS" },
151 [LXC_NS_IPC
] = { "ipc", CLONE_NEWIPC
, "CLONE_NEWIPC", "LXC_IPC_NS" },
152 [LXC_NS_NET
] = { "net", CLONE_NEWNET
, "CLONE_NEWNET", "LXC_NET_NS" },
153 [LXC_NS_CGROUP
] = { "cgroup", CLONE_NEWCGROUP
, "CLONE_NEWCGROUP", "LXC_CGROUP_NS" }
156 int lxc_namespace_2_cloneflag(const char *namespace)
159 for (i
= 0; i
< LXC_NS_MAX
; i
++)
160 if (!strcasecmp(ns_info
[i
].proc_name
, namespace))
161 return ns_info
[i
].clone_flag
;
163 ERROR("Invalid namespace name \"%s\"", namespace);
167 int lxc_namespace_2_ns_idx(const char *namespace)
170 for (i
= 0; i
< LXC_NS_MAX
; i
++)
171 if (!strcmp(ns_info
[i
].proc_name
, namespace))
174 ERROR("Invalid namespace name \"%s\"", namespace);
178 int lxc_fill_namespace_flags(char *flaglist
, int *flags
)
180 char *token
, *saveptr
= NULL
;
184 ERROR("At least one namespace is needed.");
188 token
= strtok_r(flaglist
, "|", &saveptr
);
191 aflag
= lxc_namespace_2_cloneflag(token
);
197 token
= strtok_r(NULL
, "|", &saveptr
);