***/
#include <errno.h>
-#include <sys/mount.h>
-#include <string.h>
+#include <sched.h>
#include <stdio.h>
-#include <unistd.h>
+#include <string.h>
+#include <sys/mount.h>
#include <sys/stat.h>
-#include <sys/types.h>
-#include <sched.h>
-#include <sys/syscall.h>
-#include <limits.h>
+#include <unistd.h>
#include <linux/fs.h>
-#include <sys/file.h>
+#include "dev-setup.h"
+#include "loopback-setup.h"
+#include "missing.h"
+#include "mkdir.h"
+#include "path-util.h"
+#include "selinux-util.h"
+#include "string-util.h"
#include "strv.h"
#include "util.h"
-#include "path-util.h"
#include "namespace.h"
-#include "missing.h"
-#include "execute.h"
-#include "loopback-setup.h"
-#include "mkdir.h"
-#include "dev-setup.h"
-#include "def.h"
-#include "label.h"
typedef enum MountMode {
/* This is ordered by priority! */
static int mount_path_compare(const void *a, const void *b) {
const BindMount *p = a, *q = b;
+ int d;
- if (path_equal(p->path, q->path)) {
+ d = path_compare(p->path, q->path);
+ if (d == 0) {
/* If the paths are equal, check the mode */
if (p->mode < q->mode)
return -1;
}
/* If the paths are not equal, then order prefixes first */
- if (path_startswith(p->path, q->path))
- return 1;
-
- if (path_startswith(q->path, p->path))
- return -1;
-
- return 0;
+ return d;
}
static void drop_duplicates(BindMount *m, unsigned *n) {
"/dev/tty\0";
char temporary_mount[] = "/tmp/namespace-dev-XXXXXX";
- const char *d, *dev = NULL, *devpts = NULL, *devshm = NULL, *devkdbus = NULL, *devhugepages = NULL, *devmqueue = NULL, *devlog = NULL, *devptmx = NULL;
+ const char *d, *dev = NULL, *devpts = NULL, *devshm = NULL, *devhugepages = NULL, *devmqueue = NULL, *devlog = NULL, *devptmx = NULL;
_cleanup_umask_ mode_t u;
int r;
if (!mkdtemp(temporary_mount))
return -errno;
- dev = strappenda(temporary_mount, "/dev");
- mkdir(dev, 0755);
+ dev = strjoina(temporary_mount, "/dev");
+ (void) mkdir(dev, 0755);
if (mount("tmpfs", dev, "tmpfs", MS_NOSUID|MS_STRICTATIME, "mode=755") < 0) {
r = -errno;
goto fail;
}
- devpts = strappenda(temporary_mount, "/dev/pts");
- mkdir(devpts, 0755);
+ devpts = strjoina(temporary_mount, "/dev/pts");
+ (void) mkdir(devpts, 0755);
if (mount("/dev/pts", devpts, NULL, MS_BIND, NULL) < 0) {
r = -errno;
goto fail;
}
- devptmx = strappenda(temporary_mount, "/dev/ptmx");
- symlink("pts/ptmx", devptmx);
+ devptmx = strjoina(temporary_mount, "/dev/ptmx");
+ if (symlink("pts/ptmx", devptmx) < 0) {
+ r = -errno;
+ goto fail;
+ }
- devshm = strappenda(temporary_mount, "/dev/shm");
- mkdir(devshm, 01777);
+ devshm = strjoina(temporary_mount, "/dev/shm");
+ (void) mkdir(devshm, 01777);
r = mount("/dev/shm", devshm, NULL, MS_BIND, NULL);
if (r < 0) {
r = -errno;
goto fail;
}
- devmqueue = strappenda(temporary_mount, "/dev/mqueue");
- mkdir(devmqueue, 0755);
- mount("/dev/mqueue", devmqueue, NULL, MS_BIND, NULL);
+ devmqueue = strjoina(temporary_mount, "/dev/mqueue");
+ (void) mkdir(devmqueue, 0755);
+ (void) mount("/dev/mqueue", devmqueue, NULL, MS_BIND, NULL);
- devkdbus = strappenda(temporary_mount, "/dev/kdbus");
- mkdir(devkdbus, 0755);
- mount("/dev/kdbus", devkdbus, NULL, MS_BIND, NULL);
+ devhugepages = strjoina(temporary_mount, "/dev/hugepages");
+ (void) mkdir(devhugepages, 0755);
+ (void) mount("/dev/hugepages", devhugepages, NULL, MS_BIND, NULL);
- devhugepages = strappenda(temporary_mount, "/dev/hugepages");
- mkdir(devhugepages, 0755);
- mount("/dev/hugepages", devhugepages, NULL, MS_BIND, NULL);
-
- devlog = strappenda(temporary_mount, "/dev/log");
- symlink("/run/systemd/journal/dev-log", devlog);
+ devlog = strjoina(temporary_mount, "/dev/log");
+ (void) symlink("/run/systemd/journal/dev-log", devlog);
NULSTR_FOREACH(d, devnodes) {
_cleanup_free_ char *dn = NULL;
goto fail;
}
- label_context_set(d, st.st_mode);
+ mac_selinux_create_file_prepare(d, st.st_mode);
r = mknod(dn, st.st_mode, st.st_rdev);
- label_context_clear();
+ mac_selinux_create_file_clear();
if (r < 0) {
r = -errno;
}
}
- dev_setup(temporary_mount);
+ dev_setup(temporary_mount, UID_INVALID, GID_INVALID);
+
+ /* Create the /dev directory if missing. It is more likely to be
+ * missing when the service is started with RootDirectory. This is
+ * consistent with mount units creating the mount points when missing.
+ */
+ (void) mkdir_p_label(m->path, 0755);
- if (mount(dev, "/dev/", NULL, MS_MOVE, NULL) < 0) {
+ if (mount(dev, m->path, NULL, MS_MOVE, NULL) < 0) {
r = -errno;
goto fail;
}
if (devshm)
umount(devshm);
- if (devkdbus)
- umount(devkdbus);
-
if (devhugepages)
umount(devhugepages);
u = umask(0000);
- if (!mkdtemp(temporary_mount)) {
- log_error("Failed create temp dir: %m");
- return -errno;
- }
+ if (!mkdtemp(temporary_mount))
+ return log_error_errno(errno, "Failed create temp dir: %m");
- root = strappenda(temporary_mount, "/kdbus");
- mkdir(root, 0755);
+ root = strjoina(temporary_mount, "/kdbus");
+ (void) mkdir(root, 0755);
if (mount("tmpfs", root, "tmpfs", MS_NOSUID|MS_STRICTATIME, "mode=777") < 0) {
r = -errno;
goto fail;
/* create a new /dev/null dev node copy so we have some fodder to
* bind-mount the custom endpoint over. */
if (stat("/dev/null", &st) < 0) {
- log_error("Failed to stat /dev/null: %m");
- r = -errno;
+ r = log_error_errno(errno, "Failed to stat /dev/null: %m");
goto fail;
}
- busnode = strappenda(root, "/bus");
+ busnode = strjoina(root, "/bus");
if (mknod(busnode, (st.st_mode & ~07777) | 0600, st.st_rdev) < 0) {
- log_error("mknod() for %s failed: %m", busnode);
- r = -errno;
+ r = log_error_errno(errno, "mknod() for %s failed: %m",
+ busnode);
goto fail;
}
- r = mount(m->path, busnode, "bind", MS_BIND, NULL);
+ r = mount(m->path, busnode, NULL, MS_BIND, NULL);
if (r < 0) {
- log_error("bind mount of %s failed: %m", m->path);
- r = -errno;
+ r = log_error_errno(errno, "bind mount of %s failed: %m",
+ m->path);
goto fail;
}
}
if (mount(root, basepath, NULL, MS_MOVE, NULL) < 0) {
- log_error("bind mount of %s failed: %m", basepath);
- r = -errno;
+ r = log_error_errno(errno, "bind mount of %s failed: %m",
+ basepath);
goto fail;
}
}
int setup_namespace(
+ const char* root_directory,
char** read_write_dirs,
char** read_only_dirs,
char** inaccessible_dirs,
- char* tmp_dir,
- char* var_tmp_dir,
- char* bus_endpoint_path,
+ const char* tmp_dir,
+ const char* var_tmp_dir,
+ const char* bus_endpoint_path,
bool private_dev,
ProtectHome protect_home,
ProtectSystem protect_system,
- unsigned mount_flags) {
+ unsigned long mount_flags) {
BindMount *m, *mounts = NULL;
unsigned n;
return r;
if (tmp_dir) {
- m->path = "/tmp";
+ m->path = prefix_roota(root_directory, "/tmp");
m->mode = PRIVATE_TMP;
m++;
}
if (var_tmp_dir) {
- m->path = "/var/tmp";
+ m->path = prefix_roota(root_directory, "/var/tmp");
m->mode = PRIVATE_VAR_TMP;
m++;
}
if (private_dev) {
- m->path = "/dev";
+ m->path = prefix_roota(root_directory, "/dev");
m->mode = PRIVATE_DEV;
m++;
}
if (bus_endpoint_path) {
- m->path = bus_endpoint_path;
+ m->path = prefix_roota(root_directory, bus_endpoint_path);
m->mode = PRIVATE_BUS_ENDPOINT;
m++;
}
if (protect_home != PROTECT_HOME_NO) {
- r = append_mounts(&m, STRV_MAKE("-/home", "-/run/user", "-/root"), protect_home == PROTECT_HOME_READ_ONLY ? READONLY : INACCESSIBLE);
+ const char *home_dir, *run_user_dir, *root_dir;
+
+ home_dir = prefix_roota(root_directory, "/home");
+ home_dir = strjoina("-", home_dir);
+ run_user_dir = prefix_roota(root_directory, "/run/user");
+ run_user_dir = strjoina("-", run_user_dir);
+ root_dir = prefix_roota(root_directory, "/root");
+ root_dir = strjoina("-", root_dir);
+
+ r = append_mounts(&m, STRV_MAKE(home_dir, run_user_dir, root_dir),
+ protect_home == PROTECT_HOME_READ_ONLY ? READONLY : INACCESSIBLE);
if (r < 0)
return r;
}
if (protect_system != PROTECT_SYSTEM_NO) {
- r = append_mounts(&m, protect_system == PROTECT_SYSTEM_FULL ? STRV_MAKE("/usr", "-/boot", "/etc") : STRV_MAKE("/usr", "-/boot"), READONLY);
+ const char *usr_dir, *boot_dir, *etc_dir;
+
+ usr_dir = prefix_roota(root_directory, "/usr");
+ boot_dir = prefix_roota(root_directory, "/boot");
+ boot_dir = strjoina("-", boot_dir);
+ etc_dir = prefix_roota(root_directory, "/etc");
+
+ r = append_mounts(&m, protect_system == PROTECT_SYSTEM_FULL
+ ? STRV_MAKE(usr_dir, boot_dir, etc_dir)
+ : STRV_MAKE(usr_dir, boot_dir), READONLY);
if (r < 0)
return r;
}
drop_duplicates(mounts, &n);
}
- if (n > 0) {
+ if (n > 0 || root_directory) {
/* Remount / as SLAVE so that nothing now mounted in the namespace
shows up in the parent */
if (mount(NULL, "/", NULL, MS_SLAVE|MS_REC, NULL) < 0)
return -errno;
+ }
+
+ if (root_directory) {
+ /* Turn directory into bind mount */
+ if (mount(root_directory, root_directory, NULL, MS_BIND|MS_REC, NULL) < 0)
+ return -errno;
+ }
+ if (n > 0) {
for (m = mounts; m < mounts + n; ++m) {
r = apply_mount(m, tmp_dir, var_tmp_dir);
if (r < 0)
}
}
+ if (root_directory) {
+ /* MS_MOVE does not work on MS_SHARED so the remount MS_SHARED will be done later */
+ r = mount_move_root(root_directory);
+
+ /* at this point, we cannot rollback */
+ if (r < 0)
+ return r;
+ }
+
/* Remount / as the desired mode. Not that this will not
* reestablish propagation from our side to the host, since
* what's disconnected is disconnected. */
- if (mount(NULL, "/", NULL, mount_flags | MS_REC, NULL) < 0) {
- r = -errno;
- goto fail;
- }
+ if (mount(NULL, "/", NULL, mount_flags | MS_REC, NULL) < 0)
+ /* at this point, we cannot rollback */
+ return -errno;
return 0;
if (n > 0) {
for (m = mounts; m < mounts + n; ++m)
if (m->done)
- umount2(m->path, MNT_DETACH);
+ (void) umount2(m->path, MNT_DETACH);
}
return r;
RUN_WITH_UMASK(0000) {
char *y;
- y = strappenda(x, "/tmp");
+ y = strjoina(x, "/tmp");
if (mkdir(y, 0777 | S_ISVTX) < 0)
return -errno;
if (r < 0) {
char *t;
- t = strappenda(a, "/tmp");
+ t = strjoina(a, "/tmp");
rmdir(t);
rmdir(a);
int setup_netns(int netns_storage_socket[2]) {
_cleanup_close_ int netns = -1;
- union {
- struct cmsghdr cmsghdr;
- uint8_t buf[CMSG_SPACE(sizeof(int))];
- } control = {};
- struct msghdr mh = {
- .msg_control = &control,
- .msg_controllen = sizeof(control),
- };
- struct cmsghdr *cmsg;
- int r;
+ int r, q;
assert(netns_storage_socket);
assert(netns_storage_socket[0] >= 0);
if (lockf(netns_storage_socket[0], F_LOCK, 0) < 0)
return -errno;
- if (recvmsg(netns_storage_socket[0], &mh, MSG_DONTWAIT|MSG_CMSG_CLOEXEC) < 0) {
- if (errno != EAGAIN) {
- r = -errno;
- goto fail;
- }
-
+ netns = receive_one_fd(netns_storage_socket[0], MSG_DONTWAIT);
+ if (netns == -EAGAIN) {
/* Nothing stored yet, so let's create a new namespace */
if (unshare(CLONE_NEWNET) < 0) {
}
r = 1;
- } else {
- /* Yay, found something, so let's join the namespace */
- for (cmsg = CMSG_FIRSTHDR(&mh); cmsg; cmsg = CMSG_NXTHDR(&mh, cmsg)) {
- if (cmsg->cmsg_level == SOL_SOCKET && cmsg->cmsg_type == SCM_RIGHTS) {
- assert(cmsg->cmsg_len == CMSG_LEN(sizeof(int)));
- netns = *(int*) CMSG_DATA(cmsg);
- }
- }
+ } else if (netns < 0) {
+ r = netns;
+ goto fail;
+ } else {
+ /* Yay, found something, so let's join the namespace */
if (setns(netns, CLONE_NEWNET) < 0) {
r = -errno;
goto fail;
r = 0;
}
- cmsg = CMSG_FIRSTHDR(&mh);
- cmsg->cmsg_level = SOL_SOCKET;
- cmsg->cmsg_type = SCM_RIGHTS;
- cmsg->cmsg_len = CMSG_LEN(sizeof(int));
- memcpy(CMSG_DATA(cmsg), &netns, sizeof(int));
- mh.msg_controllen = cmsg->cmsg_len;
-
- if (sendmsg(netns_storage_socket[1], &mh, MSG_DONTWAIT|MSG_NOSIGNAL) < 0) {
- r = -errno;
+ q = send_one_fd(netns_storage_socket[1], netns, MSG_DONTWAIT);
+ if (q < 0) {
+ r = q;
goto fail;
}
fail:
lockf(netns_storage_socket[0], F_ULOCK, 0);
-
return r;
}