#include <sys/stat.h>
#include <stdlib.h>
#include <string.h>
-#include <libgen.h>
#include <assert.h>
#include <unistd.h>
#include <ftw.h>
#include "missing.h"
#include "virt.h"
#include "efivars.h"
-
-#ifndef TTY_GID
-#define TTY_GID 5
-#endif
+#include "smack-util.h"
+#include "def.h"
+#include "cgroup-util.h"
typedef enum MountMode {
MNT_NONE = 0,
/* The first three entries we might need before SELinux is up. The
* fourth (securityfs) is needed by IMA to load a custom policy. The
- * other ones we can delay until SELinux and IMA are loaded. */
+ * other ones we can delay until SELinux and IMA are loaded. When
+ * SMACK is enabled we need smackfs, too, so it's a fifth one. */
+#ifdef HAVE_SMACK
#define N_EARLY_MOUNT 5
+#else
+#define N_EARLY_MOUNT 4
+#endif
static const MountPoint mount_table[] = {
- { "proc", "/proc", "proc", NULL, MS_NOSUID|MS_NOEXEC|MS_NODEV,
- NULL, MNT_FATAL|MNT_IN_CONTAINER },
- { "sysfs", "/sys", "sysfs", NULL, MS_NOSUID|MS_NOEXEC|MS_NODEV,
- NULL, MNT_FATAL|MNT_IN_CONTAINER },
- { "devtmpfs", "/dev", "devtmpfs", "mode=755", MS_NOSUID|MS_STRICTATIME,
- NULL, MNT_FATAL|MNT_IN_CONTAINER },
- { "securityfs", "/sys/kernel/security", "securityfs", NULL, MS_NOSUID|MS_NOEXEC|MS_NODEV,
- NULL, MNT_NONE },
- { "smackfs", "/sys/fs/smackfs", "smackfs", "smackfsdef=*", MS_NOSUID|MS_NOEXEC|MS_NODEV|MS_STRICTATIME,
- NULL, MNT_NONE },
- { "tmpfs", "/dev/shm", "tmpfs", "mode=1777", MS_NOSUID|MS_NODEV|MS_STRICTATIME,
- NULL, MNT_FATAL|MNT_IN_CONTAINER },
- { "devpts", "/dev/pts", "devpts", "mode=620,gid=" STRINGIFY(TTY_GID), MS_NOSUID|MS_NOEXEC,
- NULL, MNT_IN_CONTAINER },
- { "tmpfs", "/run", "tmpfs", "mode=755", MS_NOSUID|MS_NODEV|MS_STRICTATIME,
- NULL, MNT_FATAL|MNT_IN_CONTAINER },
- { "tmpfs", "/sys/fs/cgroup", "tmpfs", "mode=755", MS_NOSUID|MS_NOEXEC|MS_NODEV|MS_STRICTATIME,
- NULL, MNT_IN_CONTAINER },
- { "cgroup", "/sys/fs/cgroup/systemd", "cgroup", "none,name=systemd", MS_NOSUID|MS_NOEXEC|MS_NODEV,
- NULL, MNT_IN_CONTAINER },
- { "pstore", "/sys/fs/pstore", "pstore", NULL, MS_NOSUID|MS_NOEXEC|MS_NODEV,
- NULL, MNT_NONE },
+ { "sysfs", "/sys", "sysfs", NULL, MS_NOSUID|MS_NOEXEC|MS_NODEV,
+ NULL, MNT_FATAL|MNT_IN_CONTAINER },
+ { "proc", "/proc", "proc", NULL, MS_NOSUID|MS_NOEXEC|MS_NODEV,
+ NULL, MNT_FATAL|MNT_IN_CONTAINER },
+ { "devtmpfs", "/dev", "devtmpfs", "mode=755", MS_NOSUID|MS_STRICTATIME,
+ NULL, MNT_FATAL|MNT_IN_CONTAINER },
+ { "securityfs", "/sys/kernel/security", "securityfs", NULL, MS_NOSUID|MS_NOEXEC|MS_NODEV,
+ NULL, MNT_NONE },
+#ifdef HAVE_SMACK
+ { "smackfs", "/sys/fs/smackfs", "smackfs", "smackfsdef=*", MS_NOSUID|MS_NOEXEC|MS_NODEV,
+ mac_smack_use, MNT_FATAL },
+ { "tmpfs", "/dev/shm", "tmpfs", "mode=1777,smackfsroot=*", MS_NOSUID|MS_NODEV|MS_STRICTATIME,
+ mac_smack_use, MNT_FATAL },
+#endif
+ { "tmpfs", "/dev/shm", "tmpfs", "mode=1777", MS_NOSUID|MS_NODEV|MS_STRICTATIME,
+ NULL, MNT_FATAL|MNT_IN_CONTAINER },
+ { "devpts", "/dev/pts", "devpts", "mode=620,gid=" STRINGIFY(TTY_GID), MS_NOSUID|MS_NOEXEC,
+ NULL, MNT_IN_CONTAINER },
+#ifdef HAVE_SMACK
+ { "tmpfs", "/run", "tmpfs", "mode=755,smackfsroot=*", MS_NOSUID|MS_NODEV|MS_STRICTATIME,
+ mac_smack_use, MNT_FATAL },
+#endif
+ { "tmpfs", "/run", "tmpfs", "mode=755", MS_NOSUID|MS_NODEV|MS_STRICTATIME,
+ NULL, MNT_FATAL|MNT_IN_CONTAINER },
+ { "tmpfs", "/sys/fs/cgroup", "tmpfs", "mode=755", MS_NOSUID|MS_NOEXEC|MS_NODEV|MS_STRICTATIME,
+ NULL, MNT_FATAL|MNT_IN_CONTAINER },
+ { "cgroup", "/sys/fs/cgroup/systemd", "cgroup", "none,name=systemd,xattr", MS_NOSUID|MS_NOEXEC|MS_NODEV,
+ NULL, MNT_IN_CONTAINER },
+ { "cgroup", "/sys/fs/cgroup/systemd", "cgroup", "none,name=systemd", MS_NOSUID|MS_NOEXEC|MS_NODEV,
+ NULL, MNT_FATAL|MNT_IN_CONTAINER },
+ { "pstore", "/sys/fs/pstore", "pstore", NULL, MS_NOSUID|MS_NOEXEC|MS_NODEV,
+ NULL, MNT_NONE },
#ifdef ENABLE_EFI
- { "efivarfs", "/sys/firmware/efi/efivars", "efivarfs", NULL, MS_NOSUID|MS_NOEXEC|MS_NODEV,
- is_efi_boot, MNT_NONE },
+ { "efivarfs", "/sys/firmware/efi/efivars", "efivarfs", NULL, MS_NOSUID|MS_NOEXEC|MS_NODEV,
+ is_efi_boot, MNT_NONE },
+#endif
+#ifdef ENABLE_KDBUS
+ { "kdbusfs", "/sys/fs/kdbus", "kdbusfs", NULL, MS_NOSUID|MS_NOEXEC|MS_NODEV,
+ NULL, MNT_IN_CONTAINER },
#endif
};
static const char ignore_paths[] =
/* SELinux file systems */
"/sys/fs/selinux\0"
- "/selinux\0"
- /* Legacy cgroup mount points */
- "/dev/cgroup\0"
- "/cgroup\0"
- /* Legacy kernel file system */
- "/proc/bus/usb\0"
/* Container bind mounts */
"/proc/sys\0"
"/dev/console\0"
/* The access mode here doesn't really matter too much, since
* the mounted file system will take precedence anyway. */
- mkdir_p_label(p->where, 0755);
+ if (relabel)
+ mkdir_p_label(p->where, 0755);
+ else
+ mkdir_p(p->where, 0755);
log_debug("Mounting %s to %s of type %s with options %s.",
p->what,
p->type,
p->flags,
p->options) < 0) {
- log_full((p->mode & MNT_FATAL) ? LOG_ERR : LOG_DEBUG, "Failed to mount %s: %s", p->where, strerror(errno));
+ log_full((p->mode & MNT_FATAL) ? LOG_ERR : LOG_DEBUG, "Failed to mount %s at %s: %m", p->type, p->where);
return (p->mode & MNT_FATAL) ? -errno : 0;
}
}
int mount_cgroup_controllers(char ***join_controllers) {
+ _cleanup_set_free_free_ Set *controllers = NULL;
int r;
- FILE *f;
- char buf[LINE_MAX];
- Set *controllers;
/* Mount all available cgroup controllers that are built into the kernel. */
- f = fopen("/proc/cgroups", "re");
- if (!f) {
- log_error("Failed to enumerate cgroup controllers: %m");
- return 0;
- }
-
- controllers = set_new(string_hash_func, string_compare_func);
- if (!controllers) {
- r = log_oom();
- goto finish;
- }
+ controllers = set_new(&string_hash_ops);
+ if (!controllers)
+ return log_oom();
- /* Ignore the header line */
- (void) fgets(buf, sizeof(buf), f);
-
- for (;;) {
- char *controller;
- int enabled = 0;
-
- if (fscanf(f, "%ms %*i %*i %i", &controller, &enabled) != 2) {
-
- if (feof(f))
- break;
-
- log_error("Failed to parse /proc/cgroups.");
- r = -EIO;
- goto finish;
- }
-
- if (!enabled) {
- free(controller);
- continue;
- }
-
- r = set_put(controllers, controller);
- if (r < 0) {
- log_error("Failed to add controller to set.");
- free(controller);
- goto finish;
- }
- }
+ r = cg_kernel_controllers(controllers);
+ if (r < 0)
+ return log_error_errno(r, "Failed to enumerate cgroup controllers: %m");
for (;;) {
- MountPoint p;
- char *controller, *where, *options;
+ _cleanup_free_ char *options = NULL, *controller = NULL, *where = NULL;
+ MountPoint p = {
+ .what = "cgroup",
+ .type = "cgroup",
+ .flags = MS_NOSUID|MS_NOEXEC|MS_NODEV,
+ .mode = MNT_IN_CONTAINER,
+ };
char ***k = NULL;
controller = set_steal_first(controllers);
for (i = *k, j = *k; *i; i++) {
if (!streq(*i, controller)) {
- char *t;
+ _cleanup_free_ char *t;
t = set_remove(controllers, *i);
if (!t) {
free(*i);
continue;
}
- free(t);
}
*(j++) = *i;
*j = NULL;
options = strv_join(*k, ",");
- if (!options) {
- free(controller);
- r = log_oom();
- goto finish;
- }
-
+ if (!options)
+ return log_oom();
} else {
options = controller;
controller = NULL;
}
where = strappend("/sys/fs/cgroup/", options);
- if (!where) {
- free(options);
- r = log_oom();
- goto finish;
- }
+ if (!where)
+ return log_oom();
- zero(p);
- p.what = "cgroup";
p.where = where;
- p.type = "cgroup";
p.options = options;
- p.flags = MS_NOSUID|MS_NOEXEC|MS_NODEV;
- p.mode = MNT_IN_CONTAINER;
r = mount_one(&p, true);
- free(controller);
- free(where);
-
- if (r < 0) {
- free(options);
- goto finish;
- }
+ if (r < 0)
+ return r;
if (r > 0 && k && *k) {
char **i;
for (i = *k; *i; i++) {
- char *t;
+ _cleanup_free_ char *t = NULL;
t = strappend("/sys/fs/cgroup/", *i);
- if (!t) {
- r = log_oom();
- free(options);
- goto finish;
- }
+ if (!t)
+ return log_oom();
r = symlink(options, t);
- free(t);
-
- if (r < 0 && errno != EEXIST) {
- log_error("Failed to create symlink: %m");
- r = -errno;
- free(options);
- goto finish;
- }
+ if (r < 0 && errno != EEXIST)
+ return log_error_errno(errno, "Failed to create symlink %s: %m", t);
}
}
-
- free(options);
}
- r = 0;
-
-finish:
- set_free_free(controllers);
-
- fclose(f);
+ /* Now that we mounted everything, let's make the tmpfs the
+ * cgroup file systems are mounted into read-only. */
+ (void) mount("tmpfs", "/sys/fs/cgroup", "tmpfs", MS_REMOUNT|MS_NOSUID|MS_NOEXEC|MS_NODEV|MS_STRICTATIME|MS_RDONLY, "mode=755");
- return r;
+ return 0;
}
+#if defined(HAVE_SELINUX) || defined(HAVE_SMACK)
static int nftw_cb(
const char *fpath,
const struct stat *sb,
return FTW_CONTINUE;
};
+#endif
int mount_setup(bool loaded_policy) {
-
- static const char relabel[] =
- "/run/initramfs/root-fsck\0"
- "/run/initramfs/shutdown\0";
-
- int r;
unsigned i;
- const char *j;
+ int r = 0;
for (i = 0; i < ELEMENTSOF(mount_table); i ++) {
- r = mount_one(mount_table + i, true);
+ int j;
- if (r < 0)
- return r;
+ j = mount_one(mount_table + i, loaded_policy);
+ if (r == 0)
+ r = j;
}
+ if (r < 0)
+ return r;
+
+#if defined(HAVE_SELINUX) || defined(HAVE_SMACK)
/* Nodes in devtmpfs and /run need to be manually updated for
* the appropriate labels, after mounting. The other virtual
* API file systems like /sys and /proc do not need that, they
nftw("/dev", nftw_cb, 64, FTW_MOUNT|FTW_PHYS|FTW_ACTIONRETVAL);
nftw("/run", nftw_cb, 64, FTW_MOUNT|FTW_PHYS|FTW_ACTIONRETVAL);
- /* Explicitly relabel these */
- NULSTR_FOREACH(j, relabel)
- label_fix(j, true, false);
-
after_relabel = now(CLOCK_MONOTONIC);
log_info("Relabelled /dev and /run in %s.",
format_timespan(timespan, sizeof(timespan), after_relabel - before_relabel, 0));
}
+#endif
/* Create a few default symlinks, which are normally created
* by udevd, but some scripts might need them before we start
* propagation mode to private if needed. */
if (detect_container(NULL) <= 0)
if (mount(NULL, "/", NULL, MS_REC|MS_SHARED, NULL) < 0)
- log_warning("Failed to set up the root directory for shared mount propagation: %m");
+ log_warning_errno(errno, "Failed to set up the root directory for shared mount propagation: %m");
/* Create a few directories we always want around, Note that
* sd_booted() checks for /run/systemd/system, so this mkdir