X-Git-Url: http://www.chiark.greenend.org.uk/ucgi/~ianmdlvl/git?p=elogind.git;a=blobdiff_plain;f=src%2Fcore%2Fmount-setup.c;h=f38f1ee8e72c6ffd0800c1536bb94b5963b84931;hp=eb9641fbcbe199686576e5c87f910cc9f8bf4217;hb=109cb9ef1b95f791b8cad2fa2459882c122edc54;hpb=ecdacd4b2fca2153f35b689281ba38107c84f528

diff --git a/src/core/mount-setup.c b/src/core/mount-setup.c
index eb9641fbc..f38f1ee8e 100644
--- a/src/core/mount-setup.c
+++ b/src/core/mount-setup.c
@@ -1,5 +1,3 @@
-/*-*- Mode: C; c-basic-offset: 8; indent-tabs-mode: nil -*-*/
-
 /***
   This file is part of systemd.
 
@@ -19,31 +17,35 @@
   along with systemd; If not, see <http://www.gnu.org/licenses/>.
 ***/
 
-#include <sys/mount.h>
 #include <errno.h>
-#include <sys/stat.h>
+#include <ftw.h>
 #include <stdlib.h>
-#include <string.h>
-#include <assert.h>
+#include <sys/mount.h>
 #include <unistd.h>
-#include <ftw.h>
 
-#include "mount-setup.h"
-#include "dev-setup.h"
-#include "log.h"
-#include "macro.h"
-#include "util.h"
+#include "alloc-util.h"
+//#include "bus-util.h"
+#include "cgroup-util.h"
+//#include "dev-setup.h"
+//#include "efivars.h"
+#include "fs-util.h"
 #include "label.h"
-#include "set.h"
-#include "strv.h"
+//#include "log.h"
+#include "macro.h"
+//#include "missing.h"
 #include "mkdir.h"
+#include "mount-setup.h"
+#include "mount-util.h"
 #include "path-util.h"
-#include "missing.h"
+//#include "set.h"
+//#include "smack-util.h"
+//#include "strv.h"
+#include "user-util.h"
+//#include "util.h"
 #include "virt.h"
-#include "efivars.h"
-#include "smack-util.h"
-#include "def.h"
-#include "cgroup-util.h"
+
+/// Additional includes needed by elogind
+#include "string-util.h"
 
 typedef enum MountMode {
         MNT_NONE  =        0,
@@ -72,6 +74,7 @@ typedef struct MountPoint {
 #endif
 
 static const MountPoint mount_table[] = {
+#if 0 /// UNNEEDED by elogind
         { "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,
@@ -96,36 +99,41 @@ static const MountPoint mount_table[] = {
 #endif
         { "tmpfs",       "/run",                      "tmpfs",      "mode=755",                MS_NOSUID|MS_NODEV|MS_STRICTATIME,
           NULL,          MNT_FATAL|MNT_IN_CONTAINER },
+        { "cgroup",      "/sys/fs/cgroup",            "cgroup2",    NULL,                      MS_NOSUID|MS_NOEXEC|MS_NODEV,
+          cg_is_unified_wanted, MNT_FATAL|MNT_IN_CONTAINER },
+#endif // 0
         { "tmpfs",       "/sys/fs/cgroup",            "tmpfs",      "mode=755",                MS_NOSUID|MS_NOEXEC|MS_NODEV|MS_STRICTATIME,
-          NULL,          MNT_FATAL|MNT_IN_CONTAINER },
+          cg_is_legacy_wanted, MNT_FATAL|MNT_IN_CONTAINER },
+#if 0 /// UNNEEDED by elogind
+        { "cgroup",      "/sys/fs/cgroup/systemd",    "cgroup2",    NULL,                      MS_NOSUID|MS_NOEXEC|MS_NODEV,
+          cg_is_unified_systemd_controller_wanted, MNT_IN_CONTAINER },
         { "cgroup",      "/sys/fs/cgroup/systemd",    "cgroup",     "none,name=systemd,xattr", MS_NOSUID|MS_NOEXEC|MS_NODEV,
-          NULL,          MNT_IN_CONTAINER           },
+          cg_is_legacy_systemd_controller_wanted, MNT_IN_CONTAINER  },
         { "cgroup",      "/sys/fs/cgroup/systemd",    "cgroup",     "none,name=systemd",       MS_NOSUID|MS_NOEXEC|MS_NODEV,
-          NULL,          MNT_FATAL|MNT_IN_CONTAINER },
+          cg_is_legacy_systemd_controller_wanted, MNT_IN_CONTAINER  },
+#else
+        { "cgroup",      "/sys/fs/cgroup/elogind",    "cgroup",     "none,name=elogind,xattr", MS_NOSUID|MS_NOEXEC|MS_NODEV,
+          cg_is_legacy_systemd_controller_wanted, MNT_IN_CONTAINER  },
+        { "cgroup",      "/sys/fs/cgroup/elogind",    "cgroup",     "none,name=elogind",       MS_NOSUID|MS_NOEXEC|MS_NODEV,
+          cg_is_legacy_systemd_controller_wanted, MNT_IN_CONTAINER  },
+#endif // 0
+#if 0 /// UNNEEDED by elogind
         { "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                   },
 #endif
-#ifdef ENABLE_KDBUS
-        { "kdbusfs",    "/sys/fs/kdbus",             "kdbusfs",    NULL, MS_NOSUID|MS_NOEXEC|MS_NODEV,
-          NULL,       MNT_IN_CONTAINER },
-#endif
+#endif // 0
 };
 
+#if 0 /// UNNEEDED by elogind
 /* These are API file systems that might be mounted by other software,
  * we just list them here so that we know that we should ignore them */
 
 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"
@@ -153,6 +161,7 @@ bool mount_point_ignore(const char *path) {
 
         return false;
 }
+#endif // 0
 
 static int mount_one(const MountPoint *p, bool relabel) {
         int r;
@@ -164,25 +173,26 @@ static int mount_one(const MountPoint *p, bool relabel) {
 
         /* Relabel first, just in case */
         if (relabel)
-                label_fix(p->where, true, true);
-
-        r = path_is_mount_point(p->where, true);
-        if (r < 0)
-                return r;
+                (void) label_fix(p->where, true, true);
 
+        r = path_is_mount_point(p->where, NULL, AT_SYMLINK_FOLLOW);
+        if (r < 0 && r != -ENOENT) {
+                log_full_errno((p->mode & MNT_FATAL) ? LOG_ERR : LOG_DEBUG, r, "Failed to determine whether %s is a mount point: %m", p->where);
+                return (p->mode & MNT_FATAL) ? r : 0;
+        }
         if (r > 0)
                 return 0;
 
         /* Skip securityfs in a container */
-        if (!(p->mode & MNT_IN_CONTAINER) && detect_container(NULL) > 0)
+        if (!(p->mode & MNT_IN_CONTAINER) && detect_container() > 0)
                 return 0;
 
         /* The access mode here doesn't really matter too much, since
          * the mounted file system will take precedence anyway. */
         if (relabel)
-                mkdir_p_label(p->where, 0755);
+                (void) mkdir_p_label(p->where, 0755);
         else
-                mkdir_p(p->where, 0755);
+                (void) mkdir_p(p->where, 0755);
 
         log_debug("Mounting %s to %s of type %s with options %s.",
                   p->what,
@@ -195,40 +205,48 @@ static int mount_one(const MountPoint *p, bool relabel) {
                   p->type,
                   p->flags,
                   p->options) < 0) {
-                log_full((p->mode & MNT_FATAL) ? LOG_ERR : LOG_DEBUG, "Failed to mount %s at %s: %m", p->type, p->where);
+                log_full_errno((p->mode & MNT_FATAL) ? LOG_ERR : LOG_DEBUG, errno, "Failed to mount %s at %s: %m", p->type, p->where);
                 return (p->mode & MNT_FATAL) ? -errno : 0;
         }
 
         /* Relabel again, since we now mounted something fresh here */
         if (relabel)
-                label_fix(p->where, false, false);
+                (void) label_fix(p->where, false, false);
 
         return 1;
 }
 
-int mount_setup_early(void) {
+static int mount_points_setup(unsigned n, bool loaded_policy) {
         unsigned i;
         int r = 0;
 
-        assert_cc(N_EARLY_MOUNT <= ELEMENTSOF(mount_table));
-
-        /* Do a minimal mount of /proc and friends to enable the most
-         * basic stuff, such as SELinux */
-        for (i = 0; i < N_EARLY_MOUNT; i ++)  {
+        for (i = 0; i < n; i ++) {
                 int j;
 
-                j = mount_one(mount_table + i, false);
-                if (r == 0)
+                j = mount_one(mount_table + i, loaded_policy);
+                if (j != 0 && r >= 0)
                         r = j;
         }
 
         return r;
 }
 
+#if 0 /// UNNEEDED by elogind
+int mount_setup_early(void) {
+        assert_cc(N_EARLY_MOUNT <= ELEMENTSOF(mount_table));
+
+        /* Do a minimal mount of /proc and friends to enable the most
+         * basic stuff, such as SELinux */
+        return mount_points_setup(N_EARLY_MOUNT, false);
+}
+
 int mount_cgroup_controllers(char ***join_controllers) {
         _cleanup_set_free_free_ Set *controllers = NULL;
         int r;
 
+        if (!cg_is_legacy_wanted())
+                return 0;
+
         /* Mount all available cgroup controllers that are built into the kernel. */
 
         controllers = set_new(&string_hash_ops);
@@ -308,7 +326,17 @@ int mount_cgroup_controllers(char ***join_controllers) {
                                         return log_oom();
 
                                 r = symlink(options, t);
-                                if (r < 0 && errno != EEXIST)
+                                if (r >= 0) {
+#ifdef SMACK_RUN_LABEL
+                                        _cleanup_free_ char *src;
+                                        src = strappend("/sys/fs/cgroup/", options);
+                                        if (!src)
+                                                return log_oom();
+                                        r = mac_smack_copy(t, src);
+                                        if (r < 0 && r != -EOPNOTSUPP)
+                                                return log_error_errno(r, "Failed to copy smack label from %s to %s: %m", src, t);
+#endif
+                                } else if (errno != EEXIST)
                                         return log_error_errno(errno, "Failed to create symlink %s: %m", t);
                         }
                 }
@@ -344,22 +372,16 @@ static int nftw_cb(
         return FTW_CONTINUE;
 };
 #endif
+#endif // 0
 
 int mount_setup(bool loaded_policy) {
-        unsigned i;
         int r = 0;
 
-        for (i = 0; i < ELEMENTSOF(mount_table); i ++) {
-                int j;
-
-                j = mount_one(mount_table + i, loaded_policy);
-                if (r == 0)
-                        r = j;
-        }
-
+        r = mount_points_setup(ELEMENTSOF(mount_table), loaded_policy);
         if (r < 0)
                 return r;
 
+#if 0 /// elogind does not control /, /dev, /run and /run/systemd/* are setup elsewhere.
 #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
@@ -372,6 +394,7 @@ int mount_setup(bool loaded_policy) {
                 before_relabel = now(CLOCK_MONOTONIC);
 
                 nftw("/dev", nftw_cb, 64, FTW_MOUNT|FTW_PHYS|FTW_ACTIONRETVAL);
+                nftw("/dev/shm", nftw_cb, 64, FTW_MOUNT|FTW_PHYS|FTW_ACTIONRETVAL);
                 nftw("/run", nftw_cb, 64, FTW_MOUNT|FTW_PHYS|FTW_ACTIONRETVAL);
 
                 after_relabel = now(CLOCK_MONOTONIC);
@@ -384,26 +407,33 @@ int mount_setup(bool loaded_policy) {
         /* Create a few default symlinks, which are normally created
          * by udevd, but some scripts might need them before we start
          * udevd. */
-        dev_setup(NULL);
-
-        /* Mark the root directory as shared in regards to mount
-         * propagation. The kernel defaults to "private", but we think
-         * it makes more sense to have a default of "shared" so that
-         * nspawn and the container tools work out of the box. If
-         * specific setups need other settings they can reset the
-         * propagation mode to private if needed. */
-        if (detect_container(NULL) <= 0)
+        dev_setup(NULL, UID_INVALID, GID_INVALID);
+
+        /* Mark the root directory as shared in regards to mount propagation. The kernel defaults to "private", but we
+         * think it makes more sense to have a default of "shared" so that nspawn and the container tools work out of
+         * the box. If specific setups need other settings they can reset the propagation mode to private if
+         * needed. Note that we set this only when we are invoked directly by the kernel. If we are invoked by a
+         * container manager we assume the container manager knows what it is doing (for example, because it set up
+         * some directories with different propagation modes). */
+        if (detect_container() <= 0)
                 if (mount(NULL, "/", NULL, MS_REC|MS_SHARED, NULL) < 0)
                         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
-         * really needs to stay for good, otherwise software that
-         * copied sd-daemon.c into their sources will misdetect
-         * systemd. */
-        mkdir_label("/run/systemd", 0755);
-        mkdir_label("/run/systemd/system", 0755);
-        mkdir_label("/run/systemd/inaccessible", 0000);
+#endif // 0
+        /* Create a few directories we always want around, Note that sd_booted() checks for /run/systemd/system, so
+         * this mkdir really needs to stay for good, otherwise software that copied sd-daemon.c into their sources will
+         * misdetect systemd. */
+        (void) mkdir_label("/run/systemd", 0755);
+        (void) mkdir_label("/run/systemd/system", 0755);
+
+        /* Set up inaccessible items */
+        (void) mkdir_label("/run/systemd/inaccessible", 0000);
+        (void) mknod("/run/systemd/inaccessible/reg", S_IFREG | 0000, 0);
+        (void) mkdir_label("/run/systemd/inaccessible/dir", 0000);
+        (void) mknod("/run/systemd/inaccessible/chr", S_IFCHR | 0000, makedev(0, 0));
+        (void) mknod("/run/systemd/inaccessible/blk", S_IFBLK | 0000, makedev(0, 0));
+        (void) mkfifo("/run/systemd/inaccessible/fifo", 0000);
+        (void) mknod("/run/systemd/inaccessible/sock", S_IFSOCK | 0000, 0);
 
         return 0;
 }