README

   1 systemd System and Service Manager
   2
   3 WEB SITE:
   4         https://systemd.io
   5
   6 GIT:
   7         git@github.com:systemd/systemd.git
   8         https://github.com/systemd/systemd
   9
  10 MAILING LIST:
  11         https://lists.freedesktop.org/mailman/listinfo/systemd-devel
  12
  13 IRC:
  14         #systemd on irc.libera.chat
  15
  16 BUG REPORTS:
  17         https://github.com/systemd/systemd/issues
  18
  19 OLDER DOCUMENTATION:
  20         https://0pointer.de/blog/projects/systemd.html
  21         https://www.freedesktop.org/wiki/Software/systemd
  22
  23 AUTHOR:
  24         Lennart Poettering
  25         Kay Sievers
  26         ...and many others
  27
  28 LICENSE:
  29         LGPL-2.1-or-later for all code, exceptions noted in LICENSES/README.md
  30
  31 REQUIREMENTS:
  32         Linux kernel ≥ 3.15
  33                      ≥ 4.3 for ambient capabilities
  34                      ≥ 4.5 for pids controller in cgroup v2
  35                      ≥ 4.6 for cgroup namespaces
  36                      ≥ 4.9 for RENAME_NOREPLACE support in vfat
  37                      ≥ 4.10 for cgroup-bpf egress and ingress hooks
  38                      ≥ 4.15 for cgroup-bpf device hook and cpu controller in cgroup v2
  39                      ≥ 4.17 for cgroup-bpf socket address hooks
  40                      ≥ 4.20 for PSI (used by systemd-oomd)
  41                      ≥ 5.3 for bounded loops in BPF program
  42                      ≥ 5.4 for signed Verity images
  43                      ≥ 5.7 for BPF links and the BPF LSM hook
  44
  45         ⛔ Kernel versions below 3.15 ("minimum baseline") are not supported at
  46         all, and are missing required functionality (e.g. CLOCK_BOOTTIME
  47         support for timerfd_create()).
  48
  49         ⚠️ Kernel versions below 4.15 ("recommended baseline") have significant
  50         gaps in functionality and are not recommended for use with this version
  51         of systemd (e.g. lack sufficiently comprehensive and working cgroupv2
  52         support). Taint flag 'old-kernel' will be set. systemd will most likely
  53         still function, but upstream support and testing are limited.
  54
  55         Kernel Config Options:
  56           CONFIG_DEVTMPFS
  57           CONFIG_CGROUPS (it is OK to disable all controllers)
  58           CONFIG_INOTIFY_USER
  59           CONFIG_SIGNALFD
  60           CONFIG_TIMERFD
  61           CONFIG_EPOLL
  62           CONFIG_UNIX (it requires CONFIG_NET, but every other flag in it is not necessary)
  63           CONFIG_SYSFS
  64           CONFIG_PROC_FS
  65           CONFIG_FHANDLE (libudev, mount and bind mount handling)
  66
  67         udev will fail to work with the legacy sysfs layout:
  68           CONFIG_SYSFS_DEPRECATED=n
  69
  70         Legacy hotplug slows down the system and confuses udev:
  71           CONFIG_UEVENT_HELPER_PATH=""
  72
  73         Userspace firmware loading is not supported and should be disabled in
  74         the kernel:
  75           CONFIG_FW_LOADER_USER_HELPER=n
  76
  77         Some udev rules and virtualization detection relies on it:
  78           CONFIG_DMIID
  79
  80         Support for some SCSI devices serial number retrieval, to create
  81         additional symlinks in /dev/disk/ and /dev/tape:
  82           CONFIG_BLK_DEV_BSG
  83
  84         Required for PrivateNetwork= in service units:
  85           CONFIG_NET_NS
  86         Note that systemd-localed.service and other systemd units use
  87         PrivateNetwork so this is effectively required.
  88
  89         Required for PrivateUsers= in service units:
  90           CONFIG_USER_NS
  91
  92         Optional but strongly recommended:
  93           CONFIG_IPV6
  94           CONFIG_AUTOFS_FS
  95           CONFIG_TMPFS_XATTR
  96           CONFIG_{TMPFS,EXT4_FS,XFS,BTRFS_FS,...}_POSIX_ACL
  97           CONFIG_SECCOMP
  98           CONFIG_SECCOMP_FILTER (required for seccomp support)
  99           CONFIG_KCMP (for the kcmp() syscall, used to be under
 100                        CONFIG_CHECKPOINT_RESTORE before ~5.12)
 101
 102         Required for CPUShares= in resource control unit settings:
 103           CONFIG_CGROUP_SCHED
 104           CONFIG_FAIR_GROUP_SCHED
 105
 106         Required for CPUQuota= in resource control unit settings:
 107           CONFIG_CFS_BANDWIDTH
 108
 109         Required for IPAddressDeny=, IPAddressAllow=, IPIngressFilterPath=,
 110         IPEgressFilterPath= in resource control unit settings unit settings:
 111           CONFIG_BPF
 112           CONFIG_BPF_SYSCALL
 113           CONFIG_BPF_JIT
 114           CONFIG_HAVE_EBPF_JIT
 115           CONFIG_CGROUP_BPF
 116
 117         Required for SocketBind{Allow|Deny}=, RestrictNetworkInterfaces= in
 118         resource control unit settings:
 119           CONFIG_BPF
 120           CONFIG_BPF_SYSCALL
 121           CONFIG_BPF_JIT
 122           CONFIG_HAVE_EBPF_JIT
 123           CONFIG_CGROUP_BPF
 124
 125         For UEFI systems:
 126           CONFIG_EFIVAR_FS
 127           CONFIG_EFI_PARTITION
 128
 129         Required for signed Verity images support:
 130           CONFIG_DM_VERITY_VERIFY_ROOTHASH_SIG
 131         Required to verify signed Verity images using keys enrolled in the MoK
 132         (Machine-Owner Key) keyring:
 133           CONFIG_DM_VERITY_VERIFY_ROOTHASH_SIG_SECONDARY_KEYRING
 134           CONFIG_IMA_ARCH_POLICY
 135           CONFIG_INTEGRITY_MACHINE_KEYRING
 136
 137         Required for reading credentials from SMBIOS:
 138           CONFIG_DMI
 139           CONFIG_DMI_SYSFS
 140
 141         Required for RestrictFileSystems= in service units:
 142           CONFIG_BPF
 143           CONFIG_BPF_SYSCALL
 144           CONFIG_BPF_LSM
 145           CONFIG_DEBUG_INFO_BTF
 146           CONFIG_LSM="...,bpf" or kernel booted with lsm="...,bpf".
 147
 148         We recommend to turn off Real-Time group scheduling in the kernel when
 149         using systemd. RT group scheduling effectively makes RT scheduling
 150         unavailable for most userspace, since it requires explicit assignment of
 151         RT budgets to each unit whose processes making use of RT. As there's no
 152         sensible way to assign these budgets automatically this cannot really be
 153         fixed, and it's best to disable group scheduling hence:
 154            CONFIG_RT_GROUP_SCHED=n
 155
 156         It's a good idea to disable the implicit creation of networking bonding
 157         devices by the kernel networking bonding module, so that the
 158         automatically created "bond0" interface doesn't conflict with any such
 159         device created by systemd-networkd (or other tools). Ideally there would
 160         be a kernel compile-time option for this, but there currently isn't. The
 161         next best thing is to make this change through a modprobe.d drop-in.
 162         This is shipped by default, see modprobe.d/systemd.conf.
 163
 164         Required for systemd-nspawn:
 165           CONFIG_DEVPTS_MULTIPLE_INSTANCES or Linux kernel >= 4.7
 166
 167         Required for systemd-oomd:
 168           CONFIG_PSI
 169
 170         Note that kernel auditing is broken when used with systemd's container
 171         code. When using systemd in conjunction with containers, please make
 172         sure to either turn off auditing at runtime using the kernel command
 173         line option "audit=0", or turn it off at kernel compile time using:
 174           CONFIG_AUDIT=n
 175
 176         If systemd is compiled with libseccomp support on architectures which do
 177         not use socketcall() and where seccomp is supported (this effectively
 178         means x86-64 and ARM, but excludes 32-bit x86!), then nspawn will now
 179         install a work-around seccomp filter that makes containers boot even
 180         with audit being enabled. This works correctly only on kernels 3.14 and
 181         newer though. TL;DR: turn audit off, still.
 182
 183         glibc >= 2.16
 184         libcap
 185         libmount >= 2.30 (from util-linux)
 186                 (util-linux *must* be built without --enable-libmount-support-mtab)
 187         libseccomp >= 2.3.1 (optional)
 188         libblkid >= 2.24 (from util-linux) (optional)
 189         libkmod >= 15 (optional)
 190         PAM >= 1.1.2 (optional)
 191         libcryptsetup (optional), >= 2.3.0 required for signed Verity images support
 192         libaudit (optional)
 193         libacl (optional)
 194         libbpf >= 0.1.0 (optional)
 195         libfdisk >= 2.32 (from util-linux) (optional)
 196         libselinux (optional)
 197         liblzma (optional)
 198         liblz4 >= 1.3.0 / 130 (optional)
 199         libzstd >= 1.4.0 (optional)
 200         libgcrypt (optional)
 201         libqrencode (optional)
 202         libmicrohttpd (optional)
 203         libidn2 or libidn (optional)
 204         gnutls >= 3.1.4 (optional, >= 3.6.0 is required to support DNS-over-TLS with gnutls)
 205         openssl >= 1.1.0 (optional, required to support DNS-over-TLS with openssl)
 206         elfutils >= 158 (optional)
 207         polkit (optional)
 208         tzdata >= 2014f (optional)
 209         pkg-config
 210         gperf
 211         docbook-xsl (optional, required for documentation)
 212         xsltproc    (optional, required for documentation)
 213         python >= 3.7 (required by meson too, >= 3.9 is required for ukify)
 214         python-jinja2
 215         python-pefile (optional, required for ukify)
 216         python-lxml (optional, required to build the indices)
 217         pyelftools (optional, required for systemd-boot)
 218         meson >= 0.60.0
 219         ninja
 220         gcc >= 4.7
 221         awk, sed, grep, and similar tools
 222         clang >= 10.0, llvm >= 10.0 (optional, required to build BPF programs
 223                 from source code in C)
 224
 225         During runtime, you need the following additional
 226         dependencies:
 227
 228         util-linux >= v2.27.1 required (including but not limited to: mount,
 229                                         umount, swapon, swapoff, sulogin,
 230                                         agetty, fsck)
 231         dbus >= 1.4.0 (strictly speaking optional, but recommended)
 232                 NOTE: If using dbus < 1.9.18, you should override the default
 233                 policy directory (--with-dbuspolicydir=/etc/dbus-1/system.d).
 234         dracut (optional)
 235         polkit (optional)
 236
 237         To build in directory build/:
 238           meson setup build/ && ninja -C build/
 239
 240         Any configuration options can be specified as -Darg=value... arguments
 241         to meson. After the build directory is initially configured, meson will
 242         refuse to run again, and options must be changed with:
 243           meson configure -Darg=value build/
 244         meson configure without any arguments will print out available options and
 245         their current values.
 246
 247         Useful commands:
 248           ninja -C build -v some/target
 249           meson test -C build/
 250           sudo meson install -C build/ --no-rebuild
 251           DESTDIR=... meson install -C build/
 252
 253         A tarball can be created with:
 254           v=250 && git archive --prefix=systemd-$v/ v$v | zstd >systemd-$v.tar.zstd
 255
 256         When systemd-hostnamed is used, it is strongly recommended to install
 257         nss-myhostname to ensure that, in a world of dynamically changing
 258         hostnames, the hostname stays resolvable under all circumstances. In
 259         fact, systemd-hostnamed will warn if nss-myhostname is not installed.
 260
 261         nss-systemd must be enabled on systemd systems, as that's required for
 262         DynamicUser= to work. Note that we ship services out-of-the-box that
 263         make use of DynamicUser= now, hence enabling nss-systemd is not
 264         optional.
 265
 266         Note that the build prefix for systemd must be /usr/. (Moreover, packages
 267         systemd relies on — such as D-Bus — really should use the same prefix,
 268         otherwise you are on your own.) Split-usr and unmerged-usr systems are no
 269         longer supported, and moving everything under /usr/ is required. Systems
 270         with a separate /usr/ partition must mount it before transitioning into it
 271         (i.e.: from the initrd). For more information see:
 272         https://www.freedesktop.org/wiki/Software/systemd/separate-usr-is-broken
 273         https://www.freedesktop.org/wiki/Software/systemd/TheCaseForTheUsrMerge
 274
 275         Additional packages are necessary to run some tests:
 276         - nc                 (used by test/TEST-12-ISSUE-3171)
 277         - python             (test-udev which is installed is in python)
 278         - python-pyparsing
 279         - python-evdev       (used by hwdb parsing tests)
 280         - strace             (used by test/test-functions)
 281         - capsh              (optional, used by test-execute)
 282
 283 POLICY FOR SUPPORT OF DISTRIBUTIONS AND ARCHITECTURES:
 284         systemd main branch and latest major or stable releases are generally
 285         expected to compile on current versions of popular distributions (at
 286         least all non-EOL versions of Fedora, Debian unstable/testing/stable,
 287         latest Ubuntu LTS and non-LTS releases, openSUSE Tumbleweed/Leap,
 288         CentOS Stream 8 and 9, up-to-date Arch, etc.)  We will generally
 289         attempt to support also other non-EOL versions of various distros.
 290         Features which would break compilation on slightly older distributions
 291         will only be introduced if there are significant reasons for this
 292         (i.e. supporting them interferes with development or requires too many
 293         resources to support). In some cases backports of specific libraries or
 294         tools might be required.
 295
 296         The policy is similar for architecture support. systemd is regularly
 297         tested on popular architectures (currently amd64, i386, arm64, ppc64el,
 298         and s390x), but should compile and work also on other architectures, for
 299         which support has been added. systemd will emit warnings when
 300         architecture-specific constants are not defined.
 301
 302 STATIC COMPILATION AND "STANDALONE" BINARIES:
 303         systemd provides a public shared libraries libsystemd.so and
 304         libudev.so. The latter is deprecated, and the sd-device APIs in
 305         libsystemd should be used instead for new code. In addition, systemd is
 306         built with a private shared library, libsystemd-shared-<suffix>.so,
 307         that also includes the libsystemd code, and by default most systemd
 308         binaries are linked to it. Using shared libraries saves disk space and
 309         memory at runtime, because only one copy of the code is needed.
 310
 311         It is possible to build static versions of systemd public shared
 312         libraries (via the configuration options '-Dstatic-libsystemd' and
 313         '-Dstatic-libudev'). This allows the libsystemd and libudev code to be
 314         linked statically into programs. Note that mixing & matching different
 315         versions of libsystemd and systemd is generally not recommended, since
 316         various of its APIs wrap internal state and protocols of systemd
 317         (e.g. logind and udev databases), which are not considered
 318         stable. Hence, using static libraries is not recommended since it
 319         generally means that version of the static libsystemd linked into
 320         applications and the host systemd are not in sync, and will thus create
 321         compatibility problems.
 322
 323         In addition, it is possible to disable the use of
 324         libsystemd-shared-<suffix>.so for various components (via the
 325         configuration options '-Dlink-*-shared'). In this mode, the libsystemd
 326         and libsystemd-shared code is linked statically into selected
 327         binaries. This option is intended for systems where some of the
 328         components are intended to be delivered independently of the main
 329         systemd package. Finally, some binaries can be compiled in a second
 330         version (via the configuration option '-Dstandalone-binaries'). The
 331         version suffixed with ".standalone" has the libsystemd and
 332         libsystemd-shared code linked statically. Those binaries are intended
 333         as replacements to be used in limited installations where the full
 334         systemd suite is not installed. Yet another option is to rebuild
 335         systemd with a different '-Dshared-lib-tag' setting, allowing different
 336         systemd binaries to be linked to instances of the private shared
 337         library that can be installed in parallel.
 338
 339         Again: Using the default shared linking is recommended, building static
 340         or "standalone" versions is not. Mixing versions of systemd components
 341         that would normally be built and used together (in particular various
 342         daemons and the manager) is not recommended: we do not test such
 343         combinations upstream and cannot provide support. Distributors making
 344         use of those options are responsible if things do not work as expected.
 345
 346 USERS AND GROUPS:
 347         Default udev rules use the following standard system group names, which
 348         need to be resolvable by getgrnam() at any time, even in the very early
 349         boot stages, where no other databases and network are available:
 350
 351         audio, cdrom, dialout, disk, input, kmem, kvm, lp, render, tape, tty, video
 352
 353         During runtime, the journal daemon requires the "systemd-journal" system
 354         group to exist. New journal files will be readable by this group (but
 355         not writable), which may be used to grant specific users read access. In
 356         addition, system groups "wheel" and "adm" will be given read-only access
 357         to journal files using systemd-tmpfiles-setup.service.
 358
 359         The journal remote daemon requires the "systemd-journal-remote" system
 360         user and group to exist. During execution this network facing service
 361         will drop privileges and assume this uid/gid for security reasons.
 362
 363         Similarly, the network management daemon requires the "systemd-network"
 364         system user and group to exist.
 365
 366         Similarly, the name resolution daemon requires the "systemd-resolve"
 367         system user and group to exist.
 368
 369         Similarly, the coredump support requires the "systemd-coredump" system
 370         user and group to exist.
 371
 372 GLIBC NSS:
 373         systemd ships with four glibc NSS modules:
 374
 375         nss-myhostname resolves the local hostname to locally configured IP
 376         addresses, as well as "localhost" to 127.0.0.1/::1.
 377
 378         nss-resolve enables DNS resolution via the systemd-resolved DNS/LLMNR
 379         caching stub resolver "systemd-resolved".
 380
 381         nss-mymachines enables resolution of all local containers registered
 382         with machined to their respective IP addresses.
 383
 384         nss-systemd enables resolution of users/group registered via the
 385         User/Group Record Lookup API (https://systemd.io/USER_GROUP_API),
 386         including all dynamically allocated service users. (See the
 387         DynamicUser= setting in unit files.)
 388
 389         To make use of these NSS modules, please add them to the "hosts:",
 390         "passwd:", "group:", "shadow:" and "gshadow:" lines in
 391         /etc/nsswitch.conf.
 392
 393         The four modules should be used in the following order:
 394
 395                 passwd: files systemd
 396                 group: files [SUCCESS=merge] systemd
 397                 shadow: files systemd
 398                 gshadow: files systemd
 399                 hosts: mymachines resolve [!UNAVAIL=return] files myhostname dns
 400
 401 SYSV INIT.D SCRIPTS:
 402         When calling "systemctl enable/disable/is-enabled" on a unit which is a
 403         SysV init.d script, it calls /usr/lib/systemd/systemd-sysv-install;
 404         this needs to translate the action into the distribution specific
 405         mechanism such as chkconfig or update-rc.d. Packagers need to provide
 406         this script if you need this functionality (you don't if you disabled
 407         SysV init support).
 408
 409         Please see src/systemctl/systemd-sysv-install.SKELETON for how this
 410         needs to look like, and provide an implementation at the marked places.
 411
 412 WARNINGS and TAINT FLAGS:
 413         systemd requires that the /run mount point exists. systemd also
 414         requires that /var/run is a symlink to /run. Taint flag 'var-run-bad'
 415         will be set when this condition is detected.
 416
 417         Systemd will also warn when the cgroup support is unavailable in the
 418         kernel (taint flag 'cgroups-missing'), the system is using the old
 419         cgroup hierarchy (taint flag 'cgroupsv1'), the hardware clock is
 420         running in non-UTC mode (taint flag 'local-hwclock'), the kernel
 421         overflow UID or GID are not 65534 (taint flags 'overflowuid-not-65534'
 422         and 'overflowgid-not-65534'), the UID or GID range assigned to the
 423         running systemd instance covers less than 0…65534 (taint flags
 424         'short-uid-range' and 'short-gid-range').
 425
 426         Taint conditions are logged during boot, but may also be checked at any
 427         time with:
 428
 429           busctl get-property org.freedesktop.systemd1 /org/freedesktop/systemd1 org.freedesktop.systemd1.Manager Tainted
 430
 431         See org.freedesktop.systemd1(5) for more information.
 432
 433 VALGRIND:
 434         To run systemd under valgrind, compile systemd with the valgrind
 435         development headers available (i.e. valgrind-devel or equivalent).
 436         Otherwise, false positives will be triggered by code which violates
 437         some rules but is actually safe. Note that valgrind generates nice
 438         output only on exit(), hence on shutdown we don't execve()
 439         systemd-shutdown.
 440
 441 STABLE BRANCHES AND BACKPORTS:
 442         Stable branches with backported patches are available in the
 443         systemd-stable repo at https://github.com/systemd/systemd-stable.
 444
 445         Stable branches are started for certain releases of systemd and named
 446         after them, e.g. v238-stable. Stable branches are managed by
 447         distribution maintainers on an as needed basis. See
 448         https://www.freedesktop.org/wiki/Software/systemd/Backports for some
 449         more information and examples.