Bump version for v4.1 release

Signed-off-by: Joachim Wiberg <troglobit@gmail.com>
Revert "Travis-CI: disable clang for Coverity run"
2026-10-12 10:12:32 +00:00 · 2021-06-06 17:30:48 +02:00 · 2021-06-06 17:05:09 +02:00 · 2021-06-06 16:57:52 +02:00 · 2021-06-06 16:50:51 +02:00 · 2021-06-06 16:46:59 +02:00
122 changed files with 7199 additions and 2142 deletions
@@ -27,7 +27,7 @@ addons:
      name: "troglobit/finit"
      description: "finit | Fast & Extensible init for Linux"
    notification_email: troglobit@gmail.com
-    build_command_prepend: "./autogen.sh && ./configure --prefix=/tmp --disable-doc --disable-contrib --enable-watchdog --enable-fallback-shell"
+    build_command_prepend: "./autogen.sh && ./configure --prefix= --disable-doc --disable-contrib --with-watchdog"
    build_command: "make"
    branch_pattern: dev

@@ -42,20 +42,20 @@ install:
 # Custom build script for the time being, no "make test" yet
 script:
  - ./autogen.sh
-  - ./configure --prefix=/tmp --enable-static
+  - ./configure --prefix= --enable-static
  - make clean
  - make V=1 -j5
-  - ./configure --prefix=/tmp --enable-x11-common-plugin --enable-fallback-shell --enable-watchdog
+  - ./configure --prefix= --enable-x11-common-plugin --with-watchdog --with-keventd
  - cat config.log
  - make clean
  - make V=1 -j5
-  - make install-strip
-  - tree /tmp
-  - ldd /tmp/sbin/finit
-  - size /tmp/sbin/finit
-  - size /tmp/sbin/initctl
-  - size /tmp/sbin/reboot
-  - LD_LIBRARY_PATH=/tmp/lib /tmp/sbin/initctl -h
+  - make install-strip DESTDIR=/tmp/finit
+  - tree /tmp/finit
+  - ldd /tmp/finit/sbin/finit
+  - size /tmp/finit/sbin/finit
+  - size /tmp/finit/sbin/initctl
+  - size /tmp/finit/sbin/reboot
+  - LD_LIBRARY_PATH=/tmp/lib /tmp/finit/sbin/initctl -h
  - unshare -V
  - nsenter -V
  - chroot --version
@@ -4,12 +4,112 @@ Change Log
 All relevant changes are documented in this file.


-[4.0][UNRELEASED] - 2021-02-21
------------------------------
+[4.1][] - 2021-06-06
+--------------------
+
+Bug fix release.  Also disables handlers for `SIGINT` and `SIGPWR`, a
+new set of `sys` conditions are instead generated which can be used to
+trigger external programs.
+
+### Changes
+* Change behavior on SIGUSR1 to be compatible with sysvinit and systemd.
+  Previously SIGUSR1 caused Finit to halt, like BusyBox init.  This had
+  "interesting" side effects on Debian systems when coexisting with
+  sysvinit (upgrading/reinstalling causes scripts to `kill -USR1 1`)
+* Change how `contrib/debian/install.sh` sets up a Grub boot entry for
+  finit.  We now modify the $SUPPORTED_INITS variable in `10_linux`
+* Disable default kernel ctrl-alt-delete handler and let Finit instead
+  catch `SIGINT` from kernel to be able to perform a proper reboot.
+  There is no default command for this, you need to set up a task that
+  triggers on `<sys/key/ctrlaltdel>` to issue `initctl reboot`
+* Added keventd to provide `<sys/pwr/ac>` condition to Finit.  keventd
+  is currently only responsible for monitoring `/sys/class/power_supply`
+  for changes to active AC mains power online status.  Enable keventd
+  with `configure --with-keventd`
+* For handling power fail events (from UPS and similar) a process may
+  send `SIGPWR` to PID 1. Finit no longer redirects this to `SIGUSR1`
+  (poweroff).  There is no default command for this, you need to set up
+  a task that triggers on `<sys/pwr/fail>` to issue `initctl poweroff`
+* Built-in Finit getty is now a standalone program
+* Default termios for TTYs now enable `IUTF8` on input
+* If `/bin/login` is not found, Finit now tries `sulogin` before it
+  falls back to an unauthenticated `/bin/sh`
+* Dropped (broken) support for multiple consoles.  Finit now follows
+  the default console selected by the kernel, `/dev/console`
+* Dropped signal handlers for SIGSTOP/TSTP and SIGCONT
+* Added support for `\n`, in addition to `\r`, in "Please press Enter"
+  prompt before starting getty
+* Finit no longer parses `/proc/cmdline` for its options.  Instead all
+  options are by default now read from `argv[]`, like a normal program,
+  this is also what the kernel does by default.  Please note, this may
+  not work if your systems boots with an initramfs (ymmv), for such
+  cases, see `configure --enable-kernel-cmdline`
+* The following plugins are now possible to disable (for containers):
+  `rtc.so`, `urandom.so`, you may also want to disable `hotplug.so`.
+  They are all enabled by default, as in Finit 4.0, but may be moved
+  to external tools or entries in `finit.conf` in Finit 5.0
+* Added support for reading `PRETTY_NAME` from `/etc/os-release` to use
+  as heading in progress output, unless `--with-heading=GREET` is used.
+* Added manual pages for finit(8), initctl(8), and finit.conf(5)
+
+### Fixes
+* Stricter interface name validation in netlink plugin, modeled after
+  the kernel.  Suggested by Coverity Scan
+* Fix problem of re-registering a service as a task.  Previously, if a
+  fundamental change, like type, was made to an active service/run/task
+  it did not take.  Only possible workaround was to remove from config
+* initctl: drop warning when removing a non-existing usr condition
+* initctl: drop confusing `errno 0` when timing out waiting for reply
+* Ensure services in plugins and from Finit main belong to a cgroup
+* Ensure init top-level cgroup remains a leaf group
+* Fix tty parse error for detecting use of external getty
+* Fix default `name:` and `:ID` for tty's, e.g. `ttyS0` now gives
+  `tty:S0` as expected.  This was default for built-in getty already
+* Fix max username (32 chars) in bundled Finit getty
+* The `contrib/*/install.sh` scripts failed to run from tarball
+* Finit no longer forcibly mounts; `/dev`, `/proc`, or `/sys`, instead
+  it checks first if they are already mounted (devtmpfs or container)
+* Fix `/etc/fstab` parser to properly check for 'ro' to not remount the
+  root filesystem at boot.  The wrong field was read, so a root mounted
+  by an initramfs, or by lxc for a container, had their root remounted
+* Fix SIGCHLD handler, `waitpd()` may be interrupted by a signal
+* Reset `starting` flag of services being stopped.  When a service
+  is started and then stopped before it has created its pid file,
+  it could be left forever in the stopping state, unless we reset
+  the starting flag.
+* Fix #170: detect loss of default route when interfaces go down.  This
+  emulates the missing kernel netlink message to remove the condition
+  net/default/route to allow stopping dependent services
+* Fix #171: restore automatic mount of `/dev/shm`, `/dev/pts`, `/run`
+  and `/tmp`, unless mounted already by `/etc/fstab`.  This is what most
+  desktop systems expect PID 1 to do.  Here we also make sure to mount
+  `/run/lock` as a tmpfs as well, with write perms for regular users,
+  this prevents regular users from filling up `/run` and causing DoS.
+* Fix #173: netlink plugin runs out of socket buffer space;
+
+        finit[1]: nl_callback():recv(): No buffer space available
+
+  Fixed by adding support for resync with kernel on `ENOBUFS`.  See
+  netlink(7) for details.  As a spin-off the plugin now supports any
+  number of interfaces and routes on a system.  On resync, the following
+  message is now logged, as a warning:
+
+        finit[1]: nl_callback():busy system, resynchronizing with kernel.
+
+* Fix #174: loss of log messages using combo of prio and facility, e.g.,
+  `logit(LOG_CONSOLE | LOG_NOTICE, ...)`, by Jacques de Laval, Westermo
+* Fix #175: ensure Finit does not acquire a controlling TTY when checking
+  if a device is a TTY before starting a getty.  This fixes an old bug
+  where Ctrl-C after logout from a shell could cause PID 1 to get SIGINT,
+  which in turn could lead to a system reboot
+
+
+[4.0][] - 2021-04-26
+--------------------

 This release became v4.0, and not v3.2, because of incompatible changes
 to service conditions.  There are other significant changes as well, so
-make sure to read the whole changelog when upgrading.
+make sure to read the whole change log when upgrading.

 ### Changes
 * The stand-alone `reboot` tool has been replaced with a symlink to
@@ -18,31 +118,53 @@ make sure to read the whole changelog when upgrading.
  with a fallback to sending signals as per traditional SysV init.  The
  `-f` (force) flag remains, where `reboot(2)` is called directly
 * Introducing Finit progress 𝓜𝓸𝓭𝓮𝓻𝓷
+* The `inictl cond set|clear COND` have changed completely.  Constrained
+  to a flat `<usr/...>` namespace and automatically activated by a new
+  `usr.so` plugin that checks services for usr condition changes
 * Removed built-in inetd super server.  If you need this functionality,
  use an external inetd, like xinetd, instead.  A pull request for a
  stand-alone inetd, like watchdogd and getty, is most welcome!
-* Incompatible `configure` script changes, i.e., you must give proper
-  path arguments to the script, no more guessing just GNU defaults.
-  There are examples in the documentation and the `contrib/` section
-* Change service conditions from the non-obvious `<svc/path/to/foo>` to
-  `<pid/foo:id>`.  This to hopefully make it clear that one service's
-  'pid:!foo' pidfile is another service's `<pid/foo>` condition
-* Initial support for cgroups, every service now runs in its own cgroup
-  https://twitter.com/b0rk/status/1214341831049252870?s=20
+* Incompatible `configure` script changes, i.e., no guessing `--prefix`
+  and other paths.  Also, many options have been changed, renamed, or
+  flipped defaults, or even dropped altogether.  There are examples in
+  the documentation and the `contrib/` section
+* Service conditions change from the non-obvious `<svc/path/to/foo>` to
+  `<pid/foo:id>`.  Not only does this give simpler internal semantics,
+  it hopefully also makes it clear that one service's `pid:!foo` pidfile
+  is another service's `<pid/foo>` condition, issue #143
+* Initial support for cgroups v2:
+  - services runs in a cgroup named after their respective *.conf file
+  - top-level groups are; init, user, and system
+  - all top-level groups can be configured from finit *.conf files
+  - each service can tweak the cgroup settings
+  - Use `initctl [top|ps|cgroup]` commands to inspect runtime state
+  - https://twitter.com/b0rk/status/1214341831049252870?s=20
 * Major refactor of Finit's `main()` function to be able to start the
  event loop earlier.  This also facilitated factoring out functionality
  previously hard-coded in Finit, e.g., starting the bundled watchdogd,
  various distro packed udevd and other hotplugging tools
+* A proper rescue mode has been added.  It is started extremely early
+  and is protected with a bundled `suslogin`.  Exiting rescue mode now
+  brings up the system as a normal boot, as one expects
 * Support for `sysv` start/stop scripts as well as monitoring forking
  services, stared using `sysv` or `service` stanza
 * Support for custom `kill:DELAY`, default 3 sec.
 * Support for custom `halt:SIGNAL`, default SIGTERM
+* Support for `pre:script` and `post:script`, allows for setup and
+  teardown/cleanup before and after a service runs, issue #129
+* Support for `env:file` in `/etc/default/foo` or `/etc/conf.d/foo`, see
+  the contrib section for examples that utilize this feature.  Variables
+  expanded from env files, and the env files themselves, are tracked for
+  changes to see if a service .conf file is "dirty" and needs restart on
+  `initctl reload`
 * Support for tracking custom PID files, using `pid:!/path/to/foo.pid`,
  useful with new `sysv` or `service` which fork to background
 * Support starting run/task/services without absolute path, trust `$PATH`
 * Add support for `--disable-doc` and `--disable-contrib` to speed up
  builds and work around issue with massively parallel builds
-* Add support for `@console` also for external getty
+* Support for `@console` also for external getty
+* Support for `notty` option to built-in getty, for board bring-up
+* Support for `rescue` option to built-in getty, for rescue shells
 * Add `-b`, batch mode, for non-interactive use to `initctl`
 * Prefer udev to handle `/dev/` if mdev is also available
 * Redirect dbus daemon output to syslog
@@ -92,6 +214,8 @@ make sure to read the whole changelog when upgrading.
 * Always append `:ID` qualifier to conditions if set for a service
 * The IPC socket has moved from `/run/finit.sock` to `/run/finit/socket`
  officially only supported for use by the `initctl` tool
+* The IPC socket now uses `SOCK_SEQPACKET` instead of `SOCK_STREAM`.
+  Recommend using watchdogd v3.4, or later, which support this
 * Improved support for modern `/etc/network/interfaces`, which has
  include statements.  No more native `ifup` of individual interfaces,
  Finit now calls `ifup -a`, or `ifdown -a`, delegating all details to
@@ -109,16 +233,21 @@ make sure to read the whole changelog when upgrading.
  that do not have a watchdog
 * Fix #100: Early condition handling may not work if `/var/run` does
  not yet exist (symlink to `/run`).  Added compat layer for access
+* Fix #101: Built-in inetd removed
+* Fix #102: Start built-in watchdogd as a regular service
 * Fix #103: Register multiple getty if `@console` resolves to >1 TTY,
 * Fix #105: Only remove /etc/nologin when moving from runlevel 0, 1, 6
  Fixed by Jonas Johansson, Westermo
 * Fix #109: Support for PID files in sub-directories to `/var/run`
 * Handle rename of PID files, by Robert Andersson, Atlas Copco
+* Fix #110: automatic modprobe of RTC devices, built-in hwclock
 * Fix #120: Redirect `stdin` to `/dev/null` for services by default
 * Fix #122: Switch to `nanosleep()` to achieve "signal safe" sleep,
  fixed by Jacques de Laval, Westermo
 * Fix #124: Lingering processes in process group when session leader
  exits.  E.g., lingering `logit` processes when parent dies
+* Fix #127: Show all runparts scripts as they start, like rc.local, fixed
+  by Jacques de Laval, Westermo
 * Fix service name matching, e.g. for condition handling, may match with
  wrong service, by Jonas Holmberg, Westermo
 * Run all run-parts scripts using `/bin/sh -c foo` just like the standard
@@ -137,6 +266,8 @@ make sure to read the whole changelog when upgrading.
 * Provide service description for built-in watchdog daemon
 * Fix #138: Handle `SIGPWR` like `SIGSUR2`, i.e., power off the system
 * Drop the '%m' GNUism, for compat with older musl libc
+* Fix #139: call `tzset()` on `initctl reload` to activate system
+  timezone changes (for logging)


 [3.1][] - 2018-01-23
@@ -745,14 +876,14 @@ Major bug fix release.
 * Makefile fixes for installation, paths encoded wrong
 * Strip binaries + .so files, support for `$(CROSS)` toolchain strip
 * Default install is now to `/sbin/finit` and `/usr/`
-* Note change in `$PLUGIN_DIR` environemnt variable to `$plugindir`
+* Note change in `$PLUGIN_DIR` environment variable to `$plugindir`


 [1.2][] - 2012-09-27
 --------------------

 ### Changes
-* Update README with section on building and enviroment variables
+* Update README with section on building and environment variables

 ### Fixes
 * Fix installation paths encoded in finit binary
@@ -844,7 +975,8 @@ Major bug fix release.

 * Initial release

-[UNRELEASED]: https://github.com/troglobit/finit/compare/3.1...HEAD
+[UNRELEASED]: https://github.com/troglobit/finit/compare/4.0...HEAD
+[4.1]: https://github.com/troglobit/finit/compare/4.0...4.1
 [4.0]: https://github.com/troglobit/finit/compare/3.1...4.0
 [3.1]: https://github.com/troglobit/finit/compare/3.0...3.1
 [3.0]: https://github.com/troglobit/finit/compare/2.4...3.0
@@ -876,9 +1008,3 @@ Major bug fix release.
 [0.2]: https://github.com/troglobit/finit/compare/0.1...0.2
 [libuEv]: https://github.com/troglobit/libuev
 [dea3ae8]: https://github.com/troglobit/finit/commit/dea3ae8
-
-<!--
-  -- Local Variables:
-  -- mode: markdown
-  -- End:
-  -->
@@ -1,5 +1,5 @@
 ACLOCAL_AMFLAGS    = -I m4
-SUBDIRS            = plugins src test
+SUBDIRS            = man plugins src test
 doc_DATA           = README.md LICENSE contrib/finit.conf
 EXTRA_DIST         = README.md LICENSE ChangeLog.md contrib/finit.conf

@@ -22,13 +22,30 @@
 Introduction
 ------------

-Finit is a simple alternative to [SysV init][4] and [systemd][6].  It
-was reverse engineered from the [EeePC fastinit][] over ten years ago
-by [Claudio Matsuoka][] — "gaps filled with frog DNA …"
+Fast init for Linux systems.  Reverse engineered from the [EeePC
+fastinit][] over ten years ago by [Claudio Matsuoka][] — "gaps filled
+with frog DNA …"

 ![Alpine screenshot](img/finit4-screenshot.png)  
 *Figure 1: Screenshot showing Finit booting [Alpine Linux](https://www.alpinelinux.org/).*

+Features include:
+
+  * Runlevels, defined per service
+  * One-shot tasks, services (daemons), or [SysV init][4] start/stop scripts
+  * Runparts and `/etc/rc.local` support
+  * Process supervision similar to [systemd][]
+  * Sourcing environment files
+  * Conditions for network/process/custom dependencies
+  * Pre/Post script actions
+  * Tooling to enable/disable services
+  * Built-in getty
+  * Built-in watchdog, with support for hand-over to [watchdogd](https://troglobit.com/watchdogd.html)
+  * Built-in support for Debian/BusyBox `/etc/network/interfaces`
+  * Cgroups v2, both configuration and monitoring in `initctl top`
+  * Plugin support for customization
+  * Proper rescue mode with bundled `sulogin` for protected maintenance shell
+
 Focus is on small and embedded systems, although Finit is fully usable
 on server and desktop systems as well.  For working examples, see the
 [contrib/](contrib/) section with tutorials for:
@@ -58,10 +75,10 @@ runlevel 2
 log size=100k count=4

 # Services to be monitored and respawned as needed
-service [S12345] watchdogd -L -f                 -- System watchdog daemon
-service [S12345] syslogd -n -b 3 -D              -- System log daemon
-service [S12345] <pid/syslogd> klogd -n          -- Kernel log daemon
-service   [2345] lldpd -d -c -M1 -H0 -i          -- LLDP daemon (IEEE 802.1ab)
+service [S12345] env:-/etc/conf.d/watchdog watchdog $WATCHDOG_OPTS $WATCHDOG_DEV -- System watchdog daemon
+service [S12345] env:-/etc/conf.d/syslog syslogd -n $SYSLOGD_OPTS          -- System log daemon
+service [S12345] <pid/syslogd> env:-/etc/conf.d/klogd klogd -n $KLOGD_OPTS -- Kernel log daemon
+service   [2345] env:-/etc/conf.d/lldpd lldpd -d $LLDPD_OPTS               -- LLDP daemon (IEEE 802.1ab)

 # The BusyBox ntpd does not use syslog when running in the foreground
 # So we use this trick to redirect stdout/stderr to a log file.  The
@@ -76,24 +93,24 @@ service :80   [2345] merecat -n -p 80   /var/www -- Web server
 service :8080 [2345] merecat -n -p 8080 /var/www -- Old web server

 # Alternative method instead of below runparts, can also use /etc/rc.local
-#task [S] /etc/init.d/keyboard-setup start       -- Setting up preliminary keymap
-#task [S] /etc/init.d/acpid start                -- Starting ACPI Daemon
-#task [S] /etc/init.d/kbd start                  -- Preparing console
+#sysv [S] /etc/init.d/keyboard-setup       -- Setting up preliminary keymap
+#sysv [S] /etc/init.d/acpid                -- Starting ACPI Daemon
+#task [S] /etc/init.d/kbd                  -- Preparing console

 # Run start scripts from this directory
 # runparts /etc/start.d

 # Virtual consoles run BusyBox getty, keep kernel default speed
-tty [12345] /sbin/getty -L 0 /dev/tty1 linux
-tty [12345] /sbin/getty -L 0 /dev/tty2 linux
-tty [12345] /sbin/getty -L 0 /dev/tty3 linux
+tty [12345] /sbin/getty -L 0 /dev/tty1  linux nowait noclear
+tty [2345]  /sbin/getty -L 0 /dev/tty2  linux nowait noclear
+tty [2345]  /sbin/getty -L 0 /dev/tty3  linux nowait noclear

 # Use built-in getty for serial port and USB serial
-tty [12345] /dev/ttyAMA0 noclear nowait
-tty [12345] /dev/ttyUSB0 noclear
+#tty [12345] /dev/ttyAMA0 noclear nowait
+#tty [12345] /dev/ttyUSB0 noclear

 # Just give me a shell, I need to debug this embedded system!
-tty [12345] @console noclear nologin
+#tty [12345] console noclear nologin
 ```

 The `service` stanza, as well as `task`, `run` and others are described
@@ -117,7 +134,7 @@ Some components are optional: runlevel(s), condition(s) and description,
 making it easy to create simple start scripts and still possible for more
 advanced uses as well:

-    service sshd -D
+    service /usr/sbin/sshd -D

 Dependencies are handled using [conditions](doc/conditions.md).  One of
 the most common conditions is to wait for basic networking to become
@@ -232,8 +249,11 @@ For more information, see [doc/plugins.md](doc/plugins.md).
 By default, Finit monitors `/etc/finit.d/` and `/etc/finit.d/enabled/`
 registering any changes to `.conf` files.  To activate a change the user
 must call `initctl reload`, which reloads all modified files, stops any
-removed services, starts new ones, and restarts any modified ones, with
-SIGHUP if the process supports it.
+removed services, starts new ones, and restarts any modified ones.  If the
+command line arguments of a service have changed, the process will be
+terminated and then started again with the updated arguments. If the arguments
+have not been modified and the process supports SIGHUP, the process will
+receive a SIGHUP rather than being terminated and started.

 For some use-cases the extra step of calling `initctl reload` creates an
 unnecessary overhead, which can be removed at build-time using:
@@ -241,6 +261,46 @@ unnecessary overhead, which can be removed at build-time using:
    configure --enable-auto-reload


+**Cgroups**
+
+Finit supports cgroups v2 and comes with the following default groups
+in which services and user sessions are placed in:
+
+     /sys/fs/cgroup
+       |-- init/               # cpu.weight:100
+       |-- system/             # cpu.weight:9800
+       `-- user/               # cpu.weight:100
+
+Finit itself and its helper scripts and services are placed in the
+top-level leaf-node group `init/`, which also is _reserved_.
+
+All run/task/service/sysv processes are placed in their own sub-group
+in `system/`.  The name of each sub-group is taken from the respective
+`.conf` file from `/etc/finit.d`.
+
+All getty/tty processes are placed in their own sub-group in `user/`.
+The name of each sub-group is taken from the username.
+
+A fourth group also exists, the `root` group.  It is also _reserved_ and
+primarily intended for RT tasks.  If you have RT tasks they need to be
+declared as such in their service stanza like this:
+
+    service [...] <...> cgroup.root /path/to/foo args -- description
+
+or
+
+    cgroup.root
+    service [...] <...> /path/to/foo args -- description
+    service [...] <...> /path/to/bar args -- description
+
+See [doc/config.md](doc/config.md#cgroups) for more information, e.g.,
+how to configure per-group limits.
+
+The `initctl` tool has three commands to help debug and optimize the
+setup and monitoring of cgroups.  See the `ps`, `top`, and `cgroup`
+commands for details.
+
+
 Runparts & /etc/rc.local
 ------------------------

@@ -356,6 +416,9 @@ Options:
  -b, --batch               Batch mode, no screen size probing
  -c, --create              Create missing paths (and files) as needed
  -f, --force               Ignore missing files and arguments, never prompt
+  -1, --once                Only one lap in commands like 'top'
+  -p, --plain               Use plain table headings, no ctrl chars
+  -q, --quiet               Silent, only return status of command
  -t, --no-heading          Skip table headings
  -v, --verbose             Verbose output
  -h, --help                This help text
@@ -375,17 +438,22 @@ Commands:
  disable  <CONF>           Disable  .conf in /etc/finit.d/enabled
  reload                    Reload  *.conf in /etc/finit.d (activate changes)

-  cond     show             Show condition status
+  cond     set   <COND>     Set (assert) user-defined condition     +usr/COND
+  cond     clear <COND>     Clear (deassert) user-defined condition -usr/COND
+  cond     status           Show condition status, default cond command
  cond     dump             Dump all conditions and their status

  log      [NAME]           Show ten last Finit, or NAME, messages from syslog
  start    <NAME>[:ID]      Start service by name, with optional ID
  stop     <NAME>[:ID]      Stop/Pause a running service by name
+  reload   <NAME>[:ID]      Reload service by name (SIGHUP or restart)
  restart  <NAME>[:ID]      Restart (stop/start) service by name
  status   <NAME>[:ID]      Show service status, by name
  status                    Show status of services, default command

+  cgroup                    List cgroup config overview
  ps                        List processes based on cgroups
+  top                       Show top-like listing based on cgroups

  runlevel [0-9]            Show or set runlevel: 0 halt, 6 reboot
  reboot                    Reboot system
@@ -399,13 +467,17 @@ file must be used to tell Finit to stop and start it on `reload` and
 `runlevel` changes.  If `<>` holds more [conditions](doc/conditions.md),
 these will also affect how a service is maintained.

-**Note:** even though it is possible to start services not belonging in
-the current runlevel these services will not be respawned automatically
-by Finit if they exit (crash).  Hence, if the runlevel is 2, the below
-Dropbear SSH service will not be restarted if it is killed or exits.
+> **Note:** even though it is possible to start services not belonging in
+> the current runlevel these services will not be respawned automatically
+> by Finit if they exit (crash).  Hence, if the runlevel is 2, the below
+> Dropbear SSH service will not be restarted if it is killed or exits.
+
+The `status` command is the default, it displays a quick overview of all
+monitored run/task/services.  Here we call `initctl -p`, suitable for
+scripting and documentation:

 ```
-alpine:~# initctl 
+alpine:~# initctl -p
 PID   IDENT     STATUS   RUNLEVELS    DESCRIPTION
 ======================================================================
 1506  acpid     running  [--2345----] ACPI daemon
@@ -415,7 +487,7 @@ PID   IDENT     STATUS   RUNLEVELS    DESCRIPTION
 1512  ntpd      running  [--2345----] NTP daemon
 1473  syslogd   running  [S12345----] Syslog daemon

-alpine:~# initctl -v
+alpine:~# initctl -pv
 PID   IDENT     STATUS   RUNLEVELS    COMMAND
 ======================================================================
 1506  acpid     running  [--2345----] acpid -f
@@ -430,6 +502,40 @@ The environment variables to each of the services above are read from,
 in the case of Alpine Linux, `/etc/conf.d/`.  Other distributions may
 have other directories, e.g., Debian use `/etc/default/`.

+The `status` command takes an optional `NAME:ID` argument.  Here we
+check the status of `dropbear`, which only has one instance in this
+system:
+
+```
+alpine:~# initctl -p status dropbear
+     Status : running
+   Identity : dropbear
+Description : Dropbear SSH daemon
+     Origin : /etc/finit.d/enabled/dropbear.conf
+Environment : -/etc/conf.d/dropbear
+Condition(s):
+    Command : dropbear -R -F $DROPBEAR_OPTS
+   PID file : !/run/dropbear.pid
+        PID : 1485
+       User : root
+      Group : root
+     Uptime : 2 hour 46 min 56 sec
+  Runlevels : [--2345----]
+     Memory : 1.2M
+     CGroup : /system/dropbear cpu 0 [100, max] mem [--.--, max]
+              |- 1485 dropbear -R -F
+              |- 2634 dropbear -R -F
+              |- 2635 ash
+              `- 2652 initctl -p status dropbear
+
+Apr  8 12:19:49 alpine authpriv.info dropbear[1485]: Not backgrounding
+Apr  8 12:37:45 alpine authpriv.info dropbear[2300]: Child connection from 192.168.121.1:47834
+Apr  8 12:37:46 alpine authpriv.notice dropbear[2300]: Password auth succeeded for 'root' from 192.168.121.1:47834
+Apr  8 12:37:46 alpine authpriv.info dropbear[2300]: Exit (root) from <192.168.121.1:47834>: Disconnect received
+Apr  8 15:02:11 alpine authpriv.info dropbear[2634]: Child connection from 192.168.121.1:48576
+Apr  8 15:02:12 alpine authpriv.notice dropbear[2634]: Password auth succeeded for 'root' from 192.168.121.1:48576
+```
+

 Requirements
 ------------
@@ -475,8 +581,6 @@ and proposed extensions.
 [3]:  http://smarden.org/runit/
 [4]:  http://en.wikipedia.org/wiki/Init
 [5]:  http://en.wikipedia.org/wiki/Runlevel
-[6]:  https://www.freedesktop.org/wiki/Software/systemd/
-[9]:  https://github.com/troglobit/troglos
 [RFC862]:           https://tools.ietf.org/html/rfc862
 [RFC863]:           https://tools.ietf.org/html/rfc863
 [RFC864]:           https://tools.ietf.org/html/rfc864
@@ -1,4 +1,4 @@
-AC_INIT(Finit, 4.0-rc2, https://github.com/troglobit/finit/issues, finit, https://troglobit.com/finit.html)
+AC_INIT(Finit, 4.1, https://github.com/troglobit/finit/issues, finit, https://troglobit.com/finit.html)
 AC_CONFIG_AUX_DIR(aux)
 AM_INIT_AUTOMAKE([1.11 foreign dist-xz subdir-objects])
 AM_SILENT_RULES([yes])
@@ -8,9 +8,10 @@ AC_CONFIG_HEADER([config.h])
 AC_CONFIG_FILES([Makefile
 	contrib/Makefile
 	contrib/alpine/Makefile contrib/alpine/finit.d/Makefile contrib/alpine/finit.d/available/Makefile
-	contrib/debian/Makefile contrib/debian/finit.d/Makefile contrib/debian/finit.d/available/Makefile contrib/debian/grub.d/Makefile
+	contrib/debian/Makefile contrib/debian/finit.d/Makefile contrib/debian/finit.d/available/Makefile
 	contrib/void/Makefile   contrib/void/finit.d/Makefile   contrib/void/finit.d/available/Makefile
 	doc/Makefile
+	man/Makefile
 	plugins/Makefile
 	src/Makefile
 	test/test.env
@@ -48,30 +49,17 @@ PKG_CHECK_MODULES([uev],  [libuev >= 2.2.0])
 PKG_CHECK_MODULES([lite], [libite >= 2.2.0])

 # Check for configured Finit features
-AC_ARG_ENABLE(emergency_shell,
-	AS_HELP_STRING([--enable-emergency-shell], [Experimental emergency fallback at boot.
-	                Use only for debugging very early boot problems with Finit]),,[
-	enable_emergency_shell=no])
-
-AC_ARG_ENABLE(fallback_shell,
-        AS_HELP_STRING([--enable-fallback-shell], [Start a shell if no TTYs can start]),,[
-	enable_fallback_shell=no])
-
 AC_ARG_ENABLE(auto_reload,
        AS_HELP_STRING([--enable-auto-reload], [Automatic reload on any .conf change]),,[
 	enable_auto_reload=no])

-AC_ARG_ENABLE(watchdog,
-        AS_HELP_STRING([--enable-watchdog], [Enable built-in watchdog, uses /dev/watchdog]),,[
-	enable_watchdog=no])
+AC_ARG_ENABLE(kernel_cmdline,
+        AS_HELP_STRING([--enable-kernel-cmdline], [Parse init args from /proc/cmdline (don't use!)]),,[
+	enable_kernel_cmdline=no])

 AC_ARG_ENABLE(redirect,
-        AS_HELP_STRING([--enable-redirect], [Redirect service output to /dev/null, default: no]),,[
-	enable_redirect_output=no])
-
-AC_ARG_ENABLE(logit,
-     AS_HELP_STRING([--disable-logit], [Disable logit, used for log redirection, default: enable]),,[
-     enable_logit=yes])
+        AS_HELP_STRING([--disable-redirect], [Disable redirection of service output to /dev/null]),,[
+	enable_redirect=yes])

 AC_ARG_ENABLE(logrotate,
        AS_HELP_STRING([--disable-logrotate], [Disable built-in rotation of /var/log/wtmp, default enabled]),,[
@@ -94,10 +82,12 @@ AC_PLUGIN([resolvconf],    [no],  [Setup necessary files for resolvconf])
 AC_PLUGIN([x11-common],    [no],  [Console setup (for X)])
 AC_PLUGIN([netlink],       [yes], [Basic netlink plugin for IFUP/IFDN and GW events. Can be replaced with externally built plugin that links with libnl or similar.])
 AC_PLUGIN([hotplug],       [yes], [Setup and start udev or mdev hotplug daemon])
+AC_PLUGIN([rtc],           [yes], [Save and restore RTC using hwclock])
+AC_PLUGIN([urandom],       [yes], [Setup and save random seed at boot/shutdown])

 # Check for extra arguments or packages
 AC_ARG_WITH(heading,
-        AS_HELP_STRING([--with-heading=GREET], [Display GREET instead of Finit version at boot]),
+        AS_HELP_STRING([--with-heading=GREET], [Heading for boot progress. Default: PRETTY_NAME from /etc/os-release, fallback to "Finit vX.YY"]),
 	[heading=$withval], [heading=yes])

 AC_ARG_WITH(config,
@@ -128,28 +118,26 @@ AC_ARG_WITH(random-seed,
        AS_HELP_STRING([--with-random-seed=FILE], [Save a random seed for /dev/urandom across reboots, default /var/lib/misc/random-seed]),
 	[random_seed=$withval], [random_seed=yes])

+AC_ARG_WITH(keventd,
+        AS_HELP_STRING([--with-keventd], [Enable built-in keventd, default: no]),, [with_keventd=no])
+
+AC_ARG_WITH(watchdog,
+        AS_HELP_STRING([--with-watchdog=[DEV]], [Enable built-in watchdog, default: /dev/watchdog]),
+	[watchdog=$withval], [with_watchdog=no watchdog=])
+
 ### Enable features ###########################################################################

-# Create config.h from selected features and fallback defautls
-AS_IF([test "x$enable_emergency_shell" = "xyes"], [
-        AC_DEFINE(EMERGENCY_SHELL,  1, [Fork off init and let PID 1 wait and start /bin/sh if init crashes.])])
-
-AS_IF([test "x$enable_fallback_shell" = "xyes"], [
-        AC_DEFINE(FALLBACK_SHELL,  1, [Start /bin/sh on console if no TTYs.])])
-
+# Create config.h from selected features and fallback defaults
 AS_IF([test "x$enable_auto_reload" = "xyes"], [
        AC_DEFINE(AUTO_RELOAD, 1, [Finit automatically does `initctl reload` on any .conf change])])

-AS_IF([test "x$enable_watchdog" = "xyes"], [
-        AC_DEFINE(BUILTIN_WATCHDOG,  1, [Enable built-in watchdog, kicks on /dev/watchdog])])
+AS_IF([test "x$enable_kernel_cmdline" = "xyes"], [
+        AC_DEFINE(KERNEL_CMDLINE, 1, [Dumpster diving after init args from /proc/cmdline])])

 AS_IF([test "x$enable_redirect" = "xyes"], [
 	AC_DEFINE(REDIRECT_OUTPUT, 1, [Enable redirection of service output to /dev/null])])

 ### Disable features ###########################################################################
-AS_IF([test "x$enable_logit" != "xno"], [
-	AC_DEFINE(LOGIT_ENABLED,  1, [Enable logit tool, used for log redirection by Finit])])
-
 AS_IF([test "x$enable_logrotate" != "xno"], [
 	AC_DEFINE(LOGROTATE_ENABLED, 1, [Enable built-in rotation of /var/log/wtmp et al.])])

@@ -183,9 +171,8 @@ AS_IF([test "x$with_sysconfig" != "xno"], [
 			    [System environment, default /etc/default & /etc/conf.d])])])

 AS_IF([test "x$with_heading" != "xno"], [
-	AS_IF([test "x$heading" = "xyes"], [
-		heading="$PACKAGE_NAME v$PACKAGE_VERSION"])])
-AC_DEFINE_UNQUOTED(INIT_HEADING, "$heading", [Displayed by Finit at boot])
+	AS_IF([test "x$heading" = "xyes"], [heading=""])
+	AC_DEFINE_UNQUOTED(INIT_OSHEADING, "$heading", [Displayed by Finit at boot, defaults to /etc/os-release PRETTY_NAME])])

 AS_IF([test "x$with_hostname" != "xno"], [
 	AS_IF([test "x$hostname" = "xyes"], [
@@ -204,10 +191,19 @@ AS_IF([test "x$with_random_seed" != "xno"], [
 	AC_EXPAND_DIR(random_path, "$random_seed")
 	AC_DEFINE_UNQUOTED(RANDOMSEED, "$random_path", [Improve random at boot by seeding it with sth from before.])])

+AS_IF([test "x$with_keventd" != "xno"], [with_keventd=yes])
+
+AS_IF([test "x$with_watchdog" != "xno"], [
+	AS_IF([test "x$watchdog" = "xyes"], [
+		watchdog=/dev/watchdog])
+	with_watchdog=yes
+        AC_DEFINE_UNQUOTED(WDT_DEVNODE, "$watchdog", [Built-in watchdog device node])], [
+	watchdog=])
+
 # Control build with automake flags
 AM_CONDITIONAL(STATIC,    [test "x$enable_static" = "xyes"])
-AM_CONDITIONAL(WATCHDOGD, [test "x$enable_watchdog" = "xyes"])
-AM_CONDITIONAL(LOGIT,     [test "x$enable_logit" = "xyes"])
+AM_CONDITIONAL(KEVENTD,   [test "x$with_keventd"  != "xno"])
+AM_CONDITIONAL(WATCHDOGD, [test "x$with_watchdog" != "xno"])
 AM_CONDITIONAL(DOC,       [test "x$enable_doc" = "xyes"])
 AM_CONDITIONAL(CONTRIB,   [test "x$enable_contrib" = "xyes"])

@@ -226,6 +222,9 @@ AC_DEFINE_UNQUOTED(RESCUE_CONF, "$rescue_conf", [Finit rescue mode .conf])
 AC_EXPAND_DIR(sample_conf, "$libdir/finit/sample.conf")
 AC_DEFINE_UNQUOTED(SAMPLE_CONF, "$sample_conf", [Finit sameple.conf for initctl create])

+AS_IF([test "$heading" = ""],   [heading="<DEFAULT>"])
+AS_IF([test "$heading" = "no"], [heading="<NONE>"])
+
 # Generate all files
 AC_OUTPUT

@@ -253,11 +252,10 @@ Behavior:
 Optional features:
  Install doc/..........: $enable_doc
  Install contrib/......: $enable_contrib
-  Built-in watchdogd....: $enable_watchdog
+  Built-in keventd......: $with_keventd
+  Built-in watchdogd....: $with_watchdog $watchdog
  Built-in logrotate....: $enable_logrotate
-  Scripting tool logit..: $enable_logit
-  Emergency shell.......: $enable_emergency_shell
-  Fallback shell........: $enable_fallback_shell
+  Redirect output.......: $enable_redirect
  Default hostname......: $hostname
  Default runlevel......: $runlevel

@@ -1,5 +1,5 @@
 SUBDIRS    = finit.d
 docsdir   := @docdir@/contrib/alpine
-docs_DATA  = README.md finit.conf alpine-release build.sh install.sh rc.local
+docs_DATA  = README.md finit.conf build.sh install.sh rc.local
 EXTRA_DIST = $(docs_DATA)

@@ -1 +0,0 @@
-3.6.2
@@ -17,8 +17,8 @@ PKG_CONFIG_LIBDIR=/usr/lib/pkgconfig:/usr/local/lib/pkgconfig ./configure	\
 		 --prefix=/usr			--exec-prefix=			\
 		 --sysconfdir=/etc		--localstatedir=/var		\
 		 --enable-dbus-plugin		--enable-x11-common-plugin	\
-		 --enable-alsa-utils-plugin	--enable-watchdog		\
-		 --with-heading="Alpine Linux 3.13"
+		 --enable-alsa-utils-plugin	--with-watchdog			\
+		 --with-keventd

 echo
 echo "*** Building ..."
@@ -5,6 +5,22 @@
 # Default is 2
 #runlevel 2

+# Top-level cgroups and their default settings.  All groups mandatory
+# but more can be added, max 8 groups in total currently.  The cgroup
+# 'root' is also available, reserved for RT processes.  Settings are
+# as-is, only one shorthand 'mem.' exists, other than that it's the
+# cgroup v2 controller default names.
+#cgroup init   cpu.weight:100
+#cgroup user   cpu.weight:100
+#cgroup system cpu.weight:9700
+
+# Task to run if ctrl-alt-del is pressed.  This condition is asserted by
+# Finit upon receiving SIGINT (from the kernel).
+#task <sys/key/ctrlaltdel> [12345789] initctl reboot -- Rebooting system
+
+# Task to run if the kernel gets a power fail condition is pressed.
+# Asserted by Finit upon receiving SIGPWR (from the kernel).
+#task <sys/pwr/fail> [12345789] initctl poweroff -- Shutting down system
+
 ## Bootstrap services and tasks
 # Can be listed here or in /etc/finit.d/*.conf
-
@@ -1,5 +1,4 @@
 docsdir   := @docdir@/contrib/alpine/finit.d/available
 docs_DATA  = acpid.conf crond.conf dropbear.conf getty.conf keymap.conf	\
-             lxdm.conf modules.conf ntpd.conf klogd.conf syslog.conf	\
-	     watchdog.conf
+             lxdm.conf modules.conf ntpd.conf syslog.conf watchdog.conf
 EXTRA_DIST = $(docs_DATA)
@@ -1,4 +1,5 @@
 # Consoles to start getty on when system is up
-tty [12345] /dev/tty1 noclear nowait
-tty [2345]  /dev/tty2 noclear nowait
-tty [2345]  /dev/tty3 noclear nowait
+tty [12345] /dev/ttyS0 noclear
+tty [12345] /dev/tty1  noclear nowait
+tty [2345]  /dev/tty2  noclear nowait
+tty [2345]  /dev/tty3  noclear nowait
@@ -1,2 +0,0 @@
-# Start klogd as soon as possible, after syslogd, should always run
-service [S12345] <pid/syslogd> env:-/etc/conf.d/klogd klogd -n $KLOGD_OPTS -- Kernel log daemon
@@ -4,3 +4,4 @@
 #module evdev
 #module mousedev
 #module 8139cp
+module softdog
@@ -1,2 +1,4 @@
 # Start syslogd as soon as possible, should always run
-service [S12345] env:-/etc/conf.d/syslog syslogd -n $SYSLOGD_OPTS -- Syslog daemon
+# Start klogd as soon as possible, but after syslogd
+service [S12345] env:-/etc/conf.d/syslog syslogd -n $SYSLOGD_OPTS          -- Syslog daemon
+service [S12345] <pid/syslogd> env:-/etc/conf.d/klogd klogd -n $KLOGD_OPTS -- Kernel log daemon
@@ -8,9 +8,10 @@ if [ "x`id -u`" != "x0" ]; then
 fi

 # Adjust base directory
-if [ -e alpine-release ]; then
+grep -q alpine Makefile 2>/dev/null
+if [ $? -eq 0 ]; then
    cd ../..
-elif [ ! -e autogen.sh ]; then
+elif [ ! -e configure.ac ]; then
    echo "*** Please run this script from the Finit base directory."
    exit 1
 fi
@@ -1,4 +1,4 @@
-SUBDIRS    = finit.d grub.d
+SUBDIRS    = finit.d
 docsdir   := @docdir@/contrib/debian
-docs_DATA  = README.md finit.conf os-release build.sh install.sh
+docs_DATA  = README.md finit.conf build.sh install.sh
 EXTRA_DIST = $(docs_DATA)
@@ -20,29 +20,42 @@ tools pre-systemd work as intended.

    root@debian:~# apt install initscripts console-setup

+> **Note:** as of Debian 11 (Bullseye), libuev and libite are part of
+> the main section of Debian.  So just install the -dev packages :)
+
 The following script can then be used to configure, build, install and
 set up your system to run Finit:

    user@debian:~/finit$ contrib/debian/build.sh

 However, since `/sbin/init` already exists on your system, the script
-creates another entry in your GRUB config, in `/etc/grub.d/40_custom`,
-where `init=/sbin/finit` is added to your kernel line.  It is of course
-also possible to change the default init to Finit, if you do you can
-remove the custom Grub entry:
+creates another entry in your GRUB config by updating `$SUPPORTED_INITS`
+in `/etc/grub.d/10_inux` and run `update-grub`.  On reboot you will find
+a (finit) entry in the "Advanced options for Debian GNU/Linux" section.
+It is of course also possible to change the default init to Finit:

    user@debian:~/finit# cd /sbin
-    user@debian:/sbin# sudo rm init
+    user@debian:/sbin# sudo mv init oldinit
    user@debian:/sbin# sudo ln -s finit init

 Before rebooting, check the default [/etc/finit.conf](finit.conf) and
 `/etc/finit.d/*.conf` files.  The build + install script above provides
-a few sample `.conf` files. See `initctl list` after boot for a list of
-enabled and available services.
+a few sample `.conf` files.  See `initctl ls` after boot for a list of
+enabled and available services, you can then use `enable` and `disable`
+comands to `initctl`, followed by `reload` to activate your changes.

 You can also use a standard [/etc/rc.local](rc.local) for one-shot tasks
 and initialization like keyboard language etc.

+> **NOTE:** X Window system, you may need to `sudo apt install elogind`
+> (Bullseye and later), followed by `initctl reload` to activate it (it
+> is enabled by default), and logout/login again.  The elogind daemon
+> ensures a regular non-root user can start and interact with an X
+> session, otherwise keyboard and mouse won't work.  When you're happy,
+> you can enable the lightdm.conf, change the default runlevel to 3, and
+> presto!  you have a desktop again.
+
+
 Have fun!  
 /Joachim ツ

@@ -25,9 +25,8 @@ echo
    --prefix=/usr                     --exec-prefix=			\
    --sysconfdir=/etc                 --localstatedir=/var		\
    --enable-dbus-plugin              --enable-x11-common-plugin	\
-    --enable-alsa-utils-plugin						\
-    --with-random-seed=/var/lib/urandom/random-seed			\
-    --with-heading="Debian GNU/Linux"
+    --enable-alsa-utils-plugin	      --with-keventd			\
+    --with-random-seed=/var/lib/urandom/random-seed

 if [ $? -ne 0 ]; then
    echo
@@ -5,6 +5,22 @@
 # Default is 2
 #runlevel 2

+# Top-level cgroups and their default settings.  All groups mandatory
+# but more can be added, max 8 groups in total currently.  The cgroup
+# 'root' is also available, reserved for RT processes.  Settings are
+# as-is, only one shorthand 'mem.' exists, other than that it's the
+# cgroup v2 controller default names.
+#cgroup init   cpu.weight:100
+#cgroup user   cpu.weight:100
+#cgroup system cpu.weight:9700
+
+# Task to run if ctrl-alt-del is pressed.  This condition is asserted by
+# Finit upon receiving SIGINT (from the kernel).
+#task <sys/key/ctrlaltdel> [12345789] initctl reboot -- Rebooting system
+
+# Task to run if the kernel gets a power fail condition is pressed.
+# Asserted by Finit upon receiving SIGPWR (from the kernel).
+#task <sys/pwr/fail> [12345789] initctl poweroff -- Shutting down system
+
 ## Bootstrap services and tasks
 # Can be listed here or in /etc/finit.d/*.conf
-
@@ -1,7 +1,9 @@
 SUBDIRS    = available
 docsdir   := @docdir@/contrib/debian/finit.d
-d_LINKS    = console-setup.conf getty.conf keyboard-setup.conf	\
-	     modules.conf rsyslogd.conf sshd.conf
+d_LINKS    = acpid.conf apparmor.conf avahi-daemon.conf		\
+	     bluetooth.conf console-setup.conf elogind.conf 	\
+	     getty.conf keyboard-setup.conf modules.conf	\
+	     plymouth-quit.conf rsyslogd.conf sshd.conf

 install-exec-hook:
 	@$(MKDIR_P) $(DESTDIR)$(docsdir)/available
@@ -1,5 +1,8 @@
 docsdir   := @docdir@/contrib/debian/finit.d/available
-docs_DATA  = acpid.conf console-setup.conf getty.conf keyboard-setup.conf \
-	     lightdm.conf modules.conf rsyslogd.conf sntpd.conf		  \
-	     sshd.conf syslogd.conf xdm.conf
+docs_DATA  = acpid.conf apparmor.conf avahi-daemon.conf bluetooth.conf	\
+	     console-setup.conf elogind.conf getty.conf			\
+	     keyboard-setup.conf lightdm.conf modules.conf sntpd.conf	\
+	     plymouth-quit.conf rsyslogd.conf sshd.conf syslogd.conf	\
+	     xdm.conf
 EXTRA_DIST = $(docs_DATA)
+
@@ -0,0 +1 @@
+task [2345] <sys/pwr/ac,pid/syslogd> halt:usr1 env:-/etc/default/anacron anacron -d -q $ANACRON_ARGS -- Anacron jobs
@@ -0,0 +1,2 @@
+sysv [S] /etc/init.d/apparmor -- AppArmor initialization
+
@@ -0,0 +1 @@
+service [2345] env:-/etc/default/avahi-daemon avahi-daemon -s $AVAHI_DAEMON_OPTS -- Avahi mDNS/DNS-SD Stack
@@ -0,0 +1,3 @@
+# To ignore errors like "org.freedesktop.hostname1", update the env file with
+# "--noplugin=hostname", i.e. append to BLUETOOTHD_OPTS
+service [2345] log:null env:-/etc/default/bluetooth /usr/libexec/bluetooth/bluetoothd -n $BLUETOOTHD_OPTS -- Bluetooth service
@@ -0,0 +1 @@
+service [2345] <sys/pwr/ac,pid/syslogd> env:-/etc/default/cron cron -f $EXTRA_OPTS -- Background job processing daemon
@@ -0,0 +1,4 @@
+# On some systems, bootstrap can be really quick and cause elogind
+# to not start before a user has managed to log in.  To prevent
+# this, add <pid/elogind> as condition to your TTYs in getty.conf
+service [2345] <pid/syslogd> /lib/elogind/elogind -- Login manager
@@ -1,4 +1,7 @@
 # Consoles to start getty on when system is up
 tty [12345] /dev/tty1 linux noclear nowait
-tty [2345]  /dev/tty2 linux noclear nowait
-tty [2345]  /dev/tty3 linux noclear nowait
+tty [2345]  /dev/tty2 linux nowait
+tty [2345]  /dev/tty3 linux nowait
+tty [2345]  /dev/tty4 linux nowait
+tty [2345]  /dev/tty5 linux nowait
+tty [2345]  /dev/tty6 linux nowait
@@ -0,0 +1,2 @@
+# Need to stop plymouth, or it'll block stdin on tty1
+task [S] /usr/bin/plymouth quit -- Stopping Plymouth Boot Screen
@@ -1,14 +0,0 @@
-#!/bin/sh
-exec tail -n +3 $0
-# This file provides an easy way to add custom menu entries.  Simply type the
-# menu entries you want to add after this comment.  Be careful not to change
-# the 'exec tail' line above.
-
-menuentry '$MENU with Finit' --class $TYPE --class gnu-linux --class gnu --class os --user {
-	insmod gzio
-	insmod part_msdos
-	insmod ext2
-	search --no-floppy --fs-uuid --set=root $BOOT
-	linux	/vmlinuz root=UUID=$ROOT ro quiet init=/sbin/finit
-	initrd	/initrd.img
-}
@@ -1,3 +0,0 @@
-docsdir   := @docdir@/contrib/debian/grub.d
-docs_DATA  = 40_custom
-EXTRA_DIST = $(docs_DATA)
@@ -8,10 +8,10 @@ if [ "x`id -u`" != "x0" ]; then
 fi

 # Adjust base directory
-grep -q Debian os-release 2>/dev/null
+grep -q debian Makefile 2>/dev/null
 if [ $? -eq 0 ]; then
    cd ../..
-elif [ ! -e autogen.sh ]; then
+elif [ ! -e configure.ac ]; then
    echo "*** Please run this script from the Finit base directory."
    exit 1
 fi
@@ -23,7 +23,7 @@ echo "/sbin/finit           - PID 1"
 echo "/lib/finit/plugins/*  - All enabled Finit plugins"
 echo "/etc/finit.conf       - Finit configuration file"
 echo "/etc/finit.d/         - Finit services"
-echo "/etc/grub.d/40_custom - Add menu entry to the Grub boot loader"
+echo "/etc/grub.d/10_linux  - Add Finit to Grub's SUPPORTED_INITS"
 echo
 read -p "Do you want to continue (y/N)? " yorn
 echo
@@ -38,12 +38,18 @@ if [ "x$yorn" = "xy" -o "x$yorn" = "xY" ]; then
    cp -va finit.d /etc/

    echo "*** Setting up a GRUB boot entry ..."
-    MENU=`grep gnu-linux /boot/grub/grub.cfg |head -1 |sed "s/menuentry '\([^']*\).*/\1/"`
-    TYPE=`grep gnu-linux /boot/grub/grub.cfg |head -1 |sed "s/.*' --class \([a-z]*\).*/\1/"`
-    BOOT=`grep set=root /boot/grub/grub.cfg |head -1 |sed 's/.* \([a-f0-9]*-[a-f0-9]*-[a-f0-9]*-[a-f0-9]*-[a-f0-9]*\).*/\1/'`
-    ROOT=`grep root=UUID /boot/grub/grub.cfg |head -1 |sed 's/.*UUID=\([a-f0-9]*-[a-f0-9]*-[a-f0-9]*-[a-f0-9]*-[a-f0-9]*\).*/\1/'`
-    cat grub.d/40_custom | sed "s,\$BOOT,$BOOT,;s,\$ROOT,$ROOT,;s,\$TYPE,$TYPE,;s,\$MENU,$MENU," >/etc/grub.d/40_custom
-    update-grub
+    fn=/etc/grub.d/10_linux
+    if [ -e $fn ]; then
+	if `grep SUPPORTED_INITS $fn |head -1 |grep -q finit`; then
+	    echo "Already installed, done."
+	else
+	    echo "Adding Finit to list of SUPPORTED_INITS ..."
+	    sed -i 's/SUPPORTED_INITS="[^"]*/& finit:\/sbin\/finit/' $fn
+	    update-grub
+	fi
+    else
+	echo "Cannot find $fn, you'll have to set up your bootloader on your own."
+    fi

    echo
    echo "*** Done"
@@ -1,8 +0,0 @@
-PRETTY_NAME="Debian GNU/Linux 8 (jessie)"
-NAME="Debian GNU/Linux"
-VERSION_ID="8"
-VERSION="8 (jessie)"
-ID=debian
-HOME_URL="http://www.debian.org/"
-SUPPORT_URL="http://www.debian.org/support"
-BUG_REPORT_URL="https://bugs.debian.org/"
@@ -8,6 +8,18 @@ module evdev
 module loop
 module psmouse

+# Top-level cgroups and their default settings.  All groups mandatory
+# but more can be added, max 8 groups in total currently.  The cgroup
+# 'root' is also available, reserved for RT processes.  Settings are
+# as-is, only one shorthand 'mem.' exists, other than that it's the
+# cgroup v2 controller default names.
+cgroup init   cpu.weight:100
+cgroup user   cpu.weight:100
+cgroup system cpu.weight:9700 cpu.max:50000
+
+# Example extra cgroup
+cgroup maint  cpu.weight:100
+
 # Runlevel to start after bootstrap, runlevel 'S'
 runlevel 2

@@ -16,13 +28,21 @@ network service networking start

 # Max file size for each log file: 100 kiB, rotate max 4 copies:
 # log => log.1 => log.2.gz => log.3.gz => log.4.gz
-log size=100k count=4
+log size:100k count:4

 # Virtual consoles to start getty on
 tty /dev/tty1
 tty /dev/tty2
 tty /dev/tty3

+# Task to run if ctrl-alt-del is pressed.  This condition is asserted by
+# Finit upon receiving SIGINT (from the kernel).
+task <sys/key/ctrlaltdel> [12345789] initctl reboot -- Rebooting system
+
+# Task to run if the kernel gets a power fail condition is pressed.
+# Asserted by Finit upon receiving SIGPWR (from the kernel).
+task <sys/pwr/fail> [12345789] initctl poweroff -- Shutting down system
+
 # Alternative method instead of runparts
 #task [S] /etc/init.d/keyboard-setup start -- Setting up preliminary keymap
 #task [S] /etc/init.d/acpid start          -- Starting ACPI Daemon
@@ -30,9 +50,24 @@ tty /dev/tty3
 #run [2] /etc/init.d/networking start      -- Start networking

 # Services to be monitored and respawned as needed
+
+# klgod and syslogd are placed in the maint cgroup
+cgroup.maint
 service [2345] <pid/syslogd> klogd -n     -- Kernel logging server
 service [2345] syslogd -n                 -- Syslog server
-service [3]    gdm                        -- GNOME Display Manager
+
+# gdm in the system cgroup (default when read from it's own .conf file)
+cgroup.system
+service [3] cgroup:cpu.weight:250,mem.max:10M gdm -- GNOME Display Manager
+
+# Start SSH daemon with opts from /etc/default/ssh (if available), move
+# to user cgroup with cpu.weight:250.  Default for services is system
+# group.  Default weight in user group is 100 (above).  Max length of
+# cgroup argument is (currently) 63 chars.
+#
+# Notice alternative syntax where current cgroup is only set for this
+# particular service.
+service [2345] cgroup.user:cpu.weight:250,mem.max:10M env:-/etc/default/ssh /usr/sbin/sshd -D $SSHD_OPTS -- OpenSSH daemon

 # The BusyBox ntpd does not use syslog when running in the foreground
 # So we use this trick to redirect stdout/stderr to a log file.  The
@@ -1,4 +1,4 @@
 SUBDIRS    = finit.d
 docsdir   := @docdir@/contrib/void
-docs_DATA  = README.md finit.conf build.sh install.sh rc.local xbps
+docs_DATA  = README.md finit.conf build.sh install.sh rc.local
 EXTRA_DIST = $(docs_DATA)
@@ -17,8 +17,8 @@ PKG_CONFIG_LIBDIR=/usr/lib/pkgconfig:/usr/local/lib/pkgconfig ./configure	\
 		 --prefix=/usr                  --exec-prefix=			\
 		 --sysconfdir=/etc              --localstatedir=/var		\
                 --enable-dbus-plugin           --enable-x11-common-plugin	\
-		 --enable-alsa-utils-plugin	--enable-watchdog		\
-		 --with-heading="Void Linux"
+		 --enable-alsa-utils-plugin	--with-watchdog			\
+		 --with-keventd

 echo
 echo "*** Building ..."
@@ -5,5 +5,22 @@
 # Default is 2
 #runlevel 2

+# Top-level cgroups and their default settings.  All groups mandatory
+# but more can be added, max 8 groups in total currently.  The cgroup
+# 'root' is also available, reserved for RT processes.  Settings are
+# as-is, only one shorthand 'mem.' exists, other than that it's the
+# cgroup v2 controller default names.
+#cgroup init   cpu.weight:100
+#cgroup user   cpu.weight:100
+#cgroup system cpu.weight:9700
+
+# Task to run if ctrl-alt-del is pressed.  This condition is asserted by
+# Finit upon receiving SIGINT (from the kernel).
+#task <sys/key/ctrlaltdel> [12345789] initctl reboot -- Rebooting system
+
+# Task to run if the kernel gets a power fail condition is pressed.
+# Asserted by Finit upon receiving SIGPWR (from the kernel).
+#task <sys/pwr/fail> [12345789] initctl poweroff -- Shutting down system
+
 ## Bootstrap services and tasks
 # Can be listed here or in /etc/finit.d/*.conf
@@ -1,5 +1,5 @@
 docsdir   := @docdir@/contrib/void/finit.d/available
 docs_DATA  = acpid.conf crond.conf dhcpcd.conf dmeventd.conf getty.conf	\
-             keymap.conf klogd.conf lxdm.conf modules.conf ntpd.conf	\
-             sshd.conf syslogd.conf uuidd.conf
+             keymap.conf lxdm.conf modules.conf ntpd.conf		\
+             sshd.conf syslog.conf uuidd.conf
 EXTRA_DIST = $(docs_DATA)
@@ -1,2 +0,0 @@
-# Start klogd as soon as possible, should always run
-service [S12345] <pid/syslogd> klogd -n -- Kernel log daemon
@@ -4,3 +4,4 @@
 #module evdev
 #module mousedev
 #module 8139cp
+module softdog
@@ -0,0 +1,4 @@
+# Start syslogd as soon as possible, should always run
+# Start klogd as soon as possible, but after syslogd
+service [S12345] syslogd -n             -- Syslog daemon
+service [S12345] <pid/syslogd> klogd -n -- Kernel log daemon
@@ -1,2 +0,0 @@
-# Start syslogd as soon as possible, should always run
-service [S12345] syslogd -n -- Syslog daemon
@@ -8,9 +8,10 @@ if [ "x`id -u`" != "x0" ]; then
 fi

 # Adjust base directory
-if [ -e xbps ]; then
+grep -q void Makefile 2>/dev/null
+if [ $? -eq 0 ]; then
    cd ../..
-elif [ ! -e autogen.sh ]; then
+elif [ ! -e configure.ac ]; then
    echo "*** Please run this script from the Finit base directory."
    exit 1
 fi
@@ -19,19 +19,31 @@ Some of these are also registered in the issue tracker on GitHub.
 * Add new (complimentary) .svc file format to slowly migrate away
  from the very terse one-liner format currently used
 * Allow running as non-pid1 => read .conf and RCSD form cmdline
-* When a process dies, and Finit does not restart it, we should collect
-  a `siginfo_t` from the SIGCHLD signal and supply to the user in case
-  `initctl status <process>` is called to debug the issue.
 * Add `finit.conf` support for UPS notification (SIGPWR) to start a task
  using, e.g. <sys/power/{ok,fail,low}> conditions.  More info in sig.c
+
+  As of Finit v4.1 SIGPWR generates <sys/power/fail> condition and it
+  is up to a task to check why and take appropriate action.
 * Add `finit.conf` support for ctrl-alt-delete (SIGINT) and kbrequest,
  i.e. KeyboardSignal, (SIGWINCH) behavior.  Using conditions to a task,
  e.g, <sys/key/ctrlaltdel> and <sys/key/signal> like SIGPWR handling.
-* Write man pages for finit and `finit.conf`, steal from the excellent
-  `pimd` man pages ...
-* Add support for timed shutdown in Finit, including cancelled shutdown
+
+  As of Finit v4.1 SIGINT generates <sys/key/ctrlaltdel>, but there is
+  no check if the signal actually originates from the kernel.  To get
+  this verification, a new libuEv release is needed.  As of June 6 -21
+  that functionality is not yet released.
+
+
+Initctl
+-------
+
+* Add support for timed shutdown in Finit, including cancelled shutdown:
+
+        shutdown -h 03:35 "foobar was here"
+
 * Add support for JSON output, or similar, from `initctl show`, e.g.
  `initctl show --json`
+* Add support for `-s` to redirect all output to syslog, for scripting


 Init
@@ -9,9 +9,11 @@ Bootstrap
 6. Load all `.so` plugins
 7. Call `fsck` on file systems listed in `/etc/fstab`
 8. Populate `/dev` using either udev or mdev, depending on system type
-9. Parse `/etc/finit.conf` and all `/etc/finit.d/*.conf` files
-10. Start built-in watchdog, if enabled
-11. Set hostname
+9. Parse `/etc/finit.conf`
+10. Start built-in watchdog, if enabled and `WDT_DEVNODE` exists.  This
+    means any WDT that requires a kernel module need to be either
+    compiled into the kernel, or insmod'ed in `/etc/finit.conf`
+11. Load all `/etc/finit.d/*.conf` files and set hostname
 12. Remount `/` read-write if `/` is listed in `/etc/fstab` without `ro`
 13. Call 1st level hooks, `HOOK_ROOTFS_UP`
 14. Mount all file systems listed in `/etc/fstab` and swap, if available
@@ -34,7 +36,7 @@ Bootstrap
 24. Call 4th level hooks, `HOOK_SVC_UP`
 25. Call `/etc/rc.local`, if it exists and is an executable shell script
 26. Call 5th level (last) hooks, `HOOK_SYSTEM_UP`
-27. Start all configured TTYs, or a fallback shell on `/dev/console`
+27. Start all configured TTYs

 In (19) and (22) tasks and services defined in `/etc/finit.conf` and
 `/etc/sysctl.d/*.conf` are started.  Remember, all `service` and `task`
@@ -53,6 +53,10 @@ Below are a few of the main switches to configure:
  built-ins (.o files) and all external libraries, except the C library
  will be linked statically.

+* `--enable-kernel-cmdline`: Enable Finit pre-4.1 parsing of init args from
+  `/proc/cmdline`, this is not recommended since Finit may be running as the
+  init for container apps that can see the host's `/proc` filesystem
+
 * `--enable-alsa-utils-plugin`: Enable the optional `alsa-utils.so` sound plugin.

 * `--enable-dbus-plugin`: Enable the optional D-Bus `dbus.so` plugin.
@@ -70,16 +74,16 @@ Example
 First, unpack the archive:

 ```shell
-$ tar xf finit-3.0.tar.xz
-$ cd finit-3.0/
+$ tar xf finit-4.1.tar.gz
+$ cd finit-4.1/
 ```

 Then configure, build and install:

 ```shell
-$ ./configure --prefix=/usr                 --exec-prefix=                    \
-              --sysconfdir=/etc             --localstatedir=/var              \
-              --with-heading="Alpine Linux 3.13" --with-hostname=alpine
+$ ./configure --prefix=/usr                 --exec-prefix=         \
+              --sysconfdir=/etc             --localstatedir=/var   \
+			  --with-keventd                --with-watchdog
 $ make
 .
 .
@@ -87,13 +91,13 @@ $ make
 $ DESTDIR=/tmp/finit make install
 ```

-In this example the [finit-3.0.tar.xz][1] archive is unpacked to the
+In this example the [finit-4.1.tar.gz][1] archive is unpacked to the
 user's home directory, configured, built and installed to a temporary
 staging directory.  The environment variable `DESTDIR` controls the
 destination directory when installing, very useful for building binary
 standalone packages.

-Finit 3.0 and later can detect if it runs on an embedded system, or a
+Finit 4.1 and later can detect if it runs on an embedded system, or a
 system that use BusyBox tools instead of udev & C:o.  On such systems
 `mdev` instead of `udev` is used.  However, remember to also change the
 Linux config to:
@@ -153,21 +157,24 @@ kernel messages and add `debug` to enable Finit debug messages.
 append="init=/sbin/finit debug"
 ```

-To debug startup issues, in particular issues with getty/login, try
-`configure --enable-fallback-shell`.  When no TTYs are detected, and
-Finit is configured with this option, Finit will try to start a bare
-`/bin/sh` on the boot console.  Remember, this is only for debugging
-and would leave your production system potentially wide open.
+To debug startup issues, in particular issues with getty/login, add
+the following to your Finit .conf file:
+
+    tty [12345789] notty noclear
+
+The `notty` option ensures reusing the stdin/stdout set up by the
+kernel.  Remember, this is only for debugging and would leave your
+production system potentially wide open.

 There is also a rescue shell available, in case Finit crashes and the
 kernel usually reboots: `configure --enable-emergency-shell`.  However,
 the behavior of Finit is severely limited when this is enabled, so use
 it only for debugging start up issues when Finit crashes.

-**NOTE:** Neither of these two configure options should be enabled on
-  production systems since they can potentially give a user root access.
+**NOTE:** Neither of these options should be enabled on production
+         systems since they can potentially give a user root access.


-[1]:       ftp://troglobit.com/finit/finit-3.0.tar.xz
+[1]:       ftp://troglobit.com/finit/finit-4.1.tar.gz
 [libuEv]:  https://github.com/troglobit/libuev
 [libite]:  https://github.com/troglobit/libite
@@ -6,6 +6,8 @@ passed on the Linux kernel command line.  Not limited to Finit.

 The `bool` setting is one of `on, off, true false, 1, 0`.

+> **NOTE:** remember to use `--` to separate kernel parameters from
+> parameters to init.  E.g., `init=/sbin/finit -- finit.debug rescue`

 * `debug`: Enable kernel debug.  Debug messages are printed to the
    console until Finit starts up.
@@ -30,10 +32,12 @@ The `bool` setting is one of `on, off, true false, 1, 0`.
 * `quiet`: Suppress kernel logging to console, except for warnings and
  errors.

-* `rescue`: Start Finit rescue mode; no network, no `fsck` of file
-    systems in `/etc/fstab`, no `/etc/rc.local`, no `runparts`.  The
-    configuration is read from `/lib/finit/rescue.conf`, and can be
-    modified by the operator.
+* `rescue`: Start rescue/maintenance mode.  If your system comes with
+    the bundled `sulogin` program (Finit, or from util-linux/Busybox),
+    you will be given a root login to a maintenance shell.  However, if
+    `sulogin` is missing, the file `/lib/finit/rescue.conf` is read and
+    the system booted in a limited fallback mode.  See [config.md][]
+	for more information.

 * `single`, or `S`: Single user mode, runlevel 1, in Finit.  Useful to
    debug problems with the regular runlevel.  All services and TTYs in
@@ -42,5 +46,5 @@ The `bool` setting is one of `on, off, true false, 1, 0`.

 For more on kernel boot paramaters, see the man page [bootparam(7)][].

-
+[config.md][]:  config.md#rescue-mode
 [bootparam(7)]: https://www.man7.org/linux/man-pages/man7/bootparam.7.html
@@ -45,7 +45,31 @@ being created.
 Triggering
 ----------

-Conditions can only be triggered by built-in plugins.
+Conditions are mainly triggered (asserted) by built-in plugins, e.g.,
+`netlink.so` and `pidfile.so`, see below for built-in conditions.  Finit
+also supports user-defined conditions, sometimes referred to as static
+or one-shot conditions.  They live in the `<usr/...` namespace and are
+constrained to a flat hierarchy without any sub-directories, unlike the
+pidfile plugin, for instance.
+
+User-defined conditions are controlled using the `initctl cond set` and
+`initctl cond clear` commands:
+
+    initctl cond set foo
+    initctl cond clear foo
+
+The purpose of user-defined conditions is to be able to start or stop
+services, or run/task jobs, on external site-dependent stimuli.
+Example:
+
+    service [2345] <usr/foo> alarm --arg foo -- Foo alarm
+
+For convenience, prefixing with `usr/` is allowed, but any other slashes
+or period characters are disallowed.  E.g., to trigger the `Foo alarm`,
+the same as above, can also be achieved like this:
+
+    initctl cond set usr/foo
+    initctl cond clear usr/foo

 Conditions retain their current state until the next reconfiguration or
 runlevel change.  At that point all set conditions transition into the
@@ -57,16 +81,21 @@ unchanged) state of it.
 > **Note:** For `pid/` conditions it is expected that services "touch"
 >           or recreate their PID file on `SIGHUP`.

+Static (one-shot) conditions, like `usr/`, never enter the `flux` state.
+

 Built-in Conditions
 -------------------

-Finit comes with a `pidfile` and `netlink` plugin.  When enabled, the
-`pidfile` plugin watches `/var/run/` (recursively) for PID files created
-by the monitored services, and sets a corresponding condition in the
-`pid/` namespace.  Similarily, the `netlink` plugin provides basic
-conditions for when an interface is brought up/down and when a default
-route (gateway) is set, in the `net/` namespace.
+Finit comes with a set of plugins for conditions: `pidfile`, `netlink`,
+`sys`, and `usr`.  The `pidfile` plugin (recursively) watches `/run/`
+(recursively) for PID files created by the monitored services, and sets
+a corresponding condition in the `pid/` namespace.  Similarily, the
+`netlink` plugin provides basic conditions for when an interface is
+brought up/down and when a default route (gateway) is set, in the `net/`
+namespace.  The `sys` and `usr` plugins monitor are passive condition
+monitors where the action is provided by `keventd`, signal handlers,
+and in the case of `usr`, the end-user via the `initctl` tool

 With the example listed above, finit does not start the `/sbin/netd`
 daemon until `setupd` and `zebra` has started *and* created their PID
@@ -85,6 +114,10 @@ Built-in conditions:
 - `net/<IFNAME>/exist`
 - `net/<IFNAME>/up`
 - `net/<IFNAME>/running`
+- `sys/pwr/ac`
+- `sys/pwr/fail`
+- `sys/key/ctrlaltdel`
+- `usr/foo`

 **Note:** `up` means administratively up, the interface flag `IFF_UP`.
  `running` is the `IFF_RUNNING` flag, meaning operatively up.  The
@@ -2,7 +2,9 @@ Configuration
 =============

 * [Introduction](#introduction)
+* [Service Environment](#service-environment)
 * [Service Wrapper Scripts](#service-wrapper-scripts)
+* [Cgroups](#cgroups)
 * [Configuration File Syntax](#configuration-file-syntax)
  * [Hostname](#hostname)
  * [Kernel Modules](#kernel-modules)
@@ -19,9 +21,12 @@ Configuration
  * [TTYs and Consoles](#ttys-and-consoles)
  * [Non-privileged Services](#non-privileged-services)
  * [Redirecting Output](#redirecting-output)
+* [Rescue Mode](#rescue-mode)
 * [Limitations](#limitations)
  * [/etc/finit.conf](#etcfinitconf)
  * [/etc/finit.d](#etcfinitd)
+* [Watchdog](#watchdog)
+* [keventd](#keventd)


 Introduction
@@ -35,16 +40,16 @@ Linux distributions -- each package can provide its own "script" file.
 - `/etc/finit.d/*.conf`: snippets, usually one service per file

 Not all configuration directives are available in `/etc/finit.d/*.conf`
-and some directives are only available at bootstrap, runlevel `S`, see
-the section [Limitations](#limitations) below for details.
+and some directives are only available at [bootstrap][], runlevel `S`,
+see the section [Limitations](#limitations) below for details.

-To add a new service, simply drop a `.conf` file in `/etc/finit.d` and
-run `initctl reload`.  (It is also possible to `SIGHUP` to PID 1, or
-call `finit q`, but that has been deprecated with the `initctl` tool).
-Finit monitors all known active `.conf` files, so if you want to force
-a restart of any service you can simply touch its corresponding `.conf`
-file in `/etc/finit.d` and call `initctl reload`.  Finit handle any
-and all conditions and dependencies between services automatically.
+To add a new service, drop a `.conf` file in `/etc/finit.d` and run
+`initctl reload`.  (It is also possible to `SIGHUP` to PID 1, or call
+`finit q`, but that has been deprecated with the `initctl` tool).  Finit
+monitors all known active `.conf` files, so if you want to force a
+restart of any service you can touch its corresponding `.conf` file in
+`/etc/finit.d` and call `initctl reload`.  Finit handle any and all
+conditions and dependencies between services automatically.

 It is also possible to drop `.conf` files in `/etc/finit.d/available/`
 and use `initctl enable` to enable a service `.conf` file.  This may be
@@ -128,6 +133,67 @@ example employs a wrapper script in `/etc/start.d`.
 >           instead of sending SIGHUP at restart/reload events.


+Cgroups
+-------
+
+There are three major cgroup configuration directives:
+
+ 1. Global top-level group: init, system, user, or a custom group
+ 2. Selecting a top-level group for a set of run/task/services
+ 3. Per run/task/service limits
+
+Top-level group configuration.
+
+    # Top-level cgroups and their default settings.  All groups mandatory
+    # but more can be added, max 8 groups in total currently.  The cgroup
+    # 'root' is also available, reserved for RT processes.  Settings are
+    # as-is, only one shorthand 'mem.' exists, other than that it's the
+    # cgroup v2 controller default names.
+    cgroup init   cpu.weight:100
+    cgroup user   cpu.weight:100
+    cgroup system cpu.weight:9800
+
+Adding an extra cgroup `maint/` will require you to adjust the weight of
+the above three.  We leave `init/` and `user/` as-is reducing weight of
+`system/` to 9700.
+
+    cgroup system cpu.weight:9700
+
+    # Example extra cgroup 'maint'
+    cgroup maint  cpu.weight:100
+
+By default, the `system/` cgroup is selected for almost everything.  The
+`init/` cgroup is reserved for PID 1 itself and its closest relatives.
+The `user/` cgroup is for local TTY logins spawned by getty.
+
+To select a different top-level cgroup, e.g. `maint/`, one can either
+define it for a group of run/task/service directives in a `.conf` or per
+each stanza:
+
+    cgroup.maint
+    service [...] <...> /path/to/foo args -- description
+    service [...] <...> /path/to/bar args -- description
+
+or
+
+    service [...] <...> cgroup.maint /path/to/foo args -- description
+
+The latter form also allows per-stanza limits on the form:
+
+    service [...] <...> cgroup.maint:cpu.max:10000,mem.max:655360 /path/to/foo args -- description
+
+Notice the comma separation and the `mem.` exception to the rule: every
+cgroup setting maps directly to cgroup v2 syntax.  I.e., `cpu.max` maps
+to the file `/sys/fs/cgroup/maint/foo/cpu.max`.  There is no filtering,
+except for expanding the shorthand `mem.` to `memory.`, if the file is
+not available, either the cgroup controller is not available in your
+Linux kernel, or the name is misspelled.
+
+> Linux cgroups and details surrounding values are not explained in the
+> Finit documentation.  The Linux admin-guide cover this well:
+> <https://www.kernel.org/doc/html/latest/admin-guide/cgroup-v2.html>
+
+
 Configuration File Syntax
 -------------------------

@@ -140,7 +206,7 @@ the contents of that file is used.

 Deprecated.  We recommend using `/etc/hostname` instead.

-> **Note:** only read and executed in runlevel S (bootstrap).
+> **Note:** only read and executed in runlevel S ([bootstrap][]).


 ### Kernel Modules
@@ -158,7 +224,7 @@ BusyBox mdev tool, add to `/etc/mdev.conf`:

    $MODALIAS=.*  root:root       0660    @modprobe -b "$MODALIAS"

-> **Note:** only read and executed in runlevel S (bootstrap).
+> **Note:** only read and executed in runlevel S ([bootstrap][]).


 ### Networking
@@ -171,7 +237,7 @@ Deprecated.  We recommend using dedicated task/run stanzas per runlevel,
 or `/etc/network/interfaces` if you have a system with `ifupdown`, like
 Debian, Ubuntu, Linux Mint, or an embedded BusyBox system.

-> **Note:** only read and executed in runlevel S (bootstrap).
+> **Note:** only read and executed in runlevel S ([bootstrap][]).


 ### Resource Limits
@@ -205,13 +271,13 @@ stanzas can share the same rlimits if they are in the same .conf.

 **Syntax:** `runlevel <N>`

-Defines the system runlevel to go to after bootstrap (S) has completed.
-N is the runlevel number 0-9, where 6 is reserved for reboot and 0 for
+The system runlevel to go to after [bootstrap][] (S) has completed.  `N`
+is the runlevel number 0-9, where 6 is reserved for reboot and 0 for
 halt.

 Default: 2

-> **Note:** only read and executed in runlevel S (bootstrap).
+> **Note:** only read and executed in runlevel S ([bootstrap][]).


 ### One-shot Commands (sequence)
@@ -333,6 +399,28 @@ process using a different signal, use the option `halt:SIGNAL`, e.g.,
 use the option `kill:SEC`, e.g., `kill:10` to wait 10 seconds before
 sending `SIGKILL`.

+Services support `pre:script` and `post:script` actions as well.  These
+run as the same `@USER:GROUP` as the service itself, with any `env:file`
+sourced.  The scripts must use an absolute path, but are executed from
+the `$HOME` of the given user.  The scripts are not called with any
+argument (currently), but both get the `SERVICE_IDENT=foo` environment
+variable set.  Here `foo` denotes the identity of the service, which if
+there are multiple services named `foo`, may be `foo:1`, or any unique
+identifier specified in the .conf file.  The `post:script` is called
+with an additional set of environment variables:
+
+ - `EXIT_CODE=[exited,signal]`: set to one of `exited` or `signal`
+ - `EXIT_STATUS=[num,SIGNAME]`: set to one of exit status code from
+   the program, if it exited normally, or the signal name (`HUP`,
+   `TERM`, etc.) if it exited due to signal
+
+The scripts have a default execution time of 3 seconds before they are
+SIGKILLed, this can be adjusted using the above `kill:SEC` syntax.
+
+> **Note:** the `pre:script` _must_ be idempotent, because a service
+>           can transition between READY and HALTED states any number
+>           of times before going to RUNNING.
+

 ### Run-parts Scripts

@@ -341,7 +429,7 @@ sending `SIGKILL`.
 Call [run-parts(8)][] on `DIR` to run start scripts.  All executable
 files, or scripts, in the directory are called, in alphabetic order.
 The scripts in this directory are executed at the very end of runlevel
-`S`, bootstrap.
+`S`, [bootstrap][].

 It can be beneficial to use `S01name`, `S02othername`, etc. if there
 is a dependency order between the scripts.  Symlinks to existing
@@ -351,7 +439,10 @@ Similar to the `/etc/rc.local` shell script, make sure that all your
 services and programs either terminate or start in the background or
 you will block Finit.

-> **Note:** only read and executed in runlevel S (bootstrap).
+> **Note:** only read and executed in runlevel S ([bootstrap][]).
+
+[run-parts(8)]: http://manpages.debian.org/cgi-bin/man.cgi?query=run-parts
+

 ### Including Finit Configs

@@ -377,24 +468,40 @@ size limit is reached.

 ### TTYs and Consoles

-**Syntax:** `tty [LVLS] <DEV> [BAUD] [noclear] [nowait] [nologin] [TERM]`  
-  `tty [LVLS] <CMD> <ARGS> [noclear] [nowait]`
+**Syntax:** `tty [LVLS] <COND> DEV [BAUD] [noclear] [nowait] [nologin] [TERM]`  
+  `tty [LVLS] <COND> CMD <ARGS> [noclear] [nowait]`  
+  `tty [LVLS] <COND> [notty] [rescue]`

 The first variant of this option uses the built-in getty on the given
-TTY device DEV, in the given runlevels.  The DEV may be the special
-keyword `@console`, or `console`, useful on embedded systems.
+TTY device DEV, in the given runlevels.  DEV may be the special keyword
+`@console`, which is expanded from `/sys/class/tty/console/active`,
+useful on embedded systems.

 The default baud rate is 0, i.e., keep kernel default.

+> The `tty` stanza inherits runlevel, condition (and other feature)
+> parsing from the `service` stanza.  So TTYs can run in one or many
+> runlevels and depend on any condition supported by Finit.  This is
+> useful e.g. to depend on `<pid/elogind>` before starting a TTY.
+
 **Example:**

    tty [12345] /dev/ttyAMA0 115200 noclear vt220

 The second `tty` syntax variant is for using an external getty, like
 agetty or the BusyBox getty.
-    
-By default both variants *clear* the TTY and *wait* for the user to
-press enter before starting getty.
+
+The third variant is for board bringup and the `rescue` boot mode.  No
+device node is required in this variant, the same output that the kernel
+uses is reused for stdio.  If the `rescue` option is omitted, a shell is
+started (`nologin`, `noclear`, and `nowait` are implied), if the rescue
+option is set the bundled `/libexec/finit/sulogin` is started to present
+a bare-bones root login prompt.  If the root (uid:0, gid:0) user does
+not have a password set, no rescue is possible.  For more information,
+see the [Rescue Mode](#rescue-mode) section.
+
+By default, the first two syntax variants *clear* the TTY and *wait* for
+the user to press enter before starting getty.

 **Example:**

@@ -423,16 +530,23 @@ uses `ttyS0`, `ttyAMA0`, `ttyMXC0`, or anything else.  Finit figures
 it out by querying sysfs: `/sys/class/tty/console/active`.  The speed
 can be omitted to keep the kernel default.

+> Most systems get by fine by just using `console`, which will evaluate
+> to `/dev/console`.  If you have to use `@console` to get any output,
+> you may have some issue with your kernel config.
+
 **Example:**

    tty [12345] @console noclear vt220

-On really bare bones systems Finit offers a fallback shell, which
-should not be enabled on production systems since.  This because it
-may give a user root access without having to log in.  However, for
-bringup and system debugging it can come in handy:
+On really bare bones systems, or for board bringup, Finit can give you a
+shell prompt as soon as [bootstrap][] is done, without opening any
+device node:

-    configure --enable-fallback-shell
+    tty [12345789] notty
+
+This should of course not be enabled on production systems.  Because it
+may give a user root access without having to log in.  However, for
+board bringup and system debugging it can come in handy.

 One can also use the `service` stanza to start a stand-alone shell:

@@ -442,8 +556,8 @@ One can also use the `service` stanza to start a stand-alone shell:
 ### Non-privileged Services

 Every `run`, `task`, or `service` can also list the privileges the
-`/path/to/cmd` should be executed with.  Simply prefix the path with
-`[@USR[:GRP]]` like this:
+`/path/to/cmd` should be executed with.  Prefix the command with
+`@USR[:GRP]`, group is optional, like this:

    run [2345] @joe:users logger "Hello world"

@@ -483,28 +597,129 @@ Log rotation is controlled using the global `log` setting.

    service log:prio:user.warn,tag:ntpd /sbin/ntpd pool.ntp.org -- NTP daemon

-Worth noting is that conditions is allowed for all these stanzas.  For a
-detailed description, see the [Conditions](conditions.md) document.
+
+Rescue Mode
+-----------
+
+Finit supports a rescue mode which is activated by the `rescue` option
+on the kernel command line.  See [cmdline docs](cmdline.md) for how to
+activate it.
+
+The rescue mode comes in two flavors; *traditional* and *fallback*.
+
+
+### Traditional
+
+This is what most users expect.  A very early maintenance login prompt,
+served by the bundled `/libexec/finit/sulogin` program, or the standard
+`sulogin` from util-linux or BusyBox is searched for in the UNIX default
+`$PATH`.  If a successful login is made, or the user exits (Ctrl-D), the
+rescue mode is ended and the system boots up normally.
+
+> **Note:** if the user (UID 0 and GID 0) does not have a password, or
+> __the account is locked__, the user is presented with a password-less
+> `"Press enter to enter maintenance mode."`, prompt which opens up a
+> root shell.
+
+
+### Fallback
+
+If no `sulogin` program is found, Finit tries to bring up as much of its
+own functionality as possible, yet limiting many aspects, meaning; no
+network, no `fsck` of file systems in `/etc/fstab`, no `/etc/rc.local`,
+no `runparts`, and most plugins are skipped (except those that provide
+functionality for the condition subsystem).
+
+Instead of reading `/etc/finit.conf` et al, system configuration is read
+from `/lib/finit/rescue.conf`, which can be freely modified by the
+system administrator.
+
+The bundled default `rescue.conf` contains nothing more than:
+
+    runlevel 1
+    tty [12345] rescue
+
+The `tty` has the `rescue` option set, which works similar to the board
+bring-up tty option `notty`.  The major difference being that `sulogin`
+is started to query for root/admin password.  If `sulogin` is not found,
+`rescue` behaves like `notty` and gives a plain root shell prompt.
+
+If Finit cannot find `/lib/finit/rescue.conf` it defaults to:
+
+    tty [12345] rescue
+
+There is no way to exit the *fallback* rescue mode.


 Limitations
 -----------

 As of Finit v4 there are no limitations to where `.conf` settings can be
-placed.  Except for the system/global `rlimit`, which can only be set
-from `/etc/finit.conf`, since it is the first `.conf` file Finit reads.
+placed.  Except for the system/global `rlimit` and `cgroup` top-level
+group declarations, which can only be set from `/etc/finit.conf`, since
+it is the first `.conf` file Finit reads.

 Originally, `/etc/finit.conf` was the only way to set up a Finit system.
 Today it is mainly used for bootstrap settings like system hostname,
-network bringup and system shutdown.  These can now also be set in any
-`.conf` file in `/etc/finit.d`.
+early module loading for watchdogd, network bringup and system shutdown.
+These can now also be set in any `.conf` file in `/etc/finit.d`.

 There is, however, nothing preventing you from having all configuration
 settings in `/etc/finit.conf`.

 > **Note:** The `/etc/finit.d` directory was previously the default
->          Finit `runparts` directory.  Finit no longer has a default
->          `runparts`, make sure to update your setup, or the finit
->          configuration, accordingly.
+>          Finit [runparts](#run-parts-scripts) directory.  Finit no
+>          longer has a default `runparts`, make sure to update your
+>          setup, or the finit configuration, accordingly.

-[run-parts(8)]: http://manpages.debian.org/cgi-bin/man.cgi?query=run-parts
+
+Watchdog
+--------
+
+When built `--with-watchdog` a separate service is built and installed
+in `/libexec/finit/watchdogd`.  If this exists at runtime, and the WDT
+device node exists, Finit will start it and treat it as the elected
+watchdog service to delegate its reboot to.  See [bootstrap][] for
+details.  This delegation is to ensure that the system is rebooted by a
+hardware watchdog timer -- on many embedded systems this is crucial to
+ensure all circuits on the board are properly reset for the next boot,
+in effect ensuring the system works the same after both a power-on and
+reboot event.
+
+The delegation is performed at the very last steps of system shutdown,
+if reboot has been selected and an elected watchdog is known, first a
+`SIGPWR` is sent to advise watchdogd of the pending reboot.  Then, when
+the necessary steps of preparing the system for shutdown (umount etc.)
+are completed, Finit sends `SIGTERM` to watchdogd and puts itself in a
+10 sec timeout loop waiting for the WDT to reset the board.  If a reset
+is not done before the timeout, Finit falls back to`reboot(RB_AUTOBOOT)`
+which tells the kernel to do the reboot.
+
+An external watchdog service can also be used.  The more advanced cousin
+[watchdogd](https://github.com/troglobit/watchdogd/) is the recommended
+option here.  It can register itself with Finit using the same IPC as
+`initctl`.  If the bundled watchdogd is running a hand-over takes place,
+so it's safe to have both services installed on a system.  For the
+hand-over to work it requires that the WDT driver supports the safe exit
+functionality where `"V"` is written to the device before closing the
+device descriptor.  If the kernel driver has been built without this,
+the only option is to remove `/libexec/finit/watchdogd` or build without
+it at configure time.
+
+[bootstrap]: bootstrap.md
+
+
+keventd
+-------
+
+The kernel event daemon bundled with Finit is a simple uevent monitor
+for `/sys/class/power_supply`.  It provides the `sys/pwr/ac` condition,
+which can be useful to prevent power hungry services like anacron to run
+when a laptop is only running on battery, for instance.
+
+Since keventd is not an integral part of Finit yet it is not enabled by
+default.  Enable it using `./configuure --with-keventd`.  The bundled
+contrib build scripts for Debian, Alpine, and Void have this enabled.
+
+This daemon is planned to be extended with monitoring of other uevents,
+patches and ideas are welcome in the issue tracker.
@@ -12,28 +12,36 @@ have been adopted:
 Same effect as `finit q`, `init q`, or `initctl reload`, reloads all
 *.conf files in `/etc/finit.d/`

+This also restarts the API (initctl) socket, like SysV init and systemd
+does on USR1 with their FIFO/D-Bus.
+
+
 `SIGUSR1`
 ---------

-Calls shutdown hooks, including `HOOK_SHUTDOWN`, stopping all running
-processes, and unmounts all file systems.  Then tells kernel to halt.
-Most people these days want `SIGUSR2` though.
+Since Finit 4.1 this signal causes Finit to restart its API (initctl)
+socket, like SysV init and systemd does on USR1 with their FIFO/D-Bus.
+
+Finit <= 4.0 performed a system halt (like USR2 without power-off), but
+this caused compatibility problems with systemd and sysvinit on desktop
+systems.  Hence, since Finit 4.1 it is no longer possible to halt a
+system with a signal.

-SysV init and systemd use this to re-open their FIFO/D-Bus.

 `SIGUSR2`
 ---------

-Like SIGUSR1, but tell kernel to power-off the system, if ACPI or
-similar exists to actually do this.  If the kernel fails power-off,
-Finit falls back to halt.
- 
+Calls shutdown hooks, including `HOOK_SHUTDOWN`, stopping all running
+processes, and unmounts all file systems.  Then tells kernel to power
+off the system, if ACPI or similar exists to actually do this.  If the
+kernel fails power-off, Finit falls back to halt.
+
 SysV init N/A, systemd dumps its internal state to log.

 `SIGTERM`
 ---------

-Like `SIGUSR1`, but tell kernel to reboot the system when done.
+Like `SIGUSR2`, but tell kernel to reboot the system when done.
 
 SysV init N/A, systemd rexecutes itself.

@@ -0,0 +1,2 @@
+Makefile
+Makefile.in
@@ -0,0 +1,2 @@
+dist_man8_MANS = finit.8 initctl.8
+dist_man5_MANS = finit.conf.5
@@ -0,0 +1,178 @@
+.\"                                      Hey, EMACS: -*- nroff -*-
+.\" First parameter, NAME, should be all caps
+.\" Second parameter, SECTION, should be 1-8, maybe w/ subsection
+.\" other parameters are allowed: see man(7), man(1)
+.Dd June 6, 2021
+.\" Please adjust this date whenever revising the manpage.
+.Dt FINIT 8 SMM
+.Os Linux
+.Sh NAME
+.Nm finit
+.Nd Fast init for Linux systems
+.Sh SYNOPSIS
+.Nm /sbin/finit
+.Op finit.status_style=classic
+.Op finit.show_status=[on,off]
+.Op finit.debug
+.Op rescue | recover
+.Op single | S
+.Nm /sbin/telinit
+.Op Fl hv
+.Op q | Q | 0-9
+.Sh DESCRIPTION
+Fast init for Linux systems.  Reverse engineered from the EeePC
+fastinit, 10+ years ago by Claudio Matsuoka, "gaps filled with frog
+DNA ...".  Focus is on small and embedded systems, although Finit is
+fully usable on server and desktop systems as well.
+.Pp
+Features include:
+.Bl -bullet -width 1n -compact
+.It
+Runlevels, defined per service
+.It
+One-shot tasks, services (daemons), or SysV init start/stop scripts
+.It
+Runparts and
+.Pa /etc/rc.local
+support
+.It
+Process supervision similar to
+.Xr systemd 8
+.It
+Sourcing environment files
+.It
+Conditions for network/process/custom dependencies
+.It
+Pre/Post script actions
+.It
+Tooling to enable/disable services
+.It
+Built-in getty
+.It
+Built-in watchdog, with support for hand-over to
+.Lk https://troglobit.com/watchdogd.html watchdogd
+.It
+Built-in support for Debian/BusyBox
+.Pa /etc/network/interfaces
+.It
+Cgroups v2, both configuration and monitoring in `initctl top`
+.It
+Plugin support for customization
+.It
+Proper rescue mode with bundled `sulogin` for protected maintenance shell
+.El
+.Sh BOOTING
+.Pp
+Before laucnhing services, the system needs to be bootstrapped.  This
+involves mounting all filesystems (not already mounted by an initramfs)
+in
+.Pa /etc/fstab ,
+ensuring that necessary filesystems like:
+.Pa /dev ,
+.Pa /proc ,
+.Pa /sys ,
+and
+.Pa /run
+are available and can provide the system with the services required.
+.Pp
+With that out of the way, a set of plugins take over:
+.Pp
+.Bl -tag -width bootmisc.so -offset indent
+.It Cm bootmisc.so
+Cleans up stale files from previous boots and default directories are
+created in tmpfs filesystems
+.It Cm modprobe.so
+Probes kernel modules required by the system
+.It Cm procps.so
+Set kernel variables from
+.Pa /etc/sysctl.conf
+et al
+.El
+.Pp
+When the plugins are done, finit reads
+.Pa /etc/finit.conf
+and all other enabled services in
+.Pa /etc/finit.d/* .
+First all services, run/tasks in runlevel S (bootStrap) are started,
+When all run/tasks have completed, Finit changes to the configured
+runlevel (default 2), Processes are started in parallel, in dependency
+order according to their conditions.
+.Pp
+When all services and run/tasks have been started, the console progress
+is disabled and all configured getty services are started.
+.Sh CONDITIONS
+The condition subysystem in Finit was initially created to track PID
+files of daemons.  If service/task B requires service A to have started,
+Finit waits for the PID file of service A to be created/updated before
+it starts service B.  Conversely, if service A is stopped and its PID
+file removed, service B is also stopped.
+.Pp
+The following condition families are availble today:
+.Pp
+.Bl -tag -width pid -offset indent
+.It Cm net
+Linux netlink events, e.g. net/route/default, net/eth0/up, and
+net/et0/running
+.It Cm pid
+PID files basd on the service declaration
+.Cm name:id ,
+gives the condition pid/name:id
+.It Cm sys
+System conditions, e.g. sys/key/ctrlaltdel and sys/pwr/fail
+.It Cm usr
+User conditions, assert and deassert using
+.Cm initctl cond [set|clr] foo
+.El
+.Sh SIGNALS
+PID 1 is the only process that by default blocks all signals.  Finit
+allows for the following signals to control operation, but recommends
+instead using the
+.Xr initctl 8
+tool.
+.Pp
+.Bl -tag -width TERM
+.It HUP
+Tell
+.Nm
+to reload its configuration file(s), same as
+.Cm initctl reload
+.It USR1
+Restart API
+.Cm ( initctl )
+socket.
+.It USR2
+Tell
+.Nm
+to perform a system shutdown, same as
+.Cm initctl poweroff
+.It TERM
+Like SIGUSR2 but performs a reboot, same as
+.Cm initctl reboot
+.It INT
+Sent from kernel when Ctrl-Alt-Del is pressed (on the console), this
+asserts the sys/key/ctrlaltdel condition, which can be used to start
+tasks or services
+.It PWR
+Sent from a a power daemon on changes to UPS status, Finit asserts
+the sys/pwr/fail condition
+.El
+.Sh FILES
+.Bl -tag -width /etc/finit.d/available/*.conf -compact
+.It Pa /etc/finit.conf
+Main configuration file
+.It Pa /etc/finit.d/*.conf
+Static service definitions
+.It Pa /etc/finit.d/available/*.conf
+Available services
+.It Pa /etc/finit.d/enabled/*.conf
+Symlinks from available
+.It Pa /run/finit/
+Runtime files, including the condition subsystem
+.El
+.Sh SEE ALSO
+.Xr finit.conf 5
+.Xr initctl 8
+.Sh AUTHORS
+.Nm
+was conceived and reverse engineered by Claudio Matsuoka.  Since v1.0,
+maintained by Joachim Wiberg, with contributions by many others.
@@ -0,0 +1,814 @@
+.\"                                      Hey, EMACS: -*- nroff -*-
+.\" First parameter, NAME, should be all caps
+.\" Second parameter, SECTION, should be 1-8, maybe w/ subsection
+.\" other parameters are allowed: see man(7), man(1)
+.Dd Oct 17, 2020
+.Dt FINIT.CONF 5 SMM
+.Os Linux
+.Sh NAME
+.Nm finit.conf
+.Nd Finit configuration file format
+.Sh SYNOPSIS
+.Nm /etc/finit.conf
+.Nm /etc/finit.d/*.conf
+.Nm /etc/finit.d/available/*.conf
+.Sh DESCRIPTION
+.Nm finit
+can be configured using only the original
+.Pa /etc/finit.conf
+file or in combination with
+.Pa /etc/finit.d/*.conf .
+Useful for package-based Linux distributions \(em each package can
+provide its own "script" file.
+.Pp
+.Bl -tag -width /etc/finit.d/*.conf -offset indent -compact
+.It Pa /etc/finit.conf
+Main configuration file
+.It Pa /etc/finit.d/*.conf
+Snippets, usually one service per file
+.El
+.Pp
+Not all configuration directives are available in
+.Pa /etc/finit.d/*.conf
+and some directives are only available at bootstrap, runlevel
+.Cm S ,
+see the section
+.Sx Limitations
+below for details.
+.Sh FILES
+.Bl -tag -width /etc/finit.d/available/*.conf -compact
+.It Pa /etc/finit.conf
+Main configuration file
+.It Pa /etc/finit.d/*.conf
+Static service definitions
+.It Pa /etc/finit.d/available/*.conf
+Available services
+.It Pa /etc/finit.d/enabled/*.conf
+Enabled services, symlinks from available
+.El
+.Sh FILE FORMAT
+Lines starting with `#' are ignored. Note that end-of-line comments are
+NOT supported, comments must be on a line of their own.
+.Pp
+A configuration stanza must be on a single (long) line and
+.Em cannot
+be broken up in multiple lines by use of a backslash as the last
+character.
+.Sh DIRECTIVES
+This section lists all supported configuration directives.  There also exist
+deprecated directives, see the Markdown documentation for details on these.
+.Pp
+.Bl -tag -width 1n
+.It Cm rlimit Oo hard|soft Oc Ar RESOURCE Aq LIMIT | unlimited
+.Pp
+Set the hard or soft limit for a resource, or both if that argument is
+omitted.
+.Ar RESOURCE
+is the lower-case
+.Cm RLIMIT_
+string constants from
+.Xr setrlimit 2 ,
+without prefix.  E.g. to set
+.Cm RLIMIT_CPU ,
+use
+.Cm cpu .
+.Pp
+.Ar LIMIT
+is an integer that depends on the resource being modified, see
+the man page, or the kernel
+.Pa /proc/PID/limits
+file, for details.
+Finit versions before v3.1 used
+.Cm infinity
+for
+.Cm unlimited ,
+which is still supported, albeit deprecated.
+.Bd -unfilled -offset indent
+# No process is allowed more than 8MB of address space
+rlimit hard as 8388608
+
+# Core dumps may be arbitrarily large
+rlimit soft core infinity
+
+# CPU limit for all services, soft & hard = 10 sec
+rlimit cpu 10
+.Ed
+.Pp
+.Cm rlimit
+can be set globally, in
+.Pa /etc/finit.conf ,
+or locally per each
+.Pa /etc/finit.d/*.conf
+read.  I.e., a set of task/run/service stanzas can share the same
+rlimits if they are in the same .conf.
+.It Cm runlevel Aq N
+The system runlevel to go to after bootstrap (S) has completed.
+.Cm N
+is the runlevel number 0-9, where 6 is reserved for reboot and 0 for
+halt.  All other can be used by operating system administrators.
+Default: 2
+.Pp
+.Sy Note:
+only read and executed in runlevel S (bootstrap).
+.It Cm run Oo LVLS Oc Ao COND Ac Ar /path/to/cmd ARGS Op -- Optional description
+One-shot command to run in sequence when entering a runlevel, with
+optional arguments and description.
+.Pp  
+.Cm run
+commands are guaranteed to be completed before running the next
+command.  Highly useful if true serialization is needed.  Usually
+only used in the bootstrap (S) runlevel.
+.Pp
+.Cm Aq COND
+conditions are described in
+.Xr finit 8 ,
+see also the
+.Sx Examples
+section below.
+.It Cm task Oo LVLS Oc Ao COND Ac Ar /path/to/cmd ARGS Op -- Optional description
+One-shot like
+.Cm run ,
+but starts in parallel with the next command.
+.Pp  
+Both
+.Cm run
+and
+.Cm task
+commands are run in a shell, so pipes and redirects can be freely used:
+.Bd -unfilled -offset indent
+task [s] echo "foo" | cat >/tmp/bar
+.Ed
+.It Cm sysv Oo LVLS Oc Ao COND Ac Ar /path/to/script ARGS Op -- Optional description
+Similar to
+.Cm task
+is the
+.Cm sysv
+stanza, which can be used to call SysV style start/stop scripts.  The
+primary intention for this command is to be able to re-use much of
+existing setup and init scripts in Linux distributions.
+.Pp  
+When entering an allowed runlevel, Finit calls
+.Cm init-script start ,
+when entering a disallowed runlevel, Finit calls
+.Cm init-script stop ,
+and if the Finit .conf, where the
+.Cm sysv
+stanza is declared, is modified, Finit calls
+.Cm init-script restart
+on
+.Cm initctl reload .
+Similar to how
+.Cm service
+stanzas work.
+.Pp
+Forking services started with
+.Cm sysv
+scripts can be monitored by Finit by declaring the PID file to look for:
+.Bd -unfilled -offset indent
+sysv pid:!/path/to/pidfile.pid /path/to/script ...
+.Ed
+.Pp
+The leading '!' is to prevent Finit from managing the PID file, which is
+the default behavior for the
+.Cm pid:
+command modifier.
+.It Cm service Oo LVLS Oc Ao COND Ac Ar /path/to/daemon ARGS Op -- Optional description
+Service, or daemon, to be monitored and automatically restarted if it
+exits prematurely.  Finit tries to restart services that die 10 times
+before giving up, then you have to restart it manually:
+.Cm initctl restart NAME
+.Pp  
+For daemons that support it, we recommend appending
+.Cm --foreground , --no-background , -n , -F ,
+or similar command line argument to prevent them from forking off a
+sub-process in the background.  This is the most reliable way to monitor
+a service.
+.Pp
+However, not all daemons support running in the foreground, or they may
+start logging to the foreground as well, these are called forking
+services and are supported using the same syntax as forking
+.Cm sysv
+services, using the
+.Cm pid:!/path/to/pidfile.pid
+command modifier syntax.
+.Pp
+.Sy Example: in the case of
+.Cm ospfd
+(below), we omit the
+.Cm -d
+flag (daemonize) to prevent it from forking to the background:
+.Bd -unfilled -offset indent
+service [2345] <pid/zebra> /sbin/ospfd -- OSPF daemon
+.Ed
+.Cm [2345]
+denote the runlevels
+.Cm ospfd
+is allowed to run in, they are optional and default to runlevel 2-5 if
+omitted.
+.Pp
+.Cm <pid/zebra>
+is the condition for starting
+.Cm ospfd .
+In this example Finit waits for another service,
+.Cm zebra ,
+to have created its PID file in
+.Pa /var/run/quagga/zebra.pid
+before starting
+.Cm ospfd .
+Finit watches *all* files in
+.Pa /var/run ,
+for each file named
+.Cm *.pid ,
+.Cm */pid ,
+Finit opens it and find the matching
+.Cm NAME:ID
+using the PID.
+.Pp
+Some services do not maintain a PID file and rather than patching each
+application Finit provides a workaround.  A
+.Cm pid
+modifier keyword can be set to have Finit automatically create (when
+starting) and later remove (when stopping) the PID file.  The file is
+created in the
+.Pa /var/run
+directory using the
+.Xr basename 3
+of the service.  The full syntax of the
+.Cm pid
+modifier is:
+.Bd -unfilled -offset indent
+pid[:[!][/path/to/]filename[.pid]]
+.Ed
+.Pp
+For example, by adding
+.Cm pid:/run/foo.pid
+to the service
+.Cm /sbin/bar ,
+that PID file will, not only be created and removed automatically, but
+also be used by the Finit condition subsystem.  So a service/run/task
+can depend on the
+.Cm <pid/bar>
+condition.
+.Pp
+If a service should not be automatically started, it can be configured
+as manual with the
+.Cm manual:yes
+command modifier.  The service can then be started at any time by
+running
+.Cm initctl start NAME
+.Pp
+The name of a service, shown by the
+.Cm initctl
+tool, defaults to the basename of the service executable. It can be
+changed with the
+.Cm name:foo
+command modifier.
+.Pp
+When stopping a service (run/task/sysv/service), either manually or when
+moving to another runlevel, Finit starts by sending SIGTERM, to allow
+the process to shut down gracefully.  If the process has not been
+collected within 3 seconds, Finit sends SIGKILL.  To halt the process
+using a different signal, use the command modifier
+.Cm halt:SIGNAL ,
+e.g.,
+.Cm halt:SIGPWR .
+To change the delay between your halt signal and KILL, use the command
+modifier
+.Cm kill:SEC ,
+e.g.,
+.Cm kill:10
+to wait 10 seconds before sending SIGKILL.
+.Pp
+Services support the
+.Cm pre:script
+and
+.Cm post:script
+command actions as well.  These run as the same
+.Cm @USER:GROUP
+as the service itself, with any
+.Cm env:file
+sourced.  The scripts must use an absolute path, but are executed from
+the
+.Cm $HOME
+of the given user.  The scripts are not called with any argument
+(currently), but both get the
+.Cm SERVICE_IDENT=foo
+environment variable set.  Here
+.Cm foo
+denotes the identity of the service, which if there are multiple
+services named
+.Cm foo ,
+may be
+.Cm foo:1 ,
+or any unique identifier specified in the .conf file.  The
+.cm post:script
+is called with an additional set of environment variables:
+.Bl -tag -offset indent -width 1n
+.It Cm EXIT_CODE=[exited,signal]
+set to one of
+.Cm exited
+or
+.Cm signal
+.It Cm EXIT_STATUS=[num,SIGNAME]
+set to one of exit status code from the program, if it exited normally,
+or the signal name (HUP, TERM, etc.) if it exited due to signal
+.El
+.Pp
+The scripts have a default execution time of 3 seconds before they are
+SIGKILLed, this can be adjusted using the
+.Cm kill:SEC
+modifier syntax.
+.Pp
+.Sy Note:
+the
+.Cm pre:script
+.Em must
+be idempotent, because a service can transition between READY and HALTED
+states any number of times before going to RUNNING.
+.It Cm runparts Aq DIR
+Call
+.Xr run-parts 8
+on
+.Cm DIR
+to run start scripts.  All executable files, or scripts, in the
+directory are called, in alphabetic order.  The scripts in this
+directory are executed at the very end of bootstrap, runlevel S.
+.Pp
+It can be beneficial to use
+.Cm S01name ,
+.Cm S02othername ,
+etc. if there is a dependency order between the scripts.  Symlinks to
+existing daemons can talso be used, but make sure they daemonize by
+default.
+.Pp
+Similar to the
+.Pa /etc/rc.local
+shell script, make sure that all your services and programs either
+terminate or start in the background or you will block Finit.
+.Sy Note:
+only read and executed in runlevel S (bootstrap).
+.It Cm include Aq CONF
+Include another configuration file.  Absolute path required.
+.It Cm log size:BYTES count:NUM
+Log rotation for run/task/services using the
+.Cm log
+command modifier with redirection to a log file.  Global setting,
+applies to all services.
+.Pp
+The size can be given as bytes, without a specifier, or in `k`, `M`,
+or `G`, e.g.
+.Cm size:10M ,
+or
+.Cm size:3G .
+A value of
+.Cm size:0
+disables log rotation.  The default is
+.Cm size:200k .
+.Pp
+The count value is recommended to be between 1-5, with a default 5.
+Setting count to 0 means the logfile will be truncated when the MAX
+size limit is reached.
+.It Cm tty Oo LVLS Oc Ao COND Ac Ar DEV Oo BAUD Oc Oo noclear Oc Oo nowait Oc Oo nologin Oc Oo TERM Oc
+This form of the
+.Cm tty
+stanza uses the built-in getty on the given TTY device
+.Ar DEV ,
+in the given runlevels.
+.Ar DEV may be the special keyword
+.Cm @console ,
+or `console`, which is expanded from `/sys/class/tty/console/active`,
+useful on embedded systems.
+.Pp
+The default baud rate is 0, i.e., keep kernel default.
+.Pp
+The `tty` stanza inherits runlevel, condition (and other feature)
+parsing from the `service` stanza.  So TTYs can run in one or many
+runlevels and depend on any condition supported by Finit.  This is
+useful e.g. to depend on `<pid/elogind>` before starting a TTY.
+.Bd -unfilled -offset indent
+tty [12345] /dev/ttyAMA0 115200 noclear vt220
+.Ed
+.It Cm tty Oo LVLS Oc Ao COND Ac Ar CMD DEV Oo noclear Oc Oo nowait Oc
+This form of the
+.Cm tty
+stanza is for using an external getty, like agetty or the BusyBox getty.
+.Pp
+By default, these first two syntax variants
+.Em clear
+the TTY and
+.Em wait
+for the user to press enter before starting getty.
+.Bd -unfilled -offset indent
+tty [12345] /sbin/getty  -L 115200 /dev/ttyAMA0 vt100
+tty [12345] /sbin/agetty -L ttyAMA0 115200 vt100 nowait
+.Ed
+.Pp
+The
+.Cm noclear
+option disables clearing the TTY after each session.  Clearing the TTY
+when a user logs out is usually preferable.
+.Pp  
+The
+.Cm nowait
+option disables the
+.Cm Please press Enter to activate console
+message before actually starting the getty program.  On small and
+embedded systems running multiple unused getty wastes both memory and
+CPU cycles, so `wait` is the preferred default.
+.Pp
+The
+.Cm nologin
+option disables getty and
+.Pa /bin/login ,
+and gives the user a root (login) shell on the given TTY
+.Cm DEV
+immediately.  Needless to say, this is a rather insecure option, but can
+be very useful for developer builds, during board bringup, or similar.
+.Pp
+Notice the ordering, the
+.Cm TERM
+option to the built-in getty must be the last argument.
+.Pp
+Embedded systems may want to enable automatic `DEV` by supplying the
+special
+.Cm @console
+device.  This works regardless weather the system uses
+.Cm ttyS0 , ttyAMA0 ,  ttyMXC0 ,
+or anything else.  Finit figures it out by querying sysfs:
+.Pa /sys/class/tty/console/active .
+The speed can be omitted to keep the kernel default.
+.Pp
+Most systems get by fine by just using `console`, which will evaluate
+to
+.Pa /dev/console .
+If you have to use
+.Cm @console
+to get any output, you may have some issue with your kernel config.
+.Bd -unfilled -offset indent
+tty [12345] @console noclear vt220
+.Ed
+.Pp
+On really bare bones systems, or for board bringup, Finit can give you a
+shell prompt as soon as bootstrap is done, without opening any device
+node:
+.Bd -unfilled -offset indent
+tty [12345789] notty
+.Ed
+.Pp
+This should of course not be enabled on production systems.  Because it
+may give a user root access without having to log in.  However, for
+board bringup and system debugging it can come in handy.
+.Pp
+One can also use the
+.Cm service
+stanza to start a stand-alone shell:
+.Bd -unfilled -offset indent
+service [12345] /bin/sh -l
+.Ed
+.Pp
+.It Cm tty Oo LVLS Oc Ao COND Ac Oo notty Oc Oo rescue Oc
+The third
+.Cm tty
+form is for board bringup and the
+.Cm rescue
+boot mode.  No device node is required in this variant, the same output
+that the kernel uses is reused for stdio.  If the
+.Cm rescue
+option is omitted, a shell is started.  The flags
+.Cm nologin , noclear ,
+and
+.Cm nowait
+are implied.   If the
+.Cm rescue
+option is set the bundled
+.Pa /libexec/finit/sulogin
+is started to present a bare-bones root login prompt.  If the root
+(uid:0, gid:0) user does not have a password set, no rescue is possible.
+.El
+.Sh COMMAND MODIFIERS
+The run/task/tty/service/sysv stanzas take modifiers, or options, to
+control their behavior.  This section lists them with their limitations.
+All modifiers must be placed between the stanza and its command.
+.Bl -tag -width 1n
+.It Cm @user:group
+Every
+.Cm run , task ,
+or
+.Cm service
+can also list the privileges the
+.Cm /path/to/cmd
+should be executed with.  Prefix the command with
+.Cm @USR[:GRP] ,
+group is optional, like this:
+.Bd -unfilled -offset indent
+run [2345] @joe:users logger "Hello world"
+.Ed
+.Pp
+For multiple instances of the same command, e.g. a DHCP client or
+multiple web servers, add
+.Cm :ID
+somewhere between the
+.Cm run , task , service
+keyword and the command, like this:
+.Bd -unfilled -offset indent
+service :80  [2345] httpd -f -h /http -p 80   -- Web server
+service :8080[2345] httpd -f -h /http -p 8080 -- Old web server
+.Ed
+.Pp
+Without the
+.Cm :ID
+to the service the latter will overwrite the former and only the old web
+server would be started and supervised.
+.It Cm log:/path/to/file
+Redirect stdout/stderr of a command to the given log file.  See the
+global log directive, above, for details on log rotation.
+.It Cm log:console
+Redirect stdout/stderr of a command to
+.Pa /dev/console ,
+only use this for debugging or bringup.
+.It Cm log:null
+Redirect stdout/stderr of a command to
+.Pa /dev/null .
+.It Cm log:prio:facility.level,tag:ident
+Redirect stdout/stderr of a command to syslog using the given priority
+and tag identity.
+.Bd -unfilled -offset indent
+service log:prio:user.warn,tag:ntpd /sbin/ntpd pool.ntp.org -- NTP daemon
+.Ed
+.It Cm log
+Default
+.Cm prio
+is
+.Cm daemon.info
+and the default
+.Cm tag
+identity is the basename of the service or run/task command.
+.El
+.Sh RESCUE MODE
+Finit supports a rescue mode which is activated by the
+.Cm rescue
+option on the kernel command line.  The rescue mode comes in two
+flavors:
+.Em traditional
+and
+.Em fallback .
+.Ss Traditional
+This is what most users expect.  A very early maintenance login prompt,
+served by the bundled
+.Pa /libexec/finit/sulogin
+program, or the standard
+.Cm sulogin
+from util-linux or BusyBox is searched for in the UNIX default
+.Cm $PATH .
+If a successful login is made, or the user exits (Ctrl-D), the rescue
+mode is ended and the system boots up normally.
+.Pp
+.Sy Note:
+if the user (UID 0 and GID 0) does not have a password, or
+.Em the account is locked ,
+the user is presented with a password-less prompt:
+.Cm "Press enter to enter maintenance mode." ,
+which opens up a root shell.
+.Ss Fallback
+If no
+.Cm sulogin
+program is found, Finit tries to bring up as much of its own
+functionality as possible, yet limiting many aspects, meaning; no
+network, no`fsck` of file systems in
+.Pa /etc/fstab ,
+no
+.Pa /etc/rc.local ,
+no
+.Cm runparts ,
+and most plugins are skipped (except those that provide functionality
+for the condition subsystem).
+.Pp
+Instead of reading
+.Pa /etc/finit.conf
+et al, system configuration is read from
+.Pa /lib/finit/rescue.conf ,
+which can be freely modified by the system administrator.
+.Pp
+The bundled default `rescue.conf` contains nothing more than:
+.Bd -unfilled -offset indent
+runlevel 1
+tty [12345] rescue
+.Ed
+.Pp
+The
+.Cm tty
+has the
+.Cm rescue
+option set, which works similar to the board bring-up tty option
+.Cm notty .
+The major difference being that `sulogin` is started to query for
+root/admin password.  If
+.Cm sulogin
+is not found,
+.Cm rescue
+behaves like
+.Cm notty
+and gives a plain root shell prompt.
+.Pp
+If Finit cannot find
+.Pa /lib/finit/rescue.conf
+it defaults to:
+.Bd -unfilled -offset indent
+tty [12345] rescue
+.Ed
+.Pp
+There is no way to exit the
+.Em fallback
+rescue mode.
+.Sh SERVICE ENVIRONMENT
+Finit supports sourcing environment variables from
+.Pa /etc/default/* ,
+or similar.  This is a common pattern from SysV init scripts, where the
+start/stop script is a generic script for the given service,
+.Cm foo ,
+and the options for the service are sourced from the file
+.Pa /etc/default/foo .
+Like this:
+.Bd -unfilled -offset indent
+/etc/default/foo:
+
+    FOO_OPTIONS=--extra-arg="bar" -s -x
+
+/etc/finit.conf:
+
+    service [2345] env:-/etc/default/foo foo -n $FOO_OPTIONS -- Example foo daemon
+.Ed
+.Pp
+Here the service
+.Cm foo
+is started with
+.Op Fl -n ,
+to make sure it runs in the foreground, and the with the options found
+in the environment file.  With the
+.Cm ps
+command we can see that the process is started with:
+.Bd -unfilled -offset indent
+foo -n --extra-arg=bar -s -x
+.Ed
+.Pp
+.Sy Note:
+the leading `-` determines if Finit should treat a missing environment
+file as blocking the start of the service or not.  When `-` is used, a
+missing environment file does
+.Em not
+block the start.
+.Sh SERVICE WRAPPER SCRIPTS
+If your service requires to run additional commands, executed before the
+service is actually started, like the systemd `ExecStartPre`, you can
+use a wrapper shell script to start your service.
+.Pp
+The Finit service
+.Cm .conf
+file can be put into
+.Pa /etc/finit.d/available ,
+so you can control the service using
+.Cm initctl .
+Then use the path to the wrapper script in the Finit
+.Cm .conf
+service stanza.  The following example employs a wrapper script in
+.Pa /etc/start.d .
+.Bd -unfilled
+/etc/finit.d/available/program.conf:
+
+    service [235] <!> /etc/start.d/program -- Example Program
+
+/etc/start.d/program:
+
+    #!/bin/sh
+    # Prepare the command line options
+    OPTIONS="-u $(cat /etc/username)"
+
+    # Execute the program
+    exec /usr/bin/program $OPTIONS
+.Ed
+.Pp
+.Sy Note:
+the example sets
+.Cm <!>
+to denote that it doesn't support SIGHUP.  That way Finit will
+stop/start the service instead of sending SIGHUP at restart/reload
+events.
+.Sh CGROUPS
+There are three major cgroup configuration directives:
+.Pp
+.Bl -enum -offset indent -compact
+.It
+Global top-level group: init, system, user, or a custom group
+.It
+Selecting a top-level group for a set of run/task/services
+.It
+Per run/task/service limits
+.El
+.Pp
+Top-level group configuration.
+.Bd -unfilled -offset indent
+# Top-level cgroups and their default settings.  All groups mandatory
+# but more can be added, max 8 groups in total currently.  The cgroup
+# 'root' is also available, reserved for RT processes.  Settings are
+# as-is, only one shorthand 'mem.' exists, other than that it's the
+# cgroup v2 controller default names.
+cgroup init   cpu.weight:100
+cgroup user   cpu.weight:100
+cgroup system cpu.weight:9800
+.Ed
+.Pp
+Adding an extra cgroup
+.Cm maint/
+will require you to adjust the weight of the above three.  We leave
+.Cm init/
+and
+.Cm user/
+as-is reducing weight of
+.Cm system/
+to 9700.
+.Bd -unfilled -offset indent
+cgroup system cpu.weight:9700
+
+# Example extra cgroup 'maint'
+cgroup maint  cpu.weight:100
+.Ed
+.Pp
+By default, the
+.Cm system/
+cgroup is selected for almost everything.  The
+.Cm init/
+cgroup is reserved for PID 1 itself and its closest relatives.  The
+.Cm user/
+cgroup is for local TTY logins spawned by getty.
+.Pp
+To select a different top-level cgroup, e.g.
+.Cm maint/ ,
+one can either define it for a group of run/task/service directives in a
+.Cm .conf
+or per each stanza:
+.Bd -unfilled -offset indent
+cgroup.maint
+service [...] <...> /path/to/foo args -- description
+service [...] <...> /path/to/bar args -- description
+.Ed
+.Pp
+or
+.Bd -unfilled -offset indent
+service [...] <...> cgroup.maint /path/to/foo args -- description
+.Ed
+.Pp
+The latter form also allows per-stanza limits on the form:
+.Bd -unfilled -offset indent
+service [...] <...> cgroup.maint:cpu.max:10000,mem.max:655360 /path/to/foo args -- description
+.Ed
+.Pp
+Notice the comma separation and the
+.Cm mem.
+exception to the rule: every cgroup setting maps directly to cgroup v2
+syntax.  I.e.,
+.Cm cpu.max
+maps to the file
+.Pm /sys/fs/cgroup/maint/foo/cpu.max .
+There is no filtering, except for expanding the shorthand
+.Cm mem.
+to
+.Cm memory. ,
+if the file is not available, either the cgroup controller is not
+available in your Linux kernel, or the name is misspelled.
+.Pp
+Linux cgroups and details surrounding values are not explained in the
+Finit documentation.  The Linux admin-guide cover this well:
+.Lk https://www.kernel.org/doc/html/latest/admin-guide/cgroup-v2.html
+.Sh LIMITATIONS
+As of Finit v4 there are no limitations to where
+.Cm .conf
+settings can be placed.  Except for the system/global
+.Cm rlimit
+and
+.Cm cgroup
+top-level group declarations, which can only be set from
+.Pa /etc/finit.conf ,
+since it is the first
+.Cm .conf
+file Finit reads.
+.Pp
+Originally,
+.Pp /etc/finit.conf
+was the only way to set up a Finit system.  Today it is mainly used for
+bootstrap settings like system hostname, early module loading for
+watchdogd, network bringup and system shutdown.  These can now also be
+set in any
+.Cm .conf
+file in
+.Pa /etc/finit.d .
+.Pp
+There is, however, nothing preventing you from having all configuration
+settings in
+.Pa /etc/finit.conf .
+.Sh SEE ALSO
+.Xr finit 8 ,
+.Xr initctl 8
+.Sh AUTHORS
+.Nm finit
+was conceived and reverse engineered by Claudio Matsuoka.  Since v1.0,
+maintained by Joachim Wiberg, with contributions by many others.
@@ -0,0 +1,175 @@
+.\"                                      Hey, EMACS: -*- nroff -*-
+.\" First parameter, NAME, should be all caps
+.\" Second parameter, SECTION, should be 1-8, maybe w/ subsection
+.\" other parameters are allowed: see man(7), man(1)
+.Dd June 6, 2021
+.\" Please adjust this date whenever revising the manpage.
+.Dt INITCTL 8 SMM
+.Os Linux
+.Sh NAME
+.Nm initctl
+.Nd Control tool for Finit
+.Sh SYNOPSIS
+.Nm /sbin/initctl
+.Op Fl bcfhlpqtv
+.Op COMMAND
+.Sh DESCRIPTION
+.Nm
+is the official tool for interacting with
+.Xr finit 8 .
+It comes with functions for showing status, querying health of services,
+as well as editing, enabling and disabling services.
+.Sh OPTIONS
+This program follows the usual UNIX command line syntax, with long
+options starting with two dashes (`-').  The options are as follows:
+.Bl -tag -width Ds
+.It Fl b, -batch
+Batch mode, no screen size probing.
+.It Fl c, -create
+Create missing paths (and files) as needed.  Useful with the edit command.
+.It Fl f, -force
+Ignore missing files and arguments, never prompt.
+.It Fl h, -help
+Show built-in help text.
+.It Fl 1, -once
+Only one lap in commands like top.
+.It Fl p, -plain
+Use plain table headings, no ANSI control characters.
+.It Fl q, -quiet
+Silent, only return status of command.
+.It Fl t, -no-heading
+Skip table headings.
+.It Fl v, -verbose
+Verbose output, where applicable.
+.El
+.Sh COMMANDS
+.Bl -tag -width Ds
+.It Nm Ar debug
+Toggle finit (daemon) debug to
+.Pa /dev/console
+.It Nm Ar help
+Show built-in help text
+.It Nm Ar version
+Show Finit (daemon) version
+.It Nm Ar ls | list
+List all
+.Cm *.conf
+files in
+.Pa /etc/finit.d
+.It Nm Ar create Ar CONF
+Create
+.Cm .conf
+in
+.Pa /etc/finit.d/available
+.It Nm Ar delete Ar CONF
+Delete
+.Cm .conf
+in
+.Pa /etc/finit.d/available
+.It Nm Ar show Ar CONF
+Show
+.Cm .conf
+in
+.Pa /etc/finit.d/available
+.It Nm Ar edit Ar CONF
+Edit
+.Cm .conf
+in
+.Pa /etc/finit.d/available
+.It Nm Ar touch Ar CONF
+Change (mark as modified)
+.Cm .conf
+in
+.Pa /etc/finit.d/available ,
+like
+.Ar edit
+this tells
+.Nm finit
+that a configuration has been changed and is a candidate to be restarted
+(or SIGHUP'ed) on
+.Cm initctl reload
+.It Nm Ar enable Ar CONF
+Enable
+.Cm .conf
+in
+.Pa /etc/finit.d/available ,
+i.e., add symlink in
+.Pa /etc/finit.d/enabled
+.It Nm Ar disable Ar CONF
+Disable
+.Cm .conf
+in
+.Pa /etc/finit.d/enabled ,
+i.e., removes symlink
+.It Nm Ar reload
+Reload
+.Cm *.conf in
+.Pa /etc/finit.d ,
+i.e., activates changes
+.It Nm Ar cond set Ar COND
+Set (assert) user-defined condition,
+.Cm +usr/COND
+.It Nm Ar cond clr | clear Ar COND
+Clear (deassert) user-defined condition,
+.Cm -usr/COND
+.It Nm Ar cond status
+Show condition status, default
+.Cm cond
+command.
+.It Nm Ar cond dump
+Dump all conditions and their status
+.It Nm Ar log Op Cm NAME
+Show ten last Finit, or
+.Cm NAME ,
+messages from syslog
+.It Nm Ar start Cm NAME[:ID]
+Start service by name, with optional ID, e.g.,
+.Cm initctl start tty:1
+.It Nm Ar stop Cm NAME[:ID]
+Stop/Pause a running service by name
+.It Nm Ar reload Cm NAME[:ID]
+Reload service by name (SIGHUP or restart)
+.It Nm Ar restart Cm NAME[:ID]
+Restart (stop/start) service by name
+.It Nm Ar status Cm NAME[:ID]
+Show service status, by name
+.It Nm Ar status
+Show status of all services, default command
+.It Nm Ar cgroup
+List cgroup config overview
+.It Nm Ar ps
+List processes based on cgroups
+.It Nm Ar top
+Show top-like listing based on cgroups
+.It Nm Ar runlevel Op Ar 0-9
+Show or set runlevel: 0 halt, 6 reboot
+.It Nm Ar reboot
+Reboot system, default if
+.Cm reboot
+is symlinked to
+.Nm
+.It Nm Ar halt
+Halt system, default if
+.Cm halt
+is symlinked to
+.Nm
+.It Nm Ar poweroff
+Power-off system, default if
+.Cm poweroff
+is symlinked to
+.Nm
+.It Nm Ar suspend
+Suspend system, default if
+.Cm suspend
+is symlinked to
+.Nm
+.It Nm Ar utmp show
+Raw dump of UTMP/WTMP db
+.El
+.Sh SEE ALSO
+.Xr finit.conf 5
+.Xr finit 8
+.Sh AUTHORS
+.Nm finit
+was conceived and reverse engineered by Claudio Matsuoka.  Since v1.0,
+maintained by Joachim Wiberg, with contributions by many others.
@@ -7,7 +7,7 @@ AM_CPPFLAGS        += $(lite_CFLAGS)

 if STATIC
 noinst_LTLIBRARIES  = libplug.la
-libplug_la_SOURCES  = bootmisc.c modprobe.c rtc.c pidfile.c procps.c tty.c urandom.c
+libplug_la_SOURCES  = bootmisc.c modprobe.c pidfile.c procps.c sys.c tty.c usr.c

 if BUILD_ALSA_UTILS_PLUGIN
 libplug_la_SOURCES += alsa-utils.c
@@ -33,8 +33,20 @@ if BUILD_RESOLVCONF_PLUGIN
 libplug_la_SOURCES += resolvconf.c
 endif

+if BUILD_RTC_PLUGIN
+libplug_la_SOURCES += rtc.c
+endif
+
+if BUILD_URANDOM_PLUGIN
+libplug_la_SOURCES += urandom.c
+endif
+
+if BUILD_X11_COMMON_PLUGIN
+libplug_la_SOURCES += x11-common.c
+endif
+
 else
-pkglib_LTLIBRARIES  = bootmisc.la modprobe.la rtc.la pidfile.la procps.la tty.la urandom.la
+pkglib_LTLIBRARIES  = bootmisc.la modprobe.la pidfile.la procps.la sys.la tty.la usr.la

 if BUILD_ALSA_UTILS_PLUGIN
 pkglib_LTLIBRARIES += alsa-utils.la
@@ -60,6 +72,14 @@ if BUILD_RESOLVCONF_PLUGIN
 pkglib_LTLIBRARIES += resolvconf.la
 endif

+if BUILD_RTC_PLUGIN
+pkglib_LTLIBRARIES += rtc.la
+endif
+
+if BUILD_URANDOM_PLUGIN
+pkglib_LTLIBRARIES += urandom.la
+endif
+
 if BUILD_X11_COMMON_PLUGIN
 pkglib_LTLIBRARIES += x11-common.la
 endif
@@ -30,6 +30,11 @@

 static void save(void *arg)
 {
+	if (rescue) {
+		_d("Skipping %s plugin in rescue mode.", __FILE__);
+		return;
+	}
+
 	if (whichp(ALSACTL)) {
 		_d("Saving sound settings ...");
 		run_interactive(ALSACTL " -g store", "Saving sound settings");
@@ -38,6 +43,11 @@ static void save(void *arg)

 static void restore(void *arg)
 {
+	if (rescue) {
+		_d("Skipping %s plugin in rescue mode.", __FILE__);
+		return;
+	}
+
 	if (whichp(ALSACTL)) {
 		_d("Restoring sound settings ...");
 		run_interactive(ALSACTL " -g restore", "Restoring sound settings");
@@ -65,7 +65,7 @@ static int is_tmpfs(char *path)
 	return tmpfs;
 }

-static int do_clean(const char *fpath, const struct stat *sb, int tflag, struct FTW *ftw)
+static int bootclean(const char *fpath, const struct stat *sb, int tflag, struct FTW *ftw)
 {
 	if (ftw->level == 0)
 		return 1;
@@ -76,8 +76,14 @@ static int do_clean(const char *fpath, const struct stat *sb, int tflag, struct
 	return 0;
 }

-/* We can safely skip tmpfs, nothing to clean from previous boot there */
-static void bootclean(void)
+/*
+ * Cleanup stale files from previous boot, if any still linger on.
+ * Some systems, e.g. Alpine Linux, still have a persistent /run and
+ * /tmp, i.e. not tmpfs.
+ *
+ * We can safely skip tmpfs, nothing to clean there.
+ */
+static void clean(void *arg)
 {
 	char *dir[] = {
 		"/tmp/",
@@ -90,7 +96,7 @@ static void bootclean(void)
 		if (is_tmpfs(dir[i]))
 			continue;

-		nftw(dir[i], do_clean, 20, FTW_DEPTH);
+		nftw(dir[i], bootclean, 20, FTW_DEPTH);
 	}
 }

@@ -99,13 +105,11 @@ static void bootclean(void)
 */
 static void setup(void *arg)
 {
-	int uid, gid;
+	mode_t prev;

-	/* Cleanup stale files, if any still linger on. */
-	bootclean();
+	prev = umask(0);

 	_d("Setting up FHS structure in /var ...");
-	umask(0);
 	makedir("/var/cache",      0755);
 	makedir("/var/db",         0755); /* _PATH_VARDB on some systems */
 	makedir("/var/games",      0755);
@@ -115,15 +119,16 @@ static void setup(void *arg)
 	makedir("/var/lib/urandom",0755);
 	if (fisdir("/run")) {
 		_d("System with new /run tmpfs ...");
-		makedir("/run/lock",       1777);
+		if (!fisdir("/run/lock"))
+			makedir("/run/lock", 1777);
 		symlink("/run/lock", "/var/lock");
-		symlink("/dev/shm",  "/run/shm");
+		symlink("/dev/shm", "/run/shm");

 		/* compat only, should really be set up by OS/dist */
-		symlink("/run",      "/var/run");
+		symlink("/run", "/var/run");
 	} else {
-		makedir("/var/lock",       1777);
-		makedir("/var/run",        0755);
+		makedir("/var/lock", 1777);
+		makedir("/var/run", 0755);
 	}
 	makedir("/var/log",        0755);
 	makedir("/var/mail",       0755);
@@ -133,19 +138,21 @@ static void setup(void *arg)
 	makedir("/var/tmp",        0755);
 	makedir("/var/empty",      0755);

-	/*
-	 * If /etc/group or "utmp" group is missing,
-	 * default to "root"/"wheel" group
-	 */
-	_d("Setting up necessary UTMP files ...");
-	gid = getgroup("utmp");
-	if (gid < 0)
-		gid = 0;
-
 	/*
 	 * UTMP actually needs multiple db files
 	 */
 	if (has_utmp()) {
+		int gid;
+
+		/*
+		 * If /etc/group or "utmp" group is missing, default to
+		 * "root", or "wheel", group.
+		 */
+		_d("Setting up necessary UTMP files ...");
+		gid = getgroup("utmp");
+		if (gid < 0)
+			gid = 0;
+
 		create("/var/run/utmp",    0644, 0, gid); /* Currently logged in */
 		create("/var/log/wtmp",    0644, 0, gid); /* Login history       */
 		create("/var/log/btmp",    0600, 0, gid); /* Failed logins       */
@@ -165,39 +172,31 @@ static void setup(void *arg)
 	makedir("/var/run/network",0755); /* Needed by Debian/Ubuntu ifupdown */
 	makedir("/var/run/lldpd",  0755); /* Needed by lldpd */
 	makedir("/var/run/pluto",  0755); /* Needed by Openswan */
-	makedir("/var/run/quagga", 0755); /* Needed by Quagga */
-	makedir("/var/log/quagga", 0755); /* Needed by Quagga */
-	makedir("/var/run/frr",    0755); /* Needed by FRR */
-	makedir("/var/log/frr",    0755); /* Needed by FRR */
+	mksubsys("/var/run/quagga", 0755, "quagga", "quagga");
+	mksubsys("/var/log/quagga", 0755, "quagga", "quagga");
+	mksubsys("/var/run/frr",    0755, "frr", "frr");
+	mksubsys("/var/log/frr",    0755, "frr", "frr");
 	makedir("/var/run/sshd",  01755); /* OpenSSH  */

-	gid = getgroup("tty");
-	if (gid < 0)
-		gid = 0;
-	makefifo("/dev/xconsole",  0640); /* sysklogd */
-	chown("/dev/xconsole", 0, gid);
-
 	if (!fexist("/etc/mtab"))
 		symlink("../proc/self/mounts", "/etc/mtab");

 	/* Void Linux has a uuidd that runs as uuid:uuid and needs /run/uuid */
-	uid = getuser("uuidd", NULL);
-	if (uid >= 0) {
-		gid = getgroup("uuidd");
-		if (gid < 0)
-			gid = 0;
-		makedir("/var/run/uuidd", 0755);
-		chown("/var/run/uuidd", uid, gid);
-	}
+	mksubsys("/var/run/uuidd", 0755, "uuidd", "uuidd");

-	umask(022);
+	/* Debian has /run/sudo, ensure correct perms and SELinux label */
+	mksubsys("/var/run/sudo", 0711, "root", "root");
+	mksubsys("/var/run/sudo/ts", 0700, "root", "root");
+	if (whichp("restorecon"))
+		run("restorecon /var/run/sudo /var/run/sudo/ts");
+
+	umask(prev);
 }

 static plugin_t plugin = {
 	.name = __FILE__,
-	.hook[HOOK_BASEFS_UP] = {
-		.cb  = setup
-	},
+	.hook[HOOK_MOUNT_POST] = { .cb = clean },
+	.hook[HOOK_BASEFS_UP]  = { .cb = setup },
 	.depends = { "pidfile" },
 };

@@ -32,13 +32,21 @@
 #include "conf.h"

 #define DAEMON "dbus-daemon"
-#define ARGS   "--nofork --system"
+#define ARGS   "--nofork --system --syslog-only"
 #define DESC   "D-Bus message bus daemon"

+#define DAEMONUSER "messagebus"
+
 static void setup(void *arg)
 {
-	char *cmd;
 	char line[256];
+	mode_t prev;
+	char *cmd;
+
+	if (rescue) {
+		_d("Skipping %s plugin in rescue mode.", __FILE__);
+		return;
+	}

 	cmd = which(DAEMON);
 	if (!cmd) {
@@ -46,29 +54,28 @@ static void setup(void *arg)
 		return;
 	}

-	umask(0);
+	prev =umask(0);

-	makedir("/var/run/dbus", 0755);
-	makedir("/var/lock/subsys", 0755);
-	makedir("/var/lock/subsys/messagebus", 0755);
+	mksubsys("/var/run/dbus", 0755, DAEMONUSER, DAEMONUSER);
 	if (whichp("dbus-uuidgen"))
-		run("dbus-uuidgen --ensure");
+		run_interactive("dbus-uuidgen --ensure", "Creating machine UUID for D-Bus");

 	/* Clean up from any previous pre-bootstrap run */
-	erase("/var/run/dbus/pid");
+	remove("/var/run/dbus/pid");

 	/* Register service with Finit */
-	snprintf(line, sizeof(line), "[S12345] pid:!/run/dbus/pid log:prio:daemon.info,tag:dbus %s %s -- %s", cmd, ARGS, DESC);
+	snprintf(line, sizeof(line), "[S12345789] cgroup.system @%s:%s %s %s -- %s",
+		 DAEMONUSER, DAEMONUSER, cmd, ARGS, DESC);
 	if (service_register(SVC_TYPE_SERVICE, line, global_rlimit, NULL))
 		_pe("Failed registering %s", DAEMON);
 	free(cmd);

-	umask(022);
+	umask(prev);
 }

 static plugin_t plugin = {
 	.name = __FILE__,
-	.hook[HOOK_NETWORK_UP] = {
+	.hook[HOOK_BASEFS_UP] = {
 		.cb  = setup
 	},
 	.depends = { "bootmisc", },
@@ -34,6 +34,11 @@ static void setup(void *arg)
 	char cmd[256];
 	char *path;

+	if (rescue) {
+		_d("Skipping %s plugin in rescue mode.", __FILE__);
+		return;
+	}
+
 	/*
 	 * Populate /dev and prepare for runtime events from kernel.
 	 * Prefer udev if mdev is also available on the system.
@@ -43,17 +48,17 @@ static void setup(void *arg)
 		path = which("/lib/systemd/systemd-udevd");
 	if (path) {
 		/* Register udevd as a monitored service */
-		snprintf(cmd, sizeof(cmd), "[S12345789] pid:udevd name:udevd log %s "
+		snprintf(cmd, sizeof(cmd), "[S12345789] cgroup.system pid:udevd name:udevd log %s "
 			 "-- Device event managing daemon", path);
 		if (service_register(SVC_TYPE_SERVICE, cmd, global_rlimit, NULL)) {
 			_pe("Failed registering %s", path);
 		} else {
-			snprintf(cmd, sizeof(cmd), ":1 [S] <pid/udevd> log "
+			snprintf(cmd, sizeof(cmd), "cgroup.init :1 [S] <pid/udevd> log "
 				 "udevadm trigger -c add -t devices "
 				 "-- Requesting device events");
 			service_register(SVC_TYPE_RUN, cmd, global_rlimit, NULL);

-			snprintf(cmd, sizeof(cmd), ":2 [S] <pid/udevd> log "
+			snprintf(cmd, sizeof(cmd), "cgroup.init :2 [S] <pid/udevd> log "
 				 "udevadm trigger -c add -t subsystems "
 				 "-- Requesting subsystem events");
 			service_register(SVC_TYPE_RUN, cmd, global_rlimit, NULL);
@@ -90,6 +95,7 @@ static plugin_t plugin = {
 	.hook[HOOK_BASEFS_UP] = {
 		.cb  = setup
 	},
+	.depends = { "bootmisc" },
 };

 PLUGIN_INIT(plugin_init)
@@ -36,7 +36,7 @@
 #define MODULES_LOAD_PATH "/etc/modules-load.d"
 #endif
 #define SERVICE_LINE \
-	":%d name:modprobe.%s [2345] /sbin/modprobe %s %s -- Kernel module: %s"
+	"cgroup.init :%d name:modprobe.%s [2345] /sbin/modprobe %s %s -- Kernel module: %s"

 static void load(void *arg)
 {
@@ -71,8 +71,7 @@ static void load(void *arg)

 			mod = strtok_r(mod, " ", &args);

-			snprintf(cmd, sizeof(cmd), SERVICE_LINE,
-					 index, mod, mod, args, mod);
+			snprintf(cmd, sizeof(cmd), SERVICE_LINE, index, mod, mod, args, mod);
 			service_register(SVC_TYPE_TASK, cmd, global_rlimit, NULL);

 			index++;
@@ -30,31 +30,35 @@
 #include "cond.h"
 #include "helpers.h"
 #include "plugin.h"
+#include "service.h"

-static int nlmsg_validate(struct nlmsghdr *nh, size_t len)
-{
-	if (!NLMSG_OK(nh, len))
-		return 1;
+#define  NL_BUFSZ	4096

-	if (nh->nlmsg_type == NLMSG_DONE) {
-		_d("Done with netlink messages.");
-		return 1;
-	}
+struct nl_request {
+	struct nlmsghdr nh;
+	union {
+		struct rtmsg     rtm;
+		struct ifinfomsg ifi;
+	};
+};

-	if (nh->nlmsg_type == NLMSG_ERROR) {
-		_d("Netlink reports error.");
-		return 1;
-	}
+static int   nl_defidx;
+static int   nl_ifdown;
+static char *nl_buf;

-	return 0;
-}

 static void nl_route(struct nlmsghdr *nlmsg, ssize_t len)
 {
+	char daddr[INET_ADDRSTRLEN];
+	char gaddr[INET_ADDRSTRLEN];
+	struct in_addr ind, ing;
 	struct rtmsg *r;
 	struct rtattr *a;
+	int plen = 0;
+	int dst = 0;
+	int idx = 0;
+	int gw = 0;
 	int la;
-	int gw = 0, dst = 0, mask = 0, idx = 0;

 	if (nlmsg->nlmsg_len < NLMSG_LENGTH(sizeof(struct rtmsg))) {
 		_e("Packet too small or truncated!");
@@ -80,8 +84,8 @@ static void nl_route(struct nlmsghdr *nlmsg, ssize_t len)

 		case RTA_DST:
 			dst = *((int *)data);
-			mask = r->rtm_dst_len;
-			//_d("MASK: 0x%04x", mask);
+			plen = r->rtm_dst_len;
+			//_d("Prefix LEN: 0x%04x", plen);
 			break;

 		case RTA_OIF:
@@ -93,11 +97,20 @@ static void nl_route(struct nlmsghdr *nlmsg, ssize_t len)
 		a = RTA_NEXT(a, la);
 	}

-	if ((!dst && !mask) && (gw || idx)) {
-		if (nlmsg->nlmsg_type == RTM_DELROUTE)
+	ind.s_addr = dst;
+	ing.s_addr = gw;
+	inet_ntop(AF_INET, &ind, daddr, sizeof(daddr));
+	inet_ntop(AF_INET, &ing, gaddr, sizeof(gaddr));
+	_d("Got gw %s dst/len %s/%d ifindex %d", gaddr, daddr, plen, idx);
+
+	if ((!dst && !plen) && (gw || idx)) {
+		if (nlmsg->nlmsg_type == RTM_DELROUTE) {
 			cond_clear("net/route/default");
-		else
+			nl_defidx = 0;
+		} else {
 			cond_set("net/route/default");
+			nl_defidx = idx;
+		}
 	}
 }

@@ -117,14 +130,14 @@ static int validate_ifname(const char *ifname)
 	if (!ifname || !ifname[0])
 		return 1;

-	if (strlen(ifname) >= IFNAMSIZ)
+	if (strnlen(ifname, IFNAMSIZ) == IFNAMSIZ)
 		return 1;

-	if (strstr(ifname, ".."))
+	if (!strcmp(ifname, ".") || !strcmp(ifname, ".."))
 		return 1;

 	while (*ifname) {
-		if (*ifname == '/' || isspace(*ifname))
+		if (*ifname == '/' || *ifname == ':' || isspace(*ifname))
 			return 1;
 		ifname++;
 	}
@@ -132,12 +145,25 @@ static int validate_ifname(const char *ifname)
 	return 0;
 }

+/*
+ * Check if this interface was associated with the default route
+ * previously, or if it's been removed.  If so, trigger a recheck
+ * of the system default route.
+ */
+static void nl_check_default(char *ifname)
+{
+	int idx = (int)if_nametoindex(ifname);
+
+	if ((nl_defidx > 0 && nl_defidx == idx) || (idx == 0 && errno == ENODEV))
+		nl_ifdown = 1;
+}
+
 static void nl_link(struct nlmsghdr *nlmsg, ssize_t len)
 {
-	int la;
 	char ifname[IFNAMSIZ + 1];
-	struct rtattr *a;
 	struct ifinfomsg *i;
+	struct rtattr *a;
+	int la;

 	if (nlmsg->nlmsg_len < NLMSG_LENGTH(sizeof(struct ifinfomsg))) {
 		_e("Packet too small or truncated!");
@@ -172,6 +198,8 @@ static void nl_link(struct nlmsghdr *nlmsg, ssize_t len)
 			net_cond_set(ifname, "exist",   1);
 			net_cond_set(ifname, "up",      i->ifi_flags & IFF_UP);
 			net_cond_set(ifname, "running", i->ifi_flags & IFF_RUNNING);
+			if (!(i->ifi_flags & IFF_UP) || !(i->ifi_flags & IFF_RUNNING))
+				nl_check_default(ifname);
 			break;

 		case RTM_DELLINK:
@@ -180,6 +208,7 @@ static void nl_link(struct nlmsghdr *nlmsg, ssize_t len)
 			net_cond_set(ifname, "exist",   0);
 			net_cond_set(ifname, "up",      0);
 			net_cond_set(ifname, "running", 0);
+			nl_check_default(ifname);
 			break;

 		case RTM_NEWADDR:
@@ -197,26 +226,176 @@ static void nl_link(struct nlmsghdr *nlmsg, ssize_t len)
 	}
 }

-static void nl_callback(void *arg, int sd, int events)
+static int nl_parse(int sd)
 {
-	static char buf[4096];
-	struct nlmsghdr *nh;
-	ssize_t len;
+	while (1) {
+		struct nlmsghdr *nh;
+		ssize_t len;
+		size_t l;

-	memset(buf, 0, sizeof(buf));
-	len = recv(sd, buf, sizeof(buf), 0);
-	if (len < 0) {
-		if (errno != EINTR)	/* Signal */
-			_pe("recv()");
+		while ((len = recv(sd, nl_buf, NL_BUFSZ, 0)) < 0) {
+			switch (errno) {
+			case EAGAIN:	/* Nothing more right now. */
+				return 0;
+
+			case EINTR:	/* Signal */
+				continue;
+
+			case ENOBUFS:	/* netlink(7) */
+				break;
+
+			default:
+				_pe("recv()");
+				break;
+			}
+
+			return -1;
+		}
+
+		_d("recv %lld bytes", len);
+		l = (size_t)len;
+		for (nh = (struct nlmsghdr *)nl_buf; NLMSG_OK(nh, l); nh = NLMSG_NEXT(nh, l)) {
+			struct nlmsgerr *nle;
+
+			switch (nh->nlmsg_type) {
+			case NLMSG_DONE:
+				_d("Done with netlink messages.");
+				return 0;
+
+			case NLMSG_ERROR:
+				_d("Kernel netlink comm. error.");
+				nle = NLMSG_DATA(nh);
+				if (nle) {
+					errno = -nle->error;
+					_pe("Kernel netlink error %d", errno);
+				}
+				return -1;
+
+			case RTM_NEWROUTE:
+			case RTM_DELROUTE:
+				_d("Netlink route ...");
+				nl_route(nh, len);
+				break;
+
+			case RTM_NEWLINK:
+			case RTM_DELLINK:
+				_d("Netlink link ...");
+				nl_link(nh, len);
+				break;
+
+			default:
+				_w("unhandled netlink message, type %d", nh->nlmsg_type);
+				break;
+			}
+		}
+	}
+}
+
+static int nl_request(int sd, unsigned int seq, int type)
+{
+	struct nl_request *nlr = (struct nl_request *)nl_buf;
+
+	memset(nlr, 0, sizeof(struct nl_request));
+	nlr->nh.nlmsg_type  = type;
+	nlr->nh.nlmsg_flags = NLM_F_DUMP | NLM_F_REQUEST;
+	nlr->nh.nlmsg_seq   = seq;
+	nlr->nh.nlmsg_pid   = 1;
+
+	switch (type) {
+	case RTM_GETROUTE:
+		_d("RTM_GETROUTE");
+		nlr->rtm.rtm_family = AF_INET;
+		nlr->rtm.rtm_table  = RT_TABLE_MAIN;
+		nlr->nh.nlmsg_len   = NLMSG_LENGTH(sizeof(struct rtmsg));
+		break;
+
+	case RTM_GETLINK:
+		_d("RTM_GETLINK");
+		nlr->ifi.ifi_family = AF_UNSPEC;
+		nlr->ifi.ifi_change = 0xFFFFFFFF;
+		nlr->nh.nlmsg_len   = NLMSG_LENGTH(sizeof(struct ifinfomsg));
+		break;
+
+	default:
+		_w("Cannot resync, unhandled message type %d", type);
+		return -1;
+	}
+
+	if (send(sd, nlr, nlr->nh.nlmsg_len, 0) < 0)
+		return 1;
+
+	return nl_parse(sd);
+}
+
+static void nl_resync_routes(int sd, unsigned int seq)
+{
+	if (nl_request(sd, seq, RTM_GETROUTE))
+		_pe("Failed netlink route request");
+}
+
+static void nl_resync_ifaces(int sd, unsigned int seq)
+{
+	if (nl_request(sd, seq, RTM_GETLINK))
+		_pe("Failed netlink link request");
+}
+
+/*
+ * We've potentially lost netlink events, let's resync with kernel.
+ */
+static void nl_resync(int all)
+{
+	unsigned int seq = 0;
+	int sd;
+
+	sd = socket(AF_NETLINK, SOCK_DGRAM, NETLINK_ROUTE);
+	if (sd < 0) {
+		_pe("netlink socket");
 		return;
 	}

-	for (nh = (struct nlmsghdr *)buf; !nlmsg_validate(nh, len); nh = NLMSG_NEXT(nh, len)) {
-		//_d("Well formed netlink message received. type %d ...", nh->nlmsg_type);
-		if (nh->nlmsg_type == RTM_NEWROUTE || nh->nlmsg_type == RTM_DELROUTE)
-			nl_route(nh, len);
-		else
-			nl_link(nh, len);
+	if (all) {
+		_d("============================ RESYNC =================================");
+		/* this doesn't update condtions, and thus does not stop services */
+		cond_deassert("net/");
+
+		nl_resync_ifaces(sd, seq++);
+		nl_resync_routes(sd, seq++);
+
+		/* delayed update after we've corrected things */
+		service_step_all(SVC_TYPE_ANY);
+		_d("=========================== RESYNCED ================================");
+	} else
+		nl_resync_routes(sd, seq++);
+
+	close(sd);
+}
+
+static void nl_callback(void *arg, int sd, int events)
+{
+	if (nl_parse(sd) < 0) {
+		if (errno == ENOBUFS) {	/* netlink(7) */
+			_w("busy system, resynchronizing with kernel.");
+			nl_resync(1);
+			return;
+		}
+	}
+
+	/*
+	 * Linux doesn't send route changes when interfaces go down, so
+	 * we need to check ourselves, e.g. for loss of default route.
+	 */
+	if (nl_ifdown) {
+		_d("interface down, checking default route.");
+		if (nl_defidx > 0) {
+			nl_defidx = 0;
+			nl_resync(0);
+			if (nl_defidx <= 0) {
+				cond_clear("net/route/default");
+				nl_defidx = 0;
+			}
+		}
+
+		nl_ifdown = 0;
 	}
 }

@@ -236,8 +415,8 @@ static plugin_t plugin = {

 PLUGIN_INIT(plugin_init)
 {
-	int sd;
 	struct sockaddr_nl sa;
+	int sd;

 	sd = socket(AF_NETLINK, SOCK_RAW | SOCK_NONBLOCK | SOCK_CLOEXEC, NETLINK_ROUTE);
 	if (sd < 0) {
@@ -247,7 +426,7 @@ PLUGIN_INIT(plugin_init)

 	memset(&sa, 0, sizeof(sa));
 	sa.nl_family = AF_NETLINK;
-	sa.nl_groups = RTMGRP_IPV4_ROUTE | RTMGRP_LINK; // | RTMGRP_NOTIFY | RTMGRP_IPV4_IFADDR;
+	sa.nl_groups = RTMGRP_IPV4_ROUTE | RTMGRP_LINK;
 	sa.nl_pid    = getpid();

 	if (bind(sd, (struct sockaddr *)&sa, sizeof(sa)) < 0) {
@@ -256,6 +435,13 @@ PLUGIN_INIT(plugin_init)
 		return;
 	}

+	nl_buf = malloc(NL_BUFSZ);
+	if (!nl_buf) {
+		_pe("malloc()");
+		close(sd);
+		return;
+	}
+
 	plugin.io.fd = sd;
 	plugin_register(&plugin);
 }
@@ -58,7 +58,7 @@ static int pidfile_add_path(struct iwatch *iw, char *path)
 		}
 	}

-	return iwatch_add(iw, path, IN_ONLYDIR);
+	return iwatch_add(iw, path, IN_ONLYDIR | IN_CLOSE_WRITE);
 }

 static void pidfile_update_conds(char *dir, char *name, uint32_t mask)
@@ -68,11 +68,11 @@ static void pidfile_update_conds(char *dir, char *name, uint32_t mask)
 	svc_t *svc;

 	paste(fn, sizeof(fn), dir, name);
-	_d("path: %s, mask: %08x", fn, mask);
-
 	if (fnmatch("*\\.pid", fn, 0) && fnmatch("*/pid", fn, 0))
 		return;

+	_d("path: %s, mask: %08x", fn, mask);
+
 	svc = svc_find_by_pidfile(fn);
 	if (!svc) {
 		_d("No matching svc for %s", fn);
@@ -82,15 +82,25 @@ static void pidfile_update_conds(char *dir, char *name, uint32_t mask)
 	_d("Found svc %s for %s with pid %d", svc->name, fn, svc->pid);

 	mkcond(svc, cond, sizeof(cond));
-	if (mask & (IN_CREATE | IN_ATTRIB | IN_MODIFY | IN_MOVED_TO)) {
+	if (mask & (IN_CLOSE_WRITE | IN_ATTRIB | IN_MODIFY | IN_MOVED_TO)) {
 		svc_started(svc);
+		if (!svc_has_pidfile(svc))
+			pid_file_set(svc, fn, 1);
+
 		if (svc_is_forking(svc)) {
 			pid_t pid;

 			pid = pid_file_read(pid_file(svc));
-			_d("Forking service %s changed PID from %d to %d",
-			   svc->cmd, svc->pid, pid);
-			svc->pid = pid;
+			if (pid == -1) {
+				cond_clear(cond);
+				return;
+			}
+
+			if (pid != svc->pid) {
+				_d("Forking service %s changed PID from %d to %d",
+				   svc->cmd, svc->pid, pid);
+				svc->pid = pid;
+			}
 		}

 		cond_set(cond);
@@ -174,8 +184,6 @@ static void pidfile_callback(void *arg, int fd, int events)
 		if (!ev->mask)
 			continue;

-		_d("name %s, event: 0x%08x", ev->name, ev->mask);
-
 		/* Find base path for this event */
 		iwp = iwatch_find_by_wd(&iw_pidfile, ev->wd);
 		if (!iwp)
@@ -186,12 +194,7 @@ static void pidfile_callback(void *arg, int fd, int events)
 			continue;
 		}

-		if (ev->mask & IN_DELETE) {
-			_d("pidfile %s/%s removed ...", iwp->path, ev->name);
-			continue;
-		}
-
-		if (ev->mask & (IN_CREATE | IN_ATTRIB | IN_MODIFY | IN_MOVED_TO))
+		if (ev->mask & (IN_CLOSE_WRITE | IN_DELETE | IN_ATTRIB | IN_MODIFY | IN_MOVED_TO))
 			pidfile_update_conds(iwp->path, ev->name, ev->mask);
 	}
 }
@@ -233,8 +236,14 @@ static void pidfile_reconf(void *arg)

 static void pidfile_init(void *arg)
 {
+	char piddir[MAX_ARG_LEN];
 	char *path;

+	if (mkpath(pid_runpath(_PATH_CONDPID, piddir, sizeof(piddir)), 0755) && errno != EEXIST) {
+		_pe("Failed creating %s condition directory, %s", COND_PID, _PATH_CONDPID);
+		return;
+	}
+
 	/*
 	 * The bootmisc plugin is responsible for setting up /var/run.
 	 * and/or /run, with proper symlinks etc.  We depend on bootmisc
@@ -242,6 +251,7 @@ static void pidfile_init(void *arg)
 	 */
 	path = realpath(_PATH_VARRUN, NULL);
 	if (!path) {
+		_d("Failed querying realpath(%s): %s", _PATH_VARRUN, strerror(errno));
 		path = realpath("/run", NULL);
 		if (!path) {
 			_e("System does not have %s or /run, aborting.", _PATH_VARRUN);
@@ -251,6 +261,8 @@ static void pidfile_init(void *arg)

 	if (pidfile_add_path(&iw_pidfile, path))
 		iwatch_exit(&iw_pidfile);
+
+	free(path);
 }

 /*
@@ -23,6 +23,9 @@

 #include <glob.h>
 #include <lite/lite.h>
+
+#include "finit.h"
+#include "helpers.h"
 #include "plugin.h"

 static void setup(void *arg)
@@ -31,6 +34,11 @@ static void setup(void *arg)
 	glob_t gl;
 	char buf[160];

+	if (rescue) {
+		_d("Skipping %s plugin in rescue mode.", __FILE__);
+		return;
+	}
+
 	glob("/run/sysctl.d/*.conf",           0, NULL, &gl);
 	glob("/etc/sysctl.d/*.conf",           GLOB_APPEND, NULL, &gl);
 	glob("/usr/local/lib/sysctl.d/*.conf", GLOB_APPEND, NULL, &gl);
@@ -84,6 +84,11 @@ static void rtc_save(void *arg)
 	int fd, rc = 0;
 	char tz[128];

+	if (rescue) {
+		_d("Skipping %s plugin in rescue mode.", __FILE__);
+		return;
+	}
+
 	fd = rtc_open();
 	if (fd < 0)
 		return;
@@ -121,6 +126,11 @@ static void rtc_restore(void *arg)
 	int fd, rc = 0;
 	char tz[128];

+	if (rescue) {
+		_d("Skipping %s plugin in rescue mode.", __FILE__);
+		return;
+	}
+
 	fd = rtc_open();
 	if (fd < 0)
 		return;
@@ -0,0 +1,209 @@
+/* System condtion event monitor for the Finit condition engine
+ *
+ * Copyright (c) 2021  Joachim Wiberg <troglobit@gmail.com>
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to deal
+ * in the Software without restriction, including without limitation the rights
+ * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+ * copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
+ * THE SOFTWARE.
+ */
+
+#include <dirent.h>
+#include <fnmatch.h>
+#include <glob.h>
+#include <limits.h>
+#include <paths.h>
+
+#include "finit.h"
+#include "cond.h"
+#include "pid.h"
+#include "plugin.h"
+#include "iwatch.h"
+
+static struct iwatch iw_sys;
+
+
+static int sys_add_path(struct iwatch *iw, char *path)
+{
+	return iwatch_add(iw, path, IN_ONLYDIR | IN_CLOSE_WRITE);
+}
+
+/*
+ * unlink conditions with no active task/service, otherwise they may
+ * trigger inadvertently when calling `initctl reload` to activate
+ * a new configuration.
+ */
+static void sys_update_conds(char *dir, char *name, uint32_t mask)
+{
+	char path[256];
+	char *cond;
+	char *ptr;
+
+	ptr = strrchr(name, '/');
+	if (ptr)
+		snprintf(path, sizeof(path), "%s/%s", dir, ++ptr);
+	else
+		snprintf(path, sizeof(path), "%s/%s", dir, name);
+
+	cond = strstr(path, COND_BASE);
+	if (!cond)
+		return;
+
+	cond += strlen(COND_BASE) + 1;
+	_d("cond: %s set: %d", cond, mask & IN_CREATE ? 1 : 0);
+	if (!cond_update(cond))
+		unlink(path);
+}
+
+/* synthesize events in case of new run dirs */
+static void sys_scandir(struct iwatch *iw, char *dir, int len)
+{
+	glob_t gl;
+	size_t i;
+	char path[len + 6];
+	int rc;
+
+	snprintf(path, sizeof(path), "%s/*", dir);
+	rc = glob(path, GLOB_NOSORT, NULL, &gl);
+	if (rc && rc != GLOB_NOMATCH)
+		return;
+
+	for (i = 0; i < gl.gl_pathc; i++) {
+		_d("scan found %s", gl.gl_pathv[i]);
+		sys_update_conds(dir, gl.gl_pathv[i], IN_CREATE);
+	}
+	globfree(&gl);
+}
+
+/*
+ * create/remove sub-directory in monitored directory
+ */
+static void sys_handle_dir(struct iwatch *iw, char *dir, char *name, int mask)
+{
+	char path[strlen(dir) + strlen(name) + 2];
+	struct iwatch_path *iwp;
+
+	paste(path, sizeof(path), dir, name);
+	_d("path: %s", path);
+
+	iwp = iwatch_find_by_path(iw, path);
+
+	if (mask & IN_CREATE) {
+		if (!iwp) {
+			if (!sys_add_path(iw, path))
+				sys_scandir(iw, path, sizeof(path));
+		}
+	} else if (mask & IN_DELETE) {
+		if (iwp)
+			iwatch_del(&iw_sys, iwp);
+	}
+}
+
+static void sys_callback(void *arg, int fd, int events)
+{
+	static char ev_buf[8 *(sizeof(struct inotify_event) + NAME_MAX + 1) + 1];
+	struct inotify_event *ev;
+	ssize_t sz;
+	size_t off;
+
+	sz = read(fd, ev_buf, sizeof(ev_buf) - 1);
+	if (sz <= 0) {
+		_pe("invalid inotify event");
+		return;
+	}
+	ev_buf[sz] = 0;
+
+	for (off = 0; off < (size_t)sz; off += sizeof(*ev) + ev->len) {
+		struct iwatch_path *iwp;
+
+		if (off + sizeof(*ev) > (size_t)sz)
+			break;
+
+		ev = (struct inotify_event *)&ev_buf[off];
+		if (off + sizeof(*ev) + ev->len > (size_t)sz)
+			break;
+
+		_d("name %s, event: 0x%08x", ev->name, ev->mask);
+		if (!ev->mask)
+			continue;
+
+		/* Find base path for this event */
+		iwp = iwatch_find_by_wd(&iw_sys, ev->wd);
+		if (!iwp)
+			continue;
+
+		if (ev->mask & IN_ISDIR) {
+			sys_handle_dir(&iw_sys, iwp->path, ev->name, ev->mask);
+			continue;
+		}
+
+		if (ev->mask & (IN_CREATE | IN_DELETE))
+			sys_update_conds(iwp->path, ev->name, ev->mask);
+	}
+}
+
+static void sys_init(void *arg)
+{
+	char sysdir[MAX_ARG_LEN];
+	char *path;
+
+	if (mkpath(pid_runpath(_PATH_CONDSYS, sysdir, sizeof(sysdir)), 0755) && errno != EEXIST) {
+		_pe("Failed creating %s condition directory, %s", COND_SYS, _PATH_CONDSYS);
+		return;
+	}
+
+	path = realpath(_PATH_CONDSYS, NULL);
+	if (!path) {
+		_pe("Cannot figure out real path to %s, aborting", _PATH_CONDSYS);
+		return;
+	}
+
+	if (iwatch_add(&iw_sys, path, IN_ONLYDIR))
+		iwatch_exit(&iw_sys);
+}
+
+static plugin_t plugin = {
+	.name = __FILE__,
+	.hook[HOOK_BASEFS_UP]  = { .cb = sys_init },
+};
+
+PLUGIN_INIT(plugin_init)
+{
+	int fd;
+
+	fd = iwatch_init(&iw_sys);
+	if (fd < 0)
+		return;
+
+	plugin.io.fd = fd;
+	plugin.io.cb = sys_callback;
+	plugin.io.flags = PLUGIN_IO_READ;
+
+	plugin_register(&plugin);
+}
+
+PLUGIN_EXIT(plugin_exit)
+{
+	iwatch_exit(&iw_sys);
+	plugin_unregister(&plugin);
+}
+
+/**
+ * Local Variables:
+ *  indent-tabs-mode: t
+ *  c-file-style: "linux"
+ * End:
+ */
@@ -32,13 +32,14 @@
 #include "finit.h"
 #include "helpers.h"
 #include "plugin.h"
+#include "service.h"
 #include "tty.h"

-static void watcher(void *arg, int fd, int events);
+static void tty_watcher(void *arg, int fd, int events);

 static plugin_t plugin = {
 	.io = {
-		.cb    = watcher,
+		.cb    = tty_watcher,
 		.flags = PLUGIN_IO_READ,
 	},
 };
@@ -56,19 +57,24 @@ static void setup(void)
 static void do_tty(char *tty, size_t len, int creat)
 {
 	char name[len + 6];
-	struct tty *entry;
+	svc_t *svc;

 	snprintf(name, sizeof(name), "/dev/%s", tty);
-	entry = tty_find(name);
-	if (entry && tty_enabled(entry)) {
-		if (creat)
-			tty_start(entry);
-		else if (entry->pid)
-			tty_stop(entry);
+	svc = svc_find_by_tty(name);
+	if (svc) {
+		if (svc_is_blocked(svc) && creat) {
+			_d("%s: blocked, re-enabling", svc_ident(svc, NULL, 0));
+			svc_start(svc);
+		} else if (svc->pid) {
+			_d("%s: missing ...", svc_ident(svc, NULL, 0));
+			svc_missing(svc);
+		}
+
+		service_step_all(SVC_TYPE_TTY);
 	}
 }

-static void watcher(void *arg, int fd, int events)
+static void tty_watcher(void *arg, int fd, int events)
 {
 	static char ev_buf[8 *(sizeof(struct inotify_event) + NAME_MAX + 1) + 1];
 	struct inotify_event *ev;
@@ -1,4 +1,4 @@
-/* Setup random seed
+/* Setup and save random seed at boot/shutdown
 *
 * Copyright (c) 2012-2021  Joachim Wiberg <troglobit@gmail.com>
 *
@@ -35,12 +35,18 @@
 static void setup(void *arg)
 {
 #ifdef RANDOMSEED
+	if (rescue) {
+		_d("Skipping %s plugin in rescue mode.", __FILE__);
+		return;
+	}
+
 	if (!fexist(RANDOMSEED)) {
 		int ret = 1;
+		mode_t prev;
 		FILE *fp;

 		print_desc("Bootstrapping random seed", NULL);
-		umask(077);
+		prev = umask(077);
 		fp = fopen(RANDOMSEED, "w");
 		if (fp) {
 			int iter = 128;
@@ -57,7 +63,7 @@ static void setup(void *arg)
 			ret = fclose(fp);
 		}
 		print_result(ret);
-		umask(0);
+		umask(prev);
 	}

 	print_desc("Initializing random number generator", NULL);
@@ -68,10 +74,19 @@ static void setup(void *arg)
 static void save(void *arg)
 {
 #ifdef RANDOMSEED
-	umask(077);
+	mode_t prev;
+
+	if (rescue) {
+		_d("Skipping %s plugin in rescue mode.", __FILE__);
+		return;
+	}
+
+	prev = umask(077);
+
 	print_desc("Saving random seed", NULL);
 	print_result(512 != copyfile("/dev/urandom", RANDOMSEED, 512, 0));
-	umask(0);
+
+	umask(prev);
 #endif
 }

@@ -0,0 +1,133 @@
+/* User-defined condtion event monitor for the Finit condition engine
+ *
+ * Copyright (c) 2021  Joachim Wiberg <troglobit@gmail.com>
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to deal
+ * in the Software without restriction, including without limitation the rights
+ * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+ * copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
+ * THE SOFTWARE.
+ */
+
+#include <fnmatch.h>
+#include <glob.h>
+#include <limits.h>
+#include <paths.h>
+
+#include "finit.h"
+#include "cond.h"
+#include "pid.h"
+#include "plugin.h"
+#include "iwatch.h"
+
+static struct iwatch iw_usr;
+
+
+static void usr_cond(char *name, uint32_t mask)
+{
+	char cond[MAX_COND_LEN] = COND_USR;
+
+	strlcat(cond, name, sizeof(cond));
+	cond_update(cond);
+}
+
+static void usr_callback(void *arg, int fd, int events)
+{
+	static char ev_buf[8 *(sizeof(struct inotify_event) + NAME_MAX + 1) + 1];
+	struct inotify_event *ev;
+	ssize_t sz;
+	size_t off;
+
+	sz = read(fd, ev_buf, sizeof(ev_buf) - 1);
+	if (sz <= 0) {
+		_pe("invalid inotify event");
+		return;
+	}
+	ev_buf[sz] = 0;
+
+	for (off = 0; off < (size_t)sz; off += sizeof(*ev) + ev->len) {
+		if (off + sizeof(*ev) > (size_t)sz)
+			break;
+
+		ev = (struct inotify_event *)&ev_buf[off];
+		if (off + sizeof(*ev) + ev->len > (size_t)sz)
+			break;
+
+		_d("name %s, event: 0x%08x", ev->name, ev->mask);
+		if (!ev->mask)
+			continue;
+
+		if (ev->mask & IN_ISDIR)
+			continue;	/* unsupported */
+
+		if (ev->mask & IN_DELETE)
+			sync();		/* dont ask */
+
+		usr_cond(ev->name, ev->mask);
+	}
+}
+
+static void usr_init(void *arg)
+{
+	char usrdir[MAX_ARG_LEN];
+	char *path;
+
+	if (mkpath(pid_runpath(_PATH_CONDUSR, usrdir, sizeof(usrdir)), 0755) && errno != EEXIST) {
+		_pe("Failed creating %s condition directory, %s", COND_USR, _PATH_CONDUSR);
+		return;
+	}
+
+	path = realpath(_PATH_CONDUSR, NULL);
+	if (!path) {
+		_pe("Cannot figure out real path to %s, aborting", _PATH_CONDUSR);
+		return;
+	}
+
+	if (iwatch_add(&iw_usr, path, IN_ONLYDIR))
+		iwatch_exit(&iw_usr);
+}
+
+static plugin_t plugin = {
+	.name = __FILE__,
+	.hook[HOOK_BASEFS_UP]  = { .cb = usr_init },
+};
+
+PLUGIN_INIT(plugin_init)
+{
+	int fd;
+
+	fd = iwatch_init(&iw_usr);
+	if (fd < 0)
+		return;
+
+	plugin.io.fd = fd;
+	plugin.io.cb = usr_callback;
+	plugin.io.flags = PLUGIN_IO_READ;
+
+	plugin_register(&plugin);
+}
+
+PLUGIN_EXIT(plugin_exit)
+{
+	iwatch_exit(&iw_usr);
+	plugin_unregister(&plugin);
+}
+
+/**
+ * Local Variables:
+ *  indent-tabs-mode: t
+ *  c-file-style: "linux"
+ * End:
+ */
@@ -32,13 +32,19 @@

 static void setup(void *arg)
 {
+	const char *username = "nobody";
+	char line[LINE_SIZE];
+	mode_t prev;
 #ifdef PAM_CONSOLE
 	int fd;
 #endif
-	char line[LINE_SIZE];
-	const char *username = "nobody";

-	umask(0);
+	if (rescue) {
+		_d("Skipping %s plugin in rescue mode.", __FILE__);
+		return;
+	}
+
+	prev = umask(0);

 	makedir("/var/run/console", 01777);
 	snprintf(line, sizeof(line), "/var/run/console/%s", username);
@@ -58,7 +64,7 @@ static void setup(void *arg)
 	if (whichp("restorecon"))
 		run("restorecon /tmp/.ICE-unix /tmp/.X11-unix");

-	umask(022);
+	umask(prev);
 }

 static plugin_t plugin = {
@@ -1,8 +1,11 @@
 .deps/*
 finit
+getty
 initctl
+keventd
 logit
 reboot
+sulogin
 watchdogd
 Makefile
 Makefile.in
@@ -1,4 +1,4 @@
-EXTRA_DIST         = cgreaper.sh rescue.conf sample.conf
+EXTRA_DIST         = rescue.conf sample.conf
 AM_CPPFLAGS        = -D_XOPEN_SOURCE=600 -D_BSD_SOURCE -D_GNU_SOURCE -D_DEFAULT_SOURCE
 if STATIC
 AM_CPPFLAGS       += -DENABLE_STATIC
@@ -7,29 +7,42 @@ else
 AM_LDFLAGS         = -export-dynamic
 endif

-sbin_PROGRAMS      = finit initctl
-pkglibexec_SCRIPTS = cgreaper.sh
+sbin_PROGRAMS        = finit initctl
+pkglibexec_PROGRAMS  = getty logit sulogin
+if KEVENTD
+pkglibexec_PROGRAMS += keventd
+endif
+if WATCHDOGD
+pkglibexec_PROGRAMS += watchdogd
+endif

 finit_pkglibdir      = $(pkglibdir)
 finit_pkglib_SCRIPTS = rescue.conf sample.conf

-if WATCHDOGD
-pkglibexec_PROGRAMS = watchdogd
-endif
+getty_SOURCES        = finit.h getty.c helpers.h logrotate.c stty.c utmp-api.c utmp-api.h
+getty_CFLAGS         = -W -Wall -Wextra -std=gnu99
+getty_CFLAGS        += $(lite_CFLAGS)
+getty_LDADD          = $(lite_LIBS)

-if LOGIT
-pkglibexec_PROGRAMS = logit
-logit_SOURCES       = logit.c logrotate.c
-logit_CFLAGS        = -W -Wall -Wextra -Wno-unused-parameter -std=gnu99
-logit_CFLAGS       += $(lite_CFLAGS)
-logit_LDADD         = $(lite_LIBS)
-endif
+keventd_SOURCES      = keventd.c iwatch.c iwatch.h util.c util.h
+keventd_CFLAGS       = -W -Wall -Wextra -std=gnu99
+keventd_CFLAGS      += $(lite_CFLAGS)
+keventd_LDADD        = $(lite_LIBS)
+
+sulogin_SOURCES      = sulogin.c
+sulogin_CFLAGS       = -W -Wall -Wextra -std=gnu99
+sulogin_LDADD        = -lcrypt
+
+logit_SOURCES        = logit.c logrotate.c
+logit_CFLAGS         = -W -Wall -Wextra -Wno-unused-parameter -std=gnu99
+logit_CFLAGS        += $(lite_CFLAGS)
+logit_LDADD          = $(lite_LIBS)

 finit_SOURCES      = api.c	cgroup.c	cgroup.h	\
 		     cond.c	cond-w.c	cond.h		\
 		     conf.c	conf.h				\
 		     exec.c	finit.c		finit.h		\
-		     getty.c	stty.c				\
+		     		stty.c				\
 		     helpers.c	helpers.h			\
 		     iwatch.c   iwatch.h			\
 		     log.c	log.h				\
@@ -48,7 +61,7 @@ if LOGROTATE
 finit_SOURCES     += logrotate.c
 endif

-pkginclude_HEADERS = cond.h finit.h helpers.h log.h plugin.h svc.h
+pkginclude_HEADERS = cgroup.h cond.h finit.h helpers.h log.h plugin.h svc.h

 finit_CFLAGS       = -W -Wall -Wextra -Wno-unused-parameter -std=gnu99
 finit_CFLAGS      += $(lite_CFLAGS) $(uev_CFLAGS)
@@ -59,20 +72,19 @@ else
 finit_LDADD       += -ldl
 endif

-initctl_SOURCES    = initctl.c client.c client.h 	\
-		     reboot.c serv.c serv.h svc.h	\
-		     cond.c cond.h util.c util.h
+initctl_SOURCES    = initctl.c initctl.h cgutil.c cgutil.h		\
+		     client.c client.h cond.c cond.h reboot.c		\
+		     serv.c serv.h svc.h util.c util.h
 initctl_CFLAGS     = -W -Wall -Wextra -Wno-unused-parameter -std=gnu99
-initctl_CFLAGS    += $(lite_CFLAGS)
-initctl_LDADD      = $(lite_LIBS)
+initctl_CFLAGS    += $(lite_CFLAGS) $(uev_CFLAGS)
+initctl_LDADD      = $(lite_LIBS) $(uev_LIBS)

-INIT_LNKS          = init
+INIT_LNKS          = init telinit
 REBOOT_LNKS        = reboot shutdown halt poweroff suspend

 if WATCHDOGD
 watchdogd_SOURCES  = watchdog.c watchdog.h
 watchdogd_CFLAGS   = -W -Wall -Wextra -Wno-unused-parameter -std=gnu99
-#log.c log.h
 endif

 # Hook in install to add finit and reboot symlink(s)
@@ -74,7 +74,24 @@ static int start(svc_t *svc)
 	return 0;
 }

+/*
+ * NOTE: this does not wait for svc to be stopped first, that is the
+ *       responsibility of initctl to do.  Otherwise we'd block PID 1,
+ *       or introduce some nasty race conditions.
+ */
 static int restart(svc_t *svc)
+{
+	if (!svc)
+		return 1;
+
+	svc_mark_dirty(svc);
+	svc_start(svc);
+	service_step(svc);
+
+	return 0;
+}
+
+static int reload(svc_t *svc)
 {
 	if (!svc)
 		return 1;
@@ -88,9 +105,10 @@ static int restart(svc_t *svc)
 	return 0;
 }

-static int do_start  (char *buf, size_t len) { return call(start,   buf, len); }
 static int do_stop   (char *buf, size_t len) { return call(stop,    buf, len); }
+static int do_start  (char *buf, size_t len) { return call(start,   buf, len); }
 static int do_restart(char *buf, size_t len) { return call(restart, buf, len); }
+static int do_reload (char *buf, size_t len) { return call(reload,  buf, len); }

 static char query_buf[368];
 static int missing(char *job, char *id)
@@ -110,7 +128,7 @@ static int missing(char *job, char *id)

 static int do_query(char *buf, size_t len)
 {
-	memset(query_buf, 0, sizeof(query_buf));
+	query_buf[0] = 0;
 	if (svc_parse_jobstr(buf, len, NULL, missing)) {
 		memcpy(buf, query_buf, len);
 		return 1;
@@ -252,10 +270,10 @@ static void send_svc(int sd, svc_t *svc)

 static void api_cb(uev_t *w, void *arg, int events)
 {
-	int sd, lvl;
-	svc_t *svc;
 	static svc_t *iter = NULL;
 	struct init_request rq;
+	int sd, lvl;
+	svc_t *svc;

 	sd = accept(w->fd, NULL, NULL);
 	if (sd < 0) {
@@ -327,16 +345,22 @@ static void api_cb(uev_t *w, void *arg, int events)
 			result = do_start(rq.data, sizeof(rq.data));
 			break;

+		case INIT_CMD_RESTART_SVC:
+			_d("restart %s", rq.data);
+			strterm(rq.data, sizeof(rq.data));
+			result = do_restart(rq.data, sizeof(rq.data));
+			break;
+
 		case INIT_CMD_STOP_SVC:
 			_d("stop %s", rq.data);
 			strterm(rq.data, sizeof(rq.data));
 			result = do_stop(rq.data, sizeof(rq.data));
 			break;

-		case INIT_CMD_RESTART_SVC:
-			_d("restart %s", rq.data);
+		case INIT_CMD_RELOAD_SVC:
+			_d("reload %s", rq.data);
 			strterm(rq.data, sizeof(rq.data));
-			result = do_restart(rq.data, sizeof(rq.data));
+			result = do_reload(rq.data, sizeof(rq.data));
 			break;

 		case INIT_CMD_GET_RUNLEVEL:
@@ -364,24 +388,30 @@ static void api_cb(uev_t *w, void *arg, int events)
 			}

 			_e("Request to hand-over wdog ... to PID %d", rq.runlevel);
-			if (!svc_find_by_pid(rq.runlevel)) {
+			svc = svc_find_by_pid(rq.runlevel);
+			if (!svc) {
 				logit(LOG_ERR, "Cannot find PID %d, not registered.", rq.runlevel);
 				break;
 			}

-			/* Disable and allow Finit to collect bundled watchdog */
-			if (wdog) {
-				logit(LOG_NOTICE, "Stopping and removing %s (PID:%d)", wdog->cmd, wdog->pid);
-				stop(wdog);
+			if (wdog && wdog != svc) {
+				char name[32];
+
+				svc_ident(svc, name, sizeof(name));
+				logit(LOG_NOTICE, "Handing over wdog ctrl from %s[%d] to %s[%d]",
+				      svc_ident(wdog, NULL, 0), wdog->pid, name, svc->pid);
+
 				if (wdog->protect) {
-					wdog->protect = 0;
-					wdog->runlevels = 0;
+					logit(LOG_NOTICE, "Stopping and deleting built-in watchdog.");
+					stop(wdog);
+					svc_del(wdog);
 				}
 			}
+			wdog = svc;
 			break;

 		case INIT_CMD_SVC_ITER:
-			_d("svc iter, first: %d", rq.runlevel);
+//			_d("svc iter, first: %d", rq.runlevel);
 			/*
 			 * XXX: This severly limits the number of
 			 * simultaneous client connections, but will
@@ -444,7 +474,7 @@ int api_init(uev_ctx_t *ctx)
 	};

 	_d("Setting up external API socket ...");
-	sd = socket(AF_UNIX, SOCK_STREAM | SOCK_CLOEXEC, 0);
+	sd = socket(AF_UNIX, SOCK_SEQPACKET | SOCK_CLOEXEC, 0);
 	if (-1 == sd) {
 		_pe("Failed starting external API socket");
 		return 1;
@@ -1,11 +0,0 @@
-#!/bin/sh
-# Called by kernel when last child in cgroup exists
-
-BASE=/sys/fs/cgroup/finit
-PROC=`basename $1`
-DIR="$BASE$1"
-
-#/libexec/finit/logit -p daemon.info -t finit "$PROC stopped, cleaning up control group $DIR"
-rmdir $DIR
-
-exit 0
@@ -25,116 +25,461 @@
 #include <stdio.h>
 #include <string.h>
 #include <lite/lite.h>
+#include <lite/queue.h>
 #include <sys/mount.h>
+#include <sys/sysinfo.h>		/* get_nprocs_conf() */

+#include "cgroup.h"
 #include "finit.h"
+#include "iwatch.h"
 #include "log.h"
 #include "util.h"

-static int cg_init = 0;
+struct cg {
+	TAILQ_ENTRY(cg) link;

-/*
- * Called by Finit at early boot to mount initial cgroups
- */
-void cgroup_init(void)
+	char *name;		/* top-level group name */
+	char *cfg;		/* kernel settings */
+
+	int  active;		/* for mark & sweep */
+	int  protected;		/* for init/, user/, & system/ */
+};
+
+static TAILQ_HEAD(, cg) cgroups = TAILQ_HEAD_INITIALIZER(cgroups);
+
+static char controllers[256];
+
+static struct iwatch iw_cgroup;
+static uev_t cgw;
+
+
+static void cgset(const char *path, char *ctrl, char *prop)
 {
-	FILE *fp;
-	char buf[80];
-	int opts = MS_NODEV | MS_NOEXEC | MS_NOSUID;
+	char *val;

-	fp = fopen("/proc/cgroups", "r");
-	if (!fp) {
-		_d("No cgroup support");
+	_d("path %s, ctrl %s, prop %s", path ?: "NIL", ctrl ?: "NIL", prop ?: "NIL");
+	if (!path || !ctrl) {
+		_e("Missing path or controller, skipping!");
 		return;
 	}

-	if (mount("none", "/sys/fs/cgroup", "tmpfs", opts, NULL)) {
-		_d("Failed mounting cgroups");
-		goto fail;
+	if (!prop) {
+		prop = strchr(ctrl, '.');
+		if (!prop) {
+			_e("Invalid cgroup ctrl syntax: %s", ctrl);
+			return;
+		}
+
+		*prop++ = 0;
 	}

-	/* Skip first line, header */
-	(void)fgets(buf, sizeof(buf), fp);
-	while (fgets(buf, sizeof(buf), fp)) {
-		char *cgroup;
-		char rc[80];
+	val = strchr(prop, ':');
+	if (!val) {
+		_e("Missing cgroup ctrl value, prop %s", prop);
+		return;
+	}
+	*val++ = 0;

-		cgroup = strtok(buf, "\t ");
-		if (!cgroup)
-			continue;
-
-		snprintf(rc, sizeof(rc), "/sys/fs/cgroup/%s", cgroup);
-		if (mkdir(rc, 0755) && EEXIST != errno)
-			continue;
-		if (mount("cgroup", rc, "cgroup", opts, cgroup))
-			_d("Failed mounting %s cgroup on %s", cgroup, rc);
+	/* disallow sneaky relative paths */
+	if (strstr(ctrl, "..") || strstr(prop, "..")) {
+		_e("Possible security violation; '..' not allowed in cgroup config!");
+		return;
 	}

-	/* Default cgroups for process monitoring */
-	if (mkdir(FINIT_CGPATH, 0755) && EEXIST != errno)
-		goto fail;
-	if (mount("none", FINIT_CGPATH, "cgroup", opts, "none,name=finit")) {
-		_pe("Failed mounting Finit cgroup hierarchy");
-		goto fail;
-	}
-
-	/* Set up reaper and enable callbacks */
-	echo(FINIT_CGPATH "/release_agent", 0, FINIT_LIBPATH_ "/cgreaper.sh");
-	echo(FINIT_CGPATH "/notify_on_release", 0, "1");
-
-	/* Default groups, PID 1, services, and user/login processes */
-	if (mkdir(FINIT_CGPATH "/init", 0755) && EEXIST != errno)
-		goto fail;
-	if (mkdir(FINIT_CGPATH "/system", 0755) && EEXIST != errno)
-		goto fail;
-	if (mkdir(FINIT_CGPATH "/user", 0755) && EEXIST != errno)
-		goto fail;
-
-	/* Move ourselves to init */
-	echo(FINIT_CGPATH "/init/cgroup.procs", 0, "1");
-
-	/* We have signal, main screen turn on! */
-	cg_init = 1;
-fail:
-	fclose(fp);
+	_d("%s/%s.%s <= %s", path, ctrl, prop, val);
+	if (fnwrite(val, "%s/%s.%s", path, ctrl, prop))
+		_pe("Failed setting %s/%s.%s = %s", path, ctrl, prop, val);
 }

-static int move_pid(char *group, char *name, char *id, int pid)
+/*
+ * Settings for a cgroup are on the form: cpu.weight:1234,mem.max:4321,...
+ * Finit supports the short-form 'mem.', replacing it with 'memory.' when
+ * writing the setting to the file system.
+ */
+static void group_init(char *path, int leaf, const char *cfg)
+{
+	char *ptr, *s;
+
+	_d("path %s, leaf %d, cfg %s", path, leaf, cfg ?: "NIL");
+	if (!fisdir(path)) {
+		if (mkdir(path, 0755)) {
+			_pe("Failed creating cgroup %s", path);
+			return;
+		}
+
+		/* enable detected controllers on domain groups */
+		if (!leaf && fnwrite(controllers, "%s/cgroup.subtree_control", path))
+			_pe("Failed enabling %s for %s", controllers, path);
+	}
+
+	if (!cfg || !cfg[0])
+		return;
+
+	s = strdupa(cfg);
+	if (!s) {
+		_pe("Failed activating cgroup cfg for %s", path);
+		return;
+	}
+
+	_d("%s <=> %s", path, s);
+	ptr = strtok(s, ",");
+	while (ptr) {
+		_d("ptr: %s", ptr);
+		if (!strncmp("mem.", ptr, 4))
+			cgset(path, "memory", &ptr[4]);
+		else
+			cgset(path, ptr, NULL);
+
+		ptr = strtok(NULL, ",");
+	}
+}
+
+static int cgroup_leaf_init(char *group, char *name, int pid, const char *cfg)
 {
 	char path[256];

-	snprintf(path, sizeof(path), "/sys/fs/cgroup/%s/%s", group, name);
-	if (id && strlen(id)) {
-		strlcat(path, ":", sizeof(path));
-		strlcat(path, id, sizeof(path));
-	}
-	if (mkdir(path, 0755) && errno != EEXIST)
-		return 1;
-
-	strlcat(path, "/cgroup.procs", sizeof(path));
-
-	return echo(path, 0, "%d", pid);
-}
-
-int cgroup_user(char *name)
-{
-	if (!cg_init)
-		return 0;
-
-	return move_pid("finit/user", name, "0", getpid());
-}
-
-int cgroup_service(char *nm, char *id, int pid)
-{
-	if (!cg_init)
-		return 0;
-
-	if (pid <= 0) {
+	_d("group %s, name %s, pid %d, cfg %s", group, name, pid, cfg ?: "NIL");
+	if (pid < 0 || pid == 1) {
 		errno = EINVAL;
 		return 1;
 	}

-	return move_pid("finit/system", nm, id, pid);
+	/* create and initialize new group */
+	snprintf(path, sizeof(path), "/sys/fs/cgroup/%s/%s", group, name);
+	group_init(path, 1, cfg);
+
+	/* move process to new group */
+	if (fnwrite(str("%d", pid), "%s/cgroup.procs", path))
+		_pe("Failed moving pid %d to group %s", pid, path);
+
+	strlcat(path, "/cgroup.events", sizeof(path));
+
+	return iwatch_add(&iw_cgroup, path, 0);
+}
+
+int cgroup_user(char *name, int pid)
+{
+	return cgroup_leaf_init("user", name, pid, NULL);
+}
+
+int cgroup_service(char *name, int pid, struct cgroup *cg)
+{
+	char *group = "system";
+
+	if (cg && cg->name[0]) {
+		char path[256];
+
+		if (!strcmp(cg->name, "root"))
+			return fnwrite(str("%d", pid), FINIT_CGPATH "/cgroup.procs");
+
+		if (!strcmp(cg->name, "init"))
+			return fnwrite(str("%d", pid), FINIT_CGPATH "/init/cgroup.procs");
+
+		snprintf(path, sizeof(path), "/sys/fs/cgroup/%s", cg->name);
+		if (fisdir(path))
+			group = cg->name;
+	}
+
+	return cgroup_leaf_init(group, name, pid, cg ? cg->cfg : NULL);
+}
+
+static void append_ctrl(char *ctrl)
+{
+	if (controllers[0])
+		strlcat(controllers, " ", sizeof(controllers));
+
+	strlcat(controllers, "+", sizeof(controllers));
+	strlcat(controllers, ctrl, sizeof(controllers));
+}
+
+static void cgroup_handle_event(char *event, uint32_t mask)
+{
+	char path[strlen(event) + 1];
+	char buf[80];
+	char *ptr;
+	FILE *fp;
+
+	_d("event: '%s', mask: %08x", event, mask);
+	if (!(mask & IN_MODIFY))
+		return;
+
+	fp = fopen(event, "r");
+	if (!fp) {
+		_d("Failed opening %s, skipping ...", event);
+		return;
+	}
+
+	while (fgets(buf, sizeof(buf), fp)) {
+		if (strncmp(buf, "populated", 9))
+			continue;
+
+		chomp(buf);
+		if (atoi(&buf[10]))
+			break;
+
+		strlcpy(path, event, sizeof(path));
+		ptr = strrchr(path, '/');
+		if (ptr) {
+			*ptr = 0;
+			if (!cgroup_del(path)) {
+				/*
+				 * try with parent, top-level group, we
+				 * may get events out-of-order *sigh*
+				 */
+				ptr = strrchr(path, '/');
+				if (!ptr)
+					break;
+				*ptr = 0;
+				cgroup_del(path);
+			}
+		}
+
+		break;
+	}
+
+	fclose(fp);
+}
+
+static void cgroup_events_cb(uev_t *w, void *arg, int events)
+{
+	static char ev_buf[8 *(sizeof(struct inotify_event) + NAME_MAX + 1) + 1];
+	struct inotify_event *ev;
+	ssize_t sz;
+	size_t off;
+
+	sz = read(w->fd, ev_buf, sizeof(ev_buf) - 1);
+	if (sz <= 0) {
+		_pe("invalid inotify event");
+		return;
+	}
+	ev_buf[sz] = 0;
+
+	for (off = 0; off < (size_t)sz; off += sizeof(*ev) + ev->len) {
+		struct iwatch_path *iwp;
+
+		if (off + sizeof(*ev) > (size_t)sz)
+			break;
+
+		ev = (struct inotify_event *)&ev_buf[off];
+		if (off + sizeof(*ev) + ev->len > (size_t)sz)
+			break;
+
+		if (!ev->mask)
+			continue;
+
+		/* Find base path for this event */
+		iwp = iwatch_find_by_wd(&iw_cgroup, ev->wd);
+		if (!iwp || !iwp->path)
+			continue;
+
+		cgroup_handle_event(iwp->path, ev->mask);
+	}
+
+#ifdef ENABLE_AUTO_RELOAD
+	if (conf_any_change())
+		service_reload_dynamic();
+#endif
+}
+
+static struct cg *cgroup_find(char *name)
+{
+	struct cg *cg;
+
+	TAILQ_FOREACH(cg, &cgroups, link) {
+		if (strcmp(cg->name, name))
+			continue;
+
+		return cg;
+	}
+
+	return NULL;
+}
+
+/*
+ * Marks all unprotected cgroups for deletion (during reload)
+ */
+void cgroup_mark_all(void)
+{
+	struct cg *cg;
+
+	TAILQ_FOREACH(cg, &cgroups, link) {
+		if (cg->protected)
+			continue;
+
+		cg->active = 0;
+	}
+}
+
+/*
+ * Remove (try to) all unused cgroups
+ */
+void cgroup_cleanup(void)
+{
+	struct cg *cg, *tmp;
+	char path[256];
+
+	TAILQ_FOREACH_SAFE(cg, &cgroups, link, tmp) {
+		if (cg->active)
+			continue;
+
+		snprintf(path, sizeof(path), FINIT_CGPATH "/%s", cg->name);
+		cgroup_del(path);
+	}
+}
+
+/*
+ * Add, or update, settings for top-level cgroup
+ */
+int cgroup_add(char *name, char *cfg, int protected)
+{
+	struct cg *cg;
+
+	if (!name)
+		return -1;
+	if (!cfg)
+		cfg = "";
+
+	cg = cgroup_find(name);
+	if (!cg) {
+		cg = malloc(sizeof(struct cg));
+		if (!cg) {
+			_pe("Failed allocating 'struct cg' for %s", name);
+			return -1;
+		}
+		cg->name = strdup(name);
+		if (!cg->name) {
+			_pe("Failed setting cgroup name %s", name);
+			free(cg);
+			return -1;
+		}
+		TAILQ_INSERT_TAIL(&cgroups, cg, link);
+	} else
+		free(cg->cfg);
+
+	cg->cfg = strdup(cfg);
+	if (!cg->cfg) {
+		_pe("Failed add/update of cgroup %s", name);
+		TAILQ_REMOVE(&cgroups, cg, link);
+		free(cg->name);
+		free(cg);
+		return -1;
+	}
+	cg->protected = protected;
+	cg->active = 1;
+
+	return 0;
+}
+
+/*
+ * Remove inactive top-level cgroup
+ */
+int cgroup_del(char *dir)
+{
+	struct cg *cg;
+	char path[256];
+
+	TAILQ_FOREACH(cg, &cgroups, link) {
+		snprintf(path, sizeof(path), FINIT_CGPATH "/%s", cg->name);
+		if (strcmp(path, dir))
+			continue;
+
+		if (cg->active)
+			return -1;
+
+		break;
+	}
+
+	if (rmdir(dir) && errno != ENOENT) {
+		_d("Failed removing %s: %s", dir, strerror(errno));
+		return -1;
+	}
+
+	if (cg) {
+		TAILQ_REMOVE(&cgroups, cg, link);
+		free(cg->name);
+		free(cg->cfg);
+		free(cg);
+	}
+
+	return 0;
+}
+
+/* the top-level init cgroup is a leaf, that's ensured in cgroup_init() */
+void cgroup_config(void)
+{
+	struct cg *cg;
+
+	TAILQ_FOREACH(cg, &cgroups, link) {
+		char path[256];
+		int leaf = 0;
+
+		if (!cg->active)
+			continue;
+		if (!strcmp(cg->name, "init"))
+			leaf = 1;	/* reserved */
+
+		snprintf(path, sizeof(path), "%s/%s", FINIT_CGPATH, cg->name);
+		group_init(path, leaf, cg->cfg);
+
+		strlcat(path, "/cgroup.events", sizeof(path));
+		iwatch_add(&iw_cgroup, path, 0);
+	}
+}
+
+/*
+ * Called by Finit at early boot to mount initial cgroups
+ */
+void cgroup_init(uev_ctx_t *ctx)
+{
+	int opts = MS_NODEV | MS_NOEXEC | MS_NOSUID;
+	char buf[80];
+	FILE *fp;
+	int fd;
+
+	if (mount("none", FINIT_CGPATH, "cgroup2", opts, NULL)) {
+		_pe("Failed mounting cgroup v2");
+		return;
+	}
+
+	/* Find available controllers */
+	fp = fopen(FINIT_CGPATH "/cgroup.controllers", "r");
+	if (!fp) {
+		_pe("Failed opening %s", FINIT_CGPATH "/cgroup.controllers");
+		return;
+	}
+
+	if (fgets(buf, sizeof(buf), fp)) {
+		char *cgroup;
+
+		cgroup = strtok(chomp(buf), "\t ");
+		while (cgroup) {
+			append_ctrl(cgroup);
+			cgroup = strtok(NULL, "\t ");
+		}
+	}
+
+	fclose(fp);
+
+	/* Enable all controllers */
+	if (fnwrite(controllers, FINIT_CGPATH "/cgroup.subtree_control"))
+		_pe("Failed enabling %s for %s", controllers, FINIT_CGPATH "/cgroup.subtree_control");
+
+	/* Default (protected) groups, PID 1, services, and user/login processes */
+	cgroup_add("init",   "cpu.weight:100",  1);
+	cgroup_add("system", "cpu.weight:9800", 1);
+	cgroup_add("user",   "cpu.weight:100",  1);
+	cgroup_config();
+
+	/* Move ourselves to init (best effort, otherwise run in 'root' group */
+	if (fnwrite("1", FINIT_CGPATH "/init/cgroup.procs"))
+		_pe("Failed moving PID 1 to cgroup ", FINIT_CGPATH "/init");
+
+	/* prepare cgroup.events watcher */
+	fd = iwatch_init(&iw_cgroup);
+	if (uev_io_init(ctx, &cgw, cgroup_events_cb, NULL, fd, UEV_READ)) {
+		_pe("Failed setting up cgroup.events watcher");
+		close(fd);
+	}
 }

 /**
@@ -24,9 +24,23 @@
 #ifndef FINIT_CGROUP_H_
 #define FINIT_CGROUP_H_

-void cgroup_init   (void);
+#include <uev/uev.h>

-int cgroup_user    (char *name);
-int cgroup_service (char *cmd, char *id, int pid);
+struct cgroup {
+	char name[16];
+	char cfg[128];
+};
+
+void cgroup_mark_all(void);
+void cgroup_cleanup (void);
+
+int  cgroup_add     (char *name, char *cfg, int protected);
+int  cgroup_del     (char *dir);
+void cgroup_config  (void);
+
+void cgroup_init    (uev_ctx_t *ctx);
+
+int  cgroup_user    (char *name, int pid);
+int  cgroup_service (char *name, int pid, struct cgroup *cg);

 #endif /* FINIT_CGROUP_H_ */
@@ -0,0 +1,654 @@
+/* New client tool, replaces old /dev/initctl API and telinit tool
+ *
+ * Copyright (c) 2019-2021  Joachim Wiberg <troglobit@gmail.com>
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to deal
+ * in the Software without restriction, including without limitation the rights
+ * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+ * copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
+ * THE SOFTWARE.
+ */
+
+#include "config.h"
+
+#include <err.h>
+#include <dirent.h>
+#include <stdio.h>
+#include <signal.h>
+#include <stdlib.h>
+#include <search.h>
+#include <inttypes.h>
+#include <lite/lite.h>
+#include <sys/sysinfo.h>		/* sysinfo() */
+
+#include "cgutil.h"
+#include "initctl.h"
+
+#define CDIM plain ? "" : "\e[2m"
+#define CRST plain ? "" : "\e[0m"
+
+#define NONE " "
+#define PIPE plain ? "| " : "│ "
+#define FORK plain ? "|-" : "├─"
+#define END  plain ? "`-" : "└─"
+
+uint64_t total_ram;			/* From sysinfo() */
+
+struct cg  dummy;			/* empty result "NULL"      */
+struct cg *list;
+
+
+char *pid_cmdline(int pid, char *buf, size_t len)
+{
+	size_t i, sz;
+	char *ptr;
+	FILE *fp;
+
+	fp = fopenf("r", "/proc/%d/cmdline", pid);
+	if (!fp) {
+		buf[0] = 0;
+		return buf;
+	}
+
+	sz = fread(buf, sizeof(buf[0]), len - 1, fp);
+	fclose(fp);
+	if (!sz)
+		return NULL;		/* kernel thread */
+
+	buf[sz] = 0;
+
+	ptr = strchr(buf, 0);
+	if (ptr && ptr != buf) {
+		ptr++;
+		sz -= ptr - buf;
+		memmove(buf, ptr, sz + 1);
+	}
+
+	for (i = 0; i < sz; i++) {
+		if (buf[i] == 0)
+			buf[i] = ' ';
+	}
+
+	return buf;
+}
+
+char *pid_comm(int pid, char *buf, size_t len)
+{
+	char *ptr = NULL;
+	FILE *fp;
+
+	fp = fopenf("r", "/proc/%d/comm", pid);
+	if (!fp)
+		return NULL;
+
+	if (fgets(buf, len, fp))
+		ptr = chomp(buf);
+	fclose(fp);
+
+	return ptr;
+}
+
+char *pid_cgroup(int pid, char *buf, size_t len)
+{
+	char *ptr = NULL;
+	FILE *fp;
+
+	fp = fopenf("r", "/proc/%d/cgroup", pid);
+	if (!fp)
+		return NULL;
+
+	if (fgets(buf, len, fp))
+		ptr = chomp(buf);
+	fclose(fp);
+
+	if (ptr)
+		ptr = strchr(buf, '/');
+
+	return ptr;
+}
+
+static char *cgroup_val(char *path, char *file, char *buf, size_t len)
+{
+	char *val = NULL;
+	FILE *fp;
+
+	fp = fopenf("r", "%s/%s", path, file);
+	if (fp) {
+		if (fgets(buf, len, fp)) {
+			val = chomp(buf);
+			len = strcspn(val, " \t");
+			val[len] = 0;
+		}
+
+		fclose(fp);
+	}
+
+	return val;
+}
+
+static uint64_t cgroup_uint64(char *path, char *file)
+{
+	uint64_t val = 0;
+	char buf[42];
+
+	if (cgroup_val(path, file, buf, sizeof(buf)))
+		val = strtoull(buf, NULL, 10);
+
+	return val;
+}
+
+static char *cgroup_memval(char *path, char *file, char *buf, size_t len)
+{
+	char data[42];
+
+	if (cgroup_val(path, file, data, sizeof(data))) {
+		if (!strcmp(data, "max"))
+			strlcpy(buf, data, len);
+		else
+			memsz(strtoull(data, NULL, 10), buf, len);
+	} else
+		buf[0] = 0;
+
+	return buf;
+}
+
+static uint64_t cgroup_memuse(struct cg *cg)
+{
+	char buf[42];
+	FILE *fp;
+
+	fp = fopenf("r", "%s/memory.stat", cg->cg_path);
+	if (fp) {
+		cg->cg_rss = 0;
+		cg->cg_vmlib = 0;
+
+		while (fgets(buf, sizeof(buf), fp)) {
+			chomp(buf);
+
+			if (!strncmp(buf, "anon", 4)) {
+				cg->cg_rss += strtoull(&buf[5], NULL, 10);
+				continue;
+			}
+			if (!strncmp(buf, "slab", 4)) {
+				cg->cg_rss += strtoull(&buf[5], NULL, 10);
+				continue;
+			}
+			if (!strncmp(buf, "kernel_stack", 12)) {
+				cg->cg_rss += strtoull(&buf[5], NULL, 10);
+				continue;
+			}
+			if (!strncmp(buf, "pagetables", 10)) {
+				cg->cg_rss += strtoull(&buf[5], NULL, 10);
+				continue;
+			}
+			if (!strncmp(buf, "percpu", 6)) {
+				cg->cg_rss += strtoull(&buf[5], NULL, 10);
+				continue;
+			}
+			if (!strncmp(buf, "sock", 4)) {
+				cg->cg_rss += strtoull(&buf[5], NULL, 10);
+				continue;
+			}
+			if (!strncmp(buf, "file", 4)) {
+				cg->cg_vmlib += strtoull(&buf[5], NULL, 10);
+				continue;
+			}
+			if (!strncmp(buf, "file_mapped", 11)) {
+				cg->cg_vmlib += strtoull(&buf[5], NULL, 10);
+				continue;
+			}
+		}
+		fclose(fp);
+	}
+
+	cg->cg_memshare = (float)(cg->cg_rss * 100 / total_ram);
+
+	return cg->cg_vmsize = cgroup_uint64(cg->cg_path, "memory.current");
+}
+
+uint64_t cgroup_memory(char *group)
+{
+	char path[256];
+
+	paste(path, sizeof(path), FINIT_CGPATH, group);
+
+	return cgroup_uint64(path, "memory.current");
+}
+
+static float cgroup_cpuload(struct cg *cg)
+{
+	char fn[256];
+	char buf[64];
+	FILE *fp;
+
+	snprintf(fn, sizeof(fn), "%s/cpu.stat", cg->cg_path);
+	fp = fopen(fn, "r");
+	if (!fp)
+		err(1, "Cannot open %s", fn);
+
+	while (fgets(buf, sizeof(buf), fp)) {
+		uint64_t curr;
+
+		chomp(buf);
+		if (strncmp(buf, "usage_usec", 10))
+			continue;
+
+		curr = strtoull(&buf[11], NULL, 10);
+		if (cg->cg_prev != 0) {
+			uint64_t diff = curr - cg->cg_prev;
+
+			/* this expects 1 sec poll interval */
+			cg->cg_load = (float)(diff / 1000000);
+			cg->cg_load *= 100.0;
+		}
+		cg->cg_prev = curr;
+		break;
+	}
+
+	fclose(fp);
+
+	return cg->cg_load;
+}
+
+static struct cg *append(char *path)
+{
+	struct cg *cg;
+	char fn[256];
+	ENTRY item;
+
+	snprintf(fn, sizeof(fn), "%s/cpu.stat", path);
+	if (access(fn, F_OK)) {
+		/* older kernels, 4.19, don't have summary cpu.stat in root */
+		if (strcmp(path, FINIT_CGPATH))
+			warn("not a cgroup path with cpu controller, %s", path);
+		return NULL;
+	}
+
+	cg = calloc(1, sizeof(struct cg));
+	if (!cg)
+		err(1, "failed allocating struct cg");
+
+	cg->cg_path = strdup(path);
+	if (list)
+		cg->cg_next = list;
+	list = cg;
+
+	item.key  = cg->cg_path;
+	item.data = cg;
+	if (!hsearch(item, ENTER))
+		err(1, "failed adding to hash table");
+
+	return cg;
+}
+
+static struct cg *find(char *path)
+{
+	ENTRY *ep, item = { path, NULL };
+
+	ep = hsearch(item, FIND);
+	if (ep)
+		return ep->data;
+
+	return append(path);
+}
+
+/* update stats */
+struct cg *cg_stats(char *path)
+{
+	struct cg *cg;
+
+	cg = find(path);
+	if (!cg)
+		return &dummy;
+
+	cgroup_cpuload(cg);
+	cgroup_memuse(cg);
+
+	return cg;
+}
+
+/* query config */
+struct cg *cg_conf(char *path)
+{
+	static struct cg cg;
+
+	cgroup_val(path, "memory.min", cg.cg_mem.min, sizeof(cg.cg_mem.min));
+	cgroup_memval(path, "memory.max", cg.cg_mem.max, sizeof(cg.cg_mem.max));
+	cgroup_val(path, "cpu.weight", cg.cg_cpu.weight, sizeof(cg.cg_cpu.weight));
+	cgroup_val(path, "cpu.max",    cg.cg_cpu.max, sizeof(cg.cg_cpu.max));
+	cgroup_val(path, "cpuset.cpus.effective", cg.cg_cpu.set, sizeof(cg.cg_cpu.set));
+	cg.cg_vmsize = cgroup_uint64(path, "memory.current");
+
+	return &cg;
+}
+
+static int cgroup_filter(const struct dirent *entry)
+{
+	/* Skip current dir ".", and prev dir "..", from list of files */
+	if ((1 == strlen(entry->d_name) && entry->d_name[0] == '.') ||
+	    (2 == strlen(entry->d_name) && !strcmp(entry->d_name, "..")))
+		return 0;
+
+	if (entry->d_name[0] == '.')
+		return 0;
+
+	if (entry->d_type != DT_DIR)
+		return 0;
+
+	return 1;
+}
+
+int cgroup_tree(char *path, char *pfx, int mode, int pos)
+{
+	struct dirent **namelist = NULL;
+	char s[32], r[32], l[32];
+	char row[ttcols + 9];		/* + control codes */
+	size_t rlen = sizeof(row) - 1;
+	size_t rplen = rlen - 9;
+	struct stat st;
+	struct cg *cg;
+	char buf[512];
+	int rc = 0;
+	FILE *fp;
+	int i, n;
+	int num;
+
+	if (pos >= ttrows)
+		return 0;
+
+	if (-1 == lstat(path, &st))
+		return 1;
+
+	if ((st.st_mode & S_IFMT) != S_IFDIR) {
+		errno = ENOTDIR;
+		return -1;
+	}
+
+	fp = fopenf("r", "%s/cgroup.procs", path);
+	if (!fp)
+		return -1;
+	num = 0;
+	while (fgets(buf, sizeof(buf), fp))
+		num++;
+
+	if (!pfx) {
+		pfx = "";
+		switch (mode) {
+		case 1:
+			cg = cg_stats(path);
+			snprintf(row, rplen, "\r %6.6s  %6.6s  %6.6s %5.1f %5.1f  %s",
+				 memsz(cg->cg_vmsize, s, sizeof(s)),
+				 memsz(cg->cg_rss,    r, sizeof(r)),
+				 memsz(cg->cg_vmlib,  l, sizeof(l)),
+				 cg->cg_memshare, cg->cg_load, path);
+			break;
+		case 2:
+			cg = cg_conf(path);
+			snprintf(row, rplen, "\r%6.6s [%-6.6s%6.6s] %6s [%-6.6s%6.6s] %s",
+				 memsz(cg->cg_vmsize, s, sizeof(s)),
+				 cg->cg_mem.min, cg->cg_mem.max, cg->cg_cpu.set,
+				 cg->cg_cpu.weight, cg->cg_cpu.max, path);
+			break;
+		default:
+			strlcpy(row, "\r", rplen);
+			strlcat(row, path, rplen);
+			break;
+		}
+
+		puts(row);
+	}
+
+	if (num > 0) {
+		rewind(fp);
+
+		i = 0;
+		while (fgets(buf, sizeof(buf), fp)) {
+			char comm[80] = { 0 };
+			pid_t pid;
+
+			pid = atoi(chomp(buf));
+			if (pid <= 0)
+				continue;
+
+			/* skip kernel threads for now (no cmdline) */
+			pid_comm(pid, comm, sizeof(comm));
+			if (pid_cmdline(pid, buf, sizeof(buf))) {
+				char proc[ttcols];
+				int len;
+
+				switch (mode) {
+				case 1:
+					snprintf(row, rplen, "\r%37s", " ");
+					break;
+				case 2:
+					snprintf(row, rplen, "\r --.-- [            ]        [            ] ");
+					break;
+				default:
+					strlcpy(row, "\r", rplen);
+					break;
+				}
+
+				strlcat(row, pfx, rplen);
+				strlcat(row, ++i == num ? END : FORK, rlen);
+
+				snprintf(proc, sizeof(proc), " %d %s %s", pid, comm, buf);
+
+				if (plain) {
+					strlcat(row, proc, rplen);
+				} else {
+					strlcat(row, CDIM, rlen);
+					strlcat(row, proc, rlen);
+
+					len = strlen(row) + strlen(CRST);
+					if (len > (int)rlen)
+						row[rlen - strlen(CRST)] = 0;
+
+					strlcat(row, CRST, sizeof(row));
+				}
+
+				puts(row);
+			}
+
+			if (mode == 1) {
+				pos += i;
+				if (pos >= ttrows)
+					break;
+			}
+		}
+	}
+
+	fclose(fp);
+
+	n = scandir(path, &namelist, cgroup_filter, alphasort);
+	if (n > 0) {
+		for (i = 0; i < n; i++) {
+			char *nm = namelist[i]->d_name;
+			char prefix[80];
+
+			snprintf(buf, sizeof(buf), "%s/%s", path, nm);
+			switch (mode) {
+			case 1:
+				cg = cg_stats(buf);
+				snprintf(row, rplen,
+					 "\r %6.6s  %6.6s  %6.6s %5.1f %5.1f  ",
+					 memsz(cg->cg_vmsize, s, sizeof(s)),
+					 memsz(cg->cg_rss,    r, sizeof(r)),
+					 memsz(cg->cg_vmlib,  l, sizeof(l)),
+					 cg->cg_memshare, cg->cg_load);
+				break;
+			case 2:
+				cg = cg_conf(buf);
+				snprintf(row, rplen, "\r%6.6s [%-6.6s%6.6s] %6.6s [%-6.6s%6.6s] ",
+					 memsz(cg->cg_vmsize, s, sizeof(s)),
+					 cg->cg_mem.min, cg->cg_mem.max,
+					 cg->cg_cpu.set, cg->cg_cpu.weight, cg->cg_cpu.max);
+				break;
+			default:
+				strlcpy(row, "\r", rplen);
+				break;
+			}
+
+			strlcat(row, pfx, rplen);
+			if (i + 1 == n) {
+				strlcat(row, END, rplen);
+				snprintf(prefix, sizeof(prefix), "%s   ", pfx);
+			} else {
+				strlcat(row, FORK, rplen);
+				snprintf(prefix, sizeof(prefix), "%s%s  ", pfx, PIPE);
+			}
+			strlcat(row, " ", rplen);
+
+			strlcat(row, nm,   rplen);
+			strlcat(row, "/ ", rplen);
+
+			puts(row);
+
+			rc += cgroup_tree(buf, prefix, mode, pos + i);
+
+			free(namelist[i]);
+		}
+
+		free(namelist);
+	}
+
+	return rc;
+}
+
+int show_cgps(char *arg)
+{
+	char path[512];
+
+	if (!arg)
+		arg = FINIT_CGPATH;
+	else if (arg[0] != '/') {
+		paste(path, sizeof(path), FINIT_CGPATH, arg);
+		arg = path;
+	}
+
+	return cgroup_tree(arg, NULL, 0, 0);
+}
+
+static void cgtop(uev_t *w, void *arg, int events)
+{
+	(void)w;
+	(void)events;
+
+	fputs("\e[2J\e[1;1H", stdout);
+	if (heading)
+		print_header(" VmSIZE     RSS   VmLIB  %%MEM  %%CPU  GROUP");
+	cgroup_tree(arg, NULL, 1, 0);
+}
+
+static void cleanup(void)
+{
+	ttcooked();
+	showcursor();
+	puts("");
+}
+
+static void leave(uev_t *w, void *arg, int events)
+{
+	(void)arg;
+	(void)events;
+
+	uev_exit(w->ctx);
+}
+
+static void key(uev_t *w, void *arg, int events)
+{
+	char ch;
+
+	(void)arg;
+	(void)events;
+
+	if (read(w->fd, &ch, sizeof(ch)) != -1) {
+		switch (ch) {
+		case 'q':
+			uev_exit(w->ctx);
+			break;
+
+		default:
+			dbg("Got char 0x%02x", ch);
+			break;
+		}
+	}
+}
+
+int show_cgtop(char *arg)
+{
+	struct sysinfo si = { 0 };
+        uev_t timer, input, sigint, sigterm, sigquit;
+	char path[512];
+        uev_ctx_t ctx;
+	int flags;
+
+	if (!arg)
+		arg = FINIT_CGPATH;
+	else if (arg[0] != '/') {
+		paste(path, sizeof(path), FINIT_CGPATH, arg);
+		arg = path;
+	}
+
+	if (!hcreate(ttrows + 25))
+		err(1, "failed creating hash table");
+
+	sysinfo(&si);
+	total_ram = si.totalram * si.mem_unit;
+
+        uev_init(&ctx);
+        uev_timer_init(&ctx, &timer, cgtop, arg, 1, ionce ? 0 : 1000);
+
+	if (!ionce && !plain) {
+		atexit(cleanup);
+		ttraw();
+		hidecursor();
+
+		flags = fcntl(STDIN_FILENO, F_GETFL);
+		if (flags != -1)
+			(void)fcntl(STDIN_FILENO, F_SETFL, flags | O_NONBLOCK);
+		(void)uev_io_init(&ctx, &input, key, NULL, STDIN_FILENO, UEV_READ);
+
+		(void)uev_signal_init(&ctx, &sigint, leave, NULL, SIGINT);
+		(void)uev_signal_init(&ctx, &sigterm, leave, NULL, SIGTERM);
+		(void)uev_signal_init(&ctx, &sigquit, leave, NULL, SIGQUIT);
+	}
+
+	return uev_run(&ctx, 0);
+}
+
+int show_cgroup(char *arg)
+{
+	char path[512];
+
+	if (!arg)
+		arg = FINIT_CGPATH;
+	else if (arg[0] != '/') {
+		paste(path, sizeof(path), FINIT_CGPATH, arg);
+		arg = path;
+	}
+
+	/* memory.current memory.min memory.max cpuset.cpus cpu.weight cpu.max */
+	if (heading)
+		print_header("   MEM [MIN      MAX]    CPU [WEIGHT   MAX] GROUP");
+
+	return cgroup_tree(arg, NULL, 2, 0);
+}
+
+/**
+ * Local Variables:
+ *  indent-tabs-mode: t
+ *  c-file-style: "linux"
+ * End:
+ */
@@ -0,0 +1,77 @@
+/* Cgroup utility functions for initctl
+ *
+ * Copyright (c) 2019-2021  Joachim Wiberg <troglobit@gmail.com>
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to deal
+ * in the Software without restriction, including without limitation the rights
+ * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+ * copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
+ * THE SOFTWARE.
+ */
+
+#ifndef FINIT_CGUTIL_H_
+#define FINIT_CGUTIL_H_
+
+#include <stdint.h>
+#include <stdlib.h>
+
+struct cg {
+	struct cg *cg_next;
+
+	/* stats */
+	char      *cg_path;		/* path in /sys/fs/cgroup   */
+	uint64_t   cg_prev;		/* cpuacct.usage            */
+	uint64_t   cg_rss;		/* memory.stat              */
+	uint64_t   cg_vmlib;		/* memory.stat              */
+	uint64_t   cg_vmsize;		/* memory.current           */
+	float      cg_memshare;		/* cg_rss / total_ram * 100 */
+	float      cg_load;		/* curr - prev / 10000000.0 */
+
+	/* config */
+	struct {
+		char min[32];
+		char max[32];
+	} cg_mem;
+
+	struct {
+		char set[16];
+		char weight[32];
+		char max[32];
+	} cg_cpu;
+};
+
+char *pid_cmdline (int pid, char *buf, size_t len);
+char *pid_comm    (int pid, char *buf, size_t len);
+char *pid_cgroup  (int pid, char *buf, size_t len);
+
+uint64_t cgroup_memory(char *group);
+
+struct cg *cg_stats(char *path);
+struct cg *cg_conf (char *path);
+
+int   cgroup_tree  (char *path, char *pfx, int mode, int pos);
+
+int   show_cgroup (char *arg);
+int   show_cgtop  (char *arg);
+int   show_cgps   (char *arg);
+
+#endif /* FINIT_CGUTIL_H_ */
+
+/**
+ * Local Variables:
+ *  indent-tabs-mode: t
+ *  c-file-style: "linux"
+ * End:
+ */
@@ -21,7 +21,9 @@
 * THE SOFTWARE.
 */

+#include <err.h>
 #include <errno.h>
+#include <poll.h>
 #include <sys/socket.h>
 #include <sys/un.h>

@@ -36,19 +38,14 @@ int client_connect(void)
 		.sun_path   = INIT_SOCKET,
 	};

-	sd = socket(AF_UNIX, SOCK_STREAM, 0);
+	sd = socket(AF_UNIX, SOCK_SEQPACKET, 0);
 	if (-1 == sd)
-		goto error;
+		err(1, "Failed creating UNIX domain socket");

-	if (connect(sd, (struct sockaddr*)&sun, sizeof(sun)) == -1) {
-		close(sd);
-		goto error;
-	}
+	if (connect(sd, (struct sockaddr*)&sun, sizeof(sun)) == -1)
+		err(1, "Failed connecting to finit");

 	return sd;
-error:
-	perror("Failed connecting to finit");
-	return -1;
 }

 int client_disconnect(void)
@@ -68,27 +65,48 @@ int client_disconnect(void)

 int client_send(struct init_request *rq, ssize_t len)
 {
+	struct pollfd pfd = { 0 };
 	int sd, result = 255;
+	int rc;

 	sd = client_connect();
 	if (-1 == sd)
 		return -1;

-	if (write(sd, rq, len) != len)
-		goto error;
+	pfd.fd     = sd;
+	pfd.events = POLLOUT;
+	if (poll(&pfd, 1, 2000) <= 0) {
+		warn("Timed out waiting for Finit, errno %d", errno);
+		goto exit;
+	}

-	if (read(sd, rq, len) != len)
-		goto error;
+	if (write(sd, rq, len) != len) {
+		warn("Failed communicating with Finit, errno %d", errno);
+		goto exit;
+	}
+
+	pfd.fd = sd;
+	pfd.events = POLLIN;
+	if ((rc = poll(&pfd, 1, 2000)) <= 0) {
+		if (rc)
+			warn("poll(), errno %d", errno);
+		else
+			warnx("Timed out waiting for reply from Finit.");
+		goto exit;
+	}
+
+	if (read(sd, rq, len) != len) {
+		warn("Failed reading reply from Finit, errno %d", errno);
+		goto exit;
+	}

 	if (rq->cmd == INIT_CMD_NACK)
 		result = 1;
 	else
 		result = 0;
-	goto exit;
-error:
-	perror("Failed communicating with finit");
 exit:
 	client_disconnect();
+
 	return result;
 }

@@ -99,13 +99,14 @@ static void cond_bump_reconf(void)
 	unsigned int rgen;

 	/*
-	 * If %COND_RECONF does not exist, cond_get_gen() returns 0
+	 * If %_PATH_RECONF does not exist, cond_get_gen() returns 0
 	 * meaning that rgen++ is always what we want.
 	 */
-	rgen = cond_get_gen(COND_RECONF);
+	rgen = cond_get_gen(_PATH_RECONF);
 	rgen++;

-	cond_set_gen(COND_RECONF, rgen);
+	if (cond_set_gen(_PATH_RECONF, rgen))
+		_pe("Failed setting %s to gen %d", _PATH_RECONF, rgen);
 }

 static int cond_checkpath(const char *path)
@@ -119,7 +120,7 @@ static int cond_checkpath(const char *path)
 		return 1;
 	}

-	if (makepath(dir) && errno != EEXIST) {
+	if (mkpath(dir, 0755) && errno != EEXIST) {
 		_pe("Failed creating dir '%s' for condition '%s'", dir, path);
 		return 1;
 	}
@@ -127,22 +128,22 @@ static int cond_checkpath(const char *path)
 	return 0;
 }

-int cond_set_path(const char *path, enum cond_state new)
+int cond_set_path(const char *path, enum cond_state next)
 {
-	enum cond_state old;
+	enum cond_state prev;
 	unsigned int rgen;

-	_d("%s", path);
+	_d("%s <= %d", path, next);

-	rgen = cond_get_gen(COND_RECONF);
+	rgen = cond_get_gen(_PATH_RECONF);
 	if (!rgen) {
 		_e("Unable to read configuration generation (%s)", path);
 		return -1;
 	}

-	old = cond_get_path(path);
+	prev = cond_get_path(path);

-	switch (new) {
+	switch (next) {
 	case COND_ON:
 		if (cond_checkpath(path))
 		    return 0;
@@ -159,59 +160,95 @@ int cond_set_path(const char *path, enum cond_state new)
 		return 0;
 	}

-	return new != old;
+	return next != prev;
 }

-static void cond_update(const char *name)
+/* Should only be used by cond_set*(), cond_clear(), and usr/sys plugins! */
+int cond_update(const char *name)
 {
 	svc_t *svc, *iter = NULL;
+	int affects = 0;

 	_d("%s", name);
 	for (svc = svc_iterator(&iter, 1); svc; svc = svc_iterator(&iter, 0)) {
 		if (!svc_has_cond(svc) || !cond_affects(name, svc->cond))
 			continue;

+		affects++;
 		_d("%s: match <%s> %s(%s)", name ?: "nil", svc->cond, svc->desc, svc->cmd);
 		service_step(svc);
 	}
+
+	return affects;
+}
+
+int cond_set_noupdate(const char *name)
+{
+	_d("%s", name);
+	if (string_compare(name, "nop"))
+		return 1;
+
+	if (!cond_set_path(cond_path(name), COND_ON))
+		return 1;
+
+	return 0;
 }

 void cond_set(const char *name)
 {
 	_d("%s", name);
-	if (string_compare(name, "nop"))
-		return;
-
-	if (!cond_set_path(cond_path(name), COND_ON))
+	if (cond_set_noupdate(name))
 		return;

 	cond_update(name);
 }

-void cond_set_oneshot(const char *name)
+int cond_set_oneshot_noupdate(const char *name)
 {
 	const char *path;

 	if (string_compare(name, "nop"))
-		return;
+		return 1;

 	path = cond_path(name);
 	_d("%s => %s", name, path);

 	if (cond_checkpath(path))
+		return 1;
+
+	if (symlink(_PATH_RECONF, path) && errno != EEXIST) {
+		_pe("Failed creating onshot cond %s", name);
+		return 1;
+	}
+
+	return 0;
+}
+
+void cond_set_oneshot(const char *name)
+{
+	_d("%s", name);
+	if (cond_set_oneshot_noupdate(name))
 		return;

-	symlink(COND_RECONF, path);
 	cond_update(name);
 }

+int cond_clear_noupdate(const char *name)
+{
+	_d("%s", name);
+	if (string_compare(name, "nop"))
+		return 1;
+
+	if (!cond_set_path(cond_path(name), COND_OFF))
+		return 1;
+
+	return 0;
+}
+
 void cond_clear(const char *name)
 {
 	_d("%s", name);
-	if (string_compare(name, "nop"))
-		return;
-
-	if (!cond_set_path(cond_path(name), COND_OFF))
+	if (cond_clear_noupdate(name))
 		return;

 	cond_update(name);
@@ -224,7 +261,7 @@ void cond_reload(void)
 	cond_bump_reconf();
 }

-static int reassert(const char *fpath, const struct stat *sb, int tflg, struct FTW *ftw)
+static int do_assert(const char *fpath, const struct stat *sb, int tflg, struct FTW *ftw, int set)
 {
 	char *nm;

@@ -234,35 +271,57 @@ static int reassert(const char *fpath, const struct stat *sb, int tflg, struct F
 	if (tflg != FTW_F)
 		return 0;

-	nm = strstr((char *)fpath, COND_DIR);
+	nm = strstr((char *)fpath, COND_BASE);
 	if (!nm) {
-		_e("Incorrect condtion path %s, cannot reassert", fpath);
+		_e("Incorrect condtion path %s, cannot %sassert", fpath, set ? "re" : "de");
 		return 1;
 	}

-	nm += sizeof(COND_DIR);
-	_d("Reasserting %s => %s", fpath, nm);
-	cond_set(nm);
+	nm += strlen(COND_BASE);
+	_d("%sasserting %s => %s", set ? "Re" : "De", fpath, nm);
+	if (set)
+		cond_set(nm);
+	else
+		cond_clear_noupdate(nm); /* important, see netlink plugin! */

 	return 0;
 }

+static int reassert(const char *fpath, const struct stat *sb, int tflg, struct FTW *ftw)
+{
+	return do_assert(fpath, sb, tflg, ftw, 1);
+}
+
+static int deassert(const char *fpath, const struct stat *sb, int tflg, struct FTW *ftw)
+{
+	return do_assert(fpath, sb, tflg, ftw, 0);
+}
+
 /*
 * Used only by netlink plugin atm.
 * type: is a one of pid/, net/, etc.
 */
-void cond_reassert(const char *type)
+void cond_reassert(const char *pat)
 {
-	_d("%s", type);
-	nftw(cond_path(type), reassert, 20, FTW_DEPTH);
+	_d("%s", pat);
+	nftw(cond_path(pat), reassert, 20, FTW_DEPTH);
+}
+
+/*
+ * Used only by netlink plugin atm.
+ */
+void cond_deassert(const char *pat)
+{
+	_d("%s", pat);
+	nftw(cond_path(pat), deassert, 20, FTW_DEPTH);
 }

 void cond_init(void)
 {
 	char path[MAX_ARG_LEN];

-	if (makepath(pid_runpath(COND_PATH, path, sizeof(path))) && errno != EEXIST) {
-		_pe("Failed creating condition base directory '%s'", COND_PATH);
+	if (mkpath(pid_runpath(_PATH_COND, path, sizeof(path)), 0755) && errno != EEXIST) {
+		_pe("Failed creating condition base directory '%s'", _PATH_COND);
 		return;
 	}

@@ -45,15 +45,15 @@ const char *cond_path(const char *name)
 	static char path[256];
 	char tmp[MAX_ARG_LEN];

-	snprintf(tmp, sizeof(tmp), COND_PATH "/%s", name);
+	snprintf(tmp, sizeof(tmp), _PATH_COND "%s", name);

 	return pid_runpath(tmp, path, sizeof(path));
 }

 unsigned int cond_get_gen(const char *file)
 {
-	unsigned int gen;
 	char *ptr, path[256];
+	unsigned int gen;
 	FILE *fp;
 	int ret;

@@ -74,14 +74,14 @@ enum cond_state cond_get_path(const char *path)
 {
 	int cgen, rgen;

-	rgen = cond_get_gen(COND_RECONF);
-	if (!rgen)
-		return COND_OFF;
-
 	cgen = cond_get_gen(path);
 	if (!cgen)
 		return COND_OFF;

+	rgen = cond_get_gen(_PATH_RECONF);
+	if (!rgen)
+		return COND_OFF;
+
 	return (cgen == rgen) ? COND_ON : COND_FLUX;
 }

@@ -4,9 +4,16 @@
 #include <paths.h>
 #include <svc.h>

-#define COND_DIR      "finit/cond"
-#define COND_PATH     _PATH_VARRUN COND_DIR
-#define COND_RECONF   COND_PATH "/reconf"
+#define COND_BASE      "finit/cond"
+#define COND_PID       "pid/"
+#define COND_SYS       "sys/"
+#define COND_USR       "usr/"
+
+#define _PATH_COND     _PATH_VARRUN COND_BASE "/"
+#define _PATH_CONDPID  _PATH_COND   COND_PID
+#define _PATH_CONDSYS  _PATH_COND   COND_SYS
+#define _PATH_CONDUSR  _PATH_COND   COND_USR
+#define _PATH_RECONF   _PATH_COND   "reconf"

 typedef enum cond_state {
 	COND_OFF = 0,
@@ -23,12 +30,20 @@ enum cond_state cond_get     (const char *name);
 enum cond_state cond_get_agg (const char *names);
 int             cond_affects (const char *name, const char *names);

+int  cond_update      (const char *name);
 int  cond_set_path    (const char *path, enum cond_state new);
 void cond_set         (const char *name);
 void cond_set_oneshot (const char *name);
 void cond_clear       (const char *name);
 void cond_reload      (void);
+
+int cond_set_noupdate (const char *name);
+int cond_set_oneshot_noupdate(const char *name);
+int cond_clear_noupdate(const char *name);
+
 void cond_reassert    (const char *pat);
+void cond_deassert    (const char *pat);
+
 void cond_init        (void);

 #endif	/* FINIT_COND_H_ */
@@ -23,6 +23,7 @@

 #include "config.h"		/* Generated by configure script */

+#include <ctype.h>
 #include <dirent.h>
 #include <string.h>
 #include <sys/inotify.h>
@@ -50,6 +51,8 @@ int logfile_count_max = 5;
 struct rlimit initial_rlimit[RLIMIT_NLIMITS];
 struct rlimit global_rlimit[RLIMIT_NLIMITS];

+char cgroup_current[16]; /* cgroup.NAME sets current cgroup for a set of services */
+
 struct conf_change {
 	TAILQ_ENTRY(conf_change) link;
 	char *name;
@@ -63,22 +66,6 @@ static TAILQ_HEAD(head, conf_change) conf_change_list = TAILQ_HEAD_INITIALIZER(c
 static int  parse_conf(char *file, int is_rcsd);
 static void drop_changes(void);

-static void hide_args(int argc, char *argv[])
-{
-	/*
-	 * Hide command line arguments from ps (in particular for
-	 * forked children that don't execv()).  This is an ugly
-	 * hack that only works on Linux.
-	 * https://web.archive.org/web/20110227041321/http://netsplit.com/2007/01/10/hiding-arguments-from-ps/
-	 */
-	if (argc > 1) {
-		char *arg_end;
-
-		arg_end = argv[argc-1] + strlen (argv[argc-1]);
-		*arg_end = ' ';
-	}
-}
-
 int get_bool(char *arg, int default_value)
 {
 	if (!arg)
@@ -142,24 +129,27 @@ static void parse_arg(char *arg)
 		single = 1;
 }

-void conf_parse_cmdline(int argc, char *argv[])
+#ifdef KERNEL_CMDLINE
+/*
+ * Parse /proc/cmdline to find args for init.  Don't use this!
+ *
+ * Instead, rely on the kernel to give Finit its arguments as
+ * regular argc + argv[].  Only use this if the system you run
+ * on has a broken initramfs system that cannot forward args
+ * to Finit properly.
+ */
+static void parse_kernel_cmdline(void)
 {
-	int dbg = 0;
-	FILE *fp;
 	char line[LINE_SIZE], *cmdline, *tok;
-
-	for (int i = 1; i < argc; i++)
-		parse_arg(argv[i]);
-
-	hide_args(argc, argv);
+	FILE *fp;

 	fp = fopen("/proc/cmdline", "r");
 	if (!fp)
-		goto done;
+		return;

 	if (!fgets(line, sizeof(line), fp)) {
 		fclose(fp);
-		goto done;
+		return;
 	}

 	cmdline = chomp(line);
@@ -170,8 +160,23 @@ void conf_parse_cmdline(int argc, char *argv[])
 		parse_arg(tok);
 	}
 	fclose(fp);
+}
+#else
+#define parse_kernel_cmdline()
+#endif
+
+/*
+ * Kernel gives us all non-kernel options on our cmdline
+ */
+void conf_parse_cmdline(int argc, char *argv[])
+{
+	int dbg = 0;
+
+	for (int i = 1; i < argc; i++)
+		parse_arg(argv[i]);
+
+	parse_kernel_cmdline();

-done:
 	log_init(dbg);
 }

@@ -352,7 +357,7 @@ char *lim2str(struct rlimit *rlim)
 	char tmp[25];
 	static char buf[42];

-	memset(buf, 0, sizeof(buf));
+	buf[0] = 0;
 	if (RLIM_INFINITY == rlim->rlim_cur)
 		snprintf(tmp, sizeof(tmp), "unlimited, ");
 	else
@@ -423,6 +428,29 @@ error:
 	logit(LOG_WARNING, "rlimit: parse error");
 }

+/* cgroup NAME ctrl.prop:value,ctrl.prop:value ... */
+static void conf_parse_cgroup(char *line)
+{
+	char config[strlen(line) + 1];
+	char *ptr, *name;
+
+	name = strtok(line, " \t");
+	if (!name)
+		return;
+
+	if (strstr(name, "..") || strchr(name, '/'))
+		return;		/* illegal */
+
+	config[0] = 0;
+	while ((ptr = strtok(NULL, " \t"))) {
+		if (config[0])
+			strlcat(config, ",", sizeof(config));
+		strlcat(config, ptr, sizeof(config));
+	}
+
+	cgroup_add(name, config, 0);
+}
+
 static void parse_static(char *line, int is_rcsd)
 {
 	char cmd[CMD_SIZE];
@@ -552,9 +580,21 @@ static void parse_dynamic(char *line, struct rlimit rlimit[], char *file)
 		return;
 	}

+	/* Read contrl group limits */
+	if (MATCH_CMD(line, "cgroup ", x)) {
+		conf_parse_cgroup(x);
+		return;
+	}
+
+	/* Set current cgroup for the following services/run/tasks */
+	if (MATCH_CMD(line, "cgroup.", x)) {
+		strlcpy(cgroup_current, x, sizeof(cgroup_current));
+		return;
+	}
+
 	/* Regular or serial TTYs to run getty */
 	if (MATCH_CMD(line, "tty ", x)) {
-		tty_register(strip_line(x), rlimit, file);
+		service_register(SVC_TYPE_TTY, strip_line(x), rlimit, file);
 		return;
 	}
 }
@@ -580,9 +620,11 @@ static int parse_conf(char *file, int is_rcsd)
 	if (!fp)
 		return 1;

-	/* Prepare default limits for each service in /etc/finit.d/ */
-	if (is_rcsd)
+	/* Prepare default limits and group for each service in /etc/finit.d/ */
+	if (is_rcsd) {
 		memcpy(rlimit, global_rlimit, sizeof(rlimit));
+		cgroup_current[0] = 0;
+	}

 	_d("*** Parsing %s", file);
 	while (!feof(fp)) {
@@ -622,8 +664,8 @@ int conf_reload(void)
 	   daylight, timezone, tzname[0], tzname[1]);

 	/* Mark and sweep */
+	cgroup_mark_all();
 	svc_mark_dynamic();
-	tty_mark();

 	/*
 	 * Reset global rlimit to bootstrap values from conf_init().
@@ -632,12 +674,12 @@ int conf_reload(void)

 	if (rescue) {
 		int rc;
-		char line[80] = "tty [12345] @console noclear nologin";
+		char line[80] = "tty [12345] rescue";

 		/* If rescue.conf is missing, fall back to a root shell */
 		rc = parse_conf(RESCUE_CONF, 0);
 		if (rc)
-			tty_register(line, global_rlimit, NULL);
+			service_register(SVC_TYPE_TTY, line, global_rlimit, NULL);

 		print(rc, "Entering rescue mode");
 		goto done;
@@ -677,7 +719,7 @@ int conf_reload(void)

 		/* Check for dangling symlinks */
 		if (S_ISLNK(st.st_mode)) {
-			path = rp = realpath(path, NULL);
+			rp = realpath(path, NULL);
 			if (!rp) {
 				logit(LOG_WARNING, "Skipping %s, dangling symlink: %s", path, strerror(errno));
 				continue;
@@ -687,7 +729,7 @@ int conf_reload(void)
 		/* Check that file ends with '.conf' */
 		len = strlen(path);
 		if (len < 6 || strcmp(&path[len - 5], ".conf"))
-			_d("Skipping %s, not a valid .conf ... ", path);
+			_d("Skipping %s, not a Finit .conf file ... ", path);
 		else
 			parse_conf(path, 1);

@@ -699,7 +741,13 @@ int conf_reload(void)

 	/* Mark any reverse deps as chenaged. */
 	service_update_rdeps();
+
+	/* Set up top-level cgroups */
+	cgroup_config();
 done:
+	/* Remove all unused top-level cgroups */
+	cgroup_cleanup();
+
 	/* Drop record of all .conf changes */
 	drop_changes();

@@ -792,13 +840,21 @@ int conf_any_change(void)

 int conf_changed(char *file)
 {
+	int rc = 0;
+	char *rp;
+
 	if (!file)
 		return 0;

-	if (conf_find(file))
-		return 1;
+	rp = realpath(file, NULL);
+	if (!rp)
+		return 0;

-	return 0;
+	if (conf_find(rp))
+		rc = 1;
+	free(rp);
+
+	return rc;
 }

 static void conf_cb(uev_t *w, void *arg, int events)
@@ -880,7 +936,7 @@ int conf_monitor(void)
 }

 /*
- * Prepare .conf parser and load all .conf files
+ * Prepare .conf parser and load /etc/finit.conf for global settings
 */
 int conf_init(uev_ctx_t *ctx)
 {
@@ -900,6 +956,12 @@ int conf_init(uev_ctx_t *ctx)
 			      rlim2str(i), strerror(errno));
 	}

+	/* Initialize global rlimits, e.g. for built-in services */
+	memcpy(global_rlimit, initial_rlimit, sizeof(global_rlimit));
+
+	/* Read global rlimits and global cgroup setup from /etc/finit.conf */
+	parse_conf(FINIT_CONF, 0);
+
 	/* prepare /etc watcher */
 	fd = iwatch_init(&iw_conf);
 	if (fd < 0)
@@ -24,12 +24,14 @@
 #ifndef FINIT_CONF_H_
 #define FINIT_CONF_H_

+#include "cgroup.h"
 #include "svc.h"

 extern int logfile_size_max;
 extern int logfile_count_max;

 extern struct rlimit global_rlimit[];
+extern char cgroup_current[];

 int   str2rlim(char *str);
 char *rlim2str(int rlim);
@@ -26,6 +26,7 @@

 #include <ctype.h>		/* isdigit() */
 #include <dirent.h>
+#include <err.h>
 #include <stdarg.h>
 #include <sys/ioctl.h>
 #include <sys/prctl.h>
@@ -33,8 +34,10 @@
 #include <sys/ttydefaults.h>	/* Not included by default in musl libc */
 #include <termios.h>
 #include <lite/lite.h>
+#include <linux/vt.h>

 #include "finit.h"
+#include "cgroup.h"
 #include "conf.h"
 #include "helpers.h"
 #include "sig.h"
@@ -105,8 +108,6 @@ int run(char *cmd)
 	if (0 == pid) {
 		FILE *fp;

-		/* Reset signal handlers that were set by the parent process */
-		sig_unblock();
 		setsid();

 		/* Always redirect stdio for run() */
@@ -239,14 +240,8 @@ int exec_runtask(char *cmd, char *args[])

 static void prepare_tty(char *tty, speed_t speed, char *procname, struct rlimit rlimit[])
 {
-	struct sigaction sa;
-	struct termios term;
 	char name[80];
-	int fd;
-
-	/* Detach from initial controlling TTY and become session leader */
-	vhangup();
-	setsid();
+	int fd, dummy;

 	fd = open(tty, O_RDWR);
 	if (fd < 0) {
@@ -257,40 +252,18 @@ static void prepare_tty(char *tty, speed_t speed, char *procname, struct rlimit
 	dup2(fd, STDIN_FILENO);
 	dup2(fd, STDOUT_FILENO);
 	dup2(fd, STDERR_FILENO);
-	close(fd);

-	if (ioctl(STDIN_FILENO, TIOCSCTTY, 1) < 0)
-		logit(LOG_WARNING, "Failed TIOCSCTTY on %s: %s", tty, strerror(errno));
+	/* Set default TERM, run_getty() below allows override */
+	if (ioctl(fd, VT_OPENQRY, &dummy) == -1)
+		setenv("TERM", "vt102", 1); /* likely a serial line */
+	else
+		setenv("TERM", "linux", 1);

 	/*
 	 * Reset to sane defaults in case of messup from prev. session
 	 */
-	stty(STDIN_FILENO, speed);
-
-	/*
-	 * Disable ISIG (INTR, QUIT, SUSP) before handing over to getty.
-	 * It is up to the getty process to allow them again.
-	 */
-	if (!tcgetattr(STDIN_FILENO, &term)) {
-		term.c_lflag    &= ~ISIG;
-		term.c_cc[VEOF]  = _POSIX_VDISABLE;
-		term.c_cc[VINTR] = _POSIX_VDISABLE;
-		tcsetattr(STDIN_FILENO, TCSAFLUSH, &term);
-	}
-
-	/* Reset signal handlers that were set by the parent process */
-	sig_unblock();
-
-	/*
-	 * Ignore a few signals, needed to prevent Ctrl-C at login:
-	 * prompt and to prevent QUIT from dumping core.
-	 */
-	sigemptyset(&sa.sa_mask);
-	sa.sa_flags   = SA_RESTART;
-	sa.sa_handler = SIG_IGN;
-	sigaction(SIGHUP,  &sa, NULL);
-	sigaction(SIGINT,  &sa, NULL);
-	sigaction(SIGQUIT, &sa, NULL);
+	stty(fd, speed);
+	close(fd);

 	/* Set configured limits */
 	for (int i = 0; i < RLIMIT_NLIMITS; i++) {
@@ -308,14 +281,18 @@ static void prepare_tty(char *tty, speed_t speed, char *procname, struct rlimit

 static int activate_console(int noclear, int nowait)
 {
+	static const char msg[] = "\nPlease press Enter to activate this console.";
+	static const char clr[] = "\r\e[2K";
+	static const char cup[] = "\e[A";
 	struct termios orig;
-	int ret = 0;
+	char c;
+	int rc;

 	if (nowait || rescue)
 		return 1;

 	if (!noclear)
-		(void)write(STDERR_FILENO, "\e[r\e[H\e[J", 9);
+		dprint(STDERR_FILENO, "\e[r\e[H\e[J", 9);

 	/* Disable ECHO, XON/OFF while waiting for <CR> */
 	if (!tcgetattr(STDIN_FILENO, &orig)) {
@@ -328,36 +305,24 @@ static int activate_console(int noclear, int nowait)
 		tcsetattr(STDIN_FILENO, TCSAFLUSH, &c);
 	}

-	while (!fexist(SYNC_SHUTDOWN)) {
-		char c;
-		static const char clr[] = "\r\e[2K";
-		static const char cup[] = "\e[A";
-		static const char msg[] = "\nPlease press Enter to activate this console.";
+	dprint(STDERR_FILENO, clr, strlen(clr));
+	dprint(STDERR_FILENO, msg, strlen(msg));
+	while ((rc = read(STDIN_FILENO, &c, 1)) > 0 && c != '\r' && c != '\n')
+		continue;

-		if (fexist(SYNC_STOPPED)) {
-			sleep(5);
-			continue;
-		}
+	/* On any error (likely EINTR), we avoid starting getty */
+	if (rc == -1)
+		rc = 0;

-		(void)write(STDERR_FILENO, clr, strlen(clr));
-		(void)write(STDERR_FILENO, msg, strlen(msg));
-		while (read(STDIN_FILENO, &c, 1) == 1 && c != '\r')
-			continue;
-
-		if (fexist(SYNC_STOPPED))
-			continue;
-
-		(void)write(STDERR_FILENO, clr, strlen(clr));
-		(void)write(STDERR_FILENO, cup, strlen(cup));
-		ret = 1;
-		break;
-	}
+	/* Clear msg and move cursor up for next message/login: */
+	dprint(STDERR_FILENO, clr, strlen(clr));
+	dprint(STDERR_FILENO, cup, strlen(cup));

 	/* Restore TTY */
 	if (tcsetattr(STDIN_FILENO, TCSAFLUSH, &orig) == -1)
-		ret = 0;
+		rc = 0;		/* TTY in bad shape, restart */

-	return ret;
+	return rc;
 }

 /*
@@ -379,64 +344,57 @@ static int activate_console(int noclear, int nowait)
 * since /bin/login usually only disables ECHO until a password line has
 * been entered.  Upon starting the user's $SHELL the ISIG flag is reset
 */
-pid_t run_getty(char *tty, char *baud, char *term, int noclear, int nowait, struct rlimit rlimit[])
+pid_t run_getty(char *tty, char *cmd, char *args[], int noclear, int nowait, struct rlimit rlimit[])
 {
-	pid_t pid;
+	int rc = 1;

-	pid = fork();
-	if (!pid) {
-		speed_t speed;
-		int rc = 1;
-
-		speed = stty_parse_speed(baud);
-		logit(LOG_INFO, "Starting built-in getty on %s, speed %u", tty, speed);
-		prepare_tty(tty, speed, "tty", rlimit);
-		if (activate_console(noclear, nowait))
-			rc = getty(tty, speed, term, NULL);
-
-		_exit(rc);
+	/* Dunno speed, tell stty() to not mess with it */
+	prepare_tty(tty, B0, "getty", rlimit);
+	if (activate_console(noclear, nowait)) {
+		logit(LOG_INFO, "Starting external getty on %s", tty);
+		rc = execv(cmd, args);
 	}

-	return pid;
+	vhangup();
+
+	return rc;
 }

-pid_t run_getty2(char *tty, char *cmd, char *args[], int noclear, int nowait, struct rlimit rlimit[])
+int sh(char *tty)
 {
-	pid_t pid;
+	char *args[2] = {
+		NULL,
+		NULL
+	};
+	char *arg0;
+	size_t len;

-	pid = fork();
-	if (!pid) {
-		int rc = 1;
+	/* The getty process is usually responsible for the UTMP login record */
+	utmp_set_login(tty, NULL);

-		/* Dunno speed, tell stty() to not mess with it */
-		logit(LOG_INFO, "Starting external getty on %s, speed %u", tty, B0);
-		prepare_tty(tty, B0, "getty", rlimit);
-		if (activate_console(noclear, nowait))
-			rc = execv(cmd, args);
+	/* Start /bin/sh as a login shell, i.e. with a prefix '-' */
+	len = strlen(_PATH_BSHELL) + 2;
+	arg0 = malloc(len);
+	if (!arg0)
+		err(1, "Failed allocating memory");
+	snprintf(arg0, len, "-%s", _PATH_BSHELL);
+	args[0] = arg0;

-		vhangup();
-		_exit(rc);
-	}
+	/* Unblock signals inherited from parent */
+	sig_unblock();

-	return pid;
+	return execv(_PATH_BSHELL, args);
 }

 pid_t run_sh(char *tty, int noclear, int nowait, struct rlimit rlimit[])
 {
-	pid_t pid;
+	int rc = 1;

-	pid = fork();
-	if (!pid) {
-		int rc = 1;
+	prepare_tty(tty, B0, "finitsh", rlimit);
+	if (activate_console(noclear, nowait))
+		rc = sh(tty);

-		prepare_tty(tty, B0, "finit-sh", rlimit);
-		if (activate_console(noclear, nowait))
-			rc = sh(tty);
-
-		_exit(rc);
-	}
-
-	return pid;
+	return rc;
 }

 int run_parts(char *dir, char *cmd)
@@ -52,6 +52,7 @@
 #include "util.h"
 #include "utmp-api.h"
 #include "schedule.h"
+#include "watchdog.h"

 int   runlevel  = 0;		/* Bootstrap 'S' */
 int   cfglevel  = RUNLEVEL;	/* Fallback if no configured runlevel */
@@ -59,11 +60,13 @@ int   prevlevel = -1;
 int   debug     = 0;		/* debug mode from kernel cmdline */
 int   rescue    = 0;		/* rescue mode from kernel cmdline */
 int   single    = 0;		/* single user mode from kernel cmdline */
+int   bootstrap = 1;		/* set while bootrapping (for TTYs) */
 char *sdown     = NULL;
 char *network   = NULL;
 char *hostname  = NULL;
 char *rcsd      = FINIT_RCSD;
 char *runparts  = NULL;
+char *osheading = NULL;

 uev_ctx_t *ctx  = NULL;		/* Main loop context */
 svc_t *wdog     = NULL;		/* No watchdog by default */
@@ -88,7 +91,14 @@ static void banner(void)
 	 */
 	plugin_run_hooks(HOOK_BANNER);

-	print_banner(INIT_HEADING);
+#ifdef INIT_OSHEADING
+	osheading = INIT_OSHEADING;
+	if (osheading) {
+		if (!osheading[0])
+			osheading = release_heading();
+		print_banner(osheading);
+	}
+#endif
 }

 /*
@@ -147,6 +157,25 @@ static int fsck_all(void)
 }

 #ifndef SYSROOT
+/* If / is not listed in fstab, or listed as 'ro', leave it alone */
+static int fs_readonly_root(struct fstab *fs)
+{
+	char *tok, *str;
+
+	if (!fs)
+		return 1;
+
+	str = fs->fs_mntops;
+	while ((tok = strtok(str, ","))) {
+		if (!strcmp(tok, "ro"))
+			return 1;
+
+		str = NULL;
+	}
+
+	return 0;
+}
+
 static void fs_remount_root(int fsckerr)
 {
 	struct fstab *fs;
@@ -159,8 +188,7 @@ static void fs_remount_root(int fsckerr)
 			break;
 	}

-	/* If / is not listed in fstab, or listed as 'ro', leave it alone */
-	if (!fs || !strcmp(fs->fs_type, "ro"))
+	if (fs_readonly_root(fs))
 		goto out;

 	if (fsckerr)
@@ -184,6 +212,69 @@ static void fs_remount_root(int fsckerr)
 }
 #endif	/* SYSROOT */

+/*
+ * Opinionated file system setup.  Checks for critical mount points and
+ * mounts them as most users expect.  All file systems are checked with
+ * /proc/mounts before mounting.
+ *
+ * Embedded systems, and other people who want full control, can set up
+ * their system with /etc/fstab, which is handled before this function
+ * is called.  For systems like Debian/Ubuntu, who only have / and swap
+ * in their /etc/fstab, this function does all the magic necessary.
+ */
+static void fs_finalize(void)
+{
+	/*
+	 * Some systems rely on us to both create /dev/shm and, to mount
+	 * a tmpfs there.  Any system with dbus needs shared memory, so
+	 * mount it, unless its already mounted, but not if listed in
+	 * the /etc/fstab file already.
+	 */
+	if (!fismnt("/dev/shm")) {
+		makedir("/dev/shm", 0777);
+		mount("shm", "/dev/shm", "tmpfs", 0, "mode=0777");
+	}
+
+	/* Modern systems use /dev/pts */
+	if (!fismnt("/dev/pts")) {
+		char opts[32];
+		int mode;
+		int gid;
+
+		gid = getgroup("tty");
+		if (gid == -1)
+			gid = 0;
+
+		/* 0600 is default on Debian, use 0620 to get mesg y by default */
+		mode = 0620;
+		snprintf(opts, sizeof(opts), "gid=%d,mode=%d,ptmxmode=0666", gid, mode);
+
+		makedir("/dev/pts", 0755);
+		mount("devpts", "/dev/pts", "devpts", MS_NOSUID | MS_NOEXEC, opts);
+	}
+
+	/*
+	 * Modern systems use tmpfs for /run.  Fallback to /var/run if
+	 * /run doesn't exist is handled by the bootmisc plugin.  It
+	 * also sets up compat symlinks.
+	 *
+	 * The unconditional mount of /run/lock is for DoS prevention.
+	 * To override any of this behavior, add entries to /etc/fstab
+	 * for /run (and optionally /run/lock).
+	 */
+	if (fisdir("/run") && !fismnt("/run")) {
+		mount("tmpfs", "/run", "tmpfs", MS_NOSUID | MS_NODEV | MS_NOEXEC | MS_RELATIME, "mode=0755,size=10%");
+
+		/* This prevents user DoS of /run by filling /run/lock at the expense of another tmpfs, max 5MiB */
+		makedir("/run/lock", 1777);
+		mount("tmpfs", "/run/lock", "tmpfs", MS_NOSUID | MS_NODEV | MS_NOEXEC | MS_RELATIME, "mode=0777,size=5252880");
+	}
+
+	/* Modern systems use tmpfs for /tmp */
+	if (!fismnt("/tmp"))
+		mount("tmpfs", "/tmp", "tmpfs", MS_NOSUID | MS_NODEV, "mode=1777");
+}
+
 static void fs_mount(void)
 {
 	if (!rescue)
@@ -199,7 +290,9 @@ static void fs_mount(void)
 	plugin_run_hooks(HOOK_MOUNT_POST);

 	run("swapon -ea");
-	umask(0022);
+
+	_d("Finalize, ensure common file systems are available ...");
+	fs_finalize();
 }

 /*
@@ -222,57 +315,27 @@ static void fs_init(void)
 	};
 	size_t i;

+	/* mask writable bit for g and o */
+	umask(022);
+
 	for (i = 0; i < NELEMS(fs); i++) {
 		int rc;

+		/*
+		 * Check if already mounted, we may be running in a
+		 * container, or an initramfs ran before us.  The
+		 * function fismnt() reliles on /proc/mounts being
+		 * unique for each chroot/container.
+		 */
+		if (fismnt(fs[i].file))
+			continue;
+
 		rc = mount(fs[i].spec, fs[i].file, fs[i].type, 0, NULL);
 		if (rc && errno != EBUSY)
 			_pe("Failed mounting %s on %s", fs[i].spec, fs[i].file);
 	}
 }

-/*
- * If everything goes south we can use this to give the operator an
- * emergency shell to debug the problem -- Finit should not crash!
- *
- * Note: Only use this for debugging a new Finit setup, don't use
- *       this in production since it gives a root shell to anyone
- *       if Finit crashes.
- *
- * This emergency shell steps in to prevent "Aieee, PID 1 crashed"
- * messages from the kernel, which usually results in a reboot, so
- * that the operator instead can debug the problem.
- */
-static void emergency_shell(void)
-{
-#ifdef EMERGENCY_SHELL
-	pid_t pid;
-
-	pid = fork();
-	if (pid) {
-		while (1) {
-			pid_t id;
-
-			/* Reap 'em (prevents Zombies) */
-			id = waitpid(-1, NULL, WNOHANG);
-			if (id == pid)
-				break;
-		}
-
-		fprintf(stderr, "\n=> Embarrassingly, Finit has crashed.  Check /dev/kmsg for details.\n");
-		fprintf(stderr,   "=> To debug, add 'debug' to the kernel command line.\n\n");
-
-		/*
-		 * Become session leader and set controlling TTY
-		 * to enable Ctrl-C and job control in shell.
-		 */
-		setsid();
-		ioctl(STDIN_FILENO, TIOCSCTTY, 1);
-
-		execl(_PATH_BSHELL, _PATH_BSHELL, NULL);
-	}
-#endif /* EMERGENCY_SHELL */
-}

 /*
 * Handle bootstrap transition to configured runlevel, start TTYs
@@ -285,20 +348,8 @@ static void emergency_shell(void)
 * finit.conf, or *.conf in finit.d/, run to completion before we
 * finalize the bootstrap process by calling this function.
 */
-static void finalize(void)
+static void finalize(void *unused)
 {
-	/*
-	 * Run startup scripts in the runparts directory, if any.
-	 */
-	if (runparts && fisdir(runparts) && !rescue)
-		run_parts(runparts, NULL);
-
-	/*
-	 * Start all tasks/services in the configured runlevel
-	 */
-	_d("Change to default runlevel, start all services ...");
-	service_runlevel(cfglevel);
-
 	/* Clean up bootstrap-only tasks/services that never started */
 	_d("Clean up all bootstrap-only tasks/services ...");
 	svc_prune_bootstrap();
@@ -320,9 +371,9 @@ static void finalize(void)
 	/* Disable progress output at normal runtime */
 	enable_progress(0);

-	/* Delayed start of TTYs at bootstrap */
-	_d("Launching all getty services ...");
-	tty_runlevel();
+	/* System bootrapped, launch TTYs et al */
+	bootstrap = 0;
+	service_step_all(SVC_TYPE_RESPAWN);
 }

 /*
@@ -340,12 +391,17 @@ static void crank_worker(void *unused)

 /*
 * Wait for system bootstrap to complete, all SVC_TYPE_RUNTASK must be
- * allowed to complete their work in [S], or timeout, before we call
- * finalize(), should not take more than 120 sec.
+ * allowed to complete their work in [S], or timeout, before we switch
+ * to the configured runlevel and call finalize(), should not take more
+ * than 120 sec.
 */
-static void final_worker(void *work)
+static void bootstrap_worker(void *work)
 {
 	static int cnt = 120;
+	static struct wq final = {
+		.cb = finalize,
+		.delay = 10
+	};

 	_d("Step all services ...");
 	service_step_all(SVC_TYPE_ANY);
@@ -361,12 +417,36 @@ static void final_worker(void *work)
 	else
 		_d("Timeout, resuming bootstrap.");

-	finalize();
+	_d("Starting runlevel change finalize ...");
+	schedule_work(&final);
+
+	/*
+	 * Run startup scripts in the runparts directory, if any.
+	 */
+	if (runparts && fisdir(runparts) && !rescue)
+		run_parts(runparts, NULL);
+
+	/*
+	 * Start all tasks/services in the configured runlevel
+	 */
+	_d("Change to default runlevel(%d), starting all services ...", cfglevel);
+	service_runlevel(cfglevel);
+}
+
+static int version(int rc)
+{
+	puts(PACKAGE_STRING);
+	printf("Bug report address: %-40s\n", PACKAGE_BUGREPORT);
+#ifdef PACKAGE_URL
+	printf("Project homepage: %s\n", PACKAGE_URL);
+#endif
+
+	return rc;
 }

 static int usage(int rc)
 {
-	printf("Usage: telinit [OPTIONS] [q | Q | 0-9]\n\n"
+	printf("Usage: %s [OPTIONS] [q | Q | 0-9]\n\n"
 	       "Options:\n"
 //	       "  -a       Ignored, compat SysV init\n"
 //	       "  -b       Ignored, compat SysV init\n"
@@ -384,12 +464,7 @@ static int usage(int rc)
 	       "  q, Q     Reload /etc/finit.conf and/or any *.conf in /etc/finit.d/\n"
 	       "           if modified, same as initctl reload or SIGHUP to PID 1\n"
 	       "  1, s, S  Enter system rescue mode, runlevel 1\n"
-	       "\n");
-
-	printf("Bug report address: %-40s\n", PACKAGE_BUGREPORT);
-#ifdef PACKAGE_URL
-	printf("Project homepage: %s\n", PACKAGE_URL);
-#endif
+	       "\n", prognm);

 	return rc;
 }
@@ -402,6 +477,7 @@ static int telinit(int argc, char *argv[])
 {
 	int c;

+	progname(argv[0]);
 	while ((c = getopt(argc, argv, "abe:h?st:vVz:")) != EOF) {
 		switch(c) {
 		case 'a': case 'b': case 'e': case 's': case 'z':
@@ -411,7 +487,7 @@ static int telinit(int argc, char *argv[])
 			break;

 		case 'v': case 'V':
-			return puts("v" VERSION) == EOF;
+			return version(0);

 		case 'h':
 		case '?':
@@ -423,13 +499,13 @@ static int telinit(int argc, char *argv[])
 		int req = (int)argv[optind][0];

 		if (isdigit(req))
-			return systemf("initctl runlevel %c", req);
+			return systemf("initctl -b runlevel %c", req);

 		if (req == 'q' || req == 'Q')
-			return systemf("initctl reload");
+			return systemf("initctl -b reload");

 		if (req == 's' || req == 'S')
-			return systemf("initctl runlevel %c", req);
+			return systemf("initctl -b runlevel %c", req);
 	}

 	/* XXX: add non-pid1 process monitor here
@@ -444,11 +520,12 @@ static int telinit(int argc, char *argv[])
 int main(int argc, char *argv[])
 {
 	struct wq crank = {
-		.cb = crank_worker
+		.cb = crank_worker,
+		.delay = 10
 	};
-	struct wq final = {
-		.cb = final_worker,
-		.delay = 1000
+	struct wq bootstrap = {
+		.cb = bootstrap_worker,
+		.delay = 100
 	};
 	uev_ctx_t loop;

@@ -479,22 +556,34 @@ int main(int argc, char *argv[])
 	ctx = &loop;

 	/*
-	 * Set PATH, SHELL, PWD, and umask early to something sane
+	 * Set PATH, SHELL, and PWD early to something sane
 	 */
 	setenv("PATH", _PATH_STDPATH, 1);
 	setenv("SHELL", _PATH_BSHELL, 1);
+	setenv("LOGNAME", "root", 1);
+	setenv("USER", "root", 1);

 	if (chdir("/"))
 		_pe("Failed cd /");
-	umask(0);
-
-	/* Allow progress, if enabled */
-	enable_progress(1);

 	/*
 	 * In case of emergency.
 	 */
-	emergency_shell();
+	if (rescue) {
+		char *sulogin[] = {
+			_PATH_SULOGIN,
+			"sulogin",
+		};
+		size_t i;
+
+		for (i = 0; i < NELEMS(sulogin); i++) {
+			if (systemf(sulogin[i]))
+				continue;
+
+			rescue = 0;
+			break;
+		}
+	}

 	/*
 	 * Load plugins early, the first hook is in banner(), so we
@@ -505,6 +594,7 @@ int main(int argc, char *argv[])
 	/*
 	 * Hello world.
 	 */
+	enable_progress(1);	/* Allow progress, if enabled */
 	banner();

 	/*
@@ -515,8 +605,7 @@ int main(int argc, char *argv[])
 	/*
 	 * Initialize default control groups, if available
 	 */
-	cgroup_init();
-
+	cgroup_init(&loop);

 	/* Check and mount filesystems. */
 	fs_mount();
@@ -534,6 +623,20 @@ int main(int argc, char *argv[])
 	 */
 	conf_init(&loop);

+	/*
+	 * Start built-in watchdogd as soon as possible, if enabled
+	 */
+	if (whichp(FINIT_LIBPATH_ "/watchdogd") && fexist(WDT_DEVNODE)) {
+		service_register(SVC_TYPE_SERVICE, "[123456789] cgroup.init name:watchdog :finit " FINIT_LIBPATH_ "/watchdogd -- Finit watchdog daemon", global_rlimit, NULL);
+		wdog = svc_find_by_nameid("watchdog", "finit");
+	}
+
+	/*
+	 * Start kernel event daemon as soon as possible, if enabled
+	 */
+	if (whichp(FINIT_LIBPATH_ "/keventd"))
+		service_register(SVC_TYPE_SERVICE, "[123456789] cgroup.init " FINIT_LIBPATH_ "/keventd -- Finit kernel event daemon", global_rlimit, NULL);
+
 	/* Base FS up, enable standard SysV init signals */
 	sig_setup(&loop);

@@ -548,13 +651,12 @@ int main(int argc, char *argv[])

 	_d("Starting initctl API responder ...");
 	api_init(&loop);
-	umask(022);

 	_d("Starting the big state machine ...");
 	schedule_work(&crank);

 	_d("Starting bootstrap finalize timer ...");
-	schedule_work(&final);
+	schedule_work(&bootstrap);

 	/*
 	 * Enter main loop to monitor /dev/initctl and services
@@ -30,6 +30,7 @@
 #include <paths.h>
 #include <stdio.h>
 #include <stdlib.h>
+#include <sysexits.h>
 #include <sys/types.h>
 #include <sys/stat.h>
 #include <unistd.h>
@@ -46,10 +47,13 @@
 #endif

 #ifndef FINIT_CGPATH
-#define FINIT_CGPATH            "/sys/fs/cgroup/finit"
+#define FINIT_CGPATH            "/sys/fs/cgroup"
 #endif

-#define LOGIT_PATH              FINIT_LIBPATH_ "/logit"
+/* To use these, include config.h, or define FINIT_LIBPATH_ */
+#define _PATH_LOGIT             FINIT_LIBPATH_ "/logit"
+#define _PATH_SULOGIN           FINIT_LIBPATH_ "/sulogin"
+#define _PATH_GETTY             FINIT_LIBPATH_ "/getty"

 #define CMD_SIZE                256
 #define LINE_SIZE               1024
@@ -71,10 +75,10 @@
 /* Finit extensions over std SysV */
 #define INIT_CMD_DEBUG          8    /* Toggle Finit debug */
 #define INIT_CMD_RELOAD         9    /* Reload *.conf in /etc/finit.d/ */
-#define INIT_CMD_START_SVC      10
+#define INIT_CMD_START_SVC      10   /* START service */
 #define INIT_CMD_STOP_SVC       11
 #define INIT_CMD_RELOAD_SVC     12   /* SIGHUP service */
-#define INIT_CMD_RESTART_SVC    13   /* STOP + START service */
+#define INIT_CMD_RESTART_SVC    13   /* START service, clearing blocks */
 #define INIT_CMD_UNUSED2        14   /* Unused, was INIT_CMD_QUERY_INETD */
 #define INIT_CMD_UNUSED1        15   /* Unused, was INIT_CMD_EMIT */
 #define INIT_CMD_GET_RUNLEVEL   16
@@ -105,11 +109,13 @@ extern int    prevlevel;
 extern int    debug;
 extern int    rescue;
 extern int    single;
+extern int    bootstrap;
 extern char  *rcsd;
 extern char  *sdown;
 extern char  *network;
 extern char  *hostname;
 extern char  *runparts;
+extern char  *osheading;
 extern uev_ctx_t *ctx;

 #endif /* FINIT_H_ */
@@ -21,45 +21,38 @@
 * THE SOFTWARE.
 */

+#include "config.h"
+
 #include <err.h>
 #include <errno.h>
 #include <paths.h>
 #include <signal.h>
 #include <stdio.h>
+#include <stdlib.h>
 #include <sys/ioctl.h>
 #include <sys/utsname.h>
 #include <sys/ttydefaults.h>	/* Not included by default in musl libc */
 #include <termios.h>

 #include "finit.h"
-#include "cgroup.h"
 #include "helpers.h"
-#include "sig.h"
 #include "utmp-api.h"

 #ifndef _PATH_LOGIN
 #define _PATH_LOGIN  "/bin/login"
 #endif

-#define print(s) (void)write(STDOUT_FILENO, s, strlen(s))
-
-
 /*
 * Read one character from stdin.
 */
-static int readch(char *tty)
+static int readch(void)
 {
-	int st;
 	char ch1;
+	int st;

 	st = read(STDIN_FILENO, &ch1, 1);
-	if (st == 0) {
-		print("\n");
-		_exit(0);
-	}
-
-	if (st < 0)
-		errx(1, "getty: %s: read error", tty);
+	if (st <= 0)
+		return -1;

 	return ch1 & 0xFF;
 }
@@ -74,31 +67,32 @@ static void do_parse(char *line, struct utsname *uts, char *tty)
 	s0 = line;
 	for (s = line; *s != 0; s++) {
 		if (*s == '\\') {
-			(void)write(1, s0, s - s0);
+			if ((s - s0) > 0)
+				dprint(STDOUT_FILENO, s0, s - s0);
 			s0 = s + 2;
 			switch (*++s) {
 			case 'l':
-				print(tty);
+				dprint(STDOUT_FILENO, tty, 0);
 				break;
 			case 'm':
-				print(uts->machine);
+				dprint(STDOUT_FILENO, uts->machine, 0);
 				break;
 			case 'n':
-				print(uts->nodename);
+				dprint(STDOUT_FILENO, uts->nodename, 0);
 				break;
 #ifdef _GNU_SOURCE
 			case 'o':
-				print(uts->domainname);
+				dprint(STDOUT_FILENO, uts->domainname, 0);
 				break;
 #endif
 			case 'r':
-				print(uts->release);
+				dprint(STDOUT_FILENO, uts->release, 0);
 				break;
 			case 's':
-				print(uts->sysname);
+				dprint(STDOUT_FILENO, uts->sysname, 0);
 				break;
 			case 'v':
-				print(uts->version);
+				dprint(STDOUT_FILENO, uts->version, 0);
 				break;
 			case 0:
 				goto leave;
@@ -109,7 +103,8 @@ static void do_parse(char *line, struct utsname *uts, char *tty)
 	}

 leave:
-	(void)write(1, s0, s - s0);
+	if ((s - s0) > 0)
+		dprint(STDOUT_FILENO, s0, s - s0);
 }

 /*
@@ -142,10 +137,10 @@ static void do_issue(char *tty)
 /*
 * Handle the process of a GETTY.
 */
-static void do_getty(char *tty, char *name, size_t len)
+static int get_logname(char *tty, char *name, size_t len)
 {
-	int ch;
 	char *np;
+	int ch;

 	/*
 	 * Display prompt.
@@ -156,14 +151,9 @@ static void do_getty(char *tty, char *name, size_t len)
 		do_issue(tty);

 		np = name;
-		while ((ch = readch(tty)) != '\n') {
-			if (ch == CTRL('U')) {
-				while (np > name) {
-					(void)write(1, "\b \b", 3);
-					np--;
-				}
-				continue;
-			}
+		while ((ch = readch()) != '\n') {
+			if (ch < 0)
+				return 1;

 			if (np < name + len)
 				*np++ = ch;
@@ -175,6 +165,7 @@ static void do_getty(char *tty, char *name, size_t len)
 	}

 	name[len - 1] = 0;
+	return 0;
 }

 /*
@@ -186,26 +177,28 @@ static int do_login(char *name)
 {
 	struct stat st;

-	cgroup_user(name);
 	execl(_PATH_LOGIN, _PATH_LOGIN, name, NULL);

 	/*
-	 * Failed to exec login, should be impossible.  Try a starting a
-	 * fallback shell instead.
-	 *
-	 * Note: Add /etc/securetty handling.
+	 * Failed to exec login, should not happen on normal systems.
+	 * Try a starting a rescue shell instead.
 	 */
-	warnx("Failed exec %s, attempting fallback to %s ...", _PATH_LOGIN, _PATH_BSHELL);
-	if (fstat(0, &st) == 0 && S_ISCHR(st.st_mode))
+	if (fstat(0, &st) == 0 && S_ISCHR(st.st_mode)) {
+		warnx("Failed exec %s, attempting fallback to %s ...", _PATH_LOGIN, _PATH_SULOGIN);
+		execl(_PATH_SULOGIN, _PATH_SULOGIN, NULL);
+
+		warnx("Failed exec %s, attempting fallback to %s ...", _PATH_SULOGIN, _PATH_BSHELL);
 		execl(_PATH_BSHELL, _PATH_BSHELL, NULL);
+	}

 	return 1;	/* We shouldn't get here ... */
 }

-int getty(char *tty, speed_t speed, char *term, char *user)
+static int getty(char *tty, speed_t speed, char *term, char *user)
 {
 	const char cln[] = "\r\e[2K\n";
-	char name[30];
+	char name[33];		/* useradd(1) limit at 32 chars */
+	pid_t sid;

 	/*
 	 * Clean up tty name.
@@ -213,59 +206,64 @@ int getty(char *tty, speed_t speed, char *term, char *user)
 	if (!strncmp(tty, _PATH_DEV, strlen(_PATH_DEV)))
 		tty += 5;

-	/* Set up TTY, re-enabling ISIG et al. */
-	stty(STDIN_FILENO, speed);
-	(void)write(STDERR_FILENO, cln, strlen(cln));
-
 	/* The getty process is responsible for the UTMP login record */
 	utmp_set_login(tty, NULL);
-	if (!user)
-		do_getty(tty, name, sizeof(name));
-	else
+
+	/* Replace "Please press enter ..." with login: */
+	dprint(STDERR_FILENO, cln, strlen(cln));
+
+restart:
+	stty(STDIN_FILENO, speed);
+	if (!user) {
+		if (get_logname(tty, name, sizeof(name)))
+			goto restart;
+	} else
 		strlcpy(name, user, sizeof(name));

+	/* check current session associated with our tty */
+	sid = tcgetsid(0);
+	if (sid < 0 || getpid() != sid) {
+		if (ioctl(STDIN_FILENO, TIOCSCTTY, 1) == -1)
+			err(1, "failed stealing controlling TTY");
+	}
+
 	if (term && term[0])
 		setenv("TERM", term, 1);

 	return do_login(name);
 }

-int sh(char *tty)
+static int usage(int rc)
 {
-	struct termios term;
-	char *args[2] = {
-		NULL,
-		NULL
-	};
-	char *arg0;
-	size_t len;
+	warnx("usage: getty [-h?] tty [speed [term]]");
+	return rc;
+}

-	/* The getty process is usually responsible for the UTMP login record */
-	utmp_set_login(tty, NULL);
+int main(int argc, char *argv[])
+{
+	char *tty, *speed = "0", *term = NULL;
+	int c;

-	/* Set up TTY, re-enabling ISIG et al. */
-	stty(STDIN_FILENO, B0);
-
-	/* Start /bin/sh as a login shell, i.e. with a prefix '-' */
-	len = strlen(_PATH_BSHELL) + 2;
-	arg0 = malloc(len);
-	if (!arg0)
-		err(1, "Failed allocating memory");
-	snprintf(arg0, len, "-%s", _PATH_BSHELL);
-	args[0] = arg0;
-
-	/* Reenable Ctrl-D and Ctrl-C, and ... */
-	if (!tcgetattr(STDIN_FILENO, &term)) {
-		term.c_lflag    |= ISIG;
-		term.c_cc[VEOF]  = CEOF;
-		term.c_cc[VINTR] = CINTR;
-		tcsetattr(STDIN_FILENO, TCSAFLUSH, &term);
+	while ((c = getopt(argc, argv, "h?")) != EOF) {
+		switch(c) {
+		case 'h':
+		case '?':
+			return usage(0);
+		default:
+			return usage(1);
+		}
 	}

-	/* ... unblock signals in general */
-	sig_unblock();
+	if (optind >= argc)
+		return usage(1);

-	return execv(_PATH_BSHELL, args);
+	tty = argv[optind++];
+	if (optind < argc)
+		speed = argv[optind++];
+	if (optind < argc)
+		term = argv[optind++];
+
+	return getty(tty, atoi(speed), term, NULL);
 }

 /**
@@ -68,90 +68,14 @@ static const char *status2[] = STATUS_CLASS;
 #define CHOOSE(x,y,z) z
 static const char *color[] = STATUS_CLASS;

-/*
- * System consoles, can be given multiple times on the kernel cmdline.
- * We support only three currently.  If none are given on the cmdline
- * we probe /sys/class/tty/console/active and use the first (default).
- */
-#define MAX_CONS 3
-
-static char *consoles[MAX_CONS];
-static int fds[MAX_CONS];
-static int num_cons = 0;
-
-/*
- * System default console is always the first
- */
 char *console(void)
 {
-	return consoles[0];
-}
-
-static void add_console(char *cons)
-{
-	char path[80];
-	char *ptr;
-	int i;
-
-	if (num_cons == NELEMS(fds)) {
-		_e("too many consoles, max %d supported", NELEMS(fds));
-		return;
-	}
-
-	/* drop any options */
-	ptr = strchr(cons, ',');
-	if (ptr)
-		*ptr = 0;
-
-	snprintf(path, sizeof(path), "/dev/%s", cons);
-
-	/* check for and skip duplicates */
-	for (i = 0; i < num_cons; i++) {
-		if (string_compare(path, consoles[i]))
-			return;
-	}
-
-	/* this requires we have /dev mounted && a device there */
-	fds[num_cons] = open(path, O_WRONLY);
-	if (fds[num_cons] < 0) {
-		_pe("Failed opening console %s", path);
-		return;
-	}
-	consoles[num_cons] = strdup(path);
-	num_cons++;
-}
-
-static void find_active_consoles(void)
-{
-	char *cons, *ptr, *tok;
-	char buf[512];
-	FILE *fp;
-
-	fp = fopen("/sys/class/tty/console/active", "r");
-	if (!fp) {
-	fallback:
-		_d("Using fallback /dev/console");
-		add_console("/dev/console");
-		return;
-	}
-
-	ptr = fgets(buf, sizeof(buf), fp);
-	fclose(fp);
-	if (!ptr)
-		goto fallback;
-
-	cons = chomp(ptr);
-	_d("Using these system consoles: %s", cons);
-	while ((tok = strtok(cons, " \t"))) {
-		cons = NULL;
-		add_console(tok);
-	}
+	return _PATH_CONSOLE;
 }

 void console_init(void)
 {
-	find_active_consoles();
-	screen_init();
+	ttinit();
 }

 ssize_t cprintf(const char *fmt, ...)
@@ -160,14 +84,12 @@ ssize_t cprintf(const char *fmt, ...)
 	char buf[len < 256 ? 256 : len];
 	size_t size;
 	va_list ap;
-	int i;

 	va_start(ap, fmt);
 	size = vsnprintf(buf, sizeof(buf), fmt, ap);
 	va_end(ap);

-	for (i = 0; i < num_cons; i++)
-		(void)write(fds[i], buf, size);
+	dprint(STDERR_FILENO, buf, size);

 	return size;
 }
@@ -202,6 +124,49 @@ void show_progress(pstyle_t style)
 	progress_onoff = style;
 }

+/*
+ * Find PRETTY_NAME in /etc/os-release to use as heading at boot
+ * and on runlevel changes.  Fallback: "Finit vX.YY"
+ */
+char *release_heading(void)
+{
+	char *name = NULL;
+	char buf[256];
+	FILE *fp;
+
+	fp = fopen("/etc/os-release", "r");
+	if (fp) {
+		while (fgets(buf, sizeof(buf), fp)) {
+			char *ptr = &buf[12];
+
+			if (strncmp(buf, "PRETTY_NAME=", 12))
+				continue;
+
+			if (*ptr == '"' || *ptr == '\'') {
+				char *end = &ptr[strlen(ptr) - 1];
+				char q = *ptr;
+
+				while (end > ptr && isspace(*end))
+					*end-- = 0;
+
+				if (*end == q) {
+					*end = 0;
+					ptr++;
+				}
+			}
+
+			name = strdup(ptr);
+			break;
+		}
+		fclose(fp);
+	}
+
+	if (!name)
+		name = strdup(PACKAGE_NAME " v" PACKAGE_VERSION);
+
+	return name;
+}
+
 /*
 * Return screen length of string, not counting escape chars, and
 * accounting for unicode characters as only one screen byte wide
@@ -252,20 +217,20 @@ static char *pad(char *buf, size_t len, char *ch, size_t width)

 void print_banner(const char *heading)
 {
-	char buf[4 * SCREEN_WIDTH];
+	char buf[4 * ttcols];

 	if (progress_style == PROGRESS_SILENT)
 		return;

-	memset(buf, 0, sizeof(buf));
+	buf[0] = 0;
 	strlcat(buf, "\r\e[2K", sizeof(buf));

 	if (progress_style == PROGRESS_CLASSIC) {
 		strlcat(buf, "\e[1m", sizeof(buf));
 		strlcat(buf, heading, sizeof(buf));
-		pad(buf, sizeof(buf), "=", SCREEN_WIDTH + 10);
+		pad(buf, sizeof(buf), "=", ttcols + 10);
 	} else {
-		size_t wmax = 80 <= SCREEN_WIDTH ? 80 : SCREEN_WIDTH;
+		size_t wmax = 80 <= ttcols ? 80 : ttcols;

 		/* • • • Foo System ═══════════════════════════════════════════════ */
 		/* ⬤ ⬤ ⬤ Foo System ═══════════════════════════════════════════════ */
@@ -328,7 +293,7 @@ static char *status(int rc)

 void printv(const char *fmt, va_list ap)
 {
-	char buf[SCREEN_WIDTH];
+	char buf[ttcols];
 	size_t len;

 	if (!fmt || progress_style == PROGRESS_SILENT)
@@ -336,7 +301,7 @@ void printv(const char *fmt, va_list ap)

 	delline();

-	memset(buf, 0, sizeof(buf));
+	buf[0] = 0;
 	len = print_timestamp(buf, sizeof(buf));
 	vsnprintf(&buf[len], sizeof(buf) - len, fmt, ap);

@@ -410,6 +375,64 @@ int getgroup(char *group)
 #endif
 }

+/* get current user */
+int getcuser(char *buf, size_t len)
+{
+	char *ptr = "root";
+
+#ifndef ENABLE_STATIC
+	struct passwd *pw = getpwuid(getuid());
+
+	if (pw)
+		ptr = pw->pw_name;
+#endif
+	if (strlcpy(buf, ptr, len) > len)
+		return -1;
+
+	return 0;
+}
+
+/* get current group */
+int getcgroup(char *buf, size_t len)
+{
+	char *ptr = "root";
+
+#ifndef ENABLE_STATIC
+	struct group *gr = getgrgid(getgid());
+
+	if (gr)
+		ptr = gr->gr_name;
+#endif
+	if (strlcpy(buf, ptr, len) > len)
+		return -1;
+
+	return 0;
+}
+
+int mksubsys(const char *dir, mode_t mode, char *user, char *group)
+{
+	mode_t omask;
+	int uid, gid;
+	int rc = 0;
+
+	omask = umask(0);
+
+	uid = getuser(user, NULL);
+	if (uid >= 0) {
+		gid = getgroup(group);
+		if (gid < 0)
+			gid = 0;
+
+		rc = makedir(dir, mode);
+		if (chown(dir, uid, gid))
+			_pe("Failed chown(%s, %d, %d)", dir, uid, gid);
+	}
+
+	umask(omask);
+
+	return rc;
+}
+
 void set_hostname(char **hostname)
 {
 	FILE *fp;
@@ -438,8 +461,10 @@ void set_hostname(char **hostname)
 	}

 done:
-	if (*hostname)
-		(void)sethostname(*hostname, strlen(*hostname));
+	if (*hostname) {
+		if (sethostname(*hostname, strlen(*hostname)))
+			_pe("Failed sethostnaem(%s)", *hostname);
+	}
 }

 /*
@@ -477,6 +502,7 @@ void networking(int updown)
 				run("ifdown -a");
 			_exit(0);
 		}
+		cgroup_service("network", pid, NULL);

 		print(pid > 0 ? 0 : 1, "%s networking", updown ? "Starting" : "Stopping");
 	}
@@ -26,6 +26,7 @@
 #define FINIT_HELPERS_H_

 #include <mntent.h>
+#include <sched.h>		/* sched_yield() */
 #include <stdarg.h>
 #include <sys/resource.h>
 #include <sys/types.h>
@@ -67,11 +68,10 @@ char   *strip_line      (char *line);
 void    enable_progress (int onoff);
 void    show_progress   (pstyle_t style);

-int     getty           (char *tty, speed_t speed, char *term, char *user);
-int     sh              (char *tty);
+char   *release_heading (void);

-int     stty            (int fd, speed_t speed);
 speed_t stty_parse_speed(char *baud);
+void    stty            (int fd, speed_t speed);

 void    print_banner    (const char *heading);
 void    printv          (const char *fmt, va_list ap);
@@ -81,6 +81,12 @@ int     print_result    (int fail);

 int     getuser         (char *username, char **home);
 int     getgroup        (char *group);
+
+int     getcuser        (char *buf, size_t len);
+int     getcgroup       (char *buf, size_t len);
+
+int     mksubsys        (const char *dir, mode_t mode, char *user, char *group);
+
 void    set_hostname    (char **hostname);
 void    networking      (int updown);

@@ -88,8 +94,7 @@ int     complete        (char *cmd, int pid);
 int     run             (char *cmd);
 int     run_interactive (char *cmd, char *fmt, ...);
 int     exec_runtask    (char *cmd, char *args[]);
-pid_t   run_getty       (char *tty, char *baud, char *term,  int noclear, int nowait, struct rlimit rlimit[]);
-pid_t   run_getty2      (char *tty, char *cmd, char *args[], int noclear, int nowait, struct rlimit rlimit[]);
+pid_t   run_getty       (char *tty, char *cmd, char *args[], int noclear, int nowait, struct rlimit rlimit[]);
 pid_t   run_sh          (char *tty, int noclear, int nowait, struct rlimit rlimit[]);
 int     run_parts       (char *dir, char *cmd);

@@ -103,6 +108,24 @@ static inline int create(char *path, mode_t mode, uid_t uid, gid_t gid)
 	return 0;
 }

+static inline int dprint(int fd, const char *s, size_t len)
+{
+	size_t loop = 3;
+	int rc = -1;
+
+	if (!len)
+		len = strlen(s);
+
+	while (loop--) {
+		rc = write(fd, s, len);
+		if (rc == -1 && errno == EINTR)
+			continue;
+		break;
+	}
+
+	return rc;
+}
+
 int	ismnt		(char *file, char *dir, char *mode);
 int	fismnt		(char *dir);

@@ -0,0 +1,46 @@
+/* New client tool, replaces old /dev/initctl API and telinit tool
+ *
+ * Copyright (c) 2015-2021  Joachim Wiberg <troglobit@gmail.com>
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to deal
+ * in the Software without restriction, including without limitation the rights
+ * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+ * copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
+ * THE SOFTWARE.
+ */
+
+#ifndef FINIT_INITCTL_H_
+#define FINIT_INITCTL_H_
+
+#include "finit.h"
+#include "util.h"
+
+extern int icreate;			/* initctl -c */
+extern int iforce;			/* initctl -f */
+extern int ionce;			/* initctl -1 */
+extern int heading;			/* initctl -t */
+extern int verbose;			/* initctl -v */
+extern int plain;			/* initctl -p */
+
+extern void print_header(const char *fmt, ...);
+
+#endif /* FINIT_INITCTL_H_ */
+
+/**
+ * Local Variables:
+ *  indent-tabs-mode: t
+ *  c-file-style: "linux"
+ * End:
+ */
@@ -81,10 +81,10 @@ int iwatch_add(struct iwatch *iw, char *file, uint32_t mask)
 		return -1;

 	if (!fexist(file)) {
-		_d("No such file or directory, skipping %s", file);
+		_d("skipping %s: no such file or directory", file);
 		return 0;
 	}
-	_d("Adding new watcher for path %s", file);
+	_d("adding new watcher for path %s", file);

 	path = strdup(file);
 	if (!path) {
@@ -94,7 +94,24 @@ int iwatch_add(struct iwatch *iw, char *file, uint32_t mask)

 	wd = inotify_add_watch(iw->fd, path, IWATCH_MASK | mask);
 	if (wd < 0) {
-		_pe("Failed adding watcher for %s", path);
+		switch (errno) {
+		case EEXIST:
+			break;
+
+		case EINVAL:
+			/*
+			 * some older kernels, notably 4.19, don't like
+			 * inotify on top-level cgroups in cgroup v2
+			 */
+			if (!strncmp(path, "/sys/fs/cgroup", 14))
+				break;	/* ignore error */
+			/* fallthrough */
+
+		default:
+			_pe("Failed adding watcher for %s", path);
+			break;
+		}
+
 		free(path);
 		return -1;
 	}
@@ -0,0 +1,279 @@
+/* Listens to kernel events like AC power status and manages sys/ conditions
+ *
+ * Copyright (c) 2021  Joachim Wiberg <troglobit@gmail.com>
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to deal
+ * in the Software without restriction, including without limitation the rights
+ * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+ * copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
+ * THE SOFTWARE.
+ */
+
+#include <errno.h>
+#include <dirent.h>
+#include <poll.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <stdarg.h>
+#include <string.h>
+#include <syslog.h>
+#include <unistd.h>
+
+#include <sys/socket.h>
+#include <sys/types.h>
+
+#include <linux/types.h>
+#include <linux/netlink.h>
+
+#include "lite/lite.h"
+
+#include "cond.h"
+#include "pid.h"
+#include "util.h"
+
+#define _PATH_SYSFS_PWR  "/sys/class/power_supply"
+
+static int num_ac_online;
+static int num_ac;
+
+int debug;			/* debug in other modules as well */
+
+void logit(int prio, const char *fmt, ...)
+{
+	va_list ap;
+
+	va_start(ap, fmt);
+	if (!debug)
+		vsyslog(prio, fmt, ap);
+	else {
+		vfprintf(stderr, fmt, ap);
+		fprintf(stderr, "\n");
+	}
+	va_end(ap);
+}
+#define panic(fmt, args...) { logit(LOG_CRIT, fmt ":%s", ##args, strerror(errno)); exit(1); }
+#define warn(fmt, args...)  { logit(LOG_WARNING, fmt ":%s", ##args, strerror(errno)); }
+
+static void sys_cond(char *cond, int set)
+{
+	char oneshot[256];
+
+	snprintf(oneshot, sizeof(oneshot), "%s/%s", _PATH_CONDSYS, cond);
+	if (set) {
+		if (symlink(_PATH_RECONF, oneshot) && errno != EEXIST)
+			warn("failed asserting sys/%s", cond);
+	} else {
+		if (erase(oneshot) && errno != ENOENT)
+			warn("failed asserting sys/%s", cond);
+	}
+}
+
+static int fgetline(char *path, char *buf, size_t len)
+{
+	FILE *fp;
+
+	fp = fopen(path, "r");
+	if (!fp)
+		return -1;
+
+	if (!fgets(buf, len, fp)) {
+		fclose(fp);
+		return -1;
+	}
+
+	chomp(buf);
+	fclose(fp);
+
+	return 0;
+}
+
+static int check_online(char *online)
+{
+	int val;
+
+	if (!online)
+		return 0;
+
+	val = atoi(online);
+	logit(LOG_INFO, "AC %s", val ? "connected" : "disconnected");
+
+	return val;
+}
+
+static int is_ac(char *type)
+{
+	char *types[] = {
+		"Mains",
+		"USB",
+		"BrickID",
+		"Wireless",
+		NULL
+	};
+	int i;
+
+	for (i = 0; types[i]; i++) {
+		if (!strncmp(type, types[i], strlen(types[i])))
+			return 1;
+	}
+
+	return 0;
+}
+
+static void init(void)
+{
+	struct dirent **d = NULL;
+	char *cond_dirs[] = {
+		_PATH_CONDSYS,
+		_PATH_CONDSYS "/pwr",
+	};
+	char path[384];
+	int i, n;
+
+	for (i = 0; i < (int)NELEMS(cond_dirs); i++) {
+		if (mkpath(cond_dirs[i], 0755) && errno != EEXIST) {
+			_pe("Failed creating %s condition directory, %s", COND_SYS,
+			    cond_dirs[i]);
+			return;
+		}
+	}
+
+	n = scandir(_PATH_SYSFS_PWR, &d, NULL, alphasort);
+	for (i = 0; i < n; i++) {
+		char *nm = d[i]->d_name;
+ 		char buf[10];
+
+		snprintf(path, sizeof(path), "%s/%s/type", _PATH_SYSFS_PWR, nm);
+		if (!fgetline(path, buf, sizeof(buf)) && is_ac(buf)) {
+			num_ac++;
+
+			snprintf(path, sizeof(path), "%s/%s/online", _PATH_SYSFS_PWR, nm);
+			if (!fgetline(path, buf, sizeof(buf))) {
+				if (check_online(buf))
+					num_ac_online++;
+			}
+		}
+		free(d[i]);
+	}
+
+	if (n > 0)
+		free(d);
+
+	/* if any power_supply is online, or none can be found */
+	if (num_ac == 0 || num_ac_online > 0)
+		sys_cond("pwr/ac", 1);
+}
+
+/*
+ * Started by Finit as soon as possible when base filesystem is up,
+ * modules have been probed, or insmodded from /etc/finit.conf, so by
+ * now we should have /sys/class/power_supply/ available for probing.
+ * If none is found we assert /sys/pwr/ac condition anyway, this is what
+ * systemd does (ConditionACPower) and also makes most sense.
+ */
+int main(int argc, char *argv[])
+{
+	struct sockaddr_nl nls = { 0 };
+	struct pollfd pfd;
+	char buf[1024];
+
+	if (argc > 1) {
+		if (!strcmp(argv[1], "-d"))
+			debug = 1;
+	}
+
+	if (!debug) {
+		openlog("keventd", LOG_CONS | LOG_PID, LOG_DAEMON);
+		setlogmask(LOG_UPTO(LOG_NOTICE));
+	}
+
+	init();
+
+	pfd.events = POLLIN;
+	pfd.fd = socket(PF_NETLINK, SOCK_DGRAM, NETLINK_KOBJECT_UEVENT);
+	if (pfd.fd == -1)
+		panic("failed creating netlink socket");
+
+	nls.nl_family = AF_NETLINK;
+	nls.nl_pid    = 0;
+	nls.nl_groups = -1;
+	if (bind(pfd.fd, (void *)&nls, sizeof(struct sockaddr_nl)))
+		panic("bind failed");
+
+	logit(LOG_DEBUG, "Waiting for events ...");
+	while (-1 != poll(&pfd, 1, -1)) {
+		char *path = NULL;
+		int ac = 0;
+		int i, len;
+
+		len = recv(pfd.fd, buf, sizeof(buf), MSG_DONTWAIT);
+		if (len == -1) {
+			switch (errno) {
+			case EINTR:
+				continue;
+			case ENOBUFS:
+				warn("lost events");
+				continue;
+			default:
+				panic("unhandled");
+				continue;
+			}
+		}
+		buf[len] = 0;
+
+		/* skip libusb events, focus on kernel events/changes */
+		if (strncmp(buf, "change@", 7))
+			continue;
+
+		/* XXX: currently limited to monitoring this subsystem */
+		if (!strstr(buf, "power_supply"))
+			continue;
+
+		i = 0;
+		while (i < len) {
+			char *line = buf + i;
+
+			logit(LOG_DEBUG, "%s", line);
+			if (!strncmp(line, "DEVPATH=", 8)) {
+				path = &line[8];
+				logit(LOG_DEBUG, "Got path %s", path);
+			} else if (!strncmp(line, "POWER_SUPPLY_TYPE=", 18)) {
+				ac = is_ac(&line[18]);
+			} else if (!strncmp(line, "POWER_SUPPLY_ONLINE=", 20) && ac) {
+				if (check_online(&line[20])) {
+					if (!num_ac_online)
+						sys_cond("pwr/ac", 1);
+					num_ac_online++;
+				} else {
+					if (num_ac_online > 0)
+						num_ac_online--;
+					if (!num_ac_online)
+						sys_cond("pwr/ac", 0);
+				}
+			}
+
+			i += strlen(line) + 1;
+		}
+	}
+	close(pfd.fd);
+
+	return 0;
+}
+
+/**
+ * Local Variables:
+ *  indent-tabs-mode: t
+ *  c-file-style: "linux"
+ * End:
+ */
@@ -50,7 +50,7 @@ void log_exit(void)
 	 * terminal may have been resized at runtime
 	 */
 	if (!debug)
-		screen_init();
+		ttinit();

 	enable_progress(1);
 }
@@ -78,7 +78,7 @@ void log_debug(void)
 		loglevel = LOG_DEBUG;
 	} else {
 		loglevel = LOG_NOTICE;
-		screen_init();
+		ttinit();
 	}
 	log_open();

@@ -104,7 +104,7 @@ void logit(int prio, const char *fmt, ...)
 		goto done;
 	}

-	if (prio > loglevel)
+	if (LOG_PRI(prio) > loglevel)
 		goto done;

 	fp = fopen("/dev/kmsg", "w");
@@ -98,29 +98,38 @@ static int logit(int level, char *buf, size_t len)

 static int parse_prio(char *arg, int *f, int *l)
 {
-	char *ptr;
+	char *duparg = strdup(arg);
+	char *ptr, *prio;

-	ptr = strchr(arg, '.');
+	if (!duparg)
+		prio = arg;
+	else
+		prio = duparg;
+
+	ptr = strchr(prio, '.');
 	if (ptr) {
 		*ptr++ = 0;

 		for (int i = 0; facilitynames[i].c_name; i++) {
-			if (!strcmp(facilitynames[i].c_name, arg)) {
+			if (!strcmp(facilitynames[i].c_name, prio)) {
 				*f = facilitynames[i].c_val;
 				break;
 			}
 		}

-		arg = ptr;
+		prio = ptr;
 	}

 	for (int i = 0; prioritynames[i].c_name; i++) {
-		if (!strcmp(prioritynames[i].c_name, arg)) {
+		if (!strcmp(prioritynames[i].c_name, prio)) {
 			*l = prioritynames[i].c_val;
 			break;
 		}
 	}

+	if (duparg != arg)
+		free(duparg);
+
 	return 0;
 }

@@ -80,7 +80,7 @@ int logrotate(char *file, int num, off_t sz)
 				}

 				if (cnt == 2 && fexist(nfile)) {
-					if (systemf("gzip %s", nfile))
+					if (systemf("gzip %s 2>/dev/null", nfile))
 						continue; /* no gzip, probably */

 					(void)remove(nfile);
@@ -95,18 +95,15 @@ char *pid_get_name(pid_t pid, char *name, size_t len)

 char *pid_file(svc_t *svc)
 {
-	static char path[256];
-	char fn[MAX_ARG_LEN];
-
 	if (svc->pidfile[0]) {
 		if (svc->pidfile[0] == '!')
 			return &svc->pidfile[1];
 		return svc->pidfile;
 	}

-	snprintf(fn, sizeof(fn), "%s%s.pid", _PATH_VARRUN, basename(svc->cmd));
+	errno = ENOENT;

-	return pid_runpath(fn, path, sizeof(path));
+	return NULL;
 }

 pid_t pid_file_read(const char *fn)
@@ -143,12 +140,13 @@ int pid_file_create(svc_t *svc)
 	return fclose(fp);
 }

-static int pid_realpath(svc_t *svc, char *file)
+int pid_file_set(svc_t *svc, char *file, int not)
 {
-	int not = 0;
-
-	if (!file)
+	if (!file) {
 		file = pid_file(svc);
+		if (!file)
+			return -1;
+	}

 	if (file[0] == '!') {
 		not = 1;
@@ -202,7 +200,7 @@ int pid_file_parse(svc_t *svc, char *arg)

 		arg += 4;
 		if ((arg[0] == '!' && arg[1] == '/') || arg[0] == '/')
-			return pid_realpath(svc, arg);
+			return pid_file_set(svc, arg, 0);

 		if (arg[0] == '!') {
 			arg++;
@@ -213,12 +211,12 @@ int pid_file_parse(svc_t *svc, char *arg)
 		if (len > 4 && strcmp(&path[len - 4], ".pid"))
 			strlcat(path, ".pid", sizeof(path));

-		return pid_realpath(svc, path);
+		return pid_file_set(svc, path, 0);
 	}

 	/* 'pid' arg, no argument following */
 	if (!strcmp(arg, "pid"))
-		return pid_realpath(svc, NULL);
+		return pid_file_set(svc, NULL, 0);

 	return 1;
 }
--- a/Show More
+++ b/Show More
				`@@ -0,0 +1 @@`
				`task [2345] <sys/pwr/ac,pid/syslogd> halt:usr1 env:-/etc/default/anacron anacron -d -q $ANACRON_ARGS -- Anacron jobs`
				`@@ -0,0 +1,2 @@`
				`sysv [S] /etc/init.d/apparmor -- AppArmor initialization`
				`@@ -0,0 +1 @@`
				`service [2345] env:-/etc/default/avahi-daemon avahi-daemon -s $AVAHI_DAEMON_OPTS -- Avahi mDNS/DNS-SD Stack`
				`@@ -0,0 +1 @@`
				`service [2345] <sys/pwr/ac,pid/syslogd> env:-/etc/default/cron cron -f $EXTRA_OPTS -- Background job processing daemon`