All Execs within the service, will get mounted the same
/tmp and /var/tmp directories, if service is configured with
PrivateTmp=yes. Temporary directories are cleaned up by service
itself in addition to systemd-tmpfiles. Directory which is mounted
as inaccessible is created at runtime in /run/systemd.
CIPSO is the Common IP Security Option, an IETF standard for setting
security levels for a process sending packets. In Smack kernels,
CIPSO headers are mapped to Smack labels automatically, but can be changed.
This patch writes label/category mappings from /etc/smack/cipso/ to
/sys/fs/smackfs/cipso2. The mapping format is "%s%4d%4d"["%4d"]...
For more information about Smack and CIPSO, see:
https://kernel.org/doc/Documentation/security/Smack.txt
.mount units coming from /proc/self/mountinfo file are
unmounted after local-fs.target is reached during shutdown.
Problem: .mount units popping up in mountinfo file are
added to systemd without any dependency. For that reason,
they are the first one to be unmounted during shutdown.
Whichever program mounted the file system deserves a
chance to also unmount it. This patch ensures that
/proc/self/mountinfo units will be unmounted after
local-fs.target during shutdown (if they haven't been
unmounted already)
Previously we were testing whether /sys/fs/cgroup/systemd/ was a mount
point. This might be problematic however, when the cgroup trees are bind
mounted into a container from the host (which should be absolutely
valid), which might create the impression that the container was running
systemd, but only the host actually is.
Replace this by a check for the existance of the directory
/run/systemd/system/, which should work unconditionally, since /run can
never be a bind mount but *must* be a tmpfs on systemd systems, which is
flushed at boots. This means that data in /run always reflects
information about the current boot, and only of the local container,
which makes it the perfect choice for a check like this.
(As side effect this is nice to Ubuntu people who now use logind with
the systemd cgroup hierarchy, where the old sd_booted() check misdetects
systemd, even though they still run legacy Upstart.)
First, rename root-fs.target to initrd-root-fs.target to clarify its usage.
Mount units with "x-initrd-rootfs.mount" are now ordered before
initrd-root-fs.target. As we sometimes construct /sysroot mounts in
/etc/fstab in the initrd, we want these to be mounted before the
initrd-root-fs.target is active.
initrd.target can be the default target in the initrd.
(normal startup)
:
:
v
basic.target
|
______________________/|
/ |
| sysroot.mount
| |
| v
| initrd-root-fs.target
| |
| v
| initrd-parse-etc.service
(custom initrd services) |
| v
| (sysroot-usr.mount and
| various mounts marked
| with fstab option
| x-initrd.mount)
| |
| v
| initrd-fs.target
| |
\______________________ |
\|
v
initrd.target
|
v
initrd-cleanup.service
isolates to
initrd-switch-root.target
|
v
______________________/|
/ |
| initrd-udevadm-cleanup-db.service
| |
(custom initrd services) |
| |
\______________________ |
\|
v
initrd-switch-root.target
|
v
initrd-switch-root.service
|
v
switch-root
Back from old times when we developed systemd on non-systemd hosts we
still mounted the missing directories such as the cgroup stuff even when
running with a PID != 1. There's no point for that anymore, so let's
just do that if we are actually PID 1, and never otherwise.
https://bugs.freedesktop.org/show_bug.cgi?id=62354
There are very few differences in the implementations of the kill method in the
unit types that have one. Let's unify them.
This does not yet unify unit_kill() with unit_kill_context().
Instead of using local-fs*.target in the initrd, use root-fs.target for
sysroot.mount and initrd-fs.target for /sysroot/usr and friends.
Using local-fs.target would mean to carry over the activated
local-fs.target to the isolated initrd-switch-root.target and thus in
the real root. Having local-fs.target already active after
deserialization causes ordering problems with the real root services and
targets.
We better isolate to targets for initrd-switch-root.target, which are
only available in the initrd.
BogdanR> I think it's cool it supports SMACK and that it encourages
them to use a propper mount point for smackfs but I don't
think it's cool that it's printing on the screen even when
I parse quiet to the kernel that "SMACK support is not
enabled ...".
In order to maintain compatibility with older initrds which do not have
AllowIsolate=yes set for their target units, fallback to JOB_REPLACE if
JOB_ISOLATE doesn't work, but complain about it.
SMACK is the Simple Mandatory Access Control Kernel, a minimal
approach to Access Control implemented as a kernel LSM.
The kernel exposes the smackfs filesystem API through which access
rules can be loaded. At boot time, we want to load the access rules
as early as possible to ensure all early boot steps are checked by Smack.
This patch mounts smackfs at the new location at /sys/fs/smackfs for
kernels 3.8 and above. The /smack mountpoint is not supported.
After mounting smackfs, rules are loaded from the usual location.
For more information about Smack see:
http://www.kernel.org/doc/Documentation/security/Smack.txt
Dropping the distribution specific #ifdefs in
88516c0c95 broke the .sh suffix stripping
since we now always used the else clause of the rc. check.
We eventually want to drop the rc. prefix stripping, but for now we
assume that no sysv init script uses both an rc. prefix and .sh suffix,
so make the check for the .sh suffix and rc. prefix mutually exclusive.
If XDG_RUNTIME_DIR contains a character like ":" (for instance if it's
formed from an X11 display name), then it isn't valid to substitute
it into a D-Bus address without escaping.
http://bugs.freedesktop.org/show_bug.cgi?id=60499
This allows switch-root to work correctly if a unit is active both before and
after the switch-root, but its dependencies change. Before the patch, any
dependencies added to active units by switch-root will not be pulled, in
particular filesystems configured in /etc/fstab would not be activated if
local-fs.target was active in the initrd.
It is not clear to me if there is a bug in the REPLACE handling, or if it is
working as expected and that we really want to use ISOLATE instead as this patch
does.
Harald encountered division by zero in manager_print_jobs_in_progress.
Clearly we had the watch enabled when we shouldn't - there were no
running jobs in m->jobs, only waiting ones. This is either a deadlock,
or maybe some of them would be detected as runnable in the next dispatch
of the run queue. In any case we mustn't crash.
Fix it by starting and stopping the watch based on n_running_jobs
instead of the number of all jobs.
When watches are installed from the bottom, it is always possible
to race, and miss a file creation event. The race can be avoided
if a watch is first established for a parent directory, and then for
the file in the directory. If the file is created in the time between,
the watch on the parent directory will fire.
Some messages (mostly at debug level) are added to help diagnose
pidfile issues.
Should fix https://bugzilla.redhat.com/show_bug.cgi?id=917075.
Errors because of oom conditions or descriptor exhaustion should not
be ignored. We probably cannot recover from those conditions.
Current behaviour wrt. insufficient permissions is described in the
man page. It might make sense in case of user sessions, so I left
it as is.
The "OK" status messages should not draw attention to themselves.
It's better if they're not printed in bright/bold. Leave that
to errors and warnings.
Use a plain inconspicuous enterprisey green.
All active units will call unit_notify() during coldplug, so we just
make sure we're counting from zero again and get the correct result for
n_on_console.
For n_running_jobs we likewise reset it to zero and then count
the running jobs as we encounter them in deserialization.
unit_notify is fired in deserelization code (particulary in
service_set_state). Units passed in random order, and there is possibility,
that unit with StopWhenUnneeded=yes passed before it actual dependecies. In
that case unit will be stopped as unneeded, because deps in UNIT_INACTIVE state
yet.
So, reuse similar logic (unit.c:1421) to avoid this race
Sometimes the boot gets stuck until a timeout hits. The usual timeouts
are on the order of minutes, so users may lose patience.
Print animated status messages telling the names of units with running
jobs to make it easy to see what systemd is waiting for.
The animation looks cooler with a shorter interval, but 1 s is OK and
should not be too hard on slow serial console users.
Similar to already existing is_terminal_input().
Note that the only current user (connect_logger_as) is never called
for EXEC_OUTPUT_TTY, so it won't mind whether we accept it.
Zbigniew Jędrzejewski-Szmek
Michal Sekletar
Nathaniel Chen
Umut Tezduyar
Lennart Poettering
Harald Hoyer
Michal Schmidt
Michael Biebl
Simon McVittie
Enrico Scholz
Tom Gundersen
Lukas Nykryn
Oleksii Shevchuk
Frederic Crozat