To make debugging easier, this patches allows one to change the log target and
do reload/reexec without modifying configuration permanently, which makes
debugging easier.
Indeed if one changed the log target at runtime (via the bus or via signals),
the change was lost on the next reload/reexecution.
In order to restore back the default value (set via system.conf, environment
variables or any other means ), the empty string in the "LogTarget" property is
now supported as well as sending SIGTRMIN+26 signal.
To make debugging easier, this patches allows one to change the log level and
do reload/reexec without modifying configuration permanently, which makes
debugging easier.
Indeed if one changed the log max level at runtime (via the bus or via
signals), the change was lost on the next daemon reload/reexecution.
In order to restore the original value back (set via system.conf, environment
variables or any other means), the empty string in the "LogLevel" property is
now supported as well as sending SIGRTMIN+23 signal.
Since bb28e68477 parsing failures of
certain unit file settings will result in load failures of units. This
introduces a new load state "bad-setting" that is entered in precisely
this case.
With this addition error messages on bad settings should be a lot more
explicit, as we don't have to show some generic "errno" error in that
case, but can explicitly say that a bad setting is at fault.
Internally this unit load state is entered as soon as any configuration
loader call returns ENOEXEC. Hence: config parser calls should return
ENOEXEC now for such essential unit file settings. Turns out, they
generally already do.
Fixes: #9107
These states should never be visible to the outside, as they are used
only internally while loading unit. Hence let's drop them from the
documentation.
This adds a function sd_bus_slot_set_destroy_callback() to set a function
which can free userdata or perform other cleanups.
sd_bus_slot_get_destory_callback() queries the callback, and is included
for completeness.
Without something like this, for floating asynchronous callbacks, which might
be called or not, depending on the sequence of events, it's hard to perform
resource cleanup. The alternative would be to always perform the cleanup from
the caller too, but that requires more coordination and keeping of some shared
state. It's nicer to keep the cleanup contained between the callback and the
function that requests the callback.
The only difference is that functions are not individually listed by name,
but that seems completely pointless, since all functions that are documented
are always exported, so the generic text tells the user all she or he needs
to know.
Fedora 28 is out already, let's advertise it. While at it, drop "container"
from "f28container" — it's a subdirectory under /var/lib/machines, it's pretty
obvious that's it a container.
To make the switch easier in the future, define the number as an entity.
This "netdevsim" as implied by the name is a tool for network developers and is a simulator.
This simulated networking device is used for testing various networking APIs and at this time
is particularly focused on testing hardware offloading related interfaces.
'systemctl disable --runtime' would disable a unit, but only if it was enabled
with '--runtime', and silently do nothing if the unit was enabled persistently.
And similarly 'systemctl disable' would do nothing if the unit was enabled in
/run. This just doesn't seem useful.
This pathch changes enable/disable and mask/unmask to be asymmetrical. enable
and mask create symlinks in /etc or /run, depending on whether --runtime was
specified. disable and unmask remove symlinks from both locations. --runtime
cannot be specified for the disable and unmask verbs.
The advantage is that 'disable' now means that the unit is disabled, period.
And similarly for 'unmask', all masks are removed.
Similarly for preset and preset-all, they now cannot be called with --runtime,
and are asymmetrical: when they enable a unit, symlinks are created in /etc.
When they disable a unit, all symlinks are nuked.
$ systemctl --root=/ enable bluetooth
Created symlink /etc/systemd/system/dbus-org.bluez.service → /usr/lib/systemd/system/bluetooth.service.
Created symlink /etc/systemd/system/bluetooth.target.wants/bluetooth.service → /usr/lib/systemd/system/bluetooth.service.
$ systemctl --root=/ --runtime enable bluetooth
Created symlink /run/systemd/system/dbus-org.bluez.service → /usr/lib/systemd/system/bluetooth.service.
Created symlink /run/systemd/system/bluetooth.target.wants/bluetooth.service → /usr/lib/systemd/system/bluetooth.service.
$ systemctl --root=/ disable bluetooth
Removed /run/systemd/system/bluetooth.target.wants/bluetooth.service.
Removed /run/systemd/system/dbus-org.bluez.service.
Removed /etc/systemd/system/bluetooth.target.wants/bluetooth.service.
Removed /etc/systemd/system/dbus-org.bluez.service.
$ systemctl --root=/ disable --runtime bluetooth
--runtime cannot be used with disable
$ systemctl --root=/ mask --runtime bluetooth
Created symlink /run/systemd/system/bluetooth.service → /dev/null.
$ systemctl --root=/ mask bluetooth
Created symlink /etc/systemd/system/bluetooth.service → /dev/null.
$ systemctl --root=/ unmask bluetooth
Removed /run/systemd/system/bluetooth.service.
Removed /etc/systemd/system/bluetooth.service.
$ systemctl --root=/ unmask --runtime bluetooth
--runtime cannot be used with unmask
$ systemctl --root=/ --runtime enable bluetooth
Created symlink /run/systemd/system/dbus-org.bluez.service → /usr/lib/systemd/system/bluetooth.service.
Created symlink /run/systemd/system/bluetooth.target.wants/bluetooth.service → /usr/lib/systemd/system/bluetooth.service.
$ systemctl --root=/ enable bluetooth
Created symlink /etc/systemd/system/dbus-org.bluez.service → /usr/lib/systemd/system/bluetooth.service.
Created symlink /etc/systemd/system/bluetooth.target.wants/bluetooth.service → /usr/lib/systemd/system/bluetooth.service.
$ systemctl --root=/ preset bluetooth
Removed /run/systemd/system/bluetooth.target.wants/bluetooth.service.
Removed /run/systemd/system/dbus-org.bluez.service.
Removed /etc/systemd/system/bluetooth.target.wants/bluetooth.service.
Removed /etc/systemd/system/dbus-org.bluez.service.
$ systemctl --root=/ preset --runtime bluetooth
--runtime cannot be used with preset
$ systemctl preset-all --runtime
--runtime cannot be used with preset-all
This rewords the section, explicitly distuingishing the cases of clients
that only want a continious log stream (which can simply treat
SD_JOURNAL_INVALIDATE the same way as SD_JOURNAL_APPEND) and those which
want to represent on screen the full state of the log data on disk.
This is an alternative to a part of PR #9060, but keeps an explanation
of the destinction of handling depending on the type of client.
Fixes: #8963
We currently return -ENOMEDIUM when /etc/machine-id is empty, and -EINVAL when
it is all zeros. But -EINVAL is also used for invalid args. The distinction
between empty and all-zero is not very important, let's use the same return
code.
Also document -ENOENT and -ENOMEDIUM since they can be a bit surprising.
This corresponds nicely with the specifiers we already pass for
/var/lib, /var/cache, /run and so on.
This is particular useful to update the test-path service files to
operate without guessable files, thus allowing multiple parallel
test-path invocations to pass without issues (the idea is to set $TMPDIR
early on in the test to some private directory, and then only use the
new %T or %V specifier to refer to it).
Usually, we order our settings in our unit files in a logical order,
grouping related settings together, and putting more relevant stuff
first, instead of following a strictly alphabetical order.
For specifiers I think it makes sense to follow an alphabetical order
however, since they literally are just characters, and hence I think the
concept of alphabetical ordering is much more commanding for them. Also,
since specifiers are usually not used in combination, but mostly used
indepdently of each other I think it's not that important to group
similar ones together.
No other changes except the reordering.
When dealing with a large number of template instances, for example
when launching daemons per VRF, it is hard for operators to correlate
log lines to arguments.
Add a new with-unit mode which, if available, prefixes unit and user
unit names when displaying its log messages instead of the syslog
identifier. It will also use the full timestamp with timezones, like
the short-full mode.
sysexits.h has:
#define EX_CANTCREAT 73 /* can't create (user) output file */
EX_DATAERR is a copy-paste error from the previous sentence, which is
correct.
Similar as the other options added before, this is primarily useful to
provide comprehensive OCI runtime compatbility, but might be useful
otherwise, too.
This simply controls the PR_SET_NO_NEW_PRIVS flag for the container.
This too is primarily relevant to provide OCI runtime compaitiblity, but
might have other uses too, in particular as it nicely complements the
existing --capability= and --drop-capability= flags.
Previously, the container's hostname was exclusively initialized from
the machine name configured with --machine=, i.e. the internal name and
the external name used for and by the container was synchronized. This
adds a new option --hostname= that optionally allows the internal name
to deviate from the external name.
This new option is mainly useful to ultimately implement the OCI runtime
spec directly in nspawn, but it might be useful on its own for some
other usecases too.
This ensures we set the various resource limits of our container
explicitly on each invocation so that we inherit less from our callers
into the payload.
By default resource limits are now set to the same values Linux
generally passes to the host PID 1, thus minimizing needless differences
between host and container environments.
The limits are now also configurable using a new --rlimit= switch. This
is preparation for teaching nspawn native OCI runtime support as OCI
permits setting resource limits for container payloads, and it hence
probably makes sense if we do too.
What the man page said was different than what the code did.
save_external_coredump() will store the core temporarily for backtrace
generation, and will delete if afterwards if it is too large. So to disable
processing, it's necessary to both set
Storage=none/Storage=journal+JournalSizeMax=0/Storage=external+ExternalSizeMax=0
and ProcessSizeMax=0. This updates the man page to reflect the code.
The man pages are extended to describe that Storage=none + ProcessSizeMax=0 is
the simplest way to disable coredump processing. All the storage and processing
options make this quite complicated, so let's add a copy-and-pasteable example
of how to disable coredump. Doing it through coredump.conf has the advantage
that we still log, and the effect is immediate, unlike masking the sysconf
file.
Fixes#8788.
Commenting out "WatchdogTimeout=3min" in systemd-logind.service causes
NotifyAccess to go from "main" to "none", breaking support for logind
restart. Let's fix that.
Since StandardOutput=file:path is more similar to StandardInput= than
StandardInputText=, and only StandardInput= is actually documented above
StandardOutput= whereas StandardInputText= is documented below it, I
assume the intention was to refer to the former.
While set of systemd-journal-{gatewayd,remote,upload}.service services presents single subsystem on journald logs network transmission, systemd-journal-gatewayd.service description should also contain links to other parts of this subsystem: systemd-journal-remote.service and systemd-journal-upload.service.
* man: systemd-networkd-wait-online: systemd.service
While service type is mentioned (is a oneshot system service), link on systemd.service is added. 'See Also' section is also updated with link on systemd.service man-page.
Added short keys -u and -m for --unescape and --mangle respectively. These short keys are present in systemd-escape --help output and are absent in man systemd-escape page.
Add journal-upload.conf refentrytitle to have the same format to systemd-journal-remote.service description, which contains refentrytitle on journal-remote.conf in 'See Also' section.
This patch add support to enables to send User Class option code 77
RFC 3004.
This option MAY carry multiple User Classes.
The format of this option is as follows:
Code Len Value
+-----+-----+--------------------- . . . --+
| 77 | N | User Class Data ('Len' octets) |
+-----+-----+--------------------- . . . --+
where Value consists of one or more instances of User Class Data.
Each instance of User Class Data is formatted as follows:
UC_Len_i User_Class_Data_i
+--------+------------------------ . . . --+
| L_i | Opaque-Data ('UC_Len_i' octets) |
+--------+------------------------ . . . --+
UserClass=
A DHCPv4 client can use UserClass option to identify the type or category of user or applications
it represents. The information contained in this option is an string that represents the user class
of which the client is a member. Each class sets an identifying string of information to be used by the DHCP service to classify clients. Takes a whitespace-separated list.
UserClass= hello world how are you
Closes: RFC: #5134
Our CODING_STYLE document suggests to suffix all paths referring to dirs
rather than regular files with a "/" in our docs and log messages.
Update file-hierarchy(7) to do just that.
No other changes.
We document this further down in the text, but let's also list this
early on, where we mention the FHS as major influence too, so that it is
clear we incorporate all that thinking.
Previously, reading through systemd.exec(5) one might get the idea that
XDG_SEAT and XDG_VTNR are part of the service management logic, but they
are not, they are only set if pam_systemd is part of a PAM stack an
pam_systemd is used.
Hence, let's drop these env vars from the list of env vars, and instead
add a paragraph after the list mentioning that pam_systemd might add
more systemd-specific env vars if included in the PAM stack for a
service that uses PAMName=.
Rather than choosing to set or unset any of these flag
use kernel defaults. This patch makes following properties to unset.
UseBPDU = unset
HairPin = unset
FastLeave = unset
AllowPortToBeRoot = unset
UnicastFlood = unset
This implements similar logic as conf_files_cat(), but with slightly different
file gathering logic. I also want to add support for replacement files later on,
so it seems better to keep those two file-gathering functions separate.
This is used as 'systemd-analyze show-config systemd/logind.conf', which
will dump
/etc/systemd/system/user@.service
/etc/systemd/system/user@.service.d/*.conf
/run/systemd/system/user@.service.d/*.conf
/usr/local/lib/systemd/system/user@.service.d/*.conf
/usr/lib/systemd/system/user@.service.d/*.conf
The idea is to make it easy to dump the configuration using the same locations
and order that systemd programs use themselves (including masking, in the right
order, etc.). This is the generic variant that works with any configuration
scheme that follows the same general rules:
$ systemd-analyze cat-config systemd/system.conf
$ systemd-analyze cat-config systemd/user.conf
$ systemd-analyze cat-config systemd/logind.conf
$ systemd-analyze cat-config systemd/sleep.conf
$ systemd-analyze cat-config systemd/journald.conf
$ systemd-analyze cat-config systemd/journal-remote.conf
$ systemd-analyze cat-config systemd/journal-upload.conf
$ systemd-analyze cat-config systemd/coredump.conf
$ systemd-analyze cat-config systemd/resolved.conf
$ systemd-analyze cat-config systemd/timesyncd.conf
$ systemd-analyze cat-config udev/udev.conf
This removes the UserTasksMax= setting in logind.conf. Instead, the generic
TasksMax= setting on the slice should be used. Instead of a transient unit we
use a drop-in to tweak the default definition of a .slice. It's better to use
the normal unit mechanisms instead of creating units on the fly. This will also
make it easier to start user@.service independently of logind, or set
additional settings like MemoryMax= for user slices.
The setting in logind is removed, because otherwise we would have two sources
of "truth": the slice on disk and the logind config. Instead of trying to
coordinate those two sources of configuration (and maintainer overrides to
both), let's just convert to the new one fully.
Right now now automatic transition mechanism is provided. logind will emit a
hint when it encounters the setting, but otherwise it will be ignored.
Fixes#2556.
Those are quite similar to %i/%I, but refer to the last dash-separated
component of the name prefix.
The new functionality of dash-dropins could largely supersede the template
functionality, so it would be tempting to overload %i/%I. But that would
not be backwards compatible. So let's add the two new letters instead.
The description in the man page disagreed with the code. Let the code win,
since if anybody is using this, they are more likely to depend on actual
behaviour rather than the docs. (In Fedora workstation installation there's
only one use, and it doesn't make much sense either way: SyslogIdentifier=%N
in xfs_scrub@.service.)
Also adds dots at the end everywhere, because we have multiple sentences in
some explanations, so we need dots.
- Add a new flag --strict to tell systemd-hwdb to return a
non-zero code on error.
- Make systemd-hwdb update return an error when any parsing
error occurs (only if strict flag is set).
Newer terminals (in particular gnome-terminal) understand special escape
sequence for formatting clickable links. Let's support that to make our
tool output more clickable where that's appropriate.
For details see this:
https://gist.github.com/egmontkob/eb114294efbcd5adb1944c9f3cb5feda
The one big issue is that 'less' currently doesn't grok this, and
doesn't ignore sequence like regular terminal implementations do if they
don't support it. Hence for now, let's disable URL output if a pager is
used. We should revisit that though as soon as less added support for it
and enough time passed for it to enter various distributions.
Double newlines (i.e. one empty lines) are great to structure code. But
let's avoid triple newlines (i.e. two empty lines), quadruple newlines,
quintuple newlines, …, that's just spurious whitespace.
It's an easy way to drop 121 lines of code, and keeps the coding style
of our sources a bit tigther.
First of all, it's frickin' ugly and wrong, as IPC sockets should be
placed in /run and definitely not under a guessable name under
world-writable /tmp. Secondly, it can't even work as we set
PrivateTmp=yes on the service.
Hence, let's clean up the example, and simply use a socket in /run
instead.
Fixes: #8419
We have a common parser, but for the user it might be
completely unobvious that the same general rules apply
to all those files. Let's add a page about the basic syntax
so that the more specific pages don't have to repeat those
details.
Absolute paths make everything simple and quick, but sometimes this requirement
can be annoying. A good example is calling 'test', which will be located in
/usr/bin/ or /bin depending on the distro. The need the provide the full path
makes it harder a portable unit file in such cases.
This patch uses a fixed search path (DEFAULT_PATH which was already used as the
default value of $PATH), and if a non-absolute file name is found, it is
immediately resolved to a full path using this search path when the unit is
loaded. After that, everything behaves as if an absolute path was specified. In
particular, the executable must exist when the unit is loaded.
Systems that have an accurate real-time clock may have an initial
unsynchronized time that is close enough to the synchronized time that
the final adjustment doesn't trigger a waking "clock set" event. Have
timesyncd touch a file in its runtime directory as a secondary signal
for synchronization. Continue to support the timerfd-based trigger as a
sufficient condition when the watchfile is not present.
Closes issue #8683
Clarify the helper/checker terminology in the systemd-fsck@.service manpage to
make the description more clear about what is responsible for deciding if a filesystem
needs checking.
Files which are installed as-is (any .service and other unit files, .conf
files, .policy files, etc), are left as is. My assumption is that SPDX
identifiers are not yet that well known, so it's better to retain the
extended header to avoid any doubt.
I also kept any copyright lines. We can probably remove them, but it'd nice to
obtain explicit acks from all involved authors before doing that.
Things can fail, and we have no control over it:
- file system issues (immutable bits, file system errors, MAC refusals, etc)
- kernel refusing certain arguments when writing to /proc/sys or /sys
Let's add a new code for the case where we parsed configuration but failed
to execute it because of external errors.
Fixes#5387.
I kept the _SYSTEMD_UNIT= example because it is easy to understand and
not very verbose. _SYSTEMD_CGROUP has much longer entries which do not
fit well in the narrow man page. Instead, I added an explanation of what
-u is translated into.
Issue #6673 requests advice on backup strategy. But the right backup strategy
depends on many factors, too many to describe in a man page. So let's just
provide some general information which files are mutable and that it is always
safe to use/copy files.
Closes#6673.
Issue #6499 requests that a mention that those varibles can be set in the
environment is added. But the man page already says that. There isn't much
detail, but a man page does not need to and in this case should not include
all the details. Instead a note is added that those vars can be derived from
$DISPLAY.
Closes#6499.
We're moving towards just SPDX license identifiers, and the boilerplate
is especially annoying in a man page. Also adjust to the smaller indentation
to make the code fit better on a page.
man/.dir-locals is to keep indentation under control.
This makes it much easier to compile and run those examples, c.f. #7578.
v2:
- copy more of .dir-locals.el from the root to man/.dir-locals.el
(I though emacs would inherit from the one in the parent dir, but
it seems it just uses its own broken defaults, including
indent-tabs-mode by default.)
Support was killed in kernel 4.15 as well as ethtool 4.13.
Justification was lack of use by drivers and too much of a maintenance burden.
https://www.spinics.net/lists/netdev/msg443815.html
Also moved config_parse_warn_compat to conf-parser.[ch] to fix compile errors.
This makes users can configure DHCPv4 client with ClientIdentifier=duid-only.
If set so, then DHCP client sends only DUID as the client identifier.
This may not be RFC compliant, but some setups require this.
Closes#7828.
Suspend to Hibernate is a new sleep method that invokes suspend
for a predefined period of time before automatically waking up
and hibernating the system.
It's similar to HybridSleep however there isn't a performance
impact on every suspend cycle.
It's intended to use with systems that may have a higher power
drain in their supported suspend states to prevent battery and
data loss over an extended suspend cycle.
Signed-off-by: Mario Limonciello <mario.limonciello@dell.com>
This PR implements the first part of RFE #8046. I.e. this allows to
write:
```
u username -:300
```
Where the uid is chosen automatically but the gid is fixed.
In particular:
- drop "when it is non-zero" to avoid implying that it can be called if the
queue is not empty.
- "has been created" sounds like something happened in parallel,
but what we really mean is that *this* particular object *was* created in a
certain way.
NAME is kind of meaningless, because everything has a name. "Unit"
makes it more obvious that a name of a unit is necessary. I was always
momentarily baffled by "set-property NAME ASSIGNMENT...", where there
are two objects (the unit and the property), and it's not clear which of
the two "NAME" is supposed to signify.
Before this change all unit types would default to "private" in the
system service manager and "inherit" to in the user service manager.
With this change this is slightly altered: non-service units of the
system service manager are now run with KeyringMode=shared. This appears
to be the more appropriate choice as isolation is not as desirable for
mount tools, which regularly consume key material. After all mounts are
a shared resource themselves as they appear system-wide hence it makes a
lot of sense to share their key material too.
Fixes: #8159
After discussions with @htejun it appears it's OK now to enable memory
accounting by default for all units without affecting system performance
too badly. facebook has made good experiences with deploying memory
accounting across their infrastructure.
This hence turns MemoryAccounting= from opt-in to opt-out, similar to
how TasksAccounting= is already handled. The other accounting options
remain off, their performance impact is too big still.
This change adds support for controlling the suspend-on-lid-close
behaviour based on the power status as well as whether the machine is
docked or has an external monitor. For backwards compatibility the new
configuration file variable is ignored completely by default, and must
be set explicitly before being considered in any decisions.
So far we didn't document control, transient, dbus config, or generator paths.
But those paths are visible to users, and they need to understand why systemd
loads units from those paths, and how the precedence hierarchy looks.
The whole thing is a bit messy, since the list of paths is quite long.
I made the tables a bit shorter by combining rows for the alternatives
where $XDG_* is set and the fallback.
In various places, tags are split like <element
param="blah">
this. This is necessary to keep everyting in one logical XML line so that
docbook renders the table properly.
Replaces #8050.
There are cases that we want to trigger and settle only specific
commands. For example, let's say at boot time we want to make sure all
the graphics devices are working correctly because it's critical for
booting, but not the USB subsystem (we'll trigger USB events later). So
we do:
udevadm trigger --action="add" --subsystem-match="graphics"
udevadm settle
However, we cannot block the kernel from emitting kernel events from
discovering USB devices. So if any of the USB kernel event was emitted
before the settle command, the settle command would still wait for the
entire queue to complete. And if the USB event takes a long time to be
processed, the system slows down.
The new `settle` option allows the `trigger` command to wait for only
the triggered events, and effectively solves this problem.
CAP_ADMIN does not exist (the closest existing capability name would be
CAP_SYS_ADMIN), and according to man:open(2) and man:capabilities(7),
the capability required to specify O_NOATIME is actually CAP_FOWNER.
Remote= must be a non multicast address. ip-link(8) says:
> remote IPADDR - specifies the unicast destination IP address to
> use in outgoing packets when the destination link layer address
> is not known in the VXLAN device forwarding database.
Closes#8088.
Changed <filename>.service</filename> to <literal>.service</literal> to match style in other manual pages: man 5 systemd.socket, device, mount, automount, swap, target path, timer, slice and scope.
Changed "reboot" to "power off" in poweroff.target description. It was most likely copied and pasted from the reboot.target below, compare with e.g. halt.target
