We allocated the names set for each unit, but in the majority of cases, we'd
put only one name in the set:
$ systemctl show --value -p Names '*'|grep .|grep -v ' '|wc -l
564
$ systemctl show --value -p Names '*'|grep .|grep ' '|wc -l
16
So let's add a separate .id field, and only store aliases in the set, and only
create the set if there's at least one alias. This requires a bit of gymnastics
in the code, but I think this optimization is worth the trouble, because we
save one object for many loaded units.
In particular set_complete_move() wasn't very useful because the target
unit would always have at least one name defined, i.e. the optimization to
move the whole set over would never fire.
With cgroup v2 the cgroup freezer is implemented as a cgroup
attribute called cgroup.freeze. cgroup can be frozen by writing "1"
to the file and kernel will send us a notification through
"cgroup.events" after the operation is finished and processes in the
cgroup entered quiescent state, i.e. they are not scheduled to
run. Writing "0" to the attribute file does the inverse and process
execution is resumed.
This commit exposes above low-level functionality through systemd's DBus
API. Each unit type must provide specialized implementation for these
methods, otherwise, we return an error. So far only service, scope, and
slice unit types provide the support. It is possible to check if a
given unit has the support using CanFreeze() DBus property.
Note that DBus API has a synchronous behavior and we dispatch the reply
to freeze/thaw requests only after the kernel has notified us that
requested operation was completed.
We should allow the ones that the [Unit] section of regular unit files
may accet, but no other, in particular not the internal deps we
synthesize as reverse of explicitly configured ones, such was WantedBy=.
Fixes: #14251
Introduce support for configuring cpus and mems for processes using
cgroup v2 CPUSET controller. This allows users to limit which cpus
and memory NUMA nodes can be used by processes to better utilize
system resources.
The cgroup v2 interfaces to control it are cpuset.cpus and cpuset.mems
where the requested configuration is written. However, it doesn't mean
that the requested configuration will be actually used as parent cgroup
may limit the cpus or mems as well. In order to reflect the real
configuration cgroup v2 provides read-only files cpuset.cpus.effective
and cpuset.mems.effective which are exported to users as well.
The functions to retrieve and print process cmdlines were based on the
assumption that they contain printable ASCII, and everything else
should be filtered out. That assumption doesn't hold in today's world,
where people are free to use unicode everywhere.
This replaces the custom cmdline reading code with a more generic approach
using utf8_escape_non_printable_full().
For kernel threads, truncation is done on the parenthesized name, so we'll
get "[worker]", "[worker…]", …, "[w…]", "[…", "…" as we reduce the number of
available columns.
This implementation is most likely slower for very long cmdlines, but I don't
think this is very important. The common case is to have short commandlines,
and should print those properly. Absurdly long cmdlines are the exception,
which needs to be handled correctly and safely, but speed is not too important.
Fixes#12532.
v2:
- use size_t for the number of columns. This change propagates into various
other functions that call get_process_cmdline(), increasing the size of the
patch, but the changes are rather trivial.
This was the last kind of accounting still not exposed on for each unit.
Let's fix that.
Note that this is a relatively simplistic approach: we don't expose
per-device stats, but sum them all up, much like cgtop does. This kind
of metric is probably the most interesting for most usecases, and covers
the "systemctl status" output best. If we want per-device stats one day
we can of course always add that eventually.
The error string for operations that are not supported (e.g. "shutdown" for
user-defined units) should take two arguments, where the first one is the type
of action being defined (i.e. "FailureAction" vs. "SuccessAction") and the
second is the string that was invalid.
Currently, the code prints this:
$ systemd-run --user --wait -p SuccessAction=poweroff true
Failed to start transient service unit: EmergencyAction setting invalid for manager type: SuccessAction
Change the format string to instead print:
$ systemd-run --user --wait -p SuccessAction=poweroff true
Failed to start transient service unit: SuccessAction setting invalid for manager type: poweroff
This function returns 0 on success and a negative value on failure. On success,
it writes the parsed action to the address passed in its third argument.
`bus_set_transient_emergency_action` does this:
r = parse_emergency_action(s, system, &v);
if (v < 0)
// handle failure
However, `v` is not updated if the function fails, and this should be checking
`r` instead of `v`.
The result of this is that if an invalid failure (or success) action is
specified, systemd ends up creating the unit anyway and then misbehaves if it
tries to run the failure action because the action value comes from
uninitialized stack data. In my case, this resulted in a failed assertion:
Program received signal SIGABRT, Aborted.
0x00007fe52cca0d7f in raise () from /snap/usr/lib/libc.so.6
(gdb) bt
#0 0x00007fe52cca0d7f in raise () from /snap/usr/lib/libc.so.6
#1 0x00007fe52cc8b672 in abort () from /snap/usr/lib/libc.so.6
#2 0x00007fe52d66f169 in log_assert_failed_realm (realm=LOG_REALM_SYSTEMD, text=0x56177ab8e000 "action < _EMERGENCY_ACTION_MAX", file=0x56177ab8dfb8 "../src/core/emergency-action.c", line=33, func=0x56177ab8e2b0 <__PRETTY_FUNCTION__.14207> "emergency_action") at ../src/basic/log.c:795
#3 0x000056177aa98cf4 in emergency_action (m=0x56177c992cb0, action=2059118610, options=(unknown: 0), reboot_arg=0x0, exit_status=1, reason=0x7ffdd2df4290 "unit run-u0.service failed") at ../src/core/emergency-action.c:33
#4 0x000056177ab2b739 in unit_notify (u=0x56177c9eb340, os=UNIT_ACTIVE, ns=UNIT_FAILED, flags=(unknown: 0)) at ../src/core/unit.c:2504
#5 0x000056177aaf62ed in service_set_state (s=0x56177c9eb340, state=SERVICE_FAILED) at ../src/core/service.c:1104
#6 0x000056177aaf8a29 in service_enter_dead (s=0x56177c9eb340, f=SERVICE_SUCCESS, allow_restart=true) at ../src/core/service.c:1712
#7 0x000056177aaf9233 in service_enter_signal (s=0x56177c9eb340, state=SERVICE_FINAL_SIGKILL, f=SERVICE_SUCCESS) at ../src/core/service.c:1854
#8 0x000056177aaf921b in service_enter_signal (s=0x56177c9eb340, state=SERVICE_FINAL_SIGTERM, f=SERVICE_SUCCESS) at ../src/core/service.c:1852
#9 0x000056177aaf8eb3 in service_enter_stop_post (s=0x56177c9eb340, f=SERVICE_SUCCESS) at ../src/core/service.c:1788
#10 0x000056177aaf91eb in service_enter_signal (s=0x56177c9eb340, state=SERVICE_STOP_SIGKILL, f=SERVICE_SUCCESS) at ../src/core/service.c:1850
#11 0x000056177aaf91bc in service_enter_signal (s=0x56177c9eb340, state=SERVICE_STOP_SIGTERM, f=SERVICE_FAILURE_EXIT_CODE) at ../src/core/service.c:1848
#12 0x000056177aaf9759 in service_enter_running (s=0x56177c9eb340, f=SERVICE_FAILURE_EXIT_CODE) at ../src/core/service.c:1941
#13 0x000056177ab005b7 in service_sigchld_event (u=0x56177c9eb340, pid=112, code=1, status=1) at ../src/core/service.c:3296
#14 0x000056177aad84b5 in manager_invoke_sigchld_event (m=0x56177c992cb0, u=0x56177c9eb340, si=0x7ffdd2df48f0) at ../src/core/manager.c:2444
#15 0x000056177aad88df in manager_dispatch_sigchld (source=0x56177c994710, userdata=0x56177c992cb0) at ../src/core/manager.c:2508
#16 0x00007fe52d72f807 in source_dispatch (s=0x56177c994710) at ../src/libsystemd/sd-event/sd-event.c:2846
#17 0x00007fe52d730f7d in sd_event_dispatch (e=0x56177c993530) at ../src/libsystemd/sd-event/sd-event.c:3229
#18 0x00007fe52d73142e in sd_event_run (e=0x56177c993530, timeout=18446744073709551615) at ../src/libsystemd/sd-event/sd-event.c:3286
#19 0x000056177aad9f71 in manager_loop (m=0x56177c992cb0) at ../src/core/manager.c:2906
#20 0x000056177aa7c876 in invoke_main_loop (m=0x56177c992cb0, ret_reexecute=0x7ffdd2df4bff, ret_retval=0x7ffdd2df4c04, ret_shutdown_verb=0x7ffdd2df4c58, ret_fds=0x7ffdd2df4c70, ret_switch_root_dir=0x7ffdd2df4c48, ret_switch_root_init=0x7ffdd2df4c50, ret_error_message=0x7ffdd2df4c60) at ../src/core/main.c:1792
#21 0x000056177aa7f251 in main (argc=2, argv=0x7ffdd2df4e78) at ../src/core/main.c:2573
Fix this by checking the correct variable.
Let's inform the clients about assert/condition property changes as they
happen, it's basically for free because assert/condition property
changes generally coincide with other unit state changes (after all
these checks are done on unit_start())
When a client requests a new job, let's make sure we for out the JobNew
signals for it, before we return successfully from the method call.
After all we shouldn't return a path that is not announced yet, as
announcement of jobs should be considered part of the job setup.
This allows clients to follow our internal state changes safely.
Previously, quick state changes (for example, when we restart a unit due
to Restart= after it quickly transitioned through DEAD/FAILED states)
would be coalesced into one bus signal event, with this change there's
the guarantee that all state changes after the unit was announced ones
are reflected on th bus.
Note we only do this kind of guaranteed flushing only for unit state
changes, not for other unit property changes, where clients still have
to expect coalescing. This is because the unit state is a very
important, high-level concept.
Fixes: #10185
This adds SuccessActionExitStatus= and FailureActionExitStatus= that may
be used to configure the exit status to propagate in when
SuccessAction=exit or FailureAction=exit is used.
When not specified let's also propagate the exit status of the main
process we fork off for the unit.
This is useful for debugging client-side ref counting of units: for each
ref taken on a unit the client's sender name is listed. If a client has
multiple refs on the same unit it is listed multiple times.
"killing" is very UNIX terminology, and not really what this is about.
Let's be more correct and say "send a UNIX signal" for the operation.
Otherwise things are really weird if users call "journalctl --rotate"
from the command line, which internally asks systemd to send SIGUSR2 to
to journald: when german locale is selected this asks the user — roughly
transliterated — whether they want to "eliminate" journald, which is
definitely not the intended meaning.
We would accept e.g. FailureAction=reboot-force in user units and then do an
exit in the user manager. Let's be stricter, and define "exit"/"exit-force" as
the only supported actions in user units.
v2:
- rename 'exit' to 'exit-force' and add new 'exit'
- add test for the parsing function
These lines are generally out-of-date, incomplete and unnecessary. With
SPDX and git repository much more accurate and fine grained information
about licensing and authorship is available, hence let's drop the
per-file copyright notice. Of course, removing copyright lines of others
is problematic, hence this commit only removes my own lines and leaves
all others untouched. It might be nicer if sooner or later those could
go away too, making git the only and accurate source of authorship
information.
This part of the copyright blurb stems from the GPL use recommendations:
https://www.gnu.org/licenses/gpl-howto.en.html
The concept appears to originate in times where version control was per
file, instead of per tree, and was a way to glue the files together.
Ultimately, we nowadays don't live in that world anymore, and this
information is entirely useless anyway, as people are very welcome to
copy these files into any projects they like, and they shouldn't have to
change bits that are part of our copyright header for that.
hence, let's just get rid of this old cruft, and shorten our codebase a
bit.
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
The load_error is only valid in some load_state cases, lets generate
prettier messages for other cases too, by reusing the
bus_unit_validate_load_state() call which does jus that.
Clients (such as systemctl) ignored LoadError unles LoadState was
"error" before. With this change they could even show LoadError in other
cases and it would show a useful name.
Let's use a switch() statement, cover more cases with pretty messages.
Also let's rename it to "validate", as that's more specific that
"check", as it implies checking for a "valid"/"good" state, which is
what this function does.
