Yu Watanabe [Sat, 23 Dec 2017 10:10:24 +0000 (19:10 +0900)]
cgroup: IODeviceWeight= or friends can take device node files in /run/systemd/inaccessible/
systemd creates several device nodes in /run/systemd/inaccessible/.
This makes CGroup's settings related to IO can take device node
files in the directory.
Yu Watanabe [Tue, 19 Dec 2017 02:05:43 +0000 (11:05 +0900)]
core,seccomp: fix logic to parse RestrictAddressFamilies= in dbus-execute.c
If multiple RestrictAddressFamilies= settings, some of them are
whitelist and the others are blacklist, are sent to bus, then parsing
result was corrupted.
This fixes the parse logic, now it is the same as one used in
load-fragment.c
Yu Watanabe [Sat, 23 Dec 2017 09:45:32 +0000 (18:45 +0900)]
core,seccomp: fix logic to parse syscall filter in dbus-execute.c
If multiple SystemCallFilter= settings, some of them are whitelist
and the others are blacklist, are sent to bus, then the parse
result was corrupted.
This fixes the parse logic, now it is the same as one used in
load-fragment.c
analyze: use normal bus connection for "plot" verb (#7685)
We need to connect to hostnamed, so a private bus connection is no good.
It'd be simpler to use the normal bus connection unconditionally, but
that'd mean that e.g. systemd-analyze set-log-level might not work in
emergency mode. So let's keep trying to use the private connection except
for "plot".
cryptsetup-generator: Don't mistake NULL input as OOM (#7688)
Since systemd v236, several Arch users complained that
systemd-cryptsetup-generator exits with an OOM error and that it
prevents the boot from continuing.
Investigating the diff of cryptsetup-generator between v235 and v236 I
noticed that create_disk allowed for the `password` and `filtered`
variables to be NULL (they're handled with `strempty()`) but not their
`*_escaped` versions, and returned OOM errors in those cases.
Fix this by checking that the input string is non-NULL before deciding
that `specifier_escape` had an OOM error.
I could not test this fix myself, but some users have reported success.
Up until now, the behaviour in systemd has (mostly) been to silently
ignore failures to action unit directives that refer to an unavailble
controller. The addition of AssertControlGroupController and its
conditional counterpart allow explicit specification of the desired
behaviour when such a situation occurs.
As for how this can happen, it is possible that a particular controller
is not available in the cgroup hierarchy. One possible reason for this
is that, in the running kernel, the controller simply doesn't exist --
for example, the CPU controller in cgroup v2 has only recently been
merged and was out of tree until then. Another possibility is that the
controller exists, but has been forcibly disabled by `cgroup_disable=`
on the kernel command line.
In future this will also support whatever comes out of issue #7624,
`DefaultXAccounting=never`, or similar.
Alan Jenkins [Sat, 16 Dec 2017 10:48:12 +0000 (10:48 +0000)]
man: generalize "binary" to "program" (#7668)
Systemd services are permitted to be scripts, as well as binary
executables.
The same also applies to the underlying /sbin/mount and /sbin/swapon.
It is not necessary for the user to consider what type of program file
these are. Nor is it necessary with systemd-nspawn, to distinguish between
init as a "binary" v.s. a user-specified "program".
Also fix a couple of grammar nits in the modified sentences.
This code is executed before we parse command line/configuration
parameters, hence let's not use arg_system to figure our how to clean up
things, but instead PID == 1. Let's move that check inside of the
function, to make things a bit more robust abstract from the outside.
Also, let's add a log message about this, that was so far missing.
main: move install_crash_handler() and mount_cgroup_controllers() invocations
Let's place them in initialize_runtime(), where they appear to fit best.
Effectively this is just a move a little bit down, swapping places with
log_execution_mode(), which should require neither call to be done
first.
Note that changes the conditionalization a bit for these calls, from
(PID == 1) to (arg_system && arg_action == ACTION_RUN). At this point this is pretty much the same
however, as we don't allow PID 1 without ACTION_RUN and without
arg_system set, safety_checks() ensures that.
core: move arg_show_status fix-up into load_configuration()
It's part of finalizing our runtime parameters, hence let's move this
into load_configuration() after we loaded everything else. This is safe,
since we don't use it between the location where it was and where we
place it now yet.
We need to specify a full path to the "ip" binary and busybox "ip" has a
slightly different output than the normal ip, and won't show "DOWN".
hence instead ensure that at lest not "UP" is in there.
The kernel needs two numbers, but for the user it's most convenient to provide the
user name and have that resolved to uid and gid.
Right now the primary group of the specified user is always used. That's the most
common case anyway. In the future we can extend the --owner option to allow a group
after a colon.
[I added this before realizing that this will not be enough to be used for user
runtime directory. But this seems useful on its own, so I'm keeping this commit.]
tree-wide: add DEBUG_LOGGING macro that checks whether debug logging is on (#7645)
This makes things a bit easier to read I think, and also makes sure we
always use the _unlikely_ wrapper around it, which so far we used
sometimes and other times we didn't. Let's clean that up.