The motivation is that variants of the same distro that share the same compiled
rpm want to customize various aspects of the system, in particular the
hostname. In some sense the default hostname is part of the identity of the
system, so setting it through os-release makes sense. In particular, instead of
setting a default value in /etc/hostname, the appropriate default can be baked
into the image, leaving /etc/hostname for local overrides only.
Why make this a separate field instead of e.g. using NAME from os-release?
NAME is already used for other purposes, and it seems likely that people want
to set those independently.
Quoting https://github.com/systemd/systemd/pull/18704#discussion_r579465254:
> I'd really drop XDG_CONFIG_HOME from the docs. It's confusing enough as it
> is. Where we don't need the indirections we should not confuse people with
> it, in particular as people might then think it's actually a good idea to use
> that env var and redirect things. I'd just show the literal path everywhere,
> even if we internally use the env var.
This is useful for various variables that modify process behaviour. This makes
it easy to set it for pid1 without touching the kernel command line. Even for
the *user manager* this also can be convenient for the unprivileged user, who
cannot modify user@.service definition.
Variables that could be set like this include $SD_EVENT_PROFILE_DELAYS,
$SYSTEMD_FALLBACK_HOSTNAME, $SYSTEMD_MEMPOOL, $SYSTMED_RDRAND, etc.
manager: read ~/.config/systemd/user.conf in user mode
This changes the paths we read user manager config from in two ways:
- split-usr-root paths are dropped. The user manager is a poster boy for
non-early-boot, so reading dropins only from /usr is appropriate.
- we look at ~/.config/systemd/user.conf. Users should be allowed to override
their own config.
As user managers become more and more used, it becomes more important for users
to customize their own daemon. By reading from ~/.config, this is possible
without privileges.
udevadm: after validating action, use our internal string instead of optarg
This doesn't really change anything, but feels nicer, since it abstracts
away what device_action_from_string()/device_action_to_string() do
internally, and always uses a normalized action string (yes, there's no
ambiguity, but it's nice to stay abstract, maybe one day there is
ambiguity around this)
To make sd-device properly usable for all programs we need to provide an
API for the "action" field of an event, it's one of the most relevant
ones, and it was so far missing.
This also adds sd_device_get_seqnum(), which isn't that interesting,
except for generating pretty debug output, which we use it ourselves
for.
This also makes device_new_from_stat_rdev() public, as it is truly
useful, as we can see in our own uses of it, and I think is fairly
generic to show up in the public APIs.
Apparently, in our current public headers (i.e. those called sd-*.h) we
suffixed typedefs that we use as values with _t, but we didn't do this
for enum typedefs. Fix that while this stuff is not actually public yet.
With this scheme "value typedefs" now end systematically in _t, and
"object typedefs" (i.e. structures that are typically passed around via
pointers and not values) do not.
resolved: tweak how we calculate MTU for sending packets
Let's take all MTU info we possibly have into account, i.e. the one
reported via netlink, as before and the one the socket might now (from
PMTUD and such), clamped by our own ideas.
resolved: collect incoming fragment size when receiving UDP datagrams
We can later use this to adapt our announced EDNS buffer size in order
to avoid fragmentation to make the best of large datagrams while still
avoiding he security weaknesses of it.
I'm seeing the following with kernel-core-5.10.16-200.fc33.x86_64:
$ sudo SYSTEMD_LOG_LEVEL=debug build/systemd-rfkill
Reading struct rfkill_event: got 8 bytes.
A new rfkill device has been added with index 0 and type bluetooth.
Found cgroup2 on /sys/fs/cgroup/, full unified hierarchy
Found container virtualization none.
rfkill0: Operating on rfkill device 'tpacpi_bluetooth_sw'.
Writing struct rfkill_event successful (8 of 9 bytes).
Loaded state '0' from /var/lib/systemd/rfkill/platform-thinkpad_acpi:bluetooth.
Reading struct rfkill_event: got 8 bytes.
A new rfkill device has been added with index 1 and type wwan.
rfkill1: Operating on rfkill device 'tpacpi_wwan_sw'.
Writing struct rfkill_event successful (8 of 9 bytes).
Loaded state '0' from /var/lib/systemd/rfkill/platform-thinkpad_acpi:wwan.
Reading struct rfkill_event: got 8 bytes.
A new rfkill device has been added with index 2 and type bluetooth.
rfkill2: Operating on rfkill device 'hci0'.
Writing struct rfkill_event successful (8 of 9 bytes).
Loaded state '0' from /var/lib/systemd/rfkill/pci-0000:00:14.0-usb-0:7:1.0:bluetooth.
Reading struct rfkill_event: got 8 bytes.
A new rfkill device has been added with index 3 and type wlan.
rfkill3: Operating on rfkill device 'phy0'.
Writing struct rfkill_event successful (8 of 9 bytes).
Loaded state '0' from /var/lib/systemd/rfkill/pci-0000:04:00.0:wlan.
All events read and idle, exiting.
We were expecting a read of exactly RFKILL_EVENT_SIZE_V1==8 bytes. But the
structure has 9 after [1].
/* we don't need the 'hard' variable but accept it */
if (count < RFKILL_EVENT_SIZE_V1 - 1)
return -EINVAL;
/*
* Copy as much data as we can accept into our 'ev' buffer,
* but tell userspace how much we've copied so it can determine
* our API version even in a write() call, if it cares.
*/
count = min(count, sizeof(ev));
if (copy_from_user(&ev, buf, count))
return -EFAULT;
... so it should accept the full size. I'm not sure what is going on here.
But we don't care about the extra fields, so let's accept a write as long as
it's at least RFKILL_EVENT_SIZE_V1.
Luca Boccassi [Fri, 12 Feb 2021 15:30:10 +0000 (15:30 +0000)]
os-util: allow missing VERSION_ID on the host
Rolling releases, like ArchLinux, do not set VERSION_ID in
their os-release files, so allow matching simply on ID if the host
does not provide anything.
shell-completion: complete --legend=no for resolvectl and systemctl
I don't think it makes sense to complete --legend=yes. It is the default, and
it would be only used very rarely (and then it is easy enough to just remove
the '=no' part from the suggested string).
systemctl: hide legends with --quiet, allow overriding
--no-legend is replaced by --legend=no.
--quiet now implies --legend=no, but --legend=yes may be used to override that.
--quiet controls hints and warnings and such, and --legend controls just the
legends. I think it makes sense to allow both to controlled independently, in
particular --quiet --legend makes sense when using systemctl in a script to
provide some user-visible output.
journal-remote: convert to parse_boolean_argument() and fix type confusion
We were passing a reference to 'int arg_seal' to config_parse_bool(),
which expects a 'bool *'. Luckily, this would work, because 'bool'
is smaller than 'int', so config_parse_bool() would set the least-significant
byte of arg_seal. At least I think so. But let's use consistent types ;)
Also, modernize style a bit and don't use integers in boolean context.
This nicely covers the case when optarg is optional. The same parser can be
used when the option string passed to getopt_long() requires a parameter and
when it doesn't.
The error messages are made consistent.
Also fixes a log error c&p in --crash-reboot message.
We almost never use the named enum type, in almost all cases we use
"int" instead, since we overload it with negative errnos. To simplify
things, let's use "int" really everywhere.
Moreover, let's rename the fields for this enum to "type_or_errno", to
make the overloading clear. And let's ad some assertions that things are
in the right range.
resolved: in DNSSEC permissive mode, check if DO bit wasn't copied from request to response
If the server doesn't copy the DO bit from request to response, this is
a very early and easy indication that it doesn#t support DNSSEC
properly. Hence, let's immediately downgrade to non-DNSSEC mode if we
see this – if permissive mode is on and this is allowed.
Luca Boccassi [Tue, 16 Feb 2021 23:47:34 +0000 (23:47 +0000)]
test: avoid leaking open loop devices
When a subshell is used ('make' or 'make all') the LOOPDEV environment
variable, which is used to store the opened loop device, is lost.
So the cleanup on trap/exit doesn't do anything, and the loop
device used to mount the test image is left around.
This is an updated version of #8608 with more restrictive logic. To
quite the original bug:
Some captive portals, lie and do not respond with the captive portal
IP address, if the query is with EDNS0 enabled and D0 bit set to
zero. Thus retry "secure" domain name look ups with less secure
methods, upon NXDOMAIN.
Vito Caputo [Tue, 27 Oct 2020 06:24:34 +0000 (23:24 -0700)]
logs-show: move show_journal_by_unit _BOOT_ID match
In scrutinizing the journal overhead of `systemctl status $service`
it became apparent that the matching engine was performing the unit
matches on every journal in my system, even ones containing nothing
relevant to the current boot.
This seemed strange and likely suboptimal to me, since there's likely
far more unit data to rifle through than boot IDs in any given
journal. The _BOOT_ID match seemed like it should be serving as an
early exit match on irrelevant journals, but that wasn't what seemed
to be happening.
As a quick experiment to see if I could get the _BOOT_ID match to be
something along the lines of a higher priority when matching, and try
early exit on these unrelated journals, I moved add_match_this_boot()
to after the unit match adds, inserting a conjunction between them.
The end result seems to be a very substantial performance gain in my
simple uncached tests, and I still get the expected journal output
from the `systemctl status $service` command:
root@localhost:/# echo 2 > /proc/sys/vm/drop_caches
root@localhost:/# time systemctl --no-pager status dbus
● dbus.service - D-Bus System Message Bus
Loaded: loaded (/lib/systemd/system/dbus.service; static; vendor preset: enabled)
Active: active (running) since Sun 2020-10-25 17:03:05 PDT; 1 day 6h ago
Docs: man:dbus-daemon(1)
Main PID: 572 (dbus-daemon)
Memory: 2.8M
CPU: 110ms
CGroup: /system.slice/dbus.service
└─572 /usr/bin/dbus-daemon --system --address=systemd: --nofork --nopidfile --systemd-activation
Oct 25 17:03:05 localhost systemd[1]: Started D-Bus System Message Bus.
Oct 25 17:06:26 localhost dbus[572]: [system] Activating via systemd: service name='org.freedesktop.machine1' unit='dbus-org.freedesktop.machine1.service'
Oct 25 17:06:26 localhost dbus[572]: [system] Successfully activated service 'org.freedesktop.machine1'
real 0m0.695s
user 0m0.005s
sys 0m0.043s
root@localhost:/# echo 2 > /proc/sys/vm/drop_caches
root@localhost:/# time systemctl --no-pager status dbus
● dbus.service - D-Bus System Message Bus
Loaded: loaded (/lib/systemd/system/dbus.service; static; vendor preset: enabled)
Active: active (running) since Sun 2020-10-25 17:03:05 PDT; 1 day 6h ago
Docs: man:dbus-daemon(1)
Main PID: 572 (dbus-daemon)
Memory: 2.8M
CPU: 110ms
CGroup: /system.slice/dbus.service
└─572 /usr/bin/dbus-daemon --system --address=systemd: --nofork --nopidfile --systemd-activation
Oct 25 17:03:05 localhost systemd[1]: Started D-Bus System Message Bus.
Oct 25 17:06:26 localhost dbus[572]: [system] Activating via systemd: service name='org.freedesktop.machine1' unit='dbus-org.freedesktop.machine1.service'
Oct 25 17:06:26 localhost dbus[572]: [system] Successfully activated service 'org.freedesktop.machine1'
real 0m0.696s
user 0m0.003s
sys 0m0.046s
root@localhost:/# echo 2 > /proc/sys/vm/drop_caches
root@localhost:/# time systemctl --no-pager status dbus
● dbus.service - D-Bus System Message Bus
Loaded: loaded (/lib/systemd/system/dbus.service; static; vendor preset: enabled)
Active: active (running) since Sun 2020-10-25 17:03:05 PDT; 1 day 6h ago
Docs: man:dbus-daemon(1)
Main PID: 572 (dbus-daemon)
Memory: 2.8M
CPU: 110ms
CGroup: /system.slice/dbus.service
└─572 /usr/bin/dbus-daemon --system --address=systemd: --nofork --nopidfile --systemd-activation
Oct 25 17:03:05 localhost systemd[1]: Started D-Bus System Message Bus.
Oct 25 17:06:26 localhost dbus[572]: [system] Activating via systemd: service name='org.freedesktop.machine1' unit='dbus-org.freedesktop.machine1.service'
Oct 25 17:06:26 localhost dbus[572]: [system] Successfully activated service 'org.freedesktop.machine1'
root@localhost:/home/vc/gh/systemd/build# echo 2 > /proc/sys/vm/drop_caches
root@localhost:/home/vc/gh/systemd/build# time ./systemctl --no-pager status dbus
● dbus.service - D-Bus System Message Bus
Loaded: loaded (/lib/systemd/system/dbus.service; static)
Active: active (running) since Sun 2020-10-25 17:03:05 PDT; 1 day 6h ago
TriggeredBy: ● dbus.socket
Docs: man:dbus-daemon(1)
Main PID: 572 (dbus-daemon)
Memory: 2.8M
CPU: 110ms
CGroup: /system.slice/dbus.service
└─572 /usr/bin/dbus-daemon --system --address=systemd: --nofork --nopidfile --systemd-activation
Oct 25 17:03:05 localhost systemd[1]: Started D-Bus System Message Bus.
Oct 25 17:06:26 localhost dbus[572]: [system] Activating via systemd: service name='org.freedesktop.machine1' unit='dbus-org.freedesktop.machine1.service'
Oct 25 17:06:26 localhost dbus[572]: [system] Successfully activated service 'org.freedesktop.machine1'
real 0m0.168s
user 0m0.003s
sys 0m0.016s
root@localhost:/home/vc/gh/systemd/build# echo 2 > /proc/sys/vm/drop_caches
root@localhost:/home/vc/gh/systemd/build# time ./systemctl --no-pager status dbus
● dbus.service - D-Bus System Message Bus
Loaded: loaded (/lib/systemd/system/dbus.service; static)
Active: active (running) since Sun 2020-10-25 17:03:05 PDT; 1 day 6h ago
TriggeredBy: ● dbus.socket
Docs: man:dbus-daemon(1)
Main PID: 572 (dbus-daemon)
Memory: 2.8M
CPU: 110ms
CGroup: /system.slice/dbus.service
└─572 /usr/bin/dbus-daemon --system --address=systemd: --nofork --nopidfile --systemd-activation
Oct 25 17:03:05 localhost systemd[1]: Started D-Bus System Message Bus.
Oct 25 17:06:26 localhost dbus[572]: [system] Activating via systemd: service name='org.freedesktop.machine1' unit='dbus-org.freedesktop.machine1.service'
Oct 25 17:06:26 localhost dbus[572]: [system] Successfully activated service 'org.freedesktop.machine1'
real 0m0.167s
user 0m0.005s
sys 0m0.013s
root@localhost:/home/vc/gh/systemd/build# echo 2 > /proc/sys/vm/drop_caches
root@localhost:/home/vc/gh/systemd/build# time ./systemctl --no-pager status dbus
● dbus.service - D-Bus System Message Bus
Loaded: loaded (/lib/systemd/system/dbus.service; static)
Active: active (running) since Sun 2020-10-25 17:03:05 PDT; 1 day 6h ago
TriggeredBy: ● dbus.socket
Docs: man:dbus-daemon(1)
Main PID: 572 (dbus-daemon)
Memory: 2.8M
CPU: 110ms
CGroup: /system.slice/dbus.service
└─572 /usr/bin/dbus-daemon --system --address=systemd: --nofork --nopidfile --systemd-activation
Oct 25 17:03:05 localhost systemd[1]: Started D-Bus System Message Bus.
Oct 25 17:06:26 localhost dbus[572]: [system] Activating via systemd: service name='org.freedesktop.machine1' unit='dbus-org.freedesktop.machine1.service'
Oct 25 17:06:26 localhost dbus[572]: [system] Successfully activated service 'org.freedesktop.machine1'
This adds minimal support for RFC5001 NSID to the stub resolver. This
useful to identify systemd-resolved when talking to the stub resolver,
and distuingishing the packets resolved answers itself (where NSID is
now set) from those which it proxies 1:1 upstream (where NSID will not
be set, or set to whatever the upstream server has it set to).
The NSID chosen consist of two parts:
1. The first part is derived from /etc/machine-id and identifies the
resolved instance in a stable way.
2. The second part is the fixed string ".resolved.systemd.io".
This thus maybe used for a veriety of checks:
a. Am I talking to a resolved stub?
b. Am I talking to the same stub as last time?
c. Am I talking to the local resolved?
Given that the first part leaks the identity of the system in away two
protections are in place:
I) The NSID is only included on the main stub, not the extra stub. The
main stub has with a TTL of 1 and other protections a lot of safety
in place that the datagrams never leave the local system, thus the
identifying info is only accessible to the local system — but
/etc/machine-id is accessible to local software anyway.
II) The NSID is hashed from /etc/machine-id in a non-invertable way, so
that the machine ID itself isn't leaked, but only an identifier
derived from it.