]> git.ipfire.org Git - thirdparty/systemd.git/log
thirdparty/systemd.git
5 years agojournal: LGTM doesn't recognize suppressions in /* */
Frantisek Sumsal [Mon, 25 Mar 2019 17:22:48 +0000 (18:22 +0100)] 
journal: LGTM doesn't recognize suppressions in /* */

5 years agojournal-remote: use source's boot-id
Chris Morin [Thu, 14 Mar 2019 18:24:52 +0000 (11:24 -0700)] 
journal-remote: use source's boot-id

systemd-journal-remote always wrote the boot-id of the device it was running on
to the header of its journal files. When the source had a different boot-id
(because it was generated on a different boot, or a different device), the
boot-ids in the file were inconsistent. The _BOOT_ID field was that of the
source, but the journal file header and each entry object header were that of
the device systemd-journal-remote ran on. This breaks journalctl --list-boots
on any of these files.

Set the boot-id in the header to be that of the source. This also fixes the
entry object headers.

5 years agoipv4ll: do not reset seed generation counter on restart
Yu Watanabe [Fri, 29 Mar 2019 22:33:03 +0000 (07:33 +0900)] 
ipv4ll: do not reset seed generation counter on restart

Fixes #12145.

5 years agoMerge pull request #12007 from poettering/clock-change
Lennart Poettering [Tue, 2 Apr 2019 08:24:35 +0000 (10:24 +0200)] 
Merge pull request #12007 from poettering/clock-change

.timer OnClockChange= and OnTimezoneChange= settings

5 years agoupdate TODO 12007/head
Lennart Poettering [Thu, 14 Mar 2019 22:30:11 +0000 (23:30 +0100)] 
update TODO

5 years agoman: document the two new .timer settings
Lennart Poettering [Thu, 14 Mar 2019 22:28:46 +0000 (23:28 +0100)] 
man: document the two new .timer settings

5 years agotest: add tests for new .timer units
Lennart Poettering [Thu, 14 Mar 2019 20:40:08 +0000 (21:40 +0100)] 
test: add tests for new .timer units

5 years agocore: optionally, trigger .timer units on timezone and clock changes
Lennart Poettering [Thu, 14 Mar 2019 20:36:47 +0000 (21:36 +0100)] 
core: optionally, trigger .timer units on timezone and clock changes

Fixes: #6228
5 years agorun: rename with_timer → arg_with_timer
Lennart Poettering [Thu, 14 Mar 2019 18:55:29 +0000 (19:55 +0100)] 
run: rename with_timer → arg_with_timer

The value is directly initialized from cmdline args, hence let's name it
so, following our usual naming style.

5 years agocore: use more structured initialization
Lennart Poettering [Thu, 14 Mar 2019 18:54:46 +0000 (19:54 +0100)] 
core: use more structured initialization

5 years agobuild: install /etc/systemd/{system,user}-generators
Topi Miettinen [Mon, 1 Apr 2019 20:31:47 +0000 (23:31 +0300)] 
build: install /etc/systemd/{system,user}-generators

Manual page systemd.generators refers to /etc/systemd/{system,user}-generators,
but the paths do not exist, so let's install them.

5 years agoMerge pull request #12030 from poettering/condition-memory
Yu Watanabe [Mon, 1 Apr 2019 23:01:42 +0000 (08:01 +0900)] 
Merge pull request #12030 from poettering/condition-memory

add ConditionCPUs= + ConditionMemory=

5 years agoMerge pull request #12168 from poettering/man-fixes
Yu Watanabe [Mon, 1 Apr 2019 22:32:40 +0000 (07:32 +0900)] 
Merge pull request #12168 from poettering/man-fixes

three minor tweaks to the man pages

5 years agocore: refactor transaction.c to use fewer gotos
Lennart Poettering [Tue, 26 Mar 2019 16:05:42 +0000 (17:05 +0100)] 
core: refactor transaction.c to use fewer gotos

In particular, let's not use gotos that jump up, i.e. are loops. gotos
that jump down for the purpose of clean-up are cool, but using them for
loops is evil.

No change in behaviour, just some refactoring.

5 years agoNEWS: document the change to installation
Zbigniew Jędrzejewski-Szmek [Mon, 1 Apr 2019 13:47:15 +0000 (15:47 +0200)] 
NEWS: document the change to installation

5 years agoMerge pull request #12160 from yuwata/wait-online-allow-configuring
Yu Watanabe [Mon, 1 Apr 2019 21:10:36 +0000 (06:10 +0900)] 
Merge pull request #12160 from yuwata/wait-online-allow-configuring

wait-online: add --any option

5 years agoMerge pull request #12155 from yuwata/network-fix-and-extend-foo-over-udp-support
Yu Watanabe [Mon, 1 Apr 2019 21:10:07 +0000 (06:10 +0900)] 
Merge pull request #12155 from yuwata/network-fix-and-extend-foo-over-udp-support

network: fix and extend Foo over UDP

5 years agosystemctl: print a more accurate error message when we can
Lennart Poettering [Fri, 29 Mar 2019 19:23:16 +0000 (20:23 +0100)] 
systemctl: print a more accurate error message when we can

5 years agorm-rf: refuse combining REMOVE_ONLY_DIRECTORIES and REMOVE_SUBVOLUME for now
Lennart Poettering [Fri, 29 Mar 2019 15:09:49 +0000 (16:09 +0100)] 
rm-rf: refuse combining REMOVE_ONLY_DIRECTORIES and REMOVE_SUBVOLUME for now

It's not easy to implement such a combined operation race-freely since
dropping a subvolume will drop all its contents, including any
non-directories.

Hence, let's just not support this combination for now. Which isn't much
of a loss, since we never combine these flags anyway.

5 years agocore: export ReloadResult value on the bus
Lennart Poettering [Mon, 1 Apr 2019 16:39:25 +0000 (18:39 +0200)] 
core: export ReloadResult value on the bus

We keep track of it, but never exposed it. Let's fix that.

5 years agoshared: add some minor comments
Lennart Poettering [Mon, 1 Apr 2019 16:41:19 +0000 (18:41 +0200)] 
shared: add some minor comments

5 years agocore: add a common function for bus calls that return unit dbus path
Lennart Poettering [Mon, 1 Apr 2019 15:49:27 +0000 (17:49 +0200)] 
core: add a common function for bus calls that return unit dbus path

Let's shorten the code a bit by using a single function for similar
cases.

No change in behaviour, just some refactoring and shortening.

5 years agoMerge pull request #12167 from poettering/timer-parse-tweak
Yu Watanabe [Mon, 1 Apr 2019 20:29:53 +0000 (05:29 +0900)] 
Merge pull request #12167 from poettering/timer-parse-tweak

two tweaks for timer expression parsing

5 years agofs-util: suppress world-writable warnings if we read /dev/null
Lennart Poettering [Mon, 1 Apr 2019 18:13:36 +0000 (20:13 +0200)] 
fs-util: suppress world-writable warnings if we read /dev/null

Fixes: #12165
5 years agosd-bus: add missing empty line
Lennart Poettering [Fri, 29 Mar 2019 11:34:22 +0000 (12:34 +0100)] 
sd-bus: add missing empty line

5 years agotest-network: add tests for --any option of wait-online 12160/head
Yu Watanabe [Mon, 1 Apr 2019 01:28:32 +0000 (10:28 +0900)] 
test-network: add tests for --any option of wait-online

5 years agotest-network: fix timeout argument for wait_online()
Yu Watanabe [Sun, 31 Mar 2019 15:08:16 +0000 (00:08 +0900)] 
test-network: fix timeout argument for wait_online()

5 years agowait-online: add --any option
Yu Watanabe [Sun, 31 Mar 2019 21:43:48 +0000 (06:43 +0900)] 
wait-online: add --any option

When this option is specified, wait-online exits with success even
when several interfaces are in configuring state.

Closes #9714.

5 years agotest-network: add tests for FooOverUDP tunnels 12155/head
Yu Watanabe [Sun, 31 Mar 2019 13:04:00 +0000 (22:04 +0900)] 
test-network: add tests for FooOverUDP tunnels

5 years agoman: update FooOverUDP=
Yu Watanabe [Sun, 31 Mar 2019 07:31:25 +0000 (16:31 +0900)] 
man: update FooOverUDP=

5 years agonetwork: make FooOverUDP.Protocol= support name of ipproto
Yu Watanabe [Sun, 31 Mar 2019 13:35:44 +0000 (22:35 +0900)] 
network: make FooOverUDP.Protocol= support name of ipproto

5 years agonetwork: use asynchronous call for creating FOU tunnels
Yu Watanabe [Sun, 31 Mar 2019 09:08:32 +0000 (18:08 +0900)] 
network: use asynchronous call for creating FOU tunnels

Otherwise, multiple FOU tunnels cannot be created correctly.

5 years agonetwork: do not ignore FooOverUDP.Encapsulation= setting
Yu Watanabe [Sun, 31 Mar 2019 08:26:16 +0000 (17:26 +0900)] 
network: do not ignore FooOverUDP.Encapsulation= setting

Previously the setting is ignored and always FOU_ENCAP_GUE is sent.

5 years agonetwork: merge ipip_init() and sit_init()
Yu Watanabe [Sun, 31 Mar 2019 13:02:37 +0000 (22:02 +0900)] 
network: merge ipip_init() and sit_init()

5 years agonetwork: add FooOverUDP support for SIT and GRE tunnels
Yu Watanabe [Sun, 31 Mar 2019 07:24:48 +0000 (16:24 +0900)] 
network: add FooOverUDP support for SIT and GRE tunnels

5 years agotest-network: add more tests for SerializeTunneledPackets=, Key=, and friends
Yu Watanabe [Sun, 31 Mar 2019 06:44:42 +0000 (15:44 +0900)] 
test-network: add more tests for SerializeTunneledPackets=, Key=, and friends

5 years agohwdb: Fix micmute on ASUS FX503VD
Benjamin Berg [Mon, 1 Apr 2019 14:54:12 +0000 (16:54 +0200)] 
hwdb: Fix micmute on ASUS FX503VD

The micmute key needs to be remapped to F20 for userspace to consume it.

See https://gitlab.gnome.org/GNOME/gnome-settings-daemon/issues/121

5 years agotest: set longer StartLimitIntervalSec= and fewer StartLimitBurst=
Yu Watanabe [Mon, 1 Apr 2019 12:26:26 +0000 (21:26 +0900)] 
test: set longer StartLimitIntervalSec= and fewer StartLimitBurst=

Some test environment may be in heavy load. In that case, rate limit
never hit, and the test fails...

5 years agoMerge pull request #12164 from keszybz/units-use-presets
Lennart Poettering [Mon, 1 Apr 2019 16:33:10 +0000 (18:33 +0200)] 
Merge pull request #12164 from keszybz/units-use-presets

Enable our units using presets in the usual fashion

5 years agoman: be clearer that .timer time expressions need to be reset to override them 12168/head
Lennart Poettering [Mon, 1 Apr 2019 15:30:45 +0000 (17:30 +0200)] 
man: be clearer that .timer time expressions need to be reset to override them

let's be clearer about the overriding concept for OnCalendar= settings.

Prompted by this thread:

https://lists.freedesktop.org/archives/systemd-devel/2019-March/042351.html

5 years agoman: refer to innermost directory as innermost, not as "lowest"
Lennart Poettering [Mon, 1 Apr 2019 15:29:12 +0000 (17:29 +0200)] 
man: refer to innermost directory as innermost, not as "lowest"

Let's avoid confusion whether the root is at the top or of the bottom of
the directory tree. Moreover we use "innermost" further down for the
same concept, so let's stick to the same terminology here.

5 years agoman: tweak XyzDirectory= table a bit
Lennart Poettering [Mon, 1 Apr 2019 15:29:02 +0000 (17:29 +0200)] 
man: tweak XyzDirectory= table a bit

5 years agocore: pass parse error to log functions when parsing timer expressions 12167/head
Lennart Poettering [Mon, 1 Apr 2019 15:43:29 +0000 (17:43 +0200)] 
core: pass parse error to log functions when parsing timer expressions

5 years agocore: simply timer expression parsing by using ".ltype" field of conf-parser logic
Lennart Poettering [Mon, 1 Apr 2019 15:39:11 +0000 (17:39 +0200)] 
core: simply timer expression parsing by using ".ltype" field of conf-parser logic

No change of behaviour. Let's just not parse the lvalue all the time
with timer_base_from_string() if we can already pass it in parsed.

5 years agoudev: move udev_ctrl_cleanup() into manager_free()
Yu Watanabe [Sun, 31 Mar 2019 19:11:02 +0000 (04:11 +0900)] 
udev: move udev_ctrl_cleanup() into manager_free()

5 years agoMerge pull request #12157 from yuwata/network-netdev-name-conflict
Lennart Poettering [Mon, 1 Apr 2019 13:17:07 +0000 (15:17 +0200)] 
Merge pull request #12157 from yuwata/network-netdev-name-conflict

network: handle NetDev.Name= conflict nicely

5 years agomeson: stop creating enablement symlinks in /etc during installation 12164/head
Zbigniew Jędrzejewski-Szmek [Mon, 1 Apr 2019 11:57:24 +0000 (13:57 +0200)] 
meson: stop creating enablement symlinks in /etc during installation

This patch was initially prompted by a report on a Fedora update [1], that the
upgrade causes systemd-resolved.service and systemd-networkd.service to be
re-enabled. We generally want to preserve the enablement of all services during
upgrades, so a reset like this is not expected.

Both services declare two symlinks in their [Install] sections, for their dbus
names and for multi-user.target.wants/.  It turns out that both services were
only partially enabled, because their dbus unit symlinks
/etc/systemd/system/dbus-org.freedesktop.{resolve1,network1}.service were
created, by the symlinks in /etc/systemd/system/multi-user.target.wants/ were
not. This means that the units could be activated by dbus, but not in usual
fashion using systemctl start. Our tools make it rather hard to figure out when
something like this happens, and it is definitely an area for improvement on its
own. The symlink in .wants/ was filtered out by during packaging, but the dbus
symlink was left in (I assume by mistake).

Let's simplify things by not creating the symlinks statically during 'ninja
install'. This means that the units shipped by systemd have to be enabled in
the usual fashion, which in turns means that [Install] section and presets
become the "single source of truth" and we don't have two sets of conflicting
configuration.

Let's consider a few cases:
- developer: a developer installs systemd from git on a running system, and they
  don't want the installation to reset enablement of anything. So this change is
  either positive for them, or has no effect (if they have everything at
  defaults).

- package creation: we want to create symlinks using 'preset-all' and 'preset'
  on upgraded packages, we don't want to have any static symlinks. This change
  will remove the need to filter out symlinks in packaging and of course fix
  the original report.

- installation of systemd from scratch: this change means that without
  'preset-all' the system will not be functional. This case could be affected
  negatively by this change, but I think it's enough of a corner case to accept
  this. In practice I expect people to build a package, not installl directly
  into the file system, so this might not even matter in practice.

Creating those symlinks was probably the right thing in the beginning, but
nowadays the preset system is very well established and people expect it to
be honoured. Ignoring the presets and doing static configuration is not welcome
anymore.

Note: during package installation, either 'preset-all' or 'preset getty@.service
machines.target remote-cryptsetup.target remote-fs.target
systemd-networkd.service systemd-resolved.service
systemd-networkd-wait-online.service systemd-timesyncd.service' should be called.

[1] https://bodhi.fedoraproject.org/updates/FEDORA-2019-616045ca76

5 years agomeson: indentation
Zbigniew Jędrzejewski-Szmek [Mon, 1 Apr 2019 11:57:07 +0000 (13:57 +0200)] 
meson: indentation

5 years agoMerge pull request #12156 from yuwata/fix-bootspec-memleaks
Lennart Poettering [Mon, 1 Apr 2019 10:46:37 +0000 (12:46 +0200)] 
Merge pull request #12156 from yuwata/fix-bootspec-memleaks

bootspec: fix memleaks

5 years agopo: update ja.po
Yu Watanabe [Mon, 1 Apr 2019 02:23:51 +0000 (11:23 +0900)] 
po: update ja.po

5 years agotest-network: add test for NetDev.Name= conflict 12157/head
Yu Watanabe [Sun, 31 Mar 2019 16:06:22 +0000 (01:06 +0900)] 
test-network: add test for NetDev.Name= conflict

5 years agoudev: shorten code a bit
Yu Watanabe [Sun, 31 Mar 2019 15:55:22 +0000 (00:55 +0900)] 
udev: shorten code a bit

5 years agonetwork: add '=' to config key names in log
Yu Watanabe [Sun, 31 Mar 2019 15:37:58 +0000 (00:37 +0900)] 
network: add '=' to config key names in log

Also, long lines are wrapped.

5 years agonetwork: do not abort execution when NetDev.Name= conflicts
Yu Watanabe [Sun, 31 Mar 2019 15:24:25 +0000 (00:24 +0900)] 
network: do not abort execution when NetDev.Name= conflicts

This also changes that .netdev files are loaded in ascending order.
Otherwise, when NetDev.ifname= setting conflicts with other .netdev file,
then .netdev file with large prefix number wins.

5 years agotest-network: add test for drop-in [WireGuardPeer] section
Yu Watanabe [Sun, 31 Mar 2019 14:56:39 +0000 (23:56 +0900)] 
test-network: add test for drop-in [WireGuardPeer] section

This also merges the two wireguard tests, and use wait_online()
to speed up the test.

5 years agobootspec: fix memleak caused by setting invalid cleanup function 12156/head
Yu Watanabe [Sun, 31 Mar 2019 14:31:29 +0000 (23:31 +0900)] 
bootspec: fix memleak caused by setting invalid cleanup function

5 years agobootspec: add missing free() in boot_config_free()
Yu Watanabe [Sun, 31 Mar 2019 14:30:30 +0000 (23:30 +0900)] 
bootspec: add missing free() in boot_config_free()

5 years agoMerge pull request #12147 from yuwata/network-gre-key-12144
Yu Watanabe [Sun, 31 Mar 2019 13:12:34 +0000 (22:12 +0900)] 
Merge pull request #12147 from yuwata/network-gre-key-12144

network: make GRE and GRETAP support Key= or friends

5 years agotest-network: test stacked erspan tunnels 12147/head
Yu Watanabe [Fri, 29 Mar 2019 18:50:11 +0000 (03:50 +0900)] 
test-network: test stacked erspan tunnels

5 years agoman: update Tunnel.Key= and friends
Yu Watanabe [Fri, 29 Mar 2019 18:49:11 +0000 (03:49 +0900)] 
man: update Tunnel.Key= and friends

5 years agonetwork: make GRE and GRETAP support Key=, InputKey=, OutputKey=, and SerializeTunnel...
Yu Watanabe [Fri, 29 Mar 2019 18:43:19 +0000 (03:43 +0900)] 
network: make GRE and GRETAP support Key=, InputKey=, OutputKey=, and SerializeTunneledPackets=

This also merge netdev_gre_fill_message_create() and netdev_erspan_fill_message_create().

5 years agoMerge pull request #12048 from jengelh/master
Zbigniew Jędrzejewski-Szmek [Fri, 29 Mar 2019 21:27:31 +0000 (22:27 +0100)] 
Merge pull request #12048 from jengelh/master

rpm: avoid hiding errors from systemd commands

5 years agoMerge pull request #12146 from yuwata/test-network-wait-online
Zbigniew Jędrzejewski-Szmek [Fri, 29 Mar 2019 21:23:38 +0000 (22:23 +0100)] 
Merge pull request #12146 from yuwata/test-network-wait-online

test-network: use wait-online to speed up tests

5 years agonetwork: make erspan netdev can be specified in Network.Tunnel=
Yu Watanabe [Fri, 29 Mar 2019 18:34:03 +0000 (03:34 +0900)] 
network: make erspan netdev can be specified in Network.Tunnel=

5 years agonetwork: do not continue when appending data to netlink message fails
Yu Watanabe [Fri, 29 Mar 2019 17:47:33 +0000 (02:47 +0900)] 
network: do not continue when appending data to netlink message fails

5 years agotest-network: merge tests for [Route] section 12146/head
Yu Watanabe [Wed, 27 Mar 2019 20:53:45 +0000 (05:53 +0900)] 
test-network: merge tests for [Route] section

5 years agotest-network: use wait_online() in test_sysctl_disable_ipv6()
Yu Watanabe [Thu, 21 Mar 2019 20:56:52 +0000 (05:56 +0900)] 
test-network: use wait_online() in test_sysctl_disable_ipv6()

5 years agotest-network: use wait_online() in test_sysctl()
Yu Watanabe [Thu, 21 Mar 2019 20:53:15 +0000 (05:53 +0900)] 
test-network: use wait_online() in test_sysctl()

This also disables IPv6AcceptRA= to speed up the test.

5 years agotest-network: use wait_online() in test_link_local_addressing()
Yu Watanabe [Thu, 21 Mar 2019 20:48:12 +0000 (05:48 +0900)] 
test-network: use wait_online() in test_link_local_addressing()

This also disables IPv6AcceptRA= to speed up the test.

5 years agotest-network: fix addr_gen_mode
Yu Watanabe [Thu, 21 Mar 2019 20:41:51 +0000 (05:41 +0900)] 
test-network: fix addr_gen_mode

If stable_secret is set, then networkd sets addr_gen_mode 2.

5 years agotest-network: move tests related to bonding
Yu Watanabe [Thu, 21 Mar 2019 20:20:04 +0000 (05:20 +0900)] 
test-network: move tests related to bonding

5 years agotest-network: merge tests about static addresses
Yu Watanabe [Thu, 21 Mar 2019 20:08:20 +0000 (05:08 +0900)] 
test-network: merge tests about static addresses

And use wait_online()

5 years agotest-network: add wait_online() helper function
Yu Watanabe [Thu, 21 Mar 2019 20:06:09 +0000 (05:06 +0900)] 
test-network: add wait_online() helper function

5 years agoMerge pull request #12138 from poettering/doc-ip-allow-src-dst 12069/head
Zbigniew Jędrzejewski-Szmek [Fri, 29 Mar 2019 15:44:48 +0000 (16:44 +0100)] 
Merge pull request #12138 from poettering/doc-ip-allow-src-dst

man: expand IPAddressAllow= docs a bit

5 years agoupdate TODO 12138/head
Lennart Poettering [Thu, 28 Mar 2019 16:15:40 +0000 (17:15 +0100)] 
update TODO

5 years agoman: clarify which addresses are affected by IPAddressAllow=/IPAddressDeny=
Lennart Poettering [Thu, 28 Mar 2019 16:14:39 +0000 (17:14 +0100)] 
man: clarify which addresses are affected by IPAddressAllow=/IPAddressDeny=

For ingress traffic it's the source address of IP packets we check, for
egress traffic it's the destination address. Mention that.

5 years agopo: update Polish translation
Piotr Drąg [Mon, 25 Mar 2019 09:40:57 +0000 (10:40 +0100)] 
po: update Polish translation

5 years agoMerge pull request #12140 from poettering/copy-early
Zbigniew Jędrzejewski-Szmek [Fri, 29 Mar 2019 14:02:50 +0000 (15:02 +0100)] 
Merge pull request #12140 from poettering/copy-early

chattr/copy.c fixes

5 years agoMerge pull request #12137 from poettering/socket-var-run
Zbigniew Jędrzejewski-Szmek [Fri, 29 Mar 2019 14:00:25 +0000 (15:00 +0100)] 
Merge pull request #12137 from poettering/socket-var-run

warn about sockets in /var/run/ too

5 years agoMerge pull request #12133 from poettering/rseq-whitelist
Zbigniew Jędrzejewski-Szmek [Fri, 29 Mar 2019 13:59:00 +0000 (14:59 +0100)] 
Merge pull request #12133 from poettering/rseq-whitelist

whitelist rseq() system call in `@default` syscall group

5 years agoanalyze: check both possible mount points of tracefs
Lennart Poettering [Thu, 28 Mar 2019 08:59:24 +0000 (09:59 +0100)] 
analyze: check both possible mount points of tracefs

Let's try the new one first, the old one second.

5 years agofsck: copy out device argument from argv[] before forking
Lennart Poettering [Thu, 28 Mar 2019 18:26:21 +0000 (19:26 +0100)] 
fsck: copy out device argument from argv[] before forking

We nowadays rename our child processes, hence argv[] will be clobbered,
let's hence copy the device path to dynamic memory before forking.

This is fall-out from 60ffa37a65a96c3af857a3dfc4a6fd47b20cc90e since we
now a lot more often end up overriding the argv[] buffer than before,
simple because we know what to override.

These kind of bugs kinda suck. THere are only two options here: stop
overriding argv[] for all cases (or just these cases) or explicitly
copying out everything we need in child processes before forking. With
this patch I opt for the latter, though I am not 100% convinced this is
a great solution. Just a better solution than everything else, i.e.
allowing argv[] to remain out of sync with what others see.

Fixes: #12135
5 years agowireguard: fix exponential backoff when resolving hosts
Jörg Thalheim [Thu, 28 Mar 2019 16:20:52 +0000 (16:20 +0000)] 
wireguard: fix exponential backoff when resolving hosts

It should stop at 25s, not start.
Fixes #12134

5 years agoheaders: add missing includes
Zbigniew Jędrzejewski-Szmek [Thu, 28 Mar 2019 14:55:09 +0000 (15:55 +0100)] 
headers: add missing includes

Fixes #12125.

5 years agosd-bus: change "int" → "signed int" on bitfield
Lennart Poettering [Thu, 28 Mar 2019 08:26:50 +0000 (09:26 +0100)] 
sd-bus: change "int" → "signed int" on bitfield

Apparently by the C standard "int" bitfields can have any signedness
(unlike non-bitfield declarations which are "signed" if the signedness
is not specified).

Let's fix the LGTM warning about this hence and be explicit that we mean
"signed" here.

5 years agotmpfiles: move full chattr flag set to chattr-util.h 12140/head
Lennart Poettering [Thu, 28 Mar 2019 17:41:39 +0000 (18:41 +0100)] 
tmpfiles: move full chattr flag set to chattr-util.h

It's a pretty generic concept and fits will there, hence let's move it.

5 years agoupdate NEWS
Lennart Poettering [Thu, 28 Mar 2019 17:38:59 +0000 (18:38 +0100)] 
update NEWS

5 years agoupdate TODO
Lennart Poettering [Thu, 28 Mar 2019 17:10:17 +0000 (18:10 +0100)] 
update TODO

5 years agotmpfiles: support the FS_PROJINHERIT_FL chattr flag
Lennart Poettering [Thu, 28 Mar 2019 17:35:03 +0000 (18:35 +0100)] 
tmpfiles: support the FS_PROJINHERIT_FL chattr flag

5 years agoutil-lib: when copying files make sure to apply some chattrs early, some late
Lennart Poettering [Thu, 28 Mar 2019 16:54:04 +0000 (17:54 +0100)] 
util-lib: when copying files make sure to apply some chattrs early, some late

Some chattrs only work sensible if you set them right after opening a
file for create (think: FS_NOCOW_FL). Others only work when they are
applied when the file is fully written (think: FS_IMMUTABLE_FL). Let's
take that into account when copying files and applying a chattr to them.

5 years agomissing: add FS_PROJINHERIT_FL
Lennart Poettering [Thu, 28 Mar 2019 17:34:33 +0000 (18:34 +0100)] 
missing: add FS_PROJINHERIT_FL

It's available since kernel 4.5, but not in older kernels.

5 years agoupdate TODO 12137/head
Lennart Poettering [Thu, 28 Mar 2019 16:00:46 +0000 (17:00 +0100)] 
update TODO

5 years agocore: complain and correct /var/run/ → /run/ for listening sockets
Lennart Poettering [Thu, 28 Mar 2019 15:59:57 +0000 (16:59 +0100)] 
core: complain and correct /var/run/ → /run/ for listening sockets

We already do that for PIDFile= paths, and for tmpfiles.d/ snippets,
let's also do this for .socket paths.

5 years agoload-fragment: use TAKE_PTR() where we can
Lennart Poettering [Thu, 28 Mar 2019 15:46:27 +0000 (16:46 +0100)] 
load-fragment: use TAKE_PTR() where we can

5 years agocryptsetup-generator: set high OOM score for systemd-cryptsetup instances
Michal Sekletar [Wed, 27 Mar 2019 08:27:21 +0000 (09:27 +0100)] 
cryptsetup-generator: set high OOM score for systemd-cryptsetup instances

With new LUKS2 header format it is possible to use Argon2 key derivation
function. This function is "memory-hard" hence keyslot unlocking can
potentially use a lot of RAM as this increases resistance to massively
parallel GPU based password cracking.

However, when multiple systemd-cryptsetup binaries run at the same
time it is very likely that system using Argon2 (e.g. Fedora 30)
will encounter memory-pressure during early boot, following OOM killing
spree.

This patch aims to lower the damage done by OOM killer and sets OOMScore
for systemd-cryptsetup units to 500. Hopefully OOM killer will then
shoot us down and leave rest of the system services alone.

5 years agoMerge pull request #12130 from keszybz/fix-ndebug-builds
Lennart Poettering [Thu, 28 Mar 2019 14:52:27 +0000 (15:52 +0100)] 
Merge pull request #12130 from keszybz/fix-ndebug-builds

Fix ndebug builds

5 years agoupdate TODO 12133/head
Lennart Poettering [Thu, 28 Mar 2019 11:00:28 +0000 (12:00 +0100)] 
update TODO

5 years agoseccomp: add rseq() to default list of syscalls to whitelist
Lennart Poettering [Thu, 28 Mar 2019 09:01:09 +0000 (10:01 +0100)] 
seccomp: add rseq() to default list of syscalls to whitelist

Apparently glibc is going to call this implicitly soon, hence let's
whitelist this by default.

Fixes: #12127
5 years agocore: break overly long line
Lennart Poettering [Thu, 28 Mar 2019 11:09:23 +0000 (12:09 +0100)] 
core: break overly long line

5 years agocore: parse '@default' seccomp group permissively
Lennart Poettering [Thu, 28 Mar 2019 11:00:56 +0000 (12:00 +0100)] 
core: parse '@default' seccomp group permissively

We are about to add system calls (rseq()) not available on old
libseccomp/old kernels, and hence we need to be permissive when parsing
our definitions.