core: when propagating restart requests due to deps, downgrade restart to try-restart
Previously, if a service A depended on a service B via Requires=, and A
was not running and B restarted this would trigger a start of A as well,
since the restart was propagated as restart independently of the state
of A.
This patch ensures that a restart of B would be propagated as a
try-restart to A, thus not changing its state if it isn't up.
After all Requisite= should be close to Requires=, without the one
exception that it doesn't pull in dependencies on start. However,
reverse deps on stop/restart should be treated the same way as for
Restart=, and this is already documented in the man page, hence stick to
it.
No distro ships that old systemd versions anyway, hence let's drop
support for live-upgrades for them. Offline updates are still supported.
And live-upgrades will only lose the job queue, hence basically still
work...
core: also enforce ratelimiter if we stop a unit due to BindsTo=
This extends on bea355dac94e82697aa98e25d80ee4248263bf92, and extends
the ratelimiter to not only be used for StopWhenUnneeded=1 units but
also for units that have BindsTo= on a unit that is dead.
mount: don't claim a device is gone from /proc/self/mountinfo before it is gone from *all* lines
Devices might be referenced by multiple mount points in
/proc/self/mountinfo, hence we should consider them unmounted only after
they disappeared from all lines, not just from one.
Martin Pitt [Sun, 17 May 2015 13:07:47 +0000 (15:07 +0200)]
device: create units with intended "found" value
Change device_found_node() to also create a .device unit if a device is not
known by udev; this is the case for "tentative" devices picked up by mountinfo
(DEVICE_FOUND_MOUNT). With that we can record the "found" attribute on the
unit.
Change device_setup_unit() to also accept a NULL udev_device, and don't
add the extra udev information in that case.
Previously device_found_node() would not create a .device unit, and
unit_add_node_link() would then create a "dead" stub one via
manager_load_unit(), so we lost the "found" attribute and unmounted everything
from that device.
Martin Pitt [Tue, 19 May 2015 05:49:56 +0000 (07:49 +0200)]
hostname: Allow comments in /etc/hostname
The hostname(1) tool allows comments in /etc/hostname. Introduce a new
read_hostname_config() in hostname-util which reads a hostname configuration
file like /etc/hostname, strips out comments, whitespace, and cleans the
hostname. Use it in hostname-setup.c and hostnamed and remove duplicated code.
Jan Janssen [Fri, 1 May 2015 13:15:16 +0000 (15:15 +0200)]
journalctl: Improve boot ID lookup
This method should greatly improve offset based lookup, by simply jumping
from one boot to the next boot. It starts at the journal head to get the
a boot ID, makes a _BOOT_ID match and then comes from the opposite
journal direction (tail) to get to the end that boot. After flushing the matches
and advancing the journal from that exact position, we arrive at the start
of next boot. Rinse and repeat.
This is faster than the old method of aggregating the full boot listing just
so we can jump to a specific boot, which can be a real pain on big journals
just for a mere "-b -1" case.
As an additional benefit --list-boots should improve slightly too, because
it does less seeking.
Note that there can be a change in boot order with this lookup method
because it will use the order of boots in the journal, not the realtime stamp
stored in them. That's arguably better, though.
Another deficiency is that it will get confused with boots interleaving in the
journal, therefore, it will refuse operation in --merge, --file and --directory mode.
Tom Gundersen [Mon, 18 May 2015 15:07:04 +0000 (17:07 +0200)]
udevd: process all SIGCHLD events every time the handler is invoked
We were returning rather than continuing in some cases. The intention
was always to fully process all pending events before returning
from the SIGCHLD handler. Restore this behaviour.
Alban Crequy [Mon, 18 May 2015 14:45:30 +0000 (16:45 +0200)]
nspawn: close extra fds before execing init
When systemd-nspawn gets exec*()ed, it inherits the followings file
descriptors:
- 0, 1, 2: stdin, stdout, stderr
- SD_LISTEN_FDS_START, ... SD_LISTEN_FDS_START+LISTEN_FDS: file
descriptors passed by the system manager (useful for socket
activation). They are passed to the child process (process leader).
- extra lock fd: rkt passes a locked directory as an extra fd, so the
directory remains locked as long as the container is alive.
systemd-nspawn used to close all open fds except 0, 1, 2 and the
SD_LISTEN_FDS_START..SD_LISTEN_FDS_START+LISTEN_FDS. This patch delays
the close just before the exec so the nspawn process (parent) keeps the
extra fds open.
This patch supersedes the previous attempt ("cloexec extraneous fds"):
http://lists.freedesktop.org/archives/systemd-devel/2015-May/031608.html
Alban Crequy [Mon, 18 May 2015 10:20:28 +0000 (12:20 +0200)]
core: Private*/Protect* options with RootDirectory
When a service is chrooted with the option RootDirectory=/opt/..., then
the options PrivateDevices, PrivateTmp, ProtectHome, ProtectSystem must
mount the directories under $RootDirectory/{dev,tmp,home,usr,boot}.
The test-ns tool can test setup_namespace() with and without chroot:
$ sudo TEST_NS_PROJECTS=/home/lennart/projects ./test-ns
$ sudo TEST_NS_CHROOT=/home/alban/debian-tree TEST_NS_PROJECTS=/home/alban/debian-tree/home/alban/Documents ./test-ns
Karel Zak [Mon, 18 May 2015 10:30:37 +0000 (12:30 +0200)]
fstab-generator: add x-systemd.requires and x-systemd.requires-mounts-for
Currently we have no way how to specify dependencies between fstab
entries (or another units) in the /etc/fstab. It means that users are
forced to bypass fstab and write .mount units manually.
The patch introduces new systemd fstab options:
x-systemd.requires=<PATH>
- to specify dependence an another mount (PATH is translated to unit name)
x-systemd.requires=<UNIT>
- to specify dependence on arbitrary UNIT
x-systemd.requires-mounts-for=<PATH ...>
- to specify dependence on another paths, implemented by
RequiresMountsFor=. The option may be specified more than once.
Eric Cook [Mon, 18 May 2015 05:02:39 +0000 (01:02 -0400)]
zsh-completion: actually complete template names for subcommands enable, reenable and disable.
compadd's -a option treats non-option arguments as arrays. So
$(_systemctl_get_template_names) expands to some words that aren't
legal array names. Even if there were, they would be empty; thus adding
nothing.
Tom Gundersen [Sat, 16 May 2015 08:14:20 +0000 (10:14 +0200)]
udevd: queue - update queue state when events are queued/freed
This way it is more obvious that the queue flag file is always
up-to-date. Moreover, we only have to touch/unlink it when the
first/last event is allocated/freed.
Only if both keep_free and max_use are actually 0 we can shortcut things
and avoid vacuuming. If either are positive or -1 we need to execute the
vacuuming.
socket-util: socket_address_parse() should not log errors on its own
Given that socket_address_parse() is mostly a "library" call it
shouldn't log on its own, but leave that to its caller.
This patch removes logging from the call in case IPv6 is not available
but and IPv6 address shall be parsed. Instead a new call
socket_address_parse_and_warn() is introduced which first invokes
socket_address_parse() and then logs if necessary.
This should fix "make check" on ipv6-less kernels:
Direct leak of 928 byte(s) in 1 object(s) allocated from:
#0 0x7f782f788db1 in __interceptor_calloc (/usr/lib64/libasan.so.2+0x96db1)
#1 0x562a83ae60cf in bus_message_from_header src/libsystemd/sd-bus/bus-message.c:480
#2 0x562a83ae6f5a in bus_message_from_malloc src/libsystemd/sd-bus/bus-message.c:576
#3 0x562a83ad3cad in bus_socket_make_message src/libsystemd/sd-bus/bus-socket.c:915
#4 0x562a83ad4cfc in bus_socket_read_message src/libsystemd/sd-bus/bus-socket.c:1051
#5 0x562a83ab733f in bus_read_message src/libsystemd/sd-bus/sd-bus.c:1647
#6 0x562a83ab98ea in sd_bus_call src/libsystemd/sd-bus/sd-bus.c:2038
#7 0x562a83b1f46d in sd_bus_call_method src/libsystemd/sd-bus/bus-convenience.c:94
#8 0x562a83aab3e1 in context_read_ntp src/timedate/timedated.c:192
#9 0x562a83aae1af in main src/timedate/timedated.c:730
#10 0x7f782eb238c4 in __libc_start_main (/lib64/libc.so.6+0x208c4)
Indirect leak of 77 byte(s) in 1 object(s) allocated from:
#0 0x7f782f788f6a in realloc (/usr/lib64/libasan.so.2+0x96f6a)
#1 0x562a83ad418a in bus_socket_read_message src/libsystemd/sd-bus/bus-socket.c:963
#2 0x562a83ab733f in bus_read_message src/libsystemd/sd-bus/sd-bus.c:1647
#3 0x562a83ab98ea in sd_bus_call src/libsystemd/sd-bus/sd-bus.c:2038
#4 0x562a83b1f46d in sd_bus_call_method src/libsystemd/sd-bus/bus-convenience.c:94
#5 0x562a83aab3e1 in context_read_ntp src/timedate/timedated.c:192
#6 0x562a83aae1af in main src/timedate/timedated.c:730
#7 0x7f782eb238c4 in __libc_start_main (/lib64/libc.so.6+0x208c4)
Indirect leak of 2 byte(s) in 1 object(s) allocated from:
#0 0x7f782f75493f in strdup (/usr/lib64/libasan.so.2+0x6293f)
#1 0x562a83b0229b in bus_message_parse_fields src/libsystemd/sd-bus/bus-message.c:5382
#2 0x562a83ae7290 in bus_message_from_malloc src/libsystemd/sd-bus/bus-message.c:601
#3 0x562a83ad3cad in bus_socket_make_message src/libsystemd/sd-bus/bus-socket.c:915
#4 0x562a83ad4cfc in bus_socket_read_message src/libsystemd/sd-bus/bus-socket.c:1051
#5 0x562a83ab733f in bus_read_message src/libsystemd/sd-bus/sd-bus.c:1647
#6 0x562a83ab98ea in sd_bus_call src/libsystemd/sd-bus/sd-bus.c:2038
#7 0x562a83b1f46d in sd_bus_call_method src/libsystemd/sd-bus/bus-convenience.c:94
#8 0x562a83aab3e1 in context_read_ntp src/timedate/timedated.c:192
#9 0x562a83aae1af in main src/timedate/timedated.c:730
#10 0x7f782eb238c4 in __libc_start_main (/lib64/libc.so.6+0x208c4)
SUMMARY: AddressSanitizer: 1007 byte(s) leaked in 3 allocation(s).
This is due to missing _cleanup_bus_message_unref_ in context_read_ntp()
core: Execute first boot presets in an enable-only preset-mode.
This means any existing enabled units well be preserved and no
pre-created symlinks will be removed. This is done on first boot, when
the assumption is that /etc is not populated at all (no machine-id
setup). For minimal containers that gives a significant first boot
speed up, approximately ~20ms / ~16% in my trials.
It turns out we don't actually need to set the global ip_forward setting.
The only relevant setting is the one on each interface.
What the global toggle actually does is switch forwarding on/off for all
currently present interfaces and change the default for new ones.
That means that by setting the global ip_forward we
- Introduce a race condition, because if the interface with IPForward=yes
is brought up after one with IPForward=no, both will have forwarding
enabled, because the global switch turns it on for all interfaces.
If the other interface comes up first networkd correctly sets forward=0
and it doesn't get overridden.
- Change the forwarding setting for interfaces that networkd is not
configured to touch, even if the user disabled forwarding via sysctl,
either globally or per-interface
As forwarding works fine without this, as long as all relevant interfacest
individually set IPForward=yes: just drop it
This means that non-networkd interfaces use the global default while
networkd interfaces default to off if IPForward isn't given.
Building with address sanitizer enabled on GCC 5.1.x a memory leak
is reported because we never close the bus, fix it by using
cleanup variable attribute.
GNU memmem() requires a nonnull first parameter. Let's introduce
memmem_safe() that removes this restriction for zero-length parameters,
and make use of it where appropriate.