Yu Watanabe [Mon, 12 Nov 2018 05:57:04 +0000 (14:57 +0900)]
network: fixes related to NetworkConfigSection
- Do not allocate NetworkConfigSection when filename == NULL
- set .network element before calling hashmap_put()
- Always free NetworkConfigSection in each object.
There's a much stronger case for doing this kind of notification in
tmpfiles.d (since it helps relating lines to each other for the purpose
of merging them). Doing this for PIDFile= is mostly about being
systematic and copying tmpfiles.d/ behaviour here.
While we are at it, let's also support relative filenames in PIDFile=
now, and prefix them with /run, to make them absolute.
David Anderson [Fri, 9 Nov 2018 04:44:16 +0000 (20:44 -0800)]
Add a networkd test for 6rd interface creation.
Unfortunately we can't check the 6rd attribute, because it's not
exposed in /sys or anywhere other than netlink... But at least we
can check that networkd brings up an interface that looks right.
core: expose bus client names currently reffing a unit as property
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.
machined: rework referencing of machine scopes from machined, too
When a machine scope is registered by machined, let's add a reference to
it, and change the GC mode so that the unit is cleaned up as soon as
machined drops the reference, regardless of the fail state.
We have the machine name anyway, let's use TerminateMachine() on
machined's Manager object directly with it. That way it's a single
method call only, instead of two, to terminate the machine.
sd-bus: make "close+flush-on-exit" optional when using sd-event with sd-bus
This adds a new pair of API calls sd_bus_set_close_on_exit() and
sd_bus_get_close_on_exit(). They control whether an sd_bus object
attached to a an sd-event loop shall automatically be flushed/closed
when the event loop goes down. Usually that's a good thing, except for
very few cases where the bus connection is longer living than the event
loop it is attached on. Specifically, this is the case for nspawn, where
we run the event loop only while the container is up, but afterwards
still want to be able to use the bus connection.
pid1: add a new AbandonScope() method call on the Manager object
This is the same as Abandon() on the Scope object, but saves clients
from first translating a unit name into a unit object path. This logic
matches how all the other unit methods have counterparts on the Manager
object too (e.g. StopUnit() on the Manager object matching Stop() on the
Unit object), this one was simply forgotten so far.
travis.yml: replace test-capability with a script exiting with 77
The test is currently failing when run under ASan in a docker container:
```
--- command ---
SYSTEMD_KBD_MODEL_MAP='/build/src/locale/kbd-model-map' PATH='/build/build:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin' SYSTEMD_LANGUAGE_FALLBACK_MAP='/build/src/locale/language-fallback-map' /build/build/test-capability
--- stderr ---
have ambient caps: yes
Capabilities:= cap_chown,cap_dac_override,cap_dac_read_search,cap_fowner,cap_fsetid,cap_kill,cap_setgid,cap_setuid,cap_setpcap,cap_linux_immutable,cap_net_bind_service,cap_net_broadcast,cap_net_admin,cap_net_raw,cap_ipc_lock,cap_ipc_owner,cap_sys_module,cap_sys_rawio,cap_sys_chroot,cap_sys_ptrace,cap_sys_pacct,cap_sys_admin,cap_sys_boot,cap_sys_nice,cap_sys_resource,cap_sys_time,cap_sys_tty_config,cap_mknod,cap_lease,cap_audit_write,cap_audit_control,cap_setfcap,cap_mac_override,cap_mac_admin,cap_syslog,cap_wake_alarm,cap_block_suspend,cap_audit_read+eip
Capabilities:= cap_dac_override,cap_net_raw+ep
==7021==LeakSanitizer has encountered a fatal error.
==7021==HINT: For debugging, try setting environment variable LSAN_OPTIONS=verbosity=1:log_threads=1
==7021==HINT: LeakSanitizer does not work under ptrace (strace, gdb, etc)
Assertion 'WIFEXITED(status) && WEXITSTATUS(status) == 0' failed at ../src/test/test-capability.c:71, function fork_test(). Aborting.
-------
```
tests: drop the precondition check for inherited flag
Docker's default capability set has the inherited flag already
set - that breaks tests which expect otherwise. Let's just
drop the check and run the test anyway.
random-util: optionally allow randomness to be generated via RDRAND
We only use this when we don't require the best randomness. The primary
usecase for this is UUID generation, as this means we don't drain
randomness from the kernel pool for them. Since UUIDs are usually not
secrets RDRAND should be goot enough for them to avoid real-life
collisions.
Originally, the high_quality_required boolean argument controlled two
things: whether to extend any random data we successfully read with
pseudo-random data, and whether to return -ENODATA if we couldn't read
any data at all.
The boolean got replaced by RANDOM_EXTEND_WITH_PSEUDO, but this name
doesn't really cover the second part nicely. Moreover hiding both
changes of behaviour under a single flag is confusing. Hence, let's
split this part off under a new flag, and use it from random_bytes().
projects / thirdparty / systemd.git / log
