Richard Maw [Mon, 17 Aug 2015 08:52:13 +0000 (08:52 +0000)]
namespace helpers: Allow entering a UID namespace
To be able to use `systemd-run` or `machinectl login` on a container
that is in a private user namespace, the sub-process must have entered
the user namespace before connecting to the container's D-Bus, otherwise
the UID and GID in the peer credentials are garbage.
So we extend namespace_open and namespace_enter to support UID namespaces,
and we enter the UID namespace in bus_container_connect_{socket,kernel}.
namespace_open will degrade to a no-op if user namespaces are not enabled
in the kernel.
Special handling is required for the setns call in namespace_enter with
a user namespace, since transitioning to your own namespace is forbidden,
as it would result in re-entering your user namespace as root.
Arguably it may be valid to check this at the call site, rather than
inside namespace_enter, but it is less code to do it inside, and if the
intention of calling namespace_enter is to *be* in the target namespace,
rather than to transition to the target namespace, it is a reasonable
approach.
The check for whether the user namespace is the same must happen before
entering namespaces, as we may not be able to access /proc during the
intermediate transition stage.
We can't instead attempt to enter the user namespace and then ignore
the failure from it being the same namespace, since the error code is
not distinct, and we can't compare namespaces while mid-transition.
https://bugzilla.redhat.com/show_bug.cgi?id=1251334
is about a unit file which has
Environment=TERM=linux PS1=system-upgrade:\w\$\x20
We used to allow that, but after recent tightening of parsing
rules, we barf. Make it clear that this is intentional.
machined: rework state tracking logic for machines
This splits up the stopping logic for machines into two steps: first on
machine_stop() we begin with the shutdown of a machine by queuing the
stop method call for it. Then, in machine_finalize() we actually remove
the rest of its runtime context. This mimics closely how sessions are
handled in logind.
This also reworks the GC logic to strictly check the current state of
the machine unit, rather than shortcutting a few cases, like for example
assuming that UnitRemoved really means a machine is gone (which it isn't
since Reloading might trigger it, see #376).
machined,logind: don't generate errors on signal match functions
If we get a weird signal, then we should log about it, but not return an
error, since sd-bus will not call us again then anymore, but for these
signals we match here we actually do want to be called on the next
invocation.
execute: don't fail if we create the runtime directory from two processes simultaneously
If a service has both ExecStart= and ExecStartPost= set with
Type=simple, then it might happen that we have two children create the
runtime directory of a service (as configured with RuntimeDirectory=) at
the same time. Previously we did this with mkdir_safe() which will
create the dir only if it is missing, but if it already exists will at
least verify the access mode and ownership to match the right values.
This is problematic in this case, since it creates and then adjusts the
settings, thus it might happen that one child creates the directory with
root owner, another one then verifies it, and only afterwards the
directory ownership is fixed by the original child, while the second
child already failed.
With this change we'll now always adjust the access mode, so that we
know that it is right. In the worst case this means we adjust the
mode/ownership even though its unnecessary, but this should have no
negative effect.
Daniel Mack [Wed, 5 Aug 2015 15:47:45 +0000 (17:47 +0200)]
core: dbus: track bus names per unit
Currently, PID1 installs an unfiltered NameOwnerChanged signal match, and
dispatches the signals itself. This does not scale, as right now, PID1
wakes up every time a bus client connects.
To fix this, install individual matches once they are requested by
unit_watch_bus_name(), and remove the watches again through their slot in
unit_unwatch_bus_name().
If the bus is not available during unit_watch_bus_name(), just store
name in the 'watch_bus' hashmap, and let bus_setup_api() do the installing
later.
When the user requests to set hostname, and we are setting both
pretty and static hostnames, and the name is a valid FQDN, we
use it as the static hostname, and unset the pretty hostname.
The change is that a FQDN with a trailing dot is accepted and ignored.
hostname-util: add relax parameter to hostname_is_valid
Tests are modified to check behaviour with relax and without relax.
New tests are added for hostname_cleanup().
Tests are moved a new file (test-hostname-util) because there's
now a bunch of them.
New parameter is not used anywhere, except in tests, so there should
be no observable change.
Alex Crawford [Sat, 1 Aug 2015 03:02:22 +0000 (20:02 -0700)]
networkd: add private options to lease struct
This stores private-zone DHCP options inside of their respective DHCP
lease. These options aren't used by networkd (what would it do with
them?), but saving them will allow other programs to query the values.
To improve performance, the options are stored in ascending order by
tag.
Daniel Mack [Wed, 5 Aug 2015 13:58:39 +0000 (15:58 +0200)]
logind: switch to sd_bus_track helper
Let logind use the sd_bus_track helper object to track the controllers of
sessions. This does not only remove quite some code but also kills the
unconditional matches for all NameOwnerChanged signals.
The latter is something we should never ever do, as it wakes up the daemon
every time a client connects, which doesn't scale.
Richard Maw [Wed, 5 Aug 2015 08:40:36 +0000 (08:40 +0000)]
man: point nspawn --machine to machinectl search-path
The --machine option used to describe searching for machines in
/var/lib/machines, which is not the whole story, so let's link to where
it's described in more detail.
WaLyong Cho [Mon, 27 Jul 2015 17:55:35 +0000 (02:55 +0900)]
core: set default process label only exec label is none
When command path has access label and no SmackProcessLabel= is not
set, default process label will be set. But if the default process
label has no rule for the access label of the command path then smack
access error will be occurred.
So, if the command path has execute label then the child have to set
its label to the same of execute label of command path instead of
default process label.
WaLyong Cho [Mon, 27 Jul 2015 17:43:55 +0000 (02:43 +0900)]
smack-util: revise smack-util apis and add read smack attr apis
- Add smack xattr lookup table
- Unify all of mac_smack_apply_xxx{_fd}() to mac_smack_apply() and
mac_smack_apply_fd().
- Add smack xattr read apis similar with apply apis as
mac_smack_read{_fd}().
David Herrmann [Tue, 4 Aug 2015 11:36:03 +0000 (13:36 +0200)]
bus-proxy: discard invalid environment updates
If the caller provides invalid assignments via
UpdateActivationEnvironment=, we now drop it and warn loudly about it. We
need this to be compatible to dbus1, but we also clearly want to tell the
caller that they better ought to follow POSIX recommendations and use
valid labels only.
Alex Crawford [Tue, 4 Aug 2015 06:11:53 +0000 (23:11 -0700)]
basic: add LIST_INSERT_BEFORE
Similar in function to LIST_INSERT_AFTER, this will insert a new element
into the list before the specified position. If the specified position
is NULL, the element is added as the tail of the list.
Richard Maw [Tue, 4 Aug 2015 08:55:04 +0000 (08:55 +0000)]
machined: Relax path checks for Copy{From,To}
Absolute paths should be sufficient to prevent funny business,
and while path_is_safe() checks this, it also checks whether the path
contains . or .. components, which while odd, aren't a security risk.
terminal-util: when resetting terminals, don't wait for carrier
In case of non-CLOCAL lines (i.e. those with carrier detect configured)
we shouldnt wait for a carrier if all we try to do is reset the TTY.
Hence, whenever we open such a TTY pass O_NONBLOCK.
Note that we continue to open ttys we intend to write to without
O_ONBLOCK, we only add it in cases we invoke ioctl()s or other terminal
operations without reading or writing to the device.
projects / thirdparty / systemd.git / log
