Michal Schmidt [Fri, 20 Feb 2015 01:25:16 +0000 (02:25 +0100)]
shared: handle unnamed sockets in socket_address_equal()
Make sure we don't inspect sun_path of unnamed sockets.
Since we cannot know if two unnamed sockets' adresses refer to the same
socket, just return false.
Michal Schmidt [Thu, 19 Feb 2015 22:12:38 +0000 (23:12 +0100)]
core, shared: in deserializing, match same files reached via different paths
When dbus.socket is updated like this:
-ListenStream=/var/run/dbus/system_bus_socket
+ListenStream=/run/dbus/system_bus_socket
... and daemon-reload is performed, bad things happen.
During deserialization systemd does not recognize that the two paths
refer to the same named socket and replaces the socket file with a new
one. As a result, applications hang when they try talking to dbus.
Fix this by finding a match not only when the path names are equal, but
also when they point to the same inode.
In socket_address_equal() it is necessary to move the address size
comparison into the abstract sockets branch. For path name sockets the
comparison must not be done and for other families it is redundant
(their sizes are constant and checked by socket_address_verify()).
FIFOs and special files can also have multiple pathnames, so compare the
inodes for them as well. Note that previously the pathname checks used
streq_ptr(), but the paths cannot be NULL.
Fix result testing of is_symlink() to ignore negative results, which happen if
the file name does not exist at all. In this case we do not want a warning and
unlink the non-existing link.
Martin Pitt [Thu, 19 Feb 2015 08:09:57 +0000 (09:09 +0100)]
systemd-sysv-generator test: Adjust to dropped runlevelN.target mapping
Commit d5d8429a dropped the explicit runlevelN.target mapping. Adjust the tests
accordingly to explicitly state the expected targets instead of runlevels.
nspawn: when connected to pipes for stdin/stdout, pass them as-is to PID 1
Previously we always invoked the container PID 1 on /dev/console of the
container. With this change we do so only if nspawn was invoked
interactively (i.e. its stdin/stdout was connected to a TTY). In all other
cases we directly pass through the fds unmodified.
This has the benefit that nspawn can be added into shell pipelines.
everywhere: remove configurability of sysv runlevel to target mapping
With this change runlevel 2, 3, 4 are mapped to multi-user.target for
good, and 5 to graphical.target. This was already the previous mapping
but is now no longer reconfigurable, but hard-coded into the core.
This should generally simplify things, but also fix one bug: the
sysv-generator previously generated symlinks to runlevel[2-5].target
units, which possibly weren't picked up if these aliases were otherwise
only referenced by the real names "multi-user.target" and
"graphical.target".
We keep compat aliases "runlevel[2345].target" arround for cases where
this target name is explicitly requested.
The call iterates through cmsg list and closes all fds passed via
SCM_RIGHTS.
This patch also ensures the call is used wherever appropriate, where we
might get spurious fds sent and we should better close them, then leave
them lying around.
systemctl: allow interactive authorization for all bus calls
Make use of the new sd_bus_set_allow_interactive_authorization() call to
globally enable interactive authorization. Also, turn on PK agent for
more calls.
This allows us to make use of the sd_bus_call_method() convencience
helper at more places.
Didier Roche [Mon, 26 Jan 2015 16:12:54 +0000 (17:12 +0100)]
Translate fsckd messages for plymouth
For plymouth themes not supporting i18n (like .script), send translated
messages to display to user, which is equivalent to the sent machine
readable data.
Didier Roche [Thu, 5 Feb 2015 16:08:18 +0000 (17:08 +0100)]
Connect to plymouth and support cancellation of in progress fsck
Try to connect and send to plymouth (if running) some checked report progress,
using direct plymouth protocole.
Update message is the following:
fsckd:<num_devices>:<progress>:<string>
* num_devices corresponds to the current number of devices being checked (int)
* progress corresponds to the current minimum percentage of all devices being
checked (float, from 0 to 100)
* string is a translated message ready to be displayed by the plymouth theme
displaying the information above. It can be overriden by plymouth themes
supporting i18n.
Grab in fsckd plymouth watch key Control+C, and propagate this cancel request
to systemd-fsck which will terminate fsck.
Send a message to signal to user what key we are grabbing for fsck cancel.
Message is: fsckd-cancel-msg:<string>
Where string is a translated string ready to be displayed by the plymouth theme
indicating that Control+C can be used to cancel current checks. It can be
overriden (matching only fsckd-cancel-msg prefix) for themes supporting i18n.
Didier Roche [Mon, 26 Jan 2015 15:01:11 +0000 (16:01 +0100)]
systemd-fsck: always connect to systemd-fsckd
Remove the plymouth running or show-status checks from systemd-fsck. Instead,
always connect to systemd-fsckd socket, and let this one decide if we display
progress or not.
Didier Roche [Wed, 4 Feb 2015 15:42:47 +0000 (16:42 +0100)]
fsckd daemon for inter-fsckd communication
Add systemd-fsckd multiplexer which accepts multiple systemd-fsck
instances to connect to it and sends progress report. systemd-fsckd then
computes and writes to /dev/console the number of devices currently being
checked and the minimum fsck progress. This will be used for interactive
progress report and cancelling in plymouth.
systemd-fsckd stops on idle when no systemd-fsck is connected.
Make the necessary changes to systemd-fsck to connect to the systemd-fsckd
socket.
Tom Gundersen [Tue, 17 Feb 2015 19:12:50 +0000 (20:12 +0100)]
gummiboot/sd-boot/systemd-boot: rename galore
What used to be gummiboot, was renamed sd-boot when it was merged into
systemd. Let's try to be a bit more consistent with the rest of systemd
and rename it again as follows:
The EFI bootloader is now called 'systemd-bootx64.efi', and its sources are in
'src/boot/efi/'. The drop-in directory where bootctl will find EFI loaders
is now /usr/lib/systemd/boot/efi/.
logind: open up most bus calls for unpriviliged processes, using PolicyKit
Also, allow clients to alter their own objects without any further
priviliges. i.e. this allows clients to kill and lock their own sessions
without involving PK.
sd-bus: allow setting a per-connection default value for the "allow-interactive-authentication" message flag
Most of our client tools want to set this bit for all their method
calls, even though it defaults to off in sd-bus, and rightfully so.
Hence, to simplify thing, introduce a per sd_bus-object flag that sets
the default value for all messages created on the connection.
analyze: change behaviour of combined --to/from--pattern
We would require a match against all three: patterns specified
with --to, with --from, and as positional arguments to show an
edge. This does not seem useful. Let instead the positional args
behave like they were specified in both --to and --from, which is
fairly intuitive and should be more useful.
network-internal: chain matches with AND in net_match_config()
The test would treat the first non-empty set of matches in
match_paths, match_drivers, match_types, match_names as definitive
(essentially chaining them with OR). Make those tests instead match
like other tests and require all to pass if the set of patterns is
nonempty.
For two releases those exported with version 183 by mistake, and then
they were fixed to have version 215 (015419c0df libudev: fix symbol
version for udev_queue_flush() and udev_queue_get_fd()). But that
breaks ABI compatibility for binaries compiled with udev from before
that commit. There most likely very few such binaries, if any, but as
a matter of principle we should export the old symbols too, in order
to keep full compatibility.
David Herrmann [Sat, 14 Feb 2015 18:01:21 +0000 (19:01 +0100)]
bus-proxy: be more verbose if messages cannot be forwarded
If we cannot forward messages, include information on the peer and message
just like the xml-policy does. This helps debugging such situations and
figuring out what exactly is going wrong.
David Herrmann [Sat, 14 Feb 2015 14:13:38 +0000 (15:13 +0100)]
bus-proxy: don't fake data we don't have
UDS sockets transmit EUID+EGID only. Don't try to fake data we don't know!
Otherwise, this might be used to override user-limits by non-root setuid
programs (by faking UID==EUID).
Now that sd-bus is fixed to always use EUID even on UDS, we can safely set
all other UID/GID fields to INVALID.
Bruno Bottazzini [Fri, 13 Feb 2015 20:40:50 +0000 (18:40 -0200)]
util: use a shared lookup function for string tables
Macro DEFINE_STRING_TABLE_LOOKUP expands to a new function for each
of the almost 120 tables throghout the code.
Move the its implementation to a function (guaranteed to never be inlined),
and make the macro expand to an inlined function that calls this function.
This saves a few kilobytes from the systemd binary
(David: - fix coding-style
- use 'ssize_t' to fix 32bit to 64bit propagation
- use streq_ptr())
bus-proxy: also consider ENOTCONN a clean termination condition
Sometimes, when we try to reply to messages we don't check return
values. This means we might miss a ECONNRESET, and will get a ENOTCONN
on next command. Treat both the same hence.
Tom Gundersen [Fri, 13 Feb 2015 15:20:45 +0000 (16:20 +0100)]
exit-on-idle: only exit if actually idle
sd_event_wait() returning 0 usually means that it timed out, which means it must
have been idle. However, sd_event_wait() may return 0 in case an event was triggered
but it turned out there was nothing to do. Make the check for idle explicit to avoid
this edge-case.
bus-proxy: whenever we cannot forward a message, report this back to caller, but don't exit
Errors like EPERM from the kernel should certainly not be reason to
exit. Let's try to be defensive here, and try to continue on most send
errors, but possibly tell the sender about it.