chiark / gitweb /
Lennart Poettering [Mon, 26 Jun 2017 16:52:47 +0000 (18:52 +0200)]
sd-bus: use GetConnectionCredentials() when querying credentials, if available
Newer D-Bus versions implement the GetConnectionCredentials() driver
call to get all connection creds in one go. Make use of that to reduce
the number of bus calls we do.
When only a single credential field is queried we will still use the old
calls, which we'll also use if the new call isn't implemented.
Lennart Poettering [Mon, 26 Jun 2017 16:24:58 +0000 (18:24 +0200)]
sd-bus: when credentials of the "org.freedesktop.DBus" service are queried return the bus owner's credentials
The bus driver service is always implemented by the owner of the bus,
hence let's shortcut the credential operation and use our cached data.
This makes sure things simply work, given that dbus itself doesn't
support GetConnectionSELinuxSecurityContext() on the bus driver name
itself.
Fixes: #6120
Lennart Poettering [Mon, 26 Jun 2017 15:40:08 +0000 (17:40 +0200)]
core: make IOSchedulingClass= and IOSchedulingPriority= settable for transient units
This patch is a bit more complex thant I hoped. In particular the single
IOScheduling= property exposed on the bus is split up into
IOSchedulingClass= and IOSchedulingPriority= (though compat is
retained). Otherwise the asymmetry between setting props and getting
them is a bit too nasty.
Fixes #5613
Zbigniew Jędrzejewski-Szmek [Mon, 26 Jun 2017 07:34:03 +0000 (03:34 -0400)]
Copy defines for renameat2 from casync (#6181)
Zbigniew Jędrzejewski-Szmek [Sun, 25 Jun 2017 21:09:05 +0000 (17:09 -0400)]
basic/random-util: do not fall back to /dev/urandom if getrandom() returns short
During early boot, we'd call getrandom(), and immediately fall back to
reading from /dev/urandom unless we got the full requested number of bytes.
Those two sources are the same, so the most likely result is /dev/urandom
producing some pseudorandom numbers for us, complaining widely on the way.
Let's change our behaviour to be more conservative:
- if the numbers are only used to initialize a hash table, a short read is OK,
we don't really care if we get the first part of the seed truly random and
then some pseudorandom bytes. So just do that and return "success".
- if getrandom() returns -EAGAIN, fall back to rand() instead of querying
/dev/urandom again.
The idea with those two changes is to avoid generating a warning about
reading from an /dev/urandom when the kernel doesn't have enough entropy.
- only in the cases where we really need to make the best effort possible
(sd_id128_randomize and firstboot password hashing), fall back to
/dev/urandom.
When calling getrandom(), drop the checks whether the argument fits in an int —
getrandom() should do that for us already, and we call it with small arguments
only anyway.
Note that this does not really change the (relatively high) number of random
bytes we request from the kernel. On my laptop, during boot, PID 1 and all
other processes using this code through libelogind request:
74780 bytes with high_quality_required == false
464 bytes with high_quality_required == true
and it does not eliminate reads from /dev/urandom completely. If the kernel was
short on entropy and getrandom() would fail, we would fall back to /dev/urandom
for those 464 bytes.
When falling back to /dev/urandom, don't lose the short read we already got,
and just read the remaining bytes.
If getrandom() syscall is not available, we fall back to /dev/urandom same
as before.
Fixes #4167 (possibly partially, let's see).
Zbigniew Jędrzejewski-Szmek [Sat, 24 Jun 2017 00:43:48 +0000 (20:43 -0400)]
tests: add tests for environment serialization
Zbigniew Jędrzejewski-Szmek [Sat, 24 Jun 2017 00:27:12 +0000 (20:27 -0400)]
basic/env-util: when serializing, actually use escaping
Fixes #6152.
Zbigniew Jędrzejewski-Szmek [Fri, 23 Jun 2017 23:20:54 +0000 (19:20 -0400)]
manager: just warn about an invalid environment entry
Apart from bugs (as in #6152), this can happen if we ever make
our requirements for environment entries more stringent. As with
the rest of deserialization, we should just warn and continue.
Lennart Poettering [Thu, 22 Jun 2017 18:52:23 +0000 (20:52 +0200)]
time-util: add new call usec_shift_clock() for converting times between clocks
We use that quite often, let's implement one clean version of it.
Lennart Poettering [Thu, 22 Jun 2017 18:49:12 +0000 (20:49 +0200)]
time-util: rename usec_sub() to usec_sub_signed() and add usec_sub_unsigned()
Quite often we just want to subtract two normal usec_t values, hence
provide an implementation for that.
Evgeny Vereshchagin [Thu, 22 Jun 2017 01:10:29 +0000 (04:10 +0300)]
log: pass the correct length to vsnprintf (#6168)
This prevents log_object_internalv from corrupting the stack.
Closes #6147.
Many thanks to Walter Doekes for the code review.
Lennart Poettering [Wed, 21 Jun 2017 18:42:28 +0000 (20:42 +0200)]
sd-bus: make sure propagate all errors with vtable callbacks back to clients
Previously we'd propagate errors returned by user callbacks configured
in vtables back to the users only for method handlers and property
get/set handlers. This does the same for child enumeration and when we
check whether a fallback unit exists.
Without this the failure will be treated as a non-recoverable connection
error and result in connection termination.
Fixes: #6059
Lennart Poettering [Wed, 21 Jun 2017 17:17:38 +0000 (19:17 +0200)]
core: make NotifyAccess= and FileDescriptorStoreMax= available to transient services
This is helpful for debugging/testing #5606.
Lennart Poettering [Tue, 20 Jun 2017 21:30:30 +0000 (23:30 +0200)]
core: permit FDSTORE=1 messages with non-pollable fds
This also alters the documentation to recommend memfds rather than /run
for serializing state across reboots. That's because /run doesn't
actually have the same lifecycle as the fd store, as it is cleared out
on restarts.
Fixes: #5606
Zbigniew Jędrzejewski-Szmek [Sat, 17 Jun 2017 16:41:08 +0000 (12:41 -0400)]
basic/rm-rf: allow a symlink to / to be removed
We open the target path with O_DIRECTORY|O_NOFOLLOW, and if that doesn't work,
we call unlink() on the path. In neither case we will follow the symlink, so
we can relax our check to also not follow symlinks.
Fixes #5864.
Zbigniew Jędrzejewski-Szmek [Sat, 17 Jun 2017 16:37:16 +0000 (12:37 -0400)]
basic/path-util: allow flags for path_equal_or_files_same
No functional change, just a new parameters and the tests that
AT_SYMLINK_NOFOLLOW works as expected.
Waldemar Brodkorb [Thu, 15 Jun 2017 15:44:59 +0000 (17:44 +0200)]
Make IDN support conditional
[zj: rename HAVE_IDN to ENABLE_IDN]
Zbigniew Jędrzejewski-Szmek [Sun, 11 Jun 2017 19:24:07 +0000 (15:24 -0400)]
Use "dollar-single-quotes" to escape shell-sensitive strings
Also called "ANSI-C Quoting" in info:(bash) ANSI-C Quoting.
The escaping rules are a POSIX proposal, and are described in
http://austingroupbugs.net/view.php?id=249. There's a lot of back-and-forth on
the details of escaping of control characters, but we'll be only using a small
subset of the syntax that is common to all proposals and is widely supported.
Unfortunately dash and fish and maybe some other shells do not support it (see
the man page patch for a list).
This allows environment variables to be safely exported using show-environment
and imported into the shell. Shells which do not support this syntax will have
to do something like
export $(systemctl show-environment|grep -v '=\$')
or whatever is appropriate in their case. I think csh and fish do not support
the A=B syntax anyway, so the change is moot for them.
Fixes #5536.
v2:
- also escape newlines (which currently disallowed in shell values, so this
doesn't really matter), and tabs (as $'\t'), and ! (as $'!'). This way quoted
output can be included directly in both interactive and noninteractive bash.
Zbigniew Jędrzejewski-Szmek [Sat, 3 Jun 2017 09:41:17 +0000 (05:41 -0400)]
sd-bus: silence format warnings in kdbus code (#6072)
The code is mostly correct, but gcc is trying to outsmart us, and emits a
warning for a "llu vs lu" mismatch, even though they are the same size (on alpha):
src/libelogind/sd-bus/bus-control.c: In function ‘kernel_get_list’:
src/libelogind/sd-bus/bus-control.c:267:42: error: format ‘%llu’ expects argument of type ‘long long unsigned int’, but argument 3 has type ‘__u64 {aka long unsigned int}’ [-Werror=format=]
if (asprintf(&n, ":1.%llu", name->id) < 0) {
^
src/libelogind/sd-bus/bus-control.c: In function ‘bus_get_name_creds_kdbus’:
src/libelogind/sd-bus/bus-control.c:714:47: error: format ‘%llu’ expects argument of type ‘long long unsigned int’, but argument 3 has type ‘__u64 {aka long unsigned int}’ [-Werror=format=]
if (asprintf(&c->unique_name, ":1.%llu", conn_info->id) < 0) {
^
This is hard to work around properly, because kdbus.h uses __u64 which is
defined-differently-despite-being-the-same-size then uint64_t. Thus the simple
solution of using %PRIu64 fails on amd64:
src/libelogind/sd-bus/bus-control.c:714:47: error: format ‘%lu’ expects argument of type ‘long unsigned int’, but argument 3 has type ‘__u64 {aka long long unsigned int}’ [-Werror=format=]
if (asprintf(&c->unique_name, ":1.%"PRIu64, conn_info->id) < 0) {
^~~~~~
Let's just avoid the whole issue for now by silencing the warning.
After the next release, we should just get rid of the kdbus code.
Fixes #5561.
Zbigniew Jędrzejewski-Szmek [Thu, 1 Jun 2017 02:42:14 +0000 (22:42 -0400)]
loginctl: also use $XDG_SESSION_ID for session-status
Zbigniew Jędrzejewski-Szmek [Thu, 1 Jun 2017 02:23:30 +0000 (22:23 -0400)]
loginctl: use $XDG_SESSION_ID for "our" session
Instead of always letting logind guess what the caller's session is, let's
give it the value from $XDG_SESSION_ID when it is present in the caller's
environment.
Nowadays terminal emulators are often running as services under elogind --user,
and not as part of an actual session, so all loginctl calls which depend on
logind guessing the session will fail. I don't see a reason not to honour
$XDG_SESSION_ID.
This applies to LockSession, UnlockSession, TerminateSession, ActivateSession,
SetUserLinger.
Fixes #6032.
Zbigniew Jędrzejewski-Szmek [Thu, 1 Jun 2017 02:10:15 +0000 (22:10 -0400)]
logind: nicer error message when we cannot guess the caller's session
Partial fix for #6032.
Yu Watanabe [Wed, 31 May 2017 15:27:47 +0000 (00:27 +0900)]
man: update sd_get_seats(3)
Yu Watanabe [Wed, 31 May 2017 15:12:32 +0000 (00:12 +0900)]
sd-login: sd_get_machine_names(): do not return -EINVAL when output parameter is NULL
Other functions in sd-login generally allow the output parameter to be NULL, in
which case only the number of items that would be stored in the array is returned.
Be nice and do the same here.
Yu Watanabe [Wed, 31 May 2017 14:48:47 +0000 (23:48 +0900)]
sd-login: treat missing /run/elogind/{seats,sessions,users} the same as empty
C.f.
0543105b0fb13e4243b71a78f62f81fb9dde5d51.
This makes if /run/elogind/{seats,sessions,users} are missing, then
sd_get_seats(), sd_get_sessions() and sd_get_uids() return 0, that is,
an empty list, instead of -ENOENT.
Zbigniew Jędrzejewski-Szmek [Wed, 31 May 2017 00:21:45 +0000 (20:21 -0400)]
sd-login: translate -ENOMEDIUM to -ENODATA
The -ENOMEDIUM return value was introduced in
v232-1001-g2977724b09,
('core: make hybrid cgroup unified mode keep compat /sys/fs/cgroup/elogind hierarchy'),
and would be returned by cg_pid_get_path_shifted(), but the documented and
expected return value is -ENODATA. Let's just catch ENXIO/ENOMEDIUM and translate
it to ENODATA in all cases.
Complements
171f8f591ff27ebb5ff475b7a9d1f13a846c9331, fixes #6012.
Tom Gundersen [Thu, 25 May 2017 14:11:04 +0000 (16:11 +0200)]
logn: tests - don't compare signed with unsigned
Zbigniew Jędrzejewski-Szmek [Fri, 19 May 2017 18:49:06 +0000 (14:49 -0400)]
basic/time-util: make parsing of dual_timestamp more strict
*scanf functions set errno on i/o error. For sscanf, this doesn't really apply,
so (based on the man page), it seems that errno is unlikely to be ever set to a
useful value. So just ignore errno. The error message includes the string that
was parsed, so it should be always pretty clear why parsing failed.
On the other hand, detect trailing characters and minus prefix that weren't
converted properly. This matches what our safe_ato* functions do. Add tests to
elucidate various edge cases.
Zbigniew Jędrzejewski-Szmek [Fri, 19 May 2017 18:37:48 +0000 (14:37 -0400)]
test-parse-util: verify that ato[ui] actually rejects trailing garbage
Alexey Brodkin [Fri, 19 May 2017 14:55:24 +0000 (16:55 +0200)]
architecture: add Synopsys DesignWare ARC cores support (#5992)
DesignWare ARC Processors are a family of 32-bit CPUs from Synopsys
used extensively in SoCs of different vendors.
Matija Skala [Fri, 19 May 2017 14:01:35 +0000 (16:01 +0200)]
Fix includes (#5980)
Needed on musl.
Timothée Ravier [Fri, 19 May 2017 12:38:40 +0000 (14:38 +0200)]
core: open /proc/self/mountinfo early to allow mounts over /proc (#5985)
Enable masking the /proc folder using the 'InaccessiblePaths' unit
option.
This also slightly simplify mounts setup as the bind_remount_recursive
function will only open /proc/self/mountinfo once.
This is based on the suggestion at:
https://lists.freedesktop.org/archives/elogind-devel/2017-April/038634.html
Matija Skala [Fri, 19 May 2017 12:36:12 +0000 (14:36 +0200)]
timesync/timesyncd-manager: fix format-specifier issues
timex::time::tv_usec and timex::freq can have different sizes
depending on the host architecture. On x32 in particular,
it is 8 bytes long while the long int type is only 4 bytes
long. Hence, using li as a format specifier will trigger
a format error. Thus, introduce a new format specifier
PRI_TIMEX which is defined as PRIi64 on x32 and li
everywhere else.
Zbigniew Jędrzejewski-Szmek [Fri, 19 May 2017 11:52:00 +0000 (07:52 -0400)]
sd-login: fix querying machines when machined is not running
We should not leak the internal error from missing directory and treat
that case the same as no machines.
Zbigniew Jędrzejewski-Szmek [Fri, 19 May 2017 02:32:19 +0000 (22:32 -0400)]
sd-login,test-login: return -ENODATA from sd_pid_get_unit too
After all, we might be running on a non-elogind system.
Zbigniew Jędrzejewski-Szmek [Fri, 19 May 2017 02:11:08 +0000 (22:11 -0400)]
sd-login: fix return value of sd_pid_get_user_unit
E.g. "/user.slice/user-1000.slice/session-15.scope" would cause -ENXIO to be
returned.
Zbigniew Jędrzejewski-Szmek [Fri, 19 May 2017 01:55:44 +0000 (21:55 -0400)]
sd-login: fix return value of sd_pid_get_session
We'd return -ENXIO, even thoug -ENODATA is documented.
Zbigniew Jędrzejewski-Szmek [Fri, 19 May 2017 00:58:23 +0000 (20:58 -0400)]
sd-login: read list of uids of sessions from UIDS not ACTIVE_SESSIONS
As described by Luke Shumaker:
sd_seat_get_sessions looks at /run/elogind/seats/${seat_name}:SESSIONS to get
the list of sessions (which I believe is correct), and at
/run/elogind/seats/${seat_name}:ACTIVE_SESSIONS for the list of users (which
I believe is incorrect); I believe that it should look at the UIDS field for
the list of users. As far as I can tell, the ACTIVE_SESSIONS field is never
even present in the seats file. I also believe that this has been broken
since the function was first committed almost 6 years ago.
Fixes #5743.
Zbigniew Jędrzejewski-Szmek [Fri, 19 May 2017 00:48:09 +0000 (20:48 -0400)]
sd-login: always return two arrays of same length from sd_seat_get_sessions
sd_seat_get_sessions returns two arrays, that in principle should always match:
the session names and corresponding uids. The second array could be shorter only
if parsing or uid conversion fails. But in that case there is no way to tell
*which* uid is wrong, so they are *all* useless. It's better to simplify things and
just return an error if parsing fails.
Lennart Poettering [Thu, 18 May 2017 16:24:17 +0000 (18:24 +0200)]
man: extend documentation on sd_bus_add_match a bit()
Explain briefly how the concept of "sd_bus_slot" works.
This recently came up on the mailing list, hence let's document this for
the next time.
Ronny Chevalier [Sun, 14 May 2017 14:30:40 +0000 (16:30 +0200)]
env-util: fix memory leak (#5962)
If cunescape succeeds, but the assignment is not valid, uce is not freed.
Ronny Chevalier [Sun, 14 May 2017 11:19:11 +0000 (13:19 +0200)]
test-conf-parser: add valid and invalid utf8 test for config_parse_path
Ronny Chevalier [Sun, 14 May 2017 11:19:11 +0000 (13:19 +0200)]
conf-parser: fix wrong argument given to log_syntax_invalid_utf8
The condition is on "word", hence we give word instead of rvalue.
An assert would be triggered if !utf8_is_valid(word) is true and
rvalue == NULL, since log_syntax_invalid_utf8 calls utf8_escape_invalid
which calls assert(str).
A test case has been added to test with valid and invalid utf8.
Zbigniew Jędrzejewski-Szmek [Sat, 13 May 2017 15:44:51 +0000 (11:44 -0400)]
tree-wide: drop assert.h includes
We provide an independent reimplementation in macro.h, and that's the one
we want to use. Including the system header is unnecessary and confusing.
Zbigniew Jędrzejewski-Szmek [Sat, 13 May 2017 15:26:55 +0000 (11:26 -0400)]
tree-wide: when %m is used in log_*, always specify errno explicitly
All those uses were correct, but I think it's better to be explicit.
Using implicit errno is too error prone, and with this change we can require
(in the sense of a style guideline) that the code is always specified.
Helpful query: git grep -n -P 'log_[^s][a-z]+\(.*%m'
Zbigniew Jędrzejewski-Szmek [Fri, 12 May 2017 18:25:17 +0000 (14:25 -0400)]
test-login: make the test non-manual
test-login.c is largely rewritten to use _cleanup_ and give more meaningful
messages (function names are used instead of creative terms like "active
session", so that when something unexpected is returned, it's much easier to
see what function is responsible).
The monitoring part is only activated if '-m' is passed on the command line.
It runs against the information from /run/elogind/ in the live system, but that
should be OK: logind/sd-login interface is supposed to be stable and both
backwards and forwards compatible.
If not running in a login session, some tests are skipped.
Those two changes together mean that it's possible to run test-login in the
test suite.
Tests for sd_pid_get_{unit,user_unit,slice} are added.
Matthijs van Duin [Thu, 11 May 2017 22:55:26 +0000 (00:55 +0200)]
sd-bus: fix c++ compatibility (#5941)
g++ annoyingly requires a non-empty struct-initializer to initialize all
struct members, in order of declaration.
Signed-off-by: Matthijs van Duin <matthijsvanduin@gmail.com>
Zbigniew Jędrzejewski-Szmek [Thu, 11 May 2017 14:23:36 +0000 (10:23 -0400)]
basic/fileio: extend atomic file writing with timestamp setting
There should be no functional change.
Ian Wienand [Mon, 8 May 2017 00:23:49 +0000 (10:23 +1000)]
Add short-iso-precise for journalctl output (#5884)
This adds a short-iso-precise option for journalctl output. It is similar to
short-iso, but includes microseconds.
Zbigniew Jędrzejewski-Szmek [Sun, 7 May 2017 15:35:32 +0000 (11:35 -0400)]
Mark python scripts executable
Since all our python scripts have a proper python3 shebang, there is no benefit
to letting meson autodetect them. On linux, meson will just uses exec(), so the
shebang is used anyway. The only difference should be in how meson reports the
script and that the detection won't fail for (most likely misconfigured)
non-UTF8 locales.
Closes #5855.
Zbigniew Jędrzejewski-Szmek [Sun, 7 May 2017 15:29:40 +0000 (11:29 -0400)]
man: fix links to external man pages
linkchecker ftw!
Gary Tierney [Tue, 2 May 2017 16:42:19 +0000 (17:42 +0100)]
Revert "selinux: split up mac_selinux_have() from mac_selinux_use()"
This reverts commit
6355e75610a8d47fc3ba5ab8bd442172a2cfe574.
The previously mentioned commit inadvertently broke a lot of SELinux related
functionality for both unprivileged users and elogind instances running as
MANAGER_USER. In particular, setting the correct SELinux context after a User=
directive is used would fail to work since we attempt to set the security
context after changing UID. Additionally, it causes activated socket units to
be mislabeled for elogind --user processes since setsockcreatecon() would never
be called.
Reverting this fixes the issues with labeling outlined above, and reinstates
SELinux access checks on unprivileged user services.
Franck Bui [Tue, 2 May 2017 07:59:17 +0000 (09:59 +0200)]
core: introduce cg_mask_from_string()/cg_mask_to_string()
Susant Sahani [Thu, 27 Apr 2017 05:11:46 +0000 (10:41 +0530)]
config parser: Introduce config_parse_ip_port
Matija Skala [Thu, 27 Apr 2017 00:08:52 +0000 (02:08 +0200)]
improve readability (#5814)
Franck Bui [Wed, 26 Apr 2017 16:18:03 +0000 (18:18 +0200)]
parse-util: introduce parse_dev() helper
Franck Bui [Wed, 26 Apr 2017 12:20:41 +0000 (14:20 +0200)]
logind: when setting a new controller, don't prepare the VT if logind is restarted
When assigning a new session controller to a session, the VT is prepared so the
controller can expect the VT to be in a good default state.
However when logind is restarted and a session controller already took control
of a session, there's no need to prepare th VT otherwise logind may screw up
the VT state set by the controller.
This patch prevents the preparation of the VT in this case.
Zbigniew Jędrzejewski-Szmek [Wed, 26 Apr 2017 03:50:35 +0000 (23:50 -0400)]
tmpfiles: use safe_glob()
This filters out "." and ".." from glob results. Fixes #5655 and #5644.
Any judgements on whether the path is "safe" are removed. We will not remove
"/" under any name (including "/../" and such), but we will remove stuff that
is specified using paths that include "//", "/./" and "/../". Such paths can be
created when joining strings automatically, or for other reasons, and people
generally know what ".." and "." is.
Tests are added to make sure that the helper functions behave as expected.
Zbigniew Jędrzejewski-Szmek [Wed, 26 Apr 2017 03:44:34 +0000 (23:44 -0400)]
basic: add readdir_no_dot and safe_glob functions
safe_glob filters out "." and "..".
This converts all users of glob_extend() and glob_exists() to safe_glob.
Zbigniew Jędrzejewski-Szmek [Wed, 26 Apr 2017 01:14:59 +0000 (21:14 -0400)]
test-sizeof: print the size of an enum
David Herrmann [Wed, 26 Apr 2017 00:36:30 +0000 (02:36 +0200)]
bus: report builtins in GetManagedObjects() (#5799)
We already report builtin interfaces with InterfacesAdded and InterfacesRemoved. However,
we never reported them in GetManagedObjects(). This might end up confusing callers that
want to use those interfaces (or simply rely on the interface count to be coherent).
Report the builtins for all objects that are queried.
Susant Sahani [Tue, 25 Apr 2017 13:11:50 +0000 (18:41 +0530)]
config parser: Introduce config_parse_uint8
Matija Skala [Tue, 25 Apr 2017 12:44:29 +0000 (14:44 +0200)]
fix parenthesis in pager.c (#5798)
- result of dup2 should be compared against 0
- making parenthesis around comparisons just to be sure
Federico [Tue, 25 Apr 2017 07:32:24 +0000 (09:32 +0200)]
Export sd_bus_message_append_ap. It is renamed to sd_bus_message_appendv to follow elogind naming conventions. (#5753)
Moreover, man page for sd_bus_message_append is updated with reference to new exposed function.
Makefile-man is updated too, to reflect new alias.
Susant Sahani [Tue, 25 Apr 2017 04:13:41 +0000 (09:43 +0530)]
Add Geneve netlink properties to missing.h
Matija Skala [Mon, 24 Apr 2017 16:30:50 +0000 (18:30 +0200)]
redirect stdout/stderr back when closing the pager (#5661)
Zbigniew Jędrzejewski-Szmek [Sat, 22 Apr 2017 23:04:02 +0000 (19:04 -0400)]
shared/extract-word: replace enum with unsigned int to avoid undefined behaviour
../src/basic/extract-word.c:255:22: warning: passing an object that undergoes default argument promotion to 'va_start' has undefined behavior [-Wvarargs]
va_start(ap, flags);
^
../src/basic/extract-word.c:244:77: note: parameter of type 'ExtractFlags' (aka 'enum ExtractFlags') is declared here
int extract_many_words(const char **p, const char *separators, ExtractFlags flags, ...) {
^
../src/basic/extract-word.c:286:22: warning: passing an object that undergoes default argument promotion to 'va_start' has undefined behavior [-Wvarargs]
va_start(ap, flags);
^
../src/basic/extract-word.c:244:77: note: parameter of type 'ExtractFlags' (aka 'enum ExtractFlags') is declared here
int extract_many_words(const char **p, const char *separators, ExtractFlags flags, ...) {
^
2 warnings generated.
I think the relevant part of C99 is 6.7.2.2 Enumeration specifiers:
Each enumerated type shall be compatible with char, a signed integer type, or
an unsigned integer type. The choice of type is implementation-defined, but
shall be capable of representing the values of all the members of the
enumeration.
and 7.16.1.4:
The parameter parmN is the identifier of the rightmost parameter in the
variable parameter list in the function definition (the one just before the
...). If the parameter parmN is declared with the register storage class, with
a function or array type, or with a type that is not compatible with the type
that results after application of the default argument promotions, the behavior
is undefined.
This might cause a real issue if the compiler chooses something that is not an
integer for ExtractFlags. Rework the code to avoid the warning, but add an
assert_cc in a large-valued ExtractFlags element is ever defined and the type
is bumped to something wider than an int.
Zbigniew Jędrzejewski-Szmek [Fri, 21 Apr 2017 16:52:54 +0000 (12:52 -0400)]
basic/log: expose log_dispatch
This is useful when we want to avoid printf formatting on the message.
It's nicer than using log_struct with "%s" as the format, because printf
is slow and with a large message (like from a backtrace) this would require
extra unnecessary memory.
I'm not exposing all the fields in the wrapper: only level and errno.
Those are the most likely to be useful.
Zbigniew Jędrzejewski-Szmek [Thu, 20 Apr 2017 18:15:28 +0000 (14:15 -0400)]
tree-wide: mark log_struct with _printf_ and fix fallout
log_struct takes multiple format strings, each one followed by arguments.
The _printf_ annotation is not sufficiently flexible to express this,
but we can still annotate the first format string, though not its
arguments (because their number is unknown).
With the annotation, the places which specified the message id or similar
as the first pattern cause a warning from -Wformat-nonliteral. This can
be trivially fixed by putting the MESSAGE= first.
This change will help find issues where a non-literal is erroneously used
as the pattern.
Zbigniew Jędrzejewski-Szmek [Thu, 20 Apr 2017 18:09:47 +0000 (14:09 -0400)]
basic/log: fix _printf_ annotation on log_object_internalv
Fixup for
4b58153dd22172d817055d2a09a0cdf3f4bd9db3.
I saw this because of a clang warning. With gcc the -Wformat-nonliteral warning
doesn't seem to work as expected.
In two places, a string constructed with strjoina is used as the pattern. This
is safe, because we're taking a pattern which was already marked with _printf_
and prepending a known value to it. Those places are marked with #pragma to
silence the warning.
Zbigniew Jędrzejewski-Szmek [Wed, 19 Apr 2017 20:32:52 +0000 (16:32 -0400)]
test-exec-util: drop duplicate const
gcc-7 warns about this with -Wduplicate-decl-specifier.
Zbigniew Jędrzejewski-Szmek [Wed, 19 Apr 2017 20:13:27 +0000 (16:13 -0400)]
basic/random-util: add new header for getrandom()
There's some confusion: older man pages specify that linux/random.h
contains getrandom, but newer glibc has it in sys/random.h. Detect if
the newer header is available and include it. We still need the older
header for the flags.
Yu Watanabe [Wed, 19 Apr 2017 06:48:29 +0000 (15:48 +0900)]
tree-wide: fix wrong indent (#5757)
Fixes wrong indent introduced by the commit
43688c49d1fdb585196d94e2e30bb29755fa591b.
Zbigniew Jędrzejewski-Szmek [Tue, 18 Apr 2017 23:04:15 +0000 (19:04 -0400)]
basic/missing.h: drop inclusion of macro.h
It's not necessary for anything.
slodki [Thu, 13 Apr 2017 10:34:59 +0000 (12:34 +0200)]
loginctl: fix typo causing ignoring multiple session IDs (#5732)
Fixes #5733
Zbigniew Jędrzejewski-Szmek [Thu, 6 Apr 2017 04:12:40 +0000 (00:12 -0400)]
basic/def.h: drop TTY_GID definition
We already provide a definition through the configuration system, this
one is duplicate.
afrantzis [Wed, 5 Apr 2017 08:32:55 +0000 (11:32 +0300)]
logind: Stopped inhibitions should be considered inactive (#5698)
Zbigniew Jędrzejewski-Szmek [Thu, 30 Mar 2017 01:31:58 +0000 (21:31 -0400)]
test-sizeof: add some struct-timespec related fields
Might help with #5264.
Zbigniew Jędrzejewski-Szmek [Thu, 30 Mar 2017 01:31:21 +0000 (21:31 -0400)]
test-sizeof: do not link with libelogind-shared
This makes it much quicker to compile.
Jan Synacek [Wed, 29 Mar 2017 06:25:52 +0000 (08:25 +0200)]
basic: forbid rm_rf() to remove paths ending with ".." (#5653)
Fixes: #5644
Franck Bui [Thu, 16 Mar 2017 10:17:09 +0000 (11:17 +0100)]
logind: save/restore session devices and their respective file descriptors
This patch ensures that session devices are saved for each session.
In order to make the revokation logic work when logind is restarted, the
session devices are now saved in the session state files and their respective
file descriptors sent to PID1's fdstore in order to keep them open accross
restart.
This is mandatory in order to keep the revokation logic working. Indeed in case
of input-devices, the same file descriptors must be shared by logind and a
given session controller in order EVIOCREVOKE to work otherwise multiple
sessions can have device access in parallel.
This should be the only remaining and missing piece for making logind fully
restartable.
Fixes: #1163
Matija Skala [Wed, 15 Mar 2017 13:16:40 +0000 (14:16 +0100)]
check for _POSIX_C_SOURCE instead of __USE_POSIX*
check for _GNU_SOURCE as well as sd_event_child_handler_t needs to be
defined correctly while compiling elogind
Matija Skala [Wed, 15 Mar 2017 12:49:05 +0000 (13:49 +0100)]
assigning stdout and stderr is not allowed
Matija Skala [Wed, 15 Mar 2017 12:21:10 +0000 (13:21 +0100)]
fix includes
linux/sockios.h is needed for the SIOCGSTAMPNS macro
xlocale.h is included indirectly in glibc and doesn't even exist in
other libcs
Zbigniew Jędrzejewski-Szmek [Mon, 13 Mar 2017 07:11:24 +0000 (03:11 -0400)]
headers: check that __INCLUDE_LEVEL__ is defined before using it (#5575)
That macro is a gcc extension, and while widely supported, not ubiquitous.
In particular the coverity scanner is having trouble with it.
Michal Sekletar [Fri, 10 Mar 2017 14:16:24 +0000 (15:16 +0100)]
Fix typo in function name (#5565)
Susant Sahani [Wed, 8 Mar 2017 13:09:43 +0000 (18:39 +0530)]
missing.h: add vxlan netlink properties
AsciiWolf [Tue, 7 Mar 2017 06:51:35 +0000 (07:51 +0100)]
sd-bus: drop kdbus-related docs (#5533)
Martin Pitt [Thu, 2 Mar 2017 22:42:01 +0000 (23:42 +0100)]
test: skip instead of fail if crypto kmods are not available
Package build machines may have module loading disabled, thus AF_ALG
sockets are not available. Skip the tests that cover those (khash and
id128) instead of failing them in this case.
Fixes #5524
Zbigniew Jędrzejewski-Szmek [Thu, 23 Feb 2017 08:16:44 +0000 (03:16 -0500)]
Rip out setting of the log level from udev_new and put it in a new function
This function is internal to elogind code, so external users of libudev
will not see those log messages. I think this is better. If we want to
allow that, the function could be put in libudev and exported.
v2: check that the string is more than one char before stripping quotes
Zbigniew Jędrzejewski-Szmek [Thu, 23 Feb 2017 03:57:34 +0000 (22:57 -0500)]
basic/log: split max log level into multiple "realms"
The single log level is split into an array of log levels. Which index in the
array is used can be determined for each compilation unit separately by setting
a macro before including log.h. All compilation units use the same index
(LOG_REALM_SYSTEMD), so there should be no functional change.
v2:
- the "realm" is squished into the level (upper bits that are not used by
priority or facility), and unsquished later in functions in log.c.
v3:
- rename REALM_PLUS_LEVEL to LOG_REALM_PLUS_LEVEL and REALM to LOG_REALM_REMOVE_LEVEL.
Sven Eden [Fri, 21 Jul 2017 05:48:48 +0000 (07:48 +0200)]
Version 233.3 Release
Sven Eden [Fri, 21 Jul 2017 05:46:20 +0000 (07:46 +0200)]
Prep v233.3: Add HAVE_UTMP information to basic/build.h
Sven Eden [Fri, 21 Jul 2017 05:41:40 +0000 (07:41 +0200)]
Prep v233.3: Remove stray character.
Sven Eden [Thu, 20 Jul 2017 16:25:48 +0000 (18:25 +0200)]
Prep v233.3: Add 'loginctl list' as a shorthand for list-sessions
Sven Eden [Thu, 20 Jul 2017 16:16:13 +0000 (18:16 +0200)]
Prep v233.3: Moved four functions, that did not need adaptation in elogind-dbus.c, back to logind-dbus.c.
Sven Eden [Thu, 20 Jul 2017 15:16:50 +0000 (17:16 +0200)]
Prep v233.3: Moved shutdown_or_sleep() and run_helper() to elogind-dbus.c, so elogind-action.c is not needed any more.
Sven Eden [Thu, 20 Jul 2017 15:03:33 +0000 (17:03 +0200)]
Prep v233.3: Apply upstream updates to systemctl commands we mimic.
Sven Eden [Thu, 20 Jul 2017 06:24:36 +0000 (08:24 +0200)]
Prep v233.3: Add a traget 'cccc' to Makefile.am
Sven Eden [Wed, 19 Jul 2017 11:46:39 +0000 (13:46 +0200)]
Prep v233.3: Add missing documentation for API functions exported by elogind. (unreviewed)