Age | Commit message (Collapse) | Author | Files | Lines |
|
(int)unix_user should be (uid_t) because it may be (un)signed.
POSIX is vague about the datatype
http://pubs.opengroup.org/onlinepubs/9699919799/functions/V2_chap02.html#tag_15_12
|
|
|
|
uid and pid could be used uninitialized, CWE-457
|
|
Coverity CID 89285, CWE-484
|
|
|
|
|
|
|
|
Otherwise GDBus doesn't think we handled the call and gets angry.
|
|
Otherwise GDBus doesn't think we handled the call and gets angry.
|
|
|
|
|
|
When a session has a name-owner change and loses its owner, we
need to remove it from the manager as it's going away.
|
|
We need to set the values in the keyfile, not try to read values
in the keyfile.
|
|
We need to parse the string into a boolean or int before stuffing
it into the GVariant. Also add a check for is-local in
test-session-leader.
|
|
|
|
GDBus/GVariant dislikes NULL strings so we have to add workarounds
for that shortcoming.
|
|
So we can easily track when dbus calls happen when compiled with
--enable-debug=full
Additionally fix the session Activate signal failure handling.
|
|
Otherwise we get a warning message:
GLib-GObject-CRITICAL: g_object_unref: assertion 'G_IS_OBJECT (object)' failed
|
|
|
|
Add some checks for opening and closing sessions from the Manager.
Also unref the variants that get created and close the inhibit fd
when done.
|
|
Fix make distcheck, bump minimum glib to 2.40 since we're using
that with this port to GDBus. Properly return the error messages
when calling Inhbit and it fails and fix some other GDBus return
values.
|
|
Might as well continue to test everything and print the output to
easily compare changes before and after the port to GDBus. We
don't currently have a way to create seats on the fly to test out
more, but that will come with the multi-seat code when we get there.
|
|
Workaround the issue of not easily being able to addpend new
values to the GVariant.
|
|
Let test-session test all the public dbus methods of CkSession.
Additionally, fix the active signal on CkSession and a couple
other warning messages.
|
|
Fix the complete_get_sessions_for_unix_user and simplify CkSeat's
dbus_get_sessions.
|
|
|
|
|
|
This adds some of the missing methods that were missing during
the port. Additionally, some of them have been reworked to fix
warning messages.
|
|
|
|
|
|
Now we can start the daemon, call ck-launch-session and get info
about it from ck-list-sessions.
|
|
Need to add in the dbus stuff, the codegen doesn't help us here.
|
|
Fix some method calls, add more public dbus api handling for seats.
Added the dbus library deps back in so the ck-connector doesn't
have to be ported right away. Would like to start working on
fixing what's currently ported and testing everything.
|
|
We can't pass around the GVariantBuilder and parse it at the same
time so we'll need to fix the is-local in another way.
Fix the signal emission on the manager.
We may need to use a GDBusObjectManager to get the signals for
interface-added/interface-removed on our Sessions (and eventually
Seats?)
|
|
Port CkDevice to using GVariants rather than GValues.
Have ck-session-leader return a GVariantBuilder so is-local can
be added.
|
|
Getting a better understanding of GVariant. Added a test of
ck-session-leader and a mock session-info to feed it usable
junk data. There's still more work to do on this big GDBus
port.
|
|
Quite a bit of work here. ck-session-leader needs to be ported.
|
|
Port the CkSeat to GDBus using the gdbus-codegen. Same as with
CkSession there's now a split between public and private signals,
properties, and function calls.
CkSeat and CkSession host private signals and properties
ConsoleKitSeat and ConsoleKitSession are public D-BUS exported ones.
https://github.com/ConsoleKit2/ConsoleKit2/issues/8
|
|
Port CKSession to GDBus using gdbus-codegen. The generated
ConsoleKitSession object properties are used where possible.
A test-session binary is generated to test some of the
CkSession properties/DBUS calls. CKSession expects to be
passed a GDBusConnection in ck_session_new but this doesn't
happen because CKManager hasn't been ported yet.
https://github.com/ConsoleKit2/ConsoleKit2/issues/8
|
|
|
|
|
|
openbsd: unbreak get_kinfo_proc() by properly initializing len
|
|
No need to call sysctl(3) twice since we know exactly how many
structures it wants back: just one.
|
|
|
|
On remote ck session, there was a warning generated during log out.
This patch should fix that warning as well as add a couple debug
lines to help future troubleshooting. Thanks to hhoffstaette for
helping to troubleshoot this issue. For details see:
https://github.com/ConsoleKit2/ConsoleKit2/pull/15
|
|
This allows use of /run on tmpfs instead of /var/run
without downstream symlink magic or patches.
|
|
This reverts commit 417e387750920b9e092705863a863e71a77ea431.
Causes a regression with polkit authorizations. See:
https://github.com/ConsoleKit2/ConsoleKit2/issues/16
https://github.com/ConsoleKit2/ConsoleKit2/issues/12
|
|
We never needed to use an external helper script with the original
ConsoleKit for ck-get-x11-*, so drop these.
Make sure we install the other tools and simplify the scripts in the
process.
|
|
This way we can prevent the user applications from attempting to
perform multiple system events at the same time. The additional
attempts will fail.
|
|
This patch adds the PrepareForSleep/Shutdown signals so apps can
listed for these signals and perform operations prior to the event
such as logging out of online chatrooms.
|