manager: print ephemeral information about running jobs' timeouts

[elogind.git] / TODO

diff --git a/TODO b/TODO
index a567669aa70a6fad841ca22f3237db8b62129d12..b76db091e883db36f45d341340bc6a6829fac1b0 100644 (file)
--- a/TODO
+++ b/TODO
@@ -1,4 +1,10 @@
 Bugfixes:
+* killing nspawn with ]]] results in:
+   machine-f20.scope stopping timed out. Killing.
+   Stopped Container f20.
+   Unit machine-f20.scope entered failed state.
+  and it cannot be started again
+
 * enabling an instance unit creates a pointless link, and
   the unit will be started with getty@getty.service:
     $ systemctl enable getty@.service
@@ -21,8 +27,6 @@ Bugfixes:
 
 External:
 
-* ps should gain colums for slice
-
 * Fedora: when installing fedora with yum --installroot /var/run is a directory, not a symlink
    https://bugzilla.redhat.com/show_bug.cgi?id=975864
 
@@ -32,14 +36,20 @@ External:
 
 Features:
 
+* hookup nspawn and PrivateNetwork=yes with "ip netns"
+
+* socket units: support creating sockets in different namespace,
+  opening it up for JoinsNamespaceOf=. This would require to fork off
+  a tiny process that joins the namespace and creates/binds the socket
+  and passes this back to PID1 via SCM_RIGHTS. This also could be used
+  to allow Chown/chgrp on sockets without requiring NSS in PID 1.
+
 * sd-resolve:
   - make sure event loop integration works similar to event loop integration in other libs
   - maybe drop _free() call, introduce _unref() instead
-  - sd_resolve_freeanswer() is probably pointless, we don't define such a function for any other return values, but expect people to call free directly
   - maybe simplify function calls that take a sd_resolve_query object, to not also require the matching sd_resolve?
   - maybe drop the "n_proc" parameter to sd_resolve_new()?
   - change all functions to return "int" and negative errno errors
-  - getnext and getnqueries should be renamed to get_next() and get_n_queries(), i.e. functions that are not modelled directly after the libc counterparts don't need to be a single word...
 
 * New service property: maximum CPU and wallclock runtime for a service
 
@@ -135,18 +145,14 @@ Features:
   ReadOnlyDirectories=... for whitelisting files for a service.
 
 * sd-bus:
-  - let the activator peek the peer connection with KDBUS_CMD_MSG_SRC and log it
   - when kdbus doesn't take our message without memfds, try again with memfds
   - implement monitor logic
   - see if we can drop more message validation on the sending side
   - add API to clone sd_bus_message objects
-  - SD_BUS_COMMENT() macro for inclusion in vtables, syntax inspired by gdbus
   - systemd-bus-proxyd needs to enforce good old XML policy
   - upload minimal kdbus policy into the kernel at boot
   - kdbus: matches against source or destination pids for an "strace -p"-like feel. Problem: The PID info needs to be available in userspace too...
-  - longer term:
-    * priority queues
-    * priority inheritance
+  - longer term: priority inheritance
   - check sender of response messages
   - dbus spec updates:
        - kdbus mapping
@@ -154,10 +160,13 @@ Features:
        - GVariant
        - "const" properties (posted)
   - port exit-on-idle logic to byebye ioctl
-  - make use of "drop" ioctl in pid 1 bus activation
   - bus proxy: override unique id sender for messages from driver to match the well-known name
   - bus driver: GetNameOwner() for "org.freedesktop.DBus" should return "org.freedesktop.DBus"
   - update systemd.special(7) to mention that dbus.socket is only about the compatibility socket now
+  - bus proxy should fake seclabel when connecting to kdbus
+  - allow updating attach flags during runtime
+  - pid1: peek into activating message when activating a service
+  - when service activation failed, drop one message from the activator queue
 
 * sd-event
   - allow multiple signal handlers per signal?