X-Git-Url: http://www.chiark.greenend.org.uk/ucgi/~ianmdlvl/git?a=blobdiff_plain;f=TODO;h=27d5df6ae1438609f395c26dc3d93c545081f0bf;hb=f6940be7825755d77ade4cd42231aab9e3580623;hp=c7292150778559ef49035dac606ba441a7547a2d;hpb=241da3287d0c16f79e2f415f17543599b3e73c85;p=elogind.git diff --git a/TODO b/TODO index c72921507..27d5df6ae 100644 --- a/TODO +++ b/TODO @@ -11,6 +11,14 @@ Bugfixes: * properly handle .mount unit state tracking when two mount points are stacked one on top of another on the exact same mount point. +* stop importing kernel exported env variables. The utterly broken logic in + the kernel exports every kernel command line option which is not recognized + as a built-in module option as an env variable. Systemd should not pass-on + that nonsense, a kernel command line option is a command line option not an + env variable: + $ cat /proc/252/environ + initrd=\6a9857a393724b7a981ebb5b8495b9ea\3.10.0-2.fc20.x86_64\initrd + Fedora 19: * external: maybe it is time to patch procps so that "ps" links to @@ -26,40 +34,65 @@ Fedora 19: * when installing fedora with yum --installroot /var/run is a directory, not a symlink -Features: +CGroup Rework Completion: + +* user@.service and session-*.scope should get posession of their own cgroups + +* introduce "mainpid" for scopes + +* implement system-wide DefaultCPUAccounting=1 switch (and similar for blockio, memory?) + +* implement per-slice CPUFairScheduling=1 switch + +* handle jointly mounted controllers correctly + +* logind: implement session kill exceptions + +* fix machine regstration to forward property array + +* add implicit slice for instantiated services * make BlockIODeviceWeight=, BlockIODeviceBandwidth= runtime settable * split up BlockIOWeight= and BlockIODeviceWeight= -* how to reset dynamically changed attributes sanely? +* introduce high-level settings for RT budget, swapiness -* when reloading configuration, apply new cgroup configuration +* man: document new bus apis -* implement system-wide DefaultCPUAccounting=1 switch (and similar for blockio, memory, fair scheduling?) +Features: -* handle jointly mounted controllers correctly +* when a kernel driver logs in a tight loop we should ratelimit that too. -* split out CreateMachine into systemd-machined +* journald: when we drop syslog messages because the syslog socket is + full, make sure to write how many messages are lost as first thing + to syslog when it works again. -* "transient" units, i.e units that are not sourced from disk but - created only transiently via bus calls +* prohibit Restart= set with Type=oneshot -* introduce new Scope unit type then make logind's session and machine - registration use this to set up cgroups +* man: the documentation of Restart= currently is very misleading and suggests the tools from ExecStartPre= might get restarted. -* should Slice= be part of [Unit] or of [Service]? +* load .d/*.conf dropins for device units -* journald: make sure ratelimit is actually really per-service with the new cgroup changes +* move systemctl set-log-level to systemd-analyze? + +* move "systemctl dump" to systemd-analyze + +* add a fixed dbus path for "my own unit", "my own session", ... to PID1, logind, ... -* move systemctl dump to systemd-analyze +* service_coldplug() appears to reinstall the wrong stop timeout watch? + +* transient units: allow creating auxiliary units with the same call + +* how to reset dynamically changed attributes sanely? + +* when reloading configuration, apply new cgroup configuration + +* journald: make sure ratelimit is actually really per-service with the new cgroup changes * libsystemd-logind: sd_session_is_active() and friends: verify validity of session name before appending it to a path -* logind: when a PAM client calls ReleaseSession() start a timeout and - kill the session entirely after that is reached. - * gparted needs to disable auto-activation of mount units somehow, or maybe we should stop doing auto-activiation of this after boot entirely. https://bugzilla.gnome.org/show_bug.cgi?id=701676 @@ -160,10 +193,6 @@ Features: so that the coredump is properly written to the user's own journal file. -* move /usr/lib/modules/$(uname -r)/modules.devname parsing from udevd to - kmod static-nodes - call kmod as an early service, and drop CAP_MKNOD from udevd.service - * seems that when we follow symlinks to units we prefer the symlink destination path over /etc and /usr. We shouldn't do that. Instead /etc should always override /run+/usr and also any symlink