X-Git-Url: https://www.chiark.greenend.org.uk/ucgi/~ianmdlvl/git?p=elogind.git;a=blobdiff_plain;f=TODO;h=b5f97c7267ec70d958d1358f6007eb8de10e9dd7;hp=526926f8628f822021874734bede6eb5bbc64223;hb=18ead1b00dff180a97507d014325cb46198e6657;hpb=c7eff5ec06c3025d1ef14f955c653259ae7c615b diff --git a/TODO b/TODO index 526926f86..b5f97c726 100644 --- a/TODO +++ b/TODO @@ -1,6 +1,4 @@ Bugfixes: -* Re-enable "fsck -l" when it is ready: - https://bugs.freedesktop.org/show_bug.cgi?id=79576#c5 * Should systemctl status \* work on all unit types, not just .service? @@ -10,7 +8,18 @@ Bugfixes: * properly handle .mount unit state tracking when two mount points are stacked one on top of another on the exact same mount point. +* ExecStart with unicode characters fails in strv_split_quoted: + + [Service] + Environment=ONE='one' "TWO='two two' too" THREE= + ExecStart=/bin/python3 -c 'import sys;print(sys.argv)' $ONE $TWO $THREE + +* MEMORY return code is overloaded for syntax errors in the command line. + str_split_quoted() should return a real return code, so spawn_child can + report the failure properly. + External: + * Fedora: add an rpmlint check that verifies that all unit files in the RPM are listed in %systemd_post macros. * Fedora: post FPC ticket to move add %tmpfiles_create to the packaging guidelines @@ -26,36 +35,61 @@ External: Features: -* busctl: add a pcap writer, using LINKTYPE_DBUS/231 +* figure out when we can use the coarse timers -* man: maybe use the word "inspect" rather than "introspect"? +* sd-resolve: drop res_query wrapping, people should call via the bus to resolved instead -* introduce machines.target to order after all nspawn instances +* add "systemctl start -v foobar.service" that shows logs of a service + while the start command runs. This is non-trivial to do without + races though, since we should flush out all journal messages before + returning from the "systemctl stop". -* systemd-nspawn@.service should fail if some nspawn arg is invalid, with Type=notify +* firstboot: make it useful to be run immediately after yum --installroot to set up a machine. (most specifically, make --copy-root-password work even if /etc/passwd already exists -* "machinectl list" should probably show columns for OS version and IP addresses +* timesyncd + resolved: add ugly bus calls to set NTP and DNS servers per-interface, for usage by NM -* systemctl: if it fails, show log output? +* networkd-wait-online really should have a timeout by default -* maybe add "systemctl edit" that copies unit files from /usr/lib/systemd/system to /etc/systemd/system and invokes vim on them +* add infrastructure to allocate dynamic/transient users and UID ranges, for use in user-namespaced containers, per-seat gdm login screens and gdm guest sessions -* dbus: add new message hdr field for allowing interactive auth, write spec for it. update dbus spec to mandate that unknown flags *must* be ignored... +* machined: add an API so that libvirt-lxc can inform us about network interfaces being removed or added to an existing machine -* maybe introduce AssertXYZ= similar to ConditionXYZ= that causes a unit to fail (instead of skipping it) if some condition is not true... +* maybe add support for specifier expansion in user.conf, specifically DefaultEnvironment= -* remove multi-seat-x now +* code cleanup: retire FOREACH_WORD_QUOTED, port to unquote_first_word() loops instead -* refcounting in sd-resolve is borked +* introduce systemd-timesync-wait.service or so to sync on an NTP fix? -* exponential backoff in timesyncd and resolved when we cannot reach a server +* systemd --user should issue sd_notify() upon reaching basic.target, not on becoming idle + +* configure.ac pretends dbus was optional but actually hardcodes use of dbus' pkg-config file to determine various dbus dirs such as policy and activation dirs + +* consider showing the unit names during boot up in the status output, not just the unit descriptions + +* dhcp: do we allow configuring dhcp routes on interfaces that are not the one we got the dhcp info from? + +* maybe allow timer units with an empty Units= setting, so that they + can be used for resuming the system but nothing else. -* apply start timeout during the "initializing" manager state only, - instead of both "initializing" and "starting". maybe rename the - timeout to "initialization-timeout" then or so? +* what to do about udev db binary stability for apps? (raw access is not an option) -* journald: make use of uid-range.h to managed uid ranges to split - journals in. +* maybe provide an API to allow migration of foreign PIDs into existing scopes. + +* maybe support a new very "soft" reboot mode, that simply kills all processes, disassembles everything, flushes /run and sysvipc, and then reexecs systemd again + +* man: document that corrupted journal files is nothing to act on + +* man: maybe use the word "inspect" rather than "introspect"? + +* "machinectl list" should probably show columns for OS version and IP addresses + +* systemctl: if some operation fails, show log output? + +* systemctl edit: add commented help text to the end, like git commit + +* refcounting in sd-resolve is borked + +* exponential backoff in timesyncd and resolved when we cannot reach a server * tmpfiles: port to unquote_many_words(), similar to sysusers @@ -63,17 +97,8 @@ Features: currently uses FOREACH_WORD and friends. For example, most conf parsing callbacks should use it. -* logind: make the Suspend()/Hibernate() bus calls wait for the for - the job to be completed. before returning, so that clients can wait - for "systemctl suspend" to finish to know when the suspending is - complete. - * merge ~/.local/share and ~/.local/lib into one similar /usr/lib and /usr/share.... -* remove readahead in 217 - -* journald: allows specification of UID range for splitting up journal files - * systemd.show_status= should probably have a mode where only failed units are shown. @@ -94,6 +119,8 @@ Features: to clients. It should also pass on its own timezone information. - provide a way to define a per-network interface default metric value for all routes to it. possibly a second default for DHCP routes. + - allow Name= to be specified repeatedly in the [Match] section. Maybe also + support Name=foo*|bar*|baz ? * resolved: - put networkd events and rtnl events at a higher priority, so that @@ -110,38 +137,27 @@ Features: - edns0 - dname - cname on PTR (?) - -* Ignore .busname units on classic D-Bus boots, systemd-resolved cannot be started on kdbus - without the active policy and should get a Wants=org.freedesktop.resolve1.busname to - pull-in the policy. + - maybe randomize DNS UDP source ports + - maybe compare query section of DNS replies * Allow multiple ExecStart= for all Type= settings, so that we can cover rescue.service nicely * Add a new verb "systemctl top" -* logind: allow users to kill or lock their own sessions - * add new gpt type for btrfs volumes * support empty /etc boots nicely: - nspawn/gpt-generator: introduce new gpt partition type for /usr - fstab-generator: support systemd.volatile=yes|no|state on the kernel cmdline, too, similar to nspawn's --volatile= - - fstab-generator: add support for usr= in addition to root= on the kernel cmdline * generator that automatically discovers btrfs subvolumes, identifies their purpose based on some xattr on them. -* support setting empty environment variables with Environment= and EnvironmentFile= - * timer units: actually add extra delays to timer units with high AccuracySec values, don't start them already when we are awake... * a way for container managers to turn off getty starting via $container_headless= or so... * figure out a nice way how we can let the admin know what child/sibling unit causes cgroup membership for a specific unit -* journalctl: add the ability to look for the most recent process of a binary. journalctl /usr/bin/X11 --pid=-1 or so... - -* set NOTIFY_SOCKET also for control processes - * mount_cgroup_controllers(): symlinks need to get the label applied * For timer units: add some mechanisms so that timer units that trigger immediately on boot do not have the services @@ -151,8 +167,6 @@ Features: * sd-event: generate a failure of a default event loop is executed out-of-thread -* expose "Locked" property on logind sesison objects - * add bus api to query unit file's X fields. * consider adding RuntimeDirectoryUser= + RuntimeDirectoryGroup= @@ -164,11 +178,6 @@ Features: - Define new partition type for encrypted swap? Support probed LUKS for encrypted swap? - Make /home automount rather than mount? -* improve journalctl performance by loading journal files - lazily. Encode just enough information in the file name, so that we - do not have to open it to know that it is not interesting for us, for - the most common operations. - * add generator that pulls in systemd-network from containers when CAP_NET_ADMIN is set, more than the loopback device is defined, even when it is otherwise off @@ -236,10 +245,6 @@ Features: * refuse boot if /etc/os-release is missing or /etc/machine-id cannot be set up -* given that logind now lets PID 1 do all nasty work, we can - probably reduce the capability set it retains substantially. - (we need CAP_SYS_ADMIN for drmSetMaster(), so maybe not worth it) - * btrfs raid assembly: some .device jobs stay stuck in the queue * make sure gdm does not use multi-user-x but the new default X configuration file, and then remove multi-user-x from systemd @@ -252,12 +257,22 @@ Features: ReadOnlyDirectories=... for whitelisting files for a service. * sd-bus: + - xml policy inforcement and bus activated services is broken? + - rework errno registration logic + - how can we make the xml enforcement for native clients unnecessary? + - kdbus: the kernel should not allow messages to be delivered that have a reply serial != 0, reply-expect unset, but no appropriate window + - kdbus: when we fake creds euid being (uint32_t) -1 is weirdly translated + - kdbus: timestamps on kernel's NameOwnerChanged messages? + - kdbus: busctl monitor message ordering is wildly out-of-order? + - kdbus' busnames.target should get pulled in by basic.target + - Ignore .busname units on classic D-Bus boots, systemd-resolved cannot be started on kdbus + without the active policy and should get a Wants=org.freedesktop.resolve1.busname to + pull-in the policy. + - port to sd-resolve for connecting to TCP dbus servers + - kdbus: maybe add controlling tty and ppid metadata fields + - see if we can introduce a new sd_bus_get_owner_machine_id() call to retrieve the machine ID of the machine of the bus itself - when kdbus does not take our message without memfds, try again with memfds - - systemd-bus-proxyd needs to enforce good old XML policy - - allow updating attach flags during runtime - - pid1: peek into activating message when activating a service - introduce sd_bus_emit_object_added()/sd_bus_emit_object_removed() that automatically includes the build-in interfaces in the list - - port to sd-resolve for connecting to TCP dbus servers - see if we can drop more message validation on the sending side - add API to clone sd_bus_message objects - make AddMatch calls on dbus1 transports async? @@ -267,13 +282,19 @@ Features: - kdbus mapping - NameLost/NameAcquired obsolete - GVariant - - "const" properties (posted) - path escaping - update systemd.special(7) to mention that dbus.socket is only about the compatibility socket now - test bloom filter generation indexes + - bus-proxy: when passing messages from kdbus, make sure we properly + handle the case where a large number of fds is appended that we + cannot pass into sendmsg() of the AF_UNIX sokcet (which only accepts + 253 messages) + - kdbus: introduce a concept of "send-only" connections + - kdbus: add counter for refused unicast messages that is passed out via the RECV ioctl. SImilar to the counter for dropped multicast messages we already have. * sd-event - allow multiple signal handlers per signal? + - document chaining of signal handler for SIGCHLD and child handlers * in the final killing spree, detect processes from the root directory, and complain loudly if they have argv[0][0] == '@' set. @@ -290,8 +311,6 @@ Features: * maybe add a generator that looks for "systemd.run=" on the kernel cmdline for container usercases... -* timedatectl: print a nicer message when enabling ntp fails because ntpd/chrony are not installed - * cgtop: make cgtop useful in a container * test/: @@ -333,8 +352,6 @@ Features: * think about window-manager-run-as-user-service problem: exit 0 → activate shutdown.target; exit != 0 → restart service -* use "log level" rather than "log priority" everywhere - * merge unit_kill_common() and unit_kill_context() * introduce ExecCondition= in services @@ -368,12 +385,26 @@ Features: any session we should probably just become a NOP, since that's usually not a real user session but just some system code that just needs setuid(). + - logind: make the Suspend()/Hibernate() bus calls wait for the for + the job to be completed. before returning, so that clients can wait + for "systemctl suspend" to finish to know when the suspending is + complete. + - logind: when the power button is pressed short, just popup a + logout dialog. If it is pressed for 1s, do the usual + shutdown. Inspiration are Macs here. + - logind: allow users to kill or lock their own sessions + - expose "Locked" property on logind sesison objects + - given that logind now lets PID 1 do all nasty work, we can + probably reduce the capability set it retains substantially. + (we need CAP_SYS_ADMIN for drmSetMaster(), so maybe not worth it) + - expose orientation sensors and tablet mode through logind * exec: when deinitializating a tty device fix the perms and group, too, not only when initializing. Set access mode/gid to 0620/tty. * service: watchdog logic: for testing purposes allow ping, but do not require pong * journal: + - consider introducing implicit _TTY= + _PPID= + _EUID= + _EGID= + _FSUID= + _FSGID= fields - import and delete pstore filesystem content at startup - journald: also get thread ID from client, plus thread name - journal: when waiting for journal additions in the client always sleep at least 1s or so, in order to minimize wakeups @@ -395,17 +426,14 @@ Features: - journal-send.c, log.c: when the log socket is clogged, and we drop, count this and write a message about this when it gets unclogged again. - journal: find a way to allow dropping history early, based on priority, other rules - journal: When used on NFS, check payload hashes - - journald: check whether it is OK if the client can still modify delivered journal entries - journal live copy, based on libneon (client) and libmicrohttpd (server) - journald: add kernel cmdline option to disable ratelimiting for debug purposes - refuse taking lower-case variable names in sd_journal_send() and friends. - journald: we currently rotate only after MaxUse+MaxFilesize has been reached. - journal: deal nicely with byte-by-byte copied files, especially regards header - - journal: store euid in journal if it differs from uid - journal: sanely deal with entries which are larger than the individual file size, but where the components would fit - Replace utmp, wtmp, btmp, and lastlog completely with journal - journalctl: instead --after-cursor= maybe have a --cursor=XYZ+1 syntax? - - tmpfiles: when applying ownership to /run/log/journal, also do this for the journal fails contained in it - when a kernel driver logs in a tight loop, we should ratelimit that too. - journald: optionally, log debug messages to /run but everything else to /var - journald: when we drop syslog messages because the syslog socket is @@ -416,6 +444,15 @@ Features: boot, and causes the journal to be moved back to /run on shutdown, so that we do not keep /var busy. This needs to happen synchronously, hence doing this via signals is not going to work. + - optionally support running journald from the command line for testing purposes in external projects + - journald: allow per-priority and per-service retention times when rotating/vacuuming + - journald: make use of uid-range.h to managed uid ranges to split + journals in. + - journalctl: add the ability to look for the most recent process of a binary. journalctl /usr/bin/X11 --pid=-1 or so... + - improve journalctl performance by loading journal files + lazily. Encode just enough information in the file name, so that we + do not have to open it to know that it is not interesting for us, for + the most common operations. * document: - document that deps in [Unit] sections ignore Alias= fields in @@ -452,7 +489,6 @@ Features: - "systemctl mask" should find all names by which a unit is accessible (i.e. by scanning for symlinks to it) and link them all to /dev/null - systemctl list-unit-files should list generated files (and probably with a new state "generated" for them, or so) - - systemctl: maybe add "systemctl add-wants" or so... * timer units: - timer units should get the ability to trigger when: @@ -463,15 +499,6 @@ Features: - when parsing calendar timestamps support the UTC timezone (even if we will not support arbitrary timezone specs, support UTC itself certainly makes sense), also support syntaxes such as +0200 - Modulate timer frequency based on battery state -* update the kernel's TZ (sys_tz) when DST changes - -* sync down the system time to the RTC when: - - CLOCK_REALTIME makes jumps (the user explicitely requested a time set) - - DST/timezone changes && ntp is active && RTC-in-localtime (never do it without ntp) - This takes care of syncing ntpdate updates to the RTC, and DST updates for localtime - mode, it will never touch the RTC if the no reliable time source is active or the - user did not request anything like it. - * add libsystemd-password or so to query passwords during boot using the password agent logic * If we show an error about a unit (such as not showing up) and it has no Description string, then show a description string generated form the reverse of unit_name_mangle(). @@ -495,6 +522,9 @@ Features: - bind mount read-only the cgroup tree higher than nspawn - refuses to boot containers without /etc/machine-id (OK?), and with empty /etc/machine-id (not OK). - support taking a btrfs snapshot at startup and dropping it afterwards + - introduce machines.target to order after all nspawn instances + - systemd-nspawn@.service should fail if some nspawn arg is invalid, with Type=notify + - PID 1 doesn't apply nspawns devices cgroup policy * cryptsetup: - cryptsetup-generator: allow specification of passwords in crypttab itself @@ -509,8 +539,6 @@ Features: * make timer units go away after they elapsed -* come up with a nice way to write queue/read_ahead_kb for a block device without interfering with readahead - * move PID 1 segfaults to /var/lib/systemd/coredump? * create /sbin/init symlinks from the build system @@ -611,13 +639,6 @@ Features: * and a dbus call to generate target from current state -* readahead: - - drop /.readahead on bigger upgrades with yum - - move readahead files into /var (look for them with .path units?) - - readahead: use BTRFS_IOC_DEFRAG_RANGE instead of BTRFS_IOC_DEFRAG ioctl, with START_IO - - readahead: when bumping /sys readahead variable save mtime and compare later to detect changes - - readahead: make use of EXT4_IOC_MOVE_EXT, as used by http://e4rat.sourceforge.net/ - * GC unreferenced jobs (such as .device jobs) * write blog stories about: @@ -627,7 +648,6 @@ Features: - how to make changes to sysctl and sysfs attributes - remote access - how to pass throw-away units to systemd, or dynamically change properties of existing units - - how to integrate cgconfig and suchlike with systemd - testing with Harald's awesome test kit - auto-restart - how to develop against journal browsing APIs @@ -655,8 +675,6 @@ Features: * fingerprint.target, wireless.target, gps.target, netdevice.target -* drop cap bounding set in readahead and other services - * systemd-python: - figure out a simple way to wait for journal events in a way that works with ^C @@ -746,8 +764,10 @@ External: * register catalog database signature as file magic -* zsh shell completion: - should complete options, but currently - does not +* zsh shell completion: + - - should complete options, but currently does not + - systemctl add-wants,add-requires, edit + Regularly: