summaryrefslogtreecommitdiff
AgeCommit message (Collapse)Author
2010-01-28add new manager initializer callbacks to per-unit type vtableLennart Poettering
2010-01-28add new hexchar() callLennart Poettering
2010-01-28add unit_name_escape_path() callLennart Poettering
2010-01-28relax unit name rules a bitLennart Poettering
2010-01-28make use of unit_add_dependency_by_name() where applicableLennart Poettering
2010-01-28add new unit_add_dependency_by_name() callLennart Poettering
2010-01-28Use @ instead = as abstract namespace socket prefixLennart Poettering
@ makes more sense since the NUL addr contained in the sockaddr's sun_path field can also be written as ^@. Also, in the .socket files writing "ListenStream==foo" is kinda ugly. Finally, tools like strace decode it with an @ too.
2010-01-28the systemd logger only matters in the abstract namespaceLennart Poettering
2010-01-28set nice/oom_adjust only when asked forLennart Poettering
2010-01-28rework config file load logicLennart Poettering
2010-01-28add more special unitsLennart Poettering
2010-01-28implement proper logging for servicesLennart Poettering
2010-01-28add infrastructure for special unitsLennart Poettering
2010-01-28minor cleanupLennart Poettering
2010-01-28unit_merge() needs a proper reviewLennart Poettering
2010-01-28when shortcutting states do not skip state transitionsLennart Poettering
2010-01-28when resetting signal handlers, set them to SA_RESTARTLennart Poettering
2010-01-28add non-failing close() variantLennart Poettering
2010-01-27implement /dev/kmsg logginLennart Poettering
2010-01-27implement trivial socket activated logger daemonLennart Poettering
2010-01-27drop unused function pointerLennart Poettering
2010-01-27make sure we don't ask the same socket unit for fds more than onceLennart Poettering
2010-01-27process only one epoll event at a timeLennart Poettering
if we ask for more than one from the kernel we might need to check for the validity of the ptr element since event might be processed after its ptr was already destructed.
2010-01-27make sure the log functions don't modify errnoLennart Poettering
2010-01-27port fragment loader to new utility functionsLennart Poettering
2010-01-27port config parser to new utility functionsLennart Poettering
2010-01-27introduce strv_contains()Lennart Poettering
2010-01-27implement new utility functions strstrip() and file_in_same_dir()Lennart Poettering
2010-01-27properly implement target unitLennart Poettering
2010-01-27extend examples a bitLennart Poettering
2010-01-27notify socket unit when service unit diesLennart Poettering
2010-01-27reset signal handlers on startupLennart Poettering
2010-01-27drop O_CLOEXEC/O_NONBLOCK from files intended for forked clientsLennart Poettering
2010-01-27reset signal mask when forkingLennart Poettering
2010-01-27fix socket state machine to listen properly on socketsLennart Poettering
2010-01-27properly recognize successfull scriptsLennart Poettering
2010-01-27trap some signalsLennart Poettering
2010-01-27fix an assert when forkingLennart Poettering
2010-01-27don't allow comments at the end of linesLennart Poettering
2010-01-27don't allow zero length unit namesLennart Poettering
2010-01-27don't choke if there are no sockets for a serviceLennart Poettering
2010-01-27don't choke on invalid dropin file namesLennart Poettering
2010-01-27exit cleanly on SIGINTLennart Poettering
2010-01-27yay, we can start socket unitsLennart Poettering
2010-01-27fix quoted parserLennart Poettering
2010-01-27add some test script outputLennart Poettering
2010-01-27improve dump output for socketsLennart Poettering
2010-01-27fill commands into right fields for socketsLennart Poettering
2010-01-27properly terminate strings with NUL byteLennart Poettering
2010-01-27fully clean up job dependencies on abort, tooLennart Poettering