File _patchinfo of Package patchinfo.21260

<patchinfo incident="21260">
  <issue tracker="bnc" id="1189480">systemd-udevd.service fails with undefined symbol: fchmod_and_chown</issue>
  <issue tracker="bnc" id="1188588">L3-Question: journalctl --flush returns error when journald is not running unlike in 15sp2 and older</issue>
  <issue tracker="bnc" id="1189446">systemd  fails to use the hardware watchdog</issue>
  <issue tracker="bnc" id="1188713">L3: multipath devices default to bfq scheduler</issue>
  <issue tracker="bnc" id="1184994">elevator= makes udev skip the rest of the io-scheduler rules</issue>
  <issue tracker="bnc" id="1134353">adopting BFQ to control I/O</issue>
  <issue tracker="bnc" id="1188291">Error logged by systemd update: systemd-udevd-control.socket: Socket service systemd-udevd.service already active, refusing.</issue>
  <issue tracker="jsc" id="SLE-21032"/>
  <packager>fbui</packager>
  <rating>moderate</rating>
  <category>recommended</category>
  <summary>Recommended update for systemd</summary>
  <description>This update for systemd fixes the following issues:

- Switch I/O scheduler from 'mq-deadline' to 'bfq' for rotating disks(HD's) (jsc#SLE-21032, bsc#1134353).
- Multipath: Rules weren't applied to dm devices (bsc#1188713).
- Ignore obsolete "elevator" kernel parameter (bsc#1184994).
- Remove kernel unsupported single-queue block I/O.
- Make sure the versions of both udev and systemd packages are always the same (bsc#1189480).
- Avoid error message when updating active udev on sockets restart (bsc#1188291).

- Merge of v246.16, for a complete list of changes, visit:
   https://github.com/openSUSE/systemd/compare/8d8f5fc31eece95644b299b784bbfb8f836d0108...f5c33d9f82d3d782d28938df9ff09484360c540d

- Drop 1007-tmpfiles-follow-SUSE-policies.patch:
   Since most of the tmpfiles config files shipped by upstream are
   ignored (see previous commit "Drop most of the tmpfiles that deal
   with generic paths"), this patch is no more relevant.

Additional fixes:
- core: make sure cgroup_oom_queue is flushed on manager exit.
- cgroup: do 'catchup' for unit cgroup inotify watch files.
- journalctl: never fail at flushing when the flushed flag is set (bsc#1188588).
- manager: reexecute on SIGRTMIN+25, user instances only.
- manager: fix HW watchdog when systemd starts before driver loaded (bsc#1189446).
- pid1: watchdog modernizations.
</description>
</patchinfo>