Merge ~bryce/ubuntu/+source/chrony:merge-v4.3-1-lunar into ubuntu/+source/chrony:debian/sid

I've run the autopkgtests against the PPA, and note there's a failure or two to investigate:

Creating config file /etc/chrony/chrony.conf with new version

Creating config file /etc/chrony/chrony.keys with new version
dpkg-statoverride: warning: --update given but /var/log/chrony does not exist
Created symlink /etc/systemd/system/chronyd.service → /lib/systemd/system/chrony.service.
Created symlink /etc/systemd/system/multi-user.target.wants/chrony.service → /lib/systemd/system/chrony.service.
Could not execute systemctl: at /usr/bin/deb-systemd-invoke line 145.
Setting up autopkgtest-satdep (0) ...
Processing triggers for dbus (1.14.4-1ubuntu1) ...
Processing triggers for man-db (2.11.1-1) ...
(Reading database ... 104336 files and directories currently installed.)
Removing autopkgtest-satdep (0) ...
autopkgtest [06:52:01]: test fragmented-configuration: [-----------------------
autopkgtest [06:52:02]: test fragmented-configuration: -----------------------]
fragmented-configuration FAIL non-zero exit status 1

The run_system_tests also fail on arm64 and armhf, and upstream-simulation-test-suite fails on s390x, however I've not yet ruled out those are flaky fails. Advice welcome, I'd like to get those resolved before uploading to -proposed.

Checking this for the merge itself at first.

Most delta is retained as-is.
Differences in commit message make sense - thanks for improving my sentences :-)

The noise in it is intersting though:
Your 8ba63418dd22bdcc1d4943627903dfccbc8820f0 kills the changes of
https://salsa.debian.org/debian/chrony/-/commit/164b371d65bdde1834310776b5ec4932280b2b55
I think we'd want to follow there.

The rest of the merge LGTM.

Now to the tests ...
This might be due to the change above that I pointed out.
Without it it will run as _chrony, but try to create directories.
That would fail when starting the service.

For the s390x issue, there are two related cases I had in the past
  Testing hwclock FAIL (on line 71)
One was that the upstream testing isn't done on many platforms, but some expose different behavior.
The new version has:
      hwclock: refactor processing of PHC readings
      hwclock: improve filtering of readings
Maybe one of them breaks on the s390x clock.
The fix was as easy as runnign it on s390x and debugging into the problem.
This led to these back in the day:
- commit c4d6f98b test: accept test result if RTC can't enable RTC_UIE_ON
- commit bff3f51d rtc: extend check for RTCs that don't support interrupts

The other common issue were changes to libc which makes it use other systemcalls which is breaking for chrony as it uses hard seccomp sandboxing. So sometimes (I think it was arm) a change to the syscall behind getting the time led to this breakage.

I hope that helps to get a lead ...

> The noise in it is intersting though:
> Your 8ba63418dd22bdcc1d4943627903dfccbc8820f0 kills the changes of
> https://salsa.debian.org/debian/chrony/-/commit/164b371d65bdde1834310776b5ec4932280b2b55
> I think we'd want to follow there.

Ah, do you mean the prefixed '!'? I did miss that.

I'll see how that affects the tests.

Ok, looks like you were right, the tests are all green now.
Actually I'm still waiting on amd64 (odd that amd64 is the slow poke but whatever).

Results: (from http://autopkgtest.ubuntu.com/results/autopkgtest-lunar-bryce-chrony-merge-v4.3-1/?format=plain)
  chrony @ amd64:
  chrony @ arm64:
    12.01.23 21:00:12 Log 🗒️ ✅ Triggers: chrony/4.3-1ubuntu1~lunar3
      upstream-simulation-test-suite PASS 🟩
      time-sources-from-dhcp-servers PASS 🟩
      run_system_tests PASS 🟩
      run_destructive_system_tests PASS 🟩
      fragmented-configuration PASS 🟩
      dynamically-add-source PASS 🟩
      ntp-server-and-nts-auth PASS 🟩
  chrony @ armhf:
    12.01.23 20:07:26 Log 🗒️ ✅ Triggers: chrony/4.3-1ubuntu1~lunar3
      upstream-simulation-test-suite PASS 🟩
      run_system_tests PASS 🟩
      fragmented-configuration PASS 🟩
      dynamically-add-source PASS 🟩
      ntp-server-and-nts-auth PASS 🟩
  chrony @ ppc64el:
    12.01.23 20:32:44 Log 🗒️ ✅ Triggers: chrony/4.3-1ubuntu1~lunar3
      upstream-simulation-test-suite PASS 🟩
      time-sources-from-dhcp-servers PASS 🟩
      run_system_tests PASS 🟩
      run_destructive_system_tests PASS 🟩
      fragmented-configuration PASS 🟩
      dynamically-add-source PASS 🟩
      ntp-server-and-nts-auth PASS 🟩
  chrony @ s390x:
    12.01.23 20:42:26 Log 🗒️ ✅ Triggers: chrony/4.3-1ubuntu1~lunar3
      upstream-simulation-test-suite PASS 🟩
      time-sources-from-dhcp-servers PASS 🟩
      run_system_tests PASS 🟩
      run_destructive_system_tests PASS 🟩
      fragmented-configuration PASS 🟩
      dynamically-add-source PASS 🟩
      ntp-server-and-nts-auth PASS 🟩
Running:
    time pkg release arch ppa trigger
    1111 chrony lunar amd64 bryce/chrony-merge-v4.3-1 chrony/4.3-1ubuntu1~lunar3
    8628 chrony lunar amd64 bryce/chrony-merge-v4.3-1 chrony/4.3-1ubuntu1~lunar3
Waiting: (none)

I'll go ahead and upload, and hopefully there should be no further issues.

Vcs-Git: https://git.launchpad.net/~bryce/ubuntu/+source/chrony
Vcs-Git-Commit: 270d064713c4b0ed3ebd6b9f7d7be2924709ef39
Vcs-Git-Ref: refs/heads/merge-v4.3-1-lunar
$ dput ubuntu ../chrony_4.3-1ubuntu1_source.changes
D: Setting host argument.
Checking signature on .changes
Checking signature on .dsc
Package includes an .orig.tar.gz file although the debian revision suggests
that it might not be required. Multiple uploads of the .orig.tar.gz may be
rejected by the upload queue management software.
Uploading to ubuntu (via ftp to upload.ubuntu.com):
  Uploading chrony_4.3-1ubuntu1.dsc: done.
  Uploading chrony_4.3.orig.tar.gz: done.
  Uploading chrony_4.3-1ubuntu1.debian.tar.xz: done.
  Uploading chrony_4.3-1ubuntu1_source.buildinfo: done.
  Uploading chrony_4.3-1ubuntu1_source.changes: done.
Successfully uploaded packages.