Comment 25 for bug 1949089

Thanks ddstreet, I agree 918e6f1c0151429f5095355f4f3f74f16e79724a could be related as well, but unfortunately that commit produced quite some fall-out.

Nevertheless, I've prepared a new systemd build in my PPA and the "core18_20211118_amd64.snap" at https://people.ubuntu.com/~slyon/uc18/ cherry-picking the changes from

https://github.com/systemd/systemd/pull/8675 (incl. 918e6f1c)

As I said, there are quite some other PRs related to this, in order to clean-up after it
https://github.com/systemd/systemd/pull/8798
https://github.com/systemd/systemd/pull/9200
https://github.com/systemd/systemd/pull/9229
https://github.com/systemd/systemd/pull/9366

In total that should be approx. 1000 changed LOC... So I'm not sure how suitable this is for SRU. But let's first give this newest build a try and check if the initial commit/PR changes anything wrt to our core problem here, if so we can into the details of those backports (and hopefully skip manny commits of those).