Potential Data Loss from regression in systemd 252

Last updated 2023-02-02 02:06:45 SGT

Recently, I found (after wasting a literal whole day on debugging, including accidentally breaking grub) that systemd's suspend-then-hibernate functionality had stopped working. This turns out to be because the HibernateDelaySec setting for a RTC wakeup call in sleep.conf is no longer respected in version 252, in favour of an ACPI low-power wakeup call. Unfortunately, this is not being issued on my laptop (Framework), so it just runs out of battery rather than hibernating.

Also unfortunately, the dev responsible for this change has decided to take things personally and locked the discussion in the bug report rather than engaging with the obvious user complaints this has generated. I am living with systemd downgraded to 251 until this issue is fixed.

Edit 1/2/2023: This has been fixed in a recent pull request.

comments powered by Disqus