sys-apps/systemd-232 just landed on my ~x86 system and with it, my system would no longer boot. After a few hours of trial and error, I found the problem, and indeed it already seems to be known upstream, see: https://github.com/systemd/systemd/issues/4575 Bottom line: There's an issue on 32 bit i386 with systemd-232 that would require you to edit all unit files in /usr/lib/systemd/system which contain "RestrictAddressFamilies=" and comment these lines in order to get a bootable system. Since I believe a non-bootable system is a rather serious issue, even on ~, and since one might need a few hours to actually figure out what's wrong (like I did), I'd suggest that we might probably do good by masking systemd-232 on ~x86 for the time being. If we don't go that route, this bug report will probably at least point other ~x86 Gentoo users in the right direction if they stumble upon the same problem. :-)
Masked on x86. https://gitweb.gentoo.org/repo/gentoo.git/commit/?id=c0a74c15fb9d5f7ee67a67e60c61a16901f1658d
Just adding some more background information: https://github.com/systemd/systemd/issues/4575#issuecomment-259376909 https://bugs.launchpad.net/ubuntu/+source/glibc/+bug/1576066 http://patchwork.sourceware.org/patch/7679/ It seems that the fix will have to be made in libc. Also, according to https://github.com/systemd/systemd/issues/4575#issuecomment-259160658, the problem is also present on s390, but I don't think anything needs to be masked for that arch, as I can't find (~)s390 among the KEYWORDS in any of the existing systemd ebuilds.
In upstream report I see they point to this backports: https://anonscm.debian.org/cgit/pkg-systemd/systemd.git/commit/?id=c31b5ee9b9c79c9d2ac491920b23a8084a4ccc46
this should be fixed finally in 233