Created attachment 927587 [details] Systemd crash when linking with mold 2.38.1-r1 Latest mold 2.38.1-r1 destroys the whole installation when emerging systemd. I noticed this two days ago when performing the Python 3.13 system update. The problem is not actually related to the Python update, it is that the currently unstable mold 2.38.1-r1 can not be used to emerge systemd. This is for stable systemd 256.10 as well as for 257.5. Postinstall systemd fully crashes the box, rebooting completely fails. So I had to recover from USB stick. After restoring my box from USB, I am now testing with stable mold 2.37.1 and that version seems to work fine. This is for AMD64, with # cat /etc/portage/make.conf COMMON_FLAGS="-march=native -O3 -flto -pipe -fomit-frame-pointer" Perhaps it would work without -O3 -flto, but I can not test that now, as the whole recovery process is each time a lot of work. I will attach the systemd build log when linking with mold 2.38.1-r1 showing the postinstall issues.
Wonder if it installs a broken NSS module, which would explain chgrp and such failing immediately after.
Maybe it's another manifestation of the thing https://github.com/rui314/mold/commit/fded2d8a77cda00944b4ef982f617c07ae2ce46d fixes.
The bug has been referenced in the following commit(s): https://gitweb.gentoo.org/repo/gentoo.git/commit/?id=d98b86690097a1c9d38e0aeabf6372e9a31f7c9a commit d98b86690097a1c9d38e0aeabf6372e9a31f7c9a Author: Sam James <sam@gentoo.org> AuthorDate: 2025-05-05 06:47:18 +0000 Commit: Sam James <sam@gentoo.org> CommitDate: 2025-05-05 06:48:29 +0000 sys-devel/mold: add 2.39.0 May fix bug #955406 but not clear. Bug: https://bugs.gentoo.org/955406 Signed-off-by: Sam James <sam@gentoo.org> sys-devel/mold/Manifest | 1 + sys-devel/mold/mold-2.39.0.ebuild | 117 ++++++++++++++++++++++++++++++++++++++ 2 files changed, 118 insertions(+)
