Yu Watanabe

@spikeh and @davem330 I think the commit https://github.com/torvalds/linux/commit/8debcf5832c3e8a6baaea27c75ad8a6ba5077beb should be fixed something like the following. ```diff diff --git a/drivers/net/netdevsim/netdev.c b/drivers/net/netdevsim/netdev.c index c22897bf5509..8b8a5532a33d 100644 --- a/drivers/net/netdevsim/netdev.c +++ b/drivers/net/netdevsim/netdev.c @@ -324,7 +324,7...

Is this for #32898?

systemd-journald.service should be running at the very last stage of shutdown process. To make journald stopped only on soft-reboot, it is necessary to make soft-reboot.target is a usable synchronization point....

> softreboot is new enough and there were actual issues, but I don't think we should change these ones after RC2, unless there's a specific bug that is being solved....

> So yeah, I think this should be exclusive to soft-reboot, which is realistically more similar to switch-root rather than normal shutdown. > > If one day an actual issue...

udevadm already handles device units in some situations, e.g. `udevadm info foo.device` works.

Fix is waiting in #32850. Workaround for the issue is specifying unit path explicitly e.g. /run/systemd/transient/xyz.service.

Please provide the result of `cat ~/.config/systemd/user/podmandemo.service`.

I guess this is expected. See below: https://github.com/systemd/systemd/blob/1bd0cefcab937140a07acf0a3ee528a4900e3072/src/journal/journald-rate-limit.c#L173-L182

Please backport (at least, the last commit) to v255-stable.