]> git.ipfire.org Git - thirdparty/openembedded/openembedded-core-contrib.git/commit
systemd.bbclass: Make systemd_postinst run as intended
authorPeter Kjellerstedt <pkj@axis.com>
Mon, 8 Sep 2025 16:05:04 +0000 (18:05 +0200)
committerRichard Purdie <richard.purdie@linuxfoundation.org>
Thu, 11 Sep 2025 09:45:09 +0000 (10:45 +0100)
commit5b86efb5c9ca782fd0f8ff306f82583ec1e5e909
tree6db3ba2d841e40b5f460a6d94634adcc15724c81
parent30140cb9354fa535f68fab58e73b76f0cca342e4
systemd.bbclass: Make systemd_postinst run as intended

After the switch from using a systemctl written in Python to using the
official version of systemctl from the systemd project, the
systemd_postinst function has effectively not been executed during the
rootfs creation. The reason is that systemctl provided by
systemctl-native fails if run without arguments (as systemd_postinst
does):

  Failed to connect to system scope bus via local transport: Operation
  not permitted (consider using --machine=<user>@.host --user to connect
  to bus of other user)

This is not seen in the logs since stderr is sent to /dev/null, and the
only way to tell that there is a problem is because systemd services
that are expected to be enabled aren't running.

The reason this has gone unnoticed is because systemd_handle_machine_id
in rootfs-postcommands.bbclass will call systemctl preset-all, which in
most cases will create the missing links to enable the systemd services.

This change effectively reverts commit
a52e66762c0c51918b1ba3d4622759637b6e920a (systemd.bbclass: update
command to check systemctl available) and instead only runs systemctl
without arguments (to determine that it can communicate with systemd)
when executed on target.

Signed-off-by: Peter Kjellerstedt <peter.kjellerstedt@axis.com>
Signed-off-by: Mathieu Dubois-Briand <mathieu.dubois-briand@bootlin.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
meta/classes-recipe/systemd.bbclass