reboot on rpi4 takes longer than usual due to systemd-udevd hanging

[anonymouse64]

See the serial log:

[FAILED] Failed unmounting /run/mnt/data.
[FAILED] Failed unmounting /usr/lib/modules.
[  128.104486] [vc_sm_connected_init]: failed to connect VCHI instance (ret=1)
[  128.104556] bcm2835_audio: probe of bcm2835_audio failed with error -5
[  128.122220] vc_sm_cma: module is from the staging directory, the quality is unknown, you have been warned.
[  128.132814] bcm2835_mmal_vchiq: module is from the staging directory, the quality is unknown, you have been warned.
[  128.132957] bcm2835_mmal_vchiq: module is from the staging directory, the quality is unknown, you have been warned.
[  128.134801] bcm2835_mmal_vchiq: module is from the staging directory, the quality is unknown, you have been warned.
[  128.148335] bcm2835_codec: module is from the staging directory, the quality is unknown, you have been warned.
[  128.151684] bcm2835_v4l2: module is from the staging directory, the quality is unknown, you have been warned.
[  128.156949] bcm2835_isp: module is from the staging directory, the quality is unknown, you have been warned.
[  138.125549] systemd-shutdown[1]: Waiting for process: systemd-udevd, systemd-udevd, systemd-udevd, systemd-udevd
[  200.276215] bcm2835_mmal_vchiq: Failed to connect VCHI instance (status=1)
[  200.276358] systemd-udevd[526]: bcm2835-isp: Worker [568] processing SEQNUM=4104 killed
[  200.283260] bcm2835-isp: probe of bcm2835-isp failed with error -5
[  200.291419] bcm2835_mmal_vchiq: Failed to connect VCHI instance (status=1)
[  200.291420] systemd-udevd[526]: bcm2835-camera: Worker [549] processing SEQNUM=4102 killed
[  200.291469] bcm2835_mmal_vchiq: Failed to connect VCHI instance (status=1)
[  200.291478] systemd-udevd[526]: bcm2835-codec: Worker [548] processing SEQNUM=4103 killed
[  200.291481] bcm2835-codec: probe of bcm2835-codec failed with error -5
[  200.292198] systemd-udevd[526]: Worker [548] terminated by signal 9 (KILL)
[  200.292214] systemd-udevd[526]: bcm2835-codec: Worker [548] failed
[  200.298191] systemd-udevd[526]: Worker [568] terminated by signal 9 (KILL)
[  200.304629] bcm2835-camera: probe of bcm2835-camera failed with error -5
[  200.313042] systemd-udevd[526]: bcm2835-isp: Worker [568] failed
[  200.368169] systemd-udevd[526]: Worker [549] terminated by signal 9 (KILL)
[  200.375171] systemd-udevd[526]: bcm2835-camera: Worker [549] failed
[  200.917037] sd-umoun[1317]: Failed to unmount /oldroot: Device or resource busy
[  200.925963] sd-remou[1318]: Failed to remount '/oldroot/run/mnt/data' read-only: Device or resource busy
[  200.937105] sd-umoun[1319]: Failed to unmount /oldroot/run/mnt/data: Device or resource busy
[  200.947116] sd-umoun[1320]: Failed to unmount /oldroot/run: Device or resource busy
[  201.038273] shutdown[1]: Could not detach loopback /dev/loop0: Device or resource busy
[  201.048094] shutdown[1]: Failed to finalize  file systems, loop devices, ignoring
[  201.087635] reboot: Restarting system

This is with rev 782 on arm64 on a Raspberry Pi 4 Model B Revision 1.1. Normally, reboots will happen very quickly, within a few seconds of the [FAILED] Failed unmounting /run/mnt/data. message I see (and AIUI that particular message is expected, but if that is a bug happy to file a new one for that too) but here it took a full minute after that message before it actually rebooted.

Note that this was the same system as #80, so perhaps the issues are related.

[xnox]

once booted, can you paste output of ps -auxf or some such?

Normally there shouldn't be any systemd-udevd workers at all. It's as if something didn't get completed or got stuck already before starting to shutdown.

[xnox]

I guess i should be abel to do that via a checkout of the machine with testflinger thing.