Description of problem:
crond refuses to load any user's crontabs from /var/spool/cron directory, and complains about the SELinux contexts.

Feb 09 20:33:33 saurok systemd[1]: Starting Command Scheduler...
Feb 09 20:33:33 saurok crond[30971]: (CRON) INFO (RANDOM_DELAY will be scaled with factor 64% if used.)
Feb 09 20:33:33 saurok crond[30971]: (mirror) Unauthorized SELinux context=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 file_context=unconfined_u:object_r:user_cron_spool_t:s0 (/var/spool/cron/mirror)
Feb 09 20:33:33 saurok crond[30971]: (mirror) FAILED (loading cron table)
Feb 09 20:33:33 saurok crond[30971]: (root) Unauthorized SELinux context=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023 file_context=unconfined_u:object_r:user_cron_spool_t:s0 (/var/spool/cron/root)
Feb 09 20:33:33 saurok crond[30971]: (root) FAILED (loading cron table)

crond seems to be running with the right context:

# ps axZ | grep crond
system_u:system_r:crond_t:s0-s0:c0.c1023 334 ? Ss     0:00 /usr/sbin/crond -n

The files themselves seem to have the right contexts:

# ls -aZ /var/spool/cron
    system_u:object_r:user_cron_spool_t:s0 .
          system_u:object_r:var_spool_t:s0 ..
unconfined_u:object_r:user_cron_spool_t:s0 mirror
unconfined_u:object_r:user_cron_spool_t:s0 root

Attempting to use restorecon had no effect:

# restorecon -r -v /var/spool/cron
restorecon:  Warning no default label for /var/spool/cron/mirror
restorecon:  Warning no default label for /var/spool/cron/root

No AVCs were logged.


Version-Release number of selected component (if applicable):
cronie-1.5.0-3.fc23.x86_64
selinux-policy-targeted-3.13.1-158.4.fc23.noarch


How reproducible:
Always


Steps to Reproduce:
1. crontab -e
2. systemctl restart crond


Actual results:
cron jobs fail to run; journal entries complain about Unauthorized SELinux context.


Expected results:
cron jobs run normally.


Additional info:
This appears to have begun immediately after upgrading from F22 to F23 a couple of weeks ago. No user cron jobs have run since the upgrade took place; only jobs in /etc/cron.* are being run.