There is is shell error here is diff to fix it: --- /usr/lib64/pm-utils/sleep.d/56atd 2011-11-14 13:09:53.000000000 +0300 +++ 56atd 2011-11-15 23:46:53.484899900 +0300 @@ -7,9 +7,10 @@ hibernate|suspend) ;; thaw|resume) - if [ -f /etc/init.d/atd ]; then + if [ -f /etc/init.d/atd ] + then /etc/init.d/atd restart - elif + else systemctl try-restart atd.service fi ;;
I just reported this same issue to the test list, having rediscovered it. Jon.
What is this script even meant to be doing? Why does at want to be restarted after suspend?
(In reply to comment #2) > What is this script even meant to be doing? Why does at want to be restarted > after suspend? The script is based on bug #464393. It should be condrestart for sysvinit, but that doesn't matter now, because upstart can be hardly used in Fedora.
atd really needs to handle this case itself without relying on being restarted. The easiest approach is probably to get rid of the sleeps and use POSIX timers instead.
*** This bug has been marked as a duplicate of bug 754156 ***
Much more to the point, it ought to be impossible for an update breaking suspend to land in even rawhide. There should be a process in place to catch that and prevent the package going out.
http://fedoraproject.org/wiki/AutoQA