Description of problem: If one changes one's mind and interrupts a livecd-creator run it stops but leaves behind a mounted install_root. How reproducible: every time Steps to Reproduce: 1. run livecd-creator 2. Press Ctrl-c Actual results: # df | grep loop /dev/loop0 4128448 140664 3945844 4% /var/tmp/imgcreate-rjmEfO/install_root Expected results: loop mount to be unmounted before exit Additional info: The worse thing is that this forces the need to power-cycle when rebooting since usually the machine is not able to shutdown because of the hanging mountpoint. This has been happening for a long time.
The signal is caught deep in rpm so there's little we can do. We have a general catch-all exception handler that would be getting used if it could to clean up
Ok, I see - any hints about how to unmount those loop devices by hand afterwards? They always say they are busy...
If all the processes are dead, then they should unmount and detach the loopdev. The cleanup script I wrote at one point (before the catchall handler ;) just does something like for d in $(cat /proc/mounts |grep imgcreate |awk {'print $2;'}) ; do umount -fl $d 2>/dev/null ; done a few times
Re comment #1: The signal is caught by rpm and mapped into an exported symbol that can be tested whenever you choose.
This bug appears to have been reported against 'rawhide' during the Fedora 11 development cycle. Changing version to '11'. More information and reason for this action is here: http://fedoraproject.org/wiki/BugZappers/HouseKeeping
So back to livecd-tools?
Back to livecd-tools. It's possible to poll for signals caught by rpm with rpm.signalsCaught(). Alternatively you can just set your own SIGINT handler before rpm does, rpm wont touch handlers which are set to non-default. When operating on a real, live system overriding rpm's signal handler is not a very good idea but as livecd-tools is just creating a system image into a chroot, rpmdb consistency when interrupted is hardly an issue.
This package has changed ownership in the Fedora Package Database. Reassigning to the new owner of this component.
In reply to comment 4 I have done some testing, and since the rpm.checkSignals() happens after every package, after say hitting ctrl-C in miscutils.py sys.exit(1) happens. How are you suggesting to use rpm.signalsCaught() or rpm.checkSignals() to gracefully cleanup in livecd-creator if it exits? So far the only way I have been able to avoid this RPMN behaviour is by initializing my own signal handler, before RPM does. Do you have a better suggestion?
livecd-tools-034-1.fc14 has been submitted as an update for Fedora 14. https://admin.fedoraproject.org/updates/livecd-tools-034-1.fc14
livecd-tools-034-2.fc14 has been pushed to the Fedora 14 testing repository. If problems still persist, please make note of it in this bug report. If you want to test the update, you can install it with su -c 'yum --enablerepo=updates-testing update livecd-tools'. You can provide feedback for this update here: https://admin.fedoraproject.org/updates/livecd-tools-034-2.fc14
livecd-tools-034-7.fc14 has been submitted as an update for Fedora 14. https://admin.fedoraproject.org/updates/livecd-tools-034-7.fc14
livecd-tools-034-7.fc14 has been pushed to the Fedora 14 testing repository. If problems still persist, please make note of it in this bug report. If you want to test the update, you can install it with su -c 'yum --enablerepo=updates-testing update livecd-tools'. You can provide feedback for this update here: https://admin.fedoraproject.org/updates/livecd-tools-034-7.fc14
livecd-tools-034-7.fc14 has been pushed to the Fedora 14 stable repository. If problems still persist, please make note of it in this bug report.