On 11/07/2012 06:31 PM, Jon wrote: > One of the ill effects is that NetworkManager.service does not > work during the boot sequence. Afterward NM works with a restart, so > perhaps some context is wrong, or some race condition. Jon Masters was > speculating that perhaps the way systemd starts things with the wrong > context on the service, or something similar. Off hand there's nothing recently in the selinux-policy for mmap_zero and that's pretty basic so I assumed something changed wrt the contexts for the affected services. They seemed ok. Note that we are not using F18 Anaconda to compose images at the moment. Instead, the alphas were made using F17 tooling. Therefore, we know it's likely our "fault" but we'd still like help tracking this down. A simple relabel apparently doesn't fix things, though I've not poked directly yet. I spent only 5 min looking so far, but happy to help. Jon. _______________________________________________ arm mailing list arm@xxxxxxxxxxxxxxxxxxxxxxx https://admin.fedoraproject.org/mailman/listinfo/arm
