> yes run-init is the standard way to nuke your initramfs and > the succesor of pivot_root() it is heavily audited used on > rh, suse, debian, ubuntu, .. > no point in keeping rootfs, nuking is very quick.
I agree with run-init is the standard way to nuke an initramfs. According to the kernel documentation this is *not* required. That is why my testcase is not doing it. Furthermore you're totally wrong with pointing to pivot_root. With "normal" (Debian) initramfs pivot_root is never called. Instead run-init deletes all files from rootfs and then mounts the real root device over the rootfs in order to execute init after that. However I don't have a block device. So the system just lives inside the initramfs. > see yaird for compiling it against glibc. I don't know what you mean. Please explain this in further detail. > > In /etc/init.d/udev line 27, please insert > > > > test $(grep '^[^ ]* / ' /proc/mounts | wc -l) -eq 1 && no_static_dev=y > sorry but that is not a patch, please post a real unified diff, > thanks. http://www.debian.org/Bugs/Developer: | patch | A patch or some other easy procedure for fixing the bug is | included in the bug logs. If there's a patch, but it doesn't | resolve the bug adequately or causes some other problems, this tag | should not be used. Do I also need to file a bug against bugs.debian.org? Helmut -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
