tags 472579 patch thanks > sooo urgss you are not using run-init!?
I am not even using klibc. Am I required to use run-init? Is there any benefit from doing so when I do not want to mount another block device on /? The test case I have given makes the kernel directly execute init like it was before we had initramfs or initrd. Is this deprecated or in some other way doomed? I don't really see why I should not be doing this. The document I quoted in my previous mail also implicitly states that it is *not* required to mount a block device over /. As this looks really complicated let me propose a direct fix for this problem, so we can forget about it: In /etc/init.d/udev line 27, please insert test $(grep '^[^ ]* / ' /proc/mounts | wc -l) -eq 1 && no_static_dev=y Explanation: The grep will spit out all mounts on /. As rootfs is always mounted we cannot grep for that, so we count the things mounted on /. If it is only rootfs there is only one line and we set no_static_dev in order to avoid the failing mounts. In all other cases there will be at least two mounts, so the line does nothing. Helmut -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
