[arch-general] Archiso, archiso_early, and udev issues with live cd

[Tom medhurst]

/lib/udev/rules.d/ contains 60-cdrom_id.rules which looks for cdroms and
makes them /dev/sr#. So I'm guessing it's not seeing the disk that the
initrd is loaded from?



On Fri, Apr 8, 2011 at 12:21 PM, Tom medhurst <tom.medhurst at gmail.com>wrote:

> Thanks Heiko!
>
> That bug contains the same error message, but not sure it's for the same
> reason. /dev/sr0 doesn't appear either for me.
> For clarity, is /dev/disk/by-label/MY_LABEL supposed to point to my
> installation CD, so isomounts can be loaded? Is that what this is trying to
> do?
>
> I did get a ramfs prompt, but "udevadm trigger" does not seem to affect
> /dev for me :(
>
> -tommed
>
>
> On Fri, Apr 8, 2011 at 12:00 PM, Heiko Baums <lists at baums-on-web.de>wrote:
>
>> Am Fri, 8 Apr 2011 11:42:25 +0100
>> schrieb Tom medhurst <tom.medhurst at gmail.com>:
>>
>> > I'm attempting to create my own Arch Live CD and I can't get the
>> > archiso hook to complete successfully.
>> >
>> > In March 2010, Gerardo Exequiel Pozzi removed the archiso_early hook
>> > from archiso.git, which used to mount /dev/archiso (I'm not entirely
>> > sure what this pointed to, but it's clearly required). In the commit
>> > message he said it was removed because we could rely
>> > on /dev/disk/by-label to get the device (created by udev).
>> >
>> > However I have built a custom live CD using mkarchiso and there is no
>> > /dev/disk available to me, so this boot process fails.
>> >
>> > I was just wondering if anyone knew why this is missing, and what I
>> > can do to get this booting?
>> >
>> > My mkinitcpio.conf file looks like this:
>> >
>> > *HOOKS="base udev memdisk archiso archiso_pxe_nbd archiso_loop_mnt
>> > pata scsi sata usb fw pcmcia filesystems usbinput"*
>>
>> Is it this bug?
>> https://bugs.archlinux.org/task/17231
>>
>> If you get the ramfs$ prompt you can as a workaround run `udevadm
>> trigger` wait a few seconds and then enter `exit`.
>>
>> Heiko
>>
>
>