New Beta Bootscript Package 20040601

Ian Molton spyro at f2s.com
Wed Jun 2 06:35:42 PDT 2004


On Wed, 2 Jun 2004 07:25:47 -0600 (MDT)
"Nathan Coulson" <conathan at conet.dyndns.org> wrote:

> So hotplug calls scripts which do everything but create the device, and
> then call udev which creates the device...

was thinking about this... wouldnt it be better solved in userspace by patching hotplug so that it wont call udev until such time as the subsystem is running?

perhaps it should test for a file in the /dev directory? this should be safe as it will be a ramfs by the time udev is up, so there is no risk of it being left stale on the harddisc, especially if udev can be patched to not create the file if it finds itself not in the 'ramfs /dev'.

What do you think? better than patching the kernel, surely?



More information about the lfs-dev mailing list