devfs
[Top] [All Lists]

Re: Shutdown problems.

To: devfs@xxxxxxxxxxx
Subject: Re: Shutdown problems.
From: Juan Pablo Hierro Álvarez <BUSCAIDEAS@xxxxxxxx>
Date: Sat, 26 Jan 2002 17:21:36 +0100
In-reply-to: <200201140159.g0E1xdK11713@xxxxxxxxxxxxxxxxxxxxxxxx>
References: <20020114005121.A1034@xxxxxxxxxxxxxxxx> <200201140159.g0E1xdK11713@xxxxxxxxxxxxxxxxxxxxxxxx>
Sender: owner-devfs@xxxxxxxxxxx
El Lun 14 Ene 2002 02:59, escribió:
> > It works basically right and now /dev looks much prettier and cleaner on
> > my system. Just a couple of questions:
> >
> > 1) When loading, there is an error in "Copying /dev-state/fb to /dev/fb".
> > However, then it works fine. Maybe, is there a need to build a tarfile
> > for that devices?
>
> No, no tarfiles! Please send the complete error message and any other
> information that you might think will help identify the problem.
Here is the complete error message:
____________________________________
Started device management daemon for /dev
error copying "/dev-state/fb" to "/dev/fb"
____________________________________
Looking /dev, I find
ls -l /dev | grep fb ==> 
drwxr-xr-x    1 root     root            0 ene  1  1970 fb

While in /dev-state, I have
ls -l /dev-state | grep fb
lrwxrwxrwx    1 root     root            3 ene 10 03:16 fb -> fb0
crw-------    1 root     root      29,   0 ago 30 22:30 fb0
and so on, upto fb31

Now, I think that, maybe, in /dev-state I have the old /dev from my fresh 
RH7.2 system while in /dev I have the devices used by my new, hand-compiled 
kernel 2.2.17. Of course, I did not compile it with the same options. In 
fact, it has no framebuffer support.
>
> > 2) The most serious comes when shuting down. I get the following messages
> > almost at the end (while it is unmounting everything)
> > ///////////////////////////////////////////
> > umount2: Device or resource busy
>
> You can't unmount devfs, because it is always being used. A newer
> version of umount(8) (from util-linux) fixes the "umount -a" case.
Problem almost solved, I cannot umount devfs but the reason it could not 
shutdown properly was lack of either apm or acpi support.


<Prev in Thread] Current Thread [Next in Thread>