ts-7000
[Top] [All Lists]

Re: [ts-7000] udev problems with debian-sarge-udev-11-15-2007

To:
Subject: Re: [ts-7000] udev problems with debian-sarge-udev-11-15-2007
From: Jeff Cunningham <>
Date: Wed, 26 Dec 2007 16:29:34 -0800
Hi Triffid,

Thanks for the info. I don't have a video card so presumably I need no 
virtual consoles.
But how do I change /etc/inittab to refer terminals that actually exist? 
I'm looking at it and it is currently set up like this:

1:2345:respawn:/sbin/getty 38400 tty1
T0:23:respawn:/sbin/getty -L ttyAM0 115200 vt100
T1:23:respawn:/sbin/getty -L ttyAM1 115200 vt100

Doesn't that mean it should be generating /dev/tty1?

I don't think I'm running dropbear ssh. I installed the latest stable 
version with apt-get when running debian:

:dev# ssh -v
OpenSSH_4.3p2 Debian-9, OpenSSL 0.9.8c 05 Sep 2006

But for all I know it has the same requirement as dropbear. So how do I 
mount a tmpfs over /dev at boot before starting udev? I know just enough 
about udev to be dangerous.

--Jeff



Triffid Hunter wrote:
>
> Hi Jeff,
>
> /dev/tty{0..63} are virtual consoles that exist on the primary display
> device. Does your embedded arm have a video card? If not, there ARE no
> virtual consoles, and udev does not assign nodes for things that don't
> exist. I suggest you change /etc/inittab to reference terminals that
> actually exist on the device, or change startup scripts to create 
> sensible
> symlinks.
>
> ssh/dropbear/telnet/etc require /dev/pts to have a devpts filesystem
> mounted on it. If you mount a tmpfs over /dev at boot before starting
> udev, you'll have to add mkdir /dev/pts; mount -t devpts none /dev/pts to
> your bootscripts before udev. This special filesystem allows creation of
> almost unlimited pseudo TTYs (pty/pts) without cluttering /dev up with
> unused nodes. The ts-7800 I recieved recently failed to mount /dev/pts
> during first stage boot, preventing telnet from working, so this is a new
> one to techonolic too.
>
> Note that only /dev/console needs to exist in order to boot, so that's
> the only node you need in there if you mount a tmpfs over /dev and let
> udev handle things. It's wise to include /dev/null too, in case you
> redirect stuff to it before starting udev.
>
> 2.6 kernels have a number of differences to 2.4 ones, most of which are
> designed to make things easier as long as you know about them ;)
>
> On Wed, 26 Dec 2007, jcunningham63 wrote:
>
> > I have two 2.6 kernels working, the 2.6.21 off the TS website, and Dan
> > Smolik's 2.6.23_rc1. In both cases I am mounting the root file system
> > I downloaded from the TS website: debian-sarge-udev-11-15-2007
> >
> > I think there's something broken about the udev setup in this
> > filesystem and I don't understand udev well enough to fix it. The
> > symptoms are two-fold. First, every five minutes the console prints:
> >
> > INIT: Id "1" respawning too fast: disabled for 5 minutes
> >
> > which if you google around seems to result from /dev/tty1 being messed
> > up. When I look in /dev I don't see it, but only:
> >
> > crw-rw-rw- 1 root root 5, 0 Dec 26 13:46 tty
> > crw------- 1 root tty 204, 16 Dec 26 13:46 ttyAM0
> > crw------- 1 root root 204, 17 Dec 26 13:46 ttyAM1
> > crw-rw---- 1 root dialout 204, 18 Dec 26 13:46 ttyAM2
> >
> > I can't use makedev to create it, because this is udev and its
> > supposed to make it dynamically when needed. But I actually tried
> > using makedev and what it says is that udev is operating, so it
> > creates it in /dev/.static/dev/tty1 . I subquently made a softlink
> > from /dev/tty1 to the static one, but it gives an error that the tty1
> > device node is dysfunctional.
> >
> > The second symptom is that SSH is broken. when I try to connect via
> > SSH from another box it hangs. The auth.log shows a similar device
> > problem:
> >
> > Dec 26 13:47:34 localhost sshd[621]: debug1: server_input_channel_req:
> > channel 0 request pty-req reply 0
> > Dec 26 13:47:34 localhost sshd[621]: debug1: session_by_channel:
> > session 0 channel 0
> > Dec 26 13:47:34 localhost sshd[621]: debug1:
> > session_input_channel_req: session 0 req pty-req
> > Dec 26 13:47:34 localhost sshd[621]: debug1: Allocating pty.
> > Dec 26 13:47:34 localhost sshd[621]: error: openpty: No such file or
> > directory
> > Dec 26 13:47:34 localhost sshd[621]: error: session_pty_req: session 0
> > alloc failed
> > Dec 26 13:47:34 localhost sshd[621]: debug1: server_input_channel_req:
> > channel 0 request shell reply 0
> > Dec 26 13:47:34 localhost sshd[621]: debug1: session_by_channel:
> > session 0 channel 0
> > Dec 26 13:47:34 localhost sshd[621]: debug1:
> > session_input_channel_req: session 0 req shell
> > Dec 26 13:47:34 localhost sshd[625]: debug1: permanently_set_uid:
> > 1000/1000
> >
> >
> > Googling for these errors turns up lots of stuff that again implicates
> > udev - looks like SSH wants to use /dev/pty0 which also doesn't exist.
> >
> > So, does anyone have any idea how to fix udev?
> > --Jeff
>
> 



 
Yahoo! Groups Links

<*> To visit your group on the web, go to:
    http://groups.yahoo.com/group/ts-7000/

<*> Your email settings:
    Individual Email | Traditional

<*> To change settings online go to:
    http://groups.yahoo.com/group/ts-7000/join
    (Yahoo! ID required)

<*> To change settings via email:
     
    

<*> To unsubscribe from this group, send an email to:
    

<*> Your use of Yahoo! Groups is subject to:
    http://docs.yahoo.com/info/terms/
 

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

Disclaimer: Neither Andrew Taylor nor the University of NSW School of Computer and Engineering take any responsibility for the contents of this archive. It is purely a compilation of material sent by many people to the birding-aus mailing list. It has not been checked for accuracy nor its content verified in any way. If you wish to get material removed from the archive or have other queries about the archive e-mail Andrew Taylor at this address: andrewt@cse.unsw.EDU.AU