Barry barry at
Tue Jan 2 11:56:17 PST 2001

Perrin Aybara wrote:
> Okay, I'm starting to get really frustrated with this whole pseudo-tty
> situation.  Eterm, aterm, and rxvt refuse to start because the
> "pseudo-tty" device doesn't exist.  However, pty* and tty[a-whatever]*
> all exist, and I've made a symbolic link from ptya0 to pseudo-tty (and
> tried doing that with ttya0 too).  All the pty's and tty[a-whatever]'s
> are readable and writable by all (just for the time being trying to get
> this working), and /dev/pts is in /etc/fstab like someone suggested to
> this list earlier.  It's also readable and write-able by all.  What am I
> missing here?

the pty interface has changed.  If you have pty filesystem enabled in
your kernel, the kernel will look for Unix pty devices in /dev/pts

if you built your lfs system based on the book, then the directory
/dev/pts already exists.  if not, mkdir /dev/pts ...

then add the line:

none                    /dev/pts                devpts  gid=4,mode=620 
0 0

that should mount the /dev/pts virtual partition so that the kernel will
use that...the method of using pty's that you've listed above is
actually old.  The method I've listed will work only if Unix
pseudo-terminals is enabled in the kernel..

see the kernel docs for more info, particularly Documentation/Changes

Unsubscribe: send email to lfs-apps-request at
and put unsubscribe in the subject header of the message

More information about the blfs-support mailing list