Re: /dev/console switching: the continuing saga

[David Walter]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Marcus Brinkmann <Marcus.Brinkmann@ruhr-uni-bochum.de> writes:

> (btw, marcus@debian.org is not one of my addresses)

Apologies, don't know what that was from %-|

> On Wed, Nov 20, 2002 at 09:23:37AM -0500, David Walter wrote:
>> I've been noticing that if the initial term process running on console
>> isn't killed console output doesn't change to the new device.
>> 
>> If process 7 is killed (/hurd/term /dev/console ...) then there is a
>> lost resource.

A word of clarification.  I was killing the initial term process to be
able get output, not because I thought it was the right thing(tm).

More below.

> The purpose of the fsysopts call is exactly to change the options of the
> running term process, and thus redirect all live users of the term node.

I am using:

  fsysopts /dev/console --type=hurdio /dev/vcs/10/console

> It's completely illogical that killing the running term process helps in
> bringing a change to effect that is initiated with fsysopts.  Are you really
> using fsysopts to redirect the term process, or are you by any chance
> running settrans?  Please list the exact sequence of commands you are
> running, it's not clear from your message what you were doing.

Seems strange to do that, and I did't give enough explanation.

If I perform the following sequence:

  console -d vga -d pc_kbd /dev/vcs   

Then in one of the vt's.

   $ fsysopts /dev/console --type=hurdio /dev/vcs/10/console

Now, I have   a  passive translator for   a dirty  partition  on /opt,
starting the translator will cause an error to the console.

   $ ls /opt

No error output.

   $ settrans -fga /opt

   $ ls /opt

No error output and ls hangs.

   ^C
   $ fsysopts /dev/console --type=hurdio /dev/vcs/10/console
   $ ls /opt

Now on vt10 we see:

ext2fs: /dev/hd0s13: warning: FILESYSTEM NOT UNMOUNTED CLEANLY; PLEASE fsck
ext2fs: /dev/hd0s13: warning: MOUNTED READ-ONLY; MUST USE `fsysopts --writable'

This was working with  killing 7 and then  doing fsysopts, but you are
completely right that this doesn't seem sensible.

The reason that I had started killing pid 7 this was that fsysopts had
been hanging, and  I hadn't disassociated  the problem of the argument
copy from this other problem, which is still showing up.

At this  point  I am trying  to find  out if  there is any  difference
between being  started initially and switched, as  it seems that there
is some initialization being lost.

> The problem that some programs open the device directly is unrelated to
> that.

Right.

Thanks.

- -- 
pub  1024D/DC92AE30 2002-02-26 David Walter <dwalter@syr.edu>
fingerprint = 50A0 E513 732D 1D0F BD26  C84E A8DD 9D80 DC92 AE30
sub  2048g/51023582 2002-02-26
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.6 (GNU/Hurd)
Comment: Processed by Mailcrypt 3.5.6 and Gnu Privacy Guard 
<http://www.gnupg.org/>

iD8DBQE92+UoqN2dgNySrjARAjvKAJ9eBvlIt1ITVzyZ9V6p2U7nWJMsjwCeO3iC
4Oh3HU+8Be2J1ieIyw1CXns=
=Bw9k
-----END PGP SIGNATURE-----