[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [MiNT] OpenSSh port



Hi!

On Mon, Dec 11, 2000 at 03:51:34PM +0200, you wrote:
> First, many thanks to Thomas Binder for porting it:  this is the
> most stable and best usable SSH port to ever hit our platform.

You're welcome. I hope you've noticed that ssh-client fixes the cursor
key problem earlier ports have had :)

> SSHD:
> 1) fails to display MOTD when PrintMotd is enabled, however the undocumented
>    option KbdInteractiveAuthentication fixes it randomly.  This problem also
>    existed in Juhani's SSH 1.2.26 port, but the above undocumented option
>    did not exist in that SSH release, nor in earlier OpenSSH versions.

I've not tried that yet, will do so if I find the time.

> 2) fails to use Login; exits if enabled. Error reported on the remote end:
> <paska:q-funk>/home/q-funk$ ssh rakas
> login: No such file or directory.
> Connection to rakas closed.

As there's no login program for Sparemint yet, sshd was compiled without
support for it.

> 3) produces the following error messages, after authentication is accepted:
> Dec  7 21:21:41 rakas sshd[70]: Accepted rsa for q-funk from 192.168.1.1 port 1029

That's not an error :)

> Dec  7 21:21:41 rakas sshd[72]: error: Failed to disconnect from controlling tty.

I've not yet been able to track down the reason for that message. But as
it doesn't affect the connection, I don't think it's urgent. Just ignore
the message.

> Dec  7 21:22:39 rakas sshd[70]: syslogin_perform_logout: logout() returned an error

I've never seen that before. Is that with /bin/login enabled?

> 4) restarting sshd produces the following error:
> Dec  9 13:08:28 rakas sshd[95]: Received SIGHUP; restarting.
> Dec  9 13:08:28 rakas sshd[95]: RESTART FAILED: av0='sshd', error: No such file or directory.

I have to look at this. Never actually tried to SIGHUP sshd yet.

> SSH-KEYGEN:
> 1) needed two (2) hours to generate a default 1024 bits DSA key, while
>    ssh 1.2.26's ssh-keygen succeeded in about 10 minutes (on a standard TT).

Well, you know that DSA and RSA (as 1.2.26 only supported) are two very
different schemes? To comfort you: It takes approx. ten times longer
to create DSA keys, which is even noticable on a PentiumIII/800.

> 2) in any case, the RPM's installation script did not even attempt to verify
>    if one already existed, nor to create a new one if none was found.

Correct. This is done by /etc/init.d/sshd - it will create the keys, if
necessary. But Sparemint's init not yet provides the "functions" script
needed by the init-script, so you can't yet run it without modification
- but I decided to include it as it is, as it's likely that "functions"
will be provided some day.

> Btw, regarding the installation warning: I thought that the initial
> SpareMiNT GCC "specs" had been fixed later, so that m68020-60 no
> longer implied an FPU, but only the CPU architecture as expected?

Nothing has/had to be "fixed" here. The official docs of gcc say that
this option will automatically enable FPU usage. Period.

-- snip --
`-m68020-40'
     Generate output for a 68040, without using any of the new
     instructions.  This results in code which can run relatively
     efficiently on either a 68020/68881 or a 68030 or a 68040.  The
     generated code does use the 68881 instructions that are emulated
     on the 68040.

`-m68020-60'
     Generate output for a 68060, without using any of the new
     instructions.  This results in code which can run relatively
     efficiently on either a 68020/68881 or a 68030 or a 68040.  The
     generated code does use the 68881 instructions that are emulated
     on the 68060.
-- snap --


Ciao

Thomas


-- 
Thomas Binder (Gryf @ IRCNet)  gryf@hrzpub.tu-darmstadt.de
PGP-key available on request!  binder@rbg.informatik.tu-darmstadt.de
Vote against SPAM:             http://www.politik-digital.de/spam/

Attachment: pgpiSBNjd8C0g.pgp
Description: PGP signature