OpenSSH 3.7 testing (Re: 3.6p1 bug on SCO OpenServer)

Gary E. Miller gem at
Sun Sep 7 04:49:25 EST 2003

Yo Ben!

Looks almost ready to go.

Small problem here with: openssh-SNAP-20030905.tar.gz

I have a bastard slackware 8.0 laptop with lot's of updates.  Now running
gcc 3.3, glibc 2.3.3 and openssl 0.9.6g

Here is the configure command:

./configure --with-default-path=$PATH --with-md5-passwords \
  --with-mantype=man --with-tcp-wrappers \
  --with-ssl-dir=/usr/local/ssl/lib --with-dns

Here is the config result:

OpenSSH has been configured with the following options:
                     User binaries: /usr/local/bin
                   System binaries: /usr/local/sbin
               Configuration files: /usr/local/etc
                   Askpass program: /usr/local/libexec/ssh-askpass
                      Manual pages: /usr/local/man/manX
                          PID file: /var/run
  Privilege separation chroot path: /var/empty
            sshd default user PATH: /usr/local/qt/bin:/usr/local/sbin:/usr/sbin:/sbin:/opt/schily/bin:/usr/local/bin:/usr/bin:/bin:/usr/local/enlightenment/bin:/usr/X11R6/bin:/usr/local/samba/bin:/usr/X11R6/bin:/usr/games:.:/opt/gnome/bin:/opt/www/htdig/bin:/opt/kde/bin:/usr/local/java/bin:/usr/share/texmf/bin:/usr/openwin/bin
                    Manpage format: man
                       DNS support: yes
                       PAM support: no
                 KerberosV support: no
                 Smartcard support: no
                     S/KEY support: no
              TCP Wrappers support: yes
              MD5 password support: yes
       IP address in $DISPLAY hack: no
           Translate v4 in v6 hack: yes
                  BSD Auth support: no
              Random number source: OpenSSL internal ONLY

              Host: i686-pc-linux-gnu
          Compiler: gcc
    Compiler flags: -g -O2 -Wall -Wpointer-arith -Wno-uninitialized
Preprocessor flags: -I/usr/local/ssl/lib
      Linker flags: -L/usr/local/ssl/lib
         Libraries: -lwrap  -lresolv -lutil -lz -lnsl  -lcrypto -lcrypt

Here is the error:

gcc -g -O2 -Wall -Wpointer-arith -Wno-uninitialized -I. -I.. -I. -I./.. -I/usr/local/ssl/lib  -DHAVE_CONFIG_H -c xcrypt.c
xcrypt.c:68:3: invalid preprocessing directive #elsif
xcrypt.c: In function `xcrypt':
xcrypt.c:69: warning: implicit declaration of function `iscomsec'
xcrypt.c:70: warning: implicit declaration of function `bigcrypt'
xcrypt.c:70: warning: assignment makes pointer from integer without a cast
xcrypt.c:73:3: invalid preprocessing directive #elsif
xcrypt.c:74: warning: assignment makes pointer from integer without a cast
make[1]: *** [xcrypt.o] Error 1
make[1]: Leaving directory `/usr/local/src/openssh/openbsd-compat'
make: *** [openbsd-compat/libopenbsd-compat.a] Error 2

Changing the #elsif to #elif on lines 68 and 73 allows the SNAP to build.

make tests passes fine, except for one thing I did not dig out:

run test ...
/usr/local/src/openssh/regress/ [: too many arguments
/usr/local/src/openssh/regress/ [: too many arguments
skipped (no suitable ProxyCommand found)

Gary E. Miller Rellim 20340 Empire Blvd, Suite E-3, Bend, OR 97701
	gem at  Tel:+1(541)382-8588 Fax: +1(541)382-8676

On Sat, 6 Sep 2003, Ben Lindstrom wrote:

> Date: Sat, 6 Sep 2003 00:47:14 -0500 (CDT)
> From: Ben Lindstrom <mouring at>
> To: OpenSSH Development <openssh-unix-dev at>
> Subject: OpenSSH 3.7 testing (Re: 3.6p1 bug on SCO OpenServer)
> On this note I think it would be best to opening the floor for testing of
> the current CVS tree.  OpenSSH is in a feature lock and we should be in
> in sync with the OpenBSD tree (there may be stray patches, but hopefully
> nothing major).
> So please if we can get people to start doing tests on their platform is
> would help.
> The major things that have changed is PAM and Kerb/GSS support.  However,
> we did do a clean up of openbsd-compat/ code so I may have broken
> platforms I don't own by playing in there.  I also redid how we handle
> signal (signal() vs a stricter mysignal() version).  So that is something
> to watch out for (basicly made it consistant).
> There are a bunch more changes, but I don't have a copy of the proposed
> release notes sitting in front of me.
> BTW, regression testing should hopefully work for most platforms.  Please
> get in the habbit of using this.  It should give us a better handle on
> platform breakage.  Darren was nice enough to get that rolling recently.
> It should be as simple as ./configure [needed options] && make && make
> tests, but I've not had a chance to confirm it with an actually test on my
> Solaris test box.
> Sooooo... If you wish your platform to work then drop by
>, pick your favorate ftp server
> or grab the the source from the CVS tree.
> Thanks,
> - Ben && The OpenSSH Portable Team
> On Fri, 5 Sep 2003, Tim Rice wrote:
> > On Sat, 6 Sep 2003, Darren Tucker wrote:
> >
> > > Roger Cornelius wrote:
> > > > loginrec.c writes incorrect data into the ut_id field of the utmp file.
> > > > This has been an issue since at least openssh 3.0.2 but I never bothered
> > > > to report it.  For Openssh 3.6p1, defining WITH_ABBREV_NO_TTY corrects
> > > > the problem.
> > >
> > > [snip]
> > > Are you talking about "*-*-sco3.2v4*" or "*-*-sco3.2v5*" or both?
> >
> > My 3.2v4.2 box is dead. Can someone check the last(C) output here?
> >
> > >
> > > Tim, any issue with doing this?
> >
> > I've verified the complaint. As soon as I can get "current" to build [1]
> > on OpenServer I'll test the fix.
> >
> > 1. ssh-keygen.c uses PATH_MAX now.
> > No source file used PATH_MAX in openssh-3.6.1
> >
> > More to do this weekend. ;-)
> >
> > --
> > Tim Rice				Multitalents	(707) 887-1469
> > tim at
> >
> > _______________________________________________
> > openssh-unix-dev mailing list
> > openssh-unix-dev at
> >
> >
> _______________________________________________
> openssh-unix-dev mailing list
> openssh-unix-dev at

More information about the openssh-unix-dev mailing list