openssh unix dev - Jan 2019 - [Regression] OpenSSH 7.7p1 no longer tests on NonStop Port

On January 22, 2019 10:45, I wrote:
> On January 21, 2019 19:00, Darren Tucker wrote:
> > On Tue, 22 Jan 2019 at 12:01, Randall S. Becker
> > <rsbecker at nexbridge.com>
> > wrote:
> >
> > > Hi All,
> > >
> > > I finally got around to trying to test the 7.7p1 release on the
HPE
> > > NonStop Platform. 7.6p1 worked just fine - no issues at all.
> > >
> > > In 7.7p1, when the agent test ran, the following happened (prior
> > > tests worked perfectly):
> > >
> > >         run test agent.sh ...
> > >         Couldn't open logfile
/home/git/openssh-portable/regress/ssh.log:
> > > Permission denied
> > >         agent fwd failed (exit code 1)
> > >         failed simple agent test
> > >
> >
> > My guess is that this is due to a previous test creating ssh.log as
> > root, however offhand I can't think of any changes in that time
frame
> > that would cause it.
> >
> > Do you see the problem if you remove the log file and run the test in
> > isolation?
> >
> > $ SUDO=sudo make tests LTESTS=agent SKIP_UNIT=1
> 
> <snip>
> run test agent.sh ...
> ok simple agent test
> make[1]: Leaving directory '/home/git/openssh-portable/regress'
> all tests passed
> 
> So that's a good thing. Operator error. When running the full test:
> 
> key option expiry-time valid
> failed key options
> Makefile:206: recipe for target 't-exec' failed
> 
> /home/git/openssh-portable/regress: cat failed-regress.log, failed-ssh.log
#
> are the same
> trace: key option pty default
> FAIL: key option failed default
> 
> trace: key option pty restrict,pty
> FAIL: key option failed restrict,pty
> 
> /home/git/openssh-portable/regress: tail -7 failed-sshd.log
> debug3: do_exec_no_pty: fork() in child
> debug3: send packet: type 99
> debug3: do_child: entering do_child()
> debug1: permanently_set_uid: 25603/100
> FAIL: key option failed restrict,pty
> debug3: do_child: Shell path=/bin/ksh
> 
> So, better, but this is a new break for us.
> 
> >
> > [...]
> > > We have a rather large set of changes
> >
> > are these changes publicly available?
> 
> My GitHub fork for this is https://github.com/rsbeckerca/openssh-portable
> on the Randall branch, which is work in progress. The 7.6p1 port changes
are
> at 2abb363, representing the bulk of the work. 7.7p1 is from there upward
to
> 814c8f2.
Here's where we are in debugging. I've set +x and fail in test-exec.sh,
here's the log. I am having difficulty interpreting the output or where to
look next. To the uninitiated, this does not look bad. 

+ SSH_KEYTYPES=rsa ed25519
+ trace generate keys
+ [ ! -f /home/git/openssh-portable/regress/rsa ]
+ [ /home/git/openssh-portable/ssh-keygen -nt
/home/git/openssh-portable/regress/rsa ]
+ 1>> /home/git/openssh-portable/regress/known_hosts
+ printf localhost-with-alias,127.0.0.1,::1
+ cat /home/git/openssh-portable/regress/rsa.pub
+ cat /home/git/openssh-portable/regress/rsa.pub
+ 1>> /home/git/openssh-portable/regress/authorized_keys_randall
+ echo IdentityFile /home/git/openssh-portable/regress/rsa
+ 1>> /home/git/openssh-portable/regress/ssh_config
+ sudo cp /home/git/openssh-portable/regress/rsa
/home/git/openssh-portable/regress/host.rsa
+ echo HostKey /home/git/openssh-portable/regress/host.rsa
+ 1>> /home/git/openssh-portable/regress/sshd_config
+ echo HostKey /home/git/openssh-portable/regress/rsa
+ 1>> /home/git/openssh-portable/regress/sshd_proxy
+ [ ! -f /home/git/openssh-portable/regress/ed25519 ]
+ [ /home/git/openssh-portable/ssh-keygen -nt
/home/git/openssh-portable/regress/ed25519 ]
+ 1>> /home/git/openssh-portable/regress/known_hosts
+ printf localhost-with-alias,127.0.0.1,::1
+ cat /home/git/openssh-portable/regress/ed25519.pub
+ cat /home/git/openssh-portable/regress/ed25519.pub
+ 1>> /home/git/openssh-portable/regress/authorized_keys_randall
+ echo IdentityFile /home/git/openssh-portable/regress/ed25519
+ 1>> /home/git/openssh-portable/regress/ssh_config
+ sudo cp /home/git/openssh-portable/regress/ed25519
/home/git/openssh-portable/regress/host.ed25519
+ echo HostKey /home/git/openssh-portable/regress/host.ed25519
+ 1>> /home/git/openssh-portable/regress/sshd_config
+ echo HostKey /home/git/openssh-portable/regress/ed25519
+ 1>> /home/git/openssh-portable/regress/sshd_proxy
+ chmod 644 /home/git/openssh-portable/regress/authorized_keys_randall
+ REGRESS_INTEROP_CONCH=no
+ test -x
+ REGRESS_INTEROP_PUTTY=no
+ test -x  -a -x
+ REGRESS_INTEROP_PUTTY=no
+ test no = yes
+ 1> /home/git/openssh-portable/regress/ssh_proxy
+ cat /home/git/openssh-portable/regress/ssh_config
+ echo proxycommand sudo sh
/home/git/openssh-portable/regress/sshd-log-wrapper.sh
/home/git/openssh-portable/regress/sshd.log /home/git/openssh-portable/sshd
-i -f /home/git/openssh-portable/regress/sshd_proxy
+ /home/git/openssh-portable/sshd -t -f
/home/git/openssh-portable/regress/sshd_proxy
+ . /home/git/openssh-portable/regress/key-options.sh
+ tid=key options
+ origkeys=/home/git/openssh-portable/regress/authkeys_orig
+ authkeys=/home/git/openssh-portable/regress/authorized_keys_randall
+ cp /home/git/openssh-portable/regress/authorized_keys_randall
/home/git/openssh-portable/regress/authkeys_orig
+ sed s/.*/command="echo bar" &/
/home/git/openssh-portable/regress/authkeys_orig
+ 1> /home/git/openssh-portable/regress/authorized_keys_randall
+ verbose key option command="echo bar"
key option command="echo bar"
+ + /home/git/openssh-portable/regress/ssh-log-wrapper.sh -q -F
/home/git/openssh-portable/regress/ssh_proxy somehost echo foo
r=bar
+ [ bar = foo ]
+ [ bar != bar ]
+ sed s/.*/no-pty,command="echo bar" &/
/home/git/openssh-portable/regress/authkeys_orig
+ 1> /home/git/openssh-portable/regress/authorized_keys_randall
+ verbose key option no-pty,command="echo bar"
key option no-pty,command="echo bar"
+ + /home/git/openssh-portable/regress/ssh-log-wrapper.sh -q -F
/home/git/openssh-portable/regress/ssh_proxy somehost echo foo
r=bar
+ [ bar = foo ]
+ [ bar != bar ]
+ expect_pty_succeed default
key option pty default
key option failed default
Makefile:206: recipe for target 't-exec' failed
make[1]: *** [t-exec] Error 1

On Fri, 25 Jan 2019 at 05:23, Randall S. Becker <rsbecker at
nexbridge.com> wrote:
> Here's where we are in debugging. I've set +x and fail in
test-exec.sh,
> here's the log. I am having difficulty interpreting the output or where
to
> look next. To the uninitiated, this does not look bad.
If you add an "exit 1" into the fail() function call in test-exec.sh
it should leave all of the keys in place so that you can copy and
paste the failing command into your shell.
> + + /home/git/openssh-portable/regress/ssh-log-wrapper.sh -q -F
> /home/git/openssh-portable/regress/ssh_proxy somehost echo foo
Replace ssh-log-wrapper.sh  with the path to your newly built ssh
binary and replace the -q with -vvv.

Failing that, I'd try starting an "sshd -ddd" in one window and
the
ssh -vvv in another window and set the same key options that the
failing test has.

-- 
Darren Tucker (dtucker at dtucker.net)
GPG key 11EAA6FA / A86E 3E07 5B19 5880 E860  37F4 9357 ECEF 11EA A6FA (new)
    Good judgement comes with experience. Unfortunately, the experience
usually comes from bad judgement.

On January 24, 2019 12:11, Darren Tucker wrote:
> On Fri, 25 Jan 2019 at 05:23, Randall S. Becker <rsbecker at
nexbridge.com>
> wrote:
> > Here's where we are in debugging. I've set +x and fail in
> > test-exec.sh, here's the log. I am having difficulty interpreting
the
> > output or where to look next. To the uninitiated, this does not look
bad.
> 
> If you add an "exit 1" into the fail() function call in
test-exec.sh it
should
> leave all of the keys in place so that you can copy and paste the failing
> command into your shell.
> 
> > + + /home/git/openssh-portable/regress/ssh-log-wrapper.sh -q -F
> > /home/git/openssh-portable/regress/ssh_proxy somehost echo foo
> 
> Replace ssh-log-wrapper.sh  with the path to your newly built ssh binary
and
> replace the -q with -vvv.
> 
> Failing that, I'd try starting an "sshd -ddd" in one window
and the ssh
-vvv in
> another window and set the same key options that the failing test has.
Is there any way to bypass this one failing test and carry blindly forward?
I think this key has had issues in the past that we just did not see.

Thanks,
Randall