openssh bugs - Jul 2006 - [Bug 1208] ssh fails to remove control socket when using with subversion

http://bugzilla.mindrot.org/show_bug.cgi?id=1208

           Summary: ssh fails to remove control socket when using with
                    subversion
           Product: Portable OpenSSH
           Version: 4.3p2
          Platform: ix86
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: ssh
        AssignedTo: bitbucket at mindrot.org
        ReportedBy: daniel.subs at internode.on.net


Using svn over ssh creates a mastersocket as expected however it
doesn't remove it when completed.

The next svn command finds the socket however no process is attached to
it.




------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

http://bugzilla.mindrot.org/show_bug.cgi?id=1208





------- Comment #1 from daniel.subs at internode.on.net  2006-07-10 13:38
-------
Created an attachment (id=1159)
 --> (http://bugzilla.mindrot.org/attachment.cgi?id=1159&action=view)
debug logs + version info




------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

http://bugzilla.mindrot.org/show_bug.cgi?id=1208





------- Comment #2 from djm at mindrot.org  2006-07-10 13:59 -------
You seem to be using a modified "hpn" SSH client. Can you reproduce
this with an unpatched OpenSSH?




------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

http://bugzilla.mindrot.org/show_bug.cgi?id=1208


daniel.subs at internode.on.net changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
Attachment #1159 is|0                           |1
           obsolete|                            |




------- Comment #3 from daniel.subs at internode.on.net  2006-07-10 14:31
-------
Created an attachment (id=1160)
 --> (http://bugzilla.mindrot.org/attachment.cgi?id=1160&action=view)
debug logs + version info (without hpn patch)

yes it is reproduceable without the hpn patch.




------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

http://bugzilla.mindrot.org/show_bug.cgi?id=1208





------- Comment #4 from djm at mindrot.org  2006-07-10 14:37 -------
it looks like svn might be ungracefully killing its ssh process. Could
you try running it under strace "strace -ffo svn.strace svn up"? 

This should create a couple of svn.strace.$PID files, one of which
should contain ssh information.




------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

http://bugzilla.mindrot.org/show_bug.cgi?id=1208





------- Comment #5 from daniel.subs at internode.on.net  2006-07-10 14:59
-------
Created an attachment (id=1161)
 --> (http://bugzilla.mindrot.org/attachment.cgi?id=1161&action=view)
svn.strace

I think your right - subversion process - line 594




------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

http://bugzilla.mindrot.org/show_bug.cgi?id=1208





------- Comment #6 from daniel.subs at internode.on.net  2006-07-10 15:00
-------
Created an attachment (id=1162)
 --> (http://bugzilla.mindrot.org/attachment.cgi?id=1162&action=view)
svn.strace.12827 (ssh process 1)




------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

http://bugzilla.mindrot.org/show_bug.cgi?id=1208





------- Comment #7 from daniel.subs at internode.on.net  2006-07-10 15:03
-------
Created an attachment (id=1163)
 --> (http://bugzilla.mindrot.org/attachment.cgi?id=1163&action=view)
svn.strace.12837 (ssh process 2)

I read somewhere on the subversion bug records that two transports is
normal.

cvs-ssh works ok for me and so does normal ssh shell with ControlMaster
so I'll put a subversion bug in soon.

Thanks for your time.




------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

http://bugzilla.mindrot.org/show_bug.cgi?id=1208





------- Comment #8 from daniel.subs at internode.on.net  2006-07-10 15:26
-------
Cross references
http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=313371

http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=335528




------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

http://bugzilla.mindrot.org/show_bug.cgi?id=1208





------- Comment #9 from djm at mindrot.org  2006-07-10 15:36 -------
Yes, it looks like svn is using SIGKILL to "tidy up" its ssh
transport.

That is definitely a bug in svn: shutting a transport down using
SIGKILL gives the transport no opportunity to clean up after itself.
There is nothing that OpenSSH can do to remove the socket in this case
as SIGKILL cannot be caught.

BTW Debian bug 313371 is unrelated.




------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

http://bugzilla.mindrot.org/show_bug.cgi?id=1208





------- Comment #10 from dtucker at zip.com.au  2006-07-10 15:39 -------
(In reply to comment #8)
> Cross references
> http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=313371
That's an excellent example of throwing the baby out with the
bathwater: it would appear that svn is using SIGKILL simply because
they want the "terminated by signal" messages, but if that's the
only
reason then all they had to do was use "ssh -q"...




------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

http://bugzilla.mindrot.org/show_bug.cgi?id=1208





------- Comment #11 from dtucker at zip.com.au  2006-07-10 15:44 -------
(In reply to comment #10)
> they want the "terminated by signal" messages,
Doh.  Make that "they *don't* want" ...

Still, as Damien says there's nothing ssh can do to clean up if svn is
SIGKILL'ing it.

Maybe ssh could recover more gracefully from a stale control socket?




------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

http://bugzilla.mindrot.org/show_bug.cgi?id=1208





------- Comment #12 from djm at mindrot.org  2006-07-10 15:48 -------
It seems like the Subversion developers have known about the problem
for some time, but obviously haven't fixed it:

http://svn.haxx.se/dev/archive-2006-03/1020.shtml
http://svn.haxx.se/users/archive-2005-11/0194.shtml




------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

http://bugzilla.mindrot.org/show_bug.cgi?id=1208





------- Comment #13 from djm at mindrot.org  2006-07-10 16:03 -------
Created an attachment (id=1164)
 --> (http://bugzilla.mindrot.org/attachment.cgi?id=1164&action=view)
subversion patch

Try this patch to subversion. It changes its broken behaviour of
SIGKILL'ing ssh while avoiding printing the termination message by
setting ssh's "-q" flag. (completely untested)

If it works, please feed it back to the subversion developers.




------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

http://bugzilla.mindrot.org/show_bug.cgi?id=1208


daniel.subs at internode.on.net changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                URL|                            |http://subversion.tigris.org
                   |                            |/issues/show_bug.cgi?id=2580




------- Comment #14 from daniel.subs at internode.on.net  2006-07-10 18:21
-------
(In reply to comment #10)
Thanks Damien. Patch works beautifully. Tested well.

As per URL I've issued the subversion folks with a bug.

I've attached comparative straces on the subversion bug report.


(In reply to comment #10)
> Maybe ssh could recover more gracefully from a stale control socket?
yes probably. I'm sure subversion isn't the only thing that could kill
a ssh client.

(In reply to comment #9)
> BTW Debian bug 313371 is unrelated.
Here is Peter basically saying that ssh is overly verbose on a couple
of types of signals and provide a patch to make the obvious ones
silent.

Thanks Damian and Darren, much appreciate your help.




------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.