openssh bugs - Feb 2013 - [Bug 2067] New: lsetstat extension to sftp-server

https://bugzilla.mindrot.org/show_bug.cgi?id=2067

            Bug ID: 2067
           Summary: lsetstat extension to sftp-server
    Classification: Unclassified
           Product: Portable OpenSSH
           Version: 6.1p1
          Hardware: Other
                OS: Other
            Status: NEW
          Severity: enhancement
          Priority: P5
         Component: sftp-server
          Assignee: unassigned-bugs at mindrot.org
          Reporter: bert at bertenselena.net

Created attachment 2212
  --> https://bugzilla.mindrot.org/attachment.cgi?id=2212&action=edit
Patch for sftp-server to add lsetstat extension

At the moment sftp is not able to set timestamps on symlinks.
In order to make this possible lsetstat needs to be implemented.
Attached patch makes this addition.

Some background info. I want to run dirvish over sshfs in order to make
daily remote snapshot-backups. Currently rsync (which is used by
dirvish) reports an error when updating the timestamp of a symlink.
This in turn makes dirvish abort the backup.

In cooperation with Miklos Szeredi, the author of sshfs, I have created
attached patch. He has created the functionality in sshfs to make use
of this lsetstat extenstion in sftp. I've used this patch for 5 months
now and it behaves perfectly. Dirvish creates complete backups every
day now.

I have also created a patch for the sftp client-side to implement a
"touch [-h]" command. This command makes it possible to update the
utime of a symlink. It was mostly used to test above functionality, but
will be handy in more cases. I will add this patch here too.

-- 
You are receiving this mail because:
You are watching the assignee of the bug.

https://bugzilla.mindrot.org/show_bug.cgi?id=2067

--- Comment #1 from Bert Haverkamp <bert at bertenselena.net> ---
Created attachment 2213
  --> https://bugzilla.mindrot.org/attachment.cgi?id=2213&action=edit
Patch to sftp to add touch [-h] command

-- 
You are receiving this mail because:
You are watching the assignee of the bug.

https://bugzilla.mindrot.org/show_bug.cgi?id=2067

Bert Haverkamp <bert at bertenselena.net> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Attachment #2212|application/octet-stream    |text/plain
          mime type|                            |
   Attachment #2212|0                           |1
           is patch|                            |

-- 
You are receiving this mail because:
You are watching the assignee of the bug.

https://bugzilla.mindrot.org/show_bug.cgi?id=2067

TJ Saunders <tj at castaglia.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |tj at castaglia.org

-- 
You are receiving this mail because:
You are watching the assignee of the bug.

https://bugzilla.mindrot.org/show_bug.cgi?id=2067

--- Comment #2 from Bert Haverkamp <bert at bertenselena.net> ---
Can someone please take a look to this patch? I have been using it in
my system for over a year now. I'd love to have it included. Or at
least reviewed;-)

-- 
You are receiving this mail because:
You are watching the assignee of the bug.

https://bugzilla.mindrot.org/show_bug.cgi?id=2067

--- Comment #3 from Bert Haverkamp <bert at bertenselena.net> ---
Hopefully someone can take this patch into openssh?
I am running this happily for a long time now. But recompiling with
every  upgrade is a hassle..

-- 
You are receiving this mail because:
You are watching the assignee of the bug.

https://bugzilla.mindrot.org/show_bug.cgi?id=2067

A. James Lewis <james at fsck.co.uk> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |james at fsck.co.uk

--- Comment #4 from A. James Lewis <james at fsck.co.uk> ---
I'm afraid I don't have much to add, other than to say that the OP is
not the only person with this issue, I know at least one person who has
been lamenting the lack of this feature for some time.... I guess not
everyone is prepared to create an account to add their support!

-- 
You are receiving this mail because:
You are watching the assignee of the bug.

https://bugzilla.mindrot.org/show_bug.cgi?id=2067

Julien Picalausa <jpicalau+mindrot at elmandria.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |jpicalau+mindrot at elmandria.
                   |                            |com

--- Comment #5 from Julien Picalausa <jpicalau+mindrot at elmandria.com>
---
I just want to +1 this. Also, I want to point out that this seem like
the root cause of an issue with extracting tar archives when using
sshfs. In short, if a symlink gets extracted before its target, when
tar attempts to set the time for that symlink, it fails because the
target of the symlink doesn't (yet) exist, causing tar itself to report
an error. This results for instance in steam failing to install its
runtime when running from a home folder mounted over sshfs.

I haven't looked at the patch proposed here, so I can't vouch for it,
but lsetstat seem to be a very much needed feature, especially in a
world where sshfs is becoming a popular way to mount remote folders.

-- 
You are receiving this mail because:
You are watching the assignee of the bug.

https://bugzilla.mindrot.org/show_bug.cgi?id=2067

Damien Miller <djm at mindrot.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Attachment #2212|0                           |1
        is obsolete|                            |
   Attachment #2213|0                           |1
        is obsolete|                            |
                 CC|                            |djm at mindrot.org,
                   |                            |dtucker at dtucker.net
           Assignee|unassigned-bugs at mindrot.org |djm at mindrot.org
   Attachment #3219|                            |ok?(dtucker at dtucker.net)
              Flags|                            |

--- Comment #6 from Damien Miller <djm at mindrot.org> ---
Created attachment 3219
  --> https://bugzilla.mindrot.org/attachment.cgi?id=3219&action=edit
sftp-server lsetstat extension, sftp client chmod/chgrp/chown -h flag

This implements the lsetstat at openssh.com extension in the server, with
a few more Attrib fields supported than the previous patches here.

On the client side, it adds a -h to the chown, chmod and chgrp command
to give them no-dereference semantics.

-- 
You are receiving this mail because:
You are watching the assignee of the bug.
You are watching someone on the CC list of the bug.

https://bugzilla.mindrot.org/show_bug.cgi?id=2067

Damien Miller <djm at mindrot.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Blocks|                            |2915


Referenced Bugs:

https://bugzilla.mindrot.org/show_bug.cgi?id=2915
[Bug 2915] Tracking bug for 8.0 release
-- 
You are receiving this mail because:
You are watching the assignee of the bug.
You are watching someone on the CC list of the bug.

https://bugzilla.mindrot.org/show_bug.cgi?id=2067

Darren Tucker <dtucker at dtucker.net> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Attachment #3219|ok?(dtucker at dtucker.net)    |ok+
              Flags|                            |

--- Comment #7 from Darren Tucker <dtucker at dtucker.net> ---
Comment on attachment 3219
  --> https://bugzilla.mindrot.org/attachment.cgi?id=3219
sftp-server lsetstat extension, sftp client chmod/chgrp/chown -h flag
>-	sshbuf_reset(msg);
these look like unrelated cleanup and should probably be a separate
commit.  other than that, ok dtucker

-- 
You are receiving this mail because:
You are watching someone on the CC list of the bug.
You are watching the assignee of the bug.

https://bugzilla.mindrot.org/show_bug.cgi?id=2067

Damien Miller <djm at mindrot.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|---                         |FIXED

--- Comment #8 from Damien Miller <djm at mindrot.org> ---
This has been committed and will be in openssh-8.0.

I haven't committed the "touch" command for sftp, but if you have
a
good use case for it then I can consider it.

-- 
You are receiving this mail because:
You are watching someone on the CC list of the bug.
You are watching the assignee of the bug.

https://bugzilla.mindrot.org/show_bug.cgi?id=2067

--- Comment #9 from Bert Haverkamp <bert at bertenselena.net> ---
Thanks for including this lsetstat extention patch in V8.0!

I have no personal usecase for the touch command. I added it as it is a
logical addition, based on the added hardlink functionality. It would
be a shame to leave it out, as it is very low-impact.

-- 
You are receiving this mail because:
You are watching the assignee of the bug.
You are watching someone on the CC list of the bug.

https://bugzilla.mindrot.org/show_bug.cgi?id=2067

Damien Miller <djm at mindrot.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|RESOLVED                    |CLOSED

--- Comment #10 from Damien Miller <djm at mindrot.org> ---
closing resolved bugs as of 8.6p1 release

-- 
You are receiving this mail because:
You are watching someone on the CC list of the bug.
You are watching the assignee of the bug.