GNU bug report logs - #2912
23.0.92; Ediff over TRAMP too slow

Previous Next

To add a comment to this bug, you must first unarchive it, by sending
a message to control AT debbugs.gnu.org, with unarchive 2912 in the body.
You can then email your comments to 2912 AT debbugs.gnu.org in the normal way.

Toggle the display of automated, internal messages from the tracker.

Message #5 received at submit <at> emacsbugs.donarmstrong.com (full text, mbox):

From: David Abrahams <dave <at> boostpro.com>
To: emacs-pretest-bug <at> gnu.org
Subject: 23.0.92; Ediff over TRAMP too slow
Date: Mon, 06 Apr 2009 20:43:37 -0400

Please write in English if possible, because the Emacs maintainers
usually do not have translators to read other languages for them.

Your bug report will be posted to the emacs-pretest-bug <at> gnu.org mailing list.

Please describe exactly what actions triggered the bug
and the precise symptoms of the bug:

Ediff two remote files using /ssh:... protocol

It seems to go over the network for everything.  At first I thought it
was the auto-refining, but I turned that off and it didn't seem to make
any difference in speed.

If Emacs crashed, and you have the Emacs process in the gdb debugger,
please include the output from the following gdb commands:
    `bt full' and `xbacktrace'.
If you would like to further debug the crash, please read the file
/Applications/Emacs.app/Contents/Resources/etc/DEBUG for instructions.


In GNU Emacs 23.0.92.1 (i386-apple-darwin9.6.0, NS apple-appkit-949.43)
 of 2009-04-06 on black.local
Windowing system distributor `Apple', version 10.3.949
configured using `configure  '--with-ns''

Important settings:
  value of $LC_ALL: nil
  value of $LC_COLLATE: nil
  value of $LC_CTYPE: nil
  value of $LC_MESSAGES: nil
  value of $LC_MONETARY: nil
  value of $LC_NUMERIC: nil
  value of $LC_TIME: nil
  value of $LANG: nil
  value of $XMODIFIERS: nil
  locale-coding-system: nil
  default-enable-multibyte-characters: t

Major mode: BSDmakefile

Minor modes in effect:
  diff-auto-refine-mode: t
  shell-dirtrack-mode: t
  show-paren-mode: t
  server-mode: t
  global-auto-revert-mode: t
  delete-selection-mode: t
  tooltip-mode: t
  mouse-wheel-mode: t
  menu-bar-mode: t
  file-name-shadow-mode: t
  global-font-lock-mode: t
  font-lock-mode: t
  global-auto-composition-mode: t
  auto-composition-mode: t
  auto-encryption-mode: t
  auto-compression-mode: t
  line-number-mode: t
  transient-mark-mode: t

Recent input:
- r e v i <tab> <return> <return> <return> C-b C-e 
<return> SPC SPC SPC SPC SPC SPC C-x C-n C-x C-f <M-backspace> 
<M-backspace> <M-backspace> <backspace> <return> C-n 
C-p C-. C-. C-. C-, C-b C-SPC C-e M-w C-x C-f C-a C-. 
C-. C-, C-. C-SPC C-e p r o . c o m : . d o v e c o 
t . s i e <tab> <return> C-n C-n C-n C-n C-p C-p C-p 
C-p C-s a m a z o n C-, C-, C-, C-, C-, C-, C-, C-n 
C-n C-e C-, C-, C-, C-, C-, C-. C-. C-. C-, C-, C-, 
C-, C-, C-p C-p C-n C-n C-n C-, C-, C-, C-, C-, C-. 
C-e C-, C-, C-, C-, C-, C-p C-x C-f . d o <tab> s i 
<tab> . <tab> C-g C-n C-n C-p C-p <down-mouse-1> <mouse-1> 
<down-mouse-1> <mouse-1> <down-mouse-1> <mouse-1> C-, 
C-, C-, C-, C-, C-b C-s C-w C-s C-s C-r C-n C-p C-p 
C-n C-r C-r C-r C-r C-r C-. C-. C-, C-b C-SPC C-, C-b 
<backspace> C-n C-n C-n C-n C-n C-n C-p C-. C-. C-. 
C-, C-b <M-backspace> <backspace> C-n C-n C-n C-n C-. 
C-. C-, C-b <M-backspace> <backspace> C-x C-s C-x C-n 
C-x C-n C-x C-p C-x C-p C-x C-k C-x C-n C-n C-n C-p 
C-p C-x C-n C-x C-g C-x C-g C-x C-b C-x C-b C-x C-p 
C-x C-b C-x C-b C-x C-b C-n C-n C-n C-n C-x b * E d 
i <tab> C <tab> <return> | | <M-S-down> <M-S-down> 
C-g C-g C-a C-n SPC C-a C-e C-a C-e C-n C-b C-b C-b 
C-b C-b SPC p p ? @ p ? C-n C-p C-n C-, C-, C-, C-, 
C-. C-. C-, q y M-x r e p o r t e <backspace> - e m 
<tab> <return>

Recent messages:
Mark saved where search started [2 times]
Mark set
Saving file /ssh:boostpro.com:/home/dave/.dovecot.sieve...
Wrote /ssh:boostpro.com:/home/dave/.dovecot.sieve
Entering debugger...
Quit [2 times]
ediff-next-difference: At end of the difference list
Auto-refining is OFF
Region 4 in buffer A is empty [2 times]
Quit this Ediff session? (y or n) 

-- 
Dave Abrahams
BoostPro Computing
http://www.boostpro.com

Message #12 received at 2912 <at> emacsbugs.donarmstrong.com (full text, mbox):

From: Michael Albinus <michael.albinus <at> gmx.de>
To: David Abrahams <dave <at> boostpro.com>
Cc: 2912 <at> debbugs.gnu.org
Subject: Re: bug#2912: 23.0.92; Ediff over TRAMP too slow
Date: Tue, 07 Apr 2009 09:44:48 +0200

David Abrahams <dave <at> boostpro.com> writes:

> Ediff two remote files using /ssh:... protocol
>
> It seems to go over the network for everything.  At first I thought it
> was the auto-refining, but I turned that off and it didn't seem to make
> any difference in speed.

In case of remote files, ediff compares local copies, indeed. I'm not so
familiar with ediff, therefore I don't know, whether this can be changed
easily, running the diff command on the remote host. However, this
introduces more complexity into ediff: there might be protocols which do
not support execution of remote commands (like ftp or smb). Furthermore,
ediff would need to check, whether the files to be compared reside on
the same remote host.

I don't believe that this shall be changed before the release of Emacs
23.1. After the release, it might be a new feature somebody (with more
skills than I have on ediff) might implement.

For larger files, I recommend to use the scp method instead of ssh.

Best regards, Michael.

Message #17 received at 2912 <at> emacsbugs.donarmstrong.com (full text, mbox):

From: David Abrahams <dave <at> boostpro.com>
To: Michael Albinus <michael.albinus <at> gmx.de>
Cc: 2912 <at> debbugs.gnu.org
Subject: Re: bug#2912: 23.0.92; Ediff over TRAMP too slow
Date: Tue, 7 Apr 2009 09:17:22 -0400

On Apr 7, 2009, at 3:44 AM, Michael Albinus wrote:

> David Abrahams <dave <at> boostpro.com> writes:
>
>> Ediff two remote files using /ssh:... protocol
>>
>> It seems to go over the network for everything.  At first I thought  
>> it
>> was the auto-refining, but I turned that off and it didn't seem to  
>> make
>> any difference in speed.
>
> In case of remote files, ediff compares local copies, indeed.

If that's all it was doing, I wouldn't be complaining.  Each time I  
move to the next diff, it appears to be going over the network.

> I'm not so
> familiar with ediff, therefore I don't know, whether this can be  
> changed
> easily, running the diff command on the remote host. However, this
> introduces more complexity into ediff: there might be protocols  
> which do
> not support execution of remote commands (like ftp or smb).  
> Furthermore,
> ediff would need to check, whether the files to be compared reside on
> the same remote host.

I don't think that would be a big win.  There is, however, a lot to be  
gained somewhere *during* the ediff session.

> I don't believe that this shall be changed before the release of Emacs
> 23.1. After the release, it might be a new feature somebody (with more
> skills than I have on ediff) might implement.
>
> For larger files, I recommend to use the scp method instead of ssh.


Really!  I thought they were roughly the same.  Thanks for the hint.

--
David Abrahams
BoostPro Computing
http://boostpro.com

Message #22 received at 2912 <at> emacsbugs.donarmstrong.com (full text, mbox):

From: Michael Albinus <michael.albinus <at> gmx.de>
To: David Abrahams <dave <at> boostpro.com>
Cc: 2912 <at> debbugs.gnu.org
Subject: Re: bug#2912: 23.0.92; Ediff over TRAMP too slow
Date: Tue, 07 Apr 2009 15:29:26 +0200

David Abrahams <dave <at> boostpro.com> writes:

> I don't think that would be a big win.  There is, however, a lot to be
> gained somewhere *during* the ediff session.

As I said, I'm not familiar with ediff. But there might also be another
reason for network activity. If the files are under vc control, a lot of
checks are always performed silently. You won't see the checks for local
files, because they are fast enough. But for remote files it matters.

If you know, that your files are *not* under vc control, you might apply

  (setq vc-handled-backends nil)

Best regards, Michael.

Message #27 received at 2912 <at> emacsbugs.donarmstrong.com (full text, mbox):

From: David Abrahams <dave <at> boostpro.com>
To: Michael Albinus <michael.albinus <at> gmx.de>
Cc: 2912 <at> debbugs.gnu.org
Subject: Re: bug#2912: 23.0.92; Ediff over TRAMP too slow
Date: Tue, 7 Apr 2009 10:58:06 -0400

On Apr 7, 2009, at 9:29 AM, Michael Albinus wrote:

> David Abrahams <dave <at> boostpro.com> writes:
>
>> I don't think that would be a big win.  There is, however, a lot to  
>> be
>> gained somewhere *during* the ediff session.
>
> As I said, I'm not familiar with ediff. But there might also be  
> another
> reason for network activity. If the files are under vc control, a  
> lot of
> checks are always performed silently.

Wow, you're right!  Ediff'ing files not under vc control is very snappy!
OK, this is weird: now I can't reproduce the problem even for files  
under version control.

I'll let you know if I find out anything more, but I guess we can  
ignore this one for now.

Thanks,
--
David Abrahams
BoostPro Computing
http://boostpro.com

Message #32 received at 2912-done <at> emacsbugs.donarmstrong.com (full text, mbox):

From: Michael Albinus <michael.albinus <at> gmx.de>
To: David Abrahams <dave <at> boostpro.com>
Cc: 2912-done <at> debbugs.gnu.org
Subject: Re: bug#2912: 23.0.92; Ediff over TRAMP too slow
Date: Tue, 07 Apr 2009 17:14:42 +0200

David Abrahams <dave <at> boostpro.com> writes:

> Wow, you're right!  Ediff'ing files not under vc control is very snappy!
> OK, this is weird: now I can't reproduce the problem even for files
> under version control.
>
> I'll let you know if I find out anything more, but I guess we can
> ignore this one for now.

OK, I've closed the bug. If you find something else in this area, it
might be worth a new bug report.

Maybe one could think about enabling/disabling vc on the fly for remote
files, but I have no real use case yet how it shall go.

> Thanks,

Best regards, Michael.

Message #42 received at 2912 <at> emacsbugs.donarmstrong.com (full text, mbox):

From: Stefan Monnier <monnier <at> iro.umontreal.ca>
To: David Abrahams <dave <at> boostpro.com>
Cc: 2912 <at> debbugs.gnu.org, Michael Albinus <michael.albinus <at> gmx.de>
Subject: Re: bug#2912: 23.0.92; Ediff over TRAMP too slow
Date: Tue, 07 Apr 2009 11:26:59 -0400

>>> Ediff two remote files using /ssh:... protocol
>>> It seems to go over the network for everything.  At first I thought it
>>> was the auto-refining, but I turned that off and it didn't seem to make
>>> any difference in speed.
>> In case of remote files, ediff compares local copies, indeed.
> If that's all it was doing, I wouldn't be complaining.  Each time I move to
> the next diff, it appears to be going over the network.

I cannot reproduce this here.  I opened two remote files (via Tramp/SSH)
and called M-x ediff-buffers, then killed Tramp's internal buffer and
moved from diff to diff (via ediff's frame using keys like `n' and SPC)
and it did not try to reconnect to the remote machine.


        Stefan

Previous Next

GNU bug tracking system
Copyright (C) 1999 Darren O. Benham, 1997,2003 nCipher Corporation Ltd, 1994-97 Ian Jackson.