GNU bug report logs - #55832
28.1; Emacs crashes when using tramp from helm in emacs-29

Previous Next

Package: emacs;

Reported by: Thierry Volpiatto <thievol <at> posteo.net>

Date: Tue, 7 Jun 2022 15:24:02 UTC

Severity: normal

Found in version 28.1

Fixed in version 29.1

Done: Michael Albinus <michael.albinus <at> gmx.de>

Bug is archived. No further changes may be made.

Full log


Message #83 received at 55832 <at> debbugs.gnu.org (full text, mbox):

From: Michael Albinus <michael.albinus <at> gmx.de>
To: Thierry Volpiatto <thievol <at> posteo.net>
Cc: Eli Zaretskii <eliz <at> gnu.org>, larsi <at> gnus.org, 55832 <at> debbugs.gnu.org
Subject: Re: bug#55832: 28.1; Emacs crashes when using tramp from helm in
 emacs-29
Date: Sat, 11 Jun 2022 21:27:40 +0200
Thierry Volpiatto <thievol <at> posteo.net> writes:

Hi Thierry,

> Ah ok, I noticed I have always in ~/.emacs.d/tramp an UNKNOW entry for
> uid and gid, is it normal?

When the remote uid or gid cannot be determined, Tramp sets the default
value "UNKNOWN" (for the uid-string resp. remote-gid property), and -1
(for the uid-integer resp. gid-integer property). See
tramp-unknown-id-string and tramp-unknown-id-integer.

However, I would be surprised if these values are used for a "sudo"
connection. Checking my own ~/.emacs.d/tramp I find such entries,
indeed, for a "sudo" connection. Very strange. But although incorrect,
they haven't disturbed my use of Tramp.

After exiting Emacs, deleting ~/.emacs.d/tramp, and starting Emacs again
with a "sudo" connection, there arer proper values "root" and 0 in the
cached properties. Hmm, don't know where these values came from in
history, but I don't believe they should disturb Tramp. But who
knows. Perhaps you move the ~/.emacs.d/tramp file away (in order to use
it later if necessary), start a new Emacs instance, and see, whether it
still crashes with your scenario.

>> Sad. Since I cannot reproduce the problem locally, what happens if you
>> invoke "emacs -Q" similar to how I've done?
>
> I can't reproduce the crash using helm -P path/to/emacs though there is
> not much difference with my helm-find-files config, the difference comes
> I guess from the usage of normal emacs (or possibly only -q) and emacs
> -Q which affect tramp in some ways.
>
> What I would like to understand is why the crash appears AFTER
> tramp-cleanup-all-connections and not before on first shot.

Also a mystery to me.

If the removing of the ~/.emacs.d/tramp file does not help, you might
try the following settings:

--8<---------------cut here---------------start------------->8---
(setq tramp-verbose 10 tramp-debug-to-file t)
--8<---------------cut here---------------end--------------->8---

This increases the Tramp debug messages volume, and writes those
messages to disk at /tmp as they appear, under a file name like the
Tramp debug buffer name. So we have something to analyze even after
crash.

Well, reading the Helm sources, you'll better with setting
helm-tramp-verbose.

Best regards, Michael.




This bug report was last modified 2 years and 332 days ago.

Previous Next


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