[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
bug#31245: 27.0.50; Crash in Windows emacs
From: |
Eli Zaretskii |
Subject: |
bug#31245: 27.0.50; Crash in Windows emacs |
Date: |
Wed, 02 May 2018 17:49:04 +0300 |
> From: Andy Moreton <andrewjmoreton@gmail.com>
> Date: Tue, 01 May 2018 21:47:00 +0100
>
> On Mon 23 Apr 2018, Andy Moreton wrote:
>
> > On Mon 23 Apr 2018, Eli Zaretskii wrote:
> >> Thanks, but this kind of problems is impossible to debug without a
> >> reproducer. Can you come up with one?
>
> I still cannot use a debugger on this system, but I have collected two
> more backtraces from emacs built with the same options as before, from
> changeset 65ac27783a959a8339c2aab0f1e54d9b508a1f1f on master.
>
> In both of these I was typing in an rnc-mode buffer (for a Relax-NG
> schema), and it crashed while I typed OSVERSION as an element name in
> the schema (only the O was displayed). The file is plain ascii.
Thanks. There's no need to run Emacs under a debugger in this case,
AFAIU. What is needed is a recipe that would more or less reliably
reproduce the problem. The place where the assertion is violated and
why Emacs aborts is quite clear; what is missing is the root cause of
that impossible situation. And that can only be provided by a recipe.
Can you describe what should one do to reproduce this problem,
starting from "emacs -Q", loading rnc-mode and anything else that is
needed, then typing or evaluating something that causes the crash?