emacs frame screws up now and then on Windows XP

emacs frame screws up now and then on Windows XP

[William Xu]

[-- Attachment #1: Type: text/plain, Size: 455 bytes --]

Hi, 

See the attached screenshot.  Why does the emacs frame screw up now and
then? I only see this on Windows XP.  When this happens, I have to
restart emacs.  

Firstly I thought it might be some memory problem, but i have not seen
this problem in other applications, like firefox, outlook, etc.  So
really something goes wrong with emacs?

Any thoughts? 

GNU Emacs 23.1.90.1 (i386-mingw-nt5.1.2600) of 2010-01-07 

-- 
William

http://xwl.appspot.com

[-- Attachment #2: screw.png --]
[-- Type: image/png, Size: 7747 bytes --]

Re: emacs frame screws up now and then on Windows XP

[Eli Zaretskii]

> From: William Xu <william.xwl@gmail.com>
> Date: Wed, 03 Feb 2010 13:37:38 +0800
> 
> See the attached screenshot.  Why does the emacs frame screw up now and
> then? I only see this on Windows XP.  When this happens, I have to
> restart emacs.  

Never saw anything like that.

Re: emacs frame screws up now and then on Windows XP

[Lennart Borgman]

On Wed, Feb 3, 2010 at 6:56 PM, Eli Zaretskii <eliz@gnu.org> wrote:
>> From: William Xu <william.xwl@gmail.com>
>> Date: Wed, 03 Feb 2010 13:37:38 +0800
>>
>> See the attached screenshot.  Why does the emacs frame screw up now and
>> then? I only see this on Windows XP.  When this happens, I have to
>> restart emacs.
>
> Never saw anything like that.


I have never seen anything like that either, but I have seen Emacs
hanging and crashing. Most of the times I believe it has something to
do with frame creation, but I also suspect process creation sometimes.
I am just not sure what is happening. Most of the time it is hard to
reproduce (though I think I have given some examples of more easily
reproducible crashes too).

Since the crashes mostly happens sometime after frame creation I have
suspected that it is some lookup in the w32 message handling though it
takes a lot of time to dig into it.

I have also suspected some badly protected data with resulting data
corruption. This is not easy either (especially since there is no good
marking of which part of the code is run in which thread).

Your example looks like data corruption of course. Have you noticed
some special chain of events that leads to this state?