Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]


Groups > comp.lang.clipper.visual-objects > #4582 > unrolled thread

Windows 7 memory corruption

Started bykvstegemann@web.de
First post2013-01-14 09:11 -0800
Last post2013-02-04 02:36 -0800
Articles 10 — 6 participants

Back to article view | Back to comp.lang.clipper.visual-objects


Contents

  Windows 7 memory corruption kvstegemann@web.de - 2013-01-14 09:11 -0800
    Re: Windows 7 memory corruption D.J.W. van Kooten <public@ic2remove.this.com> - 2013-01-14 22:35 +0100
      Re: Windows 7 memory corruption kvstegemann@web.de - 2013-01-15 08:22 -0800
    Re: Windows 7 memory corruption Massimo Bighelli <info@arcasistemi.it> - 2013-01-15 18:23 +0100
    Re: Windows 7 memory corruption "Jamal" <vodotnet_nospam@yahoo.com> - 2013-01-15 16:44 -0500
      Re: Windows 7 memory corruption kvstegemann@web.de - 2013-01-16 05:06 -0800
        Re: Windows 7 memory corruption Karl Faller <k.faller_withoutthat_@onlinehome.de> - 2013-01-16 14:53 +0100
          Re: Windows 7 memory corruption kvstegemann@web.de - 2013-01-18 01:42 -0800
        Re: Windows 7 memory corruption "Jamal" <vodotnet_nospam@yahoo.com> - 2013-01-16 20:18 -0500
    Re: Windows 7 memory corruption Thomas <Thomas.Walter@ndstsk.de> - 2013-02-04 02:36 -0800

#4582 — Windows 7 memory corruption

Fromkvstegemann@web.de
Date2013-01-14 09:11 -0800
SubjectWindows 7 memory corruption
Message-ID<f4eb5757-b901-4e53-808a-907b44a1e679@googlegroups.com>
Hi out there,

we have a rather big VO28SP2 application running at several client's sites and the clients are in the process of switching from WinXP to Windows 7. This went without big problems overall but I am trying to track down a certain problem which seems very evasive.

Clients report that the application crashes at some point and I can sometimes reproduce these crashes with Win7, even in the debugger. The crash looks like some sort of dynamic memory corruption (5333 for example at ADEDataWindow:Init() line 11, but only sometimes, and sometimes somewhere else).

It was like in the old days when we tried to fix the GPFs. I introduced some _DynCheck()/DynCheckError() calls and stepped through the code but I never was able to put my finger on the place where memory corruption might be caused. I placed those two function calls in the Watch Expressions at watched them, and at some point they did not show any results any longer. I tried DynLock() but the crash occurred anyway. The crash is not always logged as a 5333, sometimes the application simply stops working and Windows says the application does not work any longer.

The funny thing is that with XP everything is fine. At the moment it SEEMS that the problem also disappears when the application is running in Windows 7 in XP compatibility mode but I'm not sure about that.

I'm running out of ideas. Anyone?

TIA
Kay-Vikor
 

[toc] | [next] | [standalone]


#4584

FromD.J.W. van Kooten <public@ic2remove.this.com>
Date2013-01-14 22:35 +0100
Message-ID<p7u8f85bg6tjvv26ov96qi182ji14c1t8g@4ax.com>
In reply to#4582
On Mon, 14 Jan 2013 09:11:04 -0800 (PST), kvstegemann@web.de wrote:

Hello Kay-Vikor,

>
>Clients report that the application crashes at some point and I can sometimes reproduce these crashes with Win7, even in the debugger. 


I doubt if what follows is at all comparable with what you report but
it may be worth a try. When we first started with W7 3 years ago, I
had unexplained problems as well. For me it usually showed as program
code not running when e.g. pressing a button, then after a bit of
waiting I got a kid stack size error.

It was never explained but what helped is to deliver an exe without
any debug status on "on". I did NOT have the problem when running from
the IDE (with debug on) fortunately, and an exe with debug on was
working fine on XP or Vista.


You may want to check if this helps you too.

Dick

[toc] | [prev] | [next] | [standalone]


#4585

Fromkvstegemann@web.de
Date2013-01-15 08:22 -0800
Message-ID<b9d1ba78-7694-47a1-a971-49b2ed797515@googlegroups.com>
In reply to#4584
Hi Dick,

thanks for the suggestion. I am using a repository with all debug switches off to create my production exes, so that is already covered.

Kay-Viktor

[toc] | [prev] | [next] | [standalone]


#4586

FromMassimo Bighelli <info@arcasistemi.it>
Date2013-01-15 18:23 +0100
Message-ID<kd43as$8vr$3@speranza.aioe.org>
In reply to#4582
Hi Kay-Vicor

do you think it's related to file accesses
(use / open / ...) ?
if so, look at SMB2 on win7

Massimo

[toc] | [prev] | [next] | [standalone]


#4587

From"Jamal" <vodotnet_nospam@yahoo.com>
Date2013-01-15 16:44 -0500
Message-ID<kd4ij5$97v$1@dont-email.me>
In reply to#4582
If it is critical to resolve the issue,  upgrading may be VO SP4 can help 
which has:

"A fix has been applied to the internal handling of dynamic memory to avoid 
a problem with the function DynProtect."

Of course there are many other fixes.

HTH,
Jamal

wrote in message 
news:f4eb5757-b901-4e53-808a-907b44a1e679@googlegroups.com...

Hi out there,

we have a rather big VO28SP2 application running at several client's sites 
and the clients are in the process of switching from WinXP to Windows 7. 
This went without big problems overall but I am trying to track down a 
certain problem which seems very evasive.

Clients report that the application crashes at some point and I can 
sometimes reproduce these crashes with Win7, even in the debugger. The crash 
looks like some sort of dynamic memory corruption (5333 for example at 
ADEDataWindow:Init() line 11, but only sometimes, and sometimes somewhere 
else).

It was like in the old days when we tried to fix the GPFs. I introduced some 
_DynCheck()/DynCheckError() calls and stepped through the code but I never 
was able to put my finger on the place where memory corruption might be 
caused. I placed those two function calls in the Watch Expressions at 
watched them, and at some point they did not show any results any longer. I 
tried DynLock() but the crash occurred anyway. The crash is not always 
logged as a 5333, sometimes the application simply stops working and Windows 
says the application does not work any longer.

The funny thing is that with XP everything is fine. At the moment it SEEMS 
that the problem also disappears when the application is running in Windows 
7 in XP compatibility mode but I'm not sure about that.

I'm running out of ideas. Anyone?

TIA
Kay-Vikor 

[toc] | [prev] | [next] | [standalone]


#4593

Fromkvstegemann@web.de
Date2013-01-16 05:06 -0800
Message-ID<bda28397-677e-463c-9fcf-6bfbd61b9ab0@googlegroups.com>
In reply to#4587
Massimo, thanks for the input. If you are talking about Xbase type file operations, we do not use xBase DB, we are using SQL databases. The application does write to text files, though.

Jamal, if I knew that SP4 would solve my problem I'd grab it anytime. I just checked but DynProtect() is used nowhere in the app, so I don't think this particular fix will help me.

Kay-Viktor

[toc] | [prev] | [next] | [standalone]


#4595

FromKarl Faller <k.faller_withoutthat_@onlinehome.de>
Date2013-01-16 14:53 +0100
Message-ID<45cdf8d7riea44ajvf8lj3sv7hol6vodrm@4ax.com>
In reply to#4593
On Wed, 16 Jan 2013 05:06:01 -0800 (PST), kvstegemann@web.de wrote:
> I just checked but DynProtect() is used nowhere in the app, so I don't think this particular fix will help me.
the question is: does the SDK code use DynProtect?

Karl

[toc] | [prev] | [next] | [standalone]


#4600

Fromkvstegemann@web.de
Date2013-01-18 01:42 -0800
Message-ID<ad8e74d6-cd0f-4c46-8029-5abd2e40a487@googlegroups.com>
In reply to#4595
> the question is: does the SDK code use DynProtect?

Karl,

I just checked and the answer is no :)

Anyway, I'll probably upgrade to SP4 and let's see if there's a difference.

Cheers
Kay-Viktor

[toc] | [prev] | [next] | [standalone]


#4596

From"Jamal" <vodotnet_nospam@yahoo.com>
Date2013-01-16 20:18 -0500
Message-ID<kd7jha$4kv$1@dont-email.me>
In reply to#4593
Kay-Viktor,

Well, you are 2 services packs behind and lots of fixes have been made in VO 
in terms of memory handling and run time, etc.
Whether an upgrade will resolve your issues, I do not know, but hopefully it 
will.

wrote in message 
news:bda28397-677e-463c-9fcf-6bfbd61b9ab0@googlegroups.com...

Massimo, thanks for the input. If you are talking about Xbase type file 
operations, we do not use xBase DB, we are using SQL databases. The 
application does write to text files, though.

Jamal, if I knew that SP4 would solve my problem I'd grab it anytime. I just 
checked but DynProtect() is used nowhere in the app, so I don't think this 
particular fix will help me.

Kay-Viktor 

[toc] | [prev] | [next] | [standalone]


#4719

FromThomas <Thomas.Walter@ndstsk.de>
Date2013-02-04 02:36 -0800
Message-ID<34916dea-c988-43aa-9f4a-d143b4348a7a@googlegroups.com>
In reply to#4582
Remove any Altd() in your sourcecode.

[toc] | [prev] | [standalone]


Back to top | Article view | comp.lang.clipper.visual-objects


csiph-web