RE: [reSIProcate] More Heap corruption..

["Matthias Moetje - TERASENS GmbH" <moetje@xxxxxxxxxxxx>]

Byron,

 

thanks a lot for your reply. I'm using VS2005 on Windows, thus I can't

really follow your suggestions.

 

I have now changed the PersistanceManager from a pointer to a member

variable and the problems are gone...!

 

Could this problem be specific to the list implementation VS is using?

It contains the following copyright line

* Copyright (c) 1992-2005 by P.J. Plauger. ALL RIGHTS RESERVED.

The code seems a bit ill (maybe correct, but looks awful), there are hundreds

of subroutines and I somewhat felt getting mad when trying to debug

through all that stuff ...

 

For now I can live with it, though the actual reason for the problem 

remains to be discovered..    

Best regards,

Matthias Moetje

TERASENS GmbH
Augustenstraße 24
80333 Munich
GERMANY

Phone: Fax: e-mail: Web:

+49.89.143370-0 +49.89.143370-22 info@xxxxxxxxxxxx www.terasens.com

 

---

From: Byron Campen [mailto:bcampen@xxxxxxxxxxxx]
Sent: Thursday, June 08, 2006 6:53 PM
To: Matthias Moetje - TERASENS GmbH
Subject: Re: [reSIProcate] More Heap corruption..

Using a heap-allocated InMemoryRegistrationDatabase should not be a problem. You might try using the debug stl (pass -D_GLIBCXX_DEBUG to the compiler through CXXFLAGS) to check whether some sort of badness is going on in the stl.

Best regards,

Byron Campen

Byron, Scott,

thanks very much for your replies. I'm using the InMemoryRegistrationDatabase.

I understand that the list elements are'nt pointers, but the std::list implementation

calls delete two times and if I skip the first, the heap corruption does not occur.

I stepped through all of the list implementation code when assigning, but I never

passed the Uri's copy constructor, shouldn't this have occured?

One difference to repro is that I am creating the InMemoryRegistrationDatabase

with "new" (so it is created on the heap itself) and keep a pointer to it as long

as the session is running. Could this be the reason for the problem?

Matthias