[gecode-users] Crash while freeing memory in connection with VS2010

[Mikael Zayenz Lagerkvist]

Hi,

You seem to be using managed C++. Have you tried running your program with
standard C++? The different memory policy might be a source of problems.

I don't have a Windows machine to test on, so I haven't tried your program
myself.

Cheers,
Mikael

2011/1/19 Alexander Kleff <alexander.kleff at ptv.de>

> Hi once again,
>
>
>
> the code below compiles and links in both versions (so I assume I did not
> make an all too obvious modeling mistake) but it crashes in VS2010 in
> contrast to VS2008(Express).
>
> The error seems to occur during the destruction of the linear expression.
> The debugger sets the break point to Heap::rfree(void* p) in heap.hpp in all
> three cases.
>
> Can you reproduce it?
>
>
>
> Thanks for help,
>
> Alexander
>
>
>
>
>
>
>
> #include <gecode/minimodel.hh>
>
>
>
> using namespace Gecode;
>
>
>
> class GecodeTest : public Space {
>
> public:
>
>   GecodeTest(void)
>
>   {
>
>    IntVar z(*this, 0, 1);
>
>
>
>    IntVar e1 = expr(*this, 1 * z);
>
>    IntVar e2 = expr(*this, z);      // error 1
>
>
>
>    IntArgs cost(2);
>
>    for (int v=0; v<2; ++v)
>
>    {
>
>       cost[v] = v;
>
>    }
>
>
>
>    LinExpr xy = element(cost, z);
>
>    IntVar e3 = expr(*this, xy);
>
>    IntVar e4 = expr(*this, element(cost, z));   // error 2
>
>   } // error 3 (in destructor)
>
>
>
>
>
>   GecodeTest(bool share, GecodeTest& s) : Space(share, s) {}
>
>   virtual Space* copy(bool share) { return new GecodeTest(share,*this); }
>
> };
>
>
>
>
>
> // main function
>
> int main(array<System::String ^> ^args)
>
> {
>
>   GecodeTest* m = new GecodeTest;
>
>   delete m;
>
>   return 0;
>
> }
>
>
>
> _______________________________________________
> Gecode users mailing list
> users at gecode.org
> https://www.gecode.org/mailman/listinfo/gecode-users
>
>


-- 
Mikael Zayenz Lagerkvist, http://www.ict.kth.se/~zayenz/
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.gecode.org/pipermail/users/attachments/20110119/b296b198/attachment-0001.htm>