[Dune] memory loss with matrix operator in dune-disc

[Oliver Sander]

Hi!
Thanks for the message.  It seems like nobody has the time to look
at this right now.  Could you please open a new entry in the bug
tracker so this doesn't get lost?

Thanks,
Oliver

Chamakuri Nagaiah schrieb:
> Dear Dune,
> I realized that there is a memory loss problem in the matrix operator.  
> For example I have tested the testadaptivity.cc problem in dune-dd  
> using valgrind debugger and copied the debugging summary .
> I ran the simulation with different number of iterations as following...
>
> 1) ./valgrind-3.3.0/bin/valgrind ./testadapt alusimplex 1 local  
> nopicture 0.0025 0.1
> summary for the valgrind :
> ==21340==    definitely lost: 37,344 bytes in 8 blocks.
> ==21340==      possibly lost: 0 bytes in 0 blocks.
> ==21340==    still reachable: 0 bytes in 0 blocks.
> ==21340==         suppressed: 0 bytes in 0 blocks.
>
> 2) ./valgrind-3.3.0/bin/valgrind ./testadapt alusimplex 2 local  
> nopicture 0.0025 0.1
> summary :
> ==21461==    definitely lost: 132,264 bytes in 20 blocks.
> ==21461==      possibly lost: 0 bytes in 0 blocks.
> ==21461==    still reachable: 0 bytes in 0 blocks.
> ==21461==         suppressed: 0 bytes in 0 blocks.
>
>
>
> 3) ./valgrind-3.3.0/bin/valgrind ./testadapt alusimplex 3 local  
> nopicture 0.0025 0.1
>
> summary :
> ==21674==    definitely lost: 303,888 bytes in 32 blocks.
> ==21674==      possibly lost: 0 bytes in 0 blocks.
> ==21674==    still reachable: 0 bytes in 0 blocks.
> ==21674==         suppressed: 0 bytes in 0 blocks.
>
>
>
> 4) ./valgrind-3.3.0/bin/valgrind ./testadapt alusimplex 4 local  
> nopicture 0.0025 0.1
> summary :
> ==21678==    definitely lost: 563,280 bytes in 44 blocks.
> ==21678==      possibly lost: 0 bytes in 0 blocks.
> ==21678==    still reachable: 0 bytes in 0 blocks.
> ==21678==         suppressed: 0 bytes in 0 blocks.
>
>
> if I increase the number of iterations, then memory lost increases  
> subsequently.
>
> I tried to figure it out but did not succeed. I would appreciate
> your response.
>
> Regards
> Nagaiah
>
>
>
>
>
>
> _______________________________________________
> Dune mailing list
> Dune at dune-project.org
> http://lists.dune-project.org/mailman/listinfo/dune
>   


-- 
************************************************************************
* Oliver Sander                ** email: sander at mi.fu-berlin.de        *
* Freie Universität Berlin     ** phone: + 49 (30) 838 75217           *
* Institut für Mathematik II   ** URL  : page.mi.fu-berlin.de/~sander  *
* Arnimallee 6                 ** -------------------------------------*
* 14195 Berlin, Germany        ** Member of MATHEON (www.matheon.de)   *
************************************************************************