Mikiya Fujii
mikiy****@gmail*****
2012年 6月 30日 (土) 15:32:32 JST
Dear ktns,
Actually, rev. 836 does not leak any memories.
But there is a bug at line 479 in MallocerFreer.h.
Please fix it.
See following diff:
ubuntu-vm:src$ svn diff base/MallocerFreer.h
Index: base/MallocerFreer.h
===================================================================
--- base/MallocerFreer.h (リビジョン 836)
+++ base/MallocerFreer.h (作業コピー)
@@ -476,7 +476,7 @@
delete [] p3d;
delete [] p2d;
delete [] p1d;
- MallocerFreer::SubtCurrentMalloced(static_cast<double>(size1*size2*size3*size4*size5*sizeof(T)));
+ MallocerFreer::SubtCurrentMalloced(static_cast<double>(size1*size2*size3*size4*size5*size6*sizeof(T)));
*matrix = NUL
Mikiya
2012/6/30 Mikiya Fujii <mikiy****@gmail*****>:
> Dear ktns,
>
> Rev. 836 may leak some memories.
> I get following output for test/ch4_mndo.in.
>
> Summary for memory usage:
> Max Heap: 0.064024[MB].
> Current Heap(Leaked): 0.038400[MB].
>
>
> >>>>> The MolDS finished normally! <<<<<
> >>>>> CPU time: 0.04[s]. <<<<<
> >>>>> Elapsed time: 0[s]. <<<<<
> >>>>> Elapsed time(OMP): 0.0469129[s]. <<<<<
> >>>>> See you. <<<<<
>
> Mikiya