When runing valgrind on "source/cblas_dtrsmx.out" compiled from within "mkl/examples/cblas", I end up with a log similar to:
==9740== 3,131,264 bytes in 1 blocks are still reachable in loss record 7 of 7
==9740== at 0x4C28C20: malloc (vg_replace_malloc.c:296)
==9740== by 0x60EAFB4: mkl_serv_allocate (in /opt/intel/composer_xe_2015.1.133/mkl/lib/intel64/libmkl_core.so)
==9740== by 0x9269DBE: mkl_blas_mc3_dgemm_get_bufs (in /opt/intel/composer_xe_2015.1.133/mkl/lib/intel64/libmkl_mc3.so)
==9740== by 0x9242B7A: mkl_blas_mc3_xdtrsm (in /opt/intel/composer_xe_2015.1.133/mkl/lib/intel64/libmkl_mc3.so)
==9740== by 0x4F466C2: DTRSM (in /opt/intel/composer_xe_2015.1.133/mkl/lib/intel64/libmkl_intel_lp64.so)
==9740== by 0x4F60E0C: cblas_dtrsm (in /opt/intel/composer_xe_2015.1.133/mkl/lib/intel64/libmkl_intel_lp64.so)
==9740== by 0x4016D5: main (in /opt/intel/composer_xe_2015.1.133/mkl/examples/cblas/_results/intel_lp64_sequential_intel64_so/cblas_dtrsmx.out)
Is that normal to have that much memory leaked ?
Thank you in advance.
it looks as a false positive generated by valgrind. I have checked the problem with Intel(R) Inspector 2015 with the widest scope memory error analysis type. Intel Inspector detected no problems.