Hi.

If I check my multithreaded program with valgrind in linux I get these memoryleaks:

==27262== 32 bytes in 1 blocks are still reachable in loss record 1 of 2 ==27262== at 0x4C27480: calloc (vg_replace_malloc.c:467) ==27262== by 0x504C31F: _dlerror_run (dlerror.c:142) ==27262== by 0x504BEE0: dlopen@@GLIBC_2.2.5 (dlopen.c:88) ==27262== by 0x43E368: CTHREADS_LOADPTHREADS$$BOOLEAN (pthread.inc:261) ==27262== by 0x43FC38: CTHREADS_CINITTHREADS$$BOOLEAN (cthreads.pp:960) ==27262== by 0x4324D0: SYSTEM_SETTHREADMANAGER$TTHREADMANAGER$$BOOLEAN (thread.inc:241) ==27262== by 0x43FF96: CTHREADS_SETCTHREADMANAGER (cthreads.pp:1025) ==27262== by 0x440023: CTHREADS_init (cthreads.pp:1036) ==27262== by 0x42E8FE: fpc_initializeunits (system.inc:794) ==27262== by 0x417F16: main (xmlrpcserver.dpr:21) ==27262==
==27262== 56 bytes in 1 blocks are still reachable in loss record 2 of 2 ==27262== at 0x4C2815C: malloc (vg_replace_malloc.c:236) ==27262== by 0x400D624: _dl_map_object_deps (dl-deps.c:506) ==27262== by 0x4013964: dl_open_worker (dl-open.c:263) ==27262== by 0x400ED35: _dl_catch_error (dl-error.c:178) ==27262== by 0x4013329: _dl_open (dl-open.c:555) ==27262== by 0x504BF65: dlopen_doit (dlopen.c:67) ==27262== by 0x400ED35: _dl_catch_error (dl-error.c:178) ==27262== by 0x504C2AB: _dlerror_run (dlerror.c:164) ==27262== by 0x504BEE0: dlopen@@GLIBC_2.2.5 (dlopen.c:88) ==27262== by 0x43E368: CTHREADS_LOADPTHREADS$$BOOLEAN (pthread.inc:261) ==27262== by 0x43FC38: CTHREADS_CINITTHREADS$$BOOLEAN (cthreads.pp:960) ==27262== by 0x4324D0: SYSTEM_SETTHREADMANAGER$TTHREADMANAGER$$BOOLEAN (thread.inc:241) ==27262== by 0x43FF96: CTHREADS_SETCTHREADMANAGER (cthreads.pp:1025) ==27262== by 0x440023: CTHREADS_init (cthreads.pp:1036) ==27262== by 0x42E8FE: fpc_initializeunits (system.inc:794) ==27262== by 0x417F16: main (xmlrpcserver.dpr:21)

IMHO it is because CThreadManager.DoneManager (calls UnLoadPthreads) is not called at the end of program.

Is any way how to shutdown threading and satisfy valgrind by dlclose()?

Thanks
Petr