[Valgrind-users] Lock address can't be shown?

[Phil L. <plo...@sa...>]

Using valgrind 3.8.0 on a freebsd system, I get this report:

==23046== Possible data race during read of size 8 at 0x12D2C5A0 by thread #65
==23046== Locks held: none
==23046==    at 0x41C11F6: memmove (in /lib/libc.so.7)
==23046==    by 0x1030375: variable_list** std::__copy_move<false, true, std::random_access_iterator_tag>::__copy_m<variable_list*>(variable_list* const*, variable_list* const*, variable_list**) (stl_algobase.h:377)
==23046==    by 0x10303BD: variable_list** std::__copy_move_a<false, variable_list**, variable_list**>(variable_list**, variable_list**, variable_list**) (stl_algobase.h:396)
==23046==    by 0x1030405: variable_list** std::__copy_move_a2<false, variable_list**, variable_list**>(variable_list**, variable_list**, variable_list**) (stl_algobase.h:435)
==23046==    by 0x1030447: variable_list** std::copy<variable_list**, variable_list**>(variable_list**, variable_list**, variable_list**) (stl_algobase.h:466)
==23046==    by 0x1030473: variable_list** std::__uninitialized_copy<true>::uninitialized_copy<variable_list**, variable_list**>(variable_list**, variable_list**, variable_list**) (stl_uninitialized.h:98)
==23046==    by 0x103049A: variable_list** std::uninitialized_copy<variable_list**, variable_list**>(variable_list**, variable_list**, variable_list**) (stl_uninitialized.h:122)
==23046==    by 0x10304C5: variable_list** std::__uninitialized_copy_a<variable_list**, variable_list**, variable_list*>(variable_list**, variable_list**, variable_list**, std::allocator<variable_list*>&) (stl_uninitialized.h:262)
==23046==    by 0x10304F4: variable_list** std::__uninitialized_move_a<variable_list**, variable_list**, std::allocator<variable_list*> >(variable_list**, variable_list**, variable_list**, std::allocator<variable_list*>&) (stl_uninitialized.h:272)
==23046==    by 0x103074A: std::vector<variable_list*, std::allocator<variable_list*> >::_M_insert_aux(__gnu_cxx::__normal_iterator<variable_list**, std::vector<variable_list*, std::allocator<variable_list*> > >, variable_list* const&) (vector.tcc:316)
==23046==    by 0x10308DC: std::vector<variable_list*, std::allocator<variable_list*> >::insert(__gnu_cxx::__normal_iterator<variable_list**, std::vector<variable_list*, std::allocator<variable_list*> > >, variable_list* const&) (vector.tcc:113)
==23046==    by 0x102E417: snmpNotification::addVarbind(snmpVarbind const&, unsigned long) (snmpNotification.cpp:392)
==23046==
==23046== This conflicts with a previous write of size 8 by thread #43
==23046== Locks held: 3, at addresses 0xFDA2148 0xFE2B020  (and 1 that can't be shown)
==23046==    at 0x11D4CA3: pdbBulkUpdateRecord::pdbBulkUpdateRecord(pdbBulkUpdateRecord const&) (BulkUpdateDefs.h:36)
==23046==    by 0x1212730: __gnu_cxx::new_allocator<pdbBulkUpdateRecord>::construct(pdbBulkUpdateRecord*, pdbBulkUpdateRecord const&) (new_allocator.h:108)
==23046==    by 0x1216A4C: std::vector<pdbBulkUpdateRecord, std::allocator<pdbBulkUpdateRecord> >::push_back(pdbBulkUpdateRecord const&) (stl_vector.h:690)

What could cause a lock to have an address that can't be shown?

Phil
-----
Phil Longstaff
Senior Software Engineer
x2904