[Ns-developers] buildbot failure in NsNam on linux

[Mathieu Lacage]

I have not had time to check yet but we have this bug report:

http://www.nsnam.org/bugzilla/show_bug.cgi?id=589

On Sun, 2009-06-14 at 18:09 +0200, Faker.Moatamri at sophia.inria.fr wrote:
> Hi,
> It looks like we have problems with valgrind no?
> Invalid write of size 8
> ==18742==    at 0x50FD012:
> __gnu_cxx::new_allocator<ns3::BufferData*>::construct(ns3::BufferData**,
> ns3::BufferData* const&) (new_allocator.h:104)
> ==18742==    by 0x50FE1BC: std::vector<ns3::BufferData*,
> std::allocator<ns3::BufferData*> >::push_back(ns3::BufferData* const&)
> (stl_vector.h:606)
> ==18742==    by 0x50F74D2: ns3::Buffer::Recycle(ns3::BufferData*)
> (buffer.cc:164)
> ==18742==    by 0x50F86F9: ns3::Buffer::~Buffer() (buffer.cc:311)
> ==18742==    by 0x512E2C8: ns3::Packet::~Packet() (packet.h:202)
> ==18742==    by 0x511F65D: ns3::Packet::Unref() const (packet.cc:123)
> ==18742==    by 0x511BDDD: ns3::Ptr<ns3::Packet>::~Ptr() (ptr.h:432)
> ==18742==    by 0x51B7910: ns3::UdpSocketImplTest::~UdpSocketImplTest()
> (udp-test.cc:80)
> ==18742==    by 0x51B755B: __static_initialization_and_destruction_0(int,
> int) (udp-test.cc:262)
> ==18742==    by 0x51B757C: _GLOBAL__D__ZN3ns317UdpSocketImplTestC2Ev
> (udp-test.cc:266)
> ==18742==    by 0x50037E1: (within
> /home/buildslave/slave/full-rahan-g++-4.1.2/build/build/debug/libns3.so)
> ==18742==    by 0x5468AA0: (within
> /home/buildslave/slave/full-rahan-g++-4.1.2/build/build/debug/libns3.so)
> ==18742==  Address 0x5e19940 is 40 bytes inside a block of size 64 free'd
> ==18742==    at 0x4A05B9D: operator delete(void*) (vg_replace_malloc.c:342)
> ==18742==    by 0x50FD542:
> __gnu_cxx::new_allocator<ns3::BufferData*>::deallocate(ns3::BufferData**,
> unsigned long) (new_allocator.h:94)
> ==18742==    by 0x50FD574: std::_Vector_base<ns3::BufferData*,
> std::allocator<ns3::BufferData*> >::_M_deallocate(ns3::BufferData**,
> unsigned long) (stl_vector.h:133)
> ==18742==    by 0x50FD5B5: std::_Vector_base<ns3::BufferData*,
> std::allocator<ns3::BufferData*> >::~_Vector_base() (stl_vector.h:119)
> ==18742==    by 0x50FD60C: std::vector<ns3::BufferData*,
> std::allocator<ns3::BufferData*> >::~vector() (stl_vector.h:272)
> ==18742==    by 0x50F6819: ns3::BufferDataList::~BufferDataList()
> (buffer.cc:125)
> ==18742==    by 0x50F68C4: __static_initialization_and_destruction_0(int,
> int) (buffer.cc:96)
> ==18742==    by 0x50F68E6: _GLOBAL__D__ZN3ns314BufferDataListD2Ev
> (buffer.cc:1483)
> ==18742==    by 0x50037E1: (within
> /home/buildslave/slave/full-rahan-g++-4.1.2/build/build/debug/libns3.so)
> ==18742==    by 0x5468AA0: (within
> /home/buildslave/slave/full-rahan-g++-4.1.2/build/build/debug/libns3.so)
> ==18742==    by 0x30CEE3659C: exit (in /lib64/libc-2.9.so)
> ==18742==    by 0x30CEE1E57C: (below main) (in /lib64/libc-2.9.so)
> The same error is happening for the linux builders, any idea what is
> happening?
> Regards
> Faker