Description
Saw this in a precommit build. Pretty sure it's within openssl, but perhaps due to how we're misusing it?
*** Aborted at 1533139477 (unix time) try "date -d @1533139477" if you are using GNU date *** PC: @ 0x7fb3cf0a0530 (unknown) *** SIGSEGV (@0x7fb3cf0a0530) received by PID 2038 (TID 0x7fb3240fe700) from PID 18446744072888124720; stack trace: *** @ 0x7fb3d80b0330 (unknown) at ??:0 @ 0x7fb3cf0a0530 (unknown) at ??:0 @ 0x7fb3cf0fdb07 (unknown) at ??:0 @ 0x7fb3cf0fa047 EVP_CIPHER_CTX_ctrl at ??:0 @ 0x7fb3cf0fde3c (unknown) at ??:0 @ 0x7fb3ce90b489 (unknown) at ??:0 @ 0x7fb3ce9013d6 (unknown) at ??:0 @ 0x7fb3ce9017d3 (unknown) at ??:0 @ 0x7fb3d75da6ec kudu::security::TlsSocket::Write() at ??:0 @ 0x7fb3d75daaba kudu::security::TlsSocket::Writev() at ??:0 @ 0x7fb3d848ae0d kudu::rpc::OutboundTransfer::SendBuffer() at ??:0 @ 0x7fb3d8414d13 kudu::rpc::Connection::WriteHandler() at ??:0 @ 0x7fb3d841975b ev::base<>::method_thunk<>() at ??:0 @ 0x7fb3cf3f1606 ev_invoke_pending at ??:0 @ 0x7fb3d8442228 kudu::rpc::ReactorThread::InvokePendingCb() at ??:0 @ 0x7fb3cf3f24f8 ev_run at ??:0 @ 0x7fb3d844897f ev::loop_ref::run() at ??:0 @ 0x7fb3d8444fb8 kudu::rpc::ReactorThread::RunThread() at ??:0 @ 0x7fb3d8450179 boost::_mfi::mf0<>::operator()() at ??:0 @ 0x7fb3d844fdea boost::_bi::list1<>::operator()<>() at ??:0 @ 0x7fb3d844f4bd boost::_bi::bind_t<>::operator()() at ??:0 @ 0x7fb3d844ebca boost::detail::function::void_function_obj_invoker0<>::invoke() at ??:0 @ 0x7fb3d843b3c2 boost::function0<>::operator()() at ??:0 @ 0x7fb3d721d8a9 kudu::Thread::SuperviseThread() at ??:0 @ 0x7fb3d80a8184 start_thread at ??:0 @ 0x7fb3d5822ffd clone at ??:0
I'm attaching the full test log.