Uploaded image for project: 'Apache Arrow'
  1. Apache Arrow
  2. ARROW-18351

[C++][Flight] Crash in UcxErrorHandlingTest.TestDoExchange

    XMLWordPrintableJSON

Details

    Description

      I get a non-deterministic crash in the Flight UCX tests.

      [----------] 3 tests from UcxErrorHandlingTest
      [ RUN      ] UcxErrorHandlingTest.TestGetFlightInfo
      [       OK ] UcxErrorHandlingTest.TestGetFlightInfo (24 ms)
      [ RUN      ] UcxErrorHandlingTest.TestDoPut
      [       OK ] UcxErrorHandlingTest.TestDoPut (15 ms)
      [ RUN      ] UcxErrorHandlingTest.TestDoExchange
      /arrow/cpp/src/arrow/util/future.cc:125:  Check failed: !IsFutureFinished(state_) Future already marked finished
      

      Here is the GDB backtrace:

      #0  __GI_raise (sig=sig@entry=6) at ../sysdeps/unix/sysv/linux/raise.c:51
      #1  0x00007f18c49cd7f1 in __GI_abort () at abort.c:79
      #2  0x00007f18c5854e00 in arrow::util::CerrLog::~CerrLog (this=0x7f18a81607b0, __in_chrg=<optimized out>) at /arrow/cpp/src/arrow/util/logging.cc:72
      #3  0x00007f18c5854e1c in arrow::util::CerrLog::~CerrLog (this=0x7f18a81607b0, __in_chrg=<optimized out>) at /arrow/cpp/src/arrow/util/logging.cc:74
      #4  0x00007f18c5855181 in arrow::util::ArrowLog::~ArrowLog (this=0x7f18c07fc380, __in_chrg=<optimized out>) at /arrow/cpp/src/arrow/util/logging.cc:250
      #5  0x00007f18c5826f86 in arrow::ConcreteFutureImpl::DoMarkFinishedOrFailed (this=0x7f18a815f030, state=arrow::FutureState::FAILURE)
          at /arrow/cpp/src/arrow/util/future.cc:125
      #6  0x00007f18c58265af in arrow::ConcreteFutureImpl::DoMarkFailed (this=0x7f18a815f030) at /arrow/cpp/src/arrow/util/future.cc:40
      #7  0x00007f18c5827660 in arrow::FutureImpl::MarkFailed (this=0x7f18a815f030) at /arrow/cpp/src/arrow/util/future.cc:195
      #8  0x00007f18c80ff8d8 in arrow::Future<std::shared_ptr<arrow::flight::transport::ucx::Frame> >::DoMarkFinished (this=0x7f18a815efb0, res=...)
          at /arrow/cpp/src/arrow/util/future.h:660
      #9  0x00007f18c80fb37d in arrow::Future<std::shared_ptr<arrow::flight::transport::ucx::Frame> >::MarkFinished (this=0x7f18a815efb0, res=...)
          at /arrow/cpp/src/arrow/util/future.h:403
      #10 0x00007f18c80f5ae3 in arrow::flight::transport::ucx::UcpCallDriver::Impl::Push (this=0x7f18a804d2d0, status=...)
          at /arrow/cpp/src/arrow/flight/transport/ucx/ucx_internal.cc:780
      #11 0x00007f18c80f5c1f in arrow::flight::transport::ucx::UcpCallDriver::Impl::RecvActiveMessage (this=0x7f18a804d2d0, header=0x7f18c8081865, header_length=12, 
          data=0x7f18c8081864, data_length=1, param=0x7f18c07fc680) at /arrow/cpp/src/arrow/flight/transport/ucx/ucx_internal.cc:791
      #12 0x00007f18c80f7d29 in arrow::flight::transport::ucx::UcpCallDriver::RecvActiveMessage (this=0x7f18b80017e0, header=0x7f18c8081865, header_length=12, 
          data=0x7f18c8081864, data_length=1, param=0x7f18c07fc680) at /arrow/cpp/src/arrow/flight/transport/ucx/ucx_internal.cc:1082
      #13 0x00007f18c80e3ea4 in arrow::flight::transport::ucx::(anonymous namespace)::UcxServerImpl::HandleIncomingActiveMessage (self=0x7f18a80259a0, 
          header=0x7f18c8081865, header_length=12, data=0x7f18c8081864, data_length=1, param=0x7f18c07fc680)
          at /arrow/cpp/src/arrow/flight/transport/ucx/ucx_server.cc:586
      #14 0x00007f18c4661a09 in ucp_am_invoke_cb (recv_flags=<optimized out>, reply_ep=<optimized out>, data_length=1, data=<optimized out>, 
          user_hdr_length=<optimized out>, user_hdr=0x7f18c8081865, am_id=4132, worker=<optimized out>) at core/ucp_am.c:1220
      #15 ucp_am_handler_common (name=<synthetic pointer>, recv_flags=<optimized out>, am_flags=0, reply_ep=<optimized out>, total_length=<optimized out>, 
          am_hdr=0x7f18c808185c, worker=<optimized out>) at core/ucp_am.c:1289
      #16 ucp_am_handler_reply (am_arg=<optimized out>, am_data=<optimized out>, am_length=<optimized out>, am_flags=<optimized out>) at core/ucp_am.c:1327
      #17 0x00007f18c28e3f1c in uct_iface_invoke_am (flags=0, length=29, data=0x7f18c808185c, id=<optimized out>, iface=0x7f18a8027e20)
          at /usr/local/src/conda/ucx-1.13.1/src/uct/base/uct_iface.h:861
      #18 uct_mm_iface_invoke_am (flags=0, length=29, data=0x7f18c808185c, am_id=<optimized out>, iface=0x7f18a8027e20) at sm/mm/base/mm_iface.h:256
      #19 uct_mm_iface_process_recv (iface=0x7f18a8027e20) at sm/mm/base/mm_iface.c:256
      #20 uct_mm_iface_poll_fifo (iface=0x7f18a8027e20) at sm/mm/base/mm_iface.c:304
      #21 uct_mm_iface_progress (tl_iface=0x7f18a8027e20) at sm/mm/base/mm_iface.c:357
      #22 0x00007f18c4686e22 in ucs_callbackq_dispatch (cbq=<optimized out>) at /usr/local/src/conda/ucx-1.13.1/src/ucs/datastruct/callbackq.h:211
      #23 uct_worker_progress (worker=<optimized out>) at /usr/local/src/conda/ucx-1.13.1/src/uct/api/uct.h:2638
      #24 ucp_worker_progress (worker=0x7f18a80008d0) at core/ucp_worker.c:2782
      #25 0x00007f18c80f586f in arrow::flight::transport::ucx::UcpCallDriver::Impl::MakeProgress (this=0x7f18a804d2d0)
          at /arrow/cpp/src/arrow/flight/transport/ucx/ucx_internal.cc:759
      #26 0x00007f18c80f2e40 in arrow::flight::transport::ucx::UcpCallDriver::Impl::ReadNextFrame (this=0x7f18a804d2d0)
          at /arrow/cpp/src/arrow/flight/transport/ucx/ucx_internal.cc:449
      #27 0x00007f18c80f7661 in arrow::flight::transport::ucx::UcpCallDriver::ReadNextFrame (this=0x7f18b80017e0)
          at /arrow/cpp/src/arrow/flight/transport/ucx/ucx_internal.cc:1037
      #28 0x00007f18c80dc7cd in arrow::flight::transport::ucx::(anonymous namespace)::PutServerStream::ReadImpl (this=0x7f18c07fcb60, data=0x7f18c07fcaf0)
          at /arrow/cpp/src/arrow/flight/transport/ucx/ucx_server.cc:140
      #29 0x00007f18c80dc525 in arrow::flight::transport::ucx::(anonymous namespace)::PutServerStream::ReadData (this=0x7f18c07fcb60, data=0x7f18c07fcaf0)
          at /arrow/cpp/src/arrow/flight/transport/ucx/ucx_server.cc:120
      #30 0x00007f18c80e18c9 in arrow::flight::transport::ucx::(anonymous namespace)::UcxServerImpl::HandleDoExchange (this=0x7f18b405c430, driver=0x7f18b80017e0)
          at /arrow/cpp/src/arrow/flight/transport/ucx/ucx_server.cc:414
      #31 0x00007f18c80e2008 in arrow::flight::transport::ucx::(anonymous namespace)::UcxServerImpl::HandleOneCall (this=0x7f18b405c430, driver=0x7f18b80017e0, 
          frame=0x7f18a804df80) at /arrow/cpp/src/arrow/flight/transport/ucx/ucx_server.cc:426
      #32 0x00007f18c80e2d21 in arrow::flight::transport::ucx::(anonymous namespace)::UcxServerImpl::WorkerLoop (this=0x7f18b405c430, request=0x7f18b80016d0)
          at /arrow/cpp/src/arrow/flight/transport/ucx/ucx_server.cc:493
      #33 0x00007f18c80e335e in operator() (__closure=0x7f18b4050c30) at /arrow/cpp/src/arrow/flight/transport/ucx/ucx_server.cc:520
      #34 0x00007f18c80ecb1a in arrow::detail::ContinueFuture::operator()<arrow::flight::transport::ucx::(anonymous namespace)::UcxServerImpl::DriveConnections()::<lambda()>&>(arrow::Future<arrow::internal::Empty>, struct {...} &) const (this=0x7f18b4050c28, next=..., f=...) at /arrow/cpp/src/arrow/util/future.h:133
      #35 0x00007f18c80ec9e1 in std::__invoke_impl<void, arrow::detail::ContinueFuture&, arrow::Future<arrow::internal::Empty>&, arrow::flight::transport::ucx::(anonymous namespace)::UcxServerImpl::DriveConnections()::<lambda()>&>(std::__invoke_other, arrow::detail::ContinueFuture &) (__f=...)
          at /opt/conda/envs/arrow/x86_64-conda-linux-gnu/include/c++/10.4.0/bits/invoke.h:60
      #36 0x00007f18c80ec883 in std::__invoke<arrow::detail::ContinueFuture&, arrow::Future<arrow::internal::Empty>&, arrow::flight::transport::ucx::(anonymous namespace)::UcxServerImpl::DriveConnections()::<lambda()>&>(arrow::detail::ContinueFuture &) (__fn=...)
          at /opt/conda/envs/arrow/x86_64-conda-linux-gnu/include/c++/10.4.0/bits/invoke.h:95
      #37 0x00007f18c80ec71c in std::_Bind<arrow::detail::ContinueFuture(arrow::Future<arrow::internal::Empty>, arrow::flight::transport::ucx::(anonymous namespace)::UcxServerImpl::DriveConnections()::<lambda()>)>::__call<void, 0, 1>(std::tuple<> &&, std::_Index_tuple<0, 1>) (this=0x7f18b4050c28, __args=...)
          at /opt/conda/envs/arrow/x86_64-conda-linux-gnu/include/c++/10.4.0/functional:416
      #38 0x00007f18c80ec5b9 in std::_Bind<arrow::detail::ContinueFuture(arrow::Future<arrow::internal::Empty>, arrow::flight::transport::ucx::(anonymous namespace)::---Type <return> to continue, or q <return> to quit---
      UcxServerImpl::DriveConnections()::<lambda()>)>::operator()<>(void) (this=0x7f18b4050c28)
          at /opt/conda/envs/arrow/x86_64-conda-linux-gnu/include/c++/10.4.0/functional:499
      #39 0x00007f18c80ec440 in arrow::internal::FnOnce<void()>::FnImpl<std::_Bind<arrow::detail::ContinueFuture(arrow::Future<arrow::internal::Empty>, arrow::flight::transport::ucx::(anonymous namespace)::UcxServerImpl::DriveConnections()::<lambda()>)> >::invoke(void) (this=0x7f18b4050c20)
          at /arrow/cpp/src/arrow/util/functional.h:152
      #40 0x00007f18c587331d in arrow::internal::FnOnce<void ()>::operator()() && (this=0x7f18c07fd110) at /arrow/cpp/src/arrow/util/functional.h:140
      #41 0x00007f18c587130f in arrow::internal::WorkerLoop (state=..., it=...) at /arrow/cpp/src/arrow/util/thread_pool.cc:262
      #42 0x00007f18c587227c in operator() (__closure=0x7f18b40517d8) at /arrow/cpp/src/arrow/util/thread_pool.cc:423
      #43 0x00007f18c587b8b6 in std::__invoke_impl<void, arrow::internal::ThreadPool::LaunchWorkersUnlocked(int)::<lambda()> >(std::__invoke_other, struct {...} &&)
          (__f=...) at /opt/conda/envs/arrow/x86_64-conda-linux-gnu/include/c++/10.4.0/bits/invoke.h:60
      #44 0x00007f18c587b86b in std::__invoke<arrow::internal::ThreadPool::LaunchWorkersUnlocked(int)::<lambda()> >(struct {...} &&) (__fn=...)
          at /opt/conda/envs/arrow/x86_64-conda-linux-gnu/include/c++/10.4.0/bits/invoke.h:95
      #45 0x00007f18c587b818 in std::thread::_Invoker<std::tuple<arrow::internal::ThreadPool::LaunchWorkersUnlocked(int)::<lambda()> > >::_M_invoke<0>(std::_Index_tuple<0>) (this=0x7f18b40517d8) at /opt/conda/envs/arrow/x86_64-conda-linux-gnu/include/c++/10.4.0/thread:264
      #46 0x00007f18c587b65c in std::thread::_Invoker<std::tuple<arrow::internal::ThreadPool::LaunchWorkersUnlocked(int)::<lambda()> > >::operator()(void) (
          this=0x7f18b40517d8) at /opt/conda/envs/arrow/x86_64-conda-linux-gnu/include/c++/10.4.0/thread:271
      #47 0x00007f18c587b580 in std::thread::_State_impl<std::thread::_Invoker<std::tuple<arrow::internal::ThreadPool::LaunchWorkersUnlocked(int)::<lambda()> > > >::_M_run(void) (this=0x7f18b40517d0) at /opt/conda/envs/arrow/x86_64-conda-linux-gnu/include/c++/10.4.0/thread:215
      #48 0x00007f18c4e4ba93 in std::execute_native_thread_routine (__p=<optimized out>)
          at /home/conda/feedstock_root/build_artifacts/gcc_compilers_1666516830325/work/build/x86_64-conda-linux-gnu/libstdc++-v3/include/new_allocator.h:82
      #49 0x00007f18c47756db in start_thread (arg=0x7f18c07fe700) at pthread_create.c:463
      #50 0x00007f18c4aae61f in clone () at ../sysdeps/unix/sysv/linux/x86_64/clone.S:95
      

      Attachments

        Issue Links

          Activity

            People

              lidavidm David Li
              apitrou Antoine Pitrou
              Votes:
              0 Vote for this issue
              Watchers:
              3 Start watching this issue

              Dates

                Created:
                Updated:
                Resolved:

                Time Tracking

                  Estimated:
                  Original Estimate - Not Specified
                  Not Specified
                  Remaining:
                  Remaining Estimate - 0h
                  0h
                  Logged:
                  Time Spent - 50m
                  50m