Antoine Pitrou created ARROW-18351:
--------------------------------------

             Summary: [C++][Flight] Crash in UcxErrorHandlingTest.TestDoExchange
                 Key: ARROW-18351
                 URL: https://issues.apache.org/jira/browse/ARROW-18351
             Project: Apache Arrow
          Issue Type: Bug
          Components: C++, FlightRPC
            Reporter: Antoine Pitrou


I get a non-deterministic crash in the Flight UCX tests.
{code}
[----------] 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
{code}

Here is the GDB backtrace:
{code}
#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
{code}




--
This message was sent by Atlassian Jira
(v8.20.10#820010)

Reply via email to