Model with custom cpu Op will hang when open BladeDISC optimize.

[tuanzhangCS]

Describe the bug
When I run a model with custom cpu ops by DeepRec, it works. But if I use BladeDISC to optimize the performance, the model running will hang.

Thread 1 (Thread 0x7f44a9d05740 (LWP 29790)):
#0  syscall () at ../sysdeps/unix/sysv/linux/x86_64/syscall.S:38
#1  0x00007f44075617f4 in nsync::nsync_mu_semaphore_p_with_deadline(nsync::nsync_semaphore_s_*, timespec) () from /opt/venv_disc/lib/python3.8/site-packages/tensorflow_core/python/_pywrap_tensorflow_internal.so
#2  0x00007f4407560e09 in nsync::nsync_sem_wait_with_cancel_(nsync::waiter*, timespec, nsync::nsync_note_s_*) () from /opt/venv_disc/lib/python3.8/site-packages/tensorflow_core/python/_pywrap_tensorflow_internal.so
#3  0x00007f440755e183 in nsync::nsync_cv_wait_with_deadline_generic(nsync::nsync_cv_s_*, void*, void (*)(void*), void (*)(void*), timespec, nsync::nsync_note_s_*) () from /opt/venv_disc/lib/python3.8/site-packages/tensorflow_core/python/_pywrap_tensorflow_internal.so
#4  0x00007f440755e667 in nsync::nsync_cv_wait_with_deadline(nsync::nsync_cv_s_*, nsync::nsync_mu_s_*, timespec, nsync::nsync_note_s_*) () from /opt/venv_disc/lib/python3.8/site-packages/tensorflow_core/python/_pywrap_tensorflow_internal.so
#5  0x00007f440bc88bbc in tensorflow::DirectSession::WaitForNotification(tensorflow::Notification*, long long) () from /opt/venv_disc/lib/python3.8/site-packages/tensorflow_core/python/_pywrap_tensorflow_internal.so
#6  0x00007f440bc88c8f in tensorflow::DirectSession::WaitForNotification(tensorflow::DirectSession::RunState*, tensorflow::CancellationManager*, long long) () from /opt/venv_disc/lib/python3.8/site-packages/tensorflow_core/python/_pywrap_tensorflow_internal.so
#7  0x00007f440bc95971 in tensorflow::DirectSession::RunInternal(long long, tensorflow::RunOptions const&, tensorflow::CallFrameInterface*, tensorflow::DirectSession::ExecutorsAndKeys*, tensorflow::RunMetadata*, tensorflow::thread::ThreadPoolOptions const&) () from /opt/venv_disc/lib/python3.8/site-packages/tensorflow_core/python/_pywrap_tensorflow_internal.so
#8  0x00007f440bca9cfc in tensorflow::DirectSession::Run(tensorflow::RunOptions const&, std::vector<std::pair<std::string, tensorflow::Tensor>, std::allocator<std::pair<std::string, tensorflow::Tensor> > > const&, std::vector<std::string, std::allocator<std::string> > const&, std::vector<std::string, std::allocator<std::string> > const&, std::vector<tensorflow::Tensor, std::allocator<tensorflow::Tensor> >*, tensorflow::RunMetadata*) () from /opt/venv_disc/lib/python3.8/site-packages/tensorflow_core/python/_pywrap_tensorflow_internal.so
#9  0x00007f4403c2dfb5 in tensorflow::SessionRef::Run(tensorflow::RunOptions const&, std::vector<std::pair<std::string, tensorflow::Tensor>, std::allocator<std::pair<std::string, tensorflow::Tensor> > > const&,

other threads:
Thread 6 (Thread 0x7f449ec98700 (LWP 29799)):
#0  futex_wait_cancelable (private=<optimized out>, expected=0, futex_word=0x7f44a8ca8d60 <thread_status+608>) at ../sysdeps/nptl/futex-internal.h:183
#1  __pthread_cond_wait_common (abstime=0x0, clockid=0, mutex=0x7f44a8ca8d10 <thread_status+528>, cond=0x7f44a8ca8d38 <thread_status+568>) at pthread_cond_wait.c:508
#2  __pthread_cond_wait (cond=0x7f44a8ca8d38 <thread_status+568>, mutex=0x7f44a8ca8d10 <thread_status+528>) at pthread_cond_wait.c:647
#3  0x00007f44a72aa60b in blas_thread_server () from /opt/venv_disc/lib/python3.8/site-packages/numpy/core/../../numpy.libs/libopenblasp-r0-34a18dc3.3.7.so
#4  0x00007f44a9eb6609 in start_thread (arg=<optimized out>) at pthread_create.c:477
#5  0x00007f44a9ff0133 in clone () at ../sysdeps/unix/sysv/linux/x86_64/clone.S:95

Thread 5 (Thread 0x7f44a1499700 (LWP 29798)):
#0  futex_wait_cancelable (private=<optimized out>, expected=0, futex_word=0x7f44a8ca8ce0 <thread_status+480>) at ../sysdeps/nptl/futex-internal.h:183
--Type <RET> for more, q to quit, c to continue without paging--
#1  __pthread_cond_wait_common (abstime=0x0, clockid=0, mutex=0x7f44a8ca8c90 <thread_status+400>, cond=0x7f44a8ca8cb8 <thread_status+440>) at pthread_cond_wait.c:508
#2  __pthread_cond_wait (cond=0x7f44a8ca8cb8 <thread_status+440>, mutex=0x7f44a8ca8c90 <thread_status+400>) at pthread_cond_wait.c:647
#3  0x00007f44a72aa60b in blas_thread_server () from /opt/venv_disc/lib/python3.8/site-packages/numpy/core/../../numpy.libs/libopenblasp-r0-34a18dc3.3.7.so
#4  0x00007f44a9eb6609 in start_thread (arg=<optimized out>) at pthread_create.c:477
#5  0x00007f44a9ff0133 in clone () at ../sysdeps/unix/sysv/linux/x86_64/clone.S:95

Thread 4 (Thread 0x7f44a3c9a700 (LWP 29797)):
#0  futex_wait_cancelable (private=<optimized out>, expected=0, futex_word=0x7f44a8ca8c60 <thread_status+352>) at ../sysdeps/nptl/futex-internal.h:183
#1  __pthread_cond_wait_common (abstime=0x0, clockid=0, mutex=0x7f44a8ca8c10 <thread_status+272>, cond=0x7f44a8ca8c38 <thread_status+312>) at pthread_cond_wait.c:508
#2  __pthread_cond_wait (cond=0x7f44a8ca8c38 <thread_status+312>, mutex=0x7f44a8ca8c10 <thread_status+272>) at pthread_cond_wait.c:647
#3  0x00007f44a72aa60b in blas_thread_server () from /opt/venv_disc/lib/python3.8/site-packages/numpy/core/../../numpy.libs/libopenblasp-r0-34a18dc3.3.7.so
#4  0x00007f44a9eb6609 in start_thread (arg=<optimized out>) at pthread_create.c:477
#5  0x00007f44a9ff0133 in clone () at ../sysdeps/unix/sysv/linux/x86_64/clone.S:95

Thread 3 (Thread 0x7f44a649b700 (LWP 29796)):
#0  futex_wait_cancelable (private=<optimized out>, expected=0, futex_word=0x7f44a8ca8be0 <thread_status+224>) at ../sysdeps/nptl/futex-internal.h:183
#1  __pthread_cond_wait_common (abstime=0x0, clockid=0, mutex=0x7f44a8ca8b90 <thread_status+144>, cond=0x7f44a8ca8bb8 <thread_status+184>) at pthread_cond_wait.c:508
#2  __pthread_cond_wait (cond=0x7f44a8ca8bb8 <thread_status+184>, mutex=0x7f44a8ca8b90 <thread_status+144>) at pthread_cond_wait.c:647
#3  0x00007f44a72aa60b in blas_thread_server () from /opt/venv_disc/lib/python3.8/site-packages/numpy/core/../../numpy.libs/libopenblasp-r0-34a18dc3.3.7.so
#4  0x00007f44a9eb6609 in start_thread (arg=<optimized out>) at pthread_create.c:477
#5  0x00007f44a9ff0133 in clone () at ../sysdeps/unix/sysv/linux/x86_64/clone.S:95

Thread 2 (Thread 0x7f44a6c9c700 (LWP 29795)):
#0  futex_wait_cancelable (private=<optimized out>, expected=0, futex_word=0x7f44a8ca8b60 <thread_status+96>) at ../sysdeps/nptl/futex-internal.h:183
#1  __pthread_cond_wait_common (abstime=0x0, clockid=0, mutex=0x7f44a8ca8b10 <thread_status+16>, cond=0x7f44a8ca8b38 <thread_status+56>) at pthread_cond_wait.c:508
#2  __pthread_cond_wait (cond=0x7f44a8ca8b38 <thread_status+56>, mutex=0x7f44a8ca8b10 <thread_status+16>) at pthread_cond_wait.c:647
#3  0x00007f44a72aa60b in blas_thread_server () from /opt/venv_disc/lib/python3.8/site-packages/numpy/core/../../numpy.libs/libopenblasp-r0-34a18dc3.3.7.so
#4  0x00007f44a9eb6609 in start_thread (arg=<optimized out>) at pthread_create.c:477
#5  0x00007f44a9ff0133 in clone () at ../sysdeps/unix/sysv/linux/x86_64/clone.S:95

To Reproduce

import blade_disc_tf as disc
disc.enable()

I only add these two lines to my code, then it will hang.

BladeDISC branch: features/deeprec2208-cu114
DeepRec: commit be62ec312595b51b74260f96a6c0872ce5f1540c (HEAD -> main, origin/main, origin/HEAD)