Hello all!
I am trying to build PyTorch-0.3.1 with intelcuda-2017b but I am hitting the
error I reported below. Has anybody seen anything similar before?
/tmp/PyTorch/0.3.1/intelcuda-2017b-Python-3.6.3/pytorch-0.3.1/torch/lib/THD/base/data_channels/Store.cpp(44):
error: no instance of overloaded function "std::vector<_Tp, _Alloc>::push_back
[with _Tp=pollfd, _Alloc=std::allocator<pollfd>]" matches the argument list
argument types are: ({...})
object type is: std::vector<pollfd, std::allocator<pollfd>>
fds.push_back({ .fd = _listen_socket, .events = POLLIN });
^
/opt/easybuild/software/Core/GCCcore/6.4.0/include/c++/6.4.0/bits/stl_vector.h(932):
note: this candidate was rejected because arguments do not match
push_back(value_type&& __x)
^
/opt/easybuild/software/Core/GCCcore/6.4.0/include/c++/6.4.0/bits/stl_vector.h(914):
note: this candidate was rejected because arguments do not match
push_back(const value_type& __x)
^
/tmp/PyTorch/0.3.1/intelcuda-2017b-Python-3.6.3/pytorch-0.3.1/torch/lib/THD/base/data_channels/Store.cpp(44):
internal error: assertion failed at: "shared/cfe/edgcpfe/exprutil.c", line 747
fds.push_back({ .fd = _listen_socket, .events = POLLIN });
^
compilation aborted for
/tmp/PyTorch/0.3.1/intelcuda-2017b-Python-3.6.3/pytorch-0.3.1/torch/lib/THD/base/data_channels/Store.cpp
(code 4)
--
Davide Vanzo, PhD
Sr. HPC Research Scientist
Advanced Computing Center for Research and Education (ACCRE)
Vanderbilt University - Hill Center 201
(615)-875-9137
www.vanderbilt.edu/accre<http://www.vanderbilt.edu/accre>