Insatllation Strucked at 'lotd'

[Harvey-Mei]

Very thanks for your job！ I got stuck while compiling LOTD, and there are already several issues the same with me. @ventusff Do you have time to help me take a look and give some suggestion? I have tested that using version V0.5.1 can compile successfully, but the commit point used in neuralsim is e1e87d1. If not checkout to this commit point, an error will be reported when running neuralsim.

see log below:

building 'nr3d_lib.bindings._lotd' extension creating /home/may/work/code/neuralsim/nr3d_lib/build/temp.linux-x86_64-cpython-38/csrc/lotd creating /home/may/work/code/neuralsim/nr3d_lib/build/temp.linux-x86_64-cpython-38/csrc/lotd/src Emitting ninja build file /home/may/work/code/neuralsim/nr3d_lib/build/temp.linux-x86_64-cpython-38/build.ninja... Compiling objects... Allowing ninja to set a default number of workers... (overridable by setting the environment variable MAX_JOBS=N) [1/5] c++ -MMD -MF /home/may/work/code/neuralsim/nr3d_lib/build/temp.linux-x86_64-cpython-38/csrc/lotd/src/lotd.o.d -pthread -B /home/may/anaconda3/envs/neuralsim/compiler_compat -Wl,--sysroot=/ -Wsign-compare -DNDEBUG -g -fwrapv -O3 -Wall -Wstrict-prototypes -fPIC -I/home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/include -I/home/may/work/code/neuralsim/nr3d_lib/csrc/forest -I/home/may/anaconda3/envs/neuralsim/lib/python3.8/site-packages/torch/include -I/home/may/anaconda3/envs/neuralsim/lib/python3.8/site-packages/torch/include/torch/csrc/api/include -I/home/may/anaconda3/envs/neuralsim/lib/python3.8/site-packages/torch/include/TH -I/home/may/anaconda3/envs/neuralsim/lib/python3.8/site-packages/torch/include/THC -I/usr/local/cuda-11.3/include -I/home/may/anaconda3/envs/neuralsim/include/python3.8 -c -c /home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/src/lotd.cpp -o /home/may/work/code/neuralsim/nr3d_lib/build/temp.linux-x86_64-cpython-38/csrc/lotd/src/lotd.o -std=c++14 -O3 -DNDEBUG -DTORCH_API_INCLUDE_EXTENSION_H '-DPYBIND11_COMPILER_TYPE="_gcc"' '-DPYBIND11_STDLIB="_libstdcpp"' '-DPYBIND11_BUILD_ABI="_cxxabi1011"' -DTORCH_EXTENSION_NAME=_lotd -D_GLIBCXX_USE_CXX11_ABI=0 cc1plus: warning: command line option ‘-Wstrict-prototypes’ is valid for C/ObjC but not for C++ [2/5] /usr/local/cuda-11.3/bin/nvcc -I/home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/include -I/home/may/work/code/neuralsim/nr3d_lib/csrc/forest -I/home/may/anaconda3/envs/neuralsim/lib/python3.8/site-packages/torch/include -I/home/may/anaconda3/envs/neuralsim/lib/python3.8/site-packages/torch/include/torch/csrc/api/include -I/home/may/anaconda3/envs/neuralsim/lib/python3.8/site-packages/torch/include/TH -I/home/may/anaconda3/envs/neuralsim/lib/python3.8/site-packages/torch/include/THC -I/usr/local/cuda-11.3/include -I/home/may/anaconda3/envs/neuralsim/include/python3.8 -c -c /home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/src/compile_split_1.cu -o /home/may/work/code/neuralsim/nr3d_lib/build/temp.linux-x86_64-cpython-38/csrc/lotd/src/compile_split_1.o -DCUDA_NO_HALF_OPERATORS -DCUDA_NO_HALF_CONVERSIONS -DCUDA_NO_BFLOAT16_CONVERSIONS -DCUDA_NO_HALF2_OPERATORS --expt-relaxed-constexpr --compiler-options ''"'"'-fPIC'"'"'' -std=c++14 --extended-lambda --expt-relaxed-constexpr -UCUDA_NO_HALF_OPERATORS -UCUDA_NO_HALF2_OPERATORS -UCUDA_NO_HALF_CONVERSIONS -UCUDA_NO_BFLOAT16_CONVERSIONS -gencode=arch=compute_86,code=compute_86 -gencode=arch=compute_86,code=sm_86 -Xcompiler=-mf16c -Xcompiler=-Wno-float-conversion -Xcompiler=-fno-strict-aliasing -Xcudafe=--diag_suppress=unrecognized_gcc_pragma -O3 -DNDEBUG -DTORCH_API_INCLUDE_EXTENSION_H '-DPYBIND11_COMPILER_TYPE="_gcc"' '-DPYBIND11_STDLIB="_libstdcpp"' '-DPYBIND11_BUILD_ABI="_cxxabi1011"' -DTORCH_EXTENSION_NAME=_lotd -D_GLIBCXX_USE_CXX11_ABI=0 /home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/include/lotd/lotd_cuda.h(885): warning: missing return statement at end of non-void function "lotd::calc_dLdx_dim_vm_impl<PARAM_T,COMPUTE_T,N_POS_DIMS,N_FEAT>(const uint32_t , uint32_t, const uint32_t , uint32_t, uint32_t, const PARAM_T , const PARAM_T , COMPUTE_T) [with PARAM_T=float, COMPUTE_T=float, N_POS_DIMS=2U, N_FEAT=2U]" detected during: instantiation of "std::enable_if<, COMPUTE_T>::type lotd::calc_dLdx_dim_vm_impl<PARAM_T,COMPUTE_T,N_POS_DIMS,N_FEAT>(const uint32_t , uint32_t, const uint32_t , uint32_t, uint32_t, const PARAM_T , const PARAM_T , COMPUTE_T) [with PARAM_T=float, COMPUTE_T=float, N_POS_DIMS=2U, N_FEAT=2U]" /home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/include/lotd/lotd_encoding.h(1260): here instantiation of "std::enable_if<std::is_same<INPUT_T, float>::value, void>::type lotd::torch::kernel_lod_backward_input_backward_input<INPUT_T,PARAM_T,COMPUTE_T,N_POS_DIMS,N_FEAT_PER_PSEUDO_LVL,N_FEAT_PER_THREAD>(uint32_t, lotd::LoDMetaRef, int32_t, const int32_t , const INPUT_T , const PARAM_T , const PARAM_T , const INPUT_T , const int64_t , const int64_t , uint32_t, INPUT_T ) [with INPUT_T=float, PARAM_T=float, COMPUTE_T=float, N_POS_DIMS=2U, N_FEAT_PER_PSEUDO_LVL=8U, N_FEAT_PER_THREAD=2U]" /home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/include/lotd/lotd_encoding.h(1746): here instantiation of "void lotd::torch::lod_bwd_bwd_input_impl_dispatched<INPUT_T,PARAM_T,COMPUTE_T,N_POS_DIMS,N_FEAT_PER_PSEUDO_LVL>(lotd::torch::LoDMeta &, at::Tensor, at::Tensor, at::Tensor, at::Tensor, c10::optional, c10::optional, c10::optional, uint32_t, int32_t, nv_bool, __nv_bool, nv_bool, at::Tensor, at::Tensor, at::Tensor) [with INPUT_T=float, PARAM_T=float, COMPUTE_T=float, N_POS_DIMS=2U, N_FEAT_PER_PSEUDO_LVL=8U]" /home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/include/lotd/lotd_encoding.h(1807): here instantiation of "void lotd::torch::lod_bwd_bwd_input_impl_templated<N_POS_DIMS,N_FEAT_PER_PSEUDO_LVL>(lotd::torch::LoDMeta &, at::Tensor, at::Tensor, at::Tensor, at::Tensor, c10::optional, c10::optional, c10::optional, uint32_t, int32_t, nv_bool, __nv_bool, nv_bool, at::Tensor, at::Tensor, at::Tensor) [with N_POS_DIMS=2U, N_FEAT_PER_PSEUDO_LVL=8U]" /home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/include/lotd/lotd_encoding.h(1836): here instantiation of "void lotd::torch::lod_bwd_bwd_input_impl(lotd::torch::LoDMeta &, at::Tensor, at::Tensor, at::Tensor, at::Tensor, c10::optional, c10::optional, c10::optional, uint32_t, int32_t, nv_bool, __nv_bool, nv_bool, at::Tensor, at::Tensor, at::Tensor) [with N_POS_DIMS=2U]" /home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/src/compile_split_1.cu(30): here

/home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/include/lotd/lotd_cuda.h(934): warning: missing return statement at end of non-void function "lotd::calc_dLdx_dim_vec_z_mat_xoy_impl<PARAM_T,COMPUTE_T,N_POS_DIMS,N_FEAT>(const uint32_t , uint32_t, const uint32_t , uint32_t, uint32_t, const PARAM_T , const PARAM_T , COMPUTE_T) [with PARAM_T=float, COMPUTE_T=float, N_POS_DIMS=2U, N_FEAT=2U]" detected during: instantiation of "std::enable_if<, COMPUTE_T>::type lotd::calc_dLdx_dim_vec_z_mat_xoy_impl<PARAM_T,COMPUTE_T,N_POS_DIMS,N_FEAT>(const uint32_t , uint32_t, const uint32_t , uint32_t, uint32_t, const PARAM_T , const PARAM_T , COMPUTE_T) [with PARAM_T=float, COMPUTE_T=float, N_POS_DIMS=2U, N_FEAT=2U]" /home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/include/lotd/lotd_encoding.h(1271): here instantiation of "std::enable_if<std::is_same<INPUT_T, float>::value, void>::type lotd::torch::kernel_lod_backward_input_backward_input<INPUT_T,PARAM_T,COMPUTE_T,N_POS_DIMS,N_FEAT_PER_PSEUDO_LVL,N_FEAT_PER_THREAD>(uint32_t, lotd::LoDMetaRef, int32_t, const int32_t , const INPUT_T , const PARAM_T , const PARAM_T , const INPUT_T , const int64_t , const int64_t , uint32_t, INPUT_T ) [with INPUT_T=float, PARAM_T=float, COMPUTE_T=float, N_POS_DIMS=2U, N_FEAT_PER_PSEUDO_LVL=8U, N_FEAT_PER_THREAD=2U]" /home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/include/lotd/lotd_encoding.h(1746): here instantiation of "void lotd::torch::lod_bwd_bwd_input_impl_dispatched<INPUT_T,PARAM_T,COMPUTE_T,N_POS_DIMS,N_FEAT_PER_PSEUDO_LVL>(lotd::torch::LoDMeta &, at::Tensor, at::Tensor, at::Tensor, at::Tensor, c10::optional, c10::optional, c10::optional, uint32_t, int32_t, nv_bool, __nv_bool, nv_bool, at::Tensor, at::Tensor, at::Tensor) [with INPUT_T=float, PARAM_T=float, COMPUTE_T=float, N_POS_DIMS=2U, N_FEAT_PER_PSEUDO_LVL=8U]" /home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/include/lotd/lotd_encoding.h(1807): here instantiation of "void lotd::torch::lod_bwd_bwd_input_impl_templated<N_POS_DIMS,N_FEAT_PER_PSEUDO_LVL>(lotd::torch::LoDMeta &, at::Tensor, at::Tensor, at::Tensor, at::Tensor, c10::optional, c10::optional, c10::optional, uint32_t, int32_t, nv_bool, __nv_bool, nv_bool, at::Tensor, at::Tensor, at::Tensor) [with N_POS_DIMS=2U, N_FEAT_PER_PSEUDO_LVL=8U]" /home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/include/lotd/lotd_encoding.h(1836): here instantiation of "void lotd::torch::lod_bwd_bwd_input_impl(lotd::torch::LoDMeta &, at::Tensor, at::Tensor, at::Tensor, at::Tensor, c10::optional, c10::optional, c10::optional, uint32_t, int32_t, nv_bool, __nv_bool, nv_bool, at::Tensor, at::Tensor, at::Tensor) [with N_POS_DIMS=2U]" /home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/src/compile_split_1.cu(30): here

/home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/include/lotd/lotd_cuda.h(885): warning: missing return statement at end of non-void function "lotd::calc_dLdx_dim_vm_impl<PARAM_T,COMPUTE_T,N_POS_DIMS,N_FEAT>(const uint32_t , uint32_t, const uint32_t , uint32_t, uint32_t, const PARAM_T , const PARAM_T , COMPUTE_T) [with PARAM_T=half, COMPUTE_T=float, N_POS_DIMS=2U, N_FEAT=2U]" detected during: instantiation of "std::enable_if<, COMPUTE_T>::type lotd::calc_dLdx_dim_vm_impl<PARAM_T,COMPUTE_T,N_POS_DIMS,N_FEAT>(const uint32_t , uint32_t, const uint32_t , uint32_t, uint32_t, const PARAM_T , const PARAM_T , COMPUTE_T) [with PARAM_T=half, COMPUTE_T=float, N_POS_DIMS=2U, N_FEAT=2U]" /home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/include/lotd/lotd_encoding.h(1260): here instantiation of "std::enable_if<std::is_same<INPUT_T, float>::value, void>::type lotd::torch::kernel_lod_backward_input_backward_input<INPUT_T,PARAM_T,COMPUTE_T,N_POS_DIMS,N_FEAT_PER_PSEUDO_LVL,N_FEAT_PER_THREAD>(uint32_t, lotd::LoDMetaRef, int32_t, const int32_t , const INPUT_T , const PARAM_T , const PARAM_T , const INPUT_T , const int64_t , const int64_t , uint32_t, INPUT_T ) [with INPUT_T=float, PARAM_T=half, COMPUTE_T=float, N_POS_DIMS=2U, N_FEAT_PER_PSEUDO_LVL=8U, N_FEAT_PER_THREAD=2U]" /home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/include/lotd/lotd_encoding.h(1746): here instantiation of "void lotd::torch::lod_bwd_bwd_input_impl_dispatched<INPUT_T,PARAM_T,COMPUTE_T,N_POS_DIMS,N_FEAT_PER_PSEUDO_LVL>(lotd::torch::LoDMeta &, at::Tensor, at::Tensor, at::Tensor, at::Tensor, c10::optional, c10::optional, c10::optional, uint32_t, int32_t, nv_bool, nv_bool, __nv_bool, at::Tensor, at::Tensor, at::Tensor) [with INPUT_T=float, PARAM_T=half, COMPUTE_T=float, N_POS_DIMS=2U, N_FEAT_PER_PSEUDO_LVL=8U]" /home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/include/lotd/lotd_encoding.h(1807): here instantiation of "void lotd::torch::lod_bwd_bwd_input_impl_templated<N_POS_DIMS,N_FEAT_PER_PSEUDO_LVL>(lotd::torch::LoDMeta &, at::Tensor, at::Tensor, at::Tensor, at::Tensor, c10::optional, c10::optional, c10::optional, uint32_t, int32_t, nv_bool, __nv_bool, nv_bool, at::Tensor, at::Tensor, at::Tensor) [with N_POS_DIMS=2U, N_FEAT_PER_PSEUDO_LVL=8U]" /home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/include/lotd/lotd_encoding.h(1836): here instantiation of "void lotd::torch::lod_bwd_bwd_input_impl(lotd::torch::LoDMeta &, at::Tensor, at::Tensor, at::Tensor, at::Tensor, c10::optional, c10::optional, c10::optional, uint32_t, int32_t, nv_bool, __nv_bool, nv_bool, at::Tensor, at::Tensor, at::Tensor) [with N_POS_DIMS=2U]" /home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/src/compile_split_1.cu(30): here

/home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/include/lotd/lotd_cuda.h(934): warning: missing return statement at end of non-void function "lotd::calc_dLdx_dim_vec_z_mat_xoy_impl<PARAM_T,COMPUTE_T,N_POS_DIMS,N_FEAT>(const uint32_t , uint32_t, const uint32_t , uint32_t, uint32_t, const PARAM_T , const PARAM_T , COMPUTE_T) [with PARAM_T=half, COMPUTE_T=float, N_POS_DIMS=2U, N_FEAT=2U]" detected during: instantiation of "std::enable_if<, COMPUTE_T>::type lotd::calc_dLdx_dim_vec_z_mat_xoy_impl<PARAM_T,COMPUTE_T,N_POS_DIMS,N_FEAT>(const uint32_t , uint32_t, const uint32_t , uint32_t, uint32_t, const PARAM_T , const PARAM_T , COMPUTE_T) [with PARAM_T=__half, COMPUTE_T=float, N_POS_DIMS=2U, N_FEAT=2U]" /home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/include/lotd/lotd_encoding.h(1271): here instantiation of "std::enable_if<std::is_same<INPUT_T, float>::value, void>::type lotd::torch::kernel_lod_backward_input_backward_input<INPUT_T,PARAM_T,COMPUTE_T,N_POS_DIMS,N_FEAT_PER_PSEUDO_LVL,N_FEAT_PER_THREAD>(uint32_t, lotd::LoDMetaRef, int32_t, const int32_t , const INPUT_T , const PARAM_T , const PARAM_T , const INPUT_T , const int64_t , const int64_t , uint32_t, INPUT_T ) [with INPUT_T=float, PARAM_T=half, COMPUTE_T=float, N_POS_DIMS=2U, N_FEAT_PER_PSEUDO_LVL=8U, N_FEAT_PER_THREAD=2U]" /home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/include/lotd/lotd_encoding.h(1746): here instantiation of "void lotd::torch::lod_bwd_bwd_input_impl_dispatched<INPUT_T,PARAM_T,COMPUTE_T,N_POS_DIMS,N_FEAT_PER_PSEUDO_LVL>(lotd::torch::LoDMeta &, at::Tensor, at::Tensor, at::Tensor, at::Tensor, c10::optional, c10::optional, c10::optional, uint32_t, int32_t, nv_bool, __nv_bool, nv_bool, at::Tensor, at::Tensor, at::Tensor) [with INPUT_T=float, PARAM_T=half, COMPUTE_T=float, N_POS_DIMS=2U, N_FEAT_PER_PSEUDO_LVL=8U]" /home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/include/lotd/lotd_encoding.h(1807): here instantiation of "void lotd::torch::lod_bwd_bwd_input_impl_templated<N_POS_DIMS,N_FEAT_PER_PSEUDO_LVL>(lotd::torch::LoDMeta &, at::Tensor, at::Tensor, at::Tensor, at::Tensor, c10::optional, c10::optional, c10::optional, uint32_t, int32_t, nv_bool, nv_bool, __nv_bool, at::Tensor, at::Tensor, at::Tensor) [with N_POS_DIMS=2U, N_FEAT_PER_PSEUDO_LVL=8U]" /home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/include/lotd/lotd_encoding.h(1836): here instantiation of "void lotd::torch::lod_bwd_bwd_input_impl(lotd::torch::LoDMeta &, at::Tensor, at::Tensor, at::Tensor, at::Tensor, c10::optional, c10::optional, c10::optional, uint32_t, int32_t, nv_bool, __nv_bool, __nv_bool, at::Tensor, at::Tensor, at::Tensor) [with N_POS_DIMS=2U]" /home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/src/compile_split_1.cu(30): here

[3/5] /usr/local/cuda-11.3/bin/nvcc -I/home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/include -I/home/may/work/code/neuralsim/nr3d_lib/csrc/forest -I/home/may/anaconda3/envs/neuralsim/lib/python3.8/site-packages/torch/include -I/home/may/anaconda3/envs/neuralsim/lib/python3.8/site-packages/torch/include/torch/csrc/api/include -I/home/may/anaconda3/envs/neuralsim/lib/python3.8/site-packages/torch/include/TH -I/home/may/anaconda3/envs/neuralsim/lib/python3.8/site-packages/torch/include/THC -I/usr/local/cuda-11.3/include -I/home/may/anaconda3/envs/neuralsim/include/python3.8 -c -c /home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/src/compile_split_2.cu -o /home/may/work/code/neuralsim/nr3d_lib/build/temp.linux-x86_64-cpython-38/csrc/lotd/src/compile_split_2.o -DCUDA_NO_HALF_OPERATORS -DCUDA_NO_HALF_CONVERSIONS -DCUDA_NO_BFLOAT16_CONVERSIONS -DCUDA_NO_HALF2_OPERATORS --expt-relaxed-constexpr --compiler-options ''"'"'-fPIC'"'"'' -std=c++14 --extended-lambda --expt-relaxed-constexpr -UCUDA_NO_HALF_OPERATORS -UCUDA_NO_HALF2_OPERATORS -UCUDA_NO_HALF_CONVERSIONS -UCUDA_NO_BFLOAT16_CONVERSIONS -gencode=arch=compute_86,code=compute_86 -gencode=arch=compute_86,code=sm_86 -Xcompiler=-mf16c -Xcompiler=-Wno-float-conversion -Xcompiler=-fno-strict-aliasing -Xcudafe=--diag_suppress=unrecognized_gcc_pragma -O3 -DNDEBUG -DTORCH_API_INCLUDE_EXTENSION_H '-DPYBIND11_COMPILER_TYPE="_gcc"' '-DPYBIND11_STDLIB="_libstdcpp"' '-DPYBIND11_BUILD_ABI="_cxxabi1011"' -DTORCH_EXTENSION_NAME=_lotd -D_GLIBCXX_USE_CXX11_ABI=0 [4/5] /usr/local/cuda-11.3/bin/nvcc -I/home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/include -I/home/may/work/code/neuralsim/nr3d_lib/csrc/forest -I/home/may/anaconda3/envs/neuralsim/lib/python3.8/site-packages/torch/include -I/home/may/anaconda3/envs/neuralsim/lib/python3.8/site-packages/torch/include/torch/csrc/api/include -I/home/may/anaconda3/envs/neuralsim/lib/python3.8/site-packages/torch/include/TH -I/home/may/anaconda3/envs/neuralsim/lib/python3.8/site-packages/torch/include/THC -I/usr/local/cuda-11.3/include -I/home/may/anaconda3/envs/neuralsim/include/python3.8 -c -c /home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/src/compile_split_3.cu -o /home/may/work/code/neuralsim/nr3d_lib/build/temp.linux-x86_64-cpython-38/csrc/lotd/src/compile_split_3.o -DCUDA_NO_HALF_OPERATORS -DCUDA_NO_HALF_CONVERSIONS -DCUDA_NO_BFLOAT16_CONVERSIONS -DCUDA_NO_HALF2_OPERATORS --expt-relaxed-constexpr --compiler-options ''"'"'-fPIC'"'"'' -std=c++14 --extended-lambda --expt-relaxed-constexpr -UCUDA_NO_HALF_OPERATORS -UCUDA_NO_HALF2_OPERATORS -UCUDA_NO_HALF_CONVERSIONS -UCUDA_NO_BFLOAT16_CONVERSIONS -gencode=arch=compute_86,code=compute_86 -gencode=arch=compute_86,code=sm_86 -Xcompiler=-mf16c -Xcompiler=-Wno-float-conversion -Xcompiler=-fno-strict-aliasing -Xcudafe=--diag_suppress=unrecognized_gcc_pragma -O3 -DNDEBUG -DTORCH_API_INCLUDE_EXTENSION_H '-DPYBIND11_COMPILER_TYPE="_gcc"' '-DPYBIND11_STDLIB="_libstdcpp"' '-DPYBIND11_BUILD_ABI="_cxxabi1011"' -DTORCH_EXTENSION_NAME=_lotd -D_GLIBCXX_USE_CXX11_ABI=0 /home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/include/lotd/lotd_cuda.h(885): warning: missing return statement at end of non-void function "lotd::calc_dLdx_dim_vm_impl<PARAM_T,COMPUTE_T,N_POS_DIMS,N_FEAT>(const uint32_t , uint32_t, const uint32_t , uint32_t, uint32_t, const PARAM_T , const PARAM_T , COMPUTE_T) [with PARAM_T=float, COMPUTE_T=float, N_POS_DIMS=4U, N_FEAT=2U]" detected during: instantiation of "std::enable_if<, COMPUTE_T>::type lotd::calc_dLdx_dim_vm_impl<PARAM_T,COMPUTE_T,N_POS_DIMS,N_FEAT>(const uint32_t , uint32_t, const uint32_t , uint32_t, uint32_t, const PARAM_T , const PARAM_T , COMPUTE_T) [with PARAM_T=float, COMPUTE_T=float, N_POS_DIMS=4U, N_FEAT=2U]" /home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/include/lotd/lotd_encoding.h(1260): here instantiation of "std::enable_if<std::is_same<INPUT_T, float>::value, void>::type lotd::torch::kernel_lod_backward_input_backward_input<INPUT_T,PARAM_T,COMPUTE_T,N_POS_DIMS,N_FEAT_PER_PSEUDO_LVL,N_FEAT_PER_THREAD>(uint32_t, lotd::LoDMetaRef, int32_t, const int32_t , const INPUT_T , const PARAM_T , const PARAM_T , const INPUT_T , const int64_t , const int64_t , uint32_t, INPUT_T ) [with INPUT_T=float, PARAM_T=float, COMPUTE_T=float, N_POS_DIMS=4U, N_FEAT_PER_PSEUDO_LVL=8U, N_FEAT_PER_THREAD=2U]" /home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/include/lotd/lotd_encoding.h(1746): here instantiation of "void lotd::torch::lod_bwd_bwd_input_impl_dispatched<INPUT_T,PARAM_T,COMPUTE_T,N_POS_DIMS,N_FEAT_PER_PSEUDO_LVL>(lotd::torch::LoDMeta &, at::Tensor, at::Tensor, at::Tensor, at::Tensor, c10::optional, c10::optional, c10::optional, uint32_t, int32_t, nv_bool, __nv_bool, nv_bool, at::Tensor, at::Tensor, at::Tensor) [with INPUT_T=float, PARAM_T=float, COMPUTE_T=float, N_POS_DIMS=4U, N_FEAT_PER_PSEUDO_LVL=8U]" /home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/include/lotd/lotd_encoding.h(1807): here instantiation of "void lotd::torch::lod_bwd_bwd_input_impl_templated<N_POS_DIMS,N_FEAT_PER_PSEUDO_LVL>(lotd::torch::LoDMeta &, at::Tensor, at::Tensor, at::Tensor, at::Tensor, c10::optional, c10::optional, c10::optional, uint32_t, int32_t, nv_bool, __nv_bool, nv_bool, at::Tensor, at::Tensor, at::Tensor) [with N_POS_DIMS=4U, N_FEAT_PER_PSEUDO_LVL=8U]" /home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/include/lotd/lotd_encoding.h(1836): here instantiation of "void lotd::torch::lod_bwd_bwd_input_impl(lotd::torch::LoDMeta &, at::Tensor, at::Tensor, at::Tensor, at::Tensor, c10::optional, c10::optional, c10::optional, uint32_t, int32_t, nv_bool, __nv_bool, nv_bool, at::Tensor, at::Tensor, at::Tensor) [with N_POS_DIMS=4U]" /home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/src/compile_split_3.cu(30): here

/home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/include/lotd/lotd_cuda.h(934): warning: missing return statement at end of non-void function "lotd::calc_dLdx_dim_vec_z_mat_xoy_impl<PARAM_T,COMPUTE_T,N_POS_DIMS,N_FEAT>(const uint32_t , uint32_t, const uint32_t , uint32_t, uint32_t, const PARAM_T , const PARAM_T , COMPUTE_T) [with PARAM_T=float, COMPUTE_T=float, N_POS_DIMS=4U, N_FEAT=2U]" detected during: instantiation of "std::enable_if<, COMPUTE_T>::type lotd::calc_dLdx_dim_vec_z_mat_xoy_impl<PARAM_T,COMPUTE_T,N_POS_DIMS,N_FEAT>(const uint32_t , uint32_t, const uint32_t , uint32_t, uint32_t, const PARAM_T , const PARAM_T , COMPUTE_T) [with PARAM_T=float, COMPUTE_T=float, N_POS_DIMS=4U, N_FEAT=2U]" /home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/include/lotd/lotd_encoding.h(1271): here instantiation of "std::enable_if<std::is_same<INPUT_T, float>::value, void>::type lotd::torch::kernel_lod_backward_input_backward_input<INPUT_T,PARAM_T,COMPUTE_T,N_POS_DIMS,N_FEAT_PER_PSEUDO_LVL,N_FEAT_PER_THREAD>(uint32_t, lotd::LoDMetaRef, int32_t, const int32_t , const INPUT_T , const PARAM_T , const PARAM_T , const INPUT_T , const int64_t , const int64_t , uint32_t, INPUT_T ) [with INPUT_T=float, PARAM_T=float, COMPUTE_T=float, N_POS_DIMS=4U, N_FEAT_PER_PSEUDO_LVL=8U, N_FEAT_PER_THREAD=2U]" /home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/include/lotd/lotd_encoding.h(1746): here instantiation of "void lotd::torch::lod_bwd_bwd_input_impl_dispatched<INPUT_T,PARAM_T,COMPUTE_T,N_POS_DIMS,N_FEAT_PER_PSEUDO_LVL>(lotd::torch::LoDMeta &, at::Tensor, at::Tensor, at::Tensor, at::Tensor, c10::optional, c10::optional, c10::optional, uint32_t, int32_t, nv_bool, __nv_bool, nv_bool, at::Tensor, at::Tensor, at::Tensor) [with INPUT_T=float, PARAM_T=float, COMPUTE_T=float, N_POS_DIMS=4U, N_FEAT_PER_PSEUDO_LVL=8U]" /home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/include/lotd/lotd_encoding.h(1807): here instantiation of "void lotd::torch::lod_bwd_bwd_input_impl_templated<N_POS_DIMS,N_FEAT_PER_PSEUDO_LVL>(lotd::torch::LoDMeta &, at::Tensor, at::Tensor, at::Tensor, at::Tensor, c10::optional, c10::optional, c10::optional, uint32_t, int32_t, nv_bool, __nv_bool, nv_bool, at::Tensor, at::Tensor, at::Tensor) [with N_POS_DIMS=4U, N_FEAT_PER_PSEUDO_LVL=8U]" /home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/include/lotd/lotd_encoding.h(1836): here instantiation of "void lotd::torch::lod_bwd_bwd_input_impl(lotd::torch::LoDMeta &, at::Tensor, at::Tensor, at::Tensor, at::Tensor, c10::optional, c10::optional, c10::optional, uint32_t, int32_t, nv_bool, __nv_bool, nv_bool, at::Tensor, at::Tensor, at::Tensor) [with N_POS_DIMS=4U]" /home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/src/compile_split_3.cu(30): here

/home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/include/lotd/lotd_cuda.h(885): warning: missing return statement at end of non-void function "lotd::calc_dLdx_dim_vm_impl<PARAM_T,COMPUTE_T,N_POS_DIMS,N_FEAT>(const uint32_t , uint32_t, const uint32_t , uint32_t, uint32_t, const PARAM_T , const PARAM_T , COMPUTE_T) [with PARAM_T=half, COMPUTE_T=float, N_POS_DIMS=4U, N_FEAT=2U]" detected during: instantiation of "std::enable_if<, COMPUTE_T>::type lotd::calc_dLdx_dim_vm_impl<PARAM_T,COMPUTE_T,N_POS_DIMS,N_FEAT>(const uint32_t , uint32_t, const uint32_t , uint32_t, uint32_t, const PARAM_T , const PARAM_T , COMPUTE_T) [with PARAM_T=half, COMPUTE_T=float, N_POS_DIMS=4U, N_FEAT=2U]" /home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/include/lotd/lotd_encoding.h(1260): here instantiation of "std::enable_if<std::is_same<INPUT_T, float>::value, void>::type lotd::torch::kernel_lod_backward_input_backward_input<INPUT_T,PARAM_T,COMPUTE_T,N_POS_DIMS,N_FEAT_PER_PSEUDO_LVL,N_FEAT_PER_THREAD>(uint32_t, lotd::LoDMetaRef, int32_t, const int32_t , const INPUT_T , const PARAM_T , const PARAM_T , const INPUT_T , const int64_t , const int64_t , uint32_t, INPUT_T ) [with INPUT_T=float, PARAM_T=half, COMPUTE_T=float, N_POS_DIMS=4U, N_FEAT_PER_PSEUDO_LVL=8U, N_FEAT_PER_THREAD=2U]" /home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/include/lotd/lotd_encoding.h(1746): here instantiation of "void lotd::torch::lod_bwd_bwd_input_impl_dispatched<INPUT_T,PARAM_T,COMPUTE_T,N_POS_DIMS,N_FEAT_PER_PSEUDO_LVL>(lotd::torch::LoDMeta &, at::Tensor, at::Tensor, at::Tensor, at::Tensor, c10::optional, c10::optional, c10::optional, uint32_t, int32_t, nv_bool, nv_bool, __nv_bool, at::Tensor, at::Tensor, at::Tensor) [with INPUT_T=float, PARAM_T=half, COMPUTE_T=float, N_POS_DIMS=4U, N_FEAT_PER_PSEUDO_LVL=8U]" /home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/include/lotd/lotd_encoding.h(1807): here instantiation of "void lotd::torch::lod_bwd_bwd_input_impl_templated<N_POS_DIMS,N_FEAT_PER_PSEUDO_LVL>(lotd::torch::LoDMeta &, at::Tensor, at::Tensor, at::Tensor, at::Tensor, c10::optional, c10::optional, c10::optional, uint32_t, int32_t, nv_bool, __nv_bool, nv_bool, at::Tensor, at::Tensor, at::Tensor) [with N_POS_DIMS=4U, N_FEAT_PER_PSEUDO_LVL=8U]" /home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/include/lotd/lotd_encoding.h(1836): here instantiation of "void lotd::torch::lod_bwd_bwd_input_impl(lotd::torch::LoDMeta &, at::Tensor, at::Tensor, at::Tensor, at::Tensor, c10::optional, c10::optional, c10::optional, uint32_t, int32_t, nv_bool, __nv_bool, nv_bool, at::Tensor, at::Tensor, at::Tensor) [with N_POS_DIMS=4U]" /home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/src/compile_split_3.cu(30): here

/home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/include/lotd/lotd_cuda.h(934): warning: missing return statement at end of non-void function "lotd::calc_dLdx_dim_vec_z_mat_xoy_impl<PARAM_T,COMPUTE_T,N_POS_DIMS,N_FEAT>(const uint32_t , uint32_t, const uint32_t , uint32_t, uint32_t, const PARAM_T , const PARAM_T , COMPUTE_T) [with PARAM_T=half, COMPUTE_T=float, N_POS_DIMS=4U, N_FEAT=2U]" detected during: instantiation of "std::enable_if<, COMPUTE_T>::type lotd::calc_dLdx_dim_vec_z_mat_xoy_impl<PARAM_T,COMPUTE_T,N_POS_DIMS,N_FEAT>(const uint32_t , uint32_t, const uint32_t , uint32_t, uint32_t, const PARAM_T , const PARAM_T , COMPUTE_T) [with PARAM_T=__half, COMPUTE_T=float, N_POS_DIMS=4U, N_FEAT=2U]" /home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/include/lotd/lotd_encoding.h(1271): here instantiation of "std::enable_if<std::is_same<INPUT_T, float>::value, void>::type lotd::torch::kernel_lod_backward_input_backward_input<INPUT_T,PARAM_T,COMPUTE_T,N_POS_DIMS,N_FEAT_PER_PSEUDO_LVL,N_FEAT_PER_THREAD>(uint32_t, lotd::LoDMetaRef, int32_t, const int32_t , const INPUT_T , const PARAM_T , const PARAM_T , const INPUT_T , const int64_t , const int64_t , uint32_t, INPUT_T ) [with INPUT_T=float, PARAM_T=half, COMPUTE_T=float, N_POS_DIMS=4U, N_FEAT_PER_PSEUDO_LVL=8U, N_FEAT_PER_THREAD=2U]" /home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/include/lotd/lotd_encoding.h(1746): here instantiation of "void lotd::torch::lod_bwd_bwd_input_impl_dispatched<INPUT_T,PARAM_T,COMPUTE_T,N_POS_DIMS,N_FEAT_PER_PSEUDO_LVL>(lotd::torch::LoDMeta &, at::Tensor, at::Tensor, at::Tensor, at::Tensor, c10::optional, c10::optional, c10::optional, uint32_t, int32_t, nv_bool, __nv_bool, nv_bool, at::Tensor, at::Tensor, at::Tensor) [with INPUT_T=float, PARAM_T=half, COMPUTE_T=float, N_POS_DIMS=4U, N_FEAT_PER_PSEUDO_LVL=8U]" /home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/include/lotd/lotd_encoding.h(1807): here instantiation of "void lotd::torch::lod_bwd_bwd_input_impl_templated<N_POS_DIMS,N_FEAT_PER_PSEUDO_LVL>(lotd::torch::LoDMeta &, at::Tensor, at::Tensor, at::Tensor, at::Tensor, c10::optional, c10::optional, c10::optional, uint32_t, int32_t, nv_bool, nv_bool, __nv_bool, at::Tensor, at::Tensor, at::Tensor) [with N_POS_DIMS=4U, N_FEAT_PER_PSEUDO_LVL=8U]" /home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/include/lotd/lotd_encoding.h(1836): here instantiation of "void lotd::torch::lod_bwd_bwd_input_impl(lotd::torch::LoDMeta &, at::Tensor, at::Tensor, at::Tensor, at::Tensor, c10::optional, c10::optional, c10::optional, uint32_t, int32_t, nv_bool, __nv_bool, __nv_bool, at::Tensor, at::Tensor, at::Tensor) [with N_POS_DIMS=4U]" /home/may/work/code/neuralsim/nr3d_lib/csrc/lotd/src/compile_split_3.cu(30): here

[ccclu]

same problem