anaconda libgomp.so.1: version `GOMP_4.0' not found 解决办法

pytorch no kernel image is available for execution on the device

/home/caocao/anaconda3/compiler_compat/ld: cannot find -lpthread
/home/caocao/anaconda3/compiler_compat/ld: cannot find -lc
collect2: error: ld returned 1 exit status

I have the same error, then I input this:
sudo ln -s /usr/local/cuda-8.0/lib64/libcufft.so /usr/local/lib/libcufft.so,
it work well.

cp /usr/lib/x86_64-linux-gnu/libc.so /home/caocao/anaconda3/lib/libc.so

cp /usr/lib/x86_64-linux-gnu/libpthread.so /home/caocao/anaconda3/lib/libpthread.so

cffi.error.VerificationError: CompileError: command 'gcc' failed with exit status 1

ln -s /usr/local/cuda-9.0/include/*  /home/caocao/anaconda3/lib/python3.6/site-packages/torch/utils/ffi/../../lib/include/THC/

ln -s /usr/local/cuda-9.0/*  /usr/local/cuda/*

 

ARCH = sm_52

export PATH = / usr / local / cuda-9.0 / bin $ {PATH：+：$ {PATH}}
export CPATH = / usr / local / cuda-9.0 / include $ {CPATH：+：$ {CPATH}}
export LD_LIBRARY_PATH =在/ usr /本地/ CUDA的9.0 / lib64下$ {LD_LIBRARY_PATH：+：$ {LD_LIBRARY_PATH}}`

看起来它没有编译nnd_cuda.cu。是否在setup.py脚本中正确指定了？您的用户是否允许在磁盘上写入文件？

export CUDA_PATH = / usr / local / cuda /
export CXXFLAGS =“ - std = c ++ 11”
export CFLAGS =“ - std = c99”

export PATH = / usr / local / cuda-9.0 / bin $ {PATH：+：$ {PATH}}
export CPATH = / usr / local / cuda-9.0 / include $ {CPATH：+：$ {CPATH}}
export LD_LIBRARY_PATH =在/ usr /本地/ CUDA的9.0 / lib64下$ {LD_LIBRARY_PATH：+：$ {LD_LIBRARY_PATH}}

ARCH= -gencode arch=compute_52,code=[sm_52,compute_52]

 

 

类似问题还有'ImportError ../lib/libstdc++.so.6:未找到版本`CXXABI_1.3.8'（xxx要求）'，'ImportError ../lib/libstdc++.so.6:version`CXXABI_1.3.9'找不到（XXX要求）'等。

cp /usr/lib/x86_64-linux-gnu/libgomp.so.1 /home/caocao/anaconda3/lib/libgomp.so.1

cp /usr/lib/x86_64-linux-gnu/libstdc++.so.6 /home/caocao/anaconda3/lib/libstdc++.so.6

cp /usr/lib/x86_64-linux-gnu/libstdc++.so.6 /home/caocao/anaconda3/libstdc++.so.6

 

ln -s /usr/lib/x86_64-linux-gnu/libgomp.so.1 /home/caocao/anaconda3/libgomp.so.1
这种报错情况一般出现在阿纳康达环境的Python的程序中，原因是水蟒对应目录下的“libgomp.so.1”和“++的libstdc.so.6”动态库文件（LIB）中相关模块的版本比较旧，不包含相应版本模块导致的。

ln -s /usr/lib/x86_64-linux-gnu/libgomp.so.1 /home/caocao/anaconda3/lib/python3.6/site-packages/torch/lib/libgomp.so.1

系统中动态库存放路径在：/ usr / lib / x86_64-linux-gnu /或/ usr / lib /

森蚺中动态库存放路径是：〜/ anaconda2 / lib目录

查看的的Linux的系统动态库文件包含的模块版本：

字符串/usr/lib/libstdc++.so.6 | grep'CXXABI”

 

strings libstdc ++。so.6 | grep'CXXABI_1.3.8'

或者

字符串/usr/lib/x86_64-linux-gnu/libstdc++.so.6 | grep'CXXABI”

输出：

    CXXABI_1.3
    CXXABI_1.3.1
    CXXABI_1.3.2
    CXXABI_1.3.3
    CXXABI_1.3.4
    CXXABI_1.3.5
    CXXABI_1.3.6
    CXXABI_1.3.7
    CXXABI_1.3.8
    CXXABI_1.3.9
    CXXABI_TM_1
    CXXABI_FLOAT128

查看蟒蛇路径下动态库文件所包含的模块版本：

字符串anaconda2 / lib / libstdc ++。so.6 | grep'CXXABI'

输出：

    CXXABI_1.3
    CXXABI_1.3.1
    CXXABI_1.3.2
    CXXABI_1.3.3
    CXXABI_1.3.4
    CXXABI_1.3.5
    CXXABI_1.3.6
    CXXABI_1.3.7
    CXXABI_TM_1

 

系统目录下的文件包含所需模块的版本 'CXXABI_1.3.8' 和 'CXXABI_1.3.9'，复制系统目录下的动态库文件替换掉水蟒目录下的动态库文件，问题就解决了。

同样可以使用字符串指令查看蟒蛇中和系统下libgomp.so.1文件中分别包含的版本：

字符串libgomp.so.1 | grep的“GOMP”

 

如果系统目录的lib中也不包含所需模块，可以尝试使用以下命令更新动态库文件：

    LD_LIBRARY_PATH = / usr / local / lib64 /：$ LD_LIBRARY_PATH
    export LD_LIBRARY_PATH
    sudo add-apt-repository ppa：ubuntu-toolchain -r / test
    sudo apt-get update
    sudo apt-get install gcc-4.9 g ++ - 4.9
--- - ----------------  
作者： - 牧野 -  
来源：CSDN  
原文：HTTPS：//blog.csdn.net/dcrmg/article/details/78358279  
版权声明：本文为博主原创文章，转载请附上博文链接！

 

/home/caocao/anaconda3/bin/python /home/caocao/Downloads/stnm.pytorch-master/script/test.py
Traceback (most recent call last):
error in BilinearSampler.updateOutput: no kernel image is available for execution on the device
  File "/home/caocao/Downloads/stnm.pytorch-master/script/test.py", line 31, in <module>
    out = s(canvas, fgimg, fggrid, fgmask)
  File "/home/caocao/anaconda3/lib/python3.6/site-packages/torch/nn/modules/module.py", line 491, in __call__
    result = self.forward(*input, **kwargs)
  File "/home/caocao/Downloads/stnm.pytorch-master/script/modules/stnm.py", line 9, in forward
    return self.f(canvas, fgimg, fggrid, fgmask)
  File "/home/caocao/Downloads/stnm.pytorch-master/script/functions/stnm.py", line 19, in forward
    stnm.BilinearSamplerBHWD_updateOutput_cuda(canvas, fgimg, fggrid, fgmask, output)
  File "/home/caocao/anaconda3/lib/python3.6/site-packages/torch/utils/ffi/__init__.py", line 197, in safe_call
    result = torch._C._safe_call(*args, **kwargs)
torch.FatalError: aborting at /home/caocao/Downloads/stnm.pytorch-master/script/src/stnm_cuda.c:60