ubuntu常见问题（持续更新）
1、nvcc warning : The ‘compute_20’, ‘sm_20’, and ‘sm_21’ architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
nvcc warning : The ‘compute_20’, ‘sm_20’, and ‘sm_21’ architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
到cmake后产生的OpencvConfig.cmake文件，找到如下代码：
# Version Compute Capability from which OpenCV has been compiled is remembered
set(OpenCV_COMPUTE_CAPABILITIES -gencode;arch=compute_11,code=sm_11;-gencode;arch=compute_12,code=sm_12;-gencode;arch=compute_13,code=sm_13;-gencode;arch=compute_20,code=sm_20;-gencode;arch=compute_20,code=sm_21;-gencode;arch=compute_30,code=sm_30;-gencode;arch=compute_35,code=sm_35;-gencode;arch=compute_30,code=compute_30)
将”-gencode ;arch=compute_11,code=sm_11”;-gencode;arch=compute_12,code=sm_12;-gencode;arch=compute_13,code=sm_13; 等删除。
主要因为此架构在CUDA过时～～将来也将不支持
以上若无果：
在cmakelist中关闭cuda~~
或者在cmake时加上参数-D BUILD_opencv_gpu=OFF
2、在编译caffe的Mkefile.config时遇到报错
nvcc warning : The ‘compute_20’, ‘sm_20’, and ‘sm_21’ architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
原因是，Makefile中采用了CUDA的compute capability 2.0和2.1，这是两种计算能力。安装的CUDA版本是8.0，但 从CUDA 8.0开始compute capability 2.0和2.1被弃用了，所以可以将-gencode arch=compute_20,code=sm_20 和-gencode arch=compute_20,code=sm_21这两行删除即可。
安装opencv3.3.1测试报错！
Traceback (most recent call last): File “opencvtest.py”, line 1, in import cv2 ImportError: No module named cv2
3、Ubuntu16.04上使用Python调用OpenCV需要安装对应的python包:
sudo apt-get install python-opencv这个还是重装Ubuntu才测试成功的。可以检查一下安装版本
pkg-config --modversion opencv
我测试的是3.3.1
但是用
my@my-by-HP:~$ python
Python 2.7.12 (default, Nov 19 2016, 06:48:10)
[GCC 5.4.0 20160609] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> import cv2
>>> cv2.__version__
'2.4.9.1'4、 import cv2
ImportError: /home/heyuxuan/.local/lib/python2.7/site-packages/torch/lib/libgomp.so.1: version `GOMP_4.0’ not found (required by /usr/lib/x86_64-linux-gnu/libsoxr.so.0)
参考http://blog.csdn.net/sinat_33621733/article/details/53710868
具体可以参考自己的错误来操作。
ImportError: /home/zdyd/anaconda2/bin/../lib/libstdc++.so.6: version `GLIBCXX_3.4.21' not found (required by /home/zdyd/caffe/python/caffe/_caffe.so)
strings /home/zdyd/anaconda2/bin/../lib/libstdc++.so.6 | grep GLIBCXX
发现确实没有，
find / -iname libstdc++.so.6.0.2*
找到最新版本在下面的文件家中存在，
strings /usr/lib/x86_64-linux-gnu/libstdc++.so.6.0.21 | grep GLIBCXX
确实有，下一步拷到相关目录
cp /usr/lib/x86_64-linux-gnu/libstdc++.so.6.0.21 /home/zdyd/anaconda2/bin/../lib/
在到之前的目录
cd /home/zdyd/anaconda2/bin/../lib/
删除原来的：
rm -rf libstdc++.so.6
建立软链接：
ln -s libstdc++.so.6.0.21 libstdc++.so.6
再次查看已存在
strings /home/zdyd/anaconda2/bin/../lib/libstdc++.so.6 | grep GLIBCXX
注意：一定要看好是自己的那个目录的相关版本不存在，不要瞎改，
5、Traceback (most recent call last):
File “test.py”, line 6, in
from src.data_loader import ImageDataLoader
File “/home/heyuxuan/ROOT/src/data_loader.py”, line 5, in
import pandas as pd
ImportError: No module named pandas
sudo apt-get install python-pandas
pandas可以通过pip安装，但要安装相关的依赖包。
6、import h5py
ImportError: No module named h5py
安装h5py的命令如下：
sudo pip install cython
sudo apt-get install libhdf5-dev
sudo pip install h5py安装完成后可以用如下命令测试：
python
import h5py7、heyuxuan@heyuxuan-by-HP:~/ROOT$ python test.py
Pre-loading the data. This may take a while…
Loaded 100 / 182
Completed laoding 182 files
THCudaCheck FAIL file=/b/wheel/pytorch-src/torch/lib/THC/generic/THCStorage.cu line=66 error=30 : unknown error
Traceback (most recent call last):
File “test.py”, line 37, in
net.cuda()
File “/home/heyuxuan/.local/lib/python2.7/site-packages/torch/nn/modules/module.py”, line 147, in cuda
return self._apply(lambda t: t.cuda(device_id))
File “/home/heyuxuan/.local/lib/python2.7/site-packages/torch/nn/modules/module.py”, line 118, in _apply
module._apply(fn)
File “/home/heyuxuan/.local/lib/python2.7/site-packages/torch/nn/modules/module.py”, line 118, in _apply
module._apply(fn)
File “/home/heyuxuan/.local/lib/python2.7/site-packages/torch/nn/modules/module.py”, line 118, in _apply
module._apply(fn)
File “/home/heyuxuan/.local/lib/python2.7/site-packages/torch/nn/modules/module.py”, line 118, in _apply
module._apply(fn)
File “/home/heyuxuan/.local/lib/python2.7/site-packages/torch/nn/modules/module.py”, line 124, in _apply
param.data = fn(param.data)
File “/home/heyuxuan/.local/lib/python2.7/site-packages/torch/nn/modules/module.py”, line 147, in
return self._apply(lambda t: t.cuda(device_id))
File “/home/heyuxuan/.local/lib/python2.7/site-packages/torch/_utils.py”, line 65, in _cuda
return new_type(self.size()).copy_(self, async)
File “/home/heyuxuan/.local/lib/python2.7/site-packages/torch/cuda/init.py”, line 280, in new
return super(_CudaBase, cls).new(cls, *args, **kwargs)
RuntimeError: cuda runtime error (30) : unknown error at /b/wheel/pytorch-src/torch/lib/THC/generic/THCStorage.cu:66
8、heyuxuan@heyuxuan-by-HP:~/ROOT$ python train.py
Pre-loading the data. This may take a while…
Loaded 100 / 2439
Loaded 200 / 2439
Loaded 300 / 2439
Loaded 400 / 2439
Loaded 500 / 2439
Loaded 600 / 2439
Loaded 700 / 2439
Loaded 800 / 2439
Loaded 900 / 2439
Loaded 1000 / 2439
Loaded 1100 / 2439
Loaded 1200 / 2439
Loaded 1300 / 2439
Loaded 1400 / 2439
Loaded 1500 / 2439
Loaded 1600 / 2439
Loaded 1700 / 2439
Loaded 1800 / 2439
Loaded 1900 / 2439
Loaded 2000 / 2439
Loaded 2100 / 2439
Loaded 2200 / 2439
Loaded 2300 / 2439
Loaded 2400 / 2439
Completed laoding 2439 files
Pre-loading the data. This may take a while…
Loaded 100 / 261
Loaded 200 / 261
Completed laoding 261 files
THCudaCheck FAIL file=/b/wheel/pytorch-src/torch/lib/THC/generic/THCStorage.cu line=66 error=30 : unknown error
Traceback (most recent call last):
File “train.py”, line 71, in
net = CrowdCounter(ce_weights=class_wts)
File “/home/heyuxuan/ROOT/src/crowd_count.py”, line 14, in init
ce_weights = ce_weights.cuda()
File “/home/heyuxuan/.local/lib/python2.7/site-packages/torch/_utils.py”, line 65, in _cuda
return new_type(self.size()).copy_(self, async)
File “/home/heyuxuan/.local/lib/python2.7/site-packages/torch/cuda/init.py”, line 280, in new
return super(_CudaBase, cls).new(cls, *args, **kwargs)
RuntimeError: cuda runtime error (30) : unknown error at /b/wheel/pytorch-src/torch/lib/THC/generic/THCStorage.cu:66