Closed amueller closed 6 years ago
Yes, that is the case. We are hoping to get in front of everything soon here, but it must be understood that this is a race and inevitably there'll be times when a distro is infront of us and things will break for people trying to do this kind of interoperation.
Could try renaming (and patchelf-ing everything that links to it) the libstdc++ that was used to build conda packages, so at least the dynamic loader has a chance to also try loading the newer system copy? I believe that was the approach recommended by the manylinux wheel-builder folks from what they've learned.
Some of the libraries installed by conda (e.g., snappy) also requires CXXABI_1.3.9
$HOME/anaconda3/envs/tf/lib/libsnappy.so: undefined reference to `operator delete(void*, unsigned long)@CXXABI_1.3.9'
We have released Anaconda Distribution 5
now which is built against libstdc++
from GCC 7.2.0
and therefore supports CXXABI_1.3.9
(up to and including CXXABI_1.3.11
) so you've run into the opposite problem; your system libstdc++
(or some other libstdc++
that ld.so
has loaded) doesn't support CXXABI_1.3.9
and you are trying to use that with Anaconda Distribution 5
software.
I am assuming you have set LD_LIBRARY_PATH
or LD_PRELOAD
here? If not can you give details of your OS and exact reproduction steps please.
The environment is Ubuntu 16.04 x64
I was trying to compile caffe completely from Anaconda, so I did something like this:
conda create -n tf anaconda python=3.6
source activate tf
conda install atlas boost gflags glog hdf5 leveldb lmdb openblas protobuf
conda install -c menpo opencv3
git clone https://github.com/BVLC/caffe
cd caffe
cp Makefile.config.example Makefile.config
and set the compiler to link against $(ANACONDA_HOME)/lib
by
5c5
< # USE_CUDNN := 1
---
> USE_CUDNN := 1
21c21
< # OPENCV_VERSION := 3
---
> OPENCV_VERSION := 3
68,69c68,69
< PYTHON_INCLUDE := /usr/include/python2.7 \
< /usr/lib/python2.7/dist-packages/numpy/core/include
---
> # PYTHON_INCLUDE := /usr/include/python2.7 \
> # /usr/lib/python2.7/dist-packages/numpy/core/include
72,75c72,75
< # ANACONDA_HOME := $(HOME)/anaconda
< # PYTHON_INCLUDE := $(ANACONDA_HOME)/include \
< # $(ANACONDA_HOME)/include/python2.7 \
< # $(ANACONDA_HOME)/lib/python2.7/site-packages/numpy/core/include
---
> ANACONDA_HOME := $(HOME)/anaconda3/envs/tf
> PYTHON_INCLUDE := $(ANACONDA_HOME)/include \
> $(ANACONDA_HOME)/include/python3.6 \
> $(ANACONDA_HOME)/lib/python3.6/site-packages/numpy/core/include
83,84c83,84
< PYTHON_LIB := /usr/lib
< # PYTHON_LIB := $(ANACONDA_HOME)/lib
---
> # PYTHON_LIB := /usr/lib
> PYTHON_LIB := $(ANACONDA_HOME)/lib
However, doing so would cause the following error:
undefined reference to `google::protobuf::internal::WireFormatLite
I suspect that this is because Ubuntu 16.04 has gcc 5.x, because I was able to compile in this way under Ubuntu 14.04 in the past.
Therefore, I decided to install gcc in Anaconda by
conda install gcc libgcc
and change the compiler to
25c25
< # CUSTOM_CXX := g++
---
> CUSTOM_CXX := $(HOME)/anaconda3/envs/tf/bin/g++
and eventually the error was produced:
$HOME/anaconda3/envs/tf/lib/libsnappy.so: undefined reference to `operator delete(void*, unsigned long)@CXXABI_1.3.9'
The gcc
and libgcc
packages are deprecated now. Please try gxx_linux-64
instead. Note these compilers have prefixed names.
@mingwandroid Thanks, but CUDA only supports GCC <= 5.x[1] and even if I turn off CUDA I still encounter the following issue:
$HOME/anaconda3/envs/tf/lib/libopencv_highgui.so: undefined reference to `gtk_widget_get_type'
I really don't know why but I found that I am able to compile now if I set:
25c25
< # CUSTOM_CXX := g++
---
> CUSTOM_CXX := /usr/bin/g++-4.9
So it appears that the snappy
package, while compiled by GCC 7.x, actually supports GCC 4.9, but does not support GCC 5.x?
[1]: I know Anaconda has a package called cudatoolkit, but it doesn't seem to include the header files.
There is an issue in anaconda where libstdc++ is not linking to version 6.0.24 but rather to 6.0.19. If you manually relink libstdc++.so and libstdc++.so.6 to libstdc++.so.6.0.24 it should work.
@8greg8 Thanks a lot! You save my life!
@8greg8 worked, thnx. Put /home/user/anaconda2/lib/libstdc++.so.6.0.24 in LD_PRELOAD enviroment variable. Doesn't anyone know if anaconda pkgs will be updated, will this olution still work? Will that problem be fixed in new update?
@8greg8 Could you elaborate on what you mean by manually relinking libstdc++.so and libstdc++.so.6 to libstdc++.so.6.0.24
@8greg8 I have a same query as @masonkirchner , Could you please explain how to relink libstdc++.so and libstdc++.so.6 to libstdc++.so.6.0.24.
@masonkirchner and @8greg8, maybe I can help explaining how to relink. This is the way I did it:
libstdc++.so.6.0.24
in ~/anaconda3/lib/:
$ ls libstdc++.so.6.0.24
libstdc++.so.6
in ~/anaconda3/lib/:
$ ls libstdc++.so.6
libstdc++.so.6 -> libstdc++.so.6.0.19
):
$ rm ~/anaconda3/lib/libstdc++.so.6
libstdc++.so.6.0.24
:
$ ln -s libstdc++.so.6.0.24 libstdc++.so.6
It worked for me.
@kmyokoyama , it worked! Thanks!
In our system there are links with libstdc++.so and libstdc++.so.6 to libstdc++.so.6.0.19. But there is no libstdc++.so.6.0.24 How can I get that lib and our system is also complaining abt the OSError: /opt/anaconda3/bin/../lib/libstdc++.so.6: version `CXXABI_1.3.9' not found (required by /opt/anaconda3/lib/python3.6/site-packages/porechop/cpp_functions.so)
So, I need help. How can I solve the issue. Thanks!
Hi, relinking libstdc did not help. I still have this issue. I have conda 4.5.1, just upgraded seconds ago. I have libstdc++.so.6.0.24 in /anaconda3/lib. I've created symlink libstdc++.so.6 pointing to the former.
You just need to make sure you are not using either of the old gcc
nor the libgcc
packages.
Can you elaborate on how to find out if I'm using old libgcc packages. My gcc is version 4.8.4
Can you elaborate on how to find out if I'm using old libgcc packages.
conda list --show-channel-urls
will show you what packages you have installed.
My gcc is version 4.8.4
Your system GCC does not matter here. Conda environments are (almost fully) self-contained.
the list is long, but i guess what matters are these two
libgcc 5.2.0 0 defaults
libgcc-ng 7.2.0 h7cc24e2_2 defaults
One update : the script that fails has these import
lines,
import tensorflow as tf
import matplotlib.pyplot as plt
However, when I reverse the order of these two lines, ie.
import matplotlib.pyplot as plt
import tensorflow as tf
The error disappears. Why?
Background info tensorflow 1.7.0 python 3.6.5 anaconda3 matplotlib 2.1.0
conda remove libgcc
Then conda install libgcc-ng --force
It does not matter that the list is long, please paste the whole thing.
I do have the same issue (I work on a system that uses Python 2) when I try to import tellurium which I installed using
pip install git+https://github.com/sys-bio/tellurium.git
When I then try
import tellurium as te
I get
ImportError: /usr/lib/x86_64-linux-gnu/libstdc++.so.6: version `CXXABI_1.3.9' not found (required by path/to/anaconda2/lib/python2.7/site-packages/matplotlib/_path.so)
When I first do import matplotlib.pyplot as plt
, the import works fine.
I checked
ls -l ~/anaconda2/lib/libstdc++.so.6
which returns
path/to/anaconda2/lib/libstdc++.so.6 -> libstdc++.so.6.0.24
so the linking should be ok.
conda list --show-channel-urls
returns
# Name Version Build Channel
_ipyw_jlab_nb_ext_conf 0.1.0 py27h08a7f0c_0 defaults
alabaster 0.7.10 py27he5a193a_0 defaults
anaconda 5.1.0 py27_2 defaults
anaconda-client 1.6.9 py27_0 defaults
anaconda-navigator 1.7.0 py27_0 defaults
anaconda-project 0.8.2 py27h236b58a_0 defaults
antimony 2.9.4 <pip>
appdirs 1.4.3 <pip>
asn1crypto 0.24.0 py27_0 defaults
astroid 1.6.1 py27_0 defaults
astropy 2.0.3 py27h14c3975_0 defaults
attrs 17.4.0 py27_0 defaults
babel 2.5.3 py27_0 defaults
backports 1.0 py27h63c9359_1 defaults
backports.functools_lru_cache 1.4 py27he8db605_1 defaults
backports.shutil_get_terminal_size 1.0.0 py27h5bc021e_2 defaults
backports_abc 0.5 py27h7b3c97b_0 defaults
beautifulsoup4 4.6.0 py27h3f86ba9_1 defaults
bitarray 0.8.1 py27h14c3975_1 defaults
bkcharts 0.2 py27h241ae91_0 defaults
blaze 0.11.3 py27h5f341da_0 defaults
bleach 2.1.2 py27_0 defaults
bokeh 0.12.13 py27h5233db4_0 defaults
boto 2.48.0 py27h9556ac2_1 defaults
bottleneck 1.2.1 py27h21b16a3_0 defaults
bzip2 1.0.6 h9a117a8_4 defaults
ca-certificates 2017.08.26 h1d4fec5_0 defaults
cairo 1.14.12 h77bcde2_0 defaults
cdecimal 2.3 py27h14c3975_3 defaults
certifi 2018.1.18 py27_0 defaults
cffi 1.11.4 py27h9745a5d_0 defaults
chardet 3.0.4 py27hfa10054_1 defaults
click 6.7 py27h4225b90_0 defaults
cloudpickle 0.5.2 py27_1 defaults
clyent 1.2.2 py27h7276e6c_1 defaults
cobra 0.11.3 <pip>
colorama 0.3.9 py27h5cde069_0 defaults
conda 4.4.10 py27_0 defaults
conda-build 3.4.1 py27_0 defaults
conda-env 2.6.0 h36134e3_1 defaults
conda-verify 2.0.0 py27hf052a9d_0 defaults
configparser 3.5.0 py27h5117587_0 defaults
contextlib2 0.5.5 py27hbf4c468_0 defaults
cryptography 2.1.4 py27h6697b16_0 defaults
curl 7.58.0 h84994c4_0 defaults
cycler 0.10.0 py27hc7354d3_0 defaults
cython 0.27.3 py27hc56b35e_0 defaults
cytoolz 0.9.0 py27h14c3975_0 defaults
dask 0.16.1 py27_0 defaults
dask-core 0.16.1 py27_0 defaults
datashape 0.5.4 py27hf507385_0 defaults
dbus 1.12.2 hc3f9b76_1 defaults
decorator 4.2.1 py27_0 defaults
distributed 1.20.2 py27_0 defaults
docutils 0.14 py27hae222c1_0 defaults
entrypoints 0.2.3 py27h502b47d_2 defaults
enum34 1.1.6 py27h99a27e9_1 defaults
Escher 1.6.0 <pip>
et_xmlfile 1.0.1 py27h75840f5_0 defaults
expat 2.2.5 he0dffb1_0 defaults
fastcache 1.0.2 py27h14c3975_2 defaults
filelock 2.0.13 py27h61a9c69_0 defaults
flask 0.12.2 py27h6d5c1cd_0 defaults
flask-cors 3.0.3 py27h1a8a27f_0 defaults
fontconfig 2.12.4 h88586e7_1 defaults
freetype 2.8 hab7d2ae_1 defaults
funcsigs 1.0.2 py27h83f16ab_0 defaults
functools32 3.2.3.2 py27h4ead58f_1 defaults
future 0.16.0 <pip>
futures 3.2.0 py27h7b459c0_0 defaults
get_terminal_size 1.0.0 haa9412d_0 defaults
gevent 1.2.2 py27h475ea6a_0 defaults
glib 2.53.6 h5d9569c_2 defaults
glob2 0.6 py27hcea9cbd_0 defaults
glpk 0.3 <pip>
gmp 6.1.2 h6c8ec71_1 defaults
gmpy2 2.0.8 py27h4cf3fa8_2 defaults
graphite2 1.3.10 hf63cedd_1 defaults
greenlet 0.4.12 py27hac09c53_0 defaults
grin 1.2.1 py27_4 defaults
gst-plugins-base 1.12.4 h33fb286_0 defaults
gstreamer 1.12.4 hb53b477_0 defaults
h5py 2.7.1 py27h2697762_0 defaults
harfbuzz 1.7.4 hc5b324e_0 defaults
hdf5 1.10.1 h9caa474_1 defaults
heapdict 1.0.0 py27_2 defaults
html5lib 1.0.1 py27h5233db4_0 defaults
icu 58.2 h9c2bf20_1 defaults
idna 2.6 py27h5722d68_1 defaults
imageio 2.2.0 py27hf108a7f_0 defaults
imagesize 0.7.1 py27hd17bf80_0 defaults
intel-openmp 2018.0.0 hc7b2577_8 defaults
ipaddress 1.0.19 py27_0 defaults
ipykernel 4.8.0 py27_0 defaults
ipython 5.4.1 py27_2 defaults
ipython_genutils 0.2.0 py27h89fb69b_0 defaults
ipywidgets 7.1.1 py27_0 defaults
isodate 0.6.0 <pip>
isort 4.2.15 py27hcfa4749_0 defaults
itsdangerous 0.24 py27hb8295c1_1 defaults
jbig 2.1 hdba287a_0 defaults
jdcal 1.3 py27h2cc5433_0 defaults
jedi 0.11.1 py27_0 defaults
jinja2 2.10 py27h4114e70_0 defaults
jpeg 9b h024ee3a_2 defaults
jsonschema 2.6.0 py27h7ed5aa4_0 defaults
jupyter 1.0.0 py27_4 defaults
jupyter_client 5.2.2 py27_0 defaults
jupyter_console 5.2.0 py27hc6bee7e_1 defaults
jupyter_core 4.4.0 py27h345911c_0 defaults
jupyterlab 0.31.5 py27_0 defaults
jupyterlab_launcher 0.10.2 py27_0 defaults
lazy-object-proxy 1.3.1 py27h682c727_0 defaults
libcurl 7.58.0 h1ad7b7a_0 defaults
libedit 3.1 heed3624_0 defaults
libffi 3.2.1 hd88cf55_4 defaults
libgcc-ng 7.2.0 h7cc24e2_2 defaults
libgfortran-ng 7.2.0 h9f7466a_2 defaults
libpng 1.6.34 hb9fc6fc_0 defaults
libroadrunner 1.4.24 <pip>
libsodium 1.0.15 hf101ebd_0 defaults
libssh2 1.8.0 h9cfc8f7_4 defaults
libstdcxx-ng 7.2.0 h7a57d05_2 defaults
libtiff 4.0.9 h28f6b97_0 defaults
libtool 2.4.6 h544aabb_3 defaults
libxcb 1.12 hcd93eb1_4 defaults
libxml2 2.9.7 h26e45fe_0 defaults
libxslt 1.1.32 h1312cb7_0 defaults
llvmlite 0.21.0 py27h6dad3f9_0 defaults
locket 0.2.0 py27h73929a2_1 defaults
lxml 4.1.1 py27hf71bdeb_1 defaults
lzo 2.10 h49e0be7_2 defaults
markupsafe 1.0 py27h97b2822_1 defaults
matplotlib 2.1.2 py27h0e671d2_0 defaults
mccabe 0.6.1 py27h0e7c7be_1 defaults
mistune 0.8.3 py27_0 defaults
mkl 2018.0.1 h19d6760_4 defaults
mkl-service 1.1.2 py27hb2d42c5_4 defaults
mpc 1.0.3 hec55b23_5 defaults
mpfr 3.1.5 h11a74b3_2 defaults
mpmath 1.0.0 py27h9669132_2 defaults
msgpack-python 0.5.1 py27h6bb024c_0 defaults
multipledispatch 0.4.9 py27h9b5f95a_0 defaults
navigator-updater 0.1.0 py27h0f9cd39_0 defaults
nbconvert 5.3.1 py27he041f76_0 defaults
nbformat 4.4.0 py27hed7f2b2_0 defaults
ncurses 6.0 h9df7e31_2 defaults
networkx 2.1 py27_0 defaults
nltk 3.2.5 py27hec5f4de_0 defaults
nose 1.3.7 py27heec2199_2 defaults
notebook 5.4.0 py27_0 defaults
numba 0.36.2 np114py27he2868da_0 defaults
numexpr 2.6.4 py27hd318778_0 defaults
numpy 1.14.0 py27h3dfced4_1 defaults
numpydoc 0.7.0 py27h9647a75_0 defaults
odo 0.5.1 py27h9170de3_0 defaults
olefile 0.45.1 py27_0 defaults
openpyxl 2.4.10 py27_0 defaults
openssl 1.0.2o h20670df_0 defaults
optlang 1.4.1 <pip>
packaging 16.8 py27h5e07c7c_1 defaults
pandas 0.22.0 py27hf484d3e_0 defaults
pandoc 1.19.2.1 hea2e7c5_1 defaults
pandocfilters 1.4.2 py27h428e1e5_1 defaults
pango 1.41.0 hd475d92_0 defaults
parso 0.1.1 py27h718acc2_0 defaults
partd 0.3.8 py27h4e55004_0 defaults
patchelf 0.9 hf79760b_2 defaults
path.py 10.5 py27hefe4bee_0 defaults
pathlib2 2.3.0 py27h6e9d198_0 defaults
patsy 0.5.0 py27_0 defaults
pcre 8.41 hc27e229_1 defaults
pep8 1.7.1 py27_0 defaults
pexpect 4.3.1 py27_0 defaults
phrasedml 1.0.9 <pip>
pickleshare 0.7.4 py27h09770e1_0 defaults
pillow 5.0.0 py27h3deb7b8_0 defaults
pip 9.0.1 py27ha730c48_4 defaults
pixman 0.34.0 hceecf20_3 defaults
pkginfo 1.4.1 py27hee1a9ad_1 defaults
plotly 2.5.1 <pip>
pluggy 0.6.0 py27h1f4f128_0 defaults
ply 3.10 py27hd6d9ae5_0 defaults
prompt_toolkit 1.0.15 py27h1b593e1_0 defaults
psutil 5.4.3 py27h14c3975_0 defaults
ptyprocess 0.5.2 py27h4ccb14c_0 defaults
py 1.5.2 py27h203d672_0 defaults
pycairo 1.15.4 py27h1b9232e_1 defaults
pycodestyle 2.3.1 py27h904819d_0 defaults
pycosat 0.6.3 py27ha4109ae_0 defaults
pycparser 2.18 py27hefa08c5_1 defaults
pycrypto 2.6.1 py27h14c3975_7 defaults
pycurl 7.43.0.1 py27hb7f436b_0 defaults
pyflakes 1.6.0 py27h904a57d_0 defaults
pygments 2.2.0 py27h4a8b6f5_0 defaults
pylint 1.8.2 py27_0 defaults
pyodbc 4.0.22 py27hf484d3e_0 defaults
pyopenssl 17.5.0 py27hcee3be0_0 defaults
pyparsing 2.2.0 py27hf1513f8_1 defaults
pyqt 5.6.0 py27h4b1e83c_5 defaults
pysocks 1.6.7 py27he2db6d2_1 defaults
pytables 3.4.2 py27h1f7bffc_2 defaults
pytest 3.3.2 py27_0 defaults
python 2.7.14 h1571d57_29 defaults
python-dateutil 2.6.1 py27h4ca5741_1 defaults
python-libsbml 5.16.0 py27hded5652_0 sbmlteam
pytz 2017.3 py27h001bace_0 defaults
pywavelets 0.5.2 py27hecda097_0 defaults
pyyaml 3.12 py27h2d70dd7_1 defaults
pyzmq 16.0.3 py27hc579512_0 defaults
qt 5.6.2 h974d657_12 defaults
qtawesome 0.4.4 py27hd7914c3_0 defaults
qtconsole 4.3.1 py27hc444b0d_0 defaults
qtpy 1.3.1 py27h63d3751_0 defaults
rdflib 4.2.2 <pip>
readline 7.0 ha6073c6_4 defaults
requests 2.18.4 py27hc5b0589_1 defaults
rope 0.10.7 py27hfe459b0_0 defaults
rrplugins 1.1.10 <pip>
ruamel.ordereddict 0.4.13 <pip>
ruamel.yaml 0.14.12 <pip>
ruamel_yaml 0.15.35 py27h14c3975_1 defaults
sbml2matlab 0.9.1 <pip>
scandir 1.6 py27hf7388dc_0 defaults
scikit-image 0.13.1 py27h14c3975_1 defaults
scikit-learn 0.19.1 py27h445a80a_0 defaults
scipy 1.0.0 py27hf5f0f52_0 defaults
seaborn 0.8.1 py27h633ea1e_0 defaults
send2trash 1.4.2 py27_0 defaults
setuptools 38.4.0 py27_0 defaults
simplegeneric 0.8.1 py27_2 defaults
singledispatch 3.4.0.3 py27h9bcb476_0 defaults
sip 4.18.1 py27he9ba0ab_2 defaults
six 1.11.0 py27h5f960f1_1 defaults
snowballstemmer 1.2.1 py27h44e2768_0 defaults
sortedcollections 0.5.3 py27h135218e_0 defaults
sortedcontainers 1.5.9 py27_0 defaults
SPARQLWrapper 1.8.1 <pip>
sphinx 1.6.6 py27_0 defaults
sphinxcontrib 1.0 py27h1512b58_1 defaults
sphinxcontrib-websupport 1.0.1 py27hf906f22_1 defaults
spyder 3.2.6 py27_0 defaults
sqlalchemy 1.2.1 py27h14c3975_0 defaults
sqlite 3.22.0 h1bed415_0 defaults
ssl_match_hostname 3.5.0.1 py27h4ec10b9_2 defaults
statsmodels 0.8.0 py27hc87d62d_0 defaults
subprocess32 3.2.7 py27h373dbce_0 defaults
swiglpk 1.4.4 <pip>
sympy 1.1.1 py27hc28188a_0 defaults
tabulate 0.8.2 <pip>
tblib 1.3.2 py27h51fe5ba_0 defaults
tecombine 0.2.2.1 <pip>
tellurium 2.0.12 <pip>
tenuml 1.1.1 <pip>
terminado 0.8.1 py27_1 defaults
tesbml 5.15.0.1 <pip>
tesedml 0.4.3 <pip>
testpath 0.3.1 py27hc38d2c4_0 defaults
tk 8.6.7 hc745277_3 defaults
toolz 0.9.0 py27_0 defaults
tornado 4.5.3 py27_0 defaults
traitlets 4.3.2 py27hd6ce930_0 defaults
typing 3.6.2 py27h66f49e2_0 defaults
unicodecsv 0.14.1 py27h5062da9_0 defaults
unixodbc 2.3.4 hc36303a_1 defaults
urllib3 1.22 py27ha55213b_0 defaults
wcwidth 0.1.7 py27h9e3e1ab_0 defaults
webencodings 0.5.1 py27hff10b21_1 defaults
werkzeug 0.14.1 py27_0 defaults
wheel 0.30.0 py27h2bc6bb2_1 defaults
widgetsnbextension 3.1.0 py27_0 defaults
wrapt 1.10.11 py27h04f6869_0 defaults
xlrd 1.1.0 py27ha77178f_1 defaults
xlsxwriter 1.0.2 py27h12cbc6b_0 defaults
xlwt 1.3.0 py27h3d85d97_0 defaults
xz 5.2.3 h55aa19d_2 defaults
yaml 0.1.7 had09818_2 defaults
zeromq 4.2.2 hbedb6e5_2 defaults
zict 0.1.3 py27h12c336c_0 defaults
zlib 1.2.11 ha838bed_2 defaults
Any idea how to fix this?
I update LD_LIBARARY_PATH in .basrc
export LD_LIBRARY_PATH="/home/myname/anaconda3/lib:$LD_LIBRARY_PATH"
This issue has not gone away. Here is my
conda list --show-channel-urls
### # packages in environment at /home/rspace/anaconda3:
#
# Name Version Build Channel
_ipyw_jlab_nb_ext_conf 0.1.0 py36he11e457_0 defaults
_nb_ext_conf 0.4.0 py36_1 defaults
absl-py 0.1.13 <pip>
alabaster 0.7.10 py36h306e16b_0 defaults
anaconda custom py36hbbc8b67_0 defaults
anaconda-client 1.6.5 py36h19c0dcd_0 defaults
anaconda-navigator 1.6.9 py36h11ddaaa_0 defaults
anaconda-project 0.8.0 py36h29abdf5_0 defaults
arctic 1.55.0 <pip>
asn1crypto 0.22.0 py36h265ca7c_1 defaults
astor 0.6.2 <pip>
astroid 1.5.3 py36hbdb9df2_0 defaults
astropy 2.0.2 py36ha51211e_4 defaults
atari-py 0.1.1 <pip>
babel 2.5.0 py36h7d14adf_0 defaults
backports 1.0 py36hfa02d7e_1 defaults
backports.shutil_get_terminal_size 1.0.0 py36hfea85ff_2 defaults
beautifulsoup4 4.6.0 py36h49b8c8c_1 defaults
bintrees 2.0.7 <pip>
bitarray 0.8.1 py36h5834eb8_0 defaults
bkcharts 0.2 py36h735825a_0 defaults
blaze 0.11.3 py36h4e06776_0 defaults
bleach 1.5.0 <pip>
bleach 2.0.0 py36h688b259_0 defaults
bokeh 0.12.10 py36hbb0e44a_0 defaults
boto 2.48.0 py36h6e4cd66_1 defaults
bottleneck 1.2.1 py36haac1ea0_0 defaults
Box2D-kengz 2.3.3 <pip>
bzip2 1.0.6 h0376d23_1 defaults
ca-certificates 2018.03.07 0 defaults
cairo 1.14.10 haa5651f_5 defaults
certifi 2018.1.18 py36_0 defaults
cffi 1.10.0 py36had8d393_1 defaults
chardet 3.0.4 py36h0f667ec_1 defaults
click 6.7 py36h5253387_0 defaults
cloudpickle 0.4.0 py36h30f8c20_0 defaults
clyent 1.2.2 py36h7e57e65_1 defaults
colorama 0.3.9 py36h489cec4_0 defaults
conda 4.5.1 py36_0 defaults
conda-build 3.0.27 py36h940a66d_0 defaults
conda-env 2.6.0 h36134e3_1 defaults
conda-verify 2.0.0 py36h98955d8_0 defaults
contextlib2 0.5.5 py36h6c84a62_0 defaults
cryptography 2.0.3 py36ha225213_1 defaults
curl 7.55.1 hcb0b314_2 defaults
cycler 0.10.0 py36h93f1223_0 defaults
cython 0.26.1 py36h21c49d0_0 defaults
cytoolz 0.8.2 py36h708bfd4_0 defaults
dask 0.15.3 py36hdc2c8aa_0 defaults
dask 0.17.2 <pip>
dask-core 0.15.3 py36h10e6167_0 defaults
datashape 0.5.4 py36h3ad6b5c_0 defaults
dbus 1.10.22 h3b5a359_0 defaults
decorator 4.1.2 py36hd076ac8_0 defaults
deepdish 0.3.6 <pip>
distributed 1.19.1 py36h25f3894_0 defaults
dnspython 1.15.0 <pip>
docutils 0.14 py36hb0f60f5_0 defaults
ecdsa 0.13 <pip>
Electrum 3.0.2 <pip>
entrypoints 0.2.3 py36h1aec115_2 defaults
enum34 1.1.6 <pip>
et_xmlfile 1.0.1 py36hd6bccc3_0 defaults
expat 2.2.4 hc00ebd1_1 defaults
fastcache 1.0.2 py36h5b0c431_0 defaults
filelock 2.0.12 py36hacfa1f5_0 defaults
flask 0.12.2 py36hb24657c_0 defaults
flask-cors 3.0.3 py36h2d857d3_0 defaults
fontconfig 2.12.4 h88586e7_1 defaults
freetype 2.8 h52ed37b_0 defaults
future 0.16.0 <pip>
gast 0.2.0 <pip>
get_terminal_size 1.0.0 haa9412d_0 defaults
gevent 1.2.2 py36h2fe25dc_0 defaults
glib 2.53.6 hc861d11_1 defaults
glob2 0.5 py36h2c1b292_1 defaults
gmp 6.1.2 hb3b607b_0 defaults
gmpy2 2.0.8 py36h55090d7_1 defaults
graphite2 1.3.10 hc526e54_0 defaults
greenlet 0.4.12 py36h2d503a6_0 defaults
grpcio 1.10.1 <pip>
gst-plugins-base 1.12.2 he3457e5_0 defaults
gstreamer 1.12.2 h4f93127_0 defaults
gym 0.9.3 <pip>
h5py 2.7.0 py36he81ebca_1 defaults
harfbuzz 1.5.0 h2545bd6_0 defaults
hdbscan 0.8.12 <pip>
hdf5 1.10.1 hb0523eb_0 defaults
heapdict 1.0.0 py36h79797d7_0 defaults
html5lib 0.999999999 py36h2cfc398_0 defaults
html5lib 0.9999999 <pip>
hypertools 0.5.0 <pip>
icu 58.2 h211956c_0 defaults
idna 2.6 py36h82fb2a8_1 defaults
imageio 2.2.0 py36he555465_0 defaults
imagesize 0.7.1 py36h52d8127_0 defaults
imgaug 0.2.5 <pip>
inflection 0.3.1 <pip>
intel-openmp 2018.0.0 h15fc484_7 defaults
ipykernel 4.6.1 py36hbf841aa_0 defaults
ipyparallel 6.1.1 <pip>
ipython 6.1.0 py36hc72a948_1 defaults
ipython_genutils 0.2.0 py36hb52b0d5_0 defaults
ipywidgets 7.0.0 py36h7b55c3a_0 defaults
isort 4.2.15 py36had401c0_0 defaults
itsdangerous 0.24 py36h93cc618_1 defaults
jbig 2.1 hdba287a_0 defaults
jdcal 1.3 py36h4c697fb_0 defaults
jedi 0.10.2 py36h552def0_0 defaults
jinja2 2.9.6 py36h489bce4_1 defaults
jpeg 9b habf39ab_1 defaults
jsonrpclib-pelix 0.3.1 <pip>
jsonschema 2.6.0 py36h006f8b5_0 defaults
jupyter 1.0.0 py36h9896ce5_0 defaults
jupyter_client 5.1.0 py36h614e9ea_0 defaults
jupyter_console 5.2.0 py36he59e554_1 defaults
jupyter_core 4.3.0 py36h357a921_0 defaults
jupyterlab 0.27.0 py36h86377d0_2 defaults
jupyterlab_launcher 0.4.0 py36h4d8058d_0 defaults
Keras 2.1.5 <pip>
lazy-object-proxy 1.3.1 py36h10fcdad_0 defaults
libedit 3.1 heed3624_0 defaults
libffi 3.2.1 h4deb6c0_3 defaults
libgcc 5.2.0 0 defaults
libgcc-ng 7.2.0 h7cc24e2_2 defaults
libgfortran-ng 7.2.0 h9f7466a_2 defaults
libpng 1.6.32 hda9c8bc_2 defaults
libsodium 1.0.13 h31c71d8_2 defaults
libssh2 1.8.0 h8c220ad_2 defaults
libstdcxx-ng 7.2.0 h7a57d05_2 defaults
libtiff 4.0.8 h90200ff_9 defaults
libtool 2.4.6 hd50d1a6_0 defaults
libxcb 1.12 h84ff03f_3 defaults
libxml2 2.9.4 h6b072ca_5 defaults
libxslt 1.1.29 hcf9102b_5 defaults
llvmlite 0.20.0 py36_0 defaults
locket 0.2.0 py36h787c0ad_1 defaults
lxml 4.1.0 py36h5b66e50_0 defaults
lzo 2.10 h1bfc0ba_1 defaults
Markdown 2.6.11 <pip>
markupsafe 1.0 py36hd9260cd_1 defaults
matplotlib 2.1.0 py36hba5de38_0 defaults
mccabe 0.6.1 py36h5ad9710_1 defaults
mistune 0.7.4 py36hbab8784_0 defaults
mkl 2018.0.0 hb491cac_4 defaults
mkl-service 1.1.2 py36h17a0993_4 defaults
mockextras 1.0.2 <pip>
more-itertools 4.0.1 <pip>
mpc 1.0.3 hf803216_4 defaults
mpfr 3.1.5 h12ff648_1 defaults
mpmath 0.19 py36h8cc018b_2 defaults
msgpack-python 0.4.8 py36hec4c5d1_0 defaults
mujoco-py 0.5.7 <pip>
multipledispatch 0.4.9 py36h41da3fb_0 defaults
navigator-updater 0.1.0 py36h14770f7_0 defaults
nb_anacondacloud 1.4.0 py36_0 defaults
nb_conda 2.2.1 py36h8118bb2_0 defaults
nb_conda_kernels 2.1.0 py36_0 defaults
nbconvert 5.3.1 py36hb41ffb7_0 defaults
nbformat 4.4.0 py36h31c9010_0 defaults
nbpresent 3.0.2 py36h5f95a39_1 defaults
ncurses 6.0 h06874d7_1 defaults
ndg-httpsclient 0.4.3 <pip>
networkx 2.0 py36h7e96fb8_0 defaults
nltk 3.2.4 py36h1a0979f_0 defaults
nose 1.3.7 py36hcdf7029_2 defaults
notebook 5.0.0 py36h0b20546_2 defaults
numba 0.35.0 np113py36_10 defaults
numexpr 2.6.2 py36hdd3393f_1 defaults
numpy 1.14.2 <pip>
numpy 1.13.3 py36ha12f23b_0 defaults
numpydoc 0.7.0 py36h18f165f_0 defaults
odo 0.5.1 py36h90ed295_0 defaults
olefile 0.44 py36h79f9f78_0 defaults
opencv-python 3.4.0.12 <pip>
openpyxl 2.4.8 py36h41dd2a8_1 defaults
openssl 1.0.2o h20670df_0 defaults
pachi-py 0.0.21 <pip>
packaging 16.8 py36ha668100_1 defaults
pandas 0.21.1 py36h20b78c2_0 defaults
pandoc 1.19.2.1 hea2e7c5_1 defaults
pandocfilters 1.4.2 py36ha6701b7_1 defaults
pango 1.40.11 h8191d47_0 defaults
partd 0.3.8 py36h36fd896_0 defaults
patchelf 0.9 hf79760b_2 defaults
path.py 10.3.1 py36he0c6f6d_0 defaults
pathlib2 2.3.0 py36h49efa8e_0 defaults
patsy 0.4.1 py36ha3be15e_0 defaults
pbkdf2 1.3 <pip>
pcre 8.41 hc71a17e_0 defaults
pep8 1.7.0 py36h26ade29_0 defaults
pexpect 4.2.1 py36h3b9d41b_0 defaults
pickleshare 0.7.4 py36h63277f8_0 defaults
pillow 4.2.1 py36h9119f52_0 defaults
pip 9.0.1 py36h8ec8b28_3 defaults
pip 9.0.3 <pip>
pixman 0.34.0 h83dc358_2 defaults
pkginfo 1.4.1 py36h215d178_1 defaults
ply 3.10 py36hed35086_0 defaults
ppca 0.0.3 <pip>
prompt_toolkit 1.0.15 py36h17d85b1_0 defaults
protobuf 3.5.2.post1 <pip>
psutil 5.4.0 py36h84c53db_0 defaults
ptyprocess 0.5.2 py36h69acd42_0 defaults
py 1.4.34 py36h0712aa3_1 defaults
pyaes 1.6.1 <pip>
pyasn1 0.4.2 <pip>
pycocotools 2.0 <pip>
pycodestyle 2.3.1 py36hf609f19_0 defaults
pycosat 0.6.3 py36h0a5515d_0 defaults
pycparser 2.18 py36hf9f622e_1 defaults
pycrypto 2.6.1 py36h6998063_1 defaults
pycurl 7.43.0 py36h5e72054_3 defaults
pyflakes 1.6.0 py36h7bd6a15_0 defaults
pygame 1.9.3 <pip>
pyglet 1.2.4 <pip>
pygments 2.2.0 py36h0d3125c_0 defaults
pylint 1.7.4 py36hb9d4533_0 defaults
pymongo 3.6.0 <pip>
pymongo 3.5.1 <pip>
pyodbc 4.0.17 py36h999153c_0 defaults
PyOpenGL 3.1.0 <pip>
pyopenssl 17.2.0 py36h5cc804b_0 defaults
pyparsing 2.2.0 py36hee85983_1 defaults
pyqt 5.6.0 py36h0386399_5 defaults
pysocks 1.6.7 py36hd97a5b1_1 defaults
pytables 3.4.2 py36h3b5282a_2 defaults
pytest 3.2.1 py36h11ad3bb_1 defaults
python 3.6.5 hc3d631a_0 defaults
python-dateutil 2.6.1 py36h88d3b88_1 defaults
pytz 2017.2 py36hc2ccc2a_1 defaults
pywavelets 0.5.2 py36he602eb0_0 defaults
pyyaml 3.12 py36hafb9ca4_1 defaults
pyzmq 16.0.2 py36h3b0cf96_2 defaults
qrcode 5.3 <pip>
qt 5.6.2 h974d657_12 defaults
qtawesome 0.4.4 py36h609ed8c_0 defaults
qtconsole 4.3.1 py36h8f73b5b_0 defaults
qtpy 1.3.1 py36h3691cc8_0 defaults
Quandl 3.2.1 <pip>
readline 7.0 hac23ff0_3 defaults
requests 2.18.4 py36he2e5f8d_1 defaults
requests 2.13.0 <pip>
rope 0.10.5 py36h1f8c17e_0 defaults
ruamel_yaml 0.11.14 py36ha2fb22d_2 defaults
scikit-image 0.13.0 py36had3c07a_1 defaults
scikit-learn 0.19.1 py36h7aa7ec6_0 defaults
scipy 1.0.1 <pip>
scipy 0.19.1 py36h9976243_3 defaults
seaborn 0.8.1 <pip>
seaborn 0.8.0 py36h197244f_0 defaults
setuptools 39.0.1 <pip>
setuptools 36.5.0 py36he42e2e1_0 defaults
simplegeneric 0.8.1 py36h2cb9092_0 defaults
singledispatch 3.4.0.3 py36h7a266c3_0 defaults
sip 4.18.1 py36h51ed4ed_2 defaults
six 1.11.0 py36h372c433_1 defaults
six 1.11.0 <pip>
snowballstemmer 1.2.1 py36h6febd40_0 defaults
sortedcollections 0.5.3 py36h3c761f9_0 defaults
sortedcontainers 1.5.7 py36hdf89491_0 defaults
sphinx 1.6.3 py36he5f0bdb_0 defaults
sphinxcontrib 1.0 py36h6d0f590_1 defaults
sphinxcontrib-websupport 1.0.1 py36hb5cb234_1 defaults
spyder 3.2.4 py36hbe6152b_0 defaults
sqlalchemy 1.1.13 py36hfb5efd7_0 defaults
sqlite 3.22.0 h1bed415_0 defaults
statsmodels 0.8.0 py36h8533d0b_0 defaults
sympy 1.1.1 py36hc6d1c1c_0 defaults
tblib 1.3.2 py36h34cf8b6_0 defaults
tensorboard 1.7.0 <pip>
tensorflow 1.7.0 <pip>
termcolor 1.1.0 <pip>
terminado 0.6 py36ha25a19f_0 defaults
testpath 0.3.1 py36h8cadb63_0 defaults
Theano 1.0.1 <pip>
tk 8.6.7 h5979e9b_1 defaults
toolz 0.8.2 py36h81f2dff_0 defaults
tornado 4.5.2 py36h1283b2a_0 defaults
traitlets 4.3.2 py36h674d592_0 defaults
typing 3.6.2 py36h7da032a_0 defaults
tzlocal 1.5.1 <pip>
umap-learn 0.2.3 <pip>
unicodecsv 0.14.1 py36ha668878_0 defaults
unixodbc 2.3.4 hc36303a_1 defaults
urllib3 1.22 py36hbe7ace6_0 defaults
virtualenv 15.1.0 py36_0 defaults
wcwidth 0.1.7 py36hdf4376a_0 defaults
webencodings 0.5.1 py36h800622e_1 defaults
websocket-client 0.40.0 <pip>
Werkzeug 0.14.1 <pip>
werkzeug 0.12.2 py36hc703753_0 defaults
wheel 0.31.0 <pip>
wheel 0.29.0 py36he7f4e38_1 defaults
widgetsnbextension 3.0.2 py36hd01bb71_1 defaults
wrapt 1.10.11 py36h28b7045_0 defaults
xlrd 1.1.0 py36h1db9f0c_1 defaults
xlsxwriter 1.0.2 py36h3de1aca_0 defaults
xlwt 1.3.0 py36h7b00a1f_0 defaults
xz 5.2.3 h2bcbf08_1 defaults
yaml 0.1.7 h96e3832_1 defaults
zeromq 4.2.2 hb0b69da_1 defaults
zict 0.1.3 py36h3a3bf81_0 defaults
zlib 1.2.11 hfbfcf68_1 defaults
Please do not do that.
@mingwandroid this occurred for me today, even though I had just updated conda (to 4.5.4-py27_0). I wonder if the problem is now back and this issue should be reopened?
As an aside, I did manage to get myself up and running by setting the env variable:
> export LD_PRELOAD=/home/duro/anaconda2/envs/cdat80/lib/libstdc++.so.6.0.24
I've run into this issue with Travis CI ... during regression testing I get the following when importing scipy, installed using miniconda.
The OS provided by Travis is Ubuntu 14.04.5 LTS. From there, I'm installing conda 4.5.8, and creating a test environment using Python 3.6.6. Full details at https://travis-ci.org/sandialabs/samlab/builds/405475819
Not sure whether this is a conda issue or Ubuntu 14 showing its age. Any suggestions welcome.
Thanks in advance, Tim
Traceback (most recent call last):
File "/home/travis/miniconda/envs/test-environment/lib/python3.6/site-packages/behave/model.py", line 1329, in run
match.run(runner.context)
File "/home/travis/miniconda/envs/test-environment/lib/python3.6/site-packages/behave/matchers.py", line 98, in run
self.func(context, *args, **kwargs)
File "features/steps/examples.py", line 15, in step_impl
samlab.example.iris.create(context.database, context.fs)
File "/home/travis/build/sandialabs/samlab/samlab/example/iris.py", line 33, in create
import sklearn.datasets
File "/home/travis/miniconda/envs/test-environment/lib/python3.6/site-packages/sklearn/__init__.py", line 134, in <module>
from .base import clone
File "/home/travis/miniconda/envs/test-environment/lib/python3.6/site-packages/sklearn/base.py", line 11, in <module>
from scipy import sparse
File "/home/travis/miniconda/envs/test-environment/lib/python3.6/site-packages/scipy/sparse/__init__.py", line 229, in <module>
from .csr import *
File "/home/travis/miniconda/envs/test-environment/lib/python3.6/site-packages/scipy/sparse/csr.py", line 15, in <module>
from ._sparsetools import csr_tocsc, csr_tobsr, csr_count_blocks, \
ImportError: /usr/lib/x86_64-linux-gnu/libstdc++.so.6: version `CXXABI_1.3.9' not found (required by /home/travis/miniconda/envs/test-environment/lib/python3.6/site-packages/scipy/sparse/_sparsetools.cpython-36m-x86_64-linux-gnu.so)
I cannot reproduce this (but I never use PIP), I suspect one of your PIP installed dependencies is being imported before scipy and is linking to your system libstdc++ and that prevents our libstdc++ from being used.
I can't really do much about PIP binaries, it's kind of like the wild-west of package building over there.
Please try to stick to using packages from the Anaconda Distribution if you can.
As a potential workaround you could try setting LD_LIBRARY_PATH
to your conda environment's lib dir and see if that causes the PIP installed things to use our libs. If they do, whether they work ok or not is pot-luck I am afraid.
BTW we do fully support Ubuntu 14.04 at present.
There is an issue in anaconda where libstdc++ is not linking to version 6.0.24 but rather to 6.0.19. If you manually relink libstdc++.so and libstdc++.so.6 to libstdc++.so.6.0.24 it should work.
how to relink manually?????
@shuvamklu, please ignore that comment it is incorrect.
Just for everyone else having the problem, here the solution:
example:
rm /home/<YOUR_USER>/anaconda3/lib/libstdc++.so.6.0.21
ln -s /home/<YOUR_USER>/anaconda3/lib/libstdc++.so.6.0.25 /home/<YOUR_USER>/anaconda3/lib/libstdc++.so.6.0.21
That did not help at all, 1) running apt-get update
to update all the package lists,
followed by apt-get upgrade
to update all the installed software to the latest versions and,
2) running pip install opencv-contrib-python
solved the problem.
For those facing this issue under linux CentOS, the solution of this thread worked for me: https://stackoverflow.com/questions/48831881/centos-7-libstdc-so-6-version-cxxabi-1-3-9-not-found
It seems to me that it's not possible to use some dynamically linked libraries (in my case /usr/lib/x86_64-linux-gnu/libmodplug.so.1) on ubuntu 17.04 because the libstdc++.so.6 shipped with conda does not include the CXXABI_1.3.9 symbols.