<html><body><p><font size="2">Current build instructions for ppc64le are here: </font><a href="https://github.com/ppc64le/build-scripts/tree/master/caffe"><font size="2">https://github.com/ppc64le/build-scripts/tree/master/caffe</font></a><br><font size="2"><br><br>gerrit<br><br>--<br>"Only those who will risk going too far can possibly find out how far one can go." ~ T. S. Eliot<br><br>Gerrit Huizenga, STSM<br>Power Open Source Ecosystem Lead<br>gerrit@us.ibm.com</font><br><br><img width="16" height="16" src="cid:1__=07BB0BFCDFC552B98f9e8a93df938690918c07B@" border="0" alt="Inactive hide details for Frederic DUMOULIN ---08/01/2017 07:38:47 AM---Hi, I'm trying to compil Caffe software following this "><font size="2" color="#424282">Frederic DUMOULIN ---08/01/2017 07:38:47 AM---Hi, I'm trying to compil Caffe software following this steps</font><br><br><font size="2" color="#5F5F5F">From: </font><font size="2">Frederic DUMOULIN <frederic.dumoulin@mipsology.com></font><br><font size="2" color="#5F5F5F">To: </font><font size="2">linuxppc-users@lists.ozlabs.org</font><br><font size="2" color="#5F5F5F">Date: </font><font size="2">08/01/2017 07:38 AM</font><br><font size="2" color="#5F5F5F">Subject: </font><font size="2">[Linuxppc-users] Caffe compilation failed</font><br><font size="2" color="#5F5F5F">Sent by: </font><font size="2">"Linuxppc-users" <linuxppc-users-bounces+gerrit=us.ibm.com@lists.ozlabs.org></font><br><hr width="100%" size="2" align="left" noshade style="color:#8091A5; "><br><br><br><tt><font size="2">Hi,<br><br>I'm trying to compil Caffe software following this steps <br></font></tt><tt><font size="2"><a href="http://caffe.berkeleyvision.org/installation.html#compilation">http://caffe.berkeleyvision.org/installation.html#compilation</a></font></tt><tt><font size="2">.<br>I need the CPU_ONLY flag to be OFF and USE_CUDNN to ON. In this mode <br>cuda files are compiled.<br><br>It seems that the cuda files can not be compiled with <br>/opt/at10.0/bin/gcc since the version is 6.3.1 whereas cuda headers <br>doen't support gcc > 5:<br>[ 1%] Building NVCC (Device) object <br>src/caffe/CMakeFiles/cuda_compile.dir/solvers/cuda_compile_generated_adagrad_solver.cu.o<br>nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are <br>deprecated, and may be removed in a future release (Use <br>-Wno-deprecated-gpu-targets to suppress warning).<br>In file included from /usr/local/cuda/include/cuda_runtime.h:78:0,<br> from <command-line>:0:<br>/usr/local/cuda/include/host_config.h:119:2: error: #error -- <br>unsupported GNU version! gcc versions later than 5 are not supported!<br> #error -- unsupported GNU version! gcc versions later than 5 are not <br>supported!<br> ^~~~~<br>CMake Error at cuda_compile_generated_adagrad_solver.cu.o.cmake:207 <br>(message):<br><br>Here is the cmake configuration:<br>-- Boost version: 1.61.0<br>-- Found the following Boost libraries:<br>-- system<br>-- thread<br>-- filesystem<br>-- chrono<br>-- date_time<br>-- atomic<br>-- Found gflags (include: /usr/include, library: <br>/usr/lib/powerpc64le-linux-gnu/libgflags.so)<br>-- Found glog (include: /usr/include, library: <br>/usr/lib/powerpc64le-linux-gnu/libglog.so)<br>-- Found PROTOBUF Compiler: /usr/bin/protoc<br>-- Found lmdb (include: /usr/include, library: <br>/usr/lib/powerpc64le-linux-gnu/liblmdb.so)<br>-- Found LevelDB (include: /usr/include, library: <br>/usr/lib/powerpc64le-linux-gnu/libleveldb.so)<br>-- Found Snappy (include: /usr/include, library: <br>/usr/lib/powerpc64le-linux-gnu/libsnappy.so)<br>-- CUDA detected: 8.0<br>-- Found cuDNN: ver. 5.0.5 found (include: /usr/local/cuda/include, <br>library: /usr/local/cuda/lib64/libcudnn.so)<br>-- Automatic GPU detection failed. Building for all known architectures.<br>-- Added CUDA NVCC flags for: sm_20 sm_21 sm_30 sm_35 sm_50<br>-- OpenCV found (/usr/share/OpenCV)<br>-- Found Atlas (include: /usr/include, library: /usr/lib/libatlas.so)<br>-- NumPy ver. 1.11.0 found (include: <br>/usr/lib/python2.7/dist-packages/numpy/core/include)<br>-- Boost version: 1.61.0<br>-- Found the following Boost libraries:<br>-- python<br>-- Detected Doxygen OUTPUT_DIRECTORY: ./doxygen/<br>-- <br>-- ******************* Caffe Configuration Summary *******************<br>-- General:<br>-- Version : 1.0.0-rc3<br>-- Git : rc3-176-g790c2db<br>-- System : Linux<br>-- C++ compiler : /opt/at10.0/bin/c++<br>-- Release CXX flags : -O3 -DNDEBUG -fPIC -Wall -Wno-sign-compare <br>-Wno-uninitialized<br>-- Debug CXX flags : -g -fPIC -Wall -Wno-sign-compare <br>-Wno-uninitialized<br>-- Build type : Release<br>-- <br>-- BUILD_SHARED_LIBS : ON<br>-- BUILD_python : ON<br>-- BUILD_matlab : OFF<br>-- BUILD_docs : ON<br>-- CPU_ONLY : OFF<br>-- USE_OPENCV : ON<br>-- USE_LEVELDB : ON<br>-- USE_LMDB : ON<br>-- ALLOW_LMDB_NOLOCK : OFF<br>-- <br>-- Dependencies:<br>-- BLAS : Yes (Atlas)<br>-- Boost : Yes (ver. 1.61)<br>-- glog : Yes<br>-- gflags : Yes<br>-- protobuf : Yes (ver. 2.6.1)<br>-- lmdb : Yes (ver. 0.9.17)<br>-- LevelDB : Yes (ver. 1.18)<br>-- Snappy : Yes (ver. 1.1.3)<br>-- OpenCV : Yes (ver. 2.4.9.1)<br>-- CUDA : Yes (ver. 8.0)<br>-- <br>-- NVIDIA CUDA:<br>-- Target GPU(s) : Auto<br>-- GPU arch(s) : sm_20 sm_21 sm_30 sm_35 sm_50<br>-- cuDNN : Yes (ver. 5.0.5)<br><br>How can it be solved ?<br><br>Thanks,<br>Fred<br><br>_______________________________________________<br>Linuxppc-users mailing list<br>Linuxppc-users@lists.ozlabs.org<br></font></tt><tt><font size="2"><a href="https://lists.ozlabs.org/listinfo/linuxppc-users">https://lists.ozlabs.org/listinfo/linuxppc-users</a></font></tt><tt><font size="2"><br></font></tt><br><BR>
</body></html>