Update documentation for OSX issues
Jon007 opened this issue · 3 comments
Effectively the instructions on http://torch.ch/docs/getting-started.html should come with a warning that they don't work on as stated on current OSX builds.
We provide a simple installation process for Torch on Mac OS X and Ubuntu 12+:
Torch can be installed to your home folder in ~/torch by running these three commands:
in a terminal, run the commands WITHOUT sudo
git clone https://github.com/torch/distro.git ~/torch --recursive
cd ~/torch; bash install-deps;
./install.sh
This gives build error:
nvcc fatal : The version ('90000') of the host compiler ('Apple clang') is not supported
There are similar issues reported for this as per pytorch/pytorch#3047
it may be necessary to download Command Line Tool for 8.3.2 from https://developer.apple.com/download/more/
Run 'sudo xcode-select --switch /Library/Developer/CommandLineTools'
and then try the torch install again, however there will be further errors eg:
error: specified alignment (4) is different from alignment (2) specified on a previous declaration
so it's not quite a simple installation process
same problem here
This is so ridiculous.
nvcc fatal : The version ('90000') of the host compiler ('Apple clang') is not supported
computer:pytorch ~$ nvcc --version
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2017 NVIDIA Corporation
Built on Fri_Nov__3_21:08:57_CDT_2017
Cuda compilation tools, release 9.1, V9.1.85
computer:pytorch ~$ pkgutil --pkg-info=com.apple.pkg.CLTools_Executables
package-id: com.apple.pkg.CLTools_Executables
version: 9.2.0.0.1.1510905681
volume: /
location: /
install-time: 1514324198
groups: com.apple.FindSystemFiles.pkg-group
computer:pytorch ~$ clang --version
Apple LLVM version 9.0.0 (clang-900.0.39.2)
Target: x86_64-apple-darwin16.7.0
Thread model: posix
InstalledDir: /Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin
Same problem, and I see tickets getting closed in related repos over false workarounds, I copy:
It's a serious issue leaving an entire platform (OS X) out of the support map and the solution above was not fully validated. I have tried all plausible downgrades (it's suggested to use 8.2 from "Dec 13" as "solution") of the CLI and none work; this "fix" was probably not a fix at all, e.g. not what made the code eventually build for the test case. This is a seriously overlooked nvidia/xcode compatibility issue I've seen in other related repos, claiming the "fix" to be in another CLI version which doesn't actually work either.
Specifically with the 8.2 CLI (actually from Dec 12, not 13):
clang --version
Apple LLVM version 8.0.0 (clang-800.0.42.1)
Target: x86_64-apple-darwin16.7.0
Thread model: posix
InstalledDir: /Library/Developer/CommandLineTools/usr/bin
Errors:
nvcc fatal : The version ('80000') of the host compiler ('Apple clang') is not supported
...
CMake Error at ATen_generated_THCStorageCopy.cu.o.cmake:207 (message):
...
etc.
Additionally, why the 10.9 target? Where was that inherited from? Should we take that as a hint as to how long ago has this code been actually tested for OS X?