spack/var/spack/repos/builtin/packages/cuda/package.py
Tim Fuller 5bc15b2d9a find_libraries searches lib and lib64 before prefix (#11958)
The default library search for a package checks the lib/ and lib64/
directories for libraries before the root prefix, in order to save
time when searching for libraries provided by externals (which e.g.
may have '/usr/' as their root).

This moves that logic into the "find_libraries" utility method so
packages implementing their own custom library search logic can
benefit from it.

This also updates packages which appear to be replicating this logic
exactly, replacing it with a single call to "find_libraries".
2019-07-12 17:46:47 -07:00

76 lines
3.7 KiB
Python

# Copyright 2013-2019 Lawrence Livermore National Security, LLC and other
# Spack Project Developers. See the top-level COPYRIGHT file for details.
#
# SPDX-License-Identifier: (Apache-2.0 OR MIT)
from spack import *
from glob import glob
from llnl.util.filesystem import LibraryList
import os
class Cuda(Package):
"""CUDA is a parallel computing platform and programming model invented
by NVIDIA. It enables dramatic increases in computing performance by
harnessing the power of the graphics processing unit (GPU).
Note: This package does not currently install the drivers necessary
to run CUDA. These will need to be installed manually. See:
https://docs.nvidia.com/cuda/ for details."""
homepage = "https://developer.nvidia.com/cuda-zone"
version('10.0.130', sha256='92351f0e4346694d0fcb4ea1539856c9eb82060c25654463bfd8574ec35ee39a', expand=False,
url="https://developer.nvidia.com/compute/cuda/10.0/Prod/local_installers/cuda_10.0.130_410.48_linux")
version('9.2.88', 'dd6e33e10d32a29914b7700c7b3d1ca0', expand=False,
url="https://developer.nvidia.com/compute/cuda/9.2/Prod/local_installers/cuda_9.2.88_396.26_linux")
version('9.1.85', '67a5c3933109507df6b68f80650b4b4a', expand=False,
url="https://developer.nvidia.com/compute/cuda/9.1/Prod/local_installers/cuda_9.1.85_387.26_linux")
version('9.0.176', '7a00187b2ce5c5e350e68882f42dd507', expand=False,
url="https://developer.nvidia.com/compute/cuda/9.0/Prod/local_installers/cuda_9.0.176_384.81_linux-run")
version('8.0.61', '33e1bd980e91af4e55f3ef835c103f9b', expand=False,
url="https://developer.nvidia.com/compute/cuda/8.0/Prod2/local_installers/cuda_8.0.61_375.26_linux-run")
version('8.0.44', '6dca912f9b7e2b7569b0074a41713640', expand=False,
url="https://developer.nvidia.com/compute/cuda/8.0/prod/local_installers/cuda_8.0.44_linux-run")
version('7.5.18', '4b3bcecf0dfc35928a0898793cf3e4c6', expand=False,
url="http://developer.download.nvidia.com/compute/cuda/7.5/Prod/local_installers/cuda_7.5.18_linux.run")
version('6.5.14', '90b1b8f77313600cc294d9271741f4da', expand=False,
url="http://developer.download.nvidia.com/compute/cuda/6_5/rel/installers/cuda_6.5.14_linux_64.run")
def setup_environment(self, spack_env, run_env):
run_env.set('CUDA_HOME', self.prefix)
def install(self, spec, prefix):
runfile = glob(join_path(self.stage.source_path, 'cuda*_linux*'))[0]
chmod = which('chmod')
chmod('+x', runfile)
runfile = which(runfile)
# Note: NVIDIA does not officially support many newer versions of
# compilers. For example, on CentOS 6, you must use GCC 4.4.7 or
# older. See:
# http://docs.nvidia.com/cuda/cuda-installation-guide-linux/#system-requirements
# https://gist.github.com/ax3l/9489132
# for details.
runfile(
'--silent', # disable interactive prompts
'--verbose', # create verbose log file
'--override', # override compiler version checks
'--toolkit', # install CUDA Toolkit
'--toolkitpath=%s' % prefix
)
@property
def libs(self):
libs = find_libraries('libcuda', root=self.prefix, shared=True,
recursive=True)
filtered_libs = []
# CUDA 10.0 provides Compatability libraries for running newer versions
# of CUDA with older drivers. These do not work with newer drivers.
for lib in libs:
if 'compat' not in lib.split(os.sep):
filtered_libs.append(lib)
return LibraryList(filtered_libs)