Skip to content

Commit

Permalink
cuda: update recipe, deploy 11.4.2 (spack#1339)
Browse files Browse the repository at this point in the history 
* cuda: new versions 11.4.1, 11.4.2 and 11.5.0.
* deploy: cuda: deploy 11.4.2, drop 10.1.243.
  • Loading branch information
@olupton
olupton committed Oct 28, 2021
1 parent 3d67e21 commit 76bfcbe
Show file tree
Hide file tree
Showing 4 changed files with 137 additions and 81 deletions.
2 changes: 0 additions & 2 deletions deploy/configs/packages.yaml
View file
Original file line number Diff line number Diff line change
Expand Up @@ -13,8 +13,6 @@ packages:
variants: +pic
cmake:
version: [3.15.7]
cuda:
version: [11.1.0]
curl:
version: [7.29.0]
externals:
Expand Down
2 changes: 1 addition & 1 deletion deploy/environments/externals.yaml
View file
Original file line number Diff line number Diff line change
Expand Up @@ -70,8 +70,8 @@ spack:
- cgal
- cli-tools
- cmake@3.15.7
- cuda@10.1.243
- cuda@11.0.2
- cuda@11.4.2
- cudnn@8.0.3.33-11.0-linux-x64
- darshan-runtime
- darshan-util
Expand Down
146 changes: 75 additions & 71 deletions lib/spack/spack/build_systems/cuda.py
View file
Original file line number Diff line number Diff line change
@@ -1,19 +1,18 @@
# Copyright 2013-2020 Lawrence Livermore National Security, LLC and other
# Copyright 2013-2021 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.package import PackageBase
from spack.directives import depends_on, variant, conflicts

import spack.variant
from spack.directives import conflicts, depends_on, variant
from spack.package import PackageBase


class CudaPackage(PackageBase):
"""Auxiliary class which contains CUDA variant, dependencies and conflicts
and is meant to unify and facilitate its usage.
Maintainers: ax3l, Rombur
Maintainers: ax3l, Rombur, davidbeckingsale
"""

# https://docs.nvidia.com/cuda/cuda-compiler-driver-nvcc/index.html#gpu-feature-list
Expand Down Expand Up @@ -79,47 +78,63 @@ def cuda_flags(arch_list):
depends_on('cuda@11.0:', when='cuda_arch=80')
depends_on('cuda@11.1:', when='cuda_arch=86')

# There are at least three cases to be aware of for compiler conflicts
# 1. Linux x86_64
# 2. Linux ppc64le
# 3. Mac OS X
# CUDA-compiler conflicts are version-to-version specific and are
# difficult to express with the current Spack conflict syntax
# From the NVIDIA install guide we know of conflicts for particular
# platforms (linux, darwin), architectures (x86, powerpc) and compilers
# (gcc, clang). We don't restrict %gcc and %clang conflicts to
# platform=linux, since they should also apply to platform=cray, and may
# apply to platform=darwin. We currently do not provide conflicts for
# platform=darwin with %apple-clang.

# Linux x86_64 compiler conflicts from here:
# https://gist.github.com/ax3l/9489132
arch_platform = ' target=x86_64: platform=linux'
conflicts('%gcc@5:', when='+cuda ^cuda@:7.5' + arch_platform)
conflicts('%gcc@6:', when='+cuda ^cuda@:8' + arch_platform)
conflicts('%gcc@7:', when='+cuda ^cuda@:9.1' + arch_platform)
conflicts('%gcc@8:', when='+cuda ^cuda@:10.0.130' + arch_platform)
conflicts('%gcc@9:', when='+cuda ^cuda@:10.2.89' + arch_platform)
conflicts('%gcc@:4', when='+cuda ^cuda@11.0.2:' + arch_platform)
conflicts('%gcc@10:', when='+cuda ^cuda@:11.0.2' + arch_platform)
conflicts('%gcc@11:', when='+cuda ^cuda@:11.1.0' + arch_platform)
conflicts('%pgi@:14.8', when='+cuda ^cuda@:7.0.27' + arch_platform)
conflicts('%pgi@:15.3,15.5:', when='+cuda ^cuda@7.5' + arch_platform)
conflicts('%pgi@:16.2,16.0:16.3', when='+cuda ^cuda@8' + arch_platform)
conflicts('%pgi@:15,18:', when='+cuda ^cuda@9.0:9.1' + arch_platform)
conflicts('%pgi@:16,19:', when='+cuda ^cuda@9.2.88:10' + arch_platform)
conflicts('%pgi@:17,20:',
when='+cuda ^cuda@10.1.105:10.2.89' + arch_platform)
conflicts('%pgi@:17,21:',
when='+cuda ^cuda@11.0.2:11.1.0' + arch_platform)
conflicts('%clang@:3.4', when='+cuda ^cuda@:7.5' + arch_platform)
conflicts('%clang@:3.7,4:',
when='+cuda ^cuda@8.0:9.0' + arch_platform)
conflicts('%clang@:3.7,4.1:',
when='+cuda ^cuda@9.1' + arch_platform)
conflicts('%clang@:3.7,5.1:', when='+cuda ^cuda@9.2' + arch_platform)
conflicts('%clang@:3.7,6.1:', when='+cuda ^cuda@10.0.130' + arch_platform)
conflicts('%clang@:3.7,7.1:', when='+cuda ^cuda@10.1.105' + arch_platform)

# GCC
# According to
# https://github.com/spack/spack/pull/25054#issuecomment-886531664
# these conflicts are valid independently from the architecture

# minimum supported versions
conflicts('%gcc@:4', when='+cuda ^cuda@11.0:')
conflicts('%gcc@:5', when='+cuda ^cuda@11.4:')

# maximum supported version
# NOTE:
# in order to not constrain future cuda version to old gcc versions,
# it has been decided to use an upper bound for the latest version.
# This implies that the last one in the list has to be updated at
# each release of a new cuda minor version.
conflicts('%gcc@10:', when='+cuda ^cuda@:11.0')
conflicts('%gcc@11:', when='+cuda ^cuda@:11.4.0')
conflicts('%gcc@12:', when='+cuda ^cuda@:11.5.0')
conflicts('%clang@12:', when='+cuda ^cuda@:11.4.0')
conflicts('%clang@13:', when='+cuda ^cuda@:11.5.0')

# https://gist.github.com/ax3l/9489132#gistcomment-3860114
conflicts('%gcc@10', when='+cuda ^cuda@:11.4.0')
conflicts('%gcc@5:', when='+cuda ^cuda@:7.5 target=x86_64:')
conflicts('%gcc@6:', when='+cuda ^cuda@:8 target=x86_64:')
conflicts('%gcc@7:', when='+cuda ^cuda@:9.1 target=x86_64:')
conflicts('%gcc@8:', when='+cuda ^cuda@:10.0.130 target=x86_64:')
conflicts('%gcc@9:', when='+cuda ^cuda@:10.2.89 target=x86_64:')
conflicts('%pgi@:14.8', when='+cuda ^cuda@:7.0.27 target=x86_64:')
conflicts('%pgi@:15.3,15.5:', when='+cuda ^cuda@7.5 target=x86_64:')
conflicts('%pgi@:16.2,16.0:16.3', when='+cuda ^cuda@8 target=x86_64:')
conflicts('%pgi@:15,18:', when='+cuda ^cuda@9.0:9.1 target=x86_64:')
conflicts('%pgi@:16,19:', when='+cuda ^cuda@9.2.88:10 target=x86_64:')
conflicts('%pgi@:17,20:', when='+cuda ^cuda@10.1.105:10.2.89 target=x86_64:')
conflicts('%pgi@:17,21:', when='+cuda ^cuda@11.0.2:11.1.0 target=x86_64:')
conflicts('%clang@:3.4', when='+cuda ^cuda@:7.5 target=x86_64:')
conflicts('%clang@:3.7,4:', when='+cuda ^cuda@8.0:9.0 target=x86_64:')
conflicts('%clang@:3.7,4.1:', when='+cuda ^cuda@9.1 target=x86_64:')
conflicts('%clang@:3.7,5.1:', when='+cuda ^cuda@9.2 target=x86_64:')
conflicts('%clang@:3.7,6.1:', when='+cuda ^cuda@10.0.130 target=x86_64:')
conflicts('%clang@:3.7,7.1:', when='+cuda ^cuda@10.1.105 target=x86_64:')
conflicts('%clang@:3.7,8.1:',
when='+cuda ^cuda@10.1.105:10.1.243' + arch_platform)
conflicts('%clang@:3.2,9:', when='+cuda ^cuda@10.2.89' + arch_platform)
conflicts('%clang@:5', when='+cuda ^cuda@11.0.2:' + arch_platform)
conflicts('%clang@10:', when='+cuda ^cuda@:11.0.2' + arch_platform)
conflicts('%clang@11:', when='+cuda ^cuda@:11.1.0' + arch_platform)
when='+cuda ^cuda@10.1.105:10.1.243 target=x86_64:')
conflicts('%clang@:3.2,9:', when='+cuda ^cuda@10.2.89 target=x86_64:')
conflicts('%clang@:5', when='+cuda ^cuda@11.0.2: target=x86_64:')
conflicts('%clang@10:', when='+cuda ^cuda@:11.0.3 target=x86_64:')
conflicts('%clang@11:', when='+cuda ^cuda@:11.1.0 target=x86_64:')

# x86_64 vs. ppc64le differ according to NVidia docs
# Linux ppc64le compiler conflicts from Table from the docs below:
Expand All @@ -129,27 +144,23 @@ def cuda_flags(arch_list):
# https://docs.nvidia.com/cuda/archive/9.0/cuda-installation-guide-linux/index.html
# https://docs.nvidia.com/cuda/archive/8.0/cuda-installation-guide-linux/index.html

arch_platform = ' target=ppc64le: platform=linux'
# information prior to CUDA 9 difficult to find
conflicts('%gcc@6:', when='+cuda ^cuda@:9' + arch_platform)
conflicts('%gcc@8:', when='+cuda ^cuda@:10.0.130' + arch_platform)
conflicts('%gcc@9:', when='+cuda ^cuda@:10.1.243' + arch_platform)
conflicts('%gcc@6:', when='+cuda ^cuda@:9 target=ppc64le:')
conflicts('%gcc@8:', when='+cuda ^cuda@:10.0.130 target=ppc64le:')
conflicts('%gcc@9:', when='+cuda ^cuda@:10.1.243 target=ppc64le:')
# officially, CUDA 11.0.2 only supports the system GCC 8.3 on ppc64le
conflicts('%gcc@:4', when='+cuda ^cuda@11.0.2:' + arch_platform)
conflicts('%gcc@10:', when='+cuda ^cuda@:11.0.2' + arch_platform)
conflicts('%gcc@11:', when='+cuda ^cuda@:11.1.0' + arch_platform)
conflicts('%pgi', when='+cuda ^cuda@:8' + arch_platform)
conflicts('%pgi@:16', when='+cuda ^cuda@:9.1.185' + arch_platform)
conflicts('%pgi@:17', when='+cuda ^cuda@:10' + arch_platform)
conflicts('%clang@4:', when='+cuda ^cuda@:9.0.176' + arch_platform)
conflicts('%clang@5:', when='+cuda ^cuda@:9.1' + arch_platform)
conflicts('%clang@6:', when='+cuda ^cuda@:9.2' + arch_platform)
conflicts('%clang@7:', when='+cuda ^cuda@10.0.130' + arch_platform)
conflicts('%clang@7.1:', when='+cuda ^cuda@:10.1.105' + arch_platform)
conflicts('%clang@8.1:', when='+cuda ^cuda@:10.2.89' + arch_platform)
conflicts('%clang@:5', when='+cuda ^cuda@11.0.2:' + arch_platform)
conflicts('%clang@10:', when='+cuda ^cuda@:11.0.2' + arch_platform)
conflicts('%clang@11:', when='+cuda ^cuda@:11.1.0' + arch_platform)
conflicts('%pgi', when='+cuda ^cuda@:8 target=ppc64le:')
conflicts('%pgi@:16', when='+cuda ^cuda@:9.1.185 target=ppc64le:')
conflicts('%pgi@:17', when='+cuda ^cuda@:10 target=ppc64le:')
conflicts('%clang@4:', when='+cuda ^cuda@:9.0.176 target=ppc64le:')
conflicts('%clang@5:', when='+cuda ^cuda@:9.1 target=ppc64le:')
conflicts('%clang@6:', when='+cuda ^cuda@:9.2 target=ppc64le:')
conflicts('%clang@7:', when='+cuda ^cuda@10.0.130 target=ppc64le:')
conflicts('%clang@7.1:', when='+cuda ^cuda@:10.1.105 target=ppc64le:')
conflicts('%clang@8.1:', when='+cuda ^cuda@:10.2.89 target=ppc64le:')
conflicts('%clang@:5', when='+cuda ^cuda@11.0.2: target=ppc64le:')
conflicts('%clang@10:', when='+cuda ^cuda@:11.0.2 target=ppc64le:')
conflicts('%clang@11:', when='+cuda ^cuda@:11.1.0 target=ppc64le:')

# Intel is mostly relevant for x86_64 Linux, even though it also
# exists for Mac OS X. No information prior to CUDA 3.2 or Intel 11.1
Expand All @@ -171,16 +182,9 @@ def cuda_flags(arch_list):
conflicts('%xl@:12,14:15,17:', when='+cuda ^cuda@9.2')
conflicts('%xl@:12,17:', when='+cuda ^cuda@:11.1.0')

# Mac OS X
# platform = ' platform=darwin'
# Apple XCode clang vs. LLVM clang are difficult to specify
# with spack syntax. Xcode clang name is `clang@x.y.z-apple`
# which precludes ranges being specified. We have proposed
# rename XCode clang to `clang@apple-x.y.z` or even
# `clang-apple@x.y.z as a possible fix.
# Compiler conflicts will be eventual taken from here:
# https://docs.nvidia.com/cuda/cuda-installation-guide-mac-os-x/index.html#abstract
conflicts('platform=darwin', when='+cuda ^cuda@11.0.2:')
# Darwin.
# TODO: add missing conflicts for %apple-clang cuda@:10
conflicts('platform=darwin', when='+cuda ^cuda@11.0.2: ')

# Make sure cuda_arch can not be used without +cuda
for value in cuda_arch_values:
Expand Down
Loading

0 comments on commit 76bfcbe

Please sign in to comment.