Skip to content
This repository was archived by the owner on Dec 17, 2020. It is now read-only.

Commit 2983b80

Browse files
committed
Merge branch 'develop' for 2.9.00
Part of Kokkos C++ Performance Portability Programming EcoSystem 2.9
2 parents 5d6e7fb + 4c6cb80 commit 2983b80

File tree

499 files changed

+37838
-4266
lines changed

Some content is hidden

Large Commits have some content hidden by default. Use the searchbox below for content that may be hidden.

499 files changed

+37838
-4266
lines changed

.jenkins

+112
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,112 @@
1+
pipeline {
2+
agent {
3+
label 'docker'
4+
}
5+
environment {
6+
CCACHE_DIR = '/tmp/ccache'
7+
CCACHE_MAXSIZE = '10G'
8+
CCACHE_CPP2 = 'true'
9+
}
10+
stages {
11+
stage('Build') {
12+
parallel {
13+
stage('CUDA-9.2-Clang') {
14+
agent {
15+
dockerfile {
16+
filename 'Dockerfile.clang'
17+
dir 'scripts/docker'
18+
additionalBuildArgs '--pull'
19+
label 'nvidia-docker && volta'
20+
args '-v /tmp/ccache.kokkos:/tmp/ccache'
21+
}
22+
}
23+
environment {
24+
OPTIONS = '--with-options=compiler_warnings --arch=Volta70 --compiler=/usr/local/bin/clang++ --with-cuda --with-cuda-options=enable_lambda'
25+
}
26+
steps {
27+
sh 'ln -s /usr/bin/ccache /usr/local/bin/clang++'
28+
sh 'ccache --zero-stats'
29+
sh 'rm -rf build && mkdir -p build && cd build && ../generate_makefile.bash $OPTIONS && make test -j8'
30+
}
31+
post {
32+
always {
33+
sh 'ccache --show-stats'
34+
}
35+
}
36+
}
37+
stage("CUDA-9.2-NVCC") {
38+
agent {
39+
dockerfile {
40+
filename 'Dockerfile.nvcc'
41+
dir 'scripts/docker'
42+
additionalBuildArgs '--pull --build-arg BASE=nvidia/cuda:9.2-devel'
43+
label 'nvidia-docker && volta'
44+
args '-v /tmp/ccache.kokkos:/tmp/ccache'
45+
}
46+
}
47+
environment {
48+
OPTIONS = '--with-options=compiler_warnings --arch=Volta70 --with-cuda --with-cuda-options=enable_lambda'
49+
}
50+
steps {
51+
sh 'ln -s /usr/bin/ccache /usr/local/bin/g++'
52+
sh 'ccache --zero-stats'
53+
sh 'rm -rf build && mkdir -p build && cd build && ../generate_makefile.bash $OPTIONS && make test -j8'
54+
}
55+
post {
56+
always {
57+
sh 'ccache --show-stats'
58+
}
59+
}
60+
}
61+
stage("CUDA-10.1-NVCC") {
62+
agent {
63+
dockerfile {
64+
filename 'Dockerfile.nvcc'
65+
dir 'scripts/docker'
66+
additionalBuildArgs '--pull --build-arg BASE=nvidia/cuda:10.1-devel'
67+
label 'nvidia-docker && volta'
68+
args '-v /tmp/ccache.kokkos:/tmp/ccache'
69+
}
70+
}
71+
environment {
72+
OPTIONS = '--with-options=compiler_warnings --arch=Volta70 --with-cuda --with-cuda-options=enable_lambda,rdc,force_uvm'
73+
}
74+
steps {
75+
sh 'ln -s /usr/bin/ccache /usr/local/bin/g++'
76+
sh 'ccache --zero-stats'
77+
sh 'rm -rf build && mkdir -p build && cd build && ../generate_makefile.bash $OPTIONS && make test -j8'
78+
}
79+
post {
80+
always {
81+
sh 'ccache --show-stats'
82+
}
83+
}
84+
}
85+
stage("CUDA-10.1-NVCC-DEBUG") {
86+
agent {
87+
dockerfile {
88+
filename 'Dockerfile.nvcc'
89+
dir 'scripts/docker'
90+
additionalBuildArgs '--pull --build-arg BASE=nvidia/cuda:10.1-devel'
91+
label 'nvidia-docker && volta'
92+
args '-v /tmp/ccache.kokkos:/tmp/ccache'
93+
}
94+
}
95+
environment {
96+
OPTIONS = '--with-options=compiler_warnings --arch=Volta70--with-cuda --with-cuda-options=enable_lambda --debug'
97+
}
98+
steps {
99+
sh 'ln -s /usr/bin/ccache /usr/local/bin/g++'
100+
sh 'ccache --zero-stats'
101+
sh 'rm -rf build && mkdir -p build && cd build && ../generate_makefile.bash $OPTIONS && make test -j8'
102+
}
103+
post {
104+
always {
105+
sh 'ccache --show-stats'
106+
}
107+
}
108+
}
109+
}
110+
}
111+
}
112+
}

.travis.yml

+1-1
Original file line numberDiff line numberDiff line change
@@ -50,7 +50,7 @@ script:
5050
fi
5151
- mkdir build
5252
- cd build
53-
- ../generate_makefile.bash --compiler=$CXX --with-$THREADING --with-options=compiler_warnings ${GENERATE_OPTS}
53+
- ../generate_makefile.bash --compiler=$CXX --with-$THREADING --with-options=compiler_warnings --cxxflags="-O3 -Werror" ${GENERATE_OPTS}
5454
- make
5555
- make test
5656

CHANGELOG.md

+34
Original file line numberDiff line numberDiff line change
@@ -1,5 +1,39 @@
11
# Change Log
22

3+
## [2.9.00](https://github.com/kokkos/kokkos/tree/2.9.00) (2019-06-24)
4+
[Full Changelog](https://github.com/kokkos/kokkos/compare/2.8.00...2.9.00)
5+
6+
**Implemented enhancements:**
7+
8+
- Capability: CUDA Streams [\#1723](https://github.com/kokkos/kokkos/issues/1723)
9+
- Capability: CUDA Stream support for parallel\_reduce [\#2061](https://github.com/kokkos/kokkos/issues/2061)
10+
- Capability: Feature Request: TeamVectorRange [\#713](https://github.com/kokkos/kokkos/issues/713)
11+
- Capability: Adding HPX backend [\#2080](https://github.com/kokkos/kokkos/issues/2080)
12+
- Capability: TaskScheduler to have multiple queues [\#565](https://github.com/kokkos/kokkos/issues/565)
13+
- Capability: Support for additional reductions in ScatterView [\#1674](https://github.com/kokkos/kokkos/issues/1674)
14+
- Capability: Request: deep\_copy within parallel regions [\#689](https://github.com/kokkos/kokkos/issues/689)
15+
- Capability: Feature Request: `create\_mirror\_view\_without\_initializing` [\#1765](https://github.com/kokkos/kokkos/issues/1765)
16+
- View: Use SFINAE to restrict possible View type conversions [\#2127](https://github.com/kokkos/kokkos/issues/2127)
17+
- Deprecation: Deprecate ExecutionSpace::fence\(\) as static function and make it non-static [\#2140](https://github.com/kokkos/kokkos/issues/2140)
18+
- Deprecation: Deprecate LayoutTileLeft [\#2122](https://github.com/kokkos/kokkos/issues/2122)
19+
- Macros: KOKKOS\_RESTRICT defined for non-Intel compilers [\#2038](https://github.com/kokkos/kokkos/issues/2038)
20+
21+
**Fixed bugs:**
22+
23+
- Cuda: TeamThreadRange loop count on device is passed by reference to host static constexpr [\#1733](https://github.com/kokkos/kokkos/issues/1733)
24+
- Cuda: Build error with relocatable device code with CUDA 10.1 GCC 7.3 [\#2134](https://github.com/kokkos/kokkos/issues/2134)
25+
- Cuda: cudaFuncSetCacheConfig is setting CachePreferShared too often [\#2066](https://github.com/kokkos/kokkos/issues/2066)
26+
- Cuda: TeamPolicy doesn't throw then created with non-viable vector length and also doesn't backscale to viable one [\#2020](https://github.com/kokkos/kokkos/issues/2020)
27+
- Cuda: cudaMemcpy error for large league sizes on V100 [\#1991](https://github.com/kokkos/kokkos/issues/1991)
28+
- Cuda: illegal warp sync in parallel\_reduce by functor on Turing 75 [\#1958](https://github.com/kokkos/kokkos/issues/1958)
29+
- TeamThreadRange: Inconsistent results from TeamThreadRange reduction [\#1905](https://github.com/kokkos/kokkos/issues/1905)
30+
- Atomics: atomic\_fetch\_oper & atomic\_oper\_fetch don't build for complex\<float\> [\#1964](https://github.com/kokkos/kokkos/issues/1964)
31+
- Views: Kokkos randomread Views leak memory [\#2155](https://github.com/kokkos/kokkos/issues/2155)
32+
- ScatterView: LayoutLeft overload currently non-functional [\#2165](https://github.com/kokkos/kokkos/issues/2165)
33+
- KNL: With intel 17.2.174 illegal instruction in random number test [\#2078](https://github.com/kokkos/kokkos/issues/2078)
34+
- Bitset: Enable copy constructor on device [\#2094](https://github.com/kokkos/kokkos/issues/2094)
35+
- Examples: do not compile due to template deduction error \(multi\_fem\) [\#1928](https://github.com/kokkos/kokkos/issues/1928)
36+
337
## [2.8.00](https://github.com/kokkos/kokkos/tree/2.8.00) (2019-02-05)
438
[Full Changelog](https://github.com/kokkos/kokkos/compare/2.7.24...2.8.00)
539

0 commit comments

Comments
 (0)