-
Notifications
You must be signed in to change notification settings - Fork 3.9k
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.
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
[CUDA] Support CUDA 9 and test different CUDA versions at CI #3880
Changes from all commits
58564c1
0fcae1d
60419e1
071caeb
67a9a37
7aa5f44
1cdb8ef
09d3c5e
304d5ef
d8cef88
9a85089
c605e87
b599761
4ff8b1a
645766f
13952b3
671e91f
98259d4
32cb83c
4a32c30
3b0d4f9
File filter
Filter by extension
Conversations
Jump to
Diff view
Diff view
There are no files selected for viewing
Original file line number | Diff line number | Diff line change |
---|---|---|
|
@@ -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: | ||
|
@@ -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" | ||
There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. GitHub Actions treats values written as There was a problem hiding this comment. Choose a reason for hiding this commentThe 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: | | ||
|
@@ -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 - | ||
|
@@ -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 $ROOT_DOCKER_FOLDER/docker-script.sh | ||
all-successful: | ||
# https://github.saobby.my.eu.orgmunity/t/is-it-possible-to-require-all-github-actions-tasks-to-pass-without-enumerating-them/117957/4?u=graingert | ||
runs-on: ubuntu-latest | ||
|
Original file line number | Diff line number | Diff line change |
---|---|---|
|
@@ -156,7 +156,14 @@ if(USE_CUDA) | |
find_package(CUDA REQUIRED) | ||
include_directories(${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
There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. cc @ChipKerchner There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. I don't think so There was a problem hiding this comment. Choose a reason for hiding this commentThe 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 And if (1), how this change can be harmful to POWER machines? There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. I think this change is OK for POWER systems. There was a problem hiding this comment. Choose a reason for hiding this commentThe 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) | ||
|
There was a problem hiding this comment.
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
There was a problem hiding this comment.
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