Skip to content

[WIP][CI] Enable DCU C++20 toolchain baseline #79421

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

Open
gouzil wants to merge 12 commits into
base: develop
Choose a base branch
from
Open

[WIP][CI] Enable DCU C++20 toolchain baseline #79421

Show file tree
Hide file tree
Changes from 7 commits
Commits
Show all changes
12 commits
Select commit Hold shift + click to select a range
6959f42
[CI] Enable DCU C++20 toolchain baseline
gouzil Jul 5, 2026
4e922db
[CI] let DCU docker image follow Dockerfile md5
gouzil Jul 8, 2026
c0adfdb
[CI] detect lib64 path when installing GCC
gouzil Jul 9, 2026
e9adce2
[CI] check libstdc++ path before GCC install
gouzil Jul 9, 2026
35e05b9
[CI] install DCU requirements from build context
gouzil Jul 9, 2026
ab2a559
[CI] restore DCU requirements download
gouzil Jul 9, 2026
7ccb519
更新 DTK 下载链接至新的源
gouzil Jul 9, 2026
dc37cc1
[CI] keep GCC runtimes isolated from system libraries
gouzil Jul 10, 2026
d9e3864
[CI] force remove stale DCU containers
gouzil Jul 13, 2026
709a5ac
[CI] clean up DCU containers after failures
gouzil Jul 13, 2026
451581e
[CI] simplify DCU container cleanup
gouzil Jul 13, 2026
d563602
[CI] run DCU cleanup commands directly
gouzil Jul 13, 2026
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
3 changes: 0 additions & 3 deletions .github/workflows/docker.yml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -109,9 +109,6 @@ jobs:
for name in "${!docker_files[@]}"
do
md5_value=`md5sum tools/dockerfile/${docker_files[$name]} | awk '{print $1}'`

@risemeup1111 risemeup1111 Jul 9, 2026

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

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

P1 优先级：P1
处理要求：请针对该评论修复并提交新的 commit。

这里仍然只用 Dockerfile.develop.dtk 的内容 md5 作为 docker_dcu 镜像 tag，但当前最新提交只修改了 tools/dockerfile/build_scripts/install_gcc.sh，Dockerfile.develop.dtk 的 md5 仍是上一轮的 881e30f7ca65be3eb3c04ea34e44c030。由于该 Dockerfile 通过 COPY tools/dockerfile/build_scripts /build_scripts 引入脚本，现有逻辑会继续拉取同一个 docker_dcu 镜像 tag；如果上一轮镜像已存在，docker pull 成功后就不会重新 build，install_gcc.sh 的 Kylin/lib64 修复不会进入实际 DCU 镜像，也无法验证本 PR 的目标环境。

建议把 docker_dcu 实际 COPY 的脚本也纳入 cache key，或采用等价方式确保 build_scripts 变更会触发 DCU 镜像重建。一个最小修复形态如下：

Suggested change
md5_value=`md5sum tools/dockerfile/${docker_files[$name]} | awk '{print $1}'`
if [ "$name" == "docker_dcu" ]; then
md5_value=`md5sum tools/dockerfile/${docker_files[$name]} tools/dockerfile/build_scripts/install_gcc.sh | md5sum | awk '{print $1}'`
else
md5_value=`md5sum tools/dockerfile/${docker_files[$name]} | awk '{print $1}'`
fi

@risemeup1111 risemeup1111 Jul 9, 2026

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

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

P1 优先级：P1
处理要求：请针对该评论修复并提交新的 commit。

补充当前 head ab2a5596fd175cdcda2edfac405d727be77b19cd 的复查结果：Dockerfile.develop.dtk 现在又回到 md5 881e30f7ca65be3eb3c04ea34e44c030，而 install_gcc.sh 仍是修改后的 a6ec0a16bc1ebde45ea6f329c87638d1。因此 docker_dcu 仍只会生成旧的 Dockerfile tag，若该 tag 已存在就会复用旧镜像，install_gcc.sh 的最新修复仍无法保证进入实际 DCU 镜像。请把 install_gcc.sh/build_scripts 纳入 docker_dcu hash，或用等价方式确保脚本变更触发新的镜像 tag。

This comment was marked as outdated.

Sign in to view

@risemeup1111 risemeup1111 Jul 9, 2026

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

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

P1 优先级：P1
处理要求：请针对该评论修复并提交新的 commit。

删除 docker_dcu 固定 md5 后，这里的 md5sum tools/dockerfile/${docker_files[$name]} 也会在 inputs.is_merge == "true" 的路径执行。该路径上面只下载 ci_dockerfile.sh 和 Dockerfile.ubuntu20，而 ci_dockerfile.sh 只生成 CPU/SOT/build/coverage 相关 Dockerfile，不会生成 Dockerfile.develop.dtk，也没有 tools/dockerfile/build_scripts；默认 task=cpu 的 docker_files 仍包含 docker_dcu。因此 merge-mode 会先得到空 md5，再尝试 docker build -f tools/dockerfile/Dockerfile.develop.dtk .，最终因为 Dockerfile/构建上下文缺失而失败。

请让 merge-mode 的 docker_dcu 路径也有完整输入，或在该模式下显式复用已有 DCU tag。一个可实现的修复形态如下：

if [ "${{ inputs.is_merge }}" == "true" ]; then
  mkdir -p tools/dockerfile/build_scripts
  cd tools/dockerfile
  wget -q --tries=5 --no-proxy \
    ${dockerfile_script} \
    ${dockerfile_origin} \
    https://raw.githubusercontent.com/PaddlePaddle/Paddle/refs/heads/develop/tools/dockerfile/Dockerfile.develop.dtk \
    --no-check-certificate
  cd build_scripts
  wget -q --tries=5 --no-proxy \
    https://raw.githubusercontent.com/PaddlePaddle/Paddle/refs/heads/develop/tools/dockerfile/build_scripts/install_gcc.sh \
    --no-check-certificate
  cd ..
else
  ...
fi

if [ $name == "docker_dcu" ]; then
md5_value="76937a563116f6008c8ca4cb4f592759"
fi
if [ $name == "docker_npu" ]; then
md5_value="a3793bdeea5ae881a0c1eaf4d7c30c64"
fi
Expand Down
2 changes: 1 addition & 1 deletion cmake/hip.cmake
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -198,7 +198,7 @@ list(APPEND HIP_CXX_FLAGS -Wno-missing-braces)
list(APPEND HIP_CXX_FLAGS -Wno-sometimes-uninitialized)
list(APPEND HIP_CXX_FLAGS -Wno-deprecated-copy)
list(APPEND HIP_CXX_FLAGS -Wno-pessimizing-move)
list(APPEND HIP_CXX_FLAGS -std=c++17)
list(APPEND HIP_CXX_FLAGS -std=c++20)

This comment was marked as outdated.

Sign in to view

list(APPEND HIP_CXX_FLAGS --gpu-max-threads-per-block=1024)
Comment on lines 200 to 202

if(CMAKE_BUILD_TYPE MATCHES Debug)
Expand Down
14 changes: 7 additions & 7 deletions tools/dockerfile/Dockerfile.develop.dtk
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -10,12 +10,12 @@ RUN yum install -y bzip2-devel openssh-server elfutils-devel diffutils libtool i
openssl-devel libjpeg-turbo-devel libpng-devel ninja-build pciutils libzstd-devel zstd

COPY tools/dockerfile/build_scripts /build_scripts
RUN bash /build_scripts/install_gcc.sh gcc82
RUN ln -sf /usr/local/gcc-8.2/bin/gcc /usr/local/bin/gcc
RUN ln -sf /usr/local/gcc-8.2/bin/g++ /usr/local/bin/g++
RUN ln -sf /usr/local/gcc-8.2/bin/gcc /usr/bin/gcc
RUN ln -sf /usr/local/gcc-8.2/bin/g++ /usr/bin/g++
ENV PATH=/usr/local/gcc-8.2/bin:$PATH
RUN bash /build_scripts/install_gcc.sh gcc122
RUN ln -sf /usr/local/gcc-12.2/bin/gcc /usr/local/bin/gcc
RUN ln -sf /usr/local/gcc-12.2/bin/g++ /usr/local/bin/g++
RUN ln -sf /usr/local/gcc-12.2/bin/gcc /usr/bin/gcc
RUN ln -sf /usr/local/gcc-12.2/bin/g++ /usr/bin/g++
ENV PATH=/usr/local/gcc-12.2/bin:$PATH

# workdir
WORKDIR /opt
Expand Down Expand Up @@ -100,7 +100,7 @@ RUN yum clean all && \
rm -rf /var/lib/yum/history

# Install DTK
RUN wget -q --no-proxy https://cancon.hpccube.com:65024/file/1/DTK-24.04.1/CentOS7.6/DTK-24.04.1-CentOS7.6-x86_64.tar.gz --no-check-certificate && \
RUN wget -q --no-proxy https://download.sourcefind.cn:65024/file/1/DTK-24.04.1/CentOS7.6/DTK-24.04.1-CentOS7.6-x86_64.tar.gz --no-check-certificate && \
tar zxf DTK-24.04.1-CentOS7.6-x86_64.tar.gz && rm -rf DTK-24.04.1-CentOS7.6-x86_64.tar.gz
# Replace if you use other device type, e.g. Z100, Z100L, K100
RUN wget -q https://paddle-device.bj.bcebos.com/dcu/hyhal-K100AI.tar.gz && \
Expand Down
12 changes: 6 additions & 6 deletions tools/dockerfile/build_scripts/install_gcc.sh
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -19,14 +19,14 @@
# Stop at any error, show all commands
set -ex

if [ -f "/etc/redhat-release" ];then
lib_so_5=/usr/lib64/libgfortran.so.5
lib_so_6=/usr/lib64/libstdc++.so.6
lib_path=/usr/lib64
else
if [ -e "/usr/lib/x86_64-linux-gnu/libstdc++.so.6" ];then
lib_so_5=/usr/lib/x86_64-linux-gnu/libstdc++.so.5
lib_so_6=/usr/lib/x86_64-linux-gnu/libstdc++.so.6
lib_path=/usr/lib/x86_64-linux-gnu
else
lib_so_5=/usr/lib64/libgfortran.so.5
lib_so_6=/usr/lib64/libstdc++.so.6
lib_path=/usr/lib64
fi

if [ "$1" == "gcc82" ]; then
Expand All @@ -43,7 +43,7 @@ if [ "$1" == "gcc82" ]; then
../gcc-8.2.0/configure --prefix=/usr/local/gcc-8.2 --enable-threads=posix --disable-checking --disable-multilib && \
make -j8 && make install
cd .. && rm -rf temp_gcc82 gcc-8.2.0 gcc-8.2.0.tar.xz
if [ -f "/etc/redhat-release" ];then
if [ "${lib_path}" == "/usr/lib64" ];then
cp ${lib_so_6} ${lib_so_6}.bak && rm -f ${lib_so_6} &&
ln -s /usr/local/gcc-8.2/lib64/libgfortran.so.5 ${lib_so_5} && \
ln -s /usr/local/gcc-8.2/lib64/libstdc++.so.6 ${lib_so_6} && \
Expand Down
Loading