Support for Ubuntu 22.04 and ROCm 5.3 (#48)

* Testing and CI support for Ubuntu 22.04

* Fixes for ROCm

- Jammy does not have ROCm installers

* Name, timeout, and python updates

- renamed ubuntu-jammy-external.yml to ubuntu-jammy.yml
- increased all 5 minute timeouts to 10 minutes
- include python 3.10 in testing

* Update dyninst to remove interposed definition of _r_debug

* Rebuild Dyninst + test install script

* Revert container change

* git safe directory

* pushd -> cd

* fix MPI include

* Fix testing step

* OMPI_ALLOW_RUN_AS_ROOT

* Test script changes

* Fix mismatched malloc / delete[]

* Jammy workflow tweaks

* CPack tweak for boost deb deps

* pthread_mutex_gotcha config returns when not enabled

* fix echoing config in CI

* USE_CLANG_OMP

- option to disable using LLVM OpenMP when building OpenMP test executables
- Jammy workflow sets USE_CLANG_OMP=OFF

* Dyninst submodule boost download

- updated containers workflow to include jammy
- updated workflow to use ci

* Updates to workflows + replace test-install.sh

- test-install.sh in this branch was replaced with one in main branch

* Expand jammy test-install.sh args

* Fix openmp-cg-sampling-duration test

* update timemory submodule

- use-after-free violation in popen::pclose

* revert some tweaks to sampling-duration test

* Fix env of test-install.sh

* cmake format

* jammy bash

* CPack install for jammy

* formatting workflow action version bump

* Update timemory submodule

- libunwind submodule via timemory sets SOVERSION to 99 to avoid ABI conflicts with v8

* Fix help menu for omnitrace-sample

* Support other boolean forms in test-install.sh

* Update docker files and build-docker.sh

- consolidated cases in build-docker.sh
- support rocm version of 0.0 (no rocm install)
- support rocm v5.3
- updated centos handling

* update opensuse actions/checkout version

* Tweaks to ubuntu-focal testing

- actions/checkout@v3
- use test-install script

* update cpack

- ubuntu 22.04
- rocm 5.3
- rename os matrix field to os-version
- remove CI_ROCM_VERSION (no longer necessary)
- remove default-rocm-version matrix field (no longer necessary)
- CentOS packaging

* fix argparsing and omnitrace-sample tests in install-tests.sh

* focal rocm test install workflow fix

* Fix omnitrace-sample build

* Dockerfile.centos + build-docker.sh updates

* Update actions/upload-artifact version

* Dockerfile.ubuntu: install rocm-device-libs

* Refactor cpack

* fix cpack if quotes

* Dockerfile.ubuntu rocm < 5 installs rocm-dev

* build-release.sh defaults to boost version 1.79.0
This commit is contained in:
Jonathan R. Madsen
2022-10-17 12:54:26 -05:00
committed by GitHub
parent a3439d5bf2
commit ede6007f9b
28 changed files with 543 additions and 313 deletions
+27 -10
View File
@@ -12,21 +12,38 @@ SHELL [ "/bin/bash", "-c" ]
ENV PATH /usr/local/bin:${PATH}
RUN if [ -n "$(cat /etc/os-release | grep 'VERSION="8"')" ]; then \
dnf -y --disablerepo '*' --enablerepo=extras swap centos-linux-repos centos-stream-repos && \
dnf -y distro-sync; \
fi
RUN yum update -y && \
yum groupinstall -y "Development Tools" && \
yum install -y centos-release-scl && \
yum install -y epel-release && \
yum install -y devtoolset-9 python3-pip openmpi3-devel zlib-devel numactl-devel papi-devel dpkg-devel dpkg-dev wget curl && \
yum groupinstall -y "Development Tools"
ARG TOOLSET_VERSION=9
RUN yum install -y epel-release && \
if [ -n "$(cat /etc/os-release | grep 'VERSION="8"')" ]; then \
yum install -y centos-release-stream && \
yum install -y gcc-toolset-${TOOLSET_VERSION} openmpi-devel; \
else \
yum install -y centos-release-scl && \
yum install -y devtoolset-${TOOLSET_VERSION} openmpi3-devel dpkg-dev; \
fi; \
yum install -y python3-pip zlib-devel numactl-devel papi-devel dpkg-devel wget curl && \
python3 -m pip install 'cmake==3.21.4'
ARG ROCM_VERSION=0.0
ARG AMDGPU_RPM=21.40.2/rhel/7.9/amdgpu-install-21.40.2.40502-1.el7.noarch.rpm
# ARG AMDGPU_RPM=latest/rhel/7.9/amdgpu-install-21.50.50000-1.el7.noarch.rpm
RUN yum install -y https://repo.radeon.com/amdgpu-install/${AMDGPU_RPM} && \
amdgpu-install --usecase=rocm,hip,hiplibsdk --no-dkms --skip-broken -y && \
yum install -y rocm-hip-sdk rocm-smi-lib roctracer-dev rocprofiler-dev rccl-dev && \
yum update -y && \
yum clean all
RUN if [ "${ROCM_VERSION}" != "0.0" ]; then \
yum install -y https://repo.radeon.com/amdgpu-install/${AMDGPU_RPM} && \
amdgpu-install --usecase=rocm,hip,hiplibsdk --no-dkms --skip-broken -y && \
yum install -y rocm-hip-sdk rocm-smi-lib roctracer-dev rocprofiler-dev rccl-dev && \
yum update -y && \
yum clean all; \
fi
ARG PYTHON_VERSIONS="6 7 8 9 10"
@@ -39,7 +56,7 @@ RUN wget https://repo.continuum.io/miniconda/Miniconda3-latest-Linux-x86_64.sh -
conda clean -a -y && \
conda init
RUN ln -s /opt/rocm-* /opt/rocm
RUN if [ "${ROCM_VERSION}" != "0.0" ]; then ln -sf /opt/rocm-${ROCM_VERSION}* /opt/rocm; fi
WORKDIR /home
SHELL [ "/bin/bash", "--login", "-c" ]
+10 -7
View File
@@ -18,15 +18,18 @@ RUN zypper update -y && \
zypper install -y python3-pip openmpi3-devel gcc-c++ git libnuma-devel dpkg-devel rpm-build wget curl binutils-gold && \
python3 -m pip install 'cmake==3.21.4'
# ARG AMDGPU_RPM=21.40.2/sle/15/amdgpu-install-21.40.2.40502-1.noarch.rpm
ARG ROCM_VERSION=0.0
ARG AMDGPU_RPM=latest/sle/15/amdgpu-install-21.50.50000-1.noarch.rpm
# ARG AMDGPU_RPM=21.40.2/sle/15/amdgpu-install-21.40.2.40502-1.noarch.rpm
RUN zypper --no-gpg-checks install -y https://repo.radeon.com/amdgpu-install/${AMDGPU_RPM} && \
zypper addrepo https://download.opensuse.org/repositories/devel:languages:perl/SLE_15/devel:languages:perl.repo && \
zypper --non-interactive --gpg-auto-import-keys refresh && \
amdgpu-install --usecase=rocm,hip,hiplibsdk --no-dkms -y && \
zypper install -y rocm-hip-sdk rocm-smi-lib roctracer-dev rocprofiler-dev rccl-devel && \
zypper clean --all
RUN if [ "${ROCM_VERSION}" != "0.0" ]; then \
zypper --no-gpg-checks install -y https://repo.radeon.com/amdgpu-install/${AMDGPU_RPM} && \
zypper addrepo https://download.opensuse.org/repositories/devel:languages:perl/SLE_15/devel:languages:perl.repo && \
zypper --non-interactive --gpg-auto-import-keys refresh && \
amdgpu-install --usecase=rocm,hip,hiplibsdk --no-dkms -y && \
zypper install -y rocm-hip-sdk rocm-smi-lib roctracer-dev rocprofiler-dev rccl-devel && \
zypper clean --all; \
fi
ARG PYTHON_VERSIONS="6 7 8 9 10"
+20 -8
View File
@@ -16,20 +16,32 @@ SHELL [ "/bin/bash", "-c" ]
ARG EXTRA_PACKAGES=""
ARG ROCM_REPO_VERSION="debian"
ARG ROCM_VERSION="0.0"
ARG ROCM_REPO_DIST="ubuntu"
ARG PYTHON_VERSIONS="6 7 8 9 10"
ENV PATH ${HOME}/.local/bin:${PATH}
RUN apt-get update && \
apt-get dist-upgrade -y && \
apt-get install -y build-essential cmake libnuma1 wget gnupg2 m4 bash-completion git-core autoconf libtool autotools-dev python3-pip lsb-release libpapi-dev libpfm4-dev libudev1 libopenmpi-dev rpm librpm-dev curl && \
python3 -m pip install 'cmake==3.21.4' && \
wget -q -O - https://repo.radeon.com/rocm/rocm.gpg.key | apt-key add - && \
echo "deb [arch=amd64] https://repo.radeon.com/rocm/apt/${ROCM_REPO_VERSION}/ ${ROCM_REPO_DIST} main" | tee /etc/apt/sources.list.d/rocm.list && \
apt-get update && \
apt-get dist-upgrade -y && \
apt-get install -y rocm-dev rocm-utils rocm-smi-lib roctracer-dev rocprofiler-dev rccl-dev hip-base hsa-amd-aqlprofile hsa-rocr-dev hsakmt-roct-dev ${EXTRA_PACKAGES} && \
apt-get autoclean
apt-get install -y build-essential cmake libnuma1 wget gnupg2 m4 bash-completion git-core autoconf libtool autotools-dev python3-pip lsb-release libpapi-dev libpfm4-dev libudev1 libopenmpi-dev rpm librpm-dev curl apt-utils && \
python3 -m pip install 'cmake==3.21.4'
RUN if [ "${ROCM_VERSION}" != "0.0" ]; then \
if [ -d /etc/apt/trusted.gpg.d ]; then \
wget -q -O - https://repo.radeon.com/rocm/rocm.gpg.key | gpg --dearmor > /etc/apt/trusted.gpg.d/rocm.gpg; \
else \
wget -q -O - https://repo.radeon.com/rocm/rocm.gpg.key | apt-key add -; \
fi && \
echo "deb [arch=amd64] https://repo.radeon.com/rocm/apt/${ROCM_REPO_VERSION}/ ${ROCM_REPO_DIST} main" | tee /etc/apt/sources.list.d/rocm.list && \
apt-get update && \
apt-get dist-upgrade -y && \
apt-get install -y hsa-amd-aqlprofile hsa-rocr-dev hsakmt-roct-dev && \
apt-get install -y hip-base hip-runtime-amd hip-dev && \
apt-get install -y rocm-llvm rocm-core rocm-smi-lib rocm-device-libs && \
apt-get install -y roctracer-dev rocprofiler-dev rccl-dev ${EXTRA_PACKAGES} && \
if [ "$(echo ${ROCM_VERSION} | awk -F '.' '{print $1}')" -lt "5" ]; then apt-get install -y rocm-dev; fi && \
apt-get autoclean; \
fi
RUN wget https://repo.continuum.io/miniconda/Miniconda3-latest-Linux-x86_64.sh -O miniconda.sh && \
bash miniconda.sh -b -p /opt/conda && \
+3 -2
View File
@@ -14,10 +14,11 @@ SHELL [ "/bin/bash", "-c" ]
ARG EXTRA_PACKAGES=""
ARG ELFUTILS_DOWNLOAD_VERSION="0.186"
ARG BOOST_DOWNLOAD_VERSION="1.79.0"
ARG NJOBS="12"
ARG PYTHON_VERSIONS="6 7 8 9 10"
ENV PATH ${HOME}/.local/bin:${PATH}
ENV PATH /usr/local/bin:${PATH}
ENV LIBRARY_PATH /usr/local/lib:/usr/local/lib64:${LIBRARY_PATH}
ENV LD_LIBRARY_PATH /usr/local/lib:/usr/local/lib64:${LD_LIBRARY_PATH}
ENV CMAKE_PREFIX_PATH /usr/local:${CMAKE_PREFIX_PATH}
@@ -31,7 +32,7 @@ RUN apt-get update && \
apt-get autoclean && \
locale -a && \
cd /tmp/dyninst && \
cmake -B build -DCMAKE_BUILD_TYPE=Release -DBUILD_BOOST=ON -DBUILD_TBB=ON -DBUILD_ELFUTILS=ON -DBUILD_LIBIBERTY=ON -DELFUTILS_DOWNLOAD_VERSION=${ELFUTILS_DOWNLOAD_VERSION} && \
cmake -B build -DCMAKE_BUILD_TYPE=Release -DBUILD_BOOST=ON -DBUILD_TBB=ON -DBUILD_ELFUTILS=ON -DBUILD_LIBIBERTY=ON -DELFUTILS_DOWNLOAD_VERSION=${ELFUTILS_DOWNLOAD_VERSION} -DBOOST_DOWNLOAD_VERSION=${BOOST_DOWNLOAD_VERSION} && \
cmake --build build --target all --parallel ${NJOBS} && \
cmake --build build --target install --parallel ${NJOBS} && \
wget https://repo.continuum.io/miniconda/Miniconda3-latest-Linux-x86_64.sh -O miniconda.sh && \
+9 -1
View File
@@ -7,6 +7,7 @@ set -e
: ${VERSIONS:=20.04}
: ${NJOBS=$(nproc)}
: ${ELFUTILS_VERSION:=0.186}
: ${BOOST_VERSION:=1.79.0}
: ${PUSH:=0}
verbose-run()
@@ -38,6 +39,7 @@ usage()
print_default_option versions "[VERSION] [VERSION...]" "Ubuntu or OpenSUSE release" "${VERSIONS}"
print_default_option "jobs -j" "[N]" "parallel build jobs" "${NJOBS}"
print_default_option elfutils-version "[0.183..0.186]" "ElfUtils version" "${ELFUTILS_VERSION}"
print_default_option boost-version "[1.67.0..1.79.0]" "Boost version" "${BOOST_VERSION}"
print_default_option user "[USERNAME]" "DockerHub username" "${USER}"
}
@@ -83,6 +85,11 @@ do
ELFUTILS_VERSION=${1}
reset-last
;;
"--boost-version")
shift
BOOST_VERSION=${1}
reset-last
;;
--user|-u)
shift
USER=${1}
@@ -131,7 +138,8 @@ do
--build-arg DISTRO=${DISTRO_IMAGE} \
--build-arg VERSION=${VERSION} \
--build-arg NJOBS=${NJOBS} \
--build-arg ELFUTILS_DOWNLOAD_VERSION=${ELFUTILS_VERSION}
--build-arg ELFUTILS_DOWNLOAD_VERSION=${ELFUTILS_VERSION} \
--build-arg BOOST_DOWNLOAD_VERSION=${BOOST_VERSION}
done
if [ "${PUSH}" -gt 0 ]; then
+58 -29
View File
@@ -111,12 +111,12 @@ if [ ! -f ${DOCKER_FILE} ]; then send-error "File \"${DOCKER_FILE}\" not found";
for VERSION in ${VERSIONS}
do
for i in ${ROCM_VERSIONS}
for ROCM_VERSION in ${ROCM_VERSIONS}
do
CONTAINER=${USER}/omnitrace:release-base-${DISTRO}-${VERSION}-rocm-${i}
ROCM_MAJOR=$(echo ${i} | sed 's/\./ /g' | awk '{print $1}')
ROCM_MINOR=$(echo ${i} | sed 's/\./ /g' | awk '{print $2}')
ROCM_PATCH=$(echo ${i} | sed 's/\./ /g' | awk '{print $3}')
CONTAINER=${USER}/omnitrace:release-base-${DISTRO}-${VERSION}-rocm-${ROCM_VERSION}
ROCM_MAJOR=$(echo ${ROCM_VERSION} | sed 's/\./ /g' | awk '{print $1}')
ROCM_MINOR=$(echo ${ROCM_VERSION} | sed 's/\./ /g' | awk '{print $2}')
ROCM_PATCH=$(echo ${ROCM_VERSION} | sed 's/\./ /g' | awk '{print $3}')
if [ -n "${ROCM_PATCH}" ]; then
ROCM_VERSN=$(( (${ROCM_MAJOR}*10000)+(${ROCM_MINOR}*100)+(${ROCM_PATCH}) ))
ROCM_SEP="."
@@ -126,46 +126,73 @@ do
fi
if [ "${DISTRO}" = "ubuntu" ]; then
ROCM_REPO_DIST="ubuntu"
ROCM_REPO_VERSION=${i}
case "${i}" in
ROCM_REPO_VERSION=${ROCM_VERSION}
case "${ROCM_VERSION}" in
4.1* | 4.0*)
ROCM_REPO_DIST="xenial"
;;
5.3*)
case "${VERSION}" in
22.04)
ROCM_REPO_DIST="jammy"
;;
20.04)
ROCM_REPO_DIST="focal"
;;
18.04)
ROCM_REPO_DIST="bionic"
;;
*)
;;
esac
;;
*)
;;
esac
verbose-run docker build . -f ${DOCKER_FILE} --tag ${CONTAINER} --build-arg DISTRO=${DISTRO} --build-arg VERSION=${VERSION} --build-arg ROCM_REPO_VERSION=${ROCM_REPO_VERSION} --build-arg ROCM_REPO_DIST=${ROCM_REPO_DIST} --build-arg PYTHON_VERSIONS=\"${PYTHON_VERSIONS}\"
verbose-run docker build . -f ${DOCKER_FILE} --tag ${CONTAINER} --build-arg DISTRO=${DISTRO} --build-arg VERSION=${VERSION} --build-arg ROCM_VERSION=${ROCM_VERSION} --build-arg ROCM_REPO_VERSION=${ROCM_REPO_VERSION} --build-arg ROCM_REPO_DIST=${ROCM_REPO_DIST} --build-arg PYTHON_VERSIONS=\"${PYTHON_VERSIONS}\"
elif [ "${DISTRO}" = "centos" ]; then
case "${VERSION}" in
7)
RPM_PATH=7.9
RPM_TAG=".el7"
TOOLSET_VERSION=9
;;
8)
RPM_PATH=8.5
RPM_TAG=".el7"
RPM_PATH=8.4
RPM_TAG=".el8"
TOOLSET_VERSION=11
;;
9)
RPM_PATH=9.0
RPM_TAG=".el9"
TOOLSET_VERSION=11
;;
*)
send-error "Invalid centos version ${VERSION}. Supported: 7, 8"
send-error "Invalid centos version ${VERSION}. Supported: 7, 8, 9"
esac
case "${i}" in
5.0*)
ROCM_RPM=latest/rhel/${RPM_PATH}/amdgpu-install-21.50.50000-1${RPM_TAG}.noarch.rpm
case "${ROCM_VERSION}" in
5.3 | 5.3.*)
ROCM_RPM=${ROCM_VERSION}/rhel/${RPM_PATH}/amdgpu-install-${ROCM_MAJOR}.${ROCM_MINOR}.${ROCM_VERSN}-1${RPM_TAG}.noarch.rpm
;;
4.5 | 4.5.2)
ROCM_RPM=21.40.2/rhel/${RPM_PATH}/amdgpu-install-21.40.2.40502-1${RPM_TAG}.noarch.rpm
5.2 | 5.2.*)
ROCM_RPM=22.20${ROCM_SEP}${ROCM_PATCH}/rhel/${RPM_PATH}/amdgpu-install-22.20.${ROCM_VERSN}-1${RPM_TAG}.noarch.rpm
;;
4.5.1)
ROCM_RPM=21.40.1/rhel/${RPM_PATH}/amdgpu-install-21.40.1.40501-1${RPM_TAG}.noarch.rpm
5.1 | 5.1.*)
ROCM_RPM=22.10${ROCM_SEP}${ROCM_PATCH}/rhel/${RPM_PATH}/amdgpu-install-22.10${ROCM_SEP}${ROCM_PATCH}.${ROCM_VERSN}-1${RPM_TAG}.noarch.rpm
;;
4.5.0)
ROCM_RPM=21.40/rhel/${RPM_PATH}/amdgpu-install-21.40.1.40501-1${RPM_TAG}.noarch.rpm
5.0 | 5.0.*)
ROCM_RPM=21.50${ROCM_SEP}${ROCM_PATCH}/rhel/${RPM_PATH}/amdgpu-install-21.50${ROCM_SEP}${ROCM_PATCH}.${ROCM_VERSN}-1${RPM_TAG}.noarch.rpm
;;
4.5 | 4.5.*)
ROCM_RPM=21.40${ROCM_SEP}${ROCM_PATCH}/rhel/${RPM_PATH}/amdgpu-install-21.40${ROCM_SEP}${ROCM_PATCH}.${ROCM_VERSN}-1.noarch.rpm
;;
0.0)
;;
*)
send-error "Unsupported combination :: ${DISTRO}-${VERSION} + ROCm ${i}"
send-error "Unsupported combination :: ${DISTRO}-${VERSION} + ROCm ${ROCM_VERSION}"
;;
esac
verbose-run docker build . -f ${DOCKER_FILE} --tag ${CONTAINER} --build-arg DISTRO=${DISTRO} --build-arg VERSION=${VERSION} --build-arg AMDGPU_RPM=${ROCM_RPM} --build-arg PYTHON_VERSIONS=\"${PYTHON_VERSIONS}\"
verbose-run docker build . -f ${DOCKER_FILE} --tag ${CONTAINER} --build-arg DISTRO=${DISTRO} --build-arg VERSION=${VERSION} --build-arg ROCM_VERSION=${ROCM_VERSION} --build-arg TOOLSET_VERSION=${TOOLSET_VERSION} --build-arg AMDGPU_RPM=${ROCM_RPM} --build-arg PYTHON_VERSIONS=\"${PYTHON_VERSIONS}\"
elif [ "${DISTRO}" = "opensuse" ]; then
case "${VERSION}" in
15.*)
@@ -176,14 +203,14 @@ do
send-error "Invalid opensuse version ${VERSION}. Supported: 15.x"
;;
esac
case "${i}" in
case "${ROCM_VERSION}" in
5.3 | 5.3.*)
ROCM_RPM=${ROCM_VERSION}/sle/${VERSION}/amdgpu-install-${ROCM_MAJOR}.${ROCM_MINOR}.${ROCM_VERSN}-1.noarch.rpm
;;
5.2 | 5.2.*)
ROCM_RPM=22.20${ROCM_SEP}${ROCM_PATCH}/sle/${VERSION}/amdgpu-install-22.20.${ROCM_VERSN}-1.noarch.rpm
;;
5.1 | 5.1.0)
ROCM_RPM=22.10/sle/15/amdgpu-install-22.10.50100-1.noarch.rpm
;;
5.1.*)
5.1 | 5.1.*)
ROCM_RPM=22.10${ROCM_SEP}${ROCM_PATCH}/sle/15/amdgpu-install-22.10${ROCM_SEP}${ROCM_PATCH}.${ROCM_VERSN}-1.noarch.rpm
;;
5.0 | 5.0.*)
@@ -192,11 +219,13 @@ do
4.5 | 4.5.*)
ROCM_RPM=21.40${ROCM_SEP}${ROCM_PATCH}/sle/15/amdgpu-install-21.40${ROCM_SEP}${ROCM_PATCH}.${ROCM_VERSN}-1.noarch.rpm
;;
0.0)
;;
*)
send-error "Unsupported combination :: ${DISTRO}-${VERSION} + ROCm ${i}"
send-error "Unsupported combination :: ${DISTRO}-${VERSION} + ROCm ${ROCM_VERSION}"
;;
esac
verbose-run docker build . -f ${DOCKER_FILE} --tag ${CONTAINER} --build-arg DISTRO=${DISTRO_IMAGE} --build-arg VERSION=${VERSION} --build-arg AMDGPU_RPM=${ROCM_RPM} --build-arg PYTHON_VERSIONS=\"${PYTHON_VERSIONS}\"
verbose-run docker build . -f ${DOCKER_FILE} --tag ${CONTAINER} --build-arg DISTRO=${DISTRO_IMAGE} --build-arg VERSION=${VERSION} --build-arg ROCM_VERSION=${ROCM_VERSION} --build-arg AMDGPU_RPM=${ROCM_RPM} --build-arg PYTHON_VERSIONS=\"${PYTHON_VERSIONS}\"
fi
if [ "${PUSH}" -ne 0 ]; then
docker push ${CONTAINER}