Merge pull request #69 from itkovian/24.05.ug-before-reduce-patches

24.05.ug before reduce patches

.github/workflows/build.yml

@@ -0,0 +1,24 @@
+name: slurm C/C++ build
+
+on:
+  push:
+    branches: [20.11.ug, 22.05.ug, 24.05.ug]
+  pull_request:
+    branches: [20.11.ug, 22.05.ug, 24.05.ug]
+
+jobs:
+  build:
+    runs-on: ubuntu-20.04
+
+    steps:
+      - uses: actions/checkout@v2
+      - name: Install deps
+        run: sudo apt-get install -y libmunge-dev
+      - name: configure
+        run: ./configure --enable-multiple-slurmd --prefix=/tmp/slurm/
+      - name: make
+        run: make -j
+      - name: make check
+        run: make -j check
+      - name: make install
+        run: make -j install

META

@@ -7,8 +7,8 @@
   Name:		slurm
   Major:	24
   Minor:	05
-  Micro:	4
-  Version:	24.05.4
+  Micro:	5
+  Version:	24.05.5
   Release:	1
 
 ##

NEWS

@@ -1,6 +1,9 @@
 This file describes changes in recent versions of Slurm. It primarily
 documents those changes that are of interest to users and administrators.
 
+* Changes in Slurm 24.05.6
+==========================
+
 * Changes in Slurm 24.05.5
 ==========================
  -- Fix issue signaling cron jobs resulting in unintended requeues.
@@ -14,6 +17,25 @@ documents those changes that are of interest to users and administrators.
     removal of a dynamic node.
  -- gpu/nvml - Attempt loading libnvidia-ml.so.1 as a fallback for failure in
     loading libnvidia-ml.so.
+ -- slurmrestd - Fix populating non-required object fields of objects as '{}' in
+    JSON/YAML instead of 'null' causing compiled OpenAPI clients to reject
+    the response to 'GET /slurm/v0.0.40/jobs' due to validation failure of
+    '.jobs[].job_resources'.
+ -- Fix sstat/sattach protocol errors for steps on higher version slurmd's
+    (regressions since 20.11.0rc1 and 16.05.1rc1 respectively).
+ -- slurmd - Avoid a crash when starting slurmd version 24.05 with
+    SlurmdSpoolDir files that have been upgraded to a newer major version of
+    Slurm. Log warnings instead.
+ -- Fix race condition in stepmgr step completion handling.
+ -- Fix slurmctld segfault with stepmgr and MpiParams when running a job array.
+ -- Fix requeued jobs keeping their priority until the decay thread happens.
+ -- slurmctld - Fix crash and possible split brain issue if the
+    backup controller handles an scontrol reconfigure while in control
+    before the primary resumes operation.
+ -- Fix stepmgr not getting dynamic node addrs from the controller
+ -- stepmgr - avoid "Unexpected missing socket" errors.
+ -- Fix `scontrol show steps` with dynamic stepmgr
+ -- Support IPv6 in configless mode.
 
 * Changes in Slurm 24.05.4
 ==========================

bdist_rpm.sh

@@ -0,0 +1,7 @@
+#!/bin/bash
+
+set -e
+set -x
+
+
+./build.sh

build.sh

@@ -0,0 +1,150 @@
+#!/bin/bash
+
+set -e  # exit whenever command in pipeline fails
+set -x  # print commands as they are executed
+
+VERSION=`grep "Version:.*[0-9]" slurm.spec | tr -s " " |  awk '{print $2;}'`
+RELEASE=`grep "%define rel.*[-1-9]" slurm.spec | tr -s " " | awk '{print $3}'`
+
+if [ "${RELEASE}" != "1" ]; then
+    SUFFIX=${VERSION}-${RELEASE}
+else
+    SUFFIX=${VERSION}
+fi
+
+GITTAG=$(git log --format=%ct.%h -1)
+
+SCRIPT=$(readlink -f "${BASH_SOURCE[0]}")
+ORIGIN=$(dirname "$SCRIPT")
+
+# which version to download from github
+SLURM_VERSION=${VERSION:-24.05.3}
+UPSTREAM_REL=${UPSTREAM_REL:-1}
+
+# which release should be used for our RPMs
+OUR_RELEASE=${RELEASE:-1}
+
+# NVML
+# allow _empty_ version, which is used in pipeline
+
+if grep "release 8.8" /etc/redhat-release; then
+   NVIDIA_DRIVER=${NVIDIA_DRIVER-555.42.06}
+   NVDRV_NVML_PKG="nvidia-driver-NVML${NVIDIA_DRIVER:+-$NVIDIA_DRIVER}"
+   CUDA_VERSION=${CUDA_VERSION:-12.6}
+   CUDA_NVML_PKG="cuda-nvml-devel-${CUDA_VERSION//./-}"
+elif grep "release 9.4" /etc/redhat-release; then
+   NVIDIA_DRIVER=${NVIDIA_DRIVER-555.42.06}
+   NVDRV_NVML_PKG="nvidia-driver-NVML${NVIDIA_DRIVER:+-$NVIDIA_DRIVER}"
+   CUDA_VERSION=${CUDA_VERSION:-12.6}
+   CUDA_NVML_PKG="cuda-nvml-devel-${CUDA_VERSION//./-}"
+fi
+
+# Prepare directory structure
+rm -Rf $ORIGIN/rpmbuild/ $ORIGIN/dist/
+mkdir -p $ORIGIN/rpmbuild/{BUILD,RPMS,SRPMS,SOURCES} $ORIGIN/dist
+echo "Building source tarball"
+
+# archive git repo
+echo "archive suffix: ${SUFFIX}"
+git archive --format=tar.gz -o "rpmbuild/SOURCES/slurm-${SUFFIX}.tar.gz" --prefix="slurm-${SUFFIX}/" HEAD
+
+cp slurm.spec "SPECS"
+
+# Patch sources
+#for src_patch in $ORIGIN/src-patches/*.patch; do
+    #echo "Patching $src_patch"
+    #git am $src_patch
+#done
+
+# Patch spec file
+#for spec_patch in $ORIGIN/spec-patches/*.patch; do
+    #echo "Patching $spec_patch"
+    #patch -p1 -b -i $spec_patch
+#done
+
+# Install dependencies
+# see https://slurm.schedmd.com/quickstart_admin.html#build_install
+
+echo "Installing specfile requires"
+# this goes first because it might install undesired stuff related to `--with` options
+sudo dnf -y builddep slurm.spec
+
+# dependecy versions
+
+if grep "release 8.8" /etc/redhat-release; then
+    UCX_VERSION="1.13.1-2.el8.x86_64"
+    PMIX_VERSION=">= 4.2.6"
+    HWLOC_VERSION=">= 2.2.0-3"
+elif grep "release 9.2" /etc/redhat-release; then
+    UCX_VERSION="1.13.1-2.el9.x86_64"
+    PMIX_VERSION=">= 4.2.7"
+    HWLOC_VERSION=">= 2.4.1-5"
+elif grep "release 9.4" /etc/redhat-release; then
+    UCX_VERSION="1.15.0-2.el9.x86_64"
+    PMIX_VERSION=">= 4.2.7"
+    HWLOC_VERSION=">= 2.4.1-5"
+else
+    echo "unsupported OS release"
+    exit 1
+fi
+
+echo "Installing dependencies"
+# - features: basic
+sudo dnf -y install lua-devel mariadb-devel lz4-devel
+# - features: authentication (MUNGE: yes, JWT: yes, PAM: yes)
+sudo dnf -y install munge-devel libjwt-devel pam-devel
+# - features: slurmrestd
+sudo dnf -y install http-parser-devel json-c-devel libyaml-devel
+# - features: Nvidia NVML
+sudo dnf -y autoremove cuda-nvml-* nvidia-driver-NVML-* nvidia-driver* libnvidia-ml*
+sudo dnf -y install "$CUDA_NVML_PKG" "$NVDRV_NVML_PKG" "nvidia-driver-devel"
+# - plugins: MPI
+sudo dnf -y install pmix "pmix-devel ${PMIX_VERSION}"  "ucx-devel-${UCX_VERSION}"
+# - plugins: cgroup/v2
+# see https://slurm.schedmd.com/cgroup_v2.html
+sudo dnf -y install kernel-headers dbus-devel
+# - plugins: task/cgroup, task/affinity
+sudo dnf -y install "hwloc-devel ${HWLOC_VERSION}" numactl-devel
+# - plugins: acct_gather_profile/hdf5
+sudo dnf -y install hdf5-devel
+
+# Build defines
+RPM_DEFINES=( --define "gittag ${GITTAG}" --define "_topdir $ORIGIN/rpmbuild" )
+
+# Build options
+SLURM_BUILDOPTS=( --with slurmrestd --without debug )
+# basic features
+SLURM_BUILDOPTS+=( --with lua --with mysql --with x11 )
+# plugins
+SLURM_BUILDOPTS+=( --with numa --with hwloc --with pmix --with ucx )
+# authentication
+SLURM_BUILDOPTS+=( --with pam --with jwt )
+
+echo "Running rpmbuild (without nvml)"
+rpmbuild -ba "${RPM_DEFINES[@]}" "${SLURM_BUILDOPTS[@]}" --without nvml \
+    slurm.spec 2>&1 | tee rpmbuild-without-nvml.out
+
+
+echo "Doing rpm rebuild (without nvml)"
+for rpm in $ORIGIN/rpmbuild/RPMS/x86_64/slurm-*$SUFFIX*.rpm ; do
+    rpmrebuild --release=${OUR_RELEASE}.${GITTAG}$(rpm -E '%dist').nogpu.ug -d $ORIGIN/dist -p $rpm
+done
+
+
+echo "Running rpmbuild (with nvml)"
+RPM_DEFINES+=( --define "_cuda_version $CUDA_VERSION" )
+rpmbuild -ba "${RPM_DEFINES[@]}" "${SLURM_BUILDOPTS[@]}" --with nvml \
+    slurm.spec 2>&1 | tee rpmbuild-with-nvml.out
+
+echo "Doing rpm rebuild (with nvml)"
+for rpm in $ORIGIN/rpmbuild/RPMS/x86_64/slurm-*$SUFFIX*.rpm ; do
+    rpmrebuild --release=${OUR_RELEASE}.${GITTAG}$(rpm -E '%dist').ug -d $ORIGIN/dist -p $rpm
+done
+
+# strip out torque binaries/wrapper from slurm-torque
+rpmrebuild -d $ORIGIN/dist --change-spec-files="sed '/\(pbsnodes\|mpiexec\|bin\/q.\+\)/d'" -p $ORIGIN/dist/x86_64/slurm-torque-*-${OUR_RELEASE}.${GITTAG}$(rpm -E '%dist').nogpu.ug*.rpm
+rpmrebuild -d $ORIGIN/dist --change-spec-files="sed '/\(pbsnodes\|mpiexec\|bin\/q.\+\)/d'" -p $ORIGIN/dist/x86_64/slurm-torque-*-${OUR_RELEASE}.${GITTAG}$(rpm -E '%dist').ug.*.rpm
+
+# get the RPMs out of the subdirectories
+find rpmbuild -type f -name "*.rpm" -exec rm {} ";"
+find $ORIGIN/dist/ -type f -name '*.rpm' -print0 | xargs -0 -I{} mv {} rpmbuild/RPMS/x86_64/