Address some code review comments

Author: jovial

ansible/.gitignore

@@ -90,3 +90,6 @@ roles/*
 !roles/gateway/**
 !roles/alertmanager/
 !roles/alertmanager/**
+!roles/slurm_recompile/**
+!roles/slurm_recompile/**
+

ansible/fatimage.yml

@@ -254,25 +254,11 @@
   hosts: cuda
   become: yes
   tasks:
-    - name: Get facts about CUDA installation
-      import_role: cuda
-      tasks_from: facts.yml
-
-    - name: Recompile and install slurm packages
-      shell: |
-        #!/bin/bash
-        set -eux
-        dnf download -y --source slurm-slurmd-ohpc
-        rpm -i slurm-ohpc-*.src.rpm
-        dnf install -y  @'Development Tools'
-        cd /root/rpmbuild/SPECS
-        dnf builddep -y slurm.spec
-        rpmbuild -bb -D "_with_nvml --with-nvml=/usr/local/cuda-{{ cuda_facts_version_short }}/targets/x86_64-linux/" slurm.spec
-        dnf reinstall -y /root/rpmbuild/RPMS/x86_64/*.rpm
-        # Workaround path issue: https://groups.google.com/g/slurm-users/c/cvGb4JnK8BY
-        if [[ -e /lib64/libnvidia-ml.so ]]; then
-          ln -s /lib64/libnvidia-ml.so.1 /lib64/libnvidia-ml.so
-        fi
+    - name: Recompile slurm
+      import_role:
+        name: recompile_slurm
+      vars:
+        recompile_slurm_nvml: "{{ groups.cuda | length > 0 }}"
 
 - name: Run post.yml hook
   vars:

ansible/roles/cuda/tasks/facts.yml

@@ -0,0 +1,5 @@
+---
+
+- name: Set cuda_facts_version_short
+  set_fact:
+    cuda_facts_version_short: "{{ cuda_version_short }}" 

ansible/roles/slurm_recompile/defaults/main.yml

@@ -0,0 +1,2 @@
+---
+slurm_recompile_nvml: false

ansible/roles/slurm_recompile/tasks/main.yml

@@ -0,0 +1,30 @@
+---
+- name: Get facts about CUDA installation
+  import_role: cuda
+  tasks_from: facts.yml
+
+- name: Recompile and install slurm packages
+  shell: |
+    #!/bin/bash
+    set -eux
+    dnf download -y --source slurm-slurmd-ohpc
+    rpm -i slurm-ohpc-*.src.rpm
+    cd /root/rpmbuild/SPECS
+    dnf builddep -y slurm.spec
+    rpmbuild -bb{% if slurm_recompile_nvml | bool %} -D "_with_nvml --with-nvml=/usr/local/cuda-{{ cuda_facts_version_short }}/targets/x86_64-linux/"{% endif %} slurm.spec
+    dnf reinstall -y /root/rpmbuild/RPMS/x86_64/*.rpm
+  become: true
+
+- name: Workaround missing symlink
+  # Workaround path issue: https://groups.google.com/g/slurm-users/c/cvGb4JnK8BY
+  command: ln -s /lib64/libnvidia-ml.so.1 /lib64/libnvidia-ml.so
+  args:
+    creates: /lib64/libnvidia-ml.so
+
+- name: Cleanup Dependencies
+  shell: |
+    #!/bin/bash
+    set -eux
+    set -o pipefail
+    dnf history list | grep Install | grep 'builddep -y slurm.spec' | head -n 1 |  awk '{print $1}' | xargs dnf history -y undo
+  become: true