Merge pull request #1055 from NVIDIA/gfd-mount-etc-nvidia-imex

Make the IMEX nodes config file available to GFD

Author: cdesiniotis

assets/gpu-feature-discovery/0500_configmap.yaml

@@ -0,0 +1,22 @@
+apiVersion: v1
+kind: ConfigMap
+metadata:
+  name: gpu-feature-discovery-entrypoint
+  namespace: "FILLED BY THE OPERATOR"
+  labels:
+    app: gpu-feature-discovery
+data:
+  entrypoint.sh: |-
+    #!/bin/bash
+
+    until [[ -f /run/nvidia/validations/driver-ready ]]
+    do
+      echo "waiting for the driver validations to be ready..."
+      sleep 5
+    done
+
+    set -o allexport
+    cat /run/nvidia/validations/driver-ready
+    . /run/nvidia/validations/driver-ready
+
+    exec gpu-feature-discovery

assets/gpu-feature-discovery/0500_daemonset.yaml renamed to assets/gpu-feature-discovery/0600_daemonset.yaml

@@ -65,7 +65,9 @@ spec:
       containers:
         - image: "FILLED BY THE OPERATOR"
           name: gpu-feature-discovery
-          command: ["gpu-feature-discovery"]
+          command: [ "/bin/bash", "-c" ]
+          args:
+            - /bin/entrypoint.sh
           env:
             - name: GFD_SLEEP_INTERVAL
               value: 60s
@@ -80,11 +82,24 @@ spec:
                 fieldRef:
                   fieldPath: spec.nodeName
           volumeMounts:
+            - name: gpu-feature-discovery-entrypoint
+              readOnly: true
+              mountPath: /bin/entrypoint.sh
+              subPath: entrypoint.sh
             - name: output-dir
               mountPath: "/etc/kubernetes/node-feature-discovery/features.d"
             - name: dmi-info-dir
               mountPath: "/sys/class/dmi/id"
               readOnly: true
+            - name: run-nvidia-validations
+              mountPath: /run/nvidia/validations
+            - name: driver-install-dir
+              mountPath: /driver-root
+              mountPropagation: HostToContainer
+            - name: host-root
+              mountPath: /host
+              readOnly: true
+              mountPropagation: HostToContainer
           securityContext:
             privileged: true
         - image: "FILLED BY THE OPERATOR"
@@ -116,6 +131,10 @@ spec:
           - name: PROCESS_TO_SIGNAL
             value: "gpu-feature-discovery"
       volumes:
+        - name: gpu-feature-discovery-entrypoint
+          configMap:
+            name: gpu-feature-discovery-entrypoint
+            defaultMode: 448
         - name: output-dir
           hostPath:
             path: "/etc/kubernetes/node-feature-discovery/features.d"
@@ -126,3 +145,14 @@ spec:
           hostPath:
             path: "/run/nvidia"
             type: Directory
+        - name: run-nvidia-validations
+          hostPath:
+            path: /run/nvidia/validations
+            type: DirectoryOrCreate
+        - name: host-root
+          hostPath:
+            path: /
+        - name: driver-install-dir
+          hostPath:
+            path: /run/nvidia/driver
+            type: DirectoryOrCreate