ci: isolate GPU runners, respect CUDA_VISIBLE_DEVICES, drop global prunes

.github/workflows/docker-bases.yml

@@ -7,10 +7,10 @@ concurrency:
 on:
   push:
     paths:
-      - '/docker/Dockerfile.nvidia'
-      - '/docker/Dockerfile.cpu'
-      - '/docker/Dockerfile.amd'
-      - '/docker/Dockerfile.intel'
+      - "/docker/Dockerfile.nvidia"
+      - "/docker/Dockerfile.cpu"
+      - "/docker/Dockerfile.amd"
+      - "/docker/Dockerfile.intel"
   workflow_dispatch:
     inputs:
       cpu:
@@ -27,15 +27,15 @@ on:
         default: false
 
       tags:
-        description: 'Build compiler bases'
+        description: "Build compiler bases"
   schedule:
     # Run once a month
     - cron: "0 0 1 * *"
 
 jobs:
-#######################################################
-############## Basic gcc CPU ##########################
-#######################################################
+  #######################################################
+  ############## Basic gcc CPU ##########################
+  #######################################################
   deploy-cpu-bases:
     if: inputs.cpu
     name: "cpu-base"
@@ -66,22 +66,18 @@ jobs:
           username: ${{ secrets.DOCKER_USERNAME }}
           password: ${{ secrets.DOCKER_PASSWORD }}
 
-      - name: cleanup
-        run: docker system prune -a -f
-
       - name: GCC image
         uses: docker/build-push-action@v6
         with:
           context: .
-          file: './docker/Dockerfile.cpu'
+          file: "./docker/Dockerfile.cpu"
           push: true
-          build-args: 'gcc=${{ matrix.gcc }}'
-          tags: 'devitocodes/bases:cpu-gcc${{ matrix.gcc }}'
-
+          build-args: "gcc=${{ matrix.gcc }}"
+          tags: "devitocodes/bases:cpu-gcc${{ matrix.gcc }}"
 
-#######################################################
-############## Intel OneApi CPU #######################
-#######################################################
+  #######################################################
+  ############## Intel OneApi CPU #######################
+  #######################################################
   deploy-oneapi-bases:
     if: inputs.intel
     name: "oneapi-base"
@@ -107,43 +103,39 @@ jobs:
         with:
           username: ${{ secrets.DOCKER_USERNAME }}
           password: ${{ secrets.DOCKER_PASSWORD }}
-
-      - name: cleanup
-        run: docker system prune -a -f
-
       - name: ICX image
         uses: docker/build-push-action@v6
         with:
           context: .
-          file: './docker/Dockerfile.intel'
+          file: "./docker/Dockerfile.intel"
           push: true
-          target: 'icx'
-          build-args: 'arch=icx'
-          tags: 'devitocodes/bases:cpu-icx'
+          target: "icx"
+          build-args: "arch=icx"
+          tags: "devitocodes/bases:cpu-icx"
 
       - name: SYCL CPU image
         uses: docker/build-push-action@v6
         with:
           context: .
-          file: './docker/Dockerfile.intel'
+          file: "./docker/Dockerfile.intel"
           push: true
-          target: 'cpu-sycl'
-          build-args: 'arch=cpu-sycl'
-          tags: 'devitocodes/bases:cpu-sycl'
+          target: "cpu-sycl"
+          build-args: "arch=cpu-sycl"
+          tags: "devitocodes/bases:cpu-sycl"
 
       - name: SYCL GPU image
         uses: docker/build-push-action@v6
         with:
           context: .
-          file: './docker/Dockerfile.intel'
+          file: "./docker/Dockerfile.intel"
           push: true
-          target: 'gpu-sycl'
-          build-args: 'arch=gpu-sycl'
-          tags: 'devitocodes/bases:gpu-sycl'
+          target: "gpu-sycl"
+          build-args: "arch=gpu-sycl"
+          tags: "devitocodes/bases:gpu-sycl"
 
-#######################################################
-################### Nvidia nvhpc ######################
-#######################################################
+  #######################################################
+  ################### Nvidia nvhpc ######################
+  #######################################################
   deploy-nvidia-bases:
     if: inputs.nvidia
     name: "nvidia-bases"
@@ -170,42 +162,43 @@ jobs:
           username: ${{ secrets.DOCKER_USERNAME }}
           password: ${{ secrets.DOCKER_PASSWORD }}
 
-      - name: cleanup
-        run: docker system prune -a -f
-
       - name: NVC image
         uses: docker/build-push-action@v6
         with:
           context: .
-          file: './docker/Dockerfile.nvidia'
+          file: "./docker/Dockerfile.nvidia"
           push: true
-          target: 'nvc'
-          build-args: 'arch=nvc'
-          tags: 'devitocodes/bases:nvidia-nvc'
+          target: "nvc"
+          build-args: "arch=nvc"
+          # Label (not tag) with runner name for traceability without changing image tags
+          labels: builder-runner=${{ runner.name }}
+          tags: "devitocodes/bases:nvidia-nvc"
 
       - name: NVCC image
         uses: docker/build-push-action@v6
         with:
           context: .
-          file: './docker/Dockerfile.nvidia'
+          file: "./docker/Dockerfile.nvidia"
           push: true
-          target: 'nvcc'
-          build-args: 'arch=nvcc'
-          tags: 'devitocodes/bases:nvidia-nvcc'
+          target: "nvcc"
+          build-args: "arch=nvcc"
+          labels: builder-runner=${{ runner.name }}
+          tags: "devitocodes/bases:nvidia-nvcc"
 
       - name: NVC host image
         uses: docker/build-push-action@v6
         with:
           context: .
-          file: './docker/Dockerfile.nvidia'
+          file: "./docker/Dockerfile.nvidia"
           push: true
-          target: 'nvc-host'
-          build-args: 'arch=nvc-host'
-          tags: 'devitocodes/bases:cpu-nvc'
-
-#######################################################
-##################### AMD #############################
-#######################################################
+          target: "nvc-host"
+          build-args: "arch=nvc-host"
+          labels: builder-runner=${{ runner.name }}
+          tags: "devitocodes/bases:cpu-nvc"
+
+  #######################################################
+  ##################### AMD #############################
+  #######################################################
   deploy-amd-bases:
     if: inputs.amd
     name: "amd-base"
@@ -232,16 +225,13 @@ jobs:
           username: ${{ secrets.DOCKER_USERNAME }}
           password: ${{ secrets.DOCKER_PASSWORD }}
 
-      - name: cleanup
-        run: docker system prune -a -f
-
       - name: AMD image
         uses: docker/build-push-action@v6
         with:
           context: .
-          file: './docker/Dockerfile.amd'
+          file: "./docker/Dockerfile.amd"
           push: true
-          target: 'amdclang'
+          target: "amdclang"
           build-args: |
             ROCM_VERSION=5.5.1
             UCX_BRANCH=v1.13.1
@@ -252,9 +242,9 @@ jobs:
         uses: docker/build-push-action@v6
         with:
           context: .
-          file: './docker/Dockerfile.amd'
+          file: "./docker/Dockerfile.amd"
           push: true
-          target: 'hip'
+          target: "hip"
           build-args: |
             ROCM_VERSION=6.3.4
           tags: devitocodes/bases:amd-hip

.github/workflows/docker-devito.yml

@@ -13,14 +13,22 @@ jobs:
     env:
       # Use buildkit https://docs.docker.com/develop/develop-images/build_enhancements/ for better build
       DOCKER_BUILDKIT: "1"
+      # Unique container name to avoid clashes across concurrent self-hosted runners
+      CONTAINER_NAME: testrun-${{ matrix.tag }}-${{ runner.name }}
+
 
     strategy:
       fail-fast: false
       matrix:
         include:
           - base: 'bases:nvidia-nvc'
             tag: 'nvidia-nvc'
-            flag: '--init --gpus all'
+            # Respect CUDA_VISIBLE_DEVICES set by the runner and hard-limit docker to that device.
+            # (--env without value forwards host var; --gpus maps only that device)
+            flag: |
+              --init
+              --env CUDA_VISIBLE_DEVICES
+              --gpus "device=${CUDA_VISIBLE_DEVICES:-all}"
             test: 'tests/test_gpu_openacc.py tests/test_gpu_common.py'
             runner: ["self-hosted", "nvidiagpu"]
 
@@ -102,8 +110,11 @@ jobs:
           build-args: base=devitocodes/${{ matrix.base }}
 
       - name: Remove dangling layers
+        if: ${{ !contains(matrix.runner, 'nvidiagpu') }}
         run: docker system prune -f
 
       - name: Run tests
         run: |
-         docker run ${{ matrix.flag }} --rm -t --name testrun 'devitocodes/devito:${{ matrix.tag }}-dev' pytest ${{ matrix.test }}
+         docker run ${{ matrix.flag }} --rm -t --name "${CONTAINER_NAME}" \
+           devitocodes/devito:${{ matrix.tag }}-dev \
+           pytest ${{ matrix.test }}