bump R550 and R535 to the latest TRD releases

Signed-off-by: Tariq Ibrahim <[email protected]>

Author: tariq1890

.common-ci.yml

@@ -19,7 +19,7 @@ default:
       command: ["--experimental"]
 
 variables:
-  DRIVER_VERSIONS: 535.183.06 550.90.12 560.35.03
+  DRIVER_VERSIONS: 535.216.01 550.127.05 560.35.03
   BUILD_MULTI_ARCH_IMAGES: "true"
 
 stages:
@@ -73,13 +73,13 @@ trigger-pipeline:
 .driver-versions:
   parallel:
     matrix:
-      - DRIVER_VERSION: [535.183.06, 550.90.12, 560.35.03]
+      - DRIVER_VERSION: [535.216.01, 550.127.05, 560.35.03]
 
 # Define the driver versions for jobs that can be run in parallel for ubuntu22.04
 .driver-versions-ubuntu22.04:
   parallel:
     matrix:
-      - DRIVER_VERSION: [535.183.06, 550.90.12, 560.35.03]
+      - DRIVER_VERSION: [535.216.01, 550.127.05, 560.35.03]
 
 # Define the matrix of precompiled jobs that can be run in parallel for ubuntu22.04
 .driver-versions-precompiled-ubuntu22.04:
@@ -92,7 +92,7 @@ trigger-pipeline:
 .driver-versions-rhel9:
   parallel:
     matrix:
-      - DRIVER_VERSION: [535.183.06, 550.90.12, 560.35.03]
+      - DRIVER_VERSION: [535.216.01, 550.127.05, 560.35.03]
 
 # Define the distribution targets
 .dist-ubuntu20.04:

.github/workflows/image.yaml

@@ -34,8 +34,8 @@ jobs:
     strategy:
       matrix:
         driver:
-          - 535.183.06
-          - 550.90.12
+          - 535.216.01
+          - 550.127.05
           - 560.35.03
         dist:
           - ubuntu20.04
@@ -49,10 +49,10 @@ jobs:
             dist: rhel8
           - ispr: true
             dist: ubuntu20.04
-            driver: 535.183.06
+            driver: 535.216.01
           - ispr: true
             dist: ubuntu20.04
-            driver: 550.90.12
+            driver: 550.127.05
       fail-fast: false
     steps:
       - uses: actions/checkout@v4

.nvidia-ci.yml

@@ -481,13 +481,13 @@ sign:ngc-ubuntu-rhel-rhcos:
     matrix:
     - SIGN_JOB_NAME: ["ubuntu"]
       VERSION: ["22.04"]
-      DRIVER_VERSION: ["535.183.06", "550.90.12", "560.35.03"]
+      DRIVER_VERSION: ["535.216.01", "550.127.05", "560.35.03"]
     - SIGN_JOB_NAME: ["ubuntu"]
       VERSION: ["20.04"]
-      DRIVER_VERSION: ["535.183.06", "550.90.12", "560.35.03"]
+      DRIVER_VERSION: ["535.216.01", "550.127.05", "560.35.03"]
     - SIGN_JOB_NAME: ["rhel"]
       VERSION: ["8.8", "8.10"]
-      DRIVER_VERSION: ["535.183.06", "550.90.12", "560.35.03"]
+      DRIVER_VERSION: ["535.216.01", "550.127.05", "560.35.03"]
     - SIGN_JOB_NAME: ["rhcos"]
       VERSION: ["4.12","4.13","4.14","4.15", "4.16", "4.17"]
-      DRIVER_VERSION: ["535.183.06", "550.90.12", "560.35.03"]
+      DRIVER_VERSION: ["535.216.01", "550.127.05", "560.35.03"]

rhel8/precompiled/Makefile

@@ -2,7 +2,7 @@ DOCKERFILE = Dockerfile
 RHEL_VERSION ?= 8.6
 CUDA_VERSION ?= 12.6.0
 CUDA_DIST ?= ubi8
-DRIVER_VERSION ?= 535.183.06
+DRIVER_VERSION ?= 535.216.01
 KERNEL_VERSION ?= 4.18.0-372.51.1.el8_6.x86_64
 TARGET_ARCH = $(shell echo "${KERNEL_VERSION}" | awk -F. '{print $$NF}')
 CONTAINER_TOOL ?= docker

rhel8/precompiled/README.md

@@ -50,7 +50,7 @@ The procedure is based on [building custom kmod packages](https://github.com/NVI
     export CUDA_VERSION=12.1.0
     export CUDA_DIST=ubi8
     export DRIVER_EPOCH=1
-    export DRIVER_VERSION=535.183.06
+    export DRIVER_VERSION=535.216.01
     export OS_TAG=rhcos4.13
 
     make image image-push
@@ -66,7 +66,7 @@ The procedure is based on [building custom kmod packages](https://github.com/NVI
 
 ## NVIDIA GPU operator
 
-In order to be used with the NVIDIA GPU Operator on Red Hat OpenShift, the image tag must follow the format `${DRIVER_VERSION}-${KERNEL_VERSION}-${OS_TAG}`, and the full name will look like `nvcr.io/nvidia/driver:535.183.06-4.18.0-513.9.1.el8_9.x86_64-rhcos4.13`.
+In order to be used with the NVIDIA GPU Operator on Red Hat OpenShift, the image tag must follow the format `${DRIVER_VERSION}-${KERNEL_VERSION}-${OS_TAG}`, and the full name will look like `nvcr.io/nvidia/driver:535.216.01-4.18.0-513.9.1.el8_9.x86_64-rhcos4.13`.
 
 
 Define the `ClusterPolicy` resource to make use of the pre-compiled driver image, e.g.:
@@ -76,7 +76,7 @@ Define the `ClusterPolicy` resource to make use of the pre-compiled driver image
     usePrecompiled: true
     image: driver
     repository: nvcr.io/nvidia
-    version: 535.183.06
+    version: 535.216.01
 ```
 
 Find more information in the [Precompiled Driver Containers](https://docs.nvidia.com/datacenter/cloud-native/gpu-operator/latest/precompiled-drivers.html) documentation.

rhel9/precompiled/README.md

@@ -105,7 +105,7 @@ The procedure is based on [building custom kmod packages](https://github.com/NVI
    ```
    export CUDA_VERSION=12.3.2
    export DRIVER_EPOCH=1
-   export DRIVER_VERSION=550.90.07
+   export DRIVER_VERSION=550.127.05
    ```
 
 6. [Optional] Build the vGPU guest driver
@@ -153,7 +153,7 @@ Define the `NVIDIADDriver` custom resource to make use of the pre-compiled drive
     usePrecompiled: true
     repository: quay.io/acme
     image: nvidia-gpu-driver
-    version: 550.90.07
+    version: 550.127.05
 ```
 
 Define the `ClusterPolicy` resource to make use of the NVIDIADriver custom resource, e.g.:

rhel9/precompiled/nvidiadriver.json

@@ -9,7 +9,7 @@
     "usePrecompiled": true,
     "repository": "quay.io/fabiendupont",
     "image": "nvidia-gpu-driver",
-    "version": "550.90.07",
+    "version": "550.127.05",
     "imagePullPolicy": "Always",
     "nodeSelector": {},
     "manager": {},

ubuntu22.04/precompiled/Dockerfile

@@ -4,7 +4,7 @@ ENV DEBIAN_FRONTEND=noninteractive
 
 ARG DRIVER_BRANCH=535
 ENV DRIVER_BRANCH=$DRIVER_BRANCH
-ARG DRIVER_VERSION=535.183.06
+ARG DRIVER_VERSION=535.216.01
 ENV DRIVER_VERSION=$DRIVER_VERSION
 
 ARG KERNEL_VERSION=5.15.0-116-generic

versions.mk

@@ -13,7 +13,7 @@
 # limitations under the License.
 
 # DRIVER_VERSIONS contains latest version in all active datacenter branches
-DRIVER_VERSIONS ?= 535.183.06 550.90.12 560.35.03
+DRIVER_VERSIONS ?= 535.216.01 550.127.05 560.35.03
 
 CUDA_VERSION := 12.6.0