Update cuda versions and remove deprecated versions.

.github/workflows/license_check.yml

@@ -1,6 +1,13 @@
 name: license check
 
-on: [push, pull_request]
+on:
+  push:
+    branches:
+    - master
+  pull_request:
+    branches:
+    - master
+
 
 defaults:
   run:

.github/workflows/link_check.yml

@@ -1,6 +1,13 @@
 name: link check
 
-on: [push, pull_request]
+on:
+  push:
+    branches:
+    - master
+  pull_request:
+    branches:
+    - master
+
 
 defaults:
   run:

.github/workflows/os_x_mklbuild.yml

@@ -1,6 +1,12 @@
 name: mkl continuous build
 
-on: [push, pull_request]
+on:
+  push:
+    branches:
+    - master
+  pull_request:
+    branches:
+    - master
 
 jobs:
   macosx-x86_64:

.github/workflows/os_x_staticbuild.yml

@@ -1,6 +1,13 @@
 name: continuous build
 
-on: [push, pull_request]
+on:
+  push:
+    branches:
+    - master
+  pull_request:
+    branches:
+    - master
+
 
 jobs:
   macosx-x86_64:

cd/Jenkinsfile_cd_pipeline

@@ -36,7 +36,7 @@ pipeline {
 
   parameters {
     // Release parameters
-    string(defaultValue: "cpu,native,cu101,cu102,cu110,cu112", description: "Comma separated list of variants", name: "MXNET_VARIANTS")
+    string(defaultValue: "cpu,native,cu118", description: "Comma separated list of variants", name: "MXNET_VARIANTS")
     booleanParam(defaultValue: false, description: 'Whether this is a release build or not', name: "RELEASE_BUILD")
   }
 

cd/Jenkinsfile_release_job

@@ -43,7 +43,7 @@ pipeline {
     // any disruption caused by different COMMIT_ID values chaning the job parameter configuration on
     // Jenkins.
     string(defaultValue: "mxnet_lib", description: "Pipeline to build", name: "RELEASE_JOB_TYPE")
-    string(defaultValue: "cpu,native,cu101,cu102,cu110,cu112", description: "Comma separated list of variants", name: "MXNET_VARIANTS")
+    string(defaultValue: "cpu,native,cu118", description: "Comma separated list of variants", name: "MXNET_VARIANTS")
     booleanParam(defaultValue: false, description: 'Whether this is a release build or not', name: "RELEASE_BUILD")
     string(defaultValue: "nightly", description: "String used for naming docker images", name: "VERSION")
   }

cd/utils/mxnet_base_image.sh

@@ -21,17 +21,14 @@
 mxnet_variant=${1:?"Please specify the mxnet variant as the first parameter"}
 
 case ${mxnet_variant} in
-    cu101*)
-    echo "nvidia/cuda:10.1-cudnn7-runtime-ubuntu18.04"
-    ;;
-    cu102*)
-    echo "nvidia/cuda:10.2-cudnn7-runtime-ubuntu18.04"
-    ;;
     cu110*)
-    echo "nvidia/cuda:11.0-cudnn8-runtime-ubuntu18.04"
+    echo "nvidia/cuda:11.0.3-cudnn8-runtime-ubuntu18.04"
     ;;
     cu112*)
-    echo "nvidia/cuda:11.2.0-cudnn8-runtime-ubuntu18.04"
+    echo "nvidia/cuda:11.2.2-cudnn8-runtime-ubuntu18.04"
+    ;;
+    cu118*)
+    echo "nvidia/cuda:11.8.0-cudnn8-runtime-ubuntu18.04"
     ;;
     cpu)
     echo "ubuntu:18.04"

ci/docker/docker-compose.yml

@@ -41,26 +41,6 @@ services:
         BASE_IMAGE: centos:7
       cache_from:
         - ${DOCKER_CACHE_REGISTRY}/build.centos7_cpu:latest
-  centos7_gpu_cu101:
-    image: ${DOCKER_CACHE_REGISTRY}/build.centos7_gpu_cu101:latest
-    build:
-      context: .
-      dockerfile: Dockerfile.build.centos7
-      target: base
-      args:
-        BASE_IMAGE: nvidia/cuda:10.1-cudnn8-devel-centos7
-      cache_from:
-        - ${DOCKER_CACHE_REGISTRY}/build.centos7_gpu_cu101:latest
-  centos7_gpu_cu102:
-    image: ${DOCKER_CACHE_REGISTRY}/build.centos7_gpu_cu102:latest
-    build:
-      context: .
-      dockerfile: Dockerfile.build.centos7
-      target: base
-      args:
-        BASE_IMAGE: nvidia/cuda:10.2-cudnn8-devel-centos7
-      cache_from:
-        - ${DOCKER_CACHE_REGISTRY}/build.centos7_gpu_cu102:latest
   centos7_gpu_cu110:
     image: ${DOCKER_CACHE_REGISTRY}/build.centos7_gpu_cu110:latest
     build:
@@ -78,9 +58,19 @@ services:
       dockerfile: Dockerfile.build.centos7
       target: base
       args:
-        BASE_IMAGE: nvidia/cuda:11.2.0-cudnn8-devel-centos7
+        BASE_IMAGE: nvidia/cuda:11.2.2-cudnn8-devel-centos7
       cache_from:
         - ${DOCKER_CACHE_REGISTRY}/build.centos7_gpu_cu112:latest
+  centos7_gpu_cu118:
+    image: ${DOCKER_CACHE_REGISTRY}/build.centos7_gpu_cu118:latest
+    build:
+      context: .
+      dockerfile: Dockerfile.build.centos7
+      target: base
+      args:
+        BASE_IMAGE: nvidia/cuda:11.8.0-cudnn8-devel-centos7
+      cache_from:
+        - ${DOCKER_CACHE_REGISTRY}/build.centos7_gpu_cu118:latest
   ###################################################################################################
   # Dockerfile.build.ubuntu based images. On Ubuntu we test more recent
   # toolchain and dependency versions compared to CentOS7. We attempt to update
@@ -105,20 +95,20 @@ services:
       dockerfile: Dockerfile.build.ubuntu
       target: gpu
       args:
-        BASE_IMAGE: nvidia/cuda:11.4.0-cudnn8-devel-ubuntu20.04
+        BASE_IMAGE: nvidia/cuda:11.4.3-cudnn8-devel-ubuntu20.04
         TRT_VERSION: 8.2.4-1+cuda11.4
       cache_from:
         - ${DOCKER_CACHE_REGISTRY}/build.ubuntu_tensorrt_cu114:latest
-  ubuntu_gpu_cu111:
-    image: ${DOCKER_CACHE_REGISTRY}/build.ubuntu_gpu_cu111:latest
+  ubuntu_gpu_cu118:
+    image: ${DOCKER_CACHE_REGISTRY}/build.ubuntu_gpu_cu118:latest
     build:
       context: .
       dockerfile: Dockerfile.build.ubuntu
       target: gpu
       args:
-        BASE_IMAGE: nvidia/cuda:11.1.1-cudnn8-devel-ubuntu20.04
+        BASE_IMAGE: nvidia/cuda:11.8.0-cudnn8-devel-ubuntu20.04
       cache_from:
-        - ${DOCKER_CACHE_REGISTRY}/build.ubuntu_gpu_cu111:latest
+        - ${DOCKER_CACHE_REGISTRY}/build.ubuntu_gpu_cu118:latest
   ###################################################################################################
   # Dockerfile.build.android based images used for testing cross-compilation for plain ARM
   ###################################################################################################
@@ -204,19 +194,12 @@ services:
       context: .
       dockerfile: Dockerfile.publish.test.centos7
       args:
-        BASE_IMAGE: nvidia/cuda:9.2-cudnn7-devel-centos7
+        BASE_IMAGE: nvidia/cuda:11.8.0-cudnn8-devel-centos7
       cache_from:
         - ${DOCKER_CACHE_REGISTRY}/publish.test.centos7_gpu:latest
   ###################################################################################################
   # Miscellaneous containers
   ###################################################################################################
-  jetson:
-    image: ${DOCKER_CACHE_REGISTRY}/build.jetson:latest
-    build:
-      context: .
-      dockerfile: Dockerfile.build.jetson
-      cache_from:
-        - ${DOCKER_CACHE_REGISTRY}/build.jetson:latest
   ubuntu_cpu_jekyll:
     image: ${DOCKER_CACHE_REGISTRY}/build.ubuntu_cpu_jekyll:latest
     build:

ci/docker/runtime_functions.sh

@@ -111,7 +111,7 @@ gather_licenses() {
 
 # Compiles the dynamic mxnet library
 # Parameters:
-# $1 -> mxnet_variant: the mxnet variant to build, e.g. cpu, native, cu101, cu102, etc.
+# $1 -> mxnet_variant: the mxnet variant to build, e.g. cpu, native, cu112, cu118, etc.
 build_dynamic_libmxnet() {
     set -ex
 
@@ -746,8 +746,7 @@ sanity_cpp() {
 }
 
 sanity_python_prospector() {
-    set -e
-    set +x
+    set -ex
 
     # Run Prospector
     python3 -m prospector --profile prospector.yaml | tee prospector-output.txt
@@ -761,8 +760,7 @@ sanity_python_prospector() {
 }
 
 sanity_clang() {
-    set -e
-    set +x
+    set -ex
     # .github/workgflows/greetings.yml passes BASE_SHA, GITHUB_RUN_ID, GITHUB_BASE_REF for pull requests.
     BASE_SHA="${GITHUB_PR_BASE_SHA}"
     GITHUB_RUN_ID="${GITHUB_PR_RUN_ID}"
@@ -1504,10 +1502,10 @@ build_static_python_cpu() {
     popd
 }
 
-build_static_python_cu102() {
+build_static_python_cu118() {
     set -ex
     pushd .
-    export mxnet_variant=cu102
+    export mxnet_variant=cu118
     source /opt/rh/devtoolset-8/enable
     source /opt/rh/rh-python38/enable
     # Opt in to newer GCC C++ ABI. devtoolset defaults to ABI Version 2.

ci/jenkins/Jenkins_steps.groovy

@@ -151,7 +151,7 @@ def compile_unix_int64_gpu(lib_name) {
         ws('workspace/build-gpu-int64') {
           timeout(time: max_time, unit: 'MINUTES') {
             utils.init_git()
-            utils.docker_run('ubuntu_gpu_cu111', 'build_ubuntu_gpu_large_tensor', false)
+            utils.docker_run('ubuntu_gpu_cu118', 'build_ubuntu_gpu_large_tensor', false)
             utils.pack_lib(lib_name, mx_cmake_lib)
           }
         }
@@ -207,7 +207,7 @@ def compile_unix_onednn_gpu(lib_name) {
         ws('workspace/build-onednn-gpu') {
           timeout(time: max_time, unit: 'MINUTES') {
             utils.init_git()
-            utils.docker_run('ubuntu_gpu_cu111', 'build_ubuntu_gpu_onednn', false)
+            utils.docker_run('ubuntu_gpu_cu118', 'build_ubuntu_gpu_onednn', false)
             utils.pack_lib(lib_name, mx_onednn_lib)
           }
         }
@@ -221,7 +221,7 @@ def compile_unix_onednn_nocudnn_gpu(lib_name) {
          ws('workspace/build-onednn-gpu-nocudnn') {
            timeout(time: max_time, unit: 'MINUTES') {
              utils.init_git()
-             utils.docker_run('ubuntu_gpu_cu111', 'build_ubuntu_gpu_onednn_nocudnn', false)
+             utils.docker_run('ubuntu_gpu_cu118', 'build_ubuntu_gpu_onednn_nocudnn', false)
              utils.pack_lib(lib_name, mx_onednn_lib)
            }
          }
@@ -235,7 +235,7 @@ def compile_unix_full_gpu(lib_name) {
         ws('workspace/build-gpu') {
           timeout(time: max_time, unit: 'MINUTES') {
             utils.init_git()
-            utils.docker_run('ubuntu_gpu_cu111', 'build_ubuntu_gpu', false)
+            utils.docker_run('ubuntu_gpu_cu118', 'build_ubuntu_gpu', false)
             utils.pack_lib(lib_name, mx_lib_cpp_examples)
           }
         }
@@ -249,7 +249,7 @@ def compile_unix_full_gpu_debug(lib_name) {
         ws('workspace/build-gpu') {
           timeout(time: max_time, unit: 'MINUTES') {
             utils.init_git()
-            utils.docker_run('ubuntu_gpu_cu111', 'build_ubuntu_gpu_debug', false)
+            utils.docker_run('ubuntu_gpu_cu118', 'build_ubuntu_gpu_debug', false)
             utils.pack_lib(lib_name, mx_lib_cpp_examples)
           }
         }
@@ -304,7 +304,7 @@ def compile_centos7_gpu(lib_name) {
         ws('workspace/build-centos7-gpu') {
           timeout(time: max_time, unit: 'MINUTES') {
             utils.init_git()
-            utils.docker_run('centos7_gpu_cu102', 'build_centos7_gpu', false)
+            utils.docker_run('centos7_gpu_cu118', 'build_centos7_gpu', false)
             utils.pack_lib(lib_name, mx_lib)
           }
         }
@@ -508,7 +508,7 @@ def compile_unix_clang10_cuda_werror(lib_name) {
         ws('workspace/build-cpu-clang10') {
           timeout(time: max_time, unit: 'MINUTES') {
             utils.init_git()
-            utils.docker_run('ubuntu_gpu_cu111', 'build_ubuntu_gpu_clang10_werror', false)
+            utils.docker_run('ubuntu_gpu_cu118', 'build_ubuntu_gpu_clang10_werror', false)
             utils.pack_lib(lib_name, mx_lib)
           }
         }
@@ -633,7 +633,7 @@ def compile_static_python_gpu() {
         ws('workspace/ut-publish-python-gpu') {
           timeout(time: max_time, unit: 'MINUTES') {
             utils.init_git()
-            utils.docker_run('centos7_gpu_cu102', 'build_static_python_cu102')
+            utils.docker_run('centos7_gpu_cu118', 'build_static_python_cu118')
           }
         }
     }
@@ -646,7 +646,7 @@ def compile_static_cd_gpu(lib_name) {
         ws('workspace/build-cd-static/gpu') {
           timeout(time: max_time, unit: 'MINUTES') {
             utils.init_git()
-            utils.docker_run('centos7_gpu_cu102', 'build_static_libmxnet cu102', false)
+            utils.docker_run('centos7_gpu_cu118', 'build_static_libmxnet cu118', false)
             utils.pack_lib(lib_name, mx_cd_lib)
           }
         }
@@ -706,7 +706,7 @@ def test_unix_python3_gpu(lib_name) {
         ws('workspace/ut-python3-gpu') {
           try {
             utils.unpack_and_init(lib_name, mx_lib_cython)
-            python3_gpu_ut_cython('ubuntu_gpu_cu111')
+            python3_gpu_ut_cython('ubuntu_gpu_cu118')
             utils.publish_test_coverage()
           } finally {
             utils.collect_test_results_unix('tests_gpu.xml', 'tests_python3_gpu.xml')
@@ -722,7 +722,7 @@ def test_unix_python3_ampere_gpu(lib_name) {
         ws('workspace/ut-python3-gpu') {
           try {
             utils.unpack_and_init(lib_name, mx_lib_cython)
-            python3_gpu_ut_cython('ubuntu_gpu_cu111')
+            python3_gpu_ut_cython('ubuntu_gpu_cu118')
             utils.publish_test_coverage()
           } finally {
             utils.collect_test_results_unix('tests_gpu.xml', 'tests_python3_ampere_gpu.xml')
@@ -820,7 +820,7 @@ def test_unix_python3_onednn_gpu(lib_name) {
         ws('workspace/ut-python3-onednn-gpu') {
           try {
             utils.unpack_and_init(lib_name, mx_onednn_lib)
-            python3_gpu_ut('ubuntu_gpu_cu111')
+            python3_gpu_ut('ubuntu_gpu_cu118')
             utils.publish_test_coverage()
           } finally {
             utils.collect_test_results_unix('tests_gpu.xml', 'tests_python3_onednn_gpu.xml')
@@ -836,7 +836,7 @@ def test_unix_python3_onednn_nocudnn_gpu(lib_name) {
         ws('workspace/ut-python3-onednn-gpu-nocudnn') {
           try {
             utils.unpack_and_init(lib_name, mx_onednn_lib)
-            python3_gpu_ut_nocudnn('ubuntu_gpu_cu111')
+            python3_gpu_ut_nocudnn('ubuntu_gpu_cu118')
             utils.publish_test_coverage()
           } finally {
             utils.collect_test_results_unix('tests_gpu.xml', 'tests_python3_onednn_gpu_nocudnn.xml')
@@ -880,7 +880,7 @@ def test_unix_byteps_gpu(lib_name) {
         ws('workspace/it-byteps') {
           timeout(time: max_time, unit: 'MINUTES') {
             utils.unpack_and_init(lib_name, mx_lib)
-            utils.docker_run('ubuntu_gpu_cu111', 'integrationtest_ubuntu_gpu_byteps', true, '32768m')
+            utils.docker_run('ubuntu_gpu_cu118', 'integrationtest_ubuntu_gpu_byteps', true, '32768m')
             utils.publish_test_coverage()
           }
         }
@@ -894,7 +894,7 @@ def test_unix_distributed_kvstore_gpu(lib_name) {
         ws('workspace/it-dist-kvstore') {
           timeout(time: max_time, unit: 'MINUTES') {
             utils.unpack_and_init(lib_name, mx_lib)
-            utils.docker_run('ubuntu_gpu_cu111', 'integrationtest_ubuntu_gpu_dist_kvstore', true)
+            utils.docker_run('ubuntu_gpu_cu118', 'integrationtest_ubuntu_gpu_dist_kvstore', true)
             utils.publish_test_coverage()
           }
         }
@@ -908,7 +908,7 @@ def test_unix_cpp_package_gpu(lib_name) {
         ws('workspace/it-cpp-package-gpu') {
           timeout(time: max_time, unit: 'MINUTES') {
             utils.unpack_and_init(lib_name, mx_lib_cpp_examples)
-            utils.docker_run('ubuntu_gpu_cu111', 'integrationtest_ubuntu_cpp_package_gpu', true)
+            utils.docker_run('ubuntu_gpu_cu118', 'integrationtest_ubuntu_cpp_package_gpu', true)
             utils.publish_test_coverage()
           }
         }
@@ -922,7 +922,7 @@ def test_unix_python3_data_interchange_gpu(lib_name) {
         ws('workspace/it-data-interchange') {
           timeout(time: max_time, unit: 'MINUTES') {
             utils.unpack_and_init(lib_name, mx_lib)
-            utils.docker_run('ubuntu_gpu_cu111', 'test_python3_data_interchange_gpu', true)
+            utils.docker_run('ubuntu_gpu_cu118', 'test_python3_data_interchange_gpu', true)
             utils.publish_test_coverage()
           }
         }
@@ -982,7 +982,7 @@ def test_centos7_python3_gpu(lib_name) {
           timeout(time: max_time, unit: 'MINUTES') {
             try {
               utils.unpack_and_init(lib_name, mx_lib)
-              utils.docker_run('centos7_gpu_cu102', 'unittest_centos7_gpu', true)
+              utils.docker_run('centos7_gpu_cu118', 'unittest_centos7_gpu', true)
               utils.publish_test_coverage()
             } finally {
               utils.collect_test_results_unix('tests_gpu.xml', 'tests_python3_centos7_gpu.xml')
@@ -999,7 +999,7 @@ def test_centos7_python3_cd_gpu(lib_name) {
         ws('workspace/test-cd-static/gpu') {
           timeout(time: max_time, unit: 'MINUTES') {
             utils.unpack_and_init(lib_name, mx_cd_lib)
-            utils.docker_run('centos7_gpu_cu102', 'cd_unittest_ubuntu cu102', true)
+            utils.docker_run('centos7_gpu_cu118', 'cd_unittest_ubuntu cu118', true)
           }
         }
       }
@@ -1012,7 +1012,7 @@ def test_centos7_pypi_package_cd_gpu(lib_name) {
         ws('workspace/test-cd-pypi/gpu') {
           timeout(time: max_time, unit: 'MINUTES') {
             utils.unpack_and_init(lib_name, mx_cd_lib)
-            utils.docker_run('centos7_gpu_cu102', 'ci_package_pypi cu102', true)
+            utils.docker_run('centos7_gpu_cu118', 'ci_package_pypi cu118', true)
           }
         }
       }
@@ -1122,7 +1122,7 @@ def docs_python(lib_name) {
         ws('workspace/docs') {
           timeout(time: max_time, unit: 'MINUTES') {
             utils.unpack_and_init(lib_name, mx_lib_cython)
-            utils.docker_run('ubuntu_gpu_cu111', 'build_python_docs', true)
+            utils.docker_run('ubuntu_gpu_cu118', 'build_python_docs', true)
             if (should_pack_website()) {
               utils.pack_lib('python-artifacts', 'docs/_build/python-artifacts.tgz', false)
             }

ci/jenkins/Jenkinsfile_edge

@@ -34,7 +34,7 @@ utils.assign_node_labels(utility: 'utility', linux_cpu: 'mxnetlinux-cpu', linux_
 utils.main_wrapper(
 core_logic: {
   utils.parallel_stage('Build', [
-    custom_steps.compile_armv8_jetson_gpu(),
+    // custom_steps.compile_armv8_jetson_gpu(),
     custom_steps.compile_armv7_cpu('armv7'),
     custom_steps.compile_armv6_cpu('armv6'),
     custom_steps.compile_armv8_cpu('armv8'),

config/distribution/linux_cu118.cmake

@@ -0,0 +1,36 @@
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more contributor license agreements.  See the NOTICE file
+# distributed with this work for additional information
+# regarding copyright ownership.  The ASF licenses this file
+# to you under the Apache License, Version 2.0 (the
+# "License"); you may not use this file except in compliance
+# with the License.  You may obtain a copy of the License at
+#
+#   http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing,
+# software distributed under the License is distributed on an
+# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+# KIND, either express or implied.  See the License for the
+# specific language governing permissions and limitations
+# under the License.
+
+set(CMAKE_BUILD_TYPE "Distribution" CACHE STRING "Build type")
+set(CFLAGS "-mno-avx" CACHE STRING "CFLAGS")
+set(CXXFLAGS "-mno-avx" CACHE STRING "CXXFLAGS")
+
+set(USE_BLAS "open" CACHE STRING "BLAS Vendor")
+set(USE_CUDA ON CACHE BOOL "Build with CUDA support")
+set(USE_CUDNN ON CACHE BOOL "Build with CUDNN support")
+set(USE_NCCL ON CACHE BOOL "Build with NCCL support")
+set(USE_OPENCV ON CACHE BOOL "Build with OpenCV support")
+set(USE_OPENMP ON CACHE BOOL "Build with Openmp support")
+set(USE_ONEDNN ON CACHE BOOL "Build with oneDNN support")
+set(USE_LAPACK ON CACHE BOOL "Build with lapack support")
+set(USE_TVM_OP OFF CACHE BOOL "Enable use of TVM operator build system.")
+set(USE_SSE ON CACHE BOOL "Build with x86 SSE instruction support")
+set(USE_F16C OFF CACHE BOOL "Build with x86 F16C instruction support")
+set(USE_LIBJPEG_TURBO ON CACHE BOOL "Build with libjpeg-turbo")
+set(USE_DIST_KVSTORE ON CACHE BOOL "Build with DIST_KVSTORE support")
+set(CUDACXX "/usr/local/cuda-11.8/bin/nvcc" CACHE STRING "Cuda compiler")
+set(MXNET_CUDA_ARCH "5.0;6.0;7.0;8.0;8.6" CACHE STRING "Cuda architectures")