Merge branch 'sima/main' into sima/develop_pr

mikeseven · Mar 10, 2022 · 379039c · 379039c
2 parents f892d5f + fc3f8be
commit 379039c
Show file tree

Hide file tree

Showing 499 changed files with 25,291 additions and 10,568 deletions.
diff --git a/.github/workflows/cc_bot.yml b/.github/workflows/cc_bot.yml
@@ -32,6 +32,7 @@ concurrency:
 
 jobs:
   cc-reviewers:
+    if: github.repository == 'apache/tvm'
     runs-on: ubuntu-latest
     steps:
       - uses: actions/checkout@v2

diff --git a/.github/workflows/ping_reviewers.yml b/.github/workflows/ping_reviewers.yml
@@ -11,6 +11,7 @@ concurrency:
 
 jobs:
   ping:
+    if: github.repository == 'apache/tvm'
     runs-on: ubuntu-20.04
     steps:
       - uses: actions/checkout@v2

diff --git a/.github/workflows/tag_teams.yml b/.github/workflows/tag_teams.yml
@@ -0,0 +1,48 @@
+# 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.
+
+# GH actions.
+# We use it to cover windows and mac builds
+# Jenkins is still the primary CI
+
+name: Teams
+
+on:
+  # See https://docs.github.com/en/actions/using-workflows/events-that-trigger-workflows#pull_request_target
+  pull_request_target:
+    types: [opened, reopened, edited, ready_for_review, labeled]
+  issues:
+    types: [opened, edited, reopened, labeled]
+
+concurrency:
+  group: Teams-${{ github.event.pull_request.number }}-${{ github.event.issue.number }}
+  cancel-in-progress: true
+
+jobs:
+  tag-teams:
+    if: github.repository == 'apache/tvm'
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v2
+      - name: Tag people from relevant teams
+        env:
+          PR: ${{ toJson(github.event.pull_request) }}
+          ISSUE: ${{ toJson(github.event.issue) }}
+          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+        run: |
+          set -eux
+          python tests/scripts/github_tag_teams.py || echo failed
diff --git a/.github/workflows/update_last_successful_branch.yml b/.github/workflows/update_last_successful_branch.yml
@@ -32,6 +32,7 @@ concurrency:
 
 jobs:
   update-last-successful-branch:
+    if: github.repository == 'apache/tvm'
     runs-on: ubuntu-20.04
     steps:
       - uses: actions/checkout@v2

diff --git a/.gitignore b/.gitignore
@@ -251,3 +251,8 @@ _docs/
 jvm/target
 .config/configstore/
 .ci-py-scripts/
+
+# Generated Hexagon files
+src/runtime/hexagon/rpc/hexagon_rpc.h
+src/runtime/hexagon/rpc/hexagon_rpc_skel.c
+src/runtime/hexagon/rpc/hexagon_rpc_stub.c
diff --git a/CMakeLists.txt b/CMakeLists.txt
@@ -26,6 +26,16 @@ endif()
 tvm_option(USE_CUDA "Build with CUDA" OFF)
 tvm_option(USE_OPENCL "Build with OpenCL" OFF)
 tvm_option(USE_VULKAN "Build with Vulkan" OFF)
+
+
+# Whether to use spirv-tools.and SPIRV-Headers from Khronos github or gitlab.
+#
+# Possible values:
+# - OFF: not to use
+# - /path/to/install: path to your khronis spirv-tools and SPIRV-Headers installation directory
+#
+tvm_option(USE_KHRONOS_SPIRV "Whether to use spirv-tools.and SPIRV-Headers from Khronos github or gitlab" OFF)
+tvm_option(USE_SPIRV_KHR_INTEGER_DOT_PRODUCT "whether enable SPIRV_KHR_DOT_PRODUCT" OFF)
 tvm_option(USE_METAL "Build with Metal" OFF)
 tvm_option(USE_ROCM "Build with ROCM" OFF)
 tvm_option(ROCM_PATH "The path to rocm" /opt/rocm)
@@ -38,6 +48,7 @@ tvm_option(USE_LLVM "Build with LLVM, can be set to specific llvm-config path" O
 tvm_option(USE_STACKVM_RUNTIME "Include stackvm into the runtime" OFF)
 tvm_option(USE_GRAPH_EXECUTOR "Build with tiny graph executor" ON)
 tvm_option(USE_GRAPH_EXECUTOR_CUDA_GRAPH "Build with tiny graph executor with CUDA Graph for GPUs" OFF)
+tvm_option(USE_AOT_EXECUTOR "Build with AOT executor" ON)
 tvm_option(USE_PROFILER "Build profiler for the VM and graph executor" ON)
 tvm_option(USE_OPENMP "Build with OpenMP thread pool implementation" OFF)
 tvm_option(USE_RELAY_DEBUG "Building Relay in debug mode..." OFF)
@@ -78,6 +89,7 @@ tvm_option(USE_MIOPEN "Build with ROCM:MIOpen" OFF)
 tvm_option(USE_ROCBLAS "Build with ROCM:RoCBLAS" OFF)
 tvm_option(USE_SORT "Build with sort support" ON)
 tvm_option(USE_NNPACK "Build with nnpack support" OFF)
+tvm_option(USE_LIBTORCH "Build with libtorch support" OFF)
 tvm_option(USE_RANDOM "Build with random support" ON)
 tvm_option(USE_MICRO_STANDALONE_RUNTIME "Build with micro.standalone_runtime support" OFF)
 tvm_option(USE_CPP_RPC "Build CPP RPC" OFF)
@@ -395,6 +407,13 @@ if(USE_PROFILER)
   list(APPEND RUNTIME_SRCS ${RUNTIME_VM_PROFILER_SRCS})
 endif(USE_PROFILER)
 
+if(USE_AOT_EXECUTOR)
+  message(STATUS "Build with AOT Executor support...")
+  file(GLOB RUNTIME_AOT_EXECUTOR_SRCS src/runtime/aot_executor/*.cc)
+  list(APPEND RUNTIME_SRCS ${RUNTIME_AOT_EXECUTOR_SRCS})
+
+endif(USE_AOT_EXECUTOR)
+
 # Enable ctest if gtest is available
 if(USE_GTEST)
   # Check env var for backward compatibility. A better way to specify package
@@ -447,6 +466,7 @@ include(cmake/modules/contrib/Posit.cmake)
 include(cmake/modules/contrib/MicroStandaloneRuntime.cmake)
 include(cmake/modules/contrib/Sort.cmake)
 include(cmake/modules/contrib/NNPack.cmake)
+include(cmake/modules/contrib/LibTorch.cmake)
 include(cmake/modules/contrib/HybridDump.cmake)
 include(cmake/modules/contrib/TFLite.cmake)
 include(cmake/modules/contrib/TF_TVMDSOOP.cmake)

diff --git a/CONTRIBUTORS.md b/CONTRIBUTORS.md
@@ -55,7 +55,8 @@ We do encourage everyone to work anything they are interested in.
 - [Thierry Moreau](https://github.com/tmoreau89) (PMC): @tmoreau89 - vta
 - [Kazutaka Morita](https://github.com/kazum): @kazum - frontends, opencl
 - [Trevor Morris](https://github.com/trevor-m): @trevor-m - byoc, compiler
-- [Leandro Nunes](https://github.com/leandron): @leandron - tvmc
+- [Leandro Nunes](https://github.com/leandron) (PMC): @leandron - tvmc
+- [Lily Orth-Smith](https://github.com/electriclilies): @electriclilies - relay
 - [Krzysztof Parzyszek](https://github.com/kparzysz-quic): @kparzysz-quic - hexagon, llvm
 - [Andrew Reusch](https://github.com/areusch): (PMC) @areusch - runtime, microTVM
 - [Jared Roesch](https://github.com/jroesch) (PMC): @jroesch - relay
@@ -107,7 +108,7 @@ We do encourage everyone to work anything they are interested in.
 - [Manupa Karunaratne](https://github.com/manupa-arm): @manupa-arm
 - [Marisa Kirisame](https://github.com/MarisaKirisame): @MarisaKirisame
 - [Tristan Konolige](https://github.com/tkonolige): @tkonolige
-- [Ruihang Lai](https://github.com/MasterJH5574): @MasterJH5574 
+- [Ruihang Lai](https://github.com/MasterJH5574): @MasterJH5574
 - [Wuwei Lin](https://github.com/vinx13): @vinx13
 - [Andrew Liu](https://github.com/hypercubestart): @hypercubestart
 - [Henry Liu](https://github.com/optima2005): @optima2005

diff --git a/Jenkinsfile b/Jenkinsfile
@@ -45,14 +45,14 @@
 import org.jenkinsci.plugins.pipeline.modeldefinition.Utils
 
 // NOTE: these lines are scanned by docker/dev_common.sh. Please update the regex as needed. -->
-ci_lint = "tlcpack/ci-lint:v0.68"
-ci_gpu = "tlcpack/ci-gpu:v0.81"
-ci_cpu = "tlcpack/ci-cpu:v0.81"
-ci_wasm = "tlcpack/ci-wasm:v0.71"
-ci_i386 = "tlcpack/ci-i386:v0.74"
-ci_qemu = "tlcpack/ci-qemu:v0.10"
-ci_arm = "tlcpack/ci-arm:v0.07"
-ci_hexagon = "tlcpack/ci-hexagon:v0.01"
+ci_lint = 'tlcpack/ci-lint:v0.68'
+ci_gpu = 'tlcpack/ci-gpu:v0.81'
+ci_cpu = 'tlcpack/ci-cpu:v0.81'
+ci_wasm = 'tlcpack/ci-wasm:v0.71'
+ci_i386 = 'tlcpack/ci-i386:v0.74'
+ci_qemu = 'tlcpack/ci-qemu:v0.10'
+ci_arm = 'tlcpack/ci-arm:v0.07'
+ci_hexagon = 'tlcpack/ci-hexagon:v0.01'
 // <--- End of regex-scanned config.
 
 // Parameters to allow overriding (in Jenkins UI), the images
@@ -105,6 +105,21 @@ def init_git() {
   }
 }
 
+def should_skip_slow_tests(pr_number) {
+  withCredentials([string(
+    credentialsId: 'tvm-bot-jenkins-reader',
+    variable: 'GITHUB_TOKEN',
+  )]) {
+    // Exit code of 1 means run slow tests, exit code of 0 means skip slow tests
+    result = sh (
+      returnStatus: true,
+      script: "./tests/scripts/should_run_slow_tests.py --pr '${pr_number}'",
+      label: 'Check if CI should run slow tests',
+    )
+  }
+  return result == 0
+}
+
 def cancel_previous_build() {
   // cancel previous build if it is not on main.
   if (env.BRANCH_NAME != 'main') {
@@ -117,6 +132,18 @@ def cancel_previous_build() {
 }
 
 def should_skip_ci(pr_number) {
+  if (!env.BRANCH_NAME.startsWith('PR-')) {
+    // never skip CI on build sourced from a branch
+    return false
+  }
+  glob_skip_ci_code = sh (
+    returnStatus: true,
+    script: "./tests/scripts/git_skip_ci_globs.py",
+    label: 'Check if CI should be skipped due to changed files',
+  )
+  if (glob_skip_ci_code == 0) {
+    return true
+  }
   withCredentials([string(
     credentialsId: 'tvm-bot-jenkins-reader',
     variable: 'TOKEN',
@@ -128,10 +155,18 @@ def should_skip_ci(pr_number) {
       script: "./tests/scripts/git_skip_ci.py --pr '${pr_number}'",
       label: 'Check if CI should be skipped',
     )
-    }
+  }
   return git_skip_ci_code == 0
 }
 
+// skips builds from branch indexing; sourced from https://www.jvt.me/posts/2020/02/23/jenkins-multibranch-skip-branch-index/
+// execute this before anything else, including requesting any time on an agent
+if (currentBuild.getBuildCauses().toString().contains('BranchIndexingCause')) {
+  print "INFO: Build skipped due to trigger being Branch Indexing"
+  currentBuild.result = 'ABORTED' // optional, gives a better hint to the user that it's been skipped, rather than the default which shows it's successful
+  return
+}
+
 cancel_previous_build()
 
 stage('Prepare') {
@@ -169,6 +204,7 @@ stage('Sanity Check') {
           label: 'Check for docs only changes',
         )
         skip_ci = should_skip_ci(env.CHANGE_ID)
+        skip_slow_tests = should_skip_slow_tests(env.CHANGE_ID)
         sh (
           script: "${docker_run} ${ci_lint}  ./tests/scripts/task_lint.sh",
           label: 'Run lint',
@@ -178,6 +214,7 @@ stage('Sanity Check') {
   }
 }
 
+
 // Run make. First try to do an incremental make from a previous workspace in hope to
 // accelerate the compilation. If something is wrong, clean the workspace and then
 // build from scratch.
@@ -238,13 +275,13 @@ def python_unittest(image) {
 def fsim_test(image) {
   sh (
     script: "${docker_run} ${image} ./tests/scripts/task_python_vta_fsim.sh",
-    label: 'Run VTA tests in FSIM ',
+    label: 'Run VTA tests in FSIM',
   )
 }
 
 def cmake_build(image, path, make_flag) {
   sh (
-    script: "${docker_run} ${image} ./tests/scripts/task_build.sh ${path} ${make_flag}",
+    script: "${docker_run} ${image} ./tests/scripts/task_build.py --num-executors ${CI_NUM_EXECUTORS} --sccache-bucket tvm-sccache-prod",
     label: 'Run cmake build',
   )
 }
@@ -257,6 +294,9 @@ def cpp_unittest(image) {
 }
 
 stage('Build') {
+  environment {
+    SKIP_SLOW_TESTS = "${skip_slow_tests}"
+  }
   parallel 'BUILD: GPU': {
     if (!skip_ci) {
       node('GPUBUILD') {
@@ -287,7 +327,7 @@ stage('Build') {
             ci_setup(ci_cpu)
             // sh "${docker_run} ${ci_cpu} ./tests/scripts/task_golang.sh"
             // TODO(@jroesch): need to resolve CI issue will turn back on in follow up patch
-            sh (script: "${docker_run} ${ci_cpu} ./tests/scripts/task_rust.sh", label: "Rust build and test")
+            sh (script: "${docker_run} ${ci_cpu} ./tests/scripts/task_rust.sh", label: 'Rust build and test')
           }
         }
       }
@@ -402,6 +442,10 @@ stage('Build') {
               script: "${docker_run} ${ci_hexagon} ./tests/scripts/task_python_hexagon.sh",
               label: 'Run Hexagon tests',
             )
+            sh (
+              script: "${docker_run} ${ci_hexagon} ./tests/scripts/task_python_hexagon_simulator.sh",
+              label: 'Run Hexagon tests on simulator',
+            )
           } finally {
             junit 'build/pytest-results/*.xml'
           }
@@ -414,6 +458,9 @@ stage('Build') {
 }
 
 stage('Test') {
+  environment {
+    SKIP_SLOW_TESTS = "${skip_slow_tests}"
+  }
   parallel 'unittest: GPU': {
     if (!skip_ci && is_docs_only_build != 1) {
       node('TensorCore') {
@@ -471,7 +518,7 @@ stage('Test') {
   'unittest: CPU': {
     if (!skip_ci && is_docs_only_build != 1) {
       node('CPU') {
-        ws(per_exec_ws("tvm/ut-python-cpu")) {
+        ws(per_exec_ws('tvm/ut-python-cpu')) {
           try {
             init_git()
             unpack_lib('cpu', tvm_multilib_tsim)
@@ -481,7 +528,7 @@ stage('Test') {
               fsim_test(ci_cpu)
               sh (
                 script: "${docker_run} ${ci_cpu} ./tests/scripts/task_python_vta_tsim.sh",
-                label: "Run VTA tests in TSIM",
+                label: 'Run VTA tests in TSIM',
               )
             }
           } finally {
@@ -532,6 +579,10 @@ stage('Test') {
                 script: "${docker_run} ${ci_arm} ./tests/scripts/task_python_arm_compute_library.sh",
                 label: 'Run test_arm_compute_lib test',
               )
+              sh (
+                script: "${docker_run} ${ci_arm} ./tests/scripts/task_python_topi.sh",
+                label: 'Run TOPI tests',
+              )
             // sh "${docker_run} ${ci_arm} ./tests/scripts/task_python_integration.sh"
             }
           } finally {
@@ -566,7 +617,7 @@ stage('Test') {
       Utils.markStageSkippedForConditional('topi: GPU')
     }
   },
-  'frontend: GPU': {
+  'frontend: GPU 1': {
     if (!skip_ci && is_docs_only_build != 1) {
       node('GPU') {
         ws(per_exec_ws('tvm/frontend-python-gpu')) {
@@ -576,8 +627,31 @@ stage('Test') {
             timeout(time: max_time, unit: 'MINUTES') {
               ci_setup(ci_gpu)
               sh (
-                script: "${docker_run} ${ci_gpu} ./tests/scripts/task_python_frontend.sh",
-                label: 'Run Python frontend tests',
+                script: "${docker_run} ${ci_gpu} ./tests/scripts/task_python_frontend.sh 1",
+                label: 'Run Python frontend tests (shard 1)',
+              )
+            }
+          } finally {
+            junit 'build/pytest-results/*.xml'
+          }
+        }
+      }
+     } else {
+      Utils.markStageSkippedForConditional('frontend: GPU 1')
+    }
+  },
+  'frontend: GPU 2': {
+    if (!skip_ci && is_docs_only_build != 1) {
+      node('GPU') {
+        ws(per_exec_ws('tvm/frontend-python-gpu')) {
+          try {
+            init_git()
+            unpack_lib('gpu', tvm_multilib)
+            timeout(time: max_time, unit: 'MINUTES') {
+              ci_setup(ci_gpu)
+              sh (
+                script: "${docker_run} ${ci_gpu} ./tests/scripts/task_python_frontend.sh 2",
+                label: 'Run Python frontend tests (shard 2)',
               )
             }
           } finally {
@@ -586,7 +660,7 @@ stage('Test') {
         }
       }
      } else {
-      Utils.markStageSkippedForConditional('frontend: GPU')
+      Utils.markStageSkippedForConditional('frontend: GPU 2')
     }
   },
   'frontend: CPU': {
-Original file line number
+Diff line change
@@ Expand Up / @@ -11,6 +11,7 @@ concurrency: @@
     jobs:
       ping:
+        if: github.repository == 'apache/tvm'
         runs-on: ubuntu-20.04
         steps:
           - uses: actions/checkout@v2
@@ Expand Down @@