Merge branch 'main' into yc/0806-exclude-shadow-from-coredump

Author: yingcong-wu

.github/workflows/benchmarks_compute.yml

@@ -12,6 +12,7 @@ on:
         default: 'level_zero'
         options:
           - level_zero
+          - level_zero_v2
       unit:
         description: Test unit (cpu/gpu)
         type: choice
@@ -57,7 +58,8 @@ jobs:
         adapter: [
           {str_name: "${{inputs.str_name}}",
           sycl_config: "${{inputs.sycl_config_params}}",
-          unit: "${{inputs.unit}}"}
+          unit: "${{inputs.unit}}"
+          }
         ]
         build_type: [Release]
         compiler: [{c: clang, cxx: clang++}]
@@ -141,8 +143,8 @@ jobs:
         --ci-defaults ${{matrix.adapter.sycl_config}}
         --cmake-opt="-DLLVM_INSTALL_UTILS=ON"
         --cmake-opt="-DSYCL_PI_TESTS=OFF"
-        --cmake-opt="-DSYCL_PI_UR_USE_FETCH_CONTENT=OFF"
-        --cmake-opt="-DSYCL_PI_UR_SOURCE_DIR=${{github.workspace}}/ur-repo/"
+        --cmake-opt="-DSYCL_UR_USE_FETCH_CONTENT=OFF"
+        --cmake-opt="-DSYCL_UR_SOURCE_DIR=${{github.workspace}}/ur-repo/"
         --cmake-opt=-DCMAKE_C_COMPILER_LAUNCHER=ccache
         --cmake-opt=-DCMAKE_CXX_COMPILER_LAUNCHER=ccache
 
@@ -155,7 +157,7 @@ jobs:
 
     - name: Run benchmarks
       id: benchmarks
-      run: numactl -N 0 ${{ github.workspace }}/ur-repo/scripts/benchmarks/main.py ~/bench_workdir ${{github.workspace}}/sycl_build ${{ inputs.bench_script_params }}
+      run: numactl -N 0 ${{ github.workspace }}/ur-repo/scripts/benchmarks/main.py ~/bench_workdir ${{github.workspace}}/sycl_build ${{github.workspace}}/ur-repo ${{ matrix.adapter.str_name }} ${{ inputs.bench_script_params }}
 
     - name: Add comment to PR
       uses: actions/github-script@60a0d83039c74a4aee543508d2ffcb1c3799cdea # v7.0.1

.github/workflows/build-hw-reusable.yml

@@ -4,7 +4,10 @@ name: Build - Adapters on HW - Reusable
 on:
   workflow_call:
     inputs:
-      name:
+      adapter_name:
+        required: true
+        type: string
+      runner_name:
         required: true
         type: string
       platform:
@@ -33,7 +36,7 @@ jobs:
     strategy:
       matrix:
         adapter: [
-          {name: "${{inputs.name}}", platform: "${{inputs.platform}}", static_Loader: "${{inputs.static_loader}}"},
+          {name: "${{inputs.adapter_name}}", platform: "${{inputs.platform}}", static_Loader: "${{inputs.static_loader}}"},
         ]
         build_type: [Debug, Release]
         compiler: [{c: gcc, cxx: g++}, {c: clang, cxx: clang++}]
@@ -47,7 +50,7 @@ jobs:
          - adapter: {static_Loader: ON}
            compiler: {c: clang, cxx: clang++}
 
-    runs-on: ${{matrix.adapter.name}}
+    runs-on: ${{inputs.runner_name}}
 
     steps:
     - uses: actions/checkout@b4ffde65f46336ab88eb53be808477a3936bae11 # v4.1.1

.github/workflows/cmake.yml

@@ -20,6 +20,7 @@ jobs:
         compiler: [{c: gcc, cxx: g++}]
         libbacktrace: ['-DVAL_USE_LIBBACKTRACE_BACKTRACE=OFF']
         pool_tracking: ['-DUMF_ENABLE_POOL_TRACKING=ON', '-DUMF_ENABLE_POOL_TRACKING=OFF']
+        latency_tracking: ['-DUMF_ENABLE_LATENCY_TRACKING=OFF']
         include:
           - os: 'ubuntu-22.04'
             build_type: Release
@@ -36,6 +37,10 @@ jobs:
           - os: 'ubuntu-20.04'
             build_type: Release
             compiler: {c: gcc-7, cxx: g++-7}
+          - os: 'ubuntu-22.04'
+            build_type: Release
+            compiler: {c: clang, cxx: clang++}
+            latency_tracking: '-DUMF_ENABLE_LATENCY_TRACKING=ON'
     runs-on: ${{ (matrix.os == 'ubuntu-22.04' && github.repository_owner == 'oneapi-src') && 'intel-ubuntu-22.04' || matrix.os }}
 
     steps:
@@ -92,6 +97,7 @@ jobs:
         -DUR_DPCXX=${{github.workspace}}/dpcpp_compiler/bin/clang++
         ${{matrix.libbacktrace}}
         ${{matrix.pool_tracking}}
+        ${{matrix.latency_tracking}}
 
     - name: Configure CMake
       if: matrix.os == 'ubuntu-20.04'
@@ -106,6 +112,7 @@ jobs:
         -DUR_FORMAT_CPP_STYLE=ON
         ${{matrix.libbacktrace}}
         ${{matrix.pool_tracking}}
+        ${{matrix.latency_tracking}}
 
     - name: Generate source from spec, check for uncommitted diff
       if: matrix.os == 'ubuntu-22.04'
@@ -131,39 +138,52 @@ jobs:
     name: Level Zero
     uses: ./.github/workflows/build-hw-reusable.yml
     with:
-      name: L0
+      adapter_name: L0
+      runner_name: L0
+
+  level-zero-v2:
+    name: Level Zero V2
+    uses: ./.github/workflows/build-hw-reusable.yml
+    with:
+      adapter_name: L0_V2
+      runner_name: L0
 
   level-zero-static:
     name: Level Zero static
     uses: ./.github/workflows/build-hw-reusable.yml
     with:
-      name: L0
+      adapter_name: L0
+      runner_name: L0
       static_loader: ON
 
   opencl:
     name: OpenCL
     uses: ./.github/workflows/build-hw-reusable.yml
     with:
-      name: OPENCL
+      adapter_name: OPENCL
+      runner_name: OPENCL
       platform: "Intel(R) OpenCL"
 
   cuda:
     name: CUDA
     uses: ./.github/workflows/build-hw-reusable.yml
     with:
-      name: CUDA
+      adapter_name: CUDA
+      runner_name: CUDA
 
   hip:
     name: HIP
     uses: ./.github/workflows/build-hw-reusable.yml
     with:
-      name: HIP
+      adapter_name: HIP
+      runner_name: HIP
 
   native-cpu:
     name: Native CPU
     uses: ./.github/workflows/build-hw-reusable.yml
     with:
-      name: NATIVE_CPU
+      adapter_name: NATIVE_CPU
+      runner_name: NATIVE_CPU
 
   e2e-level-zero:
     name: E2E L0
@@ -195,15 +215,20 @@ jobs:
       matrix:
         os: ['windows-2019', 'windows-2022']
         adapter: [
-          {name: None, var: ''}, {name: L0, var: '-DUR_BUILD_ADAPTER_L0=ON'}
+          {name: None, var: ''}, {name: L0, var: '-DUR_BUILD_ADAPTER_L0=ON'},
+          {name: None, var: ''}, {name: L0_V2, var: '-DUR_BUILD_ADAPTER_L0_V2=ON'}
         ]
 
         # TODO: building level zero loader on windows-2019 and clang-cl is currently broken
         exclude:
          - os: 'windows-2019'
            adapter: {name: L0, var: '-DUR_BUILD_ADAPTER_L0=ON'}
+         - os: 'windows-2019'
+           adapter: {name: L0_V2, var: '-DUR_BUILD_ADAPTER_L0_V2=ON'}
          - adapter: {name: L0, var: '-DUR_BUILD_ADAPTER_L0=ON'}
            compiler: {c: clang-cl, cxx: clang-cl}
+         - adapter: {name: L0_V2, var: '-DUR_BUILD_ADAPTER_L0_V2=ON'}
+           compiler: {c: clang-cl, cxx: clang-cl}
 
         build_type: [Debug, Release]
         compiler: [{c: cl, cxx: cl}, {c: clang-cl, cxx: clang-cl}]

.github/workflows/coverity.yml

@@ -1,5 +1,5 @@
 #
-# Copyright (C) 2023 Intel Corporation
+# Copyright (C) 2023-2024 Intel Corporation
 #
 # Part of the Unified-Runtime Project, under the Apache License v2.0 with LLVM Exceptions.
 # See LICENSE.TXT
@@ -38,11 +38,6 @@ jobs:
       - name: Install pip packages
         run: pip install -r third_party/requirements.txt
 
-      - name: Install apt packages
-        run: |
-          sudo apt-get update
-          sudo apt-get install -y libhwloc-dev
-
       - name: Configure CMake
         run: >
           cmake

.github/workflows/e2e_core.yml

@@ -130,8 +130,8 @@ jobs:
         --ci-defaults ${{matrix.adapter.config}}
         --cmake-opt="-DLLVM_INSTALL_UTILS=ON"
         --cmake-opt="-DSYCL_PI_TESTS=OFF"
-        --cmake-opt="-DSYCL_PI_UR_USE_FETCH_CONTENT=OFF"
-        --cmake-opt="-DSYCL_PI_UR_SOURCE_DIR=${{github.workspace}}/ur-repo/"
+        --cmake-opt="-DSYCL_UR_USE_FETCH_CONTENT=OFF"
+        --cmake-opt="-DSYCL_UR_SOURCE_DIR=${{github.workspace}}/ur-repo/"
         --cmake-opt=-DCMAKE_C_COMPILER_LAUNCHER=ccache
         --cmake-opt=-DCMAKE_CXX_COMPILER_LAUNCHER=ccache
 
@@ -157,7 +157,7 @@ jobs:
     - name: Setup SYCL variables
       run: |
         which clang++ sycl-ls
-        SYCL_PI_TRACE=-1 sycl-ls
+        SYCL_UR_TRACE=-1 sycl-ls
 
     - name: Build e2e tests
       run: >

.github/workflows/e2e_cuda.yml

@@ -20,3 +20,4 @@ jobs:
       prefix: "ext_oneapi_"
       config: "--cuda"
       unit: "gpu"
+      extra_lit_flags: "-sv --max-time=3600"

.github/workflows/e2e_level_zero.yml

@@ -21,9 +21,9 @@ jobs:
       config: ""
       unit: "gpu"
       # Failing tests
-      xfail: "Config/select_device.cpp;DeviceCodeSplit/grf.cpp;ESIMD/grf.cpp;ESIMD/named_barriers/loop_extended.cpp;KernelAndProgram/target_register_alloc_mode.cpp;Matrix/SG32/element_wise_all_ops.cpp;Matrix/SG32/get_coord_int8_matB.cpp;Matrix/get_coord_int8_matB.cpp;Matrix/joint_matrix_prefetch.cpp;Matrix/joint_matrix_rowmajorA_rowmajorB.cpp;Plugin/level_zero_barrier_optimization.cpp"
+      xfail: "DeviceCodeSplit/grf.cpp;ESIMD/grf.cpp;ESIMD/mask_expand_load.cpp;KernelAndProgram/target_register_alloc_mode.cpp;Matrix/SG32/get_coord_int8_matB.cpp;Matrix/get_coord_int8_matB.cpp;Matrix/joint_matrix_prefetch.cpp;Matrix/joint_matrix_rowmajorA_rowmajorB.cpp;  SYCL :: ESIMD/mask_expand_load.cpp;Matrix/SPVCooperativeMatrix/joint_matrix_bf16_fill_k_cache_OOB.cpp;Matrix/SPVCooperativeMatrix/joint_matrix_out_bounds.cpp;Matrix/SPVCooperativeMatrix/joint_matrix_prefetch.cpp;Matrix/SPVCooperativeMatrix/joint_matrix_rowmajorA_rowmajorB.cpp;Matrix/element_wise_all_ops_1d.cpp;Matrix/element_wise_all_ops_1d_cont.cpp;Matrix/element_wise_all_ops_scalar.cpp;Matrix/joint_matrix_bf16_fill_k_cache_OOB.cpp;Matrix/joint_matrix_bf16_fill_k_cache_prefetch.cpp;Matrix/joint_matrix_out_bounds.cpp;Matrix/joint_matrix_unaligned_k.cpp"
       # Flaky tests
-      filter_out: ""
+      filter_out: "ESIMD/named_barriers/loop_extended.cpp;ESIMD/local_accessor_copy_to_from.cpp;"
       # These runners by default spawn upwards of 260 workers.
       # We also add a time out just in case some test hangs
-      extra_lit_flags: "--param gpu-intel-pvc=True --param gpu-intel-pvc-1T=True -sv -j 100 --max-time 600"
+      extra_lit_flags: "--param gpu-intel-pvc=True --param gpu-intel-pvc-1T=True -sv -j 100 --max-time=3600"

.github/workflows/e2e_opencl.yml

@@ -20,3 +20,5 @@ jobs:
       prefix: ""
       config: ""
       unit: "cpu"
+      xfail: "AOT/double.cpp;AOT/half.cpp;AOT/reqd-sg-size.cpp;Basic/built-ins/marray_geometric.cpp;KernelCompiler/kernel_compiler_spirv.cpp;KernelCompiler/opencl_queries.cpp"
+      extra_lit_flags: "-sv --max-time=3600"

.github/workflows/nightly.yml

@@ -1,6 +1,7 @@
 name: Nightly
 
 on:
+  workflow_dispatch:
   schedule:
     # Run every day at 23:00 UTC
     - cron: '0 23 * * *'

CMakeLists.txt

@@ -4,7 +4,7 @@
 # SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
 
 cmake_minimum_required(VERSION 3.20.0 FATAL_ERROR)
-project(unified-runtime VERSION 0.10.0)
+project(unified-runtime VERSION 0.11.0)
 
 # Check if unified runtime is built as a standalone project.
 if(CMAKE_SOURCE_DIR STREQUAL CMAKE_CURRENT_SOURCE_DIR OR UR_STANDALONE_BUILD)
@@ -51,12 +51,14 @@ option(UR_BUILD_ADAPTER_CUDA "Build the CUDA adapter" OFF)
 option(UR_BUILD_ADAPTER_HIP "Build the HIP adapter" OFF)
 option(UR_BUILD_ADAPTER_NATIVE_CPU "Build the Native-CPU adapter" OFF)
 option(UR_BUILD_ADAPTER_ALL "Build all currently supported adapters" OFF)
+option(UR_BUILD_ADAPTER_L0_V2 "Build the (experimental) Level-Zero v2 adapter" OFF)
 option(UR_BUILD_EXAMPLE_CODEGEN "Build the codegen example." OFF)
 option(VAL_USE_LIBBACKTRACE_BACKTRACE "enable libbacktrace validation backtrace for linux" OFF)
 option(UR_ENABLE_ASSERTIONS "Enable assertions for all build types" OFF)
 option(UR_BUILD_XPTI_LIBS "Build the XPTI libraries when tracing is enabled" ON)
 option(UR_STATIC_LOADER "Build loader as a static library" OFF)
 option(UR_FORCE_LIBSTDCXX "Force use of libstdc++ in a build using libc++ on Linux" OFF)
+option(UR_ENABLE_LATENCY_HISTOGRAM "Enable latncy histogram" OFF)
 set(UR_DPCXX "" CACHE FILEPATH "Path of the DPC++ compiler executable")
 set(UR_DPCXX_BUILD_FLAGS "" CACHE STRING "Build flags to pass to DPC++ when compiling device programs")
 set(UR_SYCL_LIBRARY_DIR "" CACHE PATH