Merge remote-tracking branch 'origin/sycl' into llvmspirv_pulldown

Author: jsji

.github/CODEOWNERS

@@ -139,18 +139,18 @@ devops/dependencies-igc-dev.json @intel/sycl-matrix-reviewers @intel/dpcpp-esimd
 # Benchmarking scripts
 devops/scripts/benchmarks/ @intel/llvm-reviewers-benchmarking
 
-# Kernel fusion JIT compiler
-sycl-jit/ @intel/dpcpp-kernel-fusion-reviewers
-sycl/doc/design/KernelFusionJIT.md @intel/dpcpp-kernel-fusion-reviewers
-sycl/doc/extensions/experimental/sycl_ext_codeplay_kernel_fusion.asciidoc @intel/dpcpp-kernel-fusion-reviewers
-sycl/include/sycl/ext/codeplay/experimental/fusion_properties.hpp  @intel/dpcpp-kernel-fusion-reviewers
-sycl/include/sycl/ext/codeplay/experimental/fusion_wrapper.hpp @intel/dpcpp-kernel-fusion-reviewers
-sycl/source/detail/fusion/ @intel/dpcpp-kernel-fusion-reviewers
-sycl/source/detail/jit_compiler.hpp @intel/dpcpp-kernel-fusion-reviewers
-sycl/source/detail/jit_compiler.cpp @intel/dpcpp-kernel-fusion-reviewers
-sycl/source/detail/jit_device_binaries.hpp @intel/dpcpp-kernel-fusion-reviewers
-sycl/source/detail/jit_device_binaries.cpp @intel/dpcpp-kernel-fusion-reviewers
-sycl/test-e2e/KernelFusion @intel/dpcpp-kernel-fusion-reviewers
+# JIT compiler
+sycl-jit/ @intel/llvm-reviewers-runtime @cperkinsintel @aelovikov-intel
+sycl/doc/design/KernelFusionJIT.md @intel/llvm-reviewers-runtime @cperkinsintel @aelovikov-intel
+sycl/doc/extensions/experimental/sycl_ext_codeplay_kernel_fusion.asciidoc @intel/llvm-reviewers-runtime @cperkinsintel @aelovikov-intel
+sycl/include/sycl/ext/codeplay/experimental/fusion_properties.hpp  @intel/llvm-reviewers-runtime @cperkinsintel @aelovikov-intel
+sycl/include/sycl/ext/codeplay/experimental/fusion_wrapper.hpp @intel/llvm-reviewers-runtime @cperkinsintel @aelovikov-intel
+sycl/source/detail/fusion/ @intel/llvm-reviewers-runtime @cperkinsintel @aelovikov-intel
+sycl/source/detail/jit_compiler.hpp @intel/llvm-reviewers-runtime @cperkinsintel @aelovikov-intel
+sycl/source/detail/jit_compiler.cpp @intel/llvm-reviewers-runtime @cperkinsintel @aelovikov-intel
+sycl/source/detail/jit_device_binaries.hpp @intel/llvm-reviewers-runtime @cperkinsintel @aelovikov-intel
+sycl/source/detail/jit_device_binaries.cpp @intel/llvm-reviewers-runtime @cperkinsintel @aelovikov-intel
+sycl/test-e2e/KernelFusion @intel/llvm-reviewers-runtime @cperkinsintel @aelovikov-intel
 
 # Matrix
 sycl/include/sycl/ext/oneapi/matrix/ @intel/sycl-matrix-reviewers
@@ -178,12 +178,6 @@ sycl/doc/design/CommandGraph.md @intel/sycl-graphs-reviewers
 sycl/doc/extensions/**/sycl_ext_oneapi_graph.asciidoc @intel/sycl-graphs-reviewers
 sycl/doc/syclgraph/ @intel/sycl-graphs-reviewers
 
-# syclcompat library
-sycl/**/syclcompat/ @intel/syclcompat-lib-reviewers
-sycl/cmake/modules/AddSYCLLibraryUnitTest.cmake @intel/syclcompat-lib-reviewers
-sycl/include/syclcompat.hpp @intel/syclcompat-lib-reviewers
-sycl/test-e2e/syclcompat/ @intel/syclcompat-lib-reviewers
-
 # bindless images
 sycl/doc/extensions/experimental/sycl_ext_oneapi_bindless_images.asciidoc @intel/bindless-images-reviewers
 sycl/include/sycl/ext/oneapi/bindless* @intel/bindless-images-reviewers

.github/workflows/sycl-docs.yml

@@ -10,13 +10,24 @@ on:
     - '.github/workflows/sycl-docs.yml'
     - 'clang/docs/**'
     - 'sycl/doc/**'
+    - 'devops/benchmarks/scripts/html/**'
   push:
     branches:
     - sycl
     paths:
     - '.github/workflows/sycl-docs.yml'
     - 'clang/docs/**'
     - 'sycl/doc/**'
+    - 'devops/benchmarks/scripts/html/**'
+  workflow_dispatch:
+    inputs:
+      update_gh_pages:
+        type: choice
+        description: Update Github Pages
+        options:
+          - true
+          - false
+        default: true
 
 permissions:
   contents: read
@@ -62,5 +73,5 @@ jobs:
       with:
         path: ./install_docs
     - name: Deploy to GitHub Pages
-      if: ${{ github.event_name == 'push' }}
+      if: ${{ github.event_name == 'push' || inputs.update_gh_pages == 'true' }}
       uses: actions/deploy-pages@v4

.github/workflows/sycl-linux-precommit.yml

@@ -62,6 +62,42 @@ jobs:
       e2e_binaries_artifact: e2e_bin
       e2e_binaries_preview_artifact: e2e_bin_preview
 
+  # Build and run native cpu e2e tests separately as cannot currently
+  # build all the e2e tests
+  build_run_native_cpu_e2e_tests:
+    if: ${{ always() && !cancelled() && needs.build.outputs.build_conclusion == 'success' }}  
+    runs-on: [Linux, build]
+    needs: [build]
+    container:
+      image: ghcr.io/intel/llvm/sycl_ubuntu2404_nightly:latest
+      options: -u 1001:1001
+    steps:
+      - uses: actions/checkout@v4
+        with:
+          sparse-checkout: |
+            devops/
+
+      # download build artefact
+      - name: Download toolchain
+        uses: actions/download-artifact@v4
+        with:
+          name: sycl_linux_default
+      - name: Extract SYCL toolchain
+        shell: bash
+        run: |
+          mkdir toolchain
+          tar -xf llvm_sycl.tar.zst -C toolchain
+          rm llvm_sycl.tar.zst
+      - name: Build and run E2E tests
+        uses: ./devops/actions/run-tests/e2e
+        with:
+          ref: ${{ inputs.ref || github.sha }}
+          testing_mode: full
+          target_devices: native_cpu:cpu
+          sycl_compiler: $GITHUB_WORKSPACE/toolchain/bin/clang++
+          extra_lit_opts: --param sycl_build_targets="native_cpu"
+          extra_cmake_args: -DSYCL_TEST_E2E_TARGETS="native_cpu:cpu" -DSYCL_TEST_E2E_STANDALONE=ON
+
   # If a PR changes CUDA adapter, run the build on Ubuntu 22.04 as well.
   # Ubuntu 22.04 container has CUDA 12.1 installed while Ubuntu 24.0 image
   # has CUDA 12.6.1 installed.
@@ -100,14 +136,14 @@ jobs:
         echo FILTER_6_2="$(grep -v '^#\|^\W*$' devops/compat_ci_exclude.sycl-rel-6_2 | paste -sd '|')" >> $GITHUB_OUTPUT
         echo FILTER_6_3="$(grep -v '^#\|^\W*$' devops/compat_ci_exclude.sycl-rel-6_3 | paste -sd '|')" >> $GITHUB_OUTPUT
 
-  run_prebuilt_e2e_tests:
+  E2E:
     needs: [build, detect_changes, compat_read_exclude]
     if: ${{ always() && !cancelled() && needs.build.outputs.build_conclusion == 'success' }}
     strategy:
       fail-fast: false
       matrix:
         include:
-          - name: GEN 12 Integrated
+          - name: Intel / GEN 12 Integrated
             runner: '["Linux", "gen12"]'
             target_devices: level_zero:gpu;opencl:gpu;opencl:cpu
           - name: NVIDIA/CUDA
@@ -119,39 +155,39 @@ jobs:
             image_options: -u 1001 --device=/dev/dri --device=/dev/kfd
             target_devices: hip:gpu
             extra_lit_opts: -j 1
-          - name: Intel Arc A-Series Graphics
+          - name: Intel / Arc A-Series Graphics
             runner: '["Linux", "arc"]'
             target_devices: level_zero:gpu;opencl:gpu;level_zero_v2:gpu
-          - name: Dev IGC / Intel Arc A-Series Graphics
+          - name: Intel Dev IGC / Arc A-Series Graphics
             runner: '["Linux", "arc"]'
             image: ghcr.io/intel/llvm/ubuntu2404_intel_drivers:devigc
             target_devices: level_zero:gpu
             use_igc_dev: true
             env: '{"LIT_FILTER":"Matrix/"}'
-          - name: Intel Ponte Vecchio GPU
+          - name: Intel / Ponte Vecchio GPU
             runner: '["Linux", "pvc"]'
             target_devices: level_zero:gpu;opencl:gpu;level_zero_v2:gpu
-          - name: Dev IGC / Intel Ponte Vecchio GPU
+          - name: Intel Dev IGC / Ponte Vecchio GPU
             runner: '["Linux", "pvc"]'
             image: ghcr.io/intel/llvm/ubuntu2404_intel_drivers:devigc
             target_devices: level_zero:gpu
             use_igc_dev: true
             env: '{"LIT_FILTER":"Matrix/"}'
-          - name: Intel Battlemage Graphics
+          - name: Intel / Battlemage Graphics
             runner: '["Linux", "bmg"]'
             target_devices: level_zero_v1:gpu;level_zero_v2:gpu
           - name: Preview Mode
             runner: '["Linux", "gen12"]'
             target_devices: level_zero:gpu;opencl:gpu;opencl:cpu
             extra_lit_opts: --param test-preview-mode=True
             e2e_binaries_artifact: e2e_bin_preview
-          - name: ABI compatibility against sycl-rel-6_2
+          - name: ABI compatibility / sycl-rel-6_2
             runner: '["Linux", "pvc"]'
             image: ghcr.io/intel/llvm/sycl_prebuilt_tests:sycl-rel-6_2
             target_devices: level_zero:gpu
             extra_lit_opts: '--param test-preview-mode=False --filter-out "${{ needs.compat_read_exclude.outputs.FILTER_6_2 }}"'
             e2e_binaries_artifact: 'in-container'
-          - name: ABI compatibility against sycl-rel-6_3
+          - name: ABI compatibility / sycl-rel-6_3
             runner: '["Linux", "pvc"]'
             image: ghcr.io/intel/llvm/sycl_prebuilt_tests:sycl-rel-6_3
             target_devices: level_zero:gpu

.github/workflows/sycl-linux-run-tests.yml

@@ -235,7 +235,7 @@ jobs:
     - name: Reset Intel GPU
       uses: ./devops/actions/reset_gpu
     - name: Install drivers
-      if: inputs.install_igc_driver == 'true' || inputs.install_dev_igc_driver == 'true'
+      if: inputs.e2e_binaries_artifact != 'in-container' && (inputs.install_igc_driver == 'true' || inputs.install_dev_igc_driver == 'true')
       env:
         GITHUB_TOKEN: ${{ github.token }}
       run: |

.github/workflows/sycl-nightly.yml

@@ -201,11 +201,23 @@ jobs:
       always()
       && !cancelled()
       && needs.build-win.outputs.build_conclusion == 'success'
+    strategy:
+      fail-fast: false
+      matrix:
+        include:
+          - name: Intel L0 Gen12 GPU
+            runner: '["Windows", "gen12"]'
+            target_devices: level_zero:gpu
+
+          - name: Intel L0 Battlemage GPU
+            runner: '["Windows", "bmg"]'
+            target_devices: level_zero:gpu
+
     uses: ./.github/workflows/sycl-windows-run-tests.yml
     with:
-      name: Intel GEN12 Graphics with Level Zero
-      runner: '["Windows","gen12"]'
-      target_devices: level_zero:gpu
+      name: ${{ matrix.name }}
+      runner: ${{ matrix.runner }}
+      target_devices: ${{ matrix.target_devices }}
       toolchain_artifact_filename: ${{ needs.build-win.outputs.toolchain_artifact_filename }}
 
   cuda-aws-start:
@@ -320,20 +332,30 @@ jobs:
       toolchain_artifact_filename: ${{ needs.build-win.outputs.toolchain_artifact_filename }}
       sycl_cts_artifact: sycl_cts_bin_win
 
+  # Verification example:
+  # cosign-windows-amd64.exe verify-blob sycl_linux.tar.gz \
+  # --bundle sycl_linux.tar.gz.sigstore.json \
+  # --certificate-oidc-issuer https://token.actions.githubusercontent.com \
+  # --certificate-identity https://github.com/intel/llvm/.github/workflows/sycl-nightly.yml@refs/heads/sycl
   nightly_build_upload:
     name: Nightly Build Upload
     if: ${{ github.ref_name == 'sycl' }}
     needs: [ubuntu2204_build, build-win]
     runs-on: ubuntu-latest
     permissions:
       contents: write
+      id-token: write
     steps:
     - uses: actions/download-artifact@v4
       with:
         name: sycl_linux_default
     - uses: actions/download-artifact@v4
       with:
         name: sycl_windows_default
+    - name: Sign with sigstore/cosign
+      uses: sigstore/[email protected]
+      with:
+        inputs: sycl_linux.tar.gz sycl_windows.tar.gz
     - name: Compute tag
       id: tag
       run: |
@@ -349,6 +371,8 @@ jobs:
         files: |
           sycl_linux.tar.gz
           sycl_windows.tar.gz
+          sycl_linux.tar.gz.sigstore.json
+          sycl_windows.tar.gz.sigstore.json
         tag_name: nightly-${{ steps.tag.outputs.TAG }}
         name: DPC++ daily ${{ steps.tag.outputs.TAG }}
         prerelease: true

clang/include/clang/Sema/Sema.h

@@ -9184,7 +9184,13 @@ class Sema final : public SemaBase {
   };
 
   /// Compute the mangling number context for a lambda expression or
-  /// block literal. Also return the extra mangling decl if any.
+  /// block literal that appears in the specified declaration context in
+  /// consideration of the current expression evaluation and template
+  /// instantiation contexts. If the mangling context requires external linkage,
+  /// then a mangling number context is returned in the first tuple
+  /// element. If the mangling context is non-normal (specialized for
+  /// lambda and block types relative to other entities), the overriding
+  /// declaration is returned in the second tuple element.
   ///
   /// \param DC - The DeclContext containing the lambda expression or
   /// block literal.

clang/lib/AST/ASTContext.cpp

@@ -13483,6 +13483,7 @@ MangleNumberingContext &
 ASTContext::getManglingNumberContext(const DeclContext *DC) {
   assert(LangOpts.CPlusPlus);  // We don't need mangling numbers for plain C.
   std::unique_ptr<MangleNumberingContext> &MCtx = MangleNumberingContexts[DC];
+  DC = DC->getPrimaryContext();
   if (!MCtx)
     MCtx = createMangleNumberingContext();
   return *MCtx;

clang/lib/CodeGen/BackendUtil.cpp

@@ -1093,8 +1093,9 @@ void EmitAssemblyHelper::RunOptimizationPipeline(
              ThinOrFullLTOPhase) {
             MPM.addPass(createModuleToFunctionPassAdaptor(
                 InferAddressSpacesPass(clang::targets::SPIR_GENERIC_AS)));
-            MPM.addPass(
-                createModuleToFunctionPassAdaptor(SYCLOptimizeBarriersPass()));
+            if (Level != OptimizationLevel::O0)
+              MPM.addPass(createModuleToFunctionPassAdaptor(
+                  SYCLOptimizeBarriersPass()));
           });
     }
 

clang/lib/Driver/ToolChains/Clang.cpp

@@ -10066,8 +10066,13 @@ void OffloadWrapper::ConstructJob(Compilation &C, const JobAction &JA,
     // clang-offload-wrapper
     //   -o=<outputfile>.bc
     //   -host=x86_64-pc-linux-gnu -kind=sycl
+#ifndef __INTEL_PREVIEW_BREAKING_CHANGES
     //   -format=spirv <inputfile1>.spv <manifest1>(optional)
     //   -format=spirv <inputfile2>.spv <manifest2>(optional)
+#else
+    //   -format=spirv <inputfile1>.spv
+    //   -format=spirv <inputfile2>.spv
+#endif
     //  ...
     ArgStringList WrapperArgs;
 
@@ -10132,6 +10137,12 @@ void OffloadWrapper::ConstructJob(Compilation &C, const JobAction &JA,
     WrapperArgs.push_back(
         C.getArgs().MakeArgString(Twine("-kind=") + Twine(Kind)));
 
+    // Enable preview breaking changes in clang-offload-wrapper,
+    // in case it needs to introduce any ABI breaking changes.
+    // For example, changes in offload binary descriptor format.
+    if (C.getArgs().hasArg(options::OPT_fpreview_breaking_changes))
+      WrapperArgs.push_back("-fpreview-breaking-changes");
+
     assert((Inputs.size() > 0) && "no inputs for clang-offload-wrapper");
     assert(((Inputs[0].getType() != types::TY_Tempfiletable) ||
             (Inputs.size() == 1)) &&
@@ -11362,6 +11373,23 @@ void LinkerWrapper::ConstructJob(Compilation &C, const JobAction &JA,
       CmdArgs.push_back(
           Args.MakeArgString("--sycl-post-link-options=" + PostLinkOptString));
 
+    if (Args.hasArg(options::OPT_fsycl_remove_unused_external_funcs))
+      CmdArgs.push_back(
+          Args.MakeArgString("-sycl-remove-unused-external-funcs"));
+    if (Args.hasArg(options::OPT_fno_sycl_remove_unused_external_funcs))
+      CmdArgs.push_back(
+          Args.MakeArgString("-no-sycl-remove-unused-external-funcs"));
+    if (Args.hasArg(options::OPT_fsycl_device_code_split_esimd))
+      CmdArgs.push_back(Args.MakeArgString("-sycl-device-code-split-esimd"));
+    if (Args.hasArg(options::OPT_fno_sycl_device_code_split_esimd))
+      CmdArgs.push_back(Args.MakeArgString("-no-sycl-device-code-split-esimd"));
+    if (Args.hasArg(options::OPT_fsycl_add_default_spec_consts_image))
+      CmdArgs.push_back(
+          Args.MakeArgString("-sycl-add-default-spec-consts-image"));
+    if (Args.hasArg(options::OPT_fno_sycl_add_default_spec_consts_image))
+      CmdArgs.push_back(
+          Args.MakeArgString("-no-sycl-add-default-spec-consts-image"));
+
     // --llvm-spirv-options="options" provides a string of options to be passed
     // along to the llvm-spirv (translation) step during device link.
     SmallString<128> OptString;
@@ -11538,6 +11566,12 @@ void LinkerWrapper::ConstructJob(Compilation &C, const JobAction &JA,
           Args.MakeArgString("--wrapper-jobs=" + Twine(NumThreads)));
   }
 
+  // Enable preview breaking changes in clang-linker-wrapper,
+  // in case it needs to introduce any ABI breaking changes.
+  // For example, changes in offload binary descriptor format.
+  if (Args.hasArg(options::OPT_fpreview_breaking_changes))
+    CmdArgs.push_back("-fpreview-breaking-changes");
+
   const char *Exec =
       Args.MakeArgString(getToolChain().GetProgramPath("clang-linker-wrapper"));
 

clang/lib/Sema/SemaLambda.cpp

@@ -356,7 +356,14 @@ Sema::getCurrentMangleNumberContext(const DeclContext *DC) {
       return std::make_tuple(&Context.getManglingNumberContext(DC), nullptr);
     }
 
-    return std::make_tuple(nullptr, nullptr);
+    if (ManglingContextDecl) {
+      // Lambdas defined in the initializer of a local variable are mangled
+      // in the enclosing function context.
+      if (auto *VD = dyn_cast<VarDecl>(ManglingContextDecl);
+          VD && !VD->hasGlobalStorage())
+        ManglingContextDecl = nullptr;
+    }
+    return std::make_tuple(nullptr, ManglingContextDecl);
   }
 
   case NonInlineInModulePurview: