Merge branch 'sycl' into llvmspirv_pulldown

Author: fzou1

.github/CODEOWNERS

@@ -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

@@ -100,14 +100,14 @@ 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 +119,39 @@ 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-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/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

@@ -11362,6 +11362,23 @@ 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;

clang/test/CodeGenSYCL/kernel-early-optimization-pipeline.cpp

@@ -12,6 +12,7 @@
 // CHECK: InferFunctionAttrsPass
 // CHECK: AlwaysInlinerPass
 // CHECK: ModuleInlinerWrapperPass
+// CHECK: SYCLOptimizeBarriersPass
 // CHECK: ConstantMergePass
 // CHECK: SYCLMutatePrintfAddrspacePass
 // CHECK: SYCLPropagateAspectsUsagePass
@@ -22,4 +23,13 @@
 //
 // RUN: %clang_cc1 -O2 -fsycl-is-device -triple spir64-unknown-unknown %s -mdebug-pass Structure -emit-llvm -fno-sycl-early-optimizations -o /dev/null 2>&1 | FileCheck %s --check-prefix=CHECK-NOEARLYOPT
 // CHECK-NOEARLYOPT-NOT: ConstantMergePass1
+// CHECK-NOEARLYOPT-NOT: SYCLOptimizeBarriersPass
 // CHECK-NOEARLYOPT: SYCLMutatePrintfAddrspacePass
+
+// RUN: %clang_cc1 -O0 -fsycl-is-device -triple spir64-unknown-unknown %s -mdebug-pass Structure -emit-llvm -o /dev/null 2>&1 | FileCheck %s --check-prefix=CHECK-O0
+// CHECK-O0-NOT: SYCLOptimizeBarriersPass
+
+// Passes registered via registerOptimizerLastEPCallback don't run on empty
+// code
+__attribute__((sycl_device)) void bar() {
+}

clang/test/Driver/sycl-add-default-spec-consts-image-old-model.cpp

@@ -13,7 +13,7 @@
 // RUN: %clang -### -fsycl --no-offload-new-driver -fsycl-add-default-spec-consts-image -fsycl-targets=spir64_gen 2>&1  %s | FileCheck %s -check-prefix=CHECK-AOT
 // RUN: %clang -### -fsycl --no-offload-new-driver -fsycl-add-default-spec-consts-image -fsycl-targets=spir64_x86_64 2>&1 %s | FileCheck %s -check-prefix=CHECK-AOT
 // RUN: %clang -### -fsycl --no-offload-new-driver -fsycl-add-default-spec-consts-image -fsycl-targets=intel_gpu_pvc 2>&1 %s | FileCheck %s -check-prefix=CHECK-AOT
-// RUN: %clang -### -fsycl --no-offload-new-driver -fsycl-add-default-spec-consts-image -fsycl-targets=nvidia_gpu_sm_90 -nocudalib 2>&1 %s | FileCheck %s -check-prefix=CHECK-AOT
+// RUN: %clang -### -fsycl --no-offload-new-driver -fsycl-add-default-spec-consts-image -fsycl-targets=nvidia_gpu_sm_90 -fno-sycl-libspirv -nocudalib 2>&1 %s | FileCheck %s -check-prefix=CHECK-AOT
 // RUN: %clang -### -fsycl --no-offload-new-driver -fsycl-add-default-spec-consts-image -fsycl-targets=amd_gpu_gfx1034 -fno-sycl-libspirv -nogpulib  2>&1 %s | FileCheck %s -check-prefix=CHECK-AOT
 // CHECK-AOT-NOT: warning: -fsycl-add-default-spec-consts-image flag has an effect only in Ahead of Time Compilation mode (AOT)
 // CHECK-AOT: {{.*}}sycl-post-link{{.*}} "-generate-device-image-default-spec-consts"