merge update

2025-12-21 22:33:56 +00:00 · 2025-10-18 23:03:13 +08:00
parent 04431b50fa bc1a818fdc
commit cb13784a11
587 changed files with 74804 additions and 27637 deletions
--- a/.github/workflows/release.yaml
+++ b/.github/workflows/release.yaml
@@ -94,7 +94,7 @@ jobs:
            install: https://download.amd.com/developer/eula/rocm-hub/AMD-Software-PRO-Edition-24.Q4-WinSvr2022-For-HIP.exe
            rocm-version: '6.2'
            flags: '-DCMAKE_C_COMPILER=clang -DCMAKE_CXX_COMPILER=clang++ -DCMAKE_C_FLAGS="-parallel-jobs=4 -Wno-ignored-attributes -Wno-deprecated-pragma" -DCMAKE_CXX_FLAGS="-parallel-jobs=4 -Wno-ignored-attributes -Wno-deprecated-pragma"'
-            runner_dir: ''
+            runner_dir: 'rocm'
    runs-on: ${{ matrix.arch == 'arm64' && format('{0}-{1}', matrix.os, matrix.arch) || matrix.os }}
    environment: release
    env:
@@ -163,6 +163,7 @@ jobs:
          cmake --preset "${{ matrix.preset }}" ${{ matrix.flags }} -DOLLAMA_RUNNER_DIR="${{ matrix.runner_dir }}"
          cmake --build --parallel --preset "${{ matrix.preset }}"
          cmake --install build --component "${{ startsWith(matrix.preset, 'CUDA ') && 'CUDA' || startsWith(matrix.preset, 'ROCm ') && 'HIP' || 'CPU' }}" --strip --parallel 8
+          Remove-Item -Path dist\lib\ollama\rocm\rocblas\library\*gfx906* -ErrorAction SilentlyContinue
        env:
          CMAKE_GENERATOR: Ninja
      - uses: actions/upload-artifact@v4
@@ -175,19 +176,19 @@ jobs:
      matrix:
        os: [windows]
        arch: [amd64, arm64]
+        include:
+        - os: windows
+          arch: amd64
+          llvmarch: x86_64
+        - os: windows
+          arch: arm64
+          llvmarch: aarch64
    runs-on: ${{ matrix.arch == 'arm64' && format('{0}-{1}', matrix.os, matrix.arch) || matrix.os }}
    environment: release
    needs: [setup-environment]
    env:
      GOFLAGS: ${{ needs.setup-environment.outputs.GOFLAGS }}
    steps:
-      - name: Install AMD64 system dependencies
-        if: matrix.arch == 'amd64'
-        run: |
-          $ErrorActionPreference = "Stop"
-          Start-Process "C:\msys64\usr\bin\pacman.exe" -ArgumentList @("-S", "--noconfirm", "mingw-w64-clang-x86_64-gcc-compat", "mingw-w64-clang-x86_64-clang") -NoNewWindow -Wait
-          echo "C:\msys64\usr\bin" | Out-File -FilePath $env:GITHUB_PATH -Encoding utf8 -Append
-          echo "C:\msys64\clang64\bin" | Out-File -FilePath $env:GITHUB_PATH -Encoding utf8 -Append
      - name: Install ARM64 system dependencies
        if: matrix.arch == 'arm64'
        run: |
@@ -199,15 +200,29 @@ jobs:

          choco install -y --no-progress git gzip
          echo "C:\Program Files\Git\cmd" | Out-File -FilePath $env:GITHUB_PATH -Encoding utf8 -Append
-
-          Invoke-WebRequest -Uri "https://github.com/mstorsjo/llvm-mingw/releases/download/20240619/llvm-mingw-20240619-ucrt-aarch64.zip" -OutFile "${{ runner.temp }}\llvm-mingw-ucrt-aarch64.zip"
-          Expand-Archive -Path ${{ runner.temp }}\llvm-mingw-ucrt-aarch64.zip -DestinationPath "C:\Program Files\"
-          $installPath=(Resolve-Path -Path "C:\Program Files\llvm-mingw-*-ucrt-aarch64").path
-          echo $installPath\bin | Out-File -FilePath $env:GITHUB_PATH -Encoding utf8 -Append
+      - name: Install clang and gcc-compat
+        run: |
+          $ErrorActionPreference = "Stop"
+          Set-ExecutionPolicy Bypass -Scope Process -Force
+          Invoke-WebRequest -Uri "https://github.com/mstorsjo/llvm-mingw/releases/download/20240619/llvm-mingw-20240619-ucrt-${{ matrix.llvmarch }}.zip" -OutFile "${{ runner.temp }}\llvm-mingw-ucrt.zip"
+          Expand-Archive -Path ${{ runner.temp }}\llvm-mingw-ucrt.zip -DestinationPath "C:\Program Files\"
+          $installPath=(Resolve-Path -Path "C:\Program Files\llvm-mingw-*-ucrt*").path
+          echo "$installPath\bin" | Out-File -FilePath $env:GITHUB_PATH -Encoding utf8 -Append
      - uses: actions/checkout@v4
      - uses: actions/setup-go@v5
        with:
          go-version-file: go.mod
+      - name: Verify gcc is actually clang
+        run: |
+          $ErrorActionPreference='Continue'
+          $version=& gcc -v 2>&1
+          $version=$version -join "`n"
+          echo "gcc is $version"
+          if ($version -notmatch 'clang') {
+            echo "ERROR: GCC must be clang for proper utf16 handling"
+            exit 1
+          }
+          $ErrorActionPreference='Stop'
      - run: |
          go build -o dist/${{ matrix.os }}-${{ matrix.arch }}/ .
      - uses: actions/upload-artifact@v4
@@ -222,13 +237,13 @@ jobs:
        include:
          - os: linux
            arch: amd64
-            target: archive
+            target: archive_novulkan
          - os: linux
            arch: amd64
            target: rocm
          - os: linux
            arch: arm64
-            target: archive
+            target: archive_novulkan
    runs-on: ${{ matrix.arch == 'arm64' && format('{0}-{1}', matrix.os, matrix.arch) || matrix.os }}
    environment: release
    needs: setup-environment
@@ -284,12 +299,14 @@ jobs:
        include:
          - os: linux
            arch: arm64
+            target: novulkan
            build-args: |
              CGO_CFLAGS
              CGO_CXXFLAGS
              GOFLAGS
          - os: linux
            arch: amd64
+            target: novulkan
            build-args: |
              CGO_CFLAGS
              CGO_CXXFLAGS
@@ -302,6 +319,14 @@ jobs:
              CGO_CXXFLAGS
              GOFLAGS
              FLAVOR=rocm
+          - os: linux
+            arch: amd64
+            suffix: '-vulkan'
+            target: default
+            build-args: |
+              CGO_CFLAGS
+              CGO_CXXFLAGS
+              GOFLAGS
    runs-on: ${{ matrix.arch == 'arm64' && format('{0}-{1}', matrix.os, matrix.arch) || matrix.os }}
    environment: release
    needs: setup-environment
@@ -319,6 +344,7 @@ jobs:
        with:
          context: .
          platforms: ${{ matrix.os }}/${{ matrix.arch }}
+          target: ${{ matrix.target }}
          build-args: ${{ matrix.build-args }}
          outputs: type=image,name=${{ vars.DOCKER_REPO }},push-by-digest=true,name-canonical=true,push=true
          cache-from: type=registry,ref=${{ vars.DOCKER_REPO }}:latest
--- a/.github/workflows/test.yaml
+++ b/.github/workflows/test.yaml
@@ -52,6 +52,12 @@ jobs:
            container: rocm/dev-ubuntu-22.04:6.1.2
            extra-packages: rocm-libs
            flags: '-DAMDGPU_TARGETS=gfx1010 -DCMAKE_PREFIX_PATH=/opt/rocm'
+          - preset: Vulkan
+            container: ubuntu:22.04
+            extra-packages: >
+              mesa-vulkan-drivers vulkan-tools
+              libvulkan1 libvulkan-dev
+              vulkan-sdk cmake ccache g++ make
    runs-on: linux
    container: ${{ matrix.container }}
    steps:
@@ -59,7 +65,19 @@ jobs:
      - run: |
          [ -n "${{ matrix.container }}" ] || sudo=sudo
          $sudo apt-get update
+          # Add LunarG Vulkan SDK apt repo for Ubuntu 22.04
+          if [ "${{ matrix.preset }}" = "Vulkan" ]; then
+            $sudo apt-get install -y --no-install-recommends wget gnupg ca-certificates software-properties-common
+            wget -qO - https://packages.lunarg.com/lunarg-signing-key-pub.asc | $sudo gpg --dearmor -o /usr/share/keyrings/lunarg-archive-keyring.gpg
+            # Use signed-by to bind the repo to the installed keyring to avoid NO_PUBKEY
+            echo "deb [signed-by=/usr/share/keyrings/lunarg-archive-keyring.gpg]  https://packages.lunarg.com/vulkan/1.4.313 jammy main" | $sudo tee /etc/apt/sources.list.d/lunarg-vulkan-1.4.313-jammy.list > /dev/null
+            $sudo apt-get update
+          fi
          $sudo apt-get install -y cmake ccache ${{ matrix.extra-packages }}
+          # Export VULKAN_SDK if provided by LunarG package (defensive)
+          if [ -d "/usr/lib/x86_64-linux-gnu/vulkan" ] && [ "${{ matrix.preset }}" = "Vulkan" ]; then
+            echo "VULKAN_SDK=/usr" >> $GITHUB_ENV
+          fi
        env:
          DEBIAN_FRONTEND: noninteractive
      - uses: actions/cache@v4
@@ -92,18 +110,21 @@ jobs:
          - preset: ROCm
            install: https://download.amd.com/developer/eula/rocm-hub/AMD-Software-PRO-Edition-24.Q4-WinSvr2022-For-HIP.exe
            flags: '-DAMDGPU_TARGETS=gfx1010 -DCMAKE_C_COMPILER=clang -DCMAKE_CXX_COMPILER=clang++ -DCMAKE_C_FLAGS="-parallel-jobs=4 -Wno-ignored-attributes -Wno-deprecated-pragma" -DCMAKE_CXX_FLAGS="-parallel-jobs=4 -Wno-ignored-attributes -Wno-deprecated-pragma"'
+          - preset: Vulkan
+            install: https://sdk.lunarg.com/sdk/download/1.4.321.1/windows/vulkansdk-windows-X64-1.4.321.1.exe
    runs-on: windows
    steps:
      - run: |
          choco install -y --no-progress ccache ninja
          ccache -o cache_dir=${{ github.workspace }}\.ccache
-      - if: matrix.preset == 'CUDA' || matrix.preset == 'ROCm'
+      - if: matrix.preset == 'CUDA' || matrix.preset == 'ROCm' || matrix.preset == 'Vulkan'
        id: cache-install
        uses: actions/cache/restore@v4
        with:
          path: |
            C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA
            C:\Program Files\AMD\ROCm
+            C:\VulkanSDK
          key: ${{ matrix.install }}
      - if: matrix.preset == 'CUDA'
        name: Install CUDA ${{ matrix.cuda-version }}
@@ -133,6 +154,18 @@ jobs:
          echo "HIPCXX=$hipPath\bin\clang++.exe" | Out-File -FilePath $env:GITHUB_ENV -Append
          echo "HIP_PLATFORM=amd" | Out-File -FilePath $env:GITHUB_ENV -Append
          echo "CMAKE_PREFIX_PATH=$hipPath" | Out-File -FilePath $env:GITHUB_ENV -Append
+      - if: matrix.preset == 'Vulkan'
+        name: Install Vulkan ${{ matrix.rocm-version }}
+        run: |
+          $ErrorActionPreference = "Stop"
+          if ("${{ steps.cache-install.outputs.cache-hit }}" -ne 'true') {
+            Invoke-WebRequest -Uri "${{ matrix.install }}" -OutFile "install.exe"
+            Start-Process -FilePath .\install.exe -ArgumentList "-c","--am","--al","in" -NoNewWindow -Wait
+          }
+          
+          $vulkanPath = (Resolve-Path "C:\VulkanSDK\*").path
+          echo "$vulkanPath\bin" | Out-File -FilePath $env:GITHUB_PATH -Encoding utf8 -Append
+          echo "VULKAN_SDK=$vulkanPath" >> $env:GITHUB_ENV
      - if: ${{ !cancelled() && steps.cache-install.outputs.cache-hit != 'true' }}
        uses: actions/cache/save@v4
        with: