macOS: Apple Silicon support (interpreter only)

.github/workflows/build.yml

@@ -204,7 +204,7 @@ jobs:
         name: cemu-bin-windows-x64
         path: ./bin/Cemu.exe
 
-  build-macos:
+  build-macos-intel:
     runs-on: macos-12
     steps:
     - name: "Checkout repo"
@@ -239,8 +239,7 @@ jobs:
     - name: "Install system dependencies"
       run: |
         brew update
-        brew install llvm@15 ninja nasm automake libtool
-        brew install cmake python3 ninja
+        brew install llvm@15 ninja nasm automake libtool cmake python3 ninja
 
     - name: "Build and install molten-vk"
       run: |
@@ -251,11 +250,6 @@ jobs:
         make macos
         make install
 
-    - name: "Setup cmake"
-      uses: jwlawson/actions-setup-cmake@v2
-      with:
-        cmake-version: '3.29.0'
-
     - name: "Bootstrap vcpkg"
       run: |
         bash ./dependencies/vcpkg/bootstrap-vcpkg.sh
@@ -308,3 +302,101 @@ jobs:
       with:
         name: cemu-bin-macos-x64
         path: ./bin/Cemu.dmg
+
+  build-macos-arm64:
+    runs-on: macos-14
+    steps:
+      - name: "Checkout repo"
+        uses: actions/checkout@v4
+        with:
+          submodules: "recursive"
+
+      - name: "Fetch full history for vcpkg submodule"
+        run: |
+          cd dependencies/vcpkg
+          git fetch --unshallow
+
+      - name: Setup release mode parameters (for deploy)
+        if: ${{ inputs.deploymode == 'release' }}
+        run: |
+          echo "BUILD_MODE=release" >> $GITHUB_ENV
+          echo "BUILD_FLAGS=" >> $GITHUB_ENV
+          echo "Build mode is release"
+      - name: Setup debug mode parameters (for continous build)
+        if: ${{ inputs.deploymode != 'release' }}
+        run: |
+          echo "BUILD_MODE=debug" >> $GITHUB_ENV
+          echo "BUILD_FLAGS=" >> $GITHUB_ENV
+          echo "Build mode is debug"
+
+      - name: Setup version for experimental
+        if: ${{ inputs.experimentalversion != '' }}
+        run: |
+          echo "[INFO] Experimental version ${{ inputs.experimentalversion }}"
+          echo "BUILD_FLAGS=${{ env.BUILD_FLAGS }} -DEXPERIMENTAL_VERSION=${{ inputs.experimentalversion }}" >> $GITHUB_ENV
+
+      - name: "Install system dependencies"
+        run: |
+          brew update
+          brew install ninja automake libtool cmake ninja python3
+
+      - name: "Build and install molten-vk"
+        run: |
+          git clone https://github.com/KhronosGroup/MoltenVK.git
+          cd MoltenVK
+          git checkout bf097edc74ec3b6dfafdcd5a38d3ce14b11952d6
+          ./fetchDependencies --macos
+          make macos
+          sudo make install
+          cp /usr/local/lib/libMoltenVK.dylib /opt/homebrew/lib
+
+      - name: "Bootstrap vcpkg"
+        run: |
+          bash ./dependencies/vcpkg/bootstrap-vcpkg.sh
+
+      - name: 'Setup NuGet Credentials for vcpkg'
+        shell: 'bash'
+        run: |
+          mono `./dependencies/vcpkg/vcpkg fetch nuget | tail -n 1` \
+          sources add \
+          -source "https://nuget.pkg.github.com/${{ github.repository_owner }}/index.json" \
+          -storepasswordincleartext \
+          -name "GitHub" \
+          -username "${{ github.repository_owner }}" \
+          -password "${{ secrets.GITHUB_TOKEN }}"
+          mono `./dependencies/vcpkg/vcpkg fetch nuget | tail -n 1` \
+          setapikey "${{ secrets.GITHUB_TOKEN }}" \
+          -source "https://nuget.pkg.github.com/${{ github.repository_owner }}/index.json"
+
+      - name: "cmake"
+        run: |
+          mkdir build
+          cd build
+          cmake .. ${{ env.BUILD_FLAGS }} \
+          -DCMAKE_BUILD_TYPE=${{ env.BUILD_MODE }} \
+          -DMACOS_BUNDLE=ON \
+          -G Ninja
+
+      - name: "Build Cemu"
+        run: |
+          cmake --build build
+
+      - name: Prepare artifact
+        if: ${{ inputs.deploymode == 'release' }}
+        run: |
+          mkdir bin/Cemu_app
+          mv bin/Cemu_release.app bin/Cemu_app/Cemu.app
+          mv bin/Cemu_app/Cemu.app/Contents/MacOS/Cemu_release bin/Cemu_app/Cemu.app/Contents/MacOS/Cemu
+          sed -i '' 's/Cemu_release/Cemu/g' bin/Cemu_app/Cemu.app/Contents/Info.plist
+          chmod a+x bin/Cemu_app/Cemu.app/Contents/MacOS/{Cemu,update.sh}
+          ln -s /Applications bin/Cemu_app/Applications
+          hdiutil create ./bin/tmp.dmg -ov -volname "Cemu" -fs HFS+ -srcfolder "./bin/Cemu_app"
+          hdiutil convert ./bin/tmp.dmg -format UDZO -o bin/Cemu.dmg
+          rm bin/tmp.dmg
+
+      - name: Upload artifact
+        uses: actions/upload-artifact@v4
+        if: ${{ inputs.deploymode == 'release' }}
+        with:
+          name: cemu-bin-macos-arm64
+          path: ./bin/Cemu.dmg

BUILD.md

@@ -16,7 +16,6 @@
          - [Compiling Errors](#compiling-errors)
          - [Building Errors](#building-errors)
 - [macOS](#macos)
-   - [On Apple Silicon Macs, Rosetta 2 and the x86_64 version of Homebrew must be used](#on-apple-silicon-macs-rosetta-2-and-the-x86_64-version-of-homebrew-must-be-used)
    - [Installing brew](#installing-brew)
    - [Installing Dependencies](#installing-dependencies)
    - [Build Cemu using CMake and Clang](#build-cemu-using-cmake-and-clang)
@@ -142,21 +141,15 @@ If you are getting a different error than any of the errors listed above, you ma
 ## macOS
 
 To compile Cemu, a recent enough compiler and STL with C++20 support is required! LLVM 13 and
-below, built in LLVM, and Xcode LLVM don't support the C++20 feature set required. The OpenGL graphics
-API isn't support on macOS, Vulkan must be used. Additionally Vulkan must be used through the
-Molten-VK compatibility layer
+lower don't support the C++20 feature set required. Cemu doesn't support using OpenGL
+on macOS, Vulkan must be used via the Molten-Vk compatibility layer.
 
-### On Apple Silicon Macs, Rosetta 2 and the x86_64 version of Homebrew must be used
-
-You can skip this section if you have an Intel Mac. Every time you compile, you need to perform steps 2.
-
-1. `softwareupdate --install-rosetta` # Install Rosetta 2 if you don't have it. This only has to be done once
-2. `arch -x86_64 zsh` # run an x64 shell
+Note that if you are compiling on/for Intel, you must replace all references of `/opt/homebrew` to `/usr/local`.
 
 ### Installing brew
 
 1. `/bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/HEAD/install.sh)"`
-2. `eval "$(/usr/local/Homebrew/bin/brew shellenv)"` # set x86_64 brew env
+2. `eval "$(/opt/homebrew/bin/brew shellenv)"` # set brew env
 
 ### Installing Dependencies
 
@@ -165,12 +158,12 @@ You can skip this section if you have an Intel Mac. Every time you compile, you
 ### Build Cemu using CMake and Clang
 1. `git clone --recursive https://github.com/cemu-project/Cemu`
 2. `cd Cemu`
-3. `cmake -S . -B build -DCMAKE_BUILD_TYPE=release -DCMAKE_C_COMPILER=/usr/local/opt/llvm/bin/clang -DCMAKE_CXX_COMPILER=/usr/local/opt/llvm/bin/clang++ -G Ninja`
+3. `cmake -S . -B build -DCMAKE_BUILD_TYPE=release -DCMAKE_C_COMPILER=/opt/homebrew/opt/llvm/bin/clang -DCMAKE_CXX_COMPILER=/opt/homebrew/opt/llvm/bin/clang++ -G Ninja`
 4. `cmake --build build`
 5. You should now have a Cemu executable file in the /bin folder, which you can run using `./bin/Cemu_release`.
 
 #### Troubleshooting steps
-- If step 3 gives you an error about not being able to find ninja, try appending `-DCMAKE_MAKE_PROGRAM=/usr/local/bin/ninja` to the command and running it again.
+- If step 3 gives you an error about not being able to find ninja, try appending `-DCMAKE_MAKE_PROGRAM=/opt/homebrew/bin/ninja` to the command and running it again.
 
 ## Updating Cemu and source code
 1. To update your Cemu local repository, use the command `git pull --recurse-submodules` (run this command on the Cemu root).

CMakeLists.txt

@@ -1,7 +1,8 @@
 cmake_minimum_required(VERSION 3.21.1)
 
+include(CMakeDependentOption)
 option(ENABLE_VCPKG "Enable the vcpkg package manager" ON)
-option(MACOS_BUNDLE "The executable when built on macOS will be created as an application bundle" OFF)
+cmake_dependent_option(MACOS_BUNDLE "The executable when built on macOS will be created as an application bundle" ON "APPLE" OFF)
 set(EXPERIMENTAL_VERSION "" CACHE STRING "") # used by CI script to set experimental version
 
 if (EXPERIMENTAL_VERSION)

dependencies/ih264d/CMakeLists.txt

@@ -140,7 +140,7 @@ target_sources(ih264d PRIVATE
 "decoder/x86/ih264d_function_selector_sse42.c"
 "decoder/x86/ih264d_function_selector_ssse3.c"
 )
-elseif(CMAKE_SYSTEM_PROCESSOR STREQUAL "aarch64")
+elseif(CMAKE_SYSTEM_PROCESSOR STREQUAL "aarch64" OR CMAKE_SYSTEM_PROCESSOR STREQUAL "arm64")
 enable_language( C CXX ASM )
 set(LIBAVCDEC_ARM_INCLUDES "common/armv8" "decoder/arm")
 include_directories("common/" "decoder/" ${LIBAVCDEC_ARM_INCLUDES})

dependencies/ih264d/common/armv8/ih264_intra_pred_chroma_av8.s

@@ -53,10 +53,13 @@
 .text
 .p2align 2
 .include "ih264_neon_macros.s"
-
+#ifdef __APPLE__
+.extern _ih264_gai1_intrapred_chroma_plane_coeffs1
+.extern _ih264_gai1_intrapred_chroma_plane_coeffs2
+#else
 .extern ih264_gai1_intrapred_chroma_plane_coeffs1
 .extern ih264_gai1_intrapred_chroma_plane_coeffs2
-
+#endif
 
 
 ///**
@@ -429,8 +432,13 @@ ih264_intra_pred_chroma_8x8_mode_plane_av8:
     rev64     v7.4h, v2.4h
     ld1       {v3.2s}, [x10]
     sub       x5, x3, #8
+#ifdef __APPLE__
+    adrp      x12, _ih264_gai1_intrapred_chroma_plane_coeffs1@GOTPAGE
+    ldr       x12, [x12, _ih264_gai1_intrapred_chroma_plane_coeffs1@GOTPAGEOFF]
+#else
     adrp      x12, :got:ih264_gai1_intrapred_chroma_plane_coeffs1
     ldr       x12, [x12, #:got_lo12:ih264_gai1_intrapred_chroma_plane_coeffs1]
+#endif
     usubl     v10.8h, v5.8b, v1.8b
     ld1       {v8.8b, v9.8b}, [x12]     // Load multiplication factors 1 to 8 into D3
     mov       v8.d[1], v9.d[0]
@@ -484,10 +492,13 @@ ih264_intra_pred_chroma_8x8_mode_plane_av8:
     zip1      v1.8h, v0.8h, v2.8h
     zip2      v2.8h, v0.8h, v2.8h
     mov       v0.16b, v1.16b
-
+#ifdef __APPLE__
+    adrp      x12, _ih264_gai1_intrapred_chroma_plane_coeffs2@GOTPAGE
+    ldr       x12, [x12, _ih264_gai1_intrapred_chroma_plane_coeffs2@GOTPAGEOFF]
+#else
     adrp      x12, :got:ih264_gai1_intrapred_chroma_plane_coeffs2
     ldr       x12, [x12, #:got_lo12:ih264_gai1_intrapred_chroma_plane_coeffs2]
-
+#endif
     ld1       {v8.2s, v9.2s}, [x12]
     mov       v8.d[1], v9.d[0]
     mov       v10.16b, v8.16b

dependencies/ih264d/common/armv8/ih264_intra_pred_luma_16x16_av8.s

@@ -53,7 +53,11 @@
 .text
 .p2align 2
 .include "ih264_neon_macros.s"
+#ifdef __APPLE__
+.extern _ih264_gai1_intrapred_luma_plane_coeffs
+#else
 .extern ih264_gai1_intrapred_luma_plane_coeffs
+#endif
 
 
 
@@ -431,10 +435,13 @@ ih264_intra_pred_luma_16x16_mode_plane_av8:
     mov       x10, x1                   //top_left
     mov       x4, #-1
     ld1       {v2.2s}, [x1], x8
-
+#ifdef __APPLE__
+    adrp      x7, _ih264_gai1_intrapred_luma_plane_coeffs@GOTPAGE
+    ldr       x7, [x7, _ih264_gai1_intrapred_luma_plane_coeffs@GOTPAGEOFF]
+#else
     adrp      x7, :got:ih264_gai1_intrapred_luma_plane_coeffs
     ldr       x7, [x7, #:got_lo12:ih264_gai1_intrapred_luma_plane_coeffs]
-
+#endif
     ld1       {v0.2s}, [x1]
     rev64     v2.8b, v2.8b
     ld1       {v6.2s, v7.2s}, [x7]

dependencies/ih264d/common/armv8/ih264_intra_pred_luma_8x8_av8.s

@@ -56,8 +56,11 @@
 .text
 .p2align 2
 .include "ih264_neon_macros.s"
-
+#ifdef __APPLE__
+.extern _ih264_gai1_intrapred_luma_8x8_horz_u
+#else
 .extern ih264_gai1_intrapred_luma_8x8_horz_u
+#endif
 
 
 
@@ -1029,9 +1032,13 @@ ih264_intra_pred_luma_8x8_mode_horz_u_av8:
     mov       v3.d[0], v2.d[1]
     ext       v4.16b, v2.16b , v2.16b , #1
     mov       v5.d[0], v4.d[1]
-
+#ifdef __APPLE__
+    adrp      x12, _ih264_gai1_intrapred_luma_8x8_horz_u@GOTPAGE
+    ldr       x12, [x12, _ih264_gai1_intrapred_luma_8x8_horz_u@GOTPAGEOFF]
+#else
     adrp      x12, :got:ih264_gai1_intrapred_luma_8x8_horz_u
     ldr       x12, [x12, #:got_lo12:ih264_gai1_intrapred_luma_8x8_horz_u]
+#endif
     uaddl     v20.8h, v0.8b, v2.8b
     uaddl     v22.8h, v1.8b, v3.8b
     uaddl     v24.8h, v2.8b, v4.8b

dependencies/ih264d/common/armv8/ih264_weighted_bi_pred_av8.s

@@ -142,14 +142,22 @@ ih264_weighted_bi_pred_luma_av8:
     sxtw      x4, w4
     sxtw      x5, w5
     stp       x19, x20, [sp, #-16]!
+#ifndef __APPLE__
     ldr       w8, [sp, #80]             //Load wt2 in w8
     ldr       w9, [sp, #88]             //Load ofst1 in w9
-    add       w6, w6, #1                //w6  = log_WD + 1
-    neg       w10, w6                   //w10 = -(log_WD + 1)
-    dup       v0.8h, w10                //Q0  = -(log_WD + 1) (32-bit)
     ldr       w10, [sp, #96]            //Load ofst2 in w10
     ldr       w11, [sp, #104]           //Load ht in w11
     ldr       w12, [sp, #112]           //Load wd in w12
+#else
+    ldr       w8, [sp, #80]             //Load wd in w8
+    ldr       w9, [sp, #84]             //Load ht in w9
+    ldr       w10, [sp, #88]            //Load offst2 in w10
+    ldr       w11,  [sp, #92]           //Load offst1 in w11
+    ldr       w12,  [sp, #96]           //Load offst1 in w12
+#endif
+    add       w6, w6, #1                //w6  = log_WD + 1
+    neg       w10, w6                   //w10 = -(log_WD + 1)
+    dup       v0.8h, w10                //Q0  = -(log_WD + 1) (32-bit)
     add       w9, w9, #1                //w9 = ofst1 + 1
     add       w9, w9, w10               //w9 = ofst1 + ofst2 + 1
     mov       v2.s[0], w7
@@ -424,17 +432,24 @@ ih264_weighted_bi_pred_chroma_av8:
     sxtw      x5, w5
     stp       x19, x20, [sp, #-16]!
 
-
+#ifndef __APPLE__
     ldr       w8, [sp, #80]             //Load wt2 in w8
+    ldr       w9, [sp, #88]             //Load ofst1 in w9
+    ldr       w10, [sp, #96]            //Load ofst2 in w10
+    ldr       w11, [sp, #104]           //Load ht in w11
+    ldr       w12, [sp, #112]           //Load wd in w12
+#else
+    ldr       w8, [sp, #80]             //Load wd in w8
+    ldr       w9, [sp, #84]             //Load ht in w9
+    ldr       w10, [sp, #88]            //Load offst2 in w10
+    ldr       w11,  [sp, #92]           //Load offst1 in w11
+    ldr       w12,  [sp, #96]           //Load offst1 in w12
+#endif
     dup       v4.4s, w8                 //Q2 = (wt2_u, wt2_v) (32-bit)
     dup       v2.4s, w7                 //Q1 = (wt1_u, wt1_v) (32-bit)
     add       w6, w6, #1                //w6  = log_WD + 1
-    ldr       w9, [sp, #88]             //Load ofst1 in w9
-    ldr       w10, [sp, #96]            //Load ofst2 in w10
     neg       w20, w6                   //w20 = -(log_WD + 1)
     dup       v0.8h, w20                //Q0  = -(log_WD + 1) (16-bit)
-    ldr       w11, [sp, #104]           //Load ht in x11
-    ldr       w12, [sp, #112]           //Load wd in x12
     dup       v20.8h, w9                //0ffset1
     dup       v21.8h, w10               //0ffset2
     srhadd    v6.8b, v20.8b, v21.8b