hle: kernel: Migrate PageHeap/PageTable to KPageHeap/KPageTable.

externals: Update dynarmic to latest

.ci/scripts/common/post-upload.sh

@@ -15,5 +15,5 @@ mv "${REV_NAME}-source.tar.xz" $RELEASE_NAME
 7z a "$REV_NAME.7z" $RELEASE_NAME
 
 # move the compiled archive into the artifacts directory to be uploaded by travis releases
-mv "$ARCHIVE_NAME" artifacts/
-mv "$REV_NAME.7z" artifacts/
+mv "$ARCHIVE_NAME" "${ARTIFACTS_DIR}/"
+mv "$REV_NAME.7z" "${ARTIFACTS_DIR}/"

.ci/scripts/common/pre-upload.sh

@@ -2,5 +2,6 @@
 
 GITDATE="`git show -s --date=short --format='%ad' | sed 's/-//g'`"
 GITREV="`git show -s --format='%h'`"
+ARTIFACTS_DIR="artifacts"
 
-mkdir -p artifacts
+mkdir -p "${ARTIFACTS_DIR}/"

.ci/scripts/linux/docker.sh

@@ -1,14 +1,49 @@
 #!/bin/bash -ex
 
+# Exit on error, rather than continuing with the rest of the script.
+set -e
+
 cd /yuzu
 
 ccache -s
 
 mkdir build || true && cd build
-cmake .. -G Ninja -DDISPLAY_VERSION=$1 -DYUZU_USE_BUNDLED_UNICORN=ON -DYUZU_USE_QT_WEB_ENGINE=ON -DCMAKE_BUILD_TYPE=Release -DCMAKE_C_COMPILER=/usr/lib/ccache/gcc -DCMAKE_CXX_COMPILER=/usr/lib/ccache/g++ -DYUZU_ENABLE_COMPATIBILITY_REPORTING=${ENABLE_COMPATIBILITY_REPORTING:-"OFF"} -DENABLE_COMPATIBILITY_LIST_DOWNLOAD=ON -DUSE_DISCORD_PRESENCE=ON -DENABLE_QT_TRANSLATION=ON
+cmake .. -DDISPLAY_VERSION=$1 -DCMAKE_BUILD_TYPE=Release -DCMAKE_C_COMPILER=/usr/lib/ccache/gcc -DCMAKE_CXX_COMPILER=/usr/lib/ccache/g++ -DYUZU_ENABLE_COMPATIBILITY_REPORTING=${ENABLE_COMPATIBILITY_REPORTING:-"OFF"} -DENABLE_COMPATIBILITY_LIST_DOWNLOAD=ON -DUSE_DISCORD_PRESENCE=ON -DENABLE_QT_TRANSLATION=ON -DCMAKE_INSTALL_PREFIX="/usr"
 
-ninja
+make -j$(nproc)
 
 ccache -s
 
 ctest -VV -C Release
+
+make install DESTDIR=AppDir
+rm -vf AppDir/usr/bin/yuzu-cmd AppDir/usr/bin/yuzu-tester
+
+# Download tools needed to build an AppImage
+wget -nc https://github.com/linuxdeploy/linuxdeploy/releases/download/continuous/linuxdeploy-x86_64.AppImage
+wget -nc https://github.com/linuxdeploy/linuxdeploy-plugin-qt/releases/download/continuous/linuxdeploy-plugin-qt-x86_64.AppImage
+wget -nc https://github.com/darealshinji/AppImageKit-checkrt/releases/download/continuous/AppRun-patched-x86_64
+wget -nc https://github.com/darealshinji/AppImageKit-checkrt/releases/download/continuous/exec-x86_64.so
+# Set executable bit
+chmod 755 \
+    AppRun-patched-x86_64 \
+    exec-x86_64.so \
+    linuxdeploy-x86_64.AppImage \
+    linuxdeploy-plugin-qt-x86_64.AppImage
+
+# Workaround for https://github.com/AppImage/AppImageKit/issues/828
+export APPIMAGE_EXTRACT_AND_RUN=1
+
+mkdir -p AppDir/usr/optional
+mkdir -p AppDir/usr/optional/libstdc++
+mkdir -p AppDir/usr/optional/libgcc_s
+
+# Deploy yuzu's needed dependencies
+./linuxdeploy-x86_64.AppImage --appdir AppDir --plugin qt
+
+# Workaround for building yuzu with GCC 10 but also trying to distribute it to Ubuntu 18.04 et al.
+# See https://github.com/darealshinji/AppImageKit-checkrt
+cp exec-x86_64.so AppDir/usr/optional/exec.so
+cp AppRun-patched-x86_64 AppDir/AppRun
+cp --dereference /usr/lib/x86_64-linux-gnu/libstdc++.so.6 AppDir/usr/optional/libstdc++/libstdc++.so.6
+cp --dereference /lib/x86_64-linux-gnu/libgcc_s.so.1 AppDir/usr/optional/libgcc_s/libgcc_s.so.1

.ci/scripts/linux/upload.sh

@@ -2,6 +2,7 @@
 
 . .ci/scripts/common/pre-upload.sh
 
+APPIMAGE_NAME="yuzu-${GITDATE}-${GITREV}.AppImage"
 REV_NAME="yuzu-linux-${GITDATE}-${GITREV}"
 ARCHIVE_NAME="${REV_NAME}.tar.xz"
 COMPRESSION_FLAGS="-cJvf"
@@ -17,4 +18,24 @@ mkdir "$DIR_NAME"
 cp build/bin/yuzu-cmd "$DIR_NAME"
 cp build/bin/yuzu "$DIR_NAME"
 
+# Build an AppImage
+cd build
+
+wget -nc https://github.com/AppImage/AppImageKit/releases/download/continuous/appimagetool-x86_64.AppImage
+chmod 755 appimagetool-x86_64.AppImage
+
+if [ "${RELEASE_NAME}" = "mainline" ]; then
+    # Generate update information if releasing to mainline
+    ./appimagetool-x86_64.AppImage -u "gh-releases-zsync|yuzu-emu|yuzu-${RELEASE_NAME}|latest|yuzu-*.AppImage.zsync" AppDir "${APPIMAGE_NAME}"
+else
+    ./appimagetool-x86_64.AppImage AppDir "${APPIMAGE_NAME}"
+fi
+cd ..
+
+# Copy the AppImage and update info to the artifacts directory and avoid compressing it
+cp "build/${APPIMAGE_NAME}" "${ARTIFACTS_DIR}/"
+if [ -f "build/${APPIMAGE_NAME}.zsync" ]; then
+    cp "build/${APPIMAGE_NAME}.zsync" "${ARTIFACTS_DIR}/"
+fi
+
 . .ci/scripts/common/post-upload.sh

.ci/scripts/windows/docker.sh

@@ -5,7 +5,7 @@ cd /yuzu
 ccache -s
 
 mkdir build || true && cd build
-cmake .. -G Ninja -DDISPLAY_VERSION=$1 -DCMAKE_TOOLCHAIN_FILE="$(pwd)/../CMakeModules/MinGWCross.cmake" -DUSE_CCACHE=ON -DYUZU_USE_BUNDLED_UNICORN=ON -DENABLE_COMPATIBILITY_LIST_DOWNLOAD=ON -DCMAKE_BUILD_TYPE=Release -DENABLE_QT_TRANSLATION=ON
+cmake .. -G Ninja -DDISPLAY_VERSION=$1 -DCMAKE_TOOLCHAIN_FILE="$(pwd)/../CMakeModules/MinGWCross.cmake" -DUSE_CCACHE=ON -DENABLE_COMPATIBILITY_LIST_DOWNLOAD=ON -DCMAKE_BUILD_TYPE=Release -DENABLE_QT_TRANSLATION=ON
 ninja
 
 ccache -s
@@ -42,3 +42,8 @@ done
 pip3 install pefile
 python3 .ci/scripts/windows/scan_dll.py package/*.exe "package/"
 python3 .ci/scripts/windows/scan_dll.py package/imageformats/*.dll "package/"
+
+# copy FFmpeg libraries
+EXTERNALS_PATH="$(pwd)/build/externals"
+FFMPEG_DLL_PATH="$(find ${EXTERNALS_PATH} -maxdepth 1 -type d | grep ffmpeg)/bin"
+find ${FFMPEG_DLL_PATH} -type f -regex ".*\.dll" -exec cp -v {} package/ ';'

.ci/templates/build-msvc.yml

@@ -4,9 +4,11 @@ parameters:
   version: ''
 
 steps:
+- script: choco install vulkan-sdk
+  displayName: 'Install vulkan-sdk'
 - script: python -m pip install --upgrade pip conan
   displayName: 'Install conan'
-- script: mkdir build && cd build && cmake -G "Visual Studio 16 2019" -A x64 --config Release -DYUZU_USE_BUNDLED_QT=1 -DYUZU_USE_BUNDLED_UNICORN=1 -DYUZU_USE_QT_WEB_ENGINE=ON -DENABLE_COMPATIBILITY_LIST_DOWNLOAD=ON -DYUZU_ENABLE_COMPATIBILITY_REPORTING=${COMPAT} -DUSE_DISCORD_PRESENCE=ON -DDISPLAY_VERSION=${{ parameters['version'] }} .. && cd ..
+- script: refreshenv && mkdir build && cd build && cmake -G "Visual Studio 16 2019" -A x64 --config Release -DYUZU_USE_BUNDLED_QT=1 -DYUZU_USE_QT_WEB_ENGINE=ON -DENABLE_COMPATIBILITY_LIST_DOWNLOAD=ON -DYUZU_ENABLE_COMPATIBILITY_REPORTING=${COMPAT} -DUSE_DISCORD_PRESENCE=ON -DENABLE_QT_TRANSLATION=ON -DDISPLAY_VERSION=${{ parameters['version'] }} .. && cd ..
   displayName: 'Configure CMake'
 - task: MSBuild@1
   displayName: 'Build'

.gitmodules

@@ -37,3 +37,6 @@
 [submodule "opus"]
 	path = externals/opus/opus
 	url = https://github.com/xiph/opus.git
+[submodule "ffmpeg"]
+	path = externals/ffmpeg
+	url = https://git.ffmpeg.org/ffmpeg.git

.travis/linux/build.sh

@@ -1,4 +1,4 @@
 #!/bin/bash -ex
 
 mkdir -p "$HOME/.ccache"
-docker run -e ENABLE_COMPATIBILITY_REPORTING --env-file .travis/common/travis-ci.env -v $(pwd):/yuzu -v "$HOME/.ccache":/root/.ccache yuzuemu/build-environments:linux-fresh /bin/bash /yuzu/.travis/linux/docker.sh
+docker run -e ENABLE_COMPATIBILITY_REPORTING --env-file .travis/common/travis-ci.env -v $(pwd):/yuzu -v "$HOME/.ccache":/home/yuzu/.ccache yuzuemu/build-environments:linux-fresh /bin/bash /yuzu/.travis/linux/docker.sh

CMakeLists.txt

@@ -18,16 +18,20 @@ CMAKE_DEPENDENT_OPTION(YUZU_USE_BUNDLED_QT "Download bundled Qt binaries" ON "EN
 
 option(ENABLE_WEB_SERVICE "Enable web services (telemetry, etc.)" ON)
 
+CMAKE_DEPENDENT_OPTION(YUZU_USE_BUNDLED_FFMPEG "Download/Build bundled yuzu" ON "WIN32" OFF)
+
 option(YUZU_USE_QT_WEB_ENGINE "Use QtWebEngine for web applet implementation" OFF)
 
 option(YUZU_ENABLE_BOXCAT "Enable the Boxcat service, a yuzu high-level implementation of BCAT" ON)
 
 option(ENABLE_CUBEB "Enables the cubeb audio backend" ON)
 
-option(ENABLE_VULKAN "Enables Vulkan backend" ON)
-
 option(USE_DISCORD_PRESENCE "Enables Discord Rich Presence" OFF)
 
+if (NOT ENABLE_WEB_SERVICE)
+    set(YUZU_ENABLE_BOXCAT OFF)
+endif()
+
 # Default to a Release build
 get_property(IS_MULTI_CONFIG GLOBAL PROPERTY GENERATOR_IS_MULTI_CONFIG)
 if (NOT IS_MULTI_CONFIG AND NOT CMAKE_BUILD_TYPE)
@@ -113,6 +117,9 @@ if (NOT DEFINED ARCHITECTURE)
 endif()
 message(STATUS "Target architecture: ${ARCHITECTURE}")
 
+if (UNIX)
+    add_definitions(-DYUZU_UNIX=1)
+endif()
 
 # Configure C++ standard
 # ===========================
@@ -157,7 +164,6 @@ macro(yuzu_find_packages)
     # Capitalization matters here. We need the naming to match the generated paths from Conan
     set(REQUIRED_LIBS
     #    Cmake Pkg Prefix  Version     Conan Pkg
-        "Boost             1.73        boost/1.73.0"
         "Catch2            2.13        catch2/2.13.0"
         "fmt               7.1         fmt/7.1.2"
     # can't use until https://github.com/bincrafters/community/issues/1173
@@ -165,7 +171,7 @@ macro(yuzu_find_packages)
         "lz4               1.8         lz4/1.9.2"
         "nlohmann_json     3.8         nlohmann_json/3.8.0"
         "ZLIB              1.2         zlib/1.2.11"
-        "zstd              1.4         zstd/1.4.5"
+        "zstd              1.4         zstd/1.4.8"
     )
 
     foreach(PACKAGE ${REQUIRED_LIBS})
@@ -192,6 +198,22 @@ macro(yuzu_find_packages)
     unset(FN_FORCE_REQUIRED)
 endmacro()
 
+find_package(Boost 1.73.0 COMPONENTS context headers QUIET)
+if (Boost_FOUND)
+    set(Boost_LIBRARIES Boost::boost)
+    # Conditionally add Boost::context only if the active version of the Conan or system Boost package provides it
+    # The old version is missing Boost::context, so we want to avoid adding in that case
+    # The new version requires adding Boost::context to prevent linking issues
+    #
+    # This one is used by Conan on subsequent CMake configures, not the first configure.
+    if (TARGET Boost::context)
+        list(APPEND Boost_LIBRARIES Boost::context)
+    endif()
+else()
+    message(STATUS "Boost 1.73.0 or newer not found, falling back to Conan")
+    list(APPEND CONAN_REQUIRED_LIBS "boost/1.73.0")
+endif()
+
 # Attempt to locate any packages that are required and report the missing ones in CONAN_REQUIRED_LIBS
 yuzu_find_packages()
 
@@ -223,7 +245,7 @@ if(ENABLE_QT)
     if (YUZU_USE_QT_WEB_ENGINE)
         find_package(Qt5 COMPONENTS WebEngineCore WebEngineWidgets)
     endif()
-	
+
     if (ENABLE_QT_TRANSLATION)
         find_package(Qt5 REQUIRED COMPONENTS LinguistTools ${QT_PREFIX_HINT})
     endif()
@@ -241,7 +263,7 @@ if(ENABLE_SDL2)
     find_package(SDL2)
     if (NOT SDL2_FOUND)
         # otherwise add this to the list of libraries to install
-        list(APPEND CONAN_REQUIRED_LIBS "sdl2/2.0.12@bincrafters/stable")
+        list(APPEND CONAN_REQUIRED_LIBS "sdl2/2.0.14@bincrafters/stable")
     endif()
 endif()
 
@@ -296,6 +318,17 @@ if (CONAN_REQUIRED_LIBS)
     # this time with required, so we bail if its not found.
     yuzu_find_packages(FORCE_REQUIRED)
 
+    if (NOT Boost_FOUND)
+        find_package(Boost 1.73.0 REQUIRED COMPONENTS context headers)
+        set(Boost_LIBRARIES Boost::boost)
+        # Conditionally add Boost::context only if the active version of the Conan Boost package provides it
+        # The old version is missing Boost::context, so we want to avoid adding in that case
+        # The new version requires adding Boost::context to prevent linking issues
+        if (TARGET Boost::context)
+            list(APPEND Boost_LIBRARIES Boost::context)
+        endif()
+    endif()
+
     # Due to issues with variable scopes in functions, we need to also find_package(qt5) outside of the function
     if(ENABLE_QT)
         list(APPEND CMAKE_MODULE_PATH "${CONAN_QT_ROOT_RELEASE}")
@@ -353,19 +386,141 @@ if (NOT LIBUSB_FOUND)
     set(LIBUSB_LIBRARIES usb)
 endif()
 
-# Use system installed ffmpeg.
-if (NOT MSVC)
-    find_package(FFmpeg REQUIRED)
-else()
-    set(FFMPEG_EXT_NAME "ffmpeg-4.2.1")
-    set(FFMPEG_PATH "${CMAKE_BINARY_DIR}/externals/${FFMPEG_EXT_NAME}")
-    download_bundled_external("ffmpeg/" ${FFMPEG_EXT_NAME} "")
-    set(FFMPEG_FOUND YES)
-    set(FFMPEG_INCLUDE_DIR "${FFMPEG_PATH}/include" CACHE PATH "Path to FFmpeg headers" FORCE)
-    set(FFMPEG_LIBRARY_DIR "${FFMPEG_PATH}/bin" CACHE PATH "Path to FFmpeg library" FORCE)
-    set(FFMPEG_DLL_DIR "${FFMPEG_PATH}/bin" CACHE PATH "Path to FFmpeg dll's" FORCE)
+# List of all FFmpeg components required
+set(FFmpeg_COMPONENTS
+    avcodec
+    avutil
+    swscale)
+
+if (NOT YUZU_USE_BUNDLED_FFMPEG)
+    # Use system installed FFmpeg
+    find_package(FFmpeg REQUIRED COMPONENTS ${FFmpeg_COMPONENTS})
+
+    if (FFmpeg_FOUND)
+        # Overwrite aggregate defines from FFmpeg module to avoid over-linking libraries.
+        # Prevents shipping too many libraries with the AppImage.
+        set(FFmpeg_LIBRARIES "")
+        set(FFmpeg_INCLUDE_DIR "")
+
+        foreach(COMPONENT ${FFmpeg_COMPONENTS})
+            set(FFmpeg_LIBRARIES ${FFmpeg_LIBRARIES} ${FFmpeg_LIBRARY_${COMPONENT}} CACHE PATH "Paths to FFmpeg libraries" FORCE)
+            set(FFmpeg_INCLUDE_DIR ${FFmpeg_INCLUDE_DIR} ${FFmpeg_INCLUDE_${COMPONENT}} CACHE PATH "Path to FFmpeg headers" FORCE)
+        endforeach()
+    else()
+        message(WARNING "FFmpeg not found, falling back to externals")
+        set(YUZU_USE_BUNDLED_FFMPEG ON)
+    endif()
 endif()
 
+if (YUZU_USE_BUNDLED_FFMPEG)
+    if (NOT WIN32)
+        # Build FFmpeg from externals
+        message(STATUS "Using FFmpeg from externals")
+
+        # FFmpeg has source that requires one of nasm or yasm to assemble it.
+        # REQUIRED throws an error if not found here during configuration rather than during compilation.
+        find_program(ASSEMBLER NAMES nasm yasm REQUIRED)
+
+        set(FFmpeg_PREFIX ${PROJECT_SOURCE_DIR}/externals/ffmpeg)
+        set(FFmpeg_BUILD_DIR ${PROJECT_BINARY_DIR}/externals/ffmpeg)
+        set(FFmpeg_MAKEFILE ${FFmpeg_BUILD_DIR}/Makefile)
+        make_directory(${FFmpeg_BUILD_DIR})
+
+        # Read version string from external
+        file(READ ${FFmpeg_PREFIX}/RELEASE FFmpeg_VERSION)
+        set(FFmpeg_FOUND NO)
+        if (NOT FFmpeg_VERSION STREQUAL "")
+            set(FFmpeg_FOUND YES)
+        endif()
+
+        foreach(COMPONENT ${FFmpeg_COMPONENTS})
+            set(FFmpeg_${COMPONENT}_PREFIX "${FFmpeg_BUILD_DIR}/lib${COMPONENT}")
+            set(FFmpeg_${COMPONENT}_LIB_NAME "lib${COMPONENT}.a")
+            set(FFmpeg_${COMPONENT}_LIBRARY "${FFmpeg_${COMPONENT}_PREFIX}/${FFmpeg_${COMPONENT}_LIB_NAME}")
+
+            set(FFmpeg_LIBRARIES
+                ${FFmpeg_LIBRARIES}
+                ${FFmpeg_${COMPONENT}_LIBRARY}
+                CACHE PATH "Paths to FFmpeg libraries" FORCE)
+        endforeach()
+
+        set(FFmpeg_INCLUDE_DIR
+            ${FFmpeg_PREFIX}
+            CACHE PATH "Path to FFmpeg headers" FORCE)
+
+        # `configure` parameters builds only exactly what yuzu needs from FFmpeg
+        # `--disable-{vaapi,vdpau}` is needed to avoid linking issues
+        add_custom_command(
+            OUTPUT
+                ${FFmpeg_MAKEFILE}
+            COMMAND
+                /bin/bash ${FFmpeg_PREFIX}/configure
+                    --disable-avdevice
+                    --disable-avfilter
+                    --disable-avformat
+                    --disable-doc
+                    --disable-everything
+                    --disable-ffmpeg
+                    --disable-ffprobe
+                    --disable-network
+                    --disable-postproc
+                    --disable-swresample
+                    --disable-vaapi
+                    --disable-vdpau
+                    --enable-decoder=h264
+                    --enable-decoder=vp9
+            WORKING_DIRECTORY
+                ${FFmpeg_BUILD_DIR}
+        )
+
+        # Workaround for Ubuntu 18.04's older version of make not being able to call make as a child
+        # with context of the jobserver. Also helps ninja users.
+        execute_process(
+            COMMAND
+                nproc
+            OUTPUT_VARIABLE
+                SYSTEM_THREADS)
+
+        add_custom_command(
+            OUTPUT
+                ${FFmpeg_LIBRARIES}
+            COMMAND
+                make -j${SYSTEM_THREADS}
+            WORKING_DIRECTORY
+                ${FFmpeg_BUILD_DIR}
+        )
+
+        # ALL makes this custom target build every time
+        # but it won't actually build if the DEPENDS parameter is up to date
+        add_custom_target(ffmpeg-build ALL DEPENDS ${FFmpeg_LIBRARIES})
+        add_custom_target(ffmpeg-configure ALL DEPENDS ${FFmpeg_MAKEFILE})
+
+        if (FFmpeg_FOUND)
+            message(STATUS "Found FFmpeg version ${FFmpeg_VERSION}")
+
+            add_dependencies(ffmpeg-build ffmpeg-configure)
+        else()
+            message(FATAL_ERROR "FFmpeg not found")
+        endif()
+    else() # WIN32
+        # Use yuzu FFmpeg binaries
+        set(FFmpeg_EXT_NAME "ffmpeg-4.3.1")
+        set(FFmpeg_PATH "${CMAKE_BINARY_DIR}/externals/${FFmpeg_EXT_NAME}")
+        download_bundled_external("ffmpeg/" ${FFmpeg_EXT_NAME} "")
+        set(FFmpeg_FOUND YES)
+        set(FFmpeg_INCLUDE_DIR "${FFmpeg_PATH}/include" CACHE PATH "Path to FFmpeg headers" FORCE)
+        set(FFmpeg_LIBRARY_DIR "${FFmpeg_PATH}/bin" CACHE PATH "Path to FFmpeg library directory" FORCE)
+        set(FFmpeg_DLL_DIR "${FFmpeg_PATH}/bin" CACHE PATH "Path to FFmpeg dll's" FORCE)
+        set(FFmpeg_LIBRARIES
+            ${FFmpeg_LIBRARY_DIR}/swscale.lib
+            ${FFmpeg_LIBRARY_DIR}/avcodec.lib
+            ${FFmpeg_LIBRARY_DIR}/avutil.lib
+            CACHE PATH "Paths to FFmpeg libraries" FORCE)
+    endif()
+endif()
+
+unset(FFmpeg_COMPONENTS)
+
 # Prefer the -pthread flag on Linux.
 set(THREADS_PREFER_PTHREAD_FLAG ON)
 find_package(Threads REQUIRED)

CMakeModules/CopyYuzuFFmpegDeps.cmake

@@ -1,10 +1,6 @@
 function(copy_yuzu_FFmpeg_deps target_dir)
     include(WindowsCopyFiles)
     set(DLL_DEST "${CMAKE_BINARY_DIR}/bin/$<CONFIG>/")
-    windows_copy_files(${target_dir} ${FFMPEG_DLL_DIR} ${DLL_DEST}
-        avcodec-58.dll
-        avutil-56.dll
-        swresample-3.dll
-        swscale-5.dll
-    )
+    file(READ "${FFmpeg_PATH}/requirements.txt" FFmpeg_REQUIRED_DLLS)
+    windows_copy_files(${target_dir} ${FFmpeg_DLL_DIR} ${DLL_DEST} ${FFmpeg_REQUIRED_DLLS})
 endfunction(copy_yuzu_FFmpeg_deps)

README.md

@@ -30,11 +30,10 @@ If you want to contribute to the user interface translation, please check out th
 
 * __Windows__: [Windows Build](https://github.com/yuzu-emu/yuzu/wiki/Building-For-Windows)
 * __Linux__: [Linux Build](https://github.com/yuzu-emu/yuzu/wiki/Building-For-Linux)
-* __macOS__: [macOS Build](https://github.com/yuzu-emu/yuzu/wiki/Building-for-macOS)
 
 
 ### Support
-We happily accept monetary donations or donated games and hardware. Please see our [donations page](https://yuzu-emu.org/donate/) for more information on how you can contribute to yuzu. Any donations received will go towards things like:
+We happily accept monetary donations, or donated games and hardware. Please see our [donations page](https://yuzu-emu.org/donate/) for more information on how you can contribute to yuzu. Any donations received will go towards things like:
 * Switch consoles to explore and reverse-engineer the hardware
 * Switch games for testing, reverse-engineering, and implementing new features
 * Web hosting and infrastructure setup

dist/icons/controller/controller.qrc

@@ -1,26 +1,5 @@
 <RCC>
   <qresource prefix="controller">
-    <file alias="dual_joycon">dual_joycon.png</file>
-    <file alias="dual_joycon_dark">dual_joycon_dark.png</file>
-    <file alias="dual_joycon_midnight">dual_joycon_midnight.png</file>
-    <file alias="handheld">handheld.png</file>
-    <file alias="handheld_dark">handheld_dark.png</file>
-    <file alias="handheld_midnight">handheld_midnight.png</file>
-    <file alias="pro_controller">pro_controller.png</file>
-    <file alias="pro_controller_dark">pro_controller_dark.png</file>
-    <file alias="pro_controller_midnight">pro_controller_midnight.png</file>
-    <file alias="single_joycon_left">single_joycon_left.png</file>
-    <file alias="single_joycon_left_dark">single_joycon_left_dark.png</file>
-    <file alias="single_joycon_left_midnight">single_joycon_left_midnight.png</file>
-    <file alias="single_joycon_right">single_joycon_right.png</file>
-    <file alias="single_joycon_right_dark">single_joycon_right_dark.png</file>
-    <file alias="single_joycon_right_midnight">single_joycon_right_midnight.png</file>
-    <file alias="single_joycon_left_vertical">single_joycon_left_vertical.png</file>
-    <file alias="single_joycon_left_vertical_dark">single_joycon_left_vertical_dark.png</file>
-    <file alias="single_joycon_left_vertical_midnight">single_joycon_left_vertical_midnight.png</file>
-    <file alias="single_joycon_right_vertical">single_joycon_right_vertical.png</file>
-    <file alias="single_joycon_right_vertical_dark">single_joycon_right_vertical_dark.png</file>
-    <file alias="single_joycon_right_vertical_midnight">single_joycon_right_vertical_midnight.png</file>
     <file alias="applet_dual_joycon">applet_dual_joycon.png</file>
     <file alias="applet_dual_joycon_dark">applet_dual_joycon_dark.png</file>	
     <file alias="applet_dual_joycon_midnight">applet_dual_joycon_midnight.png</file>	

dist/qt_themes/colorful_dark/icons/index.theme

@@ -1,7 +1,7 @@
 [Icon Theme]
 Name=colorful_dark
 Comment=Colorful theme (Dark style)
-Inherits=default
+Inherits=colorful
 Directories=16x16
  
 [16x16]

dist/qt_themes/colorful_midnight_blue/icons/index.theme

@@ -1,7 +1,7 @@
 [Icon Theme]
 Name=colorful_midnight_blue
 Comment=Colorful theme (Midnight Blue style)
-Inherits=default
+Inherits=colorful
 Directories=16x16
 
 [16x16]

dist/qt_themes/qdarkstyle_midnight_blue/style.qss

@@ -1257,10 +1257,6 @@ QComboBox::item:alternate {
   background: #19232D;
 }
 
-QComboBox::item:checked {
-  font-weight: bold;
-}
-
 QComboBox::item:selected {
   border: 0px solid transparent;
 }

externals/CMakeLists.txt

@@ -61,9 +61,7 @@ if (USE_DISCORD_PRESENCE)
 endif()
 
 # Sirit
-if (ENABLE_VULKAN)
-    add_subdirectory(sirit)
-endif()
+add_subdirectory(sirit)
 
 # libzip
 find_package(Libzip 1.5)

externals/find-modules/FindFFmpeg.cmake

@@ -1,100 +1,187 @@
-# - Try to find ffmpeg libraries (libavcodec, libavformat and libavutil)
-# Once done this will define
+# FindFFmpeg
+# ----------
 #
-# FFMPEG_FOUND - system has ffmpeg or libav
-# FFMPEG_INCLUDE_DIR - the ffmpeg include directory
-# FFMPEG_LIBRARIES - Link these to use ffmpeg
-# FFMPEG_LIBAVCODEC
-# FFMPEG_LIBAVFORMAT
-# FFMPEG_LIBAVUTIL
+# Copyright 2019 Citra Emulator Project
+# Licensed under GPLv2 or any later version
 #
-# Copyright (c) 2008 Andreas Schneider <mail@cynapses.org>
-# Modified for other libraries by Lasse Kärkkäinen <tronic>
-# Modified for Hedgewars by Stepik777
-# Modified for FFmpeg-example Tuukka Pasanen 2018
-# Modified for yuzu toastUnlimted 2020
+# Find the native FFmpeg includes and libraries
 #
-# Redistribution and use is allowed according to the terms of the New
-# BSD license.
+# This module defines the following variables:
+#
+#  FFmpeg_INCLUDE_<component>: where to find <component>.h
+#  FFmpeg_LIBRARY_<component>: where to find the <component> library
+#  FFmpeg_INCLUDE_DIR: aggregate all the include paths
+#  FFmpeg_LIBRARIES: aggregate all the paths to the libraries
+#  FFmpeg_FOUND: True if all components have been found
+#
+# This module defines the following targets, which are prefered over variables:
+#
+#  FFmpeg::<component>: Target to use <component> directly, with include path,
+#    library and dependencies set up. If you are using a static build, you are
+#    responsible for adding any external dependencies (such as zlib, bzlib...).
+#
+# <component> can be one of:
+#   avcodec
+#   avdevice
+#   avfilter
+#   avformat
+#   avutil
+#   postproc
+#   swresample
+#   swscale
 #
 
-include(FindPackageHandleStandardArgs)
-
-find_package_handle_standard_args(FFMPEG
-  FOUND_VAR FFMPEG_FOUND
-  REQUIRED_VARS
-      FFMPEG_LIBRARY
-      FFMPEG_INCLUDE_DIR
-  VERSION_VAR FFMPEG_VERSION
+set(_FFmpeg_ALL_COMPONENTS
+    avcodec
+    avdevice
+    avfilter
+    avformat
+    avutil
+    postproc
+    swresample
+    swscale
 )
 
-if(FFMPEG_LIBRARIES AND FFMPEG_INCLUDE_DIR)
-  # in cache already
-  set(FFMPEG_FOUND TRUE)
-else()
-  # use pkg-config to get the directories and then use these values
-  # in the FIND_PATH() and FIND_LIBRARY() calls
-  find_package(PkgConfig)
-  if(PKG_CONFIG_FOUND)
-    pkg_check_modules(_FFMPEG_AVCODEC libavcodec)
-    pkg_check_modules(_FFMPEG_AVUTIL libavutil)
-    pkg_check_modules(_FFMPEG_SWSCALE libswscale)
+set(_FFmpeg_DEPS_avcodec avutil)
+set(_FFmpeg_DEPS_avdevice avcodec avformat avutil)
+set(_FFmpeg_DEPS_avfilter avutil)
+set(_FFmpeg_DEPS_avformat avcodec avutil)
+set(_FFmpeg_DEPS_postproc avutil)
+set(_FFmpeg_DEPS_swresample avutil)
+set(_FFmpeg_DEPS_swscale avutil)
+
+function(find_ffmpeg LIBNAME)
+  if(DEFINED ENV{FFMPEG_DIR})
+    set(FFMPEG_DIR $ENV{FFMPEG_DIR})
   endif()
 
-  find_path(FFMPEG_AVCODEC_INCLUDE_DIR
-    NAMES libavcodec/avcodec.h
-    PATHS ${_FFMPEG_AVCODEC_INCLUDE_DIRS}
-      /usr/include
-      /usr/local/include
-      /opt/local/include
-      /sw/include
-    PATH_SUFFIXES ffmpeg libav)
-
-  find_library(FFMPEG_LIBAVCODEC
-    NAMES avcodec
-    PATHS ${_FFMPEG_AVCODEC_LIBRARY_DIRS}
-      /usr/lib
-      /usr/local/lib
-      /opt/local/lib
-      /sw/lib)
-
-  find_library(FFMPEG_LIBAVUTIL
-    NAMES avutil
-    PATHS ${_FFMPEG_AVUTIL_LIBRARY_DIRS}
-      /usr/lib
-      /usr/local/lib
-      /opt/local/lib
-      /sw/lib)
-
-  find_library(FFMPEG_LIBSWSCALE
-    NAMES swscale
-    PATHS ${_FFMPEG_SWSCALE_LIBRARY_DIRS}
-      /usr/lib
-      /usr/local/lib
-      /opt/local/lib
-      /sw/lib)
-
-  if(FFMPEG_LIBAVCODEC AND FFMPEG_LIBAVUTIL AND FFMPEG_LIBSWSCALE)
-    set(FFMPEG_FOUND TRUE)
-  endif()
-
-  if(FFMPEG_FOUND)
-    set(FFMPEG_INCLUDE_DIR ${FFMPEG_AVCODEC_INCLUDE_DIR})
-    set(FFMPEG_LIBRARIES
-      ${FFMPEG_LIBAVCODEC}
-      ${FFMPEG_LIBAVUTIL}
-      ${FFMPEG_LIBSWSCALE})
-  endif()
-
-  if(FFMPEG_FOUND)
-    if(NOT FFMPEG_FIND_QUIETLY)
-      message(STATUS
-      "Found FFMPEG or Libav: ${FFMPEG_LIBRARIES}, ${FFMPEG_INCLUDE_DIR}")
-    endif()
+  if(FFMPEG_DIR)
+    list(APPEND INCLUDE_PATHS
+      ${FFMPEG_DIR}
+      ${FFMPEG_DIR}/ffmpeg
+      ${FFMPEG_DIR}/lib${LIBNAME}
+      ${FFMPEG_DIR}/include/lib${LIBNAME}
+      ${FFMPEG_DIR}/include/ffmpeg
+      ${FFMPEG_DIR}/include
+      NO_DEFAULT_PATH
+      NO_CMAKE_FIND_ROOT_PATH
+    )
+    list(APPEND LIB_PATHS
+      ${FFMPEG_DIR}
+      ${FFMPEG_DIR}/lib
+      ${FFMPEG_DIR}/lib${LIBNAME}
+      NO_DEFAULT_PATH
+      NO_CMAKE_FIND_ROOT_PATH
+    )
   else()
-    if(FFMPEG_FIND_REQUIRED)
-      message(FATAL_ERROR
-      "Could not find libavcodec or libavutil or libswscale")
+    list(APPEND INCLUDE_PATHS
+      /usr/local/include/ffmpeg
+      /usr/local/include/lib${LIBNAME}
+      /usr/include/ffmpeg
+      /usr/include/lib${LIBNAME}
+      /usr/include/ffmpeg/lib${LIBNAME}
+    )
+
+    list(APPEND LIB_PATHS
+      /usr/local/lib
+      /usr/lib
+    )
+  endif()
+
+  find_path(FFmpeg_INCLUDE_${LIBNAME} lib${LIBNAME}/${LIBNAME}.h
+    HINTS ${INCLUDE_PATHS}
+  )
+
+  find_library(FFmpeg_LIBRARY_${LIBNAME} ${LIBNAME}
+    HINTS ${LIB_PATHS}
+  )
+
+  if(NOT FFMPEG_DIR AND (NOT FFmpeg_LIBRARY_${LIBNAME} OR NOT FFmpeg_INCLUDE_${LIBNAME}))
+    # Didn't find it in the usual paths, try pkg-config
+    find_package(PkgConfig QUIET)
+    pkg_check_modules(FFmpeg_PKGCONFIG_${LIBNAME} QUIET lib${LIBNAME})
+
+    find_path(FFmpeg_INCLUDE_${LIBNAME} lib${LIBNAME}/${LIBNAME}.h
+      ${FFmpeg_PKGCONFIG_${LIBNAME}_INCLUDE_DIRS}
+    )
+
+    find_library(FFmpeg_LIBRARY_${LIBNAME} ${LIBNAME}
+      ${FFmpeg_PKGCONFIG_${LIBNAME}_LIBRARY_DIRS}
+    )
+  endif()
+
+  if(FFmpeg_INCLUDE_${LIBNAME} AND FFmpeg_LIBRARY_${LIBNAME})
+    set(FFmpeg_INCLUDE_${LIBNAME} "${FFmpeg_INCLUDE_${LIBNAME}}" PARENT_SCOPE)
+    set(FFmpeg_LIBRARY_${LIBNAME} "${FFmpeg_LIBRARY_${LIBNAME}}" PARENT_SCOPE)
+
+    # Extract FFmpeg version from version.h
+    foreach(v MAJOR MINOR MICRO)
+      set(FFmpeg_${LIBNAME}_VERSION_${v} 0)
+    endforeach()
+    string(TOUPPER ${LIBNAME} LIBNAME_UPPER)
+    file(STRINGS "${FFmpeg_INCLUDE_${LIBNAME}}/lib${LIBNAME}/version.h" _FFmpeg_VERSION_H_CONTENTS REGEX "#define LIB${LIBNAME_UPPER}_VERSION_(MAJOR|MINOR|MICRO) ")
+    set(_FFmpeg_VERSION_REGEX "([0-9]+)")
+    foreach(v MAJOR MINOR MICRO)
+      if("${_FFmpeg_VERSION_H_CONTENTS}" MATCHES "#define LIB${LIBNAME_UPPER}_VERSION_${v}[\\t ]+${_FFmpeg_VERSION_REGEX}")
+        set(FFmpeg_${LIBNAME}_VERSION_${v} "${CMAKE_MATCH_1}")
+      endif()
+    endforeach()
+    set(FFmpeg_${LIBNAME}_VERSION "${FFmpeg_${LIBNAME}_VERSION_MAJOR}.${FFmpeg_${LIBNAME}_VERSION_MINOR}.${FFmpeg_${LIBNAME}_VERSION_MICRO}")
+    set(FFmpeg_${c}_VERSION "${FFmpeg_${LIBNAME}_VERSION}" PARENT_SCOPE)
+    unset(_FFmpeg_VERSION_REGEX)
+    unset(_FFmpeg_VERSION_H_CONTENTS)
+
+    set(FFmpeg_${c}_FOUND TRUE PARENT_SCOPE)
+    if(NOT FFmpeg_FIND_QUIETLY)
+      message("--  Found ${LIBNAME}: ${FFmpeg_INCLUDE_${LIBNAME}} ${FFmpeg_LIBRARY_${LIBNAME}} (version: ${FFmpeg_${LIBNAME}_VERSION})")
     endif()
   endif()
+endfunction()
+
+foreach(c ${_FFmpeg_ALL_COMPONENTS})
+  find_ffmpeg(${c})
+endforeach()
+
+foreach(c ${_FFmpeg_ALL_COMPONENTS})
+  if(FFmpeg_${c}_FOUND)
+    list(APPEND FFmpeg_INCLUDE_DIR ${FFmpeg_INCLUDE_${c}})
+    list(APPEND FFmpeg_LIBRARIES ${FFmpeg_LIBRARY_${c}})
+
+    add_library(FFmpeg::${c} IMPORTED UNKNOWN)
+    set_target_properties(FFmpeg::${c} PROPERTIES
+      IMPORTED_LOCATION ${FFmpeg_LIBRARY_${c}}
+      INTERFACE_INCLUDE_DIRECTORIES ${FFmpeg_INCLUDE_${c}}
+    )
+    if(_FFmpeg_DEPS_${c})
+      set(deps)
+      foreach(dep ${_FFmpeg_DEPS_${c}})
+        list(APPEND deps FFmpeg::${dep})
+      endforeach()
+
+      set_target_properties(FFmpeg::${c} PROPERTIES
+        INTERFACE_LINK_LIBRARIES "${deps}"
+      )
+      unset(deps)
+    endif()
+  endif()
+endforeach()
+
+if(FFmpeg_INCLUDE_DIR)
+  list(REMOVE_DUPLICATES FFmpeg_INCLUDE_DIR)
 endif()
+
+foreach(c ${FFmpeg_FIND_COMPONENTS})
+  list(APPEND _FFmpeg_REQUIRED_VARS FFmpeg_INCLUDE_${c} FFmpeg_LIBRARY_${c})
+endforeach()
+
+include(FindPackageHandleStandardArgs)
+find_package_handle_standard_args(FFmpeg
+  REQUIRED_VARS ${_FFmpeg_REQUIRED_VARS}
+  HANDLE_COMPONENTS
+)
+
+foreach(c ${_FFmpeg_ALL_COMPONENTS})
+  unset(_FFmpeg_DEPS_${c})
+endforeach()
+unset(_FFmpeg_ALL_COMPONENTS)
+unset(_FFmpeg_REQUIRED_VARS)

src/CMakeLists.txt

@@ -45,10 +45,15 @@ if (MSVC)
 
         # Warnings
         /W3
+        /we4062 # enumerator 'identifier' in a switch of enum 'enumeration' is not handled
+        /we4101 # 'identifier': unreferenced local variable
+        /we4265 # 'class': class has virtual functions, but destructor is not virtual
+        /we4388 # signed/unsigned mismatch
         /we4547 # 'operator' : operator before comma has no effect; expected operator with side-effect
         /we4549 # 'operator1': operator before comma has no effect; did you intend 'operator2'?
         /we4555 # Expression has no effect; expected expression with side-effect
         /we4834 # Discarding return value of function with 'nodiscard' attribute
+        /we5038 # data member 'member1' will be initialized after data member 'member2'
     )
 
     # /GS- - No stack buffer overflow checks
@@ -59,13 +64,20 @@ if (MSVC)
 else()
     add_compile_options(
         -Wall
+        -Werror=array-bounds
         -Werror=implicit-fallthrough
         -Werror=missing-declarations
+        -Werror=missing-field-initializers
         -Werror=reorder
+        -Werror=switch
+        -Werror=uninitialized
+        -Werror=unused-function
         -Werror=unused-result
+        -Werror=unused-variable
         -Wextra
         -Wmissing-declarations
         -Wno-attributes
+        -Wno-invalid-offsetof
         -Wno-unused-parameter
     )
 
@@ -120,7 +132,6 @@ add_subdirectory(tests)
 
 if (ENABLE_SDL2)
     add_subdirectory(yuzu_cmd)
-    add_subdirectory(yuzu_tester)
 endif()
 
 if (ENABLE_QT)

src/audio_core/CMakeLists.txt

@@ -15,6 +15,8 @@ add_library(audio_core STATIC
     command_generator.cpp
     command_generator.h
     common.h
+    delay_line.cpp
+    delay_line.h
     effect_context.cpp
     effect_context.h
     info_updater.cpp
@@ -51,6 +53,8 @@ if (NOT MSVC)
         -Werror=implicit-fallthrough
         -Werror=reorder
         -Werror=sign-compare
+        -Werror=shadow
+        -Werror=unused-parameter
         -Werror=unused-variable
 
         $<$<CXX_COMPILER_ID:GNU>:-Werror=unused-but-set-parameter>

src/audio_core/algorithm/filter.cpp

@@ -31,8 +31,8 @@ Filter Filter::LowPass(double cutoff, double Q) {
 
 Filter::Filter() : Filter(1.0, 0.0, 0.0, 1.0, 0.0, 0.0) {}
 
-Filter::Filter(double a0, double a1, double a2, double b0, double b1, double b2)
-    : a1(a1 / a0), a2(a2 / a0), b0(b0 / a0), b1(b1 / a0), b2(b2 / a0) {}
+Filter::Filter(double a0_, double a1_, double a2_, double b0_, double b1_, double b2_)
+    : a1(a1_ / a0_), a2(a2_ / a0_), b0(b0_ / a0_), b1(b1_ / a0_), b2(b2_ / a0_) {}
 
 void Filter::Process(std::vector<s16>& signal) {
     const std::size_t num_frames = signal.size() / 2;
@@ -69,7 +69,7 @@ CascadingFilter CascadingFilter::LowPass(double cutoff, std::size_t cascade_size
 }
 
 CascadingFilter::CascadingFilter() = default;
-CascadingFilter::CascadingFilter(std::vector<Filter> filters) : filters(std::move(filters)) {}
+CascadingFilter::CascadingFilter(std::vector<Filter> filters_) : filters(std::move(filters_)) {}
 
 void CascadingFilter::Process(std::vector<s16>& signal) {
     for (auto& filter : filters) {

src/audio_core/algorithm/filter.h

@@ -25,7 +25,7 @@ public:
     /// Passthrough filter.
     Filter();
 
-    Filter(double a0, double a1, double a2, double b0, double b1, double b2);
+    Filter(double a0_, double a1_, double a2_, double b0_, double b1_, double b2_);
 
     void Process(std::vector<s16>& signal);
 
@@ -51,7 +51,7 @@ public:
     /// Passthrough.
     CascadingFilter();
 
-    explicit CascadingFilter(std::vector<Filter> filters);
+    explicit CascadingFilter(std::vector<Filter> filters_);
 
     void Process(std::vector<s16>& signal);
 

src/audio_core/algorithm/interpolate.cpp

@@ -218,7 +218,7 @@ void Resample(s32* output, const s32* input, s32 pitch, s32& fraction, std::size
         const auto l2 = lut[lut_index + 2];
         const auto l3 = lut[lut_index + 3];
 
-        const auto s0 = static_cast<s32>(input[index]);
+        const auto s0 = static_cast<s32>(input[index + 0]);
         const auto s1 = static_cast<s32>(input[index + 1]);
         const auto s2 = static_cast<s32>(input[index + 2]);
         const auto s3 = static_cast<s32>(input[index + 3]);

src/audio_core/audio_out.cpp

@@ -43,6 +43,10 @@ std::vector<Buffer::Tag> AudioOut::GetTagsAndReleaseBuffers(StreamPtr stream,
     return stream->GetTagsAndReleaseBuffers(max_count);
 }
 
+std::vector<Buffer::Tag> AudioOut::GetTagsAndReleaseBuffers(StreamPtr stream) {
+    return stream->GetTagsAndReleaseBuffers();
+}
+
 void AudioOut::StartStream(StreamPtr stream) {
     stream->Play();
 }

src/audio_core/audio_out.h

@@ -31,6 +31,9 @@ public:
     /// Returns a vector of recently released buffers specified by tag for the specified stream
     std::vector<Buffer::Tag> GetTagsAndReleaseBuffers(StreamPtr stream, std::size_t max_count);
 
+    /// Returns a vector of all recently released buffers specified by tag for the specified stream
+    std::vector<Buffer::Tag> GetTagsAndReleaseBuffers(StreamPtr stream);
+
     /// Starts an audio stream for playback
     void StartStream(StreamPtr stream);
 

src/audio_core/audio_renderer.cpp

@@ -2,6 +2,7 @@
 // Licensed under GPLv2 or any later version
 // Refer to the license.txt file included.
 
+#include <limits>
 #include <vector>
 
 #include "audio_core/audio_out.h"
@@ -10,32 +11,81 @@
 #include "audio_core/info_updater.h"
 #include "audio_core/voice_context.h"
 #include "common/logging/log.h"
-#include "core/hle/kernel/writable_event.h"
 #include "core/memory.h"
 #include "core/settings.h"
 
+namespace {
+[[nodiscard]] static constexpr s16 ClampToS16(s32 value) {
+    return static_cast<s16>(std::clamp(value, s32{std::numeric_limits<s16>::min()},
+                                       s32{std::numeric_limits<s16>::max()}));
+}
+
+[[nodiscard]] static constexpr s16 Mix2To1(s16 l_channel, s16 r_channel) {
+    // Mix 50% from left and 50% from right channel
+    constexpr float l_mix_amount = 50.0f / 100.0f;
+    constexpr float r_mix_amount = 50.0f / 100.0f;
+    return ClampToS16(static_cast<s32>((static_cast<float>(l_channel) * l_mix_amount) +
+                                       (static_cast<float>(r_channel) * r_mix_amount)));
+}
+
+[[nodiscard]] static constexpr std::tuple<s16, s16> Mix6To2(s16 fl_channel, s16 fr_channel,
+                                                            s16 fc_channel,
+                                                            [[maybe_unused]] s16 lf_channel,
+                                                            s16 bl_channel, s16 br_channel) {
+    // Front channels are mixed 36.94%, Center channels are mixed to be 26.12% & the back channels
+    // are mixed to be 36.94%
+
+    constexpr float front_mix_amount = 36.94f / 100.0f;
+    constexpr float center_mix_amount = 26.12f / 100.0f;
+    constexpr float back_mix_amount = 36.94f / 100.0f;
+
+    // Mix 50% from left and 50% from right channel
+    const auto left = front_mix_amount * static_cast<float>(fl_channel) +
+                      center_mix_amount * static_cast<float>(fc_channel) +
+                      back_mix_amount * static_cast<float>(bl_channel);
+
+    const auto right = front_mix_amount * static_cast<float>(fr_channel) +
+                       center_mix_amount * static_cast<float>(fc_channel) +
+                       back_mix_amount * static_cast<float>(br_channel);
+
+    return {ClampToS16(static_cast<s32>(left)), ClampToS16(static_cast<s32>(right))};
+}
+
+[[nodiscard]] static constexpr std::tuple<s16, s16> Mix6To2WithCoefficients(
+    s16 fl_channel, s16 fr_channel, s16 fc_channel, s16 lf_channel, s16 bl_channel, s16 br_channel,
+    const std::array<float_le, 4>& coeff) {
+    const auto left =
+        static_cast<float>(fl_channel) * coeff[0] + static_cast<float>(fc_channel) * coeff[1] +
+        static_cast<float>(lf_channel) * coeff[2] + static_cast<float>(bl_channel) * coeff[0];
+
+    const auto right =
+        static_cast<float>(fr_channel) * coeff[0] + static_cast<float>(fc_channel) * coeff[1] +
+        static_cast<float>(lf_channel) * coeff[2] + static_cast<float>(br_channel) * coeff[0];
+
+    return {ClampToS16(static_cast<s32>(left)), ClampToS16(static_cast<s32>(right))};
+}
+
+} // namespace
+
 namespace AudioCore {
 AudioRenderer::AudioRenderer(Core::Timing::CoreTiming& core_timing, Core::Memory::Memory& memory_,
                              AudioCommon::AudioRendererParameter params,
-                             std::shared_ptr<Kernel::WritableEvent> buffer_event,
+                             Stream::ReleaseCallback&& release_callback,
                              std::size_t instance_number)
-    : worker_params{params}, buffer_event{buffer_event},
-      memory_pool_info(params.effect_count + params.voice_count * 4),
+    : worker_params{params}, memory_pool_info(params.effect_count + params.voice_count * 4),
       voice_context(params.voice_count), effect_context(params.effect_count), mix_context(),
       sink_context(params.sink_count), splitter_context(),
       voices(params.voice_count), memory{memory_},
       command_generator(worker_params, voice_context, mix_context, splitter_context, effect_context,
-                        memory),
-      temp_mix_buffer(AudioCommon::TOTAL_TEMP_MIX_SIZE) {
+                        memory) {
     behavior_info.SetUserRevision(params.revision);
     splitter_context.Initialize(behavior_info, params.splitter_count,
                                 params.num_splitter_send_channels);
     mix_context.Initialize(behavior_info, params.submix_count + 1, params.effect_count);
     audio_out = std::make_unique<AudioCore::AudioOut>();
-    stream =
-        audio_out->OpenStream(core_timing, params.sample_rate, AudioCommon::STREAM_NUM_CHANNELS,
-                              fmt::format("AudioRenderer-Instance{}", instance_number),
-                              [=]() { buffer_event->Signal(); });
+    stream = audio_out->OpenStream(
+        core_timing, params.sample_rate, AudioCommon::STREAM_NUM_CHANNELS,
+        fmt::format("AudioRenderer-Instance{}", instance_number), std::move(release_callback));
     audio_out->StartStream(stream);
 
     QueueMixedBuffer(0);
@@ -62,10 +112,6 @@ Stream::State AudioRenderer::GetStreamState() const {
     return stream->GetState();
 }
 
-static constexpr s16 ClampToS16(s32 value) {
-    return static_cast<s16>(std::clamp(value, -32768, 32767));
-}
-
 ResultCode AudioRenderer::UpdateAudioRenderer(const std::vector<u8>& input_params,
                                               std::vector<u8>& output_params) {
 
@@ -104,8 +150,8 @@ ResultCode AudioRenderer::UpdateAudioRenderer(const std::vector<u8>& input_param
         }
     }
 
-    auto mix_result = info_updater.UpdateMixes(mix_context, worker_params.mix_buffer_count,
-                                               splitter_context, effect_context);
+    const auto mix_result = info_updater.UpdateMixes(mix_context, worker_params.mix_buffer_count,
+                                                     splitter_context, effect_context);
 
     if (mix_result.IsError()) {
         LOG_ERROR(Audio, "Failed to update mix parameters");
@@ -194,20 +240,22 @@ void AudioRenderer::QueueMixedBuffer(Buffer::Tag tag) {
         for (std::size_t i = 0; i < BUFFER_SIZE; i++) {
             if (channel_count == 1) {
                 const auto sample = ClampToS16(mix_buffers[0][i]);
-                buffer[i * stream_channel_count + 0] = sample;
-                if (stream_channel_count > 1) {
-                    buffer[i * stream_channel_count + 1] = sample;
+
+                // Place sample in all channels
+                for (u32 channel = 0; channel < stream_channel_count; channel++) {
+                    buffer[i * stream_channel_count + channel] = sample;
                 }
+
                 if (stream_channel_count == 6) {
-                    buffer[i * stream_channel_count + 2] = sample;
-                    buffer[i * stream_channel_count + 4] = sample;
-                    buffer[i * stream_channel_count + 5] = sample;
+                    // Output stream has a LF channel, mute it!
+                    buffer[i * stream_channel_count + 3] = 0;
                 }
+
             } else if (channel_count == 2) {
                 const auto l_sample = ClampToS16(mix_buffers[0][i]);
                 const auto r_sample = ClampToS16(mix_buffers[1][i]);
                 if (stream_channel_count == 1) {
-                    buffer[i * stream_channel_count + 0] = l_sample;
+                    buffer[i * stream_channel_count + 0] = Mix2To1(l_sample, r_sample);
                 } else if (stream_channel_count == 2) {
                     buffer[i * stream_channel_count + 0] = l_sample;
                     buffer[i * stream_channel_count + 1] = r_sample;
@@ -215,8 +263,8 @@ void AudioRenderer::QueueMixedBuffer(Buffer::Tag tag) {
                     buffer[i * stream_channel_count + 0] = l_sample;
                     buffer[i * stream_channel_count + 1] = r_sample;
 
-                    buffer[i * stream_channel_count + 2] =
-                        ClampToS16((static_cast<s32>(l_sample) + static_cast<s32>(r_sample)) / 2);
+                    // Combine both left and right channels to the center channel
+                    buffer[i * stream_channel_count + 2] = Mix2To1(l_sample, r_sample);
 
                     buffer[i * stream_channel_count + 4] = l_sample;
                     buffer[i * stream_channel_count + 5] = r_sample;
@@ -231,17 +279,25 @@ void AudioRenderer::QueueMixedBuffer(Buffer::Tag tag) {
                 const auto br_sample = ClampToS16(mix_buffers[5][i]);
 
                 if (stream_channel_count == 1) {
-                    buffer[i * stream_channel_count + 0] = fc_sample;
+                    // Games seem to ignore the center channel half the time, we use the front left
+                    // and right channel for mixing as that's where majority of the audio goes
+                    buffer[i * stream_channel_count + 0] = Mix2To1(fl_sample, fr_sample);
                 } else if (stream_channel_count == 2) {
-                    buffer[i * stream_channel_count + 0] =
-                        static_cast<s16>(0.3694f * static_cast<float>(fl_sample) +
-                                         0.2612f * static_cast<float>(fc_sample) +
-                                         0.3694f * static_cast<float>(bl_sample));
-                    buffer[i * stream_channel_count + 1] =
-                        static_cast<s16>(0.3694f * static_cast<float>(fr_sample) +
-                                         0.2612f * static_cast<float>(fc_sample) +
-                                         0.3694f * static_cast<float>(br_sample));
+                    // Mix all channels into 2 channels
+                    if (sink_context.HasDownMixingCoefficients()) {
+                        const auto [left, right] = Mix6To2WithCoefficients(
+                            fl_sample, fr_sample, fc_sample, lf_sample, bl_sample, br_sample,
+                            sink_context.GetDownmixCoefficients());
+                        buffer[i * stream_channel_count + 0] = left;
+                        buffer[i * stream_channel_count + 1] = right;
+                    } else {
+                        const auto [left, right] = Mix6To2(fl_sample, fr_sample, fc_sample,
+                                                           lf_sample, bl_sample, br_sample);
+                        buffer[i * stream_channel_count + 0] = left;
+                        buffer[i * stream_channel_count + 1] = right;
+                    }
                 } else if (stream_channel_count == 6) {
+                    // Pass through
                     buffer[i * stream_channel_count + 0] = fl_sample;
                     buffer[i * stream_channel_count + 1] = fr_sample;
                     buffer[i * stream_channel_count + 2] = fc_sample;
@@ -259,7 +315,7 @@ void AudioRenderer::QueueMixedBuffer(Buffer::Tag tag) {
 }
 
 void AudioRenderer::ReleaseAndQueueBuffers() {
-    const auto released_buffers{audio_out->GetTagsAndReleaseBuffers(stream, 2)};
+    const auto released_buffers{audio_out->GetTagsAndReleaseBuffers(stream)};
     for (const auto& tag : released_buffers) {
         QueueMixedBuffer(tag);
     }

src/audio_core/audio_renderer.h

@@ -27,46 +27,35 @@ namespace Core::Timing {
 class CoreTiming;
 }
 
-namespace Kernel {
-class WritableEvent;
-}
-
 namespace Core::Memory {
 class Memory;
 }
 
 namespace AudioCore {
-using DSPStateHolder = std::array<VoiceState*, 6>;
+using DSPStateHolder = std::array<VoiceState*, AudioCommon::MAX_CHANNEL_COUNT>;
 
 class AudioOut;
 
-struct RendererInfo {
-    u64_le elasped_frame_count{};
-    INSERT_PADDING_WORDS(2);
-};
-static_assert(sizeof(RendererInfo) == 0x10, "RendererInfo is an invalid size");
-
 class AudioRenderer {
 public:
     AudioRenderer(Core::Timing::CoreTiming& core_timing, Core::Memory::Memory& memory_,
                   AudioCommon::AudioRendererParameter params,
-                  std::shared_ptr<Kernel::WritableEvent> buffer_event, std::size_t instance_number);
+                  Stream::ReleaseCallback&& release_callback, std::size_t instance_number);
     ~AudioRenderer();
 
-    ResultCode UpdateAudioRenderer(const std::vector<u8>& input_params,
-                                   std::vector<u8>& output_params);
+    [[nodiscard]] ResultCode UpdateAudioRenderer(const std::vector<u8>& input_params,
+                                                 std::vector<u8>& output_params);
     void QueueMixedBuffer(Buffer::Tag tag);
     void ReleaseAndQueueBuffers();
-    u32 GetSampleRate() const;
-    u32 GetSampleCount() const;
-    u32 GetMixBufferCount() const;
-    Stream::State GetStreamState() const;
+    [[nodiscard]] u32 GetSampleRate() const;
+    [[nodiscard]] u32 GetSampleCount() const;
+    [[nodiscard]] u32 GetMixBufferCount() const;
+    [[nodiscard]] Stream::State GetStreamState() const;
 
 private:
     BehaviorInfo behavior_info{};
 
     AudioCommon::AudioRendererParameter worker_params;
-    std::shared_ptr<Kernel::WritableEvent> buffer_event;
     std::vector<ServerMemoryPoolInfo> memory_pool_info;
     VoiceContext voice_context;
     EffectContext effect_context;
@@ -79,7 +68,6 @@ private:
     Core::Memory::Memory& memory;
     CommandGenerator command_generator;
     std::size_t elapsed_frame_count{};
-    std::vector<s32> temp_mix_buffer{};
 };
 
 } // namespace AudioCore

src/audio_core/behavior_info.h

@@ -43,22 +43,22 @@ public:
     void ClearError();
     void UpdateFlags(u64_le dest_flags);
     void SetUserRevision(u32_le revision);
-    u32_le GetUserRevision() const;
-    u32_le GetProcessRevision() const;
+    [[nodiscard]] u32_le GetUserRevision() const;
+    [[nodiscard]] u32_le GetProcessRevision() const;
 
-    bool IsAdpcmLoopContextBugFixed() const;
-    bool IsSplitterSupported() const;
-    bool IsLongSizePreDelaySupported() const;
-    bool IsAudioRendererProcessingTimeLimit80PercentSupported() const;
-    bool IsAudioRendererProcessingTimeLimit75PercentSupported() const;
-    bool IsAudioRendererProcessingTimeLimit70PercentSupported() const;
-    bool IsElapsedFrameCountSupported() const;
-    bool IsMemoryPoolForceMappingEnabled() const;
-    bool IsFlushVoiceWaveBuffersSupported() const;
-    bool IsVoicePlayedSampleCountResetAtLoopPointSupported() const;
-    bool IsVoicePitchAndSrcSkippedSupported() const;
-    bool IsMixInParameterDirtyOnlyUpdateSupported() const;
-    bool IsSplitterBugFixed() const;
+    [[nodiscard]] bool IsAdpcmLoopContextBugFixed() const;
+    [[nodiscard]] bool IsSplitterSupported() const;
+    [[nodiscard]] bool IsLongSizePreDelaySupported() const;
+    [[nodiscard]] bool IsAudioRendererProcessingTimeLimit80PercentSupported() const;
+    [[nodiscard]] bool IsAudioRendererProcessingTimeLimit75PercentSupported() const;
+    [[nodiscard]] bool IsAudioRendererProcessingTimeLimit70PercentSupported() const;
+    [[nodiscard]] bool IsElapsedFrameCountSupported() const;
+    [[nodiscard]] bool IsMemoryPoolForceMappingEnabled() const;
+    [[nodiscard]] bool IsFlushVoiceWaveBuffersSupported() const;
+    [[nodiscard]] bool IsVoicePlayedSampleCountResetAtLoopPointSupported() const;
+    [[nodiscard]] bool IsVoicePitchAndSrcSkippedSupported() const;
+    [[nodiscard]] bool IsMixInParameterDirtyOnlyUpdateSupported() const;
+    [[nodiscard]] bool IsSplitterBugFixed() const;
     void CopyErrorInfo(OutParams& dst);
 
 private:

src/audio_core/buffer.h

@@ -18,7 +18,7 @@ class Buffer {
 public:
     using Tag = u64;
 
-    Buffer(Tag tag, std::vector<s16>&& samples) : tag{tag}, samples{std::move(samples)} {}
+    Buffer(Tag tag_, std::vector<s16>&& samples_) : tag{tag_}, samples{std::move(samples_)} {}
 
     /// Returns the raw audio data for the buffer
     std::vector<s16>& GetSamples() {

src/audio_core/command_generator.cpp

@@ -2,6 +2,8 @@
 // Licensed under GPLv2 or any later version
 // Refer to the license.txt file included.
 
+#include <cmath>
+#include <numbers>
 #include "audio_core/algorithm/interpolate.h"
 #include "audio_core/command_generator.h"
 #include "audio_core/effect_context.h"
@@ -13,6 +15,20 @@ namespace AudioCore {
 namespace {
 constexpr std::size_t MIX_BUFFER_SIZE = 0x3f00;
 constexpr std::size_t SCALED_MIX_BUFFER_SIZE = MIX_BUFFER_SIZE << 15ULL;
+using DelayLineTimes = std::array<f32, AudioCommon::I3DL2REVERB_DELAY_LINE_COUNT>;
+
+constexpr DelayLineTimes FDN_MIN_DELAY_LINE_TIMES{5.0f, 6.0f, 13.0f, 14.0f};
+constexpr DelayLineTimes FDN_MAX_DELAY_LINE_TIMES{45.704f, 82.782f, 149.94f, 271.58f};
+constexpr DelayLineTimes DECAY0_MAX_DELAY_LINE_TIMES{17.0f, 13.0f, 9.0f, 7.0f};
+constexpr DelayLineTimes DECAY1_MAX_DELAY_LINE_TIMES{19.0f, 11.0f, 10.0f, 6.0f};
+constexpr std::array<f32, AudioCommon::I3DL2REVERB_TAPS> EARLY_TAP_TIMES{
+    0.017136f, 0.059154f, 0.161733f, 0.390186f, 0.425262f, 0.455411f, 0.689737f,
+    0.745910f, 0.833844f, 0.859502f, 0.000000f, 0.075024f, 0.168788f, 0.299901f,
+    0.337443f, 0.371903f, 0.599011f, 0.716741f, 0.817859f, 0.851664f};
+constexpr std::array<f32, AudioCommon::I3DL2REVERB_TAPS> EARLY_GAIN{
+    0.67096f, 0.61027f, 1.0f,     0.35680f, 0.68361f, 0.65978f, 0.51939f,
+    0.24712f, 0.45945f, 0.45021f, 0.64196f, 0.54879f, 0.92925f, 0.38270f,
+    0.72867f, 0.69794f, 0.5464f,  0.24563f, 0.45214f, 0.44042f};
 
 template <std::size_t N>
 void ApplyMix(s32* output, const s32* input, s32 gain, s32 sample_count) {
@@ -65,14 +81,162 @@ s32 ApplyMixDepop(s32* output, s32 first_sample, s32 delta, s32 sample_count) {
     }
 }
 
+float Pow10(float x) {
+    if (x >= 0.0f) {
+        return 1.0f;
+    } else if (x <= -5.3f) {
+        return 0.0f;
+    }
+    return std::pow(10.0f, x);
+}
+
+float SinD(float degrees) {
+    return std::sin(degrees * std::numbers::pi_v<float> / 180.0f);
+}
+
+float CosD(float degrees) {
+    return std::cos(degrees * std::numbers::pi_v<float> / 180.0f);
+}
+
+float ToFloat(s32 sample) {
+    return static_cast<float>(sample) / 65536.f;
+}
+
+s32 ToS32(float sample) {
+    constexpr auto min = -8388608.0f;
+    constexpr auto max = 8388607.f;
+    float rescaled_sample = sample * 65536.0f;
+    if (rescaled_sample < min) {
+        rescaled_sample = min;
+    }
+    if (rescaled_sample > max) {
+        rescaled_sample = max;
+    }
+    return static_cast<s32>(rescaled_sample);
+}
+
+constexpr std::array<std::size_t, 20> REVERB_TAP_INDEX_1CH{0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+                                                           0, 0, 0, 0, 0, 0, 0, 0, 0, 0};
+
+constexpr std::array<std::size_t, 20> REVERB_TAP_INDEX_2CH{0, 0, 0, 1, 1, 1, 1, 0, 0, 0,
+                                                           1, 1, 1, 0, 0, 0, 0, 1, 1, 1};
+
+constexpr std::array<std::size_t, 20> REVERB_TAP_INDEX_4CH{0, 0, 0, 1, 1, 1, 1, 2, 2, 2,
+                                                           1, 1, 1, 0, 0, 0, 0, 3, 3, 3};
+
+constexpr std::array<std::size_t, 20> REVERB_TAP_INDEX_6CH{4, 0, 0, 1, 1, 1, 1, 2, 2, 2,
+                                                           1, 1, 1, 0, 0, 0, 0, 3, 3, 3};
+
+template <std::size_t CHANNEL_COUNT>
+void ApplyReverbGeneric(I3dl2ReverbState& state,
+                        const std::array<const s32*, AudioCommon::MAX_CHANNEL_COUNT>& input,
+                        const std::array<s32*, AudioCommon::MAX_CHANNEL_COUNT>& output,
+                        s32 sample_count) {
+
+    auto GetTapLookup = []() {
+        if constexpr (CHANNEL_COUNT == 1) {
+            return REVERB_TAP_INDEX_1CH;
+        } else if constexpr (CHANNEL_COUNT == 2) {
+            return REVERB_TAP_INDEX_2CH;
+        } else if constexpr (CHANNEL_COUNT == 4) {
+            return REVERB_TAP_INDEX_4CH;
+        } else if constexpr (CHANNEL_COUNT == 6) {
+            return REVERB_TAP_INDEX_6CH;
+        }
+    };
+
+    const auto& tap_index_lut = GetTapLookup();
+    for (s32 sample = 0; sample < sample_count; sample++) {
+        std::array<f32, CHANNEL_COUNT> out_samples{};
+        std::array<f32, AudioCommon::I3DL2REVERB_DELAY_LINE_COUNT> fsamp{};
+        std::array<f32, AudioCommon::I3DL2REVERB_DELAY_LINE_COUNT> mixed{};
+        std::array<f32, AudioCommon::I3DL2REVERB_DELAY_LINE_COUNT> osamp{};
+
+        // Mix everything into a single sample
+        s32 temp_mixed_sample = 0;
+        for (std::size_t i = 0; i < CHANNEL_COUNT; i++) {
+            temp_mixed_sample += input[i][sample];
+        }
+        const auto current_sample = ToFloat(temp_mixed_sample);
+        const auto early_tap = state.early_delay_line.TapOut(state.early_to_late_taps);
+
+        for (std::size_t i = 0; i < AudioCommon::I3DL2REVERB_TAPS; i++) {
+            const auto tapped_samp =
+                state.early_delay_line.TapOut(state.early_tap_steps[i]) * EARLY_GAIN[i];
+            out_samples[tap_index_lut[i]] += tapped_samp;
+
+            if constexpr (CHANNEL_COUNT == 6) {
+                // handle lfe
+                out_samples[5] += tapped_samp;
+            }
+        }
+
+        state.lowpass_0 = current_sample * state.lowpass_2 + state.lowpass_0 * state.lowpass_1;
+        state.early_delay_line.Tick(state.lowpass_0);
+
+        for (std::size_t i = 0; i < CHANNEL_COUNT; i++) {
+            out_samples[i] *= state.early_gain;
+        }
+
+        // Two channel seems to apply a latet gain, we require to save this
+        f32 filter{};
+        for (std::size_t i = 0; i < AudioCommon::I3DL2REVERB_DELAY_LINE_COUNT; i++) {
+            filter = state.fdn_delay_line[i].GetOutputSample();
+            const auto computed = filter * state.lpf_coefficients[0][i] + state.shelf_filter[i];
+            state.shelf_filter[i] =
+                filter * state.lpf_coefficients[1][i] + computed * state.lpf_coefficients[2][i];
+            fsamp[i] = computed;
+        }
+
+        // Mixing matrix
+        mixed[0] = fsamp[1] + fsamp[2];
+        mixed[1] = -fsamp[0] - fsamp[3];
+        mixed[2] = fsamp[0] - fsamp[3];
+        mixed[3] = fsamp[1] - fsamp[2];
+
+        if constexpr (CHANNEL_COUNT == 2) {
+            for (auto& mix : mixed) {
+                mix *= (filter * state.late_gain);
+            }
+        }
+
+        for (std::size_t i = 0; i < AudioCommon::I3DL2REVERB_DELAY_LINE_COUNT; i++) {
+            const auto late = early_tap * state.late_gain;
+            osamp[i] = state.decay_delay_line0[i].Tick(late + mixed[i]);
+            osamp[i] = state.decay_delay_line1[i].Tick(osamp[i]);
+            state.fdn_delay_line[i].Tick(osamp[i]);
+        }
+
+        if constexpr (CHANNEL_COUNT == 1) {
+            output[0][sample] = ToS32(state.dry_gain * ToFloat(input[0][sample]) +
+                                      (out_samples[0] + osamp[0] + osamp[1]));
+        } else if constexpr (CHANNEL_COUNT == 2 || CHANNEL_COUNT == 4) {
+            for (std::size_t i = 0; i < CHANNEL_COUNT; i++) {
+                output[i][sample] =
+                    ToS32(state.dry_gain * ToFloat(input[i][sample]) + (out_samples[i] + osamp[i]));
+            }
+        } else if constexpr (CHANNEL_COUNT == 6) {
+            const auto temp_center = state.center_delay_line.Tick(0.5f * (osamp[2] - osamp[3]));
+            for (std::size_t i = 0; i < 4; i++) {
+                output[i][sample] =
+                    ToS32(state.dry_gain * ToFloat(input[i][sample]) + (out_samples[i] + osamp[i]));
+            }
+            output[4][sample] =
+                ToS32(state.dry_gain * ToFloat(input[4][sample]) + (out_samples[4] + temp_center));
+            output[5][sample] =
+                ToS32(state.dry_gain * ToFloat(input[5][sample]) + (out_samples[5] + osamp[3]));
+        }
+    }
+}
+
 } // namespace
 
-CommandGenerator::CommandGenerator(AudioCommon::AudioRendererParameter& worker_params,
-                                   VoiceContext& voice_context, MixContext& mix_context,
-                                   SplitterContext& splitter_context, EffectContext& effect_context,
-                                   Core::Memory::Memory& memory)
-    : worker_params(worker_params), voice_context(voice_context), mix_context(mix_context),
-      splitter_context(splitter_context), effect_context(effect_context), memory(memory),
+CommandGenerator::CommandGenerator(AudioCommon::AudioRendererParameter& worker_params_,
+                                   VoiceContext& voice_context_, MixContext& mix_context_,
+                                   SplitterContext& splitter_context_,
+                                   EffectContext& effect_context_, Core::Memory::Memory& memory_)
+    : worker_params(worker_params_), voice_context(voice_context_), mix_context(mix_context_),
+      splitter_context(splitter_context_), effect_context(effect_context_), memory(memory_),
       mix_buffer((worker_params.mix_buffer_count + AudioCommon::MAX_CHANNEL_COUNT) *
                  worker_params.sample_count),
       sample_buffer(MIX_BUFFER_SIZE),
@@ -255,7 +419,8 @@ void CommandGenerator::GenerateDataSourceCommand(ServerVoiceInfo& voice_info, Vo
 
 void CommandGenerator::GenerateBiquadFilterCommandForVoice(ServerVoiceInfo& voice_info,
                                                            VoiceState& dsp_state,
-                                                           s32 mix_buffer_count, s32 channel) {
+                                                           [[maybe_unused]] s32 mix_buffer_count,
+                                                           [[maybe_unused]] s32 channel) {
     for (std::size_t i = 0; i < AudioCommon::MAX_BIQUAD_FILTERS; i++) {
         const auto& in_params = voice_info.GetInParams();
         auto& biquad_filter = in_params.biquad_filter[i];
@@ -270,17 +435,19 @@ void CommandGenerator::GenerateBiquadFilterCommandForVoice(ServerVoiceInfo& voic
         }
 
         // Generate biquad filter
-        //        GenerateBiquadFilterCommand(mix_buffer_count, biquad_filter,
-        //        dsp_state.biquad_filter_state,
-        //                                    mix_buffer_count + channel, mix_buffer_count +
-        //                                    channel, worker_params.sample_count,
-        //                                    voice_info.GetInParams().node_id);
+        // GenerateBiquadFilterCommand(mix_buffer_count, biquad_filter,
+        // dsp_state.biquad_filter_state,
+        //                            mix_buffer_count + channel, mix_buffer_count + channel,
+        //                            worker_params.sample_count, voice_info.GetInParams().node_id);
     }
 }
 
-void AudioCore::CommandGenerator::GenerateBiquadFilterCommand(
-    s32 mix_buffer, const BiquadFilterParameter& params, std::array<s64, 2>& state,
-    std::size_t input_offset, std::size_t output_offset, s32 sample_count, s32 node_id) {
+void CommandGenerator::GenerateBiquadFilterCommand([[maybe_unused]] s32 mix_buffer_id,
+                                                   const BiquadFilterParameter& params,
+                                                   std::array<s64, 2>& state,
+                                                   std::size_t input_offset,
+                                                   std::size_t output_offset, s32 sample_count,
+                                                   s32 node_id) {
     if (dumping_frame) {
         LOG_DEBUG(Audio,
                   "(DSP_TRACE) GenerateBiquadFilterCommand node_id={}, "
@@ -372,17 +539,53 @@ void CommandGenerator::GenerateEffectCommand(ServerMixInfo& mix_info) {
 
 void CommandGenerator::GenerateI3dl2ReverbEffectCommand(s32 mix_buffer_offset, EffectBase* info,
                                                         bool enabled) {
-    if (!enabled) {
+    auto* reverb = dynamic_cast<EffectI3dl2Reverb*>(info);
+    const auto& params = reverb->GetParams();
+    auto& state = reverb->GetState();
+    const auto channel_count = params.channel_count;
+
+    if (channel_count != 1 && channel_count != 2 && channel_count != 4 && channel_count != 6) {
         return;
     }
-    const auto& params = dynamic_cast<EffectI3dl2Reverb*>(info)->GetParams();
-    const auto channel_count = params.channel_count;
+
+    std::array<const s32*, AudioCommon::MAX_CHANNEL_COUNT> input{};
+    std::array<s32*, AudioCommon::MAX_CHANNEL_COUNT> output{};
+
+    const auto status = params.status;
     for (s32 i = 0; i < channel_count; i++) {
-        // TODO(ogniK): Actually implement reverb
-        if (params.input[i] != params.output[i]) {
-            const auto* input = GetMixBuffer(mix_buffer_offset + params.input[i]);
-            auto* output = GetMixBuffer(mix_buffer_offset + params.output[i]);
-            ApplyMix<1>(output, input, 32768, worker_params.sample_count);
+        input[i] = GetMixBuffer(mix_buffer_offset + params.input[i]);
+        output[i] = GetMixBuffer(mix_buffer_offset + params.output[i]);
+    }
+
+    if (enabled) {
+        if (status == ParameterStatus::Initialized) {
+            InitializeI3dl2Reverb(reverb->GetParams(), state, info->GetWorkBuffer());
+        } else if (status == ParameterStatus::Updating) {
+            UpdateI3dl2Reverb(reverb->GetParams(), state, false);
+        }
+    }
+
+    if (enabled) {
+        switch (channel_count) {
+        case 1:
+            ApplyReverbGeneric<1>(state, input, output, worker_params.sample_count);
+            break;
+        case 2:
+            ApplyReverbGeneric<2>(state, input, output, worker_params.sample_count);
+            break;
+        case 4:
+            ApplyReverbGeneric<4>(state, input, output, worker_params.sample_count);
+            break;
+        case 6:
+            ApplyReverbGeneric<6>(state, input, output, worker_params.sample_count);
+            break;
+        }
+    } else {
+        for (s32 i = 0; i < channel_count; i++) {
+            // Only copy if the buffer input and output do not match!
+            if ((mix_buffer_offset + params.input[i]) != (mix_buffer_offset + params.output[i])) {
+                std::memcpy(output[i], input[i], worker_params.sample_count * sizeof(s32));
+            }
         }
     }
 }
@@ -521,6 +724,133 @@ s32 CommandGenerator::ReadAuxBuffer(AuxInfoDSP& recv_info, VAddr recv_buffer, u3
     return sample_count;
 }
 
+void CommandGenerator::InitializeI3dl2Reverb(I3dl2ReverbParams& info, I3dl2ReverbState& state,
+                                             std::vector<u8>& work_buffer) {
+    // Reset state
+    state.lowpass_0 = 0.0f;
+    state.lowpass_1 = 0.0f;
+    state.lowpass_2 = 0.0f;
+
+    state.early_delay_line.Reset();
+    state.early_tap_steps.fill(0);
+    state.early_gain = 0.0f;
+    state.late_gain = 0.0f;
+    state.early_to_late_taps = 0;
+    for (std::size_t i = 0; i < AudioCommon::I3DL2REVERB_DELAY_LINE_COUNT; i++) {
+        state.fdn_delay_line[i].Reset();
+        state.decay_delay_line0[i].Reset();
+        state.decay_delay_line1[i].Reset();
+    }
+    state.last_reverb_echo = 0.0f;
+    state.center_delay_line.Reset();
+    for (auto& coef : state.lpf_coefficients) {
+        coef.fill(0.0f);
+    }
+    state.shelf_filter.fill(0.0f);
+    state.dry_gain = 0.0f;
+
+    const auto sample_rate = info.sample_rate / 1000;
+    f32* work_buffer_ptr = reinterpret_cast<f32*>(work_buffer.data());
+
+    s32 delay_samples{};
+    for (std::size_t i = 0; i < AudioCommon::I3DL2REVERB_DELAY_LINE_COUNT; i++) {
+        delay_samples =
+            AudioCommon::CalculateDelaySamples(sample_rate, FDN_MAX_DELAY_LINE_TIMES[i]);
+        state.fdn_delay_line[i].Initialize(delay_samples, work_buffer_ptr);
+        work_buffer_ptr += delay_samples + 1;
+
+        delay_samples =
+            AudioCommon::CalculateDelaySamples(sample_rate, DECAY0_MAX_DELAY_LINE_TIMES[i]);
+        state.decay_delay_line0[i].Initialize(delay_samples, 0.0f, work_buffer_ptr);
+        work_buffer_ptr += delay_samples + 1;
+
+        delay_samples =
+            AudioCommon::CalculateDelaySamples(sample_rate, DECAY1_MAX_DELAY_LINE_TIMES[i]);
+        state.decay_delay_line1[i].Initialize(delay_samples, 0.0f, work_buffer_ptr);
+        work_buffer_ptr += delay_samples + 1;
+    }
+    delay_samples = AudioCommon::CalculateDelaySamples(sample_rate, 5.0f);
+    state.center_delay_line.Initialize(delay_samples, work_buffer_ptr);
+    work_buffer_ptr += delay_samples + 1;
+
+    delay_samples = AudioCommon::CalculateDelaySamples(sample_rate, 400.0f);
+    state.early_delay_line.Initialize(delay_samples, work_buffer_ptr);
+
+    UpdateI3dl2Reverb(info, state, true);
+}
+
+void CommandGenerator::UpdateI3dl2Reverb(I3dl2ReverbParams& info, I3dl2ReverbState& state,
+                                         bool should_clear) {
+
+    state.dry_gain = info.dry_gain;
+    state.shelf_filter.fill(0.0f);
+    state.lowpass_0 = 0.0f;
+    state.early_gain = Pow10(std::min(info.room + info.reflection, 5000.0f) / 2000.0f);
+    state.late_gain = Pow10(std::min(info.room + info.reverb, 5000.0f) / 2000.0f);
+
+    const auto sample_rate = info.sample_rate / 1000;
+    const f32 hf_gain = Pow10(info.room_hf / 2000.0f);
+    if (hf_gain >= 1.0f) {
+        state.lowpass_2 = 1.0f;
+        state.lowpass_1 = 0.0f;
+    } else {
+        const auto a = 1.0f - hf_gain;
+        const auto b = 2.0f * (1.0f - hf_gain * CosD(256.0f * info.hf_reference /
+                                                     static_cast<f32>(info.sample_rate)));
+        const auto c = std::sqrt(b * b - 4.0f * a * a);
+
+        state.lowpass_1 = (b - c) / (2.0f * a);
+        state.lowpass_2 = 1.0f - state.lowpass_1;
+    }
+    state.early_to_late_taps = AudioCommon::CalculateDelaySamples(
+        sample_rate, 1000.0f * (info.reflection_delay + info.reverb_delay));
+
+    state.last_reverb_echo = 0.6f * info.diffusion * 0.01f;
+    for (std::size_t i = 0; i < AudioCommon::I3DL2REVERB_DELAY_LINE_COUNT; i++) {
+        const auto length =
+            FDN_MIN_DELAY_LINE_TIMES[i] +
+            (info.density / 100.0f) * (FDN_MAX_DELAY_LINE_TIMES[i] - FDN_MIN_DELAY_LINE_TIMES[i]);
+        state.fdn_delay_line[i].SetDelay(AudioCommon::CalculateDelaySamples(sample_rate, length));
+
+        const auto delay_sample_counts = state.fdn_delay_line[i].GetDelay() +
+                                         state.decay_delay_line0[i].GetDelay() +
+                                         state.decay_delay_line1[i].GetDelay();
+
+        float a = (-60.0f * static_cast<f32>(delay_sample_counts)) /
+                  (info.decay_time * static_cast<f32>(info.sample_rate));
+        float b = a / info.hf_decay_ratio;
+        float c = CosD(128.0f * 0.5f * info.hf_reference / static_cast<f32>(info.sample_rate)) /
+                  SinD(128.0f * 0.5f * info.hf_reference / static_cast<f32>(info.sample_rate));
+        float d = Pow10((b - a) / 40.0f);
+        float e = Pow10((b + a) / 40.0f) * 0.7071f;
+
+        state.lpf_coefficients[0][i] = e * ((d * c) + 1.0f) / (c + d);
+        state.lpf_coefficients[1][i] = e * (1.0f - (d * c)) / (c + d);
+        state.lpf_coefficients[2][i] = (c - d) / (c + d);
+
+        state.decay_delay_line0[i].SetCoefficient(state.last_reverb_echo);
+        state.decay_delay_line1[i].SetCoefficient(-0.9f * state.last_reverb_echo);
+    }
+
+    if (should_clear) {
+        for (std::size_t i = 0; i < AudioCommon::I3DL2REVERB_DELAY_LINE_COUNT; i++) {
+            state.fdn_delay_line[i].Clear();
+            state.decay_delay_line0[i].Clear();
+            state.decay_delay_line1[i].Clear();
+        }
+        state.early_delay_line.Clear();
+        state.center_delay_line.Clear();
+    }
+
+    const auto max_early_delay = state.early_delay_line.GetMaxDelay();
+    const auto reflection_time = 1000.0f * (0.0098f * info.reverb_delay + 0.02f);
+    for (std::size_t tap = 0; tap < AudioCommon::I3DL2REVERB_TAPS; tap++) {
+        const auto length = AudioCommon::CalculateDelaySamples(
+            sample_rate, 1000.0f * info.reflection_delay + reflection_time * EARLY_TAP_TIMES[tap]);
+        state.early_tap_steps[tap] = std::min(length, max_early_delay);
+    }
+}
+
 void CommandGenerator::GenerateVolumeRampCommand(float last_volume, float current_volume,
                                                  s32 channel, s32 node_id) {
     const auto last = static_cast<s32>(last_volume * 32768.0f);
@@ -714,7 +1044,8 @@ s32 CommandGenerator::DecodePcm16(ServerVoiceInfo& voice_info, VoiceState& dsp_s
 }
 
 s32 CommandGenerator::DecodeAdpcm(ServerVoiceInfo& voice_info, VoiceState& dsp_state,
-                                  s32 sample_count, s32 channel, std::size_t mix_offset) {
+                                  s32 sample_count, [[maybe_unused]] s32 channel,
+                                  std::size_t mix_offset) {
     const auto& in_params = voice_info.GetInParams();
     const auto& wave_buffer = in_params.wave_buffer[dsp_state.wave_buffer_index];
     if (wave_buffer.buffer_address == 0) {

src/audio_core/command_generator.h

@@ -21,14 +21,16 @@ class ServerMixInfo;
 class EffectContext;
 class EffectBase;
 struct AuxInfoDSP;
+struct I3dl2ReverbParams;
+struct I3dl2ReverbState;
 using MixVolumeBuffer = std::array<float, AudioCommon::MAX_MIX_BUFFERS>;
 
 class CommandGenerator {
 public:
-    explicit CommandGenerator(AudioCommon::AudioRendererParameter& worker_params,
-                              VoiceContext& voice_context, MixContext& mix_context,
-                              SplitterContext& splitter_context, EffectContext& effect_context,
-                              Core::Memory::Memory& memory);
+    explicit CommandGenerator(AudioCommon::AudioRendererParameter& worker_params_,
+                              VoiceContext& voice_context_, MixContext& mix_context_,
+                              SplitterContext& splitter_context_, EffectContext& effect_context_,
+                              Core::Memory::Memory& memory_);
     ~CommandGenerator();
 
     void ClearMixBuffers();
@@ -39,13 +41,13 @@ public:
     void PreCommand();
     void PostCommand();
 
-    s32* GetChannelMixBuffer(s32 channel);
-    const s32* GetChannelMixBuffer(s32 channel) const;
-    s32* GetMixBuffer(std::size_t index);
-    const s32* GetMixBuffer(std::size_t index) const;
-    std::size_t GetMixChannelBufferOffset(s32 channel) const;
+    [[nodiscard]] s32* GetChannelMixBuffer(s32 channel);
+    [[nodiscard]] const s32* GetChannelMixBuffer(s32 channel) const;
+    [[nodiscard]] s32* GetMixBuffer(std::size_t index);
+    [[nodiscard]] const s32* GetMixBuffer(std::size_t index) const;
+    [[nodiscard]] std::size_t GetMixChannelBufferOffset(s32 channel) const;
 
-    std::size_t GetTotalMixBufferCount() const;
+    [[nodiscard]] std::size_t GetTotalMixBufferCount() const;
 
 private:
     void GenerateDataSourceCommand(ServerVoiceInfo& voice_info, VoiceState& dsp_state, s32 channel);
@@ -73,13 +75,16 @@ private:
     void GenerateI3dl2ReverbEffectCommand(s32 mix_buffer_offset, EffectBase* info, bool enabled);
     void GenerateBiquadFilterEffectCommand(s32 mix_buffer_offset, EffectBase* info, bool enabled);
     void GenerateAuxCommand(s32 mix_buffer_offset, EffectBase* info, bool enabled);
-    ServerSplitterDestinationData* GetDestinationData(s32 splitter_id, s32 index);
+    [[nodiscard]] ServerSplitterDestinationData* GetDestinationData(s32 splitter_id, s32 index);
 
     s32 WriteAuxBuffer(AuxInfoDSP& dsp_info, VAddr send_buffer, u32 max_samples, const s32* data,
                        u32 sample_count, u32 write_offset, u32 write_count);
     s32 ReadAuxBuffer(AuxInfoDSP& recv_info, VAddr recv_buffer, u32 max_samples, s32* out_data,
                       u32 sample_count, u32 read_offset, u32 read_count);
 
+    void InitializeI3dl2Reverb(I3dl2ReverbParams& info, I3dl2ReverbState& state,
+                               std::vector<u8>& work_buffer);
+    void UpdateI3dl2Reverb(I3dl2ReverbParams& info, I3dl2ReverbState& state, bool should_clear);
     // DSP Code
     s32 DecodePcm16(ServerVoiceInfo& voice_info, VoiceState& dsp_state, s32 sample_count,
                     s32 channel, std::size_t mix_offset);

src/audio_core/common.h

@@ -22,7 +22,7 @@ constexpr std::size_t MAX_CHANNEL_COUNT = 6;
 constexpr std::size_t MAX_WAVE_BUFFERS = 4;
 constexpr std::size_t MAX_SAMPLE_HISTORY = 4;
 constexpr u32 STREAM_SAMPLE_RATE = 48000;
-constexpr u32 STREAM_NUM_CHANNELS = 6;
+constexpr u32 STREAM_NUM_CHANNELS = 2;
 constexpr s32 NO_SPLITTER = -1;
 constexpr s32 NO_MIX = 0x7fffffff;
 constexpr s32 NO_FINAL_MIX = std::numeric_limits<s32>::min();
@@ -33,6 +33,29 @@ constexpr std::size_t TEMP_MIX_BASE_SIZE = 0x3f00; // TODO(ogniK): Work out this
 // and our const ends up being 0x3f04, the 4 bytes are most
 // likely the sample history
 constexpr std::size_t TOTAL_TEMP_MIX_SIZE = TEMP_MIX_BASE_SIZE + AudioCommon::MAX_SAMPLE_HISTORY;
+constexpr f32 I3DL2REVERB_MAX_LEVEL = 5000.0f;
+constexpr f32 I3DL2REVERB_MIN_REFLECTION_DURATION = 0.02f;
+constexpr std::size_t I3DL2REVERB_TAPS = 20;
+constexpr std::size_t I3DL2REVERB_DELAY_LINE_COUNT = 4;
+using Fractional = s32;
+
+template <typename T>
+constexpr Fractional ToFractional(T x) {
+    return static_cast<Fractional>(x * static_cast<T>(0x4000));
+}
+
+constexpr Fractional MultiplyFractional(Fractional lhs, Fractional rhs) {
+    return static_cast<Fractional>(static_cast<s64>(lhs) * rhs >> 14);
+}
+
+constexpr s32 FractionalToFixed(Fractional x) {
+    const auto s = x & (1 << 13);
+    return static_cast<s32>(x >> 14) + s;
+}
+
+constexpr s32 CalculateDelaySamples(s32 sample_rate_khz, float time) {
+    return FractionalToFixed(MultiplyFractional(ToFractional(sample_rate_khz), ToFractional(time)));
+}
 
 static constexpr u32 VersionFromRevision(u32_le rev) {
     // "REV7" -> 7

src/audio_core/cubeb_sink.cpp

@@ -21,15 +21,16 @@ namespace AudioCore {
 
 class CubebSinkStream final : public SinkStream {
 public:
-    CubebSinkStream(cubeb* ctx, u32 sample_rate, u32 num_channels_, cubeb_devid output_device,
+    CubebSinkStream(cubeb* ctx_, u32 sample_rate, u32 num_channels_, cubeb_devid output_device,
                     const std::string& name)
-        : ctx{ctx}, num_channels{std::min(num_channels_, 6u)}, time_stretch{sample_rate,
-                                                                            num_channels} {
+        : ctx{ctx_}, num_channels{std::min(num_channels_, 6u)}, time_stretch{sample_rate,
+                                                                             num_channels} {
 
         cubeb_stream_params params{};
         params.rate = sample_rate;
         params.channels = num_channels;
         params.format = CUBEB_SAMPLE_S16NE;
+        params.prefs = CUBEB_STREAM_PREF_PERSIST;
         switch (num_channels) {
         case 1:
             params.layout = CUBEB_LAYOUT_MONO;
@@ -192,8 +193,9 @@ SinkStream& CubebSink::AcquireSinkStream(u32 sample_rate, u32 num_channels,
     return *sink_streams.back();
 }
 
-long CubebSinkStream::DataCallback(cubeb_stream* stream, void* user_data, const void* input_buffer,
-                                   void* output_buffer, long num_frames) {
+long CubebSinkStream::DataCallback([[maybe_unused]] cubeb_stream* stream, void* user_data,
+                                   [[maybe_unused]] const void* input_buffer, void* output_buffer,
+                                   long num_frames) {
     auto* impl = static_cast<CubebSinkStream*>(user_data);
     auto* buffer = static_cast<u8*>(output_buffer);
 
@@ -236,7 +238,9 @@ long CubebSinkStream::DataCallback(cubeb_stream* stream, void* user_data, const
     return num_frames;
 }
 
-void CubebSinkStream::StateCallback(cubeb_stream* stream, void* user_data, cubeb_state state) {}
+void CubebSinkStream::StateCallback([[maybe_unused]] cubeb_stream* stream,
+                                    [[maybe_unused]] void* user_data,
+                                    [[maybe_unused]] cubeb_state state) {}
 
 std::vector<std::string> ListCubebSinkDevices() {
     std::vector<std::string> device_list;

src/audio_core/delay_line.cpp

@@ -0,0 +1,104 @@
+#include <cstring>
+#include "audio_core/delay_line.h"
+
+namespace AudioCore {
+DelayLineBase::DelayLineBase() = default;
+DelayLineBase::~DelayLineBase() = default;
+
+void DelayLineBase::Initialize(s32 max_delay_, float* src_buffer) {
+    buffer = src_buffer;
+    buffer_end = buffer + max_delay_;
+    max_delay = max_delay_;
+    output = buffer;
+    SetDelay(max_delay_);
+    Clear();
+}
+
+void DelayLineBase::SetDelay(s32 new_delay) {
+    if (max_delay < new_delay) {
+        return;
+    }
+    delay = new_delay;
+    input = (buffer + ((output - buffer) + new_delay) % (max_delay + 1));
+}
+
+s32 DelayLineBase::GetDelay() const {
+    return delay;
+}
+
+s32 DelayLineBase::GetMaxDelay() const {
+    return max_delay;
+}
+
+f32 DelayLineBase::TapOut(s32 last_sample) {
+    const float* ptr = input - (last_sample + 1);
+    if (ptr < buffer) {
+        ptr += (max_delay + 1);
+    }
+
+    return *ptr;
+}
+
+f32 DelayLineBase::Tick(f32 sample) {
+    *(input++) = sample;
+    const auto out_sample = *(output++);
+
+    if (buffer_end < input) {
+        input = buffer;
+    }
+
+    if (buffer_end < output) {
+        output = buffer;
+    }
+
+    return out_sample;
+}
+
+float* DelayLineBase::GetInput() {
+    return input;
+}
+
+const float* DelayLineBase::GetInput() const {
+    return input;
+}
+
+f32 DelayLineBase::GetOutputSample() const {
+    return *output;
+}
+
+void DelayLineBase::Clear() {
+    std::memset(buffer, 0, sizeof(float) * max_delay);
+}
+
+void DelayLineBase::Reset() {
+    buffer = nullptr;
+    buffer_end = nullptr;
+    max_delay = 0;
+    input = nullptr;
+    output = nullptr;
+    delay = 0;
+}
+
+DelayLineAllPass::DelayLineAllPass() = default;
+DelayLineAllPass::~DelayLineAllPass() = default;
+
+void DelayLineAllPass::Initialize(u32 delay_, float coeffcient_, f32* src_buffer) {
+    DelayLineBase::Initialize(delay_, src_buffer);
+    SetCoefficient(coeffcient_);
+}
+
+void DelayLineAllPass::SetCoefficient(float coeffcient_) {
+    coefficient = coeffcient_;
+}
+
+f32 DelayLineAllPass::Tick(f32 sample) {
+    const auto temp = sample - coefficient * *output;
+    return coefficient * temp + DelayLineBase::Tick(temp);
+}
+
+void DelayLineAllPass::Reset() {
+    coefficient = 0.0f;
+    DelayLineBase::Reset();
+}
+
+} // namespace AudioCore

src/audio_core/delay_line.h

@@ -0,0 +1,46 @@
+#pragma once
+
+#include "common/common_types.h"
+
+namespace AudioCore {
+
+class DelayLineBase {
+public:
+    DelayLineBase();
+    ~DelayLineBase();
+
+    void Initialize(s32 max_delay_, float* src_buffer);
+    void SetDelay(s32 new_delay);
+    s32 GetDelay() const;
+    s32 GetMaxDelay() const;
+    f32 TapOut(s32 last_sample);
+    f32 Tick(f32 sample);
+    float* GetInput();
+    const float* GetInput() const;
+    f32 GetOutputSample() const;
+    void Clear();
+    void Reset();
+
+protected:
+    float* buffer{nullptr};
+    float* buffer_end{nullptr};
+    s32 max_delay{};
+    float* input{nullptr};
+    float* output{nullptr};
+    s32 delay{};
+};
+
+class DelayLineAllPass final : public DelayLineBase {
+public:
+    DelayLineAllPass();
+    ~DelayLineAllPass();
+
+    void Initialize(u32 delay, float coeffcient_, f32* src_buffer);
+    void SetCoefficient(float coeffcient_);
+    f32 Tick(f32 sample);
+    void Reset();
+
+private:
+    float coefficient{};
+};
+} // namespace AudioCore

src/audio_core/effect_context.cpp

@@ -12,7 +12,7 @@ bool ValidChannelCountForEffect(s32 channel_count) {
 }
 } // namespace
 
-EffectContext::EffectContext(std::size_t effect_count) : effect_count(effect_count) {
+EffectContext::EffectContext(std::size_t effect_count_) : effect_count(effect_count_) {
     effects.reserve(effect_count);
     std::generate_n(std::back_inserter(effects), effect_count,
                     [] { return std::make_unique<EffectStubbed>(); });
@@ -61,13 +61,13 @@ const EffectBase* EffectContext::GetInfo(std::size_t i) const {
     return effects.at(i).get();
 }
 
-EffectStubbed::EffectStubbed() : EffectBase::EffectBase(EffectType::Invalid) {}
+EffectStubbed::EffectStubbed() : EffectBase(EffectType::Invalid) {}
 EffectStubbed::~EffectStubbed() = default;
 
-void EffectStubbed::Update(EffectInfo::InParams& in_params) {}
+void EffectStubbed::Update([[maybe_unused]] EffectInfo::InParams& in_params) {}
 void EffectStubbed::UpdateForCommandGeneration() {}
 
-EffectBase::EffectBase(EffectType effect_type) : effect_type(effect_type) {}
+EffectBase::EffectBase(EffectType effect_type_) : effect_type(effect_type_) {}
 EffectBase::~EffectBase() = default;
 
 UsageState EffectBase::GetUsage() const {
@@ -90,33 +90,47 @@ s32 EffectBase::GetProcessingOrder() const {
     return processing_order;
 }
 
-EffectI3dl2Reverb::EffectI3dl2Reverb() : EffectGeneric::EffectGeneric(EffectType::I3dl2Reverb) {}
+std::vector<u8>& EffectBase::GetWorkBuffer() {
+    return work_buffer;
+}
+
+const std::vector<u8>& EffectBase::GetWorkBuffer() const {
+    return work_buffer;
+}
+
+EffectI3dl2Reverb::EffectI3dl2Reverb() : EffectGeneric(EffectType::I3dl2Reverb) {}
 EffectI3dl2Reverb::~EffectI3dl2Reverb() = default;
 
 void EffectI3dl2Reverb::Update(EffectInfo::InParams& in_params) {
-    auto& internal_params = GetParams();
+    auto& params = GetParams();
     const auto* reverb_params = reinterpret_cast<I3dl2ReverbParams*>(in_params.raw.data());
     if (!ValidChannelCountForEffect(reverb_params->max_channels)) {
         UNREACHABLE_MSG("Invalid reverb max channel count {}", reverb_params->max_channels);
         return;
     }
 
-    const auto last_status = internal_params.status;
+    const auto last_status = params.status;
     mix_id = in_params.mix_id;
     processing_order = in_params.processing_order;
-    internal_params = *reverb_params;
+    params = *reverb_params;
     if (!ValidChannelCountForEffect(reverb_params->channel_count)) {
-        internal_params.channel_count = internal_params.max_channels;
+        params.channel_count = params.max_channels;
     }
     enabled = in_params.is_enabled;
     if (last_status != ParameterStatus::Updated) {
-        internal_params.status = last_status;
+        params.status = last_status;
     }
 
     if (in_params.is_new || skipped) {
         usage = UsageState::Initialized;
-        internal_params.status = ParameterStatus::Initialized;
+        params.status = ParameterStatus::Initialized;
         skipped = in_params.buffer_address == 0 || in_params.buffer_size == 0;
+        if (!skipped) {
+            auto& cur_work_buffer = GetWorkBuffer();
+            // Has two buffers internally
+            cur_work_buffer.resize(in_params.buffer_size * 2);
+            std::fill(cur_work_buffer.begin(), cur_work_buffer.end(), 0);
+        }
     }
 }
 
@@ -129,15 +143,23 @@ void EffectI3dl2Reverb::UpdateForCommandGeneration() {
     GetParams().status = ParameterStatus::Updated;
 }
 
-EffectBiquadFilter::EffectBiquadFilter() : EffectGeneric::EffectGeneric(EffectType::BiquadFilter) {}
+I3dl2ReverbState& EffectI3dl2Reverb::GetState() {
+    return state;
+}
+
+const I3dl2ReverbState& EffectI3dl2Reverb::GetState() const {
+    return state;
+}
+
+EffectBiquadFilter::EffectBiquadFilter() : EffectGeneric(EffectType::BiquadFilter) {}
 EffectBiquadFilter::~EffectBiquadFilter() = default;
 
 void EffectBiquadFilter::Update(EffectInfo::InParams& in_params) {
-    auto& internal_params = GetParams();
+    auto& params = GetParams();
     const auto* biquad_params = reinterpret_cast<BiquadFilterParams*>(in_params.raw.data());
     mix_id = in_params.mix_id;
     processing_order = in_params.processing_order;
-    internal_params = *biquad_params;
+    params = *biquad_params;
     enabled = in_params.is_enabled;
 }
 
@@ -150,7 +172,7 @@ void EffectBiquadFilter::UpdateForCommandGeneration() {
     GetParams().status = ParameterStatus::Updated;
 }
 
-EffectAuxInfo::EffectAuxInfo() : EffectGeneric::EffectGeneric(EffectType::Aux) {}
+EffectAuxInfo::EffectAuxInfo() : EffectGeneric(EffectType::Aux) {}
 EffectAuxInfo::~EffectAuxInfo() = default;
 
 void EffectAuxInfo::Update(EffectInfo::InParams& in_params) {
@@ -200,32 +222,32 @@ VAddr EffectAuxInfo::GetRecvBuffer() const {
     return recv_buffer;
 }
 
-EffectDelay::EffectDelay() : EffectGeneric::EffectGeneric(EffectType::Delay) {}
+EffectDelay::EffectDelay() : EffectGeneric(EffectType::Delay) {}
 EffectDelay::~EffectDelay() = default;
 
 void EffectDelay::Update(EffectInfo::InParams& in_params) {
     const auto* delay_params = reinterpret_cast<DelayParams*>(in_params.raw.data());
-    auto& internal_params = GetParams();
+    auto& params = GetParams();
     if (!ValidChannelCountForEffect(delay_params->max_channels)) {
         return;
     }
 
-    const auto last_status = internal_params.status;
+    const auto last_status = params.status;
     mix_id = in_params.mix_id;
     processing_order = in_params.processing_order;
-    internal_params = *delay_params;
+    params = *delay_params;
     if (!ValidChannelCountForEffect(delay_params->channels)) {
-        internal_params.channels = internal_params.max_channels;
+        params.channels = params.max_channels;
     }
     enabled = in_params.is_enabled;
 
     if (last_status != ParameterStatus::Updated) {
-        internal_params.status = last_status;
+        params.status = last_status;
     }
 
     if (in_params.is_new || skipped) {
         usage = UsageState::Initialized;
-        internal_params.status = ParameterStatus::Initialized;
+        params.status = ParameterStatus::Initialized;
         skipped = in_params.buffer_address == 0 || in_params.buffer_size == 0;
     }
 }
@@ -239,7 +261,7 @@ void EffectDelay::UpdateForCommandGeneration() {
     GetParams().status = ParameterStatus::Updated;
 }
 
-EffectBufferMixer::EffectBufferMixer() : EffectGeneric::EffectGeneric(EffectType::BufferMixer) {}
+EffectBufferMixer::EffectBufferMixer() : EffectGeneric(EffectType::BufferMixer) {}
 EffectBufferMixer::~EffectBufferMixer() = default;
 
 void EffectBufferMixer::Update(EffectInfo::InParams& in_params) {
@@ -257,32 +279,32 @@ void EffectBufferMixer::UpdateForCommandGeneration() {
     }
 }
 
-EffectReverb::EffectReverb() : EffectGeneric::EffectGeneric(EffectType::Reverb) {}
+EffectReverb::EffectReverb() : EffectGeneric(EffectType::Reverb) {}
 EffectReverb::~EffectReverb() = default;
 
 void EffectReverb::Update(EffectInfo::InParams& in_params) {
     const auto* reverb_params = reinterpret_cast<ReverbParams*>(in_params.raw.data());
-    auto& internal_params = GetParams();
+    auto& params = GetParams();
     if (!ValidChannelCountForEffect(reverb_params->max_channels)) {
         return;
     }
 
-    const auto last_status = internal_params.status;
+    const auto last_status = params.status;
     mix_id = in_params.mix_id;
     processing_order = in_params.processing_order;
-    internal_params = *reverb_params;
+    params = *reverb_params;
     if (!ValidChannelCountForEffect(reverb_params->channels)) {
-        internal_params.channels = internal_params.max_channels;
+        params.channels = params.max_channels;
     }
     enabled = in_params.is_enabled;
 
     if (last_status != ParameterStatus::Updated) {
-        internal_params.status = last_status;
+        params.status = last_status;
     }
 
     if (in_params.is_new || skipped) {
         usage = UsageState::Initialized;
-        internal_params.status = ParameterStatus::Initialized;
+        params.status = ParameterStatus::Initialized;
         skipped = in_params.buffer_address == 0 || in_params.buffer_size == 0;
     }
 }

src/audio_core/effect_context.h

@@ -8,6 +8,7 @@
 #include <memory>
 #include <vector>
 #include "audio_core/common.h"
+#include "audio_core/delay_line.h"
 #include "common/common_funcs.h"
 #include "common/common_types.h"
 #include "common/swap.h"
@@ -184,16 +185,18 @@ struct AuxAddress {
 
 class EffectBase {
 public:
-    explicit EffectBase(EffectType effect_type);
+    explicit EffectBase(EffectType effect_type_);
     virtual ~EffectBase();
 
     virtual void Update(EffectInfo::InParams& in_params) = 0;
     virtual void UpdateForCommandGeneration() = 0;
-    UsageState GetUsage() const;
-    EffectType GetType() const;
-    bool IsEnabled() const;
-    s32 GetMixID() const;
-    s32 GetProcessingOrder() const;
+    [[nodiscard]] UsageState GetUsage() const;
+    [[nodiscard]] EffectType GetType() const;
+    [[nodiscard]] bool IsEnabled() const;
+    [[nodiscard]] s32 GetMixID() const;
+    [[nodiscard]] s32 GetProcessingOrder() const;
+    [[nodiscard]] std::vector<u8>& GetWorkBuffer();
+    [[nodiscard]] const std::vector<u8>& GetWorkBuffer() const;
 
 protected:
     UsageState usage{UsageState::Invalid};
@@ -201,18 +204,19 @@ protected:
     s32 mix_id{};
     s32 processing_order{};
     bool enabled = false;
+    std::vector<u8> work_buffer{};
 };
 
 template <typename T>
 class EffectGeneric : public EffectBase {
 public:
-    explicit EffectGeneric(EffectType effect_type) : EffectBase(effect_type) {}
+    explicit EffectGeneric(EffectType effect_type_) : EffectBase(effect_type_) {}
 
     T& GetParams() {
         return internal_params;
     }
 
-    const I3dl2ReverbParams& GetParams() const {
+    const T& GetParams() const {
         return internal_params;
     }
 
@@ -229,6 +233,27 @@ public:
     void UpdateForCommandGeneration() override;
 };
 
+struct I3dl2ReverbState {
+    f32 lowpass_0{};
+    f32 lowpass_1{};
+    f32 lowpass_2{};
+
+    DelayLineBase early_delay_line{};
+    std::array<u32, AudioCommon::I3DL2REVERB_TAPS> early_tap_steps{};
+    f32 early_gain{};
+    f32 late_gain{};
+
+    u32 early_to_late_taps{};
+    std::array<DelayLineBase, AudioCommon::I3DL2REVERB_DELAY_LINE_COUNT> fdn_delay_line{};
+    std::array<DelayLineAllPass, AudioCommon::I3DL2REVERB_DELAY_LINE_COUNT> decay_delay_line0{};
+    std::array<DelayLineAllPass, AudioCommon::I3DL2REVERB_DELAY_LINE_COUNT> decay_delay_line1{};
+    f32 last_reverb_echo{};
+    DelayLineBase center_delay_line{};
+    std::array<std::array<f32, AudioCommon::I3DL2REVERB_DELAY_LINE_COUNT>, 3> lpf_coefficients{};
+    std::array<f32, AudioCommon::I3DL2REVERB_DELAY_LINE_COUNT> shelf_filter{};
+    f32 dry_gain{};
+};
+
 class EffectI3dl2Reverb : public EffectGeneric<I3dl2ReverbParams> {
 public:
     explicit EffectI3dl2Reverb();
@@ -237,8 +262,12 @@ public:
     void Update(EffectInfo::InParams& in_params) override;
     void UpdateForCommandGeneration() override;
 
+    I3dl2ReverbState& GetState();
+    const I3dl2ReverbState& GetState() const;
+
 private:
     bool skipped = false;
+    I3dl2ReverbState state{};
 };
 
 class EffectBiquadFilter : public EffectGeneric<BiquadFilterParams> {
@@ -257,10 +286,10 @@ public:
 
     void Update(EffectInfo::InParams& in_params) override;
     void UpdateForCommandGeneration() override;
-    VAddr GetSendInfo() const;
-    VAddr GetSendBuffer() const;
-    VAddr GetRecvInfo() const;
-    VAddr GetRecvBuffer() const;
+    [[nodiscard]] VAddr GetSendInfo() const;
+    [[nodiscard]] VAddr GetSendBuffer() const;
+    [[nodiscard]] VAddr GetRecvInfo() const;
+    [[nodiscard]] VAddr GetRecvBuffer() const;
 
 private:
     VAddr send_info{};
@@ -306,13 +335,13 @@ private:
 
 class EffectContext {
 public:
-    explicit EffectContext(std::size_t effect_count);
+    explicit EffectContext(std::size_t effect_count_);
     ~EffectContext();
 
-    std::size_t GetCount() const;
-    EffectBase* GetInfo(std::size_t i);
-    EffectBase* RetargetEffect(std::size_t i, EffectType effect);
-    const EffectBase* GetInfo(std::size_t i) const;
+    [[nodiscard]] std::size_t GetCount() const;
+    [[nodiscard]] EffectBase* GetInfo(std::size_t i);
+    [[nodiscard]] EffectBase* RetargetEffect(std::size_t i, EffectType effect);
+    [[nodiscard]] const EffectBase* GetInfo(std::size_t i) const;
 
 private:
     std::size_t effect_count{};

src/audio_core/info_updater.cpp

@@ -14,9 +14,9 @@
 
 namespace AudioCore {
 
-InfoUpdater::InfoUpdater(const std::vector<u8>& in_params, std::vector<u8>& out_params,
-                         BehaviorInfo& behavior_info)
-    : in_params(in_params), out_params(out_params), behavior_info(behavior_info) {
+InfoUpdater::InfoUpdater(const std::vector<u8>& in_params_, std::vector<u8>& out_params_,
+                         BehaviorInfo& behavior_info_)
+    : in_params(in_params_), out_params(out_params_), behavior_info(behavior_info_) {
     ASSERT(
         AudioCommon::CanConsumeBuffer(in_params.size(), 0, sizeof(AudioCommon::UpdateDataHeader)));
     std::memcpy(&input_header, in_params.data(), sizeof(AudioCommon::UpdateDataHeader));
@@ -135,8 +135,8 @@ bool InfoUpdater::UpdateVoiceChannelResources(VoiceContext& voice_context) {
 }
 
 bool InfoUpdater::UpdateVoices(VoiceContext& voice_context,
-                               std::vector<ServerMemoryPoolInfo>& memory_pool_info,
-                               VAddr audio_codec_dsp_addr) {
+                               [[maybe_unused]] std::vector<ServerMemoryPoolInfo>& memory_pool_info,
+                               [[maybe_unused]] VAddr audio_codec_dsp_addr) {
     const auto voice_count = voice_context.GetVoiceCount();
     std::vector<VoiceInfo::InParams> voice_in(voice_count);
     std::vector<VoiceInfo::OutParams> voice_out(voice_count);
@@ -165,28 +165,28 @@ bool InfoUpdater::UpdateVoices(VoiceContext& voice_context,
 
     // Update our voices
     for (std::size_t i = 0; i < voice_count; i++) {
-        auto& in_params = voice_in[i];
-        const auto channel_count = static_cast<std::size_t>(in_params.channel_count);
+        auto& voice_in_params = voice_in[i];
+        const auto channel_count = static_cast<std::size_t>(voice_in_params.channel_count);
         // Skip if it's not currently in use
-        if (!in_params.is_in_use) {
+        if (!voice_in_params.is_in_use) {
             continue;
         }
         // Voice states for each channel
         std::array<VoiceState*, AudioCommon::MAX_CHANNEL_COUNT> voice_states{};
-        ASSERT(static_cast<std::size_t>(in_params.id) < voice_count);
+        ASSERT(static_cast<std::size_t>(voice_in_params.id) < voice_count);
 
         // Grab our current voice info
-        auto& voice_info = voice_context.GetInfo(static_cast<std::size_t>(in_params.id));
+        auto& voice_info = voice_context.GetInfo(static_cast<std::size_t>(voice_in_params.id));
 
         ASSERT(channel_count <= AudioCommon::MAX_CHANNEL_COUNT);
 
         // Get all our channel voice states
         for (std::size_t channel = 0; channel < channel_count; channel++) {
             voice_states[channel] =
-                &voice_context.GetState(in_params.voice_channel_resource_ids[channel]);
+                &voice_context.GetState(voice_in_params.voice_channel_resource_ids[channel]);
         }
 
-        if (in_params.is_new) {
+        if (voice_in_params.is_new) {
             // Default our values for our voice
             voice_info.Initialize();
             if (channel_count == 0 || channel_count > AudioCommon::MAX_CHANNEL_COUNT) {
@@ -200,12 +200,12 @@ bool InfoUpdater::UpdateVoices(VoiceContext& voice_context,
         }
 
         // Update our voice
-        voice_info.UpdateParameters(in_params, behavior_info);
+        voice_info.UpdateParameters(voice_in_params, behavior_info);
         // TODO(ogniK): Handle mapping errors with behavior info based on in params response
 
         // Update our wave buffers
-        voice_info.UpdateWaveBuffers(in_params, voice_states, behavior_info);
-        voice_info.WriteOutStatus(voice_out[i], in_params, voice_states);
+        voice_info.UpdateWaveBuffers(voice_in_params, voice_states, behavior_info);
+        voice_info.WriteOutStatus(voice_out[i], voice_in_params, voice_states);
     }
 
     if (!AudioCommon::CanConsumeBuffer(out_params.size(), output_offset, voice_out_size)) {
@@ -445,7 +445,7 @@ bool InfoUpdater::UpdatePerformanceBuffer() {
     return true;
 }
 
-bool InfoUpdater::UpdateErrorInfo(BehaviorInfo& in_behavior_info) {
+bool InfoUpdater::UpdateErrorInfo([[maybe_unused]] BehaviorInfo& in_behavior_info) {
     const auto total_beahvior_info_out = sizeof(BehaviorInfo::OutParams);
 
     if (!AudioCommon::CanConsumeBuffer(out_params.size(), output_offset, total_beahvior_info_out)) {

src/audio_core/info_updater.h

@@ -21,8 +21,8 @@ class SplitterContext;
 class InfoUpdater {
 public:
     // TODO(ogniK): Pass process handle when we support it
-    InfoUpdater(const std::vector<u8>& in_params, std::vector<u8>& out_params,
-                BehaviorInfo& behavior_info);
+    InfoUpdater(const std::vector<u8>& in_params_, std::vector<u8>& out_params_,
+                BehaviorInfo& behavior_info_);
     ~InfoUpdater();
 
     bool UpdateBehaviorInfo(BehaviorInfo& in_behavior_info);

src/audio_core/memory_pool.cpp

@@ -10,11 +10,10 @@ namespace AudioCore {
 
 ServerMemoryPoolInfo::ServerMemoryPoolInfo() = default;
 ServerMemoryPoolInfo::~ServerMemoryPoolInfo() = default;
-bool ServerMemoryPoolInfo::Update(const ServerMemoryPoolInfo::InParams& in_params,
-                                  ServerMemoryPoolInfo::OutParams& out_params) {
+
+bool ServerMemoryPoolInfo::Update(const InParams& in_params, OutParams& out_params) {
     // Our state does not need to be changed
-    if (in_params.state != ServerMemoryPoolInfo::State::RequestAttach &&
-        in_params.state != ServerMemoryPoolInfo::State::RequestDetach) {
+    if (in_params.state != State::RequestAttach && in_params.state != State::RequestDetach) {
         return true;
     }
 
@@ -32,11 +31,11 @@ bool ServerMemoryPoolInfo::Update(const ServerMemoryPoolInfo::InParams& in_param
         return false;
     }
 
-    if (in_params.state == ServerMemoryPoolInfo::State::RequestAttach) {
+    if (in_params.state == State::RequestAttach) {
         cpu_address = in_params.address;
         size = in_params.size;
         used = true;
-        out_params.state = ServerMemoryPoolInfo::State::Attached;
+        out_params.state = State::Attached;
     } else {
         // Unexpected address
         if (cpu_address != in_params.address) {
@@ -54,7 +53,7 @@ bool ServerMemoryPoolInfo::Update(const ServerMemoryPoolInfo::InParams& in_param
         cpu_address = 0;
         size = 0;
         used = false;
-        out_params.state = ServerMemoryPoolInfo::State::Detached;
+        out_params.state = State::Detached;
     }
     return true;
 }

src/audio_core/memory_pool.h

@@ -28,19 +28,18 @@ public:
     struct InParams {
         u64_le address{};
         u64_le size{};
-        ServerMemoryPoolInfo::State state{};
+        State state{};
         INSERT_PADDING_WORDS(3);
     };
-    static_assert(sizeof(ServerMemoryPoolInfo::InParams) == 0x20, "InParams are an invalid size");
+    static_assert(sizeof(InParams) == 0x20, "InParams are an invalid size");
 
     struct OutParams {
-        ServerMemoryPoolInfo::State state{};
+        State state{};
         INSERT_PADDING_WORDS(3);
     };
-    static_assert(sizeof(ServerMemoryPoolInfo::OutParams) == 0x10, "OutParams are an invalid size");
+    static_assert(sizeof(OutParams) == 0x10, "OutParams are an invalid size");
 
-    bool Update(const ServerMemoryPoolInfo::InParams& in_params,
-                ServerMemoryPoolInfo::OutParams& out_params);
+    bool Update(const InParams& in_params, OutParams& out_params);
 
 private:
     // There's another entry here which is the DSP address, however since we're not talking to the

src/audio_core/mix_context.h

@@ -62,17 +62,17 @@ public:
     ServerMixInfo();
     ~ServerMixInfo();
 
-    const ServerMixInfo::InParams& GetInParams() const;
-    ServerMixInfo::InParams& GetInParams();
+    [[nodiscard]] const ServerMixInfo::InParams& GetInParams() const;
+    [[nodiscard]] ServerMixInfo::InParams& GetInParams();
 
     bool Update(EdgeMatrix& edge_matrix, const MixInfo::InParams& mix_in,
                 BehaviorInfo& behavior_info, SplitterContext& splitter_context,
                 EffectContext& effect_context);
-    bool HasAnyConnection() const;
+    [[nodiscard]] bool HasAnyConnection() const;
     void Cleanup();
     void SetEffectCount(std::size_t count);
     void ResetEffectProcessingOrder();
-    s32 GetEffectOrder(std::size_t i) const;
+    [[nodiscard]] s32 GetEffectOrder(std::size_t i) const;
 
 private:
     std::vector<s32> effect_processing_order;
@@ -91,15 +91,15 @@ public:
     void SortInfo();
     bool TsortInfo(SplitterContext& splitter_context);
 
-    std::size_t GetCount() const;
-    ServerMixInfo& GetInfo(std::size_t i);
-    const ServerMixInfo& GetInfo(std::size_t i) const;
-    ServerMixInfo& GetSortedInfo(std::size_t i);
-    const ServerMixInfo& GetSortedInfo(std::size_t i) const;
-    ServerMixInfo& GetFinalMixInfo();
-    const ServerMixInfo& GetFinalMixInfo() const;
-    EdgeMatrix& GetEdgeMatrix();
-    const EdgeMatrix& GetEdgeMatrix() const;
+    [[nodiscard]] std::size_t GetCount() const;
+    [[nodiscard]] ServerMixInfo& GetInfo(std::size_t i);
+    [[nodiscard]] const ServerMixInfo& GetInfo(std::size_t i) const;
+    [[nodiscard]] ServerMixInfo& GetSortedInfo(std::size_t i);
+    [[nodiscard]] const ServerMixInfo& GetSortedInfo(std::size_t i) const;
+    [[nodiscard]] ServerMixInfo& GetFinalMixInfo();
+    [[nodiscard]] const ServerMixInfo& GetFinalMixInfo() const;
+    [[nodiscard]] EdgeMatrix& GetEdgeMatrix();
+    [[nodiscard]] const EdgeMatrix& GetEdgeMatrix() const;
 
 private:
     void CalcMixBufferOffset();

src/audio_core/sink_context.cpp

@@ -5,17 +5,23 @@
 #include "audio_core/sink_context.h"
 
 namespace AudioCore {
-SinkContext::SinkContext(std::size_t sink_count) : sink_count(sink_count) {}
+SinkContext::SinkContext(std::size_t sink_count_) : sink_count{sink_count_} {}
 SinkContext::~SinkContext() = default;
 
 std::size_t SinkContext::GetCount() const {
     return sink_count;
 }
 
-void SinkContext::UpdateMainSink(SinkInfo::InParams& in) {
+void SinkContext::UpdateMainSink(const SinkInfo::InParams& in) {
+    ASSERT(in.type == SinkTypes::Device);
+
+    has_downmix_coefs = in.device.down_matrix_enabled;
+    if (has_downmix_coefs) {
+        downmix_coefficients = in.device.down_matrix_coef;
+    }
     in_use = in.in_use;
     use_count = in.device.input_count;
-    std::memcpy(buffers.data(), in.device.input.data(), AudioCommon::MAX_CHANNEL_COUNT);
+    buffers = in.device.input;
 }
 
 bool SinkContext::InUse() const {
@@ -28,4 +34,12 @@ std::vector<u8> SinkContext::OutputBuffers() const {
     return buffer_ret;
 }
 
+bool SinkContext::HasDownMixingCoefficients() const {
+    return has_downmix_coefs;
+}
+
+const DownmixCoefficients& SinkContext::GetDownmixCoefficients() const {
+    return downmix_coefficients;
+}
+
 } // namespace AudioCore

src/audio_core/sink_context.h

@@ -11,6 +11,8 @@
 
 namespace AudioCore {
 
+using DownmixCoefficients = std::array<float_le, 4>;
+
 enum class SinkTypes : u8 {
     Invalid = 0,
     Device = 1,
@@ -38,21 +40,21 @@ public:
         SinkSampleFormat sample_format;
         std::array<u8, AudioCommon::MAX_CHANNEL_COUNT> input;
         bool in_use;
-        INSERT_UNION_PADDING_BYTES(5);
+        INSERT_PADDING_BYTES_NOINIT(5);
     };
-    static_assert(sizeof(SinkInfo::CircularBufferIn) == 0x28,
+    static_assert(sizeof(CircularBufferIn) == 0x28,
                   "SinkInfo::CircularBufferIn is in invalid size");
 
     struct DeviceIn {
         std::array<u8, 255> device_name;
-        INSERT_UNION_PADDING_BYTES(1);
+        INSERT_PADDING_BYTES_NOINIT(1);
         s32_le input_count;
         std::array<u8, AudioCommon::MAX_CHANNEL_COUNT> input;
-        INSERT_UNION_PADDING_BYTES(1);
+        INSERT_PADDING_BYTES_NOINIT(1);
         bool down_matrix_enabled;
-        std::array<float_le, 4> down_matrix_coef;
+        DownmixCoefficients down_matrix_coef;
     };
-    static_assert(sizeof(SinkInfo::DeviceIn) == 0x11c, "SinkInfo::DeviceIn is an invalid size");
+    static_assert(sizeof(DeviceIn) == 0x11c, "SinkInfo::DeviceIn is an invalid size");
 
     struct InParams {
         SinkTypes type{};
@@ -62,28 +64,33 @@ public:
         INSERT_PADDING_WORDS(6);
         union {
             // std::array<u8, 0x120> raw{};
-            SinkInfo::DeviceIn device;
-            SinkInfo::CircularBufferIn circular_buffer;
+            DeviceIn device;
+            CircularBufferIn circular_buffer;
         };
     };
-    static_assert(sizeof(SinkInfo::InParams) == 0x140, "SinkInfo::InParams are an invalid size!");
+    static_assert(sizeof(InParams) == 0x140, "SinkInfo::InParams are an invalid size!");
 };
 
 class SinkContext {
 public:
-    explicit SinkContext(std::size_t sink_count);
+    explicit SinkContext(std::size_t sink_count_);
     ~SinkContext();
 
-    std::size_t GetCount() const;
+    [[nodiscard]] std::size_t GetCount() const;
 
-    void UpdateMainSink(SinkInfo::InParams& in);
-    bool InUse() const;
-    std::vector<u8> OutputBuffers() const;
+    void UpdateMainSink(const SinkInfo::InParams& in);
+    [[nodiscard]] bool InUse() const;
+    [[nodiscard]] std::vector<u8> OutputBuffers() const;
+
+    [[nodiscard]] bool HasDownMixingCoefficients() const;
+    [[nodiscard]] const DownmixCoefficients& GetDownmixCoefficients() const;
 
 private:
     bool in_use{false};
     s32 use_count{};
     std::array<u8, AudioCommon::MAX_CHANNEL_COUNT> buffers{};
     std::size_t sink_count{};
+    bool has_downmix_coefs{false};
+    DownmixCoefficients downmix_coefficients{};
 };
 } // namespace AudioCore

src/audio_core/splitter_context.cpp

@@ -10,7 +10,7 @@
 
 namespace AudioCore {
 
-ServerSplitterDestinationData::ServerSplitterDestinationData(s32 id) : id(id) {}
+ServerSplitterDestinationData::ServerSplitterDestinationData(s32 id_) : id{id_} {}
 ServerSplitterDestinationData::~ServerSplitterDestinationData() = default;
 
 void ServerSplitterDestinationData::Update(SplitterInfo::InDestinationParams& header) {
@@ -87,7 +87,7 @@ void ServerSplitterDestinationData::UpdateInternalState() {
     needs_update = false;
 }
 
-ServerSplitterInfo::ServerSplitterInfo(s32 id) : id(id) {}
+ServerSplitterInfo::ServerSplitterInfo(s32 id_) : id(id_) {}
 ServerSplitterInfo::~ServerSplitterInfo() = default;
 
 void ServerSplitterInfo::InitializeInfos() {
@@ -121,7 +121,7 @@ const ServerSplitterDestinationData* ServerSplitterInfo::GetHead() const {
 }
 
 ServerSplitterDestinationData* ServerSplitterInfo::GetData(std::size_t depth) {
-    auto current_head = head;
+    auto* current_head = head;
     for (std::size_t i = 0; i < depth; i++) {
         if (current_head == nullptr) {
             return nullptr;
@@ -132,7 +132,7 @@ ServerSplitterDestinationData* ServerSplitterInfo::GetData(std::size_t depth) {
 }
 
 const ServerSplitterDestinationData* ServerSplitterInfo::GetData(std::size_t depth) const {
-    auto current_head = head;
+    auto* current_head = head;
     for (std::size_t i = 0; i < depth; i++) {
         if (current_head == nullptr) {
             return nullptr;
@@ -245,7 +245,7 @@ ServerSplitterDestinationData* SplitterContext::GetDestinationData(std::size_t i
 const ServerSplitterDestinationData* SplitterContext::GetDestinationData(std::size_t info,
                                                                          std::size_t data) const {
     ASSERT(info < info_count);
-    auto& cur_info = GetInfo(info);
+    const auto& cur_info = GetInfo(info);
     return cur_info.GetData(data);
 }
 
@@ -267,11 +267,11 @@ std::size_t SplitterContext::GetDataCount() const {
     return data_count;
 }
 
-void SplitterContext::Setup(std::size_t _info_count, std::size_t _data_count,
+void SplitterContext::Setup(std::size_t info_count_, std::size_t data_count_,
                             bool is_splitter_bug_fixed) {
 
-    info_count = _info_count;
-    data_count = _data_count;
+    info_count = info_count_;
+    data_count = data_count_;
 
     for (std::size_t i = 0; i < info_count; i++) {
         auto& splitter = infos.emplace_back(static_cast<s32>(i));
@@ -364,7 +364,7 @@ bool SplitterContext::RecomposeDestination(ServerSplitterInfo& info,
     // Clear our current destinations
     auto* current_head = info.GetHead();
     while (current_head != nullptr) {
-        auto next_head = current_head->GetNextDestination();
+        auto* next_head = current_head->GetNextDestination();
         current_head->SetNextDestination(nullptr);
         current_head = next_head;
     }
@@ -471,8 +471,8 @@ bool NodeStates::DepthFirstSearch(EdgeMatrix& edge_matrix) {
                 continue;
             }
 
-            const auto node_count = edge_matrix.GetNodeCount();
-            for (s32 j = 0; j < static_cast<s32>(node_count); j++) {
+            const auto edge_node_count = edge_matrix.GetNodeCount();
+            for (s32 j = 0; j < static_cast<s32>(edge_node_count); j++) {
                 // Check if our node is connected to our edge matrix
                 if (!edge_matrix.Connected(current_stack_index, j)) {
                     continue;

src/audio_core/splitter_context.h

@@ -63,7 +63,7 @@ public:
     NodeStates();
     ~NodeStates();
 
-    void Initialize(std::size_t _node_count);
+    void Initialize(std::size_t node_count_);
     bool Tsort(EdgeMatrix& edge_matrix);
     std::size_t GetIndexPos() const;
     const std::vector<s32>& GetIndexList() const;
@@ -72,15 +72,15 @@ private:
     void PushTsortResult(s32 index);
     bool DepthFirstSearch(EdgeMatrix& edge_matrix);
     void ResetState();
-    void UpdateState(NodeStates::State state, std::size_t i);
-    NodeStates::State GetState(std::size_t i);
+    void UpdateState(State state, std::size_t i);
+    State GetState(std::size_t i);
 
     std::size_t node_count{};
     std::vector<bool> was_node_found{};
     std::vector<bool> was_node_completed{};
     std::size_t index_pos{};
     std::vector<s32> index_list{};
-    NodeStates::Stack index_stack{};
+    Stack index_stack{};
 };
 
 enum class SplitterMagic : u32_le {
@@ -97,8 +97,7 @@ public:
         s32_le data_count{};
         INSERT_PADDING_WORDS(5);
     };
-    static_assert(sizeof(SplitterInfo::InHeader) == 0x20,
-                  "SplitterInfo::InHeader is an invalid size");
+    static_assert(sizeof(InHeader) == 0x20, "SplitterInfo::InHeader is an invalid size");
 
     struct InInfoPrams {
         SplitterMagic magic{};
@@ -107,8 +106,7 @@ public:
         s32_le length{};
         s32_le resource_id_base{};
     };
-    static_assert(sizeof(SplitterInfo::InInfoPrams) == 0x14,
-                  "SplitterInfo::InInfoPrams is an invalid size");
+    static_assert(sizeof(InInfoPrams) == 0x14, "SplitterInfo::InInfoPrams is an invalid size");
 
     struct InDestinationParams {
         SplitterMagic magic{};
@@ -118,13 +116,13 @@ public:
         bool in_use{};
         INSERT_PADDING_BYTES(3);
     };
-    static_assert(sizeof(SplitterInfo::InDestinationParams) == 0x70,
+    static_assert(sizeof(InDestinationParams) == 0x70,
                   "SplitterInfo::InDestinationParams is an invalid size");
 };
 
 class ServerSplitterDestinationData {
 public:
-    explicit ServerSplitterDestinationData(s32 id);
+    explicit ServerSplitterDestinationData(s32 id_);
     ~ServerSplitterDestinationData();
 
     void Update(SplitterInfo::InDestinationParams& header);
@@ -153,7 +151,7 @@ private:
 
 class ServerSplitterInfo {
 public:
-    explicit ServerSplitterInfo(s32 id);
+    explicit ServerSplitterInfo(s32 id_);
     ~ServerSplitterInfo();
 
     void InitializeInfos();

src/audio_core/stream.cpp

@@ -31,10 +31,10 @@ u32 Stream::GetNumChannels() const {
     return {};
 }
 
-Stream::Stream(Core::Timing::CoreTiming& core_timing, u32 sample_rate, Format format,
-               ReleaseCallback&& release_callback, SinkStream& sink_stream, std::string&& name_)
-    : sample_rate{sample_rate}, format{format}, release_callback{std::move(release_callback)},
-      sink_stream{sink_stream}, core_timing{core_timing}, name{std::move(name_)} {
+Stream::Stream(Core::Timing::CoreTiming& core_timing_, u32 sample_rate_, Format format_,
+               ReleaseCallback&& release_callback_, SinkStream& sink_stream_, std::string&& name_)
+    : sample_rate{sample_rate_}, format{format_}, release_callback{std::move(release_callback_)},
+      sink_stream{sink_stream_}, core_timing{core_timing_}, name{std::move(name_)} {
     release_event =
         Core::Timing::CreateEvent(name, [this](std::uintptr_t, std::chrono::nanoseconds ns_late) {
             ReleaseActiveBuffer(ns_late);
@@ -51,6 +51,14 @@ void Stream::Stop() {
     UNIMPLEMENTED();
 }
 
+bool Stream::Flush() {
+    const bool had_buffers = !queued_buffers.empty();
+    while (!queued_buffers.empty()) {
+        queued_buffers.pop();
+    }
+    return had_buffers;
+}
+
 void Stream::SetVolume(float volume) {
     game_volume = volume;
 }
@@ -103,7 +111,14 @@ void Stream::PlayNextBuffer(std::chrono::nanoseconds ns_late) {
 
     sink_stream.EnqueueSamples(GetNumChannels(), active_buffer->GetSamples());
 
-    core_timing.ScheduleEvent(GetBufferReleaseNS(*active_buffer) - ns_late, release_event, {});
+    const auto buffer_release_ns = GetBufferReleaseNS(*active_buffer);
+
+    // If ns_late is higher than the update rate ignore the delay
+    if (ns_late > buffer_release_ns) {
+        ns_late = {};
+    }
+
+    core_timing.ScheduleEvent(buffer_release_ns - ns_late, release_event, {});
 }
 
 void Stream::ReleaseActiveBuffer(std::chrono::nanoseconds ns_late) {
@@ -122,7 +137,7 @@ bool Stream::QueueBuffer(BufferPtr&& buffer) {
     return false;
 }
 
-bool Stream::ContainsBuffer(Buffer::Tag tag) const {
+bool Stream::ContainsBuffer([[maybe_unused]] Buffer::Tag tag) const {
     UNIMPLEMENTED();
     return {};
 }
@@ -130,7 +145,25 @@ bool Stream::ContainsBuffer(Buffer::Tag tag) const {
 std::vector<Buffer::Tag> Stream::GetTagsAndReleaseBuffers(std::size_t max_count) {
     std::vector<Buffer::Tag> tags;
     for (std::size_t count = 0; count < max_count && !released_buffers.empty(); ++count) {
-        tags.push_back(released_buffers.front()->GetTag());
+        if (released_buffers.front()) {
+            tags.push_back(released_buffers.front()->GetTag());
+        } else {
+            ASSERT_MSG(false, "Invalid tag in released_buffers!");
+        }
+        released_buffers.pop();
+    }
+    return tags;
+}
+
+std::vector<Buffer::Tag> Stream::GetTagsAndReleaseBuffers() {
+    std::vector<Buffer::Tag> tags;
+    tags.reserve(released_buffers.size());
+    while (!released_buffers.empty()) {
+        if (released_buffers.front()) {
+            tags.push_back(released_buffers.front()->GetTag());
+        } else {
+            ASSERT_MSG(false, "Invalid tag in released_buffers!");
+        }
         released_buffers.pop();
     }
     return tags;

src/audio_core/stream.h

@@ -44,8 +44,8 @@ public:
     /// Callback function type, used to change guest state on a buffer being released
     using ReleaseCallback = std::function<void()>;
 
-    Stream(Core::Timing::CoreTiming& core_timing, u32 sample_rate, Format format,
-           ReleaseCallback&& release_callback, SinkStream& sink_stream, std::string&& name_);
+    Stream(Core::Timing::CoreTiming& core_timing_, u32 sample_rate_, Format format_,
+           ReleaseCallback&& release_callback_, SinkStream& sink_stream_, std::string&& name_);
 
     /// Plays the audio stream
     void Play();
@@ -56,38 +56,44 @@ public:
     /// Queues a buffer into the audio stream, returns true on success
     bool QueueBuffer(BufferPtr&& buffer);
 
+    /// Flush audio buffers
+    bool Flush();
+
     /// Returns true if the audio stream contains a buffer with the specified tag
-    bool ContainsBuffer(Buffer::Tag tag) const;
+    [[nodiscard]] bool ContainsBuffer(Buffer::Tag tag) const;
 
     /// Returns a vector of recently released buffers specified by tag
-    std::vector<Buffer::Tag> GetTagsAndReleaseBuffers(std::size_t max_count);
+    [[nodiscard]] std::vector<Buffer::Tag> GetTagsAndReleaseBuffers(std::size_t max_count);
+
+    /// Returns a vector of all recently released buffers specified by tag
+    [[nodiscard]] std::vector<Buffer::Tag> GetTagsAndReleaseBuffers();
 
     void SetVolume(float volume);
 
-    float GetVolume() const {
+    [[nodiscard]] float GetVolume() const {
         return game_volume;
     }
 
     /// Returns true if the stream is currently playing
-    bool IsPlaying() const {
+    [[nodiscard]] bool IsPlaying() const {
         return state == State::Playing;
     }
 
     /// Returns the number of queued buffers
-    std::size_t GetQueueSize() const {
+    [[nodiscard]] std::size_t GetQueueSize() const {
         return queued_buffers.size();
     }
 
     /// Gets the sample rate
-    u32 GetSampleRate() const {
+    [[nodiscard]] u32 GetSampleRate() const {
         return sample_rate;
     }
 
     /// Gets the number of channels
-    u32 GetNumChannels() const;
+    [[nodiscard]] u32 GetNumChannels() const;
 
     /// Get the state
-    State GetState() const;
+    [[nodiscard]] State GetState() const;
 
 private:
     /// Plays the next queued buffer in the audio stream, starting playback if necessary
@@ -97,7 +103,7 @@ private:
     void ReleaseActiveBuffer(std::chrono::nanoseconds ns_late = {});
 
     /// Gets the number of core cycles when the specified buffer will be released
-    std::chrono::nanoseconds GetBufferReleaseNS(const Buffer& buffer) const;
+    [[nodiscard]] std::chrono::nanoseconds GetBufferReleaseNS(const Buffer& buffer) const;
 
     u32 sample_rate;                  ///< Sample rate of the stream
     Format format;                    ///< Format of the stream

src/audio_core/voice_context.cpp

@@ -8,7 +8,7 @@
 
 namespace AudioCore {
 
-ServerVoiceChannelResource::ServerVoiceChannelResource(s32 id) : id(id) {}
+ServerVoiceChannelResource::ServerVoiceChannelResource(s32 id_) : id(id_) {}
 ServerVoiceChannelResource::~ServerVoiceChannelResource() = default;
 
 bool ServerVoiceChannelResource::InUse() const {
@@ -209,7 +209,8 @@ void ServerVoiceInfo::UpdateWaveBuffers(
 
 void ServerVoiceInfo::UpdateWaveBuffer(ServerWaveBuffer& out_wavebuffer,
                                        const WaveBuffer& in_wave_buffer, SampleFormat sample_format,
-                                       bool is_buffer_valid, BehaviorInfo& behavior_info) {
+                                       bool is_buffer_valid,
+                                       [[maybe_unused]] BehaviorInfo& behavior_info) {
     if (!is_buffer_valid && out_wavebuffer.sent_to_dsp) {
         out_wavebuffer.buffer_address = 0;
         out_wavebuffer.buffer_size = 0;
@@ -400,7 +401,7 @@ bool ServerVoiceInfo::HasValidWaveBuffer(const VoiceState* state) const {
     return std::find(valid_wb.begin(), valid_wb.end(), true) != valid_wb.end();
 }
 
-VoiceContext::VoiceContext(std::size_t voice_count) : voice_count(voice_count) {
+VoiceContext::VoiceContext(std::size_t voice_count_) : voice_count{voice_count_} {
     for (std::size_t i = 0; i < voice_count; i++) {
         voice_channel_resources.emplace_back(static_cast<s32>(i));
         sorted_voice_info.push_back(&voice_info.emplace_back());

src/audio_core/voice_context.h

@@ -86,28 +86,28 @@ struct BehaviorFlags {
 static_assert(sizeof(BehaviorFlags) == 0x4, "BehaviorFlags is an invalid size");
 
 struct ADPCMContext {
-    u16 header{};
-    s16 yn1{};
-    s16 yn2{};
+    u16 header;
+    s16 yn1;
+    s16 yn2;
 };
 static_assert(sizeof(ADPCMContext) == 0x6, "ADPCMContext is an invalid size");
 
 struct VoiceState {
-    s64 played_sample_count{};
-    s32 offset{};
-    s32 wave_buffer_index{};
-    std::array<bool, AudioCommon::MAX_WAVE_BUFFERS> is_wave_buffer_valid{};
-    s32 wave_buffer_consumed{};
-    std::array<s32, AudioCommon::MAX_SAMPLE_HISTORY> sample_history{};
-    s32 fraction{};
-    VAddr context_address{};
-    Codec::ADPCM_Coeff coeff{};
-    ADPCMContext context{};
-    std::array<s64, 2> biquad_filter_state{};
-    std::array<s32, AudioCommon::MAX_MIX_BUFFERS> previous_samples{};
-    u32 external_context_size{};
-    bool is_external_context_used{};
-    bool voice_dropped{};
+    s64 played_sample_count;
+    s32 offset;
+    s32 wave_buffer_index;
+    std::array<bool, AudioCommon::MAX_WAVE_BUFFERS> is_wave_buffer_valid;
+    s32 wave_buffer_consumed;
+    std::array<s32, AudioCommon::MAX_SAMPLE_HISTORY> sample_history;
+    s32 fraction;
+    VAddr context_address;
+    Codec::ADPCM_Coeff coeff;
+    ADPCMContext context;
+    std::array<s64, 2> biquad_filter_state;
+    std::array<s32, AudioCommon::MAX_MIX_BUFFERS> previous_samples;
+    u32 external_context_size;
+    bool is_external_context_used;
+    bool voice_dropped;
 };
 
 class VoiceChannelResource {
@@ -118,12 +118,12 @@ public:
         bool in_use{};
         INSERT_PADDING_BYTES(11);
     };
-    static_assert(sizeof(VoiceChannelResource::InParams) == 0x70, "InParams is an invalid size");
+    static_assert(sizeof(InParams) == 0x70, "InParams is an invalid size");
 };
 
 class ServerVoiceChannelResource {
 public:
-    explicit ServerVoiceChannelResource(s32 id);
+    explicit ServerVoiceChannelResource(s32 id_);
     ~ServerVoiceChannelResource();
 
     bool InUse() const;
@@ -174,7 +174,7 @@ public:
         BehaviorFlags behavior_flags{};
         INSERT_PADDING_BYTES(16);
     };
-    static_assert(sizeof(VoiceInfo::InParams) == 0x170, "InParams is an invalid size");
+    static_assert(sizeof(InParams) == 0x170, "InParams is an invalid size");
 
     struct OutParams {
         u64_le played_sample_count{};
@@ -182,7 +182,7 @@ public:
         u8 voice_dropped{};
         INSERT_PADDING_BYTES(3);
     };
-    static_assert(sizeof(VoiceInfo::OutParams) == 0x10, "OutParams is an invalid size");
+    static_assert(sizeof(OutParams) == 0x10, "OutParams is an invalid size");
 };
 
 class ServerVoiceInfo {
@@ -263,7 +263,7 @@ private:
 
 class VoiceContext {
 public:
-    VoiceContext(std::size_t voice_count);
+    explicit VoiceContext(std::size_t voice_count_);
     ~VoiceContext();
 
     std::size_t GetVoiceCount() const;

src/common/CMakeLists.txt

@@ -98,20 +98,20 @@ add_library(common STATIC
     algorithm.h
     alignment.h
     assert.h
-    atomic_ops.cpp
     atomic_ops.h
     detached_tasks.cpp
     detached_tasks.h
     bit_cast.h
     bit_field.h
+    bit_set.h
     bit_util.h
     cityhash.cpp
     cityhash.h
-    color.h
     common_funcs.h
     common_paths.h
     common_types.h
     concepts.h
+    div_ceil.h
     dynamic_library.cpp
     dynamic_library.h
     fiber.cpp
@@ -121,6 +121,7 @@ add_library(common STATIC
     hash.h
     hex_util.cpp
     hex_util.h
+    intrusive_red_black_tree.h
     logging/backend.cpp
     logging/backend.h
     logging/filter.cpp
@@ -133,17 +134,17 @@ add_library(common STATIC
     math_util.h
     memory_detect.cpp
     memory_detect.h
-    memory_hook.cpp
-    memory_hook.h
     microprofile.cpp
     microprofile.h
     microprofileui.h
     misc.cpp
-    multi_level_queue.h
+    nvidia_flags.cpp
+    nvidia_flags.h
     page_table.cpp
     page_table.h
     param_package.cpp
     param_package.h
+    parent_of_member.h
     quaternion.h
     ring_buffer.h
     scm_rev.cpp
@@ -161,12 +162,13 @@ add_library(common STATIC
     thread.cpp
     thread.h
     thread_queue_list.h
+    thread_worker.cpp
+    thread_worker.h
     threadsafe_queue.h
     time_zone.cpp
     time_zone.h
-    timer.cpp
-    timer.h
-    uint128.cpp
+    tiny_mt.h
+    tree.h
     uint128.h
     uuid.cpp
     uuid.h
@@ -204,11 +206,12 @@ if (MSVC)
 else()
   target_compile_options(common PRIVATE
     -Werror
+
+    $<$<CXX_COMPILER_ID:Clang>:-fsized-deallocation>
   )
 endif()
 
 create_target_directory_groups(common)
-find_package(Boost 1.71 COMPONENTS context headers REQUIRED)
 
 target_link_libraries(common PUBLIC ${Boost_LIBRARIES} fmt::fmt microprofile)
 target_link_libraries(common PRIVATE lz4::lz4 xbyak)

src/common/alignment.h

@@ -9,50 +9,50 @@
 namespace Common {
 
 template <typename T>
-[[nodiscard]] constexpr T AlignUp(T value, std::size_t size) {
-    static_assert(std::is_unsigned_v<T>, "T must be an unsigned value.");
+requires std::is_unsigned_v<T>[[nodiscard]] constexpr T AlignUp(T value, size_t size) {
     auto mod{static_cast<T>(value % size)};
     value -= mod;
     return static_cast<T>(mod == T{0} ? value : value + size);
 }
 
 template <typename T>
-[[nodiscard]] constexpr T AlignDown(T value, std::size_t size) {
-    static_assert(std::is_unsigned_v<T>, "T must be an unsigned value.");
+requires std::is_unsigned_v<T>[[nodiscard]] constexpr T AlignUpLog2(T value, size_t align_log2) {
+    return static_cast<T>((value + ((1ULL << align_log2) - 1)) >> align_log2 << align_log2);
+}
+
+template <typename T>
+requires std::is_unsigned_v<T>[[nodiscard]] constexpr T AlignDown(T value, size_t size) {
     return static_cast<T>(value - value % size);
 }
 
 template <typename T>
-[[nodiscard]] constexpr T AlignBits(T value, std::size_t align) {
-    static_assert(std::is_unsigned_v<T>, "T must be an unsigned value.");
-    return static_cast<T>((value + ((1ULL << align) - 1)) >> align << align);
-}
-
-template <typename T>
-[[nodiscard]] constexpr bool Is4KBAligned(T value) {
-    static_assert(std::is_unsigned_v<T>, "T must be an unsigned value.");
+requires std::is_unsigned_v<T>[[nodiscard]] constexpr bool Is4KBAligned(T value) {
     return (value & 0xFFF) == 0;
 }
 
 template <typename T>
-[[nodiscard]] constexpr bool IsWordAligned(T value) {
-    static_assert(std::is_unsigned_v<T>, "T must be an unsigned value.");
+requires std::is_unsigned_v<T>[[nodiscard]] constexpr bool IsWordAligned(T value) {
     return (value & 0b11) == 0;
 }
 
 template <typename T>
-[[nodiscard]] constexpr bool IsAligned(T value, std::size_t alignment) {
-    using U = typename std::make_unsigned<T>::type;
+requires std::is_integral_v<T>[[nodiscard]] constexpr bool IsAligned(T value, size_t alignment) {
+    using U = typename std::make_unsigned_t<T>;
     const U mask = static_cast<U>(alignment - 1);
     return (value & mask) == 0;
 }
 
-template <typename T, std::size_t Align = 16>
+template <typename T, typename U>
+requires std::is_integral_v<T>[[nodiscard]] constexpr T DivideUp(T x, U y) {
+    return (x + (y - 1)) / y;
+}
+
+template <typename T, size_t Align = 16>
 class AlignmentAllocator {
 public:
     using value_type = T;
-    using size_type = std::size_t;
-    using difference_type = std::ptrdiff_t;
+    using size_type = size_t;
+    using difference_type = ptrdiff_t;
 
     using propagate_on_container_copy_assignment = std::true_type;
     using propagate_on_container_move_assignment = std::true_type;

src/common/atomic_ops.cpp

@@ -1,75 +0,0 @@
-// Copyright 2020 yuzu Emulator Project
-// Licensed under GPLv2 or any later version
-// Refer to the license.txt file included.
-
-#include <cstring>
-
-#include "common/atomic_ops.h"
-
-#if _MSC_VER
-#include <intrin.h>
-#endif
-
-namespace Common {
-
-#if _MSC_VER
-
-bool AtomicCompareAndSwap(volatile u8* pointer, u8 value, u8 expected) {
-    const u8 result =
-        _InterlockedCompareExchange8(reinterpret_cast<volatile char*>(pointer), value, expected);
-    return result == expected;
-}
-
-bool AtomicCompareAndSwap(volatile u16* pointer, u16 value, u16 expected) {
-    const u16 result =
-        _InterlockedCompareExchange16(reinterpret_cast<volatile short*>(pointer), value, expected);
-    return result == expected;
-}
-
-bool AtomicCompareAndSwap(volatile u32* pointer, u32 value, u32 expected) {
-    const u32 result =
-        _InterlockedCompareExchange(reinterpret_cast<volatile long*>(pointer), value, expected);
-    return result == expected;
-}
-
-bool AtomicCompareAndSwap(volatile u64* pointer, u64 value, u64 expected) {
-    const u64 result = _InterlockedCompareExchange64(reinterpret_cast<volatile __int64*>(pointer),
-                                                     value, expected);
-    return result == expected;
-}
-
-bool AtomicCompareAndSwap(volatile u64* pointer, u128 value, u128 expected) {
-    return _InterlockedCompareExchange128(reinterpret_cast<volatile __int64*>(pointer), value[1],
-                                          value[0],
-                                          reinterpret_cast<__int64*>(expected.data())) != 0;
-}
-
-#else
-
-bool AtomicCompareAndSwap(volatile u8* pointer, u8 value, u8 expected) {
-    return __sync_bool_compare_and_swap(pointer, expected, value);
-}
-
-bool AtomicCompareAndSwap(volatile u16* pointer, u16 value, u16 expected) {
-    return __sync_bool_compare_and_swap(pointer, expected, value);
-}
-
-bool AtomicCompareAndSwap(volatile u32* pointer, u32 value, u32 expected) {
-    return __sync_bool_compare_and_swap(pointer, expected, value);
-}
-
-bool AtomicCompareAndSwap(volatile u64* pointer, u64 value, u64 expected) {
-    return __sync_bool_compare_and_swap(pointer, expected, value);
-}
-
-bool AtomicCompareAndSwap(volatile u64* pointer, u128 value, u128 expected) {
-    unsigned __int128 value_a;
-    unsigned __int128 expected_a;
-    std::memcpy(&value_a, value.data(), sizeof(u128));
-    std::memcpy(&expected_a, expected.data(), sizeof(u128));
-    return __sync_bool_compare_and_swap((unsigned __int128*)pointer, expected_a, value_a);
-}
-
-#endif
-
-} // namespace Common

src/common/atomic_ops.h

@@ -4,14 +4,75 @@
 
 #pragma once
 
+#include <cstring>
+#include <memory>
+
 #include "common/common_types.h"
 
+#if _MSC_VER
+#include <intrin.h>
+#endif
+
 namespace Common {
 
-[[nodiscard]] bool AtomicCompareAndSwap(volatile u8* pointer, u8 value, u8 expected);
-[[nodiscard]] bool AtomicCompareAndSwap(volatile u16* pointer, u16 value, u16 expected);
-[[nodiscard]] bool AtomicCompareAndSwap(volatile u32* pointer, u32 value, u32 expected);
-[[nodiscard]] bool AtomicCompareAndSwap(volatile u64* pointer, u64 value, u64 expected);
-[[nodiscard]] bool AtomicCompareAndSwap(volatile u64* pointer, u128 value, u128 expected);
+#if _MSC_VER
+
+[[nodiscard]] inline bool AtomicCompareAndSwap(volatile u8* pointer, u8 value, u8 expected) {
+    const u8 result =
+        _InterlockedCompareExchange8(reinterpret_cast<volatile char*>(pointer), value, expected);
+    return result == expected;
+}
+
+[[nodiscard]] inline bool AtomicCompareAndSwap(volatile u16* pointer, u16 value, u16 expected) {
+    const u16 result =
+        _InterlockedCompareExchange16(reinterpret_cast<volatile short*>(pointer), value, expected);
+    return result == expected;
+}
+
+[[nodiscard]] inline bool AtomicCompareAndSwap(volatile u32* pointer, u32 value, u32 expected) {
+    const u32 result =
+        _InterlockedCompareExchange(reinterpret_cast<volatile long*>(pointer), value, expected);
+    return result == expected;
+}
+
+[[nodiscard]] inline bool AtomicCompareAndSwap(volatile u64* pointer, u64 value, u64 expected) {
+    const u64 result = _InterlockedCompareExchange64(reinterpret_cast<volatile __int64*>(pointer),
+                                                     value, expected);
+    return result == expected;
+}
+
+[[nodiscard]] inline bool AtomicCompareAndSwap(volatile u64* pointer, u128 value, u128 expected) {
+    return _InterlockedCompareExchange128(reinterpret_cast<volatile __int64*>(pointer), value[1],
+                                          value[0],
+                                          reinterpret_cast<__int64*>(expected.data())) != 0;
+}
+
+#else
+
+[[nodiscard]] inline bool AtomicCompareAndSwap(volatile u8* pointer, u8 value, u8 expected) {
+    return __sync_bool_compare_and_swap(pointer, expected, value);
+}
+
+[[nodiscard]] inline bool AtomicCompareAndSwap(volatile u16* pointer, u16 value, u16 expected) {
+    return __sync_bool_compare_and_swap(pointer, expected, value);
+}
+
+[[nodiscard]] inline bool AtomicCompareAndSwap(volatile u32* pointer, u32 value, u32 expected) {
+    return __sync_bool_compare_and_swap(pointer, expected, value);
+}
+
+[[nodiscard]] inline bool AtomicCompareAndSwap(volatile u64* pointer, u64 value, u64 expected) {
+    return __sync_bool_compare_and_swap(pointer, expected, value);
+}
+
+[[nodiscard]] inline bool AtomicCompareAndSwap(volatile u64* pointer, u128 value, u128 expected) {
+    unsigned __int128 value_a;
+    unsigned __int128 expected_a;
+    std::memcpy(&value_a, value.data(), sizeof(u128));
+    std::memcpy(&expected_a, expected.data(), sizeof(u128));
+    return __sync_bool_compare_and_swap((unsigned __int128*)pointer, expected_a, value_a);
+}
+
+#endif
 
 } // namespace Common

src/common/bit_set.h

@@ -0,0 +1,99 @@
+/*
+ * Copyright (c) 2018-2020 Atmosphère-NX
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#pragma once
+
+#include <array>
+#include <bit>
+
+#include "common/alignment.h"
+#include "common/bit_util.h"
+#include "common/common_types.h"
+
+namespace Common {
+
+namespace impl {
+
+template <typename Storage, size_t N>
+class BitSet {
+
+public:
+    constexpr BitSet() = default;
+
+    constexpr void SetBit(size_t i) {
+        this->words[i / FlagsPerWord] |= GetBitMask(i % FlagsPerWord);
+    }
+
+    constexpr void ClearBit(size_t i) {
+        this->words[i / FlagsPerWord] &= ~GetBitMask(i % FlagsPerWord);
+    }
+
+    constexpr size_t CountLeadingZero() const {
+        for (size_t i = 0; i < NumWords; i++) {
+            if (this->words[i]) {
+                return FlagsPerWord * i + CountLeadingZeroImpl(this->words[i]);
+            }
+        }
+        return FlagsPerWord * NumWords;
+    }
+
+    constexpr size_t GetNextSet(size_t n) const {
+        for (size_t i = (n + 1) / FlagsPerWord; i < NumWords; i++) {
+            Storage word = this->words[i];
+            if (!IsAligned(n + 1, FlagsPerWord)) {
+                word &= GetBitMask(n % FlagsPerWord) - 1;
+            }
+            if (word) {
+                return FlagsPerWord * i + CountLeadingZeroImpl(word);
+            }
+        }
+        return FlagsPerWord * NumWords;
+    }
+
+private:
+    static_assert(std::is_unsigned_v<Storage>);
+    static_assert(sizeof(Storage) <= sizeof(u64));
+
+    static constexpr size_t FlagsPerWord = BitSize<Storage>();
+    static constexpr size_t NumWords = AlignUp(N, FlagsPerWord) / FlagsPerWord;
+
+    static constexpr auto CountLeadingZeroImpl(Storage word) {
+        return std::countl_zero(static_cast<unsigned long long>(word)) -
+               (BitSize<unsigned long long>() - FlagsPerWord);
+    }
+
+    static constexpr Storage GetBitMask(size_t bit) {
+        return Storage(1) << (FlagsPerWord - 1 - bit);
+    }
+
+    std::array<Storage, NumWords> words{};
+};
+
+} // namespace impl
+
+template <size_t N>
+using BitSet8 = impl::BitSet<u8, N>;
+
+template <size_t N>
+using BitSet16 = impl::BitSet<u16, N>;
+
+template <size_t N>
+using BitSet32 = impl::BitSet<u32, N>;
+
+template <size_t N>
+using BitSet64 = impl::BitSet<u64, N>;
+
+} // namespace Common

src/common/bit_util.h

@@ -4,13 +4,10 @@
 
 #pragma once
 
+#include <bit>
 #include <climits>
 #include <cstddef>
 
-#ifdef _MSC_VER
-#include <intrin.h>
-#endif
-
 #include "common/common_types.h"
 
 namespace Common {
@@ -21,124 +18,30 @@ template <typename T>
     return sizeof(T) * CHAR_BIT;
 }
 
-#ifdef _MSC_VER
-[[nodiscard]] inline u32 CountLeadingZeroes32(u32 value) {
-    unsigned long leading_zero = 0;
-
-    if (_BitScanReverse(&leading_zero, value) != 0) {
-        return 31 - leading_zero;
-    }
-
-    return 32;
+[[nodiscard]] constexpr u32 MostSignificantBit32(const u32 value) {
+    return 31U - static_cast<u32>(std::countl_zero(value));
 }
 
-[[nodiscard]] inline u32 CountLeadingZeroes64(u64 value) {
-    unsigned long leading_zero = 0;
-
-    if (_BitScanReverse64(&leading_zero, value) != 0) {
-        return 63 - leading_zero;
-    }
-
-    return 64;
-}
-#else
-[[nodiscard]] inline u32 CountLeadingZeroes32(u32 value) {
-    if (value == 0) {
-        return 32;
-    }
-
-    return static_cast<u32>(__builtin_clz(value));
+[[nodiscard]] constexpr u32 MostSignificantBit64(const u64 value) {
+    return 63U - static_cast<u32>(std::countl_zero(value));
 }
 
-[[nodiscard]] inline u32 CountLeadingZeroes64(u64 value) {
-    if (value == 0) {
-        return 64;
-    }
-
-    return static_cast<u32>(__builtin_clzll(value));
-}
-#endif
-
-#ifdef _MSC_VER
-[[nodiscard]] inline u32 CountTrailingZeroes32(u32 value) {
-    unsigned long trailing_zero = 0;
-
-    if (_BitScanForward(&trailing_zero, value) != 0) {
-        return trailing_zero;
-    }
-
-    return 32;
-}
-
-[[nodiscard]] inline u32 CountTrailingZeroes64(u64 value) {
-    unsigned long trailing_zero = 0;
-
-    if (_BitScanForward64(&trailing_zero, value) != 0) {
-        return trailing_zero;
-    }
-
-    return 64;
-}
-#else
-[[nodiscard]] inline u32 CountTrailingZeroes32(u32 value) {
-    if (value == 0) {
-        return 32;
-    }
-
-    return static_cast<u32>(__builtin_ctz(value));
-}
-
-[[nodiscard]] inline u32 CountTrailingZeroes64(u64 value) {
-    if (value == 0) {
-        return 64;
-    }
-
-    return static_cast<u32>(__builtin_ctzll(value));
-}
-#endif
-
-#ifdef _MSC_VER
-
-[[nodiscard]] inline u32 MostSignificantBit32(const u32 value) {
-    unsigned long result;
-    _BitScanReverse(&result, value);
-    return static_cast<u32>(result);
-}
-
-[[nodiscard]] inline u32 MostSignificantBit64(const u64 value) {
-    unsigned long result;
-    _BitScanReverse64(&result, value);
-    return static_cast<u32>(result);
-}
-
-#else
-
-[[nodiscard]] inline u32 MostSignificantBit32(const u32 value) {
-    return 31U - static_cast<u32>(__builtin_clz(value));
-}
-
-[[nodiscard]] inline u32 MostSignificantBit64(const u64 value) {
-    return 63U - static_cast<u32>(__builtin_clzll(value));
-}
-
-#endif
-
-[[nodiscard]] inline u32 Log2Floor32(const u32 value) {
+[[nodiscard]] constexpr u32 Log2Floor32(const u32 value) {
     return MostSignificantBit32(value);
 }
 
-[[nodiscard]] inline u32 Log2Ceil32(const u32 value) {
-    const u32 log2_f = Log2Floor32(value);
-    return log2_f + ((value ^ (1U << log2_f)) != 0U);
-}
-
-[[nodiscard]] inline u32 Log2Floor64(const u64 value) {
+[[nodiscard]] constexpr u32 Log2Floor64(const u64 value) {
     return MostSignificantBit64(value);
 }
 
-[[nodiscard]] inline u32 Log2Ceil64(const u64 value) {
-    const u64 log2_f = static_cast<u64>(Log2Floor64(value));
-    return static_cast<u32>(log2_f + ((value ^ (1ULL << log2_f)) != 0ULL));
+[[nodiscard]] constexpr u32 Log2Ceil32(const u32 value) {
+    const u32 log2_f = Log2Floor32(value);
+    return log2_f + static_cast<u32>((value ^ (1U << log2_f)) != 0U);
+}
+
+[[nodiscard]] constexpr u32 Log2Ceil64(const u64 value) {
+    const u64 log2_f = Log2Floor64(value);
+    return static_cast<u32>(log2_f + static_cast<u64>((value ^ (1ULL << log2_f)) != 0ULL));
 }
 
 } // namespace Common

src/common/cityhash.cpp

@@ -28,8 +28,10 @@
 // compromising on hash quality.
 
 #include <algorithm>
-#include <string.h> // for memcpy and memset
-#include "cityhash.h"
+#include <cstring>
+#include <utility>
+
+#include "common/cityhash.h"
 #include "common/swap.h"
 
 // #include "config.h"
@@ -42,21 +44,17 @@
 
 using namespace std;
 
-typedef uint8_t uint8;
-typedef uint32_t uint32;
-typedef uint64_t uint64;
-
 namespace Common {
 
-static uint64 UNALIGNED_LOAD64(const char* p) {
-    uint64 result;
-    memcpy(&result, p, sizeof(result));
+static u64 unaligned_load64(const char* p) {
+    u64 result;
+    std::memcpy(&result, p, sizeof(result));
     return result;
 }
 
-static uint32 UNALIGNED_LOAD32(const char* p) {
-    uint32 result;
-    memcpy(&result, p, sizeof(result));
+static u32 unaligned_load32(const char* p) {
+    u32 result;
+    std::memcpy(&result, p, sizeof(result));
     return result;
 }
 
@@ -76,64 +74,64 @@ static uint32 UNALIGNED_LOAD32(const char* p) {
 #endif
 #endif
 
-static uint64 Fetch64(const char* p) {
-    return uint64_in_expected_order(UNALIGNED_LOAD64(p));
+static u64 Fetch64(const char* p) {
+    return uint64_in_expected_order(unaligned_load64(p));
 }
 
-static uint32 Fetch32(const char* p) {
-    return uint32_in_expected_order(UNALIGNED_LOAD32(p));
+static u32 Fetch32(const char* p) {
+    return uint32_in_expected_order(unaligned_load32(p));
 }
 
 // Some primes between 2^63 and 2^64 for various uses.
-static const uint64 k0 = 0xc3a5c85c97cb3127ULL;
-static const uint64 k1 = 0xb492b66fbe98f273ULL;
-static const uint64 k2 = 0x9ae16a3b2f90404fULL;
+static constexpr u64 k0 = 0xc3a5c85c97cb3127ULL;
+static constexpr u64 k1 = 0xb492b66fbe98f273ULL;
+static constexpr u64 k2 = 0x9ae16a3b2f90404fULL;
 
 // Bitwise right rotate.  Normally this will compile to a single
 // instruction, especially if the shift is a manifest constant.
-static uint64 Rotate(uint64 val, int shift) {
+static u64 Rotate(u64 val, int shift) {
     // Avoid shifting by 64: doing so yields an undefined result.
     return shift == 0 ? val : ((val >> shift) | (val << (64 - shift)));
 }
 
-static uint64 ShiftMix(uint64 val) {
+static u64 ShiftMix(u64 val) {
     return val ^ (val >> 47);
 }
 
-static uint64 HashLen16(uint64 u, uint64 v) {
-    return Hash128to64(uint128(u, v));
+static u64 HashLen16(u64 u, u64 v) {
+    return Hash128to64(u128{u, v});
 }
 
-static uint64 HashLen16(uint64 u, uint64 v, uint64 mul) {
+static u64 HashLen16(u64 u, u64 v, u64 mul) {
     // Murmur-inspired hashing.
-    uint64 a = (u ^ v) * mul;
+    u64 a = (u ^ v) * mul;
     a ^= (a >> 47);
-    uint64 b = (v ^ a) * mul;
+    u64 b = (v ^ a) * mul;
     b ^= (b >> 47);
     b *= mul;
     return b;
 }
 
-static uint64 HashLen0to16(const char* s, std::size_t len) {
+static u64 HashLen0to16(const char* s, size_t len) {
     if (len >= 8) {
-        uint64 mul = k2 + len * 2;
-        uint64 a = Fetch64(s) + k2;
-        uint64 b = Fetch64(s + len - 8);
-        uint64 c = Rotate(b, 37) * mul + a;
-        uint64 d = (Rotate(a, 25) + b) * mul;
+        u64 mul = k2 + len * 2;
+        u64 a = Fetch64(s) + k2;
+        u64 b = Fetch64(s + len - 8);
+        u64 c = Rotate(b, 37) * mul + a;
+        u64 d = (Rotate(a, 25) + b) * mul;
         return HashLen16(c, d, mul);
     }
     if (len >= 4) {
-        uint64 mul = k2 + len * 2;
-        uint64 a = Fetch32(s);
+        u64 mul = k2 + len * 2;
+        u64 a = Fetch32(s);
         return HashLen16(len + (a << 3), Fetch32(s + len - 4), mul);
     }
     if (len > 0) {
-        uint8 a = s[0];
-        uint8 b = s[len >> 1];
-        uint8 c = s[len - 1];
-        uint32 y = static_cast<uint32>(a) + (static_cast<uint32>(b) << 8);
-        uint32 z = static_cast<uint32>(len) + (static_cast<uint32>(c) << 2);
+        u8 a = s[0];
+        u8 b = s[len >> 1];
+        u8 c = s[len - 1];
+        u32 y = static_cast<u32>(a) + (static_cast<u32>(b) << 8);
+        u32 z = static_cast<u32>(len) + (static_cast<u32>(c) << 2);
         return ShiftMix(y * k2 ^ z * k0) * k2;
     }
     return k2;
@@ -141,22 +139,21 @@ static uint64 HashLen0to16(const char* s, std::size_t len) {
 
 // This probably works well for 16-byte strings as well, but it may be overkill
 // in that case.
-static uint64 HashLen17to32(const char* s, std::size_t len) {
-    uint64 mul = k2 + len * 2;
-    uint64 a = Fetch64(s) * k1;
-    uint64 b = Fetch64(s + 8);
-    uint64 c = Fetch64(s + len - 8) * mul;
-    uint64 d = Fetch64(s + len - 16) * k2;
+static u64 HashLen17to32(const char* s, size_t len) {
+    u64 mul = k2 + len * 2;
+    u64 a = Fetch64(s) * k1;
+    u64 b = Fetch64(s + 8);
+    u64 c = Fetch64(s + len - 8) * mul;
+    u64 d = Fetch64(s + len - 16) * k2;
     return HashLen16(Rotate(a + b, 43) + Rotate(c, 30) + d, a + Rotate(b + k2, 18) + c, mul);
 }
 
 // Return a 16-byte hash for 48 bytes.  Quick and dirty.
 // Callers do best to use "random-looking" values for a and b.
-static pair<uint64, uint64> WeakHashLen32WithSeeds(uint64 w, uint64 x, uint64 y, uint64 z, uint64 a,
-                                                   uint64 b) {
+static pair<u64, u64> WeakHashLen32WithSeeds(u64 w, u64 x, u64 y, u64 z, u64 a, u64 b) {
     a += w;
     b = Rotate(b + a + z, 21);
-    uint64 c = a;
+    u64 c = a;
     a += x;
     a += y;
     b += Rotate(a, 44);
@@ -164,34 +161,34 @@ static pair<uint64, uint64> WeakHashLen32WithSeeds(uint64 w, uint64 x, uint64 y,
 }
 
 // Return a 16-byte hash for s[0] ... s[31], a, and b.  Quick and dirty.
-static pair<uint64, uint64> WeakHashLen32WithSeeds(const char* s, uint64 a, uint64 b) {
+static pair<u64, u64> WeakHashLen32WithSeeds(const char* s, u64 a, u64 b) {
     return WeakHashLen32WithSeeds(Fetch64(s), Fetch64(s + 8), Fetch64(s + 16), Fetch64(s + 24), a,
                                   b);
 }
 
 // Return an 8-byte hash for 33 to 64 bytes.
-static uint64 HashLen33to64(const char* s, std::size_t len) {
-    uint64 mul = k2 + len * 2;
-    uint64 a = Fetch64(s) * k2;
-    uint64 b = Fetch64(s + 8);
-    uint64 c = Fetch64(s + len - 24);
-    uint64 d = Fetch64(s + len - 32);
-    uint64 e = Fetch64(s + 16) * k2;
-    uint64 f = Fetch64(s + 24) * 9;
-    uint64 g = Fetch64(s + len - 8);
-    uint64 h = Fetch64(s + len - 16) * mul;
-    uint64 u = Rotate(a + g, 43) + (Rotate(b, 30) + c) * 9;
-    uint64 v = ((a + g) ^ d) + f + 1;
-    uint64 w = swap64((u + v) * mul) + h;
-    uint64 x = Rotate(e + f, 42) + c;
-    uint64 y = (swap64((v + w) * mul) + g) * mul;
-    uint64 z = e + f + c;
+static u64 HashLen33to64(const char* s, size_t len) {
+    u64 mul = k2 + len * 2;
+    u64 a = Fetch64(s) * k2;
+    u64 b = Fetch64(s + 8);
+    u64 c = Fetch64(s + len - 24);
+    u64 d = Fetch64(s + len - 32);
+    u64 e = Fetch64(s + 16) * k2;
+    u64 f = Fetch64(s + 24) * 9;
+    u64 g = Fetch64(s + len - 8);
+    u64 h = Fetch64(s + len - 16) * mul;
+    u64 u = Rotate(a + g, 43) + (Rotate(b, 30) + c) * 9;
+    u64 v = ((a + g) ^ d) + f + 1;
+    u64 w = swap64((u + v) * mul) + h;
+    u64 x = Rotate(e + f, 42) + c;
+    u64 y = (swap64((v + w) * mul) + g) * mul;
+    u64 z = e + f + c;
     a = swap64((x + z) * mul + y) + b;
     b = ShiftMix((z + a) * mul + d + h) * mul;
     return b + x;
 }
 
-uint64 CityHash64(const char* s, std::size_t len) {
+u64 CityHash64(const char* s, size_t len) {
     if (len <= 32) {
         if (len <= 16) {
             return HashLen0to16(s, len);
@@ -204,15 +201,15 @@ uint64 CityHash64(const char* s, std::size_t len) {
 
     // For strings over 64 bytes we hash the end first, and then as we
     // loop we keep 56 bytes of state: v, w, x, y, and z.
-    uint64 x = Fetch64(s + len - 40);
-    uint64 y = Fetch64(s + len - 16) + Fetch64(s + len - 56);
-    uint64 z = HashLen16(Fetch64(s + len - 48) + len, Fetch64(s + len - 24));
-    pair<uint64, uint64> v = WeakHashLen32WithSeeds(s + len - 64, len, z);
-    pair<uint64, uint64> w = WeakHashLen32WithSeeds(s + len - 32, y + k1, x);
+    u64 x = Fetch64(s + len - 40);
+    u64 y = Fetch64(s + len - 16) + Fetch64(s + len - 56);
+    u64 z = HashLen16(Fetch64(s + len - 48) + len, Fetch64(s + len - 24));
+    pair<u64, u64> v = WeakHashLen32WithSeeds(s + len - 64, len, z);
+    pair<u64, u64> w = WeakHashLen32WithSeeds(s + len - 32, y + k1, x);
     x = x * k1 + Fetch64(s);
 
     // Decrease len to the nearest multiple of 64, and operate on 64-byte chunks.
-    len = (len - 1) & ~static_cast<std::size_t>(63);
+    len = (len - 1) & ~static_cast<size_t>(63);
     do {
         x = Rotate(x + y + v.first + Fetch64(s + 8), 37) * k1;
         y = Rotate(y + v.second + Fetch64(s + 48), 42) * k1;
@@ -229,21 +226,21 @@ uint64 CityHash64(const char* s, std::size_t len) {
                      HashLen16(v.second, w.second) + x);
 }
 
-uint64 CityHash64WithSeed(const char* s, std::size_t len, uint64 seed) {
+u64 CityHash64WithSeed(const char* s, size_t len, u64 seed) {
     return CityHash64WithSeeds(s, len, k2, seed);
 }
 
-uint64 CityHash64WithSeeds(const char* s, std::size_t len, uint64 seed0, uint64 seed1) {
+u64 CityHash64WithSeeds(const char* s, size_t len, u64 seed0, u64 seed1) {
     return HashLen16(CityHash64(s, len) - seed0, seed1);
 }
 
 // A subroutine for CityHash128().  Returns a decent 128-bit hash for strings
 // of any length representable in signed long.  Based on City and Murmur.
-static uint128 CityMurmur(const char* s, std::size_t len, uint128 seed) {
-    uint64 a = Uint128Low64(seed);
-    uint64 b = Uint128High64(seed);
-    uint64 c = 0;
-    uint64 d = 0;
+static u128 CityMurmur(const char* s, size_t len, u128 seed) {
+    u64 a = seed[0];
+    u64 b = seed[1];
+    u64 c = 0;
+    u64 d = 0;
     signed long l = static_cast<long>(len) - 16;
     if (l <= 0) { // len <= 16
         a = ShiftMix(a * k1) * k1;
@@ -266,20 +263,20 @@ static uint128 CityMurmur(const char* s, std::size_t len, uint128 seed) {
     }
     a = HashLen16(a, c);
     b = HashLen16(d, b);
-    return uint128(a ^ b, HashLen16(b, a));
+    return u128{a ^ b, HashLen16(b, a)};
 }
 
-uint128 CityHash128WithSeed(const char* s, std::size_t len, uint128 seed) {
+u128 CityHash128WithSeed(const char* s, size_t len, u128 seed) {
     if (len < 128) {
         return CityMurmur(s, len, seed);
     }
 
     // We expect len >= 128 to be the common case.  Keep 56 bytes of state:
     // v, w, x, y, and z.
-    pair<uint64, uint64> v, w;
-    uint64 x = Uint128Low64(seed);
-    uint64 y = Uint128High64(seed);
-    uint64 z = len * k1;
+    pair<u64, u64> v, w;
+    u64 x = seed[0];
+    u64 y = seed[1];
+    u64 z = len * k1;
     v.first = Rotate(y ^ k1, 49) * k1 + Fetch64(s);
     v.second = Rotate(v.first, 42) * k1 + Fetch64(s + 8);
     w.first = Rotate(y + z, 35) * k1 + x;
@@ -313,7 +310,7 @@ uint128 CityHash128WithSeed(const char* s, std::size_t len, uint128 seed) {
     w.first *= 9;
     v.first *= k0;
     // If 0 < len < 128, hash up to 4 chunks of 32 bytes each from the end of s.
-    for (std::size_t tail_done = 0; tail_done < len;) {
+    for (size_t tail_done = 0; tail_done < len;) {
         tail_done += 32;
         y = Rotate(x + y, 42) * k0 + v.second;
         w.first += Fetch64(s + len - tail_done + 16);
@@ -328,13 +325,12 @@ uint128 CityHash128WithSeed(const char* s, std::size_t len, uint128 seed) {
     // different 56-byte-to-8-byte hashes to get a 16-byte final result.
     x = HashLen16(x, v.first);
     y = HashLen16(y + z, w.first);
-    return uint128(HashLen16(x + v.second, w.second) + y, HashLen16(x + w.second, y + v.second));
+    return u128{HashLen16(x + v.second, w.second) + y, HashLen16(x + w.second, y + v.second)};
 }
 
-uint128 CityHash128(const char* s, std::size_t len) {
-    return len >= 16
-               ? CityHash128WithSeed(s + 16, len - 16, uint128(Fetch64(s), Fetch64(s + 8) + k0))
-               : CityHash128WithSeed(s, len, uint128(k0, k1));
+u128 CityHash128(const char* s, size_t len) {
+    return len >= 16 ? CityHash128WithSeed(s + 16, len - 16, u128{Fetch64(s), Fetch64(s + 8) + k0})
+                     : CityHash128WithSeed(s, len, u128{k0, k1});
 }
 
 } // namespace Common

src/common/cityhash.h

@@ -61,50 +61,38 @@
 
 #pragma once
 
-#include <cstddef>
-#include <cstdint>
-#include <utility>
+#include "common/common_types.h"
 
 namespace Common {
 
-using uint128 = std::pair<uint64_t, uint64_t>;
-
-[[nodiscard]] inline uint64_t Uint128Low64(const uint128& x) {
-    return x.first;
-}
-[[nodiscard]] inline uint64_t Uint128High64(const uint128& x) {
-    return x.second;
-}
-
 // Hash function for a byte array.
-[[nodiscard]] uint64_t CityHash64(const char* buf, std::size_t len);
+[[nodiscard]] u64 CityHash64(const char* buf, size_t len);
 
 // Hash function for a byte array.  For convenience, a 64-bit seed is also
 // hashed into the result.
-[[nodiscard]] uint64_t CityHash64WithSeed(const char* buf, std::size_t len, uint64_t seed);
+[[nodiscard]] u64 CityHash64WithSeed(const char* buf, size_t len, u64 seed);
 
 // Hash function for a byte array.  For convenience, two seeds are also
 // hashed into the result.
-[[nodiscard]] uint64_t CityHash64WithSeeds(const char* buf, std::size_t len, uint64_t seed0,
-                                           uint64_t seed1);
+[[nodiscard]] u64 CityHash64WithSeeds(const char* buf, size_t len, u64 seed0, u64 seed1);
 
 // Hash function for a byte array.
-[[nodiscard]] uint128 CityHash128(const char* s, std::size_t len);
+[[nodiscard]] u128 CityHash128(const char* s, size_t len);
 
 // Hash function for a byte array.  For convenience, a 128-bit seed is also
 // hashed into the result.
-[[nodiscard]] uint128 CityHash128WithSeed(const char* s, std::size_t len, uint128 seed);
+[[nodiscard]] u128 CityHash128WithSeed(const char* s, size_t len, u128 seed);
 
 // Hash 128 input bits down to 64 bits of output.
 // This is intended to be a reasonably good hash function.
-[[nodiscard]] inline uint64_t Hash128to64(const uint128& x) {
+[[nodiscard]] inline u64 Hash128to64(const u128& x) {
     // Murmur-inspired hashing.
-    const uint64_t kMul = 0x9ddfea08eb382d69ULL;
-    uint64_t a = (Uint128Low64(x) ^ Uint128High64(x)) * kMul;
+    const u64 mul = 0x9ddfea08eb382d69ULL;
+    u64 a = (x[0] ^ x[1]) * mul;
     a ^= (a >> 47);
-    uint64_t b = (Uint128High64(x) ^ a) * kMul;
+    u64 b = (x[1] ^ a) * mul;
     b ^= (b >> 47);
-    b *= kMul;
+    b *= mul;
     return b;
 }
 

src/common/color.h

@@ -1,271 +0,0 @@
-// Copyright 2014 Citra Emulator Project
-// Licensed under GPLv2 or any later version
-// Refer to the license.txt file included.
-
-#pragma once
-
-#include <cstring>
-
-#include "common/common_types.h"
-#include "common/swap.h"
-#include "common/vector_math.h"
-
-namespace Common::Color {
-
-/// Convert a 1-bit color component to 8 bit
-[[nodiscard]] constexpr u8 Convert1To8(u8 value) {
-    return value * 255;
-}
-
-/// Convert a 4-bit color component to 8 bit
-[[nodiscard]] constexpr u8 Convert4To8(u8 value) {
-    return (value << 4) | value;
-}
-
-/// Convert a 5-bit color component to 8 bit
-[[nodiscard]] constexpr u8 Convert5To8(u8 value) {
-    return (value << 3) | (value >> 2);
-}
-
-/// Convert a 6-bit color component to 8 bit
-[[nodiscard]] constexpr u8 Convert6To8(u8 value) {
-    return (value << 2) | (value >> 4);
-}
-
-/// Convert a 8-bit color component to 1 bit
-[[nodiscard]] constexpr u8 Convert8To1(u8 value) {
-    return value >> 7;
-}
-
-/// Convert a 8-bit color component to 4 bit
-[[nodiscard]] constexpr u8 Convert8To4(u8 value) {
-    return value >> 4;
-}
-
-/// Convert a 8-bit color component to 5 bit
-[[nodiscard]] constexpr u8 Convert8To5(u8 value) {
-    return value >> 3;
-}
-
-/// Convert a 8-bit color component to 6 bit
-[[nodiscard]] constexpr u8 Convert8To6(u8 value) {
-    return value >> 2;
-}
-
-/**
- * Decode a color stored in RGBA8 format
- * @param bytes Pointer to encoded source color
- * @return Result color decoded as Common::Vec4<u8>
- */
-[[nodiscard]] inline Common::Vec4<u8> DecodeRGBA8(const u8* bytes) {
-    return {bytes[3], bytes[2], bytes[1], bytes[0]};
-}
-
-/**
- * Decode a color stored in RGB8 format
- * @param bytes Pointer to encoded source color
- * @return Result color decoded as Common::Vec4<u8>
- */
-[[nodiscard]] inline Common::Vec4<u8> DecodeRGB8(const u8* bytes) {
-    return {bytes[2], bytes[1], bytes[0], 255};
-}
-
-/**
- * Decode a color stored in RG8 (aka HILO8) format
- * @param bytes Pointer to encoded source color
- * @return Result color decoded as Common::Vec4<u8>
- */
-[[nodiscard]] inline Common::Vec4<u8> DecodeRG8(const u8* bytes) {
-    return {bytes[1], bytes[0], 0, 255};
-}
-
-/**
- * Decode a color stored in RGB565 format
- * @param bytes Pointer to encoded source color
- * @return Result color decoded as Common::Vec4<u8>
- */
-[[nodiscard]] inline Common::Vec4<u8> DecodeRGB565(const u8* bytes) {
-    u16_le pixel;
-    std::memcpy(&pixel, bytes, sizeof(pixel));
-    return {Convert5To8((pixel >> 11) & 0x1F), Convert6To8((pixel >> 5) & 0x3F),
-            Convert5To8(pixel & 0x1F), 255};
-}
-
-/**
- * Decode a color stored in RGB5A1 format
- * @param bytes Pointer to encoded source color
- * @return Result color decoded as Common::Vec4<u8>
- */
-[[nodiscard]] inline Common::Vec4<u8> DecodeRGB5A1(const u8* bytes) {
-    u16_le pixel;
-    std::memcpy(&pixel, bytes, sizeof(pixel));
-    return {Convert5To8((pixel >> 11) & 0x1F), Convert5To8((pixel >> 6) & 0x1F),
-            Convert5To8((pixel >> 1) & 0x1F), Convert1To8(pixel & 0x1)};
-}
-
-/**
- * Decode a color stored in RGBA4 format
- * @param bytes Pointer to encoded source color
- * @return Result color decoded as Common::Vec4<u8>
- */
-[[nodiscard]] inline Common::Vec4<u8> DecodeRGBA4(const u8* bytes) {
-    u16_le pixel;
-    std::memcpy(&pixel, bytes, sizeof(pixel));
-    return {Convert4To8((pixel >> 12) & 0xF), Convert4To8((pixel >> 8) & 0xF),
-            Convert4To8((pixel >> 4) & 0xF), Convert4To8(pixel & 0xF)};
-}
-
-/**
- * Decode a depth value stored in D16 format
- * @param bytes Pointer to encoded source value
- * @return Depth value as an u32
- */
-[[nodiscard]] inline u32 DecodeD16(const u8* bytes) {
-    u16_le data;
-    std::memcpy(&data, bytes, sizeof(data));
-    return data;
-}
-
-/**
- * Decode a depth value stored in D24 format
- * @param bytes Pointer to encoded source value
- * @return Depth value as an u32
- */
-[[nodiscard]] inline u32 DecodeD24(const u8* bytes) {
-    return (bytes[2] << 16) | (bytes[1] << 8) | bytes[0];
-}
-
-/**
- * Decode a depth value and a stencil value stored in D24S8 format
- * @param bytes Pointer to encoded source values
- * @return Resulting values stored as a Common::Vec2
- */
-[[nodiscard]] inline Common::Vec2<u32> DecodeD24S8(const u8* bytes) {
-    return {static_cast<u32>((bytes[2] << 16) | (bytes[1] << 8) | bytes[0]), bytes[3]};
-}
-
-/**
- * Encode a color as RGBA8 format
- * @param color Source color to encode
- * @param bytes Destination pointer to store encoded color
- */
-inline void EncodeRGBA8(const Common::Vec4<u8>& color, u8* bytes) {
-    bytes[3] = color.r();
-    bytes[2] = color.g();
-    bytes[1] = color.b();
-    bytes[0] = color.a();
-}
-
-/**
- * Encode a color as RGB8 format
- * @param color Source color to encode
- * @param bytes Destination pointer to store encoded color
- */
-inline void EncodeRGB8(const Common::Vec4<u8>& color, u8* bytes) {
-    bytes[2] = color.r();
-    bytes[1] = color.g();
-    bytes[0] = color.b();
-}
-
-/**
- * Encode a color as RG8 (aka HILO8) format
- * @param color Source color to encode
- * @param bytes Destination pointer to store encoded color
- */
-inline void EncodeRG8(const Common::Vec4<u8>& color, u8* bytes) {
-    bytes[1] = color.r();
-    bytes[0] = color.g();
-}
-/**
- * Encode a color as RGB565 format
- * @param color Source color to encode
- * @param bytes Destination pointer to store encoded color
- */
-inline void EncodeRGB565(const Common::Vec4<u8>& color, u8* bytes) {
-    const u16_le data =
-        (Convert8To5(color.r()) << 11) | (Convert8To6(color.g()) << 5) | Convert8To5(color.b());
-
-    std::memcpy(bytes, &data, sizeof(data));
-}
-
-/**
- * Encode a color as RGB5A1 format
- * @param color Source color to encode
- * @param bytes Destination pointer to store encoded color
- */
-inline void EncodeRGB5A1(const Common::Vec4<u8>& color, u8* bytes) {
-    const u16_le data = (Convert8To5(color.r()) << 11) | (Convert8To5(color.g()) << 6) |
-                        (Convert8To5(color.b()) << 1) | Convert8To1(color.a());
-
-    std::memcpy(bytes, &data, sizeof(data));
-}
-
-/**
- * Encode a color as RGBA4 format
- * @param color Source color to encode
- * @param bytes Destination pointer to store encoded color
- */
-inline void EncodeRGBA4(const Common::Vec4<u8>& color, u8* bytes) {
-    const u16 data = (Convert8To4(color.r()) << 12) | (Convert8To4(color.g()) << 8) |
-                     (Convert8To4(color.b()) << 4) | Convert8To4(color.a());
-
-    std::memcpy(bytes, &data, sizeof(data));
-}
-
-/**
- * Encode a 16 bit depth value as D16 format
- * @param value 16 bit source depth value to encode
- * @param bytes Pointer where to store the encoded value
- */
-inline void EncodeD16(u32 value, u8* bytes) {
-    const u16_le data = static_cast<u16>(value);
-    std::memcpy(bytes, &data, sizeof(data));
-}
-
-/**
- * Encode a 24 bit depth value as D24 format
- * @param value 24 bit source depth value to encode
- * @param bytes Pointer where to store the encoded value
- */
-inline void EncodeD24(u32 value, u8* bytes) {
-    bytes[0] = value & 0xFF;
-    bytes[1] = (value >> 8) & 0xFF;
-    bytes[2] = (value >> 16) & 0xFF;
-}
-
-/**
- * Encode a 24 bit depth and 8 bit stencil values as D24S8 format
- * @param depth 24 bit source depth value to encode
- * @param stencil 8 bit source stencil value to encode
- * @param bytes Pointer where to store the encoded value
- */
-inline void EncodeD24S8(u32 depth, u8 stencil, u8* bytes) {
-    bytes[0] = depth & 0xFF;
-    bytes[1] = (depth >> 8) & 0xFF;
-    bytes[2] = (depth >> 16) & 0xFF;
-    bytes[3] = stencil;
-}
-
-/**
- * Encode a 24 bit depth value as D24X8 format (32 bits per pixel with 8 bits unused)
- * @param depth 24 bit source depth value to encode
- * @param bytes Pointer where to store the encoded value
- * @note unused bits will not be modified
- */
-inline void EncodeD24X8(u32 depth, u8* bytes) {
-    bytes[0] = depth & 0xFF;
-    bytes[1] = (depth >> 8) & 0xFF;
-    bytes[2] = (depth >> 16) & 0xFF;
-}
-
-/**
- * Encode an 8 bit stencil value as X24S8 format (32 bits per pixel with 24 bits unused)
- * @param stencil 8 bit source stencil value to encode
- * @param bytes Pointer where to store the encoded value
- * @note unused bits will not be modified
- */
-inline void EncodeX24S8(u8 stencil, u8* bytes) {
-    bytes[3] = stencil;
-}
-
-} // namespace Common::Color

src/common/common_funcs.h

@@ -24,10 +24,10 @@
 #define INSERT_PADDING_WORDS(num_words)                                                            \
     std::array<u32, num_words> CONCAT2(pad, __LINE__) {}
 
-/// These are similar to the INSERT_PADDING_* macros, but are needed for padding unions. This is
-/// because unions can only be initialized by one member.
-#define INSERT_UNION_PADDING_BYTES(num_bytes) std::array<u8, num_bytes> CONCAT2(pad, __LINE__)
-#define INSERT_UNION_PADDING_WORDS(num_words) std::array<u32, num_words> CONCAT2(pad, __LINE__)
+/// These are similar to the INSERT_PADDING_* macros but do not zero-initialize the contents.
+/// This keeps the structure trivial to construct.
+#define INSERT_PADDING_BYTES_NOINIT(num_bytes) std::array<u8, num_bytes> CONCAT2(pad, __LINE__)
+#define INSERT_PADDING_WORDS_NOINIT(num_words) std::array<u32, num_words> CONCAT2(pad, __LINE__)
 
 #ifndef _MSC_VER
 
@@ -93,6 +93,31 @@ __declspec(dllimport) void __stdcall DebugBreak(void);
         return static_cast<T>(key) == 0;                                                           \
     }
 
+/// Evaluates a boolean expression, and returns a result unless that expression is true.
+#define R_UNLESS(expr, res)                                                                        \
+    {                                                                                              \
+        if (!(expr)) {                                                                             \
+            if (res.IsError()) {                                                                   \
+                LOG_ERROR(Kernel, "Failed with result: {}", res.raw);                              \
+            }                                                                                      \
+            return res;                                                                            \
+        }                                                                                          \
+    }
+
+#define R_SUCCEEDED(res) (res.IsSuccess())
+
+/// Evaluates an expression that returns a result, and returns the result if it would fail.
+#define R_TRY(res_expr)                                                                            \
+    {                                                                                              \
+        const auto _tmp_r_try_rc = (res_expr);                                                     \
+        if (_tmp_r_try_rc.IsError()) {                                                             \
+            return _tmp_r_try_rc;                                                                  \
+        }                                                                                          \
+    }
+
+/// Evaluates a boolean expression, and succeeds if that expression is true.
+#define R_SUCCEED_IF(expr) R_UNLESS(!(expr), RESULT_SUCCESS)
+
 namespace Common {
 
 [[nodiscard]] constexpr u32 MakeMagic(char a, char b, char c, char d) {

src/common/concepts.h

@@ -31,4 +31,8 @@ concept DerivedFrom = requires {
     std::is_convertible_v<const volatile Derived*, const volatile Base*>;
 };
 
+// TODO: Replace with std::convertible_to when libc++ implements it.
+template <typename From, typename To>
+concept ConvertibleTo = std::is_convertible_v<From, To>;
+
 } // namespace Common

src/common/div_ceil.h

@@ -0,0 +1,26 @@
+// Copyright 2020 yuzu emulator team
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#pragma once
+
+#include <cstddef>
+#include <type_traits>
+
+namespace Common {
+
+/// Ceiled integer division.
+template <typename N, typename D>
+requires std::is_integral_v<N>&& std::is_unsigned_v<D>[[nodiscard]] constexpr N DivCeil(N number,
+                                                                                        D divisor) {
+    return static_cast<N>((static_cast<D>(number) + divisor - 1) / divisor);
+}
+
+/// Ceiled integer division with logarithmic divisor in base 2
+template <typename N, typename D>
+requires std::is_integral_v<N>&& std::is_unsigned_v<D>[[nodiscard]] constexpr N DivCeilLog2(
+    N value, D alignment_log2) {
+    return static_cast<N>((static_cast<D>(value) + (D(1) << alignment_log2) - 1) >> alignment_log2);
+}
+
+} // namespace Common

src/common/fiber.cpp

@@ -5,18 +5,20 @@
 #include "common/assert.h"
 #include "common/fiber.h"
 #include "common/spin_lock.h"
+#include "common/virtual_buffer.h"
 
-#if defined(_WIN32) || defined(WIN32)
-#include <windows.h>
-#else
 #include <boost/context/detail/fcontext.hpp>
-#endif
 
 namespace Common {
 
-constexpr std::size_t default_stack_size = 256 * 1024; // 256kb
+constexpr std::size_t default_stack_size = 256 * 1024;
 
 struct Fiber::FiberImpl {
+    FiberImpl() : stack{default_stack_size}, rewind_stack{default_stack_size} {}
+
+    VirtualBuffer<u8> stack;
+    VirtualBuffer<u8> rewind_stack;
+
     SpinLock guard{};
     std::function<void(void*)> entry_point;
     std::function<void(void*)> rewind_point;
@@ -26,17 +28,10 @@ struct Fiber::FiberImpl {
     bool is_thread_fiber{};
     bool released{};
 
-#if defined(_WIN32) || defined(WIN32)
-    LPVOID handle = nullptr;
-    LPVOID rewind_handle = nullptr;
-#else
-    alignas(64) std::array<u8, default_stack_size> stack;
-    alignas(64) std::array<u8, default_stack_size> rewind_stack;
-    u8* stack_limit;
-    u8* rewind_stack_limit;
-    boost::context::detail::fcontext_t context;
-    boost::context::detail::fcontext_t rewind_context;
-#endif
+    u8* stack_limit{};
+    u8* rewind_stack_limit{};
+    boost::context::detail::fcontext_t context{};
+    boost::context::detail::fcontext_t rewind_context{};
 };
 
 void Fiber::SetStartParameter(void* new_parameter) {
@@ -48,95 +43,6 @@ void Fiber::SetRewindPoint(std::function<void(void*)>&& rewind_func, void* rewin
     impl->rewind_parameter = rewind_param;
 }
 
-#if defined(_WIN32) || defined(WIN32)
-
-void Fiber::Start() {
-    ASSERT(impl->previous_fiber != nullptr);
-    impl->previous_fiber->impl->guard.unlock();
-    impl->previous_fiber.reset();
-    impl->entry_point(impl->start_parameter);
-    UNREACHABLE();
-}
-
-void Fiber::OnRewind() {
-    ASSERT(impl->handle != nullptr);
-    DeleteFiber(impl->handle);
-    impl->handle = impl->rewind_handle;
-    impl->rewind_handle = nullptr;
-    impl->rewind_point(impl->rewind_parameter);
-    UNREACHABLE();
-}
-
-void Fiber::FiberStartFunc(void* fiber_parameter) {
-    auto* fiber = static_cast<Fiber*>(fiber_parameter);
-    fiber->Start();
-}
-
-void Fiber::RewindStartFunc(void* fiber_parameter) {
-    auto* fiber = static_cast<Fiber*>(fiber_parameter);
-    fiber->OnRewind();
-}
-
-Fiber::Fiber(std::function<void(void*)>&& entry_point_func, void* start_parameter)
-    : impl{std::make_unique<FiberImpl>()} {
-    impl->entry_point = std::move(entry_point_func);
-    impl->start_parameter = start_parameter;
-    impl->handle = CreateFiber(default_stack_size, &FiberStartFunc, this);
-}
-
-Fiber::Fiber() : impl{std::make_unique<FiberImpl>()} {}
-
-Fiber::~Fiber() {
-    if (impl->released) {
-        return;
-    }
-    // Make sure the Fiber is not being used
-    const bool locked = impl->guard.try_lock();
-    ASSERT_MSG(locked, "Destroying a fiber that's still running");
-    if (locked) {
-        impl->guard.unlock();
-    }
-    DeleteFiber(impl->handle);
-}
-
-void Fiber::Exit() {
-    ASSERT_MSG(impl->is_thread_fiber, "Exitting non main thread fiber");
-    if (!impl->is_thread_fiber) {
-        return;
-    }
-    ConvertFiberToThread();
-    impl->guard.unlock();
-    impl->released = true;
-}
-
-void Fiber::Rewind() {
-    ASSERT(impl->rewind_point);
-    ASSERT(impl->rewind_handle == nullptr);
-    impl->rewind_handle = CreateFiber(default_stack_size, &RewindStartFunc, this);
-    SwitchToFiber(impl->rewind_handle);
-}
-
-void Fiber::YieldTo(std::shared_ptr<Fiber> from, std::shared_ptr<Fiber> to) {
-    ASSERT_MSG(from != nullptr, "Yielding fiber is null!");
-    ASSERT_MSG(to != nullptr, "Next fiber is null!");
-    to->impl->guard.lock();
-    to->impl->previous_fiber = from;
-    SwitchToFiber(to->impl->handle);
-    ASSERT(from->impl->previous_fiber != nullptr);
-    from->impl->previous_fiber->impl->guard.unlock();
-    from->impl->previous_fiber.reset();
-}
-
-std::shared_ptr<Fiber> Fiber::ThreadToFiber() {
-    std::shared_ptr<Fiber> fiber = std::shared_ptr<Fiber>{new Fiber()};
-    fiber->impl->guard.lock();
-    fiber->impl->handle = ConvertThreadToFiber(nullptr);
-    fiber->impl->is_thread_fiber = true;
-    return fiber;
-}
-
-#else
-
 void Fiber::Start(boost::context::detail::transfer_t& transfer) {
     ASSERT(impl->previous_fiber != nullptr);
     impl->previous_fiber->impl->context = transfer.fctx;
@@ -229,5 +135,4 @@ std::shared_ptr<Fiber> Fiber::ThreadToFiber() {
     return fiber;
 }
 
-#endif
 } // namespace Common

src/common/fiber.h

@@ -7,11 +7,9 @@
 #include <functional>
 #include <memory>
 
-#if !defined(_WIN32) && !defined(WIN32)
 namespace boost::context::detail {
 struct transfer_t;
 }
-#endif
 
 namespace Common {
 
@@ -59,17 +57,10 @@ public:
 private:
     Fiber();
 
-#if defined(_WIN32) || defined(WIN32)
-    void OnRewind();
-    void Start();
-    static void FiberStartFunc(void* fiber_parameter);
-    static void RewindStartFunc(void* fiber_parameter);
-#else
     void OnRewind(boost::context::detail::transfer_t& transfer);
     void Start(boost::context::detail::transfer_t& transfer);
     static void FiberStartFunc(boost::context::detail::transfer_t transfer);
     static void RewindStartFunc(boost::context::detail::transfer_t transfer);
-#endif
 
     struct FiberImpl;
     std::unique_ptr<FiberImpl> impl;

src/common/intrusive_red_black_tree.h

@@ -0,0 +1,602 @@
+// Copyright 2021 yuzu Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#pragma once
+
+#include "common/parent_of_member.h"
+#include "common/tree.h"
+
+namespace Common {
+
+namespace impl {
+
+class IntrusiveRedBlackTreeImpl;
+
+}
+
+struct IntrusiveRedBlackTreeNode {
+public:
+    using EntryType = RBEntry<IntrusiveRedBlackTreeNode>;
+
+    constexpr IntrusiveRedBlackTreeNode() = default;
+
+    void SetEntry(const EntryType& new_entry) {
+        entry = new_entry;
+    }
+
+    [[nodiscard]] EntryType& GetEntry() {
+        return entry;
+    }
+
+    [[nodiscard]] const EntryType& GetEntry() const {
+        return entry;
+    }
+
+private:
+    EntryType entry{};
+
+    friend class impl::IntrusiveRedBlackTreeImpl;
+
+    template <class, class, class>
+    friend class IntrusiveRedBlackTree;
+};
+
+template <class T, class Traits, class Comparator>
+class IntrusiveRedBlackTree;
+
+namespace impl {
+
+class IntrusiveRedBlackTreeImpl {
+private:
+    template <class, class, class>
+    friend class ::Common::IntrusiveRedBlackTree;
+
+    using RootType = RBHead<IntrusiveRedBlackTreeNode>;
+    RootType root;
+
+public:
+    template <bool Const>
+    class Iterator;
+
+    using value_type = IntrusiveRedBlackTreeNode;
+    using size_type = size_t;
+    using difference_type = ptrdiff_t;
+    using pointer = value_type*;
+    using const_pointer = const value_type*;
+    using reference = value_type&;
+    using const_reference = const value_type&;
+    using iterator = Iterator<false>;
+    using const_iterator = Iterator<true>;
+
+    template <bool Const>
+    class Iterator {
+    public:
+        using iterator_category = std::bidirectional_iterator_tag;
+        using value_type = typename IntrusiveRedBlackTreeImpl::value_type;
+        using difference_type = typename IntrusiveRedBlackTreeImpl::difference_type;
+        using pointer = std::conditional_t<Const, IntrusiveRedBlackTreeImpl::const_pointer,
+                                           IntrusiveRedBlackTreeImpl::pointer>;
+        using reference = std::conditional_t<Const, IntrusiveRedBlackTreeImpl::const_reference,
+                                             IntrusiveRedBlackTreeImpl::reference>;
+
+    private:
+        pointer node;
+
+    public:
+        explicit Iterator(pointer n) : node(n) {}
+
+        bool operator==(const Iterator& rhs) const {
+            return this->node == rhs.node;
+        }
+
+        bool operator!=(const Iterator& rhs) const {
+            return !(*this == rhs);
+        }
+
+        pointer operator->() const {
+            return this->node;
+        }
+
+        reference operator*() const {
+            return *this->node;
+        }
+
+        Iterator& operator++() {
+            this->node = GetNext(this->node);
+            return *this;
+        }
+
+        Iterator& operator--() {
+            this->node = GetPrev(this->node);
+            return *this;
+        }
+
+        Iterator operator++(int) {
+            const Iterator it{*this};
+            ++(*this);
+            return it;
+        }
+
+        Iterator operator--(int) {
+            const Iterator it{*this};
+            --(*this);
+            return it;
+        }
+
+        operator Iterator<true>() const {
+            return Iterator<true>(this->node);
+        }
+    };
+
+private:
+    // Define accessors using RB_* functions.
+    bool EmptyImpl() const {
+        return root.IsEmpty();
+    }
+
+    IntrusiveRedBlackTreeNode* GetMinImpl() const {
+        return RB_MIN(const_cast<RootType*>(&root));
+    }
+
+    IntrusiveRedBlackTreeNode* GetMaxImpl() const {
+        return RB_MAX(const_cast<RootType*>(&root));
+    }
+
+    IntrusiveRedBlackTreeNode* RemoveImpl(IntrusiveRedBlackTreeNode* node) {
+        return RB_REMOVE(&root, node);
+    }
+
+public:
+    static IntrusiveRedBlackTreeNode* GetNext(IntrusiveRedBlackTreeNode* node) {
+        return RB_NEXT(node);
+    }
+
+    static IntrusiveRedBlackTreeNode* GetPrev(IntrusiveRedBlackTreeNode* node) {
+        return RB_PREV(node);
+    }
+
+    static const IntrusiveRedBlackTreeNode* GetNext(const IntrusiveRedBlackTreeNode* node) {
+        return static_cast<const IntrusiveRedBlackTreeNode*>(
+            GetNext(const_cast<IntrusiveRedBlackTreeNode*>(node)));
+    }
+
+    static const IntrusiveRedBlackTreeNode* GetPrev(const IntrusiveRedBlackTreeNode* node) {
+        return static_cast<const IntrusiveRedBlackTreeNode*>(
+            GetPrev(const_cast<IntrusiveRedBlackTreeNode*>(node)));
+    }
+
+public:
+    constexpr IntrusiveRedBlackTreeImpl() {}
+
+    // Iterator accessors.
+    iterator begin() {
+        return iterator(this->GetMinImpl());
+    }
+
+    const_iterator begin() const {
+        return const_iterator(this->GetMinImpl());
+    }
+
+    iterator end() {
+        return iterator(static_cast<IntrusiveRedBlackTreeNode*>(nullptr));
+    }
+
+    const_iterator end() const {
+        return const_iterator(static_cast<const IntrusiveRedBlackTreeNode*>(nullptr));
+    }
+
+    const_iterator cbegin() const {
+        return this->begin();
+    }
+
+    const_iterator cend() const {
+        return this->end();
+    }
+
+    iterator iterator_to(reference ref) {
+        return iterator(&ref);
+    }
+
+    const_iterator iterator_to(const_reference ref) const {
+        return const_iterator(&ref);
+    }
+
+    // Content management.
+    bool empty() const {
+        return this->EmptyImpl();
+    }
+
+    reference back() {
+        return *this->GetMaxImpl();
+    }
+
+    const_reference back() const {
+        return *this->GetMaxImpl();
+    }
+
+    reference front() {
+        return *this->GetMinImpl();
+    }
+
+    const_reference front() const {
+        return *this->GetMinImpl();
+    }
+
+    iterator erase(iterator it) {
+        auto cur = std::addressof(*it);
+        auto next = GetNext(cur);
+        this->RemoveImpl(cur);
+        return iterator(next);
+    }
+};
+
+} // namespace impl
+
+template <typename T>
+concept HasLightCompareType = requires {
+    { std::is_same<typename T::LightCompareType, void>::value }
+    ->std::convertible_to<bool>;
+};
+
+namespace impl {
+
+template <typename T, typename Default>
+consteval auto* GetLightCompareType() {
+    if constexpr (HasLightCompareType<T>) {
+        return static_cast<typename T::LightCompareType*>(nullptr);
+    } else {
+        return static_cast<Default*>(nullptr);
+    }
+}
+
+} // namespace impl
+
+template <typename T, typename Default>
+using LightCompareType = std::remove_pointer_t<decltype(impl::GetLightCompareType<T, Default>())>;
+
+template <class T, class Traits, class Comparator>
+class IntrusiveRedBlackTree {
+
+public:
+    using ImplType = impl::IntrusiveRedBlackTreeImpl;
+
+private:
+    ImplType impl{};
+
+public:
+    template <bool Const>
+    class Iterator;
+
+    using value_type = T;
+    using size_type = size_t;
+    using difference_type = ptrdiff_t;
+    using pointer = T*;
+    using const_pointer = const T*;
+    using reference = T&;
+    using const_reference = const T&;
+    using iterator = Iterator<false>;
+    using const_iterator = Iterator<true>;
+
+    using light_value_type = LightCompareType<Comparator, value_type>;
+    using const_light_pointer = const light_value_type*;
+    using const_light_reference = const light_value_type&;
+
+    template <bool Const>
+    class Iterator {
+    public:
+        friend class IntrusiveRedBlackTree<T, Traits, Comparator>;
+
+        using ImplIterator =
+            std::conditional_t<Const, ImplType::const_iterator, ImplType::iterator>;
+
+        using iterator_category = std::bidirectional_iterator_tag;
+        using value_type = typename IntrusiveRedBlackTree::value_type;
+        using difference_type = typename IntrusiveRedBlackTree::difference_type;
+        using pointer = std::conditional_t<Const, IntrusiveRedBlackTree::const_pointer,
+                                           IntrusiveRedBlackTree::pointer>;
+        using reference = std::conditional_t<Const, IntrusiveRedBlackTree::const_reference,
+                                             IntrusiveRedBlackTree::reference>;
+
+    private:
+        ImplIterator iterator;
+
+    private:
+        explicit Iterator(ImplIterator it) : iterator(it) {}
+
+        explicit Iterator(typename std::conditional<Const, ImplType::const_iterator,
+                                                    ImplType::iterator>::type::pointer ptr)
+            : iterator(ptr) {}
+
+        ImplIterator GetImplIterator() const {
+            return this->iterator;
+        }
+
+    public:
+        bool operator==(const Iterator& rhs) const {
+            return this->iterator == rhs.iterator;
+        }
+
+        bool operator!=(const Iterator& rhs) const {
+            return !(*this == rhs);
+        }
+
+        pointer operator->() const {
+            return Traits::GetParent(std::addressof(*this->iterator));
+        }
+
+        reference operator*() const {
+            return *Traits::GetParent(std::addressof(*this->iterator));
+        }
+
+        Iterator& operator++() {
+            ++this->iterator;
+            return *this;
+        }
+
+        Iterator& operator--() {
+            --this->iterator;
+            return *this;
+        }
+
+        Iterator operator++(int) {
+            const Iterator it{*this};
+            ++this->iterator;
+            return it;
+        }
+
+        Iterator operator--(int) {
+            const Iterator it{*this};
+            --this->iterator;
+            return it;
+        }
+
+        operator Iterator<true>() const {
+            return Iterator<true>(this->iterator);
+        }
+    };
+
+private:
+    static int CompareImpl(const IntrusiveRedBlackTreeNode* lhs,
+                           const IntrusiveRedBlackTreeNode* rhs) {
+        return Comparator::Compare(*Traits::GetParent(lhs), *Traits::GetParent(rhs));
+    }
+
+    static int LightCompareImpl(const void* elm, const IntrusiveRedBlackTreeNode* rhs) {
+        return Comparator::Compare(*static_cast<const_light_pointer>(elm), *Traits::GetParent(rhs));
+    }
+
+    // Define accessors using RB_* functions.
+    IntrusiveRedBlackTreeNode* InsertImpl(IntrusiveRedBlackTreeNode* node) {
+        return RB_INSERT(&impl.root, node, CompareImpl);
+    }
+
+    IntrusiveRedBlackTreeNode* FindImpl(const IntrusiveRedBlackTreeNode* node) const {
+        return RB_FIND(const_cast<ImplType::RootType*>(&impl.root),
+                       const_cast<IntrusiveRedBlackTreeNode*>(node), CompareImpl);
+    }
+
+    IntrusiveRedBlackTreeNode* NFindImpl(const IntrusiveRedBlackTreeNode* node) const {
+        return RB_NFIND(const_cast<ImplType::RootType*>(&impl.root),
+                        const_cast<IntrusiveRedBlackTreeNode*>(node), CompareImpl);
+    }
+
+    IntrusiveRedBlackTreeNode* FindLightImpl(const_light_pointer lelm) const {
+        return RB_FIND_LIGHT(const_cast<ImplType::RootType*>(&impl.root),
+                             static_cast<const void*>(lelm), LightCompareImpl);
+    }
+
+    IntrusiveRedBlackTreeNode* NFindLightImpl(const_light_pointer lelm) const {
+        return RB_NFIND_LIGHT(const_cast<ImplType::RootType*>(&impl.root),
+                              static_cast<const void*>(lelm), LightCompareImpl);
+    }
+
+public:
+    constexpr IntrusiveRedBlackTree() = default;
+
+    // Iterator accessors.
+    iterator begin() {
+        return iterator(this->impl.begin());
+    }
+
+    const_iterator begin() const {
+        return const_iterator(this->impl.begin());
+    }
+
+    iterator end() {
+        return iterator(this->impl.end());
+    }
+
+    const_iterator end() const {
+        return const_iterator(this->impl.end());
+    }
+
+    const_iterator cbegin() const {
+        return this->begin();
+    }
+
+    const_iterator cend() const {
+        return this->end();
+    }
+
+    iterator iterator_to(reference ref) {
+        return iterator(this->impl.iterator_to(*Traits::GetNode(std::addressof(ref))));
+    }
+
+    const_iterator iterator_to(const_reference ref) const {
+        return const_iterator(this->impl.iterator_to(*Traits::GetNode(std::addressof(ref))));
+    }
+
+    // Content management.
+    bool empty() const {
+        return this->impl.empty();
+    }
+
+    reference back() {
+        return *Traits::GetParent(std::addressof(this->impl.back()));
+    }
+
+    const_reference back() const {
+        return *Traits::GetParent(std::addressof(this->impl.back()));
+    }
+
+    reference front() {
+        return *Traits::GetParent(std::addressof(this->impl.front()));
+    }
+
+    const_reference front() const {
+        return *Traits::GetParent(std::addressof(this->impl.front()));
+    }
+
+    iterator erase(iterator it) {
+        return iterator(this->impl.erase(it.GetImplIterator()));
+    }
+
+    iterator insert(reference ref) {
+        ImplType::pointer node = Traits::GetNode(std::addressof(ref));
+        this->InsertImpl(node);
+        return iterator(node);
+    }
+
+    iterator find(const_reference ref) const {
+        return iterator(this->FindImpl(Traits::GetNode(std::addressof(ref))));
+    }
+
+    iterator nfind(const_reference ref) const {
+        return iterator(this->NFindImpl(Traits::GetNode(std::addressof(ref))));
+    }
+
+    iterator find_light(const_light_reference ref) const {
+        return iterator(this->FindLightImpl(std::addressof(ref)));
+    }
+
+    iterator nfind_light(const_light_reference ref) const {
+        return iterator(this->NFindLightImpl(std::addressof(ref)));
+    }
+};
+
+template <auto T, class Derived = impl::GetParentType<T>>
+class IntrusiveRedBlackTreeMemberTraits;
+
+template <class Parent, IntrusiveRedBlackTreeNode Parent::*Member, class Derived>
+class IntrusiveRedBlackTreeMemberTraits<Member, Derived> {
+public:
+    template <class Comparator>
+    using TreeType = IntrusiveRedBlackTree<Derived, IntrusiveRedBlackTreeMemberTraits, Comparator>;
+    using TreeTypeImpl = impl::IntrusiveRedBlackTreeImpl;
+
+private:
+    template <class, class, class>
+    friend class IntrusiveRedBlackTree;
+
+    friend class impl::IntrusiveRedBlackTreeImpl;
+
+    static constexpr IntrusiveRedBlackTreeNode* GetNode(Derived* parent) {
+        return std::addressof(parent->*Member);
+    }
+
+    static constexpr IntrusiveRedBlackTreeNode const* GetNode(Derived const* parent) {
+        return std::addressof(parent->*Member);
+    }
+
+    static constexpr Derived* GetParent(IntrusiveRedBlackTreeNode* node) {
+        return GetParentPointer<Member, Derived>(node);
+    }
+
+    static constexpr Derived const* GetParent(const IntrusiveRedBlackTreeNode* node) {
+        return GetParentPointer<Member, Derived>(node);
+    }
+
+private:
+    static constexpr TypedStorage<Derived> DerivedStorage = {};
+    static_assert(GetParent(GetNode(GetPointer(DerivedStorage))) == GetPointer(DerivedStorage));
+};
+
+template <auto T, class Derived = impl::GetParentType<T>>
+class IntrusiveRedBlackTreeMemberTraitsDeferredAssert;
+
+template <class Parent, IntrusiveRedBlackTreeNode Parent::*Member, class Derived>
+class IntrusiveRedBlackTreeMemberTraitsDeferredAssert<Member, Derived> {
+public:
+    template <class Comparator>
+    using TreeType =
+        IntrusiveRedBlackTree<Derived, IntrusiveRedBlackTreeMemberTraitsDeferredAssert, Comparator>;
+    using TreeTypeImpl = impl::IntrusiveRedBlackTreeImpl;
+
+    static constexpr bool IsValid() {
+        TypedStorage<Derived> DerivedStorage = {};
+        return GetParent(GetNode(GetPointer(DerivedStorage))) == GetPointer(DerivedStorage);
+    }
+
+private:
+    template <class, class, class>
+    friend class IntrusiveRedBlackTree;
+
+    friend class impl::IntrusiveRedBlackTreeImpl;
+
+    static constexpr IntrusiveRedBlackTreeNode* GetNode(Derived* parent) {
+        return std::addressof(parent->*Member);
+    }
+
+    static constexpr IntrusiveRedBlackTreeNode const* GetNode(Derived const* parent) {
+        return std::addressof(parent->*Member);
+    }
+
+    static constexpr Derived* GetParent(IntrusiveRedBlackTreeNode* node) {
+        return GetParentPointer<Member, Derived>(node);
+    }
+
+    static constexpr Derived const* GetParent(const IntrusiveRedBlackTreeNode* node) {
+        return GetParentPointer<Member, Derived>(node);
+    }
+};
+
+template <class Derived>
+class IntrusiveRedBlackTreeBaseNode : public IntrusiveRedBlackTreeNode {
+public:
+    constexpr Derived* GetPrev() {
+        return static_cast<Derived*>(impl::IntrusiveRedBlackTreeImpl::GetPrev(this));
+    }
+    constexpr const Derived* GetPrev() const {
+        return static_cast<const Derived*>(impl::IntrusiveRedBlackTreeImpl::GetPrev(this));
+    }
+
+    constexpr Derived* GetNext() {
+        return static_cast<Derived*>(impl::IntrusiveRedBlackTreeImpl::GetNext(this));
+    }
+    constexpr const Derived* GetNext() const {
+        return static_cast<const Derived*>(impl::IntrusiveRedBlackTreeImpl::GetNext(this));
+    }
+};
+
+template <class Derived>
+class IntrusiveRedBlackTreeBaseTraits {
+public:
+    template <class Comparator>
+    using TreeType = IntrusiveRedBlackTree<Derived, IntrusiveRedBlackTreeBaseTraits, Comparator>;
+    using TreeTypeImpl = impl::IntrusiveRedBlackTreeImpl;
+
+private:
+    template <class, class, class>
+    friend class IntrusiveRedBlackTree;
+
+    friend class impl::IntrusiveRedBlackTreeImpl;
+
+    static constexpr IntrusiveRedBlackTreeNode* GetNode(Derived* parent) {
+        return static_cast<IntrusiveRedBlackTreeNode*>(parent);
+    }
+
+    static constexpr IntrusiveRedBlackTreeNode const* GetNode(Derived const* parent) {
+        return static_cast<const IntrusiveRedBlackTreeNode*>(parent);
+    }
+
+    static constexpr Derived* GetParent(IntrusiveRedBlackTreeNode* node) {
+        return static_cast<Derived*>(node);
+    }
+
+    static constexpr Derived const* GetParent(const IntrusiveRedBlackTreeNode* node) {
+        return static_cast<const Derived*>(node);
+    }
+};
+
+} // namespace Common

src/common/logging/backend.cpp

@@ -23,6 +23,7 @@
 #include "common/logging/text_formatter.h"
 #include "common/string_util.h"
 #include "common/threadsafe_queue.h"
+#include "core/settings.h"
 
 namespace Log {
 
@@ -144,18 +145,35 @@ void ColorConsoleBackend::Write(const Entry& entry) {
     PrintColoredMessage(entry);
 }
 
-// _SH_DENYWR allows read only access to the file for other programs.
-// It is #defined to 0 on other platforms
-FileBackend::FileBackend(const std::string& filename)
-    : file(filename, "w", _SH_DENYWR), bytes_written(0) {}
+FileBackend::FileBackend(const std::string& filename) : bytes_written(0) {
+    if (Common::FS::Exists(filename + ".old.txt")) {
+        Common::FS::Delete(filename + ".old.txt");
+    }
+    if (Common::FS::Exists(filename)) {
+        Common::FS::Rename(filename, filename + ".old.txt");
+    }
+
+    // _SH_DENYWR allows read only access to the file for other programs.
+    // It is #defined to 0 on other platforms
+    file = Common::FS::IOFile(filename, "w", _SH_DENYWR);
+}
 
 void FileBackend::Write(const Entry& entry) {
     // prevent logs from going over the maximum size (in case its spamming and the user doesn't
     // know)
-    constexpr std::size_t MAX_BYTES_WRITTEN = 50 * 1024L * 1024L;
-    if (!file.IsOpen() || bytes_written > MAX_BYTES_WRITTEN) {
+    constexpr std::size_t MAX_BYTES_WRITTEN = 100 * 1024 * 1024;
+    constexpr std::size_t MAX_BYTES_WRITTEN_EXTENDED = 1024 * 1024 * 1024;
+
+    if (!file.IsOpen()) {
         return;
     }
+
+    if (Settings::values.extended_logging && bytes_written > MAX_BYTES_WRITTEN_EXTENDED) {
+        return;
+    } else if (!Settings::values.extended_logging && bytes_written > MAX_BYTES_WRITTEN) {
+        return;
+    }
+
     bytes_written += file.WriteString(FormatLogMessage(entry).append(1, '\n'));
     if (entry.log_level >= Level::Error) {
         file.Flush();

src/common/memory_hook.cpp

@@ -1,11 +0,0 @@
-// Copyright 2018 Citra Emulator Project
-// Licensed under GPLv2 or any later version
-// Refer to the license.txt file included.
-
-#include "common/memory_hook.h"
-
-namespace Common {
-
-MemoryHook::~MemoryHook() = default;
-
-} // namespace Common

src/common/memory_hook.h

@@ -1,47 +0,0 @@
-// Copyright 2016 Citra Emulator Project
-// Licensed under GPLv2 or any later version
-// Refer to the license.txt file included.
-
-#pragma once
-
-#include <memory>
-#include <optional>
-
-#include "common/common_types.h"
-
-namespace Common {
-
-/**
- * Memory hooks have two purposes:
- * 1. To allow reads and writes to a region of memory to be intercepted. This is used to implement
- *    texture forwarding and memory breakpoints for debugging.
- * 2. To allow for the implementation of MMIO devices.
- *
- * A hook may be mapped to multiple regions of memory.
- *
- * If a std::nullopt or false is returned from a function, the read/write request is passed through
- * to the underlying memory region.
- */
-class MemoryHook {
-public:
-    virtual ~MemoryHook();
-
-    virtual std::optional<bool> IsValidAddress(VAddr addr) = 0;
-
-    virtual std::optional<u8> Read8(VAddr addr) = 0;
-    virtual std::optional<u16> Read16(VAddr addr) = 0;
-    virtual std::optional<u32> Read32(VAddr addr) = 0;
-    virtual std::optional<u64> Read64(VAddr addr) = 0;
-
-    virtual bool ReadBlock(VAddr src_addr, void* dest_buffer, std::size_t size) = 0;
-
-    virtual bool Write8(VAddr addr, u8 data) = 0;
-    virtual bool Write16(VAddr addr, u16 data) = 0;
-    virtual bool Write32(VAddr addr, u32 data) = 0;
-    virtual bool Write64(VAddr addr, u64 data) = 0;
-
-    virtual bool WriteBlock(VAddr dest_addr, const void* src_buffer, std::size_t size) = 0;
-};
-
-using MemoryHookPointer = std::shared_ptr<MemoryHook>;
-} // namespace Common

src/common/multi_level_queue.h

@@ -1,345 +0,0 @@
-// Copyright 2019 TuxSH
-// Licensed under GPLv2 or any later version
-// Refer to the license.txt file included.
-
-#pragma once
-
-#include <array>
-#include <iterator>
-#include <list>
-#include <utility>
-
-#include "common/bit_util.h"
-#include "common/common_types.h"
-
-namespace Common {
-
-/**
- * A MultiLevelQueue is a type of priority queue which has the following characteristics:
- * - iteratable through each of its elements.
- * - back can be obtained.
- * - O(1) add, lookup (both front and back)
- * - discrete priorities and a max of 64 priorities (limited domain)
- * This type of priority queue is normaly used for managing threads within an scheduler
- */
-template <typename T, std::size_t Depth>
-class MultiLevelQueue {
-public:
-    using value_type = T;
-    using reference = value_type&;
-    using const_reference = const value_type&;
-    using pointer = value_type*;
-    using const_pointer = const value_type*;
-
-    using difference_type = typename std::pointer_traits<pointer>::difference_type;
-    using size_type = std::size_t;
-
-    template <bool is_constant>
-    class iterator_impl {
-    public:
-        using iterator_category = std::bidirectional_iterator_tag;
-        using value_type = T;
-        using pointer = std::conditional_t<is_constant, T*, const T*>;
-        using reference = std::conditional_t<is_constant, const T&, T&>;
-        using difference_type = typename std::pointer_traits<pointer>::difference_type;
-
-        friend bool operator==(const iterator_impl& lhs, const iterator_impl& rhs) {
-            if (lhs.IsEnd() && rhs.IsEnd())
-                return true;
-            return std::tie(lhs.current_priority, lhs.it) == std::tie(rhs.current_priority, rhs.it);
-        }
-
-        friend bool operator!=(const iterator_impl& lhs, const iterator_impl& rhs) {
-            return !operator==(lhs, rhs);
-        }
-
-        reference operator*() const {
-            return *it;
-        }
-
-        pointer operator->() const {
-            return it.operator->();
-        }
-
-        iterator_impl& operator++() {
-            if (IsEnd()) {
-                return *this;
-            }
-
-            ++it;
-
-            if (it == GetEndItForPrio()) {
-                u64 prios = mlq.used_priorities;
-                prios &= ~((1ULL << (current_priority + 1)) - 1);
-                if (prios == 0) {
-                    current_priority = static_cast<u32>(mlq.depth());
-                } else {
-                    current_priority = CountTrailingZeroes64(prios);
-                    it = GetBeginItForPrio();
-                }
-            }
-            return *this;
-        }
-
-        iterator_impl& operator--() {
-            if (IsEnd()) {
-                if (mlq.used_priorities != 0) {
-                    current_priority = 63 - CountLeadingZeroes64(mlq.used_priorities);
-                    it = GetEndItForPrio();
-                    --it;
-                }
-            } else if (it == GetBeginItForPrio()) {
-                u64 prios = mlq.used_priorities;
-                prios &= (1ULL << current_priority) - 1;
-                if (prios != 0) {
-                    current_priority = CountTrailingZeroes64(prios);
-                    it = GetEndItForPrio();
-                    --it;
-                }
-            } else {
-                --it;
-            }
-            return *this;
-        }
-
-        iterator_impl operator++(int) {
-            const iterator_impl v{*this};
-            ++(*this);
-            return v;
-        }
-
-        iterator_impl operator--(int) {
-            const iterator_impl v{*this};
-            --(*this);
-            return v;
-        }
-
-        // allow implicit const->non-const
-        iterator_impl(const iterator_impl<false>& other)
-            : mlq(other.mlq), it(other.it), current_priority(other.current_priority) {}
-
-        iterator_impl(const iterator_impl<true>& other)
-            : mlq(other.mlq), it(other.it), current_priority(other.current_priority) {}
-
-        iterator_impl& operator=(const iterator_impl<false>& other) {
-            mlq = other.mlq;
-            it = other.it;
-            current_priority = other.current_priority;
-            return *this;
-        }
-
-        friend class iterator_impl<true>;
-        iterator_impl() = default;
-
-    private:
-        friend class MultiLevelQueue;
-        using container_ref =
-            std::conditional_t<is_constant, const MultiLevelQueue&, MultiLevelQueue&>;
-        using list_iterator = std::conditional_t<is_constant, typename std::list<T>::const_iterator,
-                                                 typename std::list<T>::iterator>;
-
-        explicit iterator_impl(container_ref mlq, list_iterator it, u32 current_priority)
-            : mlq(mlq), it(it), current_priority(current_priority) {}
-        explicit iterator_impl(container_ref mlq, u32 current_priority)
-            : mlq(mlq), it(), current_priority(current_priority) {}
-
-        bool IsEnd() const {
-            return current_priority == mlq.depth();
-        }
-
-        list_iterator GetBeginItForPrio() const {
-            return mlq.levels[current_priority].begin();
-        }
-
-        list_iterator GetEndItForPrio() const {
-            return mlq.levels[current_priority].end();
-        }
-
-        container_ref mlq;
-        list_iterator it;
-        u32 current_priority;
-    };
-
-    using iterator = iterator_impl<false>;
-    using const_iterator = iterator_impl<true>;
-
-    void add(const T& element, u32 priority, bool send_back = true) {
-        if (send_back)
-            levels[priority].push_back(element);
-        else
-            levels[priority].push_front(element);
-        used_priorities |= 1ULL << priority;
-    }
-
-    void remove(const T& element, u32 priority) {
-        auto it = ListIterateTo(levels[priority], element);
-        if (it == levels[priority].end())
-            return;
-        levels[priority].erase(it);
-        if (levels[priority].empty()) {
-            used_priorities &= ~(1ULL << priority);
-        }
-    }
-
-    void adjust(const T& element, u32 old_priority, u32 new_priority, bool adjust_front = false) {
-        remove(element, old_priority);
-        add(element, new_priority, !adjust_front);
-    }
-    void adjust(const_iterator it, u32 old_priority, u32 new_priority, bool adjust_front = false) {
-        adjust(*it, old_priority, new_priority, adjust_front);
-    }
-
-    void transfer_to_front(const T& element, u32 priority, MultiLevelQueue& other) {
-        ListSplice(other.levels[priority], other.levels[priority].begin(), levels[priority],
-                   ListIterateTo(levels[priority], element));
-
-        other.used_priorities |= 1ULL << priority;
-
-        if (levels[priority].empty()) {
-            used_priorities &= ~(1ULL << priority);
-        }
-    }
-
-    void transfer_to_front(const_iterator it, u32 priority, MultiLevelQueue& other) {
-        transfer_to_front(*it, priority, other);
-    }
-
-    void transfer_to_back(const T& element, u32 priority, MultiLevelQueue& other) {
-        ListSplice(other.levels[priority], other.levels[priority].end(), levels[priority],
-                   ListIterateTo(levels[priority], element));
-
-        other.used_priorities |= 1ULL << priority;
-
-        if (levels[priority].empty()) {
-            used_priorities &= ~(1ULL << priority);
-        }
-    }
-
-    void transfer_to_back(const_iterator it, u32 priority, MultiLevelQueue& other) {
-        transfer_to_back(*it, priority, other);
-    }
-
-    void yield(u32 priority, std::size_t n = 1) {
-        ListShiftForward(levels[priority], n);
-    }
-
-    [[nodiscard]] std::size_t depth() const {
-        return Depth;
-    }
-
-    [[nodiscard]] std::size_t size(u32 priority) const {
-        return levels[priority].size();
-    }
-
-    [[nodiscard]] std::size_t size() const {
-        u64 priorities = used_priorities;
-        std::size_t size = 0;
-        while (priorities != 0) {
-            const u64 current_priority = CountTrailingZeroes64(priorities);
-            size += levels[current_priority].size();
-            priorities &= ~(1ULL << current_priority);
-        }
-        return size;
-    }
-
-    [[nodiscard]] bool empty() const {
-        return used_priorities == 0;
-    }
-
-    [[nodiscard]] bool empty(u32 priority) const {
-        return (used_priorities & (1ULL << priority)) == 0;
-    }
-
-    [[nodiscard]] u32 highest_priority_set(u32 max_priority = 0) const {
-        const u64 priorities =
-            max_priority == 0 ? used_priorities : (used_priorities & ~((1ULL << max_priority) - 1));
-        return priorities == 0 ? Depth : static_cast<u32>(CountTrailingZeroes64(priorities));
-    }
-
-    [[nodiscard]] u32 lowest_priority_set(u32 min_priority = Depth - 1) const {
-        const u64 priorities = min_priority >= Depth - 1
-                                   ? used_priorities
-                                   : (used_priorities & ((1ULL << (min_priority + 1)) - 1));
-        return priorities == 0 ? Depth : 63 - CountLeadingZeroes64(priorities);
-    }
-
-    [[nodiscard]] const_iterator cbegin(u32 max_prio = 0) const {
-        const u32 priority = highest_priority_set(max_prio);
-        return priority == Depth ? cend()
-                                 : const_iterator{*this, levels[priority].cbegin(), priority};
-    }
-    [[nodiscard]] const_iterator begin(u32 max_prio = 0) const {
-        return cbegin(max_prio);
-    }
-    [[nodiscard]] iterator begin(u32 max_prio = 0) {
-        const u32 priority = highest_priority_set(max_prio);
-        return priority == Depth ? end() : iterator{*this, levels[priority].begin(), priority};
-    }
-
-    [[nodiscard]] const_iterator cend(u32 min_prio = Depth - 1) const {
-        return min_prio == Depth - 1 ? const_iterator{*this, Depth} : cbegin(min_prio + 1);
-    }
-    [[nodiscard]] const_iterator end(u32 min_prio = Depth - 1) const {
-        return cend(min_prio);
-    }
-    [[nodiscard]] iterator end(u32 min_prio = Depth - 1) {
-        return min_prio == Depth - 1 ? iterator{*this, Depth} : begin(min_prio + 1);
-    }
-
-    [[nodiscard]] T& front(u32 max_priority = 0) {
-        const u32 priority = highest_priority_set(max_priority);
-        return levels[priority == Depth ? 0 : priority].front();
-    }
-    [[nodiscard]] const T& front(u32 max_priority = 0) const {
-        const u32 priority = highest_priority_set(max_priority);
-        return levels[priority == Depth ? 0 : priority].front();
-    }
-
-    [[nodiscard]] T& back(u32 min_priority = Depth - 1) {
-        const u32 priority = lowest_priority_set(min_priority); // intended
-        return levels[priority == Depth ? 63 : priority].back();
-    }
-    [[nodiscard]] const T& back(u32 min_priority = Depth - 1) const {
-        const u32 priority = lowest_priority_set(min_priority); // intended
-        return levels[priority == Depth ? 63 : priority].back();
-    }
-
-    void clear() {
-        used_priorities = 0;
-        for (std::size_t i = 0; i < Depth; i++) {
-            levels[i].clear();
-        }
-    }
-
-private:
-    using const_list_iterator = typename std::list<T>::const_iterator;
-
-    static void ListShiftForward(std::list<T>& list, const std::size_t shift = 1) {
-        if (shift >= list.size()) {
-            return;
-        }
-
-        const auto begin_range = list.begin();
-        const auto end_range = std::next(begin_range, shift);
-        list.splice(list.end(), list, begin_range, end_range);
-    }
-
-    static void ListSplice(std::list<T>& in_list, const_list_iterator position,
-                           std::list<T>& out_list, const_list_iterator element) {
-        in_list.splice(position, out_list, element);
-    }
-
-    [[nodiscard]] static const_list_iterator ListIterateTo(const std::list<T>& list,
-                                                           const T& element) {
-        auto it = list.cbegin();
-        while (it != list.cend() && *it != element) {
-            ++it;
-        }
-        return it;
-    }
-
-    std::array<std::list<T>, Depth> levels;
-    u64 used_priorities = 0;
-};
-
-} // namespace Common

src/common/nvidia_flags.cpp

@@ -0,0 +1,27 @@
+// Copyright 2021 yuzu Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#include <filesystem>
+#include <stdlib.h>
+
+#include <fmt/format.h>
+
+#include "common/file_util.h"
+#include "common/nvidia_flags.h"
+
+namespace Common {
+
+void ConfigureNvidiaEnvironmentFlags() {
+#ifdef _WIN32
+    const std::string shader_path = Common::FS::SanitizePath(
+        fmt::format("{}/nvidia/", Common::FS::GetUserPath(Common::FS::UserPath::ShaderDir)));
+    const std::string windows_path =
+        Common::FS::SanitizePath(shader_path, Common::FS::DirectorySeparator::BackwardSlash);
+    void(Common::FS::CreateFullPath(shader_path + '/'));
+    void(_putenv(fmt::format("__GL_SHADER_DISK_CACHE_PATH={}", windows_path).c_str()));
+    void(_putenv("__GL_SHADER_DISK_CACHE_SKIP_CLEANUP=1"));
+#endif
+}
+
+} // namespace Common

src/common/nvidia_flags.h

@@ -0,0 +1,10 @@
+// Copyright 2021 yuzu Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+namespace Common {
+
+/// Configure platform specific flags for Nvidia's driver
+void ConfigureNvidiaEnvironmentFlags();
+
+} // namespace Common

src/common/page_table.cpp

@@ -10,16 +10,10 @@ PageTable::PageTable() = default;
 
 PageTable::~PageTable() noexcept = default;
 
-void PageTable::Resize(std::size_t address_space_width_in_bits, std::size_t page_size_in_bits,
-                       bool has_attribute) {
-    const std::size_t num_page_table_entries{1ULL
-                                             << (address_space_width_in_bits - page_size_in_bits)};
+void PageTable::Resize(size_t address_space_width_in_bits, size_t page_size_in_bits) {
+    const size_t num_page_table_entries{1ULL << (address_space_width_in_bits - page_size_in_bits)};
     pointers.resize(num_page_table_entries);
     backing_addr.resize(num_page_table_entries);
-
-    if (has_attribute) {
-        attributes.resize(num_page_table_entries);
-    }
 }
 
 } // namespace Common

src/common/page_table.h

@@ -4,10 +4,10 @@
 
 #pragma once
 
+#include <atomic>
 #include <tuple>
 
 #include "common/common_types.h"
-#include "common/memory_hook.h"
 #include "common/virtual_buffer.h"
 
 namespace Common {
@@ -20,27 +20,6 @@ enum class PageType : u8 {
     /// Page is mapped to regular memory, but also needs to check for rasterizer cache flushing and
     /// invalidation
     RasterizerCachedMemory,
-    /// Page is mapped to a I/O region. Writing and reading to this page is handled by functions.
-    Special,
-    /// Page is allocated for use.
-    Allocated,
-};
-
-struct SpecialRegion {
-    enum class Type {
-        DebugHook,
-        IODevice,
-    } type;
-
-    MemoryHookPointer handler;
-
-    [[nodiscard]] bool operator<(const SpecialRegion& other) const {
-        return std::tie(type, handler) < std::tie(other.type, other.handler);
-    }
-
-    [[nodiscard]] bool operator==(const SpecialRegion& other) const {
-        return std::tie(type, handler) == std::tie(other.type, other.handler);
-    }
 };
 
 /**
@@ -48,6 +27,59 @@ struct SpecialRegion {
  * mimics the way a real CPU page table works.
  */
 struct PageTable {
+    /// Number of bits reserved for attribute tagging.
+    /// This can be at most the guaranteed alignment of the pointers in the page table.
+    static constexpr int ATTRIBUTE_BITS = 2;
+
+    /**
+     * Pair of host pointer and page type attribute.
+     * This uses the lower bits of a given pointer to store the attribute tag.
+     * Writing and reading the pointer attribute pair is guaranteed to be atomic for the same method
+     * call. In other words, they are guaranteed to be synchronized at all times.
+     */
+    class PageInfo {
+    public:
+        /// Returns the page pointer
+        [[nodiscard]] u8* Pointer() const noexcept {
+            return ExtractPointer(raw.load(std::memory_order_relaxed));
+        }
+
+        /// Returns the page type attribute
+        [[nodiscard]] PageType Type() const noexcept {
+            return ExtractType(raw.load(std::memory_order_relaxed));
+        }
+
+        /// Returns the page pointer and attribute pair, extracted from the same atomic read
+        [[nodiscard]] std::pair<u8*, PageType> PointerType() const noexcept {
+            const uintptr_t non_atomic_raw = raw.load(std::memory_order_relaxed);
+            return {ExtractPointer(non_atomic_raw), ExtractType(non_atomic_raw)};
+        }
+
+        /// Returns the raw representation of the page information.
+        /// Use ExtractPointer and ExtractType to unpack the value.
+        [[nodiscard]] uintptr_t Raw() const noexcept {
+            return raw.load(std::memory_order_relaxed);
+        }
+
+        /// Write a page pointer and type pair atomically
+        void Store(u8* pointer, PageType type) noexcept {
+            raw.store(reinterpret_cast<uintptr_t>(pointer) | static_cast<uintptr_t>(type));
+        }
+
+        /// Unpack a pointer from a page info raw representation
+        [[nodiscard]] static u8* ExtractPointer(uintptr_t raw) noexcept {
+            return reinterpret_cast<u8*>(raw & (~uintptr_t{0} << ATTRIBUTE_BITS));
+        }
+
+        /// Unpack a page type from a page info raw representation
+        [[nodiscard]] static PageType ExtractType(uintptr_t raw) noexcept {
+            return static_cast<PageType>(raw & ((uintptr_t{1} << ATTRIBUTE_BITS) - 1));
+        }
+
+    private:
+        std::atomic<uintptr_t> raw;
+    };
+
     PageTable();
     ~PageTable() noexcept;
 
@@ -58,25 +90,21 @@ struct PageTable {
     PageTable& operator=(PageTable&&) noexcept = default;
 
     /**
-     * Resizes the page table to be able to accomodate enough pages within
+     * Resizes the page table to be able to accommodate enough pages within
      * a given address space.
      *
      * @param address_space_width_in_bits The address size width in bits.
      * @param page_size_in_bits           The page size in bits.
-     * @param has_attribute               Whether or not this page has any backing attributes.
      */
-    void Resize(std::size_t address_space_width_in_bits, std::size_t page_size_in_bits,
-                bool has_attribute);
+    void Resize(size_t address_space_width_in_bits, size_t page_size_in_bits);
 
     /**
      * Vector of memory pointers backing each page. An entry can only be non-null if the
-     * corresponding entry in the `attributes` vector is of type `Memory`.
+     * corresponding attribute element is of type `Memory`.
      */
-    VirtualBuffer<u8*> pointers;
+    VirtualBuffer<PageInfo> pointers;
 
     VirtualBuffer<u64> backing_addr;
-
-    VirtualBuffer<PageType> attributes;
 };
 
 } // namespace Common

src/common/parent_of_member.h

@@ -0,0 +1,191 @@
+// Copyright 2021 yuzu Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#pragma once
+
+#include <type_traits>
+
+#include "common/assert.h"
+#include "common/common_types.h"
+
+namespace Common {
+namespace detail {
+template <typename T, size_t Size, size_t Align>
+struct TypedStorageImpl {
+    std::aligned_storage_t<Size, Align> storage_;
+};
+} // namespace detail
+
+template <typename T>
+using TypedStorage = detail::TypedStorageImpl<T, sizeof(T), alignof(T)>;
+
+template <typename T>
+static constexpr T* GetPointer(TypedStorage<T>& ts) {
+    return static_cast<T*>(static_cast<void*>(std::addressof(ts.storage_)));
+}
+
+template <typename T>
+static constexpr const T* GetPointer(const TypedStorage<T>& ts) {
+    return static_cast<const T*>(static_cast<const void*>(std::addressof(ts.storage_)));
+}
+
+namespace impl {
+
+template <size_t MaxDepth>
+struct OffsetOfUnionHolder {
+    template <typename ParentType, typename MemberType, size_t Offset>
+    union UnionImpl {
+        using PaddingMember = char;
+        static constexpr size_t GetOffset() {
+            return Offset;
+        }
+
+#pragma pack(push, 1)
+        struct {
+            PaddingMember padding[Offset];
+            MemberType members[(sizeof(ParentType) / sizeof(MemberType)) + 1];
+        } data;
+#pragma pack(pop)
+        UnionImpl<ParentType, MemberType, Offset + 1> next_union;
+    };
+
+    template <typename ParentType, typename MemberType>
+    union UnionImpl<ParentType, MemberType, 0> {
+        static constexpr size_t GetOffset() {
+            return 0;
+        }
+
+        struct {
+            MemberType members[(sizeof(ParentType) / sizeof(MemberType)) + 1];
+        } data;
+        UnionImpl<ParentType, MemberType, 1> next_union;
+    };
+
+    template <typename ParentType, typename MemberType>
+    union UnionImpl<ParentType, MemberType, MaxDepth> {};
+};
+
+template <typename ParentType, typename MemberType>
+struct OffsetOfCalculator {
+    using UnionHolder =
+        typename OffsetOfUnionHolder<sizeof(MemberType)>::template UnionImpl<ParentType, MemberType,
+                                                                             0>;
+    union Union {
+        char c{};
+        UnionHolder first_union;
+        TypedStorage<ParentType> parent;
+
+        constexpr Union() : c() {}
+    };
+    static constexpr Union U = {};
+
+    static constexpr const MemberType* GetNextAddress(const MemberType* start,
+                                                      const MemberType* target) {
+        while (start < target) {
+            start++;
+        }
+        return start;
+    }
+
+    static constexpr std::ptrdiff_t GetDifference(const MemberType* start,
+                                                  const MemberType* target) {
+        return (target - start) * sizeof(MemberType);
+    }
+
+    template <typename CurUnion>
+    static constexpr std::ptrdiff_t OffsetOfImpl(MemberType ParentType::*member,
+                                                 CurUnion& cur_union) {
+        constexpr size_t Offset = CurUnion::GetOffset();
+        const auto target = std::addressof(GetPointer(U.parent)->*member);
+        const auto start = std::addressof(cur_union.data.members[0]);
+        const auto next = GetNextAddress(start, target);
+
+        if (next != target) {
+            if constexpr (Offset < sizeof(MemberType) - 1) {
+                return OffsetOfImpl(member, cur_union.next_union);
+            } else {
+                UNREACHABLE();
+            }
+        }
+
+        return (next - start) * sizeof(MemberType) + Offset;
+    }
+
+    static constexpr std::ptrdiff_t OffsetOf(MemberType ParentType::*member) {
+        return OffsetOfImpl(member, U.first_union);
+    }
+};
+
+template <typename T>
+struct GetMemberPointerTraits;
+
+template <typename P, typename M>
+struct GetMemberPointerTraits<M P::*> {
+    using Parent = P;
+    using Member = M;
+};
+
+template <auto MemberPtr>
+using GetParentType = typename GetMemberPointerTraits<decltype(MemberPtr)>::Parent;
+
+template <auto MemberPtr>
+using GetMemberType = typename GetMemberPointerTraits<decltype(MemberPtr)>::Member;
+
+template <auto MemberPtr, typename RealParentType = GetParentType<MemberPtr>>
+static inline std::ptrdiff_t OffsetOf = [] {
+    using DeducedParentType = GetParentType<MemberPtr>;
+    using MemberType = GetMemberType<MemberPtr>;
+    static_assert(std::is_base_of<DeducedParentType, RealParentType>::value ||
+                  std::is_same<RealParentType, DeducedParentType>::value);
+
+    return OffsetOfCalculator<RealParentType, MemberType>::OffsetOf(MemberPtr);
+}();
+
+} // namespace impl
+
+template <auto MemberPtr, typename RealParentType = impl::GetParentType<MemberPtr>>
+constexpr RealParentType& GetParentReference(impl::GetMemberType<MemberPtr>* member) {
+    std::ptrdiff_t Offset = impl::OffsetOf<MemberPtr, RealParentType>;
+    return *static_cast<RealParentType*>(
+        static_cast<void*>(static_cast<uint8_t*>(static_cast<void*>(member)) - Offset));
+}
+
+template <auto MemberPtr, typename RealParentType = impl::GetParentType<MemberPtr>>
+constexpr RealParentType const& GetParentReference(impl::GetMemberType<MemberPtr> const* member) {
+    std::ptrdiff_t Offset = impl::OffsetOf<MemberPtr, RealParentType>;
+    return *static_cast<const RealParentType*>(static_cast<const void*>(
+        static_cast<const uint8_t*>(static_cast<const void*>(member)) - Offset));
+}
+
+template <auto MemberPtr, typename RealParentType = impl::GetParentType<MemberPtr>>
+constexpr RealParentType* GetParentPointer(impl::GetMemberType<MemberPtr>* member) {
+    return std::addressof(GetParentReference<MemberPtr, RealParentType>(member));
+}
+
+template <auto MemberPtr, typename RealParentType = impl::GetParentType<MemberPtr>>
+constexpr RealParentType const* GetParentPointer(impl::GetMemberType<MemberPtr> const* member) {
+    return std::addressof(GetParentReference<MemberPtr, RealParentType>(member));
+}
+
+template <auto MemberPtr, typename RealParentType = impl::GetParentType<MemberPtr>>
+constexpr RealParentType& GetParentReference(impl::GetMemberType<MemberPtr>& member) {
+    return GetParentReference<MemberPtr, RealParentType>(std::addressof(member));
+}
+
+template <auto MemberPtr, typename RealParentType = impl::GetParentType<MemberPtr>>
+constexpr RealParentType const& GetParentReference(impl::GetMemberType<MemberPtr> const& member) {
+    return GetParentReference<MemberPtr, RealParentType>(std::addressof(member));
+}
+
+template <auto MemberPtr, typename RealParentType = impl::GetParentType<MemberPtr>>
+constexpr RealParentType* GetParentPointer(impl::GetMemberType<MemberPtr>& member) {
+    return std::addressof(GetParentReference<MemberPtr, RealParentType>(member));
+}
+
+template <auto MemberPtr, typename RealParentType = impl::GetParentType<MemberPtr>>
+constexpr RealParentType const* GetParentPointer(impl::GetMemberType<MemberPtr> const& member) {
+    return std::addressof(GetParentReference<MemberPtr, RealParentType>(member));
+}
+
+} // namespace Common

src/common/ring_buffer.h

@@ -19,15 +19,14 @@ namespace Common {
 /// SPSC ring buffer
 /// @tparam T            Element type
 /// @tparam capacity     Number of slots in ring buffer
-/// @tparam granularity  Slot size in terms of number of elements
-template <typename T, std::size_t capacity, std::size_t granularity = 1>
+template <typename T, std::size_t capacity>
 class RingBuffer {
-    /// A "slot" is made of `granularity` elements of `T`.
-    static constexpr std::size_t slot_size = granularity * sizeof(T);
+    /// A "slot" is made of a single `T`.
+    static constexpr std::size_t slot_size = sizeof(T);
     // T must be safely memcpy-able and have a trivial default constructor.
     static_assert(std::is_trivial_v<T>);
     // Ensure capacity is sensible.
-    static_assert(capacity < std::numeric_limits<std::size_t>::max() / 2 / granularity);
+    static_assert(capacity < std::numeric_limits<std::size_t>::max() / 2);
     static_assert((capacity & (capacity - 1)) == 0, "capacity must be a power of two");
     // Ensure lock-free.
     static_assert(std::atomic_size_t::is_always_lock_free);
@@ -47,7 +46,7 @@ public:
         const std::size_t second_copy = push_count - first_copy;
 
         const char* in = static_cast<const char*>(new_slots);
-        std::memcpy(m_data.data() + pos * granularity, in, first_copy * slot_size);
+        std::memcpy(m_data.data() + pos, in, first_copy * slot_size);
         in += first_copy * slot_size;
         std::memcpy(m_data.data(), in, second_copy * slot_size);
 
@@ -74,7 +73,7 @@ public:
         const std::size_t second_copy = pop_count - first_copy;
 
         char* out = static_cast<char*>(output);
-        std::memcpy(out, m_data.data() + pos * granularity, first_copy * slot_size);
+        std::memcpy(out, m_data.data() + pos, first_copy * slot_size);
         out += first_copy * slot_size;
         std::memcpy(out, m_data.data(), second_copy * slot_size);
 
@@ -84,9 +83,9 @@ public:
     }
 
     std::vector<T> Pop(std::size_t max_slots = ~std::size_t(0)) {
-        std::vector<T> out(std::min(max_slots, capacity) * granularity);
-        const std::size_t count = Pop(out.data(), out.size() / granularity);
-        out.resize(count * granularity);
+        std::vector<T> out(std::min(max_slots, capacity));
+        const std::size_t count = Pop(out.data(), out.size());
+        out.resize(count);
         return out;
     }
 
@@ -113,7 +112,7 @@ private:
     alignas(128) std::atomic_size_t m_write_index{0};
 #endif
 
-    std::array<T, granularity * capacity> m_data;
+    std::array<T, capacity> m_data;
 };
 
 } // namespace Common

src/common/scope_exit.h

@@ -10,7 +10,7 @@
 namespace detail {
 template <typename Func>
 struct ScopeExitHelper {
-    explicit ScopeExitHelper(Func&& func) : func(std::move(func)) {}
+    explicit ScopeExitHelper(Func&& func_) : func(std::move(func_)) {}
     ~ScopeExitHelper() {
         if (active) {
             func();
@@ -49,3 +49,9 @@ ScopeExitHelper<Func> ScopeExit(Func&& func) {
  * \endcode
  */
 #define SCOPE_EXIT(body) auto CONCAT2(scope_exit_helper_, __LINE__) = detail::ScopeExit([&]() body)
+
+/**
+ * This macro is similar to SCOPE_EXIT, except the object is caller managed. This is intended to be
+ * used when the caller might want to cancel the ScopeExit.
+ */
+#define SCOPE_GUARD(body) detail::ScopeExit([&]() body)