core/CMakeLists: Add web_types.h

debugger: controller: Add access key

.ci/scripts/linux/docker.sh

@@ -22,12 +22,10 @@ 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/AppImage/AppImageKit/releases/download/continuous/appimagetool-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 \
-    appimagetool-x86_64.AppImage \
     AppRun-patched-x86_64 \
     exec-x86_64.so \
     linuxdeploy-x86_64.AppImage \
@@ -49,6 +47,3 @@ 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
-
-# Build the AppImage
-./appimagetool-x86_64.AppImage AppDir

.ci/scripts/linux/upload.sh

@@ -2,8 +2,7 @@
 
 . .ci/scripts/common/pre-upload.sh
 
-APPIMAGE_NAME="yuzu-x86_64.AppImage"
-NEW_APPIMAGE_NAME="yuzu-${GITDATE}-${GITREV}-x86_64.AppImage"
+APPIMAGE_NAME="yuzu-${GITDATE}-${GITREV}.AppImage"
 REV_NAME="yuzu-linux-${GITDATE}-${GITREV}"
 ARCHIVE_NAME="${REV_NAME}.tar.xz"
 COMPRESSION_FLAGS="-cJvf"
@@ -19,7 +18,24 @@ mkdir "$DIR_NAME"
 cp build/bin/yuzu-cmd "$DIR_NAME"
 cp build/bin/yuzu "$DIR_NAME"
 
-# Copy the AppImage to the artifacts directory and avoid compressing it
-cp "build/${APPIMAGE_NAME}" "${ARTIFACTS_DIR}/${NEW_APPIMAGE_NAME}"
+# Build an AppImage
+cd build
+
+wget -nc https://github.com/AppImage/AppImageKit/releases/download/continuous/appimagetool-x86_64.AppImage
+chmod 755 appimagetool-x86_64.AppImage
+
+if [ "${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

@@ -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/ ';'

.gitmodules

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

CMakeLists.txt

@@ -18,6 +18,8 @@ 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)
@@ -261,7 +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()
 
@@ -384,19 +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 ${FFmpeg_VERSION}")
+
+            add_dependencies(ffmpeg-build ffmpeg-configure)
+        else()
+            message(FATAL_ERROR "FFmpeg not found")
+        endif()
+    else() # WIN32
+        # Use yuzu FFmpeg binaries
+        set(FFmpeg_EXT_NAME "ffmpeg-4.2.1")
+        set(FFmpeg_PATH "${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 "${FFmpeg_PATH}/bin" CACHE PATH "Path to FFmpeg library directory" FORCE)
+        set(FFmpeg_DLL_DIR "${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

@@ -1,7 +1,7 @@
 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} ${DLL_DEST}
+    windows_copy_files(${target_dir} ${FFmpeg_DLL_DIR} ${DLL_DEST}
         avcodec-58.dll
         avutil-56.dll
         swresample-3.dll

README.md

@@ -33,7 +33,7 @@ If you want to contribute to the user interface translation, please check out th
 
 
 ### 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/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

@@ -64,11 +64,16 @@ 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
@@ -127,7 +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

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) {
@@ -65,6 +81,154 @@ 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_,
@@ -271,11 +435,10 @@ 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);
     }
 }
 
@@ -376,17 +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));
+            }
         }
     }
 }
@@ -525,6 +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);

src/audio_core/command_generator.h

@@ -21,6 +21,8 @@ class ServerMixInfo;
 class EffectContext;
 class EffectBase;
 struct AuxInfoDSP;
+struct I3dl2ReverbParams;
+struct I3dl2ReverbState;
 using MixVolumeBuffer = std::array<float, AudioCommon::MAX_MIX_BUFFERS>;
 
 class CommandGenerator {
@@ -80,6 +82,9 @@ private:
     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

@@ -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/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

@@ -90,6 +90,14 @@ s32 EffectBase::GetProcessingOrder() const {
     return processing_order;
 }
 
+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;
 
@@ -117,6 +125,12 @@ void EffectI3dl2Reverb::Update(EffectInfo::InParams& in_params) {
         usage = UsageState::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);
+        }
     }
 }
 
@@ -129,6 +143,14 @@ void EffectI3dl2Reverb::UpdateForCommandGeneration() {
     GetParams().status = ParameterStatus::Updated;
 }
 
+I3dl2ReverbState& EffectI3dl2Reverb::GetState() {
+    return state;
+}
+
+const I3dl2ReverbState& EffectI3dl2Reverb::GetState() const {
+    return state;
+}
+
 EffectBiquadFilter::EffectBiquadFilter() : EffectGeneric(EffectType::BiquadFilter) {}
 EffectBiquadFilter::~EffectBiquadFilter() = default;
 

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"
@@ -194,6 +195,8 @@ public:
     [[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};
@@ -201,6 +204,7 @@ protected:
     s32 mix_id{};
     s32 processing_order{};
     bool enabled = false;
+    std::vector<u8> work_buffer{};
 };
 
 template <typename T>
@@ -212,7 +216,7 @@ public:
         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> {

src/audio_core/sink_context.h

@@ -40,17 +40,17 @@ 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(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;
         DownmixCoefficients down_matrix_coef;
     };

src/audio_core/stream.cpp

@@ -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) {

src/audio_core/stream.h

@@ -56,6 +56,9 @@ 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
     [[nodiscard]] bool ContainsBuffer(Buffer::Tag tag) const;
 

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 {

src/common/CMakeLists.txt

@@ -98,7 +98,6 @@ add_library(common STATIC
     algorithm.h
     alignment.h
     assert.h
-    atomic_ops.cpp
     atomic_ops.h
     detached_tasks.cpp
     detached_tasks.h
@@ -108,7 +107,6 @@ add_library(common STATIC
     bit_util.h
     cityhash.cpp
     cityhash.h
-    color.h
     common_funcs.h
     common_paths.h
     common_types.h
@@ -140,6 +138,8 @@ add_library(common STATIC
     microprofile.h
     microprofileui.h
     misc.cpp
+    nvidia_flags.cpp
+    nvidia_flags.h
     page_table.cpp
     page_table.h
     param_package.cpp
@@ -167,8 +167,6 @@ add_library(common STATIC
     threadsafe_queue.h
     time_zone.cpp
     time_zone.h
-    timer.cpp
-    timer.h
     tree.h
     uint128.cpp
     uint128.h
@@ -208,6 +206,8 @@ if (MSVC)
 else()
   target_compile_options(common PRIVATE
     -Werror
+
+    $<$<CXX_COMPILER_ID:Clang>:-fsized-deallocation>
   )
 endif()
 

src/common/alignment.h

@@ -9,50 +9,45 @@
 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, 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_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/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
 
@@ -97,10 +97,27 @@ __declspec(dllimport) void __stdcall DebugBreak(void);
 #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/logging/backend.cpp

@@ -145,10 +145,18 @@ 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

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/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

@@ -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)

src/common/string_util.cpp

@@ -141,27 +141,13 @@ std::string ReplaceAll(std::string result, const std::string& src, const std::st
 }
 
 std::string UTF16ToUTF8(const std::u16string& input) {
-#ifdef _MSC_VER
-    // Workaround for missing char16_t/char32_t instantiations in MSVC2017
-    std::wstring_convert<std::codecvt_utf8_utf16<__int16>, __int16> convert;
-    std::basic_string<__int16> tmp_buffer(input.cbegin(), input.cend());
-    return convert.to_bytes(tmp_buffer);
-#else
     std::wstring_convert<std::codecvt_utf8_utf16<char16_t>, char16_t> convert;
     return convert.to_bytes(input);
-#endif
 }
 
 std::u16string UTF8ToUTF16(const std::string& input) {
-#ifdef _MSC_VER
-    // Workaround for missing char16_t/char32_t instantiations in MSVC2017
-    std::wstring_convert<std::codecvt_utf8_utf16<__int16>, __int16> convert;
-    auto tmp_buffer = convert.from_bytes(input);
-    return std::u16string(tmp_buffer.cbegin(), tmp_buffer.cend());
-#else
     std::wstring_convert<std::codecvt_utf8_utf16<char16_t>, char16_t> convert;
     return convert.from_bytes(input);
-#endif
 }
 
 #ifdef _WIN32

src/common/timer.cpp

@@ -1,159 +0,0 @@
-// Copyright 2013 Dolphin Emulator Project / 2014 Citra Emulator Project
-// Licensed under GPLv2 or any later version
-// Refer to the license.txt file included.
-
-#include <ctime>
-#include <fmt/format.h>
-#include "common/common_types.h"
-#include "common/string_util.h"
-#include "common/timer.h"
-
-namespace Common {
-
-std::chrono::milliseconds Timer::GetTimeMs() {
-    return std::chrono::duration_cast<std::chrono::milliseconds>(
-        std::chrono::system_clock::now().time_since_epoch());
-}
-
-// --------------------------------------------
-// Initiate, Start, Stop, and Update the time
-// --------------------------------------------
-
-// Set initial values for the class
-Timer::Timer() : m_LastTime(0), m_StartTime(0), m_Running(false) {
-    Update();
-}
-
-// Write the starting time
-void Timer::Start() {
-    m_StartTime = GetTimeMs();
-    m_Running = true;
-}
-
-// Stop the timer
-void Timer::Stop() {
-    // Write the final time
-    m_LastTime = GetTimeMs();
-    m_Running = false;
-}
-
-// Update the last time variable
-void Timer::Update() {
-    m_LastTime = GetTimeMs();
-    // TODO(ector) - QPF
-}
-
-// -------------------------------------
-// Get time difference and elapsed time
-// -------------------------------------
-
-// Get the number of milliseconds since the last Update()
-std::chrono::milliseconds Timer::GetTimeDifference() {
-    return GetTimeMs() - m_LastTime;
-}
-
-// Add the time difference since the last Update() to the starting time.
-// This is used to compensate for a paused game.
-void Timer::AddTimeDifference() {
-    m_StartTime += GetTimeDifference();
-}
-
-// Get the time elapsed since the Start()
-std::chrono::milliseconds Timer::GetTimeElapsed() {
-    // If we have not started yet, return 1 (because then I don't
-    // have to change the FPS calculation in CoreRerecording.cpp .
-    if (m_StartTime.count() == 0)
-        return std::chrono::milliseconds(1);
-
-    // Return the final timer time if the timer is stopped
-    if (!m_Running)
-        return (m_LastTime - m_StartTime);
-
-    return (GetTimeMs() - m_StartTime);
-}
-
-// Get the formatted time elapsed since the Start()
-std::string Timer::GetTimeElapsedFormatted() const {
-    // If we have not started yet, return zero
-    if (m_StartTime.count() == 0)
-        return "00:00:00:000";
-
-    // The number of milliseconds since the start.
-    // Use a different value if the timer is stopped.
-    std::chrono::milliseconds Milliseconds;
-    if (m_Running)
-        Milliseconds = GetTimeMs() - m_StartTime;
-    else
-        Milliseconds = m_LastTime - m_StartTime;
-    // Seconds
-    std::chrono::seconds Seconds = std::chrono::duration_cast<std::chrono::seconds>(Milliseconds);
-    // Minutes
-    std::chrono::minutes Minutes = std::chrono::duration_cast<std::chrono::minutes>(Milliseconds);
-    // Hours
-    std::chrono::hours Hours = std::chrono::duration_cast<std::chrono::hours>(Milliseconds);
-
-    std::string TmpStr = fmt::format("{:02}:{:02}:{:02}:{:03}", Hours.count(), Minutes.count() % 60,
-                                     Seconds.count() % 60, Milliseconds.count() % 1000);
-    return TmpStr;
-}
-
-// Get the number of seconds since January 1 1970
-std::chrono::seconds Timer::GetTimeSinceJan1970() {
-    return std::chrono::duration_cast<std::chrono::seconds>(GetTimeMs());
-}
-
-std::chrono::seconds Timer::GetLocalTimeSinceJan1970() {
-    time_t sysTime, tzDiff, tzDST;
-    struct tm* gmTime;
-
-    time(&sysTime);
-
-    // Account for DST where needed
-    gmTime = localtime(&sysTime);
-    if (gmTime->tm_isdst == 1)
-        tzDST = 3600;
-    else
-        tzDST = 0;
-
-    // Lazy way to get local time in sec
-    gmTime = gmtime(&sysTime);
-    tzDiff = sysTime - mktime(gmTime);
-
-    return std::chrono::seconds(sysTime + tzDiff + tzDST);
-}
-
-// Return the current time formatted as Minutes:Seconds:Milliseconds
-// in the form 00:00:000.
-std::string Timer::GetTimeFormatted() {
-    time_t sysTime;
-    struct tm* gmTime;
-    char tmp[13];
-
-    time(&sysTime);
-    gmTime = localtime(&sysTime);
-
-    strftime(tmp, 6, "%M:%S", gmTime);
-
-    u64 milliseconds = static_cast<u64>(GetTimeMs().count()) % 1000;
-    return fmt::format("{}:{:03}", tmp, milliseconds);
-}
-
-// Returns a timestamp with decimals for precise time comparisons
-// ----------------
-double Timer::GetDoubleTime() {
-    // Get continuous timestamp
-    auto tmp_seconds = static_cast<u64>(GetTimeSinceJan1970().count());
-    const auto ms = static_cast<double>(static_cast<u64>(GetTimeMs().count()) % 1000);
-
-    // Remove a few years. We only really want enough seconds to make
-    // sure that we are detecting actual actions, perhaps 60 seconds is
-    // enough really, but I leave a year of seconds anyway, in case the
-    // user's clock is incorrect or something like that.
-    tmp_seconds = tmp_seconds - (38 * 365 * 24 * 60 * 60);
-
-    // Make a smaller integer that fits in the double
-    const auto seconds = static_cast<u32>(tmp_seconds);
-    return seconds + ms;
-}
-
-} // Namespace Common

src/common/timer.h

@@ -1,41 +0,0 @@
-// Copyright 2013 Dolphin Emulator Project / 2014 Citra Emulator Project
-// Licensed under GPLv2 or any later version
-// Refer to the license.txt file included.
-
-#pragma once
-
-#include <chrono>
-#include <string>
-#include "common/common_types.h"
-
-namespace Common {
-class Timer {
-public:
-    Timer();
-
-    void Start();
-    void Stop();
-    void Update();
-
-    // The time difference is always returned in milliseconds, regardless of alternative internal
-    // representation
-    [[nodiscard]] std::chrono::milliseconds GetTimeDifference();
-    void AddTimeDifference();
-
-    [[nodiscard]] static std::chrono::seconds GetTimeSinceJan1970();
-    [[nodiscard]] static std::chrono::seconds GetLocalTimeSinceJan1970();
-    [[nodiscard]] static double GetDoubleTime();
-
-    [[nodiscard]] static std::string GetTimeFormatted();
-    [[nodiscard]] std::string GetTimeElapsedFormatted() const;
-    [[nodiscard]] std::chrono::milliseconds GetTimeElapsed();
-
-    [[nodiscard]] static std::chrono::milliseconds GetTimeMs();
-
-private:
-    std::chrono::milliseconds m_LastTime;
-    std::chrono::milliseconds m_StartTime;
-    bool m_Running;
-};
-
-} // Namespace Common

src/common/uuid.h

@@ -14,8 +14,8 @@ constexpr u128 INVALID_UUID{{0, 0}};
 
 struct UUID {
     // UUIDs which are 0 are considered invalid!
-    u128 uuid = INVALID_UUID;
-    constexpr UUID() = default;
+    u128 uuid;
+    UUID() = default;
     constexpr explicit UUID(const u128& id) : uuid{id} {}
     constexpr explicit UUID(const u64 lo, const u64 hi) : uuid{{lo, hi}} {}
 

src/common/x64/native_clock.cpp

@@ -2,19 +2,74 @@
 // Licensed under GPLv2 or any later version
 // Refer to the license.txt file included.
 
+#include <array>
 #include <chrono>
+#include <limits>
 #include <mutex>
 #include <thread>
 
 #ifdef _MSC_VER
 #include <intrin.h>
+
+#pragma intrinsic(__umulh)
+#pragma intrinsic(_udiv128)
 #else
 #include <x86intrin.h>
 #endif
 
+#include "common/atomic_ops.h"
 #include "common/uint128.h"
 #include "common/x64/native_clock.h"
 
+namespace {
+
+[[nodiscard]] u64 GetFixedPoint64Factor(u64 numerator, u64 divisor) {
+#ifdef __SIZEOF_INT128__
+    const auto base = static_cast<unsigned __int128>(numerator) << 64ULL;
+    return static_cast<u64>(base / divisor);
+#elif defined(_M_X64) || defined(_M_ARM64)
+    std::array<u64, 2> r = {0, numerator};
+    u64 remainder;
+#if _MSC_VER < 1923
+    return udiv128(r[1], r[0], divisor, &remainder);
+#else
+    return _udiv128(r[1], r[0], divisor, &remainder);
+#endif
+#else
+    // This one is bit more inaccurate.
+    return MultiplyAndDivide64(std::numeric_limits<u64>::max(), numerator, divisor);
+#endif
+}
+
+[[nodiscard]] u64 MultiplyHigh(u64 a, u64 b) {
+#ifdef __SIZEOF_INT128__
+    return (static_cast<unsigned __int128>(a) * static_cast<unsigned __int128>(b)) >> 64;
+#elif defined(_M_X64) || defined(_M_ARM64)
+    return __umulh(a, b); // MSVC
+#else
+    // Generic fallback
+    const u64 a_lo = u32(a);
+    const u64 a_hi = a >> 32;
+    const u64 b_lo = u32(b);
+    const u64 b_hi = b >> 32;
+
+    const u64 a_x_b_hi = a_hi * b_hi;
+    const u64 a_x_b_mid = a_hi * b_lo;
+    const u64 b_x_a_mid = b_hi * a_lo;
+    const u64 a_x_b_lo = a_lo * b_lo;
+
+    const u64 carry_bit = (static_cast<u64>(static_cast<u32>(a_x_b_mid)) +
+                           static_cast<u64>(static_cast<u32>(b_x_a_mid)) + (a_x_b_lo >> 32)) >>
+                          32;
+
+    const u64 multhi = a_x_b_hi + (a_x_b_mid >> 32) + (b_x_a_mid >> 32) + carry_bit;
+
+    return multhi;
+#endif
+}
+
+} // namespace
+
 namespace Common {
 
 u64 EstimateRDTSCFrequency() {
@@ -48,54 +103,71 @@ NativeClock::NativeClock(u64 emulated_cpu_frequency_, u64 emulated_clock_frequen
     : WallClock(emulated_cpu_frequency_, emulated_clock_frequency_, true), rtsc_frequency{
                                                                                rtsc_frequency_} {
     _mm_mfence();
-    last_measure = __rdtsc();
-    accumulated_ticks = 0U;
+    time_point.inner.last_measure = __rdtsc();
+    time_point.inner.accumulated_ticks = 0U;
+    ns_rtsc_factor = GetFixedPoint64Factor(1000000000, rtsc_frequency);
+    us_rtsc_factor = GetFixedPoint64Factor(1000000, rtsc_frequency);
+    ms_rtsc_factor = GetFixedPoint64Factor(1000, rtsc_frequency);
+    clock_rtsc_factor = GetFixedPoint64Factor(emulated_clock_frequency, rtsc_frequency);
+    cpu_rtsc_factor = GetFixedPoint64Factor(emulated_cpu_frequency, rtsc_frequency);
 }
 
 u64 NativeClock::GetRTSC() {
-    std::scoped_lock scope{rtsc_serialize};
-    _mm_mfence();
-    const u64 current_measure = __rdtsc();
-    u64 diff = current_measure - last_measure;
-    diff = diff & ~static_cast<u64>(static_cast<s64>(diff) >> 63); // max(diff, 0)
-    if (current_measure > last_measure) {
-        last_measure = current_measure;
-    }
-    accumulated_ticks += diff;
+    TimePoint new_time_point{};
+    TimePoint current_time_point{};
+    do {
+        current_time_point.pack = time_point.pack;
+        _mm_mfence();
+        const u64 current_measure = __rdtsc();
+        u64 diff = current_measure - current_time_point.inner.last_measure;
+        diff = diff & ~static_cast<u64>(static_cast<s64>(diff) >> 63); // max(diff, 0)
+        new_time_point.inner.last_measure = current_measure > current_time_point.inner.last_measure
+                                                ? current_measure
+                                                : current_time_point.inner.last_measure;
+        new_time_point.inner.accumulated_ticks = current_time_point.inner.accumulated_ticks + diff;
+    } while (!Common::AtomicCompareAndSwap(time_point.pack.data(), new_time_point.pack,
+                                           current_time_point.pack));
     /// The clock cannot be more precise than the guest timer, remove the lower bits
-    return accumulated_ticks & inaccuracy_mask;
+    return new_time_point.inner.accumulated_ticks & inaccuracy_mask;
 }
 
 void NativeClock::Pause(bool is_paused) {
     if (!is_paused) {
-        _mm_mfence();
-        last_measure = __rdtsc();
+        TimePoint current_time_point{};
+        TimePoint new_time_point{};
+        do {
+            current_time_point.pack = time_point.pack;
+            new_time_point.pack = current_time_point.pack;
+            _mm_mfence();
+            new_time_point.inner.last_measure = __rdtsc();
+        } while (!Common::AtomicCompareAndSwap(time_point.pack.data(), new_time_point.pack,
+                                               current_time_point.pack));
     }
 }
 
 std::chrono::nanoseconds NativeClock::GetTimeNS() {
     const u64 rtsc_value = GetRTSC();
-    return std::chrono::nanoseconds{MultiplyAndDivide64(rtsc_value, 1000000000, rtsc_frequency)};
+    return std::chrono::nanoseconds{MultiplyHigh(rtsc_value, ns_rtsc_factor)};
 }
 
 std::chrono::microseconds NativeClock::GetTimeUS() {
     const u64 rtsc_value = GetRTSC();
-    return std::chrono::microseconds{MultiplyAndDivide64(rtsc_value, 1000000, rtsc_frequency)};
+    return std::chrono::microseconds{MultiplyHigh(rtsc_value, us_rtsc_factor)};
 }
 
 std::chrono::milliseconds NativeClock::GetTimeMS() {
     const u64 rtsc_value = GetRTSC();
-    return std::chrono::milliseconds{MultiplyAndDivide64(rtsc_value, 1000, rtsc_frequency)};
+    return std::chrono::milliseconds{MultiplyHigh(rtsc_value, ms_rtsc_factor)};
 }
 
 u64 NativeClock::GetClockCycles() {
     const u64 rtsc_value = GetRTSC();
-    return MultiplyAndDivide64(rtsc_value, emulated_clock_frequency, rtsc_frequency);
+    return MultiplyHigh(rtsc_value, clock_rtsc_factor);
 }
 
 u64 NativeClock::GetCPUCycles() {
     const u64 rtsc_value = GetRTSC();
-    return MultiplyAndDivide64(rtsc_value, emulated_cpu_frequency, rtsc_frequency);
+    return MultiplyHigh(rtsc_value, cpu_rtsc_factor);
 }
 
 } // namespace X64

src/common/x64/native_clock.h

@@ -6,7 +6,6 @@
 
 #include <optional>
 
-#include "common/spin_lock.h"
 #include "common/wall_clock.h"
 
 namespace Common {
@@ -32,14 +31,28 @@ public:
 private:
     u64 GetRTSC();
 
+    union alignas(16) TimePoint {
+        TimePoint() : pack{} {}
+        u128 pack{};
+        struct Inner {
+            u64 last_measure{};
+            u64 accumulated_ticks{};
+        } inner;
+    };
+
     /// value used to reduce the native clocks accuracy as some apss rely on
     /// undefined behavior where the level of accuracy in the clock shouldn't
     /// be higher.
     static constexpr u64 inaccuracy_mask = ~(UINT64_C(0x400) - 1);
 
-    SpinLock rtsc_serialize{};
-    u64 last_measure{};
-    u64 accumulated_ticks{};
+    TimePoint time_point;
+    // factors
+    u64 clock_rtsc_factor{};
+    u64 cpu_rtsc_factor{};
+    u64 ns_rtsc_factor{};
+    u64 us_rtsc_factor{};
+    u64 ms_rtsc_factor{};
+
     u64 rtsc_frequency;
 };
 } // namespace X64

src/core/CMakeLists.txt

@@ -148,7 +148,7 @@ add_library(core STATIC
     hle/kernel/client_session.h
     hle/kernel/code_set.cpp
     hle/kernel/code_set.h
-    hle/kernel/errors.h
+    hle/kernel/svc_results.h
     hle/kernel/global_scheduler_context.cpp
     hle/kernel/global_scheduler_context.h
     hle/kernel/handle_table.cpp
@@ -160,14 +160,29 @@ add_library(core STATIC
     hle/kernel/k_affinity_mask.h
     hle/kernel/k_condition_variable.cpp
     hle/kernel/k_condition_variable.h
+    hle/kernel/k_event.cpp
+    hle/kernel/k_event.h
+    hle/kernel/k_light_condition_variable.h
+    hle/kernel/k_light_lock.cpp
+    hle/kernel/k_light_lock.h
     hle/kernel/k_priority_queue.h
+    hle/kernel/k_readable_event.cpp
+    hle/kernel/k_readable_event.h
+    hle/kernel/k_resource_limit.cpp
+    hle/kernel/k_resource_limit.h
     hle/kernel/k_scheduler.cpp
     hle/kernel/k_scheduler.h
     hle/kernel/k_scheduler_lock.h
     hle/kernel/k_scoped_lock.h
+    hle/kernel/k_scoped_resource_reservation.h
     hle/kernel/k_scoped_scheduler_lock_and_sleep.h
     hle/kernel/k_synchronization_object.cpp
     hle/kernel/k_synchronization_object.h
+    hle/kernel/k_thread.cpp
+    hle/kernel/k_thread.h
+    hle/kernel/k_thread_queue.h
+    hle/kernel/k_writable_event.cpp
+    hle/kernel/k_writable_event.h
     hle/kernel/kernel.cpp
     hle/kernel/kernel.h
     hle/kernel/memory/address_space_info.cpp
@@ -196,10 +211,6 @@ add_library(core STATIC
     hle/kernel/process.h
     hle/kernel/process_capability.cpp
     hle/kernel/process_capability.h
-    hle/kernel/readable_event.cpp
-    hle/kernel/readable_event.h
-    hle/kernel/resource_limit.cpp
-    hle/kernel/resource_limit.h
     hle/kernel/server_port.cpp
     hle/kernel/server_port.h
     hle/kernel/server_session.cpp
@@ -213,17 +224,12 @@ add_library(core STATIC
     hle/kernel/svc.cpp
     hle/kernel/svc.h
     hle/kernel/svc_common.h
-    hle/kernel/svc_results.h
     hle/kernel/svc_types.h
     hle/kernel/svc_wrap.h
-    hle/kernel/thread.cpp
-    hle/kernel/thread.h
     hle/kernel/time_manager.cpp
     hle/kernel/time_manager.h
     hle/kernel/transfer_memory.cpp
     hle/kernel/transfer_memory.h
-    hle/kernel/writable_event.cpp
-    hle/kernel/writable_event.h
     hle/lock.cpp
     hle/lock.h
     hle/result.h
@@ -260,6 +266,7 @@ add_library(core STATIC
     hle/service/am/applets/software_keyboard.h
     hle/service/am/applets/web_browser.cpp
     hle/service/am/applets/web_browser.h
+    hle/service/am/applets/web_types.h
     hle/service/am/idle.cpp
     hle/service/am/idle.h
     hle/service/am/omm.cpp
@@ -394,14 +401,13 @@ add_library(core STATIC
     hle/service/hid/controllers/xpad.h
     hle/service/lbl/lbl.cpp
     hle/service/lbl/lbl.h
+    hle/service/ldn/errors.h
     hle/service/ldn/ldn.cpp
     hle/service/ldn/ldn.h
     hle/service/ldr/ldr.cpp
     hle/service/ldr/ldr.h
     hle/service/lm/lm.cpp
     hle/service/lm/lm.h
-    hle/service/lm/manager.cpp
-    hle/service/lm/manager.h
     hle/service/mig/mig.cpp
     hle/service/mig/mig.h
     hle/service/mii/manager.cpp
@@ -643,13 +649,14 @@ else()
         -Werror=conversion
         -Werror=ignored-qualifiers
         -Werror=implicit-fallthrough
-        -Werror=reorder
         -Werror=sign-compare
-        -Werror=unused-variable
 
+        $<$<CXX_COMPILER_ID:GNU>:-Werror=class-memaccess>
         $<$<CXX_COMPILER_ID:GNU>:-Werror=unused-but-set-parameter>
         $<$<CXX_COMPILER_ID:GNU>:-Werror=unused-but-set-variable>
 
+        $<$<CXX_COMPILER_ID:Clang>:-fsized-deallocation>
+
         -Wno-sign-conversion
     )
 endif()

src/core/arm/dynarmic/arm_dynarmic_32.cpp

@@ -71,8 +71,9 @@ public:
     }
 
     void ExceptionRaised(u32 pc, Dynarmic::A32::Exception exception) override {
-        LOG_CRITICAL(Core_ARM, "ExceptionRaised(exception = {}, pc = {:08X}, code = {:08X})",
-                     exception, pc, MemoryReadCode(pc));
+        LOG_CRITICAL(Core_ARM,
+                     "ExceptionRaised(exception = {}, pc = {:08X}, code = {:08X}, thumb = {})",
+                     exception, pc, MemoryReadCode(pc), parent.IsInThumbMode());
         UNIMPLEMENTED();
     }
 
@@ -255,6 +256,9 @@ void ARM_Dynarmic_32::ChangeProcessorID(std::size_t new_core_id) {
 }
 
 void ARM_Dynarmic_32::SaveContext(ThreadContext32& ctx) {
+    if (!jit) {
+        return;
+    }
     Dynarmic::A32::Context context;
     jit->SaveContext(context);
     ctx.cpu_registers = context.Regs();
@@ -264,6 +268,9 @@ void ARM_Dynarmic_32::SaveContext(ThreadContext32& ctx) {
 }
 
 void ARM_Dynarmic_32::LoadContext(const ThreadContext32& ctx) {
+    if (!jit) {
+        return;
+    }
     Dynarmic::A32::Context context;
     context.Regs() = ctx.cpu_registers;
     context.ExtRegs() = ctx.extension_registers;

src/core/arm/dynarmic/arm_dynarmic_32.h

@@ -50,6 +50,10 @@ public:
     u64 GetTPIDR_EL0() const override;
     void ChangeProcessorID(std::size_t new_core_id) override;
 
+    bool IsInThumbMode() const {
+        return (GetPSTATE() & 0x20) != 0;
+    }
+
     void SaveContext(ThreadContext32& ctx) override;
     void SaveContext(ThreadContext64& ctx) override {}
     void LoadContext(const ThreadContext32& ctx) override;

src/core/arm/dynarmic/arm_dynarmic_64.cpp

@@ -294,6 +294,9 @@ void ARM_Dynarmic_64::ChangeProcessorID(std::size_t new_core_id) {
 }
 
 void ARM_Dynarmic_64::SaveContext(ThreadContext64& ctx) {
+    if (!jit) {
+        return;
+    }
     ctx.cpu_registers = jit->GetRegisters();
     ctx.sp = jit->GetSP();
     ctx.pc = jit->GetPC();
@@ -305,6 +308,9 @@ void ARM_Dynarmic_64::SaveContext(ThreadContext64& ctx) {
 }
 
 void ARM_Dynarmic_64::LoadContext(const ThreadContext64& ctx) {
+    if (!jit) {
+        return;
+    }
     jit->SetRegisters(ctx.cpu_registers);
     jit->SetSP(ctx.sp);
     jit->SetPC(ctx.pc);

src/core/core.cpp

@@ -28,15 +28,14 @@
 #include "core/hardware_interrupt_manager.h"
 #include "core/hle/kernel/client_port.h"
 #include "core/hle/kernel/k_scheduler.h"
+#include "core/hle/kernel/k_thread.h"
 #include "core/hle/kernel/kernel.h"
 #include "core/hle/kernel/physical_core.h"
 #include "core/hle/kernel/process.h"
-#include "core/hle/kernel/thread.h"
 #include "core/hle/service/am/applets/applets.h"
 #include "core/hle/service/apm/controller.h"
 #include "core/hle/service/filesystem/filesystem.h"
 #include "core/hle/service/glue/manager.h"
-#include "core/hle/service/lm/manager.h"
 #include "core/hle/service/service.h"
 #include "core/hle/service/sm/sm.h"
 #include "core/hle/service/time/time_manager.h"
@@ -293,8 +292,6 @@ struct System::Impl {
                                         perf_stats->GetMeanFrametime());
         }
 
-        lm_manager.Flush();
-
         is_powered_on = false;
         exit_lock = false;
 
@@ -398,7 +395,6 @@ struct System::Impl {
 
     /// Service State
     Service::Glue::ARPManager arp_manager;
-    Service::LM::Manager lm_manager{reporter};
     Service::Time::TimeManager time_manager;
 
     /// Service manager
@@ -720,14 +716,6 @@ const Service::APM::Controller& System::GetAPMController() const {
     return impl->apm_controller;
 }
 
-Service::LM::Manager& System::GetLogManager() {
-    return impl->lm_manager;
-}
-
-const Service::LM::Manager& System::GetLogManager() const {
-    return impl->lm_manager;
-}
-
 Service::Time::TimeManager& System::GetTimeManager() {
     return impl->time_manager;
 }

src/core/core.h

@@ -62,10 +62,6 @@ namespace Glue {
 class ARPManager;
 }
 
-namespace LM {
-class Manager;
-} // namespace LM
-
 namespace SM {
 class ServiceManager;
 } // namespace SM
@@ -351,9 +347,6 @@ public:
     [[nodiscard]] Service::APM::Controller& GetAPMController();
     [[nodiscard]] const Service::APM::Controller& GetAPMController() const;
 
-    [[nodiscard]] Service::LM::Manager& GetLogManager();
-    [[nodiscard]] const Service::LM::Manager& GetLogManager() const;
-
     [[nodiscard]] Service::Time::TimeManager& GetTimeManager();
     [[nodiscard]] const Service::Time::TimeManager& GetTimeManager() const;
 

src/core/cpu_manager.cpp

@@ -11,9 +11,9 @@
 #include "core/core_timing.h"
 #include "core/cpu_manager.h"
 #include "core/hle/kernel/k_scheduler.h"
+#include "core/hle/kernel/k_thread.h"
 #include "core/hle/kernel/kernel.h"
 #include "core/hle/kernel/physical_core.h"
-#include "core/hle/kernel/thread.h"
 #include "video_core/gpu.h"
 
 namespace Core {
@@ -147,7 +147,7 @@ void CpuManager::MultiCoreRunSuspendThread() {
     while (true) {
         auto core = kernel.GetCurrentHostThreadID();
         auto& scheduler = *kernel.CurrentScheduler();
-        Kernel::Thread* current_thread = scheduler.GetCurrentThread();
+        Kernel::KThread* current_thread = scheduler.GetCurrentThread();
         Common::Fiber::YieldTo(current_thread->GetHostContext(), core_data[core].host_context);
         ASSERT(scheduler.ContextSwitchPending());
         ASSERT(core == kernel.GetCurrentHostThreadID());
@@ -208,7 +208,6 @@ void CpuManager::SingleCoreRunGuestThread() {
 
 void CpuManager::SingleCoreRunGuestLoop() {
     auto& kernel = system.Kernel();
-    auto* thread = kernel.CurrentScheduler()->GetCurrentThread();
     while (true) {
         auto* physical_core = &kernel.CurrentPhysicalCore();
         system.EnterDynarmicProfile();
@@ -217,9 +216,9 @@ void CpuManager::SingleCoreRunGuestLoop() {
             physical_core = &kernel.CurrentPhysicalCore();
         }
         system.ExitDynarmicProfile();
-        thread->SetPhantomMode(true);
+        kernel.SetIsPhantomModeForSingleCore(true);
         system.CoreTiming().Advance();
-        thread->SetPhantomMode(false);
+        kernel.SetIsPhantomModeForSingleCore(false);
         physical_core->ArmInterface().ClearExclusiveState();
         PreemptSingleCore();
         auto& scheduler = kernel.Scheduler(current_core);
@@ -245,7 +244,7 @@ void CpuManager::SingleCoreRunSuspendThread() {
     while (true) {
         auto core = kernel.GetCurrentHostThreadID();
         auto& scheduler = *kernel.CurrentScheduler();
-        Kernel::Thread* current_thread = scheduler.GetCurrentThread();
+        Kernel::KThread* current_thread = scheduler.GetCurrentThread();
         Common::Fiber::YieldTo(current_thread->GetHostContext(), core_data[0].host_context);
         ASSERT(scheduler.ContextSwitchPending());
         ASSERT(core == kernel.GetCurrentHostThreadID());
@@ -255,22 +254,23 @@ void CpuManager::SingleCoreRunSuspendThread() {
 
 void CpuManager::PreemptSingleCore(bool from_running_enviroment) {
     {
-        auto& scheduler = system.Kernel().Scheduler(current_core);
-        Kernel::Thread* current_thread = scheduler.GetCurrentThread();
+        auto& kernel = system.Kernel();
+        auto& scheduler = kernel.Scheduler(current_core);
+        Kernel::KThread* current_thread = scheduler.GetCurrentThread();
         if (idle_count >= 4 || from_running_enviroment) {
             if (!from_running_enviroment) {
                 system.CoreTiming().Idle();
                 idle_count = 0;
             }
-            current_thread->SetPhantomMode(true);
+            kernel.SetIsPhantomModeForSingleCore(true);
             system.CoreTiming().Advance();
-            current_thread->SetPhantomMode(false);
+            kernel.SetIsPhantomModeForSingleCore(false);
         }
         current_core.store((current_core + 1) % Core::Hardware::NUM_CPU_CORES);
         system.CoreTiming().ResetTicks();
         scheduler.Unload(scheduler.GetCurrentThread());
 
-        auto& next_scheduler = system.Kernel().Scheduler(current_core);
+        auto& next_scheduler = kernel.Scheduler(current_core);
         Common::Fiber::YieldTo(current_thread->GetHostContext(), next_scheduler.ControlContext());
     }
 
@@ -278,8 +278,7 @@ void CpuManager::PreemptSingleCore(bool from_running_enviroment) {
     {
         auto& scheduler = system.Kernel().Scheduler(current_core);
         scheduler.Reload(scheduler.GetCurrentThread());
-        auto* currrent_thread2 = scheduler.GetCurrentThread();
-        if (!currrent_thread2->IsIdleThread()) {
+        if (!scheduler.IsIdle()) {
             idle_count = 0;
         }
     }

src/core/crypto/key_manager.cpp

@@ -568,6 +568,11 @@ KeyManager::KeyManager() {
     // Initialize keys
     const std::string hactool_keys_dir = Common::FS::GetHactoolConfigurationPath();
     const std::string yuzu_keys_dir = Common::FS::GetUserPath(Common::FS::UserPath::KeysDir);
+
+    if (!Common::FS::Exists(yuzu_keys_dir)) {
+        Common::FS::CreateDir(yuzu_keys_dir);
+    }
+
     if (Settings::values.use_dev_keys) {
         dev_mode = true;
         AttemptLoadKeyFile(yuzu_keys_dir, hactool_keys_dir, "dev.keys", false);

src/core/file_sys/content_archive.cpp

@@ -43,17 +43,17 @@ static_assert(sizeof(IVFCLevel) == 0x18, "IVFCLevel has incorrect size.");
 struct IVFCHeader {
     u32_le magic;
     u32_le magic_number;
-    INSERT_UNION_PADDING_BYTES(8);
+    INSERT_PADDING_BYTES_NOINIT(8);
     std::array<IVFCLevel, 6> levels;
-    INSERT_UNION_PADDING_BYTES(64);
+    INSERT_PADDING_BYTES_NOINIT(64);
 };
 static_assert(sizeof(IVFCHeader) == 0xE0, "IVFCHeader has incorrect size.");
 
 struct NCASectionHeaderBlock {
-    INSERT_UNION_PADDING_BYTES(3);
+    INSERT_PADDING_BYTES_NOINIT(3);
     NCASectionFilesystemType filesystem_type;
     NCASectionCryptoType crypto_type;
-    INSERT_UNION_PADDING_BYTES(3);
+    INSERT_PADDING_BYTES_NOINIT(3);
 };
 static_assert(sizeof(NCASectionHeaderBlock) == 0x8, "NCASectionHeaderBlock has incorrect size.");
 
@@ -61,7 +61,7 @@ struct NCASectionRaw {
     NCASectionHeaderBlock header;
     std::array<u8, 0x138> block_data;
     std::array<u8, 0x8> section_ctr;
-    INSERT_UNION_PADDING_BYTES(0xB8);
+    INSERT_PADDING_BYTES_NOINIT(0xB8);
 };
 static_assert(sizeof(NCASectionRaw) == 0x200, "NCASectionRaw has incorrect size.");
 
@@ -69,19 +69,19 @@ struct PFS0Superblock {
     NCASectionHeaderBlock header_block;
     std::array<u8, 0x20> hash;
     u32_le size;
-    INSERT_UNION_PADDING_BYTES(4);
+    INSERT_PADDING_BYTES_NOINIT(4);
     u64_le hash_table_offset;
     u64_le hash_table_size;
     u64_le pfs0_header_offset;
     u64_le pfs0_size;
-    INSERT_UNION_PADDING_BYTES(0x1B0);
+    INSERT_PADDING_BYTES_NOINIT(0x1B0);
 };
 static_assert(sizeof(PFS0Superblock) == 0x200, "PFS0Superblock has incorrect size.");
 
 struct RomFSSuperblock {
     NCASectionHeaderBlock header_block;
     IVFCHeader ivfc;
-    INSERT_UNION_PADDING_BYTES(0x118);
+    INSERT_PADDING_BYTES_NOINIT(0x118);
 };
 static_assert(sizeof(RomFSSuperblock) == 0x200, "RomFSSuperblock has incorrect size.");
 
@@ -89,19 +89,19 @@ struct BKTRHeader {
     u64_le offset;
     u64_le size;
     u32_le magic;
-    INSERT_UNION_PADDING_BYTES(0x4);
+    INSERT_PADDING_BYTES_NOINIT(0x4);
     u32_le number_entries;
-    INSERT_UNION_PADDING_BYTES(0x4);
+    INSERT_PADDING_BYTES_NOINIT(0x4);
 };
 static_assert(sizeof(BKTRHeader) == 0x20, "BKTRHeader has incorrect size.");
 
 struct BKTRSuperblock {
     NCASectionHeaderBlock header_block;
     IVFCHeader ivfc;
-    INSERT_UNION_PADDING_BYTES(0x18);
+    INSERT_PADDING_BYTES_NOINIT(0x18);
     BKTRHeader relocation;
     BKTRHeader subsection;
-    INSERT_UNION_PADDING_BYTES(0xC0);
+    INSERT_PADDING_BYTES_NOINIT(0xC0);
 };
 static_assert(sizeof(BKTRSuperblock) == 0x200, "BKTRSuperblock has incorrect size.");
 

src/core/file_sys/savedata_factory.h

@@ -58,7 +58,7 @@ struct SaveDataAttribute {
     SaveDataType type;
     SaveDataRank rank;
     u16 index;
-    INSERT_PADDING_BYTES(4);
+    INSERT_PADDING_BYTES_NOINIT(4);
     u64 zero_1;
     u64 zero_2;
     u64 zero_3;
@@ -72,7 +72,7 @@ struct SaveDataExtraData {
     u64 owner_id;
     s64 timestamp;
     SaveDataFlags flags;
-    INSERT_PADDING_BYTES(4);
+    INSERT_PADDING_BYTES_NOINIT(4);
     s64 available_size;
     s64 journal_size;
     s64 commit_id;

src/core/file_sys/vfs_real.cpp

@@ -133,8 +133,11 @@ VirtualFile RealVfsFilesystem::MoveFile(std::string_view old_path_, std::string_
         }
 
         cache.erase(old_path);
-        file->Open(new_path, "r+b");
-        cache.insert_or_assign(new_path, std::move(file));
+        if (file->Open(new_path, "r+b")) {
+            cache.insert_or_assign(new_path, std::move(file));
+        } else {
+            LOG_ERROR(Service_FS, "Failed to open path {} in order to re-cache it", new_path);
+        }
     } else {
         UNREACHABLE();
         return nullptr;
@@ -214,9 +217,12 @@ VirtualDir RealVfsFilesystem::MoveDirectory(std::string_view old_path_,
         }
 
         auto file = cached.lock();
-        file->Open(file_new_path, "r+b");
         cache.erase(file_old_path);
-        cache.insert_or_assign(std::move(file_new_path), std::move(file));
+        if (file->Open(file_new_path, "r+b")) {
+            cache.insert_or_assign(std::move(file_new_path), std::move(file));
+        } else {
+            LOG_ERROR(Service_FS, "Failed to open path {} in order to re-cache it", file_new_path);
+        }
     }
 
     return OpenDirectory(new_path, Mode::ReadWrite);

src/core/frontend/emu_window.cpp

@@ -21,21 +21,18 @@ public:
 
     std::mutex mutex;
 
-    bool touch_pressed = false; ///< True if touchpad area is currently pressed, otherwise false
-
-    float touch_x = 0.0f; ///< Touchpad X-position
-    float touch_y = 0.0f; ///< Touchpad Y-position
+    Input::TouchStatus status;
 
 private:
     class Device : public Input::TouchDevice {
     public:
         explicit Device(std::weak_ptr<TouchState>&& touch_state) : touch_state(touch_state) {}
-        std::tuple<float, float, bool> GetStatus() const override {
+        Input::TouchStatus GetStatus() const override {
             if (auto state = touch_state.lock()) {
                 std::lock_guard guard{state->mutex};
-                return std::make_tuple(state->touch_x, state->touch_y, state->touch_pressed);
+                return state->status;
             }
-            return std::make_tuple(0.0f, 0.0f, false);
+            return {};
         }
 
     private:
@@ -79,36 +76,44 @@ std::tuple<unsigned, unsigned> EmuWindow::ClipToTouchScreen(unsigned new_x, unsi
     return std::make_tuple(new_x, new_y);
 }
 
-void EmuWindow::TouchPressed(unsigned framebuffer_x, unsigned framebuffer_y) {
-    if (!IsWithinTouchscreen(framebuffer_layout, framebuffer_x, framebuffer_y))
+void EmuWindow::TouchPressed(unsigned framebuffer_x, unsigned framebuffer_y, std::size_t id) {
+    if (!IsWithinTouchscreen(framebuffer_layout, framebuffer_x, framebuffer_y)) {
         return;
+    }
+    if (id >= touch_state->status.size()) {
+        return;
+    }
 
     std::lock_guard guard{touch_state->mutex};
-    touch_state->touch_x =
+    const float x =
         static_cast<float>(framebuffer_x - framebuffer_layout.screen.left) /
         static_cast<float>(framebuffer_layout.screen.right - framebuffer_layout.screen.left);
-    touch_state->touch_y =
+    const float y =
         static_cast<float>(framebuffer_y - framebuffer_layout.screen.top) /
         static_cast<float>(framebuffer_layout.screen.bottom - framebuffer_layout.screen.top);
 
-    touch_state->touch_pressed = true;
+    touch_state->status[id] = std::make_tuple(x, y, true);
 }
 
-void EmuWindow::TouchReleased() {
+void EmuWindow::TouchReleased(std::size_t id) {
+    if (id >= touch_state->status.size()) {
+        return;
+    }
     std::lock_guard guard{touch_state->mutex};
-    touch_state->touch_pressed = false;
-    touch_state->touch_x = 0;
-    touch_state->touch_y = 0;
+    touch_state->status[id] = std::make_tuple(0.0f, 0.0f, false);
 }
 
-void EmuWindow::TouchMoved(unsigned framebuffer_x, unsigned framebuffer_y) {
-    if (!touch_state->touch_pressed)
+void EmuWindow::TouchMoved(unsigned framebuffer_x, unsigned framebuffer_y, std::size_t id) {
+    if (id >= touch_state->status.size()) {
+        return;
+    }
+    if (!std::get<2>(touch_state->status[id]))
         return;
 
     if (!IsWithinTouchscreen(framebuffer_layout, framebuffer_x, framebuffer_y))
         std::tie(framebuffer_x, framebuffer_y) = ClipToTouchScreen(framebuffer_x, framebuffer_y);
 
-    TouchPressed(framebuffer_x, framebuffer_y);
+    TouchPressed(framebuffer_x, framebuffer_y, id);
 }
 
 void EmuWindow::UpdateCurrentFramebufferLayout(unsigned width, unsigned height) {

src/core/frontend/emu_window.h

@@ -117,18 +117,23 @@ public:
      * Signal that a touch pressed event has occurred (e.g. mouse click pressed)
      * @param framebuffer_x Framebuffer x-coordinate that was pressed
      * @param framebuffer_y Framebuffer y-coordinate that was pressed
+     * @param id Touch event ID
      */
-    void TouchPressed(unsigned framebuffer_x, unsigned framebuffer_y);
+    void TouchPressed(unsigned framebuffer_x, unsigned framebuffer_y, std::size_t id);
 
-    /// Signal that a touch released event has occurred (e.g. mouse click released)
-    void TouchReleased();
+    /**
+     * Signal that a touch released event has occurred (e.g. mouse click released)
+     * @param id Touch event ID
+     */
+    void TouchReleased(std::size_t id);
 
     /**
      * Signal that a touch movement event has occurred (e.g. mouse was moved over the emu window)
      * @param framebuffer_x Framebuffer x-coordinate
      * @param framebuffer_y Framebuffer y-coordinate
+     * @param id Touch event ID
      */
-    void TouchMoved(unsigned framebuffer_x, unsigned framebuffer_y);
+    void TouchMoved(unsigned framebuffer_x, unsigned framebuffer_y, std::size_t id);
 
     /**
      * Returns currently active configuration.

src/core/frontend/input.h

@@ -21,6 +21,11 @@ enum class AnalogDirection : u8 {
     UP,
     DOWN,
 };
+struct AnalogProperties {
+    float deadzone;
+    float range;
+    float threshold;
+};
 
 /// An abstract class template for an input device (a button, an analog input, etc.).
 template <typename StatusType>
@@ -30,6 +35,12 @@ public:
     virtual StatusType GetStatus() const {
         return {};
     }
+    virtual StatusType GetRawStatus() const {
+        return GetStatus();
+    }
+    virtual AnalogProperties GetAnalogProperties() const {
+        return {};
+    }
     virtual bool GetAnalogDirectionStatus([[maybe_unused]] AnalogDirection direction) const {
         return {};
     }
@@ -163,10 +174,11 @@ using MotionStatus = std::tuple<Common::Vec3<float>, Common::Vec3<float>, Common
 using MotionDevice = InputDevice<MotionStatus>;
 
 /**
- * A touch status is an object that returns a tuple of two floats and a bool. The floats are
- * x and y coordinates in the range 0.0 - 1.0, and the bool indicates whether it is pressed.
+ * A touch status is an object that returns an array of 16 tuple elements of two floats and a bool.
+ * The floats are x and y coordinates in the range 0.0 - 1.0, and the bool indicates whether it is
+ * pressed.
  */
-using TouchStatus = std::tuple<float, float, bool>;
+using TouchStatus = std::array<std::tuple<float, float, bool>, 16>;
 
 /**
  * A touch device is an input device that returns a touch status object

src/core/frontend/input_interpreter.cpp

@@ -25,6 +25,10 @@ void InputInterpreter::PollInput() {
     button_states[current_index] = button_state;
 }
 
+bool InputInterpreter::IsButtonPressed(HIDButton button) const {
+    return (button_states[current_index] & (1U << static_cast<u8>(button))) != 0;
+}
+
 bool InputInterpreter::IsButtonPressedOnce(HIDButton button) const {
     const bool current_press =
         (button_states[current_index] & (1U << static_cast<u8>(button))) != 0;

src/core/frontend/input_interpreter.h

@@ -66,6 +66,27 @@ public:
     /// Gets a button state from HID and inserts it into the array of button states.
     void PollInput();
 
+    /**
+     * Checks whether the button is pressed.
+     *
+     * @param button The button to check.
+     *
+     * @returns True when the button is pressed.
+     */
+    [[nodiscard]] bool IsButtonPressed(HIDButton button) const;
+
+    /**
+     * Checks whether any of the buttons in the parameter list is pressed.
+     *
+     * @tparam HIDButton The buttons to check.
+     *
+     * @returns True when at least one of the buttons is pressed.
+     */
+    template <HIDButton... T>
+    [[nodiscard]] bool IsAnyButtonPressed() {
+        return (IsButtonPressed(T) || ...);
+    }
+
     /**
      * The specified button is considered to be pressed once
      * if it is currently pressed and not pressed previously.
@@ -79,12 +100,12 @@ public:
     /**
      * Checks whether any of the buttons in the parameter list is pressed once.
      *
-     * @tparam HIDButton The buttons to check.
+     * @tparam T The buttons to check.
      *
      * @returns True when at least one of the buttons is pressed once.
      */
     template <HIDButton... T>
-    [[nodiscard]] bool IsAnyButtonPressedOnce() {
+    [[nodiscard]] bool IsAnyButtonPressedOnce() const {
         return (IsButtonPressedOnce(T) || ...);
     }
 
@@ -100,12 +121,12 @@ public:
     /**
      * Checks whether any of the buttons in the parameter list is held down.
      *
-     * @tparam HIDButton The buttons to check.
+     * @tparam T The buttons to check.
      *
      * @returns True when at least one of the buttons is held down.
      */
     template <HIDButton... T>
-    [[nodiscard]] bool IsAnyButtonHeld() {
+    [[nodiscard]] bool IsAnyButtonHeld() const {
         return (IsButtonHeld(T) || ...);
     }
 

src/core/hardware_properties.h

@@ -4,8 +4,10 @@
 
 #pragma once
 
+#include <array>
 #include <tuple>
 
+#include "common/bit_util.h"
 #include "common/common_types.h"
 
 namespace Core {
@@ -18,34 +20,12 @@ constexpr u64 BASE_CLOCK_RATE = 1019215872; // Switch cpu frequency is 1020MHz u
 constexpr u64 CNTFREQ = 19200000;           // Switch's hardware clock speed
 constexpr u32 NUM_CPU_CORES = 4;            // Number of CPU Cores
 
-} // namespace Hardware
-
-constexpr u32 INVALID_HOST_THREAD_ID = 0xFFFFFFFF;
-
-struct EmuThreadHandle {
-    u32 host_handle;
-    u32 guest_handle;
-
-    u64 GetRaw() const {
-        return (static_cast<u64>(host_handle) << 32) | guest_handle;
-    }
-
-    bool operator==(const EmuThreadHandle& rhs) const {
-        return std::tie(host_handle, guest_handle) == std::tie(rhs.host_handle, rhs.guest_handle);
-    }
-
-    bool operator!=(const EmuThreadHandle& rhs) const {
-        return !operator==(rhs);
-    }
-
-    static constexpr EmuThreadHandle InvalidHandle() {
-        constexpr u32 invalid_handle = 0xFFFFFFFF;
-        return {invalid_handle, invalid_handle};
-    }
-
-    bool IsInvalid() const {
-        return (*this) == InvalidHandle();
-    }
+// Virtual to Physical core map.
+constexpr std::array<s32, Common::BitSize<u64>()> VirtualToPhysicalCoreMap{
+    0, 1, 2, 3, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 3,
 };
 
+} // namespace Hardware
+
 } // namespace Core

src/core/hle/ipc.h

@@ -146,7 +146,7 @@ static_assert(sizeof(BufferDescriptorC) == 8, "BufferDescriptorC size is incorre
 
 struct DataPayloadHeader {
     u32_le magic;
-    INSERT_PADDING_WORDS(1);
+    INSERT_PADDING_WORDS_NOINIT(1);
 };
 static_assert(sizeof(DataPayloadHeader) == 8, "DataPayloadHeader size is incorrect");
 
@@ -160,7 +160,7 @@ struct DomainMessageHeader {
         // Used when responding to an IPC request, Server -> Client.
         struct {
             u32_le num_objects;
-            INSERT_UNION_PADDING_WORDS(3);
+            INSERT_PADDING_WORDS_NOINIT(3);
         };
 
         // Used when performing an IPC request, Client -> Server.
@@ -171,10 +171,10 @@ struct DomainMessageHeader {
                 BitField<16, 16, u32> size;
             };
             u32_le object_id;
-            INSERT_UNION_PADDING_WORDS(2);
+            INSERT_PADDING_WORDS_NOINIT(2);
         };
 
-        std::array<u32, 4> raw{};
+        std::array<u32, 4> raw;
     };
 };
 static_assert(sizeof(DomainMessageHeader) == 16, "DomainMessageHeader size is incorrect");

src/core/hle/kernel/client_port.cpp

@@ -4,11 +4,11 @@
 
 #include "core/hle/kernel/client_port.h"
 #include "core/hle/kernel/client_session.h"
-#include "core/hle/kernel/errors.h"
 #include "core/hle/kernel/hle_ipc.h"
 #include "core/hle/kernel/object.h"
 #include "core/hle/kernel/server_port.h"
 #include "core/hle/kernel/session.h"
+#include "core/hle/kernel/svc_results.h"
 
 namespace Kernel {
 
@@ -21,7 +21,7 @@ std::shared_ptr<ServerPort> ClientPort::GetServerPort() const {
 
 ResultVal<std::shared_ptr<ClientSession>> ClientPort::Connect() {
     if (active_sessions >= max_sessions) {
-        return ERR_MAX_CONNECTIONS_REACHED;
+        return ResultMaxConnectionsReached;
     }
     active_sessions++;
 

src/core/hle/kernel/client_port.h

@@ -51,6 +51,8 @@ public:
      */
     void ConnectionClosed();
 
+    void Finalize() override {}
+
 private:
     std::shared_ptr<ServerPort> server_port; ///< ServerPort associated with this client port.
     u32 max_sessions = 0;    ///< Maximum number of simultaneous sessions the port can have

src/core/hle/kernel/client_session.cpp

@@ -3,11 +3,11 @@
 // Refer to the license.txt file included.
 
 #include "core/hle/kernel/client_session.h"
-#include "core/hle/kernel/errors.h"
 #include "core/hle/kernel/hle_ipc.h"
+#include "core/hle/kernel/k_thread.h"
 #include "core/hle/kernel/server_session.h"
 #include "core/hle/kernel/session.h"
-#include "core/hle/kernel/thread.h"
+#include "core/hle/kernel/svc_results.h"
 #include "core/hle/result.h"
 
 namespace Kernel {
@@ -38,12 +38,12 @@ ResultVal<std::shared_ptr<ClientSession>> ClientSession::Create(KernelCore& kern
     return MakeResult(std::move(client_session));
 }
 
-ResultCode ClientSession::SendSyncRequest(std::shared_ptr<Thread> thread,
+ResultCode ClientSession::SendSyncRequest(std::shared_ptr<KThread> thread,
                                           Core::Memory::Memory& memory,
                                           Core::Timing::CoreTiming& core_timing) {
     // Keep ServerSession alive until we're done working with it.
     if (!parent->Server()) {
-        return ERR_SESSION_CLOSED_BY_REMOTE;
+        return ResultSessionClosedByRemote;
     }
 
     // Signal the server session that new data is available

src/core/hle/kernel/client_session.h

@@ -24,7 +24,7 @@ namespace Kernel {
 
 class KernelCore;
 class Session;
-class Thread;
+class KThread;
 
 class ClientSession final : public KSynchronizationObject {
 public:
@@ -46,11 +46,13 @@ public:
         return HANDLE_TYPE;
     }
 
-    ResultCode SendSyncRequest(std::shared_ptr<Thread> thread, Core::Memory::Memory& memory,
+    ResultCode SendSyncRequest(std::shared_ptr<KThread> thread, Core::Memory::Memory& memory,
                                Core::Timing::CoreTiming& core_timing);
 
     bool IsSignaled() const override;
 
+    void Finalize() override {}
+
 private:
     static ResultVal<std::shared_ptr<ClientSession>> Create(KernelCore& kernel,
                                                             std::shared_ptr<Session> parent,

src/core/hle/kernel/errors.h

@@ -1,43 +0,0 @@
-// Copyright 2018 yuzu emulator team
-// Licensed under GPLv2 or any later version
-// Refer to the license.txt file included.
-
-#pragma once
-
-#include "core/hle/result.h"
-
-namespace Kernel {
-
-// Confirmed Switch kernel error codes
-
-constexpr ResultCode ERR_MAX_CONNECTIONS_REACHED{ErrorModule::Kernel, 7};
-constexpr ResultCode ERR_INVALID_CAPABILITY_DESCRIPTOR{ErrorModule::Kernel, 14};
-constexpr ResultCode ERR_THREAD_TERMINATING{ErrorModule::Kernel, 59};
-constexpr ResultCode ERR_TERMINATION_REQUESTED{ErrorModule::Kernel, 59};
-constexpr ResultCode ERR_INVALID_SIZE{ErrorModule::Kernel, 101};
-constexpr ResultCode ERR_INVALID_ADDRESS{ErrorModule::Kernel, 102};
-constexpr ResultCode ERR_OUT_OF_RESOURCES{ErrorModule::Kernel, 103};
-constexpr ResultCode ERR_OUT_OF_MEMORY{ErrorModule::Kernel, 104};
-constexpr ResultCode ERR_HANDLE_TABLE_FULL{ErrorModule::Kernel, 105};
-constexpr ResultCode ERR_INVALID_ADDRESS_STATE{ErrorModule::Kernel, 106};
-constexpr ResultCode ERR_INVALID_CURRENT_MEMORY{ErrorModule::Kernel, 106};
-constexpr ResultCode ERR_INVALID_MEMORY_PERMISSIONS{ErrorModule::Kernel, 108};
-constexpr ResultCode ERR_INVALID_MEMORY_RANGE{ErrorModule::Kernel, 110};
-constexpr ResultCode ERR_INVALID_PROCESSOR_ID{ErrorModule::Kernel, 113};
-constexpr ResultCode ERR_INVALID_THREAD_PRIORITY{ErrorModule::Kernel, 112};
-constexpr ResultCode ERR_INVALID_HANDLE{ErrorModule::Kernel, 114};
-constexpr ResultCode ERR_INVALID_POINTER{ErrorModule::Kernel, 115};
-constexpr ResultCode ERR_INVALID_COMBINATION{ErrorModule::Kernel, 116};
-constexpr ResultCode RESULT_TIMEOUT{ErrorModule::Kernel, 117};
-constexpr ResultCode ERR_SYNCHRONIZATION_CANCELED{ErrorModule::Kernel, 118};
-constexpr ResultCode ERR_CANCELLED{ErrorModule::Kernel, 118};
-constexpr ResultCode ERR_OUT_OF_RANGE{ErrorModule::Kernel, 119};
-constexpr ResultCode ERR_INVALID_ENUM_VALUE{ErrorModule::Kernel, 120};
-constexpr ResultCode ERR_NOT_FOUND{ErrorModule::Kernel, 121};
-constexpr ResultCode ERR_BUSY{ErrorModule::Kernel, 122};
-constexpr ResultCode ERR_SESSION_CLOSED_BY_REMOTE{ErrorModule::Kernel, 123};
-constexpr ResultCode ERR_INVALID_STATE{ErrorModule::Kernel, 125};
-constexpr ResultCode ERR_RESERVED_VALUE{ErrorModule::Kernel, 126};
-constexpr ResultCode ERR_RESOURCE_LIMIT_EXCEEDED{ErrorModule::Kernel, 132};
-
-} // namespace Kernel

src/core/hle/kernel/global_scheduler_context.cpp

@@ -17,12 +17,12 @@ GlobalSchedulerContext::GlobalSchedulerContext(KernelCore& kernel)
 
 GlobalSchedulerContext::~GlobalSchedulerContext() = default;
 
-void GlobalSchedulerContext::AddThread(std::shared_ptr<Thread> thread) {
+void GlobalSchedulerContext::AddThread(std::shared_ptr<KThread> thread) {
     std::scoped_lock lock{global_list_guard};
     thread_list.push_back(std::move(thread));
 }
 
-void GlobalSchedulerContext::RemoveThread(std::shared_ptr<Thread> thread) {
+void GlobalSchedulerContext::RemoveThread(std::shared_ptr<KThread> thread) {
     std::scoped_lock lock{global_list_guard};
     thread_list.erase(std::remove(thread_list.begin(), thread_list.end(), thread),
                       thread_list.end());

src/core/hle/kernel/global_scheduler_context.h

@@ -12,7 +12,8 @@
 #include "core/hardware_properties.h"
 #include "core/hle/kernel/k_priority_queue.h"
 #include "core/hle/kernel/k_scheduler_lock.h"
-#include "core/hle/kernel/thread.h"
+#include "core/hle/kernel/k_thread.h"
+#include "core/hle/kernel/svc_types.h"
 
 namespace Kernel {
 
@@ -20,8 +21,12 @@ class KernelCore;
 class SchedulerLock;
 
 using KSchedulerPriorityQueue =
-    KPriorityQueue<Thread, Core::Hardware::NUM_CPU_CORES, THREADPRIO_LOWEST, THREADPRIO_HIGHEST>;
-constexpr s32 HighestCoreMigrationAllowedPriority = 2;
+    KPriorityQueue<KThread, Core::Hardware::NUM_CPU_CORES, Svc::LowestThreadPriority,
+                   Svc::HighestThreadPriority>;
+
+static constexpr s32 HighestCoreMigrationAllowedPriority = 2;
+static_assert(Svc::LowestThreadPriority >= HighestCoreMigrationAllowedPriority);
+static_assert(Svc::HighestThreadPriority <= HighestCoreMigrationAllowedPriority);
 
 class GlobalSchedulerContext final {
     friend class KScheduler;
@@ -33,13 +38,13 @@ public:
     ~GlobalSchedulerContext();
 
     /// Adds a new thread to the scheduler
-    void AddThread(std::shared_ptr<Thread> thread);
+    void AddThread(std::shared_ptr<KThread> thread);
 
     /// Removes a thread from the scheduler
-    void RemoveThread(std::shared_ptr<Thread> thread);
+    void RemoveThread(std::shared_ptr<KThread> thread);
 
     /// Returns a list of all threads managed by the scheduler
-    [[nodiscard]] const std::vector<std::shared_ptr<Thread>>& GetThreadList() const {
+    [[nodiscard]] const std::vector<std::shared_ptr<KThread>>& GetThreadList() const {
         return thread_list;
     }
 
@@ -74,7 +79,7 @@ private:
     LockType scheduler_lock;
 
     /// Lists all thread ids that aren't deleted/etc.
-    std::vector<std::shared_ptr<Thread>> thread_list;
+    std::vector<std::shared_ptr<KThread>> thread_list;
     Common::SpinLock global_list_guard{};
 };
 

src/core/hle/kernel/handle_table.cpp

@@ -6,12 +6,12 @@
 #include "common/assert.h"
 #include "common/logging/log.h"
 #include "core/core.h"
-#include "core/hle/kernel/errors.h"
 #include "core/hle/kernel/handle_table.h"
 #include "core/hle/kernel/k_scheduler.h"
+#include "core/hle/kernel/k_thread.h"
 #include "core/hle/kernel/kernel.h"
 #include "core/hle/kernel/process.h"
-#include "core/hle/kernel/thread.h"
+#include "core/hle/kernel/svc_results.h"
 
 namespace Kernel {
 namespace {
@@ -33,7 +33,7 @@ HandleTable::~HandleTable() = default;
 ResultCode HandleTable::SetSize(s32 handle_table_size) {
     if (static_cast<u32>(handle_table_size) > MAX_COUNT) {
         LOG_ERROR(Kernel, "Handle table size {} is greater than {}", handle_table_size, MAX_COUNT);
-        return ERR_OUT_OF_MEMORY;
+        return ResultOutOfMemory;
     }
 
     // Values less than or equal to zero indicate to use the maximum allowable
@@ -53,7 +53,7 @@ ResultVal<Handle> HandleTable::Create(std::shared_ptr<Object> obj) {
     const u16 slot = next_free_slot;
     if (slot >= table_size) {
         LOG_ERROR(Kernel, "Unable to allocate Handle, too many slots in use.");
-        return ERR_HANDLE_TABLE_FULL;
+        return ResultHandleTableFull;
     }
     next_free_slot = generations[slot];
 
@@ -76,7 +76,7 @@ ResultVal<Handle> HandleTable::Duplicate(Handle handle) {
     std::shared_ptr<Object> object = GetGeneric(handle);
     if (object == nullptr) {
         LOG_ERROR(Kernel, "Tried to duplicate invalid handle: {:08X}", handle);
-        return ERR_INVALID_HANDLE;
+        return ResultInvalidHandle;
     }
     return Create(std::move(object));
 }
@@ -84,11 +84,15 @@ ResultVal<Handle> HandleTable::Duplicate(Handle handle) {
 ResultCode HandleTable::Close(Handle handle) {
     if (!IsValid(handle)) {
         LOG_ERROR(Kernel, "Handle is not valid! handle={:08X}", handle);
-        return ERR_INVALID_HANDLE;
+        return ResultInvalidHandle;
     }
 
     const u16 slot = GetSlot(handle);
 
+    if (objects[slot].use_count() == 1) {
+        objects[slot]->Finalize();
+    }
+
     objects[slot] = nullptr;
 
     generations[slot] = next_free_slot;

src/core/hle/kernel/hle_ipc.cpp

@@ -14,19 +14,19 @@
 #include "common/common_types.h"
 #include "common/logging/log.h"
 #include "core/hle/ipc_helpers.h"
-#include "core/hle/kernel/errors.h"
 #include "core/hle/kernel/handle_table.h"
 #include "core/hle/kernel/hle_ipc.h"
+#include "core/hle/kernel/k_readable_event.h"
 #include "core/hle/kernel/k_scheduler.h"
 #include "core/hle/kernel/k_scoped_scheduler_lock_and_sleep.h"
+#include "core/hle/kernel/k_thread.h"
+#include "core/hle/kernel/k_writable_event.h"
 #include "core/hle/kernel/kernel.h"
 #include "core/hle/kernel/object.h"
 #include "core/hle/kernel/process.h"
-#include "core/hle/kernel/readable_event.h"
 #include "core/hle/kernel/server_session.h"
-#include "core/hle/kernel/thread.h"
+#include "core/hle/kernel/svc_results.h"
 #include "core/hle/kernel/time_manager.h"
-#include "core/hle/kernel/writable_event.h"
 #include "core/memory.h"
 
 namespace Kernel {
@@ -48,7 +48,7 @@ void SessionRequestHandler::ClientDisconnected(
 
 HLERequestContext::HLERequestContext(KernelCore& kernel, Core::Memory::Memory& memory,
                                      std::shared_ptr<ServerSession> server_session,
-                                     std::shared_ptr<Thread> thread)
+                                     std::shared_ptr<KThread> thread)
     : server_session(std::move(server_session)),
       thread(std::move(thread)), kernel{kernel}, memory{memory} {
     cmd_buf[0] = 0;
@@ -182,7 +182,7 @@ ResultCode HLERequestContext::PopulateFromIncomingCommandBuffer(const HandleTabl
     return RESULT_SUCCESS;
 }
 
-ResultCode HLERequestContext::WriteToOutgoingCommandBuffer(Thread& thread) {
+ResultCode HLERequestContext::WriteToOutgoingCommandBuffer(KThread& thread) {
     auto& owner_process = *thread.GetOwnerProcess();
     auto& handle_table = owner_process.GetHandleTable();
 
@@ -338,6 +338,28 @@ std::size_t HLERequestContext::GetWriteBufferSize(std::size_t buffer_index) cons
     return 0;
 }
 
+bool HLERequestContext::CanReadBuffer(std::size_t buffer_index) const {
+    const bool is_buffer_a{BufferDescriptorA().size() > buffer_index &&
+                           BufferDescriptorA()[buffer_index].Size()};
+
+    if (is_buffer_a) {
+        return BufferDescriptorA().size() > buffer_index;
+    } else {
+        return BufferDescriptorX().size() > buffer_index;
+    }
+}
+
+bool HLERequestContext::CanWriteBuffer(std::size_t buffer_index) const {
+    const bool is_buffer_b{BufferDescriptorB().size() > buffer_index &&
+                           BufferDescriptorB()[buffer_index].Size()};
+
+    if (is_buffer_b) {
+        return BufferDescriptorB().size() > buffer_index;
+    } else {
+        return BufferDescriptorC().size() > buffer_index;
+    }
+}
+
 std::string HLERequestContext::Description() const {
     if (!command_header) {
         return "No command header available";

src/core/hle/kernel/hle_ipc.h

@@ -40,9 +40,9 @@ class HLERequestContext;
 class KernelCore;
 class Process;
 class ServerSession;
-class Thread;
-class ReadableEvent;
-class WritableEvent;
+class KThread;
+class KReadableEvent;
+class KWritableEvent;
 
 enum class ThreadWakeupReason;
 
@@ -110,7 +110,7 @@ class HLERequestContext {
 public:
     explicit HLERequestContext(KernelCore& kernel, Core::Memory::Memory& memory,
                                std::shared_ptr<ServerSession> session,
-                               std::shared_ptr<Thread> thread);
+                               std::shared_ptr<KThread> thread);
     ~HLERequestContext();
 
     /// Returns a pointer to the IPC command buffer for this request.
@@ -126,15 +126,12 @@ public:
         return server_session;
     }
 
-    using WakeupCallback = std::function<void(
-        std::shared_ptr<Thread> thread, HLERequestContext& context, ThreadWakeupReason reason)>;
-
     /// Populates this context with data from the requesting process/thread.
     ResultCode PopulateFromIncomingCommandBuffer(const HandleTable& handle_table,
                                                  u32_le* src_cmdbuf);
 
     /// Writes data from this context back to the requesting process/thread.
-    ResultCode WriteToOutgoingCommandBuffer(Thread& thread);
+    ResultCode WriteToOutgoingCommandBuffer(KThread& thread);
 
     u32_le GetCommand() const {
         return command;
@@ -207,6 +204,12 @@ public:
     /// Helper function to get the size of the output buffer
     std::size_t GetWriteBufferSize(std::size_t buffer_index = 0) const;
 
+    /// Helper function to test whether the input buffer at buffer_index can be read
+    bool CanReadBuffer(std::size_t buffer_index = 0) const;
+
+    /// Helper function to test whether the output buffer at buffer_index can be written
+    bool CanWriteBuffer(std::size_t buffer_index = 0) const;
+
     template <typename T>
     std::shared_ptr<T> GetCopyObject(std::size_t index) {
         return DynamicObjectCast<T>(copy_objects.at(index));
@@ -261,11 +264,11 @@ public:
 
     std::string Description() const;
 
-    Thread& GetThread() {
+    KThread& GetThread() {
         return *thread;
     }
 
-    const Thread& GetThread() const {
+    const KThread& GetThread() const {
         return *thread;
     }
 
@@ -280,7 +283,7 @@ private:
 
     std::array<u32, IPC::COMMAND_BUFFER_LENGTH> cmd_buf;
     std::shared_ptr<Kernel::ServerSession> server_session;
-    std::shared_ptr<Thread> thread;
+    std::shared_ptr<KThread> thread;
     // TODO(yuriks): Check common usage of this and optimize size accordingly
     boost::container::small_vector<std::shared_ptr<Object>, 8> move_objects;
     boost::container::small_vector<std::shared_ptr<Object>, 8> copy_objects;

src/core/hle/kernel/k_address_arbiter.cpp

@@ -7,9 +7,9 @@
 #include "core/hle/kernel/k_address_arbiter.h"
 #include "core/hle/kernel/k_scheduler.h"
 #include "core/hle/kernel/k_scoped_scheduler_lock_and_sleep.h"
+#include "core/hle/kernel/k_thread.h"
 #include "core/hle/kernel/kernel.h"
 #include "core/hle/kernel/svc_results.h"
-#include "core/hle/kernel/thread.h"
 #include "core/hle/kernel/time_manager.h"
 #include "core/memory.h"
 
@@ -96,7 +96,7 @@ ResultCode KAddressArbiter::Signal(VAddr addr, s32 count) {
         auto it = thread_tree.nfind_light({addr, -1});
         while ((it != thread_tree.end()) && (count <= 0 || num_waiters < count) &&
                (it->GetAddressArbiterKey() == addr)) {
-            Thread* target_thread = std::addressof(*it);
+            KThread* target_thread = std::addressof(*it);
             target_thread->SetSyncedObject(nullptr, RESULT_SUCCESS);
 
             ASSERT(target_thread->IsWaitingForAddressArbiter());
@@ -118,14 +118,18 @@ ResultCode KAddressArbiter::SignalAndIncrementIfEqual(VAddr addr, s32 value, s32
 
         // Check the userspace value.
         s32 user_value{};
-        R_UNLESS(UpdateIfEqual(system, std::addressof(user_value), addr, value, value + 1),
-                 Svc::ResultInvalidCurrentMemory);
-        R_UNLESS(user_value == value, Svc::ResultInvalidState);
+        if (!UpdateIfEqual(system, &user_value, addr, value, value + 1)) {
+            LOG_ERROR(Kernel, "Invalid current memory!");
+            return ResultInvalidCurrentMemory;
+        }
+        if (user_value != value) {
+            return ResultInvalidState;
+        }
 
         auto it = thread_tree.nfind_light({addr, -1});
         while ((it != thread_tree.end()) && (count <= 0 || num_waiters < count) &&
                (it->GetAddressArbiterKey() == addr)) {
-            Thread* target_thread = std::addressof(*it);
+            KThread* target_thread = std::addressof(*it);
             target_thread->SetSyncedObject(nullptr, RESULT_SUCCESS);
 
             ASSERT(target_thread->IsWaitingForAddressArbiter());
@@ -143,61 +147,34 @@ ResultCode KAddressArbiter::SignalAndModifyByWaitingCountIfEqual(VAddr addr, s32
     // Perform signaling.
     s32 num_waiters{};
     {
-        KScopedSchedulerLock sl(kernel);
+        [[maybe_unused]] const KScopedSchedulerLock sl(kernel);
 
         auto it = thread_tree.nfind_light({addr, -1});
         // Determine the updated value.
         s32 new_value{};
-        if (/*GetTargetFirmware() >= TargetFirmware_7_0_0*/ true) {
-            if (count <= 0) {
-                if ((it != thread_tree.end()) && (it->GetAddressArbiterKey() == addr)) {
-                    new_value = value - 2;
-                } else {
-                    new_value = value + 1;
-                }
+        if (count <= 0) {
+            if (it != thread_tree.end() && it->GetAddressArbiterKey() == addr) {
+                new_value = value - 2;
             } else {
-                if ((it != thread_tree.end()) && (it->GetAddressArbiterKey() == addr)) {
-                    auto tmp_it = it;
-                    s32 tmp_num_waiters{};
-                    while ((++tmp_it != thread_tree.end()) &&
-                           (tmp_it->GetAddressArbiterKey() == addr)) {
-                        if ((tmp_num_waiters++) >= count) {
-                            break;
-                        }
-                    }
-
-                    if (tmp_num_waiters < count) {
-                        new_value = value - 1;
-                    } else {
-                        new_value = value;
-                    }
-                } else {
-                    new_value = value + 1;
-                }
+                new_value = value + 1;
             }
         } else {
-            if (count <= 0) {
-                if ((it != thread_tree.end()) && (it->GetAddressArbiterKey() == addr)) {
-                    new_value = value - 1;
-                } else {
-                    new_value = value + 1;
-                }
-            } else {
+            if (it != thread_tree.end() && it->GetAddressArbiterKey() == addr) {
                 auto tmp_it = it;
                 s32 tmp_num_waiters{};
-                while ((tmp_it != thread_tree.end()) && (tmp_it->GetAddressArbiterKey() == addr) &&
-                       (tmp_num_waiters < count + 1)) {
-                    ++tmp_num_waiters;
-                    ++tmp_it;
+                while (++tmp_it != thread_tree.end() && tmp_it->GetAddressArbiterKey() == addr) {
+                    if (tmp_num_waiters++ >= count) {
+                        break;
+                    }
                 }
 
-                if (tmp_num_waiters == 0) {
-                    new_value = value + 1;
-                } else if (tmp_num_waiters <= count) {
+                if (tmp_num_waiters < count) {
                     new_value = value - 1;
                 } else {
                     new_value = value;
                 }
+            } else {
+                new_value = value + 1;
             }
         }
 
@@ -205,17 +182,22 @@ ResultCode KAddressArbiter::SignalAndModifyByWaitingCountIfEqual(VAddr addr, s32
         s32 user_value{};
         bool succeeded{};
         if (value != new_value) {
-            succeeded = UpdateIfEqual(system, std::addressof(user_value), addr, value, new_value);
+            succeeded = UpdateIfEqual(system, &user_value, addr, value, new_value);
         } else {
-            succeeded = ReadFromUser(system, std::addressof(user_value), addr);
+            succeeded = ReadFromUser(system, &user_value, addr);
         }
 
-        R_UNLESS(succeeded, Svc::ResultInvalidCurrentMemory);
-        R_UNLESS(user_value == value, Svc::ResultInvalidState);
+        if (!succeeded) {
+            LOG_ERROR(Kernel, "Invalid current memory!");
+            return ResultInvalidCurrentMemory;
+        }
+        if (user_value != value) {
+            return ResultInvalidState;
+        }
 
         while ((it != thread_tree.end()) && (count <= 0 || num_waiters < count) &&
                (it->GetAddressArbiterKey() == addr)) {
-            Thread* target_thread = std::addressof(*it);
+            KThread* target_thread = std::addressof(*it);
             target_thread->SetSyncedObject(nullptr, RESULT_SUCCESS);
 
             ASSERT(target_thread->IsWaitingForAddressArbiter());
@@ -231,59 +213,55 @@ ResultCode KAddressArbiter::SignalAndModifyByWaitingCountIfEqual(VAddr addr, s32
 
 ResultCode KAddressArbiter::WaitIfLessThan(VAddr addr, s32 value, bool decrement, s64 timeout) {
     // Prepare to wait.
-    Thread* cur_thread = kernel.CurrentScheduler()->GetCurrentThread();
-    Handle timer = InvalidHandle;
+    KThread* cur_thread = kernel.CurrentScheduler()->GetCurrentThread();
 
     {
-        KScopedSchedulerLockAndSleep slp(kernel, timer, cur_thread, timeout);
+        KScopedSchedulerLockAndSleep slp{kernel, cur_thread, timeout};
 
         // Check that the thread isn't terminating.
         if (cur_thread->IsTerminationRequested()) {
             slp.CancelSleep();
-            return Svc::ResultTerminationRequested;
+            return ResultTerminationRequested;
         }
 
         // Set the synced object.
-        cur_thread->SetSyncedObject(nullptr, Svc::ResultTimedOut);
+        cur_thread->SetSyncedObject(nullptr, ResultTimedOut);
 
         // Read the value from userspace.
         s32 user_value{};
         bool succeeded{};
         if (decrement) {
-            succeeded = DecrementIfLessThan(system, std::addressof(user_value), addr, value);
+            succeeded = DecrementIfLessThan(system, &user_value, addr, value);
         } else {
-            succeeded = ReadFromUser(system, std::addressof(user_value), addr);
+            succeeded = ReadFromUser(system, &user_value, addr);
         }
 
         if (!succeeded) {
             slp.CancelSleep();
-            return Svc::ResultInvalidCurrentMemory;
+            return ResultInvalidCurrentMemory;
         }
 
         // Check that the value is less than the specified one.
         if (user_value >= value) {
             slp.CancelSleep();
-            return Svc::ResultInvalidState;
+            return ResultInvalidState;
         }
 
         // Check that the timeout is non-zero.
         if (timeout == 0) {
             slp.CancelSleep();
-            return Svc::ResultTimedOut;
+            return ResultTimedOut;
         }
 
         // Set the arbiter.
-        cur_thread->SetAddressArbiter(std::addressof(thread_tree), addr);
+        cur_thread->SetAddressArbiter(&thread_tree, addr);
         thread_tree.insert(*cur_thread);
         cur_thread->SetState(ThreadState::Waiting);
         cur_thread->SetWaitReasonForDebugging(ThreadWaitReasonForDebugging::Arbitration);
     }
 
     // Cancel the timer wait.
-    if (timer != InvalidHandle) {
-        auto& time_manager = kernel.TimeManager();
-        time_manager.UnscheduleTimeEvent(timer);
-    }
+    kernel.TimeManager().UnscheduleTimeEvent(cur_thread);
 
     // Remove from the address arbiter.
     {
@@ -297,57 +275,53 @@ ResultCode KAddressArbiter::WaitIfLessThan(VAddr addr, s32 value, bool decrement
 
     // Get the result.
     KSynchronizationObject* dummy{};
-    return cur_thread->GetWaitResult(std::addressof(dummy));
+    return cur_thread->GetWaitResult(&dummy);
 }
 
 ResultCode KAddressArbiter::WaitIfEqual(VAddr addr, s32 value, s64 timeout) {
     // Prepare to wait.
-    Thread* cur_thread = kernel.CurrentScheduler()->GetCurrentThread();
-    Handle timer = InvalidHandle;
+    KThread* cur_thread = kernel.CurrentScheduler()->GetCurrentThread();
 
     {
-        KScopedSchedulerLockAndSleep slp(kernel, timer, cur_thread, timeout);
+        KScopedSchedulerLockAndSleep slp{kernel, cur_thread, timeout};
 
         // Check that the thread isn't terminating.
         if (cur_thread->IsTerminationRequested()) {
             slp.CancelSleep();
-            return Svc::ResultTerminationRequested;
+            return ResultTerminationRequested;
         }
 
         // Set the synced object.
-        cur_thread->SetSyncedObject(nullptr, Svc::ResultTimedOut);
+        cur_thread->SetSyncedObject(nullptr, ResultTimedOut);
 
         // Read the value from userspace.
         s32 user_value{};
-        if (!ReadFromUser(system, std::addressof(user_value), addr)) {
+        if (!ReadFromUser(system, &user_value, addr)) {
             slp.CancelSleep();
-            return Svc::ResultInvalidCurrentMemory;
+            return ResultInvalidCurrentMemory;
         }
 
         // Check that the value is equal.
         if (value != user_value) {
             slp.CancelSleep();
-            return Svc::ResultInvalidState;
+            return ResultInvalidState;
         }
 
         // Check that the timeout is non-zero.
         if (timeout == 0) {
             slp.CancelSleep();
-            return Svc::ResultTimedOut;
+            return ResultTimedOut;
         }
 
         // Set the arbiter.
-        cur_thread->SetAddressArbiter(std::addressof(thread_tree), addr);
+        cur_thread->SetAddressArbiter(&thread_tree, addr);
         thread_tree.insert(*cur_thread);
         cur_thread->SetState(ThreadState::Waiting);
         cur_thread->SetWaitReasonForDebugging(ThreadWaitReasonForDebugging::Arbitration);
     }
 
     // Cancel the timer wait.
-    if (timer != InvalidHandle) {
-        auto& time_manager = kernel.TimeManager();
-        time_manager.UnscheduleTimeEvent(timer);
-    }
+    kernel.TimeManager().UnscheduleTimeEvent(cur_thread);
 
     // Remove from the address arbiter.
     {
@@ -361,7 +335,7 @@ ResultCode KAddressArbiter::WaitIfEqual(VAddr addr, s32 value, s64 timeout) {
 
     // Get the result.
     KSynchronizationObject* dummy{};
-    return cur_thread->GetWaitResult(std::addressof(dummy));
+    return cur_thread->GetWaitResult(&dummy);
 }
 
 } // namespace Kernel

src/core/hle/kernel/k_affinity_mask.h

@@ -27,7 +27,7 @@ public:
     }
 
     [[nodiscard]] constexpr bool GetAffinity(s32 core) const {
-        return this->mask & GetCoreBit(core);
+        return (this->mask & GetCoreBit(core)) != 0;
     }
 
     constexpr void SetAffinity(s32 core, bool set) {

src/core/hle/kernel/k_condition_variable.cpp

@@ -10,11 +10,11 @@
 #include "core/hle/kernel/k_scheduler.h"
 #include "core/hle/kernel/k_scoped_scheduler_lock_and_sleep.h"
 #include "core/hle/kernel/k_synchronization_object.h"
+#include "core/hle/kernel/k_thread.h"
 #include "core/hle/kernel/kernel.h"
 #include "core/hle/kernel/process.h"
 #include "core/hle/kernel/svc_common.h"
 #include "core/hle/kernel/svc_results.h"
-#include "core/hle/kernel/thread.h"
 #include "core/memory.h"
 
 namespace Kernel {
@@ -66,7 +66,7 @@ KConditionVariable::KConditionVariable(Core::System& system_)
 KConditionVariable::~KConditionVariable() = default;
 
 ResultCode KConditionVariable::SignalToAddress(VAddr addr) {
-    Thread* owner_thread = kernel.CurrentScheduler()->GetCurrentThread();
+    KThread* owner_thread = kernel.CurrentScheduler()->GetCurrentThread();
 
     // Signal the address.
     {
@@ -74,7 +74,7 @@ ResultCode KConditionVariable::SignalToAddress(VAddr addr) {
 
         // Remove waiter thread.
         s32 num_waiters{};
-        Thread* next_owner_thread =
+        KThread* next_owner_thread =
             owner_thread->RemoveWaiterByKey(std::addressof(num_waiters), addr);
 
         // Determine the next tag.
@@ -92,10 +92,10 @@ ResultCode KConditionVariable::SignalToAddress(VAddr addr) {
         // Write the value to userspace.
         if (!WriteToUser(system, addr, std::addressof(next_value))) {
             if (next_owner_thread) {
-                next_owner_thread->SetSyncedObject(nullptr, Svc::ResultInvalidCurrentMemory);
+                next_owner_thread->SetSyncedObject(nullptr, ResultInvalidCurrentMemory);
             }
 
-            return Svc::ResultInvalidCurrentMemory;
+            return ResultInvalidCurrentMemory;
         }
     }
 
@@ -103,31 +103,31 @@ ResultCode KConditionVariable::SignalToAddress(VAddr addr) {
 }
 
 ResultCode KConditionVariable::WaitForAddress(Handle handle, VAddr addr, u32 value) {
-    Thread* cur_thread = kernel.CurrentScheduler()->GetCurrentThread();
+    KThread* cur_thread = kernel.CurrentScheduler()->GetCurrentThread();
 
     // Wait for the address.
     {
-        std::shared_ptr<Thread> owner_thread;
+        std::shared_ptr<KThread> owner_thread;
         ASSERT(!owner_thread);
         {
             KScopedSchedulerLock sl(kernel);
             cur_thread->SetSyncedObject(nullptr, RESULT_SUCCESS);
 
             // Check if the thread should terminate.
-            R_UNLESS(!cur_thread->IsTerminationRequested(), Svc::ResultTerminationRequested);
+            R_UNLESS(!cur_thread->IsTerminationRequested(), ResultTerminationRequested);
 
             {
                 // Read the tag from userspace.
                 u32 test_tag{};
                 R_UNLESS(ReadFromUser(system, std::addressof(test_tag), addr),
-                         Svc::ResultInvalidCurrentMemory);
+                         ResultInvalidCurrentMemory);
 
                 // If the tag isn't the handle (with wait mask), we're done.
                 R_UNLESS(test_tag == (handle | Svc::HandleWaitMask), RESULT_SUCCESS);
 
                 // Get the lock owner thread.
-                owner_thread = kernel.CurrentProcess()->GetHandleTable().Get<Thread>(handle);
-                R_UNLESS(owner_thread, Svc::ResultInvalidHandle);
+                owner_thread = kernel.CurrentProcess()->GetHandleTable().Get<KThread>(handle);
+                R_UNLESS(owner_thread, ResultInvalidHandle);
 
                 // Update the lock.
                 cur_thread->SetAddressKey(addr, value);
@@ -143,7 +143,7 @@ ResultCode KConditionVariable::WaitForAddress(Handle handle, VAddr addr, u32 val
     // Remove the thread as a waiter from the lock owner.
     {
         KScopedSchedulerLock sl(kernel);
-        Thread* owner_thread = cur_thread->GetLockOwner();
+        KThread* owner_thread = cur_thread->GetLockOwner();
         if (owner_thread != nullptr) {
             owner_thread->RemoveWaiter(cur_thread);
         }
@@ -154,7 +154,7 @@ ResultCode KConditionVariable::WaitForAddress(Handle handle, VAddr addr, u32 val
     return cur_thread->GetWaitResult(std::addressof(dummy));
 }
 
-Thread* KConditionVariable::SignalImpl(Thread* thread) {
+KThread* KConditionVariable::SignalImpl(KThread* thread) {
     // Check pre-conditions.
     ASSERT(kernel.GlobalSchedulerContext().IsLocked());
 
@@ -174,7 +174,7 @@ Thread* KConditionVariable::SignalImpl(Thread* thread) {
         }
     }
 
-    Thread* thread_to_close = nullptr;
+    KThread* thread_to_close = nullptr;
     if (can_access) {
         if (prev_tag == InvalidHandle) {
             // If nobody held the lock previously, we're all good.
@@ -182,7 +182,7 @@ Thread* KConditionVariable::SignalImpl(Thread* thread) {
             thread->Wakeup();
         } else {
             // Get the previous owner.
-            auto owner_thread = kernel.CurrentProcess()->GetHandleTable().Get<Thread>(
+            auto owner_thread = kernel.CurrentProcess()->GetHandleTable().Get<KThread>(
                 prev_tag & ~Svc::HandleWaitMask);
 
             if (owner_thread) {
@@ -191,13 +191,13 @@ Thread* KConditionVariable::SignalImpl(Thread* thread) {
                 thread_to_close = owner_thread.get();
             } else {
                 // The lock was tagged with a thread that doesn't exist.
-                thread->SetSyncedObject(nullptr, Svc::ResultInvalidState);
+                thread->SetSyncedObject(nullptr, ResultInvalidState);
                 thread->Wakeup();
             }
         }
     } else {
         // If the address wasn't accessible, note so.
-        thread->SetSyncedObject(nullptr, Svc::ResultInvalidCurrentMemory);
+        thread->SetSyncedObject(nullptr, ResultInvalidCurrentMemory);
         thread->Wakeup();
     }
 
@@ -210,8 +210,8 @@ void KConditionVariable::Signal(u64 cv_key, s32 count) {
 
     // TODO(bunnei): This should just be Thread once we implement KAutoObject instead of using
     // std::shared_ptr.
-    std::vector<std::shared_ptr<Thread>> thread_list;
-    std::array<Thread*, MaxThreads> thread_array;
+    std::vector<std::shared_ptr<KThread>> thread_list;
+    std::array<KThread*, MaxThreads> thread_array;
     s32 num_to_close{};
 
     // Perform signaling.
@@ -222,9 +222,9 @@ void KConditionVariable::Signal(u64 cv_key, s32 count) {
         auto it = thread_tree.nfind_light({cv_key, -1});
         while ((it != thread_tree.end()) && (count <= 0 || num_waiters < count) &&
                (it->GetConditionVariableKey() == cv_key)) {
-            Thread* target_thread = std::addressof(*it);
+            KThread* target_thread = std::addressof(*it);
 
-            if (Thread* thread = SignalImpl(target_thread); thread != nullptr) {
+            if (KThread* thread = SignalImpl(target_thread); thread != nullptr) {
                 if (num_to_close < MaxThreads) {
                     thread_array[num_to_close++] = thread;
                 } else {
@@ -257,26 +257,25 @@ void KConditionVariable::Signal(u64 cv_key, s32 count) {
 
 ResultCode KConditionVariable::Wait(VAddr addr, u64 key, u32 value, s64 timeout) {
     // Prepare to wait.
-    Thread* cur_thread = kernel.CurrentScheduler()->GetCurrentThread();
-    Handle timer = InvalidHandle;
+    KThread* cur_thread = kernel.CurrentScheduler()->GetCurrentThread();
 
     {
-        KScopedSchedulerLockAndSleep slp(kernel, timer, cur_thread, timeout);
+        KScopedSchedulerLockAndSleep slp{kernel, cur_thread, timeout};
 
         // Set the synced object.
-        cur_thread->SetSyncedObject(nullptr, Svc::ResultTimedOut);
+        cur_thread->SetSyncedObject(nullptr, ResultTimedOut);
 
         // Check that the thread isn't terminating.
         if (cur_thread->IsTerminationRequested()) {
             slp.CancelSleep();
-            return Svc::ResultTerminationRequested;
+            return ResultTerminationRequested;
         }
 
         // Update the value and process for the next owner.
         {
             // Remove waiter thread.
             s32 num_waiters{};
-            Thread* next_owner_thread =
+            KThread* next_owner_thread =
                 cur_thread->RemoveWaiterByKey(std::addressof(num_waiters), addr);
 
             // Update for the next owner thread.
@@ -303,7 +302,7 @@ ResultCode KConditionVariable::Wait(VAddr addr, u64 key, u32 value, s64 timeout)
             // Write the value to userspace.
             if (!WriteToUser(system, addr, std::addressof(next_value))) {
                 slp.CancelSleep();
-                return Svc::ResultInvalidCurrentMemory;
+                return ResultInvalidCurrentMemory;
             }
         }
 
@@ -322,16 +321,13 @@ ResultCode KConditionVariable::Wait(VAddr addr, u64 key, u32 value, s64 timeout)
     }
 
     // Cancel the timer wait.
-    if (timer != InvalidHandle) {
-        auto& time_manager = kernel.TimeManager();
-        time_manager.UnscheduleTimeEvent(timer);
-    }
+    kernel.TimeManager().UnscheduleTimeEvent(cur_thread);
 
     // Remove from the condition variable.
     {
         KScopedSchedulerLock sl(kernel);
 
-        if (Thread* owner = cur_thread->GetLockOwner(); owner != nullptr) {
+        if (KThread* owner = cur_thread->GetLockOwner(); owner != nullptr) {
             owner->RemoveWaiter(cur_thread);
         }
 

src/core/hle/kernel/k_condition_variable.h

@@ -8,8 +8,8 @@
 #include "common/common_types.h"
 
 #include "core/hle/kernel/k_scheduler.h"
+#include "core/hle/kernel/k_thread.h"
 #include "core/hle/kernel/kernel.h"
-#include "core/hle/kernel/thread.h"
 #include "core/hle/result.h"
 
 namespace Core {
@@ -20,7 +20,7 @@ namespace Kernel {
 
 class KConditionVariable {
 public:
-    using ThreadTree = typename Thread::ConditionVariableThreadTreeType;
+    using ThreadTree = typename KThread::ConditionVariableThreadTreeType;
 
     explicit KConditionVariable(Core::System& system_);
     ~KConditionVariable();
@@ -34,7 +34,7 @@ public:
     [[nodiscard]] ResultCode Wait(VAddr addr, u64 key, u32 value, s64 timeout);
 
 private:
-    [[nodiscard]] Thread* SignalImpl(Thread* thread);
+    [[nodiscard]] KThread* SignalImpl(KThread* thread);
 
     ThreadTree thread_tree;
 
@@ -43,14 +43,14 @@ private:
 };
 
 inline void BeforeUpdatePriority(const KernelCore& kernel, KConditionVariable::ThreadTree* tree,
-                                 Thread* thread) {
+                                 KThread* thread) {
     ASSERT(kernel.GlobalSchedulerContext().IsLocked());
 
     tree->erase(tree->iterator_to(*thread));
 }
 
 inline void AfterUpdatePriority(const KernelCore& kernel, KConditionVariable::ThreadTree* tree,
-                                Thread* thread) {
+                                KThread* thread) {
     ASSERT(kernel.GlobalSchedulerContext().IsLocked());
 
     tree->insert(*thread);

src/core/hle/kernel/k_event.cpp

@@ -0,0 +1,32 @@
+// Copyright 2021 yuzu emulator team
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#include "core/hle/kernel/k_event.h"
+#include "core/hle/kernel/k_readable_event.h"
+#include "core/hle/kernel/k_writable_event.h"
+
+namespace Kernel {
+
+KEvent::KEvent(KernelCore& kernel, std::string&& name) : Object{kernel, std::move(name)} {}
+
+KEvent::~KEvent() = default;
+
+std::shared_ptr<KEvent> KEvent::Create(KernelCore& kernel, std::string&& name) {
+    return std::make_shared<KEvent>(kernel, std::move(name));
+}
+
+void KEvent::Initialize() {
+    // Create our sub events.
+    readable_event = std::make_shared<KReadableEvent>(kernel, GetName() + ":Readable");
+    writable_event = std::make_shared<KWritableEvent>(kernel, GetName() + ":Writable");
+
+    // Initialize our sub sessions.
+    readable_event->Initialize(this);
+    writable_event->Initialize(this);
+
+    // Mark initialized.
+    initialized = true;
+}
+
+} // namespace Kernel