shader/arithmetic: Implement FCMP immediate + register variant

Trivially add the encoding for this.

.ci/yuzu-patreon-step2.yml

@@ -9,6 +9,7 @@ stages:
   displayName: 'build'
   jobs:
   - job: build
+    timeoutInMinutes: 120
     displayName: 'windows-msvc'
     pool:
       vmImage: windows-2019

.lgtm.yml

@@ -6,8 +6,5 @@ extraction:
       packages:
       - "libsdl2-dev"
       - "qtmultimedia5-dev"
-      - "clang-format-10"
       - "libtbb-dev"
       - "libjack-jackd2-dev"
-      - "doxygen"
-      - "graphviz"

CMakeLists.txt

@@ -210,7 +210,7 @@ if(ENABLE_QT)
     set(QT_PREFIX_HINT)
     if(YUZU_USE_BUNDLED_QT)
         if ((MSVC_VERSION GREATER_EQUAL 1910 AND MSVC_VERSION LESS 1930) AND ARCHITECTURE_x86_64)
-            set(QT_VER qt-5.12.0-msvc2017_64)
+            set(QT_VER qt-5.12.8-msvc2017_64)
         else()
             message(FATAL_ERROR "No bundled Qt binaries for your toolchain. Disable YUZU_USE_BUNDLED_QT and provide your own.")
         endif()
@@ -263,6 +263,7 @@ if (CONAN_REQUIRED_LIBS)
         libzip:with_openssl=False
         libzip:enable_windows_crypto=False
     )
+
     conan_check(VERSION 1.24.0 REQUIRED)
     # Add the bincrafters remote
     conan_add_remote(NAME bincrafters
@@ -354,6 +355,19 @@ 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)
+endif()
+
 # Prefer the -pthread flag on Linux.
 set(THREADS_PREFER_PTHREAD_FLAG ON)
 find_package(Threads REQUIRED)

CMakeModules/CopyYuzuFFmpegDeps.cmake

@@ -0,0 +1,10 @@
+function(copy_yuzu_FFmpeg_deps target_dir)
+    include(WindowsCopyFiles)
+    set(DLL_DEST "${CMAKE_BINARY_DIR}/bin/$<CONFIG>/")
+    windows_copy_files(${target_dir} ${FFMPEG_DLL_DIR} ${DLL_DEST}
+        avcodec-58.dll
+        avutil-56.dll
+        swresample-3.dll
+        swscale-5.dll
+    )
+endfunction(copy_yuzu_FFmpeg_deps)

dist/icons/controller/controller.qrc

@@ -21,5 +21,35 @@
     <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>	
+    <file alias="applet_handheld">applet_handheld.png</file>
+    <file alias="applet_handheld_dark">applet_handheld_dark.png</file>	
+    <file alias="applet_handheld_midnight">applet_handheld_midnight.png</file>
+    <file alias="applet_pro_controller">applet_pro_controller.png</file>
+    <file alias="applet_pro_controller_dark">applet_pro_controller_dark.png</file>
+    <file alias="applet_pro_controller_midnight">applet_pro_controller_midnight.png</file>
+    <file alias="applet_joycon_left">applet_single_joycon_left.png</file>
+    <file alias="applet_joycon_left_dark">applet_single_joycon_left_dark.png</file>
+    <file alias="applet_joycon_left_midnight">applet_single_joycon_left_midnight.png</file>
+    <file alias="applet_joycon_right">applet_single_joycon_right.png</file>
+    <file alias="applet_joycon_right_dark">applet_single_joycon_right_dark.png</file>
+    <file alias="applet_joycon_right_midnight">applet_single_joycon_right_midnight.png</file>
+    <file alias="applet_dual_joycon_disabled">applet_dual_joycon_disabled.png</file>
+    <file alias="applet_dual_joycon_dark_disabled">applet_dual_joycon_dark_disabled.png</file>
+    <file alias="applet_dual_joycon_midnight_disabled">applet_dual_joycon_midnight_disabled.png</file>
+    <file alias="applet_handheld_disabled">applet_handheld_disabled.png</file>
+    <file alias="applet_handheld_dark_disabled">applet_handheld_dark_disabled.png</file>
+    <file alias="applet_handheld_midnight_disabled">applet_handheld_midnight_disabled.png</file>
+    <file alias="applet_pro_controller_disabled">applet_pro_controller_disabled.png</file>
+    <file alias="applet_pro_controller_dark_disabled">applet_pro_controller_dark_disabled.png</file>
+    <file alias="applet_pro_controller_midnight_disabled">applet_pro_controller_midnight_disabled.png</file>
+    <file alias="applet_joycon_left_disabled">applet_single_joycon_left_disabled.png</file>
+    <file alias="applet_joycon_left_dark_disabled">applet_single_joycon_left_dark_disabled.png</file>
+    <file alias="applet_joycon_left_midnight_disabled">applet_single_joycon_left_midnight_disabled.png</file>
+    <file alias="applet_joycon_right_disabled">applet_single_joycon_right_disabled.png</file>
+    <file alias="applet_joycon_right_dark_disabled">applet_single_joycon_right_dark_disabled.png</file>
+    <file alias="applet_joycon_right_midnight_disabled">applet_single_joycon_right_midnight_disabled.png</file>
   </qresource>
 </RCC>

dist/qt_themes/default/style.qss

@@ -41,6 +41,99 @@ QPushButton#buttonRefreshDevices {
     max-height: 20px;
 }
 
+QWidget#bottomPerGameInput,
+QWidget#topControllerApplet,
+QWidget#bottomControllerApplet,
+QGroupBox#groupPlayer1Connected:checked,
+QGroupBox#groupPlayer2Connected:checked,
+QGroupBox#groupPlayer3Connected:checked,
+QGroupBox#groupPlayer4Connected:checked,
+QGroupBox#groupPlayer5Connected:checked,
+QGroupBox#groupPlayer6Connected:checked,
+QGroupBox#groupPlayer7Connected:checked,
+QGroupBox#groupPlayer8Connected:checked {
+    background-color: #f5f5f5;
+}
+
+QWidget#topControllerApplet {
+    border-bottom: 1px solid #828790
+}
+
+QWidget#bottomPerGameInput,
+QWidget#bottomControllerApplet {
+    border-top: 1px solid #828790
+}
+
+QWidget#topPerGameInput,
+QWidget#middleControllerApplet {
+    background-color: #fff;
+}
+
+QWidget#topPerGameInput QComboBox,
+QWidget#middleControllerApplet QComboBox {
+    width: 123px;
+}
+
+QWidget#connectedControllers {
+    background: transparent;
+}
+
+QWidget#playersSupported,
+QWidget#controllersSupported,
+QWidget#controllerSupported1,
+QWidget#controllerSupported2,
+QWidget#controllerSupported3,
+QWidget#controllerSupported4,
+QWidget#controllerSupported5,
+QWidget#controllerSupported6 {
+    border: none;
+    background: transparent;
+}
+
+QGroupBox#groupPlayer1Connected,
+QGroupBox#groupPlayer2Connected,
+QGroupBox#groupPlayer3Connected,
+QGroupBox#groupPlayer4Connected,
+QGroupBox#groupPlayer5Connected,
+QGroupBox#groupPlayer6Connected,
+QGroupBox#groupPlayer7Connected,
+QGroupBox#groupPlayer8Connected {
+    border: 1px solid #828790;
+    border-radius: 3px;
+    padding: 0px;
+    min-height: 98px;
+    max-height: 98px;
+}
+
+QGroupBox#groupPlayer1Connected:unchecked,
+QGroupBox#groupPlayer2Connected:unchecked,
+QGroupBox#groupPlayer3Connected:unchecked,
+QGroupBox#groupPlayer4Connected:unchecked,
+QGroupBox#groupPlayer5Connected:unchecked,
+QGroupBox#groupPlayer6Connected:unchecked,
+QGroupBox#groupPlayer7Connected:unchecked,
+QGroupBox#groupPlayer8Connected:unchecked {
+    border: 1px solid #d9d9d9;
+}
+
+QGroupBox#groupPlayer1Connected::title,
+QGroupBox#groupPlayer2Connected::title,
+QGroupBox#groupPlayer3Connected::title,
+QGroupBox#groupPlayer4Connected::title,
+QGroupBox#groupPlayer5Connected::title,
+QGroupBox#groupPlayer6Connected::title,
+QGroupBox#groupPlayer7Connected::title,
+QGroupBox#groupPlayer8Connected::title {
+    subcontrol-origin: margin;
+    subcontrol-position: top left;
+    padding-left: 0px;
+    padding-right: 0px;
+    padding-top: 1px;
+    margin-left: 0px;
+    margin-right: -4px;
+    margin-bottom: 4px;
+}
+
 QCheckBox#checkboxPlayer1Connected,
 QCheckBox#checkboxPlayer2Connected,
 QCheckBox#checkboxPlayer3Connected,
@@ -52,6 +145,42 @@ QCheckBox#checkboxPlayer8Connected {
     spacing: 0px;
 }
 
+QWidget#Player1LEDs QCheckBox,
+QWidget#Player2LEDs QCheckBox,
+QWidget#Player3LEDs QCheckBox,
+QWidget#Player4LEDs QCheckBox,
+QWidget#Player5LEDs QCheckBox,
+QWidget#Player6LEDs QCheckBox,
+QWidget#Player7LEDs QCheckBox,
+QWidget#Player8LEDs QCheckBox {
+    spacing: 0px;
+}
+
+QWidget#Player1LEDs QCheckBox::indicator,
+QWidget#Player2LEDs QCheckBox::indicator,
+QWidget#Player3LEDs QCheckBox::indicator,
+QWidget#Player4LEDs QCheckBox::indicator,
+QWidget#Player5LEDs QCheckBox::indicator,
+QWidget#Player6LEDs QCheckBox::indicator,
+QWidget#Player7LEDs QCheckBox::indicator,
+QWidget#Player8LEDs QCheckBox::indicator {
+    width: 6px;
+    height: 6px;
+    margin-left: 0px;
+}
+
+QWidget#bottomPerGameInput QCheckBox#checkboxPlayer1Connected::indicator,
+QWidget#bottomPerGameInput QCheckBox#checkboxPlayer2Connected::indicator,
+QWidget#bottomPerGameInput QCheckBox#checkboxPlayer3Connected::indicator,
+QWidget#bottomPerGameInput QCheckBox#checkboxPlayer4Connected::indicator,
+QWidget#bottomPerGameInput QCheckBox#checkboxPlayer5Connected::indicator,
+QWidget#bottomPerGameInput QCheckBox#checkboxPlayer6Connected::indicator,
+QWidget#bottomPerGameInput QCheckBox#checkboxPlayer7Connected::indicator,
+QWidget#bottomPerGameInput QCheckBox#checkboxPlayer8Connected::indicator {
+    width: 12px;
+    height: 12px;
+}
+
 QCheckBox#checkboxPlayer1Connected::indicator,
 QCheckBox#checkboxPlayer2Connected::indicator,
 QCheckBox#checkboxPlayer3Connected::indicator,
@@ -64,6 +193,34 @@ QCheckBox#checkboxPlayer8Connected::indicator {
     height: 14px;
 }
 
+QGroupBox#groupPlayer1Connected::indicator,
+QGroupBox#groupPlayer2Connected::indicator,
+QGroupBox#groupPlayer3Connected::indicator,
+QGroupBox#groupPlayer4Connected::indicator,
+QGroupBox#groupPlayer5Connected::indicator,
+QGroupBox#groupPlayer6Connected::indicator,
+QGroupBox#groupPlayer7Connected::indicator,
+QGroupBox#groupPlayer8Connected::indicator {
+    width: 16px;
+    height: 16px;
+}
+
+QWidget#Player1LEDs QCheckBox::indicator:checked,
+QWidget#Player2LEDs QCheckBox::indicator:checked,
+QWidget#Player3LEDs QCheckBox::indicator:checked,
+QWidget#Player4LEDs QCheckBox::indicator:checked,
+QWidget#Player5LEDs QCheckBox::indicator:checked,
+QWidget#Player6LEDs QCheckBox::indicator:checked,
+QWidget#Player7LEDs QCheckBox::indicator:checked,
+QWidget#Player8LEDs QCheckBox::indicator:checked,
+QGroupBox#groupPlayer1Connected::indicator:checked,
+QGroupBox#groupPlayer2Connected::indicator:checked,
+QGroupBox#groupPlayer3Connected::indicator:checked,
+QGroupBox#groupPlayer4Connected::indicator:checked,
+QGroupBox#groupPlayer5Connected::indicator:checked,
+QGroupBox#groupPlayer6Connected::indicator:checked,
+QGroupBox#groupPlayer7Connected::indicator:checked,
+QGroupBox#groupPlayer8Connected::indicator:checked,
 QCheckBox#checkboxPlayer1Connected::indicator:checked,
 QCheckBox#checkboxPlayer2Connected::indicator:checked,
 QCheckBox#checkboxPlayer3Connected::indicator:checked,
@@ -74,11 +231,27 @@ QCheckBox#checkboxPlayer7Connected::indicator:checked,
 QCheckBox#checkboxPlayer8Connected::indicator:checked,
 QGroupBox#groupConnectedController::indicator:checked {
     border-radius: 2px;
-    border: 1px solid black;
+    border: 1px solid #929192;
     background: #39ff14;
     image: none;
 }
 
+QWidget#Player1LEDs QCheckBox::indicator:unchecked,
+QWidget#Player2LEDs QCheckBox::indicator:unchecked,
+QWidget#Player3LEDs QCheckBox::indicator:unchecked,
+QWidget#Player4LEDs QCheckBox::indicator:unchecked,
+QWidget#Player5LEDs QCheckBox::indicator:unchecked,
+QWidget#Player6LEDs QCheckBox::indicator:unchecked,
+QWidget#Player7LEDs QCheckBox::indicator:unchecked,
+QWidget#Player8LEDs QCheckBox::indicator:unchecked,
+QGroupBox#groupPlayer1Connected::indicator:unchecked,
+QGroupBox#groupPlayer2Connected::indicator:unchecked,
+QGroupBox#groupPlayer3Connected::indicator:unchecked,
+QGroupBox#groupPlayer4Connected::indicator:unchecked,
+QGroupBox#groupPlayer5Connected::indicator:unchecked,
+QGroupBox#groupPlayer6Connected::indicator:unchecked,
+QGroupBox#groupPlayer7Connected::indicator:unchecked,
+QGroupBox#groupPlayer8Connected::indicator:unchecked,
 QCheckBox#checkboxPlayer1Connected::indicator:unchecked,
 QCheckBox#checkboxPlayer2Connected::indicator:unchecked,
 QCheckBox#checkboxPlayer3Connected::indicator:unchecked,
@@ -89,7 +262,18 @@ QCheckBox#checkboxPlayer7Connected::indicator:unchecked,
 QCheckBox#checkboxPlayer8Connected::indicator:unchecked,
 QGroupBox#groupConnectedController::indicator:unchecked {
     border-radius: 2px;
-    border: 1px solid black;
+    border: 1px solid #929192;
     background: transparent;
     image: none;
 }
+
+QWidget#controllerPlayer1,
+QWidget#controllerPlayer2,
+QWidget#controllerPlayer3,
+QWidget#controllerPlayer4,
+QWidget#controllerPlayer5,
+QWidget#controllerPlayer6,
+QWidget#controllerPlayer7,
+QWidget#controllerPlayer8 {
+    background: transparent;
+}

dist/qt_themes/qdarkstyle/style.qrc

@@ -52,6 +52,6 @@
     <file>rc/radio_unchecked.png</file>
   </qresource>
   <qresource prefix="qdarkstyle">
-      <file>style.qss</file>
+    <file>style.qss</file>
   </qresource>
 </RCC>

dist/qt_themes/qdarkstyle/style.qss

@@ -1284,59 +1284,6 @@ QPushButton#buttonRefreshDevices {
     padding: 0px 0px;
 }
 
-QCheckBox#checkboxPlayer1Connected,
-QCheckBox#checkboxPlayer2Connected,
-QCheckBox#checkboxPlayer3Connected,
-QCheckBox#checkboxPlayer4Connected,
-QCheckBox#checkboxPlayer5Connected,
-QCheckBox#checkboxPlayer6Connected,
-QCheckBox#checkboxPlayer7Connected,
-QCheckBox#checkboxPlayer8Connected {
-    spacing: 0px;
-}
-
-QCheckBox#checkboxPlayer1Connected::indicator,
-QCheckBox#checkboxPlayer2Connected::indicator,
-QCheckBox#checkboxPlayer3Connected::indicator,
-QCheckBox#checkboxPlayer4Connected::indicator,
-QCheckBox#checkboxPlayer5Connected::indicator,
-QCheckBox#checkboxPlayer6Connected::indicator,
-QCheckBox#checkboxPlayer7Connected::indicator,
-QCheckBox#checkboxPlayer8Connected::indicator {
-    width: 14px;
-    height: 14px;
-}
-
-QCheckBox#checkboxPlayer1Connected::indicator:checked,
-QCheckBox#checkboxPlayer2Connected::indicator:checked,
-QCheckBox#checkboxPlayer3Connected::indicator:checked,
-QCheckBox#checkboxPlayer4Connected::indicator:checked,
-QCheckBox#checkboxPlayer5Connected::indicator:checked,
-QCheckBox#checkboxPlayer6Connected::indicator:checked,
-QCheckBox#checkboxPlayer7Connected::indicator:checked,
-QCheckBox#checkboxPlayer8Connected::indicator:checked,
-QGroupBox#groupConnectedController::indicator:checked {
-    border-radius: 2px;
-    border: 1px solid #929192;
-    background: #39ff14;
-    image: none;
-}
-
-QCheckBox#checkboxPlayer1Connected::indicator:unchecked,
-QCheckBox#checkboxPlayer2Connected::indicator:unchecked,
-QCheckBox#checkboxPlayer3Connected::indicator:unchecked,
-QCheckBox#checkboxPlayer4Connected::indicator:unchecked,
-QCheckBox#checkboxPlayer5Connected::indicator:unchecked,
-QCheckBox#checkboxPlayer6Connected::indicator:unchecked,
-QCheckBox#checkboxPlayer7Connected::indicator:unchecked,
-QCheckBox#checkboxPlayer8Connected::indicator:unchecked,
-QGroupBox#groupConnectedController::indicator:unchecked {
-    border-radius: 2px;
-    border: 1px solid #929192;
-    background: transparent;
-    image: none;
-}
-
 QSpinBox#spinboxLStickRange,
 QSpinBox#spinboxRStickRange {
     padding: 4px 0px 5px 0px;
@@ -1367,9 +1314,260 @@ QGroupBox#vibrationGroup::indicator {
 
 QGroupBox#motionGroup::title,
 QGroupBox#vibrationGroup::title {
-	spacing: 2px;
-	padding-left: 1px;
-	padding-right: 1px;
+    spacing: 2px;
+    padding-left: 1px;
+    padding-right: 1px;
+}
+
+QWidget#bottomPerGameInput,
+QWidget#topControllerApplet,
+QWidget#bottomControllerApplet,
+QGroupBox#groupPlayer1Connected:checked,
+QGroupBox#groupPlayer2Connected:checked,
+QGroupBox#groupPlayer3Connected:checked,
+QGroupBox#groupPlayer4Connected:checked,
+QGroupBox#groupPlayer5Connected:checked,
+QGroupBox#groupPlayer6Connected:checked,
+QGroupBox#groupPlayer7Connected:checked,
+QGroupBox#groupPlayer8Connected:checked {
+    background-color: #232629;
+}
+
+QWidget#topPerGameInput,
+QWidget#middleControllerApplet {
+    background-color: #31363b;
+}
+
+QWidget#topPerGameInput QComboBox,
+QWidget#middleControllerApplet QComboBox {
+    width: 119px;
+}
+
+QRadioButton#radioDocked {
+    margin-left: -3px;
+}
+
+
+QRadioButton#radioUndocked {
+    margin-right: 5px;
+}
+
+QWidget#connectedControllers {
+    background: transparent;
+}
+
+QWidget#playersSupported,
+QWidget#controllersSupported,
+QWidget#controllerSupported1,
+QWidget#controllerSupported2,
+QWidget#controllerSupported3,
+QWidget#controllerSupported4,
+QWidget#controllerSupported5,
+QWidget#controllerSupported6 {
+    border: none;
+    background: transparent;
+}
+
+QGroupBox#groupPlayer1Connected,
+QGroupBox#groupPlayer2Connected,
+QGroupBox#groupPlayer3Connected,
+QGroupBox#groupPlayer4Connected,
+QGroupBox#groupPlayer5Connected,
+QGroupBox#groupPlayer6Connected,
+QGroupBox#groupPlayer7Connected,
+QGroupBox#groupPlayer8Connected {
+    border: 1px solid #76797c;
+    border-radius: 3px;
+    padding: 0px;
+    min-height: 98px;
+    max-height: 98px;
+    margin-top: 0px;
+}
+
+QGroupBox#groupPlayer1Connected:unchecked,
+QGroupBox#groupPlayer2Connected:unchecked,
+QGroupBox#groupPlayer3Connected:unchecked,
+QGroupBox#groupPlayer4Connected:unchecked,
+QGroupBox#groupPlayer5Connected:unchecked,
+QGroupBox#groupPlayer6Connected:unchecked,
+QGroupBox#groupPlayer7Connected:unchecked,
+QGroupBox#groupPlayer8Connected:unchecked {
+    border: 1px solid #54575b;
+}
+
+QGroupBox#groupPlayer1Connected::title,
+QGroupBox#groupPlayer2Connected::title,
+QGroupBox#groupPlayer3Connected::title,
+QGroupBox#groupPlayer4Connected::title,
+QGroupBox#groupPlayer5Connected::title,
+QGroupBox#groupPlayer6Connected::title,
+QGroupBox#groupPlayer7Connected::title,
+QGroupBox#groupPlayer8Connected::title {
+    subcontrol-origin: margin;
+    subcontrol-position: top left;
+    padding-left: 0px;
+    padding-right: 0px;
+    padding-top: 1px;
+    margin-left: -2px;
+    margin-right: -4px;
+    margin-bottom: 6px;
+}
+
+QCheckBox#checkboxPlayer1Connected,
+QCheckBox#checkboxPlayer2Connected,
+QCheckBox#checkboxPlayer3Connected,
+QCheckBox#checkboxPlayer4Connected,
+QCheckBox#checkboxPlayer5Connected,
+QCheckBox#checkboxPlayer6Connected,
+QCheckBox#checkboxPlayer7Connected,
+QCheckBox#checkboxPlayer8Connected {
+    spacing: 0px;
+}
+
+QWidget#Player1LEDs,
+QWidget#Player2LEDs,
+QWidget#Player3LEDs,
+QWidget#Player4LEDs,
+QWidget#Player5LEDs,
+QWidget#Player6LEDs,
+QWidget#Player7LEDs,
+QWidget#Player8LEDs {
+    background: transparent;
+}
+
+QWidget#Player1LEDs QCheckBox,
+QWidget#Player2LEDs QCheckBox,
+QWidget#Player3LEDs QCheckBox,
+QWidget#Player4LEDs QCheckBox,
+QWidget#Player5LEDs QCheckBox,
+QWidget#Player6LEDs QCheckBox,
+QWidget#Player7LEDs QCheckBox,
+QWidget#Player8LEDs QCheckBox {
+    spacing: 0px;
+    margin-bottom: 0px;
+    margin-right: 0px;
+}
+
+QWidget#Player1LEDs QCheckBox::indicator,
+QWidget#Player2LEDs QCheckBox::indicator,
+QWidget#Player3LEDs QCheckBox::indicator,
+QWidget#Player4LEDs QCheckBox::indicator,
+QWidget#Player5LEDs QCheckBox::indicator,
+QWidget#Player6LEDs QCheckBox::indicator,
+QWidget#Player7LEDs QCheckBox::indicator,
+QWidget#Player8LEDs QCheckBox::indicator {
+    width: 6px;
+    height: 6px;
+    margin-left: 0px;
+}
+
+QWidget#bottomPerGameInput QCheckBox#checkboxPlayer1Connected::indicator,
+QWidget#bottomPerGameInput QCheckBox#checkboxPlayer2Connected::indicator,
+QWidget#bottomPerGameInput QCheckBox#checkboxPlayer3Connected::indicator,
+QWidget#bottomPerGameInput QCheckBox#checkboxPlayer4Connected::indicator,
+QWidget#bottomPerGameInput QCheckBox#checkboxPlayer5Connected::indicator,
+QWidget#bottomPerGameInput QCheckBox#checkboxPlayer6Connected::indicator,
+QWidget#bottomPerGameInput QCheckBox#checkboxPlayer7Connected::indicator,
+QWidget#bottomPerGameInput QCheckBox#checkboxPlayer8Connected::indicator {
+    width: 12px;
+    height: 12px;
+}
+
+QCheckBox#checkboxPlayer1Connected::indicator,
+QCheckBox#checkboxPlayer2Connected::indicator,
+QCheckBox#checkboxPlayer3Connected::indicator,
+QCheckBox#checkboxPlayer4Connected::indicator,
+QCheckBox#checkboxPlayer5Connected::indicator,
+QCheckBox#checkboxPlayer6Connected::indicator,
+QCheckBox#checkboxPlayer7Connected::indicator,
+QCheckBox#checkboxPlayer8Connected::indicator {
+    width: 14px;
+    height: 14px;
+}
+
+QGroupBox#groupPlayer1Connected::indicator,
+QGroupBox#groupPlayer2Connected::indicator,
+QGroupBox#groupPlayer3Connected::indicator,
+QGroupBox#groupPlayer4Connected::indicator,
+QGroupBox#groupPlayer5Connected::indicator,
+QGroupBox#groupPlayer6Connected::indicator,
+QGroupBox#groupPlayer7Connected::indicator,
+QGroupBox#groupPlayer8Connected::indicator {
+    width: 16px;
+    height: 16px;
+}
+
+QWidget#Player1LEDs QCheckBox::indicator:checked,
+QWidget#Player2LEDs QCheckBox::indicator:checked,
+QWidget#Player3LEDs QCheckBox::indicator:checked,
+QWidget#Player4LEDs QCheckBox::indicator:checked,
+QWidget#Player5LEDs QCheckBox::indicator:checked,
+QWidget#Player6LEDs QCheckBox::indicator:checked,
+QWidget#Player7LEDs QCheckBox::indicator:checked,
+QWidget#Player8LEDs QCheckBox::indicator:checked,
+QGroupBox#groupPlayer1Connected::indicator:checked,
+QGroupBox#groupPlayer2Connected::indicator:checked,
+QGroupBox#groupPlayer3Connected::indicator:checked,
+QGroupBox#groupPlayer4Connected::indicator:checked,
+QGroupBox#groupPlayer5Connected::indicator:checked,
+QGroupBox#groupPlayer6Connected::indicator:checked,
+QGroupBox#groupPlayer7Connected::indicator:checked,
+QGroupBox#groupPlayer8Connected::indicator:checked,
+QCheckBox#checkboxPlayer1Connected::indicator:checked,
+QCheckBox#checkboxPlayer2Connected::indicator:checked,
+QCheckBox#checkboxPlayer3Connected::indicator:checked,
+QCheckBox#checkboxPlayer4Connected::indicator:checked,
+QCheckBox#checkboxPlayer5Connected::indicator:checked,
+QCheckBox#checkboxPlayer6Connected::indicator:checked,
+QCheckBox#checkboxPlayer7Connected::indicator:checked,
+QCheckBox#checkboxPlayer8Connected::indicator:checked,
+QGroupBox#groupConnectedController::indicator:checked {
+    border-radius: 2px;
+    border: 1px solid #929192;
+    background: #39ff14;
+    image: none;
+}
+
+QWidget#Player1LEDs QCheckBox::indicator:unchecked,
+QWidget#Player2LEDs QCheckBox::indicator:unchecked,
+QWidget#Player3LEDs QCheckBox::indicator:unchecked,
+QWidget#Player4LEDs QCheckBox::indicator:unchecked,
+QWidget#Player5LEDs QCheckBox::indicator:unchecked,
+QWidget#Player6LEDs QCheckBox::indicator:unchecked,
+QWidget#Player7LEDs QCheckBox::indicator:unchecked,
+QWidget#Player8LEDs QCheckBox::indicator:unchecked,
+QGroupBox#groupPlayer1Connected::indicator:unchecked,
+QGroupBox#groupPlayer2Connected::indicator:unchecked,
+QGroupBox#groupPlayer3Connected::indicator:unchecked,
+QGroupBox#groupPlayer4Connected::indicator:unchecked,
+QGroupBox#groupPlayer5Connected::indicator:unchecked,
+QGroupBox#groupPlayer6Connected::indicator:unchecked,
+QGroupBox#groupPlayer7Connected::indicator:unchecked,
+QGroupBox#groupPlayer8Connected::indicator:unchecked,
+QCheckBox#checkboxPlayer1Connected::indicator:unchecked,
+QCheckBox#checkboxPlayer2Connected::indicator:unchecked,
+QCheckBox#checkboxPlayer3Connected::indicator:unchecked,
+QCheckBox#checkboxPlayer4Connected::indicator:unchecked,
+QCheckBox#checkboxPlayer5Connected::indicator:unchecked,
+QCheckBox#checkboxPlayer6Connected::indicator:unchecked,
+QCheckBox#checkboxPlayer7Connected::indicator:unchecked,
+QCheckBox#checkboxPlayer8Connected::indicator:unchecked,
+QGroupBox#groupConnectedController::indicator:unchecked {
+    border-radius: 2px;
+    border: 1px solid #929192;
+    background: transparent;
+    image: none;
+}
+
+QWidget#controllerPlayer1,
+QWidget#controllerPlayer2,
+QWidget#controllerPlayer3,
+QWidget#controllerPlayer4,
+QWidget#controllerPlayer5,
+QWidget#controllerPlayer6,
+QWidget#controllerPlayer7,
+QWidget#controllerPlayer8 {
+    background: transparent;
 }
 
 /* touchscreen mapping widget */

dist/qt_themes/qdarkstyle_midnight_blue/style.qrc

@@ -221,6 +221,6 @@
     <file>rc/window_undock_pressed@2x.png</file>
   </qresource>
   <qresource prefix="qdarkstyle_midnight_blue">
-      <file>style.qss</file>
+    <file>style.qss</file>
   </qresource>
 </RCC>

dist/qt_themes/qdarkstyle_midnight_blue/style.qss

@@ -235,19 +235,19 @@ https://doc.qt.io/qt-5/stylesheet-examples.html#customizing-qgroupbox
 
 --------------------------------------------------------------------------- */
 QGroupBox {
-    font-weight: bold;
-    border: 1px solid #32414B;
-    border-radius: 4px;
-    margin-top: 12px;
-    padding: 4px;
+  font-weight: bold;
+  border: 1px solid #32414B;
+  border-radius: 4px;
+  margin-top: 12px;
+  padding: 4px;
 }
 
 QGroupBox::title {
-    subcontrol-origin: margin;
-    subcontrol-position: top left;
-    padding-left: 3px;
-    padding-right: 5px;
-    padding-top: 4px;
+  subcontrol-origin: margin;
+  subcontrol-position: top left;
+  padding-left: 3px;
+  padding-right: 5px;
+  padding-top: 4px;
 }
 
 QGroupBox::indicator {
@@ -2205,7 +2205,179 @@ QPushButton#buttonRefreshDevices {
   padding: 0px 0px;
 }
 
+QSpinBox#spinboxLStickRange,
+QSpinBox#spinboxRStickRange {
+  min-width: 38px;
+}
+
+QGroupBox#motionGroup::indicator,
+QGroupBox#vibrationGroup::indicator {
+  margin-left: 0px;
+}
+
+QWidget#bottomPerGameInput QGroupBox#motionGroup,
+QWidget#bottomPerGameInput QGroupBox#vibrationGroup,
+QWidget#bottomPerGameInput QGroupBox#inputConfigGroup {
+  padding: 0px;
+}
+
+QGroupBox#motionGroup::title,
+QGroupBox#vibrationGroup::title {
+  spacing: 2px;
+  padding-left: 1px;
+  padding-right: 1px;
+}
+
+QListWidget#selectorList {
+  background-color: #0f1922;
+}
+
+QSpinBox,
+QLineEdit,
+QTreeView#hotkey_list,
+QScrollArea#scrollArea QTreeView {
+  background-color: #0f1922;
+}
+
+QWidget#bottomPerGameInput,
+QWidget#topControllerApplet,
+QWidget#bottomControllerApplet,
+QGroupBox#groupPlayer1Connected:checked,
+QGroupBox#groupPlayer2Connected:checked,
+QGroupBox#groupPlayer3Connected:checked,
+QGroupBox#groupPlayer4Connected:checked,
+QGroupBox#groupPlayer5Connected:checked,
+QGroupBox#groupPlayer6Connected:checked,
+QGroupBox#groupPlayer7Connected:checked,
+QGroupBox#groupPlayer8Connected:checked {
+  background-color: #0f1922;
+}
+
+QWidget#topPerGameInput,
+QWidget#middleControllerApplet {
+  background-color: #19232d;
+}
+
+QWidget#topPerGameInput QComboBox,
+QWidget#middleControllerApplet QComboBox {
+  padding-right: 2px;
+  width: 127px;
+}
+
+QGroupBox#handheldGroup {
+  padding-left: 0px;
+}
+
+QRadioButton#radioDocked {
+  margin-left: -1px;
+  padding-left: 0px;
+}
+
+QRadioButton#radioDocked::indicator {
+  margin-left: 0px;
+}
+
+
+QRadioButton#radioUndocked {
+  margin-right: 2px;
+}
+
+QWidget#connectedControllers {
+  background: transparent;
+}
+
+QWidget#playersSupported,
+QWidget#controllersSupported,
+QWidget#controllerSupported1,
+QWidget#controllerSupported2,
+QWidget#controllerSupported3,
+QWidget#controllerSupported4,
+QWidget#controllerSupported5,
+QWidget#controllerSupported6 {
+  border: none;
+  background: transparent;
+}
+
+QGroupBox#groupPlayer1Connected,
+QGroupBox#groupPlayer2Connected,
+QGroupBox#groupPlayer3Connected,
+QGroupBox#groupPlayer4Connected,
+QGroupBox#groupPlayer5Connected,
+QGroupBox#groupPlayer6Connected,
+QGroupBox#groupPlayer7Connected,
+QGroupBox#groupPlayer8Connected {
+  border: 1px solid #76797c;
+  border-radius: 3px;
+  padding: 0px;
+  min-height: 98px;
+  max-height: 98px;
+  margin-top: 0px;
+}
+
+
+QGroupBox#groupPlayer1Connected:unchecked,
+QGroupBox#groupPlayer2Connected:unchecked,
+QGroupBox#groupPlayer3Connected:unchecked,
+QGroupBox#groupPlayer4Connected:unchecked,
+QGroupBox#groupPlayer5Connected:unchecked,
+QGroupBox#groupPlayer6Connected:unchecked,
+QGroupBox#groupPlayer7Connected:unchecked,
+QGroupBox#groupPlayer8Connected:unchecked {
+  border: 1px solid #32414b;
+}
+
+QGroupBox#groupPlayer1Connected::title,
+QGroupBox#groupPlayer2Connected::title,
+QGroupBox#groupPlayer3Connected::title,
+QGroupBox#groupPlayer4Connected::title,
+QGroupBox#groupPlayer5Connected::title,
+QGroupBox#groupPlayer6Connected::title,
+QGroupBox#groupPlayer7Connected::title,
+QGroupBox#groupPlayer8Connected::title {
+  subcontrol-origin: margin;
+  subcontrol-position: top left;
+  padding-left: 0px;
+  padding-right: 0px;
+  padding-top: 1px;
+  margin-left: -2px;
+  margin-right: -4px;
+  margin-bottom: 6px;
+}
 
+QCheckBox#checkboxPlayer1Connected,
+QCheckBox#checkboxPlayer2Connected,
+QCheckBox#checkboxPlayer3Connected,
+QCheckBox#checkboxPlayer4Connected,
+QCheckBox#checkboxPlayer5Connected,
+QCheckBox#checkboxPlayer6Connected,
+QCheckBox#checkboxPlayer7Connected,
+QCheckBox#checkboxPlayer8Connected {
+    spacing: 0px;
+}
+
+QWidget#connectedControllers QLabel {
+  padding: 0px;
+}
+
+QWidget#Player1LEDs,
+QWidget#Player2LEDs,
+QWidget#Player3LEDs,
+QWidget#Player4LEDs,
+QWidget#Player5LEDs,
+QWidget#Player6LEDs,
+QWidget#Player7LEDs,
+QWidget#Player8LEDs {
+  background: transparent;
+}
+
+QWidget#Player1LEDs QCheckBox,
+QWidget#Player2LEDs QCheckBox,
+QWidget#Player3LEDs QCheckBox,
+QWidget#Player4LEDs QCheckBox,
+QWidget#Player5LEDs QCheckBox,
+QWidget#Player6LEDs QCheckBox,
+QWidget#Player7LEDs QCheckBox,
+QWidget#Player8LEDs QCheckBox,
 QCheckBox#checkboxPlayer1Connected,
 QCheckBox#checkboxPlayer2Connected,
 QCheckBox#checkboxPlayer3Connected,
@@ -2215,6 +2387,34 @@ QCheckBox#checkboxPlayer6Connected,
 QCheckBox#checkboxPlayer7Connected,
 QCheckBox#checkboxPlayer8Connected {
   spacing: 0px;
+  padding-top: 0px;
+  padding-bottom: 0px;
+  background: transparent;
+}
+
+QWidget#Player1LEDs QCheckBox::indicator,
+QWidget#Player2LEDs QCheckBox::indicator,
+QWidget#Player3LEDs QCheckBox::indicator,
+QWidget#Player4LEDs QCheckBox::indicator,
+QWidget#Player5LEDs QCheckBox::indicator,
+QWidget#Player6LEDs QCheckBox::indicator,
+QWidget#Player7LEDs QCheckBox::indicator,
+QWidget#Player8LEDs QCheckBox::indicator {
+  width: 6px;
+  height: 6px;
+  margin-left: 0px;
+}
+
+QWidget#bottomPerGameInput QCheckBox#checkboxPlayer1Connected::indicator,
+QWidget#bottomPerGameInput QCheckBox#checkboxPlayer2Connected::indicator,
+QWidget#bottomPerGameInput QCheckBox#checkboxPlayer3Connected::indicator,
+QWidget#bottomPerGameInput QCheckBox#checkboxPlayer4Connected::indicator,
+QWidget#bottomPerGameInput QCheckBox#checkboxPlayer5Connected::indicator,
+QWidget#bottomPerGameInput QCheckBox#checkboxPlayer6Connected::indicator,
+QWidget#bottomPerGameInput QCheckBox#checkboxPlayer7Connected::indicator,
+QWidget#bottomPerGameInput QCheckBox#checkboxPlayer8Connected::indicator {
+  width: 12px;
+  height: 12px;
 }
 
 QCheckBox#checkboxPlayer1Connected::indicator,
@@ -2227,8 +2427,25 @@ QCheckBox#checkboxPlayer7Connected::indicator,
 QCheckBox#checkboxPlayer8Connected::indicator {
   width: 14px;
   height: 14px;
+  margin-left: 2px;
 }
 
+QWidget#Player1LEDs QCheckBox::indicator:checked,
+QWidget#Player2LEDs QCheckBox::indicator:checked,
+QWidget#Player3LEDs QCheckBox::indicator:checked,
+QWidget#Player4LEDs QCheckBox::indicator:checked,
+QWidget#Player5LEDs QCheckBox::indicator:checked,
+QWidget#Player6LEDs QCheckBox::indicator:checked,
+QWidget#Player7LEDs QCheckBox::indicator:checked,
+QWidget#Player8LEDs QCheckBox::indicator:checked,
+QGroupBox#groupPlayer1Connected::indicator:checked,
+QGroupBox#groupPlayer2Connected::indicator:checked,
+QGroupBox#groupPlayer3Connected::indicator:checked,
+QGroupBox#groupPlayer4Connected::indicator:checked,
+QGroupBox#groupPlayer5Connected::indicator:checked,
+QGroupBox#groupPlayer6Connected::indicator:checked,
+QGroupBox#groupPlayer7Connected::indicator:checked,
+QGroupBox#groupPlayer8Connected::indicator:checked,
 QCheckBox#checkboxPlayer1Connected::indicator:checked,
 QCheckBox#checkboxPlayer2Connected::indicator:checked,
 QCheckBox#checkboxPlayer3Connected::indicator:checked,
@@ -2244,6 +2461,22 @@ QGroupBox#groupConnectedController::indicator:checked {
   image: none;
 }
 
+QWidget#Player1LEDs QCheckBox::indicator:unchecked,
+QWidget#Player2LEDs QCheckBox::indicator:unchecked,
+QWidget#Player3LEDs QCheckBox::indicator:unchecked,
+QWidget#Player4LEDs QCheckBox::indicator:unchecked,
+QWidget#Player5LEDs QCheckBox::indicator:unchecked,
+QWidget#Player6LEDs QCheckBox::indicator:unchecked,
+QWidget#Player7LEDs QCheckBox::indicator:unchecked,
+QWidget#Player8LEDs QCheckBox::indicator:unchecked,
+QGroupBox#groupPlayer1Connected::indicator:unchecked,
+QGroupBox#groupPlayer2Connected::indicator:unchecked,
+QGroupBox#groupPlayer3Connected::indicator:unchecked,
+QGroupBox#groupPlayer4Connected::indicator:unchecked,
+QGroupBox#groupPlayer5Connected::indicator:unchecked,
+QGroupBox#groupPlayer6Connected::indicator:unchecked,
+QGroupBox#groupPlayer7Connected::indicator:unchecked,
+QGroupBox#groupPlayer8Connected::indicator:unchecked,
 QCheckBox#checkboxPlayer1Connected::indicator:unchecked,
 QCheckBox#checkboxPlayer2Connected::indicator:unchecked,
 QCheckBox#checkboxPlayer3Connected::indicator:unchecked,
@@ -2255,34 +2488,17 @@ QCheckBox#checkboxPlayer8Connected::indicator:unchecked,
 QGroupBox#groupConnectedController::indicator:unchecked {
   border-radius: 2px;
   border: 1px solid #929192;
-  background: transparent;
+  background: #19232d;
   image: none;
 }
 
-QSpinBox#spinboxLStickRange,
-QSpinBox#spinboxRStickRange {
-  min-width: 38px;
+QWidget#controllerPlayer1,
+QWidget#controllerPlayer2,
+QWidget#controllerPlayer3,
+QWidget#controllerPlayer4,
+QWidget#controllerPlayer5,
+QWidget#controllerPlayer6,
+QWidget#controllerPlayer7,
+QWidget#controllerPlayer8 {
+  background: transparent;
 }
-
-QGroupBox#motionGroup::indicator,
-QGroupBox#vibrationGroup::indicator {
-  margin-left: 0px;
-}
-
-QGroupBox#motionGroup::title,
-QGroupBox#vibrationGroup::title {
-spacing: 2px;
-  padding-left: 1px;
-  padding-right: 1px;
-}
-
-QListWidget#selectorList {
-  background-color: #0f1922;
-}
-
-QSpinBox,
-QLineEdit,
-QTreeView#hotkey_list,
-QScrollArea#scrollArea QTreeView {
-  background-color: #0f1922;
-}

externals/CMakeLists.txt

@@ -73,23 +73,29 @@ if (NOT LIBZIP_FOUND)
 endif()
 
 if (ENABLE_WEB_SERVICE)
-    # LibreSSL
-    set(LIBRESSL_SKIP_INSTALL ON CACHE BOOL "")
-    add_subdirectory(libressl EXCLUDE_FROM_ALL)
-    target_include_directories(ssl INTERFACE ./libressl/include)
-    target_compile_definitions(ssl PRIVATE -DHAVE_INET_NTOP)
-    get_directory_property(OPENSSL_LIBRARIES
-        DIRECTORY libressl
-        DEFINITION OPENSSL_LIBS)
-
-    # lurlparser
-    add_subdirectory(lurlparser EXCLUDE_FROM_ALL)
+    find_package(OpenSSL 1.1)
+    if (OPENSSL_FOUND)
+        set(OPENSSL_LIBRARIES OpenSSL::SSL OpenSSL::Crypto)
+    else()
+        # LibreSSL
+        set(LIBRESSL_SKIP_INSTALL ON CACHE BOOL "")
+        set(OPENSSLDIR "/etc/ssl/")
+        add_subdirectory(libressl EXCLUDE_FROM_ALL)
+        target_include_directories(ssl INTERFACE ./libressl/include)
+        target_compile_definitions(ssl PRIVATE -DHAVE_INET_NTOP)
+        get_directory_property(OPENSSL_LIBRARIES
+            DIRECTORY libressl
+            DEFINITION OPENSSL_LIBS)
+    endif()
 
     # httplib
     add_library(httplib INTERFACE)
     target_include_directories(httplib INTERFACE ./httplib)
     target_compile_definitions(httplib INTERFACE -DCPPHTTPLIB_OPENSSL_SUPPORT)
     target_link_libraries(httplib INTERFACE ${OPENSSL_LIBRARIES})
+    if (WIN32)
+        target_link_libraries(httplib INTERFACE crypt32 cryptui ws2_32)
+    endif()
 endif()
 
 # Opus

externals/find-modules/FindFFmpeg.cmake

@@ -0,0 +1,100 @@
+# - Try to find ffmpeg libraries (libavcodec, libavformat and libavutil)
+# Once done this will define
+#
+# 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 (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
+#
+# Redistribution and use is allowed according to the terms of the New
+# BSD license.
+#
+
+include(FindPackageHandleStandardArgs)
+
+find_package_handle_standard_args(FFMPEG
+  FOUND_VAR FFMPEG_FOUND
+  REQUIRED_VARS
+      FFMPEG_LIBRARY
+      FFMPEG_INCLUDE_DIR
+  VERSION_VAR FFMPEG_VERSION
+)
+
+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)
+  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()
+  else()
+    if(FFMPEG_FIND_REQUIRED)
+      message(FATAL_ERROR
+      "Could not find libavcodec or libavutil or libswscale")
+    endif()
+  endif()
+endif()

externals/lurlparser/CMakeLists.txt

@@ -1,8 +0,0 @@
-add_library(lurlparser
-        LUrlParser.cpp
-        LUrlParser.h
-)
-
-create_target_directory_groups(lurlparser)
-
-target_include_directories(lurlparser INTERFACE ${CMAKE_CURRENT_SOURCE_DIR})

externals/lurlparser/LUrlParser.cpp

@@ -1,265 +0,0 @@
-/*
- * Lightweight URL & URI parser (RFC 1738, RFC 3986)
- * https://github.com/corporateshark/LUrlParser
- *
- * The MIT License (MIT)
- *
- * Copyright (C) 2015 Sergey Kosarevsky (sk@linderdaum.com)
- *
- * Permission is hereby granted, free of charge, to any person obtaining a copy
- * of this software and associated documentation files (the "Software"), to deal
- * in the Software without restriction, including without limitation the rights
- * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
- * copies of the Software, and to permit persons to whom the Software is
- * furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in all
- * copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
- * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
- * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
- * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
- * SOFTWARE.
- */
-
-#include "LUrlParser.h"
-
-#include <algorithm>
-#include <cstring>
-#include <stdlib.h>
-
-// check if the scheme name is valid
-static bool IsSchemeValid( const std::string& SchemeName )
-{
-    for ( auto c : SchemeName  )
-    {
-        if ( !isalpha( c ) && c != '+' && c != '-' && c != '.' ) return false;
-    }
-
-    return true;
-}
-
-bool LUrlParser::clParseURL::GetPort( int* OutPort ) const
-{
-    if ( !IsValid() ) { return false; }
-
-    int Port = atoi( m_Port.c_str() );
-
-    if ( Port <= 0 || Port > 65535 ) { return false; }
-
-    if ( OutPort ) { *OutPort = Port; }
-
-    return true;
-}
-
-// based on RFC 1738 and RFC 3986
-LUrlParser::clParseURL LUrlParser::clParseURL::ParseURL( const std::string& URL )
-{
-    LUrlParser::clParseURL Result;
-
-    const char* CurrentString = URL.c_str();
-
-    /*
-     *	<scheme>:<scheme-specific-part>
-     *	<scheme> := [a-z\+\-\.]+
-     *	For resiliency, programs interpreting URLs should treat upper case letters as equivalent to lower case in scheme names
-     */
-
-    // try to read scheme
-    {
-        const char* LocalString = strchr( CurrentString, ':' );
-
-        if ( !LocalString )
-        {
-            return clParseURL( LUrlParserError_NoUrlCharacter );
-        }
-
-        // save the scheme name
-        Result.m_Scheme = std::string( CurrentString, LocalString - CurrentString );
-
-        if ( !IsSchemeValid( Result.m_Scheme ) )
-        {
-            return clParseURL( LUrlParserError_InvalidSchemeName );
-        }
-
-        // scheme should be lowercase
-        std::transform( Result.m_Scheme.begin(), Result.m_Scheme.end(), Result.m_Scheme.begin(), ::tolower );
-
-        // skip ':'
-        CurrentString = LocalString+1;
-    }
-
-    /*
-     *	//<user>:<password>@<host>:<port>/<url-path>
-     *	any ":", "@" and "/" must be normalized
-     */
-
-    // skip "//"
-    if ( *CurrentString++ != '/' ) return clParseURL( LUrlParserError_NoDoubleSlash );
-    if ( *CurrentString++ != '/' ) return clParseURL( LUrlParserError_NoDoubleSlash );
-
-    // check if the user name and password are specified
-    bool bHasUserName = false;
-
-    const char* LocalString = CurrentString;
-
-    while ( *LocalString )
-    {
-        if ( *LocalString == '@' )
-        {
-            // user name and password are specified
-            bHasUserName = true;
-            break;
-        }
-        else if ( *LocalString == '/' )
-        {
-            // end of <host>:<port> specification
-            bHasUserName = false;
-            break;
-        }
-
-        LocalString++;
-    }
-
-    // user name and password
-    LocalString = CurrentString;
-
-    if ( bHasUserName )
-    {
-        // read user name
-        while ( *LocalString && *LocalString != ':' && *LocalString != '@' ) LocalString++;
-
-        Result.m_UserName = std::string( CurrentString, LocalString - CurrentString );
-
-        // proceed with the current pointer
-        CurrentString = LocalString;
-
-        if ( *CurrentString == ':' )
-        {
-            // skip ':'
-            CurrentString++;
-
-            // read password
-            LocalString = CurrentString;
-
-            while ( *LocalString && *LocalString != '@' ) LocalString++;
-
-            Result.m_Password = std::string( CurrentString, LocalString - CurrentString );
-
-            CurrentString = LocalString;
-        }
-
-        // skip '@'
-        if ( *CurrentString != '@' )
-        {
-            return clParseURL( LUrlParserError_NoAtSign );
-        }
-
-        CurrentString++;
-    }
-
-    bool bHasBracket = ( *CurrentString == '[' );
-
-    // go ahead, read the host name
-    LocalString = CurrentString;
-
-    while ( *LocalString )
-    {
-        if ( bHasBracket && *LocalString == ']' )
-        {
-            // end of IPv6 address
-            LocalString++;
-            break;
-        }
-        else if ( !bHasBracket && ( *LocalString == ':' || *LocalString == '/' ) )
-        {
-            // port number is specified
-            break;
-        }
-
-        LocalString++;
-    }
-
-    Result.m_Host = std::string( CurrentString, LocalString - CurrentString );
-
-    CurrentString = LocalString;
-
-    // is port number specified?
-    if ( *CurrentString == ':' )
-    {
-        CurrentString++;
-
-        // read port number
-        LocalString = CurrentString;
-
-        while ( *LocalString && *LocalString != '/' ) LocalString++;
-
-        Result.m_Port = std::string( CurrentString, LocalString - CurrentString );
-
-        CurrentString = LocalString;
-    }
-
-    // end of string
-    if ( !*CurrentString )
-    {
-        Result.m_ErrorCode = LUrlParserError_Ok;
-
-        return Result;
-    }
-
-    // skip '/'
-    if ( *CurrentString != '/' )
-    {
-        return clParseURL( LUrlParserError_NoSlash );
-    }
-
-    CurrentString++;
-
-    // parse the path
-    LocalString = CurrentString;
-
-    while ( *LocalString && *LocalString != '#' && *LocalString != '?' ) LocalString++;
-
-    Result.m_Path = std::string( CurrentString, LocalString - CurrentString );
-
-    CurrentString = LocalString;
-
-    // check for query
-    if ( *CurrentString == '?' )
-    {
-        // skip '?'
-        CurrentString++;
-
-        // read query
-        LocalString = CurrentString;
-
-        while ( *LocalString && *LocalString != '#' ) LocalString++;
-
-        Result.m_Query = std::string( CurrentString, LocalString - CurrentString );
-
-        CurrentString = LocalString;
-    }
-
-    // check for fragment
-    if ( *CurrentString == '#' )
-    {
-        // skip '#'
-        CurrentString++;
-
-        // read fragment
-        LocalString = CurrentString;
-
-        while ( *LocalString ) LocalString++;
-
-        Result.m_Fragment = std::string( CurrentString, LocalString - CurrentString );
-
-        CurrentString = LocalString;
-    }
-
-    Result.m_ErrorCode = LUrlParserError_Ok;
-
-    return Result;
-}

externals/lurlparser/LUrlParser.h

@@ -1,78 +0,0 @@
-/*
- * Lightweight URL & URI parser (RFC 1738, RFC 3986)
- * https://github.com/corporateshark/LUrlParser
- *
- * The MIT License (MIT)
- *
- * Copyright (C) 2015 Sergey Kosarevsky (sk@linderdaum.com)
- *
- * Permission is hereby granted, free of charge, to any person obtaining a copy
- * of this software and associated documentation files (the "Software"), to deal
- * in the Software without restriction, including without limitation the rights
- * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
- * copies of the Software, and to permit persons to whom the Software is
- * furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in all
- * copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
- * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
- * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
- * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
- * SOFTWARE.
- */
-
-#pragma once
-
-#include <string>
-
-namespace LUrlParser
-{
-enum LUrlParserError
-{
-    LUrlParserError_Ok = 0,
-    LUrlParserError_Uninitialized = 1,
-    LUrlParserError_NoUrlCharacter = 2,
-    LUrlParserError_InvalidSchemeName = 3,
-    LUrlParserError_NoDoubleSlash = 4,
-    LUrlParserError_NoAtSign = 5,
-    LUrlParserError_UnexpectedEndOfLine = 6,
-    LUrlParserError_NoSlash = 7,
-};
-
-class clParseURL
-{
-public:
-    LUrlParserError m_ErrorCode;
-    std::string m_Scheme;
-    std::string m_Host;
-    std::string m_Port;
-    std::string m_Path;
-    std::string m_Query;
-    std::string m_Fragment;
-    std::string m_UserName;
-    std::string m_Password;
-
-    clParseURL()
-            : m_ErrorCode( LUrlParserError_Uninitialized )
-    {}
-
-    /// return 'true' if the parsing was successful
-    bool IsValid() const { return m_ErrorCode == LUrlParserError_Ok; }
-
-    /// helper to convert the port number to int, return 'true' if the port is valid (within the 0..65535 range)
-    bool GetPort( int* OutPort ) const;
-
-    /// parse the URL
-    static clParseURL ParseURL( const std::string& URL );
-
-private:
-    explicit clParseURL( LUrlParserError ErrorCode )
-            : m_ErrorCode( ErrorCode )
-    {}
-};
-
-} // namespace LUrlParser

externals/lurlparser/README.md

@@ -1,19 +0,0 @@
-From https://github.com/corporateshark/LUrlParser/commit/455d5e2d27e3946f11ad0328fee9ee2628e6a8e2
-
-MIT License
-
-===
-
-Lightweight URL & URI parser (RFC 1738, RFC 3986)
-
-(C) Sergey Kosarevsky, 2015
-
-@corporateshark sk@linderdaum.com
-
-http://www.linderdaum.com
-
-http://blog.linderdaum.com
-
-=============================
-
-A tiny and lightweight URL & URI parser (RFC 1738, RFC 3986) written in C++.

src/audio_core/CMakeLists.txt

@@ -12,22 +12,54 @@ add_library(audio_core STATIC
     buffer.h
     codec.cpp
     codec.h
+    command_generator.cpp
+    command_generator.h
     common.h
+    effect_context.cpp
+    effect_context.h
+    info_updater.cpp
+    info_updater.h
+    memory_pool.cpp
+    memory_pool.h
+    mix_context.cpp
+    mix_context.h
     null_sink.h
     sink.h
+    sink_context.cpp
+    sink_context.h
     sink_details.cpp
     sink_details.h
     sink_stream.h
+    splitter_context.cpp
+    splitter_context.h
     stream.cpp
     stream.h
     time_stretch.cpp
     time_stretch.h
+    voice_context.cpp
+    voice_context.h
 
     $<$<BOOL:${ENABLE_CUBEB}>:cubeb_sink.cpp cubeb_sink.h>
 )
 
 create_target_directory_groups(audio_core)
 
+if (NOT MSVC)
+    target_compile_options(audio_core PRIVATE
+        -Werror=conversion
+        -Werror=ignored-qualifiers
+        -Werror=implicit-fallthrough
+        -Werror=reorder
+        -Werror=sign-compare
+        -Werror=unused-variable
+
+        $<$<CXX_COMPILER_ID:GNU>:-Werror=unused-but-set-parameter>
+        $<$<CXX_COMPILER_ID:GNU>:-Werror=unused-but-set-variable>
+
+        -Wno-sign-conversion
+    )
+endif()
+
 target_link_libraries(audio_core PUBLIC common core)
 target_link_libraries(audio_core PRIVATE SoundTouch)
 

src/audio_core/algorithm/filter.cpp

@@ -55,7 +55,8 @@ void Filter::Process(std::vector<s16>& signal) {
 /// @param total_count The total number of biquads to be cascaded.
 /// @param index 0-index of the biquad to calculate the Q value for.
 static double CascadingBiquadQ(std::size_t total_count, std::size_t index) {
-    const double pole = M_PI * (2 * index + 1) / (4.0 * total_count);
+    const auto pole =
+        M_PI * static_cast<double>(2 * index + 1) / (4.0 * static_cast<double>(total_count));
     return 1.0 / (2.0 * std::cos(pole));
 }
 

src/audio_core/algorithm/interpolate.cpp

@@ -146,7 +146,7 @@ std::vector<s16> Interpolate(InterpolationState& state, std::vector<s16> input,
         return {};
 
     if (ratio <= 0) {
-        LOG_CRITICAL(Audio, "Nonsensical interpolation ratio {}", ratio);
+        LOG_ERROR(Audio, "Nonsensical interpolation ratio {}", ratio);
         return input;
     }
 
@@ -164,7 +164,8 @@ std::vector<s16> Interpolate(InterpolationState& state, std::vector<s16> input,
     const std::size_t num_frames{input.size() / 2};
 
     std::vector<s16> output;
-    output.reserve(static_cast<std::size_t>(input.size() / ratio + InterpolationState::taps));
+    output.reserve(static_cast<std::size_t>(static_cast<double>(input.size()) / ratio +
+                                            InterpolationState::taps));
 
     for (std::size_t frame{}; frame < num_frames; ++frame) {
         const std::size_t lut_index{(state.fraction >> 8) * InterpolationState::taps};
@@ -197,4 +198,36 @@ std::vector<s16> Interpolate(InterpolationState& state, std::vector<s16> input,
     return output;
 }
 
+void Resample(s32* output, const s32* input, s32 pitch, s32& fraction, std::size_t sample_count) {
+    const std::array<s16, 512>& lut = [pitch] {
+        if (pitch > 0xaaaa) {
+            return curve_lut0;
+        }
+        if (pitch <= 0x8000) {
+            return curve_lut1;
+        }
+        return curve_lut2;
+    }();
+
+    std::size_t index{};
+
+    for (std::size_t i = 0; i < sample_count; i++) {
+        const std::size_t lut_index{(static_cast<std::size_t>(fraction) >> 8) * 4};
+        const auto l0 = lut[lut_index + 0];
+        const auto l1 = lut[lut_index + 1];
+        const auto l2 = lut[lut_index + 2];
+        const auto l3 = lut[lut_index + 3];
+
+        const auto s0 = static_cast<s32>(input[index]);
+        const auto s1 = static_cast<s32>(input[index + 1]);
+        const auto s2 = static_cast<s32>(input[index + 2]);
+        const auto s3 = static_cast<s32>(input[index + 3]);
+
+        output[i] = (l0 * s0 + l1 * s1 + l2 * s2 + l3 * s3) >> 15;
+        fraction += pitch;
+        index += (fraction >> 15);
+        fraction &= 0x7fff;
+    }
+}
+
 } // namespace AudioCore

src/audio_core/algorithm/interpolate.h

@@ -38,4 +38,7 @@ inline std::vector<s16> Interpolate(InterpolationState& state, std::vector<s16>
     return Interpolate(state, std::move(input), ratio);
 }
 
+/// Nintendo Switchs DSP resampling algorithm. Based on a single channel
+void Resample(s32* output, const s32* input, s32 pitch, s32& fraction, std::size_t sample_count);
+
 } // namespace AudioCore

src/audio_core/audio_renderer.cpp

@@ -2,95 +2,46 @@
 // Licensed under GPLv2 or any later version
 // Refer to the license.txt file included.
 
-#include "audio_core/algorithm/interpolate.h"
+#include <vector>
+
 #include "audio_core/audio_out.h"
 #include "audio_core/audio_renderer.h"
-#include "audio_core/codec.h"
 #include "audio_core/common.h"
-#include "common/assert.h"
+#include "audio_core/info_updater.h"
+#include "audio_core/voice_context.h"
 #include "common/logging/log.h"
-#include "core/core.h"
 #include "core/hle/kernel/writable_event.h"
 #include "core/memory.h"
+#include "core/settings.h"
 
 namespace AudioCore {
-
-constexpr u32 STREAM_SAMPLE_RATE{48000};
-constexpr u32 STREAM_NUM_CHANNELS{2};
-using VoiceChannelHolder = std::array<VoiceResourceInformation*, 6>;
-class AudioRenderer::VoiceState {
-public:
-    bool IsPlaying() const {
-        return is_in_use && info.play_state == PlayState::Started;
-    }
-
-    const VoiceOutStatus& GetOutStatus() const {
-        return out_status;
-    }
-
-    const VoiceInfo& GetInfo() const {
-        return info;
-    }
-
-    VoiceInfo& GetInfo() {
-        return info;
-    }
-
-    void SetWaveIndex(std::size_t index);
-    std::vector<s16> DequeueSamples(std::size_t sample_count, Core::Memory::Memory& memory,
-                                    const VoiceChannelHolder& voice_resources);
-    void UpdateState();
-    void RefreshBuffer(Core::Memory::Memory& memory, const VoiceChannelHolder& voice_resources);
-
-private:
-    bool is_in_use{};
-    bool is_refresh_pending{};
-    std::size_t wave_index{};
-    std::size_t offset{};
-    Codec::ADPCMState adpcm_state{};
-    InterpolationState interp_state{};
-    std::vector<s16> samples;
-    VoiceOutStatus out_status{};
-    VoiceInfo info{};
-};
-
-class AudioRenderer::EffectState {
-public:
-    const EffectOutStatus& GetOutStatus() const {
-        return out_status;
-    }
-
-    const EffectInStatus& GetInfo() const {
-        return info;
-    }
-
-    EffectInStatus& GetInfo() {
-        return info;
-    }
-
-    void UpdateState(Core::Memory::Memory& memory);
-
-private:
-    EffectOutStatus out_status{};
-    EffectInStatus info{};
-};
-
 AudioRenderer::AudioRenderer(Core::Timing::CoreTiming& core_timing, Core::Memory::Memory& memory_,
-                             AudioRendererParameter params,
+                             AudioCommon::AudioRendererParameter params,
                              std::shared_ptr<Kernel::WritableEvent> buffer_event,
                              std::size_t instance_number)
-    : worker_params{params}, buffer_event{buffer_event}, voices(params.voice_count),
-      voice_resources(params.voice_count), effects(params.effect_count), memory{memory_} {
+    : worker_params{params}, buffer_event{buffer_event},
+      memory_pool_info(params.effect_count + params.voice_count * 4),
+      voice_context(params.voice_count), effect_context(params.effect_count), mix_context(),
+      sink_context(params.sink_count), splitter_context(),
+      voices(params.voice_count), memory{memory_},
+      command_generator(worker_params, voice_context, mix_context, splitter_context, effect_context,
+                        memory),
+      temp_mix_buffer(AudioCommon::TOTAL_TEMP_MIX_SIZE) {
     behavior_info.SetUserRevision(params.revision);
+    splitter_context.Initialize(behavior_info, params.splitter_count,
+                                params.num_splitter_send_channels);
+    mix_context.Initialize(behavior_info, params.submix_count + 1, params.effect_count);
     audio_out = std::make_unique<AudioCore::AudioOut>();
-    stream = audio_out->OpenStream(core_timing, STREAM_SAMPLE_RATE, STREAM_NUM_CHANNELS,
-                                   fmt::format("AudioRenderer-Instance{}", instance_number),
-                                   [=]() { buffer_event->Signal(); });
+    stream =
+        audio_out->OpenStream(core_timing, params.sample_rate, AudioCommon::STREAM_NUM_CHANNELS,
+                              fmt::format("AudioRenderer-Instance{}", instance_number),
+                              [=]() { buffer_event->Signal(); });
     audio_out->StartStream(stream);
 
     QueueMixedBuffer(0);
     QueueMixedBuffer(1);
     QueueMixedBuffer(2);
+    QueueMixedBuffer(3);
 }
 
 AudioRenderer::~AudioRenderer() = default;
@@ -111,355 +62,200 @@ Stream::State AudioRenderer::GetStreamState() const {
     return stream->GetState();
 }
 
-ResultVal<std::vector<u8>> AudioRenderer::UpdateAudioRenderer(const std::vector<u8>& input_params) {
-    // Copy UpdateDataHeader struct
-    UpdateDataHeader config{};
-    std::memcpy(&config, input_params.data(), sizeof(UpdateDataHeader));
-    u32 memory_pool_count = worker_params.effect_count + (worker_params.voice_count * 4);
-
-    if (!behavior_info.UpdateInput(input_params, sizeof(UpdateDataHeader))) {
-        LOG_ERROR(Audio, "Failed to update behavior info input parameters");
-        return Audren::ERR_INVALID_PARAMETERS;
-    }
-
-    // Copy MemoryPoolInfo structs
-    std::vector<MemoryPoolInfo> mem_pool_info(memory_pool_count);
-    std::memcpy(mem_pool_info.data(),
-                input_params.data() + sizeof(UpdateDataHeader) + config.behavior_size,
-                memory_pool_count * sizeof(MemoryPoolInfo));
-
-    // Copy voice resources
-    const std::size_t voice_resource_offset{sizeof(UpdateDataHeader) + config.behavior_size +
-                                            config.memory_pools_size};
-    std::memcpy(voice_resources.data(), input_params.data() + voice_resource_offset,
-                sizeof(VoiceResourceInformation) * voice_resources.size());
-
-    // Copy VoiceInfo structs
-    std::size_t voice_offset{sizeof(UpdateDataHeader) + config.behavior_size +
-                             config.memory_pools_size + config.voice_resource_size};
-    for (auto& voice : voices) {
-        std::memcpy(&voice.GetInfo(), input_params.data() + voice_offset, sizeof(VoiceInfo));
-        voice_offset += sizeof(VoiceInfo);
-    }
-
-    std::size_t effect_offset{sizeof(UpdateDataHeader) + config.behavior_size +
-                              config.memory_pools_size + config.voice_resource_size +
-                              config.voices_size};
-    for (auto& effect : effects) {
-        std::memcpy(&effect.GetInfo(), input_params.data() + effect_offset, sizeof(EffectInStatus));
-        effect_offset += sizeof(EffectInStatus);
-    }
-
-    // Update memory pool state
-    std::vector<MemoryPoolEntry> memory_pool(memory_pool_count);
-    for (std::size_t index = 0; index < memory_pool.size(); ++index) {
-        if (mem_pool_info[index].pool_state == MemoryPoolStates::RequestAttach) {
-            memory_pool[index].state = MemoryPoolStates::Attached;
-        } else if (mem_pool_info[index].pool_state == MemoryPoolStates::RequestDetach) {
-            memory_pool[index].state = MemoryPoolStates::Detached;
-        }
-    }
-
-    // Update voices
-    for (auto& voice : voices) {
-        voice.UpdateState();
-        if (!voice.GetInfo().is_in_use) {
-            continue;
-        }
-        if (voice.GetInfo().is_new) {
-            voice.SetWaveIndex(voice.GetInfo().wave_buffer_head);
-        }
-    }
-
-    for (auto& effect : effects) {
-        effect.UpdateState(memory);
-    }
-
-    // Release previous buffers and queue next ones for playback
-    ReleaseAndQueueBuffers();
-
-    // Copy output header
-    UpdateDataHeader response_data{worker_params};
-    if (behavior_info.IsElapsedFrameCountSupported()) {
-        response_data.render_info = sizeof(RendererInfo);
-        response_data.total_size += sizeof(RendererInfo);
-    }
-
-    std::vector<u8> output_params(response_data.total_size);
-    std::memcpy(output_params.data(), &response_data, sizeof(UpdateDataHeader));
-
-    // Copy output memory pool entries
-    std::memcpy(output_params.data() + sizeof(UpdateDataHeader), memory_pool.data(),
-                response_data.memory_pools_size);
-
-    // Copy output voice status
-    std::size_t voice_out_status_offset{sizeof(UpdateDataHeader) + response_data.memory_pools_size};
-    for (const auto& voice : voices) {
-        std::memcpy(output_params.data() + voice_out_status_offset, &voice.GetOutStatus(),
-                    sizeof(VoiceOutStatus));
-        voice_out_status_offset += sizeof(VoiceOutStatus);
-    }
-
-    std::size_t effect_out_status_offset{
-        sizeof(UpdateDataHeader) + response_data.memory_pools_size + response_data.voices_size +
-        response_data.voice_resource_size};
-    for (const auto& effect : effects) {
-        std::memcpy(output_params.data() + effect_out_status_offset, &effect.GetOutStatus(),
-                    sizeof(EffectOutStatus));
-        effect_out_status_offset += sizeof(EffectOutStatus);
-    }
-
-    // Update behavior info output
-    const std::size_t behavior_out_status_offset{
-        sizeof(UpdateDataHeader) + response_data.memory_pools_size + response_data.voices_size +
-        response_data.effects_size + response_data.sinks_size +
-        response_data.performance_manager_size};
-
-    if (!behavior_info.UpdateOutput(output_params, behavior_out_status_offset)) {
-        LOG_ERROR(Audio, "Failed to update behavior info output parameters");
-        return Audren::ERR_INVALID_PARAMETERS;
-    }
-
-    if (behavior_info.IsElapsedFrameCountSupported()) {
-        const std::size_t renderer_info_offset{
-            sizeof(UpdateDataHeader) + response_data.memory_pools_size + response_data.voices_size +
-            response_data.effects_size + response_data.sinks_size +
-            response_data.performance_manager_size + response_data.behavior_size};
-        RendererInfo renderer_info{};
-        renderer_info.elasped_frame_count = elapsed_frame_count;
-        std::memcpy(output_params.data() + renderer_info_offset, &renderer_info,
-                    sizeof(RendererInfo));
-    }
-
-    return MakeResult(output_params);
-}
-
-void AudioRenderer::VoiceState::SetWaveIndex(std::size_t index) {
-    wave_index = index & 3;
-    is_refresh_pending = true;
-}
-
-std::vector<s16> AudioRenderer::VoiceState::DequeueSamples(
-    std::size_t sample_count, Core::Memory::Memory& memory,
-    const VoiceChannelHolder& voice_resources) {
-    if (!IsPlaying()) {
-        return {};
-    }
-
-    if (is_refresh_pending) {
-        RefreshBuffer(memory, voice_resources);
-    }
-
-    const std::size_t max_size{samples.size() - offset};
-    const std::size_t dequeue_offset{offset};
-    std::size_t size{sample_count * STREAM_NUM_CHANNELS};
-    if (size > max_size) {
-        size = max_size;
-    }
-
-    out_status.played_sample_count += size / STREAM_NUM_CHANNELS;
-    offset += size;
-
-    const auto& wave_buffer{info.wave_buffer[wave_index]};
-    if (offset == samples.size()) {
-        offset = 0;
-
-        if (!wave_buffer.is_looping && wave_buffer.buffer_sz) {
-            SetWaveIndex(wave_index + 1);
-        }
-
-        if (wave_buffer.buffer_sz) {
-            out_status.wave_buffer_consumed++;
-        }
-
-        if (wave_buffer.end_of_stream || wave_buffer.buffer_sz == 0) {
-            info.play_state = PlayState::Paused;
-        }
-    }
-
-    return {samples.begin() + dequeue_offset, samples.begin() + dequeue_offset + size};
-}
-
-void AudioRenderer::VoiceState::UpdateState() {
-    if (is_in_use && !info.is_in_use) {
-        // No longer in use, reset state
-        is_refresh_pending = true;
-        wave_index = 0;
-        offset = 0;
-        out_status = {};
-    }
-    is_in_use = info.is_in_use;
-}
-
-void AudioRenderer::VoiceState::RefreshBuffer(Core::Memory::Memory& memory,
-                                              const VoiceChannelHolder& voice_resources) {
-    const auto wave_buffer_address = info.wave_buffer[wave_index].buffer_addr;
-    const auto wave_buffer_size = info.wave_buffer[wave_index].buffer_sz;
-    std::vector<s16> new_samples(wave_buffer_size / sizeof(s16));
-    memory.ReadBlock(wave_buffer_address, new_samples.data(), wave_buffer_size);
-
-    switch (static_cast<Codec::PcmFormat>(info.sample_format)) {
-    case Codec::PcmFormat::Int16: {
-        // PCM16 is played as-is
-        break;
-    }
-    case Codec::PcmFormat::Adpcm: {
-        // Decode ADPCM to PCM16
-        Codec::ADPCM_Coeff coeffs;
-        memory.ReadBlock(info.additional_params_addr, coeffs.data(), sizeof(Codec::ADPCM_Coeff));
-        new_samples = Codec::DecodeADPCM(reinterpret_cast<u8*>(new_samples.data()),
-                                         new_samples.size() * sizeof(s16), coeffs, adpcm_state);
-        break;
-    }
-    default:
-        UNIMPLEMENTED_MSG("Unimplemented sample_format={}", info.sample_format);
-        break;
-    }
-
-    switch (info.channel_count) {
-    case 1: {
-        // 1 channel is upsampled to 2 channel
-        samples.resize(new_samples.size() * 2);
-
-        for (std::size_t index = 0; index < new_samples.size(); ++index) {
-            auto sample = static_cast<float>(new_samples[index]);
-            if (voice_resources[0]->in_use) {
-                sample *= voice_resources[0]->mix_volumes[0];
-            }
-
-            samples[index * 2] = static_cast<s16>(sample * info.volume);
-            samples[index * 2 + 1] = static_cast<s16>(sample * info.volume);
-        }
-        break;
-    }
-    case 2: {
-        // 2 channel is played as is
-        samples = std::move(new_samples);
-        const std::size_t sample_count = (samples.size() / 2);
-        for (std::size_t index = 0; index < sample_count; ++index) {
-            const std::size_t index_l = index * 2;
-            const std::size_t index_r = index * 2 + 1;
-
-            auto sample_l = static_cast<float>(samples[index_l]);
-            auto sample_r = static_cast<float>(samples[index_r]);
-
-            if (voice_resources[0]->in_use) {
-                sample_l *= voice_resources[0]->mix_volumes[0];
-            }
-
-            if (voice_resources[1]->in_use) {
-                sample_r *= voice_resources[1]->mix_volumes[1];
-            }
-
-            samples[index_l] = static_cast<s16>(sample_l * info.volume);
-            samples[index_r] = static_cast<s16>(sample_r * info.volume);
-        }
-        break;
-    }
-    case 6: {
-        samples.resize((new_samples.size() / 6) * 2);
-        const std::size_t sample_count = samples.size() / 2;
-
-        for (std::size_t index = 0; index < sample_count; ++index) {
-            auto FL = static_cast<float>(new_samples[index * 6]);
-            auto FR = static_cast<float>(new_samples[index * 6 + 1]);
-            auto FC = static_cast<float>(new_samples[index * 6 + 2]);
-            auto BL = static_cast<float>(new_samples[index * 6 + 4]);
-            auto BR = static_cast<float>(new_samples[index * 6 + 5]);
-
-            if (voice_resources[0]->in_use) {
-                FL *= voice_resources[0]->mix_volumes[0];
-            }
-            if (voice_resources[1]->in_use) {
-                FR *= voice_resources[1]->mix_volumes[1];
-            }
-            if (voice_resources[2]->in_use) {
-                FC *= voice_resources[2]->mix_volumes[2];
-            }
-            if (voice_resources[4]->in_use) {
-                BL *= voice_resources[4]->mix_volumes[4];
-            }
-            if (voice_resources[5]->in_use) {
-                BR *= voice_resources[5]->mix_volumes[5];
-            }
-
-            samples[index * 2] =
-                static_cast<s16>((0.3694f * FL + 0.2612f * FC + 0.3694f * BL) * info.volume);
-            samples[index * 2 + 1] =
-                static_cast<s16>((0.3694f * FR + 0.2612f * FC + 0.3694f * BR) * info.volume);
-        }
-        break;
-    }
-    default:
-        UNIMPLEMENTED_MSG("Unimplemented channel_count={}", info.channel_count);
-        break;
-    }
-
-    // Only interpolate when necessary, expensive.
-    if (GetInfo().sample_rate != STREAM_SAMPLE_RATE) {
-        samples = Interpolate(interp_state, std::move(samples), GetInfo().sample_rate,
-                              STREAM_SAMPLE_RATE);
-    }
-
-    is_refresh_pending = false;
-}
-
-void AudioRenderer::EffectState::UpdateState(Core::Memory::Memory& memory) {
-    if (info.is_new) {
-        out_status.state = EffectStatus::New;
-    } else {
-        if (info.type == Effect::Aux) {
-            ASSERT_MSG(memory.Read32(info.aux_info.return_buffer_info) == 0,
-                       "Aux buffers tried to update");
-            ASSERT_MSG(memory.Read32(info.aux_info.send_buffer_info) == 0,
-                       "Aux buffers tried to update");
-            ASSERT_MSG(memory.Read32(info.aux_info.return_buffer_base) == 0,
-                       "Aux buffers tried to update");
-            ASSERT_MSG(memory.Read32(info.aux_info.send_buffer_base) == 0,
-                       "Aux buffers tried to update");
-        }
-    }
-}
-
 static constexpr s16 ClampToS16(s32 value) {
     return static_cast<s16>(std::clamp(value, -32768, 32767));
 }
 
+ResultCode AudioRenderer::UpdateAudioRenderer(const std::vector<u8>& input_params,
+                                              std::vector<u8>& output_params) {
+
+    InfoUpdater info_updater{input_params, output_params, behavior_info};
+
+    if (!info_updater.UpdateBehaviorInfo(behavior_info)) {
+        LOG_ERROR(Audio, "Failed to update behavior info input parameters");
+        return AudioCommon::Audren::ERR_INVALID_PARAMETERS;
+    }
+
+    if (!info_updater.UpdateMemoryPools(memory_pool_info)) {
+        LOG_ERROR(Audio, "Failed to update memory pool parameters");
+        return AudioCommon::Audren::ERR_INVALID_PARAMETERS;
+    }
+
+    if (!info_updater.UpdateVoiceChannelResources(voice_context)) {
+        LOG_ERROR(Audio, "Failed to update voice channel resource parameters");
+        return AudioCommon::Audren::ERR_INVALID_PARAMETERS;
+    }
+
+    if (!info_updater.UpdateVoices(voice_context, memory_pool_info, 0)) {
+        LOG_ERROR(Audio, "Failed to update voice parameters");
+        return AudioCommon::Audren::ERR_INVALID_PARAMETERS;
+    }
+
+    // TODO(ogniK): Deal with stopped audio renderer but updates still taking place
+    if (!info_updater.UpdateEffects(effect_context, true)) {
+        LOG_ERROR(Audio, "Failed to update effect parameters");
+        return AudioCommon::Audren::ERR_INVALID_PARAMETERS;
+    }
+
+    if (behavior_info.IsSplitterSupported()) {
+        if (!info_updater.UpdateSplitterInfo(splitter_context)) {
+            LOG_ERROR(Audio, "Failed to update splitter parameters");
+            return AudioCommon::Audren::ERR_INVALID_PARAMETERS;
+        }
+    }
+
+    auto mix_result = info_updater.UpdateMixes(mix_context, worker_params.mix_buffer_count,
+                                               splitter_context, effect_context);
+
+    if (mix_result.IsError()) {
+        LOG_ERROR(Audio, "Failed to update mix parameters");
+        return mix_result;
+    }
+
+    // TODO(ogniK): Sinks
+    if (!info_updater.UpdateSinks(sink_context)) {
+        LOG_ERROR(Audio, "Failed to update sink parameters");
+        return AudioCommon::Audren::ERR_INVALID_PARAMETERS;
+    }
+
+    // TODO(ogniK): Performance buffer
+    if (!info_updater.UpdatePerformanceBuffer()) {
+        LOG_ERROR(Audio, "Failed to update performance buffer parameters");
+        return AudioCommon::Audren::ERR_INVALID_PARAMETERS;
+    }
+
+    if (!info_updater.UpdateErrorInfo(behavior_info)) {
+        LOG_ERROR(Audio, "Failed to update error info");
+        return AudioCommon::Audren::ERR_INVALID_PARAMETERS;
+    }
+
+    if (behavior_info.IsElapsedFrameCountSupported()) {
+        if (!info_updater.UpdateRendererInfo(elapsed_frame_count)) {
+            LOG_ERROR(Audio, "Failed to update renderer info");
+            return AudioCommon::Audren::ERR_INVALID_PARAMETERS;
+        }
+    }
+    // TODO(ogniK): Statistics
+
+    if (!info_updater.WriteOutputHeader()) {
+        LOG_ERROR(Audio, "Failed to write output header");
+        return AudioCommon::Audren::ERR_INVALID_PARAMETERS;
+    }
+
+    // TODO(ogniK): Check when all sections are implemented
+
+    if (!info_updater.CheckConsumedSize()) {
+        LOG_ERROR(Audio, "Audio buffers were not consumed!");
+        return AudioCommon::Audren::ERR_INVALID_PARAMETERS;
+    }
+
+    ReleaseAndQueueBuffers();
+
+    return RESULT_SUCCESS;
+}
+
 void AudioRenderer::QueueMixedBuffer(Buffer::Tag tag) {
-    constexpr std::size_t BUFFER_SIZE{512};
+    command_generator.PreCommand();
+    // Clear mix buffers before our next operation
+    command_generator.ClearMixBuffers();
+
+    // If the splitter is not in use, sort our mixes
+    if (!splitter_context.UsingSplitter()) {
+        mix_context.SortInfo();
+    }
+    // Sort our voices
+    voice_context.SortInfo();
+
+    // Handle samples
+    command_generator.GenerateVoiceCommands();
+    command_generator.GenerateSubMixCommands();
+    command_generator.GenerateFinalMixCommands();
+
+    command_generator.PostCommand();
+    // Base sample size
+    std::size_t BUFFER_SIZE{worker_params.sample_count};
+    // Samples
     std::vector<s16> buffer(BUFFER_SIZE * stream->GetNumChannels());
+    // Make sure to clear our samples
+    std::memset(buffer.data(), 0, buffer.size() * sizeof(s16));
 
-    for (auto& voice : voices) {
-        if (!voice.IsPlaying()) {
-            continue;
-        }
-        VoiceChannelHolder resources{};
-        for (u32 channel = 0; channel < voice.GetInfo().channel_count; channel++) {
-            const auto channel_resource_id = voice.GetInfo().voice_channel_resource_ids[channel];
-            resources[channel] = &voice_resources[channel_resource_id];
+    if (sink_context.InUse()) {
+        const auto stream_channel_count = stream->GetNumChannels();
+        const auto buffer_offsets = sink_context.OutputBuffers();
+        const auto channel_count = buffer_offsets.size();
+        const auto& final_mix = mix_context.GetFinalMixInfo();
+        const auto& in_params = final_mix.GetInParams();
+        std::vector<s32*> mix_buffers(channel_count);
+        for (std::size_t i = 0; i < channel_count; i++) {
+            mix_buffers[i] =
+                command_generator.GetMixBuffer(in_params.buffer_offset + buffer_offsets[i]);
         }
 
-        std::size_t offset{};
-        s64 samples_remaining{BUFFER_SIZE};
-        while (samples_remaining > 0) {
-            const std::vector<s16> samples{
-                voice.DequeueSamples(samples_remaining, memory, resources)};
+        for (std::size_t i = 0; i < BUFFER_SIZE; i++) {
+            if (channel_count == 1) {
+                const auto sample = ClampToS16(mix_buffers[0][i]);
+                buffer[i * stream_channel_count + 0] = sample;
+                if (stream_channel_count > 1) {
+                    buffer[i * stream_channel_count + 1] = sample;
+                }
+                if (stream_channel_count == 6) {
+                    buffer[i * stream_channel_count + 2] = sample;
+                    buffer[i * stream_channel_count + 4] = sample;
+                    buffer[i * stream_channel_count + 5] = sample;
+                }
+            } else if (channel_count == 2) {
+                const auto l_sample = ClampToS16(mix_buffers[0][i]);
+                const auto r_sample = ClampToS16(mix_buffers[1][i]);
+                if (stream_channel_count == 1) {
+                    buffer[i * stream_channel_count + 0] = l_sample;
+                } else if (stream_channel_count == 2) {
+                    buffer[i * stream_channel_count + 0] = l_sample;
+                    buffer[i * stream_channel_count + 1] = r_sample;
+                } else if (stream_channel_count == 6) {
+                    buffer[i * stream_channel_count + 0] = l_sample;
+                    buffer[i * stream_channel_count + 1] = r_sample;
 
-            if (samples.empty()) {
-                break;
-            }
+                    buffer[i * stream_channel_count + 2] =
+                        ClampToS16((static_cast<s32>(l_sample) + static_cast<s32>(r_sample)) / 2);
 
-            samples_remaining -= samples.size() / stream->GetNumChannels();
+                    buffer[i * stream_channel_count + 4] = l_sample;
+                    buffer[i * stream_channel_count + 5] = r_sample;
+                }
 
-            for (const auto& sample : samples) {
-                const s32 buffer_sample{buffer[offset]};
-                buffer[offset++] =
-                    ClampToS16(buffer_sample + static_cast<s32>(sample * voice.GetInfo().volume));
+            } else if (channel_count == 6) {
+                const auto fl_sample = ClampToS16(mix_buffers[0][i]);
+                const auto fr_sample = ClampToS16(mix_buffers[1][i]);
+                const auto fc_sample = ClampToS16(mix_buffers[2][i]);
+                const auto lf_sample = ClampToS16(mix_buffers[3][i]);
+                const auto bl_sample = ClampToS16(mix_buffers[4][i]);
+                const auto br_sample = ClampToS16(mix_buffers[5][i]);
+
+                if (stream_channel_count == 1) {
+                    buffer[i * stream_channel_count + 0] = fc_sample;
+                } else if (stream_channel_count == 2) {
+                    buffer[i * stream_channel_count + 0] =
+                        static_cast<s16>(0.3694f * static_cast<float>(fl_sample) +
+                                         0.2612f * static_cast<float>(fc_sample) +
+                                         0.3694f * static_cast<float>(bl_sample));
+                    buffer[i * stream_channel_count + 1] =
+                        static_cast<s16>(0.3694f * static_cast<float>(fr_sample) +
+                                         0.2612f * static_cast<float>(fc_sample) +
+                                         0.3694f * static_cast<float>(br_sample));
+                } else if (stream_channel_count == 6) {
+                    buffer[i * stream_channel_count + 0] = fl_sample;
+                    buffer[i * stream_channel_count + 1] = fr_sample;
+                    buffer[i * stream_channel_count + 2] = fc_sample;
+                    buffer[i * stream_channel_count + 3] = lf_sample;
+                    buffer[i * stream_channel_count + 4] = bl_sample;
+                    buffer[i * stream_channel_count + 5] = br_sample;
+                }
             }
         }
     }
+
     audio_out->QueueBuffer(stream, tag, std::move(buffer));
     elapsed_frame_count++;
+    voice_context.UpdateStateByDspShared();
 }
 
 void AudioRenderer::ReleaseAndQueueBuffers() {

src/audio_core/audio_renderer.h

@@ -9,12 +9,18 @@
 #include <vector>
 
 #include "audio_core/behavior_info.h"
+#include "audio_core/command_generator.h"
 #include "audio_core/common.h"
+#include "audio_core/effect_context.h"
+#include "audio_core/memory_pool.h"
+#include "audio_core/mix_context.h"
+#include "audio_core/sink_context.h"
+#include "audio_core/splitter_context.h"
 #include "audio_core/stream.h"
+#include "audio_core/voice_context.h"
 #include "common/common_funcs.h"
 #include "common/common_types.h"
 #include "common/swap.h"
-#include "core/hle/kernel/object.h"
 #include "core/hle/result.h"
 
 namespace Core::Timing {
@@ -30,220 +36,25 @@ class Memory;
 }
 
 namespace AudioCore {
+using DSPStateHolder = std::array<VoiceState*, 6>;
 
 class AudioOut;
 
-enum class PlayState : u8 {
-    Started = 0,
-    Stopped = 1,
-    Paused = 2,
-};
-
-enum class Effect : u8 {
-    None = 0,
-    Aux = 2,
-};
-
-enum class EffectStatus : u8 {
-    None = 0,
-    New = 1,
-};
-
-struct AudioRendererParameter {
-    u32_le sample_rate;
-    u32_le sample_count;
-    u32_le mix_buffer_count;
-    u32_le submix_count;
-    u32_le voice_count;
-    u32_le sink_count;
-    u32_le effect_count;
-    u32_le performance_frame_count;
-    u8 is_voice_drop_enabled;
-    u8 unknown_21;
-    u8 unknown_22;
-    u8 execution_mode;
-    u32_le splitter_count;
-    u32_le num_splitter_send_channels;
-    u32_le unknown_30;
-    u32_le revision;
-};
-static_assert(sizeof(AudioRendererParameter) == 52, "AudioRendererParameter is an invalid size");
-
-enum class MemoryPoolStates : u32 { // Should be LE
-    Invalid = 0x0,
-    Unknown = 0x1,
-    RequestDetach = 0x2,
-    Detached = 0x3,
-    RequestAttach = 0x4,
-    Attached = 0x5,
-    Released = 0x6,
-};
-
-struct MemoryPoolEntry {
-    MemoryPoolStates state;
-    u32_le unknown_4;
-    u32_le unknown_8;
-    u32_le unknown_c;
-};
-static_assert(sizeof(MemoryPoolEntry) == 0x10, "MemoryPoolEntry has wrong size");
-
-struct MemoryPoolInfo {
-    u64_le pool_address;
-    u64_le pool_size;
-    MemoryPoolStates pool_state;
-    INSERT_PADDING_WORDS(3); // Unknown
-};
-static_assert(sizeof(MemoryPoolInfo) == 0x20, "MemoryPoolInfo has wrong size");
-struct BiquadFilter {
-    u8 enable;
-    INSERT_PADDING_BYTES(1);
-    std::array<s16_le, 3> numerator;
-    std::array<s16_le, 2> denominator;
-};
-static_assert(sizeof(BiquadFilter) == 0xc, "BiquadFilter has wrong size");
-
-struct WaveBuffer {
-    u64_le buffer_addr;
-    u64_le buffer_sz;
-    s32_le start_sample_offset;
-    s32_le end_sample_offset;
-    u8 is_looping;
-    u8 end_of_stream;
-    u8 sent_to_server;
-    INSERT_PADDING_BYTES(5);
-    u64 context_addr;
-    u64 context_sz;
-    INSERT_PADDING_BYTES(8);
-};
-static_assert(sizeof(WaveBuffer) == 0x38, "WaveBuffer has wrong size");
-
-struct VoiceResourceInformation {
-    s32_le id{};
-    std::array<float_le, MAX_MIX_BUFFERS> mix_volumes{};
-    bool in_use{};
-    INSERT_PADDING_BYTES(11);
-};
-static_assert(sizeof(VoiceResourceInformation) == 0x70, "VoiceResourceInformation has wrong size");
-
-struct VoiceInfo {
-    u32_le id;
-    u32_le node_id;
-    u8 is_new;
-    u8 is_in_use;
-    PlayState play_state;
-    u8 sample_format;
-    u32_le sample_rate;
-    u32_le priority;
-    u32_le sorting_order;
-    u32_le channel_count;
-    float_le pitch;
-    float_le volume;
-    std::array<BiquadFilter, 2> biquad_filter;
-    u32_le wave_buffer_count;
-    u32_le wave_buffer_head;
-    INSERT_PADDING_WORDS(1);
-    u64_le additional_params_addr;
-    u64_le additional_params_sz;
-    u32_le mix_id;
-    u32_le splitter_info_id;
-    std::array<WaveBuffer, 4> wave_buffer;
-    std::array<u32_le, 6> voice_channel_resource_ids;
-    INSERT_PADDING_BYTES(24);
-};
-static_assert(sizeof(VoiceInfo) == 0x170, "VoiceInfo is wrong size");
-
-struct VoiceOutStatus {
-    u64_le played_sample_count;
-    u32_le wave_buffer_consumed;
-    u32_le voice_drops_count;
-};
-static_assert(sizeof(VoiceOutStatus) == 0x10, "VoiceOutStatus has wrong size");
-
-struct AuxInfo {
-    std::array<u8, 24> input_mix_buffers;
-    std::array<u8, 24> output_mix_buffers;
-    u32_le mix_buffer_count;
-    u32_le sample_rate; // Stored in the aux buffer currently
-    u32_le sample_count;
-    u64_le send_buffer_info;
-    u64_le send_buffer_base;
-
-    u64_le return_buffer_info;
-    u64_le return_buffer_base;
-};
-static_assert(sizeof(AuxInfo) == 0x60, "AuxInfo is an invalid size");
-
-struct EffectInStatus {
-    Effect type;
-    u8 is_new;
-    u8 is_enabled;
-    INSERT_PADDING_BYTES(1);
-    u32_le mix_id;
-    u64_le buffer_base;
-    u64_le buffer_sz;
-    s32_le priority;
-    INSERT_PADDING_BYTES(4);
-    union {
-        std::array<u8, 0xa0> raw;
-        AuxInfo aux_info;
-    };
-};
-static_assert(sizeof(EffectInStatus) == 0xc0, "EffectInStatus is an invalid size");
-
-struct EffectOutStatus {
-    EffectStatus state;
-    INSERT_PADDING_BYTES(0xf);
-};
-static_assert(sizeof(EffectOutStatus) == 0x10, "EffectOutStatus is an invalid size");
-
 struct RendererInfo {
     u64_le elasped_frame_count{};
     INSERT_PADDING_WORDS(2);
 };
 static_assert(sizeof(RendererInfo) == 0x10, "RendererInfo is an invalid size");
 
-struct UpdateDataHeader {
-    UpdateDataHeader() {}
-
-    explicit UpdateDataHeader(const AudioRendererParameter& config) {
-        revision = Common::MakeMagic('R', 'E', 'V', '8'); // 9.2.0 Revision
-        behavior_size = 0xb0;
-        memory_pools_size = (config.effect_count + (config.voice_count * 4)) * 0x10;
-        voices_size = config.voice_count * 0x10;
-        voice_resource_size = 0x0;
-        effects_size = config.effect_count * 0x10;
-        mixes_size = 0x0;
-        sinks_size = config.sink_count * 0x20;
-        performance_manager_size = 0x10;
-        render_info = 0;
-        total_size = sizeof(UpdateDataHeader) + behavior_size + memory_pools_size + voices_size +
-                     effects_size + sinks_size + performance_manager_size;
-    }
-
-    u32_le revision{};
-    u32_le behavior_size{};
-    u32_le memory_pools_size{};
-    u32_le voices_size{};
-    u32_le voice_resource_size{};
-    u32_le effects_size{};
-    u32_le mixes_size{};
-    u32_le sinks_size{};
-    u32_le performance_manager_size{};
-    u32_le splitter_size{};
-    u32_le render_info{};
-    INSERT_PADDING_WORDS(4);
-    u32_le total_size{};
-};
-static_assert(sizeof(UpdateDataHeader) == 0x40, "UpdateDataHeader has wrong size");
-
 class AudioRenderer {
 public:
     AudioRenderer(Core::Timing::CoreTiming& core_timing, Core::Memory::Memory& memory_,
-                  AudioRendererParameter params,
+                  AudioCommon::AudioRendererParameter params,
                   std::shared_ptr<Kernel::WritableEvent> buffer_event, std::size_t instance_number);
     ~AudioRenderer();
 
-    ResultVal<std::vector<u8>> UpdateAudioRenderer(const std::vector<u8>& input_params);
+    ResultCode UpdateAudioRenderer(const std::vector<u8>& input_params,
+                                   std::vector<u8>& output_params);
     void QueueMixedBuffer(Buffer::Tag tag);
     void ReleaseAndQueueBuffers();
     u32 GetSampleRate() const;
@@ -252,19 +63,23 @@ public:
     Stream::State GetStreamState() const;
 
 private:
-    class EffectState;
-    class VoiceState;
     BehaviorInfo behavior_info{};
 
-    AudioRendererParameter worker_params;
+    AudioCommon::AudioRendererParameter worker_params;
     std::shared_ptr<Kernel::WritableEvent> buffer_event;
+    std::vector<ServerMemoryPoolInfo> memory_pool_info;
+    VoiceContext voice_context;
+    EffectContext effect_context;
+    MixContext mix_context;
+    SinkContext sink_context;
+    SplitterContext splitter_context;
     std::vector<VoiceState> voices;
-    std::vector<VoiceResourceInformation> voice_resources;
-    std::vector<EffectState> effects;
     std::unique_ptr<AudioOut> audio_out;
     StreamPtr stream;
     Core::Memory::Memory& memory;
+    CommandGenerator command_generator;
     std::size_t elapsed_frame_count{};
+    std::vector<s32> temp_mix_buffer{};
 };
 
 } // namespace AudioCore

src/audio_core/behavior_info.cpp

@@ -9,39 +9,11 @@
 
 namespace AudioCore {
 
-BehaviorInfo::BehaviorInfo() : process_revision(CURRENT_PROCESS_REVISION) {}
+BehaviorInfo::BehaviorInfo() : process_revision(AudioCommon::CURRENT_PROCESS_REVISION) {}
 BehaviorInfo::~BehaviorInfo() = default;
 
-bool BehaviorInfo::UpdateInput(const std::vector<u8>& buffer, std::size_t offset) {
-    if (!CanConsumeBuffer(buffer.size(), offset, sizeof(InParams))) {
-        LOG_ERROR(Audio, "Buffer is an invalid size!");
-        return false;
-    }
-    InParams params{};
-    std::memcpy(&params, buffer.data() + offset, sizeof(InParams));
-
-    if (!IsValidRevision(params.revision)) {
-        LOG_ERROR(Audio, "Invalid input revision, revision=0x{:08X}", params.revision);
-        return false;
-    }
-
-    if (user_revision != params.revision) {
-        LOG_ERROR(Audio,
-                  "User revision differs from input revision, expecting 0x{:08X} but got 0x{:08X}",
-                  user_revision, params.revision);
-        return false;
-    }
-
-    ClearError();
-    UpdateFlags(params.flags);
-
-    // TODO(ogniK): Check input params size when InfoUpdater is used
-
-    return true;
-}
-
 bool BehaviorInfo::UpdateOutput(std::vector<u8>& buffer, std::size_t offset) {
-    if (!CanConsumeBuffer(buffer.size(), offset, sizeof(OutParams))) {
+    if (!AudioCommon::CanConsumeBuffer(buffer.size(), offset, sizeof(OutParams))) {
         LOG_ERROR(Audio, "Buffer is an invalid size!");
         return false;
     }
@@ -65,36 +37,69 @@ void BehaviorInfo::SetUserRevision(u32_le revision) {
     user_revision = revision;
 }
 
+u32_le BehaviorInfo::GetUserRevision() const {
+    return user_revision;
+}
+
+u32_le BehaviorInfo::GetProcessRevision() const {
+    return process_revision;
+}
+
 bool BehaviorInfo::IsAdpcmLoopContextBugFixed() const {
-    return IsRevisionSupported(2, user_revision);
+    return AudioCommon::IsRevisionSupported(2, user_revision);
 }
 
 bool BehaviorInfo::IsSplitterSupported() const {
-    return IsRevisionSupported(2, user_revision);
+    return AudioCommon::IsRevisionSupported(2, user_revision);
 }
 
 bool BehaviorInfo::IsLongSizePreDelaySupported() const {
-    return IsRevisionSupported(3, user_revision);
+    return AudioCommon::IsRevisionSupported(3, user_revision);
 }
 
-bool BehaviorInfo::IsAudioRenererProcessingTimeLimit80PercentSupported() const {
-    return IsRevisionSupported(5, user_revision);
+bool BehaviorInfo::IsAudioRendererProcessingTimeLimit80PercentSupported() const {
+    return AudioCommon::IsRevisionSupported(5, user_revision);
 }
 
-bool BehaviorInfo::IsAudioRenererProcessingTimeLimit75PercentSupported() const {
-    return IsRevisionSupported(4, user_revision);
+bool BehaviorInfo::IsAudioRendererProcessingTimeLimit75PercentSupported() const {
+    return AudioCommon::IsRevisionSupported(4, user_revision);
 }
 
-bool BehaviorInfo::IsAudioRenererProcessingTimeLimit70PercentSupported() const {
-    return IsRevisionSupported(1, user_revision);
+bool BehaviorInfo::IsAudioRendererProcessingTimeLimit70PercentSupported() const {
+    return AudioCommon::IsRevisionSupported(1, user_revision);
 }
 
 bool BehaviorInfo::IsElapsedFrameCountSupported() const {
-    return IsRevisionSupported(5, user_revision);
+    return AudioCommon::IsRevisionSupported(5, user_revision);
 }
 
 bool BehaviorInfo::IsMemoryPoolForceMappingEnabled() const {
     return (flags & 1) != 0;
 }
 
+bool BehaviorInfo::IsFlushVoiceWaveBuffersSupported() const {
+    return AudioCommon::IsRevisionSupported(5, user_revision);
+}
+
+bool BehaviorInfo::IsVoicePlayedSampleCountResetAtLoopPointSupported() const {
+    return AudioCommon::IsRevisionSupported(5, user_revision);
+}
+
+bool BehaviorInfo::IsVoicePitchAndSrcSkippedSupported() const {
+    return AudioCommon::IsRevisionSupported(5, user_revision);
+}
+
+bool BehaviorInfo::IsMixInParameterDirtyOnlyUpdateSupported() const {
+    return AudioCommon::IsRevisionSupported(7, user_revision);
+}
+
+bool BehaviorInfo::IsSplitterBugFixed() const {
+    return AudioCommon::IsRevisionSupported(5, user_revision);
+}
+
+void BehaviorInfo::CopyErrorInfo(BehaviorInfo::OutParams& dst) {
+    dst.error_count = static_cast<u32>(error_count);
+    std::copy(errors.begin(), errors.begin() + error_count, dst.errors.begin());
+}
+
 } // namespace AudioCore

src/audio_core/behavior_info.h

@@ -14,30 +14,6 @@
 namespace AudioCore {
 class BehaviorInfo {
 public:
-    explicit BehaviorInfo();
-    ~BehaviorInfo();
-
-    bool UpdateInput(const std::vector<u8>& buffer, std::size_t offset);
-    bool UpdateOutput(std::vector<u8>& buffer, std::size_t offset);
-
-    void ClearError();
-    void UpdateFlags(u64_le dest_flags);
-    void SetUserRevision(u32_le revision);
-
-    bool IsAdpcmLoopContextBugFixed() const;
-    bool IsSplitterSupported() const;
-    bool IsLongSizePreDelaySupported() const;
-    bool IsAudioRenererProcessingTimeLimit80PercentSupported() const;
-    bool IsAudioRenererProcessingTimeLimit75PercentSupported() const;
-    bool IsAudioRenererProcessingTimeLimit70PercentSupported() const;
-    bool IsElapsedFrameCountSupported() const;
-    bool IsMemoryPoolForceMappingEnabled() const;
-
-private:
-    u32_le process_revision{};
-    u32_le user_revision{};
-    u64_le flags{};
-
     struct ErrorInfo {
         u32_le result{};
         INSERT_PADDING_WORDS(1);
@@ -45,9 +21,6 @@ private:
     };
     static_assert(sizeof(ErrorInfo) == 0x10, "ErrorInfo is an invalid size");
 
-    std::array<ErrorInfo, 10> errors{};
-    std::size_t error_count{};
-
     struct InParams {
         u32_le revision{};
         u32_le padding{};
@@ -61,6 +34,39 @@ private:
         INSERT_PADDING_BYTES(12);
     };
     static_assert(sizeof(OutParams) == 0xb0, "OutParams is an invalid size");
+
+    explicit BehaviorInfo();
+    ~BehaviorInfo();
+
+    bool UpdateOutput(std::vector<u8>& buffer, std::size_t offset);
+
+    void ClearError();
+    void UpdateFlags(u64_le dest_flags);
+    void SetUserRevision(u32_le revision);
+    u32_le GetUserRevision() const;
+    u32_le GetProcessRevision() const;
+
+    bool IsAdpcmLoopContextBugFixed() const;
+    bool IsSplitterSupported() const;
+    bool IsLongSizePreDelaySupported() const;
+    bool IsAudioRendererProcessingTimeLimit80PercentSupported() const;
+    bool IsAudioRendererProcessingTimeLimit75PercentSupported() const;
+    bool IsAudioRendererProcessingTimeLimit70PercentSupported() const;
+    bool IsElapsedFrameCountSupported() const;
+    bool IsMemoryPoolForceMappingEnabled() const;
+    bool IsFlushVoiceWaveBuffersSupported() const;
+    bool IsVoicePlayedSampleCountResetAtLoopPointSupported() const;
+    bool IsVoicePitchAndSrcSkippedSupported() const;
+    bool IsMixInParameterDirtyOnlyUpdateSupported() const;
+    bool IsSplitterBugFixed() const;
+    void CopyErrorInfo(OutParams& dst);
+
+private:
+    u32_le process_revision{};
+    u32_le user_revision{};
+    u64_le flags{};
+    std::array<ErrorInfo, 10> errors{};
+    std::size_t error_count{};
 };
 
 } // namespace AudioCore

src/audio_core/codec.cpp

@@ -16,8 +16,9 @@ std::vector<s16> DecodeADPCM(const u8* const data, std::size_t size, const ADPCM
 
     constexpr std::size_t FRAME_LEN = 8;
     constexpr std::size_t SAMPLES_PER_FRAME = 14;
-    constexpr std::array<int, 16> SIGNED_NIBBLES = {
-        {0, 1, 2, 3, 4, 5, 6, 7, -8, -7, -6, -5, -4, -3, -2, -1}};
+    static constexpr std::array<int, 16> SIGNED_NIBBLES{
+        0, 1, 2, 3, 4, 5, 6, 7, -8, -7, -6, -5, -4, -3, -2, -1,
+    };
 
     const std::size_t sample_count = (size / FRAME_LEN) * SAMPLES_PER_FRAME;
     const std::size_t ret_size =

src/audio_core/codec.h

@@ -38,7 +38,7 @@ using ADPCM_Coeff = std::array<s16, 16>;
  * @param state ADPCM state, this is updated with new state
  * @return Decoded stereo signed PCM16 data, sample_count in length
  */
-std::vector<s16> DecodeADPCM(const u8* const data, std::size_t size, const ADPCM_Coeff& coeff,
+std::vector<s16> DecodeADPCM(const u8* data, std::size_t size, const ADPCM_Coeff& coeff,
                              ADPCMState& state);
 
 }; // namespace AudioCore::Codec

src/audio_core/command_generator.cpp

@@ -0,0 +1,977 @@
+// Copyright 2020 yuzu Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#include "audio_core/algorithm/interpolate.h"
+#include "audio_core/command_generator.h"
+#include "audio_core/effect_context.h"
+#include "audio_core/mix_context.h"
+#include "audio_core/voice_context.h"
+#include "core/memory.h"
+
+namespace AudioCore {
+namespace {
+constexpr std::size_t MIX_BUFFER_SIZE = 0x3f00;
+constexpr std::size_t SCALED_MIX_BUFFER_SIZE = MIX_BUFFER_SIZE << 15ULL;
+
+template <std::size_t N>
+void ApplyMix(s32* output, const s32* input, s32 gain, s32 sample_count) {
+    for (std::size_t i = 0; i < static_cast<std::size_t>(sample_count); i += N) {
+        for (std::size_t j = 0; j < N; j++) {
+            output[i + j] +=
+                static_cast<s32>((static_cast<s64>(input[i + j]) * gain + 0x4000) >> 15);
+        }
+    }
+}
+
+s32 ApplyMixRamp(s32* output, const s32* input, float gain, float delta, s32 sample_count) {
+    s32 x = 0;
+    for (s32 i = 0; i < sample_count; i++) {
+        x = static_cast<s32>(static_cast<float>(input[i]) * gain);
+        output[i] += x;
+        gain += delta;
+    }
+    return x;
+}
+
+void ApplyGain(s32* output, const s32* input, s32 gain, s32 delta, s32 sample_count) {
+    for (s32 i = 0; i < sample_count; i++) {
+        output[i] = static_cast<s32>((static_cast<s64>(input[i]) * gain + 0x4000) >> 15);
+        gain += delta;
+    }
+}
+
+void ApplyGainWithoutDelta(s32* output, const s32* input, s32 gain, s32 sample_count) {
+    for (s32 i = 0; i < sample_count; i++) {
+        output[i] = static_cast<s32>((static_cast<s64>(input[i]) * gain + 0x4000) >> 15);
+    }
+}
+
+s32 ApplyMixDepop(s32* output, s32 first_sample, s32 delta, s32 sample_count) {
+    const bool positive = first_sample > 0;
+    auto final_sample = std::abs(first_sample);
+    for (s32 i = 0; i < sample_count; i++) {
+        final_sample = static_cast<s32>((static_cast<s64>(final_sample) * delta) >> 15);
+        if (positive) {
+            output[i] += final_sample;
+        } else {
+            output[i] -= final_sample;
+        }
+    }
+    if (positive) {
+        return final_sample;
+    } else {
+        return -final_sample;
+    }
+}
+
+} // namespace
+
+CommandGenerator::CommandGenerator(AudioCommon::AudioRendererParameter& worker_params,
+                                   VoiceContext& voice_context, MixContext& mix_context,
+                                   SplitterContext& splitter_context, EffectContext& effect_context,
+                                   Core::Memory::Memory& memory)
+    : worker_params(worker_params), voice_context(voice_context), mix_context(mix_context),
+      splitter_context(splitter_context), effect_context(effect_context), memory(memory),
+      mix_buffer((worker_params.mix_buffer_count + AudioCommon::MAX_CHANNEL_COUNT) *
+                 worker_params.sample_count),
+      sample_buffer(MIX_BUFFER_SIZE),
+      depop_buffer((worker_params.mix_buffer_count + AudioCommon::MAX_CHANNEL_COUNT) *
+                   worker_params.sample_count) {}
+CommandGenerator::~CommandGenerator() = default;
+
+void CommandGenerator::ClearMixBuffers() {
+    std::fill(mix_buffer.begin(), mix_buffer.end(), 0);
+    std::fill(sample_buffer.begin(), sample_buffer.end(), 0);
+    // std::fill(depop_buffer.begin(), depop_buffer.end(), 0);
+}
+
+void CommandGenerator::GenerateVoiceCommands() {
+    if (dumping_frame) {
+        LOG_DEBUG(Audio, "(DSP_TRACE) GenerateVoiceCommands");
+    }
+    // Grab all our voices
+    const auto voice_count = voice_context.GetVoiceCount();
+    for (std::size_t i = 0; i < voice_count; i++) {
+        auto& voice_info = voice_context.GetSortedInfo(i);
+        // Update voices and check if we should queue them
+        if (voice_info.ShouldSkip() || !voice_info.UpdateForCommandGeneration(voice_context)) {
+            continue;
+        }
+
+        // Queue our voice
+        GenerateVoiceCommand(voice_info);
+    }
+    // Update our splitters
+    splitter_context.UpdateInternalState();
+}
+
+void CommandGenerator::GenerateVoiceCommand(ServerVoiceInfo& voice_info) {
+    auto& in_params = voice_info.GetInParams();
+    const auto channel_count = in_params.channel_count;
+
+    for (s32 channel = 0; channel < channel_count; channel++) {
+        const auto resource_id = in_params.voice_channel_resource_id[channel];
+        auto& dsp_state = voice_context.GetDspSharedState(resource_id);
+        auto& channel_resource = voice_context.GetChannelResource(resource_id);
+
+        // Decode our samples for our channel
+        GenerateDataSourceCommand(voice_info, dsp_state, channel);
+
+        if (in_params.should_depop) {
+            in_params.last_volume = 0.0f;
+        } else if (in_params.splitter_info_id != AudioCommon::NO_SPLITTER ||
+                   in_params.mix_id != AudioCommon::NO_MIX) {
+            // Apply a biquad filter if needed
+            GenerateBiquadFilterCommandForVoice(voice_info, dsp_state,
+                                                worker_params.mix_buffer_count, channel);
+            // Base voice volume ramping
+            GenerateVolumeRampCommand(in_params.last_volume, in_params.volume, channel,
+                                      in_params.node_id);
+            in_params.last_volume = in_params.volume;
+
+            if (in_params.mix_id != AudioCommon::NO_MIX) {
+                // If we're using a mix id
+                auto& mix_info = mix_context.GetInfo(in_params.mix_id);
+                const auto& dest_mix_params = mix_info.GetInParams();
+
+                // Voice Mixing
+                GenerateVoiceMixCommand(
+                    channel_resource.GetCurrentMixVolume(), channel_resource.GetLastMixVolume(),
+                    dsp_state, dest_mix_params.buffer_offset, dest_mix_params.buffer_count,
+                    worker_params.mix_buffer_count + channel, in_params.node_id);
+
+                // Update last mix volumes
+                channel_resource.UpdateLastMixVolumes();
+            } else if (in_params.splitter_info_id != AudioCommon::NO_SPLITTER) {
+                s32 base = channel;
+                while (auto* destination_data =
+                           GetDestinationData(in_params.splitter_info_id, base)) {
+                    base += channel_count;
+
+                    if (!destination_data->IsConfigured()) {
+                        continue;
+                    }
+                    if (destination_data->GetMixId() >= static_cast<int>(mix_context.GetCount())) {
+                        continue;
+                    }
+
+                    const auto& mix_info = mix_context.GetInfo(destination_data->GetMixId());
+                    const auto& dest_mix_params = mix_info.GetInParams();
+                    GenerateVoiceMixCommand(
+                        destination_data->CurrentMixVolumes(), destination_data->LastMixVolumes(),
+                        dsp_state, dest_mix_params.buffer_offset, dest_mix_params.buffer_count,
+                        worker_params.mix_buffer_count + channel, in_params.node_id);
+                    destination_data->MarkDirty();
+                }
+            }
+            // Update biquad filter enabled states
+            for (std::size_t i = 0; i < AudioCommon::MAX_BIQUAD_FILTERS; i++) {
+                in_params.was_biquad_filter_enabled[i] = in_params.biquad_filter[i].enabled;
+            }
+        }
+    }
+}
+
+void CommandGenerator::GenerateSubMixCommands() {
+    const auto mix_count = mix_context.GetCount();
+    for (std::size_t i = 0; i < mix_count; i++) {
+        auto& mix_info = mix_context.GetSortedInfo(i);
+        const auto& in_params = mix_info.GetInParams();
+        if (!in_params.in_use || in_params.mix_id == AudioCommon::FINAL_MIX) {
+            continue;
+        }
+        GenerateSubMixCommand(mix_info);
+    }
+}
+
+void CommandGenerator::GenerateFinalMixCommands() {
+    GenerateFinalMixCommand();
+}
+
+void CommandGenerator::PreCommand() {
+    if (!dumping_frame) {
+        return;
+    }
+    for (std::size_t i = 0; i < splitter_context.GetInfoCount(); i++) {
+        const auto& base = splitter_context.GetInfo(i);
+        std::string graph = fmt::format("b[{}]", i);
+        const auto* head = base.GetHead();
+        while (head != nullptr) {
+            graph += fmt::format("->{}", head->GetMixId());
+            head = head->GetNextDestination();
+        }
+        LOG_DEBUG(Audio, "(DSP_TRACE) SplitterGraph splitter_info={}, {}", i, graph);
+    }
+}
+
+void CommandGenerator::PostCommand() {
+    if (!dumping_frame) {
+        return;
+    }
+    dumping_frame = false;
+}
+
+void CommandGenerator::GenerateDataSourceCommand(ServerVoiceInfo& voice_info, VoiceState& dsp_state,
+                                                 s32 channel) {
+    const auto& in_params = voice_info.GetInParams();
+    const auto depop = in_params.should_depop;
+
+    if (depop) {
+        if (in_params.mix_id != AudioCommon::NO_MIX) {
+            auto& mix_info = mix_context.GetInfo(in_params.mix_id);
+            const auto& mix_in = mix_info.GetInParams();
+            GenerateDepopPrepareCommand(dsp_state, mix_in.buffer_count, mix_in.buffer_offset);
+        } else if (in_params.splitter_info_id != AudioCommon::NO_SPLITTER) {
+            s32 index{};
+            while (const auto* destination =
+                       GetDestinationData(in_params.splitter_info_id, index++)) {
+                if (!destination->IsConfigured()) {
+                    continue;
+                }
+                auto& mix_info = mix_context.GetInfo(destination->GetMixId());
+                const auto& mix_in = mix_info.GetInParams();
+                GenerateDepopPrepareCommand(dsp_state, mix_in.buffer_count, mix_in.buffer_offset);
+            }
+        }
+    } else {
+        switch (in_params.sample_format) {
+        case SampleFormat::Pcm16:
+            DecodeFromWaveBuffers(voice_info, GetChannelMixBuffer(channel), dsp_state, channel,
+                                  worker_params.sample_rate, worker_params.sample_count,
+                                  in_params.node_id);
+            break;
+        case SampleFormat::Adpcm:
+            ASSERT(channel == 0 && in_params.channel_count == 1);
+            DecodeFromWaveBuffers(voice_info, GetChannelMixBuffer(0), dsp_state, 0,
+                                  worker_params.sample_rate, worker_params.sample_count,
+                                  in_params.node_id);
+            break;
+        default:
+            UNREACHABLE_MSG("Unimplemented sample format={}", in_params.sample_format);
+        }
+    }
+}
+
+void CommandGenerator::GenerateBiquadFilterCommandForVoice(ServerVoiceInfo& voice_info,
+                                                           VoiceState& dsp_state,
+                                                           s32 mix_buffer_count, s32 channel) {
+    for (std::size_t i = 0; i < AudioCommon::MAX_BIQUAD_FILTERS; i++) {
+        const auto& in_params = voice_info.GetInParams();
+        auto& biquad_filter = in_params.biquad_filter[i];
+        // Check if biquad filter is actually used
+        if (!biquad_filter.enabled) {
+            continue;
+        }
+
+        // Reinitialize our biquad filter state if it was enabled previously
+        if (!in_params.was_biquad_filter_enabled[i]) {
+            dsp_state.biquad_filter_state.fill(0);
+        }
+
+        // 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);
+    }
+}
+
+void AudioCore::CommandGenerator::GenerateBiquadFilterCommand(
+    s32 mix_buffer, const BiquadFilterParameter& params, std::array<s64, 2>& state,
+    std::size_t input_offset, std::size_t output_offset, s32 sample_count, s32 node_id) {
+    if (dumping_frame) {
+        LOG_DEBUG(Audio,
+                  "(DSP_TRACE) GenerateBiquadFilterCommand node_id={}, "
+                  "input_mix_buffer={}, output_mix_buffer={}",
+                  node_id, input_offset, output_offset);
+    }
+    const auto* input = GetMixBuffer(input_offset);
+    auto* output = GetMixBuffer(output_offset);
+
+    // Biquad filter parameters
+    const auto [n0, n1, n2] = params.numerator;
+    const auto [d0, d1] = params.denominator;
+
+    // Biquad filter states
+    auto [s0, s1] = state;
+
+    constexpr s64 int32_min = std::numeric_limits<s32>::min();
+    constexpr s64 int32_max = std::numeric_limits<s32>::max();
+
+    for (int i = 0; i < sample_count; ++i) {
+        const auto sample = static_cast<s64>(input[i]);
+        const auto f = (sample * n0 + s0 + 0x4000) >> 15;
+        const auto y = std::clamp(f, int32_min, int32_max);
+        s0 = sample * n1 + y * d0 + s1;
+        s1 = sample * n2 + y * d1;
+        output[i] = static_cast<s32>(y);
+    }
+
+    state = {s0, s1};
+}
+
+void CommandGenerator::GenerateDepopPrepareCommand(VoiceState& dsp_state,
+                                                   std::size_t mix_buffer_count,
+                                                   std::size_t mix_buffer_offset) {
+    for (std::size_t i = 0; i < mix_buffer_count; i++) {
+        auto& sample = dsp_state.previous_samples[i];
+        if (sample != 0) {
+            depop_buffer[mix_buffer_offset + i] += sample;
+            sample = 0;
+        }
+    }
+}
+
+void CommandGenerator::GenerateDepopForMixBuffersCommand(std::size_t mix_buffer_count,
+                                                         std::size_t mix_buffer_offset,
+                                                         s32 sample_rate) {
+    const std::size_t end_offset =
+        std::min(mix_buffer_offset + mix_buffer_count, GetTotalMixBufferCount());
+    const s32 delta = sample_rate == 48000 ? 0x7B29 : 0x78CB;
+    for (std::size_t i = mix_buffer_offset; i < end_offset; i++) {
+        if (depop_buffer[i] == 0) {
+            continue;
+        }
+
+        depop_buffer[i] =
+            ApplyMixDepop(GetMixBuffer(i), depop_buffer[i], delta, worker_params.sample_count);
+    }
+}
+
+void CommandGenerator::GenerateEffectCommand(ServerMixInfo& mix_info) {
+    const std::size_t effect_count = effect_context.GetCount();
+    const auto buffer_offset = mix_info.GetInParams().buffer_offset;
+    for (std::size_t i = 0; i < effect_count; i++) {
+        const auto index = mix_info.GetEffectOrder(i);
+        if (index == AudioCommon::NO_EFFECT_ORDER) {
+            break;
+        }
+        auto* info = effect_context.GetInfo(index);
+        const auto type = info->GetType();
+
+        // TODO(ogniK): Finish remaining effects
+        switch (type) {
+        case EffectType::Aux:
+            GenerateAuxCommand(buffer_offset, info, info->IsEnabled());
+            break;
+        case EffectType::I3dl2Reverb:
+            GenerateI3dl2ReverbEffectCommand(buffer_offset, info, info->IsEnabled());
+            break;
+        case EffectType::BiquadFilter:
+            GenerateBiquadFilterEffectCommand(buffer_offset, info, info->IsEnabled());
+            break;
+        default:
+            break;
+        }
+
+        info->UpdateForCommandGeneration();
+    }
+}
+
+void CommandGenerator::GenerateI3dl2ReverbEffectCommand(s32 mix_buffer_offset, EffectBase* info,
+                                                        bool enabled) {
+    if (!enabled) {
+        return;
+    }
+    const auto& params = dynamic_cast<EffectI3dl2Reverb*>(info)->GetParams();
+    const auto channel_count = params.channel_count;
+    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);
+        }
+    }
+}
+
+void CommandGenerator::GenerateBiquadFilterEffectCommand(s32 mix_buffer_offset, EffectBase* info,
+                                                         bool enabled) {
+    if (!enabled) {
+        return;
+    }
+    const auto& params = dynamic_cast<EffectBiquadFilter*>(info)->GetParams();
+    const auto channel_count = params.channel_count;
+    for (s32 i = 0; i < channel_count; i++) {
+        // TODO(ogniK): Actually implement biquad filter
+        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);
+        }
+    }
+}
+
+void CommandGenerator::GenerateAuxCommand(s32 mix_buffer_offset, EffectBase* info, bool enabled) {
+    auto* aux = dynamic_cast<EffectAuxInfo*>(info);
+    const auto& params = aux->GetParams();
+    if (aux->GetSendBuffer() != 0 && aux->GetRecvBuffer() != 0) {
+        const auto max_channels = params.count;
+        u32 offset{};
+        for (u32 channel = 0; channel < max_channels; channel++) {
+            u32 write_count = 0;
+            if (channel == (max_channels - 1)) {
+                write_count = offset + worker_params.sample_count;
+            }
+
+            const auto input_index = params.input_mix_buffers[channel] + mix_buffer_offset;
+            const auto output_index = params.output_mix_buffers[channel] + mix_buffer_offset;
+
+            if (enabled) {
+                AuxInfoDSP send_info{};
+                AuxInfoDSP recv_info{};
+                memory.ReadBlock(aux->GetSendInfo(), &send_info, sizeof(AuxInfoDSP));
+                memory.ReadBlock(aux->GetRecvInfo(), &recv_info, sizeof(AuxInfoDSP));
+
+                WriteAuxBuffer(send_info, aux->GetSendBuffer(), params.sample_count,
+                               GetMixBuffer(input_index), worker_params.sample_count, offset,
+                               write_count);
+                memory.WriteBlock(aux->GetSendInfo(), &send_info, sizeof(AuxInfoDSP));
+
+                const auto samples_read = ReadAuxBuffer(
+                    recv_info, aux->GetRecvBuffer(), params.sample_count,
+                    GetMixBuffer(output_index), worker_params.sample_count, offset, write_count);
+                memory.WriteBlock(aux->GetRecvInfo(), &recv_info, sizeof(AuxInfoDSP));
+
+                if (samples_read != static_cast<int>(worker_params.sample_count) &&
+                    samples_read <= params.sample_count) {
+                    std::memset(GetMixBuffer(output_index), 0, params.sample_count - samples_read);
+                }
+            } else {
+                AuxInfoDSP empty{};
+                memory.WriteBlock(aux->GetSendInfo(), &empty, sizeof(AuxInfoDSP));
+                memory.WriteBlock(aux->GetRecvInfo(), &empty, sizeof(AuxInfoDSP));
+                if (output_index != input_index) {
+                    std::memcpy(GetMixBuffer(output_index), GetMixBuffer(input_index),
+                                worker_params.sample_count * sizeof(s32));
+                }
+            }
+
+            offset += worker_params.sample_count;
+        }
+    }
+}
+
+ServerSplitterDestinationData* CommandGenerator::GetDestinationData(s32 splitter_id, s32 index) {
+    if (splitter_id == AudioCommon::NO_SPLITTER) {
+        return nullptr;
+    }
+    return splitter_context.GetDestinationData(splitter_id, index);
+}
+
+s32 CommandGenerator::WriteAuxBuffer(AuxInfoDSP& dsp_info, VAddr send_buffer, u32 max_samples,
+                                     const s32* data, u32 sample_count, u32 write_offset,
+                                     u32 write_count) {
+    if (max_samples == 0) {
+        return 0;
+    }
+    u32 offset = dsp_info.write_offset + write_offset;
+    if (send_buffer == 0 || offset > max_samples) {
+        return 0;
+    }
+
+    std::size_t data_offset{};
+    u32 remaining = sample_count;
+    while (remaining > 0) {
+        // Get position in buffer
+        const auto base = send_buffer + (offset * sizeof(u32));
+        const auto samples_to_grab = std::min(max_samples - offset, remaining);
+        // Write to output
+        memory.WriteBlock(base, (data + data_offset), samples_to_grab * sizeof(u32));
+        offset = (offset + samples_to_grab) % max_samples;
+        remaining -= samples_to_grab;
+        data_offset += samples_to_grab;
+    }
+
+    if (write_count != 0) {
+        dsp_info.write_offset = (dsp_info.write_offset + write_count) % max_samples;
+    }
+    return sample_count;
+}
+
+s32 CommandGenerator::ReadAuxBuffer(AuxInfoDSP& recv_info, VAddr recv_buffer, u32 max_samples,
+                                    s32* out_data, u32 sample_count, u32 read_offset,
+                                    u32 read_count) {
+    if (max_samples == 0) {
+        return 0;
+    }
+
+    u32 offset = recv_info.read_offset + read_offset;
+    if (recv_buffer == 0 || offset > max_samples) {
+        return 0;
+    }
+
+    u32 remaining = sample_count;
+    while (remaining > 0) {
+        const auto base = recv_buffer + (offset * sizeof(u32));
+        const auto samples_to_grab = std::min(max_samples - offset, remaining);
+        std::vector<s32> buffer(samples_to_grab);
+        memory.ReadBlock(base, buffer.data(), buffer.size() * sizeof(u32));
+        std::memcpy(out_data, buffer.data(), buffer.size() * sizeof(u32));
+        out_data += samples_to_grab;
+        offset = (offset + samples_to_grab) % max_samples;
+        remaining -= samples_to_grab;
+    }
+
+    if (read_count != 0) {
+        recv_info.read_offset = (recv_info.read_offset + read_count) % max_samples;
+    }
+    return sample_count;
+}
+
+void CommandGenerator::GenerateVolumeRampCommand(float last_volume, float current_volume,
+                                                 s32 channel, s32 node_id) {
+    const auto last = static_cast<s32>(last_volume * 32768.0f);
+    const auto current = static_cast<s32>(current_volume * 32768.0f);
+    const auto delta = static_cast<s32>((static_cast<float>(current) - static_cast<float>(last)) /
+                                        static_cast<float>(worker_params.sample_count));
+
+    if (dumping_frame) {
+        LOG_DEBUG(Audio,
+                  "(DSP_TRACE) GenerateVolumeRampCommand node_id={}, input={}, output={}, "
+                  "last_volume={}, current_volume={}",
+                  node_id, GetMixChannelBufferOffset(channel), GetMixChannelBufferOffset(channel),
+                  last_volume, current_volume);
+    }
+    // Apply generic gain on samples
+    ApplyGain(GetChannelMixBuffer(channel), GetChannelMixBuffer(channel), last, delta,
+              worker_params.sample_count);
+}
+
+void CommandGenerator::GenerateVoiceMixCommand(const MixVolumeBuffer& mix_volumes,
+                                               const MixVolumeBuffer& last_mix_volumes,
+                                               VoiceState& dsp_state, s32 mix_buffer_offset,
+                                               s32 mix_buffer_count, s32 voice_index, s32 node_id) {
+    // Loop all our mix buffers
+    for (s32 i = 0; i < mix_buffer_count; i++) {
+        if (last_mix_volumes[i] != 0.0f || mix_volumes[i] != 0.0f) {
+            const auto delta = static_cast<float>((mix_volumes[i] - last_mix_volumes[i])) /
+                               static_cast<float>(worker_params.sample_count);
+
+            if (dumping_frame) {
+                LOG_DEBUG(Audio,
+                          "(DSP_TRACE) GenerateVoiceMixCommand node_id={}, input={}, "
+                          "output={}, last_volume={}, current_volume={}",
+                          node_id, voice_index, mix_buffer_offset + i, last_mix_volumes[i],
+                          mix_volumes[i]);
+            }
+
+            dsp_state.previous_samples[i] =
+                ApplyMixRamp(GetMixBuffer(mix_buffer_offset + i), GetMixBuffer(voice_index),
+                             last_mix_volumes[i], delta, worker_params.sample_count);
+        } else {
+            dsp_state.previous_samples[i] = 0;
+        }
+    }
+}
+
+void CommandGenerator::GenerateSubMixCommand(ServerMixInfo& mix_info) {
+    if (dumping_frame) {
+        LOG_DEBUG(Audio, "(DSP_TRACE) GenerateSubMixCommand");
+    }
+    const auto& in_params = mix_info.GetInParams();
+    GenerateDepopForMixBuffersCommand(in_params.buffer_count, in_params.buffer_offset,
+                                      in_params.sample_rate);
+
+    GenerateEffectCommand(mix_info);
+
+    GenerateMixCommands(mix_info);
+}
+
+void CommandGenerator::GenerateMixCommands(ServerMixInfo& mix_info) {
+    if (!mix_info.HasAnyConnection()) {
+        return;
+    }
+    const auto& in_params = mix_info.GetInParams();
+    if (in_params.dest_mix_id != AudioCommon::NO_MIX) {
+        const auto& dest_mix = mix_context.GetInfo(in_params.dest_mix_id);
+        const auto& dest_in_params = dest_mix.GetInParams();
+
+        const auto buffer_count = in_params.buffer_count;
+
+        for (s32 i = 0; i < buffer_count; i++) {
+            for (s32 j = 0; j < dest_in_params.buffer_count; j++) {
+                const auto mixed_volume = in_params.volume * in_params.mix_volume[i][j];
+                if (mixed_volume != 0.0f) {
+                    GenerateMixCommand(dest_in_params.buffer_offset + j,
+                                       in_params.buffer_offset + i, mixed_volume,
+                                       in_params.node_id);
+                }
+            }
+        }
+    } else if (in_params.splitter_id != AudioCommon::NO_SPLITTER) {
+        s32 base{};
+        while (const auto* destination_data = GetDestinationData(in_params.splitter_id, base++)) {
+            if (!destination_data->IsConfigured()) {
+                continue;
+            }
+
+            const auto& dest_mix = mix_context.GetInfo(destination_data->GetMixId());
+            const auto& dest_in_params = dest_mix.GetInParams();
+            const auto mix_index = (base - 1) % in_params.buffer_count + in_params.buffer_offset;
+            for (std::size_t i = 0; i < static_cast<std::size_t>(dest_in_params.buffer_count);
+                 i++) {
+                const auto mixed_volume = in_params.volume * destination_data->GetMixVolume(i);
+                if (mixed_volume != 0.0f) {
+                    GenerateMixCommand(dest_in_params.buffer_offset + i, mix_index, mixed_volume,
+                                       in_params.node_id);
+                }
+            }
+        }
+    }
+}
+
+void CommandGenerator::GenerateMixCommand(std::size_t output_offset, std::size_t input_offset,
+                                          float volume, s32 node_id) {
+
+    if (dumping_frame) {
+        LOG_DEBUG(Audio,
+                  "(DSP_TRACE) GenerateMixCommand node_id={}, input={}, output={}, volume={}",
+                  node_id, input_offset, output_offset, volume);
+    }
+
+    auto* output = GetMixBuffer(output_offset);
+    const auto* input = GetMixBuffer(input_offset);
+
+    const s32 gain = static_cast<s32>(volume * 32768.0f);
+    // Mix with loop unrolling
+    if (worker_params.sample_count % 4 == 0) {
+        ApplyMix<4>(output, input, gain, worker_params.sample_count);
+    } else if (worker_params.sample_count % 2 == 0) {
+        ApplyMix<2>(output, input, gain, worker_params.sample_count);
+    } else {
+        ApplyMix<1>(output, input, gain, worker_params.sample_count);
+    }
+}
+
+void CommandGenerator::GenerateFinalMixCommand() {
+    if (dumping_frame) {
+        LOG_DEBUG(Audio, "(DSP_TRACE) GenerateFinalMixCommand");
+    }
+    auto& mix_info = mix_context.GetFinalMixInfo();
+    const auto& in_params = mix_info.GetInParams();
+
+    GenerateDepopForMixBuffersCommand(in_params.buffer_count, in_params.buffer_offset,
+                                      in_params.sample_rate);
+
+    GenerateEffectCommand(mix_info);
+
+    for (s32 i = 0; i < in_params.buffer_count; i++) {
+        const s32 gain = static_cast<s32>(in_params.volume * 32768.0f);
+        if (dumping_frame) {
+            LOG_DEBUG(
+                Audio,
+                "(DSP_TRACE) ApplyGainWithoutDelta node_id={}, input={}, output={}, volume={}",
+                in_params.node_id, in_params.buffer_offset + i, in_params.buffer_offset + i,
+                in_params.volume);
+        }
+        ApplyGainWithoutDelta(GetMixBuffer(in_params.buffer_offset + i),
+                              GetMixBuffer(in_params.buffer_offset + i), gain,
+                              worker_params.sample_count);
+    }
+}
+
+s32 CommandGenerator::DecodePcm16(ServerVoiceInfo& voice_info, VoiceState& dsp_state,
+                                  s32 sample_count, s32 channel, std::size_t mix_offset) {
+    const auto& in_params = voice_info.GetInParams();
+    const auto& wave_buffer = in_params.wave_buffer[dsp_state.wave_buffer_index];
+    if (wave_buffer.buffer_address == 0) {
+        return 0;
+    }
+    if (wave_buffer.buffer_size == 0) {
+        return 0;
+    }
+    if (wave_buffer.end_sample_offset < wave_buffer.start_sample_offset) {
+        return 0;
+    }
+    const auto samples_remaining =
+        (wave_buffer.end_sample_offset - wave_buffer.start_sample_offset) - dsp_state.offset;
+    const auto start_offset =
+        ((wave_buffer.start_sample_offset + dsp_state.offset) * in_params.channel_count) *
+        sizeof(s16);
+    const auto buffer_pos = wave_buffer.buffer_address + start_offset;
+    const auto samples_processed = std::min(sample_count, samples_remaining);
+
+    if (in_params.channel_count == 1) {
+        std::vector<s16> buffer(samples_processed);
+        memory.ReadBlock(buffer_pos, buffer.data(), buffer.size() * sizeof(s16));
+        for (std::size_t i = 0; i < buffer.size(); i++) {
+            sample_buffer[mix_offset + i] = buffer[i];
+        }
+    } else {
+        const auto channel_count = in_params.channel_count;
+        std::vector<s16> buffer(samples_processed * channel_count);
+        memory.ReadBlock(buffer_pos, buffer.data(), buffer.size() * sizeof(s16));
+
+        for (std::size_t i = 0; i < static_cast<std::size_t>(samples_processed); i++) {
+            sample_buffer[mix_offset + i] = buffer[i * channel_count + channel];
+        }
+    }
+
+    return samples_processed;
+}
+
+s32 CommandGenerator::DecodeAdpcm(ServerVoiceInfo& voice_info, VoiceState& dsp_state,
+                                  s32 sample_count, s32 channel, std::size_t mix_offset) {
+    const auto& in_params = voice_info.GetInParams();
+    const auto& wave_buffer = in_params.wave_buffer[dsp_state.wave_buffer_index];
+    if (wave_buffer.buffer_address == 0) {
+        return 0;
+    }
+    if (wave_buffer.buffer_size == 0) {
+        return 0;
+    }
+    if (wave_buffer.end_sample_offset < wave_buffer.start_sample_offset) {
+        return 0;
+    }
+
+    static constexpr std::array<int, 16> SIGNED_NIBBLES{
+        0, 1, 2, 3, 4, 5, 6, 7, -8, -7, -6, -5, -4, -3, -2, -1,
+    };
+
+    constexpr std::size_t FRAME_LEN = 8;
+    constexpr std::size_t NIBBLES_PER_SAMPLE = 16;
+    constexpr std::size_t SAMPLES_PER_FRAME = 14;
+
+    auto frame_header = dsp_state.context.header;
+    s32 idx = (frame_header >> 4) & 0xf;
+    s32 scale = frame_header & 0xf;
+    s16 yn1 = dsp_state.context.yn1;
+    s16 yn2 = dsp_state.context.yn2;
+
+    Codec::ADPCM_Coeff coeffs;
+    memory.ReadBlock(in_params.additional_params_address, coeffs.data(),
+                     sizeof(Codec::ADPCM_Coeff));
+
+    s32 coef1 = coeffs[idx * 2];
+    s32 coef2 = coeffs[idx * 2 + 1];
+
+    const auto samples_remaining =
+        (wave_buffer.end_sample_offset - wave_buffer.start_sample_offset) - dsp_state.offset;
+    const auto samples_processed = std::min(sample_count, samples_remaining);
+    const auto sample_pos = wave_buffer.start_sample_offset + dsp_state.offset;
+
+    const auto samples_remaining_in_frame = sample_pos % SAMPLES_PER_FRAME;
+    auto position_in_frame = ((sample_pos / SAMPLES_PER_FRAME) * NIBBLES_PER_SAMPLE) +
+                             samples_remaining_in_frame + (samples_remaining_in_frame != 0 ? 2 : 0);
+
+    const auto decode_sample = [&](const int nibble) -> s16 {
+        const int xn = nibble * (1 << scale);
+        // We first transform everything into 11 bit fixed point, perform the second order
+        // digital filter, then transform back.
+        // 0x400 == 0.5 in 11 bit fixed point.
+        // Filter: y[n] = x[n] + 0.5 + c1 * y[n-1] + c2 * y[n-2]
+        int val = ((xn << 11) + 0x400 + coef1 * yn1 + coef2 * yn2) >> 11;
+        // Clamp to output range.
+        val = std::clamp<s32>(val, -32768, 32767);
+        // Advance output feedback.
+        yn2 = yn1;
+        yn1 = static_cast<s16>(val);
+        return yn1;
+    };
+
+    std::size_t buffer_offset{};
+    std::vector<u8> buffer(
+        std::max((samples_processed / FRAME_LEN) * SAMPLES_PER_FRAME, FRAME_LEN));
+    memory.ReadBlock(wave_buffer.buffer_address + (position_in_frame / 2), buffer.data(),
+                     buffer.size());
+    std::size_t cur_mix_offset = mix_offset;
+
+    auto remaining_samples = samples_processed;
+    while (remaining_samples > 0) {
+        if (position_in_frame % NIBBLES_PER_SAMPLE == 0) {
+            // Read header
+            frame_header = buffer[buffer_offset++];
+            idx = (frame_header >> 4) & 0xf;
+            scale = frame_header & 0xf;
+            coef1 = coeffs[idx * 2];
+            coef2 = coeffs[idx * 2 + 1];
+            position_in_frame += 2;
+
+            // Decode entire frame
+            if (remaining_samples >= static_cast<int>(SAMPLES_PER_FRAME)) {
+                for (std::size_t i = 0; i < SAMPLES_PER_FRAME / 2; i++) {
+                    // Sample 1
+                    const s32 s0 = SIGNED_NIBBLES[buffer[buffer_offset] >> 4];
+                    const s32 s1 = SIGNED_NIBBLES[buffer[buffer_offset++] & 0xf];
+                    const s16 sample_1 = decode_sample(s0);
+                    const s16 sample_2 = decode_sample(s1);
+                    sample_buffer[cur_mix_offset++] = sample_1;
+                    sample_buffer[cur_mix_offset++] = sample_2;
+                }
+                remaining_samples -= static_cast<int>(SAMPLES_PER_FRAME);
+                position_in_frame += SAMPLES_PER_FRAME;
+                continue;
+            }
+        }
+        // Decode mid frame
+        s32 current_nibble = buffer[buffer_offset];
+        if (position_in_frame++ & 0x1) {
+            current_nibble &= 0xf;
+            buffer_offset++;
+        } else {
+            current_nibble >>= 4;
+        }
+        const s16 sample = decode_sample(SIGNED_NIBBLES[current_nibble]);
+        sample_buffer[cur_mix_offset++] = sample;
+        remaining_samples--;
+    }
+
+    dsp_state.context.header = frame_header;
+    dsp_state.context.yn1 = yn1;
+    dsp_state.context.yn2 = yn2;
+
+    return samples_processed;
+}
+
+s32* CommandGenerator::GetMixBuffer(std::size_t index) {
+    return mix_buffer.data() + (index * worker_params.sample_count);
+}
+
+const s32* CommandGenerator::GetMixBuffer(std::size_t index) const {
+    return mix_buffer.data() + (index * worker_params.sample_count);
+}
+
+std::size_t CommandGenerator::GetMixChannelBufferOffset(s32 channel) const {
+    return worker_params.mix_buffer_count + channel;
+}
+
+std::size_t CommandGenerator::GetTotalMixBufferCount() const {
+    return worker_params.mix_buffer_count + AudioCommon::MAX_CHANNEL_COUNT;
+}
+
+s32* CommandGenerator::GetChannelMixBuffer(s32 channel) {
+    return GetMixBuffer(worker_params.mix_buffer_count + channel);
+}
+
+const s32* CommandGenerator::GetChannelMixBuffer(s32 channel) const {
+    return GetMixBuffer(worker_params.mix_buffer_count + channel);
+}
+
+void CommandGenerator::DecodeFromWaveBuffers(ServerVoiceInfo& voice_info, s32* output,
+                                             VoiceState& dsp_state, s32 channel,
+                                             s32 target_sample_rate, s32 sample_count,
+                                             s32 node_id) {
+    const auto& in_params = voice_info.GetInParams();
+    if (dumping_frame) {
+        LOG_DEBUG(Audio,
+                  "(DSP_TRACE) DecodeFromWaveBuffers, node_id={}, channel={}, "
+                  "format={}, sample_count={}, sample_rate={}, mix_id={}, splitter_id={}",
+                  node_id, channel, in_params.sample_format, sample_count, in_params.sample_rate,
+                  in_params.mix_id, in_params.splitter_info_id);
+    }
+    ASSERT_OR_EXECUTE(output != nullptr, { return; });
+
+    const auto resample_rate = static_cast<s32>(
+        static_cast<float>(in_params.sample_rate) / static_cast<float>(target_sample_rate) *
+        static_cast<float>(static_cast<s32>(in_params.pitch * 32768.0f)));
+    if (dsp_state.fraction + sample_count * resample_rate >
+        static_cast<s32>(SCALED_MIX_BUFFER_SIZE - 4ULL)) {
+        return;
+    }
+
+    auto min_required_samples =
+        std::min(static_cast<s32>(SCALED_MIX_BUFFER_SIZE) - dsp_state.fraction, resample_rate);
+    if (min_required_samples >= sample_count) {
+        min_required_samples = sample_count;
+    }
+
+    std::size_t temp_mix_offset{};
+    bool is_buffer_completed{false};
+    auto samples_remaining = sample_count;
+    while (samples_remaining > 0 && !is_buffer_completed) {
+        const auto samples_to_output = std::min(samples_remaining, min_required_samples);
+        const auto samples_to_read = (samples_to_output * resample_rate + dsp_state.fraction) >> 15;
+
+        if (!in_params.behavior_flags.is_pitch_and_src_skipped) {
+            // Append sample histtory for resampler
+            for (std::size_t i = 0; i < AudioCommon::MAX_SAMPLE_HISTORY; i++) {
+                sample_buffer[temp_mix_offset + i] = dsp_state.sample_history[i];
+            }
+            temp_mix_offset += 4;
+        }
+
+        s32 samples_read{};
+        while (samples_read < samples_to_read) {
+            const auto& wave_buffer = in_params.wave_buffer[dsp_state.wave_buffer_index];
+            // No more data can be read
+            if (!dsp_state.is_wave_buffer_valid[dsp_state.wave_buffer_index]) {
+                is_buffer_completed = true;
+                break;
+            }
+
+            if (in_params.sample_format == SampleFormat::Adpcm && dsp_state.offset == 0 &&
+                wave_buffer.context_address != 0 && wave_buffer.context_size != 0) {
+                // TODO(ogniK): ADPCM loop context
+            }
+
+            s32 samples_decoded{0};
+            switch (in_params.sample_format) {
+            case SampleFormat::Pcm16:
+                samples_decoded = DecodePcm16(voice_info, dsp_state, samples_to_read - samples_read,
+                                              channel, temp_mix_offset);
+                break;
+            case SampleFormat::Adpcm:
+                samples_decoded = DecodeAdpcm(voice_info, dsp_state, samples_to_read - samples_read,
+                                              channel, temp_mix_offset);
+                break;
+            default:
+                UNREACHABLE_MSG("Unimplemented sample format={}", in_params.sample_format);
+            }
+
+            temp_mix_offset += samples_decoded;
+            samples_read += samples_decoded;
+            dsp_state.offset += samples_decoded;
+            dsp_state.played_sample_count += samples_decoded;
+
+            if (dsp_state.offset >=
+                    (wave_buffer.end_sample_offset - wave_buffer.start_sample_offset) ||
+                samples_decoded == 0) {
+                // Reset our sample offset
+                dsp_state.offset = 0;
+                if (wave_buffer.is_looping) {
+                    if (samples_decoded == 0) {
+                        // End of our buffer
+                        is_buffer_completed = true;
+                        break;
+                    }
+
+                    if (in_params.behavior_flags.is_played_samples_reset_at_loop_point.Value()) {
+                        dsp_state.played_sample_count = 0;
+                    }
+                } else {
+
+                    // Update our wave buffer states
+                    dsp_state.is_wave_buffer_valid[dsp_state.wave_buffer_index] = false;
+                    dsp_state.wave_buffer_consumed++;
+                    dsp_state.wave_buffer_index =
+                        (dsp_state.wave_buffer_index + 1) % AudioCommon::MAX_WAVE_BUFFERS;
+                    if (wave_buffer.end_of_stream) {
+                        dsp_state.played_sample_count = 0;
+                    }
+                }
+            }
+        }
+
+        if (in_params.behavior_flags.is_pitch_and_src_skipped.Value()) {
+            // No need to resample
+            std::memcpy(output, sample_buffer.data(), samples_read * sizeof(s32));
+        } else {
+            std::fill(sample_buffer.begin() + temp_mix_offset,
+                      sample_buffer.begin() + temp_mix_offset + (samples_to_read - samples_read),
+                      0);
+            AudioCore::Resample(output, sample_buffer.data(), resample_rate, dsp_state.fraction,
+                                samples_to_output);
+            // Resample
+            for (std::size_t i = 0; i < AudioCommon::MAX_SAMPLE_HISTORY; i++) {
+                dsp_state.sample_history[i] = sample_buffer[samples_to_read + i];
+            }
+        }
+        output += samples_to_output;
+        samples_remaining -= samples_to_output;
+    }
+}
+
+} // namespace AudioCore

src/audio_core/command_generator.h

@@ -0,0 +1,102 @@
+// Copyright 2020 yuzu Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#pragma once
+
+#include <array>
+#include "audio_core/common.h"
+#include "audio_core/voice_context.h"
+#include "common/common_types.h"
+
+namespace Core::Memory {
+class Memory;
+}
+
+namespace AudioCore {
+class MixContext;
+class SplitterContext;
+class ServerSplitterDestinationData;
+class ServerMixInfo;
+class EffectContext;
+class EffectBase;
+struct AuxInfoDSP;
+using MixVolumeBuffer = std::array<float, AudioCommon::MAX_MIX_BUFFERS>;
+
+class CommandGenerator {
+public:
+    explicit CommandGenerator(AudioCommon::AudioRendererParameter& worker_params,
+                              VoiceContext& voice_context, MixContext& mix_context,
+                              SplitterContext& splitter_context, EffectContext& effect_context,
+                              Core::Memory::Memory& memory);
+    ~CommandGenerator();
+
+    void ClearMixBuffers();
+    void GenerateVoiceCommands();
+    void GenerateVoiceCommand(ServerVoiceInfo& voice_info);
+    void GenerateSubMixCommands();
+    void GenerateFinalMixCommands();
+    void PreCommand();
+    void PostCommand();
+
+    s32* GetChannelMixBuffer(s32 channel);
+    const s32* GetChannelMixBuffer(s32 channel) const;
+    s32* GetMixBuffer(std::size_t index);
+    const s32* GetMixBuffer(std::size_t index) const;
+    std::size_t GetMixChannelBufferOffset(s32 channel) const;
+
+    std::size_t GetTotalMixBufferCount() const;
+
+private:
+    void GenerateDataSourceCommand(ServerVoiceInfo& voice_info, VoiceState& dsp_state, s32 channel);
+    void GenerateBiquadFilterCommandForVoice(ServerVoiceInfo& voice_info, VoiceState& dsp_state,
+                                             s32 mix_buffer_count, s32 channel);
+    void GenerateVolumeRampCommand(float last_volume, float current_volume, s32 channel,
+                                   s32 node_id);
+    void GenerateVoiceMixCommand(const MixVolumeBuffer& mix_volumes,
+                                 const MixVolumeBuffer& last_mix_volumes, VoiceState& dsp_state,
+                                 s32 mix_buffer_offset, s32 mix_buffer_count, s32 voice_index,
+                                 s32 node_id);
+    void GenerateSubMixCommand(ServerMixInfo& mix_info);
+    void GenerateMixCommands(ServerMixInfo& mix_info);
+    void GenerateMixCommand(std::size_t output_offset, std::size_t input_offset, float volume,
+                            s32 node_id);
+    void GenerateFinalMixCommand();
+    void GenerateBiquadFilterCommand(s32 mix_buffer, const BiquadFilterParameter& params,
+                                     std::array<s64, 2>& state, std::size_t input_offset,
+                                     std::size_t output_offset, s32 sample_count, s32 node_id);
+    void GenerateDepopPrepareCommand(VoiceState& dsp_state, std::size_t mix_buffer_count,
+                                     std::size_t mix_buffer_offset);
+    void GenerateDepopForMixBuffersCommand(std::size_t mix_buffer_count,
+                                           std::size_t mix_buffer_offset, s32 sample_rate);
+    void GenerateEffectCommand(ServerMixInfo& mix_info);
+    void GenerateI3dl2ReverbEffectCommand(s32 mix_buffer_offset, EffectBase* info, bool enabled);
+    void GenerateBiquadFilterEffectCommand(s32 mix_buffer_offset, EffectBase* info, bool enabled);
+    void GenerateAuxCommand(s32 mix_buffer_offset, EffectBase* info, bool enabled);
+    ServerSplitterDestinationData* GetDestinationData(s32 splitter_id, s32 index);
+
+    s32 WriteAuxBuffer(AuxInfoDSP& dsp_info, VAddr send_buffer, u32 max_samples, const s32* data,
+                       u32 sample_count, u32 write_offset, u32 write_count);
+    s32 ReadAuxBuffer(AuxInfoDSP& recv_info, VAddr recv_buffer, u32 max_samples, s32* out_data,
+                      u32 sample_count, u32 read_offset, u32 read_count);
+
+    // DSP Code
+    s32 DecodePcm16(ServerVoiceInfo& voice_info, VoiceState& dsp_state, s32 sample_count,
+                    s32 channel, std::size_t mix_offset);
+    s32 DecodeAdpcm(ServerVoiceInfo& voice_info, VoiceState& dsp_state, s32 sample_count,
+                    s32 channel, std::size_t mix_offset);
+    void DecodeFromWaveBuffers(ServerVoiceInfo& voice_info, s32* output, VoiceState& dsp_state,
+                               s32 channel, s32 target_sample_rate, s32 sample_count, s32 node_id);
+
+    AudioCommon::AudioRendererParameter& worker_params;
+    VoiceContext& voice_context;
+    MixContext& mix_context;
+    SplitterContext& splitter_context;
+    EffectContext& effect_context;
+    Core::Memory::Memory& memory;
+    std::vector<s32> mix_buffer{};
+    std::vector<s32> sample_buffer{};
+    std::vector<s32> depop_buffer{};
+    bool dumping_frame{false};
+};
+} // namespace AudioCore

src/audio_core/common.h

@@ -3,18 +3,36 @@
 // Refer to the license.txt file included.
 
 #pragma once
+
 #include "common/common_funcs.h"
 #include "common/common_types.h"
 #include "common/swap.h"
 #include "core/hle/result.h"
 
-namespace AudioCore {
+namespace AudioCommon {
 namespace Audren {
 constexpr ResultCode ERR_INVALID_PARAMETERS{ErrorModule::Audio, 41};
-}
+constexpr ResultCode ERR_SPLITTER_SORT_FAILED{ErrorModule::Audio, 43};
+} // namespace Audren
 
 constexpr u32_le CURRENT_PROCESS_REVISION = Common::MakeMagic('R', 'E', 'V', '8');
 constexpr std::size_t MAX_MIX_BUFFERS = 24;
+constexpr std::size_t MAX_BIQUAD_FILTERS = 2;
+constexpr std::size_t MAX_CHANNEL_COUNT = 6;
+constexpr std::size_t MAX_WAVE_BUFFERS = 4;
+constexpr std::size_t MAX_SAMPLE_HISTORY = 4;
+constexpr u32 STREAM_SAMPLE_RATE = 48000;
+constexpr u32 STREAM_NUM_CHANNELS = 6;
+constexpr s32 NO_SPLITTER = -1;
+constexpr s32 NO_MIX = 0x7fffffff;
+constexpr s32 NO_FINAL_MIX = std::numeric_limits<s32>::min();
+constexpr s32 FINAL_MIX = 0;
+constexpr s32 NO_EFFECT_ORDER = -1;
+constexpr std::size_t TEMP_MIX_BASE_SIZE = 0x3f00; // TODO(ogniK): Work out this constant
+// Any size checks seem to take the sample history into account
+// 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;
 
 static constexpr u32 VersionFromRevision(u32_le rev) {
     // "REV7" -> 7
@@ -45,4 +63,46 @@ static constexpr bool CanConsumeBuffer(std::size_t size, std::size_t offset, std
     return true;
 }
 
-} // namespace AudioCore
+struct UpdateDataSizes {
+    u32_le behavior{};
+    u32_le memory_pool{};
+    u32_le voice{};
+    u32_le voice_channel_resource{};
+    u32_le effect{};
+    u32_le mixer{};
+    u32_le sink{};
+    u32_le performance{};
+    u32_le splitter{};
+    u32_le render_info{};
+    INSERT_PADDING_WORDS(4);
+};
+static_assert(sizeof(UpdateDataSizes) == 0x38, "UpdateDataSizes is an invalid size");
+
+struct UpdateDataHeader {
+    u32_le revision{};
+    UpdateDataSizes size{};
+    u32_le total_size{};
+};
+static_assert(sizeof(UpdateDataHeader) == 0x40, "UpdateDataHeader is an invalid size");
+
+struct AudioRendererParameter {
+    u32_le sample_rate;
+    u32_le sample_count;
+    u32_le mix_buffer_count;
+    u32_le submix_count;
+    u32_le voice_count;
+    u32_le sink_count;
+    u32_le effect_count;
+    u32_le performance_frame_count;
+    u8 is_voice_drop_enabled;
+    u8 unknown_21;
+    u8 unknown_22;
+    u8 execution_mode;
+    u32_le splitter_count;
+    u32_le num_splitter_send_channels;
+    u32_le unknown_30;
+    u32_le revision;
+};
+static_assert(sizeof(AudioRendererParameter) == 52, "AudioRendererParameter is an invalid size");
+
+} // namespace AudioCommon

src/audio_core/cubeb_sink.cpp

@@ -23,14 +23,24 @@ class CubebSinkStream final : public SinkStream {
 public:
     CubebSinkStream(cubeb* ctx, u32 sample_rate, u32 num_channels_, cubeb_devid output_device,
                     const std::string& name)
-        : ctx{ctx}, num_channels{std::min(num_channels_, 2u)}, time_stretch{sample_rate,
+        : ctx{ctx}, num_channels{std::min(num_channels_, 6u)}, time_stretch{sample_rate,
                                                                             num_channels} {
 
         cubeb_stream_params params{};
         params.rate = sample_rate;
         params.channels = num_channels;
         params.format = CUBEB_SAMPLE_S16NE;
-        params.layout = num_channels == 1 ? CUBEB_LAYOUT_MONO : CUBEB_LAYOUT_STEREO;
+        switch (num_channels) {
+        case 1:
+            params.layout = CUBEB_LAYOUT_MONO;
+            break;
+        case 2:
+            params.layout = CUBEB_LAYOUT_STEREO;
+            break;
+        case 6:
+            params.layout = CUBEB_LAYOUT_3F2_LFE;
+            break;
+        }
 
         u32 minimum_latency{};
         if (cubeb_get_min_latency(ctx, &params, &minimum_latency) != CUBEB_OK) {
@@ -83,8 +93,10 @@ public:
                 constexpr s32 clev{707}; // center mixing level coefficient
                 constexpr s32 slev{707}; // surround mixing level coefficient
 
-                buf.push_back(left + (clev * center / 1000) + (slev * surround_left / 1000));
-                buf.push_back(right + (clev * center / 1000) + (slev * surround_right / 1000));
+                buf.push_back(static_cast<s16>(left + (clev * center / 1000) +
+                                               (slev * surround_left / 1000)));
+                buf.push_back(static_cast<s16>(right + (clev * center / 1000) +
+                                               (slev * surround_right / 1000)));
             }
             queue.Push(buf);
             return;
@@ -182,8 +194,8 @@ SinkStream& CubebSink::AcquireSinkStream(u32 sample_rate, u32 num_channels,
 
 long CubebSinkStream::DataCallback(cubeb_stream* stream, void* user_data, const void* input_buffer,
                                    void* output_buffer, long num_frames) {
-    CubebSinkStream* impl = static_cast<CubebSinkStream*>(user_data);
-    u8* buffer = reinterpret_cast<u8*>(output_buffer);
+    auto* impl = static_cast<CubebSinkStream*>(user_data);
+    auto* buffer = static_cast<u8*>(output_buffer);
 
     if (!impl) {
         return {};
@@ -193,6 +205,7 @@ long CubebSinkStream::DataCallback(cubeb_stream* stream, void* user_data, const
     const std::size_t samples_to_write = num_channels * num_frames;
     std::size_t samples_written;
 
+    /*
     if (Settings::values.enable_audio_stretching.GetValue()) {
         const std::vector<s16> in{impl->queue.Pop()};
         const std::size_t num_in{in.size() / num_channels};
@@ -207,7 +220,8 @@ long CubebSinkStream::DataCallback(cubeb_stream* stream, void* user_data, const
         }
     } else {
         samples_written = impl->queue.Pop(buffer, samples_to_write);
-    }
+    }*/
+    samples_written = impl->queue.Pop(buffer, samples_to_write);
 
     if (samples_written >= num_channels) {
         std::memcpy(&impl->last_frame[0], buffer + (samples_written - num_channels) * sizeof(s16),

src/audio_core/effect_context.cpp

@@ -0,0 +1,299 @@
+// Copyright 2020 yuzu Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#include <algorithm>
+#include "audio_core/effect_context.h"
+
+namespace AudioCore {
+namespace {
+bool ValidChannelCountForEffect(s32 channel_count) {
+    return channel_count == 1 || channel_count == 2 || channel_count == 4 || channel_count == 6;
+}
+} // namespace
+
+EffectContext::EffectContext(std::size_t effect_count) : effect_count(effect_count) {
+    effects.reserve(effect_count);
+    std::generate_n(std::back_inserter(effects), effect_count,
+                    [] { return std::make_unique<EffectStubbed>(); });
+}
+EffectContext::~EffectContext() = default;
+
+std::size_t EffectContext::GetCount() const {
+    return effect_count;
+}
+
+EffectBase* EffectContext::GetInfo(std::size_t i) {
+    return effects.at(i).get();
+}
+
+EffectBase* EffectContext::RetargetEffect(std::size_t i, EffectType effect) {
+    switch (effect) {
+    case EffectType::Invalid:
+        effects[i] = std::make_unique<EffectStubbed>();
+        break;
+    case EffectType::BufferMixer:
+        effects[i] = std::make_unique<EffectBufferMixer>();
+        break;
+    case EffectType::Aux:
+        effects[i] = std::make_unique<EffectAuxInfo>();
+        break;
+    case EffectType::Delay:
+        effects[i] = std::make_unique<EffectDelay>();
+        break;
+    case EffectType::Reverb:
+        effects[i] = std::make_unique<EffectReverb>();
+        break;
+    case EffectType::I3dl2Reverb:
+        effects[i] = std::make_unique<EffectI3dl2Reverb>();
+        break;
+    case EffectType::BiquadFilter:
+        effects[i] = std::make_unique<EffectBiquadFilter>();
+        break;
+    default:
+        UNREACHABLE_MSG("Unimplemented effect {}", effect);
+        effects[i] = std::make_unique<EffectStubbed>();
+    }
+    return GetInfo(i);
+}
+
+const EffectBase* EffectContext::GetInfo(std::size_t i) const {
+    return effects.at(i).get();
+}
+
+EffectStubbed::EffectStubbed() : EffectBase::EffectBase(EffectType::Invalid) {}
+EffectStubbed::~EffectStubbed() = default;
+
+void EffectStubbed::Update(EffectInfo::InParams& in_params) {}
+void EffectStubbed::UpdateForCommandGeneration() {}
+
+EffectBase::EffectBase(EffectType effect_type) : effect_type(effect_type) {}
+EffectBase::~EffectBase() = default;
+
+UsageState EffectBase::GetUsage() const {
+    return usage;
+}
+
+EffectType EffectBase::GetType() const {
+    return effect_type;
+}
+
+bool EffectBase::IsEnabled() const {
+    return enabled;
+}
+
+s32 EffectBase::GetMixID() const {
+    return mix_id;
+}
+
+s32 EffectBase::GetProcessingOrder() const {
+    return processing_order;
+}
+
+EffectI3dl2Reverb::EffectI3dl2Reverb() : EffectGeneric::EffectGeneric(EffectType::I3dl2Reverb) {}
+EffectI3dl2Reverb::~EffectI3dl2Reverb() = default;
+
+void EffectI3dl2Reverb::Update(EffectInfo::InParams& in_params) {
+    auto& internal_params = GetParams();
+    const auto* reverb_params = reinterpret_cast<I3dl2ReverbParams*>(in_params.raw.data());
+    if (!ValidChannelCountForEffect(reverb_params->max_channels)) {
+        UNREACHABLE_MSG("Invalid reverb max channel count {}", reverb_params->max_channels);
+        return;
+    }
+
+    const auto last_status = internal_params.status;
+    mix_id = in_params.mix_id;
+    processing_order = in_params.processing_order;
+    internal_params = *reverb_params;
+    if (!ValidChannelCountForEffect(reverb_params->channel_count)) {
+        internal_params.channel_count = internal_params.max_channels;
+    }
+    enabled = in_params.is_enabled;
+    if (last_status != ParameterStatus::Updated) {
+        internal_params.status = last_status;
+    }
+
+    if (in_params.is_new || skipped) {
+        usage = UsageState::Initialized;
+        internal_params.status = ParameterStatus::Initialized;
+        skipped = in_params.buffer_address == 0 || in_params.buffer_size == 0;
+    }
+}
+
+void EffectI3dl2Reverb::UpdateForCommandGeneration() {
+    if (enabled) {
+        usage = UsageState::Running;
+    } else {
+        usage = UsageState::Stopped;
+    }
+    GetParams().status = ParameterStatus::Updated;
+}
+
+EffectBiquadFilter::EffectBiquadFilter() : EffectGeneric::EffectGeneric(EffectType::BiquadFilter) {}
+EffectBiquadFilter::~EffectBiquadFilter() = default;
+
+void EffectBiquadFilter::Update(EffectInfo::InParams& in_params) {
+    auto& internal_params = GetParams();
+    const auto* biquad_params = reinterpret_cast<BiquadFilterParams*>(in_params.raw.data());
+    mix_id = in_params.mix_id;
+    processing_order = in_params.processing_order;
+    internal_params = *biquad_params;
+    enabled = in_params.is_enabled;
+}
+
+void EffectBiquadFilter::UpdateForCommandGeneration() {
+    if (enabled) {
+        usage = UsageState::Running;
+    } else {
+        usage = UsageState::Stopped;
+    }
+    GetParams().status = ParameterStatus::Updated;
+}
+
+EffectAuxInfo::EffectAuxInfo() : EffectGeneric::EffectGeneric(EffectType::Aux) {}
+EffectAuxInfo::~EffectAuxInfo() = default;
+
+void EffectAuxInfo::Update(EffectInfo::InParams& in_params) {
+    const auto* aux_params = reinterpret_cast<AuxInfo*>(in_params.raw.data());
+    mix_id = in_params.mix_id;
+    processing_order = in_params.processing_order;
+    GetParams() = *aux_params;
+    enabled = in_params.is_enabled;
+
+    if (in_params.is_new || skipped) {
+        skipped = aux_params->send_buffer_info == 0 || aux_params->return_buffer_info == 0;
+        if (skipped) {
+            return;
+        }
+
+        // There's two AuxInfos which are an identical size, the first one is managed by the cpu,
+        // the second is managed by the dsp. All we care about is managing the DSP one
+        send_info = aux_params->send_buffer_info + sizeof(AuxInfoDSP);
+        send_buffer = aux_params->send_buffer_info + (sizeof(AuxInfoDSP) * 2);
+
+        recv_info = aux_params->return_buffer_info + sizeof(AuxInfoDSP);
+        recv_buffer = aux_params->return_buffer_info + (sizeof(AuxInfoDSP) * 2);
+    }
+}
+
+void EffectAuxInfo::UpdateForCommandGeneration() {
+    if (enabled) {
+        usage = UsageState::Running;
+    } else {
+        usage = UsageState::Stopped;
+    }
+}
+
+VAddr EffectAuxInfo::GetSendInfo() const {
+    return send_info;
+}
+
+VAddr EffectAuxInfo::GetSendBuffer() const {
+    return send_buffer;
+}
+
+VAddr EffectAuxInfo::GetRecvInfo() const {
+    return recv_info;
+}
+
+VAddr EffectAuxInfo::GetRecvBuffer() const {
+    return recv_buffer;
+}
+
+EffectDelay::EffectDelay() : EffectGeneric::EffectGeneric(EffectType::Delay) {}
+EffectDelay::~EffectDelay() = default;
+
+void EffectDelay::Update(EffectInfo::InParams& in_params) {
+    const auto* delay_params = reinterpret_cast<DelayParams*>(in_params.raw.data());
+    auto& internal_params = GetParams();
+    if (!ValidChannelCountForEffect(delay_params->max_channels)) {
+        return;
+    }
+
+    const auto last_status = internal_params.status;
+    mix_id = in_params.mix_id;
+    processing_order = in_params.processing_order;
+    internal_params = *delay_params;
+    if (!ValidChannelCountForEffect(delay_params->channels)) {
+        internal_params.channels = internal_params.max_channels;
+    }
+    enabled = in_params.is_enabled;
+
+    if (last_status != ParameterStatus::Updated) {
+        internal_params.status = last_status;
+    }
+
+    if (in_params.is_new || skipped) {
+        usage = UsageState::Initialized;
+        internal_params.status = ParameterStatus::Initialized;
+        skipped = in_params.buffer_address == 0 || in_params.buffer_size == 0;
+    }
+}
+
+void EffectDelay::UpdateForCommandGeneration() {
+    if (enabled) {
+        usage = UsageState::Running;
+    } else {
+        usage = UsageState::Stopped;
+    }
+    GetParams().status = ParameterStatus::Updated;
+}
+
+EffectBufferMixer::EffectBufferMixer() : EffectGeneric::EffectGeneric(EffectType::BufferMixer) {}
+EffectBufferMixer::~EffectBufferMixer() = default;
+
+void EffectBufferMixer::Update(EffectInfo::InParams& in_params) {
+    mix_id = in_params.mix_id;
+    processing_order = in_params.processing_order;
+    GetParams() = *reinterpret_cast<BufferMixerParams*>(in_params.raw.data());
+    enabled = in_params.is_enabled;
+}
+
+void EffectBufferMixer::UpdateForCommandGeneration() {
+    if (enabled) {
+        usage = UsageState::Running;
+    } else {
+        usage = UsageState::Stopped;
+    }
+}
+
+EffectReverb::EffectReverb() : EffectGeneric::EffectGeneric(EffectType::Reverb) {}
+EffectReverb::~EffectReverb() = default;
+
+void EffectReverb::Update(EffectInfo::InParams& in_params) {
+    const auto* reverb_params = reinterpret_cast<ReverbParams*>(in_params.raw.data());
+    auto& internal_params = GetParams();
+    if (!ValidChannelCountForEffect(reverb_params->max_channels)) {
+        return;
+    }
+
+    const auto last_status = internal_params.status;
+    mix_id = in_params.mix_id;
+    processing_order = in_params.processing_order;
+    internal_params = *reverb_params;
+    if (!ValidChannelCountForEffect(reverb_params->channels)) {
+        internal_params.channels = internal_params.max_channels;
+    }
+    enabled = in_params.is_enabled;
+
+    if (last_status != ParameterStatus::Updated) {
+        internal_params.status = last_status;
+    }
+
+    if (in_params.is_new || skipped) {
+        usage = UsageState::Initialized;
+        internal_params.status = ParameterStatus::Initialized;
+        skipped = in_params.buffer_address == 0 || in_params.buffer_size == 0;
+    }
+}
+
+void EffectReverb::UpdateForCommandGeneration() {
+    if (enabled) {
+        usage = UsageState::Running;
+    } else {
+        usage = UsageState::Stopped;
+    }
+    GetParams().status = ParameterStatus::Updated;
+}
+
+} // namespace AudioCore

src/audio_core/effect_context.h

@@ -0,0 +1,321 @@
+// Copyright 2020 yuzu Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#pragma once
+
+#include <array>
+#include <memory>
+#include <vector>
+#include "audio_core/common.h"
+#include "common/common_funcs.h"
+#include "common/common_types.h"
+#include "common/swap.h"
+
+namespace AudioCore {
+enum class EffectType : u8 {
+    Invalid = 0,
+    BufferMixer = 1,
+    Aux = 2,
+    Delay = 3,
+    Reverb = 4,
+    I3dl2Reverb = 5,
+    BiquadFilter = 6,
+};
+
+enum class UsageStatus : u8 {
+    Invalid = 0,
+    New = 1,
+    Initialized = 2,
+    Used = 3,
+    Removed = 4,
+};
+
+enum class UsageState {
+    Invalid = 0,
+    Initialized = 1,
+    Running = 2,
+    Stopped = 3,
+};
+
+enum class ParameterStatus : u8 {
+    Initialized = 0,
+    Updating = 1,
+    Updated = 2,
+};
+
+struct BufferMixerParams {
+    std::array<s8, AudioCommon::MAX_MIX_BUFFERS> input{};
+    std::array<s8, AudioCommon::MAX_MIX_BUFFERS> output{};
+    std::array<float_le, AudioCommon::MAX_MIX_BUFFERS> volume{};
+    s32_le count{};
+};
+static_assert(sizeof(BufferMixerParams) == 0x94, "BufferMixerParams is an invalid size");
+
+struct AuxInfoDSP {
+    u32_le read_offset{};
+    u32_le write_offset{};
+    u32_le remaining{};
+    INSERT_PADDING_WORDS(13);
+};
+static_assert(sizeof(AuxInfoDSP) == 0x40, "AuxInfoDSP is an invalid size");
+
+struct AuxInfo {
+    std::array<s8, AudioCommon::MAX_MIX_BUFFERS> input_mix_buffers{};
+    std::array<s8, AudioCommon::MAX_MIX_BUFFERS> output_mix_buffers{};
+    u32_le count{};
+    s32_le sample_rate{};
+    s32_le sample_count{};
+    s32_le mix_buffer_count{};
+    u64_le send_buffer_info{};
+    u64_le send_buffer_base{};
+
+    u64_le return_buffer_info{};
+    u64_le return_buffer_base{};
+};
+static_assert(sizeof(AuxInfo) == 0x60, "AuxInfo is an invalid size");
+
+struct I3dl2ReverbParams {
+    std::array<s8, AudioCommon::MAX_CHANNEL_COUNT> input{};
+    std::array<s8, AudioCommon::MAX_CHANNEL_COUNT> output{};
+    u16_le max_channels{};
+    u16_le channel_count{};
+    INSERT_PADDING_BYTES(1);
+    u32_le sample_rate{};
+    f32 room_hf{};
+    f32 hf_reference{};
+    f32 decay_time{};
+    f32 hf_decay_ratio{};
+    f32 room{};
+    f32 reflection{};
+    f32 reverb{};
+    f32 diffusion{};
+    f32 reflection_delay{};
+    f32 reverb_delay{};
+    f32 density{};
+    f32 dry_gain{};
+    ParameterStatus status{};
+    INSERT_PADDING_BYTES(3);
+};
+static_assert(sizeof(I3dl2ReverbParams) == 0x4c, "I3dl2ReverbParams is an invalid size");
+
+struct BiquadFilterParams {
+    std::array<s8, AudioCommon::MAX_CHANNEL_COUNT> input{};
+    std::array<s8, AudioCommon::MAX_CHANNEL_COUNT> output{};
+    std::array<s16_le, 3> numerator;
+    std::array<s16_le, 2> denominator;
+    s8 channel_count{};
+    ParameterStatus status{};
+};
+static_assert(sizeof(BiquadFilterParams) == 0x18, "BiquadFilterParams is an invalid size");
+
+struct DelayParams {
+    std::array<s8, AudioCommon::MAX_CHANNEL_COUNT> input{};
+    std::array<s8, AudioCommon::MAX_CHANNEL_COUNT> output{};
+    u16_le max_channels{};
+    u16_le channels{};
+    s32_le max_delay{};
+    s32_le delay{};
+    s32_le sample_rate{};
+    s32_le gain{};
+    s32_le feedback_gain{};
+    s32_le out_gain{};
+    s32_le dry_gain{};
+    s32_le channel_spread{};
+    s32_le low_pass{};
+    ParameterStatus status{};
+    INSERT_PADDING_BYTES(3);
+};
+static_assert(sizeof(DelayParams) == 0x38, "DelayParams is an invalid size");
+
+struct ReverbParams {
+    std::array<s8, AudioCommon::MAX_CHANNEL_COUNT> input{};
+    std::array<s8, AudioCommon::MAX_CHANNEL_COUNT> output{};
+    u16_le max_channels{};
+    u16_le channels{};
+    s32_le sample_rate{};
+    s32_le mode0{};
+    s32_le mode0_gain{};
+    s32_le pre_delay{};
+    s32_le mode1{};
+    s32_le mode1_gain{};
+    s32_le decay{};
+    s32_le hf_decay_ratio{};
+    s32_le coloration{};
+    s32_le reverb_gain{};
+    s32_le out_gain{};
+    s32_le dry_gain{};
+    ParameterStatus status{};
+    INSERT_PADDING_BYTES(3);
+};
+static_assert(sizeof(ReverbParams) == 0x44, "ReverbParams is an invalid size");
+
+class EffectInfo {
+public:
+    struct InParams {
+        EffectType type{};
+        u8 is_new{};
+        u8 is_enabled{};
+        INSERT_PADDING_BYTES(1);
+        s32_le mix_id{};
+        u64_le buffer_address{};
+        u64_le buffer_size{};
+        s32_le processing_order{};
+        INSERT_PADDING_BYTES(4);
+        union {
+            std::array<u8, 0xa0> raw;
+        };
+    };
+    static_assert(sizeof(InParams) == 0xc0, "InParams is an invalid size");
+
+    struct OutParams {
+        UsageStatus status{};
+        INSERT_PADDING_BYTES(15);
+    };
+    static_assert(sizeof(OutParams) == 0x10, "OutParams is an invalid size");
+};
+
+struct AuxAddress {
+    VAddr send_dsp_info{};
+    VAddr send_buffer_base{};
+    VAddr return_dsp_info{};
+    VAddr return_buffer_base{};
+};
+
+class EffectBase {
+public:
+    explicit EffectBase(EffectType effect_type);
+    virtual ~EffectBase();
+
+    virtual void Update(EffectInfo::InParams& in_params) = 0;
+    virtual void UpdateForCommandGeneration() = 0;
+    UsageState GetUsage() const;
+    EffectType GetType() const;
+    bool IsEnabled() const;
+    s32 GetMixID() const;
+    s32 GetProcessingOrder() const;
+
+protected:
+    UsageState usage{UsageState::Invalid};
+    EffectType effect_type{};
+    s32 mix_id{};
+    s32 processing_order{};
+    bool enabled = false;
+};
+
+template <typename T>
+class EffectGeneric : public EffectBase {
+public:
+    explicit EffectGeneric(EffectType effect_type) : EffectBase(effect_type) {}
+
+    T& GetParams() {
+        return internal_params;
+    }
+
+    const I3dl2ReverbParams& GetParams() const {
+        return internal_params;
+    }
+
+private:
+    T internal_params{};
+};
+
+class EffectStubbed : public EffectBase {
+public:
+    explicit EffectStubbed();
+    ~EffectStubbed() override;
+
+    void Update(EffectInfo::InParams& in_params) override;
+    void UpdateForCommandGeneration() override;
+};
+
+class EffectI3dl2Reverb : public EffectGeneric<I3dl2ReverbParams> {
+public:
+    explicit EffectI3dl2Reverb();
+    ~EffectI3dl2Reverb() override;
+
+    void Update(EffectInfo::InParams& in_params) override;
+    void UpdateForCommandGeneration() override;
+
+private:
+    bool skipped = false;
+};
+
+class EffectBiquadFilter : public EffectGeneric<BiquadFilterParams> {
+public:
+    explicit EffectBiquadFilter();
+    ~EffectBiquadFilter() override;
+
+    void Update(EffectInfo::InParams& in_params) override;
+    void UpdateForCommandGeneration() override;
+};
+
+class EffectAuxInfo : public EffectGeneric<AuxInfo> {
+public:
+    explicit EffectAuxInfo();
+    ~EffectAuxInfo() override;
+
+    void Update(EffectInfo::InParams& in_params) override;
+    void UpdateForCommandGeneration() override;
+    VAddr GetSendInfo() const;
+    VAddr GetSendBuffer() const;
+    VAddr GetRecvInfo() const;
+    VAddr GetRecvBuffer() const;
+
+private:
+    VAddr send_info{};
+    VAddr send_buffer{};
+    VAddr recv_info{};
+    VAddr recv_buffer{};
+    bool skipped = false;
+    AuxAddress addresses{};
+};
+
+class EffectDelay : public EffectGeneric<DelayParams> {
+public:
+    explicit EffectDelay();
+    ~EffectDelay() override;
+
+    void Update(EffectInfo::InParams& in_params) override;
+    void UpdateForCommandGeneration() override;
+
+private:
+    bool skipped = false;
+};
+
+class EffectBufferMixer : public EffectGeneric<BufferMixerParams> {
+public:
+    explicit EffectBufferMixer();
+    ~EffectBufferMixer() override;
+
+    void Update(EffectInfo::InParams& in_params) override;
+    void UpdateForCommandGeneration() override;
+};
+
+class EffectReverb : public EffectGeneric<ReverbParams> {
+public:
+    explicit EffectReverb();
+    ~EffectReverb() override;
+
+    void Update(EffectInfo::InParams& in_params) override;
+    void UpdateForCommandGeneration() override;
+
+private:
+    bool skipped = false;
+};
+
+class EffectContext {
+public:
+    explicit EffectContext(std::size_t effect_count);
+    ~EffectContext();
+
+    std::size_t GetCount() const;
+    EffectBase* GetInfo(std::size_t i);
+    EffectBase* RetargetEffect(std::size_t i, EffectType effect);
+    const EffectBase* GetInfo(std::size_t i) const;
+
+private:
+    std::size_t effect_count{};
+    std::vector<std::unique_ptr<EffectBase>> effects;
+};
+} // namespace AudioCore

src/audio_core/info_updater.cpp

@@ -0,0 +1,516 @@
+// Copyright 2020 yuzu Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#include "audio_core/behavior_info.h"
+#include "audio_core/effect_context.h"
+#include "audio_core/info_updater.h"
+#include "audio_core/memory_pool.h"
+#include "audio_core/mix_context.h"
+#include "audio_core/sink_context.h"
+#include "audio_core/splitter_context.h"
+#include "audio_core/voice_context.h"
+#include "common/logging/log.h"
+
+namespace AudioCore {
+
+InfoUpdater::InfoUpdater(const std::vector<u8>& in_params, std::vector<u8>& out_params,
+                         BehaviorInfo& behavior_info)
+    : in_params(in_params), out_params(out_params), behavior_info(behavior_info) {
+    ASSERT(
+        AudioCommon::CanConsumeBuffer(in_params.size(), 0, sizeof(AudioCommon::UpdateDataHeader)));
+    std::memcpy(&input_header, in_params.data(), sizeof(AudioCommon::UpdateDataHeader));
+    output_header.total_size = sizeof(AudioCommon::UpdateDataHeader);
+}
+
+InfoUpdater::~InfoUpdater() = default;
+
+bool InfoUpdater::UpdateBehaviorInfo(BehaviorInfo& in_behavior_info) {
+    if (input_header.size.behavior != sizeof(BehaviorInfo::InParams)) {
+        LOG_ERROR(Audio, "Behavior info is an invalid size, expecting 0x{:X} but got 0x{:X}",
+                  sizeof(BehaviorInfo::InParams), input_header.size.behavior);
+        return false;
+    }
+
+    if (!AudioCommon::CanConsumeBuffer(in_params.size(), input_offset,
+                                       sizeof(BehaviorInfo::InParams))) {
+        LOG_ERROR(Audio, "Buffer is an invalid size!");
+        return false;
+    }
+
+    BehaviorInfo::InParams behavior_in{};
+    std::memcpy(&behavior_in, in_params.data() + input_offset, sizeof(BehaviorInfo::InParams));
+    input_offset += sizeof(BehaviorInfo::InParams);
+
+    // Make sure it's an audio revision we can actually support
+    if (!AudioCommon::IsValidRevision(behavior_in.revision)) {
+        LOG_ERROR(Audio, "Invalid input revision, revision=0x{:08X}", behavior_in.revision);
+        return false;
+    }
+
+    // Make sure that our behavior info revision matches the input
+    if (in_behavior_info.GetUserRevision() != behavior_in.revision) {
+        LOG_ERROR(Audio,
+                  "User revision differs from input revision, expecting 0x{:08X} but got 0x{:08X}",
+                  in_behavior_info.GetUserRevision(), behavior_in.revision);
+        return false;
+    }
+
+    // Update behavior info flags
+    in_behavior_info.ClearError();
+    in_behavior_info.UpdateFlags(behavior_in.flags);
+
+    return true;
+}
+
+bool InfoUpdater::UpdateMemoryPools(std::vector<ServerMemoryPoolInfo>& memory_pool_info) {
+    const auto memory_pool_count = memory_pool_info.size();
+    const auto total_memory_pool_in = sizeof(ServerMemoryPoolInfo::InParams) * memory_pool_count;
+    const auto total_memory_pool_out = sizeof(ServerMemoryPoolInfo::OutParams) * memory_pool_count;
+
+    if (input_header.size.memory_pool != total_memory_pool_in) {
+        LOG_ERROR(Audio, "Memory pools are an invalid size, expecting 0x{:X} but got 0x{:X}",
+                  total_memory_pool_in, input_header.size.memory_pool);
+        return false;
+    }
+
+    if (!AudioCommon::CanConsumeBuffer(in_params.size(), input_offset, total_memory_pool_in)) {
+        LOG_ERROR(Audio, "Buffer is an invalid size!");
+        return false;
+    }
+
+    std::vector<ServerMemoryPoolInfo::InParams> mempool_in(memory_pool_count);
+    std::vector<ServerMemoryPoolInfo::OutParams> mempool_out(memory_pool_count);
+
+    std::memcpy(mempool_in.data(), in_params.data() + input_offset, total_memory_pool_in);
+    input_offset += total_memory_pool_in;
+
+    // Update our memory pools
+    for (std::size_t i = 0; i < memory_pool_count; i++) {
+        if (!memory_pool_info[i].Update(mempool_in[i], mempool_out[i])) {
+            LOG_ERROR(Audio, "Failed to update memory pool {}!", i);
+            return false;
+        }
+    }
+
+    if (!AudioCommon::CanConsumeBuffer(out_params.size(), output_offset,
+                                       sizeof(BehaviorInfo::InParams))) {
+        LOG_ERROR(Audio, "Buffer is an invalid size!");
+        return false;
+    }
+
+    std::memcpy(out_params.data() + output_offset, mempool_out.data(), total_memory_pool_out);
+    output_offset += total_memory_pool_out;
+    output_header.size.memory_pool = static_cast<u32>(total_memory_pool_out);
+    return true;
+}
+
+bool InfoUpdater::UpdateVoiceChannelResources(VoiceContext& voice_context) {
+    const auto voice_count = voice_context.GetVoiceCount();
+    const auto voice_size = voice_count * sizeof(VoiceChannelResource::InParams);
+    std::vector<VoiceChannelResource::InParams> resources_in(voice_count);
+
+    if (input_header.size.voice_channel_resource != voice_size) {
+        LOG_ERROR(Audio, "VoiceChannelResource is an invalid size, expecting 0x{:X} but got 0x{:X}",
+                  voice_size, input_header.size.voice_channel_resource);
+        return false;
+    }
+
+    if (!AudioCommon::CanConsumeBuffer(in_params.size(), input_offset, voice_size)) {
+        LOG_ERROR(Audio, "Buffer is an invalid size!");
+        return false;
+    }
+
+    std::memcpy(resources_in.data(), in_params.data() + input_offset, voice_size);
+    input_offset += voice_size;
+
+    // Update our channel resources
+    for (std::size_t i = 0; i < voice_count; i++) {
+        // Grab our channel resource
+        auto& resource = voice_context.GetChannelResource(i);
+        resource.Update(resources_in[i]);
+    }
+
+    return true;
+}
+
+bool InfoUpdater::UpdateVoices(VoiceContext& voice_context,
+                               std::vector<ServerMemoryPoolInfo>& memory_pool_info,
+                               VAddr audio_codec_dsp_addr) {
+    const auto voice_count = voice_context.GetVoiceCount();
+    std::vector<VoiceInfo::InParams> voice_in(voice_count);
+    std::vector<VoiceInfo::OutParams> voice_out(voice_count);
+
+    const auto voice_in_size = voice_count * sizeof(VoiceInfo::InParams);
+    const auto voice_out_size = voice_count * sizeof(VoiceInfo::OutParams);
+
+    if (input_header.size.voice != voice_in_size) {
+        LOG_ERROR(Audio, "Voices are an invalid size, expecting 0x{:X} but got 0x{:X}",
+                  voice_in_size, input_header.size.voice);
+        return false;
+    }
+
+    if (!AudioCommon::CanConsumeBuffer(in_params.size(), input_offset, voice_in_size)) {
+        LOG_ERROR(Audio, "Buffer is an invalid size!");
+        return false;
+    }
+
+    std::memcpy(voice_in.data(), in_params.data() + input_offset, voice_in_size);
+    input_offset += voice_in_size;
+
+    // Set all voices to not be in use
+    for (std::size_t i = 0; i < voice_count; i++) {
+        voice_context.GetInfo(i).GetInParams().in_use = false;
+    }
+
+    // Update our voices
+    for (std::size_t i = 0; i < voice_count; i++) {
+        auto& in_params = voice_in[i];
+        const auto channel_count = static_cast<std::size_t>(in_params.channel_count);
+        // Skip if it's not currently in use
+        if (!in_params.is_in_use) {
+            continue;
+        }
+        // Voice states for each channel
+        std::array<VoiceState*, AudioCommon::MAX_CHANNEL_COUNT> voice_states{};
+        ASSERT(static_cast<std::size_t>(in_params.id) < voice_count);
+
+        // Grab our current voice info
+        auto& voice_info = voice_context.GetInfo(static_cast<std::size_t>(in_params.id));
+
+        ASSERT(channel_count <= AudioCommon::MAX_CHANNEL_COUNT);
+
+        // Get all our channel voice states
+        for (std::size_t channel = 0; channel < channel_count; channel++) {
+            voice_states[channel] =
+                &voice_context.GetState(in_params.voice_channel_resource_ids[channel]);
+        }
+
+        if (in_params.is_new) {
+            // Default our values for our voice
+            voice_info.Initialize();
+            if (channel_count == 0 || channel_count > AudioCommon::MAX_CHANNEL_COUNT) {
+                continue;
+            }
+
+            // Zero out our voice states
+            for (std::size_t channel = 0; channel < channel_count; channel++) {
+                std::memset(voice_states[channel], 0, sizeof(VoiceState));
+            }
+        }
+
+        // Update our voice
+        voice_info.UpdateParameters(in_params, behavior_info);
+        // TODO(ogniK): Handle mapping errors with behavior info based on in params response
+
+        // Update our wave buffers
+        voice_info.UpdateWaveBuffers(in_params, voice_states, behavior_info);
+        voice_info.WriteOutStatus(voice_out[i], in_params, voice_states);
+    }
+
+    if (!AudioCommon::CanConsumeBuffer(out_params.size(), output_offset, voice_out_size)) {
+        LOG_ERROR(Audio, "Buffer is an invalid size!");
+        return false;
+    }
+    std::memcpy(out_params.data() + output_offset, voice_out.data(), voice_out_size);
+    output_offset += voice_out_size;
+    output_header.size.voice = static_cast<u32>(voice_out_size);
+    return true;
+}
+
+bool InfoUpdater::UpdateEffects(EffectContext& effect_context, bool is_active) {
+    const auto effect_count = effect_context.GetCount();
+    std::vector<EffectInfo::InParams> effect_in(effect_count);
+    std::vector<EffectInfo::OutParams> effect_out(effect_count);
+
+    const auto total_effect_in = effect_count * sizeof(EffectInfo::InParams);
+    const auto total_effect_out = effect_count * sizeof(EffectInfo::OutParams);
+
+    if (input_header.size.effect != total_effect_in) {
+        LOG_ERROR(Audio, "Effects are an invalid size, expecting 0x{:X} but got 0x{:X}",
+                  total_effect_in, input_header.size.effect);
+        return false;
+    }
+
+    if (!AudioCommon::CanConsumeBuffer(in_params.size(), input_offset, total_effect_in)) {
+        LOG_ERROR(Audio, "Buffer is an invalid size!");
+        return false;
+    }
+
+    std::memcpy(effect_in.data(), in_params.data() + input_offset, total_effect_in);
+    input_offset += total_effect_in;
+
+    // Update effects
+    for (std::size_t i = 0; i < effect_count; i++) {
+        auto* info = effect_context.GetInfo(i);
+        if (effect_in[i].type != info->GetType()) {
+            info = effect_context.RetargetEffect(i, effect_in[i].type);
+        }
+
+        info->Update(effect_in[i]);
+
+        if ((!is_active && info->GetUsage() != UsageState::Initialized) ||
+            info->GetUsage() == UsageState::Stopped) {
+            effect_out[i].status = UsageStatus::Removed;
+        } else {
+            effect_out[i].status = UsageStatus::Used;
+        }
+    }
+
+    if (!AudioCommon::CanConsumeBuffer(out_params.size(), output_offset, total_effect_out)) {
+        LOG_ERROR(Audio, "Buffer is an invalid size!");
+        return false;
+    }
+
+    std::memcpy(out_params.data() + output_offset, effect_out.data(), total_effect_out);
+    output_offset += total_effect_out;
+    output_header.size.effect = static_cast<u32>(total_effect_out);
+
+    return true;
+}
+
+bool InfoUpdater::UpdateSplitterInfo(SplitterContext& splitter_context) {
+    std::size_t start_offset = input_offset;
+    std::size_t bytes_read{};
+    // Update splitter context
+    if (!splitter_context.Update(in_params, input_offset, bytes_read)) {
+        LOG_ERROR(Audio, "Failed to update splitter context!");
+        return false;
+    }
+
+    const auto consumed = input_offset - start_offset;
+
+    if (input_header.size.splitter != consumed) {
+        LOG_ERROR(Audio, "Splitters is an invalid size, expecting 0x{:X} but got 0x{:X}",
+                  bytes_read, input_header.size.splitter);
+        return false;
+    }
+
+    return true;
+}
+
+ResultCode InfoUpdater::UpdateMixes(MixContext& mix_context, std::size_t mix_buffer_count,
+                                    SplitterContext& splitter_context,
+                                    EffectContext& effect_context) {
+    std::vector<MixInfo::InParams> mix_in_params;
+
+    if (!behavior_info.IsMixInParameterDirtyOnlyUpdateSupported()) {
+        // If we're not dirty, get ALL mix in parameters
+        const auto context_mix_count = mix_context.GetCount();
+        const auto total_mix_in = context_mix_count * sizeof(MixInfo::InParams);
+        if (input_header.size.mixer != total_mix_in) {
+            LOG_ERROR(Audio, "Mixer is an invalid size, expecting 0x{:X} but got 0x{:X}",
+                      total_mix_in, input_header.size.mixer);
+            return AudioCommon::Audren::ERR_INVALID_PARAMETERS;
+        }
+
+        if (!AudioCommon::CanConsumeBuffer(in_params.size(), input_offset, total_mix_in)) {
+            LOG_ERROR(Audio, "Buffer is an invalid size!");
+            return AudioCommon::Audren::ERR_INVALID_PARAMETERS;
+        }
+
+        mix_in_params.resize(context_mix_count);
+        std::memcpy(mix_in_params.data(), in_params.data() + input_offset, total_mix_in);
+
+        input_offset += total_mix_in;
+    } else {
+        // Only update the "dirty" mixes
+        MixInfo::DirtyHeader dirty_header{};
+        if (!AudioCommon::CanConsumeBuffer(in_params.size(), input_offset,
+                                           sizeof(MixInfo::DirtyHeader))) {
+            LOG_ERROR(Audio, "Buffer is an invalid size!");
+            return AudioCommon::Audren::ERR_INVALID_PARAMETERS;
+        }
+
+        std::memcpy(&dirty_header, in_params.data() + input_offset, sizeof(MixInfo::DirtyHeader));
+        input_offset += sizeof(MixInfo::DirtyHeader);
+
+        const auto total_mix_in =
+            dirty_header.mixer_count * sizeof(MixInfo::InParams) + sizeof(MixInfo::DirtyHeader);
+
+        if (input_header.size.mixer != total_mix_in) {
+            LOG_ERROR(Audio, "Mixer is an invalid size, expecting 0x{:X} but got 0x{:X}",
+                      total_mix_in, input_header.size.mixer);
+            return AudioCommon::Audren::ERR_INVALID_PARAMETERS;
+        }
+
+        if (dirty_header.mixer_count != 0) {
+            mix_in_params.resize(dirty_header.mixer_count);
+            std::memcpy(mix_in_params.data(), in_params.data() + input_offset,
+                        mix_in_params.size() * sizeof(MixInfo::InParams));
+            input_offset += mix_in_params.size() * sizeof(MixInfo::InParams);
+        }
+    }
+
+    // Get our total input count
+    const auto mix_count = mix_in_params.size();
+
+    if (!behavior_info.IsMixInParameterDirtyOnlyUpdateSupported()) {
+        // Only verify our buffer count if we're not dirty
+        std::size_t total_buffer_count{};
+        for (std::size_t i = 0; i < mix_count; i++) {
+            const auto& in = mix_in_params[i];
+            total_buffer_count += in.buffer_count;
+            if (static_cast<std::size_t>(in.dest_mix_id) > mix_count &&
+                in.dest_mix_id != AudioCommon::NO_MIX && in.mix_id != AudioCommon::FINAL_MIX) {
+                LOG_ERROR(
+                    Audio,
+                    "Invalid mix destination, mix_id={:X}, dest_mix_id={:X}, mix_buffer_count={:X}",
+                    in.mix_id, in.dest_mix_id, mix_buffer_count);
+                return AudioCommon::Audren::ERR_INVALID_PARAMETERS;
+            }
+        }
+
+        if (total_buffer_count > mix_buffer_count) {
+            LOG_ERROR(Audio,
+                      "Too many mix buffers used! mix_buffer_count={:X}, requesting_buffers={:X}",
+                      mix_buffer_count, total_buffer_count);
+            return AudioCommon::Audren::ERR_INVALID_PARAMETERS;
+        }
+    }
+
+    if (mix_buffer_count == 0) {
+        LOG_ERROR(Audio, "No mix buffers!");
+        return AudioCommon::Audren::ERR_INVALID_PARAMETERS;
+    }
+
+    bool should_sort = false;
+    for (std::size_t i = 0; i < mix_count; i++) {
+        const auto& mix_in = mix_in_params[i];
+        std::size_t target_mix{};
+        if (behavior_info.IsMixInParameterDirtyOnlyUpdateSupported()) {
+            target_mix = mix_in.mix_id;
+        } else {
+            // Non dirty supported games just use i instead of the actual mix_id
+            target_mix = i;
+        }
+        auto& mix_info = mix_context.GetInfo(target_mix);
+        auto& mix_info_params = mix_info.GetInParams();
+        if (mix_info_params.in_use != mix_in.in_use) {
+            mix_info_params.in_use = mix_in.in_use;
+            mix_info.ResetEffectProcessingOrder();
+            should_sort = true;
+        }
+
+        if (mix_in.in_use) {
+            should_sort |= mix_info.Update(mix_context.GetEdgeMatrix(), mix_in, behavior_info,
+                                           splitter_context, effect_context);
+        }
+    }
+
+    if (should_sort && behavior_info.IsSplitterSupported()) {
+        // Sort our splitter data
+        if (!mix_context.TsortInfo(splitter_context)) {
+            return AudioCommon::Audren::ERR_SPLITTER_SORT_FAILED;
+        }
+    }
+
+    // TODO(ogniK): Sort when splitter is suppoorted
+
+    return RESULT_SUCCESS;
+}
+
+bool InfoUpdater::UpdateSinks(SinkContext& sink_context) {
+    const auto sink_count = sink_context.GetCount();
+    std::vector<SinkInfo::InParams> sink_in_params(sink_count);
+    const auto total_sink_in = sink_count * sizeof(SinkInfo::InParams);
+
+    if (input_header.size.sink != total_sink_in) {
+        LOG_ERROR(Audio, "Sinks are an invalid size, expecting 0x{:X} but got 0x{:X}",
+                  total_sink_in, input_header.size.effect);
+        return false;
+    }
+
+    if (!AudioCommon::CanConsumeBuffer(in_params.size(), input_offset, total_sink_in)) {
+        LOG_ERROR(Audio, "Buffer is an invalid size!");
+        return false;
+    }
+
+    std::memcpy(sink_in_params.data(), in_params.data() + input_offset, total_sink_in);
+    input_offset += total_sink_in;
+
+    // TODO(ogniK): Properly update sinks
+    if (!sink_in_params.empty()) {
+        sink_context.UpdateMainSink(sink_in_params[0]);
+    }
+
+    output_header.size.sink = static_cast<u32>(0x20 * sink_count);
+    output_offset += 0x20 * sink_count;
+    return true;
+}
+
+bool InfoUpdater::UpdatePerformanceBuffer() {
+    output_header.size.performance = 0x10;
+    output_offset += 0x10;
+    return true;
+}
+
+bool InfoUpdater::UpdateErrorInfo(BehaviorInfo& in_behavior_info) {
+    const auto total_beahvior_info_out = sizeof(BehaviorInfo::OutParams);
+
+    if (!AudioCommon::CanConsumeBuffer(out_params.size(), output_offset, total_beahvior_info_out)) {
+        LOG_ERROR(Audio, "Buffer is an invalid size!");
+        return false;
+    }
+
+    BehaviorInfo::OutParams behavior_info_out{};
+    behavior_info.CopyErrorInfo(behavior_info_out);
+
+    std::memcpy(out_params.data() + output_offset, &behavior_info_out, total_beahvior_info_out);
+    output_offset += total_beahvior_info_out;
+    output_header.size.behavior = total_beahvior_info_out;
+
+    return true;
+}
+
+struct RendererInfo {
+    u64_le elasped_frame_count{};
+    INSERT_PADDING_WORDS(2);
+};
+static_assert(sizeof(RendererInfo) == 0x10, "RendererInfo is an invalid size");
+
+bool InfoUpdater::UpdateRendererInfo(std::size_t elapsed_frame_count) {
+    const auto total_renderer_info_out = sizeof(RendererInfo);
+    if (!AudioCommon::CanConsumeBuffer(out_params.size(), output_offset, total_renderer_info_out)) {
+        LOG_ERROR(Audio, "Buffer is an invalid size!");
+        return false;
+    }
+    RendererInfo out{};
+    out.elasped_frame_count = elapsed_frame_count;
+    std::memcpy(out_params.data() + output_offset, &out, total_renderer_info_out);
+    output_offset += total_renderer_info_out;
+    output_header.size.render_info = total_renderer_info_out;
+
+    return true;
+}
+
+bool InfoUpdater::CheckConsumedSize() const {
+    if (output_offset != out_params.size()) {
+        LOG_ERROR(Audio, "Output is not consumed! Consumed {}, but requires {}. {} bytes remaining",
+                  output_offset, out_params.size(), out_params.size() - output_offset);
+        return false;
+    }
+    /*if (input_offset != in_params.size()) {
+        LOG_ERROR(Audio, "Input is not consumed!");
+        return false;
+    }*/
+    return true;
+}
+
+bool InfoUpdater::WriteOutputHeader() {
+    if (!AudioCommon::CanConsumeBuffer(out_params.size(), 0,
+                                       sizeof(AudioCommon::UpdateDataHeader))) {
+        LOG_ERROR(Audio, "Buffer is an invalid size!");
+        return false;
+    }
+    output_header.revision = AudioCommon::CURRENT_PROCESS_REVISION;
+    const auto& sz = output_header.size;
+    output_header.total_size += sz.behavior + sz.memory_pool + sz.voice +
+                                sz.voice_channel_resource + sz.effect + sz.mixer + sz.sink +
+                                sz.performance + sz.splitter + sz.render_info;
+
+    std::memcpy(out_params.data(), &output_header, sizeof(AudioCommon::UpdateDataHeader));
+    return true;
+}
+
+} // namespace AudioCore

src/audio_core/info_updater.h

@@ -0,0 +1,58 @@
+// Copyright 2020 yuzu Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#pragma once
+
+#include <vector>
+#include "audio_core/common.h"
+#include "common/common_types.h"
+
+namespace AudioCore {
+
+class BehaviorInfo;
+class ServerMemoryPoolInfo;
+class VoiceContext;
+class EffectContext;
+class MixContext;
+class SinkContext;
+class SplitterContext;
+
+class InfoUpdater {
+public:
+    // TODO(ogniK): Pass process handle when we support it
+    InfoUpdater(const std::vector<u8>& in_params, std::vector<u8>& out_params,
+                BehaviorInfo& behavior_info);
+    ~InfoUpdater();
+
+    bool UpdateBehaviorInfo(BehaviorInfo& in_behavior_info);
+    bool UpdateMemoryPools(std::vector<ServerMemoryPoolInfo>& memory_pool_info);
+    bool UpdateVoiceChannelResources(VoiceContext& voice_context);
+    bool UpdateVoices(VoiceContext& voice_context,
+                      std::vector<ServerMemoryPoolInfo>& memory_pool_info,
+                      VAddr audio_codec_dsp_addr);
+    bool UpdateEffects(EffectContext& effect_context, bool is_active);
+    bool UpdateSplitterInfo(SplitterContext& splitter_context);
+    ResultCode UpdateMixes(MixContext& mix_context, std::size_t mix_buffer_count,
+                           SplitterContext& splitter_context, EffectContext& effect_context);
+    bool UpdateSinks(SinkContext& sink_context);
+    bool UpdatePerformanceBuffer();
+    bool UpdateErrorInfo(BehaviorInfo& in_behavior_info);
+    bool UpdateRendererInfo(std::size_t elapsed_frame_count);
+    bool CheckConsumedSize() const;
+
+    bool WriteOutputHeader();
+
+private:
+    const std::vector<u8>& in_params;
+    std::vector<u8>& out_params;
+    BehaviorInfo& behavior_info;
+
+    AudioCommon::UpdateDataHeader input_header{};
+    AudioCommon::UpdateDataHeader output_header{};
+
+    std::size_t input_offset{sizeof(AudioCommon::UpdateDataHeader)};
+    std::size_t output_offset{sizeof(AudioCommon::UpdateDataHeader)};
+};
+
+} // namespace AudioCore

src/audio_core/memory_pool.cpp

@@ -0,0 +1,62 @@
+
+// Copyright 2020 yuzu Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#include "audio_core/memory_pool.h"
+#include "common/logging/log.h"
+
+namespace AudioCore {
+
+ServerMemoryPoolInfo::ServerMemoryPoolInfo() = default;
+ServerMemoryPoolInfo::~ServerMemoryPoolInfo() = default;
+bool ServerMemoryPoolInfo::Update(const ServerMemoryPoolInfo::InParams& in_params,
+                                  ServerMemoryPoolInfo::OutParams& out_params) {
+    // Our state does not need to be changed
+    if (in_params.state != ServerMemoryPoolInfo::State::RequestAttach &&
+        in_params.state != ServerMemoryPoolInfo::State::RequestDetach) {
+        return true;
+    }
+
+    // Address or size is null
+    if (in_params.address == 0 || in_params.size == 0) {
+        LOG_ERROR(Audio, "Memory pool address or size is zero! address={:X}, size={:X}",
+                  in_params.address, in_params.size);
+        return false;
+    }
+
+    // Address or size is not aligned
+    if ((in_params.address % 0x1000) != 0 || (in_params.size % 0x1000) != 0) {
+        LOG_ERROR(Audio, "Memory pool address or size is not aligned! address={:X}, size={:X}",
+                  in_params.address, in_params.size);
+        return false;
+    }
+
+    if (in_params.state == ServerMemoryPoolInfo::State::RequestAttach) {
+        cpu_address = in_params.address;
+        size = in_params.size;
+        used = true;
+        out_params.state = ServerMemoryPoolInfo::State::Attached;
+    } else {
+        // Unexpected address
+        if (cpu_address != in_params.address) {
+            LOG_ERROR(Audio, "Memory pool address differs! Expecting {:X} but address is {:X}",
+                      cpu_address, in_params.address);
+            return false;
+        }
+
+        if (size != in_params.size) {
+            LOG_ERROR(Audio, "Memory pool size differs! Expecting {:X} but size is {:X}", size,
+                      in_params.size);
+            return false;
+        }
+
+        cpu_address = 0;
+        size = 0;
+        used = false;
+        out_params.state = ServerMemoryPoolInfo::State::Detached;
+    }
+    return true;
+}
+
+} // namespace AudioCore

src/audio_core/memory_pool.h

@@ -0,0 +1,53 @@
+// Copyright 2020 yuzu Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#pragma once
+
+#include "common/common_funcs.h"
+#include "common/common_types.h"
+#include "common/swap.h"
+
+namespace AudioCore {
+
+class ServerMemoryPoolInfo {
+public:
+    ServerMemoryPoolInfo();
+    ~ServerMemoryPoolInfo();
+
+    enum class State : u32_le {
+        Invalid = 0x0,
+        Aquired = 0x1,
+        RequestDetach = 0x2,
+        Detached = 0x3,
+        RequestAttach = 0x4,
+        Attached = 0x5,
+        Released = 0x6,
+    };
+
+    struct InParams {
+        u64_le address{};
+        u64_le size{};
+        ServerMemoryPoolInfo::State state{};
+        INSERT_PADDING_WORDS(3);
+    };
+    static_assert(sizeof(ServerMemoryPoolInfo::InParams) == 0x20, "InParams are an invalid size");
+
+    struct OutParams {
+        ServerMemoryPoolInfo::State state{};
+        INSERT_PADDING_WORDS(3);
+    };
+    static_assert(sizeof(ServerMemoryPoolInfo::OutParams) == 0x10, "OutParams are an invalid size");
+
+    bool Update(const ServerMemoryPoolInfo::InParams& in_params,
+                ServerMemoryPoolInfo::OutParams& out_params);
+
+private:
+    // There's another entry here which is the DSP address, however since we're not talking to the
+    // DSP we can just use the same address provided by the guest without needing to remap
+    u64_le cpu_address{};
+    u64_le size{};
+    bool used{};
+};
+
+} // namespace AudioCore

src/audio_core/mix_context.cpp

@@ -0,0 +1,296 @@
+// Copyright 2020 yuzu Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#include "audio_core/behavior_info.h"
+#include "audio_core/common.h"
+#include "audio_core/effect_context.h"
+#include "audio_core/mix_context.h"
+#include "audio_core/splitter_context.h"
+
+namespace AudioCore {
+MixContext::MixContext() = default;
+MixContext::~MixContext() = default;
+
+void MixContext::Initialize(const BehaviorInfo& behavior_info, std::size_t mix_count,
+                            std::size_t effect_count) {
+    info_count = mix_count;
+    infos.resize(info_count);
+    auto& final_mix = GetInfo(AudioCommon::FINAL_MIX);
+    final_mix.GetInParams().mix_id = AudioCommon::FINAL_MIX;
+    sorted_info.reserve(infos.size());
+    for (auto& info : infos) {
+        sorted_info.push_back(&info);
+    }
+
+    for (auto& info : infos) {
+        info.SetEffectCount(effect_count);
+    }
+
+    // Only initialize our edge matrix and node states if splitters are supported
+    if (behavior_info.IsSplitterSupported()) {
+        node_states.Initialize(mix_count);
+        edge_matrix.Initialize(mix_count);
+    }
+}
+
+void MixContext::UpdateDistancesFromFinalMix() {
+    // Set all distances to be invalid
+    for (std::size_t i = 0; i < info_count; i++) {
+        GetInfo(i).GetInParams().final_mix_distance = AudioCommon::NO_FINAL_MIX;
+    }
+
+    for (std::size_t i = 0; i < info_count; i++) {
+        auto& info = GetInfo(i);
+        auto& in_params = info.GetInParams();
+        // Populate our sorted info
+        sorted_info[i] = &info;
+
+        if (!in_params.in_use) {
+            continue;
+        }
+
+        auto mix_id = in_params.mix_id;
+        // Needs to be referenced out of scope
+        s32 distance_to_final_mix{AudioCommon::FINAL_MIX};
+        for (; distance_to_final_mix < static_cast<s32>(info_count); distance_to_final_mix++) {
+            if (mix_id == AudioCommon::FINAL_MIX) {
+                // If we're at the final mix, we're done
+                break;
+            } else if (mix_id == AudioCommon::NO_MIX) {
+                // If we have no more mix ids, we're done
+                distance_to_final_mix = AudioCommon::NO_FINAL_MIX;
+                break;
+            } else {
+                const auto& dest_mix = GetInfo(mix_id);
+                const auto dest_mix_distance = dest_mix.GetInParams().final_mix_distance;
+
+                if (dest_mix_distance == AudioCommon::NO_FINAL_MIX) {
+                    // If our current mix isn't pointing to a final mix, follow through
+                    mix_id = dest_mix.GetInParams().dest_mix_id;
+                } else {
+                    // Our current mix + 1 = final distance
+                    distance_to_final_mix = dest_mix_distance + 1;
+                    break;
+                }
+            }
+        }
+
+        // If we're out of range for our distance, mark it as no final mix
+        if (distance_to_final_mix >= static_cast<s32>(info_count)) {
+            distance_to_final_mix = AudioCommon::NO_FINAL_MIX;
+        }
+
+        in_params.final_mix_distance = distance_to_final_mix;
+    }
+}
+
+void MixContext::CalcMixBufferOffset() {
+    s32 offset{};
+    for (std::size_t i = 0; i < info_count; i++) {
+        auto& info = GetSortedInfo(i);
+        auto& in_params = info.GetInParams();
+        if (in_params.in_use) {
+            // Only update if in use
+            in_params.buffer_offset = offset;
+            offset += in_params.buffer_count;
+        }
+    }
+}
+
+void MixContext::SortInfo() {
+    // Get the distance to the final mix
+    UpdateDistancesFromFinalMix();
+
+    // Sort based on the distance to the final mix
+    std::sort(sorted_info.begin(), sorted_info.end(),
+              [](const ServerMixInfo* lhs, const ServerMixInfo* rhs) {
+                  return lhs->GetInParams().final_mix_distance >
+                         rhs->GetInParams().final_mix_distance;
+              });
+
+    // Calculate the mix buffer offset
+    CalcMixBufferOffset();
+}
+
+bool MixContext::TsortInfo(SplitterContext& splitter_context) {
+    // If we're not using mixes, just calculate the mix buffer offset
+    if (!splitter_context.UsingSplitter()) {
+        CalcMixBufferOffset();
+        return true;
+    }
+    // Sort our node states
+    if (!node_states.Tsort(edge_matrix)) {
+        return false;
+    }
+
+    // Get our sorted list
+    const auto sorted_list = node_states.GetIndexList();
+    std::size_t info_id{};
+    for (auto itr = sorted_list.rbegin(); itr != sorted_list.rend(); ++itr) {
+        // Set our sorted info
+        sorted_info[info_id++] = &GetInfo(*itr);
+    }
+
+    // Calculate the mix buffer offset
+    CalcMixBufferOffset();
+    return true;
+}
+
+std::size_t MixContext::GetCount() const {
+    return info_count;
+}
+
+ServerMixInfo& MixContext::GetInfo(std::size_t i) {
+    ASSERT(i < info_count);
+    return infos.at(i);
+}
+
+const ServerMixInfo& MixContext::GetInfo(std::size_t i) const {
+    ASSERT(i < info_count);
+    return infos.at(i);
+}
+
+ServerMixInfo& MixContext::GetSortedInfo(std::size_t i) {
+    ASSERT(i < info_count);
+    return *sorted_info.at(i);
+}
+
+const ServerMixInfo& MixContext::GetSortedInfo(std::size_t i) const {
+    ASSERT(i < info_count);
+    return *sorted_info.at(i);
+}
+
+ServerMixInfo& MixContext::GetFinalMixInfo() {
+    return infos.at(AudioCommon::FINAL_MIX);
+}
+
+const ServerMixInfo& MixContext::GetFinalMixInfo() const {
+    return infos.at(AudioCommon::FINAL_MIX);
+}
+
+EdgeMatrix& MixContext::GetEdgeMatrix() {
+    return edge_matrix;
+}
+
+const EdgeMatrix& MixContext::GetEdgeMatrix() const {
+    return edge_matrix;
+}
+
+ServerMixInfo::ServerMixInfo() {
+    Cleanup();
+}
+ServerMixInfo::~ServerMixInfo() = default;
+
+const ServerMixInfo::InParams& ServerMixInfo::GetInParams() const {
+    return in_params;
+}
+
+ServerMixInfo::InParams& ServerMixInfo::GetInParams() {
+    return in_params;
+}
+
+bool ServerMixInfo::Update(EdgeMatrix& edge_matrix, const MixInfo::InParams& mix_in,
+                           BehaviorInfo& behavior_info, SplitterContext& splitter_context,
+                           EffectContext& effect_context) {
+    in_params.volume = mix_in.volume;
+    in_params.sample_rate = mix_in.sample_rate;
+    in_params.buffer_count = mix_in.buffer_count;
+    in_params.in_use = mix_in.in_use;
+    in_params.mix_id = mix_in.mix_id;
+    in_params.node_id = mix_in.node_id;
+    for (std::size_t i = 0; i < mix_in.mix_volume.size(); i++) {
+        std::copy(mix_in.mix_volume[i].begin(), mix_in.mix_volume[i].end(),
+                  in_params.mix_volume[i].begin());
+    }
+
+    bool require_sort = false;
+
+    if (behavior_info.IsSplitterSupported()) {
+        require_sort = UpdateConnection(edge_matrix, mix_in, splitter_context);
+    } else {
+        in_params.dest_mix_id = mix_in.dest_mix_id;
+        in_params.splitter_id = AudioCommon::NO_SPLITTER;
+    }
+
+    ResetEffectProcessingOrder();
+    const auto effect_count = effect_context.GetCount();
+    for (std::size_t i = 0; i < effect_count; i++) {
+        auto* effect_info = effect_context.GetInfo(i);
+        if (effect_info->GetMixID() == in_params.mix_id) {
+            effect_processing_order[effect_info->GetProcessingOrder()] = static_cast<s32>(i);
+        }
+    }
+
+    // TODO(ogniK): Update effect processing order
+    return require_sort;
+}
+
+bool ServerMixInfo::HasAnyConnection() const {
+    return in_params.splitter_id != AudioCommon::NO_SPLITTER ||
+           in_params.mix_id != AudioCommon::NO_MIX;
+}
+
+void ServerMixInfo::Cleanup() {
+    in_params.volume = 0.0f;
+    in_params.sample_rate = 0;
+    in_params.buffer_count = 0;
+    in_params.in_use = false;
+    in_params.mix_id = AudioCommon::NO_MIX;
+    in_params.node_id = 0;
+    in_params.buffer_offset = 0;
+    in_params.dest_mix_id = AudioCommon::NO_MIX;
+    in_params.splitter_id = AudioCommon::NO_SPLITTER;
+    std::memset(in_params.mix_volume.data(), 0, sizeof(float) * in_params.mix_volume.size());
+}
+
+void ServerMixInfo::SetEffectCount(std::size_t count) {
+    effect_processing_order.resize(count);
+    ResetEffectProcessingOrder();
+}
+
+void ServerMixInfo::ResetEffectProcessingOrder() {
+    for (auto& order : effect_processing_order) {
+        order = AudioCommon::NO_EFFECT_ORDER;
+    }
+}
+
+s32 ServerMixInfo::GetEffectOrder(std::size_t i) const {
+    return effect_processing_order.at(i);
+}
+
+bool ServerMixInfo::UpdateConnection(EdgeMatrix& edge_matrix, const MixInfo::InParams& mix_in,
+                                     SplitterContext& splitter_context) {
+    // Mixes are identical
+    if (in_params.dest_mix_id == mix_in.dest_mix_id &&
+        in_params.splitter_id == mix_in.splitter_id &&
+        ((in_params.splitter_id == AudioCommon::NO_SPLITTER) ||
+         !splitter_context.GetInfo(in_params.splitter_id).HasNewConnection())) {
+        return false;
+    }
+    // Remove current edges for mix id
+    edge_matrix.RemoveEdges(in_params.mix_id);
+    if (mix_in.dest_mix_id != AudioCommon::NO_MIX) {
+        // If we have a valid destination mix id, set our edge matrix
+        edge_matrix.Connect(in_params.mix_id, mix_in.dest_mix_id);
+    } else if (mix_in.splitter_id != AudioCommon::NO_SPLITTER) {
+        // Recurse our splitter linked and set our edges
+        auto& splitter_info = splitter_context.GetInfo(mix_in.splitter_id);
+        const auto length = splitter_info.GetLength();
+        for (s32 i = 0; i < length; i++) {
+            const auto* splitter_destination =
+                splitter_context.GetDestinationData(mix_in.splitter_id, i);
+            if (splitter_destination == nullptr) {
+                continue;
+            }
+            if (splitter_destination->ValidMixId()) {
+                edge_matrix.Connect(in_params.mix_id, splitter_destination->GetMixId());
+            }
+        }
+    }
+    in_params.dest_mix_id = mix_in.dest_mix_id;
+    in_params.splitter_id = mix_in.splitter_id;
+    return true;
+}
+
+} // namespace AudioCore

src/audio_core/mix_context.h

@@ -0,0 +1,114 @@
+// Copyright 2020 yuzu Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#pragma once
+
+#include <array>
+#include <vector>
+#include "audio_core/common.h"
+#include "audio_core/splitter_context.h"
+#include "common/common_funcs.h"
+#include "common/common_types.h"
+
+namespace AudioCore {
+class BehaviorInfo;
+class EffectContext;
+
+class MixInfo {
+public:
+    struct DirtyHeader {
+        u32_le magic{};
+        u32_le mixer_count{};
+        INSERT_PADDING_BYTES(0x18);
+    };
+    static_assert(sizeof(DirtyHeader) == 0x20, "MixInfo::DirtyHeader is an invalid size");
+
+    struct InParams {
+        float_le volume{};
+        s32_le sample_rate{};
+        s32_le buffer_count{};
+        bool in_use{};
+        INSERT_PADDING_BYTES(3);
+        s32_le mix_id{};
+        s32_le effect_count{};
+        u32_le node_id{};
+        INSERT_PADDING_WORDS(2);
+        std::array<std::array<float_le, AudioCommon::MAX_MIX_BUFFERS>, AudioCommon::MAX_MIX_BUFFERS>
+            mix_volume{};
+        s32_le dest_mix_id{};
+        s32_le splitter_id{};
+        INSERT_PADDING_WORDS(1);
+    };
+    static_assert(sizeof(MixInfo::InParams) == 0x930, "MixInfo::InParams is an invalid size");
+};
+
+class ServerMixInfo {
+public:
+    struct InParams {
+        float volume{};
+        s32 sample_rate{};
+        s32 buffer_count{};
+        bool in_use{};
+        s32 mix_id{};
+        u32 node_id{};
+        std::array<std::array<float_le, AudioCommon::MAX_MIX_BUFFERS>, AudioCommon::MAX_MIX_BUFFERS>
+            mix_volume{};
+        s32 dest_mix_id{};
+        s32 splitter_id{};
+        s32 buffer_offset{};
+        s32 final_mix_distance{};
+    };
+    ServerMixInfo();
+    ~ServerMixInfo();
+
+    const ServerMixInfo::InParams& GetInParams() const;
+    ServerMixInfo::InParams& GetInParams();
+
+    bool Update(EdgeMatrix& edge_matrix, const MixInfo::InParams& mix_in,
+                BehaviorInfo& behavior_info, SplitterContext& splitter_context,
+                EffectContext& effect_context);
+    bool HasAnyConnection() const;
+    void Cleanup();
+    void SetEffectCount(std::size_t count);
+    void ResetEffectProcessingOrder();
+    s32 GetEffectOrder(std::size_t i) const;
+
+private:
+    std::vector<s32> effect_processing_order;
+    InParams in_params{};
+    bool UpdateConnection(EdgeMatrix& edge_matrix, const MixInfo::InParams& mix_in,
+                          SplitterContext& splitter_context);
+};
+
+class MixContext {
+public:
+    MixContext();
+    ~MixContext();
+
+    void Initialize(const BehaviorInfo& behavior_info, std::size_t mix_count,
+                    std::size_t effect_count);
+    void SortInfo();
+    bool TsortInfo(SplitterContext& splitter_context);
+
+    std::size_t GetCount() const;
+    ServerMixInfo& GetInfo(std::size_t i);
+    const ServerMixInfo& GetInfo(std::size_t i) const;
+    ServerMixInfo& GetSortedInfo(std::size_t i);
+    const ServerMixInfo& GetSortedInfo(std::size_t i) const;
+    ServerMixInfo& GetFinalMixInfo();
+    const ServerMixInfo& GetFinalMixInfo() const;
+    EdgeMatrix& GetEdgeMatrix();
+    const EdgeMatrix& GetEdgeMatrix() const;
+
+private:
+    void CalcMixBufferOffset();
+    void UpdateDistancesFromFinalMix();
+
+    NodeStates node_states{};
+    EdgeMatrix edge_matrix{};
+    std::size_t info_count{};
+    std::vector<ServerMixInfo> infos{};
+    std::vector<ServerMixInfo*> sorted_info{};
+};
+} // namespace AudioCore

src/audio_core/sink_context.cpp

@@ -0,0 +1,31 @@
+// Copyright 2020 yuzu Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#include "audio_core/sink_context.h"
+
+namespace AudioCore {
+SinkContext::SinkContext(std::size_t sink_count) : sink_count(sink_count) {}
+SinkContext::~SinkContext() = default;
+
+std::size_t SinkContext::GetCount() const {
+    return sink_count;
+}
+
+void SinkContext::UpdateMainSink(SinkInfo::InParams& in) {
+    in_use = in.in_use;
+    use_count = in.device.input_count;
+    std::memcpy(buffers.data(), in.device.input.data(), AudioCommon::MAX_CHANNEL_COUNT);
+}
+
+bool SinkContext::InUse() const {
+    return in_use;
+}
+
+std::vector<u8> SinkContext::OutputBuffers() const {
+    std::vector<u8> buffer_ret(use_count);
+    std::memcpy(buffer_ret.data(), buffers.data(), use_count);
+    return buffer_ret;
+}
+
+} // namespace AudioCore

src/audio_core/sink_context.h

@@ -0,0 +1,89 @@
+// Copyright 2020 yuzu Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#pragma once
+
+#include "audio_core/common.h"
+#include "common/common_funcs.h"
+#include "common/common_types.h"
+#include "common/swap.h"
+
+namespace AudioCore {
+
+enum class SinkTypes : u8 {
+    Invalid = 0,
+    Device = 1,
+    Circular = 2,
+};
+
+enum class SinkSampleFormat : u32_le {
+    None = 0,
+    Pcm8 = 1,
+    Pcm16 = 2,
+    Pcm24 = 3,
+    Pcm32 = 4,
+    PcmFloat = 5,
+    Adpcm = 6,
+};
+
+class SinkInfo {
+public:
+    struct CircularBufferIn {
+        u64_le address;
+        u32_le size;
+        u32_le input_count;
+        u32_le sample_count;
+        u32_le previous_position;
+        SinkSampleFormat sample_format;
+        std::array<u8, AudioCommon::MAX_CHANNEL_COUNT> input;
+        bool in_use;
+        INSERT_UNION_PADDING_BYTES(5);
+    };
+    static_assert(sizeof(SinkInfo::CircularBufferIn) == 0x28,
+                  "SinkInfo::CircularBufferIn is in invalid size");
+
+    struct DeviceIn {
+        std::array<u8, 255> device_name;
+        INSERT_UNION_PADDING_BYTES(1);
+        s32_le input_count;
+        std::array<u8, AudioCommon::MAX_CHANNEL_COUNT> input;
+        INSERT_UNION_PADDING_BYTES(1);
+        bool down_matrix_enabled;
+        std::array<float_le, 4> down_matrix_coef;
+    };
+    static_assert(sizeof(SinkInfo::DeviceIn) == 0x11c, "SinkInfo::DeviceIn is an invalid size");
+
+    struct InParams {
+        SinkTypes type{};
+        bool in_use{};
+        INSERT_PADDING_BYTES(2);
+        u32_le node_id{};
+        INSERT_PADDING_WORDS(6);
+        union {
+            // std::array<u8, 0x120> raw{};
+            SinkInfo::DeviceIn device;
+            SinkInfo::CircularBufferIn circular_buffer;
+        };
+    };
+    static_assert(sizeof(SinkInfo::InParams) == 0x140, "SinkInfo::InParams are an invalid size!");
+};
+
+class SinkContext {
+public:
+    explicit SinkContext(std::size_t sink_count);
+    ~SinkContext();
+
+    std::size_t GetCount() const;
+
+    void UpdateMainSink(SinkInfo::InParams& in);
+    bool InUse() const;
+    std::vector<u8> OutputBuffers() const;
+
+private:
+    bool in_use{false};
+    s32 use_count{};
+    std::array<u8, AudioCommon::MAX_CHANNEL_COUNT> buffers{};
+    std::size_t sink_count{};
+};
+} // namespace AudioCore

src/audio_core/splitter_context.cpp

@@ -0,0 +1,617 @@
+// Copyright 2020 yuzu Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#include "audio_core/behavior_info.h"
+#include "audio_core/splitter_context.h"
+#include "common/alignment.h"
+#include "common/assert.h"
+#include "common/logging/log.h"
+
+namespace AudioCore {
+
+ServerSplitterDestinationData::ServerSplitterDestinationData(s32 id) : id(id) {}
+ServerSplitterDestinationData::~ServerSplitterDestinationData() = default;
+
+void ServerSplitterDestinationData::Update(SplitterInfo::InDestinationParams& header) {
+    // Log error as these are not actually failure states
+    if (header.magic != SplitterMagic::DataHeader) {
+        LOG_ERROR(Audio, "Splitter destination header is invalid!");
+        return;
+    }
+
+    // Incorrect splitter id
+    if (header.splitter_id != id) {
+        LOG_ERROR(Audio, "Splitter destination ids do not match!");
+        return;
+    }
+
+    mix_id = header.mix_id;
+    // Copy our mix volumes
+    std::copy(header.mix_volumes.begin(), header.mix_volumes.end(), current_mix_volumes.begin());
+    if (!in_use && header.in_use) {
+        // Update mix volumes
+        std::copy(current_mix_volumes.begin(), current_mix_volumes.end(), last_mix_volumes.begin());
+        needs_update = false;
+    }
+    in_use = header.in_use;
+}
+
+ServerSplitterDestinationData* ServerSplitterDestinationData::GetNextDestination() {
+    return next;
+}
+
+const ServerSplitterDestinationData* ServerSplitterDestinationData::GetNextDestination() const {
+    return next;
+}
+
+void ServerSplitterDestinationData::SetNextDestination(ServerSplitterDestinationData* dest) {
+    next = dest;
+}
+
+bool ServerSplitterDestinationData::ValidMixId() const {
+    return GetMixId() != AudioCommon::NO_MIX;
+}
+
+s32 ServerSplitterDestinationData::GetMixId() const {
+    return mix_id;
+}
+
+bool ServerSplitterDestinationData::IsConfigured() const {
+    return in_use && ValidMixId();
+}
+
+float ServerSplitterDestinationData::GetMixVolume(std::size_t i) const {
+    ASSERT(i < AudioCommon::MAX_MIX_BUFFERS);
+    return current_mix_volumes.at(i);
+}
+
+const std::array<float, AudioCommon::MAX_MIX_BUFFERS>&
+ServerSplitterDestinationData::CurrentMixVolumes() const {
+    return current_mix_volumes;
+}
+
+const std::array<float, AudioCommon::MAX_MIX_BUFFERS>&
+ServerSplitterDestinationData::LastMixVolumes() const {
+    return last_mix_volumes;
+}
+
+void ServerSplitterDestinationData::MarkDirty() {
+    needs_update = true;
+}
+
+void ServerSplitterDestinationData::UpdateInternalState() {
+    if (in_use && needs_update) {
+        std::copy(current_mix_volumes.begin(), current_mix_volumes.end(), last_mix_volumes.begin());
+    }
+    needs_update = false;
+}
+
+ServerSplitterInfo::ServerSplitterInfo(s32 id) : id(id) {}
+ServerSplitterInfo::~ServerSplitterInfo() = default;
+
+void ServerSplitterInfo::InitializeInfos() {
+    send_length = 0;
+    head = nullptr;
+    new_connection = true;
+}
+
+void ServerSplitterInfo::ClearNewConnectionFlag() {
+    new_connection = false;
+}
+
+std::size_t ServerSplitterInfo::Update(SplitterInfo::InInfoPrams& header) {
+    if (header.send_id != id) {
+        return 0;
+    }
+
+    sample_rate = header.sample_rate;
+    new_connection = true;
+    // We need to update the size here due to the splitter bug being present and providing an
+    // incorrect size. We're suppose to also update the header here but we just ignore and continue
+    return (sizeof(s32_le) * (header.length - 1)) + (sizeof(s32_le) * 3);
+}
+
+ServerSplitterDestinationData* ServerSplitterInfo::GetHead() {
+    return head;
+}
+
+const ServerSplitterDestinationData* ServerSplitterInfo::GetHead() const {
+    return head;
+}
+
+ServerSplitterDestinationData* ServerSplitterInfo::GetData(std::size_t depth) {
+    auto current_head = head;
+    for (std::size_t i = 0; i < depth; i++) {
+        if (current_head == nullptr) {
+            return nullptr;
+        }
+        current_head = current_head->GetNextDestination();
+    }
+    return current_head;
+}
+
+const ServerSplitterDestinationData* ServerSplitterInfo::GetData(std::size_t depth) const {
+    auto current_head = head;
+    for (std::size_t i = 0; i < depth; i++) {
+        if (current_head == nullptr) {
+            return nullptr;
+        }
+        current_head = current_head->GetNextDestination();
+    }
+    return current_head;
+}
+
+bool ServerSplitterInfo::HasNewConnection() const {
+    return new_connection;
+}
+
+s32 ServerSplitterInfo::GetLength() const {
+    return send_length;
+}
+
+void ServerSplitterInfo::SetHead(ServerSplitterDestinationData* new_head) {
+    head = new_head;
+}
+
+void ServerSplitterInfo::SetHeadDepth(s32 length) {
+    send_length = length;
+}
+
+SplitterContext::SplitterContext() = default;
+SplitterContext::~SplitterContext() = default;
+
+void SplitterContext::Initialize(BehaviorInfo& behavior_info, std::size_t _info_count,
+                                 std::size_t _data_count) {
+    if (!behavior_info.IsSplitterSupported() || _data_count == 0 || _info_count == 0) {
+        Setup(0, 0, false);
+        return;
+    }
+    // Only initialize if we're using splitters
+    Setup(_info_count, _data_count, behavior_info.IsSplitterBugFixed());
+}
+
+bool SplitterContext::Update(const std::vector<u8>& input, std::size_t& input_offset,
+                             std::size_t& bytes_read) {
+    const auto UpdateOffsets = [&](std::size_t read) {
+        input_offset += read;
+        bytes_read += read;
+    };
+
+    if (info_count == 0 || data_count == 0) {
+        bytes_read = 0;
+        return true;
+    }
+
+    if (!AudioCommon::CanConsumeBuffer(input.size(), input_offset,
+                                       sizeof(SplitterInfo::InHeader))) {
+        LOG_ERROR(Audio, "Buffer is an invalid size!");
+        return false;
+    }
+    SplitterInfo::InHeader header{};
+    std::memcpy(&header, input.data() + input_offset, sizeof(SplitterInfo::InHeader));
+    UpdateOffsets(sizeof(SplitterInfo::InHeader));
+
+    if (header.magic != SplitterMagic::SplitterHeader) {
+        LOG_ERROR(Audio, "Invalid header magic! Expecting {:X} but got {:X}",
+                  SplitterMagic::SplitterHeader, header.magic);
+        return false;
+    }
+
+    // Clear all connections
+    for (auto& info : infos) {
+        info.ClearNewConnectionFlag();
+    }
+
+    UpdateInfo(input, input_offset, bytes_read, header.info_count);
+    UpdateData(input, input_offset, bytes_read, header.data_count);
+    const auto aligned_bytes_read = Common::AlignUp(bytes_read, 16);
+    input_offset += aligned_bytes_read - bytes_read;
+    bytes_read = aligned_bytes_read;
+    return true;
+}
+
+bool SplitterContext::UsingSplitter() const {
+    return info_count > 0 && data_count > 0;
+}
+
+ServerSplitterInfo& SplitterContext::GetInfo(std::size_t i) {
+    ASSERT(i < info_count);
+    return infos.at(i);
+}
+
+const ServerSplitterInfo& SplitterContext::GetInfo(std::size_t i) const {
+    ASSERT(i < info_count);
+    return infos.at(i);
+}
+
+ServerSplitterDestinationData& SplitterContext::GetData(std::size_t i) {
+    ASSERT(i < data_count);
+    return datas.at(i);
+}
+
+const ServerSplitterDestinationData& SplitterContext::GetData(std::size_t i) const {
+    ASSERT(i < data_count);
+    return datas.at(i);
+}
+
+ServerSplitterDestinationData* SplitterContext::GetDestinationData(std::size_t info,
+                                                                   std::size_t data) {
+    ASSERT(info < info_count);
+    auto& cur_info = GetInfo(info);
+    return cur_info.GetData(data);
+}
+
+const ServerSplitterDestinationData* SplitterContext::GetDestinationData(std::size_t info,
+                                                                         std::size_t data) const {
+    ASSERT(info < info_count);
+    auto& cur_info = GetInfo(info);
+    return cur_info.GetData(data);
+}
+
+void SplitterContext::UpdateInternalState() {
+    if (data_count == 0) {
+        return;
+    }
+
+    for (auto& data : datas) {
+        data.UpdateInternalState();
+    }
+}
+
+std::size_t SplitterContext::GetInfoCount() const {
+    return info_count;
+}
+
+std::size_t SplitterContext::GetDataCount() const {
+    return data_count;
+}
+
+void SplitterContext::Setup(std::size_t _info_count, std::size_t _data_count,
+                            bool is_splitter_bug_fixed) {
+
+    info_count = _info_count;
+    data_count = _data_count;
+
+    for (std::size_t i = 0; i < info_count; i++) {
+        auto& splitter = infos.emplace_back(static_cast<s32>(i));
+        splitter.InitializeInfos();
+    }
+    for (std::size_t i = 0; i < data_count; i++) {
+        datas.emplace_back(static_cast<s32>(i));
+    }
+
+    bug_fixed = is_splitter_bug_fixed;
+}
+
+bool SplitterContext::UpdateInfo(const std::vector<u8>& input, std::size_t& input_offset,
+                                 std::size_t& bytes_read, s32 in_splitter_count) {
+    const auto UpdateOffsets = [&](std::size_t read) {
+        input_offset += read;
+        bytes_read += read;
+    };
+
+    for (s32 i = 0; i < in_splitter_count; i++) {
+        if (!AudioCommon::CanConsumeBuffer(input.size(), input_offset,
+                                           sizeof(SplitterInfo::InInfoPrams))) {
+            LOG_ERROR(Audio, "Buffer is an invalid size!");
+            return false;
+        }
+        SplitterInfo::InInfoPrams header{};
+        std::memcpy(&header, input.data() + input_offset, sizeof(SplitterInfo::InInfoPrams));
+
+        // Logged as warning as these don't actually cause a bailout for some reason
+        if (header.magic != SplitterMagic::InfoHeader) {
+            LOG_ERROR(Audio, "Bad splitter data header");
+            break;
+        }
+
+        if (header.send_id < 0 || static_cast<std::size_t>(header.send_id) > info_count) {
+            LOG_ERROR(Audio, "Bad splitter data id");
+            break;
+        }
+
+        UpdateOffsets(sizeof(SplitterInfo::InInfoPrams));
+        auto& info = GetInfo(header.send_id);
+        if (!RecomposeDestination(info, header, input, input_offset)) {
+            LOG_ERROR(Audio, "Failed to recompose destination for splitter!");
+            return false;
+        }
+        const std::size_t read = info.Update(header);
+        bytes_read += read;
+        input_offset += read;
+    }
+    return true;
+}
+
+bool SplitterContext::UpdateData(const std::vector<u8>& input, std::size_t& input_offset,
+                                 std::size_t& bytes_read, s32 in_data_count) {
+    const auto UpdateOffsets = [&](std::size_t read) {
+        input_offset += read;
+        bytes_read += read;
+    };
+
+    for (s32 i = 0; i < in_data_count; i++) {
+        if (!AudioCommon::CanConsumeBuffer(input.size(), input_offset,
+                                           sizeof(SplitterInfo::InDestinationParams))) {
+            LOG_ERROR(Audio, "Buffer is an invalid size!");
+            return false;
+        }
+        SplitterInfo::InDestinationParams header{};
+        std::memcpy(&header, input.data() + input_offset,
+                    sizeof(SplitterInfo::InDestinationParams));
+        UpdateOffsets(sizeof(SplitterInfo::InDestinationParams));
+
+        // Logged as warning as these don't actually cause a bailout for some reason
+        if (header.magic != SplitterMagic::DataHeader) {
+            LOG_ERROR(Audio, "Bad splitter data header");
+            break;
+        }
+
+        if (header.splitter_id < 0 || static_cast<std::size_t>(header.splitter_id) > data_count) {
+            LOG_ERROR(Audio, "Bad splitter data id");
+            break;
+        }
+        GetData(header.splitter_id).Update(header);
+    }
+    return true;
+}
+
+bool SplitterContext::RecomposeDestination(ServerSplitterInfo& info,
+                                           SplitterInfo::InInfoPrams& header,
+                                           const std::vector<u8>& input,
+                                           const std::size_t& input_offset) {
+    // Clear our current destinations
+    auto* current_head = info.GetHead();
+    while (current_head != nullptr) {
+        auto next_head = current_head->GetNextDestination();
+        current_head->SetNextDestination(nullptr);
+        current_head = next_head;
+    }
+    info.SetHead(nullptr);
+
+    s32 size = header.length;
+    // If the splitter bug is present, calculate fixed size
+    if (!bug_fixed) {
+        if (info_count > 0) {
+            const auto factor = data_count / info_count;
+            size = std::min(header.length, static_cast<s32>(factor));
+        } else {
+            size = 0;
+        }
+    }
+
+    if (size < 1) {
+        LOG_ERROR(Audio, "Invalid splitter info size! size={:X}", size);
+        return true;
+    }
+
+    auto* start_head = &GetData(header.resource_id_base);
+    current_head = start_head;
+    std::vector<s32_le> resource_ids(size - 1);
+    if (!AudioCommon::CanConsumeBuffer(input.size(), input_offset,
+                                       resource_ids.size() * sizeof(s32_le))) {
+        LOG_ERROR(Audio, "Buffer is an invalid size!");
+        return false;
+    }
+    std::memcpy(resource_ids.data(), input.data() + input_offset,
+                resource_ids.size() * sizeof(s32_le));
+
+    for (auto resource_id : resource_ids) {
+        auto* head = &GetData(resource_id);
+        current_head->SetNextDestination(head);
+        current_head = head;
+    }
+
+    info.SetHead(start_head);
+    info.SetHeadDepth(size);
+
+    return true;
+}
+
+NodeStates::NodeStates() = default;
+NodeStates::~NodeStates() = default;
+
+void NodeStates::Initialize(std::size_t node_count_) {
+    // Setup our work parameters
+    node_count = node_count_;
+    was_node_found.resize(node_count);
+    was_node_completed.resize(node_count);
+    index_list.resize(node_count);
+    index_stack.Reset(node_count * node_count);
+}
+
+bool NodeStates::Tsort(EdgeMatrix& edge_matrix) {
+    return DepthFirstSearch(edge_matrix);
+}
+
+std::size_t NodeStates::GetIndexPos() const {
+    return index_pos;
+}
+
+const std::vector<s32>& NodeStates::GetIndexList() const {
+    return index_list;
+}
+
+void NodeStates::PushTsortResult(s32 index) {
+    ASSERT(index < static_cast<s32>(node_count));
+    index_list[index_pos++] = index;
+}
+
+bool NodeStates::DepthFirstSearch(EdgeMatrix& edge_matrix) {
+    ResetState();
+    for (std::size_t i = 0; i < node_count; i++) {
+        const auto node_id = static_cast<s32>(i);
+
+        // If we don't have a state, send to our index stack for work
+        if (GetState(i) == NodeStates::State::NoState) {
+            index_stack.push(node_id);
+        }
+
+        // While we have work to do in our stack
+        while (index_stack.Count() > 0) {
+            // Get the current node
+            const auto current_stack_index = index_stack.top();
+            // Check if we've seen the node yet
+            const auto index_state = GetState(current_stack_index);
+            if (index_state == NodeStates::State::NoState) {
+                // Mark the node as seen
+                UpdateState(NodeStates::State::InFound, current_stack_index);
+            } else if (index_state == NodeStates::State::InFound) {
+                // We've seen this node before, mark it as completed
+                UpdateState(NodeStates::State::InCompleted, current_stack_index);
+                // Update our index list
+                PushTsortResult(current_stack_index);
+                // Pop the stack
+                index_stack.pop();
+                continue;
+            } else if (index_state == NodeStates::State::InCompleted) {
+                // If our node is already sorted, clear it
+                index_stack.pop();
+                continue;
+            }
+
+            const auto node_count = edge_matrix.GetNodeCount();
+            for (s32 j = 0; j < static_cast<s32>(node_count); j++) {
+                // Check if our node is connected to our edge matrix
+                if (!edge_matrix.Connected(current_stack_index, j)) {
+                    continue;
+                }
+
+                // Check if our node exists
+                const auto node_state = GetState(j);
+                if (node_state == NodeStates::State::NoState) {
+                    // Add more work
+                    index_stack.push(j);
+                } else if (node_state == NodeStates::State::InFound) {
+                    UNREACHABLE_MSG("Node start marked as found");
+                    ResetState();
+                    return false;
+                }
+            }
+        }
+    }
+    return true;
+}
+
+void NodeStates::ResetState() {
+    // Reset to the start of our index stack
+    index_pos = 0;
+    for (std::size_t i = 0; i < node_count; i++) {
+        // Mark all nodes as not found
+        was_node_found[i] = false;
+        // Mark all nodes as uncompleted
+        was_node_completed[i] = false;
+        // Mark all indexes as invalid
+        index_list[i] = -1;
+    }
+}
+
+void NodeStates::UpdateState(NodeStates::State state, std::size_t i) {
+    switch (state) {
+    case NodeStates::State::NoState:
+        was_node_found[i] = false;
+        was_node_completed[i] = false;
+        break;
+    case NodeStates::State::InFound:
+        was_node_found[i] = true;
+        was_node_completed[i] = false;
+        break;
+    case NodeStates::State::InCompleted:
+        was_node_found[i] = false;
+        was_node_completed[i] = true;
+        break;
+    }
+}
+
+NodeStates::State NodeStates::GetState(std::size_t i) {
+    ASSERT(i < node_count);
+    if (was_node_found[i]) {
+        // If our node exists in our found list
+        return NodeStates::State::InFound;
+    } else if (was_node_completed[i]) {
+        // If node is in the completed list
+        return NodeStates::State::InCompleted;
+    } else {
+        // If in neither
+        return NodeStates::State::NoState;
+    }
+}
+
+NodeStates::Stack::Stack() = default;
+NodeStates::Stack::~Stack() = default;
+
+void NodeStates::Stack::Reset(std::size_t size) {
+    // Mark our stack as empty
+    stack.resize(size);
+    stack_size = size;
+    stack_pos = 0;
+    std::fill(stack.begin(), stack.end(), 0);
+}
+
+void NodeStates::Stack::push(s32 val) {
+    ASSERT(stack_pos < stack_size);
+    stack[stack_pos++] = val;
+}
+
+std::size_t NodeStates::Stack::Count() const {
+    return stack_pos;
+}
+
+s32 NodeStates::Stack::top() const {
+    ASSERT(stack_pos > 0);
+    return stack[stack_pos - 1];
+}
+
+s32 NodeStates::Stack::pop() {
+    ASSERT(stack_pos > 0);
+    stack_pos--;
+    return stack[stack_pos];
+}
+
+EdgeMatrix::EdgeMatrix() = default;
+EdgeMatrix::~EdgeMatrix() = default;
+
+void EdgeMatrix::Initialize(std::size_t _node_count) {
+    node_count = _node_count;
+    edge_matrix.resize(node_count * node_count);
+}
+
+bool EdgeMatrix::Connected(s32 a, s32 b) {
+    return GetState(a, b);
+}
+
+void EdgeMatrix::Connect(s32 a, s32 b) {
+    SetState(a, b, true);
+}
+
+void EdgeMatrix::Disconnect(s32 a, s32 b) {
+    SetState(a, b, false);
+}
+
+void EdgeMatrix::RemoveEdges(s32 edge) {
+    for (std::size_t i = 0; i < node_count; i++) {
+        SetState(edge, static_cast<s32>(i), false);
+    }
+}
+
+std::size_t EdgeMatrix::GetNodeCount() const {
+    return node_count;
+}
+
+void EdgeMatrix::SetState(s32 a, s32 b, bool state) {
+    ASSERT(InRange(a, b));
+    edge_matrix.at(a * node_count + b) = state;
+}
+
+bool EdgeMatrix::GetState(s32 a, s32 b) {
+    ASSERT(InRange(a, b));
+    return edge_matrix.at(a * node_count + b);
+}
+
+bool EdgeMatrix::InRange(s32 a, s32 b) const {
+    const std::size_t pos = a * node_count + b;
+    return pos < (node_count * node_count);
+}
+
+} // namespace AudioCore

src/audio_core/splitter_context.h

@@ -0,0 +1,221 @@
+// Copyright 2020 yuzu Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#pragma once
+
+#include <stack>
+#include <vector>
+#include "audio_core/common.h"
+#include "common/common_funcs.h"
+#include "common/common_types.h"
+#include "common/swap.h"
+
+namespace AudioCore {
+class BehaviorInfo;
+
+class EdgeMatrix {
+public:
+    EdgeMatrix();
+    ~EdgeMatrix();
+
+    void Initialize(std::size_t _node_count);
+    bool Connected(s32 a, s32 b);
+    void Connect(s32 a, s32 b);
+    void Disconnect(s32 a, s32 b);
+    void RemoveEdges(s32 edge);
+    std::size_t GetNodeCount() const;
+
+private:
+    void SetState(s32 a, s32 b, bool state);
+    bool GetState(s32 a, s32 b);
+
+    bool InRange(s32 a, s32 b) const;
+    std::vector<bool> edge_matrix{};
+    std::size_t node_count{};
+};
+
+class NodeStates {
+public:
+    enum class State {
+        NoState = 0,
+        InFound = 1,
+        InCompleted = 2,
+    };
+
+    // Looks to be a fixed size stack. Placed within the NodeStates class based on symbols
+    class Stack {
+    public:
+        Stack();
+        ~Stack();
+
+        void Reset(std::size_t size);
+        void push(s32 val);
+        std::size_t Count() const;
+        s32 top() const;
+        s32 pop();
+
+    private:
+        std::vector<s32> stack{};
+        std::size_t stack_size{};
+        std::size_t stack_pos{};
+    };
+    NodeStates();
+    ~NodeStates();
+
+    void Initialize(std::size_t _node_count);
+    bool Tsort(EdgeMatrix& edge_matrix);
+    std::size_t GetIndexPos() const;
+    const std::vector<s32>& GetIndexList() const;
+
+private:
+    void PushTsortResult(s32 index);
+    bool DepthFirstSearch(EdgeMatrix& edge_matrix);
+    void ResetState();
+    void UpdateState(NodeStates::State state, std::size_t i);
+    NodeStates::State GetState(std::size_t i);
+
+    std::size_t node_count{};
+    std::vector<bool> was_node_found{};
+    std::vector<bool> was_node_completed{};
+    std::size_t index_pos{};
+    std::vector<s32> index_list{};
+    NodeStates::Stack index_stack{};
+};
+
+enum class SplitterMagic : u32_le {
+    SplitterHeader = Common::MakeMagic('S', 'N', 'D', 'H'),
+    DataHeader = Common::MakeMagic('S', 'N', 'D', 'D'),
+    InfoHeader = Common::MakeMagic('S', 'N', 'D', 'I'),
+};
+
+class SplitterInfo {
+public:
+    struct InHeader {
+        SplitterMagic magic{};
+        s32_le info_count{};
+        s32_le data_count{};
+        INSERT_PADDING_WORDS(5);
+    };
+    static_assert(sizeof(SplitterInfo::InHeader) == 0x20,
+                  "SplitterInfo::InHeader is an invalid size");
+
+    struct InInfoPrams {
+        SplitterMagic magic{};
+        s32_le send_id{};
+        s32_le sample_rate{};
+        s32_le length{};
+        s32_le resource_id_base{};
+    };
+    static_assert(sizeof(SplitterInfo::InInfoPrams) == 0x14,
+                  "SplitterInfo::InInfoPrams is an invalid size");
+
+    struct InDestinationParams {
+        SplitterMagic magic{};
+        s32_le splitter_id{};
+        std::array<float_le, AudioCommon::MAX_MIX_BUFFERS> mix_volumes{};
+        s32_le mix_id{};
+        bool in_use{};
+        INSERT_PADDING_BYTES(3);
+    };
+    static_assert(sizeof(SplitterInfo::InDestinationParams) == 0x70,
+                  "SplitterInfo::InDestinationParams is an invalid size");
+};
+
+class ServerSplitterDestinationData {
+public:
+    explicit ServerSplitterDestinationData(s32 id);
+    ~ServerSplitterDestinationData();
+
+    void Update(SplitterInfo::InDestinationParams& header);
+
+    ServerSplitterDestinationData* GetNextDestination();
+    const ServerSplitterDestinationData* GetNextDestination() const;
+    void SetNextDestination(ServerSplitterDestinationData* dest);
+    bool ValidMixId() const;
+    s32 GetMixId() const;
+    bool IsConfigured() const;
+    float GetMixVolume(std::size_t i) const;
+    const std::array<float, AudioCommon::MAX_MIX_BUFFERS>& CurrentMixVolumes() const;
+    const std::array<float, AudioCommon::MAX_MIX_BUFFERS>& LastMixVolumes() const;
+    void MarkDirty();
+    void UpdateInternalState();
+
+private:
+    bool needs_update{};
+    bool in_use{};
+    s32 id{};
+    s32 mix_id{};
+    std::array<float, AudioCommon::MAX_MIX_BUFFERS> current_mix_volumes{};
+    std::array<float, AudioCommon::MAX_MIX_BUFFERS> last_mix_volumes{};
+    ServerSplitterDestinationData* next = nullptr;
+};
+
+class ServerSplitterInfo {
+public:
+    explicit ServerSplitterInfo(s32 id);
+    ~ServerSplitterInfo();
+
+    void InitializeInfos();
+    void ClearNewConnectionFlag();
+    std::size_t Update(SplitterInfo::InInfoPrams& header);
+
+    ServerSplitterDestinationData* GetHead();
+    const ServerSplitterDestinationData* GetHead() const;
+    ServerSplitterDestinationData* GetData(std::size_t depth);
+    const ServerSplitterDestinationData* GetData(std::size_t depth) const;
+
+    bool HasNewConnection() const;
+    s32 GetLength() const;
+
+    void SetHead(ServerSplitterDestinationData* new_head);
+    void SetHeadDepth(s32 length);
+
+private:
+    s32 sample_rate{};
+    s32 id{};
+    s32 send_length{};
+    ServerSplitterDestinationData* head = nullptr;
+    bool new_connection{};
+};
+
+class SplitterContext {
+public:
+    SplitterContext();
+    ~SplitterContext();
+
+    void Initialize(BehaviorInfo& behavior_info, std::size_t splitter_count,
+                    std::size_t data_count);
+
+    bool Update(const std::vector<u8>& input, std::size_t& input_offset, std::size_t& bytes_read);
+    bool UsingSplitter() const;
+
+    ServerSplitterInfo& GetInfo(std::size_t i);
+    const ServerSplitterInfo& GetInfo(std::size_t i) const;
+    ServerSplitterDestinationData& GetData(std::size_t i);
+    const ServerSplitterDestinationData& GetData(std::size_t i) const;
+    ServerSplitterDestinationData* GetDestinationData(std::size_t info, std::size_t data);
+    const ServerSplitterDestinationData* GetDestinationData(std::size_t info,
+                                                            std::size_t data) const;
+    void UpdateInternalState();
+
+    std::size_t GetInfoCount() const;
+    std::size_t GetDataCount() const;
+
+private:
+    void Setup(std::size_t info_count, std::size_t data_count, bool is_splitter_bug_fixed);
+    bool UpdateInfo(const std::vector<u8>& input, std::size_t& input_offset,
+                    std::size_t& bytes_read, s32 in_splitter_count);
+    bool UpdateData(const std::vector<u8>& input, std::size_t& input_offset,
+                    std::size_t& bytes_read, s32 in_data_count);
+    bool RecomposeDestination(ServerSplitterInfo& info, SplitterInfo::InInfoPrams& header,
+                              const std::vector<u8>& input, const std::size_t& input_offset);
+
+    std::vector<ServerSplitterInfo> infos{};
+    std::vector<ServerSplitterDestinationData> datas{};
+
+    std::size_t info_count{};
+    std::size_t data_count{};
+    bool bug_fixed{};
+};
+} // namespace AudioCore

src/audio_core/stream.cpp

@@ -12,7 +12,6 @@
 #include "common/assert.h"
 #include "common/logging/log.h"
 #include "core/core_timing.h"
-#include "core/core_timing_util.h"
 #include "core/settings.h"
 
 namespace AudioCore {
@@ -104,11 +103,7 @@ void Stream::PlayNextBuffer(std::chrono::nanoseconds ns_late) {
 
     sink_stream.EnqueueSamples(GetNumChannels(), active_buffer->GetSamples());
 
-    const auto time_stretch_delta = Settings::values.enable_audio_stretching.GetValue()
-                                        ? std::chrono::nanoseconds::zero()
-                                        : ns_late;
-    const auto future_time = GetBufferReleaseNS(*active_buffer) - time_stretch_delta;
-    core_timing.ScheduleEvent(future_time, release_event, {});
+    core_timing.ScheduleEvent(GetBufferReleaseNS(*active_buffer) - ns_late, release_event, {});
 }
 
 void Stream::ReleaseActiveBuffer(std::chrono::nanoseconds ns_late) {

src/audio_core/voice_context.cpp

@@ -0,0 +1,529 @@
+// Copyright 2020 yuzu Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#include "audio_core/behavior_info.h"
+#include "audio_core/voice_context.h"
+#include "core/memory.h"
+
+namespace AudioCore {
+
+ServerVoiceChannelResource::ServerVoiceChannelResource(s32 id) : id(id) {}
+ServerVoiceChannelResource::~ServerVoiceChannelResource() = default;
+
+bool ServerVoiceChannelResource::InUse() const {
+    return in_use;
+}
+
+float ServerVoiceChannelResource::GetCurrentMixVolumeAt(std::size_t i) const {
+    ASSERT(i < AudioCommon::MAX_MIX_BUFFERS);
+    return mix_volume.at(i);
+}
+
+float ServerVoiceChannelResource::GetLastMixVolumeAt(std::size_t i) const {
+    ASSERT(i < AudioCommon::MAX_MIX_BUFFERS);
+    return last_mix_volume.at(i);
+}
+
+void ServerVoiceChannelResource::Update(VoiceChannelResource::InParams& in_params) {
+    in_use = in_params.in_use;
+    // Update our mix volumes only if it's in use
+    if (in_params.in_use) {
+        mix_volume = in_params.mix_volume;
+    }
+}
+
+void ServerVoiceChannelResource::UpdateLastMixVolumes() {
+    last_mix_volume = mix_volume;
+}
+
+const std::array<float, AudioCommon::MAX_MIX_BUFFERS>&
+ServerVoiceChannelResource::GetCurrentMixVolume() const {
+    return mix_volume;
+}
+
+const std::array<float, AudioCommon::MAX_MIX_BUFFERS>&
+ServerVoiceChannelResource::GetLastMixVolume() const {
+    return last_mix_volume;
+}
+
+ServerVoiceInfo::ServerVoiceInfo() {
+    Initialize();
+}
+ServerVoiceInfo::~ServerVoiceInfo() = default;
+
+void ServerVoiceInfo::Initialize() {
+    in_params.in_use = false;
+    in_params.node_id = 0;
+    in_params.id = 0;
+    in_params.current_playstate = ServerPlayState::Stop;
+    in_params.priority = 255;
+    in_params.sample_rate = 0;
+    in_params.sample_format = SampleFormat::Invalid;
+    in_params.channel_count = 0;
+    in_params.pitch = 0.0f;
+    in_params.volume = 0.0f;
+    in_params.last_volume = 0.0f;
+    in_params.biquad_filter.fill({});
+    in_params.wave_buffer_count = 0;
+    in_params.wave_bufffer_head = 0;
+    in_params.mix_id = AudioCommon::NO_MIX;
+    in_params.splitter_info_id = AudioCommon::NO_SPLITTER;
+    in_params.additional_params_address = 0;
+    in_params.additional_params_size = 0;
+    in_params.is_new = false;
+    out_params.played_sample_count = 0;
+    out_params.wave_buffer_consumed = 0;
+    in_params.voice_drop_flag = false;
+    in_params.buffer_mapped = false;
+    in_params.wave_buffer_flush_request_count = 0;
+    in_params.was_biquad_filter_enabled.fill(false);
+
+    for (auto& wave_buffer : in_params.wave_buffer) {
+        wave_buffer.start_sample_offset = 0;
+        wave_buffer.end_sample_offset = 0;
+        wave_buffer.is_looping = false;
+        wave_buffer.end_of_stream = false;
+        wave_buffer.buffer_address = 0;
+        wave_buffer.buffer_size = 0;
+        wave_buffer.context_address = 0;
+        wave_buffer.context_size = 0;
+        wave_buffer.sent_to_dsp = true;
+    }
+
+    stored_samples.clear();
+}
+
+void ServerVoiceInfo::UpdateParameters(const VoiceInfo::InParams& voice_in,
+                                       BehaviorInfo& behavior_info) {
+    in_params.in_use = voice_in.is_in_use;
+    in_params.id = voice_in.id;
+    in_params.node_id = voice_in.node_id;
+    in_params.last_playstate = in_params.current_playstate;
+    switch (voice_in.play_state) {
+    case PlayState::Paused:
+        in_params.current_playstate = ServerPlayState::Paused;
+        break;
+    case PlayState::Stopped:
+        if (in_params.current_playstate != ServerPlayState::Stop) {
+            in_params.current_playstate = ServerPlayState::RequestStop;
+        }
+        break;
+    case PlayState::Started:
+        in_params.current_playstate = ServerPlayState::Play;
+        break;
+    default:
+        UNREACHABLE_MSG("Unknown playstate {}", voice_in.play_state);
+        break;
+    }
+
+    in_params.priority = voice_in.priority;
+    in_params.sorting_order = voice_in.sorting_order;
+    in_params.sample_rate = voice_in.sample_rate;
+    in_params.sample_format = voice_in.sample_format;
+    in_params.channel_count = voice_in.channel_count;
+    in_params.pitch = voice_in.pitch;
+    in_params.volume = voice_in.volume;
+    in_params.biquad_filter = voice_in.biquad_filter;
+    in_params.wave_buffer_count = voice_in.wave_buffer_count;
+    in_params.wave_bufffer_head = voice_in.wave_buffer_head;
+    if (behavior_info.IsFlushVoiceWaveBuffersSupported()) {
+        const auto in_request_count = in_params.wave_buffer_flush_request_count;
+        const auto voice_request_count = voice_in.wave_buffer_flush_request_count;
+        in_params.wave_buffer_flush_request_count =
+            static_cast<u8>(in_request_count + voice_request_count);
+    }
+    in_params.mix_id = voice_in.mix_id;
+    if (behavior_info.IsSplitterSupported()) {
+        in_params.splitter_info_id = voice_in.splitter_info_id;
+    } else {
+        in_params.splitter_info_id = AudioCommon::NO_SPLITTER;
+    }
+
+    std::memcpy(in_params.voice_channel_resource_id.data(),
+                voice_in.voice_channel_resource_ids.data(),
+                sizeof(s32) * in_params.voice_channel_resource_id.size());
+
+    if (behavior_info.IsVoicePlayedSampleCountResetAtLoopPointSupported()) {
+        in_params.behavior_flags.is_played_samples_reset_at_loop_point =
+            voice_in.behavior_flags.is_played_samples_reset_at_loop_point;
+    } else {
+        in_params.behavior_flags.is_played_samples_reset_at_loop_point.Assign(0);
+    }
+    if (behavior_info.IsVoicePitchAndSrcSkippedSupported()) {
+        in_params.behavior_flags.is_pitch_and_src_skipped =
+            voice_in.behavior_flags.is_pitch_and_src_skipped;
+    } else {
+        in_params.behavior_flags.is_pitch_and_src_skipped.Assign(0);
+    }
+
+    if (voice_in.is_voice_drop_flag_clear_requested) {
+        in_params.voice_drop_flag = false;
+    }
+
+    if (in_params.additional_params_address != voice_in.additional_params_address ||
+        in_params.additional_params_size != voice_in.additional_params_size) {
+        in_params.additional_params_address = voice_in.additional_params_address;
+        in_params.additional_params_size = voice_in.additional_params_size;
+        // TODO(ogniK): Reattach buffer, do we actually need to? Maybe just signal to the DSP that
+        // our context is new
+    }
+}
+
+void ServerVoiceInfo::UpdateWaveBuffers(
+    const VoiceInfo::InParams& voice_in,
+    std::array<VoiceState*, AudioCommon::MAX_CHANNEL_COUNT>& voice_states,
+    BehaviorInfo& behavior_info) {
+    if (voice_in.is_new) {
+        // Initialize our wave buffers
+        for (auto& wave_buffer : in_params.wave_buffer) {
+            wave_buffer.start_sample_offset = 0;
+            wave_buffer.end_sample_offset = 0;
+            wave_buffer.is_looping = false;
+            wave_buffer.end_of_stream = false;
+            wave_buffer.buffer_address = 0;
+            wave_buffer.buffer_size = 0;
+            wave_buffer.context_address = 0;
+            wave_buffer.context_size = 0;
+            wave_buffer.sent_to_dsp = true;
+        }
+
+        // Mark all our wave buffers as invalid
+        for (std::size_t channel = 0; channel < static_cast<std::size_t>(in_params.channel_count);
+             channel++) {
+            for (auto& is_valid : voice_states[channel]->is_wave_buffer_valid) {
+                is_valid = false;
+            }
+        }
+    }
+
+    // Update our wave buffers
+    for (std::size_t i = 0; i < AudioCommon::MAX_WAVE_BUFFERS; i++) {
+        // Assume that we have at least 1 channel voice state
+        const auto have_valid_wave_buffer = voice_states[0]->is_wave_buffer_valid[i];
+
+        UpdateWaveBuffer(in_params.wave_buffer[i], voice_in.wave_buffer[i], in_params.sample_format,
+                         have_valid_wave_buffer, behavior_info);
+    }
+}
+
+void ServerVoiceInfo::UpdateWaveBuffer(ServerWaveBuffer& out_wavebuffer,
+                                       const WaveBuffer& in_wave_buffer, SampleFormat sample_format,
+                                       bool is_buffer_valid, BehaviorInfo& behavior_info) {
+    if (!is_buffer_valid && out_wavebuffer.sent_to_dsp) {
+        out_wavebuffer.buffer_address = 0;
+        out_wavebuffer.buffer_size = 0;
+    }
+
+    if (!in_wave_buffer.sent_to_server || !in_params.buffer_mapped) {
+        // Validate sample offset sizings
+        if (sample_format == SampleFormat::Pcm16) {
+            const auto buffer_size = in_wave_buffer.buffer_size;
+            if (in_wave_buffer.start_sample_offset < 0 || in_wave_buffer.end_sample_offset < 0 ||
+                (buffer_size < (sizeof(s16) * in_wave_buffer.start_sample_offset)) ||
+                (buffer_size < (sizeof(s16) * in_wave_buffer.end_sample_offset))) {
+                // TODO(ogniK): Write error info
+                return;
+            }
+        }
+        // TODO(ogniK): ADPCM Size error
+
+        out_wavebuffer.sent_to_dsp = false;
+        out_wavebuffer.start_sample_offset = in_wave_buffer.start_sample_offset;
+        out_wavebuffer.end_sample_offset = in_wave_buffer.end_sample_offset;
+        out_wavebuffer.is_looping = in_wave_buffer.is_looping;
+        out_wavebuffer.end_of_stream = in_wave_buffer.end_of_stream;
+
+        out_wavebuffer.buffer_address = in_wave_buffer.buffer_address;
+        out_wavebuffer.buffer_size = in_wave_buffer.buffer_size;
+        out_wavebuffer.context_address = in_wave_buffer.context_address;
+        out_wavebuffer.context_size = in_wave_buffer.context_size;
+        in_params.buffer_mapped =
+            in_wave_buffer.buffer_address != 0 && in_wave_buffer.buffer_size != 0;
+        // TODO(ogniK): Pool mapper attachment
+        // TODO(ogniK): IsAdpcmLoopContextBugFixed
+    }
+}
+
+void ServerVoiceInfo::WriteOutStatus(
+    VoiceInfo::OutParams& voice_out, VoiceInfo::InParams& voice_in,
+    std::array<VoiceState*, AudioCommon::MAX_CHANNEL_COUNT>& voice_states) {
+    if (voice_in.is_new) {
+        in_params.is_new = true;
+        voice_out.wave_buffer_consumed = 0;
+        voice_out.played_sample_count = 0;
+        voice_out.voice_dropped = false;
+    } else if (!in_params.is_new) {
+        voice_out.wave_buffer_consumed = voice_states[0]->wave_buffer_consumed;
+        voice_out.played_sample_count = voice_states[0]->played_sample_count;
+        voice_out.voice_dropped = in_params.voice_drop_flag;
+    } else {
+        voice_out.wave_buffer_consumed = 0;
+        voice_out.played_sample_count = 0;
+        voice_out.voice_dropped = false;
+    }
+}
+
+const ServerVoiceInfo::InParams& ServerVoiceInfo::GetInParams() const {
+    return in_params;
+}
+
+ServerVoiceInfo::InParams& ServerVoiceInfo::GetInParams() {
+    return in_params;
+}
+
+const ServerVoiceInfo::OutParams& ServerVoiceInfo::GetOutParams() const {
+    return out_params;
+}
+
+ServerVoiceInfo::OutParams& ServerVoiceInfo::GetOutParams() {
+    return out_params;
+}
+
+bool ServerVoiceInfo::ShouldSkip() const {
+    // TODO(ogniK): Handle unmapped wave buffers or parameters
+    return !in_params.in_use || (in_params.wave_buffer_count == 0) || in_params.voice_drop_flag;
+}
+
+bool ServerVoiceInfo::UpdateForCommandGeneration(VoiceContext& voice_context) {
+    std::array<VoiceState*, AudioCommon::MAX_CHANNEL_COUNT> dsp_voice_states{};
+    if (in_params.is_new) {
+        ResetResources(voice_context);
+        in_params.last_volume = in_params.volume;
+        in_params.is_new = false;
+    }
+
+    const s32 channel_count = in_params.channel_count;
+    for (s32 i = 0; i < channel_count; i++) {
+        const auto channel_resource = in_params.voice_channel_resource_id[i];
+        dsp_voice_states[i] =
+            &voice_context.GetDspSharedState(static_cast<std::size_t>(channel_resource));
+    }
+    return UpdateParametersForCommandGeneration(dsp_voice_states);
+}
+
+void ServerVoiceInfo::ResetResources(VoiceContext& voice_context) {
+    const s32 channel_count = in_params.channel_count;
+    for (s32 i = 0; i < channel_count; i++) {
+        const auto channel_resource = in_params.voice_channel_resource_id[i];
+        auto& dsp_state =
+            voice_context.GetDspSharedState(static_cast<std::size_t>(channel_resource));
+        dsp_state = {};
+        voice_context.GetChannelResource(static_cast<std::size_t>(channel_resource))
+            .UpdateLastMixVolumes();
+    }
+}
+
+bool ServerVoiceInfo::UpdateParametersForCommandGeneration(
+    std::array<VoiceState*, AudioCommon::MAX_CHANNEL_COUNT>& dsp_voice_states) {
+    const s32 channel_count = in_params.channel_count;
+    if (in_params.wave_buffer_flush_request_count > 0) {
+        FlushWaveBuffers(in_params.wave_buffer_flush_request_count, dsp_voice_states,
+                         channel_count);
+        in_params.wave_buffer_flush_request_count = 0;
+    }
+
+    switch (in_params.current_playstate) {
+    case ServerPlayState::Play: {
+        for (std::size_t i = 0; i < AudioCommon::MAX_WAVE_BUFFERS; i++) {
+            if (!in_params.wave_buffer[i].sent_to_dsp) {
+                for (s32 channel = 0; channel < channel_count; channel++) {
+                    dsp_voice_states[channel]->is_wave_buffer_valid[i] = true;
+                }
+                in_params.wave_buffer[i].sent_to_dsp = true;
+            }
+        }
+        in_params.should_depop = false;
+        return HasValidWaveBuffer(dsp_voice_states[0]);
+    }
+    case ServerPlayState::Paused:
+    case ServerPlayState::Stop: {
+        in_params.should_depop = in_params.last_playstate == ServerPlayState::Play;
+        return in_params.should_depop;
+    }
+    case ServerPlayState::RequestStop: {
+        for (std::size_t i = 0; i < AudioCommon::MAX_WAVE_BUFFERS; i++) {
+            in_params.wave_buffer[i].sent_to_dsp = true;
+            for (s32 channel = 0; channel < channel_count; channel++) {
+                auto* dsp_state = dsp_voice_states[channel];
+
+                if (dsp_state->is_wave_buffer_valid[i]) {
+                    dsp_state->wave_buffer_index =
+                        (dsp_state->wave_buffer_index + 1) % AudioCommon::MAX_WAVE_BUFFERS;
+                    dsp_state->wave_buffer_consumed++;
+                }
+
+                dsp_state->is_wave_buffer_valid[i] = false;
+            }
+        }
+
+        for (s32 channel = 0; channel < channel_count; channel++) {
+            auto* dsp_state = dsp_voice_states[channel];
+            dsp_state->offset = 0;
+            dsp_state->played_sample_count = 0;
+            dsp_state->fraction = 0;
+            dsp_state->sample_history.fill(0);
+            dsp_state->context = {};
+        }
+
+        in_params.current_playstate = ServerPlayState::Stop;
+        in_params.should_depop = in_params.last_playstate == ServerPlayState::Play;
+        return in_params.should_depop;
+    }
+    default:
+        UNREACHABLE_MSG("Invalid playstate {}", in_params.current_playstate);
+    }
+
+    return false;
+}
+
+void ServerVoiceInfo::FlushWaveBuffers(
+    u8 flush_count, std::array<VoiceState*, AudioCommon::MAX_CHANNEL_COUNT>& dsp_voice_states,
+    s32 channel_count) {
+    auto wave_head = in_params.wave_bufffer_head;
+
+    for (u8 i = 0; i < flush_count; i++) {
+        in_params.wave_buffer[wave_head].sent_to_dsp = true;
+        for (s32 channel = 0; channel < channel_count; channel++) {
+            auto* dsp_state = dsp_voice_states[channel];
+            dsp_state->wave_buffer_consumed++;
+            dsp_state->is_wave_buffer_valid[wave_head] = false;
+            dsp_state->wave_buffer_index =
+                (dsp_state->wave_buffer_index + 1) % AudioCommon::MAX_WAVE_BUFFERS;
+        }
+        wave_head = (wave_head + 1) % AudioCommon::MAX_WAVE_BUFFERS;
+    }
+}
+
+bool ServerVoiceInfo::HasValidWaveBuffer(const VoiceState* state) const {
+    const auto& valid_wb = state->is_wave_buffer_valid;
+    return std::find(valid_wb.begin(), valid_wb.end(), true) != valid_wb.end();
+}
+
+VoiceContext::VoiceContext(std::size_t voice_count) : voice_count(voice_count) {
+    for (std::size_t i = 0; i < voice_count; i++) {
+        voice_channel_resources.emplace_back(static_cast<s32>(i));
+        sorted_voice_info.push_back(&voice_info.emplace_back());
+        voice_states.emplace_back();
+        dsp_voice_states.emplace_back();
+    }
+}
+
+VoiceContext::~VoiceContext() {
+    sorted_voice_info.clear();
+}
+
+std::size_t VoiceContext::GetVoiceCount() const {
+    return voice_count;
+}
+
+ServerVoiceChannelResource& VoiceContext::GetChannelResource(std::size_t i) {
+    ASSERT(i < voice_count);
+    return voice_channel_resources.at(i);
+}
+
+const ServerVoiceChannelResource& VoiceContext::GetChannelResource(std::size_t i) const {
+    ASSERT(i < voice_count);
+    return voice_channel_resources.at(i);
+}
+
+VoiceState& VoiceContext::GetState(std::size_t i) {
+    ASSERT(i < voice_count);
+    return voice_states.at(i);
+}
+
+const VoiceState& VoiceContext::GetState(std::size_t i) const {
+    ASSERT(i < voice_count);
+    return voice_states.at(i);
+}
+
+VoiceState& VoiceContext::GetDspSharedState(std::size_t i) {
+    ASSERT(i < voice_count);
+    return dsp_voice_states.at(i);
+}
+
+const VoiceState& VoiceContext::GetDspSharedState(std::size_t i) const {
+    ASSERT(i < voice_count);
+    return dsp_voice_states.at(i);
+}
+
+ServerVoiceInfo& VoiceContext::GetInfo(std::size_t i) {
+    ASSERT(i < voice_count);
+    return voice_info.at(i);
+}
+
+const ServerVoiceInfo& VoiceContext::GetInfo(std::size_t i) const {
+    ASSERT(i < voice_count);
+    return voice_info.at(i);
+}
+
+ServerVoiceInfo& VoiceContext::GetSortedInfo(std::size_t i) {
+    ASSERT(i < voice_count);
+    return *sorted_voice_info.at(i);
+}
+
+const ServerVoiceInfo& VoiceContext::GetSortedInfo(std::size_t i) const {
+    ASSERT(i < voice_count);
+    return *sorted_voice_info.at(i);
+}
+
+s32 VoiceContext::DecodePcm16(s32* output_buffer, ServerWaveBuffer* wave_buffer, s32 channel,
+                              s32 channel_count, s32 buffer_offset, s32 sample_count,
+                              Core::Memory::Memory& memory) {
+    if (wave_buffer->buffer_address == 0) {
+        return 0;
+    }
+    if (wave_buffer->buffer_size == 0) {
+        return 0;
+    }
+    if (wave_buffer->end_sample_offset < wave_buffer->start_sample_offset) {
+        return 0;
+    }
+
+    const auto samples_remaining =
+        (wave_buffer->end_sample_offset - wave_buffer->start_sample_offset) - buffer_offset;
+    const auto start_offset = (wave_buffer->start_sample_offset + buffer_offset) * channel_count;
+    const auto buffer_pos = wave_buffer->buffer_address + start_offset;
+
+    s16* buffer_data = reinterpret_cast<s16*>(memory.GetPointer(buffer_pos));
+
+    const auto samples_processed = std::min(sample_count, samples_remaining);
+
+    // Fast path
+    if (channel_count == 1) {
+        for (std::ptrdiff_t i = 0; i < samples_processed; i++) {
+            output_buffer[i] = buffer_data[i];
+        }
+    } else {
+        for (std::ptrdiff_t i = 0; i < samples_processed; i++) {
+            output_buffer[i] = buffer_data[i * channel_count + channel];
+        }
+    }
+
+    return samples_processed;
+}
+
+void VoiceContext::SortInfo() {
+    for (std::size_t i = 0; i < voice_count; i++) {
+        sorted_voice_info[i] = &voice_info[i];
+    }
+
+    std::sort(sorted_voice_info.begin(), sorted_voice_info.end(),
+              [](const ServerVoiceInfo* lhs, const ServerVoiceInfo* rhs) {
+                  const auto& lhs_in = lhs->GetInParams();
+                  const auto& rhs_in = rhs->GetInParams();
+                  // Sort by priority
+                  if (lhs_in.priority != rhs_in.priority) {
+                      return lhs_in.priority > rhs_in.priority;
+                  } else {
+                      // If the priorities match, sort by sorting order
+                      return lhs_in.sorting_order > rhs_in.sorting_order;
+                  }
+              });
+}
+
+void VoiceContext::UpdateStateByDspShared() {
+    voice_states = dsp_voice_states;
+}
+
+} // namespace AudioCore

src/audio_core/voice_context.h

@@ -0,0 +1,296 @@
+// Copyright 2020 yuzu Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#pragma once
+
+#include <array>
+#include "audio_core/algorithm/interpolate.h"
+#include "audio_core/codec.h"
+#include "audio_core/common.h"
+#include "common/bit_field.h"
+#include "common/common_funcs.h"
+#include "common/common_types.h"
+
+namespace Core::Memory {
+class Memory;
+}
+
+namespace AudioCore {
+
+class BehaviorInfo;
+class VoiceContext;
+
+enum class SampleFormat : u8 {
+    Invalid = 0,
+    Pcm8 = 1,
+    Pcm16 = 2,
+    Pcm24 = 3,
+    Pcm32 = 4,
+    PcmFloat = 5,
+    Adpcm = 6,
+};
+
+enum class PlayState : u8 {
+    Started = 0,
+    Stopped = 1,
+    Paused = 2,
+};
+
+enum class ServerPlayState {
+    Play = 0,
+    Stop = 1,
+    RequestStop = 2,
+    Paused = 3,
+};
+
+struct BiquadFilterParameter {
+    bool enabled{};
+    INSERT_PADDING_BYTES(1);
+    std::array<s16, 3> numerator{};
+    std::array<s16, 2> denominator{};
+};
+static_assert(sizeof(BiquadFilterParameter) == 0xc, "BiquadFilterParameter is an invalid size");
+
+struct WaveBuffer {
+    u64_le buffer_address{};
+    u64_le buffer_size{};
+    s32_le start_sample_offset{};
+    s32_le end_sample_offset{};
+    u8 is_looping{};
+    u8 end_of_stream{};
+    u8 sent_to_server{};
+    INSERT_PADDING_BYTES(5);
+    u64 context_address{};
+    u64 context_size{};
+    INSERT_PADDING_BYTES(8);
+};
+static_assert(sizeof(WaveBuffer) == 0x38, "WaveBuffer is an invalid size");
+
+struct ServerWaveBuffer {
+    VAddr buffer_address{};
+    std::size_t buffer_size{};
+    s32 start_sample_offset{};
+    s32 end_sample_offset{};
+    bool is_looping{};
+    bool end_of_stream{};
+    VAddr context_address{};
+    std::size_t context_size{};
+    bool sent_to_dsp{true};
+};
+
+struct BehaviorFlags {
+    BitField<0, 1, u16> is_played_samples_reset_at_loop_point;
+    BitField<1, 1, u16> is_pitch_and_src_skipped;
+};
+static_assert(sizeof(BehaviorFlags) == 0x4, "BehaviorFlags is an invalid size");
+
+struct ADPCMContext {
+    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{};
+};
+
+class VoiceChannelResource {
+public:
+    struct InParams {
+        s32_le id{};
+        std::array<float_le, AudioCommon::MAX_MIX_BUFFERS> mix_volume{};
+        bool in_use{};
+        INSERT_PADDING_BYTES(11);
+    };
+    static_assert(sizeof(VoiceChannelResource::InParams) == 0x70, "InParams is an invalid size");
+};
+
+class ServerVoiceChannelResource {
+public:
+    explicit ServerVoiceChannelResource(s32 id);
+    ~ServerVoiceChannelResource();
+
+    bool InUse() const;
+    float GetCurrentMixVolumeAt(std::size_t i) const;
+    float GetLastMixVolumeAt(std::size_t i) const;
+    void Update(VoiceChannelResource::InParams& in_params);
+    void UpdateLastMixVolumes();
+
+    const std::array<float, AudioCommon::MAX_MIX_BUFFERS>& GetCurrentMixVolume() const;
+    const std::array<float, AudioCommon::MAX_MIX_BUFFERS>& GetLastMixVolume() const;
+
+private:
+    s32 id{};
+    std::array<float, AudioCommon::MAX_MIX_BUFFERS> mix_volume{};
+    std::array<float, AudioCommon::MAX_MIX_BUFFERS> last_mix_volume{};
+    bool in_use{};
+};
+
+class VoiceInfo {
+public:
+    struct InParams {
+        s32_le id{};
+        u32_le node_id{};
+        u8 is_new{};
+        u8 is_in_use{};
+        PlayState play_state{};
+        SampleFormat sample_format{};
+        s32_le sample_rate{};
+        s32_le priority{};
+        s32_le sorting_order{};
+        s32_le channel_count{};
+        float_le pitch{};
+        float_le volume{};
+        std::array<BiquadFilterParameter, 2> biquad_filter{};
+        s32_le wave_buffer_count{};
+        s16_le wave_buffer_head{};
+        INSERT_PADDING_BYTES(6);
+        u64_le additional_params_address{};
+        u64_le additional_params_size{};
+        s32_le mix_id{};
+        s32_le splitter_info_id{};
+        std::array<WaveBuffer, 4> wave_buffer{};
+        std::array<u32_le, 6> voice_channel_resource_ids{};
+        // TODO(ogniK): Remaining flags
+        u8 is_voice_drop_flag_clear_requested{};
+        u8 wave_buffer_flush_request_count{};
+        INSERT_PADDING_BYTES(2);
+        BehaviorFlags behavior_flags{};
+        INSERT_PADDING_BYTES(16);
+    };
+    static_assert(sizeof(VoiceInfo::InParams) == 0x170, "InParams is an invalid size");
+
+    struct OutParams {
+        u64_le played_sample_count{};
+        u32_le wave_buffer_consumed{};
+        u8 voice_dropped{};
+        INSERT_PADDING_BYTES(3);
+    };
+    static_assert(sizeof(VoiceInfo::OutParams) == 0x10, "OutParams is an invalid size");
+};
+
+class ServerVoiceInfo {
+public:
+    struct InParams {
+        bool in_use{};
+        bool is_new{};
+        bool should_depop{};
+        SampleFormat sample_format{};
+        s32 sample_rate{};
+        s32 channel_count{};
+        s32 id{};
+        s32 node_id{};
+        s32 mix_id{};
+        ServerPlayState current_playstate{};
+        ServerPlayState last_playstate{};
+        s32 priority{};
+        s32 sorting_order{};
+        float pitch{};
+        float volume{};
+        float last_volume{};
+        std::array<BiquadFilterParameter, AudioCommon::MAX_BIQUAD_FILTERS> biquad_filter{};
+        s32 wave_buffer_count{};
+        s16 wave_bufffer_head{};
+        INSERT_PADDING_BYTES(2);
+        BehaviorFlags behavior_flags{};
+        VAddr additional_params_address{};
+        std::size_t additional_params_size{};
+        std::array<ServerWaveBuffer, AudioCommon::MAX_WAVE_BUFFERS> wave_buffer{};
+        std::array<s32, AudioCommon::MAX_CHANNEL_COUNT> voice_channel_resource_id{};
+        s32 splitter_info_id{};
+        u8 wave_buffer_flush_request_count{};
+        bool voice_drop_flag{};
+        bool buffer_mapped{};
+        std::array<bool, AudioCommon::MAX_BIQUAD_FILTERS> was_biquad_filter_enabled{};
+    };
+
+    struct OutParams {
+        s64 played_sample_count{};
+        s32 wave_buffer_consumed{};
+    };
+
+    ServerVoiceInfo();
+    ~ServerVoiceInfo();
+    void Initialize();
+    void UpdateParameters(const VoiceInfo::InParams& voice_in, BehaviorInfo& behavior_info);
+    void UpdateWaveBuffers(const VoiceInfo::InParams& voice_in,
+                           std::array<VoiceState*, AudioCommon::MAX_CHANNEL_COUNT>& voice_states,
+                           BehaviorInfo& behavior_info);
+    void UpdateWaveBuffer(ServerWaveBuffer& out_wavebuffer, const WaveBuffer& in_wave_buffer,
+                          SampleFormat sample_format, bool is_buffer_valid,
+                          BehaviorInfo& behavior_info);
+    void WriteOutStatus(VoiceInfo::OutParams& voice_out, VoiceInfo::InParams& voice_in,
+                        std::array<VoiceState*, AudioCommon::MAX_CHANNEL_COUNT>& voice_states);
+
+    const InParams& GetInParams() const;
+    InParams& GetInParams();
+
+    const OutParams& GetOutParams() const;
+    OutParams& GetOutParams();
+
+    bool ShouldSkip() const;
+    bool UpdateForCommandGeneration(VoiceContext& voice_context);
+    void ResetResources(VoiceContext& voice_context);
+    bool UpdateParametersForCommandGeneration(
+        std::array<VoiceState*, AudioCommon::MAX_CHANNEL_COUNT>& dsp_voice_states);
+    void FlushWaveBuffers(u8 flush_count,
+                          std::array<VoiceState*, AudioCommon::MAX_CHANNEL_COUNT>& dsp_voice_states,
+                          s32 channel_count);
+
+private:
+    std::vector<s16> stored_samples;
+    InParams in_params{};
+    OutParams out_params{};
+
+    bool HasValidWaveBuffer(const VoiceState* state) const;
+};
+
+class VoiceContext {
+public:
+    VoiceContext(std::size_t voice_count);
+    ~VoiceContext();
+
+    std::size_t GetVoiceCount() const;
+    ServerVoiceChannelResource& GetChannelResource(std::size_t i);
+    const ServerVoiceChannelResource& GetChannelResource(std::size_t i) const;
+    VoiceState& GetState(std::size_t i);
+    const VoiceState& GetState(std::size_t i) const;
+    VoiceState& GetDspSharedState(std::size_t i);
+    const VoiceState& GetDspSharedState(std::size_t i) const;
+    ServerVoiceInfo& GetInfo(std::size_t i);
+    const ServerVoiceInfo& GetInfo(std::size_t i) const;
+    ServerVoiceInfo& GetSortedInfo(std::size_t i);
+    const ServerVoiceInfo& GetSortedInfo(std::size_t i) const;
+
+    s32 DecodePcm16(s32* output_buffer, ServerWaveBuffer* wave_buffer, s32 channel,
+                    s32 channel_count, s32 buffer_offset, s32 sample_count,
+                    Core::Memory::Memory& memory);
+    void SortInfo();
+    void UpdateStateByDspShared();
+
+private:
+    std::size_t voice_count{};
+    std::vector<ServerVoiceChannelResource> voice_channel_resources{};
+    std::vector<VoiceState> voice_states{};
+    std::vector<VoiceState> dsp_voice_states{};
+    std::vector<ServerVoiceInfo> voice_info{};
+    std::vector<ServerVoiceInfo*> sorted_voice_info{};
+};
+
+} // namespace AudioCore

src/common/CMakeLists.txt

@@ -150,6 +150,8 @@ add_library(common STATIC
     scope_exit.h
     spin_lock.cpp
     spin_lock.h
+    stream.cpp
+    stream.h
     string_util.cpp
     string_util.h
     swap.h
@@ -192,4 +194,9 @@ create_target_directory_groups(common)
 find_package(Boost 1.71 COMPONENTS context headers REQUIRED)
 
 target_link_libraries(common PUBLIC ${Boost_LIBRARIES} fmt::fmt microprofile)
-target_link_libraries(common PRIVATE lz4::lz4 zstd::zstd xbyak)
+target_link_libraries(common PRIVATE lz4::lz4 xbyak)
+if (MSVC)
+  target_link_libraries(common PRIVATE zstd::zstd)
+else()
+  target_link_libraries(common PRIVATE zstd)
+endif()

src/common/fiber.cpp

@@ -91,7 +91,7 @@ void Fiber::Rewind() {
     SwitchToFiber(impl->rewind_handle);
 }
 
-void Fiber::YieldTo(std::shared_ptr<Fiber>& from, std::shared_ptr<Fiber>& to) {
+void Fiber::YieldTo(std::shared_ptr<Fiber> from, std::shared_ptr<Fiber> to) {
     ASSERT_MSG(from != nullptr, "Yielding fiber is null!");
     ASSERT_MSG(to != nullptr, "Next fiber is null!");
     to->guard.lock();
@@ -199,7 +199,7 @@ void Fiber::Rewind() {
     boost::context::detail::jump_fcontext(impl->rewind_context, this);
 }
 
-void Fiber::YieldTo(std::shared_ptr<Fiber>& from, std::shared_ptr<Fiber>& to) {
+void Fiber::YieldTo(std::shared_ptr<Fiber> from, std::shared_ptr<Fiber> to) {
     ASSERT_MSG(from != nullptr, "Yielding fiber is null!");
     ASSERT_MSG(to != nullptr, "Next fiber is null!");
     to->guard.lock();

src/common/fiber.h

@@ -46,7 +46,7 @@ public:
 
     /// Yields control from Fiber 'from' to Fiber 'to'
     /// Fiber 'from' must be the currently running fiber.
-    static void YieldTo(std::shared_ptr<Fiber>& from, std::shared_ptr<Fiber>& to);
+    static void YieldTo(std::shared_ptr<Fiber> from, std::shared_ptr<Fiber> to);
     [[nodiscard]] static std::shared_ptr<Fiber> ThreadToFiber();
 
     void SetRewindPoint(std::function<void(void*)>&& rewind_func, void* start_parameter);

src/common/hex_util.h

@@ -16,14 +16,14 @@ namespace Common {
 
 [[nodiscard]] constexpr u8 ToHexNibble(char c) {
     if (c >= 65 && c <= 70) {
-        return c - 55;
+        return static_cast<u8>(c - 55);
     }
 
     if (c >= 97 && c <= 102) {
-        return c - 87;
+        return static_cast<u8>(c - 87);
     }
 
-    return c - 48;
+    return static_cast<u8>(c - 48);
 }
 
 [[nodiscard]] std::vector<u8> HexStringToVector(std::string_view str, bool little_endian);
@@ -33,11 +33,11 @@ template <std::size_t Size, bool le = false>
     std::array<u8, Size> out{};
     if constexpr (le) {
         for (std::size_t i = 2 * Size - 2; i <= 2 * Size; i -= 2) {
-            out[i / 2] = (ToHexNibble(str[i]) << 4) | ToHexNibble(str[i + 1]);
+            out[i / 2] = static_cast<u8>((ToHexNibble(str[i]) << 4) | ToHexNibble(str[i + 1]));
         }
     } else {
         for (std::size_t i = 0; i < 2 * Size; i += 2) {
-            out[i / 2] = (ToHexNibble(str[i]) << 4) | ToHexNibble(str[i + 1]);
+            out[i / 2] = static_cast<u8>((ToHexNibble(str[i]) << 4) | ToHexNibble(str[i + 1]));
         }
     }
     return out;

src/common/math_util.h

@@ -20,14 +20,14 @@ struct Rectangle {
 
     constexpr Rectangle() = default;
 
-    constexpr Rectangle(T left, T top, T right, T bottom)
-        : left(left), top(top), right(right), bottom(bottom) {}
+    constexpr Rectangle(T left_, T top_, T right_, T bottom_)
+        : left(left_), top(top_), right(right_), bottom(bottom_) {}
 
     [[nodiscard]] T GetWidth() const {
         if constexpr (std::is_floating_point_v<T>) {
             return std::abs(right - left);
         } else {
-            return std::abs(static_cast<std::make_signed_t<T>>(right - left));
+            return static_cast<T>(std::abs(static_cast<std::make_signed_t<T>>(right - left)));
         }
     }
 
@@ -35,7 +35,7 @@ struct Rectangle {
         if constexpr (std::is_floating_point_v<T>) {
             return std::abs(bottom - top);
         } else {
-            return std::abs(static_cast<std::make_signed_t<T>>(bottom - top));
+            return static_cast<T>(std::abs(static_cast<std::make_signed_t<T>>(bottom - top)));
         }
     }
 

src/common/stream.cpp

@@ -0,0 +1,47 @@
+// Copyright 2020 yuzu Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#include <stdexcept>
+#include "common/common_types.h"
+#include "common/stream.h"
+
+namespace Common {
+
+Stream::Stream() = default;
+Stream::~Stream() = default;
+
+void Stream::Seek(s32 offset, SeekOrigin origin) {
+    if (origin == SeekOrigin::SetOrigin) {
+        if (offset < 0) {
+            position = 0;
+        } else if (position >= buffer.size()) {
+            position = buffer.size();
+        } else {
+            position = offset;
+        }
+    } else if (origin == SeekOrigin::FromCurrentPos) {
+        Seek(static_cast<s32>(position) + offset, SeekOrigin::SetOrigin);
+    } else if (origin == SeekOrigin::FromEnd) {
+        Seek(static_cast<s32>(buffer.size()) - offset, SeekOrigin::SetOrigin);
+    }
+}
+
+u8 Stream::ReadByte() {
+    if (position < buffer.size()) {
+        return buffer[position++];
+    } else {
+        throw std::out_of_range("Attempting to read a byte not within the buffer range");
+    }
+}
+
+void Stream::WriteByte(u8 byte) {
+    if (position == buffer.size()) {
+        buffer.push_back(byte);
+        position++;
+    } else {
+        buffer.insert(buffer.begin() + position, byte);
+    }
+}
+
+} // namespace Common

src/common/stream.h

@@ -0,0 +1,50 @@
+// Copyright 2020 yuzu Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#pragma once
+
+#include <vector>
+#include "common/common_types.h"
+
+namespace Common {
+
+enum class SeekOrigin {
+    SetOrigin,
+    FromCurrentPos,
+    FromEnd,
+};
+
+class Stream {
+public:
+    /// Stream creates a bitstream and provides common functionality on the stream.
+    explicit Stream();
+    ~Stream();
+
+    /// Reposition bitstream "cursor" to the specified offset from origin
+    void Seek(s32 offset, SeekOrigin origin);
+
+    /// Reads next byte in the stream buffer and increments position
+    u8 ReadByte();
+
+    /// Writes byte at current position
+    void WriteByte(u8 byte);
+
+    std::size_t GetPosition() const {
+        return position;
+    }
+
+    std::vector<u8>& GetBuffer() {
+        return buffer;
+    }
+
+    const std::vector<u8>& GetBuffer() const {
+        return buffer;
+    }
+
+private:
+    std::vector<u8> buffer;
+    std::size_t position{0};
+};
+
+} // namespace Common

src/common/vector_math.h

@@ -87,7 +87,13 @@ public:
 
     template <typename V>
     [[nodiscard]] constexpr Vec2<decltype(T{} * V{})> operator*(const V& f) const {
-        return {x * f, y * f};
+        using TV = decltype(T{} * V{});
+        using C = std::common_type_t<T, V>;
+
+        return {
+            static_cast<TV>(static_cast<C>(x) * static_cast<C>(f)),
+            static_cast<TV>(static_cast<C>(y) * static_cast<C>(f)),
+        };
     }
 
     template <typename V>
@@ -98,7 +104,13 @@ public:
 
     template <typename V>
     [[nodiscard]] constexpr Vec2<decltype(T{} / V{})> operator/(const V& f) const {
-        return {x / f, y / f};
+        using TV = decltype(T{} / V{});
+        using C = std::common_type_t<T, V>;
+
+        return {
+            static_cast<TV>(static_cast<C>(x) / static_cast<C>(f)),
+            static_cast<TV>(static_cast<C>(y) / static_cast<C>(f)),
+        };
     }
 
     template <typename V>
@@ -168,7 +180,10 @@ public:
 
 template <typename T, typename V>
 [[nodiscard]] constexpr Vec2<T> operator*(const V& f, const Vec2<T>& vec) {
-    return Vec2<T>(f * vec.x, f * vec.y);
+    using C = std::common_type_t<T, V>;
+
+    return Vec2<T>(static_cast<T>(static_cast<C>(f) * static_cast<C>(vec.x)),
+                   static_cast<T>(static_cast<C>(f) * static_cast<C>(vec.y)));
 }
 
 using Vec2f = Vec2<float>;
@@ -237,7 +252,14 @@ public:
 
     template <typename V>
     [[nodiscard]] constexpr Vec3<decltype(T{} * V{})> operator*(const V& f) const {
-        return {x * f, y * f, z * f};
+        using TV = decltype(T{} * V{});
+        using C = std::common_type_t<T, V>;
+
+        return {
+            static_cast<TV>(static_cast<C>(x) * static_cast<C>(f)),
+            static_cast<TV>(static_cast<C>(y) * static_cast<C>(f)),
+            static_cast<TV>(static_cast<C>(z) * static_cast<C>(f)),
+        };
     }
 
     template <typename V>
@@ -247,7 +269,14 @@ public:
     }
     template <typename V>
     [[nodiscard]] constexpr Vec3<decltype(T{} / V{})> operator/(const V& f) const {
-        return {x / f, y / f, z / f};
+        using TV = decltype(T{} / V{});
+        using C = std::common_type_t<T, V>;
+
+        return {
+            static_cast<TV>(static_cast<C>(x) / static_cast<C>(f)),
+            static_cast<TV>(static_cast<C>(y) / static_cast<C>(f)),
+            static_cast<TV>(static_cast<C>(z) / static_cast<C>(f)),
+        };
     }
 
     template <typename V>
@@ -367,7 +396,11 @@ public:
 
 template <typename T, typename V>
 [[nodiscard]] constexpr Vec3<T> operator*(const V& f, const Vec3<T>& vec) {
-    return Vec3<T>(f * vec.x, f * vec.y, f * vec.z);
+    using C = std::common_type_t<T, V>;
+
+    return Vec3<T>(static_cast<T>(static_cast<C>(f) * static_cast<C>(vec.x)),
+                   static_cast<T>(static_cast<C>(f) * static_cast<C>(vec.y)),
+                   static_cast<T>(static_cast<C>(f) * static_cast<C>(vec.z)));
 }
 
 template <>
@@ -446,7 +479,15 @@ public:
 
     template <typename V>
     [[nodiscard]] constexpr Vec4<decltype(T{} * V{})> operator*(const V& f) const {
-        return {x * f, y * f, z * f, w * f};
+        using TV = decltype(T{} * V{});
+        using C = std::common_type_t<T, V>;
+
+        return {
+            static_cast<TV>(static_cast<C>(x) * static_cast<C>(f)),
+            static_cast<TV>(static_cast<C>(y) * static_cast<C>(f)),
+            static_cast<TV>(static_cast<C>(z) * static_cast<C>(f)),
+            static_cast<TV>(static_cast<C>(w) * static_cast<C>(f)),
+        };
     }
 
     template <typename V>
@@ -457,7 +498,15 @@ public:
 
     template <typename V>
     [[nodiscard]] constexpr Vec4<decltype(T{} / V{})> operator/(const V& f) const {
-        return {x / f, y / f, z / f, w / f};
+        using TV = decltype(T{} / V{});
+        using C = std::common_type_t<T, V>;
+
+        return {
+            static_cast<TV>(static_cast<C>(x) / static_cast<C>(f)),
+            static_cast<TV>(static_cast<C>(y) / static_cast<C>(f)),
+            static_cast<TV>(static_cast<C>(z) / static_cast<C>(f)),
+            static_cast<TV>(static_cast<C>(w) / static_cast<C>(f)),
+        };
     }
 
     template <typename V>
@@ -582,7 +631,15 @@ public:
 
 template <typename T, typename V>
 [[nodiscard]] constexpr Vec4<decltype(V{} * T{})> operator*(const V& f, const Vec4<T>& vec) {
-    return {f * vec.x, f * vec.y, f * vec.z, f * vec.w};
+    using TV = decltype(V{} * T{});
+    using C = std::common_type_t<T, V>;
+
+    return {
+        static_cast<TV>(static_cast<C>(f) * static_cast<C>(vec.x)),
+        static_cast<TV>(static_cast<C>(f) * static_cast<C>(vec.y)),
+        static_cast<TV>(static_cast<C>(f) * static_cast<C>(vec.z)),
+        static_cast<TV>(static_cast<C>(f) * static_cast<C>(vec.w)),
+    };
 }
 
 using Vec4f = Vec4<float>;

src/common/wall_clock.cpp

@@ -15,7 +15,7 @@ namespace Common {
 using base_timer = std::chrono::steady_clock;
 using base_time_point = std::chrono::time_point<base_timer>;
 
-class StandardWallClock : public WallClock {
+class StandardWallClock final : public WallClock {
 public:
     StandardWallClock(u64 emulated_cpu_frequency, u64 emulated_clock_frequency)
         : WallClock(emulated_cpu_frequency, emulated_clock_frequency, false) {

src/common/wall_clock.h

@@ -13,6 +13,8 @@ namespace Common {
 
 class WallClock {
 public:
+    virtual ~WallClock() = default;
+
     /// Returns current wall time in nanoseconds
     [[nodiscard]] virtual std::chrono::nanoseconds GetTimeNS() = 0;
 

src/common/x64/native_clock.h

@@ -12,7 +12,7 @@
 namespace Common {
 
 namespace X64 {
-class NativeClock : public WallClock {
+class NativeClock final : public WallClock {
 public:
     NativeClock(u64 emulated_cpu_frequency, u64 emulated_clock_frequency, u64 rtsc_frequency);
 

src/core/CMakeLists.txt

@@ -1,9 +1,3 @@
-if (YUZU_ENABLE_BOXCAT)
-    set(BCAT_BOXCAT_ADDITIONAL_SOURCES hle/service/bcat/backend/boxcat.cpp hle/service/bcat/backend/boxcat.h)
-else()
-    set(BCAT_BOXCAT_ADDITIONAL_SOURCES)
-endif()
-
 add_library(core STATIC
     arm/arm_interface.h
     arm/arm_interface.cpp
@@ -126,6 +120,8 @@ add_library(core STATIC
     file_sys/vfs_vector.h
     file_sys/xts_archive.cpp
     file_sys/xts_archive.h
+    frontend/applets/controller.cpp
+    frontend/applets/controller.h
     frontend/applets/error.cpp
     frontend/applets/error.h
     frontend/applets/general_frontend.cpp
@@ -244,6 +240,8 @@ add_library(core STATIC
     hle/service/am/applet_oe.h
     hle/service/am/applets/applets.cpp
     hle/service/am/applets/applets.h
+    hle/service/am/applets/controller.cpp
+    hle/service/am/applets/controller.h
     hle/service/am/applets/error.cpp
     hle/service/am/applets/error.h
     hle/service/am/applets/general_backend.cpp
@@ -299,7 +297,6 @@ add_library(core STATIC
     hle/service/audio/hwopus.h
     hle/service/bcat/backend/backend.cpp
     hle/service/bcat/backend/backend.h
-    ${BCAT_BOXCAT_ADDITIONAL_SOURCES}
     hle/service/bcat/bcat.cpp
     hle/service/bcat/bcat.h
     hle/service/bcat/module.cpp
@@ -442,6 +439,8 @@ add_library(core STATIC
     hle/service/nvdrv/devices/nvhost_gpu.h
     hle/service/nvdrv/devices/nvhost_nvdec.cpp
     hle/service/nvdrv/devices/nvhost_nvdec.h
+    hle/service/nvdrv/devices/nvhost_nvdec_common.cpp
+    hle/service/nvdrv/devices/nvhost_nvdec_common.h
     hle/service/nvdrv/devices/nvhost_nvjpg.cpp
     hle/service/nvdrv/devices/nvhost_nvjpg.h
     hle/service/nvdrv/devices/nvhost_vic.cpp
@@ -604,6 +603,13 @@ add_library(core STATIC
     tools/freezer.h
 )
 
+if (YUZU_ENABLE_BOXCAT)
+    target_sources(core PRIVATE
+        hle/service/bcat/backend/boxcat.cpp
+        hle/service/bcat/backend/boxcat.h
+    )
+endif()
+
 if (MSVC)
     target_compile_options(core PRIVATE
         # 'expression' : signed/unsigned mismatch
@@ -619,6 +625,20 @@ if (MSVC)
         # 'context' : truncation from 'type1' to 'type2'
         /we4305
     )
+else()
+    target_compile_options(core PRIVATE
+        -Werror=conversion
+        -Werror=ignored-qualifiers
+        -Werror=implicit-fallthrough
+        -Werror=reorder
+        -Werror=sign-compare
+        -Werror=unused-variable
+
+        $<$<CXX_COMPILER_ID:GNU>:-Werror=unused-but-set-parameter>
+        $<$<CXX_COMPILER_ID:GNU>:-Werror=unused-but-set-variable>
+
+        -Wno-sign-conversion
+    )
 endif()
 
 create_target_directory_groups(core)