38 files changed, 881 insertions, 587 deletions

diff --git a/.gitmodules b/.gitmodules
index 598e4c64d..059512902 100644
--- a/.gitmodules
+++ b/.gitmodules
@@ -7,3 +7,6 @@
 [submodule "nihstro"]
         path = externals/nihstro
         url = https://github.com/neobrain/nihstro.git
+[submodule "soundtouch"]
+        path = externals/soundtouch
+        url = https://github.com/citra-emu/soundtouch.git
diff --git a/CMakeLists.txt b/CMakeLists.txt
index 019321ad8..d628ecc50 100644
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@@ -249,6 +249,9 @@ if(ENABLE_QT)
     include_directories(externals/qhexedit)
     add_subdirectory(externals/qhexedit)
 endif()
+
+add_subdirectory(externals/soundtouch)
+
 add_subdirectory(src)
 
 # Install freedesktop.org metadata files, following those specifications:
diff --git a/externals/soundtouch b/externals/soundtouch
new file mode 160000
+Subproject 5274ec4dec498bd88ccbcd28862a0f78a3b95ef
diff --git a/src/audio_core/CMakeLists.txt b/src/audio_core/CMakeLists.txt
index b1c86462f..a965af291 100644
--- a/src/audio_core/CMakeLists.txt
+++ b/src/audio_core/CMakeLists.txt
@@ -18,6 +18,9 @@ set(HEADERS
             sink.h
             )
 
+include_directories(../../externals/soundtouch/include)
+
 create_directory_groups(${SRCS} ${HEADERS})
 
-add_library(audio_core STATIC ${SRCS} ${HEADERS})
\ No newline at end of file
+add_library(audio_core STATIC ${SRCS} ${HEADERS})
+target_link_libraries(audio_core SoundTouch)
diff --git a/src/audio_core/audio_core.cpp b/src/audio_core/audio_core.cpp
index 894f46990..b512b0f9b 100644
--- a/src/audio_core/audio_core.cpp
+++ b/src/audio_core/audio_core.cpp
@@ -4,6 +4,7 @@
 
 #include "audio_core/audio_core.h"
 #include "audio_core/hle/dsp.h"
+#include "audio_core/hle/pipe.h"
 
 #include "core/core_timing.h"
 #include "core/hle/kernel/vm_manager.h"
@@ -17,10 +18,10 @@ static constexpr u64 audio_frame_ticks = 1310252ull; ///< Units: ARM11 cycles
 
 static void AudioTickCallback(u64 /*userdata*/, int cycles_late) {
     if (DSP::HLE::Tick()) {
-        // HACK: We're not signaling the interrups when they should be, but just firing them all off together.
-        // It should be only (interrupt_id = 2, channel_id = 2) that's signalled here.
-        // TODO(merry): Understand when the other interrupts are fired.
-        DSP_DSP::SignalAllInterrupts();
+        // TODO(merry): Signal all the other interrupts as appropriate.
+        DSP_DSP::SignalPipeInterrupt(DSP::HLE::DspPipe::Audio);
+        // HACK(merry): Added to prevent regressions. Will remove soon.
+        DSP_DSP::SignalPipeInterrupt(DSP::HLE::DspPipe::Binary);
     }
 
     // Reschedule recurrent event
diff --git a/src/audio_core/audio_core.h b/src/audio_core/audio_core.h
index 64c330914..b349895ea 100644
--- a/src/audio_core/audio_core.h
+++ b/src/audio_core/audio_core.h
@@ -10,8 +10,6 @@ class VMManager;
 
 namespace AudioCore {
 
-constexpr int num_sources = 24;
-constexpr int samples_per_frame = 160;     ///< Samples per audio frame at native sample rate
 constexpr int native_sample_rate = 32728;  ///< 32kHz
 
 /// Initialise Audio Core
diff --git a/src/audio_core/hle/common.h b/src/audio_core/hle/common.h
index 37d441eb2..7910f42ae 100644
--- a/src/audio_core/hle/common.h
+++ b/src/audio_core/hle/common.h
@@ -7,18 +7,19 @@
 #include <algorithm>
 #include <array>
 
-#include "audio_core/audio_core.h"
-
 #include "common/common_types.h"
 
 namespace DSP {
 namespace HLE {
 
+constexpr int num_sources = 24;
+constexpr int samples_per_frame = 160;     ///< Samples per audio frame at native sample rate
+
 /// The final output to the speakers is stereo. Preprocessing output in Source is also stereo.
-using StereoFrame16 = std::array<std::array<s16, 2>, AudioCore::samples_per_frame>;
+using StereoFrame16 = std::array<std::array<s16, 2>, samples_per_frame>;
 
 /// The DSP is quadraphonic internally.
-using QuadFrame32   = std::array<std::array<s32, 4>, AudioCore::samples_per_frame>;
+using QuadFrame32   = std::array<std::array<s32, 4>, samples_per_frame>;
 
 /**
  * This performs the filter operation defined by FilterT::ProcessSample on the frame in-place.
diff --git a/src/audio_core/hle/dsp.h b/src/audio_core/hle/dsp.h
index c15ef0b7a..c76350bdd 100644
--- a/src/audio_core/hle/dsp.h
+++ b/src/audio_core/hle/dsp.h
@@ -7,7 +7,7 @@
 #include <cstddef>
 #include <type_traits>
 
-#include "audio_core/audio_core.h"
+#include "audio_core/hle/common.h"
 
 #include "common/bit_field.h"
 #include "common/common_funcs.h"
@@ -305,7 +305,7 @@ struct SourceConfiguration {
         u16_le buffer_id;
     };
 
-    Configuration config[AudioCore::num_sources];
+    Configuration config[num_sources];
 };
 ASSERT_DSP_STRUCT(SourceConfiguration::Configuration, 192);
 ASSERT_DSP_STRUCT(SourceConfiguration::Configuration::Buffer, 20);
@@ -320,7 +320,7 @@ struct SourceStatus {
         INSERT_PADDING_DSPWORDS(1);
     };
 
-    Status status[AudioCore::num_sources];
+    Status status[num_sources];
 };
 ASSERT_DSP_STRUCT(SourceStatus::Status, 12);
 
@@ -413,7 +413,7 @@ ASSERT_DSP_STRUCT(DspConfiguration::ReverbEffect, 52);
 struct AdpcmCoefficients {
     /// Coefficients are signed fixed point with 11 fractional bits.
     /// Each source has 16 coefficients associated with it.
-    s16_le coeff[AudioCore::num_sources][16];
+    s16_le coeff[num_sources][16];
 };
 ASSERT_DSP_STRUCT(AdpcmCoefficients, 768);
 
@@ -427,7 +427,7 @@ ASSERT_DSP_STRUCT(DspStatus, 32);
 /// Final mixed output in PCM16 stereo format, what you hear out of the speakers.
 /// When the application writes to this region it has no effect.
 struct FinalMixSamples {
-    s16_le pcm16[2 * AudioCore::samples_per_frame];
+    s16_le pcm16[2 * samples_per_frame];
 };
 ASSERT_DSP_STRUCT(FinalMixSamples, 640);
 
@@ -437,7 +437,7 @@ ASSERT_DSP_STRUCT(FinalMixSamples, 640);
 /// Values that exceed s16 range will be clipped by the DSP after further processing.
 struct IntermediateMixSamples {
     struct Samples {
-        s32_le pcm32[4][AudioCore::samples_per_frame]; ///< Little-endian as opposed to DSP middle-endian.
+        s32_le pcm32[4][samples_per_frame]; ///< Little-endian as opposed to DSP middle-endian.
     };
 
     Samples mix1;
diff --git a/src/audio_core/hle/pipe.cpp b/src/audio_core/hle/pipe.cpp
index 9381883b4..03280780f 100644
--- a/src/audio_core/hle/pipe.cpp
+++ b/src/audio_core/hle/pipe.cpp
@@ -12,12 +12,14 @@
 #include "common/common_types.h"
 #include "common/logging/log.h"
 
+#include "core/hle/service/dsp_dsp.h"
+
 namespace DSP {
 namespace HLE {
 
 static DspState dsp_state = DspState::Off;
 
-static std::array<std::vector<u8>, static_cast<size_t>(DspPipe::DspPipe_MAX)> pipe_data;
+static std::array<std::vector<u8>, NUM_DSP_PIPE> pipe_data;
 
 void ResetPipes() {
     for (auto& data : pipe_data) {
@@ -27,16 +29,18 @@ void ResetPipes() {
 }
 
 std::vector<u8> PipeRead(DspPipe pipe_number, u32 length) {
-    if (pipe_number >= DspPipe::DspPipe_MAX) {
-        LOG_ERROR(Audio_DSP, "pipe_number = %u invalid", pipe_number);
+    const size_t pipe_index = static_cast<size_t>(pipe_number);
+
+    if (pipe_index >= NUM_DSP_PIPE) {
+        LOG_ERROR(Audio_DSP, "pipe_number = %zu invalid", pipe_index);
         return {};
     }
 
-    std::vector<u8>& data = pipe_data[static_cast<size_t>(pipe_number)];
+    std::vector<u8>& data = pipe_data[pipe_index];
 
     if (length > data.size()) {
-        LOG_WARNING(Audio_DSP, "pipe_number = %u is out of data, application requested read of %u but %zu remain",
-                    pipe_number, length, data.size());
+        LOG_WARNING(Audio_DSP, "pipe_number = %zu is out of data, application requested read of %u but %zu remain",
+                    pipe_index, length, data.size());
         length = data.size();
     }
 
@@ -49,16 +53,20 @@ std::vector<u8> PipeRead(DspPipe pipe_number, u32 length) {
 }
 
 size_t GetPipeReadableSize(DspPipe pipe_number) {
-    if (pipe_number >= DspPipe::DspPipe_MAX) {
-        LOG_ERROR(Audio_DSP, "pipe_number = %u invalid", pipe_number);
+    const size_t pipe_index = static_cast<size_t>(pipe_number);
+
+    if (pipe_index >= NUM_DSP_PIPE) {
+        LOG_ERROR(Audio_DSP, "pipe_number = %zu invalid", pipe_index);
         return 0;
     }
 
-    return pipe_data[static_cast<size_t>(pipe_number)].size();
+    return pipe_data[pipe_index].size();
 }
 
 static void WriteU16(DspPipe pipe_number, u16 value) {
-    std::vector<u8>& data = pipe_data[static_cast<size_t>(pipe_number)];
+    const size_t pipe_index = static_cast<size_t>(pipe_number);
+
+    std::vector<u8>& data = pipe_data.at(pipe_index);
     // Little endian
     data.emplace_back(value & 0xFF);
     data.emplace_back(value >> 8);
@@ -91,6 +99,8 @@ static void AudioPipeWriteStructAddresses() {
     for (u16 addr : struct_addresses) {
         WriteU16(DspPipe::Audio, addr);
     }
+    // Signal that we have data on this pipe.
+    DSP_DSP::SignalPipeInterrupt(DspPipe::Audio);
 }
 
 void PipeWrite(DspPipe pipe_number, const std::vector<u8>& buffer) {
@@ -145,7 +155,7 @@ void PipeWrite(DspPipe pipe_number, const std::vector<u8>& buffer) {
         return;
     }
     default:
-        LOG_CRITICAL(Audio_DSP, "pipe_number = %u unimplemented", pipe_number);
+        LOG_CRITICAL(Audio_DSP, "pipe_number = %zu unimplemented", static_cast<size_t>(pipe_number));
         UNIMPLEMENTED();
         return;
     }
diff --git a/src/audio_core/hle/pipe.h b/src/audio_core/hle/pipe.h
index 382d35e87..64d97f8ba 100644
--- a/src/audio_core/hle/pipe.h
+++ b/src/audio_core/hle/pipe.h
@@ -19,9 +19,9 @@ enum class DspPipe {
     Debug = 0,
     Dma = 1,
     Audio = 2,
-    Binary = 3,
-    DspPipe_MAX
+    Binary = 3
 };
+constexpr size_t NUM_DSP_PIPE = 8;
 
 /**
  * Read a DSP pipe.
diff --git a/src/citra_qt/CMakeLists.txt b/src/citra_qt/CMakeLists.txt
index 6660d9879..cc9e0c624 100644
--- a/src/citra_qt/CMakeLists.txt
+++ b/src/citra_qt/CMakeLists.txt
@@ -92,7 +92,7 @@ else()
 endif()
 target_link_libraries(citra-qt core video_core audio_core common qhexedit)
 target_link_libraries(citra-qt ${OPENGL_gl_LIBRARY} ${CITRA_QT_LIBS})
-target_link_libraries(citra-qt ${PLATFORM_LIBRARIES})
+target_link_libraries(citra-qt ${PLATFORM_LIBRARIES} Threads::Threads)
 
 if(${CMAKE_SYSTEM_NAME} MATCHES "Linux|FreeBSD|OpenBSD|NetBSD")
     install(TARGETS citra-qt RUNTIME DESTINATION "${CMAKE_INSTALL_PREFIX}/bin")
diff --git a/src/citra_qt/bootmanager.cpp b/src/citra_qt/bootmanager.cpp
index 8e60b9cad..01b81c11c 100644
--- a/src/citra_qt/bootmanager.cpp
+++ b/src/citra_qt/bootmanager.cpp
@@ -71,7 +71,9 @@ void EmuThread::run() {
     // Shutdown the core emulation
     System::Shutdown();
 
+#if MICROPROFILE_ENABLED
     MicroProfileOnThreadExit();
+#endif
 
     render_window->moveContext();
 }
diff --git a/src/citra_qt/debugger/profiler.cpp b/src/citra_qt/debugger/profiler.cpp
index 4f6ba0e1f..7bb010f77 100644
--- a/src/citra_qt/debugger/profiler.cpp
+++ b/src/citra_qt/debugger/profiler.cpp
@@ -9,13 +9,16 @@
 #include "citra_qt/debugger/profiler.h"
 #include "citra_qt/util/util.h"
 
+#include "common/common_types.h"
 #include "common/microprofile.h"
 #include "common/profiler_reporting.h"
 
 // Include the implementation of the UI in this file. This isn't in microprofile.cpp because the
 // non-Qt frontends don't need it (and don't implement the UI drawing hooks either).
+#if MICROPROFILE_ENABLED
 #define MICROPROFILEUI_IMPL 1
 #include "common/microprofileui.h"
+#endif
 
 using namespace Common::Profiling;
 
@@ -34,21 +37,9 @@ static QVariant GetDataForColumn(int col, const AggregatedDuration& duration)
     }
 }
 
-static const TimingCategoryInfo* GetCategoryInfo(int id)
-{
-    const auto& categories = GetProfilingManager().GetTimingCategoriesInfo();
-    if ((size_t)id >= categories.size()) {
-        return nullptr;
-    } else {
-        return &categories[id];
-    }
-}
-
 ProfilerModel::ProfilerModel(QObject* parent) : QAbstractItemModel(parent)
 {
     updateProfilingInfo();
-    const auto& categories = GetProfilingManager().GetTimingCategoriesInfo();
-    results.time_per_category.resize(categories.size());
 }
 
 QVariant ProfilerModel::headerData(int section, Qt::Orientation orientation, int role) const
@@ -85,7 +76,7 @@ int ProfilerModel::rowCount(const QModelIndex& parent) const
     if (parent.isValid()) {
         return 0;
     } else {
-        return static_cast<int>(results.time_per_category.size() + 2);
+        return 2;
     }
 }
 
@@ -104,17 +95,6 @@ QVariant ProfilerModel::data(const QModelIndex& index, int role) const
             } else {
                 return GetDataForColumn(index.column(), results.interframe_time);
             }
-        } else {
-            if (index.column() == 0) {
-                const TimingCategoryInfo* info = GetCategoryInfo(index.row() - 2);
-                return info != nullptr ? QString(info->name) : QVariant();
-            } else {
-                if (index.row() - 2 < (int)results.time_per_category.size()) {
-                    return GetDataForColumn(index.column(), results.time_per_category[index.row() - 2]);
-                } else {
-                    return QVariant();
-                }
-            }
         }
     }
 
@@ -148,6 +128,8 @@ void ProfilerWidget::setProfilingInfoUpdateEnabled(bool enable)
     }
 }
 
+#if MICROPROFILE_ENABLED
+
 class MicroProfileWidget : public QWidget {
 public:
     MicroProfileWidget(QWidget* parent = nullptr);
@@ -171,6 +153,8 @@ private:
     QTimer update_timer;
 };
 
+#endif
+
 MicroProfileDialog::MicroProfileDialog(QWidget* parent)
     : QWidget(parent, Qt::Dialog)
 {
@@ -180,6 +164,8 @@ MicroProfileDialog::MicroProfileDialog(QWidget* parent)
     // Remove the "?" button from the titlebar and enable the maximize button
     setWindowFlags(windowFlags() & ~Qt::WindowContextHelpButtonHint | Qt::WindowMaximizeButtonHint);
 
+#if MICROPROFILE_ENABLED
+
     MicroProfileWidget* widget = new MicroProfileWidget(this);
 
     QLayout* layout = new QVBoxLayout(this);
@@ -191,6 +177,7 @@ MicroProfileDialog::MicroProfileDialog(QWidget* parent)
     setFocusProxy(widget);
     widget->setFocusPolicy(Qt::StrongFocus);
     widget->setFocus();
+#endif
 }
 
 QAction* MicroProfileDialog::toggleViewAction() {
@@ -218,6 +205,9 @@ void MicroProfileDialog::hideEvent(QHideEvent* ev) {
     QWidget::hideEvent(ev);
 }
 
+
+#if MICROPROFILE_ENABLED
+
 /// There's no way to pass a user pointer to MicroProfile, so this variable is used to make the
 /// QPainter available inside the drawing callbacks.
 static QPainter* mp_painter = nullptr;
@@ -337,3 +327,4 @@ void MicroProfileDrawLine2D(u32 vertices_length, float* vertices, u32 hex_color)
     mp_painter->drawPolyline(point_buf.data(), vertices_length);
     point_buf.clear();
 }
+#endif
diff --git a/src/citra_qt/debugger/profiler.h b/src/citra_qt/debugger/profiler.h
index 036054740..3b38ed8ec 100644
--- a/src/citra_qt/debugger/profiler.h
+++ b/src/citra_qt/debugger/profiler.h
@@ -7,8 +7,10 @@
 #include <QAbstractItemModel>
 #include <QDockWidget>
 #include <QTimer>
+
 #include "ui_profiler.h"
 
+#include "common/microprofile.h"
 #include "common/profiler_reporting.h"
 
 class ProfilerModel : public QAbstractItemModel
@@ -49,6 +51,7 @@ private:
     QTimer update_timer;
 };
 
+
 class MicroProfileDialog : public QWidget {
     Q_OBJECT
 
diff --git a/src/citra_qt/main.cpp b/src/citra_qt/main.cpp
index 2ca1e51f6..f1ab29755 100644
--- a/src/citra_qt/main.cpp
+++ b/src/citra_qt/main.cpp
@@ -69,8 +69,10 @@ GMainWindow::GMainWindow() : config(new Config()), emu_thread(nullptr)
     addDockWidget(Qt::BottomDockWidgetArea, profilerWidget);
     profilerWidget->hide();
 
+#if MICROPROFILE_ENABLED
     microProfileDialog = new MicroProfileDialog(this);
     microProfileDialog->hide();
+#endif
 
     disasmWidget = new DisassemblerWidget(this, emu_thread.get());
     addDockWidget(Qt::BottomDockWidgetArea, disasmWidget);
@@ -110,7 +112,9 @@ GMainWindow::GMainWindow() : config(new Config()), emu_thread(nullptr)
 
     QMenu* debug_menu = ui.menu_View->addMenu(tr("Debugging"));
     debug_menu->addAction(profilerWidget->toggleViewAction());
+#if MICROPROFILE_ENABLED
     debug_menu->addAction(microProfileDialog->toggleViewAction());
+#endif
     debug_menu->addAction(disasmWidget->toggleViewAction());
     debug_menu->addAction(registersWidget->toggleViewAction());
     debug_menu->addAction(callstackWidget->toggleViewAction());
@@ -136,8 +140,10 @@ GMainWindow::GMainWindow() : config(new Config()), emu_thread(nullptr)
     restoreGeometry(UISettings::values.geometry);
     restoreState(UISettings::values.state);
     render_window->restoreGeometry(UISettings::values.renderwindow_geometry);
+#if MICROPROFILE_ENABLED
     microProfileDialog->restoreGeometry(UISettings::values.microprofile_geometry);
     microProfileDialog->setVisible(UISettings::values.microprofile_visible);
+#endif
 
     game_list->LoadInterfaceLayout();
 
@@ -511,9 +517,10 @@ void GMainWindow::closeEvent(QCloseEvent* event) {
     UISettings::values.geometry = saveGeometry();
     UISettings::values.state = saveState();
     UISettings::values.renderwindow_geometry = render_window->saveGeometry();
+#if MICROPROFILE_ENABLED
     UISettings::values.microprofile_geometry = microProfileDialog->saveGeometry();
     UISettings::values.microprofile_visible = microProfileDialog->isVisible();
-
+#endif
     UISettings::values.single_window_mode = ui.action_Single_Window_Mode->isChecked();
     UISettings::values.display_titlebar = ui.actionDisplay_widget_title_bars->isChecked();
     UISettings::values.first_start = false;
diff --git a/src/common/CMakeLists.txt b/src/common/CMakeLists.txt
index c839ce173..aa6eee2a3 100644
--- a/src/common/CMakeLists.txt
+++ b/src/common/CMakeLists.txt
@@ -47,7 +47,6 @@ set(HEADERS
             microprofile.h
             microprofileui.h
             platform.h
-            profiler.h
             profiler_reporting.h
             scm_rev.h
             scope_exit.h
diff --git a/src/common/microprofile.h b/src/common/microprofile.h
index d3b6cb97c..ef312c6e1 100644
--- a/src/common/microprofile.h
+++ b/src/common/microprofile.h
@@ -4,6 +4,10 @@
 
 #pragma once
 
+// Uncomment this to disable microprofile. This will get you cleaner profiles when using
+// external sampling profilers like "Very Sleepy", and will improve performance somewhat.
+// #define MICROPROFILE_ENABLED 0
+
 // Customized Citra settings.
 // This file wraps the MicroProfile header so that these are consistent everywhere.
 #define MICROPROFILE_WEBSERVER 0
diff --git a/src/common/microprofileui.h b/src/common/microprofileui.h
index 97c369bd9..41abe6b75 100644
--- a/src/common/microprofileui.h
+++ b/src/common/microprofileui.h
@@ -13,4 +13,7 @@
 #define MICROPROFILE_HELP_ALT "Right-Click"
 #define MICROPROFILE_HELP_MOD "Ctrl"
 
+// This isn't included by microprofileui.h :(
+#include <cstdlib> // For std::abs
+
 #include <microprofileui.h>
diff --git a/src/common/profiler.cpp b/src/common/profiler.cpp
index 7792edd2f..49eb3f40c 100644
--- a/src/common/profiler.cpp
+++ b/src/common/profiler.cpp
@@ -7,71 +7,16 @@
 #include <vector>
 
 #include "common/assert.h"
-#include "common/profiler.h"
 #include "common/profiler_reporting.h"
 #include "common/synchronized_wrapper.h"
 
-#if defined(_MSC_VER) && _MSC_VER <= 1800 // MSVC 2013.
-    #define WIN32_LEAN_AND_MEAN
-    #include <Windows.h> // For QueryPerformanceCounter/Frequency
-#endif
-
 namespace Common {
 namespace Profiling {
 
-#if ENABLE_PROFILING
-thread_local Timer* Timer::current_timer = nullptr;
-#endif
-
-#if defined(_MSC_VER) && _MSC_VER <= 1800 // MSVC 2013
-QPCClock::time_point QPCClock::now() {
-    static LARGE_INTEGER freq;
-    // Use this dummy local static to ensure this gets initialized once.
-    static BOOL dummy = QueryPerformanceFrequency(&freq);
-
-    LARGE_INTEGER ticks;
-    QueryPerformanceCounter(&ticks);
-
-    // This is prone to overflow when multiplying, which is why I'm using micro instead of nano. The
-    // correct way to approach this would be to just return ticks as a time_point and then subtract
-    // and do this conversion when creating a duration from two time_points, however, as far as I
-    // could tell the C++ requirements for these types are incompatible with this approach.
-    return time_point(duration(ticks.QuadPart * std::micro::den / freq.QuadPart));
-}
-#endif
-
-TimingCategory::TimingCategory(const char* name, TimingCategory* parent)
-        : accumulated_duration(0) {
-
-    ProfilingManager& manager = GetProfilingManager();
-    category_id = manager.RegisterTimingCategory(this, name);
-    if (parent != nullptr)
-        manager.SetTimingCategoryParent(category_id, parent->category_id);
-}
-
 ProfilingManager::ProfilingManager()
         : last_frame_end(Clock::now()), this_frame_start(Clock::now()) {
 }
 
-unsigned int ProfilingManager::RegisterTimingCategory(TimingCategory* category, const char* name) {
-    TimingCategoryInfo info;
-    info.category = category;
-    info.name = name;
-    info.parent = TimingCategoryInfo::NO_PARENT;
-
-    unsigned int id = (unsigned int)timing_categories.size();
-    timing_categories.push_back(std::move(info));
-
-    return id;
-}
-
-void ProfilingManager::SetTimingCategoryParent(unsigned int category, unsigned int parent) {
-    ASSERT(category < timing_categories.size());
-    ASSERT(parent < timing_categories.size());
-
-    timing_categories[category].parent = parent;
-}
-
 void ProfilingManager::BeginFrame() {
     this_frame_start = Clock::now();
 }
@@ -82,11 +27,6 @@ void ProfilingManager::FinishFrame() {
     results.interframe_time = now - last_frame_end;
     results.frame_time = now - this_frame_start;
 
-    results.time_per_category.resize(timing_categories.size());
-    for (size_t i = 0; i < timing_categories.size(); ++i) {
-        results.time_per_category[i] = timing_categories[i].category->GetAccumulatedTime();
-    }
-
     last_frame_end = now;
 }
 
@@ -100,26 +40,9 @@ void TimingResultsAggregator::Clear() {
     window_size = cursor = 0;
 }
 
-void TimingResultsAggregator::SetNumberOfCategories(size_t n) {
-    size_t old_size = times_per_category.size();
-    if (n == old_size)
-        return;
-
-    times_per_category.resize(n);
-
-    for (size_t i = old_size; i < n; ++i) {
-        times_per_category[i].resize(max_window_size, Duration::zero());
-    }
-}
-
 void TimingResultsAggregator::AddFrame(const ProfilingFrameResult& frame_result) {
-    SetNumberOfCategories(frame_result.time_per_category.size());
-
     interframe_times[cursor] = frame_result.interframe_time;
     frame_times[cursor] = frame_result.frame_time;
-    for (size_t i = 0; i < frame_result.time_per_category.size(); ++i) {
-        times_per_category[i][cursor] = frame_result.time_per_category[i];
-    }
 
     ++cursor;
     if (cursor == max_window_size)
@@ -162,11 +85,6 @@ AggregatedFrameResult TimingResultsAggregator::GetAggregatedResults() const {
         result.fps = 0.0f;
     }
 
-    result.time_per_category.resize(times_per_category.size());
-    for (size_t i = 0; i < times_per_category.size(); ++i) {
-        result.time_per_category[i] = AggregateField(times_per_category[i], window_size);
-    }
-
     return result;
 }
 
diff --git a/src/common/profiler.h b/src/common/profiler.h
deleted file mode 100644
index 3e967b4bc..000000000
--- a/src/common/profiler.h
+++ /dev/null
@@ -1,152 +0,0 @@
-// Copyright 2015 Citra Emulator Project
-// Licensed under GPLv2 or any later version
-// Refer to the license.txt file included.
-
-#pragma once
-
-#include <atomic>
-#include <chrono>
-
-#include "common/assert.h"
-#include "common/thread.h"
-
-namespace Common {
-namespace Profiling {
-
-// If this is defined to 0, it turns all Timers into no-ops.
-#ifndef ENABLE_PROFILING
-#define ENABLE_PROFILING 1
-#endif
-
-#if defined(_MSC_VER) && _MSC_VER <= 1800 // MSVC 2013
-// MSVC up to 2013 doesn't use QueryPerformanceCounter for high_resolution_clock, so it has bad
-// precision. We manually implement a clock based on QPC to get good results.
-
-struct QPCClock {
-    using duration = std::chrono::microseconds;
-    using time_point = std::chrono::time_point<QPCClock>;
-    using rep = duration::rep;
-    using period = duration::period;
-    static const bool is_steady = false;
-
-    static time_point now();
-};
-
-using Clock = QPCClock;
-#else
-using Clock = std::chrono::high_resolution_clock;
-#endif
-
-using Duration = Clock::duration;
-
-/**
- * Represents a timing category that measured time can be accounted towards. Should be declared as a
- * global variable and passed to Timers.
- */
-class TimingCategory final {
-public:
-    TimingCategory(const char* name, TimingCategory* parent = nullptr);
-
-    unsigned int GetCategoryId() const {
-        return category_id;
-    }
-
-    /// Adds some time to this category. Can safely be called from multiple threads at the same time.
-    void AddTime(Duration amount) {
-        std::atomic_fetch_add_explicit(
-                &accumulated_duration, amount.count(),
-                std::memory_order_relaxed);
-    }
-
-    /**
-     * Atomically retrieves the accumulated measured time for this category and resets the counter
-     * to zero. Can be safely called concurrently with AddTime.
-     */
-    Duration GetAccumulatedTime() {
-        return Duration(std::atomic_exchange_explicit(
-                &accumulated_duration, (Duration::rep)0,
-                std::memory_order_relaxed));
-    }
-
-private:
-    unsigned int category_id;
-    std::atomic<Duration::rep> accumulated_duration;
-};
-
-/**
- * Measures time elapsed between a call to Start and a call to Stop and attributes it to the given
- * TimingCategory. Start/Stop can be called multiple times on the same timer, but each call must be
- * appropriately paired.
- *
- * When a Timer is started, it automatically pauses a previously running timer on the same thread,
- * which is resumed when it is stopped. As such, no special action needs to be taken to avoid
- * double-accounting of time on two categories.
- */
-class Timer {
-public:
-    Timer(TimingCategory& category) : category(category) {
-    }
-
-    void Start() {
-#if ENABLE_PROFILING
-        ASSERT(!running);
-        previous_timer = current_timer;
-        current_timer = this;
-        if (previous_timer != nullptr)
-            previous_timer->StopTiming();
-
-        StartTiming();
-#endif
-    }
-
-    void Stop() {
-#if ENABLE_PROFILING
-        ASSERT(running);
-        StopTiming();
-
-        if (previous_timer != nullptr)
-            previous_timer->StartTiming();
-        current_timer = previous_timer;
-#endif
-    }
-
-private:
-#if ENABLE_PROFILING
-    void StartTiming() {
-        start = Clock::now();
-        running = true;
-    }
-
-    void StopTiming() {
-        auto duration = Clock::now() - start;
-        running = false;
-        category.AddTime(std::chrono::duration_cast<Duration>(duration));
-    }
-
-    Clock::time_point start;
-    bool running = false;
-
-    Timer* previous_timer;
-    static thread_local Timer* current_timer;
-#endif
-
-    TimingCategory& category;
-};
-
-/**
- * A Timer that automatically starts timing when created and stops at the end of the scope. Should
- * be used in the majority of cases.
- */
-class ScopeTimer : public Timer {
-public:
-    ScopeTimer(TimingCategory& category) : Timer(category) {
-        Start();
-    }
-
-    ~ScopeTimer() {
-        Stop();
-    }
-};
-
-} // namespace Profiling
-} // namespace Common
diff --git a/src/common/profiler_reporting.h b/src/common/profiler_reporting.h
index df98e05b7..fa1ac883f 100644
--- a/src/common/profiler_reporting.h
+++ b/src/common/profiler_reporting.h
@@ -4,22 +4,17 @@
 
 #pragma once
 
+#include <chrono>
 #include <cstddef>
 #include <vector>
 
-#include "common/profiler.h"
 #include "common/synchronized_wrapper.h"
 
 namespace Common {
 namespace Profiling {
 
-struct TimingCategoryInfo {
-    static const unsigned int NO_PARENT = -1;
-
-    TimingCategory* category;
-    const char* name;
-    unsigned int parent;
-};
+using Clock = std::chrono::high_resolution_clock;
+using Duration = Clock::duration;
 
 struct ProfilingFrameResult {
     /// Time since the last delivered frame
@@ -27,22 +22,12 @@ struct ProfilingFrameResult {
 
     /// Time spent processing a frame, excluding VSync
     Duration frame_time;
-
-    /// Total amount of time spent inside each category in this frame. Indexed by the category id
-    std::vector<Duration> time_per_category;
 };
 
 class ProfilingManager final {
 public:
     ProfilingManager();
 
-    unsigned int RegisterTimingCategory(TimingCategory* category, const char* name);
-    void SetTimingCategoryParent(unsigned int category, unsigned int parent);
-
-    const std::vector<TimingCategoryInfo>& GetTimingCategoriesInfo() const {
-        return timing_categories;
-    }
-
     /// This should be called after swapping screen buffers.
     void BeginFrame();
     /// This should be called before swapping screen buffers.
@@ -54,7 +39,6 @@ public:
     }
 
 private:
-    std::vector<TimingCategoryInfo> timing_categories;
     Clock::time_point last_frame_end;
     Clock::time_point this_frame_start;
 
@@ -73,9 +57,6 @@ struct AggregatedFrameResult {
     AggregatedDuration frame_time;
 
     float fps;
-
-    /// Total amount of time spent inside each category in this frame. Indexed by the category id
-    std::vector<AggregatedDuration> time_per_category;
 };
 
 class TimingResultsAggregator final {
@@ -83,7 +64,6 @@ public:
     TimingResultsAggregator(size_t window_size);
 
     void Clear();
-    void SetNumberOfCategories(size_t n);
 
     void AddFrame(const ProfilingFrameResult& frame_result);
 
@@ -95,7 +75,6 @@ public:
 
     std::vector<Duration> interframe_times;
     std::vector<Duration> frame_times;
-    std::vector<std::vector<Duration>> times_per_category;
 };
 
 ProfilingManager& GetProfilingManager();
diff --git a/src/core/arm/dyncom/arm_dyncom_interpreter.cpp b/src/core/arm/dyncom/arm_dyncom_interpreter.cpp
index 647784208..8d4b26815 100644
--- a/src/core/arm/dyncom/arm_dyncom_interpreter.cpp
+++ b/src/core/arm/dyncom/arm_dyncom_interpreter.cpp
@@ -10,7 +10,6 @@
 #include "common/common_types.h"
 #include "common/logging/log.h"
 #include "common/microprofile.h"
-#include "common/profiler.h"
 
 #include "core/memory.h"
 #include "core/hle/svc.h"
@@ -25,9 +24,6 @@
 
 #include "core/gdbstub/gdbstub.h"
 
-Common::Profiling::TimingCategory profile_execute("DynCom::Execute");
-Common::Profiling::TimingCategory profile_decode("DynCom::Decode");
-
 enum {
     COND            = (1 << 0),
     NON_BRANCH      = (1 << 1),
@@ -3496,7 +3492,6 @@ static unsigned int InterpreterTranslateInstruction(const ARMul_State* cpu, cons
 }
 
 static int InterpreterTranslateBlock(ARMul_State* cpu, int& bb_start, u32 addr) {
-    Common::Profiling::ScopeTimer timer_decode(profile_decode);
     MICROPROFILE_SCOPE(DynCom_Decode);
 
     // Decode instruction, get index
@@ -3530,7 +3525,6 @@ static int InterpreterTranslateBlock(ARMul_State* cpu, int& bb_start, u32 addr)
 }
 
 static int InterpreterTranslateSingle(ARMul_State* cpu, int& bb_start, u32 addr) {
-    Common::Profiling::ScopeTimer timer_decode(profile_decode);
     MICROPROFILE_SCOPE(DynCom_Decode);
 
     ARM_INST_PTR inst_base = nullptr;
@@ -3565,7 +3559,6 @@ static int clz(unsigned int x) {
 MICROPROFILE_DEFINE(DynCom_Execute, "DynCom", "Execute", MP_RGB(255, 0, 0));
 
 unsigned InterpreterMainLoop(ARMul_State* cpu) {
-    Common::Profiling::ScopeTimer timer_execute(profile_execute);
     MICROPROFILE_SCOPE(DynCom_Execute);
 
     GDBStub::BreakpointAddress breakpoint_data;
diff --git a/src/core/hle/result.h b/src/core/hle/result.h
index 2d22652d9..53931a106 100644
--- a/src/core/hle/result.h
+++ b/src/core/hle/result.h
@@ -18,6 +18,7 @@
 /// Detailed description of the error. This listing is likely incomplete.
 enum class ErrorDescription : u32 {
     Success = 0,
+    OS_InvalidBufferDescriptor = 48,
     WrongAddress = 53,
     FS_NotFound = 120,
     FS_AlreadyExists = 190,
diff --git a/src/core/hle/service/dsp_dsp.cpp b/src/core/hle/service/dsp_dsp.cpp
index 08e437125..995bee3f9 100644
--- a/src/core/hle/service/dsp_dsp.cpp
+++ b/src/core/hle/service/dsp_dsp.cpp
@@ -2,6 +2,7 @@
 // Licensed under GPLv2 or any later version
 // Refer to the license.txt file included.
 
+#include <algorithm>
 #include <cinttypes>
 
 #include "audio_core/hle/pipe.h"
@@ -12,37 +13,80 @@
 #include "core/hle/kernel/event.h"
 #include "core/hle/service/dsp_dsp.h"
 
+using DspPipe = DSP::HLE::DspPipe;
+
 ////////////////////////////////////////////////////////////////////////////////////////////////////
 // Namespace DSP_DSP
 
 namespace DSP_DSP {
 
-static u32 read_pipe_count;
 static Kernel::SharedPtr<Kernel::Event> semaphore_event;
 
-struct PairHash {
-    template <typename T, typename U>
-    std::size_t operator()(const std::pair<T, U> &x) const {
-        // TODO(yuriks): Replace with better hash combining function.
-        return std::hash<T>()(x.first) ^ std::hash<U>()(x.second);
+/// There are three types of interrupts
+enum class InterruptType {
+    Zero, One, Pipe
+};
+constexpr size_t NUM_INTERRUPT_TYPE = 3;
+
+class InterruptEvents final {
+public:
+    void Signal(InterruptType type, DspPipe pipe) {
+        Kernel::SharedPtr<Kernel::Event>& event = Get(type, pipe);
+        if (event) {
+            event->Signal();
+        }
     }
+
+    Kernel::SharedPtr<Kernel::Event>& Get(InterruptType type, DspPipe dsp_pipe) {
+        switch (type) {
+        case InterruptType::Zero:
+            return zero;
+        case InterruptType::One:
+            return one;
+        case InterruptType::Pipe: {
+            const size_t pipe_index = static_cast<size_t>(dsp_pipe);
+            ASSERT(pipe_index < DSP::HLE::NUM_DSP_PIPE);
+            return pipe[pipe_index];
+        }
+        }
+
+        UNREACHABLE_MSG("Invalid interrupt type = %zu", static_cast<size_t>(type));
+    }
+
+    bool HasTooManyEventsRegistered() const {
+        // Actual service implementation only has 6 'slots' for interrupts.
+        constexpr size_t max_number_of_interrupt_events = 6;
+
+        size_t number = std::count_if(pipe.begin(), pipe.end(), [](const auto& evt) {
+            return evt != nullptr;
+        });
+
+        if (zero != nullptr)
+            number++;
+        if (one != nullptr)
+            number++;
+
+        return number >= max_number_of_interrupt_events;
+    }
+
+private:
+    /// Currently unknown purpose
+    Kernel::SharedPtr<Kernel::Event> zero = nullptr;
+    /// Currently unknown purpose
+    Kernel::SharedPtr<Kernel::Event> one = nullptr;
+    /// Each DSP pipe has an associated interrupt
+    std::array<Kernel::SharedPtr<Kernel::Event>, DSP::HLE::NUM_DSP_PIPE> pipe = {{}};
 };
 
-/// Map of (audio interrupt number, channel number) to Kernel::Events. See: RegisterInterruptEvents
-static std::unordered_map<std::pair<u32, u32>, Kernel::SharedPtr<Kernel::Event>, PairHash> interrupt_events;
+static InterruptEvents interrupt_events;
 
 // DSP Interrupts:
-// Interrupt #2 occurs every frame tick. Userland programs normally have a thread that's waiting
-// for an interrupt event. Immediately after this interrupt event, userland normally updates the
-// state in the next region and increments the relevant frame counter by two.
-void SignalAllInterrupts() {
-    // HACK: The other interrupts have currently unknown purpose, we trigger them each tick in any case.
-    for (auto& interrupt_event : interrupt_events)
-        interrupt_event.second->Signal();
-}
-
-void SignalInterrupt(u32 interrupt, u32 channel) {
-    interrupt_events[std::make_pair(interrupt, channel)]->Signal();
+// The audio-pipe interrupt occurs every frame tick. Userland programs normally have a thread
+// that's waiting for an interrupt event. Immediately after this interrupt event, userland
+// normally updates the state in the next region and increments the relevant frame counter by
+// two.
+void SignalPipeInterrupt(DspPipe pipe) {
+    interrupt_events.Signal(InterruptType::Pipe, pipe);
 }
 
 /**
@@ -58,7 +102,10 @@ static void ConvertProcessAddressFromDspDram(Service::Interface* self) {
 
     u32 addr = cmd_buff[1];
 
+    cmd_buff[0] = IPC::MakeHeader(0xC, 2, 0);
     cmd_buff[1] = RESULT_SUCCESS.raw; // No error
+
+    // TODO(merry): There is a per-region offset missing in this calculation (that seems to be always zero).
     cmd_buff[2] = (addr << 1) + (Memory::DSP_RAM_VADDR + 0x40000);
 
     LOG_DEBUG(Service_DSP, "addr=0x%08X", addr);
@@ -113,7 +160,9 @@ static void LoadComponent(Service::Interface* self) {
 static void GetSemaphoreEventHandle(Service::Interface* self) {
     u32* cmd_buff = Kernel::GetCommandBuffer();
 
+    cmd_buff[0] = IPC::MakeHeader(0x16, 1, 2);
     cmd_buff[1] = RESULT_SUCCESS.raw; // No error
+    // cmd_buff[2] not set
     cmd_buff[3] = Kernel::g_handle_table.Create(semaphore_event).MoveFrom(); // Event handle
 
     LOG_WARNING(Service_DSP, "(STUBBED) called");
@@ -138,8 +187,7 @@ static void FlushDataCache(Service::Interface* self) {
     u32 size    = cmd_buff[2];
     u32 process = cmd_buff[4];
 
-    // TODO(purpasmart96): Verify return header on HW
-
+    cmd_buff[0] = IPC::MakeHeader(0x13, 1, 0);
     cmd_buff[1] = RESULT_SUCCESS.raw; // No error
 
     LOG_TRACE(Service_DSP, "called address=0x%08X, size=0x%X, process=0x%08X", address, size, process);
@@ -148,8 +196,8 @@ static void FlushDataCache(Service::Interface* self) {
 /**
  * DSP_DSP::RegisterInterruptEvents service function
  *  Inputs:
- *      1 : Interrupt Number
- *      2 : Channel Number
+ *      1 : Interrupt Type
+ *      2 : Pipe Number
  *      4 : Interrupt event handle
  *  Outputs:
  *      1 : Result of function, 0 on success, otherwise error code
@@ -157,23 +205,40 @@ static void FlushDataCache(Service::Interface* self) {
 static void RegisterInterruptEvents(Service::Interface* self) {
     u32* cmd_buff = Kernel::GetCommandBuffer();
 
-    u32 interrupt = cmd_buff[1];
-    u32 channel = cmd_buff[2];
+    u32 type_index = cmd_buff[1];
+    u32 pipe_index = cmd_buff[2];
     u32 event_handle = cmd_buff[4];
 
+    ASSERT_MSG(type_index < NUM_INTERRUPT_TYPE && pipe_index < DSP::HLE::NUM_DSP_PIPE,
+               "Invalid type or pipe: type = %u, pipe = %u", type_index, pipe_index);
+
+    InterruptType type = static_cast<InterruptType>(cmd_buff[1]);
+    DspPipe pipe = static_cast<DspPipe>(cmd_buff[2]);
+
+    cmd_buff[0] = IPC::MakeHeader(0x15, 1, 0);
+
     if (event_handle) {
         auto evt = Kernel::g_handle_table.Get<Kernel::Event>(cmd_buff[4]);
-        if (evt) {
-            interrupt_events[std::make_pair(interrupt, channel)] = evt;
-            cmd_buff[1] = RESULT_SUCCESS.raw;
-            LOG_INFO(Service_DSP, "Registered interrupt=%u, channel=%u, event_handle=0x%08X", interrupt, channel, event_handle);
-        } else {
-            LOG_CRITICAL(Service_DSP, "Invalid event handle! interrupt=%u, channel=%u, event_handle=0x%08X", interrupt, channel, event_handle);
-            ASSERT(false); // This should really be handled at a IPC translation layer.
+
+        if (!evt) {
+            LOG_INFO(Service_DSP, "Invalid event handle! type=%u, pipe=%u, event_handle=0x%08X", type_index, pipe_index, event_handle);
+            ASSERT(false); // TODO: This should really be handled at an IPC translation layer.
+        }
+
+        if (interrupt_events.HasTooManyEventsRegistered()) {
+            LOG_INFO(Service_DSP, "Ran out of space to register interrupts (Attempted to register type=%u, pipe=%u, event_handle=0x%08X)",
+                     type_index, pipe_index, event_handle);
+            cmd_buff[1] = ResultCode(ErrorDescription::InvalidResultValue, ErrorModule::DSP, ErrorSummary::OutOfResource, ErrorLevel::Status).raw;
+            return;
         }
+
+        interrupt_events.Get(type, pipe) = evt;
+        LOG_INFO(Service_DSP, "Registered type=%u, pipe=%u, event_handle=0x%08X", type_index, pipe_index, event_handle);
+        cmd_buff[1] = RESULT_SUCCESS.raw;
     } else {
-        interrupt_events.erase(std::make_pair(interrupt, channel));
-        LOG_INFO(Service_DSP, "Unregistered interrupt=%u, channel=%u, event_handle=0x%08X", interrupt, channel, event_handle);
+        interrupt_events.Get(type, pipe) = nullptr;
+        LOG_INFO(Service_DSP, "Unregistered interrupt=%u, channel=%u, event_handle=0x%08X", type_index, pipe_index, event_handle);
+        cmd_buff[1] = RESULT_SUCCESS.raw;
     }
 }
 
@@ -187,6 +252,7 @@ static void RegisterInterruptEvents(Service::Interface* self) {
 static void SetSemaphore(Service::Interface* self) {
     u32* cmd_buff = Kernel::GetCommandBuffer();
 
+    cmd_buff[0] = IPC::MakeHeader(0x7, 1, 0);
     cmd_buff[1] = RESULT_SUCCESS.raw; // No error
 
     LOG_WARNING(Service_DSP, "(STUBBED) called");
@@ -195,7 +261,7 @@ static void SetSemaphore(Service::Interface* self) {
 /**
  * DSP_DSP::WriteProcessPipe service function
  *  Inputs:
- *      1 : Channel
+ *      1 : Pipe Number
  *      2 : Size
  *      3 : (size << 14) | 0x402
  *      4 : Buffer
@@ -206,24 +272,32 @@ static void SetSemaphore(Service::Interface* self) {
 static void WriteProcessPipe(Service::Interface* self) {
     u32* cmd_buff = Kernel::GetCommandBuffer();
 
-    DSP::HLE::DspPipe pipe = static_cast<DSP::HLE::DspPipe>(cmd_buff[1]);
+    u32 pipe_index = cmd_buff[1];
     u32 size = cmd_buff[2];
     u32 buffer = cmd_buff[4];
 
-    ASSERT_MSG(IPC::StaticBufferDesc(size, 1) == cmd_buff[3], "IPC static buffer descriptor failed validation (0x%X). pipe=%u, size=0x%X, buffer=0x%08X", cmd_buff[3], pipe, size, buffer);
-    ASSERT_MSG(Memory::GetPointer(buffer) != nullptr, "Invalid Buffer: pipe=%u, size=0x%X, buffer=0x%08X", pipe, size, buffer);
+    DSP::HLE::DspPipe pipe = static_cast<DSP::HLE::DspPipe>(pipe_index);
 
-    std::vector<u8> message(size);
+    if (IPC::StaticBufferDesc(size, 1) != cmd_buff[3]) {
+        LOG_ERROR(Service_DSP, "IPC static buffer descriptor failed validation (0x%X). pipe=%u, size=0x%X, buffer=0x%08X", cmd_buff[3], pipe_index, size, buffer);
+        cmd_buff[0] = IPC::MakeHeader(0, 1, 0);
+        cmd_buff[1] = ResultCode(ErrorDescription::OS_InvalidBufferDescriptor, ErrorModule::OS, ErrorSummary::WrongArgument, ErrorLevel::Permanent).raw;
+        return;
+    }
+
+    ASSERT_MSG(Memory::GetPointer(buffer) != nullptr, "Invalid Buffer: pipe=%u, size=0x%X, buffer=0x%08X", pipe_index, size, buffer);
 
+    std::vector<u8> message(size);
     for (size_t i = 0; i < size; i++) {
         message[i] = Memory::Read8(buffer + i);
     }
 
     DSP::HLE::PipeWrite(pipe, message);
 
+    cmd_buff[0] = IPC::MakeHeader(0xD, 1, 0);
     cmd_buff[1] = RESULT_SUCCESS.raw; // No error
 
-    LOG_DEBUG(Service_DSP, "pipe=%u, size=0x%X, buffer=0x%08X", pipe, size, buffer);
+    LOG_DEBUG(Service_DSP, "pipe=%u, size=0x%X, buffer=0x%08X", pipe_index, size, buffer);
 }
 
 /**
@@ -243,13 +317,16 @@ static void WriteProcessPipe(Service::Interface* self) {
 static void ReadPipeIfPossible(Service::Interface* self) {
     u32* cmd_buff = Kernel::GetCommandBuffer();
 
-    DSP::HLE::DspPipe pipe = static_cast<DSP::HLE::DspPipe>(cmd_buff[1]);
+    u32 pipe_index = cmd_buff[1];
     u32 unknown = cmd_buff[2];
     u32 size = cmd_buff[3] & 0xFFFF; // Lower 16 bits are size
     VAddr addr = cmd_buff[0x41];
 
-    ASSERT_MSG(Memory::GetPointer(addr) != nullptr, "Invalid addr: pipe=0x%08X, unknown=0x%08X, size=0x%X, buffer=0x%08X", pipe, unknown, size, addr);
+    DSP::HLE::DspPipe pipe = static_cast<DSP::HLE::DspPipe>(pipe_index);
+
+    ASSERT_MSG(Memory::GetPointer(addr) != nullptr, "Invalid addr: pipe=%u, unknown=0x%08X, size=0x%X, buffer=0x%08X", pipe_index, unknown, size, addr);
 
+    cmd_buff[0] = IPC::MakeHeader(0x10, 1, 2);
     cmd_buff[1] = RESULT_SUCCESS.raw; // No error
     if (DSP::HLE::GetPipeReadableSize(pipe) >= size) {
         std::vector<u8> response = DSP::HLE::PipeRead(pipe, size);
@@ -260,8 +337,10 @@ static void ReadPipeIfPossible(Service::Interface* self) {
     } else {
         cmd_buff[2] = 0; // Return no data
     }
+    cmd_buff[3] = IPC::StaticBufferDesc(size, 0);
+    cmd_buff[4] = addr;
 
-    LOG_DEBUG(Service_DSP, "pipe=0x%08X, unknown=0x%08X, size=0x%X, buffer=0x%08X, return cmd_buff[2]=0x%08X", pipe, unknown, size, addr, cmd_buff[2]);
+    LOG_DEBUG(Service_DSP, "pipe=%u, unknown=0x%08X, size=0x%X, buffer=0x%08X, return cmd_buff[2]=0x%08X", pipe_index, unknown, size, addr, cmd_buff[2]);
 }
 
 /**
@@ -278,26 +357,31 @@ static void ReadPipeIfPossible(Service::Interface* self) {
 static void ReadPipe(Service::Interface* self) {
     u32* cmd_buff = Kernel::GetCommandBuffer();
 
-    DSP::HLE::DspPipe pipe = static_cast<DSP::HLE::DspPipe>(cmd_buff[1]);
+    u32 pipe_index = cmd_buff[1];
     u32 unknown = cmd_buff[2];
     u32 size = cmd_buff[3] & 0xFFFF; // Lower 16 bits are size
     VAddr addr = cmd_buff[0x41];
 
-    ASSERT_MSG(Memory::GetPointer(addr) != nullptr, "Invalid addr: pipe=0x%08X, unknown=0x%08X, size=0x%X, buffer=0x%08X", pipe, unknown, size, addr);
+    DSP::HLE::DspPipe pipe = static_cast<DSP::HLE::DspPipe>(pipe_index);
+
+    ASSERT_MSG(Memory::GetPointer(addr) != nullptr, "Invalid addr: pipe=%u, unknown=0x%08X, size=0x%X, buffer=0x%08X", pipe_index, unknown, size, addr);
 
     if (DSP::HLE::GetPipeReadableSize(pipe) >= size) {
         std::vector<u8> response = DSP::HLE::PipeRead(pipe, size);
 
         Memory::WriteBlock(addr, response.data(), response.size());
 
+        cmd_buff[0] = IPC::MakeHeader(0xE, 2, 2);
         cmd_buff[1] = RESULT_SUCCESS.raw; // No error
         cmd_buff[2] = static_cast<u32>(response.size());
+        cmd_buff[3] = IPC::StaticBufferDesc(size, 0);
+        cmd_buff[4] = addr;
     } else {
         // No more data is in pipe. Hardware hangs in this case; this should never happen.
         UNREACHABLE();
     }
 
-    LOG_DEBUG(Service_DSP, "pipe=0x%08X, unknown=0x%08X, size=0x%X, buffer=0x%08X, return cmd_buff[2]=0x%08X", pipe, unknown, size, addr, cmd_buff[2]);
+    LOG_DEBUG(Service_DSP, "pipe=%u, unknown=0x%08X, size=0x%X, buffer=0x%08X, return cmd_buff[2]=0x%08X", pipe_index, unknown, size, addr, cmd_buff[2]);
 }
 
 /**
@@ -312,13 +396,16 @@ static void ReadPipe(Service::Interface* self) {
 static void GetPipeReadableSize(Service::Interface* self) {
     u32* cmd_buff = Kernel::GetCommandBuffer();
 
-    DSP::HLE::DspPipe pipe = static_cast<DSP::HLE::DspPipe>(cmd_buff[1]);
+    u32 pipe_index = cmd_buff[1];
     u32 unknown = cmd_buff[2];
 
+    DSP::HLE::DspPipe pipe = static_cast<DSP::HLE::DspPipe>(pipe_index);
+
+    cmd_buff[0] = IPC::MakeHeader(0xF, 2, 0);
     cmd_buff[1] = RESULT_SUCCESS.raw; // No error
     cmd_buff[2] = DSP::HLE::GetPipeReadableSize(pipe);
 
-    LOG_DEBUG(Service_DSP, "pipe=0x%08X, unknown=0x%08X, return cmd_buff[2]=0x%08X", pipe, unknown, cmd_buff[2]);
+    LOG_DEBUG(Service_DSP, "pipe=%u, unknown=0x%08X, return cmd_buff[2]=0x%08X", pipe_index, unknown, cmd_buff[2]);
 }
 
 /**
@@ -333,6 +420,7 @@ static void SetSemaphoreMask(Service::Interface* self) {
 
     u32 mask = cmd_buff[1];
 
+    cmd_buff[0] = IPC::MakeHeader(0x17, 1, 0);
     cmd_buff[1] = RESULT_SUCCESS.raw; // No error
 
     LOG_WARNING(Service_DSP, "(STUBBED) called mask=0x%08X", mask);
@@ -350,6 +438,7 @@ static void SetSemaphoreMask(Service::Interface* self) {
 static void GetHeadphoneStatus(Service::Interface* self) {
     u32* cmd_buff = Kernel::GetCommandBuffer();
 
+    cmd_buff[0] = IPC::MakeHeader(0x1F, 2, 0);
     cmd_buff[1] = RESULT_SUCCESS.raw; // No error
     cmd_buff[2] = 0; // Not using headphones?
 
@@ -376,6 +465,7 @@ static void RecvData(Service::Interface* self) {
 
     // Application reads this after requesting DSP shutdown, to verify the DSP has indeed shutdown or slept.
 
+    cmd_buff[0] = IPC::MakeHeader(0x1, 2, 0);
     cmd_buff[1] = RESULT_SUCCESS.raw;
     switch (DSP::HLE::GetDspState()) {
     case DSP::HLE::DspState::On:
@@ -411,6 +501,7 @@ static void RecvDataIsReady(Service::Interface* self) {
 
     ASSERT_MSG(register_number == 0, "Unknown register_number %u", register_number);
 
+    cmd_buff[0] = IPC::MakeHeader(0x2, 2, 0);
     cmd_buff[1] = RESULT_SUCCESS.raw;
     cmd_buff[2] = 1; // Ready to read
 
@@ -458,14 +549,14 @@ const Interface::FunctionInfo FunctionTable[] = {
 
 Interface::Interface() {
     semaphore_event = Kernel::Event::Create(Kernel::ResetType::OneShot, "DSP_DSP::semaphore_event");
-    read_pipe_count = 0;
+    interrupt_events = {};
 
     Register(FunctionTable);
 }
 
 Interface::~Interface() {
     semaphore_event = nullptr;
-    interrupt_events.clear();
+    interrupt_events = {};
 }
 
 } // namespace
diff --git a/src/core/hle/service/dsp_dsp.h b/src/core/hle/service/dsp_dsp.h
index 32b89e9bb..22f6687cc 100644
--- a/src/core/hle/service/dsp_dsp.h
+++ b/src/core/hle/service/dsp_dsp.h
@@ -8,6 +8,12 @@
 
 #include "core/hle/service/service.h"
 
+namespace DSP {
+namespace HLE {
+enum class DspPipe;
+}
+}
+
 ////////////////////////////////////////////////////////////////////////////////////////////////////
 // Namespace DSP_DSP
 
@@ -23,15 +29,10 @@ public:
     }
 };
 
-/// Signal all audio related interrupts.
-void SignalAllInterrupts();
-
 /**
- * Signal a specific audio related interrupt based on interrupt id and channel id.
- * @param interrupt_id The interrupt id
- * @param channel_id The channel id
- * The significance of various values of interrupt_id and channel_id is not yet known.
+ * Signal a specific DSP related interrupt of type == InterruptType::Pipe, pipe == pipe.
+ * @param pipe The DSP pipe for which to signal an interrupt for.
  */
-void SignalInterrupt(u32 interrupt_id, u32 channel_id);
+void SignalPipeInterrupt(DSP::HLE::DspPipe pipe);
 
-} // namespace
+} // namespace DSP_DSP
diff --git a/src/core/hle/service/gsp_gpu.cpp b/src/core/hle/service/gsp_gpu.cpp
index 211fcf599..233592d7f 100644
--- a/src/core/hle/service/gsp_gpu.cpp
+++ b/src/core/hle/service/gsp_gpu.cpp
@@ -4,7 +4,6 @@
 
 #include "common/bit_field.h"
 #include "common/microprofile.h"
-#include "common/profiler.h"
 
 #include "core/memory.h"
 #include "core/hle/kernel/event.h"
diff --git a/src/core/hle/service/y2r_u.cpp b/src/core/hle/service/y2r_u.cpp
index 1672ad775..d16578f87 100644
--- a/src/core/hle/service/y2r_u.cpp
+++ b/src/core/hle/service/y2r_u.cpp
@@ -4,6 +4,7 @@
 
 #include <cstring>
 
+#include "common/common_funcs.h"
 #include "common/common_types.h"
 #include "common/logging/log.h"
 
@@ -25,13 +26,17 @@ struct ConversionParameters {
     u16 input_line_width;
     u16 input_lines;
     StandardCoefficient standard_coefficient;
-    u8 reserved;
+    u8 padding;
     u16 alpha;
 };
 static_assert(sizeof(ConversionParameters) == 12, "ConversionParameters struct has incorrect size");
 
 static Kernel::SharedPtr<Kernel::Event> completion_event;
 static ConversionConfiguration conversion;
+static DitheringWeightParams dithering_weight_params;
+static u32 temporal_dithering_enabled = 0;
+static u32 transfer_end_interrupt_enabled = 0;
+static u32 spacial_dithering_enabled = 0;
 
 static const CoefficientSet standard_coefficients[4] = {
     {{ 0x100, 0x166, 0xB6, 0x58, 0x1C5, -0x166F, 0x10EE, -0x1C5B }}, // ITU_Rec601
@@ -70,7 +75,7 @@ ResultCode ConversionConfiguration::SetInputLines(u16 lines) {
 
 ResultCode ConversionConfiguration::SetStandardCoefficient(StandardCoefficient standard_coefficient) {
     size_t index = static_cast<size_t>(standard_coefficient);
-    if (index >= 4) {
+    if (index >= ARRAY_SIZE(standard_coefficients)) {
         return ResultCode(ErrorDescription::InvalidEnumValue, ErrorModule::CAM,
             ErrorSummary::InvalidArgument, ErrorLevel::Usage); // 0xE0E053ED
     }
@@ -83,44 +88,183 @@ static void SetInputFormat(Service::Interface* self) {
     u32* cmd_buff = Kernel::GetCommandBuffer();
 
     conversion.input_format = static_cast<InputFormat>(cmd_buff[1]);
+
+    cmd_buff[0] = IPC::MakeHeader(0x1, 1, 0);
+    cmd_buff[1] = RESULT_SUCCESS.raw;
+
     LOG_DEBUG(Service_Y2R, "called input_format=%hhu", conversion.input_format);
+}
 
+static void GetInputFormat(Service::Interface* self) {
+    u32* cmd_buff = Kernel::GetCommandBuffer();
+
+    cmd_buff[0] = IPC::MakeHeader(0x2, 2, 0);
     cmd_buff[1] = RESULT_SUCCESS.raw;
+    cmd_buff[2] = static_cast<u32>(conversion.input_format);
+
+    LOG_DEBUG(Service_Y2R, "called input_format=%hhu", conversion.input_format);
 }
 
 static void SetOutputFormat(Service::Interface* self) {
     u32* cmd_buff = Kernel::GetCommandBuffer();
 
     conversion.output_format = static_cast<OutputFormat>(cmd_buff[1]);
+
+    cmd_buff[0] = IPC::MakeHeader(0x3, 1, 0);
+    cmd_buff[1] = RESULT_SUCCESS.raw;
+
     LOG_DEBUG(Service_Y2R, "called output_format=%hhu", conversion.output_format);
+}
+
+static void GetOutputFormat(Service::Interface* self) {
+    u32* cmd_buff = Kernel::GetCommandBuffer();
 
+    cmd_buff[0] = IPC::MakeHeader(0x4, 2, 0);
     cmd_buff[1] = RESULT_SUCCESS.raw;
+    cmd_buff[2] = static_cast<u32>(conversion.output_format);
+
+    LOG_DEBUG(Service_Y2R, "called output_format=%hhu", conversion.output_format);
 }
 
 static void SetRotation(Service::Interface* self) {
     u32* cmd_buff = Kernel::GetCommandBuffer();
 
     conversion.rotation = static_cast<Rotation>(cmd_buff[1]);
+
+    cmd_buff[0] = IPC::MakeHeader(0x5, 1, 0);
+    cmd_buff[1] = RESULT_SUCCESS.raw;
+
     LOG_DEBUG(Service_Y2R, "called rotation=%hhu", conversion.rotation);
+}
+
+static void GetRotation(Service::Interface* self) {
+    u32* cmd_buff = Kernel::GetCommandBuffer();
 
+    cmd_buff[0] = IPC::MakeHeader(0x6, 2, 0);
     cmd_buff[1] = RESULT_SUCCESS.raw;
+    cmd_buff[2] = static_cast<u32>(conversion.rotation);
+
+    LOG_DEBUG(Service_Y2R, "called rotation=%hhu", conversion.rotation);
 }
 
 static void SetBlockAlignment(Service::Interface* self) {
     u32* cmd_buff = Kernel::GetCommandBuffer();
 
     conversion.block_alignment = static_cast<BlockAlignment>(cmd_buff[1]);
-    LOG_DEBUG(Service_Y2R, "called alignment=%hhu", conversion.block_alignment);
 
+    cmd_buff[0] = IPC::MakeHeader(0x7, 1, 0);
+    cmd_buff[1] = RESULT_SUCCESS.raw;
+
+    LOG_DEBUG(Service_Y2R, "called block_alignment=%hhu", conversion.block_alignment);
+}
+
+static void GetBlockAlignment(Service::Interface* self) {
+    u32* cmd_buff = Kernel::GetCommandBuffer();
+
+    cmd_buff[0] = IPC::MakeHeader(0x8, 2, 0);
+    cmd_buff[1] = RESULT_SUCCESS.raw;
+    cmd_buff[2] = static_cast<u32>(conversion.block_alignment);
+
+    LOG_DEBUG(Service_Y2R, "called block_alignment=%hhu", conversion.block_alignment);
+}
+
+/**
+ * Y2R_U::SetSpacialDithering service function
+ *  Inputs:
+ *      1 : u8, 0 = Disabled, 1 = Enabled
+ *  Outputs:
+ *      1 : Result of function, 0 on success, otherwise error code
+ */
+static void SetSpacialDithering(Service::Interface* self) {
+    u32* cmd_buff = Kernel::GetCommandBuffer();
+    spacial_dithering_enabled = cmd_buff[1] & 0xF;
+
+    cmd_buff[0] = IPC::MakeHeader(0x9, 1, 0);
+    cmd_buff[1] = RESULT_SUCCESS.raw;
+
+    LOG_WARNING(Service_Y2R, "(STUBBED) called");
+}
+
+/**
+ * Y2R_U::GetSpacialDithering service function
+ *  Outputs:
+ *      1 : Result of function, 0 on success, otherwise error code
+ *      2 : u8, 0 = Disabled, 1 = Enabled
+ */
+static void GetSpacialDithering(Service::Interface* self) {
+    u32* cmd_buff = Kernel::GetCommandBuffer();
+
+    cmd_buff[0] = IPC::MakeHeader(0xA, 2, 0);
+    cmd_buff[1] = RESULT_SUCCESS.raw;
+    cmd_buff[2] = spacial_dithering_enabled;
+
+    LOG_WARNING(Service_Y2R, "(STUBBED) called");
+}
+
+/**
+ * Y2R_U::SetTemporalDithering service function
+ *  Inputs:
+ *      1 : u8, 0 = Disabled, 1 = Enabled
+ *  Outputs:
+ *      1 : Result of function, 0 on success, otherwise error code
+ */
+static void SetTemporalDithering(Service::Interface* self) {
+    u32* cmd_buff = Kernel::GetCommandBuffer();
+    temporal_dithering_enabled = cmd_buff[1] & 0xF;
+
+    cmd_buff[0] = IPC::MakeHeader(0xB, 1, 0);
+    cmd_buff[1] = RESULT_SUCCESS.raw;
+
+    LOG_WARNING(Service_Y2R, "(STUBBED) called");
+}
+
+/**
+ * Y2R_U::GetTemporalDithering service function
+ *  Outputs:
+ *      1 : Result of function, 0 on success, otherwise error code
+ *      2 : u8, 0 = Disabled, 1 = Enabled
+ */
+static void GetTemporalDithering(Service::Interface* self) {
+    u32* cmd_buff = Kernel::GetCommandBuffer();
+
+    cmd_buff[0] = IPC::MakeHeader(0xC, 2, 0);
     cmd_buff[1] = RESULT_SUCCESS.raw;
+    cmd_buff[2] = temporal_dithering_enabled;
+
+    LOG_WARNING(Service_Y2R, "(STUBBED) called");
 }
 
+/**
+ * Y2R_U::SetTransferEndInterrupt service function
+ *  Inputs:
+ *      1 : u8, 0 = Disabled, 1 = Enabled
+ *  Outputs:
+ *      1 : Result of function, 0 on success, otherwise error code
+ */
 static void SetTransferEndInterrupt(Service::Interface* self) {
     u32* cmd_buff = Kernel::GetCommandBuffer();
+    transfer_end_interrupt_enabled = cmd_buff[1] & 0xf;
 
     cmd_buff[0] = IPC::MakeHeader(0xD, 1, 0);
     cmd_buff[1] = RESULT_SUCCESS.raw;
-    LOG_DEBUG(Service_Y2R, "(STUBBED) called");
+
+    LOG_WARNING(Service_Y2R, "(STUBBED) called");
+}
+
+/**
+ * Y2R_U::GetTransferEndInterrupt service function
+ *  Outputs:
+ *      1 : Result of function, 0 on success, otherwise error code
+ *      2 : u8, 0 = Disabled, 1 = Enabled
+ */
+static void GetTransferEndInterrupt(Service::Interface* self) {
+    u32* cmd_buff = Kernel::GetCommandBuffer();
+
+    cmd_buff[0] = IPC::MakeHeader(0xE, 2, 0);
+    cmd_buff[1] = RESULT_SUCCESS.raw;
+    cmd_buff[2] = transfer_end_interrupt_enabled;
+
+    LOG_WARNING(Service_Y2R, "(STUBBED) called");
 }
 
 /**
@@ -132,8 +276,10 @@ static void SetTransferEndInterrupt(Service::Interface* self) {
 static void GetTransferEndEvent(Service::Interface* self) {
     u32* cmd_buff = Kernel::GetCommandBuffer();
 
+    cmd_buff[0] = IPC::MakeHeader(0xF, 2, 0);
     cmd_buff[1] = RESULT_SUCCESS.raw;
     cmd_buff[3] = Kernel::g_handle_table.Create(completion_event).MoveFrom();
+
     LOG_DEBUG(Service_Y2R, "called");
 }
 
@@ -144,12 +290,12 @@ static void SetSendingY(Service::Interface* self) {
     conversion.src_Y.image_size = cmd_buff[2];
     conversion.src_Y.transfer_unit = cmd_buff[3];
     conversion.src_Y.gap = cmd_buff[4];
-    u32 src_process_handle = cmd_buff[6];
-    LOG_DEBUG(Service_Y2R, "called image_size=0x%08X, transfer_unit=%hu, transfer_stride=%hu, "
-        "src_process_handle=0x%08X", conversion.src_Y.image_size,
-        conversion.src_Y.transfer_unit, conversion.src_Y.gap, src_process_handle);
 
+    cmd_buff[0] = IPC::MakeHeader(0x10, 1, 0);
     cmd_buff[1] = RESULT_SUCCESS.raw;
+
+    LOG_DEBUG(Service_Y2R, "called image_size=0x%08X, transfer_unit=%hu, transfer_stride=%hu, src_process_handle=0x%08X",
+            conversion.src_Y.image_size, conversion.src_Y.transfer_unit, conversion.src_Y.gap, cmd_buff[6]);
 }
 
 static void SetSendingU(Service::Interface* self) {
@@ -159,12 +305,12 @@ static void SetSendingU(Service::Interface* self) {
     conversion.src_U.image_size = cmd_buff[2];
     conversion.src_U.transfer_unit = cmd_buff[3];
     conversion.src_U.gap = cmd_buff[4];
-    u32 src_process_handle = cmd_buff[6];
-    LOG_DEBUG(Service_Y2R, "called image_size=0x%08X, transfer_unit=%hu, transfer_stride=%hu, "
-        "src_process_handle=0x%08X", conversion.src_U.image_size,
-        conversion.src_U.transfer_unit, conversion.src_U.gap, src_process_handle);
 
+    cmd_buff[0] = IPC::MakeHeader(0x11, 1, 0);
     cmd_buff[1] = RESULT_SUCCESS.raw;
+
+    LOG_DEBUG(Service_Y2R, "called image_size=0x%08X, transfer_unit=%hu, transfer_stride=%hu, src_process_handle=0x%08X",
+            conversion.src_U.image_size, conversion.src_U.transfer_unit, conversion.src_U.gap, cmd_buff[6]);
 }
 
 static void SetSendingV(Service::Interface* self) {
@@ -174,12 +320,12 @@ static void SetSendingV(Service::Interface* self) {
     conversion.src_V.image_size = cmd_buff[2];
     conversion.src_V.transfer_unit = cmd_buff[3];
     conversion.src_V.gap = cmd_buff[4];
-    u32 src_process_handle = cmd_buff[6];
-    LOG_DEBUG(Service_Y2R, "called image_size=0x%08X, transfer_unit=%hu, transfer_stride=%hu, "
-        "src_process_handle=0x%08X", conversion.src_V.image_size,
-        conversion.src_V.transfer_unit, conversion.src_V.gap, src_process_handle);
 
+    cmd_buff[0] = IPC::MakeHeader(0x12, 1, 0);
     cmd_buff[1] = RESULT_SUCCESS.raw;
+
+    LOG_DEBUG(Service_Y2R, "called image_size=0x%08X, transfer_unit=%hu, transfer_stride=%hu, src_process_handle=0x%08X",
+            conversion.src_V.image_size, conversion.src_V.transfer_unit, conversion.src_V.gap, cmd_buff[6]);
 }
 
 static void SetSendingYUYV(Service::Interface* self) {
@@ -189,12 +335,76 @@ static void SetSendingYUYV(Service::Interface* self) {
     conversion.src_YUYV.image_size = cmd_buff[2];
     conversion.src_YUYV.transfer_unit = cmd_buff[3];
     conversion.src_YUYV.gap = cmd_buff[4];
-    u32 src_process_handle = cmd_buff[6];
-    LOG_DEBUG(Service_Y2R, "called image_size=0x%08X, transfer_unit=%hu, transfer_stride=%hu, "
-        "src_process_handle=0x%08X", conversion.src_YUYV.image_size,
-        conversion.src_YUYV.transfer_unit, conversion.src_YUYV.gap, src_process_handle);
 
+    cmd_buff[0] = IPC::MakeHeader(0x13, 1, 0);
+    cmd_buff[1] = RESULT_SUCCESS.raw;
+
+    LOG_DEBUG(Service_Y2R, "called image_size=0x%08X, transfer_unit=%hu, transfer_stride=%hu, src_process_handle=0x%08X",
+            conversion.src_YUYV.image_size, conversion.src_YUYV.transfer_unit, conversion.src_YUYV.gap, cmd_buff[6]);
+}
+
+/**
+ * Y2R::IsFinishedSendingYuv service function
+ * Output:
+ *       1 : Result of the function, 0 on success, otherwise error code
+ *       2 : u8, 0 = Not Finished, 1 = Finished
+ */
+static void IsFinishedSendingYuv(Service::Interface* self) {
+    u32* cmd_buff = Kernel::GetCommandBuffer();
+
+    cmd_buff[0] = IPC::MakeHeader(0x14, 2, 0);
     cmd_buff[1] = RESULT_SUCCESS.raw;
+    cmd_buff[2] = 1;
+
+    LOG_WARNING(Service_Y2R, "(STUBBED) called");
+}
+
+/**
+ * Y2R::IsFinishedSendingY service function
+ * Output:
+ *       1 : Result of the function, 0 on success, otherwise error code
+ *       2 : u8, 0 = Not Finished, 1 = Finished
+ */
+static void IsFinishedSendingY(Service::Interface* self) {
+    u32* cmd_buff = Kernel::GetCommandBuffer();
+
+    cmd_buff[0] = IPC::MakeHeader(0x15, 2, 0);
+    cmd_buff[1] = RESULT_SUCCESS.raw;
+    cmd_buff[2] = 1;
+
+    LOG_WARNING(Service_Y2R, "(STUBBED) called");
+}
+
+/**
+ * Y2R::IsFinishedSendingU service function
+ * Output:
+ *       1 : Result of the function, 0 on success, otherwise error code
+ *       2 : u8, 0 = Not Finished, 1 = Finished
+ */
+static void IsFinishedSendingU(Service::Interface* self) {
+    u32* cmd_buff = Kernel::GetCommandBuffer();
+
+    cmd_buff[0] = IPC::MakeHeader(0x16, 2, 0);
+    cmd_buff[1] = RESULT_SUCCESS.raw;
+    cmd_buff[2] = 1;
+
+    LOG_WARNING(Service_Y2R, "(STUBBED) called");
+}
+
+/**
+ * Y2R::IsFinishedSendingV service function
+ * Output:
+ *       1 : Result of the function, 0 on success, otherwise error code
+ *       2 : u8, 0 = Not Finished, 1 = Finished
+ */
+static void IsFinishedSendingV(Service::Interface* self) {
+    u32* cmd_buff = Kernel::GetCommandBuffer();
+
+    cmd_buff[0] = IPC::MakeHeader(0x17, 2, 0);
+    cmd_buff[1] = RESULT_SUCCESS.raw;
+    cmd_buff[2] = 1;
+
+    LOG_WARNING(Service_Y2R, "(STUBBED) called");
 }
 
 static void SetReceiving(Service::Interface* self) {
@@ -204,27 +414,66 @@ static void SetReceiving(Service::Interface* self) {
     conversion.dst.image_size = cmd_buff[2];
     conversion.dst.transfer_unit = cmd_buff[3];
     conversion.dst.gap = cmd_buff[4];
-    u32 dst_process_handle = cmd_buff[6];
-    LOG_DEBUG(Service_Y2R, "called image_size=0x%08X, transfer_unit=%hu, transfer_stride=%hu, "
-        "dst_process_handle=0x%08X", conversion.dst.image_size,
-        conversion.dst.transfer_unit, conversion.dst.gap,
-        dst_process_handle);
 
+    cmd_buff[0] = IPC::MakeHeader(0x18, 1, 0);
+    cmd_buff[1] = RESULT_SUCCESS.raw;
+
+    LOG_DEBUG(Service_Y2R, "called image_size=0x%08X, transfer_unit=%hu, transfer_stride=%hu, dst_process_handle=0x%08X",
+            conversion.dst.image_size, conversion.dst.transfer_unit, conversion.dst.gap, cmd_buff[6]);
+}
+
+/**
+ * Y2R::IsFinishedReceiving service function
+ * Output:
+ *       1 : Result of the function, 0 on success, otherwise error code
+ *       2 : u8, 0 = Not Finished, 1 = Finished
+ */
+static void IsFinishedReceiving(Service::Interface* self) {
+    u32* cmd_buff = Kernel::GetCommandBuffer();
+
+    cmd_buff[0] = IPC::MakeHeader(0x19, 2, 0);
     cmd_buff[1] = RESULT_SUCCESS.raw;
+    cmd_buff[2] = 1;
+
+    LOG_WARNING(Service_Y2R, "(STUBBED) called");
 }
 
 static void SetInputLineWidth(Service::Interface* self) {
     u32* cmd_buff = Kernel::GetCommandBuffer();
 
-    LOG_DEBUG(Service_Y2R, "called input_line_width=%u", cmd_buff[1]);
+    cmd_buff[0] = IPC::MakeHeader(0x1A, 1, 0);
     cmd_buff[1] = conversion.SetInputLineWidth(cmd_buff[1]).raw;
+
+    LOG_DEBUG(Service_Y2R, "called input_line_width=%u", cmd_buff[1]);
+}
+
+static void GetInputLineWidth(Service::Interface* self) {
+    u32* cmd_buff = Kernel::GetCommandBuffer();
+
+    cmd_buff[0] = IPC::MakeHeader(0x1B, 2, 0);
+    cmd_buff[1] = RESULT_SUCCESS.raw;
+    cmd_buff[2] = conversion.input_line_width;
+
+    LOG_DEBUG(Service_Y2R, "called input_line_width=%u", conversion.input_line_width);
 }
 
 static void SetInputLines(Service::Interface* self) {
     u32* cmd_buff = Kernel::GetCommandBuffer();
 
-    LOG_DEBUG(Service_Y2R, "called input_line_number=%u", cmd_buff[1]);
+    cmd_buff[0] = IPC::MakeHeader(0x1C, 1, 0);
     cmd_buff[1] = conversion.SetInputLines(cmd_buff[1]).raw;
+
+    LOG_DEBUG(Service_Y2R, "called input_lines=%u", cmd_buff[1]);
+}
+
+static void GetInputLines(Service::Interface* self) {
+    u32* cmd_buff = Kernel::GetCommandBuffer();
+
+    cmd_buff[0] = IPC::MakeHeader(0x1D, 2, 0);
+    cmd_buff[1] = RESULT_SUCCESS.raw;
+    cmd_buff[2] = static_cast<u32>(conversion.input_lines);
+
+    LOG_DEBUG(Service_Y2R, "called input_lines=%u", conversion.input_lines);
 }
 
 static void SetCoefficient(Service::Interface* self) {
@@ -232,44 +481,111 @@ static void SetCoefficient(Service::Interface* self) {
 
     const u16* coefficients = reinterpret_cast<const u16*>(&cmd_buff[1]);
     std::memcpy(conversion.coefficients.data(), coefficients, sizeof(CoefficientSet));
+
+    cmd_buff[0] = IPC::MakeHeader(0x1E, 1, 0);
+    cmd_buff[1] = RESULT_SUCCESS.raw;
+
     LOG_DEBUG(Service_Y2R, "called coefficients=[%hX, %hX, %hX, %hX, %hX, %hX, %hX, %hX]",
             coefficients[0], coefficients[1], coefficients[2], coefficients[3],
             coefficients[4], coefficients[5], coefficients[6], coefficients[7]);
+}
+
+static void GetCoefficient(Service::Interface* self) {
+    u32* cmd_buff = Kernel::GetCommandBuffer();
 
+    cmd_buff[0] = IPC::MakeHeader(0x1F, 5, 0);
     cmd_buff[1] = RESULT_SUCCESS.raw;
+    std::memcpy(&cmd_buff[2], conversion.coefficients.data(), sizeof(CoefficientSet));
+
+    LOG_DEBUG(Service_Y2R, "called");
 }
 
 static void SetStandardCoefficient(Service::Interface* self) {
     u32* cmd_buff = Kernel::GetCommandBuffer();
 
-    LOG_DEBUG(Service_Y2R, "called standard_coefficient=%u", cmd_buff[1]);
+    u32 index = cmd_buff[1];
+
+    cmd_buff[0] = IPC::MakeHeader(0x20, 1, 0);
+    cmd_buff[1] = conversion.SetStandardCoefficient((StandardCoefficient)index).raw;
 
-    cmd_buff[1] = conversion.SetStandardCoefficient((StandardCoefficient)cmd_buff[1]).raw;
+    LOG_DEBUG(Service_Y2R, "called standard_coefficient=%u", index);
+}
+
+static void GetStandardCoefficient(Service::Interface* self) {
+    u32* cmd_buff = Kernel::GetCommandBuffer();
+
+    u32 index = cmd_buff[1];
+
+    if (index < ARRAY_SIZE(standard_coefficients)) {
+        cmd_buff[0] = IPC::MakeHeader(0x21, 5, 0);
+        cmd_buff[1] = RESULT_SUCCESS.raw;
+        std::memcpy(&cmd_buff[2], &standard_coefficients[index], sizeof(CoefficientSet));
+
+        LOG_DEBUG(Service_Y2R, "called standard_coefficient=%u ", index);
+    } else {
+        cmd_buff[0] = IPC::MakeHeader(0x21, 1, 0);
+        cmd_buff[1] = -1; // TODO(bunnei): Identify the correct error code for this
+
+        LOG_ERROR(Service_Y2R, "called standard_coefficient=%u  The argument is invalid!", index);
+    }
 }
 
 static void SetAlpha(Service::Interface* self) {
     u32* cmd_buff = Kernel::GetCommandBuffer();
 
     conversion.alpha = cmd_buff[1];
+
+    cmd_buff[0] = IPC::MakeHeader(0x22, 1, 0);
+    cmd_buff[1] = RESULT_SUCCESS.raw;
+
+    LOG_DEBUG(Service_Y2R, "called alpha=%hu", conversion.alpha);
+}
+
+static void GetAlpha(Service::Interface* self) {
+    u32* cmd_buff = Kernel::GetCommandBuffer();
+
+    cmd_buff[0] = IPC::MakeHeader(0x23, 2, 0);
+    cmd_buff[1] = RESULT_SUCCESS.raw;
+    cmd_buff[2] = conversion.alpha;
+
     LOG_DEBUG(Service_Y2R, "called alpha=%hu", conversion.alpha);
+}
+
+static void SetDitheringWeightParams(Service::Interface* self) {
+    u32* cmd_buff = Kernel::GetCommandBuffer();
+    std::memcpy(&dithering_weight_params, &cmd_buff[1], sizeof(DitheringWeightParams));
 
+    cmd_buff[0] = IPC::MakeHeader(0x24, 1, 0);
     cmd_buff[1] = RESULT_SUCCESS.raw;
+
+    LOG_DEBUG(Service_Y2R, "called");
+}
+
+static void GetDitheringWeightParams(Service::Interface* self) {
+    u32* cmd_buff = Kernel::GetCommandBuffer();
+
+    cmd_buff[0] = IPC::MakeHeader(0x25, 9, 0);
+    cmd_buff[1] = RESULT_SUCCESS.raw;
+    std::memcpy(&cmd_buff[2], &dithering_weight_params, sizeof(DitheringWeightParams));
+
+    LOG_DEBUG(Service_Y2R, "called");
 }
 
 static void StartConversion(Service::Interface* self) {
     u32* cmd_buff = Kernel::GetCommandBuffer();
 
     // dst_image_size would seem to be perfect for this, but it doesn't include the gap :(
-    u32 total_output_size = conversion.input_lines *
-        (conversion.dst.transfer_unit + conversion.dst.gap);
+    u32 total_output_size = conversion.input_lines * (conversion.dst.transfer_unit + conversion.dst.gap);
     Memory::RasterizerFlushAndInvalidateRegion(Memory::VirtualToPhysicalAddress(conversion.dst.address), total_output_size);
 
     HW::Y2R::PerformConversion(conversion);
 
-    LOG_DEBUG(Service_Y2R, "called");
     completion_event->Signal();
 
+    cmd_buff[0] = IPC::MakeHeader(0x26, 1, 0);
     cmd_buff[1] = RESULT_SUCCESS.raw;
+
+    LOG_DEBUG(Service_Y2R, "called");
 }
 
 static void StopConversion(Service::Interface* self) {
@@ -277,6 +593,7 @@ static void StopConversion(Service::Interface* self) {
 
     cmd_buff[0] = IPC::MakeHeader(0x27, 1, 0);
     cmd_buff[1] = RESULT_SUCCESS.raw;
+
     LOG_DEBUG(Service_Y2R, "called");
 }
 
@@ -289,50 +606,61 @@ static void StopConversion(Service::Interface* self) {
 static void IsBusyConversion(Service::Interface* self) {
     u32* cmd_buff = Kernel::GetCommandBuffer();
 
+    cmd_buff[0] = IPC::MakeHeader(0x28, 2, 0);
     cmd_buff[1] = RESULT_SUCCESS.raw;
     cmd_buff[2] = 0; // StartConversion always finishes immediately
+
     LOG_DEBUG(Service_Y2R, "called");
 }
 
 /**
- * Y2R_U::SetConversionParams service function
+ * Y2R_U::SetPackageParameter service function
  */
-static void SetConversionParams(Service::Interface* self) {
+static void SetPackageParameter(Service::Interface* self) {
     u32* cmd_buff = Kernel::GetCommandBuffer();
 
     auto params = reinterpret_cast<const ConversionParameters*>(&cmd_buff[1]);
-    LOG_DEBUG(Service_Y2R,
-        "called input_format=%hhu output_format=%hhu rotation=%hhu block_alignment=%hhu "
-        "input_line_width=%hu input_lines=%hu standard_coefficient=%hhu "
-        "reserved=%hhu alpha=%hX",
-        params->input_format, params->output_format, params->rotation, params->block_alignment,
-        params->input_line_width, params->input_lines, params->standard_coefficient,
-        params->reserved, params->alpha);
-
-    ResultCode result = RESULT_SUCCESS;
 
     conversion.input_format = params->input_format;
     conversion.output_format = params->output_format;
     conversion.rotation = params->rotation;
     conversion.block_alignment = params->block_alignment;
-    result = conversion.SetInputLineWidth(params->input_line_width);
-    if (result.IsError()) goto cleanup;
+
+    ResultCode result = conversion.SetInputLineWidth(params->input_line_width);
+
+    if (result.IsError())
+        goto cleanup;
+
     result = conversion.SetInputLines(params->input_lines);
-    if (result.IsError()) goto cleanup;
+
+    if (result.IsError())
+        goto cleanup;
+
     result = conversion.SetStandardCoefficient(params->standard_coefficient);
-    if (result.IsError()) goto cleanup;
+
+    if (result.IsError())
+        goto cleanup;
+
+    conversion.padding = params->padding;
     conversion.alpha = params->alpha;
 
 cleanup:
     cmd_buff[0] = IPC::MakeHeader(0x29, 1, 0);
     cmd_buff[1] = result.raw;
+
+    LOG_DEBUG(Service_Y2R, "called input_format=%hhu output_format=%hhu rotation=%hhu block_alignment=%hhu "
+            "input_line_width=%hu input_lines=%hu standard_coefficient=%hhu reserved=%hhu alpha=%hX",
+            params->input_format, params->output_format, params->rotation, params->block_alignment,
+            params->input_line_width, params->input_lines, params->standard_coefficient, params->padding, params->alpha);
 }
 
 static void PingProcess(Service::Interface* self) {
     u32* cmd_buff = Kernel::GetCommandBuffer();
 
+    cmd_buff[0] = IPC::MakeHeader(0x2A, 2, 0);
     cmd_buff[1] = RESULT_SUCCESS.raw;
     cmd_buff[2] = 0;
+
     LOG_WARNING(Service_Y2R, "(STUBBED) called");
 }
 
@@ -358,6 +686,7 @@ static void DriverInitialize(Service::Interface* self) {
 
     cmd_buff[0] = IPC::MakeHeader(0x2B, 1, 0);
     cmd_buff[1] = RESULT_SUCCESS.raw;
+
     LOG_DEBUG(Service_Y2R, "called");
 }
 
@@ -366,54 +695,67 @@ static void DriverFinalize(Service::Interface* self) {
 
     cmd_buff[0] = IPC::MakeHeader(0x2C, 1, 0);
     cmd_buff[1] = RESULT_SUCCESS.raw;
+
+    LOG_DEBUG(Service_Y2R, "called");
+}
+
+
+static void GetPackageParameter(Service::Interface* self) {
+    u32* cmd_buff = Kernel::GetCommandBuffer();
+
+    cmd_buff[0] = IPC::MakeHeader(0x2D, 4, 0);
+    cmd_buff[1] = RESULT_SUCCESS.raw;
+    std::memcpy(&cmd_buff[2], &conversion, sizeof(ConversionParameters));
+
     LOG_DEBUG(Service_Y2R, "called");
 }
 
 const Interface::FunctionInfo FunctionTable[] = {
     {0x00010040, SetInputFormat,          "SetInputFormat"},
-    {0x00020000, nullptr,                 "GetInputFormat"},
+    {0x00020000, GetInputFormat,          "GetInputFormat"},
     {0x00030040, SetOutputFormat,         "SetOutputFormat"},
-    {0x00040000, nullptr,                 "GetOutputFormat"},
+    {0x00040000, GetOutputFormat,         "GetOutputFormat"},
     {0x00050040, SetRotation,             "SetRotation"},
-    {0x00060000, nullptr,                 "GetRotation"},
+    {0x00060000, GetRotation,             "GetRotation"},
     {0x00070040, SetBlockAlignment,       "SetBlockAlignment"},
-    {0x00080000, nullptr,                 "GetBlockAlignment"},
-    {0x00090040, nullptr,                 "SetSpacialDithering"},
-    {0x000A0000, nullptr,                 "GetSpacialDithering"},
-    {0x000B0040, nullptr,                 "SetTemporalDithering"},
-    {0x000C0000, nullptr,                 "GetTemporalDithering"},
+    {0x00080000, GetBlockAlignment,       "GetBlockAlignment"},
+    {0x00090040, SetSpacialDithering,     "SetSpacialDithering"},
+    {0x000A0000, GetSpacialDithering,     "GetSpacialDithering"},
+    {0x000B0040, SetTemporalDithering,    "SetTemporalDithering"},
+    {0x000C0000, GetTemporalDithering,    "GetTemporalDithering"},
     {0x000D0040, SetTransferEndInterrupt, "SetTransferEndInterrupt"},
+    {0x000E0000, GetTransferEndInterrupt, "GetTransferEndInterrupt"},
     {0x000F0000, GetTransferEndEvent,     "GetTransferEndEvent"},
     {0x00100102, SetSendingY,             "SetSendingY"},
     {0x00110102, SetSendingU,             "SetSendingU"},
     {0x00120102, SetSendingV,             "SetSendingV"},
     {0x00130102, SetSendingYUYV,          "SetSendingYUYV"},
-    {0x00140000, nullptr,                 "IsFinishedSendingYuv"},
-    {0x00150000, nullptr,                 "IsFinishedSendingY"},
-    {0x00160000, nullptr,                 "IsFinishedSendingU"},
-    {0x00170000, nullptr,                 "IsFinishedSendingV"},
+    {0x00140000, IsFinishedSendingYuv,    "IsFinishedSendingYuv"},
+    {0x00150000, IsFinishedSendingY,      "IsFinishedSendingY"},
+    {0x00160000, IsFinishedSendingU,      "IsFinishedSendingU"},
+    {0x00170000, IsFinishedSendingV,      "IsFinishedSendingV"},
     {0x00180102, SetReceiving,            "SetReceiving"},
-    {0x00190000, nullptr,                 "IsFinishedReceiving"},
+    {0x00190000, IsFinishedReceiving,     "IsFinishedReceiving"},
     {0x001A0040, SetInputLineWidth,       "SetInputLineWidth"},
-    {0x001B0000, nullptr,                 "GetInputLineWidth"},
+    {0x001B0000, GetInputLineWidth,       "GetInputLineWidth"},
     {0x001C0040, SetInputLines,           "SetInputLines"},
-    {0x001D0000, nullptr,                 "GetInputLines"},
+    {0x001D0000, GetInputLines,           "GetInputLines"},
     {0x001E0100, SetCoefficient,          "SetCoefficient"},
-    {0x001F0000, nullptr,                 "GetCoefficient"},
+    {0x001F0000, GetCoefficient,          "GetCoefficient"},
     {0x00200040, SetStandardCoefficient,  "SetStandardCoefficient"},
-    {0x00210040, nullptr,                 "GetStandardCoefficientParams"},
+    {0x00210040, GetStandardCoefficient,  "GetStandardCoefficient"},
     {0x00220040, SetAlpha,                "SetAlpha"},
-    {0x00230000, nullptr,                 "GetAlpha"},
-    {0x00240200, nullptr,                 "SetDitheringWeightParams"},
-    {0x00250000, nullptr,                 "GetDitheringWeightParams"},
+    {0x00230000, GetAlpha,                "GetAlpha"},
+    {0x00240200, SetDitheringWeightParams,"SetDitheringWeightParams"},
+    {0x00250000, GetDitheringWeightParams,"GetDitheringWeightParams"},
     {0x00260000, StartConversion,         "StartConversion"},
     {0x00270000, StopConversion,          "StopConversion"},
     {0x00280000, IsBusyConversion,        "IsBusyConversion"},
-    {0x002901C0, SetConversionParams,     "SetConversionParams"},
+    {0x002901C0, SetPackageParameter,     "SetPackageParameter"},
     {0x002A0000, PingProcess,             "PingProcess"},
     {0x002B0000, DriverInitialize,        "DriverInitialize"},
     {0x002C0000, DriverFinalize,          "DriverFinalize"},
-    {0x002D0000, nullptr,                 "GetPackageParameter"},
+    {0x002D0000, GetPackageParameter,     "GetPackageParameter"},
 };
 
 ////////////////////////////////////////////////////////////////////////////////////////////////////
diff --git a/src/core/hle/service/y2r_u.h b/src/core/hle/service/y2r_u.h
index 3965a5545..95fa2fdb7 100644
--- a/src/core/hle/service/y2r_u.h
+++ b/src/core/hle/service/y2r_u.h
@@ -97,6 +97,7 @@ struct ConversionConfiguration {
     u16 input_line_width;
     u16 input_lines;
     CoefficientSet coefficients;
+    u8 padding;
     u16 alpha;
 
     /// Input parameters for the Y (luma) plane
@@ -109,6 +110,25 @@ struct ConversionConfiguration {
     ResultCode SetStandardCoefficient(StandardCoefficient standard_coefficient);
 };
 
+struct DitheringWeightParams {
+    u16 w0_xEven_yEven;
+    u16 w0_xOdd_yEven;
+    u16 w0_xEven_yOdd;
+    u16 w0_xOdd_yOdd;
+    u16 w1_xEven_yEven;
+    u16 w1_xOdd_yEven;
+    u16 w1_xEven_yOdd;
+    u16 w1_xOdd_yOdd;
+    u16 w2_xEven_yEven;
+    u16 w2_xOdd_yEven;
+    u16 w2_xEven_yOdd;
+    u16 w2_xOdd_yOdd;
+    u16 w3_xEven_yEven;
+    u16 w3_xOdd_yEven;
+    u16 w3_xEven_yOdd;
+    u16 w3_xOdd_yOdd;
+};
+
 class Interface : public Service::Interface {
 public:
     Interface();
diff --git a/src/core/hle/svc.cpp b/src/core/hle/svc.cpp
index ae54afb1c..a9a1a3244 100644
--- a/src/core/hle/svc.cpp
+++ b/src/core/hle/svc.cpp
@@ -6,7 +6,6 @@
 
 #include "common/logging/log.h"
 #include "common/microprofile.h"
-#include "common/profiler.h"
 #include "common/string_util.h"
 #include "common/symbols.h"
 
@@ -1031,8 +1030,6 @@ static const FunctionDef SVC_Table[] = {
     {0x7D, HLE::Wrap<QueryProcessMemory>,   "QueryProcessMemory"},
 };
 
-Common::Profiling::TimingCategory profiler_svc("SVC Calls");
-
 static const FunctionDef* GetSVCInfo(u32 func_num) {
     if (func_num >= ARRAY_SIZE(SVC_Table)) {
         LOG_ERROR(Kernel_SVC, "unknown svc=0x%02X", func_num);
@@ -1044,7 +1041,6 @@ static const FunctionDef* GetSVCInfo(u32 func_num) {
 MICROPROFILE_DEFINE(Kernel_SVC, "Kernel", "SVC", MP_RGB(70, 200, 70));
 
 void CallSVC(u32 immediate) {
-    Common::Profiling::ScopeTimer timer_svc(profiler_svc);
     MICROPROFILE_SCOPE(Kernel_SVC);
 
     const FunctionDef* info = GetSVCInfo(immediate);
diff --git a/src/video_core/CMakeLists.txt b/src/video_core/CMakeLists.txt
index 76cfd4f7d..de4082b1f 100644
--- a/src/video_core/CMakeLists.txt
+++ b/src/video_core/CMakeLists.txt
@@ -16,6 +16,7 @@ set(SRCS
             shader/shader_interpreter.cpp
             swrasterizer.cpp
             utils.cpp
+            vertex_loader.cpp
             video_core.cpp
             )
 
@@ -43,6 +44,7 @@ set(HEADERS
             shader/shader_interpreter.h
             swrasterizer.h
             utils.h
+            vertex_loader.h
             video_core.h
             )
 
diff --git a/src/video_core/command_processor.cpp b/src/video_core/command_processor.cpp
index 3abe79c09..58883e374 100644
--- a/src/video_core/command_processor.cpp
+++ b/src/video_core/command_processor.cpp
@@ -7,7 +7,6 @@
 
 #include "common/alignment.h"
 #include "common/microprofile.h"
-#include "common/profiler.h"
 
 #include "core/settings.h"
 #include "core/hle/service/gsp_gpu.h"
@@ -22,6 +21,7 @@
 #include "video_core/video_core.h"
 #include "video_core/debug_utils/debug_utils.h"
 #include "video_core/shader/shader_interpreter.h"
+#include "video_core/vertex_loader.h"
 
 namespace Pica {
 
@@ -35,8 +35,6 @@ static int default_attr_counter = 0;
 
 static u32 default_attr_write_buffer[3];
 
-Common::Profiling::TimingCategory category_drawing("Drawing");
-
 // Expand a 4-bit mask to 4-byte mask, e.g. 0b0101 -> 0x00FF00FF
 static const u32 expand_bits_to_bytes[] = {
     0x00000000, 0x000000ff, 0x0000ff00, 0x0000ffff,
@@ -186,60 +184,19 @@ static void WritePicaReg(u32 id, u32 value, u32 mask) {
         case PICA_REG_INDEX(trigger_draw):
         case PICA_REG_INDEX(trigger_draw_indexed):
         {
-            Common::Profiling::ScopeTimer scope_timer(category_drawing);
             MICROPROFILE_SCOPE(GPU_Drawing);
 
 #if PICA_LOG_TEV
             DebugUtils::DumpTevStageConfig(regs.GetTevStages());
 #endif
-
             if (g_debug_context)
                 g_debug_context->OnEvent(DebugContext::Event::IncomingPrimitiveBatch, nullptr);
 
-            const auto& attribute_config = regs.vertex_attributes;
-            const u32 base_address = attribute_config.GetPhysicalBaseAddress();
-
-            // Information about internal vertex attributes
-            u32 vertex_attribute_sources[16];
-            boost::fill(vertex_attribute_sources, 0xdeadbeef);
-            u32 vertex_attribute_strides[16] = {};
-            Regs::VertexAttributeFormat vertex_attribute_formats[16] = {};
-
-            u32 vertex_attribute_elements[16] = {};
-            u32 vertex_attribute_element_size[16] = {};
-
-            // Setup attribute data from loaders
-            for (int loader = 0; loader < 12; ++loader) {
-                const auto& loader_config = attribute_config.attribute_loaders[loader];
-
-                u32 offset = 0;
-
-                // TODO: What happens if a loader overwrites a previous one's data?
-                for (unsigned component = 0; component < loader_config.component_count; ++component) {
-                    if (component >= 12) {
-                        LOG_ERROR(HW_GPU, "Overflow in the vertex attribute loader %u trying to load component %u", loader, component);
-                        continue;
-                    }
-
-                    u32 attribute_index = loader_config.GetComponent(component);
-                    if (attribute_index < 12) {
-                        int element_size = attribute_config.GetElementSizeInBytes(attribute_index);
-                        offset = Common::AlignUp(offset, element_size);
-                        vertex_attribute_sources[attribute_index] = base_address + loader_config.data_offset + offset;
-                        vertex_attribute_strides[attribute_index] = static_cast<u32>(loader_config.byte_count);
-                        vertex_attribute_formats[attribute_index] = attribute_config.GetFormat(attribute_index);
-                        vertex_attribute_elements[attribute_index] = attribute_config.GetNumElements(attribute_index);
-                        vertex_attribute_element_size[attribute_index] = element_size;
-                        offset += attribute_config.GetStride(attribute_index);
-                    } else if (attribute_index < 16) {
-                        // Attribute ids 12, 13, 14 and 15 signify 4, 8, 12 and 16-byte paddings, respectively
-                        offset = Common::AlignUp(offset, 4);
-                        offset += (attribute_index - 11) * 4;
-                    } else {
-                        UNREACHABLE(); // This is truly unreachable due to the number of bits for each component
-                    }
-                }
-            }
+            // Processes information about internal vertex attributes to figure out how a vertex is loaded.
+            // Later, these can be compiled and cached.
+            VertexLoader loader;
+            const u32 base_address = regs.vertex_attributes.GetPhysicalBaseAddress();
+            loader.Setup(regs);
 
             // Load vertices
             bool is_indexed = (id == PICA_REG_INDEX(trigger_draw_indexed));
@@ -263,32 +220,7 @@ static void WritePicaReg(u32 id, u32 value, u32 mask) {
                 }
             }
 
-            class {
-                /// Combine overlapping and close ranges
-                void SimplifyRanges() {
-                    for (auto it = ranges.begin(); it != ranges.end(); ++it) {
-                        // NOTE: We add 32 to the range end address to make sure "close" ranges are combined, too
-                        auto it2 = std::next(it);
-                        while (it2 != ranges.end() && it->first + it->second + 32 >= it2->first) {
-                            it->second = std::max(it->second, it2->first + it2->second - it->first);
-                            it2 = ranges.erase(it2);
-                        }
-                    }
-                }
-
-            public:
-                /// Record a particular memory access in the list
-                void AddAccess(u32 paddr, u32 size) {
-                    // Create new range or extend existing one
-                    ranges[paddr] = std::max(ranges[paddr], size);
-
-                    // Simplify ranges...
-                    SimplifyRanges();
-                }
-
-                /// Map of accessed ranges (mapping start address to range size)
-                std::map<u32, u32> ranges;
-            } memory_accesses;
+            DebugUtils::MemoryAccessTracker memory_accesses;
 
             // Simple circular-replacement vertex cache
             // The size has been tuned for optimal balance between hit-rate and the cost of lookup
@@ -332,60 +264,13 @@ static void WritePicaReg(u32 id, u32 value, u32 mask) {
                 if (!vertex_cache_hit) {
                     // Initialize data for the current vertex
                     Shader::InputVertex input;
-
-                    for (int i = 0; i < attribute_config.GetNumTotalAttributes(); ++i) {
-                        if (vertex_attribute_elements[i] != 0) {
-                            // Default attribute values set if array elements have < 4 components. This
-                            // is *not* carried over from the default attribute settings even if they're
-                            // enabled for this attribute.
-                            static const float24 zero = float24::FromFloat32(0.0f);
-                            static const float24 one = float24::FromFloat32(1.0f);
-                            input.attr[i] = Math::Vec4<float24>(zero, zero, zero, one);
-
-                            // Load per-vertex data from the loader arrays
-                            for (unsigned int comp = 0; comp < vertex_attribute_elements[i]; ++comp) {
-                                u32 source_addr = vertex_attribute_sources[i] + vertex_attribute_strides[i] * vertex + comp * vertex_attribute_element_size[i];
-                                const u8* srcdata = Memory::GetPhysicalPointer(source_addr);
-
-                                if (g_debug_context && Pica::g_debug_context->recorder) {
-                                    memory_accesses.AddAccess(source_addr,
-                                        (vertex_attribute_formats[i] == Regs::VertexAttributeFormat::FLOAT) ? 4
-                                        : (vertex_attribute_formats[i] == Regs::VertexAttributeFormat::SHORT) ? 2 : 1);
-                                }
-
-                                const float srcval =
-                                    (vertex_attribute_formats[i] == Regs::VertexAttributeFormat::BYTE)  ? *reinterpret_cast<const s8*>(srcdata) :
-                                    (vertex_attribute_formats[i] == Regs::VertexAttributeFormat::UBYTE) ? *reinterpret_cast<const u8*>(srcdata) :
-                                    (vertex_attribute_formats[i] == Regs::VertexAttributeFormat::SHORT) ? *reinterpret_cast<const s16*>(srcdata) :
-                                    *reinterpret_cast<const float*>(srcdata);
-
-                                input.attr[i][comp] = float24::FromFloat32(srcval);
-                                LOG_TRACE(HW_GPU, "Loaded component %x of attribute %x for vertex %x (index %x) from 0x%08x + 0x%08x + 0x%04x: %f",
-                                    comp, i, vertex, index,
-                                    attribute_config.GetPhysicalBaseAddress(),
-                                    vertex_attribute_sources[i] - base_address,
-                                    vertex_attribute_strides[i] * vertex + comp * vertex_attribute_element_size[i],
-                                    input.attr[i][comp].ToFloat32());
-                            }
-                        } else if (attribute_config.IsDefaultAttribute(i)) {
-                            // Load the default attribute if we're configured to do so
-                            input.attr[i] = g_state.vs.default_attributes[i];
-                            LOG_TRACE(HW_GPU, "Loaded default attribute %x for vertex %x (index %x): (%f, %f, %f, %f)",
-                                      i, vertex, index,
-                                      input.attr[i][0].ToFloat32(), input.attr[i][1].ToFloat32(),
-                                      input.attr[i][2].ToFloat32(), input.attr[i][3].ToFloat32());
-                        } else {
-                            // TODO(yuriks): In this case, no data gets loaded and the vertex
-                            // remains with the last value it had. This isn't currently maintained
-                            // as global state, however, and so won't work in Citra yet.
-                        }
-                    }
+                    loader.LoadVertex(base_address, index, vertex, input, memory_accesses);
 
                     if (g_debug_context)
                         g_debug_context->OnEvent(DebugContext::Event::VertexLoaded, (void*)&input);
 
                     // Send to vertex shader
-                    output = Shader::Run(shader_unit, input, attribute_config.GetNumTotalAttributes());
+                    output = Shader::Run(shader_unit, input, loader.GetNumTotalAttributes());
 
                     if (is_indexed) {
                         vertex_cache[vertex_cache_pos] = output;
diff --git a/src/video_core/debug_utils/debug_utils.h b/src/video_core/debug_utils/debug_utils.h
index 56f9bd958..dd0828cee 100644
--- a/src/video_core/debug_utils/debug_utils.h
+++ b/src/video_core/debug_utils/debug_utils.h
@@ -216,6 +216,36 @@ void DumpTexture(const Pica::Regs::TextureConfig& texture_config, u8* data);
 
 void DumpTevStageConfig(const std::array<Pica::Regs::TevStageConfig,6>& stages);
 
+/**
+ * Used in the vertex loader to merge access records. TODO: Investigate if actually useful.
+ */
+class MemoryAccessTracker {
+    /// Combine overlapping and close ranges
+    void SimplifyRanges() {
+        for (auto it = ranges.begin(); it != ranges.end(); ++it) {
+            // NOTE: We add 32 to the range end address to make sure "close" ranges are combined, too
+            auto it2 = std::next(it);
+            while (it2 != ranges.end() && it->first + it->second + 32 >= it2->first) {
+                it->second = std::max(it->second, it2->first + it2->second - it->first);
+                it2 = ranges.erase(it2);
+            }
+        }
+    }
+
+public:
+    /// Record a particular memory access in the list
+    void AddAccess(u32 paddr, u32 size) {
+        // Create new range or extend existing one
+        ranges[paddr] = std::max(ranges[paddr], size);
+
+        // Simplify ranges...
+        SimplifyRanges();
+    }
+
+    /// Map of accessed ranges (mapping start address to range size)
+    std::map<u32, u32> ranges;
+};
+
 } // namespace
 
 } // namespace
diff --git a/src/video_core/rasterizer.cpp b/src/video_core/rasterizer.cpp
index 0434ad05a..9cf77b1f2 100644
--- a/src/video_core/rasterizer.cpp
+++ b/src/video_core/rasterizer.cpp
@@ -9,7 +9,6 @@
 #include "common/common_types.h"
 #include "common/math_util.h"
 #include "common/microprofile.h"
-#include "common/profiler.h"
 
 #include "core/memory.h"
 #include "core/hw/gpu.h"
@@ -287,7 +286,6 @@ static int SignedArea (const Math::Vec2<Fix12P4>& vtx1,
     return Math::Cross(vec1, vec2).z;
 };
 
-static Common::Profiling::TimingCategory rasterization_category("Rasterization");
 MICROPROFILE_DEFINE(GPU_Rasterization, "GPU", "Rasterization", MP_RGB(50, 50, 240));
 
 /**
@@ -300,7 +298,6 @@ static void ProcessTriangleInternal(const Shader::OutputVertex& v0,
                                     bool reversed = false)
 {
     const auto& regs = g_state.regs;
-    Common::Profiling::ScopeTimer timer(rasterization_category);
     MICROPROFILE_SCOPE(GPU_Rasterization);
 
     // vertex positions in rasterizer coordinates
diff --git a/src/video_core/renderer_opengl/gl_rasterizer.cpp b/src/video_core/renderer_opengl/gl_rasterizer.cpp
index 30187d4cf..a8c775c80 100644
--- a/src/video_core/renderer_opengl/gl_rasterizer.cpp
+++ b/src/video_core/renderer_opengl/gl_rasterizer.cpp
@@ -11,7 +11,6 @@
 #include "common/file_util.h"
 #include "common/math_util.h"
 #include "common/microprofile.h"
-#include "common/profiler.h"
 
 #include "core/memory.h"
 #include "core/settings.h"
diff --git a/src/video_core/shader/shader.cpp b/src/video_core/shader/shader.cpp
index 75301accd..043e99190 100644
--- a/src/video_core/shader/shader.cpp
+++ b/src/video_core/shader/shader.cpp
@@ -9,7 +9,6 @@
 
 #include "common/hash.h"
 #include "common/microprofile.h"
-#include "common/profiler.h"
 
 #include "video_core/debug_utils/debug_utils.h"
 #include "video_core/pica.h"
@@ -57,13 +56,11 @@ void Shutdown() {
 #endif // ARCHITECTURE_x86_64
 }
 
-static Common::Profiling::TimingCategory shader_category("Vertex Shader");
 MICROPROFILE_DEFINE(GPU_VertexShader, "GPU", "Vertex Shader", MP_RGB(50, 50, 240));
 
 OutputVertex Run(UnitState<false>& state, const InputVertex& input, int num_attributes) {
     auto& config = g_state.regs.vs;
 
-    Common::Profiling::ScopeTimer timer(shader_category);
     MICROPROFILE_SCOPE(GPU_VertexShader);
 
     state.program_counter = config.main_offset;
diff --git a/src/video_core/shader/shader.h b/src/video_core/shader/shader.h
index 9c5bd97bd..9ce9344d2 100644
--- a/src/video_core/shader/shader.h
+++ b/src/video_core/shader/shader.h
@@ -25,7 +25,7 @@ namespace Pica {
 namespace Shader {
 
 struct InputVertex {
-    Math::Vec4<float24> attr[16];
+    alignas(16) Math::Vec4<float24> attr[16];
 };
 
 struct OutputVertex {
diff --git a/src/video_core/vertex_loader.cpp b/src/video_core/vertex_loader.cpp
new file mode 100644
index 000000000..8a3d91896
--- /dev/null
+++ b/src/video_core/vertex_loader.cpp
@@ -0,0 +1,140 @@
+#include <cmath>
+#include <string>
+
+#include "boost/range/algorithm/fill.hpp"
+
+#include "common/assert.h"
+#include "common/alignment.h"
+#include "common/bit_field.h"
+#include "common/common_funcs.h"
+#include "common/common_types.h"
+#include "common/logging/log.h"
+
+#include "core/memory.h"
+
+#include "video_core/debug_utils/debug_utils.h"
+#include "video_core/pica.h"
+#include "video_core/pica_state.h"
+#include "video_core/pica_types.h"
+#include "video_core/vertex_loader.h"
+
+namespace Pica {
+
+void VertexLoader::Setup(const Pica::Regs& regs) {
+    const auto& attribute_config = regs.vertex_attributes;
+    num_total_attributes = attribute_config.GetNumTotalAttributes();
+
+    boost::fill(vertex_attribute_sources, 0xdeadbeef);
+
+    for (int i = 0; i < 16; i++) {
+        vertex_attribute_is_default[i] = attribute_config.IsDefaultAttribute(i);
+    }
+
+    // Setup attribute data from loaders
+    for (int loader = 0; loader < 12; ++loader) {
+        const auto& loader_config = attribute_config.attribute_loaders[loader];
+
+        u32 offset = 0;
+
+        // TODO: What happens if a loader overwrites a previous one's data?
+        for (unsigned component = 0; component < loader_config.component_count; ++component) {
+            if (component >= 12) {
+                LOG_ERROR(HW_GPU, "Overflow in the vertex attribute loader %u trying to load component %u", loader, component);
+                continue;
+            }
+
+            u32 attribute_index = loader_config.GetComponent(component);
+            if (attribute_index < 12) {
+                offset = Common::AlignUp(offset, attribute_config.GetElementSizeInBytes(attribute_index));
+                vertex_attribute_sources[attribute_index] = loader_config.data_offset + offset;
+                vertex_attribute_strides[attribute_index] = static_cast<u32>(loader_config.byte_count);
+                vertex_attribute_formats[attribute_index] = attribute_config.GetFormat(attribute_index);
+                vertex_attribute_elements[attribute_index] = attribute_config.GetNumElements(attribute_index);
+                offset += attribute_config.GetStride(attribute_index);
+            } else if (attribute_index < 16) {
+                // Attribute ids 12, 13, 14 and 15 signify 4, 8, 12 and 16-byte paddings, respectively
+                offset = Common::AlignUp(offset, 4);
+                offset += (attribute_index - 11) * 4;
+            } else {
+                UNREACHABLE(); // This is truly unreachable due to the number of bits for each component
+            }
+        }
+    }
+}
+
+void VertexLoader::LoadVertex(u32 base_address, int index, int vertex, Shader::InputVertex& input, DebugUtils::MemoryAccessTracker& memory_accesses) {
+    for (int i = 0; i < num_total_attributes; ++i) {
+        if (vertex_attribute_elements[i] != 0) {
+            // Load per-vertex data from the loader arrays
+            u32 source_addr = base_address + vertex_attribute_sources[i] + vertex_attribute_strides[i] * vertex;
+
+            if (g_debug_context && Pica::g_debug_context->recorder) {
+                memory_accesses.AddAccess(source_addr, vertex_attribute_elements[i] * (
+                    (vertex_attribute_formats[i] == Regs::VertexAttributeFormat::FLOAT) ? 4
+                    : (vertex_attribute_formats[i] == Regs::VertexAttributeFormat::SHORT) ? 2 : 1));
+            }
+
+            switch (vertex_attribute_formats[i]) {
+            case Regs::VertexAttributeFormat::BYTE:
+            {
+                const s8* srcdata = reinterpret_cast<const s8*>(Memory::GetPhysicalPointer(source_addr));
+                for (unsigned int comp = 0; comp < vertex_attribute_elements[i]; ++comp) {
+                    input.attr[i][comp] = float24::FromFloat32(srcdata[comp]);
+                }
+                break;
+            }
+            case Regs::VertexAttributeFormat::UBYTE:
+            {
+                const u8* srcdata = reinterpret_cast<const u8*>(Memory::GetPhysicalPointer(source_addr));
+                for (unsigned int comp = 0; comp < vertex_attribute_elements[i]; ++comp) {
+                    input.attr[i][comp] = float24::FromFloat32(srcdata[comp]);
+                }
+                break;
+            }
+            case Regs::VertexAttributeFormat::SHORT:
+            {
+                const s16* srcdata = reinterpret_cast<const s16*>(Memory::GetPhysicalPointer(source_addr));
+                for (unsigned int comp = 0; comp < vertex_attribute_elements[i]; ++comp) {
+                    input.attr[i][comp] = float24::FromFloat32(srcdata[comp]);
+                }
+                break;
+            }
+            case Regs::VertexAttributeFormat::FLOAT:
+            {
+                const float* srcdata = reinterpret_cast<const float*>(Memory::GetPhysicalPointer(source_addr));
+                for (unsigned int comp = 0; comp < vertex_attribute_elements[i]; ++comp) {
+                    input.attr[i][comp] = float24::FromFloat32(srcdata[comp]);
+                }
+                break;
+            }
+            }
+
+            // Default attribute values set if array elements have < 4 components. This
+            // is *not* carried over from the default attribute settings even if they're
+            // enabled for this attribute.
+            for (unsigned int comp = vertex_attribute_elements[i]; comp < 4; ++comp) {
+                input.attr[i][comp] = comp == 3 ? float24::FromFloat32(1.0f) : float24::FromFloat32(0.0f);
+            }
+
+            LOG_TRACE(HW_GPU, "Loaded %d components of attribute %x for vertex %x (index %x) from 0x%08x + 0x%08x + 0x%04x: %f %f %f %f",
+                vertex_attribute_elements[i], i, vertex, index,
+                base_address,
+                vertex_attribute_sources[i],
+                vertex_attribute_strides[i] * vertex,
+                input.attr[i][0].ToFloat32(), input.attr[i][1].ToFloat32(), input.attr[i][2].ToFloat32(), input.attr[i][3].ToFloat32());
+        } else if (vertex_attribute_is_default[i]) {
+            // Load the default attribute if we're configured to do so
+            input.attr[i] = g_state.vs.default_attributes[i];
+            LOG_TRACE(HW_GPU, "Loaded default attribute %x for vertex %x (index %x): (%f, %f, %f, %f)",
+                i, vertex, index,
+                input.attr[i][0].ToFloat32(), input.attr[i][1].ToFloat32(),
+                input.attr[i][2].ToFloat32(), input.attr[i][3].ToFloat32());
+        } else {
+            // TODO(yuriks): In this case, no data gets loaded and the vertex
+            // remains with the last value it had. This isn't currently maintained
+            // as global state, however, and so won't work in Citra yet.
+        }
+    }
+}
+
+}  // namespace Pica
\ No newline at end of file
diff --git a/src/video_core/vertex_loader.h b/src/video_core/vertex_loader.h
new file mode 100644
index 000000000..ff42d1596
--- /dev/null
+++ b/src/video_core/vertex_loader.h
@@ -0,0 +1,28 @@
+#pragma once
+
+#include <iterator>
+#include <algorithm>
+
+#include "video_core/pica.h"
+#include "video_core/shader/shader.h"
+#include "video_core/debug_utils/debug_utils.h"
+
+namespace Pica {
+
+class VertexLoader {
+public:
+    void Setup(const Pica::Regs& regs);
+    void LoadVertex(u32 base_address, int index, int vertex, Shader::InputVertex& input, DebugUtils::MemoryAccessTracker& memory_accesses);
+
+    int GetNumTotalAttributes() const { return num_total_attributes; }
+
+private:
+    u32 vertex_attribute_sources[16];
+    u32 vertex_attribute_strides[16] = {};
+    Regs::VertexAttributeFormat vertex_attribute_formats[16] = {};
+    u32 vertex_attribute_elements[16] = {};
+    bool vertex_attribute_is_default[16];
+    int num_total_attributes;
+};
+
+}  // namespace Pica