Merge branch 'master' into BGRA8-Framebuffer-Format

This reverts commit 865387f3c3.

Conflicts:
	src/video_core/gpu.h
	src/video_core/renderer_opengl/gl_rasterizer_cache.cpp
	src/video_core/surface.cpp

.gitmodules

@@ -37,6 +37,3 @@
 [submodule "discord-rpc"]
     path = externals/discord-rpc
     url = https://github.com/discordapp/discord-rpc.git
-[submodule "Vulkan-Headers"]
-    path = externals/Vulkan-Headers
-    url = https://github.com/KhronosGroup/Vulkan-Headers.git

CMakeLists.txt

@@ -23,8 +23,6 @@ option(YUZU_USE_QT_WEB_ENGINE "Use QtWebEngine for web applet implementation" OF
 
 option(ENABLE_CUBEB "Enables the cubeb audio backend" ON)
 
-option(ENABLE_VULKAN "Enables Vulkan backend" ON)
-
 option(USE_DISCORD_PRESENCE "Enables Discord Rich Presence" OFF)
 
 if(NOT EXISTS ${PROJECT_SOURCE_DIR}/.git/hooks/pre-commit)

README.md

@@ -7,7 +7,7 @@ yuzu is an experimental open-source emulator for the Nintendo Switch from the cr
 
 It is written in C++ with portability in mind, with builds actively maintained for Windows, Linux and macOS. The emulator is currently only useful for homebrew development and research purposes.
 
-yuzu only emulates a subset of Switch hardware and therefore is generally only useful for running/debugging homebrew applications. At this time, yuzu cannot play any commercial games without major problems. yuzu can boot some games, to varying degrees of success.
+yuzu only emulates a subset of Switch hardware and therefore is generally only useful for running/debugging homebrew applications. At this time, yuzu cannot play any commercial games without major problems. yuzu can boot some games, to varying degrees of success, but does not implement any of the necessary GPU features to render 3D graphics.
 
 yuzu is licensed under the GPLv2 (or any later version). Refer to the license.txt file included.
 

src/audio_core/audio_out.cpp

@@ -26,15 +26,14 @@ static Stream::Format ChannelsToStreamFormat(u32 num_channels) {
     return {};
 }
 
-StreamPtr AudioOut::OpenStream(Core::Timing::CoreTiming& core_timing, u32 sample_rate,
-                               u32 num_channels, std::string&& name,
+StreamPtr AudioOut::OpenStream(u32 sample_rate, u32 num_channels, std::string&& name,
                                Stream::ReleaseCallback&& release_callback) {
     if (!sink) {
         sink = CreateSinkFromID(Settings::values.sink_id, Settings::values.audio_device_id);
     }
 
     return std::make_shared<Stream>(
-        core_timing, sample_rate, ChannelsToStreamFormat(num_channels), std::move(release_callback),
+        sample_rate, ChannelsToStreamFormat(num_channels), std::move(release_callback),
         sink->AcquireSinkStream(sample_rate, num_channels, name), std::move(name));
 }
 

src/audio_core/audio_out.h

@@ -13,10 +13,6 @@
 #include "audio_core/stream.h"
 #include "common/common_types.h"
 
-namespace Core::Timing {
-class CoreTiming;
-}
-
 namespace AudioCore {
 
 /**
@@ -25,8 +21,8 @@ namespace AudioCore {
 class AudioOut {
 public:
     /// Opens a new audio stream
-    StreamPtr OpenStream(Core::Timing::CoreTiming& core_timing, u32 sample_rate, u32 num_channels,
-                         std::string&& name, Stream::ReleaseCallback&& release_callback);
+    StreamPtr OpenStream(u32 sample_rate, u32 num_channels, std::string&& name,
+                         Stream::ReleaseCallback&& release_callback);
 
     /// Returns a vector of recently released buffers specified by tag for the specified stream
     std::vector<Buffer::Tag> GetTagsAndReleaseBuffers(StreamPtr stream, std::size_t max_count);

src/audio_core/audio_renderer.cpp

@@ -8,7 +8,6 @@
 #include "audio_core/codec.h"
 #include "common/assert.h"
 #include "common/logging/log.h"
-#include "core/core.h"
 #include "core/hle/kernel/writable_event.h"
 #include "core/memory.h"
 
@@ -72,14 +71,14 @@ private:
     EffectOutStatus out_status{};
     EffectInStatus info{};
 };
-AudioRenderer::AudioRenderer(Core::Timing::CoreTiming& core_timing, AudioRendererParameter params,
+AudioRenderer::AudioRenderer(AudioRendererParameter params,
                              Kernel::SharedPtr<Kernel::WritableEvent> buffer_event)
     : worker_params{params}, buffer_event{buffer_event}, voices(params.voice_count),
       effects(params.effect_count) {
 
     audio_out = std::make_unique<AudioCore::AudioOut>();
-    stream = audio_out->OpenStream(core_timing, STREAM_SAMPLE_RATE, STREAM_NUM_CHANNELS,
-                                   "AudioRenderer", [=]() { buffer_event->Signal(); });
+    stream = audio_out->OpenStream(STREAM_SAMPLE_RATE, STREAM_NUM_CHANNELS, "AudioRenderer",
+                                   [=]() { buffer_event->Signal(); });
     audio_out->StartStream(stream);
 
     QueueMixedBuffer(0);

src/audio_core/audio_renderer.h

@@ -14,10 +14,6 @@
 #include "common/swap.h"
 #include "core/hle/kernel/object.h"
 
-namespace Core::Timing {
-class CoreTiming;
-}
-
 namespace Kernel {
 class WritableEvent;
 }
@@ -46,18 +42,16 @@ struct AudioRendererParameter {
     u32_le sample_rate;
     u32_le sample_count;
     u32_le mix_buffer_count;
-    u32_le submix_count;
+    u32_le unknown_c;
     u32_le voice_count;
     u32_le sink_count;
     u32_le effect_count;
-    u32_le performance_frame_count;
-    u8 is_voice_drop_enabled;
-    u8 unknown_21;
-    u8 unknown_22;
-    u8 execution_mode;
+    u32_le unknown_1c;
+    u8 unknown_20;
+    INSERT_PADDING_BYTES(3);
     u32_le splitter_count;
-    u32_le num_splitter_send_channels;
-    u32_le unknown_30;
+    u32_le unknown_2c;
+    INSERT_PADDING_WORDS(1);
     u32_le revision;
 };
 static_assert(sizeof(AudioRendererParameter) == 52, "AudioRendererParameter is an invalid size");
@@ -214,7 +208,7 @@ static_assert(sizeof(UpdateDataHeader) == 0x40, "UpdateDataHeader has wrong size
 
 class AudioRenderer {
 public:
-    AudioRenderer(Core::Timing::CoreTiming& core_timing, AudioRendererParameter params,
+    AudioRenderer(AudioRendererParameter params,
                   Kernel::SharedPtr<Kernel::WritableEvent> buffer_event);
     ~AudioRenderer();
 

src/audio_core/buffer.h

@@ -21,7 +21,7 @@ public:
     Buffer(Tag tag, std::vector<s16>&& samples) : tag{tag}, samples{std::move(samples)} {}
 
     /// Returns the raw audio data for the buffer
-    std::vector<s16>& GetSamples() {
+    std::vector<s16>& Samples() {
         return samples;
     }
 

src/audio_core/codec.cpp

@@ -68,8 +68,8 @@ std::vector<s16> DecodeADPCM(const u8* const data, std::size_t size, const ADPCM
         }
     }
 
-    state.yn1 = static_cast<s16>(yn1);
-    state.yn2 = static_cast<s16>(yn2);
+    state.yn1 = yn1;
+    state.yn2 = yn2;
 
     return ret;
 }

src/audio_core/cubeb_sink.cpp

@@ -12,10 +12,6 @@
 #include "common/ring_buffer.h"
 #include "core/settings.h"
 
-#ifdef _MSC_VER
-#include <objbase.h>
-#endif
-
 namespace AudioCore {
 
 class CubebSinkStream final : public SinkStream {
@@ -50,7 +46,7 @@ public:
         }
     }
 
-    ~CubebSinkStream() override {
+    ~CubebSinkStream() {
         if (!ctx) {
             return;
         }
@@ -79,11 +75,11 @@ public:
         queue.Push(samples);
     }
 
-    std::size_t SamplesInQueue(u32 channel_count) const override {
+    std::size_t SamplesInQueue(u32 num_channels) const override {
         if (!ctx)
             return 0;
 
-        return queue.Size() / channel_count;
+        return queue.Size() / num_channels;
     }
 
     void Flush() override {
@@ -102,7 +98,7 @@ private:
     u32 num_channels{};
 
     Common::RingBuffer<s16, 0x10000> queue;
-    std::array<s16, 2> last_frame{};
+    std::array<s16, 2> last_frame;
     std::atomic<bool> should_flush{};
     TimeStretcher time_stretch;
 
@@ -112,11 +108,6 @@ private:
 };
 
 CubebSink::CubebSink(std::string_view target_device_name) {
-    // Cubeb requires COM to be initialized on the thread calling cubeb_init on Windows
-#ifdef _MSC_VER
-    com_init_result = CoInitializeEx(nullptr, COINIT_MULTITHREADED);
-#endif
-
     if (cubeb_init(&ctx, "yuzu", nullptr) != CUBEB_OK) {
         LOG_CRITICAL(Audio_Sink, "cubeb_init failed");
         return;
@@ -151,12 +142,6 @@ CubebSink::~CubebSink() {
     }
 
     cubeb_destroy(ctx);
-
-#ifdef _MSC_VER
-    if (SUCCEEDED(com_init_result)) {
-        CoUninitialize();
-    }
-#endif
 }
 
 SinkStream& CubebSink::AcquireSinkStream(u32 sample_rate, u32 num_channels,

src/audio_core/cubeb_sink.h

@@ -25,10 +25,6 @@ private:
     cubeb* ctx{};
     cubeb_devid output_device{};
     std::vector<SinkStreamPtr> sink_streams;
-
-#ifdef _MSC_VER
-    u32 com_init_result = 0;
-#endif
 };
 
 std::vector<std::string> ListCubebSinkDevices();

src/audio_core/stream.cpp

@@ -32,12 +32,12 @@ u32 Stream::GetNumChannels() const {
     return {};
 }
 
-Stream::Stream(Core::Timing::CoreTiming& core_timing, u32 sample_rate, Format format,
-               ReleaseCallback&& release_callback, SinkStream& sink_stream, std::string&& name_)
+Stream::Stream(u32 sample_rate, Format format, ReleaseCallback&& release_callback,
+               SinkStream& sink_stream, std::string&& name_)
     : sample_rate{sample_rate}, format{format}, release_callback{std::move(release_callback)},
-      sink_stream{sink_stream}, core_timing{core_timing}, name{std::move(name_)} {
+      sink_stream{sink_stream}, name{std::move(name_)} {
 
-    release_event = core_timing.RegisterEvent(
+    release_event = Core::Timing::RegisterEvent(
         name, [this](u64 userdata, int cycles_late) { ReleaseActiveBuffer(); });
 }
 
@@ -95,11 +95,12 @@ void Stream::PlayNextBuffer() {
     active_buffer = queued_buffers.front();
     queued_buffers.pop();
 
-    VolumeAdjustSamples(active_buffer->GetSamples());
+    VolumeAdjustSamples(active_buffer->Samples());
 
     sink_stream.EnqueueSamples(GetNumChannels(), active_buffer->GetSamples());
 
-    core_timing.ScheduleEventThreadsafe(GetBufferReleaseCycles(*active_buffer), release_event, {});
+    Core::Timing::ScheduleEventThreadsafe(GetBufferReleaseCycles(*active_buffer), release_event,
+                                          {});
 }
 
 void Stream::ReleaseActiveBuffer() {

src/audio_core/stream.h

@@ -14,9 +14,8 @@
 #include "common/common_types.h"
 
 namespace Core::Timing {
-class CoreTiming;
 struct EventType;
-} // namespace Core::Timing
+}
 
 namespace AudioCore {
 
@@ -43,8 +42,8 @@ public:
     /// Callback function type, used to change guest state on a buffer being released
     using ReleaseCallback = std::function<void()>;
 
-    Stream(Core::Timing::CoreTiming& core_timing, u32 sample_rate, Format format,
-           ReleaseCallback&& release_callback, SinkStream& sink_stream, std::string&& name_);
+    Stream(u32 sample_rate, Format format, ReleaseCallback&& release_callback,
+           SinkStream& sink_stream, std::string&& name_);
 
     /// Plays the audio stream
     void Play();
@@ -101,7 +100,6 @@ private:
     std::queue<BufferPtr> queued_buffers;     ///< Buffers queued to be played in the stream
     std::queue<BufferPtr> released_buffers;   ///< Buffers recently released from the stream
     SinkStream& sink_stream;                  ///< Output sink for the stream
-    Core::Timing::CoreTiming& core_timing;    ///< Core timing instance.
     std::string name;                         ///< Name of the stream, must be unique
 };
 

src/common/color.h

@@ -55,36 +55,36 @@ constexpr u8 Convert8To6(u8 value) {
 /**
  * Decode a color stored in RGBA8 format
  * @param bytes Pointer to encoded source color
- * @return Result color decoded as Common::Vec4<u8>
+ * @return Result color decoded as Math::Vec4<u8>
  */
-inline Common::Vec4<u8> DecodeRGBA8(const u8* bytes) {
+inline Math::Vec4<u8> DecodeRGBA8(const u8* bytes) {
     return {bytes[3], bytes[2], bytes[1], bytes[0]};
 }
 
 /**
  * Decode a color stored in RGB8 format
  * @param bytes Pointer to encoded source color
- * @return Result color decoded as Common::Vec4<u8>
+ * @return Result color decoded as Math::Vec4<u8>
  */
-inline Common::Vec4<u8> DecodeRGB8(const u8* bytes) {
+inline Math::Vec4<u8> DecodeRGB8(const u8* bytes) {
     return {bytes[2], bytes[1], bytes[0], 255};
 }
 
 /**
  * Decode a color stored in RG8 (aka HILO8) format
  * @param bytes Pointer to encoded source color
- * @return Result color decoded as Common::Vec4<u8>
+ * @return Result color decoded as Math::Vec4<u8>
  */
-inline Common::Vec4<u8> DecodeRG8(const u8* bytes) {
+inline Math::Vec4<u8> DecodeRG8(const u8* bytes) {
     return {bytes[1], bytes[0], 0, 255};
 }
 
 /**
  * Decode a color stored in RGB565 format
  * @param bytes Pointer to encoded source color
- * @return Result color decoded as Common::Vec4<u8>
+ * @return Result color decoded as Math::Vec4<u8>
  */
-inline Common::Vec4<u8> DecodeRGB565(const u8* bytes) {
+inline Math::Vec4<u8> DecodeRGB565(const u8* bytes) {
     u16_le pixel;
     std::memcpy(&pixel, bytes, sizeof(pixel));
     return {Convert5To8((pixel >> 11) & 0x1F), Convert6To8((pixel >> 5) & 0x3F),
@@ -94,9 +94,9 @@ inline Common::Vec4<u8> DecodeRGB565(const u8* bytes) {
 /**
  * Decode a color stored in RGB5A1 format
  * @param bytes Pointer to encoded source color
- * @return Result color decoded as Common::Vec4<u8>
+ * @return Result color decoded as Math::Vec4<u8>
  */
-inline Common::Vec4<u8> DecodeRGB5A1(const u8* bytes) {
+inline Math::Vec4<u8> DecodeRGB5A1(const u8* bytes) {
     u16_le pixel;
     std::memcpy(&pixel, bytes, sizeof(pixel));
     return {Convert5To8((pixel >> 11) & 0x1F), Convert5To8((pixel >> 6) & 0x1F),
@@ -106,9 +106,9 @@ inline Common::Vec4<u8> DecodeRGB5A1(const u8* bytes) {
 /**
  * Decode a color stored in RGBA4 format
  * @param bytes Pointer to encoded source color
- * @return Result color decoded as Common::Vec4<u8>
+ * @return Result color decoded as Math::Vec4<u8>
  */
-inline Common::Vec4<u8> DecodeRGBA4(const u8* bytes) {
+inline Math::Vec4<u8> DecodeRGBA4(const u8* bytes) {
     u16_le pixel;
     std::memcpy(&pixel, bytes, sizeof(pixel));
     return {Convert4To8((pixel >> 12) & 0xF), Convert4To8((pixel >> 8) & 0xF),
@@ -138,9 +138,9 @@ inline u32 DecodeD24(const u8* bytes) {
 /**
  * Decode a depth value and a stencil value stored in D24S8 format
  * @param bytes Pointer to encoded source values
- * @return Resulting values stored as a Common::Vec2
+ * @return Resulting values stored as a Math::Vec2
  */
-inline Common::Vec2<u32> DecodeD24S8(const u8* bytes) {
+inline Math::Vec2<u32> DecodeD24S8(const u8* bytes) {
     return {static_cast<u32>((bytes[2] << 16) | (bytes[1] << 8) | bytes[0]), bytes[3]};
 }
 
@@ -149,7 +149,7 @@ inline Common::Vec2<u32> DecodeD24S8(const u8* bytes) {
  * @param color Source color to encode
  * @param bytes Destination pointer to store encoded color
  */
-inline void EncodeRGBA8(const Common::Vec4<u8>& color, u8* bytes) {
+inline void EncodeRGBA8(const Math::Vec4<u8>& color, u8* bytes) {
     bytes[3] = color.r();
     bytes[2] = color.g();
     bytes[1] = color.b();
@@ -161,7 +161,7 @@ inline void EncodeRGBA8(const Common::Vec4<u8>& color, u8* bytes) {
  * @param color Source color to encode
  * @param bytes Destination pointer to store encoded color
  */
-inline void EncodeRGB8(const Common::Vec4<u8>& color, u8* bytes) {
+inline void EncodeRGB8(const Math::Vec4<u8>& color, u8* bytes) {
     bytes[2] = color.r();
     bytes[1] = color.g();
     bytes[0] = color.b();
@@ -172,7 +172,7 @@ inline void EncodeRGB8(const Common::Vec4<u8>& color, u8* bytes) {
  * @param color Source color to encode
  * @param bytes Destination pointer to store encoded color
  */
-inline void EncodeRG8(const Common::Vec4<u8>& color, u8* bytes) {
+inline void EncodeRG8(const Math::Vec4<u8>& color, u8* bytes) {
     bytes[1] = color.r();
     bytes[0] = color.g();
 }
@@ -181,7 +181,7 @@ inline void EncodeRG8(const Common::Vec4<u8>& color, u8* bytes) {
  * @param color Source color to encode
  * @param bytes Destination pointer to store encoded color
  */
-inline void EncodeRGB565(const Common::Vec4<u8>& color, u8* bytes) {
+inline void EncodeRGB565(const Math::Vec4<u8>& color, u8* bytes) {
     const u16_le data =
         (Convert8To5(color.r()) << 11) | (Convert8To6(color.g()) << 5) | Convert8To5(color.b());
 
@@ -193,7 +193,7 @@ inline void EncodeRGB565(const Common::Vec4<u8>& color, u8* bytes) {
  * @param color Source color to encode
  * @param bytes Destination pointer to store encoded color
  */
-inline void EncodeRGB5A1(const Common::Vec4<u8>& color, u8* bytes) {
+inline void EncodeRGB5A1(const Math::Vec4<u8>& color, u8* bytes) {
     const u16_le data = (Convert8To5(color.r()) << 11) | (Convert8To5(color.g()) << 6) |
                         (Convert8To5(color.b()) << 1) | Convert8To1(color.a());
 
@@ -205,7 +205,7 @@ inline void EncodeRGB5A1(const Common::Vec4<u8>& color, u8* bytes) {
  * @param color Source color to encode
  * @param bytes Destination pointer to store encoded color
  */
-inline void EncodeRGBA4(const Common::Vec4<u8>& color, u8* bytes) {
+inline void EncodeRGBA4(const Math::Vec4<u8>& color, u8* bytes) {
     const u16 data = (Convert8To4(color.r()) << 12) | (Convert8To4(color.g()) << 8) |
                      (Convert8To4(color.b()) << 4) | Convert8To4(color.a());
 

src/common/logging/backend.cpp

@@ -40,7 +40,9 @@ public:
     const Impl& operator=(Impl const&) = delete;
 
     void PushEntry(Entry e) {
+        std::lock_guard<std::mutex> lock(message_mutex);
         message_queue.Push(std::move(e));
+        message_cv.notify_one();
     }
 
     void AddBackend(std::unique_ptr<Backend> backend) {
@@ -84,13 +86,15 @@ private:
                 }
             };
             while (true) {
-                entry = message_queue.PopWait();
-                if (entry.final_entry) {
+                {
+                    std::unique_lock<std::mutex> lock(message_mutex);
+                    message_cv.wait(lock, [&] { return !running || message_queue.Pop(entry); });
+                }
+                if (!running) {
                     break;
                 }
                 write_logs(entry);
             }
-
             // Drain the logging queue. Only writes out up to MAX_LOGS_TO_WRITE to prevent a case
             // where a system is repeatedly spamming logs even on close.
             const int MAX_LOGS_TO_WRITE = filter.IsDebug() ? INT_MAX : 100;
@@ -102,13 +106,14 @@ private:
     }
 
     ~Impl() {
-        Entry entry;
-        entry.final_entry = true;
-        message_queue.Push(entry);
+        running = false;
+        message_cv.notify_one();
         backend_thread.join();
     }
 
-    std::mutex writing_mutex;
+    std::atomic_bool running{true};
+    std::mutex message_mutex, writing_mutex;
+    std::condition_variable message_cv;
     std::thread backend_thread;
     std::vector<std::unique_ptr<Backend>> backends;
     Common::MPSCQueue<Log::Entry> message_queue;
@@ -227,7 +232,6 @@ void DebuggerBackend::Write(const Entry& entry) {
     CLS(Render)                                                                                    \
     SUB(Render, Software)                                                                          \
     SUB(Render, OpenGL)                                                                            \
-    SUB(Render, Vulkan)                                                                            \
     CLS(Audio)                                                                                     \
     SUB(Audio, DSP)                                                                                \
     SUB(Audio, Sink)                                                                               \

src/common/logging/backend.h

@@ -27,7 +27,6 @@ struct Entry {
     unsigned int line_num;
     std::string function;
     std::string message;
-    bool final_entry = false;
 
     Entry() = default;
     Entry(Entry&& o) = default;

src/common/logging/log.h

@@ -112,7 +112,6 @@ enum class Class : ClassType {
     Render,            ///< Emulator video output and hardware acceleration
     Render_Software,   ///< Software renderer backend
     Render_OpenGL,     ///< OpenGL backend
-    Render_Vulkan,     ///< Vulkan backend
     Audio,             ///< Audio emulation
     Audio_DSP,         ///< The HLE implementation of the DSP
     Audio_Sink,        ///< Emulator audio output backend

src/common/math_util.h

@@ -7,7 +7,7 @@
 #include <cstdlib>
 #include <type_traits>
 
-namespace Common {
+namespace MathUtil {
 
 constexpr float PI = 3.14159265f;
 
@@ -41,4 +41,4 @@ struct Rectangle {
     }
 };
 
-} // namespace Common
+} // namespace MathUtil

src/common/quaternion.h

@@ -6,12 +6,12 @@
 
 #include "common/vector_math.h"
 
-namespace Common {
+namespace Math {
 
 template <typename T>
 class Quaternion {
 public:
-    Vec3<T> xyz;
+    Math::Vec3<T> xyz;
     T w{};
 
     Quaternion<decltype(-T{})> Inverse() const {
@@ -38,12 +38,12 @@ public:
 };
 
 template <typename T>
-auto QuaternionRotate(const Quaternion<T>& q, const Vec3<T>& v) {
+auto QuaternionRotate(const Quaternion<T>& q, const Math::Vec3<T>& v) {
     return v + 2 * Cross(q.xyz, Cross(q.xyz, v) + v * q.w);
 }
 
-inline Quaternion<float> MakeQuaternion(const Vec3<float>& axis, float angle) {
+inline Quaternion<float> MakeQuaternion(const Math::Vec3<float>& axis, float angle) {
     return {axis * std::sin(angle / 2), std::cos(angle / 2)};
 }
 
-} // namespace Common
+} // namespace Math

src/common/swap.h

@@ -28,8 +28,8 @@
 #include <cstring>
 #include "common/common_types.h"
 
-// GCC
-#ifdef __GNUC__
+// GCC 4.6+
+#if __GNUC__ >= 5 || (__GNUC__ == 4 && __GNUC_MINOR__ >= 6)
 
 #if __BYTE_ORDER__ && (__BYTE_ORDER__ == __ORDER_LITTLE_ENDIAN__) && !defined(COMMON_LITTLE_ENDIAN)
 #define COMMON_LITTLE_ENDIAN 1
@@ -38,7 +38,7 @@
 #endif
 
 // LLVM/clang
-#elif defined(__clang__)
+#elif __clang__
 
 #if __LITTLE_ENDIAN__ && !defined(COMMON_LITTLE_ENDIAN)
 #define COMMON_LITTLE_ENDIAN 1

src/common/threadsafe_queue.h

@@ -7,17 +7,17 @@
 // a simple lockless thread-safe,
 // single reader, single writer queue
 
+#include <algorithm>
 #include <atomic>
-#include <condition_variable>
 #include <cstddef>
 #include <mutex>
-#include <utility>
+#include "common/common_types.h"
 
 namespace Common {
-template <typename T>
+template <typename T, bool NeedSize = true>
 class SPSCQueue {
 public:
-    SPSCQueue() {
+    SPSCQueue() : size(0) {
         write_ptr = read_ptr = new ElementPtr();
     }
     ~SPSCQueue() {
@@ -25,12 +25,13 @@ public:
         delete read_ptr;
     }
 
-    std::size_t Size() const {
+    u32 Size() const {
+        static_assert(NeedSize, "using Size() on FifoQueue without NeedSize");
         return size.load();
     }
 
     bool Empty() const {
-        return Size() == 0;
+        return !read_ptr->next.load();
     }
 
     T& Front() const {
@@ -46,14 +47,13 @@ public:
         ElementPtr* new_ptr = new ElementPtr();
         write_ptr->next.store(new_ptr, std::memory_order_release);
         write_ptr = new_ptr;
-        cv.notify_one();
-
-        ++size;
+        if (NeedSize)
+            size++;
     }
 
     void Pop() {
-        --size;
-
+        if (NeedSize)
+            size--;
         ElementPtr* tmpptr = read_ptr;
         // advance the read pointer
         read_ptr = tmpptr->next.load();
@@ -66,7 +66,8 @@ public:
         if (Empty())
             return false;
 
-        --size;
+        if (NeedSize)
+            size--;
 
         ElementPtr* tmpptr = read_ptr;
         read_ptr = tmpptr->next.load(std::memory_order_acquire);
@@ -76,16 +77,6 @@ public:
         return true;
     }
 
-    T PopWait() {
-        if (Empty()) {
-            std::unique_lock<std::mutex> lock(cv_mutex);
-            cv.wait(lock, [this]() { return !Empty(); });
-        }
-        T t;
-        Pop(t);
-        return t;
-    }
-
     // not thread-safe
     void Clear() {
         size.store(0);
@@ -98,7 +89,7 @@ private:
     // and a pointer to the next ElementPtr
     class ElementPtr {
     public:
-        ElementPtr() {}
+        ElementPtr() : next(nullptr) {}
         ~ElementPtr() {
             ElementPtr* next_ptr = next.load();
 
@@ -107,23 +98,21 @@ private:
         }
 
         T current;
-        std::atomic<ElementPtr*> next{nullptr};
+        std::atomic<ElementPtr*> next;
     };
 
     ElementPtr* write_ptr;
     ElementPtr* read_ptr;
-    std::atomic_size_t size{0};
-    std::mutex cv_mutex;
-    std::condition_variable cv;
+    std::atomic<u32> size;
 };
 
 // a simple thread-safe,
 // single reader, multiple writer queue
 
-template <typename T>
+template <typename T, bool NeedSize = true>
 class MPSCQueue {
 public:
-    std::size_t Size() const {
+    u32 Size() const {
         return spsc_queue.Size();
     }
 
@@ -149,17 +138,13 @@ public:
         return spsc_queue.Pop(t);
     }
 
-    T PopWait() {
-        return spsc_queue.PopWait();
-    }
-
     // not thread-safe
     void Clear() {
         spsc_queue.Clear();
     }
 
 private:
-    SPSCQueue<T> spsc_queue;
+    SPSCQueue<T, NeedSize> spsc_queue;
     std::mutex write_lock;
 };
 } // namespace Common

src/common/vector_math.h

@@ -33,7 +33,7 @@
 #include <cmath>
 #include <type_traits>
 
-namespace Common {
+namespace Math {
 
 template <typename T>
 class Vec2;
@@ -690,4 +690,4 @@ constexpr Vec4<T> MakeVec(const T& x, const Vec3<T>& yzw) {
     return MakeVec(x, yzw[0], yzw[1], yzw[2]);
 }
 
-} // namespace Common
+} // namespace Math

src/core/CMakeLists.txt

@@ -400,10 +400,6 @@ add_library(core STATIC
     hle/service/time/time.h
     hle/service/usb/usb.cpp
     hle/service/usb/usb.h
-    hle/service/vi/display/vi_display.cpp
-    hle/service/vi/display/vi_display.h
-    hle/service/vi/layer/vi_layer.cpp
-    hle/service/vi/layer/vi_layer.h
     hle/service/vi/vi.cpp
     hle/service/vi/vi.h
     hle/service/vi/vi_m.cpp

src/core/arm/dynarmic/arm_dynarmic.cpp

@@ -112,14 +112,14 @@ public:
         // Always execute at least one tick.
         amortized_ticks = std::max<u64>(amortized_ticks, 1);
 
-        parent.core_timing.AddTicks(amortized_ticks);
+        Timing::AddTicks(amortized_ticks);
         num_interpreted_instructions = 0;
     }
     u64 GetTicksRemaining() override {
-        return std::max(parent.core_timing.GetDowncount(), 0);
+        return std::max(Timing::GetDowncount(), 0);
     }
     u64 GetCNTPCT() override {
-        return parent.core_timing.GetTicks();
+        return Timing::GetTicks();
     }
 
     ARM_Dynarmic& parent;
@@ -172,10 +172,8 @@ void ARM_Dynarmic::Step() {
     cb->InterpreterFallback(jit->GetPC(), 1);
 }
 
-ARM_Dynarmic::ARM_Dynarmic(Timing::CoreTiming& core_timing, ExclusiveMonitor& exclusive_monitor,
-                           std::size_t core_index)
-    : cb(std::make_unique<ARM_Dynarmic_Callbacks>(*this)), inner_unicorn{core_timing},
-      core_index{core_index}, core_timing{core_timing},
+ARM_Dynarmic::ARM_Dynarmic(ExclusiveMonitor& exclusive_monitor, std::size_t core_index)
+    : cb(std::make_unique<ARM_Dynarmic_Callbacks>(*this)), core_index{core_index},
       exclusive_monitor{dynamic_cast<DynarmicExclusiveMonitor&>(exclusive_monitor)} {
     ThreadContext ctx{};
     inner_unicorn.SaveContext(ctx);

src/core/arm/dynarmic/arm_dynarmic.h

@@ -16,10 +16,6 @@ namespace Memory {
 struct PageTable;
 }
 
-namespace Core::Timing {
-class CoreTiming;
-}
-
 namespace Core {
 
 class ARM_Dynarmic_Callbacks;
@@ -27,8 +23,7 @@ class DynarmicExclusiveMonitor;
 
 class ARM_Dynarmic final : public ARM_Interface {
 public:
-    ARM_Dynarmic(Timing::CoreTiming& core_timing, ExclusiveMonitor& exclusive_monitor,
-                 std::size_t core_index);
+    ARM_Dynarmic(ExclusiveMonitor& exclusive_monitor, std::size_t core_index);
     ~ARM_Dynarmic();
 
     void MapBackingMemory(VAddr address, std::size_t size, u8* memory,
@@ -67,7 +62,6 @@ private:
     ARM_Unicorn inner_unicorn;
 
     std::size_t core_index;
-    Timing::CoreTiming& core_timing;
     DynarmicExclusiveMonitor& exclusive_monitor;
 
     Memory::PageTable* current_page_table = nullptr;

src/core/arm/unicorn/arm_unicorn.cpp

@@ -72,7 +72,7 @@ static bool UnmappedMemoryHook(uc_engine* uc, uc_mem_type type, u64 addr, int si
     return {};
 }
 
-ARM_Unicorn::ARM_Unicorn(Timing::CoreTiming& core_timing) : core_timing{core_timing} {
+ARM_Unicorn::ARM_Unicorn() {
     CHECKED(uc_open(UC_ARCH_ARM64, UC_MODE_ARM, &uc));
 
     auto fpv = 3 << 20;
@@ -177,7 +177,7 @@ void ARM_Unicorn::Run() {
     if (GDBStub::IsServerEnabled()) {
         ExecuteInstructions(std::max(4000000, 0));
     } else {
-        ExecuteInstructions(std::max(core_timing.GetDowncount(), 0));
+        ExecuteInstructions(std::max(Timing::GetDowncount(), 0));
     }
 }
 
@@ -190,7 +190,7 @@ MICROPROFILE_DEFINE(ARM_Jit_Unicorn, "ARM JIT", "Unicorn", MP_RGB(255, 64, 64));
 void ARM_Unicorn::ExecuteInstructions(int num_instructions) {
     MICROPROFILE_SCOPE(ARM_Jit_Unicorn);
     CHECKED(uc_emu_start(uc, GetPC(), 1ULL << 63, 0, num_instructions));
-    core_timing.AddTicks(num_instructions);
+    Timing::AddTicks(num_instructions);
     if (GDBStub::IsServerEnabled()) {
         if (last_bkpt_hit) {
             uc_reg_write(uc, UC_ARM64_REG_PC, &last_bkpt.address);

src/core/arm/unicorn/arm_unicorn.h

@@ -9,17 +9,12 @@
 #include "core/arm/arm_interface.h"
 #include "core/gdbstub/gdbstub.h"
 
-namespace Core::Timing {
-class CoreTiming;
-}
-
 namespace Core {
 
 class ARM_Unicorn final : public ARM_Interface {
 public:
-    explicit ARM_Unicorn(Timing::CoreTiming& core_timing);
+    ARM_Unicorn();
     ~ARM_Unicorn();
-
     void MapBackingMemory(VAddr address, std::size_t size, u8* memory,
                           Kernel::VMAPermission perms) override;
     void UnmapMemory(VAddr address, std::size_t size) override;
@@ -48,7 +43,6 @@ public:
 
 private:
     uc_engine* uc{};
-    Timing::CoreTiming& core_timing;
     GDBStub::BreakpointAddress last_bkpt{};
     bool last_bkpt_hit;
 };

src/core/core.cpp

@@ -94,8 +94,8 @@ struct System::Impl {
     ResultStatus Init(System& system, Frontend::EmuWindow& emu_window) {
         LOG_DEBUG(HW_Memory, "initialized OK");
 
-        core_timing.Initialize();
-        kernel.Initialize(core_timing);
+        Timing::Init();
+        kernel.Initialize();
 
         const auto current_time = std::chrono::duration_cast<std::chrono::seconds>(
             std::chrono::system_clock::now().time_since_epoch());
@@ -120,7 +120,7 @@ struct System::Impl {
         telemetry_session = std::make_unique<Core::TelemetrySession>();
         service_manager = std::make_shared<Service::SM::ServiceManager>();
 
-        Service::Init(service_manager, system, *virtual_filesystem);
+        Service::Init(service_manager, *virtual_filesystem);
         GDBStub::Init();
 
         renderer = VideoCore::CreateRenderer(emu_window, system);
@@ -128,7 +128,7 @@ struct System::Impl {
             return ResultStatus::ErrorVideoCore;
         }
 
-        gpu_core = std::make_unique<Tegra::GPU>(system, renderer->Rasterizer());
+        gpu_core = std::make_unique<Tegra::GPU>(renderer->Rasterizer());
 
         cpu_core_manager.Initialize(system);
         is_powered_on = true;
@@ -205,7 +205,7 @@ struct System::Impl {
 
         // Shutdown kernel and core timing
         kernel.Shutdown();
-        core_timing.Shutdown();
+        Timing::Shutdown();
 
         // Close app loader
         app_loader.reset();
@@ -232,10 +232,9 @@ struct System::Impl {
     }
 
     PerfStatsResults GetAndResetPerfStats() {
-        return perf_stats.GetAndResetStats(core_timing.GetGlobalTimeUs());
+        return perf_stats.GetAndResetStats(Timing::GetGlobalTimeUs());
     }
 
-    Timing::CoreTiming core_timing;
     Kernel::KernelCore kernel;
     /// RealVfsFilesystem instance
     FileSys::VirtualFilesystem virtual_filesystem;
@@ -397,14 +396,6 @@ const Kernel::KernelCore& System::Kernel() const {
     return impl->kernel;
 }
 
-Timing::CoreTiming& System::CoreTiming() {
-    return impl->core_timing;
-}
-
-const Timing::CoreTiming& System::CoreTiming() const {
-    return impl->core_timing;
-}
-
 Core::PerfStats& System::GetPerfStats() {
     return impl->perf_stats;
 }

src/core/core.h

@@ -47,10 +47,6 @@ namespace VideoCore {
 class RendererBase;
 } // namespace VideoCore
 
-namespace Core::Timing {
-class CoreTiming;
-}
-
 namespace Core {
 
 class ARM_Interface;
@@ -209,12 +205,6 @@ public:
     /// Provides a constant pointer to the current process.
     const Kernel::Process* CurrentProcess() const;
 
-    /// Provides a reference to the core timing instance.
-    Timing::CoreTiming& CoreTiming();
-
-    /// Provides a constant reference to the core timing instance.
-    const Timing::CoreTiming& CoreTiming() const;
-
     /// Provides a reference to the kernel instance.
     Kernel::KernelCore& Kernel();
 

src/core/core_cpu.cpp

@@ -49,18 +49,17 @@ bool CpuBarrier::Rendezvous() {
     return false;
 }
 
-Cpu::Cpu(Timing::CoreTiming& core_timing, ExclusiveMonitor& exclusive_monitor,
-         CpuBarrier& cpu_barrier, std::size_t core_index)
-    : cpu_barrier{cpu_barrier}, core_timing{core_timing}, core_index{core_index} {
+Cpu::Cpu(ExclusiveMonitor& exclusive_monitor, CpuBarrier& cpu_barrier, std::size_t core_index)
+    : cpu_barrier{cpu_barrier}, core_index{core_index} {
     if (Settings::values.use_cpu_jit) {
 #ifdef ARCHITECTURE_x86_64
-        arm_interface = std::make_unique<ARM_Dynarmic>(core_timing, exclusive_monitor, core_index);
+        arm_interface = std::make_unique<ARM_Dynarmic>(exclusive_monitor, core_index);
 #else
         arm_interface = std::make_unique<ARM_Unicorn>();
         LOG_WARNING(Core, "CPU JIT requested, but Dynarmic not available");
 #endif
     } else {
-        arm_interface = std::make_unique<ARM_Unicorn>(core_timing);
+        arm_interface = std::make_unique<ARM_Unicorn>();
     }
 
     scheduler = std::make_unique<Kernel::Scheduler>(*arm_interface);
@@ -94,14 +93,14 @@ void Cpu::RunLoop(bool tight_loop) {
 
         if (IsMainCore()) {
             // TODO(Subv): Only let CoreTiming idle if all 4 cores are idling.
-            core_timing.Idle();
-            core_timing.Advance();
+            Timing::Idle();
+            Timing::Advance();
         }
 
         PrepareReschedule();
     } else {
         if (IsMainCore()) {
-            core_timing.Advance();
+            Timing::Advance();
         }
 
         if (tight_loop) {

src/core/core_cpu.h

@@ -15,10 +15,6 @@ namespace Kernel {
 class Scheduler;
 }
 
-namespace Core::Timing {
-class CoreTiming;
-}
-
 namespace Core {
 
 class ARM_Interface;
@@ -45,8 +41,7 @@ private:
 
 class Cpu {
 public:
-    Cpu(Timing::CoreTiming& core_timing, ExclusiveMonitor& exclusive_monitor,
-        CpuBarrier& cpu_barrier, std::size_t core_index);
+    Cpu(ExclusiveMonitor& exclusive_monitor, CpuBarrier& cpu_barrier, std::size_t core_index);
     ~Cpu();
 
     void RunLoop(bool tight_loop = true);
@@ -87,7 +82,6 @@ private:
     std::unique_ptr<ARM_Interface> arm_interface;
     CpuBarrier& cpu_barrier;
     std::unique_ptr<Kernel::Scheduler> scheduler;
-    Timing::CoreTiming& core_timing;
 
     std::atomic<bool> reschedule_pending = false;
     std::size_t core_index;

src/core/core_timing.cpp

@@ -8,60 +8,71 @@
 #include <mutex>
 #include <string>
 #include <tuple>
-
+#include <unordered_map>
+#include <vector>
 #include "common/assert.h"
 #include "common/thread.h"
+#include "common/threadsafe_queue.h"
 #include "core/core_timing_util.h"
 
 namespace Core::Timing {
 
-constexpr int MAX_SLICE_LENGTH = 20000;
+static s64 global_timer;
+static int slice_length;
+static int downcount;
 
-struct CoreTiming::Event {
+struct EventType {
+    TimedCallback callback;
+    const std::string* name;
+};
+
+struct Event {
     s64 time;
     u64 fifo_order;
     u64 userdata;
     const EventType* type;
-
-    // Sort by time, unless the times are the same, in which case sort by
-    // the order added to the queue
-    friend bool operator>(const Event& left, const Event& right) {
-        return std::tie(left.time, left.fifo_order) > std::tie(right.time, right.fifo_order);
-    }
-
-    friend bool operator<(const Event& left, const Event& right) {
-        return std::tie(left.time, left.fifo_order) < std::tie(right.time, right.fifo_order);
-    }
 };
 
-CoreTiming::CoreTiming() = default;
-CoreTiming::~CoreTiming() = default;
-
-void CoreTiming::Initialize() {
-    downcount = MAX_SLICE_LENGTH;
-    slice_length = MAX_SLICE_LENGTH;
-    global_timer = 0;
-    idled_cycles = 0;
-
-    // The time between CoreTiming being initialized and the first call to Advance() is considered
-    // the slice boundary between slice -1 and slice 0. Dispatcher loops must call Advance() before
-    // executing the first cycle of each slice to prepare the slice length and downcount for
-    // that slice.
-    is_global_timer_sane = true;
-
-    event_fifo_id = 0;
-
-    const auto empty_timed_callback = [](u64, s64) {};
-    ev_lost = RegisterEvent("_lost_event", empty_timed_callback);
+// Sort by time, unless the times are the same, in which case sort by the order added to the queue
+static bool operator>(const Event& left, const Event& right) {
+    return std::tie(left.time, left.fifo_order) > std::tie(right.time, right.fifo_order);
 }
 
-void CoreTiming::Shutdown() {
-    MoveEvents();
-    ClearPendingEvents();
-    UnregisterAllEvents();
+static bool operator<(const Event& left, const Event& right) {
+    return std::tie(left.time, left.fifo_order) < std::tie(right.time, right.fifo_order);
 }
 
-EventType* CoreTiming::RegisterEvent(const std::string& name, TimedCallback callback) {
+// unordered_map stores each element separately as a linked list node so pointers to elements
+// remain stable regardless of rehashes/resizing.
+static std::unordered_map<std::string, EventType> event_types;
+
+// The queue is a min-heap using std::make_heap/push_heap/pop_heap.
+// We don't use std::priority_queue because we need to be able to serialize, unserialize and
+// erase arbitrary events (RemoveEvent()) regardless of the queue order. These aren't accomodated
+// by the standard adaptor class.
+static std::vector<Event> event_queue;
+static u64 event_fifo_id;
+// the queue for storing the events from other threads threadsafe until they will be added
+// to the event_queue by the emu thread
+static Common::MPSCQueue<Event, false> ts_queue;
+
+// the queue for unscheduling the events from other threads threadsafe
+static Common::MPSCQueue<std::pair<const EventType*, u64>, false> unschedule_queue;
+
+constexpr int MAX_SLICE_LENGTH = 20000;
+
+static s64 idled_cycles;
+
+// Are we in a function that has been called from Advance()
+// If events are sheduled from a function that gets called from Advance(),
+// don't change slice_length and downcount.
+static bool is_global_timer_sane;
+
+static EventType* ev_lost = nullptr;
+
+static void EmptyTimedCallback(u64 userdata, s64 cyclesLate) {}
+
+EventType* RegisterEvent(const std::string& name, TimedCallback callback) {
     // check for existing type with same name.
     // we want event type names to remain unique so that we can use them for serialization.
     ASSERT_MSG(event_types.find(name) == event_types.end(),
@@ -75,31 +86,71 @@ EventType* CoreTiming::RegisterEvent(const std::string& name, TimedCallback call
     return event_type;
 }
 
-void CoreTiming::UnregisterAllEvents() {
+void UnregisterAllEvents() {
     ASSERT_MSG(event_queue.empty(), "Cannot unregister events with events pending");
     event_types.clear();
 }
 
-void CoreTiming::ScheduleEvent(s64 cycles_into_future, const EventType* event_type, u64 userdata) {
-    ASSERT(event_type != nullptr);
-    const s64 timeout = GetTicks() + cycles_into_future;
+void Init() {
+    downcount = MAX_SLICE_LENGTH;
+    slice_length = MAX_SLICE_LENGTH;
+    global_timer = 0;
+    idled_cycles = 0;
 
-    // If this event needs to be scheduled before the next advance(), force one early
+    // The time between CoreTiming being intialized and the first call to Advance() is considered
+    // the slice boundary between slice -1 and slice 0. Dispatcher loops must call Advance() before
+    // executing the first cycle of each slice to prepare the slice length and downcount for
+    // that slice.
+    is_global_timer_sane = true;
+
+    event_fifo_id = 0;
+    ev_lost = RegisterEvent("_lost_event", &EmptyTimedCallback);
+}
+
+void Shutdown() {
+    MoveEvents();
+    ClearPendingEvents();
+    UnregisterAllEvents();
+}
+
+// This should only be called from the CPU thread. If you are calling
+// it from any other thread, you are doing something evil
+u64 GetTicks() {
+    u64 ticks = static_cast<u64>(global_timer);
     if (!is_global_timer_sane) {
-        ForceExceptionCheck(cycles_into_future);
+        ticks += slice_length - downcount;
     }
+    return ticks;
+}
 
+void AddTicks(u64 ticks) {
+    downcount -= static_cast<int>(ticks);
+}
+
+u64 GetIdleTicks() {
+    return static_cast<u64>(idled_cycles);
+}
+
+void ClearPendingEvents() {
+    event_queue.clear();
+}
+
+void ScheduleEvent(s64 cycles_into_future, const EventType* event_type, u64 userdata) {
+    ASSERT(event_type != nullptr);
+    s64 timeout = GetTicks() + cycles_into_future;
+    // If this event needs to be scheduled before the next advance(), force one early
+    if (!is_global_timer_sane)
+        ForceExceptionCheck(cycles_into_future);
     event_queue.emplace_back(Event{timeout, event_fifo_id++, userdata, event_type});
     std::push_heap(event_queue.begin(), event_queue.end(), std::greater<>());
 }
 
-void CoreTiming::ScheduleEventThreadsafe(s64 cycles_into_future, const EventType* event_type,
-                                         u64 userdata) {
+void ScheduleEventThreadsafe(s64 cycles_into_future, const EventType* event_type, u64 userdata) {
     ts_queue.Push(Event{global_timer + cycles_into_future, 0, userdata, event_type});
 }
 
-void CoreTiming::UnscheduleEvent(const EventType* event_type, u64 userdata) {
-    const auto itr = std::remove_if(event_queue.begin(), event_queue.end(), [&](const Event& e) {
+void UnscheduleEvent(const EventType* event_type, u64 userdata) {
+    auto itr = std::remove_if(event_queue.begin(), event_queue.end(), [&](const Event& e) {
         return e.type == event_type && e.userdata == userdata;
     });
 
@@ -110,33 +161,13 @@ void CoreTiming::UnscheduleEvent(const EventType* event_type, u64 userdata) {
     }
 }
 
-void CoreTiming::UnscheduleEventThreadsafe(const EventType* event_type, u64 userdata) {
+void UnscheduleEventThreadsafe(const EventType* event_type, u64 userdata) {
     unschedule_queue.Push(std::make_pair(event_type, userdata));
 }
 
-u64 CoreTiming::GetTicks() const {
-    u64 ticks = static_cast<u64>(global_timer);
-    if (!is_global_timer_sane) {
-        ticks += slice_length - downcount;
-    }
-    return ticks;
-}
-
-u64 CoreTiming::GetIdleTicks() const {
-    return static_cast<u64>(idled_cycles);
-}
-
-void CoreTiming::AddTicks(u64 ticks) {
-    downcount -= static_cast<int>(ticks);
-}
-
-void CoreTiming::ClearPendingEvents() {
-    event_queue.clear();
-}
-
-void CoreTiming::RemoveEvent(const EventType* event_type) {
-    const auto itr = std::remove_if(event_queue.begin(), event_queue.end(),
-                                    [&](const Event& e) { return e.type == event_type; });
+void RemoveEvent(const EventType* event_type) {
+    auto itr = std::remove_if(event_queue.begin(), event_queue.end(),
+                              [&](const Event& e) { return e.type == event_type; });
 
     // Removing random items breaks the invariant so we have to re-establish it.
     if (itr != event_queue.end()) {
@@ -145,24 +176,22 @@ void CoreTiming::RemoveEvent(const EventType* event_type) {
     }
 }
 
-void CoreTiming::RemoveNormalAndThreadsafeEvent(const EventType* event_type) {
+void RemoveNormalAndThreadsafeEvent(const EventType* event_type) {
     MoveEvents();
     RemoveEvent(event_type);
 }
 
-void CoreTiming::ForceExceptionCheck(s64 cycles) {
+void ForceExceptionCheck(s64 cycles) {
     cycles = std::max<s64>(0, cycles);
-    if (downcount <= cycles) {
-        return;
+    if (downcount > cycles) {
+        // downcount is always (much) smaller than MAX_INT so we can safely cast cycles to an int
+        // here. Account for cycles already executed by adjusting the g.slice_length
+        slice_length -= downcount - static_cast<int>(cycles);
+        downcount = static_cast<int>(cycles);
     }
-
-    // downcount is always (much) smaller than MAX_INT so we can safely cast cycles to an int
-    // here. Account for cycles already executed by adjusting the g.slice_length
-    slice_length -= downcount - static_cast<int>(cycles);
-    downcount = static_cast<int>(cycles);
 }
 
-void CoreTiming::MoveEvents() {
+void MoveEvents() {
     for (Event ev; ts_queue.Pop(ev);) {
         ev.fifo_order = event_fifo_id++;
         event_queue.emplace_back(std::move(ev));
@@ -170,13 +199,13 @@ void CoreTiming::MoveEvents() {
     }
 }
 
-void CoreTiming::Advance() {
+void Advance() {
     MoveEvents();
     for (std::pair<const EventType*, u64> ev; unschedule_queue.Pop(ev);) {
         UnscheduleEvent(ev.first, ev.second);
     }
 
-    const int cycles_executed = slice_length - downcount;
+    int cycles_executed = slice_length - downcount;
     global_timer += cycles_executed;
     slice_length = MAX_SLICE_LENGTH;
 
@@ -200,16 +229,16 @@ void CoreTiming::Advance() {
     downcount = slice_length;
 }
 
-void CoreTiming::Idle() {
+void Idle() {
     idled_cycles += downcount;
     downcount = 0;
 }
 
-std::chrono::microseconds CoreTiming::GetGlobalTimeUs() const {
+std::chrono::microseconds GetGlobalTimeUs() {
     return std::chrono::microseconds{GetTicks() * 1000000 / BASE_CLOCK_RATE};
 }
 
-int CoreTiming::GetDowncount() const {
+int GetDowncount() {
     return downcount;
 }
 

src/core/core_timing.h

@@ -4,27 +4,6 @@
 
 #pragma once
 
-#include <chrono>
-#include <functional>
-#include <string>
-#include <unordered_map>
-#include <vector>
-#include "common/common_types.h"
-#include "common/threadsafe_queue.h"
-
-namespace Core::Timing {
-
-/// A callback that may be scheduled for a particular core timing event.
-using TimedCallback = std::function<void(u64 userdata, int cycles_late)>;
-
-/// Contains the characteristics of a particular event.
-struct EventType {
-    /// The event's callback function.
-    TimedCallback callback;
-    /// A pointer to the name of the event.
-    const std::string* name;
-};
-
 /**
  * This is a system to schedule events into the emulated machine's future. Time is measured
  * in main CPU clock cycles.
@@ -37,120 +16,80 @@ struct EventType {
  * inside callback:
  *   ScheduleEvent(periodInCycles - cyclesLate, callback, "whatever")
  */
-class CoreTiming {
-public:
-    CoreTiming();
-    ~CoreTiming();
 
-    CoreTiming(const CoreTiming&) = delete;
-    CoreTiming(CoreTiming&&) = delete;
+#include <chrono>
+#include <functional>
+#include <string>
+#include "common/common_types.h"
 
-    CoreTiming& operator=(const CoreTiming&) = delete;
-    CoreTiming& operator=(CoreTiming&&) = delete;
+namespace Core::Timing {
 
-    /// CoreTiming begins at the boundary of timing slice -1. An initial call to Advance() is
-    /// required to end slice - 1 and start slice 0 before the first cycle of code is executed.
-    void Initialize();
+struct EventType;
 
-    /// Tears down all timing related functionality.
-    void Shutdown();
+using TimedCallback = std::function<void(u64 userdata, int cycles_late)>;
 
-    /// Registers a core timing event with the given name and callback.
-    ///
-    /// @param name     The name of the core timing event to register.
-    /// @param callback The callback to execute for the event.
-    ///
-    /// @returns An EventType instance representing the registered event.
-    ///
-    /// @pre The name of the event being registered must be unique among all
-    ///      registered events.
-    ///
-    EventType* RegisterEvent(const std::string& name, TimedCallback callback);
+/**
+ * CoreTiming begins at the boundary of timing slice -1. An initial call to Advance() is
+ * required to end slice -1 and start slice 0 before the first cycle of code is executed.
+ */
+void Init();
+void Shutdown();
 
-    /// Unregisters all registered events thus far.
-    void UnregisterAllEvents();
+/**
+ * This should only be called from the emu thread, if you are calling it any other thread, you are
+ * doing something evil
+ */
+u64 GetTicks();
+u64 GetIdleTicks();
+void AddTicks(u64 ticks);
 
-    /// After the first Advance, the slice lengths and the downcount will be reduced whenever an
-    /// event is scheduled earlier than the current values.
-    ///
-    /// Scheduling from a callback will not update the downcount until the Advance() completes.
-    void ScheduleEvent(s64 cycles_into_future, const EventType* event_type, u64 userdata = 0);
+/**
+ * Returns the event_type identifier. if name is not unique, it will assert.
+ */
+EventType* RegisterEvent(const std::string& name, TimedCallback callback);
+void UnregisterAllEvents();
 
-    /// This is to be called when outside of hle threads, such as the graphics thread, wants to
-    /// schedule things to be executed on the main thread.
-    ///
-    /// @note This doesn't change slice_length and thus events scheduled by this might be
-    /// called with a delay of up to MAX_SLICE_LENGTH
-    void ScheduleEventThreadsafe(s64 cycles_into_future, const EventType* event_type,
-                                 u64 userdata = 0);
+/**
+ * After the first Advance, the slice lengths and the downcount will be reduced whenever an event
+ * is scheduled earlier than the current values.
+ * Scheduling from a callback will not update the downcount until the Advance() completes.
+ */
+void ScheduleEvent(s64 cycles_into_future, const EventType* event_type, u64 userdata = 0);
 
-    void UnscheduleEvent(const EventType* event_type, u64 userdata);
-    void UnscheduleEventThreadsafe(const EventType* event_type, u64 userdata);
+/**
+ * This is to be called when outside of hle threads, such as the graphics thread, wants to
+ * schedule things to be executed on the main thread.
+ * Not that this doesn't change slice_length and thus events scheduled by this might be called
+ * with a delay of up to MAX_SLICE_LENGTH
+ */
+void ScheduleEventThreadsafe(s64 cycles_into_future, const EventType* event_type, u64 userdata);
 
-    /// We only permit one event of each type in the queue at a time.
-    void RemoveEvent(const EventType* event_type);
-    void RemoveNormalAndThreadsafeEvent(const EventType* event_type);
+void UnscheduleEvent(const EventType* event_type, u64 userdata);
+void UnscheduleEventThreadsafe(const EventType* event_type, u64 userdata);
 
-    void ForceExceptionCheck(s64 cycles);
+/// We only permit one event of each type in the queue at a time.
+void RemoveEvent(const EventType* event_type);
+void RemoveNormalAndThreadsafeEvent(const EventType* event_type);
 
-    /// This should only be called from the emu thread, if you are calling it any other thread,
-    /// you are doing something evil
-    u64 GetTicks() const;
+/** Advance must be called at the beginning of dispatcher loops, not the end. Advance() ends
+ * the previous timing slice and begins the next one, you must Advance from the previous
+ * slice to the current one before executing any cycles. CoreTiming starts in slice -1 so an
+ * Advance() is required to initialize the slice length before the first cycle of emulated
+ * instructions is executed.
+ */
+void Advance();
+void MoveEvents();
 
-    u64 GetIdleTicks() const;
+/// Pretend that the main CPU has executed enough cycles to reach the next event.
+void Idle();
 
-    void AddTicks(u64 ticks);
+/// Clear all pending events. This should ONLY be done on exit.
+void ClearPendingEvents();
 
-    /// Advance must be called at the beginning of dispatcher loops, not the end. Advance() ends
-    /// the previous timing slice and begins the next one, you must Advance from the previous
-    /// slice to the current one before executing any cycles. CoreTiming starts in slice -1 so an
-    /// Advance() is required to initialize the slice length before the first cycle of emulated
-    /// instructions is executed.
-    void Advance();
+void ForceExceptionCheck(s64 cycles);
 
-    /// Pretend that the main CPU has executed enough cycles to reach the next event.
-    void Idle();
+std::chrono::microseconds GetGlobalTimeUs();
 
-    std::chrono::microseconds GetGlobalTimeUs() const;
-
-    int GetDowncount() const;
-
-private:
-    struct Event;
-
-    /// Clear all pending events. This should ONLY be done on exit.
-    void ClearPendingEvents();
-    void MoveEvents();
-
-    s64 global_timer = 0;
-    s64 idled_cycles = 0;
-    int slice_length = 0;
-    int downcount = 0;
-
-    // Are we in a function that has been called from Advance()
-    // If events are scheduled from a function that gets called from Advance(),
-    // don't change slice_length and downcount.
-    bool is_global_timer_sane = false;
-
-    // The queue is a min-heap using std::make_heap/push_heap/pop_heap.
-    // We don't use std::priority_queue because we need to be able to serialize, unserialize and
-    // erase arbitrary events (RemoveEvent()) regardless of the queue order. These aren't
-    // accomodated by the standard adaptor class.
-    std::vector<Event> event_queue;
-    u64 event_fifo_id = 0;
-
-    // Stores each element separately as a linked list node so pointers to elements
-    // remain stable regardless of rehashes/resizing.
-    std::unordered_map<std::string, EventType> event_types;
-
-    // The queue for storing the events from other threads threadsafe until they will be added
-    // to the event_queue by the emu thread
-    Common::MPSCQueue<Event> ts_queue;
-
-    // The queue for unscheduling the events from other threads threadsafe
-    Common::MPSCQueue<std::pair<const EventType*, u64>> unschedule_queue;
-
-    EventType* ev_lost = nullptr;
-};
+int GetDowncount();
 
 } // namespace Core::Timing

src/core/cpu_core_manager.cpp

@@ -27,8 +27,7 @@ void CpuCoreManager::Initialize(System& system) {
     exclusive_monitor = Cpu::MakeExclusiveMonitor(cores.size());
 
     for (std::size_t index = 0; index < cores.size(); ++index) {
-        cores[index] =
-            std::make_unique<Cpu>(system.CoreTiming(), *exclusive_monitor, *barrier, index);
+        cores[index] = std::make_unique<Cpu>(*exclusive_monitor, *barrier, index);
     }
 
     // Create threads for CPU cores 1-3, and build thread_to_cpu map

src/core/crypto/key_manager.cpp

@@ -398,8 +398,7 @@ static bool ValidCryptoRevisionString(std::string_view base, size_t begin, size_
 }
 
 void KeyManager::LoadFromFile(const std::string& filename, bool is_title_keys) {
-    std::ifstream file;
-    OpenFStream(file, filename, std::ios_base::in);
+    std::ifstream file(filename);
     if (!file.is_open())
         return;
 

src/core/file_sys/vfs_vector.cpp

@@ -47,7 +47,7 @@ std::size_t VectorVfsFile::Write(const u8* data_, std::size_t length, std::size_
     if (offset + length > data.size())
         data.resize(offset + length);
     const auto write = std::min(length, data.size() - offset);
-    std::memcpy(data.data() + offset, data_, write);
+    std::memcpy(data.data(), data_, write);
     return write;
 }
 

src/core/frontend/emu_window.cpp

@@ -67,7 +67,7 @@ static bool IsWithinTouchscreen(const Layout::FramebufferLayout& layout, unsigne
             framebuffer_x >= layout.screen.left && framebuffer_x < layout.screen.right);
 }
 
-std::tuple<unsigned, unsigned> EmuWindow::ClipToTouchScreen(unsigned new_x, unsigned new_y) const {
+std::tuple<unsigned, unsigned> EmuWindow::ClipToTouchScreen(unsigned new_x, unsigned new_y) {
     new_x = std::max(new_x, framebuffer_layout.screen.left);
     new_x = std::min(new_x, framebuffer_layout.screen.right - 1);
 

src/core/frontend/emu_window.h

@@ -166,7 +166,7 @@ private:
     /**
      * Clip the provided coordinates to be inside the touchscreen area.
      */
-    std::tuple<unsigned, unsigned> ClipToTouchScreen(unsigned new_x, unsigned new_y) const;
+    std::tuple<unsigned, unsigned> ClipToTouchScreen(unsigned new_x, unsigned new_y);
 };
 
 } // namespace Core::Frontend

src/core/frontend/framebuffer_layout.cpp

@@ -12,12 +12,12 @@ namespace Layout {
 
 // Finds the largest size subrectangle contained in window area that is confined to the aspect ratio
 template <class T>
-static Common::Rectangle<T> MaxRectangle(Common::Rectangle<T> window_area,
-                                         float screen_aspect_ratio) {
+static MathUtil::Rectangle<T> maxRectangle(MathUtil::Rectangle<T> window_area,
+                                           float screen_aspect_ratio) {
     float scale = std::min(static_cast<float>(window_area.GetWidth()),
                            window_area.GetHeight() / screen_aspect_ratio);
-    return Common::Rectangle<T>{0, 0, static_cast<T>(std::round(scale)),
-                                static_cast<T>(std::round(scale * screen_aspect_ratio))};
+    return MathUtil::Rectangle<T>{0, 0, static_cast<T>(std::round(scale)),
+                                  static_cast<T>(std::round(scale * screen_aspect_ratio))};
 }
 
 FramebufferLayout DefaultFrameLayout(unsigned width, unsigned height) {
@@ -29,8 +29,8 @@ FramebufferLayout DefaultFrameLayout(unsigned width, unsigned height) {
 
     const float emulation_aspect_ratio{static_cast<float>(ScreenUndocked::Height) /
                                        ScreenUndocked::Width};
-    Common::Rectangle<unsigned> screen_window_area{0, 0, width, height};
-    Common::Rectangle<unsigned> screen = MaxRectangle(screen_window_area, emulation_aspect_ratio);
+    MathUtil::Rectangle<unsigned> screen_window_area{0, 0, width, height};
+    MathUtil::Rectangle<unsigned> screen = maxRectangle(screen_window_area, emulation_aspect_ratio);
 
     float window_aspect_ratio = static_cast<float>(height) / width;
 

src/core/frontend/framebuffer_layout.h

@@ -16,7 +16,7 @@ struct FramebufferLayout {
     unsigned width{ScreenUndocked::Width};
     unsigned height{ScreenUndocked::Height};
 
-    Common::Rectangle<unsigned> screen;
+    MathUtil::Rectangle<unsigned> screen;
 
     /**
      * Returns the ration of pixel size of the screen, compared to the native size of the undocked

src/core/frontend/input.h

@@ -124,7 +124,7 @@ using AnalogDevice = InputDevice<std::tuple<float, float>>;
  *   Orientation is determined by right-hand rule.
  *   Units: deg/sec
  */
-using MotionDevice = InputDevice<std::tuple<Common::Vec3<float>, Common::Vec3<float>>>;
+using MotionDevice = InputDevice<std::tuple<Math::Vec3<float>, Math::Vec3<float>>>;
 
 /**
  * A touch device is an input device that returns a tuple of two floats and a bool. The floats are

src/core/hle/kernel/address_arbiter.cpp

@@ -17,7 +17,8 @@
 #include "core/hle/result.h"
 #include "core/memory.h"
 
-namespace Kernel::AddressArbiter {
+namespace Kernel {
+namespace AddressArbiter {
 
 // Performs actual address waiting logic.
 static ResultCode WaitForAddress(VAddr address, s64 timeout) {
@@ -175,4 +176,5 @@ ResultCode WaitForAddressIfEqual(VAddr address, s32 value, s64 timeout) {
 
     return WaitForAddress(address, timeout);
 }
-} // namespace Kernel::AddressArbiter
+} // namespace AddressArbiter
+} // namespace Kernel

src/core/hle/kernel/address_arbiter.h

@@ -8,8 +8,9 @@
 
 union ResultCode;
 
-namespace Kernel::AddressArbiter {
+namespace Kernel {
 
+namespace AddressArbiter {
 enum class ArbitrationType {
     WaitIfLessThan = 0,
     DecrementAndWaitIfLessThan = 1,
@@ -28,5 +29,6 @@ ResultCode ModifyByWaitingCountAndSignalToAddressIfEqual(VAddr address, s32 valu
 
 ResultCode WaitForAddressIfLessThan(VAddr address, s32 value, s64 timeout, bool should_decrement);
 ResultCode WaitForAddressIfEqual(VAddr address, s32 value, s64 timeout);
+} // namespace AddressArbiter
 
-} // namespace Kernel::AddressArbiter
+} // namespace Kernel

src/core/hle/kernel/errors.h

@@ -14,7 +14,6 @@ constexpr ResultCode ERR_MAX_CONNECTIONS_REACHED{ErrorModule::Kernel, 7};
 constexpr ResultCode ERR_INVALID_CAPABILITY_DESCRIPTOR{ErrorModule::Kernel, 14};
 constexpr ResultCode ERR_INVALID_SIZE{ErrorModule::Kernel, 101};
 constexpr ResultCode ERR_INVALID_ADDRESS{ErrorModule::Kernel, 102};
-constexpr ResultCode ERR_OUT_OF_MEMORY{ErrorModule::Kernel, 104};
 constexpr ResultCode ERR_HANDLE_TABLE_FULL{ErrorModule::Kernel, 105};
 constexpr ResultCode ERR_INVALID_ADDRESS_STATE{ErrorModule::Kernel, 106};
 constexpr ResultCode ERR_INVALID_MEMORY_PERMISSIONS{ErrorModule::Kernel, 108};

src/core/hle/kernel/handle_table.cpp

@@ -14,47 +14,32 @@
 namespace Kernel {
 namespace {
 constexpr u16 GetSlot(Handle handle) {
-    return static_cast<u16>(handle >> 15);
+    return handle >> 15;
 }
 
 constexpr u16 GetGeneration(Handle handle) {
-    return static_cast<u16>(handle & 0x7FFF);
+    return handle & 0x7FFF;
 }
 } // Anonymous namespace
 
 HandleTable::HandleTable() {
+    next_generation = 1;
     Clear();
 }
 
 HandleTable::~HandleTable() = default;
 
-ResultCode HandleTable::SetSize(s32 handle_table_size) {
-    if (static_cast<u32>(handle_table_size) > MAX_COUNT) {
-        return ERR_OUT_OF_MEMORY;
-    }
-
-    // Values less than or equal to zero indicate to use the maximum allowable
-    // size for the handle table in the actual kernel, so we ignore the given
-    // value in that case, since we assume this by default unless this function
-    // is called.
-    if (handle_table_size > 0) {
-        table_size = static_cast<u16>(handle_table_size);
-    }
-
-    return RESULT_SUCCESS;
-}
-
 ResultVal<Handle> HandleTable::Create(SharedPtr<Object> obj) {
     DEBUG_ASSERT(obj != nullptr);
 
-    const u16 slot = next_free_slot;
-    if (slot >= table_size) {
+    u16 slot = next_free_slot;
+    if (slot >= generations.size()) {
         LOG_ERROR(Kernel, "Unable to allocate Handle, too many slots in use.");
         return ERR_HANDLE_TABLE_FULL;
     }
     next_free_slot = generations[slot];
 
-    const u16 generation = next_generation++;
+    u16 generation = next_generation++;
 
     // Overflow count so it fits in the 15 bits dedicated to the generation in the handle.
     // Horizon OS uses zero to represent an invalid handle, so skip to 1.
@@ -79,11 +64,10 @@ ResultVal<Handle> HandleTable::Duplicate(Handle handle) {
 }
 
 ResultCode HandleTable::Close(Handle handle) {
-    if (!IsValid(handle)) {
+    if (!IsValid(handle))
         return ERR_INVALID_HANDLE;
-    }
 
-    const u16 slot = GetSlot(handle);
+    u16 slot = GetSlot(handle);
 
     objects[slot] = nullptr;
 
@@ -93,10 +77,10 @@ ResultCode HandleTable::Close(Handle handle) {
 }
 
 bool HandleTable::IsValid(Handle handle) const {
-    const std::size_t slot = GetSlot(handle);
-    const u16 generation = GetGeneration(handle);
+    std::size_t slot = GetSlot(handle);
+    u16 generation = GetGeneration(handle);
 
-    return slot < table_size && objects[slot] != nullptr && generations[slot] == generation;
+    return slot < MAX_COUNT && objects[slot] != nullptr && generations[slot] == generation;
 }
 
 SharedPtr<Object> HandleTable::GetGeneric(Handle handle) const {
@@ -113,7 +97,7 @@ SharedPtr<Object> HandleTable::GetGeneric(Handle handle) const {
 }
 
 void HandleTable::Clear() {
-    for (u16 i = 0; i < table_size; ++i) {
+    for (u16 i = 0; i < MAX_COUNT; ++i) {
         generations[i] = i + 1;
         objects[i] = nullptr;
     }

src/core/hle/kernel/handle_table.h

@@ -49,20 +49,6 @@ public:
     HandleTable();
     ~HandleTable();
 
-    /**
-     * Sets the number of handles that may be in use at one time
-     * for this handle table.
-     *
-     * @param handle_table_size The desired size to limit the handle table to.
-     *
-     * @returns an error code indicating if initialization was successful.
-     *          If initialization was not successful, then ERR_OUT_OF_MEMORY
-     *          will be returned.
-     *
-     * @pre handle_table_size must be within the range [0, 1024]
-     */
-    ResultCode SetSize(s32 handle_table_size);
-
     /**
      * Allocates a handle for the given object.
      * @return The created Handle or one of the following errors:
@@ -117,21 +103,14 @@ private:
      */
     std::array<u16, MAX_COUNT> generations;
 
-    /**
-     * The limited size of the handle table. This can be specified by process
-     * capabilities in order to restrict the overall number of handles that
-     * can be created in a process instance
-     */
-    u16 table_size = static_cast<u16>(MAX_COUNT);
-
     /**
      * Global counter of the number of created handles. Stored in `generations` when a handle is
      * created, and wraps around to 1 when it hits 0x8000.
      */
-    u16 next_generation = 1;
+    u16 next_generation;
 
     /// Head of the free slots linked list.
-    u16 next_free_slot = 0;
+    u16 next_free_slot;
 };
 
 } // namespace Kernel

src/core/hle/kernel/kernel.cpp

@@ -86,11 +86,11 @@ static void ThreadWakeupCallback(u64 thread_handle, [[maybe_unused]] int cycles_
 }
 
 struct KernelCore::Impl {
-    void Initialize(KernelCore& kernel, Core::Timing::CoreTiming& core_timing) {
+    void Initialize(KernelCore& kernel) {
         Shutdown();
 
         InitializeSystemResourceLimit(kernel);
-        InitializeThreads(core_timing);
+        InitializeThreads();
     }
 
     void Shutdown() {
@@ -122,9 +122,9 @@ struct KernelCore::Impl {
         ASSERT(system_resource_limit->SetLimitValue(ResourceType::Sessions, 900).IsSuccess());
     }
 
-    void InitializeThreads(Core::Timing::CoreTiming& core_timing) {
+    void InitializeThreads() {
         thread_wakeup_event_type =
-            core_timing.RegisterEvent("ThreadWakeupCallback", ThreadWakeupCallback);
+            Core::Timing::RegisterEvent("ThreadWakeupCallback", ThreadWakeupCallback);
     }
 
     std::atomic<u32> next_object_id{0};
@@ -152,8 +152,8 @@ KernelCore::~KernelCore() {
     Shutdown();
 }
 
-void KernelCore::Initialize(Core::Timing::CoreTiming& core_timing) {
-    impl->Initialize(*this, core_timing);
+void KernelCore::Initialize() {
+    impl->Initialize(*this);
 }
 
 void KernelCore::Shutdown() {

src/core/hle/kernel/kernel.h

@@ -12,9 +12,8 @@ template <typename T>
 class ResultVal;
 
 namespace Core::Timing {
-class CoreTiming;
 struct EventType;
-} // namespace Core::Timing
+}
 
 namespace Kernel {
 
@@ -40,11 +39,7 @@ public:
     KernelCore& operator=(KernelCore&&) = delete;
 
     /// Resets the kernel to a clean slate for use.
-    ///
-    /// @param core_timing CoreTiming instance used to create any necessary
-    ///                    kernel-specific callback events.
-    ///
-    void Initialize(Core::Timing::CoreTiming& core_timing);
+    void Initialize();
 
     /// Clears all resources in use by the kernel instance.
     void Shutdown();

src/core/hle/kernel/process.cpp

@@ -99,13 +99,7 @@ ResultCode Process::LoadFromMetadata(const FileSys::ProgramMetadata& metadata) {
     vm_manager.Reset(metadata.GetAddressSpaceType());
 
     const auto& caps = metadata.GetKernelCapabilities();
-    const auto capability_init_result =
-        capabilities.InitializeForUserProcess(caps.data(), caps.size(), vm_manager);
-    if (capability_init_result.IsError()) {
-        return capability_init_result;
-    }
-
-    return handle_table.SetSize(capabilities.GetHandleTableSize());
+    return capabilities.InitializeForUserProcess(caps.data(), caps.size(), vm_manager);
 }
 
 void Process::Run(VAddr entry_point, s32 main_thread_priority, u32 stack_size) {

src/core/hle/kernel/process_capability.cpp

@@ -96,7 +96,7 @@ void ProcessCapabilities::InitializeForMetadatalessProcess() {
     interrupt_capabilities.set();
 
     // Allow using the maximum possible amount of handles
-    handle_table_size = static_cast<s32>(HandleTable::MAX_COUNT);
+    handle_table_size = static_cast<u32>(HandleTable::MAX_COUNT);
 
     // Allow all debugging capabilities.
     is_debuggable = true;
@@ -337,7 +337,7 @@ ResultCode ProcessCapabilities::HandleHandleTableFlags(u32 flags) {
         return ERR_RESERVED_VALUE;
     }
 
-    handle_table_size = static_cast<s32>((flags >> 16) & 0x3FF);
+    handle_table_size = (flags >> 16) & 0x3FF;
     return RESULT_SUCCESS;
 }
 

src/core/hle/kernel/process_capability.h

@@ -156,7 +156,7 @@ public:
     }
 
     /// Gets the number of total allowable handles for the process' handle table.
-    s32 GetHandleTableSize() const {
+    u32 GetHandleTableSize() const {
         return handle_table_size;
     }
 
@@ -252,7 +252,7 @@ private:
     u64 core_mask = 0;
     u64 priority_mask = 0;
 
-    s32 handle_table_size = 0;
+    u32 handle_table_size = 0;
     u32 kernel_version = 0;
 
     ProgramType program_type = ProgramType::SysModule;

src/core/hle/kernel/scheduler.cpp

@@ -111,7 +111,7 @@ void Scheduler::SwitchContext(Thread* new_thread) {
 
 void Scheduler::UpdateLastContextSwitchTime(Thread* thread, Process* process) {
     const u64 prev_switch_ticks = last_context_switch_time;
-    const u64 most_recent_switch_ticks = Core::System::GetInstance().CoreTiming().GetTicks();
+    const u64 most_recent_switch_ticks = Core::Timing::GetTicks();
     const u64 update_ticks = most_recent_switch_ticks - prev_switch_ticks;
 
     if (thread != nullptr) {

src/core/hle/kernel/svc.cpp

@@ -918,7 +918,6 @@ static ResultCode GetInfo(u64* result, u64 info_id, u64 handle, u64 info_sub_id)
         }
 
         const auto& system = Core::System::GetInstance();
-        const auto& core_timing = system.CoreTiming();
         const auto& scheduler = system.CurrentScheduler();
         const auto* const current_thread = scheduler.GetCurrentThread();
         const bool same_thread = current_thread == thread;
@@ -928,9 +927,9 @@ static ResultCode GetInfo(u64* result, u64 info_id, u64 handle, u64 info_sub_id)
         if (same_thread && info_sub_id == 0xFFFFFFFFFFFFFFFF) {
             const u64 thread_ticks = current_thread->GetTotalCPUTimeTicks();
 
-            out_ticks = thread_ticks + (core_timing.GetTicks() - prev_ctx_ticks);
+            out_ticks = thread_ticks + (Core::Timing::GetTicks() - prev_ctx_ticks);
         } else if (same_thread && info_sub_id == system.CurrentCoreIndex()) {
-            out_ticks = core_timing.GetTicks() - prev_ctx_ticks;
+            out_ticks = Core::Timing::GetTicks() - prev_ctx_ticks;
         }
 
         *result = out_ticks;
@@ -1547,11 +1546,10 @@ static ResultCode SignalToAddress(VAddr address, u32 type, s32 value, s32 num_to
 static u64 GetSystemTick() {
     LOG_TRACE(Kernel_SVC, "called");
 
-    auto& core_timing = Core::System::GetInstance().CoreTiming();
-    const u64 result{core_timing.GetTicks()};
+    const u64 result{Core::Timing::GetTicks()};
 
     // Advance time to defeat dumb games that busy-wait for the frame to end.
-    core_timing.AddTicks(400);
+    Core::Timing::AddTicks(400);
 
     return result;
 }

src/core/hle/kernel/thread.cpp

@@ -43,8 +43,7 @@ Thread::~Thread() = default;
 
 void Thread::Stop() {
     // Cancel any outstanding wakeup events for this thread
-    Core::System::GetInstance().CoreTiming().UnscheduleEvent(kernel.ThreadWakeupCallbackEventType(),
-                                                             callback_handle);
+    Core::Timing::UnscheduleEvent(kernel.ThreadWakeupCallbackEventType(), callback_handle);
     kernel.ThreadWakeupCallbackHandleTable().Close(callback_handle);
     callback_handle = 0;
 
@@ -86,14 +85,13 @@ void Thread::WakeAfterDelay(s64 nanoseconds) {
 
     // This function might be called from any thread so we have to be cautious and use the
     // thread-safe version of ScheduleEvent.
-    Core::System::GetInstance().CoreTiming().ScheduleEventThreadsafe(
-        Core::Timing::nsToCycles(nanoseconds), kernel.ThreadWakeupCallbackEventType(),
-        callback_handle);
+    Core::Timing::ScheduleEventThreadsafe(Core::Timing::nsToCycles(nanoseconds),
+                                          kernel.ThreadWakeupCallbackEventType(), callback_handle);
 }
 
 void Thread::CancelWakeupTimer() {
-    Core::System::GetInstance().CoreTiming().UnscheduleEventThreadsafe(
-        kernel.ThreadWakeupCallbackEventType(), callback_handle);
+    Core::Timing::UnscheduleEventThreadsafe(kernel.ThreadWakeupCallbackEventType(),
+                                            callback_handle);
 }
 
 static std::optional<s32> GetNextProcessorId(u64 mask) {
@@ -192,7 +190,6 @@ ResultVal<SharedPtr<Thread>> Thread::Create(KernelCore& kernel, std::string name
         return ResultCode(-1);
     }
 
-    auto& system = Core::System::GetInstance();
     SharedPtr<Thread> thread(new Thread(kernel));
 
     thread->thread_id = kernel.CreateNewThreadID();
@@ -201,7 +198,7 @@ ResultVal<SharedPtr<Thread>> Thread::Create(KernelCore& kernel, std::string name
     thread->stack_top = stack_top;
     thread->tpidr_el0 = 0;
     thread->nominal_priority = thread->current_priority = priority;
-    thread->last_running_ticks = system.CoreTiming().GetTicks();
+    thread->last_running_ticks = Core::Timing::GetTicks();
     thread->processor_id = processor_id;
     thread->ideal_core = processor_id;
     thread->affinity_mask = 1ULL << processor_id;
@@ -212,7 +209,7 @@ ResultVal<SharedPtr<Thread>> Thread::Create(KernelCore& kernel, std::string name
     thread->name = std::move(name);
     thread->callback_handle = kernel.ThreadWakeupCallbackHandleTable().Create(thread).Unwrap();
     thread->owner_process = &owner_process;
-    thread->scheduler = &system.Scheduler(processor_id);
+    thread->scheduler = &Core::System::GetInstance().Scheduler(processor_id);
     thread->scheduler->AddThread(thread, priority);
     thread->tls_address = thread->owner_process->MarkNextAvailableTLSSlotAsUsed(*thread);
 
@@ -261,7 +258,7 @@ void Thread::SetStatus(ThreadStatus new_status) {
     }
 
     if (status == ThreadStatus::Running) {
-        last_running_ticks = Core::System::GetInstance().CoreTiming().GetTicks();
+        last_running_ticks = Core::Timing::GetTicks();
     }
 
     status = new_status;

src/core/hle/service/audio/audout_u.cpp

@@ -68,12 +68,12 @@ public:
         RegisterHandlers(functions);
 
         // This is the event handle used to check if the audio buffer was released
-        auto& system = Core::System::GetInstance();
-        buffer_event = Kernel::WritableEvent::CreateEventPair(
-            system.Kernel(), Kernel::ResetType::Sticky, "IAudioOutBufferReleased");
+        auto& kernel = Core::System::GetInstance().Kernel();
+        buffer_event = Kernel::WritableEvent::CreateEventPair(kernel, Kernel::ResetType::Sticky,
+                                                              "IAudioOutBufferReleased");
 
-        stream = audio_core.OpenStream(system.CoreTiming(), audio_params.sample_rate,
-                                       audio_params.channel_count, std::move(unique_name),
+        stream = audio_core.OpenStream(audio_params.sample_rate, audio_params.channel_count,
+                                       std::move(unique_name),
                                        [=]() { buffer_event.writable->Signal(); });
     }
 

src/core/hle/service/audio/audren_u.cpp

@@ -37,16 +37,15 @@ public:
             {8, &IAudioRenderer::SetRenderingTimeLimit, "SetRenderingTimeLimit"},
             {9, &IAudioRenderer::GetRenderingTimeLimit, "GetRenderingTimeLimit"},
             {10, &IAudioRenderer::RequestUpdateImpl, "RequestUpdateAuto"},
-            {11, &IAudioRenderer::ExecuteAudioRendererRendering, "ExecuteAudioRendererRendering"},
+            {11, nullptr, "ExecuteAudioRendererRendering"},
         };
         // clang-format on
         RegisterHandlers(functions);
 
-        auto& system = Core::System::GetInstance();
-        system_event = Kernel::WritableEvent::CreateEventPair(
-            system.Kernel(), Kernel::ResetType::Sticky, "IAudioRenderer:SystemEvent");
-        renderer = std::make_unique<AudioCore::AudioRenderer>(system.CoreTiming(), audren_params,
-                                                              system_event.writable);
+        auto& kernel = Core::System::GetInstance().Kernel();
+        system_event = Kernel::WritableEvent::CreateEventPair(kernel, Kernel::ResetType::Sticky,
+                                                              "IAudioRenderer:SystemEvent");
+        renderer = std::make_unique<AudioCore::AudioRenderer>(audren_params, system_event.writable);
     }
 
 private:
@@ -138,17 +137,6 @@ private:
         rb.Push(rendering_time_limit_percent);
     }
 
-    void ExecuteAudioRendererRendering(Kernel::HLERequestContext& ctx) {
-        LOG_DEBUG(Service_Audio, "called");
-
-        // This service command currently only reports an unsupported operation
-        // error code, or aborts. Given that, we just always return an error
-        // code in this case.
-
-        IPC::ResponseBuilder rb{ctx, 2};
-        rb.Push(ResultCode{ErrorModule::Audio, 201});
-    }
-
     Kernel::EventPair system_event;
     std::unique_ptr<AudioCore::AudioRenderer> renderer;
     u32 rendering_time_limit_percent = 100;
@@ -246,7 +234,7 @@ AudRenU::AudRenU() : ServiceFramework("audren:u") {
         {0, &AudRenU::OpenAudioRenderer, "OpenAudioRenderer"},
         {1, &AudRenU::GetAudioRendererWorkBufferSize, "GetAudioRendererWorkBufferSize"},
         {2, &AudRenU::GetAudioDeviceService, "GetAudioDeviceService"},
-        {3, &AudRenU::OpenAudioRendererAuto, "OpenAudioRendererAuto"},
+        {3, nullptr, "OpenAudioRendererAuto"},
         {4, &AudRenU::GetAudioDeviceServiceWithRevisionInfo, "GetAudioDeviceServiceWithRevisionInfo"},
     };
     // clang-format on
@@ -259,7 +247,12 @@ AudRenU::~AudRenU() = default;
 void AudRenU::OpenAudioRenderer(Kernel::HLERequestContext& ctx) {
     LOG_DEBUG(Service_Audio, "called");
 
-    OpenAudioRendererImpl(ctx);
+    IPC::RequestParser rp{ctx};
+    auto params = rp.PopRaw<AudioCore::AudioRendererParameter>();
+    IPC::ResponseBuilder rb{ctx, 2, 0, 1};
+
+    rb.Push(RESULT_SUCCESS);
+    rb.PushIpcInterface<Audio::IAudioRenderer>(std::move(params));
 }
 
 void AudRenU::GetAudioRendererWorkBufferSize(Kernel::HLERequestContext& ctx) {
@@ -268,20 +261,20 @@ void AudRenU::GetAudioRendererWorkBufferSize(Kernel::HLERequestContext& ctx) {
     LOG_DEBUG(Service_Audio, "called");
 
     u64 buffer_sz = Common::AlignUp(4 * params.mix_buffer_count, 0x40);
-    buffer_sz += params.submix_count * 1024;
-    buffer_sz += 0x940 * (params.submix_count + 1);
+    buffer_sz += params.unknown_c * 1024;
+    buffer_sz += 0x940 * (params.unknown_c + 1);
     buffer_sz += 0x3F0 * params.voice_count;
-    buffer_sz += Common::AlignUp(8 * (params.submix_count + 1), 0x10);
+    buffer_sz += Common::AlignUp(8 * (params.unknown_c + 1), 0x10);
     buffer_sz += Common::AlignUp(8 * params.voice_count, 0x10);
-    buffer_sz += Common::AlignUp(
-        (0x3C0 * (params.sink_count + params.submix_count) + 4 * params.sample_count) *
-            (params.mix_buffer_count + 6),
-        0x40);
+    buffer_sz +=
+        Common::AlignUp((0x3C0 * (params.sink_count + params.unknown_c) + 4 * params.sample_count) *
+                            (params.mix_buffer_count + 6),
+                        0x40);
 
     if (IsFeatureSupported(AudioFeatures::Splitter, params.revision)) {
-        const u32 count = params.submix_count + 1;
+        u32 count = params.unknown_c + 1;
         u64 node_count = Common::AlignUp(count, 0x40);
-        const u64 node_state_buffer_sz =
+        u64 node_state_buffer_sz =
             4 * (node_count * node_count) + 0xC * node_count + 2 * (node_count / 8);
         u64 edge_matrix_buffer_sz = 0;
         node_count = Common::AlignUp(count * count, 0x40);
@@ -295,19 +288,19 @@ void AudRenU::GetAudioRendererWorkBufferSize(Kernel::HLERequestContext& ctx) {
 
     buffer_sz += 0x20 * (params.effect_count + 4 * params.voice_count) + 0x50;
     if (IsFeatureSupported(AudioFeatures::Splitter, params.revision)) {
-        buffer_sz += 0xE0 * params.num_splitter_send_channels;
+        buffer_sz += 0xE0 * params.unknown_2c;
         buffer_sz += 0x20 * params.splitter_count;
-        buffer_sz += Common::AlignUp(4 * params.num_splitter_send_channels, 0x10);
+        buffer_sz += Common::AlignUp(4 * params.unknown_2c, 0x10);
     }
     buffer_sz = Common::AlignUp(buffer_sz, 0x40) + 0x170 * params.sink_count;
     u64 output_sz = buffer_sz + 0x280 * params.sink_count + 0x4B0 * params.effect_count +
                     ((params.voice_count * 256) | 0x40);
 
-    if (params.performance_frame_count >= 1) {
+    if (params.unknown_1c >= 1) {
         output_sz = Common::AlignUp(((16 * params.sink_count + 16 * params.effect_count +
                                       16 * params.voice_count + 16) +
                                      0x658) *
-                                            (params.performance_frame_count + 1) +
+                                            (params.unknown_1c + 1) +
                                         0xc0,
                                     0x40) +
                     output_sz;
@@ -331,12 +324,6 @@ void AudRenU::GetAudioDeviceService(Kernel::HLERequestContext& ctx) {
     rb.PushIpcInterface<Audio::IAudioDevice>();
 }
 
-void AudRenU::OpenAudioRendererAuto(Kernel::HLERequestContext& ctx) {
-    LOG_DEBUG(Service_Audio, "called");
-
-    OpenAudioRendererImpl(ctx);
-}
-
 void AudRenU::GetAudioDeviceServiceWithRevisionInfo(Kernel::HLERequestContext& ctx) {
     LOG_WARNING(Service_Audio, "(STUBBED) called");
 
@@ -347,15 +334,6 @@ void AudRenU::GetAudioDeviceServiceWithRevisionInfo(Kernel::HLERequestContext& c
                                                 // based on the current revision
 }
 
-void AudRenU::OpenAudioRendererImpl(Kernel::HLERequestContext& ctx) {
-    IPC::RequestParser rp{ctx};
-    const auto params = rp.PopRaw<AudioCore::AudioRendererParameter>();
-    IPC::ResponseBuilder rb{ctx, 2, 0, 1};
-
-    rb.Push(RESULT_SUCCESS);
-    rb.PushIpcInterface<IAudioRenderer>(params);
-}
-
 bool AudRenU::IsFeatureSupported(AudioFeatures feature, u32_le revision) const {
     u32_be version_num = (revision - Common::MakeMagic('R', 'E', 'V', '0')); // Byte swap
     switch (feature) {

src/core/hle/service/audio/audren_u.h

@@ -21,11 +21,8 @@ private:
     void OpenAudioRenderer(Kernel::HLERequestContext& ctx);
     void GetAudioRendererWorkBufferSize(Kernel::HLERequestContext& ctx);
     void GetAudioDeviceService(Kernel::HLERequestContext& ctx);
-    void OpenAudioRendererAuto(Kernel::HLERequestContext& ctx);
     void GetAudioDeviceServiceWithRevisionInfo(Kernel::HLERequestContext& ctx);
 
-    void OpenAudioRendererImpl(Kernel::HLERequestContext& ctx);
-
     enum class AudioFeatures : u32 {
         Splitter,
     };

src/core/hle/service/hid/controllers/controller_base.h

@@ -7,10 +7,6 @@
 #include "common/common_types.h"
 #include "common/swap.h"
 
-namespace Core::Timing {
-class CoreTiming;
-}
-
 namespace Service::HID {
 class ControllerBase {
 public:
@@ -24,8 +20,7 @@ public:
     virtual void OnRelease() = 0;
 
     // When the controller is requesting an update for the shared memory
-    virtual void OnUpdate(const Core::Timing::CoreTiming& core_timing, u8* data,
-                          std::size_t size) = 0;
+    virtual void OnUpdate(u8* data, std::size_t size) = 0;
 
     // Called when input devices should be loaded
     virtual void OnLoadInputDevices() = 0;

src/core/hle/service/hid/controllers/debug_pad.cpp

@@ -21,9 +21,8 @@ void Controller_DebugPad::OnInit() {}
 
 void Controller_DebugPad::OnRelease() {}
 
-void Controller_DebugPad::OnUpdate(const Core::Timing::CoreTiming& core_timing, u8* data,
-                                   std::size_t size) {
-    shared_memory.header.timestamp = core_timing.GetTicks();
+void Controller_DebugPad::OnUpdate(u8* data, std::size_t size) {
+    shared_memory.header.timestamp = Core::Timing::GetTicks();
     shared_memory.header.total_entry_count = 17;
 
     if (!IsControllerActivated()) {

src/core/hle/service/hid/controllers/debug_pad.h

@@ -26,7 +26,7 @@ public:
     void OnRelease() override;
 
     // When the controller is requesting an update for the shared memory
-    void OnUpdate(const Core::Timing::CoreTiming& core_timing, u8* data, std::size_t size) override;
+    void OnUpdate(u8* data, std::size_t size) override;
 
     // Called when input devices should be loaded
     void OnLoadInputDevices() override;

src/core/hle/service/hid/controllers/gesture.cpp

@@ -17,9 +17,8 @@ void Controller_Gesture::OnInit() {}
 
 void Controller_Gesture::OnRelease() {}
 
-void Controller_Gesture::OnUpdate(const Core::Timing::CoreTiming& core_timing, u8* data,
-                                  std::size_t size) {
-    shared_memory.header.timestamp = core_timing.GetTicks();
+void Controller_Gesture::OnUpdate(u8* data, std::size_t size) {
+    shared_memory.header.timestamp = Core::Timing::GetTicks();
     shared_memory.header.total_entry_count = 17;
 
     if (!IsControllerActivated()) {

src/core/hle/service/hid/controllers/gesture.h

@@ -22,7 +22,7 @@ public:
     void OnRelease() override;
 
     // When the controller is requesting an update for the shared memory
-    void OnUpdate(const Core::Timing::CoreTiming& core_timing, u8* data, size_t size) override;
+    void OnUpdate(u8* data, size_t size) override;
 
     // Called when input devices should be loaded
     void OnLoadInputDevices() override;

src/core/hle/service/hid/controllers/keyboard.cpp

@@ -19,9 +19,8 @@ void Controller_Keyboard::OnInit() {}
 
 void Controller_Keyboard::OnRelease() {}
 
-void Controller_Keyboard::OnUpdate(const Core::Timing::CoreTiming& core_timing, u8* data,
-                                   std::size_t size) {
-    shared_memory.header.timestamp = core_timing.GetTicks();
+void Controller_Keyboard::OnUpdate(u8* data, std::size_t size) {
+    shared_memory.header.timestamp = Core::Timing::GetTicks();
     shared_memory.header.total_entry_count = 17;
 
     if (!IsControllerActivated()) {

src/core/hle/service/hid/controllers/keyboard.h

@@ -25,7 +25,7 @@ public:
     void OnRelease() override;
 
     // When the controller is requesting an update for the shared memory
-    void OnUpdate(const Core::Timing::CoreTiming& core_timing, u8* data, std::size_t size) override;
+    void OnUpdate(u8* data, std::size_t size) override;
 
     // Called when input devices should be loaded
     void OnLoadInputDevices() override;

src/core/hle/service/hid/controllers/mouse.cpp

@@ -17,9 +17,8 @@ Controller_Mouse::~Controller_Mouse() = default;
 void Controller_Mouse::OnInit() {}
 void Controller_Mouse::OnRelease() {}
 
-void Controller_Mouse::OnUpdate(const Core::Timing::CoreTiming& core_timing, u8* data,
-                                std::size_t size) {
-    shared_memory.header.timestamp = core_timing.GetTicks();
+void Controller_Mouse::OnUpdate(u8* data, std::size_t size) {
+    shared_memory.header.timestamp = Core::Timing::GetTicks();
     shared_memory.header.total_entry_count = 17;
 
     if (!IsControllerActivated()) {

src/core/hle/service/hid/controllers/mouse.h

@@ -24,7 +24,7 @@ public:
     void OnRelease() override;
 
     // When the controller is requesting an update for the shared memory
-    void OnUpdate(const Core::Timing::CoreTiming& core_timing, u8* data, std::size_t size) override;
+    void OnUpdate(u8* data, std::size_t size) override;
 
     // Called when input devices should be loaded
     void OnLoadInputDevices() override;

src/core/hle/service/hid/controllers/npad.cpp

@@ -288,8 +288,7 @@ void Controller_NPad::RequestPadStateUpdate(u32 npad_id) {
     rstick_entry.y = static_cast<s32>(stick_r_y_f * HID_JOYSTICK_MAX);
 }
 
-void Controller_NPad::OnUpdate(const Core::Timing::CoreTiming& core_timing, u8* data,
-                               std::size_t data_len) {
+void Controller_NPad::OnUpdate(u8* data, std::size_t data_len) {
     if (!IsControllerActivated())
         return;
     for (std::size_t i = 0; i < shared_memory_entries.size(); i++) {
@@ -309,7 +308,7 @@ void Controller_NPad::OnUpdate(const Core::Timing::CoreTiming& core_timing, u8*
             const auto& last_entry =
                 main_controller->npad[main_controller->common.last_entry_index];
 
-            main_controller->common.timestamp = core_timing.GetTicks();
+            main_controller->common.timestamp = Core::Timing::GetTicks();
             main_controller->common.last_entry_index =
                 (main_controller->common.last_entry_index + 1) % 17;
 

src/core/hle/service/hid/controllers/npad.h

@@ -30,7 +30,7 @@ public:
     void OnRelease() override;
 
     // When the controller is requesting an update for the shared memory
-    void OnUpdate(const Core::Timing::CoreTiming& core_timing, u8* data, std::size_t size) override;
+    void OnUpdate(u8* data, std::size_t size) override;
 
     // Called when input devices should be loaded
     void OnLoadInputDevices() override;

src/core/hle/service/hid/controllers/stubbed.cpp

@@ -16,14 +16,13 @@ void Controller_Stubbed::OnInit() {}
 
 void Controller_Stubbed::OnRelease() {}
 
-void Controller_Stubbed::OnUpdate(const Core::Timing::CoreTiming& core_timing, u8* data,
-                                  std::size_t size) {
+void Controller_Stubbed::OnUpdate(u8* data, std::size_t size) {
     if (!smart_update) {
         return;
     }
 
     CommonHeader header{};
-    header.timestamp = core_timing.GetTicks();
+    header.timestamp = Core::Timing::GetTicks();
     header.total_entry_count = 17;
     header.entry_count = 0;
     header.last_entry_index = 0;

src/core/hle/service/hid/controllers/stubbed.h

@@ -20,7 +20,7 @@ public:
     void OnRelease() override;
 
     // When the controller is requesting an update for the shared memory
-    void OnUpdate(const Core::Timing::CoreTiming& core_timing, u8* data, std::size_t size) override;
+    void OnUpdate(u8* data, std::size_t size) override;
 
     // Called when input devices should be loaded
     void OnLoadInputDevices() override;

src/core/hle/service/hid/controllers/touchscreen.cpp

@@ -20,9 +20,8 @@ void Controller_Touchscreen::OnInit() {}
 
 void Controller_Touchscreen::OnRelease() {}
 
-void Controller_Touchscreen::OnUpdate(const Core::Timing::CoreTiming& core_timing, u8* data,
-                                      std::size_t size) {
-    shared_memory.header.timestamp = core_timing.GetTicks();
+void Controller_Touchscreen::OnUpdate(u8* data, std::size_t size) {
+    shared_memory.header.timestamp = Core::Timing::GetTicks();
     shared_memory.header.total_entry_count = 17;
 
     if (!IsControllerActivated()) {
@@ -49,7 +48,7 @@ void Controller_Touchscreen::OnUpdate(const Core::Timing::CoreTiming& core_timin
         touch_entry.diameter_x = Settings::values.touchscreen.diameter_x;
         touch_entry.diameter_y = Settings::values.touchscreen.diameter_y;
         touch_entry.rotation_angle = Settings::values.touchscreen.rotation_angle;
-        const u64 tick = core_timing.GetTicks();
+        const u64 tick = Core::Timing::GetTicks();
         touch_entry.delta_time = tick - last_touch;
         last_touch = tick;
         touch_entry.finger = Settings::values.touchscreen.finger;

src/core/hle/service/hid/controllers/touchscreen.h

@@ -24,7 +24,7 @@ public:
     void OnRelease() override;
 
     // When the controller is requesting an update for the shared memory
-    void OnUpdate(const Core::Timing::CoreTiming& core_timing, u8* data, std::size_t size) override;
+    void OnUpdate(u8* data, std::size_t size) override;
 
     // Called when input devices should be loaded
     void OnLoadInputDevices() override;

src/core/hle/service/hid/controllers/xpad.cpp

@@ -17,10 +17,9 @@ void Controller_XPad::OnInit() {}
 
 void Controller_XPad::OnRelease() {}
 
-void Controller_XPad::OnUpdate(const Core::Timing::CoreTiming& core_timing, u8* data,
-                               std::size_t size) {
+void Controller_XPad::OnUpdate(u8* data, std::size_t size) {
     for (auto& xpad_entry : shared_memory.shared_memory_entries) {
-        xpad_entry.header.timestamp = core_timing.GetTicks();
+        xpad_entry.header.timestamp = Core::Timing::GetTicks();
         xpad_entry.header.total_entry_count = 17;
 
         if (!IsControllerActivated()) {

src/core/hle/service/hid/controllers/xpad.h

@@ -22,7 +22,7 @@ public:
     void OnRelease() override;
 
     // When the controller is requesting an update for the shared memory
-    void OnUpdate(const Core::Timing::CoreTiming& core_timing, u8* data, std::size_t size) override;
+    void OnUpdate(u8* data, std::size_t size) override;
 
     // Called when input devices should be loaded
     void OnLoadInputDevices() override;

src/core/hle/service/hid/hid.cpp

@@ -73,15 +73,13 @@ IAppletResource::IAppletResource() : ServiceFramework("IAppletResource") {
     GetController<Controller_Stubbed>(HidController::Unknown3).SetCommonHeaderOffset(0x5000);
 
     // Register update callbacks
-    auto& core_timing = Core::System::GetInstance().CoreTiming();
-    pad_update_event =
-        core_timing.RegisterEvent("HID::UpdatePadCallback", [this](u64 userdata, int cycles_late) {
-            UpdateControllers(userdata, cycles_late);
-        });
+    pad_update_event = Core::Timing::RegisterEvent(
+        "HID::UpdatePadCallback",
+        [this](u64 userdata, int cycles_late) { UpdateControllers(userdata, cycles_late); });
 
     // TODO(shinyquagsire23): Other update callbacks? (accel, gyro?)
 
-    core_timing.ScheduleEvent(pad_update_ticks, pad_update_event);
+    Core::Timing::ScheduleEvent(pad_update_ticks, pad_update_event);
 
     ReloadInputDevices();
 }
@@ -95,7 +93,7 @@ void IAppletResource::DeactivateController(HidController controller) {
 }
 
 IAppletResource ::~IAppletResource() {
-    Core::System::GetInstance().CoreTiming().UnscheduleEvent(pad_update_event, 0);
+    Core::Timing::UnscheduleEvent(pad_update_event, 0);
 }
 
 void IAppletResource::GetSharedMemoryHandle(Kernel::HLERequestContext& ctx) {
@@ -107,17 +105,15 @@ void IAppletResource::GetSharedMemoryHandle(Kernel::HLERequestContext& ctx) {
 }
 
 void IAppletResource::UpdateControllers(u64 userdata, int cycles_late) {
-    auto& core_timing = Core::System::GetInstance().CoreTiming();
-
     const bool should_reload = Settings::values.is_device_reload_pending.exchange(false);
     for (const auto& controller : controllers) {
         if (should_reload) {
             controller->OnLoadInputDevices();
         }
-        controller->OnUpdate(core_timing, shared_mem->GetPointer(), SHARED_MEMORY_SIZE);
+        controller->OnUpdate(shared_mem->GetPointer(), SHARED_MEMORY_SIZE);
     }
 
-    core_timing.ScheduleEvent(pad_update_ticks - cycles_late, pad_update_event);
+    Core::Timing::ScheduleEvent(pad_update_ticks - cycles_late, pad_update_event);
 }
 
 class IActiveVibrationDeviceList final : public ServiceFramework<IActiveVibrationDeviceList> {

src/core/hle/service/hid/hid.h

@@ -15,7 +15,7 @@ namespace Kernel {
 class SharedMemory;
 }
 
-namespace Service::SM {
+namespace SM {
 class ServiceManager;
 }
 

src/core/hle/service/hid/irs.cpp

@@ -98,7 +98,7 @@ void IRS::GetImageTransferProcessorState(Kernel::HLERequestContext& ctx) {
 
     IPC::ResponseBuilder rb{ctx, 5};
     rb.Push(RESULT_SUCCESS);
-    rb.PushRaw<u64>(Core::System::GetInstance().CoreTiming().GetTicks());
+    rb.PushRaw<u64>(Core::Timing::GetTicks());
     rb.PushRaw<u32>(0);
 }
 

src/core/hle/service/nvdrv/devices/nvdisp_disp0.cpp

@@ -23,7 +23,7 @@ u32 nvdisp_disp0::ioctl(Ioctl command, const std::vector<u8>& input, std::vector
 
 void nvdisp_disp0::flip(u32 buffer_handle, u32 offset, u32 format, u32 width, u32 height,
                         u32 stride, NVFlinger::BufferQueue::BufferTransformFlags transform,
-                        const Common::Rectangle<int>& crop_rect) {
+                        const MathUtil::Rectangle<int>& crop_rect) {
     VAddr addr = nvmap_dev->GetObjectAddress(buffer_handle);
     LOG_TRACE(Service,
               "Drawing from address {:X} offset {:08X} Width {} Height {} Stride {} Format {}",

src/core/hle/service/nvdrv/devices/nvdisp_disp0.h

@@ -25,7 +25,7 @@ public:
     /// Performs a screen flip, drawing the buffer pointed to by the handle.
     void flip(u32 buffer_handle, u32 offset, u32 format, u32 width, u32 height, u32 stride,
               NVFlinger::BufferQueue::BufferTransformFlags transform,
-              const Common::Rectangle<int>& crop_rect);
+              const MathUtil::Rectangle<int>& crop_rect);
 
 private:
     std::shared_ptr<nvmap> nvmap_dev;

src/core/hle/service/nvdrv/devices/nvhost_ctrl_gpu.cpp

@@ -5,7 +5,6 @@
 #include <cstring>
 #include "common/assert.h"
 #include "common/logging/log.h"
-#include "core/core.h"
 #include "core/core_timing.h"
 #include "core/core_timing_util.h"
 #include "core/hle/service/nvdrv/devices/nvhost_ctrl_gpu.h"
@@ -185,7 +184,7 @@ u32 nvhost_ctrl_gpu::GetGpuTime(const std::vector<u8>& input, std::vector<u8>& o
 
     IoctlGetGpuTime params{};
     std::memcpy(&params, input.data(), input.size());
-    params.gpu_time = Core::Timing::cyclesToNs(Core::System::GetInstance().CoreTiming().GetTicks());
+    params.gpu_time = Core::Timing::cyclesToNs(Core::Timing::GetTicks());
     std::memcpy(output.data(), &params, output.size());
     return 0;
 }

src/core/hle/service/nvflinger/buffer_queue.cpp

@@ -63,7 +63,7 @@ const IGBPBuffer& BufferQueue::RequestBuffer(u32 slot) const {
 }
 
 void BufferQueue::QueueBuffer(u32 slot, BufferTransformFlags transform,
-                              const Common::Rectangle<int>& crop_rect) {
+                              const MathUtil::Rectangle<int>& crop_rect) {
     auto itr = std::find_if(queue.begin(), queue.end(),
                             [&](const Buffer& buffer) { return buffer.slot == slot; });
     ASSERT(itr != queue.end());

src/core/hle/service/nvflinger/buffer_queue.h

@@ -67,14 +67,14 @@ public:
         Status status = Status::Free;
         IGBPBuffer igbp_buffer;
         BufferTransformFlags transform;
-        Common::Rectangle<int> crop_rect;
+        MathUtil::Rectangle<int> crop_rect;
     };
 
     void SetPreallocatedBuffer(u32 slot, const IGBPBuffer& igbp_buffer);
     std::optional<u32> DequeueBuffer(u32 width, u32 height);
     const IGBPBuffer& RequestBuffer(u32 slot) const;
     void QueueBuffer(u32 slot, BufferTransformFlags transform,
-                     const Common::Rectangle<int>& crop_rect);
+                     const MathUtil::Rectangle<int>& crop_rect);
     std::optional<std::reference_wrapper<const Buffer>> AcquireBuffer();
     void ReleaseBuffer(u32 slot);
     u32 Query(QueryType type);

src/core/hle/service/nvflinger/nvflinger.cpp

@@ -14,12 +14,11 @@
 #include "core/core_timing_util.h"
 #include "core/hle/kernel/kernel.h"
 #include "core/hle/kernel/readable_event.h"
+#include "core/hle/kernel/writable_event.h"
 #include "core/hle/service/nvdrv/devices/nvdisp_disp0.h"
 #include "core/hle/service/nvdrv/nvdrv.h"
 #include "core/hle/service/nvflinger/buffer_queue.h"
 #include "core/hle/service/nvflinger/nvflinger.h"
-#include "core/hle/service/vi/display/vi_display.h"
-#include "core/hle/service/vi/layer/vi_layer.h"
 #include "core/perf_stats.h"
 #include "video_core/renderer_base.h"
 
@@ -28,25 +27,19 @@ namespace Service::NVFlinger {
 constexpr std::size_t SCREEN_REFRESH_RATE = 60;
 constexpr u64 frame_ticks = static_cast<u64>(Core::Timing::BASE_CLOCK_RATE / SCREEN_REFRESH_RATE);
 
-NVFlinger::NVFlinger(Core::Timing::CoreTiming& core_timing) : core_timing{core_timing} {
-    displays.emplace_back(0, "Default");
-    displays.emplace_back(1, "External");
-    displays.emplace_back(2, "Edid");
-    displays.emplace_back(3, "Internal");
-    displays.emplace_back(4, "Null");
-
+NVFlinger::NVFlinger() {
     // Schedule the screen composition events
     composition_event =
-        core_timing.RegisterEvent("ScreenComposition", [this](u64 userdata, int cycles_late) {
+        Core::Timing::RegisterEvent("ScreenComposition", [this](u64 userdata, int cycles_late) {
             Compose();
-            this->core_timing.ScheduleEvent(frame_ticks - cycles_late, composition_event);
+            Core::Timing::ScheduleEvent(frame_ticks - cycles_late, composition_event);
         });
 
-    core_timing.ScheduleEvent(frame_ticks, composition_event);
+    Core::Timing::ScheduleEvent(frame_ticks, composition_event);
 }
 
 NVFlinger::~NVFlinger() {
-    core_timing.UnscheduleEvent(composition_event, 0);
+    Core::Timing::UnscheduleEvent(composition_event, 0);
 }
 
 void NVFlinger::SetNVDrvInstance(std::shared_ptr<Nvidia::Module> instance) {
@@ -59,14 +52,13 @@ std::optional<u64> NVFlinger::OpenDisplay(std::string_view name) {
     // TODO(Subv): Currently we only support the Default display.
     ASSERT(name == "Default");
 
-    const auto itr =
-        std::find_if(displays.begin(), displays.end(),
-                     [&](const VI::Display& display) { return display.GetName() == name; });
+    const auto itr = std::find_if(displays.begin(), displays.end(),
+                                  [&](const Display& display) { return display.name == name; });
     if (itr == displays.end()) {
         return {};
     }
 
-    return itr->GetID();
+    return itr->id;
 }
 
 std::optional<u64> NVFlinger::CreateLayer(u64 display_id) {
@@ -76,10 +68,13 @@ std::optional<u64> NVFlinger::CreateLayer(u64 display_id) {
         return {};
     }
 
+    ASSERT_MSG(display->layers.empty(), "Only one layer is supported per display at the moment");
+
     const u64 layer_id = next_layer_id++;
     const u32 buffer_queue_id = next_buffer_queue_id++;
-    buffer_queues.emplace_back(buffer_queue_id, layer_id);
-    display->CreateLayer(layer_id, buffer_queues.back());
+    auto buffer_queue = std::make_shared<BufferQueue>(buffer_queue_id, layer_id);
+    display->layers.emplace_back(layer_id, buffer_queue);
+    buffer_queues.emplace_back(std::move(buffer_queue));
     return layer_id;
 }
 
@@ -90,7 +85,7 @@ std::optional<u32> NVFlinger::FindBufferQueueId(u64 display_id, u64 layer_id) co
         return {};
     }
 
-    return layer->GetBufferQueue().GetId();
+    return layer->buffer_queue->GetId();
 }
 
 Kernel::SharedPtr<Kernel::ReadableEvent> NVFlinger::FindVsyncEvent(u64 display_id) const {
@@ -100,29 +95,20 @@ Kernel::SharedPtr<Kernel::ReadableEvent> NVFlinger::FindVsyncEvent(u64 display_i
         return nullptr;
     }
 
-    return display->GetVSyncEvent();
+    return display->vsync_event.readable;
 }
 
-BufferQueue& NVFlinger::FindBufferQueue(u32 id) {
+std::shared_ptr<BufferQueue> NVFlinger::FindBufferQueue(u32 id) const {
     const auto itr = std::find_if(buffer_queues.begin(), buffer_queues.end(),
-                                  [id](const auto& queue) { return queue.GetId() == id; });
+                                  [&](const auto& queue) { return queue->GetId() == id; });
 
     ASSERT(itr != buffer_queues.end());
     return *itr;
 }
 
-const BufferQueue& NVFlinger::FindBufferQueue(u32 id) const {
-    const auto itr = std::find_if(buffer_queues.begin(), buffer_queues.end(),
-                                  [id](const auto& queue) { return queue.GetId() == id; });
-
-    ASSERT(itr != buffer_queues.end());
-    return *itr;
-}
-
-VI::Display* NVFlinger::FindDisplay(u64 display_id) {
-    const auto itr =
-        std::find_if(displays.begin(), displays.end(),
-                     [&](const VI::Display& display) { return display.GetID() == display_id; });
+Display* NVFlinger::FindDisplay(u64 display_id) {
+    const auto itr = std::find_if(displays.begin(), displays.end(),
+                                  [&](const Display& display) { return display.id == display_id; });
 
     if (itr == displays.end()) {
         return nullptr;
@@ -131,10 +117,9 @@ VI::Display* NVFlinger::FindDisplay(u64 display_id) {
     return &*itr;
 }
 
-const VI::Display* NVFlinger::FindDisplay(u64 display_id) const {
-    const auto itr =
-        std::find_if(displays.begin(), displays.end(),
-                     [&](const VI::Display& display) { return display.GetID() == display_id; });
+const Display* NVFlinger::FindDisplay(u64 display_id) const {
+    const auto itr = std::find_if(displays.begin(), displays.end(),
+                                  [&](const Display& display) { return display.id == display_id; });
 
     if (itr == displays.end()) {
         return nullptr;
@@ -143,41 +128,57 @@ const VI::Display* NVFlinger::FindDisplay(u64 display_id) const {
     return &*itr;
 }
 
-VI::Layer* NVFlinger::FindLayer(u64 display_id, u64 layer_id) {
+Layer* NVFlinger::FindLayer(u64 display_id, u64 layer_id) {
     auto* const display = FindDisplay(display_id);
 
     if (display == nullptr) {
         return nullptr;
     }
 
-    return display->FindLayer(layer_id);
+    const auto itr = std::find_if(display->layers.begin(), display->layers.end(),
+                                  [&](const Layer& layer) { return layer.id == layer_id; });
+
+    if (itr == display->layers.end()) {
+        return nullptr;
+    }
+
+    return &*itr;
 }
 
-const VI::Layer* NVFlinger::FindLayer(u64 display_id, u64 layer_id) const {
+const Layer* NVFlinger::FindLayer(u64 display_id, u64 layer_id) const {
     const auto* const display = FindDisplay(display_id);
 
     if (display == nullptr) {
         return nullptr;
     }
 
-    return display->FindLayer(layer_id);
+    const auto itr = std::find_if(display->layers.begin(), display->layers.end(),
+                                  [&](const Layer& layer) { return layer.id == layer_id; });
+
+    if (itr == display->layers.end()) {
+        return nullptr;
+    }
+
+    return &*itr;
 }
 
 void NVFlinger::Compose() {
     for (auto& display : displays) {
         // Trigger vsync for this display at the end of drawing
-        SCOPE_EXIT({ display.SignalVSyncEvent(); });
+        SCOPE_EXIT({ display.vsync_event.writable->Signal(); });
 
         // Don't do anything for displays without layers.
-        if (!display.HasLayers())
+        if (display.layers.empty())
             continue;
 
         // TODO(Subv): Support more than 1 layer.
-        VI::Layer& layer = display.GetLayer(0);
-        auto& buffer_queue = layer.GetBufferQueue();
+        ASSERT_MSG(display.layers.size() == 1, "Max 1 layer per display is supported");
+
+        Layer& layer = display.layers[0];
+        auto& buffer_queue = layer.buffer_queue;
 
         // Search for a queued buffer and acquire it
-        auto buffer = buffer_queue.AcquireBuffer();
+        auto buffer = buffer_queue->AcquireBuffer();
 
         MicroProfileFlip();
 
@@ -202,8 +203,19 @@ void NVFlinger::Compose() {
                      igbp_buffer.width, igbp_buffer.height, igbp_buffer.stride,
                      buffer->get().transform, buffer->get().crop_rect);
 
-        buffer_queue.ReleaseBuffer(buffer->get().slot);
+        buffer_queue->ReleaseBuffer(buffer->get().slot);
     }
 }
 
+Layer::Layer(u64 id, std::shared_ptr<BufferQueue> queue) : id(id), buffer_queue(std::move(queue)) {}
+Layer::~Layer() = default;
+
+Display::Display(u64 id, std::string name) : id(id), name(std::move(name)) {
+    auto& kernel = Core::System::GetInstance().Kernel();
+    vsync_event = Kernel::WritableEvent::CreateEventPair(kernel, Kernel::ResetType::Sticky,
+                                                         fmt::format("Display VSync Event {}", id));
+}
+
+Display::~Display() = default;
+
 } // namespace Service::NVFlinger

src/core/hle/service/nvflinger/nvflinger.h

@@ -4,6 +4,7 @@
 
 #pragma once
 
+#include <array>
 #include <memory>
 #include <optional>
 #include <string>
@@ -14,9 +15,8 @@
 #include "core/hle/kernel/object.h"
 
 namespace Core::Timing {
-class CoreTiming;
 struct EventType;
-} // namespace Core::Timing
+}
 
 namespace Kernel {
 class ReadableEvent;
@@ -25,20 +25,34 @@ class WritableEvent;
 
 namespace Service::Nvidia {
 class Module;
-} // namespace Service::Nvidia
-
-namespace Service::VI {
-class Display;
-class Layer;
-} // namespace Service::VI
+}
 
 namespace Service::NVFlinger {
 
 class BufferQueue;
 
+struct Layer {
+    Layer(u64 id, std::shared_ptr<BufferQueue> queue);
+    ~Layer();
+
+    u64 id;
+    std::shared_ptr<BufferQueue> buffer_queue;
+};
+
+struct Display {
+    Display(u64 id, std::string name);
+    ~Display();
+
+    u64 id;
+    std::string name;
+
+    std::vector<Layer> layers;
+    Kernel::EventPair vsync_event;
+};
+
 class NVFlinger final {
 public:
-    explicit NVFlinger(Core::Timing::CoreTiming& core_timing);
+    NVFlinger();
     ~NVFlinger();
 
     /// Sets the NVDrv module instance to use to send buffers to the GPU.
@@ -65,10 +79,7 @@ public:
     Kernel::SharedPtr<Kernel::ReadableEvent> FindVsyncEvent(u64 display_id) const;
 
     /// Obtains a buffer queue identified by the ID.
-    BufferQueue& FindBufferQueue(u32 id);
-
-    /// Obtains a buffer queue identified by the ID.
-    const BufferQueue& FindBufferQueue(u32 id) const;
+    std::shared_ptr<BufferQueue> FindBufferQueue(u32 id) const;
 
     /// Performs a composition request to the emulated nvidia GPU and triggers the vsync events when
     /// finished.
@@ -76,21 +87,27 @@ public:
 
 private:
     /// Finds the display identified by the specified ID.
-    VI::Display* FindDisplay(u64 display_id);
+    Display* FindDisplay(u64 display_id);
 
     /// Finds the display identified by the specified ID.
-    const VI::Display* FindDisplay(u64 display_id) const;
+    const Display* FindDisplay(u64 display_id) const;
 
     /// Finds the layer identified by the specified ID in the desired display.
-    VI::Layer* FindLayer(u64 display_id, u64 layer_id);
+    Layer* FindLayer(u64 display_id, u64 layer_id);
 
     /// Finds the layer identified by the specified ID in the desired display.
-    const VI::Layer* FindLayer(u64 display_id, u64 layer_id) const;
+    const Layer* FindLayer(u64 display_id, u64 layer_id) const;
 
     std::shared_ptr<Nvidia::Module> nvdrv;
 
-    std::vector<VI::Display> displays;
-    std::vector<BufferQueue> buffer_queues;
+    std::array<Display, 5> displays{{
+        {0, "Default"},
+        {1, "External"},
+        {2, "Edid"},
+        {3, "Internal"},
+        {4, "Null"},
+    }};
+    std::vector<std::shared_ptr<BufferQueue>> buffer_queues;
 
     /// Id to use for the next layer that is created, this counter is shared among all displays.
     u64 next_layer_id = 1;
@@ -100,9 +117,6 @@ private:
 
     /// Event that handles screen composition.
     Core::Timing::EventType* composition_event;
-
-    /// Core timing instance for registering/unregistering the composition event.
-    Core::Timing::CoreTiming& core_timing;
 };
 
 } // namespace Service::NVFlinger

src/core/hle/service/service.cpp

@@ -194,11 +194,10 @@ ResultCode ServiceFrameworkBase::HandleSyncRequest(Kernel::HLERequestContext& co
 // Module interface
 
 /// Initialize ServiceManager
-void Init(std::shared_ptr<SM::ServiceManager>& sm, Core::System& system,
-          FileSys::VfsFilesystem& vfs) {
+void Init(std::shared_ptr<SM::ServiceManager>& sm, FileSys::VfsFilesystem& vfs) {
     // NVFlinger needs to be accessed by several services like Vi and AppletOE so we instantiate it
     // here and pass it into the respective InstallInterfaces functions.
-    auto nv_flinger = std::make_shared<NVFlinger::NVFlinger>(system.CoreTiming());
+    auto nv_flinger = std::make_shared<NVFlinger::NVFlinger>();
 
     SM::ServiceManager::InstallInterfaces(sm);
 

src/core/hle/service/service.h

@@ -14,14 +14,6 @@
 ////////////////////////////////////////////////////////////////////////////////////////////////////
 // Namespace Service
 
-namespace Core {
-class System;
-}
-
-namespace FileSys {
-class VfsFilesystem;
-}
-
 namespace Kernel {
 class ClientPort;
 class ServerPort;
@@ -29,6 +21,10 @@ class ServerSession;
 class HLERequestContext;
 } // namespace Kernel
 
+namespace FileSys {
+class VfsFilesystem;
+}
+
 namespace Service {
 
 namespace SM {
@@ -182,8 +178,7 @@ private:
 };
 
 /// Initialize ServiceManager
-void Init(std::shared_ptr<SM::ServiceManager>& sm, Core::System& system,
-          FileSys::VfsFilesystem& vfs);
+void Init(std::shared_ptr<SM::ServiceManager>& sm, FileSys::VfsFilesystem& vfs);
 
 /// Shutdown ServiceManager
 void Shutdown();

src/core/hle/service/time/time.cpp

@@ -5,7 +5,6 @@
 #include <chrono>
 #include <ctime>
 #include "common/logging/log.h"
-#include "core/core.h"
 #include "core/core_timing.h"
 #include "core/core_timing_util.h"
 #include "core/hle/ipc_helpers.h"
@@ -107,9 +106,8 @@ private:
     void GetCurrentTimePoint(Kernel::HLERequestContext& ctx) {
         LOG_DEBUG(Service_Time, "called");
 
-        const auto& core_timing = Core::System::GetInstance().CoreTiming();
         const SteadyClockTimePoint steady_clock_time_point{
-            Core::Timing::cyclesToMs(core_timing.GetTicks()) / 1000};
+            Core::Timing::cyclesToMs(Core::Timing::GetTicks()) / 1000};
         IPC::ResponseBuilder rb{ctx, (sizeof(SteadyClockTimePoint) / 4) + 2};
         rb.Push(RESULT_SUCCESS);
         rb.PushRaw(steady_clock_time_point);
@@ -283,9 +281,8 @@ void Module::Interface::GetClockSnapshot(Kernel::HLERequestContext& ctx) {
         return;
     }
 
-    const auto& core_timing = Core::System::GetInstance().CoreTiming();
     const SteadyClockTimePoint steady_clock_time_point{
-        Core::Timing::cyclesToMs(core_timing.GetTicks()) / 1000, {}};
+        Core::Timing::cyclesToMs(Core::Timing::GetTicks()) / 1000, {}};
 
     CalendarTime calendar_time{};
     calendar_time.year = tm->tm_year + 1900;

src/core/hle/service/vi/display/vi_display.cpp

@@ -1,71 +0,0 @@
-// Copyright 2019 yuzu emulator team
-// Licensed under GPLv2 or any later version
-// Refer to the license.txt file included.
-
-#include <algorithm>
-#include <utility>
-
-#include <fmt/format.h>
-
-#include "common/assert.h"
-#include "core/core.h"
-#include "core/hle/kernel/readable_event.h"
-#include "core/hle/service/vi/display/vi_display.h"
-#include "core/hle/service/vi/layer/vi_layer.h"
-
-namespace Service::VI {
-
-Display::Display(u64 id, std::string name) : id{id}, name{std::move(name)} {
-    auto& kernel = Core::System::GetInstance().Kernel();
-    vsync_event = Kernel::WritableEvent::CreateEventPair(kernel, Kernel::ResetType::Sticky,
-                                                         fmt::format("Display VSync Event {}", id));
-}
-
-Display::~Display() = default;
-
-Layer& Display::GetLayer(std::size_t index) {
-    return layers.at(index);
-}
-
-const Layer& Display::GetLayer(std::size_t index) const {
-    return layers.at(index);
-}
-
-Kernel::SharedPtr<Kernel::ReadableEvent> Display::GetVSyncEvent() const {
-    return vsync_event.readable;
-}
-
-void Display::SignalVSyncEvent() {
-    vsync_event.writable->Signal();
-}
-
-void Display::CreateLayer(u64 id, NVFlinger::BufferQueue& buffer_queue) {
-    // TODO(Subv): Support more than 1 layer.
-    ASSERT_MSG(layers.empty(), "Only one layer is supported per display at the moment");
-
-    layers.emplace_back(id, buffer_queue);
-}
-
-Layer* Display::FindLayer(u64 id) {
-    const auto itr = std::find_if(layers.begin(), layers.end(),
-                                  [id](const VI::Layer& layer) { return layer.GetID() == id; });
-
-    if (itr == layers.end()) {
-        return nullptr;
-    }
-
-    return &*itr;
-}
-
-const Layer* Display::FindLayer(u64 id) const {
-    const auto itr = std::find_if(layers.begin(), layers.end(),
-                                  [id](const VI::Layer& layer) { return layer.GetID() == id; });
-
-    if (itr == layers.end()) {
-        return nullptr;
-    }
-
-    return &*itr;
-}
-
-} // namespace Service::VI

src/core/hle/service/vi/display/vi_display.h

@@ -1,98 +0,0 @@
-// Copyright 2019 yuzu emulator team
-// Licensed under GPLv2 or any later version
-// Refer to the license.txt file included.
-
-#pragma once
-
-#include <string>
-#include <vector>
-
-#include "common/common_types.h"
-#include "core/hle/kernel/writable_event.h"
-
-namespace Service::NVFlinger {
-class BufferQueue;
-}
-
-namespace Service::VI {
-
-class Layer;
-
-/// Represents a single display type
-class Display {
-public:
-    /// Constructs a display with a given unique ID and name.
-    ///
-    /// @param id   The unique ID for this display.
-    /// @param name The name for this display.
-    ///
-    Display(u64 id, std::string name);
-    ~Display();
-
-    Display(const Display&) = delete;
-    Display& operator=(const Display&) = delete;
-
-    Display(Display&&) = default;
-    Display& operator=(Display&&) = default;
-
-    /// Gets the unique ID assigned to this display.
-    u64 GetID() const {
-        return id;
-    }
-
-    /// Gets the name of this display
-    const std::string& GetName() const {
-        return name;
-    }
-
-    /// Whether or not this display has any layers added to it.
-    bool HasLayers() const {
-        return !layers.empty();
-    }
-
-    /// Gets a layer for this display based off an index.
-    Layer& GetLayer(std::size_t index);
-
-    /// Gets a layer for this display based off an index.
-    const Layer& GetLayer(std::size_t index) const;
-
-    /// Gets the readable vsync event.
-    Kernel::SharedPtr<Kernel::ReadableEvent> GetVSyncEvent() const;
-
-    /// Signals the internal vsync event.
-    void SignalVSyncEvent();
-
-    /// Creates and adds a layer to this display with the given ID.
-    ///
-    /// @param id           The ID to assign to the created layer.
-    /// @param buffer_queue The buffer queue for the layer instance to use.
-    ///
-    void CreateLayer(u64 id, NVFlinger::BufferQueue& buffer_queue);
-
-    /// Attempts to find a layer with the given ID.
-    ///
-    /// @param id The layer ID.
-    ///
-    /// @returns If found, the Layer instance with the given ID.
-    ///          If not found, then nullptr is returned.
-    ///
-    Layer* FindLayer(u64 id);
-
-    /// Attempts to find a layer with the given ID.
-    ///
-    /// @param id The layer ID.
-    ///
-    /// @returns If found, the Layer instance with the given ID.
-    ///          If not found, then nullptr is returned.
-    ///
-    const Layer* FindLayer(u64 id) const;
-
-private:
-    u64 id;
-    std::string name;
-
-    std::vector<Layer> layers;
-    Kernel::EventPair vsync_event;
-};
-
-} // namespace Service::VI

src/core/hle/service/vi/layer/vi_layer.cpp

@@ -1,13 +0,0 @@
-// Copyright 2019 yuzu emulator team
-// Licensed under GPLv2 or any later version
-// Refer to the license.txt file included.
-
-#include "core/hle/service/vi/layer/vi_layer.h"
-
-namespace Service::VI {
-
-Layer::Layer(u64 id, NVFlinger::BufferQueue& queue) : id{id}, buffer_queue{queue} {}
-
-Layer::~Layer() = default;
-
-} // namespace Service::VI

src/core/hle/service/vi/layer/vi_layer.h

@@ -1,52 +0,0 @@
-// Copyright 2019 yuzu emulator team
-// Licensed under GPLv2 or any later version
-// Refer to the license.txt file included.
-
-#pragma once
-
-#include "common/common_types.h"
-
-namespace Service::NVFlinger {
-class BufferQueue;
-}
-
-namespace Service::VI {
-
-/// Represents a single display layer.
-class Layer {
-public:
-    /// Constructs a layer with a given ID and buffer queue.
-    ///
-    /// @param id    The ID to assign to this layer.
-    /// @param queue The buffer queue for this layer to use.
-    ///
-    Layer(u64 id, NVFlinger::BufferQueue& queue);
-    ~Layer();
-
-    Layer(const Layer&) = delete;
-    Layer& operator=(const Layer&) = delete;
-
-    Layer(Layer&&) = default;
-    Layer& operator=(Layer&&) = delete;
-
-    /// Gets the ID for this layer.
-    u64 GetID() const {
-        return id;
-    }
-
-    /// Gets a reference to the buffer queue this layer is using.
-    NVFlinger::BufferQueue& GetBufferQueue() {
-        return buffer_queue;
-    }
-
-    /// Gets a const reference to the buffer queue this layer is using.
-    const NVFlinger::BufferQueue& GetBufferQueue() const {
-        return buffer_queue;
-    }
-
-private:
-    u64 id;
-    NVFlinger::BufferQueue& buffer_queue;
-};
-
-} // namespace Service::VI

src/core/hle/service/vi/vi.cpp

@@ -420,7 +420,7 @@ public:
         u32_le fence_is_valid;
         std::array<Fence, 2> fences;
 
-        Common::Rectangle<int> GetCropRect() const {
+        MathUtil::Rectangle<int> GetCropRect() const {
             return {crop_left, crop_top, crop_right, crop_bottom};
         }
     };
@@ -525,7 +525,7 @@ private:
         LOG_DEBUG(Service_VI, "called. id=0x{:08X} transaction={:X}, flags=0x{:08X}", id,
                   static_cast<u32>(transaction), flags);
 
-        auto& buffer_queue = nv_flinger->FindBufferQueue(id);
+        auto buffer_queue = nv_flinger->FindBufferQueue(id);
 
         if (transaction == TransactionId::Connect) {
             IGBPConnectRequestParcel request{ctx.ReadBuffer()};
@@ -538,7 +538,7 @@ private:
         } else if (transaction == TransactionId::SetPreallocatedBuffer) {
             IGBPSetPreallocatedBufferRequestParcel request{ctx.ReadBuffer()};
 
-            buffer_queue.SetPreallocatedBuffer(request.data.slot, request.buffer);
+            buffer_queue->SetPreallocatedBuffer(request.data.slot, request.buffer);
 
             IGBPSetPreallocatedBufferResponseParcel response{};
             ctx.WriteBuffer(response.Serialize());
@@ -546,7 +546,7 @@ private:
             IGBPDequeueBufferRequestParcel request{ctx.ReadBuffer()};
             const u32 width{request.data.width};
             const u32 height{request.data.height};
-            std::optional<u32> slot = buffer_queue.DequeueBuffer(width, height);
+            std::optional<u32> slot = buffer_queue->DequeueBuffer(width, height);
 
             if (slot) {
                 // Buffer is available
@@ -559,8 +559,8 @@ private:
                     [=](Kernel::SharedPtr<Kernel::Thread> thread, Kernel::HLERequestContext& ctx,
                         Kernel::ThreadWakeupReason reason) {
                         // Repeat TransactParcel DequeueBuffer when a buffer is available
-                        auto& buffer_queue = nv_flinger->FindBufferQueue(id);
-                        std::optional<u32> slot = buffer_queue.DequeueBuffer(width, height);
+                        auto buffer_queue = nv_flinger->FindBufferQueue(id);
+                        std::optional<u32> slot = buffer_queue->DequeueBuffer(width, height);
                         ASSERT_MSG(slot != std::nullopt, "Could not dequeue buffer.");
 
                         IGBPDequeueBufferResponseParcel response{*slot};
@@ -568,28 +568,28 @@ private:
                         IPC::ResponseBuilder rb{ctx, 2};
                         rb.Push(RESULT_SUCCESS);
                     },
-                    buffer_queue.GetWritableBufferWaitEvent());
+                    buffer_queue->GetWritableBufferWaitEvent());
             }
         } else if (transaction == TransactionId::RequestBuffer) {
             IGBPRequestBufferRequestParcel request{ctx.ReadBuffer()};
 
-            auto& buffer = buffer_queue.RequestBuffer(request.slot);
+            auto& buffer = buffer_queue->RequestBuffer(request.slot);
 
             IGBPRequestBufferResponseParcel response{buffer};
             ctx.WriteBuffer(response.Serialize());
         } else if (transaction == TransactionId::QueueBuffer) {
             IGBPQueueBufferRequestParcel request{ctx.ReadBuffer()};
 
-            buffer_queue.QueueBuffer(request.data.slot, request.data.transform,
-                                     request.data.GetCropRect());
+            buffer_queue->QueueBuffer(request.data.slot, request.data.transform,
+                                      request.data.GetCropRect());
 
             IGBPQueueBufferResponseParcel response{1280, 720};
             ctx.WriteBuffer(response.Serialize());
         } else if (transaction == TransactionId::Query) {
             IGBPQueryRequestParcel request{ctx.ReadBuffer()};
 
-            const u32 value =
-                buffer_queue.Query(static_cast<NVFlinger::BufferQueue::QueryType>(request.type));
+            u32 value =
+                buffer_queue->Query(static_cast<NVFlinger::BufferQueue::QueryType>(request.type));
 
             IGBPQueryResponseParcel response{value};
             ctx.WriteBuffer(response.Serialize());
@@ -629,12 +629,12 @@ private:
 
         LOG_WARNING(Service_VI, "(STUBBED) called id={}, unknown={:08X}", id, unknown);
 
-        const auto& buffer_queue = nv_flinger->FindBufferQueue(id);
+        const auto buffer_queue = nv_flinger->FindBufferQueue(id);
 
         // TODO(Subv): Find out what this actually is.
         IPC::ResponseBuilder rb{ctx, 2, 1};
         rb.Push(RESULT_SUCCESS);
-        rb.PushCopyObjects(buffer_queue.GetBufferWaitEvent());
+        rb.PushCopyObjects(buffer_queue->GetBufferWaitEvent());
     }
 
     std::shared_ptr<NVFlinger::NVFlinger> nv_flinger;
@@ -752,7 +752,6 @@ public:
             {1102, nullptr, "GetDisplayResolution"},
             {2010, &IManagerDisplayService::CreateManagedLayer, "CreateManagedLayer"},
             {2011, nullptr, "DestroyManagedLayer"},
-            {2012, nullptr, "CreateStrayLayer"},
             {2050, nullptr, "CreateIndirectLayer"},
             {2051, nullptr, "DestroyIndirectLayer"},
             {2052, nullptr, "CreateIndirectProducerEndPoint"},

src/core/memory.cpp

@@ -71,20 +71,15 @@ static void MapPages(PageTable& page_table, VAddr base, u64 size, u8* memory, Pa
                                  FlushMode::FlushAndInvalidate);
 
     VAddr end = base + size;
-    ASSERT_MSG(end <= page_table.pointers.size(), "out of range mapping at {:016X}",
-               base + page_table.pointers.size());
+    while (base != end) {
+        ASSERT_MSG(base < page_table.pointers.size(), "out of range mapping at {:016X}", base);
 
-    std::fill(page_table.attributes.begin() + base, page_table.attributes.begin() + end, type);
+        page_table.attributes[base] = type;
+        page_table.pointers[base] = memory;
 
-    if (memory == nullptr) {
-        std::fill(page_table.pointers.begin() + base, page_table.pointers.begin() + end, memory);
-    } else {
-        while (base != end) {
-            page_table.pointers[base] = memory;
-
-            base += 1;
+        base += 1;
+        if (memory != nullptr)
             memory += PAGE_SIZE;
-        }
     }
 }
 

src/input_common/motion_emu.cpp

@@ -32,12 +32,12 @@ public:
     }
 
     void BeginTilt(int x, int y) {
-        mouse_origin = Common::MakeVec(x, y);
+        mouse_origin = Math::MakeVec(x, y);
         is_tilting = true;
     }
 
     void Tilt(int x, int y) {
-        auto mouse_move = Common::MakeVec(x, y) - mouse_origin;
+        auto mouse_move = Math::MakeVec(x, y) - mouse_origin;
         if (is_tilting) {
             std::lock_guard<std::mutex> guard(tilt_mutex);
             if (mouse_move.x == 0 && mouse_move.y == 0) {
@@ -45,7 +45,7 @@ public:
             } else {
                 tilt_direction = mouse_move.Cast<float>();
                 tilt_angle =
-                    std::clamp(tilt_direction.Normalize() * sensitivity, 0.0f, Common::PI * 0.5f);
+                    std::clamp(tilt_direction.Normalize() * sensitivity, 0.0f, MathUtil::PI * 0.5f);
             }
         }
     }
@@ -56,7 +56,7 @@ public:
         is_tilting = false;
     }
 
-    std::tuple<Common::Vec3<float>, Common::Vec3<float>> GetStatus() {
+    std::tuple<Math::Vec3<float>, Math::Vec3<float>> GetStatus() {
         std::lock_guard<std::mutex> guard(status_mutex);
         return status;
     }
@@ -66,17 +66,17 @@ private:
     const std::chrono::steady_clock::duration update_duration;
     const float sensitivity;
 
-    Common::Vec2<int> mouse_origin;
+    Math::Vec2<int> mouse_origin;
 
     std::mutex tilt_mutex;
-    Common::Vec2<float> tilt_direction;
+    Math::Vec2<float> tilt_direction;
     float tilt_angle = 0;
 
     bool is_tilting = false;
 
     Common::Event shutdown_event;
 
-    std::tuple<Common::Vec3<float>, Common::Vec3<float>> status;
+    std::tuple<Math::Vec3<float>, Math::Vec3<float>> status;
     std::mutex status_mutex;
 
     // Note: always keep the thread declaration at the end so that other objects are initialized
@@ -85,8 +85,8 @@ private:
 
     void MotionEmuThread() {
         auto update_time = std::chrono::steady_clock::now();
-        Common::Quaternion<float> q = Common::MakeQuaternion(Common::Vec3<float>(), 0);
-        Common::Quaternion<float> old_q;
+        Math::Quaternion<float> q = MakeQuaternion(Math::Vec3<float>(), 0);
+        Math::Quaternion<float> old_q;
 
         while (!shutdown_event.WaitUntil(update_time)) {
             update_time += update_duration;
@@ -96,18 +96,18 @@ private:
                 std::lock_guard<std::mutex> guard(tilt_mutex);
 
                 // Find the quaternion describing current 3DS tilting
-                q = Common::MakeQuaternion(
-                    Common::MakeVec(-tilt_direction.y, 0.0f, tilt_direction.x), tilt_angle);
+                q = MakeQuaternion(Math::MakeVec(-tilt_direction.y, 0.0f, tilt_direction.x),
+                                   tilt_angle);
             }
 
             auto inv_q = q.Inverse();
 
             // Set the gravity vector in world space
-            auto gravity = Common::MakeVec(0.0f, -1.0f, 0.0f);
+            auto gravity = Math::MakeVec(0.0f, -1.0f, 0.0f);
 
             // Find the angular rate vector in world space
             auto angular_rate = ((q - old_q) * inv_q).xyz * 2;
-            angular_rate *= 1000 / update_millisecond / Common::PI * 180;
+            angular_rate *= 1000 / update_millisecond / MathUtil::PI * 180;
 
             // Transform the two vectors from world space to 3DS space
             gravity = QuaternionRotate(inv_q, gravity);
@@ -131,7 +131,7 @@ public:
         device = std::make_shared<MotionEmuDevice>(update_millisecond, sensitivity);
     }
 
-    std::tuple<Common::Vec3<float>, Common::Vec3<float>> GetStatus() const override {
+    std::tuple<Math::Vec3<float>, Math::Vec3<float>> GetStatus() const override {
         return device->GetStatus();
     }
 

src/tests/core/core_timing.cpp

@@ -28,103 +28,100 @@ void CallbackTemplate(u64 userdata, s64 cycles_late) {
     REQUIRE(lateness == cycles_late);
 }
 
-struct ScopeInit final {
+class ScopeInit final {
+public:
     ScopeInit() {
-        core_timing.Initialize();
+        Core::Timing::Init();
     }
     ~ScopeInit() {
-        core_timing.Shutdown();
+        Core::Timing::Shutdown();
     }
-
-    Core::Timing::CoreTiming core_timing;
 };
 
-static void AdvanceAndCheck(Core::Timing::CoreTiming& core_timing, u32 idx, int downcount,
-                            int expected_lateness = 0, int cpu_downcount = 0) {
+static void AdvanceAndCheck(u32 idx, int downcount, int expected_lateness = 0,
+                            int cpu_downcount = 0) {
     callbacks_ran_flags = 0;
     expected_callback = CB_IDS[idx];
     lateness = expected_lateness;
 
     // Pretend we executed X cycles of instructions.
-    core_timing.AddTicks(core_timing.GetDowncount() - cpu_downcount);
-    core_timing.Advance();
+    Core::Timing::AddTicks(Core::Timing::GetDowncount() - cpu_downcount);
+    Core::Timing::Advance();
 
     REQUIRE(decltype(callbacks_ran_flags)().set(idx) == callbacks_ran_flags);
-    REQUIRE(downcount == core_timing.GetDowncount());
+    REQUIRE(downcount == Core::Timing::GetDowncount());
 }
 
 TEST_CASE("CoreTiming[BasicOrder]", "[core]") {
     ScopeInit guard;
-    auto& core_timing = guard.core_timing;
 
-    Core::Timing::EventType* cb_a = core_timing.RegisterEvent("callbackA", CallbackTemplate<0>);
-    Core::Timing::EventType* cb_b = core_timing.RegisterEvent("callbackB", CallbackTemplate<1>);
-    Core::Timing::EventType* cb_c = core_timing.RegisterEvent("callbackC", CallbackTemplate<2>);
-    Core::Timing::EventType* cb_d = core_timing.RegisterEvent("callbackD", CallbackTemplate<3>);
-    Core::Timing::EventType* cb_e = core_timing.RegisterEvent("callbackE", CallbackTemplate<4>);
+    Core::Timing::EventType* cb_a = Core::Timing::RegisterEvent("callbackA", CallbackTemplate<0>);
+    Core::Timing::EventType* cb_b = Core::Timing::RegisterEvent("callbackB", CallbackTemplate<1>);
+    Core::Timing::EventType* cb_c = Core::Timing::RegisterEvent("callbackC", CallbackTemplate<2>);
+    Core::Timing::EventType* cb_d = Core::Timing::RegisterEvent("callbackD", CallbackTemplate<3>);
+    Core::Timing::EventType* cb_e = Core::Timing::RegisterEvent("callbackE", CallbackTemplate<4>);
 
     // Enter slice 0
-    core_timing.Advance();
+    Core::Timing::Advance();
 
     // D -> B -> C -> A -> E
-    core_timing.ScheduleEvent(1000, cb_a, CB_IDS[0]);
-    REQUIRE(1000 == core_timing.GetDowncount());
-    core_timing.ScheduleEvent(500, cb_b, CB_IDS[1]);
-    REQUIRE(500 == core_timing.GetDowncount());
-    core_timing.ScheduleEvent(800, cb_c, CB_IDS[2]);
-    REQUIRE(500 == core_timing.GetDowncount());
-    core_timing.ScheduleEvent(100, cb_d, CB_IDS[3]);
-    REQUIRE(100 == core_timing.GetDowncount());
-    core_timing.ScheduleEvent(1200, cb_e, CB_IDS[4]);
-    REQUIRE(100 == core_timing.GetDowncount());
+    Core::Timing::ScheduleEvent(1000, cb_a, CB_IDS[0]);
+    REQUIRE(1000 == Core::Timing::GetDowncount());
+    Core::Timing::ScheduleEvent(500, cb_b, CB_IDS[1]);
+    REQUIRE(500 == Core::Timing::GetDowncount());
+    Core::Timing::ScheduleEvent(800, cb_c, CB_IDS[2]);
+    REQUIRE(500 == Core::Timing::GetDowncount());
+    Core::Timing::ScheduleEvent(100, cb_d, CB_IDS[3]);
+    REQUIRE(100 == Core::Timing::GetDowncount());
+    Core::Timing::ScheduleEvent(1200, cb_e, CB_IDS[4]);
+    REQUIRE(100 == Core::Timing::GetDowncount());
 
-    AdvanceAndCheck(core_timing, 3, 400);
-    AdvanceAndCheck(core_timing, 1, 300);
-    AdvanceAndCheck(core_timing, 2, 200);
-    AdvanceAndCheck(core_timing, 0, 200);
-    AdvanceAndCheck(core_timing, 4, MAX_SLICE_LENGTH);
+    AdvanceAndCheck(3, 400);
+    AdvanceAndCheck(1, 300);
+    AdvanceAndCheck(2, 200);
+    AdvanceAndCheck(0, 200);
+    AdvanceAndCheck(4, MAX_SLICE_LENGTH);
 }
 
 TEST_CASE("CoreTiming[Threadsave]", "[core]") {
     ScopeInit guard;
-    auto& core_timing = guard.core_timing;
 
-    Core::Timing::EventType* cb_a = core_timing.RegisterEvent("callbackA", CallbackTemplate<0>);
-    Core::Timing::EventType* cb_b = core_timing.RegisterEvent("callbackB", CallbackTemplate<1>);
-    Core::Timing::EventType* cb_c = core_timing.RegisterEvent("callbackC", CallbackTemplate<2>);
-    Core::Timing::EventType* cb_d = core_timing.RegisterEvent("callbackD", CallbackTemplate<3>);
-    Core::Timing::EventType* cb_e = core_timing.RegisterEvent("callbackE", CallbackTemplate<4>);
+    Core::Timing::EventType* cb_a = Core::Timing::RegisterEvent("callbackA", CallbackTemplate<0>);
+    Core::Timing::EventType* cb_b = Core::Timing::RegisterEvent("callbackB", CallbackTemplate<1>);
+    Core::Timing::EventType* cb_c = Core::Timing::RegisterEvent("callbackC", CallbackTemplate<2>);
+    Core::Timing::EventType* cb_d = Core::Timing::RegisterEvent("callbackD", CallbackTemplate<3>);
+    Core::Timing::EventType* cb_e = Core::Timing::RegisterEvent("callbackE", CallbackTemplate<4>);
 
     // Enter slice 0
-    core_timing.Advance();
+    Core::Timing::Advance();
 
     // D -> B -> C -> A -> E
-    core_timing.ScheduleEventThreadsafe(1000, cb_a, CB_IDS[0]);
+    Core::Timing::ScheduleEventThreadsafe(1000, cb_a, CB_IDS[0]);
     // Manually force since ScheduleEventThreadsafe doesn't call it
-    core_timing.ForceExceptionCheck(1000);
-    REQUIRE(1000 == core_timing.GetDowncount());
-    core_timing.ScheduleEventThreadsafe(500, cb_b, CB_IDS[1]);
+    Core::Timing::ForceExceptionCheck(1000);
+    REQUIRE(1000 == Core::Timing::GetDowncount());
+    Core::Timing::ScheduleEventThreadsafe(500, cb_b, CB_IDS[1]);
     // Manually force since ScheduleEventThreadsafe doesn't call it
-    core_timing.ForceExceptionCheck(500);
-    REQUIRE(500 == core_timing.GetDowncount());
-    core_timing.ScheduleEventThreadsafe(800, cb_c, CB_IDS[2]);
+    Core::Timing::ForceExceptionCheck(500);
+    REQUIRE(500 == Core::Timing::GetDowncount());
+    Core::Timing::ScheduleEventThreadsafe(800, cb_c, CB_IDS[2]);
     // Manually force since ScheduleEventThreadsafe doesn't call it
-    core_timing.ForceExceptionCheck(800);
-    REQUIRE(500 == core_timing.GetDowncount());
-    core_timing.ScheduleEventThreadsafe(100, cb_d, CB_IDS[3]);
+    Core::Timing::ForceExceptionCheck(800);
+    REQUIRE(500 == Core::Timing::GetDowncount());
+    Core::Timing::ScheduleEventThreadsafe(100, cb_d, CB_IDS[3]);
     // Manually force since ScheduleEventThreadsafe doesn't call it
-    core_timing.ForceExceptionCheck(100);
-    REQUIRE(100 == core_timing.GetDowncount());
-    core_timing.ScheduleEventThreadsafe(1200, cb_e, CB_IDS[4]);
+    Core::Timing::ForceExceptionCheck(100);
+    REQUIRE(100 == Core::Timing::GetDowncount());
+    Core::Timing::ScheduleEventThreadsafe(1200, cb_e, CB_IDS[4]);
     // Manually force since ScheduleEventThreadsafe doesn't call it
-    core_timing.ForceExceptionCheck(1200);
-    REQUIRE(100 == core_timing.GetDowncount());
+    Core::Timing::ForceExceptionCheck(1200);
+    REQUIRE(100 == Core::Timing::GetDowncount());
 
-    AdvanceAndCheck(core_timing, 3, 400);
-    AdvanceAndCheck(core_timing, 1, 300);
-    AdvanceAndCheck(core_timing, 2, 200);
-    AdvanceAndCheck(core_timing, 0, 200);
-    AdvanceAndCheck(core_timing, 4, MAX_SLICE_LENGTH);
+    AdvanceAndCheck(3, 400);
+    AdvanceAndCheck(1, 300);
+    AdvanceAndCheck(2, 200);
+    AdvanceAndCheck(0, 200);
+    AdvanceAndCheck(4, MAX_SLICE_LENGTH);
 }
 
 namespace SharedSlotTest {
@@ -145,62 +142,59 @@ TEST_CASE("CoreTiming[SharedSlot]", "[core]") {
     using namespace SharedSlotTest;
 
     ScopeInit guard;
-    auto& core_timing = guard.core_timing;
 
-    Core::Timing::EventType* cb_a = core_timing.RegisterEvent("callbackA", FifoCallback<0>);
-    Core::Timing::EventType* cb_b = core_timing.RegisterEvent("callbackB", FifoCallback<1>);
-    Core::Timing::EventType* cb_c = core_timing.RegisterEvent("callbackC", FifoCallback<2>);
-    Core::Timing::EventType* cb_d = core_timing.RegisterEvent("callbackD", FifoCallback<3>);
-    Core::Timing::EventType* cb_e = core_timing.RegisterEvent("callbackE", FifoCallback<4>);
+    Core::Timing::EventType* cb_a = Core::Timing::RegisterEvent("callbackA", FifoCallback<0>);
+    Core::Timing::EventType* cb_b = Core::Timing::RegisterEvent("callbackB", FifoCallback<1>);
+    Core::Timing::EventType* cb_c = Core::Timing::RegisterEvent("callbackC", FifoCallback<2>);
+    Core::Timing::EventType* cb_d = Core::Timing::RegisterEvent("callbackD", FifoCallback<3>);
+    Core::Timing::EventType* cb_e = Core::Timing::RegisterEvent("callbackE", FifoCallback<4>);
 
-    core_timing.ScheduleEvent(1000, cb_a, CB_IDS[0]);
-    core_timing.ScheduleEvent(1000, cb_b, CB_IDS[1]);
-    core_timing.ScheduleEvent(1000, cb_c, CB_IDS[2]);
-    core_timing.ScheduleEvent(1000, cb_d, CB_IDS[3]);
-    core_timing.ScheduleEvent(1000, cb_e, CB_IDS[4]);
+    Core::Timing::ScheduleEvent(1000, cb_a, CB_IDS[0]);
+    Core::Timing::ScheduleEvent(1000, cb_b, CB_IDS[1]);
+    Core::Timing::ScheduleEvent(1000, cb_c, CB_IDS[2]);
+    Core::Timing::ScheduleEvent(1000, cb_d, CB_IDS[3]);
+    Core::Timing::ScheduleEvent(1000, cb_e, CB_IDS[4]);
 
     // Enter slice 0
-    core_timing.Advance();
-    REQUIRE(1000 == core_timing.GetDowncount());
+    Core::Timing::Advance();
+    REQUIRE(1000 == Core::Timing::GetDowncount());
 
     callbacks_ran_flags = 0;
     counter = 0;
     lateness = 0;
-    core_timing.AddTicks(core_timing.GetDowncount());
-    core_timing.Advance();
-    REQUIRE(MAX_SLICE_LENGTH == core_timing.GetDowncount());
+    Core::Timing::AddTicks(Core::Timing::GetDowncount());
+    Core::Timing::Advance();
+    REQUIRE(MAX_SLICE_LENGTH == Core::Timing::GetDowncount());
     REQUIRE(0x1FULL == callbacks_ran_flags.to_ullong());
 }
 
 TEST_CASE("Core::Timing[PredictableLateness]", "[core]") {
     ScopeInit guard;
-    auto& core_timing = guard.core_timing;
 
-    Core::Timing::EventType* cb_a = core_timing.RegisterEvent("callbackA", CallbackTemplate<0>);
-    Core::Timing::EventType* cb_b = core_timing.RegisterEvent("callbackB", CallbackTemplate<1>);
+    Core::Timing::EventType* cb_a = Core::Timing::RegisterEvent("callbackA", CallbackTemplate<0>);
+    Core::Timing::EventType* cb_b = Core::Timing::RegisterEvent("callbackB", CallbackTemplate<1>);
 
     // Enter slice 0
-    core_timing.Advance();
+    Core::Timing::Advance();
 
-    core_timing.ScheduleEvent(100, cb_a, CB_IDS[0]);
-    core_timing.ScheduleEvent(200, cb_b, CB_IDS[1]);
+    Core::Timing::ScheduleEvent(100, cb_a, CB_IDS[0]);
+    Core::Timing::ScheduleEvent(200, cb_b, CB_IDS[1]);
 
-    AdvanceAndCheck(core_timing, 0, 90, 10, -10); // (100 - 10)
-    AdvanceAndCheck(core_timing, 1, MAX_SLICE_LENGTH, 50, -50);
+    AdvanceAndCheck(0, 90, 10, -10); // (100 - 10)
+    AdvanceAndCheck(1, MAX_SLICE_LENGTH, 50, -50);
 }
 
 namespace ChainSchedulingTest {
 static int reschedules = 0;
 
-static void RescheduleCallback(Core::Timing::CoreTiming& core_timing, u64 userdata,
-                               s64 cycles_late) {
+static void RescheduleCallback(u64 userdata, s64 cycles_late) {
     --reschedules;
     REQUIRE(reschedules >= 0);
     REQUIRE(lateness == cycles_late);
 
     if (reschedules > 0) {
-        core_timing.ScheduleEvent(1000, reinterpret_cast<Core::Timing::EventType*>(userdata),
-                                  userdata);
+        Core::Timing::ScheduleEvent(1000, reinterpret_cast<Core::Timing::EventType*>(userdata),
+                                    userdata);
     }
 }
 } // namespace ChainSchedulingTest
@@ -209,39 +203,36 @@ TEST_CASE("CoreTiming[ChainScheduling]", "[core]") {
     using namespace ChainSchedulingTest;
 
     ScopeInit guard;
-    auto& core_timing = guard.core_timing;
 
-    Core::Timing::EventType* cb_a = core_timing.RegisterEvent("callbackA", CallbackTemplate<0>);
-    Core::Timing::EventType* cb_b = core_timing.RegisterEvent("callbackB", CallbackTemplate<1>);
-    Core::Timing::EventType* cb_c = core_timing.RegisterEvent("callbackC", CallbackTemplate<2>);
-    Core::Timing::EventType* cb_rs = core_timing.RegisterEvent(
-        "callbackReschedule", [&core_timing](u64 userdata, s64 cycles_late) {
-            RescheduleCallback(core_timing, userdata, cycles_late);
-        });
+    Core::Timing::EventType* cb_a = Core::Timing::RegisterEvent("callbackA", CallbackTemplate<0>);
+    Core::Timing::EventType* cb_b = Core::Timing::RegisterEvent("callbackB", CallbackTemplate<1>);
+    Core::Timing::EventType* cb_c = Core::Timing::RegisterEvent("callbackC", CallbackTemplate<2>);
+    Core::Timing::EventType* cb_rs =
+        Core::Timing::RegisterEvent("callbackReschedule", RescheduleCallback);
 
     // Enter slice 0
-    core_timing.Advance();
+    Core::Timing::Advance();
 
-    core_timing.ScheduleEvent(800, cb_a, CB_IDS[0]);
-    core_timing.ScheduleEvent(1000, cb_b, CB_IDS[1]);
-    core_timing.ScheduleEvent(2200, cb_c, CB_IDS[2]);
-    core_timing.ScheduleEvent(1000, cb_rs, reinterpret_cast<u64>(cb_rs));
-    REQUIRE(800 == core_timing.GetDowncount());
+    Core::Timing::ScheduleEvent(800, cb_a, CB_IDS[0]);
+    Core::Timing::ScheduleEvent(1000, cb_b, CB_IDS[1]);
+    Core::Timing::ScheduleEvent(2200, cb_c, CB_IDS[2]);
+    Core::Timing::ScheduleEvent(1000, cb_rs, reinterpret_cast<u64>(cb_rs));
+    REQUIRE(800 == Core::Timing::GetDowncount());
 
     reschedules = 3;
-    AdvanceAndCheck(core_timing, 0, 200);  // cb_a
-    AdvanceAndCheck(core_timing, 1, 1000); // cb_b, cb_rs
+    AdvanceAndCheck(0, 200);  // cb_a
+    AdvanceAndCheck(1, 1000); // cb_b, cb_rs
     REQUIRE(2 == reschedules);
 
-    core_timing.AddTicks(core_timing.GetDowncount());
-    core_timing.Advance(); // cb_rs
+    Core::Timing::AddTicks(Core::Timing::GetDowncount());
+    Core::Timing::Advance(); // cb_rs
     REQUIRE(1 == reschedules);
-    REQUIRE(200 == core_timing.GetDowncount());
+    REQUIRE(200 == Core::Timing::GetDowncount());
 
-    AdvanceAndCheck(core_timing, 2, 800); // cb_c
+    AdvanceAndCheck(2, 800); // cb_c
 
-    core_timing.AddTicks(core_timing.GetDowncount());
-    core_timing.Advance(); // cb_rs
+    Core::Timing::AddTicks(Core::Timing::GetDowncount());
+    Core::Timing::Advance(); // cb_rs
     REQUIRE(0 == reschedules);
-    REQUIRE(MAX_SLICE_LENGTH == core_timing.GetDowncount());
+    REQUIRE(MAX_SLICE_LENGTH == Core::Timing::GetDowncount());
 }

src/video_core/CMakeLists.txt

@@ -101,28 +101,6 @@ add_library(video_core STATIC
     video_core.h
 )
 
-if (ENABLE_VULKAN)
-    target_sources(video_core PRIVATE
-        renderer_vulkan/declarations.h
-        renderer_vulkan/maxwell_to_vk.cpp
-        renderer_vulkan/maxwell_to_vk.h
-        renderer_vulkan/vk_buffer_cache.cpp
-        renderer_vulkan/vk_buffer_cache.h
-        renderer_vulkan/vk_device.cpp
-        renderer_vulkan/vk_device.h
-        renderer_vulkan/vk_memory_manager.cpp
-        renderer_vulkan/vk_memory_manager.h
-        renderer_vulkan/vk_resource_manager.cpp
-        renderer_vulkan/vk_resource_manager.h
-        renderer_vulkan/vk_scheduler.cpp
-        renderer_vulkan/vk_scheduler.h
-        renderer_vulkan/vk_stream_buffer.cpp
-        renderer_vulkan/vk_stream_buffer.h)
-
-    target_include_directories(video_core PRIVATE ../../externals/Vulkan-Headers/include)
-    target_compile_definitions(video_core PRIVATE HAS_VULKAN)
-endif()
-
 create_target_directory_groups(video_core)
 
 target_link_libraries(video_core PUBLIC common core)

src/video_core/dma_pusher.cpp

@@ -33,36 +33,18 @@ void DmaPusher::DispatchCalls() {
 }
 
 bool DmaPusher::Step() {
-    if (!ib_enable || dma_pushbuffer.empty()) {
-        // pushbuffer empty and IB empty or nonexistent - nothing to do
-        return false;
-    }
+    if (dma_get != dma_put) {
+        // Push buffer non-empty, read a word
+        const auto address = gpu.MemoryManager().GpuToCpuAddress(dma_get);
+        ASSERT_MSG(address, "Invalid GPU address");
 
-    const CommandList& command_list{dma_pushbuffer.front()};
-    const CommandListHeader& command_list_header{command_list[dma_pushbuffer_subindex++]};
-    GPUVAddr dma_get = command_list_header.addr;
-    GPUVAddr dma_put = dma_get + command_list_header.size * sizeof(u32);
-    bool non_main = command_list_header.is_non_main;
+        const CommandHeader command_header{Memory::Read32(*address)};
 
-    if (dma_pushbuffer_subindex >= command_list.size()) {
-        // We've gone through the current list, remove it from the queue
-        dma_pushbuffer.pop();
-        dma_pushbuffer_subindex = 0;
-    }
+        dma_get += sizeof(u32);
 
-    if (command_list_header.size == 0) {
-        return true;
-    }
-
-    // Push buffer non-empty, read a word
-    const auto address = gpu.MemoryManager().GpuToCpuAddress(dma_get);
-    ASSERT_MSG(address, "Invalid GPU address");
-
-    command_headers.resize(command_list_header.size);
-
-    Memory::ReadBlock(*address, command_headers.data(), command_list_header.size * sizeof(u32));
-
-    for (const CommandHeader& command_header : command_headers) {
+        if (!non_main) {
+            dma_mget = dma_get;
+        }
 
         // now, see if we're in the middle of a command
         if (dma_state.length_pending) {
@@ -109,11 +91,22 @@ bool DmaPusher::Step() {
                 break;
             }
         }
-    }
+    } else if (ib_enable && !dma_pushbuffer.empty()) {
+        // Current pushbuffer empty, but we have more IB entries to read
+        const CommandList& command_list{dma_pushbuffer.front()};
+        const CommandListHeader& command_list_header{command_list[dma_pushbuffer_subindex++]};
+        dma_get = command_list_header.addr;
+        dma_put = dma_get + command_list_header.size * sizeof(u32);
+        non_main = command_list_header.is_non_main;
 
-    if (!non_main) {
-        // TODO (degasus): This is dead code, as dma_mget is never read.
-        dma_mget = dma_put;
+        if (dma_pushbuffer_subindex >= command_list.size()) {
+            // We've gone through the current list, remove it from the queue
+            dma_pushbuffer.pop();
+            dma_pushbuffer_subindex = 0;
+        }
+    } else {
+        // Otherwise, pushbuffer empty and IB empty or nonexistent - nothing to do
+        return {};
     }
 
     return true;

src/video_core/dma_pusher.h

@@ -75,8 +75,6 @@ private:
 
     GPU& gpu;
 
-    std::vector<CommandHeader> command_headers; ///< Buffer for list of commands fetched at once
-
     std::queue<CommandList> dma_pushbuffer; ///< Queue of command lists to be processed
     std::size_t dma_pushbuffer_subindex{};  ///< Index within a command list within the pushbuffer
 
@@ -91,8 +89,11 @@ private:
     DmaState dma_state{};
     bool dma_increment_once{};
 
+    GPUVAddr dma_put{};   ///< pushbuffer current end address
+    GPUVAddr dma_get{};   ///< pushbuffer current read address
     GPUVAddr dma_mget{};  ///< main pushbuffer last read address
     bool ib_enable{true}; ///< IB mode enabled
+    bool non_main{};      ///< non-main pushbuffer active
 };
 
 } // namespace Tegra