Android #131

Merge PR 11535
Merge pull request #11980 from german77/moment
2023-11-13 00:57:12 +00:00 · 2023-11-13 00:57:11 +00:00 · 2023-11-11 20:02:23 -05:00 · 2023-11-11 20:02:09 -05:00 · 2023-11-11 20:02:01 -05:00 · 2023-11-11 20:01:53 -05:00
33 changed files with 661 additions and 243 deletions
--- a/README.md
+++ b/README.md
@@ -1,3 +1,12 @@
+| Pull Request | Commit | Title | Author | Merged? |
+|----|----|----|----|----|
+| [11535](https://github.com/yuzu-emu/yuzu//pull/11535) | [`50bcfa5fb`](https://github.com/yuzu-emu/yuzu//pull/11535/files) | renderer_vulkan: Introduce separate cmd buffer for uploads | [GPUCode](https://github.com/GPUCode/) | Yes |
+
+
+End of merge log. You can find the original README.md below the break.
+
+-----
+
 <!--
 SPDX-FileCopyrightText: 2018 yuzu Emulator Project
 SPDX-License-Identifier: GPL-2.0-or-later
--- a/src/android/app/src/main/jni/emu_window/emu_window.cpp
+++ b/src/android/app/src/main/jni/emu_window/emu_window.cpp
@@ -9,6 +9,7 @@
 #include "input_common/drivers/virtual_gamepad.h"
 #include "input_common/main.h"
 #include "jni/emu_window/emu_window.h"
+#include "jni/native.h"

 void EmuWindow_Android::OnSurfaceChanged(ANativeWindow* surface) {
    m_window_width = ANativeWindow_getWidth(surface);
@@ -57,6 +58,13 @@ void EmuWindow_Android::OnRemoveNfcTag() {
    m_input_subsystem->GetVirtualAmiibo()->CloseAmiibo();
 }

+void EmuWindow_Android::OnFrameDisplayed() {
+    if (!m_first_frame) {
+        EmulationSession::GetInstance().OnEmulationStarted();
+        m_first_frame = true;
+    }
+}
+
 EmuWindow_Android::EmuWindow_Android(InputCommon::InputSubsystem* input_subsystem,
                                     ANativeWindow* surface,
                                     std::shared_ptr<Common::DynamicLibrary> driver_library)
--- a/src/android/app/src/main/jni/emu_window/emu_window.h
+++ b/src/android/app/src/main/jni/emu_window/emu_window.h
@@ -45,7 +45,7 @@ public:
                              float gyro_z, float accel_x, float accel_y, float accel_z);
    void OnReadNfcTag(std::span<u8> data);
    void OnRemoveNfcTag();
-    void OnFrameDisplayed() override {}
+    void OnFrameDisplayed() override;

    std::unique_ptr<Core::Frontend::GraphicsContext> CreateSharedContext() const override {
        return {std::make_unique<GraphicsContext_Android>(m_driver_library)};
@@ -61,4 +61,6 @@ private:
    float m_window_height{};

    std::shared_ptr<Common::DynamicLibrary> m_driver_library;
+
+    bool m_first_frame = false;
 };
--- a/src/android/app/src/main/jni/native.cpp
+++ b/src/android/app/src/main/jni/native.cpp
@@ -372,8 +372,6 @@ void EmulationSession::RunEmulation() {
        m_system.InitializeDebugger();
    }

-    OnEmulationStarted();
-
    while (true) {
        {
            [[maybe_unused]] std::unique_lock lock(m_mutex);
--- a/src/android/app/src/main/jni/native.h
+++ b/src/android/app/src/main/jni/native.h
@@ -52,9 +52,10 @@ public:
    void OnGamepadDisconnectEvent([[maybe_unused]] int index);
    SoftwareKeyboard::AndroidKeyboard* SoftwareKeyboard();

+    static void OnEmulationStarted();
+
 private:
    static void LoadDiskCacheProgress(VideoCore::LoadCallbackStage stage, int progress, int max);
-    static void OnEmulationStarted();
    static void OnEmulationStopped(Core::SystemResultStatus result);

 private:
--- a/src/common/settings.h
+++ b/src/common/settings.h
@@ -358,6 +358,8 @@ struct Values {
                                          Category::RendererDebug};
    // TODO: remove this once AMDVLK supports VK_EXT_depth_bias_control
    bool renderer_amdvlk_depth_bias_workaround{};
+    Setting<bool> disable_buffer_reorder{linkage, false, "disable_buffer_reorder",
+                                         Category::RendererDebug};

    // System
    SwitchableSetting<Language, true> language_index{linkage,
--- a/src/core/debugger/gdbstub.cpp
+++ b/src/core/debugger/gdbstub.cpp
@@ -562,6 +562,120 @@ static std::string PaginateBuffer(std::string_view buffer, std::string_view requ
    }
 }

+static VAddr GetModuleEnd(Kernel::KProcessPageTable& page_table, VAddr base) {
+    Kernel::KMemoryInfo mem_info;
+    Kernel::Svc::MemoryInfo svc_mem_info;
+    Kernel::Svc::PageInfo page_info;
+    VAddr cur_addr{base};
+
+    // Expect: r-x Code (.text)
+    R_ASSERT(page_table.QueryInfo(std::addressof(mem_info), std::addressof(page_info), cur_addr));
+    svc_mem_info = mem_info.GetSvcMemoryInfo();
+    cur_addr = svc_mem_info.base_address + svc_mem_info.size;
+    if (svc_mem_info.state != Kernel::Svc::MemoryState::Code ||
+        svc_mem_info.permission != Kernel::Svc::MemoryPermission::ReadExecute) {
+        return cur_addr - 1;
+    }
+
+    // Expect: r-- Code (.rodata)
+    R_ASSERT(page_table.QueryInfo(std::addressof(mem_info), std::addressof(page_info), cur_addr));
+    svc_mem_info = mem_info.GetSvcMemoryInfo();
+    cur_addr = svc_mem_info.base_address + svc_mem_info.size;
+    if (svc_mem_info.state != Kernel::Svc::MemoryState::Code ||
+        svc_mem_info.permission != Kernel::Svc::MemoryPermission::Read) {
+        return cur_addr - 1;
+    }
+
+    // Expect: rw- CodeData (.data)
+    R_ASSERT(page_table.QueryInfo(std::addressof(mem_info), std::addressof(page_info), cur_addr));
+    svc_mem_info = mem_info.GetSvcMemoryInfo();
+    cur_addr = svc_mem_info.base_address + svc_mem_info.size;
+    return cur_addr - 1;
+}
+
+static Loader::AppLoader::Modules FindModules(Core::System& system) {
+    Loader::AppLoader::Modules modules;
+
+    auto& page_table = system.ApplicationProcess()->GetPageTable();
+    auto& memory = system.ApplicationMemory();
+    VAddr cur_addr = 0;
+
+    // Look for executable sections in Code or AliasCode regions.
+    while (true) {
+        Kernel::KMemoryInfo mem_info{};
+        Kernel::Svc::PageInfo page_info{};
+        R_ASSERT(
+            page_table.QueryInfo(std::addressof(mem_info), std::addressof(page_info), cur_addr));
+        auto svc_mem_info = mem_info.GetSvcMemoryInfo();
+
+        if (svc_mem_info.permission == Kernel::Svc::MemoryPermission::ReadExecute &&
+            (svc_mem_info.state == Kernel::Svc::MemoryState::Code ||
+             svc_mem_info.state == Kernel::Svc::MemoryState::AliasCode)) {
+            // Try to read the module name from its path.
+            constexpr s32 PathLengthMax = 0x200;
+            struct {
+                u32 zero;
+                s32 path_length;
+                std::array<char, PathLengthMax> path;
+            } module_path;
+
+            if (memory.ReadBlock(svc_mem_info.base_address + svc_mem_info.size, &module_path,
+                                 sizeof(module_path))) {
+                if (module_path.zero == 0 && module_path.path_length > 0) {
+                    // Truncate module name.
+                    module_path.path[PathLengthMax - 1] = '\0';
+
+                    // Ignore leading directories.
+                    char* path_pointer = module_path.path.data();
+
+                    for (s32 i = 0; i < std::min(PathLengthMax, module_path.path_length) &&
+                                    module_path.path[i] != '\0';
+                         i++) {
+                        if (module_path.path[i] == '/' || module_path.path[i] == '\\') {
+                            path_pointer = module_path.path.data() + i + 1;
+                        }
+                    }
+
+                    // Insert output.
+                    modules.emplace(svc_mem_info.base_address, path_pointer);
+                }
+            }
+        }
+
+        // Check if we're done.
+        const uintptr_t next_address = svc_mem_info.base_address + svc_mem_info.size;
+        if (next_address <= cur_addr) {
+            break;
+        }
+
+        cur_addr = next_address;
+    }
+
+    return modules;
+}
+
+static VAddr FindMainModuleEntrypoint(Core::System& system) {
+    Loader::AppLoader::Modules modules;
+    system.GetAppLoader().ReadNSOModules(modules);
+
+    // Do we have a module named main?
+    const auto main = std::find_if(modules.begin(), modules.end(),
+                                   [](const auto& key) { return key.second == "main"; });
+
+    if (main != modules.end()) {
+        return main->first;
+    }
+
+    // Do we have any loaded executable sections?
+    modules = FindModules(system);
+    if (!modules.empty()) {
+        return modules.begin()->first;
+    }
+
+    // As a last resort, use the start of the code region.
+    return GetInteger(system.ApplicationProcess()->GetPageTable().GetCodeRegionStart());
+}
+
 void GDBStub::HandleQuery(std::string_view command) {
    if (command.starts_with("TStatus")) {
        // no tracepoint support
@@ -573,21 +687,10 @@ void GDBStub::HandleQuery(std::string_view command) {
        const auto target_xml{arch->GetTargetXML()};
        SendReply(PaginateBuffer(target_xml, command.substr(30)));
    } else if (command.starts_with("Offsets")) {
-        Loader::AppLoader::Modules modules;
-        system.GetAppLoader().ReadNSOModules(modules);
-
-        const auto main = std::find_if(modules.begin(), modules.end(),
-                                       [](const auto& key) { return key.second == "main"; });
-        if (main != modules.end()) {
-            SendReply(fmt::format("TextSeg={:x}", main->first));
-        } else {
-            SendReply(fmt::format(
-                "TextSeg={:x}",
-                GetInteger(system.ApplicationProcess()->GetPageTable().GetCodeRegionStart())));
-        }
+        const auto main_offset = FindMainModuleEntrypoint(system);
+        SendReply(fmt::format("TextSeg={:x}", main_offset));
    } else if (command.starts_with("Xfer:libraries:read::")) {
-        Loader::AppLoader::Modules modules;
-        system.GetAppLoader().ReadNSOModules(modules);
+        auto modules = FindModules(system);

        std::string buffer;
        buffer += R"(<?xml version="1.0"?>)";
@@ -727,37 +830,6 @@ static constexpr const char* GetMemoryPermissionString(const Kernel::Svc::Memory
    }
 }

-static VAddr GetModuleEnd(Kernel::KProcessPageTable& page_table, VAddr base) {
-    Kernel::KMemoryInfo mem_info;
-    Kernel::Svc::MemoryInfo svc_mem_info;
-    Kernel::Svc::PageInfo page_info;
-    VAddr cur_addr{base};
-
-    // Expect: r-x Code (.text)
-    R_ASSERT(page_table.QueryInfo(std::addressof(mem_info), std::addressof(page_info), cur_addr));
-    svc_mem_info = mem_info.GetSvcMemoryInfo();
-    cur_addr = svc_mem_info.base_address + svc_mem_info.size;
-    if (svc_mem_info.state != Kernel::Svc::MemoryState::Code ||
-        svc_mem_info.permission != Kernel::Svc::MemoryPermission::ReadExecute) {
-        return cur_addr - 1;
-    }
-
-    // Expect: r-- Code (.rodata)
-    R_ASSERT(page_table.QueryInfo(std::addressof(mem_info), std::addressof(page_info), cur_addr));
-    svc_mem_info = mem_info.GetSvcMemoryInfo();
-    cur_addr = svc_mem_info.base_address + svc_mem_info.size;
-    if (svc_mem_info.state != Kernel::Svc::MemoryState::Code ||
-        svc_mem_info.permission != Kernel::Svc::MemoryPermission::Read) {
-        return cur_addr - 1;
-    }
-
-    // Expect: rw- CodeData (.data)
-    R_ASSERT(page_table.QueryInfo(std::addressof(mem_info), std::addressof(page_info), cur_addr));
-    svc_mem_info = mem_info.GetSvcMemoryInfo();
-    cur_addr = svc_mem_info.base_address + svc_mem_info.size;
-    return cur_addr - 1;
-}
-
 void GDBStub::HandleRcmd(const std::vector<u8>& command) {
    std::string_view command_str{reinterpret_cast<const char*>(&command[0]), command.size()};
    std::string reply;
@@ -784,8 +856,7 @@ void GDBStub::HandleRcmd(const std::vector<u8>& command) {
            reply = "Fastmem is not enabled.\n";
        }
    } else if (command_str == "get info") {
-        Loader::AppLoader::Modules modules;
-        system.GetAppLoader().ReadNSOModules(modules);
+        auto modules = FindModules(system);

        reply = fmt::format("Process:     {:#x} ({})\n"
                            "Program Id:  {:#018x}\n",
--- a/src/core/hle/kernel/k_capabilities.cpp
+++ b/src/core/hle/kernel/k_capabilities.cpp
@@ -5,6 +5,7 @@
 #include "core/hle/kernel/k_capabilities.h"
 #include "core/hle/kernel/k_memory_layout.h"
 #include "core/hle/kernel/k_process_page_table.h"
+#include "core/hle/kernel/k_trace.h"
 #include "core/hle/kernel/kernel.h"
 #include "core/hle/kernel/svc_results.h"
 #include "core/hle/kernel/svc_version.h"
@@ -329,6 +330,8 @@ Result KCapabilities::SetCapabilities(std::span<const u32> caps, KProcessPageTab

            // Map the range.
            R_TRY(this->MapRange_(cap, size_cap, page_table));
+        } else if (GetCapabilityType(cap) == CapabilityType::MapRegion && !IsKTraceEnabled) {
+            continue;
        } else {
            R_TRY(this->SetCapability(cap, set_flags, set_svc, page_table));
        }
--- a/src/core/hle/service/hid/irs.cpp
+++ b/src/core/hle/service/hid/irs.cpp
@@ -138,7 +138,7 @@ void IRS::RunMomentProcessor(HLERequestContext& ctx) {

    if (result.IsSuccess()) {
        auto& device = GetIrCameraSharedMemoryDeviceEntry(parameters.camera_handle);
-        MakeProcessor<MomentProcessor>(parameters.camera_handle, device);
+        MakeProcessorWithCoreContext<MomentProcessor>(parameters.camera_handle, device);
        auto& image_transfer_processor = GetProcessor<MomentProcessor>(parameters.camera_handle);
        image_transfer_processor.SetConfig(parameters.processor_config);
        npad_device->SetPollingMode(Core::HID::EmulatedDeviceIndex::RightIndex,
--- a/src/core/hle/service/hid/irsensor/clustering_processor.cpp
+++ b/src/core/hle/service/hid/irsensor/clustering_processor.cpp
@@ -3,16 +3,18 @@

 #include <queue>

+#include "core/core.h"
+#include "core/core_timing.h"
 #include "core/hid/emulated_controller.h"
 #include "core/hid/hid_core.h"
 #include "core/hle/service/hid/irsensor/clustering_processor.h"

 namespace Service::IRS {
-ClusteringProcessor::ClusteringProcessor(Core::HID::HIDCore& hid_core_,
+ClusteringProcessor::ClusteringProcessor(Core::System& system_,
                                         Core::IrSensor::DeviceFormat& device_format,
                                         std::size_t npad_index)
-    : device{device_format} {
-    npad_device = hid_core_.GetEmulatedControllerByIndex(npad_index);
+    : device{device_format}, system{system_} {
+    npad_device = system.HIDCore().GetEmulatedControllerByIndex(npad_index);

    device.mode = Core::IrSensor::IrSensorMode::ClusteringProcessor;
    device.camera_status = Core::IrSensor::IrCameraStatus::Unconnected;
@@ -48,7 +50,7 @@ void ClusteringProcessor::OnControllerUpdate(Core::HID::ControllerTriggerType ty
    }

    next_state = {};
-    const auto camera_data = npad_device->GetCamera();
+    const auto& camera_data = npad_device->GetCamera();
    auto filtered_image = camera_data.data;

    RemoveLowIntensityData(filtered_image);
@@ -83,7 +85,7 @@ void ClusteringProcessor::OnControllerUpdate(Core::HID::ControllerTriggerType ty
    }

    next_state.sampling_number = camera_data.sample;
-    next_state.timestamp = next_state.timestamp + 131;
+    next_state.timestamp = system.CoreTiming().GetGlobalTimeNs().count();
    next_state.ambient_noise_level = Core::IrSensor::CameraAmbientNoiseLevel::Low;
    shared_memory->clustering_lifo.WriteNextEntry(next_state);

@@ -202,14 +204,14 @@ ClusteringProcessor::ClusteringData ClusteringProcessor::MergeCluster(
 }

 u8 ClusteringProcessor::GetPixel(const std::vector<u8>& data, std::size_t x, std::size_t y) const {
-    if ((y * width) + x > data.size()) {
+    if ((y * width) + x >= data.size()) {
        return 0;
    }
    return data[(y * width) + x];
 }

 void ClusteringProcessor::SetPixel(std::vector<u8>& data, std::size_t x, std::size_t y, u8 value) {
-    if ((y * width) + x > data.size()) {
+    if ((y * width) + x >= data.size()) {
        return;
    }
    data[(y * width) + x] = value;
--- a/src/core/hle/service/hid/irsensor/clustering_processor.h
+++ b/src/core/hle/service/hid/irsensor/clustering_processor.h
@@ -8,6 +8,10 @@
 #include "core/hle/service/hid/irs_ring_lifo.h"
 #include "core/hle/service/hid/irsensor/processor_base.h"

+namespace Core {
+class System;
+}
+
 namespace Core::HID {
 class EmulatedController;
 } // namespace Core::HID
@@ -15,8 +19,7 @@ class EmulatedController;
 namespace Service::IRS {
 class ClusteringProcessor final : public ProcessorBase {
 public:
-    explicit ClusteringProcessor(Core::HID::HIDCore& hid_core_,
-                                 Core::IrSensor::DeviceFormat& device_format,
+    explicit ClusteringProcessor(Core::System& system_, Core::IrSensor::DeviceFormat& device_format,
                                 std::size_t npad_index);
    ~ClusteringProcessor() override;

@@ -106,5 +109,7 @@ private:
    Core::IrSensor::DeviceFormat& device;
    Core::HID::EmulatedController* npad_device;
    int callback_key{};
+
+    Core::System& system;
 };
 } // namespace Service::IRS
--- a/src/core/hle/service/hid/irsensor/image_transfer_processor.cpp
+++ b/src/core/hle/service/hid/irsensor/image_transfer_processor.cpp
@@ -49,7 +49,7 @@ void ImageTransferProcessor::OnControllerUpdate(Core::HID::ControllerTriggerType
        return;
    }

-    const auto camera_data = npad_device->GetCamera();
+    const auto& camera_data = npad_device->GetCamera();

    // This indicates how much ambient light is present
    processor_state.ambient_noise_level = Core::IrSensor::CameraAmbientNoiseLevel::Low;
--- a/src/core/hle/service/hid/irsensor/moment_processor.cpp
+++ b/src/core/hle/service/hid/irsensor/moment_processor.cpp
@@ -1,24 +1,137 @@
 // SPDX-FileCopyrightText: Copyright 2022 yuzu Emulator Project
 // SPDX-License-Identifier: GPL-3.0-or-later

+#include "core/core.h"
+#include "core/core_timing.h"
+#include "core/hid/emulated_controller.h"
+#include "core/hid/hid_core.h"
 #include "core/hle/service/hid/irsensor/moment_processor.h"

 namespace Service::IRS {
-MomentProcessor::MomentProcessor(Core::IrSensor::DeviceFormat& device_format)
-    : device(device_format) {
+static constexpr auto format = Core::IrSensor::ImageTransferProcessorFormat::Size40x30;
+static constexpr std::size_t ImageWidth = 40;
+static constexpr std::size_t ImageHeight = 30;
+
+MomentProcessor::MomentProcessor(Core::System& system_, Core::IrSensor::DeviceFormat& device_format,
+                                 std::size_t npad_index)
+    : device(device_format), system{system_} {
+    npad_device = system.HIDCore().GetEmulatedControllerByIndex(npad_index);
+
    device.mode = Core::IrSensor::IrSensorMode::MomentProcessor;
    device.camera_status = Core::IrSensor::IrCameraStatus::Unconnected;
    device.camera_internal_status = Core::IrSensor::IrCameraInternalStatus::Stopped;
+
+    shared_memory = std::construct_at(
+        reinterpret_cast<MomentSharedMemory*>(&device_format.state.processor_raw_data));
+
+    Core::HID::ControllerUpdateCallback engine_callback{
+        .on_change = [this](Core::HID::ControllerTriggerType type) { OnControllerUpdate(type); },
+        .is_npad_service = true,
+    };
+    callback_key = npad_device->SetCallback(engine_callback);
 }

-MomentProcessor::~MomentProcessor() = default;
+MomentProcessor::~MomentProcessor() {
+    npad_device->DeleteCallback(callback_key);
+};

-void MomentProcessor::StartProcessor() {}
+void MomentProcessor::StartProcessor() {
+    device.camera_status = Core::IrSensor::IrCameraStatus::Available;
+    device.camera_internal_status = Core::IrSensor::IrCameraInternalStatus::Ready;
+}

 void MomentProcessor::SuspendProcessor() {}

 void MomentProcessor::StopProcessor() {}

+void MomentProcessor::OnControllerUpdate(Core::HID::ControllerTriggerType type) {
+    if (type != Core::HID::ControllerTriggerType::IrSensor) {
+        return;
+    }
+
+    next_state = {};
+    const auto& camera_data = npad_device->GetCamera();
+
+    const auto window_width = static_cast<std::size_t>(current_config.window_of_interest.width);
+    const auto window_height = static_cast<std::size_t>(current_config.window_of_interest.height);
+    const auto window_start_x = static_cast<std::size_t>(current_config.window_of_interest.x);
+    const auto window_start_y = static_cast<std::size_t>(current_config.window_of_interest.y);
+
+    const std::size_t block_width = window_width / Columns;
+    const std::size_t block_height = window_height / Rows;
+
+    for (std::size_t row = 0; row < Rows; row++) {
+        for (std::size_t column = 0; column < Columns; column++) {
+            const size_t x_pos = (column * block_width) + window_start_x;
+            const size_t y_pos = (row * block_height) + window_start_y;
+            auto& statistic = next_state.statistic[column + (row * Columns)];
+            statistic = GetStatistic(camera_data.data, x_pos, y_pos, block_width, block_height);
+        }
+    }
+
+    next_state.sampling_number = camera_data.sample;
+    next_state.timestamp = system.CoreTiming().GetGlobalTimeNs().count();
+    next_state.ambient_noise_level = Core::IrSensor::CameraAmbientNoiseLevel::Low;
+    shared_memory->moment_lifo.WriteNextEntry(next_state);
+
+    if (!IsProcessorActive()) {
+        StartProcessor();
+    }
+}
+
+u8 MomentProcessor::GetPixel(const std::vector<u8>& data, std::size_t x, std::size_t y) const {
+    if ((y * ImageWidth) + x >= data.size()) {
+        return 0;
+    }
+    return data[(y * ImageWidth) + x];
+}
+
+MomentProcessor::MomentStatistic MomentProcessor::GetStatistic(const std::vector<u8>& data,
+                                                               std::size_t start_x,
+                                                               std::size_t start_y,
+                                                               std::size_t width,
+                                                               std::size_t height) const {
+    // The actual implementation is always 320x240
+    static constexpr std::size_t RealWidth = 320;
+    static constexpr std::size_t RealHeight = 240;
+    static constexpr std::size_t Threshold = 30;
+    MomentStatistic statistic{};
+    std::size_t active_points{};
+
+    // Sum all data points on the block that meet with the threshold
+    for (std::size_t y = 0; y < width; y++) {
+        for (std::size_t x = 0; x < height; x++) {
+            const size_t x_pos = x + start_x;
+            const size_t y_pos = y + start_y;
+            const auto pixel =
+                GetPixel(data, x_pos * ImageWidth / RealWidth, y_pos * ImageHeight / RealHeight);
+
+            if (pixel < Threshold) {
+                continue;
+            }
+
+            statistic.average_intensity += pixel;
+
+            statistic.centroid.x += static_cast<float>(x_pos);
+            statistic.centroid.y += static_cast<float>(y_pos);
+
+            active_points++;
+        }
+    }
+
+    // Return an empty field if no points were available
+    if (active_points == 0) {
+        return {};
+    }
+
+    // Finally calculate the actual centroid and average intensity
+    statistic.centroid.x /= static_cast<float>(active_points);
+    statistic.centroid.y /= static_cast<float>(active_points);
+    statistic.average_intensity /= static_cast<f32>(width * height);
+
+    return statistic;
+}
+
 void MomentProcessor::SetConfig(Core::IrSensor::PackedMomentProcessorConfig config) {
    current_config.camera_config.exposure_time = config.camera_config.exposure_time;
    current_config.camera_config.gain = config.camera_config.gain;
@@ -29,6 +142,8 @@ void MomentProcessor::SetConfig(Core::IrSensor::PackedMomentProcessorConfig conf
    current_config.preprocess =
        static_cast<Core::IrSensor::MomentProcessorPreprocess>(config.preprocess);
    current_config.preprocess_intensity_threshold = config.preprocess_intensity_threshold;
+
+    npad_device->SetCameraFormat(format);
 }

 } // namespace Service::IRS
--- a/src/core/hle/service/hid/irsensor/moment_processor.h
+++ b/src/core/hle/service/hid/irsensor/moment_processor.h
@@ -6,12 +6,22 @@
 #include "common/bit_field.h"
 #include "common/common_types.h"
 #include "core/hid/irs_types.h"
+#include "core/hle/service/hid/irs_ring_lifo.h"
 #include "core/hle/service/hid/irsensor/processor_base.h"

+namespace Core {
+class System;
+}
+
+namespace Core::HID {
+class EmulatedController;
+} // namespace Core::HID
+
 namespace Service::IRS {
 class MomentProcessor final : public ProcessorBase {
 public:
-    explicit MomentProcessor(Core::IrSensor::DeviceFormat& device_format);
+    explicit MomentProcessor(Core::System& system_, Core::IrSensor::DeviceFormat& device_format,
+                             std::size_t npad_index);
    ~MomentProcessor() override;

    // Called when the processor is initialized
@@ -27,6 +37,9 @@ public:
    void SetConfig(Core::IrSensor::PackedMomentProcessorConfig config);

 private:
+    static constexpr std::size_t Columns = 8;
+    static constexpr std::size_t Rows = 6;
+
    // This is nn::irsensor::MomentProcessorConfig
    struct MomentProcessorConfig {
        Core::IrSensor::CameraConfig camera_config;
@@ -50,12 +63,29 @@ private:
        u64 timestamp;
        Core::IrSensor::CameraAmbientNoiseLevel ambient_noise_level;
        INSERT_PADDING_BYTES(4);
-        std::array<MomentStatistic, 0x30> stadistic;
+        std::array<MomentStatistic, Columns * Rows> statistic;
    };
    static_assert(sizeof(MomentProcessorState) == 0x258, "MomentProcessorState is an invalid size");

+    struct MomentSharedMemory {
+        Service::IRS::Lifo<MomentProcessorState, 6> moment_lifo;
+    };
+    static_assert(sizeof(MomentSharedMemory) == 0xE20, "MomentSharedMemory is an invalid size");
+
+    void OnControllerUpdate(Core::HID::ControllerTriggerType type);
+    u8 GetPixel(const std::vector<u8>& data, std::size_t x, std::size_t y) const;
+    MomentStatistic GetStatistic(const std::vector<u8>& data, std::size_t start_x,
+                                 std::size_t start_y, std::size_t width, std::size_t height) const;
+
+    MomentSharedMemory* shared_memory = nullptr;
+    MomentProcessorState next_state{};
+
    MomentProcessorConfig current_config{};
    Core::IrSensor::DeviceFormat& device;
+    Core::HID::EmulatedController* npad_device;
+    int callback_key{};
+
+    Core::System& system;
 };

 } // namespace Service::IRS
--- a/src/video_core/CMakeLists.txt
+++ b/src/video_core/CMakeLists.txt
@@ -15,6 +15,7 @@ add_library(video_core STATIC
    buffer_cache/buffer_cache.cpp
    buffer_cache/buffer_cache.h
    buffer_cache/memory_tracker_base.h
+    buffer_cache/usage_tracker.h
    buffer_cache/word_manager.h
    cache_types.h
    cdma_pusher.cpp
--- a/src/video_core/buffer_cache/buffer_cache.h
+++ b/src/video_core/buffer_cache/buffer_cache.h
@@ -67,6 +67,7 @@ void BufferCache<P>::TickFrame() {
    if (!channel_state) {
        return;
    }
+    runtime.TickFrame(slot_buffers);

    // Calculate hits and shots and move hit bits to the right
    const u32 hits = std::reduce(channel_state->uniform_cache_hits.begin(),
@@ -230,7 +231,10 @@ bool BufferCache<P>::DMACopy(GPUVAddr src_address, GPUVAddr dest_address, u64 am
    for (const IntervalType& add_interval : tmp_intervals) {
        common_ranges.add(add_interval);
    }
-    runtime.CopyBuffer(dest_buffer, src_buffer, copies);
+    const auto& copy = copies[0];
+    src_buffer.MarkUsage(copy.src_offset, copy.size);
+    dest_buffer.MarkUsage(copy.dst_offset, copy.size);
+    runtime.CopyBuffer(dest_buffer, src_buffer, copies, true);
    if (has_new_downloads) {
        memory_tracker.MarkRegionAsGpuModified(*cpu_dest_address, amount);
    }
@@ -258,9 +262,10 @@ bool BufferCache<P>::DMAClear(GPUVAddr dst_address, u64 amount, u32 value) {
    common_ranges.subtract(subtract_interval);

    const BufferId buffer = FindBuffer(*cpu_dst_address, static_cast<u32>(size));
-    auto& dest_buffer = slot_buffers[buffer];
+    Buffer& dest_buffer = slot_buffers[buffer];
    const u32 offset = dest_buffer.Offset(*cpu_dst_address);
    runtime.ClearBuffer(dest_buffer, offset, size, value);
+    dest_buffer.MarkUsage(offset, size);
    return true;
 }

@@ -603,6 +608,7 @@ void BufferCache<P>::CommitAsyncFlushesHigh() {
            VAddr orig_cpu_addr = static_cast<VAddr>(second_copy.src_offset);
            const IntervalType base_interval{orig_cpu_addr, orig_cpu_addr + copy.size};
            async_downloads += std::make_pair(base_interval, 1);
+            buffer.MarkUsage(copy.src_offset, copy.size);
            runtime.CopyBuffer(download_staging.buffer, buffer, copies, false);
            normalized_copies.push_back(second_copy);
        }
@@ -621,8 +627,9 @@ void BufferCache<P>::CommitAsyncFlushesHigh() {
                    // Have in mind the staging buffer offset for the copy
                    copy.dst_offset += download_staging.offset;
                    const std::array copies{copy};
-                    runtime.CopyBuffer(download_staging.buffer, slot_buffers[buffer_id], copies,
-                                       false);
+                    Buffer& buffer = slot_buffers[buffer_id];
+                    buffer.MarkUsage(copy.src_offset, copy.size);
+                    runtime.CopyBuffer(download_staging.buffer, buffer, copies, false);
                }
                runtime.PostCopyBarrier();
                runtime.Finish();
@@ -742,7 +749,7 @@ void BufferCache<P>::BindHostIndexBuffer() {
                {BufferCopy{.src_offset = upload_staging.offset, .dst_offset = 0, .size = size}}};
            std::memcpy(upload_staging.mapped_span.data(),
                        draw_state.inline_index_draw_indexes.data(), size);
-            runtime.CopyBuffer(buffer, upload_staging.buffer, copies);
+            runtime.CopyBuffer(buffer, upload_staging.buffer, copies, true);
        } else {
            buffer.ImmediateUpload(0, draw_state.inline_index_draw_indexes);
        }
@@ -754,6 +761,7 @@ void BufferCache<P>::BindHostIndexBuffer() {
            offset + draw_state.index_buffer.first * draw_state.index_buffer.FormatSizeInBytes();
        runtime.BindIndexBuffer(buffer, new_offset, size);
    } else {
+        buffer.MarkUsage(offset, size);
        runtime.BindIndexBuffer(draw_state.topology, draw_state.index_buffer.format,
                                draw_state.index_buffer.first, draw_state.index_buffer.count,
                                buffer, offset, size);
@@ -790,6 +798,7 @@ void BufferCache<P>::BindHostVertexBuffers() {

            const u32 stride = maxwell3d->regs.vertex_streams[index].stride;
            const u32 offset = buffer.Offset(binding.cpu_addr);
+            buffer.MarkUsage(offset, binding.size);

            host_bindings.buffers.push_back(&buffer);
            host_bindings.offsets.push_back(offset);
@@ -895,6 +904,7 @@ void BufferCache<P>::BindHostGraphicsUniformBuffer(size_t stage, u32 index, u32
    if constexpr (HAS_PERSISTENT_UNIFORM_BUFFER_BINDINGS) {
        channel_state->uniform_buffer_binding_sizes[stage][binding_index] = size;
    }
+    buffer.MarkUsage(offset, size);
    if constexpr (NEEDS_BIND_UNIFORM_INDEX) {
        runtime.BindUniformBuffer(stage, binding_index, buffer, offset, size);
    } else {
@@ -913,6 +923,7 @@ void BufferCache<P>::BindHostGraphicsStorageBuffers(size_t stage) {
        SynchronizeBuffer(buffer, binding.cpu_addr, size);

        const u32 offset = buffer.Offset(binding.cpu_addr);
+        buffer.MarkUsage(offset, size);
        const bool is_written = ((channel_state->written_storage_buffers[stage] >> index) & 1) != 0;

        if (is_written) {
@@ -943,6 +954,7 @@ void BufferCache<P>::BindHostGraphicsTextureBuffers(size_t stage) {

        const u32 offset = buffer.Offset(binding.cpu_addr);
        const PixelFormat format = binding.format;
+        buffer.MarkUsage(offset, size);
        if constexpr (SEPARATE_IMAGE_BUFFERS_BINDINGS) {
            if (((channel_state->image_texture_buffers[stage] >> index) & 1) != 0) {
                runtime.BindImageBuffer(buffer, offset, size, format);
@@ -975,9 +987,10 @@ void BufferCache<P>::BindHostTransformFeedbackBuffers() {
        MarkWrittenBuffer(binding.buffer_id, binding.cpu_addr, size);

        const u32 offset = buffer.Offset(binding.cpu_addr);
+        buffer.MarkUsage(offset, size);
        host_bindings.buffers.push_back(&buffer);
        host_bindings.offsets.push_back(offset);
-        host_bindings.sizes.push_back(binding.size);
+        host_bindings.sizes.push_back(size);
    }
    if (host_bindings.buffers.size() > 0) {
        runtime.BindTransformFeedbackBuffers(host_bindings);
@@ -1001,6 +1014,7 @@ void BufferCache<P>::BindHostComputeUniformBuffers() {
        SynchronizeBuffer(buffer, binding.cpu_addr, size);

        const u32 offset = buffer.Offset(binding.cpu_addr);
+        buffer.MarkUsage(offset, size);
        if constexpr (NEEDS_BIND_UNIFORM_INDEX) {
            runtime.BindComputeUniformBuffer(binding_index, buffer, offset, size);
            ++binding_index;
@@ -1021,6 +1035,7 @@ void BufferCache<P>::BindHostComputeStorageBuffers() {
        SynchronizeBuffer(buffer, binding.cpu_addr, size);

        const u32 offset = buffer.Offset(binding.cpu_addr);
+        buffer.MarkUsage(offset, size);
        const bool is_written =
            ((channel_state->written_compute_storage_buffers >> index) & 1) != 0;

@@ -1053,6 +1068,7 @@ void BufferCache<P>::BindHostComputeTextureBuffers() {

        const u32 offset = buffer.Offset(binding.cpu_addr);
        const PixelFormat format = binding.format;
+        buffer.MarkUsage(offset, size);
        if constexpr (SEPARATE_IMAGE_BUFFERS_BINDINGS) {
            if (((channel_state->image_compute_texture_buffers >> index) & 1) != 0) {
                runtime.BindImageBuffer(buffer, offset, size, format);
@@ -1172,10 +1188,11 @@ void BufferCache<P>::UpdateVertexBuffer(u32 index) {
    if (!gpu_memory->IsWithinGPUAddressRange(gpu_addr_end)) {
        size = static_cast<u32>(gpu_memory->MaxContinuousRange(gpu_addr_begin, size));
    }
+    const BufferId buffer_id = FindBuffer(*cpu_addr, size);
    channel_state->vertex_buffers[index] = Binding{
        .cpu_addr = *cpu_addr,
        .size = size,
-        .buffer_id = FindBuffer(*cpu_addr, size),
+        .buffer_id = buffer_id,
    };
 }

@@ -1406,7 +1423,8 @@ void BufferCache<P>::JoinOverlap(BufferId new_buffer_id, BufferId overlap_id,
        .dst_offset = dst_base_offset,
        .size = overlap.SizeBytes(),
    });
-    runtime.CopyBuffer(new_buffer, overlap, copies);
+    new_buffer.MarkUsage(copies[0].dst_offset, copies[0].size);
+    runtime.CopyBuffer(new_buffer, overlap, copies, true);
    DeleteBuffer(overlap_id, true);
 }

@@ -1419,7 +1437,9 @@ BufferId BufferCache<P>::CreateBuffer(VAddr cpu_addr, u32 wanted_size) {
    const u32 size = static_cast<u32>(overlap.end - overlap.begin);
    const BufferId new_buffer_id = slot_buffers.insert(runtime, rasterizer, overlap.begin, size);
    auto& new_buffer = slot_buffers[new_buffer_id];
-    runtime.ClearBuffer(new_buffer, 0, new_buffer.SizeBytes(), 0);
+    const size_t size_bytes = new_buffer.SizeBytes();
+    runtime.ClearBuffer(new_buffer, 0, size_bytes, 0);
+    new_buffer.MarkUsage(0, size_bytes);
    for (const BufferId overlap_id : overlap.ids) {
        JoinOverlap(new_buffer_id, overlap_id, !overlap.has_stream_leap);
    }
@@ -1472,11 +1492,6 @@ void BufferCache<P>::TouchBuffer(Buffer& buffer, BufferId buffer_id) noexcept {

 template <class P>
 bool BufferCache<P>::SynchronizeBuffer(Buffer& buffer, VAddr cpu_addr, u32 size) {
-    return SynchronizeBufferImpl(buffer, cpu_addr, size);
-}
-
-template <class P>
-bool BufferCache<P>::SynchronizeBufferImpl(Buffer& buffer, VAddr cpu_addr, u32 size) {
    boost::container::small_vector<BufferCopy, 4> copies;
    u64 total_size_bytes = 0;
    u64 largest_copy = 0;
@@ -1498,51 +1513,6 @@ bool BufferCache<P>::SynchronizeBufferImpl(Buffer& buffer, VAddr cpu_addr, u32 s
    return false;
 }

-template <class P>
-bool BufferCache<P>::SynchronizeBufferNoModified(Buffer& buffer, VAddr cpu_addr, u32 size) {
-    boost::container::small_vector<BufferCopy, 4> copies;
-    u64 total_size_bytes = 0;
-    u64 largest_copy = 0;
-    IntervalSet found_sets{};
-    auto make_copies = [&] {
-        for (auto& interval : found_sets) {
-            const std::size_t sub_size = interval.upper() - interval.lower();
-            const VAddr cpu_addr_ = interval.lower();
-            copies.push_back(BufferCopy{
-                .src_offset = total_size_bytes,
-                .dst_offset = cpu_addr_ - buffer.CpuAddr(),
-                .size = sub_size,
-            });
-            total_size_bytes += sub_size;
-            largest_copy = std::max<u64>(largest_copy, sub_size);
-        }
-        const std::span<BufferCopy> copies_span(copies.data(), copies.size());
-        UploadMemory(buffer, total_size_bytes, largest_copy, copies_span);
-    };
-    memory_tracker.ForEachUploadRange(cpu_addr, size, [&](u64 cpu_addr_out, u64 range_size) {
-        const VAddr base_adr = cpu_addr_out;
-        const VAddr end_adr = base_adr + range_size;
-        const IntervalType add_interval{base_adr, end_adr};
-        found_sets.add(add_interval);
-    });
-    if (found_sets.empty()) {
-        return true;
-    }
-    const IntervalType search_interval{cpu_addr, cpu_addr + size};
-    auto it = common_ranges.lower_bound(search_interval);
-    auto it_end = common_ranges.upper_bound(search_interval);
-    if (it == common_ranges.end()) {
-        make_copies();
-        return false;
-    }
-    while (it != it_end) {
-        found_sets.subtract(*it);
-        it++;
-    }
-    make_copies();
-    return false;
-}
-
 template <class P>
 void BufferCache<P>::UploadMemory(Buffer& buffer, u64 total_size_bytes, u64 largest_copy,
                                  std::span<BufferCopy> copies) {
@@ -1591,7 +1561,8 @@ void BufferCache<P>::MappedUploadMemory([[maybe_unused]] Buffer& buffer,
            // Apply the staging offset
            copy.src_offset += upload_staging.offset;
        }
-        runtime.CopyBuffer(buffer, upload_staging.buffer, copies);
+        const bool can_reorder = runtime.CanReorderUpload(buffer, copies);
+        runtime.CopyBuffer(buffer, upload_staging.buffer, copies, true, can_reorder);
    }
 }

@@ -1633,7 +1604,8 @@ void BufferCache<P>::InlineMemoryImplementation(VAddr dest_address, size_t copy_
        }};
        u8* const src_pointer = upload_staging.mapped_span.data();
        std::memcpy(src_pointer, inlined_buffer.data(), copy_size);
-        runtime.CopyBuffer(buffer, upload_staging.buffer, copies);
+        const bool can_reorder = runtime.CanReorderUpload(buffer, copies);
+        runtime.CopyBuffer(buffer, upload_staging.buffer, copies, true, can_reorder);
    } else {
        buffer.ImmediateUpload(buffer.Offset(dest_address), inlined_buffer.first(copy_size));
    }
@@ -1686,8 +1658,9 @@ void BufferCache<P>::DownloadBufferMemory(Buffer& buffer, VAddr cpu_addr, u64 si
        for (BufferCopy& copy : copies) {
            // Modify copies to have the staging offset in mind
            copy.dst_offset += download_staging.offset;
+            buffer.MarkUsage(copy.src_offset, copy.size);
        }
-        runtime.CopyBuffer(download_staging.buffer, buffer, copies_span);
+        runtime.CopyBuffer(download_staging.buffer, buffer, copies_span, true);
        runtime.Finish();
        for (const BufferCopy& copy : copies) {
            const VAddr copy_cpu_addr = buffer.CpuAddr() + copy.src_offset;
--- a/src/video_core/buffer_cache/buffer_cache_base.h
+++ b/src/video_core/buffer_cache/buffer_cache_base.h
@@ -529,10 +529,6 @@ private:

    bool SynchronizeBuffer(Buffer& buffer, VAddr cpu_addr, u32 size);

-    bool SynchronizeBufferImpl(Buffer& buffer, VAddr cpu_addr, u32 size);
-
-    bool SynchronizeBufferNoModified(Buffer& buffer, VAddr cpu_addr, u32 size);
-
    void UploadMemory(Buffer& buffer, u64 total_size_bytes, u64 largest_copy,
                      std::span<BufferCopy> copies);

--- a/src/video_core/buffer_cache/usage_tracker.h
+++ b/src/video_core/buffer_cache/usage_tracker.h
@@ -0,0 +1,79 @@
+// SPDX-FileCopyrightText: Copyright 2023 yuzu Emulator Project
+// SPDX-License-Identifier: GPL-3.0-or-later
+
+#pragma once
+
+#include "common/alignment.h"
+#include "common/common_types.h"
+
+namespace VideoCommon {
+
+class UsageTracker {
+    static constexpr size_t BYTES_PER_BIT_SHIFT = 6;
+    static constexpr size_t PAGE_SHIFT = 6 + BYTES_PER_BIT_SHIFT;
+    static constexpr size_t PAGE_BYTES = 1 << PAGE_SHIFT;
+
+public:
+    explicit UsageTracker(size_t size) {
+        const size_t num_pages = (size >> PAGE_SHIFT) + 1;
+        pages.resize(num_pages, 0ULL);
+    }
+
+    void Reset() noexcept {
+        std::ranges::fill(pages, 0ULL);
+    }
+
+    void Track(u64 offset, u64 size) noexcept {
+        const size_t page = offset >> PAGE_SHIFT;
+        const size_t page_end = (offset + size) >> PAGE_SHIFT;
+        TrackPage(page, offset, size);
+        if (page == page_end) {
+            return;
+        }
+        for (size_t i = page + 1; i < page_end; i++) {
+            pages[i] = ~u64{0};
+        }
+        const size_t offset_end = offset + size;
+        const size_t offset_end_page_aligned = Common::AlignDown(offset_end, PAGE_BYTES);
+        TrackPage(page_end, offset_end_page_aligned, offset_end - offset_end_page_aligned);
+    }
+
+    [[nodiscard]] bool IsUsed(u64 offset, u64 size) const noexcept {
+        const size_t page = offset >> PAGE_SHIFT;
+        const size_t page_end = (offset + size) >> PAGE_SHIFT;
+        if (IsPageUsed(page, offset, size)) {
+            return true;
+        }
+        for (size_t i = page + 1; i < page_end; i++) {
+            if (pages[i] != 0) {
+                return true;
+            }
+        }
+        const size_t offset_end = offset + size;
+        const size_t offset_end_page_aligned = Common::AlignDown(offset_end, PAGE_BYTES);
+        return IsPageUsed(page_end, offset_end_page_aligned, offset_end - offset_end_page_aligned);
+    }
+
+private:
+    void TrackPage(u64 page, u64 offset, u64 size) noexcept {
+        const size_t offset_in_page = offset % PAGE_BYTES;
+        const size_t first_bit = offset_in_page >> BYTES_PER_BIT_SHIFT;
+        const size_t num_bits = std::min(size, PAGE_BYTES) >> BYTES_PER_BIT_SHIFT;
+        const size_t mask = ~u64{0} >> (64 - num_bits);
+        pages[page] |= (~u64{0} & mask) << first_bit;
+    }
+
+    bool IsPageUsed(u64 page, u64 offset, u64 size) const noexcept {
+        const size_t offset_in_page = offset % PAGE_BYTES;
+        const size_t first_bit = offset_in_page >> BYTES_PER_BIT_SHIFT;
+        const size_t num_bits = std::min(size, PAGE_BYTES) >> BYTES_PER_BIT_SHIFT;
+        const size_t mask = ~u64{0} >> (64 - num_bits);
+        const size_t mask2 = (~u64{0} & mask) << first_bit;
+        return (pages[page] & mask2) != 0;
+    }
+
+private:
+    std::vector<u64> pages;
+};
+
+} // namespace VideoCommon
--- a/src/video_core/renderer_opengl/gl_buffer_cache.cpp
+++ b/src/video_core/renderer_opengl/gl_buffer_cache.cpp
@@ -178,13 +178,14 @@ void BufferCacheRuntime::CopyBuffer(GLuint dst_buffer, Buffer& src_buffer,
 }

 void BufferCacheRuntime::CopyBuffer(Buffer& dst_buffer, GLuint src_buffer,
-                                    std::span<const VideoCommon::BufferCopy> copies, bool barrier) {
+                                    std::span<const VideoCommon::BufferCopy> copies, bool barrier,
+                                    bool) {
    CopyBuffer(dst_buffer.Handle(), src_buffer, copies, barrier);
 }

 void BufferCacheRuntime::CopyBuffer(Buffer& dst_buffer, Buffer& src_buffer,
-                                    std::span<const VideoCommon::BufferCopy> copies) {
-    CopyBuffer(dst_buffer.Handle(), src_buffer.Handle(), copies);
+                                    std::span<const VideoCommon::BufferCopy> copies, bool) {
+    CopyBuffer(dst_buffer.Handle(), src_buffer.Handle(), copies, true);
 }

 void BufferCacheRuntime::PreCopyBarrier() {
--- a/src/video_core/renderer_opengl/gl_buffer_cache.h
+++ b/src/video_core/renderer_opengl/gl_buffer_cache.h
@@ -30,6 +30,8 @@ public:

    void MakeResident(GLenum access) noexcept;

+    void MarkUsage(u64 offset, u64 size) {}
+
    [[nodiscard]] GLuint View(u32 offset, u32 size, VideoCore::Surface::PixelFormat format);

    [[nodiscard]] GLuint64EXT HostGpuAddr() const noexcept {
@@ -66,22 +68,29 @@ public:

    [[nodiscard]] StagingBufferMap DownloadStagingBuffer(size_t size);

+    bool CanReorderUpload(const Buffer&, std::span<const VideoCommon::BufferCopy>) {
+        return false;
+    }
+
    void CopyBuffer(GLuint dst_buffer, GLuint src_buffer,
-                    std::span<const VideoCommon::BufferCopy> copies, bool barrier = true);
+                    std::span<const VideoCommon::BufferCopy> copies, bool barrier);

    void CopyBuffer(GLuint dst_buffer, Buffer& src_buffer,
-                    std::span<const VideoCommon::BufferCopy> copies, bool barrier = true);
+                    std::span<const VideoCommon::BufferCopy> copies, bool barrier);

    void CopyBuffer(Buffer& dst_buffer, GLuint src_buffer,
-                    std::span<const VideoCommon::BufferCopy> copies, bool barrier = true);
+                    std::span<const VideoCommon::BufferCopy> copies, bool barrier,
+                    bool can_reorder_upload = false);

    void CopyBuffer(Buffer& dst_buffer, Buffer& src_buffer,
-                    std::span<const VideoCommon::BufferCopy> copies);
+                    std::span<const VideoCommon::BufferCopy> copies, bool);

    void PreCopyBarrier();
    void PostCopyBarrier();
    void Finish();

+    void TickFrame(VideoCommon::SlotVector<Buffer>&) noexcept {}
+
    void ClearBuffer(Buffer& dest_buffer, u32 offset, size_t size, u32 value);

    void BindIndexBuffer(Buffer& buffer, u32 offset, u32 size);
--- a/src/video_core/renderer_vulkan/vk_buffer_cache.cpp
+++ b/src/video_core/renderer_vulkan/vk_buffer_cache.cpp
@@ -79,13 +79,13 @@ vk::Buffer CreateBuffer(const Device& device, const MemoryAllocator& memory_allo
 } // Anonymous namespace

 Buffer::Buffer(BufferCacheRuntime&, VideoCommon::NullBufferParams null_params)
-    : VideoCommon::BufferBase<VideoCore::RasterizerInterface>(null_params) {}
+    : VideoCommon::BufferBase<VideoCore::RasterizerInterface>(null_params), tracker{4096} {}

 Buffer::Buffer(BufferCacheRuntime& runtime, VideoCore::RasterizerInterface& rasterizer_,
               VAddr cpu_addr_, u64 size_bytes_)
    : VideoCommon::BufferBase<VideoCore::RasterizerInterface>(rasterizer_, cpu_addr_, size_bytes_),
-      device{&runtime.device}, buffer{
-                                   CreateBuffer(*device, runtime.memory_allocator, SizeBytes())} {
+      device{&runtime.device}, buffer{CreateBuffer(*device, runtime.memory_allocator, SizeBytes())},
+      tracker{SizeBytes()} {
    if (runtime.device.HasDebuggingToolAttached()) {
        buffer.SetObjectNameEXT(fmt::format("Buffer 0x{:x}", CpuAddr()).c_str());
    }
@@ -355,12 +355,31 @@ bool BufferCacheRuntime::CanReportMemoryUsage() const {
    return device.CanReportMemoryUsage();
 }

+void BufferCacheRuntime::TickFrame(VideoCommon::SlotVector<Buffer>& slot_buffers) noexcept {
+    for (auto it = slot_buffers.begin(); it != slot_buffers.end(); it++) {
+        it->ResetUsageTracking();
+    }
+}
+
 void BufferCacheRuntime::Finish() {
    scheduler.Finish();
 }

+bool BufferCacheRuntime::CanReorderUpload(const Buffer& buffer,
+                                          std::span<const VideoCommon::BufferCopy> copies) {
+    if (Settings::values.disable_buffer_reorder) {
+        return false;
+    }
+    const bool can_use_upload_cmdbuf =
+        std::ranges::all_of(copies, [&](const VideoCommon::BufferCopy& copy) {
+            return !buffer.IsRegionUsed(copy.dst_offset, copy.size);
+        });
+    return can_use_upload_cmdbuf;
+}
+
 void BufferCacheRuntime::CopyBuffer(VkBuffer dst_buffer, VkBuffer src_buffer,
-                                    std::span<const VideoCommon::BufferCopy> copies, bool barrier) {
+                                    std::span<const VideoCommon::BufferCopy> copies, bool barrier,
+                                    bool can_reorder_upload) {
    if (dst_buffer == VK_NULL_HANDLE || src_buffer == VK_NULL_HANDLE) {
        return;
    }
@@ -376,9 +395,18 @@ void BufferCacheRuntime::CopyBuffer(VkBuffer dst_buffer, VkBuffer src_buffer,
        .srcAccessMask = VK_ACCESS_TRANSFER_WRITE_BIT,
        .dstAccessMask = VK_ACCESS_MEMORY_READ_BIT | VK_ACCESS_MEMORY_WRITE_BIT,
    };
+
    // Measuring a popular game, this number never exceeds the specified size once data is warmed up
    boost::container::small_vector<VkBufferCopy, 8> vk_copies(copies.size());
    std::ranges::transform(copies, vk_copies.begin(), MakeBufferCopy);
+    if (src_buffer == staging_pool.StreamBuf() && can_reorder_upload) {
+        scheduler.RecordWithUploadBuffer([src_buffer, dst_buffer, vk_copies](
+                                             vk::CommandBuffer, vk::CommandBuffer upload_cmdbuf) {
+            upload_cmdbuf.CopyBuffer(src_buffer, dst_buffer, vk_copies);
+        });
+        return;
+    }
+
    scheduler.RequestOutsideRenderPassOperationContext();
    scheduler.Record([src_buffer, dst_buffer, vk_copies, barrier](vk::CommandBuffer cmdbuf) {
        if (barrier) {
--- a/src/video_core/renderer_vulkan/vk_buffer_cache.h
+++ b/src/video_core/renderer_vulkan/vk_buffer_cache.h
@@ -5,6 +5,7 @@

 #include "video_core/buffer_cache/buffer_cache_base.h"
 #include "video_core/buffer_cache/memory_tracker_base.h"
+#include "video_core/buffer_cache/usage_tracker.h"
 #include "video_core/engines/maxwell_3d.h"
 #include "video_core/renderer_vulkan/vk_compute_pass.h"
 #include "video_core/renderer_vulkan/vk_staging_buffer_pool.h"
@@ -34,6 +35,18 @@ public:
        return *buffer;
    }

+    [[nodiscard]] bool IsRegionUsed(u64 offset, u64 size) const noexcept {
+        return tracker.IsUsed(offset, size);
+    }
+
+    void MarkUsage(u64 offset, u64 size) noexcept {
+        tracker.Track(offset, size);
+    }
+
+    void ResetUsageTracking() noexcept {
+        tracker.Reset();
+    }
+
    operator VkBuffer() const noexcept {
        return *buffer;
    }
@@ -49,6 +62,7 @@ private:
    const Device* device{};
    vk::Buffer buffer;
    std::vector<BufferView> views;
+    VideoCommon::UsageTracker tracker;
 };

 class QuadArrayIndexBuffer;
@@ -67,6 +81,8 @@ public:
                                ComputePassDescriptorQueue& compute_pass_descriptor_queue,
                                DescriptorPool& descriptor_pool);

+    void TickFrame(VideoCommon::SlotVector<Buffer>& slot_buffers) noexcept;
+
    void Finish();

    u64 GetDeviceLocalMemory() const;
@@ -79,12 +95,15 @@ public:

    [[nodiscard]] StagingBufferRef DownloadStagingBuffer(size_t size, bool deferred = false);

+    bool CanReorderUpload(const Buffer& buffer, std::span<const VideoCommon::BufferCopy> copies);
+
    void FreeDeferredStagingBuffer(StagingBufferRef& ref);

    void PreCopyBarrier();

    void CopyBuffer(VkBuffer src_buffer, VkBuffer dst_buffer,
-                    std::span<const VideoCommon::BufferCopy> copies, bool barrier = true);
+                    std::span<const VideoCommon::BufferCopy> copies, bool barrier,
+                    bool can_reorder_upload = false);

    void PostCopyBarrier();

--- a/src/video_core/renderer_vulkan/vk_master_semaphore.cpp
+++ b/src/video_core/renderer_vulkan/vk_master_semaphore.cpp
@@ -100,12 +100,14 @@ void MasterSemaphore::Wait(u64 tick) {
    Refresh();
 }

-VkResult MasterSemaphore::SubmitQueue(vk::CommandBuffer& cmdbuf, VkSemaphore signal_semaphore,
-                                      VkSemaphore wait_semaphore, u64 host_tick) {
+VkResult MasterSemaphore::SubmitQueue(vk::CommandBuffer& cmdbuf, vk::CommandBuffer& upload_cmdbuf,
+                                      VkSemaphore signal_semaphore, VkSemaphore wait_semaphore,
+                                      u64 host_tick) {
    if (semaphore) {
-        return SubmitQueueTimeline(cmdbuf, signal_semaphore, wait_semaphore, host_tick);
+        return SubmitQueueTimeline(cmdbuf, upload_cmdbuf, signal_semaphore, wait_semaphore,
+                                   host_tick);
    } else {
-        return SubmitQueueFence(cmdbuf, signal_semaphore, wait_semaphore, host_tick);
+        return SubmitQueueFence(cmdbuf, upload_cmdbuf, signal_semaphore, wait_semaphore, host_tick);
    }
 }

@@ -115,6 +117,7 @@ static constexpr std::array<VkPipelineStageFlags, 2> wait_stage_masks{
 };

 VkResult MasterSemaphore::SubmitQueueTimeline(vk::CommandBuffer& cmdbuf,
+                                              vk::CommandBuffer& upload_cmdbuf,
                                              VkSemaphore signal_semaphore,
                                              VkSemaphore wait_semaphore, u64 host_tick) {
    const VkSemaphore timeline_semaphore = *semaphore;
@@ -123,6 +126,8 @@ VkResult MasterSemaphore::SubmitQueueTimeline(vk::CommandBuffer& cmdbuf,
    const std::array signal_values{host_tick, u64(0)};
    const std::array signal_semaphores{timeline_semaphore, signal_semaphore};

+    const std::array cmdbuffers{*upload_cmdbuf, *cmdbuf};
+
    const u32 num_wait_semaphores = wait_semaphore ? 1 : 0;
    const VkTimelineSemaphoreSubmitInfo timeline_si{
        .sType = VK_STRUCTURE_TYPE_TIMELINE_SEMAPHORE_SUBMIT_INFO,
@@ -138,8 +143,8 @@ VkResult MasterSemaphore::SubmitQueueTimeline(vk::CommandBuffer& cmdbuf,
        .waitSemaphoreCount = num_wait_semaphores,
        .pWaitSemaphores = &wait_semaphore,
        .pWaitDstStageMask = wait_stage_masks.data(),
-        .commandBufferCount = 1,
-        .pCommandBuffers = cmdbuf.address(),
+        .commandBufferCount = static_cast<u32>(cmdbuffers.size()),
+        .pCommandBuffers = cmdbuffers.data(),
        .signalSemaphoreCount = num_signal_semaphores,
        .pSignalSemaphores = signal_semaphores.data(),
    };
@@ -147,19 +152,23 @@ VkResult MasterSemaphore::SubmitQueueTimeline(vk::CommandBuffer& cmdbuf,
    return device.GetGraphicsQueue().Submit(submit_info);
 }

-VkResult MasterSemaphore::SubmitQueueFence(vk::CommandBuffer& cmdbuf, VkSemaphore signal_semaphore,
-                                           VkSemaphore wait_semaphore, u64 host_tick) {
+VkResult MasterSemaphore::SubmitQueueFence(vk::CommandBuffer& cmdbuf,
+                                           vk::CommandBuffer& upload_cmdbuf,
+                                           VkSemaphore signal_semaphore, VkSemaphore wait_semaphore,
+                                           u64 host_tick) {
    const u32 num_signal_semaphores = signal_semaphore ? 1 : 0;
    const u32 num_wait_semaphores = wait_semaphore ? 1 : 0;

+    const std::array cmdbuffers{*upload_cmdbuf, *cmdbuf};
+
    const VkSubmitInfo submit_info{
        .sType = VK_STRUCTURE_TYPE_SUBMIT_INFO,
        .pNext = nullptr,
        .waitSemaphoreCount = num_wait_semaphores,
        .pWaitSemaphores = &wait_semaphore,
        .pWaitDstStageMask = wait_stage_masks.data(),
-        .commandBufferCount = 1,
-        .pCommandBuffers = cmdbuf.address(),
+        .commandBufferCount = static_cast<u32>(cmdbuffers.size()),
+        .pCommandBuffers = cmdbuffers.data(),
        .signalSemaphoreCount = num_signal_semaphores,
        .pSignalSemaphores = &signal_semaphore,
    };
--- a/src/video_core/renderer_vulkan/vk_master_semaphore.h
+++ b/src/video_core/renderer_vulkan/vk_master_semaphore.h
@@ -52,14 +52,16 @@ public:
    void Wait(u64 tick);

    /// Submits the device graphics queue, updating the tick as necessary
-    VkResult SubmitQueue(vk::CommandBuffer& cmdbuf, VkSemaphore signal_semaphore,
-                         VkSemaphore wait_semaphore, u64 host_tick);
+    VkResult SubmitQueue(vk::CommandBuffer& cmdbuf, vk::CommandBuffer& upload_cmdbuf,
+                         VkSemaphore signal_semaphore, VkSemaphore wait_semaphore, u64 host_tick);

 private:
-    VkResult SubmitQueueTimeline(vk::CommandBuffer& cmdbuf, VkSemaphore signal_semaphore,
-                                 VkSemaphore wait_semaphore, u64 host_tick);
-    VkResult SubmitQueueFence(vk::CommandBuffer& cmdbuf, VkSemaphore signal_semaphore,
-                              VkSemaphore wait_semaphore, u64 host_tick);
+    VkResult SubmitQueueTimeline(vk::CommandBuffer& cmdbuf, vk::CommandBuffer& upload_cmdbuf,
+                                 VkSemaphore signal_semaphore, VkSemaphore wait_semaphore,
+                                 u64 host_tick);
+    VkResult SubmitQueueFence(vk::CommandBuffer& cmdbuf, vk::CommandBuffer& upload_cmdbuf,
+                              VkSemaphore signal_semaphore, VkSemaphore wait_semaphore,
+                              u64 host_tick);

    void WaitThread(std::stop_token token);

--- a/src/video_core/renderer_vulkan/vk_scheduler.cpp
+++ b/src/video_core/renderer_vulkan/vk_scheduler.cpp
@@ -22,11 +22,12 @@ namespace Vulkan {

 MICROPROFILE_DECLARE(Vulkan_WaitForWorker);

-void Scheduler::CommandChunk::ExecuteAll(vk::CommandBuffer cmdbuf) {
+void Scheduler::CommandChunk::ExecuteAll(vk::CommandBuffer cmdbuf,
+                                         vk::CommandBuffer upload_cmdbuf) {
    auto command = first;
    while (command != nullptr) {
        auto next = command->GetNext();
-        command->Execute(cmdbuf);
+        command->Execute(cmdbuf, upload_cmdbuf);
        command->~Command();
        command = next;
    }
@@ -180,7 +181,7 @@ void Scheduler::WorkerThread(std::stop_token stop_token) {
            // Perform the work, tracking whether the chunk was a submission
            // before executing.
            const bool has_submit = work->HasSubmit();
-            work->ExecuteAll(current_cmdbuf);
+            work->ExecuteAll(current_cmdbuf, current_upload_cmdbuf);

            // If the chunk was a submission, reallocate the command buffer.
            if (has_submit) {
@@ -205,6 +206,13 @@ void Scheduler::AllocateWorkerCommandBuffer() {
        .flags = VK_COMMAND_BUFFER_USAGE_ONE_TIME_SUBMIT_BIT,
        .pInheritanceInfo = nullptr,
    });
+    current_upload_cmdbuf = vk::CommandBuffer(command_pool->Commit(), device.GetDispatchLoader());
+    current_upload_cmdbuf.Begin({
+        .sType = VK_STRUCTURE_TYPE_COMMAND_BUFFER_BEGIN_INFO,
+        .pNext = nullptr,
+        .flags = VK_COMMAND_BUFFER_USAGE_ONE_TIME_SUBMIT_BIT,
+        .pInheritanceInfo = nullptr,
+    });
 }

 u64 Scheduler::SubmitExecution(VkSemaphore signal_semaphore, VkSemaphore wait_semaphore) {
@@ -212,7 +220,17 @@ u64 Scheduler::SubmitExecution(VkSemaphore signal_semaphore, VkSemaphore wait_se
    InvalidateState();

    const u64 signal_value = master_semaphore->NextTick();
-    Record([signal_semaphore, wait_semaphore, signal_value, this](vk::CommandBuffer cmdbuf) {
+    RecordWithUploadBuffer([signal_semaphore, wait_semaphore, signal_value,
+                            this](vk::CommandBuffer cmdbuf, vk::CommandBuffer upload_cmdbuf) {
+        static constexpr VkMemoryBarrier WRITE_BARRIER{
+            .sType = VK_STRUCTURE_TYPE_MEMORY_BARRIER,
+            .pNext = nullptr,
+            .srcAccessMask = VK_ACCESS_TRANSFER_WRITE_BIT,
+            .dstAccessMask = VK_ACCESS_MEMORY_READ_BIT | VK_ACCESS_MEMORY_WRITE_BIT,
+        };
+        upload_cmdbuf.PipelineBarrier(VK_PIPELINE_STAGE_TRANSFER_BIT,
+                                      VK_PIPELINE_STAGE_ALL_COMMANDS_BIT, 0, WRITE_BARRIER);
+        upload_cmdbuf.End();
        cmdbuf.End();

        if (on_submit) {
@@ -221,7 +239,7 @@ u64 Scheduler::SubmitExecution(VkSemaphore signal_semaphore, VkSemaphore wait_se

        std::scoped_lock lock{submit_mutex};
        switch (const VkResult result = master_semaphore->SubmitQueue(
-                    cmdbuf, signal_semaphore, wait_semaphore, signal_value)) {
+                    cmdbuf, upload_cmdbuf, signal_semaphore, wait_semaphore, signal_value)) {
        case VK_SUCCESS:
            break;
        case VK_ERROR_DEVICE_LOST:
--- a/src/video_core/renderer_vulkan/vk_scheduler.h
+++ b/src/video_core/renderer_vulkan/vk_scheduler.h
@@ -80,7 +80,8 @@ public:

    /// Send work to a separate thread.
    template <typename T>
-    void Record(T&& command) {
+        requires std::is_invocable_v<T, vk::CommandBuffer, vk::CommandBuffer>
+    void RecordWithUploadBuffer(T&& command) {
        if (chunk->Record(command)) {
            return;
        }
@@ -88,6 +89,15 @@ public:
        (void)chunk->Record(command);
    }

+    template <typename T>
+        requires std::is_invocable_v<T, vk::CommandBuffer>
+    void Record(T&& c) {
+        this->RecordWithUploadBuffer(
+            [command = std::move(c)](vk::CommandBuffer cmdbuf, vk::CommandBuffer) {
+                command(cmdbuf);
+            });
+    }
+
    /// Returns the current command buffer tick.
    [[nodiscard]] u64 CurrentTick() const noexcept {
        return master_semaphore->CurrentTick();
@@ -119,7 +129,7 @@ private:
    public:
        virtual ~Command() = default;

-        virtual void Execute(vk::CommandBuffer cmdbuf) const = 0;
+        virtual void Execute(vk::CommandBuffer cmdbuf, vk::CommandBuffer upload_cmdbuf) const = 0;

        Command* GetNext() const {
            return next;
@@ -142,8 +152,8 @@ private:
        TypedCommand(TypedCommand&&) = delete;
        TypedCommand& operator=(TypedCommand&&) = delete;

-        void Execute(vk::CommandBuffer cmdbuf) const override {
-            command(cmdbuf);
+        void Execute(vk::CommandBuffer cmdbuf, vk::CommandBuffer upload_cmdbuf) const override {
+            command(cmdbuf, upload_cmdbuf);
        }

    private:
@@ -152,7 +162,7 @@ private:

    class CommandChunk final {
    public:
-        void ExecuteAll(vk::CommandBuffer cmdbuf);
+        void ExecuteAll(vk::CommandBuffer cmdbuf, vk::CommandBuffer upload_cmdbuf);

        template <typename T>
        bool Record(T& command) {
@@ -228,6 +238,7 @@ private:
    VideoCommon::QueryCacheBase<QueryCacheParams>* query_cache = nullptr;

    vk::CommandBuffer current_cmdbuf;
+    vk::CommandBuffer current_upload_cmdbuf;

    std::unique_ptr<CommandChunk> chunk;
    std::function<void()> on_submit;
--- a/src/video_core/renderer_vulkan/vk_smaa.cpp
+++ b/src/video_core/renderer_vulkan/vk_smaa.cpp
@@ -672,7 +672,7 @@ void SMAA::UploadImages(Scheduler& scheduler) {
    UploadImage(m_device, m_allocator, scheduler, m_static_images[Search], search_extent,
                VK_FORMAT_R8_UNORM, ARRAY_TO_SPAN(searchTexBytes));

-    scheduler.Record([&](vk::CommandBuffer& cmdbuf) {
+    scheduler.Record([&](vk::CommandBuffer cmdbuf) {
        for (auto& images : m_dynamic_images) {
            for (size_t i = 0; i < MaxDynamicImage; i++) {
                ClearColorImage(cmdbuf, *images.images[i]);
@@ -707,7 +707,7 @@ VkImageView SMAA::Draw(Scheduler& scheduler, size_t image_index, VkImage source_
    UpdateDescriptorSets(source_image_view, image_index);

    scheduler.RequestOutsideRenderPassOperationContext();
-    scheduler.Record([=, this](vk::CommandBuffer& cmdbuf) {
+    scheduler.Record([=, this](vk::CommandBuffer cmdbuf) {
        TransitionImageLayout(cmdbuf, source_image, VK_IMAGE_LAYOUT_GENERAL);
        TransitionImageLayout(cmdbuf, edges_image, VK_IMAGE_LAYOUT_GENERAL);
        BeginRenderPass(cmdbuf, m_renderpasses[EdgeDetection], edge_detection_framebuffer,
--- a/src/video_core/renderer_vulkan/vk_staging_buffer_pool.h
+++ b/src/video_core/renderer_vulkan/vk_staging_buffer_pool.h
@@ -36,6 +36,10 @@ public:
    StagingBufferRef Request(size_t size, MemoryUsage usage, bool deferred = false);
    void FreeDeferred(StagingBufferRef& ref);

+    [[nodiscard]] VkBuffer StreamBuf() const noexcept {
+        return *stream_buffer;
+    }
+
    void TickFrame();

 private:
--- a/src/video_core/texture_cache/slot_vector.h
+++ b/src/video_core/texture_cache/slot_vector.h
@@ -138,6 +138,10 @@ public:
        return Iterator(this, SlotId{SlotId::INVALID_INDEX});
    }

+    [[nodiscard]] size_t size() const noexcept {
+        return values_capacity - free_list.size();
+    }
+
 private:
    struct NonTrivialDummy {
        NonTrivialDummy() noexcept {}
--- a/src/video_core/vulkan_common/vulkan_wrapper.h
+++ b/src/video_core/vulkan_common/vulkan_wrapper.h
@@ -1101,6 +1101,10 @@ public:
        return &handle;
    }

+    VkCommandBuffer operator*() const noexcept {
+        return handle;
+    }
+
    void Begin(const VkCommandBufferBeginInfo& begin_info) const {
        Check(dld->vkBeginCommandBuffer(handle, &begin_info));
    }
--- a/src/yuzu/configuration/configure_debug.cpp
+++ b/src/yuzu/configuration/configure_debug.cpp
@@ -51,6 +51,8 @@ void ConfigureDebug::SetConfiguration() {
    ui->enable_all_controllers->setChecked(Settings::values.enable_all_controllers.GetValue());
    ui->enable_renderdoc_hotkey->setEnabled(runtime_lock);
    ui->enable_renderdoc_hotkey->setChecked(Settings::values.enable_renderdoc_hotkey.GetValue());
+    ui->disable_buffer_reorder->setEnabled(runtime_lock);
+    ui->disable_buffer_reorder->setChecked(Settings::values.disable_buffer_reorder.GetValue());
    ui->enable_graphics_debugging->setEnabled(runtime_lock);
    ui->enable_graphics_debugging->setChecked(Settings::values.renderer_debug.GetValue());
    ui->enable_shader_feedback->setEnabled(runtime_lock);
@@ -96,6 +98,7 @@ void ConfigureDebug::ApplyConfiguration() {
    Settings::values.enable_all_controllers = ui->enable_all_controllers->isChecked();
    Settings::values.renderer_debug = ui->enable_graphics_debugging->isChecked();
    Settings::values.enable_renderdoc_hotkey = ui->enable_renderdoc_hotkey->isChecked();
+    Settings::values.disable_buffer_reorder = ui->disable_buffer_reorder->isChecked();
    Settings::values.renderer_shader_feedback = ui->enable_shader_feedback->isChecked();
    Settings::values.cpu_debug_mode = ui->enable_cpu_debugging->isChecked();
    Settings::values.enable_nsight_aftermath = ui->enable_nsight_aftermath->isChecked();
--- a/src/yuzu/configuration/configure_debug.ui
+++ b/src/yuzu/configuration/configure_debug.ui
@@ -271,19 +271,6 @@
          </widget>
         </item>
         <item row="8" column="0">
-          <widget class="QCheckBox" name="disable_macro_hle">
-           <property name="enabled">
-            <bool>true</bool>
-           </property>
-           <property name="toolTip">
-            <string>When checked, it disables the macro HLE functions. Enabling this makes games run slower</string>
-           </property>
-           <property name="text">
-            <string>Disable Macro HLE</string>
-           </property>
-          </widget>
-         </item>
-         <item row="7" column="0">
          <widget class="QCheckBox" name="dump_macros">
           <property name="enabled">
            <bool>true</bool>
@@ -306,59 +293,7 @@
           </property>
          </widget>
         </item>
-         <item row="2" column="0">
-          <widget class="QCheckBox" name="enable_shader_feedback">
-           <property name="toolTip">
-            <string>When checked, yuzu will log statistics about the compiled pipeline cache</string>
-           </property>
-           <property name="text">
-            <string>Enable Shader Feedback</string>
-           </property>
-          </widget>
-         </item>
         <item row="6" column="0">
-          <widget class="QCheckBox" name="disable_macro_jit">
-           <property name="enabled">
-            <bool>true</bool>
-           </property>
-           <property name="toolTip">
-            <string>When checked, it disables the macro Just In Time compiler. Enabling this makes games run slower</string>
-           </property>
-           <property name="text">
-            <string>Disable Macro JIT</string>
-           </property>
-          </widget>
-         </item>
-         <item row="9" column="0">
-          <spacer name="verticalSpacer_5">
-           <property name="orientation">
-            <enum>Qt::Vertical</enum>
-           </property>
-           <property name="sizeType">
-            <enum>QSizePolicy::Preferred</enum>
-           </property>
-           <property name="sizeHint" stdset="0">
-            <size>
-             <width>20</width>
-             <height>0</height>
-            </size>
-           </property>
-          </spacer>
-         </item>
-         <item row="0" column="0">
-          <widget class="QCheckBox" name="enable_graphics_debugging">
-           <property name="enabled">
-            <bool>true</bool>
-           </property>
-           <property name="toolTip">
-            <string>When checked, the graphics API enters a slower debugging mode</string>
-           </property>
-           <property name="text">
-            <string>Enable Graphics Debugging</string>
-           </property>
-          </widget>
-         </item>
-         <item row="5" column="0">
          <widget class="QCheckBox" name="dump_shaders">
           <property name="enabled">
            <bool>true</bool>
@@ -378,6 +313,81 @@
           </property>
          </widget>
         </item>
+         <item row="7" column="0">
+          <widget class="QCheckBox" name="disable_macro_jit">
+           <property name="enabled">
+            <bool>true</bool>
+           </property>
+           <property name="toolTip">
+            <string>When checked, it disables the macro Just In Time compiler. Enabling this makes games run slower</string>
+           </property>
+           <property name="text">
+            <string>Disable Macro JIT</string>
+           </property>
+          </widget>
+         </item>
+         <item row="9" column="0">
+          <widget class="QCheckBox" name="disable_macro_hle">
+           <property name="enabled">
+            <bool>true</bool>
+           </property>
+           <property name="toolTip">
+            <string>When checked, it disables the macro HLE functions. Enabling this makes games run slower</string>
+           </property>
+           <property name="text">
+            <string>Disable Macro HLE</string>
+           </property>
+          </widget>
+         </item>
+         <item row="0" column="0">
+          <widget class="QCheckBox" name="enable_graphics_debugging">
+           <property name="enabled">
+            <bool>true</bool>
+           </property>
+           <property name="toolTip">
+            <string>When checked, the graphics API enters a slower debugging mode</string>
+           </property>
+           <property name="text">
+            <string>Enable Graphics Debugging</string>
+           </property>
+          </widget>
+         </item>
+         <item row="10" column="0">
+          <spacer name="verticalSpacer_5">
+           <property name="orientation">
+            <enum>Qt::Vertical</enum>
+           </property>
+           <property name="sizeType">
+            <enum>QSizePolicy::Preferred</enum>
+           </property>
+           <property name="sizeHint" stdset="0">
+            <size>
+             <width>20</width>
+             <height>0</height>
+            </size>
+           </property>
+          </spacer>
+         </item>
+         <item row="2" column="0">
+          <widget class="QCheckBox" name="enable_shader_feedback">
+           <property name="toolTip">
+            <string>When checked, yuzu will log statistics about the compiled pipeline cache</string>
+           </property>
+           <property name="text">
+            <string>Enable Shader Feedback</string>
+           </property>
+          </widget>
+         </item>
+         <item row="5" column="0">
+          <widget class="QCheckBox" name="disable_buffer_reorder">
+           <property name="toolTip">
+            <string>&lt;html&gt;&lt;head/&gt;&lt;body&gt;&lt;p&gt;When checked, disables reording of mapped memory uploads which allows to associate uploads with specific draws. May reduce performance in some cases.&lt;/p&gt;&lt;/body&gt;&lt;/html&gt;</string>
+           </property>
+           <property name="text">
+            <string>Disable Buffer Reorder</string>
+           </property>
+          </widget>
+         </item>
        </layout>
       </widget>
      </item>
--- a/src/yuzu/main.cpp
+++ b/src/yuzu/main.cpp
@@ -1909,7 +1909,8 @@ void GMainWindow::BootGame(const QString& filename, u64 program_id, std::size_t
                           StartGameType type, AmLaunchType launch_type) {
    LOG_INFO(Frontend, "yuzu starting...");

-    if (program_id > static_cast<u64>(Service::AM::Applets::AppletProgramId::MaxProgramId)) {
+    if (program_id == 0 ||
+        program_id > static_cast<u64>(Service::AM::Applets::AppletProgramId::MaxProgramId)) {
        StoreRecentFile(filename); // Put the filename on top of the list
    }

@@ -4295,7 +4296,7 @@ void GMainWindow::OnAlbum() {

    const auto filename = QString::fromStdString(album_nca->GetFullPath());
    UISettings::values.roms_path = QFileInfo(filename).path();
-    BootGame(filename);
+    BootGame(filename, AlbumId);
 }

 void GMainWindow::OnCabinet(Service::NFP::CabinetMode mode) {
@@ -4319,7 +4320,7 @@ void GMainWindow::OnCabinet(Service::NFP::CabinetMode mode) {

    const auto filename = QString::fromStdString(cabinet_nca->GetFullPath());
    UISettings::values.roms_path = QFileInfo(filename).path();
-    BootGame(filename);
+    BootGame(filename, CabinetId);
 }

 void GMainWindow::OnMiiEdit() {
@@ -4342,7 +4343,7 @@ void GMainWindow::OnMiiEdit() {

    const auto filename = QString::fromStdString((mii_applet_nca->GetFullPath()));
    UISettings::values.roms_path = QFileInfo(filename).path();
-    BootGame(filename);
+    BootGame(filename, MiiEditId);
 }

 void GMainWindow::OnCaptureScreenshot() {
Author	SHA1	Message	Date
yuzubot	d34e2c3044	Android #131	2023-11-13 00:57:12 +00:00
yuzubot	b2ed31e32b	Merge PR 11535	2023-11-13 00:57:11 +00:00
liamwhite	767c4b5a99	Merge pull request #11980 from german77/moment service: irs: Implement moment image processor	2023-11-11 20:02:23 -05:00
liamwhite	904d03b01f	Merge pull request #12005 from german77/homebrew_id yuzu: Keep homebrew on the recently played list	2023-11-11 20:02:09 -05:00
liamwhite	9f1c9599a2	Merge pull request #12004 from liamwhite/fix-hbl k_capabilities: ignore map region when KTrace is disabled	2023-11-11 20:02:01 -05:00
liamwhite	5f6666a7cd	Merge pull request #12003 from liamwhite/read-modules gdbstub: read module information from memory layout	2023-11-11 20:01:53 -05:00
liamwhite	1906e2724f	Merge pull request #11992 from t895/frame-check android: Hide loading animation on first frame	2023-11-11 20:01:32 -05:00
german77	0c032d3f2f	yuzu: Keep homebrew on the recently played list	2023-11-11 10:14:11 -06:00
Liam	a6735cba5f	k_capabilities: ignore map region when KTrace is disabled	2023-11-11 10:45:43 -05:00
Liam	3b872b89d1	gdbstub: read module information from memory layout	2023-11-11 10:41:06 -05:00
Narr the Reg	e588f341ed	service: irs: Implement moment image processor	2023-11-11 00:28:12 -06:00
t895	09f993899e	android: Hide loading animation on first frame	2023-11-09 22:27:40 -05:00