114 files changed, 1657 insertions, 615 deletions

diff --git a/src/common/CMakeLists.txt b/src/common/CMakeLists.txt
index 43ae8a9e7..5639021d3 100644
--- a/src/common/CMakeLists.txt
+++ b/src/common/CMakeLists.txt
@@ -91,6 +91,8 @@ add_library(common STATIC
     logging/log.h
     logging/text_formatter.cpp
     logging/text_formatter.h
+    lz4_compression.cpp
+    lz4_compression.h
     math_util.h
     memory_hook.cpp
     memory_hook.h
@@ -98,6 +100,7 @@ add_library(common STATIC
     microprofile.h
     microprofileui.h
     misc.cpp
+    multi_level_queue.h
     page_table.cpp
     page_table.h
     param_package.cpp
@@ -135,3 +138,4 @@ endif()
 create_target_directory_groups(common)
 
 target_link_libraries(common PUBLIC Boost::boost fmt microprofile)
+target_link_libraries(common PRIVATE lz4_static)
diff --git a/src/common/bit_util.h b/src/common/bit_util.h
index 1eea17ba1..a4f9ed4aa 100644
--- a/src/common/bit_util.h
+++ b/src/common/bit_util.h
@@ -58,4 +58,43 @@ inline u64 CountLeadingZeroes64(u64 value) {
     return __builtin_clzll(value);
 }
 #endif
+
+#ifdef _MSC_VER
+inline u32 CountTrailingZeroes32(u32 value) {
+    unsigned long trailing_zero = 0;
+
+    if (_BitScanForward(&trailing_zero, value) != 0) {
+        return trailing_zero;
+    }
+
+    return 32;
+}
+
+inline u64 CountTrailingZeroes64(u64 value) {
+    unsigned long trailing_zero = 0;
+
+    if (_BitScanForward64(&trailing_zero, value) != 0) {
+        return trailing_zero;
+    }
+
+    return 64;
+}
+#else
+inline u32 CountTrailingZeroes32(u32 value) {
+    if (value == 0) {
+        return 32;
+    }
+
+    return __builtin_ctz(value);
+}
+
+inline u64 CountTrailingZeroes64(u64 value) {
+    if (value == 0) {
+        return 64;
+    }
+
+    return __builtin_ctzll(value);
+}
+#endif
+
 } // namespace Common
diff --git a/src/common/detached_tasks.cpp b/src/common/detached_tasks.cpp
index a347d9e02..f268d6021 100644
--- a/src/common/detached_tasks.cpp
+++ b/src/common/detached_tasks.cpp
@@ -16,22 +16,22 @@ DetachedTasks::DetachedTasks() {
 }
 
 void DetachedTasks::WaitForAllTasks() {
-    std::unique_lock<std::mutex> lock(mutex);
+    std::unique_lock lock{mutex};
     cv.wait(lock, [this]() { return count == 0; });
 }
 
 DetachedTasks::~DetachedTasks() {
-    std::unique_lock<std::mutex> lock(mutex);
+    std::unique_lock lock{mutex};
     ASSERT(count == 0);
     instance = nullptr;
 }
 
 void DetachedTasks::AddTask(std::function<void()> task) {
-    std::unique_lock<std::mutex> lock(instance->mutex);
+    std::unique_lock lock{instance->mutex};
     ++instance->count;
     std::thread([task{std::move(task)}]() {
         task();
-        std::unique_lock<std::mutex> lock(instance->mutex);
+        std::unique_lock lock{instance->mutex};
         --instance->count;
         std::notify_all_at_thread_exit(instance->cv, std::move(lock));
     })
diff --git a/src/common/logging/backend.cpp b/src/common/logging/backend.cpp
index 4462ff3fb..a03179520 100644
--- a/src/common/logging/backend.cpp
+++ b/src/common/logging/backend.cpp
@@ -46,12 +46,12 @@ public:
     }
 
     void AddBackend(std::unique_ptr<Backend> backend) {
-        std::lock_guard<std::mutex> lock(writing_mutex);
+        std::lock_guard lock{writing_mutex};
         backends.push_back(std::move(backend));
     }
 
     void RemoveBackend(std::string_view backend_name) {
-        std::lock_guard<std::mutex> lock(writing_mutex);
+        std::lock_guard lock{writing_mutex};
         const auto it =
             std::remove_if(backends.begin(), backends.end(),
                            [&backend_name](const auto& i) { return backend_name == i->GetName(); });
@@ -80,7 +80,7 @@ private:
         backend_thread = std::thread([&] {
             Entry entry;
             auto write_logs = [&](Entry& e) {
-                std::lock_guard<std::mutex> lock(writing_mutex);
+                std::lock_guard lock{writing_mutex};
                 for (const auto& backend : backends) {
                     backend->Write(e);
                 }
diff --git a/src/common/lz4_compression.cpp b/src/common/lz4_compression.cpp
new file mode 100644
index 000000000..ade6759bb
--- /dev/null
+++ b/src/common/lz4_compression.cpp
@@ -0,0 +1,76 @@
+// Copyright 2019 yuzu Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#include <algorithm>
+#include <lz4hc.h>
+
+#include "common/assert.h"
+#include "common/lz4_compression.h"
+
+namespace Common::Compression {
+
+std::vector<u8> CompressDataLZ4(const u8* source, std::size_t source_size) {
+    ASSERT_MSG(source_size <= LZ4_MAX_INPUT_SIZE, "Source size exceeds LZ4 maximum input size");
+
+    const auto source_size_int = static_cast<int>(source_size);
+    const int max_compressed_size = LZ4_compressBound(source_size_int);
+    std::vector<u8> compressed(max_compressed_size);
+
+    const int compressed_size = LZ4_compress_default(reinterpret_cast<const char*>(source),
+                                                     reinterpret_cast<char*>(compressed.data()),
+                                                     source_size_int, max_compressed_size);
+
+    if (compressed_size <= 0) {
+        // Compression failed
+        return {};
+    }
+
+    compressed.resize(compressed_size);
+
+    return compressed;
+}
+
+std::vector<u8> CompressDataLZ4HC(const u8* source, std::size_t source_size,
+                                  s32 compression_level) {
+    ASSERT_MSG(source_size <= LZ4_MAX_INPUT_SIZE, "Source size exceeds LZ4 maximum input size");
+
+    compression_level = std::clamp(compression_level, LZ4HC_CLEVEL_MIN, LZ4HC_CLEVEL_MAX);
+
+    const auto source_size_int = static_cast<int>(source_size);
+    const int max_compressed_size = LZ4_compressBound(source_size_int);
+    std::vector<u8> compressed(max_compressed_size);
+
+    const int compressed_size = LZ4_compress_HC(
+        reinterpret_cast<const char*>(source), reinterpret_cast<char*>(compressed.data()),
+        source_size_int, max_compressed_size, compression_level);
+
+    if (compressed_size <= 0) {
+        // Compression failed
+        return {};
+    }
+
+    compressed.resize(compressed_size);
+
+    return compressed;
+}
+
+std::vector<u8> CompressDataLZ4HCMax(const u8* source, std::size_t source_size) {
+    return CompressDataLZ4HC(source, source_size, LZ4HC_CLEVEL_MAX);
+}
+
+std::vector<u8> DecompressDataLZ4(const std::vector<u8>& compressed,
+                                  std::size_t uncompressed_size) {
+    std::vector<u8> uncompressed(uncompressed_size);
+    const int size_check = LZ4_decompress_safe(reinterpret_cast<const char*>(compressed.data()),
+                                               reinterpret_cast<char*>(uncompressed.data()),
+                                               static_cast<int>(compressed.size()),
+                                               static_cast<int>(uncompressed.size()));
+    if (static_cast<int>(uncompressed_size) != size_check) {
+        // Decompression failed
+        return {};
+    }
+    return uncompressed;
+}
+
+} // namespace Common::Compression
diff --git a/src/common/lz4_compression.h b/src/common/lz4_compression.h
new file mode 100644
index 000000000..fe2231a6c
--- /dev/null
+++ b/src/common/lz4_compression.h
@@ -0,0 +1,55 @@
+// Copyright 2019 yuzu Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#include <vector>
+
+#include "common/common_types.h"
+
+namespace Common::Compression {
+
+/**
+ * Compresses a source memory region with LZ4 and returns the compressed data in a vector.
+ *
+ * @param source the uncompressed source memory region.
+ * @param source_size the size in bytes of the uncompressed source memory region.
+ *
+ * @return the compressed data.
+ */
+std::vector<u8> CompressDataLZ4(const u8* source, std::size_t source_size);
+
+/**
+ * Utilizes the LZ4 subalgorithm LZ4HC with the specified compression level. Higher compression
+ * levels result in a smaller compressed size, but require more CPU time for compression. The
+ * compression level has almost no impact on decompression speed. Data compressed with LZ4HC can
+ * also be decompressed with the default LZ4 decompression.
+ *
+ * @param source the uncompressed source memory region.
+ * @param source_size the size in bytes of the uncompressed source memory region.
+ * @param compression_level the used compression level. Should be between 3 and 12.
+ *
+ * @return the compressed data.
+ */
+std::vector<u8> CompressDataLZ4HC(const u8* source, std::size_t source_size, s32 compression_level);
+
+/**
+ * Utilizes the LZ4 subalgorithm LZ4HC with the highest possible compression level.
+ *
+ * @param source the uncompressed source memory region.
+ * @param source_size the size in bytes of the uncompressed source memory region.
+ *
+ * @return the compressed data.
+ */
+std::vector<u8> CompressDataLZ4HCMax(const u8* source, std::size_t source_size);
+
+/**
+ * Decompresses a source memory region with LZ4 and returns the uncompressed data in a vector.
+ *
+ * @param compressed the compressed source memory region.
+ * @param uncompressed_size the size in bytes of the uncompressed data.
+ *
+ * @return the decompressed data.
+ */
+std::vector<u8> DecompressDataLZ4(const std::vector<u8>& compressed, std::size_t uncompressed_size);
+
+} // namespace Common::Compression
\ No newline at end of file
diff --git a/src/common/multi_level_queue.h b/src/common/multi_level_queue.h
new file mode 100644
index 000000000..2b61b91e0
--- /dev/null
+++ b/src/common/multi_level_queue.h
@@ -0,0 +1,337 @@
+// Copyright 2019 TuxSH
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#pragma once
+
+#include <array>
+#include <iterator>
+#include <list>
+#include <utility>
+
+#include "common/bit_util.h"
+#include "common/common_types.h"
+
+namespace Common {
+
+/**
+ * A MultiLevelQueue is a type of priority queue which has the following characteristics:
+ * - iteratable through each of its elements.
+ * - back can be obtained.
+ * - O(1) add, lookup (both front and back)
+ * - discrete priorities and a max of 64 priorities (limited domain)
+ * This type of priority queue is normaly used for managing threads within an scheduler
+ */
+template <typename T, std::size_t Depth>
+class MultiLevelQueue {
+public:
+    using value_type = T;
+    using reference = value_type&;
+    using const_reference = const value_type&;
+    using pointer = value_type*;
+    using const_pointer = const value_type*;
+
+    using difference_type = typename std::pointer_traits<pointer>::difference_type;
+    using size_type = std::size_t;
+
+    template <bool is_constant>
+    class iterator_impl {
+    public:
+        using iterator_category = std::bidirectional_iterator_tag;
+        using value_type = T;
+        using pointer = std::conditional_t<is_constant, T*, const T*>;
+        using reference = std::conditional_t<is_constant, const T&, T&>;
+        using difference_type = typename std::pointer_traits<pointer>::difference_type;
+
+        friend bool operator==(const iterator_impl& lhs, const iterator_impl& rhs) {
+            if (lhs.IsEnd() && rhs.IsEnd())
+                return true;
+            return std::tie(lhs.current_priority, lhs.it) == std::tie(rhs.current_priority, rhs.it);
+        }
+
+        friend bool operator!=(const iterator_impl& lhs, const iterator_impl& rhs) {
+            return !operator==(lhs, rhs);
+        }
+
+        reference operator*() const {
+            return *it;
+        }
+
+        pointer operator->() const {
+            return it.operator->();
+        }
+
+        iterator_impl& operator++() {
+            if (IsEnd()) {
+                return *this;
+            }
+
+            ++it;
+
+            if (it == GetEndItForPrio()) {
+                u64 prios = mlq.used_priorities;
+                prios &= ~((1ULL << (current_priority + 1)) - 1);
+                if (prios == 0) {
+                    current_priority = mlq.depth();
+                } else {
+                    current_priority = CountTrailingZeroes64(prios);
+                    it = GetBeginItForPrio();
+                }
+            }
+            return *this;
+        }
+
+        iterator_impl& operator--() {
+            if (IsEnd()) {
+                if (mlq.used_priorities != 0) {
+                    current_priority = 63 - CountLeadingZeroes64(mlq.used_priorities);
+                    it = GetEndItForPrio();
+                    --it;
+                }
+            } else if (it == GetBeginItForPrio()) {
+                u64 prios = mlq.used_priorities;
+                prios &= (1ULL << current_priority) - 1;
+                if (prios != 0) {
+                    current_priority = CountTrailingZeroes64(prios);
+                    it = GetEndItForPrio();
+                    --it;
+                }
+            } else {
+                --it;
+            }
+            return *this;
+        }
+
+        iterator_impl operator++(int) {
+            const iterator_impl v{*this};
+            ++(*this);
+            return v;
+        }
+
+        iterator_impl operator--(int) {
+            const iterator_impl v{*this};
+            --(*this);
+            return v;
+        }
+
+        // allow implicit const->non-const
+        iterator_impl(const iterator_impl<false>& other)
+            : mlq(other.mlq), it(other.it), current_priority(other.current_priority) {}
+
+        iterator_impl(const iterator_impl<true>& other)
+            : mlq(other.mlq), it(other.it), current_priority(other.current_priority) {}
+
+        iterator_impl& operator=(const iterator_impl<false>& other) {
+            mlq = other.mlq;
+            it = other.it;
+            current_priority = other.current_priority;
+            return *this;
+        }
+
+        friend class iterator_impl<true>;
+        iterator_impl() = default;
+
+    private:
+        friend class MultiLevelQueue;
+        using container_ref =
+            std::conditional_t<is_constant, const MultiLevelQueue&, MultiLevelQueue&>;
+        using list_iterator = std::conditional_t<is_constant, typename std::list<T>::const_iterator,
+                                                 typename std::list<T>::iterator>;
+
+        explicit iterator_impl(container_ref mlq, list_iterator it, u32 current_priority)
+            : mlq(mlq), it(it), current_priority(current_priority) {}
+        explicit iterator_impl(container_ref mlq, u32 current_priority)
+            : mlq(mlq), it(), current_priority(current_priority) {}
+
+        bool IsEnd() const {
+            return current_priority == mlq.depth();
+        }
+
+        list_iterator GetBeginItForPrio() const {
+            return mlq.levels[current_priority].begin();
+        }
+
+        list_iterator GetEndItForPrio() const {
+            return mlq.levels[current_priority].end();
+        }
+
+        container_ref mlq;
+        list_iterator it;
+        u32 current_priority;
+    };
+
+    using iterator = iterator_impl<false>;
+    using const_iterator = iterator_impl<true>;
+
+    void add(const T& element, u32 priority, bool send_back = true) {
+        if (send_back)
+            levels[priority].push_back(element);
+        else
+            levels[priority].push_front(element);
+        used_priorities |= 1ULL << priority;
+    }
+
+    void remove(const T& element, u32 priority) {
+        auto it = ListIterateTo(levels[priority], element);
+        if (it == levels[priority].end())
+            return;
+        levels[priority].erase(it);
+        if (levels[priority].empty()) {
+            used_priorities &= ~(1ULL << priority);
+        }
+    }
+
+    void adjust(const T& element, u32 old_priority, u32 new_priority, bool adjust_front = false) {
+        remove(element, old_priority);
+        add(element, new_priority, !adjust_front);
+    }
+    void adjust(const_iterator it, u32 old_priority, u32 new_priority, bool adjust_front = false) {
+        adjust(*it, old_priority, new_priority, adjust_front);
+    }
+
+    void transfer_to_front(const T& element, u32 priority, MultiLevelQueue& other) {
+        ListSplice(other.levels[priority], other.levels[priority].begin(), levels[priority],
+                   ListIterateTo(levels[priority], element));
+
+        other.used_priorities |= 1ULL << priority;
+
+        if (levels[priority].empty()) {
+            used_priorities &= ~(1ULL << priority);
+        }
+    }
+
+    void transfer_to_front(const_iterator it, u32 priority, MultiLevelQueue& other) {
+        transfer_to_front(*it, priority, other);
+    }
+
+    void transfer_to_back(const T& element, u32 priority, MultiLevelQueue& other) {
+        ListSplice(other.levels[priority], other.levels[priority].end(), levels[priority],
+                   ListIterateTo(levels[priority], element));
+
+        other.used_priorities |= 1ULL << priority;
+
+        if (levels[priority].empty()) {
+            used_priorities &= ~(1ULL << priority);
+        }
+    }
+
+    void transfer_to_back(const_iterator it, u32 priority, MultiLevelQueue& other) {
+        transfer_to_back(*it, priority, other);
+    }
+
+    void yield(u32 priority, std::size_t n = 1) {
+        ListShiftForward(levels[priority], n);
+    }
+
+    std::size_t depth() const {
+        return Depth;
+    }
+
+    std::size_t size(u32 priority) const {
+        return levels[priority].size();
+    }
+
+    std::size_t size() const {
+        u64 priorities = used_priorities;
+        std::size_t size = 0;
+        while (priorities != 0) {
+            const u64 current_priority = CountTrailingZeroes64(priorities);
+            size += levels[current_priority].size();
+            priorities &= ~(1ULL << current_priority);
+        }
+        return size;
+    }
+
+    bool empty() const {
+        return used_priorities == 0;
+    }
+
+    bool empty(u32 priority) const {
+        return (used_priorities & (1ULL << priority)) == 0;
+    }
+
+    u32 highest_priority_set(u32 max_priority = 0) const {
+        const u64 priorities =
+            max_priority == 0 ? used_priorities : (used_priorities & ~((1ULL << max_priority) - 1));
+        return priorities == 0 ? Depth : static_cast<u32>(CountTrailingZeroes64(priorities));
+    }
+
+    u32 lowest_priority_set(u32 min_priority = Depth - 1) const {
+        const u64 priorities = min_priority >= Depth - 1
+                                   ? used_priorities
+                                   : (used_priorities & ((1ULL << (min_priority + 1)) - 1));
+        return priorities == 0 ? Depth : 63 - CountLeadingZeroes64(priorities);
+    }
+
+    const_iterator cbegin(u32 max_prio = 0) const {
+        const u32 priority = highest_priority_set(max_prio);
+        return priority == Depth ? cend()
+                                 : const_iterator{*this, levels[priority].cbegin(), priority};
+    }
+    const_iterator begin(u32 max_prio = 0) const {
+        return cbegin(max_prio);
+    }
+    iterator begin(u32 max_prio = 0) {
+        const u32 priority = highest_priority_set(max_prio);
+        return priority == Depth ? end() : iterator{*this, levels[priority].begin(), priority};
+    }
+
+    const_iterator cend(u32 min_prio = Depth - 1) const {
+        return min_prio == Depth - 1 ? const_iterator{*this, Depth} : cbegin(min_prio + 1);
+    }
+    const_iterator end(u32 min_prio = Depth - 1) const {
+        return cend(min_prio);
+    }
+    iterator end(u32 min_prio = Depth - 1) {
+        return min_prio == Depth - 1 ? iterator{*this, Depth} : begin(min_prio + 1);
+    }
+
+    T& front(u32 max_priority = 0) {
+        const u32 priority = highest_priority_set(max_priority);
+        return levels[priority == Depth ? 0 : priority].front();
+    }
+    const T& front(u32 max_priority = 0) const {
+        const u32 priority = highest_priority_set(max_priority);
+        return levels[priority == Depth ? 0 : priority].front();
+    }
+
+    T back(u32 min_priority = Depth - 1) {
+        const u32 priority = lowest_priority_set(min_priority); // intended
+        return levels[priority == Depth ? 63 : priority].back();
+    }
+    const T& back(u32 min_priority = Depth - 1) const {
+        const u32 priority = lowest_priority_set(min_priority); // intended
+        return levels[priority == Depth ? 63 : priority].back();
+    }
+
+private:
+    using const_list_iterator = typename std::list<T>::const_iterator;
+
+    static void ListShiftForward(std::list<T>& list, const std::size_t shift = 1) {
+        if (shift >= list.size()) {
+            return;
+        }
+
+        const auto begin_range = list.begin();
+        const auto end_range = std::next(begin_range, shift);
+        list.splice(list.end(), list, begin_range, end_range);
+    }
+
+    static void ListSplice(std::list<T>& in_list, const_list_iterator position,
+                           std::list<T>& out_list, const_list_iterator element) {
+        in_list.splice(position, out_list, element);
+    }
+
+    static const_list_iterator ListIterateTo(const std::list<T>& list, const T& element) {
+        auto it = list.cbegin();
+        while (it != list.cend() && *it != element) {
+            ++it;
+        }
+        return it;
+    }
+
+    std::array<std::list<T>, Depth> levels;
+    u64 used_priorities = 0;
+};
+
+} // namespace Common
diff --git a/src/common/thread.cpp b/src/common/thread.cpp
index 5144c0d9f..fe7a420cc 100644
--- a/src/common/thread.cpp
+++ b/src/common/thread.cpp
@@ -27,18 +27,6 @@ namespace Common {
 
 #ifdef _MSC_VER
 
-void SetThreadAffinity(std::thread::native_handle_type thread, u32 mask) {
-    SetThreadAffinityMask(thread, mask);
-}
-
-void SetCurrentThreadAffinity(u32 mask) {
-    SetThreadAffinityMask(GetCurrentThread(), mask);
-}
-
-void SwitchCurrentThread() {
-    SwitchToThread();
-}
-
 // Sets the debugger-visible name of the current thread.
 // Uses undocumented (actually, it is now documented) trick.
 // http://msdn.microsoft.com/library/default.asp?url=/library/en-us/vsdebug/html/vxtsksettingthreadname.asp
@@ -70,31 +58,6 @@ void SetCurrentThreadName(const char* name) {
 
 #else // !MSVC_VER, so must be POSIX threads
 
-void SetThreadAffinity(std::thread::native_handle_type thread, u32 mask) {
-#ifdef __APPLE__
-    thread_policy_set(pthread_mach_thread_np(thread), THREAD_AFFINITY_POLICY, (integer_t*)&mask, 1);
-#elif (defined __linux__ || defined __FreeBSD__) && !(defined ANDROID)
-    cpu_set_t cpu_set;
-    CPU_ZERO(&cpu_set);
-
-    for (int i = 0; i != sizeof(mask) * 8; ++i)
-        if ((mask >> i) & 1)
-            CPU_SET(i, &cpu_set);
-
-    pthread_setaffinity_np(thread, sizeof(cpu_set), &cpu_set);
-#endif
-}
-
-void SetCurrentThreadAffinity(u32 mask) {
-    SetThreadAffinity(pthread_self(), mask);
-}
-
-#ifndef _WIN32
-void SwitchCurrentThread() {
-    usleep(1000 * 1);
-}
-#endif
-
 // MinGW with the POSIX threading model does not support pthread_setname_np
 #if !defined(_WIN32) || defined(_MSC_VER)
 void SetCurrentThreadName(const char* name) {
diff --git a/src/common/thread.h b/src/common/thread.h
index 2cf74452d..0cfd98be6 100644
--- a/src/common/thread.h
+++ b/src/common/thread.h
@@ -9,14 +9,13 @@
 #include <cstddef>
 #include <mutex>
 #include <thread>
-#include "common/common_types.h"
 
 namespace Common {
 
 class Event {
 public:
     void Set() {
-        std::lock_guard<std::mutex> lk(mutex);
+        std::lock_guard lk{mutex};
         if (!is_set) {
             is_set = true;
             condvar.notify_one();
@@ -24,14 +23,14 @@ public:
     }
 
     void Wait() {
-        std::unique_lock<std::mutex> lk(mutex);
+        std::unique_lock lk{mutex};
         condvar.wait(lk, [&] { return is_set; });
         is_set = false;
     }
 
     template <class Clock, class Duration>
     bool WaitUntil(const std::chrono::time_point<Clock, Duration>& time) {
-        std::unique_lock<std::mutex> lk(mutex);
+        std::unique_lock lk{mutex};
         if (!condvar.wait_until(lk, time, [this] { return is_set; }))
             return false;
         is_set = false;
@@ -39,7 +38,7 @@ public:
     }
 
     void Reset() {
-        std::unique_lock<std::mutex> lk(mutex);
+        std::unique_lock lk{mutex};
         // no other action required, since wait loops on the predicate and any lingering signal will
         // get cleared on the first iteration
         is_set = false;
@@ -57,7 +56,7 @@ public:
 
     /// Blocks until all "count" threads have called Sync()
     void Sync() {
-        std::unique_lock<std::mutex> lk(mutex);
+        std::unique_lock lk{mutex};
         const std::size_t current_generation = generation;
 
         if (++waiting == count) {
@@ -78,9 +77,6 @@ private:
     std::size_t generation = 0; // Incremented once each time the barrier is used
 };
 
-void SetThreadAffinity(std::thread::native_handle_type thread, u32 mask);
-void SetCurrentThreadAffinity(u32 mask);
-void SwitchCurrentThread(); // On Linux, this is equal to sleep 1ms
 void SetCurrentThreadName(const char* name);
 
 } // namespace Common
diff --git a/src/common/threadsafe_queue.h b/src/common/threadsafe_queue.h
index 821e8536a..e714ba5b3 100644
--- a/src/common/threadsafe_queue.h
+++ b/src/common/threadsafe_queue.h
@@ -78,7 +78,7 @@ public:
 
     T PopWait() {
         if (Empty()) {
-            std::unique_lock<std::mutex> lock(cv_mutex);
+            std::unique_lock lock{cv_mutex};
             cv.wait(lock, [this]() { return !Empty(); });
         }
         T t;
@@ -137,7 +137,7 @@ public:
 
     template <typename Arg>
     void Push(Arg&& t) {
-        std::lock_guard<std::mutex> lock(write_lock);
+        std::lock_guard lock{write_lock};
         spsc_queue.Push(t);
     }
 
diff --git a/src/core/CMakeLists.txt b/src/core/CMakeLists.txt
index 9e23afe85..c59107102 100644
--- a/src/core/CMakeLists.txt
+++ b/src/core/CMakeLists.txt
@@ -458,7 +458,7 @@ add_library(core STATIC
 create_target_directory_groups(core)
 
 target_link_libraries(core PUBLIC common PRIVATE audio_core video_core)
-target_link_libraries(core PUBLIC Boost::boost PRIVATE fmt lz4_static mbedtls opus unicorn open_source_archives)
+target_link_libraries(core PUBLIC Boost::boost PRIVATE fmt mbedtls opus unicorn open_source_archives)
 if (ENABLE_WEB_SERVICE)
     target_compile_definitions(core PRIVATE -DENABLE_WEB_SERVICE)
     target_link_libraries(core PRIVATE web_service)
diff --git a/src/core/core_cpu.cpp b/src/core/core_cpu.cpp
index 1eefed6d0..e75741db0 100644
--- a/src/core/core_cpu.cpp
+++ b/src/core/core_cpu.cpp
@@ -22,7 +22,7 @@
 namespace Core {
 
 void CpuBarrier::NotifyEnd() {
-    std::unique_lock<std::mutex> lock(mutex);
+    std::unique_lock lock{mutex};
     end = true;
     condition.notify_all();
 }
@@ -34,7 +34,7 @@ bool CpuBarrier::Rendezvous() {
     }
 
     if (!end) {
-        std::unique_lock<std::mutex> lock(mutex);
+        std::unique_lock lock{mutex};
 
         --cores_waiting;
         if (!cores_waiting) {
@@ -131,7 +131,7 @@ void Cpu::Reschedule() {
 
     reschedule_pending = false;
     // Lock the global kernel mutex when we manipulate the HLE state
-    std::lock_guard<std::recursive_mutex> lock(HLE::g_hle_lock);
+    std::lock_guard lock{HLE::g_hle_lock};
     scheduler->Reschedule();
 }
 
diff --git a/src/core/frontend/emu_window.cpp b/src/core/frontend/emu_window.cpp
index e29afd630..1320bbe77 100644
--- a/src/core/frontend/emu_window.cpp
+++ b/src/core/frontend/emu_window.cpp
@@ -30,7 +30,7 @@ private:
         explicit Device(std::weak_ptr<TouchState>&& touch_state) : touch_state(touch_state) {}
         std::tuple<float, float, bool> GetStatus() const override {
             if (auto state = touch_state.lock()) {
-                std::lock_guard<std::mutex> guard(state->mutex);
+                std::lock_guard guard{state->mutex};
                 return std::make_tuple(state->touch_x, state->touch_y, state->touch_pressed);
             }
             return std::make_tuple(0.0f, 0.0f, false);
@@ -81,7 +81,7 @@ void EmuWindow::TouchPressed(unsigned framebuffer_x, unsigned framebuffer_y) {
     if (!IsWithinTouchscreen(framebuffer_layout, framebuffer_x, framebuffer_y))
         return;
 
-    std::lock_guard<std::mutex> guard(touch_state->mutex);
+    std::lock_guard guard{touch_state->mutex};
     touch_state->touch_x = static_cast<float>(framebuffer_x - framebuffer_layout.screen.left) /
                            (framebuffer_layout.screen.right - framebuffer_layout.screen.left);
     touch_state->touch_y = static_cast<float>(framebuffer_y - framebuffer_layout.screen.top) /
@@ -91,7 +91,7 @@ void EmuWindow::TouchPressed(unsigned framebuffer_x, unsigned framebuffer_y) {
 }
 
 void EmuWindow::TouchReleased() {
-    std::lock_guard<std::mutex> guard(touch_state->mutex);
+    std::lock_guard guard{touch_state->mutex};
     touch_state->touch_pressed = false;
     touch_state->touch_x = 0;
     touch_state->touch_y = 0;
diff --git a/src/core/hle/kernel/address_arbiter.cpp b/src/core/hle/kernel/address_arbiter.cpp
index 352190da8..c8842410b 100644
--- a/src/core/hle/kernel/address_arbiter.cpp
+++ b/src/core/hle/kernel/address_arbiter.cpp
@@ -26,7 +26,7 @@ void WakeThreads(const std::vector<SharedPtr<Thread>>& waiting_threads, s32 num_
     // them all.
     std::size_t last = waiting_threads.size();
     if (num_to_wake > 0) {
-        last = num_to_wake;
+        last = std::min(last, static_cast<std::size_t>(num_to_wake));
     }
 
     // Signal the waiting threads.
@@ -90,9 +90,9 @@ ResultCode AddressArbiter::ModifyByWaitingCountAndSignalToAddressIfEqual(VAddr a
     // Determine the modified value depending on the waiting count.
     s32 updated_value;
     if (waiting_threads.empty()) {
-        updated_value = value - 1;
-    } else if (num_to_wake <= 0 || waiting_threads.size() <= static_cast<u32>(num_to_wake)) {
         updated_value = value + 1;
+    } else if (num_to_wake <= 0 || waiting_threads.size() <= static_cast<u32>(num_to_wake)) {
+        updated_value = value - 1;
     } else {
         updated_value = value;
     }
diff --git a/src/core/hle/kernel/code_set.h b/src/core/hle/kernel/code_set.h
index 834fd23d2..879957dcb 100644
--- a/src/core/hle/kernel/code_set.h
+++ b/src/core/hle/kernel/code_set.h
@@ -5,7 +5,6 @@
 #pragma once
 
 #include <cstddef>
-#include <memory>
 #include <vector>
 
 #include "common/common_types.h"
@@ -78,7 +77,7 @@ struct CodeSet final {
     }
 
     /// The overall data that backs this code set.
-    std::shared_ptr<std::vector<u8>> memory;
+    std::vector<u8> memory;
 
     /// The segments that comprise this code set.
     std::array<Segment, 3> segments;
diff --git a/src/core/hle/kernel/kernel.cpp b/src/core/hle/kernel/kernel.cpp
index a7e4ddc05..3f14bfa86 100644
--- a/src/core/hle/kernel/kernel.cpp
+++ b/src/core/hle/kernel/kernel.cpp
@@ -34,7 +34,7 @@ static void ThreadWakeupCallback(u64 thread_handle, [[maybe_unused]] s64 cycles_
     const auto& system = Core::System::GetInstance();
 
     // Lock the global kernel mutex when we enter the kernel HLE.
-    std::lock_guard<std::recursive_mutex> lock(HLE::g_hle_lock);
+    std::lock_guard lock{HLE::g_hle_lock};
 
     SharedPtr<Thread> thread =
         system.Kernel().RetrieveThreadFromWakeupCallbackHandleTable(proper_handle);
@@ -62,7 +62,8 @@ static void ThreadWakeupCallback(u64 thread_handle, [[maybe_unused]] s64 cycles_
 
     if (thread->GetMutexWaitAddress() != 0 || thread->GetCondVarWaitAddress() != 0 ||
         thread->GetWaitHandle() != 0) {
-        ASSERT(thread->GetStatus() == ThreadStatus::WaitMutex);
+        ASSERT(thread->GetStatus() == ThreadStatus::WaitMutex ||
+               thread->GetStatus() == ThreadStatus::WaitCondVar);
         thread->SetMutexWaitAddress(0);
         thread->SetCondVarWaitAddress(0);
         thread->SetWaitHandle(0);
@@ -114,7 +115,7 @@ struct KernelCore::Impl {
 
     // Creates the default system resource limit
     void InitializeSystemResourceLimit(KernelCore& kernel) {
-        system_resource_limit = ResourceLimit::Create(kernel, "System");
+        system_resource_limit = ResourceLimit::Create(kernel);
 
         // If setting the default system values fails, then something seriously wrong has occurred.
         ASSERT(system_resource_limit->SetLimitValue(ResourceType::PhysicalMemory, 0x200000000)
@@ -190,6 +191,10 @@ const Process* KernelCore::CurrentProcess() const {
     return impl->current_process;
 }
 
+const std::vector<SharedPtr<Process>>& KernelCore::GetProcessList() const {
+    return impl->process_list;
+}
+
 void KernelCore::AddNamedPort(std::string name, SharedPtr<ClientPort> port) {
     impl->named_ports.emplace(std::move(name), std::move(port));
 }
diff --git a/src/core/hle/kernel/kernel.h b/src/core/hle/kernel/kernel.h
index 03ea5b659..6b8738599 100644
--- a/src/core/hle/kernel/kernel.h
+++ b/src/core/hle/kernel/kernel.h
@@ -72,6 +72,9 @@ public:
     /// Retrieves a const pointer to the current process.
     const Process* CurrentProcess() const;
 
+    /// Retrieves the list of processes.
+    const std::vector<SharedPtr<Process>>& GetProcessList() const;
+
     /// Adds a port to the named port table
     void AddNamedPort(std::string name, SharedPtr<ClientPort> port);
 
diff --git a/src/core/hle/kernel/object.cpp b/src/core/hle/kernel/object.cpp
index 217144efc..10431e94c 100644
--- a/src/core/hle/kernel/object.cpp
+++ b/src/core/hle/kernel/object.cpp
@@ -24,7 +24,6 @@ bool Object::IsWaitable() const {
     case HandleType::WritableEvent:
     case HandleType::SharedMemory:
     case HandleType::TransferMemory:
-    case HandleType::AddressArbiter:
     case HandleType::ResourceLimit:
     case HandleType::ClientPort:
     case HandleType::ClientSession:
diff --git a/src/core/hle/kernel/object.h b/src/core/hle/kernel/object.h
index 3f6baa094..332876c27 100644
--- a/src/core/hle/kernel/object.h
+++ b/src/core/hle/kernel/object.h
@@ -25,7 +25,6 @@ enum class HandleType : u32 {
     TransferMemory,
     Thread,
     Process,
-    AddressArbiter,
     ResourceLimit,
     ClientPort,
     ServerPort,
diff --git a/src/core/hle/kernel/process.cpp b/src/core/hle/kernel/process.cpp
index 0d782e4ba..041267318 100644
--- a/src/core/hle/kernel/process.cpp
+++ b/src/core/hle/kernel/process.cpp
@@ -5,6 +5,7 @@
 #include <algorithm>
 #include <memory>
 #include <random>
+#include "common/alignment.h"
 #include "common/assert.h"
 #include "common/logging/log.h"
 #include "core/core.h"
@@ -75,6 +76,18 @@ SharedPtr<ResourceLimit> Process::GetResourceLimit() const {
     return resource_limit;
 }
 
+u64 Process::GetTotalPhysicalMemoryUsed() const {
+    return vm_manager.GetCurrentHeapSize() + main_thread_stack_size + code_memory_size;
+}
+
+void Process::RegisterThread(const Thread* thread) {
+    thread_list.push_back(thread);
+}
+
+void Process::UnregisterThread(const Thread* thread) {
+    thread_list.remove(thread);
+}
+
 ResultCode Process::ClearSignalState() {
     if (status == ProcessStatus::Exited) {
         LOG_ERROR(Kernel, "called on a terminated process instance.");
@@ -107,14 +120,17 @@ ResultCode Process::LoadFromMetadata(const FileSys::ProgramMetadata& metadata) {
     return handle_table.SetSize(capabilities.GetHandleTableSize());
 }
 
-void Process::Run(VAddr entry_point, s32 main_thread_priority, u32 stack_size) {
+void Process::Run(VAddr entry_point, s32 main_thread_priority, u64 stack_size) {
+    // The kernel always ensures that the given stack size is page aligned.
+    main_thread_stack_size = Common::AlignUp(stack_size, Memory::PAGE_SIZE);
+
     // Allocate and map the main thread stack
     // TODO(bunnei): This is heap area that should be allocated by the kernel and not mapped as part
     // of the user address space.
+    const VAddr mapping_address = vm_manager.GetTLSIORegionEndAddress() - main_thread_stack_size;
     vm_manager
-        .MapMemoryBlock(vm_manager.GetTLSIORegionEndAddress() - stack_size,
-                        std::make_shared<std::vector<u8>>(stack_size, 0), 0, stack_size,
-                        MemoryState::Stack)
+        .MapMemoryBlock(mapping_address, std::make_shared<std::vector<u8>>(main_thread_stack_size),
+                        0, main_thread_stack_size, MemoryState::Stack)
         .Unwrap();
 
     vm_manager.LogLayout();
@@ -210,11 +226,13 @@ void Process::FreeTLSSlot(VAddr tls_address) {
 }
 
 void Process::LoadModule(CodeSet module_, VAddr base_addr) {
+    const auto memory = std::make_shared<std::vector<u8>>(std::move(module_.memory));
+
     const auto MapSegment = [&](const CodeSet::Segment& segment, VMAPermission permissions,
                                 MemoryState memory_state) {
         const auto vma = vm_manager
-                             .MapMemoryBlock(segment.addr + base_addr, module_.memory,
-                                             segment.offset, segment.size, memory_state)
+                             .MapMemoryBlock(segment.addr + base_addr, memory, segment.offset,
+                                             segment.size, memory_state)
                              .Unwrap();
         vm_manager.Reprotect(vma, permissions);
     };
@@ -224,6 +242,8 @@ void Process::LoadModule(CodeSet module_, VAddr base_addr) {
     MapSegment(module_.RODataSegment(), VMAPermission::Read, MemoryState::CodeData);
     MapSegment(module_.DataSegment(), VMAPermission::ReadWrite, MemoryState::CodeData);
 
+    code_memory_size += module_.memory.size();
+
     // Clear instruction cache in CPU JIT
     system.InvalidateCpuInstructionCaches();
 }
@@ -237,7 +257,7 @@ void Process::Acquire(Thread* thread) {
     ASSERT_MSG(!ShouldWait(thread), "Object unavailable!");
 }
 
-bool Process::ShouldWait(Thread* thread) const {
+bool Process::ShouldWait(const Thread* thread) const {
     return !is_signaled;
 }
 
diff --git a/src/core/hle/kernel/process.h b/src/core/hle/kernel/process.h
index 1bd7bf5c1..f060f2a3b 100644
--- a/src/core/hle/kernel/process.h
+++ b/src/core/hle/kernel/process.h
@@ -7,6 +7,7 @@
 #include <array>
 #include <bitset>
 #include <cstddef>
+#include <list>
 #include <string>
 #include <vector>
 #include <boost/container/static_vector.hpp>
@@ -35,14 +36,6 @@ class Thread;
 
 struct CodeSet;
 
-struct AddressMapping {
-    // Address and size must be page-aligned
-    VAddr address;
-    u64 size;
-    bool read_only;
-    bool unk_flag;
-};
-
 enum class MemoryRegion : u16 {
     APPLICATION = 1,
     SYSTEM = 2,
@@ -194,6 +187,22 @@ public:
         return random_entropy.at(index);
     }
 
+    /// Retrieves the total physical memory used by this process in bytes.
+    u64 GetTotalPhysicalMemoryUsed() const;
+
+    /// Gets the list of all threads created with this process as their owner.
+    const std::list<const Thread*>& GetThreadList() const {
+        return thread_list;
+    }
+
+    /// Registers a thread as being created under this process,
+    /// adding it to this process' thread list.
+    void RegisterThread(const Thread* thread);
+
+    /// Unregisters a thread from this process, removing it
+    /// from this process' thread list.
+    void UnregisterThread(const Thread* thread);
+
     /// Clears the signaled state of the process if and only if it's signaled.
     ///
     /// @pre The process must not be already terminated. If this is called on a
@@ -218,7 +227,7 @@ public:
     /**
      * Applies address space changes and launches the process main thread.
      */
-    void Run(VAddr entry_point, s32 main_thread_priority, u32 stack_size);
+    void Run(VAddr entry_point, s32 main_thread_priority, u64 stack_size);
 
     /**
      * Prepares a process for termination by stopping all of its threads
@@ -242,7 +251,7 @@ private:
     ~Process() override;
 
     /// Checks if the specified thread should wait until this process is available.
-    bool ShouldWait(Thread* thread) const override;
+    bool ShouldWait(const Thread* thread) const override;
 
     /// Acquires/locks this process for the specified thread if it's available.
     void Acquire(Thread* thread) override;
@@ -255,6 +264,12 @@ private:
     /// Memory manager for this process.
     Kernel::VMManager vm_manager;
 
+    /// Size of the main thread's stack in bytes.
+    u64 main_thread_stack_size = 0;
+
+    /// Size of the loaded code memory in bytes.
+    u64 code_memory_size = 0;
+
     /// Current status of the process
     ProcessStatus status;
 
@@ -307,6 +322,9 @@ private:
     /// Random values for svcGetInfo RandomEntropy
     std::array<u64, RANDOM_ENTROPY_SIZE> random_entropy;
 
+    /// List of threads that are running with this process as their owner.
+    std::list<const Thread*> thread_list;
+
     /// System context
     Core::System& system;
 
diff --git a/src/core/hle/kernel/readable_event.cpp b/src/core/hle/kernel/readable_event.cpp
index 0e5083f70..c2b798a4e 100644
--- a/src/core/hle/kernel/readable_event.cpp
+++ b/src/core/hle/kernel/readable_event.cpp
@@ -14,7 +14,7 @@ namespace Kernel {
 ReadableEvent::ReadableEvent(KernelCore& kernel) : WaitObject{kernel} {}
 ReadableEvent::~ReadableEvent() = default;
 
-bool ReadableEvent::ShouldWait(Thread* thread) const {
+bool ReadableEvent::ShouldWait(const Thread* thread) const {
     return !signaled;
 }
 
diff --git a/src/core/hle/kernel/readable_event.h b/src/core/hle/kernel/readable_event.h
index 77a9c362c..2eb9dcbb7 100644
--- a/src/core/hle/kernel/readable_event.h
+++ b/src/core/hle/kernel/readable_event.h
@@ -36,7 +36,7 @@ public:
         return HANDLE_TYPE;
     }
 
-    bool ShouldWait(Thread* thread) const override;
+    bool ShouldWait(const Thread* thread) const override;
     void Acquire(Thread* thread) override;
 
     /// Unconditionally clears the readable event's state.
diff --git a/src/core/hle/kernel/resource_limit.cpp b/src/core/hle/kernel/resource_limit.cpp
index 2f9695005..173f69915 100644
--- a/src/core/hle/kernel/resource_limit.cpp
+++ b/src/core/hle/kernel/resource_limit.cpp
@@ -16,11 +16,8 @@ constexpr std::size_t ResourceTypeToIndex(ResourceType type) {
 ResourceLimit::ResourceLimit(KernelCore& kernel) : Object{kernel} {}
 ResourceLimit::~ResourceLimit() = default;
 
-SharedPtr<ResourceLimit> ResourceLimit::Create(KernelCore& kernel, std::string name) {
-    SharedPtr<ResourceLimit> resource_limit(new ResourceLimit(kernel));
-
-    resource_limit->name = std::move(name);
-    return resource_limit;
+SharedPtr<ResourceLimit> ResourceLimit::Create(KernelCore& kernel) {
+    return new ResourceLimit(kernel);
 }
 
 s64 ResourceLimit::GetCurrentResourceValue(ResourceType resource) const {
diff --git a/src/core/hle/kernel/resource_limit.h b/src/core/hle/kernel/resource_limit.h
index 59dc11c22..70e09858a 100644
--- a/src/core/hle/kernel/resource_limit.h
+++ b/src/core/hle/kernel/resource_limit.h
@@ -31,16 +31,14 @@ constexpr bool IsValidResourceType(ResourceType type) {
 
 class ResourceLimit final : public Object {
 public:
-    /**
-     * Creates a resource limit object.
-     */
-    static SharedPtr<ResourceLimit> Create(KernelCore& kernel, std::string name = "Unknown");
+    /// Creates a resource limit object.
+    static SharedPtr<ResourceLimit> Create(KernelCore& kernel);
 
     std::string GetTypeName() const override {
         return "ResourceLimit";
     }
     std::string GetName() const override {
-        return name;
+        return GetTypeName();
     }
 
     static const HandleType HANDLE_TYPE = HandleType::ResourceLimit;
@@ -95,9 +93,6 @@ private:
     ResourceArray limits{};
     /// Current resource limit values.
     ResourceArray values{};
-
-    /// Name of resource limit object.
-    std::string name;
 };
 
 } // namespace Kernel
diff --git a/src/core/hle/kernel/scheduler.cpp b/src/core/hle/kernel/scheduler.cpp
index cc189cc64..ac501bf7f 100644
--- a/src/core/hle/kernel/scheduler.cpp
+++ b/src/core/hle/kernel/scheduler.cpp
@@ -29,8 +29,8 @@ Scheduler::~Scheduler() {
 }
 
 bool Scheduler::HaveReadyThreads() const {
-    std::lock_guard<std::mutex> lock(scheduler_mutex);
-    return ready_queue.get_first() != nullptr;
+    std::lock_guard lock{scheduler_mutex};
+    return !ready_queue.empty();
 }
 
 Thread* Scheduler::GetCurrentThread() const {
@@ -46,22 +46,27 @@ Thread* Scheduler::PopNextReadyThread() {
     Thread* thread = GetCurrentThread();
 
     if (thread && thread->GetStatus() == ThreadStatus::Running) {
+        if (ready_queue.empty()) {
+            return thread;
+        }
         // We have to do better than the current thread.
         // This call returns null when that's not possible.
-        next = ready_queue.pop_first_better(thread->GetPriority());
-        if (!next) {
-            // Otherwise just keep going with the current thread
+        next = ready_queue.front();
+        if (next == nullptr || next->GetPriority() >= thread->GetPriority()) {
             next = thread;
         }
     } else {
-        next = ready_queue.pop_first();
+        if (ready_queue.empty()) {
+            return nullptr;
+        }
+        next = ready_queue.front();
     }
 
     return next;
 }
 
 void Scheduler::SwitchContext(Thread* new_thread) {
-    Thread* const previous_thread = GetCurrentThread();
+    Thread* previous_thread = GetCurrentThread();
     Process* const previous_process = system.Kernel().CurrentProcess();
 
     UpdateLastContextSwitchTime(previous_thread, previous_process);
@@ -75,7 +80,7 @@ void Scheduler::SwitchContext(Thread* new_thread) {
         if (previous_thread->GetStatus() == ThreadStatus::Running) {
             // This is only the case when a reschedule is triggered without the current thread
             // yielding execution (i.e. an event triggered, system core time-sliced, etc)
-            ready_queue.push_front(previous_thread->GetPriority(), previous_thread);
+            ready_queue.add(previous_thread, previous_thread->GetPriority(), false);
             previous_thread->SetStatus(ThreadStatus::Ready);
         }
     }
@@ -90,7 +95,7 @@ void Scheduler::SwitchContext(Thread* new_thread) {
 
         current_thread = new_thread;
 
-        ready_queue.remove(new_thread->GetPriority(), new_thread);
+        ready_queue.remove(new_thread, new_thread->GetPriority());
         new_thread->SetStatus(ThreadStatus::Running);
 
         auto* const thread_owner_process = current_thread->GetOwnerProcess();
@@ -127,7 +132,7 @@ void Scheduler::UpdateLastContextSwitchTime(Thread* thread, Process* process) {
 }
 
 void Scheduler::Reschedule() {
-    std::lock_guard<std::mutex> lock(scheduler_mutex);
+    std::lock_guard lock{scheduler_mutex};
 
     Thread* cur = GetCurrentThread();
     Thread* next = PopNextReadyThread();
@@ -143,51 +148,54 @@ void Scheduler::Reschedule() {
     SwitchContext(next);
 }
 
-void Scheduler::AddThread(SharedPtr<Thread> thread, u32 priority) {
-    std::lock_guard<std::mutex> lock(scheduler_mutex);
+void Scheduler::AddThread(SharedPtr<Thread> thread) {
+    std::lock_guard lock{scheduler_mutex};
 
     thread_list.push_back(std::move(thread));
-    ready_queue.prepare(priority);
 }
 
 void Scheduler::RemoveThread(Thread* thread) {
-    std::lock_guard<std::mutex> lock(scheduler_mutex);
+    std::lock_guard lock{scheduler_mutex};
 
     thread_list.erase(std::remove(thread_list.begin(), thread_list.end(), thread),
                       thread_list.end());
 }
 
 void Scheduler::ScheduleThread(Thread* thread, u32 priority) {
-    std::lock_guard<std::mutex> lock(scheduler_mutex);
+    std::lock_guard lock{scheduler_mutex};
 
     ASSERT(thread->GetStatus() == ThreadStatus::Ready);
-    ready_queue.push_back(priority, thread);
+    ready_queue.add(thread, priority);
 }
 
 void Scheduler::UnscheduleThread(Thread* thread, u32 priority) {
-    std::lock_guard<std::mutex> lock(scheduler_mutex);
+    std::lock_guard lock{scheduler_mutex};
 
     ASSERT(thread->GetStatus() == ThreadStatus::Ready);
-    ready_queue.remove(priority, thread);
+    ready_queue.remove(thread, priority);
 }
 
 void Scheduler::SetThreadPriority(Thread* thread, u32 priority) {
-    std::lock_guard<std::mutex> lock(scheduler_mutex);
+    std::lock_guard lock{scheduler_mutex};
+    if (thread->GetPriority() == priority) {
+        return;
+    }
 
     // If thread was ready, adjust queues
     if (thread->GetStatus() == ThreadStatus::Ready)
-        ready_queue.move(thread, thread->GetPriority(), priority);
-    else
-        ready_queue.prepare(priority);
+        ready_queue.adjust(thread, thread->GetPriority(), priority);
 }
 
 Thread* Scheduler::GetNextSuggestedThread(u32 core, u32 maximum_priority) const {
-    std::lock_guard<std::mutex> lock(scheduler_mutex);
+    std::lock_guard lock{scheduler_mutex};
 
     const u32 mask = 1U << core;
-    return ready_queue.get_first_filter([mask, maximum_priority](Thread const* thread) {
-        return (thread->GetAffinityMask() & mask) != 0 && thread->GetPriority() < maximum_priority;
-    });
+    for (auto* thread : ready_queue) {
+        if ((thread->GetAffinityMask() & mask) != 0 && thread->GetPriority() < maximum_priority) {
+            return thread;
+        }
+    }
+    return nullptr;
 }
 
 void Scheduler::YieldWithoutLoadBalancing(Thread* thread) {
diff --git a/src/core/hle/kernel/scheduler.h b/src/core/hle/kernel/scheduler.h
index 1c5bf57d9..b29bf7be8 100644
--- a/src/core/hle/kernel/scheduler.h
+++ b/src/core/hle/kernel/scheduler.h
@@ -7,7 +7,7 @@
 #include <mutex>
 #include <vector>
 #include "common/common_types.h"
-#include "common/thread_queue_list.h"
+#include "common/multi_level_queue.h"
 #include "core/hle/kernel/object.h"
 #include "core/hle/kernel/thread.h"
 
@@ -38,7 +38,7 @@ public:
     u64 GetLastContextSwitchTicks() const;
 
     /// Adds a new thread to the scheduler
-    void AddThread(SharedPtr<Thread> thread, u32 priority);
+    void AddThread(SharedPtr<Thread> thread);
 
     /// Removes a thread from the scheduler
     void RemoveThread(Thread* thread);
@@ -156,7 +156,7 @@ private:
     std::vector<SharedPtr<Thread>> thread_list;
 
     /// Lists only ready thread ids.
-    Common::ThreadQueueList<Thread*, THREADPRIO_LOWEST + 1> ready_queue;
+    Common::MultiLevelQueue<Thread*, THREADPRIO_LOWEST + 1> ready_queue;
 
     SharedPtr<Thread> current_thread = nullptr;
 
diff --git a/src/core/hle/kernel/server_port.cpp b/src/core/hle/kernel/server_port.cpp
index 0e1515c89..708fdf9e1 100644
--- a/src/core/hle/kernel/server_port.cpp
+++ b/src/core/hle/kernel/server_port.cpp
@@ -30,7 +30,7 @@ void ServerPort::AppendPendingSession(SharedPtr<ServerSession> pending_session)
     pending_sessions.push_back(std::move(pending_session));
 }
 
-bool ServerPort::ShouldWait(Thread* thread) const {
+bool ServerPort::ShouldWait(const Thread* thread) const {
     // If there are no pending sessions, we wait until a new one is added.
     return pending_sessions.empty();
 }
diff --git a/src/core/hle/kernel/server_port.h b/src/core/hle/kernel/server_port.h
index 9bc667cf2..76293cb8b 100644
--- a/src/core/hle/kernel/server_port.h
+++ b/src/core/hle/kernel/server_port.h
@@ -75,7 +75,7 @@ public:
     /// waiting to be accepted by this port.
     void AppendPendingSession(SharedPtr<ServerSession> pending_session);
 
-    bool ShouldWait(Thread* thread) const override;
+    bool ShouldWait(const Thread* thread) const override;
     void Acquire(Thread* thread) override;
 
 private:
diff --git a/src/core/hle/kernel/server_session.cpp b/src/core/hle/kernel/server_session.cpp
index 4d8a337a7..40cec143e 100644
--- a/src/core/hle/kernel/server_session.cpp
+++ b/src/core/hle/kernel/server_session.cpp
@@ -46,7 +46,7 @@ ResultVal<SharedPtr<ServerSession>> ServerSession::Create(KernelCore& kernel, st
     return MakeResult(std::move(server_session));
 }
 
-bool ServerSession::ShouldWait(Thread* thread) const {
+bool ServerSession::ShouldWait(const Thread* thread) const {
     // Closed sessions should never wait, an error will be returned from svcReplyAndReceive.
     if (parent->client == nullptr)
         return false;
diff --git a/src/core/hle/kernel/server_session.h b/src/core/hle/kernel/server_session.h
index aea4ccfeb..79b84bade 100644
--- a/src/core/hle/kernel/server_session.h
+++ b/src/core/hle/kernel/server_session.h
@@ -82,7 +82,7 @@ public:
      */
     ResultCode HandleSyncRequest(SharedPtr<Thread> thread);
 
-    bool ShouldWait(Thread* thread) const override;
+    bool ShouldWait(const Thread* thread) const override;
 
     void Acquire(Thread* thread) override;
 
diff --git a/src/core/hle/kernel/shared_memory.cpp b/src/core/hle/kernel/shared_memory.cpp
index 62861da36..f15c5ee36 100644
--- a/src/core/hle/kernel/shared_memory.cpp
+++ b/src/core/hle/kernel/shared_memory.cpp
@@ -9,7 +9,6 @@
 #include "core/hle/kernel/errors.h"
 #include "core/hle/kernel/kernel.h"
 #include "core/hle/kernel/shared_memory.h"
-#include "core/memory.h"
 
 namespace Kernel {
 
@@ -119,7 +118,15 @@ ResultCode SharedMemory::Map(Process& target_process, VAddr address, MemoryPermi
                                                      ConvertPermissions(permissions));
 }
 
-ResultCode SharedMemory::Unmap(Process& target_process, VAddr address) {
+ResultCode SharedMemory::Unmap(Process& target_process, VAddr address, u64 unmap_size) {
+    if (unmap_size != size) {
+        LOG_ERROR(Kernel,
+                  "Invalid size passed to Unmap. Size must be equal to the size of the "
+                  "memory managed. Shared memory size=0x{:016X}, Unmap size=0x{:016X}",
+                  size, unmap_size);
+        return ERR_INVALID_SIZE;
+    }
+
     // TODO(Subv): Verify what happens if the application tries to unmap an address that is not
     // mapped to a SharedMemory.
     return target_process.VMManager().UnmapRange(address, size);
diff --git a/src/core/hle/kernel/shared_memory.h b/src/core/hle/kernel/shared_memory.h
index dab2a6bea..37e18c443 100644
--- a/src/core/hle/kernel/shared_memory.h
+++ b/src/core/hle/kernel/shared_memory.h
@@ -104,11 +104,17 @@ public:
 
     /**
      * Unmaps a shared memory block from the specified address in system memory
+     *
      * @param target_process Process from which to unmap the memory block.
-     * @param address Address in system memory where the shared memory block is mapped
+     * @param address        Address in system memory where the shared memory block is mapped.
+     * @param unmap_size     The amount of bytes to unmap from this shared memory instance.
+     *
      * @return Result code of the unmap operation
+     *
+     * @pre The given size to unmap must be the same size as the amount of memory managed by
+     *      the SharedMemory instance itself, otherwise ERR_INVALID_SIZE will be returned.
      */
-    ResultCode Unmap(Process& target_process, VAddr address);
+    ResultCode Unmap(Process& target_process, VAddr address, u64 unmap_size);
 
     /**
      * Gets a pointer to the shared memory block
diff --git a/src/core/hle/kernel/svc.cpp b/src/core/hle/kernel/svc.cpp
index e5e7f99e1..ab10db3df 100644
--- a/src/core/hle/kernel/svc.cpp
+++ b/src/core/hle/kernel/svc.cpp
@@ -175,11 +175,8 @@ static ResultCode SetHeapSize(VAddr* heap_addr, u64 heap_size) {
         return ERR_INVALID_SIZE;
     }
 
-    auto& vm_manager = Core::CurrentProcess()->VMManager();
-    const VAddr heap_base = vm_manager.GetHeapRegionBaseAddress();
-    const auto alloc_result =
-        vm_manager.HeapAllocate(heap_base, heap_size, VMAPermission::ReadWrite);
-
+    auto& vm_manager = Core::System::GetInstance().Kernel().CurrentProcess()->VMManager();
+    const auto alloc_result = vm_manager.SetHeapSize(heap_size);
     if (alloc_result.Failed()) {
         return alloc_result.Code();
     }
@@ -712,7 +709,7 @@ static ResultCode GetInfo(u64* result, u64 info_id, u64 handle, u64 info_sub_id)
         HeapRegionBaseAddr = 4,
         HeapRegionSize = 5,
         TotalMemoryUsage = 6,
-        TotalHeapUsage = 7,
+        TotalPhysicalMemoryUsed = 7,
         IsCurrentProcessBeingDebugged = 8,
         RegisterResourceLimit = 9,
         IdleTickCount = 10,
@@ -748,7 +745,7 @@ static ResultCode GetInfo(u64* result, u64 info_id, u64 handle, u64 info_sub_id)
     case GetInfoType::NewMapRegionBaseAddr:
     case GetInfoType::NewMapRegionSize:
     case GetInfoType::TotalMemoryUsage:
-    case GetInfoType::TotalHeapUsage:
+    case GetInfoType::TotalPhysicalMemoryUsed:
     case GetInfoType::IsVirtualAddressMemoryEnabled:
     case GetInfoType::PersonalMmHeapUsage:
     case GetInfoType::TitleId:
@@ -808,8 +805,8 @@ static ResultCode GetInfo(u64* result, u64 info_id, u64 handle, u64 info_sub_id)
             *result = process->VMManager().GetTotalMemoryUsage();
             return RESULT_SUCCESS;
 
-        case GetInfoType::TotalHeapUsage:
-            *result = process->VMManager().GetTotalHeapUsage();
+        case GetInfoType::TotalPhysicalMemoryUsed:
+            *result = process->GetTotalPhysicalMemoryUsed();
             return RESULT_SUCCESS;
 
         case GetInfoType::IsVirtualAddressMemoryEnabled:
@@ -1143,7 +1140,7 @@ static ResultCode UnmapSharedMemory(Handle shared_memory_handle, VAddr addr, u64
         return ERR_INVALID_MEMORY_RANGE;
     }
 
-    return shared_memory->Unmap(*current_process, addr);
+    return shared_memory->Unmap(*current_process, addr, size);
 }
 
 static ResultCode QueryProcessMemory(VAddr memory_info_address, VAddr page_info_address,
@@ -1356,7 +1353,7 @@ static ResultCode WaitProcessWideKeyAtomic(VAddr mutex_addr, VAddr condition_var
     current_thread->SetCondVarWaitAddress(condition_variable_addr);
     current_thread->SetMutexWaitAddress(mutex_addr);
     current_thread->SetWaitHandle(thread_handle);
-    current_thread->SetStatus(ThreadStatus::WaitMutex);
+    current_thread->SetStatus(ThreadStatus::WaitCondVar);
     current_thread->InvalidateWakeupCallback();
 
     current_thread->WakeAfterDelay(nano_seconds);
@@ -1400,10 +1397,10 @@ static ResultCode SignalProcessWideKey(VAddr condition_variable_addr, s32 target
     // them all.
     std::size_t last = waiting_threads.size();
     if (target != -1)
-        last = target;
+        last = std::min(waiting_threads.size(), static_cast<std::size_t>(target));
 
     // If there are no threads waiting on this condition variable, just exit
-    if (last > waiting_threads.size())
+    if (last == 0)
         return RESULT_SUCCESS;
 
     for (std::size_t index = 0; index < last; ++index) {
@@ -1411,6 +1408,9 @@ static ResultCode SignalProcessWideKey(VAddr condition_variable_addr, s32 target
 
         ASSERT(thread->GetCondVarWaitAddress() == condition_variable_addr);
 
+        // liberate Cond Var Thread.
+        thread->SetCondVarWaitAddress(0);
+
         std::size_t current_core = Core::System::GetInstance().CurrentCoreIndex();
 
         auto& monitor = Core::System::GetInstance().Monitor();
@@ -1429,10 +1429,9 @@ static ResultCode SignalProcessWideKey(VAddr condition_variable_addr, s32 target
             }
         } while (!monitor.ExclusiveWrite32(current_core, thread->GetMutexWaitAddress(),
                                            thread->GetWaitHandle()));
-
         if (mutex_val == 0) {
             // We were able to acquire the mutex, resume this thread.
-            ASSERT(thread->GetStatus() == ThreadStatus::WaitMutex);
+            ASSERT(thread->GetStatus() == ThreadStatus::WaitCondVar);
             thread->ResumeFromWait();
 
             auto* const lock_owner = thread->GetLockOwner();
@@ -1442,8 +1441,8 @@ static ResultCode SignalProcessWideKey(VAddr condition_variable_addr, s32 target
 
             thread->SetLockOwner(nullptr);
             thread->SetMutexWaitAddress(0);
-            thread->SetCondVarWaitAddress(0);
             thread->SetWaitHandle(0);
+            Core::System::GetInstance().CpuCore(thread->GetProcessorID()).PrepareReschedule();
         } else {
             // Atomically signal that the mutex now has a waiting thread.
             do {
@@ -1462,12 +1461,11 @@ static ResultCode SignalProcessWideKey(VAddr condition_variable_addr, s32 target
             const auto& handle_table = Core::CurrentProcess()->GetHandleTable();
             auto owner = handle_table.Get<Thread>(owner_handle);
             ASSERT(owner);
-            ASSERT(thread->GetStatus() == ThreadStatus::WaitMutex);
+            ASSERT(thread->GetStatus() == ThreadStatus::WaitCondVar);
             thread->InvalidateWakeupCallback();
+            thread->SetStatus(ThreadStatus::WaitMutex);
 
             owner->AddMutexWaiter(thread);
-
-            Core::System::GetInstance().CpuCore(thread->GetProcessorID()).PrepareReschedule();
         }
     }
 
@@ -1985,6 +1983,83 @@ static ResultCode SetResourceLimitLimitValue(Handle resource_limit, u32 resource
     return RESULT_SUCCESS;
 }
 
+static ResultCode GetProcessList(u32* out_num_processes, VAddr out_process_ids,
+                                 u32 out_process_ids_size) {
+    LOG_DEBUG(Kernel_SVC, "called. out_process_ids=0x{:016X}, out_process_ids_size={}",
+              out_process_ids, out_process_ids_size);
+
+    // If the supplied size is negative or greater than INT32_MAX / sizeof(u64), bail.
+    if ((out_process_ids_size & 0xF0000000) != 0) {
+        LOG_ERROR(Kernel_SVC,
+                  "Supplied size outside [0, 0x0FFFFFFF] range. out_process_ids_size={}",
+                  out_process_ids_size);
+        return ERR_OUT_OF_RANGE;
+    }
+
+    const auto& kernel = Core::System::GetInstance().Kernel();
+    const auto& vm_manager = kernel.CurrentProcess()->VMManager();
+    const auto total_copy_size = out_process_ids_size * sizeof(u64);
+
+    if (out_process_ids_size > 0 &&
+        !vm_manager.IsWithinAddressSpace(out_process_ids, total_copy_size)) {
+        LOG_ERROR(Kernel_SVC, "Address range outside address space. begin=0x{:016X}, end=0x{:016X}",
+                  out_process_ids, out_process_ids + total_copy_size);
+        return ERR_INVALID_ADDRESS_STATE;
+    }
+
+    const auto& process_list = kernel.GetProcessList();
+    const auto num_processes = process_list.size();
+    const auto copy_amount = std::min(std::size_t{out_process_ids_size}, num_processes);
+
+    for (std::size_t i = 0; i < copy_amount; ++i) {
+        Memory::Write64(out_process_ids, process_list[i]->GetProcessID());
+        out_process_ids += sizeof(u64);
+    }
+
+    *out_num_processes = static_cast<u32>(num_processes);
+    return RESULT_SUCCESS;
+}
+
+ResultCode GetThreadList(u32* out_num_threads, VAddr out_thread_ids, u32 out_thread_ids_size,
+                         Handle debug_handle) {
+    // TODO: Handle this case when debug events are supported.
+    UNIMPLEMENTED_IF(debug_handle != InvalidHandle);
+
+    LOG_DEBUG(Kernel_SVC, "called. out_thread_ids=0x{:016X}, out_thread_ids_size={}",
+              out_thread_ids, out_thread_ids_size);
+
+    // If the size is negative or larger than INT32_MAX / sizeof(u64)
+    if ((out_thread_ids_size & 0xF0000000) != 0) {
+        LOG_ERROR(Kernel_SVC, "Supplied size outside [0, 0x0FFFFFFF] range. size={}",
+                  out_thread_ids_size);
+        return ERR_OUT_OF_RANGE;
+    }
+
+    const auto* const current_process = Core::System::GetInstance().Kernel().CurrentProcess();
+    const auto& vm_manager = current_process->VMManager();
+    const auto total_copy_size = out_thread_ids_size * sizeof(u64);
+
+    if (out_thread_ids_size > 0 &&
+        !vm_manager.IsWithinAddressSpace(out_thread_ids, total_copy_size)) {
+        LOG_ERROR(Kernel_SVC, "Address range outside address space. begin=0x{:016X}, end=0x{:016X}",
+                  out_thread_ids, out_thread_ids + total_copy_size);
+        return ERR_INVALID_ADDRESS_STATE;
+    }
+
+    const auto& thread_list = current_process->GetThreadList();
+    const auto num_threads = thread_list.size();
+    const auto copy_amount = std::min(std::size_t{out_thread_ids_size}, num_threads);
+
+    auto list_iter = thread_list.cbegin();
+    for (std::size_t i = 0; i < copy_amount; ++i, ++list_iter) {
+        Memory::Write64(out_thread_ids, (*list_iter)->GetThreadID());
+        out_thread_ids += sizeof(u64);
+    }
+
+    *out_num_threads = static_cast<u32>(num_threads);
+    return RESULT_SUCCESS;
+}
+
 namespace {
 struct FunctionDef {
     using Func = void();
@@ -2097,8 +2172,8 @@ static const FunctionDef SVC_Table[] = {
     {0x62, nullptr, "TerminateDebugProcess"},
     {0x63, nullptr, "GetDebugEvent"},
     {0x64, nullptr, "ContinueDebugEvent"},
-    {0x65, nullptr, "GetProcessList"},
-    {0x66, nullptr, "GetThreadList"},
+    {0x65, SvcWrap<GetProcessList>, "GetProcessList"},
+    {0x66, SvcWrap<GetThreadList>, "GetThreadList"},
     {0x67, nullptr, "GetDebugThreadContext"},
     {0x68, nullptr, "SetDebugThreadContext"},
     {0x69, nullptr, "QueryDebugProcessMemory"},
@@ -2140,7 +2215,7 @@ void CallSVC(u32 immediate) {
     MICROPROFILE_SCOPE(Kernel_SVC);
 
     // Lock the global kernel mutex when we enter the kernel HLE.
-    std::lock_guard<std::recursive_mutex> lock(HLE::g_hle_lock);
+    std::lock_guard lock{HLE::g_hle_lock};
 
     const FunctionDef* info = GetSVCInfo(immediate);
     if (info) {
diff --git a/src/core/hle/kernel/svc_wrap.h b/src/core/hle/kernel/svc_wrap.h
index 2a2c2c5ea..b3733680f 100644
--- a/src/core/hle/kernel/svc_wrap.h
+++ b/src/core/hle/kernel/svc_wrap.h
@@ -78,6 +78,14 @@ void SvcWrap() {
     FuncReturn(retval);
 }
 
+template <ResultCode func(u32*, u64, u32)>
+void SvcWrap() {
+    u32 param_1 = 0;
+    const u32 retval = func(&param_1, Param(1), static_cast<u32>(Param(2))).raw;
+    Core::CurrentArmInterface().SetReg(1, param_1);
+    FuncReturn(retval);
+}
+
 template <ResultCode func(u64*, u32)>
 void SvcWrap() {
     u64 param_1 = 0;
diff --git a/src/core/hle/kernel/thread.cpp b/src/core/hle/kernel/thread.cpp
index 3b22e8e0d..1b891f632 100644
--- a/src/core/hle/kernel/thread.cpp
+++ b/src/core/hle/kernel/thread.cpp
@@ -28,7 +28,7 @@
 
 namespace Kernel {
 
-bool Thread::ShouldWait(Thread* thread) const {
+bool Thread::ShouldWait(const Thread* thread) const {
     return status != ThreadStatus::Dead;
 }
 
@@ -62,6 +62,8 @@ void Thread::Stop() {
     }
     wait_objects.clear();
 
+    owner_process->UnregisterThread(this);
+
     // Mark the TLS slot in the thread's page as free.
     owner_process->FreeTLSSlot(tls_address);
 }
@@ -105,6 +107,7 @@ void Thread::ResumeFromWait() {
     case ThreadStatus::WaitSleep:
     case ThreadStatus::WaitIPC:
     case ThreadStatus::WaitMutex:
+    case ThreadStatus::WaitCondVar:
     case ThreadStatus::WaitArb:
         break;
 
@@ -198,9 +201,11 @@ ResultVal<SharedPtr<Thread>> Thread::Create(KernelCore& kernel, std::string name
     thread->callback_handle = kernel.ThreadWakeupCallbackHandleTable().Create(thread).Unwrap();
     thread->owner_process = &owner_process;
     thread->scheduler = &system.Scheduler(processor_id);
-    thread->scheduler->AddThread(thread, priority);
+    thread->scheduler->AddThread(thread);
     thread->tls_address = thread->owner_process->MarkNextAvailableTLSSlotAsUsed(*thread);
 
+    thread->owner_process->RegisterThread(thread.get());
+
     // TODO(peachum): move to ScheduleThread() when scheduler is added so selected core is used
     // to initialize the context
     ResetThreadContext(thread->context, stack_top, entry_point, arg);
@@ -228,16 +233,16 @@ void Thread::SetWaitSynchronizationOutput(s32 output) {
     context.cpu_registers[1] = output;
 }
 
-s32 Thread::GetWaitObjectIndex(WaitObject* object) const {
+s32 Thread::GetWaitObjectIndex(const WaitObject* object) const {
     ASSERT_MSG(!wait_objects.empty(), "Thread is not waiting for anything");
-    auto match = std::find(wait_objects.rbegin(), wait_objects.rend(), object);
+    const auto match = std::find(wait_objects.rbegin(), wait_objects.rend(), object);
     return static_cast<s32>(std::distance(match, wait_objects.rend()) - 1);
 }
 
 VAddr Thread::GetCommandBufferAddress() const {
     // Offset from the start of TLS at which the IPC command buffer begins.
-    static constexpr int CommandHeaderOffset = 0x80;
-    return GetTLSAddress() + CommandHeaderOffset;
+    constexpr u64 command_header_offset = 0x80;
+    return GetTLSAddress() + command_header_offset;
 }
 
 void Thread::SetStatus(ThreadStatus new_status) {
@@ -351,7 +356,7 @@ void Thread::ChangeScheduler() {
     if (*new_processor_id != processor_id) {
         // Remove thread from previous core's scheduler
         scheduler->RemoveThread(this);
-        next_scheduler.AddThread(this, current_priority);
+        next_scheduler.AddThread(this);
     }
 
     processor_id = *new_processor_id;
@@ -366,7 +371,7 @@ void Thread::ChangeScheduler() {
     system.CpuCore(processor_id).PrepareReschedule();
 }
 
-bool Thread::AllWaitObjectsReady() {
+bool Thread::AllWaitObjectsReady() const {
     return std::none_of(
         wait_objects.begin(), wait_objects.end(),
         [this](const SharedPtr<WaitObject>& object) { return object->ShouldWait(this); });
diff --git a/src/core/hle/kernel/thread.h b/src/core/hle/kernel/thread.h
index faad5f391..73e5d1bb4 100644
--- a/src/core/hle/kernel/thread.h
+++ b/src/core/hle/kernel/thread.h
@@ -51,7 +51,8 @@ enum class ThreadStatus {
     WaitIPC,      ///< Waiting for the reply from an IPC request
     WaitSynchAny, ///< Waiting due to WaitSynch1 or WaitSynchN with wait_all = false
     WaitSynchAll, ///< Waiting due to WaitSynchronizationN with wait_all = true
-    WaitMutex,    ///< Waiting due to an ArbitrateLock/WaitProcessWideKey svc
+    WaitMutex,    ///< Waiting due to an ArbitrateLock svc
+    WaitCondVar,  ///< Waiting due to an WaitProcessWideKey svc
     WaitArb,      ///< Waiting due to a SignalToAddress/WaitForAddress svc
     Dormant,      ///< Created but not yet made ready
     Dead          ///< Run to completion, or forcefully terminated
@@ -110,7 +111,7 @@ public:
         return HANDLE_TYPE;
     }
 
-    bool ShouldWait(Thread* thread) const override;
+    bool ShouldWait(const Thread* thread) const override;
     void Acquire(Thread* thread) override;
 
     /**
@@ -204,7 +205,7 @@ public:
      * object in the list.
      * @param object Object to query the index of.
      */
-    s32 GetWaitObjectIndex(WaitObject* object) const;
+    s32 GetWaitObjectIndex(const WaitObject* object) const;
 
     /**
      * Stops a thread, invalidating it from further use
@@ -298,7 +299,7 @@ public:
     }
 
     /// Determines whether all the objects this thread is waiting on are ready.
-    bool AllWaitObjectsReady();
+    bool AllWaitObjectsReady() const;
 
     const MutexWaitingThreads& GetMutexWaitingThreads() const {
         return wait_mutex_threads;
diff --git a/src/core/hle/kernel/transfer_memory.cpp b/src/core/hle/kernel/transfer_memory.cpp
index 23228e1b5..26c4e5e67 100644
--- a/src/core/hle/kernel/transfer_memory.cpp
+++ b/src/core/hle/kernel/transfer_memory.cpp
@@ -14,8 +14,8 @@ namespace Kernel {
 TransferMemory::TransferMemory(KernelCore& kernel) : Object{kernel} {}
 TransferMemory::~TransferMemory() = default;
 
-SharedPtr<TransferMemory> TransferMemory::Create(KernelCore& kernel, VAddr base_address,
-                                                 size_t size, MemoryPermission permissions) {
+SharedPtr<TransferMemory> TransferMemory::Create(KernelCore& kernel, VAddr base_address, u64 size,
+                                                 MemoryPermission permissions) {
     SharedPtr<TransferMemory> transfer_memory{new TransferMemory(kernel)};
 
     transfer_memory->base_address = base_address;
@@ -26,7 +26,15 @@ SharedPtr<TransferMemory> TransferMemory::Create(KernelCore& kernel, VAddr base_
     return transfer_memory;
 }
 
-ResultCode TransferMemory::MapMemory(VAddr address, size_t size, MemoryPermission permissions) {
+const u8* TransferMemory::GetPointer() const {
+    return backing_block.get()->data();
+}
+
+u64 TransferMemory::GetSize() const {
+    return memory_size;
+}
+
+ResultCode TransferMemory::MapMemory(VAddr address, u64 size, MemoryPermission permissions) {
     if (memory_size != size) {
         return ERR_INVALID_SIZE;
     }
@@ -39,13 +47,13 @@ ResultCode TransferMemory::MapMemory(VAddr address, size_t size, MemoryPermissio
         return ERR_INVALID_STATE;
     }
 
+    backing_block = std::make_shared<std::vector<u8>>(size);
+
     const auto map_state = owner_permissions == MemoryPermission::None
                                ? MemoryState::TransferMemoryIsolated
                                : MemoryState::TransferMemory;
     auto& vm_manager = owner_process->VMManager();
-    const auto map_result = vm_manager.MapMemoryBlock(
-        address, std::make_shared<std::vector<u8>>(size), 0, size, map_state);
-
+    const auto map_result = vm_manager.MapMemoryBlock(address, backing_block, 0, size, map_state);
     if (map_result.Failed()) {
         return map_result.Code();
     }
@@ -54,7 +62,7 @@ ResultCode TransferMemory::MapMemory(VAddr address, size_t size, MemoryPermissio
     return RESULT_SUCCESS;
 }
 
-ResultCode TransferMemory::UnmapMemory(VAddr address, size_t size) {
+ResultCode TransferMemory::UnmapMemory(VAddr address, u64 size) {
     if (memory_size != size) {
         return ERR_INVALID_SIZE;
     }
diff --git a/src/core/hle/kernel/transfer_memory.h b/src/core/hle/kernel/transfer_memory.h
index ec294951e..a140b1e2b 100644
--- a/src/core/hle/kernel/transfer_memory.h
+++ b/src/core/hle/kernel/transfer_memory.h
@@ -4,6 +4,9 @@
 
 #pragma once
 
+#include <memory>
+#include <vector>
+
 #include "core/hle/kernel/object.h"
 
 union ResultCode;
@@ -25,7 +28,7 @@ class TransferMemory final : public Object {
 public:
     static constexpr HandleType HANDLE_TYPE = HandleType::TransferMemory;
 
-    static SharedPtr<TransferMemory> Create(KernelCore& kernel, VAddr base_address, size_t size,
+    static SharedPtr<TransferMemory> Create(KernelCore& kernel, VAddr base_address, u64 size,
                                             MemoryPermission permissions);
 
     TransferMemory(const TransferMemory&) = delete;
@@ -46,6 +49,12 @@ public:
         return HANDLE_TYPE;
     }
 
+    /// Gets a pointer to the backing block of this instance.
+    const u8* GetPointer() const;
+
+    /// Gets the size of the memory backing this instance in bytes.
+    u64 GetSize() const;
+
     /// Attempts to map transfer memory with the given range and memory permissions.
     ///
     /// @param address     The base address to being mapping memory at.
@@ -56,7 +65,7 @@ public:
     ///      the same values that were given when creating the transfer memory
     ///      instance.
     ///
-    ResultCode MapMemory(VAddr address, size_t size, MemoryPermission permissions);
+    ResultCode MapMemory(VAddr address, u64 size, MemoryPermission permissions);
 
     /// Unmaps the transfer memory with the given range
     ///
@@ -66,17 +75,20 @@ public:
     /// @pre The given address and size must be the same as the ones used
     ///      to create the transfer memory instance.
     ///
-    ResultCode UnmapMemory(VAddr address, size_t size);
+    ResultCode UnmapMemory(VAddr address, u64 size);
 
 private:
     explicit TransferMemory(KernelCore& kernel);
     ~TransferMemory() override;
 
+    /// Memory block backing this instance.
+    std::shared_ptr<std::vector<u8>> backing_block;
+
     /// The base address for the memory managed by this instance.
     VAddr base_address = 0;
 
     /// Size of the memory, in bytes, that this instance manages.
-    size_t memory_size = 0;
+    u64 memory_size = 0;
 
     /// The memory permissions that are applied to this instance.
     MemoryPermission owner_permissions{};
diff --git a/src/core/hle/kernel/vm_manager.cpp b/src/core/hle/kernel/vm_manager.cpp
index 22bf55ce7..ec0a480ce 100644
--- a/src/core/hle/kernel/vm_manager.cpp
+++ b/src/core/hle/kernel/vm_manager.cpp
@@ -256,57 +256,50 @@ ResultCode VMManager::ReprotectRange(VAddr target, u64 size, VMAPermission new_p
     return RESULT_SUCCESS;
 }
 
-ResultVal<VAddr> VMManager::HeapAllocate(VAddr target, u64 size, VMAPermission perms) {
-    if (!IsWithinHeapRegion(target, size)) {
-        return ERR_INVALID_ADDRESS;
+ResultVal<VAddr> VMManager::SetHeapSize(u64 size) {
+    if (size > GetHeapRegionSize()) {
+        return ERR_OUT_OF_MEMORY;
+    }
+
+    // No need to do any additional work if the heap is already the given size.
+    if (size == GetCurrentHeapSize()) {
+        return MakeResult(heap_region_base);
     }
 
     if (heap_memory == nullptr) {
         // Initialize heap
-        heap_memory = std::make_shared<std::vector<u8>>();
-        heap_start = heap_end = target;
+        heap_memory = std::make_shared<std::vector<u8>>(size);
+        heap_end = heap_region_base + size;
     } else {
-        UnmapRange(heap_start, heap_end - heap_start);
-    }
-
-    // If necessary, expand backing vector to cover new heap extents.
-    if (target < heap_start) {
-        heap_memory->insert(begin(*heap_memory), heap_start - target, 0);
-        heap_start = target;
-        RefreshMemoryBlockMappings(heap_memory.get());
-    }
-    if (target + size > heap_end) {
-        heap_memory->insert(end(*heap_memory), (target + size) - heap_end, 0);
-        heap_end = target + size;
-        RefreshMemoryBlockMappings(heap_memory.get());
+        UnmapRange(heap_region_base, GetCurrentHeapSize());
     }
-    ASSERT(heap_end - heap_start == heap_memory->size());
 
-    CASCADE_RESULT(auto vma, MapMemoryBlock(target, heap_memory, target - heap_start, size,
-                                            MemoryState::Heap));
-    Reprotect(vma, perms);
+    // If necessary, expand backing vector to cover new heap extents in
+    // the case of allocating. Otherwise, shrink the backing memory,
+    // if a smaller heap has been requested.
+    const u64 old_heap_size = GetCurrentHeapSize();
+    if (size > old_heap_size) {
+        const u64 alloc_size = size - old_heap_size;
 
-    heap_used = size;
-
-    return MakeResult<VAddr>(heap_end - size);
-}
+        heap_memory->insert(heap_memory->end(), alloc_size, 0);
+        RefreshMemoryBlockMappings(heap_memory.get());
+    } else if (size < old_heap_size) {
+        heap_memory->resize(size);
+        heap_memory->shrink_to_fit();
 
-ResultCode VMManager::HeapFree(VAddr target, u64 size) {
-    if (!IsWithinHeapRegion(target, size)) {
-        return ERR_INVALID_ADDRESS;
+        RefreshMemoryBlockMappings(heap_memory.get());
     }
 
-    if (size == 0) {
-        return RESULT_SUCCESS;
-    }
+    heap_end = heap_region_base + size;
+    ASSERT(GetCurrentHeapSize() == heap_memory->size());
 
-    const ResultCode result = UnmapRange(target, size);
-    if (result.IsError()) {
-        return result;
+    const auto mapping_result =
+        MapMemoryBlock(heap_region_base, heap_memory, 0, size, MemoryState::Heap);
+    if (mapping_result.Failed()) {
+        return mapping_result.Code();
     }
 
-    heap_used -= size;
-    return RESULT_SUCCESS;
+    return MakeResult<VAddr>(heap_region_base);
 }
 
 MemoryInfo VMManager::QueryMemory(VAddr address) const {
@@ -598,6 +591,7 @@ void VMManager::InitializeMemoryRegionRanges(FileSys::ProgramAddressSpaceType ty
 
     heap_region_base = map_region_end;
     heap_region_end = heap_region_base + heap_region_size;
+    heap_end = heap_region_base;
 
     new_map_region_base = heap_region_end;
     new_map_region_end = new_map_region_base + new_map_region_size;
@@ -692,10 +686,6 @@ u64 VMManager::GetTotalMemoryUsage() const {
     return 0xF8000000;
 }
 
-u64 VMManager::GetTotalHeapUsage() const {
-    return heap_used;
-}
-
 VAddr VMManager::GetAddressSpaceBaseAddress() const {
     return address_space_base;
 }
@@ -778,6 +768,10 @@ u64 VMManager::GetHeapRegionSize() const {
     return heap_region_end - heap_region_base;
 }
 
+u64 VMManager::GetCurrentHeapSize() const {
+    return heap_end - heap_region_base;
+}
+
 bool VMManager::IsWithinHeapRegion(VAddr address, u64 size) const {
     return IsInsideAddressRange(address, size, GetHeapRegionBaseAddress(),
                                 GetHeapRegionEndAddress());
diff --git a/src/core/hle/kernel/vm_manager.h b/src/core/hle/kernel/vm_manager.h
index 7cdff6094..6f484b7bf 100644
--- a/src/core/hle/kernel/vm_manager.h
+++ b/src/core/hle/kernel/vm_manager.h
@@ -380,11 +380,41 @@ public:
     /// Changes the permissions of a range of addresses, splitting VMAs as necessary.
     ResultCode ReprotectRange(VAddr target, u64 size, VMAPermission new_perms);
 
-    ResultVal<VAddr> HeapAllocate(VAddr target, u64 size, VMAPermission perms);
-    ResultCode HeapFree(VAddr target, u64 size);
-
     ResultCode MirrorMemory(VAddr dst_addr, VAddr src_addr, u64 size, MemoryState state);
 
+    /// Attempts to allocate a heap with the given size.
+    ///
+    /// @param size The size of the heap to allocate in bytes.
+    ///
+    /// @note If a heap is currently allocated, and this is called
+    ///       with a size that is equal to the size of the current heap,
+    ///       then this function will do nothing and return the current
+    ///       heap's starting address, as there's no need to perform
+    ///       any additional heap allocation work.
+    ///
+    /// @note If a heap is currently allocated, and this is called
+    ///       with a size less than the current heap's size, then
+    ///       this function will attempt to shrink the heap.
+    ///
+    /// @note If a heap is currently allocated, and this is called
+    ///       with a size larger than the current heap's size, then
+    ///       this function will attempt to extend the size of the heap.
+    ///
+    /// @returns A result indicating either success or failure.
+    ///          <p>
+    ///          If successful, this function will return a result
+    ///          containing the starting address to the allocated heap.
+    ///          <p>
+    ///          If unsuccessful, this function will return a result
+    ///          containing an error code.
+    ///
+    /// @pre The given size must lie within the allowable heap
+    ///      memory region managed by this VMManager instance.
+    ///      Failure to abide by this will result in ERR_OUT_OF_MEMORY
+    ///      being returned as the result.
+    ///
+    ResultVal<VAddr> SetHeapSize(u64 size);
+
     /// Queries the memory manager for information about the given address.
     ///
     /// @param address The address to query the memory manager about for information.
@@ -418,9 +448,6 @@ public:
     /// Gets the total memory usage, used by svcGetInfo
     u64 GetTotalMemoryUsage() const;
 
-    /// Gets the total heap usage, used by svcGetInfo
-    u64 GetTotalHeapUsage() const;
-
     /// Gets the address space base address
     VAddr GetAddressSpaceBaseAddress() const;
 
@@ -469,6 +496,13 @@ public:
     /// Gets the total size of the heap region in bytes.
     u64 GetHeapRegionSize() const;
 
+    /// Gets the total size of the current heap in bytes.
+    ///
+    /// @note This is the current allocated heap size, not the size
+    ///       of the region it's allowed to exist within.
+    ///
+    u64 GetCurrentHeapSize() const;
+
     /// Determines whether or not the specified range is within the heap region.
     bool IsWithinHeapRegion(VAddr address, u64 size) const;
 
@@ -617,9 +651,6 @@ private:
     VAddr new_map_region_base = 0;
     VAddr new_map_region_end = 0;
 
-    VAddr main_code_region_base = 0;
-    VAddr main_code_region_end = 0;
-
     VAddr tls_io_region_base = 0;
     VAddr tls_io_region_end = 0;
 
@@ -628,9 +659,9 @@ private:
     // This makes deallocation and reallocation of holes fast and keeps process memory contiguous
     // in the emulator address space, allowing Memory::GetPointer to be reasonably safe.
     std::shared_ptr<std::vector<u8>> heap_memory;
-    // The left/right bounds of the address space covered by heap_memory.
-    VAddr heap_start = 0;
+
+    // The end of the currently allocated heap. This is not an inclusive
+    // end of the range. This is essentially 'base_address + current_size'.
     VAddr heap_end = 0;
-    u64 heap_used = 0;
 };
 } // namespace Kernel
diff --git a/src/core/hle/kernel/wait_object.h b/src/core/hle/kernel/wait_object.h
index 5987fb971..04464a51a 100644
--- a/src/core/hle/kernel/wait_object.h
+++ b/src/core/hle/kernel/wait_object.h
@@ -24,7 +24,7 @@ public:
      * @param thread The thread about which we're deciding.
      * @return True if the current thread should wait due to this object being unavailable
      */
-    virtual bool ShouldWait(Thread* thread) const = 0;
+    virtual bool ShouldWait(const Thread* thread) const = 0;
 
     /// Acquire/lock the object for the specified thread if it is available
     virtual void Acquire(Thread* thread) = 0;
diff --git a/src/core/hle/service/am/am.cpp b/src/core/hle/service/am/am.cpp
index 9c44e27c6..85271d418 100644
--- a/src/core/hle/service/am/am.cpp
+++ b/src/core/hle/service/am/am.cpp
@@ -13,7 +13,7 @@
 #include "core/hle/kernel/kernel.h"
 #include "core/hle/kernel/process.h"
 #include "core/hle/kernel/readable_event.h"
-#include "core/hle/kernel/shared_memory.h"
+#include "core/hle/kernel/transfer_memory.h"
 #include "core/hle/kernel/writable_event.h"
 #include "core/hle/service/acc/profile_manager.h"
 #include "core/hle/service/am/am.h"
@@ -239,8 +239,8 @@ ISelfController::ISelfController(std::shared_ptr<NVFlinger::NVFlinger> nvflinger
         {0, nullptr, "Exit"},
         {1, &ISelfController::LockExit, "LockExit"},
         {2, &ISelfController::UnlockExit, "UnlockExit"},
-        {3, nullptr, "EnterFatalSection"},
-        {4, nullptr, "LeaveFatalSection"},
+        {3, &ISelfController::EnterFatalSection, "EnterFatalSection"},
+        {4, &ISelfController::LeaveFatalSection, "LeaveFatalSection"},
         {9, &ISelfController::GetLibraryAppletLaunchableEvent, "GetLibraryAppletLaunchableEvent"},
         {10, &ISelfController::SetScreenShotPermission, "SetScreenShotPermission"},
         {11, &ISelfController::SetOperationModeChangedNotification, "SetOperationModeChangedNotification"},
@@ -285,41 +285,54 @@ ISelfController::ISelfController(std::shared_ptr<NVFlinger::NVFlinger> nvflinger
 
 ISelfController::~ISelfController() = default;
 
-void ISelfController::SetFocusHandlingMode(Kernel::HLERequestContext& ctx) {
-    // Takes 3 input u8s with each field located immediately after the previous
-    // u8, these are bool flags. No output.
+void ISelfController::LockExit(Kernel::HLERequestContext& ctx) {
     LOG_WARNING(Service_AM, "(STUBBED) called");
 
-    IPC::RequestParser rp{ctx};
+    IPC::ResponseBuilder rb{ctx, 2};
+    rb.Push(RESULT_SUCCESS);
+}
 
-    struct FocusHandlingModeParams {
-        u8 unknown0;
-        u8 unknown1;
-        u8 unknown2;
-    };
-    auto flags = rp.PopRaw<FocusHandlingModeParams>();
+void ISelfController::UnlockExit(Kernel::HLERequestContext& ctx) {
+    LOG_WARNING(Service_AM, "(STUBBED) called");
 
     IPC::ResponseBuilder rb{ctx, 2};
     rb.Push(RESULT_SUCCESS);
 }
 
-void ISelfController::SetRestartMessageEnabled(Kernel::HLERequestContext& ctx) {
-    LOG_WARNING(Service_AM, "(STUBBED) called");
+void ISelfController::EnterFatalSection(Kernel::HLERequestContext& ctx) {
+    ++num_fatal_sections_entered;
+    LOG_DEBUG(Service_AM, "called. Num fatal sections entered: {}", num_fatal_sections_entered);
 
     IPC::ResponseBuilder rb{ctx, 2};
     rb.Push(RESULT_SUCCESS);
 }
 
-void ISelfController::SetPerformanceModeChangedNotification(Kernel::HLERequestContext& ctx) {
-    IPC::RequestParser rp{ctx};
+void ISelfController::LeaveFatalSection(Kernel::HLERequestContext& ctx) {
+    LOG_DEBUG(Service_AM, "called.");
 
-    bool flag = rp.Pop<bool>();
-    LOG_WARNING(Service_AM, "(STUBBED) called flag={}", flag);
+    // Entry and exit of fatal sections must be balanced.
+    if (num_fatal_sections_entered == 0) {
+        IPC::ResponseBuilder rb{ctx, 2};
+        rb.Push(ResultCode{ErrorModule::AM, 512});
+        return;
+    }
+
+    --num_fatal_sections_entered;
 
     IPC::ResponseBuilder rb{ctx, 2};
     rb.Push(RESULT_SUCCESS);
 }
 
+void ISelfController::GetLibraryAppletLaunchableEvent(Kernel::HLERequestContext& ctx) {
+    LOG_WARNING(Service_AM, "(STUBBED) called");
+
+    launchable_event.writable->Signal();
+
+    IPC::ResponseBuilder rb{ctx, 2, 1};
+    rb.Push(RESULT_SUCCESS);
+    rb.PushCopyObjects(launchable_event.readable);
+}
+
 void ISelfController::SetScreenShotPermission(Kernel::HLERequestContext& ctx) {
     LOG_WARNING(Service_AM, "(STUBBED) called");
 
@@ -337,40 +350,52 @@ void ISelfController::SetOperationModeChangedNotification(Kernel::HLERequestCont
     rb.Push(RESULT_SUCCESS);
 }
 
-void ISelfController::SetOutOfFocusSuspendingEnabled(Kernel::HLERequestContext& ctx) {
-    // Takes 3 input u8s with each field located immediately after the previous
-    // u8, these are bool flags. No output.
+void ISelfController::SetPerformanceModeChangedNotification(Kernel::HLERequestContext& ctx) {
     IPC::RequestParser rp{ctx};
 
-    bool enabled = rp.Pop<bool>();
-    LOG_WARNING(Service_AM, "(STUBBED) called enabled={}", enabled);
+    bool flag = rp.Pop<bool>();
+    LOG_WARNING(Service_AM, "(STUBBED) called flag={}", flag);
 
     IPC::ResponseBuilder rb{ctx, 2};
     rb.Push(RESULT_SUCCESS);
 }
 
-void ISelfController::LockExit(Kernel::HLERequestContext& ctx) {
-    LOG_WARNING(Service_AM, "(STUBBED) called");
+void ISelfController::SetFocusHandlingMode(Kernel::HLERequestContext& ctx) {
+    // Takes 3 input u8s with each field located immediately after the previous
+    // u8, these are bool flags. No output.
+    IPC::RequestParser rp{ctx};
+
+    struct FocusHandlingModeParams {
+        u8 unknown0;
+        u8 unknown1;
+        u8 unknown2;
+    };
+    const auto flags = rp.PopRaw<FocusHandlingModeParams>();
+
+    LOG_WARNING(Service_AM, "(STUBBED) called. unknown0={}, unknown1={}, unknown2={}",
+                flags.unknown0, flags.unknown1, flags.unknown2);
 
     IPC::ResponseBuilder rb{ctx, 2};
     rb.Push(RESULT_SUCCESS);
 }
 
-void ISelfController::UnlockExit(Kernel::HLERequestContext& ctx) {
+void ISelfController::SetRestartMessageEnabled(Kernel::HLERequestContext& ctx) {
     LOG_WARNING(Service_AM, "(STUBBED) called");
 
     IPC::ResponseBuilder rb{ctx, 2};
     rb.Push(RESULT_SUCCESS);
 }
 
-void ISelfController::GetLibraryAppletLaunchableEvent(Kernel::HLERequestContext& ctx) {
-    LOG_WARNING(Service_AM, "(STUBBED) called");
+void ISelfController::SetOutOfFocusSuspendingEnabled(Kernel::HLERequestContext& ctx) {
+    // Takes 3 input u8s with each field located immediately after the previous
+    // u8, these are bool flags. No output.
+    IPC::RequestParser rp{ctx};
 
-    launchable_event.writable->Signal();
+    bool enabled = rp.Pop<bool>();
+    LOG_WARNING(Service_AM, "(STUBBED) called enabled={}", enabled);
 
-    IPC::ResponseBuilder rb{ctx, 2, 1};
+    IPC::ResponseBuilder rb{ctx, 2};
     rb.Push(RESULT_SUCCESS);
-    rb.PushCopyObjects(launchable_event.readable);
 }
 
 void ISelfController::SetScreenShotImageOrientation(Kernel::HLERequestContext& ctx) {
@@ -907,19 +932,19 @@ void ILibraryAppletCreator::CreateTransferMemoryStorage(Kernel::HLERequestContex
     rp.SetCurrentOffset(3);
     const auto handle{rp.Pop<Kernel::Handle>()};
 
-    const auto shared_mem =
-        Core::System::GetInstance().CurrentProcess()->GetHandleTable().Get<Kernel::SharedMemory>(
+    const auto transfer_mem =
+        Core::System::GetInstance().CurrentProcess()->GetHandleTable().Get<Kernel::TransferMemory>(
             handle);
 
-    if (shared_mem == nullptr) {
+    if (transfer_mem == nullptr) {
         LOG_ERROR(Service_AM, "shared_mem is a nullpr for handle={:08X}", handle);
         IPC::ResponseBuilder rb{ctx, 2};
         rb.Push(ResultCode(-1));
         return;
     }
 
-    const u8* mem_begin = shared_mem->GetPointer();
-    const u8* mem_end = mem_begin + shared_mem->GetSize();
+    const u8* const mem_begin = transfer_mem->GetPointer();
+    const u8* const mem_end = mem_begin + transfer_mem->GetSize();
     std::vector<u8> memory{mem_begin, mem_end};
 
     IPC::ResponseBuilder rb{ctx, 2, 0, 1};
diff --git a/src/core/hle/service/am/am.h b/src/core/hle/service/am/am.h
index 565dd8e9e..991b7d47c 100644
--- a/src/core/hle/service/am/am.h
+++ b/src/core/hle/service/am/am.h
@@ -117,17 +117,19 @@ public:
     ~ISelfController() override;
 
 private:
-    void SetFocusHandlingMode(Kernel::HLERequestContext& ctx);
-    void SetRestartMessageEnabled(Kernel::HLERequestContext& ctx);
-    void SetPerformanceModeChangedNotification(Kernel::HLERequestContext& ctx);
-    void SetOperationModeChangedNotification(Kernel::HLERequestContext& ctx);
-    void SetOutOfFocusSuspendingEnabled(Kernel::HLERequestContext& ctx);
     void LockExit(Kernel::HLERequestContext& ctx);
     void UnlockExit(Kernel::HLERequestContext& ctx);
+    void EnterFatalSection(Kernel::HLERequestContext& ctx);
+    void LeaveFatalSection(Kernel::HLERequestContext& ctx);
     void GetLibraryAppletLaunchableEvent(Kernel::HLERequestContext& ctx);
+    void SetScreenShotPermission(Kernel::HLERequestContext& ctx);
+    void SetOperationModeChangedNotification(Kernel::HLERequestContext& ctx);
+    void SetPerformanceModeChangedNotification(Kernel::HLERequestContext& ctx);
+    void SetFocusHandlingMode(Kernel::HLERequestContext& ctx);
+    void SetRestartMessageEnabled(Kernel::HLERequestContext& ctx);
+    void SetOutOfFocusSuspendingEnabled(Kernel::HLERequestContext& ctx);
     void SetScreenShotImageOrientation(Kernel::HLERequestContext& ctx);
     void CreateManagedDisplayLayer(Kernel::HLERequestContext& ctx);
-    void SetScreenShotPermission(Kernel::HLERequestContext& ctx);
     void SetHandlesRequestToDisplay(Kernel::HLERequestContext& ctx);
     void SetIdleTimeDetectionExtension(Kernel::HLERequestContext& ctx);
     void GetIdleTimeDetectionExtension(Kernel::HLERequestContext& ctx);
@@ -135,6 +137,7 @@ private:
     std::shared_ptr<NVFlinger::NVFlinger> nvflinger;
     Kernel::EventPair launchable_event;
     u32 idle_time_detection_extension = 0;
+    u64 num_fatal_sections_entered = 0;
 };
 
 class ICommonStateGetter final : public ServiceFramework<ICommonStateGetter> {
diff --git a/src/core/hle/service/audio/audout_u.cpp b/src/core/hle/service/audio/audout_u.cpp
index 21f5e64c7..39acb7b23 100644
--- a/src/core/hle/service/audio/audout_u.cpp
+++ b/src/core/hle/service/audio/audout_u.cpp
@@ -150,7 +150,6 @@ private:
     void GetReleasedAudioOutBufferImpl(Kernel::HLERequestContext& ctx) {
         LOG_DEBUG(Service_Audio, "called {}", ctx.Description());
 
-        IPC::RequestParser rp{ctx};
         const u64 max_count{ctx.GetWriteBufferSize() / sizeof(u64)};
         const auto released_buffers{audio_core.GetTagsAndReleaseBuffers(stream, max_count)};
 
@@ -194,12 +193,9 @@ private:
 void AudOutU::ListAudioOutsImpl(Kernel::HLERequestContext& ctx) {
     LOG_DEBUG(Service_Audio, "called");
 
-    IPC::RequestParser rp{ctx};
-
     ctx.WriteBuffer(DefaultDevice);
 
     IPC::ResponseBuilder rb{ctx, 3};
-
     rb.Push(RESULT_SUCCESS);
     rb.Push<u32>(1); // Amount of audio devices
 }
diff --git a/src/core/hle/service/audio/audren_u.cpp b/src/core/hle/service/audio/audren_u.cpp
index c9de10a24..1dde6edb7 100644
--- a/src/core/hle/service/audio/audren_u.cpp
+++ b/src/core/hle/service/audio/audren_u.cpp
@@ -10,6 +10,7 @@
 #include "common/alignment.h"
 #include "common/common_funcs.h"
 #include "common/logging/log.h"
+#include "common/string_util.h"
 #include "core/core.h"
 #include "core/hle/ipc_helpers.h"
 #include "core/hle/kernel/hle_ipc.h"
@@ -184,7 +185,6 @@ public:
 private:
     void ListAudioDeviceName(Kernel::HLERequestContext& ctx) {
         LOG_WARNING(Service_Audio, "(STUBBED) called");
-        IPC::RequestParser rp{ctx};
 
         constexpr std::array<char, 15> audio_interface{{"AudioInterface"}};
         ctx.WriteBuffer(audio_interface);
@@ -195,13 +195,13 @@ private:
     }
 
     void SetAudioDeviceOutputVolume(Kernel::HLERequestContext& ctx) {
-        LOG_WARNING(Service_Audio, "(STUBBED) called");
-
         IPC::RequestParser rp{ctx};
-        f32 volume = static_cast<f32>(rp.Pop<u32>());
+        const f32 volume = rp.Pop<f32>();
 
-        auto file_buffer = ctx.ReadBuffer();
-        auto end = std::find(file_buffer.begin(), file_buffer.end(), '\0');
+        const auto device_name_buffer = ctx.ReadBuffer();
+        const std::string name = Common::StringFromBuffer(device_name_buffer);
+
+        LOG_WARNING(Service_Audio, "(STUBBED) called. name={}, volume={}", name, volume);
 
         IPC::ResponseBuilder rb{ctx, 2};
         rb.Push(RESULT_SUCCESS);
@@ -209,7 +209,6 @@ private:
 
     void GetActiveAudioDeviceName(Kernel::HLERequestContext& ctx) {
         LOG_WARNING(Service_Audio, "(STUBBED) called");
-        IPC::RequestParser rp{ctx};
 
         constexpr std::array<char, 12> audio_interface{{"AudioDevice"}};
         ctx.WriteBuffer(audio_interface);
diff --git a/src/core/hle/service/fatal/fatal.cpp b/src/core/hle/service/fatal/fatal.cpp
index 770590d0b..2c229bcad 100644
--- a/src/core/hle/service/fatal/fatal.cpp
+++ b/src/core/hle/service/fatal/fatal.cpp
@@ -25,21 +25,34 @@ Module::Interface::Interface(std::shared_ptr<Module> module, const char* name)
 Module::Interface::~Interface() = default;
 
 struct FatalInfo {
-    std::array<u64_le, 31> registers{}; // TODO(ogniK): See if this actually is registers or
-                                        // not(find a game which has non zero valeus)
-    u64_le unk0{};
-    u64_le unk1{};
-    u64_le unk2{};
-    u64_le unk3{};
-    u64_le unk4{};
-    u64_le unk5{};
-    u64_le unk6{};
+    enum class Architecture : s32 {
+        AArch64,
+        AArch32,
+    };
+
+    const char* ArchAsString() const {
+        return arch == Architecture::AArch64 ? "AArch64" : "AArch32";
+    }
+
+    std::array<u64_le, 31> registers{};
+    u64_le sp{};
+    u64_le pc{};
+    u64_le pstate{};
+    u64_le afsr0{};
+    u64_le afsr1{};
+    u64_le esr{};
+    u64_le far{};
 
     std::array<u64_le, 32> backtrace{};
-    u64_le unk7{};
-    u64_le unk8{};
+    u64_le program_entry_point{};
+
+    // Bit flags that indicate which registers have been set with values
+    // for this context. The service itself uses these to determine which
+    // registers to specifically print out.
+    u64_le set_flags{};
+
     u32_le backtrace_size{};
-    u32_le unk9{};
+    Architecture arch{};
     u32_le unk10{}; // TODO(ogniK): Is this even used or is it just padding?
 };
 static_assert(sizeof(FatalInfo) == 0x250, "FatalInfo is an invalid size");
@@ -52,36 +65,36 @@ enum class FatalType : u32 {
 
 static void GenerateErrorReport(ResultCode error_code, const FatalInfo& info) {
     const auto title_id = Core::CurrentProcess()->GetTitleID();
-    std::string crash_report =
-        fmt::format("Yuzu {}-{} crash report\n"
-                    "Title ID:                        {:016x}\n"
-                    "Result:                          0x{:X} ({:04}-{:04d})\n"
-                    "\n",
-                    Common::g_scm_branch, Common::g_scm_desc, title_id, error_code.raw,
-                    2000 + static_cast<u32>(error_code.module.Value()),
-                    static_cast<u32>(error_code.description.Value()), info.unk8, info.unk7);
+    std::string crash_report = fmt::format(
+        "Yuzu {}-{} crash report\n"
+        "Title ID:                        {:016x}\n"
+        "Result:                          0x{:X} ({:04}-{:04d})\n"
+        "Set flags:                       0x{:16X}\n"
+        "Program entry point:             0x{:16X}\n"
+        "\n",
+        Common::g_scm_branch, Common::g_scm_desc, title_id, error_code.raw,
+        2000 + static_cast<u32>(error_code.module.Value()),
+        static_cast<u32>(error_code.description.Value()), info.set_flags, info.program_entry_point);
     if (info.backtrace_size != 0x0) {
         crash_report += "Registers:\n";
-        // TODO(ogniK): This is just a guess, find a game which actually has non zero values
         for (size_t i = 0; i < info.registers.size(); i++) {
             crash_report +=
                 fmt::format("    X[{:02d}]:                       {:016x}\n", i, info.registers[i]);
         }
-        crash_report += fmt::format("    Unknown 0:                   {:016x}\n", info.unk0);
-        crash_report += fmt::format("    Unknown 1:                   {:016x}\n", info.unk1);
-        crash_report += fmt::format("    Unknown 2:                   {:016x}\n", info.unk2);
-        crash_report += fmt::format("    Unknown 3:                   {:016x}\n", info.unk3);
-        crash_report += fmt::format("    Unknown 4:                   {:016x}\n", info.unk4);
-        crash_report += fmt::format("    Unknown 5:                   {:016x}\n", info.unk5);
-        crash_report += fmt::format("    Unknown 6:                   {:016x}\n", info.unk6);
+        crash_report += fmt::format("    SP:                          {:016x}\n", info.sp);
+        crash_report += fmt::format("    PC:                          {:016x}\n", info.pc);
+        crash_report += fmt::format("    PSTATE:                      {:016x}\n", info.pstate);
+        crash_report += fmt::format("    AFSR0:                       {:016x}\n", info.afsr0);
+        crash_report += fmt::format("    AFSR1:                       {:016x}\n", info.afsr1);
+        crash_report += fmt::format("    ESR:                         {:016x}\n", info.esr);
+        crash_report += fmt::format("    FAR:                         {:016x}\n", info.far);
         crash_report += "\nBacktrace:\n";
         for (size_t i = 0; i < info.backtrace_size; i++) {
             crash_report +=
                 fmt::format("    Backtrace[{:02d}]:               {:016x}\n", i, info.backtrace[i]);
         }
-        crash_report += fmt::format("\nUnknown 7:                       0x{:016x}\n", info.unk7);
-        crash_report += fmt::format("Unknown 8:                       0x{:016x}\n", info.unk8);
-        crash_report += fmt::format("Unknown 9:                       0x{:016x}\n", info.unk9);
+
+        crash_report += fmt::format("Architecture:                    {}\n", info.ArchAsString());
         crash_report += fmt::format("Unknown 10:                      0x{:016x}\n", info.unk10);
     }
 
@@ -125,13 +138,13 @@ static void ThrowFatalError(ResultCode error_code, FatalType fatal_type, const F
     case FatalType::ErrorReport:
         GenerateErrorReport(error_code, info);
         break;
-    };
+    }
 }
 
 void Module::Interface::ThrowFatal(Kernel::HLERequestContext& ctx) {
     LOG_ERROR(Service_Fatal, "called");
     IPC::RequestParser rp{ctx};
-    auto error_code = rp.Pop<ResultCode>();
+    const auto error_code = rp.Pop<ResultCode>();
 
     ThrowFatalError(error_code, FatalType::ErrorScreen, {});
     IPC::ResponseBuilder rb{ctx, 2};
@@ -141,8 +154,8 @@ void Module::Interface::ThrowFatal(Kernel::HLERequestContext& ctx) {
 void Module::Interface::ThrowFatalWithPolicy(Kernel::HLERequestContext& ctx) {
     LOG_ERROR(Service_Fatal, "called");
     IPC::RequestParser rp(ctx);
-    auto error_code = rp.Pop<ResultCode>();
-    auto fatal_type = rp.PopEnum<FatalType>();
+    const auto error_code = rp.Pop<ResultCode>();
+    const auto fatal_type = rp.PopEnum<FatalType>();
 
     ThrowFatalError(error_code, fatal_type, {}); // No info is passed with ThrowFatalWithPolicy
     IPC::ResponseBuilder rb{ctx, 2};
@@ -152,9 +165,9 @@ void Module::Interface::ThrowFatalWithPolicy(Kernel::HLERequestContext& ctx) {
 void Module::Interface::ThrowFatalWithCpuContext(Kernel::HLERequestContext& ctx) {
     LOG_ERROR(Service_Fatal, "called");
     IPC::RequestParser rp(ctx);
-    auto error_code = rp.Pop<ResultCode>();
-    auto fatal_type = rp.PopEnum<FatalType>();
-    auto fatal_info = ctx.ReadBuffer();
+    const auto error_code = rp.Pop<ResultCode>();
+    const auto fatal_type = rp.PopEnum<FatalType>();
+    const auto fatal_info = ctx.ReadBuffer();
     FatalInfo info{};
 
     ASSERT_MSG(fatal_info.size() == sizeof(FatalInfo), "Invalid fatal info buffer size!");
diff --git a/src/core/hle/service/filesystem/fsp_srv.cpp b/src/core/hle/service/filesystem/fsp_srv.cpp
index f03fb629c..592dce31a 100644
--- a/src/core/hle/service/filesystem/fsp_srv.cpp
+++ b/src/core/hle/service/filesystem/fsp_srv.cpp
@@ -315,61 +315,53 @@ public:
     void CreateFile(Kernel::HLERequestContext& ctx) {
         IPC::RequestParser rp{ctx};
 
-        auto file_buffer = ctx.ReadBuffer();
-        std::string name = Common::StringFromBuffer(file_buffer);
+        const auto file_buffer = ctx.ReadBuffer();
+        const std::string name = Common::StringFromBuffer(file_buffer);
 
-        u64 mode = rp.Pop<u64>();
-        u32 size = rp.Pop<u32>();
+        const u64 mode = rp.Pop<u64>();
+        const u32 size = rp.Pop<u32>();
 
-        LOG_DEBUG(Service_FS, "called file {} mode 0x{:X} size 0x{:08X}", name, mode, size);
+        LOG_DEBUG(Service_FS, "called. file={}, mode=0x{:X}, size=0x{:08X}", name, mode, size);
 
         IPC::ResponseBuilder rb{ctx, 2};
         rb.Push(backend.CreateFile(name, size));
     }
 
     void DeleteFile(Kernel::HLERequestContext& ctx) {
-        IPC::RequestParser rp{ctx};
-
-        auto file_buffer = ctx.ReadBuffer();
-        std::string name = Common::StringFromBuffer(file_buffer);
+        const auto file_buffer = ctx.ReadBuffer();
+        const std::string name = Common::StringFromBuffer(file_buffer);
 
-        LOG_DEBUG(Service_FS, "called file {}", name);
+        LOG_DEBUG(Service_FS, "called. file={}", name);
 
         IPC::ResponseBuilder rb{ctx, 2};
         rb.Push(backend.DeleteFile(name));
     }
 
     void CreateDirectory(Kernel::HLERequestContext& ctx) {
-        IPC::RequestParser rp{ctx};
-
-        auto file_buffer = ctx.ReadBuffer();
-        std::string name = Common::StringFromBuffer(file_buffer);
+        const auto file_buffer = ctx.ReadBuffer();
+        const std::string name = Common::StringFromBuffer(file_buffer);
 
-        LOG_DEBUG(Service_FS, "called directory {}", name);
+        LOG_DEBUG(Service_FS, "called. directory={}", name);
 
         IPC::ResponseBuilder rb{ctx, 2};
         rb.Push(backend.CreateDirectory(name));
     }
 
     void DeleteDirectory(Kernel::HLERequestContext& ctx) {
-        const IPC::RequestParser rp{ctx};
-
         const auto file_buffer = ctx.ReadBuffer();
-        std::string name = Common::StringFromBuffer(file_buffer);
+        const std::string name = Common::StringFromBuffer(file_buffer);
 
-        LOG_DEBUG(Service_FS, "called directory {}", name);
+        LOG_DEBUG(Service_FS, "called. directory={}", name);
 
         IPC::ResponseBuilder rb{ctx, 2};
         rb.Push(backend.DeleteDirectory(name));
     }
 
     void DeleteDirectoryRecursively(Kernel::HLERequestContext& ctx) {
-        const IPC::RequestParser rp{ctx};
-
         const auto file_buffer = ctx.ReadBuffer();
-        std::string name = Common::StringFromBuffer(file_buffer);
+        const std::string name = Common::StringFromBuffer(file_buffer);
 
-        LOG_DEBUG(Service_FS, "called directory {}", name);
+        LOG_DEBUG(Service_FS, "called. directory={}", name);
 
         IPC::ResponseBuilder rb{ctx, 2};
         rb.Push(backend.DeleteDirectoryRecursively(name));
@@ -386,18 +378,16 @@ public:
     }
 
     void RenameFile(Kernel::HLERequestContext& ctx) {
-        IPC::RequestParser rp{ctx};
-
         std::vector<u8> buffer;
         buffer.resize(ctx.BufferDescriptorX()[0].Size());
         Memory::ReadBlock(ctx.BufferDescriptorX()[0].Address(), buffer.data(), buffer.size());
-        std::string src_name = Common::StringFromBuffer(buffer);
+        const std::string src_name = Common::StringFromBuffer(buffer);
 
         buffer.resize(ctx.BufferDescriptorX()[1].Size());
         Memory::ReadBlock(ctx.BufferDescriptorX()[1].Address(), buffer.data(), buffer.size());
-        std::string dst_name = Common::StringFromBuffer(buffer);
+        const std::string dst_name = Common::StringFromBuffer(buffer);
 
-        LOG_DEBUG(Service_FS, "called file '{}' to file '{}'", src_name, dst_name);
+        LOG_DEBUG(Service_FS, "called. file '{}' to file '{}'", src_name, dst_name);
 
         IPC::ResponseBuilder rb{ctx, 2};
         rb.Push(backend.RenameFile(src_name, dst_name));
@@ -406,12 +396,12 @@ public:
     void OpenFile(Kernel::HLERequestContext& ctx) {
         IPC::RequestParser rp{ctx};
 
-        auto file_buffer = ctx.ReadBuffer();
-        std::string name = Common::StringFromBuffer(file_buffer);
+        const auto file_buffer = ctx.ReadBuffer();
+        const std::string name = Common::StringFromBuffer(file_buffer);
 
-        auto mode = static_cast<FileSys::Mode>(rp.Pop<u32>());
+        const auto mode = static_cast<FileSys::Mode>(rp.Pop<u32>());
 
-        LOG_DEBUG(Service_FS, "called file {} mode {}", name, static_cast<u32>(mode));
+        LOG_DEBUG(Service_FS, "called. file={}, mode={}", name, static_cast<u32>(mode));
 
         auto result = backend.OpenFile(name, mode);
         if (result.Failed()) {
@@ -430,13 +420,13 @@ public:
     void OpenDirectory(Kernel::HLERequestContext& ctx) {
         IPC::RequestParser rp{ctx};
 
-        auto file_buffer = ctx.ReadBuffer();
-        std::string name = Common::StringFromBuffer(file_buffer);
+        const auto file_buffer = ctx.ReadBuffer();
+        const std::string name = Common::StringFromBuffer(file_buffer);
 
         // TODO(Subv): Implement this filter.
-        u32 filter_flags = rp.Pop<u32>();
+        const u32 filter_flags = rp.Pop<u32>();
 
-        LOG_DEBUG(Service_FS, "called directory {} filter {}", name, filter_flags);
+        LOG_DEBUG(Service_FS, "called. directory={}, filter={}", name, filter_flags);
 
         auto result = backend.OpenDirectory(name);
         if (result.Failed()) {
@@ -453,12 +443,10 @@ public:
     }
 
     void GetEntryType(Kernel::HLERequestContext& ctx) {
-        IPC::RequestParser rp{ctx};
-
-        auto file_buffer = ctx.ReadBuffer();
-        std::string name = Common::StringFromBuffer(file_buffer);
+        const auto file_buffer = ctx.ReadBuffer();
+        const std::string name = Common::StringFromBuffer(file_buffer);
 
-        LOG_DEBUG(Service_FS, "called file {}", name);
+        LOG_DEBUG(Service_FS, "called. file={}", name);
 
         auto result = backend.GetEntryType(name);
         if (result.Failed()) {
diff --git a/src/core/hle/service/nfc/nfc.cpp b/src/core/hle/service/nfc/nfc.cpp
index 5c62d42ba..ca88bf97f 100644
--- a/src/core/hle/service/nfc/nfc.cpp
+++ b/src/core/hle/service/nfc/nfc.cpp
@@ -150,7 +150,7 @@ private:
 
         IPC::ResponseBuilder rb{ctx, 3};
         rb.Push(RESULT_SUCCESS);
-        rb.PushRaw<u8>(Settings::values.enable_nfc);
+        rb.PushRaw<u8>(true);
     }
 
     void GetStateOld(Kernel::HLERequestContext& ctx) {
diff --git a/src/core/hle/service/nfp/nfp.cpp b/src/core/hle/service/nfp/nfp.cpp
index 1c4482e47..c6babdd4d 100644
--- a/src/core/hle/service/nfp/nfp.cpp
+++ b/src/core/hle/service/nfp/nfp.cpp
@@ -335,7 +335,7 @@ void Module::Interface::CreateUserInterface(Kernel::HLERequestContext& ctx) {
 }
 
 bool Module::Interface::LoadAmiibo(const std::vector<u8>& buffer) {
-    std::lock_guard<std::recursive_mutex> lock(HLE::g_hle_lock);
+    std::lock_guard lock{HLE::g_hle_lock};
     if (buffer.size() < sizeof(AmiiboFile)) {
         return false;
     }
diff --git a/src/core/hle/service/sockets/sfdnsres.cpp b/src/core/hle/service/sockets/sfdnsres.cpp
index 13ab1d31e..852e71e4b 100644
--- a/src/core/hle/service/sockets/sfdnsres.cpp
+++ b/src/core/hle/service/sockets/sfdnsres.cpp
@@ -8,12 +8,20 @@
 namespace Service::Sockets {
 
 void SFDNSRES::GetAddrInfo(Kernel::HLERequestContext& ctx) {
+    struct Parameters {
+        u8 use_nsd_resolve;
+        u32 unknown;
+        u64 process_id;
+    };
+
     IPC::RequestParser rp{ctx};
+    const auto parameters = rp.PopRaw<Parameters>();
 
-    LOG_WARNING(Service, "(STUBBED) called");
+    LOG_WARNING(Service,
+                "(STUBBED) called. use_nsd_resolve={}, unknown=0x{:08X}, process_id=0x{:016X}",
+                parameters.use_nsd_resolve, parameters.unknown, parameters.process_id);
 
     IPC::ResponseBuilder rb{ctx, 2};
-
     rb.Push(RESULT_SUCCESS);
 }
 
diff --git a/src/core/hle/service/spl/module.cpp b/src/core/hle/service/spl/module.cpp
index 8db0c2f13..e724d4ab8 100644
--- a/src/core/hle/service/spl/module.cpp
+++ b/src/core/hle/service/spl/module.cpp
@@ -26,9 +26,7 @@ Module::Interface::~Interface() = default;
 void Module::Interface::GetRandomBytes(Kernel::HLERequestContext& ctx) {
     LOG_DEBUG(Service_SPL, "called");
 
-    IPC::RequestParser rp{ctx};
-
-    std::size_t size = ctx.GetWriteBufferSize();
+    const std::size_t size = ctx.GetWriteBufferSize();
 
     std::uniform_int_distribution<u16> distribution(0, std::numeric_limits<u8>::max());
     std::vector<u8> data(size);
diff --git a/src/core/hle/service/ssl/ssl.cpp b/src/core/hle/service/ssl/ssl.cpp
index af40a1815..f671f355e 100644
--- a/src/core/hle/service/ssl/ssl.cpp
+++ b/src/core/hle/service/ssl/ssl.cpp
@@ -68,9 +68,16 @@ public:
 
 private:
     void SetOption(Kernel::HLERequestContext& ctx) {
-        LOG_WARNING(Service_SSL, "(STUBBED) called");
+        struct Parameters {
+            u8 enable;
+            u32 option;
+        };
 
         IPC::RequestParser rp{ctx};
+        const auto parameters = rp.PopRaw<Parameters>();
+
+        LOG_WARNING(Service_SSL, "(STUBBED) called. enable={}, option={}", parameters.enable,
+                    parameters.option);
 
         IPC::ResponseBuilder rb{ctx, 2};
         rb.Push(RESULT_SUCCESS);
diff --git a/src/core/hle/service/vi/vi.cpp b/src/core/hle/service/vi/vi.cpp
index 566cd6006..b77cb495d 100644
--- a/src/core/hle/service/vi/vi.cpp
+++ b/src/core/hle/service/vi/vi.cpp
@@ -1037,7 +1037,6 @@ private:
     void ListDisplays(Kernel::HLERequestContext& ctx) {
         LOG_WARNING(Service_VI, "(STUBBED) called");
 
-        IPC::RequestParser rp{ctx};
         DisplayInfo display_info;
         display_info.width *= static_cast<u64>(Settings::values.resolution_factor);
         display_info.height *= static_cast<u64>(Settings::values.resolution_factor);
diff --git a/src/core/loader/elf.cpp b/src/core/loader/elf.cpp
index 8b1920f22..46ac372f6 100644
--- a/src/core/loader/elf.cpp
+++ b/src/core/loader/elf.cpp
@@ -341,7 +341,7 @@ Kernel::CodeSet ElfReader::LoadInto(VAddr vaddr) {
     }
 
     codeset.entrypoint = base_addr + header->e_entry;
-    codeset.memory = std::make_shared<std::vector<u8>>(std::move(program_image));
+    codeset.memory = std::move(program_image);
 
     LOG_DEBUG(Loader, "Done loading.");
 
diff --git a/src/core/loader/nro.cpp b/src/core/loader/nro.cpp
index 5de02a94b..31e4a0c84 100644
--- a/src/core/loader/nro.cpp
+++ b/src/core/loader/nro.cpp
@@ -187,7 +187,7 @@ static bool LoadNroImpl(Kernel::Process& process, const std::vector<u8>& data,
     program_image.resize(static_cast<u32>(program_image.size()) + bss_size);
 
     // Load codeset for current process
-    codeset.memory = std::make_shared<std::vector<u8>>(std::move(program_image));
+    codeset.memory = std::move(program_image);
     process.LoadModule(std::move(codeset), load_base);
 
     // Register module with GDBStub
diff --git a/src/core/loader/nso.cpp b/src/core/loader/nso.cpp
index 714d85a59..ffe2eea8a 100644
--- a/src/core/loader/nso.cpp
+++ b/src/core/loader/nso.cpp
@@ -4,11 +4,12 @@
 
 #include <cinttypes>
 #include <vector>
-#include <lz4.h>
+
 #include "common/common_funcs.h"
 #include "common/file_util.h"
 #include "common/hex_util.h"
 #include "common/logging/log.h"
+#include "common/lz4_compression.h"
 #include "common/swap.h"
 #include "core/core.h"
 #include "core/file_sys/patch_manager.h"
@@ -35,15 +36,11 @@ static_assert(sizeof(MODHeader) == 0x1c, "MODHeader has incorrect size.");
 
 std::vector<u8> DecompressSegment(const std::vector<u8>& compressed_data,
                                   const NSOSegmentHeader& header) {
-    std::vector<u8> uncompressed_data(header.size);
-    const int bytes_uncompressed =
-        LZ4_decompress_safe(reinterpret_cast<const char*>(compressed_data.data()),
-                            reinterpret_cast<char*>(uncompressed_data.data()),
-                            static_cast<int>(compressed_data.size()), header.size);
+    const std::vector<u8> uncompressed_data =
+        Common::Compression::DecompressDataLZ4(compressed_data, header.size);
 
-    ASSERT_MSG(bytes_uncompressed == static_cast<int>(header.size) &&
-                   bytes_uncompressed == static_cast<int>(uncompressed_data.size()),
-               "{} != {} != {}", bytes_uncompressed, header.size, uncompressed_data.size());
+    ASSERT_MSG(uncompressed_data.size() == static_cast<int>(header.size), "{} != {}", header.size,
+               uncompressed_data.size());
 
     return uncompressed_data;
 }
@@ -161,7 +158,7 @@ std::optional<VAddr> AppLoader_NSO::LoadModule(Kernel::Process& process,
     }
 
     // Load codeset for current process
-    codeset.memory = std::make_shared<std::vector<u8>>(std::move(program_image));
+    codeset.memory = std::move(program_image);
     process.LoadModule(std::move(codeset), load_base);
 
     // Register module with GDBStub
diff --git a/src/core/perf_stats.cpp b/src/core/perf_stats.cpp
index c716a462b..4afd6c8a3 100644
--- a/src/core/perf_stats.cpp
+++ b/src/core/perf_stats.cpp
@@ -18,13 +18,13 @@ using std::chrono::microseconds;
 namespace Core {
 
 void PerfStats::BeginSystemFrame() {
-    std::lock_guard<std::mutex> lock(object_mutex);
+    std::lock_guard lock{object_mutex};
 
     frame_begin = Clock::now();
 }
 
 void PerfStats::EndSystemFrame() {
-    std::lock_guard<std::mutex> lock(object_mutex);
+    std::lock_guard lock{object_mutex};
 
     auto frame_end = Clock::now();
     accumulated_frametime += frame_end - frame_begin;
@@ -35,13 +35,13 @@ void PerfStats::EndSystemFrame() {
 }
 
 void PerfStats::EndGameFrame() {
-    std::lock_guard<std::mutex> lock(object_mutex);
+    std::lock_guard lock{object_mutex};
 
     game_frames += 1;
 }
 
 PerfStatsResults PerfStats::GetAndResetStats(microseconds current_system_time_us) {
-    std::lock_guard<std::mutex> lock(object_mutex);
+    std::lock_guard lock{object_mutex};
 
     const auto now = Clock::now();
     // Walltime elapsed since stats were reset
@@ -67,7 +67,7 @@ PerfStatsResults PerfStats::GetAndResetStats(microseconds current_system_time_us
 }
 
 double PerfStats::GetLastFrameTimeScale() {
-    std::lock_guard<std::mutex> lock(object_mutex);
+    std::lock_guard lock{object_mutex};
 
     constexpr double FRAME_LENGTH = 1.0 / 60;
     return duration_cast<DoubleSecs>(previous_frame_length).count() / FRAME_LENGTH;
diff --git a/src/core/settings.cpp b/src/core/settings.cpp
index 6dd3139cc..6d32ebea3 100644
--- a/src/core/settings.cpp
+++ b/src/core/settings.cpp
@@ -82,7 +82,6 @@ void LogSetting(const std::string& name, const T& value) {
 void LogSettings() {
     LOG_INFO(Config, "yuzu Configuration:");
     LogSetting("System_UseDockedMode", Settings::values.use_docked_mode);
-    LogSetting("System_EnableNfc", Settings::values.enable_nfc);
     LogSetting("System_RngSeed", Settings::values.rng_seed.value_or(0));
     LogSetting("System_CurrentUser", Settings::values.current_user);
     LogSetting("System_LanguageIndex", Settings::values.language_index);
diff --git a/src/core/settings.h b/src/core/settings.h
index cdfb2f742..d543eb32f 100644
--- a/src/core/settings.h
+++ b/src/core/settings.h
@@ -349,7 +349,6 @@ struct TouchscreenInput {
 struct Values {
     // System
     bool use_docked_mode;
-    bool enable_nfc;
     std::optional<u32> rng_seed;
     // Measured in seconds since epoch
     std::optional<std::chrono::seconds> custom_rtc;
diff --git a/src/input_common/keyboard.cpp b/src/input_common/keyboard.cpp
index 525fe6abc..078374be5 100644
--- a/src/input_common/keyboard.cpp
+++ b/src/input_common/keyboard.cpp
@@ -36,18 +36,18 @@ struct KeyButtonPair {
 class KeyButtonList {
 public:
     void AddKeyButton(int key_code, KeyButton* key_button) {
-        std::lock_guard<std::mutex> guard(mutex);
+        std::lock_guard guard{mutex};
         list.push_back(KeyButtonPair{key_code, key_button});
     }
 
     void RemoveKeyButton(const KeyButton* key_button) {
-        std::lock_guard<std::mutex> guard(mutex);
+        std::lock_guard guard{mutex};
         list.remove_if(
             [key_button](const KeyButtonPair& pair) { return pair.key_button == key_button; });
     }
 
     void ChangeKeyStatus(int key_code, bool pressed) {
-        std::lock_guard<std::mutex> guard(mutex);
+        std::lock_guard guard{mutex};
         for (const KeyButtonPair& pair : list) {
             if (pair.key_code == key_code)
                 pair.key_button->status.store(pressed);
@@ -55,7 +55,7 @@ public:
     }
 
     void ChangeAllKeyStatus(bool pressed) {
-        std::lock_guard<std::mutex> guard(mutex);
+        std::lock_guard guard{mutex};
         for (const KeyButtonPair& pair : list) {
             pair.key_button->status.store(pressed);
         }
diff --git a/src/input_common/motion_emu.cpp b/src/input_common/motion_emu.cpp
index 6d96d4019..868251628 100644
--- a/src/input_common/motion_emu.cpp
+++ b/src/input_common/motion_emu.cpp
@@ -39,7 +39,7 @@ public:
     void Tilt(int x, int y) {
         auto mouse_move = Common::MakeVec(x, y) - mouse_origin;
         if (is_tilting) {
-            std::lock_guard<std::mutex> guard(tilt_mutex);
+            std::lock_guard guard{tilt_mutex};
             if (mouse_move.x == 0 && mouse_move.y == 0) {
                 tilt_angle = 0;
             } else {
@@ -51,13 +51,13 @@ public:
     }
 
     void EndTilt() {
-        std::lock_guard<std::mutex> guard(tilt_mutex);
+        std::lock_guard guard{tilt_mutex};
         tilt_angle = 0;
         is_tilting = false;
     }
 
     std::tuple<Common::Vec3<float>, Common::Vec3<float>> GetStatus() {
-        std::lock_guard<std::mutex> guard(status_mutex);
+        std::lock_guard guard{status_mutex};
         return status;
     }
 
@@ -93,7 +93,7 @@ private:
             old_q = q;
 
             {
-                std::lock_guard<std::mutex> guard(tilt_mutex);
+                std::lock_guard guard{tilt_mutex};
 
                 // Find the quaternion describing current 3DS tilting
                 q = Common::MakeQuaternion(
@@ -115,7 +115,7 @@ private:
 
             // Update the sensor state
             {
-                std::lock_guard<std::mutex> guard(status_mutex);
+                std::lock_guard guard{status_mutex};
                 status = std::make_tuple(gravity, angular_rate);
             }
         }
diff --git a/src/input_common/sdl/sdl_impl.cpp b/src/input_common/sdl/sdl_impl.cpp
index b132d77f5..5949ecbae 100644
--- a/src/input_common/sdl/sdl_impl.cpp
+++ b/src/input_common/sdl/sdl_impl.cpp
@@ -55,22 +55,22 @@ public:
         : guid{std::move(guid_)}, port{port_}, sdl_joystick{joystick, deleter} {}
 
     void SetButton(int button, bool value) {
-        std::lock_guard<std::mutex> lock(mutex);
+        std::lock_guard lock{mutex};
         state.buttons[button] = value;
     }
 
     bool GetButton(int button) const {
-        std::lock_guard<std::mutex> lock(mutex);
+        std::lock_guard lock{mutex};
         return state.buttons.at(button);
     }
 
     void SetAxis(int axis, Sint16 value) {
-        std::lock_guard<std::mutex> lock(mutex);
+        std::lock_guard lock{mutex};
         state.axes[axis] = value;
     }
 
     float GetAxis(int axis) const {
-        std::lock_guard<std::mutex> lock(mutex);
+        std::lock_guard lock{mutex};
         return state.axes.at(axis) / 32767.0f;
     }
 
@@ -92,12 +92,12 @@ public:
     }
 
     void SetHat(int hat, Uint8 direction) {
-        std::lock_guard<std::mutex> lock(mutex);
+        std::lock_guard lock{mutex};
         state.hats[hat] = direction;
     }
 
     bool GetHatDirection(int hat, Uint8 direction) const {
-        std::lock_guard<std::mutex> lock(mutex);
+        std::lock_guard lock{mutex};
         return (state.hats.at(hat) & direction) != 0;
     }
     /**
@@ -140,7 +140,7 @@ private:
  * Get the nth joystick with the corresponding GUID
  */
 std::shared_ptr<SDLJoystick> SDLState::GetSDLJoystickByGUID(const std::string& guid, int port) {
-    std::lock_guard<std::mutex> lock(joystick_map_mutex);
+    std::lock_guard lock{joystick_map_mutex};
     const auto it = joystick_map.find(guid);
     if (it != joystick_map.end()) {
         while (it->second.size() <= port) {
@@ -161,7 +161,8 @@ std::shared_ptr<SDLJoystick> SDLState::GetSDLJoystickByGUID(const std::string& g
 std::shared_ptr<SDLJoystick> SDLState::GetSDLJoystickBySDLID(SDL_JoystickID sdl_id) {
     auto sdl_joystick = SDL_JoystickFromInstanceID(sdl_id);
     const std::string guid = GetGUID(sdl_joystick);
-    std::lock_guard<std::mutex> lock(joystick_map_mutex);
+
+    std::lock_guard lock{joystick_map_mutex};
     auto map_it = joystick_map.find(guid);
     if (map_it != joystick_map.end()) {
         auto vec_it = std::find_if(map_it->second.begin(), map_it->second.end(),
@@ -198,8 +199,9 @@ void SDLState::InitJoystick(int joystick_index) {
         LOG_ERROR(Input, "failed to open joystick {}", joystick_index);
         return;
     }
-    std::string guid = GetGUID(sdl_joystick);
-    std::lock_guard<std::mutex> lock(joystick_map_mutex);
+    const std::string guid = GetGUID(sdl_joystick);
+
+    std::lock_guard lock{joystick_map_mutex};
     if (joystick_map.find(guid) == joystick_map.end()) {
         auto joystick = std::make_shared<SDLJoystick>(guid, 0, sdl_joystick);
         joystick_map[guid].emplace_back(std::move(joystick));
@@ -221,7 +223,7 @@ void SDLState::CloseJoystick(SDL_Joystick* sdl_joystick) {
     std::string guid = GetGUID(sdl_joystick);
     std::shared_ptr<SDLJoystick> joystick;
     {
-        std::lock_guard<std::mutex> lock(joystick_map_mutex);
+        std::lock_guard lock{joystick_map_mutex};
         // This call to guid is safe since the joystick is guaranteed to be in the map
         auto& joystick_guid_list = joystick_map[guid];
         const auto joystick_it =
@@ -274,7 +276,7 @@ void SDLState::HandleGameControllerEvent(const SDL_Event& event) {
 }
 
 void SDLState::CloseJoysticks() {
-    std::lock_guard<std::mutex> lock(joystick_map_mutex);
+    std::lock_guard lock{joystick_map_mutex};
     joystick_map.clear();
 }
 
diff --git a/src/tests/CMakeLists.txt b/src/tests/CMakeLists.txt
index d0284bdf4..c7038b217 100644
--- a/src/tests/CMakeLists.txt
+++ b/src/tests/CMakeLists.txt
@@ -1,5 +1,7 @@
 add_executable(tests
     common/bit_field.cpp
+    common/bit_utils.cpp
+    common/multi_level_queue.cpp
     common/param_package.cpp
     common/ring_buffer.cpp
     core/arm/arm_test_common.cpp
diff --git a/src/tests/common/bit_utils.cpp b/src/tests/common/bit_utils.cpp
new file mode 100644
index 000000000..479b5995a
--- /dev/null
+++ b/src/tests/common/bit_utils.cpp
@@ -0,0 +1,23 @@
+// Copyright 2017 Citra Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#include <catch2/catch.hpp>
+#include <math.h>
+#include "common/bit_util.h"
+
+namespace Common {
+
+TEST_CASE("BitUtils::CountTrailingZeroes", "[common]") {
+    REQUIRE(Common::CountTrailingZeroes32(0) == 32);
+    REQUIRE(Common::CountTrailingZeroes64(0) == 64);
+    REQUIRE(Common::CountTrailingZeroes32(9) == 0);
+    REQUIRE(Common::CountTrailingZeroes32(8) == 3);
+    REQUIRE(Common::CountTrailingZeroes32(0x801000) == 12);
+    REQUIRE(Common::CountTrailingZeroes64(9) == 0);
+    REQUIRE(Common::CountTrailingZeroes64(8) == 3);
+    REQUIRE(Common::CountTrailingZeroes64(0x801000) == 12);
+    REQUIRE(Common::CountTrailingZeroes64(0x801000000000UL) == 36);
+}
+
+} // namespace Common
diff --git a/src/tests/common/multi_level_queue.cpp b/src/tests/common/multi_level_queue.cpp
new file mode 100644
index 000000000..cca7ec7da
--- /dev/null
+++ b/src/tests/common/multi_level_queue.cpp
@@ -0,0 +1,55 @@
+// Copyright 2019 Yuzu Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#include <catch2/catch.hpp>
+#include <math.h>
+#include "common/common_types.h"
+#include "common/multi_level_queue.h"
+
+namespace Common {
+
+TEST_CASE("MultiLevelQueue", "[common]") {
+    std::array<f32, 8> values = {0.0, 5.0, 1.0, 9.0, 8.0, 2.0, 6.0, 7.0};
+    Common::MultiLevelQueue<f32, 64> mlq;
+    REQUIRE(mlq.empty());
+    mlq.add(values[2], 2);
+    mlq.add(values[7], 7);
+    mlq.add(values[3], 3);
+    mlq.add(values[4], 4);
+    mlq.add(values[0], 0);
+    mlq.add(values[5], 5);
+    mlq.add(values[6], 6);
+    mlq.add(values[1], 1);
+    u32 index = 0;
+    bool all_set = true;
+    for (auto& f : mlq) {
+        all_set &= (f == values[index]);
+        index++;
+    }
+    REQUIRE(all_set);
+    REQUIRE(!mlq.empty());
+    f32 v = 8.0;
+    mlq.add(v, 2);
+    v = -7.0;
+    mlq.add(v, 2, false);
+    REQUIRE(mlq.front(2) == -7.0);
+    mlq.yield(2);
+    REQUIRE(mlq.front(2) == values[2]);
+    REQUIRE(mlq.back(2) == -7.0);
+    REQUIRE(mlq.empty(8));
+    v = 10.0;
+    mlq.add(v, 8);
+    mlq.adjust(v, 8, 9);
+    REQUIRE(mlq.front(9) == v);
+    REQUIRE(mlq.empty(8));
+    REQUIRE(!mlq.empty(9));
+    mlq.adjust(values[0], 0, 9);
+    REQUIRE(mlq.highest_priority_set() == 1);
+    REQUIRE(mlq.lowest_priority_set() == 9);
+    mlq.remove(values[1], 1);
+    REQUIRE(mlq.highest_priority_set() == 2);
+    REQUIRE(mlq.empty(1));
+}
+
+} // namespace Common
diff --git a/src/video_core/CMakeLists.txt b/src/video_core/CMakeLists.txt
index 14b76680f..242a0d1cd 100644
--- a/src/video_core/CMakeLists.txt
+++ b/src/video_core/CMakeLists.txt
@@ -128,7 +128,9 @@ if (ENABLE_VULKAN)
         renderer_vulkan/vk_scheduler.cpp
         renderer_vulkan/vk_scheduler.h
         renderer_vulkan/vk_stream_buffer.cpp
-        renderer_vulkan/vk_stream_buffer.h)
+        renderer_vulkan/vk_stream_buffer.h
+        renderer_vulkan/vk_swapchain.cpp
+        renderer_vulkan/vk_swapchain.h)
 
     target_include_directories(video_core PRIVATE ../../externals/Vulkan-Headers/include)
     target_compile_definitions(video_core PRIVATE HAS_VULKAN)
@@ -137,4 +139,4 @@ endif()
 create_target_directory_groups(video_core)
 
 target_link_libraries(video_core PUBLIC common core)
-target_link_libraries(video_core PRIVATE glad lz4_static)
+target_link_libraries(video_core PRIVATE glad)
diff --git a/src/video_core/debug_utils/debug_utils.cpp b/src/video_core/debug_utils/debug_utils.cpp
index 5ffb492ea..f0ef67535 100644
--- a/src/video_core/debug_utils/debug_utils.cpp
+++ b/src/video_core/debug_utils/debug_utils.cpp
@@ -10,7 +10,7 @@ namespace Tegra {
 
 void DebugContext::DoOnEvent(Event event, void* data) {
     {
-        std::unique_lock<std::mutex> lock(breakpoint_mutex);
+        std::unique_lock lock{breakpoint_mutex};
 
         // TODO(Subv): Commit the rasterizer's caches so framebuffers, render targets, etc. will
         // show on debug widgets
@@ -32,7 +32,7 @@ void DebugContext::DoOnEvent(Event event, void* data) {
 
 void DebugContext::Resume() {
     {
-        std::lock_guard<std::mutex> lock(breakpoint_mutex);
+        std::lock_guard lock{breakpoint_mutex};
 
         // Tell all observers that we are about to resume
         for (auto& breakpoint_observer : breakpoint_observers) {
diff --git a/src/video_core/debug_utils/debug_utils.h b/src/video_core/debug_utils/debug_utils.h
index c235faf46..ac3a2eb01 100644
--- a/src/video_core/debug_utils/debug_utils.h
+++ b/src/video_core/debug_utils/debug_utils.h
@@ -40,7 +40,7 @@ public:
         /// Constructs the object such that it observes events of the given DebugContext.
         explicit BreakPointObserver(std::shared_ptr<DebugContext> debug_context)
             : context_weak(debug_context) {
-            std::unique_lock<std::mutex> lock(debug_context->breakpoint_mutex);
+            std::unique_lock lock{debug_context->breakpoint_mutex};
             debug_context->breakpoint_observers.push_back(this);
         }
 
@@ -48,7 +48,7 @@ public:
             auto context = context_weak.lock();
             if (context) {
                 {
-                    std::unique_lock<std::mutex> lock(context->breakpoint_mutex);
+                    std::unique_lock lock{context->breakpoint_mutex};
                     context->breakpoint_observers.remove(this);
                 }
 
diff --git a/src/video_core/gpu.cpp b/src/video_core/gpu.cpp
index 267a03f2d..30b29e14d 100644
--- a/src/video_core/gpu.cpp
+++ b/src/video_core/gpu.cpp
@@ -286,9 +286,10 @@ void GPU::ProcessSemaphoreTriggerMethod() {
         // TODO(Kmather73): Generate a real GPU timestamp and write it here instead of
         // CoreTiming
         block.timestamp = Core::System::GetInstance().CoreTiming().GetTicks();
-        memory_manager->WriteBlock(regs.smaphore_address.SmaphoreAddress(), &block, sizeof(block));
+        memory_manager->WriteBlock(regs.semaphore_address.SemaphoreAddress(), &block,
+                                   sizeof(block));
     } else {
-        const u32 word{memory_manager->Read<u32>(regs.smaphore_address.SmaphoreAddress())};
+        const u32 word{memory_manager->Read<u32>(regs.semaphore_address.SemaphoreAddress())};
         if ((op == GpuSemaphoreOperation::AcquireEqual && word == regs.semaphore_sequence) ||
             (op == GpuSemaphoreOperation::AcquireGequal &&
              static_cast<s32>(word - regs.semaphore_sequence) > 0) ||
@@ -315,11 +316,11 @@ void GPU::ProcessSemaphoreTriggerMethod() {
 }
 
 void GPU::ProcessSemaphoreRelease() {
-    memory_manager->Write<u32>(regs.smaphore_address.SmaphoreAddress(), regs.semaphore_release);
+    memory_manager->Write<u32>(regs.semaphore_address.SemaphoreAddress(), regs.semaphore_release);
 }
 
 void GPU::ProcessSemaphoreAcquire() {
-    const u32 word = memory_manager->Read<u32>(regs.smaphore_address.SmaphoreAddress());
+    const u32 word = memory_manager->Read<u32>(regs.semaphore_address.SemaphoreAddress());
     const auto value = regs.semaphore_acquire;
     if (word != value) {
         regs.acquire_active = true;
diff --git a/src/video_core/gpu.h b/src/video_core/gpu.h
index c1830ac8d..de30ea354 100644
--- a/src/video_core/gpu.h
+++ b/src/video_core/gpu.h
@@ -177,11 +177,11 @@ public:
                     u32 address_high;
                     u32 address_low;
 
-                    GPUVAddr SmaphoreAddress() const {
+                    GPUVAddr SemaphoreAddress() const {
                         return static_cast<GPUVAddr>((static_cast<GPUVAddr>(address_high) << 32) |
                                                      address_low);
                     }
-                } smaphore_address;
+                } semaphore_address;
 
                 u32 semaphore_sequence;
                 u32 semaphore_trigger;
@@ -263,7 +263,7 @@ private:
     static_assert(offsetof(GPU::Regs, field_name) == position * 4,                                 \
                   "Field " #field_name " has invalid position")
 
-ASSERT_REG_POSITION(smaphore_address, 0x4);
+ASSERT_REG_POSITION(semaphore_address, 0x4);
 ASSERT_REG_POSITION(semaphore_sequence, 0x6);
 ASSERT_REG_POSITION(semaphore_trigger, 0x7);
 ASSERT_REG_POSITION(reference_count, 0x14);
diff --git a/src/video_core/gpu_thread.cpp b/src/video_core/gpu_thread.cpp
index 086b2f625..c5dc199c5 100644
--- a/src/video_core/gpu_thread.cpp
+++ b/src/video_core/gpu_thread.cpp
@@ -52,8 +52,8 @@ static void RunThread(VideoCore::RendererBase& renderer, Tegra::DmaPusher& dma_p
 }
 
 ThreadManager::ThreadManager(VideoCore::RendererBase& renderer, Tegra::DmaPusher& dma_pusher)
-    : renderer{renderer}, dma_pusher{dma_pusher}, thread{RunThread, std::ref(renderer),
-                                                         std::ref(dma_pusher), std::ref(state)} {}
+    : renderer{renderer}, thread{RunThread, std::ref(renderer), std::ref(dma_pusher),
+                                 std::ref(state)} {}
 
 ThreadManager::~ThreadManager() {
     // Notify GPU thread that a shutdown is pending
diff --git a/src/video_core/gpu_thread.h b/src/video_core/gpu_thread.h
index 8cd7db1c6..70acb2e79 100644
--- a/src/video_core/gpu_thread.h
+++ b/src/video_core/gpu_thread.h
@@ -4,10 +4,8 @@
 
 #pragma once
 
-#include <array>
 #include <atomic>
 #include <condition_variable>
-#include <memory>
 #include <mutex>
 #include <optional>
 #include <thread>
@@ -97,13 +95,13 @@ struct SynchState final {
     std::condition_variable frames_condition;
 
     void IncrementFramesCounter() {
-        std::lock_guard<std::mutex> lock{frames_mutex};
+        std::lock_guard lock{frames_mutex};
         ++queued_frame_count;
     }
 
     void DecrementFramesCounter() {
         {
-            std::lock_guard<std::mutex> lock{frames_mutex};
+            std::lock_guard lock{frames_mutex};
             --queued_frame_count;
 
             if (queued_frame_count) {
@@ -115,7 +113,7 @@ struct SynchState final {
 
     void WaitForFrames() {
         {
-            std::lock_guard<std::mutex> lock{frames_mutex};
+            std::lock_guard lock{frames_mutex};
             if (!queued_frame_count) {
                 return;
             }
@@ -123,14 +121,14 @@ struct SynchState final {
 
         // Wait for the GPU to be idle (all commands to be executed)
         {
-            std::unique_lock<std::mutex> lock{frames_mutex};
+            std::unique_lock lock{frames_mutex};
             frames_condition.wait(lock, [this] { return !queued_frame_count; });
         }
     }
 
     void SignalCommands() {
         {
-            std::unique_lock<std::mutex> lock{commands_mutex};
+            std::unique_lock lock{commands_mutex};
             if (queue.Empty()) {
                 return;
             }
@@ -140,7 +138,7 @@ struct SynchState final {
     }
 
     void WaitForCommands() {
-        std::unique_lock<std::mutex> lock{commands_mutex};
+        std::unique_lock lock{commands_mutex};
         commands_condition.wait(lock, [this] { return !queue.Empty(); });
     }
 
@@ -177,7 +175,6 @@ private:
 private:
     SynchState state;
     VideoCore::RendererBase& renderer;
-    Tegra::DmaPusher& dma_pusher;
     std::thread thread;
     std::thread::id thread_id;
 };
diff --git a/src/video_core/rasterizer_cache.h b/src/video_core/rasterizer_cache.h
index 9fc9f3056..291772186 100644
--- a/src/video_core/rasterizer_cache.h
+++ b/src/video_core/rasterizer_cache.h
@@ -71,8 +71,8 @@ private:
     bool is_registered{};      ///< Whether the object is currently registered with the cache
     bool is_dirty{};           ///< Whether the object is dirty (out of sync with guest memory)
     u64 last_modified_ticks{}; ///< When the object was last modified, used for in-order flushing
-    CacheAddr cache_addr{};    ///< Cache address memory, unique from emulated virtual address space
     const u8* host_ptr{};      ///< Pointer to the memory backing this cached region
+    CacheAddr cache_addr{};    ///< Cache address memory, unique from emulated virtual address space
 };
 
 template <class T>
@@ -84,7 +84,7 @@ public:
 
     /// Write any cached resources overlapping the specified region back to memory
     void FlushRegion(CacheAddr addr, std::size_t size) {
-        std::lock_guard<std::recursive_mutex> lock{mutex};
+        std::lock_guard lock{mutex};
 
         const auto& objects{GetSortedObjectsFromRegion(addr, size)};
         for (auto& object : objects) {
@@ -94,7 +94,7 @@ public:
 
     /// Mark the specified region as being invalidated
     void InvalidateRegion(CacheAddr addr, u64 size) {
-        std::lock_guard<std::recursive_mutex> lock{mutex};
+        std::lock_guard lock{mutex};
 
         const auto& objects{GetSortedObjectsFromRegion(addr, size)};
         for (auto& object : objects) {
@@ -108,7 +108,7 @@ public:
 
     /// Invalidates everything in the cache
     void InvalidateAll() {
-        std::lock_guard<std::recursive_mutex> lock{mutex};
+        std::lock_guard lock{mutex};
 
         while (interval_cache.begin() != interval_cache.end()) {
             Unregister(*interval_cache.begin()->second.begin());
@@ -133,7 +133,7 @@ protected:
 
     /// Register an object into the cache
     virtual void Register(const T& object) {
-        std::lock_guard<std::recursive_mutex> lock{mutex};
+        std::lock_guard lock{mutex};
 
         object->SetIsRegistered(true);
         interval_cache.add({GetInterval(object), ObjectSet{object}});
@@ -143,7 +143,7 @@ protected:
 
     /// Unregisters an object from the cache
     virtual void Unregister(const T& object) {
-        std::lock_guard<std::recursive_mutex> lock{mutex};
+        std::lock_guard lock{mutex};
 
         object->SetIsRegistered(false);
         rasterizer.UpdatePagesCachedCount(object->GetCpuAddr(), object->GetSizeInBytes(), -1);
@@ -153,14 +153,14 @@ protected:
 
     /// Returns a ticks counter used for tracking when cached objects were last modified
     u64 GetModifiedTicks() {
-        std::lock_guard<std::recursive_mutex> lock{mutex};
+        std::lock_guard lock{mutex};
 
         return ++modified_ticks;
     }
 
     /// Flushes the specified object, updating appropriate cache state as needed
     void FlushObject(const T& object) {
-        std::lock_guard<std::recursive_mutex> lock{mutex};
+        std::lock_guard lock{mutex};
 
         if (!object->IsDirty()) {
             return;
diff --git a/src/video_core/renderer_opengl/gl_buffer_cache.cpp b/src/video_core/renderer_opengl/gl_buffer_cache.cpp
index f75c65825..7989ec11b 100644
--- a/src/video_core/renderer_opengl/gl_buffer_cache.cpp
+++ b/src/video_core/renderer_opengl/gl_buffer_cache.cpp
@@ -7,7 +7,6 @@
 
 #include "common/alignment.h"
 #include "core/core.h"
-#include "core/memory.h"
 #include "video_core/renderer_opengl/gl_buffer_cache.h"
 #include "video_core/renderer_opengl/gl_rasterizer.h"
 
@@ -15,8 +14,8 @@ namespace OpenGL {
 
 CachedBufferEntry::CachedBufferEntry(VAddr cpu_addr, std::size_t size, GLintptr offset,
                                      std::size_t alignment, u8* host_ptr)
-    : cpu_addr{cpu_addr}, size{size}, offset{offset}, alignment{alignment}, RasterizerCacheObject{
-                                                                                host_ptr} {}
+    : RasterizerCacheObject{host_ptr}, cpu_addr{cpu_addr}, size{size}, offset{offset},
+      alignment{alignment} {}
 
 OGLBufferCache::OGLBufferCache(RasterizerOpenGL& rasterizer, std::size_t size)
     : RasterizerCache{rasterizer}, stream_buffer(size, true) {}
diff --git a/src/video_core/renderer_opengl/gl_global_cache.cpp b/src/video_core/renderer_opengl/gl_global_cache.cpp
index 0fbfbad55..5842d6213 100644
--- a/src/video_core/renderer_opengl/gl_global_cache.cpp
+++ b/src/video_core/renderer_opengl/gl_global_cache.cpp
@@ -4,7 +4,6 @@
 
 #include <glad/glad.h>
 
-#include "common/assert.h"
 #include "common/logging/log.h"
 #include "core/core.h"
 #include "video_core/renderer_opengl/gl_global_cache.h"
@@ -15,7 +14,7 @@
 namespace OpenGL {
 
 CachedGlobalRegion::CachedGlobalRegion(VAddr cpu_addr, u32 size, u8* host_ptr)
-    : cpu_addr{cpu_addr}, size{size}, RasterizerCacheObject{host_ptr} {
+    : RasterizerCacheObject{host_ptr}, cpu_addr{cpu_addr}, size{size} {
     buffer.Create();
     // Bind and unbind the buffer so it gets allocated by the driver
     glBindBuffer(GL_SHADER_STORAGE_BUFFER, buffer.handle);
diff --git a/src/video_core/renderer_opengl/gl_primitive_assembler.cpp b/src/video_core/renderer_opengl/gl_primitive_assembler.cpp
index 2bcbd3da2..c3e94d917 100644
--- a/src/video_core/renderer_opengl/gl_primitive_assembler.cpp
+++ b/src/video_core/renderer_opengl/gl_primitive_assembler.cpp
@@ -7,7 +7,7 @@
 #include "common/assert.h"
 #include "common/common_types.h"
 #include "core/core.h"
-#include "core/memory.h"
+#include "video_core/memory_manager.h"
 #include "video_core/renderer_opengl/gl_buffer_cache.h"
 #include "video_core/renderer_opengl/gl_primitive_assembler.h"
 
diff --git a/src/video_core/renderer_opengl/gl_primitive_assembler.h b/src/video_core/renderer_opengl/gl_primitive_assembler.h
index 0e2e7dc36..4e87ce4d6 100644
--- a/src/video_core/renderer_opengl/gl_primitive_assembler.h
+++ b/src/video_core/renderer_opengl/gl_primitive_assembler.h
@@ -4,11 +4,9 @@
 
 #pragma once
 
-#include <vector>
 #include <glad/glad.h>
 
 #include "common/common_types.h"
-#include "video_core/memory_manager.h"
 
 namespace OpenGL {
 
diff --git a/src/video_core/renderer_opengl/gl_rasterizer.cpp b/src/video_core/renderer_opengl/gl_rasterizer.cpp
index e06dfe43f..7ff1e6737 100644
--- a/src/video_core/renderer_opengl/gl_rasterizer.cpp
+++ b/src/video_core/renderer_opengl/gl_rasterizer.cpp
@@ -17,7 +17,6 @@
 #include "common/microprofile.h"
 #include "common/scope_exit.h"
 #include "core/core.h"
-#include "core/frontend/emu_window.h"
 #include "core/hle/kernel/process.h"
 #include "core/settings.h"
 #include "video_core/engines/maxwell_3d.h"
@@ -26,7 +25,6 @@
 #include "video_core/renderer_opengl/gl_shader_gen.h"
 #include "video_core/renderer_opengl/maxwell_to_gl.h"
 #include "video_core/renderer_opengl/renderer_opengl.h"
-#include "video_core/video_core.h"
 
 namespace OpenGL {
 
@@ -100,11 +98,9 @@ struct FramebufferCacheKey {
     }
 };
 
-RasterizerOpenGL::RasterizerOpenGL(Core::Frontend::EmuWindow& window, Core::System& system,
-                                   ScreenInfo& info)
-    : res_cache{*this}, shader_cache{*this, system}, global_cache{*this},
-      emu_window{window}, system{system}, screen_info{info},
-      buffer_cache(*this, STREAM_BUFFER_SIZE) {
+RasterizerOpenGL::RasterizerOpenGL(Core::System& system, ScreenInfo& info)
+    : res_cache{*this}, shader_cache{*this, system}, global_cache{*this}, system{system},
+      screen_info{info}, buffer_cache(*this, STREAM_BUFFER_SIZE) {
     // Create sampler objects
     for (std::size_t i = 0; i < texture_samplers.size(); ++i) {
         texture_samplers[i].Create();
@@ -320,7 +316,7 @@ void RasterizerOpenGL::SetupShaders(GLenum primitive_mode) {
         const std::size_t stage{index == 0 ? 0 : index - 1}; // Stage indices are 0 - 5
 
         GLShader::MaxwellUniformData ubo{};
-        ubo.SetFromRegs(gpu.state.shader_stages[stage]);
+        ubo.SetFromRegs(gpu, stage);
         const GLintptr offset = buffer_cache.UploadHostMemory(
             &ubo, sizeof(ubo), static_cast<std::size_t>(uniform_buffer_alignment));
 
diff --git a/src/video_core/renderer_opengl/gl_rasterizer.h b/src/video_core/renderer_opengl/gl_rasterizer.h
index 30f3e8acb..54fbf48aa 100644
--- a/src/video_core/renderer_opengl/gl_rasterizer.h
+++ b/src/video_core/renderer_opengl/gl_rasterizer.h
@@ -12,15 +12,12 @@
 #include <optional>
 #include <tuple>
 #include <utility>
-#include <vector>
 
 #include <boost/icl/interval_map.hpp>
-#include <boost/range/iterator_range.hpp>
 #include <glad/glad.h>
 
 #include "common/common_types.h"
 #include "video_core/engines/maxwell_3d.h"
-#include "video_core/memory_manager.h"
 #include "video_core/rasterizer_cache.h"
 #include "video_core/rasterizer_interface.h"
 #include "video_core/renderer_opengl/gl_buffer_cache.h"
@@ -29,10 +26,8 @@
 #include "video_core/renderer_opengl/gl_rasterizer_cache.h"
 #include "video_core/renderer_opengl/gl_resource_manager.h"
 #include "video_core/renderer_opengl/gl_shader_cache.h"
-#include "video_core/renderer_opengl/gl_shader_gen.h"
 #include "video_core/renderer_opengl/gl_shader_manager.h"
 #include "video_core/renderer_opengl/gl_state.h"
-#include "video_core/renderer_opengl/gl_stream_buffer.h"
 
 namespace Core {
 class System;
@@ -50,8 +45,7 @@ struct FramebufferCacheKey;
 
 class RasterizerOpenGL : public VideoCore::RasterizerInterface {
 public:
-    explicit RasterizerOpenGL(Core::Frontend::EmuWindow& window, Core::System& system,
-                              ScreenInfo& info);
+    explicit RasterizerOpenGL(Core::System& system, ScreenInfo& info);
     ~RasterizerOpenGL() override;
 
     void DrawArrays() override;
@@ -214,7 +208,6 @@ private:
     ShaderCacheOpenGL shader_cache;
     GlobalRegionCacheOpenGL global_cache;
 
-    Core::Frontend::EmuWindow& emu_window;
     Core::System& system;
 
     ScreenInfo& screen_info;
diff --git a/src/video_core/renderer_opengl/gl_rasterizer_cache.cpp b/src/video_core/renderer_opengl/gl_rasterizer_cache.cpp
index 0235317c0..7a3280620 100644
--- a/src/video_core/renderer_opengl/gl_rasterizer_cache.cpp
+++ b/src/video_core/renderer_opengl/gl_rasterizer_cache.cpp
@@ -13,7 +13,6 @@
 #include "common/scope_exit.h"
 #include "core/core.h"
 #include "core/hle/kernel/process.h"
-#include "core/memory.h"
 #include "core/settings.h"
 #include "video_core/engines/maxwell_3d.h"
 #include "video_core/morton.h"
@@ -562,8 +561,8 @@ void RasterizerCacheOpenGL::CopySurface(const Surface& src_surface, const Surfac
 }
 
 CachedSurface::CachedSurface(const SurfaceParams& params)
-    : params{params}, gl_target{SurfaceTargetToGL(params.target)},
-      cached_size_in_bytes{params.size_in_bytes}, RasterizerCacheObject{params.host_ptr} {
+    : RasterizerCacheObject{params.host_ptr}, params{params},
+      gl_target{SurfaceTargetToGL(params.target)}, cached_size_in_bytes{params.size_in_bytes} {
 
     const auto optional_cpu_addr{
         Core::System::GetInstance().GPU().MemoryManager().GpuToCpuAddress(params.gpu_addr)};
diff --git a/src/video_core/renderer_opengl/gl_rasterizer_cache.h b/src/video_core/renderer_opengl/gl_rasterizer_cache.h
index c644271d0..ad4fd3ad2 100644
--- a/src/video_core/renderer_opengl/gl_rasterizer_cache.h
+++ b/src/video_core/renderer_opengl/gl_rasterizer_cache.h
@@ -5,10 +5,9 @@
 #pragma once
 
 #include <array>
-#include <map>
 #include <memory>
 #include <string>
-#include <unordered_set>
+#include <tuple>
 #include <vector>
 
 #include "common/alignment.h"
@@ -538,12 +537,12 @@ private:
         return nullptr;
     }
 
-    void Register(const Surface& object) {
+    void Register(const Surface& object) override {
         RasterizerCache<Surface>::Register(object);
     }
 
     /// Unregisters an object from the cache
-    void Unregister(const Surface& object) {
+    void Unregister(const Surface& object) override {
         if (object->IsReinterpreted()) {
             auto interval = GetReinterpretInterval(object);
             reinterpreted_surfaces.erase(interval);
diff --git a/src/video_core/renderer_opengl/gl_shader_cache.cpp b/src/video_core/renderer_opengl/gl_shader_cache.cpp
index 1f8eca6f0..7030db365 100644
--- a/src/video_core/renderer_opengl/gl_shader_cache.cpp
+++ b/src/video_core/renderer_opengl/gl_shader_cache.cpp
@@ -6,13 +6,11 @@
 #include "common/assert.h"
 #include "common/hash.h"
 #include "core/core.h"
-#include "core/memory.h"
 #include "video_core/engines/maxwell_3d.h"
 #include "video_core/renderer_opengl/gl_rasterizer.h"
 #include "video_core/renderer_opengl/gl_shader_cache.h"
 #include "video_core/renderer_opengl/gl_shader_decompiler.h"
 #include "video_core/renderer_opengl/gl_shader_disk_cache.h"
-#include "video_core/renderer_opengl/gl_shader_manager.h"
 #include "video_core/renderer_opengl/utils.h"
 #include "video_core/shader/shader_ir.h"
 
@@ -215,9 +213,9 @@ CachedShader::CachedShader(VAddr cpu_addr, u64 unique_identifier,
                            Maxwell::ShaderProgram program_type, ShaderDiskCacheOpenGL& disk_cache,
                            const PrecompiledPrograms& precompiled_programs,
                            ProgramCode&& program_code, ProgramCode&& program_code_b, u8* host_ptr)
-    : host_ptr{host_ptr}, cpu_addr{cpu_addr}, unique_identifier{unique_identifier},
-      program_type{program_type}, disk_cache{disk_cache},
-      precompiled_programs{precompiled_programs}, RasterizerCacheObject{host_ptr} {
+    : RasterizerCacheObject{host_ptr}, host_ptr{host_ptr}, cpu_addr{cpu_addr},
+      unique_identifier{unique_identifier}, program_type{program_type}, disk_cache{disk_cache},
+      precompiled_programs{precompiled_programs} {
 
     const std::size_t code_size = CalculateProgramSize(program_code);
     const std::size_t code_size_b =
@@ -245,9 +243,9 @@ CachedShader::CachedShader(VAddr cpu_addr, u64 unique_identifier,
                            Maxwell::ShaderProgram program_type, ShaderDiskCacheOpenGL& disk_cache,
                            const PrecompiledPrograms& precompiled_programs,
                            GLShader::ProgramResult result, u8* host_ptr)
-    : cpu_addr{cpu_addr}, unique_identifier{unique_identifier}, program_type{program_type},
-      disk_cache{disk_cache}, precompiled_programs{precompiled_programs}, RasterizerCacheObject{
-                                                                              host_ptr} {
+    : RasterizerCacheObject{host_ptr}, cpu_addr{cpu_addr}, unique_identifier{unique_identifier},
+      program_type{program_type}, disk_cache{disk_cache}, precompiled_programs{
+                                                              precompiled_programs} {
 
     code = std::move(result.first);
     entries = result.second;
diff --git a/src/video_core/renderer_opengl/gl_shader_cache.h b/src/video_core/renderer_opengl/gl_shader_cache.h
index fd1c85115..0cf8e0b3d 100644
--- a/src/video_core/renderer_opengl/gl_shader_cache.h
+++ b/src/video_core/renderer_opengl/gl_shader_cache.h
@@ -5,21 +5,20 @@
 #pragma once
 
 #include <array>
+#include <atomic>
 #include <memory>
 #include <set>
 #include <tuple>
 #include <unordered_map>
+#include <vector>
 
 #include <glad/glad.h>
 
-#include "common/assert.h"
 #include "common/common_types.h"
 #include "video_core/rasterizer_cache.h"
-#include "video_core/renderer_base.h"
 #include "video_core/renderer_opengl/gl_resource_manager.h"
 #include "video_core/renderer_opengl/gl_shader_decompiler.h"
 #include "video_core/renderer_opengl/gl_shader_disk_cache.h"
-#include "video_core/renderer_opengl/gl_shader_gen.h"
 
 namespace Core {
 class System;
diff --git a/src/video_core/renderer_opengl/gl_shader_decompiler.cpp b/src/video_core/renderer_opengl/gl_shader_decompiler.cpp
index 11d1169f0..7300a4037 100644
--- a/src/video_core/renderer_opengl/gl_shader_decompiler.cpp
+++ b/src/video_core/renderer_opengl/gl_shader_decompiler.cpp
@@ -1196,11 +1196,12 @@ private:
         switch (meta->element) {
         case 0:
         case 1:
-            return "textureSize(" + sampler + ", " + lod + ')' + GetSwizzle(meta->element);
+            return "itof(int(textureSize(" + sampler + ", " + lod + ')' +
+                   GetSwizzle(meta->element) + "))";
         case 2:
             return "0";
         case 3:
-            return "textureQueryLevels(" + sampler + ')';
+            return "itof(textureQueryLevels(" + sampler + "))";
         }
         UNREACHABLE();
         return "0";
diff --git a/src/video_core/renderer_opengl/gl_shader_decompiler.h b/src/video_core/renderer_opengl/gl_shader_decompiler.h
index 72aca4938..4e04ab2f8 100644
--- a/src/video_core/renderer_opengl/gl_shader_decompiler.h
+++ b/src/video_core/renderer_opengl/gl_shader_decompiler.h
@@ -5,7 +5,6 @@
 #pragma once
 
 #include <array>
-#include <set>
 #include <string>
 #include <utility>
 #include <vector>
diff --git a/src/video_core/renderer_opengl/gl_shader_disk_cache.cpp b/src/video_core/renderer_opengl/gl_shader_disk_cache.cpp
index 82fc4d44b..d2d979997 100644
--- a/src/video_core/renderer_opengl/gl_shader_disk_cache.cpp
+++ b/src/video_core/renderer_opengl/gl_shader_disk_cache.cpp
@@ -4,13 +4,13 @@
 
 #include <cstring>
 #include <fmt/format.h>
-#include <lz4.h>
 
 #include "common/assert.h"
 #include "common/common_paths.h"
 #include "common/common_types.h"
 #include "common/file_util.h"
 #include "common/logging/log.h"
+#include "common/lz4_compression.h"
 #include "common/scm_rev.h"
 
 #include "core/core.h"
@@ -49,39 +49,6 @@ ShaderCacheVersionHash GetShaderCacheVersionHash() {
     return hash;
 }
 
-template <typename T>
-std::vector<u8> CompressData(const T* source, std::size_t source_size) {
-    if (source_size > LZ4_MAX_INPUT_SIZE) {
-        // Source size exceeds LZ4 maximum input size
-        return {};
-    }
-    const auto source_size_int = static_cast<int>(source_size);
-    const int max_compressed_size = LZ4_compressBound(source_size_int);
-    std::vector<u8> compressed(max_compressed_size);
-    const int compressed_size = LZ4_compress_default(reinterpret_cast<const char*>(source),
-                                                     reinterpret_cast<char*>(compressed.data()),
-                                                     source_size_int, max_compressed_size);
-    if (compressed_size <= 0) {
-        // Compression failed
-        return {};
-    }
-    compressed.resize(compressed_size);
-    return compressed;
-}
-
-std::vector<u8> DecompressData(const std::vector<u8>& compressed, std::size_t uncompressed_size) {
-    std::vector<u8> uncompressed(uncompressed_size);
-    const int size_check = LZ4_decompress_safe(reinterpret_cast<const char*>(compressed.data()),
-                                               reinterpret_cast<char*>(uncompressed.data()),
-                                               static_cast<int>(compressed.size()),
-                                               static_cast<int>(uncompressed.size()));
-    if (static_cast<int>(uncompressed_size) != size_check) {
-        // Decompression failed
-        return {};
-    }
-    return uncompressed;
-}
-
 } // namespace
 
 ShaderDiskCacheRaw::ShaderDiskCacheRaw(u64 unique_identifier, Maxwell::ShaderProgram program_type,
@@ -292,7 +259,7 @@ ShaderDiskCacheOpenGL::LoadPrecompiledFile(FileUtil::IOFile& file) {
                 return {};
             }
 
-            dump.binary = DecompressData(compressed_binary, binary_length);
+            dump.binary = Common::Compression::DecompressDataLZ4(compressed_binary, binary_length);
             if (dump.binary.empty()) {
                 return {};
             }
@@ -321,7 +288,7 @@ std::optional<ShaderDiskCacheDecompiled> ShaderDiskCacheOpenGL::LoadDecompiledEn
         return {};
     }
 
-    const std::vector<u8> code = DecompressData(compressed_code, code_size);
+    const std::vector<u8> code = Common::Compression::DecompressDataLZ4(compressed_code, code_size);
     if (code.empty()) {
         return {};
     }
@@ -507,7 +474,8 @@ void ShaderDiskCacheOpenGL::SaveDecompiled(u64 unique_identifier, const std::str
     if (!IsUsable())
         return;
 
-    const std::vector<u8> compressed_code{CompressData(code.data(), code.size())};
+    const std::vector<u8> compressed_code{Common::Compression::CompressDataLZ4HC(
+        reinterpret_cast<const u8*>(code.data()), code.size(), 9)};
     if (compressed_code.empty()) {
         LOG_ERROR(Render_OpenGL, "Failed to compress GLSL code - skipping shader {:016x}",
                   unique_identifier);
@@ -537,7 +505,9 @@ void ShaderDiskCacheOpenGL::SaveDump(const ShaderDiskCacheUsage& usage, GLuint p
     std::vector<u8> binary(binary_length);
     glGetProgramBinary(program, binary_length, nullptr, &binary_format, binary.data());
 
-    const std::vector<u8> compressed_binary = CompressData(binary.data(), binary.size());
+    const std::vector<u8> compressed_binary =
+        Common::Compression::CompressDataLZ4HC(binary.data(), binary.size(), 9);
+
     if (compressed_binary.empty()) {
         LOG_ERROR(Render_OpenGL, "Failed to compress binary program in shader={:016x}",
                   usage.unique_identifier);
diff --git a/src/video_core/renderer_opengl/gl_shader_gen.cpp b/src/video_core/renderer_opengl/gl_shader_gen.cpp
index 7d96649af..8763d9c71 100644
--- a/src/video_core/renderer_opengl/gl_shader_gen.cpp
+++ b/src/video_core/renderer_opengl/gl_shader_gen.cpp
@@ -3,7 +3,6 @@
 // Refer to the license.txt file included.
 
 #include <fmt/format.h>
-#include "common/assert.h"
 #include "video_core/engines/maxwell_3d.h"
 #include "video_core/renderer_opengl/gl_shader_decompiler.h"
 #include "video_core/renderer_opengl/gl_shader_gen.h"
diff --git a/src/video_core/renderer_opengl/gl_shader_gen.h b/src/video_core/renderer_opengl/gl_shader_gen.h
index fba8e681b..fad346b48 100644
--- a/src/video_core/renderer_opengl/gl_shader_gen.h
+++ b/src/video_core/renderer_opengl/gl_shader_gen.h
@@ -4,12 +4,9 @@
 
 #pragma once
 
-#include <array>
-#include <string>
 #include <vector>
 
 #include "common/common_types.h"
-#include "video_core/engines/shader_bytecode.h"
 #include "video_core/renderer_opengl/gl_shader_decompiler.h"
 #include "video_core/shader/shader_ir.h"
 
diff --git a/src/video_core/renderer_opengl/gl_shader_manager.cpp b/src/video_core/renderer_opengl/gl_shader_manager.cpp
index 6a30c28d2..eaf3e03a0 100644
--- a/src/video_core/renderer_opengl/gl_shader_manager.cpp
+++ b/src/video_core/renderer_opengl/gl_shader_manager.cpp
@@ -2,15 +2,15 @@
 // Licensed under GPLv2 or any later version
 // Refer to the license.txt file included.
 
-#include "core/core.h"
 #include "video_core/renderer_opengl/gl_shader_manager.h"
 
 namespace OpenGL::GLShader {
 
-void MaxwellUniformData::SetFromRegs(const Maxwell3D::State::ShaderStageInfo& shader_stage) {
-    const auto& gpu = Core::System::GetInstance().GPU().Maxwell3D();
-    const auto& regs = gpu.regs;
-    const auto& state = gpu.state;
+using Tegra::Engines::Maxwell3D;
+
+void MaxwellUniformData::SetFromRegs(const Maxwell3D& maxwell, std::size_t shader_stage) {
+    const auto& regs = maxwell.regs;
+    const auto& state = maxwell.state;
 
     // TODO(bunnei): Support more than one viewport
     viewport_flip[0] = regs.viewport_transform[0].scale_x < 0.0 ? -1.0f : 1.0f;
@@ -18,7 +18,7 @@ void MaxwellUniformData::SetFromRegs(const Maxwell3D::State::ShaderStageInfo& sh
 
     u32 func = static_cast<u32>(regs.alpha_test_func);
     // Normalize the gl variants of opCompare to be the same as the normal variants
-    u32 op_gl_variant_base = static_cast<u32>(Tegra::Engines::Maxwell3D::Regs::ComparisonOp::Never);
+    const u32 op_gl_variant_base = static_cast<u32>(Maxwell3D::Regs::ComparisonOp::Never);
     if (func >= op_gl_variant_base) {
         func = func - op_gl_variant_base + 1U;
     }
@@ -31,8 +31,9 @@ void MaxwellUniformData::SetFromRegs(const Maxwell3D::State::ShaderStageInfo& sh
 
     // Assign in which stage the position has to be flipped
     // (the last stage before the fragment shader).
-    if (gpu.regs.shader_config[static_cast<u32>(Maxwell3D::Regs::ShaderProgram::Geometry)].enable) {
-        flip_stage = static_cast<u32>(Maxwell3D::Regs::ShaderProgram::Geometry);
+    constexpr u32 geometry_index = static_cast<u32>(Maxwell3D::Regs::ShaderProgram::Geometry);
+    if (maxwell.regs.shader_config[geometry_index].enable) {
+        flip_stage = geometry_index;
     } else {
         flip_stage = static_cast<u32>(Maxwell3D::Regs::ShaderProgram::VertexB);
     }
diff --git a/src/video_core/renderer_opengl/gl_shader_manager.h b/src/video_core/renderer_opengl/gl_shader_manager.h
index 4970aafed..8eef2a920 100644
--- a/src/video_core/renderer_opengl/gl_shader_manager.h
+++ b/src/video_core/renderer_opengl/gl_shader_manager.h
@@ -12,14 +12,13 @@
 
 namespace OpenGL::GLShader {
 
-using Tegra::Engines::Maxwell3D;
-
 /// Uniform structure for the Uniform Buffer Object, all vectors must be 16-byte aligned
-// NOTE: Always keep a vec4 at the end. The GL spec is not clear whether the alignment at
-//       the end of a uniform block is included in UNIFORM_BLOCK_DATA_SIZE or not.
-//       Not following that rule will cause problems on some AMD drivers.
+/// @note Always keep a vec4 at the end. The GL spec is not clear whether the alignment at
+///       the end of a uniform block is included in UNIFORM_BLOCK_DATA_SIZE or not.
+///       Not following that rule will cause problems on some AMD drivers.
 struct MaxwellUniformData {
-    void SetFromRegs(const Maxwell3D::State::ShaderStageInfo& shader_stage);
+    void SetFromRegs(const Tegra::Engines::Maxwell3D& maxwell, std::size_t shader_stage);
+
     alignas(16) GLvec4 viewport_flip;
     struct alignas(16) {
         GLuint instance_id;
diff --git a/src/video_core/renderer_opengl/renderer_opengl.cpp b/src/video_core/renderer_opengl/renderer_opengl.cpp
index 5e3d862c6..d69cba9c3 100644
--- a/src/video_core/renderer_opengl/renderer_opengl.cpp
+++ b/src/video_core/renderer_opengl/renderer_opengl.cpp
@@ -5,7 +5,6 @@
 #include <algorithm>
 #include <cstddef>
 #include <cstdlib>
-#include <cstring>
 #include <memory>
 #include <glad/glad.h>
 #include "common/assert.h"
@@ -266,7 +265,7 @@ void RendererOpenGL::CreateRasterizer() {
     }
     // Initialize sRGB Usage
     OpenGLState::ClearsRGBUsed();
-    rasterizer = std::make_unique<RasterizerOpenGL>(render_window, system, screen_info);
+    rasterizer = std::make_unique<RasterizerOpenGL>(system, screen_info);
 }
 
 void RendererOpenGL::ConfigureFramebufferTexture(TextureInfo& texture,
diff --git a/src/video_core/renderer_vulkan/vk_buffer_cache.cpp b/src/video_core/renderer_vulkan/vk_buffer_cache.cpp
index eac51ecb3..388b5ffd5 100644
--- a/src/video_core/renderer_vulkan/vk_buffer_cache.cpp
+++ b/src/video_core/renderer_vulkan/vk_buffer_cache.cpp
@@ -19,8 +19,8 @@ namespace Vulkan {
 
 CachedBufferEntry::CachedBufferEntry(VAddr cpu_addr, std::size_t size, u64 offset,
                                      std::size_t alignment, u8* host_ptr)
-    : cpu_addr{cpu_addr}, size{size}, offset{offset}, alignment{alignment}, RasterizerCacheObject{
-                                                                                host_ptr} {}
+    : RasterizerCacheObject{host_ptr}, cpu_addr{cpu_addr}, size{size}, offset{offset},
+      alignment{alignment} {}
 
 VKBufferCache::VKBufferCache(Tegra::MemoryManager& tegra_memory_manager,
                              VideoCore::RasterizerInterface& rasterizer, const VKDevice& device,
diff --git a/src/video_core/renderer_vulkan/vk_resource_manager.cpp b/src/video_core/renderer_vulkan/vk_resource_manager.cpp
index a1e117443..13c46e5b8 100644
--- a/src/video_core/renderer_vulkan/vk_resource_manager.cpp
+++ b/src/video_core/renderer_vulkan/vk_resource_manager.cpp
@@ -21,7 +21,7 @@ public:
     CommandBufferPool(const VKDevice& device)
         : VKFencedPool(COMMAND_BUFFER_POOL_SIZE), device{device} {}
 
-    void Allocate(std::size_t begin, std::size_t end) {
+    void Allocate(std::size_t begin, std::size_t end) override {
         const auto dev = device.GetLogical();
         const auto& dld = device.GetDispatchLoader();
         const u32 graphics_family = device.GetGraphicsFamily();
diff --git a/src/video_core/renderer_vulkan/vk_resource_manager.h b/src/video_core/renderer_vulkan/vk_resource_manager.h
index 5bfe4cead..08ee86fa6 100644
--- a/src/video_core/renderer_vulkan/vk_resource_manager.h
+++ b/src/video_core/renderer_vulkan/vk_resource_manager.h
@@ -97,7 +97,7 @@ private:
 class VKFenceWatch final : public VKResource {
 public:
     explicit VKFenceWatch();
-    ~VKFenceWatch();
+    ~VKFenceWatch() override;
 
     /// Waits for the fence to be released.
     void Wait();
diff --git a/src/video_core/renderer_vulkan/vk_swapchain.cpp b/src/video_core/renderer_vulkan/vk_swapchain.cpp
new file mode 100644
index 000000000..08279e562
--- /dev/null
+++ b/src/video_core/renderer_vulkan/vk_swapchain.cpp
@@ -0,0 +1,210 @@
+// Copyright 2019 yuzu Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#include <algorithm>
+#include <array>
+#include <limits>
+#include <vector>
+
+#include "common/assert.h"
+#include "common/logging/log.h"
+#include "core/core.h"
+#include "core/frontend/framebuffer_layout.h"
+#include "video_core/renderer_vulkan/declarations.h"
+#include "video_core/renderer_vulkan/vk_device.h"
+#include "video_core/renderer_vulkan/vk_resource_manager.h"
+#include "video_core/renderer_vulkan/vk_swapchain.h"
+
+namespace Vulkan {
+
+namespace {
+vk::SurfaceFormatKHR ChooseSwapSurfaceFormat(const std::vector<vk::SurfaceFormatKHR>& formats) {
+    if (formats.size() == 1 && formats[0].format == vk::Format::eUndefined) {
+        return {vk::Format::eB8G8R8A8Unorm, vk::ColorSpaceKHR::eSrgbNonlinear};
+    }
+    const auto& found = std::find_if(formats.begin(), formats.end(), [](const auto& format) {
+        return format.format == vk::Format::eB8G8R8A8Unorm &&
+               format.colorSpace == vk::ColorSpaceKHR::eSrgbNonlinear;
+    });
+    return found != formats.end() ? *found : formats[0];
+}
+
+vk::PresentModeKHR ChooseSwapPresentMode(const std::vector<vk::PresentModeKHR>& modes) {
+    // Mailbox doesn't lock the application like fifo (vsync), prefer it
+    const auto& found = std::find_if(modes.begin(), modes.end(), [](const auto& mode) {
+        return mode == vk::PresentModeKHR::eMailbox;
+    });
+    return found != modes.end() ? *found : vk::PresentModeKHR::eFifo;
+}
+
+vk::Extent2D ChooseSwapExtent(const vk::SurfaceCapabilitiesKHR& capabilities, u32 width,
+                              u32 height) {
+    constexpr auto undefined_size{std::numeric_limits<u32>::max()};
+    if (capabilities.currentExtent.width != undefined_size) {
+        return capabilities.currentExtent;
+    }
+    vk::Extent2D extent = {width, height};
+    extent.width = std::max(capabilities.minImageExtent.width,
+                            std::min(capabilities.maxImageExtent.width, extent.width));
+    extent.height = std::max(capabilities.minImageExtent.height,
+                             std::min(capabilities.maxImageExtent.height, extent.height));
+    return extent;
+}
+} // namespace
+
+VKSwapchain::VKSwapchain(vk::SurfaceKHR surface, const VKDevice& device)
+    : surface{surface}, device{device} {}
+
+VKSwapchain::~VKSwapchain() = default;
+
+void VKSwapchain::Create(u32 width, u32 height) {
+    const auto dev = device.GetLogical();
+    const auto& dld = device.GetDispatchLoader();
+    const auto physical_device = device.GetPhysical();
+
+    const vk::SurfaceCapabilitiesKHR capabilities{
+        physical_device.getSurfaceCapabilitiesKHR(surface, dld)};
+    if (capabilities.maxImageExtent.width == 0 || capabilities.maxImageExtent.height == 0) {
+        return;
+    }
+
+    dev.waitIdle(dld);
+    Destroy();
+
+    CreateSwapchain(capabilities, width, height);
+    CreateSemaphores();
+    CreateImageViews();
+
+    fences.resize(image_count, nullptr);
+}
+
+void VKSwapchain::AcquireNextImage() {
+    const auto dev{device.GetLogical()};
+    const auto& dld{device.GetDispatchLoader()};
+    dev.acquireNextImageKHR(*swapchain, std::numeric_limits<u64>::max(),
+                            *present_semaphores[frame_index], {}, &image_index, dld);
+
+    if (auto& fence = fences[image_index]; fence) {
+        fence->Wait();
+        fence->Release();
+        fence = nullptr;
+    }
+}
+
+bool VKSwapchain::Present(vk::Semaphore render_semaphore, VKFence& fence) {
+    const vk::Semaphore present_semaphore{*present_semaphores[frame_index]};
+    const std::array<vk::Semaphore, 2> semaphores{present_semaphore, render_semaphore};
+    const u32 wait_semaphore_count{render_semaphore ? 2U : 1U};
+    const auto& dld{device.GetDispatchLoader()};
+    const auto present_queue{device.GetPresentQueue()};
+    bool recreated = false;
+
+    const vk::PresentInfoKHR present_info(wait_semaphore_count, semaphores.data(), 1,
+                                          &swapchain.get(), &image_index, {});
+    switch (const auto result = present_queue.presentKHR(&present_info, dld); result) {
+    case vk::Result::eSuccess:
+        break;
+    case vk::Result::eErrorOutOfDateKHR:
+        if (current_width > 0 && current_height > 0) {
+            Create(current_width, current_height);
+            recreated = true;
+        }
+        break;
+    default:
+        LOG_CRITICAL(Render_Vulkan, "Vulkan failed to present swapchain due to {}!",
+                     vk::to_string(result));
+        UNREACHABLE();
+    }
+
+    ASSERT(fences[image_index] == nullptr);
+    fences[image_index] = &fence;
+    frame_index = (frame_index + 1) % image_count;
+    return recreated;
+}
+
+bool VKSwapchain::HasFramebufferChanged(const Layout::FramebufferLayout& framebuffer) const {
+    // TODO(Rodrigo): Handle framebuffer pixel format changes
+    return framebuffer.width != current_width || framebuffer.height != current_height;
+}
+
+void VKSwapchain::CreateSwapchain(const vk::SurfaceCapabilitiesKHR& capabilities, u32 width,
+                                  u32 height) {
+    const auto dev{device.GetLogical()};
+    const auto& dld{device.GetDispatchLoader()};
+    const auto physical_device{device.GetPhysical()};
+
+    const std::vector<vk::SurfaceFormatKHR> formats{
+        physical_device.getSurfaceFormatsKHR(surface, dld)};
+
+    const std::vector<vk::PresentModeKHR> present_modes{
+        physical_device.getSurfacePresentModesKHR(surface, dld)};
+
+    const vk::SurfaceFormatKHR surface_format{ChooseSwapSurfaceFormat(formats)};
+    const vk::PresentModeKHR present_mode{ChooseSwapPresentMode(present_modes)};
+    extent = ChooseSwapExtent(capabilities, width, height);
+
+    current_width = extent.width;
+    current_height = extent.height;
+
+    u32 requested_image_count{capabilities.minImageCount + 1};
+    if (capabilities.maxImageCount > 0 && requested_image_count > capabilities.maxImageCount) {
+        requested_image_count = capabilities.maxImageCount;
+    }
+
+    vk::SwapchainCreateInfoKHR swapchain_ci(
+        {}, surface, requested_image_count, surface_format.format, surface_format.colorSpace,
+        extent, 1, vk::ImageUsageFlagBits::eColorAttachment, {}, {}, {},
+        capabilities.currentTransform, vk::CompositeAlphaFlagBitsKHR::eOpaque, present_mode, false,
+        {});
+
+    const u32 graphics_family{device.GetGraphicsFamily()};
+    const u32 present_family{device.GetPresentFamily()};
+    const std::array<u32, 2> queue_indices{graphics_family, present_family};
+    if (graphics_family != present_family) {
+        swapchain_ci.imageSharingMode = vk::SharingMode::eConcurrent;
+        swapchain_ci.queueFamilyIndexCount = static_cast<u32>(queue_indices.size());
+        swapchain_ci.pQueueFamilyIndices = queue_indices.data();
+    } else {
+        swapchain_ci.imageSharingMode = vk::SharingMode::eExclusive;
+    }
+
+    swapchain = dev.createSwapchainKHRUnique(swapchain_ci, nullptr, dld);
+
+    images = dev.getSwapchainImagesKHR(*swapchain, dld);
+    image_count = static_cast<u32>(images.size());
+    image_format = surface_format.format;
+}
+
+void VKSwapchain::CreateSemaphores() {
+    const auto dev{device.GetLogical()};
+    const auto& dld{device.GetDispatchLoader()};
+
+    present_semaphores.resize(image_count);
+    for (std::size_t i = 0; i < image_count; i++) {
+        present_semaphores[i] = dev.createSemaphoreUnique({}, nullptr, dld);
+    }
+}
+
+void VKSwapchain::CreateImageViews() {
+    const auto dev{device.GetLogical()};
+    const auto& dld{device.GetDispatchLoader()};
+
+    image_views.resize(image_count);
+    for (std::size_t i = 0; i < image_count; i++) {
+        const vk::ImageViewCreateInfo image_view_ci({}, images[i], vk::ImageViewType::e2D,
+                                                    image_format, {},
+                                                    {vk::ImageAspectFlagBits::eColor, 0, 1, 0, 1});
+        image_views[i] = dev.createImageViewUnique(image_view_ci, nullptr, dld);
+    }
+}
+
+void VKSwapchain::Destroy() {
+    frame_index = 0;
+    present_semaphores.clear();
+    framebuffers.clear();
+    image_views.clear();
+    swapchain.reset();
+}
+
+} // namespace Vulkan
diff --git a/src/video_core/renderer_vulkan/vk_swapchain.h b/src/video_core/renderer_vulkan/vk_swapchain.h
new file mode 100644
index 000000000..2ad84f185
--- /dev/null
+++ b/src/video_core/renderer_vulkan/vk_swapchain.h
@@ -0,0 +1,92 @@
+// Copyright 2019 yuzu Emulator Project
+// Licensed under GPLv2 or any later version
+// Refer to the license.txt file included.
+
+#pragma once
+
+#include <vector>
+
+#include "common/common_types.h"
+#include "video_core/renderer_vulkan/declarations.h"
+
+namespace Layout {
+struct FramebufferLayout;
+}
+
+namespace Vulkan {
+
+class VKDevice;
+class VKFence;
+
+class VKSwapchain {
+public:
+    explicit VKSwapchain(vk::SurfaceKHR surface, const VKDevice& device);
+    ~VKSwapchain();
+
+    /// Creates (or recreates) the swapchain with a given size.
+    void Create(u32 width, u32 height);
+
+    /// Acquires the next image in the swapchain, waits as needed.
+    void AcquireNextImage();
+
+    /// Presents the rendered image to the swapchain. Returns true when the swapchains had to be
+    /// recreated. Takes responsability for the ownership of fence.
+    bool Present(vk::Semaphore render_semaphore, VKFence& fence);
+
+    /// Returns true when the framebuffer layout has changed.
+    bool HasFramebufferChanged(const Layout::FramebufferLayout& framebuffer) const;
+
+    const vk::Extent2D& GetSize() const {
+        return extent;
+    }
+
+    u32 GetImageCount() const {
+        return image_count;
+    }
+
+    u32 GetImageIndex() const {
+        return image_index;
+    }
+
+    vk::Image GetImageIndex(u32 index) const {
+        return images[index];
+    }
+
+    vk::ImageView GetImageViewIndex(u32 index) const {
+        return *image_views[index];
+    }
+
+    vk::Format GetImageFormat() const {
+        return image_format;
+    }
+
+private:
+    void CreateSwapchain(const vk::SurfaceCapabilitiesKHR& capabilities, u32 width, u32 height);
+    void CreateSemaphores();
+    void CreateImageViews();
+
+    void Destroy();
+
+    const vk::SurfaceKHR surface;
+    const VKDevice& device;
+
+    UniqueSwapchainKHR swapchain;
+
+    u32 image_count{};
+    std::vector<vk::Image> images;
+    std::vector<UniqueImageView> image_views;
+    std::vector<UniqueFramebuffer> framebuffers;
+    std::vector<VKFence*> fences;
+    std::vector<UniqueSemaphore> present_semaphores;
+
+    u32 image_index{};
+    u32 frame_index{};
+
+    vk::Format image_format{};
+    vk::Extent2D extent{};
+
+    u32 current_width{};
+    u32 current_height{};
+};
+
+} // namespace Vulkan
diff --git a/src/web_service/web_backend.cpp b/src/web_service/web_backend.cpp
index 40da1a4e2..dc149d2ed 100644
--- a/src/web_service/web_backend.cpp
+++ b/src/web_service/web_backend.cpp
@@ -24,7 +24,7 @@ constexpr u32 TIMEOUT_SECONDS = 30;
 struct Client::Impl {
     Impl(std::string host, std::string username, std::string token)
         : host{std::move(host)}, username{std::move(username)}, token{std::move(token)} {
-        std::lock_guard<std::mutex> lock(jwt_cache.mutex);
+        std::lock_guard lock{jwt_cache.mutex};
         if (this->username == jwt_cache.username && this->token == jwt_cache.token) {
             jwt = jwt_cache.jwt;
         }
@@ -151,7 +151,7 @@ struct Client::Impl {
         if (result.result_code != Common::WebResult::Code::Success) {
             LOG_ERROR(WebService, "UpdateJWT failed");
         } else {
-            std::lock_guard<std::mutex> lock(jwt_cache.mutex);
+            std::lock_guard lock{jwt_cache.mutex};
             jwt_cache.username = username;
             jwt_cache.token = token;
             jwt_cache.jwt = jwt = result.returned_data;
diff --git a/src/yuzu/applets/profile_select.cpp b/src/yuzu/applets/profile_select.cpp
index 5c1b65a2c..f95f7fe3c 100644
--- a/src/yuzu/applets/profile_select.cpp
+++ b/src/yuzu/applets/profile_select.cpp
@@ -58,10 +58,7 @@ QtProfileSelectionDialog::QtProfileSelectionDialog(QWidget* parent)
 
     scroll_area = new QScrollArea;
 
-    buttons = new QDialogButtonBox;
-    buttons->addButton(tr("Cancel"), QDialogButtonBox::RejectRole);
-    buttons->addButton(tr("OK"), QDialogButtonBox::AcceptRole);
-
+    buttons = new QDialogButtonBox(QDialogButtonBox::Cancel | QDialogButtonBox::Ok);
     connect(buttons, &QDialogButtonBox::accepted, this, &QtProfileSelectionDialog::accept);
     connect(buttons, &QDialogButtonBox::rejected, this, &QtProfileSelectionDialog::reject);
 
@@ -163,6 +160,6 @@ void QtProfileSelector::SelectProfile(
 
 void QtProfileSelector::MainWindowFinishedSelection(std::optional<Service::Account::UUID> uuid) {
     // Acquire the HLE mutex
-    std::lock_guard<std::recursive_mutex> lock(HLE::g_hle_lock);
+    std::lock_guard lock{HLE::g_hle_lock};
     callback(uuid);
 }
diff --git a/src/yuzu/applets/software_keyboard.cpp b/src/yuzu/applets/software_keyboard.cpp
index 8a26fdff1..f3eb29b25 100644
--- a/src/yuzu/applets/software_keyboard.cpp
+++ b/src/yuzu/applets/software_keyboard.cpp
@@ -75,13 +75,13 @@ QtSoftwareKeyboardDialog::QtSoftwareKeyboardDialog(
         length_label->setText(QStringLiteral("%1/%2").arg(text.size()).arg(parameters.max_length));
     });
 
-    buttons = new QDialogButtonBox;
-    buttons->addButton(tr("Cancel"), QDialogButtonBox::RejectRole);
-    buttons->addButton(parameters.submit_text.empty()
-                           ? tr("OK")
-                           : QString::fromStdU16String(parameters.submit_text),
-                       QDialogButtonBox::AcceptRole);
-
+    buttons = new QDialogButtonBox(QDialogButtonBox::Cancel);
+    if (parameters.submit_text.empty()) {
+        buttons->addButton(QDialogButtonBox::Ok);
+    } else {
+        buttons->addButton(QString::fromStdU16String(parameters.submit_text),
+                           QDialogButtonBox::AcceptRole);
+    }
     connect(buttons, &QDialogButtonBox::accepted, this, &QtSoftwareKeyboardDialog::accept);
     connect(buttons, &QDialogButtonBox::rejected, this, &QtSoftwareKeyboardDialog::reject);
     layout->addWidget(header_label);
@@ -141,12 +141,12 @@ void QtSoftwareKeyboard::SendTextCheckDialog(std::u16string error_message,
 
 void QtSoftwareKeyboard::MainWindowFinishedText(std::optional<std::u16string> text) {
     // Acquire the HLE mutex
-    std::lock_guard<std::recursive_mutex> lock(HLE::g_hle_lock);
+    std::lock_guard lock{HLE::g_hle_lock};
     text_output(text);
 }
 
 void QtSoftwareKeyboard::MainWindowFinishedCheckDialog() {
     // Acquire the HLE mutex
-    std::lock_guard<std::recursive_mutex> lock(HLE::g_hle_lock);
+    std::lock_guard lock{HLE::g_hle_lock};
     finished_check();
 }
diff --git a/src/yuzu/applets/web_browser.cpp b/src/yuzu/applets/web_browser.cpp
index 979b9ec14..ac80b2fa2 100644
--- a/src/yuzu/applets/web_browser.cpp
+++ b/src/yuzu/applets/web_browser.cpp
@@ -104,12 +104,12 @@ void QtWebBrowser::OpenPage(std::string_view url, std::function<void()> unpack_r
 
 void QtWebBrowser::MainWindowUnpackRomFS() {
     // Acquire the HLE mutex
-    std::lock_guard<std::recursive_mutex> lock(HLE::g_hle_lock);
+    std::lock_guard lock{HLE::g_hle_lock};
     unpack_romfs_callback();
 }
 
 void QtWebBrowser::MainWindowFinishedBrowsing() {
     // Acquire the HLE mutex
-    std::lock_guard<std::recursive_mutex> lock(HLE::g_hle_lock);
+    std::lock_guard lock{HLE::g_hle_lock};
     finished_callback();
 }
diff --git a/src/yuzu/bootmanager.cpp b/src/yuzu/bootmanager.cpp
index 05ad19e1d..7438fbc0a 100644
--- a/src/yuzu/bootmanager.cpp
+++ b/src/yuzu/bootmanager.cpp
@@ -67,7 +67,7 @@ void EmuThread::run() {
 
             was_active = false;
         } else {
-            std::unique_lock<std::mutex> lock(running_mutex);
+            std::unique_lock lock{running_mutex};
             running_cv.wait(lock, [this] { return IsRunning() || exec_step || stop_run; });
         }
     }
diff --git a/src/yuzu/bootmanager.h b/src/yuzu/bootmanager.h
index 7226e690e..3183621bc 100644
--- a/src/yuzu/bootmanager.h
+++ b/src/yuzu/bootmanager.h
@@ -53,7 +53,7 @@ public:
      * @note This function is thread-safe
      */
     void SetRunning(bool running) {
-        std::unique_lock<std::mutex> lock(running_mutex);
+        std::unique_lock lock{running_mutex};
         this->running = running;
         lock.unlock();
         running_cv.notify_all();
diff --git a/src/yuzu/configuration/config.cpp b/src/yuzu/configuration/config.cpp
index 4650f96a3..dead9f807 100644
--- a/src/yuzu/configuration/config.cpp
+++ b/src/yuzu/configuration/config.cpp
@@ -413,7 +413,6 @@ void Config::ReadValues() {
 
     qt_config->beginGroup("System");
     Settings::values.use_docked_mode = ReadSetting("use_docked_mode", false).toBool();
-    Settings::values.enable_nfc = ReadSetting("enable_nfc", true).toBool();
 
     Settings::values.current_user =
         std::clamp<int>(ReadSetting("current_user", 0).toInt(), 0, Service::Account::MAX_USERS - 1);
@@ -675,7 +674,6 @@ void Config::SaveValues() {
 
     qt_config->beginGroup("System");
     WriteSetting("use_docked_mode", Settings::values.use_docked_mode, false);
-    WriteSetting("enable_nfc", Settings::values.enable_nfc, true);
     WriteSetting("current_user", Settings::values.current_user, 0);
     WriteSetting("language_index", Settings::values.language_index, 1);
 
diff --git a/src/yuzu/configuration/configure_general.cpp b/src/yuzu/configuration/configure_general.cpp
index 4116b6cd7..389fcf667 100644
--- a/src/yuzu/configuration/configure_general.cpp
+++ b/src/yuzu/configuration/configure_general.cpp
@@ -33,7 +33,6 @@ void ConfigureGeneral::setConfiguration() {
     ui->toggle_user_on_boot->setChecked(UISettings::values.select_user_on_boot);
     ui->theme_combobox->setCurrentIndex(ui->theme_combobox->findData(UISettings::values.theme));
     ui->use_cpu_jit->setChecked(Settings::values.use_cpu_jit);
-    ui->enable_nfc->setChecked(Settings::values.enable_nfc);
 }
 
 void ConfigureGeneral::PopulateHotkeyList(const HotkeyRegistry& registry) {
@@ -48,5 +47,4 @@ void ConfigureGeneral::applyConfiguration() {
         ui->theme_combobox->itemData(ui->theme_combobox->currentIndex()).toString();
 
     Settings::values.use_cpu_jit = ui->use_cpu_jit->isChecked();
-    Settings::values.enable_nfc = ui->enable_nfc->isChecked();
 }
diff --git a/src/yuzu/configuration/configure_general.ui b/src/yuzu/configuration/configure_general.ui
index dff0ad5d0..01d1c0b8e 100644
--- a/src/yuzu/configuration/configure_general.ui
+++ b/src/yuzu/configuration/configure_general.ui
@@ -71,26 +71,6 @@
       </widget>
      </item>
      <item>
-      <widget class="QGroupBox" name="EmulationGroupBox">
-       <property name="title">
-        <string>Emulation</string>
-       </property>
-       <layout class="QHBoxLayout" name="EmulationHorizontalLayout">
-        <item>
-         <layout class="QVBoxLayout" name="EmulationVerticalLayout">
-          <item>
-           <widget class="QCheckBox" name="enable_nfc">
-            <property name="text">
-             <string>Enable NFC</string>
-            </property>
-           </widget>
-          </item>
-         </layout>
-        </item>
-       </layout>
-      </widget>
-     </item>
-     <item>
       <widget class="QGroupBox" name="theme_group_box">
        <property name="title">
         <string>Theme</string>
diff --git a/src/yuzu/debugger/profiler.cpp b/src/yuzu/debugger/profiler.cpp
index 8b30e0a85..86e03e46d 100644
--- a/src/yuzu/debugger/profiler.cpp
+++ b/src/yuzu/debugger/profiler.cpp
@@ -7,6 +7,7 @@
 #include <QMouseEvent>
 #include <QPainter>
 #include <QString>
+#include <QTimer>
 #include "common/common_types.h"
 #include "common/microprofile.h"
 #include "yuzu/debugger/profiler.h"
diff --git a/src/yuzu/debugger/profiler.h b/src/yuzu/debugger/profiler.h
index eae1e9e3c..8e69fdb06 100644
--- a/src/yuzu/debugger/profiler.h
+++ b/src/yuzu/debugger/profiler.h
@@ -4,10 +4,11 @@
 
 #pragma once
 
-#include <QAbstractItemModel>
-#include <QDockWidget>
-#include <QTimer>
-#include "common/microprofile.h"
+#include <QWidget>
+
+class QAction;
+class QHideEvent;
+class QShowEvent;
 
 class MicroProfileDialog : public QWidget {
     Q_OBJECT
diff --git a/src/yuzu/debugger/wait_tree.cpp b/src/yuzu/debugger/wait_tree.cpp
index 06ad74ffe..593bb681f 100644
--- a/src/yuzu/debugger/wait_tree.cpp
+++ b/src/yuzu/debugger/wait_tree.cpp
@@ -234,6 +234,9 @@ QString WaitTreeThread::GetText() const {
     case Kernel::ThreadStatus::WaitMutex:
         status = tr("waiting for mutex");
         break;
+    case Kernel::ThreadStatus::WaitCondVar:
+        status = tr("waiting for condition variable");
+        break;
     case Kernel::ThreadStatus::WaitArb:
         status = tr("waiting for address arbiter");
         break;
@@ -269,6 +272,7 @@ QColor WaitTreeThread::GetColor() const {
     case Kernel::ThreadStatus::WaitSynchAll:
     case Kernel::ThreadStatus::WaitSynchAny:
     case Kernel::ThreadStatus::WaitMutex:
+    case Kernel::ThreadStatus::WaitCondVar:
     case Kernel::ThreadStatus::WaitArb:
         return QColor(Qt::GlobalColor::red);
     case Kernel::ThreadStatus::Dormant:
diff --git a/src/yuzu/game_list.cpp b/src/yuzu/game_list.cpp
index c0e3c5fa9..4422a572b 100644
--- a/src/yuzu/game_list.cpp
+++ b/src/yuzu/game_list.cpp
@@ -329,6 +329,8 @@ void GameList::PopupContextMenu(const QPoint& menu_location) {
     QMenu context_menu;
     QAction* open_save_location = context_menu.addAction(tr("Open Save Data Location"));
     QAction* open_lfs_location = context_menu.addAction(tr("Open Mod Data Location"));
+    QAction* open_transferable_shader_cache =
+        context_menu.addAction(tr("Open Transferable Shader Cache"));
     context_menu.addSeparator();
     QAction* dump_romfs = context_menu.addAction(tr("Dump RomFS"));
     QAction* copy_tid = context_menu.addAction(tr("Copy Title ID to Clipboard"));
@@ -344,6 +346,8 @@ void GameList::PopupContextMenu(const QPoint& menu_location) {
             [&]() { emit OpenFolderRequested(program_id, GameListOpenTarget::SaveData); });
     connect(open_lfs_location, &QAction::triggered,
             [&]() { emit OpenFolderRequested(program_id, GameListOpenTarget::ModData); });
+    connect(open_transferable_shader_cache, &QAction::triggered,
+            [&]() { emit OpenTransferableShaderCacheRequested(program_id); });
     connect(dump_romfs, &QAction::triggered, [&]() { emit DumpRomFSRequested(program_id, path); });
     connect(copy_tid, &QAction::triggered, [&]() { emit CopyTIDRequested(program_id); });
     connect(navigate_to_gamedb_entry, &QAction::triggered,
diff --git a/src/yuzu/game_list.h b/src/yuzu/game_list.h
index b317eb2fc..8ea5cbaaa 100644
--- a/src/yuzu/game_list.h
+++ b/src/yuzu/game_list.h
@@ -66,6 +66,7 @@ signals:
     void GameChosen(QString game_path);
     void ShouldCancelWorker();
     void OpenFolderRequested(u64 program_id, GameListOpenTarget target);
+    void OpenTransferableShaderCacheRequested(u64 program_id);
     void DumpRomFSRequested(u64 program_id, const std::string& game_path);
     void CopyTIDRequested(u64 program_id);
     void NavigateToGamedbEntryRequested(u64 program_id,
diff --git a/src/yuzu/main.cpp b/src/yuzu/main.cpp
index 41ba3c4c6..2b9db69a3 100644
--- a/src/yuzu/main.cpp
+++ b/src/yuzu/main.cpp
@@ -37,14 +37,20 @@ static FileSys::VirtualFile VfsDirectoryCreateFileWrapper(const FileSys::Virtual
 #include <glad/glad.h>
 
 #define QT_NO_OPENGL
+#include <QClipboard>
+#include <QDesktopServices>
 #include <QDesktopWidget>
 #include <QDialogButtonBox>
 #include <QFile>
 #include <QFileDialog>
+#include <QInputDialog>
 #include <QMessageBox>
+#include <QProgressBar>
+#include <QProgressDialog>
+#include <QShortcut>
+#include <QStatusBar>
 #include <QtConcurrent/QtConcurrent>
-#include <QtGui>
-#include <QtWidgets>
+
 #include <fmt/format.h>
 #include "common/common_paths.h"
 #include "common/detached_tasks.h"
@@ -55,11 +61,9 @@ static FileSys::VirtualFile VfsDirectoryCreateFileWrapper(const FileSys::Virtual
 #include "common/microprofile.h"
 #include "common/scm_rev.h"
 #include "common/scope_exit.h"
-#include "common/string_util.h"
 #include "common/telemetry.h"
 #include "core/core.h"
 #include "core/crypto/key_manager.h"
-#include "core/file_sys/bis_factory.h"
 #include "core/file_sys/card_image.h"
 #include "core/file_sys/content_archive.h"
 #include "core/file_sys/control_metadata.h"
@@ -71,7 +75,6 @@ static FileSys::VirtualFile VfsDirectoryCreateFileWrapper(const FileSys::Virtual
 #include "core/frontend/applets/software_keyboard.h"
 #include "core/hle/kernel/process.h"
 #include "core/hle/service/filesystem/filesystem.h"
-#include "core/hle/service/filesystem/fsp_ldr.h"
 #include "core/hle/service/nfp/nfp.h"
 #include "core/hle/service/sm/sm.h"
 #include "core/loader/loader.h"
@@ -648,6 +651,8 @@ void GMainWindow::RestoreUIState() {
 void GMainWindow::ConnectWidgetEvents() {
     connect(game_list, &GameList::GameChosen, this, &GMainWindow::OnGameListLoadFile);
     connect(game_list, &GameList::OpenFolderRequested, this, &GMainWindow::OnGameListOpenFolder);
+    connect(game_list, &GameList::OpenTransferableShaderCacheRequested, this,
+            &GMainWindow::OnTransferableShaderCacheOpenFile);
     connect(game_list, &GameList::DumpRomFSRequested, this, &GMainWindow::OnGameListDumpRomFS);
     connect(game_list, &GameList::CopyTIDRequested, this, &GMainWindow::OnGameListCopyTID);
     connect(game_list, &GameList::NavigateToGamedbEntryRequested, this,
@@ -1082,6 +1087,39 @@ void GMainWindow::OnGameListOpenFolder(u64 program_id, GameListOpenTarget target
     QDesktopServices::openUrl(QUrl::fromLocalFile(qpath));
 }
 
+void GMainWindow::OnTransferableShaderCacheOpenFile(u64 program_id) {
+    ASSERT(program_id != 0);
+
+    const QString tranferable_shader_cache_folder_path =
+        QString::fromStdString(FileUtil::GetUserPath(FileUtil::UserPath::ShaderDir)) + "opengl" +
+        DIR_SEP + "transferable";
+
+    const QString transferable_shader_cache_file_path =
+        tranferable_shader_cache_folder_path + DIR_SEP +
+        QString::fromStdString(fmt::format("{:016X}.bin", program_id));
+
+    if (!QFile::exists(transferable_shader_cache_file_path)) {
+        QMessageBox::warning(this, tr("Error Opening Transferable Shader Cache"),
+                             tr("A shader cache for this title does not exist."));
+        return;
+    }
+
+    // Windows supports opening a folder with selecting a specified file in explorer. On every other
+    // OS we just open the transferable shader cache folder without preselecting the transferable
+    // shader cache file for the selected game.
+#if defined(Q_OS_WIN)
+    const QString explorer = QStringLiteral("explorer");
+    QStringList param;
+    if (!QFileInfo(transferable_shader_cache_file_path).isDir()) {
+        param << QStringLiteral("/select,");
+    }
+    param << QDir::toNativeSeparators(transferable_shader_cache_file_path);
+    QProcess::startDetached(explorer, param);
+#else
+    QDesktopServices::openUrl(QUrl::fromLocalFile(tranferable_shader_cache_folder_path));
+#endif
+}
+
 static std::size_t CalculateRomFSEntrySize(const FileSys::VirtualDir& dir, bool full) {
     std::size_t out = 0;
 
diff --git a/src/yuzu/main.h b/src/yuzu/main.h
index e07c892cf..7f3aa998e 100644
--- a/src/yuzu/main.h
+++ b/src/yuzu/main.h
@@ -176,6 +176,7 @@ private slots:
     /// Called whenever a user selects a game in the game list widget.
     void OnGameListLoadFile(QString game_path);
     void OnGameListOpenFolder(u64 program_id, GameListOpenTarget target);
+    void OnTransferableShaderCacheOpenFile(u64 program_id);
     void OnGameListDumpRomFS(u64 program_id, const std::string& game_path);
     void OnGameListCopyTID(u64 program_id);
     void OnGameListNavigateToGamedbEntry(u64 program_id,
diff --git a/src/yuzu_cmd/config.cpp b/src/yuzu_cmd/config.cpp
index 32e78049c..f24cc77fe 100644
--- a/src/yuzu_cmd/config.cpp
+++ b/src/yuzu_cmd/config.cpp
@@ -319,7 +319,6 @@ void Config::ReadValues() {
 
     // System
     Settings::values.use_docked_mode = sdl2_config->GetBoolean("System", "use_docked_mode", false);
-    Settings::values.enable_nfc = sdl2_config->GetBoolean("System", "enable_nfc", true);
     const auto size = sdl2_config->GetInteger("System", "users_size", 0);
 
     Settings::values.current_user = std::clamp<int>(