154 files changed, 2689 insertions, 1093 deletions

diff --git a/src/common/address_space.h b/src/common/address_space.h
index 9222b2fdc..8683c23c3 100644
--- a/src/common/address_space.h
+++ b/src/common/address_space.h
@@ -12,7 +12,8 @@
 
 namespace Common {
 template <typename VaType, size_t AddressSpaceBits>
-concept AddressSpaceValid = std::is_unsigned_v<VaType> && sizeof(VaType) * 8 >= AddressSpaceBits;
+concept AddressSpaceValid = std::is_unsigned_v<VaType> && sizeof(VaType) * 8 >=
+AddressSpaceBits;
 
 struct EmptyStruct {};
 
@@ -21,7 +22,7 @@ struct EmptyStruct {};
  */
 template <typename VaType, VaType UnmappedVa, typename PaType, PaType UnmappedPa,
           bool PaContigSplit, size_t AddressSpaceBits, typename ExtraBlockInfo = EmptyStruct>
-requires AddressSpaceValid<VaType, AddressSpaceBits>
+    requires AddressSpaceValid<VaType, AddressSpaceBits>
 class FlatAddressSpaceMap {
 public:
     /// The maximum VA that this AS can technically reach
@@ -109,7 +110,7 @@ private:
  * initial, fast linear pass and a subsequent slower pass that iterates until it finds a free block
  */
 template <typename VaType, VaType UnmappedVa, size_t AddressSpaceBits>
-requires AddressSpaceValid<VaType, AddressSpaceBits>
+    requires AddressSpaceValid<VaType, AddressSpaceBits>
 class FlatAllocator
     : public FlatAddressSpaceMap<VaType, UnmappedVa, bool, false, false, AddressSpaceBits> {
 private:
diff --git a/src/common/alignment.h b/src/common/alignment.h
index 7e897334b..fa715d497 100644
--- a/src/common/alignment.h
+++ b/src/common/alignment.h
@@ -10,7 +10,7 @@
 namespace Common {
 
 template <typename T>
-requires std::is_unsigned_v<T>
+    requires std::is_unsigned_v<T>
 [[nodiscard]] constexpr T AlignUp(T value, size_t size) {
     auto mod{static_cast<T>(value % size)};
     value -= mod;
@@ -18,31 +18,31 @@ requires std::is_unsigned_v<T>
 }
 
 template <typename T>
-requires std::is_unsigned_v<T>
+    requires std::is_unsigned_v<T>
 [[nodiscard]] constexpr T AlignUpLog2(T value, size_t align_log2) {
     return static_cast<T>((value + ((1ULL << align_log2) - 1)) >> align_log2 << align_log2);
 }
 
 template <typename T>
-requires std::is_unsigned_v<T>
+    requires std::is_unsigned_v<T>
 [[nodiscard]] constexpr T AlignDown(T value, size_t size) {
     return static_cast<T>(value - value % size);
 }
 
 template <typename T>
-requires std::is_unsigned_v<T>
+    requires std::is_unsigned_v<T>
 [[nodiscard]] constexpr bool Is4KBAligned(T value) {
     return (value & 0xFFF) == 0;
 }
 
 template <typename T>
-requires std::is_unsigned_v<T>
+    requires std::is_unsigned_v<T>
 [[nodiscard]] constexpr bool IsWordAligned(T value) {
     return (value & 0b11) == 0;
 }
 
 template <typename T>
-requires std::is_integral_v<T>
+    requires std::is_integral_v<T>
 [[nodiscard]] constexpr bool IsAligned(T value, size_t alignment) {
     using U = typename std::make_unsigned_t<T>;
     const U mask = static_cast<U>(alignment - 1);
@@ -50,7 +50,7 @@ requires std::is_integral_v<T>
 }
 
 template <typename T, typename U>
-requires std::is_integral_v<T>
+    requires std::is_integral_v<T>
 [[nodiscard]] constexpr T DivideUp(T x, U y) {
     return (x + (y - 1)) / y;
 }
@@ -73,11 +73,11 @@ public:
     constexpr AlignmentAllocator(const AlignmentAllocator<T2, Align>&) noexcept {}
 
     [[nodiscard]] T* allocate(size_type n) {
-        return static_cast<T*>(::operator new (n * sizeof(T), std::align_val_t{Align}));
+        return static_cast<T*>(::operator new(n * sizeof(T), std::align_val_t{Align}));
     }
 
     void deallocate(T* p, size_type n) {
-        ::operator delete (p, n * sizeof(T), std::align_val_t{Align});
+        ::operator delete(p, n * sizeof(T), std::align_val_t{Align});
     }
 
     template <typename T2>
diff --git a/src/common/atomic_helpers.h b/src/common/atomic_helpers.h
index aef3b66a4..d997f10ba 100644
--- a/src/common/atomic_helpers.h
+++ b/src/common/atomic_helpers.h
@@ -75,7 +75,7 @@ extern "C" void AnnotateHappensAfter(const char*, int, void*);
 #if defined(AE_VCPP) || defined(AE_ICC)
 #define AE_FORCEINLINE __forceinline
 #elif defined(AE_GCC)
-//#define AE_FORCEINLINE __attribute__((always_inline))
+// #define AE_FORCEINLINE __attribute__((always_inline))
 #define AE_FORCEINLINE inline
 #else
 #define AE_FORCEINLINE inline
diff --git a/src/common/bit_util.h b/src/common/bit_util.h
index e4e6287f3..13368b439 100644
--- a/src/common/bit_util.h
+++ b/src/common/bit_util.h
@@ -45,19 +45,19 @@ template <typename T>
 }
 
 template <typename T>
-requires std::is_unsigned_v<T>
+    requires std::is_unsigned_v<T>
 [[nodiscard]] constexpr bool IsPow2(T value) {
     return std::has_single_bit(value);
 }
 
 template <typename T>
-requires std::is_integral_v<T>
+    requires std::is_integral_v<T>
 [[nodiscard]] T NextPow2(T value) {
     return static_cast<T>(1ULL << ((8U * sizeof(T)) - std::countl_zero(value - 1U)));
 }
 
 template <size_t bit_index, typename T>
-requires std::is_integral_v<T>
+    requires std::is_integral_v<T>
 [[nodiscard]] constexpr bool Bit(const T value) {
     static_assert(bit_index < BitSize<T>(), "bit_index must be smaller than size of T");
     return ((value >> bit_index) & T(1)) == T(1);
diff --git a/src/common/concepts.h b/src/common/concepts.h
index a9acff3e7..61df1d32a 100644
--- a/src/common/concepts.h
+++ b/src/common/concepts.h
@@ -16,9 +16,9 @@ concept IsContiguousContainer = std::contiguous_iterator<typename T::iterator>;
 //       is available on all supported platforms.
 template <typename Derived, typename Base>
 concept DerivedFrom = requires {
-    std::is_base_of_v<Base, Derived>;
-    std::is_convertible_v<const volatile Derived*, const volatile Base*>;
-};
+                          std::is_base_of_v<Base, Derived>;
+                          std::is_convertible_v<const volatile Derived*, const volatile Base*>;
+                      };
 
 // TODO: Replace with std::convertible_to when libc++ implements it.
 template <typename From, typename To>
diff --git a/src/common/div_ceil.h b/src/common/div_ceil.h
index eebc279c2..c12477d42 100644
--- a/src/common/div_ceil.h
+++ b/src/common/div_ceil.h
@@ -10,14 +10,14 @@ namespace Common {
 
 /// Ceiled integer division.
 template <typename N, typename D>
-requires std::is_integral_v<N> && std::is_unsigned_v<D>
+    requires std::is_integral_v<N> && std::is_unsigned_v<D>
 [[nodiscard]] constexpr N DivCeil(N number, D divisor) {
     return static_cast<N>((static_cast<D>(number) + divisor - 1) / divisor);
 }
 
 /// Ceiled integer division with logarithmic divisor in base 2
 template <typename N, typename D>
-requires std::is_integral_v<N> && std::is_unsigned_v<D>
+    requires std::is_integral_v<N> && std::is_unsigned_v<D>
 [[nodiscard]] constexpr N DivCeilLog2(N value, D alignment_log2) {
     return static_cast<N>((static_cast<D>(value) + (D(1) << alignment_log2) - 1) >> alignment_log2);
 }
diff --git a/src/common/expected.h b/src/common/expected.h
index 6e6c86ee7..5fccfbcbd 100644
--- a/src/common/expected.h
+++ b/src/common/expected.h
@@ -64,7 +64,7 @@ struct no_init_t {
  * Additionally, this requires E to be trivially destructible
  */
 template <typename T, typename E, bool = std::is_trivially_destructible_v<T>>
-requires std::is_trivially_destructible_v<E>
+    requires std::is_trivially_destructible_v<E>
 struct expected_storage_base {
     constexpr expected_storage_base() : m_val{T{}}, m_has_val{true} {}
 
@@ -111,7 +111,7 @@ struct expected_storage_base {
  * Additionally, this requires E to be trivially destructible
  */
 template <typename T, typename E>
-requires std::is_trivially_destructible_v<E>
+    requires std::is_trivially_destructible_v<E>
 struct expected_storage_base<T, E, true> {
     constexpr expected_storage_base() : m_val{T{}}, m_has_val{true} {}
 
@@ -251,7 +251,7 @@ struct expected_operations_base : expected_storage_base<T, E> {
  * Additionally, this requires E to be trivially copy constructible
  */
 template <typename T, typename E, bool = std::is_trivially_copy_constructible_v<T>>
-requires std::is_trivially_copy_constructible_v<E>
+    requires std::is_trivially_copy_constructible_v<E>
 struct expected_copy_base : expected_operations_base<T, E> {
     using expected_operations_base<T, E>::expected_operations_base;
 };
@@ -261,7 +261,7 @@ struct expected_copy_base : expected_operations_base<T, E> {
  * Additionally, this requires E to be trivially copy constructible
  */
 template <typename T, typename E>
-requires std::is_trivially_copy_constructible_v<E>
+    requires std::is_trivially_copy_constructible_v<E>
 struct expected_copy_base<T, E, false> : expected_operations_base<T, E> {
     using expected_operations_base<T, E>::expected_operations_base;
 
@@ -289,7 +289,7 @@ struct expected_copy_base<T, E, false> : expected_operations_base<T, E> {
  * Additionally, this requires E to be trivially move constructible
  */
 template <typename T, typename E, bool = std::is_trivially_move_constructible_v<T>>
-requires std::is_trivially_move_constructible_v<E>
+    requires std::is_trivially_move_constructible_v<E>
 struct expected_move_base : expected_copy_base<T, E> {
     using expected_copy_base<T, E>::expected_copy_base;
 };
@@ -299,7 +299,7 @@ struct expected_move_base : expected_copy_base<T, E> {
  * Additionally, this requires E to be trivially move constructible
  */
 template <typename T, typename E>
-requires std::is_trivially_move_constructible_v<E>
+    requires std::is_trivially_move_constructible_v<E>
 struct expected_move_base<T, E, false> : expected_copy_base<T, E> {
     using expected_copy_base<T, E>::expected_copy_base;
 
@@ -330,9 +330,9 @@ template <typename T, typename E,
           bool = std::conjunction_v<std::is_trivially_copy_assignable<T>,
                                     std::is_trivially_copy_constructible<T>,
                                     std::is_trivially_destructible<T>>>
-requires std::conjunction_v<std::is_trivially_copy_assignable<E>,
-                            std::is_trivially_copy_constructible<E>,
-                            std::is_trivially_destructible<E>>
+    requires std::conjunction_v<std::is_trivially_copy_assignable<E>,
+                                std::is_trivially_copy_constructible<E>,
+                                std::is_trivially_destructible<E>>
 struct expected_copy_assign_base : expected_move_base<T, E> {
     using expected_move_base<T, E>::expected_move_base;
 };
@@ -342,9 +342,9 @@ struct expected_copy_assign_base : expected_move_base<T, E> {
  * Additionally, this requires E to be trivially copy assignable
  */
 template <typename T, typename E>
-requires std::conjunction_v<std::is_trivially_copy_assignable<E>,
-                            std::is_trivially_copy_constructible<E>,
-                            std::is_trivially_destructible<E>>
+    requires std::conjunction_v<std::is_trivially_copy_assignable<E>,
+                                std::is_trivially_copy_constructible<E>,
+                                std::is_trivially_destructible<E>>
 struct expected_copy_assign_base<T, E, false> : expected_move_base<T, E> {
     using expected_move_base<T, E>::expected_move_base;
 
@@ -371,9 +371,9 @@ template <typename T, typename E,
           bool = std::conjunction_v<std::is_trivially_move_assignable<T>,
                                     std::is_trivially_move_constructible<T>,
                                     std::is_trivially_destructible<T>>>
-requires std::conjunction_v<std::is_trivially_move_assignable<E>,
-                            std::is_trivially_move_constructible<E>,
-                            std::is_trivially_destructible<E>>
+    requires std::conjunction_v<std::is_trivially_move_assignable<E>,
+                                std::is_trivially_move_constructible<E>,
+                                std::is_trivially_destructible<E>>
 struct expected_move_assign_base : expected_copy_assign_base<T, E> {
     using expected_copy_assign_base<T, E>::expected_copy_assign_base;
 };
@@ -383,9 +383,9 @@ struct expected_move_assign_base : expected_copy_assign_base<T, E> {
  * Additionally, this requires E to be trivially move assignable
  */
 template <typename T, typename E>
-requires std::conjunction_v<std::is_trivially_move_assignable<E>,
-                            std::is_trivially_move_constructible<E>,
-                            std::is_trivially_destructible<E>>
+    requires std::conjunction_v<std::is_trivially_move_assignable<E>,
+                                std::is_trivially_move_constructible<E>,
+                                std::is_trivially_destructible<E>>
 struct expected_move_assign_base<T, E, false> : expected_copy_assign_base<T, E> {
     using expected_copy_assign_base<T, E>::expected_copy_assign_base;
 
@@ -412,7 +412,7 @@ struct expected_move_assign_base<T, E, false> : expected_copy_assign_base<T, E>
  */
 template <typename T, typename E, bool EnableCopy = std::is_copy_constructible_v<T>,
           bool EnableMove = std::is_move_constructible_v<T>>
-requires std::conjunction_v<std::is_copy_constructible<E>, std::is_move_constructible<E>>
+    requires std::conjunction_v<std::is_copy_constructible<E>, std::is_move_constructible<E>>
 struct expected_delete_ctor_base {
     expected_delete_ctor_base() = default;
     expected_delete_ctor_base(const expected_delete_ctor_base&) = default;
@@ -422,7 +422,7 @@ struct expected_delete_ctor_base {
 };
 
 template <typename T, typename E>
-requires std::conjunction_v<std::is_copy_constructible<E>, std::is_move_constructible<E>>
+    requires std::conjunction_v<std::is_copy_constructible<E>, std::is_move_constructible<E>>
 struct expected_delete_ctor_base<T, E, true, false> {
     expected_delete_ctor_base() = default;
     expected_delete_ctor_base(const expected_delete_ctor_base&) = default;
@@ -432,7 +432,7 @@ struct expected_delete_ctor_base<T, E, true, false> {
 };
 
 template <typename T, typename E>
-requires std::conjunction_v<std::is_copy_constructible<E>, std::is_move_constructible<E>>
+    requires std::conjunction_v<std::is_copy_constructible<E>, std::is_move_constructible<E>>
 struct expected_delete_ctor_base<T, E, false, true> {
     expected_delete_ctor_base() = default;
     expected_delete_ctor_base(const expected_delete_ctor_base&) = delete;
@@ -442,7 +442,7 @@ struct expected_delete_ctor_base<T, E, false, true> {
 };
 
 template <typename T, typename E>
-requires std::conjunction_v<std::is_copy_constructible<E>, std::is_move_constructible<E>>
+    requires std::conjunction_v<std::is_copy_constructible<E>, std::is_move_constructible<E>>
 struct expected_delete_ctor_base<T, E, false, false> {
     expected_delete_ctor_base() = default;
     expected_delete_ctor_base(const expected_delete_ctor_base&) = delete;
@@ -460,8 +460,8 @@ template <
     typename T, typename E,
     bool EnableCopy = std::conjunction_v<std::is_copy_constructible<T>, std::is_copy_assignable<T>>,
     bool EnableMove = std::conjunction_v<std::is_move_constructible<T>, std::is_move_assignable<T>>>
-requires std::conjunction_v<std::is_copy_constructible<E>, std::is_move_constructible<E>,
-                            std::is_copy_assignable<E>, std::is_move_assignable<E>>
+    requires std::conjunction_v<std::is_copy_constructible<E>, std::is_move_constructible<E>,
+                                std::is_copy_assignable<E>, std::is_move_assignable<E>>
 struct expected_delete_assign_base {
     expected_delete_assign_base() = default;
     expected_delete_assign_base(const expected_delete_assign_base&) = default;
@@ -471,8 +471,8 @@ struct expected_delete_assign_base {
 };
 
 template <typename T, typename E>
-requires std::conjunction_v<std::is_copy_constructible<E>, std::is_move_constructible<E>,
-                            std::is_copy_assignable<E>, std::is_move_assignable<E>>
+    requires std::conjunction_v<std::is_copy_constructible<E>, std::is_move_constructible<E>,
+                                std::is_copy_assignable<E>, std::is_move_assignable<E>>
 struct expected_delete_assign_base<T, E, true, false> {
     expected_delete_assign_base() = default;
     expected_delete_assign_base(const expected_delete_assign_base&) = default;
@@ -482,8 +482,8 @@ struct expected_delete_assign_base<T, E, true, false> {
 };
 
 template <typename T, typename E>
-requires std::conjunction_v<std::is_copy_constructible<E>, std::is_move_constructible<E>,
-                            std::is_copy_assignable<E>, std::is_move_assignable<E>>
+    requires std::conjunction_v<std::is_copy_constructible<E>, std::is_move_constructible<E>,
+                                std::is_copy_assignable<E>, std::is_move_assignable<E>>
 struct expected_delete_assign_base<T, E, false, true> {
     expected_delete_assign_base() = default;
     expected_delete_assign_base(const expected_delete_assign_base&) = default;
@@ -493,8 +493,8 @@ struct expected_delete_assign_base<T, E, false, true> {
 };
 
 template <typename T, typename E>
-requires std::conjunction_v<std::is_copy_constructible<E>, std::is_move_constructible<E>,
-                            std::is_copy_assignable<E>, std::is_move_assignable<E>>
+    requires std::conjunction_v<std::is_copy_constructible<E>, std::is_move_constructible<E>,
+                                std::is_copy_assignable<E>, std::is_move_assignable<E>>
 struct expected_delete_assign_base<T, E, false, false> {
     expected_delete_assign_base() = default;
     expected_delete_assign_base(const expected_delete_assign_base&) = default;
diff --git a/src/common/input.h b/src/common/input.h
index d61cd7ca8..b5748a6c8 100644
--- a/src/common/input.h
+++ b/src/common/input.h
@@ -130,6 +130,8 @@ struct ButtonStatus {
     bool inverted{};
     // Press once to activate, press again to release
     bool toggle{};
+    // Spams the button when active
+    bool turbo{};
     // Internal lock for the toggle status
     bool locked{};
 };
diff --git a/src/common/intrusive_red_black_tree.h b/src/common/intrusive_red_black_tree.h
index 93046615e..5f6b34e82 100644
--- a/src/common/intrusive_red_black_tree.h
+++ b/src/common/intrusive_red_black_tree.h
@@ -242,19 +242,21 @@ public:
 
 template <typename T>
 concept HasRedBlackKeyType = requires {
-    { std::is_same<typename T::RedBlackKeyType, void>::value } -> std::convertible_to<bool>;
-};
+                                 {
+                                     std::is_same<typename T::RedBlackKeyType, void>::value
+                                     } -> std::convertible_to<bool>;
+                             };
 
 namespace impl {
 
-    template <typename T, typename Default>
-    consteval auto* GetRedBlackKeyType() {
-        if constexpr (HasRedBlackKeyType<T>) {
-            return static_cast<typename T::RedBlackKeyType*>(nullptr);
-        } else {
-            return static_cast<Default*>(nullptr);
-        }
+template <typename T, typename Default>
+consteval auto* GetRedBlackKeyType() {
+    if constexpr (HasRedBlackKeyType<T>) {
+        return static_cast<typename T::RedBlackKeyType*>(nullptr);
+    } else {
+        return static_cast<Default*>(nullptr);
     }
+}
 
 } // namespace impl
 
diff --git a/src/common/make_unique_for_overwrite.h b/src/common/make_unique_for_overwrite.h
index c7413cf51..17f81bba4 100644
--- a/src/common/make_unique_for_overwrite.h
+++ b/src/common/make_unique_for_overwrite.h
@@ -9,17 +9,19 @@
 namespace Common {
 
 template <class T>
-requires(!std::is_array_v<T>) std::unique_ptr<T> make_unique_for_overwrite() {
+    requires(!std::is_array_v<T>)
+std::unique_ptr<T> make_unique_for_overwrite() {
     return std::unique_ptr<T>(new T);
 }
 
 template <class T>
-requires std::is_unbounded_array_v<T> std::unique_ptr<T> make_unique_for_overwrite(std::size_t n) {
+    requires std::is_unbounded_array_v<T>
+std::unique_ptr<T> make_unique_for_overwrite(std::size_t n) {
     return std::unique_ptr<T>(new std::remove_extent_t<T>[n]);
 }
 
 template <class T, class... Args>
-requires std::is_bounded_array_v<T>
+    requires std::is_bounded_array_v<T>
 void make_unique_for_overwrite(Args&&...) = delete;
 
 } // namespace Common
diff --git a/src/common/polyfill_ranges.h b/src/common/polyfill_ranges.h
index ca44bfaef..512dbcbcb 100644
--- a/src/common/polyfill_ranges.h
+++ b/src/common/polyfill_ranges.h
@@ -18,9 +18,9 @@ namespace ranges {
 
 template <typename T>
 concept range = requires(T& t) {
-    begin(t);
-    end(t);
-};
+                    begin(t);
+                    end(t);
+                };
 
 template <typename T>
 concept input_range = range<T>;
@@ -421,7 +421,7 @@ struct generate_fn {
     }
 
     template <typename R, std::copy_constructible F>
-    requires std::invocable<F&> && ranges::output_range<R>
+        requires std::invocable<F&> && ranges::output_range<R>
     constexpr ranges::iterator_t<R> operator()(R&& r, F gen) const {
         return operator()(ranges::begin(r), ranges::end(r), std::move(gen));
     }
diff --git a/src/common/polyfill_thread.h b/src/common/polyfill_thread.h
index b2c929d2f..b5ef055db 100644
--- a/src/common/polyfill_thread.h
+++ b/src/common/polyfill_thread.h
@@ -41,17 +41,18 @@ bool StoppableTimedWait(std::stop_token token, const std::chrono::duration<Rep,
 #include <chrono>
 #include <condition_variable>
 #include <functional>
-#include <list>
+#include <map>
 #include <memory>
 #include <mutex>
 #include <optional>
 #include <thread>
 #include <type_traits>
+#include <utility>
 
 namespace std {
 namespace polyfill {
 
-using stop_state_callbacks = list<function<void()>>;
+using stop_state_callback = size_t;
 
 class stop_state {
 public:
@@ -59,61 +60,69 @@ public:
     ~stop_state() = default;
 
     bool request_stop() {
-        stop_state_callbacks callbacks;
+        unique_lock lk{m_lock};
 
-        {
-            scoped_lock lk{m_lock};
+        if (m_stop_requested) {
+            // Already set, nothing to do.
+            return false;
+        }
 
-            if (m_stop_requested.load()) {
-                // Already set, nothing to do
-                return false;
-            }
+        // Mark stop requested.
+        m_stop_requested = true;
 
-            // Set as requested
-            m_stop_requested = true;
+        while (!m_callbacks.empty()) {
+            // Get an iterator to the first element.
+            const auto it = m_callbacks.begin();
 
-            // Copy callback list
-            callbacks = m_callbacks;
-        }
+            // Move the callback function out of the map.
+            function<void()> f;
+            swap(it->second, f);
+
+            // Erase the now-empty map element.
+            m_callbacks.erase(it);
 
-        for (auto callback : callbacks) {
-            callback();
+            // Run the callback.
+            if (f) {
+                f();
+            }
         }
 
         return true;
     }
 
     bool stop_requested() const {
-        return m_stop_requested.load();
+        unique_lock lk{m_lock};
+        return m_stop_requested;
     }
 
-    stop_state_callbacks::const_iterator insert_callback(function<void()> f) {
-        stop_state_callbacks::const_iterator ret{};
-        bool should_run{};
-
-        {
-            scoped_lock lk{m_lock};
-            should_run = m_stop_requested.load();
-            m_callbacks.push_front(f);
-            ret = m_callbacks.begin();
-        }
+    stop_state_callback insert_callback(function<void()> f) {
+        unique_lock lk{m_lock};
 
-        if (should_run) {
-            f();
+        if (m_stop_requested) {
+            // Stop already requested. Don't insert anything,
+            // just run the callback synchronously.
+            if (f) {
+                f();
+            }
+            return 0;
         }
 
+        // Insert the callback.
+        stop_state_callback ret = ++m_next_callback;
+        m_callbacks.emplace(ret, move(f));
         return ret;
     }
 
-    void remove_callback(stop_state_callbacks::const_iterator it) {
-        scoped_lock lk{m_lock};
-        m_callbacks.erase(it);
+    void remove_callback(stop_state_callback cb) {
+        unique_lock lk{m_lock};
+        m_callbacks.erase(cb);
     }
 
 private:
-    mutex m_lock;
-    atomic<bool> m_stop_requested;
-    stop_state_callbacks m_callbacks;
+    mutable recursive_mutex m_lock;
+    map<stop_state_callback, function<void()>> m_callbacks;
+    stop_state_callback m_next_callback{0};
+    bool m_stop_requested{false};
 };
 
 } // namespace polyfill
@@ -204,7 +213,7 @@ public:
     using callback_type = Callback;
 
     template <typename C>
-    requires constructible_from<Callback, C>
+        requires constructible_from<Callback, C>
     explicit stop_callback(const stop_token& st,
                            C&& cb) noexcept(is_nothrow_constructible_v<Callback, C>)
         : m_stop_state(st.m_stop_state) {
@@ -213,7 +222,7 @@ public:
         }
     }
     template <typename C>
-    requires constructible_from<Callback, C>
+        requires constructible_from<Callback, C>
     explicit stop_callback(stop_token&& st,
                            C&& cb) noexcept(is_nothrow_constructible_v<Callback, C>)
         : m_stop_state(move(st.m_stop_state)) {
@@ -223,7 +232,7 @@ public:
     }
     ~stop_callback() {
         if (m_stop_state && m_callback) {
-            m_stop_state->remove_callback(*m_callback);
+            m_stop_state->remove_callback(m_callback);
         }
     }
 
@@ -234,7 +243,7 @@ public:
 
 private:
     shared_ptr<polyfill::stop_state> m_stop_state;
-    optional<polyfill::stop_state_callbacks::const_iterator> m_callback;
+    polyfill::stop_state_callback m_callback;
 };
 
 template <typename Callback>
diff --git a/src/common/settings.h b/src/common/settings.h
index 4b4da4da2..64db66f37 100644
--- a/src/common/settings.h
+++ b/src/common/settings.h
@@ -131,7 +131,8 @@ public:
      * @param default_val Intial value of the setting, and default value of the setting
      * @param name Label for the setting
      */
-    explicit Setting(const Type& default_val, const std::string& name) requires(!ranged)
+    explicit Setting(const Type& default_val, const std::string& name)
+        requires(!ranged)
         : value{default_val}, default_value{default_val}, label{name} {}
     virtual ~Setting() = default;
 
@@ -144,7 +145,8 @@ public:
      * @param name Label for the setting
      */
     explicit Setting(const Type& default_val, const Type& min_val, const Type& max_val,
-                     const std::string& name) requires(ranged)
+                     const std::string& name)
+        requires(ranged)
         : value{default_val},
           default_value{default_val}, maximum{max_val}, minimum{min_val}, label{name} {}
 
@@ -232,7 +234,8 @@ public:
      * @param default_val Intial value of the setting, and default value of the setting
      * @param name Label for the setting
      */
-    explicit SwitchableSetting(const Type& default_val, const std::string& name) requires(!ranged)
+    explicit SwitchableSetting(const Type& default_val, const std::string& name)
+        requires(!ranged)
         : Setting<Type>{default_val, name} {}
     virtual ~SwitchableSetting() = default;
 
@@ -245,7 +248,8 @@ public:
      * @param name Label for the setting
      */
     explicit SwitchableSetting(const Type& default_val, const Type& min_val, const Type& max_val,
-                               const std::string& name) requires(ranged)
+                               const std::string& name)
+        requires(ranged)
         : Setting<Type, true>{default_val, min_val, max_val, name} {}
 
     /**
diff --git a/src/common/string_util.cpp b/src/common/string_util.cpp
index b26db4796..e0b6180c5 100644
--- a/src/common/string_util.cpp
+++ b/src/common/string_util.cpp
@@ -30,7 +30,7 @@ std::string ToUpper(std::string str) {
     return str;
 }
 
-std::string StringFromBuffer(const std::vector<u8>& data) {
+std::string StringFromBuffer(std::span<const u8> data) {
     return std::string(data.begin(), std::find(data.begin(), data.end(), '\0'));
 }
 
diff --git a/src/common/string_util.h b/src/common/string_util.h
index ce18a33cf..f8aecc875 100644
--- a/src/common/string_util.h
+++ b/src/common/string_util.h
@@ -5,6 +5,7 @@
 #pragma once
 
 #include <cstddef>
+#include <span>
 #include <string>
 #include <vector>
 #include "common/common_types.h"
@@ -17,7 +18,7 @@ namespace Common {
 /// Make a string uppercase
 [[nodiscard]] std::string ToUpper(std::string str);
 
-[[nodiscard]] std::string StringFromBuffer(const std::vector<u8>& data);
+[[nodiscard]] std::string StringFromBuffer(std::span<const u8> data);
 
 [[nodiscard]] std::string StripSpaces(const std::string& s);
 [[nodiscard]] std::string StripQuotes(const std::string& s);
diff --git a/src/common/tree.h b/src/common/tree.h
index f77859209..f4fc43de3 100644
--- a/src/common/tree.h
+++ b/src/common/tree.h
@@ -103,12 +103,12 @@ concept IsRBEntry = CheckRBEntry<T>::value;
 
 template <typename T>
 concept HasRBEntry = requires(T& t, const T& ct) {
-    { t.GetRBEntry() } -> std::same_as<RBEntry<T>&>;
-    { ct.GetRBEntry() } -> std::same_as<const RBEntry<T>&>;
-};
+                         { t.GetRBEntry() } -> std::same_as<RBEntry<T>&>;
+                         { ct.GetRBEntry() } -> std::same_as<const RBEntry<T>&>;
+                     };
 
 template <typename T>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 class RBHead {
 private:
     T* m_rbh_root = nullptr;
@@ -130,90 +130,90 @@ public:
 };
 
 template <typename T>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 [[nodiscard]] constexpr RBEntry<T>& RB_ENTRY(T* t) {
     return t->GetRBEntry();
 }
 template <typename T>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 [[nodiscard]] constexpr const RBEntry<T>& RB_ENTRY(const T* t) {
     return t->GetRBEntry();
 }
 
 template <typename T>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 [[nodiscard]] constexpr T* RB_LEFT(T* t) {
     return RB_ENTRY(t).Left();
 }
 template <typename T>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 [[nodiscard]] constexpr const T* RB_LEFT(const T* t) {
     return RB_ENTRY(t).Left();
 }
 
 template <typename T>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 [[nodiscard]] constexpr T* RB_RIGHT(T* t) {
     return RB_ENTRY(t).Right();
 }
 template <typename T>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 [[nodiscard]] constexpr const T* RB_RIGHT(const T* t) {
     return RB_ENTRY(t).Right();
 }
 
 template <typename T>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 [[nodiscard]] constexpr T* RB_PARENT(T* t) {
     return RB_ENTRY(t).Parent();
 }
 template <typename T>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 [[nodiscard]] constexpr const T* RB_PARENT(const T* t) {
     return RB_ENTRY(t).Parent();
 }
 
 template <typename T>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 constexpr void RB_SET_LEFT(T* t, T* e) {
     RB_ENTRY(t).SetLeft(e);
 }
 template <typename T>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 constexpr void RB_SET_RIGHT(T* t, T* e) {
     RB_ENTRY(t).SetRight(e);
 }
 template <typename T>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 constexpr void RB_SET_PARENT(T* t, T* e) {
     RB_ENTRY(t).SetParent(e);
 }
 
 template <typename T>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 [[nodiscard]] constexpr bool RB_IS_BLACK(const T* t) {
     return RB_ENTRY(t).IsBlack();
 }
 template <typename T>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 [[nodiscard]] constexpr bool RB_IS_RED(const T* t) {
     return RB_ENTRY(t).IsRed();
 }
 
 template <typename T>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 [[nodiscard]] constexpr RBColor RB_COLOR(const T* t) {
     return RB_ENTRY(t).Color();
 }
 
 template <typename T>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 constexpr void RB_SET_COLOR(T* t, RBColor c) {
     RB_ENTRY(t).SetColor(c);
 }
 
 template <typename T>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 constexpr void RB_SET(T* elm, T* parent) {
     auto& rb_entry = RB_ENTRY(elm);
     rb_entry.SetParent(parent);
@@ -223,14 +223,14 @@ constexpr void RB_SET(T* elm, T* parent) {
 }
 
 template <typename T>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 constexpr void RB_SET_BLACKRED(T* black, T* red) {
     RB_SET_COLOR(black, RBColor::RB_BLACK);
     RB_SET_COLOR(red, RBColor::RB_RED);
 }
 
 template <typename T>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 constexpr void RB_ROTATE_LEFT(RBHead<T>& head, T* elm, T*& tmp) {
     tmp = RB_RIGHT(elm);
     if (RB_SET_RIGHT(elm, RB_LEFT(tmp)); RB_RIGHT(elm) != nullptr) {
@@ -252,7 +252,7 @@ constexpr void RB_ROTATE_LEFT(RBHead<T>& head, T* elm, T*& tmp) {
 }
 
 template <typename T>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 constexpr void RB_ROTATE_RIGHT(RBHead<T>& head, T* elm, T*& tmp) {
     tmp = RB_LEFT(elm);
     if (RB_SET_LEFT(elm, RB_RIGHT(tmp)); RB_LEFT(elm) != nullptr) {
@@ -274,7 +274,7 @@ constexpr void RB_ROTATE_RIGHT(RBHead<T>& head, T* elm, T*& tmp) {
 }
 
 template <typename T>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 constexpr void RB_REMOVE_COLOR(RBHead<T>& head, T* parent, T* elm) {
     T* tmp;
     while ((elm == nullptr || RB_IS_BLACK(elm)) && elm != head.Root()) {
@@ -358,7 +358,7 @@ constexpr void RB_REMOVE_COLOR(RBHead<T>& head, T* parent, T* elm) {
 }
 
 template <typename T>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 constexpr T* RB_REMOVE(RBHead<T>& head, T* elm) {
     T* child = nullptr;
     T* parent = nullptr;
@@ -451,7 +451,7 @@ constexpr T* RB_REMOVE(RBHead<T>& head, T* elm) {
 }
 
 template <typename T>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 constexpr void RB_INSERT_COLOR(RBHead<T>& head, T* elm) {
     T *parent = nullptr, *tmp = nullptr;
     while ((parent = RB_PARENT(elm)) != nullptr && RB_IS_RED(parent)) {
@@ -499,7 +499,7 @@ constexpr void RB_INSERT_COLOR(RBHead<T>& head, T* elm) {
 }
 
 template <typename T, typename Compare>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 constexpr T* RB_INSERT(RBHead<T>& head, T* elm, Compare cmp) {
     T* parent = nullptr;
     T* tmp = head.Root();
@@ -534,7 +534,7 @@ constexpr T* RB_INSERT(RBHead<T>& head, T* elm, Compare cmp) {
 }
 
 template <typename T, typename Compare>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 constexpr T* RB_FIND(RBHead<T>& head, T* elm, Compare cmp) {
     T* tmp = head.Root();
 
@@ -553,7 +553,7 @@ constexpr T* RB_FIND(RBHead<T>& head, T* elm, Compare cmp) {
 }
 
 template <typename T, typename Compare>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 constexpr T* RB_NFIND(RBHead<T>& head, T* elm, Compare cmp) {
     T* tmp = head.Root();
     T* res = nullptr;
@@ -574,7 +574,7 @@ constexpr T* RB_NFIND(RBHead<T>& head, T* elm, Compare cmp) {
 }
 
 template <typename T, typename U, typename Compare>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 constexpr T* RB_FIND_KEY(RBHead<T>& head, const U& key, Compare cmp) {
     T* tmp = head.Root();
 
@@ -593,7 +593,7 @@ constexpr T* RB_FIND_KEY(RBHead<T>& head, const U& key, Compare cmp) {
 }
 
 template <typename T, typename U, typename Compare>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 constexpr T* RB_NFIND_KEY(RBHead<T>& head, const U& key, Compare cmp) {
     T* tmp = head.Root();
     T* res = nullptr;
@@ -614,7 +614,7 @@ constexpr T* RB_NFIND_KEY(RBHead<T>& head, const U& key, Compare cmp) {
 }
 
 template <typename T, typename Compare>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 constexpr T* RB_FIND_EXISTING(RBHead<T>& head, T* elm, Compare cmp) {
     T* tmp = head.Root();
 
@@ -631,7 +631,7 @@ constexpr T* RB_FIND_EXISTING(RBHead<T>& head, T* elm, Compare cmp) {
 }
 
 template <typename T, typename U, typename Compare>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 constexpr T* RB_FIND_EXISTING_KEY(RBHead<T>& head, const U& key, Compare cmp) {
     T* tmp = head.Root();
 
@@ -648,7 +648,7 @@ constexpr T* RB_FIND_EXISTING_KEY(RBHead<T>& head, const U& key, Compare cmp) {
 }
 
 template <typename T>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 constexpr T* RB_NEXT(T* elm) {
     if (RB_RIGHT(elm)) {
         elm = RB_RIGHT(elm);
@@ -669,7 +669,7 @@ constexpr T* RB_NEXT(T* elm) {
 }
 
 template <typename T>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 constexpr T* RB_PREV(T* elm) {
     if (RB_LEFT(elm)) {
         elm = RB_LEFT(elm);
@@ -690,7 +690,7 @@ constexpr T* RB_PREV(T* elm) {
 }
 
 template <typename T>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 constexpr T* RB_MIN(RBHead<T>& head) {
     T* tmp = head.Root();
     T* parent = nullptr;
@@ -704,7 +704,7 @@ constexpr T* RB_MIN(RBHead<T>& head) {
 }
 
 template <typename T>
-requires HasRBEntry<T>
+    requires HasRBEntry<T>
 constexpr T* RB_MAX(RBHead<T>& head) {
     T* tmp = head.Root();
     T* parent = nullptr;
diff --git a/src/common/vector_math.h b/src/common/vector_math.h
index e62eeea2e..0e2095c45 100644
--- a/src/common/vector_math.h
+++ b/src/common/vector_math.h
@@ -348,9 +348,7 @@ public:
 // _DEFINE_SWIZZLER2 defines a single such function, DEFINE_SWIZZLER2 defines all of them for all
 // component names (x<->r) and permutations (xy<->yx)
 #define _DEFINE_SWIZZLER2(a, b, name)                                                              \
-    [[nodiscard]] constexpr Vec2<T> name() const {                                                 \
-        return Vec2<T>(a, b);                                                                      \
-    }
+    [[nodiscard]] constexpr Vec2<T> name() const { return Vec2<T>(a, b); }
 #define DEFINE_SWIZZLER2(a, b, a2, b2, a3, b3, a4, b4)                                             \
     _DEFINE_SWIZZLER2(a, b, a##b);                                                                 \
     _DEFINE_SWIZZLER2(a, b, a2##b2);                                                               \
@@ -543,9 +541,7 @@ public:
 // DEFINE_SWIZZLER2_COMP2 defines two component functions for all component names (x<->r) and
 // permutations (xy<->yx)
 #define _DEFINE_SWIZZLER2(a, b, name)                                                              \
-    [[nodiscard]] constexpr Vec2<T> name() const {                                                 \
-        return Vec2<T>(a, b);                                                                      \
-    }
+    [[nodiscard]] constexpr Vec2<T> name() const { return Vec2<T>(a, b); }
 #define DEFINE_SWIZZLER2_COMP1(a, a2)                                                              \
     _DEFINE_SWIZZLER2(a, a, a##a);                                                                 \
     _DEFINE_SWIZZLER2(a, a, a2##a2)
@@ -570,9 +566,7 @@ public:
 #undef _DEFINE_SWIZZLER2
 
 #define _DEFINE_SWIZZLER3(a, b, c, name)                                                           \
-    [[nodiscard]] constexpr Vec3<T> name() const {                                                 \
-        return Vec3<T>(a, b, c);                                                                   \
-    }
+    [[nodiscard]] constexpr Vec3<T> name() const { return Vec3<T>(a, b, c); }
 #define DEFINE_SWIZZLER3_COMP1(a, a2)                                                              \
     _DEFINE_SWIZZLER3(a, a, a, a##a##a);                                                           \
     _DEFINE_SWIZZLER3(a, a, a, a2##a2##a2)
@@ -641,8 +635,8 @@ template <typename T>
 
 // linear interpolation via float: 0.0=begin, 1.0=end
 template <typename X>
-[[nodiscard]] constexpr decltype(X{} * float{} + X{} * float{})
-    Lerp(const X& begin, const X& end, const float t) {
+[[nodiscard]] constexpr decltype(X{} * float{} + X{} * float{}) Lerp(const X& begin, const X& end,
+                                                                     const float t) {
     return begin * (1.f - t) + end * t;
 }
 
diff --git a/src/core/CMakeLists.txt b/src/core/CMakeLists.txt
index 5afdeb5ff..112c61b80 100644
--- a/src/core/CMakeLists.txt
+++ b/src/core/CMakeLists.txt
@@ -182,6 +182,8 @@ add_library(core STATIC
     hle/kernel/k_auto_object_container.cpp
     hle/kernel/k_auto_object_container.h
     hle/kernel/k_affinity_mask.h
+    hle/kernel/k_capabilities.cpp
+    hle/kernel/k_capabilities.h
     hle/kernel/k_class_token.cpp
     hle/kernel/k_class_token.h
     hle/kernel/k_client_port.cpp
@@ -193,6 +195,8 @@ add_library(core STATIC
     hle/kernel/k_condition_variable.cpp
     hle/kernel/k_condition_variable.h
     hle/kernel/k_debug.h
+    hle/kernel/k_device_address_space.cpp
+    hle/kernel/k_device_address_space.h
     hle/kernel/k_dynamic_page_manager.h
     hle/kernel/k_dynamic_resource_manager.h
     hle/kernel/k_dynamic_slab_heap.h
diff --git a/src/core/hardware_properties.h b/src/core/hardware_properties.h
index 13cbdb734..45567b840 100644
--- a/src/core/hardware_properties.h
+++ b/src/core/hardware_properties.h
@@ -25,6 +25,26 @@ constexpr std::array<s32, Common::BitSize<u64>()> VirtualToPhysicalCoreMap{
     0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 3,
 };
 
+static constexpr inline size_t NumVirtualCores = Common::BitSize<u64>();
+
+static constexpr inline u64 VirtualCoreMask = [] {
+    u64 mask = 0;
+    for (size_t i = 0; i < NumVirtualCores; ++i) {
+        mask |= (UINT64_C(1) << i);
+    }
+    return mask;
+}();
+
+static constexpr inline u64 ConvertVirtualCoreMaskToPhysical(u64 v_core_mask) {
+    u64 p_core_mask = 0;
+    while (v_core_mask != 0) {
+        const u64 next = std::countr_zero(v_core_mask);
+        v_core_mask &= ~(static_cast<u64>(1) << next);
+        p_core_mask |= (static_cast<u64>(1) << VirtualToPhysicalCoreMap[next]);
+    }
+    return p_core_mask;
+}
+
 // Cortex-A57 supports 4 memory watchpoints
 constexpr u64 NUM_WATCHPOINTS = 4;
 
diff --git a/src/core/hid/emulated_controller.cpp b/src/core/hid/emulated_controller.cpp
index 0e06468da..631aa6ad2 100644
--- a/src/core/hid/emulated_controller.cpp
+++ b/src/core/hid/emulated_controller.cpp
@@ -12,6 +12,7 @@
 namespace Core::HID {
 constexpr s32 HID_JOYSTICK_MAX = 0x7fff;
 constexpr s32 HID_TRIGGER_MAX = 0x7fff;
+constexpr u32 TURBO_BUTTON_DELAY = 4;
 // Use a common UUID for TAS and Virtual Gamepad
 constexpr Common::UUID TAS_UUID =
     Common::UUID{{0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x7, 0xA5, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0}};
@@ -447,6 +448,7 @@ void EmulatedController::ReloadInput() {
                 },
         });
     }
+    turbo_button_state = 0;
 }
 
 void EmulatedController::UnloadInput() {
@@ -687,6 +689,7 @@ void EmulatedController::SetButton(const Common::Input::CallbackStatus& callback
     }
 
     current_status.toggle = new_status.toggle;
+    current_status.turbo = new_status.turbo;
     current_status.uuid = uuid;
 
     // Update button status with current
@@ -1548,7 +1551,7 @@ NpadButtonState EmulatedController::GetNpadButtons() const {
     if (is_configuring) {
         return {};
     }
-    return controller.npad_button_state;
+    return {controller.npad_button_state.raw & GetTurboButtonMask()};
 }
 
 DebugPadButton EmulatedController::GetDebugPadButtons() const {
@@ -1656,4 +1659,74 @@ void EmulatedController::DeleteCallback(int key) {
     }
     callback_list.erase(iterator);
 }
+
+void EmulatedController::TurboButtonUpdate() {
+    turbo_button_state = (turbo_button_state + 1) % (TURBO_BUTTON_DELAY * 2);
+}
+
+NpadButton EmulatedController::GetTurboButtonMask() const {
+    // Apply no mask when disabled
+    if (turbo_button_state < TURBO_BUTTON_DELAY) {
+        return {NpadButton::All};
+    }
+
+    NpadButtonState button_mask{};
+    for (std::size_t index = 0; index < controller.button_values.size(); ++index) {
+        if (!controller.button_values[index].turbo) {
+            continue;
+        }
+
+        switch (index) {
+        case Settings::NativeButton::A:
+            button_mask.a.Assign(1);
+            break;
+        case Settings::NativeButton::B:
+            button_mask.b.Assign(1);
+            break;
+        case Settings::NativeButton::X:
+            button_mask.x.Assign(1);
+            break;
+        case Settings::NativeButton::Y:
+            button_mask.y.Assign(1);
+            break;
+        case Settings::NativeButton::L:
+            button_mask.l.Assign(1);
+            break;
+        case Settings::NativeButton::R:
+            button_mask.r.Assign(1);
+            break;
+        case Settings::NativeButton::ZL:
+            button_mask.zl.Assign(1);
+            break;
+        case Settings::NativeButton::ZR:
+            button_mask.zr.Assign(1);
+            break;
+        case Settings::NativeButton::DLeft:
+            button_mask.left.Assign(1);
+            break;
+        case Settings::NativeButton::DUp:
+            button_mask.up.Assign(1);
+            break;
+        case Settings::NativeButton::DRight:
+            button_mask.right.Assign(1);
+            break;
+        case Settings::NativeButton::DDown:
+            button_mask.down.Assign(1);
+            break;
+        case Settings::NativeButton::SL:
+            button_mask.left_sl.Assign(1);
+            button_mask.right_sl.Assign(1);
+            break;
+        case Settings::NativeButton::SR:
+            button_mask.left_sr.Assign(1);
+            button_mask.right_sr.Assign(1);
+            break;
+        default:
+            break;
+        }
+    }
+
+    return static_cast<NpadButton>(~button_mask.raw);
+}
+
 } // namespace Core::HID
diff --git a/src/core/hid/emulated_controller.h b/src/core/hid/emulated_controller.h
index 3ac77b2b5..b02bf35c4 100644
--- a/src/core/hid/emulated_controller.h
+++ b/src/core/hid/emulated_controller.h
@@ -411,6 +411,9 @@ public:
      */
     void DeleteCallback(int key);
 
+    /// Swaps the state of the turbo buttons
+    void TurboButtonUpdate();
+
 private:
     /// creates input devices from params
     void LoadDevices();
@@ -511,6 +514,8 @@ private:
      */
     void TriggerOnChange(ControllerTriggerType type, bool is_service_update);
 
+    NpadButton GetTurboButtonMask() const;
+
     const NpadIdType npad_id_type;
     NpadStyleIndex npad_type{NpadStyleIndex::None};
     NpadStyleIndex original_npad_type{NpadStyleIndex::None};
@@ -520,6 +525,7 @@ private:
     bool system_buttons_enabled{true};
     f32 motion_sensitivity{0.01f};
     bool force_update_motion{false};
+    u32 turbo_button_state{0};
 
     // Temporary values to avoid doing changes while the controller is in configuring mode
     NpadStyleIndex tmp_npad_type{NpadStyleIndex::None};
diff --git a/src/core/hle/kernel/hle_ipc.cpp b/src/core/hle/kernel/hle_ipc.cpp
index 738b6d0f1..494151eef 100644
--- a/src/core/hle/kernel/hle_ipc.cpp
+++ b/src/core/hle/kernel/hle_ipc.cpp
@@ -11,6 +11,7 @@
 #include "common/common_funcs.h"
 #include "common/common_types.h"
 #include "common/logging/log.h"
+#include "common/scratch_buffer.h"
 #include "core/hle/ipc_helpers.h"
 #include "core/hle/kernel/hle_ipc.h"
 #include "core/hle/kernel/k_auto_object.h"
@@ -325,7 +326,7 @@ Result HLERequestContext::WriteToOutgoingCommandBuffer(KThread& requesting_threa
     return ResultSuccess;
 }
 
-std::vector<u8> HLERequestContext::ReadBuffer(std::size_t buffer_index) const {
+std::vector<u8> HLERequestContext::ReadBufferCopy(std::size_t buffer_index) const {
     const bool is_buffer_a{BufferDescriptorA().size() > buffer_index &&
                            BufferDescriptorA()[buffer_index].Size()};
     if (is_buffer_a) {
@@ -345,6 +346,33 @@ std::vector<u8> HLERequestContext::ReadBuffer(std::size_t buffer_index) const {
     }
 }
 
+std::span<const u8> HLERequestContext::ReadBuffer(std::size_t buffer_index) const {
+    static thread_local std::array<Common::ScratchBuffer<u8>, 2> read_buffer_a;
+    static thread_local std::array<Common::ScratchBuffer<u8>, 2> read_buffer_x;
+
+    const bool is_buffer_a{BufferDescriptorA().size() > buffer_index &&
+                           BufferDescriptorA()[buffer_index].Size()};
+    if (is_buffer_a) {
+        ASSERT_OR_EXECUTE_MSG(
+            BufferDescriptorA().size() > buffer_index, { return {}; },
+            "BufferDescriptorA invalid buffer_index {}", buffer_index);
+        auto& read_buffer = read_buffer_a[buffer_index];
+        read_buffer.resize_destructive(BufferDescriptorA()[buffer_index].Size());
+        memory.ReadBlock(BufferDescriptorA()[buffer_index].Address(), read_buffer.data(),
+                         read_buffer.size());
+        return read_buffer;
+    } else {
+        ASSERT_OR_EXECUTE_MSG(
+            BufferDescriptorX().size() > buffer_index, { return {}; },
+            "BufferDescriptorX invalid buffer_index {}", buffer_index);
+        auto& read_buffer = read_buffer_x[buffer_index];
+        read_buffer.resize_destructive(BufferDescriptorX()[buffer_index].Size());
+        memory.ReadBlock(BufferDescriptorX()[buffer_index].Address(), read_buffer.data(),
+                         read_buffer.size());
+        return read_buffer;
+    }
+}
+
 std::size_t HLERequestContext::WriteBuffer(const void* buffer, std::size_t size,
                                            std::size_t buffer_index) const {
     if (size == 0) {
diff --git a/src/core/hle/kernel/hle_ipc.h b/src/core/hle/kernel/hle_ipc.h
index e252b5f4b..5bf4f171b 100644
--- a/src/core/hle/kernel/hle_ipc.h
+++ b/src/core/hle/kernel/hle_ipc.h
@@ -7,6 +7,7 @@
 #include <functional>
 #include <memory>
 #include <optional>
+#include <span>
 #include <string>
 #include <type_traits>
 #include <vector>
@@ -270,8 +271,11 @@ public:
         return domain_message_header.has_value();
     }
 
-    /// Helper function to read a buffer using the appropriate buffer descriptor
-    [[nodiscard]] std::vector<u8> ReadBuffer(std::size_t buffer_index = 0) const;
+    /// Helper function to get a span of a buffer using the appropriate buffer descriptor
+    [[nodiscard]] std::span<const u8> ReadBuffer(std::size_t buffer_index = 0) const;
+
+    /// Helper function to read a copy of a buffer using the appropriate buffer descriptor
+    [[nodiscard]] std::vector<u8> ReadBufferCopy(std::size_t buffer_index = 0) const;
 
     /// Helper function to write a buffer using the appropriate buffer descriptor
     std::size_t WriteBuffer(const void* buffer, std::size_t size,
diff --git a/src/core/hle/kernel/init/init_slab_setup.cpp b/src/core/hle/kernel/init/init_slab_setup.cpp
index 7b363eb1e..571acf4b2 100644
--- a/src/core/hle/kernel/init/init_slab_setup.cpp
+++ b/src/core/hle/kernel/init/init_slab_setup.cpp
@@ -11,6 +11,7 @@
 #include "core/hle/kernel/init/init_slab_setup.h"
 #include "core/hle/kernel/k_code_memory.h"
 #include "core/hle/kernel/k_debug.h"
+#include "core/hle/kernel/k_device_address_space.h"
 #include "core/hle/kernel/k_event.h"
 #include "core/hle/kernel/k_event_info.h"
 #include "core/hle/kernel/k_memory_layout.h"
@@ -43,6 +44,7 @@ namespace Kernel::Init {
     HANDLER(KSharedMemoryInfo, (SLAB_COUNT(KSharedMemory) * 8), ##__VA_ARGS__)                     \
     HANDLER(KTransferMemory, (SLAB_COUNT(KTransferMemory)), ##__VA_ARGS__)                         \
     HANDLER(KCodeMemory, (SLAB_COUNT(KCodeMemory)), ##__VA_ARGS__)                                 \
+    HANDLER(KDeviceAddressSpace, (SLAB_COUNT(KDeviceAddressSpace)), ##__VA_ARGS__)                 \
     HANDLER(KSession, (SLAB_COUNT(KSession)), ##__VA_ARGS__)                                       \
     HANDLER(KThreadLocalPage,                                                                      \
             (SLAB_COUNT(KProcess) + (SLAB_COUNT(KProcess) + SLAB_COUNT(KThread)) / 8),             \
diff --git a/src/core/hle/kernel/k_auto_object.h b/src/core/hle/kernel/k_auto_object.h
index 2827763d5..e8118c2b8 100644
--- a/src/core/hle/kernel/k_auto_object.h
+++ b/src/core/hle/kernel/k_auto_object.h
@@ -24,9 +24,7 @@ private:
     friend class ::Kernel::KClassTokenGenerator;                                                   \
     static constexpr inline auto ObjectType = ::Kernel::KClassTokenGenerator::ObjectType::CLASS;   \
     static constexpr inline const char* const TypeName = #CLASS;                                   \
-    static constexpr inline ClassTokenType ClassToken() {                                          \
-        return ::Kernel::ClassToken<CLASS>;                                                        \
-    }                                                                                              \
+    static constexpr inline ClassTokenType ClassToken() { return ::Kernel::ClassToken<CLASS>; }    \
                                                                                                    \
 public:                                                                                            \
     YUZU_NON_COPYABLE(CLASS);                                                                      \
@@ -37,15 +35,9 @@ public:
         constexpr ClassTokenType Token = ClassToken();                                             \
         return TypeObj(TypeName, Token);                                                           \
     }                                                                                              \
-    static constexpr const char* GetStaticTypeName() {                                             \
-        return TypeName;                                                                           \
-    }                                                                                              \
-    virtual TypeObj GetTypeObj() ATTRIBUTE {                                                       \
-        return GetStaticTypeObj();                                                                 \
-    }                                                                                              \
-    virtual const char* GetTypeName() ATTRIBUTE {                                                  \
-        return GetStaticTypeName();                                                                \
-    }                                                                                              \
+    static constexpr const char* GetStaticTypeName() { return TypeName; }                          \
+    virtual TypeObj GetTypeObj() ATTRIBUTE { return GetStaticTypeObj(); }                          \
+    virtual const char* GetTypeName() ATTRIBUTE { return GetStaticTypeName(); }                    \
                                                                                                    \
 private:                                                                                           \
     constexpr bool operator!=(const TypeObj& rhs)
@@ -245,8 +237,8 @@ public:
     }
 
     template <typename U>
-    requires(std::derived_from<T, U> ||
-             std::derived_from<U, T>) constexpr KScopedAutoObject(KScopedAutoObject<U>&& rhs) {
+        requires(std::derived_from<T, U> || std::derived_from<U, T>)
+    constexpr KScopedAutoObject(KScopedAutoObject<U>&& rhs) {
         if constexpr (std::derived_from<U, T>) {
             // Upcast.
             m_obj = rhs.m_obj;
diff --git a/src/core/hle/kernel/k_capabilities.cpp b/src/core/hle/kernel/k_capabilities.cpp
new file mode 100644
index 000000000..64f1d7371
--- /dev/null
+++ b/src/core/hle/kernel/k_capabilities.cpp
@@ -0,0 +1,358 @@
+// SPDX-FileCopyrightText: Copyright 2023 yuzu Emulator Project
+// SPDX-License-Identifier: GPL-2.0-or-later
+
+#include "core/hardware_properties.h"
+#include "core/hle/kernel/k_capabilities.h"
+#include "core/hle/kernel/k_memory_layout.h"
+#include "core/hle/kernel/k_page_table.h"
+#include "core/hle/kernel/kernel.h"
+#include "core/hle/kernel/svc_results.h"
+#include "core/hle/kernel/svc_version.h"
+
+namespace Kernel {
+
+Result KCapabilities::InitializeForKIP(std::span<const u32> kern_caps, KPageTable* page_table) {
+    // We're initializing an initial process.
+    m_svc_access_flags.reset();
+    m_irq_access_flags.reset();
+    m_debug_capabilities = 0;
+    m_handle_table_size = 0;
+    m_intended_kernel_version = 0;
+    m_program_type = 0;
+
+    // Initial processes may run on all cores.
+    constexpr u64 VirtMask = Core::Hardware::VirtualCoreMask;
+    constexpr u64 PhysMask = Core::Hardware::ConvertVirtualCoreMaskToPhysical(VirtMask);
+
+    m_core_mask = VirtMask;
+    m_phys_core_mask = PhysMask;
+
+    // Initial processes may use any user priority they like.
+    m_priority_mask = ~0xFULL;
+
+    // Here, Nintendo sets the kernel version to the current kernel version.
+    // We will follow suit and set the version to the highest supported kernel version.
+    KernelVersion intended_kernel_version{};
+    intended_kernel_version.major_version.Assign(Svc::SupportedKernelMajorVersion);
+    intended_kernel_version.minor_version.Assign(Svc::SupportedKernelMinorVersion);
+    m_intended_kernel_version = intended_kernel_version.raw;
+
+    // Parse the capabilities array.
+    R_RETURN(this->SetCapabilities(kern_caps, page_table));
+}
+
+Result KCapabilities::InitializeForUser(std::span<const u32> user_caps, KPageTable* page_table) {
+    // We're initializing a user process.
+    m_svc_access_flags.reset();
+    m_irq_access_flags.reset();
+    m_debug_capabilities = 0;
+    m_handle_table_size = 0;
+    m_intended_kernel_version = 0;
+    m_program_type = 0;
+
+    // User processes must specify what cores/priorities they can use.
+    m_core_mask = 0;
+    m_priority_mask = 0;
+
+    // Parse the user capabilities array.
+    R_RETURN(this->SetCapabilities(user_caps, page_table));
+}
+
+Result KCapabilities::SetCorePriorityCapability(const u32 cap) {
+    // We can't set core/priority if we've already set them.
+    R_UNLESS(m_core_mask == 0, ResultInvalidArgument);
+    R_UNLESS(m_priority_mask == 0, ResultInvalidArgument);
+
+    // Validate the core/priority.
+    CorePriority pack{cap};
+    const u32 min_core = pack.minimum_core_id;
+    const u32 max_core = pack.maximum_core_id;
+    const u32 max_prio = pack.lowest_thread_priority;
+    const u32 min_prio = pack.highest_thread_priority;
+
+    R_UNLESS(min_core <= max_core, ResultInvalidCombination);
+    R_UNLESS(min_prio <= max_prio, ResultInvalidCombination);
+    R_UNLESS(max_core < Core::Hardware::NumVirtualCores, ResultInvalidCoreId);
+
+    ASSERT(max_prio < Common::BitSize<u64>());
+
+    // Set core mask.
+    for (auto core_id = min_core; core_id <= max_core; core_id++) {
+        m_core_mask |= (1ULL << core_id);
+    }
+    ASSERT((m_core_mask & Core::Hardware::VirtualCoreMask) == m_core_mask);
+
+    // Set physical core mask.
+    m_phys_core_mask = Core::Hardware::ConvertVirtualCoreMaskToPhysical(m_core_mask);
+
+    // Set priority mask.
+    for (auto prio = min_prio; prio <= max_prio; prio++) {
+        m_priority_mask |= (1ULL << prio);
+    }
+
+    // We must have some core/priority we can use.
+    R_UNLESS(m_core_mask != 0, ResultInvalidArgument);
+    R_UNLESS(m_priority_mask != 0, ResultInvalidArgument);
+
+    // Processes must not have access to kernel thread priorities.
+    R_UNLESS((m_priority_mask & 0xF) == 0, ResultInvalidArgument);
+
+    R_SUCCEED();
+}
+
+Result KCapabilities::SetSyscallMaskCapability(const u32 cap, u32& set_svc) {
+    // Validate the index.
+    SyscallMask pack{cap};
+    const u32 mask = pack.mask;
+    const u32 index = pack.index;
+
+    const u32 index_flag = (1U << index);
+    R_UNLESS((set_svc & index_flag) == 0, ResultInvalidCombination);
+    set_svc |= index_flag;
+
+    // Set SVCs.
+    for (size_t i = 0; i < decltype(SyscallMask::mask)::bits; i++) {
+        const u32 svc_id = static_cast<u32>(decltype(SyscallMask::mask)::bits * index + i);
+        if (mask & (1U << i)) {
+            R_UNLESS(this->SetSvcAllowed(svc_id), ResultOutOfRange);
+        }
+    }
+
+    R_SUCCEED();
+}
+
+Result KCapabilities::MapRange_(const u32 cap, const u32 size_cap, KPageTable* page_table) {
+    const auto range_pack = MapRange{cap};
+    const auto size_pack = MapRangeSize{size_cap};
+
+    // Get/validate address/size
+    const u64 phys_addr = range_pack.address.Value() * PageSize;
+
+    // Validate reserved bits are unused.
+    R_UNLESS(size_pack.reserved.Value() == 0, ResultOutOfRange);
+
+    const size_t num_pages = size_pack.pages;
+    const size_t size = num_pages * PageSize;
+    R_UNLESS(num_pages != 0, ResultInvalidSize);
+    R_UNLESS(phys_addr < phys_addr + size, ResultInvalidAddress);
+    R_UNLESS(((phys_addr + size - 1) & ~PhysicalMapAllowedMask) == 0, ResultInvalidAddress);
+
+    // Do the mapping.
+    [[maybe_unused]] const KMemoryPermission perm = range_pack.read_only.Value()
+                                                        ? KMemoryPermission::UserRead
+                                                        : KMemoryPermission::UserReadWrite;
+    if (MapRangeSize{size_cap}.normal) {
+        // R_RETURN(page_table->MapStatic(phys_addr, size, perm));
+    } else {
+        // R_RETURN(page_table->MapIo(phys_addr, size, perm));
+    }
+
+    UNIMPLEMENTED();
+    R_SUCCEED();
+}
+
+Result KCapabilities::MapIoPage_(const u32 cap, KPageTable* page_table) {
+    // Get/validate address/size
+    const u64 phys_addr = MapIoPage{cap}.address.Value() * PageSize;
+    const size_t num_pages = 1;
+    const size_t size = num_pages * PageSize;
+    R_UNLESS(num_pages != 0, ResultInvalidSize);
+    R_UNLESS(phys_addr < phys_addr + size, ResultInvalidAddress);
+    R_UNLESS(((phys_addr + size - 1) & ~PhysicalMapAllowedMask) == 0, ResultInvalidAddress);
+
+    // Do the mapping.
+    // R_RETURN(page_table->MapIo(phys_addr, size, KMemoryPermission_UserReadWrite));
+
+    UNIMPLEMENTED();
+    R_SUCCEED();
+}
+
+template <typename F>
+Result KCapabilities::ProcessMapRegionCapability(const u32 cap, F f) {
+    // Define the allowed memory regions.
+    constexpr std::array<KMemoryRegionType, 4> MemoryRegions{
+        KMemoryRegionType_None,
+        KMemoryRegionType_KernelTraceBuffer,
+        KMemoryRegionType_OnMemoryBootImage,
+        KMemoryRegionType_DTB,
+    };
+
+    // Extract regions/read only.
+    const MapRegion pack{cap};
+    const std::array<RegionType, 3> types{pack.region0, pack.region1, pack.region2};
+    const std::array<u32, 3> ro{pack.read_only0, pack.read_only1, pack.read_only2};
+
+    for (size_t i = 0; i < types.size(); i++) {
+        const auto type = types[i];
+        const auto perm = ro[i] ? KMemoryPermission::UserRead : KMemoryPermission::UserReadWrite;
+        switch (type) {
+        case RegionType::NoMapping:
+            break;
+        case RegionType::KernelTraceBuffer:
+        case RegionType::OnMemoryBootImage:
+        case RegionType::DTB:
+            R_TRY(f(MemoryRegions[static_cast<u32>(type)], perm));
+            break;
+        default:
+            R_THROW(ResultNotFound);
+        }
+    }
+
+    R_SUCCEED();
+}
+
+Result KCapabilities::MapRegion_(const u32 cap, KPageTable* page_table) {
+    // Map each region into the process's page table.
+    R_RETURN(ProcessMapRegionCapability(
+        cap, [](KMemoryRegionType region_type, KMemoryPermission perm) -> Result {
+            // R_RETURN(page_table->MapRegion(region_type, perm));
+            UNIMPLEMENTED();
+            R_SUCCEED();
+        }));
+}
+
+Result KCapabilities::CheckMapRegion(KernelCore& kernel, const u32 cap) {
+    // Check that each region has a physical backing store.
+    R_RETURN(ProcessMapRegionCapability(
+        cap, [&](KMemoryRegionType region_type, KMemoryPermission perm) -> Result {
+            R_UNLESS(kernel.MemoryLayout().GetPhysicalMemoryRegionTree().FindFirstDerived(
+                         region_type) != nullptr,
+                     ResultOutOfRange);
+            R_SUCCEED();
+        }));
+}
+
+Result KCapabilities::SetInterruptPairCapability(const u32 cap) {
+    // Extract interrupts.
+    const InterruptPair pack{cap};
+    const std::array<u32, 2> ids{pack.interrupt_id0, pack.interrupt_id1};
+
+    for (size_t i = 0; i < ids.size(); i++) {
+        if (ids[i] != PaddingInterruptId) {
+            UNIMPLEMENTED();
+            // R_UNLESS(Kernel::GetInterruptManager().IsInterruptDefined(ids[i]), ResultOutOfRange);
+            // R_UNLESS(this->SetInterruptPermitted(ids[i]), ResultOutOfRange);
+        }
+    }
+
+    R_SUCCEED();
+}
+
+Result KCapabilities::SetProgramTypeCapability(const u32 cap) {
+    // Validate.
+    const ProgramType pack{cap};
+    R_UNLESS(pack.reserved == 0, ResultReservedUsed);
+
+    m_program_type = pack.type;
+    R_SUCCEED();
+}
+
+Result KCapabilities::SetKernelVersionCapability(const u32 cap) {
+    // Ensure we haven't set our version before.
+    R_UNLESS(KernelVersion{m_intended_kernel_version}.major_version == 0, ResultInvalidArgument);
+
+    // Set, ensure that we set a valid version.
+    m_intended_kernel_version = cap;
+    R_UNLESS(KernelVersion{m_intended_kernel_version}.major_version != 0, ResultInvalidArgument);
+
+    R_SUCCEED();
+}
+
+Result KCapabilities::SetHandleTableCapability(const u32 cap) {
+    // Validate.
+    const HandleTable pack{cap};
+    R_UNLESS(pack.reserved == 0, ResultReservedUsed);
+
+    m_handle_table_size = pack.size;
+    R_SUCCEED();
+}
+
+Result KCapabilities::SetDebugFlagsCapability(const u32 cap) {
+    // Validate.
+    const DebugFlags pack{cap};
+    R_UNLESS(pack.reserved == 0, ResultReservedUsed);
+
+    DebugFlags debug_capabilities{m_debug_capabilities};
+    debug_capabilities.allow_debug.Assign(pack.allow_debug);
+    debug_capabilities.force_debug.Assign(pack.force_debug);
+    m_debug_capabilities = debug_capabilities.raw;
+
+    R_SUCCEED();
+}
+
+Result KCapabilities::SetCapability(const u32 cap, u32& set_flags, u32& set_svc,
+                                    KPageTable* page_table) {
+    // Validate this is a capability we can act on.
+    const auto type = GetCapabilityType(cap);
+    R_UNLESS(type != CapabilityType::Invalid, ResultInvalidArgument);
+
+    // If the type is padding, we have no work to do.
+    R_SUCCEED_IF(type == CapabilityType::Padding);
+
+    // Check that we haven't already processed this capability.
+    const auto flag = GetCapabilityFlag(type);
+    R_UNLESS(((set_flags & InitializeOnceFlags) & flag) == 0, ResultInvalidCombination);
+    set_flags |= flag;
+
+    // Process the capability.
+    switch (type) {
+    case CapabilityType::CorePriority:
+        R_RETURN(this->SetCorePriorityCapability(cap));
+    case CapabilityType::SyscallMask:
+        R_RETURN(this->SetSyscallMaskCapability(cap, set_svc));
+    case CapabilityType::MapIoPage:
+        R_RETURN(this->MapIoPage_(cap, page_table));
+    case CapabilityType::MapRegion:
+        R_RETURN(this->MapRegion_(cap, page_table));
+    case CapabilityType::InterruptPair:
+        R_RETURN(this->SetInterruptPairCapability(cap));
+    case CapabilityType::ProgramType:
+        R_RETURN(this->SetProgramTypeCapability(cap));
+    case CapabilityType::KernelVersion:
+        R_RETURN(this->SetKernelVersionCapability(cap));
+    case CapabilityType::HandleTable:
+        R_RETURN(this->SetHandleTableCapability(cap));
+    case CapabilityType::DebugFlags:
+        R_RETURN(this->SetDebugFlagsCapability(cap));
+    default:
+        R_THROW(ResultInvalidArgument);
+    }
+}
+
+Result KCapabilities::SetCapabilities(std::span<const u32> caps, KPageTable* page_table) {
+    u32 set_flags = 0, set_svc = 0;
+
+    for (size_t i = 0; i < caps.size(); i++) {
+        const u32 cap{caps[i]};
+
+        if (GetCapabilityType(cap) == CapabilityType::MapRange) {
+            // Check that the pair cap exists.
+            R_UNLESS((++i) < caps.size(), ResultInvalidCombination);
+
+            // Check the pair cap is a map range cap.
+            const u32 size_cap{caps[i]};
+            R_UNLESS(GetCapabilityType(size_cap) == CapabilityType::MapRange,
+                     ResultInvalidCombination);
+
+            // Map the range.
+            R_TRY(this->MapRange_(cap, size_cap, page_table));
+        } else {
+            R_TRY(this->SetCapability(cap, set_flags, set_svc, page_table));
+        }
+    }
+
+    R_SUCCEED();
+}
+
+Result KCapabilities::CheckCapabilities(KernelCore& kernel, std::span<const u32> caps) {
+    for (auto cap : caps) {
+        // Check the capability refers to a valid region.
+        if (GetCapabilityType(cap) == CapabilityType::MapRegion) {
+            R_TRY(CheckMapRegion(kernel, cap));
+        }
+    }
+
+    R_SUCCEED();
+}
+
+} // namespace Kernel
diff --git a/src/core/hle/kernel/k_capabilities.h b/src/core/hle/kernel/k_capabilities.h
new file mode 100644
index 000000000..cd96f8d23
--- /dev/null
+++ b/src/core/hle/kernel/k_capabilities.h
@@ -0,0 +1,295 @@
+
+// SPDX-FileCopyrightText: Copyright 2023 yuzu Emulator Project
+// SPDX-License-Identifier: GPL-2.0-or-later
+
+#pragma once
+
+#include <bitset>
+#include <span>
+
+#include "common/bit_field.h"
+#include "common/common_types.h"
+
+#include "core/hle/kernel/svc_types.h"
+#include "core/hle/result.h"
+
+namespace Kernel {
+
+class KPageTable;
+class KernelCore;
+
+class KCapabilities {
+public:
+    constexpr explicit KCapabilities() = default;
+
+    Result InitializeForKIP(std::span<const u32> kern_caps, KPageTable* page_table);
+    Result InitializeForUser(std::span<const u32> user_caps, KPageTable* page_table);
+
+    static Result CheckCapabilities(KernelCore& kernel, std::span<const u32> user_caps);
+
+    constexpr u64 GetCoreMask() const {
+        return m_core_mask;
+    }
+
+    constexpr u64 GetPhysicalCoreMask() const {
+        return m_phys_core_mask;
+    }
+
+    constexpr u64 GetPriorityMask() const {
+        return m_priority_mask;
+    }
+
+    constexpr s32 GetHandleTableSize() const {
+        return m_handle_table_size;
+    }
+
+    constexpr const Svc::SvcAccessFlagSet& GetSvcPermissions() const {
+        return m_svc_access_flags;
+    }
+
+    constexpr bool IsPermittedSvc(u32 id) const {
+        return (id < m_svc_access_flags.size()) && m_svc_access_flags[id];
+    }
+
+    constexpr bool IsPermittedInterrupt(u32 id) const {
+        return (id < m_irq_access_flags.size()) && m_irq_access_flags[id];
+    }
+
+    constexpr bool IsPermittedDebug() const {
+        return DebugFlags{m_debug_capabilities}.allow_debug.Value() != 0;
+    }
+
+    constexpr bool CanForceDebug() const {
+        return DebugFlags{m_debug_capabilities}.force_debug.Value() != 0;
+    }
+
+    constexpr u32 GetIntendedKernelMajorVersion() const {
+        return KernelVersion{m_intended_kernel_version}.major_version;
+    }
+
+    constexpr u32 GetIntendedKernelMinorVersion() const {
+        return KernelVersion{m_intended_kernel_version}.minor_version;
+    }
+
+private:
+    static constexpr size_t InterruptIdCount = 0x400;
+    using InterruptFlagSet = std::bitset<InterruptIdCount>;
+
+    enum class CapabilityType : u32 {
+        CorePriority = (1U << 3) - 1,
+        SyscallMask = (1U << 4) - 1,
+        MapRange = (1U << 6) - 1,
+        MapIoPage = (1U << 7) - 1,
+        MapRegion = (1U << 10) - 1,
+        InterruptPair = (1U << 11) - 1,
+        ProgramType = (1U << 13) - 1,
+        KernelVersion = (1U << 14) - 1,
+        HandleTable = (1U << 15) - 1,
+        DebugFlags = (1U << 16) - 1,
+
+        Invalid = 0U,
+        Padding = ~0U,
+    };
+
+    using RawCapabilityValue = u32;
+
+    static constexpr CapabilityType GetCapabilityType(const RawCapabilityValue value) {
+        return static_cast<CapabilityType>((~value & (value + 1)) - 1);
+    }
+
+    static constexpr u32 GetCapabilityFlag(CapabilityType type) {
+        return static_cast<u32>(type) + 1;
+    }
+
+    template <CapabilityType Type>
+    static constexpr inline u32 CapabilityFlag = static_cast<u32>(Type) + 1;
+
+    template <CapabilityType Type>
+    static constexpr inline u32 CapabilityId = std::countr_zero(CapabilityFlag<Type>);
+
+    union CorePriority {
+        static_assert(CapabilityId<CapabilityType::CorePriority> + 1 == 4);
+
+        RawCapabilityValue raw;
+        BitField<0, 4, CapabilityType> id;
+        BitField<4, 6, u32> lowest_thread_priority;
+        BitField<10, 6, u32> highest_thread_priority;
+        BitField<16, 8, u32> minimum_core_id;
+        BitField<24, 8, u32> maximum_core_id;
+    };
+
+    union SyscallMask {
+        static_assert(CapabilityId<CapabilityType::SyscallMask> + 1 == 5);
+
+        RawCapabilityValue raw;
+        BitField<0, 5, CapabilityType> id;
+        BitField<5, 24, u32> mask;
+        BitField<29, 3, u32> index;
+    };
+
+    // #undef MESOSPHERE_ENABLE_LARGE_PHYSICAL_ADDRESS_CAPABILITIES
+    static constexpr u64 PhysicalMapAllowedMask = (1ULL << 36) - 1;
+
+    union MapRange {
+        static_assert(CapabilityId<CapabilityType::MapRange> + 1 == 7);
+
+        RawCapabilityValue raw;
+        BitField<0, 7, CapabilityType> id;
+        BitField<7, 24, u32> address;
+        BitField<31, 1, u32> read_only;
+    };
+
+    union MapRangeSize {
+        static_assert(CapabilityId<CapabilityType::MapRange> + 1 == 7);
+
+        RawCapabilityValue raw;
+        BitField<0, 7, CapabilityType> id;
+        BitField<7, 20, u32> pages;
+        BitField<27, 4, u32> reserved;
+        BitField<31, 1, u32> normal;
+    };
+
+    union MapIoPage {
+        static_assert(CapabilityId<CapabilityType::MapIoPage> + 1 == 8);
+
+        RawCapabilityValue raw;
+        BitField<0, 8, CapabilityType> id;
+        BitField<8, 24, u32> address;
+    };
+
+    enum class RegionType : u32 {
+        NoMapping = 0,
+        KernelTraceBuffer = 1,
+        OnMemoryBootImage = 2,
+        DTB = 3,
+    };
+
+    union MapRegion {
+        static_assert(CapabilityId<CapabilityType::MapRegion> + 1 == 11);
+
+        RawCapabilityValue raw;
+        BitField<0, 11, CapabilityType> id;
+        BitField<11, 6, RegionType> region0;
+        BitField<17, 1, u32> read_only0;
+        BitField<18, 6, RegionType> region1;
+        BitField<24, 1, u32> read_only1;
+        BitField<25, 6, RegionType> region2;
+        BitField<31, 1, u32> read_only2;
+    };
+
+    union InterruptPair {
+        static_assert(CapabilityId<CapabilityType::InterruptPair> + 1 == 12);
+
+        RawCapabilityValue raw;
+        BitField<0, 12, CapabilityType> id;
+        BitField<12, 10, u32> interrupt_id0;
+        BitField<22, 10, u32> interrupt_id1;
+    };
+
+    union ProgramType {
+        static_assert(CapabilityId<CapabilityType::ProgramType> + 1 == 14);
+
+        RawCapabilityValue raw;
+        BitField<0, 14, CapabilityType> id;
+        BitField<14, 3, u32> type;
+        BitField<17, 15, u32> reserved;
+    };
+
+    union KernelVersion {
+        static_assert(CapabilityId<CapabilityType::KernelVersion> + 1 == 15);
+
+        RawCapabilityValue raw;
+        BitField<0, 15, CapabilityType> id;
+        BitField<15, 4, u32> major_version;
+        BitField<19, 13, u32> minor_version;
+    };
+
+    union HandleTable {
+        static_assert(CapabilityId<CapabilityType::HandleTable> + 1 == 16);
+
+        RawCapabilityValue raw;
+        BitField<0, 16, CapabilityType> id;
+        BitField<16, 10, u32> size;
+        BitField<26, 6, u32> reserved;
+    };
+
+    union DebugFlags {
+        static_assert(CapabilityId<CapabilityType::DebugFlags> + 1 == 17);
+
+        RawCapabilityValue raw;
+        BitField<0, 17, CapabilityType> id;
+        BitField<17, 1, u32> allow_debug;
+        BitField<18, 1, u32> force_debug;
+        BitField<19, 13, u32> reserved;
+    };
+
+    static_assert(sizeof(CorePriority) == 4);
+    static_assert(sizeof(SyscallMask) == 4);
+    static_assert(sizeof(MapRange) == 4);
+    static_assert(sizeof(MapRangeSize) == 4);
+    static_assert(sizeof(MapIoPage) == 4);
+    static_assert(sizeof(MapRegion) == 4);
+    static_assert(sizeof(InterruptPair) == 4);
+    static_assert(sizeof(ProgramType) == 4);
+    static_assert(sizeof(KernelVersion) == 4);
+    static_assert(sizeof(HandleTable) == 4);
+    static_assert(sizeof(DebugFlags) == 4);
+
+    static constexpr u32 InitializeOnceFlags =
+        CapabilityFlag<CapabilityType::CorePriority> | CapabilityFlag<CapabilityType::ProgramType> |
+        CapabilityFlag<CapabilityType::KernelVersion> |
+        CapabilityFlag<CapabilityType::HandleTable> | CapabilityFlag<CapabilityType::DebugFlags>;
+
+    static const u32 PaddingInterruptId = 0x3FF;
+    static_assert(PaddingInterruptId < InterruptIdCount);
+
+private:
+    constexpr bool SetSvcAllowed(u32 id) {
+        if (id < m_svc_access_flags.size()) [[likely]] {
+            m_svc_access_flags[id] = true;
+            return true;
+        } else {
+            return false;
+        }
+    }
+
+    constexpr bool SetInterruptPermitted(u32 id) {
+        if (id < m_irq_access_flags.size()) [[likely]] {
+            m_irq_access_flags[id] = true;
+            return true;
+        } else {
+            return false;
+        }
+    }
+
+    Result SetCorePriorityCapability(const u32 cap);
+    Result SetSyscallMaskCapability(const u32 cap, u32& set_svc);
+    Result MapRange_(const u32 cap, const u32 size_cap, KPageTable* page_table);
+    Result MapIoPage_(const u32 cap, KPageTable* page_table);
+    Result MapRegion_(const u32 cap, KPageTable* page_table);
+    Result SetInterruptPairCapability(const u32 cap);
+    Result SetProgramTypeCapability(const u32 cap);
+    Result SetKernelVersionCapability(const u32 cap);
+    Result SetHandleTableCapability(const u32 cap);
+    Result SetDebugFlagsCapability(const u32 cap);
+
+    template <typename F>
+    static Result ProcessMapRegionCapability(const u32 cap, F f);
+    static Result CheckMapRegion(KernelCore& kernel, const u32 cap);
+
+    Result SetCapability(const u32 cap, u32& set_flags, u32& set_svc, KPageTable* page_table);
+    Result SetCapabilities(std::span<const u32> caps, KPageTable* page_table);
+
+private:
+    Svc::SvcAccessFlagSet m_svc_access_flags{};
+    InterruptFlagSet m_irq_access_flags{};
+    u64 m_core_mask{};
+    u64 m_phys_core_mask{};
+    u64 m_priority_mask{};
+    u32 m_debug_capabilities{};
+    s32 m_handle_table_size{};
+    u32 m_intended_kernel_version{};
+    u32 m_program_type{};
+};
+
+} // namespace Kernel
diff --git a/src/core/hle/kernel/k_device_address_space.cpp b/src/core/hle/kernel/k_device_address_space.cpp
new file mode 100644
index 000000000..27659ea3b
--- /dev/null
+++ b/src/core/hle/kernel/k_device_address_space.cpp
@@ -0,0 +1,150 @@
+// SPDX-FileCopyrightText: Copyright 2023 yuzu Emulator Project
+// SPDX-License-Identifier: GPL-2.0-or-later
+
+#include "common/assert.h"
+#include "core/core.h"
+#include "core/hle/kernel/k_device_address_space.h"
+#include "core/hle/kernel/kernel.h"
+#include "core/hle/kernel/svc_results.h"
+
+namespace Kernel {
+
+KDeviceAddressSpace::KDeviceAddressSpace(KernelCore& kernel_)
+    : KAutoObjectWithSlabHeapAndContainer(kernel_), m_lock(kernel_), m_is_initialized(false) {}
+KDeviceAddressSpace::~KDeviceAddressSpace() = default;
+
+void KDeviceAddressSpace::Initialize() {
+    // This just forwards to the device page table manager.
+    // KDevicePageTable::Initialize();
+}
+
+// Member functions.
+Result KDeviceAddressSpace::Initialize(u64 address, u64 size) {
+    // Initialize the device page table.
+    // R_TRY(m_table.Initialize(address, size));
+
+    // Set member variables.
+    m_space_address = address;
+    m_space_size = size;
+    m_is_initialized = true;
+
+    R_SUCCEED();
+}
+
+void KDeviceAddressSpace::Finalize() {
+    // Finalize the table.
+    // m_table.Finalize();
+}
+
+Result KDeviceAddressSpace::Attach(Svc::DeviceName device_name) {
+    // Lock the address space.
+    KScopedLightLock lk(m_lock);
+
+    // Attach.
+    // R_RETURN(m_table.Attach(device_name, m_space_address, m_space_size));
+    R_SUCCEED();
+}
+
+Result KDeviceAddressSpace::Detach(Svc::DeviceName device_name) {
+    // Lock the address space.
+    KScopedLightLock lk(m_lock);
+
+    // Detach.
+    // R_RETURN(m_table.Detach(device_name));
+    R_SUCCEED();
+}
+
+Result KDeviceAddressSpace::Map(KPageTable* page_table, VAddr process_address, size_t size,
+                                u64 device_address, u32 option, bool is_aligned) {
+    // Check that the address falls within the space.
+    R_UNLESS((m_space_address <= device_address &&
+              device_address + size - 1 <= m_space_address + m_space_size - 1),
+             ResultInvalidCurrentMemory);
+
+    // Decode the option.
+    const Svc::MapDeviceAddressSpaceOption option_pack{option};
+    const auto device_perm = option_pack.permission.Value();
+    const auto flags = option_pack.flags.Value();
+    const auto reserved = option_pack.reserved.Value();
+
+    // Validate the option.
+    // TODO: It is likely that this check for flags == none is only on NX board.
+    R_UNLESS(flags == Svc::MapDeviceAddressSpaceFlag::None, ResultInvalidEnumValue);
+    R_UNLESS(reserved == 0, ResultInvalidEnumValue);
+
+    // Lock the address space.
+    KScopedLightLock lk(m_lock);
+
+    // Lock the page table to prevent concurrent device mapping operations.
+    // KScopedLightLock pt_lk = page_table->AcquireDeviceMapLock();
+
+    // Lock the pages.
+    bool is_io{};
+    R_TRY(page_table->LockForMapDeviceAddressSpace(std::addressof(is_io), process_address, size,
+                                                   ConvertToKMemoryPermission(device_perm),
+                                                   is_aligned, true));
+
+    // Ensure that if we fail, we don't keep unmapped pages locked.
+    ON_RESULT_FAILURE {
+        ASSERT(page_table->UnlockForDeviceAddressSpace(process_address, size) == ResultSuccess);
+    };
+
+    // Check that the io status is allowable.
+    if (is_io) {
+        R_UNLESS(static_cast<u32>(flags & Svc::MapDeviceAddressSpaceFlag::NotIoRegister) == 0,
+                 ResultInvalidCombination);
+    }
+
+    // Map the pages.
+    {
+        // Perform the mapping.
+        // R_TRY(m_table.Map(page_table, process_address, size, device_address, device_perm,
+        //                   is_aligned, is_io));
+
+        // Ensure that we unmap the pages if we fail to update the protections.
+        // NOTE: Nintendo does not check the result of this unmap call.
+        // ON_RESULT_FAILURE { m_table.Unmap(device_address, size); };
+
+        // Update the protections in accordance with how much we mapped.
+        // R_TRY(page_table->UnlockForDeviceAddressSpacePartialMap(process_address, size));
+    }
+
+    // We succeeded.
+    R_SUCCEED();
+}
+
+Result KDeviceAddressSpace::Unmap(KPageTable* page_table, VAddr process_address, size_t size,
+                                  u64 device_address) {
+    // Check that the address falls within the space.
+    R_UNLESS((m_space_address <= device_address &&
+              device_address + size - 1 <= m_space_address + m_space_size - 1),
+             ResultInvalidCurrentMemory);
+
+    // Lock the address space.
+    KScopedLightLock lk(m_lock);
+
+    // Lock the page table to prevent concurrent device mapping operations.
+    // KScopedLightLock pt_lk = page_table->AcquireDeviceMapLock();
+
+    // Lock the pages.
+    R_TRY(page_table->LockForUnmapDeviceAddressSpace(process_address, size, true));
+
+    // Unmap the pages.
+    {
+        // If we fail to unmap, we want to do a partial unlock.
+        // ON_RESULT_FAILURE {
+        //     ASSERT(page_table->UnlockForDeviceAddressSpacePartialMap(process_address, size) ==
+        //            ResultSuccess);
+        // };
+
+        // Perform the unmap.
+        // R_TRY(m_table.Unmap(page_table, process_address, size, device_address));
+    }
+
+    // Unlock the pages.
+    ASSERT(page_table->UnlockForDeviceAddressSpace(process_address, size) == ResultSuccess);
+
+    R_SUCCEED();
+}
+
+} // namespace Kernel
diff --git a/src/core/hle/kernel/k_device_address_space.h b/src/core/hle/kernel/k_device_address_space.h
new file mode 100644
index 000000000..4709df995
--- /dev/null
+++ b/src/core/hle/kernel/k_device_address_space.h
@@ -0,0 +1,60 @@
+// SPDX-FileCopyrightText: Copyright 2023 yuzu Emulator Project
+// SPDX-License-Identifier: GPL-2.0-or-later
+
+#pragma once
+
+#include <string>
+
+#include "common/common_types.h"
+#include "core/hle/kernel/k_page_table.h"
+#include "core/hle/kernel/slab_helpers.h"
+#include "core/hle/result.h"
+
+namespace Kernel {
+
+class KDeviceAddressSpace final
+    : public KAutoObjectWithSlabHeapAndContainer<KDeviceAddressSpace, KAutoObjectWithList> {
+    KERNEL_AUTOOBJECT_TRAITS(KDeviceAddressSpace, KAutoObject);
+
+public:
+    explicit KDeviceAddressSpace(KernelCore& kernel);
+    ~KDeviceAddressSpace();
+
+    Result Initialize(u64 address, u64 size);
+    void Finalize();
+
+    bool IsInitialized() const {
+        return m_is_initialized;
+    }
+    static void PostDestroy(uintptr_t arg) {}
+
+    Result Attach(Svc::DeviceName device_name);
+    Result Detach(Svc::DeviceName device_name);
+
+    Result MapByForce(KPageTable* page_table, VAddr process_address, size_t size,
+                      u64 device_address, u32 option) {
+        R_RETURN(this->Map(page_table, process_address, size, device_address, option, false));
+    }
+
+    Result MapAligned(KPageTable* page_table, VAddr process_address, size_t size,
+                      u64 device_address, u32 option) {
+        R_RETURN(this->Map(page_table, process_address, size, device_address, option, true));
+    }
+
+    Result Unmap(KPageTable* page_table, VAddr process_address, size_t size, u64 device_address);
+
+    static void Initialize();
+
+private:
+    Result Map(KPageTable* page_table, VAddr process_address, size_t size, u64 device_address,
+               u32 option, bool is_aligned);
+
+private:
+    KLightLock m_lock;
+    // KDevicePageTable m_table;
+    u64 m_space_address{};
+    u64 m_space_size{};
+    bool m_is_initialized{};
+};
+
+} // namespace Kernel
diff --git a/src/core/hle/kernel/k_priority_queue.h b/src/core/hle/kernel/k_priority_queue.h
index cb2512b0b..645c5b531 100644
--- a/src/core/hle/kernel/k_priority_queue.h
+++ b/src/core/hle/kernel/k_priority_queue.h
@@ -17,35 +17,41 @@ namespace Kernel {
 class KThread;
 
 template <typename T>
-concept KPriorityQueueAffinityMask = !std::is_reference_v<T> && requires(T & t) {
-    { t.GetAffinityMask() } -> Common::ConvertibleTo<u64>;
-    {t.SetAffinityMask(0)};
+concept KPriorityQueueAffinityMask = !
+std::is_reference_v<T>&& requires(T& t) {
+                             { t.GetAffinityMask() } -> Common::ConvertibleTo<u64>;
+                             { t.SetAffinityMask(0) };
 
-    { t.GetAffinity(0) } -> std::same_as<bool>;
-    {t.SetAffinity(0, false)};
-    {t.SetAll()};
-};
+                             { t.GetAffinity(0) } -> std::same_as<bool>;
+                             { t.SetAffinity(0, false) };
+                             { t.SetAll() };
+                         };
 
 template <typename T>
-concept KPriorityQueueMember = !std::is_reference_v<T> && requires(T & t) {
-    {typename T::QueueEntry()};
-    {(typename T::QueueEntry()).Initialize()};
-    {(typename T::QueueEntry()).SetPrev(std::addressof(t))};
-    {(typename T::QueueEntry()).SetNext(std::addressof(t))};
-    { (typename T::QueueEntry()).GetNext() } -> std::same_as<T*>;
-    { (typename T::QueueEntry()).GetPrev() } -> std::same_as<T*>;
-    { t.GetPriorityQueueEntry(0) } -> std::same_as<typename T::QueueEntry&>;
-
-    {t.GetAffinityMask()};
-    { std::remove_cvref_t<decltype(t.GetAffinityMask())>() } -> KPriorityQueueAffinityMask;
-
-    { t.GetActiveCore() } -> Common::ConvertibleTo<s32>;
-    { t.GetPriority() } -> Common::ConvertibleTo<s32>;
-    { t.IsDummyThread() } -> Common::ConvertibleTo<bool>;
-};
+concept KPriorityQueueMember = !
+std::is_reference_v<T>&& requires(T& t) {
+                             { typename T::QueueEntry() };
+                             { (typename T::QueueEntry()).Initialize() };
+                             { (typename T::QueueEntry()).SetPrev(std::addressof(t)) };
+                             { (typename T::QueueEntry()).SetNext(std::addressof(t)) };
+                             { (typename T::QueueEntry()).GetNext() } -> std::same_as<T*>;
+                             { (typename T::QueueEntry()).GetPrev() } -> std::same_as<T*>;
+                             {
+                                 t.GetPriorityQueueEntry(0)
+                                 } -> std::same_as<typename T::QueueEntry&>;
+
+                             { t.GetAffinityMask() };
+                             {
+                                 std::remove_cvref_t<decltype(t.GetAffinityMask())>()
+                                 } -> KPriorityQueueAffinityMask;
+
+                             { t.GetActiveCore() } -> Common::ConvertibleTo<s32>;
+                             { t.GetPriority() } -> Common::ConvertibleTo<s32>;
+                             { t.IsDummyThread() } -> Common::ConvertibleTo<bool>;
+                         };
 
 template <typename Member, size_t NumCores_, int LowestPriority, int HighestPriority>
-requires KPriorityQueueMember<Member>
+    requires KPriorityQueueMember<Member>
 class KPriorityQueue {
 public:
     using AffinityMaskType = std::remove_cv_t<
diff --git a/src/core/hle/kernel/k_scoped_lock.h b/src/core/hle/kernel/k_scoped_lock.h
index 857e21156..59b3e32ae 100644
--- a/src/core/hle/kernel/k_scoped_lock.h
+++ b/src/core/hle/kernel/k_scoped_lock.h
@@ -9,13 +9,14 @@
 namespace Kernel {
 
 template <typename T>
-concept KLockable = !std::is_reference_v<T> && requires(T & t) {
-    { t.Lock() } -> std::same_as<void>;
-    { t.Unlock() } -> std::same_as<void>;
-};
+concept KLockable = !
+std::is_reference_v<T>&& requires(T& t) {
+                             { t.Lock() } -> std::same_as<void>;
+                             { t.Unlock() } -> std::same_as<void>;
+                         };
 
 template <typename T>
-requires KLockable<T>
+    requires KLockable<T>
 class [[nodiscard]] KScopedLock {
 public:
     explicit KScopedLock(T* l) : lock_ptr(l) {
diff --git a/src/core/hle/kernel/k_thread.h b/src/core/hle/kernel/k_thread.h
index 9d771de0e..8b8dc51be 100644
--- a/src/core/hle/kernel/k_thread.h
+++ b/src/core/hle/kernel/k_thread.h
@@ -677,7 +677,7 @@ private:
     union SyncObjectBuffer {
         std::array<KSynchronizationObject*, Svc::ArgumentHandleCountMax> sync_objects{};
         std::array<Handle,
-                   Svc::ArgumentHandleCountMax*(sizeof(KSynchronizationObject*) / sizeof(Handle))>
+                   Svc::ArgumentHandleCountMax * (sizeof(KSynchronizationObject*) / sizeof(Handle))>
             handles;
         constexpr SyncObjectBuffer() {}
     };
@@ -698,10 +698,8 @@ private:
         };
 
         template <typename T>
-        requires(
-            std::same_as<T, KThread> ||
-            std::same_as<T, RedBlackKeyType>) static constexpr int Compare(const T& lhs,
-                                                                           const KThread& rhs) {
+            requires(std::same_as<T, KThread> || std::same_as<T, RedBlackKeyType>)
+        static constexpr int Compare(const T& lhs, const KThread& rhs) {
             const u64 l_key = lhs.GetConditionVariableKey();
             const u64 r_key = rhs.GetConditionVariableKey();
 
diff --git a/src/core/hle/kernel/k_thread_local_page.h b/src/core/hle/kernel/k_thread_local_page.h
index fe0cff084..71254eb55 100644
--- a/src/core/hle/kernel/k_thread_local_page.h
+++ b/src/core/hle/kernel/k_thread_local_page.h
@@ -70,10 +70,8 @@ public:
     }
 
     template <typename T>
-    requires(std::same_as<T, KThreadLocalPage> ||
-             std::same_as<T, RedBlackKeyType>) static constexpr int Compare(const T& lhs,
-                                                                            const KThreadLocalPage&
-                                                                                rhs) {
+        requires(std::same_as<T, KThreadLocalPage> || std::same_as<T, RedBlackKeyType>)
+    static constexpr int Compare(const T& lhs, const KThreadLocalPage& rhs) {
         const VAddr lval = GetRedBlackKey(lhs);
         const VAddr rval = GetRedBlackKey(rhs);
 
diff --git a/src/core/hle/kernel/kernel.h b/src/core/hle/kernel/kernel.h
index 8d22f8d2c..5f52e1e95 100644
--- a/src/core/hle/kernel/kernel.h
+++ b/src/core/hle/kernel/kernel.h
@@ -35,6 +35,7 @@ class GlobalSchedulerContext;
 class KAutoObjectWithListContainer;
 class KClientSession;
 class KDebug;
+class KDeviceAddressSpace;
 class KDynamicPageManager;
 class KEvent;
 class KEventInfo;
@@ -359,6 +360,8 @@ public:
             return slab_heap_container->transfer_memory;
         } else if constexpr (std::is_same_v<T, KCodeMemory>) {
             return slab_heap_container->code_memory;
+        } else if constexpr (std::is_same_v<T, KDeviceAddressSpace>) {
+            return slab_heap_container->device_address_space;
         } else if constexpr (std::is_same_v<T, KPageBuffer>) {
             return slab_heap_container->page_buffer;
         } else if constexpr (std::is_same_v<T, KThreadLocalPage>) {
@@ -431,6 +434,7 @@ private:
         KSlabHeap<KThread> thread;
         KSlabHeap<KTransferMemory> transfer_memory;
         KSlabHeap<KCodeMemory> code_memory;
+        KSlabHeap<KDeviceAddressSpace> device_address_space;
         KSlabHeap<KPageBuffer> page_buffer;
         KSlabHeap<KThreadLocalPage> thread_local_page;
         KSlabHeap<KSessionRequest> session_request;
diff --git a/src/core/hle/kernel/svc_types.h b/src/core/hle/kernel/svc_types.h
index 33eebcef6..e90c35601 100644
--- a/src/core/hle/kernel/svc_types.h
+++ b/src/core/hle/kernel/svc_types.h
@@ -3,6 +3,9 @@
 
 #pragma once
 
+#include <bitset>
+
+#include "common/bit_field.h"
 #include "common/common_funcs.h"
 #include "common/common_types.h"
 
@@ -496,6 +499,19 @@ enum class MemoryMapping : u32 {
     Memory = 2,
 };
 
+enum class MapDeviceAddressSpaceFlag : u32 {
+    None = (0U << 0),
+    NotIoRegister = (1U << 0),
+};
+DECLARE_ENUM_FLAG_OPERATORS(MapDeviceAddressSpaceFlag);
+
+union MapDeviceAddressSpaceOption {
+    u32 raw;
+    BitField<0, 16, MemoryPermission> permission;
+    BitField<16, 1, MapDeviceAddressSpaceFlag> flags;
+    BitField<17, 15, u32> reserved;
+};
+
 enum class KernelDebugType : u32 {
     Thread = 0,
     ThreadCallStack = 1,
@@ -592,4 +608,7 @@ struct CreateProcessParameter {
 };
 static_assert(sizeof(CreateProcessParameter) == 0x30);
 
+constexpr size_t NumSupervisorCalls = 0xC0;
+using SvcAccessFlagSet = std::bitset<NumSupervisorCalls>;
+
 } // namespace Kernel::Svc
diff --git a/src/core/hle/kernel/svc_version.h b/src/core/hle/kernel/svc_version.h
new file mode 100644
index 000000000..e4f47b34b
--- /dev/null
+++ b/src/core/hle/kernel/svc_version.h
@@ -0,0 +1,58 @@
+// SPDX-FileCopyrightText: Copyright 2023 yuzu Emulator Project
+// SPDX-License-Identifier: GPL-2.0-or-later
+
+#pragma once
+
+#include "common/bit_field.h"
+#include "common/common_types.h"
+#include "common/literals.h"
+
+namespace Kernel::Svc {
+
+constexpr inline u32 ConvertToSvcMajorVersion(u32 sdk) {
+    return sdk + 4;
+}
+constexpr inline u32 ConvertToSdkMajorVersion(u32 svc) {
+    return svc - 4;
+}
+
+constexpr inline u32 ConvertToSvcMinorVersion(u32 sdk) {
+    return sdk;
+}
+constexpr inline u32 ConvertToSdkMinorVersion(u32 svc) {
+    return svc;
+}
+
+union KernelVersion {
+    u32 value;
+    BitField<0, 4, u32> minor_version;
+    BitField<4, 13, u32> major_version;
+};
+
+constexpr inline u32 EncodeKernelVersion(u32 major, u32 minor) {
+    return decltype(KernelVersion::minor_version)::FormatValue(minor) |
+           decltype(KernelVersion::major_version)::FormatValue(major);
+}
+
+constexpr inline u32 GetKernelMajorVersion(u32 encoded) {
+    return std::bit_cast<decltype(KernelVersion::major_version)>(encoded).Value();
+}
+
+constexpr inline u32 GetKernelMinorVersion(u32 encoded) {
+    return std::bit_cast<decltype(KernelVersion::minor_version)>(encoded).Value();
+}
+
+// Nintendo doesn't support programs targeting SVC versions < 3.0.
+constexpr inline u32 RequiredKernelMajorVersion = 3;
+constexpr inline u32 RequiredKernelMinorVersion = 0;
+constexpr inline u32 RequiredKernelVersion =
+    EncodeKernelVersion(RequiredKernelMajorVersion, RequiredKernelMinorVersion);
+
+// This is the highest SVC version supported, to be updated on new kernel releases.
+// NOTE: Official kernel versions have SVC major = SDK major + 4, SVC minor = SDK minor.
+constexpr inline u32 SupportedKernelMajorVersion = ConvertToSvcMajorVersion(15);
+constexpr inline u32 SupportedKernelMinorVersion = ConvertToSvcMinorVersion(3);
+constexpr inline u32 SupportedKernelVersion =
+    EncodeKernelVersion(SupportedKernelMajorVersion, SupportedKernelMinorVersion);
+
+} // namespace Kernel::Svc
diff --git a/src/core/hle/service/am/am.cpp b/src/core/hle/service/am/am.cpp
index 22999c942..ebcf6e164 100644
--- a/src/core/hle/service/am/am.cpp
+++ b/src/core/hle/service/am/am.cpp
@@ -1124,7 +1124,7 @@ void IStorageAccessor::Write(Kernel::HLERequestContext& ctx) {
     IPC::RequestParser rp{ctx};
 
     const u64 offset{rp.Pop<u64>()};
-    const std::vector<u8> data{ctx.ReadBuffer()};
+    const auto data{ctx.ReadBuffer()};
     const std::size_t size{std::min<u64>(data.size(), backing.GetSize() - offset)};
 
     LOG_DEBUG(Service_AM, "called, offset={}, size={}", offset, size);
diff --git a/src/core/hle/service/audio/audren_u.cpp b/src/core/hle/service/audio/audren_u.cpp
index 3a1c231b6..0ee28752c 100644
--- a/src/core/hle/service/audio/audren_u.cpp
+++ b/src/core/hle/service/audio/audren_u.cpp
@@ -112,7 +112,7 @@ private:
     void RequestUpdate(Kernel::HLERequestContext& ctx) {
         LOG_TRACE(Service_Audio, "called");
 
-        std::vector<u8> input{ctx.ReadBuffer(0)};
+        const auto input{ctx.ReadBuffer(0)};
 
         // These buffers are written manually to avoid an issue with WriteBuffer throwing errors for
         // checking size 0. Performance size is 0 for most games.
diff --git a/src/core/hle/service/audio/hwopus.cpp b/src/core/hle/service/audio/hwopus.cpp
index 825fb8bcc..e01f87356 100644
--- a/src/core/hle/service/audio/hwopus.cpp
+++ b/src/core/hle/service/audio/hwopus.cpp
@@ -93,7 +93,7 @@ private:
         ctx.WriteBuffer(samples);
     }
 
-    bool DecodeOpusData(u32& consumed, u32& sample_count, const std::vector<u8>& input,
+    bool DecodeOpusData(u32& consumed, u32& sample_count, std::span<const u8> input,
                         std::vector<opus_int16>& output, u64* out_performance_time) const {
         const auto start_time = std::chrono::steady_clock::now();
         const std::size_t raw_output_sz = output.size() * sizeof(opus_int16);
diff --git a/src/core/hle/service/es/es.cpp b/src/core/hle/service/es/es.cpp
index d183e5829..fb8686859 100644
--- a/src/core/hle/service/es/es.cpp
+++ b/src/core/hle/service/es/es.cpp
@@ -122,7 +122,7 @@ private:
 
     void ImportTicket(Kernel::HLERequestContext& ctx) {
         const auto ticket = ctx.ReadBuffer();
-        const auto cert = ctx.ReadBuffer(1);
+        [[maybe_unused]] const auto cert = ctx.ReadBuffer(1);
 
         if (ticket.size() < sizeof(Core::Crypto::Ticket)) {
             LOG_ERROR(Service_ETicket, "The input buffer is not large enough!");
diff --git a/src/core/hle/service/filesystem/fsp_srv.cpp b/src/core/hle/service/filesystem/fsp_srv.cpp
index fbb16a7da..447d624e1 100644
--- a/src/core/hle/service/filesystem/fsp_srv.cpp
+++ b/src/core/hle/service/filesystem/fsp_srv.cpp
@@ -190,7 +190,7 @@ private:
             return;
         }
 
-        const std::vector<u8> data = ctx.ReadBuffer();
+        const auto data = ctx.ReadBuffer();
 
         ASSERT_MSG(
             static_cast<s64>(data.size()) <= length,
@@ -401,11 +401,8 @@ public:
     }
 
     void RenameFile(Kernel::HLERequestContext& ctx) {
-        std::vector<u8> buffer = ctx.ReadBuffer(0);
-        const std::string src_name = Common::StringFromBuffer(buffer);
-
-        buffer = ctx.ReadBuffer(1);
-        const std::string dst_name = Common::StringFromBuffer(buffer);
+        const std::string src_name = Common::StringFromBuffer(ctx.ReadBuffer(0));
+        const std::string dst_name = Common::StringFromBuffer(ctx.ReadBuffer(1));
 
         LOG_DEBUG(Service_FS, "called. file '{}' to file '{}'", src_name, dst_name);
 
@@ -1086,7 +1083,7 @@ void FSP_SRV::GetGlobalAccessLogMode(Kernel::HLERequestContext& ctx) {
 }
 
 void FSP_SRV::OutputAccessLogToSdCard(Kernel::HLERequestContext& ctx) {
-    const auto raw = ctx.ReadBuffer();
+    const auto raw = ctx.ReadBufferCopy();
     auto log = Common::StringFromFixedZeroTerminatedBuffer(
         reinterpret_cast<const char*>(raw.data()), raw.size());
 
diff --git a/src/core/hle/service/glue/arp.cpp b/src/core/hle/service/glue/arp.cpp
index 49b6d45fe..ce21b69e3 100644
--- a/src/core/hle/service/glue/arp.cpp
+++ b/src/core/hle/service/glue/arp.cpp
@@ -228,7 +228,8 @@ private:
             return;
         }
 
-        control = ctx.ReadBuffer();
+        // TODO: Can this be a span?
+        control = ctx.ReadBufferCopy();
 
         IPC::ResponseBuilder rb{ctx, 2};
         rb.Push(ResultSuccess);
diff --git a/src/core/hle/service/hid/controllers/npad.cpp b/src/core/hle/service/hid/controllers/npad.cpp
index 5713f1288..513ea485a 100644
--- a/src/core/hle/service/hid/controllers/npad.cpp
+++ b/src/core/hle/service/hid/controllers/npad.cpp
@@ -428,6 +428,9 @@ void Controller_NPad::RequestPadStateUpdate(Core::HID::NpadIdType npad_id) {
         return;
     }
 
+    // This function is unique to yuzu for the turbo buttons to work properly
+    controller.device->TurboButtonUpdate();
+
     auto& pad_entry = controller.npad_pad_state;
     auto& trigger_entry = controller.npad_trigger_state;
     const auto button_state = controller.device->GetNpadButtons();
@@ -755,11 +758,12 @@ Core::HID::NpadStyleTag Controller_NPad::GetSupportedStyleSet() const {
     return hid_core.GetSupportedStyleTag();
 }
 
-void Controller_NPad::SetSupportedNpadIdTypes(u8* data, std::size_t length) {
+void Controller_NPad::SetSupportedNpadIdTypes(std::span<const u8> data) {
+    const auto length = data.size();
     ASSERT(length > 0 && (length % sizeof(u32)) == 0);
     supported_npad_id_types.clear();
     supported_npad_id_types.resize(length / sizeof(u32));
-    std::memcpy(supported_npad_id_types.data(), data, length);
+    std::memcpy(supported_npad_id_types.data(), data.data(), length);
 }
 
 void Controller_NPad::GetSupportedNpadIdTypes(u32* data, std::size_t max_length) {
diff --git a/src/core/hle/service/hid/controllers/npad.h b/src/core/hle/service/hid/controllers/npad.h
index 1a589cca2..1f7d33459 100644
--- a/src/core/hle/service/hid/controllers/npad.h
+++ b/src/core/hle/service/hid/controllers/npad.h
@@ -6,6 +6,7 @@
 #include <array>
 #include <atomic>
 #include <mutex>
+#include <span>
 
 #include "common/bit_field.h"
 #include "common/common_types.h"
@@ -95,7 +96,7 @@ public:
     void SetSupportedStyleSet(Core::HID::NpadStyleTag style_set);
     Core::HID::NpadStyleTag GetSupportedStyleSet() const;
 
-    void SetSupportedNpadIdTypes(u8* data, std::size_t length);
+    void SetSupportedNpadIdTypes(std::span<const u8> data);
     void GetSupportedNpadIdTypes(u32* data, std::size_t max_length);
     std::size_t GetSupportedNpadIdTypesSize() const;
 
diff --git a/src/core/hle/service/hid/hid.cpp b/src/core/hle/service/hid/hid.cpp
index bf28440c6..f15f1a6bb 100644
--- a/src/core/hle/service/hid/hid.cpp
+++ b/src/core/hle/service/hid/hid.cpp
@@ -1026,7 +1026,7 @@ void Hid::SetSupportedNpadIdType(Kernel::HLERequestContext& ctx) {
     const auto applet_resource_user_id{rp.Pop<u64>()};
 
     applet_resource->GetController<Controller_NPad>(HidController::NPad)
-        .SetSupportedNpadIdTypes(ctx.ReadBuffer().data(), ctx.GetReadBufferSize());
+        .SetSupportedNpadIdTypes(ctx.ReadBuffer());
 
     LOG_DEBUG(Service_HID, "called, applet_resource_user_id={}", applet_resource_user_id);
 
@@ -2104,7 +2104,7 @@ void Hid::WritePalmaRgbLedPatternEntry(Kernel::HLERequestContext& ctx) {
     const auto connection_handle{rp.PopRaw<Controller_Palma::PalmaConnectionHandle>()};
     const auto unknown{rp.Pop<u64>()};
 
-    const auto buffer = ctx.ReadBuffer();
+    [[maybe_unused]] const auto buffer = ctx.ReadBuffer();
 
     LOG_WARNING(Service_HID, "(STUBBED) called, connection_handle={}, unknown={}",
                 connection_handle.npad_id, unknown);
diff --git a/src/core/hle/service/hid/hidbus/hidbus_base.h b/src/core/hle/service/hid/hidbus/hidbus_base.h
index d3960f506..65e301137 100644
--- a/src/core/hle/service/hid/hidbus/hidbus_base.h
+++ b/src/core/hle/service/hid/hidbus/hidbus_base.h
@@ -4,6 +4,7 @@
 #pragma once
 
 #include <array>
+#include <span>
 #include "common/common_types.h"
 #include "core/hle/result.h"
 
@@ -150,7 +151,7 @@ public:
     }
 
     // Assigns a command from data
-    virtual bool SetCommand(const std::vector<u8>& data) {
+    virtual bool SetCommand(std::span<const u8> data) {
         return {};
     }
 
diff --git a/src/core/hle/service/hid/hidbus/ringcon.cpp b/src/core/hle/service/hid/hidbus/ringcon.cpp
index 78ed47014..35847cbdd 100644
--- a/src/core/hle/service/hid/hidbus/ringcon.cpp
+++ b/src/core/hle/service/hid/hidbus/ringcon.cpp
@@ -116,7 +116,7 @@ std::vector<u8> RingController::GetReply() const {
     }
 }
 
-bool RingController::SetCommand(const std::vector<u8>& data) {
+bool RingController::SetCommand(std::span<const u8> data) {
     if (data.size() < 4) {
         LOG_ERROR(Service_HID, "Command size not supported {}", data.size());
         command = RingConCommands::Error;
diff --git a/src/core/hle/service/hid/hidbus/ringcon.h b/src/core/hle/service/hid/hidbus/ringcon.h
index 845ce85a5..c2fb386b1 100644
--- a/src/core/hle/service/hid/hidbus/ringcon.h
+++ b/src/core/hle/service/hid/hidbus/ringcon.h
@@ -4,6 +4,7 @@
 #pragma once
 
 #include <array>
+#include <span>
 
 #include "common/common_types.h"
 #include "core/hle/service/hid/hidbus/hidbus_base.h"
@@ -31,7 +32,7 @@ public:
     u8 GetDeviceId() const override;
 
     // Assigns a command from data
-    bool SetCommand(const std::vector<u8>& data) override;
+    bool SetCommand(std::span<const u8> data) override;
 
     // Returns a reply from a command
     std::vector<u8> GetReply() const override;
diff --git a/src/core/hle/service/hid/hidbus/starlink.cpp b/src/core/hle/service/hid/hidbus/starlink.cpp
index dd439f60a..d0e760314 100644
--- a/src/core/hle/service/hid/hidbus/starlink.cpp
+++ b/src/core/hle/service/hid/hidbus/starlink.cpp
@@ -42,7 +42,7 @@ std::vector<u8> Starlink::GetReply() const {
     return {};
 }
 
-bool Starlink::SetCommand(const std::vector<u8>& data) {
+bool Starlink::SetCommand(std::span<const u8> data) {
     LOG_ERROR(Service_HID, "Command not implemented");
     return false;
 }
diff --git a/src/core/hle/service/hid/hidbus/starlink.h b/src/core/hle/service/hid/hidbus/starlink.h
index 0b1b7ba49..07c800e6e 100644
--- a/src/core/hle/service/hid/hidbus/starlink.h
+++ b/src/core/hle/service/hid/hidbus/starlink.h
@@ -29,7 +29,7 @@ public:
     u8 GetDeviceId() const override;
 
     // Assigns a command from data
-    bool SetCommand(const std::vector<u8>& data) override;
+    bool SetCommand(std::span<const u8> data) override;
 
     // Returns a reply from a command
     std::vector<u8> GetReply() const override;
diff --git a/src/core/hle/service/hid/hidbus/stubbed.cpp b/src/core/hle/service/hid/hidbus/stubbed.cpp
index e477443e3..07632c872 100644
--- a/src/core/hle/service/hid/hidbus/stubbed.cpp
+++ b/src/core/hle/service/hid/hidbus/stubbed.cpp
@@ -43,7 +43,7 @@ std::vector<u8> HidbusStubbed::GetReply() const {
     return {};
 }
 
-bool HidbusStubbed::SetCommand(const std::vector<u8>& data) {
+bool HidbusStubbed::SetCommand(std::span<const u8> data) {
     LOG_ERROR(Service_HID, "Command not implemented");
     return false;
 }
diff --git a/src/core/hle/service/hid/hidbus/stubbed.h b/src/core/hle/service/hid/hidbus/stubbed.h
index 91165ceff..38eaa0ecc 100644
--- a/src/core/hle/service/hid/hidbus/stubbed.h
+++ b/src/core/hle/service/hid/hidbus/stubbed.h
@@ -29,7 +29,7 @@ public:
     u8 GetDeviceId() const override;
 
     // Assigns a command from data
-    bool SetCommand(const std::vector<u8>& data) override;
+    bool SetCommand(std::span<const u8> data) override;
 
     // Returns a reply from a command
     std::vector<u8> GetReply() const override;
diff --git a/src/core/hle/service/jit/jit.cpp b/src/core/hle/service/jit/jit.cpp
index 8f2920c51..1295a44c7 100644
--- a/src/core/hle/service/jit/jit.cpp
+++ b/src/core/hle/service/jit/jit.cpp
@@ -62,7 +62,7 @@ public:
         const auto parameters{rp.PopRaw<InputParameters>()};
 
         // Optional input/output buffers
-        std::vector<u8> input_buffer{ctx.CanReadBuffer() ? ctx.ReadBuffer() : std::vector<u8>()};
+        const auto input_buffer{ctx.CanReadBuffer() ? ctx.ReadBuffer() : std::span<const u8>()};
         std::vector<u8> output_buffer(ctx.CanWriteBuffer() ? ctx.GetWriteBufferSize() : 0);
 
         // Function call prototype:
@@ -132,7 +132,7 @@ public:
         const auto command{rp.PopRaw<u64>()};
 
         // Optional input/output buffers
-        std::vector<u8> input_buffer{ctx.CanReadBuffer() ? ctx.ReadBuffer() : std::vector<u8>()};
+        const auto input_buffer{ctx.CanReadBuffer() ? ctx.ReadBuffer() : std::span<const u8>()};
         std::vector<u8> output_buffer(ctx.CanWriteBuffer() ? ctx.GetWriteBufferSize() : 0);
 
         // Function call prototype:
diff --git a/src/core/hle/service/ldn/ldn.cpp b/src/core/hle/service/ldn/ldn.cpp
index c49c61cff..e5099d61f 100644
--- a/src/core/hle/service/ldn/ldn.cpp
+++ b/src/core/hle/service/ldn/ldn.cpp
@@ -412,7 +412,7 @@ public:
     }
 
     void SetAdvertiseData(Kernel::HLERequestContext& ctx) {
-        std::vector<u8> read_buffer = ctx.ReadBuffer();
+        const auto read_buffer = ctx.ReadBuffer();
 
         IPC::ResponseBuilder rb{ctx, 2};
         rb.Push(lan_discovery.SetAdvertiseData(read_buffer));
@@ -464,7 +464,7 @@ public:
                  parameters.security_config.passphrase_size,
                  parameters.security_config.security_mode, parameters.local_communication_version);
 
-        const std::vector<u8> read_buffer = ctx.ReadBuffer();
+        const auto read_buffer = ctx.ReadBuffer();
         if (read_buffer.size() != sizeof(NetworkInfo)) {
             LOG_ERROR(Frontend, "NetworkInfo doesn't match read_buffer size!");
             IPC::ResponseBuilder rb{ctx, 2};
diff --git a/src/core/hle/service/nvdrv/devices/nvdevice.h b/src/core/hle/service/nvdrv/devices/nvdevice.h
index 204b0e757..c562e04d2 100644
--- a/src/core/hle/service/nvdrv/devices/nvdevice.h
+++ b/src/core/hle/service/nvdrv/devices/nvdevice.h
@@ -3,7 +3,9 @@
 
 #pragma once
 
+#include <span>
 #include <vector>
+
 #include "common/common_types.h"
 #include "core/hle/service/nvdrv/nvdata.h"
 
@@ -31,7 +33,7 @@ public:
      * @param output A buffer where the output data will be written to.
      * @returns The result code of the ioctl.
      */
-    virtual NvResult Ioctl1(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+    virtual NvResult Ioctl1(DeviceFD fd, Ioctl command, std::span<const u8> input,
                             std::vector<u8>& output) = 0;
 
     /**
@@ -42,8 +44,8 @@ public:
      * @param output A buffer where the output data will be written to.
      * @returns The result code of the ioctl.
      */
-    virtual NvResult Ioctl2(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
-                            const std::vector<u8>& inline_input, std::vector<u8>& output) = 0;
+    virtual NvResult Ioctl2(DeviceFD fd, Ioctl command, std::span<const u8> input,
+                            std::span<const u8> inline_input, std::vector<u8>& output) = 0;
 
     /**
      * Handles an ioctl3 request.
@@ -53,7 +55,7 @@ public:
      * @param inline_output A buffer where the inlined output data will be written to.
      * @returns The result code of the ioctl.
      */
-    virtual NvResult Ioctl3(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+    virtual NvResult Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input,
                             std::vector<u8>& output, std::vector<u8>& inline_output) = 0;
 
     /**
diff --git a/src/core/hle/service/nvdrv/devices/nvdisp_disp0.cpp b/src/core/hle/service/nvdrv/devices/nvdisp_disp0.cpp
index 4122fc98d..5a5b2e305 100644
--- a/src/core/hle/service/nvdrv/devices/nvdisp_disp0.cpp
+++ b/src/core/hle/service/nvdrv/devices/nvdisp_disp0.cpp
@@ -17,19 +17,19 @@ nvdisp_disp0::nvdisp_disp0(Core::System& system_, NvCore::Container& core)
     : nvdevice{system_}, container{core}, nvmap{core.GetNvMapFile()} {}
 nvdisp_disp0::~nvdisp_disp0() = default;
 
-NvResult nvdisp_disp0::Ioctl1(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+NvResult nvdisp_disp0::Ioctl1(DeviceFD fd, Ioctl command, std::span<const u8> input,
                               std::vector<u8>& output) {
     UNIMPLEMENTED_MSG("Unimplemented ioctl={:08X}", command.raw);
     return NvResult::NotImplemented;
 }
 
-NvResult nvdisp_disp0::Ioctl2(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
-                              const std::vector<u8>& inline_input, std::vector<u8>& output) {
+NvResult nvdisp_disp0::Ioctl2(DeviceFD fd, Ioctl command, std::span<const u8> input,
+                              std::span<const u8> inline_input, std::vector<u8>& output) {
     UNIMPLEMENTED_MSG("Unimplemented ioctl={:08X}", command.raw);
     return NvResult::NotImplemented;
 }
 
-NvResult nvdisp_disp0::Ioctl3(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+NvResult nvdisp_disp0::Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input,
                               std::vector<u8>& output, std::vector<u8>& inline_output) {
     UNIMPLEMENTED_MSG("Unimplemented ioctl={:08X}", command.raw);
     return NvResult::NotImplemented;
diff --git a/src/core/hle/service/nvdrv/devices/nvdisp_disp0.h b/src/core/hle/service/nvdrv/devices/nvdisp_disp0.h
index 04217ab12..81bd7960a 100644
--- a/src/core/hle/service/nvdrv/devices/nvdisp_disp0.h
+++ b/src/core/hle/service/nvdrv/devices/nvdisp_disp0.h
@@ -25,12 +25,12 @@ public:
     explicit nvdisp_disp0(Core::System& system_, NvCore::Container& core);
     ~nvdisp_disp0() override;
 
-    NvResult Ioctl1(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+    NvResult Ioctl1(DeviceFD fd, Ioctl command, std::span<const u8> input,
                     std::vector<u8>& output) override;
-    NvResult Ioctl2(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
-                    const std::vector<u8>& inline_input, std::vector<u8>& output) override;
-    NvResult Ioctl3(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
-                    std::vector<u8>& output, std::vector<u8>& inline_output) override;
+    NvResult Ioctl2(DeviceFD fd, Ioctl command, std::span<const u8> input,
+                    std::span<const u8> inline_input, std::vector<u8>& output) override;
+    NvResult Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input, std::vector<u8>& output,
+                    std::vector<u8>& inline_output) override;
 
     void OnOpen(DeviceFD fd) override;
     void OnClose(DeviceFD fd) override;
diff --git a/src/core/hle/service/nvdrv/devices/nvhost_as_gpu.cpp b/src/core/hle/service/nvdrv/devices/nvhost_as_gpu.cpp
index b635e6ed1..681bd0867 100644
--- a/src/core/hle/service/nvdrv/devices/nvhost_as_gpu.cpp
+++ b/src/core/hle/service/nvdrv/devices/nvhost_as_gpu.cpp
@@ -27,7 +27,7 @@ nvhost_as_gpu::nvhost_as_gpu(Core::System& system_, Module& module_, NvCore::Con
 
 nvhost_as_gpu::~nvhost_as_gpu() = default;
 
-NvResult nvhost_as_gpu::Ioctl1(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+NvResult nvhost_as_gpu::Ioctl1(DeviceFD fd, Ioctl command, std::span<const u8> input,
                                std::vector<u8>& output) {
     switch (command.group) {
     case 'A':
@@ -60,13 +60,13 @@ NvResult nvhost_as_gpu::Ioctl1(DeviceFD fd, Ioctl command, const std::vector<u8>
     return NvResult::NotImplemented;
 }
 
-NvResult nvhost_as_gpu::Ioctl2(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
-                               const std::vector<u8>& inline_input, std::vector<u8>& output) {
+NvResult nvhost_as_gpu::Ioctl2(DeviceFD fd, Ioctl command, std::span<const u8> input,
+                               std::span<const u8> inline_input, std::vector<u8>& output) {
     UNIMPLEMENTED_MSG("Unimplemented ioctl={:08X}", command.raw);
     return NvResult::NotImplemented;
 }
 
-NvResult nvhost_as_gpu::Ioctl3(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+NvResult nvhost_as_gpu::Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input,
                                std::vector<u8>& output, std::vector<u8>& inline_output) {
     switch (command.group) {
     case 'A':
@@ -87,7 +87,7 @@ NvResult nvhost_as_gpu::Ioctl3(DeviceFD fd, Ioctl command, const std::vector<u8>
 void nvhost_as_gpu::OnOpen(DeviceFD fd) {}
 void nvhost_as_gpu::OnClose(DeviceFD fd) {}
 
-NvResult nvhost_as_gpu::AllocAsEx(const std::vector<u8>& input, std::vector<u8>& output) {
+NvResult nvhost_as_gpu::AllocAsEx(std::span<const u8> input, std::vector<u8>& output) {
     IoctlAllocAsEx params{};
     std::memcpy(&params, input.data(), input.size());
 
@@ -141,7 +141,7 @@ NvResult nvhost_as_gpu::AllocAsEx(const std::vector<u8>& input, std::vector<u8>&
     return NvResult::Success;
 }
 
-NvResult nvhost_as_gpu::AllocateSpace(const std::vector<u8>& input, std::vector<u8>& output) {
+NvResult nvhost_as_gpu::AllocateSpace(std::span<const u8> input, std::vector<u8>& output) {
     IoctlAllocSpace params{};
     std::memcpy(&params, input.data(), input.size());
 
@@ -220,7 +220,7 @@ void nvhost_as_gpu::FreeMappingLocked(u64 offset) {
     mapping_map.erase(offset);
 }
 
-NvResult nvhost_as_gpu::FreeSpace(const std::vector<u8>& input, std::vector<u8>& output) {
+NvResult nvhost_as_gpu::FreeSpace(std::span<const u8> input, std::vector<u8>& output) {
     IoctlFreeSpace params{};
     std::memcpy(&params, input.data(), input.size());
 
@@ -266,7 +266,7 @@ NvResult nvhost_as_gpu::FreeSpace(const std::vector<u8>& input, std::vector<u8>&
     return NvResult::Success;
 }
 
-NvResult nvhost_as_gpu::Remap(const std::vector<u8>& input, std::vector<u8>& output) {
+NvResult nvhost_as_gpu::Remap(std::span<const u8> input, std::vector<u8>& output) {
     const auto num_entries = input.size() / sizeof(IoctlRemapEntry);
 
     LOG_DEBUG(Service_NVDRV, "called, num_entries=0x{:X}", num_entries);
@@ -320,7 +320,7 @@ NvResult nvhost_as_gpu::Remap(const std::vector<u8>& input, std::vector<u8>& out
     return NvResult::Success;
 }
 
-NvResult nvhost_as_gpu::MapBufferEx(const std::vector<u8>& input, std::vector<u8>& output) {
+NvResult nvhost_as_gpu::MapBufferEx(std::span<const u8> input, std::vector<u8>& output) {
     IoctlMapBufferEx params{};
     std::memcpy(&params, input.data(), input.size());
 
@@ -424,7 +424,7 @@ NvResult nvhost_as_gpu::MapBufferEx(const std::vector<u8>& input, std::vector<u8
     return NvResult::Success;
 }
 
-NvResult nvhost_as_gpu::UnmapBuffer(const std::vector<u8>& input, std::vector<u8>& output) {
+NvResult nvhost_as_gpu::UnmapBuffer(std::span<const u8> input, std::vector<u8>& output) {
     IoctlUnmapBuffer params{};
     std::memcpy(&params, input.data(), input.size());
 
@@ -463,7 +463,7 @@ NvResult nvhost_as_gpu::UnmapBuffer(const std::vector<u8>& input, std::vector<u8
     return NvResult::Success;
 }
 
-NvResult nvhost_as_gpu::BindChannel(const std::vector<u8>& input, std::vector<u8>& output) {
+NvResult nvhost_as_gpu::BindChannel(std::span<const u8> input, std::vector<u8>& output) {
     IoctlBindChannel params{};
     std::memcpy(&params, input.data(), input.size());
     LOG_DEBUG(Service_NVDRV, "called, fd={:X}", params.fd);
@@ -492,7 +492,7 @@ void nvhost_as_gpu::GetVARegionsImpl(IoctlGetVaRegions& params) {
     };
 }
 
-NvResult nvhost_as_gpu::GetVARegions(const std::vector<u8>& input, std::vector<u8>& output) {
+NvResult nvhost_as_gpu::GetVARegions(std::span<const u8> input, std::vector<u8>& output) {
     IoctlGetVaRegions params{};
     std::memcpy(&params, input.data(), input.size());
 
@@ -511,7 +511,7 @@ NvResult nvhost_as_gpu::GetVARegions(const std::vector<u8>& input, std::vector<u
     return NvResult::Success;
 }
 
-NvResult nvhost_as_gpu::GetVARegions(const std::vector<u8>& input, std::vector<u8>& output,
+NvResult nvhost_as_gpu::GetVARegions(std::span<const u8> input, std::vector<u8>& output,
                                      std::vector<u8>& inline_output) {
     IoctlGetVaRegions params{};
     std::memcpy(&params, input.data(), input.size());
diff --git a/src/core/hle/service/nvdrv/devices/nvhost_as_gpu.h b/src/core/hle/service/nvdrv/devices/nvhost_as_gpu.h
index 86fe71c75..1aba8d579 100644
--- a/src/core/hle/service/nvdrv/devices/nvhost_as_gpu.h
+++ b/src/core/hle/service/nvdrv/devices/nvhost_as_gpu.h
@@ -47,12 +47,12 @@ public:
     explicit nvhost_as_gpu(Core::System& system_, Module& module, NvCore::Container& core);
     ~nvhost_as_gpu() override;
 
-    NvResult Ioctl1(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+    NvResult Ioctl1(DeviceFD fd, Ioctl command, std::span<const u8> input,
                     std::vector<u8>& output) override;
-    NvResult Ioctl2(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
-                    const std::vector<u8>& inline_input, std::vector<u8>& output) override;
-    NvResult Ioctl3(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
-                    std::vector<u8>& output, std::vector<u8>& inline_output) override;
+    NvResult Ioctl2(DeviceFD fd, Ioctl command, std::span<const u8> input,
+                    std::span<const u8> inline_input, std::vector<u8>& output) override;
+    NvResult Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input, std::vector<u8>& output,
+                    std::vector<u8>& inline_output) override;
 
     void OnOpen(DeviceFD fd) override;
     void OnClose(DeviceFD fd) override;
@@ -138,17 +138,17 @@ private:
     static_assert(sizeof(IoctlGetVaRegions) == 16 + sizeof(VaRegion) * 2,
                   "IoctlGetVaRegions is incorrect size");
 
-    NvResult AllocAsEx(const std::vector<u8>& input, std::vector<u8>& output);
-    NvResult AllocateSpace(const std::vector<u8>& input, std::vector<u8>& output);
-    NvResult Remap(const std::vector<u8>& input, std::vector<u8>& output);
-    NvResult MapBufferEx(const std::vector<u8>& input, std::vector<u8>& output);
-    NvResult UnmapBuffer(const std::vector<u8>& input, std::vector<u8>& output);
-    NvResult FreeSpace(const std::vector<u8>& input, std::vector<u8>& output);
-    NvResult BindChannel(const std::vector<u8>& input, std::vector<u8>& output);
+    NvResult AllocAsEx(std::span<const u8> input, std::vector<u8>& output);
+    NvResult AllocateSpace(std::span<const u8> input, std::vector<u8>& output);
+    NvResult Remap(std::span<const u8> input, std::vector<u8>& output);
+    NvResult MapBufferEx(std::span<const u8> input, std::vector<u8>& output);
+    NvResult UnmapBuffer(std::span<const u8> input, std::vector<u8>& output);
+    NvResult FreeSpace(std::span<const u8> input, std::vector<u8>& output);
+    NvResult BindChannel(std::span<const u8> input, std::vector<u8>& output);
 
     void GetVARegionsImpl(IoctlGetVaRegions& params);
-    NvResult GetVARegions(const std::vector<u8>& input, std::vector<u8>& output);
-    NvResult GetVARegions(const std::vector<u8>& input, std::vector<u8>& output,
+    NvResult GetVARegions(std::span<const u8> input, std::vector<u8>& output);
+    NvResult GetVARegions(std::span<const u8> input, std::vector<u8>& output,
                           std::vector<u8>& inline_output);
 
     void FreeMappingLocked(u64 offset);
diff --git a/src/core/hle/service/nvdrv/devices/nvhost_ctrl.cpp b/src/core/hle/service/nvdrv/devices/nvhost_ctrl.cpp
index eee11fab8..0cdde82a7 100644
--- a/src/core/hle/service/nvdrv/devices/nvhost_ctrl.cpp
+++ b/src/core/hle/service/nvdrv/devices/nvhost_ctrl.cpp
@@ -34,7 +34,7 @@ nvhost_ctrl::~nvhost_ctrl() {
     }
 }
 
-NvResult nvhost_ctrl::Ioctl1(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+NvResult nvhost_ctrl::Ioctl1(DeviceFD fd, Ioctl command, std::span<const u8> input,
                              std::vector<u8>& output) {
     switch (command.group) {
     case 0x0:
@@ -63,13 +63,13 @@ NvResult nvhost_ctrl::Ioctl1(DeviceFD fd, Ioctl command, const std::vector<u8>&
     return NvResult::NotImplemented;
 }
 
-NvResult nvhost_ctrl::Ioctl2(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
-                             const std::vector<u8>& inline_input, std::vector<u8>& output) {
+NvResult nvhost_ctrl::Ioctl2(DeviceFD fd, Ioctl command, std::span<const u8> input,
+                             std::span<const u8> inline_input, std::vector<u8>& output) {
     UNIMPLEMENTED_MSG("Unimplemented ioctl={:08X}", command.raw);
     return NvResult::NotImplemented;
 }
 
-NvResult nvhost_ctrl::Ioctl3(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+NvResult nvhost_ctrl::Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input,
                              std::vector<u8>& output, std::vector<u8>& inline_outpu) {
     UNIMPLEMENTED_MSG("Unimplemented ioctl={:08X}", command.raw);
     return NvResult::NotImplemented;
@@ -79,7 +79,7 @@ void nvhost_ctrl::OnOpen(DeviceFD fd) {}
 
 void nvhost_ctrl::OnClose(DeviceFD fd) {}
 
-NvResult nvhost_ctrl::NvOsGetConfigU32(const std::vector<u8>& input, std::vector<u8>& output) {
+NvResult nvhost_ctrl::NvOsGetConfigU32(std::span<const u8> input, std::vector<u8>& output) {
     IocGetConfigParams params{};
     std::memcpy(&params, input.data(), sizeof(params));
     LOG_TRACE(Service_NVDRV, "called, setting={}!{}", params.domain_str.data(),
@@ -87,7 +87,7 @@ NvResult nvhost_ctrl::NvOsGetConfigU32(const std::vector<u8>& input, std::vector
     return NvResult::ConfigVarNotFound; // Returns error on production mode
 }
 
-NvResult nvhost_ctrl::IocCtrlEventWait(const std::vector<u8>& input, std::vector<u8>& output,
+NvResult nvhost_ctrl::IocCtrlEventWait(std::span<const u8> input, std::vector<u8>& output,
                                        bool is_allocation) {
     IocCtrlEventWaitParams params{};
     std::memcpy(&params, input.data(), sizeof(params));
@@ -231,7 +231,7 @@ NvResult nvhost_ctrl::FreeEvent(u32 slot) {
     return NvResult::Success;
 }
 
-NvResult nvhost_ctrl::IocCtrlEventRegister(const std::vector<u8>& input, std::vector<u8>& output) {
+NvResult nvhost_ctrl::IocCtrlEventRegister(std::span<const u8> input, std::vector<u8>& output) {
     IocCtrlEventRegisterParams params{};
     std::memcpy(&params, input.data(), sizeof(params));
     const u32 event_id = params.user_event_id;
@@ -252,8 +252,7 @@ NvResult nvhost_ctrl::IocCtrlEventRegister(const std::vector<u8>& input, std::ve
     return NvResult::Success;
 }
 
-NvResult nvhost_ctrl::IocCtrlEventUnregister(const std::vector<u8>& input,
-                                             std::vector<u8>& output) {
+NvResult nvhost_ctrl::IocCtrlEventUnregister(std::span<const u8> input, std::vector<u8>& output) {
     IocCtrlEventUnregisterParams params{};
     std::memcpy(&params, input.data(), sizeof(params));
     const u32 event_id = params.user_event_id & 0x00FF;
@@ -263,7 +262,7 @@ NvResult nvhost_ctrl::IocCtrlEventUnregister(const std::vector<u8>& input,
     return FreeEvent(event_id);
 }
 
-NvResult nvhost_ctrl::IocCtrlEventUnregisterBatch(const std::vector<u8>& input,
+NvResult nvhost_ctrl::IocCtrlEventUnregisterBatch(std::span<const u8> input,
                                                   std::vector<u8>& output) {
     IocCtrlEventUnregisterBatchParams params{};
     std::memcpy(&params, input.data(), sizeof(params));
@@ -282,7 +281,7 @@ NvResult nvhost_ctrl::IocCtrlEventUnregisterBatch(const std::vector<u8>& input,
     return NvResult::Success;
 }
 
-NvResult nvhost_ctrl::IocCtrlClearEventWait(const std::vector<u8>& input, std::vector<u8>& output) {
+NvResult nvhost_ctrl::IocCtrlClearEventWait(std::span<const u8> input, std::vector<u8>& output) {
     IocCtrlEventClearParams params{};
     std::memcpy(&params, input.data(), sizeof(params));
 
diff --git a/src/core/hle/service/nvdrv/devices/nvhost_ctrl.h b/src/core/hle/service/nvdrv/devices/nvhost_ctrl.h
index 0b56d7070..dd2e7888a 100644
--- a/src/core/hle/service/nvdrv/devices/nvhost_ctrl.h
+++ b/src/core/hle/service/nvdrv/devices/nvhost_ctrl.h
@@ -25,12 +25,12 @@ public:
                          NvCore::Container& core);
     ~nvhost_ctrl() override;
 
-    NvResult Ioctl1(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+    NvResult Ioctl1(DeviceFD fd, Ioctl command, std::span<const u8> input,
                     std::vector<u8>& output) override;
-    NvResult Ioctl2(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
-                    const std::vector<u8>& inline_input, std::vector<u8>& output) override;
-    NvResult Ioctl3(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
-                    std::vector<u8>& output, std::vector<u8>& inline_output) override;
+    NvResult Ioctl2(DeviceFD fd, Ioctl command, std::span<const u8> input,
+                    std::span<const u8> inline_input, std::vector<u8>& output) override;
+    NvResult Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input, std::vector<u8>& output,
+                    std::vector<u8>& inline_output) override;
 
     void OnOpen(DeviceFD fd) override;
     void OnClose(DeviceFD fd) override;
@@ -186,13 +186,13 @@ private:
     static_assert(sizeof(IocCtrlEventUnregisterBatchParams) == 8,
                   "IocCtrlEventKill is incorrect size");
 
-    NvResult NvOsGetConfigU32(const std::vector<u8>& input, std::vector<u8>& output);
-    NvResult IocCtrlEventWait(const std::vector<u8>& input, std::vector<u8>& output,
+    NvResult NvOsGetConfigU32(std::span<const u8> input, std::vector<u8>& output);
+    NvResult IocCtrlEventWait(std::span<const u8> input, std::vector<u8>& output,
                               bool is_allocation);
-    NvResult IocCtrlEventRegister(const std::vector<u8>& input, std::vector<u8>& output);
-    NvResult IocCtrlEventUnregister(const std::vector<u8>& input, std::vector<u8>& output);
-    NvResult IocCtrlEventUnregisterBatch(const std::vector<u8>& input, std::vector<u8>& output);
-    NvResult IocCtrlClearEventWait(const std::vector<u8>& input, std::vector<u8>& output);
+    NvResult IocCtrlEventRegister(std::span<const u8> input, std::vector<u8>& output);
+    NvResult IocCtrlEventUnregister(std::span<const u8> input, std::vector<u8>& output);
+    NvResult IocCtrlEventUnregisterBatch(std::span<const u8> input, std::vector<u8>& output);
+    NvResult IocCtrlClearEventWait(std::span<const u8> input, std::vector<u8>& output);
 
     NvResult FreeEvent(u32 slot);
 
diff --git a/src/core/hle/service/nvdrv/devices/nvhost_ctrl_gpu.cpp b/src/core/hle/service/nvdrv/devices/nvhost_ctrl_gpu.cpp
index b97813fbc..be3c083db 100644
--- a/src/core/hle/service/nvdrv/devices/nvhost_ctrl_gpu.cpp
+++ b/src/core/hle/service/nvdrv/devices/nvhost_ctrl_gpu.cpp
@@ -21,7 +21,7 @@ nvhost_ctrl_gpu::~nvhost_ctrl_gpu() {
     events_interface.FreeEvent(unknown_event);
 }
 
-NvResult nvhost_ctrl_gpu::Ioctl1(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+NvResult nvhost_ctrl_gpu::Ioctl1(DeviceFD fd, Ioctl command, std::span<const u8> input,
                                  std::vector<u8>& output) {
     switch (command.group) {
     case 'G':
@@ -53,13 +53,13 @@ NvResult nvhost_ctrl_gpu::Ioctl1(DeviceFD fd, Ioctl command, const std::vector<u
     return NvResult::NotImplemented;
 }
 
-NvResult nvhost_ctrl_gpu::Ioctl2(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
-                                 const std::vector<u8>& inline_input, std::vector<u8>& output) {
+NvResult nvhost_ctrl_gpu::Ioctl2(DeviceFD fd, Ioctl command, std::span<const u8> input,
+                                 std::span<const u8> inline_input, std::vector<u8>& output) {
     UNIMPLEMENTED_MSG("Unimplemented ioctl={:08X}", command.raw);
     return NvResult::NotImplemented;
 }
 
-NvResult nvhost_ctrl_gpu::Ioctl3(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+NvResult nvhost_ctrl_gpu::Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input,
                                  std::vector<u8>& output, std::vector<u8>& inline_output) {
     switch (command.group) {
     case 'G':
@@ -82,8 +82,7 @@ NvResult nvhost_ctrl_gpu::Ioctl3(DeviceFD fd, Ioctl command, const std::vector<u
 void nvhost_ctrl_gpu::OnOpen(DeviceFD fd) {}
 void nvhost_ctrl_gpu::OnClose(DeviceFD fd) {}
 
-NvResult nvhost_ctrl_gpu::GetCharacteristics(const std::vector<u8>& input,
-                                             std::vector<u8>& output) {
+NvResult nvhost_ctrl_gpu::GetCharacteristics(std::span<const u8> input, std::vector<u8>& output) {
     LOG_DEBUG(Service_NVDRV, "called");
     IoctlCharacteristics params{};
     std::memcpy(&params, input.data(), input.size());
@@ -128,7 +127,7 @@ NvResult nvhost_ctrl_gpu::GetCharacteristics(const std::vector<u8>& input,
     return NvResult::Success;
 }
 
-NvResult nvhost_ctrl_gpu::GetCharacteristics(const std::vector<u8>& input, std::vector<u8>& output,
+NvResult nvhost_ctrl_gpu::GetCharacteristics(std::span<const u8> input, std::vector<u8>& output,
                                              std::vector<u8>& inline_output) {
     LOG_DEBUG(Service_NVDRV, "called");
     IoctlCharacteristics params{};
@@ -176,7 +175,7 @@ NvResult nvhost_ctrl_gpu::GetCharacteristics(const std::vector<u8>& input, std::
     return NvResult::Success;
 }
 
-NvResult nvhost_ctrl_gpu::GetTPCMasks(const std::vector<u8>& input, std::vector<u8>& output) {
+NvResult nvhost_ctrl_gpu::GetTPCMasks(std::span<const u8> input, std::vector<u8>& output) {
     IoctlGpuGetTpcMasksArgs params{};
     std::memcpy(&params, input.data(), input.size());
     LOG_DEBUG(Service_NVDRV, "called, mask_buffer_size=0x{:X}", params.mask_buffer_size);
@@ -187,7 +186,7 @@ NvResult nvhost_ctrl_gpu::GetTPCMasks(const std::vector<u8>& input, std::vector<
     return NvResult::Success;
 }
 
-NvResult nvhost_ctrl_gpu::GetTPCMasks(const std::vector<u8>& input, std::vector<u8>& output,
+NvResult nvhost_ctrl_gpu::GetTPCMasks(std::span<const u8> input, std::vector<u8>& output,
                                       std::vector<u8>& inline_output) {
     IoctlGpuGetTpcMasksArgs params{};
     std::memcpy(&params, input.data(), input.size());
@@ -200,7 +199,7 @@ NvResult nvhost_ctrl_gpu::GetTPCMasks(const std::vector<u8>& input, std::vector<
     return NvResult::Success;
 }
 
-NvResult nvhost_ctrl_gpu::GetActiveSlotMask(const std::vector<u8>& input, std::vector<u8>& output) {
+NvResult nvhost_ctrl_gpu::GetActiveSlotMask(std::span<const u8> input, std::vector<u8>& output) {
     LOG_DEBUG(Service_NVDRV, "called");
 
     IoctlActiveSlotMask params{};
@@ -213,7 +212,7 @@ NvResult nvhost_ctrl_gpu::GetActiveSlotMask(const std::vector<u8>& input, std::v
     return NvResult::Success;
 }
 
-NvResult nvhost_ctrl_gpu::ZCullGetCtxSize(const std::vector<u8>& input, std::vector<u8>& output) {
+NvResult nvhost_ctrl_gpu::ZCullGetCtxSize(std::span<const u8> input, std::vector<u8>& output) {
     LOG_DEBUG(Service_NVDRV, "called");
 
     IoctlZcullGetCtxSize params{};
@@ -225,7 +224,7 @@ NvResult nvhost_ctrl_gpu::ZCullGetCtxSize(const std::vector<u8>& input, std::vec
     return NvResult::Success;
 }
 
-NvResult nvhost_ctrl_gpu::ZCullGetInfo(const std::vector<u8>& input, std::vector<u8>& output) {
+NvResult nvhost_ctrl_gpu::ZCullGetInfo(std::span<const u8> input, std::vector<u8>& output) {
     LOG_DEBUG(Service_NVDRV, "called");
 
     IoctlNvgpuGpuZcullGetInfoArgs params{};
@@ -248,7 +247,7 @@ NvResult nvhost_ctrl_gpu::ZCullGetInfo(const std::vector<u8>& input, std::vector
     return NvResult::Success;
 }
 
-NvResult nvhost_ctrl_gpu::ZBCSetTable(const std::vector<u8>& input, std::vector<u8>& output) {
+NvResult nvhost_ctrl_gpu::ZBCSetTable(std::span<const u8> input, std::vector<u8>& output) {
     LOG_WARNING(Service_NVDRV, "(STUBBED) called");
 
     IoctlZbcSetTable params{};
@@ -264,7 +263,7 @@ NvResult nvhost_ctrl_gpu::ZBCSetTable(const std::vector<u8>& input, std::vector<
     return NvResult::Success;
 }
 
-NvResult nvhost_ctrl_gpu::ZBCQueryTable(const std::vector<u8>& input, std::vector<u8>& output) {
+NvResult nvhost_ctrl_gpu::ZBCQueryTable(std::span<const u8> input, std::vector<u8>& output) {
     LOG_WARNING(Service_NVDRV, "(STUBBED) called");
 
     IoctlZbcQueryTable params{};
@@ -274,7 +273,7 @@ NvResult nvhost_ctrl_gpu::ZBCQueryTable(const std::vector<u8>& input, std::vecto
     return NvResult::Success;
 }
 
-NvResult nvhost_ctrl_gpu::FlushL2(const std::vector<u8>& input, std::vector<u8>& output) {
+NvResult nvhost_ctrl_gpu::FlushL2(std::span<const u8> input, std::vector<u8>& output) {
     LOG_WARNING(Service_NVDRV, "(STUBBED) called");
 
     IoctlFlushL2 params{};
@@ -284,7 +283,7 @@ NvResult nvhost_ctrl_gpu::FlushL2(const std::vector<u8>& input, std::vector<u8>&
     return NvResult::Success;
 }
 
-NvResult nvhost_ctrl_gpu::GetGpuTime(const std::vector<u8>& input, std::vector<u8>& output) {
+NvResult nvhost_ctrl_gpu::GetGpuTime(std::span<const u8> input, std::vector<u8>& output) {
     LOG_DEBUG(Service_NVDRV, "called");
 
     IoctlGetGpuTime params{};
diff --git a/src/core/hle/service/nvdrv/devices/nvhost_ctrl_gpu.h b/src/core/hle/service/nvdrv/devices/nvhost_ctrl_gpu.h
index 1e8f254e2..b9333d9d3 100644
--- a/src/core/hle/service/nvdrv/devices/nvhost_ctrl_gpu.h
+++ b/src/core/hle/service/nvdrv/devices/nvhost_ctrl_gpu.h
@@ -21,12 +21,12 @@ public:
     explicit nvhost_ctrl_gpu(Core::System& system_, EventInterface& events_interface_);
     ~nvhost_ctrl_gpu() override;
 
-    NvResult Ioctl1(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+    NvResult Ioctl1(DeviceFD fd, Ioctl command, std::span<const u8> input,
                     std::vector<u8>& output) override;
-    NvResult Ioctl2(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
-                    const std::vector<u8>& inline_input, std::vector<u8>& output) override;
-    NvResult Ioctl3(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
-                    std::vector<u8>& output, std::vector<u8>& inline_output) override;
+    NvResult Ioctl2(DeviceFD fd, Ioctl command, std::span<const u8> input,
+                    std::span<const u8> inline_input, std::vector<u8>& output) override;
+    NvResult Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input, std::vector<u8>& output,
+                    std::vector<u8>& inline_output) override;
 
     void OnOpen(DeviceFD fd) override;
     void OnClose(DeviceFD fd) override;
@@ -151,21 +151,21 @@ private:
     };
     static_assert(sizeof(IoctlGetGpuTime) == 0x10, "IoctlGetGpuTime is incorrect size");
 
-    NvResult GetCharacteristics(const std::vector<u8>& input, std::vector<u8>& output);
-    NvResult GetCharacteristics(const std::vector<u8>& input, std::vector<u8>& output,
+    NvResult GetCharacteristics(std::span<const u8> input, std::vector<u8>& output);
+    NvResult GetCharacteristics(std::span<const u8> input, std::vector<u8>& output,
                                 std::vector<u8>& inline_output);
 
-    NvResult GetTPCMasks(const std::vector<u8>& input, std::vector<u8>& output);
-    NvResult GetTPCMasks(const std::vector<u8>& input, std::vector<u8>& output,
+    NvResult GetTPCMasks(std::span<const u8> input, std::vector<u8>& output);
+    NvResult GetTPCMasks(std::span<const u8> input, std::vector<u8>& output,
                          std::vector<u8>& inline_output);
 
-    NvResult GetActiveSlotMask(const std::vector<u8>& input, std::vector<u8>& output);
-    NvResult ZCullGetCtxSize(const std::vector<u8>& input, std::vector<u8>& output);
-    NvResult ZCullGetInfo(const std::vector<u8>& input, std::vector<u8>& output);
-    NvResult ZBCSetTable(const std::vector<u8>& input, std::vector<u8>& output);
-    NvResult ZBCQueryTable(const std::vector<u8>& input, std::vector<u8>& output);
-    NvResult FlushL2(const std::vector<u8>& input, std::vector<u8>& output);
-    NvResult GetGpuTime(const std::vector<u8>& input, std::vector<u8>& output);
+    NvResult GetActiveSlotMask(std::span<const u8> input, std::vector<u8>& output);
+    NvResult ZCullGetCtxSize(std::span<const u8> input, std::vector<u8>& output);
+    NvResult ZCullGetInfo(std::span<const u8> input, std::vector<u8>& output);
+    NvResult ZBCSetTable(std::span<const u8> input, std::vector<u8>& output);
+    NvResult ZBCQueryTable(std::span<const u8> input, std::vector<u8>& output);
+    NvResult FlushL2(std::span<const u8> input, std::vector<u8>& output);
+    NvResult GetGpuTime(std::span<const u8> input, std::vector<u8>& output);
 
     EventInterface& events_interface;
 
diff --git a/src/core/hle/service/nvdrv/devices/nvhost_gpu.cpp b/src/core/hle/service/nvdrv/devices/nvhost_gpu.cpp
index e123564c6..d2308fffc 100644
--- a/src/core/hle/service/nvdrv/devices/nvhost_gpu.cpp
+++ b/src/core/hle/service/nvdrv/devices/nvhost_gpu.cpp
@@ -46,7 +46,7 @@ nvhost_gpu::~nvhost_gpu() {
     syncpoint_manager.FreeSyncpoint(channel_syncpoint);
 }
 
-NvResult nvhost_gpu::Ioctl1(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+NvResult nvhost_gpu::Ioctl1(DeviceFD fd, Ioctl command, std::span<const u8> input,
                             std::vector<u8>& output) {
     switch (command.group) {
     case 0x0:
@@ -98,8 +98,8 @@ NvResult nvhost_gpu::Ioctl1(DeviceFD fd, Ioctl command, const std::vector<u8>& i
     return NvResult::NotImplemented;
 };
 
-NvResult nvhost_gpu::Ioctl2(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
-                            const std::vector<u8>& inline_input, std::vector<u8>& output) {
+NvResult nvhost_gpu::Ioctl2(DeviceFD fd, Ioctl command, std::span<const u8> input,
+                            std::span<const u8> inline_input, std::vector<u8>& output) {
     switch (command.group) {
     case 'H':
         switch (command.cmd) {
@@ -112,7 +112,7 @@ NvResult nvhost_gpu::Ioctl2(DeviceFD fd, Ioctl command, const std::vector<u8>& i
     return NvResult::NotImplemented;
 }
 
-NvResult nvhost_gpu::Ioctl3(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+NvResult nvhost_gpu::Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input,
                             std::vector<u8>& output, std::vector<u8>& inline_output) {
     UNIMPLEMENTED_MSG("Unimplemented ioctl={:08X}", command.raw);
     return NvResult::NotImplemented;
@@ -121,7 +121,7 @@ NvResult nvhost_gpu::Ioctl3(DeviceFD fd, Ioctl command, const std::vector<u8>& i
 void nvhost_gpu::OnOpen(DeviceFD fd) {}
 void nvhost_gpu::OnClose(DeviceFD fd) {}
 
-NvResult nvhost_gpu::SetNVMAPfd(const std::vector<u8>& input, std::vector<u8>& output) {
+NvResult nvhost_gpu::SetNVMAPfd(std::span<const u8> input, std::vector<u8>& output) {
     IoctlSetNvmapFD params{};
     std::memcpy(&params, input.data(), input.size());
     LOG_DEBUG(Service_NVDRV, "called, fd={}", params.nvmap_fd);
@@ -130,7 +130,7 @@ NvResult nvhost_gpu::SetNVMAPfd(const std::vector<u8>& input, std::vector<u8>& o
     return NvResult::Success;
 }
 
-NvResult nvhost_gpu::SetClientData(const std::vector<u8>& input, std::vector<u8>& output) {
+NvResult nvhost_gpu::SetClientData(std::span<const u8> input, std::vector<u8>& output) {
     LOG_DEBUG(Service_NVDRV, "called");
 
     IoctlClientData params{};
@@ -139,7 +139,7 @@ NvResult nvhost_gpu::SetClientData(const std::vector<u8>& input, std::vector<u8>
     return NvResult::Success;
 }
 
-NvResult nvhost_gpu::GetClientData(const std::vector<u8>& input, std::vector<u8>& output) {
+NvResult nvhost_gpu::GetClientData(std::span<const u8> input, std::vector<u8>& output) {
     LOG_DEBUG(Service_NVDRV, "called");
 
     IoctlClientData params{};
@@ -149,7 +149,7 @@ NvResult nvhost_gpu::GetClientData(const std::vector<u8>& input, std::vector<u8>
     return NvResult::Success;
 }
 
-NvResult nvhost_gpu::ZCullBind(const std::vector<u8>& input, std::vector<u8>& output) {
+NvResult nvhost_gpu::ZCullBind(std::span<const u8> input, std::vector<u8>& output) {
     std::memcpy(&zcull_params, input.data(), input.size());
     LOG_DEBUG(Service_NVDRV, "called, gpu_va={:X}, mode={:X}", zcull_params.gpu_va,
               zcull_params.mode);
@@ -158,7 +158,7 @@ NvResult nvhost_gpu::ZCullBind(const std::vector<u8>& input, std::vector<u8>& ou
     return NvResult::Success;
 }
 
-NvResult nvhost_gpu::SetErrorNotifier(const std::vector<u8>& input, std::vector<u8>& output) {
+NvResult nvhost_gpu::SetErrorNotifier(std::span<const u8> input, std::vector<u8>& output) {
     IoctlSetErrorNotifier params{};
     std::memcpy(&params, input.data(), input.size());
     LOG_WARNING(Service_NVDRV, "(STUBBED) called, offset={:X}, size={:X}, mem={:X}", params.offset,
@@ -168,14 +168,14 @@ NvResult nvhost_gpu::SetErrorNotifier(const std::vector<u8>& input, std::vector<
     return NvResult::Success;
 }
 
-NvResult nvhost_gpu::SetChannelPriority(const std::vector<u8>& input, std::vector<u8>& output) {
+NvResult nvhost_gpu::SetChannelPriority(std::span<const u8> input, std::vector<u8>& output) {
     std::memcpy(&channel_priority, input.data(), input.size());
     LOG_DEBUG(Service_NVDRV, "(STUBBED) called, priority={:X}", channel_priority);
 
     return NvResult::Success;
 }
 
-NvResult nvhost_gpu::AllocGPFIFOEx2(const std::vector<u8>& input, std::vector<u8>& output) {
+NvResult nvhost_gpu::AllocGPFIFOEx2(std::span<const u8> input, std::vector<u8>& output) {
     IoctlAllocGpfifoEx2 params{};
     std::memcpy(&params, input.data(), input.size());
     LOG_WARNING(Service_NVDRV,
@@ -197,7 +197,7 @@ NvResult nvhost_gpu::AllocGPFIFOEx2(const std::vector<u8>& input, std::vector<u8
     return NvResult::Success;
 }
 
-NvResult nvhost_gpu::AllocateObjectContext(const std::vector<u8>& input, std::vector<u8>& output) {
+NvResult nvhost_gpu::AllocateObjectContext(std::span<const u8> input, std::vector<u8>& output) {
     IoctlAllocObjCtx params{};
     std::memcpy(&params, input.data(), input.size());
     LOG_WARNING(Service_NVDRV, "(STUBBED) called, class_num={:X}, flags={:X}", params.class_num,
@@ -293,7 +293,7 @@ NvResult nvhost_gpu::SubmitGPFIFOImpl(IoctlSubmitGpfifo& params, std::vector<u8>
     return NvResult::Success;
 }
 
-NvResult nvhost_gpu::SubmitGPFIFOBase(const std::vector<u8>& input, std::vector<u8>& output,
+NvResult nvhost_gpu::SubmitGPFIFOBase(std::span<const u8> input, std::vector<u8>& output,
                                       bool kickoff) {
     if (input.size() < sizeof(IoctlSubmitGpfifo)) {
         UNIMPLEMENTED();
@@ -314,8 +314,7 @@ NvResult nvhost_gpu::SubmitGPFIFOBase(const std::vector<u8>& input, std::vector<
     return SubmitGPFIFOImpl(params, output, std::move(entries));
 }
 
-NvResult nvhost_gpu::SubmitGPFIFOBase(const std::vector<u8>& input,
-                                      const std::vector<u8>& input_inline,
+NvResult nvhost_gpu::SubmitGPFIFOBase(std::span<const u8> input, std::span<const u8> input_inline,
                                       std::vector<u8>& output) {
     if (input.size() < sizeof(IoctlSubmitGpfifo)) {
         UNIMPLEMENTED();
@@ -328,7 +327,7 @@ NvResult nvhost_gpu::SubmitGPFIFOBase(const std::vector<u8>& input,
     return SubmitGPFIFOImpl(params, output, std::move(entries));
 }
 
-NvResult nvhost_gpu::GetWaitbase(const std::vector<u8>& input, std::vector<u8>& output) {
+NvResult nvhost_gpu::GetWaitbase(std::span<const u8> input, std::vector<u8>& output) {
     IoctlGetWaitbase params{};
     std::memcpy(&params, input.data(), sizeof(IoctlGetWaitbase));
     LOG_INFO(Service_NVDRV, "called, unknown=0x{:X}", params.unknown);
@@ -338,7 +337,7 @@ NvResult nvhost_gpu::GetWaitbase(const std::vector<u8>& input, std::vector<u8>&
     return NvResult::Success;
 }
 
-NvResult nvhost_gpu::ChannelSetTimeout(const std::vector<u8>& input, std::vector<u8>& output) {
+NvResult nvhost_gpu::ChannelSetTimeout(std::span<const u8> input, std::vector<u8>& output) {
     IoctlChannelSetTimeout params{};
     std::memcpy(&params, input.data(), sizeof(IoctlChannelSetTimeout));
     LOG_INFO(Service_NVDRV, "called, timeout=0x{:X}", params.timeout);
@@ -346,7 +345,7 @@ NvResult nvhost_gpu::ChannelSetTimeout(const std::vector<u8>& input, std::vector
     return NvResult::Success;
 }
 
-NvResult nvhost_gpu::ChannelSetTimeslice(const std::vector<u8>& input, std::vector<u8>& output) {
+NvResult nvhost_gpu::ChannelSetTimeslice(std::span<const u8> input, std::vector<u8>& output) {
     IoctlSetTimeslice params{};
     std::memcpy(&params, input.data(), sizeof(IoctlSetTimeslice));
     LOG_INFO(Service_NVDRV, "called, timeslice=0x{:X}", params.timeslice);
diff --git a/src/core/hle/service/nvdrv/devices/nvhost_gpu.h b/src/core/hle/service/nvdrv/devices/nvhost_gpu.h
index 1e4ecd55b..3ca58202d 100644
--- a/src/core/hle/service/nvdrv/devices/nvhost_gpu.h
+++ b/src/core/hle/service/nvdrv/devices/nvhost_gpu.h
@@ -40,12 +40,12 @@ public:
                         NvCore::Container& core);
     ~nvhost_gpu() override;
 
-    NvResult Ioctl1(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+    NvResult Ioctl1(DeviceFD fd, Ioctl command, std::span<const u8> input,
                     std::vector<u8>& output) override;
-    NvResult Ioctl2(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
-                    const std::vector<u8>& inline_input, std::vector<u8>& output) override;
-    NvResult Ioctl3(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
-                    std::vector<u8>& output, std::vector<u8>& inline_output) override;
+    NvResult Ioctl2(DeviceFD fd, Ioctl command, std::span<const u8> input,
+                    std::span<const u8> inline_input, std::vector<u8>& output) override;
+    NvResult Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input, std::vector<u8>& output,
+                    std::vector<u8>& inline_output) override;
 
     void OnOpen(DeviceFD fd) override;
     void OnClose(DeviceFD fd) override;
@@ -186,23 +186,23 @@ private:
     u32_le channel_priority{};
     u32_le channel_timeslice{};
 
-    NvResult SetNVMAPfd(const std::vector<u8>& input, std::vector<u8>& output);
-    NvResult SetClientData(const std::vector<u8>& input, std::vector<u8>& output);
-    NvResult GetClientData(const std::vector<u8>& input, std::vector<u8>& output);
-    NvResult ZCullBind(const std::vector<u8>& input, std::vector<u8>& output);
-    NvResult SetErrorNotifier(const std::vector<u8>& input, std::vector<u8>& output);
-    NvResult SetChannelPriority(const std::vector<u8>& input, std::vector<u8>& output);
-    NvResult AllocGPFIFOEx2(const std::vector<u8>& input, std::vector<u8>& output);
-    NvResult AllocateObjectContext(const std::vector<u8>& input, std::vector<u8>& output);
+    NvResult SetNVMAPfd(std::span<const u8> input, std::vector<u8>& output);
+    NvResult SetClientData(std::span<const u8> input, std::vector<u8>& output);
+    NvResult GetClientData(std::span<const u8> input, std::vector<u8>& output);
+    NvResult ZCullBind(std::span<const u8> input, std::vector<u8>& output);
+    NvResult SetErrorNotifier(std::span<const u8> input, std::vector<u8>& output);
+    NvResult SetChannelPriority(std::span<const u8> input, std::vector<u8>& output);
+    NvResult AllocGPFIFOEx2(std::span<const u8> input, std::vector<u8>& output);
+    NvResult AllocateObjectContext(std::span<const u8> input, std::vector<u8>& output);
     NvResult SubmitGPFIFOImpl(IoctlSubmitGpfifo& params, std::vector<u8>& output,
                               Tegra::CommandList&& entries);
-    NvResult SubmitGPFIFOBase(const std::vector<u8>& input, std::vector<u8>& output,
+    NvResult SubmitGPFIFOBase(std::span<const u8> input, std::vector<u8>& output,
                               bool kickoff = false);
-    NvResult SubmitGPFIFOBase(const std::vector<u8>& input, const std::vector<u8>& input_inline,
+    NvResult SubmitGPFIFOBase(std::span<const u8> input, std::span<const u8> input_inline,
                               std::vector<u8>& output);
-    NvResult GetWaitbase(const std::vector<u8>& input, std::vector<u8>& output);
-    NvResult ChannelSetTimeout(const std::vector<u8>& input, std::vector<u8>& output);
-    NvResult ChannelSetTimeslice(const std::vector<u8>& input, std::vector<u8>& output);
+    NvResult GetWaitbase(std::span<const u8> input, std::vector<u8>& output);
+    NvResult ChannelSetTimeout(std::span<const u8> input, std::vector<u8>& output);
+    NvResult ChannelSetTimeslice(std::span<const u8> input, std::vector<u8>& output);
 
     EventInterface& events_interface;
     NvCore::Container& core;
diff --git a/src/core/hle/service/nvdrv/devices/nvhost_nvdec.cpp b/src/core/hle/service/nvdrv/devices/nvhost_nvdec.cpp
index 1703f9cc3..0c7aee1b8 100644
--- a/src/core/hle/service/nvdrv/devices/nvhost_nvdec.cpp
+++ b/src/core/hle/service/nvdrv/devices/nvhost_nvdec.cpp
@@ -15,7 +15,7 @@ nvhost_nvdec::nvhost_nvdec(Core::System& system_, NvCore::Container& core_)
     : nvhost_nvdec_common{system_, core_, NvCore::ChannelType::NvDec} {}
 nvhost_nvdec::~nvhost_nvdec() = default;
 
-NvResult nvhost_nvdec::Ioctl1(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+NvResult nvhost_nvdec::Ioctl1(DeviceFD fd, Ioctl command, std::span<const u8> input,
                               std::vector<u8>& output) {
     switch (command.group) {
     case 0x0:
@@ -55,13 +55,13 @@ NvResult nvhost_nvdec::Ioctl1(DeviceFD fd, Ioctl command, const std::vector<u8>&
     return NvResult::NotImplemented;
 }
 
-NvResult nvhost_nvdec::Ioctl2(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
-                              const std::vector<u8>& inline_input, std::vector<u8>& output) {
+NvResult nvhost_nvdec::Ioctl2(DeviceFD fd, Ioctl command, std::span<const u8> input,
+                              std::span<const u8> inline_input, std::vector<u8>& output) {
     UNIMPLEMENTED_MSG("Unimplemented ioctl={:08X}", command.raw);
     return NvResult::NotImplemented;
 }
 
-NvResult nvhost_nvdec::Ioctl3(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+NvResult nvhost_nvdec::Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input,
                               std::vector<u8>& output, std::vector<u8>& inline_output) {
     UNIMPLEMENTED_MSG("Unimplemented ioctl={:08X}", command.raw);
     return NvResult::NotImplemented;
diff --git a/src/core/hle/service/nvdrv/devices/nvhost_nvdec.h b/src/core/hle/service/nvdrv/devices/nvhost_nvdec.h
index c1b4e53e8..0d615bbcb 100644
--- a/src/core/hle/service/nvdrv/devices/nvhost_nvdec.h
+++ b/src/core/hle/service/nvdrv/devices/nvhost_nvdec.h
@@ -13,12 +13,12 @@ public:
     explicit nvhost_nvdec(Core::System& system_, NvCore::Container& core);
     ~nvhost_nvdec() override;
 
-    NvResult Ioctl1(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+    NvResult Ioctl1(DeviceFD fd, Ioctl command, std::span<const u8> input,
                     std::vector<u8>& output) override;
-    NvResult Ioctl2(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
-                    const std::vector<u8>& inline_input, std::vector<u8>& output) override;
-    NvResult Ioctl3(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
-                    std::vector<u8>& output, std::vector<u8>& inline_output) override;
+    NvResult Ioctl2(DeviceFD fd, Ioctl command, std::span<const u8> input,
+                    std::span<const u8> inline_input, std::vector<u8>& output) override;
+    NvResult Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input, std::vector<u8>& output,
+                    std::vector<u8>& inline_output) override;
 
     void OnOpen(DeviceFD fd) override;
     void OnClose(DeviceFD fd) override;
diff --git a/src/core/hle/service/nvdrv/devices/nvhost_nvdec_common.cpp b/src/core/hle/service/nvdrv/devices/nvhost_nvdec_common.cpp
index 99eede702..7bcef105b 100644
--- a/src/core/hle/service/nvdrv/devices/nvhost_nvdec_common.cpp
+++ b/src/core/hle/service/nvdrv/devices/nvhost_nvdec_common.cpp
@@ -23,7 +23,7 @@ namespace {
 // Copies count amount of type T from the input vector into the dst vector.
 // Returns the number of bytes written into dst.
 template <typename T>
-std::size_t SliceVectors(const std::vector<u8>& input, std::vector<T>& dst, std::size_t count,
+std::size_t SliceVectors(std::span<const u8> input, std::vector<T>& dst, std::size_t count,
                          std::size_t offset) {
     if (dst.empty()) {
         return 0;
@@ -63,7 +63,7 @@ nvhost_nvdec_common::~nvhost_nvdec_common() {
     core.Host1xDeviceFile().syncpts_accumulated.push_back(channel_syncpoint);
 }
 
-NvResult nvhost_nvdec_common::SetNVMAPfd(const std::vector<u8>& input) {
+NvResult nvhost_nvdec_common::SetNVMAPfd(std::span<const u8> input) {
     IoctlSetNvmapFD params{};
     std::memcpy(&params, input.data(), sizeof(IoctlSetNvmapFD));
     LOG_DEBUG(Service_NVDRV, "called, fd={}", params.nvmap_fd);
@@ -72,7 +72,7 @@ NvResult nvhost_nvdec_common::SetNVMAPfd(const std::vector<u8>& input) {
     return NvResult::Success;
 }
 
-NvResult nvhost_nvdec_common::Submit(DeviceFD fd, const std::vector<u8>& input,
+NvResult nvhost_nvdec_common::Submit(DeviceFD fd, std::span<const u8> input,
                                      std::vector<u8>& output) {
     IoctlSubmit params{};
     std::memcpy(&params, input.data(), sizeof(IoctlSubmit));
@@ -121,7 +121,7 @@ NvResult nvhost_nvdec_common::Submit(DeviceFD fd, const std::vector<u8>& input,
     return NvResult::Success;
 }
 
-NvResult nvhost_nvdec_common::GetSyncpoint(const std::vector<u8>& input, std::vector<u8>& output) {
+NvResult nvhost_nvdec_common::GetSyncpoint(std::span<const u8> input, std::vector<u8>& output) {
     IoctlGetSyncpoint params{};
     std::memcpy(&params, input.data(), sizeof(IoctlGetSyncpoint));
     LOG_DEBUG(Service_NVDRV, "called GetSyncpoint, id={}", params.param);
@@ -133,7 +133,7 @@ NvResult nvhost_nvdec_common::GetSyncpoint(const std::vector<u8>& input, std::ve
     return NvResult::Success;
 }
 
-NvResult nvhost_nvdec_common::GetWaitbase(const std::vector<u8>& input, std::vector<u8>& output) {
+NvResult nvhost_nvdec_common::GetWaitbase(std::span<const u8> input, std::vector<u8>& output) {
     IoctlGetWaitbase params{};
     LOG_CRITICAL(Service_NVDRV, "called WAITBASE");
     std::memcpy(&params, input.data(), sizeof(IoctlGetWaitbase));
@@ -142,7 +142,7 @@ NvResult nvhost_nvdec_common::GetWaitbase(const std::vector<u8>& input, std::vec
     return NvResult::Success;
 }
 
-NvResult nvhost_nvdec_common::MapBuffer(const std::vector<u8>& input, std::vector<u8>& output) {
+NvResult nvhost_nvdec_common::MapBuffer(std::span<const u8> input, std::vector<u8>& output) {
     IoctlMapBuffer params{};
     std::memcpy(&params, input.data(), sizeof(IoctlMapBuffer));
     std::vector<MapBufferEntry> cmd_buffer_handles(params.num_entries);
@@ -159,7 +159,7 @@ NvResult nvhost_nvdec_common::MapBuffer(const std::vector<u8>& input, std::vecto
     return NvResult::Success;
 }
 
-NvResult nvhost_nvdec_common::UnmapBuffer(const std::vector<u8>& input, std::vector<u8>& output) {
+NvResult nvhost_nvdec_common::UnmapBuffer(std::span<const u8> input, std::vector<u8>& output) {
     IoctlMapBuffer params{};
     std::memcpy(&params, input.data(), sizeof(IoctlMapBuffer));
     std::vector<MapBufferEntry> cmd_buffer_handles(params.num_entries);
@@ -173,8 +173,7 @@ NvResult nvhost_nvdec_common::UnmapBuffer(const std::vector<u8>& input, std::vec
     return NvResult::Success;
 }
 
-NvResult nvhost_nvdec_common::SetSubmitTimeout(const std::vector<u8>& input,
-                                               std::vector<u8>& output) {
+NvResult nvhost_nvdec_common::SetSubmitTimeout(std::span<const u8> input, std::vector<u8>& output) {
     std::memcpy(&submit_timeout, input.data(), input.size());
     LOG_WARNING(Service_NVDRV, "(STUBBED) called");
     return NvResult::Success;
diff --git a/src/core/hle/service/nvdrv/devices/nvhost_nvdec_common.h b/src/core/hle/service/nvdrv/devices/nvhost_nvdec_common.h
index fe76100c8..5af26a26f 100644
--- a/src/core/hle/service/nvdrv/devices/nvhost_nvdec_common.h
+++ b/src/core/hle/service/nvdrv/devices/nvhost_nvdec_common.h
@@ -107,13 +107,13 @@ protected:
     static_assert(sizeof(IoctlMapBuffer) == 0x0C, "IoctlMapBuffer is incorrect size");
 
     /// Ioctl command implementations
-    NvResult SetNVMAPfd(const std::vector<u8>& input);
-    NvResult Submit(DeviceFD fd, const std::vector<u8>& input, std::vector<u8>& output);
-    NvResult GetSyncpoint(const std::vector<u8>& input, std::vector<u8>& output);
-    NvResult GetWaitbase(const std::vector<u8>& input, std::vector<u8>& output);
-    NvResult MapBuffer(const std::vector<u8>& input, std::vector<u8>& output);
-    NvResult UnmapBuffer(const std::vector<u8>& input, std::vector<u8>& output);
-    NvResult SetSubmitTimeout(const std::vector<u8>& input, std::vector<u8>& output);
+    NvResult SetNVMAPfd(std::span<const u8> input);
+    NvResult Submit(DeviceFD fd, std::span<const u8> input, std::vector<u8>& output);
+    NvResult GetSyncpoint(std::span<const u8> input, std::vector<u8>& output);
+    NvResult GetWaitbase(std::span<const u8> input, std::vector<u8>& output);
+    NvResult MapBuffer(std::span<const u8> input, std::vector<u8>& output);
+    NvResult UnmapBuffer(std::span<const u8> input, std::vector<u8>& output);
+    NvResult SetSubmitTimeout(std::span<const u8> input, std::vector<u8>& output);
 
     Kernel::KEvent* QueryEvent(u32 event_id) override;
 
diff --git a/src/core/hle/service/nvdrv/devices/nvhost_nvjpg.cpp b/src/core/hle/service/nvdrv/devices/nvhost_nvjpg.cpp
index bdbc2f9e1..39f30e7c8 100644
--- a/src/core/hle/service/nvdrv/devices/nvhost_nvjpg.cpp
+++ b/src/core/hle/service/nvdrv/devices/nvhost_nvjpg.cpp
@@ -12,7 +12,7 @@ namespace Service::Nvidia::Devices {
 nvhost_nvjpg::nvhost_nvjpg(Core::System& system_) : nvdevice{system_} {}
 nvhost_nvjpg::~nvhost_nvjpg() = default;
 
-NvResult nvhost_nvjpg::Ioctl1(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+NvResult nvhost_nvjpg::Ioctl1(DeviceFD fd, Ioctl command, std::span<const u8> input,
                               std::vector<u8>& output) {
     switch (command.group) {
     case 'H':
@@ -31,13 +31,13 @@ NvResult nvhost_nvjpg::Ioctl1(DeviceFD fd, Ioctl command, const std::vector<u8>&
     return NvResult::NotImplemented;
 }
 
-NvResult nvhost_nvjpg::Ioctl2(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
-                              const std::vector<u8>& inline_input, std::vector<u8>& output) {
+NvResult nvhost_nvjpg::Ioctl2(DeviceFD fd, Ioctl command, std::span<const u8> input,
+                              std::span<const u8> inline_input, std::vector<u8>& output) {
     UNIMPLEMENTED_MSG("Unimplemented ioctl={:08X}", command.raw);
     return NvResult::NotImplemented;
 }
 
-NvResult nvhost_nvjpg::Ioctl3(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+NvResult nvhost_nvjpg::Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input,
                               std::vector<u8>& output, std::vector<u8>& inline_output) {
     UNIMPLEMENTED_MSG("Unimplemented ioctl={:08X}", command.raw);
     return NvResult::NotImplemented;
@@ -46,7 +46,7 @@ NvResult nvhost_nvjpg::Ioctl3(DeviceFD fd, Ioctl command, const std::vector<u8>&
 void nvhost_nvjpg::OnOpen(DeviceFD fd) {}
 void nvhost_nvjpg::OnClose(DeviceFD fd) {}
 
-NvResult nvhost_nvjpg::SetNVMAPfd(const std::vector<u8>& input, std::vector<u8>& output) {
+NvResult nvhost_nvjpg::SetNVMAPfd(std::span<const u8> input, std::vector<u8>& output) {
     IoctlSetNvmapFD params{};
     std::memcpy(&params, input.data(), input.size());
     LOG_DEBUG(Service_NVDRV, "called, fd={}", params.nvmap_fd);
diff --git a/src/core/hle/service/nvdrv/devices/nvhost_nvjpg.h b/src/core/hle/service/nvdrv/devices/nvhost_nvjpg.h
index 440e7d371..41b57e872 100644
--- a/src/core/hle/service/nvdrv/devices/nvhost_nvjpg.h
+++ b/src/core/hle/service/nvdrv/devices/nvhost_nvjpg.h
@@ -15,12 +15,12 @@ public:
     explicit nvhost_nvjpg(Core::System& system_);
     ~nvhost_nvjpg() override;
 
-    NvResult Ioctl1(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+    NvResult Ioctl1(DeviceFD fd, Ioctl command, std::span<const u8> input,
                     std::vector<u8>& output) override;
-    NvResult Ioctl2(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
-                    const std::vector<u8>& inline_input, std::vector<u8>& output) override;
-    NvResult Ioctl3(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
-                    std::vector<u8>& output, std::vector<u8>& inline_output) override;
+    NvResult Ioctl2(DeviceFD fd, Ioctl command, std::span<const u8> input,
+                    std::span<const u8> inline_input, std::vector<u8>& output) override;
+    NvResult Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input, std::vector<u8>& output,
+                    std::vector<u8>& inline_output) override;
 
     void OnOpen(DeviceFD fd) override;
     void OnClose(DeviceFD fd) override;
@@ -33,7 +33,7 @@ private:
 
     s32_le nvmap_fd{};
 
-    NvResult SetNVMAPfd(const std::vector<u8>& input, std::vector<u8>& output);
+    NvResult SetNVMAPfd(std::span<const u8> input, std::vector<u8>& output);
 };
 
 } // namespace Service::Nvidia::Devices
diff --git a/src/core/hle/service/nvdrv/devices/nvhost_vic.cpp b/src/core/hle/service/nvdrv/devices/nvhost_vic.cpp
index 73f97136e..b0ea402a7 100644
--- a/src/core/hle/service/nvdrv/devices/nvhost_vic.cpp
+++ b/src/core/hle/service/nvdrv/devices/nvhost_vic.cpp
@@ -15,7 +15,7 @@ nvhost_vic::nvhost_vic(Core::System& system_, NvCore::Container& core_)
 
 nvhost_vic::~nvhost_vic() = default;
 
-NvResult nvhost_vic::Ioctl1(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+NvResult nvhost_vic::Ioctl1(DeviceFD fd, Ioctl command, std::span<const u8> input,
                             std::vector<u8>& output) {
     switch (command.group) {
     case 0x0:
@@ -55,13 +55,13 @@ NvResult nvhost_vic::Ioctl1(DeviceFD fd, Ioctl command, const std::vector<u8>& i
     return NvResult::NotImplemented;
 }
 
-NvResult nvhost_vic::Ioctl2(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
-                            const std::vector<u8>& inline_input, std::vector<u8>& output) {
+NvResult nvhost_vic::Ioctl2(DeviceFD fd, Ioctl command, std::span<const u8> input,
+                            std::span<const u8> inline_input, std::vector<u8>& output) {
     UNIMPLEMENTED_MSG("Unimplemented ioctl={:08X}", command.raw);
     return NvResult::NotImplemented;
 }
 
-NvResult nvhost_vic::Ioctl3(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+NvResult nvhost_vic::Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input,
                             std::vector<u8>& output, std::vector<u8>& inline_output) {
     UNIMPLEMENTED_MSG("Unimplemented ioctl={:08X}", command.raw);
     return NvResult::NotImplemented;
diff --git a/src/core/hle/service/nvdrv/devices/nvhost_vic.h b/src/core/hle/service/nvdrv/devices/nvhost_vic.h
index f164caafb..b5e350a83 100644
--- a/src/core/hle/service/nvdrv/devices/nvhost_vic.h
+++ b/src/core/hle/service/nvdrv/devices/nvhost_vic.h
@@ -12,12 +12,12 @@ public:
     explicit nvhost_vic(Core::System& system_, NvCore::Container& core);
     ~nvhost_vic();
 
-    NvResult Ioctl1(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+    NvResult Ioctl1(DeviceFD fd, Ioctl command, std::span<const u8> input,
                     std::vector<u8>& output) override;
-    NvResult Ioctl2(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
-                    const std::vector<u8>& inline_input, std::vector<u8>& output) override;
-    NvResult Ioctl3(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
-                    std::vector<u8>& output, std::vector<u8>& inline_output) override;
+    NvResult Ioctl2(DeviceFD fd, Ioctl command, std::span<const u8> input,
+                    std::span<const u8> inline_input, std::vector<u8>& output) override;
+    NvResult Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input, std::vector<u8>& output,
+                    std::vector<u8>& inline_output) override;
 
     void OnOpen(DeviceFD fd) override;
     void OnClose(DeviceFD fd) override;
diff --git a/src/core/hle/service/nvdrv/devices/nvmap.cpp b/src/core/hle/service/nvdrv/devices/nvmap.cpp
index fa29db758..29c1e0f01 100644
--- a/src/core/hle/service/nvdrv/devices/nvmap.cpp
+++ b/src/core/hle/service/nvdrv/devices/nvmap.cpp
@@ -25,7 +25,7 @@ nvmap::nvmap(Core::System& system_, NvCore::Container& container_)
 
 nvmap::~nvmap() = default;
 
-NvResult nvmap::Ioctl1(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+NvResult nvmap::Ioctl1(DeviceFD fd, Ioctl command, std::span<const u8> input,
                        std::vector<u8>& output) {
     switch (command.group) {
     case 0x1:
@@ -54,13 +54,13 @@ NvResult nvmap::Ioctl1(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
     return NvResult::NotImplemented;
 }
 
-NvResult nvmap::Ioctl2(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
-                       const std::vector<u8>& inline_input, std::vector<u8>& output) {
+NvResult nvmap::Ioctl2(DeviceFD fd, Ioctl command, std::span<const u8> input,
+                       std::span<const u8> inline_input, std::vector<u8>& output) {
     UNIMPLEMENTED_MSG("Unimplemented ioctl={:08X}", command.raw);
     return NvResult::NotImplemented;
 }
 
-NvResult nvmap::Ioctl3(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+NvResult nvmap::Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input,
                        std::vector<u8>& output, std::vector<u8>& inline_output) {
     UNIMPLEMENTED_MSG("Unimplemented ioctl={:08X}", command.raw);
     return NvResult::NotImplemented;
@@ -69,7 +69,7 @@ NvResult nvmap::Ioctl3(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
 void nvmap::OnOpen(DeviceFD fd) {}
 void nvmap::OnClose(DeviceFD fd) {}
 
-NvResult nvmap::IocCreate(const std::vector<u8>& input, std::vector<u8>& output) {
+NvResult nvmap::IocCreate(std::span<const u8> input, std::vector<u8>& output) {
     IocCreateParams params;
     std::memcpy(&params, input.data(), sizeof(params));
     LOG_DEBUG(Service_NVDRV, "called, size=0x{:08X}", params.size);
@@ -89,7 +89,7 @@ NvResult nvmap::IocCreate(const std::vector<u8>& input, std::vector<u8>& output)
     return NvResult::Success;
 }
 
-NvResult nvmap::IocAlloc(const std::vector<u8>& input, std::vector<u8>& output) {
+NvResult nvmap::IocAlloc(std::span<const u8> input, std::vector<u8>& output) {
     IocAllocParams params;
     std::memcpy(&params, input.data(), sizeof(params));
     LOG_DEBUG(Service_NVDRV, "called, addr={:X}", params.address);
@@ -137,7 +137,7 @@ NvResult nvmap::IocAlloc(const std::vector<u8>& input, std::vector<u8>& output)
     return result;
 }
 
-NvResult nvmap::IocGetId(const std::vector<u8>& input, std::vector<u8>& output) {
+NvResult nvmap::IocGetId(std::span<const u8> input, std::vector<u8>& output) {
     IocGetIdParams params;
     std::memcpy(&params, input.data(), sizeof(params));
 
@@ -161,7 +161,7 @@ NvResult nvmap::IocGetId(const std::vector<u8>& input, std::vector<u8>& output)
     return NvResult::Success;
 }
 
-NvResult nvmap::IocFromId(const std::vector<u8>& input, std::vector<u8>& output) {
+NvResult nvmap::IocFromId(std::span<const u8> input, std::vector<u8>& output) {
     IocFromIdParams params;
     std::memcpy(&params, input.data(), sizeof(params));
 
@@ -192,7 +192,7 @@ NvResult nvmap::IocFromId(const std::vector<u8>& input, std::vector<u8>& output)
     return NvResult::Success;
 }
 
-NvResult nvmap::IocParam(const std::vector<u8>& input, std::vector<u8>& output) {
+NvResult nvmap::IocParam(std::span<const u8> input, std::vector<u8>& output) {
     enum class ParamTypes { Size = 1, Alignment = 2, Base = 3, Heap = 4, Kind = 5, Compr = 6 };
 
     IocParamParams params;
@@ -241,7 +241,7 @@ NvResult nvmap::IocParam(const std::vector<u8>& input, std::vector<u8>& output)
     return NvResult::Success;
 }
 
-NvResult nvmap::IocFree(const std::vector<u8>& input, std::vector<u8>& output) {
+NvResult nvmap::IocFree(std::span<const u8> input, std::vector<u8>& output) {
     IocFreeParams params;
     std::memcpy(&params, input.data(), sizeof(params));
 
diff --git a/src/core/hle/service/nvdrv/devices/nvmap.h b/src/core/hle/service/nvdrv/devices/nvmap.h
index e9bfd0358..82bd3b118 100644
--- a/src/core/hle/service/nvdrv/devices/nvmap.h
+++ b/src/core/hle/service/nvdrv/devices/nvmap.h
@@ -26,12 +26,12 @@ public:
     nvmap(const nvmap&) = delete;
     nvmap& operator=(const nvmap&) = delete;
 
-    NvResult Ioctl1(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+    NvResult Ioctl1(DeviceFD fd, Ioctl command, std::span<const u8> input,
                     std::vector<u8>& output) override;
-    NvResult Ioctl2(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
-                    const std::vector<u8>& inline_input, std::vector<u8>& output) override;
-    NvResult Ioctl3(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
-                    std::vector<u8>& output, std::vector<u8>& inline_output) override;
+    NvResult Ioctl2(DeviceFD fd, Ioctl command, std::span<const u8> input,
+                    std::span<const u8> inline_input, std::vector<u8>& output) override;
+    NvResult Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input, std::vector<u8>& output,
+                    std::vector<u8>& inline_output) override;
 
     void OnOpen(DeviceFD fd) override;
     void OnClose(DeviceFD fd) override;
@@ -106,12 +106,12 @@ private:
     };
     static_assert(sizeof(IocGetIdParams) == 8, "IocGetIdParams has wrong size");
 
-    NvResult IocCreate(const std::vector<u8>& input, std::vector<u8>& output);
-    NvResult IocAlloc(const std::vector<u8>& input, std::vector<u8>& output);
-    NvResult IocGetId(const std::vector<u8>& input, std::vector<u8>& output);
-    NvResult IocFromId(const std::vector<u8>& input, std::vector<u8>& output);
-    NvResult IocParam(const std::vector<u8>& input, std::vector<u8>& output);
-    NvResult IocFree(const std::vector<u8>& input, std::vector<u8>& output);
+    NvResult IocCreate(std::span<const u8> input, std::vector<u8>& output);
+    NvResult IocAlloc(std::span<const u8> input, std::vector<u8>& output);
+    NvResult IocGetId(std::span<const u8> input, std::vector<u8>& output);
+    NvResult IocFromId(std::span<const u8> input, std::vector<u8>& output);
+    NvResult IocParam(std::span<const u8> input, std::vector<u8>& output);
+    NvResult IocFree(std::span<const u8> input, std::vector<u8>& output);
 
     NvCore::Container& container;
     NvCore::NvMap& file;
diff --git a/src/core/hle/service/nvdrv/nvdrv.cpp b/src/core/hle/service/nvdrv/nvdrv.cpp
index 6fc8565c0..52d27e755 100644
--- a/src/core/hle/service/nvdrv/nvdrv.cpp
+++ b/src/core/hle/service/nvdrv/nvdrv.cpp
@@ -124,7 +124,7 @@ DeviceFD Module::Open(const std::string& device_name) {
     return fd;
 }
 
-NvResult Module::Ioctl1(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+NvResult Module::Ioctl1(DeviceFD fd, Ioctl command, std::span<const u8> input,
                         std::vector<u8>& output) {
     if (fd < 0) {
         LOG_ERROR(Service_NVDRV, "Invalid DeviceFD={}!", fd);
@@ -141,8 +141,8 @@ NvResult Module::Ioctl1(DeviceFD fd, Ioctl command, const std::vector<u8>& input
     return itr->second->Ioctl1(fd, command, input, output);
 }
 
-NvResult Module::Ioctl2(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
-                        const std::vector<u8>& inline_input, std::vector<u8>& output) {
+NvResult Module::Ioctl2(DeviceFD fd, Ioctl command, std::span<const u8> input,
+                        std::span<const u8> inline_input, std::vector<u8>& output) {
     if (fd < 0) {
         LOG_ERROR(Service_NVDRV, "Invalid DeviceFD={}!", fd);
         return NvResult::InvalidState;
@@ -158,7 +158,7 @@ NvResult Module::Ioctl2(DeviceFD fd, Ioctl command, const std::vector<u8>& input
     return itr->second->Ioctl2(fd, command, input, inline_input, output);
 }
 
-NvResult Module::Ioctl3(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
+NvResult Module::Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input,
                         std::vector<u8>& output, std::vector<u8>& inline_output) {
     if (fd < 0) {
         LOG_ERROR(Service_NVDRV, "Invalid DeviceFD={}!", fd);
diff --git a/src/core/hle/service/nvdrv/nvdrv.h b/src/core/hle/service/nvdrv/nvdrv.h
index f3c81bd88..b09b6e585 100644
--- a/src/core/hle/service/nvdrv/nvdrv.h
+++ b/src/core/hle/service/nvdrv/nvdrv.h
@@ -7,6 +7,7 @@
 #include <functional>
 #include <list>
 #include <memory>
+#include <span>
 #include <string>
 #include <unordered_map>
 #include <vector>
@@ -79,14 +80,13 @@ public:
     DeviceFD Open(const std::string& device_name);
 
     /// Sends an ioctl command to the specified file descriptor.
-    NvResult Ioctl1(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
-                    std::vector<u8>& output);
+    NvResult Ioctl1(DeviceFD fd, Ioctl command, std::span<const u8> input, std::vector<u8>& output);
 
-    NvResult Ioctl2(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
-                    const std::vector<u8>& inline_input, std::vector<u8>& output);
+    NvResult Ioctl2(DeviceFD fd, Ioctl command, std::span<const u8> input,
+                    std::span<const u8> inline_input, std::vector<u8>& output);
 
-    NvResult Ioctl3(DeviceFD fd, Ioctl command, const std::vector<u8>& input,
-                    std::vector<u8>& output, std::vector<u8>& inline_output);
+    NvResult Ioctl3(DeviceFD fd, Ioctl command, std::span<const u8> input, std::vector<u8>& output,
+                    std::vector<u8>& inline_output);
 
     /// Closes a device file descriptor and returns operation success.
     NvResult Close(DeviceFD fd);
diff --git a/src/core/hle/service/nvflinger/buffer_queue_producer.cpp b/src/core/hle/service/nvflinger/buffer_queue_producer.cpp
index e601b5da1..bcbe05b0d 100644
--- a/src/core/hle/service/nvflinger/buffer_queue_producer.cpp
+++ b/src/core/hle/service/nvflinger/buffer_queue_producer.cpp
@@ -815,8 +815,8 @@ Status BufferQueueProducer::SetPreallocatedBuffer(s32 slot,
 
 void BufferQueueProducer::Transact(Kernel::HLERequestContext& ctx, TransactionId code, u32 flags) {
     Status status{Status::NoError};
-    Parcel parcel_in{ctx.ReadBuffer()};
-    Parcel parcel_out{};
+    InputParcel parcel_in{ctx.ReadBuffer()};
+    OutputParcel parcel_out{};
 
     switch (code) {
     case TransactionId::Connect: {
diff --git a/src/core/hle/service/nvflinger/graphic_buffer_producer.cpp b/src/core/hle/service/nvflinger/graphic_buffer_producer.cpp
index 4043c91f1..769e8c0a3 100644
--- a/src/core/hle/service/nvflinger/graphic_buffer_producer.cpp
+++ b/src/core/hle/service/nvflinger/graphic_buffer_producer.cpp
@@ -9,7 +9,7 @@
 
 namespace Service::android {
 
-QueueBufferInput::QueueBufferInput(Parcel& parcel) {
+QueueBufferInput::QueueBufferInput(InputParcel& parcel) {
     parcel.ReadFlattened(*this);
 }
 
diff --git a/src/core/hle/service/nvflinger/graphic_buffer_producer.h b/src/core/hle/service/nvflinger/graphic_buffer_producer.h
index 6ea327bbe..2969f0fd5 100644
--- a/src/core/hle/service/nvflinger/graphic_buffer_producer.h
+++ b/src/core/hle/service/nvflinger/graphic_buffer_producer.h
@@ -14,11 +14,11 @@
 
 namespace Service::android {
 
-class Parcel;
+class InputParcel;
 
 #pragma pack(push, 1)
 struct QueueBufferInput final {
-    explicit QueueBufferInput(Parcel& parcel);
+    explicit QueueBufferInput(InputParcel& parcel);
 
     void Deflate(s64* timestamp_, bool* is_auto_timestamp_, Common::Rectangle<s32>* crop_,
                  NativeWindowScalingMode* scaling_mode_, NativeWindowTransform* transform_,
diff --git a/src/core/hle/service/nvflinger/parcel.h b/src/core/hle/service/nvflinger/parcel.h
index f3fa2587d..d1b6201e0 100644
--- a/src/core/hle/service/nvflinger/parcel.h
+++ b/src/core/hle/service/nvflinger/parcel.h
@@ -4,6 +4,7 @@
 #pragma once
 
 #include <memory>
+#include <span>
 #include <vector>
 
 #include "common/alignment.h"
@@ -12,18 +13,17 @@
 
 namespace Service::android {
 
-class Parcel final {
-public:
-    static constexpr std::size_t DefaultBufferSize = 0x40;
-
-    Parcel() : buffer(DefaultBufferSize) {}
-
-    template <typename T>
-    explicit Parcel(const T& out_data) : buffer(DefaultBufferSize) {
-        Write(out_data);
-    }
+struct ParcelHeader {
+    u32 data_size;
+    u32 data_offset;
+    u32 objects_size;
+    u32 objects_offset;
+};
+static_assert(sizeof(ParcelHeader) == 16, "ParcelHeader has wrong size");
 
-    explicit Parcel(std::vector<u8> in_data) : buffer(std::move(in_data)) {
+class InputParcel final {
+public:
+    explicit InputParcel(std::span<const u8> in_data) : read_buffer(std::move(in_data)) {
         DeserializeHeader();
         [[maybe_unused]] const std::u16string token = ReadInterfaceToken();
     }
@@ -31,9 +31,9 @@ public:
     template <typename T>
     void Read(T& val) {
         static_assert(std::is_trivially_copyable_v<T>, "T must be trivially copyable.");
-        ASSERT(read_index + sizeof(T) <= buffer.size());
+        ASSERT(read_index + sizeof(T) <= read_buffer.size());
 
-        std::memcpy(&val, buffer.data() + read_index, sizeof(T));
+        std::memcpy(&val, read_buffer.data() + read_index, sizeof(T));
         read_index += sizeof(T);
         read_index = Common::AlignUp(read_index, 4);
     }
@@ -62,10 +62,10 @@ public:
     template <typename T>
     T ReadUnaligned() {
         static_assert(std::is_trivially_copyable_v<T>, "T must be trivially copyable.");
-        ASSERT(read_index + sizeof(T) <= buffer.size());
+        ASSERT(read_index + sizeof(T) <= read_buffer.size());
 
         T val;
-        std::memcpy(&val, buffer.data() + read_index, sizeof(T));
+        std::memcpy(&val, read_buffer.data() + read_index, sizeof(T));
         read_index += sizeof(T);
         return val;
     }
@@ -101,6 +101,31 @@ public:
         return token;
     }
 
+    void DeserializeHeader() {
+        ASSERT(read_buffer.size() > sizeof(ParcelHeader));
+
+        ParcelHeader header{};
+        std::memcpy(&header, read_buffer.data(), sizeof(ParcelHeader));
+
+        read_index = header.data_offset;
+    }
+
+private:
+    std::span<const u8> read_buffer;
+    std::size_t read_index = 0;
+};
+
+class OutputParcel final {
+public:
+    static constexpr std::size_t DefaultBufferSize = 0x40;
+
+    OutputParcel() : buffer(DefaultBufferSize) {}
+
+    template <typename T>
+    explicit OutputParcel(const T& out_data) : buffer(DefaultBufferSize) {
+        Write(out_data);
+    }
+
     template <typename T>
     void Write(const T& val) {
         static_assert(std::is_trivially_copyable_v<T>, "T must be trivially copyable.");
@@ -133,40 +158,20 @@ public:
         WriteObject(ptr.get());
     }
 
-    void DeserializeHeader() {
-        ASSERT(buffer.size() > sizeof(Header));
-
-        Header header{};
-        std::memcpy(&header, buffer.data(), sizeof(Header));
-
-        read_index = header.data_offset;
-    }
-
     std::vector<u8> Serialize() const {
-        ASSERT(read_index == 0);
-
-        Header header{};
-        header.data_size = static_cast<u32>(write_index - sizeof(Header));
-        header.data_offset = sizeof(Header);
+        ParcelHeader header{};
+        header.data_size = static_cast<u32>(write_index - sizeof(ParcelHeader));
+        header.data_offset = sizeof(ParcelHeader);
         header.objects_size = 4;
-        header.objects_offset = static_cast<u32>(sizeof(Header) + header.data_size);
-        std::memcpy(buffer.data(), &header, sizeof(Header));
+        header.objects_offset = static_cast<u32>(sizeof(ParcelHeader) + header.data_size);
+        std::memcpy(buffer.data(), &header, sizeof(ParcelHeader));
 
         return buffer;
     }
 
 private:
-    struct Header {
-        u32 data_size;
-        u32 data_offset;
-        u32 objects_size;
-        u32 objects_offset;
-    };
-    static_assert(sizeof(Header) == 16, "ParcelHeader has wrong size");
-
     mutable std::vector<u8> buffer;
-    std::size_t read_index = 0;
-    std::size_t write_index = sizeof(Header);
+    std::size_t write_index = sizeof(ParcelHeader);
 };
 
 } // namespace Service::android
diff --git a/src/core/hle/service/prepo/prepo.cpp b/src/core/hle/service/prepo/prepo.cpp
index 78f897d3e..01040b32a 100644
--- a/src/core/hle/service/prepo/prepo.cpp
+++ b/src/core/hle/service/prepo/prepo.cpp
@@ -63,7 +63,7 @@ private:
                 return ctx.ReadBuffer(1);
             }
 
-            return std::vector<u8>{};
+            return std::span<const u8>{};
         }();
 
         LOG_DEBUG(Service_PREPO,
@@ -90,7 +90,7 @@ private:
                 return ctx.ReadBuffer(1);
             }
 
-            return std::vector<u8>{};
+            return std::span<const u8>{};
         }();
 
         LOG_DEBUG(Service_PREPO,
@@ -142,7 +142,7 @@ private:
                 return ctx.ReadBuffer(1);
             }
 
-            return std::vector<u8>{};
+            return std::span<const u8>{};
         }();
 
         LOG_DEBUG(Service_PREPO, "called, title_id={:016X}, data1_size={:016X}, data2_size={:016X}",
@@ -166,7 +166,7 @@ private:
                 return ctx.ReadBuffer(1);
             }
 
-            return std::vector<u8>{};
+            return std::span<const u8>{};
         }();
 
         LOG_DEBUG(Service_PREPO,
diff --git a/src/core/hle/service/sockets/bsd.cpp b/src/core/hle/service/sockets/bsd.cpp
index 9e94a462f..bdb499268 100644
--- a/src/core/hle/service/sockets/bsd.cpp
+++ b/src/core/hle/service/sockets/bsd.cpp
@@ -208,7 +208,6 @@ void BSD::Bind(Kernel::HLERequestContext& ctx) {
     const s32 fd = rp.Pop<s32>();
 
     LOG_DEBUG(Service, "called. fd={} addrlen={}", fd, ctx.GetReadBufferSize());
-
     BuildErrnoResponse(ctx, BindImpl(fd, ctx.ReadBuffer()));
 }
 
@@ -312,7 +311,7 @@ void BSD::SetSockOpt(Kernel::HLERequestContext& ctx) {
     const u32 level = rp.Pop<u32>();
     const OptName optname = static_cast<OptName>(rp.Pop<u32>());
 
-    const std::vector<u8> buffer = ctx.ReadBuffer();
+    const auto buffer = ctx.ReadBuffer();
     const u8* optval = buffer.empty() ? nullptr : buffer.data();
     size_t optlen = buffer.size();
 
@@ -489,7 +488,7 @@ std::pair<s32, Errno> BSD::SocketImpl(Domain domain, Type type, Protocol protoco
     return {fd, Errno::SUCCESS};
 }
 
-std::pair<s32, Errno> BSD::PollImpl(std::vector<u8>& write_buffer, std::vector<u8> read_buffer,
+std::pair<s32, Errno> BSD::PollImpl(std::vector<u8>& write_buffer, std::span<const u8> read_buffer,
                                     s32 nfds, s32 timeout) {
     if (write_buffer.size() < nfds * sizeof(PollFD)) {
         return {-1, Errno::INVAL};
@@ -584,7 +583,7 @@ std::pair<s32, Errno> BSD::AcceptImpl(s32 fd, std::vector<u8>& write_buffer) {
     return {new_fd, Errno::SUCCESS};
 }
 
-Errno BSD::BindImpl(s32 fd, const std::vector<u8>& addr) {
+Errno BSD::BindImpl(s32 fd, std::span<const u8> addr) {
     if (!IsFileDescriptorValid(fd)) {
         return Errno::BADF;
     }
@@ -595,7 +594,7 @@ Errno BSD::BindImpl(s32 fd, const std::vector<u8>& addr) {
     return Translate(file_descriptors[fd]->socket->Bind(Translate(addr_in)));
 }
 
-Errno BSD::ConnectImpl(s32 fd, const std::vector<u8>& addr) {
+Errno BSD::ConnectImpl(s32 fd, std::span<const u8> addr) {
     if (!IsFileDescriptorValid(fd)) {
         return Errno::BADF;
     }
@@ -800,15 +799,15 @@ std::pair<s32, Errno> BSD::RecvFromImpl(s32 fd, u32 flags, std::vector<u8>& mess
     return {ret, bsd_errno};
 }
 
-std::pair<s32, Errno> BSD::SendImpl(s32 fd, u32 flags, const std::vector<u8>& message) {
+std::pair<s32, Errno> BSD::SendImpl(s32 fd, u32 flags, std::span<const u8> message) {
     if (!IsFileDescriptorValid(fd)) {
         return {-1, Errno::BADF};
     }
     return Translate(file_descriptors[fd]->socket->Send(message, flags));
 }
 
-std::pair<s32, Errno> BSD::SendToImpl(s32 fd, u32 flags, const std::vector<u8>& message,
-                                      const std::vector<u8>& addr) {
+std::pair<s32, Errno> BSD::SendToImpl(s32 fd, u32 flags, std::span<const u8> message,
+                                      std::span<const u8> addr) {
     if (!IsFileDescriptorValid(fd)) {
         return {-1, Errno::BADF};
     }
diff --git a/src/core/hle/service/sockets/bsd.h b/src/core/hle/service/sockets/bsd.h
index 81e855e0f..56bb3f8b1 100644
--- a/src/core/hle/service/sockets/bsd.h
+++ b/src/core/hle/service/sockets/bsd.h
@@ -4,6 +4,7 @@
 #pragma once
 
 #include <memory>
+#include <span>
 #include <vector>
 
 #include "common/common_types.h"
@@ -44,7 +45,7 @@ private:
 
         s32 nfds;
         s32 timeout;
-        std::vector<u8> read_buffer;
+        std::span<const u8> read_buffer;
         std::vector<u8> write_buffer;
         s32 ret{};
         Errno bsd_errno{};
@@ -65,7 +66,7 @@ private:
         void Response(Kernel::HLERequestContext& ctx);
 
         s32 fd;
-        std::vector<u8> addr;
+        std::span<const u8> addr;
         Errno bsd_errno{};
     };
 
@@ -98,7 +99,7 @@ private:
 
         s32 fd;
         u32 flags;
-        std::vector<u8> message;
+        std::span<const u8> message;
         s32 ret{};
         Errno bsd_errno{};
     };
@@ -109,8 +110,8 @@ private:
 
         s32 fd;
         u32 flags;
-        std::vector<u8> message;
-        std::vector<u8> addr;
+        std::span<const u8> message;
+        std::span<const u8> addr;
         s32 ret{};
         Errno bsd_errno{};
     };
@@ -143,11 +144,11 @@ private:
     void ExecuteWork(Kernel::HLERequestContext& ctx, Work work);
 
     std::pair<s32, Errno> SocketImpl(Domain domain, Type type, Protocol protocol);
-    std::pair<s32, Errno> PollImpl(std::vector<u8>& write_buffer, std::vector<u8> read_buffer,
+    std::pair<s32, Errno> PollImpl(std::vector<u8>& write_buffer, std::span<const u8> read_buffer,
                                    s32 nfds, s32 timeout);
     std::pair<s32, Errno> AcceptImpl(s32 fd, std::vector<u8>& write_buffer);
-    Errno BindImpl(s32 fd, const std::vector<u8>& addr);
-    Errno ConnectImpl(s32 fd, const std::vector<u8>& addr);
+    Errno BindImpl(s32 fd, std::span<const u8> addr);
+    Errno ConnectImpl(s32 fd, std::span<const u8> addr);
     Errno GetPeerNameImpl(s32 fd, std::vector<u8>& write_buffer);
     Errno GetSockNameImpl(s32 fd, std::vector<u8>& write_buffer);
     Errno ListenImpl(s32 fd, s32 backlog);
@@ -157,9 +158,9 @@ private:
     std::pair<s32, Errno> RecvImpl(s32 fd, u32 flags, std::vector<u8>& message);
     std::pair<s32, Errno> RecvFromImpl(s32 fd, u32 flags, std::vector<u8>& message,
                                        std::vector<u8>& addr);
-    std::pair<s32, Errno> SendImpl(s32 fd, u32 flags, const std::vector<u8>& message);
-    std::pair<s32, Errno> SendToImpl(s32 fd, u32 flags, const std::vector<u8>& message,
-                                     const std::vector<u8>& addr);
+    std::pair<s32, Errno> SendImpl(s32 fd, u32 flags, std::span<const u8> message);
+    std::pair<s32, Errno> SendToImpl(s32 fd, u32 flags, std::span<const u8> message,
+                                     std::span<const u8> addr);
     Errno CloseImpl(s32 fd);
 
     s32 FindFreeFileDescriptorHandle() noexcept;
diff --git a/src/core/hle/service/sockets/sfdnsres.cpp b/src/core/hle/service/sockets/sfdnsres.cpp
index 097c37d7a..e96eda7f3 100644
--- a/src/core/hle/service/sockets/sfdnsres.cpp
+++ b/src/core/hle/service/sockets/sfdnsres.cpp
@@ -243,4 +243,4 @@ void SFDNSRES::GetAddrInfoRequestWithOptions(Kernel::HLERequestContext& ctx) {
     rb.Push(0);
 }
 
-} // namespace Service::Sockets
\ No newline at end of file
+} // namespace Service::Sockets
diff --git a/src/core/hle/service/ssl/ssl.cpp b/src/core/hle/service/ssl/ssl.cpp
index 3735e0452..dcf47083f 100644
--- a/src/core/hle/service/ssl/ssl.cpp
+++ b/src/core/hle/service/ssl/ssl.cpp
@@ -101,7 +101,7 @@ private:
     void ImportServerPki(Kernel::HLERequestContext& ctx) {
         IPC::RequestParser rp{ctx};
         const auto certificate_format = rp.PopEnum<CertificateFormat>();
-        const auto pkcs_12_certificates = ctx.ReadBuffer(0);
+        [[maybe_unused]] const auto pkcs_12_certificates = ctx.ReadBuffer(0);
 
         constexpr u64 server_id = 0;
 
@@ -113,13 +113,13 @@ private:
     }
 
     void ImportClientPki(Kernel::HLERequestContext& ctx) {
-        const auto pkcs_12_certificate = ctx.ReadBuffer(0);
-        const auto ascii_password = [&ctx] {
+        [[maybe_unused]] const auto pkcs_12_certificate = ctx.ReadBuffer(0);
+        [[maybe_unused]] const auto ascii_password = [&ctx] {
             if (ctx.CanReadBuffer(1)) {
                 return ctx.ReadBuffer(1);
             }
 
-            return std::vector<u8>{};
+            return std::span<const u8>{};
         }();
 
         constexpr u64 client_id = 0;
diff --git a/src/core/hle/service/vi/vi.cpp b/src/core/hle/service/vi/vi.cpp
index bb283e74e..2fb631183 100644
--- a/src/core/hle/service/vi/vi.cpp
+++ b/src/core/hle/service/vi/vi.cpp
@@ -603,7 +603,7 @@ private:
             return;
         }
 
-        const auto parcel = android::Parcel{NativeWindow{*buffer_queue_id}};
+        const auto parcel = android::OutputParcel{NativeWindow{*buffer_queue_id}};
         const auto buffer_size = ctx.WriteBuffer(parcel.Serialize());
 
         IPC::ResponseBuilder rb{ctx, 4};
@@ -649,7 +649,7 @@ private:
             return;
         }
 
-        const auto parcel = android::Parcel{NativeWindow{*buffer_queue_id}};
+        const auto parcel = android::OutputParcel{NativeWindow{*buffer_queue_id}};
         const auto buffer_size = ctx.WriteBuffer(parcel.Serialize());
 
         IPC::ResponseBuilder rb{ctx, 6};
diff --git a/src/core/internal_network/network.cpp b/src/core/internal_network/network.cpp
index 282ea1ff9..7494fb62d 100644
--- a/src/core/internal_network/network.cpp
+++ b/src/core/internal_network/network.cpp
@@ -550,7 +550,7 @@ std::pair<s32, Errno> Socket::RecvFrom(int flags, std::vector<u8>& message, Sock
     return {-1, GetAndLogLastError()};
 }
 
-std::pair<s32, Errno> Socket::Send(const std::vector<u8>& message, int flags) {
+std::pair<s32, Errno> Socket::Send(std::span<const u8> message, int flags) {
     ASSERT(message.size() < static_cast<size_t>(std::numeric_limits<int>::max()));
     ASSERT(flags == 0);
 
@@ -563,7 +563,7 @@ std::pair<s32, Errno> Socket::Send(const std::vector<u8>& message, int flags) {
     return {-1, GetAndLogLastError()};
 }
 
-std::pair<s32, Errno> Socket::SendTo(u32 flags, const std::vector<u8>& message,
+std::pair<s32, Errno> Socket::SendTo(u32 flags, std::span<const u8> message,
                                      const SockAddrIn* addr) {
     ASSERT(flags == 0);
 
diff --git a/src/core/internal_network/socket_proxy.cpp b/src/core/internal_network/socket_proxy.cpp
index 1e1c42cea..7a77171c2 100644
--- a/src/core/internal_network/socket_proxy.cpp
+++ b/src/core/internal_network/socket_proxy.cpp
@@ -182,7 +182,7 @@ std::pair<s32, Errno> ProxySocket::ReceivePacket(int flags, std::vector<u8>& mes
     return {static_cast<u32>(read_bytes), Errno::SUCCESS};
 }
 
-std::pair<s32, Errno> ProxySocket::Send(const std::vector<u8>& message, int flags) {
+std::pair<s32, Errno> ProxySocket::Send(std::span<const u8> message, int flags) {
     LOG_WARNING(Network, "(STUBBED) called");
     ASSERT(message.size() < static_cast<size_t>(std::numeric_limits<int>::max()));
     ASSERT(flags == 0);
@@ -200,7 +200,7 @@ void ProxySocket::SendPacket(ProxyPacket& packet) {
     }
 }
 
-std::pair<s32, Errno> ProxySocket::SendTo(u32 flags, const std::vector<u8>& message,
+std::pair<s32, Errno> ProxySocket::SendTo(u32 flags, std::span<const u8> message,
                                           const SockAddrIn* addr) {
     ASSERT(flags == 0);
 
diff --git a/src/core/internal_network/socket_proxy.h b/src/core/internal_network/socket_proxy.h
index f12b5f567..9421492bc 100644
--- a/src/core/internal_network/socket_proxy.h
+++ b/src/core/internal_network/socket_proxy.h
@@ -4,6 +4,7 @@
 #pragma once
 
 #include <mutex>
+#include <span>
 #include <vector>
 #include <queue>
 
@@ -48,11 +49,11 @@ public:
     std::pair<s32, Errno> ReceivePacket(int flags, std::vector<u8>& message, SockAddrIn* addr,
                                         std::size_t max_length);
 
-    std::pair<s32, Errno> Send(const std::vector<u8>& message, int flags) override;
+    std::pair<s32, Errno> Send(std::span<const u8> message, int flags) override;
 
     void SendPacket(ProxyPacket& packet);
 
-    std::pair<s32, Errno> SendTo(u32 flags, const std::vector<u8>& message,
+    std::pair<s32, Errno> SendTo(u32 flags, std::span<const u8> message,
                                  const SockAddrIn* addr) override;
 
     Errno SetLinger(bool enable, u32 linger) override;
diff --git a/src/core/internal_network/sockets.h b/src/core/internal_network/sockets.h
index 2e328c645..4c7489258 100644
--- a/src/core/internal_network/sockets.h
+++ b/src/core/internal_network/sockets.h
@@ -5,6 +5,7 @@
 
 #include <map>
 #include <memory>
+#include <span>
 #include <utility>
 
 #if defined(_WIN32)
@@ -66,9 +67,9 @@ public:
     virtual std::pair<s32, Errno> RecvFrom(int flags, std::vector<u8>& message,
                                            SockAddrIn* addr) = 0;
 
-    virtual std::pair<s32, Errno> Send(const std::vector<u8>& message, int flags) = 0;
+    virtual std::pair<s32, Errno> Send(std::span<const u8> message, int flags) = 0;
 
-    virtual std::pair<s32, Errno> SendTo(u32 flags, const std::vector<u8>& message,
+    virtual std::pair<s32, Errno> SendTo(u32 flags, std::span<const u8> message,
                                          const SockAddrIn* addr) = 0;
 
     virtual Errno SetLinger(bool enable, u32 linger) = 0;
@@ -138,9 +139,9 @@ public:
 
     std::pair<s32, Errno> RecvFrom(int flags, std::vector<u8>& message, SockAddrIn* addr) override;
 
-    std::pair<s32, Errno> Send(const std::vector<u8>& message, int flags) override;
+    std::pair<s32, Errno> Send(std::span<const u8> message, int flags) override;
 
-    std::pair<s32, Errno> SendTo(u32 flags, const std::vector<u8>& message,
+    std::pair<s32, Errno> SendTo(u32 flags, std::span<const u8> message,
                                  const SockAddrIn* addr) override;
 
     Errno SetLinger(bool enable, u32 linger) override;
diff --git a/src/core/reporter.cpp b/src/core/reporter.cpp
index 77821e047..59dfb8767 100644
--- a/src/core/reporter.cpp
+++ b/src/core/reporter.cpp
@@ -312,7 +312,7 @@ void Reporter::SaveUnimplementedAppletReport(
 }
 
 void Reporter::SavePlayReport(PlayReportType type, u64 title_id,
-                              const std::vector<std::vector<u8>>& data,
+                              const std::vector<std::span<const u8>>& data,
                               std::optional<u64> process_id, std::optional<u128> user_id) const {
     if (!IsReportingEnabled()) {
         return;
diff --git a/src/core/reporter.h b/src/core/reporter.h
index 9fdb9d6c1..bb11f8e7c 100644
--- a/src/core/reporter.h
+++ b/src/core/reporter.h
@@ -5,6 +5,7 @@
 
 #include <array>
 #include <optional>
+#include <span>
 #include <string>
 #include <vector>
 #include "common/common_types.h"
@@ -56,7 +57,8 @@ public:
         System,
     };
 
-    void SavePlayReport(PlayReportType type, u64 title_id, const std::vector<std::vector<u8>>& data,
+    void SavePlayReport(PlayReportType type, u64 title_id,
+                        const std::vector<std::span<const u8>>& data,
                         std::optional<u64> process_id = {}, std::optional<u128> user_id = {}) const;
 
     // Used by error applet
diff --git a/src/input_common/drivers/joycon.cpp b/src/input_common/drivers/joycon.cpp
index cedc94e63..4fcfb4510 100644
--- a/src/input_common/drivers/joycon.cpp
+++ b/src/input_common/drivers/joycon.cpp
@@ -668,12 +668,10 @@ std::string Joycons::JoyconName(Joycon::ControllerType type) const {
         return "Right Joycon";
     case Joycon::ControllerType::Pro:
         return "Pro Controller";
-    case Joycon::ControllerType::Grip:
-        return "Grip Controller";
     case Joycon::ControllerType::Dual:
         return "Dual Joycon";
     default:
-        return "Unknown Joycon";
+        return "Unknown Switch Controller";
     }
 }
 } // namespace InputCommon
diff --git a/src/input_common/drivers/joycon.h b/src/input_common/drivers/joycon.h
index 316d383d8..2149ab7fd 100644
--- a/src/input_common/drivers/joycon.h
+++ b/src/input_common/drivers/joycon.h
@@ -15,7 +15,7 @@ using SerialNumber = std::array<u8, 15>;
 struct Battery;
 struct Color;
 struct MotionData;
-enum class ControllerType;
+enum class ControllerType : u8;
 enum class DriverResult;
 enum class IrsResolution;
 class JoyconDriver;
diff --git a/src/input_common/helpers/joycon_driver.cpp b/src/input_common/helpers/joycon_driver.cpp
index 3775e2d35..8f94c9f45 100644
--- a/src/input_common/helpers/joycon_driver.cpp
+++ b/src/input_common/helpers/joycon_driver.cpp
@@ -162,14 +162,14 @@ void JoyconDriver::InputThread(std::stop_token stop_token) {
 }
 
 void JoyconDriver::OnNewData(std::span<u8> buffer) {
-    const auto report_mode = static_cast<InputReport>(buffer[0]);
+    const auto report_mode = static_cast<ReportMode>(buffer[0]);
 
     // Packages can be a litte bit inconsistent. Average the delta time to provide a smoother motion
     // experience
     switch (report_mode) {
-    case InputReport::STANDARD_FULL_60HZ:
-    case InputReport::NFC_IR_MODE_60HZ:
-    case InputReport::SIMPLE_HID_MODE: {
+    case ReportMode::STANDARD_FULL_60HZ:
+    case ReportMode::NFC_IR_MODE_60HZ:
+    case ReportMode::SIMPLE_HID_MODE: {
         const auto now = std::chrono::steady_clock::now();
         const auto new_delta_time = static_cast<u64>(
             std::chrono::duration_cast<std::chrono::microseconds>(now - last_update).count());
@@ -190,7 +190,7 @@ void JoyconDriver::OnNewData(std::span<u8> buffer) {
     };
 
     // TODO: Remove this when calibration is properly loaded and not calculated
-    if (ring_connected && report_mode == InputReport::STANDARD_FULL_60HZ) {
+    if (ring_connected && report_mode == ReportMode::STANDARD_FULL_60HZ) {
         InputReportActive data{};
         memcpy(&data, buffer.data(), sizeof(InputReportActive));
         calibration_protocol->GetRingCalibration(ring_calibration, data.ring_input);
@@ -228,16 +228,16 @@ void JoyconDriver::OnNewData(std::span<u8> buffer) {
     }
 
     switch (report_mode) {
-    case InputReport::STANDARD_FULL_60HZ:
+    case ReportMode::STANDARD_FULL_60HZ:
         joycon_poller->ReadActiveMode(buffer, motion_status, ring_status);
         break;
-    case InputReport::NFC_IR_MODE_60HZ:
+    case ReportMode::NFC_IR_MODE_60HZ:
         joycon_poller->ReadNfcIRMode(buffer, motion_status);
         break;
-    case InputReport::SIMPLE_HID_MODE:
+    case ReportMode::SIMPLE_HID_MODE:
         joycon_poller->ReadPassiveMode(buffer);
         break;
-    case InputReport::SUBCMD_REPLY:
+    case ReportMode::SUBCMD_REPLY:
         LOG_DEBUG(Input, "Unhandled command reply");
         break;
     default:
diff --git a/src/input_common/helpers/joycon_protocol/calibration.cpp b/src/input_common/helpers/joycon_protocol/calibration.cpp
index f6e7e97d5..d8f040f75 100644
--- a/src/input_common/helpers/joycon_protocol/calibration.cpp
+++ b/src/input_common/helpers/joycon_protocol/calibration.cpp
@@ -13,33 +13,33 @@ CalibrationProtocol::CalibrationProtocol(std::shared_ptr<JoyconHandle> handle)
 
 DriverResult CalibrationProtocol::GetLeftJoyStickCalibration(JoyStickCalibration& calibration) {
     ScopedSetBlocking sb(this);
-    std::vector<u8> buffer;
     DriverResult result{DriverResult::Success};
+    JoystickLeftSpiCalibration spi_calibration{};
+    bool has_user_calibration = false;
     calibration = {};
 
-    result = ReadSPI(CalAddr::USER_LEFT_MAGIC, sizeof(u16), buffer);
-
     if (result == DriverResult::Success) {
-        const bool has_user_calibration = buffer[0] == 0xB2 && buffer[1] == 0xA1;
-        if (has_user_calibration) {
-            result = ReadSPI(CalAddr::USER_LEFT_DATA, 9, buffer);
-        } else {
-            result = ReadSPI(CalAddr::FACT_LEFT_DATA, 9, buffer);
-        }
+        result = HasUserCalibration(SpiAddress::USER_LEFT_MAGIC, has_user_calibration);
     }
 
-    if (result == DriverResult::Success) {
-        calibration.x.max = static_cast<u16>(((buffer[1] & 0x0F) << 8) | buffer[0]);
-        calibration.y.max = static_cast<u16>((buffer[2] << 4) | (buffer[1] >> 4));
-        calibration.x.center = static_cast<u16>(((buffer[4] & 0x0F) << 8) | buffer[3]);
-        calibration.y.center = static_cast<u16>((buffer[5] << 4) | (buffer[4] >> 4));
-        calibration.x.min = static_cast<u16>(((buffer[7] & 0x0F) << 8) | buffer[6]);
-        calibration.y.min = static_cast<u16>((buffer[8] << 4) | (buffer[7] >> 4));
+    // Read User defined calibration
+    if (result == DriverResult::Success && has_user_calibration) {
+        result = ReadSPI(SpiAddress::USER_LEFT_DATA, spi_calibration);
     }
 
-    // Nintendo fix for drifting stick
-    // result = ReadSPI(0x60, 0x86 ,buffer, 16);
-    // calibration.deadzone = (u16)((buffer[4] << 8) & 0xF00 | buffer[3]);
+    // Read Factory calibration
+    if (result == DriverResult::Success && !has_user_calibration) {
+        result = ReadSPI(SpiAddress::FACT_LEFT_DATA, spi_calibration);
+    }
+
+    if (result == DriverResult::Success) {
+        calibration.x.center = GetXAxisCalibrationValue(spi_calibration.center);
+        calibration.y.center = GetYAxisCalibrationValue(spi_calibration.center);
+        calibration.x.min = GetXAxisCalibrationValue(spi_calibration.min);
+        calibration.y.min = GetYAxisCalibrationValue(spi_calibration.min);
+        calibration.x.max = GetXAxisCalibrationValue(spi_calibration.max);
+        calibration.y.max = GetYAxisCalibrationValue(spi_calibration.max);
+    }
 
     // Set a valid default calibration if data is missing
     ValidateCalibration(calibration);
@@ -49,33 +49,33 @@ DriverResult CalibrationProtocol::GetLeftJoyStickCalibration(JoyStickCalibration
 
 DriverResult CalibrationProtocol::GetRightJoyStickCalibration(JoyStickCalibration& calibration) {
     ScopedSetBlocking sb(this);
-    std::vector<u8> buffer;
     DriverResult result{DriverResult::Success};
+    JoystickRightSpiCalibration spi_calibration{};
+    bool has_user_calibration = false;
     calibration = {};
 
-    result = ReadSPI(CalAddr::USER_RIGHT_MAGIC, sizeof(u16), buffer);
-
     if (result == DriverResult::Success) {
-        const bool has_user_calibration = buffer[0] == 0xB2 && buffer[1] == 0xA1;
-        if (has_user_calibration) {
-            result = ReadSPI(CalAddr::USER_RIGHT_DATA, 9, buffer);
-        } else {
-            result = ReadSPI(CalAddr::FACT_RIGHT_DATA, 9, buffer);
-        }
+        result = HasUserCalibration(SpiAddress::USER_RIGHT_MAGIC, has_user_calibration);
     }
 
-    if (result == DriverResult::Success) {
-        calibration.x.center = static_cast<u16>(((buffer[1] & 0x0F) << 8) | buffer[0]);
-        calibration.y.center = static_cast<u16>((buffer[2] << 4) | (buffer[1] >> 4));
-        calibration.x.min = static_cast<u16>(((buffer[4] & 0x0F) << 8) | buffer[3]);
-        calibration.y.min = static_cast<u16>((buffer[5] << 4) | (buffer[4] >> 4));
-        calibration.x.max = static_cast<u16>(((buffer[7] & 0x0F) << 8) | buffer[6]);
-        calibration.y.max = static_cast<u16>((buffer[8] << 4) | (buffer[7] >> 4));
+    // Read User defined calibration
+    if (result == DriverResult::Success && has_user_calibration) {
+        result = ReadSPI(SpiAddress::USER_RIGHT_DATA, spi_calibration);
+    }
+
+    // Read Factory calibration
+    if (result == DriverResult::Success && !has_user_calibration) {
+        result = ReadSPI(SpiAddress::FACT_RIGHT_DATA, spi_calibration);
     }
 
-    // Nintendo fix for drifting stick
-    // buffer = ReadSPI(0x60, 0x98 , 16);
-    // joystick.deadzone = (u16)((buffer[4] << 8) & 0xF00 | buffer[3]);
+    if (result == DriverResult::Success) {
+        calibration.x.center = GetXAxisCalibrationValue(spi_calibration.center);
+        calibration.y.center = GetYAxisCalibrationValue(spi_calibration.center);
+        calibration.x.min = GetXAxisCalibrationValue(spi_calibration.min);
+        calibration.y.min = GetYAxisCalibrationValue(spi_calibration.min);
+        calibration.x.max = GetXAxisCalibrationValue(spi_calibration.max);
+        calibration.y.max = GetYAxisCalibrationValue(spi_calibration.max);
+    }
 
     // Set a valid default calibration if data is missing
     ValidateCalibration(calibration);
@@ -85,39 +85,41 @@ DriverResult CalibrationProtocol::GetRightJoyStickCalibration(JoyStickCalibratio
 
 DriverResult CalibrationProtocol::GetImuCalibration(MotionCalibration& calibration) {
     ScopedSetBlocking sb(this);
-    std::vector<u8> buffer;
     DriverResult result{DriverResult::Success};
+    ImuSpiCalibration spi_calibration{};
+    bool has_user_calibration = false;
     calibration = {};
 
-    result = ReadSPI(CalAddr::USER_IMU_MAGIC, sizeof(u16), buffer);
-
     if (result == DriverResult::Success) {
-        const bool has_user_calibration = buffer[0] == 0xB2 && buffer[1] == 0xA1;
-        if (has_user_calibration) {
-            result = ReadSPI(CalAddr::USER_IMU_DATA, sizeof(IMUCalibration), buffer);
-        } else {
-            result = ReadSPI(CalAddr::FACT_IMU_DATA, sizeof(IMUCalibration), buffer);
-        }
+        result = HasUserCalibration(SpiAddress::USER_IMU_MAGIC, has_user_calibration);
+    }
+
+    // Read User defined calibration
+    if (result == DriverResult::Success && has_user_calibration) {
+        result = ReadSPI(SpiAddress::USER_IMU_DATA, spi_calibration);
+    }
+
+    // Read Factory calibration
+    if (result == DriverResult::Success && !has_user_calibration) {
+        result = ReadSPI(SpiAddress::FACT_IMU_DATA, spi_calibration);
     }
 
     if (result == DriverResult::Success) {
-        IMUCalibration device_calibration{};
-        memcpy(&device_calibration, buffer.data(), sizeof(IMUCalibration));
-        calibration.accelerometer[0].offset = device_calibration.accelerometer_offset[0];
-        calibration.accelerometer[1].offset = device_calibration.accelerometer_offset[1];
-        calibration.accelerometer[2].offset = device_calibration.accelerometer_offset[2];
+        calibration.accelerometer[0].offset = spi_calibration.accelerometer_offset[0];
+        calibration.accelerometer[1].offset = spi_calibration.accelerometer_offset[1];
+        calibration.accelerometer[2].offset = spi_calibration.accelerometer_offset[2];
 
-        calibration.accelerometer[0].scale = device_calibration.accelerometer_scale[0];
-        calibration.accelerometer[1].scale = device_calibration.accelerometer_scale[1];
-        calibration.accelerometer[2].scale = device_calibration.accelerometer_scale[2];
+        calibration.accelerometer[0].scale = spi_calibration.accelerometer_scale[0];
+        calibration.accelerometer[1].scale = spi_calibration.accelerometer_scale[1];
+        calibration.accelerometer[2].scale = spi_calibration.accelerometer_scale[2];
 
-        calibration.gyro[0].offset = device_calibration.gyroscope_offset[0];
-        calibration.gyro[1].offset = device_calibration.gyroscope_offset[1];
-        calibration.gyro[2].offset = device_calibration.gyroscope_offset[2];
+        calibration.gyro[0].offset = spi_calibration.gyroscope_offset[0];
+        calibration.gyro[1].offset = spi_calibration.gyroscope_offset[1];
+        calibration.gyro[2].offset = spi_calibration.gyroscope_offset[2];
 
-        calibration.gyro[0].scale = device_calibration.gyroscope_scale[0];
-        calibration.gyro[1].scale = device_calibration.gyroscope_scale[1];
-        calibration.gyro[2].scale = device_calibration.gyroscope_scale[2];
+        calibration.gyro[0].scale = spi_calibration.gyroscope_scale[0];
+        calibration.gyro[1].scale = spi_calibration.gyroscope_scale[1];
+        calibration.gyro[2].scale = spi_calibration.gyroscope_scale[2];
     }
 
     ValidateCalibration(calibration);
@@ -127,10 +129,12 @@ DriverResult CalibrationProtocol::GetImuCalibration(MotionCalibration& calibrati
 
 DriverResult CalibrationProtocol::GetRingCalibration(RingCalibration& calibration,
                                                      s16 current_value) {
+    constexpr s16 DefaultRingRange{800};
+
     // TODO: Get default calibration form ring itself
     if (ring_data_max == 0 && ring_data_min == 0) {
-        ring_data_max = current_value + 800;
-        ring_data_min = current_value - 800;
+        ring_data_max = current_value + DefaultRingRange;
+        ring_data_min = current_value - DefaultRingRange;
         ring_data_default = current_value;
     }
     ring_data_max = std::max(ring_data_max, current_value);
@@ -143,42 +147,72 @@ DriverResult CalibrationProtocol::GetRingCalibration(RingCalibration& calibratio
     return DriverResult::Success;
 }
 
+DriverResult CalibrationProtocol::HasUserCalibration(SpiAddress address,
+                                                     bool& has_user_calibration) {
+    MagicSpiCalibration spi_magic{};
+    const DriverResult result{ReadSPI(address, spi_magic)};
+    has_user_calibration = false;
+    if (result == DriverResult::Success) {
+        has_user_calibration = spi_magic.first == CalibrationMagic::USR_MAGIC_0 &&
+                               spi_magic.second == CalibrationMagic::USR_MAGIC_1;
+    }
+    return result;
+}
+
+u16 CalibrationProtocol::GetXAxisCalibrationValue(std::span<u8> block) const {
+    return static_cast<u16>(((block[1] & 0x0F) << 8) | block[0]);
+}
+
+u16 CalibrationProtocol::GetYAxisCalibrationValue(std::span<u8> block) const {
+    return static_cast<u16>((block[2] << 4) | (block[1] >> 4));
+}
+
 void CalibrationProtocol::ValidateCalibration(JoyStickCalibration& calibration) {
-    constexpr u16 DefaultStickCenter{2048};
-    constexpr u16 DefaultStickRange{1740};
+    constexpr u16 DefaultStickCenter{0x800};
+    constexpr u16 DefaultStickRange{0x6cc};
 
-    if (calibration.x.center == 0xFFF || calibration.x.center == 0) {
-        calibration.x.center = DefaultStickCenter;
-    }
-    if (calibration.x.max == 0xFFF || calibration.x.max == 0) {
-        calibration.x.max = DefaultStickRange;
+    calibration.x.center = ValidateValue(calibration.x.center, DefaultStickCenter);
+    calibration.x.max = ValidateValue(calibration.x.max, DefaultStickRange);
+    calibration.x.min = ValidateValue(calibration.x.min, DefaultStickRange);
+
+    calibration.y.center = ValidateValue(calibration.y.center, DefaultStickCenter);
+    calibration.y.max = ValidateValue(calibration.y.max, DefaultStickRange);
+    calibration.y.min = ValidateValue(calibration.y.min, DefaultStickRange);
+}
+
+void CalibrationProtocol::ValidateCalibration(MotionCalibration& calibration) {
+    constexpr s16 DefaultAccelerometerScale{0x4000};
+    constexpr s16 DefaultGyroScale{0x3be7};
+    constexpr s16 DefaultOffset{0};
+
+    for (auto& sensor : calibration.accelerometer) {
+        sensor.scale = ValidateValue(sensor.scale, DefaultAccelerometerScale);
+        sensor.offset = ValidateValue(sensor.offset, DefaultOffset);
     }
-    if (calibration.x.min == 0xFFF || calibration.x.min == 0) {
-        calibration.x.min = DefaultStickRange;
+    for (auto& sensor : calibration.gyro) {
+        sensor.scale = ValidateValue(sensor.scale, DefaultGyroScale);
+        sensor.offset = ValidateValue(sensor.offset, DefaultOffset);
     }
+}
 
-    if (calibration.y.center == 0xFFF || calibration.y.center == 0) {
-        calibration.y.center = DefaultStickCenter;
-    }
-    if (calibration.y.max == 0xFFF || calibration.y.max == 0) {
-        calibration.y.max = DefaultStickRange;
+u16 CalibrationProtocol::ValidateValue(u16 value, u16 default_value) const {
+    if (value == 0) {
+        return default_value;
     }
-    if (calibration.y.min == 0xFFF || calibration.y.min == 0) {
-        calibration.y.min = DefaultStickRange;
+    if (value == 0xFFF) {
+        return default_value;
     }
+    return value;
 }
 
-void CalibrationProtocol::ValidateCalibration(MotionCalibration& calibration) {
-    for (auto& sensor : calibration.accelerometer) {
-        if (sensor.scale == 0) {
-            sensor.scale = 0x4000;
-        }
+s16 CalibrationProtocol::ValidateValue(s16 value, s16 default_value) const {
+    if (value == 0) {
+        return default_value;
     }
-    for (auto& sensor : calibration.gyro) {
-        if (sensor.scale == 0) {
-            sensor.scale = 0x3be7;
-        }
+    if (value == 0xFFF) {
+        return default_value;
     }
+    return value;
 }
 
 } // namespace InputCommon::Joycon
diff --git a/src/input_common/helpers/joycon_protocol/calibration.h b/src/input_common/helpers/joycon_protocol/calibration.h
index afb52a36a..c6fd0f729 100644
--- a/src/input_common/helpers/joycon_protocol/calibration.h
+++ b/src/input_common/helpers/joycon_protocol/calibration.h
@@ -53,9 +53,27 @@ public:
     DriverResult GetRingCalibration(RingCalibration& calibration, s16 current_value);
 
 private:
+    /// Returns true if the specified address corresponds to the magic value of user calibration
+    DriverResult HasUserCalibration(SpiAddress address, bool& has_user_calibration);
+
+    /// Converts a raw calibration block to an u16 value containing the x axis value
+    u16 GetXAxisCalibrationValue(std::span<u8> block) const;
+
+    /// Converts a raw calibration block to an u16 value containing the y axis value
+    u16 GetYAxisCalibrationValue(std::span<u8> block) const;
+
+    /// Ensures that all joystick calibration values are set
     void ValidateCalibration(JoyStickCalibration& calibration);
+
+    /// Ensures that all motion calibration values are set
     void ValidateCalibration(MotionCalibration& calibration);
 
+    /// Returns the default value if the value is either zero or 0xFFF
+    u16 ValidateValue(u16 value, u16 default_value) const;
+
+    /// Returns the default value if the value is either zero or 0xFFF
+    s16 ValidateValue(s16 value, s16 default_value) const;
+
     s16 ring_data_max = 0;
     s16 ring_data_default = 0;
     s16 ring_data_min = 0;
diff --git a/src/input_common/helpers/joycon_protocol/common_protocol.cpp b/src/input_common/helpers/joycon_protocol/common_protocol.cpp
index 417d0dcc5..2b42a4555 100644
--- a/src/input_common/helpers/joycon_protocol/common_protocol.cpp
+++ b/src/input_common/helpers/joycon_protocol/common_protocol.cpp
@@ -22,12 +22,9 @@ void JoyconCommonProtocol::SetNonBlocking() {
 }
 
 DriverResult JoyconCommonProtocol::GetDeviceType(ControllerType& controller_type) {
-    std::vector<u8> buffer;
-    const auto result = ReadSPI(CalAddr::DEVICE_TYPE, 1, buffer);
-    controller_type = ControllerType::None;
+    const auto result = ReadSPI(SpiAddress::DEVICE_TYPE, controller_type);
 
     if (result == DriverResult::Success) {
-        controller_type = static_cast<ControllerType>(buffer[0]);
         // Fallback to 3rd party pro controllers
         if (controller_type == ControllerType::None) {
             controller_type = ControllerType::Pro;
@@ -40,6 +37,7 @@ DriverResult JoyconCommonProtocol::GetDeviceType(ControllerType& controller_type
 DriverResult JoyconCommonProtocol::CheckDeviceAccess(SDL_hid_device_info* device_info) {
     ControllerType controller_type{ControllerType::None};
     const auto result = GetDeviceType(controller_type);
+
     if (result != DriverResult::Success || controller_type == ControllerType::None) {
         return DriverResult::UnsupportedControllerType;
     }
@@ -62,7 +60,7 @@ DriverResult JoyconCommonProtocol::SetReportMode(ReportMode report_mode) {
     return SendSubCommand(SubCommand::SET_REPORT_MODE, buffer);
 }
 
-DriverResult JoyconCommonProtocol::SendData(std::span<const u8> buffer) {
+DriverResult JoyconCommonProtocol::SendRawData(std::span<const u8> buffer) {
     const auto result = SDL_hid_write(hidapi_handle->handle, buffer.data(), buffer.size());
 
     if (result == -1) {
@@ -72,15 +70,15 @@ DriverResult JoyconCommonProtocol::SendData(std::span<const u8> buffer) {
     return DriverResult::Success;
 }
 
-DriverResult JoyconCommonProtocol::GetSubCommandResponse(SubCommand sc, std::vector<u8>& output) {
+DriverResult JoyconCommonProtocol::GetSubCommandResponse(SubCommand sc,
+                                                         SubCommandResponse& output) {
     constexpr int timeout_mili = 66;
     constexpr int MaxTries = 15;
     int tries = 0;
-    output.resize(MaxSubCommandResponseSize);
 
     do {
-        int result = SDL_hid_read_timeout(hidapi_handle->handle, output.data(),
-                                          MaxSubCommandResponseSize, timeout_mili);
+        int result = SDL_hid_read_timeout(hidapi_handle->handle, reinterpret_cast<u8*>(&output),
+                                          sizeof(SubCommandResponse), timeout_mili);
 
         if (result < 1) {
             LOG_ERROR(Input, "No response from joycon");
@@ -88,27 +86,28 @@ DriverResult JoyconCommonProtocol::GetSubCommandResponse(SubCommand sc, std::vec
         if (tries++ > MaxTries) {
             return DriverResult::Timeout;
         }
-    } while (output[0] != 0x21 && output[14] != static_cast<u8>(sc));
-
-    if (output[0] != 0x21 && output[14] != static_cast<u8>(sc)) {
-        return DriverResult::WrongReply;
-    }
+    } while (output.input_report.report_mode != ReportMode::SUBCMD_REPLY &&
+             output.sub_command != sc);
 
     return DriverResult::Success;
 }
 
 DriverResult JoyconCommonProtocol::SendSubCommand(SubCommand sc, std::span<const u8> buffer,
-                                                  std::vector<u8>& output) {
-    std::vector<u8> local_buffer(MaxResponseSize);
-
-    local_buffer[0] = static_cast<u8>(OutputReport::RUMBLE_AND_SUBCMD);
-    local_buffer[1] = GetCounter();
-    local_buffer[10] = static_cast<u8>(sc);
-    for (std::size_t i = 0; i < buffer.size(); ++i) {
-        local_buffer[11 + i] = buffer[i];
+                                                  SubCommandResponse& output) {
+    SubCommandPacket packet{
+        .output_report = OutputReport::RUMBLE_AND_SUBCMD,
+        .packet_counter = GetCounter(),
+        .sub_command = sc,
+        .command_data = {},
+    };
+
+    if (buffer.size() > packet.command_data.size()) {
+        return DriverResult::InvalidParameters;
     }
 
-    auto result = SendData(local_buffer);
+    memcpy(packet.command_data.data(), buffer.data(), buffer.size());
+
+    auto result = SendData(packet);
 
     if (result != DriverResult::Success) {
         return result;
@@ -120,44 +119,57 @@ DriverResult JoyconCommonProtocol::SendSubCommand(SubCommand sc, std::span<const
 }
 
 DriverResult JoyconCommonProtocol::SendSubCommand(SubCommand sc, std::span<const u8> buffer) {
-    std::vector<u8> output;
+    SubCommandResponse output{};
     return SendSubCommand(sc, buffer, output);
 }
 
 DriverResult JoyconCommonProtocol::SendMCUCommand(SubCommand sc, std::span<const u8> buffer) {
-    std::vector<u8> local_buffer(MaxResponseSize);
-
-    local_buffer[0] = static_cast<u8>(OutputReport::MCU_DATA);
-    local_buffer[1] = GetCounter();
-    local_buffer[10] = static_cast<u8>(sc);
-    for (std::size_t i = 0; i < buffer.size(); ++i) {
-        local_buffer[11 + i] = buffer[i];
+    SubCommandPacket packet{
+        .output_report = OutputReport::MCU_DATA,
+        .packet_counter = GetCounter(),
+        .sub_command = sc,
+        .command_data = {},
+    };
+
+    if (buffer.size() > packet.command_data.size()) {
+        return DriverResult::InvalidParameters;
     }
 
-    return SendData(local_buffer);
+    memcpy(packet.command_data.data(), buffer.data(), buffer.size());
+
+    return SendData(packet);
 }
 
 DriverResult JoyconCommonProtocol::SendVibrationReport(std::span<const u8> buffer) {
-    std::vector<u8> local_buffer(MaxResponseSize);
-
-    local_buffer[0] = static_cast<u8>(Joycon::OutputReport::RUMBLE_ONLY);
-    local_buffer[1] = GetCounter();
+    VibrationPacket packet{
+        .output_report = OutputReport::RUMBLE_ONLY,
+        .packet_counter = GetCounter(),
+        .vibration_data = {},
+    };
+
+    if (buffer.size() > packet.vibration_data.size()) {
+        return DriverResult::InvalidParameters;
+    }
 
-    memcpy(local_buffer.data() + 2, buffer.data(), buffer.size());
+    memcpy(packet.vibration_data.data(), buffer.data(), buffer.size());
 
-    return SendData(local_buffer);
+    return SendData(packet);
 }
 
-DriverResult JoyconCommonProtocol::ReadSPI(CalAddr addr, u8 size, std::vector<u8>& output) {
+DriverResult JoyconCommonProtocol::ReadRawSPI(SpiAddress addr, std::span<u8> output) {
+    constexpr std::size_t HeaderSize = 5;
     constexpr std::size_t MaxTries = 10;
     std::size_t tries = 0;
-    std::array<u8, 5> buffer = {0x00, 0x00, 0x00, 0x00, size};
-    std::vector<u8> local_buffer(size + 20);
-
-    buffer[0] = static_cast<u8>(static_cast<u16>(addr) & 0x00FF);
-    buffer[1] = static_cast<u8>((static_cast<u16>(addr) & 0xFF00) >> 8);
+    SubCommandResponse response{};
+    std::array<u8, sizeof(ReadSpiPacket)> buffer{};
+    const ReadSpiPacket packet_data{
+        .spi_address = addr,
+        .size = static_cast<u8>(output.size()),
+    };
+
+    memcpy(buffer.data(), &packet_data, sizeof(ReadSpiPacket));
     do {
-        const auto result = SendSubCommand(SubCommand::SPI_FLASH_READ, buffer, local_buffer);
+        const auto result = SendSubCommand(SubCommand::SPI_FLASH_READ, buffer, response);
         if (result != DriverResult::Success) {
             return result;
         }
@@ -165,10 +177,14 @@ DriverResult JoyconCommonProtocol::ReadSPI(CalAddr addr, u8 size, std::vector<u8
         if (tries++ > MaxTries) {
             return DriverResult::Timeout;
         }
-    } while (local_buffer[15] != buffer[0] || local_buffer[16] != buffer[1]);
+    } while (response.spi_address != addr);
+
+    if (response.command_data.size() < packet_data.size + HeaderSize) {
+        return DriverResult::WrongReply;
+    }
 
     // Remove header from output
-    output = std::vector<u8>(local_buffer.begin() + 20, local_buffer.begin() + 20 + size);
+    memcpy(output.data(), response.command_data.data() + HeaderSize, packet_data.size);
     return DriverResult::Success;
 }
 
@@ -177,7 +193,7 @@ DriverResult JoyconCommonProtocol::EnableMCU(bool enable) {
     const auto result = SendSubCommand(SubCommand::SET_MCU_STATE, mcu_state);
 
     if (result != DriverResult::Success) {
-        LOG_ERROR(Input, "SendMCUData failed with error {}", result);
+        LOG_ERROR(Input, "Failed with error {}", result);
     }
 
     return result;
@@ -192,22 +208,21 @@ DriverResult JoyconCommonProtocol::ConfigureMCU(const MCUConfig& config) {
     const auto result = SendSubCommand(SubCommand::SET_MCU_CONFIG, config_buffer);
 
     if (result != DriverResult::Success) {
-        LOG_ERROR(Input, "Set MCU config failed with error {}", result);
+        LOG_ERROR(Input, "Failed with error {}", result);
     }
 
     return result;
 }
 
-DriverResult JoyconCommonProtocol::GetMCUDataResponse(ReportMode report_mode_,
-                                                      std::vector<u8>& output) {
-    const int report_mode = static_cast<u8>(report_mode_);
+DriverResult JoyconCommonProtocol::GetMCUDataResponse(ReportMode report_mode,
+                                                      MCUCommandResponse& output) {
     constexpr int TimeoutMili = 200;
     constexpr int MaxTries = 9;
     int tries = 0;
-    output.resize(0x170);
 
     do {
-        int result = SDL_hid_read_timeout(hidapi_handle->handle, output.data(), 0x170, TimeoutMili);
+        int result = SDL_hid_read_timeout(hidapi_handle->handle, reinterpret_cast<u8*>(&output),
+                                          sizeof(MCUCommandResponse), TimeoutMili);
 
         if (result < 1) {
             LOG_ERROR(Input, "No response from joycon attempt {}", tries);
@@ -215,28 +230,29 @@ DriverResult JoyconCommonProtocol::GetMCUDataResponse(ReportMode report_mode_,
         if (tries++ > MaxTries) {
             return DriverResult::Timeout;
         }
-    } while (output[0] != report_mode || output[49] == 0xFF);
-
-    if (output[0] != report_mode || output[49] == 0xFF) {
-        return DriverResult::WrongReply;
-    }
+    } while (output.input_report.report_mode != report_mode ||
+             output.mcu_report == MCUReport::EmptyAwaitingCmd);
 
     return DriverResult::Success;
 }
 
 DriverResult JoyconCommonProtocol::SendMCUData(ReportMode report_mode, SubCommand sc,
                                                std::span<const u8> buffer,
-                                               std::vector<u8>& output) {
-    std::vector<u8> local_buffer(MaxResponseSize);
-
-    local_buffer[0] = static_cast<u8>(OutputReport::MCU_DATA);
-    local_buffer[1] = GetCounter();
-    local_buffer[9] = static_cast<u8>(sc);
-    for (std::size_t i = 0; i < buffer.size(); ++i) {
-        local_buffer[10 + i] = buffer[i];
+                                               MCUCommandResponse& output) {
+    SubCommandPacket packet{
+        .output_report = OutputReport::MCU_DATA,
+        .packet_counter = GetCounter(),
+        .sub_command = sc,
+        .command_data = {},
+    };
+
+    if (buffer.size() > packet.command_data.size()) {
+        return DriverResult::InvalidParameters;
     }
 
-    auto result = SendData(local_buffer);
+    memcpy(packet.command_data.data(), buffer.data(), buffer.size());
+
+    auto result = SendData(packet);
 
     if (result != DriverResult::Success) {
         return result;
@@ -248,7 +264,7 @@ DriverResult JoyconCommonProtocol::SendMCUData(ReportMode report_mode, SubComman
 }
 
 DriverResult JoyconCommonProtocol::WaitSetMCUMode(ReportMode report_mode, MCUMode mode) {
-    std::vector<u8> output;
+    MCUCommandResponse output{};
     constexpr std::size_t MaxTries{8};
     std::size_t tries{};
 
@@ -263,7 +279,8 @@ DriverResult JoyconCommonProtocol::WaitSetMCUMode(ReportMode report_mode, MCUMod
         if (tries++ > MaxTries) {
             return DriverResult::WrongReply;
         }
-    } while (output[49] != 1 || output[56] != static_cast<u8>(mode));
+    } while (output.mcu_report != MCUReport::StateReport ||
+             output.mcu_data[6] != static_cast<u8>(mode));
 
     return DriverResult::Success;
 }
diff --git a/src/input_common/helpers/joycon_protocol/common_protocol.h b/src/input_common/helpers/joycon_protocol/common_protocol.h
index 903bcf402..f44f73ba4 100644
--- a/src/input_common/helpers/joycon_protocol/common_protocol.h
+++ b/src/input_common/helpers/joycon_protocol/common_protocol.h
@@ -57,22 +57,31 @@ public:
      * Sends data to the joycon device
      * @param buffer data to be send
      */
-    DriverResult SendData(std::span<const u8> buffer);
+    DriverResult SendRawData(std::span<const u8> buffer);
+
+    template <typename Output>
+        requires std::is_trivially_copyable_v<Output>
+    DriverResult SendData(const Output& output) {
+        std::array<u8, sizeof(Output)> buffer;
+        std::memcpy(buffer.data(), &output, sizeof(Output));
+        return SendRawData(buffer);
+    }
 
     /**
      * Waits for incoming data of the joycon device that matchs the subcommand
      * @param sub_command type of data to be returned
-     * @returns a buffer containing the responce
+     * @returns a buffer containing the response
      */
-    DriverResult GetSubCommandResponse(SubCommand sub_command, std::vector<u8>& output);
+    DriverResult GetSubCommandResponse(SubCommand sub_command, SubCommandResponse& output);
 
     /**
      * Sends a sub command to the device and waits for it's reply
      * @param sc sub command to be send
      * @param buffer data to be send
-     * @returns output buffer containing the responce
+     * @returns output buffer containing the response
      */
-    DriverResult SendSubCommand(SubCommand sc, std::span<const u8> buffer, std::vector<u8>& output);
+    DriverResult SendSubCommand(SubCommand sc, std::span<const u8> buffer,
+                                SubCommandResponse& output);
 
     /**
      * Sends a sub command to the device and waits for it's reply and ignores the output
@@ -97,10 +106,29 @@ public:
     /**
      * Reads the SPI memory stored on the joycon
      * @param Initial address location
-     * @param size in bytes to be read
-     * @returns output buffer containing the responce
+     * @returns output buffer containing the response
      */
-    DriverResult ReadSPI(CalAddr addr, u8 size, std::vector<u8>& output);
+    DriverResult ReadRawSPI(SpiAddress addr, std::span<u8> output);
+
+    /**
+     * Reads the SPI memory stored on the joycon
+     * @param Initial address location
+     * @returns output object containing the response
+     */
+    template <typename Output>
+        requires std::is_trivially_copyable_v<Output>
+    DriverResult ReadSPI(SpiAddress addr, Output& output) {
+        std::array<u8, sizeof(Output)> buffer;
+        output = {};
+
+        const auto result = ReadRawSPI(addr, buffer);
+        if (result != DriverResult::Success) {
+            return result;
+        }
+
+        std::memcpy(&output, buffer.data(), sizeof(Output));
+        return DriverResult::Success;
+    }
 
     /**
      * Enables MCU chip on the joycon
@@ -117,19 +145,19 @@ public:
     /**
      * Waits until there's MCU data available. On timeout returns error
      * @param report mode of the expected reply
-     * @returns a buffer containing the responce
+     * @returns a buffer containing the response
      */
-    DriverResult GetMCUDataResponse(ReportMode report_mode_, std::vector<u8>& output);
+    DriverResult GetMCUDataResponse(ReportMode report_mode_, MCUCommandResponse& output);
 
     /**
      * Sends data to the MCU chip and waits for it's reply
      * @param report mode of the expected reply
      * @param sub command to be send
      * @param buffer data to be send
-     * @returns output buffer containing the responce
+     * @returns output buffer containing the response
      */
     DriverResult SendMCUData(ReportMode report_mode, SubCommand sc, std::span<const u8> buffer,
-                             std::vector<u8>& output);
+                             MCUCommandResponse& output);
 
     /**
      * Wait's until the MCU chip is on the specified mode
diff --git a/src/input_common/helpers/joycon_protocol/generic_functions.cpp b/src/input_common/helpers/joycon_protocol/generic_functions.cpp
index 63cfb1369..548a4b9e3 100644
--- a/src/input_common/helpers/joycon_protocol/generic_functions.cpp
+++ b/src/input_common/helpers/joycon_protocol/generic_functions.cpp
@@ -32,13 +32,13 @@ DriverResult GenericProtocol::TriggersElapsed() {
 
 DriverResult GenericProtocol::GetDeviceInfo(DeviceInfo& device_info) {
     ScopedSetBlocking sb(this);
-    std::vector<u8> output;
+    SubCommandResponse output{};
 
     const auto result = SendSubCommand(SubCommand::REQ_DEV_INFO, {}, output);
 
     device_info = {};
     if (result == DriverResult::Success) {
-        memcpy(&device_info, output.data(), sizeof(DeviceInfo));
+        device_info = output.device_info;
     }
 
     return result;
@@ -71,8 +71,8 @@ DriverResult GenericProtocol::GetBattery(u32& battery_level) {
 
 DriverResult GenericProtocol::GetColor(Color& color) {
     ScopedSetBlocking sb(this);
-    std::vector<u8> buffer;
-    const auto result = ReadSPI(CalAddr::COLOR_DATA, 12, buffer);
+    std::array<u8, 12> buffer{};
+    const auto result = ReadRawSPI(SpiAddress::COLOR_DATA, buffer);
 
     color = {};
     if (result == DriverResult::Success) {
@@ -87,8 +87,8 @@ DriverResult GenericProtocol::GetColor(Color& color) {
 
 DriverResult GenericProtocol::GetSerialNumber(SerialNumber& serial_number) {
     ScopedSetBlocking sb(this);
-    std::vector<u8> buffer;
-    const auto result = ReadSPI(CalAddr::SERIAL_NUMBER, 16, buffer);
+    std::array<u8, 16> buffer{};
+    const auto result = ReadRawSPI(SpiAddress::SERIAL_NUMBER, buffer);
 
     serial_number = {};
     if (result == DriverResult::Success) {
diff --git a/src/input_common/helpers/joycon_protocol/irs.cpp b/src/input_common/helpers/joycon_protocol/irs.cpp
index 09e17bc5b..731fd5981 100644
--- a/src/input_common/helpers/joycon_protocol/irs.cpp
+++ b/src/input_common/helpers/joycon_protocol/irs.cpp
@@ -132,7 +132,7 @@ DriverResult IrsProtocol::RequestImage(std::span<u8> buffer) {
 DriverResult IrsProtocol::ConfigureIrs() {
     LOG_DEBUG(Input, "Configure IRS");
     constexpr std::size_t max_tries = 28;
-    std::vector<u8> output;
+    SubCommandResponse output{};
     std::size_t tries = 0;
 
     const IrsConfigure irs_configuration{
@@ -158,7 +158,7 @@ DriverResult IrsProtocol::ConfigureIrs() {
         if (tries++ >= max_tries) {
             return DriverResult::WrongReply;
         }
-    } while (output[15] != 0x0b);
+    } while (output.command_data[0] != 0x0b);
 
     return DriverResult::Success;
 }
@@ -167,7 +167,7 @@ DriverResult IrsProtocol::WriteRegistersStep1() {
     LOG_DEBUG(Input, "WriteRegistersStep1");
     DriverResult result{DriverResult::Success};
     constexpr std::size_t max_tries = 28;
-    std::vector<u8> output;
+    SubCommandResponse output{};
     std::size_t tries = 0;
 
     const IrsWriteRegisters irs_registers{
@@ -218,7 +218,8 @@ DriverResult IrsProtocol::WriteRegistersStep1() {
         if (tries++ >= max_tries) {
             return DriverResult::WrongReply;
         }
-    } while (!(output[15] == 0x13 && output[17] == 0x07) && output[15] != 0x23);
+    } while (!(output.command_data[0] == 0x13 && output.command_data[2] == 0x07) &&
+             output.command_data[0] != 0x23);
 
     return DriverResult::Success;
 }
@@ -226,7 +227,7 @@ DriverResult IrsProtocol::WriteRegistersStep1() {
 DriverResult IrsProtocol::WriteRegistersStep2() {
     LOG_DEBUG(Input, "WriteRegistersStep2");
     constexpr std::size_t max_tries = 28;
-    std::vector<u8> output;
+    SubCommandResponse output{};
     std::size_t tries = 0;
 
     const IrsWriteRegisters irs_registers{
@@ -260,7 +261,7 @@ DriverResult IrsProtocol::WriteRegistersStep2() {
         if (tries++ >= max_tries) {
             return DriverResult::WrongReply;
         }
-    } while (output[15] != 0x13 && output[15] != 0x23);
+    } while (output.command_data[0] != 0x13 && output.command_data[0] != 0x23);
 
     return DriverResult::Success;
 }
diff --git a/src/input_common/helpers/joycon_protocol/joycon_types.h b/src/input_common/helpers/joycon_protocol/joycon_types.h
index 182d2c15b..b91934990 100644
--- a/src/input_common/helpers/joycon_protocol/joycon_types.h
+++ b/src/input_common/helpers/joycon_protocol/joycon_types.h
@@ -19,20 +19,24 @@
 namespace InputCommon::Joycon {
 constexpr u32 MaxErrorCount = 50;
 constexpr u32 MaxBufferSize = 368;
-constexpr u32 MaxResponseSize = 49;
-constexpr u32 MaxSubCommandResponseSize = 64;
 constexpr std::array<u8, 8> DefaultVibrationBuffer{0x0, 0x1, 0x40, 0x40, 0x0, 0x1, 0x40, 0x40};
 
 using MacAddress = std::array<u8, 6>;
 using SerialNumber = std::array<u8, 15>;
 
-enum class ControllerType {
-    None,
-    Left,
-    Right,
-    Pro,
-    Grip,
-    Dual,
+enum class ControllerType : u8 {
+    None = 0x00,
+    Left = 0x01,
+    Right = 0x02,
+    Pro = 0x03,
+    Dual = 0x05, // TODO: Verify this id
+    LarkHvc1 = 0x07,
+    LarkHvc2 = 0x08,
+    LarkNesLeft = 0x09,
+    LarkNesRight = 0x0A,
+    Lucia = 0x0B,
+    Lagon = 0x0C,
+    Lager = 0x0D,
 };
 
 enum class PadAxes {
@@ -99,14 +103,6 @@ enum class OutputReport : u8 {
     USB_CMD = 0x80,
 };
 
-enum class InputReport : u8 {
-    SUBCMD_REPLY = 0x21,
-    STANDARD_FULL_60HZ = 0x30,
-    NFC_IR_MODE_60HZ = 0x31,
-    SIMPLE_HID_MODE = 0x3F,
-    INPUT_USB_RESPONSE = 0x81,
-};
-
 enum class FeatureReport : u8 {
     Last_SUBCMD = 0x02,
     OTA_GW_UPGRADE = 0x70,
@@ -143,9 +139,10 @@ enum class SubCommand : u8 {
     ENABLE_VIBRATION = 0x48,
     GET_REGULATED_VOLTAGE = 0x50,
     SET_EXTERNAL_CONFIG = 0x58,
-    UNKNOWN_RINGCON = 0x59,
-    UNKNOWN_RINGCON2 = 0x5A,
-    UNKNOWN_RINGCON3 = 0x5C,
+    GET_EXTERNAL_DEVICE_INFO = 0x59,
+    ENABLE_EXTERNAL_POLLING = 0x5A,
+    DISABLE_EXTERNAL_POLLING = 0x5B,
+    SET_EXTERNAL_FORMAT_CONFIG = 0x5C,
 };
 
 enum class UsbSubCommand : u8 {
@@ -159,26 +156,31 @@ enum class UsbSubCommand : u8 {
     SEND_UART = 0x92,
 };
 
-enum class CalMagic : u8 {
+enum class CalibrationMagic : u8 {
     USR_MAGIC_0 = 0xB2,
     USR_MAGIC_1 = 0xA1,
-    USRR_MAGI_SIZE = 2,
 };
 
-enum class CalAddr {
-    SERIAL_NUMBER = 0X6000,
-    DEVICE_TYPE = 0X6012,
-    COLOR_EXIST = 0X601B,
-    FACT_LEFT_DATA = 0X603d,
-    FACT_RIGHT_DATA = 0X6046,
-    COLOR_DATA = 0X6050,
-    FACT_IMU_DATA = 0X6020,
-    USER_LEFT_MAGIC = 0X8010,
-    USER_LEFT_DATA = 0X8012,
-    USER_RIGHT_MAGIC = 0X801B,
-    USER_RIGHT_DATA = 0X801D,
-    USER_IMU_MAGIC = 0X8026,
-    USER_IMU_DATA = 0X8028,
+enum class SpiAddress : u16 {
+    MAGIC = 0x0000,
+    MAC_ADDRESS = 0x0015,
+    PAIRING_INFO = 0x2000,
+    SHIPMENT = 0x5000,
+    SERIAL_NUMBER = 0x6000,
+    DEVICE_TYPE = 0x6012,
+    FORMAT_VERSION = 0x601B,
+    FACT_IMU_DATA = 0x6020,
+    FACT_LEFT_DATA = 0x603d,
+    FACT_RIGHT_DATA = 0x6046,
+    COLOR_DATA = 0x6050,
+    DESIGN_VARIATION = 0x605C,
+    SENSOR_DATA = 0x6080,
+    USER_LEFT_MAGIC = 0x8010,
+    USER_LEFT_DATA = 0x8012,
+    USER_RIGHT_MAGIC = 0x801B,
+    USER_RIGHT_DATA = 0x801D,
+    USER_IMU_MAGIC = 0x8026,
+    USER_IMU_DATA = 0x8028,
 };
 
 enum class ReportMode : u8 {
@@ -186,10 +188,12 @@ enum class ReportMode : u8 {
     ACTIVE_POLLING_NFC_IR_CAMERA_CONFIGURATION = 0x01,
     ACTIVE_POLLING_NFC_IR_CAMERA_DATA_CONFIGURATION = 0x02,
     ACTIVE_POLLING_IR_CAMERA_DATA = 0x03,
+    SUBCMD_REPLY = 0x21,
     MCU_UPDATE_STATE = 0x23,
     STANDARD_FULL_60HZ = 0x30,
     NFC_IR_MODE_60HZ = 0x31,
     SIMPLE_HID_MODE = 0x3F,
+    INPUT_USB_RESPONSE = 0x81,
 };
 
 enum class GyroSensitivity : u8 {
@@ -360,10 +364,16 @@ enum class IrRegistersAddress : u16 {
     DenoiseColor = 0x6901,
 };
 
+enum class ExternalDeviceId : u16 {
+    RingController = 0x2000,
+    Starlink = 0x2800,
+};
+
 enum class DriverResult {
     Success,
     WrongReply,
     Timeout,
+    InvalidParameters,
     UnsupportedControllerType,
     HandleInUse,
     ErrorReadingData,
@@ -396,10 +406,35 @@ struct MotionData {
     u64 delta_timestamp{};
 };
 
+// Output from SPI read command containing user calibration magic
+struct MagicSpiCalibration {
+    CalibrationMagic first;
+    CalibrationMagic second;
+};
+static_assert(sizeof(MagicSpiCalibration) == 0x2, "MagicSpiCalibration is an invalid size");
+
+// Output from SPI read command containing left joystick calibration
+struct JoystickLeftSpiCalibration {
+    std::array<u8, 3> max;
+    std::array<u8, 3> center;
+    std::array<u8, 3> min;
+};
+static_assert(sizeof(JoystickLeftSpiCalibration) == 0x9,
+              "JoystickLeftSpiCalibration is an invalid size");
+
+// Output from SPI read command containing right joystick calibration
+struct JoystickRightSpiCalibration {
+    std::array<u8, 3> center;
+    std::array<u8, 3> min;
+    std::array<u8, 3> max;
+};
+static_assert(sizeof(JoystickRightSpiCalibration) == 0x9,
+              "JoystickRightSpiCalibration is an invalid size");
+
 struct JoyStickAxisCalibration {
-    u16 max{1};
-    u16 min{1};
-    u16 center{0};
+    u16 max;
+    u16 min;
+    u16 center;
 };
 
 struct JoyStickCalibration {
@@ -407,6 +442,14 @@ struct JoyStickCalibration {
     JoyStickAxisCalibration y;
 };
 
+struct ImuSpiCalibration {
+    std::array<s16, 3> accelerometer_offset;
+    std::array<s16, 3> accelerometer_scale;
+    std::array<s16, 3> gyroscope_offset;
+    std::array<s16, 3> gyroscope_scale;
+};
+static_assert(sizeof(ImuSpiCalibration) == 0x18, "ImuSpiCalibration is an invalid size");
+
 struct RingCalibration {
     s16 default_value;
     s16 max_value;
@@ -453,7 +496,7 @@ static_assert(sizeof(MCUConfig) == 0x26, "MCUConfig is an invalid size");
 
 #pragma pack(push, 1)
 struct InputReportPassive {
-    InputReport report_mode;
+    ReportMode report_mode;
     u16 button_input;
     u8 stick_state;
     std::array<u8, 10> unknown_data;
@@ -461,7 +504,7 @@ struct InputReportPassive {
 static_assert(sizeof(InputReportPassive) == 0xE, "InputReportPassive is an invalid size");
 
 struct InputReportActive {
-    InputReport report_mode;
+    ReportMode report_mode;
     u8 packet_id;
     Battery battery_status;
     std::array<u8, 3> button_input;
@@ -475,7 +518,7 @@ struct InputReportActive {
 static_assert(sizeof(InputReportActive) == 0x29, "InputReportActive is an invalid size");
 
 struct InputReportNfcIr {
-    InputReport report_mode;
+    ReportMode report_mode;
     u8 packet_id;
     Battery battery_status;
     std::array<u8, 3> button_input;
@@ -488,14 +531,6 @@ struct InputReportNfcIr {
 static_assert(sizeof(InputReportNfcIr) == 0x29, "InputReportNfcIr is an invalid size");
 #pragma pack(pop)
 
-struct IMUCalibration {
-    std::array<s16, 3> accelerometer_offset;
-    std::array<s16, 3> accelerometer_scale;
-    std::array<s16, 3> gyroscope_offset;
-    std::array<s16, 3> gyroscope_scale;
-};
-static_assert(sizeof(IMUCalibration) == 0x18, "IMUCalibration is an invalid size");
-
 struct NFCReadBlock {
     u8 start;
     u8 end;
@@ -581,9 +616,11 @@ static_assert(sizeof(FirmwareVersion) == 0x2, "FirmwareVersion is an invalid siz
 
 struct DeviceInfo {
     FirmwareVersion firmware;
+    std::array<u8, 2> unknown_1;
     MacAddress mac_address;
+    std::array<u8, 2> unknown_2;
 };
-static_assert(sizeof(DeviceInfo) == 0x8, "DeviceInfo is an invalid size");
+static_assert(sizeof(DeviceInfo) == 0xC, "DeviceInfo is an invalid size");
 
 struct MotionStatus {
     bool is_enabled;
@@ -599,6 +636,53 @@ struct RingStatus {
     s16 min_value;
 };
 
+struct VibrationPacket {
+    OutputReport output_report;
+    u8 packet_counter;
+    std::array<u8, 0x8> vibration_data;
+};
+static_assert(sizeof(VibrationPacket) == 0xA, "VibrationPacket is an invalid size");
+
+struct SubCommandPacket {
+    OutputReport output_report;
+    u8 packet_counter;
+    INSERT_PADDING_BYTES(0x8); // This contains vibration data
+    SubCommand sub_command;
+    std::array<u8, 0x26> command_data;
+};
+static_assert(sizeof(SubCommandPacket) == 0x31, "SubCommandPacket is an invalid size");
+
+#pragma pack(push, 1)
+struct ReadSpiPacket {
+    SpiAddress spi_address;
+    INSERT_PADDING_BYTES(0x2);
+    u8 size;
+};
+static_assert(sizeof(ReadSpiPacket) == 0x5, "ReadSpiPacket is an invalid size");
+
+struct SubCommandResponse {
+    InputReportPassive input_report;
+    SubCommand sub_command;
+    union {
+        std::array<u8, 0x30> command_data;
+        SpiAddress spi_address;              // Reply from SPI_FLASH_READ subcommand
+        ExternalDeviceId external_device_id; // Reply from GET_EXTERNAL_DEVICE_INFO subcommand
+        DeviceInfo device_info;              // Reply from REQ_DEV_INFO subcommand
+    };
+    u8 crc; // This is never used
+};
+static_assert(sizeof(SubCommandResponse) == 0x40, "SubCommandResponse is an invalid size");
+#pragma pack(pop)
+
+struct MCUCommandResponse {
+    InputReportNfcIr input_report;
+    INSERT_PADDING_BYTES(0x8);
+    MCUReport mcu_report;
+    std::array<u8, 0x13D> mcu_data;
+    u8 crc;
+};
+static_assert(sizeof(MCUCommandResponse) == 0x170, "MCUCommandResponse is an invalid size");
+
 struct JoyconCallbacks {
     std::function<void(Battery)> on_battery_data;
     std::function<void(Color)> on_color_data;
diff --git a/src/input_common/helpers/joycon_protocol/nfc.cpp b/src/input_common/helpers/joycon_protocol/nfc.cpp
index 5c0f71722..eeba82986 100644
--- a/src/input_common/helpers/joycon_protocol/nfc.cpp
+++ b/src/input_common/helpers/joycon_protocol/nfc.cpp
@@ -110,7 +110,7 @@ bool NfcProtocol::HasAmiibo() {
 
 DriverResult NfcProtocol::WaitUntilNfcIsReady() {
     constexpr std::size_t timeout_limit = 10;
-    std::vector<u8> output;
+    MCUCommandResponse output{};
     std::size_t tries = 0;
 
     do {
@@ -122,8 +122,9 @@ DriverResult NfcProtocol::WaitUntilNfcIsReady() {
         if (tries++ > timeout_limit) {
             return DriverResult::Timeout;
         }
-    } while (output[49] != 0x2a || (output[51] << 8) + output[50] != 0x0500 || output[55] != 0x31 ||
-             output[56] != 0x00);
+    } while (output.mcu_report != MCUReport::NFCState ||
+             (output.mcu_data[1] << 8) + output.mcu_data[0] != 0x0500 ||
+             output.mcu_data[5] != 0x31 || output.mcu_data[6] != 0x00);
 
     return DriverResult::Success;
 }
@@ -131,7 +132,7 @@ DriverResult NfcProtocol::WaitUntilNfcIsReady() {
 DriverResult NfcProtocol::StartPolling(TagFoundData& data) {
     LOG_DEBUG(Input, "Start Polling for tag");
     constexpr std::size_t timeout_limit = 7;
-    std::vector<u8> output;
+    MCUCommandResponse output{};
     std::size_t tries = 0;
 
     do {
@@ -142,18 +143,20 @@ DriverResult NfcProtocol::StartPolling(TagFoundData& data) {
         if (tries++ > timeout_limit) {
             return DriverResult::Timeout;
         }
-    } while (output[49] != 0x2a || (output[51] << 8) + output[50] != 0x0500 || output[56] != 0x09);
+    } while (output.mcu_report != MCUReport::NFCState ||
+             (output.mcu_data[1] << 8) + output.mcu_data[0] != 0x0500 ||
+             output.mcu_data[6] != 0x09);
 
-    data.type = output[62];
-    data.uuid.resize(output[64]);
-    memcpy(data.uuid.data(), output.data() + 65, data.uuid.size());
+    data.type = output.mcu_data[12];
+    data.uuid.resize(output.mcu_data[14]);
+    memcpy(data.uuid.data(), output.mcu_data.data() + 15, data.uuid.size());
 
     return DriverResult::Success;
 }
 
 DriverResult NfcProtocol::ReadTag(const TagFoundData& data) {
     constexpr std::size_t timeout_limit = 10;
-    std::vector<u8> output;
+    MCUCommandResponse output{};
     std::size_t tries = 0;
 
     std::string uuid_string;
@@ -168,23 +171,24 @@ DriverResult NfcProtocol::ReadTag(const TagFoundData& data) {
     // Read Tag data
     while (true) {
         auto result = SendReadAmiiboRequest(output, ntag_pages);
-        const auto mcu_report = static_cast<MCUReport>(output[49]);
-        const auto nfc_status = static_cast<NFCStatus>(output[56]);
+        const auto nfc_status = static_cast<NFCStatus>(output.mcu_data[6]);
 
         if (result != DriverResult::Success) {
             return result;
         }
 
-        if ((mcu_report == MCUReport::NFCReadData || mcu_report == MCUReport::NFCState) &&
+        if ((output.mcu_report == MCUReport::NFCReadData ||
+             output.mcu_report == MCUReport::NFCState) &&
             nfc_status == NFCStatus::TagLost) {
             return DriverResult::ErrorReadingData;
         }
 
-        if (mcu_report == MCUReport::NFCReadData && output[51] == 0x07 && output[52] == 0x01) {
+        if (output.mcu_report == MCUReport::NFCReadData && output.mcu_data[1] == 0x07 &&
+            output.mcu_data[2] == 0x01) {
             if (data.type != 2) {
                 continue;
             }
-            switch (output[74]) {
+            switch (output.mcu_data[24]) {
             case 0:
                 ntag_pages = NFCPages::Block135;
                 break;
@@ -200,14 +204,14 @@ DriverResult NfcProtocol::ReadTag(const TagFoundData& data) {
             continue;
         }
 
-        if (mcu_report == MCUReport::NFCState && nfc_status == NFCStatus::LastPackage) {
+        if (output.mcu_report == MCUReport::NFCState && nfc_status == NFCStatus::LastPackage) {
             // finished
             SendStopPollingRequest(output);
             return DriverResult::Success;
         }
 
         // Ignore other state reports
-        if (mcu_report == MCUReport::NFCState) {
+        if (output.mcu_report == MCUReport::NFCState) {
             continue;
         }
 
@@ -221,7 +225,7 @@ DriverResult NfcProtocol::ReadTag(const TagFoundData& data) {
 
 DriverResult NfcProtocol::GetAmiiboData(std::vector<u8>& ntag_data) {
     constexpr std::size_t timeout_limit = 10;
-    std::vector<u8> output;
+    MCUCommandResponse output{};
     std::size_t tries = 0;
 
     NFCPages ntag_pages = NFCPages::Block135;
@@ -229,36 +233,38 @@ DriverResult NfcProtocol::GetAmiiboData(std::vector<u8>& ntag_data) {
     // Read Tag data
     while (true) {
         auto result = SendReadAmiiboRequest(output, ntag_pages);
-        const auto mcu_report = static_cast<MCUReport>(output[49]);
-        const auto nfc_status = static_cast<NFCStatus>(output[56]);
+        const auto nfc_status = static_cast<NFCStatus>(output.mcu_data[6]);
 
         if (result != DriverResult::Success) {
             return result;
         }
 
-        if ((mcu_report == MCUReport::NFCReadData || mcu_report == MCUReport::NFCState) &&
+        if ((output.mcu_report == MCUReport::NFCReadData ||
+             output.mcu_report == MCUReport::NFCState) &&
             nfc_status == NFCStatus::TagLost) {
             return DriverResult::ErrorReadingData;
         }
 
-        if (mcu_report == MCUReport::NFCReadData && output[51] == 0x07) {
-            std::size_t payload_size = (output[54] << 8 | output[55]) & 0x7FF;
-            if (output[52] == 0x01) {
-                memcpy(ntag_data.data() + ntag_buffer_pos, output.data() + 116, payload_size - 60);
+        if (output.mcu_report == MCUReport::NFCReadData && output.mcu_data[1] == 0x07) {
+            std::size_t payload_size = (output.mcu_data[4] << 8 | output.mcu_data[5]) & 0x7FF;
+            if (output.mcu_data[2] == 0x01) {
+                memcpy(ntag_data.data() + ntag_buffer_pos, output.mcu_data.data() + 66,
+                       payload_size - 60);
                 ntag_buffer_pos += payload_size - 60;
             } else {
-                memcpy(ntag_data.data() + ntag_buffer_pos, output.data() + 56, payload_size);
+                memcpy(ntag_data.data() + ntag_buffer_pos, output.mcu_data.data() + 6,
+                       payload_size);
             }
             continue;
         }
 
-        if (mcu_report == MCUReport::NFCState && nfc_status == NFCStatus::LastPackage) {
+        if (output.mcu_report == MCUReport::NFCState && nfc_status == NFCStatus::LastPackage) {
             LOG_INFO(Input, "Finished reading amiibo");
             return DriverResult::Success;
         }
 
         // Ignore other state reports
-        if (mcu_report == MCUReport::NFCState) {
+        if (output.mcu_report == MCUReport::NFCState) {
             continue;
         }
 
@@ -270,7 +276,7 @@ DriverResult NfcProtocol::GetAmiiboData(std::vector<u8>& ntag_data) {
     return DriverResult::Success;
 }
 
-DriverResult NfcProtocol::SendStartPollingRequest(std::vector<u8>& output) {
+DriverResult NfcProtocol::SendStartPollingRequest(MCUCommandResponse& output) {
     NFCRequestState request{
         .sub_command = MCUSubCommand::ReadDeviceMode,
         .command_argument = NFCReadCommand::StartPolling,
@@ -294,7 +300,7 @@ DriverResult NfcProtocol::SendStartPollingRequest(std::vector<u8>& output) {
     return SendMCUData(ReportMode::NFC_IR_MODE_60HZ, SubCommand::STATE, request_data, output);
 }
 
-DriverResult NfcProtocol::SendStopPollingRequest(std::vector<u8>& output) {
+DriverResult NfcProtocol::SendStopPollingRequest(MCUCommandResponse& output) {
     NFCRequestState request{
         .sub_command = MCUSubCommand::ReadDeviceMode,
         .command_argument = NFCReadCommand::StopPolling,
@@ -311,7 +317,7 @@ DriverResult NfcProtocol::SendStopPollingRequest(std::vector<u8>& output) {
     return SendMCUData(ReportMode::NFC_IR_MODE_60HZ, SubCommand::STATE, request_data, output);
 }
 
-DriverResult NfcProtocol::SendStartWaitingRecieveRequest(std::vector<u8>& output) {
+DriverResult NfcProtocol::SendStartWaitingRecieveRequest(MCUCommandResponse& output) {
     NFCRequestState request{
         .sub_command = MCUSubCommand::ReadDeviceMode,
         .command_argument = NFCReadCommand::StartWaitingRecieve,
@@ -328,7 +334,7 @@ DriverResult NfcProtocol::SendStartWaitingRecieveRequest(std::vector<u8>& output
     return SendMCUData(ReportMode::NFC_IR_MODE_60HZ, SubCommand::STATE, request_data, output);
 }
 
-DriverResult NfcProtocol::SendReadAmiiboRequest(std::vector<u8>& output, NFCPages ntag_pages) {
+DriverResult NfcProtocol::SendReadAmiiboRequest(MCUCommandResponse& output, NFCPages ntag_pages) {
     NFCRequestState request{
         .sub_command = MCUSubCommand::ReadDeviceMode,
         .command_argument = NFCReadCommand::Ntag,
diff --git a/src/input_common/helpers/joycon_protocol/nfc.h b/src/input_common/helpers/joycon_protocol/nfc.h
index e63665aa9..11e263e07 100644
--- a/src/input_common/helpers/joycon_protocol/nfc.h
+++ b/src/input_common/helpers/joycon_protocol/nfc.h
@@ -45,13 +45,13 @@ private:
 
     DriverResult GetAmiiboData(std::vector<u8>& data);
 
-    DriverResult SendStartPollingRequest(std::vector<u8>& output);
+    DriverResult SendStartPollingRequest(MCUCommandResponse& output);
 
-    DriverResult SendStopPollingRequest(std::vector<u8>& output);
+    DriverResult SendStopPollingRequest(MCUCommandResponse& output);
 
-    DriverResult SendStartWaitingRecieveRequest(std::vector<u8>& output);
+    DriverResult SendStartWaitingRecieveRequest(MCUCommandResponse& output);
 
-    DriverResult SendReadAmiiboRequest(std::vector<u8>& output, NFCPages ntag_pages);
+    DriverResult SendReadAmiiboRequest(MCUCommandResponse& output, NFCPages ntag_pages);
 
     NFCReadBlockCommand GetReadBlockCommand(NFCPages pages) const;
 
diff --git a/src/input_common/helpers/joycon_protocol/poller.cpp b/src/input_common/helpers/joycon_protocol/poller.cpp
index 7f8e093fa..9bb15e935 100644
--- a/src/input_common/helpers/joycon_protocol/poller.cpp
+++ b/src/input_common/helpers/joycon_protocol/poller.cpp
@@ -31,9 +31,7 @@ void JoyconPoller::ReadActiveMode(std::span<u8> buffer, const MotionStatus& moti
     case Joycon::ControllerType::Pro:
         UpdateActiveProPadInput(data, motion_status);
         break;
-    case Joycon::ControllerType::Grip:
-    case Joycon::ControllerType::Dual:
-    case Joycon::ControllerType::None:
+    default:
         break;
     }
 
@@ -58,9 +56,7 @@ void JoyconPoller::ReadPassiveMode(std::span<u8> buffer) {
     case Joycon::ControllerType::Pro:
         UpdatePasiveProPadInput(data);
         break;
-    case Joycon::ControllerType::Grip:
-    case Joycon::ControllerType::Dual:
-    case Joycon::ControllerType::None:
+    default:
         break;
     }
 }
diff --git a/src/input_common/helpers/joycon_protocol/ringcon.cpp b/src/input_common/helpers/joycon_protocol/ringcon.cpp
index 12f81309e..190cef812 100644
--- a/src/input_common/helpers/joycon_protocol/ringcon.cpp
+++ b/src/input_common/helpers/joycon_protocol/ringcon.cpp
@@ -70,14 +70,12 @@ DriverResult RingConProtocol::StartRingconPolling() {
 DriverResult RingConProtocol::IsRingConnected(bool& is_connected) {
     LOG_DEBUG(Input, "IsRingConnected");
     constexpr std::size_t max_tries = 28;
-    constexpr u8 ring_controller_id = 0x20;
-    std::vector<u8> output;
+    SubCommandResponse output{};
     std::size_t tries = 0;
     is_connected = false;
 
     do {
-        std::array<u8, 1> empty_data{};
-        const auto result = SendSubCommand(SubCommand::UNKNOWN_RINGCON, empty_data, output);
+        const auto result = SendSubCommand(SubCommand::GET_EXTERNAL_DEVICE_INFO, {}, output);
 
         if (result != DriverResult::Success) {
             return result;
@@ -86,7 +84,7 @@ DriverResult RingConProtocol::IsRingConnected(bool& is_connected) {
         if (tries++ >= max_tries) {
             return DriverResult::NoDeviceDetected;
         }
-    } while (output[16] != ring_controller_id);
+    } while (output.external_device_id != ExternalDeviceId::RingController);
 
     is_connected = true;
     return DriverResult::Success;
@@ -100,14 +98,14 @@ DriverResult RingConProtocol::ConfigureRing() {
         0x00, 0x00, 0x00, 0x0A, 0x64, 0x0B, 0xE6, 0xA9, 0x22, 0x00, 0x00, 0x04, 0x00,
         0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x90, 0xA8, 0xE1, 0x34, 0x36};
 
-    const DriverResult result = SendSubCommand(SubCommand::UNKNOWN_RINGCON3, ring_config);
+    const DriverResult result = SendSubCommand(SubCommand::SET_EXTERNAL_FORMAT_CONFIG, ring_config);
 
     if (result != DriverResult::Success) {
         return result;
     }
 
     static constexpr std::array<u8, 4> ringcon_data{0x04, 0x01, 0x01, 0x02};
-    return SendSubCommand(SubCommand::UNKNOWN_RINGCON2, ringcon_data);
+    return SendSubCommand(SubCommand::ENABLE_EXTERNAL_POLLING, ringcon_data);
 }
 
 bool RingConProtocol::IsEnabled() const {
diff --git a/src/input_common/helpers/stick_from_buttons.cpp b/src/input_common/helpers/stick_from_buttons.cpp
index 096c23b07..a6be6dac1 100644
--- a/src/input_common/helpers/stick_from_buttons.cpp
+++ b/src/input_common/helpers/stick_from_buttons.cpp
@@ -15,6 +15,9 @@ public:
     // do not play nicely with the theoretical maximum range.
     // Using a value one lower from the maximum emulates real stick behavior.
     static constexpr float MAX_RANGE = 32766.0f / 32767.0f;
+    static constexpr float TAU = Common::PI * 2.0f;
+    // Use wider angle to ease the transition.
+    static constexpr float APERTURE = TAU * 0.15f;
 
     using Button = std::unique_ptr<Common::Input::InputDevice>;
 
@@ -61,30 +64,23 @@ public:
     }
 
     bool IsAngleGreater(float old_angle, float new_angle) const {
-        constexpr float TAU = Common::PI * 2.0f;
-        // Use wider angle to ease the transition.
-        constexpr float aperture = TAU * 0.15f;
-        const float top_limit = new_angle + aperture;
+        const float top_limit = new_angle + APERTURE;
         return (old_angle > new_angle && old_angle <= top_limit) ||
                (old_angle + TAU > new_angle && old_angle + TAU <= top_limit);
     }
 
     bool IsAngleSmaller(float old_angle, float new_angle) const {
-        constexpr float TAU = Common::PI * 2.0f;
-        // Use wider angle to ease the transition.
-        constexpr float aperture = TAU * 0.15f;
-        const float bottom_limit = new_angle - aperture;
+        const float bottom_limit = new_angle - APERTURE;
         return (old_angle >= bottom_limit && old_angle < new_angle) ||
                (old_angle - TAU >= bottom_limit && old_angle - TAU < new_angle);
     }
 
     float GetAngle(std::chrono::time_point<std::chrono::steady_clock> now) const {
-        constexpr float TAU = Common::PI * 2.0f;
         float new_angle = angle;
 
         auto time_difference = static_cast<float>(
-            std::chrono::duration_cast<std::chrono::microseconds>(now - last_update).count());
-        time_difference /= 1000.0f * 1000.0f;
+            std::chrono::duration_cast<std::chrono::milliseconds>(now - last_update).count());
+        time_difference /= 1000.0f;
         if (time_difference > 0.5f) {
             time_difference = 0.5f;
         }
@@ -201,8 +197,6 @@ public:
     }
 
     void UpdateStatus() {
-        const float coef = modifier_status.value ? modifier_scale : MAX_RANGE;
-
         bool r = right_status;
         bool l = left_status;
         bool u = up_status;
@@ -220,7 +214,7 @@ public:
 
         // Move if a key is pressed
         if (r || l || u || d) {
-            amplitude = coef;
+            amplitude = modifier_status.value ? modifier_scale : MAX_RANGE;
         } else {
             amplitude = 0;
         }
@@ -274,30 +268,17 @@ public:
         Common::Input::StickStatus status{};
         status.x.properties = properties;
         status.y.properties = properties;
+
         if (Settings::values.emulate_analog_keyboard) {
             const auto now = std::chrono::steady_clock::now();
-            float angle_ = GetAngle(now);
+            const float angle_ = GetAngle(now);
             status.x.raw_value = std::cos(angle_) * amplitude;
             status.y.raw_value = std::sin(angle_) * amplitude;
             return status;
         }
-        constexpr float SQRT_HALF = 0.707106781f;
-        int x = 0, y = 0;
-        if (right_status) {
-            ++x;
-        }
-        if (left_status) {
-            --x;
-        }
-        if (up_status) {
-            ++y;
-        }
-        if (down_status) {
-            --y;
-        }
-        const float coef = modifier_status.value ? modifier_scale : MAX_RANGE;
-        status.x.raw_value = static_cast<float>(x) * coef * (y == 0 ? 1.0f : SQRT_HALF);
-        status.y.raw_value = static_cast<float>(y) * coef * (x == 0 ? 1.0f : SQRT_HALF);
+
+        status.x.raw_value = std::cos(goal_angle) * amplitude;
+        status.y.raw_value = std::sin(goal_angle) * amplitude;
         return status;
     }
 
diff --git a/src/input_common/input_poller.cpp b/src/input_common/input_poller.cpp
index 15cbf7e5f..8c6a6521a 100644
--- a/src/input_common/input_poller.cpp
+++ b/src/input_common/input_poller.cpp
@@ -16,10 +16,10 @@ public:
 
 class InputFromButton final : public Common::Input::InputDevice {
 public:
-    explicit InputFromButton(PadIdentifier identifier_, int button_, bool toggle_, bool inverted_,
-                             InputEngine* input_engine_)
-        : identifier(identifier_), button(button_), toggle(toggle_), inverted(inverted_),
-          input_engine(input_engine_) {
+    explicit InputFromButton(PadIdentifier identifier_, int button_, bool turbo_, bool toggle_,
+                             bool inverted_, InputEngine* input_engine_)
+        : identifier(identifier_), button(button_), turbo(turbo_), toggle(toggle_),
+          inverted(inverted_), input_engine(input_engine_) {
         UpdateCallback engine_callback{[this]() { OnChange(); }};
         const InputIdentifier input_identifier{
             .identifier = identifier,
@@ -40,6 +40,7 @@ public:
             .value = input_engine->GetButton(identifier, button),
             .inverted = inverted,
             .toggle = toggle,
+            .turbo = turbo,
         };
     }
 
@@ -68,6 +69,7 @@ public:
 private:
     const PadIdentifier identifier;
     const int button;
+    const bool turbo;
     const bool toggle;
     const bool inverted;
     int callback_key;
@@ -77,10 +79,10 @@ private:
 
 class InputFromHatButton final : public Common::Input::InputDevice {
 public:
-    explicit InputFromHatButton(PadIdentifier identifier_, int button_, u8 direction_, bool toggle_,
-                                bool inverted_, InputEngine* input_engine_)
-        : identifier(identifier_), button(button_), direction(direction_), toggle(toggle_),
-          inverted(inverted_), input_engine(input_engine_) {
+    explicit InputFromHatButton(PadIdentifier identifier_, int button_, u8 direction_, bool turbo_,
+                                bool toggle_, bool inverted_, InputEngine* input_engine_)
+        : identifier(identifier_), button(button_), direction(direction_), turbo(turbo_),
+          toggle(toggle_), inverted(inverted_), input_engine(input_engine_) {
         UpdateCallback engine_callback{[this]() { OnChange(); }};
         const InputIdentifier input_identifier{
             .identifier = identifier,
@@ -101,6 +103,7 @@ public:
             .value = input_engine->GetHatButton(identifier, button, direction),
             .inverted = inverted,
             .toggle = toggle,
+            .turbo = turbo,
         };
     }
 
@@ -130,6 +133,7 @@ private:
     const PadIdentifier identifier;
     const int button;
     const u8 direction;
+    const bool turbo;
     const bool toggle;
     const bool inverted;
     int callback_key;
@@ -853,14 +857,15 @@ std::unique_ptr<Common::Input::InputDevice> InputFactory::CreateButtonDevice(
     const auto keyboard_key = params.Get("code", 0);
     const auto toggle = params.Get("toggle", false) != 0;
     const auto inverted = params.Get("inverted", false) != 0;
+    const auto turbo = params.Get("turbo", false) != 0;
     input_engine->PreSetController(identifier);
     input_engine->PreSetButton(identifier, button_id);
     input_engine->PreSetButton(identifier, keyboard_key);
     if (keyboard_key != 0) {
-        return std::make_unique<InputFromButton>(identifier, keyboard_key, toggle, inverted,
+        return std::make_unique<InputFromButton>(identifier, keyboard_key, turbo, toggle, inverted,
                                                  input_engine.get());
     }
-    return std::make_unique<InputFromButton>(identifier, button_id, toggle, inverted,
+    return std::make_unique<InputFromButton>(identifier, button_id, turbo, toggle, inverted,
                                              input_engine.get());
 }
 
@@ -876,11 +881,12 @@ std::unique_ptr<Common::Input::InputDevice> InputFactory::CreateHatButtonDevice(
     const auto direction = input_engine->GetHatButtonId(params.Get("direction", ""));
     const auto toggle = params.Get("toggle", false) != 0;
     const auto inverted = params.Get("inverted", false) != 0;
+    const auto turbo = params.Get("turbo", false) != 0;
 
     input_engine->PreSetController(identifier);
     input_engine->PreSetHatButton(identifier, button_id);
-    return std::make_unique<InputFromHatButton>(identifier, button_id, direction, toggle, inverted,
-                                                input_engine.get());
+    return std::make_unique<InputFromHatButton>(identifier, button_id, direction, turbo, toggle,
+                                                inverted, input_engine.get());
 }
 
 std::unique_ptr<Common::Input::InputDevice> InputFactory::CreateStickDevice(
diff --git a/src/shader_recompiler/backend/glasm/emit_glasm.cpp b/src/shader_recompiler/backend/glasm/emit_glasm.cpp
index 0cb1e193e..fd4a61a4d 100644
--- a/src/shader_recompiler/backend/glasm/emit_glasm.cpp
+++ b/src/shader_recompiler/backend/glasm/emit_glasm.cpp
@@ -279,6 +279,8 @@ void SetupOptions(const IR::Program& program, const Profile& profile,
     header += "OPTION NV_internal;"
               "OPTION NV_shader_storage_buffer;"
               "OPTION NV_gpu_program_fp64;";
+    // TODO: Enable only when MS is used
+    header += "OPTION NV_texture_multisample;";
     if (info.uses_int64_bit_atomics) {
         header += "OPTION NV_shader_atomic_int64;";
     }
diff --git a/src/shader_recompiler/backend/glasm/emit_glasm_image.cpp b/src/shader_recompiler/backend/glasm/emit_glasm_image.cpp
index e67e80fac..85ee27333 100644
--- a/src/shader_recompiler/backend/glasm/emit_glasm_image.cpp
+++ b/src/shader_recompiler/backend/glasm/emit_glasm_image.cpp
@@ -59,7 +59,14 @@ std::string Image(EmitContext& ctx, IR::TextureInstInfo info,
     }
 }
 
-std::string_view TextureType(IR::TextureInstInfo info) {
+bool IsTextureMsaa(EmitContext& ctx, const IR::TextureInstInfo& info) {
+    if (info.type == TextureType::Buffer) {
+        return false;
+    }
+    return ctx.info.texture_descriptors.at(info.descriptor_index).is_multisample;
+}
+
+std::string_view TextureType(IR::TextureInstInfo info, bool is_ms = false) {
     if (info.is_depth) {
         switch (info.type) {
         case TextureType::Color1D:
@@ -88,9 +95,9 @@ std::string_view TextureType(IR::TextureInstInfo info) {
             return "ARRAY1D";
         case TextureType::Color2D:
         case TextureType::Color2DRect:
-            return "2D";
+            return is_ms ? "2DMS" : "2D";
         case TextureType::ColorArray2D:
-            return "ARRAY2D";
+            return is_ms ? "ARRAY2DMS" : "ARRAY2D";
         case TextureType::Color3D:
             return "3D";
         case TextureType::ColorCube:
@@ -510,15 +517,16 @@ void EmitImageFetch(EmitContext& ctx, IR::Inst& inst, const IR::Value& index,
                     const IR::Value& coord, const IR::Value& offset, ScalarS32 lod, ScalarS32 ms) {
     const auto info{inst.Flags<IR::TextureInstInfo>()};
     const auto sparse_inst{PrepareSparse(inst)};
+    const bool is_multisample{ms.type != Type::Void};
     const std::string_view sparse_mod{sparse_inst ? ".SPARSE" : ""};
-    const std::string_view type{TextureType(info)};
+    const std::string_view type{TextureType(info, is_multisample)};
     const std::string texture{Texture(ctx, info, index)};
     const std::string offset_vec{Offset(ctx, offset)};
     const auto [coord_vec, coord_alloc]{Coord(ctx, coord)};
     const Register ret{ctx.reg_alloc.Define(inst)};
     if (info.type == TextureType::Buffer) {
         ctx.Add("TXF.F{} {},{},{},{}{};", sparse_mod, ret, coord_vec, texture, type, offset_vec);
-    } else if (ms.type != Type::Void) {
+    } else if (is_multisample) {
         ctx.Add("MOV.S {}.w,{};"
                 "TXFMS.F{} {},{},{},{}{};",
                 coord_vec, ms, sparse_mod, ret, coord_vec, texture, type, offset_vec);
@@ -531,10 +539,11 @@ void EmitImageFetch(EmitContext& ctx, IR::Inst& inst, const IR::Value& index,
 }
 
 void EmitImageQueryDimensions(EmitContext& ctx, IR::Inst& inst, const IR::Value& index,
-                              ScalarS32 lod) {
+                              ScalarS32 lod, [[maybe_unused]] const IR::Value& skip_mips) {
     const auto info{inst.Flags<IR::TextureInstInfo>()};
     const std::string texture{Texture(ctx, info, index)};
-    const std::string_view type{TextureType(info)};
+    const bool is_msaa{IsTextureMsaa(ctx, info)};
+    const std::string_view type{TextureType(info, is_msaa)};
     ctx.Add("TXQ {},{},{},{};", inst, lod, texture, type);
 }
 
diff --git a/src/shader_recompiler/backend/glasm/emit_glasm_instructions.h b/src/shader_recompiler/backend/glasm/emit_glasm_instructions.h
index 415a249e4..1a1ea61d5 100644
--- a/src/shader_recompiler/backend/glasm/emit_glasm_instructions.h
+++ b/src/shader_recompiler/backend/glasm/emit_glasm_instructions.h
@@ -581,7 +581,7 @@ void EmitImageGatherDref(EmitContext& ctx, IR::Inst& inst, const IR::Value& inde
 void EmitImageFetch(EmitContext& ctx, IR::Inst& inst, const IR::Value& index,
                     const IR::Value& coord, const IR::Value& offset, ScalarS32 lod, ScalarS32 ms);
 void EmitImageQueryDimensions(EmitContext& ctx, IR::Inst& inst, const IR::Value& index,
-                              ScalarS32 lod);
+                              ScalarS32 lod, const IR::Value& skip_mips);
 void EmitImageQueryLod(EmitContext& ctx, IR::Inst& inst, const IR::Value& index, Register coord);
 void EmitImageGradient(EmitContext& ctx, IR::Inst& inst, const IR::Value& index,
                        const IR::Value& coord, const IR::Value& derivatives,
diff --git a/src/shader_recompiler/backend/glsl/emit_glsl_image.cpp b/src/shader_recompiler/backend/glsl/emit_glsl_image.cpp
index cecdbb9d6..f335c8af0 100644
--- a/src/shader_recompiler/backend/glsl/emit_glsl_image.cpp
+++ b/src/shader_recompiler/backend/glsl/emit_glsl_image.cpp
@@ -25,6 +25,13 @@ std::string Image(EmitContext& ctx, const IR::TextureInstInfo& info, const IR::V
     return fmt::format("img{}{}", def.binding, index_offset);
 }
 
+bool IsTextureMsaa(EmitContext& ctx, const IR::TextureInstInfo& info) {
+    if (info.type == TextureType::Buffer) {
+        return false;
+    }
+    return ctx.info.texture_descriptors.at(info.descriptor_index).is_multisample;
+}
+
 std::string CastToIntVec(std::string_view value, const IR::TextureInstInfo& info) {
     switch (info.type) {
     case TextureType::Color1D:
@@ -414,7 +421,7 @@ void EmitImageGatherDref(EmitContext& ctx, IR::Inst& inst, const IR::Value& inde
 
 void EmitImageFetch(EmitContext& ctx, IR::Inst& inst, const IR::Value& index,
                     std::string_view coords, std::string_view offset, std::string_view lod,
-                    [[maybe_unused]] std::string_view ms) {
+                    std::string_view ms) {
     const auto info{inst.Flags<IR::TextureInstInfo>()};
     if (info.has_bias) {
         throw NotImplementedException("EmitImageFetch Bias texture samples");
@@ -431,19 +438,24 @@ void EmitImageFetch(EmitContext& ctx, IR::Inst& inst, const IR::Value& index,
         ctx.AddU1("{}=true;", *sparse_inst);
     }
     if (!sparse_inst || !supports_sparse) {
-        if (!offset.empty()) {
-            ctx.Add("{}=texelFetchOffset({},{},int({}),{});", texel, texture,
-                    CoordsCastToInt(coords, info), lod, CoordsCastToInt(offset, info));
+        const auto int_coords{CoordsCastToInt(coords, info)};
+        if (!ms.empty()) {
+            ctx.Add("{}=texelFetch({},{},int({}));", texel, texture, int_coords, ms);
+        } else if (!offset.empty()) {
+            ctx.Add("{}=texelFetchOffset({},{},int({}),{});", texel, texture, int_coords, lod,
+                    CoordsCastToInt(offset, info));
         } else {
             if (info.type == TextureType::Buffer) {
                 ctx.Add("{}=texelFetch({},int({}));", texel, texture, coords);
             } else {
-                ctx.Add("{}=texelFetch({},{},int({}));", texel, texture,
-                        CoordsCastToInt(coords, info), lod);
+                ctx.Add("{}=texelFetch({},{},int({}));", texel, texture, int_coords, lod);
             }
         }
         return;
     }
+    if (!ms.empty()) {
+        throw NotImplementedException("EmitImageFetch Sparse MSAA samples");
+    }
     if (!offset.empty()) {
         ctx.AddU1("{}=sparseTexelsResidentARB(sparseTexelFetchOffsetARB({},{},int({}),{},{}));",
                   *sparse_inst, texture, CastToIntVec(coords, info), lod,
@@ -455,29 +467,36 @@ void EmitImageFetch(EmitContext& ctx, IR::Inst& inst, const IR::Value& index,
 }
 
 void EmitImageQueryDimensions(EmitContext& ctx, IR::Inst& inst, const IR::Value& index,
-                              std::string_view lod) {
+                              std::string_view lod, const IR::Value& skip_mips_val) {
     const auto info{inst.Flags<IR::TextureInstInfo>()};
     const auto texture{Texture(ctx, info, index)};
+    const bool is_msaa{IsTextureMsaa(ctx, info)};
+    const bool skip_mips{skip_mips_val.U1()};
+    const auto mips{skip_mips ? "0u" : fmt::format("uint(textureQueryLevels({}))", texture)};
+    if (is_msaa && !skip_mips) {
+        throw NotImplementedException("EmitImageQueryDimensions MSAA QueryLevels");
+    }
+    if (info.type == TextureType::Buffer && !skip_mips) {
+        throw NotImplementedException("EmitImageQueryDimensions TextureType::Buffer QueryLevels");
+    }
+    const bool uses_lod{!is_msaa && info.type != TextureType::Buffer};
+    const auto lod_str{uses_lod ? fmt::format(",int({})", lod) : ""};
     switch (info.type) {
     case TextureType::Color1D:
-        return ctx.AddU32x4(
-            "{}=uvec4(uint(textureSize({},int({}))),0u,0u,uint(textureQueryLevels({})));", inst,
-            texture, lod, texture);
+        return ctx.AddU32x4("{}=uvec4(uint(textureSize({}{})),0u,0u,{});", inst, texture, lod_str,
+                            mips);
     case TextureType::ColorArray1D:
     case TextureType::Color2D:
     case TextureType::ColorCube:
     case TextureType::Color2DRect:
-        return ctx.AddU32x4(
-            "{}=uvec4(uvec2(textureSize({},int({}))),0u,uint(textureQueryLevels({})));", inst,
-            texture, lod, texture);
+        return ctx.AddU32x4("{}=uvec4(uvec2(textureSize({}{})),0u,{});", inst, texture, lod_str,
+                            mips);
     case TextureType::ColorArray2D:
     case TextureType::Color3D:
     case TextureType::ColorArrayCube:
-        return ctx.AddU32x4(
-            "{}=uvec4(uvec3(textureSize({},int({}))),uint(textureQueryLevels({})));", inst, texture,
-            lod, texture);
+        return ctx.AddU32x4("{}=uvec4(uvec3(textureSize({}{})),{});", inst, texture, lod_str, mips);
     case TextureType::Buffer:
-        throw NotImplementedException("EmitImageQueryDimensions Texture buffers");
+        return ctx.AddU32x4("{}=uvec4(uint(textureSize({})),0u,0u,{});", inst, texture, mips);
     }
     throw LogicError("Unspecified image type {}", info.type.Value());
 }
diff --git a/src/shader_recompiler/backend/glsl/emit_glsl_instructions.h b/src/shader_recompiler/backend/glsl/emit_glsl_instructions.h
index c6df1dba7..8d0a65047 100644
--- a/src/shader_recompiler/backend/glsl/emit_glsl_instructions.h
+++ b/src/shader_recompiler/backend/glsl/emit_glsl_instructions.h
@@ -654,7 +654,7 @@ void EmitImageFetch(EmitContext& ctx, IR::Inst& inst, const IR::Value& index,
                     std::string_view coords, std::string_view offset, std::string_view lod,
                     std::string_view ms);
 void EmitImageQueryDimensions(EmitContext& ctx, IR::Inst& inst, const IR::Value& index,
-                              std::string_view lod);
+                              std::string_view lod, const IR::Value& skip_mips);
 void EmitImageQueryLod(EmitContext& ctx, IR::Inst& inst, const IR::Value& index,
                        std::string_view coords);
 void EmitImageGradient(EmitContext& ctx, IR::Inst& inst, const IR::Value& index,
diff --git a/src/shader_recompiler/backend/glsl/glsl_emit_context.cpp b/src/shader_recompiler/backend/glsl/glsl_emit_context.cpp
index 5d01ec0cd..1b006e811 100644
--- a/src/shader_recompiler/backend/glsl/glsl_emit_context.cpp
+++ b/src/shader_recompiler/backend/glsl/glsl_emit_context.cpp
@@ -61,24 +61,28 @@ std::string OutputDecorator(Stage stage, u32 size) {
     }
 }
 
-std::string_view SamplerType(TextureType type, bool is_depth) {
-    if (is_depth) {
-        switch (type) {
-        case TextureType::Color1D:
-            return "sampler1DShadow";
-        case TextureType::ColorArray1D:
-            return "sampler1DArrayShadow";
-        case TextureType::Color2D:
-            return "sampler2DShadow";
-        case TextureType::ColorArray2D:
-            return "sampler2DArrayShadow";
-        case TextureType::ColorCube:
-            return "samplerCubeShadow";
-        case TextureType::ColorArrayCube:
-            return "samplerCubeArrayShadow";
-        default:
-            throw NotImplementedException("Texture type: {}", type);
-        }
+std::string_view DepthSamplerType(TextureType type) {
+    switch (type) {
+    case TextureType::Color1D:
+        return "sampler1DShadow";
+    case TextureType::ColorArray1D:
+        return "sampler1DArrayShadow";
+    case TextureType::Color2D:
+        return "sampler2DShadow";
+    case TextureType::ColorArray2D:
+        return "sampler2DArrayShadow";
+    case TextureType::ColorCube:
+        return "samplerCubeShadow";
+    case TextureType::ColorArrayCube:
+        return "samplerCubeArrayShadow";
+    default:
+        throw NotImplementedException("Texture type: {}", type);
+    }
+}
+
+std::string_view ColorSamplerType(TextureType type, bool is_multisample = false) {
+    if (is_multisample) {
+        ASSERT(type == TextureType::Color2D || type == TextureType::ColorArray2D);
     }
     switch (type) {
     case TextureType::Color1D:
@@ -87,9 +91,9 @@ std::string_view SamplerType(TextureType type, bool is_depth) {
         return "sampler1DArray";
     case TextureType::Color2D:
     case TextureType::Color2DRect:
-        return "sampler2D";
+        return is_multisample ? "sampler2DMS" : "sampler2D";
     case TextureType::ColorArray2D:
-        return "sampler2DArray";
+        return is_multisample ? "sampler2DMSArray" : "sampler2DArray";
     case TextureType::Color3D:
         return "sampler3D";
     case TextureType::ColorCube:
@@ -677,7 +681,7 @@ void EmitContext::SetupTextures(Bindings& bindings) {
     texture_buffers.reserve(info.texture_buffer_descriptors.size());
     for (const auto& desc : info.texture_buffer_descriptors) {
         texture_buffers.push_back({bindings.texture, desc.count});
-        const auto sampler_type{SamplerType(TextureType::Buffer, false)};
+        const auto sampler_type{ColorSamplerType(TextureType::Buffer)};
         const auto array_decorator{desc.count > 1 ? fmt::format("[{}]", desc.count) : ""};
         header += fmt::format("layout(binding={}) uniform {} tex{}{};", bindings.texture,
                               sampler_type, bindings.texture, array_decorator);
@@ -686,7 +690,8 @@ void EmitContext::SetupTextures(Bindings& bindings) {
     textures.reserve(info.texture_descriptors.size());
     for (const auto& desc : info.texture_descriptors) {
         textures.push_back({bindings.texture, desc.count});
-        const auto sampler_type{SamplerType(desc.type, desc.is_depth)};
+        const auto sampler_type{desc.is_depth ? DepthSamplerType(desc.type)
+                                              : ColorSamplerType(desc.type, desc.is_multisample)};
         const auto array_decorator{desc.count > 1 ? fmt::format("[{}]", desc.count) : ""};
         header += fmt::format("layout(binding={}) uniform {} tex{}{};", bindings.texture,
                               sampler_type, bindings.texture, array_decorator);
diff --git a/src/shader_recompiler/backend/spirv/emit_spirv_image.cpp b/src/shader_recompiler/backend/spirv/emit_spirv_image.cpp
index c898ce12f..02073c420 100644
--- a/src/shader_recompiler/backend/spirv/emit_spirv_image.cpp
+++ b/src/shader_recompiler/backend/spirv/emit_spirv_image.cpp
@@ -201,6 +201,13 @@ Id Image(EmitContext& ctx, const IR::Value& index, IR::TextureInstInfo info) {
     }
 }
 
+bool IsTextureMsaa(EmitContext& ctx, const IR::TextureInstInfo& info) {
+    if (info.type == TextureType::Buffer) {
+        return false;
+    }
+    return ctx.textures.at(info.descriptor_index).is_multisample;
+}
+
 Id Decorate(EmitContext& ctx, IR::Inst* inst, Id sample) {
     const auto info{inst->Flags<IR::TextureInstInfo>()};
     if (info.relaxed_precision != 0) {
@@ -445,29 +452,33 @@ Id EmitImageFetch(EmitContext& ctx, IR::Inst* inst, const IR::Value& index, Id c
                 TextureImage(ctx, info, index), coords, operands.MaskOptional(), operands.Span());
 }
 
-Id EmitImageQueryDimensions(EmitContext& ctx, IR::Inst* inst, const IR::Value& index, Id lod) {
+Id EmitImageQueryDimensions(EmitContext& ctx, IR::Inst* inst, const IR::Value& index, Id lod,
+                            const IR::Value& skip_mips_val) {
     const auto info{inst->Flags<IR::TextureInstInfo>()};
     const Id image{TextureImage(ctx, info, index)};
     const Id zero{ctx.u32_zero_value};
-    const auto mips{[&] { return ctx.OpImageQueryLevels(ctx.U32[1], image); }};
+    const bool skip_mips{skip_mips_val.U1()};
+    const auto mips{[&] { return skip_mips ? zero : ctx.OpImageQueryLevels(ctx.U32[1], image); }};
+    const bool is_msaa{IsTextureMsaa(ctx, info)};
+    const bool uses_lod{!is_msaa && info.type != TextureType::Buffer};
+    const auto query{[&](Id type) {
+        return uses_lod ? ctx.OpImageQuerySizeLod(type, image, lod)
+                        : ctx.OpImageQuerySize(type, image);
+    }};
     switch (info.type) {
     case TextureType::Color1D:
-        return ctx.OpCompositeConstruct(ctx.U32[4], ctx.OpImageQuerySizeLod(ctx.U32[1], image, lod),
-                                        zero, zero, mips());
+        return ctx.OpCompositeConstruct(ctx.U32[4], query(ctx.U32[1]), zero, zero, mips());
     case TextureType::ColorArray1D:
     case TextureType::Color2D:
     case TextureType::ColorCube:
     case TextureType::Color2DRect:
-        return ctx.OpCompositeConstruct(ctx.U32[4], ctx.OpImageQuerySizeLod(ctx.U32[2], image, lod),
-                                        zero, mips());
+        return ctx.OpCompositeConstruct(ctx.U32[4], query(ctx.U32[2]), zero, mips());
     case TextureType::ColorArray2D:
     case TextureType::Color3D:
     case TextureType::ColorArrayCube:
-        return ctx.OpCompositeConstruct(ctx.U32[4], ctx.OpImageQuerySizeLod(ctx.U32[3], image, lod),
-                                        mips());
+        return ctx.OpCompositeConstruct(ctx.U32[4], query(ctx.U32[3]), mips());
     case TextureType::Buffer:
-        return ctx.OpCompositeConstruct(ctx.U32[4], ctx.OpImageQuerySize(ctx.U32[1], image), zero,
-                                        zero, mips());
+        return ctx.OpCompositeConstruct(ctx.U32[4], query(ctx.U32[1]), zero, zero, mips());
     }
     throw LogicError("Unspecified image type {}", info.type.Value());
 }
diff --git a/src/shader_recompiler/backend/spirv/emit_spirv_instructions.h b/src/shader_recompiler/backend/spirv/emit_spirv_instructions.h
index db12e8176..a440b557d 100644
--- a/src/shader_recompiler/backend/spirv/emit_spirv_instructions.h
+++ b/src/shader_recompiler/backend/spirv/emit_spirv_instructions.h
@@ -539,7 +539,8 @@ Id EmitImageGatherDref(EmitContext& ctx, IR::Inst* inst, const IR::Value& index,
                        const IR::Value& offset, const IR::Value& offset2, Id dref);
 Id EmitImageFetch(EmitContext& ctx, IR::Inst* inst, const IR::Value& index, Id coords, Id offset,
                   Id lod, Id ms);
-Id EmitImageQueryDimensions(EmitContext& ctx, IR::Inst* inst, const IR::Value& index, Id lod);
+Id EmitImageQueryDimensions(EmitContext& ctx, IR::Inst* inst, const IR::Value& index, Id lod,
+                            const IR::Value& skip_mips);
 Id EmitImageQueryLod(EmitContext& ctx, IR::Inst* inst, const IR::Value& index, Id coords);
 Id EmitImageGradient(EmitContext& ctx, IR::Inst* inst, const IR::Value& index, Id coords,
                      Id derivates, Id offset, Id lod_clamp);
diff --git a/src/shader_recompiler/backend/spirv/spirv_emit_context.cpp b/src/shader_recompiler/backend/spirv/spirv_emit_context.cpp
index 3b97721e1..d48d4860e 100644
--- a/src/shader_recompiler/backend/spirv/spirv_emit_context.cpp
+++ b/src/shader_recompiler/backend/spirv/spirv_emit_context.cpp
@@ -1288,6 +1288,7 @@ void EmitContext::DefineTextures(const Info& info, u32& binding, u32& scaling_in
             .pointer_type = pointer_type,
             .image_type = image_type,
             .count = desc.count,
+            .is_multisample = desc.is_multisample,
         });
         if (profile.supported_spirv >= 0x00010400) {
             interfaces.push_back(id);
diff --git a/src/shader_recompiler/backend/spirv/spirv_emit_context.h b/src/shader_recompiler/backend/spirv/spirv_emit_context.h
index dbc5c55b9..768a4fbb5 100644
--- a/src/shader_recompiler/backend/spirv/spirv_emit_context.h
+++ b/src/shader_recompiler/backend/spirv/spirv_emit_context.h
@@ -35,6 +35,7 @@ struct TextureDefinition {
     Id pointer_type;
     Id image_type;
     u32 count;
+    bool is_multisample;
 };
 
 struct TextureBufferDefinition {
diff --git a/src/shader_recompiler/frontend/ir/ir_emitter.cpp b/src/shader_recompiler/frontend/ir/ir_emitter.cpp
index 430797d23..b7caa4246 100644
--- a/src/shader_recompiler/frontend/ir/ir_emitter.cpp
+++ b/src/shader_recompiler/frontend/ir/ir_emitter.cpp
@@ -1846,15 +1846,16 @@ Value IREmitter::ImageFetch(const Value& handle, const Value& coords, const Valu
     return Inst(op, Flags{info}, handle, coords, offset, lod, multisampling);
 }
 
-Value IREmitter::ImageQueryDimension(const Value& handle, const IR::U32& lod) {
+Value IREmitter::ImageQueryDimension(const Value& handle, const IR::U32& lod,
+                                     const IR::U1& skip_mips) {
     const Opcode op{handle.IsImmediate() ? Opcode::BoundImageQueryDimensions
                                          : Opcode::BindlessImageQueryDimensions};
-    return Inst(op, handle, lod);
+    return Inst(op, handle, lod, skip_mips);
 }
 
 Value IREmitter::ImageQueryDimension(const Value& handle, const IR::U32& lod,
-                                     TextureInstInfo info) {
-    return Inst(Opcode::ImageQueryDimensions, Flags{info}, handle, lod);
+                                     const IR::U1& skip_mips, TextureInstInfo info) {
+    return Inst(Opcode::ImageQueryDimensions, Flags{info}, handle, lod, skip_mips);
 }
 
 Value IREmitter::ImageQueryLod(const Value& handle, const Value& coords, TextureInstInfo info) {
diff --git a/src/shader_recompiler/frontend/ir/ir_emitter.h b/src/shader_recompiler/frontend/ir/ir_emitter.h
index 7aaaa4ab0..f3c81dbe1 100644
--- a/src/shader_recompiler/frontend/ir/ir_emitter.h
+++ b/src/shader_recompiler/frontend/ir/ir_emitter.h
@@ -320,9 +320,10 @@ public:
     [[nodiscard]] F32 ImageSampleDrefExplicitLod(const Value& handle, const Value& coords,
                                                  const F32& dref, const F32& lod,
                                                  const Value& offset, TextureInstInfo info);
-    [[nodiscard]] Value ImageQueryDimension(const Value& handle, const IR::U32& lod);
     [[nodiscard]] Value ImageQueryDimension(const Value& handle, const IR::U32& lod,
-                                            TextureInstInfo info);
+                                            const IR::U1& skip_mips);
+    [[nodiscard]] Value ImageQueryDimension(const Value& handle, const IR::U32& lod,
+                                            const IR::U1& skip_mips, TextureInstInfo info);
 
     [[nodiscard]] Value ImageQueryLod(const Value& handle, const Value& coords,
                                       TextureInstInfo info);
@@ -408,7 +409,8 @@ private:
     }
 
     template <typename T>
-    requires(sizeof(T) <= sizeof(u32) && std::is_trivially_copyable_v<T>) struct Flags {
+        requires(sizeof(T) <= sizeof(u32) && std::is_trivially_copyable_v<T>)
+    struct Flags {
         Flags() = default;
         Flags(T proxy_) : proxy{proxy_} {}
 
diff --git a/src/shader_recompiler/frontend/ir/opcodes.inc b/src/shader_recompiler/frontend/ir/opcodes.inc
index 24e82f802..4447d67b0 100644
--- a/src/shader_recompiler/frontend/ir/opcodes.inc
+++ b/src/shader_recompiler/frontend/ir/opcodes.inc
@@ -482,7 +482,7 @@ OPCODE(BindlessImageSampleDrefExplicitLod,                  F32,            U32,
 OPCODE(BindlessImageGather,                                 F32x4,          U32,            Opaque,         Opaque,         Opaque,                         )
 OPCODE(BindlessImageGatherDref,                             F32x4,          U32,            Opaque,         Opaque,         Opaque,         F32,            )
 OPCODE(BindlessImageFetch,                                  F32x4,          U32,            Opaque,         Opaque,         U32,            Opaque,         )
-OPCODE(BindlessImageQueryDimensions,                        U32x4,          U32,            U32,                                                            )
+OPCODE(BindlessImageQueryDimensions,                        U32x4,          U32,            U32,            U1,                                             )
 OPCODE(BindlessImageQueryLod,                               F32x4,          U32,            Opaque,                                                         )
 OPCODE(BindlessImageGradient,                               F32x4,          U32,            Opaque,         Opaque,         Opaque,         Opaque,         )
 OPCODE(BindlessImageRead,                                   U32x4,          U32,            Opaque,                                                         )
@@ -495,7 +495,7 @@ OPCODE(BoundImageSampleDrefExplicitLod,                     F32,            U32,
 OPCODE(BoundImageGather,                                    F32x4,          U32,            Opaque,         Opaque,         Opaque,                         )
 OPCODE(BoundImageGatherDref,                                F32x4,          U32,            Opaque,         Opaque,         Opaque,         F32,            )
 OPCODE(BoundImageFetch,                                     F32x4,          U32,            Opaque,         Opaque,         U32,            Opaque,         )
-OPCODE(BoundImageQueryDimensions,                           U32x4,          U32,            U32,                                                            )
+OPCODE(BoundImageQueryDimensions,                           U32x4,          U32,            U32,            U1,                                             )
 OPCODE(BoundImageQueryLod,                                  F32x4,          U32,            Opaque,                                                         )
 OPCODE(BoundImageGradient,                                  F32x4,          U32,            Opaque,         Opaque,         Opaque,         Opaque,         )
 OPCODE(BoundImageRead,                                      U32x4,          U32,            Opaque,                                                         )
@@ -508,7 +508,7 @@ OPCODE(ImageSampleDrefExplicitLod,                          F32,            Opaq
 OPCODE(ImageGather,                                         F32x4,          Opaque,         Opaque,         Opaque,         Opaque,                         )
 OPCODE(ImageGatherDref,                                     F32x4,          Opaque,         Opaque,         Opaque,         Opaque,         F32,            )
 OPCODE(ImageFetch,                                          F32x4,          Opaque,         Opaque,         Opaque,         U32,            Opaque,         )
-OPCODE(ImageQueryDimensions,                                U32x4,          Opaque,         U32,                                                            )
+OPCODE(ImageQueryDimensions,                                U32x4,          Opaque,         U32,            U1,                                             )
 OPCODE(ImageQueryLod,                                       F32x4,          Opaque,         Opaque,                                                         )
 OPCODE(ImageGradient,                                       F32x4,          Opaque,         Opaque,         Opaque,         Opaque,         Opaque,         )
 OPCODE(ImageRead,                                           U32x4,          Opaque,         Opaque,                                                         )
diff --git a/src/shader_recompiler/frontend/ir/value.h b/src/shader_recompiler/frontend/ir/value.h
index 883dfa24e..c27546b0e 100644
--- a/src/shader_recompiler/frontend/ir/value.h
+++ b/src/shader_recompiler/frontend/ir/value.h
@@ -43,7 +43,6 @@ public:
     explicit Value(u8 value) noexcept;
     explicit Value(u16 value) noexcept;
     explicit Value(u32 value) noexcept;
-    explicit Value(s32 value) noexcept;
     explicit Value(f32 value) noexcept;
     explicit Value(u64 value) noexcept;
     explicit Value(f64 value) noexcept;
@@ -66,7 +65,6 @@ public:
     [[nodiscard]] u8 U8() const;
     [[nodiscard]] u16 U16() const;
     [[nodiscard]] u32 U32() const;
-    [[nodiscard]] s32 S32() const;
     [[nodiscard]] f32 F32() const;
     [[nodiscard]] u64 U64() const;
     [[nodiscard]] f64 F64() const;
@@ -86,7 +84,6 @@ private:
         u8 imm_u8;
         u16 imm_u16;
         u32 imm_u32;
-        s32 imm_s32;
         f32 imm_f32;
         u64 imm_u64;
         f64 imm_f64;
@@ -101,9 +98,8 @@ public:
     TypedValue() = default;
 
     template <IR::Type other_type>
-    requires((other_type & type_) != IR::Type::Void) explicit(false)
-        TypedValue(const TypedValue<other_type>& value)
-        : Value(value) {}
+        requires((other_type & type_) != IR::Type::Void)
+    explicit(false) TypedValue(const TypedValue<other_type>& value) : Value(value) {}
 
     explicit TypedValue(const Value& value) : Value(value) {
         if ((value.Type() & type_) == IR::Type::Void) {
@@ -194,16 +190,16 @@ public:
     void ReplaceOpcode(IR::Opcode opcode);
 
     template <typename FlagsType>
-    requires(sizeof(FlagsType) <= sizeof(u32) && std::is_trivially_copyable_v<FlagsType>)
-        [[nodiscard]] FlagsType Flags() const noexcept {
+        requires(sizeof(FlagsType) <= sizeof(u32) && std::is_trivially_copyable_v<FlagsType>)
+    [[nodiscard]] FlagsType Flags() const noexcept {
         FlagsType ret;
         std::memcpy(reinterpret_cast<char*>(&ret), &flags, sizeof(ret));
         return ret;
     }
 
     template <typename FlagsType>
-    requires(sizeof(FlagsType) <= sizeof(u32) &&
-             std::is_trivially_copyable_v<FlagsType>) void SetFlags(FlagsType value) noexcept {
+        requires(sizeof(FlagsType) <= sizeof(u32) && std::is_trivially_copyable_v<FlagsType>)
+    void SetFlags(FlagsType value) noexcept {
         std::memcpy(&flags, &value, sizeof(value));
     }
 
@@ -379,14 +375,6 @@ inline u32 Value::U32() const {
     return imm_u32;
 }
 
-inline s32 Value::S32() const {
-    if (IsIdentity()) {
-        return inst->Arg(0).S32();
-    }
-    DEBUG_ASSERT(type == Type::S32);
-    return imm_s32;
-}
-
 inline f32 Value::F32() const {
     if (IsIdentity()) {
         return inst->Arg(0).F32();
diff --git a/src/shader_recompiler/frontend/maxwell/translate/impl/texture_query.cpp b/src/shader_recompiler/frontend/maxwell/translate/impl/texture_query.cpp
index f8cfd4ab6..39af62559 100644
--- a/src/shader_recompiler/frontend/maxwell/translate/impl/texture_query.cpp
+++ b/src/shader_recompiler/frontend/maxwell/translate/impl/texture_query.cpp
@@ -15,11 +15,13 @@ enum class Mode : u64 {
     SamplePos = 5,
 };
 
-IR::Value Query(TranslatorVisitor& v, const IR::U32& handle, Mode mode, IR::Reg src_reg) {
+IR::Value Query(TranslatorVisitor& v, const IR::U32& handle, Mode mode, IR::Reg src_reg, u64 mask) {
     switch (mode) {
     case Mode::Dimension: {
+        const bool needs_num_mips{((mask >> 3) & 1) != 0};
+        const IR::U1 skip_mips{v.ir.Imm1(!needs_num_mips)};
         const IR::U32 lod{v.X(src_reg)};
-        return v.ir.ImageQueryDimension(handle, lod);
+        return v.ir.ImageQueryDimension(handle, lod, skip_mips);
     }
     case Mode::TextureType:
     case Mode::SamplePos:
@@ -46,7 +48,7 @@ void Impl(TranslatorVisitor& v, u64 insn, std::optional<u32> cbuf_offset) {
         handle = v.X(src_reg);
         ++src_reg;
     }
-    const IR::Value query{Query(v, handle, txq.mode, src_reg)};
+    const IR::Value query{Query(v, handle, txq.mode, src_reg, txq.mask)};
     IR::Reg dest_reg{txq.dest_reg};
     for (int element = 0; element < 4; ++element) {
         if (((txq.mask >> element) & 1) == 0) {
diff --git a/src/shader_recompiler/ir_opt/texture_pass.cpp b/src/shader_recompiler/ir_opt/texture_pass.cpp
index 70b620dcb..d374c976a 100644
--- a/src/shader_recompiler/ir_opt/texture_pass.cpp
+++ b/src/shader_recompiler/ir_opt/texture_pass.cpp
@@ -355,21 +355,21 @@ TextureInst MakeInst(Environment& env, IR::Block* block, IR::Inst& inst) {
     };
 }
 
-TextureType ReadTextureType(Environment& env, const ConstBufferAddr& cbuf) {
+u32 GetTextureHandle(Environment& env, const ConstBufferAddr& cbuf) {
     const u32 secondary_index{cbuf.has_secondary ? cbuf.secondary_index : cbuf.index};
     const u32 secondary_offset{cbuf.has_secondary ? cbuf.secondary_offset : cbuf.offset};
     const u32 lhs_raw{env.ReadCbufValue(cbuf.index, cbuf.offset) << cbuf.shift_left};
     const u32 rhs_raw{env.ReadCbufValue(secondary_index, secondary_offset)
                       << cbuf.secondary_shift_left};
-    return env.ReadTextureType(lhs_raw | rhs_raw);
+    return lhs_raw | rhs_raw;
+}
+
+TextureType ReadTextureType(Environment& env, const ConstBufferAddr& cbuf) {
+    return env.ReadTextureType(GetTextureHandle(env, cbuf));
 }
 
 TexturePixelFormat ReadTexturePixelFormat(Environment& env, const ConstBufferAddr& cbuf) {
-    const u32 secondary_index{cbuf.has_secondary ? cbuf.secondary_index : cbuf.index};
-    const u32 secondary_offset{cbuf.has_secondary ? cbuf.secondary_offset : cbuf.offset};
-    const u32 lhs_raw{env.ReadCbufValue(cbuf.index, cbuf.offset)};
-    const u32 rhs_raw{env.ReadCbufValue(secondary_index, secondary_offset)};
-    return env.ReadTexturePixelFormat(lhs_raw | rhs_raw);
+    return env.ReadTexturePixelFormat(GetTextureHandle(env, cbuf));
 }
 
 class Descriptors {
@@ -386,8 +386,10 @@ public:
         return Add(texture_buffer_descriptors, desc, [&desc](const auto& existing) {
             return desc.cbuf_index == existing.cbuf_index &&
                    desc.cbuf_offset == existing.cbuf_offset &&
+                   desc.shift_left == existing.shift_left &&
                    desc.secondary_cbuf_index == existing.secondary_cbuf_index &&
                    desc.secondary_cbuf_offset == existing.secondary_cbuf_offset &&
+                   desc.secondary_shift_left == existing.secondary_shift_left &&
                    desc.count == existing.count && desc.size_shift == existing.size_shift &&
                    desc.has_secondary == existing.has_secondary;
         });
@@ -405,15 +407,20 @@ public:
     }
 
     u32 Add(const TextureDescriptor& desc) {
-        return Add(texture_descriptors, desc, [&desc](const auto& existing) {
+        const u32 index{Add(texture_descriptors, desc, [&desc](const auto& existing) {
             return desc.type == existing.type && desc.is_depth == existing.is_depth &&
                    desc.has_secondary == existing.has_secondary &&
                    desc.cbuf_index == existing.cbuf_index &&
                    desc.cbuf_offset == existing.cbuf_offset &&
+                   desc.shift_left == existing.shift_left &&
                    desc.secondary_cbuf_index == existing.secondary_cbuf_index &&
                    desc.secondary_cbuf_offset == existing.secondary_cbuf_offset &&
+                   desc.secondary_shift_left == existing.secondary_shift_left &&
                    desc.count == existing.count && desc.size_shift == existing.size_shift;
-        });
+        })};
+        // TODO: Read this from TIC
+        texture_descriptors[index].is_multisample |= desc.is_multisample;
+        return index;
     }
 
     u32 Add(const ImageDescriptor& desc) {
@@ -452,7 +459,8 @@ void PatchImageSampleImplicitLod(IR::Block& block, IR::Inst& inst) {
     const IR::Value coord(inst.Arg(1));
     const IR::Value handle(ir.Imm32(0));
     const IR::U32 lod{ir.Imm32(0)};
-    const IR::Value texture_size = ir.ImageQueryDimension(handle, lod, info);
+    const IR::U1 skip_mips{ir.Imm1(true)};
+    const IR::Value texture_size = ir.ImageQueryDimension(handle, lod, skip_mips, info);
     inst.SetArg(
         1, ir.CompositeConstruct(
                ir.FPMul(IR::F32(ir.CompositeExtract(coord, 0)),
diff --git a/src/shader_recompiler/object_pool.h b/src/shader_recompiler/object_pool.h
index 2b42c4ba2..5d648b159 100644
--- a/src/shader_recompiler/object_pool.h
+++ b/src/shader_recompiler/object_pool.h
@@ -10,7 +10,7 @@
 namespace Shader {
 
 template <typename T>
-requires std::is_destructible_v<T>
+    requires std::is_destructible_v<T>
 class ObjectPool {
 public:
     explicit ObjectPool(size_t chunk_size = 8192) : new_chunk_size{chunk_size} {
@@ -18,7 +18,7 @@ public:
     }
 
     template <typename... Args>
-    requires std::is_constructible_v<T, Args...>
+        requires std::is_constructible_v<T, Args...>
     [[nodiscard]] T* Create(Args&&... args) {
         return std::construct_at(Memory(), std::forward<Args>(args)...);
     }
diff --git a/src/video_core/host_shaders/CMakeLists.txt b/src/video_core/host_shaders/CMakeLists.txt
index dad7b07d4..52cd5bb81 100644
--- a/src/video_core/host_shaders/CMakeLists.txt
+++ b/src/video_core/host_shaders/CMakeLists.txt
@@ -45,6 +45,8 @@ set(SHADER_FILES
     smaa_neighborhood_blending.vert
     smaa_neighborhood_blending.frag
     vulkan_blit_depth_stencil.frag
+    vulkan_color_clear.frag
+    vulkan_color_clear.vert
     vulkan_fidelityfx_fsr_easu_fp16.comp
     vulkan_fidelityfx_fsr_easu_fp32.comp
     vulkan_fidelityfx_fsr_rcas_fp16.comp
diff --git a/src/video_core/host_shaders/vulkan_color_clear.frag b/src/video_core/host_shaders/vulkan_color_clear.frag
new file mode 100644
index 000000000..617bf01e1
--- /dev/null
+++ b/src/video_core/host_shaders/vulkan_color_clear.frag
@@ -0,0 +1,14 @@
+// SPDX-FileCopyrightText: Copyright 2023 yuzu Emulator Project
+// SPDX-License-Identifier: GPL-2.0-or-later
+
+#version 460 core
+
+layout (push_constant) uniform PushConstants {
+    vec4 clear_color;
+};
+
+layout(location = 0) out vec4 color;
+
+void main() {
+    color = clear_color;
+}
diff --git a/src/video_core/host_shaders/vulkan_color_clear.vert b/src/video_core/host_shaders/vulkan_color_clear.vert
new file mode 100644
index 000000000..d85883141
--- /dev/null
+++ b/src/video_core/host_shaders/vulkan_color_clear.vert
@@ -0,0 +1,10 @@
+// SPDX-FileCopyrightText: Copyright 2023 yuzu Emulator Project
+// SPDX-License-Identifier: GPL-2.0-or-later
+
+#version 460 core
+
+void main() {
+    float x = float((gl_VertexIndex & 1) << 2);
+    float y = float((gl_VertexIndex & 2) << 1);
+    gl_Position = vec4(x - 1.0, y - 1.0, 0.0, 1.0);
+}
diff --git a/src/video_core/renderer_opengl/gl_compute_pipeline.cpp b/src/video_core/renderer_opengl/gl_compute_pipeline.cpp
index 26d066004..1a0cea9b7 100644
--- a/src/video_core/renderer_opengl/gl_compute_pipeline.cpp
+++ b/src/video_core/renderer_opengl/gl_compute_pipeline.cpp
@@ -30,7 +30,7 @@ bool ComputePipelineKey::operator==(const ComputePipelineKey& rhs) const noexcep
 ComputePipeline::ComputePipeline(const Device& device, TextureCache& texture_cache_,
                                  BufferCache& buffer_cache_, ProgramManager& program_manager_,
                                  const Shader::Info& info_, std::string code,
-                                 std::vector<u32> code_v)
+                                 std::vector<u32> code_v, bool force_context_flush)
     : texture_cache{texture_cache_}, buffer_cache{buffer_cache_},
       program_manager{program_manager_}, info{info_} {
     switch (device.GetShaderBackend()) {
@@ -63,6 +63,15 @@ ComputePipeline::ComputePipeline(const Device& device, TextureCache& texture_cac
     writes_global_memory = !use_storage_buffers &&
                            std::ranges::any_of(info.storage_buffers_descriptors,
                                                [](const auto& desc) { return desc.is_written; });
+    if (force_context_flush) {
+        std::scoped_lock lock{built_mutex};
+        built_fence.Create();
+        // Flush this context to ensure compilation commands and fence are in the GPU pipe.
+        glFlush();
+        built_condvar.notify_one();
+    } else {
+        is_built = true;
+    }
 }
 
 void ComputePipeline::Configure() {
@@ -142,6 +151,9 @@ void ComputePipeline::Configure() {
     }
     texture_cache.FillComputeImageViews(std::span(views.data(), views.size()));
 
+    if (!is_built) {
+        WaitForBuild();
+    }
     if (assembly_program.handle != 0) {
         program_manager.BindComputeAssemblyProgram(assembly_program.handle);
     } else {
@@ -223,4 +235,13 @@ void ComputePipeline::Configure() {
     }
 }
 
+void ComputePipeline::WaitForBuild() {
+    if (built_fence.handle == 0) {
+        std::unique_lock lock{built_mutex};
+        built_condvar.wait(lock, [this] { return built_fence.handle != 0; });
+    }
+    ASSERT(glClientWaitSync(built_fence.handle, 0, GL_TIMEOUT_IGNORED) != GL_WAIT_FAILED);
+    is_built = true;
+}
+
 } // namespace OpenGL
diff --git a/src/video_core/renderer_opengl/gl_compute_pipeline.h b/src/video_core/renderer_opengl/gl_compute_pipeline.h
index 6534dec32..9bcc72b59 100644
--- a/src/video_core/renderer_opengl/gl_compute_pipeline.h
+++ b/src/video_core/renderer_opengl/gl_compute_pipeline.h
@@ -50,7 +50,8 @@ class ComputePipeline {
 public:
     explicit ComputePipeline(const Device& device, TextureCache& texture_cache_,
                              BufferCache& buffer_cache_, ProgramManager& program_manager_,
-                             const Shader::Info& info_, std::string code, std::vector<u32> code_v);
+                             const Shader::Info& info_, std::string code, std::vector<u32> code_v,
+                             bool force_context_flush = false);
 
     void Configure();
 
@@ -65,6 +66,8 @@ public:
     }
 
 private:
+    void WaitForBuild();
+
     TextureCache& texture_cache;
     BufferCache& buffer_cache;
     Tegra::MemoryManager* gpu_memory;
@@ -81,6 +84,11 @@ private:
 
     bool use_storage_buffers{};
     bool writes_global_memory{};
+
+    std::mutex built_mutex;
+    std::condition_variable built_condvar;
+    OGLSync built_fence{};
+    bool is_built{false};
 };
 
 } // namespace OpenGL
diff --git a/src/video_core/renderer_opengl/gl_graphics_pipeline.cpp b/src/video_core/renderer_opengl/gl_graphics_pipeline.cpp
index c115dabe1..29491e762 100644
--- a/src/video_core/renderer_opengl/gl_graphics_pipeline.cpp
+++ b/src/video_core/renderer_opengl/gl_graphics_pipeline.cpp
@@ -176,7 +176,7 @@ GraphicsPipeline::GraphicsPipeline(const Device& device, TextureCache& texture_c
                                    std::array<std::string, 5> sources,
                                    std::array<std::vector<u32>, 5> sources_spirv,
                                    const std::array<const Shader::Info*, 5>& infos,
-                                   const GraphicsPipelineKey& key_)
+                                   const GraphicsPipelineKey& key_, bool force_context_flush)
     : texture_cache{texture_cache_}, buffer_cache{buffer_cache_}, program_manager{program_manager_},
       state_tracker{state_tracker_}, key{key_} {
     if (shader_notify) {
@@ -231,7 +231,8 @@ GraphicsPipeline::GraphicsPipeline(const Device& device, TextureCache& texture_c
     const bool in_parallel = thread_worker != nullptr;
     const auto backend = device.GetShaderBackend();
     auto func{[this, sources = std::move(sources), sources_spirv = std::move(sources_spirv),
-               shader_notify, backend, in_parallel](ShaderContext::Context*) mutable {
+               shader_notify, backend, in_parallel,
+               force_context_flush](ShaderContext::Context*) mutable {
         for (size_t stage = 0; stage < 5; ++stage) {
             switch (backend) {
             case Settings::ShaderBackend::GLSL:
@@ -251,7 +252,7 @@ GraphicsPipeline::GraphicsPipeline(const Device& device, TextureCache& texture_c
                 break;
             }
         }
-        if (in_parallel) {
+        if (force_context_flush || in_parallel) {
             std::scoped_lock lock{built_mutex};
             built_fence.Create();
             // Flush this context to ensure compilation commands and fence are in the GPU pipe.
diff --git a/src/video_core/renderer_opengl/gl_graphics_pipeline.h b/src/video_core/renderer_opengl/gl_graphics_pipeline.h
index 1c06b3655..7bab3be0a 100644
--- a/src/video_core/renderer_opengl/gl_graphics_pipeline.h
+++ b/src/video_core/renderer_opengl/gl_graphics_pipeline.h
@@ -78,7 +78,7 @@ public:
                               std::array<std::string, 5> sources,
                               std::array<std::vector<u32>, 5> sources_spirv,
                               const std::array<const Shader::Info*, 5>& infos,
-                              const GraphicsPipelineKey& key_);
+                              const GraphicsPipelineKey& key_, bool force_context_flush = false);
 
     void Configure(bool is_indexed) {
         configure_func(this, is_indexed);
diff --git a/src/video_core/renderer_opengl/gl_shader_cache.cpp b/src/video_core/renderer_opengl/gl_shader_cache.cpp
index 7dd854e0f..626ea7dcb 100644
--- a/src/video_core/renderer_opengl/gl_shader_cache.cpp
+++ b/src/video_core/renderer_opengl/gl_shader_cache.cpp
@@ -286,7 +286,7 @@ void ShaderCache::LoadDiskResources(u64 title_id, std::stop_token stop_loading,
         file.read(reinterpret_cast<char*>(&key), sizeof(key));
         queue_work([this, key, env = std::move(env), &state, &callback](Context* ctx) mutable {
             ctx->pools.ReleaseContents();
-            auto pipeline{CreateComputePipeline(ctx->pools, key, env)};
+            auto pipeline{CreateComputePipeline(ctx->pools, key, env, true)};
             std::scoped_lock lock{state.mutex};
             if (pipeline) {
                 compute_cache.emplace(key, std::move(pipeline));
@@ -307,7 +307,7 @@ void ShaderCache::LoadDiskResources(u64 title_id, std::stop_token stop_loading,
                 env_ptrs.push_back(&env);
             }
             ctx->pools.ReleaseContents();
-            auto pipeline{CreateGraphicsPipeline(ctx->pools, key, MakeSpan(env_ptrs), false)};
+            auto pipeline{CreateGraphicsPipeline(ctx->pools, key, MakeSpan(env_ptrs), false, true)};
             std::scoped_lock lock{state.mutex};
             if (pipeline) {
                 graphics_cache.emplace(key, std::move(pipeline));
@@ -439,7 +439,8 @@ std::unique_ptr<GraphicsPipeline> ShaderCache::CreateGraphicsPipeline() {
 
 std::unique_ptr<GraphicsPipeline> ShaderCache::CreateGraphicsPipeline(
     ShaderContext::ShaderPools& pools, const GraphicsPipelineKey& key,
-    std::span<Shader::Environment* const> envs, bool build_in_parallel) try {
+    std::span<Shader::Environment* const> envs, bool use_shader_workers,
+    bool force_context_flush) try {
     LOG_INFO(Render_OpenGL, "0x{:016x}", key.Hash());
     size_t env_index{};
     u32 total_storage_buffers{};
@@ -531,10 +532,10 @@ std::unique_ptr<GraphicsPipeline> ShaderCache::CreateGraphicsPipeline(
         }
         previous_program = &program;
     }
-    auto* const thread_worker{build_in_parallel ? workers.get() : nullptr};
+    auto* const thread_worker{use_shader_workers ? workers.get() : nullptr};
     return std::make_unique<GraphicsPipeline>(device, texture_cache, buffer_cache, program_manager,
                                               state_tracker, thread_worker, &shader_notify, sources,
-                                              sources_spirv, infos, key);
+                                              sources_spirv, infos, key, force_context_flush);
 
 } catch (Shader::Exception& exception) {
     LOG_ERROR(Render_OpenGL, "{}", exception.what());
@@ -559,8 +560,8 @@ std::unique_ptr<ComputePipeline> ShaderCache::CreateComputePipeline(
 }
 
 std::unique_ptr<ComputePipeline> ShaderCache::CreateComputePipeline(
-    ShaderContext::ShaderPools& pools, const ComputePipelineKey& key,
-    Shader::Environment& env) try {
+    ShaderContext::ShaderPools& pools, const ComputePipelineKey& key, Shader::Environment& env,
+    bool force_context_flush) try {
     LOG_INFO(Render_OpenGL, "0x{:016x}", key.Hash());
 
     Shader::Maxwell::Flow::CFG cfg{env, pools.flow_block, env.StartAddress()};
@@ -589,7 +590,7 @@ std::unique_ptr<ComputePipeline> ShaderCache::CreateComputePipeline(
     }
 
     return std::make_unique<ComputePipeline>(device, texture_cache, buffer_cache, program_manager,
-                                             program.info, code, code_spirv);
+                                             program.info, code, code_spirv, force_context_flush);
 } catch (Shader::Exception& exception) {
     LOG_ERROR(Render_OpenGL, "{}", exception.what());
     return nullptr;
diff --git a/src/video_core/renderer_opengl/gl_shader_cache.h b/src/video_core/renderer_opengl/gl_shader_cache.h
index f82420592..6b9732fca 100644
--- a/src/video_core/renderer_opengl/gl_shader_cache.h
+++ b/src/video_core/renderer_opengl/gl_shader_cache.h
@@ -50,14 +50,16 @@ private:
 
     std::unique_ptr<GraphicsPipeline> CreateGraphicsPipeline(
         ShaderContext::ShaderPools& pools, const GraphicsPipelineKey& key,
-        std::span<Shader::Environment* const> envs, bool build_in_parallel);
+        std::span<Shader::Environment* const> envs, bool use_shader_workers,
+        bool force_context_flush = false);
 
     std::unique_ptr<ComputePipeline> CreateComputePipeline(const ComputePipelineKey& key,
                                                            const VideoCommon::ShaderInfo* shader);
 
     std::unique_ptr<ComputePipeline> CreateComputePipeline(ShaderContext::ShaderPools& pools,
                                                            const ComputePipelineKey& key,
-                                                           Shader::Environment& env);
+                                                           Shader::Environment& env,
+                                                           bool force_context_flush = false);
 
     std::unique_ptr<ShaderWorker> CreateWorkers() const;
 
diff --git a/src/video_core/renderer_vulkan/blit_image.cpp b/src/video_core/renderer_vulkan/blit_image.cpp
index dd00d3edf..cf2964a3f 100644
--- a/src/video_core/renderer_vulkan/blit_image.cpp
+++ b/src/video_core/renderer_vulkan/blit_image.cpp
@@ -12,6 +12,8 @@
 #include "video_core/host_shaders/convert_s8d24_to_abgr8_frag_spv.h"
 #include "video_core/host_shaders/full_screen_triangle_vert_spv.h"
 #include "video_core/host_shaders/vulkan_blit_depth_stencil_frag_spv.h"
+#include "video_core/host_shaders/vulkan_color_clear_frag_spv.h"
+#include "video_core/host_shaders/vulkan_color_clear_vert_spv.h"
 #include "video_core/renderer_vulkan/blit_image.h"
 #include "video_core/renderer_vulkan/maxwell_to_vk.h"
 #include "video_core/renderer_vulkan/vk_scheduler.h"
@@ -69,10 +71,11 @@ constexpr VkDescriptorSetLayoutCreateInfo TWO_TEXTURES_DESCRIPTOR_SET_LAYOUT_CRE
     .bindingCount = static_cast<u32>(TWO_TEXTURES_DESCRIPTOR_SET_LAYOUT_BINDINGS.size()),
     .pBindings = TWO_TEXTURES_DESCRIPTOR_SET_LAYOUT_BINDINGS.data(),
 };
-constexpr VkPushConstantRange PUSH_CONSTANT_RANGE{
-    .stageFlags = VK_SHADER_STAGE_VERTEX_BIT,
+template <VkShaderStageFlags stageFlags, size_t size>
+inline constexpr VkPushConstantRange PUSH_CONSTANT_RANGE{
+    .stageFlags = stageFlags,
     .offset = 0,
-    .size = sizeof(PushConstants),
+    .size = static_cast<u32>(size),
 };
 constexpr VkPipelineVertexInputStateCreateInfo PIPELINE_VERTEX_INPUT_STATE_CREATE_INFO{
     .sType = VK_STRUCTURE_TYPE_PIPELINE_VERTEX_INPUT_STATE_CREATE_INFO,
@@ -125,10 +128,8 @@ constexpr VkPipelineMultisampleStateCreateInfo PIPELINE_MULTISAMPLE_STATE_CREATE
     .alphaToCoverageEnable = VK_FALSE,
     .alphaToOneEnable = VK_FALSE,
 };
-constexpr std::array DYNAMIC_STATES{
-    VK_DYNAMIC_STATE_VIEWPORT,
-    VK_DYNAMIC_STATE_SCISSOR,
-};
+constexpr std::array DYNAMIC_STATES{VK_DYNAMIC_STATE_VIEWPORT, VK_DYNAMIC_STATE_SCISSOR,
+                                    VK_DYNAMIC_STATE_BLEND_CONSTANTS};
 constexpr VkPipelineDynamicStateCreateInfo PIPELINE_DYNAMIC_STATE_CREATE_INFO{
     .sType = VK_STRUCTURE_TYPE_PIPELINE_DYNAMIC_STATE_CREATE_INFO,
     .pNext = nullptr,
@@ -205,15 +206,15 @@ inline constexpr VkSamplerCreateInfo SAMPLER_CREATE_INFO{
 };
 
 constexpr VkPipelineLayoutCreateInfo PipelineLayoutCreateInfo(
-    const VkDescriptorSetLayout* set_layout) {
+    const VkDescriptorSetLayout* set_layout, vk::Span<VkPushConstantRange> push_constants) {
     return VkPipelineLayoutCreateInfo{
         .sType = VK_STRUCTURE_TYPE_PIPELINE_LAYOUT_CREATE_INFO,
         .pNext = nullptr,
         .flags = 0,
-        .setLayoutCount = 1,
+        .setLayoutCount = (set_layout != nullptr ? 1u : 0u),
         .pSetLayouts = set_layout,
-        .pushConstantRangeCount = 1,
-        .pPushConstantRanges = &PUSH_CONSTANT_RANGE,
+        .pushConstantRangeCount = push_constants.size(),
+        .pPushConstantRanges = push_constants.data(),
     };
 }
 
@@ -302,8 +303,7 @@ void UpdateTwoTexturesDescriptorSet(const Device& device, VkDescriptorSet descri
     device.GetLogical().UpdateDescriptorSets(write_descriptor_sets, nullptr);
 }
 
-void BindBlitState(vk::CommandBuffer cmdbuf, VkPipelineLayout layout, const Region2D& dst_region,
-                   const Region2D& src_region, const Extent3D& src_size = {1, 1, 1}) {
+void BindBlitState(vk::CommandBuffer cmdbuf, const Region2D& dst_region) {
     const VkOffset2D offset{
         .x = std::min(dst_region.start.x, dst_region.end.x),
         .y = std::min(dst_region.start.y, dst_region.end.y),
@@ -325,6 +325,13 @@ void BindBlitState(vk::CommandBuffer cmdbuf, VkPipelineLayout layout, const Regi
         .offset = offset,
         .extent = extent,
     };
+    cmdbuf.SetViewport(0, viewport);
+    cmdbuf.SetScissor(0, scissor);
+}
+
+void BindBlitState(vk::CommandBuffer cmdbuf, VkPipelineLayout layout, const Region2D& dst_region,
+                   const Region2D& src_region, const Extent3D& src_size = {1, 1, 1}) {
+    BindBlitState(cmdbuf, dst_region);
     const float scale_x = static_cast<float>(src_region.end.x - src_region.start.x) /
                           static_cast<float>(src_size.width);
     const float scale_y = static_cast<float>(src_region.end.y - src_region.start.y) /
@@ -335,8 +342,6 @@ void BindBlitState(vk::CommandBuffer cmdbuf, VkPipelineLayout layout, const Regi
                        static_cast<float>(src_region.start.y) /
                            static_cast<float>(src_size.height)},
     };
-    cmdbuf.SetViewport(0, viewport);
-    cmdbuf.SetScissor(0, scissor);
     cmdbuf.PushConstants(layout, VK_SHADER_STAGE_VERTEX_BIT, push_constants);
 }
 
@@ -408,13 +413,20 @@ BlitImageHelper::BlitImageHelper(const Device& device_, Scheduler& scheduler_,
           descriptor_pool.Allocator(*one_texture_set_layout, TEXTURE_DESCRIPTOR_BANK_INFO<1>)},
       two_textures_descriptor_allocator{
           descriptor_pool.Allocator(*two_textures_set_layout, TEXTURE_DESCRIPTOR_BANK_INFO<2>)},
-      one_texture_pipeline_layout(device.GetLogical().CreatePipelineLayout(
-          PipelineLayoutCreateInfo(one_texture_set_layout.address()))),
-      two_textures_pipeline_layout(device.GetLogical().CreatePipelineLayout(
-          PipelineLayoutCreateInfo(two_textures_set_layout.address()))),
+      one_texture_pipeline_layout(device.GetLogical().CreatePipelineLayout(PipelineLayoutCreateInfo(
+          one_texture_set_layout.address(),
+          PUSH_CONSTANT_RANGE<VK_SHADER_STAGE_VERTEX_BIT, sizeof(PushConstants)>))),
+      two_textures_pipeline_layout(
+          device.GetLogical().CreatePipelineLayout(PipelineLayoutCreateInfo(
+              two_textures_set_layout.address(),
+              PUSH_CONSTANT_RANGE<VK_SHADER_STAGE_VERTEX_BIT, sizeof(PushConstants)>))),
+      clear_color_pipeline_layout(device.GetLogical().CreatePipelineLayout(PipelineLayoutCreateInfo(
+          nullptr, PUSH_CONSTANT_RANGE<VK_SHADER_STAGE_FRAGMENT_BIT, sizeof(float) * 4>))),
       full_screen_vert(BuildShader(device, FULL_SCREEN_TRIANGLE_VERT_SPV)),
       blit_color_to_color_frag(BuildShader(device, BLIT_COLOR_FLOAT_FRAG_SPV)),
       blit_depth_stencil_frag(BuildShader(device, VULKAN_BLIT_DEPTH_STENCIL_FRAG_SPV)),
+      clear_color_vert(BuildShader(device, VULKAN_COLOR_CLEAR_VERT_SPV)),
+      clear_color_frag(BuildShader(device, VULKAN_COLOR_CLEAR_FRAG_SPV)),
       convert_depth_to_float_frag(BuildShader(device, CONVERT_DEPTH_TO_FLOAT_FRAG_SPV)),
       convert_float_to_depth_frag(BuildShader(device, CONVERT_FLOAT_TO_DEPTH_FRAG_SPV)),
       convert_abgr8_to_d24s8_frag(BuildShader(device, CONVERT_ABGR8_TO_D24S8_FRAG_SPV)),
@@ -553,6 +565,30 @@ void BlitImageHelper::ConvertS8D24ToABGR8(const Framebuffer* dst_framebuffer,
     ConvertDepthStencil(*convert_s8d24_to_abgr8_pipeline, dst_framebuffer, src_image_view);
 }
 
+void BlitImageHelper::ClearColor(const Framebuffer* dst_framebuffer, u8 color_mask,
+                                 const std::array<f32, 4>& clear_color,
+                                 const Region2D& dst_region) {
+    const BlitImagePipelineKey key{
+        .renderpass = dst_framebuffer->RenderPass(),
+        .operation = Tegra::Engines::Fermi2D::Operation::BlendPremult,
+    };
+    const VkPipeline pipeline = FindOrEmplaceClearColorPipeline(key);
+    const VkPipelineLayout layout = *clear_color_pipeline_layout;
+    scheduler.RequestRenderpass(dst_framebuffer);
+    scheduler.Record(
+        [pipeline, layout, color_mask, clear_color, dst_region](vk::CommandBuffer cmdbuf) {
+            cmdbuf.BindPipeline(VK_PIPELINE_BIND_POINT_GRAPHICS, pipeline);
+            const std::array blend_color = {
+                (color_mask & 0x1) ? 1.0f : 0.0f, (color_mask & 0x2) ? 1.0f : 0.0f,
+                (color_mask & 0x4) ? 1.0f : 0.0f, (color_mask & 0x8) ? 1.0f : 0.0f};
+            cmdbuf.SetBlendConstants(blend_color.data());
+            BindBlitState(cmdbuf, dst_region);
+            cmdbuf.PushConstants(layout, VK_SHADER_STAGE_FRAGMENT_BIT, clear_color);
+            cmdbuf.Draw(3, 1, 0, 0);
+        });
+    scheduler.InvalidateState();
+}
+
 void BlitImageHelper::Convert(VkPipeline pipeline, const Framebuffer* dst_framebuffer,
                               const ImageView& src_image_view) {
     const VkPipelineLayout layout = *one_texture_pipeline_layout;
@@ -728,6 +764,58 @@ VkPipeline BlitImageHelper::FindOrEmplaceDepthStencilPipeline(const BlitImagePip
     return *blit_depth_stencil_pipelines.back();
 }
 
+VkPipeline BlitImageHelper::FindOrEmplaceClearColorPipeline(const BlitImagePipelineKey& key) {
+    const auto it = std::ranges::find(clear_color_keys, key);
+    if (it != clear_color_keys.end()) {
+        return *clear_color_pipelines[std::distance(clear_color_keys.begin(), it)];
+    }
+    clear_color_keys.push_back(key);
+    const std::array stages = MakeStages(*clear_color_vert, *clear_color_frag);
+    const VkPipelineColorBlendAttachmentState color_blend_attachment_state{
+        .blendEnable = VK_TRUE,
+        .srcColorBlendFactor = VK_BLEND_FACTOR_CONSTANT_COLOR,
+        .dstColorBlendFactor = VK_BLEND_FACTOR_ONE_MINUS_CONSTANT_COLOR,
+        .colorBlendOp = VK_BLEND_OP_ADD,
+        .srcAlphaBlendFactor = VK_BLEND_FACTOR_CONSTANT_ALPHA,
+        .dstAlphaBlendFactor = VK_BLEND_FACTOR_ONE_MINUS_CONSTANT_ALPHA,
+        .alphaBlendOp = VK_BLEND_OP_ADD,
+        .colorWriteMask = VK_COLOR_COMPONENT_R_BIT | VK_COLOR_COMPONENT_G_BIT |
+                          VK_COLOR_COMPONENT_B_BIT | VK_COLOR_COMPONENT_A_BIT,
+    };
+    const VkPipelineColorBlendStateCreateInfo color_blend_state_generic_create_info{
+        .sType = VK_STRUCTURE_TYPE_PIPELINE_COLOR_BLEND_STATE_CREATE_INFO,
+        .pNext = nullptr,
+        .flags = 0,
+        .logicOpEnable = VK_FALSE,
+        .logicOp = VK_LOGIC_OP_CLEAR,
+        .attachmentCount = 1,
+        .pAttachments = &color_blend_attachment_state,
+        .blendConstants = {0.0f, 0.0f, 0.0f, 0.0f},
+    };
+    clear_color_pipelines.push_back(device.GetLogical().CreateGraphicsPipeline({
+        .sType = VK_STRUCTURE_TYPE_GRAPHICS_PIPELINE_CREATE_INFO,
+        .pNext = nullptr,
+        .flags = 0,
+        .stageCount = static_cast<u32>(stages.size()),
+        .pStages = stages.data(),
+        .pVertexInputState = &PIPELINE_VERTEX_INPUT_STATE_CREATE_INFO,
+        .pInputAssemblyState = &PIPELINE_INPUT_ASSEMBLY_STATE_CREATE_INFO,
+        .pTessellationState = nullptr,
+        .pViewportState = &PIPELINE_VIEWPORT_STATE_CREATE_INFO,
+        .pRasterizationState = &PIPELINE_RASTERIZATION_STATE_CREATE_INFO,
+        .pMultisampleState = &PIPELINE_MULTISAMPLE_STATE_CREATE_INFO,
+        .pDepthStencilState = &PIPELINE_DEPTH_STENCIL_STATE_CREATE_INFO,
+        .pColorBlendState = &color_blend_state_generic_create_info,
+        .pDynamicState = &PIPELINE_DYNAMIC_STATE_CREATE_INFO,
+        .layout = *clear_color_pipeline_layout,
+        .renderPass = key.renderpass,
+        .subpass = 0,
+        .basePipelineHandle = VK_NULL_HANDLE,
+        .basePipelineIndex = 0,
+    }));
+    return *clear_color_pipelines.back();
+}
+
 void BlitImageHelper::ConvertPipeline(vk::Pipeline& pipeline, VkRenderPass renderpass,
                                       bool is_target_depth) {
     if (pipeline) {
diff --git a/src/video_core/renderer_vulkan/blit_image.h b/src/video_core/renderer_vulkan/blit_image.h
index be8a9a2f6..2976a7d91 100644
--- a/src/video_core/renderer_vulkan/blit_image.h
+++ b/src/video_core/renderer_vulkan/blit_image.h
@@ -61,6 +61,9 @@ public:
 
     void ConvertS8D24ToABGR8(const Framebuffer* dst_framebuffer, ImageView& src_image_view);
 
+    void ClearColor(const Framebuffer* dst_framebuffer, u8 color_mask,
+                    const std::array<f32, 4>& clear_color, const Region2D& dst_region);
+
 private:
     void Convert(VkPipeline pipeline, const Framebuffer* dst_framebuffer,
                  const ImageView& src_image_view);
@@ -72,6 +75,8 @@ private:
 
     [[nodiscard]] VkPipeline FindOrEmplaceDepthStencilPipeline(const BlitImagePipelineKey& key);
 
+    [[nodiscard]] VkPipeline FindOrEmplaceClearColorPipeline(const BlitImagePipelineKey& key);
+
     void ConvertPipeline(vk::Pipeline& pipeline, VkRenderPass renderpass, bool is_target_depth);
 
     void ConvertDepthToColorPipeline(vk::Pipeline& pipeline, VkRenderPass renderpass);
@@ -97,9 +102,12 @@ private:
     DescriptorAllocator two_textures_descriptor_allocator;
     vk::PipelineLayout one_texture_pipeline_layout;
     vk::PipelineLayout two_textures_pipeline_layout;
+    vk::PipelineLayout clear_color_pipeline_layout;
     vk::ShaderModule full_screen_vert;
     vk::ShaderModule blit_color_to_color_frag;
     vk::ShaderModule blit_depth_stencil_frag;
+    vk::ShaderModule clear_color_vert;
+    vk::ShaderModule clear_color_frag;
     vk::ShaderModule convert_depth_to_float_frag;
     vk::ShaderModule convert_float_to_depth_frag;
     vk::ShaderModule convert_abgr8_to_d24s8_frag;
@@ -112,6 +120,8 @@ private:
     std::vector<vk::Pipeline> blit_color_pipelines;
     std::vector<BlitImagePipelineKey> blit_depth_stencil_keys;
     std::vector<vk::Pipeline> blit_depth_stencil_pipelines;
+    std::vector<BlitImagePipelineKey> clear_color_keys;
+    std::vector<vk::Pipeline> clear_color_pipelines;
     vk::Pipeline convert_d32_to_r32_pipeline;
     vk::Pipeline convert_r32_to_d32_pipeline;
     vk::Pipeline convert_d16_to_r16_pipeline;
diff --git a/src/video_core/renderer_vulkan/vk_rasterizer.cpp b/src/video_core/renderer_vulkan/vk_rasterizer.cpp
index 86ef0daeb..719edbcfb 100644
--- a/src/video_core/renderer_vulkan/vk_rasterizer.cpp
+++ b/src/video_core/renderer_vulkan/vk_rasterizer.cpp
@@ -394,7 +394,15 @@ void RasterizerVulkan::Clear(u32 layer_count) {
                 cmdbuf.ClearAttachments(attachment, clear_rect);
             });
         } else {
-            UNIMPLEMENTED_MSG("Unimplemented Clear only the specified channel");
+            u8 color_mask = static_cast<u8>(regs.clear_surface.R | regs.clear_surface.G << 1 |
+                                            regs.clear_surface.B << 2 | regs.clear_surface.A << 3);
+            Region2D dst_region = {
+                Offset2D{.x = clear_rect.rect.offset.x, .y = clear_rect.rect.offset.y},
+                Offset2D{.x = clear_rect.rect.offset.x +
+                              static_cast<s32>(clear_rect.rect.extent.width),
+                         .y = clear_rect.rect.offset.y +
+                              static_cast<s32>(clear_rect.rect.extent.height)}};
+            blit_image.ClearColor(framebuffer, color_mask, regs.clear_color, dst_region);
         }
     }
 
diff --git a/src/video_core/texture_cache/descriptor_table.h b/src/video_core/texture_cache/descriptor_table.h
index ee4240288..1bad83fb4 100644
--- a/src/video_core/texture_cache/descriptor_table.h
+++ b/src/video_core/texture_cache/descriptor_table.h
@@ -19,9 +19,7 @@ public:
     explicit DescriptorTable(Tegra::MemoryManager& gpu_memory_) : gpu_memory{gpu_memory_} {}
 
     [[nodiscard]] bool Synchronize(GPUVAddr gpu_addr, u32 limit) {
-        [[likely]] if (current_gpu_addr == gpu_addr && current_limit == limit) {
-            return false;
-        }
+        [[likely]] if (current_gpu_addr == gpu_addr && current_limit == limit) { return false; }
         Refresh(gpu_addr, limit);
         return true;
     }
diff --git a/src/video_core/texture_cache/image_info.cpp b/src/video_core/texture_cache/image_info.cpp
index 852ec2519..e9100091e 100644
--- a/src/video_core/texture_cache/image_info.cpp
+++ b/src/video_core/texture_cache/image_info.cpp
@@ -100,6 +100,10 @@ ImageInfo::ImageInfo(const TICEntry& config) noexcept {
         ASSERT_MSG(false, "Invalid texture_type={}", static_cast<int>(config.texture_type.Value()));
         break;
     }
+    if (num_samples > 1) {
+        size.width *= NumSamplesX(config.msaa_mode);
+        size.height *= NumSamplesY(config.msaa_mode);
+    }
     if (type != ImageType::Linear) {
         // FIXME: Call this without passing *this
         layer_stride = CalculateLayerStride(*this);
diff --git a/src/video_core/texture_cache/samples_helper.h b/src/video_core/texture_cache/samples_helper.h
index d552bccf0..203ac1b11 100644
--- a/src/video_core/texture_cache/samples_helper.h
+++ b/src/video_core/texture_cache/samples_helper.h
@@ -51,4 +51,48 @@ namespace VideoCommon {
     return 1;
 }
 
+[[nodiscard]] inline int NumSamplesX(Tegra::Texture::MsaaMode msaa_mode) {
+    using Tegra::Texture::MsaaMode;
+    switch (msaa_mode) {
+    case MsaaMode::Msaa1x1:
+        return 1;
+    case MsaaMode::Msaa2x1:
+    case MsaaMode::Msaa2x1_D3D:
+    case MsaaMode::Msaa2x2:
+    case MsaaMode::Msaa2x2_VC4:
+    case MsaaMode::Msaa2x2_VC12:
+        return 2;
+    case MsaaMode::Msaa4x2:
+    case MsaaMode::Msaa4x2_D3D:
+    case MsaaMode::Msaa4x2_VC8:
+    case MsaaMode::Msaa4x2_VC24:
+    case MsaaMode::Msaa4x4:
+        return 4;
+    }
+    ASSERT_MSG(false, "Invalid MSAA mode={}", static_cast<int>(msaa_mode));
+    return 1;
+}
+
+[[nodiscard]] inline int NumSamplesY(Tegra::Texture::MsaaMode msaa_mode) {
+    using Tegra::Texture::MsaaMode;
+    switch (msaa_mode) {
+    case MsaaMode::Msaa1x1:
+    case MsaaMode::Msaa2x1:
+    case MsaaMode::Msaa2x1_D3D:
+        return 1;
+    case MsaaMode::Msaa2x2:
+    case MsaaMode::Msaa2x2_VC4:
+    case MsaaMode::Msaa2x2_VC12:
+    case MsaaMode::Msaa4x2:
+    case MsaaMode::Msaa4x2_D3D:
+    case MsaaMode::Msaa4x2_VC8:
+    case MsaaMode::Msaa4x2_VC24:
+        return 2;
+    case MsaaMode::Msaa4x4:
+        return 4;
+    }
+    ASSERT_MSG(false, "Invalid MSAA mode={}", static_cast<int>(msaa_mode));
+    return 1;
+}
+
 } // namespace VideoCommon
diff --git a/src/video_core/texture_cache/slot_vector.h b/src/video_core/texture_cache/slot_vector.h
index 1e2aad76a..9df6a2903 100644
--- a/src/video_core/texture_cache/slot_vector.h
+++ b/src/video_core/texture_cache/slot_vector.h
@@ -29,7 +29,7 @@ struct SlotId {
 };
 
 template <class T>
-requires std::is_nothrow_move_assignable_v<T> && std::is_nothrow_move_constructible_v<T>
+    requires std::is_nothrow_move_assignable_v<T> && std::is_nothrow_move_constructible_v<T>
 class SlotVector {
 public:
     class Iterator {
diff --git a/src/yuzu/configuration/configure_input_player.cpp b/src/yuzu/configuration/configure_input_player.cpp
index 4b7e3b01b..723690e71 100644
--- a/src/yuzu/configuration/configure_input_player.cpp
+++ b/src/yuzu/configuration/configure_input_player.cpp
@@ -182,12 +182,13 @@ QString ConfigureInputPlayer::ButtonToText(const Common::ParamPackage& param) {
     const QString toggle = QString::fromStdString(param.Get("toggle", false) ? "~" : "");
     const QString inverted = QString::fromStdString(param.Get("inverted", false) ? "!" : "");
     const QString invert = QString::fromStdString(param.Get("invert", "+") == "-" ? "-" : "");
+    const QString turbo = QString::fromStdString(param.Get("turbo", false) ? "$" : "");
     const auto common_button_name = input_subsystem->GetButtonName(param);
 
     // Retrieve the names from Qt
     if (param.Get("engine", "") == "keyboard") {
         const QString button_str = GetKeyName(param.Get("code", 0));
-        return QObject::tr("%1%2%3").arg(toggle, inverted, button_str);
+        return QObject::tr("%1%2%3%4").arg(turbo, toggle, inverted, button_str);
     }
 
     if (common_button_name == Common::Input::ButtonNames::Invalid) {
@@ -201,7 +202,7 @@ QString ConfigureInputPlayer::ButtonToText(const Common::ParamPackage& param) {
     if (common_button_name == Common::Input::ButtonNames::Value) {
         if (param.Has("hat")) {
             const QString hat = GetDirectionName(param.Get("direction", ""));
-            return QObject::tr("%1%2Hat %3").arg(toggle, inverted, hat);
+            return QObject::tr("%1%2%3Hat %4").arg(turbo, toggle, inverted, hat);
         }
         if (param.Has("axis")) {
             const QString axis = QString::fromStdString(param.Get("axis", ""));
@@ -219,13 +220,13 @@ QString ConfigureInputPlayer::ButtonToText(const Common::ParamPackage& param) {
         }
         if (param.Has("button")) {
             const QString button = QString::fromStdString(param.Get("button", ""));
-            return QObject::tr("%1%2Button %3").arg(toggle, inverted, button);
+            return QObject::tr("%1%2%3Button %4").arg(turbo, toggle, inverted, button);
         }
     }
 
     QString button_name = GetButtonName(common_button_name);
     if (param.Has("hat")) {
-        return QObject::tr("%1%2Hat %3").arg(toggle, inverted, button_name);
+        return QObject::tr("%1%2%3Hat %4").arg(turbo, toggle, inverted, button_name);
     }
     if (param.Has("axis")) {
         return QObject::tr("%1%2Axis %3").arg(toggle, inverted, button_name);
@@ -234,7 +235,7 @@ QString ConfigureInputPlayer::ButtonToText(const Common::ParamPackage& param) {
         return QObject::tr("%1%2Axis %3").arg(toggle, inverted, button_name);
     }
     if (param.Has("button")) {
-        return QObject::tr("%1%2Button %3").arg(toggle, inverted, button_name);
+        return QObject::tr("%1%2%3Button %4").arg(turbo, toggle, inverted, button_name);
     }
 
     return QObject::tr("[unknown]");
@@ -395,6 +396,12 @@ ConfigureInputPlayer::ConfigureInputPlayer(QWidget* parent, std::size_t player_i
                             button_map[button_id]->setText(ButtonToText(param));
                             emulated_controller->SetButtonParam(button_id, param);
                         });
+                        context_menu.addAction(tr("Turbo button"), [&] {
+                            const bool turbo_value = !param.Get("turbo", false);
+                            param.Set("turbo", turbo_value);
+                            button_map[button_id]->setText(ButtonToText(param));
+                            emulated_controller->SetButtonParam(button_id, param);
+                        });
                     }
                     if (param.Has("axis")) {
                         context_menu.addAction(tr("Invert axis"), [&] {
diff --git a/src/yuzu/configuration/configure_input_player_widget.cpp b/src/yuzu/configuration/configure_input_player_widget.cpp
index 68af6c20c..c287220fc 100644
--- a/src/yuzu/configuration/configure_input_player_widget.cpp
+++ b/src/yuzu/configuration/configure_input_player_widget.cpp
@@ -81,7 +81,6 @@ void PlayerControlPreview::UpdateColors() {
         colors.outline = QColor(0, 0, 0);
         colors.primary = QColor(225, 225, 225);
         colors.button = QColor(109, 111, 114);
-        colors.button2 = QColor(109, 111, 114);
         colors.button2 = QColor(77, 80, 84);
         colors.slider_arrow = QColor(65, 68, 73);
         colors.font2 = QColor(0, 0, 0);
@@ -100,6 +99,7 @@ void PlayerControlPreview::UpdateColors() {
     colors.led_off = QColor(170, 238, 255);
     colors.indicator2 = QColor(59, 165, 93);
     colors.charging = QColor(250, 168, 26);
+    colors.button_turbo = QColor(217, 158, 4);
 
     colors.left = colors.primary;
     colors.right = colors.primary;
@@ -2469,7 +2469,6 @@ void PlayerControlPreview::DrawJoystickDot(QPainter& p, const QPointF center,
 void PlayerControlPreview::DrawRoundButton(QPainter& p, QPointF center,
                                            const Common::Input::ButtonStatus& pressed, float width,
                                            float height, Direction direction, float radius) {
-    p.setBrush(button_color);
     if (pressed.value) {
         switch (direction) {
         case Direction::Left:
@@ -2487,16 +2486,16 @@ void PlayerControlPreview::DrawRoundButton(QPainter& p, QPointF center,
         case Direction::None:
             break;
         }
-        p.setBrush(colors.highlight);
     }
     QRectF rect = {center.x() - width, center.y() - height, width * 2.0f, height * 2.0f};
+    p.setBrush(GetButtonColor(button_color, pressed.value, pressed.turbo));
     p.drawRoundedRect(rect, radius, radius);
 }
 void PlayerControlPreview::DrawMinusButton(QPainter& p, const QPointF center,
                                            const Common::Input::ButtonStatus& pressed,
                                            int button_size) {
     p.setPen(colors.outline);
-    p.setBrush(pressed.value ? colors.highlight : colors.button);
+    p.setBrush(GetButtonColor(colors.button, pressed.value, pressed.turbo));
     DrawRectangle(p, center, button_size, button_size / 3.0f);
 }
 void PlayerControlPreview::DrawPlusButton(QPainter& p, const QPointF center,
@@ -2504,7 +2503,7 @@ void PlayerControlPreview::DrawPlusButton(QPainter& p, const QPointF center,
                                           int button_size) {
     // Draw outer line
     p.setPen(colors.outline);
-    p.setBrush(pressed.value ? colors.highlight : colors.button);
+    p.setBrush(GetButtonColor(colors.button, pressed.value, pressed.turbo));
     DrawRectangle(p, center, button_size, button_size / 3.0f);
     DrawRectangle(p, center, button_size / 3.0f, button_size);
 
@@ -2526,7 +2525,7 @@ void PlayerControlPreview::DrawGCButtonX(QPainter& p, const QPointF center,
     }
 
     p.setPen(colors.outline);
-    p.setBrush(pressed.value ? colors.highlight : colors.button);
+    p.setBrush(GetButtonColor(colors.button, pressed.value, pressed.turbo));
     DrawPolygon(p, button_x);
 }
 
@@ -2539,7 +2538,7 @@ void PlayerControlPreview::DrawGCButtonY(QPainter& p, const QPointF center,
     }
 
     p.setPen(colors.outline);
-    p.setBrush(pressed.value ? colors.highlight : colors.button);
+    p.setBrush(GetButtonColor(colors.button, pressed.value, pressed.turbo));
     DrawPolygon(p, button_x);
 }
 
@@ -2553,17 +2552,15 @@ void PlayerControlPreview::DrawGCButtonZ(QPainter& p, const QPointF center,
     }
 
     p.setPen(colors.outline);
-    p.setBrush(pressed.value ? colors.highlight : colors.button2);
+    p.setBrush(GetButtonColor(colors.button2, pressed.value, pressed.turbo));
     DrawPolygon(p, button_x);
 }
 
 void PlayerControlPreview::DrawCircleButton(QPainter& p, const QPointF center,
                                             const Common::Input::ButtonStatus& pressed,
                                             float button_size) {
-    p.setBrush(button_color);
-    if (pressed.value) {
-        p.setBrush(colors.highlight);
-    }
+
+    p.setBrush(GetButtonColor(button_color, pressed.value, pressed.turbo));
     p.drawEllipse(center, button_size, button_size);
 }
 
@@ -2620,7 +2617,7 @@ void PlayerControlPreview::DrawArrowButton(QPainter& p, const QPointF center,
 
     // Draw arrow button
     p.setPen(pressed.value ? colors.highlight : colors.button);
-    p.setBrush(pressed.value ? colors.highlight : colors.button);
+    p.setBrush(GetButtonColor(colors.button, pressed.value, pressed.turbo));
     DrawPolygon(p, arrow_button);
 
     switch (direction) {
@@ -2672,10 +2669,20 @@ void PlayerControlPreview::DrawTriggerButton(QPainter& p, const QPointF center,
 
     // Draw arrow button
     p.setPen(colors.outline);
-    p.setBrush(pressed.value ? colors.highlight : colors.button);
+    p.setBrush(GetButtonColor(colors.button, pressed.value, pressed.turbo));
     DrawPolygon(p, qtrigger_button);
 }
 
+QColor PlayerControlPreview::GetButtonColor(QColor default_color, bool is_pressed, bool turbo) {
+    if (is_pressed && turbo) {
+        return colors.button_turbo;
+    }
+    if (is_pressed) {
+        return colors.highlight;
+    }
+    return default_color;
+}
+
 void PlayerControlPreview::DrawBattery(QPainter& p, QPointF center,
                                        Common::Input::BatteryLevel battery) {
     if (battery == Common::Input::BatteryLevel::None) {
diff --git a/src/yuzu/configuration/configure_input_player_widget.h b/src/yuzu/configuration/configure_input_player_widget.h
index b258c6d77..0e9e95e85 100644
--- a/src/yuzu/configuration/configure_input_player_widget.h
+++ b/src/yuzu/configuration/configure_input_player_widget.h
@@ -81,6 +81,7 @@ private:
         QColor right{};
         QColor button{};
         QColor button2{};
+        QColor button_turbo{};
         QColor font{};
         QColor font2{};
         QColor highlight{};
@@ -183,6 +184,7 @@ private:
                          const Common::Input::ButtonStatus& pressed, float size = 1.0f);
     void DrawTriggerButton(QPainter& p, QPointF center, Direction direction,
                            const Common::Input::ButtonStatus& pressed);
+    QColor GetButtonColor(QColor default_color, bool is_pressed, bool turbo);
 
     // Draw battery functions
     void DrawBattery(QPainter& p, QPointF center, Common::Input::BatteryLevel battery);
diff --git a/src/yuzu/configuration/input_profiles.cpp b/src/yuzu/configuration/input_profiles.cpp
index 9bb69cab1..41ef4250a 100644
--- a/src/yuzu/configuration/input_profiles.cpp
+++ b/src/yuzu/configuration/input_profiles.cpp
@@ -58,13 +58,16 @@ std::vector<std::string> InputProfiles::GetInputProfileNames() {
     std::vector<std::string> profile_names;
     profile_names.reserve(map_profiles.size());
 
-    for (const auto& [profile_name, config] : map_profiles) {
+    auto it = map_profiles.cbegin();
+    while (it != map_profiles.cend()) {
+        const auto& [profile_name, config] = *it;
         if (!ProfileExistsInFilesystem(profile_name)) {
-            DeleteProfile(profile_name);
+            it = map_profiles.erase(it);
             continue;
         }
 
         profile_names.push_back(profile_name);
+        ++it;
     }
 
     std::stable_sort(profile_names.begin(), profile_names.end());
diff --git a/src/yuzu/multiplayer/lobby.cpp b/src/yuzu/multiplayer/lobby.cpp
index 08c275696..6c93e3511 100644
--- a/src/yuzu/multiplayer/lobby.cpp
+++ b/src/yuzu/multiplayer/lobby.cpp
@@ -77,6 +77,7 @@ Lobby::Lobby(QWidget* parent, QStandardItemModel* list,
     // UI Buttons
     connect(ui->refresh_list, &QPushButton::clicked, this, &Lobby::RefreshLobby);
     connect(ui->games_owned, &QCheckBox::toggled, proxy, &LobbyFilterProxyModel::SetFilterOwned);
+    connect(ui->hide_empty, &QCheckBox::toggled, proxy, &LobbyFilterProxyModel::SetFilterEmpty);
     connect(ui->hide_full, &QCheckBox::toggled, proxy, &LobbyFilterProxyModel::SetFilterFull);
     connect(ui->search, &QLineEdit::textChanged, proxy, &LobbyFilterProxyModel::SetFilterSearch);
     connect(ui->room_list, &QTreeView::doubleClicked, this, &Lobby::OnJoinRoom);
@@ -329,6 +330,16 @@ bool LobbyFilterProxyModel::filterAcceptsRow(int sourceRow, const QModelIndex& s
         return true;
     }
 
+    // filter by empty rooms
+    if (filter_empty) {
+        QModelIndex member_list = sourceModel()->index(sourceRow, Column::MEMBER, sourceParent);
+        int player_count =
+            sourceModel()->data(member_list, LobbyItemMemberList::MemberListRole).toList().size();
+        if (player_count == 0) {
+            return false;
+        }
+    }
+
     // filter by filled rooms
     if (filter_full) {
         QModelIndex member_list = sourceModel()->index(sourceRow, Column::MEMBER, sourceParent);
@@ -399,6 +410,11 @@ void LobbyFilterProxyModel::SetFilterOwned(bool filter) {
     invalidate();
 }
 
+void LobbyFilterProxyModel::SetFilterEmpty(bool filter) {
+    filter_empty = filter;
+    invalidate();
+}
+
 void LobbyFilterProxyModel::SetFilterFull(bool filter) {
     filter_full = filter;
     invalidate();
diff --git a/src/yuzu/multiplayer/lobby.h b/src/yuzu/multiplayer/lobby.h
index 300dad13e..2674ae7c3 100644
--- a/src/yuzu/multiplayer/lobby.h
+++ b/src/yuzu/multiplayer/lobby.h
@@ -130,12 +130,14 @@ public:
 
 public slots:
     void SetFilterOwned(bool);
+    void SetFilterEmpty(bool);
     void SetFilterFull(bool);
     void SetFilterSearch(const QString&);
 
 private:
     QStandardItemModel* game_list;
     bool filter_owned = false;
+    bool filter_empty = false;
     bool filter_full = false;
     QString filter_search;
 };
diff --git a/src/yuzu/multiplayer/lobby.ui b/src/yuzu/multiplayer/lobby.ui
index 4c9901c9a..0ef0ef762 100644
--- a/src/yuzu/multiplayer/lobby.ui
+++ b/src/yuzu/multiplayer/lobby.ui
@@ -78,6 +78,13 @@
           </widget>
          </item>
          <item>
+          <widget class="QCheckBox" name="hide_empty">
+           <property name="text">
+            <string>Hide Empty Rooms</string>
+           </property>
+          </widget>
+         </item>
+         <item>
           <widget class="QCheckBox" name="hide_full">
            <property name="text">
             <string>Hide Full Rooms</string>
diff --git a/src/yuzu_cmd/config.cpp b/src/yuzu_cmd/config.cpp
index 527017282..9c34cdc6e 100644
--- a/src/yuzu_cmd/config.cpp
+++ b/src/yuzu_cmd/config.cpp
@@ -176,6 +176,9 @@ void Config::ReadValues() {
             Settings::values.debug_pad_analogs[i] = default_param;
     }
 
+    ReadSetting("ControlsGeneral", Settings::values.enable_raw_input);
+    ReadSetting("ControlsGeneral", Settings::values.enable_joycon_driver);
+    ReadSetting("ControlsGeneral", Settings::values.emulate_analog_keyboard);
     ReadSetting("ControlsGeneral", Settings::values.vibration_enabled);
     ReadSetting("ControlsGeneral", Settings::values.enable_accurate_vibrations);
     ReadSetting("ControlsGeneral", Settings::values.motion_enabled);
diff --git a/src/yuzu_cmd/default_ini.h b/src/yuzu_cmd/default_ini.h
index 67d230462..3f3651dbe 100644
--- a/src/yuzu_cmd/default_ini.h
+++ b/src/yuzu_cmd/default_ini.h
@@ -14,6 +14,7 @@ const char* sdl2_config_file =
 # Escape characters $0 (for ':'), $1 (for ',') and $2 (for '$') can be used in values
 
 # Indicates if this player should be connected at boot
+# 0 (default): Disabled, 1: Enabled
 connected=
 
 # for button input, the following devices are available:
@@ -94,6 +95,18 @@ motionright=
 # 0 (default): Disabled, 1: Enabled
 debug_pad_enabled =
 
+# Enable sdl raw input. Allows to configure up to 8 xinput controllers.
+# 0 (default): Disabled, 1: Enabled
+enable_raw_input =
+
+# Enable yuzu joycon driver instead of SDL drive.
+# 0: Disabled, 1 (default): Enabled
+enable_joycon_driver =
+
+# Emulates an analog input from buttons. Allowing to dial any angle.
+# 0 (default): Disabled, 1: Enabled
+emulate_analog_keyboard =
+
 # Whether to enable or disable vibration
 # 0: Disabled, 1 (default): Enabled
 vibration_enabled=
diff --git a/src/yuzu_cmd/emu_window/emu_window_sdl2.cpp b/src/yuzu_cmd/emu_window/emu_window_sdl2.cpp
index 31f28a507..5450b8c38 100644
--- a/src/yuzu_cmd/emu_window/emu_window_sdl2.cpp
+++ b/src/yuzu_cmd/emu_window/emu_window_sdl2.cpp
@@ -18,11 +18,11 @@
 
 EmuWindow_SDL2::EmuWindow_SDL2(InputCommon::InputSubsystem* input_subsystem_, Core::System& system_)
     : input_subsystem{input_subsystem_}, system{system_} {
-    if (SDL_Init(SDL_INIT_VIDEO | SDL_INIT_JOYSTICK) < 0) {
+    input_subsystem->Initialize();
+    if (SDL_Init(SDL_INIT_VIDEO | SDL_INIT_JOYSTICK | SDL_INIT_GAMECONTROLLER) < 0) {
         LOG_CRITICAL(Frontend, "Failed to initialize SDL2! Exiting...");
         exit(1);
     }
-    input_subsystem->Initialize();
     SDL_SetMainReady();
 }
 
@@ -32,10 +32,6 @@ EmuWindow_SDL2::~EmuWindow_SDL2() {
     SDL_Quit();
 }
 
-void EmuWindow_SDL2::OnMouseMotion(s32 x, s32 y) {
-    input_subsystem->GetMouse()->MouseMove(x, y, 0, 0, 0, 0);
-}
-
 InputCommon::MouseButton EmuWindow_SDL2::SDLButtonToMouseButton(u32 button) const {
     switch (button) {
     case SDL_BUTTON_LEFT:
@@ -53,44 +49,36 @@ InputCommon::MouseButton EmuWindow_SDL2::SDLButtonToMouseButton(u32 button) cons
     }
 }
 
+std::pair<float, float> EmuWindow_SDL2::MouseToTouchPos(s32 touch_x, s32 touch_y) const {
+    int w, h;
+    SDL_GetWindowSize(render_window, &w, &h);
+    const float fx = static_cast<float>(touch_x) / w;
+    const float fy = static_cast<float>(touch_y) / h;
+
+    return {std::clamp<float>(fx, 0.0f, 1.0f), std::clamp<float>(fy, 0.0f, 1.0f)};
+}
+
 void EmuWindow_SDL2::OnMouseButton(u32 button, u8 state, s32 x, s32 y) {
     const auto mouse_button = SDLButtonToMouseButton(button);
     if (state == SDL_PRESSED) {
-        input_subsystem->GetMouse()->PressButton(x, y, 0, 0, mouse_button);
+        const auto [touch_x, touch_y] = MouseToTouchPos(x, y);
+        input_subsystem->GetMouse()->PressButton(x, y, touch_x, touch_y, mouse_button);
     } else {
         input_subsystem->GetMouse()->ReleaseButton(mouse_button);
     }
 }
 
-std::pair<unsigned, unsigned> EmuWindow_SDL2::TouchToPixelPos(float touch_x, float touch_y) const {
-    int w, h;
-    SDL_GetWindowSize(render_window, &w, &h);
-
-    touch_x *= w;
-    touch_y *= h;
-
-    return {static_cast<unsigned>(std::max(std::round(touch_x), 0.0f)),
-            static_cast<unsigned>(std::max(std::round(touch_y), 0.0f))};
+void EmuWindow_SDL2::OnMouseMotion(s32 x, s32 y) {
+    const auto [touch_x, touch_y] = MouseToTouchPos(x, y);
+    input_subsystem->GetMouse()->MouseMove(x, y, touch_x, touch_y, 0, 0);
 }
 
 void EmuWindow_SDL2::OnFingerDown(float x, float y, std::size_t id) {
-    int width, height;
-    SDL_GetWindowSize(render_window, &width, &height);
-    const auto [px, py] = TouchToPixelPos(x, y);
-    const float fx = px * 1.0f / width;
-    const float fy = py * 1.0f / height;
-
-    input_subsystem->GetTouchScreen()->TouchPressed(fx, fy, id);
+    input_subsystem->GetTouchScreen()->TouchPressed(x, y, id);
 }
 
 void EmuWindow_SDL2::OnFingerMotion(float x, float y, std::size_t id) {
-    int width, height;
-    SDL_GetWindowSize(render_window, &width, &height);
-    const auto [px, py] = TouchToPixelPos(x, y);
-    const float fx = px * 1.0f / width;
-    const float fy = py * 1.0f / height;
-
-    input_subsystem->GetTouchScreen()->TouchMoved(fx, fy, id);
+    input_subsystem->GetTouchScreen()->TouchMoved(x, y, id);
 }
 
 void EmuWindow_SDL2::OnFingerUp() {
diff --git a/src/yuzu_cmd/emu_window/emu_window_sdl2.h b/src/yuzu_cmd/emu_window/emu_window_sdl2.h
index 25c23e2a5..d9b453dee 100644
--- a/src/yuzu_cmd/emu_window/emu_window_sdl2.h
+++ b/src/yuzu_cmd/emu_window/emu_window_sdl2.h
@@ -38,17 +38,17 @@ protected:
     /// Called by WaitEvent when a key is pressed or released.
     void OnKeyEvent(int key, u8 state);
 
-    /// Called by WaitEvent when the mouse moves.
-    void OnMouseMotion(s32 x, s32 y);
-
     /// Converts a SDL mouse button into MouseInput mouse button
     InputCommon::MouseButton SDLButtonToMouseButton(u32 button) const;
 
+    /// Translates pixel position to float position
+    std::pair<float, float> MouseToTouchPos(s32 touch_x, s32 touch_y) const;
+
     /// Called by WaitEvent when a mouse button is pressed or released
     void OnMouseButton(u32 button, u8 state, s32 x, s32 y);
 
-    /// Translates pixel position (0..1) to pixel positions
-    std::pair<unsigned, unsigned> TouchToPixelPos(float touch_x, float touch_y) const;
+    /// Called by WaitEvent when the mouse moves.
+    void OnMouseMotion(s32 x, s32 y);
 
     /// Called by WaitEvent when a finger starts touching the touchscreen
     void OnFingerDown(float x, float y, std::size_t id);
diff --git a/src/yuzu_cmd/yuzu.cpp b/src/yuzu_cmd/yuzu.cpp
index 91133569d..d1f7b1d49 100644
--- a/src/yuzu_cmd/yuzu.cpp
+++ b/src/yuzu_cmd/yuzu.cpp
@@ -62,13 +62,15 @@ __declspec(dllexport) int AmdPowerXpressRequestHighPerformance = 1;
 static void PrintHelp(const char* argv0) {
     std::cout << "Usage: " << argv0
               << " [options] <filename>\n"
-                 "-m, --multiplayer=nick:password@address:port"
-                 " Nickname, password, address and port for multiplayer\n"
+                 "-c, --config          Load the specified configuration file\n"
                  "-f, --fullscreen      Start in fullscreen mode\n"
+                 "-g, --game            File path of the game to load\n"
                  "-h, --help            Display this help and exit\n"
-                 "-v, --version         Output version information and exit\n"
+                 "-m, --multiplayer=nick:password@address:port"
+                 " Nickname, password, address and port for multiplayer\n"
                  "-p, --program         Pass following string as arguments to executable\n"
-                 "-c, --config          Load the specified configuration file\n";
+                 "-u, --user            Select a specific user profile from 0 to 7\n"
+                 "-v, --version         Output version information and exit\n";
 }
 
 static void PrintVersion() {
@@ -199,6 +201,7 @@ int main(int argc, char** argv) {
     std::string filepath;
     std::optional<std::string> config_path;
     std::string program_args;
+    std::optional<int> selected_user;
 
     bool use_multiplayer = false;
     bool fullscreen = false;
@@ -209,12 +212,14 @@ int main(int argc, char** argv) {
 
     static struct option long_options[] = {
         // clang-format off
-        {"multiplayer", required_argument, 0, 'm'},
+        {"config", required_argument, 0, 'c'},
         {"fullscreen", no_argument, 0, 'f'},
         {"help", no_argument, 0, 'h'},
-        {"version", no_argument, 0, 'v'},
+        {"game", required_argument, 0, 'g'},
+        {"multiplayer", required_argument, 0, 'm'},
         {"program", optional_argument, 0, 'p'},
-        {"config", required_argument, 0, 'c'},
+        {"user", required_argument, 0, 'u'},
+        {"version", no_argument, 0, 'v'},
         {0, 0, 0, 0},
         // clang-format on
     };
@@ -223,6 +228,21 @@ int main(int argc, char** argv) {
         int arg = getopt_long(argc, argv, "g:fhvp::c:", long_options, &option_index);
         if (arg != -1) {
             switch (static_cast<char>(arg)) {
+            case 'c':
+                config_path = optarg;
+                break;
+            case 'f':
+                fullscreen = true;
+                LOG_INFO(Frontend, "Starting in fullscreen mode...");
+                break;
+            case 'h':
+                PrintHelp(argv[0]);
+                return 0;
+            case 'g': {
+                const std::string str_arg(optarg);
+                filepath = str_arg;
+                break;
+            }
             case 'm': {
                 use_multiplayer = true;
                 const std::string str_arg(optarg);
@@ -255,23 +275,16 @@ int main(int argc, char** argv) {
                 }
                 break;
             }
-            case 'f':
-                fullscreen = true;
-                LOG_INFO(Frontend, "Starting in fullscreen mode...");
+            case 'p':
+                program_args = argv[optind];
+                ++optind;
                 break;
-            case 'h':
-                PrintHelp(argv[0]);
+            case 'u':
+                selected_user = atoi(optarg);
                 return 0;
             case 'v':
                 PrintVersion();
                 return 0;
-            case 'p':
-                program_args = argv[optind];
-                ++optind;
-                break;
-            case 'c':
-                config_path = optarg;
-                break;
             }
         } else {
 #ifdef _WIN32
@@ -295,6 +308,10 @@ int main(int argc, char** argv) {
         Settings::values.program_args = program_args;
     }
 
+    if (selected_user.has_value()) {
+        Settings::values.current_user = std::clamp(*selected_user, 0, 7);
+    }
+
 #ifdef _WIN32
     LocalFree(argv_w);
 #endif