diff options
Diffstat (limited to 'src')
54 files changed, 653 insertions, 488 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/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/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/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_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/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..0ef240344 100644 --- a/src/input_common/helpers/joycon_protocol/common_protocol.cpp +++ b/src/input_common/helpers/joycon_protocol/common_protocol.cpp @@ -22,8 +22,8 @@ void JoyconCommonProtocol::SetNonBlocking() { } DriverResult JoyconCommonProtocol::GetDeviceType(ControllerType& controller_type) { - std::vector<u8> buffer; - const auto result = ReadSPI(CalAddr::DEVICE_TYPE, 1, buffer); + std::array<u8, 1> buffer{}; + const auto result = ReadRawSPI(SpiAddress::DEVICE_TYPE, buffer); controller_type = ControllerType::None; if (result == DriverResult::Success) { @@ -148,11 +148,13 @@ DriverResult JoyconCommonProtocol::SendVibrationReport(std::span<const u8> buffe return SendData(local_buffer); } -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 = 20; constexpr std::size_t MaxTries = 10; + const auto size = output.size(); std::size_t tries = 0; - std::array<u8, 5> buffer = {0x00, 0x00, 0x00, 0x00, size}; - std::vector<u8> local_buffer(size + 20); + std::array<u8, 5> buffer = {0x00, 0x00, 0x00, 0x00, static_cast<u8>(size)}; + std::vector<u8> local_buffer{}; buffer[0] = static_cast<u8>(static_cast<u16>(addr) & 0x00FF); buffer[1] = static_cast<u8>((static_cast<u16>(addr) & 0xFF00) >> 8); @@ -167,8 +169,12 @@ DriverResult JoyconCommonProtocol::ReadSPI(CalAddr addr, u8 size, std::vector<u8 } } while (local_buffer[15] != buffer[0] || local_buffer[16] != buffer[1]); + if (local_buffer.size() < 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(), local_buffer.data() + HeaderSize, size); 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..188f6ecfa 100644 --- a/src/input_common/helpers/joycon_protocol/common_protocol.h +++ b/src/input_common/helpers/joycon_protocol/common_protocol.h @@ -97,10 +97,29 @@ 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 */ - 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 responce + */ + 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 diff --git a/src/input_common/helpers/joycon_protocol/generic_functions.cpp b/src/input_common/helpers/joycon_protocol/generic_functions.cpp index 63cfb1369..484c208e6 100644 --- a/src/input_common/helpers/joycon_protocol/generic_functions.cpp +++ b/src/input_common/helpers/joycon_protocol/generic_functions.cpp @@ -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/joycon_types.h b/src/input_common/helpers/joycon_protocol/joycon_types.h index 182d2c15b..14b07bfb5 100644 --- a/src/input_common/helpers/joycon_protocol/joycon_types.h +++ b/src/input_common/helpers/joycon_protocol/joycon_types.h @@ -159,13 +159,12 @@ 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 { +enum class SpiAddress { SERIAL_NUMBER = 0X6000, DEVICE_TYPE = 0X6012, COLOR_EXIST = 0X601B, @@ -396,10 +395,35 @@ 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 +431,14 @@ 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; @@ -488,14 +520,6 @@ 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; 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_bitwise_conversion.cpp b/src/shader_recompiler/backend/glasm/emit_glasm_bitwise_conversion.cpp index 5bfdecc09..2fc2a0ac6 100644 --- a/src/shader_recompiler/backend/glasm/emit_glasm_bitwise_conversion.cpp +++ b/src/shader_recompiler/backend/glasm/emit_glasm_bitwise_conversion.cpp @@ -43,10 +43,6 @@ void EmitBitCastU64F64(EmitContext&, IR::Inst& inst, const IR::Value& value) { Alias(inst, value); } -void EmitBitCastS32F32(EmitContext&, IR::Inst& inst, const IR::Value& value) { - Alias(inst, value); -} - void EmitBitCastF16U16(EmitContext&, IR::Inst& inst, const IR::Value& value) { Alias(inst, value); } diff --git a/src/shader_recompiler/backend/glasm/emit_glasm_image.cpp b/src/shader_recompiler/backend/glasm/emit_glasm_image.cpp index e67e80fac..b7bc11416 100644 --- a/src/shader_recompiler/backend/glasm/emit_glasm_image.cpp +++ b/src/shader_recompiler/backend/glasm/emit_glasm_image.cpp @@ -59,7 +59,7 @@ std::string Image(EmitContext& ctx, IR::TextureInstInfo info, } } -std::string_view TextureType(IR::TextureInstInfo info) { +std::string_view TextureType(IR::TextureInstInfo info, bool is_ms = false) { if (info.is_depth) { switch (info.type) { case TextureType::Color1D: @@ -88,9 +88,9 @@ 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 +510,16 @@ 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); @@ -531,7 +532,7 @@ 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)}; diff --git a/src/shader_recompiler/backend/glasm/emit_glasm_instructions.h b/src/shader_recompiler/backend/glasm/emit_glasm_instructions.h index eaaf9ba39..1a1ea61d5 100644 --- a/src/shader_recompiler/backend/glasm/emit_glasm_instructions.h +++ b/src/shader_recompiler/backend/glasm/emit_glasm_instructions.h @@ -197,7 +197,6 @@ void EmitSelectF64(EmitContext& ctx, ScalarS32 cond, Register true_value, Regist void EmitBitCastU16F16(EmitContext& ctx, IR::Inst& inst, const IR::Value& value); void EmitBitCastU32F32(EmitContext& ctx, IR::Inst& inst, const IR::Value& value); void EmitBitCastU64F64(EmitContext& ctx, IR::Inst& inst, const IR::Value& value); -void EmitBitCastS32F32(EmitContext& ctx, IR::Inst& inst, const IR::Value& value); void EmitBitCastF16U16(EmitContext& ctx, IR::Inst& inst, const IR::Value& value); void EmitBitCastF32U32(EmitContext& ctx, IR::Inst& inst, const IR::Value& value); void EmitBitCastF64U64(EmitContext& ctx, IR::Inst& inst, const IR::Value& value); @@ -582,7 +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_bitwise_conversion.cpp b/src/shader_recompiler/backend/glsl/emit_glsl_bitwise_conversion.cpp index 8e5e6cf1f..1be4a0f59 100644 --- a/src/shader_recompiler/backend/glsl/emit_glsl_bitwise_conversion.cpp +++ b/src/shader_recompiler/backend/glsl/emit_glsl_bitwise_conversion.cpp @@ -48,10 +48,6 @@ void EmitBitCastU64F64(EmitContext& ctx, IR::Inst& inst, std::string_view value) ctx.AddU64("{}=doubleBitsToUint64({});", inst, value); } -void EmitBitCastS32F32(EmitContext& ctx, IR::Inst& inst, std::string_view value) { - ctx.AddF32("{}=ftoi({});", inst, value); -} - void EmitBitCastF16U16([[maybe_unused]] EmitContext& ctx, [[maybe_unused]] IR::Inst& inst) { NotImplemented(); } diff --git a/src/shader_recompiler/backend/glsl/emit_glsl_image.cpp b/src/shader_recompiler/backend/glsl/emit_glsl_image.cpp index cecdbb9d6..4be2c25ec 100644 --- a/src/shader_recompiler/backend/glsl/emit_glsl_image.cpp +++ b/src/shader_recompiler/backend/glsl/emit_glsl_image.cpp @@ -414,7 +414,7 @@ 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 +431,24 @@ 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, @@ -455,27 +460,27 @@ 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 skip_mips{skip_mips_val.U1()}; + const auto mips{ + [&] { return skip_mips ? "0u" : fmt::format("uint(textureQueryLevels({}))", texture); }}; 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({},int({}))),0u,0u,{});", inst, texture, lod, + 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({},int({}))),0u,{});", inst, texture, lod, + 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({},int({}))),{});", inst, texture, lod, + mips()); case TextureType::Buffer: throw NotImplementedException("EmitImageQueryDimensions Texture buffers"); } diff --git a/src/shader_recompiler/backend/glsl/emit_glsl_instructions.h b/src/shader_recompiler/backend/glsl/emit_glsl_instructions.h index 4151c89de..8d0a65047 100644 --- a/src/shader_recompiler/backend/glsl/emit_glsl_instructions.h +++ b/src/shader_recompiler/backend/glsl/emit_glsl_instructions.h @@ -231,7 +231,6 @@ void EmitSelectF64(EmitContext& ctx, IR::Inst& inst, std::string_view cond, void EmitBitCastU16F16(EmitContext& ctx, IR::Inst& inst); void EmitBitCastU32F32(EmitContext& ctx, IR::Inst& inst, std::string_view value); void EmitBitCastU64F64(EmitContext& ctx, IR::Inst& inst, std::string_view value); -void EmitBitCastS32F32(EmitContext& ctx, IR::Inst& inst, std::string_view value); void EmitBitCastF16U16(EmitContext& ctx, IR::Inst& inst); void EmitBitCastF32U32(EmitContext& ctx, IR::Inst& inst, std::string_view value); void EmitBitCastF64U64(EmitContext& ctx, IR::Inst& inst, std::string_view value); @@ -655,7 +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_bitwise_conversion.cpp b/src/shader_recompiler/backend/spirv/emit_spirv_bitwise_conversion.cpp index 50daacd95..c4ca28d11 100644 --- a/src/shader_recompiler/backend/spirv/emit_spirv_bitwise_conversion.cpp +++ b/src/shader_recompiler/backend/spirv/emit_spirv_bitwise_conversion.cpp @@ -18,10 +18,6 @@ void EmitBitCastU64F64(EmitContext&) { throw NotImplementedException("SPIR-V Instruction"); } -void EmitBitCastS32F32(EmitContext&) { - throw NotImplementedException("SPIR-V Instruction"); -} - void EmitBitCastF16U16(EmitContext&) { throw NotImplementedException("SPIR-V Instruction"); } diff --git a/src/shader_recompiler/backend/spirv/emit_spirv_image.cpp b/src/shader_recompiler/backend/spirv/emit_spirv_image.cpp index c898ce12f..3b969d915 100644 --- a/src/shader_recompiler/backend/spirv/emit_spirv_image.cpp +++ b/src/shader_recompiler/backend/spirv/emit_spirv_image.cpp @@ -445,11 +445,13 @@ 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); }}; switch (info.type) { case TextureType::Color1D: return ctx.OpCompositeConstruct(ctx.U32[4], ctx.OpImageQuerySizeLod(ctx.U32[1], image, lod), diff --git a/src/shader_recompiler/backend/spirv/emit_spirv_instructions.h b/src/shader_recompiler/backend/spirv/emit_spirv_instructions.h index e31cdc5e8..a440b557d 100644 --- a/src/shader_recompiler/backend/spirv/emit_spirv_instructions.h +++ b/src/shader_recompiler/backend/spirv/emit_spirv_instructions.h @@ -179,7 +179,6 @@ Id EmitSelectF64(EmitContext& ctx, Id cond, Id true_value, Id false_value); void EmitBitCastU16F16(EmitContext& ctx); Id EmitBitCastU32F32(EmitContext& ctx, Id value); void EmitBitCastU64F64(EmitContext& ctx); -void EmitBitCastS32F32(EmitContext& ctx); void EmitBitCastF16U16(EmitContext&); Id EmitBitCastF32U32(EmitContext& ctx, Id value); void EmitBitCastF64U64(EmitContext& ctx); @@ -540,7 +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/frontend/ir/ir_emitter.cpp b/src/shader_recompiler/frontend/ir/ir_emitter.cpp index eb2e49a68..b7caa4246 100644 --- a/src/shader_recompiler/frontend/ir/ir_emitter.cpp +++ b/src/shader_recompiler/frontend/ir/ir_emitter.cpp @@ -704,11 +704,6 @@ IR::U32 IREmitter::BitCast<IR::U32, IR::F32>(const IR::F32& value) { } template <> -IR::S32 IREmitter::BitCast<IR::S32, IR::F32>(const IR::F32& value) { - return Inst<IR::S32>(Opcode::BitCastS32F32, value); -} - -template <> IR::F32 IREmitter::BitCast<IR::F32, IR::U32>(const IR::U32& value) { return Inst<IR::F32>(Opcode::BitCastF32U32, value); } @@ -1851,15 +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.h b/src/shader_recompiler/frontend/ir/opcodes.h index d155afd0f..e300714f3 100644 --- a/src/shader_recompiler/frontend/ir/opcodes.h +++ b/src/shader_recompiler/frontend/ir/opcodes.h @@ -38,7 +38,6 @@ constexpr Type U8{Type::U8}; constexpr Type U16{Type::U16}; constexpr Type U32{Type::U32}; constexpr Type U64{Type::U64}; -constexpr Type S32{Type::S32}; constexpr Type F16{Type::F16}; constexpr Type F32{Type::F32}; constexpr Type F64{Type::F64}; diff --git a/src/shader_recompiler/frontend/ir/opcodes.inc b/src/shader_recompiler/frontend/ir/opcodes.inc index 1fe3749cc..4447d67b0 100644 --- a/src/shader_recompiler/frontend/ir/opcodes.inc +++ b/src/shader_recompiler/frontend/ir/opcodes.inc @@ -175,7 +175,6 @@ OPCODE(SelectF64, F64, U1, OPCODE(BitCastU16F16, U16, F16, ) OPCODE(BitCastU32F32, U32, F32, ) OPCODE(BitCastU64F64, U64, F64, ) -OPCODE(BitCastS32F32, S32, F32, ) OPCODE(BitCastF16U16, F16, U16, ) OPCODE(BitCastF32U32, F32, U32, ) OPCODE(BitCastF64U64, F64, U64, ) @@ -483,7 +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, ) @@ -496,7 +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, ) @@ -509,7 +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/type.h b/src/shader_recompiler/frontend/ir/type.h index 5a7c706ad..04c8c4ddb 100644 --- a/src/shader_recompiler/frontend/ir/type.h +++ b/src/shader_recompiler/frontend/ir/type.h @@ -24,22 +24,21 @@ enum class Type { U16 = 1 << 7, U32 = 1 << 8, U64 = 1 << 9, - S32 = 1 << 10, - F16 = 1 << 11, - F32 = 1 << 12, - F64 = 1 << 13, - U32x2 = 1 << 14, - U32x3 = 1 << 15, - U32x4 = 1 << 16, - F16x2 = 1 << 17, - F16x3 = 1 << 18, - F16x4 = 1 << 19, - F32x2 = 1 << 20, - F32x3 = 1 << 21, - F32x4 = 1 << 22, - F64x2 = 1 << 23, - F64x3 = 1 << 24, - F64x4 = 1 << 25, + F16 = 1 << 10, + F32 = 1 << 11, + F64 = 1 << 12, + U32x2 = 1 << 13, + U32x3 = 1 << 14, + U32x4 = 1 << 15, + F16x2 = 1 << 16, + F16x3 = 1 << 17, + F16x4 = 1 << 18, + F32x2 = 1 << 19, + F32x3 = 1 << 20, + F32x4 = 1 << 21, + F64x2 = 1 << 22, + F64x3 = 1 << 23, + F64x4 = 1 << 24, }; DECLARE_ENUM_FLAG_OPERATORS(Type) diff --git a/src/shader_recompiler/frontend/ir/value.cpp b/src/shader_recompiler/frontend/ir/value.cpp index 30ba12316..346169328 100644 --- a/src/shader_recompiler/frontend/ir/value.cpp +++ b/src/shader_recompiler/frontend/ir/value.cpp @@ -23,8 +23,6 @@ Value::Value(u16 value) noexcept : type{Type::U16}, imm_u16{value} {} Value::Value(u32 value) noexcept : type{Type::U32}, imm_u32{value} {} -Value::Value(s32 value) noexcept : type{Type::S32}, imm_s32{value} {} - Value::Value(f32 value) noexcept : type{Type::F32}, imm_f32{value} {} Value::Value(u64 value) noexcept : type{Type::U64}, imm_u64{value} {} @@ -71,7 +69,6 @@ bool Value::operator==(const Value& other) const { return imm_u16 == other.imm_u16; case Type::U32: case Type::F32: - case Type::S32: return imm_u32 == other.imm_u32; case Type::U64: case Type::F64: diff --git a/src/shader_recompiler/frontend/ir/value.h b/src/shader_recompiler/frontend/ir/value.h index 8b34356fd..22e89dd1b 100644 --- a/src/shader_recompiler/frontend/ir/value.h +++ b/src/shader_recompiler/frontend/ir/value.h @@ -101,9 +101,8 @@ 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 +193,16 @@ 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)); } @@ -268,7 +267,6 @@ using U8 = TypedValue<Type::U8>; using U16 = TypedValue<Type::U16>; using U32 = TypedValue<Type::U32>; using U64 = TypedValue<Type::U64>; -using S32 = TypedValue<Type::S32>; using F16 = TypedValue<Type::F16>; using F32 = TypedValue<Type::F32>; using F64 = TypedValue<Type::F64>; 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 9718c6921..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)), @@ -486,10 +494,10 @@ void PatchTexelFetch(IR::Block& block, IR::Inst& inst, TexturePixelFormat pixel_ const IR::F32 w(ir.CompositeExtract(new_inst, 3)); const IR::F16F32F64 max_value(ir.Imm32(get_max_value())); const IR::Value converted = - ir.CompositeConstruct(ir.FPMul(ir.ConvertSToF(32, 32, ir.BitCast<IR::S32>(x)), max_value), - ir.FPMul(ir.ConvertSToF(32, 32, ir.BitCast<IR::S32>(y)), max_value), - ir.FPMul(ir.ConvertSToF(32, 32, ir.BitCast<IR::S32>(z)), max_value), - ir.FPMul(ir.ConvertSToF(32, 32, ir.BitCast<IR::S32>(w)), max_value)); + ir.CompositeConstruct(ir.FPMul(ir.ConvertSToF(32, 32, ir.BitCast<IR::U32>(x)), max_value), + ir.FPMul(ir.ConvertSToF(32, 32, ir.BitCast<IR::U32>(y)), max_value), + ir.FPMul(ir.ConvertSToF(32, 32, ir.BitCast<IR::U32>(z)), max_value), + ir.FPMul(ir.ConvertSToF(32, 32, ir.BitCast<IR::U32>(w)), max_value)); inst.ReplaceUsesWith(converted); } } // Anonymous namespace 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/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/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/direct_connect.cpp b/src/yuzu/multiplayer/direct_connect.cpp index cbd52da85..d71cc23a7 100644 --- a/src/yuzu/multiplayer/direct_connect.cpp +++ b/src/yuzu/multiplayer/direct_connect.cpp @@ -81,20 +81,13 @@ void DirectConnectWindow::Connect() { } } } - switch (static_cast<ConnectionType>(ui->connection_type->currentIndex())) { - case ConnectionType::TraversalServer: - break; - case ConnectionType::IP: - if (!ui->ip->hasAcceptableInput()) { - NetworkMessage::ErrorManager::ShowError( - NetworkMessage::ErrorManager::IP_ADDRESS_NOT_VALID); - return; - } - if (!ui->port->hasAcceptableInput()) { - NetworkMessage::ErrorManager::ShowError(NetworkMessage::ErrorManager::PORT_NOT_VALID); - return; - } - break; + if (!ui->ip->hasAcceptableInput()) { + NetworkMessage::ErrorManager::ShowError(NetworkMessage::ErrorManager::IP_ADDRESS_NOT_VALID); + return; + } + if (!ui->port->hasAcceptableInput()) { + NetworkMessage::ErrorManager::ShowError(NetworkMessage::ErrorManager::PORT_NOT_VALID); + return; } // Store settings diff --git a/src/yuzu/multiplayer/direct_connect.ui b/src/yuzu/multiplayer/direct_connect.ui index 57d6ec25a..0dd4e6829 100644 --- a/src/yuzu/multiplayer/direct_connect.ui +++ b/src/yuzu/multiplayer/direct_connect.ui @@ -27,19 +27,10 @@ <number>0</number> </property> <item> - <widget class="QComboBox" name="connection_type"> - <item> - <property name="text"> - <string>IP Address</string> - </property> - </item> - </widget> - </item> - <item> <widget class="QWidget" name="ip_container" native="true"> <layout class="QHBoxLayout" name="ip_layout"> <property name="leftMargin"> - <number>5</number> + <number>0</number> </property> <property name="topMargin"> <number>0</number> @@ -53,17 +44,17 @@ <item> <widget class="QLabel" name="label_2"> <property name="text"> - <string>IP</string> + <string>Server Address</string> </property> </widget> </item> <item> <widget class="QLineEdit" name="ip"> <property name="toolTip"> - <string><html><head/><body><p>IPv4 address of the host</p></body></html></string> + <string><html><head/><body><p>Server address of the host</p></body></html></string> </property> <property name="maxLength"> - <number>16</number> + <number>253</number> </property> </widget> </item> @@ -85,6 +76,12 @@ <property name="placeholderText"> <string notr="true" extracomment="placeholder string that tells user default port">24872</string> </property> + <property name="maximumSize"> + <size> + <width>65</width> + <height>50</height> + </size> + </property> </widget> </item> </layout> diff --git a/src/yuzu/multiplayer/validation.h b/src/yuzu/multiplayer/validation.h index dd25af280..cbbe6757b 100644 --- a/src/yuzu/multiplayer/validation.h +++ b/src/yuzu/multiplayer/validation.h @@ -38,11 +38,28 @@ private: QRegularExpression(QStringLiteral("^[a-zA-Z0-9._ -]{4,20}")); QRegularExpressionValidator nickname; - /// ipv4 address only - // TODO remove this when we support hostnames in direct connect + /// ipv4 / ipv6 / hostnames QRegularExpression ip_regex = QRegularExpression(QStringLiteral( - "(([0-9]|[1-9][0-9]|1[0-9]{2}|2[0-4][0-9]|25[0-5])\\.){3}([0-9]|[1-9][0-9]|1[0-9]{2}|" - "2[0-4][0-9]|25[0-5])")); + // IPv4 regex + "^((25[0-5]|2[0-4][0-9]|[01]?[0-9][0-9]?)\\.){3}(25[0-5]|2[0-4][0-9]|[01]?[0-9][0-9]?)$|" + // IPv6 regex + "^((([0-9A-Fa-f]{1,4}:){7}([0-9A-Fa-f]{1,4}|:))|" + "(([0-9A-Fa-f]{1,4}:){6}(:[0-9A-Fa-f]{1,4}|((25[0-5]|2[0-4]\\d|1\\d\\d|[1-9]?\\d)(\\.(25[0-" + "5]|2[0-4]\\d|1\\d\\d|[1-9]?\\d)){3})|:))|" + "(([0-9A-Fa-f]{1,4}:){5}(((:[0-9A-Fa-f]{1,4}){1,2})|:((25[0-5]|2[0-4]\\d|1\\d\\d|[1-9]?\\d)" + "(\\.(25[0-5]|2[0-4]\\d|1\\d\\d|[1-9]?\\d)){3})|:))|" + "(([0-9A-Fa-f]{1,4}:){4}(((:[0-9A-Fa-f]{1,4}){1,3})|((:[0-9A-Fa-f]{1,4})?:((25[0-5]|2[0-4]" + "\\d|1\\d\\d|[1-9]?\\d)(\\.(25[0-5]|2[0-4]\\d|1\\d\\d|[1-9]?\\d)){3}))|:))|" + "(([0-9A-Fa-f]{1,4}:){3}(((:[0-9A-Fa-f]{1,4}){1,4})|((:[0-9A-Fa-f]{1,4}){0,2}:((25[0-5]|2[" + "0-4]\\d|1\\d\\d|[1-9]?\\d)(\\.(25[0-5]|2[0-4]\\d|1\\d\\d|[1-9]?\\d)){3}))|:))|" + "(([0-9A-Fa-f]{1,4}:){2}(((:[0-9A-Fa-f]{1,4}){1,5})|((:[0-9A-Fa-f]{1,4}){0,3}:((25[0-5]|2[" + "0-4]\\d|1\\d\\d|[1-9]?\\d)(\\.(25[0-5]|2[0-4]\\d|1\\d\\d|[1-9]?\\d)){3}))|:))|" + "(([0-9A-Fa-f]{1,4}:){1}(((:[0-9A-Fa-f]{1,4}){1,6})|((:[0-9A-Fa-f]{1,4}){0,4}:((25[0-5]|2[" + "0-4]\\d|1\\d\\d|[1-9]?\\d)(\\.(25[0-5]|2[0-4]\\d|1\\d\\d|[1-9]?\\d)){3}))|:))|" + "(:(((:[0-9A-Fa-f]{1,4}){1,7})|((:[0-9A-Fa-f]{1,4}){0,5}:((25[0-5]|2[0-4]\\d|1\\d\\d|[1-9]?" + "\\d)(\\.(25[0-5]|2[0-4]\\d|1\\d\\d|[1-9]?\\d)){3}))|:)))(%.+)?$|" + // Hostname regex + "^([a-zA-Z0-9]+(-[a-zA-Z0-9]+)*\\.)+[a-zA-Z]{2,}$")); QRegularExpressionValidator ip; /// port must be between 0 and 65535 |