Remove OnCommandListEndCommand

Call rasterizer->ReleaseFences() directly

src/audio_core/renderer/command/effect/biquad_filter.cpp

@@ -4,7 +4,6 @@
 #include "audio_core/renderer/adsp/command_list_processor.h"
 #include "audio_core/renderer/command/effect/biquad_filter.h"
 #include "audio_core/renderer/voice/voice_state.h"
-#include "common/bit_cast.h"
 
 namespace AudioCore::AudioRenderer {
 /**
@@ -20,21 +19,21 @@ namespace AudioCore::AudioRenderer {
 void ApplyBiquadFilterFloat(std::span<s32> output, std::span<const s32> input,
                             std::array<s16, 3>& b_, std::array<s16, 2>& a_,
                             VoiceState::BiquadFilterState& state, const u32 sample_count) {
-    constexpr f64 min{std::numeric_limits<s32>::min()};
-    constexpr f64 max{std::numeric_limits<s32>::max()};
+    constexpr s64 min{std::numeric_limits<s32>::min()};
+    constexpr s64 max{std::numeric_limits<s32>::max()};
     std::array<f64, 3> b{Common::FixedPoint<50, 14>::from_base(b_[0]).to_double(),
                          Common::FixedPoint<50, 14>::from_base(b_[1]).to_double(),
                          Common::FixedPoint<50, 14>::from_base(b_[2]).to_double()};
     std::array<f64, 2> a{Common::FixedPoint<50, 14>::from_base(a_[0]).to_double(),
                          Common::FixedPoint<50, 14>::from_base(a_[1]).to_double()};
-    std::array<f64, 4> s{Common::BitCast<f64>(state.s0), Common::BitCast<f64>(state.s1),
-                         Common::BitCast<f64>(state.s2), Common::BitCast<f64>(state.s3)};
+    std::array<f64, 4> s{state.s0.to_double(), state.s1.to_double(), state.s2.to_double(),
+                         state.s3.to_double()};
 
     for (u32 i = 0; i < sample_count; i++) {
         f64 in_sample{static_cast<f64>(input[i])};
         auto sample{in_sample * b[0] + s[0] * b[1] + s[1] * b[2] + s[2] * a[0] + s[3] * a[1]};
 
-        output[i] = static_cast<s32>(std::clamp(sample, min, max));
+        output[i] = static_cast<s32>(std::clamp(static_cast<s64>(sample), min, max));
 
         s[1] = s[0];
         s[0] = in_sample;
@@ -42,10 +41,10 @@ void ApplyBiquadFilterFloat(std::span<s32> output, std::span<const s32> input,
         s[2] = sample;
     }
 
-    state.s0 = Common::BitCast<s64>(s[0]);
-    state.s1 = Common::BitCast<s64>(s[1]);
-    state.s2 = Common::BitCast<s64>(s[2]);
-    state.s3 = Common::BitCast<s64>(s[3]);
+    state.s0 = s[0];
+    state.s1 = s[1];
+    state.s2 = s[2];
+    state.s3 = s[3];
 }
 
 /**
@@ -59,20 +58,29 @@ void ApplyBiquadFilterFloat(std::span<s32> output, std::span<const s32> input,
  * @param sample_count - Number of samples to process.
  */
 static void ApplyBiquadFilterInt(std::span<s32> output, std::span<const s32> input,
-                                 std::array<s16, 3>& b, std::array<s16, 2>& a,
+                                 std::array<s16, 3>& b_, std::array<s16, 2>& a_,
                                  VoiceState::BiquadFilterState& state, const u32 sample_count) {
     constexpr s64 min{std::numeric_limits<s32>::min()};
     constexpr s64 max{std::numeric_limits<s32>::max()};
+    std::array<Common::FixedPoint<50, 14>, 3> b{
+        Common::FixedPoint<50, 14>::from_base(b_[0]),
+        Common::FixedPoint<50, 14>::from_base(b_[1]),
+        Common::FixedPoint<50, 14>::from_base(b_[2]),
+    };
+    std::array<Common::FixedPoint<50, 14>, 3> a{
+        Common::FixedPoint<50, 14>::from_base(a_[0]),
+        Common::FixedPoint<50, 14>::from_base(a_[1]),
+    };
 
     for (u32 i = 0; i < sample_count; i++) {
-        const s64 in_sample{input[i]};
-        const s64 sample{in_sample * b[0] + state.s0};
-        const s64 out_sample{std::clamp<s64>((sample + (1 << 13)) >> 14, min, max)};
+        s64 in_sample{input[i]};
+        auto sample{in_sample * b[0] + state.s0};
+        const auto out_sample{std::clamp(sample.to_long(), min, max)};
 
         output[i] = static_cast<s32>(out_sample);
 
         state.s0 = state.s1 + b[1] * in_sample + a[0] * out_sample;
-        state.s1 = b[2] * in_sample + a[1] * out_sample;
+        state.s1 = 0 + b[2] * in_sample + a[1] * out_sample;
     }
 }
 

src/audio_core/renderer/voice/voice_state.h

@@ -19,10 +19,10 @@ struct VoiceState {
      * State of the voice's biquad filter.
      */
     struct BiquadFilterState {
-        s64 s0;
-        s64 s1;
-        s64 s2;
-        s64 s3;
+        Common::FixedPoint<50, 14> s0;
+        Common::FixedPoint<50, 14> s1;
+        Common::FixedPoint<50, 14> s2;
+        Common::FixedPoint<50, 14> s3;
     };
 
     /**

src/audio_core/sink/cubeb_sink.cpp

@@ -302,21 +302,11 @@ std::vector<std::string> ListCubebSinkDevices(bool capture) {
     std::vector<std::string> device_list;
     cubeb* ctx;
 
-#ifdef _WIN32
-    auto com_init_result = CoInitializeEx(nullptr, COINIT_MULTITHREADED);
-#endif
-
     if (cubeb_init(&ctx, "yuzu Device Enumerator", nullptr) != CUBEB_OK) {
         LOG_CRITICAL(Audio_Sink, "cubeb_init failed");
         return {};
     }
 
-#ifdef _WIN32
-    if (SUCCEEDED(com_init_result)) {
-        CoUninitialize();
-    }
-#endif
-
     auto type{capture ? CUBEB_DEVICE_TYPE_INPUT : CUBEB_DEVICE_TYPE_OUTPUT};
     cubeb_device_collection collection;
     if (cubeb_enumerate_devices(ctx, type, &collection) != CUBEB_OK) {
@@ -339,22 +329,12 @@ std::vector<std::string> ListCubebSinkDevices(bool capture) {
 u32 GetCubebLatency() {
     cubeb* ctx;
 
-#ifdef _WIN32
-    auto com_init_result = CoInitializeEx(nullptr, COINIT_MULTITHREADED);
-#endif
-
     if (cubeb_init(&ctx, "yuzu Latency Getter", nullptr) != CUBEB_OK) {
         LOG_CRITICAL(Audio_Sink, "cubeb_init failed");
         // Return a large latency so we choose SDL instead.
         return 10000u;
     }
 
-#ifdef _WIN32
-    if (SUCCEEDED(com_init_result)) {
-        CoUninitialize();
-    }
-#endif
-
     cubeb_stream_params params{};
     params.rate = TargetSampleRate;
     params.channels = 2;

src/core/hle/kernel/k_capabilities.cpp

@@ -203,23 +203,23 @@ Result KCapabilities::ProcessMapRegionCapability(const u32 cap, F f) {
 
 Result KCapabilities::MapRegion_(const u32 cap, KPageTable* page_table) {
     // Map each region into the process's page table.
-    return ProcessMapRegionCapability(
+    R_RETURN(ProcessMapRegionCapability(
         cap, [](KMemoryRegionType region_type, KMemoryPermission perm) -> Result {
             // R_RETURN(page_table->MapRegion(region_type, perm));
             UNIMPLEMENTED();
             R_SUCCEED();
-        });
+        }));
 }
 
 Result KCapabilities::CheckMapRegion(KernelCore& kernel, const u32 cap) {
     // Check that each region has a physical backing store.
-    return ProcessMapRegionCapability(
+    R_RETURN(ProcessMapRegionCapability(
         cap, [&](KMemoryRegionType region_type, KMemoryPermission perm) -> Result {
             R_UNLESS(kernel.MemoryLayout().GetPhysicalMemoryRegionTree().FindFirstDerived(
                          region_type) != nullptr,
                      ResultOutOfRange);
             R_SUCCEED();
-        });
+        }));
 }
 
 Result KCapabilities::SetInterruptPairCapability(const u32 cap) {

src/core/hle/kernel/physical_core.h

@@ -3,7 +3,6 @@
 
 #pragma once
 
-#include <condition_variable>
 #include <cstddef>
 #include <memory>
 #include <mutex>

src/core/hle/kernel/svc_version.h

@@ -35,11 +35,11 @@ constexpr inline u32 EncodeKernelVersion(u32 major, u32 minor) {
 }
 
 constexpr inline u32 GetKernelMajorVersion(u32 encoded) {
-    return decltype(KernelVersion::major_version)::ExtractValue(encoded);
+    return std::bit_cast<decltype(KernelVersion::major_version)>(encoded).Value();
 }
 
 constexpr inline u32 GetKernelMinorVersion(u32 encoded) {
-    return decltype(KernelVersion::minor_version)::ExtractValue(encoded);
+    return std::bit_cast<decltype(KernelVersion::minor_version)>(encoded).Value();
 }
 
 // Nintendo doesn't support programs targeting SVC versions < 3.0.

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

@@ -758,20 +758,12 @@ Core::HID::NpadStyleTag Controller_NPad::GetSupportedStyleSet() const {
     return hid_core.GetSupportedStyleTag();
 }
 
-Result Controller_NPad::SetSupportedNpadIdTypes(std::span<const u8> data) {
-    constexpr std::size_t max_number_npad_ids = 0xa;
+void Controller_NPad::SetSupportedNpadIdTypes(std::span<const u8> data) {
     const auto length = data.size();
     ASSERT(length > 0 && (length % sizeof(u32)) == 0);
-    const std::size_t elements = length / sizeof(u32);
-
-    if (elements > max_number_npad_ids) {
-        return InvalidArraySize;
-    }
-
     supported_npad_id_types.clear();
-    supported_npad_id_types.resize(elements);
+    supported_npad_id_types.resize(length / sizeof(u32));
     std::memcpy(supported_npad_id_types.data(), data.data(), length);
-    return ResultSuccess;
 }
 
 void Controller_NPad::GetSupportedNpadIdTypes(u32* data, std::size_t max_length) {

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

@@ -96,7 +96,7 @@ public:
     void SetSupportedStyleSet(Core::HID::NpadStyleTag style_set);
     Core::HID::NpadStyleTag GetSupportedStyleSet() const;
 
-    Result SetSupportedNpadIdTypes(std::span<const u8> data);
+    void SetSupportedNpadIdTypes(std::span<const u8> data);
     void GetSupportedNpadIdTypes(u32* data, std::size_t max_length);
     std::size_t GetSupportedNpadIdTypesSize() const;
 

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

@@ -18,7 +18,6 @@ constexpr Result NpadIsDualJoycon{ErrorModule::HID, 601};
 constexpr Result NpadIsSameType{ErrorModule::HID, 602};
 constexpr Result InvalidNpadId{ErrorModule::HID, 709};
 constexpr Result NpadNotConnected{ErrorModule::HID, 710};
-constexpr Result InvalidArraySize{ErrorModule::HID, 715};
 constexpr Result InvalidPalmaHandle{ErrorModule::HID, 3302};
 
 } // namespace Service::HID

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

@@ -1025,13 +1025,13 @@ void Hid::SetSupportedNpadIdType(Kernel::HLERequestContext& ctx) {
     IPC::RequestParser rp{ctx};
     const auto applet_resource_user_id{rp.Pop<u64>()};
 
-    const auto result = applet_resource->GetController<Controller_NPad>(HidController::NPad)
-                            .SetSupportedNpadIdTypes(ctx.ReadBuffer());
+    applet_resource->GetController<Controller_NPad>(HidController::NPad)
+        .SetSupportedNpadIdTypes(ctx.ReadBuffer());
 
     LOG_DEBUG(Service_HID, "called, applet_resource_user_id={}", applet_resource_user_id);
 
     IPC::ResponseBuilder rb{ctx, 2};
-    rb.Push(result);
+    rb.Push(ResultSuccess);
 }
 
 void Hid::ActivateNpad(Kernel::HLERequestContext& ctx) {

src/shader_recompiler/backend/glsl/glsl_emit_context.cpp

@@ -310,6 +310,12 @@ EmitContext::EmitContext(IR::Program& program, Bindings& bindings, const Profile
         if (runtime_info.force_early_z) {
             header += "layout(early_fragment_tests)in;";
         }
+        if (info.uses_sample_id) {
+            header += "in int gl_SampleID;";
+        }
+        if (info.stores_sample_mask) {
+            header += "out int gl_SampleMask[];";
+        }
         break;
     case Stage::Compute:
         stage_name = "cs";

src/tests/video_core/buffer_base.cpp

@@ -538,7 +538,7 @@ TEST_CASE("BufferBase: Cached write downloads") {
     int num = 0;
     buffer.ForEachDownloadRangeAndClear(c, WORD, [&](u64 offset, u64 size) { ++num; });
     buffer.ForEachUploadRange(c, WORD, [&](u64 offset, u64 size) { ++num; });
-    REQUIRE(num == 0);
+    REQUIRE(num == 1);
     REQUIRE(!buffer.IsRegionCpuModified(c + PAGE, PAGE));
     REQUIRE(!buffer.IsRegionGpuModified(c + PAGE, PAGE));
     buffer.FlushCachedWrites();

src/video_core/buffer_cache/buffer_base.h

@@ -430,7 +430,7 @@ private:
         if (query_begin >= SizeBytes() || size < 0) {
             return;
         }
-        u64* const untracked_words = Array<Type::Untracked>();
+        [[maybe_unused]] u64* const untracked_words = Array<Type::Untracked>();
         u64* const state_words = Array<type>();
         const u64 query_end = query_begin + std::min(static_cast<u64>(size), SizeBytes());
         u64* const words_begin = state_words + query_begin / BYTES_PER_WORD;
@@ -483,7 +483,7 @@ private:
                 NotifyRasterizer<true>(word_index, current_bits, ~u64{0});
             }
             // Exclude CPU modified pages when visiting GPU pages
-            const u64 word = current_word & ~(type == Type::GPU ? untracked_words[word_index] : 0);
+            const u64 word = current_word;
             u64 page = page_begin;
             page_begin = 0;
 
@@ -531,7 +531,7 @@ private:
     [[nodiscard]] bool IsRegionModified(u64 offset, u64 size) const noexcept {
         static_assert(type != Type::Untracked);
 
-        const u64* const untracked_words = Array<Type::Untracked>();
+        [[maybe_unused]] const u64* const untracked_words = Array<Type::Untracked>();
         const u64* const state_words = Array<type>();
         const u64 num_query_words = size / BYTES_PER_WORD + 1;
         const u64 word_begin = offset / BYTES_PER_WORD;
@@ -539,8 +539,7 @@ private:
         const u64 page_limit = Common::DivCeil(offset + size, BYTES_PER_PAGE);
         u64 page_index = (offset / BYTES_PER_PAGE) % PAGES_PER_WORD;
         for (u64 word_index = word_begin; word_index < word_end; ++word_index, page_index = 0) {
-            const u64 off_word = type == Type::GPU ? untracked_words[word_index] : 0;
-            const u64 word = state_words[word_index] & ~off_word;
+            const u64 word = state_words[word_index];
             if (word == 0) {
                 continue;
             }
@@ -564,7 +563,7 @@ private:
     [[nodiscard]] std::pair<u64, u64> ModifiedRegion(u64 offset, u64 size) const noexcept {
         static_assert(type != Type::Untracked);
 
-        const u64* const untracked_words = Array<Type::Untracked>();
+        [[maybe_unused]] const u64* const untracked_words = Array<Type::Untracked>();
         const u64* const state_words = Array<type>();
         const u64 num_query_words = size / BYTES_PER_WORD + 1;
         const u64 word_begin = offset / BYTES_PER_WORD;
@@ -574,8 +573,7 @@ private:
         u64 begin = std::numeric_limits<u64>::max();
         u64 end = 0;
         for (u64 word_index = word_begin; word_index < word_end; ++word_index) {
-            const u64 off_word = type == Type::GPU ? untracked_words[word_index] : 0;
-            const u64 word = state_words[word_index] & ~off_word;
+            const u64 word = state_words[word_index];
             if (word == 0) {
                 continue;
             }

src/video_core/gpu.cpp

@@ -99,7 +99,7 @@ struct GPU::Impl {
 
     /// Signal the ending of command list.
     void OnCommandListEnd() {
-        gpu_thread.OnCommandListEnd();
+        rasterizer->ReleaseFences();
     }
 
     /// Request a host GPU memory flush from the CPU.

src/video_core/gpu_thread.cpp

@@ -40,8 +40,6 @@ static void RunThread(std::stop_token stop_token, Core::System& system,
             scheduler.Push(submit_list->channel, std::move(submit_list->entries));
         } else if (const auto* data = std::get_if<SwapBuffersCommand>(&next.data)) {
             renderer.SwapBuffers(data->framebuffer ? &*data->framebuffer : nullptr);
-        } else if (std::holds_alternative<OnCommandListEndCommand>(next.data)) {
-            rasterizer->ReleaseFences();
         } else if (std::holds_alternative<GPUTickCommand>(next.data)) {
             system.GPU().TickWork();
         } else if (const auto* flush = std::get_if<FlushRegionCommand>(&next.data)) {
@@ -110,10 +108,6 @@ void ThreadManager::FlushAndInvalidateRegion(VAddr addr, u64 size) {
     rasterizer->OnCPUWrite(addr, size);
 }
 
-void ThreadManager::OnCommandListEnd() {
-    PushCommand(OnCommandListEndCommand());
-}
-
 u64 ThreadManager::PushCommand(CommandData&& command_data, bool block) {
     if (!is_async) {
         // In synchronous GPU mode, block the caller until the command has executed

src/video_core/gpu_thread.h

@@ -77,16 +77,12 @@ struct FlushAndInvalidateRegionCommand final {
     u64 size;
 };
 
-/// Command called within the gpu, to schedule actions after a command list end
-struct OnCommandListEndCommand final {};
-
 /// Command to make the gpu look into pending requests
 struct GPUTickCommand final {};
 
 using CommandData =
     std::variant<std::monostate, SubmitListCommand, SwapBuffersCommand, FlushRegionCommand,
-                 InvalidateRegionCommand, FlushAndInvalidateRegionCommand, OnCommandListEndCommand,
-                 GPUTickCommand>;
+                 InvalidateRegionCommand, FlushAndInvalidateRegionCommand, GPUTickCommand>;
 
 struct CommandDataContainer {
     CommandDataContainer() = default;
@@ -134,8 +130,6 @@ public:
     /// Notify rasterizer that any caches of the specified region should be flushed and invalidated
     void FlushAndInvalidateRegion(VAddr addr, u64 size);
 
-    void OnCommandListEnd();
-
     void TickGPU();
 
 private:

src/video_core/renderer_vulkan/vk_graphics_pipeline.cpp

@@ -548,7 +548,31 @@ void GraphicsPipeline::MakePipeline(VkRenderPass render_pass) {
     static_vector<VkVertexInputBindingDescription, 32> vertex_bindings;
     static_vector<VkVertexInputBindingDivisorDescriptionEXT, 32> vertex_binding_divisors;
     static_vector<VkVertexInputAttributeDescription, 32> vertex_attributes;
-    if (!key.state.dynamic_vertex_input) {
+    if (key.state.dynamic_vertex_input) {
+        const size_t num_vertex_arrays = std::min(
+            key.state.attributes.size(), static_cast<size_t>(device.GetMaxVertexInputBindings()));
+        for (size_t index = 0; index < num_vertex_arrays; ++index) {
+            const u32 type = key.state.DynamicAttributeType(index);
+            if (!stage_infos[0].loads.Generic(index) || type == 0) {
+                continue;
+            }
+            vertex_attributes.push_back({
+                .location = static_cast<u32>(index),
+                .binding = 0,
+                .format = type == 1   ? VK_FORMAT_R32_SFLOAT
+                          : type == 2 ? VK_FORMAT_R32_SINT
+                                      : VK_FORMAT_R32_UINT,
+                .offset = 0,
+            });
+        }
+        if (!vertex_attributes.empty()) {
+            vertex_bindings.push_back({
+                .binding = 0,
+                .stride = 4,
+                .inputRate = VK_VERTEX_INPUT_RATE_VERTEX,
+            });
+        }
+    } else {
         const size_t num_vertex_arrays = std::min(
             Maxwell::NumVertexArrays, static_cast<size_t>(device.GetMaxVertexInputBindings()));
         for (size_t index = 0; index < num_vertex_arrays; ++index) {

src/yuzu/main.cpp

@@ -805,8 +805,6 @@ void GMainWindow::WebBrowserOpenWebPage(const std::string& main_url,
                                 layout.screen.GetHeight() / scale_ratio);
         web_browser_view.move(layout.screen.left / scale_ratio,
                               (layout.screen.top / scale_ratio) + menuBar()->height());
-        web_browser_view.setZoomFactor(static_cast<qreal>(layout.screen.GetWidth() / scale_ratio) /
-                                       static_cast<qreal>(Layout::ScreenUndocked::Width));
 
         web_browser_view.setFocus();
         web_browser_view.show();