vk_device: remove left over from other branch

Please enter the commit message for your changes. Lines starting

CMakeLists.txt

@@ -157,8 +157,14 @@ if (ENABLE_SDL2)
         target_include_directories(SDL2 INTERFACE "${SDL2_INCLUDE_DIR}")
     else()
         find_package(SDL2 REQUIRED)
-        include_directories(${SDL2_INCLUDE_DIRS})
 
+        # Some installations don't set SDL2_LIBRARIES
+        if("${SDL2_LIBRARIES}" STREQUAL "")
+            message(WARNING "SDL2_LIBRARIES wasn't set, manually setting to SDL2::SDL2")
+            set(SDL2_LIBRARIES "SDL2::SDL2")
+        endif()
+
+        include_directories(${SDL2_INCLUDE_DIRS})
         add_library(SDL2 INTERFACE)
         target_link_libraries(SDL2 INTERFACE "${SDL2_LIBRARIES}")
     endif()

src/core/frontend/framebuffer_layout.cpp

@@ -27,9 +27,9 @@ FramebufferLayout DefaultFrameLayout(u32 width, u32 height) {
     // so just calculate them both even if the other isn't showing.
     FramebufferLayout res{width, height};
 
-    const float emulation_aspect_ratio{static_cast<float>(ScreenUndocked::Height) /
-                                       ScreenUndocked::Width};
-    const auto window_aspect_ratio = static_cast<float>(height) / width;
+    const float window_aspect_ratio = static_cast<float>(height) / width;
+    const float emulation_aspect_ratio = EmulationAspectRatio(
+        static_cast<AspectRatio>(Settings::values.aspect_ratio), window_aspect_ratio);
 
     const Common::Rectangle<u32> screen_window_area{0, 0, width, height};
     Common::Rectangle<u32> screen = MaxRectangle(screen_window_area, emulation_aspect_ratio);
@@ -58,4 +58,19 @@ FramebufferLayout FrameLayoutFromResolutionScale(u32 res_scale) {
     return DefaultFrameLayout(width, height);
 }
 
+float EmulationAspectRatio(AspectRatio aspect, float window_aspect_ratio) {
+    switch (aspect) {
+    case AspectRatio::Default:
+        return static_cast<float>(ScreenUndocked::Height) / ScreenUndocked::Width;
+    case AspectRatio::R4_3:
+        return 3.0f / 4.0f;
+    case AspectRatio::R21_9:
+        return 9.0f / 21.0f;
+    case AspectRatio::StretchToWindow:
+        return window_aspect_ratio;
+    default:
+        return static_cast<float>(ScreenUndocked::Height) / ScreenUndocked::Width;
+    }
+}
+
 } // namespace Layout

src/core/frontend/framebuffer_layout.h

@@ -18,6 +18,13 @@ enum ScreenDocked : u32 {
     HeightDocked = 1080,
 };
 
+enum class AspectRatio {
+    Default,
+    R4_3,
+    R21_9,
+    StretchToWindow,
+};
+
 /// Describes the layout of the window framebuffer
 struct FramebufferLayout {
     u32 width{ScreenUndocked::Width};
@@ -48,4 +55,12 @@ FramebufferLayout DefaultFrameLayout(u32 width, u32 height);
  */
 FramebufferLayout FrameLayoutFromResolutionScale(u32 res_scale);
 
+/**
+ * Convenience method to determine emulation aspect ratio
+ * @param aspect Represents the index of aspect ratio stored in Settings::values.aspect_ratio
+ * @param window_aspect_ratio Current window aspect ratio
+ * @return Emulation render window aspect ratio
+ */
+float EmulationAspectRatio(AspectRatio aspect, float window_aspect_ratio);
+
 } // namespace Layout

src/core/hle/service/ldn/ldn.cpp

@@ -129,12 +129,20 @@ public:
             {304, nullptr, "Disconnect"},
             {400, nullptr, "Initialize"},
             {401, nullptr, "Finalize"},
-            {402, nullptr, "SetOperationMode"},
+            {402, &IUserLocalCommunicationService::Initialize2, "Initialize2"}, // 7.0.0+
         };
         // clang-format on
 
         RegisterHandlers(functions);
     }
+
+    void Initialize2(Kernel::HLERequestContext& ctx) {
+        LOG_WARNING(Service_LDN, "(STUBBED) called");
+        // Result success seem make this services start network and continue.
+        // If we just pass result error then it will stop and maybe try again and again.
+        IPC::ResponseBuilder rb{ctx, 2};
+        rb.Push(RESULT_UNKNOWN);
+    }
 };
 
 class LDNS final : public ServiceFramework<LDNS> {

src/core/hle/service/nvdrv/devices/nvhost_gpu.cpp

@@ -44,6 +44,8 @@ u32 nvhost_gpu::ioctl(Ioctl command, const std::vector<u8>& input, const std::ve
         return GetWaitbase(input, output);
     case IoctlCommand::IocChannelSetTimeoutCommand:
         return ChannelSetTimeout(input, output);
+    case IoctlCommand::IocChannelSetTimeslice:
+        return ChannelSetTimeslice(input, output);
     default:
         break;
     }
@@ -228,4 +230,14 @@ u32 nvhost_gpu::ChannelSetTimeout(const std::vector<u8>& input, std::vector<u8>&
     return 0;
 }
 
+u32 nvhost_gpu::ChannelSetTimeslice(const std::vector<u8>& input, std::vector<u8>& output) {
+    IoctlSetTimeslice params{};
+    std::memcpy(&params, input.data(), sizeof(IoctlSetTimeslice));
+    LOG_INFO(Service_NVDRV, "called, timeslice=0x{:X}", params.timeslice);
+
+    channel_timeslice = params.timeslice;
+
+    return 0;
+}
+
 } // namespace Service::Nvidia::Devices

src/core/hle/service/nvdrv/devices/nvhost_gpu.h

@@ -48,6 +48,7 @@ private:
         IocAllocObjCtxCommand = 0xC0104809,
         IocChannelGetWaitbaseCommand = 0xC0080003,
         IocChannelSetTimeoutCommand = 0x40044803,
+        IocChannelSetTimeslice = 0xC004481D,
     };
 
     enum class CtxObjects : u32_le {
@@ -101,6 +102,11 @@ private:
     static_assert(sizeof(IoctlChannelSetPriority) == 4,
                   "IoctlChannelSetPriority is incorrect size");
 
+    struct IoctlSetTimeslice {
+        u32_le timeslice;
+    };
+    static_assert(sizeof(IoctlSetTimeslice) == 4, "IoctlSetTimeslice is incorrect size");
+
     struct IoctlEventIdControl {
         u32_le cmd; // 0=disable, 1=enable, 2=clear
         u32_le id;
@@ -174,6 +180,7 @@ private:
     u64_le user_data{};
     IoctlZCullBind zcull_params{};
     u32_le channel_priority{};
+    u32_le channel_timeslice{};
 
     u32 SetNVMAPfd(const std::vector<u8>& input, std::vector<u8>& output);
     u32 SetClientData(const std::vector<u8>& input, std::vector<u8>& output);
@@ -188,6 +195,7 @@ private:
                   const std::vector<u8>& input2, IoctlVersion version);
     u32 GetWaitbase(const std::vector<u8>& input, std::vector<u8>& output);
     u32 ChannelSetTimeout(const std::vector<u8>& input, std::vector<u8>& output);
+    u32 ChannelSetTimeslice(const std::vector<u8>& input, std::vector<u8>& output);
 
     std::shared_ptr<nvmap> nvmap_dev;
     u32 assigned_syncpoints{};

src/core/settings.h

@@ -429,6 +429,7 @@ struct Values {
     int vulkan_device;
 
     float resolution_factor;
+    int aspect_ratio;
     bool use_frame_limit;
     u16 frame_limit;
     bool use_disk_shader_cache;

src/video_core/gpu.cpp

@@ -24,7 +24,7 @@ MICROPROFILE_DEFINE(GPU_wait, "GPU", "Wait for the GPU", MP_RGB(128, 128, 192));
 GPU::GPU(Core::System& system, VideoCore::RendererBase& renderer, bool is_async)
     : system{system}, renderer{renderer}, is_async{is_async} {
     auto& rasterizer{renderer.Rasterizer()};
-    memory_manager = std::make_unique<Tegra::MemoryManager>(system);
+    memory_manager = std::make_unique<Tegra::MemoryManager>(system, rasterizer);
     dma_pusher = std::make_unique<Tegra::DmaPusher>(*this);
     maxwell_3d = std::make_unique<Engines::Maxwell3D>(system, rasterizer, *memory_manager);
     fermi_2d = std::make_unique<Engines::Fermi2D>(rasterizer);

src/video_core/memory_manager.cpp

@@ -9,12 +9,13 @@
 #include "core/hle/kernel/process.h"
 #include "core/hle/kernel/vm_manager.h"
 #include "core/memory.h"
-#include "video_core/gpu.h"
 #include "video_core/memory_manager.h"
+#include "video_core/rasterizer_interface.h"
 
 namespace Tegra {
 
-MemoryManager::MemoryManager(Core::System& system) : system{system} {
+MemoryManager::MemoryManager(Core::System& system, VideoCore::RasterizerInterface& rasterizer)
+    : rasterizer{rasterizer}, system{system} {
     std::fill(page_table.pointers.begin(), page_table.pointers.end(), nullptr);
     std::fill(page_table.attributes.begin(), page_table.attributes.end(),
               Common::PageType::Unmapped);
@@ -83,8 +84,7 @@ GPUVAddr MemoryManager::UnmapBuffer(GPUVAddr gpu_addr, u64 size) {
     const auto cpu_addr = GpuToCpuAddress(gpu_addr);
     ASSERT(cpu_addr);
 
-    system.GPU().FlushAndInvalidateRegion(cache_addr, aligned_size);
-
+    rasterizer.FlushAndInvalidateRegion(cache_addr, aligned_size);
     UnmapRange(gpu_addr, aligned_size);
     ASSERT(system.CurrentProcess()
                ->VMManager()
@@ -242,7 +242,7 @@ void MemoryManager::ReadBlock(GPUVAddr src_addr, void* dest_buffer, const std::s
         switch (page_table.attributes[page_index]) {
         case Common::PageType::Memory: {
             const u8* src_ptr{page_table.pointers[page_index] + page_offset};
-            system.GPU().FlushRegion(ToCacheAddr(src_ptr), copy_amount);
+            rasterizer.FlushRegion(ToCacheAddr(src_ptr), copy_amount);
             std::memcpy(dest_buffer, src_ptr, copy_amount);
             break;
         }
@@ -292,7 +292,7 @@ void MemoryManager::WriteBlock(GPUVAddr dest_addr, const void* src_buffer, const
         switch (page_table.attributes[page_index]) {
         case Common::PageType::Memory: {
             u8* dest_ptr{page_table.pointers[page_index] + page_offset};
-            system.GPU().InvalidateRegion(ToCacheAddr(dest_ptr), copy_amount);
+            rasterizer.InvalidateRegion(ToCacheAddr(dest_ptr), copy_amount);
             std::memcpy(dest_ptr, src_buffer, copy_amount);
             break;
         }
@@ -340,7 +340,7 @@ void MemoryManager::CopyBlock(GPUVAddr dest_addr, GPUVAddr src_addr, const std::
         switch (page_table.attributes[page_index]) {
         case Common::PageType::Memory: {
             const u8* src_ptr{page_table.pointers[page_index] + page_offset};
-            system.GPU().FlushRegion(ToCacheAddr(src_ptr), copy_amount);
+            rasterizer.FlushRegion(ToCacheAddr(src_ptr), copy_amount);
             WriteBlock(dest_addr, src_ptr, copy_amount);
             break;
         }

src/video_core/memory_manager.h

@@ -10,6 +10,10 @@
 #include "common/common_types.h"
 #include "common/page_table.h"
 
+namespace VideoCore {
+class RasterizerInterface;
+}
+
 namespace Core {
 class System;
 }
@@ -47,7 +51,7 @@ struct VirtualMemoryArea {
 
 class MemoryManager final {
 public:
-    explicit MemoryManager(Core::System& system);
+    explicit MemoryManager(Core::System& system, VideoCore::RasterizerInterface& rasterizer);
     ~MemoryManager();
 
     GPUVAddr AllocateSpace(u64 size, u64 align);
@@ -172,6 +176,7 @@ private:
 
     Common::PageTable page_table{page_bits};
     VMAMap vma_map;
+    VideoCore::RasterizerInterface& rasterizer;
 
     Core::System& system;
 };

src/video_core/renderer_vulkan/vk_device.cpp

@@ -106,6 +106,8 @@ bool VKDevice::Create(const vk::DispatchLoaderDynamic& dldi, vk::Instance instan
     features.tessellationShader = true;
     features.fragmentStoresAndAtomics = true;
     features.shaderImageGatherExtended = true;
+    features.shaderStorageImageReadWithoutFormat =
+        is_shader_storage_img_read_without_format_supported;
     features.shaderStorageImageWriteWithoutFormat = true;
     features.textureCompressionASTC_LDR = is_optimal_astc_supported;
 
@@ -457,6 +459,8 @@ void VKDevice::SetupFamilies(const vk::DispatchLoaderDynamic& dldi, vk::SurfaceK
 
 void VKDevice::SetupFeatures(const vk::DispatchLoaderDynamic& dldi) {
     const auto supported_features{physical.getFeatures(dldi)};
+    is_shader_storage_img_read_without_format_supported =
+        supported_features.shaderStorageImageReadWithoutFormat;
     is_optimal_astc_supported = IsOptimalAstcSupported(supported_features, dldi);
 }
 

src/video_core/renderer_vulkan/vk_device.h

@@ -122,6 +122,11 @@ public:
         return properties.limits.maxPushConstantsSize;
     }
 
+    /// Returns true if Shader storage Image Read Without Format supported.
+    bool IsShaderStorageImageReadWithoutFormatSupported() const {
+        return is_shader_storage_img_read_without_format_supported;
+    }
+
     /// Returns true if ASTC is natively supported.
     bool IsOptimalAstcSupported() const {
         return is_optimal_astc_supported;
@@ -227,6 +232,8 @@ private:
     bool ext_depth_range_unrestricted{};       ///< Support for VK_EXT_depth_range_unrestricted.
     bool ext_shader_viewport_index_layer{};    ///< Support for VK_EXT_shader_viewport_index_layer.
     bool nv_device_diagnostic_checkpoints{};   ///< Support for VK_NV_device_diagnostic_checkpoints.
+    bool is_shader_storage_img_read_without_format_supported{}; ///< Support for shader storage
+                                                                ///< image read without format
 
     // Telemetry parameters
     std::string vendor_name;                      ///< Device's driver name.

src/video_core/renderer_vulkan/vk_shader_decompiler.cpp

@@ -86,7 +86,6 @@ struct AttributeType {
 
 struct VertexIndices {
     std::optional<u32> position;
-    std::optional<u32> layer;
     std::optional<u32> viewport;
     std::optional<u32> point_size;
     std::optional<u32> clip_distances;
@@ -283,16 +282,18 @@ public:
         AddExtension("SPV_KHR_storage_buffer_storage_class");
         AddExtension("SPV_KHR_variable_pointers");
 
-        if (ir.UsesLayer() || ir.UsesViewportIndex()) {
-            if (ir.UsesViewportIndex()) {
-                AddCapability(spv::Capability::MultiViewport);
-            }
-            if (stage != ShaderType::Geometry && device.IsExtShaderViewportIndexLayerSupported()) {
+        if (ir.UsesViewportIndex()) {
+            AddCapability(spv::Capability::MultiViewport);
+            if (device.IsExtShaderViewportIndexLayerSupported()) {
                 AddExtension("SPV_EXT_shader_viewport_index_layer");
                 AddCapability(spv::Capability::ShaderViewportIndexLayerEXT);
             }
         }
 
+        if (device.IsShaderStorageImageReadWithoutFormatSupported()) {
+            AddCapability(spv::Capability::StorageImageReadWithoutFormat);
+        }
+
         if (device.IsFloat16Supported()) {
             AddCapability(spv::Capability::Float16);
         }
@@ -923,22 +924,13 @@ private:
         VertexIndices indices;
         indices.position = AddBuiltIn(t_float4, spv::BuiltIn::Position, "position");
 
-        if (ir.UsesLayer()) {
-            if (stage != ShaderType::Vertex || device.IsExtShaderViewportIndexLayerSupported()) {
-                indices.layer = AddBuiltIn(t_int, spv::BuiltIn::Layer, "layer");
-            } else {
-                LOG_ERROR(
-                    Render_Vulkan,
-                    "Shader requires Layer but it's not supported on this stage with this device.");
-            }
-        }
-
         if (ir.UsesViewportIndex()) {
             if (stage != ShaderType::Vertex || device.IsExtShaderViewportIndexLayerSupported()) {
                 indices.viewport = AddBuiltIn(t_int, spv::BuiltIn::ViewportIndex, "viewport_index");
             } else {
-                LOG_ERROR(Render_Vulkan, "Shader requires ViewportIndex but it's not supported on "
-                                         "this stage with this device.");
+                LOG_ERROR(Render_Vulkan,
+                          "Shader requires ViewportIndex but it's not supported on this "
+                          "stage with this device.");
             }
         }
 
@@ -1297,13 +1289,6 @@ private:
                 }
                 case Attribute::Index::LayerViewportPointSize:
                     switch (element) {
-                    case 1: {
-                        if (!out_indices.layer) {
-                            return {};
-                        }
-                        const u32 index = out_indices.layer.value();
-                        return {AccessElement(t_out_int, out_vertex, index), Type::Int};
-                    }
                     case 2: {
                         if (!out_indices.viewport) {
                             return {};
@@ -1374,11 +1359,6 @@ private:
             UNIMPLEMENTED();
         }
 
-        if (!target.id) {
-            // On failure we return a nullptr target.id, skip these stores.
-            return {};
-        }
-
         OpStore(target.id, As(Visit(src), target.type));
         return {};
     }
@@ -1772,8 +1752,16 @@ private:
     }
 
     Expression ImageLoad(Operation operation) {
-        UNIMPLEMENTED();
-        return {};
+        if (!device.IsShaderStorageImageReadWithoutFormatSupported()) {
+            return {v_float_zero, Type::Float};
+        }
+
+        const auto& meta{std::get<MetaImage>(operation.GetMeta())};
+
+        const Id coords = GetCoordinates(operation, Type::Int);
+        const Id texel = OpImageRead(t_uint4, GetImage(operation), coords);
+
+        return {OpCompositeExtract(t_uint, texel, meta.element), Type::Uint};
     }
 
     Expression ImageStore(Operation operation) {

src/yuzu/configuration/config.cpp

@@ -630,6 +630,7 @@ void Config::ReadRendererValues() {
     Settings::values.vulkan_device = ReadSetting(QStringLiteral("vulkan_device"), 0).toInt();
     Settings::values.resolution_factor =
         ReadSetting(QStringLiteral("resolution_factor"), 1.0).toFloat();
+    Settings::values.aspect_ratio = ReadSetting(QStringLiteral("aspect_ratio"), 0).toInt();
     Settings::values.use_frame_limit =
         ReadSetting(QStringLiteral("use_frame_limit"), true).toBool();
     Settings::values.frame_limit = ReadSetting(QStringLiteral("frame_limit"), 100).toInt();
@@ -1064,6 +1065,7 @@ void Config::SaveRendererValues() {
     WriteSetting(QStringLiteral("vulkan_device"), Settings::values.vulkan_device, 0);
     WriteSetting(QStringLiteral("resolution_factor"),
                  static_cast<double>(Settings::values.resolution_factor), 1.0);
+    WriteSetting(QStringLiteral("aspect_ratio"), Settings::values.aspect_ratio, 0);
     WriteSetting(QStringLiteral("use_frame_limit"), Settings::values.use_frame_limit, true);
     WriteSetting(QStringLiteral("frame_limit"), Settings::values.frame_limit, 100);
     WriteSetting(QStringLiteral("use_disk_shader_cache"), Settings::values.use_disk_shader_cache,

src/yuzu/configuration/configure_graphics.cpp

@@ -97,6 +97,7 @@ void ConfigureGraphics::SetConfiguration() {
     ui->api->setCurrentIndex(static_cast<int>(Settings::values.renderer_backend));
     ui->resolution_factor_combobox->setCurrentIndex(
         static_cast<int>(FromResolutionFactor(Settings::values.resolution_factor)));
+    ui->aspect_ratio_combobox->setCurrentIndex(Settings::values.aspect_ratio);
     ui->use_disk_shader_cache->setEnabled(runtime_lock);
     ui->use_disk_shader_cache->setChecked(Settings::values.use_disk_shader_cache);
     ui->use_accurate_gpu_emulation->setChecked(Settings::values.use_accurate_gpu_emulation);
@@ -114,6 +115,7 @@ void ConfigureGraphics::ApplyConfiguration() {
     Settings::values.vulkan_device = vulkan_device;
     Settings::values.resolution_factor =
         ToResolutionFactor(static_cast<Resolution>(ui->resolution_factor_combobox->currentIndex()));
+    Settings::values.aspect_ratio = ui->aspect_ratio_combobox->currentIndex();
     Settings::values.use_disk_shader_cache = ui->use_disk_shader_cache->isChecked();
     Settings::values.use_accurate_gpu_emulation = ui->use_accurate_gpu_emulation->isChecked();
     Settings::values.use_asynchronous_gpu_emulation =

src/yuzu/configuration/configure_graphics.ui

@@ -138,6 +138,41 @@
           </item>
          </layout>
         </item>
+        <item>
+         <layout class="QHBoxLayout" name="horizontalLayout_6">
+          <item>
+           <widget class="QLabel" name="ar_label">
+            <property name="text">
+             <string>Aspect Ratio:</string>
+            </property>
+           </widget>
+          </item>
+          <item>
+           <widget class="QComboBox" name="aspect_ratio_combobox">
+            <item>
+             <property name="text">
+              <string>Default (16:9)</string>
+             </property>
+            </item>
+            <item>
+             <property name="text">
+              <string>Force 4:3</string>
+             </property>
+            </item>
+            <item>
+             <property name="text">
+              <string>Force 21:9</string>
+             </property>
+            </item>
+            <item>
+             <property name="text">
+              <string>Stretch to Window</string>
+             </property>
+            </item>
+           </widget>
+          </item>
+         </layout>
+        </item>
         <item>
          <layout class="QHBoxLayout" name="horizontalLayout_3">
           <item>

src/yuzu_cmd/config.cpp

@@ -379,6 +379,8 @@ void Config::ReadValues() {
 
     Settings::values.resolution_factor =
         static_cast<float>(sdl2_config->GetReal("Renderer", "resolution_factor", 1.0));
+    Settings::values.aspect_ratio =
+        static_cast<int>(sdl2_config->GetInteger("Renderer", "aspect_ratio", 0));
     Settings::values.use_frame_limit = sdl2_config->GetBoolean("Renderer", "use_frame_limit", true);
     Settings::values.frame_limit =
         static_cast<u16>(sdl2_config->GetInteger("Renderer", "frame_limit", 100));

src/yuzu_cmd/default_ini.h

@@ -122,6 +122,10 @@ use_shader_jit =
 # factor for the Switch resolution
 resolution_factor =
 
+# Aspect ratio
+# 0: Default (16:9), 1: Force 4:3, 2: Force 21:9, 3: Stretch to Window
+aspect_ratio =
+
 # Whether to enable V-Sync (caps the framerate at 60FPS) or not.
 # 0 (default): Off, 1: On
 use_vsync =

src/yuzu_tester/config.cpp

@@ -118,6 +118,8 @@ void Config::ReadValues() {
     // Renderer
     Settings::values.resolution_factor =
         static_cast<float>(sdl2_config->GetReal("Renderer", "resolution_factor", 1.0));
+    Settings::values.aspect_ratio =
+        static_cast<int>(sdl2_config->GetInteger("Renderer", "aspect_ratio", 0));
     Settings::values.use_frame_limit = false;
     Settings::values.frame_limit = 100;
     Settings::values.use_disk_shader_cache =

src/yuzu_tester/default_ini.h

@@ -26,6 +26,10 @@ use_shader_jit =
 # factor for the Switch resolution
 resolution_factor =
 
+# Aspect ratio
+# 0: Default (16:9), 1: Force 4:3, 2: Force 21:9, 3: Stretch to Window
+aspect_ratio =
+
 # Whether to enable V-Sync (caps the framerate at 60FPS) or not.
 # 0 (default): Off, 1: On
 use_vsync =