diff options
83 files changed, 650 insertions, 467 deletions
diff --git a/.ci/scripts/linux/docker.sh b/.ci/scripts/linux/docker.sh index f11878128..5559a527c 100644..100755 --- a/.ci/scripts/linux/docker.sh +++ b/.ci/scripts/linux/docker.sh @@ -5,7 +5,7 @@ cd /yuzu ccache -s mkdir build || true && cd build -cmake .. -G Ninja -DDISPLAY_VERSION=$1 -DYUZU_USE_BUNDLED_UNICORN=ON -DYUZU_USE_QT_WEB_ENGINE=ON -DCMAKE_BUILD_TYPE=Release -DCMAKE_C_COMPILER=/usr/lib/ccache/gcc -DCMAKE_CXX_COMPILER=/usr/lib/ccache/g++ -DYUZU_ENABLE_COMPATIBILITY_REPORTING=${ENABLE_COMPATIBILITY_REPORTING:-"OFF"} -DENABLE_COMPATIBILITY_LIST_DOWNLOAD=ON -DUSE_DISCORD_PRESENCE=ON -DENABLE_VULKAN=No +cmake .. -G Ninja -DDISPLAY_VERSION=$1 -DYUZU_USE_BUNDLED_UNICORN=ON -DYUZU_USE_QT_WEB_ENGINE=ON -DCMAKE_BUILD_TYPE=Release -DCMAKE_C_COMPILER=/usr/lib/ccache/gcc -DCMAKE_CXX_COMPILER=/usr/lib/ccache/g++ -DYUZU_ENABLE_COMPATIBILITY_REPORTING=${ENABLE_COMPATIBILITY_REPORTING:-"OFF"} -DENABLE_COMPATIBILITY_LIST_DOWNLOAD=ON -DUSE_DISCORD_PRESENCE=ON ninja diff --git a/.ci/scripts/windows/docker.sh b/.ci/scripts/windows/docker.sh index beb554b65..5d5650d3b 100644..100755 --- a/.ci/scripts/windows/docker.sh +++ b/.ci/scripts/windows/docker.sh @@ -13,7 +13,7 @@ echo '' >> /bin/cmd chmod +x /bin/cmd mkdir build || true && cd build -cmake .. -G Ninja -DDISPLAY_VERSION=$1 -DCMAKE_TOOLCHAIN_FILE="$(pwd)/../CMakeModules/MinGWCross.cmake" -DUSE_CCACHE=ON -DYUZU_USE_BUNDLED_UNICORN=ON -DENABLE_COMPATIBILITY_LIST_DOWNLOAD=ON -DCMAKE_BUILD_TYPE=Release -DENABLE_VULKAN=No +cmake .. -G Ninja -DDISPLAY_VERSION=$1 -DCMAKE_TOOLCHAIN_FILE="$(pwd)/../CMakeModules/MinGWCross.cmake" -DUSE_CCACHE=ON -DYUZU_USE_BUNDLED_UNICORN=ON -DENABLE_COMPATIBILITY_LIST_DOWNLOAD=ON -DCMAKE_BUILD_TYPE=Release ninja # Clean up the dirty hacks @@ -29,7 +29,13 @@ echo 'Prepare binaries...' cd .. mkdir package -QT_PLATFORM_DLL_PATH='/usr/x86_64-w64-mingw32/lib/qt5/plugins/platforms/' +if [ -d "/usr/x86_64-w64-mingw32/lib/qt5/plugins/platforms/" ]; then + QT_PLATFORM_DLL_PATH='/usr/x86_64-w64-mingw32/lib/qt5/plugins/platforms/' +else + #fallback to qt + QT_PLATFORM_DLL_PATH='/usr/x86_64-w64-mingw32/lib/qt/plugins/platforms/' +fi + find build/ -name "yuzu*.exe" -exec cp {} 'package' \; # copy Qt plugins diff --git a/CMakeModules/MinGWCross.cmake b/CMakeModules/MinGWCross.cmake index 29ecd1ac4..b268e72d8 100644 --- a/CMakeModules/MinGWCross.cmake +++ b/CMakeModules/MinGWCross.cmake @@ -10,8 +10,8 @@ set(SDL2_PATH ${MINGW_PREFIX}) set(MINGW_TOOL_PREFIX ${CMAKE_SYSTEM_PROCESSOR}-w64-mingw32-) # Specify the cross compiler -set(CMAKE_C_COMPILER ${MINGW_TOOL_PREFIX}gcc-posix) -set(CMAKE_CXX_COMPILER ${MINGW_TOOL_PREFIX}g++-posix) +set(CMAKE_C_COMPILER ${MINGW_TOOL_PREFIX}gcc) +set(CMAKE_CXX_COMPILER ${MINGW_TOOL_PREFIX}g++) set(CMAKE_RC_COMPILER ${MINGW_TOOL_PREFIX}windres) # Mingw tools diff --git a/externals/cubeb b/externals/cubeb -Subproject 6f2420de8f155b10330cf973900ac7bdbfee589 +Subproject 616d773441b5355800ce64197a699e6cd6b3617 diff --git a/src/common/uuid.h b/src/common/uuid.h index f6ad064fb..4d3af8cec 100644 --- a/src/common/uuid.h +++ b/src/common/uuid.h @@ -40,6 +40,11 @@ struct UUID { uuid = INVALID_UUID; } + // TODO(ogniK): Properly generate a Nintendo ID + constexpr u64 GetNintendoID() const { + return uuid[0]; + } + std::string Format() const; std::string FormatSwitch() const; }; diff --git a/src/core/hle/kernel/handle_table.cpp b/src/core/hle/kernel/handle_table.cpp index e441a27fc..35448b576 100644 --- a/src/core/hle/kernel/handle_table.cpp +++ b/src/core/hle/kernel/handle_table.cpp @@ -30,6 +30,7 @@ HandleTable::~HandleTable() = default; ResultCode HandleTable::SetSize(s32 handle_table_size) { if (static_cast<u32>(handle_table_size) > MAX_COUNT) { + LOG_ERROR(Kernel, "Handle table size {} is greater than {}", handle_table_size, MAX_COUNT); return ERR_OUT_OF_MEMORY; } @@ -80,6 +81,7 @@ ResultVal<Handle> HandleTable::Duplicate(Handle handle) { ResultCode HandleTable::Close(Handle handle) { if (!IsValid(handle)) { + LOG_ERROR(Kernel, "Handle is not valid! handle={:08X}", handle); return ERR_INVALID_HANDLE; } diff --git a/src/core/hle/kernel/mutex.cpp b/src/core/hle/kernel/mutex.cpp index eff4e45b0..7869eb32b 100644 --- a/src/core/hle/kernel/mutex.cpp +++ b/src/core/hle/kernel/mutex.cpp @@ -7,6 +7,7 @@ #include <vector> #include "common/assert.h" +#include "common/logging/log.h" #include "core/core.h" #include "core/hle/kernel/errors.h" #include "core/hle/kernel/handle_table.h" @@ -67,6 +68,7 @@ ResultCode Mutex::TryAcquire(VAddr address, Handle holding_thread_handle, Handle requesting_thread_handle) { // The mutex address must be 4-byte aligned if ((address % sizeof(u32)) != 0) { + LOG_ERROR(Kernel, "Address is not 4-byte aligned! address={:016X}", address); return ERR_INVALID_ADDRESS; } @@ -88,6 +90,8 @@ ResultCode Mutex::TryAcquire(VAddr address, Handle holding_thread_handle, } if (holding_thread == nullptr) { + LOG_ERROR(Kernel, "Holding thread does not exist! thread_handle={:08X}", + holding_thread_handle); return ERR_INVALID_HANDLE; } @@ -109,6 +113,7 @@ ResultCode Mutex::TryAcquire(VAddr address, Handle holding_thread_handle, ResultCode Mutex::Release(VAddr address) { // The mutex address must be 4-byte aligned if ((address % sizeof(u32)) != 0) { + LOG_ERROR(Kernel, "Address is not 4-byte aligned! address={:016X}", address); return ERR_INVALID_ADDRESS; } diff --git a/src/core/hle/kernel/process_capability.cpp b/src/core/hle/kernel/process_capability.cpp index 48e5ae682..63880f13d 100644 --- a/src/core/hle/kernel/process_capability.cpp +++ b/src/core/hle/kernel/process_capability.cpp @@ -3,6 +3,7 @@ // Refer to the license.txt file included. #include "common/bit_util.h" +#include "common/logging/log.h" #include "core/hle/kernel/errors.h" #include "core/hle/kernel/handle_table.h" #include "core/hle/kernel/memory/page_table.h" @@ -119,22 +120,30 @@ ResultCode ProcessCapabilities::ParseCapabilities(const u32* capabilities, // The MapPhysical type uses two descriptor flags for its parameters. // If there's only one, then there's a problem. if (i >= num_capabilities) { + LOG_ERROR(Kernel, "Invalid combination! i={}", i); return ERR_INVALID_COMBINATION; } const auto size_flags = capabilities[i]; if (GetCapabilityType(size_flags) != CapabilityType::MapPhysical) { + LOG_ERROR(Kernel, "Invalid capability type! size_flags={}", size_flags); return ERR_INVALID_COMBINATION; } const auto result = HandleMapPhysicalFlags(descriptor, size_flags, page_table); if (result.IsError()) { + LOG_ERROR(Kernel, "Failed to map physical flags! descriptor={}, size_flags={}", + descriptor, size_flags); return result; } } else { const auto result = ParseSingleFlagCapability(set_flags, set_svc_bits, descriptor, page_table); if (result.IsError()) { + LOG_ERROR( + Kernel, + "Failed to parse capability flag! set_flags={}, set_svc_bits={}, descriptor={}", + set_flags, set_svc_bits, descriptor); return result; } } @@ -162,6 +171,9 @@ ResultCode ProcessCapabilities::ParseSingleFlagCapability(u32& set_flags, u32& s const u32 flag_length = GetFlagBitOffset(type); const u32 set_flag = 1U << flag_length; if ((set_flag & set_flags & InitializeOnceMask) != 0) { + LOG_ERROR(Kernel, + "Attempted to initialize flags that may only be initialized once. set_flags={}", + set_flags); return ERR_INVALID_COMBINATION; } set_flags |= set_flag; @@ -187,6 +199,7 @@ ResultCode ProcessCapabilities::ParseSingleFlagCapability(u32& set_flags, u32& s break; } + LOG_ERROR(Kernel, "Invalid capability type! type={}", static_cast<u32>(type)); return ERR_INVALID_CAPABILITY_DESCRIPTOR; } @@ -208,23 +221,31 @@ void ProcessCapabilities::Clear() { ResultCode ProcessCapabilities::HandlePriorityCoreNumFlags(u32 flags) { if (priority_mask != 0 || core_mask != 0) { + LOG_ERROR(Kernel, "Core or priority mask are not zero! priority_mask={}, core_mask={}", + priority_mask, core_mask); return ERR_INVALID_CAPABILITY_DESCRIPTOR; } const u32 core_num_min = (flags >> 16) & 0xFF; const u32 core_num_max = (flags >> 24) & 0xFF; if (core_num_min > core_num_max) { + LOG_ERROR(Kernel, "Core min is greater than core max! core_num_min={}, core_num_max={}", + core_num_min, core_num_max); return ERR_INVALID_COMBINATION; } const u32 priority_min = (flags >> 10) & 0x3F; const u32 priority_max = (flags >> 4) & 0x3F; if (priority_min > priority_max) { + LOG_ERROR(Kernel, + "Priority min is greater than priority max! priority_min={}, priority_max={}", + core_num_min, priority_max); return ERR_INVALID_COMBINATION; } // The switch only has 4 usable cores. if (core_num_max >= 4) { + LOG_ERROR(Kernel, "Invalid max cores specified! core_num_max={}", core_num_max); return ERR_INVALID_PROCESSOR_ID; } @@ -259,6 +280,7 @@ ResultCode ProcessCapabilities::HandleSyscallFlags(u32& set_svc_bits, u32 flags) } if (svc_number >= svc_capabilities.size()) { + LOG_ERROR(Kernel, "Process svc capability is out of range! svc_number={}", svc_number); return ERR_OUT_OF_RANGE; } @@ -295,6 +317,8 @@ ResultCode ProcessCapabilities::HandleInterruptFlags(u32 flags) { // emulate that, it's sufficient to mark every interrupt as defined. if (interrupt >= interrupt_capabilities.size()) { + LOG_ERROR(Kernel, "Process interrupt capability is out of range! svc_number={}", + interrupt); return ERR_OUT_OF_RANGE; } @@ -307,6 +331,7 @@ ResultCode ProcessCapabilities::HandleInterruptFlags(u32 flags) { ResultCode ProcessCapabilities::HandleProgramTypeFlags(u32 flags) { const u32 reserved = flags >> 17; if (reserved != 0) { + LOG_ERROR(Kernel, "Reserved value is non-zero! reserved={}", reserved); return ERR_RESERVED_VALUE; } @@ -324,6 +349,9 @@ ResultCode ProcessCapabilities::HandleKernelVersionFlags(u32 flags) { const u32 major_version = kernel_version >> 19; if (major_version != 0 || flags < 0x80000) { + LOG_ERROR(Kernel, + "Kernel version is non zero or flags are too small! major_version={}, flags={}", + major_version, flags); return ERR_INVALID_CAPABILITY_DESCRIPTOR; } @@ -334,6 +362,7 @@ ResultCode ProcessCapabilities::HandleKernelVersionFlags(u32 flags) { ResultCode ProcessCapabilities::HandleHandleTableFlags(u32 flags) { const u32 reserved = flags >> 26; if (reserved != 0) { + LOG_ERROR(Kernel, "Reserved value is non-zero! reserved={}", reserved); return ERR_RESERVED_VALUE; } @@ -344,6 +373,7 @@ ResultCode ProcessCapabilities::HandleHandleTableFlags(u32 flags) { ResultCode ProcessCapabilities::HandleDebugFlags(u32 flags) { const u32 reserved = flags >> 19; if (reserved != 0) { + LOG_ERROR(Kernel, "Reserved value is non-zero! reserved={}", reserved); return ERR_RESERVED_VALUE; } diff --git a/src/core/hle/kernel/readable_event.cpp b/src/core/hle/kernel/readable_event.cpp index 9d3d3a81b..00860fcbd 100644 --- a/src/core/hle/kernel/readable_event.cpp +++ b/src/core/hle/kernel/readable_event.cpp @@ -4,6 +4,7 @@ #include <algorithm> #include "common/assert.h" +#include "common/logging/log.h" #include "core/hle/kernel/errors.h" #include "core/hle/kernel/object.h" #include "core/hle/kernel/readable_event.h" @@ -23,10 +24,12 @@ void ReadableEvent::Acquire(Thread* thread) { } void ReadableEvent::Signal() { - if (!is_signaled) { - is_signaled = true; - SynchronizationObject::Signal(); - }; + if (is_signaled) { + return; + } + + is_signaled = true; + SynchronizationObject::Signal(); } void ReadableEvent::Clear() { @@ -35,6 +38,8 @@ void ReadableEvent::Clear() { ResultCode ReadableEvent::Reset() { if (!is_signaled) { + LOG_ERROR(Kernel, "Handle is not signaled! object_id={}, object_type={}, object_name={}", + GetObjectId(), GetTypeName(), GetName()); return ERR_INVALID_STATE; } diff --git a/src/core/hle/kernel/resource_limit.cpp b/src/core/hle/kernel/resource_limit.cpp index 96e5b9892..d9beaa3a4 100644 --- a/src/core/hle/kernel/resource_limit.cpp +++ b/src/core/hle/kernel/resource_limit.cpp @@ -69,6 +69,8 @@ ResultCode ResourceLimit::SetLimitValue(ResourceType resource, s64 value) { limit[index] = value; return RESULT_SUCCESS; } else { + LOG_ERROR(Kernel, "Limit value is too large! resource={}, value={}, index={}", + static_cast<u32>(resource), value, index); return ERR_INVALID_STATE; } } diff --git a/src/core/hle/kernel/svc.cpp b/src/core/hle/kernel/svc.cpp index 41ef2caf6..4ae4529f5 100644 --- a/src/core/hle/kernel/svc.cpp +++ b/src/core/hle/kernel/svc.cpp @@ -685,6 +685,8 @@ static ResultCode GetInfo(Core::System& system, u64* result, u64 info_id, u64 ha case GetInfoType::TotalPhysicalMemoryAvailableWithoutSystemResource: case GetInfoType::TotalPhysicalMemoryUsedWithoutSystemResource: { if (info_sub_id != 0) { + LOG_ERROR(Kernel_SVC, "Info sub id is non zero! info_id={}, info_sub_id={}", info_id, + info_sub_id); return ERR_INVALID_ENUM_VALUE; } @@ -692,6 +694,8 @@ static ResultCode GetInfo(Core::System& system, u64* result, u64 info_id, u64 ha system.Kernel().CurrentProcess()->GetHandleTable(); const auto process = current_process_handle_table.Get<Process>(static_cast<Handle>(handle)); if (!process) { + LOG_ERROR(Kernel_SVC, "Process is not valid! info_id={}, info_sub_id={}, handle={:08X}", + info_id, info_sub_id, handle); return ERR_INVALID_HANDLE; } @@ -783,10 +787,13 @@ static ResultCode GetInfo(Core::System& system, u64* result, u64 info_id, u64 ha case GetInfoType::RegisterResourceLimit: { if (handle != 0) { + LOG_ERROR(Kernel, "Handle is non zero! handle={:08X}", handle); return ERR_INVALID_HANDLE; } if (info_sub_id != 0) { + LOG_ERROR(Kernel, "Info sub id is non zero! info_id={}, info_sub_id={}", info_id, + info_sub_id); return ERR_INVALID_COMBINATION; } diff --git a/src/core/hle/kernel/thread.cpp b/src/core/hle/kernel/thread.cpp index a919750a6..db7f379ac 100644 --- a/src/core/hle/kernel/thread.cpp +++ b/src/core/hle/kernel/thread.cpp @@ -423,6 +423,8 @@ ResultCode Thread::SetCoreAndAffinityMask(s32 new_core, u64 new_affinity_mask) { if (new_core == THREADPROCESSORID_DONT_UPDATE) { new_core = use_override ? ideal_core_override : ideal_core; if ((new_affinity_mask & (1ULL << new_core)) == 0) { + LOG_ERROR(Kernel, "New affinity mask is incorrect! new_core={}, new_affinity_mask={}", + new_core, new_affinity_mask); return ERR_INVALID_COMBINATION; } } diff --git a/src/core/hle/service/acc/acc.cpp b/src/core/hle/service/acc/acc.cpp index cfac8ca9a..630a8b048 100644 --- a/src/core/hle/service/acc/acc.cpp +++ b/src/core/hle/service/acc/acc.cpp @@ -228,7 +228,8 @@ public: class IManagerForApplication final : public ServiceFramework<IManagerForApplication> { public: - IManagerForApplication() : ServiceFramework("IManagerForApplication") { + explicit IManagerForApplication(Common::UUID user_id) + : ServiceFramework("IManagerForApplication"), user_id(user_id) { // clang-format off static const FunctionInfo functions[] = { {0, &IManagerForApplication::CheckAvailability, "CheckAvailability"}, @@ -254,12 +255,14 @@ private: } void GetAccountId(Kernel::HLERequestContext& ctx) { - LOG_WARNING(Service_ACC, "(STUBBED) called"); - // Should return a nintendo account ID + LOG_DEBUG(Service_ACC, "called"); + IPC::ResponseBuilder rb{ctx, 4}; rb.Push(RESULT_SUCCESS); - rb.PushRaw<u64>(1); + rb.PushRaw<u64>(user_id.GetNintendoID()); } + + Common::UUID user_id; }; void Module::Interface::GetUserCount(Kernel::HLERequestContext& ctx) { @@ -319,46 +322,37 @@ void Module::Interface::IsUserRegistrationRequestPermitted(Kernel::HLERequestCon void Module::Interface::InitializeApplicationInfo(Kernel::HLERequestContext& ctx) { IPC::RequestParser rp{ctx}; - auto pid = rp.Pop<u64>(); - LOG_DEBUG(Service_ACC, "called, process_id={}", pid); + LOG_DEBUG(Service_ACC, "called"); IPC::ResponseBuilder rb{ctx, 2}; - rb.Push(InitializeApplicationInfoBase(pid)); + rb.Push(InitializeApplicationInfoBase()); } void Module::Interface::InitializeApplicationInfoRestricted(Kernel::HLERequestContext& ctx) { IPC::RequestParser rp{ctx}; - auto pid = rp.Pop<u64>(); - LOG_WARNING(Service_ACC, "(Partial implementation) called, process_id={}", pid); + LOG_WARNING(Service_ACC, "(Partial implementation) called"); // TODO(ogniK): We require checking if the user actually owns the title and what not. As of // currently, we assume the user owns the title. InitializeApplicationInfoBase SHOULD be called // first then we do extra checks if the game is a digital copy. IPC::ResponseBuilder rb{ctx, 2}; - rb.Push(InitializeApplicationInfoBase(pid)); + rb.Push(InitializeApplicationInfoBase()); } -ResultCode Module::Interface::InitializeApplicationInfoBase(u64 process_id) { +ResultCode Module::Interface::InitializeApplicationInfoBase() { if (application_info) { LOG_ERROR(Service_ACC, "Application already initialized"); return ERR_ACCOUNTINFO_ALREADY_INITIALIZED; } - const auto& list = system.Kernel().GetProcessList(); - const auto iter = std::find_if(list.begin(), list.end(), [&process_id](const auto& process) { - return process->GetProcessID() == process_id; - }); - - if (iter == list.end()) { - LOG_ERROR(Service_ACC, "Failed to find process ID"); - application_info.application_type = ApplicationType::Unknown; - - return ERR_ACCOUNTINFO_BAD_APPLICATION; - } - - const auto launch_property = system.GetARPManager().GetLaunchProperty((*iter)->GetTitleID()); + // TODO(ogniK): This should be changed to reflect the target process for when we have multiple + // processes emulated. As we don't actually have pid support we should assume we're just using + // our own process + const auto& current_process = system.Kernel().CurrentProcess(); + const auto launch_property = + system.GetARPManager().GetLaunchProperty(current_process->GetTitleID()); if (launch_property.Failed()) { LOG_ERROR(Service_ACC, "Failed to get launch property"); @@ -372,10 +366,12 @@ ResultCode Module::Interface::InitializeApplicationInfoBase(u64 process_id) { case FileSys::StorageId::Host: case FileSys::StorageId::NandUser: case FileSys::StorageId::SdCard: + case FileSys::StorageId::None: // Yuzu specific, differs from hardware application_info.application_type = ApplicationType::Digital; break; default: - LOG_ERROR(Service_ACC, "Invalid game storage ID"); + LOG_ERROR(Service_ACC, "Invalid game storage ID! storage_id={}", + launch_property->base_game_storage_id); return ERR_ACCOUNTINFO_BAD_APPLICATION; } @@ -389,7 +385,7 @@ void Module::Interface::GetBaasAccountManagerForApplication(Kernel::HLERequestCo LOG_DEBUG(Service_ACC, "called"); IPC::ResponseBuilder rb{ctx, 2, 0, 1}; rb.Push(RESULT_SUCCESS); - rb.PushIpcInterface<IManagerForApplication>(); + rb.PushIpcInterface<IManagerForApplication>(profile_manager->GetLastOpenedUser()); } void Module::Interface::IsUserAccountSwitchLocked(Kernel::HLERequestContext& ctx) { @@ -428,6 +424,17 @@ void Module::Interface::GetProfileEditor(Kernel::HLERequestContext& ctx) { rb.PushIpcInterface<IProfileEditor>(user_id, *profile_manager); } +void Module::Interface::ListQualifiedUsers(Kernel::HLERequestContext& ctx) { + LOG_DEBUG(Service_ACC, "called"); + + // All users should be qualified. We don't actually have parental control or anything to do with + // nintendo online currently. We're just going to assume the user running the game has access to + // the game regardless of parental control settings. + ctx.WriteBuffer(profile_manager->GetAllUsers()); + IPC::ResponseBuilder rb{ctx, 2}; + rb.Push(RESULT_SUCCESS); +} + void Module::Interface::TrySelectUserWithoutInteraction(Kernel::HLERequestContext& ctx) { LOG_DEBUG(Service_ACC, "called"); // A u8 is passed into this function which we can safely ignore. It's to determine if we have diff --git a/src/core/hle/service/acc/acc.h b/src/core/hle/service/acc/acc.h index 7a7dc9ec6..74ca39d6e 100644 --- a/src/core/hle/service/acc/acc.h +++ b/src/core/hle/service/acc/acc.h @@ -33,9 +33,10 @@ public: void TrySelectUserWithoutInteraction(Kernel::HLERequestContext& ctx); void IsUserAccountSwitchLocked(Kernel::HLERequestContext& ctx); void GetProfileEditor(Kernel::HLERequestContext& ctx); + void ListQualifiedUsers(Kernel::HLERequestContext& ctx); private: - ResultCode InitializeApplicationInfoBase(u64 process_id); + ResultCode InitializeApplicationInfoBase(); enum class ApplicationType : u32_le { GameCard = 0, diff --git a/src/core/hle/service/acc/acc_su.cpp b/src/core/hle/service/acc/acc_su.cpp index ae88deda5..2eefc6df5 100644 --- a/src/core/hle/service/acc/acc_su.cpp +++ b/src/core/hle/service/acc/acc_su.cpp @@ -35,7 +35,7 @@ ACC_SU::ACC_SU(std::shared_ptr<Module> module, std::shared_ptr<ProfileManager> p {113, nullptr, "GetSaveDataThumbnailExistence"}, {120, nullptr, "ListOpenUsersInApplication"}, {130, nullptr, "ActivateOpenContextRetention"}, - {140, nullptr, "ListQualifiedUsers"}, + {140, &ACC_SU::ListQualifiedUsers, "ListQualifiedUsers"}, {150, nullptr, "AuthenticateApplicationAsync"}, {190, nullptr, "GetUserLastOpenedApplication"}, {191, nullptr, "ActivateOpenContextHolder"}, diff --git a/src/core/hle/service/acc/acc_u0.cpp b/src/core/hle/service/acc/acc_u0.cpp index 0ac19f4ff..fb4e7e772 100644 --- a/src/core/hle/service/acc/acc_u0.cpp +++ b/src/core/hle/service/acc/acc_u0.cpp @@ -32,7 +32,7 @@ ACC_U0::ACC_U0(std::shared_ptr<Module> module, std::shared_ptr<ProfileManager> p {130, nullptr, "LoadOpenContext"}, {131, nullptr, "ListOpenContextStoredUsers"}, {140, &ACC_U0::InitializeApplicationInfoRestricted, "InitializeApplicationInfoRestricted"}, - {141, nullptr, "ListQualifiedUsers"}, + {141, &ACC_U0::ListQualifiedUsers, "ListQualifiedUsers"}, {150, &ACC_U0::IsUserAccountSwitchLocked, "IsUserAccountSwitchLocked"}, }; // clang-format on diff --git a/src/core/hle/service/acc/acc_u1.cpp b/src/core/hle/service/acc/acc_u1.cpp index 2b9c11928..9f29cdc82 100644 --- a/src/core/hle/service/acc/acc_u1.cpp +++ b/src/core/hle/service/acc/acc_u1.cpp @@ -34,7 +34,7 @@ ACC_U1::ACC_U1(std::shared_ptr<Module> module, std::shared_ptr<ProfileManager> p {112, nullptr, "LoadSaveDataThumbnail"}, {113, nullptr, "GetSaveDataThumbnailExistence"}, {130, nullptr, "ActivateOpenContextRetention"}, - {140, nullptr, "ListQualifiedUsers"}, + {140, &ACC_U1::ListQualifiedUsers, "ListQualifiedUsers"}, {150, nullptr, "AuthenticateApplicationAsync"}, {190, nullptr, "GetUserLastOpenedApplication"}, {191, nullptr, "ActivateOpenContextHolder"}, diff --git a/src/core/hle/service/am/am.cpp b/src/core/hle/service/am/am.cpp index 30cd73252..4df74c4f9 100644 --- a/src/core/hle/service/am/am.cpp +++ b/src/core/hle/service/am/am.cpp @@ -1335,12 +1335,23 @@ void IApplicationFunctions::SetTerminateResult(Kernel::HLERequestContext& ctx) { } void IApplicationFunctions::GetDisplayVersion(Kernel::HLERequestContext& ctx) { - LOG_WARNING(Service_AM, "(STUBBED) called"); + LOG_DEBUG(Service_AM, "called"); + + std::array<u8, 0x10> version_string{}; + + FileSys::PatchManager pm{system.CurrentProcess()->GetTitleID()}; + const auto res = pm.GetControlMetadata(); + if (res.first != nullptr) { + const auto& version = res.first->GetVersionString(); + std::copy(version.begin(), version.end(), version_string.begin()); + } else { + constexpr u128 default_version = {1, 0}; + std::memcpy(version_string.data(), default_version.data(), sizeof(u128)); + } IPC::ResponseBuilder rb{ctx, 6}; rb.Push(RESULT_SUCCESS); - rb.Push<u64>(1); - rb.Push<u64>(0); + rb.PushRaw(version_string); } void IApplicationFunctions::GetDesiredLanguage(Kernel::HLERequestContext& ctx) { diff --git a/src/core/hle/service/audio/audin_u.cpp b/src/core/hle/service/audio/audin_u.cpp index d7f1d348d..3e2299426 100644 --- a/src/core/hle/service/audio/audin_u.cpp +++ b/src/core/hle/service/audio/audin_u.cpp @@ -2,6 +2,9 @@ // Licensed under GPLv2 or any later version // Refer to the license.txt file included. +#include "common/logging/log.h" +#include "core/hle/ipc_helpers.h" +#include "core/hle/kernel/hle_ipc.h" #include "core/hle/service/audio/audin_u.h" namespace Service::Audio { @@ -36,11 +39,12 @@ public: AudInU::AudInU() : ServiceFramework("audin:u") { // clang-format off static const FunctionInfo functions[] = { - {0, nullptr, "ListAudioIns"}, - {1, nullptr, "OpenAudioIn"}, - {2, nullptr, "Unknown"}, - {3, nullptr, "OpenAudioInAuto"}, - {4, nullptr, "ListAudioInsAuto"}, + {0, &AudInU::ListAudioIns, "ListAudioIns"}, + {1, &AudInU::OpenAudioIn, "OpenAudioIn"}, + {2, &AudInU::ListAudioIns, "ListAudioInsAuto"}, + {3, &AudInU::OpenAudioIn, "OpenAudioInAuto"}, + {4, &AudInU::ListAudioInsAutoFiltered, "ListAudioInsAutoFiltered"}, + {5, &AudInU::OpenAudioInProtocolSpecified, "OpenAudioInProtocolSpecified"}, }; // clang-format on @@ -49,4 +53,60 @@ AudInU::AudInU() : ServiceFramework("audin:u") { AudInU::~AudInU() = default; +void AudInU::ListAudioIns(Kernel::HLERequestContext& ctx) { + LOG_DEBUG(Service_Audio, "called"); + const std::size_t count = ctx.GetWriteBufferSize() / sizeof(AudioInDeviceName); + + const std::size_t device_count = std::min(count, audio_device_names.size()); + std::vector<AudioInDeviceName> device_names; + device_names.reserve(device_count); + + for (std::size_t i = 0; i < device_count; i++) { + const auto& device_name = audio_device_names[i]; + auto& entry = device_names.emplace_back(); + device_name.copy(entry.data(), device_name.size()); + } + + ctx.WriteBuffer(device_names); + + IPC::ResponseBuilder rb{ctx, 3}; + rb.Push(RESULT_SUCCESS); + rb.Push(static_cast<u32>(device_names.size())); +} + +void AudInU::ListAudioInsAutoFiltered(Kernel::HLERequestContext& ctx) { + LOG_DEBUG(Service_Audio, "called"); + constexpr u32 device_count = 0; + + // Since we don't actually use any other audio input devices, we return 0 devices. Filtered + // device listing just omits the default input device + + IPC::ResponseBuilder rb{ctx, 3}; + rb.Push(RESULT_SUCCESS); + rb.Push(static_cast<u32>(device_count)); +} + +void AudInU::OpenInOutImpl(Kernel::HLERequestContext& ctx) { + AudInOutParams params{}; + params.channel_count = 2; + params.sample_format = SampleFormat::PCM16; + params.sample_rate = 48000; + params.state = State::Started; + + IPC::ResponseBuilder rb{ctx, 6, 0, 1}; + rb.Push(RESULT_SUCCESS); + rb.PushRaw<AudInOutParams>(params); + rb.PushIpcInterface<IAudioIn>(); +} + +void AudInU::OpenAudioIn(Kernel::HLERequestContext& ctx) { + LOG_WARNING(Service_Audio, "(STUBBED) called"); + OpenInOutImpl(ctx); +} + +void AudInU::OpenAudioInProtocolSpecified(Kernel::HLERequestContext& ctx) { + LOG_WARNING(Service_Audio, "(STUBBED) called"); + OpenInOutImpl(ctx); +} + } // namespace Service::Audio diff --git a/src/core/hle/service/audio/audin_u.h b/src/core/hle/service/audio/audin_u.h index 0538b9560..a599f4a64 100644 --- a/src/core/hle/service/audio/audin_u.h +++ b/src/core/hle/service/audio/audin_u.h @@ -16,6 +16,35 @@ class AudInU final : public ServiceFramework<AudInU> { public: explicit AudInU(); ~AudInU() override; + +private: + enum class SampleFormat : u32_le { + PCM16 = 2, + }; + + enum class State : u32_le { + Started = 0, + Stopped = 1, + }; + + struct AudInOutParams { + u32_le sample_rate{}; + u32_le channel_count{}; + SampleFormat sample_format{}; + State state{}; + }; + static_assert(sizeof(AudInOutParams) == 0x10, "AudInOutParams is an invalid size"); + + using AudioInDeviceName = std::array<char, 256>; + static constexpr std::array<std::string_view, 1> audio_device_names{{ + "BuiltInHeadset", + }}; + + void ListAudioIns(Kernel::HLERequestContext& ctx); + void ListAudioInsAutoFiltered(Kernel::HLERequestContext& ctx); + void OpenInOutImpl(Kernel::HLERequestContext& ctx); + void OpenAudioIn(Kernel::HLERequestContext& ctx); + void OpenAudioInProtocolSpecified(Kernel::HLERequestContext& ctx); }; } // namespace Service::Audio diff --git a/src/core/hle/service/caps/caps_su.cpp b/src/core/hle/service/caps/caps_su.cpp index b4d9355ef..e8b0698e8 100644 --- a/src/core/hle/service/caps/caps_su.cpp +++ b/src/core/hle/service/caps/caps_su.cpp @@ -2,6 +2,8 @@ // Licensed under GPLv2 or any later version // Refer to the license.txt file included. +#include "common/logging/log.h" +#include "core/hle/ipc_helpers.h" #include "core/hle/service/caps/caps_su.h" namespace Service::Capture { @@ -9,7 +11,7 @@ namespace Service::Capture { CAPS_SU::CAPS_SU() : ServiceFramework("caps:su") { // clang-format off static const FunctionInfo functions[] = { - {32, nullptr, "SetShimLibraryVersion"}, + {32, &CAPS_SU::SetShimLibraryVersion, "SetShimLibraryVersion"}, {201, nullptr, "SaveScreenShot"}, {203, nullptr, "SaveScreenShotEx0"}, {205, nullptr, "SaveScreenShotEx1"}, @@ -22,4 +24,11 @@ CAPS_SU::CAPS_SU() : ServiceFramework("caps:su") { CAPS_SU::~CAPS_SU() = default; +void CAPS_SU::SetShimLibraryVersion(Kernel::HLERequestContext& ctx) { + LOG_WARNING(Service_Capture, "(STUBBED) called"); + + IPC::ResponseBuilder rb{ctx, 2}; + rb.Push(RESULT_SUCCESS); +} + } // namespace Service::Capture diff --git a/src/core/hle/service/caps/caps_su.h b/src/core/hle/service/caps/caps_su.h index cb11f7c9a..c494d7c84 100644 --- a/src/core/hle/service/caps/caps_su.h +++ b/src/core/hle/service/caps/caps_su.h @@ -16,6 +16,9 @@ class CAPS_SU final : public ServiceFramework<CAPS_SU> { public: explicit CAPS_SU(); ~CAPS_SU() override; + +private: + void SetShimLibraryVersion(Kernel::HLERequestContext& ctx); }; } // namespace Service::Capture diff --git a/src/core/hle/service/nvdrv/devices/nvhost_ctrl_gpu.h b/src/core/hle/service/nvdrv/devices/nvhost_ctrl_gpu.h index 642b0a2cb..07b644ec5 100644 --- a/src/core/hle/service/nvdrv/devices/nvhost_ctrl_gpu.h +++ b/src/core/hle/service/nvdrv/devices/nvhost_ctrl_gpu.h @@ -159,9 +159,10 @@ private: static_assert(sizeof(IoctlFlushL2) == 8, "IoctlFlushL2 is incorrect size"); struct IoctlGetGpuTime { - u64_le gpu_time; + u64_le gpu_time{}; + INSERT_PADDING_WORDS(2); }; - static_assert(sizeof(IoctlGetGpuTime) == 8, "IoctlGetGpuTime is incorrect size"); + static_assert(sizeof(IoctlGetGpuTime) == 0x10, "IoctlGetGpuTime is incorrect size"); u32 GetCharacteristics(const std::vector<u8>& input, std::vector<u8>& output, std::vector<u8>& output2, IoctlVersion version); diff --git a/src/core/settings.cpp b/src/core/settings.cpp index cd6c257f5..2b0bdc4d3 100644 --- a/src/core/settings.cpp +++ b/src/core/settings.cpp @@ -96,6 +96,7 @@ void LogSettings() { LogSetting("Renderer_UseAsynchronousGpuEmulation", Settings::values.use_asynchronous_gpu_emulation); LogSetting("Renderer_UseVsync", Settings::values.use_vsync); + LogSetting("Renderer_AnisotropicFilteringLevel", Settings::values.max_anisotropy); LogSetting("Audio_OutputEngine", Settings::values.sink_id); LogSetting("Audio_EnableAudioStretching", Settings::values.enable_audio_stretching); LogSetting("Audio_OutputDevice", Settings::values.audio_device_id); diff --git a/src/input_common/main.cpp b/src/input_common/main.cpp index c98c848cf..95e351e24 100644 --- a/src/input_common/main.cpp +++ b/src/input_common/main.cpp @@ -18,7 +18,9 @@ namespace InputCommon { static std::shared_ptr<Keyboard> keyboard; static std::shared_ptr<MotionEmu> motion_emu; +#ifdef HAVE_SDL2 static std::unique_ptr<SDL::State> sdl; +#endif static std::unique_ptr<CemuhookUDP::State> udp; void Init() { @@ -29,7 +31,9 @@ void Init() { motion_emu = std::make_shared<MotionEmu>(); Input::RegisterFactory<Input::MotionDevice>("motion_emu", motion_emu); +#ifdef HAVE_SDL2 sdl = SDL::Init(); +#endif udp = CemuhookUDP::Init(); } @@ -40,7 +44,9 @@ void Shutdown() { Input::UnregisterFactory<Input::AnalogDevice>("analog_from_button"); Input::UnregisterFactory<Input::MotionDevice>("motion_emu"); motion_emu.reset(); +#ifdef HAVE_SDL2 sdl.reset(); +#endif udp.reset(); } diff --git a/src/video_core/buffer_cache/buffer_cache.h b/src/video_core/buffer_cache/buffer_cache.h index 510f11089..56e570994 100644 --- a/src/video_core/buffer_cache/buffer_cache.h +++ b/src/video_core/buffer_cache/buffer_cache.h @@ -4,7 +4,6 @@ #pragma once -#include <array> #include <list> #include <memory> #include <mutex> @@ -89,10 +88,6 @@ public: map->MarkAsWritten(true); MarkRegionAsWritten(map->GetStart(), map->GetEnd() - 1); } - } else { - if (map->IsWritten()) { - WriteBarrier(); - } } return {ToHandle(block), static_cast<u64>(block->GetOffset(cpu_addr))}; @@ -254,8 +249,6 @@ protected: virtual BufferType ToHandle(const OwnerBuffer& storage) = 0; - virtual void WriteBarrier() = 0; - virtual OwnerBuffer CreateBlock(VAddr cpu_addr, std::size_t size) = 0; virtual void UploadBlockData(const OwnerBuffer& buffer, std::size_t offset, std::size_t size, diff --git a/src/video_core/engines/maxwell_3d.cpp b/src/video_core/engines/maxwell_3d.cpp index 39e3b66a2..7db055ea0 100644 --- a/src/video_core/engines/maxwell_3d.cpp +++ b/src/video_core/engines/maxwell_3d.cpp @@ -184,6 +184,10 @@ void Maxwell3D::CallMethod(const GPU::MethodCall& method_call) { } switch (method) { + case MAXWELL3D_REG_INDEX(wait_for_idle): { + rasterizer.WaitForIdle(); + break; + } case MAXWELL3D_REG_INDEX(shadow_ram_control): { shadow_state.shadow_ram_control = static_cast<Regs::ShadowRamControl>(method_call.argument); break; diff --git a/src/video_core/engines/maxwell_3d.h b/src/video_core/engines/maxwell_3d.h index 5e522e0d2..864924ff3 100644 --- a/src/video_core/engines/maxwell_3d.h +++ b/src/video_core/engines/maxwell_3d.h @@ -709,7 +709,9 @@ public: union { struct { - INSERT_UNION_PADDING_WORDS(0x45); + INSERT_UNION_PADDING_WORDS(0x44); + + u32 wait_for_idle; struct { u32 upload_address; @@ -1536,6 +1538,7 @@ private: static_assert(offsetof(Maxwell3D::Regs, field_name) == position * 4, \ "Field " #field_name " has invalid position") +ASSERT_REG_POSITION(wait_for_idle, 0x44); ASSERT_REG_POSITION(macros, 0x45); ASSERT_REG_POSITION(shadow_ram_control, 0x49); ASSERT_REG_POSITION(upload, 0x60); diff --git a/src/video_core/fence_manager.h b/src/video_core/fence_manager.h index dabd1588c..8b2a6a42c 100644 --- a/src/video_core/fence_manager.h +++ b/src/video_core/fence_manager.h @@ -88,7 +88,8 @@ public: } PopAsyncFlushes(); if (current_fence->IsSemaphore()) { - memory_manager.Write<u32>(current_fence->GetAddress(), current_fence->GetPayload()); + memory_manager.template Write<u32>(current_fence->GetAddress(), + current_fence->GetPayload()); } else { gpu.IncrementSyncPoint(current_fence->GetPayload()); } @@ -134,7 +135,8 @@ private: } PopAsyncFlushes(); if (current_fence->IsSemaphore()) { - memory_manager.Write<u32>(current_fence->GetAddress(), current_fence->GetPayload()); + memory_manager.template Write<u32>(current_fence->GetAddress(), + current_fence->GetPayload()); } else { gpu.IncrementSyncPoint(current_fence->GetPayload()); } diff --git a/src/video_core/rasterizer_interface.h b/src/video_core/rasterizer_interface.h index 603f61952..3cbdac8e7 100644 --- a/src/video_core/rasterizer_interface.h +++ b/src/video_core/rasterizer_interface.h @@ -80,6 +80,9 @@ public: /// and invalidated virtual void FlushAndInvalidateRegion(VAddr addr, u64 size) = 0; + /// Notify the host renderer to wait for previous primitive and compute operations. + virtual void WaitForIdle() = 0; + /// Notify the rasterizer to send all written commands to the host GPU. virtual void FlushCommands() = 0; diff --git a/src/video_core/renderer_opengl/gl_buffer_cache.cpp b/src/video_core/renderer_opengl/gl_buffer_cache.cpp index 4efce0de7..d2cab50bd 100644 --- a/src/video_core/renderer_opengl/gl_buffer_cache.cpp +++ b/src/video_core/renderer_opengl/gl_buffer_cache.cpp @@ -51,10 +51,6 @@ Buffer OGLBufferCache::CreateBlock(VAddr cpu_addr, std::size_t size) { return std::make_shared<CachedBufferBlock>(cpu_addr, size); } -void OGLBufferCache::WriteBarrier() { - glMemoryBarrier(GL_SHADER_STORAGE_BARRIER_BIT); -} - GLuint OGLBufferCache::ToHandle(const Buffer& buffer) { return buffer->GetHandle(); } diff --git a/src/video_core/renderer_opengl/gl_buffer_cache.h b/src/video_core/renderer_opengl/gl_buffer_cache.h index a74817857..a9e86cfc7 100644 --- a/src/video_core/renderer_opengl/gl_buffer_cache.h +++ b/src/video_core/renderer_opengl/gl_buffer_cache.h @@ -59,8 +59,6 @@ protected: GLuint ToHandle(const Buffer& buffer) override; - void WriteBarrier() override; - void UploadBlockData(const Buffer& buffer, std::size_t offset, std::size_t size, const u8* data) override; diff --git a/src/video_core/renderer_opengl/gl_rasterizer.cpp b/src/video_core/renderer_opengl/gl_rasterizer.cpp index f33c4a8f9..8b3b3ce92 100644 --- a/src/video_core/renderer_opengl/gl_rasterizer.cpp +++ b/src/video_core/renderer_opengl/gl_rasterizer.cpp @@ -59,14 +59,12 @@ constexpr std::size_t NumSupportedVertexAttributes = 16; template <typename Engine, typename Entry> Tegra::Texture::FullTextureInfo GetTextureInfo(const Engine& engine, const Entry& entry, ShaderType shader_type, std::size_t index = 0) { - if (entry.IsBindless()) { - const Tegra::Texture::TextureHandle tex_handle = - engine.AccessConstBuffer32(shader_type, entry.GetBuffer(), entry.GetOffset()); + if (entry.is_bindless) { + const auto tex_handle = engine.AccessConstBuffer32(shader_type, entry.buffer, entry.offset); return engine.GetTextureInfo(tex_handle); } const auto& gpu_profile = engine.AccessGuestDriverProfile(); - const u32 offset = - entry.GetOffset() + static_cast<u32>(index * gpu_profile.GetTextureHandlerSize()); + const u32 offset = entry.offset + static_cast<u32>(index * gpu_profile.GetTextureHandlerSize()); if constexpr (std::is_same_v<Engine, Tegra::Engines::Maxwell3D>) { return engine.GetStageTexture(shader_type, offset); } else { @@ -748,6 +746,17 @@ void RasterizerOpenGL::FlushAndInvalidateRegion(VAddr addr, u64 size) { InvalidateRegion(addr, size); } +void RasterizerOpenGL::WaitForIdle() { + // Place a barrier on everything that is not framebuffer related. + // This is related to another flag that is not currently implemented. + glMemoryBarrier(GL_VERTEX_ATTRIB_ARRAY_BARRIER_BIT | GL_ELEMENT_ARRAY_BARRIER_BIT | + GL_UNIFORM_BARRIER_BIT | GL_TEXTURE_FETCH_BARRIER_BIT | + GL_SHADER_IMAGE_ACCESS_BARRIER_BIT | GL_COMMAND_BARRIER_BIT | + GL_PIXEL_BUFFER_BARRIER_BIT | GL_TEXTURE_UPDATE_BARRIER_BIT | + GL_BUFFER_UPDATE_BARRIER_BIT | GL_TRANSFORM_FEEDBACK_BARRIER_BIT | + GL_SHADER_STORAGE_BARRIER_BIT | GL_QUERY_BUFFER_BARRIER_BIT); +} + void RasterizerOpenGL::FlushCommands() { // Only flush when we have commands queued to OpenGL. if (num_queued_commands == 0) { @@ -856,9 +865,9 @@ void RasterizerOpenGL::SetupDrawGlobalMemory(std::size_t stage_index, const Shad u32 binding = device.GetBaseBindings(stage_index).shader_storage_buffer; for (const auto& entry : shader->GetEntries().global_memory_entries) { - const auto addr{cbufs.const_buffers[entry.GetCbufIndex()].address + entry.GetCbufOffset()}; - const auto gpu_addr{memory_manager.Read<u64>(addr)}; - const auto size{memory_manager.Read<u32>(addr + 8)}; + const GPUVAddr addr{cbufs.const_buffers[entry.cbuf_index].address + entry.cbuf_offset}; + const GPUVAddr gpu_addr{memory_manager.Read<u64>(addr)}; + const u32 size{memory_manager.Read<u32>(addr + 8)}; SetupGlobalMemory(binding++, entry, gpu_addr, size); } } @@ -870,7 +879,7 @@ void RasterizerOpenGL::SetupComputeGlobalMemory(const Shader& kernel) { u32 binding = 0; for (const auto& entry : kernel->GetEntries().global_memory_entries) { - const auto addr{cbufs[entry.GetCbufIndex()].Address() + entry.GetCbufOffset()}; + const auto addr{cbufs[entry.cbuf_index].Address() + entry.cbuf_offset}; const auto gpu_addr{memory_manager.Read<u64>(addr)}; const auto size{memory_manager.Read<u32>(addr + 8)}; SetupGlobalMemory(binding++, entry, gpu_addr, size); @@ -881,7 +890,7 @@ void RasterizerOpenGL::SetupGlobalMemory(u32 binding, const GlobalMemoryEntry& e GPUVAddr gpu_addr, std::size_t size) { const auto alignment{device.GetShaderStorageBufferAlignment()}; const auto [ssbo, buffer_offset] = - buffer_cache.UploadMemory(gpu_addr, size, alignment, entry.IsWritten()); + buffer_cache.UploadMemory(gpu_addr, size, alignment, entry.is_written); glBindBufferRange(GL_SHADER_STORAGE_BUFFER, binding, ssbo, buffer_offset, static_cast<GLsizeiptr>(size)); } @@ -892,7 +901,7 @@ void RasterizerOpenGL::SetupDrawTextures(std::size_t stage_index, const Shader& u32 binding = device.GetBaseBindings(stage_index).sampler; for (const auto& entry : shader->GetEntries().samplers) { const auto shader_type = static_cast<ShaderType>(stage_index); - for (std::size_t i = 0; i < entry.Size(); ++i) { + for (std::size_t i = 0; i < entry.size; ++i) { const auto texture = GetTextureInfo(maxwell3d, entry, shader_type, i); SetupTexture(binding++, texture, entry); } @@ -904,7 +913,7 @@ void RasterizerOpenGL::SetupComputeTextures(const Shader& kernel) { const auto& compute = system.GPU().KeplerCompute(); u32 binding = 0; for (const auto& entry : kernel->GetEntries().samplers) { - for (std::size_t i = 0; i < entry.Size(); ++i) { + for (std::size_t i = 0; i < entry.size; ++i) { const auto texture = GetTextureInfo(compute, entry, ShaderType::Compute, i); SetupTexture(binding++, texture, entry); } @@ -961,7 +970,7 @@ void RasterizerOpenGL::SetupImage(u32 binding, const Tegra::Texture::TICEntry& t if (!tic.IsBuffer()) { view->ApplySwizzle(tic.x_source, tic.y_source, tic.z_source, tic.w_source); } - if (entry.IsWritten()) { + if (entry.is_written) { view->MarkAsModified(texture_cache.Tick()); } glBindImageTexture(binding, view->GetTexture(), 0, GL_TRUE, 0, GL_READ_WRITE, diff --git a/src/video_core/renderer_opengl/gl_rasterizer.h b/src/video_core/renderer_opengl/gl_rasterizer.h index 87249fb6f..b94c65907 100644 --- a/src/video_core/renderer_opengl/gl_rasterizer.h +++ b/src/video_core/renderer_opengl/gl_rasterizer.h @@ -75,6 +75,7 @@ public: void SignalSyncPoint(u32 value) override; void ReleaseFences() override; void FlushAndInvalidateRegion(VAddr addr, u64 size) override; + void WaitForIdle() override; void FlushCommands() override; void TickFrame() override; bool AccelerateSurfaceCopy(const Tegra::Engines::Fermi2D::Regs::Surface& src, diff --git a/src/video_core/renderer_opengl/gl_shader_decompiler.cpp b/src/video_core/renderer_opengl/gl_shader_decompiler.cpp index 3803a6f3a..99fd4ae2c 100644 --- a/src/video_core/renderer_opengl/gl_shader_decompiler.cpp +++ b/src/video_core/renderer_opengl/gl_shader_decompiler.cpp @@ -870,13 +870,13 @@ private: for (const auto& sampler : ir.GetSamplers()) { const std::string name = GetSampler(sampler); const std::string description = fmt::format("layout (binding = {}) uniform", binding); - binding += sampler.IsIndexed() ? sampler.Size() : 1; + binding += sampler.is_indexed ? sampler.size : 1; std::string sampler_type = [&]() { - if (sampler.IsBuffer()) { + if (sampler.is_buffer) { return "samplerBuffer"; } - switch (sampler.GetType()) { + switch (sampler.type) { case Tegra::Shader::TextureType::Texture1D: return "sampler1D"; case Tegra::Shader::TextureType::Texture2D: @@ -890,17 +890,17 @@ private: return "sampler2D"; } }(); - if (sampler.IsArray()) { + if (sampler.is_array) { sampler_type += "Array"; } - if (sampler.IsShadow()) { + if (sampler.is_shadow) { sampler_type += "Shadow"; } - if (!sampler.IsIndexed()) { + if (!sampler.is_indexed) { code.AddLine("{} {} {};", description, sampler_type, name); } else { - code.AddLine("{} {} {}[{}];", description, sampler_type, name, sampler.Size()); + code.AddLine("{} {} {}[{}];", description, sampler_type, name, sampler.size); } } if (!ir.GetSamplers().empty()) { @@ -946,14 +946,14 @@ private: u32 binding = device.GetBaseBindings(stage).image; for (const auto& image : ir.GetImages()) { std::string qualifier = "coherent volatile"; - if (image.IsRead() && !image.IsWritten()) { + if (image.is_read && !image.is_written) { qualifier += " readonly"; - } else if (image.IsWritten() && !image.IsRead()) { + } else if (image.is_written && !image.is_read) { qualifier += " writeonly"; } - const char* format = image.IsAtomic() ? "r32ui, " : ""; - const char* type_declaration = GetImageTypeDeclaration(image.GetType()); + const char* format = image.is_atomic ? "r32ui, " : ""; + const char* type_declaration = GetImageTypeDeclaration(image.type); code.AddLine("layout ({}binding = {}) {} uniform uimage{} {};", format, binding++, qualifier, type_declaration, GetImage(image)); } @@ -1337,8 +1337,8 @@ private: ASSERT(meta); const std::size_t count = operation.GetOperandsCount(); - const bool has_array = meta->sampler.IsArray(); - const bool has_shadow = meta->sampler.IsShadow(); + const bool has_array = meta->sampler.is_array; + const bool has_shadow = meta->sampler.is_shadow; std::string expr = "texture" + function_suffix; if (!meta->aoffi.empty()) { @@ -1346,7 +1346,7 @@ private: } else if (!meta->ptp.empty()) { expr += "Offsets"; } - if (!meta->sampler.IsIndexed()) { + if (!meta->sampler.is_indexed) { expr += '(' + GetSampler(meta->sampler) + ", "; } else { expr += '(' + GetSampler(meta->sampler) + '[' + Visit(meta->index).AsUint() + "], "; @@ -1982,7 +1982,7 @@ private: std::string expr = GenerateTexture( operation, "", {TextureOffset{}, TextureArgument{Type::Float, meta->bias}}); - if (meta->sampler.IsShadow()) { + if (meta->sampler.is_shadow) { expr = "vec4(" + expr + ')'; } return {expr + GetSwizzle(meta->element), Type::Float}; @@ -1994,7 +1994,7 @@ private: std::string expr = GenerateTexture( operation, "Lod", {TextureArgument{Type::Float, meta->lod}, TextureOffset{}}); - if (meta->sampler.IsShadow()) { + if (meta->sampler.is_shadow) { expr = "vec4(" + expr + ')'; } return {expr + GetSwizzle(meta->element), Type::Float}; @@ -2003,11 +2003,11 @@ private: Expression TextureGather(Operation operation) { const auto& meta = std::get<MetaTexture>(operation.GetMeta()); - const auto type = meta.sampler.IsShadow() ? Type::Float : Type::Int; - const bool separate_dc = meta.sampler.IsShadow(); + const auto type = meta.sampler.is_shadow ? Type::Float : Type::Int; + const bool separate_dc = meta.sampler.is_shadow; std::vector<TextureIR> ir; - if (meta.sampler.IsShadow()) { + if (meta.sampler.is_shadow) { ir = {TextureOffset{}}; } else { ir = {TextureOffset{}, TextureArgument{type, meta.component}}; @@ -2052,7 +2052,7 @@ private: constexpr std::array constructors = {"int", "ivec2", "ivec3", "ivec4"}; const auto meta = std::get_if<MetaTexture>(&operation.GetMeta()); ASSERT(meta); - UNIMPLEMENTED_IF(meta->sampler.IsArray()); + UNIMPLEMENTED_IF(meta->sampler.is_array); const std::size_t count = operation.GetOperandsCount(); std::string expr = "texelFetch("; @@ -2073,7 +2073,7 @@ private: } expr += ')'; - if (meta->lod && !meta->sampler.IsBuffer()) { + if (meta->lod && !meta->sampler.is_buffer) { expr += ", "; expr += Visit(meta->lod).AsInt(); } @@ -2084,12 +2084,10 @@ private: } Expression TextureGradient(Operation operation) { - const auto meta = std::get_if<MetaTexture>(&operation.GetMeta()); - ASSERT(meta); - + const auto& meta = std::get<MetaTexture>(operation.GetMeta()); std::string expr = GenerateTexture(operation, "Grad", {TextureDerivates{}, TextureOffset{}}); - return {std::move(expr) + GetSwizzle(meta->element), Type::Float}; + return {std::move(expr) + GetSwizzle(meta.element), Type::Float}; } Expression ImageLoad(Operation operation) { @@ -2608,11 +2606,11 @@ private: } std::string GetSampler(const Sampler& sampler) const { - return AppendSuffix(static_cast<u32>(sampler.GetIndex()), "sampler"); + return AppendSuffix(sampler.index, "sampler"); } std::string GetImage(const Image& image) const { - return AppendSuffix(static_cast<u32>(image.GetIndex()), "image"); + return AppendSuffix(image.index, "image"); } std::string AppendSuffix(u32 index, std::string_view name) const { diff --git a/src/video_core/renderer_opengl/gl_shader_decompiler.h b/src/video_core/renderer_opengl/gl_shader_decompiler.h index e7dbd810c..e8a178764 100644 --- a/src/video_core/renderer_opengl/gl_shader_decompiler.h +++ b/src/video_core/renderer_opengl/gl_shader_decompiler.h @@ -33,36 +33,19 @@ public: } private: - u32 index{}; + u32 index = 0; }; -class GlobalMemoryEntry { -public: - explicit GlobalMemoryEntry(u32 cbuf_index, u32 cbuf_offset, bool is_read, bool is_written) +struct GlobalMemoryEntry { + constexpr explicit GlobalMemoryEntry(u32 cbuf_index, u32 cbuf_offset, bool is_read, + bool is_written) : cbuf_index{cbuf_index}, cbuf_offset{cbuf_offset}, is_read{is_read}, is_written{ is_written} {} - u32 GetCbufIndex() const { - return cbuf_index; - } - - u32 GetCbufOffset() const { - return cbuf_offset; - } - - bool IsRead() const { - return is_read; - } - - bool IsWritten() const { - return is_written; - } - -private: - u32 cbuf_index{}; - u32 cbuf_offset{}; - bool is_read{}; - bool is_written{}; + u32 cbuf_index = 0; + u32 cbuf_offset = 0; + bool is_read = false; + bool is_written = false; }; struct ShaderEntries { diff --git a/src/video_core/renderer_vulkan/renderer_vulkan.cpp b/src/video_core/renderer_vulkan/renderer_vulkan.cpp index 04532f8f8..59b441943 100644 --- a/src/video_core/renderer_vulkan/renderer_vulkan.cpp +++ b/src/video_core/renderer_vulkan/renderer_vulkan.cpp @@ -12,15 +12,12 @@ #include <fmt/format.h> -#include "common/assert.h" #include "common/dynamic_library.h" #include "common/logging/log.h" #include "common/telemetry.h" #include "core/core.h" #include "core/core_timing.h" #include "core/frontend/emu_window.h" -#include "core/memory.h" -#include "core/perf_stats.h" #include "core/settings.h" #include "core/telemetry_session.h" #include "video_core/gpu.h" diff --git a/src/video_core/renderer_vulkan/renderer_vulkan.h b/src/video_core/renderer_vulkan/renderer_vulkan.h index 18270909b..522b5bff8 100644 --- a/src/video_core/renderer_vulkan/renderer_vulkan.h +++ b/src/video_core/renderer_vulkan/renderer_vulkan.h @@ -5,7 +5,6 @@ #pragma once #include <memory> -#include <optional> #include <string> #include <vector> diff --git a/src/video_core/renderer_vulkan/vk_blit_screen.h b/src/video_core/renderer_vulkan/vk_blit_screen.h index 5eb544aea..243640fab 100644 --- a/src/video_core/renderer_vulkan/vk_blit_screen.h +++ b/src/video_core/renderer_vulkan/vk_blit_screen.h @@ -4,7 +4,6 @@ #pragma once -#include <array> #include <memory> #include <tuple> diff --git a/src/video_core/renderer_vulkan/vk_buffer_cache.cpp b/src/video_core/renderer_vulkan/vk_buffer_cache.cpp index 81e1de2be..5b494da8c 100644 --- a/src/video_core/renderer_vulkan/vk_buffer_cache.cpp +++ b/src/video_core/renderer_vulkan/vk_buffer_cache.cpp @@ -5,11 +5,7 @@ #include <algorithm> #include <cstring> #include <memory> -#include <optional> -#include <tuple> -#include "common/assert.h" -#include "common/bit_util.h" #include "core/core.h" #include "video_core/renderer_vulkan/vk_buffer_cache.h" #include "video_core/renderer_vulkan/vk_device.h" diff --git a/src/video_core/renderer_vulkan/vk_buffer_cache.h b/src/video_core/renderer_vulkan/vk_buffer_cache.h index 3cd2e2774..a54583e7d 100644 --- a/src/video_core/renderer_vulkan/vk_buffer_cache.h +++ b/src/video_core/renderer_vulkan/vk_buffer_cache.h @@ -5,14 +5,11 @@ #pragma once #include <memory> -#include <unordered_map> -#include <vector> #include "common/common_types.h" #include "video_core/buffer_cache/buffer_cache.h" #include "video_core/rasterizer_cache.h" #include "video_core/renderer_vulkan/vk_memory_manager.h" -#include "video_core/renderer_vulkan/vk_resource_manager.h" #include "video_core/renderer_vulkan/vk_staging_buffer_pool.h" #include "video_core/renderer_vulkan/vk_stream_buffer.h" #include "video_core/renderer_vulkan/wrapper.h" @@ -55,8 +52,6 @@ public: protected: VkBuffer ToHandle(const Buffer& buffer) override; - void WriteBarrier() override {} - Buffer CreateBlock(VAddr cpu_addr, std::size_t size) override; void UploadBlockData(const Buffer& buffer, std::size_t offset, std::size_t size, diff --git a/src/video_core/renderer_vulkan/vk_compute_pass.cpp b/src/video_core/renderer_vulkan/vk_compute_pass.cpp index 7b0268033..da71e710c 100644 --- a/src/video_core/renderer_vulkan/vk_compute_pass.cpp +++ b/src/video_core/renderer_vulkan/vk_compute_pass.cpp @@ -6,7 +6,7 @@ #include <memory> #include <optional> #include <utility> -#include <vector> + #include "common/alignment.h" #include "common/assert.h" #include "common/common_types.h" diff --git a/src/video_core/renderer_vulkan/vk_compute_pass.h b/src/video_core/renderer_vulkan/vk_compute_pass.h index 26bf834de..230b526bc 100644 --- a/src/video_core/renderer_vulkan/vk_compute_pass.h +++ b/src/video_core/renderer_vulkan/vk_compute_pass.h @@ -6,7 +6,7 @@ #include <optional> #include <utility> -#include <vector> + #include "common/common_types.h" #include "video_core/engines/maxwell_3d.h" #include "video_core/renderer_vulkan/vk_descriptor_pool.h" diff --git a/src/video_core/renderer_vulkan/vk_compute_pipeline.cpp b/src/video_core/renderer_vulkan/vk_compute_pipeline.cpp index 52566bb79..8e1b46277 100644 --- a/src/video_core/renderer_vulkan/vk_compute_pipeline.cpp +++ b/src/video_core/renderer_vulkan/vk_compute_pipeline.cpp @@ -2,14 +2,12 @@ // Licensed under GPLv2 or any later version // Refer to the license.txt file included. -#include <memory> #include <vector> #include "video_core/renderer_vulkan/vk_compute_pipeline.h" #include "video_core/renderer_vulkan/vk_descriptor_pool.h" #include "video_core/renderer_vulkan/vk_device.h" #include "video_core/renderer_vulkan/vk_pipeline_cache.h" -#include "video_core/renderer_vulkan/vk_resource_manager.h" #include "video_core/renderer_vulkan/vk_scheduler.h" #include "video_core/renderer_vulkan/vk_shader_decompiler.h" #include "video_core/renderer_vulkan/vk_update_descriptor.h" diff --git a/src/video_core/renderer_vulkan/vk_compute_pipeline.h b/src/video_core/renderer_vulkan/vk_compute_pipeline.h index 33b9af29e..6e2f22a4a 100644 --- a/src/video_core/renderer_vulkan/vk_compute_pipeline.h +++ b/src/video_core/renderer_vulkan/vk_compute_pipeline.h @@ -4,8 +4,6 @@ #pragma once -#include <memory> - #include "common/common_types.h" #include "video_core/renderer_vulkan/vk_descriptor_pool.h" #include "video_core/renderer_vulkan/vk_shader_decompiler.h" diff --git a/src/video_core/renderer_vulkan/vk_descriptor_pool.cpp b/src/video_core/renderer_vulkan/vk_descriptor_pool.cpp index e9d528aa6..890fd52cf 100644 --- a/src/video_core/renderer_vulkan/vk_descriptor_pool.cpp +++ b/src/video_core/renderer_vulkan/vk_descriptor_pool.cpp @@ -2,7 +2,6 @@ // Licensed under GPLv2 or any later version // Refer to the license.txt file included. -#include <memory> #include <vector> #include "common/common_types.h" diff --git a/src/video_core/renderer_vulkan/vk_descriptor_pool.h b/src/video_core/renderer_vulkan/vk_descriptor_pool.h index ab40c70f0..9efa66bef 100644 --- a/src/video_core/renderer_vulkan/vk_descriptor_pool.h +++ b/src/video_core/renderer_vulkan/vk_descriptor_pool.h @@ -4,10 +4,8 @@ #pragma once -#include <memory> #include <vector> -#include "common/common_types.h" #include "video_core/renderer_vulkan/vk_resource_manager.h" #include "video_core/renderer_vulkan/wrapper.h" diff --git a/src/video_core/renderer_vulkan/vk_device.cpp b/src/video_core/renderer_vulkan/vk_device.cpp index e90c76492..0e4bbca97 100644 --- a/src/video_core/renderer_vulkan/vk_device.cpp +++ b/src/video_core/renderer_vulkan/vk_device.cpp @@ -4,7 +4,6 @@ #include <bitset> #include <chrono> -#include <cstdlib> #include <optional> #include <string_view> #include <thread> diff --git a/src/video_core/renderer_vulkan/vk_graphics_pipeline.cpp b/src/video_core/renderer_vulkan/vk_graphics_pipeline.cpp index 852a17a70..1ac981974 100644 --- a/src/video_core/renderer_vulkan/vk_graphics_pipeline.cpp +++ b/src/video_core/renderer_vulkan/vk_graphics_pipeline.cpp @@ -6,7 +6,6 @@ #include <cstring> #include <vector> -#include "common/assert.h" #include "common/common_types.h" #include "common/microprofile.h" #include "video_core/renderer_vulkan/fixed_pipeline_state.h" diff --git a/src/video_core/renderer_vulkan/vk_graphics_pipeline.h b/src/video_core/renderer_vulkan/vk_graphics_pipeline.h index 7aba70960..a1d699a6c 100644 --- a/src/video_core/renderer_vulkan/vk_graphics_pipeline.h +++ b/src/video_core/renderer_vulkan/vk_graphics_pipeline.h @@ -5,16 +5,13 @@ #pragma once #include <array> -#include <memory> #include <optional> -#include <unordered_map> #include <vector> #include "video_core/engines/maxwell_3d.h" #include "video_core/renderer_vulkan/fixed_pipeline_state.h" #include "video_core/renderer_vulkan/vk_descriptor_pool.h" #include "video_core/renderer_vulkan/vk_renderpass_cache.h" -#include "video_core/renderer_vulkan/vk_resource_manager.h" #include "video_core/renderer_vulkan/vk_shader_decompiler.h" #include "video_core/renderer_vulkan/wrapper.h" diff --git a/src/video_core/renderer_vulkan/vk_pipeline_cache.cpp b/src/video_core/renderer_vulkan/vk_pipeline_cache.cpp index 9b703a2f0..fe45ed269 100644 --- a/src/video_core/renderer_vulkan/vk_pipeline_cache.cpp +++ b/src/video_core/renderer_vulkan/vk_pipeline_cache.cpp @@ -22,7 +22,6 @@ #include "video_core/renderer_vulkan/vk_pipeline_cache.h" #include "video_core/renderer_vulkan/vk_rasterizer.h" #include "video_core/renderer_vulkan/vk_renderpass_cache.h" -#include "video_core/renderer_vulkan/vk_resource_manager.h" #include "video_core/renderer_vulkan/vk_scheduler.h" #include "video_core/renderer_vulkan/vk_update_descriptor.h" #include "video_core/renderer_vulkan/wrapper.h" @@ -85,7 +84,7 @@ void AddBindings(std::vector<VkDescriptorSetLayoutBinding>& bindings, u32& bindi u32 count = 1; if constexpr (descriptor_type == VK_DESCRIPTOR_TYPE_COMBINED_IMAGE_SAMPLER) { // Combined image samplers can be arrayed. - count = container[i].Size(); + count = container[i].size; } VkDescriptorSetLayoutBinding& entry = bindings.emplace_back(); entry.binding = binding++; @@ -362,7 +361,7 @@ void AddEntry(std::vector<VkDescriptorUpdateTemplateEntry>& template_entries, u3 if constexpr (descriptor_type == COMBINED_IMAGE_SAMPLER) { for (u32 i = 0; i < count; ++i) { - const u32 num_samplers = container[i].Size(); + const u32 num_samplers = container[i].size; VkDescriptorUpdateTemplateEntry& entry = template_entries.emplace_back(); entry.dstBinding = binding; entry.dstArrayElement = 0; diff --git a/src/video_core/renderer_vulkan/vk_pipeline_cache.h b/src/video_core/renderer_vulkan/vk_pipeline_cache.h index ebddafb73..0b5796fef 100644 --- a/src/video_core/renderer_vulkan/vk_pipeline_cache.h +++ b/src/video_core/renderer_vulkan/vk_pipeline_cache.h @@ -21,13 +21,11 @@ #include "video_core/renderer_vulkan/fixed_pipeline_state.h" #include "video_core/renderer_vulkan/vk_graphics_pipeline.h" #include "video_core/renderer_vulkan/vk_renderpass_cache.h" -#include "video_core/renderer_vulkan/vk_resource_manager.h" #include "video_core/renderer_vulkan/vk_shader_decompiler.h" #include "video_core/renderer_vulkan/wrapper.h" #include "video_core/shader/memory_util.h" #include "video_core/shader/registry.h" #include "video_core/shader/shader_ir.h" -#include "video_core/surface.h" namespace Core { class System; diff --git a/src/video_core/renderer_vulkan/vk_query_cache.cpp b/src/video_core/renderer_vulkan/vk_query_cache.cpp index 813f7c162..bc91c48cc 100644 --- a/src/video_core/renderer_vulkan/vk_query_cache.cpp +++ b/src/video_core/renderer_vulkan/vk_query_cache.cpp @@ -4,7 +4,6 @@ #include <algorithm> #include <cstddef> -#include <cstdint> #include <utility> #include <vector> diff --git a/src/video_core/renderer_vulkan/vk_query_cache.h b/src/video_core/renderer_vulkan/vk_query_cache.h index b63784f4b..40119e6d3 100644 --- a/src/video_core/renderer_vulkan/vk_query_cache.h +++ b/src/video_core/renderer_vulkan/vk_query_cache.h @@ -5,7 +5,6 @@ #pragma once #include <cstddef> -#include <cstdint> #include <memory> #include <utility> #include <vector> diff --git a/src/video_core/renderer_vulkan/vk_rasterizer.cpp b/src/video_core/renderer_vulkan/vk_rasterizer.cpp index 776053de5..8b009fc22 100644 --- a/src/video_core/renderer_vulkan/vk_rasterizer.cpp +++ b/src/video_core/renderer_vulkan/vk_rasterizer.cpp @@ -9,14 +9,12 @@ #include <vector> #include <boost/container/static_vector.hpp> -#include <boost/functional/hash.hpp> #include "common/alignment.h" #include "common/assert.h" #include "common/logging/log.h" #include "common/microprofile.h" #include "core/core.h" -#include "core/memory.h" #include "core/settings.h" #include "video_core/engines/kepler_compute.h" #include "video_core/engines/maxwell_3d.h" @@ -119,14 +117,13 @@ template <typename Engine, typename Entry> Tegra::Texture::FullTextureInfo GetTextureInfo(const Engine& engine, const Entry& entry, std::size_t stage, std::size_t index = 0) { const auto stage_type = static_cast<Tegra::Engines::ShaderType>(stage); - if (entry.IsBindless()) { - const Tegra::Texture::TextureHandle tex_handle = - engine.AccessConstBuffer32(stage_type, entry.GetBuffer(), entry.GetOffset()); + if (entry.is_bindless) { + const auto tex_handle = engine.AccessConstBuffer32(stage_type, entry.buffer, entry.offset); return engine.GetTextureInfo(tex_handle); } const auto& gpu_profile = engine.AccessGuestDriverProfile(); const u32 entry_offset = static_cast<u32>(index * gpu_profile.GetTextureHandlerSize()); - const u32 offset = entry.GetOffset() + entry_offset; + const u32 offset = entry.offset + entry_offset; if constexpr (std::is_same_v<Engine, Tegra::Engines::Maxwell3D>) { return engine.GetStageTexture(stage_type, offset); } else { @@ -302,7 +299,7 @@ RasterizerVulkan::RasterizerVulkan(Core::System& system, Core::Frontend::EmuWind buffer_cache(*this, system, device, memory_manager, scheduler, staging_pool), sampler_cache(device), fence_manager(system, *this, device, scheduler, texture_cache, buffer_cache, query_cache), - query_cache(system, *this, device, scheduler) { + query_cache(system, *this, device, scheduler), wfi_event{device.GetLogical().CreateEvent()} { scheduler.SetQueryCache(query_cache); } @@ -576,6 +573,26 @@ void RasterizerVulkan::FlushAndInvalidateRegion(VAddr addr, u64 size) { InvalidateRegion(addr, size); } +void RasterizerVulkan::WaitForIdle() { + // Everything but wait pixel operations. This intentionally includes FRAGMENT_SHADER_BIT because + // fragment shaders can still write storage buffers. + VkPipelineStageFlags flags = + VK_PIPELINE_STAGE_DRAW_INDIRECT_BIT | VK_PIPELINE_STAGE_VERTEX_INPUT_BIT | + VK_PIPELINE_STAGE_VERTEX_SHADER_BIT | VK_PIPELINE_STAGE_TESSELLATION_CONTROL_SHADER_BIT | + VK_PIPELINE_STAGE_TESSELLATION_EVALUATION_SHADER_BIT | + VK_PIPELINE_STAGE_GEOMETRY_SHADER_BIT | VK_PIPELINE_STAGE_FRAGMENT_SHADER_BIT | + VK_PIPELINE_STAGE_COMPUTE_SHADER_BIT | VK_PIPELINE_STAGE_TRANSFER_BIT; + if (device.IsExtTransformFeedbackSupported()) { + flags |= VK_PIPELINE_STAGE_TRANSFORM_FEEDBACK_BIT_EXT; + } + + scheduler.RequestOutsideRenderPassOperationContext(); + scheduler.Record([event = *wfi_event, flags](vk::CommandBuffer cmdbuf) { + cmdbuf.SetEvent(event, flags); + cmdbuf.WaitEvents(event, flags, VK_PIPELINE_STAGE_TOP_OF_PIPE_BIT, {}, {}, {}); + }); +} + void RasterizerVulkan::FlushCommands() { if (draw_counter > 0) { draw_counter = 0; @@ -896,6 +913,9 @@ void RasterizerVulkan::SetupVertexArrays(FixedPipelineState::VertexInput& vertex void RasterizerVulkan::SetupIndexBuffer(BufferBindings& buffer_bindings, DrawParameters& params, bool is_indexed) { + if (params.num_vertices == 0) { + return; + } const auto& regs = system.GPU().Maxwell3D().regs; switch (regs.draw.topology) { case Maxwell::PrimitiveTopology::Quads: { @@ -971,7 +991,7 @@ void RasterizerVulkan::SetupGraphicsTextures(const ShaderEntries& entries, std:: MICROPROFILE_SCOPE(Vulkan_Textures); const auto& gpu = system.GPU().Maxwell3D(); for (const auto& entry : entries.samplers) { - for (std::size_t i = 0; i < entry.Size(); ++i) { + for (std::size_t i = 0; i < entry.size; ++i) { const auto texture = GetTextureInfo(gpu, entry, stage, i); SetupTexture(texture, entry); } @@ -1023,7 +1043,7 @@ void RasterizerVulkan::SetupComputeTextures(const ShaderEntries& entries) { MICROPROFILE_SCOPE(Vulkan_Textures); const auto& gpu = system.GPU().KeplerCompute(); for (const auto& entry : entries.samplers) { - for (std::size_t i = 0; i < entry.Size(); ++i) { + for (std::size_t i = 0; i < entry.size; ++i) { const auto texture = GetTextureInfo(gpu, entry, ComputeShaderIndex, i); SetupTexture(texture, entry); } @@ -1105,7 +1125,7 @@ void RasterizerVulkan::SetupTexture(const Tegra::Texture::FullTextureInfo& textu void RasterizerVulkan::SetupImage(const Tegra::Texture::TICEntry& tic, const ImageEntry& entry) { auto view = texture_cache.GetImageSurface(tic, entry); - if (entry.IsWritten()) { + if (entry.is_written) { view->MarkAsModified(texture_cache.Tick()); } diff --git a/src/video_core/renderer_vulkan/vk_rasterizer.h b/src/video_core/renderer_vulkan/vk_rasterizer.h index d41a7929e..0ed0e48c6 100644 --- a/src/video_core/renderer_vulkan/vk_rasterizer.h +++ b/src/video_core/renderer_vulkan/vk_rasterizer.h @@ -14,7 +14,6 @@ #include <boost/functional/hash.hpp> #include "common/common_types.h" -#include "video_core/memory_manager.h" #include "video_core/rasterizer_accelerated.h" #include "video_core/rasterizer_interface.h" #include "video_core/renderer_vulkan/fixed_pipeline_state.h" @@ -127,6 +126,7 @@ public: void SignalSyncPoint(u32 value) override; void ReleaseFences() override; void FlushAndInvalidateRegion(VAddr addr, u64 size) override; + void WaitForIdle() override; void FlushCommands() override; void TickFrame() override; bool AccelerateSurfaceCopy(const Tegra::Engines::Fermi2D::Regs::Surface& src, @@ -276,6 +276,7 @@ private: vk::Buffer default_buffer; VKMemoryCommit default_buffer_commit; + vk::Event wfi_event; std::array<View, Maxwell::NumRenderTargets> color_attachments; View zeta_attachment; diff --git a/src/video_core/renderer_vulkan/vk_sampler_cache.cpp b/src/video_core/renderer_vulkan/vk_sampler_cache.cpp index 07bbcf520..2687d8d95 100644 --- a/src/video_core/renderer_vulkan/vk_sampler_cache.cpp +++ b/src/video_core/renderer_vulkan/vk_sampler_cache.cpp @@ -2,11 +2,8 @@ // Licensed under GPLv2 or any later version // Refer to the license.txt file included. -#include <cstring> -#include <optional> #include <unordered_map> -#include "common/assert.h" #include "video_core/renderer_vulkan/maxwell_to_vk.h" #include "video_core/renderer_vulkan/vk_sampler_cache.h" #include "video_core/renderer_vulkan/wrapper.h" diff --git a/src/video_core/renderer_vulkan/vk_scheduler.cpp b/src/video_core/renderer_vulkan/vk_scheduler.cpp index ae7ba3eb5..82ec9180e 100644 --- a/src/video_core/renderer_vulkan/vk_scheduler.cpp +++ b/src/video_core/renderer_vulkan/vk_scheduler.cpp @@ -8,7 +8,6 @@ #include <thread> #include <utility> -#include "common/assert.h" #include "common/microprofile.h" #include "video_core/renderer_vulkan/vk_device.h" #include "video_core/renderer_vulkan/vk_query_cache.h" diff --git a/src/video_core/renderer_vulkan/vk_scheduler.h b/src/video_core/renderer_vulkan/vk_scheduler.h index 82a8adc69..970a65566 100644 --- a/src/video_core/renderer_vulkan/vk_scheduler.h +++ b/src/video_core/renderer_vulkan/vk_scheduler.h @@ -7,7 +7,6 @@ #include <atomic> #include <condition_variable> #include <memory> -#include <optional> #include <stack> #include <thread> #include <utility> diff --git a/src/video_core/renderer_vulkan/vk_shader_decompiler.cpp b/src/video_core/renderer_vulkan/vk_shader_decompiler.cpp index 20b6ca0ad..18678968c 100644 --- a/src/video_core/renderer_vulkan/vk_shader_decompiler.cpp +++ b/src/video_core/renderer_vulkan/vk_shader_decompiler.cpp @@ -103,8 +103,8 @@ struct GenericVaryingDescription { }; spv::Dim GetSamplerDim(const Sampler& sampler) { - ASSERT(!sampler.IsBuffer()); - switch (sampler.GetType()) { + ASSERT(!sampler.is_buffer); + switch (sampler.type) { case Tegra::Shader::TextureType::Texture1D: return spv::Dim::Dim1D; case Tegra::Shader::TextureType::Texture2D: @@ -114,13 +114,13 @@ spv::Dim GetSamplerDim(const Sampler& sampler) { case Tegra::Shader::TextureType::TextureCube: return spv::Dim::Cube; default: - UNIMPLEMENTED_MSG("Unimplemented sampler type={}", static_cast<u32>(sampler.GetType())); + UNIMPLEMENTED_MSG("Unimplemented sampler type={}", static_cast<int>(sampler.type)); return spv::Dim::Dim2D; } } std::pair<spv::Dim, bool> GetImageDim(const Image& image) { - switch (image.GetType()) { + switch (image.type) { case Tegra::Shader::ImageType::Texture1D: return {spv::Dim::Dim1D, false}; case Tegra::Shader::ImageType::TextureBuffer: @@ -134,7 +134,7 @@ std::pair<spv::Dim, bool> GetImageDim(const Image& image) { case Tegra::Shader::ImageType::Texture3D: return {spv::Dim::Dim3D, false}; default: - UNIMPLEMENTED_MSG("Unimplemented image type={}", static_cast<u32>(image.GetType())); + UNIMPLEMENTED_MSG("Unimplemented image type={}", static_cast<int>(image.type)); return {spv::Dim::Dim2D, false}; } } @@ -879,11 +879,11 @@ private: u32 DeclareTexelBuffers(u32 binding) { for (const auto& sampler : ir.GetSamplers()) { - if (!sampler.IsBuffer()) { + if (!sampler.is_buffer) { continue; } - ASSERT(!sampler.IsArray()); - ASSERT(!sampler.IsShadow()); + ASSERT(!sampler.is_array); + ASSERT(!sampler.is_shadow); constexpr auto dim = spv::Dim::Buffer; constexpr int depth = 0; @@ -894,23 +894,23 @@ private: const Id image_type = TypeImage(t_float, dim, depth, arrayed, ms, sampled, format); const Id pointer_type = TypePointer(spv::StorageClass::UniformConstant, image_type); const Id id = OpVariable(pointer_type, spv::StorageClass::UniformConstant); - AddGlobalVariable(Name(id, fmt::format("sampler_{}", sampler.GetIndex()))); + AddGlobalVariable(Name(id, fmt::format("sampler_{}", sampler.index))); Decorate(id, spv::Decoration::Binding, binding++); Decorate(id, spv::Decoration::DescriptorSet, DESCRIPTOR_SET); - texel_buffers.emplace(sampler.GetIndex(), TexelBuffer{image_type, id}); + texel_buffers.emplace(sampler.index, TexelBuffer{image_type, id}); } return binding; } u32 DeclareSamplers(u32 binding) { for (const auto& sampler : ir.GetSamplers()) { - if (sampler.IsBuffer()) { + if (sampler.is_buffer) { continue; } const auto dim = GetSamplerDim(sampler); - const int depth = sampler.IsShadow() ? 1 : 0; - const int arrayed = sampler.IsArray() ? 1 : 0; + const int depth = sampler.is_shadow ? 1 : 0; + const int arrayed = sampler.is_array ? 1 : 0; constexpr bool ms = false; constexpr int sampled = 1; constexpr auto format = spv::ImageFormat::Unknown; @@ -918,17 +918,17 @@ private: const Id sampler_type = TypeSampledImage(image_type); const Id sampler_pointer_type = TypePointer(spv::StorageClass::UniformConstant, sampler_type); - const Id type = sampler.IsIndexed() - ? TypeArray(sampler_type, Constant(t_uint, sampler.Size())) + const Id type = sampler.is_indexed + ? TypeArray(sampler_type, Constant(t_uint, sampler.size)) : sampler_type; const Id pointer_type = TypePointer(spv::StorageClass::UniformConstant, type); const Id id = OpVariable(pointer_type, spv::StorageClass::UniformConstant); - AddGlobalVariable(Name(id, fmt::format("sampler_{}", sampler.GetIndex()))); + AddGlobalVariable(Name(id, fmt::format("sampler_{}", sampler.index))); Decorate(id, spv::Decoration::Binding, binding++); Decorate(id, spv::Decoration::DescriptorSet, DESCRIPTOR_SET); - sampled_images.emplace(sampler.GetIndex(), SampledImage{image_type, sampler_type, - sampler_pointer_type, id}); + sampled_images.emplace( + sampler.index, SampledImage{image_type, sampler_type, sampler_pointer_type, id}); } return binding; } @@ -943,17 +943,17 @@ private: const Id image_type = TypeImage(t_uint, dim, depth, arrayed, ms, sampled, format, {}); const Id pointer_type = TypePointer(spv::StorageClass::UniformConstant, image_type); const Id id = OpVariable(pointer_type, spv::StorageClass::UniformConstant); - AddGlobalVariable(Name(id, fmt::format("image_{}", image.GetIndex()))); + AddGlobalVariable(Name(id, fmt::format("image_{}", image.index))); Decorate(id, spv::Decoration::Binding, binding++); Decorate(id, spv::Decoration::DescriptorSet, DESCRIPTOR_SET); - if (image.IsRead() && !image.IsWritten()) { + if (image.is_read && !image.is_written) { Decorate(id, spv::Decoration::NonWritable); - } else if (image.IsWritten() && !image.IsRead()) { + } else if (image.is_written && !image.is_read) { Decorate(id, spv::Decoration::NonReadable); } - images.emplace(static_cast<u32>(image.GetIndex()), StorageImage{image_type, id}); + images.emplace(image.index, StorageImage{image_type, id}); } return binding; } @@ -1620,11 +1620,11 @@ private: Id GetTextureSampler(Operation operation) { const auto& meta = std::get<MetaTexture>(operation.GetMeta()); - ASSERT(!meta.sampler.IsBuffer()); + ASSERT(!meta.sampler.is_buffer); - const auto& entry = sampled_images.at(meta.sampler.GetIndex()); + const auto& entry = sampled_images.at(meta.sampler.index); Id sampler = entry.variable; - if (meta.sampler.IsIndexed()) { + if (meta.sampler.is_indexed) { const Id index = AsInt(Visit(meta.index)); sampler = OpAccessChain(entry.sampler_pointer_type, sampler, index); } @@ -1633,8 +1633,8 @@ private: Id GetTextureImage(Operation operation) { const auto& meta = std::get<MetaTexture>(operation.GetMeta()); - const u32 index = meta.sampler.GetIndex(); - if (meta.sampler.IsBuffer()) { + const u32 index = meta.sampler.index; + if (meta.sampler.is_buffer) { const auto& entry = texel_buffers.at(index); return OpLoad(entry.image_type, entry.image); } else { @@ -1645,7 +1645,7 @@ private: Id GetImage(Operation operation) { const auto& meta = std::get<MetaImage>(operation.GetMeta()); - const auto entry = images.at(meta.image.GetIndex()); + const auto entry = images.at(meta.image.index); return OpLoad(entry.image_type, entry.image); } @@ -1661,7 +1661,7 @@ private: } if (const auto meta = std::get_if<MetaTexture>(&operation.GetMeta())) { // Add array coordinate for textures - if (meta->sampler.IsArray()) { + if (meta->sampler.is_array) { Id array = AsInt(Visit(meta->array)); if (type == Type::Float) { array = OpConvertSToF(t_float, array); @@ -1767,7 +1767,7 @@ private: operands.push_back(GetOffsetCoordinates(operation)); } - if (meta.sampler.IsShadow()) { + if (meta.sampler.is_shadow) { const Id dref = AsFloat(Visit(meta.depth_compare)); return {OpImageSampleDrefExplicitLod(t_float, sampler, coords, dref, mask, operands), Type::Float}; @@ -1782,7 +1782,7 @@ private: const Id coords = GetCoordinates(operation, Type::Float); Id texture{}; - if (meta.sampler.IsShadow()) { + if (meta.sampler.is_shadow) { texture = OpImageDrefGather(t_float4, GetTextureSampler(operation), coords, AsFloat(Visit(meta.depth_compare))); } else { @@ -1809,8 +1809,8 @@ private: } const Id lod = AsUint(Visit(operation[0])); - const std::size_t coords_count = [&]() { - switch (const auto type = meta.sampler.GetType(); type) { + const std::size_t coords_count = [&meta] { + switch (const auto type = meta.sampler.type) { case Tegra::Shader::TextureType::Texture1D: return 1; case Tegra::Shader::TextureType::Texture2D: @@ -1819,7 +1819,7 @@ private: case Tegra::Shader::TextureType::Texture3D: return 3; default: - UNREACHABLE_MSG("Invalid texture type={}", static_cast<u32>(type)); + UNREACHABLE_MSG("Invalid texture type={}", static_cast<int>(type)); return 2; } }(); @@ -1862,7 +1862,7 @@ private: const Id image = GetTextureImage(operation); const Id coords = GetCoordinates(operation, Type::Int); Id fetch; - if (meta.lod && !meta.sampler.IsBuffer()) { + if (meta.lod && !meta.sampler.is_buffer) { fetch = OpImageFetch(t_float4, image, coords, spv::ImageOperandsMask::Lod, AsInt(Visit(meta.lod))); } else { @@ -2980,7 +2980,7 @@ ShaderEntries GenerateShaderEntries(const VideoCommon::Shader::ShaderIR& ir) { entries.global_buffers.emplace_back(base.cbuf_index, base.cbuf_offset, usage.is_written); } for (const auto& sampler : ir.GetSamplers()) { - if (sampler.IsBuffer()) { + if (sampler.is_buffer) { entries.texel_buffers.emplace_back(sampler); } else { entries.samplers.emplace_back(sampler); diff --git a/src/video_core/renderer_vulkan/vk_shader_decompiler.h b/src/video_core/renderer_vulkan/vk_shader_decompiler.h index ffea4709e..f4c05ac3c 100644 --- a/src/video_core/renderer_vulkan/vk_shader_decompiler.h +++ b/src/video_core/renderer_vulkan/vk_shader_decompiler.h @@ -5,11 +5,7 @@ #pragma once #include <array> -#include <bitset> -#include <memory> #include <set> -#include <type_traits> -#include <utility> #include <vector> #include "common/common_types.h" diff --git a/src/video_core/renderer_vulkan/vk_shader_util.cpp b/src/video_core/renderer_vulkan/vk_shader_util.cpp index 784839327..112df9c71 100644 --- a/src/video_core/renderer_vulkan/vk_shader_util.cpp +++ b/src/video_core/renderer_vulkan/vk_shader_util.cpp @@ -4,8 +4,7 @@ #include <cstring> #include <memory> -#include <vector> -#include "common/alignment.h" + #include "common/assert.h" #include "common/common_types.h" #include "video_core/renderer_vulkan/vk_device.h" diff --git a/src/video_core/renderer_vulkan/vk_shader_util.h b/src/video_core/renderer_vulkan/vk_shader_util.h index be38d6697..d1d3f3cae 100644 --- a/src/video_core/renderer_vulkan/vk_shader_util.h +++ b/src/video_core/renderer_vulkan/vk_shader_util.h @@ -4,7 +4,6 @@ #pragma once -#include <vector> #include "common/common_types.h" #include "video_core/renderer_vulkan/wrapper.h" diff --git a/src/video_core/renderer_vulkan/vk_staging_buffer_pool.h b/src/video_core/renderer_vulkan/vk_staging_buffer_pool.h index faf6418fd..3c4901437 100644 --- a/src/video_core/renderer_vulkan/vk_staging_buffer_pool.h +++ b/src/video_core/renderer_vulkan/vk_staging_buffer_pool.h @@ -5,8 +5,6 @@ #pragma once #include <climits> -#include <unordered_map> -#include <utility> #include <vector> #include "common/common_types.h" diff --git a/src/video_core/renderer_vulkan/vk_texture_cache.cpp b/src/video_core/renderer_vulkan/vk_texture_cache.cpp index de4c23120..55f43e61b 100644 --- a/src/video_core/renderer_vulkan/vk_texture_cache.cpp +++ b/src/video_core/renderer_vulkan/vk_texture_cache.cpp @@ -10,11 +10,9 @@ #include <variant> #include <vector> -#include "common/alignment.h" #include "common/assert.h" #include "common/common_types.h" #include "core/core.h" -#include "core/memory.h" #include "video_core/engines/maxwell_3d.h" #include "video_core/morton.h" #include "video_core/renderer_vulkan/maxwell_to_vk.h" @@ -26,7 +24,6 @@ #include "video_core/renderer_vulkan/vk_texture_cache.h" #include "video_core/renderer_vulkan/wrapper.h" #include "video_core/surface.h" -#include "video_core/textures/convert.h" namespace Vulkan { diff --git a/src/video_core/renderer_vulkan/vk_texture_cache.h b/src/video_core/renderer_vulkan/vk_texture_cache.h index 115595f28..f211ccb1e 100644 --- a/src/video_core/renderer_vulkan/vk_texture_cache.h +++ b/src/video_core/renderer_vulkan/vk_texture_cache.h @@ -7,19 +7,13 @@ #include <memory> #include <unordered_map> -#include "common/assert.h" #include "common/common_types.h" -#include "common/logging/log.h" -#include "common/math_util.h" -#include "video_core/gpu.h" -#include "video_core/rasterizer_cache.h" #include "video_core/renderer_vulkan/vk_image.h" #include "video_core/renderer_vulkan/vk_memory_manager.h" #include "video_core/renderer_vulkan/vk_scheduler.h" #include "video_core/renderer_vulkan/wrapper.h" #include "video_core/texture_cache/surface_base.h" #include "video_core/texture_cache/texture_cache.h" -#include "video_core/textures/decoders.h" namespace Core { class System; diff --git a/src/video_core/renderer_vulkan/vk_update_descriptor.h b/src/video_core/renderer_vulkan/vk_update_descriptor.h index 6ba2c9997..cc7e3dff4 100644 --- a/src/video_core/renderer_vulkan/vk_update_descriptor.h +++ b/src/video_core/renderer_vulkan/vk_update_descriptor.h @@ -4,7 +4,6 @@ #pragma once -#include <type_traits> #include <variant> #include <boost/container/static_vector.hpp> diff --git a/src/video_core/renderer_vulkan/wrapper.cpp b/src/video_core/renderer_vulkan/wrapper.cpp index 7f5bc1404..2ce9b0626 100644 --- a/src/video_core/renderer_vulkan/wrapper.cpp +++ b/src/video_core/renderer_vulkan/wrapper.cpp @@ -87,6 +87,7 @@ void Load(VkDevice device, DeviceDispatch& dld) noexcept { X(vkCmdSetStencilReference); X(vkCmdSetStencilWriteMask); X(vkCmdSetViewport); + X(vkCmdWaitEvents); X(vkCreateBuffer); X(vkCreateBufferView); X(vkCreateCommandPool); diff --git a/src/video_core/renderer_vulkan/wrapper.h b/src/video_core/renderer_vulkan/wrapper.h index bda16a2cb..98937a77a 100644 --- a/src/video_core/renderer_vulkan/wrapper.h +++ b/src/video_core/renderer_vulkan/wrapper.h @@ -205,6 +205,7 @@ struct DeviceDispatch : public InstanceDispatch { PFN_vkCmdSetStencilReference vkCmdSetStencilReference; PFN_vkCmdSetStencilWriteMask vkCmdSetStencilWriteMask; PFN_vkCmdSetViewport vkCmdSetViewport; + PFN_vkCmdWaitEvents vkCmdWaitEvents; PFN_vkCreateBuffer vkCreateBuffer; PFN_vkCreateBufferView vkCreateBufferView; PFN_vkCreateCommandPool vkCreateCommandPool; @@ -958,6 +959,15 @@ public: dld->vkCmdSetEvent(handle, event, stage_flags); } + void WaitEvents(Span<VkEvent> events, VkPipelineStageFlags src_stage_mask, + VkPipelineStageFlags dst_stage_mask, Span<VkMemoryBarrier> memory_barriers, + Span<VkBufferMemoryBarrier> buffer_barriers, + Span<VkImageMemoryBarrier> image_barriers) const noexcept { + dld->vkCmdWaitEvents(handle, events.size(), events.data(), src_stage_mask, dst_stage_mask, + memory_barriers.size(), memory_barriers.data(), buffer_barriers.size(), + buffer_barriers.data(), image_barriers.size(), image_barriers.data()); + } + void BindTransformFeedbackBuffersEXT(u32 first, u32 count, const VkBuffer* buffers, const VkDeviceSize* offsets, const VkDeviceSize* sizes) const noexcept { diff --git a/src/video_core/shader/decode.cpp b/src/video_core/shader/decode.cpp index 1167ff4ec..a75a5cc63 100644 --- a/src/video_core/shader/decode.cpp +++ b/src/video_core/shader/decode.cpp @@ -32,11 +32,11 @@ void DeduceTextureHandlerSize(VideoCore::GuestDriverProfile& gpu_driver, u32 count{}; std::vector<u32> bound_offsets; for (const auto& sampler : used_samplers) { - if (sampler.IsBindless()) { + if (sampler.is_bindless) { continue; } ++count; - bound_offsets.emplace_back(sampler.GetOffset()); + bound_offsets.emplace_back(sampler.offset); } if (count > 1) { gpu_driver.DeduceTextureHandlerSize(std::move(bound_offsets)); @@ -46,14 +46,14 @@ void DeduceTextureHandlerSize(VideoCore::GuestDriverProfile& gpu_driver, std::optional<u32> TryDeduceSamplerSize(const Sampler& sampler_to_deduce, VideoCore::GuestDriverProfile& gpu_driver, const std::list<Sampler>& used_samplers) { - const u32 base_offset = sampler_to_deduce.GetOffset(); + const u32 base_offset = sampler_to_deduce.offset; u32 max_offset{std::numeric_limits<u32>::max()}; for (const auto& sampler : used_samplers) { - if (sampler.IsBindless()) { + if (sampler.is_bindless) { continue; } - if (sampler.GetOffset() > base_offset) { - max_offset = std::min(sampler.GetOffset(), max_offset); + if (sampler.offset > base_offset) { + max_offset = std::min(sampler.offset, max_offset); } } if (max_offset == std::numeric_limits<u32>::max()) { @@ -353,14 +353,14 @@ void ShaderIR::PostDecode() { return; } for (auto& sampler : used_samplers) { - if (!sampler.IsIndexed()) { + if (!sampler.is_indexed) { continue; } if (const auto size = TryDeduceSamplerSize(sampler, gpu_driver, used_samplers)) { - sampler.SetSize(*size); + sampler.size = *size; } else { LOG_CRITICAL(HW_GPU, "Failed to deduce size of indexed sampler"); - sampler.SetSize(1); + sampler.size = 1; } } } diff --git a/src/video_core/shader/decode/image.cpp b/src/video_core/shader/decode/image.cpp index 85ee9aa5e..60b6ad72a 100644 --- a/src/video_core/shader/decode/image.cpp +++ b/src/video_core/shader/decode/image.cpp @@ -485,11 +485,10 @@ u32 ShaderIR::DecodeImage(NodeBlock& bb, u32 pc) { Image& ShaderIR::GetImage(Tegra::Shader::Image image, Tegra::Shader::ImageType type) { const auto offset = static_cast<u32>(image.index.Value()); - const auto it = - std::find_if(std::begin(used_images), std::end(used_images), - [offset](const Image& entry) { return entry.GetOffset() == offset; }); + const auto it = std::find_if(std::begin(used_images), std::end(used_images), + [offset](const Image& entry) { return entry.offset == offset; }); if (it != std::end(used_images)) { - ASSERT(!it->IsBindless() && it->GetType() == it->GetType()); + ASSERT(!it->is_bindless && it->type == type); return *it; } @@ -505,13 +504,12 @@ Image& ShaderIR::GetBindlessImage(Tegra::Shader::Register reg, Tegra::Shader::Im const auto buffer = std::get<1>(result); const auto offset = std::get<2>(result); - const auto it = - std::find_if(std::begin(used_images), std::end(used_images), - [buffer = buffer, offset = offset](const Image& entry) { - return entry.GetBuffer() == buffer && entry.GetOffset() == offset; - }); + const auto it = std::find_if(std::begin(used_images), std::end(used_images), + [buffer, offset](const Image& entry) { + return entry.buffer == buffer && entry.offset == offset; + }); if (it != std::end(used_images)) { - ASSERT(it->IsBindless() && it->GetType() == it->GetType()); + ASSERT(it->is_bindless && it->type == type); return *it; } diff --git a/src/video_core/shader/decode/texture.cpp b/src/video_core/shader/decode/texture.cpp index e68f1d305..8f0bb996e 100644 --- a/src/video_core/shader/decode/texture.cpp +++ b/src/video_core/shader/decode/texture.cpp @@ -139,15 +139,15 @@ u32 ShaderIR::DecodeTexture(NodeBlock& bb, u32 pc) { } const Node component = Immediate(static_cast<u32>(instr.tld4s.component)); - const SamplerInfo info{TextureType::Texture2D, false, is_depth_compare, false}; - const Sampler& sampler = *GetSampler(instr.sampler, info); + SamplerInfo info; + info.is_shadow = is_depth_compare; + const std::optional<Sampler> sampler = GetSampler(instr.sampler, info); Node4 values; for (u32 element = 0; element < values.size(); ++element) { - auto coords_copy = coords; - MetaTexture meta{sampler, {}, depth_compare, aoffi, {}, {}, - {}, {}, component, element, {}}; - values[element] = Operation(OperationCode::TextureGather, meta, std::move(coords_copy)); + MetaTexture meta{*sampler, {}, depth_compare, aoffi, {}, {}, + {}, {}, component, element, {}}; + values[element] = Operation(OperationCode::TextureGather, meta, coords); } if (instr.tld4s.fp16_flag) { @@ -165,18 +165,20 @@ u32 ShaderIR::DecodeTexture(NodeBlock& bb, u32 pc) { "AOFFI is not implemented"); const bool is_array = instr.txd.is_array != 0; - u64 base_reg = instr.gpr8.Value(); const auto derivate_reg = instr.gpr20.Value(); const auto texture_type = instr.txd.texture_type.Value(); const auto coord_count = GetCoordCount(texture_type); - Node index_var{}; - const Sampler* sampler = - is_bindless - ? GetBindlessSampler(base_reg, index_var, {{texture_type, is_array, false, false}}) - : GetSampler(instr.sampler, {{texture_type, is_array, false, false}}); + u64 base_reg = instr.gpr8.Value(); + Node index_var; + SamplerInfo info; + info.type = texture_type; + info.is_array = is_array; + const std::optional<Sampler> sampler = is_bindless + ? GetBindlessSampler(base_reg, info, index_var) + : GetSampler(instr.sampler, info); Node4 values; - if (sampler == nullptr) { - std::generate(values.begin(), values.end(), [] { return Immediate(0); }); + if (!sampler) { + std::generate(values.begin(), values.end(), [this] { return Immediate(0); }); WriteTexInstructionFloat(bb, instr, values); break; } @@ -214,14 +216,12 @@ u32 ShaderIR::DecodeTexture(NodeBlock& bb, u32 pc) { is_bindless = true; [[fallthrough]]; case OpCode::Id::TXQ: { - // TODO: The new commits on the texture refactor, change the way samplers work. - // Sadly, not all texture instructions specify the type of texture their sampler - // uses. This must be fixed at a later instance. - Node index_var{}; - const Sampler* sampler = - is_bindless ? GetBindlessSampler(instr.gpr8, index_var) : GetSampler(instr.sampler); - - if (sampler == nullptr) { + Node index_var; + const std::optional<Sampler> sampler = is_bindless + ? GetBindlessSampler(instr.gpr8, {}, index_var) + : GetSampler(instr.sampler, {}); + + if (!sampler) { u32 indexer = 0; for (u32 element = 0; element < 4; ++element) { if (!instr.txq.IsComponentEnabled(element)) { @@ -267,12 +267,17 @@ u32 ShaderIR::DecodeTexture(NodeBlock& bb, u32 pc) { UNIMPLEMENTED_IF_MSG(instr.tmml.UsesMiscMode(Tegra::Shader::TextureMiscMode::NDV), "NDV is not implemented"); - auto texture_type = instr.tmml.texture_type.Value(); - Node index_var{}; - const Sampler* sampler = - is_bindless ? GetBindlessSampler(instr.gpr20, index_var) : GetSampler(instr.sampler); - - if (sampler == nullptr) { + const auto texture_type = instr.tmml.texture_type.Value(); + const bool is_array = instr.tmml.array != 0; + SamplerInfo info; + info.type = texture_type; + info.is_array = is_array; + Node index_var; + const std::optional<Sampler> sampler = + is_bindless ? GetBindlessSampler(instr.gpr20, info, index_var) + : GetSampler(instr.sampler, info); + + if (!sampler) { u32 indexer = 0; for (u32 element = 0; element < 2; ++element) { if (!instr.tmml.IsComponentEnabled(element)) { @@ -299,12 +304,11 @@ u32 ShaderIR::DecodeTexture(NodeBlock& bb, u32 pc) { coords.push_back(GetRegister(instr.gpr8.Value() + 1)); break; default: - UNIMPLEMENTED_MSG("Unhandled texture type {}", static_cast<u32>(texture_type)); + UNIMPLEMENTED_MSG("Unhandled texture type {}", static_cast<int>(texture_type)); // Fallback to interpreting as a 2D texture for now coords.push_back(GetRegister(instr.gpr8.Value() + 0)); coords.push_back(GetRegister(instr.gpr8.Value() + 1)); - texture_type = TextureType::Texture2D; } u32 indexer = 0; for (u32 element = 0; element < 2; ++element) { @@ -353,98 +357,103 @@ u32 ShaderIR::DecodeTexture(NodeBlock& bb, u32 pc) { return pc; } -ShaderIR::SamplerInfo ShaderIR::GetSamplerInfo(std::optional<SamplerInfo> sampler_info, u32 offset, +ShaderIR::SamplerInfo ShaderIR::GetSamplerInfo(SamplerInfo info, u32 offset, std::optional<u32> buffer) { - if (sampler_info) { - return *sampler_info; + if (info.IsComplete()) { + return info; } const auto sampler = buffer ? registry.ObtainBindlessSampler(*buffer, offset) : registry.ObtainBoundSampler(offset); if (!sampler) { LOG_WARNING(HW_GPU, "Unknown sampler info"); - return SamplerInfo{TextureType::Texture2D, false, false, false}; - } - return SamplerInfo{sampler->texture_type, sampler->is_array != 0, sampler->is_shadow != 0, - sampler->is_buffer != 0}; + info.type = info.type.value_or(Tegra::Shader::TextureType::Texture2D); + info.is_array = info.is_array.value_or(false); + info.is_shadow = info.is_shadow.value_or(false); + info.is_buffer = info.is_buffer.value_or(false); + return info; + } + info.type = info.type.value_or(sampler->texture_type); + info.is_array = info.is_array.value_or(sampler->is_array != 0); + info.is_shadow = info.is_shadow.value_or(sampler->is_shadow != 0); + info.is_buffer = info.is_buffer.value_or(sampler->is_buffer != 0); + return info; } -const Sampler* ShaderIR::GetSampler(const Tegra::Shader::Sampler& sampler, - std::optional<SamplerInfo> sampler_info) { +std::optional<Sampler> ShaderIR::GetSampler(Tegra::Shader::Sampler sampler, + SamplerInfo sampler_info) { const auto offset = static_cast<u32>(sampler.index.Value()); const auto info = GetSamplerInfo(sampler_info, offset); // If this sampler has already been used, return the existing mapping. - const auto it = - std::find_if(used_samplers.begin(), used_samplers.end(), - [offset](const Sampler& entry) { return entry.GetOffset() == offset; }); + const auto it = std::find_if(used_samplers.begin(), used_samplers.end(), + [offset](const Sampler& entry) { return entry.offset == offset; }); if (it != used_samplers.end()) { - ASSERT(!it->IsBindless() && it->GetType() == info.type && it->IsArray() == info.is_array && - it->IsShadow() == info.is_shadow && it->IsBuffer() == info.is_buffer); - return &*it; + ASSERT(!it->is_bindless && it->type == info.type && it->is_array == info.is_array && + it->is_shadow == info.is_shadow && it->is_buffer == info.is_buffer); + return *it; } // Otherwise create a new mapping for this sampler const auto next_index = static_cast<u32>(used_samplers.size()); - return &used_samplers.emplace_back(next_index, offset, info.type, info.is_array, info.is_shadow, - info.is_buffer, false); + return used_samplers.emplace_back(next_index, offset, *info.type, *info.is_array, + *info.is_shadow, *info.is_buffer, false); } -const Sampler* ShaderIR::GetBindlessSampler(Tegra::Shader::Register reg, Node& index_var, - std::optional<SamplerInfo> sampler_info) { +std::optional<Sampler> ShaderIR::GetBindlessSampler(Tegra::Shader::Register reg, SamplerInfo info, + Node& index_var) { const Node sampler_register = GetRegister(reg); const auto [base_node, tracked_sampler_info] = TrackBindlessSampler(sampler_register, global_code, static_cast<s64>(global_code.size())); ASSERT(base_node != nullptr); if (base_node == nullptr) { - return nullptr; + return std::nullopt; } if (const auto bindless_sampler_info = std::get_if<BindlessSamplerNode>(&*tracked_sampler_info)) { const u32 buffer = bindless_sampler_info->GetIndex(); const u32 offset = bindless_sampler_info->GetOffset(); - const auto info = GetSamplerInfo(sampler_info, offset, buffer); + info = GetSamplerInfo(info, offset, buffer); // If this sampler has already been used, return the existing mapping. - const auto it = - std::find_if(used_samplers.begin(), used_samplers.end(), - [buffer = buffer, offset = offset](const Sampler& entry) { - return entry.GetBuffer() == buffer && entry.GetOffset() == offset; - }); + const auto it = std::find_if(used_samplers.begin(), used_samplers.end(), + [buffer = buffer, offset = offset](const Sampler& entry) { + return entry.buffer == buffer && entry.offset == offset; + }); if (it != used_samplers.end()) { - ASSERT(it->IsBindless() && it->GetType() == info.type && - it->IsArray() == info.is_array && it->IsShadow() == info.is_shadow); - return &*it; + ASSERT(it->is_bindless && it->type == info.type && it->is_array == info.is_array && + it->is_shadow == info.is_shadow); + return *it; } // Otherwise create a new mapping for this sampler const auto next_index = static_cast<u32>(used_samplers.size()); - return &used_samplers.emplace_back(next_index, offset, buffer, info.type, info.is_array, - info.is_shadow, info.is_buffer, false); - } else if (const auto array_sampler_info = - std::get_if<ArraySamplerNode>(&*tracked_sampler_info)) { + return used_samplers.emplace_back(next_index, offset, buffer, *info.type, *info.is_array, + *info.is_shadow, *info.is_buffer, false); + } + if (const auto array_sampler_info = std::get_if<ArraySamplerNode>(&*tracked_sampler_info)) { const u32 base_offset = array_sampler_info->GetBaseOffset() / 4; index_var = GetCustomVariable(array_sampler_info->GetIndexVar()); - const auto info = GetSamplerInfo(sampler_info, base_offset); + info = GetSamplerInfo(info, base_offset); // If this sampler has already been used, return the existing mapping. const auto it = std::find_if( used_samplers.begin(), used_samplers.end(), - [base_offset](const Sampler& entry) { return entry.GetOffset() == base_offset; }); + [base_offset](const Sampler& entry) { return entry.offset == base_offset; }); if (it != used_samplers.end()) { - ASSERT(!it->IsBindless() && it->GetType() == info.type && - it->IsArray() == info.is_array && it->IsShadow() == info.is_shadow && - it->IsBuffer() == info.is_buffer && it->IsIndexed()); - return &*it; + ASSERT(!it->is_bindless && it->type == info.type && it->is_array == info.is_array && + it->is_shadow == info.is_shadow && it->is_buffer == info.is_buffer && + it->is_indexed); + return *it; } uses_indexed_samplers = true; // Otherwise create a new mapping for this sampler const auto next_index = static_cast<u32>(used_samplers.size()); - return &used_samplers.emplace_back(next_index, base_offset, info.type, info.is_array, - info.is_shadow, info.is_buffer, true); + return used_samplers.emplace_back(next_index, base_offset, *info.type, *info.is_array, + *info.is_shadow, *info.is_buffer, true); } - return nullptr; + return std::nullopt; } void ShaderIR::WriteTexInstructionFloat(NodeBlock& bb, Instruction instr, const Node4& components) { @@ -529,10 +538,16 @@ Node4 ShaderIR::GetTextureCode(Instruction instr, TextureType texture_type, ASSERT_MSG(texture_type != TextureType::Texture3D || !is_array || !is_shadow, "Illegal texture type"); - const SamplerInfo info{texture_type, is_array, is_shadow, false}; + SamplerInfo info; + info.type = texture_type; + info.is_array = is_array; + info.is_shadow = is_shadow; + info.is_buffer = false; + Node index_var; - const Sampler* sampler = is_bindless ? GetBindlessSampler(*bindless_reg, index_var, info) - : GetSampler(instr.sampler, info); + const std::optional<Sampler> sampler = is_bindless + ? GetBindlessSampler(*bindless_reg, info, index_var) + : GetSampler(instr.sampler, info); if (!sampler) { return {Immediate(0), Immediate(0), Immediate(0), Immediate(0)}; } @@ -683,12 +698,17 @@ Node4 ShaderIR::GetTld4Code(Instruction instr, TextureType texture_type, bool de u64 parameter_register = instr.gpr20.Value(); - const SamplerInfo info{texture_type, is_array, depth_compare, false}; - Node index_var{}; - const Sampler* sampler = is_bindless ? GetBindlessSampler(parameter_register++, index_var, info) - : GetSampler(instr.sampler, info); + SamplerInfo info; + info.type = texture_type; + info.is_array = is_array; + info.is_shadow = depth_compare; + + Node index_var; + const std::optional<Sampler> sampler = + is_bindless ? GetBindlessSampler(parameter_register++, info, index_var) + : GetSampler(instr.sampler, info); Node4 values; - if (sampler == nullptr) { + if (!sampler) { for (u32 element = 0; element < values.size(); ++element) { values[element] = Immediate(0); } @@ -743,12 +763,12 @@ Node4 ShaderIR::GetTldCode(Tegra::Shader::Instruction instr) { // const Node aoffi_register{is_aoffi ? GetRegister(gpr20_cursor++) : nullptr}; // const Node multisample{is_multisample ? GetRegister(gpr20_cursor++) : nullptr}; - const auto& sampler = *GetSampler(instr.sampler); + const std::optional<Sampler> sampler = GetSampler(instr.sampler, {}); Node4 values; for (u32 element = 0; element < values.size(); ++element) { auto coords_copy = coords; - MetaTexture meta{sampler, array_register, {}, {}, {}, {}, {}, lod, {}, element, {}}; + MetaTexture meta{*sampler, array_register, {}, {}, {}, {}, {}, lod, {}, element, {}}; values[element] = Operation(OperationCode::TexelFetch, meta, std::move(coords_copy)); } @@ -756,7 +776,11 @@ Node4 ShaderIR::GetTldCode(Tegra::Shader::Instruction instr) { } Node4 ShaderIR::GetTldsCode(Instruction instr, TextureType texture_type, bool is_array) { - const Sampler& sampler = *GetSampler(instr.sampler); + SamplerInfo info; + info.type = texture_type; + info.is_array = is_array; + info.is_shadow = false; + const std::optional<Sampler> sampler = GetSampler(instr.sampler, info); const std::size_t type_coord_count = GetCoordCount(texture_type); const bool lod_enabled = instr.tlds.GetTextureProcessMode() == TextureProcessMode::LL; @@ -784,7 +808,7 @@ Node4 ShaderIR::GetTldsCode(Instruction instr, TextureType texture_type, bool is Node4 values; for (u32 element = 0; element < values.size(); ++element) { auto coords_copy = coords; - MetaTexture meta{sampler, array, {}, {}, {}, {}, {}, lod, {}, element, {}}; + MetaTexture meta{*sampler, array, {}, {}, {}, {}, {}, lod, {}, element, {}}; values[element] = Operation(OperationCode::TexelFetch, meta, std::move(coords_copy)); } return values; diff --git a/src/video_core/shader/node.h b/src/video_core/shader/node.h index 3f5a7bc7a..601c822d2 100644 --- a/src/video_core/shader/node.h +++ b/src/video_core/shader/node.h @@ -267,76 +267,30 @@ class ArraySamplerNode; using TrackSamplerData = std::variant<BindlessSamplerNode, ArraySamplerNode>; using TrackSampler = std::shared_ptr<TrackSamplerData>; -class Sampler { -public: - /// This constructor is for bound samplers +struct Sampler { + /// Bound samplers constructor constexpr explicit Sampler(u32 index, u32 offset, Tegra::Shader::TextureType type, bool is_array, bool is_shadow, bool is_buffer, bool is_indexed) : index{index}, offset{offset}, type{type}, is_array{is_array}, is_shadow{is_shadow}, is_buffer{is_buffer}, is_indexed{is_indexed} {} - /// This constructor is for bindless samplers + /// Bindless samplers constructor constexpr explicit Sampler(u32 index, u32 offset, u32 buffer, Tegra::Shader::TextureType type, bool is_array, bool is_shadow, bool is_buffer, bool is_indexed) : index{index}, offset{offset}, buffer{buffer}, type{type}, is_array{is_array}, is_shadow{is_shadow}, is_buffer{is_buffer}, is_bindless{true}, is_indexed{is_indexed} {} - constexpr u32 GetIndex() const { - return index; - } - - constexpr u32 GetOffset() const { - return offset; - } - - constexpr u32 GetBuffer() const { - return buffer; - } - - constexpr Tegra::Shader::TextureType GetType() const { - return type; - } - - constexpr bool IsArray() const { - return is_array; - } - - constexpr bool IsShadow() const { - return is_shadow; - } - - constexpr bool IsBuffer() const { - return is_buffer; - } - - constexpr bool IsBindless() const { - return is_bindless; - } - - constexpr bool IsIndexed() const { - return is_indexed; - } - - constexpr u32 Size() const { - return size; - } - - constexpr void SetSize(u32 new_size) { - size = new_size; - } - -private: - u32 index{}; ///< Emulated index given for the this sampler. - u32 offset{}; ///< Offset in the const buffer from where the sampler is being read. - u32 buffer{}; ///< Buffer where the bindless sampler is being read (unused on bound samplers). - u32 size{1}; ///< Size of the sampler. + u32 index = 0; ///< Emulated index given for the this sampler. + u32 offset = 0; ///< Offset in the const buffer from where the sampler is being read. + u32 buffer = 0; ///< Buffer where the bindless sampler is being read (unused on bound samplers). + u32 size = 1; ///< Size of the sampler. Tegra::Shader::TextureType type{}; ///< The type used to sample this texture (Texture2D, etc) - bool is_array{}; ///< Whether the texture is being sampled as an array texture or not. - bool is_shadow{}; ///< Whether the texture is being sampled as a depth texture or not. - bool is_buffer{}; ///< Whether the texture is a texture buffer without sampler. - bool is_bindless{}; ///< Whether this sampler belongs to a bindless texture or not. - bool is_indexed{}; ///< Whether this sampler is an indexed array of textures. + bool is_array = false; ///< Whether the texture is being sampled as an array texture or not. + bool is_shadow = false; ///< Whether the texture is being sampled as a depth texture or not. + bool is_buffer = false; ///< Whether the texture is a texture buffer without sampler. + bool is_bindless = false; ///< Whether this sampler belongs to a bindless texture or not. + bool is_indexed = false; ///< Whether this sampler is an indexed array of textures. }; /// Represents a tracked bindless sampler into a direct const buffer @@ -381,13 +335,13 @@ private: u32 offset; }; -class Image final { +struct Image { public: - /// This constructor is for bound images + /// Bound images constructor constexpr explicit Image(u32 index, u32 offset, Tegra::Shader::ImageType type) : index{index}, offset{offset}, type{type} {} - /// This constructor is for bindless samplers + /// Bindless samplers constructor constexpr explicit Image(u32 index, u32 offset, u32 buffer, Tegra::Shader::ImageType type) : index{index}, offset{offset}, buffer{buffer}, type{type}, is_bindless{true} {} @@ -405,53 +359,20 @@ public: is_atomic = true; } - constexpr u32 GetIndex() const { - return index; - } - - constexpr u32 GetOffset() const { - return offset; - } - - constexpr u32 GetBuffer() const { - return buffer; - } - - constexpr Tegra::Shader::ImageType GetType() const { - return type; - } - - constexpr bool IsBindless() const { - return is_bindless; - } - - constexpr bool IsWritten() const { - return is_written; - } - - constexpr bool IsRead() const { - return is_read; - } - - constexpr bool IsAtomic() const { - return is_atomic; - } - -private: - u32 index{}; - u32 offset{}; - u32 buffer{}; + u32 index = 0; + u32 offset = 0; + u32 buffer = 0; Tegra::Shader::ImageType type{}; - bool is_bindless{}; - bool is_written{}; - bool is_read{}; - bool is_atomic{}; + bool is_bindless = false; + bool is_written = false; + bool is_read = false; + bool is_atomic = false; }; struct GlobalMemoryBase { - u32 cbuf_index{}; - u32 cbuf_offset{}; + u32 cbuf_index = 0; + u32 cbuf_offset = 0; bool operator<(const GlobalMemoryBase& rhs) const { return std::tie(cbuf_index, cbuf_offset) < std::tie(rhs.cbuf_index, rhs.cbuf_offset); @@ -465,7 +386,7 @@ struct MetaArithmetic { /// Parameters describing a texture sampler struct MetaTexture { - const Sampler& sampler; + Sampler sampler; Node array; Node depth_compare; std::vector<Node> aoffi; diff --git a/src/video_core/shader/shader_ir.h b/src/video_core/shader/shader_ir.h index 69de5e68b..15ae152f2 100644 --- a/src/video_core/shader/shader_ir.h +++ b/src/video_core/shader/shader_ir.h @@ -28,12 +28,11 @@ struct ShaderBlock; constexpr u32 MAX_PROGRAM_LENGTH = 0x1000; -class ConstBuffer { -public: - explicit ConstBuffer(u32 max_offset, bool is_indirect) +struct ConstBuffer { + constexpr explicit ConstBuffer(u32 max_offset, bool is_indirect) : max_offset{max_offset}, is_indirect{is_indirect} {} - ConstBuffer() = default; + constexpr ConstBuffer() = default; void MarkAsUsed(u64 offset) { max_offset = std::max(max_offset, static_cast<u32>(offset)); @@ -56,8 +55,8 @@ public: } private: - u32 max_offset{}; - bool is_indirect{}; + u32 max_offset = 0; + bool is_indirect = false; }; struct GlobalMemoryUsage { @@ -191,10 +190,14 @@ private: friend class ASTDecoder; struct SamplerInfo { - Tegra::Shader::TextureType type; - bool is_array; - bool is_shadow; - bool is_buffer; + std::optional<Tegra::Shader::TextureType> type; + std::optional<bool> is_array; + std::optional<bool> is_shadow; + std::optional<bool> is_buffer; + + constexpr bool IsComplete() const noexcept { + return type && is_array && is_shadow && is_buffer; + } }; void Decode(); @@ -327,16 +330,15 @@ private: OperationCode GetPredicateCombiner(Tegra::Shader::PredOperation operation); /// Queries the missing sampler info from the execution context. - SamplerInfo GetSamplerInfo(std::optional<SamplerInfo> sampler_info, u32 offset, + SamplerInfo GetSamplerInfo(SamplerInfo info, u32 offset, std::optional<u32> buffer = std::nullopt); - /// Accesses a texture sampler - const Sampler* GetSampler(const Tegra::Shader::Sampler& sampler, - std::optional<SamplerInfo> sampler_info = std::nullopt); + /// Accesses a texture sampler. + std::optional<Sampler> GetSampler(Tegra::Shader::Sampler sampler, SamplerInfo info); /// Accesses a texture sampler for a bindless texture. - const Sampler* GetBindlessSampler(Tegra::Shader::Register reg, Node& index_var, - std::optional<SamplerInfo> sampler_info = std::nullopt); + std::optional<Sampler> GetBindlessSampler(Tegra::Shader::Register reg, SamplerInfo info, + Node& index_var); /// Accesses an image. Image& GetImage(Tegra::Shader::Image image, Tegra::Shader::ImageType type); diff --git a/src/video_core/texture_cache/surface_params.cpp b/src/video_core/texture_cache/surface_params.cpp index 0de499946..884fabffe 100644 --- a/src/video_core/texture_cache/surface_params.cpp +++ b/src/video_core/texture_cache/surface_params.cpp @@ -81,7 +81,7 @@ SurfaceParams SurfaceParams::CreateForTexture(const FormatLookupTable& lookup_ta params.pixel_format = lookup_table.GetPixelFormat( tic.format, params.srgb_conversion, tic.r_type, tic.g_type, tic.b_type, tic.a_type); params.type = GetFormatType(params.pixel_format); - if (entry.IsShadow() && params.type == SurfaceType::ColorTexture) { + if (entry.is_shadow && params.type == SurfaceType::ColorTexture) { switch (params.pixel_format) { case PixelFormat::R16U: case PixelFormat::R16F: @@ -108,7 +108,7 @@ SurfaceParams SurfaceParams::CreateForTexture(const FormatLookupTable& lookup_ta params.emulated_levels = 1; params.is_layered = false; } else { - params.target = TextureTypeToSurfaceTarget(entry.GetType(), entry.IsArray()); + params.target = TextureTypeToSurfaceTarget(entry.type, entry.is_array); params.width = tic.Width(); params.height = tic.Height(); params.depth = tic.Depth(); @@ -138,7 +138,7 @@ SurfaceParams SurfaceParams::CreateForImage(const FormatLookupTable& lookup_tabl tic.format, params.srgb_conversion, tic.r_type, tic.g_type, tic.b_type, tic.a_type); params.type = GetFormatType(params.pixel_format); params.type = GetFormatType(params.pixel_format); - params.target = ImageTypeToSurfaceTarget(entry.GetType()); + params.target = ImageTypeToSurfaceTarget(entry.type); // TODO: on 1DBuffer we should use the tic info. if (tic.IsBuffer()) { params.target = SurfaceTarget::TextureBuffer; @@ -248,12 +248,12 @@ SurfaceParams SurfaceParams::CreateForFermiCopySurface( VideoCore::Surface::SurfaceTarget SurfaceParams::ExpectedTarget( const VideoCommon::Shader::Sampler& entry) { - return TextureTypeToSurfaceTarget(entry.GetType(), entry.IsArray()); + return TextureTypeToSurfaceTarget(entry.type, entry.is_array); } VideoCore::Surface::SurfaceTarget SurfaceParams::ExpectedTarget( const VideoCommon::Shader::Image& entry) { - return ImageTypeToSurfaceTarget(entry.GetType()); + return ImageTypeToSurfaceTarget(entry.type); } bool SurfaceParams::IsLayered() const { diff --git a/src/video_core/texture_cache/texture_cache.h b/src/video_core/texture_cache/texture_cache.h index d2d2846e6..d6efc34b2 100644 --- a/src/video_core/texture_cache/texture_cache.h +++ b/src/video_core/texture_cache/texture_cache.h @@ -1172,7 +1172,7 @@ private: /// Returns true the shader sampler entry is compatible with the TIC texture type. static bool IsTypeCompatible(Tegra::Texture::TextureType tic_type, const VideoCommon::Shader::Sampler& entry) { - const auto shader_type = entry.GetType(); + const auto shader_type = entry.type; switch (tic_type) { case Tegra::Texture::TextureType::Texture1D: case Tegra::Texture::TextureType::Texture1DArray: @@ -1193,7 +1193,7 @@ private: if (shader_type == Tegra::Shader::TextureType::TextureCube) { return true; } - return shader_type == Tegra::Shader::TextureType::Texture2D && entry.IsArray(); + return shader_type == Tegra::Shader::TextureType::Texture2D && entry.is_array; } UNREACHABLE(); return true; diff --git a/src/yuzu/bootmanager.cpp b/src/yuzu/bootmanager.cpp index 1cac2f942..3d759f77b 100644 --- a/src/yuzu/bootmanager.cpp +++ b/src/yuzu/bootmanager.cpp @@ -292,6 +292,8 @@ GRenderWindow::GRenderWindow(GMainWindow* parent_, EmuThread* emu_thread_) setLayout(layout); InputCommon::Init(); + this->setMouseTracking(true); + connect(this, &GRenderWindow::FirstFrameDisplayed, parent_, &GMainWindow::OnLoadComplete); } @@ -385,6 +387,7 @@ void GRenderWindow::mousePressEvent(QMouseEvent* event) { } else if (event->button() == Qt::RightButton) { InputCommon::GetMotionEmu()->BeginTilt(pos.x(), pos.y()); } + QWidget::mousePressEvent(event); } void GRenderWindow::mouseMoveEvent(QMouseEvent* event) { @@ -397,6 +400,7 @@ void GRenderWindow::mouseMoveEvent(QMouseEvent* event) { const auto [x, y] = ScaleTouch(pos); this->TouchMoved(x, y); InputCommon::GetMotionEmu()->Tilt(pos.x(), pos.y()); + QWidget::mouseMoveEvent(event); } void GRenderWindow::mouseReleaseEvent(QMouseEvent* event) { diff --git a/src/yuzu/configuration/config.cpp b/src/yuzu/configuration/config.cpp index a44eed047..75c6cf20b 100644 --- a/src/yuzu/configuration/config.cpp +++ b/src/yuzu/configuration/config.cpp @@ -743,6 +743,8 @@ void Config::ReadUIValues() { UISettings::values.profile_index = ReadSetting(QStringLiteral("profileIndex"), 0).toUInt(); UISettings::values.pause_when_in_background = ReadSetting(QStringLiteral("pauseWhenInBackground"), false).toBool(); + UISettings::values.hide_mouse = + ReadSetting(QStringLiteral("hideInactiveMouse"), false).toBool(); ApplyDefaultProfileIfInputInvalid(); @@ -1169,6 +1171,7 @@ void Config::SaveUIValues() { WriteSetting(QStringLiteral("profileIndex"), UISettings::values.profile_index, 0); WriteSetting(QStringLiteral("pauseWhenInBackground"), UISettings::values.pause_when_in_background, false); + WriteSetting(QStringLiteral("hideInactiveMouse"), UISettings::values.hide_mouse, false); qt_config->endGroup(); } diff --git a/src/yuzu/configuration/configure_general.cpp b/src/yuzu/configuration/configure_general.cpp index 5ef927114..cb95423e0 100644 --- a/src/yuzu/configuration/configure_general.cpp +++ b/src/yuzu/configuration/configure_general.cpp @@ -26,6 +26,7 @@ void ConfigureGeneral::SetConfiguration() { ui->toggle_check_exit->setChecked(UISettings::values.confirm_before_closing); ui->toggle_user_on_boot->setChecked(UISettings::values.select_user_on_boot); ui->toggle_background_pause->setChecked(UISettings::values.pause_when_in_background); + ui->toggle_hide_mouse->setChecked(UISettings::values.hide_mouse); ui->toggle_frame_limit->setChecked(Settings::values.use_frame_limit); ui->frame_limit->setEnabled(ui->toggle_frame_limit->isChecked()); @@ -36,6 +37,7 @@ void ConfigureGeneral::ApplyConfiguration() { UISettings::values.confirm_before_closing = ui->toggle_check_exit->isChecked(); UISettings::values.select_user_on_boot = ui->toggle_user_on_boot->isChecked(); UISettings::values.pause_when_in_background = ui->toggle_background_pause->isChecked(); + UISettings::values.hide_mouse = ui->toggle_hide_mouse->isChecked(); Settings::values.use_frame_limit = ui->toggle_frame_limit->isChecked(); Settings::values.frame_limit = ui->frame_limit->value(); diff --git a/src/yuzu/configuration/configure_general.ui b/src/yuzu/configuration/configure_general.ui index 857119bb3..fc3b7e65a 100644 --- a/src/yuzu/configuration/configure_general.ui +++ b/src/yuzu/configuration/configure_general.ui @@ -72,6 +72,13 @@ </property> </widget> </item> + <item> + <widget class="QCheckBox" name="toggle_hide_mouse"> + <property name="text"> + <string>Hide mouse on inactivity</string> + </property> + </widget> + </item> </layout> </item> </layout> diff --git a/src/yuzu/main.cpp b/src/yuzu/main.cpp index b44b4276c..0a6839b2d 100644 --- a/src/yuzu/main.cpp +++ b/src/yuzu/main.cpp @@ -135,6 +135,8 @@ __declspec(dllexport) int AmdPowerXpressRequestHighPerformance = 1; } #endif +constexpr int default_mouse_timeout = 2500; + constexpr u64 DLC_BASE_TITLE_ID_MASK = 0xFFFFFFFFFFFFE000; /** @@ -236,6 +238,14 @@ GMainWindow::GMainWindow() // Show one-time "callout" messages to the user ShowTelemetryCallout(); + // make sure menubar has the arrow cursor instead of inheriting from this + ui.menubar->setCursor(QCursor()); + statusBar()->setCursor(QCursor()); + + mouse_hide_timer.setInterval(default_mouse_timeout); + connect(&mouse_hide_timer, &QTimer::timeout, this, &GMainWindow::HideMouseCursor); + connect(ui.menubar, &QMenuBar::hovered, this, &GMainWindow::ShowMouseCursor); + QStringList args = QApplication::arguments(); if (args.length() >= 2) { BootGame(args[1]); @@ -1012,6 +1022,13 @@ void GMainWindow::BootGame(const QString& filename) { async_status_button->setDisabled(true); renderer_status_button->setDisabled(true); + if (UISettings::values.hide_mouse) { + mouse_hide_timer.start(); + setMouseTracking(true); + ui.centralwidget->setMouseTracking(true); + ui.menubar->setMouseTracking(true); + } + const u64 title_id = Core::System::GetInstance().CurrentProcess()->GetTitleID(); std::string title_name; @@ -1080,6 +1097,10 @@ void GMainWindow::ShutdownGame() { game_list->show(); game_list->setFilterFocus(); + setMouseTracking(false); + ui.centralwidget->setMouseTracking(false); + ui.menubar->setMouseTracking(false); + UpdateWindowTitle(); // Disable status bar updates @@ -1837,6 +1858,17 @@ void GMainWindow::OnConfigure() { config->Save(); + if (UISettings::values.hide_mouse && emulation_running) { + setMouseTracking(true); + ui.centralwidget->setMouseTracking(true); + ui.menubar->setMouseTracking(true); + mouse_hide_timer.start(); + } else { + setMouseTracking(false); + ui.centralwidget->setMouseTracking(false); + ui.menubar->setMouseTracking(false); + } + dock_status_button->setChecked(Settings::values.use_docked_mode); async_status_button->setChecked(Settings::values.use_asynchronous_gpu_emulation); #ifdef HAS_VULKAN @@ -1970,6 +2002,30 @@ void GMainWindow::UpdateStatusBar() { emu_frametime_label->setVisible(true); } +void GMainWindow::HideMouseCursor() { + if (emu_thread == nullptr || UISettings::values.hide_mouse == false) { + mouse_hide_timer.stop(); + ShowMouseCursor(); + return; + } + setCursor(QCursor(Qt::BlankCursor)); +} + +void GMainWindow::ShowMouseCursor() { + unsetCursor(); + if (emu_thread != nullptr && UISettings::values.hide_mouse) { + mouse_hide_timer.start(); + } +} + +void GMainWindow::mouseMoveEvent(QMouseEvent* event) { + ShowMouseCursor(); +} + +void GMainWindow::mousePressEvent(QMouseEvent* event) { + ShowMouseCursor(); +} + void GMainWindow::OnCoreError(Core::System::ResultStatus result, std::string details) { QMessageBox::StandardButton answer; QString status_message; diff --git a/src/yuzu/main.h b/src/yuzu/main.h index 0b750689d..60b17c54a 100644 --- a/src/yuzu/main.h +++ b/src/yuzu/main.h @@ -216,6 +216,8 @@ private: std::optional<u64> SelectRomFSDumpTarget(const FileSys::ContentProvider&, u64 program_id); void UpdateWindowTitle(const QString& title_name = {}); void UpdateStatusBar(); + void HideMouseCursor(); + void ShowMouseCursor(); Ui::MainWindow ui; @@ -244,6 +246,7 @@ private: QString game_path; bool auto_paused = false; + QTimer mouse_hide_timer; // FS std::shared_ptr<FileSys::VfsFilesystem> vfs; @@ -265,4 +268,6 @@ protected: void dropEvent(QDropEvent* event) override; void dragEnterEvent(QDragEnterEvent* event) override; void dragMoveEvent(QDragMoveEvent* event) override; + void mouseMoveEvent(QMouseEvent* event) override; + void mousePressEvent(QMouseEvent* event) override; }; diff --git a/src/yuzu/uisettings.h b/src/yuzu/uisettings.h index a675ecf4d..830932d45 100644 --- a/src/yuzu/uisettings.h +++ b/src/yuzu/uisettings.h @@ -59,6 +59,7 @@ struct Values { bool confirm_before_closing; bool first_start; bool pause_when_in_background; + bool hide_mouse; bool select_user_on_boot; |