diff options
Diffstat (limited to '')
48 files changed, 1004 insertions, 472 deletions
diff --git a/src/core/hle/kernel/memory/memory_manager.cpp b/src/core/hle/kernel/memory/memory_manager.cpp index 616148190..acf13585c 100644 --- a/src/core/hle/kernel/memory/memory_manager.cpp +++ b/src/core/hle/kernel/memory/memory_manager.cpp @@ -139,7 +139,6 @@ ResultCode MemoryManager::Allocate(PageLinkedList& page_list, std::size_t num_pa } // Only succeed if we allocated as many pages as we wanted - ASSERT(num_pages >= 0); if (num_pages) { return ERR_OUT_OF_MEMORY; } diff --git a/src/core/hle/service/am/am.cpp b/src/core/hle/service/am/am.cpp index 4df74c4f9..20f366635 100644 --- a/src/core/hle/service/am/am.cpp +++ b/src/core/hle/service/am/am.cpp @@ -68,6 +68,7 @@ IWindowController::IWindowController(Core::System& system_) static const FunctionInfo functions[] = { {0, nullptr, "CreateWindow"}, {1, &IWindowController::GetAppletResourceUserId, "GetAppletResourceUserId"}, + {2, nullptr, "GetAppletResourceUserIdOfCallerApplet"}, {10, &IWindowController::AcquireForegroundRights, "AcquireForegroundRights"}, {11, nullptr, "ReleaseForegroundRights"}, {12, nullptr, "RejectToChangeIntoBackground"}, @@ -189,8 +190,8 @@ IDisplayController::IDisplayController() : ServiceFramework("IDisplayController" {5, nullptr, "GetLastForegroundCaptureImageEx"}, {6, nullptr, "GetLastApplicationCaptureImageEx"}, {7, nullptr, "GetCallerAppletCaptureImageEx"}, - {8, nullptr, "TakeScreenShotOfOwnLayer"}, // 2.0.0+ - {9, nullptr, "CopyBetweenCaptureBuffers"}, // 5.0.0+ + {8, nullptr, "TakeScreenShotOfOwnLayer"}, + {9, nullptr, "CopyBetweenCaptureBuffers"}, {10, nullptr, "AcquireLastApplicationCaptureBuffer"}, {11, nullptr, "ReleaseLastApplicationCaptureBuffer"}, {12, nullptr, "AcquireLastForegroundCaptureBuffer"}, @@ -200,17 +201,14 @@ IDisplayController::IDisplayController() : ServiceFramework("IDisplayController" {16, nullptr, "AcquireLastApplicationCaptureBufferEx"}, {17, nullptr, "AcquireLastForegroundCaptureBufferEx"}, {18, nullptr, "AcquireCallerAppletCaptureBufferEx"}, - // 2.0.0+ {20, nullptr, "ClearCaptureBuffer"}, {21, nullptr, "ClearAppletTransitionBuffer"}, - // 4.0.0+ {22, nullptr, "AcquireLastApplicationCaptureSharedBuffer"}, {23, nullptr, "ReleaseLastApplicationCaptureSharedBuffer"}, {24, nullptr, "AcquireLastForegroundCaptureSharedBuffer"}, {25, nullptr, "ReleaseLastForegroundCaptureSharedBuffer"}, {26, nullptr, "AcquireCallerAppletCaptureSharedBuffer"}, {27, nullptr, "ReleaseCallerAppletCaptureSharedBuffer"}, - // 6.0.0+ {28, nullptr, "TakeScreenShotOfOwnLayerEx"}, }; // clang-format on @@ -225,7 +223,7 @@ IDebugFunctions::IDebugFunctions() : ServiceFramework{"IDebugFunctions"} { static const FunctionInfo functions[] = { {0, nullptr, "NotifyMessageToHomeMenuForDebug"}, {1, nullptr, "OpenMainApplication"}, - {10, nullptr, "EmulateButtonEvent"}, + {10, nullptr, "PerformSystemButtonPressing"}, {20, nullptr, "InvalidateTransitionLayer"}, {30, nullptr, "RequestLaunchApplicationWithUserAndArgumentForDebug"}, {40, nullptr, "GetAppletResourceUsageInfo"}, @@ -267,7 +265,7 @@ ISelfController::ISelfController(Core::System& system, {16, &ISelfController::SetOutOfFocusSuspendingEnabled, "SetOutOfFocusSuspendingEnabled"}, {17, nullptr, "SetControllerFirmwareUpdateSection"}, {18, nullptr, "SetRequiresCaptureButtonShortPressedMessage"}, - {19, &ISelfController::SetScreenShotImageOrientation, "SetScreenShotImageOrientation"}, + {19, &ISelfController::SetAlbumImageOrientation, "SetAlbumImageOrientation"}, {20, nullptr, "SetDesirableKeyboardLayout"}, {40, &ISelfController::CreateManagedDisplayLayer, "CreateManagedDisplayLayer"}, {41, nullptr, "IsSystemBufferSharingEnabled"}, @@ -443,7 +441,7 @@ void ISelfController::SetOutOfFocusSuspendingEnabled(Kernel::HLERequestContext& rb.Push(RESULT_SUCCESS); } -void ISelfController::SetScreenShotImageOrientation(Kernel::HLERequestContext& ctx) { +void ISelfController::SetAlbumImageOrientation(Kernel::HLERequestContext& ctx) { LOG_WARNING(Service_AM, "(STUBBED) called"); IPC::ResponseBuilder rb{ctx, 2}; @@ -607,6 +605,7 @@ ICommonStateGetter::ICommonStateGetter(Core::System& system, {20, nullptr, "PushToGeneralChannel"}, {30, nullptr, "GetHomeButtonReaderLockAccessor"}, {31, nullptr, "GetReaderLockAccessorEx"}, + {32, nullptr, "GetWriterLockAccessorEx"}, {40, nullptr, "GetCradleFwVersion"}, {50, &ICommonStateGetter::IsVrModeEnabled, "IsVrModeEnabled"}, {51, &ICommonStateGetter::SetVrModeEnabled, "SetVrModeEnabled"}, @@ -1132,6 +1131,7 @@ IApplicationFunctions::IApplicationFunctions(Core::System& system_) {24, nullptr, "GetLaunchStorageInfoForDebug"}, {25, &IApplicationFunctions::ExtendSaveData, "ExtendSaveData"}, {26, &IApplicationFunctions::GetSaveDataSize, "GetSaveDataSize"}, + {27, nullptr, "CreateCacheStorage"}, {30, &IApplicationFunctions::BeginBlockingHomeButtonShortAndLongPressed, "BeginBlockingHomeButtonShortAndLongPressed"}, {31, &IApplicationFunctions::EndBlockingHomeButtonShortAndLongPressed, "EndBlockingHomeButtonShortAndLongPressed"}, {32, &IApplicationFunctions::BeginBlockingHomeButton, "BeginBlockingHomeButton"}, @@ -1157,6 +1157,8 @@ IApplicationFunctions::IApplicationFunctions(Core::System& system_) {120, nullptr, "ExecuteProgram"}, {121, nullptr, "ClearUserChannel"}, {122, nullptr, "UnpopToUserChannel"}, + {123, nullptr, "GetPreviousProgramIndex"}, + {124, nullptr, "EnableApplicationAllThreadDumpOnCrash"}, {130, &IApplicationFunctions::GetGpuErrorDetectedSystemEvent, "GetGpuErrorDetectedSystemEvent"}, {140, &IApplicationFunctions::GetFriendInvitationStorageChannelEvent, "GetFriendInvitationStorageChannelEvent"}, {141, nullptr, "TryPopFromFriendInvitationStorageChannel"}, diff --git a/src/core/hle/service/am/am.h b/src/core/hle/service/am/am.h index 469f7f814..2f69466ec 100644 --- a/src/core/hle/service/am/am.h +++ b/src/core/hle/service/am/am.h @@ -138,7 +138,7 @@ private: void SetFocusHandlingMode(Kernel::HLERequestContext& ctx); void SetRestartMessageEnabled(Kernel::HLERequestContext& ctx); void SetOutOfFocusSuspendingEnabled(Kernel::HLERequestContext& ctx); - void SetScreenShotImageOrientation(Kernel::HLERequestContext& ctx); + void SetAlbumImageOrientation(Kernel::HLERequestContext& ctx); void CreateManagedDisplayLayer(Kernel::HLERequestContext& ctx); void SetHandlesRequestToDisplay(Kernel::HLERequestContext& ctx); void SetIdleTimeDetectionExtension(Kernel::HLERequestContext& ctx); diff --git a/src/core/hle/service/am/applets/software_keyboard.cpp b/src/core/hle/service/am/applets/software_keyboard.cpp index d14076b02..fbe3686ae 100644 --- a/src/core/hle/service/am/applets/software_keyboard.cpp +++ b/src/core/hle/service/am/applets/software_keyboard.cpp @@ -60,7 +60,7 @@ void SoftwareKeyboard::Initialize() { std::memcpy(&config, keyboard_config.data(), sizeof(KeyboardConfig)); const auto work_buffer_storage = broker.PopNormalDataToApplet(); - ASSERT(work_buffer_storage != nullptr); + ASSERT_OR_EXECUTE(work_buffer_storage != nullptr, { return; }); const auto& work_buffer = work_buffer_storage->GetData(); if (config.initial_string_size == 0) diff --git a/src/core/hle/service/am/spsm.cpp b/src/core/hle/service/am/spsm.cpp index 003ee8667..f27729ce7 100644 --- a/src/core/hle/service/am/spsm.cpp +++ b/src/core/hle/service/am/spsm.cpp @@ -10,17 +10,17 @@ SPSM::SPSM() : ServiceFramework{"spsm"} { // clang-format off static const FunctionInfo functions[] = { {0, nullptr, "GetState"}, - {1, nullptr, "SleepSystemAndWaitAwake"}, - {2, nullptr, "Unknown1"}, - {3, nullptr, "Unknown2"}, + {1, nullptr, "EnterSleep"}, + {2, nullptr, "GetLastWakeReason"}, + {3, nullptr, "Shutdown"}, {4, nullptr, "GetNotificationMessageEventHandle"}, - {5, nullptr, "Unknown3"}, - {6, nullptr, "Unknown4"}, - {7, nullptr, "Unknown5"}, + {5, nullptr, "ReceiveNotificationMessage"}, + {6, nullptr, "AnalyzeLogForLastSleepWakeSequence"}, + {7, nullptr, "ResetEventLog"}, {8, nullptr, "AnalyzePerformanceLogForLastSleepWakeSequence"}, {9, nullptr, "ChangeHomeButtonLongPressingTime"}, - {10, nullptr, "Unknown6"}, - {11, nullptr, "Unknown7"}, + {10, nullptr, "PutErrorState"}, + {11, nullptr, "InvalidateCurrentHomeButtonPressing"}, }; // clang-format on diff --git a/src/core/hle/service/aoc/aoc_u.cpp b/src/core/hle/service/aoc/aoc_u.cpp index 4227a4adf..8e79f707b 100644 --- a/src/core/hle/service/aoc/aoc_u.cpp +++ b/src/core/hle/service/aoc/aoc_u.cpp @@ -60,6 +60,7 @@ AOC_U::AOC_U(Core::System& system) {6, nullptr, "PrepareAddOnContentByApplicationId"}, {7, &AOC_U::PrepareAddOnContent, "PrepareAddOnContent"}, {8, &AOC_U::GetAddOnContentListChangedEvent, "GetAddOnContentListChangedEvent"}, + {9, nullptr, "GetAddOnContentLostErrorCode"}, {100, nullptr, "CreateEcPurchasedEventManager"}, {101, nullptr, "CreatePermanentEcPurchasedEventManager"}, }; diff --git a/src/core/hle/service/bcat/bcat.cpp b/src/core/hle/service/bcat/bcat.cpp index 8bb2528c9..b31766212 100644 --- a/src/core/hle/service/bcat/bcat.cpp +++ b/src/core/hle/service/bcat/bcat.cpp @@ -14,6 +14,8 @@ BCAT::BCAT(Core::System& system, std::shared_ptr<Module> module, {0, &BCAT::CreateBcatService, "CreateBcatService"}, {1, &BCAT::CreateDeliveryCacheStorageService, "CreateDeliveryCacheStorageService"}, {2, &BCAT::CreateDeliveryCacheStorageServiceWithApplicationId, "CreateDeliveryCacheStorageServiceWithApplicationId"}, + {3, nullptr, "CreateDeliveryCacheProgressService"}, + {4, nullptr, "CreateDeliveryCacheProgressServiceWithApplicationId"}, }; // clang-format on RegisterHandlers(functions); diff --git a/src/core/hle/service/bcat/module.cpp b/src/core/hle/service/bcat/module.cpp index 34aba7a27..603b64d4f 100644 --- a/src/core/hle/service/bcat/module.cpp +++ b/src/core/hle/service/bcat/module.cpp @@ -143,10 +143,13 @@ public: {20401, nullptr, "UnregisterSystemApplicationDeliveryTask"}, {20410, nullptr, "SetSystemApplicationDeliveryTaskTimer"}, {30100, &IBcatService::SetPassphrase, "SetPassphrase"}, + {30101, nullptr, "Unknown"}, + {30102, nullptr, "Unknown2"}, {30200, nullptr, "RegisterBackgroundDeliveryTask"}, {30201, nullptr, "UnregisterBackgroundDeliveryTask"}, {30202, nullptr, "BlockDeliveryTask"}, {30203, nullptr, "UnblockDeliveryTask"}, + {30210, nullptr, "SetDeliveryTaskTimer"}, {30300, nullptr, "RegisterSystemApplicationDeliveryTasks"}, {90100, nullptr, "EnumerateBackgroundDeliveryTask"}, {90200, nullptr, "GetDeliveryList"}, diff --git a/src/core/hle/service/bpc/bpc.cpp b/src/core/hle/service/bpc/bpc.cpp index 1c1ecdb60..fac6b2f9c 100644 --- a/src/core/hle/service/bpc/bpc.cpp +++ b/src/core/hle/service/bpc/bpc.cpp @@ -23,9 +23,14 @@ public: {5, nullptr, "GetBoardPowerControlEvent"}, {6, nullptr, "GetSleepButtonState"}, {7, nullptr, "GetPowerEvent"}, - {8, nullptr, "Unknown1"}, - {9, nullptr, "Unknown2"}, - {10, nullptr, "Unknown3"}, + {8, nullptr, "CreateWakeupTimer"}, + {9, nullptr, "CancelWakeupTimer"}, + {10, nullptr, "EnableWakeupTimerOnDevice"}, + {11, nullptr, "CreateWakeupTimerEx"}, + {12, nullptr, "GetLastEnabledWakeupTimerType"}, + {13, nullptr, "CleanAllWakeupTimers"}, + {14, nullptr, "Unknown"}, + {15, nullptr, "Unknown2"}, }; // clang-format on @@ -38,10 +43,11 @@ public: explicit BPC_R() : ServiceFramework{"bpc:r"} { // clang-format off static const FunctionInfo functions[] = { - {0, nullptr, "GetExternalRtcValue"}, - {1, nullptr, "SetExternalRtcValue"}, - {2, nullptr, "ReadExternalRtcResetFlag"}, - {3, nullptr, "ClearExternalRtcResetFlag"}, + {0, nullptr, "GetRtcTime"}, + {1, nullptr, "SetRtcTime"}, + {2, nullptr, "GetRtcResetDetected"}, + {3, nullptr, "ClearRtcResetDetected"}, + {4, nullptr, "SetUpRtcResetOnShutdown"}, }; // clang-format on diff --git a/src/core/hle/service/btdrv/btdrv.cpp b/src/core/hle/service/btdrv/btdrv.cpp index 40a06c9fd..f311afa2f 100644 --- a/src/core/hle/service/btdrv/btdrv.cpp +++ b/src/core/hle/service/btdrv/btdrv.cpp @@ -58,102 +58,103 @@ public: {1, nullptr, "InitializeBluetooth"}, {2, nullptr, "EnableBluetooth"}, {3, nullptr, "DisableBluetooth"}, - {4, nullptr, "CleanupBluetooth"}, + {4, nullptr, "FinalizeBluetooth"}, {5, nullptr, "GetAdapterProperties"}, {6, nullptr, "GetAdapterProperty"}, {7, nullptr, "SetAdapterProperty"}, - {8, nullptr, "StartDiscovery"}, - {9, nullptr, "CancelDiscovery"}, + {8, nullptr, "StartInquiry"}, + {9, nullptr, "StopInquiry"}, {10, nullptr, "CreateBond"}, {11, nullptr, "RemoveBond"}, {12, nullptr, "CancelBond"}, - {13, nullptr, "PinReply"}, - {14, nullptr, "SspReply"}, + {13, nullptr, "RespondToPinRequest"}, + {14, nullptr, "RespondToSspRequest"}, {15, nullptr, "GetEventInfo"}, {16, nullptr, "InitializeHid"}, - {17, nullptr, "HidConnect"}, - {18, nullptr, "HidDisconnect"}, - {19, nullptr, "HidSendData"}, - {20, nullptr, "HidSendData2"}, - {21, nullptr, "HidSetReport"}, - {22, nullptr, "HidGetReport"}, - {23, nullptr, "HidWakeController"}, - {24, nullptr, "HidAddPairedDevice"}, - {25, nullptr, "HidGetPairedDevice"}, - {26, nullptr, "CleanupHid"}, - {27, nullptr, "HidGetEventInfo"}, - {28, nullptr, "ExtSetTsi"}, - {29, nullptr, "ExtSetBurstMode"}, - {30, nullptr, "ExtSetZeroRetran"}, - {31, nullptr, "ExtSetMcMode"}, - {32, nullptr, "ExtStartLlrMode"}, - {33, nullptr, "ExtExitLlrMode"}, - {34, nullptr, "ExtSetRadio"}, - {35, nullptr, "ExtSetVisibility"}, - {36, nullptr, "ExtSetTbfcScan"}, + {17, nullptr, "OpenHidConnection"}, + {18, nullptr, "CloseHidConnection"}, + {19, nullptr, "WriteHidData"}, + {20, nullptr, "WriteHidData2"}, + {21, nullptr, "SetHidReport"}, + {22, nullptr, "GetHidReport"}, + {23, nullptr, "TriggerConnection"}, + {24, nullptr, "AddPairedDeviceInfo"}, + {25, nullptr, "GetPairedDeviceInfo"}, + {26, nullptr, "FinalizeHid"}, + {27, nullptr, "GetHidEventInfo"}, + {28, nullptr, "SetTsi"}, + {29, nullptr, "EnableBurstMode"}, + {30, nullptr, "SetZeroRetransmission"}, + {31, nullptr, "EnableMcMode"}, + {32, nullptr, "EnableLlrScan"}, + {33, nullptr, "DisableLlrScan"}, + {34, nullptr, "EnableRadio"}, + {35, nullptr, "SetVisibility"}, + {36, nullptr, "EnableTbfcScan"}, {37, nullptr, "RegisterHidReportEvent"}, - {38, nullptr, "HidGetReportEventInfo"}, + {38, nullptr, "GetHidReportEventInfo"}, {39, nullptr, "GetLatestPlr"}, - {40, nullptr, "ExtGetPendingConnections"}, + {40, nullptr, "GetPendingConnections"}, {41, nullptr, "GetChannelMap"}, - {42, nullptr, "EnableBluetoothBoostSetting"}, - {43, nullptr, "IsBluetoothBoostSettingEnabled"}, - {44, nullptr, "EnableBluetoothAfhSetting"}, - {45, nullptr, "IsBluetoothAfhSettingEnabled"}, - {46, nullptr, "InitializeBluetoothLe"}, - {47, nullptr, "EnableBluetoothLe"}, - {48, nullptr, "DisableBluetoothLe"}, - {49, nullptr, "CleanupBluetoothLe"}, - {50, nullptr, "SetLeVisibility"}, - {51, nullptr, "SetLeConnectionParameter"}, - {52, nullptr, "SetLeDefaultConnectionParameter"}, - {53, nullptr, "SetLeAdvertiseData"}, - {54, nullptr, "SetLeAdvertiseParameter"}, - {55, nullptr, "StartLeScan"}, - {56, nullptr, "StopLeScan"}, - {57, nullptr, "AddLeScanFilterCondition"}, - {58, nullptr, "DeleteLeScanFilterCondition"}, - {59, nullptr, "DeleteLeScanFilter"}, - {60, nullptr, "ClearLeScanFilters"}, - {61, nullptr, "EnableLeScanFilter"}, - {62, nullptr, "RegisterLeClient"}, - {63, nullptr, "UnregisterLeClient"}, - {64, nullptr, "UnregisterLeClientAll"}, - {65, nullptr, "LeClientConnect"}, - {66, nullptr, "LeClientCancelConnection"}, - {67, nullptr, "LeClientDisconnect"}, - {68, nullptr, "LeClientGetAttributes"}, - {69, nullptr, "LeClientDiscoverService"}, - {70, nullptr, "LeClientConfigureMtu"}, - {71, nullptr, "RegisterLeServer"}, - {72, nullptr, "UnregisterLeServer"}, - {73, nullptr, "LeServerConnect"}, - {74, nullptr, "LeServerDisconnect"}, - {75, nullptr, "CreateLeService"}, - {76, nullptr, "StartLeService"}, - {77, nullptr, "AddLeCharacteristic"}, - {78, nullptr, "AddLeDescriptor"}, - {79, nullptr, "GetLeCoreEventInfo"}, - {80, nullptr, "LeGetFirstCharacteristic"}, - {81, nullptr, "LeGetNextCharacteristic"}, - {82, nullptr, "LeGetFirstDescriptor"}, - {83, nullptr, "LeGetNextDescriptor"}, - {84, nullptr, "RegisterLeCoreDataPath"}, - {85, nullptr, "UnregisterLeCoreDataPath"}, - {86, nullptr, "RegisterLeHidDataPath"}, - {87, nullptr, "UnregisterLeHidDataPath"}, - {88, nullptr, "RegisterLeDataPath"}, - {89, nullptr, "UnregisterLeDataPath"}, - {90, nullptr, "LeClientReadCharacteristic"}, - {91, nullptr, "LeClientReadDescriptor"}, - {92, nullptr, "LeClientWriteCharacteristic"}, - {93, nullptr, "LeClientWriteDescriptor"}, - {94, nullptr, "LeClientRegisterNotification"}, - {95, nullptr, "LeClientDeregisterNotification"}, + {42, nullptr, "EnableTxPowerBoostSetting"}, + {43, nullptr, "IsTxPowerBoostSettingEnabled"}, + {44, nullptr, "EnableAfhSetting"}, + {45, nullptr, "IsAfhSettingEnabled"}, + {46, nullptr, "InitializeBle"}, + {47, nullptr, "EnableBle"}, + {48, nullptr, "DisableBle"}, + {49, nullptr, "FinalizeBle"}, + {50, nullptr, "SetBleVisibility"}, + {51, nullptr, "SetBleConnectionParameter"}, + {52, nullptr, "SetBleDefaultConnectionParameter"}, + {53, nullptr, "SetBleAdvertiseData"}, + {54, nullptr, "SetBleAdvertiseParameter"}, + {55, nullptr, "StartBleScan"}, + {56, nullptr, "StopBleScan"}, + {57, nullptr, "AddBleScanFilterCondition"}, + {58, nullptr, "DeleteBleScanFilterCondition"}, + {59, nullptr, "DeleteBleScanFilter"}, + {60, nullptr, "ClearBleScanFilters"}, + {61, nullptr, "EnableBleScanFilter"}, + {62, nullptr, "RegisterGattClient"}, + {63, nullptr, "UnregisterGattClient"}, + {64, nullptr, "UnregisterAllGattClients"}, + {65, nullptr, "ConnectGattServer"}, + {66, nullptr, "CancelConnectGattServer"}, + {67, nullptr, "DisconnectGattServer"}, + {68, nullptr, "GetGattAttribute"}, + {69, nullptr, "GetGattService"}, + {70, nullptr, "ConfigureAttMtu"}, + {71, nullptr, "RegisterGattServer"}, + {72, nullptr, "UnregisterGattServer"}, + {73, nullptr, "ConnectGattClient"}, + {74, nullptr, "DisconnectGattClient"}, + {75, nullptr, "AddGattService"}, + {76, nullptr, "EnableGattService"}, + {77, nullptr, "AddGattCharacteristic"}, + {78, nullptr, "AddGattDescriptor"}, + {79, nullptr, "GetBleManagedEventInfo"}, + {80, nullptr, "GetGattFirstCharacteristic"}, + {81, nullptr, "GetGattNextCharacteristic"}, + {82, nullptr, "GetGattFirstDescriptor"}, + {83, nullptr, "GetGattNextDescriptor"}, + {84, nullptr, "RegisterGattManagedDataPath"}, + {85, nullptr, "UnregisterGattManagedDataPath"}, + {86, nullptr, "RegisterGattHidDataPath"}, + {87, nullptr, "UnregisterGattHidDataPath"}, + {88, nullptr, "RegisterGattDataPath"}, + {89, nullptr, "UnregisterGattDataPath"}, + {90, nullptr, "ReadGattCharacteristic"}, + {91, nullptr, "ReadGattDescriptor"}, + {92, nullptr, "WriteGattCharacteristic"}, + {93, nullptr, "WriteGattDescriptor"}, + {94, nullptr, "RegisterGattNotification"}, + {95, nullptr, "UnregisterGattNotification"}, {96, nullptr, "GetLeHidEventInfo"}, {97, nullptr, "RegisterBleHidEvent"}, - {98, nullptr, "SetLeScanParameter"}, - {256, nullptr, "GetIsManufacturingMode"}, + {98, nullptr, "SetBleScanParameter"}, + {99, nullptr, "MoveToSecondaryPiconet"}, + {256, nullptr, "IsManufacturingMode"}, {257, nullptr, "EmulateBluetoothCrash"}, {258, nullptr, "GetBleChannelMap"}, }; diff --git a/src/core/hle/service/btm/btm.cpp b/src/core/hle/service/btm/btm.cpp index 251b3c9df..0d251c6d0 100644 --- a/src/core/hle/service/btm/btm.cpp +++ b/src/core/hle/service/btm/btm.cpp @@ -132,66 +132,71 @@ public: explicit BTM() : ServiceFramework{"btm"} { // clang-format off static const FunctionInfo functions[] = { - {0, nullptr, "Unknown1"}, - {1, nullptr, "Unknown2"}, - {2, nullptr, "RegisterSystemEventForConnectedDeviceCondition"}, - {3, nullptr, "Unknown3"}, - {4, nullptr, "Unknown4"}, - {5, nullptr, "Unknown5"}, - {6, nullptr, "Unknown6"}, - {7, nullptr, "Unknown7"}, - {8, nullptr, "RegisterSystemEventForRegisteredDeviceInfo"}, - {9, nullptr, "Unknown8"}, - {10, nullptr, "Unknown9"}, - {11, nullptr, "Unknown10"}, - {12, nullptr, "Unknown11"}, - {13, nullptr, "Unknown12"}, + {0, nullptr, "GetState"}, + {1, nullptr, "GetHostDeviceProperty"}, + {2, nullptr, "AcquireDeviceConditionEvent"}, + {3, nullptr, "GetDeviceCondition"}, + {4, nullptr, "SetBurstMode"}, + {5, nullptr, "SetSlotMode"}, + {6, nullptr, "SetBluetoothMode"}, + {7, nullptr, "SetWlanMode"}, + {8, nullptr, "AcquireDeviceInfoEvent"}, + {9, nullptr, "GetDeviceInfo"}, + {10, nullptr, "AddDeviceInfo"}, + {11, nullptr, "RemoveDeviceInfo"}, + {12, nullptr, "IncreaseDeviceInfoOrder"}, + {13, nullptr, "LlrNotify"}, {14, nullptr, "EnableRadio"}, {15, nullptr, "DisableRadio"}, - {16, nullptr, "Unknown13"}, - {17, nullptr, "Unknown14"}, - {18, nullptr, "Unknown15"}, - {19, nullptr, "Unknown16"}, - {20, nullptr, "Unknown17"}, - {21, nullptr, "Unknown18"}, - {22, nullptr, "Unknown19"}, - {23, nullptr, "Unknown20"}, - {24, nullptr, "Unknown21"}, - {25, nullptr, "Unknown22"}, - {26, nullptr, "Unknown23"}, - {27, nullptr, "Unknown24"}, - {28, nullptr, "Unknown25"}, - {29, nullptr, "Unknown26"}, - {30, nullptr, "Unknown27"}, - {31, nullptr, "Unknown28"}, - {32, nullptr, "Unknown29"}, - {33, nullptr, "Unknown30"}, - {34, nullptr, "Unknown31"}, - {35, nullptr, "Unknown32"}, - {36, nullptr, "Unknown33"}, - {37, nullptr, "Unknown34"}, - {38, nullptr, "Unknown35"}, - {39, nullptr, "Unknown36"}, - {40, nullptr, "Unknown37"}, - {41, nullptr, "Unknown38"}, - {42, nullptr, "Unknown39"}, - {43, nullptr, "Unknown40"}, - {44, nullptr, "Unknown41"}, - {45, nullptr, "Unknown42"}, - {46, nullptr, "Unknown43"}, - {47, nullptr, "Unknown44"}, - {48, nullptr, "Unknown45"}, - {49, nullptr, "Unknown46"}, - {50, nullptr, "Unknown47"}, - {51, nullptr, "Unknown48"}, - {52, nullptr, "Unknown49"}, - {53, nullptr, "Unknown50"}, - {54, nullptr, "Unknown51"}, - {55, nullptr, "Unknown52"}, - {56, nullptr, "Unknown53"}, - {57, nullptr, "Unknown54"}, - {58, nullptr, "Unknown55"}, - {59, nullptr, "Unknown56"}, + {16, nullptr, "HidDisconnect"}, + {17, nullptr, "HidSetRetransmissionMode"}, + {18, nullptr, "AcquireAwakeReqEvent"}, + {19, nullptr, "AcquireLlrStateEvent"}, + {20, nullptr, "IsLlrStarted"}, + {21, nullptr, "EnableSlotSaving"}, + {22, nullptr, "ProtectDeviceInfo"}, + {23, nullptr, "AcquireBleScanEvent"}, + {24, nullptr, "GetBleScanParameterGeneral"}, + {25, nullptr, "GetBleScanParameterSmartDevice"}, + {26, nullptr, "StartBleScanForGeneral"}, + {27, nullptr, "StopBleScanForGeneral"}, + {28, nullptr, "GetBleScanResultsForGeneral"}, + {29, nullptr, "StartBleScanForPairedDevice"}, + {30, nullptr, "StopBleScanForPairedDevice"}, + {31, nullptr, "StartBleScanForSmartDevice"}, + {32, nullptr, "StopBleScanForSmartDevice"}, + {33, nullptr, "GetBleScanResultsForSmartDevice"}, + {34, nullptr, "AcquireBleConnectionEvent"}, + {35, nullptr, "BleConnect"}, + {36, nullptr, "BleOverrideConnection"}, + {37, nullptr, "BleDisconnect"}, + {38, nullptr, "BleGetConnectionState"}, + {39, nullptr, "BleGetGattClientConditionList"}, + {40, nullptr, "AcquireBlePairingEvent"}, + {41, nullptr, "BlePairDevice"}, + {42, nullptr, "BleUnpairDeviceOnBoth"}, + {43, nullptr, "BleUnpairDevice"}, + {44, nullptr, "BleGetPairedAddresses"}, + {45, nullptr, "AcquireBleServiceDiscoveryEvent"}, + {46, nullptr, "GetGattServices"}, + {47, nullptr, "GetGattService"}, + {48, nullptr, "GetGattIncludedServices"}, + {49, nullptr, "GetBelongingService"}, + {50, nullptr, "GetGattCharacteristics"}, + {51, nullptr, "GetGattDescriptors"}, + {52, nullptr, "AcquireBleMtuConfigEvent"}, + {53, nullptr, "ConfigureBleMtu"}, + {54, nullptr, "GetBleMtu"}, + {55, nullptr, "RegisterBleGattDataPath"}, + {56, nullptr, "UnregisterBleGattDataPath"}, + {57, nullptr, "RegisterAppletResourceUserId"}, + {58, nullptr, "UnregisterAppletResourceUserId"}, + {59, nullptr, "SetAppletResourceUserId"}, + {60, nullptr, "Unknown60"}, + {61, nullptr, "Unknown61"}, + {62, nullptr, "Unknown62"}, + {63, nullptr, "Unknown63"}, + {64, nullptr, "Unknown64"}, }; // clang-format on @@ -204,19 +209,19 @@ public: explicit BTM_DBG() : ServiceFramework{"btm:dbg"} { // clang-format off static const FunctionInfo functions[] = { - {0, nullptr, "RegisterSystemEventForDiscovery"}, - {1, nullptr, "Unknown1"}, - {2, nullptr, "Unknown2"}, - {3, nullptr, "Unknown3"}, - {4, nullptr, "Unknown4"}, - {5, nullptr, "Unknown5"}, - {6, nullptr, "Unknown6"}, - {7, nullptr, "Unknown7"}, - {8, nullptr, "Unknown8"}, - {9, nullptr, "Unknown9"}, - {10, nullptr, "Unknown10"}, - {11, nullptr, "Unknown11"}, - {12, nullptr, "Unknown11"}, + {0, nullptr, "AcquireDiscoveryEvent"}, + {1, nullptr, "StartDiscovery"}, + {2, nullptr, "CancelDiscovery"}, + {3, nullptr, "GetDeviceProperty"}, + {4, nullptr, "CreateBond"}, + {5, nullptr, "CancelBond"}, + {6, nullptr, "SetTsiMode"}, + {7, nullptr, "GeneralTest"}, + {8, nullptr, "HidConnect"}, + {9, nullptr, "GeneralGet"}, + {10, nullptr, "GetGattClientDisconnectionReason"}, + {11, nullptr, "GetBleConnectionParameter"}, + {12, nullptr, "GetBleConnectionParameterRequest"}, }; // clang-format on diff --git a/src/core/hle/service/es/es.cpp b/src/core/hle/service/es/es.cpp index f8e9df4b1..9365f27e1 100644 --- a/src/core/hle/service/es/es.cpp +++ b/src/core/hle/service/es/es.cpp @@ -27,8 +27,8 @@ public: {8, &ETicket::GetTitleKey, "GetTitleKey"}, {9, &ETicket::CountCommonTicket, "CountCommonTicket"}, {10, &ETicket::CountPersonalizedTicket, "CountPersonalizedTicket"}, - {11, &ETicket::ListCommonTicket, "ListCommonTicket"}, - {12, &ETicket::ListPersonalizedTicket, "ListPersonalizedTicket"}, + {11, &ETicket::ListCommonTicketRightsIds, "ListCommonTicketRightsIds"}, + {12, &ETicket::ListPersonalizedTicketRightsIds, "ListPersonalizedTicketRightsIds"}, {13, nullptr, "ListMissingPersonalizedTicket"}, {14, &ETicket::GetCommonTicketSize, "GetCommonTicketSize"}, {15, &ETicket::GetPersonalizedTicketSize, "GetPersonalizedTicketSize"}, @@ -55,7 +55,46 @@ public: {36, nullptr, "DeleteAllInactiveELicenseRequiredPersonalizedTicket"}, {37, nullptr, "OwnTicket2"}, {38, nullptr, "OwnTicket3"}, + {501, nullptr, "Unknown501"}, + {502, nullptr, "Unknown502"}, {503, nullptr, "GetTitleKey"}, + {504, nullptr, "Unknown504"}, + {508, nullptr, "Unknown508"}, + {509, nullptr, "Unknown509"}, + {510, nullptr, "Unknown510"}, + {511, nullptr, "Unknown511"}, + {1001, nullptr, "Unknown1001"}, + {1002, nullptr, "Unknown1001"}, + {1003, nullptr, "Unknown1003"}, + {1004, nullptr, "Unknown1004"}, + {1005, nullptr, "Unknown1005"}, + {1006, nullptr, "Unknown1006"}, + {1007, nullptr, "Unknown1007"}, + {1009, nullptr, "Unknown1009"}, + {1010, nullptr, "Unknown1010"}, + {1011, nullptr, "Unknown1011"}, + {1012, nullptr, "Unknown1012"}, + {1013, nullptr, "Unknown1013"}, + {1014, nullptr, "Unknown1014"}, + {1015, nullptr, "Unknown1015"}, + {1016, nullptr, "Unknown1016"}, + {1017, nullptr, "Unknown1017"}, + {1018, nullptr, "Unknown1018"}, + {1019, nullptr, "Unknown1019"}, + {1020, nullptr, "Unknown1020"}, + {1021, nullptr, "Unknown1021"}, + {1501, nullptr, "Unknown1501"}, + {1502, nullptr, "Unknown1502"}, + {1503, nullptr, "Unknown1503"}, + {1504, nullptr, "Unknown1504"}, + {1505, nullptr, "Unknown1505"}, + {2000, nullptr, "Unknown2000"}, + {2001, nullptr, "Unknown2001"}, + {2100, nullptr, "Unknown2100"}, + {2501, nullptr, "Unknown2501"}, + {2502, nullptr, "Unknown2502"}, + {3001, nullptr, "Unknown3001"}, + {3002, nullptr, "Unknown3002"}, }; // clang-format on RegisterHandlers(functions); @@ -147,7 +186,7 @@ private: rb.Push<u32>(count); } - void ListCommonTicket(Kernel::HLERequestContext& ctx) { + void ListCommonTicketRightsIds(Kernel::HLERequestContext& ctx) { u32 out_entries; if (keys.GetCommonTickets().empty()) out_entries = 0; @@ -170,7 +209,7 @@ private: rb.Push<u32>(out_entries); } - void ListPersonalizedTicket(Kernel::HLERequestContext& ctx) { + void ListPersonalizedTicketRightsIds(Kernel::HLERequestContext& ctx) { u32 out_entries; if (keys.GetPersonalizedTickets().empty()) out_entries = 0; diff --git a/src/core/hle/service/eupld/eupld.cpp b/src/core/hle/service/eupld/eupld.cpp index 2df30acee..0d6d244f4 100644 --- a/src/core/hle/service/eupld/eupld.cpp +++ b/src/core/hle/service/eupld/eupld.cpp @@ -19,6 +19,7 @@ public: {1, nullptr, "ImportCrt"}, {2, nullptr, "ImportPki"}, {3, nullptr, "SetAutoUpload"}, + {4, nullptr, "GetAutoUpload"}, }; // clang-format on diff --git a/src/core/hle/service/hid/hid.cpp b/src/core/hle/service/hid/hid.cpp index 415c2829e..57d5edea7 100644 --- a/src/core/hle/service/hid/hid.cpp +++ b/src/core/hle/service/hid/hid.cpp @@ -230,15 +230,15 @@ Hid::Hid(Core::System& system) : ServiceFramework("hid"), system(system) { {211, nullptr, "IsVibrationDeviceMounted"}, {300, &Hid::ActivateConsoleSixAxisSensor, "ActivateConsoleSixAxisSensor"}, {301, &Hid::StartConsoleSixAxisSensor, "StartConsoleSixAxisSensor"}, - {302, nullptr, "StopConsoleSixAxisSensor"}, - {303, nullptr, "ActivateSevenSixAxisSensor"}, - {304, nullptr, "StartSevenSixAxisSensor"}, + {302, &Hid::StopConsoleSixAxisSensor, "StopConsoleSixAxisSensor"}, + {303, &Hid::ActivateSevenSixAxisSensor, "ActivateSevenSixAxisSensor"}, + {304, &Hid::StartSevenSixAxisSensor, "StartSevenSixAxisSensor"}, {305, &Hid::StopSevenSixAxisSensor, "StopSevenSixAxisSensor"}, {306, &Hid::InitializeSevenSixAxisSensor, "InitializeSevenSixAxisSensor"}, - {307, nullptr, "FinalizeSevenSixAxisSensor"}, + {307, &Hid::FinalizeSevenSixAxisSensor, "FinalizeSevenSixAxisSensor"}, {308, nullptr, "SetSevenSixAxisSensorFusionStrength"}, {309, nullptr, "GetSevenSixAxisSensorFusionStrength"}, - {310, nullptr, "ResetSevenSixAxisSensorTimestamp"}, + {310, &Hid::ResetSevenSixAxisSensorTimestamp, "ResetSevenSixAxisSensorTimestamp"}, {400, nullptr, "IsUsbFullKeyControllerEnabled"}, {401, nullptr, "EnableUsbFullKeyController"}, {402, nullptr, "IsUsbFullKeyControllerConnected"}, @@ -374,6 +374,15 @@ void Hid::ActivateKeyboard(Kernel::HLERequestContext& ctx) { rb.Push(RESULT_SUCCESS); } +void Hid::SendKeyboardLockKeyEvent(Kernel::HLERequestContext& ctx) { + IPC::RequestParser rp{ctx}; + const auto flags{rp.Pop<u32>()}; + LOG_WARNING(Service_HID, "(STUBBED) called. flags={}", flags); + + IPC::ResponseBuilder rb{ctx, 2}; + rb.Push(RESULT_SUCCESS); +} + void Hid::ActivateGesture(Kernel::HLERequestContext& ctx) { IPC::RequestParser rp{ctx}; const auto unknown{rp.Pop<u32>()}; @@ -413,6 +422,18 @@ void Hid::StartSixAxisSensor(Kernel::HLERequestContext& ctx) { rb.Push(RESULT_SUCCESS); } +void Hid::StopSixAxisSensor(Kernel::HLERequestContext& ctx) { + IPC::RequestParser rp{ctx}; + const auto handle{rp.Pop<u32>()}; + const auto applet_resource_user_id{rp.Pop<u64>()}; + + LOG_WARNING(Service_HID, "(STUBBED) called, handle={}, applet_resource_user_id={}", handle, + applet_resource_user_id); + + IPC::ResponseBuilder rb{ctx, 2}; + rb.Push(RESULT_SUCCESS); +} + void Hid::SetGyroscopeZeroDriftMode(Kernel::HLERequestContext& ctx) { IPC::RequestParser rp{ctx}; const auto handle{rp.Pop<u32>()}; @@ -864,33 +885,35 @@ void Hid::StartConsoleSixAxisSensor(Kernel::HLERequestContext& ctx) { rb.Push(RESULT_SUCCESS); } -void Hid::StopSixAxisSensor(Kernel::HLERequestContext& ctx) { +void Hid::StopConsoleSixAxisSensor(Kernel::HLERequestContext& ctx) { IPC::RequestParser rp{ctx}; const auto handle{rp.Pop<u32>()}; + const auto applet_resource_user_id{rp.Pop<u64>()}; - LOG_WARNING(Service_HID, "(STUBBED) called, handle={}", handle); + LOG_WARNING(Service_HID, "(STUBBED) called, handle={}, applet_resource_user_id={}", handle, + applet_resource_user_id); IPC::ResponseBuilder rb{ctx, 2}; rb.Push(RESULT_SUCCESS); } -void Hid::SetIsPalmaAllConnectable(Kernel::HLERequestContext& ctx) { +void Hid::ActivateSevenSixAxisSensor(Kernel::HLERequestContext& ctx) { IPC::RequestParser rp{ctx}; const auto applet_resource_user_id{rp.Pop<u64>()}; - const auto unknown{rp.Pop<u32>()}; - LOG_WARNING(Service_HID, "(STUBBED) called, applet_resource_user_id={}, unknown={}", - applet_resource_user_id, unknown); + LOG_WARNING(Service_HID, "(STUBBED) called, applet_resource_user_id={}", + applet_resource_user_id); IPC::ResponseBuilder rb{ctx, 2}; rb.Push(RESULT_SUCCESS); } -void Hid::SetPalmaBoostMode(Kernel::HLERequestContext& ctx) { +void Hid::StartSevenSixAxisSensor(Kernel::HLERequestContext& ctx) { IPC::RequestParser rp{ctx}; - const auto unknown{rp.Pop<u32>()}; + const auto applet_resource_user_id{rp.Pop<u64>()}; - LOG_WARNING(Service_HID, "(STUBBED) called, unknown={}", unknown); + LOG_WARNING(Service_HID, "(STUBBED) called, applet_resource_user_id={}", + applet_resource_user_id); IPC::ResponseBuilder rb{ctx, 2}; rb.Push(RESULT_SUCCESS); @@ -914,10 +937,46 @@ void Hid::InitializeSevenSixAxisSensor(Kernel::HLERequestContext& ctx) { rb.Push(RESULT_SUCCESS); } -void Hid::SendKeyboardLockKeyEvent(Kernel::HLERequestContext& ctx) { +void Hid::FinalizeSevenSixAxisSensor(Kernel::HLERequestContext& ctx) { IPC::RequestParser rp{ctx}; - const auto flags{rp.Pop<u32>()}; - LOG_WARNING(Service_HID, "(STUBBED) called. flags={}", flags); + const auto applet_resource_user_id{rp.Pop<u64>()}; + + LOG_WARNING(Service_HID, "(STUBBED) called, applet_resource_user_id={}", + applet_resource_user_id); + + IPC::ResponseBuilder rb{ctx, 2}; + rb.Push(RESULT_SUCCESS); +} + +void Hid::ResetSevenSixAxisSensorTimestamp(Kernel::HLERequestContext& ctx) { + IPC::RequestParser rp{ctx}; + const auto applet_resource_user_id{rp.Pop<u64>()}; + + LOG_WARNING(Service_HID, "(STUBBED) called, applet_resource_user_id={}", + applet_resource_user_id); + + IPC::ResponseBuilder rb{ctx, 2}; + rb.Push(RESULT_SUCCESS); +} + +void Hid::SetIsPalmaAllConnectable(Kernel::HLERequestContext& ctx) { + IPC::RequestParser rp{ctx}; + const auto applet_resource_user_id{rp.Pop<u64>()}; + const auto is_palma_all_connectable{rp.Pop<bool>()}; + + LOG_WARNING(Service_HID, + "(STUBBED) called, applet_resource_user_id={}, is_palma_all_connectable={}", + applet_resource_user_id, is_palma_all_connectable); + + IPC::ResponseBuilder rb{ctx, 2}; + rb.Push(RESULT_SUCCESS); +} + +void Hid::SetPalmaBoostMode(Kernel::HLERequestContext& ctx) { + IPC::RequestParser rp{ctx}; + const auto palma_boost_mode{rp.Pop<bool>()}; + + LOG_WARNING(Service_HID, "(STUBBED) called, palma_boost_mode={}", palma_boost_mode); IPC::ResponseBuilder rb{ctx, 2}; rb.Push(RESULT_SUCCESS); diff --git a/src/core/hle/service/hid/hid.h b/src/core/hle/service/hid/hid.h index 41b330fa9..6fb048360 100644 --- a/src/core/hle/service/hid/hid.h +++ b/src/core/hle/service/hid/hid.h @@ -91,9 +91,11 @@ private: void ActivateTouchScreen(Kernel::HLERequestContext& ctx); void ActivateMouse(Kernel::HLERequestContext& ctx); void ActivateKeyboard(Kernel::HLERequestContext& ctx); + void SendKeyboardLockKeyEvent(Kernel::HLERequestContext& ctx); void ActivateGesture(Kernel::HLERequestContext& ctx); void ActivateNpadWithRevision(Kernel::HLERequestContext& ctx); void StartSixAxisSensor(Kernel::HLERequestContext& ctx); + void StopSixAxisSensor(Kernel::HLERequestContext& ctx); void SetGyroscopeZeroDriftMode(Kernel::HLERequestContext& ctx); void GetGyroscopeZeroDriftMode(Kernel::HLERequestContext& ctx); void ResetGyroscopeZeroDriftMode(Kernel::HLERequestContext& ctx); @@ -128,12 +130,15 @@ private: void IsVibrationPermitted(Kernel::HLERequestContext& ctx); void ActivateConsoleSixAxisSensor(Kernel::HLERequestContext& ctx); void StartConsoleSixAxisSensor(Kernel::HLERequestContext& ctx); - void StopSixAxisSensor(Kernel::HLERequestContext& ctx); - void SetIsPalmaAllConnectable(Kernel::HLERequestContext& ctx); - void SetPalmaBoostMode(Kernel::HLERequestContext& ctx); + void StopConsoleSixAxisSensor(Kernel::HLERequestContext& ctx); + void ActivateSevenSixAxisSensor(Kernel::HLERequestContext& ctx); + void StartSevenSixAxisSensor(Kernel::HLERequestContext& ctx); void StopSevenSixAxisSensor(Kernel::HLERequestContext& ctx); void InitializeSevenSixAxisSensor(Kernel::HLERequestContext& ctx); - void SendKeyboardLockKeyEvent(Kernel::HLERequestContext& ctx); + void FinalizeSevenSixAxisSensor(Kernel::HLERequestContext& ctx); + void ResetSevenSixAxisSensorTimestamp(Kernel::HLERequestContext& ctx); + void SetIsPalmaAllConnectable(Kernel::HLERequestContext& ctx); + void SetPalmaBoostMode(Kernel::HLERequestContext& ctx); std::shared_ptr<IAppletResource> applet_resource; Core::System& system; diff --git a/src/video_core/CMakeLists.txt b/src/video_core/CMakeLists.txt index 099bb446e..2dc752aa9 100644 --- a/src/video_core/CMakeLists.txt +++ b/src/video_core/CMakeLists.txt @@ -27,6 +27,8 @@ add_library(video_core STATIC engines/shader_type.h macro/macro.cpp macro/macro.h + macro/macro_hle.cpp + macro/macro_hle.h macro/macro_interpreter.cpp macro/macro_interpreter.h macro/macro_jit_x64.cpp diff --git a/src/video_core/buffer_cache/buffer_cache.h b/src/video_core/buffer_cache/buffer_cache.h index bae1d527c..cf8bdd021 100644 --- a/src/video_core/buffer_cache/buffer_cache.h +++ b/src/video_core/buffer_cache/buffer_cache.h @@ -41,7 +41,11 @@ class BufferCache { static constexpr u64 BLOCK_PAGE_SIZE = 1ULL << BLOCK_PAGE_BITS; public: - using BufferInfo = std::pair<BufferType, u64>; + struct BufferInfo { + BufferType handle; + u64 offset; + u64 address; + }; BufferInfo UploadMemory(GPUVAddr gpu_addr, std::size_t size, std::size_t alignment = 4, bool is_written = false, bool use_fast_cbuf = false) { @@ -50,7 +54,7 @@ public: auto& memory_manager = system.GPU().MemoryManager(); const std::optional<VAddr> cpu_addr_opt = memory_manager.GpuToCpuAddress(gpu_addr); if (!cpu_addr_opt) { - return {GetEmptyBuffer(size), 0}; + return GetEmptyBuffer(size); } const VAddr cpu_addr = *cpu_addr_opt; @@ -88,7 +92,7 @@ public: Buffer* const block = GetBlock(cpu_addr, size); MapInterval* const map = MapAddress(block, gpu_addr, cpu_addr, size); if (!map) { - return {GetEmptyBuffer(size), 0}; + return GetEmptyBuffer(size); } if (is_written) { map->MarkAsModified(true, GetModifiedTicks()); @@ -101,7 +105,7 @@ public: } } - return {block->Handle(), static_cast<u64>(block->Offset(cpu_addr))}; + return BufferInfo{block->Handle(), block->Offset(cpu_addr), block->Address()}; } /// Uploads from a host memory. Returns the OpenGL buffer where it's located and its offset. @@ -254,27 +258,17 @@ public: committed_flushes.pop_front(); } - virtual BufferType GetEmptyBuffer(std::size_t size) = 0; + virtual BufferInfo GetEmptyBuffer(std::size_t size) = 0; protected: explicit BufferCache(VideoCore::RasterizerInterface& rasterizer, Core::System& system, - std::unique_ptr<StreamBuffer> stream_buffer_) - : rasterizer{rasterizer}, system{system}, stream_buffer{std::move(stream_buffer_)}, - stream_buffer_handle{stream_buffer->Handle()} {} + std::unique_ptr<StreamBuffer> stream_buffer) + : rasterizer{rasterizer}, system{system}, stream_buffer{std::move(stream_buffer)} {} ~BufferCache() = default; virtual std::shared_ptr<Buffer> CreateBlock(VAddr cpu_addr, std::size_t size) = 0; - virtual void UploadBlockData(const Buffer& buffer, std::size_t offset, std::size_t size, - const u8* data) = 0; - - virtual void DownloadBlockData(const Buffer& buffer, std::size_t offset, std::size_t size, - u8* data) = 0; - - virtual void CopyBlock(const Buffer& src, const Buffer& dst, std::size_t src_offset, - std::size_t dst_offset, std::size_t size) = 0; - virtual BufferInfo ConstBufferUpload(const void* raw_pointer, std::size_t size) { return {}; } @@ -336,11 +330,11 @@ private: const VAddr cpu_addr_end = cpu_addr + size; if (memory_manager.IsGranularRange(gpu_addr, size)) { u8* host_ptr = memory_manager.GetPointer(gpu_addr); - UploadBlockData(*block, block->Offset(cpu_addr), size, host_ptr); + block->Upload(block->Offset(cpu_addr), size, host_ptr); } else { staging_buffer.resize(size); memory_manager.ReadBlockUnsafe(gpu_addr, staging_buffer.data(), size); - UploadBlockData(*block, block->Offset(cpu_addr), size, staging_buffer.data()); + block->Upload(block->Offset(cpu_addr), size, staging_buffer.data()); } return Register(MapInterval(cpu_addr, cpu_addr_end, gpu_addr)); } @@ -399,7 +393,7 @@ private: } staging_buffer.resize(size); system.Memory().ReadBlockUnsafe(interval.lower(), staging_buffer.data(), size); - UploadBlockData(*block, block->Offset(interval.lower()), size, staging_buffer.data()); + block->Upload(block->Offset(interval.lower()), size, staging_buffer.data()); } } @@ -436,7 +430,7 @@ private: const std::size_t size = map->end - map->start; staging_buffer.resize(size); - DownloadBlockData(*block, block->Offset(map->start), size, staging_buffer.data()); + block->Download(block->Offset(map->start), size, staging_buffer.data()); system.Memory().WriteBlockUnsafe(map->start, staging_buffer.data(), size); map->MarkAsModified(false, 0); } @@ -449,7 +443,7 @@ private: buffer_ptr += size; buffer_offset += size; - return {stream_buffer_handle, uploaded_offset}; + return BufferInfo{stream_buffer->Handle(), uploaded_offset, stream_buffer->Address()}; } void AlignBuffer(std::size_t alignment) { @@ -464,7 +458,7 @@ private: const std::size_t new_size = old_size + BLOCK_PAGE_SIZE; const VAddr cpu_addr = buffer->CpuAddr(); std::shared_ptr<Buffer> new_buffer = CreateBlock(cpu_addr, new_size); - CopyBlock(*buffer, *new_buffer, 0, 0, old_size); + new_buffer->CopyFrom(*buffer, 0, 0, old_size); QueueDestruction(std::move(buffer)); const VAddr cpu_addr_end = cpu_addr + new_size - 1; @@ -486,8 +480,8 @@ private: const std::size_t new_size = size_1 + size_2; std::shared_ptr<Buffer> new_buffer = CreateBlock(new_addr, new_size); - CopyBlock(*first, *new_buffer, 0, new_buffer->Offset(first_addr), size_1); - CopyBlock(*second, *new_buffer, 0, new_buffer->Offset(second_addr), size_2); + new_buffer->CopyFrom(*first, 0, new_buffer->Offset(first_addr), size_1); + new_buffer->CopyFrom(*second, 0, new_buffer->Offset(second_addr), size_2); QueueDestruction(std::move(first)); QueueDestruction(std::move(second)); diff --git a/src/video_core/engines/maxwell_3d.cpp b/src/video_core/engines/maxwell_3d.cpp index ea3c8a963..c01436295 100644 --- a/src/video_core/engines/maxwell_3d.cpp +++ b/src/video_core/engines/maxwell_3d.cpp @@ -128,7 +128,7 @@ void Maxwell3D::CallMacroMethod(u32 method, const std::vector<u32>& parameters) ((method - MacroRegistersStart) >> 1) % static_cast<u32>(macro_positions.size()); // Execute the current macro. - macro_engine->Execute(macro_positions[entry], parameters); + macro_engine->Execute(*this, macro_positions[entry], parameters); if (mme_draw.current_mode != MMEDrawMode::Undefined) { FlushMMEInlineDraw(); } diff --git a/src/video_core/engines/maxwell_3d.h b/src/video_core/engines/maxwell_3d.h index d5fe25065..ef1618990 100644 --- a/src/video_core/engines/maxwell_3d.h +++ b/src/video_core/engines/maxwell_3d.h @@ -1418,6 +1418,14 @@ public: return execute_on; } + VideoCore::RasterizerInterface& GetRasterizer() { + return rasterizer; + } + + const VideoCore::RasterizerInterface& GetRasterizer() const { + return rasterizer; + } + /// Notify a memory write has happened. void OnMemoryWrite() { dirty.flags |= dirty.on_write_stores; diff --git a/src/video_core/engines/shader_bytecode.h b/src/video_core/engines/shader_bytecode.h index e7cb87589..d374b73cf 100644 --- a/src/video_core/engines/shader_bytecode.h +++ b/src/video_core/engines/shader_bytecode.h @@ -661,6 +661,10 @@ union Instruction { constexpr Instruction(u64 value) : value{value} {} constexpr Instruction(const Instruction& instr) : value(instr.value) {} + constexpr bool Bit(u64 offset) const { + return ((value >> offset) & 1) != 0; + } + BitField<0, 8, Register> gpr0; BitField<8, 8, Register> gpr8; union { @@ -1874,7 +1878,9 @@ public: HSETP2_C, HSETP2_R, HSETP2_IMM, + HSET2_C, HSET2_R, + HSET2_IMM, POPC_C, POPC_R, POPC_IMM, @@ -2194,7 +2200,9 @@ private: INST("0111111-1-------", Id::HSETP2_C, Type::HalfSetPredicate, "HSETP2_C"), INST("0101110100100---", Id::HSETP2_R, Type::HalfSetPredicate, "HSETP2_R"), INST("0111111-0-------", Id::HSETP2_IMM, Type::HalfSetPredicate, "HSETP2_IMM"), + INST("0111110-1-------", Id::HSET2_C, Type::HalfSet, "HSET2_C"), INST("0101110100011---", Id::HSET2_R, Type::HalfSet, "HSET2_R"), + INST("0111110-0-------", Id::HSET2_IMM, Type::HalfSet, "HSET2_IMM"), INST("010110111010----", Id::FCMP_RR, Type::Arithmetic, "FCMP_RR"), INST("010010111010----", Id::FCMP_RC, Type::Arithmetic, "FCMP_RC"), INST("0101000010000---", Id::MUFU, Type::Arithmetic, "MUFU"), diff --git a/src/video_core/macro/macro.cpp b/src/video_core/macro/macro.cpp index 89077a2d8..ef7dad349 100644 --- a/src/video_core/macro/macro.cpp +++ b/src/video_core/macro/macro.cpp @@ -2,23 +2,37 @@ // Licensed under GPLv2 or any later version // Refer to the license.txt file included. +#include <boost/container_hash/hash.hpp> #include "common/assert.h" #include "common/logging/log.h" #include "core/settings.h" +#include "video_core/engines/maxwell_3d.h" #include "video_core/macro/macro.h" +#include "video_core/macro/macro_hle.h" #include "video_core/macro/macro_interpreter.h" #include "video_core/macro/macro_jit_x64.h" namespace Tegra { +MacroEngine::MacroEngine(Engines::Maxwell3D& maxwell3d) + : hle_macros{std::make_unique<Tegra::HLEMacro>(maxwell3d)} {} + +MacroEngine::~MacroEngine() = default; + void MacroEngine::AddCode(u32 method, u32 data) { uploaded_macro_code[method].push_back(data); } -void MacroEngine::Execute(u32 method, const std::vector<u32>& parameters) { +void MacroEngine::Execute(Engines::Maxwell3D& maxwell3d, u32 method, + const std::vector<u32>& parameters) { auto compiled_macro = macro_cache.find(method); if (compiled_macro != macro_cache.end()) { - compiled_macro->second->Execute(parameters, method); + const auto& cache_info = compiled_macro->second; + if (cache_info.has_hle_program) { + cache_info.hle_program->Execute(parameters, method); + } else { + cache_info.lle_program->Execute(parameters, method); + } } else { // Macro not compiled, check if it's uploaded and if so, compile it auto macro_code = uploaded_macro_code.find(method); @@ -26,8 +40,21 @@ void MacroEngine::Execute(u32 method, const std::vector<u32>& parameters) { UNREACHABLE_MSG("Macro 0x{0:x} was not uploaded", method); return; } - macro_cache[method] = Compile(macro_code->second); - macro_cache[method]->Execute(parameters, method); + auto& cache_info = macro_cache[method]; + cache_info.hash = boost::hash_value(macro_code->second); + cache_info.lle_program = Compile(macro_code->second); + + auto hle_program = hle_macros->GetHLEProgram(cache_info.hash); + if (hle_program.has_value()) { + cache_info.has_hle_program = true; + cache_info.hle_program = std::move(hle_program.value()); + } + + if (cache_info.has_hle_program) { + cache_info.hle_program->Execute(parameters, method); + } else { + cache_info.lle_program->Execute(parameters, method); + } } } diff --git a/src/video_core/macro/macro.h b/src/video_core/macro/macro.h index b76ed891f..4d00b84b0 100644 --- a/src/video_core/macro/macro.h +++ b/src/video_core/macro/macro.h @@ -11,9 +11,11 @@ #include "common/common_types.h" namespace Tegra { + namespace Engines { class Maxwell3D; } + namespace Macro { constexpr std::size_t NUM_MACRO_REGISTERS = 8; enum class Operation : u32 { @@ -94,6 +96,8 @@ union MethodAddress { } // namespace Macro +class HLEMacro; + class CachedMacro { public: virtual ~CachedMacro() = default; @@ -107,20 +111,29 @@ public: class MacroEngine { public: - virtual ~MacroEngine() = default; + explicit MacroEngine(Engines::Maxwell3D& maxwell3d); + virtual ~MacroEngine(); // Store the uploaded macro code to compile them when they're called. void AddCode(u32 method, u32 data); // Compiles the macro if its not in the cache, and executes the compiled macro - void Execute(u32 method, const std::vector<u32>& parameters); + void Execute(Engines::Maxwell3D& maxwell3d, u32 method, const std::vector<u32>& parameters); protected: virtual std::unique_ptr<CachedMacro> Compile(const std::vector<u32>& code) = 0; private: - std::unordered_map<u32, std::unique_ptr<CachedMacro>> macro_cache; + struct CacheInfo { + std::unique_ptr<CachedMacro> lle_program{}; + std::unique_ptr<CachedMacro> hle_program{}; + u64 hash{}; + bool has_hle_program{}; + }; + + std::unordered_map<u32, CacheInfo> macro_cache; std::unordered_map<u32, std::vector<u32>> uploaded_macro_code; + std::unique_ptr<HLEMacro> hle_macros; }; std::unique_ptr<MacroEngine> GetMacroEngine(Engines::Maxwell3D& maxwell3d); diff --git a/src/video_core/macro/macro_hle.cpp b/src/video_core/macro/macro_hle.cpp new file mode 100644 index 000000000..410f99018 --- /dev/null +++ b/src/video_core/macro/macro_hle.cpp @@ -0,0 +1,113 @@ +// Copyright 2020 yuzu Emulator Project +// Licensed under GPLv2 or any later version +// Refer to the license.txt file included. + +#include <array> +#include <vector> +#include "video_core/engines/maxwell_3d.h" +#include "video_core/macro/macro_hle.h" +#include "video_core/rasterizer_interface.h" + +namespace Tegra { + +namespace { +// HLE'd functions +static void HLE_771BB18C62444DA0(Engines::Maxwell3D& maxwell3d, + const std::vector<u32>& parameters) { + const u32 instance_count = parameters[2] & maxwell3d.GetRegisterValue(0xD1B); + + maxwell3d.regs.draw.topology.Assign( + static_cast<Tegra::Engines::Maxwell3D::Regs::PrimitiveTopology>(parameters[0] & + ~(0x3ffffff << 26))); + maxwell3d.regs.vb_base_instance = parameters[5]; + maxwell3d.mme_draw.instance_count = instance_count; + maxwell3d.regs.vb_element_base = parameters[3]; + maxwell3d.regs.index_array.count = parameters[1]; + maxwell3d.regs.index_array.first = parameters[4]; + + if (maxwell3d.ShouldExecute()) { + maxwell3d.GetRasterizer().Draw(true, true); + } + maxwell3d.regs.index_array.count = 0; + maxwell3d.mme_draw.instance_count = 0; + maxwell3d.mme_draw.current_mode = Engines::Maxwell3D::MMEDrawMode::Undefined; +} + +static void HLE_0D61FC9FAAC9FCAD(Engines::Maxwell3D& maxwell3d, + const std::vector<u32>& parameters) { + const u32 count = (maxwell3d.GetRegisterValue(0xD1B) & parameters[2]); + + maxwell3d.regs.vertex_buffer.first = parameters[3]; + maxwell3d.regs.vertex_buffer.count = parameters[1]; + maxwell3d.regs.vb_base_instance = parameters[4]; + maxwell3d.regs.draw.topology.Assign( + static_cast<Tegra::Engines::Maxwell3D::Regs::PrimitiveTopology>(parameters[0])); + maxwell3d.mme_draw.instance_count = count; + + if (maxwell3d.ShouldExecute()) { + maxwell3d.GetRasterizer().Draw(false, true); + } + maxwell3d.regs.vertex_buffer.count = 0; + maxwell3d.mme_draw.instance_count = 0; + maxwell3d.mme_draw.current_mode = Engines::Maxwell3D::MMEDrawMode::Undefined; +} + +static void HLE_0217920100488FF7(Engines::Maxwell3D& maxwell3d, + const std::vector<u32>& parameters) { + const u32 instance_count = (maxwell3d.GetRegisterValue(0xD1B) & parameters[2]); + const u32 element_base = parameters[4]; + const u32 base_instance = parameters[5]; + maxwell3d.regs.index_array.first = parameters[3]; + maxwell3d.regs.reg_array[0x446] = element_base; // vertex id base? + maxwell3d.regs.index_array.count = parameters[1]; + maxwell3d.regs.vb_element_base = element_base; + maxwell3d.regs.vb_base_instance = base_instance; + maxwell3d.mme_draw.instance_count = instance_count; + maxwell3d.CallMethodFromMME(0x8e3, 0x640); + maxwell3d.CallMethodFromMME(0x8e4, element_base); + maxwell3d.CallMethodFromMME(0x8e5, base_instance); + maxwell3d.regs.draw.topology.Assign( + static_cast<Tegra::Engines::Maxwell3D::Regs::PrimitiveTopology>(parameters[0])); + if (maxwell3d.ShouldExecute()) { + maxwell3d.GetRasterizer().Draw(true, true); + } + maxwell3d.regs.reg_array[0x446] = 0x0; // vertex id base? + maxwell3d.regs.index_array.count = 0; + maxwell3d.regs.vb_element_base = 0x0; + maxwell3d.regs.vb_base_instance = 0x0; + maxwell3d.mme_draw.instance_count = 0; + maxwell3d.CallMethodFromMME(0x8e3, 0x640); + maxwell3d.CallMethodFromMME(0x8e4, 0x0); + maxwell3d.CallMethodFromMME(0x8e5, 0x0); + maxwell3d.mme_draw.current_mode = Engines::Maxwell3D::MMEDrawMode::Undefined; +} +} // namespace + +constexpr std::array<std::pair<u64, HLEFunction>, 3> hle_funcs{{ + std::make_pair<u64, HLEFunction>(0x771BB18C62444DA0, &HLE_771BB18C62444DA0), + std::make_pair<u64, HLEFunction>(0x0D61FC9FAAC9FCAD, &HLE_0D61FC9FAAC9FCAD), + std::make_pair<u64, HLEFunction>(0x0217920100488FF7, &HLE_0217920100488FF7), +}}; + +HLEMacro::HLEMacro(Engines::Maxwell3D& maxwell3d) : maxwell3d(maxwell3d) {} +HLEMacro::~HLEMacro() = default; + +std::optional<std::unique_ptr<CachedMacro>> HLEMacro::GetHLEProgram(u64 hash) const { + const auto it = std::find_if(hle_funcs.cbegin(), hle_funcs.cend(), + [hash](const auto& pair) { return pair.first == hash; }); + if (it == hle_funcs.end()) { + return std::nullopt; + } + return std::make_unique<HLEMacroImpl>(maxwell3d, it->second); +} + +HLEMacroImpl::~HLEMacroImpl() = default; + +HLEMacroImpl::HLEMacroImpl(Engines::Maxwell3D& maxwell3d, HLEFunction func) + : maxwell3d(maxwell3d), func(func) {} + +void HLEMacroImpl::Execute(const std::vector<u32>& parameters, u32 method) { + func(maxwell3d, parameters); +} + +} // namespace Tegra diff --git a/src/video_core/macro/macro_hle.h b/src/video_core/macro/macro_hle.h new file mode 100644 index 000000000..37af875a0 --- /dev/null +++ b/src/video_core/macro/macro_hle.h @@ -0,0 +1,44 @@ +// Copyright 2020 yuzu Emulator Project +// Licensed under GPLv2 or any later version +// Refer to the license.txt file included. + +#pragma once + +#include <memory> +#include <optional> +#include <vector> +#include "common/common_types.h" +#include "video_core/macro/macro.h" + +namespace Tegra { + +namespace Engines { +class Maxwell3D; +} + +using HLEFunction = void (*)(Engines::Maxwell3D& maxwell3d, const std::vector<u32>& parameters); + +class HLEMacro { +public: + explicit HLEMacro(Engines::Maxwell3D& maxwell3d); + ~HLEMacro(); + + std::optional<std::unique_ptr<CachedMacro>> GetHLEProgram(u64 hash) const; + +private: + Engines::Maxwell3D& maxwell3d; +}; + +class HLEMacroImpl : public CachedMacro { +public: + explicit HLEMacroImpl(Engines::Maxwell3D& maxwell3d, HLEFunction func); + ~HLEMacroImpl(); + + void Execute(const std::vector<u32>& parameters, u32 method) override; + +private: + Engines::Maxwell3D& maxwell3d; + HLEFunction func; +}; + +} // namespace Tegra diff --git a/src/video_core/macro/macro_interpreter.cpp b/src/video_core/macro/macro_interpreter.cpp index 5edff27aa..aa5256419 100644 --- a/src/video_core/macro/macro_interpreter.cpp +++ b/src/video_core/macro/macro_interpreter.cpp @@ -11,7 +11,8 @@ MICROPROFILE_DEFINE(MacroInterp, "GPU", "Execute macro interpreter", MP_RGB(128, 128, 192)); namespace Tegra { -MacroInterpreter::MacroInterpreter(Engines::Maxwell3D& maxwell3d) : maxwell3d(maxwell3d) {} +MacroInterpreter::MacroInterpreter(Engines::Maxwell3D& maxwell3d) + : MacroEngine::MacroEngine(maxwell3d), maxwell3d(maxwell3d) {} std::unique_ptr<CachedMacro> MacroInterpreter::Compile(const std::vector<u32>& code) { return std::make_unique<MacroInterpreterImpl>(maxwell3d, code); diff --git a/src/video_core/macro/macro_jit_x64.cpp b/src/video_core/macro/macro_jit_x64.cpp index 30abb66e5..07292702f 100644 --- a/src/video_core/macro/macro_jit_x64.cpp +++ b/src/video_core/macro/macro_jit_x64.cpp @@ -28,7 +28,8 @@ static const std::bitset<32> PERSISTENT_REGISTERS = Common::X64::BuildRegSet({ BRANCH_HOLDER, }); -MacroJITx64::MacroJITx64(Engines::Maxwell3D& maxwell3d) : maxwell3d(maxwell3d) {} +MacroJITx64::MacroJITx64(Engines::Maxwell3D& maxwell3d) + : MacroEngine::MacroEngine(maxwell3d), maxwell3d(maxwell3d) {} std::unique_ptr<CachedMacro> MacroJITx64::Compile(const std::vector<u32>& code) { return std::make_unique<MacroJITx64Impl>(maxwell3d, code); diff --git a/src/video_core/renderer_opengl/gl_buffer_cache.cpp b/src/video_core/renderer_opengl/gl_buffer_cache.cpp index ad0577a4f..d9f7b4cc6 100644 --- a/src/video_core/renderer_opengl/gl_buffer_cache.cpp +++ b/src/video_core/renderer_opengl/gl_buffer_cache.cpp @@ -22,21 +22,46 @@ using Maxwell = Tegra::Engines::Maxwell3D::Regs; MICROPROFILE_DEFINE(OpenGL_Buffer_Download, "OpenGL", "Buffer Download", MP_RGB(192, 192, 128)); -Buffer::Buffer(VAddr cpu_addr, const std::size_t size) : VideoCommon::BufferBlock{cpu_addr, size} { +Buffer::Buffer(const Device& device, VAddr cpu_addr, std::size_t size) + : VideoCommon::BufferBlock{cpu_addr, size} { gl_buffer.Create(); glNamedBufferData(gl_buffer.handle, static_cast<GLsizeiptr>(size), nullptr, GL_DYNAMIC_DRAW); + if (device.HasVertexBufferUnifiedMemory()) { + glMakeNamedBufferResidentNV(gl_buffer.handle, GL_READ_WRITE); + glGetNamedBufferParameterui64vNV(gl_buffer.handle, GL_BUFFER_GPU_ADDRESS_NV, &gpu_address); + } } Buffer::~Buffer() = default; +void Buffer::Upload(std::size_t offset, std::size_t size, const u8* data) const { + glNamedBufferSubData(Handle(), static_cast<GLintptr>(offset), static_cast<GLsizeiptr>(size), + data); +} + +void Buffer::Download(std::size_t offset, std::size_t size, u8* data) const { + MICROPROFILE_SCOPE(OpenGL_Buffer_Download); + glMemoryBarrier(GL_BUFFER_UPDATE_BARRIER_BIT); + glGetNamedBufferSubData(Handle(), static_cast<GLintptr>(offset), static_cast<GLsizeiptr>(size), + data); +} + +void Buffer::CopyFrom(const Buffer& src, std::size_t src_offset, std::size_t dst_offset, + std::size_t size) const { + glCopyNamedBufferSubData(src.Handle(), Handle(), static_cast<GLintptr>(src_offset), + static_cast<GLintptr>(dst_offset), static_cast<GLsizeiptr>(size)); +} + OGLBufferCache::OGLBufferCache(RasterizerOpenGL& rasterizer, Core::System& system, - const Device& device, std::size_t stream_size) - : GenericBufferCache{rasterizer, system, std::make_unique<OGLStreamBuffer>(stream_size, true)} { + const Device& device_, std::size_t stream_size) + : GenericBufferCache{rasterizer, system, + std::make_unique<OGLStreamBuffer>(device_, stream_size, true)}, + device{device_} { if (!device.HasFastBufferSubData()) { return; } - static constexpr auto size = static_cast<GLsizeiptr>(Maxwell::MaxConstBufferSize); + static constexpr GLsizeiptr size = static_cast<GLsizeiptr>(Maxwell::MaxConstBufferSize); glCreateBuffers(static_cast<GLsizei>(std::size(cbufs)), std::data(cbufs)); for (const GLuint cbuf : cbufs) { glNamedBufferData(cbuf, size, nullptr, GL_STREAM_DRAW); @@ -48,39 +73,20 @@ OGLBufferCache::~OGLBufferCache() { } std::shared_ptr<Buffer> OGLBufferCache::CreateBlock(VAddr cpu_addr, std::size_t size) { - return std::make_shared<Buffer>(cpu_addr, size); + return std::make_shared<Buffer>(device, cpu_addr, size); } -GLuint OGLBufferCache::GetEmptyBuffer(std::size_t) { - return 0; -} - -void OGLBufferCache::UploadBlockData(const Buffer& buffer, std::size_t offset, std::size_t size, - const u8* data) { - glNamedBufferSubData(buffer.Handle(), static_cast<GLintptr>(offset), - static_cast<GLsizeiptr>(size), data); -} - -void OGLBufferCache::DownloadBlockData(const Buffer& buffer, std::size_t offset, std::size_t size, - u8* data) { - MICROPROFILE_SCOPE(OpenGL_Buffer_Download); - glMemoryBarrier(GL_BUFFER_UPDATE_BARRIER_BIT); - glGetNamedBufferSubData(buffer.Handle(), static_cast<GLintptr>(offset), - static_cast<GLsizeiptr>(size), data); -} - -void OGLBufferCache::CopyBlock(const Buffer& src, const Buffer& dst, std::size_t src_offset, - std::size_t dst_offset, std::size_t size) { - glCopyNamedBufferSubData(src.Handle(), dst.Handle(), static_cast<GLintptr>(src_offset), - static_cast<GLintptr>(dst_offset), static_cast<GLsizeiptr>(size)); +OGLBufferCache::BufferInfo OGLBufferCache::GetEmptyBuffer(std::size_t) { + return {0, 0, 0}; } OGLBufferCache::BufferInfo OGLBufferCache::ConstBufferUpload(const void* raw_pointer, std::size_t size) { DEBUG_ASSERT(cbuf_cursor < std::size(cbufs)); const GLuint cbuf = cbufs[cbuf_cursor++]; + glNamedBufferSubData(cbuf, 0, static_cast<GLsizeiptr>(size), raw_pointer); - return {cbuf, 0}; + return {cbuf, 0, 0}; } } // namespace OpenGL diff --git a/src/video_core/renderer_opengl/gl_buffer_cache.h b/src/video_core/renderer_opengl/gl_buffer_cache.h index a49aaf9c4..59d95adbc 100644 --- a/src/video_core/renderer_opengl/gl_buffer_cache.h +++ b/src/video_core/renderer_opengl/gl_buffer_cache.h @@ -25,15 +25,27 @@ class RasterizerOpenGL; class Buffer : public VideoCommon::BufferBlock { public: - explicit Buffer(VAddr cpu_addr, const std::size_t size); + explicit Buffer(const Device& device, VAddr cpu_addr, std::size_t size); ~Buffer(); - GLuint Handle() const { + void Upload(std::size_t offset, std::size_t size, const u8* data) const; + + void Download(std::size_t offset, std::size_t size, u8* data) const; + + void CopyFrom(const Buffer& src, std::size_t src_offset, std::size_t dst_offset, + std::size_t size) const; + + GLuint Handle() const noexcept { return gl_buffer.handle; } + u64 Address() const noexcept { + return gpu_address; + } + private: OGLBuffer gl_buffer; + u64 gpu_address = 0; }; using GenericBufferCache = VideoCommon::BufferCache<Buffer, GLuint, OGLStreamBuffer>; @@ -43,7 +55,7 @@ public: const Device& device, std::size_t stream_size); ~OGLBufferCache(); - GLuint GetEmptyBuffer(std::size_t) override; + BufferInfo GetEmptyBuffer(std::size_t) override; void Acquire() noexcept { cbuf_cursor = 0; @@ -52,22 +64,16 @@ public: protected: std::shared_ptr<Buffer> CreateBlock(VAddr cpu_addr, std::size_t size) override; - void UploadBlockData(const Buffer& buffer, std::size_t offset, std::size_t size, - const u8* data) override; - - void DownloadBlockData(const Buffer& buffer, std::size_t offset, std::size_t size, - u8* data) override; - - void CopyBlock(const Buffer& src, const Buffer& dst, std::size_t src_offset, - std::size_t dst_offset, std::size_t size) override; - BufferInfo ConstBufferUpload(const void* raw_pointer, std::size_t size) override; private: + static constexpr std::size_t NUM_CBUFS = Tegra::Engines::Maxwell3D::Regs::MaxConstBuffers * + Tegra::Engines::Maxwell3D::Regs::MaxShaderProgram; + + const Device& device; + std::size_t cbuf_cursor = 0; - std::array<GLuint, Tegra::Engines::Maxwell3D::Regs::MaxConstBuffers * - Tegra::Engines::Maxwell3D::Regs::MaxShaderProgram> - cbufs; + std::array<GLuint, NUM_CBUFS> cbufs{}; }; } // namespace OpenGL diff --git a/src/video_core/renderer_opengl/gl_device.cpp b/src/video_core/renderer_opengl/gl_device.cpp index 1011c7738..b6b6659c1 100644 --- a/src/video_core/renderer_opengl/gl_device.cpp +++ b/src/video_core/renderer_opengl/gl_device.cpp @@ -178,7 +178,7 @@ bool IsASTCSupported() { for (const GLenum format : formats) { for (const GLenum support : required_support) { GLint value; - glGetInternalformativ(GL_TEXTURE_2D, format, support, 1, &value); + glGetInternalformativ(target, format, support, 1, &value); if (value != GL_FULL_SUPPORT) { return false; } @@ -188,16 +188,32 @@ bool IsASTCSupported() { return true; } +/// @brief Returns true when a GL_RENDERER is a Turing GPU +/// @param renderer GL_RENDERER string +bool IsTuring(std::string_view renderer) { + static constexpr std::array<std::string_view, 12> TURING_GPUS = { + "GTX 1650", "GTX 1660", "RTX 2060", "RTX 2070", + "RTX 2080", "TITAN RTX", "Quadro RTX 3000", "Quadro RTX 4000", + "Quadro RTX 5000", "Quadro RTX 6000", "Quadro RTX 8000", "Tesla T4", + }; + return std::any_of(TURING_GPUS.begin(), TURING_GPUS.end(), + [renderer](std::string_view candidate) { + return renderer.find(candidate) != std::string_view::npos; + }); +} + } // Anonymous namespace Device::Device() : max_uniform_buffers{BuildMaxUniformBuffers()}, base_bindings{BuildBaseBindings()} { const std::string_view vendor = reinterpret_cast<const char*>(glGetString(GL_VENDOR)); + const std::string_view renderer = reinterpret_cast<const char*>(glGetString(GL_RENDERER)); const std::string_view version = reinterpret_cast<const char*>(glGetString(GL_VERSION)); const std::vector extensions = GetExtensions(); const bool is_nvidia = vendor == "NVIDIA Corporation"; const bool is_amd = vendor == "ATI Technologies Inc."; + const bool is_turing = is_nvidia && IsTuring(renderer); bool disable_fast_buffer_sub_data = false; if (is_nvidia && version == "4.6.0 NVIDIA 443.24") { @@ -221,8 +237,16 @@ Device::Device() has_variable_aoffi = TestVariableAoffi(); has_component_indexing_bug = is_amd; has_precise_bug = TestPreciseBug(); - has_fast_buffer_sub_data = is_nvidia && !disable_fast_buffer_sub_data; has_nv_viewport_array2 = GLAD_GL_NV_viewport_array2; + + // At the moment of writing this, only Nvidia's driver optimizes BufferSubData on exclusive + // uniform buffers as "push constants" + has_fast_buffer_sub_data = is_nvidia && !disable_fast_buffer_sub_data; + + // Nvidia's driver on Turing GPUs randomly crashes when the buffer is made resident, or on + // DeleteBuffers. Disable unified memory on these devices. + has_vertex_buffer_unified_memory = GLAD_GL_NV_vertex_buffer_unified_memory && !is_turing; + use_assembly_shaders = Settings::values.use_assembly_shaders && GLAD_GL_NV_gpu_program5 && GLAD_GL_NV_compute_program5 && GLAD_GL_NV_transform_feedback && GLAD_GL_NV_transform_feedback2; diff --git a/src/video_core/renderer_opengl/gl_device.h b/src/video_core/renderer_opengl/gl_device.h index c86e709b1..e1d811966 100644 --- a/src/video_core/renderer_opengl/gl_device.h +++ b/src/video_core/renderer_opengl/gl_device.h @@ -72,6 +72,10 @@ public: return has_texture_shadow_lod; } + bool HasVertexBufferUnifiedMemory() const { + return has_vertex_buffer_unified_memory; + } + bool HasASTC() const { return has_astc; } @@ -115,6 +119,7 @@ private: bool has_vertex_viewport_layer{}; bool has_image_load_formatted{}; bool has_texture_shadow_lod{}; + bool has_vertex_buffer_unified_memory{}; bool has_astc{}; bool has_variable_aoffi{}; bool has_component_indexing_bug{}; diff --git a/src/video_core/renderer_opengl/gl_rasterizer.cpp b/src/video_core/renderer_opengl/gl_rasterizer.cpp index 2d6c11320..362457ffe 100644 --- a/src/video_core/renderer_opengl/gl_rasterizer.cpp +++ b/src/video_core/renderer_opengl/gl_rasterizer.cpp @@ -61,7 +61,8 @@ constexpr std::size_t NUM_CONST_BUFFERS_BYTES_PER_STAGE = constexpr std::size_t TOTAL_CONST_BUFFER_BYTES = NUM_CONST_BUFFERS_BYTES_PER_STAGE * Maxwell::MaxShaderStage; -constexpr std::size_t NumSupportedVertexAttributes = 16; +constexpr std::size_t NUM_SUPPORTED_VERTEX_ATTRIBUTES = 16; +constexpr std::size_t NUM_SUPPORTED_VERTEX_BINDINGS = 16; template <typename Engine, typename Entry> Tegra::Texture::FullTextureInfo GetTextureInfo(const Engine& engine, const Entry& entry, @@ -193,7 +194,7 @@ void RasterizerOpenGL::SetupVertexFormat() { // avoid OpenGL errors. // TODO(Subv): Analyze the shader to identify which attributes are actually used and don't // assume every shader uses them all. - for (std::size_t index = 0; index < NumSupportedVertexAttributes; ++index) { + for (std::size_t index = 0; index < NUM_SUPPORTED_VERTEX_ATTRIBUTES; ++index) { if (!flags[Dirty::VertexFormat0 + index]) { continue; } @@ -231,9 +232,11 @@ void RasterizerOpenGL::SetupVertexBuffer() { MICROPROFILE_SCOPE(OpenGL_VB); + const bool use_unified_memory = device.HasVertexBufferUnifiedMemory(); + // Upload all guest vertex arrays sequentially to our buffer const auto& regs = gpu.regs; - for (std::size_t index = 0; index < Maxwell::NumVertexArrays; ++index) { + for (std::size_t index = 0; index < NUM_SUPPORTED_VERTEX_BINDINGS; ++index) { if (!flags[Dirty::VertexBuffer0 + index]) { continue; } @@ -246,16 +249,25 @@ void RasterizerOpenGL::SetupVertexBuffer() { const GPUVAddr start = vertex_array.StartAddress(); const GPUVAddr end = regs.vertex_array_limit[index].LimitAddress(); - ASSERT(end >= start); + + const GLuint gl_index = static_cast<GLuint>(index); const u64 size = end - start; if (size == 0) { - glBindVertexBuffer(static_cast<GLuint>(index), 0, 0, vertex_array.stride); + glBindVertexBuffer(gl_index, 0, 0, vertex_array.stride); + if (use_unified_memory) { + glBufferAddressRangeNV(GL_VERTEX_ATTRIB_ARRAY_ADDRESS_NV, gl_index, 0, 0); + } continue; } - const auto [vertex_buffer, vertex_buffer_offset] = buffer_cache.UploadMemory(start, size); - glBindVertexBuffer(static_cast<GLuint>(index), vertex_buffer, vertex_buffer_offset, - vertex_array.stride); + const auto info = buffer_cache.UploadMemory(start, size); + if (use_unified_memory) { + glBindVertexBuffer(gl_index, 0, 0, vertex_array.stride); + glBufferAddressRangeNV(GL_VERTEX_ATTRIB_ARRAY_ADDRESS_NV, gl_index, + info.address + info.offset, size); + } else { + glBindVertexBuffer(gl_index, info.handle, info.offset, vertex_array.stride); + } } } @@ -268,7 +280,7 @@ void RasterizerOpenGL::SetupVertexInstances() { flags[Dirty::VertexInstances] = false; const auto& regs = gpu.regs; - for (std::size_t index = 0; index < NumSupportedVertexAttributes; ++index) { + for (std::size_t index = 0; index < NUM_SUPPORTED_VERTEX_ATTRIBUTES; ++index) { if (!flags[Dirty::VertexInstance0 + index]) { continue; } @@ -285,9 +297,9 @@ GLintptr RasterizerOpenGL::SetupIndexBuffer() { MICROPROFILE_SCOPE(OpenGL_Index); const auto& regs = system.GPU().Maxwell3D().regs; const std::size_t size = CalculateIndexBufferSize(); - const auto [buffer, offset] = buffer_cache.UploadMemory(regs.index_array.IndexStart(), size); - glBindBuffer(GL_ELEMENT_ARRAY_BUFFER, buffer); - return offset; + const auto info = buffer_cache.UploadMemory(regs.index_array.IndexStart(), size); + glBindBuffer(GL_ELEMENT_ARRAY_BUFFER, info.handle); + return info.offset; } void RasterizerOpenGL::SetupShaders(GLenum primitive_mode) { @@ -643,9 +655,9 @@ void RasterizerOpenGL::Draw(bool is_indexed, bool is_instanced) { if (!device.UseAssemblyShaders()) { MaxwellUniformData ubo; ubo.SetFromRegs(gpu); - const auto [buffer, offset] = + const auto info = buffer_cache.UploadHostMemory(&ubo, sizeof(ubo), device.GetUniformBufferAlignment()); - glBindBufferRange(GL_UNIFORM_BUFFER, EmulationUniformBlockBinding, buffer, offset, + glBindBufferRange(GL_UNIFORM_BUFFER, EmulationUniformBlockBinding, info.handle, info.offset, static_cast<GLsizeiptr>(sizeof(ubo))); } @@ -956,8 +968,7 @@ void RasterizerOpenGL::SetupConstBuffer(GLenum stage, u32 binding, if (device.UseAssemblyShaders()) { glBindBufferRangeNV(stage, entry.GetIndex(), 0, 0, 0); } else { - glBindBufferRange(GL_UNIFORM_BUFFER, binding, - buffer_cache.GetEmptyBuffer(sizeof(float)), 0, sizeof(float)); + glBindBufferRange(GL_UNIFORM_BUFFER, binding, 0, 0, sizeof(float)); } return; } @@ -970,24 +981,25 @@ void RasterizerOpenGL::SetupConstBuffer(GLenum stage, u32 binding, const std::size_t alignment = use_unified ? 4 : device.GetUniformBufferAlignment(); const GPUVAddr gpu_addr = buffer.address; - auto [cbuf, offset] = buffer_cache.UploadMemory(gpu_addr, size, alignment, false, fast_upload); + auto info = buffer_cache.UploadMemory(gpu_addr, size, alignment, false, fast_upload); if (device.UseAssemblyShaders()) { UNIMPLEMENTED_IF(use_unified); - if (offset != 0) { + if (info.offset != 0) { const GLuint staging_cbuf = staging_cbufs[current_cbuf++]; - glCopyNamedBufferSubData(cbuf, staging_cbuf, offset, 0, size); - cbuf = staging_cbuf; - offset = 0; + glCopyNamedBufferSubData(info.handle, staging_cbuf, info.offset, 0, size); + info.handle = staging_cbuf; + info.offset = 0; } - glBindBufferRangeNV(stage, binding, cbuf, offset, size); + glBindBufferRangeNV(stage, binding, info.handle, info.offset, size); return; } if (use_unified) { - glCopyNamedBufferSubData(cbuf, unified_uniform_buffer.handle, offset, unified_offset, size); + glCopyNamedBufferSubData(info.handle, unified_uniform_buffer.handle, info.offset, + unified_offset, size); } else { - glBindBufferRange(GL_UNIFORM_BUFFER, binding, cbuf, offset, size); + glBindBufferRange(GL_UNIFORM_BUFFER, binding, info.handle, info.offset, size); } } @@ -1023,9 +1035,8 @@ void RasterizerOpenGL::SetupComputeGlobalMemory(Shader* kernel) { void RasterizerOpenGL::SetupGlobalMemory(u32 binding, const GlobalMemoryEntry& entry, 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.is_written); - glBindBufferRange(GL_SHADER_STORAGE_BUFFER, binding, ssbo, buffer_offset, + const auto info = buffer_cache.UploadMemory(gpu_addr, size, alignment, entry.is_written); + glBindBufferRange(GL_SHADER_STORAGE_BUFFER, binding, info.handle, info.offset, static_cast<GLsizeiptr>(size)); } @@ -1712,8 +1723,9 @@ void RasterizerOpenGL::EndTransformFeedback() { const GLuint handle = transform_feedback_buffers[index].handle; const GPUVAddr gpu_addr = binding.Address(); const std::size_t size = binding.buffer_size; - const auto [dest_buffer, offset] = buffer_cache.UploadMemory(gpu_addr, size, 4, true); - glCopyNamedBufferSubData(handle, dest_buffer, 0, offset, static_cast<GLsizeiptr>(size)); + const auto info = buffer_cache.UploadMemory(gpu_addr, size, 4, true); + glCopyNamedBufferSubData(handle, info.handle, 0, info.offset, + static_cast<GLsizeiptr>(size)); } } diff --git a/src/video_core/renderer_opengl/gl_shader_cache.cpp b/src/video_core/renderer_opengl/gl_shader_cache.cpp index 46e780a06..c6a3bf3a1 100644 --- a/src/video_core/renderer_opengl/gl_shader_cache.cpp +++ b/src/video_core/renderer_opengl/gl_shader_cache.cpp @@ -460,8 +460,9 @@ Shader* ShaderCacheOpenGL::GetStageProgram(Maxwell::ShaderProgram program) { const u8* host_ptr_b = memory_manager.GetPointer(address_b); code_b = GetShaderCode(memory_manager, address_b, host_ptr_b, false); } + const std::size_t code_size = code.size() * sizeof(u64); - const auto unique_identifier = GetUniqueIdentifier( + const u64 unique_identifier = GetUniqueIdentifier( GetShaderType(program), program == Maxwell::ShaderProgram::VertexA, code, code_b); const ShaderParameters params{system, disk_cache, device, @@ -477,7 +478,7 @@ Shader* ShaderCacheOpenGL::GetStageProgram(Maxwell::ShaderProgram program) { Shader* const result = shader.get(); if (cpu_addr) { - Register(std::move(shader), *cpu_addr, code.size() * sizeof(u64)); + Register(std::move(shader), *cpu_addr, code_size); } else { null_shader = std::move(shader); } @@ -495,8 +496,9 @@ Shader* ShaderCacheOpenGL::GetComputeKernel(GPUVAddr code_addr) { const auto host_ptr{memory_manager.GetPointer(code_addr)}; // No kernel found, create a new one - auto code{GetShaderCode(memory_manager, code_addr, host_ptr, true)}; - const auto unique_identifier{GetUniqueIdentifier(ShaderType::Compute, false, code)}; + ProgramCode code{GetShaderCode(memory_manager, code_addr, host_ptr, true)}; + const std::size_t code_size{code.size() * sizeof(u64)}; + const u64 unique_identifier{GetUniqueIdentifier(ShaderType::Compute, false, code)}; const ShaderParameters params{system, disk_cache, device, *cpu_addr, host_ptr, unique_identifier}; @@ -511,7 +513,7 @@ Shader* ShaderCacheOpenGL::GetComputeKernel(GPUVAddr code_addr) { Shader* const result = kernel.get(); if (cpu_addr) { - Register(std::move(kernel), *cpu_addr, code.size() * sizeof(u64)); + Register(std::move(kernel), *cpu_addr, code_size); } else { null_kernel = std::move(kernel); } diff --git a/src/video_core/renderer_opengl/gl_shader_cache.h b/src/video_core/renderer_opengl/gl_shader_cache.h index 6848f1388..994aaeaf2 100644 --- a/src/video_core/renderer_opengl/gl_shader_cache.h +++ b/src/video_core/renderer_opengl/gl_shader_cache.h @@ -37,7 +37,6 @@ namespace OpenGL { class Device; class RasterizerOpenGL; -struct UnspecializedShader; using Maxwell = Tegra::Engines::Maxwell3D::Regs; diff --git a/src/video_core/renderer_opengl/gl_stream_buffer.cpp b/src/video_core/renderer_opengl/gl_stream_buffer.cpp index 932a2f69e..3655ff629 100644 --- a/src/video_core/renderer_opengl/gl_stream_buffer.cpp +++ b/src/video_core/renderer_opengl/gl_stream_buffer.cpp @@ -2,11 +2,13 @@ // Licensed under GPLv2 or any later version // Refer to the license.txt file included. -#include <deque> +#include <tuple> #include <vector> + #include "common/alignment.h" #include "common/assert.h" #include "common/microprofile.h" +#include "video_core/renderer_opengl/gl_device.h" #include "video_core/renderer_opengl/gl_stream_buffer.h" MICROPROFILE_DEFINE(OpenGL_StreamBuffer, "OpenGL", "Stream Buffer Orphaning", @@ -14,8 +16,7 @@ MICROPROFILE_DEFINE(OpenGL_StreamBuffer, "OpenGL", "Stream Buffer Orphaning", namespace OpenGL { -OGLStreamBuffer::OGLStreamBuffer(GLsizeiptr size, bool vertex_data_usage, bool prefer_coherent, - bool use_persistent) +OGLStreamBuffer::OGLStreamBuffer(const Device& device, GLsizeiptr size, bool vertex_data_usage) : buffer_size(size) { gl_buffer.Create(); @@ -29,23 +30,19 @@ OGLStreamBuffer::OGLStreamBuffer(GLsizeiptr size, bool vertex_data_usage, bool p allocate_size *= 2; } - if (use_persistent) { - persistent = true; - coherent = prefer_coherent; - const GLbitfield flags = - GL_MAP_WRITE_BIT | GL_MAP_PERSISTENT_BIT | (coherent ? GL_MAP_COHERENT_BIT : 0); - glNamedBufferStorage(gl_buffer.handle, allocate_size, nullptr, flags); - mapped_ptr = static_cast<u8*>(glMapNamedBufferRange( - gl_buffer.handle, 0, buffer_size, flags | (coherent ? 0 : GL_MAP_FLUSH_EXPLICIT_BIT))); - } else { - glNamedBufferData(gl_buffer.handle, allocate_size, nullptr, GL_STREAM_DRAW); + static constexpr GLbitfield flags = GL_MAP_WRITE_BIT | GL_MAP_PERSISTENT_BIT; + glNamedBufferStorage(gl_buffer.handle, allocate_size, nullptr, flags); + mapped_ptr = static_cast<u8*>( + glMapNamedBufferRange(gl_buffer.handle, 0, buffer_size, flags | GL_MAP_FLUSH_EXPLICIT_BIT)); + + if (device.HasVertexBufferUnifiedMemory()) { + glMakeNamedBufferResidentNV(gl_buffer.handle, GL_READ_ONLY); + glGetNamedBufferParameterui64vNV(gl_buffer.handle, GL_BUFFER_GPU_ADDRESS_NV, &gpu_address); } } OGLStreamBuffer::~OGLStreamBuffer() { - if (persistent) { - glUnmapNamedBuffer(gl_buffer.handle); - } + glUnmapNamedBuffer(gl_buffer.handle); gl_buffer.Release(); } @@ -60,36 +57,21 @@ std::tuple<u8*, GLintptr, bool> OGLStreamBuffer::Map(GLsizeiptr size, GLintptr a bool invalidate = false; if (buffer_pos + size > buffer_size) { + MICROPROFILE_SCOPE(OpenGL_StreamBuffer); + glInvalidateBufferData(gl_buffer.handle); + buffer_pos = 0; invalidate = true; - - if (persistent) { - glUnmapNamedBuffer(gl_buffer.handle); - } } - if (invalidate || !persistent) { - MICROPROFILE_SCOPE(OpenGL_StreamBuffer); - GLbitfield flags = GL_MAP_WRITE_BIT | (persistent ? GL_MAP_PERSISTENT_BIT : 0) | - (coherent ? GL_MAP_COHERENT_BIT : GL_MAP_FLUSH_EXPLICIT_BIT) | - (invalidate ? GL_MAP_INVALIDATE_BUFFER_BIT : GL_MAP_UNSYNCHRONIZED_BIT); - mapped_ptr = static_cast<u8*>( - glMapNamedBufferRange(gl_buffer.handle, buffer_pos, buffer_size - buffer_pos, flags)); - mapped_offset = buffer_pos; - } - - return std::make_tuple(mapped_ptr + buffer_pos - mapped_offset, buffer_pos, invalidate); + return std::make_tuple(mapped_ptr + buffer_pos, buffer_pos, invalidate); } void OGLStreamBuffer::Unmap(GLsizeiptr size) { ASSERT(size <= mapped_size); - if (!coherent && size > 0) { - glFlushMappedNamedBufferRange(gl_buffer.handle, buffer_pos - mapped_offset, size); - } - - if (!persistent) { - glUnmapNamedBuffer(gl_buffer.handle); + if (size > 0) { + glFlushMappedNamedBufferRange(gl_buffer.handle, buffer_pos, size); } buffer_pos += size; diff --git a/src/video_core/renderer_opengl/gl_stream_buffer.h b/src/video_core/renderer_opengl/gl_stream_buffer.h index 866da3594..307a67113 100644 --- a/src/video_core/renderer_opengl/gl_stream_buffer.h +++ b/src/video_core/renderer_opengl/gl_stream_buffer.h @@ -11,10 +11,11 @@ namespace OpenGL { +class Device; + class OGLStreamBuffer : private NonCopyable { public: - explicit OGLStreamBuffer(GLsizeiptr size, bool vertex_data_usage, bool prefer_coherent = false, - bool use_persistent = true); + explicit OGLStreamBuffer(const Device& device, GLsizeiptr size, bool vertex_data_usage); ~OGLStreamBuffer(); /* @@ -33,19 +34,20 @@ public: return gl_buffer.handle; } - GLsizeiptr Size() const { + u64 Address() const { + return gpu_address; + } + + GLsizeiptr Size() const noexcept { return buffer_size; } private: OGLBuffer gl_buffer; - bool coherent = false; - bool persistent = false; - + GLuint64EXT gpu_address = 0; GLintptr buffer_pos = 0; GLsizeiptr buffer_size = 0; - GLintptr mapped_offset = 0; GLsizeiptr mapped_size = 0; u8* mapped_ptr = nullptr; }; diff --git a/src/video_core/renderer_opengl/renderer_opengl.cpp b/src/video_core/renderer_opengl/renderer_opengl.cpp index 6214fcbc3..c40adb6e7 100644 --- a/src/video_core/renderer_opengl/renderer_opengl.cpp +++ b/src/video_core/renderer_opengl/renderer_opengl.cpp @@ -488,6 +488,15 @@ void RendererOpenGL::InitOpenGLObjects() { // Clear screen to black LoadColorToActiveGLTexture(0, 0, 0, 0, screen_info.texture); + + // Enable unified vertex attributes and query vertex buffer address when the driver supports it + if (device.HasVertexBufferUnifiedMemory()) { + glEnableClientState(GL_VERTEX_ATTRIB_ARRAY_UNIFIED_NV); + + glMakeNamedBufferResidentNV(vertex_buffer.handle, GL_READ_ONLY); + glGetNamedBufferParameterui64vNV(vertex_buffer.handle, GL_BUFFER_GPU_ADDRESS_NV, + &vertex_buffer_address); + } } void RendererOpenGL::AddTelemetryFields() { @@ -656,7 +665,13 @@ void RendererOpenGL::DrawScreen(const Layout::FramebufferLayout& layout) { offsetof(ScreenRectVertex, tex_coord)); glVertexAttribBinding(PositionLocation, 0); glVertexAttribBinding(TexCoordLocation, 0); - glBindVertexBuffer(0, vertex_buffer.handle, 0, sizeof(ScreenRectVertex)); + if (device.HasVertexBufferUnifiedMemory()) { + glBindVertexBuffer(0, 0, 0, sizeof(ScreenRectVertex)); + glBufferAddressRangeNV(GL_VERTEX_ATTRIB_ARRAY_ADDRESS_NV, 0, vertex_buffer_address, + sizeof(vertices)); + } else { + glBindVertexBuffer(0, vertex_buffer.handle, 0, sizeof(ScreenRectVertex)); + } glBindTextureUnit(0, screen_info.display_texture); glBindSampler(0, 0); diff --git a/src/video_core/renderer_opengl/renderer_opengl.h b/src/video_core/renderer_opengl/renderer_opengl.h index 61bf507f4..8b18d32e6 100644 --- a/src/video_core/renderer_opengl/renderer_opengl.h +++ b/src/video_core/renderer_opengl/renderer_opengl.h @@ -107,6 +107,9 @@ private: OGLPipeline pipeline; OGLFramebuffer screenshot_framebuffer; + // GPU address of the vertex buffer + GLuint64EXT vertex_buffer_address = 0; + /// Display information for Switch screen ScreenInfo screen_info; diff --git a/src/video_core/renderer_vulkan/vk_buffer_cache.cpp b/src/video_core/renderer_vulkan/vk_buffer_cache.cpp index 1fde38328..f10f96cd8 100644 --- a/src/video_core/renderer_vulkan/vk_buffer_cache.cpp +++ b/src/video_core/renderer_vulkan/vk_buffer_cache.cpp @@ -37,9 +37,9 @@ std::unique_ptr<VKStreamBuffer> CreateStreamBuffer(const VKDevice& device, VKSch } // Anonymous namespace -Buffer::Buffer(const VKDevice& device, VKMemoryManager& memory_manager, VAddr cpu_addr, - std::size_t size) - : VideoCommon::BufferBlock{cpu_addr, size} { +Buffer::Buffer(const VKDevice& device, VKMemoryManager& memory_manager, VKScheduler& scheduler_, + VKStagingBufferPool& staging_pool_, VAddr cpu_addr, std::size_t size) + : VideoCommon::BufferBlock{cpu_addr, size}, scheduler{scheduler_}, staging_pool{staging_pool_} { VkBufferCreateInfo ci; ci.sType = VK_STRUCTURE_TYPE_BUFFER_CREATE_INFO; ci.pNext = nullptr; @@ -56,40 +56,15 @@ Buffer::Buffer(const VKDevice& device, VKMemoryManager& memory_manager, VAddr cp Buffer::~Buffer() = default; -VKBufferCache::VKBufferCache(VideoCore::RasterizerInterface& rasterizer, Core::System& system, - const VKDevice& device, VKMemoryManager& memory_manager, - VKScheduler& scheduler, VKStagingBufferPool& staging_pool) - : VideoCommon::BufferCache<Buffer, VkBuffer, VKStreamBuffer>{rasterizer, system, - CreateStreamBuffer(device, - scheduler)}, - device{device}, memory_manager{memory_manager}, scheduler{scheduler}, staging_pool{ - staging_pool} {} - -VKBufferCache::~VKBufferCache() = default; - -std::shared_ptr<Buffer> VKBufferCache::CreateBlock(VAddr cpu_addr, std::size_t size) { - return std::make_shared<Buffer>(device, memory_manager, cpu_addr, size); -} - -VkBuffer VKBufferCache::GetEmptyBuffer(std::size_t size) { - size = std::max(size, std::size_t(4)); - const auto& empty = staging_pool.GetUnusedBuffer(size, false); - scheduler.RequestOutsideRenderPassOperationContext(); - scheduler.Record([size, buffer = *empty.handle](vk::CommandBuffer cmdbuf) { - cmdbuf.FillBuffer(buffer, 0, size, 0); - }); - return *empty.handle; -} - -void VKBufferCache::UploadBlockData(const Buffer& buffer, std::size_t offset, std::size_t size, - const u8* data) { +void Buffer::Upload(std::size_t offset, std::size_t size, const u8* data) const { const auto& staging = staging_pool.GetUnusedBuffer(size, true); std::memcpy(staging.commit->Map(size), data, size); scheduler.RequestOutsideRenderPassOperationContext(); - scheduler.Record([staging = *staging.handle, buffer = buffer.Handle(), offset, - size](vk::CommandBuffer cmdbuf) { - cmdbuf.CopyBuffer(staging, buffer, VkBufferCopy{0, offset, size}); + + const VkBuffer handle = Handle(); + scheduler.Record([staging = *staging.handle, handle, offset, size](vk::CommandBuffer cmdbuf) { + cmdbuf.CopyBuffer(staging, handle, VkBufferCopy{0, offset, size}); VkBufferMemoryBarrier barrier; barrier.sType = VK_STRUCTURE_TYPE_BUFFER_MEMORY_BARRIER; @@ -98,7 +73,7 @@ void VKBufferCache::UploadBlockData(const Buffer& buffer, std::size_t offset, st barrier.dstAccessMask = UPLOAD_ACCESS_BARRIERS; barrier.srcQueueFamilyIndex = VK_QUEUE_FAMILY_IGNORED; barrier.dstQueueFamilyIndex = VK_QUEUE_FAMILY_IGNORED; - barrier.buffer = buffer; + barrier.buffer = handle; barrier.offset = offset; barrier.size = size; cmdbuf.PipelineBarrier(VK_PIPELINE_STAGE_TRANSFER_BIT, UPLOAD_PIPELINE_STAGE, 0, {}, @@ -106,12 +81,12 @@ void VKBufferCache::UploadBlockData(const Buffer& buffer, std::size_t offset, st }); } -void VKBufferCache::DownloadBlockData(const Buffer& buffer, std::size_t offset, std::size_t size, - u8* data) { +void Buffer::Download(std::size_t offset, std::size_t size, u8* data) const { const auto& staging = staging_pool.GetUnusedBuffer(size, true); scheduler.RequestOutsideRenderPassOperationContext(); - scheduler.Record([staging = *staging.handle, buffer = buffer.Handle(), offset, - size](vk::CommandBuffer cmdbuf) { + + const VkBuffer handle = Handle(); + scheduler.Record([staging = *staging.handle, handle, offset, size](vk::CommandBuffer cmdbuf) { VkBufferMemoryBarrier barrier; barrier.sType = VK_STRUCTURE_TYPE_BUFFER_MEMORY_BARRIER; barrier.pNext = nullptr; @@ -119,7 +94,7 @@ void VKBufferCache::DownloadBlockData(const Buffer& buffer, std::size_t offset, barrier.dstAccessMask = VK_ACCESS_TRANSFER_READ_BIT; barrier.srcQueueFamilyIndex = VK_QUEUE_FAMILY_IGNORED; barrier.dstQueueFamilyIndex = VK_QUEUE_FAMILY_IGNORED; - barrier.buffer = buffer; + barrier.buffer = handle; barrier.offset = offset; barrier.size = size; @@ -127,17 +102,19 @@ void VKBufferCache::DownloadBlockData(const Buffer& buffer, std::size_t offset, VK_PIPELINE_STAGE_FRAGMENT_SHADER_BIT | VK_PIPELINE_STAGE_COMPUTE_SHADER_BIT, VK_PIPELINE_STAGE_TRANSFER_BIT, 0, {}, barrier, {}); - cmdbuf.CopyBuffer(buffer, staging, VkBufferCopy{offset, 0, size}); + cmdbuf.CopyBuffer(handle, staging, VkBufferCopy{offset, 0, size}); }); scheduler.Finish(); std::memcpy(data, staging.commit->Map(size), size); } -void VKBufferCache::CopyBlock(const Buffer& src, const Buffer& dst, std::size_t src_offset, - std::size_t dst_offset, std::size_t size) { +void Buffer::CopyFrom(const Buffer& src, std::size_t src_offset, std::size_t dst_offset, + std::size_t size) const { scheduler.RequestOutsideRenderPassOperationContext(); - scheduler.Record([src_buffer = src.Handle(), dst_buffer = dst.Handle(), src_offset, dst_offset, + + const VkBuffer dst_buffer = Handle(); + scheduler.Record([src_buffer = src.Handle(), dst_buffer, src_offset, dst_offset, size](vk::CommandBuffer cmdbuf) { cmdbuf.CopyBuffer(src_buffer, dst_buffer, VkBufferCopy{src_offset, dst_offset, size}); @@ -165,4 +142,30 @@ void VKBufferCache::CopyBlock(const Buffer& src, const Buffer& dst, std::size_t }); } +VKBufferCache::VKBufferCache(VideoCore::RasterizerInterface& rasterizer, Core::System& system, + const VKDevice& device, VKMemoryManager& memory_manager, + VKScheduler& scheduler, VKStagingBufferPool& staging_pool) + : VideoCommon::BufferCache<Buffer, VkBuffer, VKStreamBuffer>{rasterizer, system, + CreateStreamBuffer(device, + scheduler)}, + device{device}, memory_manager{memory_manager}, scheduler{scheduler}, staging_pool{ + staging_pool} {} + +VKBufferCache::~VKBufferCache() = default; + +std::shared_ptr<Buffer> VKBufferCache::CreateBlock(VAddr cpu_addr, std::size_t size) { + return std::make_shared<Buffer>(device, memory_manager, scheduler, staging_pool, cpu_addr, + size); +} + +VKBufferCache::BufferInfo VKBufferCache::GetEmptyBuffer(std::size_t size) { + size = std::max(size, std::size_t(4)); + const auto& empty = staging_pool.GetUnusedBuffer(size, false); + scheduler.RequestOutsideRenderPassOperationContext(); + scheduler.Record([size, buffer = *empty.handle](vk::CommandBuffer cmdbuf) { + cmdbuf.FillBuffer(buffer, 0, size, 0); + }); + return {*empty.handle, 0, 0}; +} + } // namespace Vulkan diff --git a/src/video_core/renderer_vulkan/vk_buffer_cache.h b/src/video_core/renderer_vulkan/vk_buffer_cache.h index 9ebbef835..3630aca77 100644 --- a/src/video_core/renderer_vulkan/vk_buffer_cache.h +++ b/src/video_core/renderer_vulkan/vk_buffer_cache.h @@ -25,15 +25,29 @@ class VKScheduler; class Buffer final : public VideoCommon::BufferBlock { public: - explicit Buffer(const VKDevice& device, VKMemoryManager& memory_manager, VAddr cpu_addr, - std::size_t size); + explicit Buffer(const VKDevice& device, VKMemoryManager& memory_manager, VKScheduler& scheduler, + VKStagingBufferPool& staging_pool, VAddr cpu_addr, std::size_t size); ~Buffer(); + void Upload(std::size_t offset, std::size_t size, const u8* data) const; + + void Download(std::size_t offset, std::size_t size, u8* data) const; + + void CopyFrom(const Buffer& src, std::size_t src_offset, std::size_t dst_offset, + std::size_t size) const; + VkBuffer Handle() const { return *buffer.handle; } + u64 Address() const { + return 0; + } + private: + VKScheduler& scheduler; + VKStagingBufferPool& staging_pool; + VKBuffer buffer; }; @@ -44,20 +58,11 @@ public: VKScheduler& scheduler, VKStagingBufferPool& staging_pool); ~VKBufferCache(); - VkBuffer GetEmptyBuffer(std::size_t size) override; + BufferInfo GetEmptyBuffer(std::size_t size) override; protected: std::shared_ptr<Buffer> CreateBlock(VAddr cpu_addr, std::size_t size) override; - void UploadBlockData(const Buffer& buffer, std::size_t offset, std::size_t size, - const u8* data) override; - - void DownloadBlockData(const Buffer& buffer, std::size_t offset, std::size_t size, - u8* data) override; - - void CopyBlock(const Buffer& src, const Buffer& dst, std::size_t src_offset, - std::size_t dst_offset, std::size_t size) override; - private: const VKDevice& device; VKMemoryManager& memory_manager; diff --git a/src/video_core/renderer_vulkan/vk_rasterizer.cpp b/src/video_core/renderer_vulkan/vk_rasterizer.cpp index 29001953c..a8d94eac3 100644 --- a/src/video_core/renderer_vulkan/vk_rasterizer.cpp +++ b/src/video_core/renderer_vulkan/vk_rasterizer.cpp @@ -143,6 +143,49 @@ Tegra::Texture::FullTextureInfo GetTextureInfo(const Engine& engine, const Entry } } +/// @brief Determine if an attachment to be updated has to preserve contents +/// @param is_clear True when a clear is being executed +/// @param regs 3D registers +/// @return True when the contents have to be preserved +bool HasToPreserveColorContents(bool is_clear, const Maxwell& regs) { + if (!is_clear) { + return true; + } + // First we have to make sure all clear masks are enabled. + if (!regs.clear_buffers.R || !regs.clear_buffers.G || !regs.clear_buffers.B || + !regs.clear_buffers.A) { + return true; + } + // If scissors are disabled, the whole screen is cleared + if (!regs.clear_flags.scissor) { + return false; + } + // Then we have to confirm scissor testing clears the whole image + const std::size_t index = regs.clear_buffers.RT; + const auto& scissor = regs.scissor_test[0]; + return scissor.min_x > 0 || scissor.min_y > 0 || scissor.max_x < regs.rt[index].width || + scissor.max_y < regs.rt[index].height; +} + +/// @brief Determine if an attachment to be updated has to preserve contents +/// @param is_clear True when a clear is being executed +/// @param regs 3D registers +/// @return True when the contents have to be preserved +bool HasToPreserveDepthContents(bool is_clear, const Maxwell& regs) { + // If we are not clearing, the contents have to be preserved + if (!is_clear) { + return true; + } + // For depth stencil clears we only have to confirm scissor test covers the whole image + if (!regs.clear_flags.scissor) { + return false; + } + // Make sure the clear cover the whole image + const auto& scissor = regs.scissor_test[0]; + return scissor.min_x > 0 || scissor.min_y > 0 || scissor.max_x < regs.zeta_width || + scissor.max_y < regs.zeta_height; +} + } // Anonymous namespace class BufferBindings final { @@ -344,7 +387,7 @@ void RasterizerVulkan::Draw(bool is_indexed, bool is_instanced) { buffer_cache.Unmap(); - const Texceptions texceptions = UpdateAttachments(); + const Texceptions texceptions = UpdateAttachments(false); SetupImageTransitions(texceptions, color_attachments, zeta_attachment); key.renderpass_params = GetRenderPassParams(texceptions); @@ -400,7 +443,7 @@ void RasterizerVulkan::Clear() { return; } - [[maybe_unused]] const auto texceptions = UpdateAttachments(); + [[maybe_unused]] const auto texceptions = UpdateAttachments(true); DEBUG_ASSERT(texceptions.none()); SetupImageTransitions(0, color_attachments, zeta_attachment); @@ -677,9 +720,12 @@ void RasterizerVulkan::FlushWork() { draw_counter = 0; } -RasterizerVulkan::Texceptions RasterizerVulkan::UpdateAttachments() { +RasterizerVulkan::Texceptions RasterizerVulkan::UpdateAttachments(bool is_clear) { MICROPROFILE_SCOPE(Vulkan_RenderTargets); - auto& dirty = system.GPU().Maxwell3D().dirty.flags; + auto& maxwell3d = system.GPU().Maxwell3D(); + auto& dirty = maxwell3d.dirty.flags; + auto& regs = maxwell3d.regs; + const bool update_rendertargets = dirty[VideoCommon::Dirty::RenderTargets]; dirty[VideoCommon::Dirty::RenderTargets] = false; @@ -688,7 +734,8 @@ RasterizerVulkan::Texceptions RasterizerVulkan::UpdateAttachments() { Texceptions texceptions; for (std::size_t rt = 0; rt < Maxwell::NumRenderTargets; ++rt) { if (update_rendertargets) { - color_attachments[rt] = texture_cache.GetColorBufferSurface(rt, true); + const bool preserve_contents = HasToPreserveColorContents(is_clear, regs); + color_attachments[rt] = texture_cache.GetColorBufferSurface(rt, preserve_contents); } if (color_attachments[rt] && WalkAttachmentOverlaps(*color_attachments[rt])) { texceptions[rt] = true; @@ -696,7 +743,8 @@ RasterizerVulkan::Texceptions RasterizerVulkan::UpdateAttachments() { } if (update_rendertargets) { - zeta_attachment = texture_cache.GetDepthBufferSurface(true); + const bool preserve_contents = HasToPreserveDepthContents(is_clear, regs); + zeta_attachment = texture_cache.GetDepthBufferSurface(preserve_contents); } if (zeta_attachment && WalkAttachmentOverlaps(*zeta_attachment)) { texceptions[ZETA_TEXCEPTION_INDEX] = true; @@ -870,10 +918,10 @@ void RasterizerVulkan::BeginTransformFeedback() { UNIMPLEMENTED_IF(binding.buffer_offset != 0); const GPUVAddr gpu_addr = binding.Address(); - const auto size = static_cast<VkDeviceSize>(binding.buffer_size); - const auto [buffer, offset] = buffer_cache.UploadMemory(gpu_addr, size, 4, true); + const VkDeviceSize size = static_cast<VkDeviceSize>(binding.buffer_size); + const auto info = buffer_cache.UploadMemory(gpu_addr, size, 4, true); - scheduler.Record([buffer = buffer, offset = offset, size](vk::CommandBuffer cmdbuf) { + scheduler.Record([buffer = info.handle, offset = info.offset, size](vk::CommandBuffer cmdbuf) { cmdbuf.BindTransformFeedbackBuffersEXT(0, 1, &buffer, &offset, &size); cmdbuf.BeginTransformFeedbackEXT(0, 0, nullptr, nullptr); }); @@ -925,8 +973,8 @@ void RasterizerVulkan::SetupVertexArrays(FixedPipelineState::VertexInput& vertex buffer_bindings.AddVertexBinding(DefaultBuffer(), 0); continue; } - const auto [buffer, offset] = buffer_cache.UploadMemory(start, size); - buffer_bindings.AddVertexBinding(buffer, offset); + const auto info = buffer_cache.UploadMemory(start, size); + buffer_bindings.AddVertexBinding(info.handle, info.offset); } } @@ -948,7 +996,9 @@ void RasterizerVulkan::SetupIndexBuffer(BufferBindings& buffer_bindings, DrawPar break; } const GPUVAddr gpu_addr = regs.index_array.IndexStart(); - auto [buffer, offset] = buffer_cache.UploadMemory(gpu_addr, CalculateIndexBufferSize()); + const auto info = buffer_cache.UploadMemory(gpu_addr, CalculateIndexBufferSize()); + VkBuffer buffer = info.handle; + u64 offset = info.offset; std::tie(buffer, offset) = quad_indexed_pass.Assemble( regs.index_array.format, params.num_vertices, params.base_vertex, buffer, offset); @@ -962,7 +1012,9 @@ void RasterizerVulkan::SetupIndexBuffer(BufferBindings& buffer_bindings, DrawPar break; } const GPUVAddr gpu_addr = regs.index_array.IndexStart(); - auto [buffer, offset] = buffer_cache.UploadMemory(gpu_addr, CalculateIndexBufferSize()); + const auto info = buffer_cache.UploadMemory(gpu_addr, CalculateIndexBufferSize()); + VkBuffer buffer = info.handle; + u64 offset = info.offset; auto format = regs.index_array.format; const bool is_uint8 = format == Maxwell::IndexFormat::UnsignedByte; @@ -1109,10 +1161,9 @@ void RasterizerVulkan::SetupConstBuffer(const ConstBufferEntry& entry, Common::AlignUp(CalculateConstBufferSize(entry, buffer), 4 * sizeof(float)); ASSERT(size <= MaxConstbufferSize); - const auto [buffer_handle, offset] = + const auto info = buffer_cache.UploadMemory(buffer.address, size, device.GetUniformBufferAlignment()); - - update_descriptor_queue.AddBuffer(buffer_handle, offset, size); + update_descriptor_queue.AddBuffer(info.handle, info.offset, size); } void RasterizerVulkan::SetupGlobalBuffer(const GlobalBufferEntry& entry, GPUVAddr address) { @@ -1126,14 +1177,14 @@ void RasterizerVulkan::SetupGlobalBuffer(const GlobalBufferEntry& entry, GPUVAdd // Note: Do *not* use DefaultBuffer() here, storage buffers can be written breaking the // default buffer. static constexpr std::size_t dummy_size = 4; - const auto buffer = buffer_cache.GetEmptyBuffer(dummy_size); - update_descriptor_queue.AddBuffer(buffer, 0, dummy_size); + const auto info = buffer_cache.GetEmptyBuffer(dummy_size); + update_descriptor_queue.AddBuffer(info.handle, info.offset, dummy_size); return; } - const auto [buffer, offset] = buffer_cache.UploadMemory( + const auto info = buffer_cache.UploadMemory( actual_addr, size, device.GetStorageBufferAlignment(), entry.IsWritten()); - update_descriptor_queue.AddBuffer(buffer, offset, size); + update_descriptor_queue.AddBuffer(info.handle, info.offset, size); } void RasterizerVulkan::SetupUniformTexels(const Tegra::Texture::TICEntry& tic, diff --git a/src/video_core/renderer_vulkan/vk_rasterizer.h b/src/video_core/renderer_vulkan/vk_rasterizer.h index c8c187606..83e00e7e9 100644 --- a/src/video_core/renderer_vulkan/vk_rasterizer.h +++ b/src/video_core/renderer_vulkan/vk_rasterizer.h @@ -159,7 +159,10 @@ private: void FlushWork(); - Texceptions UpdateAttachments(); + /// @brief Updates the currently bound attachments + /// @param is_clear True when the framebuffer is updated as a clear + /// @return Bitfield of attachments being used as sampled textures + Texceptions UpdateAttachments(bool is_clear); std::tuple<VkFramebuffer, VkExtent2D> ConfigureFramebuffers(VkRenderPass renderpass); diff --git a/src/video_core/renderer_vulkan/vk_stream_buffer.h b/src/video_core/renderer_vulkan/vk_stream_buffer.h index c765c60a0..689f0d276 100644 --- a/src/video_core/renderer_vulkan/vk_stream_buffer.h +++ b/src/video_core/renderer_vulkan/vk_stream_buffer.h @@ -35,10 +35,14 @@ public: /// Ensures that "size" bytes of memory are available to the GPU, potentially recording a copy. void Unmap(u64 size); - VkBuffer Handle() const { + VkBuffer Handle() const noexcept { return *buffer; } + u64 Address() const noexcept { + return 0; + } + private: struct Watch final { VKFenceWatch fence; diff --git a/src/video_core/shader/decode/half_set.cpp b/src/video_core/shader/decode/half_set.cpp index 848e46874..b2e88fa20 100644 --- a/src/video_core/shader/decode/half_set.cpp +++ b/src/video_core/shader/decode/half_set.cpp @@ -13,55 +13,101 @@ namespace VideoCommon::Shader { +using std::move; using Tegra::Shader::Instruction; using Tegra::Shader::OpCode; +using Tegra::Shader::PredCondition; u32 ShaderIR::DecodeHalfSet(NodeBlock& bb, u32 pc) { const Instruction instr = {program_code[pc]}; const auto opcode = OpCode::Decode(instr); - if (instr.hset2.ftz == 0) { - LOG_DEBUG(HW_GPU, "{} without FTZ is not implemented", opcode->get().GetName()); + PredCondition cond; + bool bf; + bool ftz; + bool neg_a; + bool abs_a; + bool neg_b; + bool abs_b; + switch (opcode->get().GetId()) { + case OpCode::Id::HSET2_C: + case OpCode::Id::HSET2_IMM: + cond = instr.hsetp2.cbuf_and_imm.cond; + bf = instr.Bit(53); + ftz = instr.Bit(54); + neg_a = instr.Bit(43); + abs_a = instr.Bit(44); + neg_b = instr.Bit(56); + abs_b = instr.Bit(54); + break; + case OpCode::Id::HSET2_R: + cond = instr.hsetp2.reg.cond; + bf = instr.Bit(49); + ftz = instr.Bit(50); + neg_a = instr.Bit(43); + abs_a = instr.Bit(44); + neg_b = instr.Bit(31); + abs_b = instr.Bit(30); + break; + default: + UNREACHABLE(); } - Node op_a = UnpackHalfFloat(GetRegister(instr.gpr8), instr.hset2.type_a); - op_a = GetOperandAbsNegHalf(op_a, instr.hset2.abs_a, instr.hset2.negate_a); - - Node op_b = [&]() { + Node op_b = [this, instr, opcode] { switch (opcode->get().GetId()) { + case OpCode::Id::HSET2_C: + // Inform as unimplemented as this is not tested. + UNIMPLEMENTED_MSG("HSET2_C is not implemented"); + return GetConstBuffer(instr.cbuf34.index, instr.cbuf34.GetOffset()); case OpCode::Id::HSET2_R: return GetRegister(instr.gpr20); + case OpCode::Id::HSET2_IMM: + return UnpackHalfImmediate(instr, true); default: UNREACHABLE(); - return Immediate(0); + return Node{}; } }(); - op_b = UnpackHalfFloat(op_b, instr.hset2.type_b); - op_b = GetOperandAbsNegHalf(op_b, instr.hset2.abs_b, instr.hset2.negate_b); - const Node second_pred = GetPredicate(instr.hset2.pred39, instr.hset2.neg_pred); + if (!ftz) { + LOG_DEBUG(HW_GPU, "{} without FTZ is not implemented", opcode->get().GetName()); + } + + Node op_a = UnpackHalfFloat(GetRegister(instr.gpr8), instr.hset2.type_a); + op_a = GetOperandAbsNegHalf(op_a, abs_a, neg_a); + + switch (opcode->get().GetId()) { + case OpCode::Id::HSET2_R: + op_b = GetOperandAbsNegHalf(move(op_b), abs_b, neg_b); + [[fallthrough]]; + case OpCode::Id::HSET2_C: + op_b = UnpackHalfFloat(move(op_b), instr.hset2.type_b); + break; + default: + break; + } - const Node comparison_pair = GetPredicateComparisonHalf(instr.hset2.cond, op_a, op_b); + Node second_pred = GetPredicate(instr.hset2.pred39, instr.hset2.neg_pred); + + Node comparison_pair = GetPredicateComparisonHalf(cond, op_a, op_b); const OperationCode combiner = GetPredicateCombiner(instr.hset2.op); // HSET2 operates on each half float in the pack. std::array<Node, 2> values; for (u32 i = 0; i < 2; ++i) { - const u32 raw_value = instr.hset2.bf ? 0x3c00 : 0xffff; - const Node true_value = Immediate(raw_value << (i * 16)); - const Node false_value = Immediate(0); - - const Node comparison = - Operation(OperationCode::LogicalPick2, comparison_pair, Immediate(i)); - const Node predicate = Operation(combiner, comparison, second_pred); + const u32 raw_value = bf ? 0x3c00 : 0xffff; + Node true_value = Immediate(raw_value << (i * 16)); + Node false_value = Immediate(0); + Node comparison = Operation(OperationCode::LogicalPick2, comparison_pair, Immediate(i)); + Node predicate = Operation(combiner, comparison, second_pred); values[i] = - Operation(OperationCode::Select, NO_PRECISE, predicate, true_value, false_value); + Operation(OperationCode::Select, predicate, move(true_value), move(false_value)); } - const Node value = Operation(OperationCode::UBitwiseOr, NO_PRECISE, values[0], values[1]); - SetRegister(bb, instr.gpr0, value); + Node value = Operation(OperationCode::UBitwiseOr, values[0], values[1]); + SetRegister(bb, instr.gpr0, move(value)); return pc; } diff --git a/src/video_core/shader/decode/image.cpp b/src/video_core/shader/decode/image.cpp index 60b6ad72a..07778dc3e 100644 --- a/src/video_core/shader/decode/image.cpp +++ b/src/video_core/shader/decode/image.cpp @@ -97,6 +97,7 @@ ComponentType GetComponentType(Tegra::Engines::SamplerDescriptor descriptor, break; case TextureFormat::B5G6R5: case TextureFormat::B6G5R5: + case TextureFormat::BF10GF11RF11: if (component == 0) { return descriptor.b_type; } @@ -119,7 +120,7 @@ ComponentType GetComponentType(Tegra::Engines::SamplerDescriptor descriptor, } break; } - UNIMPLEMENTED_MSG("texture format not implement={}", format); + UNIMPLEMENTED_MSG("Texture format not implemented={}", format); return ComponentType::FLOAT; } @@ -191,6 +192,14 @@ u32 GetComponentSize(TextureFormat format, std::size_t component) { return 6; } return 0; + case TextureFormat::BF10GF11RF11: + if (component == 1 || component == 2) { + return 11; + } + if (component == 0) { + return 10; + } + return 0; case TextureFormat::G8R24: if (component == 0) { return 8; @@ -211,10 +220,9 @@ u32 GetComponentSize(TextureFormat format, std::size_t component) { return (component == 0 || component == 1) ? 8 : 0; case TextureFormat::G4R4: return (component == 0 || component == 1) ? 4 : 0; - default: - UNIMPLEMENTED_MSG("texture format not implement={}", format); - return 0; } + UNIMPLEMENTED_MSG("Texture format not implemented={}", format); + return 0; } std::size_t GetImageComponentMask(TextureFormat format) { @@ -235,6 +243,7 @@ std::size_t GetImageComponentMask(TextureFormat format) { case TextureFormat::R32_B24G8: case TextureFormat::B5G6R5: case TextureFormat::B6G5R5: + case TextureFormat::BF10GF11RF11: return std::size_t{R | G | B}; case TextureFormat::R32_G32: case TextureFormat::R16_G16: @@ -248,10 +257,9 @@ std::size_t GetImageComponentMask(TextureFormat format) { case TextureFormat::R8: case TextureFormat::R1: return std::size_t{R}; - default: - UNIMPLEMENTED_MSG("texture format not implement={}", format); - return std::size_t{R | G | B | A}; } + UNIMPLEMENTED_MSG("Texture format not implemented={}", format); + return std::size_t{R | G | B | A}; } std::size_t GetImageTypeNumCoordinates(Tegra::Shader::ImageType image_type) { @@ -299,7 +307,7 @@ std::pair<Node, bool> ShaderIR::GetComponentValue(ComponentType component_type, return {std::move(original_value), true}; } default: - UNIMPLEMENTED_MSG("Unimplement component type={}", component_type); + UNIMPLEMENTED_MSG("Unimplemented component type={}", component_type); return {std::move(original_value), true}; } } @@ -459,7 +467,7 @@ u32 ShaderIR::DecodeImage(NodeBlock& bb, u32 pc) { default: break; } - UNIMPLEMENTED_MSG("Unimplemented operation={} type={}", + UNIMPLEMENTED_MSG("Unimplemented operation={}, type={}", static_cast<u64>(instr.suatom_d.operation.Value()), static_cast<u64>(instr.suatom_d.operation_type.Value())); return OperationCode::AtomicImageAdd; diff --git a/src/video_core/texture_cache/surface_base.cpp b/src/video_core/texture_cache/surface_base.cpp index 94d3a6ae5..0caf3b4f0 100644 --- a/src/video_core/texture_cache/surface_base.cpp +++ b/src/video_core/texture_cache/surface_base.cpp @@ -120,6 +120,9 @@ std::optional<std::pair<u32, u32>> SurfaceBaseImpl::GetLayerMipmap( } const auto relative_address{static_cast<GPUVAddr>(candidate_gpu_addr - gpu_addr)}; const auto layer{static_cast<u32>(relative_address / layer_size)}; + if (layer >= params.depth) { + return {}; + } const GPUVAddr mipmap_address = relative_address - layer_size * layer; const auto mipmap_it = Common::BinaryFind(mipmap_offsets.begin(), mipmap_offsets.end(), mipmap_address); diff --git a/src/yuzu/main.cpp b/src/yuzu/main.cpp index 059b96e70..4dedb2549 100644 --- a/src/yuzu/main.cpp +++ b/src/yuzu/main.cpp @@ -56,6 +56,7 @@ static FileSys::VirtualFile VfsDirectoryCreateFileWrapper(const FileSys::Virtual #include <QShortcut> #include <QStatusBar> #include <QSysInfo> +#include <QUrl> #include <QtConcurrent/QtConcurrent> #include <fmt/format.h> @@ -836,6 +837,7 @@ void GMainWindow::ConnectMenuEvents() { connect(ui.action_Stop, &QAction::triggered, this, &GMainWindow::OnStopGame); connect(ui.action_Report_Compatibility, &QAction::triggered, this, &GMainWindow::OnMenuReportCompatibility); + connect(ui.action_Open_Mods_Page, &QAction::triggered, this, &GMainWindow::OnOpenModsPage); connect(ui.action_Restart, &QAction::triggered, this, [this] { BootGame(QString(game_path)); }); connect(ui.action_Configure, &QAction::triggered, this, &GMainWindow::OnConfigure); @@ -1807,6 +1809,16 @@ void GMainWindow::OnMenuReportCompatibility() { } } +void GMainWindow::OnOpenModsPage() { + const auto mods_page_url = QStringLiteral("https://github.com/yuzu-emu/yuzu/wiki/Switch-Mods"); + const QUrl mods_page(mods_page_url); + const bool open = QDesktopServices::openUrl(mods_page); + if (!open) { + QMessageBox::warning(this, tr("Error opening URL"), + tr("Unable to open the URL \"%1\".").arg(mods_page_url)); + } +} + void GMainWindow::ToggleFullscreen() { if (!emulation_running) { return; diff --git a/src/yuzu/main.h b/src/yuzu/main.h index 4f4c8ddbe..d55e55cc6 100644 --- a/src/yuzu/main.h +++ b/src/yuzu/main.h @@ -181,6 +181,7 @@ private slots: void OnPauseGame(); void OnStopGame(); void OnMenuReportCompatibility(); + void OnOpenModsPage(); /// Called whenever a user selects a game in the game list widget. void OnGameListLoadFile(QString game_path); void OnGameListOpenFolder(GameListOpenTarget target, const std::string& game_path); diff --git a/src/yuzu/main.ui b/src/yuzu/main.ui index 97c90f50b..b5745dfd5 100644 --- a/src/yuzu/main.ui +++ b/src/yuzu/main.ui @@ -113,6 +113,7 @@ <string>&Help</string> </property> <addaction name="action_Report_Compatibility"/> + <addaction name="action_Open_Mods_Page"/> <addaction name="separator"/> <addaction name="action_About"/> </widget> @@ -256,6 +257,11 @@ <bool>false</bool> </property> </action> + <action name="action_Open_Mods_Page"> + <property name="text"> + <string>Open Mods Page</string> + </property> + </action> <action name="action_Open_yuzu_Folder"> <property name="text"> <string>Open yuzu Folder</string> |