diff options
32 files changed, 474 insertions, 299 deletions
diff --git a/src/citra_qt/debugger/graphics_breakpoints.cpp b/src/citra_qt/debugger/graphics_breakpoints.cpp index 1da64f61..5202c168 100644 --- a/src/citra_qt/debugger/graphics_breakpoints.cpp +++ b/src/citra_qt/debugger/graphics_breakpoints.cpp @@ -4,7 +4,7 @@ #include <QMetaType> #include <QPushButton> -#include <QTreeWidget> +#include <QTreeView> #include <QVBoxLayout> #include <QLabel> @@ -23,7 +23,7 @@ BreakPointModel::BreakPointModel(std::shared_ptr<Pica::DebugContext> debug_conte int BreakPointModel::columnCount(const QModelIndex& parent) const { - return 2; + return 1; } int BreakPointModel::rowCount(const QModelIndex& parent) const @@ -38,29 +38,29 @@ QVariant BreakPointModel::data(const QModelIndex& index, int role) const switch (role) { case Qt::DisplayRole: { - switch (index.column()) { - case 0: - { + if (index.column() == 0) { static const std::map<Pica::DebugContext::Event, QString> map = { - { Pica::DebugContext::Event::CommandLoaded, tr("Pica command loaded") }, - { Pica::DebugContext::Event::CommandProcessed, tr("Pica command processed") }, + { Pica::DebugContext::Event::PicaCommandLoaded, tr("Pica command loaded") }, + { Pica::DebugContext::Event::PicaCommandProcessed, tr("Pica command processed") }, { Pica::DebugContext::Event::IncomingPrimitiveBatch, tr("Incoming primitive batch") }, { Pica::DebugContext::Event::FinishedPrimitiveBatch, tr("Finished primitive batch") }, - { Pica::DebugContext::Event::VertexLoaded, tr("Vertex loaded") } + { Pica::DebugContext::Event::VertexLoaded, tr("Vertex loaded") }, + { Pica::DebugContext::Event::IncomingDisplayTransfer, tr("Incoming display transfer") }, + { Pica::DebugContext::Event::GSPCommandProcessed, tr("GSP command processed") }, + { Pica::DebugContext::Event::BufferSwapped, tr("Buffers swapped") } }; DEBUG_ASSERT(map.size() == static_cast<size_t>(Pica::DebugContext::Event::NumEvents)); - return (map.find(event) != map.end()) ? map.at(event) : QString(); } - case 1: - return data(index, Role_IsEnabled).toBool() ? tr("Enabled") : tr("Disabled"); - - default: - break; - } + break; + } + case Qt::CheckStateRole: + { + if (index.column() == 0) + return data(index, Role_IsEnabled).toBool() ? Qt::Checked : Qt::Unchecked; break; } @@ -84,37 +84,34 @@ QVariant BreakPointModel::data(const QModelIndex& index, int role) const return QVariant(); } -QVariant BreakPointModel::headerData(int section, Qt::Orientation orientation, int role) const +Qt::ItemFlags BreakPointModel::flags(const QModelIndex &index) const { - switch(role) { - case Qt::DisplayRole: - { - if (section == 0) { - return tr("Event"); - } else if (section == 1) { - return tr("Status"); - } - - break; - } - } + if (!index.isValid()) + return 0; - return QVariant(); + Qt::ItemFlags flags = Qt::ItemIsEnabled; + if (index.column() == 0) + flags |= Qt::ItemIsUserCheckable; + return flags; } + bool BreakPointModel::setData(const QModelIndex& index, const QVariant& value, int role) { const auto event = static_cast<Pica::DebugContext::Event>(index.row()); switch (role) { - case Role_IsEnabled: + case Qt::CheckStateRole: { + if (index.column() != 0) + return false; + auto context = context_weak.lock(); if (!context) return false; - context->breakpoints[event].enabled = value.toBool(); - QModelIndex changed_index = createIndex(index.row(), 1); + context->breakpoints[event].enabled = value == Qt::Checked; + QModelIndex changed_index = createIndex(index.row(), 0); emit dataChanged(changed_index, changed_index); return true; } @@ -133,7 +130,7 @@ void BreakPointModel::OnBreakPointHit(Pica::DebugContext::Event event) active_breakpoint = context->active_breakpoint; at_breakpoint = context->at_breakpoint; emit dataChanged(createIndex(static_cast<int>(event), 0), - createIndex(static_cast<int>(event), 1)); + createIndex(static_cast<int>(event), 0)); } void BreakPointModel::OnResumed() @@ -144,7 +141,7 @@ void BreakPointModel::OnResumed() at_breakpoint = context->at_breakpoint; emit dataChanged(createIndex(static_cast<int>(active_breakpoint), 0), - createIndex(static_cast<int>(active_breakpoint), 1)); + createIndex(static_cast<int>(active_breakpoint), 0)); active_breakpoint = context->active_breakpoint; } @@ -162,13 +159,15 @@ GraphicsBreakPointsWidget::GraphicsBreakPointsWidget(std::shared_ptr<Pica::Debug breakpoint_model = new BreakPointModel(debug_context, this); breakpoint_list = new QTreeView; + breakpoint_list->setRootIsDecorated(false); + breakpoint_list->setHeaderHidden(true); breakpoint_list->setModel(breakpoint_model); - toggle_breakpoint_button = new QPushButton(tr("Enable")); - toggle_breakpoint_button->setEnabled(false); - qRegisterMetaType<Pica::DebugContext::Event>("Pica::DebugContext::Event"); + connect(breakpoint_list, SIGNAL(doubleClicked(const QModelIndex&)), + this, SLOT(OnItemDoubleClicked(const QModelIndex&))); + connect(resume_button, SIGNAL(clicked()), this, SLOT(OnResumeRequested())); connect(this, SIGNAL(BreakPointHit(Pica::DebugContext::Event,void*)), @@ -184,11 +183,6 @@ GraphicsBreakPointsWidget::GraphicsBreakPointsWidget(std::shared_ptr<Pica::Debug connect(this, SIGNAL(BreakPointsChanged(const QModelIndex&,const QModelIndex&)), breakpoint_model, SIGNAL(dataChanged(const QModelIndex&,const QModelIndex&))); - connect(breakpoint_list->selectionModel(), SIGNAL(currentChanged(QModelIndex,QModelIndex)), - this, SLOT(OnBreakpointSelectionChanged(QModelIndex))); - - connect(toggle_breakpoint_button, SIGNAL(clicked()), this, SLOT(OnToggleBreakpointEnabled())); - QWidget* main_widget = new QWidget; auto main_layout = new QVBoxLayout; { @@ -198,7 +192,6 @@ GraphicsBreakPointsWidget::GraphicsBreakPointsWidget(std::shared_ptr<Pica::Debug main_layout->addLayout(sub_layout); } main_layout->addWidget(breakpoint_list); - main_layout->addWidget(toggle_breakpoint_button); main_widget->setLayout(main_layout); setWidget(main_widget); @@ -234,32 +227,15 @@ void GraphicsBreakPointsWidget::OnResumeRequested() context->Resume(); } -void GraphicsBreakPointsWidget::OnBreakpointSelectionChanged(const QModelIndex& index) +void GraphicsBreakPointsWidget::OnItemDoubleClicked(const QModelIndex& index) { - if (!index.isValid()) { - toggle_breakpoint_button->setEnabled(false); + if (!index.isValid()) return; - } - toggle_breakpoint_button->setEnabled(true); - UpdateToggleBreakpointButton(index); -} - -void GraphicsBreakPointsWidget::OnToggleBreakpointEnabled() -{ - QModelIndex index = breakpoint_list->selectionModel()->currentIndex(); - bool new_state = !(breakpoint_model->data(index, BreakPointModel::Role_IsEnabled).toBool()); - - breakpoint_model->setData(index, new_state, - BreakPointModel::Role_IsEnabled); - UpdateToggleBreakpointButton(index); -} - -void GraphicsBreakPointsWidget::UpdateToggleBreakpointButton(const QModelIndex& index) -{ - if (true == breakpoint_model->data(index, BreakPointModel::Role_IsEnabled).toBool()) { - toggle_breakpoint_button->setText(tr("Disable")); - } else { - toggle_breakpoint_button->setText(tr("Enable")); - } + QModelIndex check_index = breakpoint_list->model()->index(index.row(), 0); + QVariant enabled = breakpoint_list->model()->data(check_index, Qt::CheckStateRole); + QVariant new_state = Qt::Unchecked; + if (enabled == Qt::Unchecked) + new_state = Qt::Checked; + breakpoint_list->model()->setData(check_index, new_state, Qt::CheckStateRole); } diff --git a/src/citra_qt/debugger/graphics_breakpoints.h b/src/citra_qt/debugger/graphics_breakpoints.h index 5b9ba324..d900729d 100644 --- a/src/citra_qt/debugger/graphics_breakpoints.h +++ b/src/citra_qt/debugger/graphics_breakpoints.h @@ -31,10 +31,9 @@ public: public slots: void OnBreakPointHit(Pica::DebugContext::Event event, void* data); + void OnItemDoubleClicked(const QModelIndex&); void OnResumeRequested(); void OnResumed(); - void OnBreakpointSelectionChanged(const QModelIndex&); - void OnToggleBreakpointEnabled(); signals: void Resumed(); @@ -42,11 +41,8 @@ signals: void BreakPointsChanged(const QModelIndex& topLeft, const QModelIndex& bottomRight); private: - void UpdateToggleBreakpointButton(const QModelIndex& index); - QLabel* status_text; QPushButton* resume_button; - QPushButton* toggle_breakpoint_button; BreakPointModel* breakpoint_model; QTreeView* breakpoint_list; diff --git a/src/citra_qt/debugger/graphics_breakpoints_p.h b/src/citra_qt/debugger/graphics_breakpoints_p.h index 34e72e85..00d8d510 100644 --- a/src/citra_qt/debugger/graphics_breakpoints_p.h +++ b/src/citra_qt/debugger/graphics_breakpoints_p.h @@ -23,7 +23,7 @@ public: int columnCount(const QModelIndex& parent = QModelIndex()) const override; int rowCount(const QModelIndex& parent = QModelIndex()) const override; QVariant data(const QModelIndex& index, int role = Qt::DisplayRole) const override; - QVariant headerData(int section, Qt::Orientation orientation, int role = Qt::DisplayRole) const override; + Qt::ItemFlags flags(const QModelIndex &index) const; bool setData(const QModelIndex& index, const QVariant& value, int role = Qt::EditRole) override; diff --git a/src/citra_qt/debugger/graphics_cmdlists.cpp b/src/citra_qt/debugger/graphics_cmdlists.cpp index e15ed7de..de10bce1 100644 --- a/src/citra_qt/debugger/graphics_cmdlists.cpp +++ b/src/citra_qt/debugger/graphics_cmdlists.cpp @@ -2,6 +2,8 @@ // Licensed under GPLv2 or any later version // Refer to the license.txt file included. +#include <QApplication> +#include <QClipboard> #include <QLabel> #include <QListView> #include <QMainWindow> @@ -74,7 +76,7 @@ TextureInfoDockWidget::TextureInfoDockWidget(const Pica::DebugUtils::TextureInfo format_choice->addItem(tr("I8")); format_choice->addItem(tr("A8")); format_choice->addItem(tr("IA4")); - format_choice->addItem(tr("UNK10")); + format_choice->addItem(tr("I4")); format_choice->addItem(tr("A4")); format_choice->addItem(tr("ETC1")); format_choice->addItem(tr("ETC1A4")); @@ -304,16 +306,24 @@ GPUCommandListWidget::GPUCommandListWidget(QWidget* parent) : QDockWidget(tr("Pi this, SLOT(OnCommandDoubleClicked(const QModelIndex&))); toggle_tracing = new QPushButton(tr("Start Tracing")); + QPushButton* copy_all = new QPushButton(tr("Copy All")); connect(toggle_tracing, SIGNAL(clicked()), this, SLOT(OnToggleTracing())); connect(this, SIGNAL(TracingFinished(const Pica::DebugUtils::PicaTrace&)), model, SLOT(OnPicaTraceFinished(const Pica::DebugUtils::PicaTrace&))); + connect(copy_all, SIGNAL(clicked()), this, SLOT(CopyAllToClipboard())); + command_info_widget = new QWidget; QVBoxLayout* main_layout = new QVBoxLayout; main_layout->addWidget(list_widget); - main_layout->addWidget(toggle_tracing); + { + QHBoxLayout* sub_layout = new QHBoxLayout; + sub_layout->addWidget(toggle_tracing); + sub_layout->addWidget(copy_all); + main_layout->addLayout(sub_layout); + } main_layout->addWidget(command_info_widget); main_widget->setLayout(main_layout); @@ -330,3 +340,21 @@ void GPUCommandListWidget::OnToggleTracing() { toggle_tracing->setText(tr("Start Tracing")); } } + +void GPUCommandListWidget::CopyAllToClipboard() { + QClipboard* clipboard = QApplication::clipboard(); + QString text; + + QAbstractItemModel* model = static_cast<QAbstractListModel*>(list_widget->model()); + + for (int row = 0; row < model->rowCount({}); ++row) { + for (int col = 0; col < model->columnCount({}); ++col) { + QModelIndex index = model->index(row, col); + text += model->data(index).value<QString>(); + text += '\t'; + } + text += '\n'; + } + + clipboard->setText(text); +} diff --git a/src/citra_qt/debugger/graphics_cmdlists.h b/src/citra_qt/debugger/graphics_cmdlists.h index a465d044..4859b6ec 100644 --- a/src/citra_qt/debugger/graphics_cmdlists.h +++ b/src/citra_qt/debugger/graphics_cmdlists.h @@ -49,6 +49,8 @@ public slots: void SetCommandInfo(const QModelIndex&); + void CopyAllToClipboard(); + signals: void TracingFinished(const Pica::DebugUtils::PicaTrace&); diff --git a/src/citra_qt/debugger/graphics_vertex_shader.cpp b/src/citra_qt/debugger/graphics_vertex_shader.cpp index 07441882..f42a2f4c 100644 --- a/src/citra_qt/debugger/graphics_vertex_shader.cpp +++ b/src/citra_qt/debugger/graphics_vertex_shader.cpp @@ -259,7 +259,7 @@ void GraphicsVertexShaderModel::OnUpdate() for (auto pattern : Pica::g_state.vs.swizzle_data) info.swizzle_info.push_back({pattern}); - info.labels.insert({ Pica::g_state.regs.vs_main_offset, "main" }); + info.labels.insert({ Pica::g_state.regs.vs.main_offset, "main" }); endResetModel(); } diff --git a/src/core/arm/dyncom/arm_dyncom_interpreter.cpp b/src/core/arm/dyncom/arm_dyncom_interpreter.cpp index 34cfb8cb..785f3956 100644 --- a/src/core/arm/dyncom/arm_dyncom_interpreter.cpp +++ b/src/core/arm/dyncom/arm_dyncom_interpreter.cpp @@ -4144,11 +4144,13 @@ unsigned InterpreterMainLoop(ARMul_State* cpu) { if (inst_base->cond == 0xE || CondPassed(cpu, inst_base->cond)) { bx_inst* const inst_cream = (bx_inst*)inst_base->component; + u32 address = RM; + if (inst_cream->Rm == 15) - LOG_WARNING(Core_ARM11, "BX at pc %x: use of Rm = R15 is discouraged", cpu->Reg[15]); + address += 2 * GET_INST_SIZE(cpu); - cpu->TFlag = cpu->Reg[inst_cream->Rm] & 0x1; - cpu->Reg[15] = cpu->Reg[inst_cream->Rm] & 0xfffffffe; + cpu->TFlag = address & 1; + cpu->Reg[15] = address & 0xfffffffe; INC_PC(sizeof(bx_inst)); goto DISPATCH; } @@ -6246,7 +6248,8 @@ unsigned InterpreterMainLoop(ARMul_State* cpu) { SWI_INST: { if (inst_base->cond == 0xE || CondPassed(cpu, inst_base->cond)) { - SVC::CallSVC(Memory::Read32(cpu->Reg[15])); + swi_inst* const inst_cream = (swi_inst*)inst_base->component; + SVC::CallSVC(inst_cream->num & 0xFFFF); } cpu->Reg[15] += GET_INST_SIZE(cpu); diff --git a/src/core/hle/function_wrappers.h b/src/core/hle/function_wrappers.h index 1ac6032a..9294789e 100644 --- a/src/core/hle/function_wrappers.h +++ b/src/core/hle/function_wrappers.h @@ -99,6 +99,18 @@ template<ResultCode func(MemoryInfo*, PageInfo*, u32)> void Wrap() { FuncReturn(retval); } +template<ResultCode func(MemoryInfo*, PageInfo*, Handle, u32)> void Wrap() { + MemoryInfo memory_info = {}; + PageInfo page_info = {}; + u32 retval = func(&memory_info, &page_info, PARAM(2), PARAM(3)).raw; + Core::g_app_core->SetReg(1, memory_info.base_address); + Core::g_app_core->SetReg(2, memory_info.size); + Core::g_app_core->SetReg(3, memory_info.permission); + Core::g_app_core->SetReg(4, memory_info.state); + Core::g_app_core->SetReg(5, page_info.flags); + FuncReturn(retval); +} + template<ResultCode func(s32*, u32)> void Wrap(){ s32 param_1 = 0; u32 retval = func(¶m_1, PARAM(1)).raw; diff --git a/src/core/hle/kernel/thread.cpp b/src/core/hle/kernel/thread.cpp index 64166ab9..8b49fc7d 100644 --- a/src/core/hle/kernel/thread.cpp +++ b/src/core/hle/kernel/thread.cpp @@ -217,6 +217,14 @@ static void SwitchContext(Thread* new_thread) { new_thread->context.pc -= thumb_mode ? 2 : 4; } + // Clean up the thread's wait_objects, they'll be restored if needed during + // the svcWaitSynchronization call + for (int i = 0; i < new_thread->wait_objects.size(); ++i) { + SharedPtr<WaitObject> object = new_thread->wait_objects[i]; + object->RemoveWaitingThread(new_thread); + } + new_thread->wait_objects.clear(); + ready_queue.remove(new_thread->current_priority, new_thread); new_thread->status = THREADSTATUS_RUNNING; diff --git a/src/core/hle/service/am/am.cpp b/src/core/hle/service/am/am.cpp index 57dc1ece..7332478f 100644 --- a/src/core/hle/service/am/am.cpp +++ b/src/core/hle/service/am/am.cpp @@ -38,6 +38,15 @@ void GetTitleIDList(Service::Interface* self) { LOG_WARNING(Service_AM, "(STUBBED) Requested %u titles from media type %u", num_titles, media_type); } +void GetNumContentInfos(Service::Interface* self) { + u32* cmd_buff = Kernel::GetCommandBuffer(); + + cmd_buff[1] = RESULT_SUCCESS.raw; + cmd_buff[2] = 1; // Number of content infos plus one + + LOG_WARNING(Service_AM, "(STUBBED) called"); +} + void Init() { using namespace Kernel; diff --git a/src/core/hle/service/am/am.h b/src/core/hle/service/am/am.h index 063b8bd0..0b78f539 100644 --- a/src/core/hle/service/am/am.h +++ b/src/core/hle/service/am/am.h @@ -37,6 +37,19 @@ void TitleIDListGetTotal(Service::Interface* self); */ void GetTitleIDList(Service::Interface* self); +/** + * AM::GetNumContentInfos service function + * Inputs: + * 0 : Command header (0x100100C0) + * 1 : Unknown + * 2 : Unknown + * 3 : Unknown + * Outputs: + * 1 : Result, 0 on success, otherwise error code + * 2 : Number of content infos plus one + */ +void GetNumContentInfos(Service::Interface* self); + /// Initialize AM service void Init(); diff --git a/src/core/hle/service/am/am_app.cpp b/src/core/hle/service/am/am_app.cpp index c6fc81bc..f40a87cb 100644 --- a/src/core/hle/service/am/am_app.cpp +++ b/src/core/hle/service/am/am_app.cpp @@ -9,11 +9,19 @@ namespace Service { namespace AM { -// Empty arrays are illegal -- commented out until an entry is added. -//const Interface::FunctionInfo FunctionTable[] = { }; +const Interface::FunctionInfo FunctionTable[] = { + {0x100100C0, GetNumContentInfos, "GetNumContentInfos"}, + {0x10020104, nullptr, "FindContentInfos"}, + {0x10030142, nullptr, "ListContentInfos"}, + {0x10040102, nullptr, "DeleteContents"}, + {0x10050084, nullptr, "GetDataTitleInfos"}, + {0x10070102, nullptr, "ListDataTitleTicketInfos"}, + {0x100900C0, nullptr, "IsDataTitleInUse"}, + {0x100A0000, nullptr, "IsExternalTitleDatabaseInitialized"}, +}; AM_APP_Interface::AM_APP_Interface() { - //Register(FunctionTable); + Register(FunctionTable); } } // namespace AM diff --git a/src/core/hle/service/cfg/cfg_s.cpp b/src/core/hle/service/cfg/cfg_s.cpp index af4adba8..a329514a 100644 --- a/src/core/hle/service/cfg/cfg_s.cpp +++ b/src/core/hle/service/cfg/cfg_s.cpp @@ -11,7 +11,9 @@ namespace CFG { const Interface::FunctionInfo FunctionTable[] = { {0x00010082, GetConfigInfoBlk2, "GetConfigInfoBlk2"}, - {0x00020000, nullptr, "SecureInfoGetRegion"}, + {0x00020000, SecureInfoGetRegion, "SecureInfoGetRegion"}, + {0x00030040, GenHashConsoleUnique, "GenHashConsoleUnique"}, + {0x00050000, GetSystemModel, "GetSystemModel"}, {0x04010082, GetConfigInfoBlk8, "GetConfigInfoBlk8"}, {0x04020082, nullptr, "SetConfigInfoBlk4"}, {0x04030000, UpdateConfigNANDSavegame, "UpdateConfigNANDSavegame"}, diff --git a/src/core/hle/service/frd/frd_u.cpp b/src/core/hle/service/frd/frd_u.cpp index 439c7282..3a5897d0 100644 --- a/src/core/hle/service/frd/frd_u.cpp +++ b/src/core/hle/service/frd/frd_u.cpp @@ -10,15 +10,26 @@ namespace Service { namespace FRD { const Interface::FunctionInfo FunctionTable[] = { + {0x00010000, nullptr, "HasLoggedIn"}, + {0x00030000, nullptr, "Login"}, + {0x00040000, nullptr, "Logout"}, {0x00050000, nullptr, "GetFriendKey"}, {0x00080000, nullptr, "GetMyPresence"}, + {0x00090000, nullptr, "GetMyScreenName"}, {0x00100040, nullptr, "GetPassword"}, + {0x00110080, nullptr, "GetFriendKeyList"}, {0x00190042, nullptr, "GetFriendFavoriteGame"}, {0x001A00C4, nullptr, "GetFriendInfo"}, {0x001B0080, nullptr, "IsOnFriendList"}, {0x001C0042, nullptr, "DecodeLocalFriendCode"}, {0x001D0002, nullptr, "SetCurrentlyPlayingText"}, - {0x00320042, nullptr, "SetClientSdkVersion"} + {0x00230000, nullptr, "GetLastResponseResult"}, + {0x00270040, nullptr, "ResultToErrorCode"}, + {0x00280244, nullptr, "RequestGameAuthentication"}, + {0x00290000, nullptr, "GetGameAuthenticationData"}, + {0x002A0204, nullptr, "RequestServiceLocator"}, + {0x002B0000, nullptr, "GetServiceLocatorData"}, + {0x00320042, nullptr, "SetClientSdkVersion"}, }; FRD_U_Interface::FRD_U_Interface() { diff --git a/src/core/hle/service/fs/fs_user.cpp b/src/core/hle/service/fs/fs_user.cpp index 0ad44e55..ae52083f 100644 --- a/src/core/hle/service/fs/fs_user.cpp +++ b/src/core/hle/service/fs/fs_user.cpp @@ -115,7 +115,8 @@ static void OpenFileDirectly(Service::Interface* self) { ResultVal<ArchiveHandle> archive_handle = OpenArchive(archive_id, archive_path); if (archive_handle.Failed()) { - LOG_ERROR(Service_FS, "failed to get a handle for archive"); + LOG_ERROR(Service_FS, "failed to get a handle for archive archive_id=0x%08X archive_path=%s", + archive_id, archive_path.DebugStr().c_str()); cmd_buff[1] = archive_handle.Code().raw; cmd_buff[3] = 0; return; @@ -128,7 +129,8 @@ static void OpenFileDirectly(Service::Interface* self) { cmd_buff[3] = Kernel::g_handle_table.Create(*file_res).MoveFrom(); } else { cmd_buff[3] = 0; - LOG_ERROR(Service_FS, "failed to get a handle for file %s", file_path.DebugStr().c_str()); + LOG_ERROR(Service_FS, "failed to get a handle for file %s mode=%u attributes=%d", + file_path.DebugStr().c_str(), mode.hex, attributes); } } @@ -347,7 +349,8 @@ static void OpenDirectory(Service::Interface* self) { if (dir_res.Succeeded()) { cmd_buff[3] = Kernel::g_handle_table.Create(*dir_res).MoveFrom(); } else { - LOG_ERROR(Service_FS, "failed to get a handle for directory"); + LOG_ERROR(Service_FS, "failed to get a handle for directory type=%d size=%d data=%s", + dirname_type, dirname_size, dir_path.DebugStr().c_str()); } } @@ -382,7 +385,8 @@ static void OpenArchive(Service::Interface* self) { cmd_buff[3] = (*handle >> 32) & 0xFFFFFFFF; } else { cmd_buff[2] = cmd_buff[3] = 0; - LOG_ERROR(Service_FS, "failed to get a handle for archive"); + LOG_ERROR(Service_FS, "failed to get a handle for archive archive_id=0x%08X archive_path=%s", + archive_id, archive_path.DebugStr().c_str()); } } diff --git a/src/core/hle/service/gsp_gpu.cpp b/src/core/hle/service/gsp_gpu.cpp index 8b40ba37..e93c1b43 100644 --- a/src/core/hle/service/gsp_gpu.cpp +++ b/src/core/hle/service/gsp_gpu.cpp @@ -14,6 +14,7 @@ #include "core/hw/lcd.h" #include "video_core/gpu_debugger.h" +#include "video_core/debug_utils/debug_utils.h" #include "video_core/renderer_base.h" #include "video_core/video_core.h" @@ -226,6 +227,9 @@ void SetBufferSwap(u32 screen_id, const FrameBufferInfo& info) { &info.format); WriteHWRegs(base_address + 4 * static_cast<u32>(GPU_REG_INDEX(framebuffer_config[screen_id].active_fb)), 4, &info.shown_fb); + + if (Pica::g_debug_context) + Pica::g_debug_context->OnEvent(Pica::DebugContext::Event::BufferSwapped, nullptr); } /** @@ -391,19 +395,24 @@ static void ExecuteCommand(const Command& command, u32 thread_id) { case CommandId::SET_MEMORY_FILL: { auto& params = command.memory_fill; - WriteGPURegister(static_cast<u32>(GPU_REG_INDEX(memory_fill_config[0].address_start)), - Memory::VirtualToPhysicalAddress(params.start1) >> 3); - WriteGPURegister(static_cast<u32>(GPU_REG_INDEX(memory_fill_config[0].address_end)), - Memory::VirtualToPhysicalAddress(params.end1) >> 3); - WriteGPURegister(static_cast<u32>(GPU_REG_INDEX(memory_fill_config[0].value_32bit)), params.value1); - WriteGPURegister(static_cast<u32>(GPU_REG_INDEX(memory_fill_config[0].control)), params.control1); - - WriteGPURegister(static_cast<u32>(GPU_REG_INDEX(memory_fill_config[1].address_start)), - Memory::VirtualToPhysicalAddress(params.start2) >> 3); - WriteGPURegister(static_cast<u32>(GPU_REG_INDEX(memory_fill_config[1].address_end)), - Memory::VirtualToPhysicalAddress(params.end2) >> 3); - WriteGPURegister(static_cast<u32>(GPU_REG_INDEX(memory_fill_config[1].value_32bit)), params.value2); - WriteGPURegister(static_cast<u32>(GPU_REG_INDEX(memory_fill_config[1].control)), params.control2); + + if (params.start1 != 0) { + WriteGPURegister(static_cast<u32>(GPU_REG_INDEX(memory_fill_config[0].address_start)), + Memory::VirtualToPhysicalAddress(params.start1) >> 3); + WriteGPURegister(static_cast<u32>(GPU_REG_INDEX(memory_fill_config[0].address_end)), + Memory::VirtualToPhysicalAddress(params.end1) >> 3); + WriteGPURegister(static_cast<u32>(GPU_REG_INDEX(memory_fill_config[0].value_32bit)), params.value1); + WriteGPURegister(static_cast<u32>(GPU_REG_INDEX(memory_fill_config[0].control)), params.control1); + } + + if (params.start2 != 0) { + WriteGPURegister(static_cast<u32>(GPU_REG_INDEX(memory_fill_config[1].address_start)), + Memory::VirtualToPhysicalAddress(params.start2) >> 3); + WriteGPURegister(static_cast<u32>(GPU_REG_INDEX(memory_fill_config[1].address_end)), + Memory::VirtualToPhysicalAddress(params.end2) >> 3); + WriteGPURegister(static_cast<u32>(GPU_REG_INDEX(memory_fill_config[1].value_32bit)), params.value2); + WriteGPURegister(static_cast<u32>(GPU_REG_INDEX(memory_fill_config[1].control)), params.control2); + } break; } @@ -448,6 +457,9 @@ static void ExecuteCommand(const Command& command, u32 thread_id) { default: LOG_ERROR(Service_GSP, "unknown command 0x%08X", (int)command.id.Value()); } + + if (Pica::g_debug_context) + Pica::g_debug_context->OnEvent(Pica::DebugContext::Event::GSPCommandProcessed, (void*)&command); } /** diff --git a/src/core/hle/service/service.cpp b/src/core/hle/service/service.cpp index d681cc3d..0de0b13a 100644 --- a/src/core/hle/service/service.cpp +++ b/src/core/hle/service/service.cpp @@ -54,7 +54,7 @@ static std::string MakeFunctionString(const char* name, const char* port_name, c std::string function_string = Common::StringFromFormat("function '%s': port=%s", name, port_name); for (int i = 1; i <= num_params; ++i) { - function_string += Common::StringFromFormat(", cmd_buff[%i]=%u", i, cmd_buff[i]); + function_string += Common::StringFromFormat(", cmd_buff[%i]=0x%X", i, cmd_buff[i]); } return function_string; } diff --git a/src/core/hle/svc.cpp b/src/core/hle/svc.cpp index 9d441ccf..bdede964 100644 --- a/src/core/hle/svc.cpp +++ b/src/core/hle/svc.cpp @@ -530,11 +530,16 @@ static ResultCode ReleaseSemaphore(s32* count, Handle handle, s32 release_count) return RESULT_SUCCESS; } -/// Query memory -static ResultCode QueryMemory(MemoryInfo* memory_info, PageInfo* page_info, u32 addr) { - auto vma = Kernel::g_current_process->address_space->FindVMA(addr); +/// Query process memory +static ResultCode QueryProcessMemory(MemoryInfo* memory_info, PageInfo* page_info, Handle process_handle, u32 addr) { + using Kernel::Process; + Kernel::SharedPtr<Process> process = Kernel::g_handle_table.Get<Process>(process_handle); + if (process == nullptr) + return ERR_INVALID_HANDLE; + + auto vma = process->address_space->FindVMA(addr); - if (vma == Kernel::g_current_process->address_space->vma_map.end()) + if (vma == process->address_space->vma_map.end()) return ResultCode(ErrorDescription::InvalidAddress, ErrorModule::OS, ErrorSummary::InvalidArgument, ErrorLevel::Usage); memory_info->base_address = vma->second.base; @@ -543,10 +548,15 @@ static ResultCode QueryMemory(MemoryInfo* memory_info, PageInfo* page_info, u32 memory_info->state = static_cast<u32>(vma->second.meminfo_state); page_info->flags = 0; - LOG_TRACE(Kernel_SVC, "called addr=0x%08X", addr); + LOG_TRACE(Kernel_SVC, "called process=0x%08X addr=0x%08X", process_handle, addr); return RESULT_SUCCESS; } +/// Query memory +static ResultCode QueryMemory(MemoryInfo* memory_info, PageInfo* page_info, u32 addr) { + return QueryProcessMemory(memory_info, page_info, Kernel::CurrentProcess, addr); +} + /// Create an event static ResultCode CreateEvent(Handle* out_handle, u32 reset_type) { using Kernel::Event; @@ -818,13 +828,12 @@ static const FunctionDef SVC_Table[] = { {0x7A, nullptr, "AddCodeSegment"}, {0x7B, nullptr, "Backdoor"}, {0x7C, nullptr, "KernelSetState"}, - {0x7D, nullptr, "QueryProcessMemory"}, + {0x7D, HLE::Wrap<QueryProcessMemory>, "QueryProcessMemory"}, }; Common::Profiling::TimingCategory profiler_svc("SVC Calls"); -static const FunctionDef* GetSVCInfo(u32 opcode) { - u32 func_num = opcode & 0xFFFFFF; // 8 bits +static const FunctionDef* GetSVCInfo(u32 func_num) { if (func_num >= ARRAY_SIZE(SVC_Table)) { LOG_ERROR(Kernel_SVC, "unknown svc=0x%02X", func_num); return nullptr; @@ -832,10 +841,10 @@ static const FunctionDef* GetSVCInfo(u32 opcode) { return &SVC_Table[func_num]; } -void CallSVC(u32 opcode) { +void CallSVC(u32 immediate) { Common::Profiling::ScopeTimer timer_svc(profiler_svc); - const FunctionDef *info = GetSVCInfo(opcode); + const FunctionDef* info = GetSVCInfo(immediate); if (info) { if (info->func) { info->func(); diff --git a/src/core/hle/svc.h b/src/core/hle/svc.h index 4389aa73..12de9ffb 100644 --- a/src/core/hle/svc.h +++ b/src/core/hle/svc.h @@ -41,6 +41,6 @@ enum ArbitrationType { namespace SVC { -void CallSVC(u32 opcode); +void CallSVC(u32 immediate); } // namespace diff --git a/src/core/hw/gpu.cpp b/src/core/hw/gpu.cpp index a3a7d128..3ccbc03b 100644 --- a/src/core/hw/gpu.cpp +++ b/src/core/hw/gpu.cpp @@ -151,6 +151,10 @@ inline void Write(u32 addr, const T data) { { const auto& config = g_regs.display_transfer_config; if (config.trigger & 1) { + + if (Pica::g_debug_context) + Pica::g_debug_context->OnEvent(Pica::DebugContext::Event::IncomingDisplayTransfer, nullptr); + u8* src_pointer = Memory::GetPhysicalPointer(config.GetPhysicalInputAddress()); u8* dst_pointer = Memory::GetPhysicalPointer(config.GetPhysicalOutputAddress()); @@ -217,19 +221,37 @@ inline void Write(u32 addr, const T data) { u32 dst_offset; if (config.output_tiled) { - // Interpret the input as linear and the output as tiled - u32 coarse_y = y & ~7; - u32 stride = output_width * dst_bytes_per_pixel; - - src_offset = (input_x + input_y * config.input_width) * src_bytes_per_pixel; - dst_offset = VideoCore::GetMortonOffset(x, y, dst_bytes_per_pixel) + coarse_y * stride; + if (!config.dont_swizzle) { + // Interpret the input as linear and the output as tiled + u32 coarse_y = y & ~7; + u32 stride = output_width * dst_bytes_per_pixel; + + src_offset = (input_x + input_y * config.input_width) * src_bytes_per_pixel; + dst_offset = VideoCore::GetMortonOffset(x, y, dst_bytes_per_pixel) + coarse_y * stride; + } else { + // Both input and output are linear + src_offset = (input_x + input_y * config.input_width) * src_bytes_per_pixel; + dst_offset = (x + y * output_width) * dst_bytes_per_pixel; + } } else { - // Interpret the input as tiled and the output as linear - u32 coarse_y = input_y & ~7; - u32 stride = config.input_width * src_bytes_per_pixel; - - src_offset = VideoCore::GetMortonOffset(input_x, input_y, src_bytes_per_pixel) + coarse_y * stride; - dst_offset = (x + y * output_width) * dst_bytes_per_pixel; + if (!config.dont_swizzle) { + // Interpret the input as tiled and the output as linear + u32 coarse_y = input_y & ~7; + u32 stride = config.input_width * src_bytes_per_pixel; + + src_offset = VideoCore::GetMortonOffset(input_x, input_y, src_bytes_per_pixel) + coarse_y * stride; + dst_offset = (x + y * output_width) * dst_bytes_per_pixel; + } else { + // Both input and output are tiled + u32 out_coarse_y = y & ~7; + u32 out_stride = output_width * dst_bytes_per_pixel; + + u32 in_coarse_y = input_y & ~7; + u32 in_stride = config.input_width * src_bytes_per_pixel; + + src_offset = VideoCore::GetMortonOffset(input_x, input_y, src_bytes_per_pixel) + in_coarse_y * in_stride; + dst_offset = VideoCore::GetMortonOffset(x, y, dst_bytes_per_pixel) + out_coarse_y * out_stride; + } } const u8* src_pixel = src_pointer + src_offset; diff --git a/src/core/hw/gpu.h b/src/core/hw/gpu.h index 5b8c43f8..daad506f 100644 --- a/src/core/hw/gpu.h +++ b/src/core/hw/gpu.h @@ -203,6 +203,7 @@ struct Regs { BitField< 0, 1, u32> flip_vertically; // flips input data vertically BitField< 1, 1, u32> output_tiled; // Converts from linear to tiled format BitField< 3, 1, u32> raw_copy; // Copies the data without performing any processing + BitField< 5, 1, u32> dont_swizzle; BitField< 8, 3, PixelFormat> input_format; BitField<12, 3, PixelFormat> output_format; diff --git a/src/video_core/command_processor.cpp b/src/video_core/command_processor.cpp index bfb00019..43ae0618 100644 --- a/src/video_core/command_processor.cpp +++ b/src/video_core/command_processor.cpp @@ -45,12 +45,12 @@ static inline void WritePicaReg(u32 id, u32 value, u32 mask) { if (GPU::g_skip_frame && id != PICA_REG_INDEX(trigger_irq)) return; - // TODO: Figure out how register masking acts on e.g. vs_uniform_setup.set_value + // TODO: Figure out how register masking acts on e.g. vs.uniform_setup.set_value u32 old_value = regs[id]; regs[id] = (old_value & ~mask) | (value & mask); if (g_debug_context) - g_debug_context->OnEvent(DebugContext::Event::CommandLoaded, reinterpret_cast<void*>(&id)); + g_debug_context->OnEvent(DebugContext::Event::PicaCommandLoaded, reinterpret_cast<void*>(&id)); DebugUtils::OnPicaRegWrite(id, regs[id]); @@ -282,7 +282,7 @@ static inline void WritePicaReg(u32 id, u32 value, u32 mask) { &geometry_dumper, _1, _2, _3)); // Send to vertex shader - VertexShader::OutputVertex output = VertexShader::RunShader(input, attribute_config.GetNumTotalAttributes()); + VertexShader::OutputVertex output = VertexShader::RunShader(input, attribute_config.GetNumTotalAttributes(), g_state.regs.vs, g_state.vs); if (is_indexed) { // TODO: Add processed vertex to vertex cache! @@ -321,35 +321,35 @@ static inline void WritePicaReg(u32 id, u32 value, u32 mask) { break; } - case PICA_REG_INDEX(vs_bool_uniforms): + case PICA_REG_INDEX(vs.bool_uniforms): for (unsigned i = 0; i < 16; ++i) - g_state.vs.uniforms.b[i] = (regs.vs_bool_uniforms.Value() & (1 << i)) != 0; + g_state.vs.uniforms.b[i] = (regs.vs.bool_uniforms.Value() & (1 << i)) != 0; break; - case PICA_REG_INDEX_WORKAROUND(vs_int_uniforms[0], 0x2b1): - case PICA_REG_INDEX_WORKAROUND(vs_int_uniforms[1], 0x2b2): - case PICA_REG_INDEX_WORKAROUND(vs_int_uniforms[2], 0x2b3): - case PICA_REG_INDEX_WORKAROUND(vs_int_uniforms[3], 0x2b4): + case PICA_REG_INDEX_WORKAROUND(vs.int_uniforms[0], 0x2b1): + case PICA_REG_INDEX_WORKAROUND(vs.int_uniforms[1], 0x2b2): + case PICA_REG_INDEX_WORKAROUND(vs.int_uniforms[2], 0x2b3): + case PICA_REG_INDEX_WORKAROUND(vs.int_uniforms[3], 0x2b4): { - int index = (id - PICA_REG_INDEX_WORKAROUND(vs_int_uniforms[0], 0x2b1)); - auto values = regs.vs_int_uniforms[index]; + int index = (id - PICA_REG_INDEX_WORKAROUND(vs.int_uniforms[0], 0x2b1)); + auto values = regs.vs.int_uniforms[index]; g_state.vs.uniforms.i[index] = Math::Vec4<u8>(values.x, values.y, values.z, values.w); LOG_TRACE(HW_GPU, "Set integer uniform %d to %02x %02x %02x %02x", index, values.x.Value(), values.y.Value(), values.z.Value(), values.w.Value()); break; } - case PICA_REG_INDEX_WORKAROUND(vs_uniform_setup.set_value[0], 0x2c1): - case PICA_REG_INDEX_WORKAROUND(vs_uniform_setup.set_value[1], 0x2c2): - case PICA_REG_INDEX_WORKAROUND(vs_uniform_setup.set_value[2], 0x2c3): - case PICA_REG_INDEX_WORKAROUND(vs_uniform_setup.set_value[3], 0x2c4): - case PICA_REG_INDEX_WORKAROUND(vs_uniform_setup.set_value[4], 0x2c5): - case PICA_REG_INDEX_WORKAROUND(vs_uniform_setup.set_value[5], 0x2c6): - case PICA_REG_INDEX_WORKAROUND(vs_uniform_setup.set_value[6], 0x2c7): - case PICA_REG_INDEX_WORKAROUND(vs_uniform_setup.set_value[7], 0x2c8): + case PICA_REG_INDEX_WORKAROUND(vs.uniform_setup.set_value[0], 0x2c1): + case PICA_REG_INDEX_WORKAROUND(vs.uniform_setup.set_value[1], 0x2c2): + case PICA_REG_INDEX_WORKAROUND(vs.uniform_setup.set_value[2], 0x2c3): + case PICA_REG_INDEX_WORKAROUND(vs.uniform_setup.set_value[3], 0x2c4): + case PICA_REG_INDEX_WORKAROUND(vs.uniform_setup.set_value[4], 0x2c5): + case PICA_REG_INDEX_WORKAROUND(vs.uniform_setup.set_value[5], 0x2c6): + case PICA_REG_INDEX_WORKAROUND(vs.uniform_setup.set_value[6], 0x2c7): + case PICA_REG_INDEX_WORKAROUND(vs.uniform_setup.set_value[7], 0x2c8): { - auto& uniform_setup = regs.vs_uniform_setup; + auto& uniform_setup = regs.vs.uniform_setup; // TODO: Does actual hardware indeed keep an intermediate buffer or does // it directly write the values? @@ -392,32 +392,32 @@ static inline void WritePicaReg(u32 id, u32 value, u32 mask) { } // Load shader program code - case PICA_REG_INDEX_WORKAROUND(vs_program.set_word[0], 0x2cc): - case PICA_REG_INDEX_WORKAROUND(vs_program.set_word[1], 0x2cd): - case PICA_REG_INDEX_WORKAROUND(vs_program.set_word[2], 0x2ce): - case PICA_REG_INDEX_WORKAROUND(vs_program.set_word[3], 0x2cf): - case PICA_REG_INDEX_WORKAROUND(vs_program.set_word[4], 0x2d0): - case PICA_REG_INDEX_WORKAROUND(vs_program.set_word[5], 0x2d1): - case PICA_REG_INDEX_WORKAROUND(vs_program.set_word[6], 0x2d2): - case PICA_REG_INDEX_WORKAROUND(vs_program.set_word[7], 0x2d3): + case PICA_REG_INDEX_WORKAROUND(vs.program.set_word[0], 0x2cc): + case PICA_REG_INDEX_WORKAROUND(vs.program.set_word[1], 0x2cd): + case PICA_REG_INDEX_WORKAROUND(vs.program.set_word[2], 0x2ce): + case PICA_REG_INDEX_WORKAROUND(vs.program.set_word[3], 0x2cf): + case PICA_REG_INDEX_WORKAROUND(vs.program.set_word[4], 0x2d0): + case PICA_REG_INDEX_WORKAROUND(vs.program.set_word[5], 0x2d1): + case PICA_REG_INDEX_WORKAROUND(vs.program.set_word[6], 0x2d2): + case PICA_REG_INDEX_WORKAROUND(vs.program.set_word[7], 0x2d3): { - g_state.vs.program_code[regs.vs_program.offset] = value; - regs.vs_program.offset++; + g_state.vs.program_code[regs.vs.program.offset] = value; + regs.vs.program.offset++; break; } // Load swizzle pattern data - case PICA_REG_INDEX_WORKAROUND(vs_swizzle_patterns.set_word[0], 0x2d6): - case PICA_REG_INDEX_WORKAROUND(vs_swizzle_patterns.set_word[1], 0x2d7): - case PICA_REG_INDEX_WORKAROUND(vs_swizzle_patterns.set_word[2], 0x2d8): - case PICA_REG_INDEX_WORKAROUND(vs_swizzle_patterns.set_word[3], 0x2d9): - case PICA_REG_INDEX_WORKAROUND(vs_swizzle_patterns.set_word[4], 0x2da): - case PICA_REG_INDEX_WORKAROUND(vs_swizzle_patterns.set_word[5], 0x2db): - case PICA_REG_INDEX_WORKAROUND(vs_swizzle_patterns.set_word[6], 0x2dc): - case PICA_REG_INDEX_WORKAROUND(vs_swizzle_patterns.set_word[7], 0x2dd): + case PICA_REG_INDEX_WORKAROUND(vs.swizzle_patterns.set_word[0], 0x2d6): + case PICA_REG_INDEX_WORKAROUND(vs.swizzle_patterns.set_word[1], 0x2d7): + case PICA_REG_INDEX_WORKAROUND(vs.swizzle_patterns.set_word[2], 0x2d8): + case PICA_REG_INDEX_WORKAROUND(vs.swizzle_patterns.set_word[3], 0x2d9): + case PICA_REG_INDEX_WORKAROUND(vs.swizzle_patterns.set_word[4], 0x2da): + case PICA_REG_INDEX_WORKAROUND(vs.swizzle_patterns.set_word[5], 0x2db): + case PICA_REG_INDEX_WORKAROUND(vs.swizzle_patterns.set_word[6], 0x2dc): + case PICA_REG_INDEX_WORKAROUND(vs.swizzle_patterns.set_word[7], 0x2dd): { - g_state.vs.swizzle_data[regs.vs_swizzle_patterns.offset] = value; - regs.vs_swizzle_patterns.offset++; + g_state.vs.swizzle_data[regs.vs.swizzle_patterns.offset] = value; + regs.vs.swizzle_patterns.offset++; break; } @@ -428,7 +428,7 @@ static inline void WritePicaReg(u32 id, u32 value, u32 mask) { VideoCore::g_renderer->hw_rasterizer->NotifyPicaRegisterChanged(id); if (g_debug_context) - g_debug_context->OnEvent(DebugContext::Event::CommandProcessed, reinterpret_cast<void*>(&id)); + g_debug_context->OnEvent(DebugContext::Event::PicaCommandProcessed, reinterpret_cast<void*>(&id)); } void ProcessCommandList(const u32* list, u32 size) { diff --git a/src/video_core/debug_utils/debug_utils.h b/src/video_core/debug_utils/debug_utils.h index 2573292e..3f109dcb 100644 --- a/src/video_core/debug_utils/debug_utils.h +++ b/src/video_core/debug_utils/debug_utils.h @@ -25,11 +25,14 @@ public: enum class Event { FirstEvent = 0, - CommandLoaded = FirstEvent, - CommandProcessed, + PicaCommandLoaded = FirstEvent, + PicaCommandProcessed, IncomingPrimitiveBatch, FinishedPrimitiveBatch, VertexLoaded, + IncomingDisplayTransfer, + GSPCommandProcessed, + BufferSwapped, NumEvents }; diff --git a/src/video_core/pica.h b/src/video_core/pica.h index 2b7ad206..38599a7a 100644 --- a/src/video_core/pica.h +++ b/src/video_core/pica.h @@ -119,6 +119,11 @@ struct Regs { MirroredRepeat = 3, }; + enum TextureFilter : u32 { + Nearest = 0, + Linear = 1 + }; + union { BitField< 0, 8, u32> r; BitField< 8, 8, u32> g; @@ -132,8 +137,10 @@ struct Regs { }; union { - BitField< 8, 2, WrapMode> wrap_s; - BitField<12, 2, WrapMode> wrap_t; + BitField< 1, 1, TextureFilter> mag_filter; + BitField< 2, 1, TextureFilter> min_filter; + BitField< 8, 2, WrapMode> wrap_t; + BitField<12, 2, WrapMode> wrap_s; }; INSERT_PADDING_WORDS(0x1); @@ -200,6 +207,7 @@ struct Regs { case TextureFormat::IA8: return 4; + case TextureFormat::I4: case TextureFormat::A4: return 1; @@ -784,112 +792,119 @@ struct Regs { INSERT_PADDING_WORDS(0x20); enum class TriangleTopology : u32 { - List = 0, - Strip = 1, - Fan = 2, - ListIndexed = 3, // TODO: No idea if this is correct + List = 0, + Strip = 1, + Fan = 2, + Shader = 3, // Programmable setup unit implemented in a geometry shader }; BitField<8, 2, TriangleTopology> triangle_topology; - INSERT_PADDING_WORDS(0x51); + INSERT_PADDING_WORDS(0x21); - BitField<0, 16, u32> vs_bool_uniforms; - union { - BitField< 0, 8, u32> x; - BitField< 8, 8, u32> y; - BitField<16, 8, u32> z; - BitField<24, 8, u32> w; - } vs_int_uniforms[4]; + struct ShaderConfig { + BitField<0, 16, u32> bool_uniforms; - INSERT_PADDING_WORDS(0x5); + union { + BitField< 0, 8, u32> x; + BitField< 8, 8, u32> y; + BitField<16, 8, u32> z; + BitField<24, 8, u32> w; + } int_uniforms[4]; - // Offset to shader program entry point (in words) - BitField<0, 16, u32> vs_main_offset; + INSERT_PADDING_WORDS(0x5); - union { - BitField< 0, 4, u64> attribute0_register; - BitField< 4, 4, u64> attribute1_register; - BitField< 8, 4, u64> attribute2_register; - BitField<12, 4, u64> attribute3_register; - BitField<16, 4, u64> attribute4_register; - BitField<20, 4, u64> attribute5_register; - BitField<24, 4, u64> attribute6_register; - BitField<28, 4, u64> attribute7_register; - BitField<32, 4, u64> attribute8_register; - BitField<36, 4, u64> attribute9_register; - BitField<40, 4, u64> attribute10_register; - BitField<44, 4, u64> attribute11_register; - BitField<48, 4, u64> attribute12_register; - BitField<52, 4, u64> attribute13_register; - BitField<56, 4, u64> attribute14_register; - BitField<60, 4, u64> attribute15_register; - - int GetRegisterForAttribute(int attribute_index) const { - u64 fields[] = { - attribute0_register, attribute1_register, attribute2_register, attribute3_register, - attribute4_register, attribute5_register, attribute6_register, attribute7_register, - attribute8_register, attribute9_register, attribute10_register, attribute11_register, - attribute12_register, attribute13_register, attribute14_register, attribute15_register, + // Offset to shader program entry point (in words) + BitField<0, 16, u32> main_offset; + + union { + BitField< 0, 4, u64> attribute0_register; + BitField< 4, 4, u64> attribute1_register; + BitField< 8, 4, u64> attribute2_register; + BitField<12, 4, u64> attribute3_register; + BitField<16, 4, u64> attribute4_register; + BitField<20, 4, u64> attribute5_register; + BitField<24, 4, u64> attribute6_register; + BitField<28, 4, u64> attribute7_register; + BitField<32, 4, u64> attribute8_register; + BitField<36, 4, u64> attribute9_register; + BitField<40, 4, u64> attribute10_register; + BitField<44, 4, u64> attribute11_register; + BitField<48, 4, u64> attribute12_register; + BitField<52, 4, u64> attribute13_register; + BitField<56, 4, u64> attribute14_register; + BitField<60, 4, u64> attribute15_register; + + int GetRegisterForAttribute(int attribute_index) const { + u64 fields[] = { + attribute0_register, attribute1_register, attribute2_register, attribute3_register, + attribute4_register, attribute5_register, attribute6_register, attribute7_register, + attribute8_register, attribute9_register, attribute10_register, attribute11_register, + attribute12_register, attribute13_register, attribute14_register, attribute15_register, + }; + return (int)fields[attribute_index]; + } + } input_register_map; + + // OUTMAP_MASK, 0x28E, CODETRANSFER_END + INSERT_PADDING_WORDS(0x3); + + struct { + enum Format : u32 + { + FLOAT24 = 0, + FLOAT32 = 1 }; - return (int)fields[attribute_index]; - } - } vs_input_register_map; - INSERT_PADDING_WORDS(0x3); + bool IsFloat32() const { + return format == FLOAT32; + } - struct { - enum Format : u32 - { - FLOAT24 = 0, - FLOAT32 = 1 - }; + union { + // Index of the next uniform to write to + // TODO: ctrulib uses 8 bits for this, however that seems to yield lots of invalid indices + // TODO: Maybe the uppermost index is for the geometry shader? Investigate! + BitField<0, 7, u32> index; - bool IsFloat32() const { - return format == FLOAT32; - } + BitField<31, 1, Format> format; + }; - union { - // Index of the next uniform to write to - // TODO: ctrulib uses 8 bits for this, however that seems to yield lots of invalid indices - BitField<0, 7, u32> index; + // Writing to these registers sets the current uniform. + u32 set_value[8]; - BitField<31, 1, Format> format; - }; + } uniform_setup; - // Writing to these registers sets the "current" uniform. - // TODO: It's not clear how the hardware stores what the "current" uniform is. - u32 set_value[8]; + INSERT_PADDING_WORDS(0x2); - } vs_uniform_setup; + struct { + // Offset of the next instruction to write code to. + // Incremented with each instruction write. + u32 offset; - INSERT_PADDING_WORDS(0x2); + // Writing to these registers sets the "current" word in the shader program. + u32 set_word[8]; + } program; - struct { - // Offset of the next instruction to write code to. - // Incremented with each instruction write. - u32 offset; + INSERT_PADDING_WORDS(0x1); - // Writing to these registers sets the "current" word in the shader program. - // TODO: It's not clear how the hardware stores what the "current" word is. - u32 set_word[8]; - } vs_program; + // This register group is used to load an internal table of swizzling patterns, + // which are indexed by each shader instruction to specify vector component swizzling. + struct { + // Offset of the next swizzle pattern to write code to. + // Incremented with each instruction write. + u32 offset; - INSERT_PADDING_WORDS(0x1); + // Writing to these registers sets the current swizzle pattern in the table. + u32 set_word[8]; + } swizzle_patterns; - // This register group is used to load an internal table of swizzling patterns, - // which are indexed by each shader instruction to specify vector component swizzling. - struct { - // Offset of the next swizzle pattern to write code to. - // Incremented with each instruction write. - u32 offset; + INSERT_PADDING_WORDS(0x2); + }; - // Writing to these registers sets the "current" swizzle pattern in the table. - // TODO: It's not clear how the hardware stores what the "current" swizzle pattern is. - u32 set_word[8]; - } vs_swizzle_patterns; + ShaderConfig gs; + ShaderConfig vs; - INSERT_PADDING_WORDS(0x22); + INSERT_PADDING_WORDS(0x20); // Map register indices to names readable by humans // Used for debugging purposes, so performance is not an issue here @@ -936,13 +951,20 @@ struct Regs { ADD_FIELD(vs_default_attributes_setup); ADD_FIELD(command_buffer); ADD_FIELD(triangle_topology); - ADD_FIELD(vs_bool_uniforms); - ADD_FIELD(vs_int_uniforms); - ADD_FIELD(vs_main_offset); - ADD_FIELD(vs_input_register_map); - ADD_FIELD(vs_uniform_setup); - ADD_FIELD(vs_program); - ADD_FIELD(vs_swizzle_patterns); + ADD_FIELD(gs.bool_uniforms); + ADD_FIELD(gs.int_uniforms); + ADD_FIELD(gs.main_offset); + ADD_FIELD(gs.input_register_map); + ADD_FIELD(gs.uniform_setup); + ADD_FIELD(gs.program); + ADD_FIELD(gs.swizzle_patterns); + ADD_FIELD(vs.bool_uniforms); + ADD_FIELD(vs.int_uniforms); + ADD_FIELD(vs.main_offset); + ADD_FIELD(vs.input_register_map); + ADD_FIELD(vs.uniform_setup); + ADD_FIELD(vs.program); + ADD_FIELD(vs.swizzle_patterns); #undef ADD_FIELD @@ -1014,17 +1036,14 @@ ASSERT_REG_POSITION(trigger_draw_indexed, 0x22f); ASSERT_REG_POSITION(vs_default_attributes_setup, 0x232); ASSERT_REG_POSITION(command_buffer, 0x238); ASSERT_REG_POSITION(triangle_topology, 0x25e); -ASSERT_REG_POSITION(vs_bool_uniforms, 0x2b0); -ASSERT_REG_POSITION(vs_int_uniforms, 0x2b1); -ASSERT_REG_POSITION(vs_main_offset, 0x2ba); -ASSERT_REG_POSITION(vs_input_register_map, 0x2bb); -ASSERT_REG_POSITION(vs_uniform_setup, 0x2c0); -ASSERT_REG_POSITION(vs_program, 0x2cb); -ASSERT_REG_POSITION(vs_swizzle_patterns, 0x2d5); +ASSERT_REG_POSITION(gs, 0x280); +ASSERT_REG_POSITION(vs, 0x2b0); #undef ASSERT_REG_POSITION #endif // !defined(_MSC_VER) +static_assert(sizeof(Regs::ShaderConfig) == 0x30 * sizeof(u32), "ShaderConfig structure has incorrect size"); + // The total number of registers is chosen arbitrarily, but let's make sure it's not some odd value anyway. static_assert(sizeof(Regs) <= 0x300 * sizeof(u32), "Register set structure larger than it should be"); static_assert(sizeof(Regs) >= 0x300 * sizeof(u32), "Register set structure smaller than it should be"); @@ -1134,7 +1153,7 @@ struct State { Regs regs; /// Vertex shader memory - struct { + struct ShaderSetup { struct { Math::Vec4<float24> f[96]; std::array<bool, 16> b; @@ -1145,7 +1164,10 @@ struct State { std::array<u32, 1024> program_code; std::array<u32, 1024> swizzle_data; - } vs; + }; + + ShaderSetup vs; + ShaderSetup gs; /// Current Pica command list struct { diff --git a/src/video_core/primitive_assembly.cpp b/src/video_core/primitive_assembly.cpp index 0120f289..2f22bdcc 100644 --- a/src/video_core/primitive_assembly.cpp +++ b/src/video_core/primitive_assembly.cpp @@ -20,8 +20,9 @@ template<typename VertexType> void PrimitiveAssembler<VertexType>::SubmitVertex(VertexType& vtx, TriangleHandler triangle_handler) { switch (topology) { + // TODO: Figure out what's different with TriangleTopology::Shader. case Regs::TriangleTopology::List: - case Regs::TriangleTopology::ListIndexed: + case Regs::TriangleTopology::Shader: if (buffer_index < 2) { buffer[buffer_index++] = vtx; } else { diff --git a/src/video_core/rasterizer.cpp b/src/video_core/rasterizer.cpp index a6b7997c..e2b90ad1 100644 --- a/src/video_core/rasterizer.cpp +++ b/src/video_core/rasterizer.cpp @@ -460,6 +460,7 @@ static void ProcessTriangleInternal(const VertexShader::OutputVertex& v0, u8* texture_data = Memory::GetPhysicalPointer(texture.config.GetPhysicalAddress()); auto info = DebugUtils::TextureInfo::FromPicaRegister(texture.config, texture.format); + // TODO: Apply the min and mag filters to the texture texture_color[i] = DebugUtils::LookupTexture(texture_data, s, t, info); DebugUtils::DumpTexture(texture.config, texture_data); } diff --git a/src/video_core/renderer_opengl/gl_rasterizer.cpp b/src/video_core/renderer_opengl/gl_rasterizer.cpp index 935a9f28..2db845da 100644 --- a/src/video_core/renderer_opengl/gl_rasterizer.cpp +++ b/src/video_core/renderer_opengl/gl_rasterizer.cpp @@ -816,12 +816,16 @@ void RasterizerOpenGL::ReloadColorBuffer() { } void RasterizerOpenGL::ReloadDepthBuffer() { + PAddr depth_buffer_addr = Pica::g_state.regs.framebuffer.GetDepthBufferPhysicalAddress(); + + if (depth_buffer_addr == 0) + return; + // TODO: Appears to work, but double-check endianness of depth values and order of depth-stencil - u8* depth_buffer = Memory::GetPhysicalPointer(Pica::g_state.regs.framebuffer.GetDepthBufferPhysicalAddress()); + u8* depth_buffer = Memory::GetPhysicalPointer(depth_buffer_addr); - if (depth_buffer == nullptr) { + if (depth_buffer == nullptr) return; - } u32 bytes_per_pixel = Pica::Regs::BytesPerDepthPixel(fb_depth_texture.format); diff --git a/src/video_core/renderer_opengl/gl_rasterizer_cache.cpp b/src/video_core/renderer_opengl/gl_rasterizer_cache.cpp index 2e4110a8..dc3ffdf2 100644 --- a/src/video_core/renderer_opengl/gl_rasterizer_cache.cpp +++ b/src/video_core/renderer_opengl/gl_rasterizer_cache.cpp @@ -31,12 +31,18 @@ void RasterizerCacheOpenGL::LoadAndBindTexture(OpenGLState &state, unsigned text state.texture_units[texture_unit].texture_2d = new_texture->texture.handle; state.Apply(); - // TODO: Need to choose filters that correspond to PICA once register is declared - glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MAG_FILTER, GL_LINEAR); - glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MIN_FILTER, GL_LINEAR); + glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MAG_FILTER, PicaToGL::TextureFilterMode(config.config.mag_filter)); + glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MIN_FILTER, PicaToGL::TextureFilterMode(config.config.min_filter)); - glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_WRAP_S, PicaToGL::WrapMode(config.config.wrap_s)); - glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_WRAP_T, PicaToGL::WrapMode(config.config.wrap_t)); + GLenum wrap_s = PicaToGL::WrapMode(config.config.wrap_s); + GLenum wrap_t = PicaToGL::WrapMode(config.config.wrap_t); + glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_WRAP_S, wrap_s); + glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_WRAP_T, wrap_t); + + if (wrap_s == GL_CLAMP_TO_BORDER || wrap_t == GL_CLAMP_TO_BORDER) { + auto border_color = PicaToGL::ColorRGBA8((u8*)&config.config.border_color.r); + glTexParameterfv(GL_TEXTURE_2D, GL_TEXTURE_BORDER_COLOR, border_color.data()); + } const auto info = Pica::DebugUtils::TextureInfo::FromPicaRegister(config.config, config.format); diff --git a/src/video_core/renderer_opengl/gl_state.cpp b/src/video_core/renderer_opengl/gl_state.cpp index 3526e16d..9efc1533 100644 --- a/src/video_core/renderer_opengl/gl_state.cpp +++ b/src/video_core/renderer_opengl/gl_state.cpp @@ -147,20 +147,17 @@ void OpenGLState::Apply() { // Textures for (unsigned texture_index = 0; texture_index < ARRAY_SIZE(texture_units); ++texture_index) { - if (texture_units[texture_index].enabled_2d != cur_state.texture_units[texture_index].enabled_2d) { + if (texture_units[texture_index].enabled_2d != cur_state.texture_units[texture_index].enabled_2d || + texture_units[texture_index].texture_2d != cur_state.texture_units[texture_index].texture_2d) { + glActiveTexture(GL_TEXTURE0 + texture_index); if (texture_units[texture_index].enabled_2d) { - glEnable(GL_TEXTURE_2D); + glBindTexture(GL_TEXTURE_2D, texture_units[texture_index].texture_2d); } else { - glDisable(GL_TEXTURE_2D); + glBindTexture(GL_TEXTURE_2D, 0); } } - - if (texture_units[texture_index].texture_2d != cur_state.texture_units[texture_index].texture_2d) { - glActiveTexture(GL_TEXTURE0 + texture_index); - glBindTexture(GL_TEXTURE_2D, texture_units[texture_index].texture_2d); - } } // Framebuffer diff --git a/src/video_core/renderer_opengl/pica_to_gl.h b/src/video_core/renderer_opengl/pica_to_gl.h index 73f63c55..3b562da8 100644 --- a/src/video_core/renderer_opengl/pica_to_gl.h +++ b/src/video_core/renderer_opengl/pica_to_gl.h @@ -12,6 +12,33 @@ namespace PicaToGL { +inline GLenum TextureFilterMode(Pica::Regs::TextureConfig::TextureFilter mode) { + static const GLenum filter_mode_table[] = { + GL_NEAREST, // TextureFilter::Nearest + GL_LINEAR // TextureFilter::Linear + }; + + // Range check table for input + if (mode >= ARRAY_SIZE(filter_mode_table)) { + LOG_CRITICAL(Render_OpenGL, "Unknown texture filtering mode %d", mode); + UNREACHABLE(); + + return GL_LINEAR; + } + + GLenum gl_mode = filter_mode_table[mode]; + + // Check for dummy values indicating an unknown mode + if (gl_mode == 0) { + LOG_CRITICAL(Render_OpenGL, "Unknown texture filtering mode %d", mode); + UNIMPLEMENTED(); + + return GL_LINEAR; + } + + return gl_mode; +} + inline GLenum WrapMode(Pica::Regs::TextureConfig::WrapMode mode) { static const GLenum wrap_mode_table[] = { GL_CLAMP_TO_EDGE, // WrapMode::ClampToEdge diff --git a/src/video_core/vertex_shader.cpp b/src/video_core/vertex_shader.cpp index 87006a83..b7750380 100644 --- a/src/video_core/vertex_shader.cpp +++ b/src/video_core/vertex_shader.cpp @@ -221,7 +221,7 @@ static void ProcessShaderCode(VertexShaderState& state) { for (int i = 0; i < num_components; ++i) dot = dot + src1[i] * src2[i]; - for (int i = 0; i < num_components; ++i) { + for (int i = 0; i < 4; ++i) { if (!swizzle.DestComponentEnabled(i)) continue; @@ -546,20 +546,18 @@ static void ProcessShaderCode(VertexShaderState& state) { static Common::Profiling::TimingCategory shader_category("Vertex Shader"); -OutputVertex RunShader(const InputVertex& input, int num_attributes) { +OutputVertex RunShader(const InputVertex& input, int num_attributes, const Regs::ShaderConfig& config, const State::ShaderSetup& setup) { Common::Profiling::ScopeTimer timer(shader_category); - const auto& regs = g_state.regs; - const auto& vs = g_state.vs; VertexShaderState state; - const u32* main = &vs.program_code[regs.vs_main_offset]; + const u32* main = &setup.program_code[config.main_offset]; state.program_counter = (u32*)main; state.debug.max_offset = 0; state.debug.max_opdesc_id = 0; // Setup input register table - const auto& attribute_register_map = regs.vs_input_register_map; + const auto& attribute_register_map = config.input_register_map; float24 dummy_register; boost::fill(state.input_register_table, &dummy_register); @@ -584,16 +582,16 @@ OutputVertex RunShader(const InputVertex& input, int num_attributes) { state.conditional_code[1] = false; ProcessShaderCode(state); - DebugUtils::DumpShader(vs.program_code.data(), state.debug.max_offset, vs.swizzle_data.data(), - state.debug.max_opdesc_id, regs.vs_main_offset, - regs.vs_output_attributes); + DebugUtils::DumpShader(setup.program_code.data(), state.debug.max_offset, setup.swizzle_data.data(), + state.debug.max_opdesc_id, config.main_offset, + g_state.regs.vs_output_attributes); // TODO: Don't hardcode VS here // Setup output data OutputVertex ret; // TODO(neobrain): Under some circumstances, up to 16 attributes may be output. We need to // figure out what those circumstances are and enable the remaining outputs then. for (int i = 0; i < 7; ++i) { - const auto& output_register_map = regs.vs_output_attributes[i]; + const auto& output_register_map = g_state.regs.vs_output_attributes[i]; // TODO: Don't hardcode VS here u32 semantics[4] = { output_register_map.map_x, output_register_map.map_y, diff --git a/src/video_core/vertex_shader.h b/src/video_core/vertex_shader.h index c997e6a7..97f9250d 100644 --- a/src/video_core/vertex_shader.h +++ b/src/video_core/vertex_shader.h @@ -65,7 +65,7 @@ struct OutputVertex { static_assert(std::is_pod<OutputVertex>::value, "Structure is not POD"); static_assert(sizeof(OutputVertex) == 32 * sizeof(float), "OutputVertex has invalid size"); -OutputVertex RunShader(const InputVertex& input, int num_attributes); +OutputVertex RunShader(const InputVertex& input, int num_attributes, const Regs::ShaderConfig& config, const State::ShaderSetup& setup); } // namespace |