diff options
44 files changed, 605 insertions, 655 deletions
diff --git a/CMakeLists.txt b/CMakeLists.txt index dc782e252..44ed4196d 100644 --- a/CMakeLists.txt +++ b/CMakeLists.txt @@ -151,15 +151,16 @@ if (ENABLE_SDL2) set(SDL2_INCLUDE_DIR "${SDL2_PREFIX}/include" CACHE PATH "Path to SDL2 headers") set(SDL2_LIBRARY "${SDL2_PREFIX}/lib/x64/SDL2.lib" CACHE PATH "Path to SDL2 library") set(SDL2_DLL_DIR "${SDL2_PREFIX}/lib/x64/" CACHE PATH "Path to SDL2.dll") - else() - find_package(SDL2 REQUIRED) - endif() - if (SDL2_FOUND) - # TODO(yuriks): Make FindSDL2.cmake export an IMPORTED library instead add_library(SDL2 INTERFACE) target_link_libraries(SDL2 INTERFACE "${SDL2_LIBRARY}") target_include_directories(SDL2 INTERFACE "${SDL2_INCLUDE_DIR}") + else() + find_package(SDL2 REQUIRED) + include_directories(${SDL2_INCLUDE_DIRS}) + + add_library(SDL2 INTERFACE) + target_link_libraries(SDL2 INTERFACE "${SDL2_LIBRARIES}") endif() else() set(SDL2_FOUND NO) diff --git a/externals/cmake-modules/FindSDL2.cmake b/externals/cmake-modules/FindSDL2.cmake deleted file mode 100644 index 22ce752c5..000000000 --- a/externals/cmake-modules/FindSDL2.cmake +++ /dev/null @@ -1,239 +0,0 @@ - -# This module defines -# SDL2_LIBRARY, the name of the library to link against -# SDL2_FOUND, if false, do not try to link to SDL2 -# SDL2_INCLUDE_DIR, where to find SDL.h -# SDL2_DLL_DIR, where to find SDL2.dll if it exists -# -# This module responds to the the flag: -# SDL2_BUILDING_LIBRARY -# If this is defined, then no SDL2main will be linked in because -# only applications need main(). -# Otherwise, it is assumed you are building an application and this -# module will attempt to locate and set the the proper link flags -# as part of the returned SDL2_LIBRARY variable. -# -# Don't forget to include SDLmain.h and SDLmain.m your project for the -# OS X framework based version. (Other versions link to -lSDL2main which -# this module will try to find on your behalf.) Also for OS X, this -# module will automatically add the -framework Cocoa on your behalf. -# -# -# Additional Note: If you see an empty SDL2_LIBRARY_TEMP in your configuration -# and no SDL2_LIBRARY, it means CMake did not find your SDL2 library -# (SDL2.dll, libsdl2.so, SDL2.framework, etc). -# Set SDL2_LIBRARY_TEMP to point to your SDL2 library, and configure again. -# Similarly, if you see an empty SDL2MAIN_LIBRARY, you should set this value -# as appropriate. These values are used to generate the final SDL2_LIBRARY -# variable, but when these values are unset, SDL2_LIBRARY does not get created. -# -# -# $SDL2DIR is an environment variable that would -# correspond to the ./configure --prefix=$SDL2DIR -# used in building SDL2. -# l.e.galup 9-20-02 -# -# Modified by Eric Wing. -# Added code to assist with automated building by using environmental variables -# and providing a more controlled/consistent search behavior. -# Added new modifications to recognize OS X frameworks and -# additional Unix paths (FreeBSD, etc). -# Also corrected the header search path to follow "proper" SDL guidelines. -# Added a search for SDL2main which is needed by some platforms. -# Added a search for threads which is needed by some platforms. -# Added needed compile switches for MinGW. -# -# On OSX, this will prefer the Framework version (if found) over others. -# People will have to manually change the cache values of -# SDL2_LIBRARY to override this selection or set the CMake environment -# CMAKE_INCLUDE_PATH to modify the search paths. -# -# Note that the header path has changed from SDL2/SDL.h to just SDL.h -# This needed to change because "proper" SDL convention -# is #include "SDL.h", not <SDL2/SDL.h>. This is done for portability -# reasons because not all systems place things in SDL2/ (see FreeBSD). - -#============================================================================= -# Copyright 2003-2009 Kitware, Inc. -# -# Distributed under the OSI-approved BSD License (the "License"). -# -# This software is distributed WITHOUT ANY WARRANTY; without even the -# implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. -# See the License for more information. -#============================================================================= -# CMake - Cross Platform Makefile Generator -# Copyright 2000-2016 Kitware, Inc. -# Copyright 2000-2011 Insight Software Consortium -# All rights reserved. -# -# Redistribution and use in source and binary forms, with or without -# modification, are permitted provided that the following conditions -# are met: -# -# * Redistributions of source code must retain the above copyright -# notice, this list of conditions and the following disclaimer. -# -# * Redistributions in binary form must reproduce the above copyright -# notice, this list of conditions and the following disclaimer in the -# documentation and/or other materials provided with the distribution. -# -# * Neither the names of Kitware, Inc., the Insight Software Consortium, -# nor the names of their contributors may be used to endorse or promote -# products derived from this software without specific prior written -# permission. -# -# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS -# "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT -# LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR -# A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT -# HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, -# SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT -# LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, -# DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY -# THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT -# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE -# OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. -# -# ------------------------------------------------------------------------------ -# -# The above copyright and license notice applies to distributions of -# CMake in source and binary form. Some source files contain additional -# notices of original copyright by their contributors; see each source -# for details. Third-party software packages supplied with CMake under -# compatible licenses provide their own copyright notices documented in -# corresponding subdirectories. -# -# ------------------------------------------------------------------------------ -# -# CMake was initially developed by Kitware with the following sponsorship: -# -# * National Library of Medicine at the National Institutes of Health -# as part of the Insight Segmentation and Registration Toolkit (ITK). -# -# * US National Labs (Los Alamos, Livermore, Sandia) ASC Parallel -# Visualization Initiative. -# -# * National Alliance for Medical Image Computing (NAMIC) is funded by the -# National Institutes of Health through the NIH Roadmap for Medical Research, -# Grant U54 EB005149. -# -# * Kitware, Inc. -# - -message("<FindSDL2.cmake>") - -SET(SDL2_SEARCH_PATHS - ~/Library/Frameworks - /Library/Frameworks - /usr/local - /usr - /sw # Fink - /opt/local # DarwinPorts - /opt/csw # Blastwave - /opt - ${SDL2_PATH} -) - -if(CMAKE_SIZEOF_VOID_P EQUAL 8) - set(VC_LIB_PATH_SUFFIX lib/x64) -else() - set(VC_LIB_PATH_SUFFIX lib/x86) -endif() - -FIND_LIBRARY(SDL2_LIBRARY_TEMP - NAMES SDL2 - HINTS - $ENV{SDL2DIR} - PATH_SUFFIXES lib64 lib ${VC_LIB_PATH_SUFFIX} - PATHS ${SDL2_SEARCH_PATHS} -) - -IF(SDL2_LIBRARY_TEMP) - if(MSVC) - get_filename_component(SDL2_DLL_DIR_TEMP ${SDL2_LIBRARY_TEMP} DIRECTORY) - if(EXISTS ${SDL2_DLL_DIR_TEMP}/SDL2.dll) - set(SDL2_DLL_DIR ${SDL2_DLL_DIR_TEMP}) - unset(SDL2_DLL_DIR_TEMP) - endif() - endif() - - FIND_PATH(SDL2_INCLUDE_DIR SDL.h - HINTS - $ENV{SDL2DIR} - PATH_SUFFIXES include/SDL2 include - PATHS ${SDL2_SEARCH_PATHS} - ) - - IF(NOT SDL2_BUILDING_LIBRARY) - IF(NOT ${SDL2_INCLUDE_DIR} MATCHES ".framework") - # Non-OS X framework versions expect you to also dynamically link to - # SDL2main. This is mainly for Windows and OS X. Other (Unix) platforms - # seem to provide SDL2main for compatibility even though they don't - # necessarily need it. - FIND_LIBRARY(SDL2MAIN_LIBRARY - NAMES SDL2main - HINTS - $ENV{SDL2DIR} - PATH_SUFFIXES lib64 lib - PATHS ${SDL2_SEARCH_PATHS} - ) - ENDIF(NOT ${SDL2_INCLUDE_DIR} MATCHES ".framework") - ENDIF(NOT SDL2_BUILDING_LIBRARY) - - # SDL2 may require threads on your system. - # The Apple build may not need an explicit flag because one of the - # frameworks may already provide it. - # But for non-OSX systems, I will use the CMake Threads package. - IF(NOT APPLE) - FIND_PACKAGE(Threads) - ENDIF(NOT APPLE) - - # MinGW needs an additional library, mwindows - # It's total link flags should look like -lmingw32 -lSDL2main -lSDL2 -lmwindows - # (Actually on second look, I think it only needs one of the m* libraries.) - IF(MINGW) - SET(MINGW32_LIBRARY mingw32 CACHE STRING "mwindows for MinGW") - ENDIF(MINGW) - - # For SDL2main - IF(NOT SDL2_BUILDING_LIBRARY) - IF(SDL2MAIN_LIBRARY) - SET(SDL2_LIBRARY_TEMP ${SDL2MAIN_LIBRARY} ${SDL2_LIBRARY_TEMP}) - ENDIF(SDL2MAIN_LIBRARY) - ENDIF(NOT SDL2_BUILDING_LIBRARY) - - # For OS X, SDL2 uses Cocoa as a backend so it must link to Cocoa. - # CMake doesn't display the -framework Cocoa string in the UI even - # though it actually is there if I modify a pre-used variable. - # I think it has something to do with the CACHE STRING. - # So I use a temporary variable until the end so I can set the - # "real" variable in one-shot. - IF(APPLE) - SET(SDL2_LIBRARY_TEMP ${SDL2_LIBRARY_TEMP} "-framework Cocoa") - ENDIF(APPLE) - - # For threads, as mentioned Apple doesn't need this. - # In fact, there seems to be a problem if I used the Threads package - # and try using this line, so I'm just skipping it entirely for OS X. - IF(NOT APPLE) - SET(SDL2_LIBRARY_TEMP ${SDL2_LIBRARY_TEMP} ${CMAKE_THREAD_LIBS_INIT}) - ENDIF(NOT APPLE) - - # For MinGW library - IF(MINGW) - SET(SDL2_LIBRARY_TEMP ${MINGW32_LIBRARY} ${SDL2_LIBRARY_TEMP}) - ENDIF(MINGW) - - # Set the final string here so the GUI reflects the final state. - SET(SDL2_LIBRARY ${SDL2_LIBRARY_TEMP} CACHE STRING "Where the SDL2 Library can be found") - - # Unset the temp variable to INTERNAL so it is not seen in the CMake GUI - UNSET(SDL2_LIBRARY_TEMP) -ENDIF(SDL2_LIBRARY_TEMP) - -message("</FindSDL2.cmake>") - -INCLUDE(FindPackageHandleStandardArgs) - -FIND_PACKAGE_HANDLE_STANDARD_ARGS(SDL2 REQUIRED_VARS SDL2_LIBRARY SDL2_INCLUDE_DIR) diff --git a/src/core/hle/kernel/address_arbiter.cpp b/src/core/hle/kernel/address_arbiter.cpp index 2ea3dcb61..8475b698c 100644 --- a/src/core/hle/kernel/address_arbiter.cpp +++ b/src/core/hle/kernel/address_arbiter.cpp @@ -201,42 +201,39 @@ void AddressArbiter::HandleWakeupThread(std::shared_ptr<Thread> thread) { void AddressArbiter::InsertThread(std::shared_ptr<Thread> thread) { const VAddr arb_addr = thread->GetArbiterWaitAddress(); std::list<std::shared_ptr<Thread>>& thread_list = arb_threads[arb_addr]; - auto it = thread_list.begin(); - while (it != thread_list.end()) { - const std::shared_ptr<Thread>& current_thread = *it; - if (current_thread->GetPriority() >= thread->GetPriority()) { - thread_list.insert(it, thread); - return; - } - ++it; + + const auto iter = + std::find_if(thread_list.cbegin(), thread_list.cend(), [&thread](const auto& entry) { + return entry->GetPriority() >= thread->GetPriority(); + }); + + if (iter == thread_list.cend()) { + thread_list.push_back(std::move(thread)); + } else { + thread_list.insert(iter, std::move(thread)); } - thread_list.push_back(std::move(thread)); } void AddressArbiter::RemoveThread(std::shared_ptr<Thread> thread) { const VAddr arb_addr = thread->GetArbiterWaitAddress(); std::list<std::shared_ptr<Thread>>& thread_list = arb_threads[arb_addr]; - auto it = thread_list.begin(); - while (it != thread_list.end()) { - const std::shared_ptr<Thread>& current_thread = *it; - if (current_thread.get() == thread.get()) { - thread_list.erase(it); - return; - } - ++it; - } - UNREACHABLE(); + + const auto iter = std::find_if(thread_list.cbegin(), thread_list.cend(), + [&thread](const auto& entry) { return thread == entry; }); + + ASSERT(iter != thread_list.cend()); + + thread_list.erase(iter); } -std::vector<std::shared_ptr<Thread>> AddressArbiter::GetThreadsWaitingOnAddress(VAddr address) { - std::vector<std::shared_ptr<Thread>> result; - std::list<std::shared_ptr<Thread>>& thread_list = arb_threads[address]; - auto it = thread_list.begin(); - while (it != thread_list.end()) { - std::shared_ptr<Thread> current_thread = *it; - result.push_back(std::move(current_thread)); - ++it; +std::vector<std::shared_ptr<Thread>> AddressArbiter::GetThreadsWaitingOnAddress( + VAddr address) const { + const auto iter = arb_threads.find(address); + if (iter == arb_threads.cend()) { + return {}; } - return result; + + const std::list<std::shared_ptr<Thread>>& thread_list = iter->second; + return {thread_list.cbegin(), thread_list.cend()}; } } // namespace Kernel diff --git a/src/core/hle/kernel/address_arbiter.h b/src/core/hle/kernel/address_arbiter.h index 386983e54..f958eee5a 100644 --- a/src/core/hle/kernel/address_arbiter.h +++ b/src/core/hle/kernel/address_arbiter.h @@ -86,7 +86,7 @@ private: void RemoveThread(std::shared_ptr<Thread> thread); // Gets the threads waiting on an address. - std::vector<std::shared_ptr<Thread>> GetThreadsWaitingOnAddress(VAddr address); + std::vector<std::shared_ptr<Thread>> GetThreadsWaitingOnAddress(VAddr address) const; /// List of threads waiting for a address arbiter std::unordered_map<VAddr, std::list<std::shared_ptr<Thread>>> arb_threads; diff --git a/src/core/hle/kernel/hle_ipc.cpp b/src/core/hle/kernel/hle_ipc.cpp index 2db28dcf0..ab05788d7 100644 --- a/src/core/hle/kernel/hle_ipc.cpp +++ b/src/core/hle/kernel/hle_ipc.cpp @@ -284,13 +284,18 @@ ResultCode HLERequestContext::WriteToOutgoingCommandBuffer(Thread& thread) { std::vector<u8> HLERequestContext::ReadBuffer(int buffer_index) const { std::vector<u8> buffer; - const bool is_buffer_a{BufferDescriptorA().size() && BufferDescriptorA()[buffer_index].Size()}; + const bool is_buffer_a{BufferDescriptorA().size() > buffer_index && + BufferDescriptorA()[buffer_index].Size()}; auto& memory = Core::System::GetInstance().Memory(); if (is_buffer_a) { + ASSERT_MSG(BufferDescriptorA().size() > buffer_index, + "BufferDescriptorA invalid buffer_index {}", buffer_index); buffer.resize(BufferDescriptorA()[buffer_index].Size()); memory.ReadBlock(BufferDescriptorA()[buffer_index].Address(), buffer.data(), buffer.size()); } else { + ASSERT_MSG(BufferDescriptorX().size() > buffer_index, + "BufferDescriptorX invalid buffer_index {}", buffer_index); buffer.resize(BufferDescriptorX()[buffer_index].Size()); memory.ReadBlock(BufferDescriptorX()[buffer_index].Address(), buffer.data(), buffer.size()); } @@ -305,7 +310,8 @@ std::size_t HLERequestContext::WriteBuffer(const void* buffer, std::size_t size, return 0; } - const bool is_buffer_b{BufferDescriptorB().size() && BufferDescriptorB()[buffer_index].Size()}; + const bool is_buffer_b{BufferDescriptorB().size() > buffer_index && + BufferDescriptorB()[buffer_index].Size()}; const std::size_t buffer_size{GetWriteBufferSize(buffer_index)}; if (size > buffer_size) { LOG_CRITICAL(Core, "size ({:016X}) is greater than buffer_size ({:016X})", size, @@ -315,8 +321,16 @@ std::size_t HLERequestContext::WriteBuffer(const void* buffer, std::size_t size, auto& memory = Core::System::GetInstance().Memory(); if (is_buffer_b) { + ASSERT_MSG(BufferDescriptorB().size() > buffer_index, + "BufferDescriptorB invalid buffer_index {}", buffer_index); + ASSERT_MSG(BufferDescriptorB()[buffer_index].Size() >= size, + "BufferDescriptorB buffer_index {} is not large enough", buffer_index); memory.WriteBlock(BufferDescriptorB()[buffer_index].Address(), buffer, size); } else { + ASSERT_MSG(BufferDescriptorC().size() > buffer_index, + "BufferDescriptorC invalid buffer_index {}", buffer_index); + ASSERT_MSG(BufferDescriptorC()[buffer_index].Size() >= size, + "BufferDescriptorC buffer_index {} is not large enough", buffer_index); memory.WriteBlock(BufferDescriptorC()[buffer_index].Address(), buffer, size); } @@ -324,15 +338,35 @@ std::size_t HLERequestContext::WriteBuffer(const void* buffer, std::size_t size, } std::size_t HLERequestContext::GetReadBufferSize(int buffer_index) const { - const bool is_buffer_a{BufferDescriptorA().size() && BufferDescriptorA()[buffer_index].Size()}; - return is_buffer_a ? BufferDescriptorA()[buffer_index].Size() - : BufferDescriptorX()[buffer_index].Size(); + const bool is_buffer_a{BufferDescriptorA().size() > buffer_index && + BufferDescriptorA()[buffer_index].Size()}; + if (is_buffer_a) { + ASSERT_MSG(BufferDescriptorA().size() > buffer_index, + "BufferDescriptorA invalid buffer_index {}", buffer_index); + ASSERT_MSG(BufferDescriptorA()[buffer_index].Size() > 0, + "BufferDescriptorA buffer_index {} is empty", buffer_index); + return BufferDescriptorA()[buffer_index].Size(); + } else { + ASSERT_MSG(BufferDescriptorX().size() > buffer_index, + "BufferDescriptorX invalid buffer_index {}", buffer_index); + ASSERT_MSG(BufferDescriptorX()[buffer_index].Size() > 0, + "BufferDescriptorX buffer_index {} is empty", buffer_index); + return BufferDescriptorX()[buffer_index].Size(); + } } std::size_t HLERequestContext::GetWriteBufferSize(int buffer_index) const { - const bool is_buffer_b{BufferDescriptorB().size() && BufferDescriptorB()[buffer_index].Size()}; - return is_buffer_b ? BufferDescriptorB()[buffer_index].Size() - : BufferDescriptorC()[buffer_index].Size(); + const bool is_buffer_b{BufferDescriptorB().size() > buffer_index && + BufferDescriptorB()[buffer_index].Size()}; + if (is_buffer_b) { + ASSERT_MSG(BufferDescriptorB().size() > buffer_index, + "BufferDescriptorB invalid buffer_index {}", buffer_index); + return BufferDescriptorB()[buffer_index].Size(); + } else { + ASSERT_MSG(BufferDescriptorC().size() > buffer_index, + "BufferDescriptorC invalid buffer_index {}", buffer_index); + return BufferDescriptorC()[buffer_index].Size(); + } } std::string HLERequestContext::Description() const { diff --git a/src/core/hle/kernel/svc.cpp b/src/core/hle/kernel/svc.cpp index 1d99bf7a2..9cae5c73d 100644 --- a/src/core/hle/kernel/svc.cpp +++ b/src/core/hle/kernel/svc.cpp @@ -1863,10 +1863,14 @@ static ResultCode CreateTransferMemory(Core::System& system, Handle* handle, VAd } auto& kernel = system.Kernel(); - auto transfer_mem_handle = TransferMemory::Create(kernel, addr, size, perms); + auto transfer_mem_handle = TransferMemory::Create(kernel, system.Memory(), addr, size, perms); + + if (const auto reserve_result{transfer_mem_handle->Reserve()}; reserve_result.IsError()) { + return reserve_result; + } auto& handle_table = kernel.CurrentProcess()->GetHandleTable(); - const auto result = handle_table.Create(std::move(transfer_mem_handle)); + const auto result{handle_table.Create(std::move(transfer_mem_handle))}; if (result.Failed()) { return result.Code(); } diff --git a/src/core/hle/kernel/thread.cpp b/src/core/hle/kernel/thread.cpp index e965b5b04..ad464e03b 100644 --- a/src/core/hle/kernel/thread.cpp +++ b/src/core/hle/kernel/thread.cpp @@ -474,7 +474,7 @@ void Thread::AdjustSchedulingOnPriority(u32 old_priority) { if (GetSchedulingStatus() != ThreadSchedStatus::Runnable) { return; } - auto& scheduler = Core::System::GetInstance().GlobalScheduler(); + auto& scheduler = kernel.GlobalScheduler(); if (processor_id >= 0) { scheduler.Unschedule(old_priority, static_cast<u32>(processor_id), this); } @@ -506,7 +506,7 @@ void Thread::AdjustSchedulingOnPriority(u32 old_priority) { } void Thread::AdjustSchedulingOnAffinity(u64 old_affinity_mask, s32 old_core) { - auto& scheduler = Core::System::GetInstance().GlobalScheduler(); + auto& scheduler = kernel.GlobalScheduler(); if (GetSchedulingStatus() != ThreadSchedStatus::Runnable || current_priority >= THREADPRIO_COUNT) { return; diff --git a/src/core/hle/kernel/transfer_memory.cpp b/src/core/hle/kernel/transfer_memory.cpp index f0e73f57b..f2d3f8b49 100644 --- a/src/core/hle/kernel/transfer_memory.cpp +++ b/src/core/hle/kernel/transfer_memory.cpp @@ -8,15 +8,23 @@ #include "core/hle/kernel/shared_memory.h" #include "core/hle/kernel/transfer_memory.h" #include "core/hle/result.h" +#include "core/memory.h" namespace Kernel { -TransferMemory::TransferMemory(KernelCore& kernel) : Object{kernel} {} -TransferMemory::~TransferMemory() = default; +TransferMemory::TransferMemory(KernelCore& kernel, Memory::Memory& memory) + : Object{kernel}, memory{memory} {} -std::shared_ptr<TransferMemory> TransferMemory::Create(KernelCore& kernel, VAddr base_address, - u64 size, MemoryPermission permissions) { - std::shared_ptr<TransferMemory> transfer_memory{std::make_shared<TransferMemory>(kernel)}; +TransferMemory::~TransferMemory() { + // Release memory region when transfer memory is destroyed + Reset(); +} + +std::shared_ptr<TransferMemory> TransferMemory::Create(KernelCore& kernel, Memory::Memory& memory, + VAddr base_address, u64 size, + MemoryPermission permissions) { + std::shared_ptr<TransferMemory> transfer_memory{ + std::make_shared<TransferMemory>(kernel, memory)}; transfer_memory->base_address = base_address; transfer_memory->memory_size = size; @@ -27,7 +35,7 @@ std::shared_ptr<TransferMemory> TransferMemory::Create(KernelCore& kernel, VAddr } const u8* TransferMemory::GetPointer() const { - return backing_block.get()->data(); + return memory.GetPointer(base_address); } u64 TransferMemory::GetSize() const { @@ -62,6 +70,52 @@ ResultCode TransferMemory::MapMemory(VAddr address, u64 size, MemoryPermission p return RESULT_SUCCESS; } +ResultCode TransferMemory::Reserve() { + auto& vm_manager{owner_process->VMManager()}; + const auto check_range_result{vm_manager.CheckRangeState( + base_address, memory_size, MemoryState::FlagTransfer | MemoryState::FlagMemoryPoolAllocated, + MemoryState::FlagTransfer | MemoryState::FlagMemoryPoolAllocated, VMAPermission::All, + VMAPermission::ReadWrite, MemoryAttribute::Mask, MemoryAttribute::None, + MemoryAttribute::IpcAndDeviceMapped)}; + + if (check_range_result.Failed()) { + return check_range_result.Code(); + } + + auto [state_, permissions_, attribute] = *check_range_result; + + if (const auto result{vm_manager.ReprotectRange( + base_address, memory_size, SharedMemory::ConvertPermissions(owner_permissions))}; + result.IsError()) { + return result; + } + + return vm_manager.SetMemoryAttribute(base_address, memory_size, MemoryAttribute::Mask, + attribute | MemoryAttribute::Locked); +} + +ResultCode TransferMemory::Reset() { + auto& vm_manager{owner_process->VMManager()}; + if (const auto result{vm_manager.CheckRangeState( + base_address, memory_size, + MemoryState::FlagTransfer | MemoryState::FlagMemoryPoolAllocated, + MemoryState::FlagTransfer | MemoryState::FlagMemoryPoolAllocated, VMAPermission::None, + VMAPermission::None, MemoryAttribute::Mask, MemoryAttribute::Locked, + MemoryAttribute::IpcAndDeviceMapped)}; + result.Failed()) { + return result.Code(); + } + + if (const auto result{ + vm_manager.ReprotectRange(base_address, memory_size, VMAPermission::ReadWrite)}; + result.IsError()) { + return result; + } + + return vm_manager.SetMemoryAttribute(base_address, memory_size, MemoryAttribute::Mask, + MemoryAttribute::None); +} + ResultCode TransferMemory::UnmapMemory(VAddr address, u64 size) { if (memory_size != size) { return ERR_INVALID_SIZE; diff --git a/src/core/hle/kernel/transfer_memory.h b/src/core/hle/kernel/transfer_memory.h index 0a6e15d18..6e388536a 100644 --- a/src/core/hle/kernel/transfer_memory.h +++ b/src/core/hle/kernel/transfer_memory.h @@ -11,6 +11,10 @@ union ResultCode; +namespace Memory { +class Memory; +} + namespace Kernel { class KernelCore; @@ -26,12 +30,13 @@ enum class MemoryPermission : u32; /// class TransferMemory final : public Object { public: - explicit TransferMemory(KernelCore& kernel); + explicit TransferMemory(KernelCore& kernel, Memory::Memory& memory); ~TransferMemory() override; static constexpr HandleType HANDLE_TYPE = HandleType::TransferMemory; - static std::shared_ptr<TransferMemory> Create(KernelCore& kernel, VAddr base_address, u64 size, + static std::shared_ptr<TransferMemory> Create(KernelCore& kernel, Memory::Memory& memory, + VAddr base_address, u64 size, MemoryPermission permissions); TransferMemory(const TransferMemory&) = delete; @@ -80,6 +85,14 @@ public: /// ResultCode UnmapMemory(VAddr address, u64 size); + /// Reserves the region to be used for the transfer memory, called after the transfer memory is + /// created. + ResultCode Reserve(); + + /// Resets the region previously used for the transfer memory, called after the transfer memory + /// is closed. + ResultCode Reset(); + private: /// Memory block backing this instance. std::shared_ptr<PhysicalMemory> backing_block; @@ -98,6 +111,8 @@ private: /// Whether or not this transfer memory instance has mapped memory. bool is_mapped = false; + + Memory::Memory& memory; }; } // namespace Kernel diff --git a/src/core/hle/kernel/vm_manager.cpp b/src/core/hle/kernel/vm_manager.cpp index 0b3500fce..024c22901 100644 --- a/src/core/hle/kernel/vm_manager.cpp +++ b/src/core/hle/kernel/vm_manager.cpp @@ -544,7 +544,8 @@ MemoryInfo VMManager::QueryMemory(VAddr address) const { ResultCode VMManager::SetMemoryAttribute(VAddr address, u64 size, MemoryAttribute mask, MemoryAttribute attribute) { - constexpr auto ignore_mask = MemoryAttribute::Uncached | MemoryAttribute::DeviceMapped; + constexpr auto ignore_mask = + MemoryAttribute::Uncached | MemoryAttribute::DeviceMapped | MemoryAttribute::Locked; constexpr auto attribute_mask = ~ignore_mask; const auto result = CheckRangeState( diff --git a/src/core/hle/kernel/vm_manager.h b/src/core/hle/kernel/vm_manager.h index 850a7ebc3..90b4b006a 100644 --- a/src/core/hle/kernel/vm_manager.h +++ b/src/core/hle/kernel/vm_manager.h @@ -98,6 +98,8 @@ enum class MemoryAttribute : u32 { DeviceMapped = 4, /// Uncached memory Uncached = 8, + + IpcAndDeviceMapped = LockedForIPC | DeviceMapped, }; constexpr MemoryAttribute operator|(MemoryAttribute lhs, MemoryAttribute rhs) { @@ -654,6 +656,35 @@ public: /// is scheduled. Common::PageTable page_table{Memory::PAGE_BITS}; + using CheckResults = ResultVal<std::tuple<MemoryState, VMAPermission, MemoryAttribute>>; + + /// Checks if an address range adheres to the specified states provided. + /// + /// @param address The starting address of the address range. + /// @param size The size of the address range. + /// @param state_mask The memory state mask. + /// @param state The state to compare the individual VMA states against, + /// which is done in the form of: (vma.state & state_mask) != state. + /// @param permission_mask The memory permissions mask. + /// @param permissions The permission to compare the individual VMA permissions against, + /// which is done in the form of: + /// (vma.permission & permission_mask) != permission. + /// @param attribute_mask The memory attribute mask. + /// @param attribute The memory attributes to compare the individual VMA attributes + /// against, which is done in the form of: + /// (vma.attributes & attribute_mask) != attribute. + /// @param ignore_mask The memory attributes to ignore during the check. + /// + /// @returns If successful, returns a tuple containing the memory attributes + /// (with ignored bits specified by ignore_mask unset), memory permissions, and + /// memory state across the memory range. + /// @returns If not successful, returns ERR_INVALID_ADDRESS_STATE. + /// + CheckResults CheckRangeState(VAddr address, u64 size, MemoryState state_mask, MemoryState state, + VMAPermission permission_mask, VMAPermission permissions, + MemoryAttribute attribute_mask, MemoryAttribute attribute, + MemoryAttribute ignore_mask) const; + private: using VMAIter = VMAMap::iterator; @@ -707,35 +738,6 @@ private: /// Clears out the page table void ClearPageTable(); - using CheckResults = ResultVal<std::tuple<MemoryState, VMAPermission, MemoryAttribute>>; - - /// Checks if an address range adheres to the specified states provided. - /// - /// @param address The starting address of the address range. - /// @param size The size of the address range. - /// @param state_mask The memory state mask. - /// @param state The state to compare the individual VMA states against, - /// which is done in the form of: (vma.state & state_mask) != state. - /// @param permission_mask The memory permissions mask. - /// @param permissions The permission to compare the individual VMA permissions against, - /// which is done in the form of: - /// (vma.permission & permission_mask) != permission. - /// @param attribute_mask The memory attribute mask. - /// @param attribute The memory attributes to compare the individual VMA attributes - /// against, which is done in the form of: - /// (vma.attributes & attribute_mask) != attribute. - /// @param ignore_mask The memory attributes to ignore during the check. - /// - /// @returns If successful, returns a tuple containing the memory attributes - /// (with ignored bits specified by ignore_mask unset), memory permissions, and - /// memory state across the memory range. - /// @returns If not successful, returns ERR_INVALID_ADDRESS_STATE. - /// - CheckResults CheckRangeState(VAddr address, u64 size, MemoryState state_mask, MemoryState state, - VMAPermission permission_mask, VMAPermission permissions, - MemoryAttribute attribute_mask, MemoryAttribute attribute, - MemoryAttribute ignore_mask) const; - /// Gets the amount of memory currently mapped (state != Unmapped) in a range. ResultVal<std::size_t> SizeOfAllocatedVMAsInRange(VAddr address, std::size_t size) const; diff --git a/src/core/hle/kernel/wait_object.cpp b/src/core/hle/kernel/wait_object.cpp index a0c806e8f..1838260fd 100644 --- a/src/core/hle/kernel/wait_object.cpp +++ b/src/core/hle/kernel/wait_object.cpp @@ -50,17 +50,8 @@ std::shared_ptr<Thread> WaitObject::GetHighestPriorityReadyThread() const { if (ShouldWait(thread.get())) continue; - // A thread is ready to run if it's either in ThreadStatus::WaitSynch - // and the rest of the objects it is waiting on are ready. - bool ready_to_run = true; - if (thread_status == ThreadStatus::WaitSynch) { - ready_to_run = thread->AllWaitObjectsReady(); - } - - if (ready_to_run) { - candidate = thread.get(); - candidate_priority = thread->GetPriority(); - } + candidate = thread.get(); + candidate_priority = thread->GetPriority(); } return SharedFrom(candidate); diff --git a/src/core/hle/service/am/am.cpp b/src/core/hle/service/am/am.cpp index 95aa5d23d..cc978713b 100644 --- a/src/core/hle/service/am/am.cpp +++ b/src/core/hle/service/am/am.cpp @@ -709,8 +709,34 @@ void ICommonStateGetter::SetCpuBoostMode(Kernel::HLERequestContext& ctx) { apm_sys->SetCpuBoostMode(ctx); } -IStorage::IStorage(std::vector<u8> buffer) - : ServiceFramework("IStorage"), buffer(std::move(buffer)) { +IStorageImpl::~IStorageImpl() = default; + +class StorageDataImpl final : public IStorageImpl { +public: + explicit StorageDataImpl(std::vector<u8>&& buffer) : buffer{std::move(buffer)} {} + + std::vector<u8>& GetData() override { + return buffer; + } + + const std::vector<u8>& GetData() const override { + return buffer; + } + + std::size_t GetSize() const override { + return buffer.size(); + } + +private: + std::vector<u8> buffer; +}; + +IStorage::IStorage(std::vector<u8>&& buffer) + : ServiceFramework("IStorage"), impl{std::make_shared<StorageDataImpl>(std::move(buffer))} { + Register(); +} + +void IStorage::Register() { // clang-format off static const FunctionInfo functions[] = { {0, &IStorage::Open, "Open"}, @@ -723,8 +749,13 @@ IStorage::IStorage(std::vector<u8> buffer) IStorage::~IStorage() = default; -const std::vector<u8>& IStorage::GetData() const { - return buffer; +void IStorage::Open(Kernel::HLERequestContext& ctx) { + LOG_DEBUG(Service_AM, "called"); + + IPC::ResponseBuilder rb{ctx, 2, 0, 1}; + + rb.Push(RESULT_SUCCESS); + rb.PushIpcInterface<IStorageAccessor>(*this); } void ICommonStateGetter::GetOperationMode(Kernel::HLERequestContext& ctx) { @@ -816,7 +847,7 @@ private: LOG_DEBUG(Service_AM, "called"); IPC::RequestParser rp{ctx}; - applet->GetBroker().PushNormalDataFromGame(*rp.PopIpcInterface<IStorage>()); + applet->GetBroker().PushNormalDataFromGame(rp.PopIpcInterface<IStorage>()); IPC::ResponseBuilder rb{ctx, 2}; rb.Push(RESULT_SUCCESS); @@ -825,26 +856,25 @@ private: void PopOutData(Kernel::HLERequestContext& ctx) { LOG_DEBUG(Service_AM, "called"); - IPC::ResponseBuilder rb{ctx, 2, 0, 1}; - const auto storage = applet->GetBroker().PopNormalDataToGame(); if (storage == nullptr) { LOG_ERROR(Service_AM, "storage is a nullptr. There is no data in the current normal channel"); - + IPC::ResponseBuilder rb{ctx, 2}; rb.Push(ERR_NO_DATA_IN_CHANNEL); return; } + IPC::ResponseBuilder rb{ctx, 2, 0, 1}; rb.Push(RESULT_SUCCESS); - rb.PushIpcInterface<IStorage>(std::move(*storage)); + rb.PushIpcInterface<IStorage>(std::move(storage)); } void PushInteractiveInData(Kernel::HLERequestContext& ctx) { LOG_DEBUG(Service_AM, "called"); IPC::RequestParser rp{ctx}; - applet->GetBroker().PushInteractiveDataFromGame(*rp.PopIpcInterface<IStorage>()); + applet->GetBroker().PushInteractiveDataFromGame(rp.PopIpcInterface<IStorage>()); ASSERT(applet->IsInitialized()); applet->ExecuteInteractive(); @@ -857,19 +887,18 @@ private: void PopInteractiveOutData(Kernel::HLERequestContext& ctx) { LOG_DEBUG(Service_AM, "called"); - IPC::ResponseBuilder rb{ctx, 2, 0, 1}; - const auto storage = applet->GetBroker().PopInteractiveDataToGame(); if (storage == nullptr) { LOG_ERROR(Service_AM, "storage is a nullptr. There is no data in the current interactive channel"); - + IPC::ResponseBuilder rb{ctx, 2}; rb.Push(ERR_NO_DATA_IN_CHANNEL); return; } + IPC::ResponseBuilder rb{ctx, 2, 0, 1}; rb.Push(RESULT_SUCCESS); - rb.PushIpcInterface<IStorage>(std::move(*storage)); + rb.PushIpcInterface<IStorage>(std::move(storage)); } void GetPopOutDataEvent(Kernel::HLERequestContext& ctx) { @@ -891,15 +920,6 @@ private: std::shared_ptr<Applets::Applet> applet; }; -void IStorage::Open(Kernel::HLERequestContext& ctx) { - LOG_DEBUG(Service_AM, "called"); - - IPC::ResponseBuilder rb{ctx, 2, 0, 1}; - - rb.Push(RESULT_SUCCESS); - rb.PushIpcInterface<IStorageAccessor>(*this); -} - IStorageAccessor::IStorageAccessor(IStorage& storage) : ServiceFramework("IStorageAccessor"), backing(storage) { // clang-format off @@ -921,7 +941,7 @@ void IStorageAccessor::GetSize(Kernel::HLERequestContext& ctx) { IPC::ResponseBuilder rb{ctx, 4}; rb.Push(RESULT_SUCCESS); - rb.Push(static_cast<u64>(backing.buffer.size())); + rb.Push(static_cast<u64>(backing.GetSize())); } void IStorageAccessor::Write(Kernel::HLERequestContext& ctx) { @@ -932,17 +952,17 @@ void IStorageAccessor::Write(Kernel::HLERequestContext& ctx) { LOG_DEBUG(Service_AM, "called, offset={}, size={}", offset, data.size()); - if (data.size() > backing.buffer.size() - offset) { + if (data.size() > backing.GetSize() - offset) { LOG_ERROR(Service_AM, "offset is out of bounds, backing_buffer_sz={}, data_size={}, offset={}", - backing.buffer.size(), data.size(), offset); + backing.GetSize(), data.size(), offset); IPC::ResponseBuilder rb{ctx, 2}; rb.Push(ERR_SIZE_OUT_OF_BOUNDS); return; } - std::memcpy(backing.buffer.data() + offset, data.data(), data.size()); + std::memcpy(backing.GetData().data() + offset, data.data(), data.size()); IPC::ResponseBuilder rb{ctx, 2}; rb.Push(RESULT_SUCCESS); @@ -956,16 +976,16 @@ void IStorageAccessor::Read(Kernel::HLERequestContext& ctx) { LOG_DEBUG(Service_AM, "called, offset={}, size={}", offset, size); - if (size > backing.buffer.size() - offset) { + if (size > backing.GetSize() - offset) { LOG_ERROR(Service_AM, "offset is out of bounds, backing_buffer_sz={}, size={}, offset={}", - backing.buffer.size(), size, offset); + backing.GetSize(), size, offset); IPC::ResponseBuilder rb{ctx, 2}; rb.Push(ERR_SIZE_OUT_OF_BOUNDS); return; } - ctx.WriteBuffer(backing.buffer.data() + offset, size); + ctx.WriteBuffer(backing.GetData().data() + offset, size); IPC::ResponseBuilder rb{ctx, 2}; rb.Push(RESULT_SUCCESS); @@ -1031,7 +1051,7 @@ void ILibraryAppletCreator::CreateTransferMemoryStorage(Kernel::HLERequestContex rp.SetCurrentOffset(3); const auto handle{rp.Pop<Kernel::Handle>()}; - const auto transfer_mem = + auto transfer_mem = system.CurrentProcess()->GetHandleTable().Get<Kernel::TransferMemory>(handle); if (transfer_mem == nullptr) { @@ -1047,7 +1067,7 @@ void ILibraryAppletCreator::CreateTransferMemoryStorage(Kernel::HLERequestContex IPC::ResponseBuilder rb{ctx, 2, 0, 1}; rb.Push(RESULT_SUCCESS); - rb.PushIpcInterface(std::make_shared<IStorage>(std::move(memory))); + rb.PushIpcInterface<IStorage>(std::move(memory)); } IApplicationFunctions::IApplicationFunctions(Core::System& system_) @@ -1189,13 +1209,11 @@ void IApplicationFunctions::PopLaunchParameter(Kernel::HLERequestContext& ctx) { u64 build_id{}; std::memcpy(&build_id, build_id_full.data(), sizeof(u64)); - const auto data = - backend->GetLaunchParameter({system.CurrentProcess()->GetTitleID(), build_id}); - + auto data = backend->GetLaunchParameter({system.CurrentProcess()->GetTitleID(), build_id}); if (data.has_value()) { IPC::ResponseBuilder rb{ctx, 2, 0, 1}; rb.Push(RESULT_SUCCESS); - rb.PushIpcInterface<AM::IStorage>(*data); + rb.PushIpcInterface<IStorage>(std::move(*data)); launch_popped_application_specific = true; return; } @@ -1218,7 +1236,7 @@ void IApplicationFunctions::PopLaunchParameter(Kernel::HLERequestContext& ctx) { std::vector<u8> buffer(sizeof(LaunchParameterAccountPreselectedUser)); std::memcpy(buffer.data(), ¶ms, buffer.size()); - rb.PushIpcInterface<AM::IStorage>(buffer); + rb.PushIpcInterface<IStorage>(std::move(buffer)); launch_popped_account_preselect = true; return; } diff --git a/src/core/hle/service/am/am.h b/src/core/hle/service/am/am.h index 448817be9..0b9a4332d 100644 --- a/src/core/hle/service/am/am.h +++ b/src/core/hle/service/am/am.h @@ -12,7 +12,8 @@ namespace Kernel { class KernelCore; -} +class TransferMemory; +} // namespace Kernel namespace Service::NVFlinger { class NVFlinger; @@ -188,19 +189,36 @@ private: std::shared_ptr<AppletMessageQueue> msg_queue; }; +class IStorageImpl { +public: + virtual ~IStorageImpl(); + virtual std::vector<u8>& GetData() = 0; + virtual const std::vector<u8>& GetData() const = 0; + virtual std::size_t GetSize() const = 0; +}; + class IStorage final : public ServiceFramework<IStorage> { public: - explicit IStorage(std::vector<u8> buffer); + explicit IStorage(std::vector<u8>&& buffer); ~IStorage() override; - const std::vector<u8>& GetData() const; + std::vector<u8>& GetData() { + return impl->GetData(); + } + + const std::vector<u8>& GetData() const { + return impl->GetData(); + } + + std::size_t GetSize() const { + return impl->GetSize(); + } private: + void Register(); void Open(Kernel::HLERequestContext& ctx); - std::vector<u8> buffer; - - friend class IStorageAccessor; + std::shared_ptr<IStorageImpl> impl; }; class IStorageAccessor final : public ServiceFramework<IStorageAccessor> { diff --git a/src/core/hle/service/am/applets/applets.cpp b/src/core/hle/service/am/applets/applets.cpp index 92f995f8f..c3261f3e6 100644 --- a/src/core/hle/service/am/applets/applets.cpp +++ b/src/core/hle/service/am/applets/applets.cpp @@ -50,16 +50,17 @@ AppletDataBroker::RawChannelData AppletDataBroker::PeekDataToAppletForDebug() co return {std::move(out_normal), std::move(out_interactive)}; } -std::unique_ptr<IStorage> AppletDataBroker::PopNormalDataToGame() { +std::shared_ptr<IStorage> AppletDataBroker::PopNormalDataToGame() { if (out_channel.empty()) return nullptr; auto out = std::move(out_channel.front()); out_channel.pop_front(); + pop_out_data_event.writable->Clear(); return out; } -std::unique_ptr<IStorage> AppletDataBroker::PopNormalDataToApplet() { +std::shared_ptr<IStorage> AppletDataBroker::PopNormalDataToApplet() { if (in_channel.empty()) return nullptr; @@ -68,16 +69,17 @@ std::unique_ptr<IStorage> AppletDataBroker::PopNormalDataToApplet() { return out; } -std::unique_ptr<IStorage> AppletDataBroker::PopInteractiveDataToGame() { +std::shared_ptr<IStorage> AppletDataBroker::PopInteractiveDataToGame() { if (out_interactive_channel.empty()) return nullptr; auto out = std::move(out_interactive_channel.front()); out_interactive_channel.pop_front(); + pop_interactive_out_data_event.writable->Clear(); return out; } -std::unique_ptr<IStorage> AppletDataBroker::PopInteractiveDataToApplet() { +std::shared_ptr<IStorage> AppletDataBroker::PopInteractiveDataToApplet() { if (in_interactive_channel.empty()) return nullptr; @@ -86,21 +88,21 @@ std::unique_ptr<IStorage> AppletDataBroker::PopInteractiveDataToApplet() { return out; } -void AppletDataBroker::PushNormalDataFromGame(IStorage storage) { - in_channel.push_back(std::make_unique<IStorage>(storage)); +void AppletDataBroker::PushNormalDataFromGame(std::shared_ptr<IStorage>&& storage) { + in_channel.emplace_back(std::move(storage)); } -void AppletDataBroker::PushNormalDataFromApplet(IStorage storage) { - out_channel.push_back(std::make_unique<IStorage>(storage)); +void AppletDataBroker::PushNormalDataFromApplet(std::shared_ptr<IStorage>&& storage) { + out_channel.emplace_back(std::move(storage)); pop_out_data_event.writable->Signal(); } -void AppletDataBroker::PushInteractiveDataFromGame(IStorage storage) { - in_interactive_channel.push_back(std::make_unique<IStorage>(storage)); +void AppletDataBroker::PushInteractiveDataFromGame(std::shared_ptr<IStorage>&& storage) { + in_interactive_channel.emplace_back(std::move(storage)); } -void AppletDataBroker::PushInteractiveDataFromApplet(IStorage storage) { - out_interactive_channel.push_back(std::make_unique<IStorage>(storage)); +void AppletDataBroker::PushInteractiveDataFromApplet(std::shared_ptr<IStorage>&& storage) { + out_interactive_channel.emplace_back(std::move(storage)); pop_interactive_out_data_event.writable->Signal(); } diff --git a/src/core/hle/service/am/applets/applets.h b/src/core/hle/service/am/applets/applets.h index 16e61fc6f..e75be86a2 100644 --- a/src/core/hle/service/am/applets/applets.h +++ b/src/core/hle/service/am/applets/applets.h @@ -72,17 +72,17 @@ public: // Retrieves but does not pop the data sent to applet. RawChannelData PeekDataToAppletForDebug() const; - std::unique_ptr<IStorage> PopNormalDataToGame(); - std::unique_ptr<IStorage> PopNormalDataToApplet(); + std::shared_ptr<IStorage> PopNormalDataToGame(); + std::shared_ptr<IStorage> PopNormalDataToApplet(); - std::unique_ptr<IStorage> PopInteractiveDataToGame(); - std::unique_ptr<IStorage> PopInteractiveDataToApplet(); + std::shared_ptr<IStorage> PopInteractiveDataToGame(); + std::shared_ptr<IStorage> PopInteractiveDataToApplet(); - void PushNormalDataFromGame(IStorage storage); - void PushNormalDataFromApplet(IStorage storage); + void PushNormalDataFromGame(std::shared_ptr<IStorage>&& storage); + void PushNormalDataFromApplet(std::shared_ptr<IStorage>&& storage); - void PushInteractiveDataFromGame(IStorage storage); - void PushInteractiveDataFromApplet(IStorage storage); + void PushInteractiveDataFromGame(std::shared_ptr<IStorage>&& storage); + void PushInteractiveDataFromApplet(std::shared_ptr<IStorage>&& storage); void SignalStateChanged() const; @@ -94,16 +94,16 @@ private: // Queues are named from applet's perspective // PopNormalDataToApplet and PushNormalDataFromGame - std::deque<std::unique_ptr<IStorage>> in_channel; + std::deque<std::shared_ptr<IStorage>> in_channel; // PopNormalDataToGame and PushNormalDataFromApplet - std::deque<std::unique_ptr<IStorage>> out_channel; + std::deque<std::shared_ptr<IStorage>> out_channel; // PopInteractiveDataToApplet and PushInteractiveDataFromGame - std::deque<std::unique_ptr<IStorage>> in_interactive_channel; + std::deque<std::shared_ptr<IStorage>> in_interactive_channel; // PopInteractiveDataToGame and PushInteractiveDataFromApplet - std::deque<std::unique_ptr<IStorage>> out_interactive_channel; + std::deque<std::shared_ptr<IStorage>> out_interactive_channel; Kernel::EventPair state_changed_event; diff --git a/src/core/hle/service/am/applets/error.cpp b/src/core/hle/service/am/applets/error.cpp index eab0d42c9..f12fd7f89 100644 --- a/src/core/hle/service/am/applets/error.cpp +++ b/src/core/hle/service/am/applets/error.cpp @@ -186,7 +186,7 @@ void Error::Execute() { void Error::DisplayCompleted() { complete = true; - broker.PushNormalDataFromApplet(IStorage{{}}); + broker.PushNormalDataFromApplet(std::make_shared<IStorage>(std::vector<u8>{})); broker.SignalStateChanged(); } diff --git a/src/core/hle/service/am/applets/general_backend.cpp b/src/core/hle/service/am/applets/general_backend.cpp index 328438a1d..104501ac5 100644 --- a/src/core/hle/service/am/applets/general_backend.cpp +++ b/src/core/hle/service/am/applets/general_backend.cpp @@ -20,7 +20,7 @@ namespace Service::AM::Applets { constexpr ResultCode ERROR_INVALID_PIN{ErrorModule::PCTL, 221}; static void LogCurrentStorage(AppletDataBroker& broker, std::string_view prefix) { - std::unique_ptr<IStorage> storage = broker.PopNormalDataToApplet(); + std::shared_ptr<IStorage> storage = broker.PopNormalDataToApplet(); for (; storage != nullptr; storage = broker.PopNormalDataToApplet()) { const auto data = storage->GetData(); LOG_INFO(Service_AM, @@ -148,7 +148,7 @@ void Auth::AuthFinished(bool successful) { std::vector<u8> out(sizeof(Return)); std::memcpy(out.data(), &return_, sizeof(Return)); - broker.PushNormalDataFromApplet(IStorage{out}); + broker.PushNormalDataFromApplet(std::make_shared<IStorage>(std::move(out))); broker.SignalStateChanged(); } @@ -198,7 +198,7 @@ void PhotoViewer::Execute() { } void PhotoViewer::ViewFinished() { - broker.PushNormalDataFromApplet(IStorage{{}}); + broker.PushNormalDataFromApplet(std::make_shared<IStorage>(std::vector<u8>{})); broker.SignalStateChanged(); } @@ -234,8 +234,8 @@ void StubApplet::ExecuteInteractive() { LOG_WARNING(Service_AM, "called (STUBBED)"); LogCurrentStorage(broker, "ExecuteInteractive"); - broker.PushNormalDataFromApplet(IStorage{std::vector<u8>(0x1000)}); - broker.PushInteractiveDataFromApplet(IStorage{std::vector<u8>(0x1000)}); + broker.PushNormalDataFromApplet(std::make_shared<IStorage>(std::vector<u8>(0x1000))); + broker.PushInteractiveDataFromApplet(std::make_shared<IStorage>(std::vector<u8>(0x1000))); broker.SignalStateChanged(); } @@ -243,8 +243,8 @@ void StubApplet::Execute() { LOG_WARNING(Service_AM, "called (STUBBED)"); LogCurrentStorage(broker, "Execute"); - broker.PushNormalDataFromApplet(IStorage{std::vector<u8>(0x1000)}); - broker.PushInteractiveDataFromApplet(IStorage{std::vector<u8>(0x1000)}); + broker.PushNormalDataFromApplet(std::make_shared<IStorage>(std::vector<u8>(0x1000))); + broker.PushInteractiveDataFromApplet(std::make_shared<IStorage>(std::vector<u8>(0x1000))); broker.SignalStateChanged(); } diff --git a/src/core/hle/service/am/applets/profile_select.cpp b/src/core/hle/service/am/applets/profile_select.cpp index 3eba696ca..70cc23552 100644 --- a/src/core/hle/service/am/applets/profile_select.cpp +++ b/src/core/hle/service/am/applets/profile_select.cpp @@ -50,7 +50,7 @@ void ProfileSelect::ExecuteInteractive() { void ProfileSelect::Execute() { if (complete) { - broker.PushNormalDataFromApplet(IStorage{final_data}); + broker.PushNormalDataFromApplet(std::make_shared<IStorage>(std::move(final_data))); return; } @@ -71,7 +71,7 @@ void ProfileSelect::SelectionComplete(std::optional<Common::UUID> uuid) { final_data = std::vector<u8>(sizeof(UserSelectionOutput)); std::memcpy(final_data.data(), &output, final_data.size()); - broker.PushNormalDataFromApplet(IStorage{final_data}); + broker.PushNormalDataFromApplet(std::make_shared<IStorage>(std::move(final_data))); broker.SignalStateChanged(); } diff --git a/src/core/hle/service/am/applets/software_keyboard.cpp b/src/core/hle/service/am/applets/software_keyboard.cpp index 748559cd0..54e63c138 100644 --- a/src/core/hle/service/am/applets/software_keyboard.cpp +++ b/src/core/hle/service/am/applets/software_keyboard.cpp @@ -102,7 +102,8 @@ void SoftwareKeyboard::ExecuteInteractive() { void SoftwareKeyboard::Execute() { if (complete) { - broker.PushNormalDataFromApplet(IStorage{final_data}); + broker.PushNormalDataFromApplet(std::make_shared<IStorage>(std::move(final_data))); + broker.SignalStateChanged(); return; } @@ -119,7 +120,7 @@ void SoftwareKeyboard::WriteText(std::optional<std::u16string> text) { std::vector<u8> output_sub(SWKBD_OUTPUT_BUFFER_SIZE); if (config.utf_8) { - const u64 size = text->size() + 8; + const u64 size = text->size() + sizeof(u64); const auto new_text = Common::UTF16ToUTF8(*text); std::memcpy(output_sub.data(), &size, sizeof(u64)); @@ -130,7 +131,7 @@ void SoftwareKeyboard::WriteText(std::optional<std::u16string> text) { std::memcpy(output_main.data() + 4, new_text.data(), std::min(new_text.size(), SWKBD_OUTPUT_BUFFER_SIZE - 4)); } else { - const u64 size = text->size() * 2 + 8; + const u64 size = text->size() * 2 + sizeof(u64); std::memcpy(output_sub.data(), &size, sizeof(u64)); std::memcpy(output_sub.data() + 8, text->data(), std::min(text->size() * 2, SWKBD_OUTPUT_BUFFER_SIZE - 8)); @@ -144,15 +145,15 @@ void SoftwareKeyboard::WriteText(std::optional<std::u16string> text) { final_data = output_main; if (complete) { - broker.PushNormalDataFromApplet(IStorage{output_main}); + broker.PushNormalDataFromApplet(std::make_shared<IStorage>(std::move(output_main))); broker.SignalStateChanged(); } else { - broker.PushInteractiveDataFromApplet(IStorage{output_sub}); + broker.PushInteractiveDataFromApplet(std::make_shared<IStorage>(std::move(output_sub))); } } else { output_main[0] = 1; complete = true; - broker.PushNormalDataFromApplet(IStorage{output_main}); + broker.PushNormalDataFromApplet(std::make_shared<IStorage>(std::move(output_main))); broker.SignalStateChanged(); } } diff --git a/src/core/hle/service/am/applets/web_browser.cpp b/src/core/hle/service/am/applets/web_browser.cpp index 5546ef6e8..12443c910 100644 --- a/src/core/hle/service/am/applets/web_browser.cpp +++ b/src/core/hle/service/am/applets/web_browser.cpp @@ -284,7 +284,7 @@ void WebBrowser::Finalize() { std::vector<u8> data(sizeof(WebCommonReturnValue)); std::memcpy(data.data(), &out, sizeof(WebCommonReturnValue)); - broker.PushNormalDataFromApplet(IStorage{data}); + broker.PushNormalDataFromApplet(std::make_shared<IStorage>(std::move(data))); broker.SignalStateChanged(); if (!temporary_dir.empty() && FileUtil::IsDirectory(temporary_dir)) { diff --git a/src/core/hle/service/audio/hwopus.cpp b/src/core/hle/service/audio/hwopus.cpp index cb839e4a2..d19513cbb 100644 --- a/src/core/hle/service/audio/hwopus.cpp +++ b/src/core/hle/service/audio/hwopus.cpp @@ -170,8 +170,10 @@ public: {3, nullptr, "SetContextForMultiStream"}, {4, &IHardwareOpusDecoderManager::DecodeInterleavedWithPerfOld, "DecodeInterleavedWithPerfOld"}, {5, nullptr, "DecodeInterleavedForMultiStreamWithPerfOld"}, - {6, &IHardwareOpusDecoderManager::DecodeInterleaved, "DecodeInterleaved"}, - {7, nullptr, "DecodeInterleavedForMultiStream"}, + {6, &IHardwareOpusDecoderManager::DecodeInterleaved, "DecodeInterleavedWithPerfAndResetOld"}, + {7, nullptr, "DecodeInterleavedForMultiStreamWithPerfAndResetOld"}, + {8, &IHardwareOpusDecoderManager::DecodeInterleaved, "DecodeInterleaved"}, + {9, nullptr, "DecodeInterleavedForMultiStream"}, }; // clang-format on diff --git a/src/core/hle/service/bcat/backend/backend.cpp b/src/core/hle/service/bcat/backend/backend.cpp index 6f5ea095a..def3410cc 100644 --- a/src/core/hle/service/bcat/backend/backend.cpp +++ b/src/core/hle/service/bcat/backend/backend.cpp @@ -117,13 +117,13 @@ bool NullBackend::SynchronizeDirectory(TitleIDVersion title, std::string name, } bool NullBackend::Clear(u64 title_id) { - LOG_DEBUG(Service_BCAT, "called, title_id={:016X}"); + LOG_DEBUG(Service_BCAT, "called, title_id={:016X}", title_id); return true; } void NullBackend::SetPassphrase(u64 title_id, const Passphrase& passphrase) { - LOG_DEBUG(Service_BCAT, "called, title_id={:016X}, passphrase = {}", title_id, + LOG_DEBUG(Service_BCAT, "called, title_id={:016X}, passphrase={}", title_id, Common::HexToString(passphrase)); } diff --git a/src/core/hle/service/filesystem/fsp_srv.cpp b/src/core/hle/service/filesystem/fsp_srv.cpp index 55d62fc5e..e6811d5b5 100644 --- a/src/core/hle/service/filesystem/fsp_srv.cpp +++ b/src/core/hle/service/filesystem/fsp_srv.cpp @@ -420,7 +420,7 @@ public: return; } - IFile file(result.Unwrap()); + auto file = std::make_shared<IFile>(result.Unwrap()); IPC::ResponseBuilder rb{ctx, 2, 0, 1}; rb.Push(RESULT_SUCCESS); @@ -445,7 +445,7 @@ public: return; } - IDirectory directory(result.Unwrap()); + auto directory = std::make_shared<IDirectory>(result.Unwrap()); IPC::ResponseBuilder rb{ctx, 2, 0, 1}; rb.Push(RESULT_SUCCESS); @@ -794,8 +794,8 @@ void FSP_SRV::OpenFileSystemWithPatch(Kernel::HLERequestContext& ctx) { void FSP_SRV::OpenSdCardFileSystem(Kernel::HLERequestContext& ctx) { LOG_DEBUG(Service_FS, "called"); - IFileSystem filesystem(fsc.OpenSDMC().Unwrap(), - SizeGetter::FromStorageId(fsc, FileSys::StorageId::SdCard)); + auto filesystem = std::make_shared<IFileSystem>( + fsc.OpenSDMC().Unwrap(), SizeGetter::FromStorageId(fsc, FileSys::StorageId::SdCard)); IPC::ResponseBuilder rb{ctx, 2, 0, 1}; rb.Push(RESULT_SUCCESS); @@ -846,7 +846,8 @@ void FSP_SRV::OpenSaveDataFileSystem(Kernel::HLERequestContext& ctx) { id = FileSys::StorageId::NandSystem; } - IFileSystem filesystem(std::move(dir.Unwrap()), SizeGetter::FromStorageId(fsc, id)); + auto filesystem = + std::make_shared<IFileSystem>(std::move(dir.Unwrap()), SizeGetter::FromStorageId(fsc, id)); IPC::ResponseBuilder rb{ctx, 2, 0, 1}; rb.Push(RESULT_SUCCESS); @@ -898,7 +899,7 @@ void FSP_SRV::OpenDataStorageByCurrentProcess(Kernel::HLERequestContext& ctx) { return; } - IStorage storage(std::move(romfs.Unwrap())); + auto storage = std::make_shared<IStorage>(std::move(romfs.Unwrap())); IPC::ResponseBuilder rb{ctx, 2, 0, 1}; rb.Push(RESULT_SUCCESS); @@ -937,7 +938,8 @@ void FSP_SRV::OpenDataStorageByDataId(Kernel::HLERequestContext& ctx) { FileSys::PatchManager pm{title_id}; - IStorage storage(pm.PatchRomFS(std::move(data.Unwrap()), 0, FileSys::ContentRecordType::Data)); + auto storage = std::make_shared<IStorage>( + pm.PatchRomFS(std::move(data.Unwrap()), 0, FileSys::ContentRecordType::Data)); IPC::ResponseBuilder rb{ctx, 2, 0, 1}; rb.Push(RESULT_SUCCESS); diff --git a/src/core/hle/service/prepo/prepo.cpp b/src/core/hle/service/prepo/prepo.cpp index 5eb26caf8..8f1be0e48 100644 --- a/src/core/hle/service/prepo/prepo.cpp +++ b/src/core/hle/service/prepo/prepo.cpp @@ -50,16 +50,16 @@ private: IPC::RequestParser rp{ctx}; const auto process_id = rp.PopRaw<u64>(); - const auto data1 = ctx.ReadBuffer(0); - const auto data2 = ctx.ReadBuffer(1); + std::vector<std::vector<u8>> data{ctx.ReadBuffer(0)}; + if (Type == Core::Reporter::PlayReportType::New) { + data.emplace_back(ctx.ReadBuffer(1)); + } - LOG_DEBUG(Service_PREPO, - "called, type={:02X}, process_id={:016X}, data1_size={:016X}, data2_size={:016X}", - static_cast<u8>(Type), process_id, data1.size(), data2.size()); + LOG_DEBUG(Service_PREPO, "called, type={:02X}, process_id={:016X}, data1_size={:016X}", + static_cast<u8>(Type), process_id, data[0].size()); const auto& reporter{system.GetReporter()}; - reporter.SavePlayReport(Type, system.CurrentProcess()->GetTitleID(), {data1, data2}, - process_id); + reporter.SavePlayReport(Type, system.CurrentProcess()->GetTitleID(), data, process_id); IPC::ResponseBuilder rb{ctx, 2}; rb.Push(RESULT_SUCCESS); @@ -70,19 +70,19 @@ private: IPC::RequestParser rp{ctx}; const auto user_id = rp.PopRaw<u128>(); const auto process_id = rp.PopRaw<u64>(); - - const auto data1 = ctx.ReadBuffer(0); - const auto data2 = ctx.ReadBuffer(1); + std::vector<std::vector<u8>> data{ctx.ReadBuffer(0)}; + if (Type == Core::Reporter::PlayReportType::New) { + data.emplace_back(ctx.ReadBuffer(1)); + } LOG_DEBUG( Service_PREPO, - "called, type={:02X}, user_id={:016X}{:016X}, process_id={:016X}, data1_size={:016X}, " - "data2_size={:016X}", - static_cast<u8>(Type), user_id[1], user_id[0], process_id, data1.size(), data2.size()); + "called, type={:02X}, user_id={:016X}{:016X}, process_id={:016X}, data1_size={:016X}", + static_cast<u8>(Type), user_id[1], user_id[0], process_id, data[0].size()); const auto& reporter{system.GetReporter()}; - reporter.SavePlayReport(Type, system.CurrentProcess()->GetTitleID(), {data1, data2}, - process_id, user_id); + reporter.SavePlayReport(Type, system.CurrentProcess()->GetTitleID(), data, process_id, + user_id); IPC::ResponseBuilder rb{ctx, 2}; rb.Push(RESULT_SUCCESS); diff --git a/src/video_core/engines/maxwell_3d.cpp b/src/video_core/engines/maxwell_3d.cpp index 7cea146f0..0b3e8749b 100644 --- a/src/video_core/engines/maxwell_3d.cpp +++ b/src/video_core/engines/maxwell_3d.cpp @@ -9,6 +9,7 @@ #include "core/core_timing.h" #include "video_core/engines/maxwell_3d.h" #include "video_core/engines/shader_type.h" +#include "video_core/gpu.h" #include "video_core/memory_manager.h" #include "video_core/rasterizer_interface.h" #include "video_core/textures/texture.h" @@ -519,61 +520,63 @@ void Maxwell3D::ProcessFirmwareCall4() { regs.reg_array[0xd00] = 1; } -void Maxwell3D::ProcessQueryGet() { +void Maxwell3D::StampQueryResult(u64 payload, bool long_query) { + struct LongQueryResult { + u64_le value; + u64_le timestamp; + }; + static_assert(sizeof(LongQueryResult) == 16, "LongQueryResult has wrong size"); const GPUVAddr sequence_address{regs.query.QueryAddress()}; - // Since the sequence address is given as a GPU VAddr, we have to convert it to an application - // VAddr before writing. + if (long_query) { + // Write the 128-bit result structure in long mode. Note: We emulate an infinitely fast + // GPU, this command may actually take a while to complete in real hardware due to GPU + // wait queues. + LongQueryResult query_result{payload, system.GPU().GetTicks()}; + memory_manager.WriteBlock(sequence_address, &query_result, sizeof(query_result)); + } else { + memory_manager.Write<u32>(sequence_address, static_cast<u32>(payload)); + } +} +void Maxwell3D::ProcessQueryGet() { // TODO(Subv): Support the other query units. ASSERT_MSG(regs.query.query_get.unit == Regs::QueryUnit::Crop, "Units other than CROP are unimplemented"); - u64 result = 0; - - // TODO(Subv): Support the other query variables - switch (regs.query.query_get.select) { - case Regs::QuerySelect::Zero: - // This seems to actually write the query sequence to the query address. - result = regs.query.query_sequence; + switch (regs.query.query_get.operation) { + case Regs::QueryOperation::Release: { + const u64 result = regs.query.query_sequence; + StampQueryResult(result, regs.query.query_get.short_query == 0); break; - default: - result = 1; - UNIMPLEMENTED_MSG("Unimplemented query select type {}", - static_cast<u32>(regs.query.query_get.select.Value())); } - - // TODO(Subv): Research and implement how query sync conditions work. - - struct LongQueryResult { - u64_le value; - u64_le timestamp; - }; - static_assert(sizeof(LongQueryResult) == 16, "LongQueryResult has wrong size"); - - switch (regs.query.query_get.mode) { - case Regs::QueryMode::Write: - case Regs::QueryMode::Write2: { - u32 sequence = regs.query.query_sequence; - if (regs.query.query_get.short_query) { - // Write the current query sequence to the sequence address. - // TODO(Subv): Find out what happens if you use a long query type but mark it as a short - // query. - memory_manager.Write<u32>(sequence_address, sequence); - } else { - // Write the 128-bit result structure in long mode. Note: We emulate an infinitely fast - // GPU, this command may actually take a while to complete in real hardware due to GPU - // wait queues. - LongQueryResult query_result{}; - query_result.value = result; - // TODO(Subv): Generate a real GPU timestamp and write it here instead of CoreTiming - query_result.timestamp = system.CoreTiming().GetTicks(); - memory_manager.WriteBlock(sequence_address, &query_result, sizeof(query_result)); + case Regs::QueryOperation::Acquire: { + // Todo(Blinkhawk): Under this operation, the GPU waits for the CPU + // to write a value that matches the current payload. + UNIMPLEMENTED_MSG("Unimplemented query operation ACQUIRE"); + break; + } + case Regs::QueryOperation::Counter: { + u64 result{}; + switch (regs.query.query_get.select) { + case Regs::QuerySelect::Zero: + result = 0; + break; + default: + result = 1; + UNIMPLEMENTED_MSG("Unimplemented query select type {}", + static_cast<u32>(regs.query.query_get.select.Value())); } + StampQueryResult(result, regs.query.query_get.short_query == 0); + break; + } + case Regs::QueryOperation::Trap: { + UNIMPLEMENTED_MSG("Unimplemented query operation TRAP"); + break; + } + default: { + UNIMPLEMENTED_MSG("Unknown query operation"); break; } - default: - UNIMPLEMENTED_MSG("Query mode {} not implemented", - static_cast<u32>(regs.query.query_get.mode.Value())); } } diff --git a/src/video_core/engines/maxwell_3d.h b/src/video_core/engines/maxwell_3d.h index 8808bbf76..0a2af54e5 100644 --- a/src/video_core/engines/maxwell_3d.h +++ b/src/video_core/engines/maxwell_3d.h @@ -71,12 +71,11 @@ public: static constexpr std::size_t MaxConstBuffers = 18; static constexpr std::size_t MaxConstBufferSize = 0x10000; - enum class QueryMode : u32 { - Write = 0, - Sync = 1, - // TODO(Subv): It is currently unknown what the difference between method 2 and method 0 - // is. - Write2 = 2, + enum class QueryOperation : u32 { + Release = 0, + Acquire = 1, + Counter = 2, + Trap = 3, }; enum class QueryUnit : u32 { @@ -704,8 +703,8 @@ public: INSERT_UNION_PADDING_WORDS(0x15); s32 stencil_back_func_ref; - u32 stencil_back_func_mask; u32 stencil_back_mask; + u32 stencil_back_func_mask; INSERT_UNION_PADDING_WORDS(0xC); @@ -862,7 +861,11 @@ public: float point_size; - INSERT_UNION_PADDING_WORDS(0x7); + INSERT_UNION_PADDING_WORDS(0x1); + + u32 point_sprite_enable; + + INSERT_UNION_PADDING_WORDS(0x5); u32 zeta_enable; @@ -1077,7 +1080,7 @@ public: u32 query_sequence; union { u32 raw; - BitField<0, 2, QueryMode> mode; + BitField<0, 2, QueryOperation> operation; BitField<4, 1, u32> fence; BitField<12, 4, QueryUnit> unit; BitField<16, 1, QuerySyncCondition> sync_cond; @@ -1409,6 +1412,9 @@ private: /// Handles a write to the QUERY_GET register. void ProcessQueryGet(); + // Writes the query result accordingly + void StampQueryResult(u64 payload, bool long_query); + // Handles Conditional Rendering void ProcessQueryCondition(); @@ -1458,8 +1464,8 @@ ASSERT_REG_POSITION(polygon_offset_fill_enable, 0x372); ASSERT_REG_POSITION(patch_vertices, 0x373); ASSERT_REG_POSITION(scissor_test, 0x380); ASSERT_REG_POSITION(stencil_back_func_ref, 0x3D5); -ASSERT_REG_POSITION(stencil_back_func_mask, 0x3D6); -ASSERT_REG_POSITION(stencil_back_mask, 0x3D7); +ASSERT_REG_POSITION(stencil_back_mask, 0x3D6); +ASSERT_REG_POSITION(stencil_back_func_mask, 0x3D7); ASSERT_REG_POSITION(color_mask_common, 0x3E4); ASSERT_REG_POSITION(rt_separate_frag_data, 0x3EB); ASSERT_REG_POSITION(depth_bounds, 0x3E7); @@ -1494,6 +1500,7 @@ ASSERT_REG_POSITION(vb_element_base, 0x50D); ASSERT_REG_POSITION(vb_base_instance, 0x50E); ASSERT_REG_POSITION(clip_distance_enabled, 0x544); ASSERT_REG_POSITION(point_size, 0x546); +ASSERT_REG_POSITION(point_sprite_enable, 0x548); ASSERT_REG_POSITION(zeta_enable, 0x54E); ASSERT_REG_POSITION(multisample_control, 0x54F); ASSERT_REG_POSITION(condition, 0x554); diff --git a/src/video_core/engines/shader_bytecode.h b/src/video_core/engines/shader_bytecode.h index 955f981bd..c9bc83cd7 100644 --- a/src/video_core/engines/shader_bytecode.h +++ b/src/video_core/engines/shader_bytecode.h @@ -624,6 +624,19 @@ enum class ShuffleOperation : u64 { Bfly = 3, // shuffleXorNV }; +enum class ShfType : u64 { + Bits32 = 0, + U64 = 2, + S64 = 3, +}; + +enum class ShfXmode : u64 { + None = 0, + HI = 1, + X = 2, + XHI = 3, +}; + union Instruction { constexpr Instruction& operator=(const Instruction& instr) { value = instr.value; @@ -776,6 +789,13 @@ union Instruction { } shr; union { + BitField<37, 2, ShfType> type; + BitField<48, 2, ShfXmode> xmode; + BitField<50, 1, u64> wrap; + BitField<20, 6, u64> immediate; + } shf; + + union { BitField<39, 5, u64> shift_amount; BitField<48, 1, u64> negate_b; BitField<49, 1, u64> negate_a; diff --git a/src/video_core/gpu.cpp b/src/video_core/gpu.cpp index b9c5c41a2..4419ab735 100644 --- a/src/video_core/gpu.cpp +++ b/src/video_core/gpu.cpp @@ -6,6 +6,7 @@ #include "common/microprofile.h" #include "core/core.h" #include "core/core_timing.h" +#include "core/core_timing_util.h" #include "core/memory.h" #include "video_core/engines/fermi_2d.h" #include "video_core/engines/kepler_compute.h" @@ -23,7 +24,7 @@ MICROPROFILE_DEFINE(GPU_wait, "GPU", "Wait for the GPU", MP_RGB(128, 128, 192)); GPU::GPU(Core::System& system, VideoCore::RendererBase& renderer, bool is_async) : system{system}, renderer{renderer}, is_async{is_async} { auto& rasterizer{renderer.Rasterizer()}; - memory_manager = std::make_unique<Tegra::MemoryManager>(system, rasterizer); + memory_manager = std::make_unique<Tegra::MemoryManager>(system); dma_pusher = std::make_unique<Tegra::DmaPusher>(*this); maxwell_3d = std::make_unique<Engines::Maxwell3D>(system, rasterizer, *memory_manager); fermi_2d = std::make_unique<Engines::Fermi2D>(rasterizer); @@ -122,6 +123,19 @@ bool GPU::CancelSyncptInterrupt(const u32 syncpoint_id, const u32 value) { return true; } +u64 GPU::GetTicks() const { + // This values were reversed engineered by fincs from NVN + // The gpu clock is reported in units of 385/625 nanoseconds + constexpr u64 gpu_ticks_num = 384; + constexpr u64 gpu_ticks_den = 625; + + const u64 cpu_ticks = system.CoreTiming().GetTicks(); + const u64 nanoseconds = Core::Timing::CyclesToNs(cpu_ticks).count(); + const u64 nanoseconds_num = nanoseconds / gpu_ticks_den; + const u64 nanoseconds_rem = nanoseconds % gpu_ticks_den; + return nanoseconds_num * gpu_ticks_num + (nanoseconds_rem * gpu_ticks_num) / gpu_ticks_den; +} + void GPU::FlushCommands() { renderer.Rasterizer().FlushCommands(); } @@ -340,7 +354,7 @@ void GPU::ProcessSemaphoreTriggerMethod() { block.sequence = regs.semaphore_sequence; // TODO(Kmather73): Generate a real GPU timestamp and write it here instead of // CoreTiming - block.timestamp = system.CoreTiming().GetTicks(); + block.timestamp = GetTicks(); memory_manager->WriteBlock(regs.semaphore_address.SemaphoreAddress(), &block, sizeof(block)); } else { diff --git a/src/video_core/gpu.h b/src/video_core/gpu.h index b648317bb..07727210c 100644 --- a/src/video_core/gpu.h +++ b/src/video_core/gpu.h @@ -192,6 +192,8 @@ public: bool CancelSyncptInterrupt(u32 syncpoint_id, u32 value); + u64 GetTicks() const; + std::unique_lock<std::mutex> LockSync() { return std::unique_lock{sync_mutex}; } diff --git a/src/video_core/gpu_thread.h b/src/video_core/gpu_thread.h index 08dc96bb3..882e2d9c7 100644 --- a/src/video_core/gpu_thread.h +++ b/src/video_core/gpu_thread.h @@ -86,7 +86,7 @@ struct CommandDataContainer { struct SynchState final { std::atomic_bool is_running{true}; - using CommandQueue = Common::SPSCQueue<CommandDataContainer>; + using CommandQueue = Common::MPSCQueue<CommandDataContainer>; CommandQueue queue; u64 last_fence{}; std::atomic<u64> signaled_fence{}; diff --git a/src/video_core/memory_manager.cpp b/src/video_core/memory_manager.cpp index 11848fbce..f1d50be3e 100644 --- a/src/video_core/memory_manager.cpp +++ b/src/video_core/memory_manager.cpp @@ -9,13 +9,12 @@ #include "core/hle/kernel/process.h" #include "core/hle/kernel/vm_manager.h" #include "core/memory.h" +#include "video_core/gpu.h" #include "video_core/memory_manager.h" -#include "video_core/rasterizer_interface.h" namespace Tegra { -MemoryManager::MemoryManager(Core::System& system, VideoCore::RasterizerInterface& rasterizer) - : rasterizer{rasterizer}, system{system} { +MemoryManager::MemoryManager(Core::System& system) : system{system} { std::fill(page_table.pointers.begin(), page_table.pointers.end(), nullptr); std::fill(page_table.attributes.begin(), page_table.attributes.end(), Common::PageType::Unmapped); @@ -84,7 +83,8 @@ GPUVAddr MemoryManager::UnmapBuffer(GPUVAddr gpu_addr, u64 size) { const auto cpu_addr = GpuToCpuAddress(gpu_addr); ASSERT(cpu_addr); - rasterizer.FlushAndInvalidateRegion(cache_addr, aligned_size); + system.GPU().FlushAndInvalidateRegion(cache_addr, aligned_size); + UnmapRange(gpu_addr, aligned_size); ASSERT(system.CurrentProcess() ->VMManager() @@ -242,7 +242,7 @@ void MemoryManager::ReadBlock(GPUVAddr src_addr, void* dest_buffer, const std::s switch (page_table.attributes[page_index]) { case Common::PageType::Memory: { const u8* src_ptr{page_table.pointers[page_index] + page_offset}; - rasterizer.FlushRegion(ToCacheAddr(src_ptr), copy_amount); + system.GPU().FlushRegion(ToCacheAddr(src_ptr), copy_amount); std::memcpy(dest_buffer, src_ptr, copy_amount); break; } @@ -292,7 +292,7 @@ void MemoryManager::WriteBlock(GPUVAddr dest_addr, const void* src_buffer, const switch (page_table.attributes[page_index]) { case Common::PageType::Memory: { u8* dest_ptr{page_table.pointers[page_index] + page_offset}; - rasterizer.InvalidateRegion(ToCacheAddr(dest_ptr), copy_amount); + system.GPU().InvalidateRegion(ToCacheAddr(dest_ptr), copy_amount); std::memcpy(dest_ptr, src_buffer, copy_amount); break; } @@ -340,7 +340,7 @@ void MemoryManager::CopyBlock(GPUVAddr dest_addr, GPUVAddr src_addr, const std:: switch (page_table.attributes[page_index]) { case Common::PageType::Memory: { const u8* src_ptr{page_table.pointers[page_index] + page_offset}; - rasterizer.FlushRegion(ToCacheAddr(src_ptr), copy_amount); + system.GPU().FlushRegion(ToCacheAddr(src_ptr), copy_amount); WriteBlock(dest_addr, src_ptr, copy_amount); break; } diff --git a/src/video_core/memory_manager.h b/src/video_core/memory_manager.h index aea010087..393447eb4 100644 --- a/src/video_core/memory_manager.h +++ b/src/video_core/memory_manager.h @@ -10,10 +10,6 @@ #include "common/common_types.h" #include "common/page_table.h" -namespace VideoCore { -class RasterizerInterface; -} - namespace Core { class System; } @@ -51,7 +47,7 @@ struct VirtualMemoryArea { class MemoryManager final { public: - explicit MemoryManager(Core::System& system, VideoCore::RasterizerInterface& rasterizer); + explicit MemoryManager(Core::System& system); ~MemoryManager(); GPUVAddr AllocateSpace(u64 size, u64 align); @@ -176,7 +172,6 @@ private: Common::PageTable page_table{page_bits}; VMAMap vma_map; - VideoCore::RasterizerInterface& rasterizer; Core::System& system; }; diff --git a/src/video_core/renderer_opengl/gl_rasterizer.cpp b/src/video_core/renderer_opengl/gl_rasterizer.cpp index 362942e09..b0eb14c8b 100644 --- a/src/video_core/renderer_opengl/gl_rasterizer.cpp +++ b/src/video_core/renderer_opengl/gl_rasterizer.cpp @@ -248,9 +248,6 @@ void RasterizerOpenGL::SetupVertexInstances(GLuint vao) { } GLintptr RasterizerOpenGL::SetupIndexBuffer() { - if (accelerate_draw != AccelDraw::Indexed) { - return 0; - } MICROPROFILE_SCOPE(OpenGL_Index); const auto& regs = system.GPU().Maxwell3D().regs; const std::size_t size = CalculateIndexBufferSize(); @@ -546,7 +543,8 @@ void RasterizerOpenGL::Clear() { } } -void RasterizerOpenGL::DrawPrelude() { +void RasterizerOpenGL::Draw(bool is_indexed, bool is_instanced) { + MICROPROFILE_SCOPE(OpenGL_Drawing); auto& gpu = system.GPU().Maxwell3D(); SyncRasterizeEnable(state); @@ -567,9 +565,6 @@ void RasterizerOpenGL::DrawPrelude() { buffer_cache.Acquire(); - // Draw the vertex batch - const bool is_indexed = accelerate_draw == AccelDraw::Indexed; - std::size_t buffer_size = CalculateVertexArraysSize(); // Add space for index buffer @@ -596,7 +591,11 @@ void RasterizerOpenGL::DrawPrelude() { // Upload vertex and index data. SetupVertexBuffer(vao); SetupVertexInstances(vao); - index_buffer_offset = SetupIndexBuffer(); + + GLintptr index_buffer_offset; + if (is_indexed) { + index_buffer_offset = SetupIndexBuffer(); + } // Prepare packed bindings. bind_ubo_pushbuffer.Setup(); @@ -630,6 +629,7 @@ void RasterizerOpenGL::DrawPrelude() { // As all cached buffers are invalidated, we need to recheck their state. gpu.dirty.ResetVertexArrays(); } + gpu.dirty.memory_general = false; shader_program_manager->ApplyTo(state); state.Apply(); @@ -637,106 +637,33 @@ void RasterizerOpenGL::DrawPrelude() { if (texture_cache.TextureBarrier()) { glTextureBarrier(); } -} - -struct DrawParams { - bool is_indexed{}; - bool is_instanced{}; - GLenum primitive_mode{}; - GLint count{}; - GLint base_vertex{}; - - // Indexed settings - GLenum index_format{}; - GLintptr index_buffer_offset{}; - - // Instanced setting - GLint num_instances{}; - GLint base_instance{}; - - void DispatchDraw() { - if (is_indexed) { - const auto index_buffer_ptr = reinterpret_cast<const void*>(index_buffer_offset); - if (is_instanced) { - glDrawElementsInstancedBaseVertexBaseInstance(primitive_mode, count, index_format, - index_buffer_ptr, num_instances, - base_vertex, base_instance); - } else { - glDrawElementsBaseVertex(primitive_mode, count, index_format, index_buffer_ptr, - base_vertex); - } - } else { - if (is_instanced) { - glDrawArraysInstancedBaseInstance(primitive_mode, base_vertex, count, num_instances, - base_instance); - } else { - glDrawArrays(primitive_mode, base_vertex, count); - } - } - } -}; - -bool RasterizerOpenGL::DrawBatch(bool is_indexed) { - accelerate_draw = is_indexed ? AccelDraw::Indexed : AccelDraw::Arrays; - MICROPROFILE_SCOPE(OpenGL_Drawing); - - DrawPrelude(); - - auto& maxwell3d = system.GPU().Maxwell3D(); - const auto& regs = maxwell3d.regs; - const auto current_instance = maxwell3d.state.current_instance; - DrawParams draw_call{}; - draw_call.is_indexed = is_indexed; - draw_call.num_instances = static_cast<GLint>(1); - draw_call.base_instance = static_cast<GLint>(current_instance); - draw_call.is_instanced = current_instance > 0; - draw_call.primitive_mode = MaxwellToGL::PrimitiveTopology(regs.draw.topology); - if (draw_call.is_indexed) { - draw_call.count = static_cast<GLint>(regs.index_array.count); - draw_call.base_vertex = static_cast<GLint>(regs.vb_element_base); - draw_call.index_format = MaxwellToGL::IndexFormat(regs.index_array.format); - draw_call.index_buffer_offset = index_buffer_offset; + const GLuint base_instance = static_cast<GLuint>(gpu.regs.vb_base_instance); + const GLsizei num_instances = + static_cast<GLsizei>(is_instanced ? gpu.mme_draw.instance_count : 1); + if (is_indexed) { + const GLenum index_format = MaxwellToGL::IndexFormat(gpu.regs.index_array.format); + const GLint base_vertex = static_cast<GLint>(gpu.regs.vb_element_base); + const GLsizei num_vertices = static_cast<GLsizei>(gpu.regs.index_array.count); + glDrawElementsInstancedBaseVertexBaseInstance( + primitive_mode, num_vertices, index_format, + reinterpret_cast<const void*>(index_buffer_offset), num_instances, base_vertex, + base_instance); } else { - draw_call.count = static_cast<GLint>(regs.vertex_buffer.count); - draw_call.base_vertex = static_cast<GLint>(regs.vertex_buffer.first); + const GLint base_vertex = static_cast<GLint>(gpu.regs.vertex_buffer.first); + const GLsizei num_vertices = static_cast<GLsizei>(gpu.regs.vertex_buffer.count); + glDrawArraysInstancedBaseInstance(primitive_mode, base_vertex, num_vertices, num_instances, + base_instance); } - draw_call.DispatchDraw(); +} - maxwell3d.dirty.memory_general = false; - accelerate_draw = AccelDraw::Disabled; +bool RasterizerOpenGL::DrawBatch(bool is_indexed) { + Draw(is_indexed, false); return true; } bool RasterizerOpenGL::DrawMultiBatch(bool is_indexed) { - accelerate_draw = is_indexed ? AccelDraw::Indexed : AccelDraw::Arrays; - - MICROPROFILE_SCOPE(OpenGL_Drawing); - - DrawPrelude(); - - auto& maxwell3d = system.GPU().Maxwell3D(); - const auto& regs = maxwell3d.regs; - const auto& draw_setup = maxwell3d.mme_draw; - DrawParams draw_call{}; - draw_call.is_indexed = is_indexed; - draw_call.num_instances = static_cast<GLint>(draw_setup.instance_count); - draw_call.base_instance = static_cast<GLint>(regs.vb_base_instance); - draw_call.is_instanced = draw_setup.instance_count > 1; - draw_call.primitive_mode = MaxwellToGL::PrimitiveTopology(regs.draw.topology); - if (draw_call.is_indexed) { - draw_call.count = static_cast<GLint>(regs.index_array.count); - draw_call.base_vertex = static_cast<GLint>(regs.vb_element_base); - draw_call.index_format = MaxwellToGL::IndexFormat(regs.index_array.format); - draw_call.index_buffer_offset = index_buffer_offset; - } else { - draw_call.count = static_cast<GLint>(regs.vertex_buffer.count); - draw_call.base_vertex = static_cast<GLint>(regs.vertex_buffer.first); - } - draw_call.DispatchDraw(); - - maxwell3d.dirty.memory_general = false; - accelerate_draw = AccelDraw::Disabled; + Draw(is_indexed, true); return true; } @@ -1293,6 +1220,7 @@ void RasterizerOpenGL::SyncPointState() { // Limit the point size to 1 since nouveau sometimes sets a point size of 0 (and that's invalid // in OpenGL). state.point.program_control = regs.vp_point_size.enable != 0; + state.point.sprite = regs.point_sprite_enable != 0; state.point.size = std::max(1.0f, regs.point_size); } diff --git a/src/video_core/renderer_opengl/gl_rasterizer.h b/src/video_core/renderer_opengl/gl_rasterizer.h index 6a27cf497..0501f3828 100644 --- a/src/video_core/renderer_opengl/gl_rasterizer.h +++ b/src/video_core/renderer_opengl/gl_rasterizer.h @@ -103,7 +103,7 @@ private: std::size_t size); /// Syncs all the state, shaders, render targets and textures setting before a draw call. - void DrawPrelude(); + void Draw(bool is_indexed, bool is_instanced); /// Configures the current textures to use for the draw command. void SetupDrawTextures(std::size_t stage_index, const Shader& shader); @@ -220,12 +220,7 @@ private: GLintptr SetupIndexBuffer(); - GLintptr index_buffer_offset; - void SetupShaders(GLenum primitive_mode); - - enum class AccelDraw { Disabled, Arrays, Indexed }; - AccelDraw accelerate_draw = AccelDraw::Disabled; }; } // namespace OpenGL diff --git a/src/video_core/renderer_opengl/gl_state.cpp b/src/video_core/renderer_opengl/gl_state.cpp index cc185e9e1..ab1f7983c 100644 --- a/src/video_core/renderer_opengl/gl_state.cpp +++ b/src/video_core/renderer_opengl/gl_state.cpp @@ -128,6 +128,7 @@ void OpenGLState::ApplyClipDistances() { void OpenGLState::ApplyPointSize() { Enable(GL_PROGRAM_POINT_SIZE, cur_state.point.program_control, point.program_control); + Enable(GL_POINT_SPRITE, cur_state.point.sprite, point.sprite); if (UpdateValue(cur_state.point.size, point.size)) { glPointSize(point.size); } diff --git a/src/video_core/renderer_opengl/gl_state.h b/src/video_core/renderer_opengl/gl_state.h index 678e5cd89..4953eeda2 100644 --- a/src/video_core/renderer_opengl/gl_state.h +++ b/src/video_core/renderer_opengl/gl_state.h @@ -132,6 +132,7 @@ public: struct { bool program_control = false; // GL_PROGRAM_POINT_SIZE + bool sprite = false; // GL_POINT_SPRITE GLfloat size = 1.0f; // GL_POINT_SIZE } point; diff --git a/src/video_core/renderer_opengl/maxwell_to_gl.h b/src/video_core/renderer_opengl/maxwell_to_gl.h index ea4f35663..7ed505628 100644 --- a/src/video_core/renderer_opengl/maxwell_to_gl.h +++ b/src/video_core/renderer_opengl/maxwell_to_gl.h @@ -47,8 +47,7 @@ inline GLenum VertexType(Maxwell::VertexAttribute attrib) { case Maxwell::VertexAttribute::Size::Size_10_10_10_2: return GL_UNSIGNED_INT_2_10_10_10_REV; default: - LOG_CRITICAL(Render_OpenGL, "Unimplemented vertex size={}", attrib.SizeString()); - UNREACHABLE(); + LOG_ERROR(Render_OpenGL, "Unimplemented vertex size={}", attrib.SizeString()); return {}; } case Maxwell::VertexAttribute::Type::SignedInt: @@ -72,8 +71,7 @@ inline GLenum VertexType(Maxwell::VertexAttribute attrib) { case Maxwell::VertexAttribute::Size::Size_10_10_10_2: return GL_INT_2_10_10_10_REV; default: - LOG_CRITICAL(Render_OpenGL, "Unimplemented vertex size={}", attrib.SizeString()); - UNREACHABLE(); + LOG_ERROR(Render_OpenGL, "Unimplemented vertex size={}", attrib.SizeString()); return {}; } case Maxwell::VertexAttribute::Type::Float: @@ -89,13 +87,19 @@ inline GLenum VertexType(Maxwell::VertexAttribute attrib) { case Maxwell::VertexAttribute::Size::Size_32_32_32_32: return GL_FLOAT; default: - LOG_CRITICAL(Render_OpenGL, "Unimplemented vertex size={}", attrib.SizeString()); - UNREACHABLE(); + LOG_ERROR(Render_OpenGL, "Unimplemented vertex size={}", attrib.SizeString()); + return {}; + } + case Maxwell::VertexAttribute::Type::UnsignedScaled: + switch (attrib.size) { + case Maxwell::VertexAttribute::Size::Size_8_8: + return GL_UNSIGNED_BYTE; + default: + LOG_ERROR(Render_OpenGL, "Unimplemented vertex size={}", attrib.SizeString()); return {}; } default: - LOG_CRITICAL(Render_OpenGL, "Unimplemented vertex type={}", attrib.TypeString()); - UNREACHABLE(); + LOG_ERROR(Render_OpenGL, "Unimplemented vertex type={}", attrib.TypeString()); return {}; } } diff --git a/src/video_core/shader/decode/shift.cpp b/src/video_core/shader/decode/shift.cpp index d419e9c45..3b391d3e6 100644 --- a/src/video_core/shader/decode/shift.cpp +++ b/src/video_core/shader/decode/shift.cpp @@ -10,8 +10,80 @@ namespace VideoCommon::Shader { +using std::move; using Tegra::Shader::Instruction; using Tegra::Shader::OpCode; +using Tegra::Shader::ShfType; +using Tegra::Shader::ShfXmode; + +namespace { + +Node IsFull(Node shift) { + return Operation(OperationCode::LogicalIEqual, move(shift), Immediate(32)); +} + +Node Shift(OperationCode opcode, Node value, Node shift) { + Node is_full = Operation(OperationCode::LogicalIEqual, shift, Immediate(32)); + Node shifted = Operation(opcode, move(value), shift); + return Operation(OperationCode::Select, IsFull(move(shift)), Immediate(0), move(shifted)); +} + +Node ClampShift(Node shift, s32 size = 32) { + shift = Operation(OperationCode::IMax, move(shift), Immediate(0)); + return Operation(OperationCode::IMin, move(shift), Immediate(size)); +} + +Node WrapShift(Node shift, s32 size = 32) { + return Operation(OperationCode::UBitwiseAnd, move(shift), Immediate(size - 1)); +} + +Node ShiftRight(Node low, Node high, Node shift, Node low_shift, ShfType type) { + // These values are used when the shift value is less than 32 + Node less_low = Shift(OperationCode::ILogicalShiftRight, low, shift); + Node less_high = Shift(OperationCode::ILogicalShiftLeft, high, low_shift); + Node less = Operation(OperationCode::IBitwiseOr, move(less_high), move(less_low)); + + if (type == ShfType::Bits32) { + // On 32 bit shifts we are either full (shifting 32) or shifting less than 32 bits + return Operation(OperationCode::Select, IsFull(move(shift)), move(high), move(less)); + } + + // And these when it's larger than or 32 + const bool is_signed = type == ShfType::S64; + const auto opcode = SignedToUnsignedCode(OperationCode::IArithmeticShiftRight, is_signed); + Node reduced = Operation(OperationCode::IAdd, shift, Immediate(-32)); + Node greater = Shift(opcode, high, move(reduced)); + + Node is_less = Operation(OperationCode::LogicalILessThan, shift, Immediate(32)); + Node is_zero = Operation(OperationCode::LogicalIEqual, move(shift), Immediate(0)); + + Node value = Operation(OperationCode::Select, move(is_less), move(less), move(greater)); + return Operation(OperationCode::Select, move(is_zero), move(high), move(value)); +} + +Node ShiftLeft(Node low, Node high, Node shift, Node low_shift, ShfType type) { + // These values are used when the shift value is less than 32 + Node less_low = Operation(OperationCode::ILogicalShiftRight, low, low_shift); + Node less_high = Operation(OperationCode::ILogicalShiftLeft, high, shift); + Node less = Operation(OperationCode::IBitwiseOr, move(less_low), move(less_high)); + + if (type == ShfType::Bits32) { + // On 32 bit shifts we are either full (shifting 32) or shifting less than 32 bits + return Operation(OperationCode::Select, IsFull(move(shift)), move(low), move(less)); + } + + // And these when it's larger than or 32 + Node reduced = Operation(OperationCode::IAdd, shift, Immediate(-32)); + Node greater = Shift(OperationCode::ILogicalShiftLeft, move(low), move(reduced)); + + Node is_less = Operation(OperationCode::LogicalILessThan, shift, Immediate(32)); + Node is_zero = Operation(OperationCode::LogicalIEqual, move(shift), Immediate(0)); + + Node value = Operation(OperationCode::Select, move(is_less), move(less), move(greater)); + return Operation(OperationCode::Select, move(is_zero), move(high), move(value)); +} + +} // Anonymous namespace u32 ShaderIR::DecodeShift(NodeBlock& bb, u32 pc) { const Instruction instr = {program_code[pc]}; @@ -28,29 +100,48 @@ u32 ShaderIR::DecodeShift(NodeBlock& bb, u32 pc) { } }(); - switch (opcode->get().GetId()) { + switch (const auto opid = opcode->get().GetId(); opid) { case OpCode::Id::SHR_C: case OpCode::Id::SHR_R: case OpCode::Id::SHR_IMM: { - if (instr.shr.wrap) { - op_b = Operation(OperationCode::UBitwiseAnd, std::move(op_b), Immediate(0x1f)); - } else { - op_b = Operation(OperationCode::IMax, std::move(op_b), Immediate(0)); - op_b = Operation(OperationCode::IMin, std::move(op_b), Immediate(31)); - } + op_b = instr.shr.wrap ? WrapShift(move(op_b)) : ClampShift(move(op_b)); Node value = SignedOperation(OperationCode::IArithmeticShiftRight, instr.shift.is_signed, - std::move(op_a), std::move(op_b)); + move(op_a), move(op_b)); SetInternalFlagsFromInteger(bb, value, instr.generates_cc); - SetRegister(bb, instr.gpr0, std::move(value)); + SetRegister(bb, instr.gpr0, move(value)); break; } case OpCode::Id::SHL_C: case OpCode::Id::SHL_R: case OpCode::Id::SHL_IMM: { - const Node value = Operation(OperationCode::ILogicalShiftLeft, op_a, op_b); + Node value = Operation(OperationCode::ILogicalShiftLeft, op_a, op_b); SetInternalFlagsFromInteger(bb, value, instr.generates_cc); - SetRegister(bb, instr.gpr0, value); + SetRegister(bb, instr.gpr0, move(value)); + break; + } + case OpCode::Id::SHF_RIGHT_R: + case OpCode::Id::SHF_RIGHT_IMM: + case OpCode::Id::SHF_LEFT_R: + case OpCode::Id::SHF_LEFT_IMM: { + UNIMPLEMENTED_IF(instr.generates_cc); + UNIMPLEMENTED_IF_MSG(instr.shf.xmode != ShfXmode::None, "xmode={}", + static_cast<int>(instr.shf.xmode.Value())); + + if (instr.is_b_imm) { + op_b = Immediate(static_cast<u32>(instr.shf.immediate)); + } + const s32 size = instr.shf.type == ShfType::Bits32 ? 32 : 64; + Node shift = instr.shf.wrap ? WrapShift(move(op_b), size) : ClampShift(move(op_b), size); + + Node negated_shift = Operation(OperationCode::INegate, shift); + Node low_shift = Operation(OperationCode::IAdd, move(negated_shift), Immediate(32)); + + const bool is_right = opid == OpCode::Id::SHF_RIGHT_R || opid == OpCode::Id::SHF_RIGHT_IMM; + Node value = (is_right ? ShiftRight : ShiftLeft)( + move(op_a), GetRegister(instr.gpr39), move(shift), move(low_shift), instr.shf.type); + + SetRegister(bb, instr.gpr0, move(value)); break; } default: diff --git a/src/yuzu/configuration/config.cpp b/src/yuzu/configuration/config.cpp index 280d81ba9..cd94693c1 100644 --- a/src/yuzu/configuration/config.cpp +++ b/src/yuzu/configuration/config.cpp @@ -742,7 +742,6 @@ void Config::ReadUIValues() { void Config::ReadUIGamelistValues() { qt_config->beginGroup(QStringLiteral("UIGameList")); - UISettings::values.show_unknown = ReadSetting(QStringLiteral("show_unknown"), true).toBool(); UISettings::values.show_add_ons = ReadSetting(QStringLiteral("show_add_ons"), true).toBool(); UISettings::values.icon_size = ReadSetting(QStringLiteral("icon_size"), 64).toUInt(); UISettings::values.row_1_text_id = ReadSetting(QStringLiteral("row_1_text_id"), 3).toUInt(); @@ -1159,7 +1158,6 @@ void Config::SaveUIValues() { void Config::SaveUIGamelistValues() { qt_config->beginGroup(QStringLiteral("UIGameList")); - WriteSetting(QStringLiteral("show_unknown"), UISettings::values.show_unknown, true); WriteSetting(QStringLiteral("show_add_ons"), UISettings::values.show_add_ons, true); WriteSetting(QStringLiteral("icon_size"), UISettings::values.icon_size, 64); WriteSetting(QStringLiteral("row_1_text_id"), UISettings::values.row_1_text_id, 3); diff --git a/src/yuzu/configuration/configure_ui.cpp b/src/yuzu/configuration/configure_ui.cpp index c4a84cc67..94424ee44 100644 --- a/src/yuzu/configuration/configure_ui.cpp +++ b/src/yuzu/configuration/configure_ui.cpp @@ -40,7 +40,6 @@ ConfigureUi::ConfigureUi(QWidget* parent) : QWidget(parent), ui(new Ui::Configur SetConfiguration(); // Force game list reload if any of the relevant settings are changed. - connect(ui->show_unknown, &QCheckBox::stateChanged, this, &ConfigureUi::RequestGameListUpdate); connect(ui->icon_size_combobox, QOverload<int>::of(&QComboBox::currentIndexChanged), this, &ConfigureUi::RequestGameListUpdate); connect(ui->row_1_text_combobox, QOverload<int>::of(&QComboBox::currentIndexChanged), this, @@ -60,7 +59,6 @@ ConfigureUi::~ConfigureUi() = default; void ConfigureUi::ApplyConfiguration() { UISettings::values.theme = ui->theme_combobox->itemData(ui->theme_combobox->currentIndex()).toString(); - UISettings::values.show_unknown = ui->show_unknown->isChecked(); UISettings::values.show_add_ons = ui->show_add_ons->isChecked(); UISettings::values.icon_size = ui->icon_size_combobox->currentData().toUInt(); UISettings::values.row_1_text_id = ui->row_1_text_combobox->currentData().toUInt(); @@ -74,7 +72,6 @@ void ConfigureUi::RequestGameListUpdate() { void ConfigureUi::SetConfiguration() { ui->theme_combobox->setCurrentIndex(ui->theme_combobox->findData(UISettings::values.theme)); - ui->show_unknown->setChecked(UISettings::values.show_unknown); ui->show_add_ons->setChecked(UISettings::values.show_add_ons); ui->icon_size_combobox->setCurrentIndex( ui->icon_size_combobox->findData(UISettings::values.icon_size)); diff --git a/src/yuzu/configuration/configure_ui.ui b/src/yuzu/configuration/configure_ui.ui index aa36bd112..bd5c5d3c2 100644 --- a/src/yuzu/configuration/configure_ui.ui +++ b/src/yuzu/configuration/configure_ui.ui @@ -52,13 +52,6 @@ <item> <layout class="QVBoxLayout" name="GeneralVerticalLayout"> <item> - <widget class="QCheckBox" name="show_unknown"> - <property name="text"> - <string>Show files with type 'Unknown'</string> - </property> - </widget> - </item> - <item> <widget class="QCheckBox" name="show_add_ons"> <property name="text"> <string>Show Add-Ons Column</string> diff --git a/src/yuzu/game_list_worker.cpp b/src/yuzu/game_list_worker.cpp index 4c81ef12b..da2c27aa2 100644 --- a/src/yuzu/game_list_worker.cpp +++ b/src/yuzu/game_list_worker.cpp @@ -298,8 +298,7 @@ void GameListWorker::ScanFileSystem(ScanTarget target, const std::string& dir_pa } const auto file_type = loader->GetFileType(); - if ((file_type == Loader::FileType::Unknown || file_type == Loader::FileType::Error) && - !UISettings::values.show_unknown) { + if (file_type == Loader::FileType::Unknown || file_type == Loader::FileType::Error) { return true; } diff --git a/src/yuzu/uisettings.h b/src/yuzu/uisettings.h index bc7725a01..a675ecf4d 100644 --- a/src/yuzu/uisettings.h +++ b/src/yuzu/uisettings.h @@ -89,7 +89,6 @@ struct Values { int profile_index; // Game List - bool show_unknown; bool show_add_ons; uint32_t icon_size; uint8_t row_1_text_id; |