diff options
Diffstat (limited to 'src')
95 files changed, 957 insertions, 517 deletions
diff --git a/src/common/CMakeLists.txt b/src/common/CMakeLists.txt index 850ce8006..1e8e1b215 100644 --- a/src/common/CMakeLists.txt +++ b/src/common/CMakeLists.txt @@ -91,6 +91,8 @@ add_library(common STATIC logging/log.h logging/text_formatter.cpp logging/text_formatter.h + lz4_compression.cpp + lz4_compression.h math_util.h memory_hook.cpp memory_hook.h @@ -123,6 +125,8 @@ add_library(common STATIC uint128.h vector_math.h web_result.h + zstd_compression.cpp + zstd_compression.h ) if(ARCHITECTURE_x86_64) @@ -136,3 +140,4 @@ endif() create_target_directory_groups(common) target_link_libraries(common PUBLIC Boost::boost fmt microprofile) +target_link_libraries(common PRIVATE lz4_static libzstd_static) diff --git a/src/common/bit_util.h b/src/common/bit_util.h index a4f9ed4aa..d032df413 100644 --- a/src/common/bit_util.h +++ b/src/common/bit_util.h @@ -32,7 +32,7 @@ inline u32 CountLeadingZeroes32(u32 value) { return 32; } -inline u64 CountLeadingZeroes64(u64 value) { +inline u32 CountLeadingZeroes64(u64 value) { unsigned long leading_zero = 0; if (_BitScanReverse64(&leading_zero, value) != 0) { @@ -47,15 +47,15 @@ inline u32 CountLeadingZeroes32(u32 value) { return 32; } - return __builtin_clz(value); + return static_cast<u32>(__builtin_clz(value)); } -inline u64 CountLeadingZeroes64(u64 value) { +inline u32 CountLeadingZeroes64(u64 value) { if (value == 0) { return 64; } - return __builtin_clzll(value); + return static_cast<u32>(__builtin_clzll(value)); } #endif @@ -70,7 +70,7 @@ inline u32 CountTrailingZeroes32(u32 value) { return 32; } -inline u64 CountTrailingZeroes64(u64 value) { +inline u32 CountTrailingZeroes64(u64 value) { unsigned long trailing_zero = 0; if (_BitScanForward64(&trailing_zero, value) != 0) { @@ -85,15 +85,15 @@ inline u32 CountTrailingZeroes32(u32 value) { return 32; } - return __builtin_ctz(value); + return static_cast<u32>(__builtin_ctz(value)); } -inline u64 CountTrailingZeroes64(u64 value) { +inline u32 CountTrailingZeroes64(u64 value) { if (value == 0) { return 64; } - return __builtin_ctzll(value); + return static_cast<u32>(__builtin_ctzll(value)); } #endif diff --git a/src/common/lz4_compression.cpp b/src/common/lz4_compression.cpp new file mode 100644 index 000000000..ade6759bb --- /dev/null +++ b/src/common/lz4_compression.cpp @@ -0,0 +1,76 @@ +// Copyright 2019 yuzu Emulator Project +// Licensed under GPLv2 or any later version +// Refer to the license.txt file included. + +#include <algorithm> +#include <lz4hc.h> + +#include "common/assert.h" +#include "common/lz4_compression.h" + +namespace Common::Compression { + +std::vector<u8> CompressDataLZ4(const u8* source, std::size_t source_size) { + ASSERT_MSG(source_size <= LZ4_MAX_INPUT_SIZE, "Source size exceeds LZ4 maximum input size"); + + const auto source_size_int = static_cast<int>(source_size); + const int max_compressed_size = LZ4_compressBound(source_size_int); + std::vector<u8> compressed(max_compressed_size); + + const int compressed_size = LZ4_compress_default(reinterpret_cast<const char*>(source), + reinterpret_cast<char*>(compressed.data()), + source_size_int, max_compressed_size); + + if (compressed_size <= 0) { + // Compression failed + return {}; + } + + compressed.resize(compressed_size); + + return compressed; +} + +std::vector<u8> CompressDataLZ4HC(const u8* source, std::size_t source_size, + s32 compression_level) { + ASSERT_MSG(source_size <= LZ4_MAX_INPUT_SIZE, "Source size exceeds LZ4 maximum input size"); + + compression_level = std::clamp(compression_level, LZ4HC_CLEVEL_MIN, LZ4HC_CLEVEL_MAX); + + const auto source_size_int = static_cast<int>(source_size); + const int max_compressed_size = LZ4_compressBound(source_size_int); + std::vector<u8> compressed(max_compressed_size); + + const int compressed_size = LZ4_compress_HC( + reinterpret_cast<const char*>(source), reinterpret_cast<char*>(compressed.data()), + source_size_int, max_compressed_size, compression_level); + + if (compressed_size <= 0) { + // Compression failed + return {}; + } + + compressed.resize(compressed_size); + + return compressed; +} + +std::vector<u8> CompressDataLZ4HCMax(const u8* source, std::size_t source_size) { + return CompressDataLZ4HC(source, source_size, LZ4HC_CLEVEL_MAX); +} + +std::vector<u8> DecompressDataLZ4(const std::vector<u8>& compressed, + std::size_t uncompressed_size) { + std::vector<u8> uncompressed(uncompressed_size); + const int size_check = LZ4_decompress_safe(reinterpret_cast<const char*>(compressed.data()), + reinterpret_cast<char*>(uncompressed.data()), + static_cast<int>(compressed.size()), + static_cast<int>(uncompressed.size())); + if (static_cast<int>(uncompressed_size) != size_check) { + // Decompression failed + return {}; + } + return uncompressed; +} + +} // namespace Common::Compression diff --git a/src/common/lz4_compression.h b/src/common/lz4_compression.h new file mode 100644 index 000000000..fe2231a6c --- /dev/null +++ b/src/common/lz4_compression.h @@ -0,0 +1,55 @@ +// Copyright 2019 yuzu Emulator Project +// Licensed under GPLv2 or any later version +// Refer to the license.txt file included. + +#include <vector> + +#include "common/common_types.h" + +namespace Common::Compression { + +/** + * Compresses a source memory region with LZ4 and returns the compressed data in a vector. + * + * @param source the uncompressed source memory region. + * @param source_size the size in bytes of the uncompressed source memory region. + * + * @return the compressed data. + */ +std::vector<u8> CompressDataLZ4(const u8* source, std::size_t source_size); + +/** + * Utilizes the LZ4 subalgorithm LZ4HC with the specified compression level. Higher compression + * levels result in a smaller compressed size, but require more CPU time for compression. The + * compression level has almost no impact on decompression speed. Data compressed with LZ4HC can + * also be decompressed with the default LZ4 decompression. + * + * @param source the uncompressed source memory region. + * @param source_size the size in bytes of the uncompressed source memory region. + * @param compression_level the used compression level. Should be between 3 and 12. + * + * @return the compressed data. + */ +std::vector<u8> CompressDataLZ4HC(const u8* source, std::size_t source_size, s32 compression_level); + +/** + * Utilizes the LZ4 subalgorithm LZ4HC with the highest possible compression level. + * + * @param source the uncompressed source memory region. + * @param source_size the size in bytes of the uncompressed source memory region. + * + * @return the compressed data. + */ +std::vector<u8> CompressDataLZ4HCMax(const u8* source, std::size_t source_size); + +/** + * Decompresses a source memory region with LZ4 and returns the uncompressed data in a vector. + * + * @param compressed the compressed source memory region. + * @param uncompressed_size the size in bytes of the uncompressed data. + * + * @return the decompressed data. + */ +std::vector<u8> DecompressDataLZ4(const std::vector<u8>& compressed, std::size_t uncompressed_size); + +} // namespace Common::Compression
\ No newline at end of file diff --git a/src/common/multi_level_queue.h b/src/common/multi_level_queue.h index 2b61b91e0..9cb448f56 100644 --- a/src/common/multi_level_queue.h +++ b/src/common/multi_level_queue.h @@ -72,7 +72,7 @@ public: u64 prios = mlq.used_priorities; prios &= ~((1ULL << (current_priority + 1)) - 1); if (prios == 0) { - current_priority = mlq.depth(); + current_priority = static_cast<u32>(mlq.depth()); } else { current_priority = CountTrailingZeroes64(prios); it = GetBeginItForPrio(); diff --git a/src/common/zstd_compression.cpp b/src/common/zstd_compression.cpp new file mode 100644 index 000000000..60a35c67c --- /dev/null +++ b/src/common/zstd_compression.cpp @@ -0,0 +1,53 @@ +// Copyright 2019 yuzu Emulator Project +// Licensed under GPLv2 or any later version +// Refer to the license.txt file included. + +#pragma once + +#include <algorithm> +#include <zstd.h> + +#include "common/assert.h" +#include "common/zstd_compression.h" + +namespace Common::Compression { + +std::vector<u8> CompressDataZSTD(const u8* source, std::size_t source_size, s32 compression_level) { + compression_level = std::clamp(compression_level, 1, ZSTD_maxCLevel()); + + const std::size_t max_compressed_size = ZSTD_compressBound(source_size); + std::vector<u8> compressed(max_compressed_size); + + const std::size_t compressed_size = + ZSTD_compress(compressed.data(), compressed.size(), source, source_size, compression_level); + + if (ZSTD_isError(compressed_size)) { + // Compression failed + return {}; + } + + compressed.resize(compressed_size); + + return compressed; +} + +std::vector<u8> CompressDataZSTDDefault(const u8* source, std::size_t source_size) { + return CompressDataZSTD(source, source_size, ZSTD_CLEVEL_DEFAULT); +} + +std::vector<u8> DecompressDataZSTD(const std::vector<u8>& compressed) { + const std::size_t decompressed_size = + ZSTD_getDecompressedSize(compressed.data(), compressed.size()); + std::vector<u8> decompressed(decompressed_size); + + const std::size_t uncompressed_result_size = ZSTD_decompress( + decompressed.data(), decompressed.size(), compressed.data(), compressed.size()); + + if (decompressed_size != uncompressed_result_size || ZSTD_isError(uncompressed_result_size)) { + // Decompression failed + return {}; + } + return decompressed; +} + +} // namespace Common::Compression diff --git a/src/common/zstd_compression.h b/src/common/zstd_compression.h new file mode 100644 index 000000000..e0a64b035 --- /dev/null +++ b/src/common/zstd_compression.h @@ -0,0 +1,42 @@ +// Copyright 2019 yuzu Emulator Project +// Licensed under GPLv2 or any later version +// Refer to the license.txt file included. + +#include <vector> + +#include "common/common_types.h" + +namespace Common::Compression { + +/** + * Compresses a source memory region with Zstandard and returns the compressed data in a vector. + * + * @param source the uncompressed source memory region. + * @param source_size the size in bytes of the uncompressed source memory region. + * @param compression_level the used compression level. Should be between 1 and 22. + * + * @return the compressed data. + */ +std::vector<u8> CompressDataZSTD(const u8* source, std::size_t source_size, s32 compression_level); + +/** + * Compresses a source memory region with Zstandard with the default compression level and returns + * the compressed data in a vector. + * + * @param source the uncompressed source memory region. + * @param source_size the size in bytes of the uncompressed source memory region. + * + * @return the compressed data. + */ +std::vector<u8> CompressDataZSTDDefault(const u8* source, std::size_t source_size); + +/** + * Decompresses a source memory region with Zstandard and returns the uncompressed data in a vector. + * + * @param compressed the compressed source memory region. + * + * @return the decompressed data. + */ +std::vector<u8> DecompressDataZSTD(const std::vector<u8>& compressed); + +} // namespace Common::Compression
\ No newline at end of file diff --git a/src/core/CMakeLists.txt b/src/core/CMakeLists.txt index 9e23afe85..c59107102 100644 --- a/src/core/CMakeLists.txt +++ b/src/core/CMakeLists.txt @@ -458,7 +458,7 @@ add_library(core STATIC create_target_directory_groups(core) target_link_libraries(core PUBLIC common PRIVATE audio_core video_core) -target_link_libraries(core PUBLIC Boost::boost PRIVATE fmt lz4_static mbedtls opus unicorn open_source_archives) +target_link_libraries(core PUBLIC Boost::boost PRIVATE fmt mbedtls opus unicorn open_source_archives) if (ENABLE_WEB_SERVICE) target_compile_definitions(core PRIVATE -DENABLE_WEB_SERVICE) target_link_libraries(core PRIVATE web_service) diff --git a/src/core/arm/dynarmic/arm_dynarmic.cpp b/src/core/arm/dynarmic/arm_dynarmic.cpp index 4fdc12f11..f64e4c6a6 100644 --- a/src/core/arm/dynarmic/arm_dynarmic.cpp +++ b/src/core/arm/dynarmic/arm_dynarmic.cpp @@ -26,7 +26,6 @@ using Vector = Dynarmic::A64::Vector; class ARM_Dynarmic_Callbacks : public Dynarmic::A64::UserCallbacks { public: explicit ARM_Dynarmic_Callbacks(ARM_Dynarmic& parent) : parent(parent) {} - ~ARM_Dynarmic_Callbacks() = default; u8 MemoryRead8(u64 vaddr) override { return Memory::Read8(vaddr); diff --git a/src/core/arm/dynarmic/arm_dynarmic.h b/src/core/arm/dynarmic/arm_dynarmic.h index aada1e862..81e0b4ac0 100644 --- a/src/core/arm/dynarmic/arm_dynarmic.h +++ b/src/core/arm/dynarmic/arm_dynarmic.h @@ -29,7 +29,7 @@ class ARM_Dynarmic final : public ARM_Interface { public: ARM_Dynarmic(Timing::CoreTiming& core_timing, ExclusiveMonitor& exclusive_monitor, std::size_t core_index); - ~ARM_Dynarmic(); + ~ARM_Dynarmic() override; void MapBackingMemory(VAddr address, std::size_t size, u8* memory, Kernel::VMAPermission perms) override; @@ -76,7 +76,7 @@ private: class DynarmicExclusiveMonitor final : public ExclusiveMonitor { public: explicit DynarmicExclusiveMonitor(std::size_t core_count); - ~DynarmicExclusiveMonitor(); + ~DynarmicExclusiveMonitor() override; void SetExclusive(std::size_t core_index, VAddr addr) override; void ClearExclusive() override; diff --git a/src/core/arm/unicorn/arm_unicorn.cpp b/src/core/arm/unicorn/arm_unicorn.cpp index a542a098b..27309280c 100644 --- a/src/core/arm/unicorn/arm_unicorn.cpp +++ b/src/core/arm/unicorn/arm_unicorn.cpp @@ -192,12 +192,13 @@ void ARM_Unicorn::ExecuteInstructions(int num_instructions) { CHECKED(uc_emu_start(uc, GetPC(), 1ULL << 63, 0, num_instructions)); core_timing.AddTicks(num_instructions); if (GDBStub::IsServerEnabled()) { - if (last_bkpt_hit) { + if (last_bkpt_hit && last_bkpt.type == GDBStub::BreakpointType::Execute) { uc_reg_write(uc, UC_ARM64_REG_PC, &last_bkpt.address); } + Kernel::Thread* thread = Kernel::GetCurrentThread(); SaveContext(thread->GetContext()); - if (last_bkpt_hit || GDBStub::GetCpuStepFlag()) { + if (last_bkpt_hit || GDBStub::IsMemoryBreak() || GDBStub::GetCpuStepFlag()) { last_bkpt_hit = false; GDBStub::Break(); GDBStub::SendTrap(thread, 5); diff --git a/src/core/arm/unicorn/arm_unicorn.h b/src/core/arm/unicorn/arm_unicorn.h index dbd6955ea..1e44f0736 100644 --- a/src/core/arm/unicorn/arm_unicorn.h +++ b/src/core/arm/unicorn/arm_unicorn.h @@ -18,7 +18,7 @@ namespace Core { class ARM_Unicorn final : public ARM_Interface { public: explicit ARM_Unicorn(Timing::CoreTiming& core_timing); - ~ARM_Unicorn(); + ~ARM_Unicorn() override; void MapBackingMemory(VAddr address, std::size_t size, u8* memory, Kernel::VMAPermission perms) override; @@ -50,7 +50,7 @@ private: uc_engine* uc{}; Timing::CoreTiming& core_timing; GDBStub::BreakpointAddress last_bkpt{}; - bool last_bkpt_hit; + bool last_bkpt_hit = false; }; } // namespace Core diff --git a/src/core/file_sys/control_metadata.cpp b/src/core/file_sys/control_metadata.cpp index 83c184750..60ea9ad12 100644 --- a/src/core/file_sys/control_metadata.cpp +++ b/src/core/file_sys/control_metadata.cpp @@ -67,7 +67,7 @@ std::string NACP::GetDeveloperName(Language language) const { } u64 NACP::GetTitleId() const { - return raw.title_id; + return raw.save_data_owner_id; } u64 NACP::GetDLCBaseTitleId() const { @@ -80,11 +80,11 @@ std::string NACP::GetVersionString() const { } u64 NACP::GetDefaultNormalSaveSize() const { - return raw.normal_save_data_size; + return raw.user_account_save_data_size; } u64 NACP::GetDefaultJournalSaveSize() const { - return raw.journal_sava_data_size; + return raw.user_account_save_data_journal_size; } std::vector<u8> NACP::GetRawBytes() const { diff --git a/src/core/file_sys/control_metadata.h b/src/core/file_sys/control_metadata.h index 7b9cdc910..280710ddf 100644 --- a/src/core/file_sys/control_metadata.h +++ b/src/core/file_sys/control_metadata.h @@ -38,23 +38,35 @@ struct RawNACP { u8 video_capture_mode; bool data_loss_confirmation; INSERT_PADDING_BYTES(1); - u64_le title_id; + u64_le presence_group_id; std::array<u8, 0x20> rating_age; std::array<char, 0x10> version_string; u64_le dlc_base_title_id; - u64_le title_id_2; - u64_le normal_save_data_size; - u64_le journal_sava_data_size; - INSERT_PADDING_BYTES(0x18); - u64_le product_code; + u64_le save_data_owner_id; + u64_le user_account_save_data_size; + u64_le user_account_save_data_journal_size; + u64_le device_save_data_size; + u64_le device_save_data_journal_size; + u64_le bcat_delivery_cache_storage_size; + char application_error_code_category[8]; std::array<u64_le, 0x8> local_communication; u8 logo_type; u8 logo_handling; bool runtime_add_on_content_install; INSERT_PADDING_BYTES(5); - u64_le title_id_update; - std::array<u8, 0x40> bcat_passphrase; - INSERT_PADDING_BYTES(0xEC0); + u64_le seed_for_pseudo_device_id; + std::array<u8, 0x41> bcat_passphrase; + INSERT_PADDING_BYTES(7); + u64_le user_account_save_data_max_size; + u64_le user_account_save_data_max_journal_size; + u64_le device_save_data_max_size; + u64_le device_save_data_max_journal_size; + u64_le temporary_storage_size; + u64_le cache_storage_size; + u64_le cache_storage_journal_size; + u64_le cache_storage_data_and_journal_max_size; + u64_le cache_storage_max_index; + INSERT_PADDING_BYTES(0xE70); }; static_assert(sizeof(RawNACP) == 0x4000, "RawNACP has incorrect size."); diff --git a/src/core/file_sys/fsmitm_romfsbuild.cpp b/src/core/file_sys/fsmitm_romfsbuild.cpp index 47b7526c7..d126ae8dd 100644 --- a/src/core/file_sys/fsmitm_romfsbuild.cpp +++ b/src/core/file_sys/fsmitm_romfsbuild.cpp @@ -23,6 +23,7 @@ */ #include <cstring> +#include <string_view> #include "common/alignment.h" #include "common/assert.h" #include "core/file_sys/fsmitm_romfsbuild.h" @@ -97,7 +98,8 @@ struct RomFSBuildFileContext { VirtualFile source; }; -static u32 romfs_calc_path_hash(u32 parent, std::string path, u32 start, std::size_t path_len) { +static u32 romfs_calc_path_hash(u32 parent, std::string_view path, u32 start, + std::size_t path_len) { u32 hash = parent ^ 123456789; for (u32 i = 0; i < path_len; i++) { hash = (hash >> 5) | (hash << 27); diff --git a/src/core/file_sys/nca_metadata.cpp b/src/core/file_sys/nca_metadata.cpp index 6f34b7836..93d0df6b9 100644 --- a/src/core/file_sys/nca_metadata.cpp +++ b/src/core/file_sys/nca_metadata.cpp @@ -10,14 +10,6 @@ namespace FileSys { -bool operator>=(TitleType lhs, TitleType rhs) { - return static_cast<std::size_t>(lhs) >= static_cast<std::size_t>(rhs); -} - -bool operator<=(TitleType lhs, TitleType rhs) { - return static_cast<std::size_t>(lhs) <= static_cast<std::size_t>(rhs); -} - CNMT::CNMT(VirtualFile file) { if (file->ReadObject(&header) != sizeof(CNMTHeader)) return; diff --git a/src/core/file_sys/nca_metadata.h b/src/core/file_sys/nca_metadata.h index a05d155f4..50bf38471 100644 --- a/src/core/file_sys/nca_metadata.h +++ b/src/core/file_sys/nca_metadata.h @@ -29,9 +29,6 @@ enum class TitleType : u8 { DeltaTitle = 0x83, }; -bool operator>=(TitleType lhs, TitleType rhs); -bool operator<=(TitleType lhs, TitleType rhs); - enum class ContentRecordType : u8 { Meta = 0, Program = 1, diff --git a/src/core/file_sys/program_metadata.cpp b/src/core/file_sys/program_metadata.cpp index d3e00437f..d863253f8 100644 --- a/src/core/file_sys/program_metadata.cpp +++ b/src/core/file_sys/program_metadata.cpp @@ -3,7 +3,6 @@ // Refer to the license.txt file included. #include <cstddef> -#include <cstring> #include <vector> #include "common/logging/log.h" @@ -17,28 +16,30 @@ ProgramMetadata::ProgramMetadata() = default; ProgramMetadata::~ProgramMetadata() = default; Loader::ResultStatus ProgramMetadata::Load(VirtualFile file) { - std::size_t total_size = static_cast<std::size_t>(file->GetSize()); - if (total_size < sizeof(Header)) + const std::size_t total_size = file->GetSize(); + if (total_size < sizeof(Header)) { return Loader::ResultStatus::ErrorBadNPDMHeader; + } - // TODO(DarkLordZach): Use ReadObject when Header/AcidHeader becomes trivially copyable. - std::vector<u8> npdm_header_data = file->ReadBytes(sizeof(Header)); - if (sizeof(Header) != npdm_header_data.size()) + if (sizeof(Header) != file->ReadObject(&npdm_header)) { return Loader::ResultStatus::ErrorBadNPDMHeader; - std::memcpy(&npdm_header, npdm_header_data.data(), sizeof(Header)); + } - std::vector<u8> acid_header_data = file->ReadBytes(sizeof(AcidHeader), npdm_header.acid_offset); - if (sizeof(AcidHeader) != acid_header_data.size()) + if (sizeof(AcidHeader) != file->ReadObject(&acid_header, npdm_header.acid_offset)) { return Loader::ResultStatus::ErrorBadACIDHeader; - std::memcpy(&acid_header, acid_header_data.data(), sizeof(AcidHeader)); + } - if (sizeof(AciHeader) != file->ReadObject(&aci_header, npdm_header.aci_offset)) + if (sizeof(AciHeader) != file->ReadObject(&aci_header, npdm_header.aci_offset)) { return Loader::ResultStatus::ErrorBadACIHeader; + } - if (sizeof(FileAccessControl) != file->ReadObject(&acid_file_access, acid_header.fac_offset)) + if (sizeof(FileAccessControl) != file->ReadObject(&acid_file_access, acid_header.fac_offset)) { return Loader::ResultStatus::ErrorBadFileAccessControl; - if (sizeof(FileAccessHeader) != file->ReadObject(&aci_file_access, aci_header.fah_offset)) + } + + if (sizeof(FileAccessHeader) != file->ReadObject(&aci_file_access, aci_header.fah_offset)) { return Loader::ResultStatus::ErrorBadFileAccessHeader; + } aci_kernel_capabilities.resize(aci_header.kac_size / sizeof(u32)); const u64 read_size = aci_header.kac_size; diff --git a/src/core/file_sys/program_metadata.h b/src/core/file_sys/program_metadata.h index 0033ba347..7de5b9cf9 100644 --- a/src/core/file_sys/program_metadata.h +++ b/src/core/file_sys/program_metadata.h @@ -58,7 +58,6 @@ public: void Print() const; private: - // TODO(DarkLordZach): BitField is not trivially copyable. struct Header { std::array<char, 4> magic; std::array<u8, 8> reserved; @@ -85,7 +84,6 @@ private: static_assert(sizeof(Header) == 0x80, "NPDM header structure size is wrong"); - // TODO(DarkLordZach): BitField is not trivially copyable. struct AcidHeader { std::array<u8, 0x100> signature; std::array<u8, 0x100> nca_modulus; diff --git a/src/core/file_sys/savedata_factory.cpp b/src/core/file_sys/savedata_factory.cpp index 1913dc956..7974b031d 100644 --- a/src/core/file_sys/savedata_factory.cpp +++ b/src/core/file_sys/savedata_factory.cpp @@ -16,8 +16,10 @@ namespace FileSys { constexpr char SAVE_DATA_SIZE_FILENAME[] = ".yuzu_save_size"; std::string SaveDataDescriptor::DebugInfo() const { - return fmt::format("[type={:02X}, title_id={:016X}, user_id={:016X}{:016X}, save_id={:016X}]", - static_cast<u8>(type), title_id, user_id[1], user_id[0], save_id); + return fmt::format("[type={:02X}, title_id={:016X}, user_id={:016X}{:016X}, save_id={:016X}, " + "rank={}, index={}]", + static_cast<u8>(type), title_id, user_id[1], user_id[0], save_id, + static_cast<u8>(rank), index); } SaveDataFactory::SaveDataFactory(VirtualDir save_directory) : dir(std::move(save_directory)) { @@ -28,7 +30,7 @@ SaveDataFactory::SaveDataFactory(VirtualDir save_directory) : dir(std::move(save SaveDataFactory::~SaveDataFactory() = default; -ResultVal<VirtualDir> SaveDataFactory::Open(SaveDataSpaceId space, SaveDataDescriptor meta) { +ResultVal<VirtualDir> SaveDataFactory::Open(SaveDataSpaceId space, const SaveDataDescriptor& meta) { if (meta.type == SaveDataType::SystemSaveData || meta.type == SaveDataType::SaveData) { if (meta.zero_1 != 0) { LOG_WARNING(Service_FS, diff --git a/src/core/file_sys/savedata_factory.h b/src/core/file_sys/savedata_factory.h index 3a1caf292..b73654571 100644 --- a/src/core/file_sys/savedata_factory.h +++ b/src/core/file_sys/savedata_factory.h @@ -32,12 +32,19 @@ enum class SaveDataType : u8 { CacheStorage = 5, }; +enum class SaveDataRank : u8 { + Primary, + Secondary, +}; + struct SaveDataDescriptor { u64_le title_id; u128 user_id; u64_le save_id; SaveDataType type; - INSERT_PADDING_BYTES(7); + SaveDataRank rank; + u16_le index; + INSERT_PADDING_BYTES(4); u64_le zero_1; u64_le zero_2; u64_le zero_3; @@ -57,7 +64,7 @@ public: explicit SaveDataFactory(VirtualDir dir); ~SaveDataFactory(); - ResultVal<VirtualDir> Open(SaveDataSpaceId space, SaveDataDescriptor meta); + ResultVal<VirtualDir> Open(SaveDataSpaceId space, const SaveDataDescriptor& meta); VirtualDir GetSaveDataSpaceDirectory(SaveDataSpaceId space) const; diff --git a/src/core/gdbstub/gdbstub.cpp b/src/core/gdbstub/gdbstub.cpp index dafb32aae..afa812598 100644 --- a/src/core/gdbstub/gdbstub.cpp +++ b/src/core/gdbstub/gdbstub.cpp @@ -1030,7 +1030,7 @@ static void Step() { /// Tell the CPU if we hit a memory breakpoint. bool IsMemoryBreak() { - if (IsConnected()) { + if (!IsConnected()) { return false; } diff --git a/src/core/hle/ipc_helpers.h b/src/core/hle/ipc_helpers.h index 68406eb63..ac0e1d796 100644 --- a/src/core/hle/ipc_helpers.h +++ b/src/core/hle/ipc_helpers.h @@ -139,10 +139,8 @@ public: context->AddDomainObject(std::move(iface)); } else { auto& kernel = Core::System::GetInstance().Kernel(); - auto sessions = + auto [server, client] = Kernel::ServerSession::CreateSessionPair(kernel, iface->GetServiceName()); - auto server = std::get<Kernel::SharedPtr<Kernel::ServerSession>>(sessions); - auto client = std::get<Kernel::SharedPtr<Kernel::ClientSession>>(sessions); iface->ClientConnected(server); context->AddMoveObject(std::move(client)); } diff --git a/src/core/hle/kernel/client_port.cpp b/src/core/hle/kernel/client_port.cpp index aa432658e..744b1697d 100644 --- a/src/core/hle/kernel/client_port.cpp +++ b/src/core/hle/kernel/client_port.cpp @@ -2,8 +2,6 @@ // Licensed under GPLv2 or any later version // Refer to the license.txt file included. -#include <tuple> - #include "core/hle/kernel/client_port.h" #include "core/hle/kernel/client_session.h" #include "core/hle/kernel/errors.h" @@ -31,18 +29,18 @@ ResultVal<SharedPtr<ClientSession>> ClientPort::Connect() { active_sessions++; // Create a new session pair, let the created sessions inherit the parent port's HLE handler. - auto sessions = ServerSession::CreateSessionPair(kernel, server_port->GetName(), this); + auto [server, client] = ServerSession::CreateSessionPair(kernel, server_port->GetName(), this); if (server_port->HasHLEHandler()) { - server_port->GetHLEHandler()->ClientConnected(std::get<SharedPtr<ServerSession>>(sessions)); + server_port->GetHLEHandler()->ClientConnected(server); } else { - server_port->AppendPendingSession(std::get<SharedPtr<ServerSession>>(sessions)); + server_port->AppendPendingSession(server); } // Wake the threads waiting on the ServerPort server_port->WakeupAllWaitingThreads(); - return MakeResult(std::get<SharedPtr<ClientSession>>(sessions)); + return MakeResult(client); } void ClientPort::ConnectionClosed() { diff --git a/src/core/hle/kernel/object.cpp b/src/core/hle/kernel/object.cpp index 217144efc..10431e94c 100644 --- a/src/core/hle/kernel/object.cpp +++ b/src/core/hle/kernel/object.cpp @@ -24,7 +24,6 @@ bool Object::IsWaitable() const { case HandleType::WritableEvent: case HandleType::SharedMemory: case HandleType::TransferMemory: - case HandleType::AddressArbiter: case HandleType::ResourceLimit: case HandleType::ClientPort: case HandleType::ClientSession: diff --git a/src/core/hle/kernel/object.h b/src/core/hle/kernel/object.h index 3f6baa094..332876c27 100644 --- a/src/core/hle/kernel/object.h +++ b/src/core/hle/kernel/object.h @@ -25,7 +25,6 @@ enum class HandleType : u32 { TransferMemory, Thread, Process, - AddressArbiter, ResourceLimit, ClientPort, ServerPort, diff --git a/src/core/hle/kernel/server_port.cpp b/src/core/hle/kernel/server_port.cpp index 708fdf9e1..02e7c60e6 100644 --- a/src/core/hle/kernel/server_port.cpp +++ b/src/core/hle/kernel/server_port.cpp @@ -39,9 +39,8 @@ void ServerPort::Acquire(Thread* thread) { ASSERT_MSG(!ShouldWait(thread), "object unavailable!"); } -std::tuple<SharedPtr<ServerPort>, SharedPtr<ClientPort>> ServerPort::CreatePortPair( - KernelCore& kernel, u32 max_sessions, std::string name) { - +ServerPort::PortPair ServerPort::CreatePortPair(KernelCore& kernel, u32 max_sessions, + std::string name) { SharedPtr<ServerPort> server_port(new ServerPort(kernel)); SharedPtr<ClientPort> client_port(new ClientPort(kernel)); @@ -51,7 +50,7 @@ std::tuple<SharedPtr<ServerPort>, SharedPtr<ClientPort>> ServerPort::CreatePortP client_port->max_sessions = max_sessions; client_port->active_sessions = 0; - return std::make_tuple(std::move(server_port), std::move(client_port)); + return std::make_pair(std::move(server_port), std::move(client_port)); } } // namespace Kernel diff --git a/src/core/hle/kernel/server_port.h b/src/core/hle/kernel/server_port.h index 76293cb8b..fef573b71 100644 --- a/src/core/hle/kernel/server_port.h +++ b/src/core/hle/kernel/server_port.h @@ -6,7 +6,7 @@ #include <memory> #include <string> -#include <tuple> +#include <utility> #include <vector> #include "common/common_types.h" #include "core/hle/kernel/object.h" @@ -23,6 +23,7 @@ class SessionRequestHandler; class ServerPort final : public WaitObject { public: using HLEHandler = std::shared_ptr<SessionRequestHandler>; + using PortPair = std::pair<SharedPtr<ServerPort>, SharedPtr<ClientPort>>; /** * Creates a pair of ServerPort and an associated ClientPort. @@ -32,8 +33,8 @@ public: * @param name Optional name of the ports * @return The created port tuple */ - static std::tuple<SharedPtr<ServerPort>, SharedPtr<ClientPort>> CreatePortPair( - KernelCore& kernel, u32 max_sessions, std::string name = "UnknownPort"); + static PortPair CreatePortPair(KernelCore& kernel, u32 max_sessions, + std::string name = "UnknownPort"); std::string GetTypeName() const override { return "ServerPort"; diff --git a/src/core/hle/kernel/server_session.cpp b/src/core/hle/kernel/server_session.cpp index 40cec143e..a6b2cf06a 100644 --- a/src/core/hle/kernel/server_session.cpp +++ b/src/core/hle/kernel/server_session.cpp @@ -204,6 +204,6 @@ ServerSession::SessionPair ServerSession::CreateSessionPair(KernelCore& kernel, client_session->parent = parent; server_session->parent = parent; - return std::make_tuple(std::move(server_session), std::move(client_session)); + return std::make_pair(std::move(server_session), std::move(client_session)); } } // namespace Kernel diff --git a/src/core/hle/kernel/server_session.h b/src/core/hle/kernel/server_session.h index 79b84bade..09b835ff8 100644 --- a/src/core/hle/kernel/server_session.h +++ b/src/core/hle/kernel/server_session.h @@ -6,6 +6,7 @@ #include <memory> #include <string> +#include <utility> #include <vector> #include "core/hle/kernel/object.h" @@ -41,6 +42,10 @@ public: return "ServerSession"; } + std::string GetName() const override { + return name; + } + static const HandleType HANDLE_TYPE = HandleType::ServerSession; HandleType GetHandleType() const override { return HANDLE_TYPE; @@ -54,7 +59,7 @@ public: return parent.get(); } - using SessionPair = std::tuple<SharedPtr<ServerSession>, SharedPtr<ClientSession>>; + using SessionPair = std::pair<SharedPtr<ServerSession>, SharedPtr<ClientSession>>; /** * Creates a pair of ServerSession and an associated ClientSession. diff --git a/src/core/hle/kernel/svc.cpp b/src/core/hle/kernel/svc.cpp index ab10db3df..2fd07ab34 100644 --- a/src/core/hle/kernel/svc.cpp +++ b/src/core/hle/kernel/svc.cpp @@ -1339,6 +1339,20 @@ static ResultCode WaitProcessWideKeyAtomic(VAddr mutex_addr, VAddr condition_var "called mutex_addr={:X}, condition_variable_addr={:X}, thread_handle=0x{:08X}, timeout={}", mutex_addr, condition_variable_addr, thread_handle, nano_seconds); + if (Memory::IsKernelVirtualAddress(mutex_addr)) { + LOG_ERROR( + Kernel_SVC, + "Given mutex address must not be within the kernel address space. address=0x{:016X}", + mutex_addr); + return ERR_INVALID_ADDRESS_STATE; + } + + if (!Common::IsWordAligned(mutex_addr)) { + LOG_ERROR(Kernel_SVC, "Given mutex address must be word-aligned. address=0x{:016X}", + mutex_addr); + return ERR_INVALID_ADDRESS; + } + auto* const current_process = Core::System::GetInstance().Kernel().CurrentProcess(); const auto& handle_table = current_process->GetHandleTable(); SharedPtr<Thread> thread = handle_table.Get<Thread>(thread_handle); diff --git a/src/core/hle/kernel/transfer_memory.cpp b/src/core/hle/kernel/transfer_memory.cpp index 23228e1b5..26c4e5e67 100644 --- a/src/core/hle/kernel/transfer_memory.cpp +++ b/src/core/hle/kernel/transfer_memory.cpp @@ -14,8 +14,8 @@ namespace Kernel { TransferMemory::TransferMemory(KernelCore& kernel) : Object{kernel} {} TransferMemory::~TransferMemory() = default; -SharedPtr<TransferMemory> TransferMemory::Create(KernelCore& kernel, VAddr base_address, - size_t size, MemoryPermission permissions) { +SharedPtr<TransferMemory> TransferMemory::Create(KernelCore& kernel, VAddr base_address, u64 size, + MemoryPermission permissions) { SharedPtr<TransferMemory> transfer_memory{new TransferMemory(kernel)}; transfer_memory->base_address = base_address; @@ -26,7 +26,15 @@ SharedPtr<TransferMemory> TransferMemory::Create(KernelCore& kernel, VAddr base_ return transfer_memory; } -ResultCode TransferMemory::MapMemory(VAddr address, size_t size, MemoryPermission permissions) { +const u8* TransferMemory::GetPointer() const { + return backing_block.get()->data(); +} + +u64 TransferMemory::GetSize() const { + return memory_size; +} + +ResultCode TransferMemory::MapMemory(VAddr address, u64 size, MemoryPermission permissions) { if (memory_size != size) { return ERR_INVALID_SIZE; } @@ -39,13 +47,13 @@ ResultCode TransferMemory::MapMemory(VAddr address, size_t size, MemoryPermissio return ERR_INVALID_STATE; } + backing_block = std::make_shared<std::vector<u8>>(size); + const auto map_state = owner_permissions == MemoryPermission::None ? MemoryState::TransferMemoryIsolated : MemoryState::TransferMemory; auto& vm_manager = owner_process->VMManager(); - const auto map_result = vm_manager.MapMemoryBlock( - address, std::make_shared<std::vector<u8>>(size), 0, size, map_state); - + const auto map_result = vm_manager.MapMemoryBlock(address, backing_block, 0, size, map_state); if (map_result.Failed()) { return map_result.Code(); } @@ -54,7 +62,7 @@ ResultCode TransferMemory::MapMemory(VAddr address, size_t size, MemoryPermissio return RESULT_SUCCESS; } -ResultCode TransferMemory::UnmapMemory(VAddr address, size_t size) { +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 ec294951e..a140b1e2b 100644 --- a/src/core/hle/kernel/transfer_memory.h +++ b/src/core/hle/kernel/transfer_memory.h @@ -4,6 +4,9 @@ #pragma once +#include <memory> +#include <vector> + #include "core/hle/kernel/object.h" union ResultCode; @@ -25,7 +28,7 @@ class TransferMemory final : public Object { public: static constexpr HandleType HANDLE_TYPE = HandleType::TransferMemory; - static SharedPtr<TransferMemory> Create(KernelCore& kernel, VAddr base_address, size_t size, + static SharedPtr<TransferMemory> Create(KernelCore& kernel, VAddr base_address, u64 size, MemoryPermission permissions); TransferMemory(const TransferMemory&) = delete; @@ -46,6 +49,12 @@ public: return HANDLE_TYPE; } + /// Gets a pointer to the backing block of this instance. + const u8* GetPointer() const; + + /// Gets the size of the memory backing this instance in bytes. + u64 GetSize() const; + /// Attempts to map transfer memory with the given range and memory permissions. /// /// @param address The base address to being mapping memory at. @@ -56,7 +65,7 @@ public: /// the same values that were given when creating the transfer memory /// instance. /// - ResultCode MapMemory(VAddr address, size_t size, MemoryPermission permissions); + ResultCode MapMemory(VAddr address, u64 size, MemoryPermission permissions); /// Unmaps the transfer memory with the given range /// @@ -66,17 +75,20 @@ public: /// @pre The given address and size must be the same as the ones used /// to create the transfer memory instance. /// - ResultCode UnmapMemory(VAddr address, size_t size); + ResultCode UnmapMemory(VAddr address, u64 size); private: explicit TransferMemory(KernelCore& kernel); ~TransferMemory() override; + /// Memory block backing this instance. + std::shared_ptr<std::vector<u8>> backing_block; + /// The base address for the memory managed by this instance. VAddr base_address = 0; /// Size of the memory, in bytes, that this instance manages. - size_t memory_size = 0; + u64 memory_size = 0; /// The memory permissions that are applied to this instance. MemoryPermission owner_permissions{}; diff --git a/src/core/hle/result.h b/src/core/hle/result.h index ab84f5ddc..8a3701151 100644 --- a/src/core/hle/result.h +++ b/src/core/hle/result.h @@ -119,10 +119,6 @@ union ResultCode { BitField<0, 9, ErrorModule> module; BitField<9, 13, u32> description; - // The last bit of `level` is checked by apps and the kernel to determine if a result code is an - // error - BitField<31, 1, u32> is_error; - constexpr explicit ResultCode(u32 raw) : raw(raw) {} constexpr ResultCode(ErrorModule module_, u32 description_) diff --git a/src/core/hle/service/am/am.cpp b/src/core/hle/service/am/am.cpp index d31ab7970..85271d418 100644 --- a/src/core/hle/service/am/am.cpp +++ b/src/core/hle/service/am/am.cpp @@ -13,7 +13,7 @@ #include "core/hle/kernel/kernel.h" #include "core/hle/kernel/process.h" #include "core/hle/kernel/readable_event.h" -#include "core/hle/kernel/shared_memory.h" +#include "core/hle/kernel/transfer_memory.h" #include "core/hle/kernel/writable_event.h" #include "core/hle/service/acc/profile_manager.h" #include "core/hle/service/am/am.h" @@ -363,8 +363,6 @@ void ISelfController::SetPerformanceModeChangedNotification(Kernel::HLERequestCo void ISelfController::SetFocusHandlingMode(Kernel::HLERequestContext& ctx) { // Takes 3 input u8s with each field located immediately after the previous // u8, these are bool flags. No output. - LOG_WARNING(Service_AM, "(STUBBED) called"); - IPC::RequestParser rp{ctx}; struct FocusHandlingModeParams { @@ -372,7 +370,10 @@ void ISelfController::SetFocusHandlingMode(Kernel::HLERequestContext& ctx) { u8 unknown1; u8 unknown2; }; - auto flags = rp.PopRaw<FocusHandlingModeParams>(); + const auto flags = rp.PopRaw<FocusHandlingModeParams>(); + + LOG_WARNING(Service_AM, "(STUBBED) called. unknown0={}, unknown1={}, unknown2={}", + flags.unknown0, flags.unknown1, flags.unknown2); IPC::ResponseBuilder rb{ctx, 2}; rb.Push(RESULT_SUCCESS); @@ -931,19 +932,19 @@ void ILibraryAppletCreator::CreateTransferMemoryStorage(Kernel::HLERequestContex rp.SetCurrentOffset(3); const auto handle{rp.Pop<Kernel::Handle>()}; - const auto shared_mem = - Core::System::GetInstance().CurrentProcess()->GetHandleTable().Get<Kernel::SharedMemory>( + const auto transfer_mem = + Core::System::GetInstance().CurrentProcess()->GetHandleTable().Get<Kernel::TransferMemory>( handle); - if (shared_mem == nullptr) { + if (transfer_mem == nullptr) { LOG_ERROR(Service_AM, "shared_mem is a nullpr for handle={:08X}", handle); IPC::ResponseBuilder rb{ctx, 2}; rb.Push(ResultCode(-1)); return; } - const u8* mem_begin = shared_mem->GetPointer(); - const u8* mem_end = mem_begin + shared_mem->GetSize(); + const u8* const mem_begin = transfer_mem->GetPointer(); + const u8* const mem_end = mem_begin + transfer_mem->GetSize(); std::vector<u8> memory{mem_begin, mem_end}; IPC::ResponseBuilder rb{ctx, 2, 0, 1}; diff --git a/src/core/hle/service/audio/audin_u.cpp b/src/core/hle/service/audio/audin_u.cpp index 088410564..e5daefdde 100644 --- a/src/core/hle/service/audio/audin_u.cpp +++ b/src/core/hle/service/audio/audin_u.cpp @@ -2,9 +2,6 @@ // Licensed under GPLv2 or any later version // Refer to the license.txt file included. -#include "common/logging/log.h" -#include "core/hle/ipc_helpers.h" -#include "core/hle/kernel/hle_ipc.h" #include "core/hle/service/audio/audin_u.h" namespace Service::Audio { @@ -33,7 +30,6 @@ public: RegisterHandlers(functions); } - ~IAudioIn() = default; }; AudInU::AudInU() : ServiceFramework("audin:u") { diff --git a/src/core/hle/service/audio/audout_u.cpp b/src/core/hle/service/audio/audout_u.cpp index 21f5e64c7..39acb7b23 100644 --- a/src/core/hle/service/audio/audout_u.cpp +++ b/src/core/hle/service/audio/audout_u.cpp @@ -150,7 +150,6 @@ private: void GetReleasedAudioOutBufferImpl(Kernel::HLERequestContext& ctx) { LOG_DEBUG(Service_Audio, "called {}", ctx.Description()); - IPC::RequestParser rp{ctx}; const u64 max_count{ctx.GetWriteBufferSize() / sizeof(u64)}; const auto released_buffers{audio_core.GetTagsAndReleaseBuffers(stream, max_count)}; @@ -194,12 +193,9 @@ private: void AudOutU::ListAudioOutsImpl(Kernel::HLERequestContext& ctx) { LOG_DEBUG(Service_Audio, "called"); - IPC::RequestParser rp{ctx}; - ctx.WriteBuffer(DefaultDevice); IPC::ResponseBuilder rb{ctx, 3}; - rb.Push(RESULT_SUCCESS); rb.Push<u32>(1); // Amount of audio devices } diff --git a/src/core/hle/service/audio/audrec_u.cpp b/src/core/hle/service/audio/audrec_u.cpp index 6956a2e64..1a5aed9ed 100644 --- a/src/core/hle/service/audio/audrec_u.cpp +++ b/src/core/hle/service/audio/audrec_u.cpp @@ -2,9 +2,6 @@ // Licensed under GPLv2 or any later version // Refer to the license.txt file included. -#include "common/logging/log.h" -#include "core/hle/ipc_helpers.h" -#include "core/hle/kernel/hle_ipc.h" #include "core/hle/service/audio/audrec_u.h" namespace Service::Audio { @@ -30,7 +27,6 @@ public: RegisterHandlers(functions); } - ~IFinalOutputRecorder() = default; }; AudRecU::AudRecU() : ServiceFramework("audrec:u") { diff --git a/src/core/hle/service/audio/audren_u.cpp b/src/core/hle/service/audio/audren_u.cpp index c9de10a24..1dde6edb7 100644 --- a/src/core/hle/service/audio/audren_u.cpp +++ b/src/core/hle/service/audio/audren_u.cpp @@ -10,6 +10,7 @@ #include "common/alignment.h" #include "common/common_funcs.h" #include "common/logging/log.h" +#include "common/string_util.h" #include "core/core.h" #include "core/hle/ipc_helpers.h" #include "core/hle/kernel/hle_ipc.h" @@ -184,7 +185,6 @@ public: private: void ListAudioDeviceName(Kernel::HLERequestContext& ctx) { LOG_WARNING(Service_Audio, "(STUBBED) called"); - IPC::RequestParser rp{ctx}; constexpr std::array<char, 15> audio_interface{{"AudioInterface"}}; ctx.WriteBuffer(audio_interface); @@ -195,13 +195,13 @@ private: } void SetAudioDeviceOutputVolume(Kernel::HLERequestContext& ctx) { - LOG_WARNING(Service_Audio, "(STUBBED) called"); - IPC::RequestParser rp{ctx}; - f32 volume = static_cast<f32>(rp.Pop<u32>()); + const f32 volume = rp.Pop<f32>(); - auto file_buffer = ctx.ReadBuffer(); - auto end = std::find(file_buffer.begin(), file_buffer.end(), '\0'); + const auto device_name_buffer = ctx.ReadBuffer(); + const std::string name = Common::StringFromBuffer(device_name_buffer); + + LOG_WARNING(Service_Audio, "(STUBBED) called. name={}, volume={}", name, volume); IPC::ResponseBuilder rb{ctx, 2}; rb.Push(RESULT_SUCCESS); @@ -209,7 +209,6 @@ private: void GetActiveAudioDeviceName(Kernel::HLERequestContext& ctx) { LOG_WARNING(Service_Audio, "(STUBBED) called"); - IPC::RequestParser rp{ctx}; constexpr std::array<char, 12> audio_interface{{"AudioDevice"}}; ctx.WriteBuffer(audio_interface); diff --git a/src/core/hle/service/filesystem/filesystem.cpp b/src/core/hle/service/filesystem/filesystem.cpp index c6da2df43..4c2b371c3 100644 --- a/src/core/hle/service/filesystem/filesystem.cpp +++ b/src/core/hle/service/filesystem/filesystem.cpp @@ -197,13 +197,16 @@ ResultCode VfsDirectoryServiceWrapper::RenameDirectory(const std::string& src_pa ResultVal<FileSys::VirtualFile> VfsDirectoryServiceWrapper::OpenFile(const std::string& path_, FileSys::Mode mode) const { - std::string path(FileUtil::SanitizePath(path_)); - auto npath = path; - while (npath.size() > 0 && (npath[0] == '/' || npath[0] == '\\')) - npath = npath.substr(1); + const std::string path(FileUtil::SanitizePath(path_)); + std::string_view npath = path; + while (!npath.empty() && (npath[0] == '/' || npath[0] == '\\')) { + npath.remove_prefix(1); + } + auto file = backing->GetFileRelative(npath); - if (file == nullptr) + if (file == nullptr) { return FileSys::ERROR_PATH_NOT_FOUND; + } if (mode == FileSys::Mode::Append) { return MakeResult<FileSys::VirtualFile>( @@ -319,15 +322,15 @@ ResultVal<FileSys::VirtualFile> OpenRomFS(u64 title_id, FileSys::StorageId stora } ResultVal<FileSys::VirtualDir> OpenSaveData(FileSys::SaveDataSpaceId space, - FileSys::SaveDataDescriptor save_struct) { + const FileSys::SaveDataDescriptor& descriptor) { LOG_TRACE(Service_FS, "Opening Save Data for space_id={:01X}, save_struct={}", - static_cast<u8>(space), save_struct.DebugInfo()); + static_cast<u8>(space), descriptor.DebugInfo()); if (save_data_factory == nullptr) { return FileSys::ERROR_ENTITY_NOT_FOUND; } - return save_data_factory->Open(space, save_struct); + return save_data_factory->Open(space, descriptor); } ResultVal<FileSys::VirtualDir> OpenSaveDataSpace(FileSys::SaveDataSpaceId space) { diff --git a/src/core/hle/service/filesystem/filesystem.h b/src/core/hle/service/filesystem/filesystem.h index 6fd5e7b23..7cfc0d902 100644 --- a/src/core/hle/service/filesystem/filesystem.h +++ b/src/core/hle/service/filesystem/filesystem.h @@ -46,7 +46,7 @@ ResultVal<FileSys::VirtualFile> OpenRomFSCurrentProcess(); ResultVal<FileSys::VirtualFile> OpenRomFS(u64 title_id, FileSys::StorageId storage_id, FileSys::ContentRecordType type); ResultVal<FileSys::VirtualDir> OpenSaveData(FileSys::SaveDataSpaceId space, - FileSys::SaveDataDescriptor save_struct); + const FileSys::SaveDataDescriptor& descriptor); ResultVal<FileSys::VirtualDir> OpenSaveDataSpace(FileSys::SaveDataSpaceId space); ResultVal<FileSys::VirtualDir> OpenSDMC(); diff --git a/src/core/hle/service/filesystem/fsp_srv.cpp b/src/core/hle/service/filesystem/fsp_srv.cpp index f03fb629c..657baddb8 100644 --- a/src/core/hle/service/filesystem/fsp_srv.cpp +++ b/src/core/hle/service/filesystem/fsp_srv.cpp @@ -315,61 +315,53 @@ public: void CreateFile(Kernel::HLERequestContext& ctx) { IPC::RequestParser rp{ctx}; - auto file_buffer = ctx.ReadBuffer(); - std::string name = Common::StringFromBuffer(file_buffer); + const auto file_buffer = ctx.ReadBuffer(); + const std::string name = Common::StringFromBuffer(file_buffer); - u64 mode = rp.Pop<u64>(); - u32 size = rp.Pop<u32>(); + const u64 mode = rp.Pop<u64>(); + const u32 size = rp.Pop<u32>(); - LOG_DEBUG(Service_FS, "called file {} mode 0x{:X} size 0x{:08X}", name, mode, size); + LOG_DEBUG(Service_FS, "called. file={}, mode=0x{:X}, size=0x{:08X}", name, mode, size); IPC::ResponseBuilder rb{ctx, 2}; rb.Push(backend.CreateFile(name, size)); } void DeleteFile(Kernel::HLERequestContext& ctx) { - IPC::RequestParser rp{ctx}; - - auto file_buffer = ctx.ReadBuffer(); - std::string name = Common::StringFromBuffer(file_buffer); + const auto file_buffer = ctx.ReadBuffer(); + const std::string name = Common::StringFromBuffer(file_buffer); - LOG_DEBUG(Service_FS, "called file {}", name); + LOG_DEBUG(Service_FS, "called. file={}", name); IPC::ResponseBuilder rb{ctx, 2}; rb.Push(backend.DeleteFile(name)); } void CreateDirectory(Kernel::HLERequestContext& ctx) { - IPC::RequestParser rp{ctx}; - - auto file_buffer = ctx.ReadBuffer(); - std::string name = Common::StringFromBuffer(file_buffer); + const auto file_buffer = ctx.ReadBuffer(); + const std::string name = Common::StringFromBuffer(file_buffer); - LOG_DEBUG(Service_FS, "called directory {}", name); + LOG_DEBUG(Service_FS, "called. directory={}", name); IPC::ResponseBuilder rb{ctx, 2}; rb.Push(backend.CreateDirectory(name)); } void DeleteDirectory(Kernel::HLERequestContext& ctx) { - const IPC::RequestParser rp{ctx}; - const auto file_buffer = ctx.ReadBuffer(); - std::string name = Common::StringFromBuffer(file_buffer); + const std::string name = Common::StringFromBuffer(file_buffer); - LOG_DEBUG(Service_FS, "called directory {}", name); + LOG_DEBUG(Service_FS, "called. directory={}", name); IPC::ResponseBuilder rb{ctx, 2}; rb.Push(backend.DeleteDirectory(name)); } void DeleteDirectoryRecursively(Kernel::HLERequestContext& ctx) { - const IPC::RequestParser rp{ctx}; - const auto file_buffer = ctx.ReadBuffer(); - std::string name = Common::StringFromBuffer(file_buffer); + const std::string name = Common::StringFromBuffer(file_buffer); - LOG_DEBUG(Service_FS, "called directory {}", name); + LOG_DEBUG(Service_FS, "called. directory={}", name); IPC::ResponseBuilder rb{ctx, 2}; rb.Push(backend.DeleteDirectoryRecursively(name)); @@ -386,18 +378,16 @@ public: } void RenameFile(Kernel::HLERequestContext& ctx) { - IPC::RequestParser rp{ctx}; - std::vector<u8> buffer; buffer.resize(ctx.BufferDescriptorX()[0].Size()); Memory::ReadBlock(ctx.BufferDescriptorX()[0].Address(), buffer.data(), buffer.size()); - std::string src_name = Common::StringFromBuffer(buffer); + const std::string src_name = Common::StringFromBuffer(buffer); buffer.resize(ctx.BufferDescriptorX()[1].Size()); Memory::ReadBlock(ctx.BufferDescriptorX()[1].Address(), buffer.data(), buffer.size()); - std::string dst_name = Common::StringFromBuffer(buffer); + const std::string dst_name = Common::StringFromBuffer(buffer); - LOG_DEBUG(Service_FS, "called file '{}' to file '{}'", src_name, dst_name); + LOG_DEBUG(Service_FS, "called. file '{}' to file '{}'", src_name, dst_name); IPC::ResponseBuilder rb{ctx, 2}; rb.Push(backend.RenameFile(src_name, dst_name)); @@ -406,12 +396,12 @@ public: void OpenFile(Kernel::HLERequestContext& ctx) { IPC::RequestParser rp{ctx}; - auto file_buffer = ctx.ReadBuffer(); - std::string name = Common::StringFromBuffer(file_buffer); + const auto file_buffer = ctx.ReadBuffer(); + const std::string name = Common::StringFromBuffer(file_buffer); - auto mode = static_cast<FileSys::Mode>(rp.Pop<u32>()); + const auto mode = static_cast<FileSys::Mode>(rp.Pop<u32>()); - LOG_DEBUG(Service_FS, "called file {} mode {}", name, static_cast<u32>(mode)); + LOG_DEBUG(Service_FS, "called. file={}, mode={}", name, static_cast<u32>(mode)); auto result = backend.OpenFile(name, mode); if (result.Failed()) { @@ -430,13 +420,13 @@ public: void OpenDirectory(Kernel::HLERequestContext& ctx) { IPC::RequestParser rp{ctx}; - auto file_buffer = ctx.ReadBuffer(); - std::string name = Common::StringFromBuffer(file_buffer); + const auto file_buffer = ctx.ReadBuffer(); + const std::string name = Common::StringFromBuffer(file_buffer); // TODO(Subv): Implement this filter. - u32 filter_flags = rp.Pop<u32>(); + const u32 filter_flags = rp.Pop<u32>(); - LOG_DEBUG(Service_FS, "called directory {} filter {}", name, filter_flags); + LOG_DEBUG(Service_FS, "called. directory={}, filter={}", name, filter_flags); auto result = backend.OpenDirectory(name); if (result.Failed()) { @@ -453,12 +443,10 @@ public: } void GetEntryType(Kernel::HLERequestContext& ctx) { - IPC::RequestParser rp{ctx}; - - auto file_buffer = ctx.ReadBuffer(); - std::string name = Common::StringFromBuffer(file_buffer); + const auto file_buffer = ctx.ReadBuffer(); + const std::string name = Common::StringFromBuffer(file_buffer); - LOG_DEBUG(Service_FS, "called file {}", name); + LOG_DEBUG(Service_FS, "called. file={}", name); auto result = backend.GetEntryType(name); if (result.Failed()) { @@ -616,7 +604,9 @@ private: u64_le save_id; u64_le title_id; u64_le save_image_size; - INSERT_PADDING_BYTES(0x28); + u16_le index; + FileSys::SaveDataRank rank; + INSERT_PADDING_BYTES(0x25); }; static_assert(sizeof(SaveDataInfo) == 0x60, "SaveDataInfo has incorrect size."); @@ -779,16 +769,17 @@ void FSP_SRV::CreateSaveDataFileSystem(Kernel::HLERequestContext& ctx) { } void FSP_SRV::OpenSaveDataFileSystem(Kernel::HLERequestContext& ctx) { - IPC::RequestParser rp{ctx}; - - auto space_id = rp.PopRaw<FileSys::SaveDataSpaceId>(); - auto unk = rp.Pop<u32>(); - LOG_INFO(Service_FS, "called with unknown={:08X}", unk); + LOG_INFO(Service_FS, "called."); - auto save_struct = rp.PopRaw<FileSys::SaveDataDescriptor>(); + struct Parameters { + FileSys::SaveDataSpaceId save_data_space_id; + FileSys::SaveDataDescriptor descriptor; + }; - auto dir = OpenSaveData(space_id, save_struct); + IPC::RequestParser rp{ctx}; + const auto parameters = rp.PopRaw<Parameters>(); + auto dir = OpenSaveData(parameters.save_data_space_id, parameters.descriptor); if (dir.Failed()) { IPC::ResponseBuilder rb{ctx, 2, 0, 0}; rb.Push(FileSys::ERROR_ENTITY_NOT_FOUND); diff --git a/src/core/hle/service/nvdrv/devices/nvdisp_disp0.h b/src/core/hle/service/nvdrv/devices/nvdisp_disp0.h index ace71169f..12f3ef825 100644 --- a/src/core/hle/service/nvdrv/devices/nvdisp_disp0.h +++ b/src/core/hle/service/nvdrv/devices/nvdisp_disp0.h @@ -18,7 +18,7 @@ class nvmap; class nvdisp_disp0 final : public nvdevice { public: explicit nvdisp_disp0(std::shared_ptr<nvmap> nvmap_dev); - ~nvdisp_disp0(); + ~nvdisp_disp0() override; u32 ioctl(Ioctl command, const std::vector<u8>& input, std::vector<u8>& output) override; diff --git a/src/core/hle/service/nvdrv/interface.h b/src/core/hle/service/nvdrv/interface.h index fe311b069..5b4889910 100644 --- a/src/core/hle/service/nvdrv/interface.h +++ b/src/core/hle/service/nvdrv/interface.h @@ -17,7 +17,7 @@ namespace Service::Nvidia { class NVDRV final : public ServiceFramework<NVDRV> { public: NVDRV(std::shared_ptr<Module> nvdrv, const char* name); - ~NVDRV(); + ~NVDRV() override; private: void Open(Kernel::HLERequestContext& ctx); diff --git a/src/core/hle/service/nvdrv/nvmemp.h b/src/core/hle/service/nvdrv/nvmemp.h index 5a4dfc1f9..6eafb1346 100644 --- a/src/core/hle/service/nvdrv/nvmemp.h +++ b/src/core/hle/service/nvdrv/nvmemp.h @@ -11,7 +11,7 @@ namespace Service::Nvidia { class NVMEMP final : public ServiceFramework<NVMEMP> { public: NVMEMP(); - ~NVMEMP(); + ~NVMEMP() override; private: void Cmd0(Kernel::HLERequestContext& ctx); diff --git a/src/core/hle/service/service.h b/src/core/hle/service/service.h index 830790269..abbfe5524 100644 --- a/src/core/hle/service/service.h +++ b/src/core/hle/service/service.h @@ -90,7 +90,7 @@ private: Kernel::HLERequestContext& ctx); ServiceFrameworkBase(const char* service_name, u32 max_sessions, InvokerFn* handler_invoker); - ~ServiceFrameworkBase(); + ~ServiceFrameworkBase() override; void RegisterHandlersBase(const FunctionInfoBase* functions, std::size_t n); void ReportUnimplementedFunction(Kernel::HLERequestContext& ctx, const FunctionInfoBase* info); diff --git a/src/core/hle/service/set/set_cal.h b/src/core/hle/service/set/set_cal.h index 583036eac..a0677e815 100644 --- a/src/core/hle/service/set/set_cal.h +++ b/src/core/hle/service/set/set_cal.h @@ -11,7 +11,7 @@ namespace Service::Set { class SET_CAL final : public ServiceFramework<SET_CAL> { public: explicit SET_CAL(); - ~SET_CAL(); + ~SET_CAL() override; }; } // namespace Service::Set diff --git a/src/core/hle/service/sockets/sfdnsres.cpp b/src/core/hle/service/sockets/sfdnsres.cpp index 13ab1d31e..852e71e4b 100644 --- a/src/core/hle/service/sockets/sfdnsres.cpp +++ b/src/core/hle/service/sockets/sfdnsres.cpp @@ -8,12 +8,20 @@ namespace Service::Sockets { void SFDNSRES::GetAddrInfo(Kernel::HLERequestContext& ctx) { + struct Parameters { + u8 use_nsd_resolve; + u32 unknown; + u64 process_id; + }; + IPC::RequestParser rp{ctx}; + const auto parameters = rp.PopRaw<Parameters>(); - LOG_WARNING(Service, "(STUBBED) called"); + LOG_WARNING(Service, + "(STUBBED) called. use_nsd_resolve={}, unknown=0x{:08X}, process_id=0x{:016X}", + parameters.use_nsd_resolve, parameters.unknown, parameters.process_id); IPC::ResponseBuilder rb{ctx, 2}; - rb.Push(RESULT_SUCCESS); } diff --git a/src/core/hle/service/spl/module.cpp b/src/core/hle/service/spl/module.cpp index 8db0c2f13..e724d4ab8 100644 --- a/src/core/hle/service/spl/module.cpp +++ b/src/core/hle/service/spl/module.cpp @@ -26,9 +26,7 @@ Module::Interface::~Interface() = default; void Module::Interface::GetRandomBytes(Kernel::HLERequestContext& ctx) { LOG_DEBUG(Service_SPL, "called"); - IPC::RequestParser rp{ctx}; - - std::size_t size = ctx.GetWriteBufferSize(); + const std::size_t size = ctx.GetWriteBufferSize(); std::uniform_int_distribution<u16> distribution(0, std::numeric_limits<u8>::max()); std::vector<u8> data(size); diff --git a/src/core/hle/service/ssl/ssl.cpp b/src/core/hle/service/ssl/ssl.cpp index af40a1815..f7f87a958 100644 --- a/src/core/hle/service/ssl/ssl.cpp +++ b/src/core/hle/service/ssl/ssl.cpp @@ -64,13 +64,19 @@ public: }; RegisterHandlers(functions); } - ~ISslContext() = default; private: void SetOption(Kernel::HLERequestContext& ctx) { - LOG_WARNING(Service_SSL, "(STUBBED) called"); + struct Parameters { + u8 enable; + u32 option; + }; IPC::RequestParser rp{ctx}; + const auto parameters = rp.PopRaw<Parameters>(); + + LOG_WARNING(Service_SSL, "(STUBBED) called. enable={}, option={}", parameters.enable, + parameters.option); IPC::ResponseBuilder rb{ctx, 2}; rb.Push(RESULT_SUCCESS); diff --git a/src/core/hle/service/vi/vi.cpp b/src/core/hle/service/vi/vi.cpp index 566cd6006..4e17249a9 100644 --- a/src/core/hle/service/vi/vi.cpp +++ b/src/core/hle/service/vi/vi.cpp @@ -498,7 +498,6 @@ public: }; RegisterHandlers(functions); } - ~IHOSBinderDriver() = default; private: enum class TransactionId { @@ -692,7 +691,6 @@ public: }; RegisterHandlers(functions); } - ~ISystemDisplayService() = default; private: void SetLayerZ(Kernel::HLERequestContext& ctx) { @@ -818,7 +816,6 @@ public: }; RegisterHandlers(functions); } - ~IManagerDisplayService() = default; private: void CloseDisplay(Kernel::HLERequestContext& ctx) { @@ -884,7 +881,6 @@ private: class IApplicationDisplayService final : public ServiceFramework<IApplicationDisplayService> { public: explicit IApplicationDisplayService(std::shared_ptr<NVFlinger::NVFlinger> nv_flinger); - ~IApplicationDisplayService() = default; private: enum class ConvertedScaleMode : u64 { @@ -1037,7 +1033,6 @@ private: void ListDisplays(Kernel::HLERequestContext& ctx) { LOG_WARNING(Service_VI, "(STUBBED) called"); - IPC::RequestParser rp{ctx}; DisplayInfo display_info; display_info.width *= static_cast<u64>(Settings::values.resolution_factor); display_info.height *= static_cast<u64>(Settings::values.resolution_factor); diff --git a/src/core/loader/nso.cpp b/src/core/loader/nso.cpp index babc7e646..ffe2eea8a 100644 --- a/src/core/loader/nso.cpp +++ b/src/core/loader/nso.cpp @@ -4,11 +4,12 @@ #include <cinttypes> #include <vector> -#include <lz4.h> + #include "common/common_funcs.h" #include "common/file_util.h" #include "common/hex_util.h" #include "common/logging/log.h" +#include "common/lz4_compression.h" #include "common/swap.h" #include "core/core.h" #include "core/file_sys/patch_manager.h" @@ -35,15 +36,11 @@ static_assert(sizeof(MODHeader) == 0x1c, "MODHeader has incorrect size."); std::vector<u8> DecompressSegment(const std::vector<u8>& compressed_data, const NSOSegmentHeader& header) { - std::vector<u8> uncompressed_data(header.size); - const int bytes_uncompressed = - LZ4_decompress_safe(reinterpret_cast<const char*>(compressed_data.data()), - reinterpret_cast<char*>(uncompressed_data.data()), - static_cast<int>(compressed_data.size()), header.size); - - ASSERT_MSG(bytes_uncompressed == static_cast<int>(header.size) && - bytes_uncompressed == static_cast<int>(uncompressed_data.size()), - "{} != {} != {}", bytes_uncompressed, header.size, uncompressed_data.size()); + const std::vector<u8> uncompressed_data = + Common::Compression::DecompressDataLZ4(compressed_data, header.size); + + ASSERT_MSG(uncompressed_data.size() == static_cast<int>(header.size), "{} != {}", header.size, + uncompressed_data.size()); return uncompressed_data; } diff --git a/src/core/loader/xci.h b/src/core/loader/xci.h index d6995b61e..436f7387c 100644 --- a/src/core/loader/xci.h +++ b/src/core/loader/xci.h @@ -22,7 +22,7 @@ class AppLoader_NCA; class AppLoader_XCI final : public AppLoader { public: explicit AppLoader_XCI(FileSys::VirtualFile file); - ~AppLoader_XCI(); + ~AppLoader_XCI() override; /** * Returns the type of the file diff --git a/src/core/memory.h b/src/core/memory.h index 1d38cdca8..9153e4954 100644 --- a/src/core/memory.h +++ b/src/core/memory.h @@ -28,16 +28,6 @@ constexpr u64 PAGE_MASK = PAGE_SIZE - 1; /// Virtual user-space memory regions enum : VAddr { - /// Read-only page containing kernel and system configuration values. - CONFIG_MEMORY_VADDR = 0x1FF80000, - CONFIG_MEMORY_SIZE = 0x00001000, - CONFIG_MEMORY_VADDR_END = CONFIG_MEMORY_VADDR + CONFIG_MEMORY_SIZE, - - /// Usually read-only page containing mostly values read from hardware. - SHARED_PAGE_VADDR = 0x1FF81000, - SHARED_PAGE_SIZE = 0x00001000, - SHARED_PAGE_VADDR_END = SHARED_PAGE_VADDR + SHARED_PAGE_SIZE, - /// TLS (Thread-Local Storage) related. TLS_ENTRY_SIZE = 0x200, diff --git a/src/video_core/CMakeLists.txt b/src/video_core/CMakeLists.txt index 44c761d3e..242a0d1cd 100644 --- a/src/video_core/CMakeLists.txt +++ b/src/video_core/CMakeLists.txt @@ -139,4 +139,4 @@ endif() create_target_directory_groups(video_core) target_link_libraries(video_core PUBLIC common core) -target_link_libraries(video_core PRIVATE glad lz4_static) +target_link_libraries(video_core PRIVATE glad) diff --git a/src/video_core/engines/fermi_2d.cpp b/src/video_core/engines/fermi_2d.cpp index 03b7ee5d8..55966eef1 100644 --- a/src/video_core/engines/fermi_2d.cpp +++ b/src/video_core/engines/fermi_2d.cpp @@ -6,12 +6,13 @@ #include "common/logging/log.h" #include "common/math_util.h" #include "video_core/engines/fermi_2d.h" +#include "video_core/memory_manager.h" #include "video_core/rasterizer_interface.h" namespace Tegra::Engines { Fermi2D::Fermi2D(VideoCore::RasterizerInterface& rasterizer, MemoryManager& memory_manager) - : memory_manager(memory_manager), rasterizer{rasterizer} {} + : rasterizer{rasterizer}, memory_manager{memory_manager} {} void Fermi2D::CallMethod(const GPU::MethodCall& method_call) { ASSERT_MSG(method_call.method < Regs::NUM_REGS, diff --git a/src/video_core/engines/fermi_2d.h b/src/video_core/engines/fermi_2d.h index 80523e320..2e51b7f13 100644 --- a/src/video_core/engines/fermi_2d.h +++ b/src/video_core/engines/fermi_2d.h @@ -10,7 +10,10 @@ #include "common/common_funcs.h" #include "common/common_types.h" #include "video_core/gpu.h" -#include "video_core/memory_manager.h" + +namespace Tegra { +class MemoryManager; +} namespace VideoCore { class RasterizerInterface; @@ -115,10 +118,9 @@ public: }; } regs{}; - MemoryManager& memory_manager; - private: VideoCore::RasterizerInterface& rasterizer; + MemoryManager& memory_manager; /// Performs the copy from the source surface to the destination surface as configured in the /// registers. diff --git a/src/video_core/engines/kepler_compute.h b/src/video_core/engines/kepler_compute.h index 6575afd0f..fb6cdf432 100644 --- a/src/video_core/engines/kepler_compute.h +++ b/src/video_core/engines/kepler_compute.h @@ -9,7 +9,10 @@ #include "common/common_funcs.h" #include "common/common_types.h" #include "video_core/gpu.h" -#include "video_core/memory_manager.h" + +namespace Tegra { +class MemoryManager; +} namespace Tegra::Engines { @@ -40,10 +43,11 @@ public: static_assert(sizeof(Regs) == Regs::NUM_REGS * sizeof(u32), "KeplerCompute Regs has wrong size"); - MemoryManager& memory_manager; - /// Write the value to the register identified by method. void CallMethod(const GPU::MethodCall& method_call); + +private: + MemoryManager& memory_manager; }; #define ASSERT_REG_POSITION(field_name, position) \ diff --git a/src/video_core/engines/kepler_memory.cpp b/src/video_core/engines/kepler_memory.cpp index e259bf46b..cd51a31d7 100644 --- a/src/video_core/engines/kepler_memory.cpp +++ b/src/video_core/engines/kepler_memory.cpp @@ -5,9 +5,9 @@ #include "common/assert.h" #include "common/logging/log.h" #include "core/core.h" -#include "core/memory.h" #include "video_core/engines/kepler_memory.h" #include "video_core/engines/maxwell_3d.h" +#include "video_core/memory_manager.h" #include "video_core/rasterizer_interface.h" #include "video_core/renderer_base.h" @@ -15,7 +15,7 @@ namespace Tegra::Engines { KeplerMemory::KeplerMemory(Core::System& system, VideoCore::RasterizerInterface& rasterizer, MemoryManager& memory_manager) - : system{system}, memory_manager(memory_manager), rasterizer{rasterizer} {} + : system{system}, rasterizer{rasterizer}, memory_manager{memory_manager} {} KeplerMemory::~KeplerMemory() = default; diff --git a/src/video_core/engines/kepler_memory.h b/src/video_core/engines/kepler_memory.h index 9181e9d80..78b6c3e45 100644 --- a/src/video_core/engines/kepler_memory.h +++ b/src/video_core/engines/kepler_memory.h @@ -10,12 +10,15 @@ #include "common/common_funcs.h" #include "common/common_types.h" #include "video_core/gpu.h" -#include "video_core/memory_manager.h" namespace Core { class System; } +namespace Tegra { +class MemoryManager; +} + namespace VideoCore { class RasterizerInterface; } @@ -82,8 +85,8 @@ public: private: Core::System& system; - MemoryManager& memory_manager; VideoCore::RasterizerInterface& rasterizer; + MemoryManager& memory_manager; void ProcessData(u32 data); }; diff --git a/src/video_core/engines/maxwell_3d.cpp b/src/video_core/engines/maxwell_3d.cpp index defcfbd3f..8194a4b4a 100644 --- a/src/video_core/engines/maxwell_3d.cpp +++ b/src/video_core/engines/maxwell_3d.cpp @@ -7,11 +7,10 @@ #include "common/assert.h" #include "core/core.h" #include "core/core_timing.h" -#include "core/memory.h" #include "video_core/debug_utils/debug_utils.h" #include "video_core/engines/maxwell_3d.h" +#include "video_core/memory_manager.h" #include "video_core/rasterizer_interface.h" -#include "video_core/renderer_base.h" #include "video_core/textures/texture.h" namespace Tegra::Engines { @@ -21,8 +20,8 @@ constexpr u32 MacroRegistersStart = 0xE00; Maxwell3D::Maxwell3D(Core::System& system, VideoCore::RasterizerInterface& rasterizer, MemoryManager& memory_manager) - : memory_manager(memory_manager), system{system}, rasterizer{rasterizer}, - macro_interpreter(*this) { + : system{system}, rasterizer{rasterizer}, memory_manager{memory_manager}, macro_interpreter{ + *this} { InitializeRegisterDefaults(); } @@ -250,6 +249,10 @@ void Maxwell3D::CallMethod(const GPU::MethodCall& method_call) { ProcessQueryGet(); break; } + case MAXWELL3D_REG_INDEX(sync_info): { + ProcessSyncPoint(); + break; + } default: break; } @@ -327,6 +330,14 @@ void Maxwell3D::ProcessQueryGet() { } } +void Maxwell3D::ProcessSyncPoint() { + const u32 sync_point = regs.sync_info.sync_point.Value(); + const u32 increment = regs.sync_info.increment.Value(); + const u32 cache_flush = regs.sync_info.unknown.Value(); + UNIMPLEMENTED_MSG("Syncpoint Set {}, increment: {}, unk: {}", sync_point, increment, + cache_flush); +} + void Maxwell3D::DrawArrays() { LOG_DEBUG(HW_GPU, "called, topology={}, count={}", static_cast<u32>(regs.draw.topology.Value()), regs.vertex_buffer.count); diff --git a/src/video_core/engines/maxwell_3d.h b/src/video_core/engines/maxwell_3d.h index 7fbf1026e..321af3297 100644 --- a/src/video_core/engines/maxwell_3d.h +++ b/src/video_core/engines/maxwell_3d.h @@ -16,13 +16,16 @@ #include "common/math_util.h" #include "video_core/gpu.h" #include "video_core/macro_interpreter.h" -#include "video_core/memory_manager.h" #include "video_core/textures/texture.h" namespace Core { class System; } +namespace Tegra { +class MemoryManager; +} + namespace VideoCore { class RasterizerInterface; } @@ -576,7 +579,17 @@ public: u32 bind; } macros; - INSERT_PADDING_WORDS(0x188); + INSERT_PADDING_WORDS(0x69); + + struct { + union { + BitField<0, 16, u32> sync_point; + BitField<16, 1, u32> unknown; + BitField<20, 1, u32> increment; + }; + } sync_info; + + INSERT_PADDING_WORDS(0x11E); u32 tfb_enabled; @@ -1093,7 +1106,6 @@ public: }; State state{}; - MemoryManager& memory_manager; struct DirtyFlags { std::bitset<8> color_buffer{0xFF}; @@ -1141,6 +1153,8 @@ private: VideoCore::RasterizerInterface& rasterizer; + MemoryManager& memory_manager; + /// Start offsets of each macro in macro_memory std::unordered_map<u32, u32> macro_offsets; @@ -1180,6 +1194,9 @@ private: /// Handles a write to the QUERY_GET register. void ProcessQueryGet(); + /// Handles writes to syncing register. + void ProcessSyncPoint(); + /// Handles a write to the CB_DATA[i] register. void ProcessCBData(u32 value); @@ -1195,6 +1212,7 @@ private: "Field " #field_name " has invalid position") ASSERT_REG_POSITION(macros, 0x45); +ASSERT_REG_POSITION(sync_info, 0xB2); ASSERT_REG_POSITION(tfb_enabled, 0x1D1); ASSERT_REG_POSITION(rt, 0x200); ASSERT_REG_POSITION(viewport_transform, 0x280); diff --git a/src/video_core/engines/maxwell_dma.cpp b/src/video_core/engines/maxwell_dma.cpp index 5cca5c29a..2426d0067 100644 --- a/src/video_core/engines/maxwell_dma.cpp +++ b/src/video_core/engines/maxwell_dma.cpp @@ -5,9 +5,9 @@ #include "common/assert.h" #include "common/logging/log.h" #include "core/core.h" -#include "core/memory.h" #include "video_core/engines/maxwell_3d.h" #include "video_core/engines/maxwell_dma.h" +#include "video_core/memory_manager.h" #include "video_core/rasterizer_interface.h" #include "video_core/renderer_base.h" #include "video_core/textures/decoders.h" @@ -16,7 +16,7 @@ namespace Tegra::Engines { MaxwellDMA::MaxwellDMA(Core::System& system, VideoCore::RasterizerInterface& rasterizer, MemoryManager& memory_manager) - : memory_manager(memory_manager), system{system}, rasterizer{rasterizer} {} + : system{system}, rasterizer{rasterizer}, memory_manager{memory_manager} {} void MaxwellDMA::CallMethod(const GPU::MethodCall& method_call) { ASSERT_MSG(method_call.method < Regs::NUM_REGS, diff --git a/src/video_core/engines/maxwell_dma.h b/src/video_core/engines/maxwell_dma.h index 34c369320..c6b649842 100644 --- a/src/video_core/engines/maxwell_dma.h +++ b/src/video_core/engines/maxwell_dma.h @@ -10,12 +10,15 @@ #include "common/common_funcs.h" #include "common/common_types.h" #include "video_core/gpu.h" -#include "video_core/memory_manager.h" namespace Core { class System; } +namespace Tegra { +class MemoryManager; +} + namespace VideoCore { class RasterizerInterface; } @@ -139,13 +142,13 @@ public: }; } regs{}; - MemoryManager& memory_manager; - private: Core::System& system; VideoCore::RasterizerInterface& rasterizer; + MemoryManager& memory_manager; + /// Performs the copy from the source buffer to the destination buffer as configured in the /// registers. void HandleCopy(); diff --git a/src/video_core/gpu.cpp b/src/video_core/gpu.cpp index 30b29e14d..4461083ff 100644 --- a/src/video_core/gpu.cpp +++ b/src/video_core/gpu.cpp @@ -31,7 +31,7 @@ u32 FramebufferConfig::BytesPerPixel(PixelFormat format) { GPU::GPU(Core::System& system, VideoCore::RendererBase& renderer) : renderer{renderer} { auto& rasterizer{renderer.Rasterizer()}; - memory_manager = std::make_unique<Tegra::MemoryManager>(); + memory_manager = std::make_unique<Tegra::MemoryManager>(rasterizer); dma_pusher = std::make_unique<Tegra::DmaPusher>(*this); maxwell_3d = std::make_unique<Engines::Maxwell3D>(system, rasterizer, *memory_manager); fermi_2d = std::make_unique<Engines::Fermi2D>(rasterizer, *memory_manager); diff --git a/src/video_core/gpu_asynch.cpp b/src/video_core/gpu_asynch.cpp index 8b355cf7b..db507cf04 100644 --- a/src/video_core/gpu_asynch.cpp +++ b/src/video_core/gpu_asynch.cpp @@ -9,7 +9,7 @@ namespace VideoCommon { GPUAsynch::GPUAsynch(Core::System& system, VideoCore::RendererBase& renderer) - : Tegra::GPU(system, renderer), gpu_thread{renderer, *dma_pusher} {} + : Tegra::GPU(system, renderer), gpu_thread{system, renderer, *dma_pusher} {} GPUAsynch::~GPUAsynch() = default; diff --git a/src/video_core/gpu_thread.cpp b/src/video_core/gpu_thread.cpp index c5dc199c5..cc56cf467 100644 --- a/src/video_core/gpu_thread.cpp +++ b/src/video_core/gpu_thread.cpp @@ -4,6 +4,9 @@ #include "common/assert.h" #include "common/microprofile.h" +#include "core/core.h" +#include "core/core_timing.h" +#include "core/core_timing_util.h" #include "core/frontend/scope_acquire_window_context.h" #include "video_core/dma_pusher.h" #include "video_core/gpu.h" @@ -36,7 +39,6 @@ static void RunThread(VideoCore::RendererBase& renderer, Tegra::DmaPusher& dma_p dma_pusher.Push(std::move(submit_list->entries)); dma_pusher.DispatchCalls(); } else if (const auto data = std::get_if<SwapBuffersCommand>(&next.data)) { - state.DecrementFramesCounter(); renderer.SwapBuffers(std::move(data->framebuffer)); } else if (const auto data = std::get_if<FlushRegionCommand>(&next.data)) { renderer.Rasterizer().FlushRegion(data->addr, data->size); @@ -47,13 +49,18 @@ static void RunThread(VideoCore::RendererBase& renderer, Tegra::DmaPusher& dma_p } else { UNREACHABLE(); } + state.signaled_fence = next.fence; + state.TrySynchronize(); } } } -ThreadManager::ThreadManager(VideoCore::RendererBase& renderer, Tegra::DmaPusher& dma_pusher) - : renderer{renderer}, thread{RunThread, std::ref(renderer), std::ref(dma_pusher), - std::ref(state)} {} +ThreadManager::ThreadManager(Core::System& system, VideoCore::RendererBase& renderer, + Tegra::DmaPusher& dma_pusher) + : system{system}, thread{RunThread, std::ref(renderer), std::ref(dma_pusher), std::ref(state)} { + synchronization_event = system.CoreTiming().RegisterEvent( + "GPUThreadSynch", [this](u64 fence, s64) { state.WaitForSynchronization(fence); }); +} ThreadManager::~ThreadManager() { // Notify GPU thread that a shutdown is pending @@ -62,14 +69,14 @@ ThreadManager::~ThreadManager() { } void ThreadManager::SubmitList(Tegra::CommandList&& entries) { - PushCommand(SubmitListCommand(std::move(entries))); + const u64 fence{PushCommand(SubmitListCommand(std::move(entries)))}; + const s64 synchronization_ticks{Core::Timing::usToCycles(9000)}; + system.CoreTiming().ScheduleEvent(synchronization_ticks, synchronization_event, fence); } void ThreadManager::SwapBuffers( std::optional<std::reference_wrapper<const Tegra::FramebufferConfig>> framebuffer) { - state.IncrementFramesCounter(); PushCommand(SwapBuffersCommand(std::move(framebuffer))); - state.WaitForFrames(); } void ThreadManager::FlushRegion(CacheAddr addr, u64 size) { @@ -79,7 +86,7 @@ void ThreadManager::FlushRegion(CacheAddr addr, u64 size) { void ThreadManager::InvalidateRegion(CacheAddr addr, u64 size) { if (state.queue.Empty()) { // It's quicker to invalidate a single region on the CPU if the queue is already empty - renderer.Rasterizer().InvalidateRegion(addr, size); + system.Renderer().Rasterizer().InvalidateRegion(addr, size); } else { PushCommand(InvalidateRegionCommand(addr, size)); } @@ -90,9 +97,25 @@ void ThreadManager::FlushAndInvalidateRegion(CacheAddr addr, u64 size) { InvalidateRegion(addr, size); } -void ThreadManager::PushCommand(CommandData&& command_data) { - state.queue.Push(CommandDataContainer(std::move(command_data))); +u64 ThreadManager::PushCommand(CommandData&& command_data) { + const u64 fence{++state.last_fence}; + state.queue.Push(CommandDataContainer(std::move(command_data), fence)); state.SignalCommands(); + return fence; +} + +MICROPROFILE_DEFINE(GPU_wait, "GPU", "Wait for the GPU", MP_RGB(128, 128, 192)); +void SynchState::WaitForSynchronization(u64 fence) { + if (signaled_fence >= fence) { + return; + } + + // Wait for the GPU to be idle (all commands to be executed) + { + MICROPROFILE_SCOPE(GPU_wait); + std::unique_lock<std::mutex> lock{synchronization_mutex}; + synchronization_condition.wait(lock, [this, fence] { return signaled_fence >= fence; }); + } } } // namespace VideoCommon::GPUThread diff --git a/src/video_core/gpu_thread.h b/src/video_core/gpu_thread.h index 70acb2e79..62bcea5bb 100644 --- a/src/video_core/gpu_thread.h +++ b/src/video_core/gpu_thread.h @@ -19,9 +19,12 @@ struct FramebufferConfig; class DmaPusher; } // namespace Tegra -namespace VideoCore { -class RendererBase; -} // namespace VideoCore +namespace Core { +class System; +namespace Timing { +struct EventType; +} // namespace Timing +} // namespace Core namespace VideoCommon::GPUThread { @@ -75,63 +78,47 @@ using CommandData = struct CommandDataContainer { CommandDataContainer() = default; - CommandDataContainer(CommandData&& data) : data{std::move(data)} {} + CommandDataContainer(CommandData&& data, u64 next_fence) + : data{std::move(data)}, fence{next_fence} {} CommandDataContainer& operator=(const CommandDataContainer& t) { data = std::move(t.data); + fence = t.fence; return *this; } CommandData data; + u64 fence{}; }; /// Struct used to synchronize the GPU thread struct SynchState final { std::atomic_bool is_running{true}; std::atomic_int queued_frame_count{}; - std::mutex frames_mutex; + std::mutex synchronization_mutex; std::mutex commands_mutex; std::condition_variable commands_condition; - std::condition_variable frames_condition; + std::condition_variable synchronization_condition; - void IncrementFramesCounter() { - std::lock_guard lock{frames_mutex}; - ++queued_frame_count; + /// Returns true if the gap in GPU commands is small enough that we can consider the CPU and GPU + /// synchronized. This is entirely empirical. + bool IsSynchronized() const { + constexpr std::size_t max_queue_gap{5}; + return queue.Size() <= max_queue_gap; } - void DecrementFramesCounter() { - { - std::lock_guard lock{frames_mutex}; - --queued_frame_count; - - if (queued_frame_count) { - return; - } + void TrySynchronize() { + if (IsSynchronized()) { + std::lock_guard<std::mutex> lock{synchronization_mutex}; + synchronization_condition.notify_one(); } - frames_condition.notify_one(); } - void WaitForFrames() { - { - std::lock_guard lock{frames_mutex}; - if (!queued_frame_count) { - return; - } - } - - // Wait for the GPU to be idle (all commands to be executed) - { - std::unique_lock lock{frames_mutex}; - frames_condition.wait(lock, [this] { return !queued_frame_count; }); - } - } + void WaitForSynchronization(u64 fence); void SignalCommands() { - { - std::unique_lock lock{commands_mutex}; - if (queue.Empty()) { - return; - } + if (queue.Empty()) { + return; } commands_condition.notify_one(); @@ -144,12 +131,15 @@ struct SynchState final { using CommandQueue = Common::SPSCQueue<CommandDataContainer>; CommandQueue queue; + u64 last_fence{}; + std::atomic<u64> signaled_fence{}; }; /// Class used to manage the GPU thread class ThreadManager final { public: - explicit ThreadManager(VideoCore::RendererBase& renderer, Tegra::DmaPusher& dma_pusher); + explicit ThreadManager(Core::System& system, VideoCore::RendererBase& renderer, + Tegra::DmaPusher& dma_pusher); ~ThreadManager(); /// Push GPU command entries to be processed @@ -170,11 +160,12 @@ public: private: /// Pushes a command to be executed by the GPU thread - void PushCommand(CommandData&& command_data); + u64 PushCommand(CommandData&& command_data); private: SynchState state; - VideoCore::RendererBase& renderer; + Core::System& system; + Core::Timing::EventType* synchronization_event{}; std::thread thread; std::thread::id thread_id; }; diff --git a/src/video_core/macro_interpreter.cpp b/src/video_core/macro_interpreter.cpp index 64f75db43..524d9ea5a 100644 --- a/src/video_core/macro_interpreter.cpp +++ b/src/video_core/macro_interpreter.cpp @@ -223,27 +223,21 @@ void MacroInterpreter::ProcessResult(ResultOperation operation, u32 reg, u32 res } u32 MacroInterpreter::FetchParameter() { - ASSERT(next_parameter_index < parameters.size()); - return parameters[next_parameter_index++]; + return parameters.at(next_parameter_index++); } u32 MacroInterpreter::GetRegister(u32 register_id) const { - // Register 0 is supposed to always return 0. - if (register_id == 0) - return 0; - - ASSERT(register_id < registers.size()); - return registers[register_id]; + return registers.at(register_id); } void MacroInterpreter::SetRegister(u32 register_id, u32 value) { - // Register 0 is supposed to always return 0. NOP is implemented as a store to the zero - // register. - if (register_id == 0) + // Register 0 is hardwired as the zero register. + // Ensure no writes to it actually occur. + if (register_id == 0) { return; + } - ASSERT(register_id < registers.size()); - registers[register_id] = value; + registers.at(register_id) = value; } void MacroInterpreter::SetMethodAddress(u32 address) { diff --git a/src/video_core/memory_manager.cpp b/src/video_core/memory_manager.cpp index e76b59842..0f4e820aa 100644 --- a/src/video_core/memory_manager.cpp +++ b/src/video_core/memory_manager.cpp @@ -5,16 +5,13 @@ #include "common/alignment.h" #include "common/assert.h" #include "common/logging/log.h" -#include "core/core.h" #include "core/memory.h" -#include "video_core/gpu.h" #include "video_core/memory_manager.h" #include "video_core/rasterizer_interface.h" -#include "video_core/renderer_base.h" namespace Tegra { -MemoryManager::MemoryManager() { +MemoryManager::MemoryManager(VideoCore::RasterizerInterface& rasterizer) : rasterizer{rasterizer} { std::fill(page_table.pointers.begin(), page_table.pointers.end(), nullptr); std::fill(page_table.attributes.begin(), page_table.attributes.end(), Common::PageType::Unmapped); @@ -70,23 +67,23 @@ GPUVAddr MemoryManager::UnmapBuffer(GPUVAddr gpu_addr, u64 size) { const u64 aligned_size{Common::AlignUp(size, page_size)}; const CacheAddr cache_addr{ToCacheAddr(GetPointer(gpu_addr))}; - Core::System::GetInstance().Renderer().Rasterizer().FlushAndInvalidateRegion(cache_addr, - aligned_size); + rasterizer.FlushAndInvalidateRegion(cache_addr, aligned_size); UnmapRange(gpu_addr, aligned_size); return gpu_addr; } -GPUVAddr MemoryManager::FindFreeRegion(GPUVAddr region_start, u64 size) { +GPUVAddr MemoryManager::FindFreeRegion(GPUVAddr region_start, u64 size) const { // Find the first Free VMA. - const VMAHandle vma_handle{std::find_if(vma_map.begin(), vma_map.end(), [&](const auto& vma) { - if (vma.second.type != VirtualMemoryArea::Type::Unmapped) { - return false; - } + const VMAHandle vma_handle{ + std::find_if(vma_map.begin(), vma_map.end(), [region_start, size](const auto& vma) { + if (vma.second.type != VirtualMemoryArea::Type::Unmapped) { + return false; + } - const VAddr vma_end{vma.second.base + vma.second.size}; - return vma_end > region_start && vma_end >= region_start + size; - })}; + const VAddr vma_end{vma.second.base + vma.second.size}; + return vma_end > region_start && vma_end >= region_start + size; + })}; if (vma_handle == vma_map.end()) { return {}; @@ -99,12 +96,12 @@ bool MemoryManager::IsAddressValid(GPUVAddr addr) const { return (addr >> page_bits) < page_table.pointers.size(); } -std::optional<VAddr> MemoryManager::GpuToCpuAddress(GPUVAddr addr) { +std::optional<VAddr> MemoryManager::GpuToCpuAddress(GPUVAddr addr) const { if (!IsAddressValid(addr)) { return {}; } - VAddr cpu_addr{page_table.backing_addr[addr >> page_bits]}; + const VAddr cpu_addr{page_table.backing_addr[addr >> page_bits]}; if (cpu_addr) { return cpu_addr + (addr & page_mask); } @@ -113,7 +110,7 @@ std::optional<VAddr> MemoryManager::GpuToCpuAddress(GPUVAddr addr) { } template <typename T> -T MemoryManager::Read(GPUVAddr addr) { +T MemoryManager::Read(GPUVAddr addr) const { if (!IsAddressValid(addr)) { return {}; } @@ -165,10 +162,10 @@ void MemoryManager::Write(GPUVAddr addr, T data) { } } -template u8 MemoryManager::Read<u8>(GPUVAddr addr); -template u16 MemoryManager::Read<u16>(GPUVAddr addr); -template u32 MemoryManager::Read<u32>(GPUVAddr addr); -template u64 MemoryManager::Read<u64>(GPUVAddr addr); +template u8 MemoryManager::Read<u8>(GPUVAddr addr) const; +template u16 MemoryManager::Read<u16>(GPUVAddr addr) const; +template u32 MemoryManager::Read<u32>(GPUVAddr addr) const; +template u64 MemoryManager::Read<u64>(GPUVAddr addr) const; template void MemoryManager::Write<u8>(GPUVAddr addr, u8 data); template void MemoryManager::Write<u16>(GPUVAddr addr, u16 data); template void MemoryManager::Write<u32>(GPUVAddr addr, u32 data); @@ -179,8 +176,22 @@ u8* MemoryManager::GetPointer(GPUVAddr addr) { return {}; } - u8* page_pointer{page_table.pointers[addr >> page_bits]}; - if (page_pointer) { + u8* const page_pointer{page_table.pointers[addr >> page_bits]}; + if (page_pointer != nullptr) { + return page_pointer + (addr & page_mask); + } + + LOG_ERROR(HW_GPU, "Unknown GetPointer @ 0x{:016X}", addr); + return {}; +} + +const u8* MemoryManager::GetPointer(GPUVAddr addr) const { + if (!IsAddressValid(addr)) { + return {}; + } + + const u8* const page_pointer{page_table.pointers[addr >> page_bits]}; + if (page_pointer != nullptr) { return page_pointer + (addr & page_mask); } @@ -188,15 +199,86 @@ u8* MemoryManager::GetPointer(GPUVAddr addr) { return {}; } -void MemoryManager::ReadBlock(GPUVAddr src_addr, void* dest_buffer, std::size_t size) { - std::memcpy(dest_buffer, GetPointer(src_addr), size); +void MemoryManager::ReadBlock(GPUVAddr src_addr, void* dest_buffer, std::size_t size) const { + std::size_t remaining_size{size}; + std::size_t page_index{src_addr >> page_bits}; + std::size_t page_offset{src_addr & page_mask}; + + while (remaining_size > 0) { + const std::size_t copy_amount{ + std::min(static_cast<std::size_t>(page_size) - page_offset, remaining_size)}; + + 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); + std::memcpy(dest_buffer, src_ptr, copy_amount); + break; + } + default: + UNREACHABLE(); + } + + page_index++; + page_offset = 0; + dest_buffer = static_cast<u8*>(dest_buffer) + copy_amount; + remaining_size -= copy_amount; + } } + void MemoryManager::WriteBlock(GPUVAddr dest_addr, const void* src_buffer, std::size_t size) { - std::memcpy(GetPointer(dest_addr), src_buffer, size); + std::size_t remaining_size{size}; + std::size_t page_index{dest_addr >> page_bits}; + std::size_t page_offset{dest_addr & page_mask}; + + while (remaining_size > 0) { + const std::size_t copy_amount{ + std::min(static_cast<std::size_t>(page_size) - page_offset, remaining_size)}; + + 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); + std::memcpy(dest_ptr, src_buffer, copy_amount); + break; + } + default: + UNREACHABLE(); + } + + page_index++; + page_offset = 0; + src_buffer = static_cast<const u8*>(src_buffer) + copy_amount; + remaining_size -= copy_amount; + } } void MemoryManager::CopyBlock(GPUVAddr dest_addr, GPUVAddr src_addr, std::size_t size) { - std::memcpy(GetPointer(dest_addr), GetPointer(src_addr), size); + std::size_t remaining_size{size}; + std::size_t page_index{src_addr >> page_bits}; + std::size_t page_offset{src_addr & page_mask}; + + while (remaining_size > 0) { + const std::size_t copy_amount{ + std::min(static_cast<std::size_t>(page_size) - page_offset, remaining_size)}; + + 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); + WriteBlock(dest_addr, src_ptr, copy_amount); + break; + } + default: + UNREACHABLE(); + } + + page_index++; + page_offset = 0; + dest_addr += static_cast<VAddr>(copy_amount); + src_addr += static_cast<VAddr>(copy_amount); + remaining_size -= copy_amount; + } } void MemoryManager::MapPages(GPUVAddr base, u64 size, u8* memory, Common::PageType type, @@ -336,7 +418,7 @@ MemoryManager::VMAIter MemoryManager::CarveVMA(GPUVAddr base, u64 size) { const VirtualMemoryArea& vma{vma_handle->second}; if (vma.type == VirtualMemoryArea::Type::Mapped) { // Region is already allocated - return {}; + return vma_handle; } const VAddr start_in_vma{base - vma.base}; diff --git a/src/video_core/memory_manager.h b/src/video_core/memory_manager.h index 34744bb27..647cbf93a 100644 --- a/src/video_core/memory_manager.h +++ b/src/video_core/memory_manager.h @@ -10,6 +10,10 @@ #include "common/common_types.h" #include "common/page_table.h" +namespace VideoCore { +class RasterizerInterface; +} + namespace Tegra { /** @@ -43,24 +47,25 @@ struct VirtualMemoryArea { class MemoryManager final { public: - MemoryManager(); + MemoryManager(VideoCore::RasterizerInterface& rasterizer); GPUVAddr AllocateSpace(u64 size, u64 align); GPUVAddr AllocateSpace(GPUVAddr addr, u64 size, u64 align); GPUVAddr MapBufferEx(VAddr cpu_addr, u64 size); GPUVAddr MapBufferEx(VAddr cpu_addr, GPUVAddr addr, u64 size); GPUVAddr UnmapBuffer(GPUVAddr addr, u64 size); - std::optional<VAddr> GpuToCpuAddress(GPUVAddr addr); + std::optional<VAddr> GpuToCpuAddress(GPUVAddr addr) const; template <typename T> - T Read(GPUVAddr addr); + T Read(GPUVAddr addr) const; template <typename T> void Write(GPUVAddr addr, T data); u8* GetPointer(GPUVAddr addr); + const u8* GetPointer(GPUVAddr addr) const; - void ReadBlock(GPUVAddr src_addr, void* dest_buffer, std::size_t size); + void ReadBlock(GPUVAddr src_addr, void* dest_buffer, std::size_t size) const; void WriteBlock(GPUVAddr dest_addr, const void* src_buffer, std::size_t size); void CopyBlock(GPUVAddr dest_addr, GPUVAddr src_addr, std::size_t size); @@ -127,7 +132,7 @@ private: void UpdatePageTableForVMA(const VirtualMemoryArea& vma); /// Finds a free (unmapped region) of the specified size starting at the specified address. - GPUVAddr FindFreeRegion(GPUVAddr region_start, u64 size); + GPUVAddr FindFreeRegion(GPUVAddr region_start, u64 size) const; private: static constexpr u64 page_bits{16}; @@ -143,6 +148,7 @@ private: Common::PageTable page_table{page_bits}; VMAMap vma_map; + VideoCore::RasterizerInterface& rasterizer; }; } // namespace Tegra diff --git a/src/video_core/renderer_opengl/gl_buffer_cache.cpp b/src/video_core/renderer_opengl/gl_buffer_cache.cpp index fd091c84c..7989ec11b 100644 --- a/src/video_core/renderer_opengl/gl_buffer_cache.cpp +++ b/src/video_core/renderer_opengl/gl_buffer_cache.cpp @@ -7,7 +7,6 @@ #include "common/alignment.h" #include "core/core.h" -#include "core/memory.h" #include "video_core/renderer_opengl/gl_buffer_cache.h" #include "video_core/renderer_opengl/gl_rasterizer.h" diff --git a/src/video_core/renderer_opengl/gl_global_cache.cpp b/src/video_core/renderer_opengl/gl_global_cache.cpp index da9326253..5842d6213 100644 --- a/src/video_core/renderer_opengl/gl_global_cache.cpp +++ b/src/video_core/renderer_opengl/gl_global_cache.cpp @@ -4,7 +4,6 @@ #include <glad/glad.h> -#include "common/assert.h" #include "common/logging/log.h" #include "core/core.h" #include "video_core/renderer_opengl/gl_global_cache.h" diff --git a/src/video_core/renderer_opengl/gl_primitive_assembler.cpp b/src/video_core/renderer_opengl/gl_primitive_assembler.cpp index 2bcbd3da2..c3e94d917 100644 --- a/src/video_core/renderer_opengl/gl_primitive_assembler.cpp +++ b/src/video_core/renderer_opengl/gl_primitive_assembler.cpp @@ -7,7 +7,7 @@ #include "common/assert.h" #include "common/common_types.h" #include "core/core.h" -#include "core/memory.h" +#include "video_core/memory_manager.h" #include "video_core/renderer_opengl/gl_buffer_cache.h" #include "video_core/renderer_opengl/gl_primitive_assembler.h" diff --git a/src/video_core/renderer_opengl/gl_primitive_assembler.h b/src/video_core/renderer_opengl/gl_primitive_assembler.h index 0e2e7dc36..4e87ce4d6 100644 --- a/src/video_core/renderer_opengl/gl_primitive_assembler.h +++ b/src/video_core/renderer_opengl/gl_primitive_assembler.h @@ -4,11 +4,9 @@ #pragma once -#include <vector> #include <glad/glad.h> #include "common/common_types.h" -#include "video_core/memory_manager.h" namespace OpenGL { diff --git a/src/video_core/renderer_opengl/gl_rasterizer.cpp b/src/video_core/renderer_opengl/gl_rasterizer.cpp index 046fc935b..7ff1e6737 100644 --- a/src/video_core/renderer_opengl/gl_rasterizer.cpp +++ b/src/video_core/renderer_opengl/gl_rasterizer.cpp @@ -17,7 +17,6 @@ #include "common/microprofile.h" #include "common/scope_exit.h" #include "core/core.h" -#include "core/frontend/emu_window.h" #include "core/hle/kernel/process.h" #include "core/settings.h" #include "video_core/engines/maxwell_3d.h" @@ -26,7 +25,6 @@ #include "video_core/renderer_opengl/gl_shader_gen.h" #include "video_core/renderer_opengl/maxwell_to_gl.h" #include "video_core/renderer_opengl/renderer_opengl.h" -#include "video_core/video_core.h" namespace OpenGL { @@ -318,7 +316,7 @@ void RasterizerOpenGL::SetupShaders(GLenum primitive_mode) { const std::size_t stage{index == 0 ? 0 : index - 1}; // Stage indices are 0 - 5 GLShader::MaxwellUniformData ubo{}; - ubo.SetFromRegs(gpu.state.shader_stages[stage]); + ubo.SetFromRegs(gpu, stage); const GLintptr offset = buffer_cache.UploadHostMemory( &ubo, sizeof(ubo), static_cast<std::size_t>(uniform_buffer_alignment)); diff --git a/src/video_core/renderer_opengl/gl_rasterizer.h b/src/video_core/renderer_opengl/gl_rasterizer.h index 4de565321..54fbf48aa 100644 --- a/src/video_core/renderer_opengl/gl_rasterizer.h +++ b/src/video_core/renderer_opengl/gl_rasterizer.h @@ -12,15 +12,12 @@ #include <optional> #include <tuple> #include <utility> -#include <vector> #include <boost/icl/interval_map.hpp> -#include <boost/range/iterator_range.hpp> #include <glad/glad.h> #include "common/common_types.h" #include "video_core/engines/maxwell_3d.h" -#include "video_core/memory_manager.h" #include "video_core/rasterizer_cache.h" #include "video_core/rasterizer_interface.h" #include "video_core/renderer_opengl/gl_buffer_cache.h" @@ -29,10 +26,8 @@ #include "video_core/renderer_opengl/gl_rasterizer_cache.h" #include "video_core/renderer_opengl/gl_resource_manager.h" #include "video_core/renderer_opengl/gl_shader_cache.h" -#include "video_core/renderer_opengl/gl_shader_gen.h" #include "video_core/renderer_opengl/gl_shader_manager.h" #include "video_core/renderer_opengl/gl_state.h" -#include "video_core/renderer_opengl/gl_stream_buffer.h" namespace Core { class System; diff --git a/src/video_core/renderer_opengl/gl_rasterizer_cache.cpp b/src/video_core/renderer_opengl/gl_rasterizer_cache.cpp index aba6ce731..7a3280620 100644 --- a/src/video_core/renderer_opengl/gl_rasterizer_cache.cpp +++ b/src/video_core/renderer_opengl/gl_rasterizer_cache.cpp @@ -13,7 +13,6 @@ #include "common/scope_exit.h" #include "core/core.h" #include "core/hle/kernel/process.h" -#include "core/memory.h" #include "core/settings.h" #include "video_core/engines/maxwell_3d.h" #include "video_core/morton.h" diff --git a/src/video_core/renderer_opengl/gl_rasterizer_cache.h b/src/video_core/renderer_opengl/gl_rasterizer_cache.h index e8073579f..ad4fd3ad2 100644 --- a/src/video_core/renderer_opengl/gl_rasterizer_cache.h +++ b/src/video_core/renderer_opengl/gl_rasterizer_cache.h @@ -5,10 +5,9 @@ #pragma once #include <array> -#include <map> #include <memory> #include <string> -#include <unordered_set> +#include <tuple> #include <vector> #include "common/alignment.h" diff --git a/src/video_core/renderer_opengl/gl_shader_cache.cpp b/src/video_core/renderer_opengl/gl_shader_cache.cpp index 290e654bc..7030db365 100644 --- a/src/video_core/renderer_opengl/gl_shader_cache.cpp +++ b/src/video_core/renderer_opengl/gl_shader_cache.cpp @@ -6,13 +6,11 @@ #include "common/assert.h" #include "common/hash.h" #include "core/core.h" -#include "core/memory.h" #include "video_core/engines/maxwell_3d.h" #include "video_core/renderer_opengl/gl_rasterizer.h" #include "video_core/renderer_opengl/gl_shader_cache.h" #include "video_core/renderer_opengl/gl_shader_decompiler.h" #include "video_core/renderer_opengl/gl_shader_disk_cache.h" -#include "video_core/renderer_opengl/gl_shader_manager.h" #include "video_core/renderer_opengl/utils.h" #include "video_core/shader/shader_ir.h" diff --git a/src/video_core/renderer_opengl/gl_shader_cache.h b/src/video_core/renderer_opengl/gl_shader_cache.h index fd1c85115..0cf8e0b3d 100644 --- a/src/video_core/renderer_opengl/gl_shader_cache.h +++ b/src/video_core/renderer_opengl/gl_shader_cache.h @@ -5,21 +5,20 @@ #pragma once #include <array> +#include <atomic> #include <memory> #include <set> #include <tuple> #include <unordered_map> +#include <vector> #include <glad/glad.h> -#include "common/assert.h" #include "common/common_types.h" #include "video_core/rasterizer_cache.h" -#include "video_core/renderer_base.h" #include "video_core/renderer_opengl/gl_resource_manager.h" #include "video_core/renderer_opengl/gl_shader_decompiler.h" #include "video_core/renderer_opengl/gl_shader_disk_cache.h" -#include "video_core/renderer_opengl/gl_shader_gen.h" namespace Core { class System; diff --git a/src/video_core/renderer_opengl/gl_shader_decompiler.cpp b/src/video_core/renderer_opengl/gl_shader_decompiler.cpp index 11d1169f0..a1a51f226 100644 --- a/src/video_core/renderer_opengl/gl_shader_decompiler.cpp +++ b/src/video_core/renderer_opengl/gl_shader_decompiler.cpp @@ -69,10 +69,10 @@ public: shader_source += '\n'; } - std::string GenerateTemporal() { - std::string temporal = "tmp"; - temporal += std::to_string(temporal_index++); - return temporal; + std::string GenerateTemporary() { + std::string temporary = "tmp"; + temporary += std::to_string(temporary_index++); + return temporary; } std::string GetResult() { @@ -87,7 +87,7 @@ private: } std::string shader_source; - u32 temporal_index = 1; + u32 temporary_index = 1; }; /// Generates code to use for a swizzle operation. @@ -426,9 +426,14 @@ private: std::string Visit(Node node) { if (const auto operation = std::get_if<OperationNode>(node)) { const auto operation_index = static_cast<std::size_t>(operation->GetCode()); + if (operation_index >= operation_decompilers.size()) { + UNREACHABLE_MSG("Out of bounds operation: {}", operation_index); + return {}; + } const auto decompiler = operation_decompilers[operation_index]; if (decompiler == nullptr) { - UNREACHABLE_MSG("Operation decompiler {} not defined", operation_index); + UNREACHABLE_MSG("Undefined operation: {}", operation_index); + return {}; } return (this->*decompiler)(*operation); @@ -540,7 +545,7 @@ private: } else if (std::holds_alternative<OperationNode>(*offset)) { // Indirect access - const std::string final_offset = code.GenerateTemporal(); + const std::string final_offset = code.GenerateTemporary(); code.AddLine("uint " + final_offset + " = (ftou(" + Visit(offset) + ") / 4) & " + std::to_string(MAX_CONSTBUFFER_ELEMENTS - 1) + ';'); return fmt::format("{}[{} / 4][{} % 4]", GetConstBuffer(cbuf->GetIndex()), @@ -587,9 +592,9 @@ private: // There's a bug in NVidia's proprietary drivers that makes precise fail on fragment shaders const std::string precise = stage != ShaderStage::Fragment ? "precise " : ""; - const std::string temporal = code.GenerateTemporal(); - code.AddLine(precise + "float " + temporal + " = " + value + ';'); - return temporal; + const std::string temporary = code.GenerateTemporary(); + code.AddLine(precise + "float " + temporary + " = " + value + ';'); + return temporary; } std::string VisitOperand(Operation operation, std::size_t operand_index) { @@ -601,9 +606,9 @@ private: return Visit(operand); } - const std::string temporal = code.GenerateTemporal(); - code.AddLine("float " + temporal + " = " + Visit(operand) + ';'); - return temporal; + const std::string temporary = code.GenerateTemporary(); + code.AddLine("float " + temporary + " = " + Visit(operand) + ';'); + return temporary; } std::string VisitOperand(Operation operation, std::size_t operand_index, Type type) { @@ -1196,11 +1201,12 @@ private: switch (meta->element) { case 0: case 1: - return "textureSize(" + sampler + ", " + lod + ')' + GetSwizzle(meta->element); + return "itof(int(textureSize(" + sampler + ", " + lod + ')' + + GetSwizzle(meta->element) + "))"; case 2: return "0"; case 3: - return "textureQueryLevels(" + sampler + ')'; + return "itof(textureQueryLevels(" + sampler + "))"; } UNREACHABLE(); return "0"; diff --git a/src/video_core/renderer_opengl/gl_shader_decompiler.h b/src/video_core/renderer_opengl/gl_shader_decompiler.h index 72aca4938..4e04ab2f8 100644 --- a/src/video_core/renderer_opengl/gl_shader_decompiler.h +++ b/src/video_core/renderer_opengl/gl_shader_decompiler.h @@ -5,7 +5,6 @@ #pragma once #include <array> -#include <set> #include <string> #include <utility> #include <vector> diff --git a/src/video_core/renderer_opengl/gl_shader_disk_cache.cpp b/src/video_core/renderer_opengl/gl_shader_disk_cache.cpp index 82fc4d44b..8a43eb157 100644 --- a/src/video_core/renderer_opengl/gl_shader_disk_cache.cpp +++ b/src/video_core/renderer_opengl/gl_shader_disk_cache.cpp @@ -4,7 +4,6 @@ #include <cstring> #include <fmt/format.h> -#include <lz4.h> #include "common/assert.h" #include "common/common_paths.h" @@ -12,6 +11,7 @@ #include "common/file_util.h" #include "common/logging/log.h" #include "common/scm_rev.h" +#include "common/zstd_compression.h" #include "core/core.h" #include "core/hle/kernel/process.h" @@ -49,39 +49,6 @@ ShaderCacheVersionHash GetShaderCacheVersionHash() { return hash; } -template <typename T> -std::vector<u8> CompressData(const T* source, std::size_t source_size) { - if (source_size > LZ4_MAX_INPUT_SIZE) { - // Source size exceeds LZ4 maximum input size - return {}; - } - const auto source_size_int = static_cast<int>(source_size); - const int max_compressed_size = LZ4_compressBound(source_size_int); - std::vector<u8> compressed(max_compressed_size); - const int compressed_size = LZ4_compress_default(reinterpret_cast<const char*>(source), - reinterpret_cast<char*>(compressed.data()), - source_size_int, max_compressed_size); - if (compressed_size <= 0) { - // Compression failed - return {}; - } - compressed.resize(compressed_size); - return compressed; -} - -std::vector<u8> DecompressData(const std::vector<u8>& compressed, std::size_t uncompressed_size) { - std::vector<u8> uncompressed(uncompressed_size); - const int size_check = LZ4_decompress_safe(reinterpret_cast<const char*>(compressed.data()), - reinterpret_cast<char*>(uncompressed.data()), - static_cast<int>(compressed.size()), - static_cast<int>(uncompressed.size())); - if (static_cast<int>(uncompressed_size) != size_check) { - // Decompression failed - return {}; - } - return uncompressed; -} - } // namespace ShaderDiskCacheRaw::ShaderDiskCacheRaw(u64 unique_identifier, Maxwell::ShaderProgram program_type, @@ -292,7 +259,7 @@ ShaderDiskCacheOpenGL::LoadPrecompiledFile(FileUtil::IOFile& file) { return {}; } - dump.binary = DecompressData(compressed_binary, binary_length); + dump.binary = Common::Compression::DecompressDataZSTD(compressed_binary); if (dump.binary.empty()) { return {}; } @@ -321,7 +288,7 @@ std::optional<ShaderDiskCacheDecompiled> ShaderDiskCacheOpenGL::LoadDecompiledEn return {}; } - const std::vector<u8> code = DecompressData(compressed_code, code_size); + const std::vector<u8> code = Common::Compression::DecompressDataZSTD(compressed_code); if (code.empty()) { return {}; } @@ -507,7 +474,8 @@ void ShaderDiskCacheOpenGL::SaveDecompiled(u64 unique_identifier, const std::str if (!IsUsable()) return; - const std::vector<u8> compressed_code{CompressData(code.data(), code.size())}; + const std::vector<u8> compressed_code{Common::Compression::CompressDataZSTDDefault( + reinterpret_cast<const u8*>(code.data()), code.size())}; if (compressed_code.empty()) { LOG_ERROR(Render_OpenGL, "Failed to compress GLSL code - skipping shader {:016x}", unique_identifier); @@ -537,7 +505,9 @@ void ShaderDiskCacheOpenGL::SaveDump(const ShaderDiskCacheUsage& usage, GLuint p std::vector<u8> binary(binary_length); glGetProgramBinary(program, binary_length, nullptr, &binary_format, binary.data()); - const std::vector<u8> compressed_binary = CompressData(binary.data(), binary.size()); + const std::vector<u8> compressed_binary = + Common::Compression::CompressDataZSTDDefault(binary.data(), binary.size()); + if (compressed_binary.empty()) { LOG_ERROR(Render_OpenGL, "Failed to compress binary program in shader={:016x}", usage.unique_identifier); diff --git a/src/video_core/renderer_opengl/gl_shader_gen.cpp b/src/video_core/renderer_opengl/gl_shader_gen.cpp index 7d96649af..8763d9c71 100644 --- a/src/video_core/renderer_opengl/gl_shader_gen.cpp +++ b/src/video_core/renderer_opengl/gl_shader_gen.cpp @@ -3,7 +3,6 @@ // Refer to the license.txt file included. #include <fmt/format.h> -#include "common/assert.h" #include "video_core/engines/maxwell_3d.h" #include "video_core/renderer_opengl/gl_shader_decompiler.h" #include "video_core/renderer_opengl/gl_shader_gen.h" diff --git a/src/video_core/renderer_opengl/gl_shader_gen.h b/src/video_core/renderer_opengl/gl_shader_gen.h index fba8e681b..fad346b48 100644 --- a/src/video_core/renderer_opengl/gl_shader_gen.h +++ b/src/video_core/renderer_opengl/gl_shader_gen.h @@ -4,12 +4,9 @@ #pragma once -#include <array> -#include <string> #include <vector> #include "common/common_types.h" -#include "video_core/engines/shader_bytecode.h" #include "video_core/renderer_opengl/gl_shader_decompiler.h" #include "video_core/shader/shader_ir.h" diff --git a/src/video_core/renderer_opengl/gl_shader_manager.cpp b/src/video_core/renderer_opengl/gl_shader_manager.cpp index 6a30c28d2..eaf3e03a0 100644 --- a/src/video_core/renderer_opengl/gl_shader_manager.cpp +++ b/src/video_core/renderer_opengl/gl_shader_manager.cpp @@ -2,15 +2,15 @@ // Licensed under GPLv2 or any later version // Refer to the license.txt file included. -#include "core/core.h" #include "video_core/renderer_opengl/gl_shader_manager.h" namespace OpenGL::GLShader { -void MaxwellUniformData::SetFromRegs(const Maxwell3D::State::ShaderStageInfo& shader_stage) { - const auto& gpu = Core::System::GetInstance().GPU().Maxwell3D(); - const auto& regs = gpu.regs; - const auto& state = gpu.state; +using Tegra::Engines::Maxwell3D; + +void MaxwellUniformData::SetFromRegs(const Maxwell3D& maxwell, std::size_t shader_stage) { + const auto& regs = maxwell.regs; + const auto& state = maxwell.state; // TODO(bunnei): Support more than one viewport viewport_flip[0] = regs.viewport_transform[0].scale_x < 0.0 ? -1.0f : 1.0f; @@ -18,7 +18,7 @@ void MaxwellUniformData::SetFromRegs(const Maxwell3D::State::ShaderStageInfo& sh u32 func = static_cast<u32>(regs.alpha_test_func); // Normalize the gl variants of opCompare to be the same as the normal variants - u32 op_gl_variant_base = static_cast<u32>(Tegra::Engines::Maxwell3D::Regs::ComparisonOp::Never); + const u32 op_gl_variant_base = static_cast<u32>(Maxwell3D::Regs::ComparisonOp::Never); if (func >= op_gl_variant_base) { func = func - op_gl_variant_base + 1U; } @@ -31,8 +31,9 @@ void MaxwellUniformData::SetFromRegs(const Maxwell3D::State::ShaderStageInfo& sh // Assign in which stage the position has to be flipped // (the last stage before the fragment shader). - if (gpu.regs.shader_config[static_cast<u32>(Maxwell3D::Regs::ShaderProgram::Geometry)].enable) { - flip_stage = static_cast<u32>(Maxwell3D::Regs::ShaderProgram::Geometry); + constexpr u32 geometry_index = static_cast<u32>(Maxwell3D::Regs::ShaderProgram::Geometry); + if (maxwell.regs.shader_config[geometry_index].enable) { + flip_stage = geometry_index; } else { flip_stage = static_cast<u32>(Maxwell3D::Regs::ShaderProgram::VertexB); } diff --git a/src/video_core/renderer_opengl/gl_shader_manager.h b/src/video_core/renderer_opengl/gl_shader_manager.h index 316b47910..37dcfefdb 100644 --- a/src/video_core/renderer_opengl/gl_shader_manager.h +++ b/src/video_core/renderer_opengl/gl_shader_manager.h @@ -12,14 +12,13 @@ namespace OpenGL::GLShader { -using Tegra::Engines::Maxwell3D; - /// Uniform structure for the Uniform Buffer Object, all vectors must be 16-byte aligned -// NOTE: Always keep a vec4 at the end. The GL spec is not clear whether the alignment at -// the end of a uniform block is included in UNIFORM_BLOCK_DATA_SIZE or not. -// Not following that rule will cause problems on some AMD drivers. +/// @note Always keep a vec4 at the end. The GL spec is not clear whether the alignment at +/// the end of a uniform block is included in UNIFORM_BLOCK_DATA_SIZE or not. +/// Not following that rule will cause problems on some AMD drivers. struct MaxwellUniformData { - void SetFromRegs(const Maxwell3D::State::ShaderStageInfo& shader_stage); + void SetFromRegs(const Tegra::Engines::Maxwell3D& maxwell, std::size_t shader_stage); + alignas(16) GLvec4 viewport_flip; struct alignas(16) { GLuint instance_id; diff --git a/src/video_core/renderer_opengl/renderer_opengl.cpp b/src/video_core/renderer_opengl/renderer_opengl.cpp index a01efeb05..d69cba9c3 100644 --- a/src/video_core/renderer_opengl/renderer_opengl.cpp +++ b/src/video_core/renderer_opengl/renderer_opengl.cpp @@ -5,7 +5,6 @@ #include <algorithm> #include <cstddef> #include <cstdlib> -#include <cstring> #include <memory> #include <glad/glad.h> #include "common/assert.h" diff --git a/src/video_core/renderer_vulkan/vk_buffer_cache.cpp b/src/video_core/renderer_vulkan/vk_buffer_cache.cpp index 388b5ffd5..02a9f5ecb 100644 --- a/src/video_core/renderer_vulkan/vk_buffer_cache.cpp +++ b/src/video_core/renderer_vulkan/vk_buffer_cache.cpp @@ -10,6 +10,7 @@ #include "common/alignment.h" #include "common/assert.h" #include "core/memory.h" +#include "video_core/memory_manager.h" #include "video_core/renderer_vulkan/declarations.h" #include "video_core/renderer_vulkan/vk_buffer_cache.h" #include "video_core/renderer_vulkan/vk_scheduler.h" diff --git a/src/yuzu/debugger/graphics/graphics_surface.cpp b/src/yuzu/debugger/graphics/graphics_surface.cpp index 11023ed63..f2d14becf 100644 --- a/src/yuzu/debugger/graphics/graphics_surface.cpp +++ b/src/yuzu/debugger/graphics/graphics_surface.cpp @@ -7,6 +7,7 @@ #include <QDebug> #include <QFileDialog> #include <QLabel> +#include <QMessageBox> #include <QMouseEvent> #include <QPushButton> #include <QScrollArea> @@ -95,50 +96,91 @@ GraphicsSurfaceWidget::GraphicsSurfaceWidget(std::shared_ptr<Tegra::DebugContext surface_picker_y_control = new QSpinBox; surface_picker_y_control->setRange(0, max_dimension - 1); - surface_format_control = new QComboBox; - + // clang-format off // Color formats sorted by Maxwell texture format index - surface_format_control->addItem(tr("None")); - surface_format_control->addItem(tr("Unknown")); - surface_format_control->addItem(tr("Unknown")); - surface_format_control->addItem(tr("Unknown")); - surface_format_control->addItem(tr("Unknown")); - surface_format_control->addItem(tr("Unknown")); - surface_format_control->addItem(tr("Unknown")); - surface_format_control->addItem(tr("Unknown")); - surface_format_control->addItem(tr("A8R8G8B8")); - surface_format_control->addItem(tr("Unknown")); - surface_format_control->addItem(tr("Unknown")); - surface_format_control->addItem(tr("Unknown")); - surface_format_control->addItem(tr("Unknown")); - surface_format_control->addItem(tr("Unknown")); - surface_format_control->addItem(tr("Unknown")); - surface_format_control->addItem(tr("Unknown")); - surface_format_control->addItem(tr("Unknown")); - surface_format_control->addItem(tr("Unknown")); - surface_format_control->addItem(tr("Unknown")); - surface_format_control->addItem(tr("Unknown")); - surface_format_control->addItem(tr("Unknown")); - surface_format_control->addItem(tr("Unknown")); - surface_format_control->addItem(tr("Unknown")); - surface_format_control->addItem(tr("Unknown")); - surface_format_control->addItem(tr("Unknown")); - surface_format_control->addItem(tr("Unknown")); - surface_format_control->addItem(tr("Unknown")); - surface_format_control->addItem(tr("Unknown")); - surface_format_control->addItem(tr("Unknown")); - surface_format_control->addItem(tr("Unknown")); - surface_format_control->addItem(tr("Unknown")); - surface_format_control->addItem(tr("Unknown")); - surface_format_control->addItem(tr("Unknown")); - surface_format_control->addItem(tr("Unknown")); - surface_format_control->addItem(tr("Unknown")); - surface_format_control->addItem(tr("Unknown")); - surface_format_control->addItem(tr("DXT1")); - surface_format_control->addItem(tr("DXT23")); - surface_format_control->addItem(tr("DXT45")); - surface_format_control->addItem(tr("DXN1")); - surface_format_control->addItem(tr("DXN2")); + const QStringList surface_formats{ + tr("None"), + QStringLiteral("R32_G32_B32_A32"), + QStringLiteral("R32_G32_B32"), + QStringLiteral("R16_G16_B16_A16"), + QStringLiteral("R32_G32"), + QStringLiteral("R32_B24G8"), + QStringLiteral("ETC2_RGB"), + QStringLiteral("X8B8G8R8"), + QStringLiteral("A8R8G8B8"), + QStringLiteral("A2B10G10R10"), + QStringLiteral("ETC2_RGB_PTA"), + QStringLiteral("ETC2_RGBA"), + QStringLiteral("R16_G16"), + QStringLiteral("G8R24"), + QStringLiteral("G24R8"), + QStringLiteral("R32"), + QStringLiteral("BC6H_SF16"), + QStringLiteral("BC6H_UF16"), + QStringLiteral("A4B4G4R4"), + QStringLiteral("A5B5G5R1"), + QStringLiteral("A1B5G5R5"), + QStringLiteral("B5G6R5"), + QStringLiteral("B6G5R5"), + QStringLiteral("BC7U"), + QStringLiteral("G8R8"), + QStringLiteral("EAC"), + QStringLiteral("EACX2"), + QStringLiteral("R16"), + QStringLiteral("Y8_VIDEO"), + QStringLiteral("R8"), + QStringLiteral("G4R4"), + QStringLiteral("R1"), + QStringLiteral("E5B9G9R9_SHAREDEXP"), + QStringLiteral("BF10GF11RF11"), + QStringLiteral("G8B8G8R8"), + QStringLiteral("B8G8R8G8"), + QStringLiteral("DXT1"), + QStringLiteral("DXT23"), + QStringLiteral("DXT45"), + QStringLiteral("DXN1"), + QStringLiteral("DXN2"), + QStringLiteral("Z24S8"), + QStringLiteral("X8Z24"), + QStringLiteral("S8Z24"), + QStringLiteral("X4V4Z24__COV4R4V"), + QStringLiteral("X4V4Z24__COV8R8V"), + QStringLiteral("V8Z24__COV4R12V"), + QStringLiteral("ZF32"), + QStringLiteral("ZF32_X24S8"), + QStringLiteral("X8Z24_X20V4S8__COV4R4V"), + QStringLiteral("X8Z24_X20V4S8__COV8R8V"), + QStringLiteral("ZF32_X20V4X8__COV4R4V"), + QStringLiteral("ZF32_X20V4X8__COV8R8V"), + QStringLiteral("ZF32_X20V4S8__COV4R4V"), + QStringLiteral("ZF32_X20V4S8__COV8R8V"), + QStringLiteral("X8Z24_X16V8S8__COV4R12V"), + QStringLiteral("ZF32_X16V8X8__COV4R12V"), + QStringLiteral("ZF32_X16V8S8__COV4R12V"), + QStringLiteral("Z16"), + QStringLiteral("V8Z24__COV8R24V"), + QStringLiteral("X8Z24_X16V8S8__COV8R24V"), + QStringLiteral("ZF32_X16V8X8__COV8R24V"), + QStringLiteral("ZF32_X16V8S8__COV8R24V"), + QStringLiteral("ASTC_2D_4X4"), + QStringLiteral("ASTC_2D_5X5"), + QStringLiteral("ASTC_2D_6X6"), + QStringLiteral("ASTC_2D_8X8"), + QStringLiteral("ASTC_2D_10X10"), + QStringLiteral("ASTC_2D_12X12"), + QStringLiteral("ASTC_2D_5X4"), + QStringLiteral("ASTC_2D_6X5"), + QStringLiteral("ASTC_2D_8X6"), + QStringLiteral("ASTC_2D_10X8"), + QStringLiteral("ASTC_2D_12X10"), + QStringLiteral("ASTC_2D_8X5"), + QStringLiteral("ASTC_2D_10X5"), + QStringLiteral("ASTC_2D_10X6"), + }; + // clang-format on + + surface_format_control = new QComboBox; + surface_format_control->addItems(surface_formats); surface_info_label = new QLabel(); surface_info_label->setWordWrap(true); @@ -157,22 +199,20 @@ GraphicsSurfaceWidget::GraphicsSurfaceWidget(std::shared_ptr<Tegra::DebugContext // Connections connect(this, &GraphicsSurfaceWidget::Update, this, &GraphicsSurfaceWidget::OnUpdate); - connect(surface_source_list, - static_cast<void (QComboBox::*)(int)>(&QComboBox::currentIndexChanged), this, + connect(surface_source_list, qOverload<int>(&QComboBox::currentIndexChanged), this, &GraphicsSurfaceWidget::OnSurfaceSourceChanged); connect(surface_address_control, &CSpinBox::ValueChanged, this, &GraphicsSurfaceWidget::OnSurfaceAddressChanged); - connect(surface_width_control, static_cast<void (QSpinBox::*)(int)>(&QSpinBox::valueChanged), - this, &GraphicsSurfaceWidget::OnSurfaceWidthChanged); - connect(surface_height_control, static_cast<void (QSpinBox::*)(int)>(&QSpinBox::valueChanged), - this, &GraphicsSurfaceWidget::OnSurfaceHeightChanged); - connect(surface_format_control, - static_cast<void (QComboBox::*)(int)>(&QComboBox::currentIndexChanged), this, + connect(surface_width_control, qOverload<int>(&QSpinBox::valueChanged), this, + &GraphicsSurfaceWidget::OnSurfaceWidthChanged); + connect(surface_height_control, qOverload<int>(&QSpinBox::valueChanged), this, + &GraphicsSurfaceWidget::OnSurfaceHeightChanged); + connect(surface_format_control, qOverload<int>(&QComboBox::currentIndexChanged), this, &GraphicsSurfaceWidget::OnSurfaceFormatChanged); - connect(surface_picker_x_control, static_cast<void (QSpinBox::*)(int)>(&QSpinBox::valueChanged), - this, &GraphicsSurfaceWidget::OnSurfacePickerXChanged); - connect(surface_picker_y_control, static_cast<void (QSpinBox::*)(int)>(&QSpinBox::valueChanged), - this, &GraphicsSurfaceWidget::OnSurfacePickerYChanged); + connect(surface_picker_x_control, qOverload<int>(&QSpinBox::valueChanged), this, + &GraphicsSurfaceWidget::OnSurfacePickerXChanged); + connect(surface_picker_y_control, qOverload<int>(&QSpinBox::valueChanged), this, + &GraphicsSurfaceWidget::OnSurfacePickerYChanged); connect(save_surface, &QPushButton::clicked, this, &GraphicsSurfaceWidget::SaveSurface); auto main_widget = new QWidget; @@ -420,40 +460,56 @@ void GraphicsSurfaceWidget::OnUpdate() { } void GraphicsSurfaceWidget::SaveSurface() { - QString png_filter = tr("Portable Network Graphic (*.png)"); - QString bin_filter = tr("Binary data (*.bin)"); + const QString png_filter = tr("Portable Network Graphic (*.png)"); + const QString bin_filter = tr("Binary data (*.bin)"); - QString selectedFilter; - QString filename = QFileDialog::getSaveFileName( + QString selected_filter; + const QString filename = QFileDialog::getSaveFileName( this, tr("Save Surface"), - QString("texture-0x%1.png").arg(QString::number(surface_address, 16)), - QString("%1;;%2").arg(png_filter, bin_filter), &selectedFilter); + QStringLiteral("texture-0x%1.png").arg(QString::number(surface_address, 16)), + QStringLiteral("%1;;%2").arg(png_filter, bin_filter), &selected_filter); if (filename.isEmpty()) { // If the user canceled the dialog, don't save anything. return; } - if (selectedFilter == png_filter) { - const QPixmap* pixmap = surface_picture_label->pixmap(); + if (selected_filter == png_filter) { + const QPixmap* const pixmap = surface_picture_label->pixmap(); ASSERT_MSG(pixmap != nullptr, "No pixmap set"); - QFile file(filename); - file.open(QIODevice::WriteOnly); - if (pixmap) - pixmap->save(&file, "PNG"); - } else if (selectedFilter == bin_filter) { + QFile file{filename}; + if (!file.open(QIODevice::WriteOnly)) { + QMessageBox::warning(this, tr("Error"), tr("Failed to open file '%1'").arg(filename)); + return; + } + + if (!pixmap->save(&file, "PNG")) { + QMessageBox::warning(this, tr("Error"), + tr("Failed to save surface data to file '%1'").arg(filename)); + } + } else if (selected_filter == bin_filter) { auto& gpu = Core::System::GetInstance().GPU(); - std::optional<VAddr> address = gpu.MemoryManager().GpuToCpuAddress(surface_address); + const std::optional<VAddr> address = gpu.MemoryManager().GpuToCpuAddress(surface_address); - const u8* buffer = Memory::GetPointer(*address); + const u8* const buffer = Memory::GetPointer(*address); ASSERT_MSG(buffer != nullptr, "Memory not accessible"); - QFile file(filename); - file.open(QIODevice::WriteOnly); - int size = surface_width * surface_height * Tegra::Texture::BytesPerPixel(surface_format); - QByteArray data(reinterpret_cast<const char*>(buffer), size); - file.write(data); + QFile file{filename}; + if (!file.open(QIODevice::WriteOnly)) { + QMessageBox::warning(this, tr("Error"), tr("Failed to open file '%1'").arg(filename)); + return; + } + + const int size = + surface_width * surface_height * Tegra::Texture::BytesPerPixel(surface_format); + const QByteArray data(reinterpret_cast<const char*>(buffer), size); + if (file.write(data) != data.size()) { + QMessageBox::warning( + this, tr("Error"), + tr("Failed to completely write surface data to file. The saved data will " + "likely be corrupt.")); + } } else { UNREACHABLE_MSG("Unhandled filter selected"); } diff --git a/src/yuzu/game_list.cpp b/src/yuzu/game_list.cpp index c0e3c5fa9..4422a572b 100644 --- a/src/yuzu/game_list.cpp +++ b/src/yuzu/game_list.cpp @@ -329,6 +329,8 @@ void GameList::PopupContextMenu(const QPoint& menu_location) { QMenu context_menu; QAction* open_save_location = context_menu.addAction(tr("Open Save Data Location")); QAction* open_lfs_location = context_menu.addAction(tr("Open Mod Data Location")); + QAction* open_transferable_shader_cache = + context_menu.addAction(tr("Open Transferable Shader Cache")); context_menu.addSeparator(); QAction* dump_romfs = context_menu.addAction(tr("Dump RomFS")); QAction* copy_tid = context_menu.addAction(tr("Copy Title ID to Clipboard")); @@ -344,6 +346,8 @@ void GameList::PopupContextMenu(const QPoint& menu_location) { [&]() { emit OpenFolderRequested(program_id, GameListOpenTarget::SaveData); }); connect(open_lfs_location, &QAction::triggered, [&]() { emit OpenFolderRequested(program_id, GameListOpenTarget::ModData); }); + connect(open_transferable_shader_cache, &QAction::triggered, + [&]() { emit OpenTransferableShaderCacheRequested(program_id); }); connect(dump_romfs, &QAction::triggered, [&]() { emit DumpRomFSRequested(program_id, path); }); connect(copy_tid, &QAction::triggered, [&]() { emit CopyTIDRequested(program_id); }); connect(navigate_to_gamedb_entry, &QAction::triggered, diff --git a/src/yuzu/game_list.h b/src/yuzu/game_list.h index b317eb2fc..8ea5cbaaa 100644 --- a/src/yuzu/game_list.h +++ b/src/yuzu/game_list.h @@ -66,6 +66,7 @@ signals: void GameChosen(QString game_path); void ShouldCancelWorker(); void OpenFolderRequested(u64 program_id, GameListOpenTarget target); + void OpenTransferableShaderCacheRequested(u64 program_id); void DumpRomFSRequested(u64 program_id, const std::string& game_path); void CopyTIDRequested(u64 program_id); void NavigateToGamedbEntryRequested(u64 program_id, diff --git a/src/yuzu/main.cpp b/src/yuzu/main.cpp index 41ba3c4c6..2b9db69a3 100644 --- a/src/yuzu/main.cpp +++ b/src/yuzu/main.cpp @@ -37,14 +37,20 @@ static FileSys::VirtualFile VfsDirectoryCreateFileWrapper(const FileSys::Virtual #include <glad/glad.h> #define QT_NO_OPENGL +#include <QClipboard> +#include <QDesktopServices> #include <QDesktopWidget> #include <QDialogButtonBox> #include <QFile> #include <QFileDialog> +#include <QInputDialog> #include <QMessageBox> +#include <QProgressBar> +#include <QProgressDialog> +#include <QShortcut> +#include <QStatusBar> #include <QtConcurrent/QtConcurrent> -#include <QtGui> -#include <QtWidgets> + #include <fmt/format.h> #include "common/common_paths.h" #include "common/detached_tasks.h" @@ -55,11 +61,9 @@ static FileSys::VirtualFile VfsDirectoryCreateFileWrapper(const FileSys::Virtual #include "common/microprofile.h" #include "common/scm_rev.h" #include "common/scope_exit.h" -#include "common/string_util.h" #include "common/telemetry.h" #include "core/core.h" #include "core/crypto/key_manager.h" -#include "core/file_sys/bis_factory.h" #include "core/file_sys/card_image.h" #include "core/file_sys/content_archive.h" #include "core/file_sys/control_metadata.h" @@ -71,7 +75,6 @@ static FileSys::VirtualFile VfsDirectoryCreateFileWrapper(const FileSys::Virtual #include "core/frontend/applets/software_keyboard.h" #include "core/hle/kernel/process.h" #include "core/hle/service/filesystem/filesystem.h" -#include "core/hle/service/filesystem/fsp_ldr.h" #include "core/hle/service/nfp/nfp.h" #include "core/hle/service/sm/sm.h" #include "core/loader/loader.h" @@ -648,6 +651,8 @@ void GMainWindow::RestoreUIState() { void GMainWindow::ConnectWidgetEvents() { connect(game_list, &GameList::GameChosen, this, &GMainWindow::OnGameListLoadFile); connect(game_list, &GameList::OpenFolderRequested, this, &GMainWindow::OnGameListOpenFolder); + connect(game_list, &GameList::OpenTransferableShaderCacheRequested, this, + &GMainWindow::OnTransferableShaderCacheOpenFile); connect(game_list, &GameList::DumpRomFSRequested, this, &GMainWindow::OnGameListDumpRomFS); connect(game_list, &GameList::CopyTIDRequested, this, &GMainWindow::OnGameListCopyTID); connect(game_list, &GameList::NavigateToGamedbEntryRequested, this, @@ -1082,6 +1087,39 @@ void GMainWindow::OnGameListOpenFolder(u64 program_id, GameListOpenTarget target QDesktopServices::openUrl(QUrl::fromLocalFile(qpath)); } +void GMainWindow::OnTransferableShaderCacheOpenFile(u64 program_id) { + ASSERT(program_id != 0); + + const QString tranferable_shader_cache_folder_path = + QString::fromStdString(FileUtil::GetUserPath(FileUtil::UserPath::ShaderDir)) + "opengl" + + DIR_SEP + "transferable"; + + const QString transferable_shader_cache_file_path = + tranferable_shader_cache_folder_path + DIR_SEP + + QString::fromStdString(fmt::format("{:016X}.bin", program_id)); + + if (!QFile::exists(transferable_shader_cache_file_path)) { + QMessageBox::warning(this, tr("Error Opening Transferable Shader Cache"), + tr("A shader cache for this title does not exist.")); + return; + } + + // Windows supports opening a folder with selecting a specified file in explorer. On every other + // OS we just open the transferable shader cache folder without preselecting the transferable + // shader cache file for the selected game. +#if defined(Q_OS_WIN) + const QString explorer = QStringLiteral("explorer"); + QStringList param; + if (!QFileInfo(transferable_shader_cache_file_path).isDir()) { + param << QStringLiteral("/select,"); + } + param << QDir::toNativeSeparators(transferable_shader_cache_file_path); + QProcess::startDetached(explorer, param); +#else + QDesktopServices::openUrl(QUrl::fromLocalFile(tranferable_shader_cache_folder_path)); +#endif +} + static std::size_t CalculateRomFSEntrySize(const FileSys::VirtualDir& dir, bool full) { std::size_t out = 0; diff --git a/src/yuzu/main.h b/src/yuzu/main.h index e07c892cf..7f3aa998e 100644 --- a/src/yuzu/main.h +++ b/src/yuzu/main.h @@ -176,6 +176,7 @@ private slots: /// Called whenever a user selects a game in the game list widget. void OnGameListLoadFile(QString game_path); void OnGameListOpenFolder(u64 program_id, GameListOpenTarget target); + void OnTransferableShaderCacheOpenFile(u64 program_id); void OnGameListDumpRomFS(u64 program_id, const std::string& game_path); void OnGameListCopyTID(u64 program_id); void OnGameListNavigateToGamedbEntry(u64 program_id, |