diff options
Diffstat (limited to '')
26 files changed, 220 insertions, 300 deletions
diff --git a/src/common/CMakeLists.txt b/src/common/CMakeLists.txt index 7534eb8f1..a6fa9a85d 100644 --- a/src/common/CMakeLists.txt +++ b/src/common/CMakeLists.txt @@ -110,7 +110,6 @@ add_library(common STATIC cityhash.cpp cityhash.h common_funcs.h - common_sizes.h common_types.h concepts.h div_ceil.h @@ -134,6 +133,7 @@ add_library(common STATIC host_memory.cpp host_memory.h intrusive_red_black_tree.h + literals.h logging/backend.cpp logging/backend.h logging/filter.cpp diff --git a/src/common/common_sizes.h b/src/common/common_sizes.h deleted file mode 100644 index d07b7ee5a..000000000 --- a/src/common/common_sizes.h +++ /dev/null @@ -1,44 +0,0 @@ -// Copyright 2021 yuzu Emulator Project -// Licensed under GPLv2 or any later version -// Refer to the license.txt file included. - -#pragma once - -#include <limits> - -#include "common/common_types.h" - -namespace Common { - -enum : u64 { - Size_1_KB = 0x400ULL, - Size_64_KB = 64ULL * Size_1_KB, - Size_128_KB = 128ULL * Size_1_KB, - Size_1_MB = 0x100000ULL, - Size_2_MB = 2ULL * Size_1_MB, - Size_4_MB = 4ULL * Size_1_MB, - Size_5_MB = 5ULL * Size_1_MB, - Size_14_MB = 14ULL * Size_1_MB, - Size_32_MB = 32ULL * Size_1_MB, - Size_33_MB = 33ULL * Size_1_MB, - Size_128_MB = 128ULL * Size_1_MB, - Size_448_MB = 448ULL * Size_1_MB, - Size_507_MB = 507ULL * Size_1_MB, - Size_512_MB = 512ULL * Size_1_MB, - Size_562_MB = 562ULL * Size_1_MB, - Size_1554_MB = 1554ULL * Size_1_MB, - Size_2048_MB = 2048ULL * Size_1_MB, - Size_2193_MB = 2193ULL * Size_1_MB, - Size_3285_MB = 3285ULL * Size_1_MB, - Size_4916_MB = 4916ULL * Size_1_MB, - Size_1_GB = 0x40000000ULL, - Size_2_GB = 2ULL * Size_1_GB, - Size_4_GB = 4ULL * Size_1_GB, - Size_6_GB = 6ULL * Size_1_GB, - Size_8_GB = 8ULL * Size_1_GB, - Size_64_GB = 64ULL * Size_1_GB, - Size_512_GB = 512ULL * Size_1_GB, - Size_Invalid = std::numeric_limits<u64>::max(), -}; - -} // namespace Common diff --git a/src/common/literals.h b/src/common/literals.h new file mode 100644 index 000000000..d55fed40b --- /dev/null +++ b/src/common/literals.h @@ -0,0 +1,31 @@ +// Copyright 2021 yuzu emulator team +// Licensed under GPLv2 or any later version +// Refer to the license.txt file included. + +#pragma once + +#include "common/common_types.h" + +namespace Common::Literals { + +constexpr u64 operator""_KiB(unsigned long long int x) { + return 1024ULL * x; +} + +constexpr u64 operator""_MiB(unsigned long long int x) { + return 1024_KiB * x; +} + +constexpr u64 operator""_GiB(unsigned long long int x) { + return 1024_MiB * x; +} + +constexpr u64 operator""_TiB(unsigned long long int x) { + return 1024_GiB * x; +} + +constexpr u64 operator""_PiB(unsigned long long int x) { + return 1024_TiB * x; +} + +} // namespace Common::Literals diff --git a/src/common/logging/backend.cpp b/src/common/logging/backend.cpp index 47ce06478..b6fa4affb 100644 --- a/src/common/logging/backend.cpp +++ b/src/common/logging/backend.cpp @@ -19,6 +19,8 @@ #include "common/assert.h" #include "common/fs/file.h" #include "common/fs/fs.h" +#include "common/literals.h" + #include "common/logging/backend.h" #include "common/logging/log.h" #include "common/logging/text_formatter.h" @@ -98,8 +100,8 @@ private: write_logs(entry); } - // Drain the logging queue. Only writes out up to MAX_LOGS_TO_WRITE to prevent a case - // where a system is repeatedly spamming logs even on close. + // Drain the logging queue. Only writes out up to MAX_LOGS_TO_WRITE to prevent a + // case where a system is repeatedly spamming logs even on close. const int MAX_LOGS_TO_WRITE = filter.IsDebug() ? INT_MAX : 100; int logs_written = 0; while (logs_written++ < MAX_LOGS_TO_WRITE && message_queue.Pop(entry)) { @@ -169,10 +171,11 @@ FileBackend::FileBackend(const std::filesystem::path& filename) { FileBackend::~FileBackend() = default; void FileBackend::Write(const Entry& entry) { + using namespace Common::Literals; // prevent logs from going over the maximum size (in case its spamming and the user doesn't // know) - constexpr std::size_t MAX_BYTES_WRITTEN = 100 * 1024 * 1024; - constexpr std::size_t MAX_BYTES_WRITTEN_EXTENDED = 1024 * 1024 * 1024; + constexpr std::size_t MAX_BYTES_WRITTEN = 100_MiB; + constexpr std::size_t MAX_BYTES_WRITTEN_EXTENDED = 1_GiB; if (!file->IsOpen()) { return; diff --git a/src/core/arm/dynarmic/arm_dynarmic_32.cpp b/src/core/arm/dynarmic/arm_dynarmic_32.cpp index f871f7bf4..77a44f862 100644 --- a/src/core/arm/dynarmic/arm_dynarmic_32.cpp +++ b/src/core/arm/dynarmic/arm_dynarmic_32.cpp @@ -8,6 +8,7 @@ #include <dynarmic/interface/A32/config.h> #include <dynarmic/interface/A32/context.h> #include "common/assert.h" +#include "common/literals.h" #include "common/logging/log.h" #include "common/page_table.h" #include "common/settings.h" @@ -22,6 +23,8 @@ namespace Core { +using namespace Common::Literals; + class DynarmicCallbacks32 : public Dynarmic::A32::UserCallbacks { public: explicit DynarmicCallbacks32(ARM_Dynarmic_32& parent_) @@ -143,8 +146,8 @@ std::shared_ptr<Dynarmic::A32::Jit> ARM_Dynarmic_32::MakeJit(Common::PageTable* config.wall_clock_cntpct = uses_wall_clock; // Code cache size - config.code_cache_size = 512 * 1024 * 1024; - config.far_code_offset = 400 * 1024 * 1024; + config.code_cache_size = 512_MiB; + config.far_code_offset = 400_MiB; // Safe optimizations if (Settings::values.cpu_accuracy.GetValue() == Settings::CPUAccuracy::DebugMode) { diff --git a/src/core/arm/dynarmic/arm_dynarmic_64.cpp b/src/core/arm/dynarmic/arm_dynarmic_64.cpp index ba524cd05..75332e348 100644 --- a/src/core/arm/dynarmic/arm_dynarmic_64.cpp +++ b/src/core/arm/dynarmic/arm_dynarmic_64.cpp @@ -7,6 +7,7 @@ #include <dynarmic/interface/A64/a64.h> #include <dynarmic/interface/A64/config.h> #include "common/assert.h" +#include "common/literals.h" #include "common/logging/log.h" #include "common/page_table.h" #include "common/settings.h" @@ -24,6 +25,7 @@ namespace Core { using Vector = Dynarmic::A64::Vector; +using namespace Common::Literals; class DynarmicCallbacks64 : public Dynarmic::A64::UserCallbacks { public: @@ -184,8 +186,8 @@ std::shared_ptr<Dynarmic::A64::Jit> ARM_Dynarmic_64::MakeJit(Common::PageTable* config.wall_clock_cntpct = uses_wall_clock; // Code cache size - config.code_cache_size = 512 * 1024 * 1024; - config.far_code_offset = 400 * 1024 * 1024; + config.code_cache_size = 512_MiB; + config.far_code_offset = 400_MiB; // Safe optimizations if (Settings::values.cpu_accuracy.GetValue() == Settings::CPUAccuracy::DebugMode) { diff --git a/src/core/hle/kernel/board/nintendo/nx/k_system_control.cpp b/src/core/hle/kernel/board/nintendo/nx/k_system_control.cpp index 86472b5ce..6f335c251 100644 --- a/src/core/hle/kernel/board/nintendo/nx/k_system_control.cpp +++ b/src/core/hle/kernel/board/nintendo/nx/k_system_control.cpp @@ -4,7 +4,8 @@ #include <random> -#include "common/common_sizes.h" +#include "common/literals.h" + #include "core/hle/kernel/board/nintendo/nx/k_system_control.h" #include "core/hle/kernel/board/nintendo/nx/secure_monitor.h" #include "core/hle/kernel/k_trace.h" @@ -25,6 +26,8 @@ constexpr const std::size_t RequiredNonSecureSystemMemorySize = namespace { +using namespace Common::Literals; + u32 GetMemoryModeForInit() { return 0x01; } @@ -57,11 +60,11 @@ size_t KSystemControl::Init::GetIntendedMemorySize() { switch (GetMemorySizeForInit()) { case Smc::MemorySize_4GB: default: // All invalid modes should go to 4GB. - return Common::Size_4_GB; + return 4_GiB; case Smc::MemorySize_6GB: - return Common::Size_6_GB; + return 6_GiB; case Smc::MemorySize_8GB: - return Common::Size_8_GB; + return 8_GiB; } } @@ -79,17 +82,17 @@ std::size_t KSystemControl::Init::GetApplicationPoolSize() { switch (GetMemoryArrangeForInit()) { case Smc::MemoryArrangement_4GB: default: - return Common::Size_3285_MB; + return 3285_MiB; case Smc::MemoryArrangement_4GBForAppletDev: - return Common::Size_2048_MB; + return 2048_MiB; case Smc::MemoryArrangement_4GBForSystemDev: - return Common::Size_3285_MB; + return 3285_MiB; case Smc::MemoryArrangement_6GB: - return Common::Size_4916_MB; + return 4916_MiB; case Smc::MemoryArrangement_6GBForAppletDev: - return Common::Size_3285_MB; + return 3285_MiB; case Smc::MemoryArrangement_8GB: - return Common::Size_4916_MB; + return 4916_MiB; } }(); @@ -103,22 +106,22 @@ size_t KSystemControl::Init::GetAppletPoolSize() { switch (GetMemoryArrangeForInit()) { case Smc::MemoryArrangement_4GB: default: - return Common::Size_507_MB; + return 507_MiB; case Smc::MemoryArrangement_4GBForAppletDev: - return Common::Size_1554_MB; + return 1554_MiB; case Smc::MemoryArrangement_4GBForSystemDev: - return Common::Size_448_MB; + return 448_MiB; case Smc::MemoryArrangement_6GB: - return Common::Size_562_MB; + return 562_MiB; case Smc::MemoryArrangement_6GBForAppletDev: - return Common::Size_2193_MB; + return 2193_MiB; case Smc::MemoryArrangement_8GB: - return Common::Size_2193_MB; + return 2193_MiB; } }(); // Return (possibly) adjusted size. - constexpr size_t ExtraSystemMemoryForAtmosphere = Common::Size_33_MB; + constexpr size_t ExtraSystemMemoryForAtmosphere = 33_MiB; return base_pool_size - ExtraSystemMemoryForAtmosphere - KTraceBufferSize; } diff --git a/src/core/hle/kernel/k_address_space_info.cpp b/src/core/hle/kernel/k_address_space_info.cpp index c7549f7a2..ca29edc88 100644 --- a/src/core/hle/kernel/k_address_space_info.cpp +++ b/src/core/hle/kernel/k_address_space_info.cpp @@ -5,34 +5,37 @@ #include <array> #include "common/assert.h" -#include "common/common_sizes.h" +#include "common/literals.h" #include "core/hle/kernel/k_address_space_info.h" namespace Kernel { namespace { +using namespace Common::Literals; + +constexpr u64 Size_Invalid = UINT64_MAX; + // clang-format off constexpr std::array<KAddressSpaceInfo, 13> AddressSpaceInfos{{ - { .bit_width = 32, .address = Common::Size_2_MB , .size = Common::Size_1_GB - Common::Size_2_MB , .type = KAddressSpaceInfo::Type::MapSmall, }, - { .bit_width = 32, .address = Common::Size_1_GB , .size = Common::Size_4_GB - Common::Size_1_GB , .type = KAddressSpaceInfo::Type::MapLarge, }, - { .bit_width = 32, .address = Common::Size_Invalid, .size = Common::Size_1_GB , .type = KAddressSpaceInfo::Type::Alias, }, - { .bit_width = 32, .address = Common::Size_Invalid, .size = Common::Size_1_GB , .type = KAddressSpaceInfo::Type::Heap, }, - { .bit_width = 36, .address = Common::Size_128_MB , .size = Common::Size_2_GB - Common::Size_128_MB, .type = KAddressSpaceInfo::Type::MapSmall, }, - { .bit_width = 36, .address = Common::Size_2_GB , .size = Common::Size_64_GB - Common::Size_2_GB , .type = KAddressSpaceInfo::Type::MapLarge, }, - { .bit_width = 36, .address = Common::Size_Invalid, .size = Common::Size_6_GB , .type = KAddressSpaceInfo::Type::Heap, }, - { .bit_width = 36, .address = Common::Size_Invalid, .size = Common::Size_6_GB , .type = KAddressSpaceInfo::Type::Alias, }, - { .bit_width = 39, .address = Common::Size_128_MB , .size = Common::Size_512_GB - Common::Size_128_MB, .type = KAddressSpaceInfo::Type::Map39Bit, }, - { .bit_width = 39, .address = Common::Size_Invalid, .size = Common::Size_64_GB , .type = KAddressSpaceInfo::Type::MapSmall }, - { .bit_width = 39, .address = Common::Size_Invalid, .size = Common::Size_6_GB , .type = KAddressSpaceInfo::Type::Heap, }, - { .bit_width = 39, .address = Common::Size_Invalid, .size = Common::Size_64_GB , .type = KAddressSpaceInfo::Type::Alias, }, - { .bit_width = 39, .address = Common::Size_Invalid, .size = Common::Size_2_GB , .type = KAddressSpaceInfo::Type::Stack, }, + { .bit_width = 32, .address = 2_MiB , .size = 1_GiB - 2_MiB , .type = KAddressSpaceInfo::Type::MapSmall, }, + { .bit_width = 32, .address = 1_GiB , .size = 4_GiB - 1_GiB , .type = KAddressSpaceInfo::Type::MapLarge, }, + { .bit_width = 32, .address = Size_Invalid, .size = 1_GiB , .type = KAddressSpaceInfo::Type::Alias, }, + { .bit_width = 32, .address = Size_Invalid, .size = 1_GiB , .type = KAddressSpaceInfo::Type::Heap, }, + { .bit_width = 36, .address = 128_MiB , .size = 2_GiB - 128_MiB, .type = KAddressSpaceInfo::Type::MapSmall, }, + { .bit_width = 36, .address = 2_GiB , .size = 64_GiB - 2_GiB , .type = KAddressSpaceInfo::Type::MapLarge, }, + { .bit_width = 36, .address = Size_Invalid, .size = 6_GiB , .type = KAddressSpaceInfo::Type::Heap, }, + { .bit_width = 36, .address = Size_Invalid, .size = 6_GiB , .type = KAddressSpaceInfo::Type::Alias, }, + { .bit_width = 39, .address = 128_MiB , .size = 512_GiB - 128_MiB, .type = KAddressSpaceInfo::Type::Map39Bit, }, + { .bit_width = 39, .address = Size_Invalid, .size = 64_GiB , .type = KAddressSpaceInfo::Type::MapSmall }, + { .bit_width = 39, .address = Size_Invalid, .size = 6_GiB , .type = KAddressSpaceInfo::Type::Heap, }, + { .bit_width = 39, .address = Size_Invalid, .size = 64_GiB , .type = KAddressSpaceInfo::Type::Alias, }, + { .bit_width = 39, .address = Size_Invalid, .size = 2_GiB , .type = KAddressSpaceInfo::Type::Stack, }, }}; // clang-format on constexpr bool IsAllowedIndexForAddress(std::size_t index) { - return index < AddressSpaceInfos.size() && - AddressSpaceInfos[index].address != Common::Size_Invalid; + return index < AddressSpaceInfos.size() && AddressSpaceInfos[index].address != Size_Invalid; } using IndexArray = diff --git a/src/core/hle/kernel/k_memory_layout.board.nintendo_nx.cpp b/src/core/hle/kernel/k_memory_layout.board.nintendo_nx.cpp index a78551291..af652af58 100644 --- a/src/core/hle/kernel/k_memory_layout.board.nintendo_nx.cpp +++ b/src/core/hle/kernel/k_memory_layout.board.nintendo_nx.cpp @@ -3,6 +3,7 @@ // Refer to the license.txt file included. #include "common/alignment.h" +#include "common/literals.h" #include "core/hle/kernel/k_memory_layout.h" #include "core/hle/kernel/k_memory_manager.h" #include "core/hle/kernel/k_system_control.h" @@ -12,8 +13,10 @@ namespace Kernel { namespace { +using namespace Common::Literals; + constexpr size_t CarveoutAlignment = 0x20000; -constexpr size_t CarveoutSizeMax = (512ULL * 1024 * 1024) - CarveoutAlignment; +constexpr size_t CarveoutSizeMax = (512_MiB) - CarveoutAlignment; bool SetupPowerManagementControllerMemoryRegion(KMemoryLayout& memory_layout) { // Above firmware 2.0.0, the PMC is not mappable. diff --git a/src/core/hle/kernel/k_memory_layout.h b/src/core/hle/kernel/k_memory_layout.h index 288642d9a..57ff538cc 100644 --- a/src/core/hle/kernel/k_memory_layout.h +++ b/src/core/hle/kernel/k_memory_layout.h @@ -7,8 +7,7 @@ #include <utility> #include "common/alignment.h" -#include "common/common_sizes.h" -#include "common/common_types.h" +#include "common/literals.h" #include "core/device_memory.h" #include "core/hle/kernel/k_memory_region.h" #include "core/hle/kernel/k_memory_region_type.h" @@ -16,20 +15,22 @@ namespace Kernel { -constexpr std::size_t L1BlockSize = Common::Size_1_GB; -constexpr std::size_t L2BlockSize = Common::Size_2_MB; +using namespace Common::Literals; + +constexpr std::size_t L1BlockSize = 1_GiB; +constexpr std::size_t L2BlockSize = 2_MiB; constexpr std::size_t GetMaximumOverheadSize(std::size_t size) { return (Common::DivideUp(size, L1BlockSize) + Common::DivideUp(size, L2BlockSize)) * PageSize; } -constexpr std::size_t MainMemorySize = Common::Size_4_GB; -constexpr std::size_t MainMemorySizeMax = Common::Size_8_GB; +constexpr std::size_t MainMemorySize = 4_GiB; +constexpr std::size_t MainMemorySizeMax = 8_GiB; -constexpr std::size_t ReservedEarlyDramSize = 0x60000; +constexpr std::size_t ReservedEarlyDramSize = 384_KiB; constexpr std::size_t DramPhysicalAddress = 0x80000000; -constexpr std::size_t KernelAslrAlignment = Common::Size_2_MB; +constexpr std::size_t KernelAslrAlignment = 2_MiB; constexpr std::size_t KernelVirtualAddressSpaceWidth = 1ULL << 39; constexpr std::size_t KernelPhysicalAddressSpaceWidth = 1ULL << 48; @@ -40,7 +41,7 @@ constexpr std::size_t KernelVirtualAddressSpaceLast = KernelVirtualAddressSpaceE constexpr std::size_t KernelVirtualAddressSpaceSize = KernelVirtualAddressSpaceEnd - KernelVirtualAddressSpaceBase; constexpr std::size_t KernelVirtualAddressCodeBase = KernelVirtualAddressSpaceBase; -constexpr std::size_t KernelVirtualAddressCodeSize = 0x62000; +constexpr std::size_t KernelVirtualAddressCodeSize = 392_KiB; constexpr std::size_t KernelVirtualAddressCodeEnd = KernelVirtualAddressCodeBase + KernelVirtualAddressCodeSize; @@ -53,14 +54,14 @@ constexpr std::size_t KernelPhysicalAddressSpaceSize = constexpr std::size_t KernelPhysicalAddressCodeBase = DramPhysicalAddress + ReservedEarlyDramSize; constexpr std::size_t KernelPageTableHeapSize = GetMaximumOverheadSize(MainMemorySizeMax); -constexpr std::size_t KernelInitialPageHeapSize = Common::Size_128_KB; +constexpr std::size_t KernelInitialPageHeapSize = 128_KiB; -constexpr std::size_t KernelSlabHeapDataSize = Common::Size_5_MB; -constexpr std::size_t KernelSlabHeapGapsSize = Common::Size_2_MB - Common::Size_64_KB; +constexpr std::size_t KernelSlabHeapDataSize = 5_MiB; +constexpr std::size_t KernelSlabHeapGapsSize = 2_MiB - 64_KiB; constexpr std::size_t KernelSlabHeapSize = KernelSlabHeapDataSize + KernelSlabHeapGapsSize; // NOTE: This is calculated from KThread slab counts, assuming KThread size <= 0x860. -constexpr std::size_t KernelSlabHeapAdditionalSize = 0x68000ULL; +constexpr std::size_t KernelSlabHeapAdditionalSize = 416_KiB; constexpr std::size_t KernelResourceSize = KernelPageTableHeapSize + KernelInitialPageHeapSize + KernelSlabHeapSize; diff --git a/src/core/hle/kernel/k_page_table.cpp b/src/core/hle/kernel/k_page_table.cpp index 66d260635..701268545 100644 --- a/src/core/hle/kernel/k_page_table.cpp +++ b/src/core/hle/kernel/k_page_table.cpp @@ -4,6 +4,7 @@ #include "common/alignment.h" #include "common/assert.h" +#include "common/literals.h" #include "common/scope_exit.h" #include "core/core.h" #include "core/hle/kernel/k_address_space_info.h" @@ -23,6 +24,8 @@ namespace Kernel { namespace { +using namespace Common::Literals; + constexpr std::size_t GetAddressSpaceWidthFromType(FileSys::ProgramAddressSpaceType as_type) { switch (as_type) { case FileSys::ProgramAddressSpaceType::Is32Bit: @@ -89,7 +92,7 @@ ResultCode KPageTable::InitializeForProcess(FileSys::ProgramAddressSpaceType as_ } // Set code regions and determine remaining - constexpr std::size_t RegionAlignment{2 * 1024 * 1024}; + constexpr std::size_t RegionAlignment{2_MiB}; VAddr process_code_start{}; VAddr process_code_end{}; std::size_t stack_region_size{}; diff --git a/src/core/hle/kernel/k_trace.h b/src/core/hle/kernel/k_trace.h index 91ebf9ab2..79391bccb 100644 --- a/src/core/hle/kernel/k_trace.h +++ b/src/core/hle/kernel/k_trace.h @@ -4,9 +4,13 @@ #pragma once +#include "common/common_funcs.h" + namespace Kernel { +using namespace Common::Literals; + constexpr bool IsKTraceEnabled = false; -constexpr std::size_t KTraceBufferSize = IsKTraceEnabled ? 16 * 1024 * 1024 : 0; +constexpr std::size_t KTraceBufferSize = IsKTraceEnabled ? 16_MiB : 0; } // namespace Kernel diff --git a/src/core/hle/kernel/kernel.cpp b/src/core/hle/kernel/kernel.cpp index 2ceeaeb5f..64bd0c494 100644 --- a/src/core/hle/kernel/kernel.cpp +++ b/src/core/hle/kernel/kernel.cpp @@ -12,7 +12,6 @@ #include <utility> #include "common/assert.h" -#include "common/common_sizes.h" #include "common/logging/log.h" #include "common/microprofile.h" #include "common/thread.h" @@ -180,7 +179,7 @@ struct KernelCore::Impl { system_resource_limit->Reserve(LimitableResource::PhysicalMemory, kernel_size); // Reserve secure applet memory, introduced in firmware 5.0.0 - constexpr u64 secure_applet_memory_size{Common::Size_4_MB}; + constexpr u64 secure_applet_memory_size{4_MiB}; ASSERT(system_resource_limit->Reserve(LimitableResource::PhysicalMemory, secure_applet_memory_size)); @@ -320,8 +319,8 @@ struct KernelCore::Impl { const VAddr code_end_virt_addr = KernelVirtualAddressCodeEnd; // Setup the containing kernel region. - constexpr size_t KernelRegionSize = Common::Size_1_GB; - constexpr size_t KernelRegionAlign = Common::Size_1_GB; + constexpr size_t KernelRegionSize = 1_GiB; + constexpr size_t KernelRegionAlign = 1_GiB; constexpr VAddr kernel_region_start = Common::AlignDown(code_start_virt_addr, KernelRegionAlign); size_t kernel_region_size = KernelRegionSize; @@ -368,7 +367,7 @@ struct KernelCore::Impl { // Decide on the actual size for the misc region. constexpr size_t MiscRegionAlign = KernelAslrAlignment; - constexpr size_t MiscRegionMinimumSize = Common::Size_32_MB; + constexpr size_t MiscRegionMinimumSize = 32_MiB; const size_t misc_region_size = Common::AlignUp( std::max(misc_region_needed_size, MiscRegionMinimumSize), MiscRegionAlign); ASSERT(misc_region_size > 0); @@ -381,7 +380,7 @@ struct KernelCore::Impl { misc_region_start, misc_region_size, KMemoryRegionType_KernelMisc)); // Setup the stack region. - constexpr size_t StackRegionSize = Common::Size_14_MB; + constexpr size_t StackRegionSize = 14_MiB; constexpr size_t StackRegionAlign = KernelAslrAlignment; const VAddr stack_region_start = memory_layout.GetVirtualMemoryRegionTree().GetRandomAlignedRegion( @@ -414,7 +413,7 @@ struct KernelCore::Impl { slab_region_start, slab_region_size, KMemoryRegionType_KernelSlab)); // Setup the temp region. - constexpr size_t TempRegionSize = Common::Size_128_MB; + constexpr size_t TempRegionSize = 128_MiB; constexpr size_t TempRegionAlign = KernelAslrAlignment; const VAddr temp_region_start = memory_layout.GetVirtualMemoryRegionTree().GetRandomAlignedRegion( @@ -470,7 +469,7 @@ struct KernelCore::Impl { // Determine size available for kernel page table heaps, requiring > 8 MB. const PAddr resource_end_phys_addr = slab_start_phys_addr + resource_region_size; const size_t page_table_heap_size = resource_end_phys_addr - slab_end_phys_addr; - ASSERT(page_table_heap_size / Common::Size_4_MB > 2); + ASSERT(page_table_heap_size / 4_MiB > 2); // Insert a physical region for the kernel page table heap region ASSERT(memory_layout.GetPhysicalMemoryRegionTree().Insert( @@ -495,7 +494,7 @@ struct KernelCore::Impl { ASSERT(linear_extents.GetEndAddress() != 0); // Setup the linear mapping region. - constexpr size_t LinearRegionAlign = Common::Size_1_GB; + constexpr size_t LinearRegionAlign = 1_GiB; const PAddr aligned_linear_phys_start = Common::AlignDown(linear_extents.GetAddress(), LinearRegionAlign); const size_t linear_region_size = diff --git a/src/tests/common/host_memory.cpp b/src/tests/common/host_memory.cpp index e241f8be5..2dc7b5d5e 100644 --- a/src/tests/common/host_memory.cpp +++ b/src/tests/common/host_memory.cpp @@ -5,11 +5,13 @@ #include <catch2/catch.hpp> #include "common/host_memory.h" +#include "common/literals.h" using Common::HostMemory; +using namespace Common::Literals; static constexpr size_t VIRTUAL_SIZE = 1ULL << 39; -static constexpr size_t BACKING_SIZE = 4ULL * 1024 * 1024 * 1024; +static constexpr size_t BACKING_SIZE = 4_GiB; TEST_CASE("HostMemory: Initialize and deinitialize", "[common]") { { HostMemory mem(BACKING_SIZE, VIRTUAL_SIZE); } diff --git a/src/video_core/buffer_cache/buffer_cache.h b/src/video_core/buffer_cache/buffer_cache.h index 6d04d00da..9d726a6fb 100644 --- a/src/video_core/buffer_cache/buffer_cache.h +++ b/src/video_core/buffer_cache/buffer_cache.h @@ -16,9 +16,9 @@ #include <boost/container/small_vector.hpp> -#include "common/common_sizes.h" #include "common/common_types.h" #include "common/div_ceil.h" +#include "common/literals.h" #include "common/microprofile.h" #include "common/scope_exit.h" #include "common/settings.h" @@ -48,8 +48,11 @@ constexpr u32 NUM_COMPUTE_UNIFORM_BUFFERS = 8; constexpr u32 NUM_STORAGE_BUFFERS = 16; constexpr u32 NUM_STAGES = 5; +using namespace Common::Literals; + template <typename P> class BufferCache { + // Page size for caching purposes. // This is unrelated to the CPU page size and it can be changed as it seems optimal. static constexpr u32 PAGE_BITS = 16; @@ -66,8 +69,8 @@ class BufferCache { static constexpr BufferId NULL_BUFFER_ID{0}; - static constexpr u64 EXPECTED_MEMORY = Common::Size_512_MB; - static constexpr u64 CRITICAL_MEMORY = Common::Size_1_GB; + static constexpr u64 EXPECTED_MEMORY = 512_MiB; + static constexpr u64 CRITICAL_MEMORY = 1_GiB; using Maxwell = Tegra::Engines::Maxwell3D::Regs; @@ -96,7 +99,7 @@ class BufferCache { }; public: - static constexpr u32 DEFAULT_SKIP_CACHE_SIZE = 4096; + static constexpr u32 DEFAULT_SKIP_CACHE_SIZE = 4_KiB; explicit BufferCache(VideoCore::RasterizerInterface& rasterizer_, Tegra::Engines::Maxwell3D& maxwell3d_, diff --git a/src/video_core/host_shaders/astc_decoder.comp b/src/video_core/host_shaders/astc_decoder.comp index eaba1b103..c37f15bfd 100644 --- a/src/video_core/host_shaders/astc_decoder.comp +++ b/src/video_core/host_shaders/astc_decoder.comp @@ -11,12 +11,8 @@ #define UNIFORM(n) #define BINDING_INPUT_BUFFER 0 #define BINDING_ENC_BUFFER 1 -#define BINDING_6_TO_8_BUFFER 2 -#define BINDING_7_TO_8_BUFFER 3 -#define BINDING_8_TO_8_BUFFER 4 -#define BINDING_BYTE_TO_16_BUFFER 5 -#define BINDING_SWIZZLE_BUFFER 6 -#define BINDING_OUTPUT_IMAGE 7 +#define BINDING_SWIZZLE_BUFFER 2 +#define BINDING_OUTPUT_IMAGE 3 #else // ^^^ Vulkan ^^^ // vvv OpenGL vvv @@ -26,10 +22,6 @@ #define BINDING_SWIZZLE_BUFFER 0 #define BINDING_INPUT_BUFFER 1 #define BINDING_ENC_BUFFER 2 -#define BINDING_6_TO_8_BUFFER 3 -#define BINDING_7_TO_8_BUFFER 4 -#define BINDING_8_TO_8_BUFFER 5 -#define BINDING_BYTE_TO_16_BUFFER 6 #define BINDING_OUTPUT_IMAGE 0 #endif @@ -76,19 +68,6 @@ layout(binding = BINDING_INPUT_BUFFER, std430) readonly buffer InputBufferU32 { layout(binding = BINDING_ENC_BUFFER, std430) readonly buffer EncodingsValues { EncodingData encoding_values[]; }; -// ASTC Precompiled tables -layout(binding = BINDING_6_TO_8_BUFFER, std430) readonly buffer REPLICATE_6_BIT_TO_8 { - uint REPLICATE_6_BIT_TO_8_TABLE[]; -}; -layout(binding = BINDING_7_TO_8_BUFFER, std430) readonly buffer REPLICATE_7_BIT_TO_8 { - uint REPLICATE_7_BIT_TO_8_TABLE[]; -}; -layout(binding = BINDING_8_TO_8_BUFFER, std430) readonly buffer REPLICATE_8_BIT_TO_8 { - uint REPLICATE_8_BIT_TO_8_TABLE[]; -}; -layout(binding = BINDING_BYTE_TO_16_BUFFER, std430) readonly buffer REPLICATE_BYTE_TO_16 { - uint REPLICATE_BYTE_TO_16_TABLE[]; -}; layout(binding = BINDING_OUTPUT_IMAGE, rgba8) uniform writeonly image2DArray dest_image; @@ -139,6 +118,19 @@ const uint REPLICATE_4_BIT_TO_6_TABLE[16] = const uint REPLICATE_5_BIT_TO_6_TABLE[32] = uint[](0, 2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 22, 24, 26, 28, 30, 33, 35, 37, 39, 41, 43, 45, 47, 49, 51, 53, 55, 57, 59, 61, 63); +const uint REPLICATE_6_BIT_TO_8_TABLE[64] = + uint[](0, 4, 8, 12, 16, 20, 24, 28, 32, 36, 40, 44, 48, 52, 56, 60, 65, 69, 73, 77, 81, 85, 89, + 93, 97, 101, 105, 109, 113, 117, 121, 125, 130, 134, 138, 142, 146, 150, 154, 158, 162, + 166, 170, 174, 178, 182, 186, 190, 195, 199, 203, 207, 211, 215, 219, 223, 227, 231, 235, + 239, 243, 247, 251, 255); +const uint REPLICATE_7_BIT_TO_8_TABLE[128] = + uint[](0, 2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 22, 24, 26, 28, 30, 32, 34, 36, 38, 40, 42, 44, + 46, 48, 50, 52, 54, 56, 58, 60, 62, 64, 66, 68, 70, 72, 74, 76, 78, 80, 82, 84, 86, 88, + 90, 92, 94, 96, 98, 100, 102, 104, 106, 108, 110, 112, 114, 116, 118, 120, 122, 124, 126, + 129, 131, 133, 135, 137, 139, 141, 143, 145, 147, 149, 151, 153, 155, 157, 159, 161, 163, + 165, 167, 169, 171, 173, 175, 177, 179, 181, 183, 185, 187, 189, 191, 193, 195, 197, 199, + 201, 203, 205, 207, 209, 211, 213, 215, 217, 219, 221, 223, 225, 227, 229, 231, 233, 235, + 237, 239, 241, 243, 245, 247, 249, 251, 253, 255); // Input ASTC texture globals uint current_index = 0; @@ -207,8 +199,7 @@ uint Replicate(uint val, uint num_bits, uint to_bit) { } uvec4 ReplicateByteTo16(uvec4 value) { - return uvec4(REPLICATE_BYTE_TO_16_TABLE[value.x], REPLICATE_BYTE_TO_16_TABLE[value.y], - REPLICATE_BYTE_TO_16_TABLE[value.z], REPLICATE_BYTE_TO_16_TABLE[value.w]); + return value * 0x101; } uint ReplicateBitTo7(uint value) { @@ -236,7 +227,7 @@ uint FastReplicateTo8(uint value, uint num_bits) { case 7: return REPLICATE_7_BIT_TO_8_TABLE[value]; case 8: - return REPLICATE_8_BIT_TO_8_TABLE[value]; + return value; } return Replicate(value, num_bits, 8); } @@ -1327,6 +1318,9 @@ void main() { offset += swizzle; const ivec3 coord = ivec3(gl_GlobalInvocationID * uvec3(block_dims, 1)); + if (any(greaterThanEqual(coord, imageSize(dest_image)))) { + return; + } uint block_index = pos.z * gl_WorkGroupSize.x * gl_WorkGroupSize.y + pos.y * gl_WorkGroupSize.x + pos.x; diff --git a/src/video_core/renderer_opengl/gl_stream_buffer.h b/src/video_core/renderer_opengl/gl_stream_buffer.h index 6dbb6bfba..2e67922a6 100644 --- a/src/video_core/renderer_opengl/gl_stream_buffer.h +++ b/src/video_core/renderer_opengl/gl_stream_buffer.h @@ -12,12 +12,15 @@ #include <glad/glad.h> #include "common/common_types.h" +#include "common/literals.h" #include "video_core/renderer_opengl/gl_resource_manager.h" namespace OpenGL { +using namespace Common::Literals; + class StreamBuffer { - static constexpr size_t STREAM_BUFFER_SIZE = 64 * 1024 * 1024; + static constexpr size_t STREAM_BUFFER_SIZE = 64_MiB; static constexpr size_t NUM_SYNCS = 16; static constexpr size_t REGION_SIZE = STREAM_BUFFER_SIZE / NUM_SYNCS; static constexpr size_t MAX_ALIGNMENT = 256; diff --git a/src/video_core/renderer_opengl/util_shaders.cpp b/src/video_core/renderer_opengl/util_shaders.cpp index 47fddcb6e..abaf1ee6a 100644 --- a/src/video_core/renderer_opengl/util_shaders.cpp +++ b/src/video_core/renderer_opengl/util_shaders.cpp @@ -69,7 +69,8 @@ UtilShaders::UtilShaders(ProgramManager& program_manager_) swizzle_table_buffer.Create(); astc_buffer.Create(); glNamedBufferStorage(swizzle_table_buffer.handle, sizeof(swizzle_table), &swizzle_table, 0); - glNamedBufferStorage(astc_buffer.handle, sizeof(ASTC_BUFFER_DATA), &ASTC_BUFFER_DATA, 0); + glNamedBufferStorage(astc_buffer.handle, sizeof(ASTC_ENCODINGS_VALUES), &ASTC_ENCODINGS_VALUES, + 0); } UtilShaders::~UtilShaders() = default; @@ -79,12 +80,6 @@ void UtilShaders::ASTCDecode(Image& image, const ImageBufferMap& map, static constexpr GLuint BINDING_SWIZZLE_BUFFER = 0; static constexpr GLuint BINDING_INPUT_BUFFER = 1; static constexpr GLuint BINDING_ENC_BUFFER = 2; - - static constexpr GLuint BINDING_6_TO_8_BUFFER = 3; - static constexpr GLuint BINDING_7_TO_8_BUFFER = 4; - static constexpr GLuint BINDING_8_TO_8_BUFFER = 5; - static constexpr GLuint BINDING_BYTE_TO_16_BUFFER = 6; - static constexpr GLuint BINDING_OUTPUT_IMAGE = 0; const Extent2D tile_size{ @@ -93,21 +88,7 @@ void UtilShaders::ASTCDecode(Image& image, const ImageBufferMap& map, }; program_manager.BindHostCompute(astc_decoder_program.handle); glBindBufferBase(GL_SHADER_STORAGE_BUFFER, BINDING_SWIZZLE_BUFFER, swizzle_table_buffer.handle); - glBindBufferRange(GL_SHADER_STORAGE_BUFFER, BINDING_ENC_BUFFER, astc_buffer.handle, - offsetof(AstcBufferData, encoding_values), - sizeof(AstcBufferData::encoding_values)); - glBindBufferRange(GL_SHADER_STORAGE_BUFFER, BINDING_6_TO_8_BUFFER, astc_buffer.handle, - offsetof(AstcBufferData, replicate_6_to_8), - sizeof(AstcBufferData::replicate_6_to_8)); - glBindBufferRange(GL_SHADER_STORAGE_BUFFER, BINDING_7_TO_8_BUFFER, astc_buffer.handle, - offsetof(AstcBufferData, replicate_7_to_8), - sizeof(AstcBufferData::replicate_7_to_8)); - glBindBufferRange(GL_SHADER_STORAGE_BUFFER, BINDING_8_TO_8_BUFFER, astc_buffer.handle, - offsetof(AstcBufferData, replicate_8_to_8), - sizeof(AstcBufferData::replicate_8_to_8)); - glBindBufferRange(GL_SHADER_STORAGE_BUFFER, BINDING_BYTE_TO_16_BUFFER, astc_buffer.handle, - offsetof(AstcBufferData, replicate_byte_to_16), - sizeof(AstcBufferData::replicate_byte_to_16)); + glBindBufferBase(GL_SHADER_STORAGE_BUFFER, BINDING_ENC_BUFFER, astc_buffer.handle); glFlushMappedNamedBufferRange(map.buffer, map.offset, image.guest_size_bytes); glUniform2ui(1, tile_size.width, tile_size.height); @@ -137,6 +118,12 @@ void UtilShaders::ASTCDecode(Image& image, const ImageBufferMap& map, glDispatchCompute(num_dispatches_x, num_dispatches_y, image.info.resources.layers); } + // Precautionary barrier to ensure the compute shader is done decoding prior to texture access. + // GL_TEXTURE_FETCH_BARRIER_BIT and GL_SHADER_IMAGE_ACCESS_BARRIER_BIT are used in a separate + // glMemoryBarrier call by the texture cache runtime + glMemoryBarrier(GL_UNIFORM_BARRIER_BIT | GL_COMMAND_BARRIER_BIT | GL_PIXEL_BUFFER_BARRIER_BIT | + GL_TEXTURE_UPDATE_BARRIER_BIT | GL_BUFFER_UPDATE_BARRIER_BIT | + GL_SHADER_STORAGE_BARRIER_BIT | GL_CLIENT_MAPPED_BUFFER_BARRIER_BIT); program_manager.RestoreGuestCompute(); } diff --git a/src/video_core/renderer_vulkan/vk_compute_pass.cpp b/src/video_core/renderer_vulkan/vk_compute_pass.cpp index e11406e58..205cd3b05 100644 --- a/src/video_core/renderer_vulkan/vk_compute_pass.cpp +++ b/src/video_core/renderer_vulkan/vk_compute_pass.cpp @@ -30,19 +30,16 @@ namespace Vulkan { using Tegra::Texture::SWIZZLE_TABLE; -using Tegra::Texture::ASTC::EncodingsValues; +using Tegra::Texture::ASTC::ASTC_ENCODINGS_VALUES; using namespace Tegra::Texture::ASTC; namespace { constexpr u32 ASTC_BINDING_INPUT_BUFFER = 0; constexpr u32 ASTC_BINDING_ENC_BUFFER = 1; -constexpr u32 ASTC_BINDING_6_TO_8_BUFFER = 2; -constexpr u32 ASTC_BINDING_7_TO_8_BUFFER = 3; -constexpr u32 ASTC_BINDING_8_TO_8_BUFFER = 4; -constexpr u32 ASTC_BINDING_BYTE_TO_16_BUFFER = 5; -constexpr u32 ASTC_BINDING_SWIZZLE_BUFFER = 6; -constexpr u32 ASTC_BINDING_OUTPUT_IMAGE = 7; +constexpr u32 ASTC_BINDING_SWIZZLE_BUFFER = 2; +constexpr u32 ASTC_BINDING_OUTPUT_IMAGE = 3; +constexpr size_t ASTC_NUM_BINDINGS = 4; VkPushConstantRange BuildComputePushConstantRange(std::size_t size) { return { @@ -71,7 +68,7 @@ std::array<VkDescriptorSetLayoutBinding, 2> BuildInputOutputDescriptorSetBinding }}; } -std::array<VkDescriptorSetLayoutBinding, 8> BuildASTCDescriptorSetBindings() { +std::array<VkDescriptorSetLayoutBinding, ASTC_NUM_BINDINGS> BuildASTCDescriptorSetBindings() { return {{ { .binding = ASTC_BINDING_INPUT_BUFFER, @@ -88,34 +85,6 @@ std::array<VkDescriptorSetLayoutBinding, 8> BuildASTCDescriptorSetBindings() { .pImmutableSamplers = nullptr, }, { - .binding = ASTC_BINDING_6_TO_8_BUFFER, - .descriptorType = VK_DESCRIPTOR_TYPE_STORAGE_BUFFER, - .descriptorCount = 1, - .stageFlags = VK_SHADER_STAGE_COMPUTE_BIT, - .pImmutableSamplers = nullptr, - }, - { - .binding = ASTC_BINDING_7_TO_8_BUFFER, - .descriptorType = VK_DESCRIPTOR_TYPE_STORAGE_BUFFER, - .descriptorCount = 1, - .stageFlags = VK_SHADER_STAGE_COMPUTE_BIT, - .pImmutableSamplers = nullptr, - }, - { - .binding = ASTC_BINDING_8_TO_8_BUFFER, - .descriptorType = VK_DESCRIPTOR_TYPE_STORAGE_BUFFER, - .descriptorCount = 1, - .stageFlags = VK_SHADER_STAGE_COMPUTE_BIT, - .pImmutableSamplers = nullptr, - }, - { - .binding = ASTC_BINDING_BYTE_TO_16_BUFFER, - .descriptorType = VK_DESCRIPTOR_TYPE_STORAGE_BUFFER, - .descriptorCount = 1, - .stageFlags = VK_SHADER_STAGE_COMPUTE_BIT, - .pImmutableSamplers = nullptr, - }, - { .binding = ASTC_BINDING_SWIZZLE_BUFFER, .descriptorType = VK_DESCRIPTOR_TYPE_STORAGE_BUFFER, .descriptorCount = 1, @@ -143,7 +112,8 @@ VkDescriptorUpdateTemplateEntryKHR BuildInputOutputDescriptorUpdateTemplate() { }; } -std::array<VkDescriptorUpdateTemplateEntryKHR, 8> BuildASTCPassDescriptorUpdateTemplateEntry() { +std::array<VkDescriptorUpdateTemplateEntryKHR, ASTC_NUM_BINDINGS> +BuildASTCPassDescriptorUpdateTemplateEntry() { return {{ { .dstBinding = ASTC_BINDING_INPUT_BUFFER, @@ -162,38 +132,6 @@ std::array<VkDescriptorUpdateTemplateEntryKHR, 8> BuildASTCPassDescriptorUpdateT .stride = sizeof(DescriptorUpdateEntry), }, { - .dstBinding = ASTC_BINDING_6_TO_8_BUFFER, - .dstArrayElement = 0, - .descriptorCount = 1, - .descriptorType = VK_DESCRIPTOR_TYPE_STORAGE_BUFFER, - .offset = ASTC_BINDING_6_TO_8_BUFFER * sizeof(DescriptorUpdateEntry), - .stride = sizeof(DescriptorUpdateEntry), - }, - { - .dstBinding = ASTC_BINDING_7_TO_8_BUFFER, - .dstArrayElement = 0, - .descriptorCount = 1, - .descriptorType = VK_DESCRIPTOR_TYPE_STORAGE_BUFFER, - .offset = ASTC_BINDING_7_TO_8_BUFFER * sizeof(DescriptorUpdateEntry), - .stride = sizeof(DescriptorUpdateEntry), - }, - { - .dstBinding = ASTC_BINDING_8_TO_8_BUFFER, - .dstArrayElement = 0, - .descriptorCount = 1, - .descriptorType = VK_DESCRIPTOR_TYPE_STORAGE_BUFFER, - .offset = ASTC_BINDING_8_TO_8_BUFFER * sizeof(DescriptorUpdateEntry), - .stride = sizeof(DescriptorUpdateEntry), - }, - { - .dstBinding = ASTC_BINDING_BYTE_TO_16_BUFFER, - .dstArrayElement = 0, - .descriptorCount = 1, - .descriptorType = VK_DESCRIPTOR_TYPE_STORAGE_BUFFER, - .offset = ASTC_BINDING_BYTE_TO_16_BUFFER * sizeof(DescriptorUpdateEntry), - .stride = sizeof(DescriptorUpdateEntry), - }, - { .dstBinding = ASTC_BINDING_SWIZZLE_BUFFER, .dstArrayElement = 0, .descriptorCount = 1, @@ -222,15 +160,6 @@ struct AstcPushConstants { u32 block_height_mask; }; -struct AstcBufferData { - decltype(SWIZZLE_TABLE) swizzle_table_buffer = SWIZZLE_TABLE; - decltype(EncodingsValues) encoding_values = EncodingsValues; - decltype(REPLICATE_6_BIT_TO_8_TABLE) replicate_6_to_8 = REPLICATE_6_BIT_TO_8_TABLE; - decltype(REPLICATE_7_BIT_TO_8_TABLE) replicate_7_to_8 = REPLICATE_7_BIT_TO_8_TABLE; - decltype(REPLICATE_8_BIT_TO_8_TABLE) replicate_8_to_8 = REPLICATE_8_BIT_TO_8_TABLE; - decltype(REPLICATE_BYTE_TO_16_TABLE) replicate_byte_to_16 = REPLICATE_BYTE_TO_16_TABLE; -} constexpr ASTC_BUFFER_DATA; - } // Anonymous namespace VKComputePass::VKComputePass(const Device& device, VKDescriptorPool& descriptor_pool, @@ -423,7 +352,7 @@ ASTCDecoderPass::ASTCDecoderPass(const Device& device_, VKScheduler& scheduler_, ASTCDecoderPass::~ASTCDecoderPass() = default; void ASTCDecoderPass::MakeDataBuffer() { - constexpr size_t TOTAL_BUFFER_SIZE = sizeof(ASTC_BUFFER_DATA) + sizeof(SWIZZLE_TABLE); + constexpr size_t TOTAL_BUFFER_SIZE = sizeof(ASTC_ENCODINGS_VALUES) + sizeof(SWIZZLE_TABLE); data_buffer = device.GetLogical().CreateBuffer(VkBufferCreateInfo{ .sType = VK_STRUCTURE_TYPE_BUFFER_CREATE_INFO, .pNext = nullptr, @@ -437,9 +366,10 @@ void ASTCDecoderPass::MakeDataBuffer() { data_buffer_commit = memory_allocator.Commit(data_buffer, MemoryUsage::Upload); const auto staging_ref = staging_buffer_pool.Request(TOTAL_BUFFER_SIZE, MemoryUsage::Upload); - std::memcpy(staging_ref.mapped_span.data(), &ASTC_BUFFER_DATA, sizeof(ASTC_BUFFER_DATA)); + std::memcpy(staging_ref.mapped_span.data(), &ASTC_ENCODINGS_VALUES, + sizeof(ASTC_ENCODINGS_VALUES)); // Tack on the swizzle table at the end of the buffer - std::memcpy(staging_ref.mapped_span.data() + sizeof(ASTC_BUFFER_DATA), &SWIZZLE_TABLE, + std::memcpy(staging_ref.mapped_span.data() + sizeof(ASTC_ENCODINGS_VALUES), &SWIZZLE_TABLE, sizeof(SWIZZLE_TABLE)); scheduler.Record([src = staging_ref.buffer, offset = staging_ref.offset, dst = *data_buffer, @@ -509,18 +439,8 @@ void ASTCDecoderPass::Assemble(Image& image, const StagingBufferRef& map, update_descriptor_queue.Acquire(); update_descriptor_queue.AddBuffer(map.buffer, input_offset, image.guest_size_bytes - swizzle.buffer_offset); - update_descriptor_queue.AddBuffer(*data_buffer, offsetof(AstcBufferData, encoding_values), - sizeof(AstcBufferData::encoding_values)); - update_descriptor_queue.AddBuffer(*data_buffer, offsetof(AstcBufferData, replicate_6_to_8), - sizeof(AstcBufferData::replicate_6_to_8)); - update_descriptor_queue.AddBuffer(*data_buffer, offsetof(AstcBufferData, replicate_7_to_8), - sizeof(AstcBufferData::replicate_7_to_8)); - update_descriptor_queue.AddBuffer(*data_buffer, offsetof(AstcBufferData, replicate_8_to_8), - sizeof(AstcBufferData::replicate_8_to_8)); - update_descriptor_queue.AddBuffer(*data_buffer, - offsetof(AstcBufferData, replicate_byte_to_16), - sizeof(AstcBufferData::replicate_byte_to_16)); - update_descriptor_queue.AddBuffer(*data_buffer, sizeof(AstcBufferData), + update_descriptor_queue.AddBuffer(*data_buffer, 0, sizeof(ASTC_ENCODINGS_VALUES)); + update_descriptor_queue.AddBuffer(*data_buffer, sizeof(ASTC_ENCODINGS_VALUES), sizeof(SWIZZLE_TABLE)); update_descriptor_queue.AddImage(image.StorageImageView(swizzle.level)); @@ -569,6 +489,7 @@ void ASTCDecoderPass::Assemble(Image& image, const StagingBufferRef& map, cmdbuf.PipelineBarrier(VK_PIPELINE_STAGE_COMPUTE_SHADER_BIT, VK_PIPELINE_STAGE_ALL_COMMANDS_BIT, 0, image_barrier); }); + scheduler.Finish(); } } // namespace Vulkan diff --git a/src/video_core/renderer_vulkan/vk_staging_buffer_pool.cpp b/src/video_core/renderer_vulkan/vk_staging_buffer_pool.cpp index 7a1232497..0412b5234 100644 --- a/src/video_core/renderer_vulkan/vk_staging_buffer_pool.cpp +++ b/src/video_core/renderer_vulkan/vk_staging_buffer_pool.cpp @@ -12,6 +12,7 @@ #include "common/assert.h" #include "common/bit_util.h" #include "common/common_types.h" +#include "common/literals.h" #include "video_core/renderer_vulkan/vk_scheduler.h" #include "video_core/renderer_vulkan/vk_staging_buffer_pool.h" #include "video_core/vulkan_common/vulkan_device.h" @@ -19,12 +20,15 @@ namespace Vulkan { namespace { + +using namespace Common::Literals; + // Maximum potential alignment of a Vulkan buffer constexpr VkDeviceSize MAX_ALIGNMENT = 256; // Maximum size to put elements in the stream buffer -constexpr VkDeviceSize MAX_STREAM_BUFFER_REQUEST_SIZE = 8 * 1024 * 1024; +constexpr VkDeviceSize MAX_STREAM_BUFFER_REQUEST_SIZE = 8_MiB; // Stream buffer size in bytes -constexpr VkDeviceSize STREAM_BUFFER_SIZE = 128 * 1024 * 1024; +constexpr VkDeviceSize STREAM_BUFFER_SIZE = 128_MiB; constexpr VkDeviceSize REGION_SIZE = STREAM_BUFFER_SIZE / StagingBufferPool::NUM_SYNCS; constexpr VkMemoryPropertyFlags HOST_FLAGS = diff --git a/src/video_core/renderer_vulkan/vk_stream_buffer.cpp b/src/video_core/renderer_vulkan/vk_stream_buffer.cpp index a09fe084e..7b4875d0e 100644 --- a/src/video_core/renderer_vulkan/vk_stream_buffer.cpp +++ b/src/video_core/renderer_vulkan/vk_stream_buffer.cpp @@ -10,6 +10,7 @@ #include "common/alignment.h" #include "common/assert.h" +#include "common/literals.h" #include "video_core/renderer_vulkan/vk_scheduler.h" #include "video_core/renderer_vulkan/vk_stream_buffer.h" #include "video_core/vulkan_common/vulkan_device.h" @@ -19,6 +20,8 @@ namespace Vulkan { namespace { +using namespace Common::Literals; + constexpr VkBufferUsageFlags BUFFER_USAGE = VK_BUFFER_USAGE_VERTEX_BUFFER_BIT | VK_BUFFER_USAGE_INDEX_BUFFER_BIT | VK_BUFFER_USAGE_UNIFORM_BUFFER_BIT | VK_BUFFER_USAGE_STORAGE_BUFFER_BIT; @@ -26,7 +29,7 @@ constexpr VkBufferUsageFlags BUFFER_USAGE = constexpr u64 WATCHES_INITIAL_RESERVE = 0x4000; constexpr u64 WATCHES_RESERVE_CHUNK = 0x1000; -constexpr u64 PREFERRED_STREAM_BUFFER_SIZE = 256 * 1024 * 1024; +constexpr u64 PREFERRED_STREAM_BUFFER_SIZE = 256_MiB; /// Find a memory type with the passed requirements std::optional<u32> FindMemoryType(const VkPhysicalDeviceMemoryProperties& properties, diff --git a/src/video_core/texture_cache/texture_cache.h b/src/video_core/texture_cache/texture_cache.h index e7f8478b4..84530a179 100644 --- a/src/video_core/texture_cache/texture_cache.h +++ b/src/video_core/texture_cache/texture_cache.h @@ -19,9 +19,8 @@ #include <boost/container/small_vector.hpp> #include "common/alignment.h" -#include "common/common_funcs.h" -#include "common/common_sizes.h" #include "common/common_types.h" +#include "common/literals.h" #include "common/logging/log.h" #include "common/settings.h" #include "video_core/compatible_formats.h" @@ -59,6 +58,7 @@ using VideoCore::Surface::PixelFormat; using VideoCore::Surface::PixelFormatFromDepthFormat; using VideoCore::Surface::PixelFormatFromRenderTargetFormat; using VideoCore::Surface::SurfaceType; +using namespace Common::Literals; template <class P> class TextureCache { @@ -79,8 +79,8 @@ class TextureCache { /// Sampler ID for bugged sampler ids static constexpr SamplerId NULL_SAMPLER_ID{0}; - static constexpr u64 DEFAULT_EXPECTED_MEMORY = Common::Size_1_GB; - static constexpr u64 DEFAULT_CRITICAL_MEMORY = Common::Size_2_GB; + static constexpr u64 DEFAULT_EXPECTED_MEMORY = 1_GiB; + static constexpr u64 DEFAULT_CRITICAL_MEMORY = 2_GiB; using Runtime = typename P::Runtime; using Image = typename P::Image; @@ -400,8 +400,8 @@ TextureCache<P>::TextureCache(Runtime& runtime_, VideoCore::RasterizerInterface& minimum_memory = 0; } else { // on OGL we can be more conservatives as the driver takes care. - expected_memory = DEFAULT_EXPECTED_MEMORY + Common::Size_512_MB; - critical_memory = DEFAULT_CRITICAL_MEMORY + Common::Size_1_GB; + expected_memory = DEFAULT_EXPECTED_MEMORY + 512_MiB; + critical_memory = DEFAULT_CRITICAL_MEMORY + 1_GiB; minimum_memory = expected_memory; } } diff --git a/src/video_core/textures/astc.cpp b/src/video_core/textures/astc.cpp index 9b2177ebd..7b756ba41 100644 --- a/src/video_core/textures/astc.cpp +++ b/src/video_core/textures/astc.cpp @@ -269,7 +269,7 @@ static void DecodeQuintBlock(InputBitStream& bits, IntegerEncodedVector& result, static void DecodeIntegerSequence(IntegerEncodedVector& result, InputBitStream& bits, u32 maxRange, u32 nValues) { // Determine encoding parameters - IntegerEncodedValue val = EncodingsValues[maxRange]; + IntegerEncodedValue val = ASTC_ENCODINGS_VALUES[maxRange]; // Start decoding u32 nValsDecoded = 0; @@ -310,7 +310,7 @@ struct TexelWeightParams { nIdxs *= 2; } - return EncodingsValues[m_MaxWeight].GetBitLength(nIdxs); + return ASTC_ENCODINGS_VALUES[m_MaxWeight].GetBitLength(nIdxs); } u32 GetNumWeightValues() const { @@ -551,6 +551,8 @@ static void FillError(std::span<u32> outBuf, u32 blockWidth, u32 blockHeight) { } } } + +static constexpr auto REPLICATE_BYTE_TO_16_TABLE = MakeReplicateTable<u32, 8, 16>(); static constexpr u32 ReplicateByteTo16(std::size_t value) { return REPLICATE_BYTE_TO_16_TABLE[value]; } @@ -753,12 +755,12 @@ static void DecodeColorValues(u32* out, std::span<u8> data, const u32* modes, co // figure out the max value for each of them... u32 range = 256; while (--range > 0) { - IntegerEncodedValue val = EncodingsValues[range]; + IntegerEncodedValue val = ASTC_ENCODINGS_VALUES[range]; u32 bitLength = val.GetBitLength(nValues); if (bitLength <= nBitsForColorData) { // Find the smallest possible range that matches the given encoding while (--range > 0) { - IntegerEncodedValue newval = EncodingsValues[range]; + IntegerEncodedValue newval = ASTC_ENCODINGS_VALUES[range]; if (!newval.MatchesEncoding(val)) { break; } diff --git a/src/video_core/textures/astc.h b/src/video_core/textures/astc.h index c1c37dfe7..0229ae122 100644 --- a/src/video_core/textures/astc.h +++ b/src/video_core/textures/astc.h @@ -77,7 +77,7 @@ constexpr std::array<IntegerEncodedValue, 256> MakeEncodedValues() { return encodings; } -constexpr std::array<IntegerEncodedValue, 256> EncodingsValues = MakeEncodedValues(); +constexpr std::array<IntegerEncodedValue, 256> ASTC_ENCODINGS_VALUES = MakeEncodedValues(); // Replicates low num_bits such that [(to_bit - 1):(to_bit - 1 - from_bit)] // is the same as [(num_bits - 1):0] and repeats all the way down. @@ -116,19 +116,10 @@ constexpr auto MakeReplicateTable() { return table; } -constexpr auto REPLICATE_BYTE_TO_16_TABLE = MakeReplicateTable<u32, 8, 16>(); constexpr auto REPLICATE_6_BIT_TO_8_TABLE = MakeReplicateTable<u32, 6, 8>(); constexpr auto REPLICATE_7_BIT_TO_8_TABLE = MakeReplicateTable<u32, 7, 8>(); constexpr auto REPLICATE_8_BIT_TO_8_TABLE = MakeReplicateTable<u32, 8, 8>(); -struct AstcBufferData { - decltype(EncodingsValues) encoding_values = EncodingsValues; - decltype(REPLICATE_6_BIT_TO_8_TABLE) replicate_6_to_8 = REPLICATE_6_BIT_TO_8_TABLE; - decltype(REPLICATE_7_BIT_TO_8_TABLE) replicate_7_to_8 = REPLICATE_7_BIT_TO_8_TABLE; - decltype(REPLICATE_8_BIT_TO_8_TABLE) replicate_8_to_8 = REPLICATE_8_BIT_TO_8_TABLE; - decltype(REPLICATE_BYTE_TO_16_TABLE) replicate_byte_to_16 = REPLICATE_BYTE_TO_16_TABLE; -} constexpr ASTC_BUFFER_DATA; - void Decompress(std::span<const uint8_t> data, uint32_t width, uint32_t height, uint32_t depth, uint32_t block_width, uint32_t block_height, std::span<uint8_t> output); diff --git a/src/video_core/vulkan_common/vulkan_device.cpp b/src/video_core/vulkan_common/vulkan_device.cpp index 707a8b8fb..23814afd2 100644 --- a/src/video_core/vulkan_common/vulkan_device.cpp +++ b/src/video_core/vulkan_common/vulkan_device.cpp @@ -821,9 +821,9 @@ void Device::CollectTelemetryParameters() { void Device::CollectPhysicalMemoryInfo() { const auto mem_properties = physical.GetMemoryProperties(); - const std::size_t num_properties = mem_properties.memoryHeapCount; + const size_t num_properties = mem_properties.memoryHeapCount; device_access_memory = 0; - for (std::size_t element = 0; element < num_properties; element++) { + for (size_t element = 0; element < num_properties; ++element) { if ((mem_properties.memoryHeaps[element].flags & VK_MEMORY_HEAP_DEVICE_LOCAL_BIT) != 0) { device_access_memory += mem_properties.memoryHeaps[element].size; } diff --git a/src/video_core/vulkan_common/vulkan_device.h b/src/video_core/vulkan_common/vulkan_device.h index a1aba973b..88b298196 100644 --- a/src/video_core/vulkan_common/vulkan_device.h +++ b/src/video_core/vulkan_common/vulkan_device.h @@ -264,21 +264,22 @@ private: bool IsFormatSupported(VkFormat wanted_format, VkFormatFeatureFlags wanted_usage, FormatType format_type) const; - VkInstance instance; ///< Vulkan instance. - vk::DeviceDispatch dld; ///< Device function pointers. - vk::PhysicalDevice physical; ///< Physical device. - VkPhysicalDeviceProperties properties; ///< Device properties. - vk::Device logical; ///< Logical device. - vk::Queue graphics_queue; ///< Main graphics queue. - vk::Queue present_queue; ///< Main present queue. - u32 instance_version{}; ///< Vulkan onstance version. - u32 graphics_family{}; ///< Main graphics queue family index. - u32 present_family{}; ///< Main present queue family index. - VkDriverIdKHR driver_id{}; ///< Driver ID. - VkShaderStageFlags guest_warp_stages{}; ///< Stages where the guest warp size can be forced.ed - bool is_optimal_astc_supported{}; ///< Support for native ASTC. - bool is_float16_supported{}; ///< Support for float16 arithmetics. - bool is_warp_potentially_bigger{}; ///< Host warp size can be bigger than guest. + VkInstance instance; ///< Vulkan instance. + vk::DeviceDispatch dld; ///< Device function pointers. + vk::PhysicalDevice physical; ///< Physical device. + VkPhysicalDeviceProperties properties; ///< Device properties. + vk::Device logical; ///< Logical device. + vk::Queue graphics_queue; ///< Main graphics queue. + vk::Queue present_queue; ///< Main present queue. + u32 instance_version{}; ///< Vulkan onstance version. + u32 graphics_family{}; ///< Main graphics queue family index. + u32 present_family{}; ///< Main present queue family index. + VkDriverIdKHR driver_id{}; ///< Driver ID. + VkShaderStageFlags guest_warp_stages{}; ///< Stages where the guest warp size can be forced. + u64 device_access_memory{}; ///< Total size of device local memory in bytes. + bool is_optimal_astc_supported{}; ///< Support for native ASTC. + bool is_float16_supported{}; ///< Support for float16 arithmetics. + bool is_warp_potentially_bigger{}; ///< Host warp size can be bigger than guest. bool is_formatless_image_load_supported{}; ///< Support for shader image read without format. bool is_shader_storage_image_multisample{}; ///< Support for image operations on MSAA images. bool is_blit_depth_stencil_supported{}; ///< Support for blitting from and to depth stencil. @@ -309,8 +310,6 @@ private: /// Nsight Aftermath GPU crash tracker std::unique_ptr<NsightAftermathTracker> nsight_aftermath_tracker; - - u64 device_access_memory; }; } // namespace Vulkan |