Memory: move memory chunk into pImpl and make them dynamically allocated

Otherwise MSVC would give out-of-memory error on compile time

src/core/hle/kernel/process.cpp

@@ -188,10 +188,10 @@ ResultVal<VAddr> Process::HeapAllocate(VAddr target, u32 size, VMAPermission per
         u32 interval_size = interval.upper() - interval.lower();
         LOG_DEBUG(Kernel, "Allocated FCRAM region lower={:08X}, upper={:08X}", interval.lower(),
                   interval.upper());
-        std::fill(kernel.memory.fcram.begin() + interval.lower(),
+        std::fill(kernel.memory.GetFCRAMPointer(interval.lower()),
-                  kernel.memory.fcram.begin() + interval.upper(), 0);
+                  kernel.memory.GetFCRAMPointer(interval.upper()), 0);
         auto vma = vm_manager.MapBackingMemory(interval_target,
-                                               kernel.memory.fcram.data() + interval.lower(),
+                                               kernel.memory.GetFCRAMPointer(interval.lower()),
                                                interval_size, memory_state);
         ASSERT(vma.Succeeded());
         vm_manager.Reprotect(vma.Unwrap(), perms);
@@ -263,7 +263,7 @@ ResultVal<VAddr> Process::LinearAllocate(VAddr target, u32 size, VMAPermission p
         }
     }
 
-    u8* backing_memory = kernel.memory.fcram.data() + physical_offset;
+    u8* backing_memory = kernel.memory.GetFCRAMPointer(physical_offset);
 
     std::fill(backing_memory, backing_memory + size, 0);
     auto vma = vm_manager.MapBackingMemory(target, backing_memory, size, MemoryState::Continuous);

src/core/hle/kernel/shared_memory.cpp

@@ -43,8 +43,8 @@ ResultVal<SharedPtr<SharedMemory>> KernelSystem::CreateSharedMemory(
 
         ASSERT_MSG(offset, "Not enough space in region to allocate shared memory!");
 
-        std::fill(memory.fcram.data() + *offset, memory.fcram.data() + *offset + size, 0);
+        std::fill(memory.GetFCRAMPointer(*offset), memory.GetFCRAMPointer(*offset + size), 0);
-        shared_memory->backing_blocks = {{memory.fcram.data() + *offset, size}};
+        shared_memory->backing_blocks = {{memory.GetFCRAMPointer(*offset), size}};
         shared_memory->holding_memory += MemoryRegionInfo::Interval(*offset, *offset + size);
         shared_memory->linear_heap_phys_offset = *offset;
 
@@ -86,8 +86,8 @@ SharedPtr<SharedMemory> KernelSystem::CreateSharedMemoryForApplet(
     shared_memory->other_permissions = other_permissions;
     for (const auto& interval : backing_blocks) {
         shared_memory->backing_blocks.push_back(
-            {memory.fcram.data() + interval.lower(), interval.upper() - interval.lower()});
+            {memory.GetFCRAMPointer(interval.lower()), interval.upper() - interval.lower()});
-        std::fill(memory.fcram.data() + interval.lower(), memory.fcram.data() + interval.upper(),
+        std::fill(memory.GetFCRAMPointer(interval.lower()), memory.GetFCRAMPointer(interval.upper()),
                   0);
     }
     shared_memory->base_address = Memory::HEAP_VADDR + offset;

src/core/hle/kernel/thread.cpp

@@ -355,7 +355,7 @@ ResultVal<SharedPtr<Thread>> KernelSystem::CreateThread(std::string name, VAddr
 
         // Map the page to the current process' address space.
         vm_manager.MapBackingMemory(Memory::TLS_AREA_VADDR + available_page * Memory::PAGE_SIZE,
-                                    memory.fcram.data() + *offset, Memory::PAGE_SIZE,
+                                    memory.GetFCRAMPointer(*offset), Memory::PAGE_SIZE,
                                     MemoryState::Locked);
     }
 

src/core/memory.cpp

@@ -21,15 +21,34 @@
 
 namespace Memory {
 
+class MemorySystem::Impl {
+public:
+    Impl() {
+        std::fill(fcram.get(), fcram.get() + Memory::FCRAM_N3DS_SIZE, 0);
+        std::fill(vram.get(), vram.get() + Memory::VRAM_SIZE, 0);
+        std::fill(n3ds_extra_ram.get(), n3ds_extra_ram.get() + Memory::N3DS_EXTRA_RAM_SIZE, 0);
+    }
+
+    // Visual Studio would try to allocate these on compile time if they are std::array, which would exceed the memory limit.
+    std::unique_ptr<u8[]> fcram = std::make_unique<u8[]>(Memory::FCRAM_N3DS_SIZE);
+    std::unique_ptr<u8[]> vram = std::make_unique<u8[]>(Memory::VRAM_SIZE);
+    std::unique_ptr<u8[]> n3ds_extra_ram = std::make_unique<u8[]>(Memory::N3DS_EXTRA_RAM_SIZE);
+
+    PageTable* current_page_table = nullptr;
+};
+
+MemorySystem::MemorySystem() : impl(std::make_unique<Impl>()) {}
+MemorySystem::~MemorySystem() = default;
+
 void MemorySystem::SetCurrentPageTable(PageTable* page_table) {
-    current_page_table = page_table;
+    impl->current_page_table = page_table;
     if (Core::System::GetInstance().IsPoweredOn()) {
         Core::CPU().PageTableChanged();
     }
 }
 
 PageTable* MemorySystem::GetCurrentPageTable() const {
-    return current_page_table;
+    return impl->current_page_table;
 }
 
 static void MapPages(PageTable& page_table, u32 base, u32 size, u8* memory, PageType type) {
@@ -74,13 +93,13 @@ void UnmapRegion(PageTable& page_table, VAddr base, u32 size) {
 
 u8* MemorySystem::GetPointerForRasterizerCache(VAddr addr) {
     if (addr >= LINEAR_HEAP_VADDR && addr < LINEAR_HEAP_VADDR_END) {
-        return fcram.data() + (addr - LINEAR_HEAP_VADDR);
+        return impl->fcram.get() + (addr - LINEAR_HEAP_VADDR);
     }
     if (addr >= NEW_LINEAR_HEAP_VADDR && addr < NEW_LINEAR_HEAP_VADDR_END) {
-        return fcram.data() + (addr - NEW_LINEAR_HEAP_VADDR);
+        return impl->fcram.get() + (addr - NEW_LINEAR_HEAP_VADDR);
     }
     if (addr >= VRAM_VADDR && addr < VRAM_VADDR_END) {
-        return vram.data() + (addr - VRAM_VADDR);
+        return impl->vram.get() + (addr - VRAM_VADDR);
     }
     UNREACHABLE();
 }
@@ -103,7 +122,7 @@ T ReadMMIO(MMIORegionPointer mmio_handler, VAddr addr);
 
 template <typename T>
 T MemorySystem::Read(const VAddr vaddr) {
-    const u8* page_pointer = current_page_table->pointers[vaddr >> PAGE_BITS];
+    const u8* page_pointer = impl->current_page_table->pointers[vaddr >> PAGE_BITS];
     if (page_pointer) {
         // NOTE: Avoid adding any extra logic to this fast-path block
         T value;
@@ -114,7 +133,7 @@ T MemorySystem::Read(const VAddr vaddr) {
     // The memory access might do an MMIO or cached access, so we have to lock the HLE kernel state
     std::lock_guard<std::recursive_mutex> lock(HLE::g_hle_lock);
 
-    PageType type = current_page_table->attributes[vaddr >> PAGE_BITS];
+    PageType type = impl->current_page_table->attributes[vaddr >> PAGE_BITS];
     switch (type) {
     case PageType::Unmapped:
         LOG_ERROR(HW_Memory, "unmapped Read{} @ 0x{:08X}", sizeof(T) * 8, vaddr);
@@ -130,7 +149,7 @@ T MemorySystem::Read(const VAddr vaddr) {
         return value;
     }
     case PageType::Special:
-        return ReadMMIO<T>(GetMMIOHandler(*current_page_table, vaddr), vaddr);
+        return ReadMMIO<T>(GetMMIOHandler(*impl->current_page_table, vaddr), vaddr);
     default:
         UNREACHABLE();
     }
@@ -141,7 +160,7 @@ void WriteMMIO(MMIORegionPointer mmio_handler, VAddr addr, const T data);
 
 template <typename T>
 void MemorySystem::Write(const VAddr vaddr, const T data) {
-    u8* page_pointer = current_page_table->pointers[vaddr >> PAGE_BITS];
+    u8* page_pointer = impl->current_page_table->pointers[vaddr >> PAGE_BITS];
     if (page_pointer) {
         // NOTE: Avoid adding any extra logic to this fast-path block
         std::memcpy(&page_pointer[vaddr & PAGE_MASK], &data, sizeof(T));
@@ -151,7 +170,7 @@ void MemorySystem::Write(const VAddr vaddr, const T data) {
     // The memory access might do an MMIO or cached access, so we have to lock the HLE kernel state
     std::lock_guard<std::recursive_mutex> lock(HLE::g_hle_lock);
 
-    PageType type = current_page_table->attributes[vaddr >> PAGE_BITS];
+    PageType type = impl->current_page_table->attributes[vaddr >> PAGE_BITS];
     switch (type) {
     case PageType::Unmapped:
         LOG_ERROR(HW_Memory, "unmapped Write{} 0x{:08X} @ 0x{:08X}", sizeof(data) * 8, (u32)data,
@@ -166,7 +185,7 @@ void MemorySystem::Write(const VAddr vaddr, const T data) {
         break;
     }
     case PageType::Special:
-        WriteMMIO<T>(GetMMIOHandler(*current_page_table, vaddr), vaddr, data);
+        WriteMMIO<T>(GetMMIOHandler(*impl->current_page_table, vaddr), vaddr, data);
         break;
     default:
         UNREACHABLE();
@@ -199,12 +218,12 @@ bool MemorySystem::IsValidPhysicalAddress(const PAddr paddr) {
 }
 
 u8* MemorySystem::GetPointer(const VAddr vaddr) {
-    u8* page_pointer = current_page_table->pointers[vaddr >> PAGE_BITS];
+    u8* page_pointer = impl->current_page_table->pointers[vaddr >> PAGE_BITS];
     if (page_pointer) {
         return page_pointer + (vaddr & PAGE_MASK);
     }
 
-    if (current_page_table->attributes[vaddr >> PAGE_BITS] == PageType::RasterizerCachedMemory) {
+    if (impl->current_page_table->attributes[vaddr >> PAGE_BITS] == PageType::RasterizerCachedMemory) {
         return GetPointerForRasterizerCache(vaddr);
     }
 
@@ -256,16 +275,16 @@ u8* MemorySystem::GetPhysicalPointer(PAddr address) {
     u8* target_pointer = nullptr;
     switch (area->paddr_base) {
     case VRAM_PADDR:
-        target_pointer = vram.data() + offset_into_region;
+        target_pointer = impl->vram.get() + offset_into_region;
         break;
     case DSP_RAM_PADDR:
         target_pointer = Core::DSP().GetDspMemory().data() + offset_into_region;
         break;
     case FCRAM_PADDR:
-        target_pointer = fcram.data() + offset_into_region;
+        target_pointer = impl->fcram.get() + offset_into_region;
         break;
     case N3DS_EXTRA_RAM_PADDR:
-        target_pointer = n3ds_extra_ram.data() + offset_into_region;
+        target_pointer = impl->n3ds_extra_ram.get() + offset_into_region;
         break;
     default:
         UNREACHABLE();
@@ -303,7 +322,7 @@ void MemorySystem::RasterizerMarkRegionCached(PAddr start, u32 size, bool cached
 
     for (unsigned i = 0; i < num_pages; ++i, paddr += PAGE_SIZE) {
         for (VAddr vaddr : PhysicalToVirtualAddressForRasterizer(paddr)) {
-            PageType& page_type = current_page_table->attributes[vaddr >> PAGE_BITS];
+            PageType& page_type = impl->current_page_table->attributes[vaddr >> PAGE_BITS];
 
             if (cached) {
                 // Switch page type to cached if now cached
@@ -314,7 +333,7 @@ void MemorySystem::RasterizerMarkRegionCached(PAddr start, u32 size, bool cached
                     break;
                 case PageType::Memory:
                     page_type = PageType::RasterizerCachedMemory;
-                    current_page_table->pointers[vaddr >> PAGE_BITS] = nullptr;
+                    impl->current_page_table->pointers[vaddr >> PAGE_BITS] = nullptr;
                     break;
                 default:
                     UNREACHABLE();
@@ -328,7 +347,7 @@ void MemorySystem::RasterizerMarkRegionCached(PAddr start, u32 size, bool cached
                     break;
                 case PageType::RasterizerCachedMemory: {
                     page_type = PageType::Memory;
-                    current_page_table->pointers[vaddr >> PAGE_BITS] =
+                    impl->current_page_table->pointers[vaddr >> PAGE_BITS] =
                         GetPointerForRasterizerCache(vaddr & ~PAGE_MASK);
                     break;
                 }
@@ -730,8 +749,13 @@ void WriteMMIO<u64>(MMIORegionPointer mmio_handler, VAddr addr, const u64 data)
 }
 
 u32 MemorySystem::GetFCRAMOffset(u8* pointer) {
-    ASSERT(pointer >= fcram.data() && pointer < fcram.data() + fcram.size());
+    ASSERT(pointer >= impl->fcram.get() && pointer <= impl->fcram.get() + Memory::FCRAM_N3DS_SIZE);
-    return pointer - fcram.data();
+    return pointer - impl->fcram.get();
+}
+
+u8* MemorySystem::GetFCRAMPointer(u32 offset) {
+    ASSERT(offset <= Memory::FCRAM_N3DS_SIZE);
+    return impl->fcram.get() + offset;
 }
 
 } // namespace Memory

src/core/memory.h

@@ -6,6 +6,7 @@
 
 #include <array>
 #include <cstddef>
+#include <memory>
 #include <string>
 #include <vector>
 #include "common/common_types.h"
@@ -210,6 +211,9 @@ void RasterizerFlushVirtualRegion(VAddr start, u32 size, FlushMode mode);
 
 class MemorySystem {
 public:
+    MemorySystem();
+    ~MemorySystem();
+
     /// Currently active page table
     void SetCurrentPageTable(PageTable* page_table);
     PageTable* GetCurrentPageTable() const;
@@ -248,13 +252,14 @@ public:
     /// Gets offset in FCRAM from a pointer inside FCRAM range
     u32 GetFCRAMOffset(u8* pointer);
 
+    /// Gets pointer in FCRAM with given offset
+    u8* GetFCRAMPointer(u32 offset);
+
     /**
      * Mark each page touching the region as cached.
      */
     void RasterizerMarkRegionCached(PAddr start, u32 size, bool cached);
 
-    std::array<u8, Memory::FCRAM_N3DS_SIZE> fcram{};
-
 private:
     template <typename T>
     T Read(const VAddr vaddr);
@@ -270,10 +275,9 @@ private:
      */
     u8* GetPointerForRasterizerCache(VAddr addr);
 
-    std::array<u8, Memory::VRAM_SIZE> vram{};
+    class Impl;
-    std::array<u8, Memory::N3DS_EXTRA_RAM_SIZE> n3ds_extra_ram{};
 
-    PageTable* current_page_table = nullptr;
+    std::unique_ptr<Impl> impl;
 };
 
 /// Determines if the given VAddr is valid for the specified process.

src/video_core/video_core.h

@@ -12,7 +12,7 @@
 class EmuWindow;
 class RendererBase;
 
-namespace Memory{
+namespace Memory {
 class MemorySystem;
 }