mirror of
https://github.com/shadps4-emu/shadPS4.git
synced 2025-06-26 04:16:18 +00:00
Some checks are pending
Build and Release / macos-sdl (push) Blocked by required conditions
Build and Release / macos-qt (push) Blocked by required conditions
Build and Release / reuse (push) Waiting to run
Build and Release / clang-format (push) Waiting to run
Build and Release / get-info (push) Waiting to run
Build and Release / windows-sdl (push) Blocked by required conditions
Build and Release / windows-qt (push) Blocked by required conditions
Build and Release / linux-sdl (push) Blocked by required conditions
Build and Release / linux-qt (push) Blocked by required conditions
Build and Release / linux-sdl-gcc (push) Blocked by required conditions
Build and Release / linux-qt-gcc (push) Blocked by required conditions
Build and Release / pre-release (push) Blocked by required conditions
* Update sceKernelMemoryPoolExpand Hardware tests show that this function is basically the same as sceKernelAllocateDirectMemory, with some minor differences. Update the memory searching code to match my updated AllocateDirectMemory code, with appropriate error conditions. * Update MemoryPoolReserve Only difference between real hw and our code is behavior with addr = 0. * Don't coalesce PoolReserved areas. Real hardware doesn't coalesce them. * Update PoolCommit Plenty of edge case behaviors to handle here. Addresses are treated as fixed, EINVAL is returned for bad mappings, name should be preserved from PoolReserving, committed areas should coalesce, reserved areas get their phys_base updated * Formatting * Adjust fixed PoolReserve path Hardware tests suggest this will overwrite all VMAs in the range. Run UnmapMemoryImpl on the full area, then reserve. Same logic applies to normal reservations too. Also adjusts logic of the non-fixed path to more closely align with hardware observations. * Remove phys_base modifications This can be handled later. Doing the logic properly would likely take work in MergeAdjacent, and would probably need to be applied to normal dmem mappings too. * Use VMAHandle.Contains() Why do extra math when we have a function specifically for this? * Update memory.cpp * Remove unnecessary code Since I've removed those two asserts, these two lines of code effectively do nothing. * Clang * Fix names * Fix PoolDecommit Should fix the address space regressions in UE titles on Windows. * Fix error log Should make the cause of this clearer? * Clang * Oops * Remove coalesce on PoolCommit Windows makes this more difficult. * Track pool budgets If you try to commit more pooled memory than is allocated, PoolCommit returns ENOMEM. Also fixes error conditions for PoolDecommit, that should return EINVAL if given an address that isn't part of the pool. Note: Seems like the pool budget can't hit zero? I used a <= comparison based on hardware tests, otherwise we're able to make more mappings than real hardware can.
285 lines
7.8 KiB
C++
285 lines
7.8 KiB
C++
// SPDX-FileCopyrightText: Copyright 2024 shadPS4 Emulator Project
|
|
// SPDX-License-Identifier: GPL-2.0-or-later
|
|
|
|
#pragma once
|
|
|
|
#include <map>
|
|
#include <mutex>
|
|
#include <string_view>
|
|
#include "common/enum.h"
|
|
#include "common/singleton.h"
|
|
#include "common/types.h"
|
|
#include "core/address_space.h"
|
|
#include "core/libraries/kernel/memory.h"
|
|
|
|
namespace Vulkan {
|
|
class Rasterizer;
|
|
}
|
|
|
|
namespace Libraries::Kernel {
|
|
struct OrbisQueryInfo;
|
|
}
|
|
|
|
namespace Core::Devtools::Widget {
|
|
class MemoryMapViewer;
|
|
}
|
|
|
|
namespace Core {
|
|
|
|
enum class MemoryProt : u32 {
|
|
NoAccess = 0,
|
|
CpuRead = 1,
|
|
CpuReadWrite = 2,
|
|
GpuRead = 16,
|
|
GpuWrite = 32,
|
|
GpuReadWrite = 48,
|
|
};
|
|
DECLARE_ENUM_FLAG_OPERATORS(MemoryProt)
|
|
|
|
enum class MemoryMapFlags : u32 {
|
|
NoFlags = 0,
|
|
Shared = 1,
|
|
Private = 2,
|
|
Fixed = 0x10,
|
|
NoOverwrite = 0x0080,
|
|
NoSync = 0x800,
|
|
NoCore = 0x20000,
|
|
NoCoalesce = 0x400000,
|
|
};
|
|
DECLARE_ENUM_FLAG_OPERATORS(MemoryMapFlags)
|
|
|
|
enum class VMAType : u32 {
|
|
Free = 0,
|
|
Reserved = 1,
|
|
Direct = 2,
|
|
Flexible = 3,
|
|
Pooled = 4,
|
|
PoolReserved = 5,
|
|
Stack = 6,
|
|
Code = 7,
|
|
File = 8,
|
|
};
|
|
|
|
struct DirectMemoryArea {
|
|
PAddr base = 0;
|
|
size_t size = 0;
|
|
int memory_type = 0;
|
|
bool is_pooled = false;
|
|
bool is_free = true;
|
|
|
|
PAddr GetEnd() const {
|
|
return base + size;
|
|
}
|
|
|
|
bool CanMergeWith(const DirectMemoryArea& next) const {
|
|
if (base + size != next.base) {
|
|
return false;
|
|
}
|
|
if (is_free != next.is_free) {
|
|
return false;
|
|
}
|
|
return true;
|
|
}
|
|
};
|
|
|
|
struct VirtualMemoryArea {
|
|
VAddr base = 0;
|
|
size_t size = 0;
|
|
PAddr phys_base = 0;
|
|
VMAType type = VMAType::Free;
|
|
MemoryProt prot = MemoryProt::NoAccess;
|
|
bool disallow_merge = false;
|
|
std::string name = "";
|
|
uintptr_t fd = 0;
|
|
bool is_exec = false;
|
|
|
|
bool Contains(VAddr addr, size_t size) const {
|
|
return addr >= base && (addr + size) <= (base + this->size);
|
|
}
|
|
|
|
bool IsFree() const noexcept {
|
|
return type == VMAType::Free;
|
|
}
|
|
|
|
bool IsMapped() const noexcept {
|
|
return type != VMAType::Free && type != VMAType::Reserved && type != VMAType::PoolReserved;
|
|
}
|
|
|
|
bool CanMergeWith(const VirtualMemoryArea& next) const {
|
|
if (disallow_merge || next.disallow_merge) {
|
|
return false;
|
|
}
|
|
if (base + size != next.base) {
|
|
return false;
|
|
}
|
|
if (type == VMAType::Direct && phys_base + size != next.phys_base) {
|
|
return false;
|
|
}
|
|
if (prot != next.prot || type != next.type) {
|
|
return false;
|
|
}
|
|
return true;
|
|
}
|
|
};
|
|
|
|
class MemoryManager {
|
|
using DMemMap = std::map<PAddr, DirectMemoryArea>;
|
|
using DMemHandle = DMemMap::iterator;
|
|
|
|
using VMAMap = std::map<VAddr, VirtualMemoryArea>;
|
|
using VMAHandle = VMAMap::iterator;
|
|
|
|
public:
|
|
explicit MemoryManager();
|
|
~MemoryManager();
|
|
|
|
void SetRasterizer(Vulkan::Rasterizer* rasterizer_) {
|
|
rasterizer = rasterizer_;
|
|
}
|
|
|
|
AddressSpace& GetAddressSpace() {
|
|
return impl;
|
|
}
|
|
|
|
u64 GetTotalDirectSize() const {
|
|
return total_direct_size;
|
|
}
|
|
|
|
u64 GetTotalFlexibleSize() const {
|
|
return total_flexible_size;
|
|
}
|
|
|
|
u64 GetAvailableFlexibleSize() const {
|
|
return total_flexible_size - flexible_usage;
|
|
}
|
|
|
|
VAddr SystemReservedVirtualBase() noexcept {
|
|
return impl.SystemReservedVirtualBase();
|
|
}
|
|
|
|
bool IsValidGpuMapping(VAddr virtual_addr, u64 size) {
|
|
// The PS4's GPU can only handle 40 bit addresses.
|
|
const VAddr max_gpu_address{0x10000000000};
|
|
return virtual_addr + size < max_gpu_address;
|
|
}
|
|
|
|
bool IsValidAddress(const void* addr) const noexcept {
|
|
const VAddr virtual_addr = reinterpret_cast<VAddr>(addr);
|
|
const auto end_it = std::prev(vma_map.end());
|
|
const VAddr end_addr = end_it->first + end_it->second.size;
|
|
return virtual_addr >= vma_map.begin()->first && virtual_addr < end_addr;
|
|
}
|
|
|
|
u64 ClampRangeSize(VAddr virtual_addr, u64 size);
|
|
|
|
bool TryWriteBacking(void* address, const void* data, u32 num_bytes);
|
|
|
|
void SetupMemoryRegions(u64 flexible_size, bool use_extended_mem1, bool use_extended_mem2);
|
|
|
|
PAddr PoolExpand(PAddr search_start, PAddr search_end, size_t size, u64 alignment);
|
|
|
|
PAddr Allocate(PAddr search_start, PAddr search_end, size_t size, u64 alignment,
|
|
int memory_type);
|
|
|
|
void Free(PAddr phys_addr, size_t size);
|
|
|
|
int PoolReserve(void** out_addr, VAddr virtual_addr, size_t size, MemoryMapFlags flags,
|
|
u64 alignment = 0);
|
|
|
|
int Reserve(void** out_addr, VAddr virtual_addr, size_t size, MemoryMapFlags flags,
|
|
u64 alignment = 0);
|
|
|
|
int PoolCommit(VAddr virtual_addr, size_t size, MemoryProt prot);
|
|
|
|
int MapMemory(void** out_addr, VAddr virtual_addr, size_t size, MemoryProt prot,
|
|
MemoryMapFlags flags, VMAType type, std::string_view name = "anon",
|
|
bool is_exec = false, PAddr phys_addr = -1, u64 alignment = 0);
|
|
|
|
int MapFile(void** out_addr, VAddr virtual_addr, size_t size, MemoryProt prot,
|
|
MemoryMapFlags flags, uintptr_t fd, size_t offset);
|
|
|
|
s32 PoolDecommit(VAddr virtual_addr, size_t size);
|
|
|
|
s32 UnmapMemory(VAddr virtual_addr, size_t size);
|
|
|
|
int QueryProtection(VAddr addr, void** start, void** end, u32* prot);
|
|
|
|
s32 Protect(VAddr addr, size_t size, MemoryProt prot);
|
|
|
|
s64 ProtectBytes(VAddr addr, VirtualMemoryArea vma_base, size_t size, MemoryProt prot);
|
|
|
|
int VirtualQuery(VAddr addr, int flags, ::Libraries::Kernel::OrbisVirtualQueryInfo* info);
|
|
|
|
int DirectMemoryQuery(PAddr addr, bool find_next,
|
|
::Libraries::Kernel::OrbisQueryInfo* out_info);
|
|
|
|
int DirectQueryAvailable(PAddr search_start, PAddr search_end, size_t alignment,
|
|
PAddr* phys_addr_out, size_t* size_out);
|
|
|
|
int GetDirectMemoryType(PAddr addr, int* directMemoryTypeOut, void** directMemoryStartOut,
|
|
void** directMemoryEndOut);
|
|
|
|
void NameVirtualRange(VAddr virtual_addr, size_t size, std::string_view name);
|
|
|
|
void InvalidateMemory(VAddr addr, u64 size) const;
|
|
|
|
private:
|
|
VMAHandle FindVMA(VAddr target) {
|
|
return std::prev(vma_map.upper_bound(target));
|
|
}
|
|
|
|
DMemHandle FindDmemArea(PAddr target) {
|
|
return std::prev(dmem_map.upper_bound(target));
|
|
}
|
|
|
|
template <typename Handle>
|
|
Handle MergeAdjacent(auto& handle_map, Handle iter) {
|
|
const auto next_vma = std::next(iter);
|
|
if (next_vma != handle_map.end() && iter->second.CanMergeWith(next_vma->second)) {
|
|
iter->second.size += next_vma->second.size;
|
|
handle_map.erase(next_vma);
|
|
}
|
|
|
|
if (iter != handle_map.begin()) {
|
|
auto prev_vma = std::prev(iter);
|
|
if (prev_vma->second.CanMergeWith(iter->second)) {
|
|
prev_vma->second.size += iter->second.size;
|
|
handle_map.erase(iter);
|
|
iter = prev_vma;
|
|
}
|
|
}
|
|
|
|
return iter;
|
|
}
|
|
|
|
VAddr SearchFree(VAddr virtual_addr, size_t size, u32 alignment = 0);
|
|
|
|
VMAHandle CarveVMA(VAddr virtual_addr, size_t size);
|
|
|
|
DMemHandle CarveDmemArea(PAddr addr, size_t size);
|
|
|
|
VMAHandle Split(VMAHandle vma_handle, size_t offset_in_vma);
|
|
|
|
DMemHandle Split(DMemHandle dmem_handle, size_t offset_in_area);
|
|
|
|
u64 UnmapBytesFromEntry(VAddr virtual_addr, VirtualMemoryArea vma_base, u64 size);
|
|
|
|
s32 UnmapMemoryImpl(VAddr virtual_addr, u64 size);
|
|
|
|
private:
|
|
AddressSpace impl;
|
|
DMemMap dmem_map;
|
|
VMAMap vma_map;
|
|
std::mutex mutex;
|
|
size_t total_direct_size{};
|
|
size_t total_flexible_size{};
|
|
size_t flexible_usage{};
|
|
size_t pool_budget{};
|
|
Vulkan::Rasterizer* rasterizer{};
|
|
|
|
friend class ::Core::Devtools::Widget::MemoryMapViewer;
|
|
};
|
|
|
|
using Memory = Common::Singleton<MemoryManager>;
|
|
|
|
} // namespace Core
|