1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306
|
/* This file is part of the Spring engine (GPL v2 or later), see LICENSE.html */
#ifndef MEMPOOL_TYPES_H
#define MEMPOOL_TYPES_H
#include <cassert>
#include <cstring> // memset
#include <array>
#include <deque>
#include <vector>
#include <memory>
#include "System/UnorderedMap.hpp"
#include "System/ContainerUtil.h"
#include "System/SafeUtil.h"
template<size_t S> struct DynMemPool {
public:
void* allocMem(size_t size) {
assert(size <= PAGE_SIZE());
uint8_t* m = nullptr;
size_t i = 0;
if (indcs.empty()) {
pages.emplace_back();
i = pages.size() - 1;
} else {
// must pop before ctor runs; objects can be created recursively
i = spring::VectorBackPop(indcs);
}
m = pages[curr_page_index = i].data();
table.emplace(m, i);
return m;
}
template<typename T, typename... A> T* alloc(A&&... a) {
static_assert(sizeof(T) <= PAGE_SIZE(), "");
return new (allocMem(sizeof(T))) T(std::forward<A>(a)...);
}
void freeMem(void* m) {
assert(mapped(m));
const auto iter = table.find(m);
const auto pair = std::pair<void*, size_t>{iter->first, iter->second};
std::memset(pages[pair.second].data(), 0, PAGE_SIZE());
indcs.push_back(pair.second);
table.erase(pair.first);
}
template<typename T> void free(T*& p) {
assert(mapped(p));
void* m = p;
spring::SafeDestruct(p);
// must free after dtor runs, since that can trigger *another* ctor call
// by proxy (~CUnit -> ~CObject -> DependentDied -> CommandAI::FinishCmd
// -> CBuilderCAI::ExecBuildCmd -> UnitLoader::LoadUnit -> CUnit e.g.)
freeMem(m);
}
static constexpr size_t PAGE_SIZE() { return S; }
size_t alloc_size() const { return (pages.size() * PAGE_SIZE()); } // size of total number of pages added over the pool's lifetime
size_t freed_size() const { return (indcs.size() * PAGE_SIZE()); } // size of number of pages that were freed and are awaiting reuse
bool mapped(void* p) const { return (table.find(p) != table.end()); }
bool alloced(void* p) const { return ((curr_page_index < pages.size()) && (pages[curr_page_index].data() == p)); }
void clear() {
pages.clear();
indcs.clear();
table.clear();
curr_page_index = 0;
}
void reserve(size_t n) {
indcs.reserve(n);
table.reserve(n);
}
private:
std::deque<std::array<uint8_t, S>> pages;
std::vector<size_t> indcs;
// <pointer, page index> (non-intrusive)
spring::unsynced_map<void*, size_t> table;
size_t curr_page_index = 0;
};
// fixed-size dynamic version
// page size per chunk, number of chunks, number of pages per chunk
// at most <N * K> simultaneous allocations can be made from a pool
// of size NxK, each of which consumes S bytes (N chunks with every
// chunk consuming S * K bytes) excluding overhead
template<size_t S, size_t N, size_t K> struct FixedDynMemPool {
public:
template<typename T, typename... A> T* alloc(A&&... a) {
static_assert(sizeof(T) <= PAGE_SIZE(), "");
return (new (allocMem(sizeof(T))) T(std::forward<A>(a)...));
}
void* allocMem(size_t size) {
uint8_t* ptr = nullptr;
if (indcs.empty()) {
// pool is full
if (num_chunks == N)
return ptr;
assert(chunks[num_chunks] == nullptr);
chunks[num_chunks].reset(new t_chunk_mem());
// reserve new indices; in reverse order since each will be popped from the back
indcs.reserve(K);
for (size_t j = 0; j < K; j++) {
indcs.push_back(static_cast<uint32_t>((num_chunks + 1) * K - j - 1));
}
num_chunks += 1;
}
const uint32_t idx = spring::VectorBackPop(indcs);
assert(size <= PAGE_SIZE());
memcpy(ptr = page_mem(page_index = idx), &idx, sizeof(idx));
return (ptr + sizeof(idx));
}
template<typename T> void free(T*& ptr) {
static_assert(sizeof(T) <= PAGE_SIZE(), "");
T* tmp = ptr;
spring::SafeDestruct(ptr);
freeMem(tmp);
}
void freeMem(void* ptr) {
const uint32_t idx = page_idx(ptr);
// zero-fill page
assert(idx < (N * K));
memset(page_mem(idx), 0, sizeof(idx) + S);
indcs.push_back(idx);
}
void reserve(size_t n) { indcs.reserve(n); }
void clear() {
indcs.clear();
// for every allocated chunk, add back all indices
// (objects are assumed to have already been freed)
for (size_t i = 0; i < num_chunks; i++) {
for (size_t j = 0; j < K; j++) {
indcs.push_back(static_cast<uint32_t>((i + 1) * K - j - 1));
}
}
page_index = 0;
}
static constexpr size_t NUM_CHUNKS() { return N; } // size K*S
static constexpr size_t NUM_PAGES() { return K; } // per chunk
static constexpr size_t PAGE_SIZE() { return S; }
const uint8_t* page_mem(size_t idx, size_t ofs = 0) const {
const t_chunk_ptr& chunk_ptr = chunks[idx / K];
const t_chunk_mem& chunk_mem = *chunk_ptr;
return (&chunk_mem[idx % K][0] + ofs);
}
uint8_t* page_mem(size_t idx, size_t ofs = 0) {
t_chunk_ptr& chunk_ptr = chunks[idx / K];
t_chunk_mem& chunk_mem = *chunk_ptr;
return (&chunk_mem[idx % K][0] + ofs);
}
uint32_t page_idx(void* ptr) const {
const uint8_t* raw_ptr = reinterpret_cast<const uint8_t*>(ptr);
const uint8_t* idx_ptr = raw_ptr - sizeof(uint32_t);
return (*reinterpret_cast<const uint32_t*>(idx_ptr));
}
size_t alloc_size() const { return (num_chunks * NUM_PAGES() * PAGE_SIZE()); } // size of total number of pages added over the pool's lifetime
size_t freed_size() const { return (indcs.size() * PAGE_SIZE()); } // size of number of pages that were freed and are awaiting reuse
bool mapped(void* ptr) const { return ((page_idx(ptr) < (num_chunks * K)) && (page_mem(page_idx(ptr), sizeof(uint32_t)) == ptr)); }
bool alloced(void* ptr) const { return ((page_index < (num_chunks * K)) && (page_mem(page_index, sizeof(uint32_t)) == ptr)); }
private:
// first sizeof(uint32_t) bytes are reserved for index
typedef std::array<uint8_t[sizeof(uint32_t) + S], K> t_chunk_mem;
typedef std::unique_ptr<t_chunk_mem> t_chunk_ptr;
std::array<t_chunk_ptr, N> chunks;
std::vector<uint32_t> indcs;
size_t num_chunks = 0;
size_t page_index = 0;
};
// fixed-size version
template<size_t N, size_t S> struct StaticMemPool {
public:
StaticMemPool() { clear(); }
void* allocMem(size_t size) {
assert(size <= PAGE_SIZE());
static_assert(NUM_PAGES() != 0, "");
size_t i = 0;
assert(can_alloc());
if (free_page_count == 0) {
i = used_page_count++;
} else {
i = indcs[--free_page_count];
}
return (pages[curr_page_index = i].data());
}
template<typename T, typename... A> T* alloc(A&&... a) {
static_assert(sizeof(T) <= PAGE_SIZE(), "");
return new (allocMem(sizeof(T))) T(std::forward<A>(a)...);
}
void freeMem(void* m) {
assert(can_free());
assert(mapped(m));
std::memset(m, 0, PAGE_SIZE());
// mark page as free
indcs[free_page_count++] = base_offset(m) / PAGE_SIZE();
}
template<typename T> void free(T*& p) {
assert(mapped(p));
void* m = p;
spring::SafeDestruct(p);
freeMem(m);
}
static constexpr size_t NUM_PAGES() { return N; }
static constexpr size_t PAGE_SIZE() { return S; }
size_t alloc_size() const { return (used_page_count * PAGE_SIZE()); } // size of total number of pages added over the pool's lifetime
size_t freed_size() const { return (free_page_count * PAGE_SIZE()); } // size of number of pages that were freed and are awaiting reuse
size_t total_size() const { return (NUM_PAGES() * PAGE_SIZE()); }
size_t base_offset(const void* p) const { return (reinterpret_cast<const uint8_t*>(p) - reinterpret_cast<const uint8_t*>(pages[0].data())); }
bool mapped(const void* p) const { return (((base_offset(p) / PAGE_SIZE()) < total_size()) && ((base_offset(p) % PAGE_SIZE()) == 0)); }
bool alloced(const void* p) const { return (pages[curr_page_index].data() == p); }
bool can_alloc() const { return (used_page_count < NUM_PAGES() || free_page_count > 0); }
bool can_free() const { return (free_page_count < NUM_PAGES()); }
void reserve(size_t) {} // no-op
void clear() {
std::memset(pages.data(), 0, total_size());
std::memset(indcs.data(), 0, NUM_PAGES());
used_page_count = 0;
free_page_count = 0;
curr_page_index = 0;
}
private:
std::array<std::array<uint8_t, S>, N> pages;
std::array<size_t, N> indcs;
size_t used_page_count = 0;
size_t free_page_count = 0; // indcs[fpc-1] is the last recycled page
size_t curr_page_index = 0;
};
#endif
|