python-doc-tw.github.io/c-api/memory.html at master · python-doc-tw/python-doc-tw.github.io

History

797 lines (747 loc) · 62 KB

Raw

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

307

308

309

310

311

312

313

314

315

316

317

318

319

320

321

322

323

324

325

326

327

328

329

330

331

332

333

334

335

336

337

338

339

340

341

342

343

344

345

346

347

348

349

350

351

352

353

354

355

356

357

358

359

360

361

362

363

364

365

366

367

368

369

370

371

372

373

374

375

376

377

378

379

380

381

382

383

384

385

386

387

388

389

390

391

392

393

394

395

396

397

398

399

400

401

402

403

404

405

406

407

408

409

410

411

412

413

414

415

416

417

418

419

420

421

422

423

424

425

426

427

428

429

430

431

432

433

434

435

436

437

438

439

440

441

442

443

444

445

446

447

448

449

450

451

452

453

454

455

456

457

458

459

460

461

462

463

464

465

466

467

468

469

470

471

472

473

474

475

476

477

478

479

480

481

482

483

484

485

486

487

488

489

490

491

492

493

494

495

496

497

498

499

500

501

502

503

504

505

506

507

508

509

510

511

512

513

514

515

516

517

518

519

520

521

522

523

524

525

526

527

528

529

530

531

532

533

534

535

536

537

538

539

540

541

542

543

544

545

546

547

548

549

550

551

552

553

554

555

556

557

558

559

560

561

562

563

564

565

566

567

568

569

570

571

572

573

574

575

576

577

578

579

580

581

582

583

584

585

586

587

588

589

590

591

592

593

594

595

596

597

598

599

600

601

602

603

604

605

606

607

608

609

610

611

612

613

614

615

616

617

618

619

620

621

622

623

624

625

626

627

628

629

630

631

632

633

634

635

636

637

638

639

640

641

642

643

644

645

646

647

648

649

650

651

652

653

654

655

656

657

658

659

660

661

662

663

664

665

666

667

668

669

670

671

672

673

674

675

676

677

678

679

680

681

682

683

684

685

686

687

688

689

690

691

692

693

694

695

696

697

698

699

700

701

702

703

704

705

706

707

708

709

710

711

712

713

714

715

716

717

718

719

720

721

722

723

724

725

726

727

728

729

730

731

732

733

734

735

736

737

738

739

740

741

742

743

744

745

746

747

748

749

750

751

752

753

754

755

756

757

758

759

760

761

762

763

764

765

766

767

768

769

770

771

772

773

774

775

776

777

778

779

780

781

782

783

784

785

786

787

788

789

790

791

792

793

794

795

796

797

<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"

"http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">

<head>

<title>記憶體管理 — Python 3.7.0 說明文件</title>

title="在 Python 3.7.0 說明文件中搜尋"

href="../_static/opensearch.xml"/>

</head><body>

<ul>

<a href="../genindex.html" title="General Index"

accesskey="I">索引</a></li>

<a href="../py-modindex.html" title="Python 模組索引"

>模組</a> |</li>

<a href="objimpl.html" title="Object Implementation Support"

accesskey="N">下一頁</a> |</li>

<a href="init.html" title="Initialization, Finalization, and Threads"

accesskey="P">上一頁</a> |</li>

<li><img src="../_static/py.png" alt=""

style="vertical-align: middle; margin-top: -1px"/></li>

<li><a href="https://www.python.org/">Python</a> »</li>

<li>

zh_TW

3.7.0

<a href="../index.html">Documentation </a> »

</li>

<li class="nav-item nav-item-1"><a href="index.html" accesskey="U">Python / C API 參考手冊</a> »</li>

</form>

</div>

</li>

</ul>

</div>

<h1>記憶體管理<a class="headerlink" href="#memory-management" title="本標題的永久連結">¶</a></h1>

Memory management in Python involves a private heap containing all Python

objects and data structures. The management of this private heap is ensured

internally by the Python memory manager. The Python memory manager has

different components which deal with various dynamic storage management aspects,

like sharing, segmentation, preallocation or caching.

At the lowest level, a raw memory allocator ensures that there is enough room in

the private heap for storing all Python-related data by interacting with the

memory manager of the operating system. On top of the raw memory allocator,

several object-specific allocators operate on the same heap and implement

distinct memory management policies adapted to the peculiarities of every object

type. For example, integer objects are managed differently within the heap than

strings, tuples or dictionaries because integers imply different storage

requirements and speed/space tradeoffs. The Python memory manager thus delegates

some of the work to the object-specific allocators, but ensures that the latter

operate within the bounds of the private heap.

It is important to understand that the management of the Python heap is

performed by the interpreter itself and that the user has no control over it,

even if they regularly manipulate object pointers to memory blocks inside that

heap. The allocation of heap space for Python objects and other internal

buffers is performed on demand by the Python memory manager through the Python/C

API functions listed in this document.

To avoid memory corruption, extension writers should never try to operate on

Python objects with the functions exported by the C library: <code class="xref c c-func docutils literal notranslate">malloc()</code>,

<code class="xref c c-func docutils literal notranslate">calloc()</code>, <code class="xref c c-func docutils literal notranslate">realloc()</code> and <code class="xref c c-func docutils literal notranslate">free()</code>. This will result in mixed

calls between the C allocator and the Python memory manager with fatal

consequences, because they implement different algorithms and operate on

different heaps. However, one may safely allocate and release memory blocks

with the C library allocator for individual purposes, as shown in the following

example:

<div class="highlight-c notranslate"><div class="highlight"><pre>PyObject *res;

char *buf = (char *) malloc(BUFSIZ); /* for I/O */

if (buf == NULL)

return PyErr_NoMemory();

...Do some I/O operation involving buf...

res = PyBytes_FromString(buf);

free(buf); /* malloc'ed */

return res;

</pre></div>

</div>

In this example, the memory request for the I/O buffer is handled by the C

library allocator. The Python memory manager is involved only in the allocation

of the string object returned as a result.

In most situations, however, it is recommended to allocate memory from the

Python heap specifically because the latter is under control of the Python

memory manager. For example, this is required when the interpreter is extended

with new object types written in C. Another reason for using the Python heap is

the desire to inform the Python memory manager about the memory needs of the

extension module. Even when the requested memory is used exclusively for

internal, highly-specific purposes, delegating all memory requests to the Python

memory manager causes the interpreter to have a more accurate image of its

memory footprint as a whole. Consequently, under certain circumstances, the

Python memory manager may or may not trigger appropriate actions, like garbage

collection, memory compaction or other preventive procedures. Note that by using

the C library allocator as shown in the previous example, the allocated memory

for the I/O buffer escapes completely the Python memory manager.

也參考

The <a class="reference internal" href="../using/cmdline.html#envvar-PYTHONMALLOC"><code class="xref std std-envvar docutils literal notranslate">PYTHONMALLOC</code></a> environment variable can be used to configure

the memory allocators used by Python.

The <a class="reference internal" href="../using/cmdline.html#envvar-PYTHONMALLOCSTATS"><code class="xref std std-envvar docutils literal notranslate">PYTHONMALLOCSTATS</code></a> environment variable can be used to print

statistics of the <a class="reference internal" href="#pymalloc">pymalloc memory allocator</a> every time a

new pymalloc object arena is created, and on shutdown.

</div>

<h2>Raw Memory Interface<a class="headerlink" href="#raw-memory-interface" title="本標題的永久連結">¶</a></h2>

The following function sets are wrappers to the system allocator. These

functions are thread-safe, the <a class="reference internal" href="../glossary.html#term-global-interpreter-lock">GIL</a> does not

need to be held.

The <a class="reference internal" href="#default-memory-allocators">default raw memory allocator</a> uses

the following functions: <code class="xref c c-func docutils literal notranslate">malloc()</code>, <code class="xref c c-func docutils literal notranslate">calloc()</code>, <code class="xref c c-func docutils literal notranslate">realloc()</code>

and <code class="xref c c-func docutils literal notranslate">free()</code>; call <code class="docutils literal notranslate">malloc(1)</code> (or <code class="docutils literal notranslate">calloc(1, 1)</code>) when requesting

zero bytes.

3.4 版新加入.

</div>

void* <code class="descname">PyMem_RawMalloc</code>(size_t n)<a class="headerlink" href="#c.PyMem_RawMalloc" title="本定義的永久連結">¶</a></dt>

<dd>Allocates n bytes and returns a pointer of type <code class="xref c c-type docutils literal notranslate">void*</code> to the

allocated memory, or NULL if the request fails.

Requesting zero bytes returns a distinct non-NULL pointer if possible, as

if <code class="docutils literal notranslate">PyMem_RawMalloc(1)</code> had been called instead. The memory will not have

been initialized in any way.

</dd></dl>

void* <code class="descname">PyMem_RawCalloc</code>(size_t nelem, size_t elsize)<a class="headerlink" href="#c.PyMem_RawCalloc" title="本定義的永久連結">¶</a></dt>

<dd>Allocates nelem elements each whose size in bytes is elsize and returns

a pointer of type <code class="xref c c-type docutils literal notranslate">void*</code> to the allocated memory, or NULL if the

request fails. The memory is initialized to zeros.

Requesting zero elements or elements of size zero bytes returns a distinct

non-NULL pointer if possible, as if <code class="docutils literal notranslate">PyMem_RawCalloc(1, 1)</code> had been

called instead.

3.5 版新加入.

</div>

</dd></dl>

void* <code class="descname">PyMem_RawRealloc</code>(void *p, size_t n)<a class="headerlink" href="#c.PyMem_RawRealloc" title="本定義的永久連結">¶</a></dt>

<dd>Resizes the memory block pointed to by p to n bytes. The contents will

be unchanged to the minimum of the old and the new sizes.

If p is NULL, the call is equivalent to <code class="docutils literal notranslate">PyMem_RawMalloc(n)</code>; else if

n is equal to zero, the memory block is resized but is not freed, and the

returned pointer is non-NULL.

Unless p is NULL, it must have been returned by a previous call to

<a class="reference internal" href="#c.PyMem_RawMalloc" title="PyMem_RawMalloc"><code class="xref c c-func docutils literal notranslate">PyMem_RawMalloc()</code></a>, <a class="reference internal" href="#c.PyMem_RawRealloc" title="PyMem_RawRealloc"><code class="xref c c-func docutils literal notranslate">PyMem_RawRealloc()</code></a> or

<a class="reference internal" href="#c.PyMem_RawCalloc" title="PyMem_RawCalloc"><code class="xref c c-func docutils literal notranslate">PyMem_RawCalloc()</code></a>.

If the request fails, <a class="reference internal" href="#c.PyMem_RawRealloc" title="PyMem_RawRealloc"><code class="xref c c-func docutils literal notranslate">PyMem_RawRealloc()</code></a> returns NULL and p

remains a valid pointer to the previous memory area.

</dd></dl>

void <code class="descname">PyMem_RawFree</code>(void *p)<a class="headerlink" href="#c.PyMem_RawFree" title="本定義的永久連結">¶</a></dt>

<dd>Frees the memory block pointed to by p, which must have been returned by a

previous call to <a class="reference internal" href="#c.PyMem_RawMalloc" title="PyMem_RawMalloc"><code class="xref c c-func docutils literal notranslate">PyMem_RawMalloc()</code></a>, <a class="reference internal" href="#c.PyMem_RawRealloc" title="PyMem_RawRealloc"><code class="xref c c-func docutils literal notranslate">PyMem_RawRealloc()</code></a> or

<a class="reference internal" href="#c.PyMem_RawCalloc" title="PyMem_RawCalloc"><code class="xref c c-func docutils literal notranslate">PyMem_RawCalloc()</code></a>. Otherwise, or if <code class="docutils literal notranslate">PyMem_RawFree(p)</code> has been

called before, undefined behavior occurs.

If p is NULL, no operation is performed.

</dd></dl>

</div>

<h2>Memory Interface<a class="headerlink" href="#memory-interface" title="本標題的永久連結">¶</a></h2>

The following function sets, modeled after the ANSI C standard, but specifying

behavior when requesting zero bytes, are available for allocating and releasing

memory from the Python heap.

The <a class="reference internal" href="#default-memory-allocators">default memory allocator</a> uses the

<a class="reference internal" href="#pymalloc">pymalloc memory allocator</a>.

警告

The <a class="reference internal" href="../glossary.html#term-global-interpreter-lock">GIL</a> must be held when using these

functions.

</div>

3.6 版更變: The default allocator is now pymalloc instead of system <code class="xref c c-func docutils literal notranslate">malloc()</code>.

</div>

void* <code class="descname">PyMem_Malloc</code>(size_t n)<a class="headerlink" href="#c.PyMem_Malloc" title="本定義的永久連結">¶</a></dt>

<dd>Allocates n bytes and returns a pointer of type <code class="xref c c-type docutils literal notranslate">void*</code> to the

allocated memory, or NULL if the request fails.

Requesting zero bytes returns a distinct non-NULL pointer if possible, as

if <code class="docutils literal notranslate">PyMem_Malloc(1)</code> had been called instead. The memory will not have

been initialized in any way.

</dd></dl>

void* <code class="descname">PyMem_Calloc</code>(size_t nelem, size_t elsize)<a class="headerlink" href="#c.PyMem_Calloc" title="本定義的永久連結">¶</a></dt>

<dd>Allocates nelem elements each whose size in bytes is elsize and returns

a pointer of type <code class="xref c c-type docutils literal notranslate">void*</code> to the allocated memory, or NULL if the

request fails. The memory is initialized to zeros.

Requesting zero elements or elements of size zero bytes returns a distinct

non-NULL pointer if possible, as if <code class="docutils literal notranslate">PyMem_Calloc(1, 1)</code> had been called

instead.

3.5 版新加入.

</div>

</dd></dl>

void* <code class="descname">PyMem_Realloc</code>(void *p, size_t n)<a class="headerlink" href="#c.PyMem_Realloc" title="本定義的永久連結">¶</a></dt>

<dd>Resizes the memory block pointed to by p to n bytes. The contents will be

unchanged to the minimum of the old and the new sizes.

If p is NULL, the call is equivalent to <code class="docutils literal notranslate">PyMem_Malloc(n)</code>; else if n

is equal to zero, the memory block is resized but is not freed, and the

returned pointer is non-NULL.

Unless p is NULL, it must have been returned by a previous call to

<a class="reference internal" href="#c.PyMem_Malloc" title="PyMem_Malloc"><code class="xref c c-func docutils literal notranslate">PyMem_Malloc()</code></a>, <a class="reference internal" href="#c.PyMem_Realloc" title="PyMem_Realloc"><code class="xref c c-func docutils literal notranslate">PyMem_Realloc()</code></a> or <a class="reference internal" href="#c.PyMem_Calloc" title="PyMem_Calloc"><code class="xref c c-func docutils literal notranslate">PyMem_Calloc()</code></a>.

If the request fails, <a class="reference internal" href="#c.PyMem_Realloc" title="PyMem_Realloc"><code class="xref c c-func docutils literal notranslate">PyMem_Realloc()</code></a> returns NULL and p remains

a valid pointer to the previous memory area.

</dd></dl>

void <code class="descname">PyMem_Free</code>(void *p)<a class="headerlink" href="#c.PyMem_Free" title="本定義的永久連結">¶</a></dt>

<dd>Frees the memory block pointed to by p, which must have been returned by a

previous call to <a class="reference internal" href="#c.PyMem_Malloc" title="PyMem_Malloc"><code class="xref c c-func docutils literal notranslate">PyMem_Malloc()</code></a>, <a class="reference internal" href="#c.PyMem_Realloc" title="PyMem_Realloc"><code class="xref c c-func docutils literal notranslate">PyMem_Realloc()</code></a> or

<a class="reference internal" href="#c.PyMem_Calloc" title="PyMem_Calloc"><code class="xref c c-func docutils literal notranslate">PyMem_Calloc()</code></a>. Otherwise, or if <code class="docutils literal notranslate">PyMem_Free(p)</code> has been called

before, undefined behavior occurs.

If p is NULL, no operation is performed.

</dd></dl>

The following type-oriented macros are provided for convenience. Note that

TYPE refers to any C type.

TYPE* <code class="descname">PyMem_New</code>(TYPE, size_t n)<a class="headerlink" href="#c.PyMem_New" title="本定義的永久連結">¶</a></dt>

<dd>Same as <a class="reference internal" href="#c.PyMem_Malloc" title="PyMem_Malloc"><code class="xref c c-func docutils literal notranslate">PyMem_Malloc()</code></a>, but allocates <code class="docutils literal notranslate">(n * sizeof(TYPE))</code> bytes of

memory. Returns a pointer cast to <code class="xref c c-type docutils literal notranslate">TYPE*</code>. The memory will not have

been initialized in any way.

</dd></dl>

TYPE* <code class="descname">PyMem_Resize</code>(void *p, TYPE, size_t n)<a class="headerlink" href="#c.PyMem_Resize" title="本定義的永久連結">¶</a></dt>

<dd>Same as <a class="reference internal" href="#c.PyMem_Realloc" title="PyMem_Realloc"><code class="xref c c-func docutils literal notranslate">PyMem_Realloc()</code></a>, but the memory block is resized to <code class="docutils literal notranslate">(n *

sizeof(TYPE))</code> bytes. Returns a pointer cast to <code class="xref c c-type docutils literal notranslate">TYPE*</code>. On return,

p will be a pointer to the new memory area, or NULL in the event of

failure.

This is a C preprocessor macro; p is always reassigned. Save the original

value of p to avoid losing memory when handling errors.

</dd></dl>

void <code class="descname">PyMem_Del</code>(void *p)<a class="headerlink" href="#c.PyMem_Del" title="本定義的永久連結">¶</a></dt>

<dd>Same as <a class="reference internal" href="#c.PyMem_Free" title="PyMem_Free"><code class="xref c c-func docutils literal notranslate">PyMem_Free()</code></a>.

</dd></dl>

In addition, the following macro sets are provided for calling the Python memory

allocator directly, without involving the C API functions listed above. However,

note that their use does not preserve binary compatibility across Python

versions and is therefore deprecated in extension modules.

<li><code class="docutils literal notranslate">PyMem_MALLOC(size)</code></li>

<li><code class="docutils literal notranslate">PyMem_NEW(type, size)</code></li>

<li><code class="docutils literal notranslate">PyMem_REALLOC(ptr, size)</code></li>

<li><code class="docutils literal notranslate">PyMem_RESIZE(ptr, type, size)</code></li>

<li><code class="docutils literal notranslate">PyMem_FREE(ptr)</code></li>

<li><code class="docutils literal notranslate">PyMem_DEL(ptr)</code></li>

</ul>

</div>

<h2>Object allocators<a class="headerlink" href="#object-allocators" title="本標題的永久連結">¶</a></h2>

The following function sets, modeled after the ANSI C standard, but specifying

behavior when requesting zero bytes, are available for allocating and releasing

memory from the Python heap.

The <a class="reference internal" href="#default-memory-allocators">default object allocator</a> uses the

<a class="reference internal" href="#pymalloc">pymalloc memory allocator</a>.

警告

The <a class="reference internal" href="../glossary.html#term-global-interpreter-lock">GIL</a> must be held when using these

functions.

</div>

void* <code class="descname">PyObject_Malloc</code>(size_t n)<a class="headerlink" href="#c.PyObject_Malloc" title="本定義的永久連結">¶</a></dt>

<dd>Allocates n bytes and returns a pointer of type <code class="xref c c-type docutils literal notranslate">void*</code> to the

allocated memory, or NULL if the request fails.

Requesting zero bytes returns a distinct non-NULL pointer if possible, as

if <code class="docutils literal notranslate">PyObject_Malloc(1)</code> had been called instead. The memory will not have

been initialized in any way.

</dd></dl>

void* <code class="descname">PyObject_Calloc</code>(size_t nelem, size_t elsize)<a class="headerlink" href="#c.PyObject_Calloc" title="本定義的永久連結">¶</a></dt>

<dd>Allocates nelem elements each whose size in bytes is elsize and returns

a pointer of type <code class="xref c c-type docutils literal notranslate">void*</code> to the allocated memory, or NULL if the

request fails. The memory is initialized to zeros.

Requesting zero elements or elements of size zero bytes returns a distinct

non-NULL pointer if possible, as if <code class="docutils literal notranslate">PyObject_Calloc(1, 1)</code> had been called

instead.

3.5 版新加入.

</div>

</dd></dl>

void* <code class="descname">PyObject_Realloc</code>(void *p, size_t n)<a class="headerlink" href="#c.PyObject_Realloc" title="本定義的永久連結">¶</a></dt>

<dd>Resizes the memory block pointed to by p to n bytes. The contents will be

unchanged to the minimum of the old and the new sizes.

If p is NULL, the call is equivalent to <code class="docutils literal notranslate">PyObject_Malloc(n)</code>; else if n

is equal to zero, the memory block is resized but is not freed, and the

returned pointer is non-NULL.

Unless p is NULL, it must have been returned by a previous call to

<a class="reference internal" href="#c.PyObject_Malloc" title="PyObject_Malloc"><code class="xref c c-func docutils literal notranslate">PyObject_Malloc()</code></a>, <a class="reference internal" href="#c.PyObject_Realloc" title="PyObject_Realloc"><code class="xref c c-func docutils literal notranslate">PyObject_Realloc()</code></a> or <a class="reference internal" href="#c.PyObject_Calloc" title="PyObject_Calloc"><code class="xref c c-func docutils literal notranslate">PyObject_Calloc()</code></a>.

If the request fails, <a class="reference internal" href="#c.PyObject_Realloc" title="PyObject_Realloc"><code class="xref c c-func docutils literal notranslate">PyObject_Realloc()</code></a> returns NULL and p remains

a valid pointer to the previous memory area.

</dd></dl>

void <code class="descname">PyObject_Free</code>(void *p)<a class="headerlink" href="#c.PyObject_Free" title="本定義的永久連結">¶</a></dt>

<dd>Frees the memory block pointed to by p, which must have been returned by a

previous call to <a class="reference internal" href="#c.PyObject_Malloc" title="PyObject_Malloc"><code class="xref c c-func docutils literal notranslate">PyObject_Malloc()</code></a>, <a class="reference internal" href="#c.PyObject_Realloc" title="PyObject_Realloc"><code class="xref c c-func docutils literal notranslate">PyObject_Realloc()</code></a> or

<a class="reference internal" href="#c.PyObject_Calloc" title="PyObject_Calloc"><code class="xref c c-func docutils literal notranslate">PyObject_Calloc()</code></a>. Otherwise, or if <code class="docutils literal notranslate">PyObject_Free(p)</code> has been called

before, undefined behavior occurs.

If p is NULL, no operation is performed.

</dd></dl>

</div>

<h2>Default Memory Allocators<a class="headerlink" href="#default-memory-allocators" title="本標題的永久連結">¶</a></h2>

Default memory allocators:

</colgroup>

<tr class="row-odd"><th class="head">Configuration</th>

<th class="head">PyMem_RawMalloc</th>

<th class="head">PyMem_Malloc</th>

<th class="head">PyObject_Malloc</th>

</tr>

</thead>

<tr class="row-even"><td>Release build</td>

<td><code class="docutils literal notranslate">"pymalloc"</code></td>

<td><code class="docutils literal notranslate">malloc</code></td>

<td><code class="docutils literal notranslate">pymalloc</code></td>

</tr>

<tr class="row-odd"><td>Debug build</td>

<td><code class="docutils literal notranslate">"pymalloc_debug"</code></td>

<td><code class="docutils literal notranslate">malloc</code> + debug</td>

<td><code class="docutils literal notranslate">pymalloc</code> + debug</td>

</tr>

<tr class="row-even"><td>Release build, without pymalloc</td>

<td><code class="docutils literal notranslate">"malloc"</code></td>

<td><code class="docutils literal notranslate">malloc</code></td>

</tr>

<tr class="row-odd"><td>Release build, without pymalloc</td>

<td><code class="docutils literal notranslate">"malloc_debug"</code></td>

<td><code class="docutils literal notranslate">malloc</code> + debug</td>

</tr>

</tbody>

</table>

Legend:

<li>Name: value for <a class="reference internal" href="../using/cmdline.html#envvar-PYTHONMALLOC"><code class="xref std std-envvar docutils literal notranslate">PYTHONMALLOC</code></a> environment variable</li>

<li><code class="docutils literal notranslate">malloc</code>: system allocators from the standard C library, C functions:

<code class="xref c c-func docutils literal notranslate">malloc()</code>, <code class="xref c c-func docutils literal notranslate">calloc()</code>, <code class="xref c c-func docutils literal notranslate">realloc()</code> and <code class="xref c c-func docutils literal notranslate">free()</code></li>

<li><code class="docutils literal notranslate">pymalloc</code>: <a class="reference internal" href="#pymalloc">pymalloc memory allocator</a></li>

<li>「+ debug」: with debug hooks installed by <a class="reference internal" href="#c.PyMem_SetupDebugHooks" title="PyMem_SetupDebugHooks"><code class="xref c c-func docutils literal notranslate">PyMem_SetupDebugHooks()</code></a></li>

</ul>

</div>

<h2>Customize Memory Allocators<a class="headerlink" href="#customize-memory-allocators" title="本標題的永久連結">¶</a></h2>

3.4 版新加入.

</div>

<code class="descname">PyMemAllocatorEx</code><a class="headerlink" href="#c.PyMemAllocatorEx" title="本定義的永久連結">¶</a></dt>

<dd>Structure used to describe a memory block allocator. The structure has

four fields:

</colgroup>

<tr class="row-odd"><th class="head">Field</th>

<th class="head">Meaning</th>

</tr>

</thead>

<td>user context passed as first argument</td>

</tr>

<tr class="row-odd"><td><code class="docutils literal notranslate">void* malloc(void *ctx, size_t size)</code></td>

<td>allocate a memory block</td>

</tr>

<tr class="row-even"><td><code class="docutils literal notranslate">void* calloc(void *ctx, size_t nelem, size_t elsize)</code></td>

<td>allocate a memory block initialized

with zeros</td>

</tr>

<tr class="row-odd"><td><code class="docutils literal notranslate">void* realloc(void *ctx, void *ptr, size_t new_size)</code></td>

<td>allocate or resize a memory block</td>

</tr>

<td>free a memory block</td>

</tr>

</tbody>

</table>

3.5 版更變: The <code class="xref c c-type docutils literal notranslate">PyMemAllocator</code> structure was renamed to

<a class="reference internal" href="#c.PyMemAllocatorEx" title="PyMemAllocatorEx"><code class="xref c c-type docutils literal notranslate">PyMemAllocatorEx</code></a> and a new <code class="docutils literal notranslate">calloc</code> field was added.

</div>

</dd></dl>

<code class="descname">PyMemAllocatorDomain</code><a class="headerlink" href="#c.PyMemAllocatorDomain" title="本定義的永久連結">¶</a></dt>

<dd>Enum used to identify an allocator domain. Domains:

<code class="descname">PYMEM_DOMAIN_RAW</code><a class="headerlink" href="#c.PYMEM_DOMAIN_RAW" title="本定義的永久連結">¶</a></dt>

<dd>Functions:

<li><a class="reference internal" href="#c.PyMem_RawMalloc" title="PyMem_RawMalloc"><code class="xref c c-func docutils literal notranslate">PyMem_RawMalloc()</code></a></li>

<li><a class="reference internal" href="#c.PyMem_RawRealloc" title="PyMem_RawRealloc"><code class="xref c c-func docutils literal notranslate">PyMem_RawRealloc()</code></a></li>

<li><a class="reference internal" href="#c.PyMem_RawCalloc" title="PyMem_RawCalloc"><code class="xref c c-func docutils literal notranslate">PyMem_RawCalloc()</code></a></li>

<li><a class="reference internal" href="#c.PyMem_RawFree" title="PyMem_RawFree"><code class="xref c c-func docutils literal notranslate">PyMem_RawFree()</code></a></li>

</ul>

</dd></dl>

<code class="descname">PYMEM_DOMAIN_MEM</code><a class="headerlink" href="#c.PYMEM_DOMAIN_MEM" title="本定義的永久連結">¶</a></dt>

<dd>Functions:

<li><a class="reference internal" href="#c.PyMem_Malloc" title="PyMem_Malloc"><code class="xref c c-func docutils literal notranslate">PyMem_Malloc()</code></a>,</li>

<li><a class="reference internal" href="#c.PyMem_Realloc" title="PyMem_Realloc"><code class="xref c c-func docutils literal notranslate">PyMem_Realloc()</code></a></li>

<li><a class="reference internal" href="#c.PyMem_Calloc" title="PyMem_Calloc"><code class="xref c c-func docutils literal notranslate">PyMem_Calloc()</code></a></li>

<li><a class="reference internal" href="#c.PyMem_Free" title="PyMem_Free"><code class="xref c c-func docutils literal notranslate">PyMem_Free()</code></a></li>

</ul>

</dd></dl>

<code class="descname">PYMEM_DOMAIN_OBJ</code><a class="headerlink" href="#c.PYMEM_DOMAIN_OBJ" title="本定義的永久連結">¶</a></dt>

<dd>Functions:

<li><a class="reference internal" href="#c.PyObject_Malloc" title="PyObject_Malloc"><code class="xref c c-func docutils literal notranslate">PyObject_Malloc()</code></a></li>

<li><a class="reference internal" href="#c.PyObject_Realloc" title="PyObject_Realloc"><code class="xref c c-func docutils literal notranslate">PyObject_Realloc()</code></a></li>

<li><a class="reference internal" href="#c.PyObject_Calloc" title="PyObject_Calloc"><code class="xref c c-func docutils literal notranslate">PyObject_Calloc()</code></a></li>

<li><a class="reference internal" href="#c.PyObject_Free" title="PyObject_Free"><code class="xref c c-func docutils literal notranslate">PyObject_Free()</code></a></li>

</ul>

</dd></dl>

void <code class="descname">PyMem_GetAllocator</code>(<a class="reference internal" href="#c.PyMemAllocatorDomain" title="PyMemAllocatorDomain">PyMemAllocatorDomain</a> domain, <a class="reference internal" href="#c.PyMemAllocatorEx" title="PyMemAllocatorEx">PyMemAllocatorEx</a> *allocator)<a class="headerlink" href="#c.PyMem_GetAllocator" title="本定義的永久連結">¶</a></dt>

<dd>Get the memory block allocator of the specified domain.

</dd></dl>

void <code class="descname">PyMem_SetAllocator</code>(<a class="reference internal" href="#c.PyMemAllocatorDomain" title="PyMemAllocatorDomain">PyMemAllocatorDomain</a> domain, <a class="reference internal" href="#c.PyMemAllocatorEx" title="PyMemAllocatorEx">PyMemAllocatorEx</a> *allocator)<a class="headerlink" href="#c.PyMem_SetAllocator" title="本定義的永久連結">¶</a></dt>

<dd>Set the memory block allocator of the specified domain.

The new allocator must return a distinct non-NULL pointer when requesting

zero bytes.

For the <a class="reference internal" href="#c.PYMEM_DOMAIN_RAW" title="PYMEM_DOMAIN_RAW"><code class="xref c c-data docutils literal notranslate">PYMEM_DOMAIN_RAW</code></a> domain, the allocator must be

thread-safe: the <a class="reference internal" href="../glossary.html#term-global-interpreter-lock">GIL</a> is not held when the

allocator is called.

If the new allocator is not a hook (does not call the previous allocator),

the <a class="reference internal" href="#c.PyMem_SetupDebugHooks" title="PyMem_SetupDebugHooks"><code class="xref c c-func docutils literal notranslate">PyMem_SetupDebugHooks()</code></a> function must be called to reinstall the

debug hooks on top on the new allocator.

</dd></dl>

void <code class="descname">PyMem_SetupDebugHooks</code>(void)<a class="headerlink" href="#c.PyMem_SetupDebugHooks" title="本定義的永久連結">¶</a></dt>

<dd>Setup hooks to detect bugs in the Python memory allocator functions.

Newly allocated memory is filled with the byte <code class="docutils literal notranslate">0xCB</code>, freed memory is

filled with the byte <code class="docutils literal notranslate">0xDB</code>.

Runtime checks:

<li>Detect API violations, ex: <a class="reference internal" href="#c.PyObject_Free" title="PyObject_Free"><code class="xref c c-func docutils literal notranslate">PyObject_Free()</code></a> called on a buffer

allocated by <a class="reference internal" href="#c.PyMem_Malloc" title="PyMem_Malloc"><code class="xref c c-func docutils literal notranslate">PyMem_Malloc()</code></a></li>

<li>Detect write before the start of the buffer (buffer underflow)</li>

<li>Detect write after the end of the buffer (buffer overflow)</li>

<li>Check that the <a class="reference internal" href="../glossary.html#term-global-interpreter-lock">GIL</a> is held when

allocator functions of <a class="reference internal" href="#c.PYMEM_DOMAIN_OBJ" title="PYMEM_DOMAIN_OBJ"><code class="xref c c-data docutils literal notranslate">PYMEM_DOMAIN_OBJ</code></a> (ex:

</ul>

On error, the debug hooks use the <a class="reference internal" href="../library/tracemalloc.html#module-tracemalloc" title="tracemalloc: Trace memory allocations."><code class="xref py py-mod docutils literal notranslate">tracemalloc</code></a> module to get the

traceback where a memory block was allocated. The traceback is only

displayed if <a class="reference internal" href="../library/tracemalloc.html#module-tracemalloc" title="tracemalloc: Trace memory allocations."><code class="xref py py-mod docutils literal notranslate">tracemalloc</code></a> is tracing Python memory allocations and the

memory block was traced.

These hooks are <a class="reference internal" href="#default-memory-allocators">installed by default</a> if

Python is compiled in debug

mode. The <a class="reference internal" href="../using/cmdline.html#envvar-PYTHONMALLOC"><code class="xref std std-envvar docutils literal notranslate">PYTHONMALLOC</code></a> environment variable can be used to install

debug hooks on a Python compiled in release mode.

3.6 版更變: This function now also works on Python compiled in release mode.

On error, the debug hooks now use <a class="reference internal" href="../library/tracemalloc.html#module-tracemalloc" title="tracemalloc: Trace memory allocations."><code class="xref py py-mod docutils literal notranslate">tracemalloc</code></a> to get the traceback

where a memory block was allocated. The debug hooks now also check

if the GIL is held when functions of <a class="reference internal" href="#c.PYMEM_DOMAIN_OBJ" title="PYMEM_DOMAIN_OBJ"><code class="xref c c-data docutils literal notranslate">PYMEM_DOMAIN_OBJ</code></a> and

</div>

</dd></dl>

</div>

<h2>The pymalloc allocator<a class="headerlink" href="#the-pymalloc-allocator" title="本標題的永久連結">¶</a></h2>

Python has a pymalloc allocator optimized for small objects (smaller or equal

to 512 bytes) with a short lifetime. It uses memory mappings called 「arenas」

with a fixed size of 256 KiB. It falls back to <a class="reference internal" href="#c.PyMem_RawMalloc" title="PyMem_RawMalloc"><code class="xref c c-func docutils literal notranslate">PyMem_RawMalloc()</code></a> and

<a class="reference internal" href="#c.PyMem_RawRealloc" title="PyMem_RawRealloc"><code class="xref c c-func docutils literal notranslate">PyMem_RawRealloc()</code></a> for allocations larger than 512 bytes.

pymalloc is the <a class="reference internal" href="#default-memory-allocators">default allocator</a> of the

<a class="reference internal" href="#c.PYMEM_DOMAIN_MEM" title="PYMEM_DOMAIN_MEM"><code class="xref c c-data docutils literal notranslate">PYMEM_DOMAIN_MEM</code></a> (ex: <a class="reference internal" href="#c.PyMem_Malloc" title="PyMem_Malloc"><code class="xref c c-func docutils literal notranslate">PyMem_Malloc()</code></a>) and

<a class="reference internal" href="#c.PYMEM_DOMAIN_OBJ" title="PYMEM_DOMAIN_OBJ"><code class="xref c c-data docutils literal notranslate">PYMEM_DOMAIN_OBJ</code></a> (ex: <a class="reference internal" href="#c.PyObject_Malloc" title="PyObject_Malloc"><code class="xref c c-func docutils literal notranslate">PyObject_Malloc()</code></a>) domains.

The arena allocator uses the following functions:

<li><code class="xref c c-func docutils literal notranslate">VirtualAlloc()</code> and <code class="xref c c-func docutils literal notranslate">VirtualFree()</code> on Windows,</li>

<li><code class="xref c c-func docutils literal notranslate">mmap()</code> and <code class="xref c c-func docutils literal notranslate">munmap()</code> if available,</li>

<li><code class="xref c c-func docutils literal notranslate">malloc()</code> and <code class="xref c c-func docutils literal notranslate">free()</code> otherwise.</li>

</ul>

<h3>Customize pymalloc Arena Allocator<a class="headerlink" href="#customize-pymalloc-arena-allocator" title="本標題的永久連結">¶</a></h3>

3.4 版新加入.

</div>

<code class="descname">PyObjectArenaAllocator</code><a class="headerlink" href="#c.PyObjectArenaAllocator" title="本定義的永久連結">¶</a></dt>

<dd>Structure used to describe an arena allocator. The structure has

three fields:

</colgroup>

<tr class="row-odd"><th class="head">Field</th>

<th class="head">Meaning</th>

</tr>

</thead>

<td>user context passed as first argument</td>

</tr>

<tr class="row-odd"><td><code class="docutils literal notranslate">void* alloc(void *ctx, size_t size)</code></td>

<td>allocate an arena of size bytes</td>

</tr>

<td>free an arena</td>

</tr>

</tbody>

</table>

</dd></dl>

<code class="descname">PyObject_GetArenaAllocator</code>(<a class="reference internal" href="#c.PyObjectArenaAllocator" title="PyObjectArenaAllocator">PyObjectArenaAllocator</a> *allocator)<a class="headerlink" href="#c.PyObject_GetArenaAllocator" title="本定義的永久連結">¶</a></dt>

<dd>Get the arena allocator.

</dd></dl>

<code class="descname">PyObject_SetArenaAllocator</code>(<a class="reference internal" href="#c.PyObjectArenaAllocator" title="PyObjectArenaAllocator">PyObjectArenaAllocator</a> *allocator)<a class="headerlink" href="#c.PyObject_SetArenaAllocator" title="本定義的永久連結">¶</a></dt>

<dd>Set the arena allocator.

</dd></dl>

</div>

<h2>tracemalloc C API<a class="headerlink" href="#tracemalloc-c-api" title="本標題的永久連結">¶</a></h2>

3.7 版新加入.

</div>

<h2>Examples<a class="headerlink" href="#examples" title="本標題的永久連結">¶</a></h2>

Here is the example from section <a class="reference internal" href="#memoryoverview">總覽</a>, rewritten so that the

I/O buffer is allocated from the Python heap by using the first function set:

<div class="highlight-c notranslate"><div class="highlight"><pre>PyObject *res;

char *buf = (char *) PyMem_Malloc(BUFSIZ); /* for I/O */

if (buf == NULL)

return PyErr_NoMemory();

/* ...Do some I/O operation involving buf... */

res = PyBytes_FromString(buf);

PyMem_Free(buf); /* allocated with PyMem_Malloc */

return res;

</pre></div>

</div>

The same code using the type-oriented function set:

<div class="highlight-c notranslate"><div class="highlight"><pre>PyObject *res;

char *buf = PyMem_New(char, BUFSIZ); /* for I/O */

if (buf == NULL)

return PyErr_NoMemory();

/* ...Do some I/O operation involving buf... */

res = PyBytes_FromString(buf);

PyMem_Del(buf); /* allocated with PyMem_New */

return res;

</pre></div>

</div>

Note that in the two examples above, the buffer is always manipulated via

functions belonging to the same set. Indeed, it is required to use the same

memory API family for a given memory block, so that the risk of mixing different

allocators is reduced to a minimum. The following code sequence contains two

errors, one of which is labeled as fatal because it mixes two different

allocators operating on different heaps.

<div class="highlight-c notranslate"><div class="highlight"><pre>char *buf1 = PyMem_New(char, BUFSIZ);

char *buf2 = (char *) malloc(BUFSIZ);

char *buf3 = (char *) PyMem_Malloc(BUFSIZ);

...

PyMem_Del(buf3); /* Wrong -- should be PyMem_Free() */

free(buf2); /* Right -- allocated via malloc() */

free(buf1); /* Fatal -- should be PyMem_Del() */

</pre></div>

</div>

In addition to the functions aimed at handling raw memory blocks from the Python

heap, objects in Python are allocated and released with <a class="reference internal" href="allocation.html#c.PyObject_New" title="PyObject_New"><code class="xref c c-func docutils literal notranslate">PyObject_New()</code></a>,

<a class="reference internal" href="allocation.html#c.PyObject_NewVar" title="PyObject_NewVar"><code class="xref c c-func docutils literal notranslate">PyObject_NewVar()</code></a> and <a class="reference internal" href="allocation.html#c.PyObject_Del" title="PyObject_Del"><code class="xref c c-func docutils literal notranslate">PyObject_Del()</code></a>.

These will be explained in the next chapter on defining and implementing new

object types in C.

</div>

<ul>

<li><a class="reference internal" href="#">記憶體管理</a><ul>

<li><a class="reference internal" href="#raw-memory-interface">Raw Memory Interface</a></li>

<li><a class="reference internal" href="#memory-interface">Memory Interface</a></li>

<li><a class="reference internal" href="#object-allocators">Object allocators</a></li>

<li><a class="reference internal" href="#default-memory-allocators">Default Memory Allocators</a></li>

<li><a class="reference internal" href="#customize-memory-allocators">Customize Memory Allocators</a></li>

<li><a class="reference internal" href="#the-pymalloc-allocator">The pymalloc allocator</a><ul>

<li><a class="reference internal" href="#customize-pymalloc-arena-allocator">Customize pymalloc Arena Allocator</a></li>

</ul>

</li>

<li><a class="reference internal" href="#tracemalloc-c-api">tracemalloc C API</a></li>

<li><a class="reference internal" href="#examples">Examples</a></li>

</ul>

</li>

</ul>

<a href="init.html"

title="上一章">Initialization, Finalization, and Threads</a>

<a href="objimpl.html"

title="下一章">Object Implementation Support</a>

<li><a href="../bugs.html">Report a Bug</a></li>

<li>

<a href="https://github.com/python/cpython/blob/3.7/Doc/c-api/memory.rst"

rel="nofollow">Show Source

</a>

</li>

</ul>

</div>

</div>

<ul>

<a href="../genindex.html" title="General Index"

>索引</a></li>

<a href="../py-modindex.html" title="Python 模組索引"

>模組</a> |</li>

<a href="objimpl.html" title="Object Implementation Support"

>下一頁</a> |</li>

<a href="init.html" title="Initialization, Finalization, and Threads"

>上一頁</a> |</li>

<li><img src="../_static/py.png" alt=""

style="vertical-align: middle; margin-top: -1px"/></li>

<li><a href="https://www.python.org/">Python</a> »</li>

<li>

zh_TW

3.7.0

<a href="../index.html">Documentation </a> »

</li>

<li class="nav-item nav-item-1"><a href="index.html" >Python / C API 參考手冊</a> »</li>

</form>

</div>

</li>

</ul>

</div>

The Python Software Foundation is a non-profit corporation.

<a href="https://www.python.org/psf/donations/">Please donate.</a>

Last updated on 8月 22, 2018.

<a href="../bugs.html">Found a bug</a>?

Created using <a href="http://sphinx.pocoo.org/">Sphinx</a> 1.7.7.

</div>

</body>

</html>

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

FilesExpand file tree

memory.html

Latest commit

History

memory.html

File metadata and controls