Based on kernel version 4.16.1. Page generated on 2018-04-09 11:53 EST.
1 Please note that the "What is RCU?" LWN series is an excellent place 2 to start learning about RCU: 3 4 1. What is RCU, Fundamentally? http://lwn.net/Articles/262464/ 5 2. What is RCU? Part 2: Usage http://lwn.net/Articles/263130/ 6 3. RCU part 3: the RCU API http://lwn.net/Articles/264090/ 7 4. The RCU API, 2010 Edition http://lwn.net/Articles/418853/ 8 2010 Big API Table http://lwn.net/Articles/419086/ 9 5. The RCU API, 2014 Edition http://lwn.net/Articles/609904/ 10 2014 Big API Table http://lwn.net/Articles/609973/ 11 12 13 What is RCU? 14 15 RCU is a synchronization mechanism that was added to the Linux kernel 16 during the 2.5 development effort that is optimized for read-mostly 17 situations. Although RCU is actually quite simple once you understand it, 18 getting there can sometimes be a challenge. Part of the problem is that 19 most of the past descriptions of RCU have been written with the mistaken 20 assumption that there is "one true way" to describe RCU. Instead, 21 the experience has been that different people must take different paths 22 to arrive at an understanding of RCU. This document provides several 23 different paths, as follows: 24 25 1. RCU OVERVIEW 26 2. WHAT IS RCU'S CORE API? 27 3. WHAT ARE SOME EXAMPLE USES OF CORE RCU API? 28 4. WHAT IF MY UPDATING THREAD CANNOT BLOCK? 29 5. WHAT ARE SOME SIMPLE IMPLEMENTATIONS OF RCU? 30 6. ANALOGY WITH READER-WRITER LOCKING 31 7. FULL LIST OF RCU APIs 32 8. ANSWERS TO QUICK QUIZZES 33 34 People who prefer starting with a conceptual overview should focus on 35 Section 1, though most readers will profit by reading this section at 36 some point. People who prefer to start with an API that they can then 37 experiment with should focus on Section 2. People who prefer to start 38 with example uses should focus on Sections 3 and 4. People who need to 39 understand the RCU implementation should focus on Section 5, then dive 40 into the kernel source code. People who reason best by analogy should 41 focus on Section 6. Section 7 serves as an index to the docbook API 42 documentation, and Section 8 is the traditional answer key. 43 44 So, start with the section that makes the most sense to you and your 45 preferred method of learning. If you need to know everything about 46 everything, feel free to read the whole thing -- but if you are really 47 that type of person, you have perused the source code and will therefore 48 never need this document anyway. ;-) 49 50 51 1. RCU OVERVIEW 52 53 The basic idea behind RCU is to split updates into "removal" and 54 "reclamation" phases. The removal phase removes references to data items 55 within a data structure (possibly by replacing them with references to 56 new versions of these data items), and can run concurrently with readers. 57 The reason that it is safe to run the removal phase concurrently with 58 readers is the semantics of modern CPUs guarantee that readers will see 59 either the old or the new version of the data structure rather than a 60 partially updated reference. The reclamation phase does the work of reclaiming 61 (e.g., freeing) the data items removed from the data structure during the 62 removal phase. Because reclaiming data items can disrupt any readers 63 concurrently referencing those data items, the reclamation phase must 64 not start until readers no longer hold references to those data items. 65 66 Splitting the update into removal and reclamation phases permits the 67 updater to perform the removal phase immediately, and to defer the 68 reclamation phase until all readers active during the removal phase have 69 completed, either by blocking until they finish or by registering a 70 callback that is invoked after they finish. Only readers that are active 71 during the removal phase need be considered, because any reader starting 72 after the removal phase will be unable to gain a reference to the removed 73 data items, and therefore cannot be disrupted by the reclamation phase. 74 75 So the typical RCU update sequence goes something like the following: 76 77 a. Remove pointers to a data structure, so that subsequent 78 readers cannot gain a reference to it. 79 80 b. Wait for all previous readers to complete their RCU read-side 81 critical sections. 82 83 c. At this point, there cannot be any readers who hold references 84 to the data structure, so it now may safely be reclaimed 85 (e.g., kfree()d). 86 87 Step (b) above is the key idea underlying RCU's deferred destruction. 88 The ability to wait until all readers are done allows RCU readers to 89 use much lighter-weight synchronization, in some cases, absolutely no 90 synchronization at all. In contrast, in more conventional lock-based 91 schemes, readers must use heavy-weight synchronization in order to 92 prevent an updater from deleting the data structure out from under them. 93 This is because lock-based updaters typically update data items in place, 94 and must therefore exclude readers. In contrast, RCU-based updaters 95 typically take advantage of the fact that writes to single aligned 96 pointers are atomic on modern CPUs, allowing atomic insertion, removal, 97 and replacement of data items in a linked structure without disrupting 98 readers. Concurrent RCU readers can then continue accessing the old 99 versions, and can dispense with the atomic operations, memory barriers, 100 and communications cache misses that are so expensive on present-day 101 SMP computer systems, even in absence of lock contention. 102 103 In the three-step procedure shown above, the updater is performing both 104 the removal and the reclamation step, but it is often helpful for an 105 entirely different thread to do the reclamation, as is in fact the case 106 in the Linux kernel's directory-entry cache (dcache). Even if the same 107 thread performs both the update step (step (a) above) and the reclamation 108 step (step (c) above), it is often helpful to think of them separately. 109 For example, RCU readers and updaters need not communicate at all, 110 but RCU provides implicit low-overhead communication between readers 111 and reclaimers, namely, in step (b) above. 112 113 So how the heck can a reclaimer tell when a reader is done, given 114 that readers are not doing any sort of synchronization operations??? 115 Read on to learn about how RCU's API makes this easy. 116 117 118 2. WHAT IS RCU'S CORE API? 119 120 The core RCU API is quite small: 121 122 a. rcu_read_lock() 123 b. rcu_read_unlock() 124 c. synchronize_rcu() / call_rcu() 125 d. rcu_assign_pointer() 126 e. rcu_dereference() 127 128 There are many other members of the RCU API, but the rest can be 129 expressed in terms of these five, though most implementations instead 130 express synchronize_rcu() in terms of the call_rcu() callback API. 131 132 The five core RCU APIs are described below, the other 18 will be enumerated 133 later. See the kernel docbook documentation for more info, or look directly 134 at the function header comments. 135 136 rcu_read_lock() 137 138 void rcu_read_lock(void); 139 140 Used by a reader to inform the reclaimer that the reader is 141 entering an RCU read-side critical section. It is illegal 142 to block while in an RCU read-side critical section, though 143 kernels built with CONFIG_PREEMPT_RCU can preempt RCU 144 read-side critical sections. Any RCU-protected data structure 145 accessed during an RCU read-side critical section is guaranteed to 146 remain unreclaimed for the full duration of that critical section. 147 Reference counts may be used in conjunction with RCU to maintain 148 longer-term references to data structures. 149 150 rcu_read_unlock() 151 152 void rcu_read_unlock(void); 153 154 Used by a reader to inform the reclaimer that the reader is 155 exiting an RCU read-side critical section. Note that RCU 156 read-side critical sections may be nested and/or overlapping. 157 158 synchronize_rcu() 159 160 void synchronize_rcu(void); 161 162 Marks the end of updater code and the beginning of reclaimer 163 code. It does this by blocking until all pre-existing RCU 164 read-side critical sections on all CPUs have completed. 165 Note that synchronize_rcu() will -not- necessarily wait for 166 any subsequent RCU read-side critical sections to complete. 167 For example, consider the following sequence of events: 168 169 CPU 0 CPU 1 CPU 2 170 ----------------- ------------------------- --------------- 171 1. rcu_read_lock() 172 2. enters synchronize_rcu() 173 3. rcu_read_lock() 174 4. rcu_read_unlock() 175 5. exits synchronize_rcu() 176 6. rcu_read_unlock() 177 178 To reiterate, synchronize_rcu() waits only for ongoing RCU 179 read-side critical sections to complete, not necessarily for 180 any that begin after synchronize_rcu() is invoked. 181 182 Of course, synchronize_rcu() does not necessarily return 183 -immediately- after the last pre-existing RCU read-side critical 184 section completes. For one thing, there might well be scheduling 185 delays. For another thing, many RCU implementations process 186 requests in batches in order to improve efficiencies, which can 187 further delay synchronize_rcu(). 188 189 Since synchronize_rcu() is the API that must figure out when 190 readers are done, its implementation is key to RCU. For RCU 191 to be useful in all but the most read-intensive situations, 192 synchronize_rcu()'s overhead must also be quite small. 193 194 The call_rcu() API is a callback form of synchronize_rcu(), 195 and is described in more detail in a later section. Instead of 196 blocking, it registers a function and argument which are invoked 197 after all ongoing RCU read-side critical sections have completed. 198 This callback variant is particularly useful in situations where 199 it is illegal to block or where update-side performance is 200 critically important. 201 202 However, the call_rcu() API should not be used lightly, as use 203 of the synchronize_rcu() API generally results in simpler code. 204 In addition, the synchronize_rcu() API has the nice property 205 of automatically limiting update rate should grace periods 206 be delayed. This property results in system resilience in face 207 of denial-of-service attacks. Code using call_rcu() should limit 208 update rate in order to gain this same sort of resilience. See 209 checklist.txt for some approaches to limiting the update rate. 210 211 rcu_assign_pointer() 212 213 typeof(p) rcu_assign_pointer(p, typeof(p) v); 214 215 Yes, rcu_assign_pointer() -is- implemented as a macro, though it 216 would be cool to be able to declare a function in this manner. 217 (Compiler experts will no doubt disagree.) 218 219 The updater uses this function to assign a new value to an 220 RCU-protected pointer, in order to safely communicate the change 221 in value from the updater to the reader. This function returns 222 the new value, and also executes any memory-barrier instructions 223 required for a given CPU architecture. 224 225 Perhaps just as important, it serves to document (1) which 226 pointers are protected by RCU and (2) the point at which a 227 given structure becomes accessible to other CPUs. That said, 228 rcu_assign_pointer() is most frequently used indirectly, via 229 the _rcu list-manipulation primitives such as list_add_rcu(). 230 231 rcu_dereference() 232 233 typeof(p) rcu_dereference(p); 234 235 Like rcu_assign_pointer(), rcu_dereference() must be implemented 236 as a macro. 237 238 The reader uses rcu_dereference() to fetch an RCU-protected 239 pointer, which returns a value that may then be safely 240 dereferenced. Note that rcu_dereference() does not actually 241 dereference the pointer, instead, it protects the pointer for 242 later dereferencing. It also executes any needed memory-barrier 243 instructions for a given CPU architecture. Currently, only Alpha 244 needs memory barriers within rcu_dereference() -- on other CPUs, 245 it compiles to nothing, not even a compiler directive. 246 247 Common coding practice uses rcu_dereference() to copy an 248 RCU-protected pointer to a local variable, then dereferences 249 this local variable, for example as follows: 250 251 p = rcu_dereference(head.next); 252 return p->data; 253 254 However, in this case, one could just as easily combine these 255 into one statement: 256 257 return rcu_dereference(head.next)->data; 258 259 If you are going to be fetching multiple fields from the 260 RCU-protected structure, using the local variable is of 261 course preferred. Repeated rcu_dereference() calls look 262 ugly, do not guarantee that the same pointer will be returned 263 if an update happened while in the critical section, and incur 264 unnecessary overhead on Alpha CPUs. 265 266 Note that the value returned by rcu_dereference() is valid 267 only within the enclosing RCU read-side critical section. 268 For example, the following is -not- legal: 269 270 rcu_read_lock(); 271 p = rcu_dereference(head.next); 272 rcu_read_unlock(); 273 x = p->address; /* BUG!!! */ 274 rcu_read_lock(); 275 y = p->data; /* BUG!!! */ 276 rcu_read_unlock(); 277 278 Holding a reference from one RCU read-side critical section 279 to another is just as illegal as holding a reference from 280 one lock-based critical section to another! Similarly, 281 using a reference outside of the critical section in which 282 it was acquired is just as illegal as doing so with normal 283 locking. 284 285 As with rcu_assign_pointer(), an important function of 286 rcu_dereference() is to document which pointers are protected by 287 RCU, in particular, flagging a pointer that is subject to changing 288 at any time, including immediately after the rcu_dereference(). 289 And, again like rcu_assign_pointer(), rcu_dereference() is 290 typically used indirectly, via the _rcu list-manipulation 291 primitives, such as list_for_each_entry_rcu(). 292 293 The following diagram shows how each API communicates among the 294 reader, updater, and reclaimer. 295 296 297 rcu_assign_pointer() 298 +--------+ 299 +---------------------->| reader |---------+ 300 | +--------+ | 301 | | | 302 | | | Protect: 303 | | | rcu_read_lock() 304 | | | rcu_read_unlock() 305 | rcu_dereference() | | 306 +---------+ | | 307 | updater |<---------------------+ | 308 +---------+ V 309 | +-----------+ 310 +----------------------------------->| reclaimer | 311 +-----------+ 312 Defer: 313 synchronize_rcu() & call_rcu() 314 315 316 The RCU infrastructure observes the time sequence of rcu_read_lock(), 317 rcu_read_unlock(), synchronize_rcu(), and call_rcu() invocations in 318 order to determine when (1) synchronize_rcu() invocations may return 319 to their callers and (2) call_rcu() callbacks may be invoked. Efficient 320 implementations of the RCU infrastructure make heavy use of batching in 321 order to amortize their overhead over many uses of the corresponding APIs. 322 323 There are no fewer than three RCU mechanisms in the Linux kernel; the 324 diagram above shows the first one, which is by far the most commonly used. 325 The rcu_dereference() and rcu_assign_pointer() primitives are used for 326 all three mechanisms, but different defer and protect primitives are 327 used as follows: 328 329 Defer Protect 330 331 a. synchronize_rcu() rcu_read_lock() / rcu_read_unlock() 332 call_rcu() rcu_dereference() 333 334 b. synchronize_rcu_bh() rcu_read_lock_bh() / rcu_read_unlock_bh() 335 call_rcu_bh() rcu_dereference_bh() 336 337 c. synchronize_sched() rcu_read_lock_sched() / rcu_read_unlock_sched() 338 call_rcu_sched() preempt_disable() / preempt_enable() 339 local_irq_save() / local_irq_restore() 340 hardirq enter / hardirq exit 341 NMI enter / NMI exit 342 rcu_dereference_sched() 343 344 These three mechanisms are used as follows: 345 346 a. RCU applied to normal data structures. 347 348 b. RCU applied to networking data structures that may be subjected 349 to remote denial-of-service attacks. 350 351 c. RCU applied to scheduler and interrupt/NMI-handler tasks. 352 353 Again, most uses will be of (a). The (b) and (c) cases are important 354 for specialized uses, but are relatively uncommon. 355 356 357 3. WHAT ARE SOME EXAMPLE USES OF CORE RCU API? 358 359 This section shows a simple use of the core RCU API to protect a 360 global pointer to a dynamically allocated structure. More-typical 361 uses of RCU may be found in listRCU.txt, arrayRCU.txt, and NMI-RCU.txt. 362 363 struct foo { 364 int a; 365 char b; 366 long c; 367 }; 368 DEFINE_SPINLOCK(foo_mutex); 369 370 struct foo __rcu *gbl_foo; 371 372 /* 373 * Create a new struct foo that is the same as the one currently 374 * pointed to by gbl_foo, except that field "a" is replaced 375 * with "new_a". Points gbl_foo to the new structure, and 376 * frees up the old structure after a grace period. 377 * 378 * Uses rcu_assign_pointer() to ensure that concurrent readers 379 * see the initialized version of the new structure. 380 * 381 * Uses synchronize_rcu() to ensure that any readers that might 382 * have references to the old structure complete before freeing 383 * the old structure. 384 */ 385 void foo_update_a(int new_a) 386 { 387 struct foo *new_fp; 388 struct foo *old_fp; 389 390 new_fp = kmalloc(sizeof(*new_fp), GFP_KERNEL); 391 spin_lock(&foo_mutex); 392 old_fp = rcu_dereference_protected(gbl_foo, lockdep_is_held(&foo_mutex)); 393 *new_fp = *old_fp; 394 new_fp->a = new_a; 395 rcu_assign_pointer(gbl_foo, new_fp); 396 spin_unlock(&foo_mutex); 397 synchronize_rcu(); 398 kfree(old_fp); 399 } 400 401 /* 402 * Return the value of field "a" of the current gbl_foo 403 * structure. Use rcu_read_lock() and rcu_read_unlock() 404 * to ensure that the structure does not get deleted out 405 * from under us, and use rcu_dereference() to ensure that 406 * we see the initialized version of the structure (important 407 * for DEC Alpha and for people reading the code). 408 */ 409 int foo_get_a(void) 410 { 411 int retval; 412 413 rcu_read_lock(); 414 retval = rcu_dereference(gbl_foo)->a; 415 rcu_read_unlock(); 416 return retval; 417 } 418 419 So, to sum up: 420 421 o Use rcu_read_lock() and rcu_read_unlock() to guard RCU 422 read-side critical sections. 423 424 o Within an RCU read-side critical section, use rcu_dereference() 425 to dereference RCU-protected pointers. 426 427 o Use some solid scheme (such as locks or semaphores) to 428 keep concurrent updates from interfering with each other. 429 430 o Use rcu_assign_pointer() to update an RCU-protected pointer. 431 This primitive protects concurrent readers from the updater, 432 -not- concurrent updates from each other! You therefore still 433 need to use locking (or something similar) to keep concurrent 434 rcu_assign_pointer() primitives from interfering with each other. 435 436 o Use synchronize_rcu() -after- removing a data element from an 437 RCU-protected data structure, but -before- reclaiming/freeing 438 the data element, in order to wait for the completion of all 439 RCU read-side critical sections that might be referencing that 440 data item. 441 442 See checklist.txt for additional rules to follow when using RCU. 443 And again, more-typical uses of RCU may be found in listRCU.txt, 444 arrayRCU.txt, and NMI-RCU.txt. 445 446 447 4. WHAT IF MY UPDATING THREAD CANNOT BLOCK? 448 449 In the example above, foo_update_a() blocks until a grace period elapses. 450 This is quite simple, but in some cases one cannot afford to wait so 451 long -- there might be other high-priority work to be done. 452 453 In such cases, one uses call_rcu() rather than synchronize_rcu(). 454 The call_rcu() API is as follows: 455 456 void call_rcu(struct rcu_head * head, 457 void (*func)(struct rcu_head *head)); 458 459 This function invokes func(head) after a grace period has elapsed. 460 This invocation might happen from either softirq or process context, 461 so the function is not permitted to block. The foo struct needs to 462 have an rcu_head structure added, perhaps as follows: 463 464 struct foo { 465 int a; 466 char b; 467 long c; 468 struct rcu_head rcu; 469 }; 470 471 The foo_update_a() function might then be written as follows: 472 473 /* 474 * Create a new struct foo that is the same as the one currently 475 * pointed to by gbl_foo, except that field "a" is replaced 476 * with "new_a". Points gbl_foo to the new structure, and 477 * frees up the old structure after a grace period. 478 * 479 * Uses rcu_assign_pointer() to ensure that concurrent readers 480 * see the initialized version of the new structure. 481 * 482 * Uses call_rcu() to ensure that any readers that might have 483 * references to the old structure complete before freeing the 484 * old structure. 485 */ 486 void foo_update_a(int new_a) 487 { 488 struct foo *new_fp; 489 struct foo *old_fp; 490 491 new_fp = kmalloc(sizeof(*new_fp), GFP_KERNEL); 492 spin_lock(&foo_mutex); 493 old_fp = rcu_dereference_protected(gbl_foo, lockdep_is_held(&foo_mutex)); 494 *new_fp = *old_fp; 495 new_fp->a = new_a; 496 rcu_assign_pointer(gbl_foo, new_fp); 497 spin_unlock(&foo_mutex); 498 call_rcu(&old_fp->rcu, foo_reclaim); 499 } 500 501 The foo_reclaim() function might appear as follows: 502 503 void foo_reclaim(struct rcu_head *rp) 504 { 505 struct foo *fp = container_of(rp, struct foo, rcu); 506 507 foo_cleanup(fp->a); 508 509 kfree(fp); 510 } 511 512 The container_of() primitive is a macro that, given a pointer into a 513 struct, the type of the struct, and the pointed-to field within the 514 struct, returns a pointer to the beginning of the struct. 515 516 The use of call_rcu() permits the caller of foo_update_a() to 517 immediately regain control, without needing to worry further about the 518 old version of the newly updated element. It also clearly shows the 519 RCU distinction between updater, namely foo_update_a(), and reclaimer, 520 namely foo_reclaim(). 521 522 The summary of advice is the same as for the previous section, except 523 that we are now using call_rcu() rather than synchronize_rcu(): 524 525 o Use call_rcu() -after- removing a data element from an 526 RCU-protected data structure in order to register a callback 527 function that will be invoked after the completion of all RCU 528 read-side critical sections that might be referencing that 529 data item. 530 531 If the callback for call_rcu() is not doing anything more than calling 532 kfree() on the structure, you can use kfree_rcu() instead of call_rcu() 533 to avoid having to write your own callback: 534 535 kfree_rcu(old_fp, rcu); 536 537 Again, see checklist.txt for additional rules governing the use of RCU. 538 539 540 5. WHAT ARE SOME SIMPLE IMPLEMENTATIONS OF RCU? 541 542 One of the nice things about RCU is that it has extremely simple "toy" 543 implementations that are a good first step towards understanding the 544 production-quality implementations in the Linux kernel. This section 545 presents two such "toy" implementations of RCU, one that is implemented 546 in terms of familiar locking primitives, and another that more closely 547 resembles "classic" RCU. Both are way too simple for real-world use, 548 lacking both functionality and performance. However, they are useful 549 in getting a feel for how RCU works. See kernel/rcupdate.c for a 550 production-quality implementation, and see: 551 552 http://www.rdrop.com/users/paulmck/RCU 553 554 for papers describing the Linux kernel RCU implementation. The OLS'01 555 and OLS'02 papers are a good introduction, and the dissertation provides 556 more details on the current implementation as of early 2004. 557 558 559 5A. "TOY" IMPLEMENTATION #1: LOCKING 560 561 This section presents a "toy" RCU implementation that is based on 562 familiar locking primitives. Its overhead makes it a non-starter for 563 real-life use, as does its lack of scalability. It is also unsuitable 564 for realtime use, since it allows scheduling latency to "bleed" from 565 one read-side critical section to another. It also assumes recursive 566 reader-writer locks: If you try this with non-recursive locks, and 567 you allow nested rcu_read_lock() calls, you can deadlock. 568 569 However, it is probably the easiest implementation to relate to, so is 570 a good starting point. 571 572 It is extremely simple: 573 574 static DEFINE_RWLOCK(rcu_gp_mutex); 575 576 void rcu_read_lock(void) 577 { 578 read_lock(&rcu_gp_mutex); 579 } 580 581 void rcu_read_unlock(void) 582 { 583 read_unlock(&rcu_gp_mutex); 584 } 585 586 void synchronize_rcu(void) 587 { 588 write_lock(&rcu_gp_mutex); 589 write_unlock(&rcu_gp_mutex); 590 } 591 592 [You can ignore rcu_assign_pointer() and rcu_dereference() without missing 593 much. But here are simplified versions anyway. And whatever you do, 594 don't forget about them when submitting patches making use of RCU!] 595 596 #define rcu_assign_pointer(p, v) \ 597 ({ \ 598 smp_store_release(&(p), (v)); \ 599 }) 600 601 #define rcu_dereference(p) \ 602 ({ \ 603 typeof(p) _________p1 = READ_ONCE(p); \ 604 (_________p1); \ 605 }) 606 607 608 The rcu_read_lock() and rcu_read_unlock() primitive read-acquire 609 and release a global reader-writer lock. The synchronize_rcu() 610 primitive write-acquires this same lock, then immediately releases 611 it. This means that once synchronize_rcu() exits, all RCU read-side 612 critical sections that were in progress before synchronize_rcu() was 613 called are guaranteed to have completed -- there is no way that 614 synchronize_rcu() would have been able to write-acquire the lock 615 otherwise. 616 617 It is possible to nest rcu_read_lock(), since reader-writer locks may 618 be recursively acquired. Note also that rcu_read_lock() is immune 619 from deadlock (an important property of RCU). The reason for this is 620 that the only thing that can block rcu_read_lock() is a synchronize_rcu(). 621 But synchronize_rcu() does not acquire any locks while holding rcu_gp_mutex, 622 so there can be no deadlock cycle. 623 624 Quick Quiz #1: Why is this argument naive? How could a deadlock 625 occur when using this algorithm in a real-world Linux 626 kernel? How could this deadlock be avoided? 627 628 629 5B. "TOY" EXAMPLE #2: CLASSIC RCU 630 631 This section presents a "toy" RCU implementation that is based on 632 "classic RCU". It is also short on performance (but only for updates) and 633 on features such as hotplug CPU and the ability to run in CONFIG_PREEMPT 634 kernels. The definitions of rcu_dereference() and rcu_assign_pointer() 635 are the same as those shown in the preceding section, so they are omitted. 636 637 void rcu_read_lock(void) { } 638 639 void rcu_read_unlock(void) { } 640 641 void synchronize_rcu(void) 642 { 643 int cpu; 644 645 for_each_possible_cpu(cpu) 646 run_on(cpu); 647 } 648 649 Note that rcu_read_lock() and rcu_read_unlock() do absolutely nothing. 650 This is the great strength of classic RCU in a non-preemptive kernel: 651 read-side overhead is precisely zero, at least on non-Alpha CPUs. 652 And there is absolutely no way that rcu_read_lock() can possibly 653 participate in a deadlock cycle! 654 655 The implementation of synchronize_rcu() simply schedules itself on each 656 CPU in turn. The run_on() primitive can be implemented straightforwardly 657 in terms of the sched_setaffinity() primitive. Of course, a somewhat less 658 "toy" implementation would restore the affinity upon completion rather 659 than just leaving all tasks running on the last CPU, but when I said 660 "toy", I meant -toy-! 661 662 So how the heck is this supposed to work??? 663 664 Remember that it is illegal to block while in an RCU read-side critical 665 section. Therefore, if a given CPU executes a context switch, we know 666 that it must have completed all preceding RCU read-side critical sections. 667 Once -all- CPUs have executed a context switch, then -all- preceding 668 RCU read-side critical sections will have completed. 669 670 So, suppose that we remove a data item from its structure and then invoke 671 synchronize_rcu(). Once synchronize_rcu() returns, we are guaranteed 672 that there are no RCU read-side critical sections holding a reference 673 to that data item, so we can safely reclaim it. 674 675 Quick Quiz #2: Give an example where Classic RCU's read-side 676 overhead is -negative-. 677 678 Quick Quiz #3: If it is illegal to block in an RCU read-side 679 critical section, what the heck do you do in 680 PREEMPT_RT, where normal spinlocks can block??? 681 682 683 6. ANALOGY WITH READER-WRITER LOCKING 684 685 Although RCU can be used in many different ways, a very common use of 686 RCU is analogous to reader-writer locking. The following unified 687 diff shows how closely related RCU and reader-writer locking can be. 688 689 @@ -5,5 +5,5 @@ struct el { 690 int data; 691 /* Other data fields */ 692 }; 693 -rwlock_t listmutex; 694 +spinlock_t listmutex; 695 struct el head; 696 697 @@ -13,15 +14,15 @@ 698 struct list_head *lp; 699 struct el *p; 700 701 - read_lock(&listmutex); 702 - list_for_each_entry(p, head, lp) { 703 + rcu_read_lock(); 704 + list_for_each_entry_rcu(p, head, lp) { 705 if (p->key == key) { 706 *result = p->data; 707 - read_unlock(&listmutex); 708 + rcu_read_unlock(); 709 return 1; 710 } 711 } 712 - read_unlock(&listmutex); 713 + rcu_read_unlock(); 714 return 0; 715 } 716 717 @@ -29,15 +30,16 @@ 718 { 719 struct el *p; 720 721 - write_lock(&listmutex); 722 + spin_lock(&listmutex); 723 list_for_each_entry(p, head, lp) { 724 if (p->key == key) { 725 - list_del(&p->list); 726 - write_unlock(&listmutex); 727 + list_del_rcu(&p->list); 728 + spin_unlock(&listmutex); 729 + synchronize_rcu(); 730 kfree(p); 731 return 1; 732 } 733 } 734 - write_unlock(&listmutex); 735 + spin_unlock(&listmutex); 736 return 0; 737 } 738 739 Or, for those who prefer a side-by-side listing: 740 741 1 struct el { 1 struct el { 742 2 struct list_head list; 2 struct list_head list; 743 3 long key; 3 long key; 744 4 spinlock_t mutex; 4 spinlock_t mutex; 745 5 int data; 5 int data; 746 6 /* Other data fields */ 6 /* Other data fields */ 747 7 }; 7 }; 748 8 rwlock_t listmutex; 8 spinlock_t listmutex; 749 9 struct el head; 9 struct el head; 750 751 1 int search(long key, int *result) 1 int search(long key, int *result) 752 2 { 2 { 753 3 struct list_head *lp; 3 struct list_head *lp; 754 4 struct el *p; 4 struct el *p; 755 5 5 756 6 read_lock(&listmutex); 6 rcu_read_lock(); 757 7 list_for_each_entry(p, head, lp) { 7 list_for_each_entry_rcu(p, head, lp) { 758 8 if (p->key == key) { 8 if (p->key == key) { 759 9 *result = p->data; 9 *result = p->data; 760 10 read_unlock(&listmutex); 10 rcu_read_unlock(); 761 11 return 1; 11 return 1; 762 12 } 12 } 763 13 } 13 } 764 14 read_unlock(&listmutex); 14 rcu_read_unlock(); 765 15 return 0; 15 return 0; 766 16 } 16 } 767 768 1 int delete(long key) 1 int delete(long key) 769 2 { 2 { 770 3 struct el *p; 3 struct el *p; 771 4 4 772 5 write_lock(&listmutex); 5 spin_lock(&listmutex); 773 6 list_for_each_entry(p, head, lp) { 6 list_for_each_entry(p, head, lp) { 774 7 if (p->key == key) { 7 if (p->key == key) { 775 8 list_del(&p->list); 8 list_del_rcu(&p->list); 776 9 write_unlock(&listmutex); 9 spin_unlock(&listmutex); 777 10 synchronize_rcu(); 778 10 kfree(p); 11 kfree(p); 779 11 return 1; 12 return 1; 780 12 } 13 } 781 13 } 14 } 782 14 write_unlock(&listmutex); 15 spin_unlock(&listmutex); 783 15 return 0; 16 return 0; 784 16 } 17 } 785 786 Either way, the differences are quite small. Read-side locking moves 787 to rcu_read_lock() and rcu_read_unlock, update-side locking moves from 788 a reader-writer lock to a simple spinlock, and a synchronize_rcu() 789 precedes the kfree(). 790 791 However, there is one potential catch: the read-side and update-side 792 critical sections can now run concurrently. In many cases, this will 793 not be a problem, but it is necessary to check carefully regardless. 794 For example, if multiple independent list updates must be seen as 795 a single atomic update, converting to RCU will require special care. 796 797 Also, the presence of synchronize_rcu() means that the RCU version of 798 delete() can now block. If this is a problem, there is a callback-based 799 mechanism that never blocks, namely call_rcu() or kfree_rcu(), that can 800 be used in place of synchronize_rcu(). 801 802 803 7. FULL LIST OF RCU APIs 804 805 The RCU APIs are documented in docbook-format header comments in the 806 Linux-kernel source code, but it helps to have a full list of the 807 APIs, since there does not appear to be a way to categorize them 808 in docbook. Here is the list, by category. 809 810 RCU list traversal: 811 812 list_entry_rcu 813 list_first_entry_rcu 814 list_next_rcu 815 list_for_each_entry_rcu 816 list_for_each_entry_continue_rcu 817 hlist_first_rcu 818 hlist_next_rcu 819 hlist_pprev_rcu 820 hlist_for_each_entry_rcu 821 hlist_for_each_entry_rcu_bh 822 hlist_for_each_entry_continue_rcu 823 hlist_for_each_entry_continue_rcu_bh 824 hlist_nulls_first_rcu 825 hlist_nulls_for_each_entry_rcu 826 hlist_bl_first_rcu 827 hlist_bl_for_each_entry_rcu 828 829 RCU pointer/list update: 830 831 rcu_assign_pointer 832 list_add_rcu 833 list_add_tail_rcu 834 list_del_rcu 835 list_replace_rcu 836 hlist_add_behind_rcu 837 hlist_add_before_rcu 838 hlist_add_head_rcu 839 hlist_del_rcu 840 hlist_del_init_rcu 841 hlist_replace_rcu 842 list_splice_init_rcu() 843 hlist_nulls_del_init_rcu 844 hlist_nulls_del_rcu 845 hlist_nulls_add_head_rcu 846 hlist_bl_add_head_rcu 847 hlist_bl_del_init_rcu 848 hlist_bl_del_rcu 849 hlist_bl_set_first_rcu 850 851 RCU: Critical sections Grace period Barrier 852 853 rcu_read_lock synchronize_net rcu_barrier 854 rcu_read_unlock synchronize_rcu 855 rcu_dereference synchronize_rcu_expedited 856 rcu_read_lock_held call_rcu 857 rcu_dereference_check kfree_rcu 858 rcu_dereference_protected 859 860 bh: Critical sections Grace period Barrier 861 862 rcu_read_lock_bh call_rcu_bh rcu_barrier_bh 863 rcu_read_unlock_bh synchronize_rcu_bh 864 rcu_dereference_bh synchronize_rcu_bh_expedited 865 rcu_dereference_bh_check 866 rcu_dereference_bh_protected 867 rcu_read_lock_bh_held 868 869 sched: Critical sections Grace period Barrier 870 871 rcu_read_lock_sched synchronize_sched rcu_barrier_sched 872 rcu_read_unlock_sched call_rcu_sched 873 [preempt_disable] synchronize_sched_expedited 874 [and friends] 875 rcu_read_lock_sched_notrace 876 rcu_read_unlock_sched_notrace 877 rcu_dereference_sched 878 rcu_dereference_sched_check 879 rcu_dereference_sched_protected 880 rcu_read_lock_sched_held 881 882 883 SRCU: Critical sections Grace period Barrier 884 885 srcu_read_lock synchronize_srcu srcu_barrier 886 srcu_read_unlock call_srcu 887 srcu_dereference synchronize_srcu_expedited 888 srcu_dereference_check 889 srcu_read_lock_held 890 891 SRCU: Initialization/cleanup 892 DEFINE_SRCU 893 DEFINE_STATIC_SRCU 894 init_srcu_struct 895 cleanup_srcu_struct 896 897 All: lockdep-checked RCU-protected pointer access 898 899 rcu_access_pointer 900 rcu_dereference_raw 901 RCU_LOCKDEP_WARN 902 rcu_sleep_check 903 RCU_NONIDLE 904 905 See the comment headers in the source code (or the docbook generated 906 from them) for more information. 907 908 However, given that there are no fewer than four families of RCU APIs 909 in the Linux kernel, how do you choose which one to use? The following 910 list can be helpful: 911 912 a. Will readers need to block? If so, you need SRCU. 913 914 b. What about the -rt patchset? If readers would need to block 915 in an non-rt kernel, you need SRCU. If readers would block 916 in a -rt kernel, but not in a non-rt kernel, SRCU is not 917 necessary. (The -rt patchset turns spinlocks into sleeplocks, 918 hence this distinction.) 919 920 c. Do you need to treat NMI handlers, hardirq handlers, 921 and code segments with preemption disabled (whether 922 via preempt_disable(), local_irq_save(), local_bh_disable(), 923 or some other mechanism) as if they were explicit RCU readers? 924 If so, RCU-sched is the only choice that will work for you. 925 926 d. Do you need RCU grace periods to complete even in the face 927 of softirq monopolization of one or more of the CPUs? For 928 example, is your code subject to network-based denial-of-service 929 attacks? If so, you need RCU-bh. 930 931 e. Is your workload too update-intensive for normal use of 932 RCU, but inappropriate for other synchronization mechanisms? 933 If so, consider SLAB_TYPESAFE_BY_RCU (which was originally 934 named SLAB_DESTROY_BY_RCU). But please be careful! 935 936 f. Do you need read-side critical sections that are respected 937 even though they are in the middle of the idle loop, during 938 user-mode execution, or on an offlined CPU? If so, SRCU is the 939 only choice that will work for you. 940 941 g. Otherwise, use RCU. 942 943 Of course, this all assumes that you have determined that RCU is in fact 944 the right tool for your job. 945 946 947 8. ANSWERS TO QUICK QUIZZES 948 949 Quick Quiz #1: Why is this argument naive? How could a deadlock 950 occur when using this algorithm in a real-world Linux 951 kernel? [Referring to the lock-based "toy" RCU 952 algorithm.] 953 954 Answer: Consider the following sequence of events: 955 956 1. CPU 0 acquires some unrelated lock, call it 957 "problematic_lock", disabling irq via 958 spin_lock_irqsave(). 959 960 2. CPU 1 enters synchronize_rcu(), write-acquiring 961 rcu_gp_mutex. 962 963 3. CPU 0 enters rcu_read_lock(), but must wait 964 because CPU 1 holds rcu_gp_mutex. 965 966 4. CPU 1 is interrupted, and the irq handler 967 attempts to acquire problematic_lock. 968 969 The system is now deadlocked. 970 971 One way to avoid this deadlock is to use an approach like 972 that of CONFIG_PREEMPT_RT, where all normal spinlocks 973 become blocking locks, and all irq handlers execute in 974 the context of special tasks. In this case, in step 4 975 above, the irq handler would block, allowing CPU 1 to 976 release rcu_gp_mutex, avoiding the deadlock. 977 978 Even in the absence of deadlock, this RCU implementation 979 allows latency to "bleed" from readers to other 980 readers through synchronize_rcu(). To see this, 981 consider task A in an RCU read-side critical section 982 (thus read-holding rcu_gp_mutex), task B blocked 983 attempting to write-acquire rcu_gp_mutex, and 984 task C blocked in rcu_read_lock() attempting to 985 read_acquire rcu_gp_mutex. Task A's RCU read-side 986 latency is holding up task C, albeit indirectly via 987 task B. 988 989 Realtime RCU implementations therefore use a counter-based 990 approach where tasks in RCU read-side critical sections 991 cannot be blocked by tasks executing synchronize_rcu(). 992 993 Quick Quiz #2: Give an example where Classic RCU's read-side 994 overhead is -negative-. 995 996 Answer: Imagine a single-CPU system with a non-CONFIG_PREEMPT 997 kernel where a routing table is used by process-context 998 code, but can be updated by irq-context code (for example, 999 by an "ICMP REDIRECT" packet). The usual way of handling 1000 this would be to have the process-context code disable 1001 interrupts while searching the routing table. Use of 1002 RCU allows such interrupt-disabling to be dispensed with. 1003 Thus, without RCU, you pay the cost of disabling interrupts, 1004 and with RCU you don't. 1005 1006 One can argue that the overhead of RCU in this 1007 case is negative with respect to the single-CPU 1008 interrupt-disabling approach. Others might argue that 1009 the overhead of RCU is merely zero, and that replacing 1010 the positive overhead of the interrupt-disabling scheme 1011 with the zero-overhead RCU scheme does not constitute 1012 negative overhead. 1013 1014 In real life, of course, things are more complex. But 1015 even the theoretical possibility of negative overhead for 1016 a synchronization primitive is a bit unexpected. ;-) 1017 1018 Quick Quiz #3: If it is illegal to block in an RCU read-side 1019 critical section, what the heck do you do in 1020 PREEMPT_RT, where normal spinlocks can block??? 1021 1022 Answer: Just as PREEMPT_RT permits preemption of spinlock 1023 critical sections, it permits preemption of RCU 1024 read-side critical sections. It also permits 1025 spinlocks blocking while in RCU read-side critical 1026 sections. 1027 1028 Why the apparent inconsistency? Because it is it 1029 possible to use priority boosting to keep the RCU 1030 grace periods short if need be (for example, if running 1031 short of memory). In contrast, if blocking waiting 1032 for (say) network reception, there is no way to know 1033 what should be boosted. Especially given that the 1034 process we need to boost might well be a human being 1035 who just went out for a pizza or something. And although 1036 a computer-operated cattle prod might arouse serious 1037 interest, it might also provoke serious objections. 1038 Besides, how does the computer know what pizza parlor 1039 the human being went to??? 1040 1041 1042 ACKNOWLEDGEMENTS 1043 1044 My thanks to the people who helped make this human-readable, including 1045 Jon Walpole, Josh Triplett, Serge Hallyn, Suzanne Wood, and Alan Stern. 1046 1047 1048 For more information, see http://www.rdrop.com/users/paulmck/RCU.